diff --git "a/trainer_state.json" "b/trainer_state.json"
new file mode 100644--- /dev/null
+++ "b/trainer_state.json"
@@ -0,0 +1,35034 @@
+{
+  "best_global_step": null,
+  "best_metric": null,
+  "best_model_checkpoint": null,
+  "epoch": 357.14285714285717,
+  "eval_steps": 500,
+  "global_step": 5000,
+  "is_hyper_param_search": false,
+  "is_local_process_zero": true,
+  "is_world_process_zero": true,
+  "log_history": [
+    {
+      "epoch": 0.07142857142857142,
+      "grad_norm": 1.234375,
+      "learning_rate": 0.0,
+      "loss": 0.0242,
+      "step": 1
+    },
+    {
+      "epoch": 0.14285714285714285,
+      "grad_norm": 1.234375,
+      "learning_rate": 5.0000000000000004e-08,
+      "loss": 0.0276,
+      "step": 2
+    },
+    {
+      "epoch": 0.21428571428571427,
+      "grad_norm": 1.078125,
+      "learning_rate": 1.0000000000000001e-07,
+      "loss": 0.0269,
+      "step": 3
+    },
+    {
+      "epoch": 0.2857142857142857,
+      "grad_norm": 0.734375,
+      "learning_rate": 1.5000000000000002e-07,
+      "loss": 0.0254,
+      "step": 4
+    },
+    {
+      "epoch": 0.35714285714285715,
+      "grad_norm": 1.078125,
+      "learning_rate": 2.0000000000000002e-07,
+      "loss": 0.0265,
+      "step": 5
+    },
+    {
+      "epoch": 0.42857142857142855,
+      "grad_norm": 0.609375,
+      "learning_rate": 2.5000000000000004e-07,
+      "loss": 0.0266,
+      "step": 6
+    },
+    {
+      "epoch": 0.5,
+      "grad_norm": 1.03125,
+      "learning_rate": 3.0000000000000004e-07,
+      "loss": 0.0265,
+      "step": 7
+    },
+    {
+      "epoch": 0.5714285714285714,
+      "grad_norm": 0.5859375,
+      "learning_rate": 3.5000000000000004e-07,
+      "loss": 0.027,
+      "step": 8
+    },
+    {
+      "epoch": 0.6428571428571429,
+      "grad_norm": 0.40625,
+      "learning_rate": 4.0000000000000003e-07,
+      "loss": 0.0271,
+      "step": 9
+    },
+    {
+      "epoch": 0.7142857142857143,
+      "grad_norm": 0.419921875,
+      "learning_rate": 4.5e-07,
+      "loss": 0.0246,
+      "step": 10
+    },
+    {
+      "epoch": 0.7857142857142857,
+      "grad_norm": 0.453125,
+      "learning_rate": 5.000000000000001e-07,
+      "loss": 0.0246,
+      "step": 11
+    },
+    {
+      "epoch": 0.8571428571428571,
+      "grad_norm": 0.52734375,
+      "learning_rate": 5.5e-07,
+      "loss": 0.0262,
+      "step": 12
+    },
+    {
+      "epoch": 0.9285714285714286,
+      "grad_norm": 0.91015625,
+      "learning_rate": 6.000000000000001e-07,
+      "loss": 0.0241,
+      "step": 13
+    },
+    {
+      "epoch": 1.0,
+      "grad_norm": 0.9375,
+      "learning_rate": 6.5e-07,
+      "loss": 0.0243,
+      "step": 14
+    },
+    {
+      "epoch": 1.0714285714285714,
+      "grad_norm": 0.9453125,
+      "learning_rate": 7.000000000000001e-07,
+      "loss": 0.0272,
+      "step": 15
+    },
+    {
+      "epoch": 1.1428571428571428,
+      "grad_norm": 1.3828125,
+      "learning_rate": 7.5e-07,
+      "loss": 0.0269,
+      "step": 16
+    },
+    {
+      "epoch": 1.2142857142857142,
+      "grad_norm": 0.99609375,
+      "learning_rate": 8.000000000000001e-07,
+      "loss": 0.0252,
+      "step": 17
+    },
+    {
+      "epoch": 1.2857142857142856,
+      "grad_norm": 0.34375,
+      "learning_rate": 8.500000000000001e-07,
+      "loss": 0.0241,
+      "step": 18
+    },
+    {
+      "epoch": 1.3571428571428572,
+      "grad_norm": 0.8046875,
+      "learning_rate": 9e-07,
+      "loss": 0.0252,
+      "step": 19
+    },
+    {
+      "epoch": 1.4285714285714286,
+      "grad_norm": 1.1640625,
+      "learning_rate": 9.5e-07,
+      "loss": 0.0253,
+      "step": 20
+    },
+    {
+      "epoch": 1.5,
+      "grad_norm": 0.333984375,
+      "learning_rate": 1.0000000000000002e-06,
+      "loss": 0.0264,
+      "step": 21
+    },
+    {
+      "epoch": 1.5714285714285714,
+      "grad_norm": 0.236328125,
+      "learning_rate": 1.0500000000000001e-06,
+      "loss": 0.0248,
+      "step": 22
+    },
+    {
+      "epoch": 1.6428571428571428,
+      "grad_norm": 0.275390625,
+      "learning_rate": 1.1e-06,
+      "loss": 0.0227,
+      "step": 23
+    },
+    {
+      "epoch": 1.7142857142857144,
+      "grad_norm": 0.50390625,
+      "learning_rate": 1.15e-06,
+      "loss": 0.0247,
+      "step": 24
+    },
+    {
+      "epoch": 1.7857142857142856,
+      "grad_norm": 0.423828125,
+      "learning_rate": 1.2000000000000002e-06,
+      "loss": 0.0256,
+      "step": 25
+    },
+    {
+      "epoch": 1.8571428571428572,
+      "grad_norm": 0.25,
+      "learning_rate": 1.25e-06,
+      "loss": 0.0251,
+      "step": 26
+    },
+    {
+      "epoch": 1.9285714285714286,
+      "grad_norm": 0.404296875,
+      "learning_rate": 1.3e-06,
+      "loss": 0.0233,
+      "step": 27
+    },
+    {
+      "epoch": 2.0,
+      "grad_norm": 1.1171875,
+      "learning_rate": 1.35e-06,
+      "loss": 0.0238,
+      "step": 28
+    },
+    {
+      "epoch": 2.0714285714285716,
+      "grad_norm": 0.287109375,
+      "learning_rate": 1.4000000000000001e-06,
+      "loss": 0.0238,
+      "step": 29
+    },
+    {
+      "epoch": 2.142857142857143,
+      "grad_norm": 0.62109375,
+      "learning_rate": 1.45e-06,
+      "loss": 0.0259,
+      "step": 30
+    },
+    {
+      "epoch": 2.2142857142857144,
+      "grad_norm": 0.5078125,
+      "learning_rate": 1.5e-06,
+      "loss": 0.0256,
+      "step": 31
+    },
+    {
+      "epoch": 2.2857142857142856,
+      "grad_norm": 0.369140625,
+      "learning_rate": 1.55e-06,
+      "loss": 0.0242,
+      "step": 32
+    },
+    {
+      "epoch": 2.357142857142857,
+      "grad_norm": 0.65625,
+      "learning_rate": 1.6000000000000001e-06,
+      "loss": 0.0263,
+      "step": 33
+    },
+    {
+      "epoch": 2.4285714285714284,
+      "grad_norm": 0.38671875,
+      "learning_rate": 1.65e-06,
+      "loss": 0.0246,
+      "step": 34
+    },
+    {
+      "epoch": 2.5,
+      "grad_norm": 0.265625,
+      "learning_rate": 1.7000000000000002e-06,
+      "loss": 0.0259,
+      "step": 35
+    },
+    {
+      "epoch": 2.571428571428571,
+      "grad_norm": 0.2734375,
+      "learning_rate": 1.7500000000000002e-06,
+      "loss": 0.0227,
+      "step": 36
+    },
+    {
+      "epoch": 2.642857142857143,
+      "grad_norm": 0.357421875,
+      "learning_rate": 1.8e-06,
+      "loss": 0.0256,
+      "step": 37
+    },
+    {
+      "epoch": 2.7142857142857144,
+      "grad_norm": 0.671875,
+      "learning_rate": 1.85e-06,
+      "loss": 0.0263,
+      "step": 38
+    },
+    {
+      "epoch": 2.7857142857142856,
+      "grad_norm": 0.63671875,
+      "learning_rate": 1.9e-06,
+      "loss": 0.024,
+      "step": 39
+    },
+    {
+      "epoch": 2.857142857142857,
+      "grad_norm": 0.236328125,
+      "learning_rate": 1.95e-06,
+      "loss": 0.0256,
+      "step": 40
+    },
+    {
+      "epoch": 2.928571428571429,
+      "grad_norm": 0.359375,
+      "learning_rate": 2.0000000000000003e-06,
+      "loss": 0.0244,
+      "step": 41
+    },
+    {
+      "epoch": 3.0,
+      "grad_norm": 0.3046875,
+      "learning_rate": 2.0500000000000003e-06,
+      "loss": 0.0246,
+      "step": 42
+    },
+    {
+      "epoch": 3.0714285714285716,
+      "grad_norm": 0.2216796875,
+      "learning_rate": 2.1000000000000002e-06,
+      "loss": 0.0234,
+      "step": 43
+    },
+    {
+      "epoch": 3.142857142857143,
+      "grad_norm": 0.267578125,
+      "learning_rate": 2.1499999999999997e-06,
+      "loss": 0.0231,
+      "step": 44
+    },
+    {
+      "epoch": 3.2142857142857144,
+      "grad_norm": 0.279296875,
+      "learning_rate": 2.2e-06,
+      "loss": 0.0244,
+      "step": 45
+    },
+    {
+      "epoch": 3.2857142857142856,
+      "grad_norm": 0.25390625,
+      "learning_rate": 2.25e-06,
+      "loss": 0.0237,
+      "step": 46
+    },
+    {
+      "epoch": 3.357142857142857,
+      "grad_norm": 0.349609375,
+      "learning_rate": 2.3e-06,
+      "loss": 0.0236,
+      "step": 47
+    },
+    {
+      "epoch": 3.4285714285714284,
+      "grad_norm": 0.34375,
+      "learning_rate": 2.35e-06,
+      "loss": 0.0247,
+      "step": 48
+    },
+    {
+      "epoch": 3.5,
+      "grad_norm": 0.330078125,
+      "learning_rate": 2.4000000000000003e-06,
+      "loss": 0.0244,
+      "step": 49
+    },
+    {
+      "epoch": 3.571428571428571,
+      "grad_norm": 0.6640625,
+      "learning_rate": 2.4500000000000003e-06,
+      "loss": 0.024,
+      "step": 50
+    },
+    {
+      "epoch": 3.642857142857143,
+      "grad_norm": 0.4453125,
+      "learning_rate": 2.5e-06,
+      "loss": 0.022,
+      "step": 51
+    },
+    {
+      "epoch": 3.7142857142857144,
+      "grad_norm": 0.2578125,
+      "learning_rate": 2.55e-06,
+      "loss": 0.0232,
+      "step": 52
+    },
+    {
+      "epoch": 3.7857142857142856,
+      "grad_norm": 0.310546875,
+      "learning_rate": 2.6e-06,
+      "loss": 0.0255,
+      "step": 53
+    },
+    {
+      "epoch": 3.857142857142857,
+      "grad_norm": 0.404296875,
+      "learning_rate": 2.65e-06,
+      "loss": 0.0227,
+      "step": 54
+    },
+    {
+      "epoch": 3.928571428571429,
+      "grad_norm": 0.314453125,
+      "learning_rate": 2.7e-06,
+      "loss": 0.0232,
+      "step": 55
+    },
+    {
+      "epoch": 4.0,
+      "grad_norm": 0.2255859375,
+      "learning_rate": 2.7500000000000004e-06,
+      "loss": 0.0254,
+      "step": 56
+    },
+    {
+      "epoch": 4.071428571428571,
+      "grad_norm": 0.5390625,
+      "learning_rate": 2.8000000000000003e-06,
+      "loss": 0.0252,
+      "step": 57
+    },
+    {
+      "epoch": 4.142857142857143,
+      "grad_norm": 0.349609375,
+      "learning_rate": 2.8500000000000002e-06,
+      "loss": 0.0257,
+      "step": 58
+    },
+    {
+      "epoch": 4.214285714285714,
+      "grad_norm": 0.22265625,
+      "learning_rate": 2.9e-06,
+      "loss": 0.0235,
+      "step": 59
+    },
+    {
+      "epoch": 4.285714285714286,
+      "grad_norm": 0.330078125,
+      "learning_rate": 2.95e-06,
+      "loss": 0.0239,
+      "step": 60
+    },
+    {
+      "epoch": 4.357142857142857,
+      "grad_norm": 0.326171875,
+      "learning_rate": 3e-06,
+      "loss": 0.0238,
+      "step": 61
+    },
+    {
+      "epoch": 4.428571428571429,
+      "grad_norm": 0.6953125,
+      "learning_rate": 3.05e-06,
+      "loss": 0.0234,
+      "step": 62
+    },
+    {
+      "epoch": 4.5,
+      "grad_norm": 0.431640625,
+      "learning_rate": 3.1e-06,
+      "loss": 0.0224,
+      "step": 63
+    },
+    {
+      "epoch": 4.571428571428571,
+      "grad_norm": 0.357421875,
+      "learning_rate": 3.1500000000000003e-06,
+      "loss": 0.0221,
+      "step": 64
+    },
+    {
+      "epoch": 4.642857142857143,
+      "grad_norm": 0.271484375,
+      "learning_rate": 3.2000000000000003e-06,
+      "loss": 0.0257,
+      "step": 65
+    },
+    {
+      "epoch": 4.714285714285714,
+      "grad_norm": 0.77734375,
+      "learning_rate": 3.2500000000000002e-06,
+      "loss": 0.0228,
+      "step": 66
+    },
+    {
+      "epoch": 4.785714285714286,
+      "grad_norm": 0.443359375,
+      "learning_rate": 3.3e-06,
+      "loss": 0.026,
+      "step": 67
+    },
+    {
+      "epoch": 4.857142857142857,
+      "grad_norm": 0.26953125,
+      "learning_rate": 3.3500000000000005e-06,
+      "loss": 0.0223,
+      "step": 68
+    },
+    {
+      "epoch": 4.928571428571429,
+      "grad_norm": 0.28125,
+      "learning_rate": 3.4000000000000005e-06,
+      "loss": 0.0231,
+      "step": 69
+    },
+    {
+      "epoch": 5.0,
+      "grad_norm": 0.22265625,
+      "learning_rate": 3.4500000000000004e-06,
+      "loss": 0.0239,
+      "step": 70
+    },
+    {
+      "epoch": 5.071428571428571,
+      "grad_norm": 0.546875,
+      "learning_rate": 3.5000000000000004e-06,
+      "loss": 0.0252,
+      "step": 71
+    },
+    {
+      "epoch": 5.142857142857143,
+      "grad_norm": 0.2734375,
+      "learning_rate": 3.55e-06,
+      "loss": 0.0253,
+      "step": 72
+    },
+    {
+      "epoch": 5.214285714285714,
+      "grad_norm": 0.341796875,
+      "learning_rate": 3.6e-06,
+      "loss": 0.0243,
+      "step": 73
+    },
+    {
+      "epoch": 5.285714285714286,
+      "grad_norm": 0.345703125,
+      "learning_rate": 3.6499999999999998e-06,
+      "loss": 0.023,
+      "step": 74
+    },
+    {
+      "epoch": 5.357142857142857,
+      "grad_norm": 0.361328125,
+      "learning_rate": 3.7e-06,
+      "loss": 0.0229,
+      "step": 75
+    },
+    {
+      "epoch": 5.428571428571429,
+      "grad_norm": 0.1904296875,
+      "learning_rate": 3.75e-06,
+      "loss": 0.0219,
+      "step": 76
+    },
+    {
+      "epoch": 5.5,
+      "grad_norm": 0.439453125,
+      "learning_rate": 3.8e-06,
+      "loss": 0.0238,
+      "step": 77
+    },
+    {
+      "epoch": 5.571428571428571,
+      "grad_norm": 0.3125,
+      "learning_rate": 3.85e-06,
+      "loss": 0.0245,
+      "step": 78
+    },
+    {
+      "epoch": 5.642857142857143,
+      "grad_norm": 0.71484375,
+      "learning_rate": 3.9e-06,
+      "loss": 0.0235,
+      "step": 79
+    },
+    {
+      "epoch": 5.714285714285714,
+      "grad_norm": 0.8671875,
+      "learning_rate": 3.95e-06,
+      "loss": 0.0224,
+      "step": 80
+    },
+    {
+      "epoch": 5.785714285714286,
+      "grad_norm": 0.41015625,
+      "learning_rate": 4.000000000000001e-06,
+      "loss": 0.0244,
+      "step": 81
+    },
+    {
+      "epoch": 5.857142857142857,
+      "grad_norm": 0.56640625,
+      "learning_rate": 4.05e-06,
+      "loss": 0.0227,
+      "step": 82
+    },
+    {
+      "epoch": 5.928571428571429,
+      "grad_norm": 0.890625,
+      "learning_rate": 4.1000000000000006e-06,
+      "loss": 0.0228,
+      "step": 83
+    },
+    {
+      "epoch": 6.0,
+      "grad_norm": 0.58203125,
+      "learning_rate": 4.15e-06,
+      "loss": 0.0239,
+      "step": 84
+    },
+    {
+      "epoch": 6.071428571428571,
+      "grad_norm": 0.365234375,
+      "learning_rate": 4.2000000000000004e-06,
+      "loss": 0.0236,
+      "step": 85
+    },
+    {
+      "epoch": 6.142857142857143,
+      "grad_norm": 0.71484375,
+      "learning_rate": 4.250000000000001e-06,
+      "loss": 0.0246,
+      "step": 86
+    },
+    {
+      "epoch": 6.214285714285714,
+      "grad_norm": 0.5234375,
+      "learning_rate": 4.2999999999999995e-06,
+      "loss": 0.0227,
+      "step": 87
+    },
+    {
+      "epoch": 6.285714285714286,
+      "grad_norm": 0.40234375,
+      "learning_rate": 4.35e-06,
+      "loss": 0.0228,
+      "step": 88
+    },
+    {
+      "epoch": 6.357142857142857,
+      "grad_norm": 0.265625,
+      "learning_rate": 4.4e-06,
+      "loss": 0.0233,
+      "step": 89
+    },
+    {
+      "epoch": 6.428571428571429,
+      "grad_norm": 0.435546875,
+      "learning_rate": 4.45e-06,
+      "loss": 0.0227,
+      "step": 90
+    },
+    {
+      "epoch": 6.5,
+      "grad_norm": 0.84375,
+      "learning_rate": 4.5e-06,
+      "loss": 0.0237,
+      "step": 91
+    },
+    {
+      "epoch": 6.571428571428571,
+      "grad_norm": 0.353515625,
+      "learning_rate": 4.5500000000000005e-06,
+      "loss": 0.0232,
+      "step": 92
+    },
+    {
+      "epoch": 6.642857142857143,
+      "grad_norm": 0.6015625,
+      "learning_rate": 4.6e-06,
+      "loss": 0.0229,
+      "step": 93
+    },
+    {
+      "epoch": 6.714285714285714,
+      "grad_norm": 0.28125,
+      "learning_rate": 4.65e-06,
+      "loss": 0.0221,
+      "step": 94
+    },
+    {
+      "epoch": 6.785714285714286,
+      "grad_norm": 0.44921875,
+      "learning_rate": 4.7e-06,
+      "loss": 0.0223,
+      "step": 95
+    },
+    {
+      "epoch": 6.857142857142857,
+      "grad_norm": 0.373046875,
+      "learning_rate": 4.75e-06,
+      "loss": 0.0225,
+      "step": 96
+    },
+    {
+      "epoch": 6.928571428571429,
+      "grad_norm": 0.23046875,
+      "learning_rate": 4.800000000000001e-06,
+      "loss": 0.0251,
+      "step": 97
+    },
+    {
+      "epoch": 7.0,
+      "grad_norm": 0.310546875,
+      "learning_rate": 4.85e-06,
+      "loss": 0.0208,
+      "step": 98
+    },
+    {
+      "epoch": 7.071428571428571,
+      "grad_norm": 0.427734375,
+      "learning_rate": 4.9000000000000005e-06,
+      "loss": 0.0233,
+      "step": 99
+    },
+    {
+      "epoch": 7.142857142857143,
+      "grad_norm": 0.28515625,
+      "learning_rate": 4.950000000000001e-06,
+      "loss": 0.0228,
+      "step": 100
+    },
+    {
+      "epoch": 7.214285714285714,
+      "grad_norm": 0.42578125,
+      "learning_rate": 5e-06,
+      "loss": 0.0216,
+      "step": 101
+    },
+    {
+      "epoch": 7.285714285714286,
+      "grad_norm": 0.322265625,
+      "learning_rate": 5.050000000000001e-06,
+      "loss": 0.0225,
+      "step": 102
+    },
+    {
+      "epoch": 7.357142857142857,
+      "grad_norm": 0.251953125,
+      "learning_rate": 5.1e-06,
+      "loss": 0.0214,
+      "step": 103
+    },
+    {
+      "epoch": 7.428571428571429,
+      "grad_norm": 0.62109375,
+      "learning_rate": 5.15e-06,
+      "loss": 0.0226,
+      "step": 104
+    },
+    {
+      "epoch": 7.5,
+      "grad_norm": 0.86328125,
+      "learning_rate": 5.2e-06,
+      "loss": 0.0232,
+      "step": 105
+    },
+    {
+      "epoch": 7.571428571428571,
+      "grad_norm": 0.451171875,
+      "learning_rate": 5.25e-06,
+      "loss": 0.0245,
+      "step": 106
+    },
+    {
+      "epoch": 7.642857142857143,
+      "grad_norm": 0.609375,
+      "learning_rate": 5.3e-06,
+      "loss": 0.0229,
+      "step": 107
+    },
+    {
+      "epoch": 7.714285714285714,
+      "grad_norm": 0.380859375,
+      "learning_rate": 5.3500000000000004e-06,
+      "loss": 0.022,
+      "step": 108
+    },
+    {
+      "epoch": 7.785714285714286,
+      "grad_norm": 0.2890625,
+      "learning_rate": 5.4e-06,
+      "loss": 0.0217,
+      "step": 109
+    },
+    {
+      "epoch": 7.857142857142857,
+      "grad_norm": 0.34765625,
+      "learning_rate": 5.45e-06,
+      "loss": 0.0234,
+      "step": 110
+    },
+    {
+      "epoch": 7.928571428571429,
+      "grad_norm": 0.69140625,
+      "learning_rate": 5.500000000000001e-06,
+      "loss": 0.0241,
+      "step": 111
+    },
+    {
+      "epoch": 8.0,
+      "grad_norm": 0.890625,
+      "learning_rate": 5.55e-06,
+      "loss": 0.0242,
+      "step": 112
+    },
+    {
+      "epoch": 8.071428571428571,
+      "grad_norm": 0.306640625,
+      "learning_rate": 5.600000000000001e-06,
+      "loss": 0.0219,
+      "step": 113
+    },
+    {
+      "epoch": 8.142857142857142,
+      "grad_norm": 0.53515625,
+      "learning_rate": 5.65e-06,
+      "loss": 0.025,
+      "step": 114
+    },
+    {
+      "epoch": 8.214285714285714,
+      "grad_norm": 0.3828125,
+      "learning_rate": 5.7000000000000005e-06,
+      "loss": 0.0217,
+      "step": 115
+    },
+    {
+      "epoch": 8.285714285714286,
+      "grad_norm": 0.46484375,
+      "learning_rate": 5.750000000000001e-06,
+      "loss": 0.0211,
+      "step": 116
+    },
+    {
+      "epoch": 8.357142857142858,
+      "grad_norm": 0.52734375,
+      "learning_rate": 5.8e-06,
+      "loss": 0.0215,
+      "step": 117
+    },
+    {
+      "epoch": 8.428571428571429,
+      "grad_norm": 0.671875,
+      "learning_rate": 5.850000000000001e-06,
+      "loss": 0.0253,
+      "step": 118
+    },
+    {
+      "epoch": 8.5,
+      "grad_norm": 0.462890625,
+      "learning_rate": 5.9e-06,
+      "loss": 0.0215,
+      "step": 119
+    },
+    {
+      "epoch": 8.571428571428571,
+      "grad_norm": 1.1953125,
+      "learning_rate": 5.95e-06,
+      "loss": 0.0244,
+      "step": 120
+    },
+    {
+      "epoch": 8.642857142857142,
+      "grad_norm": 0.953125,
+      "learning_rate": 6e-06,
+      "loss": 0.0222,
+      "step": 121
+    },
+    {
+      "epoch": 8.714285714285714,
+      "grad_norm": 0.77734375,
+      "learning_rate": 6.0500000000000005e-06,
+      "loss": 0.0238,
+      "step": 122
+    },
+    {
+      "epoch": 8.785714285714286,
+      "grad_norm": 1.0390625,
+      "learning_rate": 6.1e-06,
+      "loss": 0.0233,
+      "step": 123
+    },
+    {
+      "epoch": 8.857142857142858,
+      "grad_norm": 0.40234375,
+      "learning_rate": 6.15e-06,
+      "loss": 0.0232,
+      "step": 124
+    },
+    {
+      "epoch": 8.928571428571429,
+      "grad_norm": 0.37890625,
+      "learning_rate": 6.2e-06,
+      "loss": 0.021,
+      "step": 125
+    },
+    {
+      "epoch": 9.0,
+      "grad_norm": 0.4140625,
+      "learning_rate": 6.25e-06,
+      "loss": 0.0221,
+      "step": 126
+    },
+    {
+      "epoch": 9.071428571428571,
+      "grad_norm": 1.2109375,
+      "learning_rate": 6.300000000000001e-06,
+      "loss": 0.0251,
+      "step": 127
+    },
+    {
+      "epoch": 9.142857142857142,
+      "grad_norm": 0.376953125,
+      "learning_rate": 6.35e-06,
+      "loss": 0.0232,
+      "step": 128
+    },
+    {
+      "epoch": 9.214285714285714,
+      "grad_norm": 0.42578125,
+      "learning_rate": 6.4000000000000006e-06,
+      "loss": 0.0219,
+      "step": 129
+    },
+    {
+      "epoch": 9.285714285714286,
+      "grad_norm": 0.3671875,
+      "learning_rate": 6.45e-06,
+      "loss": 0.0224,
+      "step": 130
+    },
+    {
+      "epoch": 9.357142857142858,
+      "grad_norm": 0.271484375,
+      "learning_rate": 6.5000000000000004e-06,
+      "loss": 0.0224,
+      "step": 131
+    },
+    {
+      "epoch": 9.428571428571429,
+      "grad_norm": 0.2158203125,
+      "learning_rate": 6.550000000000001e-06,
+      "loss": 0.0211,
+      "step": 132
+    },
+    {
+      "epoch": 9.5,
+      "grad_norm": 0.30078125,
+      "learning_rate": 6.6e-06,
+      "loss": 0.0223,
+      "step": 133
+    },
+    {
+      "epoch": 9.571428571428571,
+      "grad_norm": 0.412109375,
+      "learning_rate": 6.650000000000001e-06,
+      "loss": 0.0233,
+      "step": 134
+    },
+    {
+      "epoch": 9.642857142857142,
+      "grad_norm": 0.46484375,
+      "learning_rate": 6.700000000000001e-06,
+      "loss": 0.0212,
+      "step": 135
+    },
+    {
+      "epoch": 9.714285714285714,
+      "grad_norm": 0.4453125,
+      "learning_rate": 6.750000000000001e-06,
+      "loss": 0.0214,
+      "step": 136
+    },
+    {
+      "epoch": 9.785714285714286,
+      "grad_norm": 0.291015625,
+      "learning_rate": 6.800000000000001e-06,
+      "loss": 0.0216,
+      "step": 137
+    },
+    {
+      "epoch": 9.857142857142858,
+      "grad_norm": 0.55078125,
+      "learning_rate": 6.8500000000000005e-06,
+      "loss": 0.0222,
+      "step": 138
+    },
+    {
+      "epoch": 9.928571428571429,
+      "grad_norm": 0.765625,
+      "learning_rate": 6.900000000000001e-06,
+      "loss": 0.0215,
+      "step": 139
+    },
+    {
+      "epoch": 10.0,
+      "grad_norm": 0.82421875,
+      "learning_rate": 6.950000000000001e-06,
+      "loss": 0.0221,
+      "step": 140
+    },
+    {
+      "epoch": 10.071428571428571,
+      "grad_norm": 0.306640625,
+      "learning_rate": 7.000000000000001e-06,
+      "loss": 0.0219,
+      "step": 141
+    },
+    {
+      "epoch": 10.142857142857142,
+      "grad_norm": 0.73828125,
+      "learning_rate": 7.049999999999999e-06,
+      "loss": 0.022,
+      "step": 142
+    },
+    {
+      "epoch": 10.214285714285714,
+      "grad_norm": 0.392578125,
+      "learning_rate": 7.1e-06,
+      "loss": 0.0197,
+      "step": 143
+    },
+    {
+      "epoch": 10.285714285714286,
+      "grad_norm": 0.435546875,
+      "learning_rate": 7.15e-06,
+      "loss": 0.0174,
+      "step": 144
+    },
+    {
+      "epoch": 10.357142857142858,
+      "grad_norm": 0.6875,
+      "learning_rate": 7.2e-06,
+      "loss": 0.0224,
+      "step": 145
+    },
+    {
+      "epoch": 10.428571428571429,
+      "grad_norm": 0.404296875,
+      "learning_rate": 7.25e-06,
+      "loss": 0.0217,
+      "step": 146
+    },
+    {
+      "epoch": 10.5,
+      "grad_norm": 0.41015625,
+      "learning_rate": 7.2999999999999996e-06,
+      "loss": 0.0196,
+      "step": 147
+    },
+    {
+      "epoch": 10.571428571428571,
+      "grad_norm": 0.43359375,
+      "learning_rate": 7.35e-06,
+      "loss": 0.021,
+      "step": 148
+    },
+    {
+      "epoch": 10.642857142857142,
+      "grad_norm": 1.0390625,
+      "learning_rate": 7.4e-06,
+      "loss": 0.0205,
+      "step": 149
+    },
+    {
+      "epoch": 10.714285714285714,
+      "grad_norm": 0.7109375,
+      "learning_rate": 7.45e-06,
+      "loss": 0.0216,
+      "step": 150
+    },
+    {
+      "epoch": 10.785714285714286,
+      "grad_norm": 0.65234375,
+      "learning_rate": 7.5e-06,
+      "loss": 0.0214,
+      "step": 151
+    },
+    {
+      "epoch": 10.857142857142858,
+      "grad_norm": 0.455078125,
+      "learning_rate": 7.55e-06,
+      "loss": 0.0205,
+      "step": 152
+    },
+    {
+      "epoch": 10.928571428571429,
+      "grad_norm": 0.353515625,
+      "learning_rate": 7.6e-06,
+      "loss": 0.0201,
+      "step": 153
+    },
+    {
+      "epoch": 11.0,
+      "grad_norm": 0.53125,
+      "learning_rate": 7.65e-06,
+      "loss": 0.0206,
+      "step": 154
+    },
+    {
+      "epoch": 11.071428571428571,
+      "grad_norm": 0.275390625,
+      "learning_rate": 7.7e-06,
+      "loss": 0.0183,
+      "step": 155
+    },
+    {
+      "epoch": 11.142857142857142,
+      "grad_norm": 0.8828125,
+      "learning_rate": 7.75e-06,
+      "loss": 0.0211,
+      "step": 156
+    },
+    {
+      "epoch": 11.214285714285714,
+      "grad_norm": 0.462890625,
+      "learning_rate": 7.8e-06,
+      "loss": 0.0214,
+      "step": 157
+    },
+    {
+      "epoch": 11.285714285714286,
+      "grad_norm": 0.291015625,
+      "learning_rate": 7.850000000000001e-06,
+      "loss": 0.0234,
+      "step": 158
+    },
+    {
+      "epoch": 11.357142857142858,
+      "grad_norm": 0.3125,
+      "learning_rate": 7.9e-06,
+      "loss": 0.018,
+      "step": 159
+    },
+    {
+      "epoch": 11.428571428571429,
+      "grad_norm": 0.80859375,
+      "learning_rate": 7.95e-06,
+      "loss": 0.0202,
+      "step": 160
+    },
+    {
+      "epoch": 11.5,
+      "grad_norm": 0.265625,
+      "learning_rate": 8.000000000000001e-06,
+      "loss": 0.0192,
+      "step": 161
+    },
+    {
+      "epoch": 11.571428571428571,
+      "grad_norm": 0.2734375,
+      "learning_rate": 8.050000000000001e-06,
+      "loss": 0.0192,
+      "step": 162
+    },
+    {
+      "epoch": 11.642857142857142,
+      "grad_norm": 0.359375,
+      "learning_rate": 8.1e-06,
+      "loss": 0.0212,
+      "step": 163
+    },
+    {
+      "epoch": 11.714285714285714,
+      "grad_norm": 0.400390625,
+      "learning_rate": 8.15e-06,
+      "loss": 0.0202,
+      "step": 164
+    },
+    {
+      "epoch": 11.785714285714286,
+      "grad_norm": 0.443359375,
+      "learning_rate": 8.200000000000001e-06,
+      "loss": 0.0194,
+      "step": 165
+    },
+    {
+      "epoch": 11.857142857142858,
+      "grad_norm": 0.28125,
+      "learning_rate": 8.25e-06,
+      "loss": 0.0187,
+      "step": 166
+    },
+    {
+      "epoch": 11.928571428571429,
+      "grad_norm": 0.3515625,
+      "learning_rate": 8.3e-06,
+      "loss": 0.0224,
+      "step": 167
+    },
+    {
+      "epoch": 12.0,
+      "grad_norm": 0.296875,
+      "learning_rate": 8.350000000000001e-06,
+      "loss": 0.019,
+      "step": 168
+    },
+    {
+      "epoch": 12.071428571428571,
+      "grad_norm": 0.90625,
+      "learning_rate": 8.400000000000001e-06,
+      "loss": 0.0214,
+      "step": 169
+    },
+    {
+      "epoch": 12.142857142857142,
+      "grad_norm": 0.384765625,
+      "learning_rate": 8.45e-06,
+      "loss": 0.0199,
+      "step": 170
+    },
+    {
+      "epoch": 12.214285714285714,
+      "grad_norm": 0.5,
+      "learning_rate": 8.500000000000002e-06,
+      "loss": 0.0205,
+      "step": 171
+    },
+    {
+      "epoch": 12.285714285714286,
+      "grad_norm": 0.41015625,
+      "learning_rate": 8.550000000000001e-06,
+      "loss": 0.0183,
+      "step": 172
+    },
+    {
+      "epoch": 12.357142857142858,
+      "grad_norm": 0.345703125,
+      "learning_rate": 8.599999999999999e-06,
+      "loss": 0.0166,
+      "step": 173
+    },
+    {
+      "epoch": 12.428571428571429,
+      "grad_norm": 0.7890625,
+      "learning_rate": 8.65e-06,
+      "loss": 0.0214,
+      "step": 174
+    },
+    {
+      "epoch": 12.5,
+      "grad_norm": 1.140625,
+      "learning_rate": 8.7e-06,
+      "loss": 0.0219,
+      "step": 175
+    },
+    {
+      "epoch": 12.571428571428571,
+      "grad_norm": 0.828125,
+      "learning_rate": 8.75e-06,
+      "loss": 0.0168,
+      "step": 176
+    },
+    {
+      "epoch": 12.642857142857142,
+      "grad_norm": 0.83203125,
+      "learning_rate": 8.8e-06,
+      "loss": 0.0203,
+      "step": 177
+    },
+    {
+      "epoch": 12.714285714285714,
+      "grad_norm": 1.03125,
+      "learning_rate": 8.85e-06,
+      "loss": 0.0199,
+      "step": 178
+    },
+    {
+      "epoch": 12.785714285714286,
+      "grad_norm": 0.392578125,
+      "learning_rate": 8.9e-06,
+      "loss": 0.0184,
+      "step": 179
+    },
+    {
+      "epoch": 12.857142857142858,
+      "grad_norm": 0.43359375,
+      "learning_rate": 8.95e-06,
+      "loss": 0.0186,
+      "step": 180
+    },
+    {
+      "epoch": 12.928571428571429,
+      "grad_norm": 0.5703125,
+      "learning_rate": 9e-06,
+      "loss": 0.0191,
+      "step": 181
+    },
+    {
+      "epoch": 13.0,
+      "grad_norm": 1.109375,
+      "learning_rate": 9.05e-06,
+      "loss": 0.0211,
+      "step": 182
+    },
+    {
+      "epoch": 13.071428571428571,
+      "grad_norm": 1.1328125,
+      "learning_rate": 9.100000000000001e-06,
+      "loss": 0.0204,
+      "step": 183
+    },
+    {
+      "epoch": 13.142857142857142,
+      "grad_norm": 0.87109375,
+      "learning_rate": 9.15e-06,
+      "loss": 0.0164,
+      "step": 184
+    },
+    {
+      "epoch": 13.214285714285714,
+      "grad_norm": 0.67578125,
+      "learning_rate": 9.2e-06,
+      "loss": 0.0209,
+      "step": 185
+    },
+    {
+      "epoch": 13.285714285714286,
+      "grad_norm": 0.8671875,
+      "learning_rate": 9.25e-06,
+      "loss": 0.0206,
+      "step": 186
+    },
+    {
+      "epoch": 13.357142857142858,
+      "grad_norm": 0.953125,
+      "learning_rate": 9.3e-06,
+      "loss": 0.0199,
+      "step": 187
+    },
+    {
+      "epoch": 13.428571428571429,
+      "grad_norm": 0.392578125,
+      "learning_rate": 9.35e-06,
+      "loss": 0.0199,
+      "step": 188
+    },
+    {
+      "epoch": 13.5,
+      "grad_norm": 0.6796875,
+      "learning_rate": 9.4e-06,
+      "loss": 0.0191,
+      "step": 189
+    },
+    {
+      "epoch": 13.571428571428571,
+      "grad_norm": 0.88671875,
+      "learning_rate": 9.450000000000001e-06,
+      "loss": 0.0209,
+      "step": 190
+    },
+    {
+      "epoch": 13.642857142857142,
+      "grad_norm": 0.77734375,
+      "learning_rate": 9.5e-06,
+      "loss": 0.0237,
+      "step": 191
+    },
+    {
+      "epoch": 13.714285714285714,
+      "grad_norm": 0.5546875,
+      "learning_rate": 9.55e-06,
+      "loss": 0.0193,
+      "step": 192
+    },
+    {
+      "epoch": 13.785714285714286,
+      "grad_norm": 0.345703125,
+      "learning_rate": 9.600000000000001e-06,
+      "loss": 0.0186,
+      "step": 193
+    },
+    {
+      "epoch": 13.857142857142858,
+      "grad_norm": 0.95703125,
+      "learning_rate": 9.65e-06,
+      "loss": 0.0177,
+      "step": 194
+    },
+    {
+      "epoch": 13.928571428571429,
+      "grad_norm": 0.265625,
+      "learning_rate": 9.7e-06,
+      "loss": 0.0206,
+      "step": 195
+    },
+    {
+      "epoch": 14.0,
+      "grad_norm": 0.55078125,
+      "learning_rate": 9.750000000000002e-06,
+      "loss": 0.0181,
+      "step": 196
+    },
+    {
+      "epoch": 14.071428571428571,
+      "grad_norm": 0.89453125,
+      "learning_rate": 9.800000000000001e-06,
+      "loss": 0.0194,
+      "step": 197
+    },
+    {
+      "epoch": 14.142857142857142,
+      "grad_norm": 0.279296875,
+      "learning_rate": 9.85e-06,
+      "loss": 0.0212,
+      "step": 198
+    },
+    {
+      "epoch": 14.214285714285714,
+      "grad_norm": 0.369140625,
+      "learning_rate": 9.900000000000002e-06,
+      "loss": 0.0178,
+      "step": 199
+    },
+    {
+      "epoch": 14.285714285714286,
+      "grad_norm": 0.75,
+      "learning_rate": 9.950000000000001e-06,
+      "loss": 0.02,
+      "step": 200
+    },
+    {
+      "epoch": 14.357142857142858,
+      "grad_norm": 0.50390625,
+      "learning_rate": 1e-05,
+      "loss": 0.0184,
+      "step": 201
+    },
+    {
+      "epoch": 14.428571428571429,
+      "grad_norm": 0.275390625,
+      "learning_rate": 1.005e-05,
+      "loss": 0.0196,
+      "step": 202
+    },
+    {
+      "epoch": 14.5,
+      "grad_norm": 0.50390625,
+      "learning_rate": 1.0100000000000002e-05,
+      "loss": 0.018,
+      "step": 203
+    },
+    {
+      "epoch": 14.571428571428571,
+      "grad_norm": 0.3125,
+      "learning_rate": 1.0150000000000001e-05,
+      "loss": 0.0175,
+      "step": 204
+    },
+    {
+      "epoch": 14.642857142857142,
+      "grad_norm": 0.2890625,
+      "learning_rate": 1.02e-05,
+      "loss": 0.0192,
+      "step": 205
+    },
+    {
+      "epoch": 14.714285714285714,
+      "grad_norm": 0.326171875,
+      "learning_rate": 1.025e-05,
+      "loss": 0.0177,
+      "step": 206
+    },
+    {
+      "epoch": 14.785714285714286,
+      "grad_norm": 0.46484375,
+      "learning_rate": 1.03e-05,
+      "loss": 0.0177,
+      "step": 207
+    },
+    {
+      "epoch": 14.857142857142858,
+      "grad_norm": 0.28125,
+      "learning_rate": 1.035e-05,
+      "loss": 0.0189,
+      "step": 208
+    },
+    {
+      "epoch": 14.928571428571429,
+      "grad_norm": 0.416015625,
+      "learning_rate": 1.04e-05,
+      "loss": 0.0188,
+      "step": 209
+    },
+    {
+      "epoch": 15.0,
+      "grad_norm": 0.296875,
+      "learning_rate": 1.045e-05,
+      "loss": 0.017,
+      "step": 210
+    },
+    {
+      "epoch": 15.071428571428571,
+      "grad_norm": 0.28125,
+      "learning_rate": 1.05e-05,
+      "loss": 0.0157,
+      "step": 211
+    },
+    {
+      "epoch": 15.142857142857142,
+      "grad_norm": 0.3828125,
+      "learning_rate": 1.055e-05,
+      "loss": 0.0179,
+      "step": 212
+    },
+    {
+      "epoch": 15.214285714285714,
+      "grad_norm": 0.25,
+      "learning_rate": 1.06e-05,
+      "loss": 0.0173,
+      "step": 213
+    },
+    {
+      "epoch": 15.285714285714286,
+      "grad_norm": 1.4765625,
+      "learning_rate": 1.065e-05,
+      "loss": 0.0196,
+      "step": 214
+    },
+    {
+      "epoch": 15.357142857142858,
+      "grad_norm": 0.46484375,
+      "learning_rate": 1.0700000000000001e-05,
+      "loss": 0.0167,
+      "step": 215
+    },
+    {
+      "epoch": 15.428571428571429,
+      "grad_norm": 0.29296875,
+      "learning_rate": 1.075e-05,
+      "loss": 0.0205,
+      "step": 216
+    },
+    {
+      "epoch": 15.5,
+      "grad_norm": 0.8046875,
+      "learning_rate": 1.08e-05,
+      "loss": 0.0189,
+      "step": 217
+    },
+    {
+      "epoch": 15.571428571428571,
+      "grad_norm": 0.80078125,
+      "learning_rate": 1.0850000000000001e-05,
+      "loss": 0.0174,
+      "step": 218
+    },
+    {
+      "epoch": 15.642857142857142,
+      "grad_norm": 0.66796875,
+      "learning_rate": 1.09e-05,
+      "loss": 0.0186,
+      "step": 219
+    },
+    {
+      "epoch": 15.714285714285714,
+      "grad_norm": 0.30859375,
+      "learning_rate": 1.095e-05,
+      "loss": 0.0159,
+      "step": 220
+    },
+    {
+      "epoch": 15.785714285714286,
+      "grad_norm": 0.404296875,
+      "learning_rate": 1.1000000000000001e-05,
+      "loss": 0.0169,
+      "step": 221
+    },
+    {
+      "epoch": 15.857142857142858,
+      "grad_norm": 1.28125,
+      "learning_rate": 1.1050000000000001e-05,
+      "loss": 0.0175,
+      "step": 222
+    },
+    {
+      "epoch": 15.928571428571429,
+      "grad_norm": 0.376953125,
+      "learning_rate": 1.11e-05,
+      "loss": 0.0168,
+      "step": 223
+    },
+    {
+      "epoch": 16.0,
+      "grad_norm": 1.21875,
+      "learning_rate": 1.115e-05,
+      "loss": 0.0225,
+      "step": 224
+    },
+    {
+      "epoch": 16.071428571428573,
+      "grad_norm": 0.27734375,
+      "learning_rate": 1.1200000000000001e-05,
+      "loss": 0.0176,
+      "step": 225
+    },
+    {
+      "epoch": 16.142857142857142,
+      "grad_norm": 0.37890625,
+      "learning_rate": 1.125e-05,
+      "loss": 0.0195,
+      "step": 226
+    },
+    {
+      "epoch": 16.214285714285715,
+      "grad_norm": 0.54296875,
+      "learning_rate": 1.13e-05,
+      "loss": 0.0181,
+      "step": 227
+    },
+    {
+      "epoch": 16.285714285714285,
+      "grad_norm": 0.83984375,
+      "learning_rate": 1.1350000000000001e-05,
+      "loss": 0.0177,
+      "step": 228
+    },
+    {
+      "epoch": 16.357142857142858,
+      "grad_norm": 0.609375,
+      "learning_rate": 1.1400000000000001e-05,
+      "loss": 0.0201,
+      "step": 229
+    },
+    {
+      "epoch": 16.428571428571427,
+      "grad_norm": 0.416015625,
+      "learning_rate": 1.145e-05,
+      "loss": 0.0176,
+      "step": 230
+    },
+    {
+      "epoch": 16.5,
+      "grad_norm": 0.578125,
+      "learning_rate": 1.1500000000000002e-05,
+      "loss": 0.0196,
+      "step": 231
+    },
+    {
+      "epoch": 16.571428571428573,
+      "grad_norm": 0.470703125,
+      "learning_rate": 1.1550000000000001e-05,
+      "loss": 0.0187,
+      "step": 232
+    },
+    {
+      "epoch": 16.642857142857142,
+      "grad_norm": 0.73046875,
+      "learning_rate": 1.16e-05,
+      "loss": 0.0177,
+      "step": 233
+    },
+    {
+      "epoch": 16.714285714285715,
+      "grad_norm": 0.263671875,
+      "learning_rate": 1.1650000000000002e-05,
+      "loss": 0.0184,
+      "step": 234
+    },
+    {
+      "epoch": 16.785714285714285,
+      "grad_norm": 1.296875,
+      "learning_rate": 1.1700000000000001e-05,
+      "loss": 0.0202,
+      "step": 235
+    },
+    {
+      "epoch": 16.857142857142858,
+      "grad_norm": 0.5078125,
+      "learning_rate": 1.175e-05,
+      "loss": 0.0189,
+      "step": 236
+    },
+    {
+      "epoch": 16.928571428571427,
+      "grad_norm": 0.341796875,
+      "learning_rate": 1.18e-05,
+      "loss": 0.017,
+      "step": 237
+    },
+    {
+      "epoch": 17.0,
+      "grad_norm": 0.33984375,
+      "learning_rate": 1.185e-05,
+      "loss": 0.0167,
+      "step": 238
+    },
+    {
+      "epoch": 17.071428571428573,
+      "grad_norm": 0.349609375,
+      "learning_rate": 1.19e-05,
+      "loss": 0.0161,
+      "step": 239
+    },
+    {
+      "epoch": 17.142857142857142,
+      "grad_norm": 0.357421875,
+      "learning_rate": 1.195e-05,
+      "loss": 0.0179,
+      "step": 240
+    },
+    {
+      "epoch": 17.214285714285715,
+      "grad_norm": 0.35546875,
+      "learning_rate": 1.2e-05,
+      "loss": 0.0211,
+      "step": 241
+    },
+    {
+      "epoch": 17.285714285714285,
+      "grad_norm": 0.6953125,
+      "learning_rate": 1.205e-05,
+      "loss": 0.015,
+      "step": 242
+    },
+    {
+      "epoch": 17.357142857142858,
+      "grad_norm": 0.69921875,
+      "learning_rate": 1.2100000000000001e-05,
+      "loss": 0.0197,
+      "step": 243
+    },
+    {
+      "epoch": 17.428571428571427,
+      "grad_norm": 0.4375,
+      "learning_rate": 1.215e-05,
+      "loss": 0.0181,
+      "step": 244
+    },
+    {
+      "epoch": 17.5,
+      "grad_norm": 0.41796875,
+      "learning_rate": 1.22e-05,
+      "loss": 0.017,
+      "step": 245
+    },
+    {
+      "epoch": 17.571428571428573,
+      "grad_norm": 0.212890625,
+      "learning_rate": 1.225e-05,
+      "loss": 0.0161,
+      "step": 246
+    },
+    {
+      "epoch": 17.642857142857142,
+      "grad_norm": 0.275390625,
+      "learning_rate": 1.23e-05,
+      "loss": 0.0195,
+      "step": 247
+    },
+    {
+      "epoch": 17.714285714285715,
+      "grad_norm": 0.39453125,
+      "learning_rate": 1.235e-05,
+      "loss": 0.0189,
+      "step": 248
+    },
+    {
+      "epoch": 17.785714285714285,
+      "grad_norm": 0.365234375,
+      "learning_rate": 1.24e-05,
+      "loss": 0.0191,
+      "step": 249
+    },
+    {
+      "epoch": 17.857142857142858,
+      "grad_norm": 0.3359375,
+      "learning_rate": 1.2450000000000001e-05,
+      "loss": 0.0175,
+      "step": 250
+    },
+    {
+      "epoch": 17.928571428571427,
+      "grad_norm": 0.26171875,
+      "learning_rate": 1.25e-05,
+      "loss": 0.0179,
+      "step": 251
+    },
+    {
+      "epoch": 18.0,
+      "grad_norm": 0.45703125,
+      "learning_rate": 1.255e-05,
+      "loss": 0.0169,
+      "step": 252
+    },
+    {
+      "epoch": 18.071428571428573,
+      "grad_norm": 0.34765625,
+      "learning_rate": 1.2600000000000001e-05,
+      "loss": 0.0176,
+      "step": 253
+    },
+    {
+      "epoch": 18.142857142857142,
+      "grad_norm": 0.5390625,
+      "learning_rate": 1.2650000000000001e-05,
+      "loss": 0.019,
+      "step": 254
+    },
+    {
+      "epoch": 18.214285714285715,
+      "grad_norm": 0.34375,
+      "learning_rate": 1.27e-05,
+      "loss": 0.0153,
+      "step": 255
+    },
+    {
+      "epoch": 18.285714285714285,
+      "grad_norm": 0.44140625,
+      "learning_rate": 1.2750000000000002e-05,
+      "loss": 0.0171,
+      "step": 256
+    },
+    {
+      "epoch": 18.357142857142858,
+      "grad_norm": 0.6640625,
+      "learning_rate": 1.2800000000000001e-05,
+      "loss": 0.0174,
+      "step": 257
+    },
+    {
+      "epoch": 18.428571428571427,
+      "grad_norm": 0.74609375,
+      "learning_rate": 1.285e-05,
+      "loss": 0.0179,
+      "step": 258
+    },
+    {
+      "epoch": 18.5,
+      "grad_norm": 0.49609375,
+      "learning_rate": 1.29e-05,
+      "loss": 0.0178,
+      "step": 259
+    },
+    {
+      "epoch": 18.571428571428573,
+      "grad_norm": 0.578125,
+      "learning_rate": 1.2950000000000001e-05,
+      "loss": 0.0187,
+      "step": 260
+    },
+    {
+      "epoch": 18.642857142857142,
+      "grad_norm": 0.6484375,
+      "learning_rate": 1.3000000000000001e-05,
+      "loss": 0.019,
+      "step": 261
+    },
+    {
+      "epoch": 18.714285714285715,
+      "grad_norm": 0.33203125,
+      "learning_rate": 1.305e-05,
+      "loss": 0.0174,
+      "step": 262
+    },
+    {
+      "epoch": 18.785714285714285,
+      "grad_norm": 0.296875,
+      "learning_rate": 1.3100000000000002e-05,
+      "loss": 0.0191,
+      "step": 263
+    },
+    {
+      "epoch": 18.857142857142858,
+      "grad_norm": 0.72265625,
+      "learning_rate": 1.3150000000000001e-05,
+      "loss": 0.0179,
+      "step": 264
+    },
+    {
+      "epoch": 18.928571428571427,
+      "grad_norm": 0.88671875,
+      "learning_rate": 1.32e-05,
+      "loss": 0.017,
+      "step": 265
+    },
+    {
+      "epoch": 19.0,
+      "grad_norm": 0.37890625,
+      "learning_rate": 1.3250000000000002e-05,
+      "loss": 0.0149,
+      "step": 266
+    },
+    {
+      "epoch": 19.071428571428573,
+      "grad_norm": 0.447265625,
+      "learning_rate": 1.3300000000000001e-05,
+      "loss": 0.0202,
+      "step": 267
+    },
+    {
+      "epoch": 19.142857142857142,
+      "grad_norm": 0.30859375,
+      "learning_rate": 1.3350000000000001e-05,
+      "loss": 0.018,
+      "step": 268
+    },
+    {
+      "epoch": 19.214285714285715,
+      "grad_norm": 0.52734375,
+      "learning_rate": 1.3400000000000002e-05,
+      "loss": 0.017,
+      "step": 269
+    },
+    {
+      "epoch": 19.285714285714285,
+      "grad_norm": 0.3125,
+      "learning_rate": 1.3450000000000002e-05,
+      "loss": 0.0185,
+      "step": 270
+    },
+    {
+      "epoch": 19.357142857142858,
+      "grad_norm": 0.7890625,
+      "learning_rate": 1.3500000000000001e-05,
+      "loss": 0.0186,
+      "step": 271
+    },
+    {
+      "epoch": 19.428571428571427,
+      "grad_norm": 0.271484375,
+      "learning_rate": 1.3550000000000002e-05,
+      "loss": 0.0147,
+      "step": 272
+    },
+    {
+      "epoch": 19.5,
+      "grad_norm": 0.419921875,
+      "learning_rate": 1.3600000000000002e-05,
+      "loss": 0.0162,
+      "step": 273
+    },
+    {
+      "epoch": 19.571428571428573,
+      "grad_norm": 0.80859375,
+      "learning_rate": 1.3650000000000001e-05,
+      "loss": 0.0171,
+      "step": 274
+    },
+    {
+      "epoch": 19.642857142857142,
+      "grad_norm": 0.86328125,
+      "learning_rate": 1.3700000000000001e-05,
+      "loss": 0.0171,
+      "step": 275
+    },
+    {
+      "epoch": 19.714285714285715,
+      "grad_norm": 0.427734375,
+      "learning_rate": 1.3750000000000002e-05,
+      "loss": 0.0184,
+      "step": 276
+    },
+    {
+      "epoch": 19.785714285714285,
+      "grad_norm": 0.44140625,
+      "learning_rate": 1.3800000000000002e-05,
+      "loss": 0.0151,
+      "step": 277
+    },
+    {
+      "epoch": 19.857142857142858,
+      "grad_norm": 0.283203125,
+      "learning_rate": 1.3850000000000001e-05,
+      "loss": 0.0162,
+      "step": 278
+    },
+    {
+      "epoch": 19.928571428571427,
+      "grad_norm": 0.6953125,
+      "learning_rate": 1.3900000000000002e-05,
+      "loss": 0.0195,
+      "step": 279
+    },
+    {
+      "epoch": 20.0,
+      "grad_norm": 0.578125,
+      "learning_rate": 1.3950000000000002e-05,
+      "loss": 0.0177,
+      "step": 280
+    },
+    {
+      "epoch": 20.071428571428573,
+      "grad_norm": 0.65625,
+      "learning_rate": 1.4000000000000001e-05,
+      "loss": 0.0169,
+      "step": 281
+    },
+    {
+      "epoch": 20.142857142857142,
+      "grad_norm": 0.259765625,
+      "learning_rate": 1.4050000000000003e-05,
+      "loss": 0.0165,
+      "step": 282
+    },
+    {
+      "epoch": 20.214285714285715,
+      "grad_norm": 0.69140625,
+      "learning_rate": 1.4099999999999999e-05,
+      "loss": 0.0166,
+      "step": 283
+    },
+    {
+      "epoch": 20.285714285714285,
+      "grad_norm": 0.306640625,
+      "learning_rate": 1.415e-05,
+      "loss": 0.0181,
+      "step": 284
+    },
+    {
+      "epoch": 20.357142857142858,
+      "grad_norm": 0.494140625,
+      "learning_rate": 1.42e-05,
+      "loss": 0.0168,
+      "step": 285
+    },
+    {
+      "epoch": 20.428571428571427,
+      "grad_norm": 0.421875,
+      "learning_rate": 1.4249999999999999e-05,
+      "loss": 0.0176,
+      "step": 286
+    },
+    {
+      "epoch": 20.5,
+      "grad_norm": 0.3515625,
+      "learning_rate": 1.43e-05,
+      "loss": 0.0179,
+      "step": 287
+    },
+    {
+      "epoch": 20.571428571428573,
+      "grad_norm": 0.578125,
+      "learning_rate": 1.435e-05,
+      "loss": 0.0145,
+      "step": 288
+    },
+    {
+      "epoch": 20.642857142857142,
+      "grad_norm": 0.244140625,
+      "learning_rate": 1.44e-05,
+      "loss": 0.0168,
+      "step": 289
+    },
+    {
+      "epoch": 20.714285714285715,
+      "grad_norm": 0.9140625,
+      "learning_rate": 1.4449999999999999e-05,
+      "loss": 0.0173,
+      "step": 290
+    },
+    {
+      "epoch": 20.785714285714285,
+      "grad_norm": 0.23828125,
+      "learning_rate": 1.45e-05,
+      "loss": 0.0142,
+      "step": 291
+    },
+    {
+      "epoch": 20.857142857142858,
+      "grad_norm": 0.34375,
+      "learning_rate": 1.455e-05,
+      "loss": 0.014,
+      "step": 292
+    },
+    {
+      "epoch": 20.928571428571427,
+      "grad_norm": 0.35546875,
+      "learning_rate": 1.4599999999999999e-05,
+      "loss": 0.0189,
+      "step": 293
+    },
+    {
+      "epoch": 21.0,
+      "grad_norm": 0.353515625,
+      "learning_rate": 1.465e-05,
+      "loss": 0.0146,
+      "step": 294
+    },
+    {
+      "epoch": 21.071428571428573,
+      "grad_norm": 0.392578125,
+      "learning_rate": 1.47e-05,
+      "loss": 0.0153,
+      "step": 295
+    },
+    {
+      "epoch": 21.142857142857142,
+      "grad_norm": 0.337890625,
+      "learning_rate": 1.475e-05,
+      "loss": 0.0188,
+      "step": 296
+    },
+    {
+      "epoch": 21.214285714285715,
+      "grad_norm": 0.515625,
+      "learning_rate": 1.48e-05,
+      "loss": 0.0156,
+      "step": 297
+    },
+    {
+      "epoch": 21.285714285714285,
+      "grad_norm": 0.515625,
+      "learning_rate": 1.485e-05,
+      "loss": 0.0163,
+      "step": 298
+    },
+    {
+      "epoch": 21.357142857142858,
+      "grad_norm": 0.37109375,
+      "learning_rate": 1.49e-05,
+      "loss": 0.017,
+      "step": 299
+    },
+    {
+      "epoch": 21.428571428571427,
+      "grad_norm": 0.482421875,
+      "learning_rate": 1.4950000000000001e-05,
+      "loss": 0.0162,
+      "step": 300
+    },
+    {
+      "epoch": 21.5,
+      "grad_norm": 0.5390625,
+      "learning_rate": 1.5e-05,
+      "loss": 0.0142,
+      "step": 301
+    },
+    {
+      "epoch": 21.571428571428573,
+      "grad_norm": 0.28125,
+      "learning_rate": 1.505e-05,
+      "loss": 0.0149,
+      "step": 302
+    },
+    {
+      "epoch": 21.642857142857142,
+      "grad_norm": 0.36328125,
+      "learning_rate": 1.51e-05,
+      "loss": 0.0156,
+      "step": 303
+    },
+    {
+      "epoch": 21.714285714285715,
+      "grad_norm": 0.8671875,
+      "learning_rate": 1.515e-05,
+      "loss": 0.0189,
+      "step": 304
+    },
+    {
+      "epoch": 21.785714285714285,
+      "grad_norm": 0.451171875,
+      "learning_rate": 1.52e-05,
+      "loss": 0.0177,
+      "step": 305
+    },
+    {
+      "epoch": 21.857142857142858,
+      "grad_norm": 0.640625,
+      "learning_rate": 1.525e-05,
+      "loss": 0.0159,
+      "step": 306
+    },
+    {
+      "epoch": 21.928571428571427,
+      "grad_norm": 0.314453125,
+      "learning_rate": 1.53e-05,
+      "loss": 0.014,
+      "step": 307
+    },
+    {
+      "epoch": 22.0,
+      "grad_norm": 0.359375,
+      "learning_rate": 1.535e-05,
+      "loss": 0.0153,
+      "step": 308
+    },
+    {
+      "epoch": 22.071428571428573,
+      "grad_norm": 0.578125,
+      "learning_rate": 1.54e-05,
+      "loss": 0.0147,
+      "step": 309
+    },
+    {
+      "epoch": 22.142857142857142,
+      "grad_norm": 0.298828125,
+      "learning_rate": 1.545e-05,
+      "loss": 0.0159,
+      "step": 310
+    },
+    {
+      "epoch": 22.214285714285715,
+      "grad_norm": 0.609375,
+      "learning_rate": 1.55e-05,
+      "loss": 0.0154,
+      "step": 311
+    },
+    {
+      "epoch": 22.285714285714285,
+      "grad_norm": 0.259765625,
+      "learning_rate": 1.5550000000000002e-05,
+      "loss": 0.0162,
+      "step": 312
+    },
+    {
+      "epoch": 22.357142857142858,
+      "grad_norm": 0.2578125,
+      "learning_rate": 1.56e-05,
+      "loss": 0.0164,
+      "step": 313
+    },
+    {
+      "epoch": 22.428571428571427,
+      "grad_norm": 0.28515625,
+      "learning_rate": 1.565e-05,
+      "loss": 0.015,
+      "step": 314
+    },
+    {
+      "epoch": 22.5,
+      "grad_norm": 0.8125,
+      "learning_rate": 1.5700000000000002e-05,
+      "loss": 0.0172,
+      "step": 315
+    },
+    {
+      "epoch": 22.571428571428573,
+      "grad_norm": 0.8515625,
+      "learning_rate": 1.575e-05,
+      "loss": 0.0178,
+      "step": 316
+    },
+    {
+      "epoch": 22.642857142857142,
+      "grad_norm": 0.91015625,
+      "learning_rate": 1.58e-05,
+      "loss": 0.0176,
+      "step": 317
+    },
+    {
+      "epoch": 22.714285714285715,
+      "grad_norm": 0.8203125,
+      "learning_rate": 1.5850000000000002e-05,
+      "loss": 0.0179,
+      "step": 318
+    },
+    {
+      "epoch": 22.785714285714285,
+      "grad_norm": 0.72265625,
+      "learning_rate": 1.59e-05,
+      "loss": 0.0153,
+      "step": 319
+    },
+    {
+      "epoch": 22.857142857142858,
+      "grad_norm": 0.423828125,
+      "learning_rate": 1.595e-05,
+      "loss": 0.0171,
+      "step": 320
+    },
+    {
+      "epoch": 22.928571428571427,
+      "grad_norm": 0.64453125,
+      "learning_rate": 1.6000000000000003e-05,
+      "loss": 0.0173,
+      "step": 321
+    },
+    {
+      "epoch": 23.0,
+      "grad_norm": 0.5859375,
+      "learning_rate": 1.605e-05,
+      "loss": 0.0164,
+      "step": 322
+    },
+    {
+      "epoch": 23.071428571428573,
+      "grad_norm": 1.125,
+      "learning_rate": 1.6100000000000002e-05,
+      "loss": 0.017,
+      "step": 323
+    },
+    {
+      "epoch": 23.142857142857142,
+      "grad_norm": 0.2470703125,
+      "learning_rate": 1.6150000000000003e-05,
+      "loss": 0.0172,
+      "step": 324
+    },
+    {
+      "epoch": 23.214285714285715,
+      "grad_norm": 0.27734375,
+      "learning_rate": 1.62e-05,
+      "loss": 0.0175,
+      "step": 325
+    },
+    {
+      "epoch": 23.285714285714285,
+      "grad_norm": 0.88671875,
+      "learning_rate": 1.6250000000000002e-05,
+      "loss": 0.0152,
+      "step": 326
+    },
+    {
+      "epoch": 23.357142857142858,
+      "grad_norm": 0.95703125,
+      "learning_rate": 1.63e-05,
+      "loss": 0.0189,
+      "step": 327
+    },
+    {
+      "epoch": 23.428571428571427,
+      "grad_norm": 0.6953125,
+      "learning_rate": 1.635e-05,
+      "loss": 0.0184,
+      "step": 328
+    },
+    {
+      "epoch": 23.5,
+      "grad_norm": 0.294921875,
+      "learning_rate": 1.6400000000000002e-05,
+      "loss": 0.0174,
+      "step": 329
+    },
+    {
+      "epoch": 23.571428571428573,
+      "grad_norm": 1.171875,
+      "learning_rate": 1.645e-05,
+      "loss": 0.0179,
+      "step": 330
+    },
+    {
+      "epoch": 23.642857142857142,
+      "grad_norm": 0.9375,
+      "learning_rate": 1.65e-05,
+      "loss": 0.0181,
+      "step": 331
+    },
+    {
+      "epoch": 23.714285714285715,
+      "grad_norm": 0.494140625,
+      "learning_rate": 1.6550000000000002e-05,
+      "loss": 0.0182,
+      "step": 332
+    },
+    {
+      "epoch": 23.785714285714285,
+      "grad_norm": 0.283203125,
+      "learning_rate": 1.66e-05,
+      "loss": 0.017,
+      "step": 333
+    },
+    {
+      "epoch": 23.857142857142858,
+      "grad_norm": 0.62890625,
+      "learning_rate": 1.665e-05,
+      "loss": 0.0164,
+      "step": 334
+    },
+    {
+      "epoch": 23.928571428571427,
+      "grad_norm": 0.91796875,
+      "learning_rate": 1.6700000000000003e-05,
+      "loss": 0.0159,
+      "step": 335
+    },
+    {
+      "epoch": 24.0,
+      "grad_norm": 0.4765625,
+      "learning_rate": 1.675e-05,
+      "loss": 0.0182,
+      "step": 336
+    },
+    {
+      "epoch": 24.071428571428573,
+      "grad_norm": 0.455078125,
+      "learning_rate": 1.6800000000000002e-05,
+      "loss": 0.0149,
+      "step": 337
+    },
+    {
+      "epoch": 24.142857142857142,
+      "grad_norm": 0.2177734375,
+      "learning_rate": 1.6850000000000003e-05,
+      "loss": 0.0152,
+      "step": 338
+    },
+    {
+      "epoch": 24.214285714285715,
+      "grad_norm": 0.9609375,
+      "learning_rate": 1.69e-05,
+      "loss": 0.0168,
+      "step": 339
+    },
+    {
+      "epoch": 24.285714285714285,
+      "grad_norm": 0.7421875,
+      "learning_rate": 1.6950000000000002e-05,
+      "loss": 0.0167,
+      "step": 340
+    },
+    {
+      "epoch": 24.357142857142858,
+      "grad_norm": 0.443359375,
+      "learning_rate": 1.7000000000000003e-05,
+      "loss": 0.0143,
+      "step": 341
+    },
+    {
+      "epoch": 24.428571428571427,
+      "grad_norm": 0.375,
+      "learning_rate": 1.705e-05,
+      "loss": 0.0168,
+      "step": 342
+    },
+    {
+      "epoch": 24.5,
+      "grad_norm": 0.515625,
+      "learning_rate": 1.7100000000000002e-05,
+      "loss": 0.0169,
+      "step": 343
+    },
+    {
+      "epoch": 24.571428571428573,
+      "grad_norm": 0.2236328125,
+      "learning_rate": 1.7150000000000004e-05,
+      "loss": 0.0176,
+      "step": 344
+    },
+    {
+      "epoch": 24.642857142857142,
+      "grad_norm": 0.2158203125,
+      "learning_rate": 1.7199999999999998e-05,
+      "loss": 0.016,
+      "step": 345
+    },
+    {
+      "epoch": 24.714285714285715,
+      "grad_norm": 0.5234375,
+      "learning_rate": 1.725e-05,
+      "loss": 0.0169,
+      "step": 346
+    },
+    {
+      "epoch": 24.785714285714285,
+      "grad_norm": 0.478515625,
+      "learning_rate": 1.73e-05,
+      "loss": 0.0163,
+      "step": 347
+    },
+    {
+      "epoch": 24.857142857142858,
+      "grad_norm": 0.2109375,
+      "learning_rate": 1.7349999999999998e-05,
+      "loss": 0.0143,
+      "step": 348
+    },
+    {
+      "epoch": 24.928571428571427,
+      "grad_norm": 0.263671875,
+      "learning_rate": 1.74e-05,
+      "loss": 0.0196,
+      "step": 349
+    },
+    {
+      "epoch": 25.0,
+      "grad_norm": 0.2353515625,
+      "learning_rate": 1.745e-05,
+      "loss": 0.0173,
+      "step": 350
+    },
+    {
+      "epoch": 25.071428571428573,
+      "grad_norm": 0.24609375,
+      "learning_rate": 1.75e-05,
+      "loss": 0.0159,
+      "step": 351
+    },
+    {
+      "epoch": 25.142857142857142,
+      "grad_norm": 0.486328125,
+      "learning_rate": 1.755e-05,
+      "loss": 0.0168,
+      "step": 352
+    },
+    {
+      "epoch": 25.214285714285715,
+      "grad_norm": 0.251953125,
+      "learning_rate": 1.76e-05,
+      "loss": 0.0145,
+      "step": 353
+    },
+    {
+      "epoch": 25.285714285714285,
+      "grad_norm": 0.49609375,
+      "learning_rate": 1.765e-05,
+      "loss": 0.0147,
+      "step": 354
+    },
+    {
+      "epoch": 25.357142857142858,
+      "grad_norm": 0.251953125,
+      "learning_rate": 1.77e-05,
+      "loss": 0.0159,
+      "step": 355
+    },
+    {
+      "epoch": 25.428571428571427,
+      "grad_norm": 0.26953125,
+      "learning_rate": 1.775e-05,
+      "loss": 0.017,
+      "step": 356
+    },
+    {
+      "epoch": 25.5,
+      "grad_norm": 0.28515625,
+      "learning_rate": 1.78e-05,
+      "loss": 0.0147,
+      "step": 357
+    },
+    {
+      "epoch": 25.571428571428573,
+      "grad_norm": 0.298828125,
+      "learning_rate": 1.785e-05,
+      "loss": 0.0185,
+      "step": 358
+    },
+    {
+      "epoch": 25.642857142857142,
+      "grad_norm": 0.365234375,
+      "learning_rate": 1.79e-05,
+      "loss": 0.0182,
+      "step": 359
+    },
+    {
+      "epoch": 25.714285714285715,
+      "grad_norm": 0.322265625,
+      "learning_rate": 1.795e-05,
+      "loss": 0.0183,
+      "step": 360
+    },
+    {
+      "epoch": 25.785714285714285,
+      "grad_norm": 0.53125,
+      "learning_rate": 1.8e-05,
+      "loss": 0.0184,
+      "step": 361
+    },
+    {
+      "epoch": 25.857142857142858,
+      "grad_norm": 0.60546875,
+      "learning_rate": 1.805e-05,
+      "loss": 0.0191,
+      "step": 362
+    },
+    {
+      "epoch": 25.928571428571427,
+      "grad_norm": 0.205078125,
+      "learning_rate": 1.81e-05,
+      "loss": 0.0187,
+      "step": 363
+    },
+    {
+      "epoch": 26.0,
+      "grad_norm": 0.1826171875,
+      "learning_rate": 1.815e-05,
+      "loss": 0.0156,
+      "step": 364
+    },
+    {
+      "epoch": 26.071428571428573,
+      "grad_norm": 0.408203125,
+      "learning_rate": 1.8200000000000002e-05,
+      "loss": 0.0125,
+      "step": 365
+    },
+    {
+      "epoch": 26.142857142857142,
+      "grad_norm": 0.333984375,
+      "learning_rate": 1.825e-05,
+      "loss": 0.0159,
+      "step": 366
+    },
+    {
+      "epoch": 26.214285714285715,
+      "grad_norm": 0.1982421875,
+      "learning_rate": 1.83e-05,
+      "loss": 0.0157,
+      "step": 367
+    },
+    {
+      "epoch": 26.285714285714285,
+      "grad_norm": 0.486328125,
+      "learning_rate": 1.8350000000000002e-05,
+      "loss": 0.0173,
+      "step": 368
+    },
+    {
+      "epoch": 26.357142857142858,
+      "grad_norm": 0.765625,
+      "learning_rate": 1.84e-05,
+      "loss": 0.016,
+      "step": 369
+    },
+    {
+      "epoch": 26.428571428571427,
+      "grad_norm": 0.265625,
+      "learning_rate": 1.845e-05,
+      "loss": 0.0169,
+      "step": 370
+    },
+    {
+      "epoch": 26.5,
+      "grad_norm": 0.376953125,
+      "learning_rate": 1.85e-05,
+      "loss": 0.0169,
+      "step": 371
+    },
+    {
+      "epoch": 26.571428571428573,
+      "grad_norm": 0.8515625,
+      "learning_rate": 1.855e-05,
+      "loss": 0.0157,
+      "step": 372
+    },
+    {
+      "epoch": 26.642857142857142,
+      "grad_norm": 0.216796875,
+      "learning_rate": 1.86e-05,
+      "loss": 0.0174,
+      "step": 373
+    },
+    {
+      "epoch": 26.714285714285715,
+      "grad_norm": 0.2333984375,
+      "learning_rate": 1.865e-05,
+      "loss": 0.0167,
+      "step": 374
+    },
+    {
+      "epoch": 26.785714285714285,
+      "grad_norm": 0.474609375,
+      "learning_rate": 1.87e-05,
+      "loss": 0.0208,
+      "step": 375
+    },
+    {
+      "epoch": 26.857142857142858,
+      "grad_norm": 0.5546875,
+      "learning_rate": 1.8750000000000002e-05,
+      "loss": 0.0184,
+      "step": 376
+    },
+    {
+      "epoch": 26.928571428571427,
+      "grad_norm": 0.5546875,
+      "learning_rate": 1.88e-05,
+      "loss": 0.0207,
+      "step": 377
+    },
+    {
+      "epoch": 27.0,
+      "grad_norm": 0.5859375,
+      "learning_rate": 1.885e-05,
+      "loss": 0.0133,
+      "step": 378
+    },
+    {
+      "epoch": 27.071428571428573,
+      "grad_norm": 0.349609375,
+      "learning_rate": 1.8900000000000002e-05,
+      "loss": 0.0168,
+      "step": 379
+    },
+    {
+      "epoch": 27.142857142857142,
+      "grad_norm": 0.59765625,
+      "learning_rate": 1.895e-05,
+      "loss": 0.0161,
+      "step": 380
+    },
+    {
+      "epoch": 27.214285714285715,
+      "grad_norm": 0.53125,
+      "learning_rate": 1.9e-05,
+      "loss": 0.016,
+      "step": 381
+    },
+    {
+      "epoch": 27.285714285714285,
+      "grad_norm": 0.2236328125,
+      "learning_rate": 1.9050000000000002e-05,
+      "loss": 0.0154,
+      "step": 382
+    },
+    {
+      "epoch": 27.357142857142858,
+      "grad_norm": 0.318359375,
+      "learning_rate": 1.91e-05,
+      "loss": 0.0155,
+      "step": 383
+    },
+    {
+      "epoch": 27.428571428571427,
+      "grad_norm": 0.318359375,
+      "learning_rate": 1.915e-05,
+      "loss": 0.0178,
+      "step": 384
+    },
+    {
+      "epoch": 27.5,
+      "grad_norm": 0.88671875,
+      "learning_rate": 1.9200000000000003e-05,
+      "loss": 0.0181,
+      "step": 385
+    },
+    {
+      "epoch": 27.571428571428573,
+      "grad_norm": 0.65625,
+      "learning_rate": 1.925e-05,
+      "loss": 0.0161,
+      "step": 386
+    },
+    {
+      "epoch": 27.642857142857142,
+      "grad_norm": 0.232421875,
+      "learning_rate": 1.93e-05,
+      "loss": 0.0171,
+      "step": 387
+    },
+    {
+      "epoch": 27.714285714285715,
+      "grad_norm": 0.287109375,
+      "learning_rate": 1.9350000000000003e-05,
+      "loss": 0.0157,
+      "step": 388
+    },
+    {
+      "epoch": 27.785714285714285,
+      "grad_norm": 0.5390625,
+      "learning_rate": 1.94e-05,
+      "loss": 0.0145,
+      "step": 389
+    },
+    {
+      "epoch": 27.857142857142858,
+      "grad_norm": 0.58203125,
+      "learning_rate": 1.9450000000000002e-05,
+      "loss": 0.0168,
+      "step": 390
+    },
+    {
+      "epoch": 27.928571428571427,
+      "grad_norm": 0.90234375,
+      "learning_rate": 1.9500000000000003e-05,
+      "loss": 0.018,
+      "step": 391
+    },
+    {
+      "epoch": 28.0,
+      "grad_norm": 0.275390625,
+      "learning_rate": 1.955e-05,
+      "loss": 0.0164,
+      "step": 392
+    },
+    {
+      "epoch": 28.071428571428573,
+      "grad_norm": 0.28125,
+      "learning_rate": 1.9600000000000002e-05,
+      "loss": 0.0159,
+      "step": 393
+    },
+    {
+      "epoch": 28.142857142857142,
+      "grad_norm": 0.6484375,
+      "learning_rate": 1.9650000000000003e-05,
+      "loss": 0.0168,
+      "step": 394
+    },
+    {
+      "epoch": 28.214285714285715,
+      "grad_norm": 0.80078125,
+      "learning_rate": 1.97e-05,
+      "loss": 0.0172,
+      "step": 395
+    },
+    {
+      "epoch": 28.285714285714285,
+      "grad_norm": 0.50390625,
+      "learning_rate": 1.9750000000000002e-05,
+      "loss": 0.017,
+      "step": 396
+    },
+    {
+      "epoch": 28.357142857142858,
+      "grad_norm": 0.412109375,
+      "learning_rate": 1.9800000000000004e-05,
+      "loss": 0.0181,
+      "step": 397
+    },
+    {
+      "epoch": 28.428571428571427,
+      "grad_norm": 0.26953125,
+      "learning_rate": 1.985e-05,
+      "loss": 0.0167,
+      "step": 398
+    },
+    {
+      "epoch": 28.5,
+      "grad_norm": 0.5703125,
+      "learning_rate": 1.9900000000000003e-05,
+      "loss": 0.02,
+      "step": 399
+    },
+    {
+      "epoch": 28.571428571428573,
+      "grad_norm": 0.392578125,
+      "learning_rate": 1.995e-05,
+      "loss": 0.018,
+      "step": 400
+    },
+    {
+      "epoch": 28.642857142857142,
+      "grad_norm": 0.33984375,
+      "learning_rate": 2e-05,
+      "loss": 0.0189,
+      "step": 401
+    },
+    {
+      "epoch": 28.714285714285715,
+      "grad_norm": 0.365234375,
+      "learning_rate": 2.0050000000000003e-05,
+      "loss": 0.0171,
+      "step": 402
+    },
+    {
+      "epoch": 28.785714285714285,
+      "grad_norm": 0.28125,
+      "learning_rate": 2.01e-05,
+      "loss": 0.0154,
+      "step": 403
+    },
+    {
+      "epoch": 28.857142857142858,
+      "grad_norm": 0.1982421875,
+      "learning_rate": 2.0150000000000002e-05,
+      "loss": 0.0149,
+      "step": 404
+    },
+    {
+      "epoch": 28.928571428571427,
+      "grad_norm": 0.21875,
+      "learning_rate": 2.0200000000000003e-05,
+      "loss": 0.0187,
+      "step": 405
+    },
+    {
+      "epoch": 29.0,
+      "grad_norm": 0.5859375,
+      "learning_rate": 2.025e-05,
+      "loss": 0.0193,
+      "step": 406
+    },
+    {
+      "epoch": 29.071428571428573,
+      "grad_norm": 0.2138671875,
+      "learning_rate": 2.0300000000000002e-05,
+      "loss": 0.0199,
+      "step": 407
+    },
+    {
+      "epoch": 29.142857142857142,
+      "grad_norm": 0.1689453125,
+      "learning_rate": 2.035e-05,
+      "loss": 0.0132,
+      "step": 408
+    },
+    {
+      "epoch": 29.214285714285715,
+      "grad_norm": 0.224609375,
+      "learning_rate": 2.04e-05,
+      "loss": 0.017,
+      "step": 409
+    },
+    {
+      "epoch": 29.285714285714285,
+      "grad_norm": 0.216796875,
+      "learning_rate": 2.045e-05,
+      "loss": 0.0165,
+      "step": 410
+    },
+    {
+      "epoch": 29.357142857142858,
+      "grad_norm": 0.2197265625,
+      "learning_rate": 2.05e-05,
+      "loss": 0.0163,
+      "step": 411
+    },
+    {
+      "epoch": 29.428571428571427,
+      "grad_norm": 0.34375,
+      "learning_rate": 2.055e-05,
+      "loss": 0.0163,
+      "step": 412
+    },
+    {
+      "epoch": 29.5,
+      "grad_norm": 0.25,
+      "learning_rate": 2.06e-05,
+      "loss": 0.0173,
+      "step": 413
+    },
+    {
+      "epoch": 29.571428571428573,
+      "grad_norm": 0.3125,
+      "learning_rate": 2.065e-05,
+      "loss": 0.0137,
+      "step": 414
+    },
+    {
+      "epoch": 29.642857142857142,
+      "grad_norm": 0.22265625,
+      "learning_rate": 2.07e-05,
+      "loss": 0.0152,
+      "step": 415
+    },
+    {
+      "epoch": 29.714285714285715,
+      "grad_norm": 0.38671875,
+      "learning_rate": 2.075e-05,
+      "loss": 0.0171,
+      "step": 416
+    },
+    {
+      "epoch": 29.785714285714285,
+      "grad_norm": 0.31640625,
+      "learning_rate": 2.08e-05,
+      "loss": 0.0168,
+      "step": 417
+    },
+    {
+      "epoch": 29.857142857142858,
+      "grad_norm": 0.3828125,
+      "learning_rate": 2.085e-05,
+      "loss": 0.0165,
+      "step": 418
+    },
+    {
+      "epoch": 29.928571428571427,
+      "grad_norm": 0.33984375,
+      "learning_rate": 2.09e-05,
+      "loss": 0.0167,
+      "step": 419
+    },
+    {
+      "epoch": 30.0,
+      "grad_norm": 0.55078125,
+      "learning_rate": 2.095e-05,
+      "loss": 0.0184,
+      "step": 420
+    },
+    {
+      "epoch": 30.071428571428573,
+      "grad_norm": 0.61328125,
+      "learning_rate": 2.1e-05,
+      "loss": 0.0164,
+      "step": 421
+    },
+    {
+      "epoch": 30.142857142857142,
+      "grad_norm": 0.494140625,
+      "learning_rate": 2.105e-05,
+      "loss": 0.0169,
+      "step": 422
+    },
+    {
+      "epoch": 30.214285714285715,
+      "grad_norm": 0.28125,
+      "learning_rate": 2.11e-05,
+      "loss": 0.0157,
+      "step": 423
+    },
+    {
+      "epoch": 30.285714285714285,
+      "grad_norm": 0.3515625,
+      "learning_rate": 2.115e-05,
+      "loss": 0.016,
+      "step": 424
+    },
+    {
+      "epoch": 30.357142857142858,
+      "grad_norm": 0.419921875,
+      "learning_rate": 2.12e-05,
+      "loss": 0.0178,
+      "step": 425
+    },
+    {
+      "epoch": 30.428571428571427,
+      "grad_norm": 0.30078125,
+      "learning_rate": 2.125e-05,
+      "loss": 0.0146,
+      "step": 426
+    },
+    {
+      "epoch": 30.5,
+      "grad_norm": 0.380859375,
+      "learning_rate": 2.13e-05,
+      "loss": 0.0153,
+      "step": 427
+    },
+    {
+      "epoch": 30.571428571428573,
+      "grad_norm": 0.203125,
+      "learning_rate": 2.135e-05,
+      "loss": 0.0161,
+      "step": 428
+    },
+    {
+      "epoch": 30.642857142857142,
+      "grad_norm": 0.1806640625,
+      "learning_rate": 2.1400000000000002e-05,
+      "loss": 0.0163,
+      "step": 429
+    },
+    {
+      "epoch": 30.714285714285715,
+      "grad_norm": 0.27734375,
+      "learning_rate": 2.145e-05,
+      "loss": 0.0166,
+      "step": 430
+    },
+    {
+      "epoch": 30.785714285714285,
+      "grad_norm": 0.328125,
+      "learning_rate": 2.15e-05,
+      "loss": 0.0163,
+      "step": 431
+    },
+    {
+      "epoch": 30.857142857142858,
+      "grad_norm": 0.26953125,
+      "learning_rate": 2.1550000000000002e-05,
+      "loss": 0.016,
+      "step": 432
+    },
+    {
+      "epoch": 30.928571428571427,
+      "grad_norm": 0.2001953125,
+      "learning_rate": 2.16e-05,
+      "loss": 0.0164,
+      "step": 433
+    },
+    {
+      "epoch": 31.0,
+      "grad_norm": 0.2001953125,
+      "learning_rate": 2.165e-05,
+      "loss": 0.0149,
+      "step": 434
+    },
+    {
+      "epoch": 31.071428571428573,
+      "grad_norm": 0.1875,
+      "learning_rate": 2.1700000000000002e-05,
+      "loss": 0.0142,
+      "step": 435
+    },
+    {
+      "epoch": 31.142857142857142,
+      "grad_norm": 0.2734375,
+      "learning_rate": 2.175e-05,
+      "loss": 0.0183,
+      "step": 436
+    },
+    {
+      "epoch": 31.214285714285715,
+      "grad_norm": 0.25390625,
+      "learning_rate": 2.18e-05,
+      "loss": 0.0128,
+      "step": 437
+    },
+    {
+      "epoch": 31.285714285714285,
+      "grad_norm": 0.2138671875,
+      "learning_rate": 2.1850000000000003e-05,
+      "loss": 0.0157,
+      "step": 438
+    },
+    {
+      "epoch": 31.357142857142858,
+      "grad_norm": 0.251953125,
+      "learning_rate": 2.19e-05,
+      "loss": 0.0149,
+      "step": 439
+    },
+    {
+      "epoch": 31.428571428571427,
+      "grad_norm": 0.365234375,
+      "learning_rate": 2.195e-05,
+      "loss": 0.0153,
+      "step": 440
+    },
+    {
+      "epoch": 31.5,
+      "grad_norm": 0.1884765625,
+      "learning_rate": 2.2000000000000003e-05,
+      "loss": 0.0159,
+      "step": 441
+    },
+    {
+      "epoch": 31.571428571428573,
+      "grad_norm": 0.455078125,
+      "learning_rate": 2.205e-05,
+      "loss": 0.019,
+      "step": 442
+    },
+    {
+      "epoch": 31.642857142857142,
+      "grad_norm": 0.291015625,
+      "learning_rate": 2.2100000000000002e-05,
+      "loss": 0.0179,
+      "step": 443
+    },
+    {
+      "epoch": 31.714285714285715,
+      "grad_norm": 0.38671875,
+      "learning_rate": 2.215e-05,
+      "loss": 0.017,
+      "step": 444
+    },
+    {
+      "epoch": 31.785714285714285,
+      "grad_norm": 0.419921875,
+      "learning_rate": 2.22e-05,
+      "loss": 0.0151,
+      "step": 445
+    },
+    {
+      "epoch": 31.857142857142858,
+      "grad_norm": 0.1962890625,
+      "learning_rate": 2.2250000000000002e-05,
+      "loss": 0.0161,
+      "step": 446
+    },
+    {
+      "epoch": 31.928571428571427,
+      "grad_norm": 0.41015625,
+      "learning_rate": 2.23e-05,
+      "loss": 0.0175,
+      "step": 447
+    },
+    {
+      "epoch": 32.0,
+      "grad_norm": 0.61328125,
+      "learning_rate": 2.235e-05,
+      "loss": 0.0186,
+      "step": 448
+    },
+    {
+      "epoch": 32.07142857142857,
+      "grad_norm": 0.193359375,
+      "learning_rate": 2.2400000000000002e-05,
+      "loss": 0.0162,
+      "step": 449
+    },
+    {
+      "epoch": 32.142857142857146,
+      "grad_norm": 0.203125,
+      "learning_rate": 2.245e-05,
+      "loss": 0.0161,
+      "step": 450
+    },
+    {
+      "epoch": 32.214285714285715,
+      "grad_norm": 0.36328125,
+      "learning_rate": 2.25e-05,
+      "loss": 0.015,
+      "step": 451
+    },
+    {
+      "epoch": 32.285714285714285,
+      "grad_norm": 0.2119140625,
+      "learning_rate": 2.2550000000000003e-05,
+      "loss": 0.016,
+      "step": 452
+    },
+    {
+      "epoch": 32.357142857142854,
+      "grad_norm": 0.25390625,
+      "learning_rate": 2.26e-05,
+      "loss": 0.0152,
+      "step": 453
+    },
+    {
+      "epoch": 32.42857142857143,
+      "grad_norm": 0.2158203125,
+      "learning_rate": 2.265e-05,
+      "loss": 0.0189,
+      "step": 454
+    },
+    {
+      "epoch": 32.5,
+      "grad_norm": 0.1533203125,
+      "learning_rate": 2.2700000000000003e-05,
+      "loss": 0.0141,
+      "step": 455
+    },
+    {
+      "epoch": 32.57142857142857,
+      "grad_norm": 0.294921875,
+      "learning_rate": 2.275e-05,
+      "loss": 0.0169,
+      "step": 456
+    },
+    {
+      "epoch": 32.642857142857146,
+      "grad_norm": 0.314453125,
+      "learning_rate": 2.2800000000000002e-05,
+      "loss": 0.0134,
+      "step": 457
+    },
+    {
+      "epoch": 32.714285714285715,
+      "grad_norm": 0.2021484375,
+      "learning_rate": 2.2850000000000003e-05,
+      "loss": 0.014,
+      "step": 458
+    },
+    {
+      "epoch": 32.785714285714285,
+      "grad_norm": 0.703125,
+      "learning_rate": 2.29e-05,
+      "loss": 0.0168,
+      "step": 459
+    },
+    {
+      "epoch": 32.857142857142854,
+      "grad_norm": 0.255859375,
+      "learning_rate": 2.2950000000000002e-05,
+      "loss": 0.0153,
+      "step": 460
+    },
+    {
+      "epoch": 32.92857142857143,
+      "grad_norm": 0.4765625,
+      "learning_rate": 2.3000000000000003e-05,
+      "loss": 0.0148,
+      "step": 461
+    },
+    {
+      "epoch": 33.0,
+      "grad_norm": 0.23046875,
+      "learning_rate": 2.305e-05,
+      "loss": 0.0169,
+      "step": 462
+    },
+    {
+      "epoch": 33.07142857142857,
+      "grad_norm": 0.31640625,
+      "learning_rate": 2.3100000000000002e-05,
+      "loss": 0.0165,
+      "step": 463
+    },
+    {
+      "epoch": 33.142857142857146,
+      "grad_norm": 0.306640625,
+      "learning_rate": 2.3150000000000004e-05,
+      "loss": 0.02,
+      "step": 464
+    },
+    {
+      "epoch": 33.214285714285715,
+      "grad_norm": 0.6171875,
+      "learning_rate": 2.32e-05,
+      "loss": 0.015,
+      "step": 465
+    },
+    {
+      "epoch": 33.285714285714285,
+      "grad_norm": 0.2421875,
+      "learning_rate": 2.3250000000000003e-05,
+      "loss": 0.0179,
+      "step": 466
+    },
+    {
+      "epoch": 33.357142857142854,
+      "grad_norm": 0.28515625,
+      "learning_rate": 2.3300000000000004e-05,
+      "loss": 0.0136,
+      "step": 467
+    },
+    {
+      "epoch": 33.42857142857143,
+      "grad_norm": 0.3671875,
+      "learning_rate": 2.3350000000000002e-05,
+      "loss": 0.0162,
+      "step": 468
+    },
+    {
+      "epoch": 33.5,
+      "grad_norm": 0.98046875,
+      "learning_rate": 2.3400000000000003e-05,
+      "loss": 0.0208,
+      "step": 469
+    },
+    {
+      "epoch": 33.57142857142857,
+      "grad_norm": 0.435546875,
+      "learning_rate": 2.345e-05,
+      "loss": 0.0144,
+      "step": 470
+    },
+    {
+      "epoch": 33.642857142857146,
+      "grad_norm": 0.205078125,
+      "learning_rate": 2.35e-05,
+      "loss": 0.0163,
+      "step": 471
+    },
+    {
+      "epoch": 33.714285714285715,
+      "grad_norm": 0.208984375,
+      "learning_rate": 2.355e-05,
+      "loss": 0.0156,
+      "step": 472
+    },
+    {
+      "epoch": 33.785714285714285,
+      "grad_norm": 0.2119140625,
+      "learning_rate": 2.36e-05,
+      "loss": 0.0166,
+      "step": 473
+    },
+    {
+      "epoch": 33.857142857142854,
+      "grad_norm": 0.298828125,
+      "learning_rate": 2.365e-05,
+      "loss": 0.015,
+      "step": 474
+    },
+    {
+      "epoch": 33.92857142857143,
+      "grad_norm": 0.34765625,
+      "learning_rate": 2.37e-05,
+      "loss": 0.0172,
+      "step": 475
+    },
+    {
+      "epoch": 34.0,
+      "grad_norm": 0.21875,
+      "learning_rate": 2.375e-05,
+      "loss": 0.0167,
+      "step": 476
+    },
+    {
+      "epoch": 34.07142857142857,
+      "grad_norm": 0.208984375,
+      "learning_rate": 2.38e-05,
+      "loss": 0.0161,
+      "step": 477
+    },
+    {
+      "epoch": 34.142857142857146,
+      "grad_norm": 0.625,
+      "learning_rate": 2.385e-05,
+      "loss": 0.0158,
+      "step": 478
+    },
+    {
+      "epoch": 34.214285714285715,
+      "grad_norm": 0.23828125,
+      "learning_rate": 2.39e-05,
+      "loss": 0.0153,
+      "step": 479
+    },
+    {
+      "epoch": 34.285714285714285,
+      "grad_norm": 0.412109375,
+      "learning_rate": 2.395e-05,
+      "loss": 0.0184,
+      "step": 480
+    },
+    {
+      "epoch": 34.357142857142854,
+      "grad_norm": 0.376953125,
+      "learning_rate": 2.4e-05,
+      "loss": 0.0157,
+      "step": 481
+    },
+    {
+      "epoch": 34.42857142857143,
+      "grad_norm": 0.26171875,
+      "learning_rate": 2.4050000000000002e-05,
+      "loss": 0.0147,
+      "step": 482
+    },
+    {
+      "epoch": 34.5,
+      "grad_norm": 0.439453125,
+      "learning_rate": 2.41e-05,
+      "loss": 0.0133,
+      "step": 483
+    },
+    {
+      "epoch": 34.57142857142857,
+      "grad_norm": 0.26171875,
+      "learning_rate": 2.415e-05,
+      "loss": 0.0147,
+      "step": 484
+    },
+    {
+      "epoch": 34.642857142857146,
+      "grad_norm": 0.275390625,
+      "learning_rate": 2.4200000000000002e-05,
+      "loss": 0.0151,
+      "step": 485
+    },
+    {
+      "epoch": 34.714285714285715,
+      "grad_norm": 0.2255859375,
+      "learning_rate": 2.425e-05,
+      "loss": 0.0175,
+      "step": 486
+    },
+    {
+      "epoch": 34.785714285714285,
+      "grad_norm": 0.345703125,
+      "learning_rate": 2.43e-05,
+      "loss": 0.0162,
+      "step": 487
+    },
+    {
+      "epoch": 34.857142857142854,
+      "grad_norm": 0.384765625,
+      "learning_rate": 2.435e-05,
+      "loss": 0.0139,
+      "step": 488
+    },
+    {
+      "epoch": 34.92857142857143,
+      "grad_norm": 0.1953125,
+      "learning_rate": 2.44e-05,
+      "loss": 0.0158,
+      "step": 489
+    },
+    {
+      "epoch": 35.0,
+      "grad_norm": 0.375,
+      "learning_rate": 2.445e-05,
+      "loss": 0.0163,
+      "step": 490
+    },
+    {
+      "epoch": 35.07142857142857,
+      "grad_norm": 0.197265625,
+      "learning_rate": 2.45e-05,
+      "loss": 0.0135,
+      "step": 491
+    },
+    {
+      "epoch": 35.142857142857146,
+      "grad_norm": 0.1982421875,
+      "learning_rate": 2.455e-05,
+      "loss": 0.0163,
+      "step": 492
+    },
+    {
+      "epoch": 35.214285714285715,
+      "grad_norm": 0.6484375,
+      "learning_rate": 2.46e-05,
+      "loss": 0.0155,
+      "step": 493
+    },
+    {
+      "epoch": 35.285714285714285,
+      "grad_norm": 0.232421875,
+      "learning_rate": 2.465e-05,
+      "loss": 0.0153,
+      "step": 494
+    },
+    {
+      "epoch": 35.357142857142854,
+      "grad_norm": 0.353515625,
+      "learning_rate": 2.47e-05,
+      "loss": 0.0161,
+      "step": 495
+    },
+    {
+      "epoch": 35.42857142857143,
+      "grad_norm": 0.30859375,
+      "learning_rate": 2.4750000000000002e-05,
+      "loss": 0.0151,
+      "step": 496
+    },
+    {
+      "epoch": 35.5,
+      "grad_norm": 0.3671875,
+      "learning_rate": 2.48e-05,
+      "loss": 0.0182,
+      "step": 497
+    },
+    {
+      "epoch": 35.57142857142857,
+      "grad_norm": 0.39453125,
+      "learning_rate": 2.485e-05,
+      "loss": 0.0146,
+      "step": 498
+    },
+    {
+      "epoch": 35.642857142857146,
+      "grad_norm": 0.412109375,
+      "learning_rate": 2.4900000000000002e-05,
+      "loss": 0.0161,
+      "step": 499
+    },
+    {
+      "epoch": 35.714285714285715,
+      "grad_norm": 0.1669921875,
+      "learning_rate": 2.495e-05,
+      "loss": 0.0173,
+      "step": 500
+    },
+    {
+      "epoch": 35.785714285714285,
+      "grad_norm": 0.421875,
+      "learning_rate": 2.5e-05,
+      "loss": 0.014,
+      "step": 501
+    },
+    {
+      "epoch": 35.857142857142854,
+      "grad_norm": 0.478515625,
+      "learning_rate": 2.4999996953825926e-05,
+      "loss": 0.0166,
+      "step": 502
+    },
+    {
+      "epoch": 35.92857142857143,
+      "grad_norm": 0.201171875,
+      "learning_rate": 2.499998781530519e-05,
+      "loss": 0.0139,
+      "step": 503
+    },
+    {
+      "epoch": 36.0,
+      "grad_norm": 0.1533203125,
+      "learning_rate": 2.4999972584442245e-05,
+      "loss": 0.0118,
+      "step": 504
+    },
+    {
+      "epoch": 36.07142857142857,
+      "grad_norm": 0.1982421875,
+      "learning_rate": 2.4999951261244508e-05,
+      "loss": 0.0147,
+      "step": 505
+    },
+    {
+      "epoch": 36.142857142857146,
+      "grad_norm": 0.314453125,
+      "learning_rate": 2.499992384572238e-05,
+      "loss": 0.0151,
+      "step": 506
+    },
+    {
+      "epoch": 36.214285714285715,
+      "grad_norm": 0.1904296875,
+      "learning_rate": 2.499989033788922e-05,
+      "loss": 0.0161,
+      "step": 507
+    },
+    {
+      "epoch": 36.285714285714285,
+      "grad_norm": 0.189453125,
+      "learning_rate": 2.499985073776136e-05,
+      "loss": 0.0172,
+      "step": 508
+    },
+    {
+      "epoch": 36.357142857142854,
+      "grad_norm": 0.32421875,
+      "learning_rate": 2.4999805045358104e-05,
+      "loss": 0.0162,
+      "step": 509
+    },
+    {
+      "epoch": 36.42857142857143,
+      "grad_norm": 0.265625,
+      "learning_rate": 2.4999753260701714e-05,
+      "loss": 0.0176,
+      "step": 510
+    },
+    {
+      "epoch": 36.5,
+      "grad_norm": 0.173828125,
+      "learning_rate": 2.4999695383817435e-05,
+      "loss": 0.016,
+      "step": 511
+    },
+    {
+      "epoch": 36.57142857142857,
+      "grad_norm": 0.208984375,
+      "learning_rate": 2.4999631414733476e-05,
+      "loss": 0.0156,
+      "step": 512
+    },
+    {
+      "epoch": 36.642857142857146,
+      "grad_norm": 0.390625,
+      "learning_rate": 2.499956135348101e-05,
+      "loss": 0.0145,
+      "step": 513
+    },
+    {
+      "epoch": 36.714285714285715,
+      "grad_norm": 0.298828125,
+      "learning_rate": 2.4999485200094186e-05,
+      "loss": 0.0167,
+      "step": 514
+    },
+    {
+      "epoch": 36.785714285714285,
+      "grad_norm": 0.4296875,
+      "learning_rate": 2.4999402954610125e-05,
+      "loss": 0.0165,
+      "step": 515
+    },
+    {
+      "epoch": 36.857142857142854,
+      "grad_norm": 0.228515625,
+      "learning_rate": 2.4999314617068904e-05,
+      "loss": 0.0141,
+      "step": 516
+    },
+    {
+      "epoch": 36.92857142857143,
+      "grad_norm": 0.291015625,
+      "learning_rate": 2.4999220187513583e-05,
+      "loss": 0.016,
+      "step": 517
+    },
+    {
+      "epoch": 37.0,
+      "grad_norm": 0.55078125,
+      "learning_rate": 2.4999119665990183e-05,
+      "loss": 0.0159,
+      "step": 518
+    },
+    {
+      "epoch": 37.07142857142857,
+      "grad_norm": 0.2421875,
+      "learning_rate": 2.4999013052547703e-05,
+      "loss": 0.018,
+      "step": 519
+    },
+    {
+      "epoch": 37.142857142857146,
+      "grad_norm": 0.2470703125,
+      "learning_rate": 2.4998900347238096e-05,
+      "loss": 0.0143,
+      "step": 520
+    },
+    {
+      "epoch": 37.214285714285715,
+      "grad_norm": 0.388671875,
+      "learning_rate": 2.4998781550116305e-05,
+      "loss": 0.0155,
+      "step": 521
+    },
+    {
+      "epoch": 37.285714285714285,
+      "grad_norm": 0.1748046875,
+      "learning_rate": 2.499865666124022e-05,
+      "loss": 0.016,
+      "step": 522
+    },
+    {
+      "epoch": 37.357142857142854,
+      "grad_norm": 0.271484375,
+      "learning_rate": 2.4998525680670712e-05,
+      "loss": 0.0135,
+      "step": 523
+    },
+    {
+      "epoch": 37.42857142857143,
+      "grad_norm": 0.216796875,
+      "learning_rate": 2.4998388608471625e-05,
+      "loss": 0.0136,
+      "step": 524
+    },
+    {
+      "epoch": 37.5,
+      "grad_norm": 0.5,
+      "learning_rate": 2.4998245444709757e-05,
+      "loss": 0.016,
+      "step": 525
+    },
+    {
+      "epoch": 37.57142857142857,
+      "grad_norm": 0.1318359375,
+      "learning_rate": 2.4998096189454893e-05,
+      "loss": 0.0138,
+      "step": 526
+    },
+    {
+      "epoch": 37.642857142857146,
+      "grad_norm": 0.1904296875,
+      "learning_rate": 2.499794084277977e-05,
+      "loss": 0.0125,
+      "step": 527
+    },
+    {
+      "epoch": 37.714285714285715,
+      "grad_norm": 0.392578125,
+      "learning_rate": 2.4997779404760114e-05,
+      "loss": 0.0161,
+      "step": 528
+    },
+    {
+      "epoch": 37.785714285714285,
+      "grad_norm": 0.1796875,
+      "learning_rate": 2.4997611875474595e-05,
+      "loss": 0.0144,
+      "step": 529
+    },
+    {
+      "epoch": 37.857142857142854,
+      "grad_norm": 0.298828125,
+      "learning_rate": 2.4997438255004876e-05,
+      "loss": 0.0132,
+      "step": 530
+    },
+    {
+      "epoch": 37.92857142857143,
+      "grad_norm": 0.55078125,
+      "learning_rate": 2.499725854343557e-05,
+      "loss": 0.0134,
+      "step": 531
+    },
+    {
+      "epoch": 38.0,
+      "grad_norm": 0.2177734375,
+      "learning_rate": 2.4997072740854265e-05,
+      "loss": 0.0162,
+      "step": 532
+    },
+    {
+      "epoch": 38.07142857142857,
+      "grad_norm": 0.283203125,
+      "learning_rate": 2.499688084735153e-05,
+      "loss": 0.0132,
+      "step": 533
+    },
+    {
+      "epoch": 38.142857142857146,
+      "grad_norm": 0.45703125,
+      "learning_rate": 2.4996682863020875e-05,
+      "loss": 0.0201,
+      "step": 534
+    },
+    {
+      "epoch": 38.214285714285715,
+      "grad_norm": 0.30078125,
+      "learning_rate": 2.499647878795881e-05,
+      "loss": 0.016,
+      "step": 535
+    },
+    {
+      "epoch": 38.285714285714285,
+      "grad_norm": 0.2734375,
+      "learning_rate": 2.499626862226479e-05,
+      "loss": 0.0157,
+      "step": 536
+    },
+    {
+      "epoch": 38.357142857142854,
+      "grad_norm": 0.53515625,
+      "learning_rate": 2.499605236604125e-05,
+      "loss": 0.0137,
+      "step": 537
+    },
+    {
+      "epoch": 38.42857142857143,
+      "grad_norm": 0.275390625,
+      "learning_rate": 2.499583001939359e-05,
+      "loss": 0.018,
+      "step": 538
+    },
+    {
+      "epoch": 38.5,
+      "grad_norm": 0.224609375,
+      "learning_rate": 2.4995601582430183e-05,
+      "loss": 0.0123,
+      "step": 539
+    },
+    {
+      "epoch": 38.57142857142857,
+      "grad_norm": 0.19921875,
+      "learning_rate": 2.499536705526236e-05,
+      "loss": 0.0115,
+      "step": 540
+    },
+    {
+      "epoch": 38.642857142857146,
+      "grad_norm": 0.33984375,
+      "learning_rate": 2.499512643800443e-05,
+      "loss": 0.0151,
+      "step": 541
+    },
+    {
+      "epoch": 38.714285714285715,
+      "grad_norm": 0.373046875,
+      "learning_rate": 2.499487973077367e-05,
+      "loss": 0.0155,
+      "step": 542
+    },
+    {
+      "epoch": 38.785714285714285,
+      "grad_norm": 0.38671875,
+      "learning_rate": 2.4994626933690314e-05,
+      "loss": 0.0147,
+      "step": 543
+    },
+    {
+      "epoch": 38.857142857142854,
+      "grad_norm": 0.2431640625,
+      "learning_rate": 2.499436804687758e-05,
+      "loss": 0.0184,
+      "step": 544
+    },
+    {
+      "epoch": 38.92857142857143,
+      "grad_norm": 0.1416015625,
+      "learning_rate": 2.499410307046164e-05,
+      "loss": 0.0143,
+      "step": 545
+    },
+    {
+      "epoch": 39.0,
+      "grad_norm": 0.19921875,
+      "learning_rate": 2.4993832004571646e-05,
+      "loss": 0.0174,
+      "step": 546
+    },
+    {
+      "epoch": 39.07142857142857,
+      "grad_norm": 0.1748046875,
+      "learning_rate": 2.499355484933971e-05,
+      "loss": 0.0149,
+      "step": 547
+    },
+    {
+      "epoch": 39.142857142857146,
+      "grad_norm": 0.25390625,
+      "learning_rate": 2.4993271604900913e-05,
+      "loss": 0.016,
+      "step": 548
+    },
+    {
+      "epoch": 39.214285714285715,
+      "grad_norm": 0.412109375,
+      "learning_rate": 2.4992982271393307e-05,
+      "loss": 0.0184,
+      "step": 549
+    },
+    {
+      "epoch": 39.285714285714285,
+      "grad_norm": 0.216796875,
+      "learning_rate": 2.4992686848957905e-05,
+      "loss": 0.0149,
+      "step": 550
+    },
+    {
+      "epoch": 39.357142857142854,
+      "grad_norm": 0.1875,
+      "learning_rate": 2.49923853377387e-05,
+      "loss": 0.0156,
+      "step": 551
+    },
+    {
+      "epoch": 39.42857142857143,
+      "grad_norm": 0.2080078125,
+      "learning_rate": 2.4992077737882636e-05,
+      "loss": 0.0147,
+      "step": 552
+    },
+    {
+      "epoch": 39.5,
+      "grad_norm": 0.330078125,
+      "learning_rate": 2.499176404953964e-05,
+      "loss": 0.017,
+      "step": 553
+    },
+    {
+      "epoch": 39.57142857142857,
+      "grad_norm": 0.2158203125,
+      "learning_rate": 2.49914442728626e-05,
+      "loss": 0.0163,
+      "step": 554
+    },
+    {
+      "epoch": 39.642857142857146,
+      "grad_norm": 0.447265625,
+      "learning_rate": 2.4991118408007366e-05,
+      "loss": 0.0167,
+      "step": 555
+    },
+    {
+      "epoch": 39.714285714285715,
+      "grad_norm": 0.228515625,
+      "learning_rate": 2.4990786455132764e-05,
+      "loss": 0.0122,
+      "step": 556
+    },
+    {
+      "epoch": 39.785714285714285,
+      "grad_norm": 0.353515625,
+      "learning_rate": 2.4990448414400585e-05,
+      "loss": 0.016,
+      "step": 557
+    },
+    {
+      "epoch": 39.857142857142854,
+      "grad_norm": 0.287109375,
+      "learning_rate": 2.4990104285975586e-05,
+      "loss": 0.0145,
+      "step": 558
+    },
+    {
+      "epoch": 39.92857142857143,
+      "grad_norm": 0.259765625,
+      "learning_rate": 2.498975407002549e-05,
+      "loss": 0.0187,
+      "step": 559
+    },
+    {
+      "epoch": 40.0,
+      "grad_norm": 0.267578125,
+      "learning_rate": 2.4989397766720983e-05,
+      "loss": 0.0138,
+      "step": 560
+    },
+    {
+      "epoch": 40.07142857142857,
+      "grad_norm": 0.21484375,
+      "learning_rate": 2.498903537623573e-05,
+      "loss": 0.0142,
+      "step": 561
+    },
+    {
+      "epoch": 40.142857142857146,
+      "grad_norm": 0.478515625,
+      "learning_rate": 2.4988666898746356e-05,
+      "loss": 0.0177,
+      "step": 562
+    },
+    {
+      "epoch": 40.214285714285715,
+      "grad_norm": 0.19140625,
+      "learning_rate": 2.4988292334432444e-05,
+      "loss": 0.0162,
+      "step": 563
+    },
+    {
+      "epoch": 40.285714285714285,
+      "grad_norm": 0.58203125,
+      "learning_rate": 2.4987911683476562e-05,
+      "loss": 0.0174,
+      "step": 564
+    },
+    {
+      "epoch": 40.357142857142854,
+      "grad_norm": 0.197265625,
+      "learning_rate": 2.4987524946064228e-05,
+      "loss": 0.0158,
+      "step": 565
+    },
+    {
+      "epoch": 40.42857142857143,
+      "grad_norm": 0.2431640625,
+      "learning_rate": 2.4987132122383936e-05,
+      "loss": 0.0139,
+      "step": 566
+    },
+    {
+      "epoch": 40.5,
+      "grad_norm": 0.3828125,
+      "learning_rate": 2.4986733212627143e-05,
+      "loss": 0.0154,
+      "step": 567
+    },
+    {
+      "epoch": 40.57142857142857,
+      "grad_norm": 0.189453125,
+      "learning_rate": 2.498632821698827e-05,
+      "loss": 0.0157,
+      "step": 568
+    },
+    {
+      "epoch": 40.642857142857146,
+      "grad_norm": 0.51953125,
+      "learning_rate": 2.4985917135664714e-05,
+      "loss": 0.0153,
+      "step": 569
+    },
+    {
+      "epoch": 40.714285714285715,
+      "grad_norm": 0.19140625,
+      "learning_rate": 2.4985499968856824e-05,
+      "loss": 0.0163,
+      "step": 570
+    },
+    {
+      "epoch": 40.785714285714285,
+      "grad_norm": 0.31640625,
+      "learning_rate": 2.4985076716767927e-05,
+      "loss": 0.0169,
+      "step": 571
+    },
+    {
+      "epoch": 40.857142857142854,
+      "grad_norm": 0.341796875,
+      "learning_rate": 2.498464737960431e-05,
+      "loss": 0.0179,
+      "step": 572
+    },
+    {
+      "epoch": 40.92857142857143,
+      "grad_norm": 0.322265625,
+      "learning_rate": 2.498421195757522e-05,
+      "loss": 0.0143,
+      "step": 573
+    },
+    {
+      "epoch": 41.0,
+      "grad_norm": 0.50390625,
+      "learning_rate": 2.4983770450892884e-05,
+      "loss": 0.0184,
+      "step": 574
+    },
+    {
+      "epoch": 41.07142857142857,
+      "grad_norm": 0.2099609375,
+      "learning_rate": 2.4983322859772484e-05,
+      "loss": 0.0167,
+      "step": 575
+    },
+    {
+      "epoch": 41.142857142857146,
+      "grad_norm": 0.21484375,
+      "learning_rate": 2.4982869184432174e-05,
+      "loss": 0.0167,
+      "step": 576
+    },
+    {
+      "epoch": 41.214285714285715,
+      "grad_norm": 0.2041015625,
+      "learning_rate": 2.4982409425093067e-05,
+      "loss": 0.0152,
+      "step": 577
+    },
+    {
+      "epoch": 41.285714285714285,
+      "grad_norm": 0.1875,
+      "learning_rate": 2.498194358197924e-05,
+      "loss": 0.015,
+      "step": 578
+    },
+    {
+      "epoch": 41.357142857142854,
+      "grad_norm": 0.353515625,
+      "learning_rate": 2.498147165531775e-05,
+      "loss": 0.0153,
+      "step": 579
+    },
+    {
+      "epoch": 41.42857142857143,
+      "grad_norm": 0.4140625,
+      "learning_rate": 2.4980993645338595e-05,
+      "loss": 0.0138,
+      "step": 580
+    },
+    {
+      "epoch": 41.5,
+      "grad_norm": 0.345703125,
+      "learning_rate": 2.4980509552274765e-05,
+      "loss": 0.0142,
+      "step": 581
+    },
+    {
+      "epoch": 41.57142857142857,
+      "grad_norm": 0.38671875,
+      "learning_rate": 2.498001937636219e-05,
+      "loss": 0.0174,
+      "step": 582
+    },
+    {
+      "epoch": 41.642857142857146,
+      "grad_norm": 0.1630859375,
+      "learning_rate": 2.4979523117839783e-05,
+      "loss": 0.014,
+      "step": 583
+    },
+    {
+      "epoch": 41.714285714285715,
+      "grad_norm": 0.1962890625,
+      "learning_rate": 2.4979020776949405e-05,
+      "loss": 0.0163,
+      "step": 584
+    },
+    {
+      "epoch": 41.785714285714285,
+      "grad_norm": 0.1611328125,
+      "learning_rate": 2.4978512353935903e-05,
+      "loss": 0.018,
+      "step": 585
+    },
+    {
+      "epoch": 41.857142857142854,
+      "grad_norm": 0.1943359375,
+      "learning_rate": 2.497799784904707e-05,
+      "loss": 0.0146,
+      "step": 586
+    },
+    {
+      "epoch": 41.92857142857143,
+      "grad_norm": 0.20703125,
+      "learning_rate": 2.4977477262533667e-05,
+      "loss": 0.0148,
+      "step": 587
+    },
+    {
+      "epoch": 42.0,
+      "grad_norm": 0.2294921875,
+      "learning_rate": 2.4976950594649423e-05,
+      "loss": 0.015,
+      "step": 588
+    },
+    {
+      "epoch": 42.07142857142857,
+      "grad_norm": 0.41015625,
+      "learning_rate": 2.4976417845651033e-05,
+      "loss": 0.0153,
+      "step": 589
+    },
+    {
+      "epoch": 42.142857142857146,
+      "grad_norm": 0.255859375,
+      "learning_rate": 2.4975879015798152e-05,
+      "loss": 0.0167,
+      "step": 590
+    },
+    {
+      "epoch": 42.214285714285715,
+      "grad_norm": 0.1982421875,
+      "learning_rate": 2.4975334105353396e-05,
+      "loss": 0.0146,
+      "step": 591
+    },
+    {
+      "epoch": 42.285714285714285,
+      "grad_norm": 0.400390625,
+      "learning_rate": 2.4974783114582352e-05,
+      "loss": 0.0115,
+      "step": 592
+    },
+    {
+      "epoch": 42.357142857142854,
+      "grad_norm": 0.173828125,
+      "learning_rate": 2.4974226043753558e-05,
+      "loss": 0.0129,
+      "step": 593
+    },
+    {
+      "epoch": 42.42857142857143,
+      "grad_norm": 0.2197265625,
+      "learning_rate": 2.497366289313853e-05,
+      "loss": 0.0192,
+      "step": 594
+    },
+    {
+      "epoch": 42.5,
+      "grad_norm": 0.251953125,
+      "learning_rate": 2.4973093663011743e-05,
+      "loss": 0.0186,
+      "step": 595
+    },
+    {
+      "epoch": 42.57142857142857,
+      "grad_norm": 0.16015625,
+      "learning_rate": 2.4972518353650626e-05,
+      "loss": 0.015,
+      "step": 596
+    },
+    {
+      "epoch": 42.642857142857146,
+      "grad_norm": 0.162109375,
+      "learning_rate": 2.4971936965335585e-05,
+      "loss": 0.0147,
+      "step": 597
+    },
+    {
+      "epoch": 42.714285714285715,
+      "grad_norm": 0.1455078125,
+      "learning_rate": 2.497134949834997e-05,
+      "loss": 0.0154,
+      "step": 598
+    },
+    {
+      "epoch": 42.785714285714285,
+      "grad_norm": 0.306640625,
+      "learning_rate": 2.4970755952980125e-05,
+      "loss": 0.0172,
+      "step": 599
+    },
+    {
+      "epoch": 42.857142857142854,
+      "grad_norm": 0.146484375,
+      "learning_rate": 2.4970156329515315e-05,
+      "loss": 0.0156,
+      "step": 600
+    },
+    {
+      "epoch": 42.92857142857143,
+      "grad_norm": 0.271484375,
+      "learning_rate": 2.4969550628247805e-05,
+      "loss": 0.0146,
+      "step": 601
+    },
+    {
+      "epoch": 43.0,
+      "grad_norm": 0.333984375,
+      "learning_rate": 2.4968938849472797e-05,
+      "loss": 0.0151,
+      "step": 602
+    },
+    {
+      "epoch": 43.07142857142857,
+      "grad_norm": 0.353515625,
+      "learning_rate": 2.496832099348847e-05,
+      "loss": 0.0142,
+      "step": 603
+    },
+    {
+      "epoch": 43.142857142857146,
+      "grad_norm": 0.2197265625,
+      "learning_rate": 2.4967697060595958e-05,
+      "loss": 0.0144,
+      "step": 604
+    },
+    {
+      "epoch": 43.214285714285715,
+      "grad_norm": 0.205078125,
+      "learning_rate": 2.4967067051099356e-05,
+      "loss": 0.0175,
+      "step": 605
+    },
+    {
+      "epoch": 43.285714285714285,
+      "grad_norm": 0.26953125,
+      "learning_rate": 2.4966430965305727e-05,
+      "loss": 0.0142,
+      "step": 606
+    },
+    {
+      "epoch": 43.357142857142854,
+      "grad_norm": 0.2578125,
+      "learning_rate": 2.4965788803525088e-05,
+      "loss": 0.0169,
+      "step": 607
+    },
+    {
+      "epoch": 43.42857142857143,
+      "grad_norm": 0.2099609375,
+      "learning_rate": 2.4965140566070423e-05,
+      "loss": 0.0149,
+      "step": 608
+    },
+    {
+      "epoch": 43.5,
+      "grad_norm": 0.1591796875,
+      "learning_rate": 2.4964486253257674e-05,
+      "loss": 0.0147,
+      "step": 609
+    },
+    {
+      "epoch": 43.57142857142857,
+      "grad_norm": 0.205078125,
+      "learning_rate": 2.4963825865405747e-05,
+      "loss": 0.0136,
+      "step": 610
+    },
+    {
+      "epoch": 43.642857142857146,
+      "grad_norm": 0.185546875,
+      "learning_rate": 2.4963159402836506e-05,
+      "loss": 0.0148,
+      "step": 611
+    },
+    {
+      "epoch": 43.714285714285715,
+      "grad_norm": 0.14453125,
+      "learning_rate": 2.4962486865874774e-05,
+      "loss": 0.0148,
+      "step": 612
+    },
+    {
+      "epoch": 43.785714285714285,
+      "grad_norm": 0.1591796875,
+      "learning_rate": 2.4961808254848342e-05,
+      "loss": 0.0148,
+      "step": 613
+    },
+    {
+      "epoch": 43.857142857142854,
+      "grad_norm": 0.2109375,
+      "learning_rate": 2.496112357008795e-05,
+      "loss": 0.0166,
+      "step": 614
+    },
+    {
+      "epoch": 43.92857142857143,
+      "grad_norm": 0.35546875,
+      "learning_rate": 2.4960432811927308e-05,
+      "loss": 0.0121,
+      "step": 615
+    },
+    {
+      "epoch": 44.0,
+      "grad_norm": 0.228515625,
+      "learning_rate": 2.495973598070309e-05,
+      "loss": 0.0112,
+      "step": 616
+    },
+    {
+      "epoch": 44.07142857142857,
+      "grad_norm": 0.435546875,
+      "learning_rate": 2.4959033076754917e-05,
+      "loss": 0.0147,
+      "step": 617
+    },
+    {
+      "epoch": 44.142857142857146,
+      "grad_norm": 0.125,
+      "learning_rate": 2.495832410042537e-05,
+      "loss": 0.0153,
+      "step": 618
+    },
+    {
+      "epoch": 44.214285714285715,
+      "grad_norm": 0.37109375,
+      "learning_rate": 2.495760905206001e-05,
+      "loss": 0.0138,
+      "step": 619
+    },
+    {
+      "epoch": 44.285714285714285,
+      "grad_norm": 0.14453125,
+      "learning_rate": 2.4956887932007332e-05,
+      "loss": 0.0151,
+      "step": 620
+    },
+    {
+      "epoch": 44.357142857142854,
+      "grad_norm": 0.173828125,
+      "learning_rate": 2.4956160740618806e-05,
+      "loss": 0.0128,
+      "step": 621
+    },
+    {
+      "epoch": 44.42857142857143,
+      "grad_norm": 0.416015625,
+      "learning_rate": 2.4955427478248848e-05,
+      "loss": 0.0165,
+      "step": 622
+    },
+    {
+      "epoch": 44.5,
+      "grad_norm": 0.1591796875,
+      "learning_rate": 2.4954688145254855e-05,
+      "loss": 0.0144,
+      "step": 623
+    },
+    {
+      "epoch": 44.57142857142857,
+      "grad_norm": 0.328125,
+      "learning_rate": 2.495394274199716e-05,
+      "loss": 0.0175,
+      "step": 624
+    },
+    {
+      "epoch": 44.642857142857146,
+      "grad_norm": 0.1533203125,
+      "learning_rate": 2.495319126883906e-05,
+      "loss": 0.0149,
+      "step": 625
+    },
+    {
+      "epoch": 44.714285714285715,
+      "grad_norm": 0.275390625,
+      "learning_rate": 2.495243372614682e-05,
+      "loss": 0.0154,
+      "step": 626
+    },
+    {
+      "epoch": 44.785714285714285,
+      "grad_norm": 0.169921875,
+      "learning_rate": 2.495167011428966e-05,
+      "loss": 0.0151,
+      "step": 627
+    },
+    {
+      "epoch": 44.857142857142854,
+      "grad_norm": 0.333984375,
+      "learning_rate": 2.4950900433639746e-05,
+      "loss": 0.0159,
+      "step": 628
+    },
+    {
+      "epoch": 44.92857142857143,
+      "grad_norm": 0.208984375,
+      "learning_rate": 2.4950124684572217e-05,
+      "loss": 0.0157,
+      "step": 629
+    },
+    {
+      "epoch": 45.0,
+      "grad_norm": 0.1611328125,
+      "learning_rate": 2.494934286746516e-05,
+      "loss": 0.0136,
+      "step": 630
+    },
+    {
+      "epoch": 45.07142857142857,
+      "grad_norm": 0.2412109375,
+      "learning_rate": 2.494855498269963e-05,
+      "loss": 0.014,
+      "step": 631
+    },
+    {
+      "epoch": 45.142857142857146,
+      "grad_norm": 0.150390625,
+      "learning_rate": 2.4947761030659627e-05,
+      "loss": 0.0147,
+      "step": 632
+    },
+    {
+      "epoch": 45.214285714285715,
+      "grad_norm": 0.16015625,
+      "learning_rate": 2.4946961011732118e-05,
+      "loss": 0.0164,
+      "step": 633
+    },
+    {
+      "epoch": 45.285714285714285,
+      "grad_norm": 0.1455078125,
+      "learning_rate": 2.4946154926307014e-05,
+      "loss": 0.0167,
+      "step": 634
+    },
+    {
+      "epoch": 45.357142857142854,
+      "grad_norm": 0.29296875,
+      "learning_rate": 2.4945342774777196e-05,
+      "loss": 0.0124,
+      "step": 635
+    },
+    {
+      "epoch": 45.42857142857143,
+      "grad_norm": 0.2265625,
+      "learning_rate": 2.4944524557538503e-05,
+      "loss": 0.0131,
+      "step": 636
+    },
+    {
+      "epoch": 45.5,
+      "grad_norm": 0.482421875,
+      "learning_rate": 2.4943700274989714e-05,
+      "loss": 0.0144,
+      "step": 637
+    },
+    {
+      "epoch": 45.57142857142857,
+      "grad_norm": 0.236328125,
+      "learning_rate": 2.494286992753258e-05,
+      "loss": 0.0165,
+      "step": 638
+    },
+    {
+      "epoch": 45.642857142857146,
+      "grad_norm": 0.322265625,
+      "learning_rate": 2.49420335155718e-05,
+      "loss": 0.0134,
+      "step": 639
+    },
+    {
+      "epoch": 45.714285714285715,
+      "grad_norm": 0.439453125,
+      "learning_rate": 2.4941191039515032e-05,
+      "loss": 0.0139,
+      "step": 640
+    },
+    {
+      "epoch": 45.785714285714285,
+      "grad_norm": 0.337890625,
+      "learning_rate": 2.494034249977289e-05,
+      "loss": 0.0168,
+      "step": 641
+    },
+    {
+      "epoch": 45.857142857142854,
+      "grad_norm": 0.1787109375,
+      "learning_rate": 2.493948789675894e-05,
+      "loss": 0.0115,
+      "step": 642
+    },
+    {
+      "epoch": 45.92857142857143,
+      "grad_norm": 0.1474609375,
+      "learning_rate": 2.4938627230889705e-05,
+      "loss": 0.0146,
+      "step": 643
+    },
+    {
+      "epoch": 46.0,
+      "grad_norm": 0.314453125,
+      "learning_rate": 2.4937760502584666e-05,
+      "loss": 0.0159,
+      "step": 644
+    },
+    {
+      "epoch": 46.07142857142857,
+      "grad_norm": 0.298828125,
+      "learning_rate": 2.4936887712266254e-05,
+      "loss": 0.0176,
+      "step": 645
+    },
+    {
+      "epoch": 46.142857142857146,
+      "grad_norm": 0.447265625,
+      "learning_rate": 2.4936008860359854e-05,
+      "loss": 0.016,
+      "step": 646
+    },
+    {
+      "epoch": 46.214285714285715,
+      "grad_norm": 0.251953125,
+      "learning_rate": 2.493512394729381e-05,
+      "loss": 0.0168,
+      "step": 647
+    },
+    {
+      "epoch": 46.285714285714285,
+      "grad_norm": 0.1982421875,
+      "learning_rate": 2.4934232973499415e-05,
+      "loss": 0.0166,
+      "step": 648
+    },
+    {
+      "epoch": 46.357142857142854,
+      "grad_norm": 0.470703125,
+      "learning_rate": 2.493333593941093e-05,
+      "loss": 0.0138,
+      "step": 649
+    },
+    {
+      "epoch": 46.42857142857143,
+      "grad_norm": 0.546875,
+      "learning_rate": 2.4932432845465543e-05,
+      "loss": 0.013,
+      "step": 650
+    },
+    {
+      "epoch": 46.5,
+      "grad_norm": 0.2138671875,
+      "learning_rate": 2.4931523692103418e-05,
+      "loss": 0.0142,
+      "step": 651
+    },
+    {
+      "epoch": 46.57142857142857,
+      "grad_norm": 0.32421875,
+      "learning_rate": 2.4930608479767667e-05,
+      "loss": 0.0133,
+      "step": 652
+    },
+    {
+      "epoch": 46.642857142857146,
+      "grad_norm": 0.404296875,
+      "learning_rate": 2.492968720890435e-05,
+      "loss": 0.0154,
+      "step": 653
+    },
+    {
+      "epoch": 46.714285714285715,
+      "grad_norm": 0.5234375,
+      "learning_rate": 2.4928759879962487e-05,
+      "loss": 0.0172,
+      "step": 654
+    },
+    {
+      "epoch": 46.785714285714285,
+      "grad_norm": 0.2294921875,
+      "learning_rate": 2.4927826493394046e-05,
+      "loss": 0.0158,
+      "step": 655
+    },
+    {
+      "epoch": 46.857142857142854,
+      "grad_norm": 0.287109375,
+      "learning_rate": 2.4926887049653943e-05,
+      "loss": 0.013,
+      "step": 656
+    },
+    {
+      "epoch": 46.92857142857143,
+      "grad_norm": 0.193359375,
+      "learning_rate": 2.492594154920006e-05,
+      "loss": 0.0179,
+      "step": 657
+    },
+    {
+      "epoch": 47.0,
+      "grad_norm": 0.419921875,
+      "learning_rate": 2.4924989992493213e-05,
+      "loss": 0.0158,
+      "step": 658
+    },
+    {
+      "epoch": 47.07142857142857,
+      "grad_norm": 0.330078125,
+      "learning_rate": 2.4924032379997186e-05,
+      "loss": 0.0157,
+      "step": 659
+    },
+    {
+      "epoch": 47.142857142857146,
+      "grad_norm": 0.240234375,
+      "learning_rate": 2.492306871217871e-05,
+      "loss": 0.0153,
+      "step": 660
+    },
+    {
+      "epoch": 47.214285714285715,
+      "grad_norm": 0.134765625,
+      "learning_rate": 2.4922098989507454e-05,
+      "loss": 0.0145,
+      "step": 661
+    },
+    {
+      "epoch": 47.285714285714285,
+      "grad_norm": 0.197265625,
+      "learning_rate": 2.492112321245606e-05,
+      "loss": 0.0149,
+      "step": 662
+    },
+    {
+      "epoch": 47.357142857142854,
+      "grad_norm": 0.65234375,
+      "learning_rate": 2.4920141381500107e-05,
+      "loss": 0.0183,
+      "step": 663
+    },
+    {
+      "epoch": 47.42857142857143,
+      "grad_norm": 0.380859375,
+      "learning_rate": 2.4919153497118126e-05,
+      "loss": 0.0153,
+      "step": 664
+    },
+    {
+      "epoch": 47.5,
+      "grad_norm": 0.337890625,
+      "learning_rate": 2.4918159559791596e-05,
+      "loss": 0.0151,
+      "step": 665
+    },
+    {
+      "epoch": 47.57142857142857,
+      "grad_norm": 0.1923828125,
+      "learning_rate": 2.4917159570004954e-05,
+      "loss": 0.018,
+      "step": 666
+    },
+    {
+      "epoch": 47.642857142857146,
+      "grad_norm": 0.26953125,
+      "learning_rate": 2.491615352824559e-05,
+      "loss": 0.0171,
+      "step": 667
+    },
+    {
+      "epoch": 47.714285714285715,
+      "grad_norm": 0.5703125,
+      "learning_rate": 2.4915141435003823e-05,
+      "loss": 0.0147,
+      "step": 668
+    },
+    {
+      "epoch": 47.785714285714285,
+      "grad_norm": 0.265625,
+      "learning_rate": 2.4914123290772945e-05,
+      "loss": 0.0161,
+      "step": 669
+    },
+    {
+      "epoch": 47.857142857142854,
+      "grad_norm": 0.244140625,
+      "learning_rate": 2.491309909604918e-05,
+      "loss": 0.0145,
+      "step": 670
+    },
+    {
+      "epoch": 47.92857142857143,
+      "grad_norm": 0.33984375,
+      "learning_rate": 2.491206885133171e-05,
+      "loss": 0.0142,
+      "step": 671
+    },
+    {
+      "epoch": 48.0,
+      "grad_norm": 0.291015625,
+      "learning_rate": 2.491103255712267e-05,
+      "loss": 0.0172,
+      "step": 672
+    },
+    {
+      "epoch": 48.07142857142857,
+      "grad_norm": 0.2255859375,
+      "learning_rate": 2.4909990213927135e-05,
+      "loss": 0.0128,
+      "step": 673
+    },
+    {
+      "epoch": 48.142857142857146,
+      "grad_norm": 0.2578125,
+      "learning_rate": 2.4908941822253124e-05,
+      "loss": 0.0141,
+      "step": 674
+    },
+    {
+      "epoch": 48.214285714285715,
+      "grad_norm": 0.68359375,
+      "learning_rate": 2.4907887382611613e-05,
+      "loss": 0.0169,
+      "step": 675
+    },
+    {
+      "epoch": 48.285714285714285,
+      "grad_norm": 0.162109375,
+      "learning_rate": 2.4906826895516528e-05,
+      "loss": 0.0124,
+      "step": 676
+    },
+    {
+      "epoch": 48.357142857142854,
+      "grad_norm": 0.318359375,
+      "learning_rate": 2.4905760361484733e-05,
+      "loss": 0.015,
+      "step": 677
+    },
+    {
+      "epoch": 48.42857142857143,
+      "grad_norm": 0.1982421875,
+      "learning_rate": 2.490468778103604e-05,
+      "loss": 0.0151,
+      "step": 678
+    },
+    {
+      "epoch": 48.5,
+      "grad_norm": 0.248046875,
+      "learning_rate": 2.4903609154693222e-05,
+      "loss": 0.0176,
+      "step": 679
+    },
+    {
+      "epoch": 48.57142857142857,
+      "grad_norm": 0.5078125,
+      "learning_rate": 2.490252448298198e-05,
+      "loss": 0.0167,
+      "step": 680
+    },
+    {
+      "epoch": 48.642857142857146,
+      "grad_norm": 0.1875,
+      "learning_rate": 2.4901433766430975e-05,
+      "loss": 0.0169,
+      "step": 681
+    },
+    {
+      "epoch": 48.714285714285715,
+      "grad_norm": 0.41015625,
+      "learning_rate": 2.4900337005571807e-05,
+      "loss": 0.0147,
+      "step": 682
+    },
+    {
+      "epoch": 48.785714285714285,
+      "grad_norm": 0.400390625,
+      "learning_rate": 2.489923420093902e-05,
+      "loss": 0.0152,
+      "step": 683
+    },
+    {
+      "epoch": 48.857142857142854,
+      "grad_norm": 0.1513671875,
+      "learning_rate": 2.4898125353070112e-05,
+      "loss": 0.0163,
+      "step": 684
+    },
+    {
+      "epoch": 48.92857142857143,
+      "grad_norm": 0.2001953125,
+      "learning_rate": 2.4897010462505522e-05,
+      "loss": 0.0138,
+      "step": 685
+    },
+    {
+      "epoch": 49.0,
+      "grad_norm": 0.1865234375,
+      "learning_rate": 2.489588952978863e-05,
+      "loss": 0.0126,
+      "step": 686
+    },
+    {
+      "epoch": 49.07142857142857,
+      "grad_norm": 0.60546875,
+      "learning_rate": 2.4894762555465772e-05,
+      "loss": 0.0172,
+      "step": 687
+    },
+    {
+      "epoch": 49.142857142857146,
+      "grad_norm": 0.4140625,
+      "learning_rate": 2.4893629540086216e-05,
+      "loss": 0.0152,
+      "step": 688
+    },
+    {
+      "epoch": 49.214285714285715,
+      "grad_norm": 0.205078125,
+      "learning_rate": 2.489249048420218e-05,
+      "loss": 0.0152,
+      "step": 689
+    },
+    {
+      "epoch": 49.285714285714285,
+      "grad_norm": 0.2177734375,
+      "learning_rate": 2.4891345388368834e-05,
+      "loss": 0.0169,
+      "step": 690
+    },
+    {
+      "epoch": 49.357142857142854,
+      "grad_norm": 0.419921875,
+      "learning_rate": 2.489019425314427e-05,
+      "loss": 0.0143,
+      "step": 691
+    },
+    {
+      "epoch": 49.42857142857143,
+      "grad_norm": 0.21484375,
+      "learning_rate": 2.488903707908955e-05,
+      "loss": 0.0142,
+      "step": 692
+    },
+    {
+      "epoch": 49.5,
+      "grad_norm": 0.349609375,
+      "learning_rate": 2.488787386676866e-05,
+      "loss": 0.0145,
+      "step": 693
+    },
+    {
+      "epoch": 49.57142857142857,
+      "grad_norm": 0.232421875,
+      "learning_rate": 2.4886704616748536e-05,
+      "loss": 0.0148,
+      "step": 694
+    },
+    {
+      "epoch": 49.642857142857146,
+      "grad_norm": 0.21875,
+      "learning_rate": 2.488552932959906e-05,
+      "loss": 0.0133,
+      "step": 695
+    },
+    {
+      "epoch": 49.714285714285715,
+      "grad_norm": 0.2041015625,
+      "learning_rate": 2.4884348005893045e-05,
+      "loss": 0.0144,
+      "step": 696
+    },
+    {
+      "epoch": 49.785714285714285,
+      "grad_norm": 0.203125,
+      "learning_rate": 2.488316064620626e-05,
+      "loss": 0.0164,
+      "step": 697
+    },
+    {
+      "epoch": 49.857142857142854,
+      "grad_norm": 0.17578125,
+      "learning_rate": 2.488196725111741e-05,
+      "loss": 0.0157,
+      "step": 698
+    },
+    {
+      "epoch": 49.92857142857143,
+      "grad_norm": 0.33984375,
+      "learning_rate": 2.488076782120814e-05,
+      "loss": 0.0144,
+      "step": 699
+    },
+    {
+      "epoch": 50.0,
+      "grad_norm": 0.1416015625,
+      "learning_rate": 2.4879562357063037e-05,
+      "loss": 0.0127,
+      "step": 700
+    },
+    {
+      "epoch": 50.07142857142857,
+      "grad_norm": 0.2451171875,
+      "learning_rate": 2.487835085926963e-05,
+      "loss": 0.0172,
+      "step": 701
+    },
+    {
+      "epoch": 50.142857142857146,
+      "grad_norm": 0.259765625,
+      "learning_rate": 2.487713332841839e-05,
+      "loss": 0.0153,
+      "step": 702
+    },
+    {
+      "epoch": 50.214285714285715,
+      "grad_norm": 0.275390625,
+      "learning_rate": 2.487590976510272e-05,
+      "loss": 0.015,
+      "step": 703
+    },
+    {
+      "epoch": 50.285714285714285,
+      "grad_norm": 0.478515625,
+      "learning_rate": 2.4874680169918976e-05,
+      "loss": 0.0128,
+      "step": 704
+    },
+    {
+      "epoch": 50.357142857142854,
+      "grad_norm": 0.1630859375,
+      "learning_rate": 2.4873444543466448e-05,
+      "loss": 0.0134,
+      "step": 705
+    },
+    {
+      "epoch": 50.42857142857143,
+      "grad_norm": 0.181640625,
+      "learning_rate": 2.4872202886347362e-05,
+      "loss": 0.0142,
+      "step": 706
+    },
+    {
+      "epoch": 50.5,
+      "grad_norm": 0.1806640625,
+      "learning_rate": 2.487095519916689e-05,
+      "loss": 0.0179,
+      "step": 707
+    },
+    {
+      "epoch": 50.57142857142857,
+      "grad_norm": 0.326171875,
+      "learning_rate": 2.4869701482533138e-05,
+      "loss": 0.015,
+      "step": 708
+    },
+    {
+      "epoch": 50.642857142857146,
+      "grad_norm": 0.2255859375,
+      "learning_rate": 2.4868441737057153e-05,
+      "loss": 0.0165,
+      "step": 709
+    },
+    {
+      "epoch": 50.714285714285715,
+      "grad_norm": 0.49609375,
+      "learning_rate": 2.486717596335292e-05,
+      "loss": 0.0129,
+      "step": 710
+    },
+    {
+      "epoch": 50.785714285714285,
+      "grad_norm": 0.375,
+      "learning_rate": 2.4865904162037358e-05,
+      "loss": 0.0133,
+      "step": 711
+    },
+    {
+      "epoch": 50.857142857142854,
+      "grad_norm": 0.1904296875,
+      "learning_rate": 2.4864626333730328e-05,
+      "loss": 0.0142,
+      "step": 712
+    },
+    {
+      "epoch": 50.92857142857143,
+      "grad_norm": 0.427734375,
+      "learning_rate": 2.4863342479054633e-05,
+      "loss": 0.0151,
+      "step": 713
+    },
+    {
+      "epoch": 51.0,
+      "grad_norm": 0.291015625,
+      "learning_rate": 2.4862052598636005e-05,
+      "loss": 0.0137,
+      "step": 714
+    },
+    {
+      "epoch": 51.07142857142857,
+      "grad_norm": 0.388671875,
+      "learning_rate": 2.4860756693103114e-05,
+      "loss": 0.0167,
+      "step": 715
+    },
+    {
+      "epoch": 51.142857142857146,
+      "grad_norm": 0.1767578125,
+      "learning_rate": 2.4859454763087577e-05,
+      "loss": 0.0126,
+      "step": 716
+    },
+    {
+      "epoch": 51.214285714285715,
+      "grad_norm": 0.154296875,
+      "learning_rate": 2.4858146809223925e-05,
+      "loss": 0.015,
+      "step": 717
+    },
+    {
+      "epoch": 51.285714285714285,
+      "grad_norm": 0.2119140625,
+      "learning_rate": 2.485683283214965e-05,
+      "loss": 0.017,
+      "step": 718
+    },
+    {
+      "epoch": 51.357142857142854,
+      "grad_norm": 0.1474609375,
+      "learning_rate": 2.485551283250517e-05,
+      "loss": 0.0155,
+      "step": 719
+    },
+    {
+      "epoch": 51.42857142857143,
+      "grad_norm": 0.2275390625,
+      "learning_rate": 2.4854186810933828e-05,
+      "loss": 0.0141,
+      "step": 720
+    },
+    {
+      "epoch": 51.5,
+      "grad_norm": 0.25390625,
+      "learning_rate": 2.4852854768081912e-05,
+      "loss": 0.0161,
+      "step": 721
+    },
+    {
+      "epoch": 51.57142857142857,
+      "grad_norm": 0.306640625,
+      "learning_rate": 2.485151670459865e-05,
+      "loss": 0.0145,
+      "step": 722
+    },
+    {
+      "epoch": 51.642857142857146,
+      "grad_norm": 0.392578125,
+      "learning_rate": 2.485017262113619e-05,
+      "loss": 0.0124,
+      "step": 723
+    },
+    {
+      "epoch": 51.714285714285715,
+      "grad_norm": 0.1455078125,
+      "learning_rate": 2.484882251834963e-05,
+      "loss": 0.0145,
+      "step": 724
+    },
+    {
+      "epoch": 51.785714285714285,
+      "grad_norm": 0.138671875,
+      "learning_rate": 2.484746639689699e-05,
+      "loss": 0.0162,
+      "step": 725
+    },
+    {
+      "epoch": 51.857142857142854,
+      "grad_norm": 0.263671875,
+      "learning_rate": 2.4846104257439222e-05,
+      "loss": 0.0137,
+      "step": 726
+    },
+    {
+      "epoch": 51.92857142857143,
+      "grad_norm": 0.5234375,
+      "learning_rate": 2.4844736100640224e-05,
+      "loss": 0.0163,
+      "step": 727
+    },
+    {
+      "epoch": 52.0,
+      "grad_norm": 0.60546875,
+      "learning_rate": 2.484336192716682e-05,
+      "loss": 0.0163,
+      "step": 728
+    },
+    {
+      "epoch": 52.07142857142857,
+      "grad_norm": 0.2236328125,
+      "learning_rate": 2.4841981737688754e-05,
+      "loss": 0.0148,
+      "step": 729
+    },
+    {
+      "epoch": 52.142857142857146,
+      "grad_norm": 0.234375,
+      "learning_rate": 2.4840595532878727e-05,
+      "loss": 0.0122,
+      "step": 730
+    },
+    {
+      "epoch": 52.214285714285715,
+      "grad_norm": 0.1611328125,
+      "learning_rate": 2.483920331341235e-05,
+      "loss": 0.015,
+      "step": 731
+    },
+    {
+      "epoch": 52.285714285714285,
+      "grad_norm": 0.1298828125,
+      "learning_rate": 2.4837805079968176e-05,
+      "loss": 0.0156,
+      "step": 732
+    },
+    {
+      "epoch": 52.357142857142854,
+      "grad_norm": 0.271484375,
+      "learning_rate": 2.483640083322769e-05,
+      "loss": 0.0165,
+      "step": 733
+    },
+    {
+      "epoch": 52.42857142857143,
+      "grad_norm": 0.55859375,
+      "learning_rate": 2.48349905738753e-05,
+      "loss": 0.0152,
+      "step": 734
+    },
+    {
+      "epoch": 52.5,
+      "grad_norm": 0.5546875,
+      "learning_rate": 2.483357430259835e-05,
+      "loss": 0.0131,
+      "step": 735
+    },
+    {
+      "epoch": 52.57142857142857,
+      "grad_norm": 0.1669921875,
+      "learning_rate": 2.483215202008712e-05,
+      "loss": 0.0147,
+      "step": 736
+    },
+    {
+      "epoch": 52.642857142857146,
+      "grad_norm": 0.11474609375,
+      "learning_rate": 2.4830723727034805e-05,
+      "loss": 0.01,
+      "step": 737
+    },
+    {
+      "epoch": 52.714285714285715,
+      "grad_norm": 0.39453125,
+      "learning_rate": 2.4829289424137538e-05,
+      "loss": 0.0137,
+      "step": 738
+    },
+    {
+      "epoch": 52.785714285714285,
+      "grad_norm": 0.1796875,
+      "learning_rate": 2.4827849112094386e-05,
+      "loss": 0.0133,
+      "step": 739
+    },
+    {
+      "epoch": 52.857142857142854,
+      "grad_norm": 0.58203125,
+      "learning_rate": 2.482640279160734e-05,
+      "loss": 0.0162,
+      "step": 740
+    },
+    {
+      "epoch": 52.92857142857143,
+      "grad_norm": 0.3828125,
+      "learning_rate": 2.4824950463381314e-05,
+      "loss": 0.0139,
+      "step": 741
+    },
+    {
+      "epoch": 53.0,
+      "grad_norm": 0.462890625,
+      "learning_rate": 2.4823492128124154e-05,
+      "loss": 0.0181,
+      "step": 742
+    },
+    {
+      "epoch": 53.07142857142857,
+      "grad_norm": 0.1337890625,
+      "learning_rate": 2.482202778654664e-05,
+      "loss": 0.0132,
+      "step": 743
+    },
+    {
+      "epoch": 53.142857142857146,
+      "grad_norm": 0.171875,
+      "learning_rate": 2.4820557439362478e-05,
+      "loss": 0.0134,
+      "step": 744
+    },
+    {
+      "epoch": 53.214285714285715,
+      "grad_norm": 0.275390625,
+      "learning_rate": 2.4819081087288286e-05,
+      "loss": 0.0145,
+      "step": 745
+    },
+    {
+      "epoch": 53.285714285714285,
+      "grad_norm": 0.279296875,
+      "learning_rate": 2.481759873104363e-05,
+      "loss": 0.0134,
+      "step": 746
+    },
+    {
+      "epoch": 53.357142857142854,
+      "grad_norm": 0.232421875,
+      "learning_rate": 2.4816110371350982e-05,
+      "loss": 0.0145,
+      "step": 747
+    },
+    {
+      "epoch": 53.42857142857143,
+      "grad_norm": 0.33984375,
+      "learning_rate": 2.4814616008935762e-05,
+      "loss": 0.013,
+      "step": 748
+    },
+    {
+      "epoch": 53.5,
+      "grad_norm": 0.458984375,
+      "learning_rate": 2.4813115644526298e-05,
+      "loss": 0.0152,
+      "step": 749
+    },
+    {
+      "epoch": 53.57142857142857,
+      "grad_norm": 0.12890625,
+      "learning_rate": 2.4811609278853847e-05,
+      "loss": 0.0128,
+      "step": 750
+    },
+    {
+      "epoch": 53.642857142857146,
+      "grad_norm": 0.16015625,
+      "learning_rate": 2.4810096912652604e-05,
+      "loss": 0.018,
+      "step": 751
+    },
+    {
+      "epoch": 53.714285714285715,
+      "grad_norm": 0.310546875,
+      "learning_rate": 2.4808578546659663e-05,
+      "loss": 0.0157,
+      "step": 752
+    },
+    {
+      "epoch": 53.785714285714285,
+      "grad_norm": 0.12451171875,
+      "learning_rate": 2.4807054181615068e-05,
+      "loss": 0.0113,
+      "step": 753
+    },
+    {
+      "epoch": 53.857142857142854,
+      "grad_norm": 0.1865234375,
+      "learning_rate": 2.4805523818261774e-05,
+      "loss": 0.0145,
+      "step": 754
+    },
+    {
+      "epoch": 53.92857142857143,
+      "grad_norm": 0.46875,
+      "learning_rate": 2.4803987457345658e-05,
+      "loss": 0.0175,
+      "step": 755
+    },
+    {
+      "epoch": 54.0,
+      "grad_norm": 0.12353515625,
+      "learning_rate": 2.4802445099615525e-05,
+      "loss": 0.0126,
+      "step": 756
+    },
+    {
+      "epoch": 54.07142857142857,
+      "grad_norm": 0.1533203125,
+      "learning_rate": 2.4800896745823105e-05,
+      "loss": 0.0142,
+      "step": 757
+    },
+    {
+      "epoch": 54.142857142857146,
+      "grad_norm": 0.267578125,
+      "learning_rate": 2.4799342396723043e-05,
+      "loss": 0.0148,
+      "step": 758
+    },
+    {
+      "epoch": 54.214285714285715,
+      "grad_norm": 0.189453125,
+      "learning_rate": 2.479778205307291e-05,
+      "loss": 0.0172,
+      "step": 759
+    },
+    {
+      "epoch": 54.285714285714285,
+      "grad_norm": 0.236328125,
+      "learning_rate": 2.4796215715633196e-05,
+      "loss": 0.014,
+      "step": 760
+    },
+    {
+      "epoch": 54.357142857142854,
+      "grad_norm": 0.201171875,
+      "learning_rate": 2.4794643385167327e-05,
+      "loss": 0.0163,
+      "step": 761
+    },
+    {
+      "epoch": 54.42857142857143,
+      "grad_norm": 0.291015625,
+      "learning_rate": 2.479306506244162e-05,
+      "loss": 0.0161,
+      "step": 762
+    },
+    {
+      "epoch": 54.5,
+      "grad_norm": 0.1376953125,
+      "learning_rate": 2.4791480748225344e-05,
+      "loss": 0.015,
+      "step": 763
+    },
+    {
+      "epoch": 54.57142857142857,
+      "grad_norm": 0.1875,
+      "learning_rate": 2.478989044329067e-05,
+      "loss": 0.0133,
+      "step": 764
+    },
+    {
+      "epoch": 54.642857142857146,
+      "grad_norm": 0.1572265625,
+      "learning_rate": 2.4788294148412693e-05,
+      "loss": 0.0123,
+      "step": 765
+    },
+    {
+      "epoch": 54.714285714285715,
+      "grad_norm": 0.146484375,
+      "learning_rate": 2.478669186436943e-05,
+      "loss": 0.014,
+      "step": 766
+    },
+    {
+      "epoch": 54.785714285714285,
+      "grad_norm": 0.1328125,
+      "learning_rate": 2.478508359194181e-05,
+      "loss": 0.0159,
+      "step": 767
+    },
+    {
+      "epoch": 54.857142857142854,
+      "grad_norm": 0.189453125,
+      "learning_rate": 2.4783469331913688e-05,
+      "loss": 0.0147,
+      "step": 768
+    },
+    {
+      "epoch": 54.92857142857143,
+      "grad_norm": 0.2001953125,
+      "learning_rate": 2.4781849085071833e-05,
+      "loss": 0.0147,
+      "step": 769
+    },
+    {
+      "epoch": 55.0,
+      "grad_norm": 0.13671875,
+      "learning_rate": 2.478022285220594e-05,
+      "loss": 0.0142,
+      "step": 770
+    },
+    {
+      "epoch": 55.07142857142857,
+      "grad_norm": 0.55078125,
+      "learning_rate": 2.4778590634108613e-05,
+      "loss": 0.0163,
+      "step": 771
+    },
+    {
+      "epoch": 55.142857142857146,
+      "grad_norm": 0.25390625,
+      "learning_rate": 2.4776952431575364e-05,
+      "loss": 0.0159,
+      "step": 772
+    },
+    {
+      "epoch": 55.214285714285715,
+      "grad_norm": 0.546875,
+      "learning_rate": 2.477530824540465e-05,
+      "loss": 0.0193,
+      "step": 773
+    },
+    {
+      "epoch": 55.285714285714285,
+      "grad_norm": 0.22265625,
+      "learning_rate": 2.4773658076397816e-05,
+      "loss": 0.0137,
+      "step": 774
+    },
+    {
+      "epoch": 55.357142857142854,
+      "grad_norm": 0.36328125,
+      "learning_rate": 2.4772001925359133e-05,
+      "loss": 0.0166,
+      "step": 775
+    },
+    {
+      "epoch": 55.42857142857143,
+      "grad_norm": 0.34375,
+      "learning_rate": 2.4770339793095802e-05,
+      "loss": 0.0162,
+      "step": 776
+    },
+    {
+      "epoch": 55.5,
+      "grad_norm": 0.220703125,
+      "learning_rate": 2.4768671680417914e-05,
+      "loss": 0.0152,
+      "step": 777
+    },
+    {
+      "epoch": 55.57142857142857,
+      "grad_norm": 0.232421875,
+      "learning_rate": 2.476699758813849e-05,
+      "loss": 0.0124,
+      "step": 778
+    },
+    {
+      "epoch": 55.642857142857146,
+      "grad_norm": 0.1435546875,
+      "learning_rate": 2.476531751707346e-05,
+      "loss": 0.0144,
+      "step": 779
+    },
+    {
+      "epoch": 55.714285714285715,
+      "grad_norm": 0.1572265625,
+      "learning_rate": 2.4763631468041675e-05,
+      "loss": 0.0166,
+      "step": 780
+    },
+    {
+      "epoch": 55.785714285714285,
+      "grad_norm": 0.12109375,
+      "learning_rate": 2.4761939441864895e-05,
+      "loss": 0.0137,
+      "step": 781
+    },
+    {
+      "epoch": 55.857142857142854,
+      "grad_norm": 0.2177734375,
+      "learning_rate": 2.476024143936779e-05,
+      "loss": 0.0167,
+      "step": 782
+    },
+    {
+      "epoch": 55.92857142857143,
+      "grad_norm": 0.17578125,
+      "learning_rate": 2.4758537461377947e-05,
+      "loss": 0.0175,
+      "step": 783
+    },
+    {
+      "epoch": 56.0,
+      "grad_norm": 0.166015625,
+      "learning_rate": 2.475682750872586e-05,
+      "loss": 0.0098,
+      "step": 784
+    },
+    {
+      "epoch": 56.07142857142857,
+      "grad_norm": 0.1435546875,
+      "learning_rate": 2.475511158224495e-05,
+      "loss": 0.0153,
+      "step": 785
+    },
+    {
+      "epoch": 56.142857142857146,
+      "grad_norm": 0.1318359375,
+      "learning_rate": 2.4753389682771523e-05,
+      "loss": 0.0131,
+      "step": 786
+    },
+    {
+      "epoch": 56.214285714285715,
+      "grad_norm": 0.392578125,
+      "learning_rate": 2.4751661811144828e-05,
+      "loss": 0.0154,
+      "step": 787
+    },
+    {
+      "epoch": 56.285714285714285,
+      "grad_norm": 0.39453125,
+      "learning_rate": 2.4749927968206996e-05,
+      "loss": 0.0173,
+      "step": 788
+    },
+    {
+      "epoch": 56.357142857142854,
+      "grad_norm": 0.12890625,
+      "learning_rate": 2.4748188154803086e-05,
+      "loss": 0.0139,
+      "step": 789
+    },
+    {
+      "epoch": 56.42857142857143,
+      "grad_norm": 0.2236328125,
+      "learning_rate": 2.4746442371781066e-05,
+      "loss": 0.0164,
+      "step": 790
+    },
+    {
+      "epoch": 56.5,
+      "grad_norm": 0.322265625,
+      "learning_rate": 2.47446906199918e-05,
+      "loss": 0.0129,
+      "step": 791
+    },
+    {
+      "epoch": 56.57142857142857,
+      "grad_norm": 0.181640625,
+      "learning_rate": 2.474293290028908e-05,
+      "loss": 0.0151,
+      "step": 792
+    },
+    {
+      "epoch": 56.642857142857146,
+      "grad_norm": 0.224609375,
+      "learning_rate": 2.4741169213529594e-05,
+      "loss": 0.0157,
+      "step": 793
+    },
+    {
+      "epoch": 56.714285714285715,
+      "grad_norm": 0.16796875,
+      "learning_rate": 2.473939956057294e-05,
+      "loss": 0.0168,
+      "step": 794
+    },
+    {
+      "epoch": 56.785714285714285,
+      "grad_norm": 0.1650390625,
+      "learning_rate": 2.4737623942281624e-05,
+      "loss": 0.0159,
+      "step": 795
+    },
+    {
+      "epoch": 56.857142857142854,
+      "grad_norm": 0.2265625,
+      "learning_rate": 2.4735842359521064e-05,
+      "loss": 0.016,
+      "step": 796
+    },
+    {
+      "epoch": 56.92857142857143,
+      "grad_norm": 0.302734375,
+      "learning_rate": 2.473405481315958e-05,
+      "loss": 0.0124,
+      "step": 797
+    },
+    {
+      "epoch": 57.0,
+      "grad_norm": 0.1455078125,
+      "learning_rate": 2.4732261304068403e-05,
+      "loss": 0.0139,
+      "step": 798
+    },
+    {
+      "epoch": 57.07142857142857,
+      "grad_norm": 0.310546875,
+      "learning_rate": 2.4730461833121664e-05,
+      "loss": 0.0149,
+      "step": 799
+    },
+    {
+      "epoch": 57.142857142857146,
+      "grad_norm": 0.37890625,
+      "learning_rate": 2.4728656401196406e-05,
+      "loss": 0.0151,
+      "step": 800
+    },
+    {
+      "epoch": 57.214285714285715,
+      "grad_norm": 0.5390625,
+      "learning_rate": 2.4726845009172572e-05,
+      "loss": 0.0191,
+      "step": 801
+    },
+    {
+      "epoch": 57.285714285714285,
+      "grad_norm": 0.1513671875,
+      "learning_rate": 2.4725027657933013e-05,
+      "loss": 0.015,
+      "step": 802
+    },
+    {
+      "epoch": 57.357142857142854,
+      "grad_norm": 0.134765625,
+      "learning_rate": 2.4723204348363487e-05,
+      "loss": 0.0124,
+      "step": 803
+    },
+    {
+      "epoch": 57.42857142857143,
+      "grad_norm": 0.41015625,
+      "learning_rate": 2.4721375081352648e-05,
+      "loss": 0.015,
+      "step": 804
+    },
+    {
+      "epoch": 57.5,
+      "grad_norm": 0.25390625,
+      "learning_rate": 2.4719539857792068e-05,
+      "loss": 0.0154,
+      "step": 805
+    },
+    {
+      "epoch": 57.57142857142857,
+      "grad_norm": 0.2890625,
+      "learning_rate": 2.47176986785762e-05,
+      "loss": 0.0153,
+      "step": 806
+    },
+    {
+      "epoch": 57.642857142857146,
+      "grad_norm": 0.208984375,
+      "learning_rate": 2.471585154460242e-05,
+      "loss": 0.0193,
+      "step": 807
+    },
+    {
+      "epoch": 57.714285714285715,
+      "grad_norm": 0.2001953125,
+      "learning_rate": 2.4713998456771e-05,
+      "loss": 0.0119,
+      "step": 808
+    },
+    {
+      "epoch": 57.785714285714285,
+      "grad_norm": 0.11962890625,
+      "learning_rate": 2.4712139415985104e-05,
+      "loss": 0.0135,
+      "step": 809
+    },
+    {
+      "epoch": 57.857142857142854,
+      "grad_norm": 0.1484375,
+      "learning_rate": 2.471027442315081e-05,
+      "loss": 0.0143,
+      "step": 810
+    },
+    {
+      "epoch": 57.92857142857143,
+      "grad_norm": 0.1904296875,
+      "learning_rate": 2.47084034791771e-05,
+      "loss": 0.0139,
+      "step": 811
+    },
+    {
+      "epoch": 58.0,
+      "grad_norm": 0.1533203125,
+      "learning_rate": 2.4706526584975843e-05,
+      "loss": 0.0136,
+      "step": 812
+    },
+    {
+      "epoch": 58.07142857142857,
+      "grad_norm": 0.2294921875,
+      "learning_rate": 2.4704643741461815e-05,
+      "loss": 0.0178,
+      "step": 813
+    },
+    {
+      "epoch": 58.142857142857146,
+      "grad_norm": 0.15234375,
+      "learning_rate": 2.4702754949552686e-05,
+      "loss": 0.0136,
+      "step": 814
+    },
+    {
+      "epoch": 58.214285714285715,
+      "grad_norm": 0.251953125,
+      "learning_rate": 2.470086021016904e-05,
+      "loss": 0.0141,
+      "step": 815
+    },
+    {
+      "epoch": 58.285714285714285,
+      "grad_norm": 0.4296875,
+      "learning_rate": 2.4698959524234346e-05,
+      "loss": 0.0187,
+      "step": 816
+    },
+    {
+      "epoch": 58.357142857142854,
+      "grad_norm": 0.140625,
+      "learning_rate": 2.4697052892674972e-05,
+      "loss": 0.0132,
+      "step": 817
+    },
+    {
+      "epoch": 58.42857142857143,
+      "grad_norm": 0.185546875,
+      "learning_rate": 2.469514031642019e-05,
+      "loss": 0.0158,
+      "step": 818
+    },
+    {
+      "epoch": 58.5,
+      "grad_norm": 0.2177734375,
+      "learning_rate": 2.4693221796402166e-05,
+      "loss": 0.0157,
+      "step": 819
+    },
+    {
+      "epoch": 58.57142857142857,
+      "grad_norm": 0.2099609375,
+      "learning_rate": 2.4691297333555964e-05,
+      "loss": 0.0175,
+      "step": 820
+    },
+    {
+      "epoch": 58.642857142857146,
+      "grad_norm": 0.1630859375,
+      "learning_rate": 2.468936692881954e-05,
+      "loss": 0.0145,
+      "step": 821
+    },
+    {
+      "epoch": 58.714285714285715,
+      "grad_norm": 0.2119140625,
+      "learning_rate": 2.468743058313376e-05,
+      "loss": 0.0127,
+      "step": 822
+    },
+    {
+      "epoch": 58.785714285714285,
+      "grad_norm": 0.28515625,
+      "learning_rate": 2.4685488297442362e-05,
+      "loss": 0.0147,
+      "step": 823
+    },
+    {
+      "epoch": 58.857142857142854,
+      "grad_norm": 0.1923828125,
+      "learning_rate": 2.4683540072692003e-05,
+      "loss": 0.0134,
+      "step": 824
+    },
+    {
+      "epoch": 58.92857142857143,
+      "grad_norm": 0.1787109375,
+      "learning_rate": 2.4681585909832217e-05,
+      "loss": 0.0152,
+      "step": 825
+    },
+    {
+      "epoch": 59.0,
+      "grad_norm": 0.189453125,
+      "learning_rate": 2.4679625809815443e-05,
+      "loss": 0.0145,
+      "step": 826
+    },
+    {
+      "epoch": 59.07142857142857,
+      "grad_norm": 0.138671875,
+      "learning_rate": 2.4677659773597007e-05,
+      "loss": 0.0151,
+      "step": 827
+    },
+    {
+      "epoch": 59.142857142857146,
+      "grad_norm": 0.126953125,
+      "learning_rate": 2.4675687802135135e-05,
+      "loss": 0.013,
+      "step": 828
+    },
+    {
+      "epoch": 59.214285714285715,
+      "grad_norm": 0.2060546875,
+      "learning_rate": 2.4673709896390938e-05,
+      "loss": 0.0177,
+      "step": 829
+    },
+    {
+      "epoch": 59.285714285714285,
+      "grad_norm": 0.1455078125,
+      "learning_rate": 2.467172605732843e-05,
+      "loss": 0.0154,
+      "step": 830
+    },
+    {
+      "epoch": 59.357142857142854,
+      "grad_norm": 0.16015625,
+      "learning_rate": 2.4669736285914505e-05,
+      "loss": 0.0124,
+      "step": 831
+    },
+    {
+      "epoch": 59.42857142857143,
+      "grad_norm": 0.2099609375,
+      "learning_rate": 2.4667740583118952e-05,
+      "loss": 0.0162,
+      "step": 832
+    },
+    {
+      "epoch": 59.5,
+      "grad_norm": 0.1494140625,
+      "learning_rate": 2.466573894991445e-05,
+      "loss": 0.0139,
+      "step": 833
+    },
+    {
+      "epoch": 59.57142857142857,
+      "grad_norm": 0.279296875,
+      "learning_rate": 2.4663731387276583e-05,
+      "loss": 0.0185,
+      "step": 834
+    },
+    {
+      "epoch": 59.642857142857146,
+      "grad_norm": 0.2216796875,
+      "learning_rate": 2.46617178961838e-05,
+      "loss": 0.016,
+      "step": 835
+    },
+    {
+      "epoch": 59.714285714285715,
+      "grad_norm": 0.1474609375,
+      "learning_rate": 2.465969847761746e-05,
+      "loss": 0.0183,
+      "step": 836
+    },
+    {
+      "epoch": 59.785714285714285,
+      "grad_norm": 0.33984375,
+      "learning_rate": 2.4657673132561797e-05,
+      "loss": 0.0153,
+      "step": 837
+    },
+    {
+      "epoch": 59.857142857142854,
+      "grad_norm": 0.1416015625,
+      "learning_rate": 2.4655641862003943e-05,
+      "loss": 0.0147,
+      "step": 838
+    },
+    {
+      "epoch": 59.92857142857143,
+      "grad_norm": 0.2734375,
+      "learning_rate": 2.465360466693391e-05,
+      "loss": 0.016,
+      "step": 839
+    },
+    {
+      "epoch": 60.0,
+      "grad_norm": 0.3125,
+      "learning_rate": 2.465156154834461e-05,
+      "loss": 0.013,
+      "step": 840
+    },
+    {
+      "epoch": 60.07142857142857,
+      "grad_norm": 0.2353515625,
+      "learning_rate": 2.464951250723183e-05,
+      "loss": 0.0171,
+      "step": 841
+    },
+    {
+      "epoch": 60.142857142857146,
+      "grad_norm": 0.1220703125,
+      "learning_rate": 2.4647457544594246e-05,
+      "loss": 0.0161,
+      "step": 842
+    },
+    {
+      "epoch": 60.214285714285715,
+      "grad_norm": 0.1455078125,
+      "learning_rate": 2.4645396661433424e-05,
+      "loss": 0.0151,
+      "step": 843
+    },
+    {
+      "epoch": 60.285714285714285,
+      "grad_norm": 0.1669921875,
+      "learning_rate": 2.4643329858753818e-05,
+      "loss": 0.0147,
+      "step": 844
+    },
+    {
+      "epoch": 60.357142857142854,
+      "grad_norm": 0.451171875,
+      "learning_rate": 2.464125713756275e-05,
+      "loss": 0.0145,
+      "step": 845
+    },
+    {
+      "epoch": 60.42857142857143,
+      "grad_norm": 0.310546875,
+      "learning_rate": 2.4639178498870452e-05,
+      "loss": 0.0148,
+      "step": 846
+    },
+    {
+      "epoch": 60.5,
+      "grad_norm": 0.1435546875,
+      "learning_rate": 2.463709394369002e-05,
+      "loss": 0.0188,
+      "step": 847
+    },
+    {
+      "epoch": 60.57142857142857,
+      "grad_norm": 0.11767578125,
+      "learning_rate": 2.4635003473037444e-05,
+      "loss": 0.0124,
+      "step": 848
+    },
+    {
+      "epoch": 60.642857142857146,
+      "grad_norm": 0.240234375,
+      "learning_rate": 2.4632907087931593e-05,
+      "loss": 0.0143,
+      "step": 849
+    },
+    {
+      "epoch": 60.714285714285715,
+      "grad_norm": 0.13671875,
+      "learning_rate": 2.4630804789394218e-05,
+      "loss": 0.0125,
+      "step": 850
+    },
+    {
+      "epoch": 60.785714285714285,
+      "grad_norm": 0.208984375,
+      "learning_rate": 2.4628696578449956e-05,
+      "loss": 0.0136,
+      "step": 851
+    },
+    {
+      "epoch": 60.857142857142854,
+      "grad_norm": 0.158203125,
+      "learning_rate": 2.4626582456126324e-05,
+      "loss": 0.0128,
+      "step": 852
+    },
+    {
+      "epoch": 60.92857142857143,
+      "grad_norm": 0.1494140625,
+      "learning_rate": 2.4624462423453716e-05,
+      "loss": 0.0143,
+      "step": 853
+    },
+    {
+      "epoch": 61.0,
+      "grad_norm": 0.1337890625,
+      "learning_rate": 2.4622336481465414e-05,
+      "loss": 0.0173,
+      "step": 854
+    },
+    {
+      "epoch": 61.07142857142857,
+      "grad_norm": 0.302734375,
+      "learning_rate": 2.4620204631197574e-05,
+      "loss": 0.0139,
+      "step": 855
+    },
+    {
+      "epoch": 61.142857142857146,
+      "grad_norm": 0.2421875,
+      "learning_rate": 2.4618066873689238e-05,
+      "loss": 0.0142,
+      "step": 856
+    },
+    {
+      "epoch": 61.214285714285715,
+      "grad_norm": 0.1181640625,
+      "learning_rate": 2.4615923209982316e-05,
+      "loss": 0.0137,
+      "step": 857
+    },
+    {
+      "epoch": 61.285714285714285,
+      "grad_norm": 0.166015625,
+      "learning_rate": 2.4613773641121607e-05,
+      "loss": 0.0141,
+      "step": 858
+    },
+    {
+      "epoch": 61.357142857142854,
+      "grad_norm": 0.1435546875,
+      "learning_rate": 2.461161816815479e-05,
+      "loss": 0.0119,
+      "step": 859
+    },
+    {
+      "epoch": 61.42857142857143,
+      "grad_norm": 0.2109375,
+      "learning_rate": 2.4609456792132407e-05,
+      "loss": 0.016,
+      "step": 860
+    },
+    {
+      "epoch": 61.5,
+      "grad_norm": 0.1845703125,
+      "learning_rate": 2.4607289514107888e-05,
+      "loss": 0.0171,
+      "step": 861
+    },
+    {
+      "epoch": 61.57142857142857,
+      "grad_norm": 0.2333984375,
+      "learning_rate": 2.4605116335137547e-05,
+      "loss": 0.0146,
+      "step": 862
+    },
+    {
+      "epoch": 61.642857142857146,
+      "grad_norm": 0.2099609375,
+      "learning_rate": 2.4602937256280558e-05,
+      "loss": 0.0167,
+      "step": 863
+    },
+    {
+      "epoch": 61.714285714285715,
+      "grad_norm": 0.205078125,
+      "learning_rate": 2.4600752278598978e-05,
+      "loss": 0.0119,
+      "step": 864
+    },
+    {
+      "epoch": 61.785714285714285,
+      "grad_norm": 0.1337890625,
+      "learning_rate": 2.4598561403157734e-05,
+      "loss": 0.0186,
+      "step": 865
+    },
+    {
+      "epoch": 61.857142857142854,
+      "grad_norm": 0.181640625,
+      "learning_rate": 2.4596364631024643e-05,
+      "loss": 0.0142,
+      "step": 866
+    },
+    {
+      "epoch": 61.92857142857143,
+      "grad_norm": 0.28515625,
+      "learning_rate": 2.4594161963270373e-05,
+      "loss": 0.0148,
+      "step": 867
+    },
+    {
+      "epoch": 62.0,
+      "grad_norm": 0.1484375,
+      "learning_rate": 2.459195340096849e-05,
+      "loss": 0.015,
+      "step": 868
+    },
+    {
+      "epoch": 62.07142857142857,
+      "grad_norm": 0.177734375,
+      "learning_rate": 2.4589738945195413e-05,
+      "loss": 0.0152,
+      "step": 869
+    },
+    {
+      "epoch": 62.142857142857146,
+      "grad_norm": 0.1611328125,
+      "learning_rate": 2.458751859703044e-05,
+      "loss": 0.014,
+      "step": 870
+    },
+    {
+      "epoch": 62.214285714285715,
+      "grad_norm": 0.2119140625,
+      "learning_rate": 2.4585292357555746e-05,
+      "loss": 0.016,
+      "step": 871
+    },
+    {
+      "epoch": 62.285714285714285,
+      "grad_norm": 0.130859375,
+      "learning_rate": 2.4583060227856368e-05,
+      "loss": 0.0133,
+      "step": 872
+    },
+    {
+      "epoch": 62.357142857142854,
+      "grad_norm": 0.1357421875,
+      "learning_rate": 2.458082220902022e-05,
+      "loss": 0.0146,
+      "step": 873
+    },
+    {
+      "epoch": 62.42857142857143,
+      "grad_norm": 0.15625,
+      "learning_rate": 2.4578578302138087e-05,
+      "loss": 0.0129,
+      "step": 874
+    },
+    {
+      "epoch": 62.5,
+      "grad_norm": 0.130859375,
+      "learning_rate": 2.457632850830363e-05,
+      "loss": 0.0161,
+      "step": 875
+    },
+    {
+      "epoch": 62.57142857142857,
+      "grad_norm": 0.14453125,
+      "learning_rate": 2.4574072828613354e-05,
+      "loss": 0.014,
+      "step": 876
+    },
+    {
+      "epoch": 62.642857142857146,
+      "grad_norm": 0.140625,
+      "learning_rate": 2.457181126416666e-05,
+      "loss": 0.0144,
+      "step": 877
+    },
+    {
+      "epoch": 62.714285714285715,
+      "grad_norm": 0.13671875,
+      "learning_rate": 2.456954381606581e-05,
+      "loss": 0.0133,
+      "step": 878
+    },
+    {
+      "epoch": 62.785714285714285,
+      "grad_norm": 0.197265625,
+      "learning_rate": 2.4567270485415926e-05,
+      "loss": 0.0118,
+      "step": 879
+    },
+    {
+      "epoch": 62.857142857142854,
+      "grad_norm": 0.1591796875,
+      "learning_rate": 2.4564991273325e-05,
+      "loss": 0.0162,
+      "step": 880
+    },
+    {
+      "epoch": 62.92857142857143,
+      "grad_norm": 0.1650390625,
+      "learning_rate": 2.4562706180903894e-05,
+      "loss": 0.0154,
+      "step": 881
+    },
+    {
+      "epoch": 63.0,
+      "grad_norm": 0.294921875,
+      "learning_rate": 2.4560415209266334e-05,
+      "loss": 0.0135,
+      "step": 882
+    },
+    {
+      "epoch": 63.07142857142857,
+      "grad_norm": 0.1259765625,
+      "learning_rate": 2.455811835952892e-05,
+      "loss": 0.0128,
+      "step": 883
+    },
+    {
+      "epoch": 63.142857142857146,
+      "grad_norm": 0.369140625,
+      "learning_rate": 2.4555815632811092e-05,
+      "loss": 0.0191,
+      "step": 884
+    },
+    {
+      "epoch": 63.214285714285715,
+      "grad_norm": 0.302734375,
+      "learning_rate": 2.4553507030235184e-05,
+      "loss": 0.0158,
+      "step": 885
+    },
+    {
+      "epoch": 63.285714285714285,
+      "grad_norm": 0.3046875,
+      "learning_rate": 2.455119255292638e-05,
+      "loss": 0.0139,
+      "step": 886
+    },
+    {
+      "epoch": 63.357142857142854,
+      "grad_norm": 0.1630859375,
+      "learning_rate": 2.4548872202012718e-05,
+      "loss": 0.017,
+      "step": 887
+    },
+    {
+      "epoch": 63.42857142857143,
+      "grad_norm": 0.158203125,
+      "learning_rate": 2.4546545978625126e-05,
+      "loss": 0.0125,
+      "step": 888
+    },
+    {
+      "epoch": 63.5,
+      "grad_norm": 0.2333984375,
+      "learning_rate": 2.454421388389736e-05,
+      "loss": 0.0166,
+      "step": 889
+    },
+    {
+      "epoch": 63.57142857142857,
+      "grad_norm": 0.1767578125,
+      "learning_rate": 2.4541875918966063e-05,
+      "loss": 0.0162,
+      "step": 890
+    },
+    {
+      "epoch": 63.642857142857146,
+      "grad_norm": 0.181640625,
+      "learning_rate": 2.453953208497073e-05,
+      "loss": 0.0128,
+      "step": 891
+    },
+    {
+      "epoch": 63.714285714285715,
+      "grad_norm": 0.1328125,
+      "learning_rate": 2.4537182383053713e-05,
+      "loss": 0.0139,
+      "step": 892
+    },
+    {
+      "epoch": 63.785714285714285,
+      "grad_norm": 0.1748046875,
+      "learning_rate": 2.4534826814360236e-05,
+      "loss": 0.0153,
+      "step": 893
+    },
+    {
+      "epoch": 63.857142857142854,
+      "grad_norm": 0.15234375,
+      "learning_rate": 2.453246538003837e-05,
+      "loss": 0.0148,
+      "step": 894
+    },
+    {
+      "epoch": 63.92857142857143,
+      "grad_norm": 0.171875,
+      "learning_rate": 2.453009808123905e-05,
+      "loss": 0.0128,
+      "step": 895
+    },
+    {
+      "epoch": 64.0,
+      "grad_norm": 0.150390625,
+      "learning_rate": 2.452772491911607e-05,
+      "loss": 0.015,
+      "step": 896
+    },
+    {
+      "epoch": 64.07142857142857,
+      "grad_norm": 0.2431640625,
+      "learning_rate": 2.4525345894826073e-05,
+      "loss": 0.0167,
+      "step": 897
+    },
+    {
+      "epoch": 64.14285714285714,
+      "grad_norm": 0.1318359375,
+      "learning_rate": 2.4522961009528577e-05,
+      "loss": 0.0172,
+      "step": 898
+    },
+    {
+      "epoch": 64.21428571428571,
+      "grad_norm": 0.1337890625,
+      "learning_rate": 2.452057026438594e-05,
+      "loss": 0.0131,
+      "step": 899
+    },
+    {
+      "epoch": 64.28571428571429,
+      "grad_norm": 0.130859375,
+      "learning_rate": 2.451817366056339e-05,
+      "loss": 0.012,
+      "step": 900
+    },
+    {
+      "epoch": 64.35714285714286,
+      "grad_norm": 0.21484375,
+      "learning_rate": 2.4515771199228987e-05,
+      "loss": 0.0142,
+      "step": 901
+    },
+    {
+      "epoch": 64.42857142857143,
+      "grad_norm": 0.251953125,
+      "learning_rate": 2.4513362881553674e-05,
+      "loss": 0.0129,
+      "step": 902
+    },
+    {
+      "epoch": 64.5,
+      "grad_norm": 0.330078125,
+      "learning_rate": 2.4510948708711234e-05,
+      "loss": 0.016,
+      "step": 903
+    },
+    {
+      "epoch": 64.57142857142857,
+      "grad_norm": 0.19921875,
+      "learning_rate": 2.45085286818783e-05,
+      "loss": 0.0138,
+      "step": 904
+    },
+    {
+      "epoch": 64.64285714285714,
+      "grad_norm": 0.1787109375,
+      "learning_rate": 2.4506102802234372e-05,
+      "loss": 0.0146,
+      "step": 905
+    },
+    {
+      "epoch": 64.71428571428571,
+      "grad_norm": 0.1103515625,
+      "learning_rate": 2.450367107096179e-05,
+      "loss": 0.0149,
+      "step": 906
+    },
+    {
+      "epoch": 64.78571428571429,
+      "grad_norm": 0.150390625,
+      "learning_rate": 2.450123348924575e-05,
+      "loss": 0.0135,
+      "step": 907
+    },
+    {
+      "epoch": 64.85714285714286,
+      "grad_norm": 0.14453125,
+      "learning_rate": 2.4498790058274297e-05,
+      "loss": 0.0148,
+      "step": 908
+    },
+    {
+      "epoch": 64.92857142857143,
+      "grad_norm": 0.1669921875,
+      "learning_rate": 2.4496340779238335e-05,
+      "loss": 0.0158,
+      "step": 909
+    },
+    {
+      "epoch": 65.0,
+      "grad_norm": 0.1298828125,
+      "learning_rate": 2.449388565333161e-05,
+      "loss": 0.0134,
+      "step": 910
+    },
+    {
+      "epoch": 65.07142857142857,
+      "grad_norm": 0.1298828125,
+      "learning_rate": 2.449142468175072e-05,
+      "loss": 0.012,
+      "step": 911
+    },
+    {
+      "epoch": 65.14285714285714,
+      "grad_norm": 0.431640625,
+      "learning_rate": 2.4488957865695114e-05,
+      "loss": 0.0196,
+      "step": 912
+    },
+    {
+      "epoch": 65.21428571428571,
+      "grad_norm": 0.146484375,
+      "learning_rate": 2.4486485206367088e-05,
+      "loss": 0.0151,
+      "step": 913
+    },
+    {
+      "epoch": 65.28571428571429,
+      "grad_norm": 0.125,
+      "learning_rate": 2.4484006704971784e-05,
+      "loss": 0.012,
+      "step": 914
+    },
+    {
+      "epoch": 65.35714285714286,
+      "grad_norm": 0.330078125,
+      "learning_rate": 2.44815223627172e-05,
+      "loss": 0.0149,
+      "step": 915
+    },
+    {
+      "epoch": 65.42857142857143,
+      "grad_norm": 0.20703125,
+      "learning_rate": 2.4479032180814166e-05,
+      "loss": 0.0148,
+      "step": 916
+    },
+    {
+      "epoch": 65.5,
+      "grad_norm": 0.345703125,
+      "learning_rate": 2.4476536160476372e-05,
+      "loss": 0.0158,
+      "step": 917
+    },
+    {
+      "epoch": 65.57142857142857,
+      "grad_norm": 0.1533203125,
+      "learning_rate": 2.4474034302920347e-05,
+      "loss": 0.0145,
+      "step": 918
+    },
+    {
+      "epoch": 65.64285714285714,
+      "grad_norm": 0.12890625,
+      "learning_rate": 2.447152660936546e-05,
+      "loss": 0.0144,
+      "step": 919
+    },
+    {
+      "epoch": 65.71428571428571,
+      "grad_norm": 0.31640625,
+      "learning_rate": 2.4469013081033936e-05,
+      "loss": 0.0205,
+      "step": 920
+    },
+    {
+      "epoch": 65.78571428571429,
+      "grad_norm": 0.30078125,
+      "learning_rate": 2.446649371915084e-05,
+      "loss": 0.0167,
+      "step": 921
+    },
+    {
+      "epoch": 65.85714285714286,
+      "grad_norm": 0.1826171875,
+      "learning_rate": 2.446396852494408e-05,
+      "loss": 0.0141,
+      "step": 922
+    },
+    {
+      "epoch": 65.92857142857143,
+      "grad_norm": 0.150390625,
+      "learning_rate": 2.4461437499644394e-05,
+      "loss": 0.0151,
+      "step": 923
+    },
+    {
+      "epoch": 66.0,
+      "grad_norm": 0.16015625,
+      "learning_rate": 2.4458900644485385e-05,
+      "loss": 0.0145,
+      "step": 924
+    },
+    {
+      "epoch": 66.07142857142857,
+      "grad_norm": 0.1396484375,
+      "learning_rate": 2.4456357960703475e-05,
+      "loss": 0.0168,
+      "step": 925
+    },
+    {
+      "epoch": 66.14285714285714,
+      "grad_norm": 0.482421875,
+      "learning_rate": 2.4453809449537947e-05,
+      "loss": 0.013,
+      "step": 926
+    },
+    {
+      "epoch": 66.21428571428571,
+      "grad_norm": 0.1943359375,
+      "learning_rate": 2.4451255112230904e-05,
+      "loss": 0.015,
+      "step": 927
+    },
+    {
+      "epoch": 66.28571428571429,
+      "grad_norm": 0.1748046875,
+      "learning_rate": 2.444869495002731e-05,
+      "loss": 0.0197,
+      "step": 928
+    },
+    {
+      "epoch": 66.35714285714286,
+      "grad_norm": 0.298828125,
+      "learning_rate": 2.4446128964174954e-05,
+      "loss": 0.0172,
+      "step": 929
+    },
+    {
+      "epoch": 66.42857142857143,
+      "grad_norm": 0.138671875,
+      "learning_rate": 2.4443557155924457e-05,
+      "loss": 0.0129,
+      "step": 930
+    },
+    {
+      "epoch": 66.5,
+      "grad_norm": 0.11865234375,
+      "learning_rate": 2.4440979526529295e-05,
+      "loss": 0.0148,
+      "step": 931
+    },
+    {
+      "epoch": 66.57142857142857,
+      "grad_norm": 0.361328125,
+      "learning_rate": 2.4438396077245778e-05,
+      "loss": 0.0181,
+      "step": 932
+    },
+    {
+      "epoch": 66.64285714285714,
+      "grad_norm": 0.1806640625,
+      "learning_rate": 2.443580680933304e-05,
+      "loss": 0.0124,
+      "step": 933
+    },
+    {
+      "epoch": 66.71428571428571,
+      "grad_norm": 0.2412109375,
+      "learning_rate": 2.4433211724053055e-05,
+      "loss": 0.0161,
+      "step": 934
+    },
+    {
+      "epoch": 66.78571428571429,
+      "grad_norm": 0.146484375,
+      "learning_rate": 2.4430610822670654e-05,
+      "loss": 0.0162,
+      "step": 935
+    },
+    {
+      "epoch": 66.85714285714286,
+      "grad_norm": 0.1279296875,
+      "learning_rate": 2.4428004106453462e-05,
+      "loss": 0.0165,
+      "step": 936
+    },
+    {
+      "epoch": 66.92857142857143,
+      "grad_norm": 0.25390625,
+      "learning_rate": 2.4425391576671974e-05,
+      "loss": 0.012,
+      "step": 937
+    },
+    {
+      "epoch": 67.0,
+      "grad_norm": 0.216796875,
+      "learning_rate": 2.4422773234599506e-05,
+      "loss": 0.0141,
+      "step": 938
+    },
+    {
+      "epoch": 67.07142857142857,
+      "grad_norm": 0.2333984375,
+      "learning_rate": 2.44201490815122e-05,
+      "loss": 0.0113,
+      "step": 939
+    },
+    {
+      "epoch": 67.14285714285714,
+      "grad_norm": 0.1611328125,
+      "learning_rate": 2.441751911868904e-05,
+      "loss": 0.0163,
+      "step": 940
+    },
+    {
+      "epoch": 67.21428571428571,
+      "grad_norm": 0.212890625,
+      "learning_rate": 2.4414883347411836e-05,
+      "loss": 0.0135,
+      "step": 941
+    },
+    {
+      "epoch": 67.28571428571429,
+      "grad_norm": 0.1376953125,
+      "learning_rate": 2.441224176896524e-05,
+      "loss": 0.0141,
+      "step": 942
+    },
+    {
+      "epoch": 67.35714285714286,
+      "grad_norm": 0.1337890625,
+      "learning_rate": 2.4409594384636707e-05,
+      "loss": 0.0142,
+      "step": 943
+    },
+    {
+      "epoch": 67.42857142857143,
+      "grad_norm": 0.1337890625,
+      "learning_rate": 2.4406941195716555e-05,
+      "loss": 0.0139,
+      "step": 944
+    },
+    {
+      "epoch": 67.5,
+      "grad_norm": 0.375,
+      "learning_rate": 2.440428220349791e-05,
+      "loss": 0.0166,
+      "step": 945
+    },
+    {
+      "epoch": 67.57142857142857,
+      "grad_norm": 0.12158203125,
+      "learning_rate": 2.4401617409276735e-05,
+      "loss": 0.0138,
+      "step": 946
+    },
+    {
+      "epoch": 67.64285714285714,
+      "grad_norm": 0.1357421875,
+      "learning_rate": 2.4398946814351818e-05,
+      "loss": 0.015,
+      "step": 947
+    },
+    {
+      "epoch": 67.71428571428571,
+      "grad_norm": 0.255859375,
+      "learning_rate": 2.4396270420024774e-05,
+      "loss": 0.0156,
+      "step": 948
+    },
+    {
+      "epoch": 67.78571428571429,
+      "grad_norm": 0.11865234375,
+      "learning_rate": 2.439358822760004e-05,
+      "loss": 0.0134,
+      "step": 949
+    },
+    {
+      "epoch": 67.85714285714286,
+      "grad_norm": 0.318359375,
+      "learning_rate": 2.439090023838489e-05,
+      "loss": 0.0121,
+      "step": 950
+    },
+    {
+      "epoch": 67.92857142857143,
+      "grad_norm": 0.2119140625,
+      "learning_rate": 2.438820645368942e-05,
+      "loss": 0.0151,
+      "step": 951
+    },
+    {
+      "epoch": 68.0,
+      "grad_norm": 0.1279296875,
+      "learning_rate": 2.438550687482654e-05,
+      "loss": 0.013,
+      "step": 952
+    },
+    {
+      "epoch": 68.07142857142857,
+      "grad_norm": 0.388671875,
+      "learning_rate": 2.4382801503112003e-05,
+      "loss": 0.0119,
+      "step": 953
+    },
+    {
+      "epoch": 68.14285714285714,
+      "grad_norm": 0.1552734375,
+      "learning_rate": 2.438009033986436e-05,
+      "loss": 0.014,
+      "step": 954
+    },
+    {
+      "epoch": 68.21428571428571,
+      "grad_norm": 0.1357421875,
+      "learning_rate": 2.437737338640501e-05,
+      "loss": 0.0141,
+      "step": 955
+    },
+    {
+      "epoch": 68.28571428571429,
+      "grad_norm": 0.55078125,
+      "learning_rate": 2.4374650644058156e-05,
+      "loss": 0.0176,
+      "step": 956
+    },
+    {
+      "epoch": 68.35714285714286,
+      "grad_norm": 0.4296875,
+      "learning_rate": 2.4371922114150835e-05,
+      "loss": 0.0155,
+      "step": 957
+    },
+    {
+      "epoch": 68.42857142857143,
+      "grad_norm": 0.1474609375,
+      "learning_rate": 2.43691877980129e-05,
+      "loss": 0.0157,
+      "step": 958
+    },
+    {
+      "epoch": 68.5,
+      "grad_norm": 0.142578125,
+      "learning_rate": 2.4366447696977018e-05,
+      "loss": 0.0117,
+      "step": 959
+    },
+    {
+      "epoch": 68.57142857142857,
+      "grad_norm": 0.201171875,
+      "learning_rate": 2.4363701812378683e-05,
+      "loss": 0.0135,
+      "step": 960
+    },
+    {
+      "epoch": 68.64285714285714,
+      "grad_norm": 0.10498046875,
+      "learning_rate": 2.4360950145556208e-05,
+      "loss": 0.0134,
+      "step": 961
+    },
+    {
+      "epoch": 68.71428571428571,
+      "grad_norm": 0.12890625,
+      "learning_rate": 2.4358192697850718e-05,
+      "loss": 0.0153,
+      "step": 962
+    },
+    {
+      "epoch": 68.78571428571429,
+      "grad_norm": 0.169921875,
+      "learning_rate": 2.4355429470606166e-05,
+      "loss": 0.0154,
+      "step": 963
+    },
+    {
+      "epoch": 68.85714285714286,
+      "grad_norm": 0.2138671875,
+      "learning_rate": 2.4352660465169307e-05,
+      "loss": 0.014,
+      "step": 964
+    },
+    {
+      "epoch": 68.92857142857143,
+      "grad_norm": 0.361328125,
+      "learning_rate": 2.4349885682889727e-05,
+      "loss": 0.0158,
+      "step": 965
+    },
+    {
+      "epoch": 69.0,
+      "grad_norm": 0.146484375,
+      "learning_rate": 2.4347105125119824e-05,
+      "loss": 0.0151,
+      "step": 966
+    },
+    {
+      "epoch": 69.07142857142857,
+      "grad_norm": 0.193359375,
+      "learning_rate": 2.4344318793214797e-05,
+      "loss": 0.0135,
+      "step": 967
+    },
+    {
+      "epoch": 69.14285714285714,
+      "grad_norm": 0.203125,
+      "learning_rate": 2.4341526688532678e-05,
+      "loss": 0.0127,
+      "step": 968
+    },
+    {
+      "epoch": 69.21428571428571,
+      "grad_norm": 0.1103515625,
+      "learning_rate": 2.4338728812434305e-05,
+      "loss": 0.0135,
+      "step": 969
+    },
+    {
+      "epoch": 69.28571428571429,
+      "grad_norm": 0.484375,
+      "learning_rate": 2.4335925166283325e-05,
+      "loss": 0.0169,
+      "step": 970
+    },
+    {
+      "epoch": 69.35714285714286,
+      "grad_norm": 0.19921875,
+      "learning_rate": 2.4333115751446208e-05,
+      "loss": 0.0166,
+      "step": 971
+    },
+    {
+      "epoch": 69.42857142857143,
+      "grad_norm": 0.1591796875,
+      "learning_rate": 2.433030056929222e-05,
+      "loss": 0.0172,
+      "step": 972
+    },
+    {
+      "epoch": 69.5,
+      "grad_norm": 0.11962890625,
+      "learning_rate": 2.4327479621193448e-05,
+      "loss": 0.0127,
+      "step": 973
+    },
+    {
+      "epoch": 69.57142857142857,
+      "grad_norm": 0.3046875,
+      "learning_rate": 2.4324652908524798e-05,
+      "loss": 0.0164,
+      "step": 974
+    },
+    {
+      "epoch": 69.64285714285714,
+      "grad_norm": 0.1806640625,
+      "learning_rate": 2.432182043266396e-05,
+      "loss": 0.0153,
+      "step": 975
+    },
+    {
+      "epoch": 69.71428571428571,
+      "grad_norm": 0.259765625,
+      "learning_rate": 2.4318982194991463e-05,
+      "loss": 0.0152,
+      "step": 976
+    },
+    {
+      "epoch": 69.78571428571429,
+      "grad_norm": 0.357421875,
+      "learning_rate": 2.431613819689062e-05,
+      "loss": 0.0143,
+      "step": 977
+    },
+    {
+      "epoch": 69.85714285714286,
+      "grad_norm": 0.134765625,
+      "learning_rate": 2.431328843974756e-05,
+      "loss": 0.0153,
+      "step": 978
+    },
+    {
+      "epoch": 69.92857142857143,
+      "grad_norm": 0.1689453125,
+      "learning_rate": 2.4310432924951228e-05,
+      "loss": 0.0139,
+      "step": 979
+    },
+    {
+      "epoch": 70.0,
+      "grad_norm": 0.2080078125,
+      "learning_rate": 2.4307571653893364e-05,
+      "loss": 0.013,
+      "step": 980
+    },
+    {
+      "epoch": 70.07142857142857,
+      "grad_norm": 0.16015625,
+      "learning_rate": 2.4304704627968515e-05,
+      "loss": 0.0144,
+      "step": 981
+    },
+    {
+      "epoch": 70.14285714285714,
+      "grad_norm": 0.134765625,
+      "learning_rate": 2.4301831848574035e-05,
+      "loss": 0.0148,
+      "step": 982
+    },
+    {
+      "epoch": 70.21428571428571,
+      "grad_norm": 0.150390625,
+      "learning_rate": 2.429895331711008e-05,
+      "loss": 0.0169,
+      "step": 983
+    },
+    {
+      "epoch": 70.28571428571429,
+      "grad_norm": 0.138671875,
+      "learning_rate": 2.429606903497962e-05,
+      "loss": 0.0142,
+      "step": 984
+    },
+    {
+      "epoch": 70.35714285714286,
+      "grad_norm": 0.166015625,
+      "learning_rate": 2.4293179003588407e-05,
+      "loss": 0.0137,
+      "step": 985
+    },
+    {
+      "epoch": 70.42857142857143,
+      "grad_norm": 0.2294921875,
+      "learning_rate": 2.429028322434501e-05,
+      "loss": 0.0136,
+      "step": 986
+    },
+    {
+      "epoch": 70.5,
+      "grad_norm": 0.12158203125,
+      "learning_rate": 2.4287381698660807e-05,
+      "loss": 0.0133,
+      "step": 987
+    },
+    {
+      "epoch": 70.57142857142857,
+      "grad_norm": 0.1640625,
+      "learning_rate": 2.4284474427949956e-05,
+      "loss": 0.0144,
+      "step": 988
+    },
+    {
+      "epoch": 70.64285714285714,
+      "grad_norm": 0.14453125,
+      "learning_rate": 2.428156141362943e-05,
+      "loss": 0.0139,
+      "step": 989
+    },
+    {
+      "epoch": 70.71428571428571,
+      "grad_norm": 0.306640625,
+      "learning_rate": 2.427864265711899e-05,
+      "loss": 0.0148,
+      "step": 990
+    },
+    {
+      "epoch": 70.78571428571429,
+      "grad_norm": 0.1630859375,
+      "learning_rate": 2.427571815984121e-05,
+      "loss": 0.0149,
+      "step": 991
+    },
+    {
+      "epoch": 70.85714285714286,
+      "grad_norm": 0.1982421875,
+      "learning_rate": 2.427278792322145e-05,
+      "loss": 0.0143,
+      "step": 992
+    },
+    {
+      "epoch": 70.92857142857143,
+      "grad_norm": 0.1953125,
+      "learning_rate": 2.4269851948687877e-05,
+      "loss": 0.0129,
+      "step": 993
+    },
+    {
+      "epoch": 71.0,
+      "grad_norm": 0.1767578125,
+      "learning_rate": 2.426691023767145e-05,
+      "loss": 0.014,
+      "step": 994
+    },
+    {
+      "epoch": 71.07142857142857,
+      "grad_norm": 0.12255859375,
+      "learning_rate": 2.426396279160591e-05,
+      "loss": 0.0144,
+      "step": 995
+    },
+    {
+      "epoch": 71.14285714285714,
+      "grad_norm": 0.130859375,
+      "learning_rate": 2.426100961192782e-05,
+      "loss": 0.0161,
+      "step": 996
+    },
+    {
+      "epoch": 71.21428571428571,
+      "grad_norm": 0.33203125,
+      "learning_rate": 2.425805070007652e-05,
+      "loss": 0.0135,
+      "step": 997
+    },
+    {
+      "epoch": 71.28571428571429,
+      "grad_norm": 0.1240234375,
+      "learning_rate": 2.4255086057494146e-05,
+      "loss": 0.0165,
+      "step": 998
+    },
+    {
+      "epoch": 71.35714285714286,
+      "grad_norm": 0.16796875,
+      "learning_rate": 2.4252115685625627e-05,
+      "loss": 0.0144,
+      "step": 999
+    },
+    {
+      "epoch": 71.42857142857143,
+      "grad_norm": 0.1806640625,
+      "learning_rate": 2.4249139585918695e-05,
+      "loss": 0.0154,
+      "step": 1000
+    },
+    {
+      "epoch": 71.5,
+      "grad_norm": 0.2373046875,
+      "learning_rate": 2.4246157759823855e-05,
+      "loss": 0.0145,
+      "step": 1001
+    },
+    {
+      "epoch": 71.57142857142857,
+      "grad_norm": 0.2578125,
+      "learning_rate": 2.4243170208794418e-05,
+      "loss": 0.0137,
+      "step": 1002
+    },
+    {
+      "epoch": 71.64285714285714,
+      "grad_norm": 0.2099609375,
+      "learning_rate": 2.424017693428648e-05,
+      "loss": 0.0147,
+      "step": 1003
+    },
+    {
+      "epoch": 71.71428571428571,
+      "grad_norm": 0.119140625,
+      "learning_rate": 2.4237177937758927e-05,
+      "loss": 0.0143,
+      "step": 1004
+    },
+    {
+      "epoch": 71.78571428571429,
+      "grad_norm": 0.212890625,
+      "learning_rate": 2.423417322067343e-05,
+      "loss": 0.0156,
+      "step": 1005
+    },
+    {
+      "epoch": 71.85714285714286,
+      "grad_norm": 0.1650390625,
+      "learning_rate": 2.423116278449445e-05,
+      "loss": 0.0156,
+      "step": 1006
+    },
+    {
+      "epoch": 71.92857142857143,
+      "grad_norm": 0.11669921875,
+      "learning_rate": 2.4228146630689245e-05,
+      "loss": 0.0117,
+      "step": 1007
+    },
+    {
+      "epoch": 72.0,
+      "grad_norm": 0.1591796875,
+      "learning_rate": 2.4225124760727847e-05,
+      "loss": 0.014,
+      "step": 1008
+    },
+    {
+      "epoch": 72.07142857142857,
+      "grad_norm": 0.1630859375,
+      "learning_rate": 2.4222097176083074e-05,
+      "loss": 0.0137,
+      "step": 1009
+    },
+    {
+      "epoch": 72.14285714285714,
+      "grad_norm": 0.166015625,
+      "learning_rate": 2.421906387823055e-05,
+      "loss": 0.0163,
+      "step": 1010
+    },
+    {
+      "epoch": 72.21428571428571,
+      "grad_norm": 0.265625,
+      "learning_rate": 2.4216024868648644e-05,
+      "loss": 0.0146,
+      "step": 1011
+    },
+    {
+      "epoch": 72.28571428571429,
+      "grad_norm": 0.2412109375,
+      "learning_rate": 2.421298014881855e-05,
+      "loss": 0.0159,
+      "step": 1012
+    },
+    {
+      "epoch": 72.35714285714286,
+      "grad_norm": 0.1875,
+      "learning_rate": 2.4209929720224223e-05,
+      "loss": 0.0147,
+      "step": 1013
+    },
+    {
+      "epoch": 72.42857142857143,
+      "grad_norm": 0.171875,
+      "learning_rate": 2.4206873584352404e-05,
+      "loss": 0.0136,
+      "step": 1014
+    },
+    {
+      "epoch": 72.5,
+      "grad_norm": 0.2294921875,
+      "learning_rate": 2.4203811742692617e-05,
+      "loss": 0.0154,
+      "step": 1015
+    },
+    {
+      "epoch": 72.57142857142857,
+      "grad_norm": 0.28125,
+      "learning_rate": 2.420074419673717e-05,
+      "loss": 0.0143,
+      "step": 1016
+    },
+    {
+      "epoch": 72.64285714285714,
+      "grad_norm": 0.2138671875,
+      "learning_rate": 2.419767094798114e-05,
+      "loss": 0.012,
+      "step": 1017
+    },
+    {
+      "epoch": 72.71428571428571,
+      "grad_norm": 0.1484375,
+      "learning_rate": 2.4194591997922393e-05,
+      "loss": 0.0135,
+      "step": 1018
+    },
+    {
+      "epoch": 72.78571428571429,
+      "grad_norm": 0.212890625,
+      "learning_rate": 2.4191507348061572e-05,
+      "loss": 0.0137,
+      "step": 1019
+    },
+    {
+      "epoch": 72.85714285714286,
+      "grad_norm": 0.23828125,
+      "learning_rate": 2.4188416999902104e-05,
+      "loss": 0.0152,
+      "step": 1020
+    },
+    {
+      "epoch": 72.92857142857143,
+      "grad_norm": 0.1435546875,
+      "learning_rate": 2.418532095495018e-05,
+      "loss": 0.0163,
+      "step": 1021
+    },
+    {
+      "epoch": 73.0,
+      "grad_norm": 0.138671875,
+      "learning_rate": 2.418221921471478e-05,
+      "loss": 0.0182,
+      "step": 1022
+    },
+    {
+      "epoch": 73.07142857142857,
+      "grad_norm": 0.181640625,
+      "learning_rate": 2.4179111780707652e-05,
+      "loss": 0.0142,
+      "step": 1023
+    },
+    {
+      "epoch": 73.14285714285714,
+      "grad_norm": 0.23828125,
+      "learning_rate": 2.417599865444332e-05,
+      "loss": 0.0129,
+      "step": 1024
+    },
+    {
+      "epoch": 73.21428571428571,
+      "grad_norm": 0.1513671875,
+      "learning_rate": 2.4172879837439086e-05,
+      "loss": 0.0162,
+      "step": 1025
+    },
+    {
+      "epoch": 73.28571428571429,
+      "grad_norm": 0.419921875,
+      "learning_rate": 2.4169755331215023e-05,
+      "loss": 0.016,
+      "step": 1026
+    },
+    {
+      "epoch": 73.35714285714286,
+      "grad_norm": 0.1572265625,
+      "learning_rate": 2.4166625137293976e-05,
+      "loss": 0.0162,
+      "step": 1027
+    },
+    {
+      "epoch": 73.42857142857143,
+      "grad_norm": 0.173828125,
+      "learning_rate": 2.4163489257201564e-05,
+      "loss": 0.0154,
+      "step": 1028
+    },
+    {
+      "epoch": 73.5,
+      "grad_norm": 0.1865234375,
+      "learning_rate": 2.4160347692466184e-05,
+      "loss": 0.0134,
+      "step": 1029
+    },
+    {
+      "epoch": 73.57142857142857,
+      "grad_norm": 0.38671875,
+      "learning_rate": 2.4157200444618985e-05,
+      "loss": 0.015,
+      "step": 1030
+    },
+    {
+      "epoch": 73.64285714285714,
+      "grad_norm": 0.123046875,
+      "learning_rate": 2.4154047515193904e-05,
+      "loss": 0.0147,
+      "step": 1031
+    },
+    {
+      "epoch": 73.71428571428571,
+      "grad_norm": 0.498046875,
+      "learning_rate": 2.4150888905727638e-05,
+      "loss": 0.0173,
+      "step": 1032
+    },
+    {
+      "epoch": 73.78571428571429,
+      "grad_norm": 0.23046875,
+      "learning_rate": 2.4147724617759656e-05,
+      "loss": 0.0167,
+      "step": 1033
+    },
+    {
+      "epoch": 73.85714285714286,
+      "grad_norm": 0.287109375,
+      "learning_rate": 2.4144554652832196e-05,
+      "loss": 0.0141,
+      "step": 1034
+    },
+    {
+      "epoch": 73.92857142857143,
+      "grad_norm": 0.44921875,
+      "learning_rate": 2.4141379012490255e-05,
+      "loss": 0.0172,
+      "step": 1035
+    },
+    {
+      "epoch": 74.0,
+      "grad_norm": 0.1552734375,
+      "learning_rate": 2.4138197698281606e-05,
+      "loss": 0.0163,
+      "step": 1036
+    },
+    {
+      "epoch": 74.07142857142857,
+      "grad_norm": 0.3046875,
+      "learning_rate": 2.4135010711756777e-05,
+      "loss": 0.0151,
+      "step": 1037
+    },
+    {
+      "epoch": 74.14285714285714,
+      "grad_norm": 0.365234375,
+      "learning_rate": 2.4131818054469078e-05,
+      "loss": 0.0166,
+      "step": 1038
+    },
+    {
+      "epoch": 74.21428571428571,
+      "grad_norm": 0.197265625,
+      "learning_rate": 2.4128619727974562e-05,
+      "loss": 0.016,
+      "step": 1039
+    },
+    {
+      "epoch": 74.28571428571429,
+      "grad_norm": 0.169921875,
+      "learning_rate": 2.412541573383205e-05,
+      "loss": 0.0129,
+      "step": 1040
+    },
+    {
+      "epoch": 74.35714285714286,
+      "grad_norm": 0.216796875,
+      "learning_rate": 2.4122206073603142e-05,
+      "loss": 0.0153,
+      "step": 1041
+    },
+    {
+      "epoch": 74.42857142857143,
+      "grad_norm": 0.30859375,
+      "learning_rate": 2.4118990748852183e-05,
+      "loss": 0.0158,
+      "step": 1042
+    },
+    {
+      "epoch": 74.5,
+      "grad_norm": 0.1552734375,
+      "learning_rate": 2.4115769761146282e-05,
+      "loss": 0.0163,
+      "step": 1043
+    },
+    {
+      "epoch": 74.57142857142857,
+      "grad_norm": 0.2138671875,
+      "learning_rate": 2.4112543112055308e-05,
+      "loss": 0.0146,
+      "step": 1044
+    },
+    {
+      "epoch": 74.64285714285714,
+      "grad_norm": 0.2177734375,
+      "learning_rate": 2.410931080315189e-05,
+      "loss": 0.0156,
+      "step": 1045
+    },
+    {
+      "epoch": 74.71428571428571,
+      "grad_norm": 0.25,
+      "learning_rate": 2.4106072836011422e-05,
+      "loss": 0.014,
+      "step": 1046
+    },
+    {
+      "epoch": 74.78571428571429,
+      "grad_norm": 0.11669921875,
+      "learning_rate": 2.410282921221204e-05,
+      "loss": 0.0153,
+      "step": 1047
+    },
+    {
+      "epoch": 74.85714285714286,
+      "grad_norm": 0.125,
+      "learning_rate": 2.409957993333466e-05,
+      "loss": 0.0159,
+      "step": 1048
+    },
+    {
+      "epoch": 74.92857142857143,
+      "grad_norm": 0.1298828125,
+      "learning_rate": 2.4096325000962927e-05,
+      "loss": 0.0144,
+      "step": 1049
+    },
+    {
+      "epoch": 75.0,
+      "grad_norm": 0.267578125,
+      "learning_rate": 2.409306441668327e-05,
+      "loss": 0.0148,
+      "step": 1050
+    },
+    {
+      "epoch": 75.07142857142857,
+      "grad_norm": 0.1845703125,
+      "learning_rate": 2.4089798182084845e-05,
+      "loss": 0.0144,
+      "step": 1051
+    },
+    {
+      "epoch": 75.14285714285714,
+      "grad_norm": 0.22265625,
+      "learning_rate": 2.408652629875958e-05,
+      "loss": 0.0142,
+      "step": 1052
+    },
+    {
+      "epoch": 75.21428571428571,
+      "grad_norm": 0.11572265625,
+      "learning_rate": 2.408324876830215e-05,
+      "loss": 0.0152,
+      "step": 1053
+    },
+    {
+      "epoch": 75.28571428571429,
+      "grad_norm": 0.1640625,
+      "learning_rate": 2.407996559230999e-05,
+      "loss": 0.0138,
+      "step": 1054
+    },
+    {
+      "epoch": 75.35714285714286,
+      "grad_norm": 0.1318359375,
+      "learning_rate": 2.4076676772383277e-05,
+      "loss": 0.0149,
+      "step": 1055
+    },
+    {
+      "epoch": 75.42857142857143,
+      "grad_norm": 0.181640625,
+      "learning_rate": 2.407338231012494e-05,
+      "loss": 0.0142,
+      "step": 1056
+    },
+    {
+      "epoch": 75.5,
+      "grad_norm": 0.130859375,
+      "learning_rate": 2.4070082207140653e-05,
+      "loss": 0.0116,
+      "step": 1057
+    },
+    {
+      "epoch": 75.57142857142857,
+      "grad_norm": 0.1728515625,
+      "learning_rate": 2.4066776465038858e-05,
+      "loss": 0.0144,
+      "step": 1058
+    },
+    {
+      "epoch": 75.64285714285714,
+      "grad_norm": 0.1328125,
+      "learning_rate": 2.4063465085430725e-05,
+      "loss": 0.0141,
+      "step": 1059
+    },
+    {
+      "epoch": 75.71428571428571,
+      "grad_norm": 0.1435546875,
+      "learning_rate": 2.4060148069930188e-05,
+      "loss": 0.0151,
+      "step": 1060
+    },
+    {
+      "epoch": 75.78571428571429,
+      "grad_norm": 0.283203125,
+      "learning_rate": 2.4056825420153917e-05,
+      "loss": 0.0159,
+      "step": 1061
+    },
+    {
+      "epoch": 75.85714285714286,
+      "grad_norm": 0.18359375,
+      "learning_rate": 2.4053497137721324e-05,
+      "loss": 0.0136,
+      "step": 1062
+    },
+    {
+      "epoch": 75.92857142857143,
+      "grad_norm": 0.224609375,
+      "learning_rate": 2.4050163224254586e-05,
+      "loss": 0.017,
+      "step": 1063
+    },
+    {
+      "epoch": 76.0,
+      "grad_norm": 0.16796875,
+      "learning_rate": 2.4046823681378603e-05,
+      "loss": 0.0148,
+      "step": 1064
+    },
+    {
+      "epoch": 76.07142857142857,
+      "grad_norm": 0.357421875,
+      "learning_rate": 2.404347851072103e-05,
+      "loss": 0.0162,
+      "step": 1065
+    },
+    {
+      "epoch": 76.14285714285714,
+      "grad_norm": 0.126953125,
+      "learning_rate": 2.4040127713912264e-05,
+      "loss": 0.0151,
+      "step": 1066
+    },
+    {
+      "epoch": 76.21428571428571,
+      "grad_norm": 0.32421875,
+      "learning_rate": 2.4036771292585438e-05,
+      "loss": 0.0162,
+      "step": 1067
+    },
+    {
+      "epoch": 76.28571428571429,
+      "grad_norm": 0.294921875,
+      "learning_rate": 2.4033409248376437e-05,
+      "loss": 0.0165,
+      "step": 1068
+    },
+    {
+      "epoch": 76.35714285714286,
+      "grad_norm": 0.306640625,
+      "learning_rate": 2.4030041582923877e-05,
+      "loss": 0.0159,
+      "step": 1069
+    },
+    {
+      "epoch": 76.42857142857143,
+      "grad_norm": 0.28515625,
+      "learning_rate": 2.4026668297869118e-05,
+      "loss": 0.0144,
+      "step": 1070
+    },
+    {
+      "epoch": 76.5,
+      "grad_norm": 0.33984375,
+      "learning_rate": 2.4023289394856257e-05,
+      "loss": 0.0139,
+      "step": 1071
+    },
+    {
+      "epoch": 76.57142857142857,
+      "grad_norm": 0.27734375,
+      "learning_rate": 2.4019904875532134e-05,
+      "loss": 0.0144,
+      "step": 1072
+    },
+    {
+      "epoch": 76.64285714285714,
+      "grad_norm": 0.13671875,
+      "learning_rate": 2.401651474154632e-05,
+      "loss": 0.0126,
+      "step": 1073
+    },
+    {
+      "epoch": 76.71428571428571,
+      "grad_norm": 0.12451171875,
+      "learning_rate": 2.401311899455112e-05,
+      "loss": 0.0154,
+      "step": 1074
+    },
+    {
+      "epoch": 76.78571428571429,
+      "grad_norm": 0.09765625,
+      "learning_rate": 2.400971763620159e-05,
+      "loss": 0.0119,
+      "step": 1075
+    },
+    {
+      "epoch": 76.85714285714286,
+      "grad_norm": 0.216796875,
+      "learning_rate": 2.4006310668155508e-05,
+      "loss": 0.0161,
+      "step": 1076
+    },
+    {
+      "epoch": 76.92857142857143,
+      "grad_norm": 0.484375,
+      "learning_rate": 2.400289809207338e-05,
+      "loss": 0.0161,
+      "step": 1077
+    },
+    {
+      "epoch": 77.0,
+      "grad_norm": 0.14453125,
+      "learning_rate": 2.3999479909618466e-05,
+      "loss": 0.0132,
+      "step": 1078
+    },
+    {
+      "epoch": 77.07142857142857,
+      "grad_norm": 0.291015625,
+      "learning_rate": 2.399605612245674e-05,
+      "loss": 0.0141,
+      "step": 1079
+    },
+    {
+      "epoch": 77.14285714285714,
+      "grad_norm": 0.2216796875,
+      "learning_rate": 2.3992626732256912e-05,
+      "loss": 0.0153,
+      "step": 1080
+    },
+    {
+      "epoch": 77.21428571428571,
+      "grad_norm": 0.2333984375,
+      "learning_rate": 2.398919174069043e-05,
+      "loss": 0.0142,
+      "step": 1081
+    },
+    {
+      "epoch": 77.28571428571429,
+      "grad_norm": 0.27734375,
+      "learning_rate": 2.3985751149431468e-05,
+      "loss": 0.0146,
+      "step": 1082
+    },
+    {
+      "epoch": 77.35714285714286,
+      "grad_norm": 0.1572265625,
+      "learning_rate": 2.3982304960156925e-05,
+      "loss": 0.0171,
+      "step": 1083
+    },
+    {
+      "epoch": 77.42857142857143,
+      "grad_norm": 0.3125,
+      "learning_rate": 2.3978853174546433e-05,
+      "loss": 0.0161,
+      "step": 1084
+    },
+    {
+      "epoch": 77.5,
+      "grad_norm": 0.1279296875,
+      "learning_rate": 2.397539579428235e-05,
+      "loss": 0.0144,
+      "step": 1085
+    },
+    {
+      "epoch": 77.57142857142857,
+      "grad_norm": 0.28125,
+      "learning_rate": 2.3971932821049765e-05,
+      "loss": 0.0141,
+      "step": 1086
+    },
+    {
+      "epoch": 77.64285714285714,
+      "grad_norm": 0.26953125,
+      "learning_rate": 2.3968464256536482e-05,
+      "loss": 0.0141,
+      "step": 1087
+    },
+    {
+      "epoch": 77.71428571428571,
+      "grad_norm": 0.1103515625,
+      "learning_rate": 2.3964990102433044e-05,
+      "loss": 0.0113,
+      "step": 1088
+    },
+    {
+      "epoch": 77.78571428571429,
+      "grad_norm": 0.140625,
+      "learning_rate": 2.3961510360432707e-05,
+      "loss": 0.0142,
+      "step": 1089
+    },
+    {
+      "epoch": 77.85714285714286,
+      "grad_norm": 0.2275390625,
+      "learning_rate": 2.3958025032231455e-05,
+      "loss": 0.0156,
+      "step": 1090
+    },
+    {
+      "epoch": 77.92857142857143,
+      "grad_norm": 0.154296875,
+      "learning_rate": 2.3954534119527996e-05,
+      "loss": 0.0154,
+      "step": 1091
+    },
+    {
+      "epoch": 78.0,
+      "grad_norm": 0.259765625,
+      "learning_rate": 2.3951037624023758e-05,
+      "loss": 0.0124,
+      "step": 1092
+    },
+    {
+      "epoch": 78.07142857142857,
+      "grad_norm": 0.255859375,
+      "learning_rate": 2.3947535547422896e-05,
+      "loss": 0.0147,
+      "step": 1093
+    },
+    {
+      "epoch": 78.14285714285714,
+      "grad_norm": 0.291015625,
+      "learning_rate": 2.394402789143227e-05,
+      "loss": 0.0146,
+      "step": 1094
+    },
+    {
+      "epoch": 78.21428571428571,
+      "grad_norm": 0.435546875,
+      "learning_rate": 2.394051465776147e-05,
+      "loss": 0.0171,
+      "step": 1095
+    },
+    {
+      "epoch": 78.28571428571429,
+      "grad_norm": 0.1376953125,
+      "learning_rate": 2.3936995848122812e-05,
+      "loss": 0.0153,
+      "step": 1096
+    },
+    {
+      "epoch": 78.35714285714286,
+      "grad_norm": 0.384765625,
+      "learning_rate": 2.3933471464231312e-05,
+      "loss": 0.0135,
+      "step": 1097
+    },
+    {
+      "epoch": 78.42857142857143,
+      "grad_norm": 0.453125,
+      "learning_rate": 2.392994150780472e-05,
+      "loss": 0.0138,
+      "step": 1098
+    },
+    {
+      "epoch": 78.5,
+      "grad_norm": 0.279296875,
+      "learning_rate": 2.392640598056349e-05,
+      "loss": 0.0153,
+      "step": 1099
+    },
+    {
+      "epoch": 78.57142857142857,
+      "grad_norm": 0.357421875,
+      "learning_rate": 2.3922864884230788e-05,
+      "loss": 0.0147,
+      "step": 1100
+    },
+    {
+      "epoch": 78.64285714285714,
+      "grad_norm": 0.28515625,
+      "learning_rate": 2.391931822053251e-05,
+      "loss": 0.0127,
+      "step": 1101
+    },
+    {
+      "epoch": 78.71428571428571,
+      "grad_norm": 0.1455078125,
+      "learning_rate": 2.391576599119726e-05,
+      "loss": 0.0145,
+      "step": 1102
+    },
+    {
+      "epoch": 78.78571428571429,
+      "grad_norm": 0.25,
+      "learning_rate": 2.3912208197956343e-05,
+      "loss": 0.0152,
+      "step": 1103
+    },
+    {
+      "epoch": 78.85714285714286,
+      "grad_norm": 0.2373046875,
+      "learning_rate": 2.3908644842543788e-05,
+      "loss": 0.0135,
+      "step": 1104
+    },
+    {
+      "epoch": 78.92857142857143,
+      "grad_norm": 0.208984375,
+      "learning_rate": 2.390507592669633e-05,
+      "loss": 0.013,
+      "step": 1105
+    },
+    {
+      "epoch": 79.0,
+      "grad_norm": 0.271484375,
+      "learning_rate": 2.390150145215341e-05,
+      "loss": 0.0157,
+      "step": 1106
+    },
+    {
+      "epoch": 79.07142857142857,
+      "grad_norm": 0.2255859375,
+      "learning_rate": 2.38979214206572e-05,
+      "loss": 0.014,
+      "step": 1107
+    },
+    {
+      "epoch": 79.14285714285714,
+      "grad_norm": 0.26171875,
+      "learning_rate": 2.3894335833952548e-05,
+      "loss": 0.015,
+      "step": 1108
+    },
+    {
+      "epoch": 79.21428571428571,
+      "grad_norm": 0.146484375,
+      "learning_rate": 2.389074469378703e-05,
+      "loss": 0.0159,
+      "step": 1109
+    },
+    {
+      "epoch": 79.28571428571429,
+      "grad_norm": 0.13671875,
+      "learning_rate": 2.3887148001910925e-05,
+      "loss": 0.0177,
+      "step": 1110
+    },
+    {
+      "epoch": 79.35714285714286,
+      "grad_norm": 0.13671875,
+      "learning_rate": 2.3883545760077215e-05,
+      "loss": 0.0122,
+      "step": 1111
+    },
+    {
+      "epoch": 79.42857142857143,
+      "grad_norm": 0.263671875,
+      "learning_rate": 2.3879937970041595e-05,
+      "loss": 0.0151,
+      "step": 1112
+    },
+    {
+      "epoch": 79.5,
+      "grad_norm": 0.11669921875,
+      "learning_rate": 2.3876324633562446e-05,
+      "loss": 0.0162,
+      "step": 1113
+    },
+    {
+      "epoch": 79.57142857142857,
+      "grad_norm": 0.125,
+      "learning_rate": 2.3872705752400874e-05,
+      "loss": 0.0146,
+      "step": 1114
+    },
+    {
+      "epoch": 79.64285714285714,
+      "grad_norm": 0.275390625,
+      "learning_rate": 2.386908132832068e-05,
+      "loss": 0.0141,
+      "step": 1115
+    },
+    {
+      "epoch": 79.71428571428571,
+      "grad_norm": 0.1181640625,
+      "learning_rate": 2.386545136308836e-05,
+      "loss": 0.0147,
+      "step": 1116
+    },
+    {
+      "epoch": 79.78571428571429,
+      "grad_norm": 0.1796875,
+      "learning_rate": 2.386181585847311e-05,
+      "loss": 0.0169,
+      "step": 1117
+    },
+    {
+      "epoch": 79.85714285714286,
+      "grad_norm": 0.1201171875,
+      "learning_rate": 2.3858174816246837e-05,
+      "loss": 0.013,
+      "step": 1118
+    },
+    {
+      "epoch": 79.92857142857143,
+      "grad_norm": 0.11767578125,
+      "learning_rate": 2.385452823818414e-05,
+      "loss": 0.0133,
+      "step": 1119
+    },
+    {
+      "epoch": 80.0,
+      "grad_norm": 0.1298828125,
+      "learning_rate": 2.3850876126062314e-05,
+      "loss": 0.0125,
+      "step": 1120
+    },
+    {
+      "epoch": 80.07142857142857,
+      "grad_norm": 0.232421875,
+      "learning_rate": 2.384721848166136e-05,
+      "loss": 0.0136,
+      "step": 1121
+    },
+    {
+      "epoch": 80.14285714285714,
+      "grad_norm": 0.2275390625,
+      "learning_rate": 2.3843555306763963e-05,
+      "loss": 0.0144,
+      "step": 1122
+    },
+    {
+      "epoch": 80.21428571428571,
+      "grad_norm": 0.29296875,
+      "learning_rate": 2.3839886603155516e-05,
+      "loss": 0.0115,
+      "step": 1123
+    },
+    {
+      "epoch": 80.28571428571429,
+      "grad_norm": 0.2373046875,
+      "learning_rate": 2.3836212372624088e-05,
+      "loss": 0.0144,
+      "step": 1124
+    },
+    {
+      "epoch": 80.35714285714286,
+      "grad_norm": 0.23828125,
+      "learning_rate": 2.3832532616960473e-05,
+      "loss": 0.0162,
+      "step": 1125
+    },
+    {
+      "epoch": 80.42857142857143,
+      "grad_norm": 0.1611328125,
+      "learning_rate": 2.3828847337958126e-05,
+      "loss": 0.018,
+      "step": 1126
+    },
+    {
+      "epoch": 80.5,
+      "grad_norm": 0.1552734375,
+      "learning_rate": 2.3825156537413213e-05,
+      "loss": 0.0126,
+      "step": 1127
+    },
+    {
+      "epoch": 80.57142857142857,
+      "grad_norm": 0.150390625,
+      "learning_rate": 2.382146021712458e-05,
+      "loss": 0.0132,
+      "step": 1128
+    },
+    {
+      "epoch": 80.64285714285714,
+      "grad_norm": 0.169921875,
+      "learning_rate": 2.3817758378893774e-05,
+      "loss": 0.0124,
+      "step": 1129
+    },
+    {
+      "epoch": 80.71428571428571,
+      "grad_norm": 0.12890625,
+      "learning_rate": 2.3814051024525018e-05,
+      "loss": 0.0128,
+      "step": 1130
+    },
+    {
+      "epoch": 80.78571428571429,
+      "grad_norm": 0.1513671875,
+      "learning_rate": 2.3810338155825245e-05,
+      "loss": 0.0142,
+      "step": 1131
+    },
+    {
+      "epoch": 80.85714285714286,
+      "grad_norm": 0.306640625,
+      "learning_rate": 2.3806619774604053e-05,
+      "loss": 0.013,
+      "step": 1132
+    },
+    {
+      "epoch": 80.92857142857143,
+      "grad_norm": 0.2138671875,
+      "learning_rate": 2.3802895882673734e-05,
+      "loss": 0.0155,
+      "step": 1133
+    },
+    {
+      "epoch": 81.0,
+      "grad_norm": 0.1474609375,
+      "learning_rate": 2.3799166481849272e-05,
+      "loss": 0.014,
+      "step": 1134
+    },
+    {
+      "epoch": 81.07142857142857,
+      "grad_norm": 0.19921875,
+      "learning_rate": 2.3795431573948333e-05,
+      "loss": 0.0129,
+      "step": 1135
+    },
+    {
+      "epoch": 81.14285714285714,
+      "grad_norm": 0.181640625,
+      "learning_rate": 2.379169116079126e-05,
+      "loss": 0.0149,
+      "step": 1136
+    },
+    {
+      "epoch": 81.21428571428571,
+      "grad_norm": 0.2001953125,
+      "learning_rate": 2.3787945244201092e-05,
+      "loss": 0.015,
+      "step": 1137
+    },
+    {
+      "epoch": 81.28571428571429,
+      "grad_norm": 0.1328125,
+      "learning_rate": 2.3784193826003538e-05,
+      "loss": 0.0135,
+      "step": 1138
+    },
+    {
+      "epoch": 81.35714285714286,
+      "grad_norm": 0.1953125,
+      "learning_rate": 2.3780436908026992e-05,
+      "loss": 0.0144,
+      "step": 1139
+    },
+    {
+      "epoch": 81.42857142857143,
+      "grad_norm": 0.125,
+      "learning_rate": 2.3776674492102535e-05,
+      "loss": 0.0147,
+      "step": 1140
+    },
+    {
+      "epoch": 81.5,
+      "grad_norm": 0.384765625,
+      "learning_rate": 2.3772906580063924e-05,
+      "loss": 0.0164,
+      "step": 1141
+    },
+    {
+      "epoch": 81.57142857142857,
+      "grad_norm": 0.1279296875,
+      "learning_rate": 2.376913317374759e-05,
+      "loss": 0.0122,
+      "step": 1142
+    },
+    {
+      "epoch": 81.64285714285714,
+      "grad_norm": 0.1357421875,
+      "learning_rate": 2.3765354274992643e-05,
+      "loss": 0.0154,
+      "step": 1143
+    },
+    {
+      "epoch": 81.71428571428571,
+      "grad_norm": 0.1865234375,
+      "learning_rate": 2.3761569885640877e-05,
+      "loss": 0.015,
+      "step": 1144
+    },
+    {
+      "epoch": 81.78571428571429,
+      "grad_norm": 0.14453125,
+      "learning_rate": 2.3757780007536757e-05,
+      "loss": 0.0154,
+      "step": 1145
+    },
+    {
+      "epoch": 81.85714285714286,
+      "grad_norm": 0.244140625,
+      "learning_rate": 2.3753984642527423e-05,
+      "loss": 0.0124,
+      "step": 1146
+    },
+    {
+      "epoch": 81.92857142857143,
+      "grad_norm": 0.1591796875,
+      "learning_rate": 2.3750183792462686e-05,
+      "loss": 0.0145,
+      "step": 1147
+    },
+    {
+      "epoch": 82.0,
+      "grad_norm": 0.17578125,
+      "learning_rate": 2.3746377459195037e-05,
+      "loss": 0.0145,
+      "step": 1148
+    },
+    {
+      "epoch": 82.07142857142857,
+      "grad_norm": 0.169921875,
+      "learning_rate": 2.3742565644579638e-05,
+      "loss": 0.0161,
+      "step": 1149
+    },
+    {
+      "epoch": 82.14285714285714,
+      "grad_norm": 0.1728515625,
+      "learning_rate": 2.373874835047432e-05,
+      "loss": 0.0142,
+      "step": 1150
+    },
+    {
+      "epoch": 82.21428571428571,
+      "grad_norm": 0.1064453125,
+      "learning_rate": 2.3734925578739588e-05,
+      "loss": 0.0123,
+      "step": 1151
+    },
+    {
+      "epoch": 82.28571428571429,
+      "grad_norm": 0.1591796875,
+      "learning_rate": 2.373109733123861e-05,
+      "loss": 0.012,
+      "step": 1152
+    },
+    {
+      "epoch": 82.35714285714286,
+      "grad_norm": 0.1201171875,
+      "learning_rate": 2.372726360983723e-05,
+      "loss": 0.017,
+      "step": 1153
+    },
+    {
+      "epoch": 82.42857142857143,
+      "grad_norm": 0.30859375,
+      "learning_rate": 2.372342441640396e-05,
+      "loss": 0.0151,
+      "step": 1154
+    },
+    {
+      "epoch": 82.5,
+      "grad_norm": 0.140625,
+      "learning_rate": 2.371957975280997e-05,
+      "loss": 0.0103,
+      "step": 1155
+    },
+    {
+      "epoch": 82.57142857142857,
+      "grad_norm": 0.2373046875,
+      "learning_rate": 2.3715729620929106e-05,
+      "loss": 0.0119,
+      "step": 1156
+    },
+    {
+      "epoch": 82.64285714285714,
+      "grad_norm": 0.267578125,
+      "learning_rate": 2.3711874022637875e-05,
+      "loss": 0.0148,
+      "step": 1157
+    },
+    {
+      "epoch": 82.71428571428571,
+      "grad_norm": 0.1416015625,
+      "learning_rate": 2.3708012959815452e-05,
+      "loss": 0.0156,
+      "step": 1158
+    },
+    {
+      "epoch": 82.78571428571429,
+      "grad_norm": 0.1962890625,
+      "learning_rate": 2.3704146434343664e-05,
+      "loss": 0.0124,
+      "step": 1159
+    },
+    {
+      "epoch": 82.85714285714286,
+      "grad_norm": 0.1201171875,
+      "learning_rate": 2.3700274448107012e-05,
+      "loss": 0.015,
+      "step": 1160
+    },
+    {
+      "epoch": 82.92857142857143,
+      "grad_norm": 0.154296875,
+      "learning_rate": 2.3696397002992663e-05,
+      "loss": 0.0127,
+      "step": 1161
+    },
+    {
+      "epoch": 83.0,
+      "grad_norm": 0.2578125,
+      "learning_rate": 2.3692514100890425e-05,
+      "loss": 0.0149,
+      "step": 1162
+    },
+    {
+      "epoch": 83.07142857142857,
+      "grad_norm": 0.19921875,
+      "learning_rate": 2.3688625743692784e-05,
+      "loss": 0.0139,
+      "step": 1163
+    },
+    {
+      "epoch": 83.14285714285714,
+      "grad_norm": 0.1162109375,
+      "learning_rate": 2.368473193329488e-05,
+      "loss": 0.0128,
+      "step": 1164
+    },
+    {
+      "epoch": 83.21428571428571,
+      "grad_norm": 0.1650390625,
+      "learning_rate": 2.36808326715945e-05,
+      "loss": 0.011,
+      "step": 1165
+    },
+    {
+      "epoch": 83.28571428571429,
+      "grad_norm": 0.1455078125,
+      "learning_rate": 2.36769279604921e-05,
+      "loss": 0.0158,
+      "step": 1166
+    },
+    {
+      "epoch": 83.35714285714286,
+      "grad_norm": 0.2373046875,
+      "learning_rate": 2.36730178018908e-05,
+      "loss": 0.0147,
+      "step": 1167
+    },
+    {
+      "epoch": 83.42857142857143,
+      "grad_norm": 0.291015625,
+      "learning_rate": 2.3669102197696347e-05,
+      "loss": 0.0159,
+      "step": 1168
+    },
+    {
+      "epoch": 83.5,
+      "grad_norm": 0.25390625,
+      "learning_rate": 2.3665181149817173e-05,
+      "loss": 0.0104,
+      "step": 1169
+    },
+    {
+      "epoch": 83.57142857142857,
+      "grad_norm": 0.1298828125,
+      "learning_rate": 2.3661254660164337e-05,
+      "loss": 0.0143,
+      "step": 1170
+    },
+    {
+      "epoch": 83.64285714285714,
+      "grad_norm": 0.173828125,
+      "learning_rate": 2.365732273065157e-05,
+      "loss": 0.0152,
+      "step": 1171
+    },
+    {
+      "epoch": 83.71428571428571,
+      "grad_norm": 0.146484375,
+      "learning_rate": 2.3653385363195248e-05,
+      "loss": 0.0174,
+      "step": 1172
+    },
+    {
+      "epoch": 83.78571428571429,
+      "grad_norm": 0.2236328125,
+      "learning_rate": 2.3649442559714387e-05,
+      "loss": 0.0145,
+      "step": 1173
+    },
+    {
+      "epoch": 83.85714285714286,
+      "grad_norm": 0.31640625,
+      "learning_rate": 2.3645494322130668e-05,
+      "loss": 0.0155,
+      "step": 1174
+    },
+    {
+      "epoch": 83.92857142857143,
+      "grad_norm": 0.1376953125,
+      "learning_rate": 2.3641540652368415e-05,
+      "loss": 0.0127,
+      "step": 1175
+    },
+    {
+      "epoch": 84.0,
+      "grad_norm": 0.23046875,
+      "learning_rate": 2.36375815523546e-05,
+      "loss": 0.0142,
+      "step": 1176
+    },
+    {
+      "epoch": 84.07142857142857,
+      "grad_norm": 0.1611328125,
+      "learning_rate": 2.3633617024018832e-05,
+      "loss": 0.0134,
+      "step": 1177
+    },
+    {
+      "epoch": 84.14285714285714,
+      "grad_norm": 0.13671875,
+      "learning_rate": 2.3629647069293384e-05,
+      "loss": 0.0129,
+      "step": 1178
+    },
+    {
+      "epoch": 84.21428571428571,
+      "grad_norm": 0.138671875,
+      "learning_rate": 2.3625671690113155e-05,
+      "loss": 0.0126,
+      "step": 1179
+    },
+    {
+      "epoch": 84.28571428571429,
+      "grad_norm": 0.130859375,
+      "learning_rate": 2.3621690888415705e-05,
+      "loss": 0.0135,
+      "step": 1180
+    },
+    {
+      "epoch": 84.35714285714286,
+      "grad_norm": 0.2197265625,
+      "learning_rate": 2.361770466614122e-05,
+      "loss": 0.0159,
+      "step": 1181
+    },
+    {
+      "epoch": 84.42857142857143,
+      "grad_norm": 0.1240234375,
+      "learning_rate": 2.3613713025232542e-05,
+      "loss": 0.0136,
+      "step": 1182
+    },
+    {
+      "epoch": 84.5,
+      "grad_norm": 0.1875,
+      "learning_rate": 2.3609715967635145e-05,
+      "loss": 0.0173,
+      "step": 1183
+    },
+    {
+      "epoch": 84.57142857142857,
+      "grad_norm": 0.111328125,
+      "learning_rate": 2.3605713495297148e-05,
+      "loss": 0.0129,
+      "step": 1184
+    },
+    {
+      "epoch": 84.64285714285714,
+      "grad_norm": 0.2158203125,
+      "learning_rate": 2.360170561016931e-05,
+      "loss": 0.0158,
+      "step": 1185
+    },
+    {
+      "epoch": 84.71428571428571,
+      "grad_norm": 0.1669921875,
+      "learning_rate": 2.3597692314205016e-05,
+      "loss": 0.0135,
+      "step": 1186
+    },
+    {
+      "epoch": 84.78571428571429,
+      "grad_norm": 0.1875,
+      "learning_rate": 2.359367360936031e-05,
+      "loss": 0.0132,
+      "step": 1187
+    },
+    {
+      "epoch": 84.85714285714286,
+      "grad_norm": 0.169921875,
+      "learning_rate": 2.358964949759385e-05,
+      "loss": 0.0139,
+      "step": 1188
+    },
+    {
+      "epoch": 84.92857142857143,
+      "grad_norm": 0.35546875,
+      "learning_rate": 2.358561998086695e-05,
+      "loss": 0.0134,
+      "step": 1189
+    },
+    {
+      "epoch": 85.0,
+      "grad_norm": 0.150390625,
+      "learning_rate": 2.3581585061143534e-05,
+      "loss": 0.018,
+      "step": 1190
+    },
+    {
+      "epoch": 85.07142857142857,
+      "grad_norm": 0.1728515625,
+      "learning_rate": 2.3577544740390184e-05,
+      "loss": 0.0172,
+      "step": 1191
+    },
+    {
+      "epoch": 85.14285714285714,
+      "grad_norm": 0.263671875,
+      "learning_rate": 2.3573499020576102e-05,
+      "loss": 0.0125,
+      "step": 1192
+    },
+    {
+      "epoch": 85.21428571428571,
+      "grad_norm": 0.17578125,
+      "learning_rate": 2.3569447903673116e-05,
+      "loss": 0.0121,
+      "step": 1193
+    },
+    {
+      "epoch": 85.28571428571429,
+      "grad_norm": 0.169921875,
+      "learning_rate": 2.356539139165569e-05,
+      "loss": 0.0145,
+      "step": 1194
+    },
+    {
+      "epoch": 85.35714285714286,
+      "grad_norm": 0.2177734375,
+      "learning_rate": 2.356132948650093e-05,
+      "loss": 0.015,
+      "step": 1195
+    },
+    {
+      "epoch": 85.42857142857143,
+      "grad_norm": 0.25,
+      "learning_rate": 2.355726219018855e-05,
+      "loss": 0.0153,
+      "step": 1196
+    },
+    {
+      "epoch": 85.5,
+      "grad_norm": 0.2890625,
+      "learning_rate": 2.3553189504700905e-05,
+      "loss": 0.0155,
+      "step": 1197
+    },
+    {
+      "epoch": 85.57142857142857,
+      "grad_norm": 0.2080078125,
+      "learning_rate": 2.3549111432022973e-05,
+      "loss": 0.0121,
+      "step": 1198
+    },
+    {
+      "epoch": 85.64285714285714,
+      "grad_norm": 0.279296875,
+      "learning_rate": 2.3545027974142353e-05,
+      "loss": 0.0172,
+      "step": 1199
+    },
+    {
+      "epoch": 85.71428571428571,
+      "grad_norm": 0.1181640625,
+      "learning_rate": 2.3540939133049273e-05,
+      "loss": 0.0128,
+      "step": 1200
+    },
+    {
+      "epoch": 85.78571428571429,
+      "grad_norm": 0.130859375,
+      "learning_rate": 2.353684491073659e-05,
+      "loss": 0.0137,
+      "step": 1201
+    },
+    {
+      "epoch": 85.85714285714286,
+      "grad_norm": 0.2080078125,
+      "learning_rate": 2.353274530919977e-05,
+      "loss": 0.013,
+      "step": 1202
+    },
+    {
+      "epoch": 85.92857142857143,
+      "grad_norm": 0.15625,
+      "learning_rate": 2.3528640330436917e-05,
+      "loss": 0.0182,
+      "step": 1203
+    },
+    {
+      "epoch": 86.0,
+      "grad_norm": 0.1455078125,
+      "learning_rate": 2.3524529976448744e-05,
+      "loss": 0.0139,
+      "step": 1204
+    },
+    {
+      "epoch": 86.07142857142857,
+      "grad_norm": 0.140625,
+      "learning_rate": 2.352041424923859e-05,
+      "loss": 0.0155,
+      "step": 1205
+    },
+    {
+      "epoch": 86.14285714285714,
+      "grad_norm": 0.154296875,
+      "learning_rate": 2.35162931508124e-05,
+      "loss": 0.012,
+      "step": 1206
+    },
+    {
+      "epoch": 86.21428571428571,
+      "grad_norm": 0.1875,
+      "learning_rate": 2.3512166683178766e-05,
+      "loss": 0.0135,
+      "step": 1207
+    },
+    {
+      "epoch": 86.28571428571429,
+      "grad_norm": 0.126953125,
+      "learning_rate": 2.3508034848348866e-05,
+      "loss": 0.0154,
+      "step": 1208
+    },
+    {
+      "epoch": 86.35714285714286,
+      "grad_norm": 0.169921875,
+      "learning_rate": 2.3503897648336503e-05,
+      "loss": 0.0135,
+      "step": 1209
+    },
+    {
+      "epoch": 86.42857142857143,
+      "grad_norm": 0.10791015625,
+      "learning_rate": 2.3499755085158106e-05,
+      "loss": 0.0131,
+      "step": 1210
+    },
+    {
+      "epoch": 86.5,
+      "grad_norm": 0.2470703125,
+      "learning_rate": 2.3495607160832707e-05,
+      "loss": 0.0155,
+      "step": 1211
+    },
+    {
+      "epoch": 86.57142857142857,
+      "grad_norm": 0.3125,
+      "learning_rate": 2.3491453877381953e-05,
+      "loss": 0.0133,
+      "step": 1212
+    },
+    {
+      "epoch": 86.64285714285714,
+      "grad_norm": 0.1787109375,
+      "learning_rate": 2.3487295236830108e-05,
+      "loss": 0.0133,
+      "step": 1213
+    },
+    {
+      "epoch": 86.71428571428571,
+      "grad_norm": 0.1513671875,
+      "learning_rate": 2.3483131241204037e-05,
+      "loss": 0.0141,
+      "step": 1214
+    },
+    {
+      "epoch": 86.78571428571429,
+      "grad_norm": 0.34765625,
+      "learning_rate": 2.3478961892533223e-05,
+      "loss": 0.0137,
+      "step": 1215
+    },
+    {
+      "epoch": 86.85714285714286,
+      "grad_norm": 0.1396484375,
+      "learning_rate": 2.3474787192849756e-05,
+      "loss": 0.0142,
+      "step": 1216
+    },
+    {
+      "epoch": 86.92857142857143,
+      "grad_norm": 0.15625,
+      "learning_rate": 2.347060714418834e-05,
+      "loss": 0.0141,
+      "step": 1217
+    },
+    {
+      "epoch": 87.0,
+      "grad_norm": 0.142578125,
+      "learning_rate": 2.346642174858627e-05,
+      "loss": 0.0121,
+      "step": 1218
+    },
+    {
+      "epoch": 87.07142857142857,
+      "grad_norm": 0.291015625,
+      "learning_rate": 2.346223100808346e-05,
+      "loss": 0.0139,
+      "step": 1219
+    },
+    {
+      "epoch": 87.14285714285714,
+      "grad_norm": 0.1787109375,
+      "learning_rate": 2.345803492472243e-05,
+      "loss": 0.0153,
+      "step": 1220
+    },
+    {
+      "epoch": 87.21428571428571,
+      "grad_norm": 0.353515625,
+      "learning_rate": 2.3453833500548295e-05,
+      "loss": 0.0128,
+      "step": 1221
+    },
+    {
+      "epoch": 87.28571428571429,
+      "grad_norm": 0.1728515625,
+      "learning_rate": 2.3449626737608784e-05,
+      "loss": 0.0124,
+      "step": 1222
+    },
+    {
+      "epoch": 87.35714285714286,
+      "grad_norm": 0.1337890625,
+      "learning_rate": 2.3445414637954217e-05,
+      "loss": 0.0143,
+      "step": 1223
+    },
+    {
+      "epoch": 87.42857142857143,
+      "grad_norm": 0.220703125,
+      "learning_rate": 2.3441197203637522e-05,
+      "loss": 0.0142,
+      "step": 1224
+    },
+    {
+      "epoch": 87.5,
+      "grad_norm": 0.1533203125,
+      "learning_rate": 2.3436974436714224e-05,
+      "loss": 0.0134,
+      "step": 1225
+    },
+    {
+      "epoch": 87.57142857142857,
+      "grad_norm": 0.255859375,
+      "learning_rate": 2.3432746339242448e-05,
+      "loss": 0.0162,
+      "step": 1226
+    },
+    {
+      "epoch": 87.64285714285714,
+      "grad_norm": 0.11279296875,
+      "learning_rate": 2.342851291328292e-05,
+      "loss": 0.0106,
+      "step": 1227
+    },
+    {
+      "epoch": 87.71428571428571,
+      "grad_norm": 0.2392578125,
+      "learning_rate": 2.342427416089896e-05,
+      "loss": 0.0141,
+      "step": 1228
+    },
+    {
+      "epoch": 87.78571428571429,
+      "grad_norm": 0.19921875,
+      "learning_rate": 2.3420030084156486e-05,
+      "loss": 0.0168,
+      "step": 1229
+    },
+    {
+      "epoch": 87.85714285714286,
+      "grad_norm": 0.14453125,
+      "learning_rate": 2.3415780685124004e-05,
+      "loss": 0.0169,
+      "step": 1230
+    },
+    {
+      "epoch": 87.92857142857143,
+      "grad_norm": 0.15234375,
+      "learning_rate": 2.341152596587262e-05,
+      "loss": 0.0189,
+      "step": 1231
+    },
+    {
+      "epoch": 88.0,
+      "grad_norm": 0.1708984375,
+      "learning_rate": 2.3407265928476042e-05,
+      "loss": 0.0135,
+      "step": 1232
+    },
+    {
+      "epoch": 88.07142857142857,
+      "grad_norm": 0.267578125,
+      "learning_rate": 2.340300057501055e-05,
+      "loss": 0.0128,
+      "step": 1233
+    },
+    {
+      "epoch": 88.14285714285714,
+      "grad_norm": 0.1748046875,
+      "learning_rate": 2.3398729907555026e-05,
+      "loss": 0.0148,
+      "step": 1234
+    },
+    {
+      "epoch": 88.21428571428571,
+      "grad_norm": 0.251953125,
+      "learning_rate": 2.3394453928190947e-05,
+      "loss": 0.0151,
+      "step": 1235
+    },
+    {
+      "epoch": 88.28571428571429,
+      "grad_norm": 0.1337890625,
+      "learning_rate": 2.339017263900237e-05,
+      "loss": 0.0149,
+      "step": 1236
+    },
+    {
+      "epoch": 88.35714285714286,
+      "grad_norm": 0.15234375,
+      "learning_rate": 2.3385886042075942e-05,
+      "loss": 0.0148,
+      "step": 1237
+    },
+    {
+      "epoch": 88.42857142857143,
+      "grad_norm": 0.29296875,
+      "learning_rate": 2.3381594139500902e-05,
+      "loss": 0.0173,
+      "step": 1238
+    },
+    {
+      "epoch": 88.5,
+      "grad_norm": 0.240234375,
+      "learning_rate": 2.3377296933369074e-05,
+      "loss": 0.0155,
+      "step": 1239
+    },
+    {
+      "epoch": 88.57142857142857,
+      "grad_norm": 0.1376953125,
+      "learning_rate": 2.337299442577485e-05,
+      "loss": 0.0138,
+      "step": 1240
+    },
+    {
+      "epoch": 88.64285714285714,
+      "grad_norm": 0.1005859375,
+      "learning_rate": 2.3368686618815238e-05,
+      "loss": 0.0147,
+      "step": 1241
+    },
+    {
+      "epoch": 88.71428571428571,
+      "grad_norm": 0.306640625,
+      "learning_rate": 2.33643735145898e-05,
+      "loss": 0.0129,
+      "step": 1242
+    },
+    {
+      "epoch": 88.78571428571429,
+      "grad_norm": 0.18359375,
+      "learning_rate": 2.336005511520069e-05,
+      "loss": 0.0121,
+      "step": 1243
+    },
+    {
+      "epoch": 88.85714285714286,
+      "grad_norm": 0.37109375,
+      "learning_rate": 2.3355731422752652e-05,
+      "loss": 0.0146,
+      "step": 1244
+    },
+    {
+      "epoch": 88.92857142857143,
+      "grad_norm": 0.294921875,
+      "learning_rate": 2.335140243935299e-05,
+      "loss": 0.0164,
+      "step": 1245
+    },
+    {
+      "epoch": 89.0,
+      "grad_norm": 0.32421875,
+      "learning_rate": 2.334706816711161e-05,
+      "loss": 0.0169,
+      "step": 1246
+    },
+    {
+      "epoch": 89.07142857142857,
+      "grad_norm": 0.111328125,
+      "learning_rate": 2.3342728608140968e-05,
+      "loss": 0.0122,
+      "step": 1247
+    },
+    {
+      "epoch": 89.14285714285714,
+      "grad_norm": 0.1611328125,
+      "learning_rate": 2.3338383764556126e-05,
+      "loss": 0.012,
+      "step": 1248
+    },
+    {
+      "epoch": 89.21428571428571,
+      "grad_norm": 0.11083984375,
+      "learning_rate": 2.3334033638474707e-05,
+      "loss": 0.0146,
+      "step": 1249
+    },
+    {
+      "epoch": 89.28571428571429,
+      "grad_norm": 0.23046875,
+      "learning_rate": 2.3329678232016898e-05,
+      "loss": 0.0157,
+      "step": 1250
+    },
+    {
+      "epoch": 89.35714285714286,
+      "grad_norm": 0.1650390625,
+      "learning_rate": 2.3325317547305485e-05,
+      "loss": 0.0146,
+      "step": 1251
+    },
+    {
+      "epoch": 89.42857142857143,
+      "grad_norm": 0.26953125,
+      "learning_rate": 2.3320951586465804e-05,
+      "loss": 0.0143,
+      "step": 1252
+    },
+    {
+      "epoch": 89.5,
+      "grad_norm": 0.125,
+      "learning_rate": 2.3316580351625777e-05,
+      "loss": 0.013,
+      "step": 1253
+    },
+    {
+      "epoch": 89.57142857142857,
+      "grad_norm": 0.18359375,
+      "learning_rate": 2.3312203844915885e-05,
+      "loss": 0.0122,
+      "step": 1254
+    },
+    {
+      "epoch": 89.64285714285714,
+      "grad_norm": 0.125,
+      "learning_rate": 2.3307822068469187e-05,
+      "loss": 0.0135,
+      "step": 1255
+    },
+    {
+      "epoch": 89.71428571428571,
+      "grad_norm": 0.140625,
+      "learning_rate": 2.3303435024421312e-05,
+      "loss": 0.0151,
+      "step": 1256
+    },
+    {
+      "epoch": 89.78571428571429,
+      "grad_norm": 0.2890625,
+      "learning_rate": 2.329904271491044e-05,
+      "loss": 0.0132,
+      "step": 1257
+    },
+    {
+      "epoch": 89.85714285714286,
+      "grad_norm": 0.11279296875,
+      "learning_rate": 2.329464514207734e-05,
+      "loss": 0.0141,
+      "step": 1258
+    },
+    {
+      "epoch": 89.92857142857143,
+      "grad_norm": 0.1767578125,
+      "learning_rate": 2.329024230806534e-05,
+      "loss": 0.0136,
+      "step": 1259
+    },
+    {
+      "epoch": 90.0,
+      "grad_norm": 0.34375,
+      "learning_rate": 2.3285834215020313e-05,
+      "loss": 0.0162,
+      "step": 1260
+    },
+    {
+      "epoch": 90.07142857142857,
+      "grad_norm": 0.5390625,
+      "learning_rate": 2.328142086509072e-05,
+      "loss": 0.0164,
+      "step": 1261
+    },
+    {
+      "epoch": 90.14285714285714,
+      "grad_norm": 0.294921875,
+      "learning_rate": 2.327700226042757e-05,
+      "loss": 0.0141,
+      "step": 1262
+    },
+    {
+      "epoch": 90.21428571428571,
+      "grad_norm": 0.12890625,
+      "learning_rate": 2.327257840318444e-05,
+      "loss": 0.0121,
+      "step": 1263
+    },
+    {
+      "epoch": 90.28571428571429,
+      "grad_norm": 0.1171875,
+      "learning_rate": 2.326814929551746e-05,
+      "loss": 0.013,
+      "step": 1264
+    },
+    {
+      "epoch": 90.35714285714286,
+      "grad_norm": 0.12109375,
+      "learning_rate": 2.326371493958533e-05,
+      "loss": 0.0142,
+      "step": 1265
+    },
+    {
+      "epoch": 90.42857142857143,
+      "grad_norm": 0.2177734375,
+      "learning_rate": 2.32592753375493e-05,
+      "loss": 0.0128,
+      "step": 1266
+    },
+    {
+      "epoch": 90.5,
+      "grad_norm": 0.283203125,
+      "learning_rate": 2.3254830491573167e-05,
+      "loss": 0.0153,
+      "step": 1267
+    },
+    {
+      "epoch": 90.57142857142857,
+      "grad_norm": 0.322265625,
+      "learning_rate": 2.325038040382331e-05,
+      "loss": 0.0137,
+      "step": 1268
+    },
+    {
+      "epoch": 90.64285714285714,
+      "grad_norm": 0.23828125,
+      "learning_rate": 2.324592507646864e-05,
+      "loss": 0.0143,
+      "step": 1269
+    },
+    {
+      "epoch": 90.71428571428571,
+      "grad_norm": 0.23046875,
+      "learning_rate": 2.324146451168063e-05,
+      "loss": 0.0119,
+      "step": 1270
+    },
+    {
+      "epoch": 90.78571428571429,
+      "grad_norm": 0.11669921875,
+      "learning_rate": 2.3236998711633307e-05,
+      "loss": 0.0144,
+      "step": 1271
+    },
+    {
+      "epoch": 90.85714285714286,
+      "grad_norm": 0.263671875,
+      "learning_rate": 2.3232527678503246e-05,
+      "loss": 0.014,
+      "step": 1272
+    },
+    {
+      "epoch": 90.92857142857143,
+      "grad_norm": 0.431640625,
+      "learning_rate": 2.3228051414469574e-05,
+      "loss": 0.0166,
+      "step": 1273
+    },
+    {
+      "epoch": 91.0,
+      "grad_norm": 0.140625,
+      "learning_rate": 2.3223569921713966e-05,
+      "loss": 0.0109,
+      "step": 1274
+    },
+    {
+      "epoch": 91.07142857142857,
+      "grad_norm": 0.1640625,
+      "learning_rate": 2.3219083202420654e-05,
+      "loss": 0.0127,
+      "step": 1275
+    },
+    {
+      "epoch": 91.14285714285714,
+      "grad_norm": 0.1982421875,
+      "learning_rate": 2.3214591258776404e-05,
+      "loss": 0.0151,
+      "step": 1276
+    },
+    {
+      "epoch": 91.21428571428571,
+      "grad_norm": 0.12158203125,
+      "learning_rate": 2.3210094092970536e-05,
+      "loss": 0.0148,
+      "step": 1277
+    },
+    {
+      "epoch": 91.28571428571429,
+      "grad_norm": 0.3671875,
+      "learning_rate": 2.3205591707194918e-05,
+      "loss": 0.0161,
+      "step": 1278
+    },
+    {
+      "epoch": 91.35714285714286,
+      "grad_norm": 0.203125,
+      "learning_rate": 2.320108410364395e-05,
+      "loss": 0.0164,
+      "step": 1279
+    },
+    {
+      "epoch": 91.42857142857143,
+      "grad_norm": 0.18359375,
+      "learning_rate": 2.3196571284514594e-05,
+      "loss": 0.0141,
+      "step": 1280
+    },
+    {
+      "epoch": 91.5,
+      "grad_norm": 0.1455078125,
+      "learning_rate": 2.3192053252006335e-05,
+      "loss": 0.0149,
+      "step": 1281
+    },
+    {
+      "epoch": 91.57142857142857,
+      "grad_norm": 0.1328125,
+      "learning_rate": 2.3187530008321208e-05,
+      "loss": 0.0132,
+      "step": 1282
+    },
+    {
+      "epoch": 91.64285714285714,
+      "grad_norm": 0.11083984375,
+      "learning_rate": 2.318300155566379e-05,
+      "loss": 0.0138,
+      "step": 1283
+    },
+    {
+      "epoch": 91.71428571428571,
+      "grad_norm": 0.12451171875,
+      "learning_rate": 2.317846789624119e-05,
+      "loss": 0.0127,
+      "step": 1284
+    },
+    {
+      "epoch": 91.78571428571429,
+      "grad_norm": 0.2080078125,
+      "learning_rate": 2.3173929032263063e-05,
+      "loss": 0.0132,
+      "step": 1285
+    },
+    {
+      "epoch": 91.85714285714286,
+      "grad_norm": 0.181640625,
+      "learning_rate": 2.3169384965941592e-05,
+      "loss": 0.0134,
+      "step": 1286
+    },
+    {
+      "epoch": 91.92857142857143,
+      "grad_norm": 0.11767578125,
+      "learning_rate": 2.3164835699491505e-05,
+      "loss": 0.0139,
+      "step": 1287
+    },
+    {
+      "epoch": 92.0,
+      "grad_norm": 0.15625,
+      "learning_rate": 2.3160281235130055e-05,
+      "loss": 0.0148,
+      "step": 1288
+    },
+    {
+      "epoch": 92.07142857142857,
+      "grad_norm": 0.10888671875,
+      "learning_rate": 2.3155721575077034e-05,
+      "loss": 0.0116,
+      "step": 1289
+    },
+    {
+      "epoch": 92.14285714285714,
+      "grad_norm": 0.12109375,
+      "learning_rate": 2.3151156721554763e-05,
+      "loss": 0.013,
+      "step": 1290
+    },
+    {
+      "epoch": 92.21428571428571,
+      "grad_norm": 0.1376953125,
+      "learning_rate": 2.3146586676788095e-05,
+      "loss": 0.0118,
+      "step": 1291
+    },
+    {
+      "epoch": 92.28571428571429,
+      "grad_norm": 0.1923828125,
+      "learning_rate": 2.3142011443004418e-05,
+      "loss": 0.0138,
+      "step": 1292
+    },
+    {
+      "epoch": 92.35714285714286,
+      "grad_norm": 0.310546875,
+      "learning_rate": 2.3137431022433652e-05,
+      "loss": 0.0153,
+      "step": 1293
+    },
+    {
+      "epoch": 92.42857142857143,
+      "grad_norm": 0.1376953125,
+      "learning_rate": 2.3132845417308223e-05,
+      "loss": 0.0164,
+      "step": 1294
+    },
+    {
+      "epoch": 92.5,
+      "grad_norm": 0.185546875,
+      "learning_rate": 2.312825462986311e-05,
+      "loss": 0.0137,
+      "step": 1295
+    },
+    {
+      "epoch": 92.57142857142857,
+      "grad_norm": 0.1455078125,
+      "learning_rate": 2.3123658662335802e-05,
+      "loss": 0.0174,
+      "step": 1296
+    },
+    {
+      "epoch": 92.64285714285714,
+      "grad_norm": 0.1435546875,
+      "learning_rate": 2.3119057516966318e-05,
+      "loss": 0.0143,
+      "step": 1297
+    },
+    {
+      "epoch": 92.71428571428571,
+      "grad_norm": 0.287109375,
+      "learning_rate": 2.3114451195997206e-05,
+      "loss": 0.0156,
+      "step": 1298
+    },
+    {
+      "epoch": 92.78571428571429,
+      "grad_norm": 0.171875,
+      "learning_rate": 2.3109839701673525e-05,
+      "loss": 0.0145,
+      "step": 1299
+    },
+    {
+      "epoch": 92.85714285714286,
+      "grad_norm": 0.11962890625,
+      "learning_rate": 2.3105223036242863e-05,
+      "loss": 0.0146,
+      "step": 1300
+    },
+    {
+      "epoch": 92.92857142857143,
+      "grad_norm": 0.12255859375,
+      "learning_rate": 2.3100601201955324e-05,
+      "loss": 0.0126,
+      "step": 1301
+    },
+    {
+      "epoch": 93.0,
+      "grad_norm": 0.2490234375,
+      "learning_rate": 2.309597420106354e-05,
+      "loss": 0.0117,
+      "step": 1302
+    },
+    {
+      "epoch": 93.07142857142857,
+      "grad_norm": 0.1943359375,
+      "learning_rate": 2.309134203582265e-05,
+      "loss": 0.0148,
+      "step": 1303
+    },
+    {
+      "epoch": 93.14285714285714,
+      "grad_norm": 0.24609375,
+      "learning_rate": 2.308670470849032e-05,
+      "loss": 0.0126,
+      "step": 1304
+    },
+    {
+      "epoch": 93.21428571428571,
+      "grad_norm": 0.12109375,
+      "learning_rate": 2.3082062221326724e-05,
+      "loss": 0.012,
+      "step": 1305
+    },
+    {
+      "epoch": 93.28571428571429,
+      "grad_norm": 0.11328125,
+      "learning_rate": 2.3077414576594553e-05,
+      "loss": 0.0147,
+      "step": 1306
+    },
+    {
+      "epoch": 93.35714285714286,
+      "grad_norm": 0.1484375,
+      "learning_rate": 2.307276177655901e-05,
+      "loss": 0.0152,
+      "step": 1307
+    },
+    {
+      "epoch": 93.42857142857143,
+      "grad_norm": 0.1318359375,
+      "learning_rate": 2.306810382348782e-05,
+      "loss": 0.0156,
+      "step": 1308
+    },
+    {
+      "epoch": 93.5,
+      "grad_norm": 0.302734375,
+      "learning_rate": 2.306344071965121e-05,
+      "loss": 0.0141,
+      "step": 1309
+    },
+    {
+      "epoch": 93.57142857142857,
+      "grad_norm": 0.11279296875,
+      "learning_rate": 2.3058772467321914e-05,
+      "loss": 0.0121,
+      "step": 1310
+    },
+    {
+      "epoch": 93.64285714285714,
+      "grad_norm": 0.1123046875,
+      "learning_rate": 2.305409906877519e-05,
+      "loss": 0.0153,
+      "step": 1311
+    },
+    {
+      "epoch": 93.71428571428571,
+      "grad_norm": 0.251953125,
+      "learning_rate": 2.3049420526288792e-05,
+      "loss": 0.0103,
+      "step": 1312
+    },
+    {
+      "epoch": 93.78571428571429,
+      "grad_norm": 0.240234375,
+      "learning_rate": 2.3044736842142987e-05,
+      "loss": 0.0141,
+      "step": 1313
+    },
+    {
+      "epoch": 93.85714285714286,
+      "grad_norm": 0.166015625,
+      "learning_rate": 2.3040048018620543e-05,
+      "loss": 0.0135,
+      "step": 1314
+    },
+    {
+      "epoch": 93.92857142857143,
+      "grad_norm": 0.12060546875,
+      "learning_rate": 2.3035354058006738e-05,
+      "loss": 0.013,
+      "step": 1315
+    },
+    {
+      "epoch": 94.0,
+      "grad_norm": 0.1962890625,
+      "learning_rate": 2.3030654962589346e-05,
+      "loss": 0.0174,
+      "step": 1316
+    },
+    {
+      "epoch": 94.07142857142857,
+      "grad_norm": 0.09912109375,
+      "learning_rate": 2.3025950734658654e-05,
+      "loss": 0.0124,
+      "step": 1317
+    },
+    {
+      "epoch": 94.14285714285714,
+      "grad_norm": 0.1357421875,
+      "learning_rate": 2.3021241376507447e-05,
+      "loss": 0.0141,
+      "step": 1318
+    },
+    {
+      "epoch": 94.21428571428571,
+      "grad_norm": 0.12353515625,
+      "learning_rate": 2.3016526890431006e-05,
+      "loss": 0.0133,
+      "step": 1319
+    },
+    {
+      "epoch": 94.28571428571429,
+      "grad_norm": 0.130859375,
+      "learning_rate": 2.3011807278727116e-05,
+      "loss": 0.0132,
+      "step": 1320
+    },
+    {
+      "epoch": 94.35714285714286,
+      "grad_norm": 0.154296875,
+      "learning_rate": 2.3007082543696055e-05,
+      "loss": 0.0156,
+      "step": 1321
+    },
+    {
+      "epoch": 94.42857142857143,
+      "grad_norm": 0.236328125,
+      "learning_rate": 2.3002352687640605e-05,
+      "loss": 0.0138,
+      "step": 1322
+    },
+    {
+      "epoch": 94.5,
+      "grad_norm": 0.291015625,
+      "learning_rate": 2.2997617712866037e-05,
+      "loss": 0.0174,
+      "step": 1323
+    },
+    {
+      "epoch": 94.57142857142857,
+      "grad_norm": 0.1875,
+      "learning_rate": 2.2992877621680125e-05,
+      "loss": 0.0142,
+      "step": 1324
+    },
+    {
+      "epoch": 94.64285714285714,
+      "grad_norm": 0.1318359375,
+      "learning_rate": 2.2988132416393125e-05,
+      "loss": 0.0144,
+      "step": 1325
+    },
+    {
+      "epoch": 94.71428571428571,
+      "grad_norm": 0.134765625,
+      "learning_rate": 2.2983382099317803e-05,
+      "loss": 0.0134,
+      "step": 1326
+    },
+    {
+      "epoch": 94.78571428571429,
+      "grad_norm": 0.15234375,
+      "learning_rate": 2.2978626672769395e-05,
+      "loss": 0.0164,
+      "step": 1327
+    },
+    {
+      "epoch": 94.85714285714286,
+      "grad_norm": 0.3203125,
+      "learning_rate": 2.2973866139065643e-05,
+      "loss": 0.0135,
+      "step": 1328
+    },
+    {
+      "epoch": 94.92857142857143,
+      "grad_norm": 0.1328125,
+      "learning_rate": 2.2969100500526775e-05,
+      "loss": 0.0179,
+      "step": 1329
+    },
+    {
+      "epoch": 95.0,
+      "grad_norm": 0.251953125,
+      "learning_rate": 2.29643297594755e-05,
+      "loss": 0.0119,
+      "step": 1330
+    },
+    {
+      "epoch": 95.07142857142857,
+      "grad_norm": 0.10400390625,
+      "learning_rate": 2.295955391823702e-05,
+      "loss": 0.013,
+      "step": 1331
+    },
+    {
+      "epoch": 95.14285714285714,
+      "grad_norm": 0.1787109375,
+      "learning_rate": 2.295477297913903e-05,
+      "loss": 0.0132,
+      "step": 1332
+    },
+    {
+      "epoch": 95.21428571428571,
+      "grad_norm": 0.2001953125,
+      "learning_rate": 2.294998694451169e-05,
+      "loss": 0.015,
+      "step": 1333
+    },
+    {
+      "epoch": 95.28571428571429,
+      "grad_norm": 0.09814453125,
+      "learning_rate": 2.294519581668767e-05,
+      "loss": 0.012,
+      "step": 1334
+    },
+    {
+      "epoch": 95.35714285714286,
+      "grad_norm": 0.1298828125,
+      "learning_rate": 2.2940399598002095e-05,
+      "loss": 0.0133,
+      "step": 1335
+    },
+    {
+      "epoch": 95.42857142857143,
+      "grad_norm": 0.11767578125,
+      "learning_rate": 2.2935598290792583e-05,
+      "loss": 0.0152,
+      "step": 1336
+    },
+    {
+      "epoch": 95.5,
+      "grad_norm": 0.1181640625,
+      "learning_rate": 2.293079189739924e-05,
+      "loss": 0.0117,
+      "step": 1337
+    },
+    {
+      "epoch": 95.57142857142857,
+      "grad_norm": 0.205078125,
+      "learning_rate": 2.292598042016464e-05,
+      "loss": 0.013,
+      "step": 1338
+    },
+    {
+      "epoch": 95.64285714285714,
+      "grad_norm": 0.29296875,
+      "learning_rate": 2.292116386143384e-05,
+      "loss": 0.0156,
+      "step": 1339
+    },
+    {
+      "epoch": 95.71428571428571,
+      "grad_norm": 0.21484375,
+      "learning_rate": 2.2916342223554374e-05,
+      "loss": 0.0138,
+      "step": 1340
+    },
+    {
+      "epoch": 95.78571428571429,
+      "grad_norm": 0.2255859375,
+      "learning_rate": 2.2911515508876243e-05,
+      "loss": 0.0161,
+      "step": 1341
+    },
+    {
+      "epoch": 95.85714285714286,
+      "grad_norm": 0.09375,
+      "learning_rate": 2.2906683719751938e-05,
+      "loss": 0.012,
+      "step": 1342
+    },
+    {
+      "epoch": 95.92857142857143,
+      "grad_norm": 0.11328125,
+      "learning_rate": 2.2901846858536408e-05,
+      "loss": 0.0128,
+      "step": 1343
+    },
+    {
+      "epoch": 96.0,
+      "grad_norm": 0.10693359375,
+      "learning_rate": 2.289700492758708e-05,
+      "loss": 0.0145,
+      "step": 1344
+    },
+    {
+      "epoch": 96.07142857142857,
+      "grad_norm": 0.12890625,
+      "learning_rate": 2.2892157929263858e-05,
+      "loss": 0.0144,
+      "step": 1345
+    },
+    {
+      "epoch": 96.14285714285714,
+      "grad_norm": 0.1337890625,
+      "learning_rate": 2.2887305865929104e-05,
+      "loss": 0.0147,
+      "step": 1346
+    },
+    {
+      "epoch": 96.21428571428571,
+      "grad_norm": 0.1328125,
+      "learning_rate": 2.2882448739947658e-05,
+      "loss": 0.0131,
+      "step": 1347
+    },
+    {
+      "epoch": 96.28571428571429,
+      "grad_norm": 0.1728515625,
+      "learning_rate": 2.2877586553686824e-05,
+      "loss": 0.0153,
+      "step": 1348
+    },
+    {
+      "epoch": 96.35714285714286,
+      "grad_norm": 0.248046875,
+      "learning_rate": 2.287271930951637e-05,
+      "loss": 0.013,
+      "step": 1349
+    },
+    {
+      "epoch": 96.42857142857143,
+      "grad_norm": 0.173828125,
+      "learning_rate": 2.286784700980854e-05,
+      "loss": 0.0132,
+      "step": 1350
+    },
+    {
+      "epoch": 96.5,
+      "grad_norm": 0.126953125,
+      "learning_rate": 2.286296965693802e-05,
+      "loss": 0.0176,
+      "step": 1351
+    },
+    {
+      "epoch": 96.57142857142857,
+      "grad_norm": 0.1240234375,
+      "learning_rate": 2.2858087253281986e-05,
+      "loss": 0.0136,
+      "step": 1352
+    },
+    {
+      "epoch": 96.64285714285714,
+      "grad_norm": 0.357421875,
+      "learning_rate": 2.2853199801220053e-05,
+      "loss": 0.0125,
+      "step": 1353
+    },
+    {
+      "epoch": 96.71428571428571,
+      "grad_norm": 0.193359375,
+      "learning_rate": 2.284830730313431e-05,
+      "loss": 0.0139,
+      "step": 1354
+    },
+    {
+      "epoch": 96.78571428571429,
+      "grad_norm": 0.251953125,
+      "learning_rate": 2.2843409761409297e-05,
+      "loss": 0.0134,
+      "step": 1355
+    },
+    {
+      "epoch": 96.85714285714286,
+      "grad_norm": 0.427734375,
+      "learning_rate": 2.283850717843202e-05,
+      "loss": 0.014,
+      "step": 1356
+    },
+    {
+      "epoch": 96.92857142857143,
+      "grad_norm": 0.10009765625,
+      "learning_rate": 2.2833599556591945e-05,
+      "loss": 0.0124,
+      "step": 1357
+    },
+    {
+      "epoch": 97.0,
+      "grad_norm": 0.2158203125,
+      "learning_rate": 2.2828686898280974e-05,
+      "loss": 0.0143,
+      "step": 1358
+    },
+    {
+      "epoch": 97.07142857142857,
+      "grad_norm": 0.365234375,
+      "learning_rate": 2.2823769205893484e-05,
+      "loss": 0.0179,
+      "step": 1359
+    },
+    {
+      "epoch": 97.14285714285714,
+      "grad_norm": 0.1982421875,
+      "learning_rate": 2.2818846481826298e-05,
+      "loss": 0.0131,
+      "step": 1360
+    },
+    {
+      "epoch": 97.21428571428571,
+      "grad_norm": 0.11279296875,
+      "learning_rate": 2.281391872847869e-05,
+      "loss": 0.0121,
+      "step": 1361
+    },
+    {
+      "epoch": 97.28571428571429,
+      "grad_norm": 0.310546875,
+      "learning_rate": 2.280898594825239e-05,
+      "loss": 0.0124,
+      "step": 1362
+    },
+    {
+      "epoch": 97.35714285714286,
+      "grad_norm": 0.458984375,
+      "learning_rate": 2.2804048143551575e-05,
+      "loss": 0.0129,
+      "step": 1363
+    },
+    {
+      "epoch": 97.42857142857143,
+      "grad_norm": 0.2470703125,
+      "learning_rate": 2.279910531678287e-05,
+      "loss": 0.0139,
+      "step": 1364
+    },
+    {
+      "epoch": 97.5,
+      "grad_norm": 0.333984375,
+      "learning_rate": 2.2794157470355344e-05,
+      "loss": 0.0128,
+      "step": 1365
+    },
+    {
+      "epoch": 97.57142857142857,
+      "grad_norm": 0.201171875,
+      "learning_rate": 2.2789204606680524e-05,
+      "loss": 0.0159,
+      "step": 1366
+    },
+    {
+      "epoch": 97.64285714285714,
+      "grad_norm": 0.166015625,
+      "learning_rate": 2.2784246728172375e-05,
+      "loss": 0.0133,
+      "step": 1367
+    },
+    {
+      "epoch": 97.71428571428571,
+      "grad_norm": 0.1640625,
+      "learning_rate": 2.2779283837247305e-05,
+      "loss": 0.0112,
+      "step": 1368
+    },
+    {
+      "epoch": 97.78571428571429,
+      "grad_norm": 0.12353515625,
+      "learning_rate": 2.2774315936324163e-05,
+      "loss": 0.0154,
+      "step": 1369
+    },
+    {
+      "epoch": 97.85714285714286,
+      "grad_norm": 0.4296875,
+      "learning_rate": 2.2769343027824253e-05,
+      "loss": 0.0166,
+      "step": 1370
+    },
+    {
+      "epoch": 97.92857142857143,
+      "grad_norm": 0.296875,
+      "learning_rate": 2.2764365114171303e-05,
+      "loss": 0.0152,
+      "step": 1371
+    },
+    {
+      "epoch": 98.0,
+      "grad_norm": 0.1953125,
+      "learning_rate": 2.275938219779149e-05,
+      "loss": 0.0128,
+      "step": 1372
+    },
+    {
+      "epoch": 98.07142857142857,
+      "grad_norm": 0.146484375,
+      "learning_rate": 2.2754394281113424e-05,
+      "loss": 0.0139,
+      "step": 1373
+    },
+    {
+      "epoch": 98.14285714285714,
+      "grad_norm": 0.177734375,
+      "learning_rate": 2.2749401366568154e-05,
+      "loss": 0.0146,
+      "step": 1374
+    },
+    {
+      "epoch": 98.21428571428571,
+      "grad_norm": 0.1552734375,
+      "learning_rate": 2.2744403456589173e-05,
+      "loss": 0.0144,
+      "step": 1375
+    },
+    {
+      "epoch": 98.28571428571429,
+      "grad_norm": 0.1435546875,
+      "learning_rate": 2.27394005536124e-05,
+      "loss": 0.0173,
+      "step": 1376
+    },
+    {
+      "epoch": 98.35714285714286,
+      "grad_norm": 0.267578125,
+      "learning_rate": 2.2734392660076183e-05,
+      "loss": 0.0137,
+      "step": 1377
+    },
+    {
+      "epoch": 98.42857142857143,
+      "grad_norm": 0.16796875,
+      "learning_rate": 2.2729379778421312e-05,
+      "loss": 0.0124,
+      "step": 1378
+    },
+    {
+      "epoch": 98.5,
+      "grad_norm": 0.1865234375,
+      "learning_rate": 2.272436191109101e-05,
+      "loss": 0.0136,
+      "step": 1379
+    },
+    {
+      "epoch": 98.57142857142857,
+      "grad_norm": 0.291015625,
+      "learning_rate": 2.2719339060530914e-05,
+      "loss": 0.0145,
+      "step": 1380
+    },
+    {
+      "epoch": 98.64285714285714,
+      "grad_norm": 0.2158203125,
+      "learning_rate": 2.2714311229189105e-05,
+      "loss": 0.0136,
+      "step": 1381
+    },
+    {
+      "epoch": 98.71428571428571,
+      "grad_norm": 0.2080078125,
+      "learning_rate": 2.2709278419516096e-05,
+      "loss": 0.0161,
+      "step": 1382
+    },
+    {
+      "epoch": 98.78571428571429,
+      "grad_norm": 0.134765625,
+      "learning_rate": 2.2704240633964802e-05,
+      "loss": 0.0124,
+      "step": 1383
+    },
+    {
+      "epoch": 98.85714285714286,
+      "grad_norm": 0.1748046875,
+      "learning_rate": 2.2699197874990584e-05,
+      "loss": 0.0123,
+      "step": 1384
+    },
+    {
+      "epoch": 98.92857142857143,
+      "grad_norm": 0.134765625,
+      "learning_rate": 2.2694150145051228e-05,
+      "loss": 0.0113,
+      "step": 1385
+    },
+    {
+      "epoch": 99.0,
+      "grad_norm": 0.287109375,
+      "learning_rate": 2.268909744660693e-05,
+      "loss": 0.0149,
+      "step": 1386
+    },
+    {
+      "epoch": 99.07142857142857,
+      "grad_norm": 0.1435546875,
+      "learning_rate": 2.2684039782120315e-05,
+      "loss": 0.0141,
+      "step": 1387
+    },
+    {
+      "epoch": 99.14285714285714,
+      "grad_norm": 0.1884765625,
+      "learning_rate": 2.2678977154056426e-05,
+      "loss": 0.013,
+      "step": 1388
+    },
+    {
+      "epoch": 99.21428571428571,
+      "grad_norm": 0.2421875,
+      "learning_rate": 2.267390956488273e-05,
+      "loss": 0.0127,
+      "step": 1389
+    },
+    {
+      "epoch": 99.28571428571429,
+      "grad_norm": 0.1708984375,
+      "learning_rate": 2.26688370170691e-05,
+      "loss": 0.0127,
+      "step": 1390
+    },
+    {
+      "epoch": 99.35714285714286,
+      "grad_norm": 0.1044921875,
+      "learning_rate": 2.2663759513087846e-05,
+      "loss": 0.0123,
+      "step": 1391
+    },
+    {
+      "epoch": 99.42857142857143,
+      "grad_norm": 0.10400390625,
+      "learning_rate": 2.2658677055413676e-05,
+      "loss": 0.0116,
+      "step": 1392
+    },
+    {
+      "epoch": 99.5,
+      "grad_norm": 0.271484375,
+      "learning_rate": 2.2653589646523714e-05,
+      "loss": 0.0136,
+      "step": 1393
+    },
+    {
+      "epoch": 99.57142857142857,
+      "grad_norm": 0.130859375,
+      "learning_rate": 2.2648497288897514e-05,
+      "loss": 0.0163,
+      "step": 1394
+    },
+    {
+      "epoch": 99.64285714285714,
+      "grad_norm": 0.1533203125,
+      "learning_rate": 2.2643399985017015e-05,
+      "loss": 0.0134,
+      "step": 1395
+    },
+    {
+      "epoch": 99.71428571428571,
+      "grad_norm": 0.2275390625,
+      "learning_rate": 2.2638297737366583e-05,
+      "loss": 0.0151,
+      "step": 1396
+    },
+    {
+      "epoch": 99.78571428571429,
+      "grad_norm": 0.1796875,
+      "learning_rate": 2.2633190548433e-05,
+      "loss": 0.0136,
+      "step": 1397
+    },
+    {
+      "epoch": 99.85714285714286,
+      "grad_norm": 0.197265625,
+      "learning_rate": 2.2628078420705443e-05,
+      "loss": 0.0129,
+      "step": 1398
+    },
+    {
+      "epoch": 99.92857142857143,
+      "grad_norm": 0.140625,
+      "learning_rate": 2.2622961356675498e-05,
+      "loss": 0.0155,
+      "step": 1399
+    },
+    {
+      "epoch": 100.0,
+      "grad_norm": 0.263671875,
+      "learning_rate": 2.261783935883717e-05,
+      "loss": 0.0122,
+      "step": 1400
+    },
+    {
+      "epoch": 100.07142857142857,
+      "grad_norm": 0.134765625,
+      "learning_rate": 2.2612712429686845e-05,
+      "loss": 0.0125,
+      "step": 1401
+    },
+    {
+      "epoch": 100.14285714285714,
+      "grad_norm": 0.134765625,
+      "learning_rate": 2.260758057172333e-05,
+      "loss": 0.0143,
+      "step": 1402
+    },
+    {
+      "epoch": 100.21428571428571,
+      "grad_norm": 0.2490234375,
+      "learning_rate": 2.2602443787447842e-05,
+      "loss": 0.0159,
+      "step": 1403
+    },
+    {
+      "epoch": 100.28571428571429,
+      "grad_norm": 0.171875,
+      "learning_rate": 2.259730207936397e-05,
+      "loss": 0.0179,
+      "step": 1404
+    },
+    {
+      "epoch": 100.35714285714286,
+      "grad_norm": 0.107421875,
+      "learning_rate": 2.2592155449977733e-05,
+      "loss": 0.0132,
+      "step": 1405
+    },
+    {
+      "epoch": 100.42857142857143,
+      "grad_norm": 0.1494140625,
+      "learning_rate": 2.2587003901797528e-05,
+      "loss": 0.0157,
+      "step": 1406
+    },
+    {
+      "epoch": 100.5,
+      "grad_norm": 0.1572265625,
+      "learning_rate": 2.258184743733416e-05,
+      "loss": 0.0144,
+      "step": 1407
+    },
+    {
+      "epoch": 100.57142857142857,
+      "grad_norm": 0.1865234375,
+      "learning_rate": 2.2576686059100825e-05,
+      "loss": 0.0161,
+      "step": 1408
+    },
+    {
+      "epoch": 100.64285714285714,
+      "grad_norm": 0.10009765625,
+      "learning_rate": 2.2571519769613117e-05,
+      "loss": 0.0136,
+      "step": 1409
+    },
+    {
+      "epoch": 100.71428571428571,
+      "grad_norm": 0.189453125,
+      "learning_rate": 2.2566348571389025e-05,
+      "loss": 0.0149,
+      "step": 1410
+    },
+    {
+      "epoch": 100.78571428571429,
+      "grad_norm": 0.1337890625,
+      "learning_rate": 2.2561172466948925e-05,
+      "loss": 0.015,
+      "step": 1411
+    },
+    {
+      "epoch": 100.85714285714286,
+      "grad_norm": 0.1328125,
+      "learning_rate": 2.2555991458815597e-05,
+      "loss": 0.014,
+      "step": 1412
+    },
+    {
+      "epoch": 100.92857142857143,
+      "grad_norm": 0.11865234375,
+      "learning_rate": 2.2550805549514184e-05,
+      "loss": 0.0111,
+      "step": 1413
+    },
+    {
+      "epoch": 101.0,
+      "grad_norm": 0.10400390625,
+      "learning_rate": 2.2545614741572248e-05,
+      "loss": 0.0134,
+      "step": 1414
+    },
+    {
+      "epoch": 101.07142857142857,
+      "grad_norm": 0.1845703125,
+      "learning_rate": 2.2540419037519725e-05,
+      "loss": 0.0151,
+      "step": 1415
+    },
+    {
+      "epoch": 101.14285714285714,
+      "grad_norm": 0.1279296875,
+      "learning_rate": 2.2535218439888933e-05,
+      "loss": 0.0151,
+      "step": 1416
+    },
+    {
+      "epoch": 101.21428571428571,
+      "grad_norm": 0.1513671875,
+      "learning_rate": 2.2530012951214587e-05,
+      "loss": 0.0144,
+      "step": 1417
+    },
+    {
+      "epoch": 101.28571428571429,
+      "grad_norm": 0.1162109375,
+      "learning_rate": 2.2524802574033765e-05,
+      "loss": 0.0155,
+      "step": 1418
+    },
+    {
+      "epoch": 101.35714285714286,
+      "grad_norm": 0.21484375,
+      "learning_rate": 2.251958731088596e-05,
+      "loss": 0.0129,
+      "step": 1419
+    },
+    {
+      "epoch": 101.42857142857143,
+      "grad_norm": 0.1689453125,
+      "learning_rate": 2.2514367164313017e-05,
+      "loss": 0.0161,
+      "step": 1420
+    },
+    {
+      "epoch": 101.5,
+      "grad_norm": 0.134765625,
+      "learning_rate": 2.2509142136859168e-05,
+      "loss": 0.0158,
+      "step": 1421
+    },
+    {
+      "epoch": 101.57142857142857,
+      "grad_norm": 0.1181640625,
+      "learning_rate": 2.2503912231071038e-05,
+      "loss": 0.0148,
+      "step": 1422
+    },
+    {
+      "epoch": 101.64285714285714,
+      "grad_norm": 0.12451171875,
+      "learning_rate": 2.249867744949762e-05,
+      "loss": 0.0148,
+      "step": 1423
+    },
+    {
+      "epoch": 101.71428571428571,
+      "grad_norm": 0.125,
+      "learning_rate": 2.2493437794690274e-05,
+      "loss": 0.0149,
+      "step": 1424
+    },
+    {
+      "epoch": 101.78571428571429,
+      "grad_norm": 0.109375,
+      "learning_rate": 2.2488193269202748e-05,
+      "loss": 0.0139,
+      "step": 1425
+    },
+    {
+      "epoch": 101.85714285714286,
+      "grad_norm": 0.208984375,
+      "learning_rate": 2.248294387559116e-05,
+      "loss": 0.013,
+      "step": 1426
+    },
+    {
+      "epoch": 101.92857142857143,
+      "grad_norm": 0.154296875,
+      "learning_rate": 2.2477689616414004e-05,
+      "loss": 0.0138,
+      "step": 1427
+    },
+    {
+      "epoch": 102.0,
+      "grad_norm": 0.1474609375,
+      "learning_rate": 2.2472430494232138e-05,
+      "loss": 0.0119,
+      "step": 1428
+    },
+    {
+      "epoch": 102.07142857142857,
+      "grad_norm": 0.107421875,
+      "learning_rate": 2.2467166511608796e-05,
+      "loss": 0.0158,
+      "step": 1429
+    },
+    {
+      "epoch": 102.14285714285714,
+      "grad_norm": 0.2216796875,
+      "learning_rate": 2.246189767110958e-05,
+      "loss": 0.0121,
+      "step": 1430
+    },
+    {
+      "epoch": 102.21428571428571,
+      "grad_norm": 0.357421875,
+      "learning_rate": 2.2456623975302454e-05,
+      "loss": 0.0162,
+      "step": 1431
+    },
+    {
+      "epoch": 102.28571428571429,
+      "grad_norm": 0.1484375,
+      "learning_rate": 2.245134542675776e-05,
+      "loss": 0.0134,
+      "step": 1432
+    },
+    {
+      "epoch": 102.35714285714286,
+      "grad_norm": 0.185546875,
+      "learning_rate": 2.2446062028048193e-05,
+      "loss": 0.0156,
+      "step": 1433
+    },
+    {
+      "epoch": 102.42857142857143,
+      "grad_norm": 0.1376953125,
+      "learning_rate": 2.244077378174882e-05,
+      "loss": 0.0118,
+      "step": 1434
+    },
+    {
+      "epoch": 102.5,
+      "grad_norm": 0.1416015625,
+      "learning_rate": 2.243548069043707e-05,
+      "loss": 0.0142,
+      "step": 1435
+    },
+    {
+      "epoch": 102.57142857142857,
+      "grad_norm": 0.134765625,
+      "learning_rate": 2.2430182756692728e-05,
+      "loss": 0.0105,
+      "step": 1436
+    },
+    {
+      "epoch": 102.64285714285714,
+      "grad_norm": 0.263671875,
+      "learning_rate": 2.2424879983097943e-05,
+      "loss": 0.0148,
+      "step": 1437
+    },
+    {
+      "epoch": 102.71428571428571,
+      "grad_norm": 0.310546875,
+      "learning_rate": 2.2419572372237226e-05,
+      "loss": 0.015,
+      "step": 1438
+    },
+    {
+      "epoch": 102.78571428571429,
+      "grad_norm": 0.12353515625,
+      "learning_rate": 2.2414259926697434e-05,
+      "loss": 0.0154,
+      "step": 1439
+    },
+    {
+      "epoch": 102.85714285714286,
+      "grad_norm": 0.1865234375,
+      "learning_rate": 2.2408942649067795e-05,
+      "loss": 0.014,
+      "step": 1440
+    },
+    {
+      "epoch": 102.92857142857143,
+      "grad_norm": 0.13671875,
+      "learning_rate": 2.2403620541939884e-05,
+      "loss": 0.0136,
+      "step": 1441
+    },
+    {
+      "epoch": 103.0,
+      "grad_norm": 0.18359375,
+      "learning_rate": 2.2398293607907634e-05,
+      "loss": 0.0132,
+      "step": 1442
+    },
+    {
+      "epoch": 103.07142857142857,
+      "grad_norm": 0.228515625,
+      "learning_rate": 2.2392961849567325e-05,
+      "loss": 0.013,
+      "step": 1443
+    },
+    {
+      "epoch": 103.14285714285714,
+      "grad_norm": 0.134765625,
+      "learning_rate": 2.2387625269517587e-05,
+      "loss": 0.0133,
+      "step": 1444
+    },
+    {
+      "epoch": 103.21428571428571,
+      "grad_norm": 0.2265625,
+      "learning_rate": 2.2382283870359412e-05,
+      "loss": 0.0137,
+      "step": 1445
+    },
+    {
+      "epoch": 103.28571428571429,
+      "grad_norm": 0.25390625,
+      "learning_rate": 2.2376937654696133e-05,
+      "loss": 0.012,
+      "step": 1446
+    },
+    {
+      "epoch": 103.35714285714286,
+      "grad_norm": 0.1865234375,
+      "learning_rate": 2.237158662513342e-05,
+      "loss": 0.0166,
+      "step": 1447
+    },
+    {
+      "epoch": 103.42857142857143,
+      "grad_norm": 0.10302734375,
+      "learning_rate": 2.2366230784279313e-05,
+      "loss": 0.0129,
+      "step": 1448
+    },
+    {
+      "epoch": 103.5,
+      "grad_norm": 0.193359375,
+      "learning_rate": 2.2360870134744178e-05,
+      "loss": 0.0133,
+      "step": 1449
+    },
+    {
+      "epoch": 103.57142857142857,
+      "grad_norm": 0.1376953125,
+      "learning_rate": 2.2355504679140728e-05,
+      "loss": 0.0133,
+      "step": 1450
+    },
+    {
+      "epoch": 103.64285714285714,
+      "grad_norm": 0.12255859375,
+      "learning_rate": 2.2350134420084023e-05,
+      "loss": 0.0125,
+      "step": 1451
+    },
+    {
+      "epoch": 103.71428571428571,
+      "grad_norm": 0.1611328125,
+      "learning_rate": 2.2344759360191468e-05,
+      "loss": 0.0134,
+      "step": 1452
+    },
+    {
+      "epoch": 103.78571428571429,
+      "grad_norm": 0.130859375,
+      "learning_rate": 2.2339379502082792e-05,
+      "loss": 0.0129,
+      "step": 1453
+    },
+    {
+      "epoch": 103.85714285714286,
+      "grad_norm": 0.1416015625,
+      "learning_rate": 2.2333994848380075e-05,
+      "loss": 0.0138,
+      "step": 1454
+    },
+    {
+      "epoch": 103.92857142857143,
+      "grad_norm": 0.1533203125,
+      "learning_rate": 2.2328605401707735e-05,
+      "loss": 0.0148,
+      "step": 1455
+    },
+    {
+      "epoch": 104.0,
+      "grad_norm": 0.177734375,
+      "learning_rate": 2.2323211164692526e-05,
+      "loss": 0.0152,
+      "step": 1456
+    },
+    {
+      "epoch": 104.07142857142857,
+      "grad_norm": 0.13671875,
+      "learning_rate": 2.2317812139963522e-05,
+      "loss": 0.0151,
+      "step": 1457
+    },
+    {
+      "epoch": 104.14285714285714,
+      "grad_norm": 0.1259765625,
+      "learning_rate": 2.2312408330152157e-05,
+      "loss": 0.0142,
+      "step": 1458
+    },
+    {
+      "epoch": 104.21428571428571,
+      "grad_norm": 0.1533203125,
+      "learning_rate": 2.2306999737892168e-05,
+      "loss": 0.0122,
+      "step": 1459
+    },
+    {
+      "epoch": 104.28571428571429,
+      "grad_norm": 0.16796875,
+      "learning_rate": 2.2301586365819644e-05,
+      "loss": 0.0123,
+      "step": 1460
+    },
+    {
+      "epoch": 104.35714285714286,
+      "grad_norm": 0.12451171875,
+      "learning_rate": 2.2296168216573e-05,
+      "loss": 0.0121,
+      "step": 1461
+    },
+    {
+      "epoch": 104.42857142857143,
+      "grad_norm": 0.177734375,
+      "learning_rate": 2.229074529279297e-05,
+      "loss": 0.0143,
+      "step": 1462
+    },
+    {
+      "epoch": 104.5,
+      "grad_norm": 0.10693359375,
+      "learning_rate": 2.2285317597122623e-05,
+      "loss": 0.0125,
+      "step": 1463
+    },
+    {
+      "epoch": 104.57142857142857,
+      "grad_norm": 0.1474609375,
+      "learning_rate": 2.227988513220735e-05,
+      "loss": 0.0148,
+      "step": 1464
+    },
+    {
+      "epoch": 104.64285714285714,
+      "grad_norm": 0.18359375,
+      "learning_rate": 2.2274447900694876e-05,
+      "loss": 0.0143,
+      "step": 1465
+    },
+    {
+      "epoch": 104.71428571428571,
+      "grad_norm": 0.1376953125,
+      "learning_rate": 2.2269005905235234e-05,
+      "loss": 0.0136,
+      "step": 1466
+    },
+    {
+      "epoch": 104.78571428571429,
+      "grad_norm": 0.1328125,
+      "learning_rate": 2.226355914848079e-05,
+      "loss": 0.012,
+      "step": 1467
+    },
+    {
+      "epoch": 104.85714285714286,
+      "grad_norm": 0.1103515625,
+      "learning_rate": 2.2258107633086227e-05,
+      "loss": 0.0136,
+      "step": 1468
+    },
+    {
+      "epoch": 104.92857142857143,
+      "grad_norm": 0.26953125,
+      "learning_rate": 2.2252651361708544e-05,
+      "loss": 0.0134,
+      "step": 1469
+    },
+    {
+      "epoch": 105.0,
+      "grad_norm": 0.1474609375,
+      "learning_rate": 2.2247190337007066e-05,
+      "loss": 0.0122,
+      "step": 1470
+    },
+    {
+      "epoch": 105.07142857142857,
+      "grad_norm": 0.1474609375,
+      "learning_rate": 2.224172456164343e-05,
+      "loss": 0.0121,
+      "step": 1471
+    },
+    {
+      "epoch": 105.14285714285714,
+      "grad_norm": 0.2333984375,
+      "learning_rate": 2.2236254038281583e-05,
+      "loss": 0.0163,
+      "step": 1472
+    },
+    {
+      "epoch": 105.21428571428571,
+      "grad_norm": 0.236328125,
+      "learning_rate": 2.2230778769587797e-05,
+      "loss": 0.0143,
+      "step": 1473
+    },
+    {
+      "epoch": 105.28571428571429,
+      "grad_norm": 0.19140625,
+      "learning_rate": 2.2225298758230643e-05,
+      "loss": 0.015,
+      "step": 1474
+    },
+    {
+      "epoch": 105.35714285714286,
+      "grad_norm": 0.2080078125,
+      "learning_rate": 2.2219814006881027e-05,
+      "loss": 0.0111,
+      "step": 1475
+    },
+    {
+      "epoch": 105.42857142857143,
+      "grad_norm": 0.1396484375,
+      "learning_rate": 2.221432451821214e-05,
+      "loss": 0.0132,
+      "step": 1476
+    },
+    {
+      "epoch": 105.5,
+      "grad_norm": 0.126953125,
+      "learning_rate": 2.220883029489949e-05,
+      "loss": 0.0122,
+      "step": 1477
+    },
+    {
+      "epoch": 105.57142857142857,
+      "grad_norm": 0.1435546875,
+      "learning_rate": 2.22033313396209e-05,
+      "loss": 0.0149,
+      "step": 1478
+    },
+    {
+      "epoch": 105.64285714285714,
+      "grad_norm": 0.19140625,
+      "learning_rate": 2.2197827655056494e-05,
+      "loss": 0.0151,
+      "step": 1479
+    },
+    {
+      "epoch": 105.71428571428571,
+      "grad_norm": 0.283203125,
+      "learning_rate": 2.21923192438887e-05,
+      "loss": 0.0153,
+      "step": 1480
+    },
+    {
+      "epoch": 105.78571428571429,
+      "grad_norm": 0.1640625,
+      "learning_rate": 2.2186806108802248e-05,
+      "loss": 0.0118,
+      "step": 1481
+    },
+    {
+      "epoch": 105.85714285714286,
+      "grad_norm": 0.126953125,
+      "learning_rate": 2.218128825248417e-05,
+      "loss": 0.0152,
+      "step": 1482
+    },
+    {
+      "epoch": 105.92857142857143,
+      "grad_norm": 0.421875,
+      "learning_rate": 2.2175765677623816e-05,
+      "loss": 0.0154,
+      "step": 1483
+    },
+    {
+      "epoch": 106.0,
+      "grad_norm": 0.2060546875,
+      "learning_rate": 2.2170238386912813e-05,
+      "loss": 0.018,
+      "step": 1484
+    },
+    {
+      "epoch": 106.07142857142857,
+      "grad_norm": 0.212890625,
+      "learning_rate": 2.216470638304509e-05,
+      "loss": 0.0156,
+      "step": 1485
+    },
+    {
+      "epoch": 106.14285714285714,
+      "grad_norm": 0.255859375,
+      "learning_rate": 2.215916966871689e-05,
+      "loss": 0.0135,
+      "step": 1486
+    },
+    {
+      "epoch": 106.21428571428571,
+      "grad_norm": 0.162109375,
+      "learning_rate": 2.2153628246626728e-05,
+      "loss": 0.0161,
+      "step": 1487
+    },
+    {
+      "epoch": 106.28571428571429,
+      "grad_norm": 0.11572265625,
+      "learning_rate": 2.2148082119475437e-05,
+      "loss": 0.0129,
+      "step": 1488
+    },
+    {
+      "epoch": 106.35714285714286,
+      "grad_norm": 0.2099609375,
+      "learning_rate": 2.214253128996613e-05,
+      "loss": 0.0154,
+      "step": 1489
+    },
+    {
+      "epoch": 106.42857142857143,
+      "grad_norm": 0.1025390625,
+      "learning_rate": 2.2136975760804204e-05,
+      "loss": 0.011,
+      "step": 1490
+    },
+    {
+      "epoch": 106.5,
+      "grad_norm": 0.1923828125,
+      "learning_rate": 2.213141553469737e-05,
+      "loss": 0.0114,
+      "step": 1491
+    },
+    {
+      "epoch": 106.57142857142857,
+      "grad_norm": 0.2373046875,
+      "learning_rate": 2.21258506143556e-05,
+      "loss": 0.0141,
+      "step": 1492
+    },
+    {
+      "epoch": 106.64285714285714,
+      "grad_norm": 0.11865234375,
+      "learning_rate": 2.2120281002491182e-05,
+      "loss": 0.0124,
+      "step": 1493
+    },
+    {
+      "epoch": 106.71428571428571,
+      "grad_norm": 0.341796875,
+      "learning_rate": 2.2114706701818672e-05,
+      "loss": 0.0154,
+      "step": 1494
+    },
+    {
+      "epoch": 106.78571428571429,
+      "grad_norm": 0.11376953125,
+      "learning_rate": 2.2109127715054916e-05,
+      "loss": 0.0126,
+      "step": 1495
+    },
+    {
+      "epoch": 106.85714285714286,
+      "grad_norm": 0.2177734375,
+      "learning_rate": 2.2103544044919045e-05,
+      "loss": 0.0134,
+      "step": 1496
+    },
+    {
+      "epoch": 106.92857142857143,
+      "grad_norm": 0.1298828125,
+      "learning_rate": 2.2097955694132473e-05,
+      "loss": 0.0122,
+      "step": 1497
+    },
+    {
+      "epoch": 107.0,
+      "grad_norm": 0.2373046875,
+      "learning_rate": 2.2092362665418887e-05,
+      "loss": 0.0136,
+      "step": 1498
+    },
+    {
+      "epoch": 107.07142857142857,
+      "grad_norm": 0.18359375,
+      "learning_rate": 2.208676496150427e-05,
+      "loss": 0.0139,
+      "step": 1499
+    },
+    {
+      "epoch": 107.14285714285714,
+      "grad_norm": 0.1005859375,
+      "learning_rate": 2.2081162585116872e-05,
+      "loss": 0.0121,
+      "step": 1500
+    },
+    {
+      "epoch": 107.21428571428571,
+      "grad_norm": 0.21484375,
+      "learning_rate": 2.2075555538987227e-05,
+      "loss": 0.0133,
+      "step": 1501
+    },
+    {
+      "epoch": 107.28571428571429,
+      "grad_norm": 0.11669921875,
+      "learning_rate": 2.2069943825848133e-05,
+      "loss": 0.0138,
+      "step": 1502
+    },
+    {
+      "epoch": 107.35714285714286,
+      "grad_norm": 0.1689453125,
+      "learning_rate": 2.206432744843468e-05,
+      "loss": 0.0167,
+      "step": 1503
+    },
+    {
+      "epoch": 107.42857142857143,
+      "grad_norm": 0.2158203125,
+      "learning_rate": 2.2058706409484213e-05,
+      "loss": 0.0156,
+      "step": 1504
+    },
+    {
+      "epoch": 107.5,
+      "grad_norm": 0.2255859375,
+      "learning_rate": 2.2053080711736364e-05,
+      "loss": 0.0123,
+      "step": 1505
+    },
+    {
+      "epoch": 107.57142857142857,
+      "grad_norm": 0.1884765625,
+      "learning_rate": 2.2047450357933032e-05,
+      "loss": 0.0142,
+      "step": 1506
+    },
+    {
+      "epoch": 107.64285714285714,
+      "grad_norm": 0.2138671875,
+      "learning_rate": 2.2041815350818374e-05,
+      "loss": 0.0152,
+      "step": 1507
+    },
+    {
+      "epoch": 107.71428571428571,
+      "grad_norm": 0.1279296875,
+      "learning_rate": 2.2036175693138827e-05,
+      "loss": 0.0151,
+      "step": 1508
+    },
+    {
+      "epoch": 107.78571428571429,
+      "grad_norm": 0.11376953125,
+      "learning_rate": 2.20305313876431e-05,
+      "loss": 0.0137,
+      "step": 1509
+    },
+    {
+      "epoch": 107.85714285714286,
+      "grad_norm": 0.11865234375,
+      "learning_rate": 2.202488243708215e-05,
+      "loss": 0.0139,
+      "step": 1510
+    },
+    {
+      "epoch": 107.92857142857143,
+      "grad_norm": 0.1640625,
+      "learning_rate": 2.201922884420921e-05,
+      "loss": 0.0144,
+      "step": 1511
+    },
+    {
+      "epoch": 108.0,
+      "grad_norm": 0.171875,
+      "learning_rate": 2.2013570611779775e-05,
+      "loss": 0.0147,
+      "step": 1512
+    },
+    {
+      "epoch": 108.07142857142857,
+      "grad_norm": 0.1474609375,
+      "learning_rate": 2.200790774255159e-05,
+      "loss": 0.0132,
+      "step": 1513
+    },
+    {
+      "epoch": 108.14285714285714,
+      "grad_norm": 0.146484375,
+      "learning_rate": 2.2002240239284677e-05,
+      "loss": 0.0149,
+      "step": 1514
+    },
+    {
+      "epoch": 108.21428571428571,
+      "grad_norm": 0.125,
+      "learning_rate": 2.199656810474131e-05,
+      "loss": 0.0129,
+      "step": 1515
+    },
+    {
+      "epoch": 108.28571428571429,
+      "grad_norm": 0.16796875,
+      "learning_rate": 2.1990891341686008e-05,
+      "loss": 0.0119,
+      "step": 1516
+    },
+    {
+      "epoch": 108.35714285714286,
+      "grad_norm": 0.11865234375,
+      "learning_rate": 2.1985209952885573e-05,
+      "loss": 0.0143,
+      "step": 1517
+    },
+    {
+      "epoch": 108.42857142857143,
+      "grad_norm": 0.15625,
+      "learning_rate": 2.1979523941109026e-05,
+      "loss": 0.0166,
+      "step": 1518
+    },
+    {
+      "epoch": 108.5,
+      "grad_norm": 0.392578125,
+      "learning_rate": 2.1973833309127666e-05,
+      "loss": 0.0134,
+      "step": 1519
+    },
+    {
+      "epoch": 108.57142857142857,
+      "grad_norm": 0.16796875,
+      "learning_rate": 2.1968138059715048e-05,
+      "loss": 0.0132,
+      "step": 1520
+    },
+    {
+      "epoch": 108.64285714285714,
+      "grad_norm": 0.126953125,
+      "learning_rate": 2.1962438195646958e-05,
+      "loss": 0.013,
+      "step": 1521
+    },
+    {
+      "epoch": 108.71428571428571,
+      "grad_norm": 0.177734375,
+      "learning_rate": 2.195673371970144e-05,
+      "loss": 0.013,
+      "step": 1522
+    },
+    {
+      "epoch": 108.78571428571429,
+      "grad_norm": 0.1494140625,
+      "learning_rate": 2.195102463465879e-05,
+      "loss": 0.0144,
+      "step": 1523
+    },
+    {
+      "epoch": 108.85714285714286,
+      "grad_norm": 0.138671875,
+      "learning_rate": 2.1945310943301544e-05,
+      "loss": 0.0121,
+      "step": 1524
+    },
+    {
+      "epoch": 108.92857142857143,
+      "grad_norm": 0.28515625,
+      "learning_rate": 2.193959264841449e-05,
+      "loss": 0.0151,
+      "step": 1525
+    },
+    {
+      "epoch": 109.0,
+      "grad_norm": 0.138671875,
+      "learning_rate": 2.1933869752784654e-05,
+      "loss": 0.014,
+      "step": 1526
+    },
+    {
+      "epoch": 109.07142857142857,
+      "grad_norm": 0.3203125,
+      "learning_rate": 2.1928142259201297e-05,
+      "loss": 0.0161,
+      "step": 1527
+    },
+    {
+      "epoch": 109.14285714285714,
+      "grad_norm": 0.173828125,
+      "learning_rate": 2.1922410170455944e-05,
+      "loss": 0.0139,
+      "step": 1528
+    },
+    {
+      "epoch": 109.21428571428571,
+      "grad_norm": 0.337890625,
+      "learning_rate": 2.1916673489342337e-05,
+      "loss": 0.0157,
+      "step": 1529
+    },
+    {
+      "epoch": 109.28571428571429,
+      "grad_norm": 0.11865234375,
+      "learning_rate": 2.1910932218656464e-05,
+      "loss": 0.0134,
+      "step": 1530
+    },
+    {
+      "epoch": 109.35714285714286,
+      "grad_norm": 0.1259765625,
+      "learning_rate": 2.1905186361196556e-05,
+      "loss": 0.0156,
+      "step": 1531
+    },
+    {
+      "epoch": 109.42857142857143,
+      "grad_norm": 0.11083984375,
+      "learning_rate": 2.189943591976307e-05,
+      "loss": 0.0135,
+      "step": 1532
+    },
+    {
+      "epoch": 109.5,
+      "grad_norm": 0.26171875,
+      "learning_rate": 2.1893680897158702e-05,
+      "loss": 0.0125,
+      "step": 1533
+    },
+    {
+      "epoch": 109.57142857142857,
+      "grad_norm": 0.1328125,
+      "learning_rate": 2.188792129618838e-05,
+      "loss": 0.015,
+      "step": 1534
+    },
+    {
+      "epoch": 109.64285714285714,
+      "grad_norm": 0.150390625,
+      "learning_rate": 2.1882157119659265e-05,
+      "loss": 0.0149,
+      "step": 1535
+    },
+    {
+      "epoch": 109.71428571428571,
+      "grad_norm": 0.10595703125,
+      "learning_rate": 2.1876388370380745e-05,
+      "loss": 0.013,
+      "step": 1536
+    },
+    {
+      "epoch": 109.78571428571429,
+      "grad_norm": 0.12890625,
+      "learning_rate": 2.187061505116444e-05,
+      "loss": 0.0132,
+      "step": 1537
+    },
+    {
+      "epoch": 109.85714285714286,
+      "grad_norm": 0.140625,
+      "learning_rate": 2.186483716482419e-05,
+      "loss": 0.0164,
+      "step": 1538
+    },
+    {
+      "epoch": 109.92857142857143,
+      "grad_norm": 0.1171875,
+      "learning_rate": 2.1859054714176076e-05,
+      "loss": 0.015,
+      "step": 1539
+    },
+    {
+      "epoch": 110.0,
+      "grad_norm": 0.2275390625,
+      "learning_rate": 2.1853267702038392e-05,
+      "loss": 0.0109,
+      "step": 1540
+    },
+    {
+      "epoch": 110.07142857142857,
+      "grad_norm": 0.201171875,
+      "learning_rate": 2.184747613123165e-05,
+      "loss": 0.0125,
+      "step": 1541
+    },
+    {
+      "epoch": 110.14285714285714,
+      "grad_norm": 0.1533203125,
+      "learning_rate": 2.1841680004578596e-05,
+      "loss": 0.016,
+      "step": 1542
+    },
+    {
+      "epoch": 110.21428571428571,
+      "grad_norm": 0.1298828125,
+      "learning_rate": 2.1835879324904195e-05,
+      "loss": 0.0146,
+      "step": 1543
+    },
+    {
+      "epoch": 110.28571428571429,
+      "grad_norm": 0.107421875,
+      "learning_rate": 2.183007409503562e-05,
+      "loss": 0.0117,
+      "step": 1544
+    },
+    {
+      "epoch": 110.35714285714286,
+      "grad_norm": 0.22265625,
+      "learning_rate": 2.1824264317802278e-05,
+      "loss": 0.0184,
+      "step": 1545
+    },
+    {
+      "epoch": 110.42857142857143,
+      "grad_norm": 0.107421875,
+      "learning_rate": 2.181844999603578e-05,
+      "loss": 0.0142,
+      "step": 1546
+    },
+    {
+      "epoch": 110.5,
+      "grad_norm": 0.2578125,
+      "learning_rate": 2.1812631132569953e-05,
+      "loss": 0.0136,
+      "step": 1547
+    },
+    {
+      "epoch": 110.57142857142857,
+      "grad_norm": 0.275390625,
+      "learning_rate": 2.1806807730240845e-05,
+      "loss": 0.0171,
+      "step": 1548
+    },
+    {
+      "epoch": 110.64285714285714,
+      "grad_norm": 0.10595703125,
+      "learning_rate": 2.1800979791886707e-05,
+      "loss": 0.0136,
+      "step": 1549
+    },
+    {
+      "epoch": 110.71428571428571,
+      "grad_norm": 0.171875,
+      "learning_rate": 2.1795147320348012e-05,
+      "loss": 0.0147,
+      "step": 1550
+    },
+    {
+      "epoch": 110.78571428571429,
+      "grad_norm": 0.1787109375,
+      "learning_rate": 2.178931031846743e-05,
+      "loss": 0.0147,
+      "step": 1551
+    },
+    {
+      "epoch": 110.85714285714286,
+      "grad_norm": 0.1875,
+      "learning_rate": 2.1783468789089844e-05,
+      "loss": 0.017,
+      "step": 1552
+    },
+    {
+      "epoch": 110.92857142857143,
+      "grad_norm": 0.1611328125,
+      "learning_rate": 2.1777622735062354e-05,
+      "loss": 0.0118,
+      "step": 1553
+    },
+    {
+      "epoch": 111.0,
+      "grad_norm": 0.361328125,
+      "learning_rate": 2.1771772159234243e-05,
+      "loss": 0.0181,
+      "step": 1554
+    },
+    {
+      "epoch": 111.07142857142857,
+      "grad_norm": 0.1708984375,
+      "learning_rate": 2.1765917064457013e-05,
+      "loss": 0.0149,
+      "step": 1555
+    },
+    {
+      "epoch": 111.14285714285714,
+      "grad_norm": 0.193359375,
+      "learning_rate": 2.1760057453584376e-05,
+      "loss": 0.0143,
+      "step": 1556
+    },
+    {
+      "epoch": 111.21428571428571,
+      "grad_norm": 0.392578125,
+      "learning_rate": 2.175419332947223e-05,
+      "loss": 0.011,
+      "step": 1557
+    },
+    {
+      "epoch": 111.28571428571429,
+      "grad_norm": 0.26953125,
+      "learning_rate": 2.174832469497867e-05,
+      "loss": 0.0158,
+      "step": 1558
+    },
+    {
+      "epoch": 111.35714285714286,
+      "grad_norm": 0.1357421875,
+      "learning_rate": 2.1742451552964006e-05,
+      "loss": 0.0151,
+      "step": 1559
+    },
+    {
+      "epoch": 111.42857142857143,
+      "grad_norm": 0.1123046875,
+      "learning_rate": 2.1736573906290736e-05,
+      "loss": 0.0149,
+      "step": 1560
+    },
+    {
+      "epoch": 111.5,
+      "grad_norm": 0.2255859375,
+      "learning_rate": 2.1730691757823553e-05,
+      "loss": 0.0151,
+      "step": 1561
+    },
+    {
+      "epoch": 111.57142857142857,
+      "grad_norm": 0.28125,
+      "learning_rate": 2.1724805110429335e-05,
+      "loss": 0.0163,
+      "step": 1562
+    },
+    {
+      "epoch": 111.64285714285714,
+      "grad_norm": 0.1259765625,
+      "learning_rate": 2.1718913966977175e-05,
+      "loss": 0.0154,
+      "step": 1563
+    },
+    {
+      "epoch": 111.71428571428571,
+      "grad_norm": 0.1572265625,
+      "learning_rate": 2.1713018330338343e-05,
+      "loss": 0.0168,
+      "step": 1564
+    },
+    {
+      "epoch": 111.78571428571429,
+      "grad_norm": 0.11962890625,
+      "learning_rate": 2.1707118203386292e-05,
+      "loss": 0.0166,
+      "step": 1565
+    },
+    {
+      "epoch": 111.85714285714286,
+      "grad_norm": 0.1962890625,
+      "learning_rate": 2.1701213588996683e-05,
+      "loss": 0.0141,
+      "step": 1566
+    },
+    {
+      "epoch": 111.92857142857143,
+      "grad_norm": 0.12109375,
+      "learning_rate": 2.1695304490047344e-05,
+      "loss": 0.0148,
+      "step": 1567
+    },
+    {
+      "epoch": 112.0,
+      "grad_norm": 0.1875,
+      "learning_rate": 2.1689390909418305e-05,
+      "loss": 0.0144,
+      "step": 1568
+    },
+    {
+      "epoch": 112.07142857142857,
+      "grad_norm": 0.130859375,
+      "learning_rate": 2.168347284999177e-05,
+      "loss": 0.0153,
+      "step": 1569
+    },
+    {
+      "epoch": 112.14285714285714,
+      "grad_norm": 0.1337890625,
+      "learning_rate": 2.1677550314652127e-05,
+      "loss": 0.0157,
+      "step": 1570
+    },
+    {
+      "epoch": 112.21428571428571,
+      "grad_norm": 0.21875,
+      "learning_rate": 2.1671623306285956e-05,
+      "loss": 0.0128,
+      "step": 1571
+    },
+    {
+      "epoch": 112.28571428571429,
+      "grad_norm": 0.189453125,
+      "learning_rate": 2.1665691827782003e-05,
+      "loss": 0.0171,
+      "step": 1572
+    },
+    {
+      "epoch": 112.35714285714286,
+      "grad_norm": 0.1669921875,
+      "learning_rate": 2.1659755882031196e-05,
+      "loss": 0.0152,
+      "step": 1573
+    },
+    {
+      "epoch": 112.42857142857143,
+      "grad_norm": 0.228515625,
+      "learning_rate": 2.1653815471926646e-05,
+      "loss": 0.0113,
+      "step": 1574
+    },
+    {
+      "epoch": 112.5,
+      "grad_norm": 0.154296875,
+      "learning_rate": 2.164787060036364e-05,
+      "loss": 0.0159,
+      "step": 1575
+    },
+    {
+      "epoch": 112.57142857142857,
+      "grad_norm": 0.189453125,
+      "learning_rate": 2.1641921270239632e-05,
+      "loss": 0.0119,
+      "step": 1576
+    },
+    {
+      "epoch": 112.64285714285714,
+      "grad_norm": 0.1142578125,
+      "learning_rate": 2.1635967484454257e-05,
+      "loss": 0.0132,
+      "step": 1577
+    },
+    {
+      "epoch": 112.71428571428571,
+      "grad_norm": 0.1484375,
+      "learning_rate": 2.163000924590931e-05,
+      "loss": 0.0141,
+      "step": 1578
+    },
+    {
+      "epoch": 112.78571428571429,
+      "grad_norm": 0.099609375,
+      "learning_rate": 2.162404655750877e-05,
+      "loss": 0.0122,
+      "step": 1579
+    },
+    {
+      "epoch": 112.85714285714286,
+      "grad_norm": 0.19921875,
+      "learning_rate": 2.161807942215878e-05,
+      "loss": 0.0138,
+      "step": 1580
+    },
+    {
+      "epoch": 112.92857142857143,
+      "grad_norm": 0.1904296875,
+      "learning_rate": 2.1612107842767647e-05,
+      "loss": 0.0117,
+      "step": 1581
+    },
+    {
+      "epoch": 113.0,
+      "grad_norm": 0.2890625,
+      "learning_rate": 2.1606131822245844e-05,
+      "loss": 0.0149,
+      "step": 1582
+    },
+    {
+      "epoch": 113.07142857142857,
+      "grad_norm": 0.2294921875,
+      "learning_rate": 2.1600151363506015e-05,
+      "loss": 0.015,
+      "step": 1583
+    },
+    {
+      "epoch": 113.14285714285714,
+      "grad_norm": 0.1279296875,
+      "learning_rate": 2.159416646946296e-05,
+      "loss": 0.0138,
+      "step": 1584
+    },
+    {
+      "epoch": 113.21428571428571,
+      "grad_norm": 0.1162109375,
+      "learning_rate": 2.158817714303365e-05,
+      "loss": 0.013,
+      "step": 1585
+    },
+    {
+      "epoch": 113.28571428571429,
+      "grad_norm": 0.1513671875,
+      "learning_rate": 2.15821833871372e-05,
+      "loss": 0.0162,
+      "step": 1586
+    },
+    {
+      "epoch": 113.35714285714286,
+      "grad_norm": 0.1806640625,
+      "learning_rate": 2.1576185204694898e-05,
+      "loss": 0.0164,
+      "step": 1587
+    },
+    {
+      "epoch": 113.42857142857143,
+      "grad_norm": 0.25,
+      "learning_rate": 2.157018259863019e-05,
+      "loss": 0.0137,
+      "step": 1588
+    },
+    {
+      "epoch": 113.5,
+      "grad_norm": 0.1357421875,
+      "learning_rate": 2.1564175571868668e-05,
+      "loss": 0.0138,
+      "step": 1589
+    },
+    {
+      "epoch": 113.57142857142857,
+      "grad_norm": 0.2421875,
+      "learning_rate": 2.1558164127338086e-05,
+      "loss": 0.0165,
+      "step": 1590
+    },
+    {
+      "epoch": 113.64285714285714,
+      "grad_norm": 0.353515625,
+      "learning_rate": 2.1552148267968347e-05,
+      "loss": 0.0146,
+      "step": 1591
+    },
+    {
+      "epoch": 113.71428571428571,
+      "grad_norm": 0.185546875,
+      "learning_rate": 2.1546127996691506e-05,
+      "loss": 0.017,
+      "step": 1592
+    },
+    {
+      "epoch": 113.78571428571429,
+      "grad_norm": 0.1552734375,
+      "learning_rate": 2.1540103316441777e-05,
+      "loss": 0.014,
+      "step": 1593
+    },
+    {
+      "epoch": 113.85714285714286,
+      "grad_norm": 0.1669921875,
+      "learning_rate": 2.153407423015551e-05,
+      "loss": 0.0154,
+      "step": 1594
+    },
+    {
+      "epoch": 113.92857142857143,
+      "grad_norm": 0.171875,
+      "learning_rate": 2.152804074077121e-05,
+      "loss": 0.0145,
+      "step": 1595
+    },
+    {
+      "epoch": 114.0,
+      "grad_norm": 0.2177734375,
+      "learning_rate": 2.152200285122953e-05,
+      "loss": 0.0151,
+      "step": 1596
+    },
+    {
+      "epoch": 114.07142857142857,
+      "grad_norm": 0.171875,
+      "learning_rate": 2.151596056447326e-05,
+      "loss": 0.013,
+      "step": 1597
+    },
+    {
+      "epoch": 114.14285714285714,
+      "grad_norm": 0.337890625,
+      "learning_rate": 2.1509913883447335e-05,
+      "loss": 0.0093,
+      "step": 1598
+    },
+    {
+      "epoch": 114.21428571428571,
+      "grad_norm": 0.146484375,
+      "learning_rate": 2.1503862811098837e-05,
+      "loss": 0.0132,
+      "step": 1599
+    },
+    {
+      "epoch": 114.28571428571429,
+      "grad_norm": 0.2578125,
+      "learning_rate": 2.1497807350376986e-05,
+      "loss": 0.0192,
+      "step": 1600
+    },
+    {
+      "epoch": 114.35714285714286,
+      "grad_norm": 0.142578125,
+      "learning_rate": 2.1491747504233138e-05,
+      "loss": 0.0162,
+      "step": 1601
+    },
+    {
+      "epoch": 114.42857142857143,
+      "grad_norm": 0.162109375,
+      "learning_rate": 2.1485683275620794e-05,
+      "loss": 0.0116,
+      "step": 1602
+    },
+    {
+      "epoch": 114.5,
+      "grad_norm": 0.189453125,
+      "learning_rate": 2.1479614667495578e-05,
+      "loss": 0.0117,
+      "step": 1603
+    },
+    {
+      "epoch": 114.57142857142857,
+      "grad_norm": 0.1328125,
+      "learning_rate": 2.1473541682815257e-05,
+      "loss": 0.0117,
+      "step": 1604
+    },
+    {
+      "epoch": 114.64285714285714,
+      "grad_norm": 0.12109375,
+      "learning_rate": 2.1467464324539734e-05,
+      "loss": 0.0152,
+      "step": 1605
+    },
+    {
+      "epoch": 114.71428571428571,
+      "grad_norm": 0.30078125,
+      "learning_rate": 2.1461382595631036e-05,
+      "loss": 0.0166,
+      "step": 1606
+    },
+    {
+      "epoch": 114.78571428571429,
+      "grad_norm": 0.1904296875,
+      "learning_rate": 2.1455296499053324e-05,
+      "loss": 0.0126,
+      "step": 1607
+    },
+    {
+      "epoch": 114.85714285714286,
+      "grad_norm": 0.1201171875,
+      "learning_rate": 2.144920603777289e-05,
+      "loss": 0.0123,
+      "step": 1608
+    },
+    {
+      "epoch": 114.92857142857143,
+      "grad_norm": 0.138671875,
+      "learning_rate": 2.1443111214758148e-05,
+      "loss": 0.0127,
+      "step": 1609
+    },
+    {
+      "epoch": 115.0,
+      "grad_norm": 0.1669921875,
+      "learning_rate": 2.1437012032979646e-05,
+      "loss": 0.0117,
+      "step": 1610
+    },
+    {
+      "epoch": 115.07142857142857,
+      "grad_norm": 0.130859375,
+      "learning_rate": 2.1430908495410042e-05,
+      "loss": 0.0148,
+      "step": 1611
+    },
+    {
+      "epoch": 115.14285714285714,
+      "grad_norm": 0.11865234375,
+      "learning_rate": 2.1424800605024137e-05,
+      "loss": 0.0137,
+      "step": 1612
+    },
+    {
+      "epoch": 115.21428571428571,
+      "grad_norm": 0.11572265625,
+      "learning_rate": 2.141868836479883e-05,
+      "loss": 0.0142,
+      "step": 1613
+    },
+    {
+      "epoch": 115.28571428571429,
+      "grad_norm": 0.28515625,
+      "learning_rate": 2.1412571777713163e-05,
+      "loss": 0.0154,
+      "step": 1614
+    },
+    {
+      "epoch": 115.35714285714286,
+      "grad_norm": 0.16796875,
+      "learning_rate": 2.1406450846748282e-05,
+      "loss": 0.0167,
+      "step": 1615
+    },
+    {
+      "epoch": 115.42857142857143,
+      "grad_norm": 0.2021484375,
+      "learning_rate": 2.140032557488746e-05,
+      "loss": 0.0131,
+      "step": 1616
+    },
+    {
+      "epoch": 115.5,
+      "grad_norm": 0.1171875,
+      "learning_rate": 2.139419596511607e-05,
+      "loss": 0.0129,
+      "step": 1617
+    },
+    {
+      "epoch": 115.57142857142857,
+      "grad_norm": 0.208984375,
+      "learning_rate": 2.138806202042161e-05,
+      "loss": 0.0135,
+      "step": 1618
+    },
+    {
+      "epoch": 115.64285714285714,
+      "grad_norm": 0.10546875,
+      "learning_rate": 2.1381923743793704e-05,
+      "loss": 0.0129,
+      "step": 1619
+    },
+    {
+      "epoch": 115.71428571428571,
+      "grad_norm": 0.1396484375,
+      "learning_rate": 2.1375781138224062e-05,
+      "loss": 0.0147,
+      "step": 1620
+    },
+    {
+      "epoch": 115.78571428571429,
+      "grad_norm": 0.166015625,
+      "learning_rate": 2.136963420670651e-05,
+      "loss": 0.0151,
+      "step": 1621
+    },
+    {
+      "epoch": 115.85714285714286,
+      "grad_norm": 0.13671875,
+      "learning_rate": 2.1363482952237e-05,
+      "loss": 0.0136,
+      "step": 1622
+    },
+    {
+      "epoch": 115.92857142857143,
+      "grad_norm": 0.1357421875,
+      "learning_rate": 2.135732737781358e-05,
+      "loss": 0.016,
+      "step": 1623
+    },
+    {
+      "epoch": 116.0,
+      "grad_norm": 0.10888671875,
+      "learning_rate": 2.1351167486436388e-05,
+      "loss": 0.0146,
+      "step": 1624
+    },
+    {
+      "epoch": 116.07142857142857,
+      "grad_norm": 0.169921875,
+      "learning_rate": 2.1345003281107692e-05,
+      "loss": 0.0159,
+      "step": 1625
+    },
+    {
+      "epoch": 116.14285714285714,
+      "grad_norm": 0.412109375,
+      "learning_rate": 2.1338834764831845e-05,
+      "loss": 0.015,
+      "step": 1626
+    },
+    {
+      "epoch": 116.21428571428571,
+      "grad_norm": 0.1875,
+      "learning_rate": 2.1332661940615312e-05,
+      "loss": 0.0162,
+      "step": 1627
+    },
+    {
+      "epoch": 116.28571428571429,
+      "grad_norm": 0.1259765625,
+      "learning_rate": 2.1326484811466646e-05,
+      "loss": 0.0142,
+      "step": 1628
+    },
+    {
+      "epoch": 116.35714285714286,
+      "grad_norm": 0.12451171875,
+      "learning_rate": 2.132030338039651e-05,
+      "loss": 0.0155,
+      "step": 1629
+    },
+    {
+      "epoch": 116.42857142857143,
+      "grad_norm": 0.11865234375,
+      "learning_rate": 2.1314117650417657e-05,
+      "loss": 0.0137,
+      "step": 1630
+    },
+    {
+      "epoch": 116.5,
+      "grad_norm": 0.1591796875,
+      "learning_rate": 2.1307927624544934e-05,
+      "loss": 0.0145,
+      "step": 1631
+    },
+    {
+      "epoch": 116.57142857142857,
+      "grad_norm": 0.30859375,
+      "learning_rate": 2.130173330579529e-05,
+      "loss": 0.0151,
+      "step": 1632
+    },
+    {
+      "epoch": 116.64285714285714,
+      "grad_norm": 0.1455078125,
+      "learning_rate": 2.1295534697187756e-05,
+      "loss": 0.0124,
+      "step": 1633
+    },
+    {
+      "epoch": 116.71428571428571,
+      "grad_norm": 0.30078125,
+      "learning_rate": 2.128933180174346e-05,
+      "loss": 0.0137,
+      "step": 1634
+    },
+    {
+      "epoch": 116.78571428571429,
+      "grad_norm": 0.25,
+      "learning_rate": 2.1283124622485616e-05,
+      "loss": 0.0114,
+      "step": 1635
+    },
+    {
+      "epoch": 116.85714285714286,
+      "grad_norm": 0.1630859375,
+      "learning_rate": 2.1276913162439532e-05,
+      "loss": 0.0131,
+      "step": 1636
+    },
+    {
+      "epoch": 116.92857142857143,
+      "grad_norm": 0.267578125,
+      "learning_rate": 2.1270697424632593e-05,
+      "loss": 0.0124,
+      "step": 1637
+    },
+    {
+      "epoch": 117.0,
+      "grad_norm": 0.29296875,
+      "learning_rate": 2.126447741209428e-05,
+      "loss": 0.0156,
+      "step": 1638
+    },
+    {
+      "epoch": 117.07142857142857,
+      "grad_norm": 0.37109375,
+      "learning_rate": 2.1258253127856148e-05,
+      "loss": 0.0164,
+      "step": 1639
+    },
+    {
+      "epoch": 117.14285714285714,
+      "grad_norm": 0.228515625,
+      "learning_rate": 2.1252024574951834e-05,
+      "loss": 0.0133,
+      "step": 1640
+    },
+    {
+      "epoch": 117.21428571428571,
+      "grad_norm": 0.2265625,
+      "learning_rate": 2.124579175641707e-05,
+      "loss": 0.0127,
+      "step": 1641
+    },
+    {
+      "epoch": 117.28571428571429,
+      "grad_norm": 0.2470703125,
+      "learning_rate": 2.1239554675289645e-05,
+      "loss": 0.0138,
+      "step": 1642
+    },
+    {
+      "epoch": 117.35714285714286,
+      "grad_norm": 0.271484375,
+      "learning_rate": 2.123331333460944e-05,
+      "loss": 0.0146,
+      "step": 1643
+    },
+    {
+      "epoch": 117.42857142857143,
+      "grad_norm": 0.1103515625,
+      "learning_rate": 2.122706773741841e-05,
+      "loss": 0.0146,
+      "step": 1644
+    },
+    {
+      "epoch": 117.5,
+      "grad_norm": 0.14453125,
+      "learning_rate": 2.122081788676058e-05,
+      "loss": 0.0131,
+      "step": 1645
+    },
+    {
+      "epoch": 117.57142857142857,
+      "grad_norm": 0.1171875,
+      "learning_rate": 2.121456378568206e-05,
+      "loss": 0.0132,
+      "step": 1646
+    },
+    {
+      "epoch": 117.64285714285714,
+      "grad_norm": 0.2353515625,
+      "learning_rate": 2.120830543723101e-05,
+      "loss": 0.0141,
+      "step": 1647
+    },
+    {
+      "epoch": 117.71428571428571,
+      "grad_norm": 0.17578125,
+      "learning_rate": 2.1202042844457682e-05,
+      "loss": 0.0151,
+      "step": 1648
+    },
+    {
+      "epoch": 117.78571428571429,
+      "grad_norm": 0.15234375,
+      "learning_rate": 2.1195776010414388e-05,
+      "loss": 0.012,
+      "step": 1649
+    },
+    {
+      "epoch": 117.85714285714286,
+      "grad_norm": 0.248046875,
+      "learning_rate": 2.11895049381555e-05,
+      "loss": 0.0117,
+      "step": 1650
+    },
+    {
+      "epoch": 117.92857142857143,
+      "grad_norm": 0.11767578125,
+      "learning_rate": 2.1183229630737467e-05,
+      "loss": 0.0142,
+      "step": 1651
+    },
+    {
+      "epoch": 118.0,
+      "grad_norm": 0.150390625,
+      "learning_rate": 2.11769500912188e-05,
+      "loss": 0.0115,
+      "step": 1652
+    },
+    {
+      "epoch": 118.07142857142857,
+      "grad_norm": 0.162109375,
+      "learning_rate": 2.117066632266006e-05,
+      "loss": 0.0159,
+      "step": 1653
+    },
+    {
+      "epoch": 118.14285714285714,
+      "grad_norm": 0.154296875,
+      "learning_rate": 2.1164378328123898e-05,
+      "loss": 0.0141,
+      "step": 1654
+    },
+    {
+      "epoch": 118.21428571428571,
+      "grad_norm": 0.1611328125,
+      "learning_rate": 2.1158086110674987e-05,
+      "loss": 0.0149,
+      "step": 1655
+    },
+    {
+      "epoch": 118.28571428571429,
+      "grad_norm": 0.12353515625,
+      "learning_rate": 2.1151789673380086e-05,
+      "loss": 0.013,
+      "step": 1656
+    },
+    {
+      "epoch": 118.35714285714286,
+      "grad_norm": 0.1298828125,
+      "learning_rate": 2.1145489019308e-05,
+      "loss": 0.014,
+      "step": 1657
+    },
+    {
+      "epoch": 118.42857142857143,
+      "grad_norm": 0.1416015625,
+      "learning_rate": 2.11391841515296e-05,
+      "loss": 0.0143,
+      "step": 1658
+    },
+    {
+      "epoch": 118.5,
+      "grad_norm": 0.2470703125,
+      "learning_rate": 2.1132875073117785e-05,
+      "loss": 0.0129,
+      "step": 1659
+    },
+    {
+      "epoch": 118.57142857142857,
+      "grad_norm": 0.27734375,
+      "learning_rate": 2.112656178714754e-05,
+      "loss": 0.015,
+      "step": 1660
+    },
+    {
+      "epoch": 118.64285714285714,
+      "grad_norm": 0.2431640625,
+      "learning_rate": 2.1120244296695874e-05,
+      "loss": 0.013,
+      "step": 1661
+    },
+    {
+      "epoch": 118.71428571428571,
+      "grad_norm": 0.12109375,
+      "learning_rate": 2.1113922604841858e-05,
+      "loss": 0.0137,
+      "step": 1662
+    },
+    {
+      "epoch": 118.78571428571429,
+      "grad_norm": 0.134765625,
+      "learning_rate": 2.1107596714666604e-05,
+      "loss": 0.0152,
+      "step": 1663
+    },
+    {
+      "epoch": 118.85714285714286,
+      "grad_norm": 0.24609375,
+      "learning_rate": 2.1101266629253286e-05,
+      "loss": 0.011,
+      "step": 1664
+    },
+    {
+      "epoch": 118.92857142857143,
+      "grad_norm": 0.125,
+      "learning_rate": 2.1094932351687095e-05,
+      "loss": 0.0145,
+      "step": 1665
+    },
+    {
+      "epoch": 119.0,
+      "grad_norm": 0.28515625,
+      "learning_rate": 2.1088593885055288e-05,
+      "loss": 0.0154,
+      "step": 1666
+    },
+    {
+      "epoch": 119.07142857142857,
+      "grad_norm": 0.404296875,
+      "learning_rate": 2.1082251232447158e-05,
+      "loss": 0.0127,
+      "step": 1667
+    },
+    {
+      "epoch": 119.14285714285714,
+      "grad_norm": 0.2001953125,
+      "learning_rate": 2.107590439695404e-05,
+      "loss": 0.0176,
+      "step": 1668
+    },
+    {
+      "epoch": 119.21428571428571,
+      "grad_norm": 0.2158203125,
+      "learning_rate": 2.1069553381669296e-05,
+      "loss": 0.0133,
+      "step": 1669
+    },
+    {
+      "epoch": 119.28571428571429,
+      "grad_norm": 0.1513671875,
+      "learning_rate": 2.106319818968834e-05,
+      "loss": 0.0144,
+      "step": 1670
+    },
+    {
+      "epoch": 119.35714285714286,
+      "grad_norm": 0.19140625,
+      "learning_rate": 2.105683882410861e-05,
+      "loss": 0.0149,
+      "step": 1671
+    },
+    {
+      "epoch": 119.42857142857143,
+      "grad_norm": 0.185546875,
+      "learning_rate": 2.1050475288029586e-05,
+      "loss": 0.0108,
+      "step": 1672
+    },
+    {
+      "epoch": 119.5,
+      "grad_norm": 0.126953125,
+      "learning_rate": 2.1044107584552785e-05,
+      "loss": 0.0123,
+      "step": 1673
+    },
+    {
+      "epoch": 119.57142857142857,
+      "grad_norm": 0.123046875,
+      "learning_rate": 2.1037735716781736e-05,
+      "loss": 0.0146,
+      "step": 1674
+    },
+    {
+      "epoch": 119.64285714285714,
+      "grad_norm": 0.259765625,
+      "learning_rate": 2.1031359687822023e-05,
+      "loss": 0.0154,
+      "step": 1675
+    },
+    {
+      "epoch": 119.71428571428571,
+      "grad_norm": 0.09716796875,
+      "learning_rate": 2.1024979500781232e-05,
+      "loss": 0.0129,
+      "step": 1676
+    },
+    {
+      "epoch": 119.78571428571429,
+      "grad_norm": 0.134765625,
+      "learning_rate": 2.1018595158769e-05,
+      "loss": 0.0141,
+      "step": 1677
+    },
+    {
+      "epoch": 119.85714285714286,
+      "grad_norm": 0.1259765625,
+      "learning_rate": 2.101220666489697e-05,
+      "loss": 0.0145,
+      "step": 1678
+    },
+    {
+      "epoch": 119.92857142857143,
+      "grad_norm": 0.197265625,
+      "learning_rate": 2.1005814022278816e-05,
+      "loss": 0.0138,
+      "step": 1679
+    },
+    {
+      "epoch": 120.0,
+      "grad_norm": 0.107421875,
+      "learning_rate": 2.099941723403024e-05,
+      "loss": 0.0119,
+      "step": 1680
+    },
+    {
+      "epoch": 120.07142857142857,
+      "grad_norm": 0.3984375,
+      "learning_rate": 2.099301630326896e-05,
+      "loss": 0.0174,
+      "step": 1681
+    },
+    {
+      "epoch": 120.14285714285714,
+      "grad_norm": 0.328125,
+      "learning_rate": 2.0986611233114702e-05,
+      "loss": 0.0121,
+      "step": 1682
+    },
+    {
+      "epoch": 120.21428571428571,
+      "grad_norm": 0.1064453125,
+      "learning_rate": 2.0980202026689227e-05,
+      "loss": 0.0131,
+      "step": 1683
+    },
+    {
+      "epoch": 120.28571428571429,
+      "grad_norm": 0.1640625,
+      "learning_rate": 2.0973788687116304e-05,
+      "loss": 0.0157,
+      "step": 1684
+    },
+    {
+      "epoch": 120.35714285714286,
+      "grad_norm": 0.1015625,
+      "learning_rate": 2.0967371217521716e-05,
+      "loss": 0.0122,
+      "step": 1685
+    },
+    {
+      "epoch": 120.42857142857143,
+      "grad_norm": 0.1337890625,
+      "learning_rate": 2.096094962103326e-05,
+      "loss": 0.0119,
+      "step": 1686
+    },
+    {
+      "epoch": 120.5,
+      "grad_norm": 0.1337890625,
+      "learning_rate": 2.0954523900780742e-05,
+      "loss": 0.0134,
+      "step": 1687
+    },
+    {
+      "epoch": 120.57142857142857,
+      "grad_norm": 0.10498046875,
+      "learning_rate": 2.0948094059895984e-05,
+      "loss": 0.0115,
+      "step": 1688
+    },
+    {
+      "epoch": 120.64285714285714,
+      "grad_norm": 0.1328125,
+      "learning_rate": 2.0941660101512806e-05,
+      "loss": 0.0151,
+      "step": 1689
+    },
+    {
+      "epoch": 120.71428571428571,
+      "grad_norm": 0.287109375,
+      "learning_rate": 2.0935222028767047e-05,
+      "loss": 0.0135,
+      "step": 1690
+    },
+    {
+      "epoch": 120.78571428571429,
+      "grad_norm": 0.248046875,
+      "learning_rate": 2.092877984479654e-05,
+      "loss": 0.0141,
+      "step": 1691
+    },
+    {
+      "epoch": 120.85714285714286,
+      "grad_norm": 0.1552734375,
+      "learning_rate": 2.0922333552741136e-05,
+      "loss": 0.0157,
+      "step": 1692
+    },
+    {
+      "epoch": 120.92857142857143,
+      "grad_norm": 0.1533203125,
+      "learning_rate": 2.091588315574267e-05,
+      "loss": 0.0164,
+      "step": 1693
+    },
+    {
+      "epoch": 121.0,
+      "grad_norm": 0.1494140625,
+      "learning_rate": 2.0909428656944995e-05,
+      "loss": 0.0135,
+      "step": 1694
+    },
+    {
+      "epoch": 121.07142857142857,
+      "grad_norm": 0.1845703125,
+      "learning_rate": 2.0902970059493955e-05,
+      "loss": 0.0132,
+      "step": 1695
+    },
+    {
+      "epoch": 121.14285714285714,
+      "grad_norm": 0.18359375,
+      "learning_rate": 2.089650736653738e-05,
+      "loss": 0.0153,
+      "step": 1696
+    },
+    {
+      "epoch": 121.21428571428571,
+      "grad_norm": 0.1435546875,
+      "learning_rate": 2.089004058122512e-05,
+      "loss": 0.0157,
+      "step": 1697
+    },
+    {
+      "epoch": 121.28571428571429,
+      "grad_norm": 0.1083984375,
+      "learning_rate": 2.0883569706709e-05,
+      "loss": 0.0117,
+      "step": 1698
+    },
+    {
+      "epoch": 121.35714285714286,
+      "grad_norm": 0.32421875,
+      "learning_rate": 2.0877094746142857e-05,
+      "loss": 0.0142,
+      "step": 1699
+    },
+    {
+      "epoch": 121.42857142857143,
+      "grad_norm": 0.173828125,
+      "learning_rate": 2.0870615702682494e-05,
+      "loss": 0.015,
+      "step": 1700
+    },
+    {
+      "epoch": 121.5,
+      "grad_norm": 0.31640625,
+      "learning_rate": 2.086413257948573e-05,
+      "loss": 0.0157,
+      "step": 1701
+    },
+    {
+      "epoch": 121.57142857142857,
+      "grad_norm": 0.1181640625,
+      "learning_rate": 2.0857645379712353e-05,
+      "loss": 0.0133,
+      "step": 1702
+    },
+    {
+      "epoch": 121.64285714285714,
+      "grad_norm": 0.1328125,
+      "learning_rate": 2.0851154106524146e-05,
+      "loss": 0.0137,
+      "step": 1703
+    },
+    {
+      "epoch": 121.71428571428571,
+      "grad_norm": 0.10791015625,
+      "learning_rate": 2.0844658763084883e-05,
+      "loss": 0.0117,
+      "step": 1704
+    },
+    {
+      "epoch": 121.78571428571429,
+      "grad_norm": 0.267578125,
+      "learning_rate": 2.0838159352560308e-05,
+      "loss": 0.0136,
+      "step": 1705
+    },
+    {
+      "epoch": 121.85714285714286,
+      "grad_norm": 0.10693359375,
+      "learning_rate": 2.0831655878118155e-05,
+      "loss": 0.0145,
+      "step": 1706
+    },
+    {
+      "epoch": 121.92857142857143,
+      "grad_norm": 0.1474609375,
+      "learning_rate": 2.0825148342928146e-05,
+      "loss": 0.0149,
+      "step": 1707
+    },
+    {
+      "epoch": 122.0,
+      "grad_norm": 0.263671875,
+      "learning_rate": 2.081863675016197e-05,
+      "loss": 0.0144,
+      "step": 1708
+    },
+    {
+      "epoch": 122.07142857142857,
+      "grad_norm": 0.2255859375,
+      "learning_rate": 2.081212110299329e-05,
+      "loss": 0.0139,
+      "step": 1709
+    },
+    {
+      "epoch": 122.14285714285714,
+      "grad_norm": 0.1357421875,
+      "learning_rate": 2.080560140459777e-05,
+      "loss": 0.0139,
+      "step": 1710
+    },
+    {
+      "epoch": 122.21428571428571,
+      "grad_norm": 0.11279296875,
+      "learning_rate": 2.0799077658153022e-05,
+      "loss": 0.014,
+      "step": 1711
+    },
+    {
+      "epoch": 122.28571428571429,
+      "grad_norm": 0.1728515625,
+      "learning_rate": 2.079254986683864e-05,
+      "loss": 0.0123,
+      "step": 1712
+    },
+    {
+      "epoch": 122.35714285714286,
+      "grad_norm": 0.10400390625,
+      "learning_rate": 2.078601803383619e-05,
+      "loss": 0.0129,
+      "step": 1713
+    },
+    {
+      "epoch": 122.42857142857143,
+      "grad_norm": 0.1904296875,
+      "learning_rate": 2.077948216232921e-05,
+      "loss": 0.0141,
+      "step": 1714
+    },
+    {
+      "epoch": 122.5,
+      "grad_norm": 0.1640625,
+      "learning_rate": 2.077294225550321e-05,
+      "loss": 0.0113,
+      "step": 1715
+    },
+    {
+      "epoch": 122.57142857142857,
+      "grad_norm": 0.263671875,
+      "learning_rate": 2.0766398316545648e-05,
+      "loss": 0.015,
+      "step": 1716
+    },
+    {
+      "epoch": 122.64285714285714,
+      "grad_norm": 0.23828125,
+      "learning_rate": 2.0759850348645972e-05,
+      "loss": 0.0164,
+      "step": 1717
+    },
+    {
+      "epoch": 122.71428571428571,
+      "grad_norm": 0.265625,
+      "learning_rate": 2.075329835499558e-05,
+      "loss": 0.0158,
+      "step": 1718
+    },
+    {
+      "epoch": 122.78571428571429,
+      "grad_norm": 0.34375,
+      "learning_rate": 2.074674233878783e-05,
+      "loss": 0.0159,
+      "step": 1719
+    },
+    {
+      "epoch": 122.85714285714286,
+      "grad_norm": 0.130859375,
+      "learning_rate": 2.074018230321804e-05,
+      "loss": 0.0131,
+      "step": 1720
+    },
+    {
+      "epoch": 122.92857142857143,
+      "grad_norm": 0.134765625,
+      "learning_rate": 2.0733618251483506e-05,
+      "loss": 0.0137,
+      "step": 1721
+    },
+    {
+      "epoch": 123.0,
+      "grad_norm": 0.125,
+      "learning_rate": 2.072705018678346e-05,
+      "loss": 0.014,
+      "step": 1722
+    },
+    {
+      "epoch": 123.07142857142857,
+      "grad_norm": 0.1796875,
+      "learning_rate": 2.0720478112319092e-05,
+      "loss": 0.0144,
+      "step": 1723
+    },
+    {
+      "epoch": 123.14285714285714,
+      "grad_norm": 0.1279296875,
+      "learning_rate": 2.0713902031293558e-05,
+      "loss": 0.0137,
+      "step": 1724
+    },
+    {
+      "epoch": 123.21428571428571,
+      "grad_norm": 0.1318359375,
+      "learning_rate": 2.0707321946911957e-05,
+      "loss": 0.014,
+      "step": 1725
+    },
+    {
+      "epoch": 123.28571428571429,
+      "grad_norm": 0.22265625,
+      "learning_rate": 2.070073786238134e-05,
+      "loss": 0.0131,
+      "step": 1726
+    },
+    {
+      "epoch": 123.35714285714286,
+      "grad_norm": 0.1591796875,
+      "learning_rate": 2.069414978091072e-05,
+      "loss": 0.0135,
+      "step": 1727
+    },
+    {
+      "epoch": 123.42857142857143,
+      "grad_norm": 0.162109375,
+      "learning_rate": 2.0687557705711033e-05,
+      "loss": 0.0145,
+      "step": 1728
+    },
+    {
+      "epoch": 123.5,
+      "grad_norm": 0.1484375,
+      "learning_rate": 2.0680961639995185e-05,
+      "loss": 0.0122,
+      "step": 1729
+    },
+    {
+      "epoch": 123.57142857142857,
+      "grad_norm": 0.1328125,
+      "learning_rate": 2.067436158697802e-05,
+      "loss": 0.0152,
+      "step": 1730
+    },
+    {
+      "epoch": 123.64285714285714,
+      "grad_norm": 0.09521484375,
+      "learning_rate": 2.066775754987632e-05,
+      "loss": 0.0134,
+      "step": 1731
+    },
+    {
+      "epoch": 123.71428571428571,
+      "grad_norm": 0.16015625,
+      "learning_rate": 2.0661149531908812e-05,
+      "loss": 0.0163,
+      "step": 1732
+    },
+    {
+      "epoch": 123.78571428571429,
+      "grad_norm": 0.1435546875,
+      "learning_rate": 2.0654537536296164e-05,
+      "loss": 0.0152,
+      "step": 1733
+    },
+    {
+      "epoch": 123.85714285714286,
+      "grad_norm": 0.12890625,
+      "learning_rate": 2.0647921566260987e-05,
+      "loss": 0.0137,
+      "step": 1734
+    },
+    {
+      "epoch": 123.92857142857143,
+      "grad_norm": 0.138671875,
+      "learning_rate": 2.064130162502782e-05,
+      "loss": 0.0137,
+      "step": 1735
+    },
+    {
+      "epoch": 124.0,
+      "grad_norm": 0.10400390625,
+      "learning_rate": 2.0634677715823137e-05,
+      "loss": 0.0143,
+      "step": 1736
+    },
+    {
+      "epoch": 124.07142857142857,
+      "grad_norm": 0.1376953125,
+      "learning_rate": 2.062804984187536e-05,
+      "loss": 0.0154,
+      "step": 1737
+    },
+    {
+      "epoch": 124.14285714285714,
+      "grad_norm": 0.1328125,
+      "learning_rate": 2.0621418006414833e-05,
+      "loss": 0.0131,
+      "step": 1738
+    },
+    {
+      "epoch": 124.21428571428571,
+      "grad_norm": 0.20703125,
+      "learning_rate": 2.0614782212673827e-05,
+      "loss": 0.0121,
+      "step": 1739
+    },
+    {
+      "epoch": 124.28571428571429,
+      "grad_norm": 0.154296875,
+      "learning_rate": 2.0608142463886553e-05,
+      "loss": 0.0145,
+      "step": 1740
+    },
+    {
+      "epoch": 124.35714285714286,
+      "grad_norm": 0.2265625,
+      "learning_rate": 2.0601498763289138e-05,
+      "loss": 0.0144,
+      "step": 1741
+    },
+    {
+      "epoch": 124.42857142857143,
+      "grad_norm": 0.1484375,
+      "learning_rate": 2.0594851114119645e-05,
+      "loss": 0.0141,
+      "step": 1742
+    },
+    {
+      "epoch": 124.5,
+      "grad_norm": 0.2138671875,
+      "learning_rate": 2.0588199519618058e-05,
+      "loss": 0.0147,
+      "step": 1743
+    },
+    {
+      "epoch": 124.57142857142857,
+      "grad_norm": 0.109375,
+      "learning_rate": 2.0581543983026276e-05,
+      "loss": 0.013,
+      "step": 1744
+    },
+    {
+      "epoch": 124.64285714285714,
+      "grad_norm": 0.18359375,
+      "learning_rate": 2.0574884507588137e-05,
+      "loss": 0.0135,
+      "step": 1745
+    },
+    {
+      "epoch": 124.71428571428571,
+      "grad_norm": 0.1904296875,
+      "learning_rate": 2.0568221096549384e-05,
+      "loss": 0.0127,
+      "step": 1746
+    },
+    {
+      "epoch": 124.78571428571429,
+      "grad_norm": 0.12890625,
+      "learning_rate": 2.056155375315768e-05,
+      "loss": 0.0153,
+      "step": 1747
+    },
+    {
+      "epoch": 124.85714285714286,
+      "grad_norm": 0.19921875,
+      "learning_rate": 2.055488248066261e-05,
+      "loss": 0.0148,
+      "step": 1748
+    },
+    {
+      "epoch": 124.92857142857143,
+      "grad_norm": 0.162109375,
+      "learning_rate": 2.0548207282315675e-05,
+      "loss": 0.0157,
+      "step": 1749
+    },
+    {
+      "epoch": 125.0,
+      "grad_norm": 0.185546875,
+      "learning_rate": 2.0541528161370276e-05,
+      "loss": 0.0135,
+      "step": 1750
+    },
+    {
+      "epoch": 125.07142857142857,
+      "grad_norm": 0.12109375,
+      "learning_rate": 2.0534845121081742e-05,
+      "loss": 0.0142,
+      "step": 1751
+    },
+    {
+      "epoch": 125.14285714285714,
+      "grad_norm": 0.1669921875,
+      "learning_rate": 2.0528158164707308e-05,
+      "loss": 0.0134,
+      "step": 1752
+    },
+    {
+      "epoch": 125.21428571428571,
+      "grad_norm": 0.1396484375,
+      "learning_rate": 2.052146729550611e-05,
+      "loss": 0.0132,
+      "step": 1753
+    },
+    {
+      "epoch": 125.28571428571429,
+      "grad_norm": 0.193359375,
+      "learning_rate": 2.0514772516739194e-05,
+      "loss": 0.012,
+      "step": 1754
+    },
+    {
+      "epoch": 125.35714285714286,
+      "grad_norm": 0.10986328125,
+      "learning_rate": 2.050807383166952e-05,
+      "loss": 0.0132,
+      "step": 1755
+    },
+    {
+      "epoch": 125.42857142857143,
+      "grad_norm": 0.10986328125,
+      "learning_rate": 2.0501371243561946e-05,
+      "loss": 0.0126,
+      "step": 1756
+    },
+    {
+      "epoch": 125.5,
+      "grad_norm": 0.1630859375,
+      "learning_rate": 2.0494664755683226e-05,
+      "loss": 0.0149,
+      "step": 1757
+    },
+    {
+      "epoch": 125.57142857142857,
+      "grad_norm": 0.193359375,
+      "learning_rate": 2.0487954371302027e-05,
+      "loss": 0.0118,
+      "step": 1758
+    },
+    {
+      "epoch": 125.64285714285714,
+      "grad_norm": 0.09716796875,
+      "learning_rate": 2.0481240093688904e-05,
+      "loss": 0.0118,
+      "step": 1759
+    },
+    {
+      "epoch": 125.71428571428571,
+      "grad_norm": 0.1064453125,
+      "learning_rate": 2.047452192611632e-05,
+      "loss": 0.0118,
+      "step": 1760
+    },
+    {
+      "epoch": 125.78571428571429,
+      "grad_norm": 0.296875,
+      "learning_rate": 2.0467799871858624e-05,
+      "loss": 0.0149,
+      "step": 1761
+    },
+    {
+      "epoch": 125.85714285714286,
+      "grad_norm": 0.10693359375,
+      "learning_rate": 2.0461073934192064e-05,
+      "loss": 0.0107,
+      "step": 1762
+    },
+    {
+      "epoch": 125.92857142857143,
+      "grad_norm": 0.12255859375,
+      "learning_rate": 2.045434411639478e-05,
+      "loss": 0.0143,
+      "step": 1763
+    },
+    {
+      "epoch": 126.0,
+      "grad_norm": 0.15625,
+      "learning_rate": 2.0447610421746803e-05,
+      "loss": 0.0149,
+      "step": 1764
+    },
+    {
+      "epoch": 126.07142857142857,
+      "grad_norm": 0.1787109375,
+      "learning_rate": 2.0440872853530055e-05,
+      "loss": 0.0147,
+      "step": 1765
+    },
+    {
+      "epoch": 126.14285714285714,
+      "grad_norm": 0.1396484375,
+      "learning_rate": 2.0434131415028346e-05,
+      "loss": 0.014,
+      "step": 1766
+    },
+    {
+      "epoch": 126.21428571428571,
+      "grad_norm": 0.2216796875,
+      "learning_rate": 2.042738610952737e-05,
+      "loss": 0.013,
+      "step": 1767
+    },
+    {
+      "epoch": 126.28571428571429,
+      "grad_norm": 0.10546875,
+      "learning_rate": 2.0420636940314708e-05,
+      "loss": 0.0114,
+      "step": 1768
+    },
+    {
+      "epoch": 126.35714285714286,
+      "grad_norm": 0.130859375,
+      "learning_rate": 2.041388391067982e-05,
+      "loss": 0.0158,
+      "step": 1769
+    },
+    {
+      "epoch": 126.42857142857143,
+      "grad_norm": 0.11572265625,
+      "learning_rate": 2.0407127023914058e-05,
+      "loss": 0.0135,
+      "step": 1770
+    },
+    {
+      "epoch": 126.5,
+      "grad_norm": 0.1767578125,
+      "learning_rate": 2.0400366283310636e-05,
+      "loss": 0.0149,
+      "step": 1771
+    },
+    {
+      "epoch": 126.57142857142857,
+      "grad_norm": 0.10009765625,
+      "learning_rate": 2.0393601692164665e-05,
+      "loss": 0.0122,
+      "step": 1772
+    },
+    {
+      "epoch": 126.64285714285714,
+      "grad_norm": 0.12890625,
+      "learning_rate": 2.038683325377312e-05,
+      "loss": 0.014,
+      "step": 1773
+    },
+    {
+      "epoch": 126.71428571428571,
+      "grad_norm": 0.126953125,
+      "learning_rate": 2.0380060971434862e-05,
+      "loss": 0.0139,
+      "step": 1774
+    },
+    {
+      "epoch": 126.78571428571429,
+      "grad_norm": 0.125,
+      "learning_rate": 2.037328484845061e-05,
+      "loss": 0.0154,
+      "step": 1775
+    },
+    {
+      "epoch": 126.85714285714286,
+      "grad_norm": 0.1181640625,
+      "learning_rate": 2.036650488812297e-05,
+      "loss": 0.013,
+      "step": 1776
+    },
+    {
+      "epoch": 126.92857142857143,
+      "grad_norm": 0.11279296875,
+      "learning_rate": 2.035972109375641e-05,
+      "loss": 0.0137,
+      "step": 1777
+    },
+    {
+      "epoch": 127.0,
+      "grad_norm": 0.099609375,
+      "learning_rate": 2.035293346865727e-05,
+      "loss": 0.0126,
+      "step": 1778
+    },
+    {
+      "epoch": 127.07142857142857,
+      "grad_norm": 0.107421875,
+      "learning_rate": 2.0346142016133758e-05,
+      "loss": 0.0131,
+      "step": 1779
+    },
+    {
+      "epoch": 127.14285714285714,
+      "grad_norm": 0.1279296875,
+      "learning_rate": 2.033934673949594e-05,
+      "loss": 0.0119,
+      "step": 1780
+    },
+    {
+      "epoch": 127.21428571428571,
+      "grad_norm": 0.11572265625,
+      "learning_rate": 2.033254764205576e-05,
+      "loss": 0.0121,
+      "step": 1781
+    },
+    {
+      "epoch": 127.28571428571429,
+      "grad_norm": 0.09033203125,
+      "learning_rate": 2.0325744727127005e-05,
+      "loss": 0.0133,
+      "step": 1782
+    },
+    {
+      "epoch": 127.35714285714286,
+      "grad_norm": 0.140625,
+      "learning_rate": 2.031893799802534e-05,
+      "loss": 0.0156,
+      "step": 1783
+    },
+    {
+      "epoch": 127.42857142857143,
+      "grad_norm": 0.154296875,
+      "learning_rate": 2.031212745806828e-05,
+      "loss": 0.0131,
+      "step": 1784
+    },
+    {
+      "epoch": 127.5,
+      "grad_norm": 0.275390625,
+      "learning_rate": 2.0305313110575197e-05,
+      "loss": 0.0157,
+      "step": 1785
+    },
+    {
+      "epoch": 127.57142857142857,
+      "grad_norm": 0.1298828125,
+      "learning_rate": 2.029849495886733e-05,
+      "loss": 0.0162,
+      "step": 1786
+    },
+    {
+      "epoch": 127.64285714285714,
+      "grad_norm": 0.166015625,
+      "learning_rate": 2.0291673006267753e-05,
+      "loss": 0.0154,
+      "step": 1787
+    },
+    {
+      "epoch": 127.71428571428571,
+      "grad_norm": 0.12060546875,
+      "learning_rate": 2.028484725610141e-05,
+      "loss": 0.0162,
+      "step": 1788
+    },
+    {
+      "epoch": 127.78571428571429,
+      "grad_norm": 0.166015625,
+      "learning_rate": 2.027801771169508e-05,
+      "loss": 0.0135,
+      "step": 1789
+    },
+    {
+      "epoch": 127.85714285714286,
+      "grad_norm": 0.287109375,
+      "learning_rate": 2.0271184376377413e-05,
+      "loss": 0.0118,
+      "step": 1790
+    },
+    {
+      "epoch": 127.92857142857143,
+      "grad_norm": 0.20703125,
+      "learning_rate": 2.026434725347888e-05,
+      "loss": 0.0133,
+      "step": 1791
+    },
+    {
+      "epoch": 128.0,
+      "grad_norm": 0.265625,
+      "learning_rate": 2.0257506346331823e-05,
+      "loss": 0.013,
+      "step": 1792
+    },
+    {
+      "epoch": 128.07142857142858,
+      "grad_norm": 0.177734375,
+      "learning_rate": 2.025066165827041e-05,
+      "loss": 0.0135,
+      "step": 1793
+    },
+    {
+      "epoch": 128.14285714285714,
+      "grad_norm": 0.1259765625,
+      "learning_rate": 2.024381319263066e-05,
+      "loss": 0.0141,
+      "step": 1794
+    },
+    {
+      "epoch": 128.21428571428572,
+      "grad_norm": 0.12353515625,
+      "learning_rate": 2.0236960952750435e-05,
+      "loss": 0.0122,
+      "step": 1795
+    },
+    {
+      "epoch": 128.28571428571428,
+      "grad_norm": 0.10693359375,
+      "learning_rate": 2.0230104941969426e-05,
+      "loss": 0.0133,
+      "step": 1796
+    },
+    {
+      "epoch": 128.35714285714286,
+      "grad_norm": 0.13671875,
+      "learning_rate": 2.022324516362918e-05,
+      "loss": 0.0127,
+      "step": 1797
+    },
+    {
+      "epoch": 128.42857142857142,
+      "grad_norm": 0.1494140625,
+      "learning_rate": 2.0216381621073064e-05,
+      "loss": 0.0152,
+      "step": 1798
+    },
+    {
+      "epoch": 128.5,
+      "grad_norm": 0.162109375,
+      "learning_rate": 2.020951431764628e-05,
+      "loss": 0.0128,
+      "step": 1799
+    },
+    {
+      "epoch": 128.57142857142858,
+      "grad_norm": 0.1376953125,
+      "learning_rate": 2.020264325669588e-05,
+      "loss": 0.0123,
+      "step": 1800
+    },
+    {
+      "epoch": 128.64285714285714,
+      "grad_norm": 0.1591796875,
+      "learning_rate": 2.0195768441570727e-05,
+      "loss": 0.0108,
+      "step": 1801
+    },
+    {
+      "epoch": 128.71428571428572,
+      "grad_norm": 0.31640625,
+      "learning_rate": 2.018888987562153e-05,
+      "loss": 0.0141,
+      "step": 1802
+    },
+    {
+      "epoch": 128.78571428571428,
+      "grad_norm": 0.21875,
+      "learning_rate": 2.0182007562200805e-05,
+      "loss": 0.0157,
+      "step": 1803
+    },
+    {
+      "epoch": 128.85714285714286,
+      "grad_norm": 0.314453125,
+      "learning_rate": 2.017512150466292e-05,
+      "loss": 0.0134,
+      "step": 1804
+    },
+    {
+      "epoch": 128.92857142857142,
+      "grad_norm": 0.19921875,
+      "learning_rate": 2.0168231706364054e-05,
+      "loss": 0.0155,
+      "step": 1805
+    },
+    {
+      "epoch": 129.0,
+      "grad_norm": 0.158203125,
+      "learning_rate": 2.0161338170662208e-05,
+      "loss": 0.0148,
+      "step": 1806
+    },
+    {
+      "epoch": 129.07142857142858,
+      "grad_norm": 0.21484375,
+      "learning_rate": 2.0154440900917206e-05,
+      "loss": 0.0139,
+      "step": 1807
+    },
+    {
+      "epoch": 129.14285714285714,
+      "grad_norm": 0.302734375,
+      "learning_rate": 2.0147539900490697e-05,
+      "loss": 0.0178,
+      "step": 1808
+    },
+    {
+      "epoch": 129.21428571428572,
+      "grad_norm": 0.265625,
+      "learning_rate": 2.0140635172746146e-05,
+      "loss": 0.0149,
+      "step": 1809
+    },
+    {
+      "epoch": 129.28571428571428,
+      "grad_norm": 0.236328125,
+      "learning_rate": 2.0133726721048832e-05,
+      "loss": 0.0142,
+      "step": 1810
+    },
+    {
+      "epoch": 129.35714285714286,
+      "grad_norm": 0.16015625,
+      "learning_rate": 2.012681454876585e-05,
+      "loss": 0.0148,
+      "step": 1811
+    },
+    {
+      "epoch": 129.42857142857142,
+      "grad_norm": 0.2197265625,
+      "learning_rate": 2.01198986592661e-05,
+      "loss": 0.0156,
+      "step": 1812
+    },
+    {
+      "epoch": 129.5,
+      "grad_norm": 0.1796875,
+      "learning_rate": 2.011297905592032e-05,
+      "loss": 0.0125,
+      "step": 1813
+    },
+    {
+      "epoch": 129.57142857142858,
+      "grad_norm": 0.1640625,
+      "learning_rate": 2.0106055742101026e-05,
+      "loss": 0.0135,
+      "step": 1814
+    },
+    {
+      "epoch": 129.64285714285714,
+      "grad_norm": 0.1875,
+      "learning_rate": 2.009912872118257e-05,
+      "loss": 0.0138,
+      "step": 1815
+    },
+    {
+      "epoch": 129.71428571428572,
+      "grad_norm": 0.109375,
+      "learning_rate": 2.0092197996541085e-05,
+      "loss": 0.0132,
+      "step": 1816
+    },
+    {
+      "epoch": 129.78571428571428,
+      "grad_norm": 0.10009765625,
+      "learning_rate": 2.008526357155453e-05,
+      "loss": 0.0131,
+      "step": 1817
+    },
+    {
+      "epoch": 129.85714285714286,
+      "grad_norm": 0.146484375,
+      "learning_rate": 2.0078325449602655e-05,
+      "loss": 0.016,
+      "step": 1818
+    },
+    {
+      "epoch": 129.92857142857142,
+      "grad_norm": 0.1943359375,
+      "learning_rate": 2.007138363406702e-05,
+      "loss": 0.0134,
+      "step": 1819
+    },
+    {
+      "epoch": 130.0,
+      "grad_norm": 0.10888671875,
+      "learning_rate": 2.0064438128330975e-05,
+      "loss": 0.0143,
+      "step": 1820
+    },
+    {
+      "epoch": 130.07142857142858,
+      "grad_norm": 0.1689453125,
+      "learning_rate": 2.005748893577969e-05,
+      "loss": 0.0141,
+      "step": 1821
+    },
+    {
+      "epoch": 130.14285714285714,
+      "grad_norm": 0.1484375,
+      "learning_rate": 2.0050536059800098e-05,
+      "loss": 0.0134,
+      "step": 1822
+    },
+    {
+      "epoch": 130.21428571428572,
+      "grad_norm": 0.169921875,
+      "learning_rate": 2.0043579503780958e-05,
+      "loss": 0.0155,
+      "step": 1823
+    },
+    {
+      "epoch": 130.28571428571428,
+      "grad_norm": 0.119140625,
+      "learning_rate": 2.003661927111281e-05,
+      "loss": 0.0109,
+      "step": 1824
+    },
+    {
+      "epoch": 130.35714285714286,
+      "grad_norm": 0.1298828125,
+      "learning_rate": 2.002965536518798e-05,
+      "loss": 0.0129,
+      "step": 1825
+    },
+    {
+      "epoch": 130.42857142857142,
+      "grad_norm": 0.177734375,
+      "learning_rate": 2.0022687789400607e-05,
+      "loss": 0.0142,
+      "step": 1826
+    },
+    {
+      "epoch": 130.5,
+      "grad_norm": 0.255859375,
+      "learning_rate": 2.0015716547146586e-05,
+      "loss": 0.0175,
+      "step": 1827
+    },
+    {
+      "epoch": 130.57142857142858,
+      "grad_norm": 0.1787109375,
+      "learning_rate": 2.000874164182362e-05,
+      "loss": 0.0139,
+      "step": 1828
+    },
+    {
+      "epoch": 130.64285714285714,
+      "grad_norm": 0.150390625,
+      "learning_rate": 2.0001763076831195e-05,
+      "loss": 0.0118,
+      "step": 1829
+    },
+    {
+      "epoch": 130.71428571428572,
+      "grad_norm": 0.1748046875,
+      "learning_rate": 1.999478085557058e-05,
+      "loss": 0.0148,
+      "step": 1830
+    },
+    {
+      "epoch": 130.78571428571428,
+      "grad_norm": 0.263671875,
+      "learning_rate": 1.9987794981444823e-05,
+      "loss": 0.0131,
+      "step": 1831
+    },
+    {
+      "epoch": 130.85714285714286,
+      "grad_norm": 0.255859375,
+      "learning_rate": 1.9980805457858758e-05,
+      "loss": 0.0135,
+      "step": 1832
+    },
+    {
+      "epoch": 130.92857142857142,
+      "grad_norm": 0.134765625,
+      "learning_rate": 1.9973812288218987e-05,
+      "loss": 0.0135,
+      "step": 1833
+    },
+    {
+      "epoch": 131.0,
+      "grad_norm": 0.1298828125,
+      "learning_rate": 1.9966815475933902e-05,
+      "loss": 0.0139,
+      "step": 1834
+    },
+    {
+      "epoch": 131.07142857142858,
+      "grad_norm": 0.1591796875,
+      "learning_rate": 1.9959815024413657e-05,
+      "loss": 0.0125,
+      "step": 1835
+    },
+    {
+      "epoch": 131.14285714285714,
+      "grad_norm": 0.1201171875,
+      "learning_rate": 1.99528109370702e-05,
+      "loss": 0.0165,
+      "step": 1836
+    },
+    {
+      "epoch": 131.21428571428572,
+      "grad_norm": 0.1552734375,
+      "learning_rate": 1.9945803217317226e-05,
+      "loss": 0.0167,
+      "step": 1837
+    },
+    {
+      "epoch": 131.28571428571428,
+      "grad_norm": 0.1533203125,
+      "learning_rate": 1.993879186857022e-05,
+      "loss": 0.0153,
+      "step": 1838
+    },
+    {
+      "epoch": 131.35714285714286,
+      "grad_norm": 0.1162109375,
+      "learning_rate": 1.9931776894246417e-05,
+      "loss": 0.0146,
+      "step": 1839
+    },
+    {
+      "epoch": 131.42857142857142,
+      "grad_norm": 0.1552734375,
+      "learning_rate": 1.992475829776484e-05,
+      "loss": 0.0132,
+      "step": 1840
+    },
+    {
+      "epoch": 131.5,
+      "grad_norm": 0.1318359375,
+      "learning_rate": 1.991773608254627e-05,
+      "loss": 0.0149,
+      "step": 1841
+    },
+    {
+      "epoch": 131.57142857142858,
+      "grad_norm": 0.1484375,
+      "learning_rate": 1.9910710252013244e-05,
+      "loss": 0.0147,
+      "step": 1842
+    },
+    {
+      "epoch": 131.64285714285714,
+      "grad_norm": 0.216796875,
+      "learning_rate": 1.9903680809590065e-05,
+      "loss": 0.0136,
+      "step": 1843
+    },
+    {
+      "epoch": 131.71428571428572,
+      "grad_norm": 0.1259765625,
+      "learning_rate": 1.98966477587028e-05,
+      "loss": 0.0126,
+      "step": 1844
+    },
+    {
+      "epoch": 131.78571428571428,
+      "grad_norm": 0.107421875,
+      "learning_rate": 1.9889611102779276e-05,
+      "loss": 0.0128,
+      "step": 1845
+    },
+    {
+      "epoch": 131.85714285714286,
+      "grad_norm": 0.1416015625,
+      "learning_rate": 1.988257084524907e-05,
+      "loss": 0.0132,
+      "step": 1846
+    },
+    {
+      "epoch": 131.92857142857142,
+      "grad_norm": 0.28125,
+      "learning_rate": 1.9875526989543517e-05,
+      "loss": 0.0132,
+      "step": 1847
+    },
+    {
+      "epoch": 132.0,
+      "grad_norm": 0.1572265625,
+      "learning_rate": 1.986847953909571e-05,
+      "loss": 0.0152,
+      "step": 1848
+    },
+    {
+      "epoch": 132.07142857142858,
+      "grad_norm": 0.189453125,
+      "learning_rate": 1.9861428497340483e-05,
+      "loss": 0.0113,
+      "step": 1849
+    },
+    {
+      "epoch": 132.14285714285714,
+      "grad_norm": 0.1884765625,
+      "learning_rate": 1.9854373867714443e-05,
+      "loss": 0.0127,
+      "step": 1850
+    },
+    {
+      "epoch": 132.21428571428572,
+      "grad_norm": 0.1025390625,
+      "learning_rate": 1.9847315653655915e-05,
+      "loss": 0.0138,
+      "step": 1851
+    },
+    {
+      "epoch": 132.28571428571428,
+      "grad_norm": 0.146484375,
+      "learning_rate": 1.9840253858604995e-05,
+      "loss": 0.0112,
+      "step": 1852
+    },
+    {
+      "epoch": 132.35714285714286,
+      "grad_norm": 0.1337890625,
+      "learning_rate": 1.9833188486003516e-05,
+      "loss": 0.0144,
+      "step": 1853
+    },
+    {
+      "epoch": 132.42857142857142,
+      "grad_norm": 0.11669921875,
+      "learning_rate": 1.9826119539295054e-05,
+      "loss": 0.0114,
+      "step": 1854
+    },
+    {
+      "epoch": 132.5,
+      "grad_norm": 0.298828125,
+      "learning_rate": 1.9819047021924926e-05,
+      "loss": 0.0141,
+      "step": 1855
+    },
+    {
+      "epoch": 132.57142857142858,
+      "grad_norm": 0.10205078125,
+      "learning_rate": 1.9811970937340196e-05,
+      "loss": 0.0129,
+      "step": 1856
+    },
+    {
+      "epoch": 132.64285714285714,
+      "grad_norm": 0.1015625,
+      "learning_rate": 1.9804891288989653e-05,
+      "loss": 0.0131,
+      "step": 1857
+    },
+    {
+      "epoch": 132.71428571428572,
+      "grad_norm": 0.125,
+      "learning_rate": 1.9797808080323837e-05,
+      "loss": 0.016,
+      "step": 1858
+    },
+    {
+      "epoch": 132.78571428571428,
+      "grad_norm": 0.19921875,
+      "learning_rate": 1.979072131479502e-05,
+      "loss": 0.0118,
+      "step": 1859
+    },
+    {
+      "epoch": 132.85714285714286,
+      "grad_norm": 0.111328125,
+      "learning_rate": 1.9783630995857202e-05,
+      "loss": 0.0126,
+      "step": 1860
+    },
+    {
+      "epoch": 132.92857142857142,
+      "grad_norm": 0.23828125,
+      "learning_rate": 1.977653712696612e-05,
+      "loss": 0.0148,
+      "step": 1861
+    },
+    {
+      "epoch": 133.0,
+      "grad_norm": 0.1357421875,
+      "learning_rate": 1.9769439711579234e-05,
+      "loss": 0.0142,
+      "step": 1862
+    },
+    {
+      "epoch": 133.07142857142858,
+      "grad_norm": 0.12451171875,
+      "learning_rate": 1.976233875315575e-05,
+      "loss": 0.012,
+      "step": 1863
+    },
+    {
+      "epoch": 133.14285714285714,
+      "grad_norm": 0.12890625,
+      "learning_rate": 1.9755234255156576e-05,
+      "loss": 0.0145,
+      "step": 1864
+    },
+    {
+      "epoch": 133.21428571428572,
+      "grad_norm": 0.1533203125,
+      "learning_rate": 1.9748126221044367e-05,
+      "loss": 0.013,
+      "step": 1865
+    },
+    {
+      "epoch": 133.28571428571428,
+      "grad_norm": 0.197265625,
+      "learning_rate": 1.9741014654283486e-05,
+      "loss": 0.0173,
+      "step": 1866
+    },
+    {
+      "epoch": 133.35714285714286,
+      "grad_norm": 0.2001953125,
+      "learning_rate": 1.973389955834003e-05,
+      "loss": 0.0124,
+      "step": 1867
+    },
+    {
+      "epoch": 133.42857142857142,
+      "grad_norm": 0.12353515625,
+      "learning_rate": 1.9726780936681807e-05,
+      "loss": 0.0143,
+      "step": 1868
+    },
+    {
+      "epoch": 133.5,
+      "grad_norm": 0.1435546875,
+      "learning_rate": 1.9719658792778344e-05,
+      "loss": 0.0124,
+      "step": 1869
+    },
+    {
+      "epoch": 133.57142857142858,
+      "grad_norm": 0.1474609375,
+      "learning_rate": 1.9712533130100898e-05,
+      "loss": 0.0153,
+      "step": 1870
+    },
+    {
+      "epoch": 133.64285714285714,
+      "grad_norm": 0.11669921875,
+      "learning_rate": 1.970540395212242e-05,
+      "loss": 0.0129,
+      "step": 1871
+    },
+    {
+      "epoch": 133.71428571428572,
+      "grad_norm": 0.146484375,
+      "learning_rate": 1.9698271262317587e-05,
+      "loss": 0.0132,
+      "step": 1872
+    },
+    {
+      "epoch": 133.78571428571428,
+      "grad_norm": 0.12060546875,
+      "learning_rate": 1.9691135064162784e-05,
+      "loss": 0.013,
+      "step": 1873
+    },
+    {
+      "epoch": 133.85714285714286,
+      "grad_norm": 0.267578125,
+      "learning_rate": 1.968399536113611e-05,
+      "loss": 0.011,
+      "step": 1874
+    },
+    {
+      "epoch": 133.92857142857142,
+      "grad_norm": 0.1796875,
+      "learning_rate": 1.9676852156717375e-05,
+      "loss": 0.0118,
+      "step": 1875
+    },
+    {
+      "epoch": 134.0,
+      "grad_norm": 0.33203125,
+      "learning_rate": 1.9669705454388076e-05,
+      "loss": 0.0151,
+      "step": 1876
+    },
+    {
+      "epoch": 134.07142857142858,
+      "grad_norm": 0.265625,
+      "learning_rate": 1.9662555257631445e-05,
+      "loss": 0.0177,
+      "step": 1877
+    },
+    {
+      "epoch": 134.14285714285714,
+      "grad_norm": 0.103515625,
+      "learning_rate": 1.9655401569932388e-05,
+      "loss": 0.0146,
+      "step": 1878
+    },
+    {
+      "epoch": 134.21428571428572,
+      "grad_norm": 0.10205078125,
+      "learning_rate": 1.9648244394777533e-05,
+      "loss": 0.0115,
+      "step": 1879
+    },
+    {
+      "epoch": 134.28571428571428,
+      "grad_norm": 0.1201171875,
+      "learning_rate": 1.9641083735655193e-05,
+      "loss": 0.0116,
+      "step": 1880
+    },
+    {
+      "epoch": 134.35714285714286,
+      "grad_norm": 0.11376953125,
+      "learning_rate": 1.96339195960554e-05,
+      "loss": 0.015,
+      "step": 1881
+    },
+    {
+      "epoch": 134.42857142857142,
+      "grad_norm": 0.1953125,
+      "learning_rate": 1.9626751979469855e-05,
+      "loss": 0.013,
+      "step": 1882
+    },
+    {
+      "epoch": 134.5,
+      "grad_norm": 0.11767578125,
+      "learning_rate": 1.961958088939197e-05,
+      "loss": 0.0142,
+      "step": 1883
+    },
+    {
+      "epoch": 134.57142857142858,
+      "grad_norm": 0.30859375,
+      "learning_rate": 1.961240632931685e-05,
+      "loss": 0.015,
+      "step": 1884
+    },
+    {
+      "epoch": 134.64285714285714,
+      "grad_norm": 0.34765625,
+      "learning_rate": 1.9605228302741288e-05,
+      "loss": 0.0134,
+      "step": 1885
+    },
+    {
+      "epoch": 134.71428571428572,
+      "grad_norm": 0.125,
+      "learning_rate": 1.9598046813163766e-05,
+      "loss": 0.0146,
+      "step": 1886
+    },
+    {
+      "epoch": 134.78571428571428,
+      "grad_norm": 0.1328125,
+      "learning_rate": 1.9590861864084458e-05,
+      "loss": 0.0142,
+      "step": 1887
+    },
+    {
+      "epoch": 134.85714285714286,
+      "grad_norm": 0.11669921875,
+      "learning_rate": 1.9583673459005212e-05,
+      "loss": 0.0127,
+      "step": 1888
+    },
+    {
+      "epoch": 134.92857142857142,
+      "grad_norm": 0.158203125,
+      "learning_rate": 1.9576481601429584e-05,
+      "loss": 0.0148,
+      "step": 1889
+    },
+    {
+      "epoch": 135.0,
+      "grad_norm": 0.140625,
+      "learning_rate": 1.9569286294862788e-05,
+      "loss": 0.0125,
+      "step": 1890
+    },
+    {
+      "epoch": 135.07142857142858,
+      "grad_norm": 0.10693359375,
+      "learning_rate": 1.9562087542811725e-05,
+      "loss": 0.0128,
+      "step": 1891
+    },
+    {
+      "epoch": 135.14285714285714,
+      "grad_norm": 0.1279296875,
+      "learning_rate": 1.955488534878499e-05,
+      "loss": 0.0126,
+      "step": 1892
+    },
+    {
+      "epoch": 135.21428571428572,
+      "grad_norm": 0.310546875,
+      "learning_rate": 1.9547679716292834e-05,
+      "loss": 0.0164,
+      "step": 1893
+    },
+    {
+      "epoch": 135.28571428571428,
+      "grad_norm": 0.1240234375,
+      "learning_rate": 1.954047064884721e-05,
+      "loss": 0.0131,
+      "step": 1894
+    },
+    {
+      "epoch": 135.35714285714286,
+      "grad_norm": 0.1357421875,
+      "learning_rate": 1.953325814996171e-05,
+      "loss": 0.0112,
+      "step": 1895
+    },
+    {
+      "epoch": 135.42857142857142,
+      "grad_norm": 0.1962890625,
+      "learning_rate": 1.9526042223151634e-05,
+      "loss": 0.0127,
+      "step": 1896
+    },
+    {
+      "epoch": 135.5,
+      "grad_norm": 0.12890625,
+      "learning_rate": 1.951882287193393e-05,
+      "loss": 0.011,
+      "step": 1897
+    },
+    {
+      "epoch": 135.57142857142858,
+      "grad_norm": 0.42578125,
+      "learning_rate": 1.9511600099827225e-05,
+      "loss": 0.019,
+      "step": 1898
+    },
+    {
+      "epoch": 135.64285714285714,
+      "grad_norm": 0.11962890625,
+      "learning_rate": 1.950437391035181e-05,
+      "loss": 0.0114,
+      "step": 1899
+    },
+    {
+      "epoch": 135.71428571428572,
+      "grad_norm": 0.1328125,
+      "learning_rate": 1.949714430702964e-05,
+      "loss": 0.0142,
+      "step": 1900
+    },
+    {
+      "epoch": 135.78571428571428,
+      "grad_norm": 0.12255859375,
+      "learning_rate": 1.9489911293384337e-05,
+      "loss": 0.0134,
+      "step": 1901
+    },
+    {
+      "epoch": 135.85714285714286,
+      "grad_norm": 0.15625,
+      "learning_rate": 1.9482674872941183e-05,
+      "loss": 0.0114,
+      "step": 1902
+    },
+    {
+      "epoch": 135.92857142857142,
+      "grad_norm": 0.11865234375,
+      "learning_rate": 1.9475435049227125e-05,
+      "loss": 0.0112,
+      "step": 1903
+    },
+    {
+      "epoch": 136.0,
+      "grad_norm": 0.2431640625,
+      "learning_rate": 1.946819182577076e-05,
+      "loss": 0.0115,
+      "step": 1904
+    },
+    {
+      "epoch": 136.07142857142858,
+      "grad_norm": 0.10205078125,
+      "learning_rate": 1.9460945206102352e-05,
+      "loss": 0.0134,
+      "step": 1905
+    },
+    {
+      "epoch": 136.14285714285714,
+      "grad_norm": 0.265625,
+      "learning_rate": 1.9453695193753812e-05,
+      "loss": 0.0132,
+      "step": 1906
+    },
+    {
+      "epoch": 136.21428571428572,
+      "grad_norm": 0.10693359375,
+      "learning_rate": 1.9446441792258707e-05,
+      "loss": 0.0123,
+      "step": 1907
+    },
+    {
+      "epoch": 136.28571428571428,
+      "grad_norm": 0.1669921875,
+      "learning_rate": 1.943918500515226e-05,
+      "loss": 0.0131,
+      "step": 1908
+    },
+    {
+      "epoch": 136.35714285714286,
+      "grad_norm": 0.20703125,
+      "learning_rate": 1.9431924835971342e-05,
+      "loss": 0.0139,
+      "step": 1909
+    },
+    {
+      "epoch": 136.42857142857142,
+      "grad_norm": 0.138671875,
+      "learning_rate": 1.9424661288254467e-05,
+      "loss": 0.0119,
+      "step": 1910
+    },
+    {
+      "epoch": 136.5,
+      "grad_norm": 0.1552734375,
+      "learning_rate": 1.9417394365541803e-05,
+      "loss": 0.0144,
+      "step": 1911
+    },
+    {
+      "epoch": 136.57142857142858,
+      "grad_norm": 0.248046875,
+      "learning_rate": 1.9410124071375158e-05,
+      "loss": 0.0149,
+      "step": 1912
+    },
+    {
+      "epoch": 136.64285714285714,
+      "grad_norm": 0.2080078125,
+      "learning_rate": 1.9402850409297986e-05,
+      "loss": 0.017,
+      "step": 1913
+    },
+    {
+      "epoch": 136.71428571428572,
+      "grad_norm": 0.1630859375,
+      "learning_rate": 1.939557338285538e-05,
+      "loss": 0.0127,
+      "step": 1914
+    },
+    {
+      "epoch": 136.78571428571428,
+      "grad_norm": 0.134765625,
+      "learning_rate": 1.9388292995594076e-05,
+      "loss": 0.0155,
+      "step": 1915
+    },
+    {
+      "epoch": 136.85714285714286,
+      "grad_norm": 0.134765625,
+      "learning_rate": 1.9381009251062447e-05,
+      "loss": 0.0147,
+      "step": 1916
+    },
+    {
+      "epoch": 136.92857142857142,
+      "grad_norm": 0.2216796875,
+      "learning_rate": 1.93737221528105e-05,
+      "loss": 0.0155,
+      "step": 1917
+    },
+    {
+      "epoch": 137.0,
+      "grad_norm": 0.1171875,
+      "learning_rate": 1.9366431704389874e-05,
+      "loss": 0.0134,
+      "step": 1918
+    },
+    {
+      "epoch": 137.07142857142858,
+      "grad_norm": 0.302734375,
+      "learning_rate": 1.9359137909353857e-05,
+      "loss": 0.0144,
+      "step": 1919
+    },
+    {
+      "epoch": 137.14285714285714,
+      "grad_norm": 0.1650390625,
+      "learning_rate": 1.9351840771257345e-05,
+      "loss": 0.0124,
+      "step": 1920
+    },
+    {
+      "epoch": 137.21428571428572,
+      "grad_norm": 0.1513671875,
+      "learning_rate": 1.934454029365688e-05,
+      "loss": 0.0145,
+      "step": 1921
+    },
+    {
+      "epoch": 137.28571428571428,
+      "grad_norm": 0.173828125,
+      "learning_rate": 1.933723648011062e-05,
+      "loss": 0.0126,
+      "step": 1922
+    },
+    {
+      "epoch": 137.35714285714286,
+      "grad_norm": 0.1552734375,
+      "learning_rate": 1.9329929334178366e-05,
+      "loss": 0.0167,
+      "step": 1923
+    },
+    {
+      "epoch": 137.42857142857142,
+      "grad_norm": 0.1103515625,
+      "learning_rate": 1.9322618859421522e-05,
+      "loss": 0.0126,
+      "step": 1924
+    },
+    {
+      "epoch": 137.5,
+      "grad_norm": 0.1064453125,
+      "learning_rate": 1.9315305059403127e-05,
+      "loss": 0.0127,
+      "step": 1925
+    },
+    {
+      "epoch": 137.57142857142858,
+      "grad_norm": 0.12890625,
+      "learning_rate": 1.930798793768784e-05,
+      "loss": 0.0136,
+      "step": 1926
+    },
+    {
+      "epoch": 137.64285714285714,
+      "grad_norm": 0.150390625,
+      "learning_rate": 1.9300667497841934e-05,
+      "loss": 0.0135,
+      "step": 1927
+    },
+    {
+      "epoch": 137.71428571428572,
+      "grad_norm": 0.1396484375,
+      "learning_rate": 1.929334374343331e-05,
+      "loss": 0.0129,
+      "step": 1928
+    },
+    {
+      "epoch": 137.78571428571428,
+      "grad_norm": 0.12109375,
+      "learning_rate": 1.9286016678031472e-05,
+      "loss": 0.0148,
+      "step": 1929
+    },
+    {
+      "epoch": 137.85714285714286,
+      "grad_norm": 0.171875,
+      "learning_rate": 1.9278686305207537e-05,
+      "loss": 0.0127,
+      "step": 1930
+    },
+    {
+      "epoch": 137.92857142857142,
+      "grad_norm": 0.17578125,
+      "learning_rate": 1.927135262853425e-05,
+      "loss": 0.0123,
+      "step": 1931
+    },
+    {
+      "epoch": 138.0,
+      "grad_norm": 0.1650390625,
+      "learning_rate": 1.9264015651585948e-05,
+      "loss": 0.0125,
+      "step": 1932
+    },
+    {
+      "epoch": 138.07142857142858,
+      "grad_norm": 0.2275390625,
+      "learning_rate": 1.925667537793859e-05,
+      "loss": 0.0143,
+      "step": 1933
+    },
+    {
+      "epoch": 138.14285714285714,
+      "grad_norm": 0.11572265625,
+      "learning_rate": 1.9249331811169735e-05,
+      "loss": 0.0155,
+      "step": 1934
+    },
+    {
+      "epoch": 138.21428571428572,
+      "grad_norm": 0.130859375,
+      "learning_rate": 1.9241984954858543e-05,
+      "loss": 0.0136,
+      "step": 1935
+    },
+    {
+      "epoch": 138.28571428571428,
+      "grad_norm": 0.177734375,
+      "learning_rate": 1.9234634812585788e-05,
+      "loss": 0.0154,
+      "step": 1936
+    },
+    {
+      "epoch": 138.35714285714286,
+      "grad_norm": 0.2734375,
+      "learning_rate": 1.9227281387933842e-05,
+      "loss": 0.0148,
+      "step": 1937
+    },
+    {
+      "epoch": 138.42857142857142,
+      "grad_norm": 0.09619140625,
+      "learning_rate": 1.921992468448667e-05,
+      "loss": 0.0123,
+      "step": 1938
+    },
+    {
+      "epoch": 138.5,
+      "grad_norm": 0.1396484375,
+      "learning_rate": 1.921256470582984e-05,
+      "loss": 0.0141,
+      "step": 1939
+    },
+    {
+      "epoch": 138.57142857142858,
+      "grad_norm": 0.3359375,
+      "learning_rate": 1.9205201455550514e-05,
+      "loss": 0.0126,
+      "step": 1940
+    },
+    {
+      "epoch": 138.64285714285714,
+      "grad_norm": 0.123046875,
+      "learning_rate": 1.9197834937237457e-05,
+      "loss": 0.0131,
+      "step": 1941
+    },
+    {
+      "epoch": 138.71428571428572,
+      "grad_norm": 0.1171875,
+      "learning_rate": 1.9190465154481018e-05,
+      "loss": 0.0121,
+      "step": 1942
+    },
+    {
+      "epoch": 138.78571428571428,
+      "grad_norm": 0.2138671875,
+      "learning_rate": 1.918309211087314e-05,
+      "loss": 0.0139,
+      "step": 1943
+    },
+    {
+      "epoch": 138.85714285714286,
+      "grad_norm": 0.19921875,
+      "learning_rate": 1.917571581000736e-05,
+      "loss": 0.0156,
+      "step": 1944
+    },
+    {
+      "epoch": 138.92857142857142,
+      "grad_norm": 0.11181640625,
+      "learning_rate": 1.9168336255478787e-05,
+      "loss": 0.0123,
+      "step": 1945
+    },
+    {
+      "epoch": 139.0,
+      "grad_norm": 0.22265625,
+      "learning_rate": 1.9160953450884125e-05,
+      "loss": 0.0126,
+      "step": 1946
+    },
+    {
+      "epoch": 139.07142857142858,
+      "grad_norm": 0.10595703125,
+      "learning_rate": 1.9153567399821674e-05,
+      "loss": 0.0118,
+      "step": 1947
+    },
+    {
+      "epoch": 139.14285714285714,
+      "grad_norm": 0.22265625,
+      "learning_rate": 1.91461781058913e-05,
+      "loss": 0.0118,
+      "step": 1948
+    },
+    {
+      "epoch": 139.21428571428572,
+      "grad_norm": 0.1328125,
+      "learning_rate": 1.913878557269446e-05,
+      "loss": 0.0133,
+      "step": 1949
+    },
+    {
+      "epoch": 139.28571428571428,
+      "grad_norm": 0.1611328125,
+      "learning_rate": 1.9131389803834177e-05,
+      "loss": 0.0118,
+      "step": 1950
+    },
+    {
+      "epoch": 139.35714285714286,
+      "grad_norm": 0.1357421875,
+      "learning_rate": 1.912399080291506e-05,
+      "loss": 0.0152,
+      "step": 1951
+    },
+    {
+      "epoch": 139.42857142857142,
+      "grad_norm": 0.1884765625,
+      "learning_rate": 1.9116588573543305e-05,
+      "loss": 0.0133,
+      "step": 1952
+    },
+    {
+      "epoch": 139.5,
+      "grad_norm": 0.13671875,
+      "learning_rate": 1.9109183119326644e-05,
+      "loss": 0.0101,
+      "step": 1953
+    },
+    {
+      "epoch": 139.57142857142858,
+      "grad_norm": 0.158203125,
+      "learning_rate": 1.9101774443874432e-05,
+      "loss": 0.0129,
+      "step": 1954
+    },
+    {
+      "epoch": 139.64285714285714,
+      "grad_norm": 0.10498046875,
+      "learning_rate": 1.9094362550797547e-05,
+      "loss": 0.0154,
+      "step": 1955
+    },
+    {
+      "epoch": 139.71428571428572,
+      "grad_norm": 0.1318359375,
+      "learning_rate": 1.9086947443708472e-05,
+      "loss": 0.0129,
+      "step": 1956
+    },
+    {
+      "epoch": 139.78571428571428,
+      "grad_norm": 0.1328125,
+      "learning_rate": 1.9079529126221223e-05,
+      "loss": 0.0133,
+      "step": 1957
+    },
+    {
+      "epoch": 139.85714285714286,
+      "grad_norm": 0.11669921875,
+      "learning_rate": 1.9072107601951417e-05,
+      "loss": 0.0139,
+      "step": 1958
+    },
+    {
+      "epoch": 139.92857142857142,
+      "grad_norm": 0.177734375,
+      "learning_rate": 1.90646828745162e-05,
+      "loss": 0.0151,
+      "step": 1959
+    },
+    {
+      "epoch": 140.0,
+      "grad_norm": 0.158203125,
+      "learning_rate": 1.9057254947534293e-05,
+      "loss": 0.0131,
+      "step": 1960
+    },
+    {
+      "epoch": 140.07142857142858,
+      "grad_norm": 0.16796875,
+      "learning_rate": 1.9049823824625993e-05,
+      "loss": 0.0136,
+      "step": 1961
+    },
+    {
+      "epoch": 140.14285714285714,
+      "grad_norm": 0.2421875,
+      "learning_rate": 1.904238950941312e-05,
+      "loss": 0.0158,
+      "step": 1962
+    },
+    {
+      "epoch": 140.21428571428572,
+      "grad_norm": 0.1240234375,
+      "learning_rate": 1.9034952005519082e-05,
+      "loss": 0.0118,
+      "step": 1963
+    },
+    {
+      "epoch": 140.28571428571428,
+      "grad_norm": 0.146484375,
+      "learning_rate": 1.9027511316568824e-05,
+      "loss": 0.0146,
+      "step": 1964
+    },
+    {
+      "epoch": 140.35714285714286,
+      "grad_norm": 0.2353515625,
+      "learning_rate": 1.902006744618885e-05,
+      "loss": 0.0165,
+      "step": 1965
+    },
+    {
+      "epoch": 140.42857142857142,
+      "grad_norm": 0.2060546875,
+      "learning_rate": 1.901262039800721e-05,
+      "loss": 0.0128,
+      "step": 1966
+    },
+    {
+      "epoch": 140.5,
+      "grad_norm": 0.1494140625,
+      "learning_rate": 1.9005170175653498e-05,
+      "loss": 0.0135,
+      "step": 1967
+    },
+    {
+      "epoch": 140.57142857142858,
+      "grad_norm": 0.2265625,
+      "learning_rate": 1.8997716782758867e-05,
+      "loss": 0.0128,
+      "step": 1968
+    },
+    {
+      "epoch": 140.64285714285714,
+      "grad_norm": 0.240234375,
+      "learning_rate": 1.899026022295602e-05,
+      "loss": 0.0137,
+      "step": 1969
+    },
+    {
+      "epoch": 140.71428571428572,
+      "grad_norm": 0.10595703125,
+      "learning_rate": 1.8982800499879178e-05,
+      "loss": 0.0153,
+      "step": 1970
+    },
+    {
+      "epoch": 140.78571428571428,
+      "grad_norm": 0.1328125,
+      "learning_rate": 1.897533761716413e-05,
+      "loss": 0.0136,
+      "step": 1971
+    },
+    {
+      "epoch": 140.85714285714286,
+      "grad_norm": 0.10986328125,
+      "learning_rate": 1.8967871578448187e-05,
+      "loss": 0.0149,
+      "step": 1972
+    },
+    {
+      "epoch": 140.92857142857142,
+      "grad_norm": 0.1962890625,
+      "learning_rate": 1.8960402387370214e-05,
+      "loss": 0.0132,
+      "step": 1973
+    },
+    {
+      "epoch": 141.0,
+      "grad_norm": 0.142578125,
+      "learning_rate": 1.8952930047570595e-05,
+      "loss": 0.0122,
+      "step": 1974
+    },
+    {
+      "epoch": 141.07142857142858,
+      "grad_norm": 0.322265625,
+      "learning_rate": 1.8945454562691266e-05,
+      "loss": 0.0159,
+      "step": 1975
+    },
+    {
+      "epoch": 141.14285714285714,
+      "grad_norm": 0.1806640625,
+      "learning_rate": 1.893797593637568e-05,
+      "loss": 0.0152,
+      "step": 1976
+    },
+    {
+      "epoch": 141.21428571428572,
+      "grad_norm": 0.2060546875,
+      "learning_rate": 1.893049417226883e-05,
+      "loss": 0.012,
+      "step": 1977
+    },
+    {
+      "epoch": 141.28571428571428,
+      "grad_norm": 0.12255859375,
+      "learning_rate": 1.8923009274017245e-05,
+      "loss": 0.012,
+      "step": 1978
+    },
+    {
+      "epoch": 141.35714285714286,
+      "grad_norm": 0.173828125,
+      "learning_rate": 1.8915521245268964e-05,
+      "loss": 0.0138,
+      "step": 1979
+    },
+    {
+      "epoch": 141.42857142857142,
+      "grad_norm": 0.1669921875,
+      "learning_rate": 1.8908030089673563e-05,
+      "loss": 0.0114,
+      "step": 1980
+    },
+    {
+      "epoch": 141.5,
+      "grad_norm": 0.1572265625,
+      "learning_rate": 1.8900535810882146e-05,
+      "loss": 0.0142,
+      "step": 1981
+    },
+    {
+      "epoch": 141.57142857142858,
+      "grad_norm": 0.1240234375,
+      "learning_rate": 1.8893038412547327e-05,
+      "loss": 0.0128,
+      "step": 1982
+    },
+    {
+      "epoch": 141.64285714285714,
+      "grad_norm": 0.1201171875,
+      "learning_rate": 1.888553789832325e-05,
+      "loss": 0.0135,
+      "step": 1983
+    },
+    {
+      "epoch": 141.71428571428572,
+      "grad_norm": 0.26953125,
+      "learning_rate": 1.8878034271865576e-05,
+      "loss": 0.0157,
+      "step": 1984
+    },
+    {
+      "epoch": 141.78571428571428,
+      "grad_norm": 0.26171875,
+      "learning_rate": 1.8870527536831473e-05,
+      "loss": 0.0171,
+      "step": 1985
+    },
+    {
+      "epoch": 141.85714285714286,
+      "grad_norm": 0.1298828125,
+      "learning_rate": 1.8863017696879642e-05,
+      "loss": 0.0125,
+      "step": 1986
+    },
+    {
+      "epoch": 141.92857142857142,
+      "grad_norm": 0.1337890625,
+      "learning_rate": 1.8855504755670287e-05,
+      "loss": 0.0169,
+      "step": 1987
+    },
+    {
+      "epoch": 142.0,
+      "grad_norm": 0.2294921875,
+      "learning_rate": 1.8847988716865115e-05,
+      "loss": 0.0141,
+      "step": 1988
+    },
+    {
+      "epoch": 142.07142857142858,
+      "grad_norm": 0.1318359375,
+      "learning_rate": 1.8840469584127367e-05,
+      "loss": 0.0158,
+      "step": 1989
+    },
+    {
+      "epoch": 142.14285714285714,
+      "grad_norm": 0.16796875,
+      "learning_rate": 1.8832947361121763e-05,
+      "loss": 0.013,
+      "step": 1990
+    },
+    {
+      "epoch": 142.21428571428572,
+      "grad_norm": 0.1181640625,
+      "learning_rate": 1.8825422051514546e-05,
+      "loss": 0.0127,
+      "step": 1991
+    },
+    {
+      "epoch": 142.28571428571428,
+      "grad_norm": 0.10205078125,
+      "learning_rate": 1.8817893658973475e-05,
+      "loss": 0.0111,
+      "step": 1992
+    },
+    {
+      "epoch": 142.35714285714286,
+      "grad_norm": 0.2021484375,
+      "learning_rate": 1.8810362187167774e-05,
+      "loss": 0.0183,
+      "step": 1993
+    },
+    {
+      "epoch": 142.42857142857142,
+      "grad_norm": 0.173828125,
+      "learning_rate": 1.8802827639768206e-05,
+      "loss": 0.016,
+      "step": 1994
+    },
+    {
+      "epoch": 142.5,
+      "grad_norm": 0.1650390625,
+      "learning_rate": 1.879529002044701e-05,
+      "loss": 0.0139,
+      "step": 1995
+    },
+    {
+      "epoch": 142.57142857142858,
+      "grad_norm": 0.203125,
+      "learning_rate": 1.878774933287794e-05,
+      "loss": 0.0137,
+      "step": 1996
+    },
+    {
+      "epoch": 142.64285714285714,
+      "grad_norm": 0.12890625,
+      "learning_rate": 1.8780205580736228e-05,
+      "loss": 0.0142,
+      "step": 1997
+    },
+    {
+      "epoch": 142.71428571428572,
+      "grad_norm": 0.1630859375,
+      "learning_rate": 1.8772658767698607e-05,
+      "loss": 0.0118,
+      "step": 1998
+    },
+    {
+      "epoch": 142.78571428571428,
+      "grad_norm": 0.134765625,
+      "learning_rate": 1.8765108897443303e-05,
+      "loss": 0.0145,
+      "step": 1999
+    },
+    {
+      "epoch": 142.85714285714286,
+      "grad_norm": 0.10302734375,
+      "learning_rate": 1.8757555973650036e-05,
+      "loss": 0.0148,
+      "step": 2000
+    },
+    {
+      "epoch": 142.92857142857142,
+      "grad_norm": 0.1435546875,
+      "learning_rate": 1.8750000000000002e-05,
+      "loss": 0.0142,
+      "step": 2001
+    },
+    {
+      "epoch": 143.0,
+      "grad_norm": 0.109375,
+      "learning_rate": 1.8742440980175896e-05,
+      "loss": 0.0128,
+      "step": 2002
+    },
+    {
+      "epoch": 143.07142857142858,
+      "grad_norm": 0.1201171875,
+      "learning_rate": 1.8734878917861885e-05,
+      "loss": 0.0127,
+      "step": 2003
+    },
+    {
+      "epoch": 143.14285714285714,
+      "grad_norm": 0.1708984375,
+      "learning_rate": 1.8727313816743637e-05,
+      "loss": 0.0128,
+      "step": 2004
+    },
+    {
+      "epoch": 143.21428571428572,
+      "grad_norm": 0.1279296875,
+      "learning_rate": 1.8719745680508282e-05,
+      "loss": 0.0158,
+      "step": 2005
+    },
+    {
+      "epoch": 143.28571428571428,
+      "grad_norm": 0.208984375,
+      "learning_rate": 1.8712174512844445e-05,
+      "loss": 0.0152,
+      "step": 2006
+    },
+    {
+      "epoch": 143.35714285714286,
+      "grad_norm": 0.248046875,
+      "learning_rate": 1.8704600317442213e-05,
+      "loss": 0.0163,
+      "step": 2007
+    },
+    {
+      "epoch": 143.42857142857142,
+      "grad_norm": 0.326171875,
+      "learning_rate": 1.869702309799316e-05,
+      "loss": 0.0115,
+      "step": 2008
+    },
+    {
+      "epoch": 143.5,
+      "grad_norm": 0.11279296875,
+      "learning_rate": 1.8689442858190332e-05,
+      "loss": 0.0138,
+      "step": 2009
+    },
+    {
+      "epoch": 143.57142857142858,
+      "grad_norm": 0.1572265625,
+      "learning_rate": 1.8681859601728248e-05,
+      "loss": 0.014,
+      "step": 2010
+    },
+    {
+      "epoch": 143.64285714285714,
+      "grad_norm": 0.1513671875,
+      "learning_rate": 1.8674273332302888e-05,
+      "loss": 0.0145,
+      "step": 2011
+    },
+    {
+      "epoch": 143.71428571428572,
+      "grad_norm": 0.10009765625,
+      "learning_rate": 1.8666684053611712e-05,
+      "loss": 0.013,
+      "step": 2012
+    },
+    {
+      "epoch": 143.78571428571428,
+      "grad_norm": 0.1494140625,
+      "learning_rate": 1.8659091769353644e-05,
+      "loss": 0.012,
+      "step": 2013
+    },
+    {
+      "epoch": 143.85714285714286,
+      "grad_norm": 0.1201171875,
+      "learning_rate": 1.865149648322907e-05,
+      "loss": 0.0104,
+      "step": 2014
+    },
+    {
+      "epoch": 143.92857142857142,
+      "grad_norm": 0.1728515625,
+      "learning_rate": 1.8643898198939837e-05,
+      "loss": 0.0111,
+      "step": 2015
+    },
+    {
+      "epoch": 144.0,
+      "grad_norm": 0.1376953125,
+      "learning_rate": 1.863629692018926e-05,
+      "loss": 0.0143,
+      "step": 2016
+    },
+    {
+      "epoch": 144.07142857142858,
+      "grad_norm": 0.302734375,
+      "learning_rate": 1.862869265068211e-05,
+      "loss": 0.0164,
+      "step": 2017
+    },
+    {
+      "epoch": 144.14285714285714,
+      "grad_norm": 0.12890625,
+      "learning_rate": 1.8621085394124616e-05,
+      "loss": 0.0111,
+      "step": 2018
+    },
+    {
+      "epoch": 144.21428571428572,
+      "grad_norm": 0.115234375,
+      "learning_rate": 1.8613475154224454e-05,
+      "loss": 0.0117,
+      "step": 2019
+    },
+    {
+      "epoch": 144.28571428571428,
+      "grad_norm": 0.1796875,
+      "learning_rate": 1.8605861934690775e-05,
+      "loss": 0.0138,
+      "step": 2020
+    },
+    {
+      "epoch": 144.35714285714286,
+      "grad_norm": 0.166015625,
+      "learning_rate": 1.859824573923416e-05,
+      "loss": 0.0161,
+      "step": 2021
+    },
+    {
+      "epoch": 144.42857142857142,
+      "grad_norm": 0.1220703125,
+      "learning_rate": 1.8590626571566654e-05,
+      "loss": 0.0109,
+      "step": 2022
+    },
+    {
+      "epoch": 144.5,
+      "grad_norm": 0.115234375,
+      "learning_rate": 1.858300443540175e-05,
+      "loss": 0.0154,
+      "step": 2023
+    },
+    {
+      "epoch": 144.57142857142858,
+      "grad_norm": 0.12890625,
+      "learning_rate": 1.857537933445438e-05,
+      "loss": 0.0164,
+      "step": 2024
+    },
+    {
+      "epoch": 144.64285714285714,
+      "grad_norm": 0.146484375,
+      "learning_rate": 1.8567751272440925e-05,
+      "loss": 0.0118,
+      "step": 2025
+    },
+    {
+      "epoch": 144.71428571428572,
+      "grad_norm": 0.1337890625,
+      "learning_rate": 1.8560120253079215e-05,
+      "loss": 0.0146,
+      "step": 2026
+    },
+    {
+      "epoch": 144.78571428571428,
+      "grad_norm": 0.11962890625,
+      "learning_rate": 1.855248628008851e-05,
+      "loss": 0.0134,
+      "step": 2027
+    },
+    {
+      "epoch": 144.85714285714286,
+      "grad_norm": 0.10595703125,
+      "learning_rate": 1.854484935718952e-05,
+      "loss": 0.0151,
+      "step": 2028
+    },
+    {
+      "epoch": 144.92857142857142,
+      "grad_norm": 0.1455078125,
+      "learning_rate": 1.8537209488104383e-05,
+      "loss": 0.0133,
+      "step": 2029
+    },
+    {
+      "epoch": 145.0,
+      "grad_norm": 0.126953125,
+      "learning_rate": 1.852956667655669e-05,
+      "loss": 0.0129,
+      "step": 2030
+    },
+    {
+      "epoch": 145.07142857142858,
+      "grad_norm": 0.1669921875,
+      "learning_rate": 1.852192092627144e-05,
+      "loss": 0.0107,
+      "step": 2031
+    },
+    {
+      "epoch": 145.14285714285714,
+      "grad_norm": 0.11376953125,
+      "learning_rate": 1.8514272240975093e-05,
+      "loss": 0.0125,
+      "step": 2032
+    },
+    {
+      "epoch": 145.21428571428572,
+      "grad_norm": 0.23046875,
+      "learning_rate": 1.850662062439551e-05,
+      "loss": 0.016,
+      "step": 2033
+    },
+    {
+      "epoch": 145.28571428571428,
+      "grad_norm": 0.11328125,
+      "learning_rate": 1.849896608026201e-05,
+      "loss": 0.0118,
+      "step": 2034
+    },
+    {
+      "epoch": 145.35714285714286,
+      "grad_norm": 0.1005859375,
+      "learning_rate": 1.8491308612305314e-05,
+      "loss": 0.014,
+      "step": 2035
+    },
+    {
+      "epoch": 145.42857142857142,
+      "grad_norm": 0.1416015625,
+      "learning_rate": 1.848364822425759e-05,
+      "loss": 0.0121,
+      "step": 2036
+    },
+    {
+      "epoch": 145.5,
+      "grad_norm": 0.2236328125,
+      "learning_rate": 1.8475984919852404e-05,
+      "loss": 0.0134,
+      "step": 2037
+    },
+    {
+      "epoch": 145.57142857142858,
+      "grad_norm": 0.1376953125,
+      "learning_rate": 1.8468318702824765e-05,
+      "loss": 0.0162,
+      "step": 2038
+    },
+    {
+      "epoch": 145.64285714285714,
+      "grad_norm": 0.1162109375,
+      "learning_rate": 1.84606495769111e-05,
+      "loss": 0.0106,
+      "step": 2039
+    },
+    {
+      "epoch": 145.71428571428572,
+      "grad_norm": 0.130859375,
+      "learning_rate": 1.845297754584924e-05,
+      "loss": 0.0146,
+      "step": 2040
+    },
+    {
+      "epoch": 145.78571428571428,
+      "grad_norm": 0.15625,
+      "learning_rate": 1.8445302613378444e-05,
+      "loss": 0.0129,
+      "step": 2041
+    },
+    {
+      "epoch": 145.85714285714286,
+      "grad_norm": 0.115234375,
+      "learning_rate": 1.8437624783239372e-05,
+      "loss": 0.0129,
+      "step": 2042
+    },
+    {
+      "epoch": 145.92857142857142,
+      "grad_norm": 0.126953125,
+      "learning_rate": 1.8429944059174116e-05,
+      "loss": 0.0139,
+      "step": 2043
+    },
+    {
+      "epoch": 146.0,
+      "grad_norm": 0.177734375,
+      "learning_rate": 1.8422260444926165e-05,
+      "loss": 0.014,
+      "step": 2044
+    },
+    {
+      "epoch": 146.07142857142858,
+      "grad_norm": 0.11279296875,
+      "learning_rate": 1.8414573944240414e-05,
+      "loss": 0.0119,
+      "step": 2045
+    },
+    {
+      "epoch": 146.14285714285714,
+      "grad_norm": 0.1416015625,
+      "learning_rate": 1.840688456086318e-05,
+      "loss": 0.0154,
+      "step": 2046
+    },
+    {
+      "epoch": 146.21428571428572,
+      "grad_norm": 0.1142578125,
+      "learning_rate": 1.8399192298542163e-05,
+      "loss": 0.0118,
+      "step": 2047
+    },
+    {
+      "epoch": 146.28571428571428,
+      "grad_norm": 0.1748046875,
+      "learning_rate": 1.839149716102648e-05,
+      "loss": 0.0149,
+      "step": 2048
+    },
+    {
+      "epoch": 146.35714285714286,
+      "grad_norm": 0.11181640625,
+      "learning_rate": 1.8383799152066662e-05,
+      "loss": 0.0126,
+      "step": 2049
+    },
+    {
+      "epoch": 146.42857142857142,
+      "grad_norm": 0.310546875,
+      "learning_rate": 1.8376098275414606e-05,
+      "loss": 0.0132,
+      "step": 2050
+    },
+    {
+      "epoch": 146.5,
+      "grad_norm": 0.298828125,
+      "learning_rate": 1.8368394534823635e-05,
+      "loss": 0.0103,
+      "step": 2051
+    },
+    {
+      "epoch": 146.57142857142858,
+      "grad_norm": 0.1640625,
+      "learning_rate": 1.8360687934048462e-05,
+      "loss": 0.0148,
+      "step": 2052
+    },
+    {
+      "epoch": 146.64285714285714,
+      "grad_norm": 0.23828125,
+      "learning_rate": 1.8352978476845182e-05,
+      "loss": 0.0143,
+      "step": 2053
+    },
+    {
+      "epoch": 146.71428571428572,
+      "grad_norm": 0.111328125,
+      "learning_rate": 1.83452661669713e-05,
+      "loss": 0.0156,
+      "step": 2054
+    },
+    {
+      "epoch": 146.78571428571428,
+      "grad_norm": 0.26953125,
+      "learning_rate": 1.8337551008185694e-05,
+      "loss": 0.0162,
+      "step": 2055
+    },
+    {
+      "epoch": 146.85714285714286,
+      "grad_norm": 0.1123046875,
+      "learning_rate": 1.832983300424864e-05,
+      "loss": 0.0148,
+      "step": 2056
+    },
+    {
+      "epoch": 146.92857142857142,
+      "grad_norm": 0.119140625,
+      "learning_rate": 1.8322112158921804e-05,
+      "loss": 0.0121,
+      "step": 2057
+    },
+    {
+      "epoch": 147.0,
+      "grad_norm": 0.1533203125,
+      "learning_rate": 1.831438847596823e-05,
+      "loss": 0.0144,
+      "step": 2058
+    },
+    {
+      "epoch": 147.07142857142858,
+      "grad_norm": 0.220703125,
+      "learning_rate": 1.830666195915235e-05,
+      "loss": 0.0129,
+      "step": 2059
+    },
+    {
+      "epoch": 147.14285714285714,
+      "grad_norm": 0.150390625,
+      "learning_rate": 1.8298932612239964e-05,
+      "loss": 0.015,
+      "step": 2060
+    },
+    {
+      "epoch": 147.21428571428572,
+      "grad_norm": 0.146484375,
+      "learning_rate": 1.8291200438998273e-05,
+      "loss": 0.0135,
+      "step": 2061
+    },
+    {
+      "epoch": 147.28571428571428,
+      "grad_norm": 0.1025390625,
+      "learning_rate": 1.8283465443195837e-05,
+      "loss": 0.0118,
+      "step": 2062
+    },
+    {
+      "epoch": 147.35714285714286,
+      "grad_norm": 0.1875,
+      "learning_rate": 1.827572762860261e-05,
+      "loss": 0.0129,
+      "step": 2063
+    },
+    {
+      "epoch": 147.42857142857142,
+      "grad_norm": 0.1982421875,
+      "learning_rate": 1.8267986998989896e-05,
+      "loss": 0.0127,
+      "step": 2064
+    },
+    {
+      "epoch": 147.5,
+      "grad_norm": 0.1513671875,
+      "learning_rate": 1.8260243558130388e-05,
+      "loss": 0.0127,
+      "step": 2065
+    },
+    {
+      "epoch": 147.57142857142858,
+      "grad_norm": 0.1337890625,
+      "learning_rate": 1.8252497309798147e-05,
+      "loss": 0.0155,
+      "step": 2066
+    },
+    {
+      "epoch": 147.64285714285714,
+      "grad_norm": 0.154296875,
+      "learning_rate": 1.8244748257768602e-05,
+      "loss": 0.0141,
+      "step": 2067
+    },
+    {
+      "epoch": 147.71428571428572,
+      "grad_norm": 0.2080078125,
+      "learning_rate": 1.823699640581854e-05,
+      "loss": 0.0134,
+      "step": 2068
+    },
+    {
+      "epoch": 147.78571428571428,
+      "grad_norm": 0.259765625,
+      "learning_rate": 1.8229241757726124e-05,
+      "loss": 0.0161,
+      "step": 2069
+    },
+    {
+      "epoch": 147.85714285714286,
+      "grad_norm": 0.255859375,
+      "learning_rate": 1.8221484317270876e-05,
+      "loss": 0.0118,
+      "step": 2070
+    },
+    {
+      "epoch": 147.92857142857142,
+      "grad_norm": 0.2890625,
+      "learning_rate": 1.821372408823368e-05,
+      "loss": 0.0127,
+      "step": 2071
+    },
+    {
+      "epoch": 148.0,
+      "grad_norm": 0.216796875,
+      "learning_rate": 1.8205961074396772e-05,
+      "loss": 0.0154,
+      "step": 2072
+    },
+    {
+      "epoch": 148.07142857142858,
+      "grad_norm": 0.162109375,
+      "learning_rate": 1.8198195279543754e-05,
+      "loss": 0.0158,
+      "step": 2073
+    },
+    {
+      "epoch": 148.14285714285714,
+      "grad_norm": 0.1123046875,
+      "learning_rate": 1.8190426707459572e-05,
+      "loss": 0.0148,
+      "step": 2074
+    },
+    {
+      "epoch": 148.21428571428572,
+      "grad_norm": 0.12353515625,
+      "learning_rate": 1.818265536193055e-05,
+      "loss": 0.0125,
+      "step": 2075
+    },
+    {
+      "epoch": 148.28571428571428,
+      "grad_norm": 0.1396484375,
+      "learning_rate": 1.8174881246744335e-05,
+      "loss": 0.0114,
+      "step": 2076
+    },
+    {
+      "epoch": 148.35714285714286,
+      "grad_norm": 0.244140625,
+      "learning_rate": 1.816710436568994e-05,
+      "loss": 0.0135,
+      "step": 2077
+    },
+    {
+      "epoch": 148.42857142857142,
+      "grad_norm": 0.234375,
+      "learning_rate": 1.8159324722557725e-05,
+      "loss": 0.0157,
+      "step": 2078
+    },
+    {
+      "epoch": 148.5,
+      "grad_norm": 0.220703125,
+      "learning_rate": 1.8151542321139385e-05,
+      "loss": 0.0161,
+      "step": 2079
+    },
+    {
+      "epoch": 148.57142857142858,
+      "grad_norm": 0.0908203125,
+      "learning_rate": 1.814375716522798e-05,
+      "loss": 0.0105,
+      "step": 2080
+    },
+    {
+      "epoch": 148.64285714285714,
+      "grad_norm": 0.17578125,
+      "learning_rate": 1.8135969258617884e-05,
+      "loss": 0.0132,
+      "step": 2081
+    },
+    {
+      "epoch": 148.71428571428572,
+      "grad_norm": 0.099609375,
+      "learning_rate": 1.8128178605104844e-05,
+      "loss": 0.0143,
+      "step": 2082
+    },
+    {
+      "epoch": 148.78571428571428,
+      "grad_norm": 0.11474609375,
+      "learning_rate": 1.8120385208485924e-05,
+      "loss": 0.0141,
+      "step": 2083
+    },
+    {
+      "epoch": 148.85714285714286,
+      "grad_norm": 0.11669921875,
+      "learning_rate": 1.811258907255953e-05,
+      "loss": 0.0134,
+      "step": 2084
+    },
+    {
+      "epoch": 148.92857142857142,
+      "grad_norm": 0.1474609375,
+      "learning_rate": 1.8104790201125408e-05,
+      "loss": 0.0173,
+      "step": 2085
+    },
+    {
+      "epoch": 149.0,
+      "grad_norm": 0.16015625,
+      "learning_rate": 1.809698859798462e-05,
+      "loss": 0.0164,
+      "step": 2086
+    },
+    {
+      "epoch": 149.07142857142858,
+      "grad_norm": 0.1357421875,
+      "learning_rate": 1.808918426693959e-05,
+      "loss": 0.0117,
+      "step": 2087
+    },
+    {
+      "epoch": 149.14285714285714,
+      "grad_norm": 0.1982421875,
+      "learning_rate": 1.8081377211794042e-05,
+      "loss": 0.012,
+      "step": 2088
+    },
+    {
+      "epoch": 149.21428571428572,
+      "grad_norm": 0.11279296875,
+      "learning_rate": 1.8073567436353054e-05,
+      "loss": 0.0129,
+      "step": 2089
+    },
+    {
+      "epoch": 149.28571428571428,
+      "grad_norm": 0.1201171875,
+      "learning_rate": 1.8065754944422996e-05,
+      "loss": 0.0121,
+      "step": 2090
+    },
+    {
+      "epoch": 149.35714285714286,
+      "grad_norm": 0.11328125,
+      "learning_rate": 1.8057939739811595e-05,
+      "loss": 0.0132,
+      "step": 2091
+    },
+    {
+      "epoch": 149.42857142857142,
+      "grad_norm": 0.10546875,
+      "learning_rate": 1.8050121826327883e-05,
+      "loss": 0.0116,
+      "step": 2092
+    },
+    {
+      "epoch": 149.5,
+      "grad_norm": 0.125,
+      "learning_rate": 1.8042301207782214e-05,
+      "loss": 0.0144,
+      "step": 2093
+    },
+    {
+      "epoch": 149.57142857142858,
+      "grad_norm": 0.1953125,
+      "learning_rate": 1.8034477887986274e-05,
+      "loss": 0.0162,
+      "step": 2094
+    },
+    {
+      "epoch": 149.64285714285714,
+      "grad_norm": 0.1298828125,
+      "learning_rate": 1.8026651870753035e-05,
+      "loss": 0.011,
+      "step": 2095
+    },
+    {
+      "epoch": 149.71428571428572,
+      "grad_norm": 0.11083984375,
+      "learning_rate": 1.8018823159896816e-05,
+      "loss": 0.0118,
+      "step": 2096
+    },
+    {
+      "epoch": 149.78571428571428,
+      "grad_norm": 0.1103515625,
+      "learning_rate": 1.8010991759233235e-05,
+      "loss": 0.0114,
+      "step": 2097
+    },
+    {
+      "epoch": 149.85714285714286,
+      "grad_norm": 0.1962890625,
+      "learning_rate": 1.8003157672579217e-05,
+      "loss": 0.0152,
+      "step": 2098
+    },
+    {
+      "epoch": 149.92857142857142,
+      "grad_norm": 0.1396484375,
+      "learning_rate": 1.7995320903753e-05,
+      "loss": 0.013,
+      "step": 2099
+    },
+    {
+      "epoch": 150.0,
+      "grad_norm": 0.12255859375,
+      "learning_rate": 1.7987481456574135e-05,
+      "loss": 0.0138,
+      "step": 2100
+    },
+    {
+      "epoch": 150.07142857142858,
+      "grad_norm": 0.1435546875,
+      "learning_rate": 1.7979639334863467e-05,
+      "loss": 0.0144,
+      "step": 2101
+    },
+    {
+      "epoch": 150.14285714285714,
+      "grad_norm": 0.2138671875,
+      "learning_rate": 1.797179454244316e-05,
+      "loss": 0.0168,
+      "step": 2102
+    },
+    {
+      "epoch": 150.21428571428572,
+      "grad_norm": 0.146484375,
+      "learning_rate": 1.796394708313666e-05,
+      "loss": 0.0129,
+      "step": 2103
+    },
+    {
+      "epoch": 150.28571428571428,
+      "grad_norm": 0.169921875,
+      "learning_rate": 1.7956096960768738e-05,
+      "loss": 0.0123,
+      "step": 2104
+    },
+    {
+      "epoch": 150.35714285714286,
+      "grad_norm": 0.11865234375,
+      "learning_rate": 1.7948244179165428e-05,
+      "loss": 0.0153,
+      "step": 2105
+    },
+    {
+      "epoch": 150.42857142857142,
+      "grad_norm": 0.21484375,
+      "learning_rate": 1.7940388742154096e-05,
+      "loss": 0.0172,
+      "step": 2106
+    },
+    {
+      "epoch": 150.5,
+      "grad_norm": 0.1298828125,
+      "learning_rate": 1.793253065356338e-05,
+      "loss": 0.013,
+      "step": 2107
+    },
+    {
+      "epoch": 150.57142857142858,
+      "grad_norm": 0.1845703125,
+      "learning_rate": 1.792466991722322e-05,
+      "loss": 0.0121,
+      "step": 2108
+    },
+    {
+      "epoch": 150.64285714285714,
+      "grad_norm": 0.2216796875,
+      "learning_rate": 1.7916806536964842e-05,
+      "loss": 0.0145,
+      "step": 2109
+    },
+    {
+      "epoch": 150.71428571428572,
+      "grad_norm": 0.2578125,
+      "learning_rate": 1.790894051662076e-05,
+      "loss": 0.0134,
+      "step": 2110
+    },
+    {
+      "epoch": 150.78571428571428,
+      "grad_norm": 0.138671875,
+      "learning_rate": 1.790107186002478e-05,
+      "loss": 0.0124,
+      "step": 2111
+    },
+    {
+      "epoch": 150.85714285714286,
+      "grad_norm": 0.208984375,
+      "learning_rate": 1.7893200571011988e-05,
+      "loss": 0.0129,
+      "step": 2112
+    },
+    {
+      "epoch": 150.92857142857142,
+      "grad_norm": 0.1875,
+      "learning_rate": 1.7885326653418757e-05,
+      "loss": 0.0117,
+      "step": 2113
+    },
+    {
+      "epoch": 151.0,
+      "grad_norm": 0.28125,
+      "learning_rate": 1.7877450111082733e-05,
+      "loss": 0.0174,
+      "step": 2114
+    },
+    {
+      "epoch": 151.07142857142858,
+      "grad_norm": 0.30859375,
+      "learning_rate": 1.7869570947842855e-05,
+      "loss": 0.0156,
+      "step": 2115
+    },
+    {
+      "epoch": 151.14285714285714,
+      "grad_norm": 0.1171875,
+      "learning_rate": 1.7861689167539324e-05,
+      "loss": 0.0129,
+      "step": 2116
+    },
+    {
+      "epoch": 151.21428571428572,
+      "grad_norm": 0.31640625,
+      "learning_rate": 1.785380477401363e-05,
+      "loss": 0.0154,
+      "step": 2117
+    },
+    {
+      "epoch": 151.28571428571428,
+      "grad_norm": 0.1943359375,
+      "learning_rate": 1.7845917771108527e-05,
+      "loss": 0.0172,
+      "step": 2118
+    },
+    {
+      "epoch": 151.35714285714286,
+      "grad_norm": 0.12353515625,
+      "learning_rate": 1.783802816266804e-05,
+      "loss": 0.0131,
+      "step": 2119
+    },
+    {
+      "epoch": 151.42857142857142,
+      "grad_norm": 0.1650390625,
+      "learning_rate": 1.7830135952537487e-05,
+      "loss": 0.0125,
+      "step": 2120
+    },
+    {
+      "epoch": 151.5,
+      "grad_norm": 0.11962890625,
+      "learning_rate": 1.782224114456341e-05,
+      "loss": 0.0122,
+      "step": 2121
+    },
+    {
+      "epoch": 151.57142857142858,
+      "grad_norm": 0.1533203125,
+      "learning_rate": 1.781434374259366e-05,
+      "loss": 0.0156,
+      "step": 2122
+    },
+    {
+      "epoch": 151.64285714285714,
+      "grad_norm": 0.1845703125,
+      "learning_rate": 1.7806443750477324e-05,
+      "loss": 0.0133,
+      "step": 2123
+    },
+    {
+      "epoch": 151.71428571428572,
+      "grad_norm": 0.1328125,
+      "learning_rate": 1.779854117206477e-05,
+      "loss": 0.0134,
+      "step": 2124
+    },
+    {
+      "epoch": 151.78571428571428,
+      "grad_norm": 0.1708984375,
+      "learning_rate": 1.7790636011207622e-05,
+      "loss": 0.0129,
+      "step": 2125
+    },
+    {
+      "epoch": 151.85714285714286,
+      "grad_norm": 0.1953125,
+      "learning_rate": 1.7782728271758746e-05,
+      "loss": 0.0119,
+      "step": 2126
+    },
+    {
+      "epoch": 151.92857142857142,
+      "grad_norm": 0.10791015625,
+      "learning_rate": 1.7774817957572286e-05,
+      "loss": 0.014,
+      "step": 2127
+    },
+    {
+      "epoch": 152.0,
+      "grad_norm": 0.1123046875,
+      "learning_rate": 1.7766905072503636e-05,
+      "loss": 0.014,
+      "step": 2128
+    },
+    {
+      "epoch": 152.07142857142858,
+      "grad_norm": 0.12451171875,
+      "learning_rate": 1.7758989620409435e-05,
+      "loss": 0.0145,
+      "step": 2129
+    },
+    {
+      "epoch": 152.14285714285714,
+      "grad_norm": 0.1279296875,
+      "learning_rate": 1.7751071605147582e-05,
+      "loss": 0.0144,
+      "step": 2130
+    },
+    {
+      "epoch": 152.21428571428572,
+      "grad_norm": 0.140625,
+      "learning_rate": 1.7743151030577214e-05,
+      "loss": 0.0165,
+      "step": 2131
+    },
+    {
+      "epoch": 152.28571428571428,
+      "grad_norm": 0.1865234375,
+      "learning_rate": 1.773522790055873e-05,
+      "loss": 0.0137,
+      "step": 2132
+    },
+    {
+      "epoch": 152.35714285714286,
+      "grad_norm": 0.10009765625,
+      "learning_rate": 1.7727302218953766e-05,
+      "loss": 0.0132,
+      "step": 2133
+    },
+    {
+      "epoch": 152.42857142857142,
+      "grad_norm": 0.2421875,
+      "learning_rate": 1.771937398962521e-05,
+      "loss": 0.0133,
+      "step": 2134
+    },
+    {
+      "epoch": 152.5,
+      "grad_norm": 0.208984375,
+      "learning_rate": 1.771144321643717e-05,
+      "loss": 0.012,
+      "step": 2135
+    },
+    {
+      "epoch": 152.57142857142858,
+      "grad_norm": 0.2353515625,
+      "learning_rate": 1.7703509903255015e-05,
+      "loss": 0.016,
+      "step": 2136
+    },
+    {
+      "epoch": 152.64285714285714,
+      "grad_norm": 0.1162109375,
+      "learning_rate": 1.769557405394535e-05,
+      "loss": 0.0133,
+      "step": 2137
+    },
+    {
+      "epoch": 152.71428571428572,
+      "grad_norm": 0.451171875,
+      "learning_rate": 1.7687635672376006e-05,
+      "loss": 0.0154,
+      "step": 2138
+    },
+    {
+      "epoch": 152.78571428571428,
+      "grad_norm": 0.11865234375,
+      "learning_rate": 1.7679694762416056e-05,
+      "loss": 0.0126,
+      "step": 2139
+    },
+    {
+      "epoch": 152.85714285714286,
+      "grad_norm": 0.158203125,
+      "learning_rate": 1.7671751327935797e-05,
+      "loss": 0.014,
+      "step": 2140
+    },
+    {
+      "epoch": 152.92857142857142,
+      "grad_norm": 0.2490234375,
+      "learning_rate": 1.7663805372806775e-05,
+      "loss": 0.0134,
+      "step": 2141
+    },
+    {
+      "epoch": 153.0,
+      "grad_norm": 0.154296875,
+      "learning_rate": 1.765585690090174e-05,
+      "loss": 0.0133,
+      "step": 2142
+    },
+    {
+      "epoch": 153.07142857142858,
+      "grad_norm": 0.181640625,
+      "learning_rate": 1.7647905916094685e-05,
+      "loss": 0.0099,
+      "step": 2143
+    },
+    {
+      "epoch": 153.14285714285714,
+      "grad_norm": 0.1953125,
+      "learning_rate": 1.763995242226083e-05,
+      "loss": 0.014,
+      "step": 2144
+    },
+    {
+      "epoch": 153.21428571428572,
+      "grad_norm": 0.09423828125,
+      "learning_rate": 1.76319964232766e-05,
+      "loss": 0.0113,
+      "step": 2145
+    },
+    {
+      "epoch": 153.28571428571428,
+      "grad_norm": 0.23046875,
+      "learning_rate": 1.762403792301966e-05,
+      "loss": 0.014,
+      "step": 2146
+    },
+    {
+      "epoch": 153.35714285714286,
+      "grad_norm": 0.115234375,
+      "learning_rate": 1.7616076925368885e-05,
+      "loss": 0.0131,
+      "step": 2147
+    },
+    {
+      "epoch": 153.42857142857142,
+      "grad_norm": 0.1796875,
+      "learning_rate": 1.7608113434204366e-05,
+      "loss": 0.0122,
+      "step": 2148
+    },
+    {
+      "epoch": 153.5,
+      "grad_norm": 0.1103515625,
+      "learning_rate": 1.7600147453407412e-05,
+      "loss": 0.0133,
+      "step": 2149
+    },
+    {
+      "epoch": 153.57142857142858,
+      "grad_norm": 0.1201171875,
+      "learning_rate": 1.7592178986860544e-05,
+      "loss": 0.0156,
+      "step": 2150
+    },
+    {
+      "epoch": 153.64285714285714,
+      "grad_norm": 0.1396484375,
+      "learning_rate": 1.7584208038447505e-05,
+      "loss": 0.0107,
+      "step": 2151
+    },
+    {
+      "epoch": 153.71428571428572,
+      "grad_norm": 0.1728515625,
+      "learning_rate": 1.7576234612053227e-05,
+      "loss": 0.0117,
+      "step": 2152
+    },
+    {
+      "epoch": 153.78571428571428,
+      "grad_norm": 0.2431640625,
+      "learning_rate": 1.756825871156387e-05,
+      "loss": 0.013,
+      "step": 2153
+    },
+    {
+      "epoch": 153.85714285714286,
+      "grad_norm": 0.140625,
+      "learning_rate": 1.756028034086678e-05,
+      "loss": 0.0114,
+      "step": 2154
+    },
+    {
+      "epoch": 153.92857142857142,
+      "grad_norm": 0.23046875,
+      "learning_rate": 1.7552299503850527e-05,
+      "loss": 0.016,
+      "step": 2155
+    },
+    {
+      "epoch": 154.0,
+      "grad_norm": 0.26953125,
+      "learning_rate": 1.7544316204404877e-05,
+      "loss": 0.0129,
+      "step": 2156
+    },
+    {
+      "epoch": 154.07142857142858,
+      "grad_norm": 0.296875,
+      "learning_rate": 1.7536330446420785e-05,
+      "loss": 0.0139,
+      "step": 2157
+    },
+    {
+      "epoch": 154.14285714285714,
+      "grad_norm": 0.1572265625,
+      "learning_rate": 1.7528342233790415e-05,
+      "loss": 0.016,
+      "step": 2158
+    },
+    {
+      "epoch": 154.21428571428572,
+      "grad_norm": 0.14453125,
+      "learning_rate": 1.7520351570407127e-05,
+      "loss": 0.013,
+      "step": 2159
+    },
+    {
+      "epoch": 154.28571428571428,
+      "grad_norm": 0.10400390625,
+      "learning_rate": 1.751235846016547e-05,
+      "loss": 0.0115,
+      "step": 2160
+    },
+    {
+      "epoch": 154.35714285714286,
+      "grad_norm": 0.1220703125,
+      "learning_rate": 1.7504362906961193e-05,
+      "loss": 0.0125,
+      "step": 2161
+    },
+    {
+      "epoch": 154.42857142857142,
+      "grad_norm": 0.2412109375,
+      "learning_rate": 1.749636491469122e-05,
+      "loss": 0.0137,
+      "step": 2162
+    },
+    {
+      "epoch": 154.5,
+      "grad_norm": 0.125,
+      "learning_rate": 1.7488364487253692e-05,
+      "loss": 0.0146,
+      "step": 2163
+    },
+    {
+      "epoch": 154.57142857142858,
+      "grad_norm": 0.154296875,
+      "learning_rate": 1.748036162854791e-05,
+      "loss": 0.0134,
+      "step": 2164
+    },
+    {
+      "epoch": 154.64285714285714,
+      "grad_norm": 0.1298828125,
+      "learning_rate": 1.7472356342474376e-05,
+      "loss": 0.013,
+      "step": 2165
+    },
+    {
+      "epoch": 154.71428571428572,
+      "grad_norm": 0.1455078125,
+      "learning_rate": 1.7464348632934758e-05,
+      "loss": 0.0137,
+      "step": 2166
+    },
+    {
+      "epoch": 154.78571428571428,
+      "grad_norm": 0.11376953125,
+      "learning_rate": 1.745633850383193e-05,
+      "loss": 0.0121,
+      "step": 2167
+    },
+    {
+      "epoch": 154.85714285714286,
+      "grad_norm": 0.1513671875,
+      "learning_rate": 1.7448325959069923e-05,
+      "loss": 0.0134,
+      "step": 2168
+    },
+    {
+      "epoch": 154.92857142857142,
+      "grad_norm": 0.1630859375,
+      "learning_rate": 1.7440311002553957e-05,
+      "loss": 0.0119,
+      "step": 2169
+    },
+    {
+      "epoch": 155.0,
+      "grad_norm": 0.2578125,
+      "learning_rate": 1.7432293638190425e-05,
+      "loss": 0.0153,
+      "step": 2170
+    },
+    {
+      "epoch": 155.07142857142858,
+      "grad_norm": 0.2734375,
+      "learning_rate": 1.742427386988689e-05,
+      "loss": 0.0132,
+      "step": 2171
+    },
+    {
+      "epoch": 155.14285714285714,
+      "grad_norm": 0.126953125,
+      "learning_rate": 1.741625170155209e-05,
+      "loss": 0.0117,
+      "step": 2172
+    },
+    {
+      "epoch": 155.21428571428572,
+      "grad_norm": 0.1484375,
+      "learning_rate": 1.740822713709594e-05,
+      "loss": 0.0142,
+      "step": 2173
+    },
+    {
+      "epoch": 155.28571428571428,
+      "grad_norm": 0.1533203125,
+      "learning_rate": 1.7400200180429503e-05,
+      "loss": 0.0124,
+      "step": 2174
+    },
+    {
+      "epoch": 155.35714285714286,
+      "grad_norm": 0.150390625,
+      "learning_rate": 1.7392170835465032e-05,
+      "loss": 0.0143,
+      "step": 2175
+    },
+    {
+      "epoch": 155.42857142857142,
+      "grad_norm": 0.15234375,
+      "learning_rate": 1.738413910611592e-05,
+      "loss": 0.0163,
+      "step": 2176
+    },
+    {
+      "epoch": 155.5,
+      "grad_norm": 0.2294921875,
+      "learning_rate": 1.7376104996296746e-05,
+      "loss": 0.0126,
+      "step": 2177
+    },
+    {
+      "epoch": 155.57142857142858,
+      "grad_norm": 0.16796875,
+      "learning_rate": 1.7368068509923233e-05,
+      "loss": 0.0134,
+      "step": 2178
+    },
+    {
+      "epoch": 155.64285714285714,
+      "grad_norm": 0.115234375,
+      "learning_rate": 1.7360029650912263e-05,
+      "loss": 0.0137,
+      "step": 2179
+    },
+    {
+      "epoch": 155.71428571428572,
+      "grad_norm": 0.162109375,
+      "learning_rate": 1.7351988423181885e-05,
+      "loss": 0.0123,
+      "step": 2180
+    },
+    {
+      "epoch": 155.78571428571428,
+      "grad_norm": 0.20703125,
+      "learning_rate": 1.7343944830651288e-05,
+      "loss": 0.0142,
+      "step": 2181
+    },
+    {
+      "epoch": 155.85714285714286,
+      "grad_norm": 0.2001953125,
+      "learning_rate": 1.733589887724083e-05,
+      "loss": 0.0108,
+      "step": 2182
+    },
+    {
+      "epoch": 155.92857142857142,
+      "grad_norm": 0.1474609375,
+      "learning_rate": 1.7327850566872e-05,
+      "loss": 0.0138,
+      "step": 2183
+    },
+    {
+      "epoch": 156.0,
+      "grad_norm": 0.2060546875,
+      "learning_rate": 1.731979990346746e-05,
+      "loss": 0.0147,
+      "step": 2184
+    },
+    {
+      "epoch": 156.07142857142858,
+      "grad_norm": 0.138671875,
+      "learning_rate": 1.7311746890950994e-05,
+      "loss": 0.0166,
+      "step": 2185
+    },
+    {
+      "epoch": 156.14285714285714,
+      "grad_norm": 0.1474609375,
+      "learning_rate": 1.7303691533247546e-05,
+      "loss": 0.0171,
+      "step": 2186
+    },
+    {
+      "epoch": 156.21428571428572,
+      "grad_norm": 0.17578125,
+      "learning_rate": 1.7295633834283205e-05,
+      "loss": 0.0116,
+      "step": 2187
+    },
+    {
+      "epoch": 156.28571428571428,
+      "grad_norm": 0.11279296875,
+      "learning_rate": 1.728757379798519e-05,
+      "loss": 0.0141,
+      "step": 2188
+    },
+    {
+      "epoch": 156.35714285714286,
+      "grad_norm": 0.12060546875,
+      "learning_rate": 1.7279511428281864e-05,
+      "loss": 0.0148,
+      "step": 2189
+    },
+    {
+      "epoch": 156.42857142857142,
+      "grad_norm": 0.11962890625,
+      "learning_rate": 1.727144672910273e-05,
+      "loss": 0.0138,
+      "step": 2190
+    },
+    {
+      "epoch": 156.5,
+      "grad_norm": 0.1494140625,
+      "learning_rate": 1.726337970437843e-05,
+      "loss": 0.013,
+      "step": 2191
+    },
+    {
+      "epoch": 156.57142857142858,
+      "grad_norm": 0.208984375,
+      "learning_rate": 1.7255310358040723e-05,
+      "loss": 0.0145,
+      "step": 2192
+    },
+    {
+      "epoch": 156.64285714285714,
+      "grad_norm": 0.1435546875,
+      "learning_rate": 1.7247238694022517e-05,
+      "loss": 0.014,
+      "step": 2193
+    },
+    {
+      "epoch": 156.71428571428572,
+      "grad_norm": 0.125,
+      "learning_rate": 1.7239164716257837e-05,
+      "loss": 0.0142,
+      "step": 2194
+    },
+    {
+      "epoch": 156.78571428571428,
+      "grad_norm": 0.171875,
+      "learning_rate": 1.7231088428681852e-05,
+      "loss": 0.0118,
+      "step": 2195
+    },
+    {
+      "epoch": 156.85714285714286,
+      "grad_norm": 0.310546875,
+      "learning_rate": 1.722300983523084e-05,
+      "loss": 0.0102,
+      "step": 2196
+    },
+    {
+      "epoch": 156.92857142857142,
+      "grad_norm": 0.248046875,
+      "learning_rate": 1.721492893984221e-05,
+      "loss": 0.0136,
+      "step": 2197
+    },
+    {
+      "epoch": 157.0,
+      "grad_norm": 0.1611328125,
+      "learning_rate": 1.7206845746454492e-05,
+      "loss": 0.0136,
+      "step": 2198
+    },
+    {
+      "epoch": 157.07142857142858,
+      "grad_norm": 0.1259765625,
+      "learning_rate": 1.7198760259007336e-05,
+      "loss": 0.0136,
+      "step": 2199
+    },
+    {
+      "epoch": 157.14285714285714,
+      "grad_norm": 0.173828125,
+      "learning_rate": 1.719067248144151e-05,
+      "loss": 0.0148,
+      "step": 2200
+    },
+    {
+      "epoch": 157.21428571428572,
+      "grad_norm": 0.40234375,
+      "learning_rate": 1.7182582417698903e-05,
+      "loss": 0.0172,
+      "step": 2201
+    },
+    {
+      "epoch": 157.28571428571428,
+      "grad_norm": 0.10595703125,
+      "learning_rate": 1.7174490071722504e-05,
+      "loss": 0.0113,
+      "step": 2202
+    },
+    {
+      "epoch": 157.35714285714286,
+      "grad_norm": 0.255859375,
+      "learning_rate": 1.716639544745644e-05,
+      "loss": 0.0132,
+      "step": 2203
+    },
+    {
+      "epoch": 157.42857142857142,
+      "grad_norm": 0.26953125,
+      "learning_rate": 1.7158298548845914e-05,
+      "loss": 0.0157,
+      "step": 2204
+    },
+    {
+      "epoch": 157.5,
+      "grad_norm": 0.123046875,
+      "learning_rate": 1.715019937983727e-05,
+      "loss": 0.0126,
+      "step": 2205
+    },
+    {
+      "epoch": 157.57142857142858,
+      "grad_norm": 0.109375,
+      "learning_rate": 1.714209794437794e-05,
+      "loss": 0.0142,
+      "step": 2206
+    },
+    {
+      "epoch": 157.64285714285714,
+      "grad_norm": 0.142578125,
+      "learning_rate": 1.7133994246416456e-05,
+      "loss": 0.0139,
+      "step": 2207
+    },
+    {
+      "epoch": 157.71428571428572,
+      "grad_norm": 0.1376953125,
+      "learning_rate": 1.712588828990248e-05,
+      "loss": 0.0152,
+      "step": 2208
+    },
+    {
+      "epoch": 157.78571428571428,
+      "grad_norm": 0.16796875,
+      "learning_rate": 1.7117780078786745e-05,
+      "loss": 0.014,
+      "step": 2209
+    },
+    {
+      "epoch": 157.85714285714286,
+      "grad_norm": 0.294921875,
+      "learning_rate": 1.7109669617021095e-05,
+      "loss": 0.0147,
+      "step": 2210
+    },
+    {
+      "epoch": 157.92857142857142,
+      "grad_norm": 0.275390625,
+      "learning_rate": 1.7101556908558478e-05,
+      "loss": 0.0128,
+      "step": 2211
+    },
+    {
+      "epoch": 158.0,
+      "grad_norm": 0.123046875,
+      "learning_rate": 1.7093441957352917e-05,
+      "loss": 0.0141,
+      "step": 2212
+    },
+    {
+      "epoch": 158.07142857142858,
+      "grad_norm": 0.35546875,
+      "learning_rate": 1.7085324767359554e-05,
+      "loss": 0.0144,
+      "step": 2213
+    },
+    {
+      "epoch": 158.14285714285714,
+      "grad_norm": 0.2080078125,
+      "learning_rate": 1.7077205342534598e-05,
+      "loss": 0.0152,
+      "step": 2214
+    },
+    {
+      "epoch": 158.21428571428572,
+      "grad_norm": 0.12060546875,
+      "learning_rate": 1.7069083686835363e-05,
+      "loss": 0.014,
+      "step": 2215
+    },
+    {
+      "epoch": 158.28571428571428,
+      "grad_norm": 0.1611328125,
+      "learning_rate": 1.7060959804220246e-05,
+      "loss": 0.0124,
+      "step": 2216
+    },
+    {
+      "epoch": 158.35714285714286,
+      "grad_norm": 0.205078125,
+      "learning_rate": 1.705283369864873e-05,
+      "loss": 0.0127,
+      "step": 2217
+    },
+    {
+      "epoch": 158.42857142857142,
+      "grad_norm": 0.171875,
+      "learning_rate": 1.7044705374081374e-05,
+      "loss": 0.012,
+      "step": 2218
+    },
+    {
+      "epoch": 158.5,
+      "grad_norm": 0.109375,
+      "learning_rate": 1.703657483447983e-05,
+      "loss": 0.0147,
+      "step": 2219
+    },
+    {
+      "epoch": 158.57142857142858,
+      "grad_norm": 0.08740234375,
+      "learning_rate": 1.702844208380682e-05,
+      "loss": 0.0127,
+      "step": 2220
+    },
+    {
+      "epoch": 158.64285714285714,
+      "grad_norm": 0.376953125,
+      "learning_rate": 1.7020307126026152e-05,
+      "loss": 0.0153,
+      "step": 2221
+    },
+    {
+      "epoch": 158.71428571428572,
+      "grad_norm": 0.2177734375,
+      "learning_rate": 1.7012169965102708e-05,
+      "loss": 0.0137,
+      "step": 2222
+    },
+    {
+      "epoch": 158.78571428571428,
+      "grad_norm": 0.21484375,
+      "learning_rate": 1.7004030605002434e-05,
+      "loss": 0.0138,
+      "step": 2223
+    },
+    {
+      "epoch": 158.85714285714286,
+      "grad_norm": 0.330078125,
+      "learning_rate": 1.699588904969236e-05,
+      "loss": 0.0171,
+      "step": 2224
+    },
+    {
+      "epoch": 158.92857142857142,
+      "grad_norm": 0.1181640625,
+      "learning_rate": 1.698774530314058e-05,
+      "loss": 0.0147,
+      "step": 2225
+    },
+    {
+      "epoch": 159.0,
+      "grad_norm": 0.13671875,
+      "learning_rate": 1.697959936931625e-05,
+      "loss": 0.0122,
+      "step": 2226
+    },
+    {
+      "epoch": 159.07142857142858,
+      "grad_norm": 0.31640625,
+      "learning_rate": 1.6971451252189617e-05,
+      "loss": 0.0127,
+      "step": 2227
+    },
+    {
+      "epoch": 159.14285714285714,
+      "grad_norm": 0.1806640625,
+      "learning_rate": 1.6963300955731958e-05,
+      "loss": 0.0134,
+      "step": 2228
+    },
+    {
+      "epoch": 159.21428571428572,
+      "grad_norm": 0.2021484375,
+      "learning_rate": 1.6955148483915633e-05,
+      "loss": 0.0153,
+      "step": 2229
+    },
+    {
+      "epoch": 159.28571428571428,
+      "grad_norm": 0.1611328125,
+      "learning_rate": 1.694699384071406e-05,
+      "loss": 0.0157,
+      "step": 2230
+    },
+    {
+      "epoch": 159.35714285714286,
+      "grad_norm": 0.234375,
+      "learning_rate": 1.6938837030101714e-05,
+      "loss": 0.0133,
+      "step": 2231
+    },
+    {
+      "epoch": 159.42857142857142,
+      "grad_norm": 0.13671875,
+      "learning_rate": 1.6930678056054123e-05,
+      "loss": 0.0126,
+      "step": 2232
+    },
+    {
+      "epoch": 159.5,
+      "grad_norm": 0.1708984375,
+      "learning_rate": 1.692251692254787e-05,
+      "loss": 0.0144,
+      "step": 2233
+    },
+    {
+      "epoch": 159.57142857142858,
+      "grad_norm": 0.21875,
+      "learning_rate": 1.6914353633560594e-05,
+      "loss": 0.014,
+      "step": 2234
+    },
+    {
+      "epoch": 159.64285714285714,
+      "grad_norm": 0.146484375,
+      "learning_rate": 1.690618819307098e-05,
+      "loss": 0.012,
+      "step": 2235
+    },
+    {
+      "epoch": 159.71428571428572,
+      "grad_norm": 0.12353515625,
+      "learning_rate": 1.6898020605058773e-05,
+      "loss": 0.0134,
+      "step": 2236
+    },
+    {
+      "epoch": 159.78571428571428,
+      "grad_norm": 0.1240234375,
+      "learning_rate": 1.6889850873504752e-05,
+      "loss": 0.0137,
+      "step": 2237
+    },
+    {
+      "epoch": 159.85714285714286,
+      "grad_norm": 0.11865234375,
+      "learning_rate": 1.6881679002390738e-05,
+      "loss": 0.0136,
+      "step": 2238
+    },
+    {
+      "epoch": 159.92857142857142,
+      "grad_norm": 0.103515625,
+      "learning_rate": 1.6873504995699614e-05,
+      "loss": 0.0119,
+      "step": 2239
+    },
+    {
+      "epoch": 160.0,
+      "grad_norm": 0.1181640625,
+      "learning_rate": 1.686532885741528e-05,
+      "loss": 0.0146,
+      "step": 2240
+    },
+    {
+      "epoch": 160.07142857142858,
+      "grad_norm": 0.21875,
+      "learning_rate": 1.6857150591522692e-05,
+      "loss": 0.0183,
+      "step": 2241
+    },
+    {
+      "epoch": 160.14285714285714,
+      "grad_norm": 0.19140625,
+      "learning_rate": 1.6848970202007836e-05,
+      "loss": 0.0133,
+      "step": 2242
+    },
+    {
+      "epoch": 160.21428571428572,
+      "grad_norm": 0.1513671875,
+      "learning_rate": 1.684078769285774e-05,
+      "loss": 0.0135,
+      "step": 2243
+    },
+    {
+      "epoch": 160.28571428571428,
+      "grad_norm": 0.1240234375,
+      "learning_rate": 1.6832603068060453e-05,
+      "loss": 0.0123,
+      "step": 2244
+    },
+    {
+      "epoch": 160.35714285714286,
+      "grad_norm": 0.181640625,
+      "learning_rate": 1.6824416331605062e-05,
+      "loss": 0.0122,
+      "step": 2245
+    },
+    {
+      "epoch": 160.42857142857142,
+      "grad_norm": 0.1787109375,
+      "learning_rate": 1.6816227487481685e-05,
+      "loss": 0.0145,
+      "step": 2246
+    },
+    {
+      "epoch": 160.5,
+      "grad_norm": 0.125,
+      "learning_rate": 1.6808036539681463e-05,
+      "loss": 0.0141,
+      "step": 2247
+    },
+    {
+      "epoch": 160.57142857142858,
+      "grad_norm": 0.126953125,
+      "learning_rate": 1.6799843492196567e-05,
+      "loss": 0.0123,
+      "step": 2248
+    },
+    {
+      "epoch": 160.64285714285714,
+      "grad_norm": 0.2734375,
+      "learning_rate": 1.6791648349020192e-05,
+      "loss": 0.0144,
+      "step": 2249
+    },
+    {
+      "epoch": 160.71428571428572,
+      "grad_norm": 0.1484375,
+      "learning_rate": 1.6783451114146545e-05,
+      "loss": 0.0119,
+      "step": 2250
+    },
+    {
+      "epoch": 160.78571428571428,
+      "grad_norm": 0.1962890625,
+      "learning_rate": 1.677525179157086e-05,
+      "loss": 0.0149,
+      "step": 2251
+    },
+    {
+      "epoch": 160.85714285714286,
+      "grad_norm": 0.2138671875,
+      "learning_rate": 1.6767050385289386e-05,
+      "loss": 0.0138,
+      "step": 2252
+    },
+    {
+      "epoch": 160.92857142857142,
+      "grad_norm": 0.125,
+      "learning_rate": 1.67588468992994e-05,
+      "loss": 0.0145,
+      "step": 2253
+    },
+    {
+      "epoch": 161.0,
+      "grad_norm": 0.13671875,
+      "learning_rate": 1.675064133759916e-05,
+      "loss": 0.0119,
+      "step": 2254
+    },
+    {
+      "epoch": 161.07142857142858,
+      "grad_norm": 0.1162109375,
+      "learning_rate": 1.6742433704187977e-05,
+      "loss": 0.0143,
+      "step": 2255
+    },
+    {
+      "epoch": 161.14285714285714,
+      "grad_norm": 0.26953125,
+      "learning_rate": 1.6734224003066146e-05,
+      "loss": 0.0133,
+      "step": 2256
+    },
+    {
+      "epoch": 161.21428571428572,
+      "grad_norm": 0.12890625,
+      "learning_rate": 1.6726012238234967e-05,
+      "loss": 0.0152,
+      "step": 2257
+    },
+    {
+      "epoch": 161.28571428571428,
+      "grad_norm": 0.1787109375,
+      "learning_rate": 1.6717798413696765e-05,
+      "loss": 0.0159,
+      "step": 2258
+    },
+    {
+      "epoch": 161.35714285714286,
+      "grad_norm": 0.154296875,
+      "learning_rate": 1.6709582533454853e-05,
+      "loss": 0.0143,
+      "step": 2259
+    },
+    {
+      "epoch": 161.42857142857142,
+      "grad_norm": 0.126953125,
+      "learning_rate": 1.6701364601513557e-05,
+      "loss": 0.0133,
+      "step": 2260
+    },
+    {
+      "epoch": 161.5,
+      "grad_norm": 0.1640625,
+      "learning_rate": 1.669314462187819e-05,
+      "loss": 0.0104,
+      "step": 2261
+    },
+    {
+      "epoch": 161.57142857142858,
+      "grad_norm": 0.11767578125,
+      "learning_rate": 1.6684922598555074e-05,
+      "loss": 0.0113,
+      "step": 2262
+    },
+    {
+      "epoch": 161.64285714285714,
+      "grad_norm": 0.11083984375,
+      "learning_rate": 1.6676698535551523e-05,
+      "loss": 0.0138,
+      "step": 2263
+    },
+    {
+      "epoch": 161.71428571428572,
+      "grad_norm": 0.2109375,
+      "learning_rate": 1.666847243687584e-05,
+      "loss": 0.0136,
+      "step": 2264
+    },
+    {
+      "epoch": 161.78571428571428,
+      "grad_norm": 0.11474609375,
+      "learning_rate": 1.6660244306537335e-05,
+      "loss": 0.0141,
+      "step": 2265
+    },
+    {
+      "epoch": 161.85714285714286,
+      "grad_norm": 0.111328125,
+      "learning_rate": 1.665201414854629e-05,
+      "loss": 0.0141,
+      "step": 2266
+    },
+    {
+      "epoch": 161.92857142857142,
+      "grad_norm": 0.138671875,
+      "learning_rate": 1.6643781966913997e-05,
+      "loss": 0.0138,
+      "step": 2267
+    },
+    {
+      "epoch": 162.0,
+      "grad_norm": 0.1357421875,
+      "learning_rate": 1.6635547765652704e-05,
+      "loss": 0.0109,
+      "step": 2268
+    },
+    {
+      "epoch": 162.07142857142858,
+      "grad_norm": 0.1162109375,
+      "learning_rate": 1.662731154877567e-05,
+      "loss": 0.0139,
+      "step": 2269
+    },
+    {
+      "epoch": 162.14285714285714,
+      "grad_norm": 0.11572265625,
+      "learning_rate": 1.6619073320297128e-05,
+      "loss": 0.0129,
+      "step": 2270
+    },
+    {
+      "epoch": 162.21428571428572,
+      "grad_norm": 0.138671875,
+      "learning_rate": 1.661083308423229e-05,
+      "loss": 0.0125,
+      "step": 2271
+    },
+    {
+      "epoch": 162.28571428571428,
+      "grad_norm": 0.1328125,
+      "learning_rate": 1.660259084459735e-05,
+      "loss": 0.0132,
+      "step": 2272
+    },
+    {
+      "epoch": 162.35714285714286,
+      "grad_norm": 0.12255859375,
+      "learning_rate": 1.659434660540946e-05,
+      "loss": 0.0133,
+      "step": 2273
+    },
+    {
+      "epoch": 162.42857142857142,
+      "grad_norm": 0.140625,
+      "learning_rate": 1.6586100370686782e-05,
+      "loss": 0.0137,
+      "step": 2274
+    },
+    {
+      "epoch": 162.5,
+      "grad_norm": 0.12255859375,
+      "learning_rate": 1.657785214444842e-05,
+      "loss": 0.0133,
+      "step": 2275
+    },
+    {
+      "epoch": 162.57142857142858,
+      "grad_norm": 0.1337890625,
+      "learning_rate": 1.656960193071446e-05,
+      "loss": 0.0129,
+      "step": 2276
+    },
+    {
+      "epoch": 162.64285714285714,
+      "grad_norm": 0.1259765625,
+      "learning_rate": 1.6561349733505956e-05,
+      "loss": 0.014,
+      "step": 2277
+    },
+    {
+      "epoch": 162.71428571428572,
+      "grad_norm": 0.1689453125,
+      "learning_rate": 1.655309555684493e-05,
+      "loss": 0.015,
+      "step": 2278
+    },
+    {
+      "epoch": 162.78571428571428,
+      "grad_norm": 0.15625,
+      "learning_rate": 1.654483940475437e-05,
+      "loss": 0.014,
+      "step": 2279
+    },
+    {
+      "epoch": 162.85714285714286,
+      "grad_norm": 0.162109375,
+      "learning_rate": 1.653658128125822e-05,
+      "loss": 0.0143,
+      "step": 2280
+    },
+    {
+      "epoch": 162.92857142857142,
+      "grad_norm": 0.12255859375,
+      "learning_rate": 1.652832119038139e-05,
+      "loss": 0.0131,
+      "step": 2281
+    },
+    {
+      "epoch": 163.0,
+      "grad_norm": 0.142578125,
+      "learning_rate": 1.6520059136149747e-05,
+      "loss": 0.0129,
+      "step": 2282
+    },
+    {
+      "epoch": 163.07142857142858,
+      "grad_norm": 0.10986328125,
+      "learning_rate": 1.651179512259012e-05,
+      "loss": 0.0168,
+      "step": 2283
+    },
+    {
+      "epoch": 163.14285714285714,
+      "grad_norm": 0.150390625,
+      "learning_rate": 1.6503529153730285e-05,
+      "loss": 0.0121,
+      "step": 2284
+    },
+    {
+      "epoch": 163.21428571428572,
+      "grad_norm": 0.1318359375,
+      "learning_rate": 1.6495261233598976e-05,
+      "loss": 0.0139,
+      "step": 2285
+    },
+    {
+      "epoch": 163.28571428571428,
+      "grad_norm": 0.16796875,
+      "learning_rate": 1.6486991366225878e-05,
+      "loss": 0.0141,
+      "step": 2286
+    },
+    {
+      "epoch": 163.35714285714286,
+      "grad_norm": 0.1396484375,
+      "learning_rate": 1.647871955564162e-05,
+      "loss": 0.0137,
+      "step": 2287
+    },
+    {
+      "epoch": 163.42857142857142,
+      "grad_norm": 0.259765625,
+      "learning_rate": 1.6470445805877782e-05,
+      "loss": 0.0133,
+      "step": 2288
+    },
+    {
+      "epoch": 163.5,
+      "grad_norm": 0.162109375,
+      "learning_rate": 1.64621701209669e-05,
+      "loss": 0.0123,
+      "step": 2289
+    },
+    {
+      "epoch": 163.57142857142858,
+      "grad_norm": 0.234375,
+      "learning_rate": 1.6453892504942425e-05,
+      "loss": 0.012,
+      "step": 2290
+    },
+    {
+      "epoch": 163.64285714285714,
+      "grad_norm": 0.212890625,
+      "learning_rate": 1.6445612961838783e-05,
+      "loss": 0.0135,
+      "step": 2291
+    },
+    {
+      "epoch": 163.71428571428572,
+      "grad_norm": 0.13671875,
+      "learning_rate": 1.643733149569131e-05,
+      "loss": 0.0157,
+      "step": 2292
+    },
+    {
+      "epoch": 163.78571428571428,
+      "grad_norm": 0.2578125,
+      "learning_rate": 1.64290481105363e-05,
+      "loss": 0.0151,
+      "step": 2293
+    },
+    {
+      "epoch": 163.85714285714286,
+      "grad_norm": 0.169921875,
+      "learning_rate": 1.6420762810410973e-05,
+      "loss": 0.0145,
+      "step": 2294
+    },
+    {
+      "epoch": 163.92857142857142,
+      "grad_norm": 0.1640625,
+      "learning_rate": 1.6412475599353477e-05,
+      "loss": 0.0113,
+      "step": 2295
+    },
+    {
+      "epoch": 164.0,
+      "grad_norm": 0.1259765625,
+      "learning_rate": 1.6404186481402908e-05,
+      "loss": 0.0126,
+      "step": 2296
+    },
+    {
+      "epoch": 164.07142857142858,
+      "grad_norm": 0.482421875,
+      "learning_rate": 1.6395895460599274e-05,
+      "loss": 0.0181,
+      "step": 2297
+    },
+    {
+      "epoch": 164.14285714285714,
+      "grad_norm": 0.1357421875,
+      "learning_rate": 1.6387602540983525e-05,
+      "loss": 0.0131,
+      "step": 2298
+    },
+    {
+      "epoch": 164.21428571428572,
+      "grad_norm": 0.25,
+      "learning_rate": 1.6379307726597517e-05,
+      "loss": 0.0147,
+      "step": 2299
+    },
+    {
+      "epoch": 164.28571428571428,
+      "grad_norm": 0.126953125,
+      "learning_rate": 1.6371011021484052e-05,
+      "loss": 0.0143,
+      "step": 2300
+    },
+    {
+      "epoch": 164.35714285714286,
+      "grad_norm": 0.1474609375,
+      "learning_rate": 1.6362712429686846e-05,
+      "loss": 0.0138,
+      "step": 2301
+    },
+    {
+      "epoch": 164.42857142857142,
+      "grad_norm": 0.1962890625,
+      "learning_rate": 1.635441195525052e-05,
+      "loss": 0.0112,
+      "step": 2302
+    },
+    {
+      "epoch": 164.5,
+      "grad_norm": 0.212890625,
+      "learning_rate": 1.6346109602220636e-05,
+      "loss": 0.0112,
+      "step": 2303
+    },
+    {
+      "epoch": 164.57142857142858,
+      "grad_norm": 0.28125,
+      "learning_rate": 1.633780537464365e-05,
+      "loss": 0.0115,
+      "step": 2304
+    },
+    {
+      "epoch": 164.64285714285714,
+      "grad_norm": 0.1376953125,
+      "learning_rate": 1.6329499276566948e-05,
+      "loss": 0.0136,
+      "step": 2305
+    },
+    {
+      "epoch": 164.71428571428572,
+      "grad_norm": 0.197265625,
+      "learning_rate": 1.632119131203882e-05,
+      "loss": 0.0174,
+      "step": 2306
+    },
+    {
+      "epoch": 164.78571428571428,
+      "grad_norm": 0.1142578125,
+      "learning_rate": 1.6312881485108473e-05,
+      "loss": 0.0133,
+      "step": 2307
+    },
+    {
+      "epoch": 164.85714285714286,
+      "grad_norm": 0.15234375,
+      "learning_rate": 1.6304569799826004e-05,
+      "loss": 0.0133,
+      "step": 2308
+    },
+    {
+      "epoch": 164.92857142857142,
+      "grad_norm": 0.10888671875,
+      "learning_rate": 1.6296256260242428e-05,
+      "loss": 0.0134,
+      "step": 2309
+    },
+    {
+      "epoch": 165.0,
+      "grad_norm": 0.19921875,
+      "learning_rate": 1.6287940870409677e-05,
+      "loss": 0.0167,
+      "step": 2310
+    },
+    {
+      "epoch": 165.07142857142858,
+      "grad_norm": 0.1181640625,
+      "learning_rate": 1.6279623634380557e-05,
+      "loss": 0.0162,
+      "step": 2311
+    },
+    {
+      "epoch": 165.14285714285714,
+      "grad_norm": 0.1552734375,
+      "learning_rate": 1.6271304556208795e-05,
+      "loss": 0.0106,
+      "step": 2312
+    },
+    {
+      "epoch": 165.21428571428572,
+      "grad_norm": 0.2236328125,
+      "learning_rate": 1.6262983639949008e-05,
+      "loss": 0.0151,
+      "step": 2313
+    },
+    {
+      "epoch": 165.28571428571428,
+      "grad_norm": 0.2578125,
+      "learning_rate": 1.62546608896567e-05,
+      "loss": 0.0125,
+      "step": 2314
+    },
+    {
+      "epoch": 165.35714285714286,
+      "grad_norm": 0.240234375,
+      "learning_rate": 1.62463363093883e-05,
+      "loss": 0.0157,
+      "step": 2315
+    },
+    {
+      "epoch": 165.42857142857142,
+      "grad_norm": 0.2451171875,
+      "learning_rate": 1.6238009903201085e-05,
+      "loss": 0.012,
+      "step": 2316
+    },
+    {
+      "epoch": 165.5,
+      "grad_norm": 0.12158203125,
+      "learning_rate": 1.6229681675153255e-05,
+      "loss": 0.0147,
+      "step": 2317
+    },
+    {
+      "epoch": 165.57142857142858,
+      "grad_norm": 0.0986328125,
+      "learning_rate": 1.6221351629303883e-05,
+      "loss": 0.0122,
+      "step": 2318
+    },
+    {
+      "epoch": 165.64285714285714,
+      "grad_norm": 0.2158203125,
+      "learning_rate": 1.6213019769712938e-05,
+      "loss": 0.0146,
+      "step": 2319
+    },
+    {
+      "epoch": 165.71428571428572,
+      "grad_norm": 0.1162109375,
+      "learning_rate": 1.6204686100441264e-05,
+      "loss": 0.0127,
+      "step": 2320
+    },
+    {
+      "epoch": 165.78571428571428,
+      "grad_norm": 0.130859375,
+      "learning_rate": 1.6196350625550585e-05,
+      "loss": 0.0124,
+      "step": 2321
+    },
+    {
+      "epoch": 165.85714285714286,
+      "grad_norm": 0.216796875,
+      "learning_rate": 1.618801334910352e-05,
+      "loss": 0.0155,
+      "step": 2322
+    },
+    {
+      "epoch": 165.92857142857142,
+      "grad_norm": 0.1142578125,
+      "learning_rate": 1.6179674275163545e-05,
+      "loss": 0.0125,
+      "step": 2323
+    },
+    {
+      "epoch": 166.0,
+      "grad_norm": 0.12109375,
+      "learning_rate": 1.6171333407795036e-05,
+      "loss": 0.0139,
+      "step": 2324
+    },
+    {
+      "epoch": 166.07142857142858,
+      "grad_norm": 0.126953125,
+      "learning_rate": 1.6162990751063217e-05,
+      "loss": 0.0156,
+      "step": 2325
+    },
+    {
+      "epoch": 166.14285714285714,
+      "grad_norm": 0.1630859375,
+      "learning_rate": 1.615464630903421e-05,
+      "loss": 0.012,
+      "step": 2326
+    },
+    {
+      "epoch": 166.21428571428572,
+      "grad_norm": 0.130859375,
+      "learning_rate": 1.6146300085774986e-05,
+      "loss": 0.0144,
+      "step": 2327
+    },
+    {
+      "epoch": 166.28571428571428,
+      "grad_norm": 0.1103515625,
+      "learning_rate": 1.6137952085353398e-05,
+      "loss": 0.0137,
+      "step": 2328
+    },
+    {
+      "epoch": 166.35714285714286,
+      "grad_norm": 0.1259765625,
+      "learning_rate": 1.612960231183816e-05,
+      "loss": 0.0136,
+      "step": 2329
+    },
+    {
+      "epoch": 166.42857142857142,
+      "grad_norm": 0.79296875,
+      "learning_rate": 1.6121250769298844e-05,
+      "loss": 0.0142,
+      "step": 2330
+    },
+    {
+      "epoch": 166.5,
+      "grad_norm": 0.1103515625,
+      "learning_rate": 1.6112897461805895e-05,
+      "loss": 0.0121,
+      "step": 2331
+    },
+    {
+      "epoch": 166.57142857142858,
+      "grad_norm": 0.138671875,
+      "learning_rate": 1.6104542393430615e-05,
+      "loss": 0.0131,
+      "step": 2332
+    },
+    {
+      "epoch": 166.64285714285714,
+      "grad_norm": 0.1455078125,
+      "learning_rate": 1.609618556824516e-05,
+      "loss": 0.0118,
+      "step": 2333
+    },
+    {
+      "epoch": 166.71428571428572,
+      "grad_norm": 0.1875,
+      "learning_rate": 1.6087826990322546e-05,
+      "loss": 0.014,
+      "step": 2334
+    },
+    {
+      "epoch": 166.78571428571428,
+      "grad_norm": 0.158203125,
+      "learning_rate": 1.607946666373664e-05,
+      "loss": 0.0135,
+      "step": 2335
+    },
+    {
+      "epoch": 166.85714285714286,
+      "grad_norm": 0.1708984375,
+      "learning_rate": 1.607110459256217e-05,
+      "loss": 0.0138,
+      "step": 2336
+    },
+    {
+      "epoch": 166.92857142857142,
+      "grad_norm": 0.17578125,
+      "learning_rate": 1.60627407808747e-05,
+      "loss": 0.0151,
+      "step": 2337
+    },
+    {
+      "epoch": 167.0,
+      "grad_norm": 0.134765625,
+      "learning_rate": 1.605437523275066e-05,
+      "loss": 0.0141,
+      "step": 2338
+    },
+    {
+      "epoch": 167.07142857142858,
+      "grad_norm": 0.11474609375,
+      "learning_rate": 1.6046007952267305e-05,
+      "loss": 0.0121,
+      "step": 2339
+    },
+    {
+      "epoch": 167.14285714285714,
+      "grad_norm": 0.1328125,
+      "learning_rate": 1.6037638943502754e-05,
+      "loss": 0.0135,
+      "step": 2340
+    },
+    {
+      "epoch": 167.21428571428572,
+      "grad_norm": 0.171875,
+      "learning_rate": 1.602926821053596e-05,
+      "loss": 0.0152,
+      "step": 2341
+    },
+    {
+      "epoch": 167.28571428571428,
+      "grad_norm": 0.1923828125,
+      "learning_rate": 1.6020895757446708e-05,
+      "loss": 0.0125,
+      "step": 2342
+    },
+    {
+      "epoch": 167.35714285714286,
+      "grad_norm": 0.1259765625,
+      "learning_rate": 1.601252158831564e-05,
+      "loss": 0.0129,
+      "step": 2343
+    },
+    {
+      "epoch": 167.42857142857142,
+      "grad_norm": 0.3046875,
+      "learning_rate": 1.600414570722422e-05,
+      "loss": 0.0154,
+      "step": 2344
+    },
+    {
+      "epoch": 167.5,
+      "grad_norm": 0.1259765625,
+      "learning_rate": 1.599576811825475e-05,
+      "loss": 0.0121,
+      "step": 2345
+    },
+    {
+      "epoch": 167.57142857142858,
+      "grad_norm": 0.3125,
+      "learning_rate": 1.5987388825490368e-05,
+      "loss": 0.0159,
+      "step": 2346
+    },
+    {
+      "epoch": 167.64285714285714,
+      "grad_norm": 0.126953125,
+      "learning_rate": 1.5979007833015036e-05,
+      "loss": 0.0138,
+      "step": 2347
+    },
+    {
+      "epoch": 167.71428571428572,
+      "grad_norm": 0.16015625,
+      "learning_rate": 1.5970625144913552e-05,
+      "loss": 0.0134,
+      "step": 2348
+    },
+    {
+      "epoch": 167.78571428571428,
+      "grad_norm": 0.134765625,
+      "learning_rate": 1.596224076527153e-05,
+      "loss": 0.0158,
+      "step": 2349
+    },
+    {
+      "epoch": 167.85714285714286,
+      "grad_norm": 0.126953125,
+      "learning_rate": 1.5953854698175422e-05,
+      "loss": 0.0147,
+      "step": 2350
+    },
+    {
+      "epoch": 167.92857142857142,
+      "grad_norm": 0.24609375,
+      "learning_rate": 1.5945466947712488e-05,
+      "loss": 0.0156,
+      "step": 2351
+    },
+    {
+      "epoch": 168.0,
+      "grad_norm": 0.1357421875,
+      "learning_rate": 1.593707751797083e-05,
+      "loss": 0.0128,
+      "step": 2352
+    },
+    {
+      "epoch": 168.07142857142858,
+      "grad_norm": 0.2216796875,
+      "learning_rate": 1.5928686413039334e-05,
+      "loss": 0.013,
+      "step": 2353
+    },
+    {
+      "epoch": 168.14285714285714,
+      "grad_norm": 0.12109375,
+      "learning_rate": 1.592029363700773e-05,
+      "loss": 0.0136,
+      "step": 2354
+    },
+    {
+      "epoch": 168.21428571428572,
+      "grad_norm": 0.12890625,
+      "learning_rate": 1.5911899193966567e-05,
+      "loss": 0.013,
+      "step": 2355
+    },
+    {
+      "epoch": 168.28571428571428,
+      "grad_norm": 0.158203125,
+      "learning_rate": 1.590350308800718e-05,
+      "loss": 0.0123,
+      "step": 2356
+    },
+    {
+      "epoch": 168.35714285714286,
+      "grad_norm": 0.216796875,
+      "learning_rate": 1.5895105323221735e-05,
+      "loss": 0.0126,
+      "step": 2357
+    },
+    {
+      "epoch": 168.42857142857142,
+      "grad_norm": 0.1162109375,
+      "learning_rate": 1.58867059037032e-05,
+      "loss": 0.0142,
+      "step": 2358
+    },
+    {
+      "epoch": 168.5,
+      "grad_norm": 0.1630859375,
+      "learning_rate": 1.5878304833545348e-05,
+      "loss": 0.0137,
+      "step": 2359
+    },
+    {
+      "epoch": 168.57142857142858,
+      "grad_norm": 0.255859375,
+      "learning_rate": 1.5869902116842765e-05,
+      "loss": 0.0153,
+      "step": 2360
+    },
+    {
+      "epoch": 168.64285714285714,
+      "grad_norm": 0.11328125,
+      "learning_rate": 1.586149775769082e-05,
+      "loss": 0.0126,
+      "step": 2361
+    },
+    {
+      "epoch": 168.71428571428572,
+      "grad_norm": 0.2001953125,
+      "learning_rate": 1.585309176018571e-05,
+      "loss": 0.0135,
+      "step": 2362
+    },
+    {
+      "epoch": 168.78571428571428,
+      "grad_norm": 0.15625,
+      "learning_rate": 1.584468412842441e-05,
+      "loss": 0.0125,
+      "step": 2363
+    },
+    {
+      "epoch": 168.85714285714286,
+      "grad_norm": 0.142578125,
+      "learning_rate": 1.5836274866504698e-05,
+      "loss": 0.0164,
+      "step": 2364
+    },
+    {
+      "epoch": 168.92857142857142,
+      "grad_norm": 0.10009765625,
+      "learning_rate": 1.5827863978525143e-05,
+      "loss": 0.0126,
+      "step": 2365
+    },
+    {
+      "epoch": 169.0,
+      "grad_norm": 0.162109375,
+      "learning_rate": 1.581945146858511e-05,
+      "loss": 0.0118,
+      "step": 2366
+    },
+    {
+      "epoch": 169.07142857142858,
+      "grad_norm": 0.162109375,
+      "learning_rate": 1.581103734078476e-05,
+      "loss": 0.0146,
+      "step": 2367
+    },
+    {
+      "epoch": 169.14285714285714,
+      "grad_norm": 0.1279296875,
+      "learning_rate": 1.5802621599225033e-05,
+      "loss": 0.0132,
+      "step": 2368
+    },
+    {
+      "epoch": 169.21428571428572,
+      "grad_norm": 0.146484375,
+      "learning_rate": 1.579420424800766e-05,
+      "loss": 0.0121,
+      "step": 2369
+    },
+    {
+      "epoch": 169.28571428571428,
+      "grad_norm": 0.234375,
+      "learning_rate": 1.578578529123515e-05,
+      "loss": 0.0135,
+      "step": 2370
+    },
+    {
+      "epoch": 169.35714285714286,
+      "grad_norm": 0.2177734375,
+      "learning_rate": 1.577736473301081e-05,
+      "loss": 0.0128,
+      "step": 2371
+    },
+    {
+      "epoch": 169.42857142857142,
+      "grad_norm": 0.271484375,
+      "learning_rate": 1.576894257743871e-05,
+      "loss": 0.0138,
+      "step": 2372
+    },
+    {
+      "epoch": 169.5,
+      "grad_norm": 0.1142578125,
+      "learning_rate": 1.5760518828623715e-05,
+      "loss": 0.0143,
+      "step": 2373
+    },
+    {
+      "epoch": 169.57142857142858,
+      "grad_norm": 0.11572265625,
+      "learning_rate": 1.5752093490671447e-05,
+      "loss": 0.0127,
+      "step": 2374
+    },
+    {
+      "epoch": 169.64285714285714,
+      "grad_norm": 0.13671875,
+      "learning_rate": 1.574366656768832e-05,
+      "loss": 0.0141,
+      "step": 2375
+    },
+    {
+      "epoch": 169.71428571428572,
+      "grad_norm": 0.2099609375,
+      "learning_rate": 1.573523806378151e-05,
+      "loss": 0.0148,
+      "step": 2376
+    },
+    {
+      "epoch": 169.78571428571428,
+      "grad_norm": 0.1201171875,
+      "learning_rate": 1.5726807983058974e-05,
+      "loss": 0.013,
+      "step": 2377
+    },
+    {
+      "epoch": 169.85714285714286,
+      "grad_norm": 0.13671875,
+      "learning_rate": 1.5718376329629424e-05,
+      "loss": 0.0143,
+      "step": 2378
+    },
+    {
+      "epoch": 169.92857142857142,
+      "grad_norm": 0.14453125,
+      "learning_rate": 1.5709943107602347e-05,
+      "loss": 0.0116,
+      "step": 2379
+    },
+    {
+      "epoch": 170.0,
+      "grad_norm": 0.265625,
+      "learning_rate": 1.5701508321087995e-05,
+      "loss": 0.0163,
+      "step": 2380
+    },
+    {
+      "epoch": 170.07142857142858,
+      "grad_norm": 0.1279296875,
+      "learning_rate": 1.569307197419738e-05,
+      "loss": 0.0145,
+      "step": 2381
+    },
+    {
+      "epoch": 170.14285714285714,
+      "grad_norm": 0.1689453125,
+      "learning_rate": 1.5684634071042278e-05,
+      "loss": 0.013,
+      "step": 2382
+    },
+    {
+      "epoch": 170.21428571428572,
+      "grad_norm": 0.15234375,
+      "learning_rate": 1.5676194615735217e-05,
+      "loss": 0.0156,
+      "step": 2383
+    },
+    {
+      "epoch": 170.28571428571428,
+      "grad_norm": 0.125,
+      "learning_rate": 1.566775361238948e-05,
+      "loss": 0.0139,
+      "step": 2384
+    },
+    {
+      "epoch": 170.35714285714286,
+      "grad_norm": 0.130859375,
+      "learning_rate": 1.5659311065119118e-05,
+      "loss": 0.015,
+      "step": 2385
+    },
+    {
+      "epoch": 170.42857142857142,
+      "grad_norm": 0.1298828125,
+      "learning_rate": 1.5650866978038925e-05,
+      "loss": 0.0126,
+      "step": 2386
+    },
+    {
+      "epoch": 170.5,
+      "grad_norm": 0.1142578125,
+      "learning_rate": 1.5642421355264438e-05,
+      "loss": 0.0142,
+      "step": 2387
+    },
+    {
+      "epoch": 170.57142857142858,
+      "grad_norm": 0.1376953125,
+      "learning_rate": 1.563397420091196e-05,
+      "loss": 0.014,
+      "step": 2388
+    },
+    {
+      "epoch": 170.64285714285714,
+      "grad_norm": 0.154296875,
+      "learning_rate": 1.5625525519098527e-05,
+      "loss": 0.0125,
+      "step": 2389
+    },
+    {
+      "epoch": 170.71428571428572,
+      "grad_norm": 0.1318359375,
+      "learning_rate": 1.5617075313941925e-05,
+      "loss": 0.0153,
+      "step": 2390
+    },
+    {
+      "epoch": 170.78571428571428,
+      "grad_norm": 0.138671875,
+      "learning_rate": 1.560862358956069e-05,
+      "loss": 0.015,
+      "step": 2391
+    },
+    {
+      "epoch": 170.85714285714286,
+      "grad_norm": 0.154296875,
+      "learning_rate": 1.560017035007407e-05,
+      "loss": 0.0147,
+      "step": 2392
+    },
+    {
+      "epoch": 170.92857142857142,
+      "grad_norm": 0.1494140625,
+      "learning_rate": 1.559171559960209e-05,
+      "loss": 0.0165,
+      "step": 2393
+    },
+    {
+      "epoch": 171.0,
+      "grad_norm": 0.255859375,
+      "learning_rate": 1.5583259342265483e-05,
+      "loss": 0.0127,
+      "step": 2394
+    },
+    {
+      "epoch": 171.07142857142858,
+      "grad_norm": 0.1689453125,
+      "learning_rate": 1.5574801582185726e-05,
+      "loss": 0.0125,
+      "step": 2395
+    },
+    {
+      "epoch": 171.14285714285714,
+      "grad_norm": 0.173828125,
+      "learning_rate": 1.5566342323485032e-05,
+      "loss": 0.0162,
+      "step": 2396
+    },
+    {
+      "epoch": 171.21428571428572,
+      "grad_norm": 0.1806640625,
+      "learning_rate": 1.555788157028634e-05,
+      "loss": 0.0123,
+      "step": 2397
+    },
+    {
+      "epoch": 171.28571428571428,
+      "grad_norm": 0.1806640625,
+      "learning_rate": 1.5549419326713323e-05,
+      "loss": 0.016,
+      "step": 2398
+    },
+    {
+      "epoch": 171.35714285714286,
+      "grad_norm": 0.15625,
+      "learning_rate": 1.5540955596890365e-05,
+      "loss": 0.0141,
+      "step": 2399
+    },
+    {
+      "epoch": 171.42857142857142,
+      "grad_norm": 0.1533203125,
+      "learning_rate": 1.5532490384942593e-05,
+      "loss": 0.0126,
+      "step": 2400
+    },
+    {
+      "epoch": 171.5,
+      "grad_norm": 0.1611328125,
+      "learning_rate": 1.5524023694995844e-05,
+      "loss": 0.0155,
+      "step": 2401
+    },
+    {
+      "epoch": 171.57142857142858,
+      "grad_norm": 0.1611328125,
+      "learning_rate": 1.5515555531176687e-05,
+      "loss": 0.0128,
+      "step": 2402
+    },
+    {
+      "epoch": 171.64285714285714,
+      "grad_norm": 0.2470703125,
+      "learning_rate": 1.5507085897612396e-05,
+      "loss": 0.0123,
+      "step": 2403
+    },
+    {
+      "epoch": 171.71428571428572,
+      "grad_norm": 0.1435546875,
+      "learning_rate": 1.5498614798430972e-05,
+      "loss": 0.0122,
+      "step": 2404
+    },
+    {
+      "epoch": 171.78571428571428,
+      "grad_norm": 0.134765625,
+      "learning_rate": 1.549014223776112e-05,
+      "loss": 0.0151,
+      "step": 2405
+    },
+    {
+      "epoch": 171.85714285714286,
+      "grad_norm": 0.11669921875,
+      "learning_rate": 1.5481668219732263e-05,
+      "loss": 0.0117,
+      "step": 2406
+    },
+    {
+      "epoch": 171.92857142857142,
+      "grad_norm": 0.1181640625,
+      "learning_rate": 1.547319274847454e-05,
+      "loss": 0.0149,
+      "step": 2407
+    },
+    {
+      "epoch": 172.0,
+      "grad_norm": 0.279296875,
+      "learning_rate": 1.546471582811879e-05,
+      "loss": 0.0167,
+      "step": 2408
+    },
+    {
+      "epoch": 172.07142857142858,
+      "grad_norm": 0.1123046875,
+      "learning_rate": 1.545623746279656e-05,
+      "loss": 0.0123,
+      "step": 2409
+    },
+    {
+      "epoch": 172.14285714285714,
+      "grad_norm": 0.1982421875,
+      "learning_rate": 1.5447757656640104e-05,
+      "loss": 0.0122,
+      "step": 2410
+    },
+    {
+      "epoch": 172.21428571428572,
+      "grad_norm": 0.40234375,
+      "learning_rate": 1.5439276413782375e-05,
+      "loss": 0.0131,
+      "step": 2411
+    },
+    {
+      "epoch": 172.28571428571428,
+      "grad_norm": 0.2734375,
+      "learning_rate": 1.543079373835703e-05,
+      "loss": 0.0149,
+      "step": 2412
+    },
+    {
+      "epoch": 172.35714285714286,
+      "grad_norm": 0.259765625,
+      "learning_rate": 1.5422309634498423e-05,
+      "loss": 0.0143,
+      "step": 2413
+    },
+    {
+      "epoch": 172.42857142857142,
+      "grad_norm": 0.1435546875,
+      "learning_rate": 1.54138241063416e-05,
+      "loss": 0.0146,
+      "step": 2414
+    },
+    {
+      "epoch": 172.5,
+      "grad_norm": 0.12255859375,
+      "learning_rate": 1.54053371580223e-05,
+      "loss": 0.0114,
+      "step": 2415
+    },
+    {
+      "epoch": 172.57142857142858,
+      "grad_norm": 0.1318359375,
+      "learning_rate": 1.5396848793676966e-05,
+      "loss": 0.0124,
+      "step": 2416
+    },
+    {
+      "epoch": 172.64285714285714,
+      "grad_norm": 0.1259765625,
+      "learning_rate": 1.5388359017442727e-05,
+      "loss": 0.0131,
+      "step": 2417
+    },
+    {
+      "epoch": 172.71428571428572,
+      "grad_norm": 0.1669921875,
+      "learning_rate": 1.5379867833457386e-05,
+      "loss": 0.0155,
+      "step": 2418
+    },
+    {
+      "epoch": 172.78571428571428,
+      "grad_norm": 0.1435546875,
+      "learning_rate": 1.5371375245859446e-05,
+      "loss": 0.0138,
+      "step": 2419
+    },
+    {
+      "epoch": 172.85714285714286,
+      "grad_norm": 0.1123046875,
+      "learning_rate": 1.5362881258788098e-05,
+      "loss": 0.0121,
+      "step": 2420
+    },
+    {
+      "epoch": 172.92857142857142,
+      "grad_norm": 0.1328125,
+      "learning_rate": 1.5354385876383197e-05,
+      "loss": 0.0129,
+      "step": 2421
+    },
+    {
+      "epoch": 173.0,
+      "grad_norm": 0.13671875,
+      "learning_rate": 1.53458891027853e-05,
+      "loss": 0.0109,
+      "step": 2422
+    },
+    {
+      "epoch": 173.07142857142858,
+      "grad_norm": 0.12451171875,
+      "learning_rate": 1.5337390942135627e-05,
+      "loss": 0.0136,
+      "step": 2423
+    },
+    {
+      "epoch": 173.14285714285714,
+      "grad_norm": 0.1318359375,
+      "learning_rate": 1.5328891398576077e-05,
+      "loss": 0.0116,
+      "step": 2424
+    },
+    {
+      "epoch": 173.21428571428572,
+      "grad_norm": 0.201171875,
+      "learning_rate": 1.532039047624922e-05,
+      "loss": 0.013,
+      "step": 2425
+    },
+    {
+      "epoch": 173.28571428571428,
+      "grad_norm": 0.1484375,
+      "learning_rate": 1.5311888179298313e-05,
+      "loss": 0.0139,
+      "step": 2426
+    },
+    {
+      "epoch": 173.35714285714286,
+      "grad_norm": 0.1484375,
+      "learning_rate": 1.5303384511867268e-05,
+      "loss": 0.0131,
+      "step": 2427
+    },
+    {
+      "epoch": 173.42857142857142,
+      "grad_norm": 0.166015625,
+      "learning_rate": 1.5294879478100663e-05,
+      "loss": 0.0141,
+      "step": 2428
+    },
+    {
+      "epoch": 173.5,
+      "grad_norm": 0.1328125,
+      "learning_rate": 1.5286373082143757e-05,
+      "loss": 0.0145,
+      "step": 2429
+    },
+    {
+      "epoch": 173.57142857142858,
+      "grad_norm": 0.1357421875,
+      "learning_rate": 1.5277865328142457e-05,
+      "loss": 0.0115,
+      "step": 2430
+    },
+    {
+      "epoch": 173.64285714285714,
+      "grad_norm": 0.158203125,
+      "learning_rate": 1.5269356220243345e-05,
+      "loss": 0.0142,
+      "step": 2431
+    },
+    {
+      "epoch": 173.71428571428572,
+      "grad_norm": 0.162109375,
+      "learning_rate": 1.5260845762593647e-05,
+      "loss": 0.0147,
+      "step": 2432
+    },
+    {
+      "epoch": 173.78571428571428,
+      "grad_norm": 0.1357421875,
+      "learning_rate": 1.525233395934127e-05,
+      "loss": 0.0144,
+      "step": 2433
+    },
+    {
+      "epoch": 173.85714285714286,
+      "grad_norm": 0.1201171875,
+      "learning_rate": 1.5243820814634758e-05,
+      "loss": 0.0124,
+      "step": 2434
+    },
+    {
+      "epoch": 173.92857142857142,
+      "grad_norm": 0.1533203125,
+      "learning_rate": 1.5235306332623309e-05,
+      "loss": 0.0118,
+      "step": 2435
+    },
+    {
+      "epoch": 174.0,
+      "grad_norm": 0.1845703125,
+      "learning_rate": 1.5226790517456785e-05,
+      "loss": 0.0126,
+      "step": 2436
+    },
+    {
+      "epoch": 174.07142857142858,
+      "grad_norm": 0.1591796875,
+      "learning_rate": 1.5218273373285686e-05,
+      "loss": 0.0139,
+      "step": 2437
+    },
+    {
+      "epoch": 174.14285714285714,
+      "grad_norm": 0.236328125,
+      "learning_rate": 1.5209754904261167e-05,
+      "loss": 0.0136,
+      "step": 2438
+    },
+    {
+      "epoch": 174.21428571428572,
+      "grad_norm": 0.11767578125,
+      "learning_rate": 1.5201235114535028e-05,
+      "loss": 0.0127,
+      "step": 2439
+    },
+    {
+      "epoch": 174.28571428571428,
+      "grad_norm": 0.1650390625,
+      "learning_rate": 1.5192714008259706e-05,
+      "loss": 0.015,
+      "step": 2440
+    },
+    {
+      "epoch": 174.35714285714286,
+      "grad_norm": 0.1748046875,
+      "learning_rate": 1.5184191589588292e-05,
+      "loss": 0.0139,
+      "step": 2441
+    },
+    {
+      "epoch": 174.42857142857142,
+      "grad_norm": 0.130859375,
+      "learning_rate": 1.51756678626745e-05,
+      "loss": 0.0118,
+      "step": 2442
+    },
+    {
+      "epoch": 174.5,
+      "grad_norm": 0.21484375,
+      "learning_rate": 1.5167142831672702e-05,
+      "loss": 0.0157,
+      "step": 2443
+    },
+    {
+      "epoch": 174.57142857142858,
+      "grad_norm": 0.1279296875,
+      "learning_rate": 1.5158616500737885e-05,
+      "loss": 0.0135,
+      "step": 2444
+    },
+    {
+      "epoch": 174.64285714285714,
+      "grad_norm": 0.203125,
+      "learning_rate": 1.5150088874025686e-05,
+      "loss": 0.0117,
+      "step": 2445
+    },
+    {
+      "epoch": 174.71428571428572,
+      "grad_norm": 0.1591796875,
+      "learning_rate": 1.514155995569236e-05,
+      "loss": 0.0118,
+      "step": 2446
+    },
+    {
+      "epoch": 174.78571428571428,
+      "grad_norm": 0.1171875,
+      "learning_rate": 1.51330297498948e-05,
+      "loss": 0.0135,
+      "step": 2447
+    },
+    {
+      "epoch": 174.85714285714286,
+      "grad_norm": 0.162109375,
+      "learning_rate": 1.5124498260790535e-05,
+      "loss": 0.015,
+      "step": 2448
+    },
+    {
+      "epoch": 174.92857142857142,
+      "grad_norm": 0.1328125,
+      "learning_rate": 1.511596549253769e-05,
+      "loss": 0.0154,
+      "step": 2449
+    },
+    {
+      "epoch": 175.0,
+      "grad_norm": 0.1416015625,
+      "learning_rate": 1.5107431449295047e-05,
+      "loss": 0.0134,
+      "step": 2450
+    },
+    {
+      "epoch": 175.07142857142858,
+      "grad_norm": 0.208984375,
+      "learning_rate": 1.5098896135221991e-05,
+      "loss": 0.0161,
+      "step": 2451
+    },
+    {
+      "epoch": 175.14285714285714,
+      "grad_norm": 0.125,
+      "learning_rate": 1.5090359554478531e-05,
+      "loss": 0.0117,
+      "step": 2452
+    },
+    {
+      "epoch": 175.21428571428572,
+      "grad_norm": 0.1376953125,
+      "learning_rate": 1.5081821711225295e-05,
+      "loss": 0.0125,
+      "step": 2453
+    },
+    {
+      "epoch": 175.28571428571428,
+      "grad_norm": 0.1328125,
+      "learning_rate": 1.5073282609623518e-05,
+      "loss": 0.0134,
+      "step": 2454
+    },
+    {
+      "epoch": 175.35714285714286,
+      "grad_norm": 0.1396484375,
+      "learning_rate": 1.5064742253835057e-05,
+      "loss": 0.0122,
+      "step": 2455
+    },
+    {
+      "epoch": 175.42857142857142,
+      "grad_norm": 0.125,
+      "learning_rate": 1.5056200648022378e-05,
+      "loss": 0.0137,
+      "step": 2456
+    },
+    {
+      "epoch": 175.5,
+      "grad_norm": 0.1611328125,
+      "learning_rate": 1.5047657796348563e-05,
+      "loss": 0.0127,
+      "step": 2457
+    },
+    {
+      "epoch": 175.57142857142858,
+      "grad_norm": 0.1611328125,
+      "learning_rate": 1.5039113702977278e-05,
+      "loss": 0.013,
+      "step": 2458
+    },
+    {
+      "epoch": 175.64285714285714,
+      "grad_norm": 0.14453125,
+      "learning_rate": 1.5030568372072826e-05,
+      "loss": 0.0156,
+      "step": 2459
+    },
+    {
+      "epoch": 175.71428571428572,
+      "grad_norm": 0.1484375,
+      "learning_rate": 1.5022021807800087e-05,
+      "loss": 0.015,
+      "step": 2460
+    },
+    {
+      "epoch": 175.78571428571428,
+      "grad_norm": 0.181640625,
+      "learning_rate": 1.501347401432456e-05,
+      "loss": 0.0105,
+      "step": 2461
+    },
+    {
+      "epoch": 175.85714285714286,
+      "grad_norm": 0.3515625,
+      "learning_rate": 1.5004924995812331e-05,
+      "loss": 0.0135,
+      "step": 2462
+    },
+    {
+      "epoch": 175.92857142857142,
+      "grad_norm": 0.140625,
+      "learning_rate": 1.4996374756430087e-05,
+      "loss": 0.0135,
+      "step": 2463
+    },
+    {
+      "epoch": 176.0,
+      "grad_norm": 0.1806640625,
+      "learning_rate": 1.4987823300345117e-05,
+      "loss": 0.0105,
+      "step": 2464
+    },
+    {
+      "epoch": 176.07142857142858,
+      "grad_norm": 0.11181640625,
+      "learning_rate": 1.4979270631725292e-05,
+      "loss": 0.0131,
+      "step": 2465
+    },
+    {
+      "epoch": 176.14285714285714,
+      "grad_norm": 0.1533203125,
+      "learning_rate": 1.4970716754739078e-05,
+      "loss": 0.0126,
+      "step": 2466
+    },
+    {
+      "epoch": 176.21428571428572,
+      "grad_norm": 0.1708984375,
+      "learning_rate": 1.4962161673555537e-05,
+      "loss": 0.0127,
+      "step": 2467
+    },
+    {
+      "epoch": 176.28571428571428,
+      "grad_norm": 0.13671875,
+      "learning_rate": 1.49536053923443e-05,
+      "loss": 0.0147,
+      "step": 2468
+    },
+    {
+      "epoch": 176.35714285714286,
+      "grad_norm": 0.306640625,
+      "learning_rate": 1.4945047915275604e-05,
+      "loss": 0.0147,
+      "step": 2469
+    },
+    {
+      "epoch": 176.42857142857142,
+      "grad_norm": 0.1201171875,
+      "learning_rate": 1.4936489246520257e-05,
+      "loss": 0.0129,
+      "step": 2470
+    },
+    {
+      "epoch": 176.5,
+      "grad_norm": 0.2470703125,
+      "learning_rate": 1.492792939024965e-05,
+      "loss": 0.0133,
+      "step": 2471
+    },
+    {
+      "epoch": 176.57142857142858,
+      "grad_norm": 0.126953125,
+      "learning_rate": 1.4919368350635757e-05,
+      "loss": 0.0116,
+      "step": 2472
+    },
+    {
+      "epoch": 176.64285714285714,
+      "grad_norm": 0.244140625,
+      "learning_rate": 1.4910806131851113e-05,
+      "loss": 0.0138,
+      "step": 2473
+    },
+    {
+      "epoch": 176.71428571428572,
+      "grad_norm": 0.142578125,
+      "learning_rate": 1.4902242738068853e-05,
+      "loss": 0.0129,
+      "step": 2474
+    },
+    {
+      "epoch": 176.78571428571428,
+      "grad_norm": 0.1787109375,
+      "learning_rate": 1.4893678173462664e-05,
+      "loss": 0.0169,
+      "step": 2475
+    },
+    {
+      "epoch": 176.85714285714286,
+      "grad_norm": 0.26171875,
+      "learning_rate": 1.4885112442206812e-05,
+      "loss": 0.0142,
+      "step": 2476
+    },
+    {
+      "epoch": 176.92857142857142,
+      "grad_norm": 0.1171875,
+      "learning_rate": 1.4876545548476128e-05,
+      "loss": 0.0162,
+      "step": 2477
+    },
+    {
+      "epoch": 177.0,
+      "grad_norm": 0.1552734375,
+      "learning_rate": 1.4867977496446017e-05,
+      "loss": 0.0145,
+      "step": 2478
+    },
+    {
+      "epoch": 177.07142857142858,
+      "grad_norm": 0.193359375,
+      "learning_rate": 1.4859408290292442e-05,
+      "loss": 0.0152,
+      "step": 2479
+    },
+    {
+      "epoch": 177.14285714285714,
+      "grad_norm": 0.259765625,
+      "learning_rate": 1.4850837934191924e-05,
+      "loss": 0.0116,
+      "step": 2480
+    },
+    {
+      "epoch": 177.21428571428572,
+      "grad_norm": 0.291015625,
+      "learning_rate": 1.4842266432321561e-05,
+      "loss": 0.0131,
+      "step": 2481
+    },
+    {
+      "epoch": 177.28571428571428,
+      "grad_norm": 0.1533203125,
+      "learning_rate": 1.4833693788858988e-05,
+      "loss": 0.0114,
+      "step": 2482
+    },
+    {
+      "epoch": 177.35714285714286,
+      "grad_norm": 0.279296875,
+      "learning_rate": 1.4825120007982413e-05,
+      "loss": 0.0133,
+      "step": 2483
+    },
+    {
+      "epoch": 177.42857142857142,
+      "grad_norm": 0.3125,
+      "learning_rate": 1.4816545093870593e-05,
+      "loss": 0.0138,
+      "step": 2484
+    },
+    {
+      "epoch": 177.5,
+      "grad_norm": 0.1416015625,
+      "learning_rate": 1.4807969050702838e-05,
+      "loss": 0.015,
+      "step": 2485
+    },
+    {
+      "epoch": 177.57142857142858,
+      "grad_norm": 0.20703125,
+      "learning_rate": 1.4799391882659004e-05,
+      "loss": 0.0143,
+      "step": 2486
+    },
+    {
+      "epoch": 177.64285714285714,
+      "grad_norm": 0.125,
+      "learning_rate": 1.4790813593919497e-05,
+      "loss": 0.0148,
+      "step": 2487
+    },
+    {
+      "epoch": 177.71428571428572,
+      "grad_norm": 0.12890625,
+      "learning_rate": 1.4782234188665278e-05,
+      "loss": 0.0146,
+      "step": 2488
+    },
+    {
+      "epoch": 177.78571428571428,
+      "grad_norm": 0.169921875,
+      "learning_rate": 1.4773653671077836e-05,
+      "loss": 0.013,
+      "step": 2489
+    },
+    {
+      "epoch": 177.85714285714286,
+      "grad_norm": 0.1435546875,
+      "learning_rate": 1.4765072045339218e-05,
+      "loss": 0.0126,
+      "step": 2490
+    },
+    {
+      "epoch": 177.92857142857142,
+      "grad_norm": 0.162109375,
+      "learning_rate": 1.4756489315632003e-05,
+      "loss": 0.0117,
+      "step": 2491
+    },
+    {
+      "epoch": 178.0,
+      "grad_norm": 0.1806640625,
+      "learning_rate": 1.4747905486139305e-05,
+      "loss": 0.013,
+      "step": 2492
+    },
+    {
+      "epoch": 178.07142857142858,
+      "grad_norm": 0.13671875,
+      "learning_rate": 1.4739320561044781e-05,
+      "loss": 0.0152,
+      "step": 2493
+    },
+    {
+      "epoch": 178.14285714285714,
+      "grad_norm": 0.150390625,
+      "learning_rate": 1.4730734544532617e-05,
+      "loss": 0.0129,
+      "step": 2494
+    },
+    {
+      "epoch": 178.21428571428572,
+      "grad_norm": 0.2001953125,
+      "learning_rate": 1.472214744078754e-05,
+      "loss": 0.0159,
+      "step": 2495
+    },
+    {
+      "epoch": 178.28571428571428,
+      "grad_norm": 0.20703125,
+      "learning_rate": 1.4713559253994793e-05,
+      "loss": 0.0126,
+      "step": 2496
+    },
+    {
+      "epoch": 178.35714285714286,
+      "grad_norm": 0.146484375,
+      "learning_rate": 1.4704969988340156e-05,
+      "loss": 0.0135,
+      "step": 2497
+    },
+    {
+      "epoch": 178.42857142857142,
+      "grad_norm": 0.1171875,
+      "learning_rate": 1.4696379648009933e-05,
+      "loss": 0.0125,
+      "step": 2498
+    },
+    {
+      "epoch": 178.5,
+      "grad_norm": 0.1259765625,
+      "learning_rate": 1.4687788237190952e-05,
+      "loss": 0.0137,
+      "step": 2499
+    },
+    {
+      "epoch": 178.57142857142858,
+      "grad_norm": 0.1376953125,
+      "learning_rate": 1.4679195760070561e-05,
+      "loss": 0.0124,
+      "step": 2500
+    },
+    {
+      "epoch": 178.64285714285714,
+      "grad_norm": 0.2255859375,
+      "learning_rate": 1.4670602220836633e-05,
+      "loss": 0.0156,
+      "step": 2501
+    },
+    {
+      "epoch": 178.71428571428572,
+      "grad_norm": 0.1357421875,
+      "learning_rate": 1.4662007623677549e-05,
+      "loss": 0.014,
+      "step": 2502
+    },
+    {
+      "epoch": 178.78571428571428,
+      "grad_norm": 0.47265625,
+      "learning_rate": 1.4653411972782211e-05,
+      "loss": 0.0176,
+      "step": 2503
+    },
+    {
+      "epoch": 178.85714285714286,
+      "grad_norm": 0.12060546875,
+      "learning_rate": 1.4644815272340037e-05,
+      "loss": 0.0121,
+      "step": 2504
+    },
+    {
+      "epoch": 178.92857142857142,
+      "grad_norm": 0.11767578125,
+      "learning_rate": 1.4636217526540961e-05,
+      "loss": 0.0122,
+      "step": 2505
+    },
+    {
+      "epoch": 179.0,
+      "grad_norm": 0.1318359375,
+      "learning_rate": 1.4627618739575408e-05,
+      "loss": 0.0152,
+      "step": 2506
+    },
+    {
+      "epoch": 179.07142857142858,
+      "grad_norm": 0.1298828125,
+      "learning_rate": 1.461901891563433e-05,
+      "loss": 0.0127,
+      "step": 2507
+    },
+    {
+      "epoch": 179.14285714285714,
+      "grad_norm": 0.12158203125,
+      "learning_rate": 1.4610418058909172e-05,
+      "loss": 0.0145,
+      "step": 2508
+    },
+    {
+      "epoch": 179.21428571428572,
+      "grad_norm": 0.119140625,
+      "learning_rate": 1.4601816173591895e-05,
+      "loss": 0.0143,
+      "step": 2509
+    },
+    {
+      "epoch": 179.28571428571428,
+      "grad_norm": 0.1396484375,
+      "learning_rate": 1.4593213263874944e-05,
+      "loss": 0.0127,
+      "step": 2510
+    },
+    {
+      "epoch": 179.35714285714286,
+      "grad_norm": 0.1240234375,
+      "learning_rate": 1.458460933395128e-05,
+      "loss": 0.0135,
+      "step": 2511
+    },
+    {
+      "epoch": 179.42857142857142,
+      "grad_norm": 0.1630859375,
+      "learning_rate": 1.457600438801435e-05,
+      "loss": 0.0153,
+      "step": 2512
+    },
+    {
+      "epoch": 179.5,
+      "grad_norm": 0.279296875,
+      "learning_rate": 1.4567398430258097e-05,
+      "loss": 0.0131,
+      "step": 2513
+    },
+    {
+      "epoch": 179.57142857142858,
+      "grad_norm": 0.2412109375,
+      "learning_rate": 1.4558791464876966e-05,
+      "loss": 0.0136,
+      "step": 2514
+    },
+    {
+      "epoch": 179.64285714285714,
+      "grad_norm": 0.140625,
+      "learning_rate": 1.455018349606589e-05,
+      "loss": 0.0127,
+      "step": 2515
+    },
+    {
+      "epoch": 179.71428571428572,
+      "grad_norm": 0.287109375,
+      "learning_rate": 1.454157452802028e-05,
+      "loss": 0.0157,
+      "step": 2516
+    },
+    {
+      "epoch": 179.78571428571428,
+      "grad_norm": 0.2490234375,
+      "learning_rate": 1.4532964564936047e-05,
+      "loss": 0.0119,
+      "step": 2517
+    },
+    {
+      "epoch": 179.85714285714286,
+      "grad_norm": 0.10986328125,
+      "learning_rate": 1.4524353611009577e-05,
+      "loss": 0.0114,
+      "step": 2518
+    },
+    {
+      "epoch": 179.92857142857142,
+      "grad_norm": 0.328125,
+      "learning_rate": 1.451574167043776e-05,
+      "loss": 0.0144,
+      "step": 2519
+    },
+    {
+      "epoch": 180.0,
+      "grad_norm": 0.13671875,
+      "learning_rate": 1.4507128747417931e-05,
+      "loss": 0.0137,
+      "step": 2520
+    },
+    {
+      "epoch": 180.07142857142858,
+      "grad_norm": 0.15234375,
+      "learning_rate": 1.4498514846147937e-05,
+      "loss": 0.0112,
+      "step": 2521
+    },
+    {
+      "epoch": 180.14285714285714,
+      "grad_norm": 0.1279296875,
+      "learning_rate": 1.4489899970826085e-05,
+      "loss": 0.0125,
+      "step": 2522
+    },
+    {
+      "epoch": 180.21428571428572,
+      "grad_norm": 0.1279296875,
+      "learning_rate": 1.4481284125651162e-05,
+      "loss": 0.0111,
+      "step": 2523
+    },
+    {
+      "epoch": 180.28571428571428,
+      "grad_norm": 0.27734375,
+      "learning_rate": 1.4472667314822425e-05,
+      "loss": 0.0169,
+      "step": 2524
+    },
+    {
+      "epoch": 180.35714285714286,
+      "grad_norm": 0.298828125,
+      "learning_rate": 1.4464049542539598e-05,
+      "loss": 0.0154,
+      "step": 2525
+    },
+    {
+      "epoch": 180.42857142857142,
+      "grad_norm": 0.1474609375,
+      "learning_rate": 1.4455430813002888e-05,
+      "loss": 0.0129,
+      "step": 2526
+    },
+    {
+      "epoch": 180.5,
+      "grad_norm": 0.1123046875,
+      "learning_rate": 1.4446811130412952e-05,
+      "loss": 0.015,
+      "step": 2527
+    },
+    {
+      "epoch": 180.57142857142858,
+      "grad_norm": 0.16015625,
+      "learning_rate": 1.4438190498970924e-05,
+      "loss": 0.0134,
+      "step": 2528
+    },
+    {
+      "epoch": 180.64285714285714,
+      "grad_norm": 0.1474609375,
+      "learning_rate": 1.4429568922878392e-05,
+      "loss": 0.0128,
+      "step": 2529
+    },
+    {
+      "epoch": 180.71428571428572,
+      "grad_norm": 0.10595703125,
+      "learning_rate": 1.4420946406337404e-05,
+      "loss": 0.0134,
+      "step": 2530
+    },
+    {
+      "epoch": 180.78571428571428,
+      "grad_norm": 0.150390625,
+      "learning_rate": 1.4412322953550477e-05,
+      "loss": 0.0161,
+      "step": 2531
+    },
+    {
+      "epoch": 180.85714285714286,
+      "grad_norm": 0.259765625,
+      "learning_rate": 1.4403698568720572e-05,
+      "loss": 0.0117,
+      "step": 2532
+    },
+    {
+      "epoch": 180.92857142857142,
+      "grad_norm": 0.140625,
+      "learning_rate": 1.4395073256051114e-05,
+      "loss": 0.0117,
+      "step": 2533
+    },
+    {
+      "epoch": 181.0,
+      "grad_norm": 0.126953125,
+      "learning_rate": 1.438644701974597e-05,
+      "loss": 0.0143,
+      "step": 2534
+    },
+    {
+      "epoch": 181.07142857142858,
+      "grad_norm": 0.12353515625,
+      "learning_rate": 1.4377819864009465e-05,
+      "loss": 0.0123,
+      "step": 2535
+    },
+    {
+      "epoch": 181.14285714285714,
+      "grad_norm": 0.11474609375,
+      "learning_rate": 1.4369191793046372e-05,
+      "loss": 0.0133,
+      "step": 2536
+    },
+    {
+      "epoch": 181.21428571428572,
+      "grad_norm": 0.2099609375,
+      "learning_rate": 1.4360562811061903e-05,
+      "loss": 0.0161,
+      "step": 2537
+    },
+    {
+      "epoch": 181.28571428571428,
+      "grad_norm": 0.14453125,
+      "learning_rate": 1.4351932922261726e-05,
+      "loss": 0.0136,
+      "step": 2538
+    },
+    {
+      "epoch": 181.35714285714286,
+      "grad_norm": 0.1298828125,
+      "learning_rate": 1.4343302130851938e-05,
+      "loss": 0.0145,
+      "step": 2539
+    },
+    {
+      "epoch": 181.42857142857142,
+      "grad_norm": 0.1171875,
+      "learning_rate": 1.4334670441039083e-05,
+      "loss": 0.0119,
+      "step": 2540
+    },
+    {
+      "epoch": 181.5,
+      "grad_norm": 0.11181640625,
+      "learning_rate": 1.4326037857030148e-05,
+      "loss": 0.0113,
+      "step": 2541
+    },
+    {
+      "epoch": 181.57142857142858,
+      "grad_norm": 0.1416015625,
+      "learning_rate": 1.431740438303254e-05,
+      "loss": 0.015,
+      "step": 2542
+    },
+    {
+      "epoch": 181.64285714285714,
+      "grad_norm": 0.2216796875,
+      "learning_rate": 1.4308770023254114e-05,
+      "loss": 0.0138,
+      "step": 2543
+    },
+    {
+      "epoch": 181.71428571428572,
+      "grad_norm": 0.125,
+      "learning_rate": 1.4300134781903151e-05,
+      "loss": 0.015,
+      "step": 2544
+    },
+    {
+      "epoch": 181.78571428571428,
+      "grad_norm": 0.12060546875,
+      "learning_rate": 1.429149866318837e-05,
+      "loss": 0.0143,
+      "step": 2545
+    },
+    {
+      "epoch": 181.85714285714286,
+      "grad_norm": 0.2451171875,
+      "learning_rate": 1.4282861671318898e-05,
+      "loss": 0.0173,
+      "step": 2546
+    },
+    {
+      "epoch": 181.92857142857142,
+      "grad_norm": 0.1640625,
+      "learning_rate": 1.4274223810504305e-05,
+      "loss": 0.012,
+      "step": 2547
+    },
+    {
+      "epoch": 182.0,
+      "grad_norm": 0.166015625,
+      "learning_rate": 1.4265585084954586e-05,
+      "loss": 0.0138,
+      "step": 2548
+    },
+    {
+      "epoch": 182.07142857142858,
+      "grad_norm": 0.13671875,
+      "learning_rate": 1.4256945498880136e-05,
+      "loss": 0.0148,
+      "step": 2549
+    },
+    {
+      "epoch": 182.14285714285714,
+      "grad_norm": 0.1669921875,
+      "learning_rate": 1.4248305056491804e-05,
+      "loss": 0.0125,
+      "step": 2550
+    },
+    {
+      "epoch": 182.21428571428572,
+      "grad_norm": 0.1787109375,
+      "learning_rate": 1.4239663762000818e-05,
+      "loss": 0.0152,
+      "step": 2551
+    },
+    {
+      "epoch": 182.28571428571428,
+      "grad_norm": 0.16796875,
+      "learning_rate": 1.4231021619618854e-05,
+      "loss": 0.02,
+      "step": 2552
+    },
+    {
+      "epoch": 182.35714285714286,
+      "grad_norm": 0.11865234375,
+      "learning_rate": 1.4222378633557976e-05,
+      "loss": 0.0101,
+      "step": 2553
+    },
+    {
+      "epoch": 182.42857142857142,
+      "grad_norm": 0.12060546875,
+      "learning_rate": 1.4213734808030682e-05,
+      "loss": 0.0129,
+      "step": 2554
+    },
+    {
+      "epoch": 182.5,
+      "grad_norm": 0.1865234375,
+      "learning_rate": 1.4205090147249856e-05,
+      "loss": 0.0149,
+      "step": 2555
+    },
+    {
+      "epoch": 182.57142857142858,
+      "grad_norm": 0.11865234375,
+      "learning_rate": 1.4196444655428806e-05,
+      "loss": 0.0125,
+      "step": 2556
+    },
+    {
+      "epoch": 182.64285714285714,
+      "grad_norm": 0.1591796875,
+      "learning_rate": 1.4187798336781241e-05,
+      "loss": 0.0139,
+      "step": 2557
+    },
+    {
+      "epoch": 182.71428571428572,
+      "grad_norm": 0.1796875,
+      "learning_rate": 1.4179151195521268e-05,
+      "loss": 0.014,
+      "step": 2558
+    },
+    {
+      "epoch": 182.78571428571428,
+      "grad_norm": 0.1337890625,
+      "learning_rate": 1.4170503235863408e-05,
+      "loss": 0.0158,
+      "step": 2559
+    },
+    {
+      "epoch": 182.85714285714286,
+      "grad_norm": 0.12158203125,
+      "learning_rate": 1.4161854462022556e-05,
+      "loss": 0.013,
+      "step": 2560
+    },
+    {
+      "epoch": 182.92857142857142,
+      "grad_norm": 0.150390625,
+      "learning_rate": 1.4153204878214032e-05,
+      "loss": 0.0117,
+      "step": 2561
+    },
+    {
+      "epoch": 183.0,
+      "grad_norm": 0.11572265625,
+      "learning_rate": 1.4144554488653533e-05,
+      "loss": 0.0144,
+      "step": 2562
+    },
+    {
+      "epoch": 183.07142857142858,
+      "grad_norm": 0.189453125,
+      "learning_rate": 1.4135903297557155e-05,
+      "loss": 0.0113,
+      "step": 2563
+    },
+    {
+      "epoch": 183.14285714285714,
+      "grad_norm": 0.11376953125,
+      "learning_rate": 1.4127251309141381e-05,
+      "loss": 0.0106,
+      "step": 2564
+    },
+    {
+      "epoch": 183.21428571428572,
+      "grad_norm": 0.11279296875,
+      "learning_rate": 1.4118598527623086e-05,
+      "loss": 0.012,
+      "step": 2565
+    },
+    {
+      "epoch": 183.28571428571428,
+      "grad_norm": 0.1630859375,
+      "learning_rate": 1.4109944957219537e-05,
+      "loss": 0.0138,
+      "step": 2566
+    },
+    {
+      "epoch": 183.35714285714286,
+      "grad_norm": 0.1171875,
+      "learning_rate": 1.4101290602148371e-05,
+      "loss": 0.0141,
+      "step": 2567
+    },
+    {
+      "epoch": 183.42857142857142,
+      "grad_norm": 0.2353515625,
+      "learning_rate": 1.4092635466627618e-05,
+      "loss": 0.0143,
+      "step": 2568
+    },
+    {
+      "epoch": 183.5,
+      "grad_norm": 0.20703125,
+      "learning_rate": 1.4083979554875685e-05,
+      "loss": 0.014,
+      "step": 2569
+    },
+    {
+      "epoch": 183.57142857142858,
+      "grad_norm": 0.1181640625,
+      "learning_rate": 1.4075322871111357e-05,
+      "loss": 0.0134,
+      "step": 2570
+    },
+    {
+      "epoch": 183.64285714285714,
+      "grad_norm": 0.1806640625,
+      "learning_rate": 1.4066665419553804e-05,
+      "loss": 0.0153,
+      "step": 2571
+    },
+    {
+      "epoch": 183.71428571428572,
+      "grad_norm": 0.1962890625,
+      "learning_rate": 1.4058007204422558e-05,
+      "loss": 0.0124,
+      "step": 2572
+    },
+    {
+      "epoch": 183.78571428571428,
+      "grad_norm": 0.17578125,
+      "learning_rate": 1.4049348229937523e-05,
+      "loss": 0.0152,
+      "step": 2573
+    },
+    {
+      "epoch": 183.85714285714286,
+      "grad_norm": 0.2412109375,
+      "learning_rate": 1.4040688500318985e-05,
+      "loss": 0.0153,
+      "step": 2574
+    },
+    {
+      "epoch": 183.92857142857142,
+      "grad_norm": 0.1455078125,
+      "learning_rate": 1.4032028019787585e-05,
+      "loss": 0.0162,
+      "step": 2575
+    },
+    {
+      "epoch": 184.0,
+      "grad_norm": 0.1640625,
+      "learning_rate": 1.4023366792564343e-05,
+      "loss": 0.0144,
+      "step": 2576
+    },
+    {
+      "epoch": 184.07142857142858,
+      "grad_norm": 0.1396484375,
+      "learning_rate": 1.4014704822870634e-05,
+      "loss": 0.0137,
+      "step": 2577
+    },
+    {
+      "epoch": 184.14285714285714,
+      "grad_norm": 0.2177734375,
+      "learning_rate": 1.4006042114928191e-05,
+      "loss": 0.014,
+      "step": 2578
+    },
+    {
+      "epoch": 184.21428571428572,
+      "grad_norm": 0.26953125,
+      "learning_rate": 1.3997378672959122e-05,
+      "loss": 0.0115,
+      "step": 2579
+    },
+    {
+      "epoch": 184.28571428571428,
+      "grad_norm": 0.26171875,
+      "learning_rate": 1.3988714501185873e-05,
+      "loss": 0.0135,
+      "step": 2580
+    },
+    {
+      "epoch": 184.35714285714286,
+      "grad_norm": 0.138671875,
+      "learning_rate": 1.3980049603831266e-05,
+      "loss": 0.0126,
+      "step": 2581
+    },
+    {
+      "epoch": 184.42857142857142,
+      "grad_norm": 0.1455078125,
+      "learning_rate": 1.3971383985118455e-05,
+      "loss": 0.0136,
+      "step": 2582
+    },
+    {
+      "epoch": 184.5,
+      "grad_norm": 0.11181640625,
+      "learning_rate": 1.396271764927097e-05,
+      "loss": 0.0133,
+      "step": 2583
+    },
+    {
+      "epoch": 184.57142857142858,
+      "grad_norm": 0.123046875,
+      "learning_rate": 1.3954050600512669e-05,
+      "loss": 0.0129,
+      "step": 2584
+    },
+    {
+      "epoch": 184.64285714285714,
+      "grad_norm": 0.1796875,
+      "learning_rate": 1.3945382843067772e-05,
+      "loss": 0.0113,
+      "step": 2585
+    },
+    {
+      "epoch": 184.71428571428572,
+      "grad_norm": 0.28515625,
+      "learning_rate": 1.3936714381160835e-05,
+      "loss": 0.0112,
+      "step": 2586
+    },
+    {
+      "epoch": 184.78571428571428,
+      "grad_norm": 0.181640625,
+      "learning_rate": 1.3928045219016761e-05,
+      "loss": 0.0133,
+      "step": 2587
+    },
+    {
+      "epoch": 184.85714285714286,
+      "grad_norm": 0.185546875,
+      "learning_rate": 1.3919375360860795e-05,
+      "loss": 0.0144,
+      "step": 2588
+    },
+    {
+      "epoch": 184.92857142857142,
+      "grad_norm": 0.111328125,
+      "learning_rate": 1.391070481091852e-05,
+      "loss": 0.0131,
+      "step": 2589
+    },
+    {
+      "epoch": 185.0,
+      "grad_norm": 0.29296875,
+      "learning_rate": 1.3902033573415862e-05,
+      "loss": 0.0137,
+      "step": 2590
+    },
+    {
+      "epoch": 185.07142857142858,
+      "grad_norm": 0.30078125,
+      "learning_rate": 1.3893361652579068e-05,
+      "loss": 0.0189,
+      "step": 2591
+    },
+    {
+      "epoch": 185.14285714285714,
+      "grad_norm": 0.12109375,
+      "learning_rate": 1.3884689052634733e-05,
+      "loss": 0.0153,
+      "step": 2592
+    },
+    {
+      "epoch": 185.21428571428572,
+      "grad_norm": 0.2451171875,
+      "learning_rate": 1.3876015777809778e-05,
+      "loss": 0.013,
+      "step": 2593
+    },
+    {
+      "epoch": 185.28571428571428,
+      "grad_norm": 0.1845703125,
+      "learning_rate": 1.3867341832331446e-05,
+      "loss": 0.013,
+      "step": 2594
+    },
+    {
+      "epoch": 185.35714285714286,
+      "grad_norm": 0.12158203125,
+      "learning_rate": 1.3858667220427315e-05,
+      "loss": 0.0139,
+      "step": 2595
+    },
+    {
+      "epoch": 185.42857142857142,
+      "grad_norm": 0.10791015625,
+      "learning_rate": 1.3849991946325287e-05,
+      "loss": 0.0134,
+      "step": 2596
+    },
+    {
+      "epoch": 185.5,
+      "grad_norm": 0.1220703125,
+      "learning_rate": 1.3841316014253584e-05,
+      "loss": 0.0091,
+      "step": 2597
+    },
+    {
+      "epoch": 185.57142857142858,
+      "grad_norm": 0.1494140625,
+      "learning_rate": 1.383263942844075e-05,
+      "loss": 0.0132,
+      "step": 2598
+    },
+    {
+      "epoch": 185.64285714285714,
+      "grad_norm": 0.150390625,
+      "learning_rate": 1.382396219311565e-05,
+      "loss": 0.0128,
+      "step": 2599
+    },
+    {
+      "epoch": 185.71428571428572,
+      "grad_norm": 0.15625,
+      "learning_rate": 1.3815284312507459e-05,
+      "loss": 0.0163,
+      "step": 2600
+    },
+    {
+      "epoch": 185.78571428571428,
+      "grad_norm": 0.11376953125,
+      "learning_rate": 1.380660579084567e-05,
+      "loss": 0.0126,
+      "step": 2601
+    },
+    {
+      "epoch": 185.85714285714286,
+      "grad_norm": 0.1259765625,
+      "learning_rate": 1.3797926632360092e-05,
+      "loss": 0.0136,
+      "step": 2602
+    },
+    {
+      "epoch": 185.92857142857142,
+      "grad_norm": 0.1318359375,
+      "learning_rate": 1.3789246841280842e-05,
+      "loss": 0.0138,
+      "step": 2603
+    },
+    {
+      "epoch": 186.0,
+      "grad_norm": 0.119140625,
+      "learning_rate": 1.3780566421838346e-05,
+      "loss": 0.0125,
+      "step": 2604
+    },
+    {
+      "epoch": 186.07142857142858,
+      "grad_norm": 0.123046875,
+      "learning_rate": 1.377188537826333e-05,
+      "loss": 0.0142,
+      "step": 2605
+    },
+    {
+      "epoch": 186.14285714285714,
+      "grad_norm": 0.12255859375,
+      "learning_rate": 1.3763203714786829e-05,
+      "loss": 0.0123,
+      "step": 2606
+    },
+    {
+      "epoch": 186.21428571428572,
+      "grad_norm": 0.1435546875,
+      "learning_rate": 1.3754521435640186e-05,
+      "loss": 0.0128,
+      "step": 2607
+    },
+    {
+      "epoch": 186.28571428571428,
+      "grad_norm": 0.1552734375,
+      "learning_rate": 1.3745838545055038e-05,
+      "loss": 0.0127,
+      "step": 2608
+    },
+    {
+      "epoch": 186.35714285714286,
+      "grad_norm": 0.1708984375,
+      "learning_rate": 1.3737155047263315e-05,
+      "loss": 0.0112,
+      "step": 2609
+    },
+    {
+      "epoch": 186.42857142857142,
+      "grad_norm": 0.12109375,
+      "learning_rate": 1.3728470946497248e-05,
+      "loss": 0.0127,
+      "step": 2610
+    },
+    {
+      "epoch": 186.5,
+      "grad_norm": 0.11376953125,
+      "learning_rate": 1.3719786246989368e-05,
+      "loss": 0.0121,
+      "step": 2611
+    },
+    {
+      "epoch": 186.57142857142858,
+      "grad_norm": 0.203125,
+      "learning_rate": 1.3711100952972491e-05,
+      "loss": 0.0125,
+      "step": 2612
+    },
+    {
+      "epoch": 186.64285714285714,
+      "grad_norm": 0.10693359375,
+      "learning_rate": 1.3702415068679722e-05,
+      "loss": 0.0125,
+      "step": 2613
+    },
+    {
+      "epoch": 186.71428571428572,
+      "grad_norm": 0.1513671875,
+      "learning_rate": 1.3693728598344454e-05,
+      "loss": 0.0155,
+      "step": 2614
+    },
+    {
+      "epoch": 186.78571428571428,
+      "grad_norm": 0.21875,
+      "learning_rate": 1.3685041546200367e-05,
+      "loss": 0.0125,
+      "step": 2615
+    },
+    {
+      "epoch": 186.85714285714286,
+      "grad_norm": 0.134765625,
+      "learning_rate": 1.3676353916481433e-05,
+      "loss": 0.0144,
+      "step": 2616
+    },
+    {
+      "epoch": 186.92857142857142,
+      "grad_norm": 0.11669921875,
+      "learning_rate": 1.3667665713421884e-05,
+      "loss": 0.0138,
+      "step": 2617
+    },
+    {
+      "epoch": 187.0,
+      "grad_norm": 0.1552734375,
+      "learning_rate": 1.3658976941256252e-05,
+      "loss": 0.0131,
+      "step": 2618
+    },
+    {
+      "epoch": 187.07142857142858,
+      "grad_norm": 0.26953125,
+      "learning_rate": 1.3650287604219342e-05,
+      "loss": 0.0108,
+      "step": 2619
+    },
+    {
+      "epoch": 187.14285714285714,
+      "grad_norm": 0.2119140625,
+      "learning_rate": 1.3641597706546223e-05,
+      "loss": 0.0118,
+      "step": 2620
+    },
+    {
+      "epoch": 187.21428571428572,
+      "grad_norm": 0.1259765625,
+      "learning_rate": 1.3632907252472255e-05,
+      "loss": 0.0114,
+      "step": 2621
+    },
+    {
+      "epoch": 187.28571428571428,
+      "grad_norm": 0.1171875,
+      "learning_rate": 1.3624216246233048e-05,
+      "loss": 0.0136,
+      "step": 2622
+    },
+    {
+      "epoch": 187.35714285714286,
+      "grad_norm": 0.12060546875,
+      "learning_rate": 1.3615524692064508e-05,
+      "loss": 0.0123,
+      "step": 2623
+    },
+    {
+      "epoch": 187.42857142857142,
+      "grad_norm": 0.177734375,
+      "learning_rate": 1.3606832594202778e-05,
+      "loss": 0.0132,
+      "step": 2624
+    },
+    {
+      "epoch": 187.5,
+      "grad_norm": 0.1181640625,
+      "learning_rate": 1.3598139956884292e-05,
+      "loss": 0.0107,
+      "step": 2625
+    },
+    {
+      "epoch": 187.57142857142858,
+      "grad_norm": 0.21484375,
+      "learning_rate": 1.358944678434573e-05,
+      "loss": 0.0134,
+      "step": 2626
+    },
+    {
+      "epoch": 187.64285714285714,
+      "grad_norm": 0.1875,
+      "learning_rate": 1.3580753080824038e-05,
+      "loss": 0.0139,
+      "step": 2627
+    },
+    {
+      "epoch": 187.71428571428572,
+      "grad_norm": 0.1474609375,
+      "learning_rate": 1.3572058850556423e-05,
+      "loss": 0.0123,
+      "step": 2628
+    },
+    {
+      "epoch": 187.78571428571428,
+      "grad_norm": 0.125,
+      "learning_rate": 1.3563364097780351e-05,
+      "loss": 0.0135,
+      "step": 2629
+    },
+    {
+      "epoch": 187.85714285714286,
+      "grad_norm": 0.1201171875,
+      "learning_rate": 1.3554668826733535e-05,
+      "loss": 0.0126,
+      "step": 2630
+    },
+    {
+      "epoch": 187.92857142857142,
+      "grad_norm": 0.1357421875,
+      "learning_rate": 1.3545973041653948e-05,
+      "loss": 0.0122,
+      "step": 2631
+    },
+    {
+      "epoch": 188.0,
+      "grad_norm": 0.197265625,
+      "learning_rate": 1.35372767467798e-05,
+      "loss": 0.0124,
+      "step": 2632
+    },
+    {
+      "epoch": 188.07142857142858,
+      "grad_norm": 0.126953125,
+      "learning_rate": 1.352857994634957e-05,
+      "loss": 0.0144,
+      "step": 2633
+    },
+    {
+      "epoch": 188.14285714285714,
+      "grad_norm": 0.130859375,
+      "learning_rate": 1.3519882644601973e-05,
+      "loss": 0.0144,
+      "step": 2634
+    },
+    {
+      "epoch": 188.21428571428572,
+      "grad_norm": 0.1396484375,
+      "learning_rate": 1.3511184845775959e-05,
+      "loss": 0.0149,
+      "step": 2635
+    },
+    {
+      "epoch": 188.28571428571428,
+      "grad_norm": 0.1923828125,
+      "learning_rate": 1.3502486554110736e-05,
+      "loss": 0.0129,
+      "step": 2636
+    },
+    {
+      "epoch": 188.35714285714286,
+      "grad_norm": 0.13671875,
+      "learning_rate": 1.3493787773845749e-05,
+      "loss": 0.0115,
+      "step": 2637
+    },
+    {
+      "epoch": 188.42857142857142,
+      "grad_norm": 0.12109375,
+      "learning_rate": 1.348508850922067e-05,
+      "loss": 0.0118,
+      "step": 2638
+    },
+    {
+      "epoch": 188.5,
+      "grad_norm": 0.140625,
+      "learning_rate": 1.347638876447542e-05,
+      "loss": 0.0152,
+      "step": 2639
+    },
+    {
+      "epoch": 188.57142857142858,
+      "grad_norm": 0.18359375,
+      "learning_rate": 1.3467688543850149e-05,
+      "loss": 0.0134,
+      "step": 2640
+    },
+    {
+      "epoch": 188.64285714285714,
+      "grad_norm": 0.10791015625,
+      "learning_rate": 1.3458987851585233e-05,
+      "loss": 0.0124,
+      "step": 2641
+    },
+    {
+      "epoch": 188.71428571428572,
+      "grad_norm": 0.1826171875,
+      "learning_rate": 1.3450286691921291e-05,
+      "loss": 0.0133,
+      "step": 2642
+    },
+    {
+      "epoch": 188.78571428571428,
+      "grad_norm": 0.16796875,
+      "learning_rate": 1.3441585069099161e-05,
+      "loss": 0.0115,
+      "step": 2643
+    },
+    {
+      "epoch": 188.85714285714286,
+      "grad_norm": 0.1064453125,
+      "learning_rate": 1.3432882987359905e-05,
+      "loss": 0.0132,
+      "step": 2644
+    },
+    {
+      "epoch": 188.92857142857142,
+      "grad_norm": 0.162109375,
+      "learning_rate": 1.3424180450944815e-05,
+      "loss": 0.0125,
+      "step": 2645
+    },
+    {
+      "epoch": 189.0,
+      "grad_norm": 0.134765625,
+      "learning_rate": 1.3415477464095397e-05,
+      "loss": 0.0166,
+      "step": 2646
+    },
+    {
+      "epoch": 189.07142857142858,
+      "grad_norm": 0.1220703125,
+      "learning_rate": 1.3406774031053387e-05,
+      "loss": 0.0133,
+      "step": 2647
+    },
+    {
+      "epoch": 189.14285714285714,
+      "grad_norm": 0.1884765625,
+      "learning_rate": 1.3398070156060729e-05,
+      "loss": 0.0158,
+      "step": 2648
+    },
+    {
+      "epoch": 189.21428571428572,
+      "grad_norm": 0.22265625,
+      "learning_rate": 1.3389365843359589e-05,
+      "loss": 0.015,
+      "step": 2649
+    },
+    {
+      "epoch": 189.28571428571428,
+      "grad_norm": 0.142578125,
+      "learning_rate": 1.3380661097192337e-05,
+      "loss": 0.0136,
+      "step": 2650
+    },
+    {
+      "epoch": 189.35714285714286,
+      "grad_norm": 0.1669921875,
+      "learning_rate": 1.3371955921801566e-05,
+      "loss": 0.0135,
+      "step": 2651
+    },
+    {
+      "epoch": 189.42857142857142,
+      "grad_norm": 0.1162109375,
+      "learning_rate": 1.3363250321430077e-05,
+      "loss": 0.0129,
+      "step": 2652
+    },
+    {
+      "epoch": 189.5,
+      "grad_norm": 0.1259765625,
+      "learning_rate": 1.3354544300320859e-05,
+      "loss": 0.0142,
+      "step": 2653
+    },
+    {
+      "epoch": 189.57142857142858,
+      "grad_norm": 0.1513671875,
+      "learning_rate": 1.3345837862717136e-05,
+      "loss": 0.0109,
+      "step": 2654
+    },
+    {
+      "epoch": 189.64285714285714,
+      "grad_norm": 0.169921875,
+      "learning_rate": 1.3337131012862313e-05,
+      "loss": 0.015,
+      "step": 2655
+    },
+    {
+      "epoch": 189.71428571428572,
+      "grad_norm": 0.107421875,
+      "learning_rate": 1.3328423755000006e-05,
+      "loss": 0.0137,
+      "step": 2656
+    },
+    {
+      "epoch": 189.78571428571428,
+      "grad_norm": 0.11767578125,
+      "learning_rate": 1.3319716093374021e-05,
+      "loss": 0.0124,
+      "step": 2657
+    },
+    {
+      "epoch": 189.85714285714286,
+      "grad_norm": 0.140625,
+      "learning_rate": 1.3311008032228369e-05,
+      "loss": 0.0113,
+      "step": 2658
+    },
+    {
+      "epoch": 189.92857142857142,
+      "grad_norm": 0.1279296875,
+      "learning_rate": 1.3302299575807257e-05,
+      "loss": 0.0119,
+      "step": 2659
+    },
+    {
+      "epoch": 190.0,
+      "grad_norm": 0.1259765625,
+      "learning_rate": 1.3293590728355077e-05,
+      "loss": 0.0129,
+      "step": 2660
+    },
+    {
+      "epoch": 190.07142857142858,
+      "grad_norm": 0.1171875,
+      "learning_rate": 1.328488149411642e-05,
+      "loss": 0.0124,
+      "step": 2661
+    },
+    {
+      "epoch": 190.14285714285714,
+      "grad_norm": 0.12451171875,
+      "learning_rate": 1.3276171877336051e-05,
+      "loss": 0.0131,
+      "step": 2662
+    },
+    {
+      "epoch": 190.21428571428572,
+      "grad_norm": 0.1474609375,
+      "learning_rate": 1.3267461882258945e-05,
+      "loss": 0.0143,
+      "step": 2663
+    },
+    {
+      "epoch": 190.28571428571428,
+      "grad_norm": 0.1328125,
+      "learning_rate": 1.3258751513130239e-05,
+      "loss": 0.0137,
+      "step": 2664
+    },
+    {
+      "epoch": 190.35714285714286,
+      "grad_norm": 0.11865234375,
+      "learning_rate": 1.325004077419526e-05,
+      "loss": 0.0121,
+      "step": 2665
+    },
+    {
+      "epoch": 190.42857142857142,
+      "grad_norm": 0.1513671875,
+      "learning_rate": 1.3241329669699523e-05,
+      "loss": 0.0136,
+      "step": 2666
+    },
+    {
+      "epoch": 190.5,
+      "grad_norm": 0.1015625,
+      "learning_rate": 1.3232618203888705e-05,
+      "loss": 0.0136,
+      "step": 2667
+    },
+    {
+      "epoch": 190.57142857142858,
+      "grad_norm": 0.1640625,
+      "learning_rate": 1.322390638100868e-05,
+      "loss": 0.0141,
+      "step": 2668
+    },
+    {
+      "epoch": 190.64285714285714,
+      "grad_norm": 0.201171875,
+      "learning_rate": 1.3215194205305478e-05,
+      "loss": 0.0147,
+      "step": 2669
+    },
+    {
+      "epoch": 190.71428571428572,
+      "grad_norm": 0.177734375,
+      "learning_rate": 1.3206481681025309e-05,
+      "loss": 0.0128,
+      "step": 2670
+    },
+    {
+      "epoch": 190.78571428571428,
+      "grad_norm": 0.134765625,
+      "learning_rate": 1.3197768812414548e-05,
+      "loss": 0.0127,
+      "step": 2671
+    },
+    {
+      "epoch": 190.85714285714286,
+      "grad_norm": 0.185546875,
+      "learning_rate": 1.3189055603719747e-05,
+      "loss": 0.0148,
+      "step": 2672
+    },
+    {
+      "epoch": 190.92857142857142,
+      "grad_norm": 0.2138671875,
+      "learning_rate": 1.3180342059187614e-05,
+      "loss": 0.0144,
+      "step": 2673
+    },
+    {
+      "epoch": 191.0,
+      "grad_norm": 0.09814453125,
+      "learning_rate": 1.3171628183065027e-05,
+      "loss": 0.0116,
+      "step": 2674
+    },
+    {
+      "epoch": 191.07142857142858,
+      "grad_norm": 0.19921875,
+      "learning_rate": 1.3162913979599022e-05,
+      "loss": 0.0135,
+      "step": 2675
+    },
+    {
+      "epoch": 191.14285714285714,
+      "grad_norm": 0.1640625,
+      "learning_rate": 1.31541994530368e-05,
+      "loss": 0.0132,
+      "step": 2676
+    },
+    {
+      "epoch": 191.21428571428572,
+      "grad_norm": 0.361328125,
+      "learning_rate": 1.3145484607625708e-05,
+      "loss": 0.0161,
+      "step": 2677
+    },
+    {
+      "epoch": 191.28571428571428,
+      "grad_norm": 0.30859375,
+      "learning_rate": 1.3136769447613261e-05,
+      "loss": 0.015,
+      "step": 2678
+    },
+    {
+      "epoch": 191.35714285714286,
+      "grad_norm": 0.11669921875,
+      "learning_rate": 1.3128053977247121e-05,
+      "loss": 0.0128,
+      "step": 2679
+    },
+    {
+      "epoch": 191.42857142857142,
+      "grad_norm": 0.248046875,
+      "learning_rate": 1.3119338200775103e-05,
+      "loss": 0.0152,
+      "step": 2680
+    },
+    {
+      "epoch": 191.5,
+      "grad_norm": 0.2412109375,
+      "learning_rate": 1.3110622122445168e-05,
+      "loss": 0.0133,
+      "step": 2681
+    },
+    {
+      "epoch": 191.57142857142858,
+      "grad_norm": 0.1201171875,
+      "learning_rate": 1.3101905746505428e-05,
+      "loss": 0.0121,
+      "step": 2682
+    },
+    {
+      "epoch": 191.64285714285714,
+      "grad_norm": 0.20703125,
+      "learning_rate": 1.3093189077204143e-05,
+      "loss": 0.0137,
+      "step": 2683
+    },
+    {
+      "epoch": 191.71428571428572,
+      "grad_norm": 0.2197265625,
+      "learning_rate": 1.3084472118789703e-05,
+      "loss": 0.0154,
+      "step": 2684
+    },
+    {
+      "epoch": 191.78571428571428,
+      "grad_norm": 0.169921875,
+      "learning_rate": 1.3075754875510654e-05,
+      "loss": 0.0135,
+      "step": 2685
+    },
+    {
+      "epoch": 191.85714285714286,
+      "grad_norm": 0.1376953125,
+      "learning_rate": 1.3067037351615672e-05,
+      "loss": 0.0139,
+      "step": 2686
+    },
+    {
+      "epoch": 191.92857142857142,
+      "grad_norm": 0.2431640625,
+      "learning_rate": 1.3058319551353578e-05,
+      "loss": 0.0124,
+      "step": 2687
+    },
+    {
+      "epoch": 192.0,
+      "grad_norm": 0.162109375,
+      "learning_rate": 1.3049601478973312e-05,
+      "loss": 0.0144,
+      "step": 2688
+    },
+    {
+      "epoch": 192.07142857142858,
+      "grad_norm": 0.1416015625,
+      "learning_rate": 1.3040883138723964e-05,
+      "loss": 0.0152,
+      "step": 2689
+    },
+    {
+      "epoch": 192.14285714285714,
+      "grad_norm": 0.1748046875,
+      "learning_rate": 1.3032164534854747e-05,
+      "loss": 0.0129,
+      "step": 2690
+    },
+    {
+      "epoch": 192.21428571428572,
+      "grad_norm": 0.181640625,
+      "learning_rate": 1.3023445671614995e-05,
+      "loss": 0.0135,
+      "step": 2691
+    },
+    {
+      "epoch": 192.28571428571428,
+      "grad_norm": 0.11572265625,
+      "learning_rate": 1.301472655325419e-05,
+      "loss": 0.0124,
+      "step": 2692
+    },
+    {
+      "epoch": 192.35714285714286,
+      "grad_norm": 0.16796875,
+      "learning_rate": 1.3006007184021907e-05,
+      "loss": 0.0124,
+      "step": 2693
+    },
+    {
+      "epoch": 192.42857142857142,
+      "grad_norm": 0.11328125,
+      "learning_rate": 1.2997287568167874e-05,
+      "loss": 0.0166,
+      "step": 2694
+    },
+    {
+      "epoch": 192.5,
+      "grad_norm": 0.14453125,
+      "learning_rate": 1.2988567709941921e-05,
+      "loss": 0.0127,
+      "step": 2695
+    },
+    {
+      "epoch": 192.57142857142858,
+      "grad_norm": 0.109375,
+      "learning_rate": 1.2979847613594002e-05,
+      "loss": 0.0118,
+      "step": 2696
+    },
+    {
+      "epoch": 192.64285714285714,
+      "grad_norm": 0.11572265625,
+      "learning_rate": 1.2971127283374185e-05,
+      "loss": 0.0132,
+      "step": 2697
+    },
+    {
+      "epoch": 192.71428571428572,
+      "grad_norm": 0.2119140625,
+      "learning_rate": 1.296240672353265e-05,
+      "loss": 0.016,
+      "step": 2698
+    },
+    {
+      "epoch": 192.78571428571428,
+      "grad_norm": 0.1865234375,
+      "learning_rate": 1.2953685938319699e-05,
+      "loss": 0.0134,
+      "step": 2699
+    },
+    {
+      "epoch": 192.85714285714286,
+      "grad_norm": 0.11376953125,
+      "learning_rate": 1.2944964931985731e-05,
+      "loss": 0.0147,
+      "step": 2700
+    },
+    {
+      "epoch": 192.92857142857142,
+      "grad_norm": 0.126953125,
+      "learning_rate": 1.2936243708781266e-05,
+      "loss": 0.013,
+      "step": 2701
+    },
+    {
+      "epoch": 193.0,
+      "grad_norm": 0.138671875,
+      "learning_rate": 1.2927522272956912e-05,
+      "loss": 0.0136,
+      "step": 2702
+    },
+    {
+      "epoch": 193.07142857142858,
+      "grad_norm": 0.173828125,
+      "learning_rate": 1.2918800628763394e-05,
+      "loss": 0.013,
+      "step": 2703
+    },
+    {
+      "epoch": 193.14285714285714,
+      "grad_norm": 0.140625,
+      "learning_rate": 1.291007878045154e-05,
+      "loss": 0.0154,
+      "step": 2704
+    },
+    {
+      "epoch": 193.21428571428572,
+      "grad_norm": 0.1220703125,
+      "learning_rate": 1.2901356732272271e-05,
+      "loss": 0.0116,
+      "step": 2705
+    },
+    {
+      "epoch": 193.28571428571428,
+      "grad_norm": 0.23828125,
+      "learning_rate": 1.2892634488476605e-05,
+      "loss": 0.0152,
+      "step": 2706
+    },
+    {
+      "epoch": 193.35714285714286,
+      "grad_norm": 0.10693359375,
+      "learning_rate": 1.288391205331566e-05,
+      "loss": 0.0136,
+      "step": 2707
+    },
+    {
+      "epoch": 193.42857142857142,
+      "grad_norm": 0.119140625,
+      "learning_rate": 1.287518943104064e-05,
+      "loss": 0.0153,
+      "step": 2708
+    },
+    {
+      "epoch": 193.5,
+      "grad_norm": 0.126953125,
+      "learning_rate": 1.2866466625902857e-05,
+      "loss": 0.0132,
+      "step": 2709
+    },
+    {
+      "epoch": 193.57142857142858,
+      "grad_norm": 0.111328125,
+      "learning_rate": 1.2857743642153685e-05,
+      "loss": 0.0126,
+      "step": 2710
+    },
+    {
+      "epoch": 193.64285714285714,
+      "grad_norm": 0.130859375,
+      "learning_rate": 1.2849020484044613e-05,
+      "loss": 0.0115,
+      "step": 2711
+    },
+    {
+      "epoch": 193.71428571428572,
+      "grad_norm": 0.16796875,
+      "learning_rate": 1.284029715582719e-05,
+      "loss": 0.0136,
+      "step": 2712
+    },
+    {
+      "epoch": 193.78571428571428,
+      "grad_norm": 0.169921875,
+      "learning_rate": 1.2831573661753074e-05,
+      "loss": 0.0149,
+      "step": 2713
+    },
+    {
+      "epoch": 193.85714285714286,
+      "grad_norm": 0.2392578125,
+      "learning_rate": 1.2822850006073986e-05,
+      "loss": 0.0138,
+      "step": 2714
+    },
+    {
+      "epoch": 193.92857142857142,
+      "grad_norm": 0.181640625,
+      "learning_rate": 1.281412619304172e-05,
+      "loss": 0.0133,
+      "step": 2715
+    },
+    {
+      "epoch": 194.0,
+      "grad_norm": 0.1259765625,
+      "learning_rate": 1.2805402226908166e-05,
+      "loss": 0.0111,
+      "step": 2716
+    },
+    {
+      "epoch": 194.07142857142858,
+      "grad_norm": 0.234375,
+      "learning_rate": 1.2796678111925276e-05,
+      "loss": 0.0153,
+      "step": 2717
+    },
+    {
+      "epoch": 194.14285714285714,
+      "grad_norm": 0.1142578125,
+      "learning_rate": 1.278795385234508e-05,
+      "loss": 0.0148,
+      "step": 2718
+    },
+    {
+      "epoch": 194.21428571428572,
+      "grad_norm": 0.30859375,
+      "learning_rate": 1.2779229452419673e-05,
+      "loss": 0.0162,
+      "step": 2719
+    },
+    {
+      "epoch": 194.28571428571428,
+      "grad_norm": 0.1533203125,
+      "learning_rate": 1.2770504916401224e-05,
+      "loss": 0.0145,
+      "step": 2720
+    },
+    {
+      "epoch": 194.35714285714286,
+      "grad_norm": 0.1298828125,
+      "learning_rate": 1.2761780248541963e-05,
+      "loss": 0.013,
+      "step": 2721
+    },
+    {
+      "epoch": 194.42857142857142,
+      "grad_norm": 0.208984375,
+      "learning_rate": 1.275305545309419e-05,
+      "loss": 0.0158,
+      "step": 2722
+    },
+    {
+      "epoch": 194.5,
+      "grad_norm": 0.1787109375,
+      "learning_rate": 1.2744330534310267e-05,
+      "loss": 0.013,
+      "step": 2723
+    },
+    {
+      "epoch": 194.57142857142858,
+      "grad_norm": 0.134765625,
+      "learning_rate": 1.2735605496442604e-05,
+      "loss": 0.0134,
+      "step": 2724
+    },
+    {
+      "epoch": 194.64285714285714,
+      "grad_norm": 0.265625,
+      "learning_rate": 1.2726880343743685e-05,
+      "loss": 0.016,
+      "step": 2725
+    },
+    {
+      "epoch": 194.71428571428572,
+      "grad_norm": 0.322265625,
+      "learning_rate": 1.2718155080466046e-05,
+      "loss": 0.014,
+      "step": 2726
+    },
+    {
+      "epoch": 194.78571428571428,
+      "grad_norm": 0.26953125,
+      "learning_rate": 1.2709429710862269e-05,
+      "loss": 0.013,
+      "step": 2727
+    },
+    {
+      "epoch": 194.85714285714286,
+      "grad_norm": 0.291015625,
+      "learning_rate": 1.2700704239184995e-05,
+      "loss": 0.0124,
+      "step": 2728
+    },
+    {
+      "epoch": 194.92857142857142,
+      "grad_norm": 0.140625,
+      "learning_rate": 1.2691978669686913e-05,
+      "loss": 0.0126,
+      "step": 2729
+    },
+    {
+      "epoch": 195.0,
+      "grad_norm": 0.1689453125,
+      "learning_rate": 1.268325300662076e-05,
+      "loss": 0.0131,
+      "step": 2730
+    },
+    {
+      "epoch": 195.07142857142858,
+      "grad_norm": 0.1171875,
+      "learning_rate": 1.2674527254239318e-05,
+      "loss": 0.0139,
+      "step": 2731
+    },
+    {
+      "epoch": 195.14285714285714,
+      "grad_norm": 0.2392578125,
+      "learning_rate": 1.2665801416795412e-05,
+      "loss": 0.013,
+      "step": 2732
+    },
+    {
+      "epoch": 195.21428571428572,
+      "grad_norm": 0.166015625,
+      "learning_rate": 1.265707549854191e-05,
+      "loss": 0.0136,
+      "step": 2733
+    },
+    {
+      "epoch": 195.28571428571428,
+      "grad_norm": 0.1962890625,
+      "learning_rate": 1.2648349503731716e-05,
+      "loss": 0.0143,
+      "step": 2734
+    },
+    {
+      "epoch": 195.35714285714286,
+      "grad_norm": 0.1318359375,
+      "learning_rate": 1.263962343661778e-05,
+      "loss": 0.0146,
+      "step": 2735
+    },
+    {
+      "epoch": 195.42857142857142,
+      "grad_norm": 0.1171875,
+      "learning_rate": 1.2630897301453073e-05,
+      "loss": 0.013,
+      "step": 2736
+    },
+    {
+      "epoch": 195.5,
+      "grad_norm": 0.171875,
+      "learning_rate": 1.2622171102490615e-05,
+      "loss": 0.0137,
+      "step": 2737
+    },
+    {
+      "epoch": 195.57142857142858,
+      "grad_norm": 0.13671875,
+      "learning_rate": 1.2613444843983441e-05,
+      "loss": 0.0103,
+      "step": 2738
+    },
+    {
+      "epoch": 195.64285714285714,
+      "grad_norm": 0.23828125,
+      "learning_rate": 1.2604718530184634e-05,
+      "loss": 0.0121,
+      "step": 2739
+    },
+    {
+      "epoch": 195.71428571428572,
+      "grad_norm": 0.28515625,
+      "learning_rate": 1.2595992165347287e-05,
+      "loss": 0.0146,
+      "step": 2740
+    },
+    {
+      "epoch": 195.78571428571428,
+      "grad_norm": 0.1591796875,
+      "learning_rate": 1.258726575372452e-05,
+      "loss": 0.0154,
+      "step": 2741
+    },
+    {
+      "epoch": 195.85714285714286,
+      "grad_norm": 0.150390625,
+      "learning_rate": 1.2578539299569488e-05,
+      "loss": 0.0131,
+      "step": 2742
+    },
+    {
+      "epoch": 195.92857142857142,
+      "grad_norm": 0.1259765625,
+      "learning_rate": 1.2569812807135353e-05,
+      "loss": 0.0128,
+      "step": 2743
+    },
+    {
+      "epoch": 196.0,
+      "grad_norm": 0.2578125,
+      "learning_rate": 1.2561086280675308e-05,
+      "loss": 0.0149,
+      "step": 2744
+    },
+    {
+      "epoch": 196.07142857142858,
+      "grad_norm": 0.1142578125,
+      "learning_rate": 1.255235972444255e-05,
+      "loss": 0.0135,
+      "step": 2745
+    },
+    {
+      "epoch": 196.14285714285714,
+      "grad_norm": 0.1328125,
+      "learning_rate": 1.2543633142690298e-05,
+      "loss": 0.0129,
+      "step": 2746
+    },
+    {
+      "epoch": 196.21428571428572,
+      "grad_norm": 0.134765625,
+      "learning_rate": 1.2534906539671785e-05,
+      "loss": 0.0133,
+      "step": 2747
+    },
+    {
+      "epoch": 196.28571428571428,
+      "grad_norm": 0.1455078125,
+      "learning_rate": 1.252617991964024e-05,
+      "loss": 0.0121,
+      "step": 2748
+    },
+    {
+      "epoch": 196.35714285714286,
+      "grad_norm": 0.126953125,
+      "learning_rate": 1.2517453286848932e-05,
+      "loss": 0.0134,
+      "step": 2749
+    },
+    {
+      "epoch": 196.42857142857142,
+      "grad_norm": 0.1455078125,
+      "learning_rate": 1.2508726645551094e-05,
+      "loss": 0.0133,
+      "step": 2750
+    },
+    {
+      "epoch": 196.5,
+      "grad_norm": 0.1171875,
+      "learning_rate": 1.25e-05,
+      "loss": 0.0136,
+      "step": 2751
+    },
+    {
+      "epoch": 196.57142857142858,
+      "grad_norm": 0.1689453125,
+      "learning_rate": 1.2491273354448907e-05,
+      "loss": 0.0156,
+      "step": 2752
+    },
+    {
+      "epoch": 196.64285714285714,
+      "grad_norm": 0.15625,
+      "learning_rate": 1.248254671315107e-05,
+      "loss": 0.0138,
+      "step": 2753
+    },
+    {
+      "epoch": 196.71428571428572,
+      "grad_norm": 0.1943359375,
+      "learning_rate": 1.2473820080359758e-05,
+      "loss": 0.0132,
+      "step": 2754
+    },
+    {
+      "epoch": 196.78571428571428,
+      "grad_norm": 0.193359375,
+      "learning_rate": 1.246509346032822e-05,
+      "loss": 0.0162,
+      "step": 2755
+    },
+    {
+      "epoch": 196.85714285714286,
+      "grad_norm": 0.1787109375,
+      "learning_rate": 1.2456366857309709e-05,
+      "loss": 0.0127,
+      "step": 2756
+    },
+    {
+      "epoch": 196.92857142857142,
+      "grad_norm": 0.1474609375,
+      "learning_rate": 1.2447640275557453e-05,
+      "loss": 0.0145,
+      "step": 2757
+    },
+    {
+      "epoch": 197.0,
+      "grad_norm": 0.1494140625,
+      "learning_rate": 1.2438913719324696e-05,
+      "loss": 0.0129,
+      "step": 2758
+    },
+    {
+      "epoch": 197.07142857142858,
+      "grad_norm": 0.1318359375,
+      "learning_rate": 1.243018719286465e-05,
+      "loss": 0.0124,
+      "step": 2759
+    },
+    {
+      "epoch": 197.14285714285714,
+      "grad_norm": 0.12451171875,
+      "learning_rate": 1.2421460700430514e-05,
+      "loss": 0.0129,
+      "step": 2760
+    },
+    {
+      "epoch": 197.21428571428572,
+      "grad_norm": 0.177734375,
+      "learning_rate": 1.2412734246275482e-05,
+      "loss": 0.0115,
+      "step": 2761
+    },
+    {
+      "epoch": 197.28571428571428,
+      "grad_norm": 0.2314453125,
+      "learning_rate": 1.2404007834652716e-05,
+      "loss": 0.0143,
+      "step": 2762
+    },
+    {
+      "epoch": 197.35714285714286,
+      "grad_norm": 0.16015625,
+      "learning_rate": 1.2395281469815367e-05,
+      "loss": 0.0142,
+      "step": 2763
+    },
+    {
+      "epoch": 197.42857142857142,
+      "grad_norm": 0.201171875,
+      "learning_rate": 1.2386555156016558e-05,
+      "loss": 0.0144,
+      "step": 2764
+    },
+    {
+      "epoch": 197.5,
+      "grad_norm": 0.1669921875,
+      "learning_rate": 1.237782889750939e-05,
+      "loss": 0.0134,
+      "step": 2765
+    },
+    {
+      "epoch": 197.57142857142858,
+      "grad_norm": 0.126953125,
+      "learning_rate": 1.2369102698546932e-05,
+      "loss": 0.0141,
+      "step": 2766
+    },
+    {
+      "epoch": 197.64285714285714,
+      "grad_norm": 0.224609375,
+      "learning_rate": 1.2360376563382224e-05,
+      "loss": 0.0139,
+      "step": 2767
+    },
+    {
+      "epoch": 197.71428571428572,
+      "grad_norm": 0.169921875,
+      "learning_rate": 1.2351650496268287e-05,
+      "loss": 0.0145,
+      "step": 2768
+    },
+    {
+      "epoch": 197.78571428571428,
+      "grad_norm": 0.208984375,
+      "learning_rate": 1.2342924501458091e-05,
+      "loss": 0.0152,
+      "step": 2769
+    },
+    {
+      "epoch": 197.85714285714286,
+      "grad_norm": 0.263671875,
+      "learning_rate": 1.233419858320459e-05,
+      "loss": 0.0139,
+      "step": 2770
+    },
+    {
+      "epoch": 197.92857142857142,
+      "grad_norm": 0.11669921875,
+      "learning_rate": 1.2325472745760686e-05,
+      "loss": 0.013,
+      "step": 2771
+    },
+    {
+      "epoch": 198.0,
+      "grad_norm": 0.12890625,
+      "learning_rate": 1.2316746993379242e-05,
+      "loss": 0.0121,
+      "step": 2772
+    },
+    {
+      "epoch": 198.07142857142858,
+      "grad_norm": 0.1826171875,
+      "learning_rate": 1.2308021330313089e-05,
+      "loss": 0.0109,
+      "step": 2773
+    },
+    {
+      "epoch": 198.14285714285714,
+      "grad_norm": 0.173828125,
+      "learning_rate": 1.229929576081501e-05,
+      "loss": 0.0117,
+      "step": 2774
+    },
+    {
+      "epoch": 198.21428571428572,
+      "grad_norm": 0.154296875,
+      "learning_rate": 1.2290570289137736e-05,
+      "loss": 0.0135,
+      "step": 2775
+    },
+    {
+      "epoch": 198.28571428571428,
+      "grad_norm": 0.12890625,
+      "learning_rate": 1.2281844919533957e-05,
+      "loss": 0.0119,
+      "step": 2776
+    },
+    {
+      "epoch": 198.35714285714286,
+      "grad_norm": 0.169921875,
+      "learning_rate": 1.2273119656256318e-05,
+      "loss": 0.0133,
+      "step": 2777
+    },
+    {
+      "epoch": 198.42857142857142,
+      "grad_norm": 0.140625,
+      "learning_rate": 1.22643945035574e-05,
+      "loss": 0.0158,
+      "step": 2778
+    },
+    {
+      "epoch": 198.5,
+      "grad_norm": 0.16796875,
+      "learning_rate": 1.2255669465689736e-05,
+      "loss": 0.0144,
+      "step": 2779
+    },
+    {
+      "epoch": 198.57142857142858,
+      "grad_norm": 0.23828125,
+      "learning_rate": 1.2246944546905811e-05,
+      "loss": 0.0157,
+      "step": 2780
+    },
+    {
+      "epoch": 198.64285714285714,
+      "grad_norm": 0.171875,
+      "learning_rate": 1.223821975145804e-05,
+      "loss": 0.0117,
+      "step": 2781
+    },
+    {
+      "epoch": 198.71428571428572,
+      "grad_norm": 0.16015625,
+      "learning_rate": 1.2229495083598777e-05,
+      "loss": 0.0128,
+      "step": 2782
+    },
+    {
+      "epoch": 198.78571428571428,
+      "grad_norm": 0.134765625,
+      "learning_rate": 1.222077054758033e-05,
+      "loss": 0.0111,
+      "step": 2783
+    },
+    {
+      "epoch": 198.85714285714286,
+      "grad_norm": 0.2265625,
+      "learning_rate": 1.2212046147654926e-05,
+      "loss": 0.0151,
+      "step": 2784
+    },
+    {
+      "epoch": 198.92857142857142,
+      "grad_norm": 0.1357421875,
+      "learning_rate": 1.2203321888074729e-05,
+      "loss": 0.0113,
+      "step": 2785
+    },
+    {
+      "epoch": 199.0,
+      "grad_norm": 0.1337890625,
+      "learning_rate": 1.2194597773091837e-05,
+      "loss": 0.0131,
+      "step": 2786
+    },
+    {
+      "epoch": 199.07142857142858,
+      "grad_norm": 0.0888671875,
+      "learning_rate": 1.2185873806958283e-05,
+      "loss": 0.013,
+      "step": 2787
+    },
+    {
+      "epoch": 199.14285714285714,
+      "grad_norm": 0.201171875,
+      "learning_rate": 1.2177149993926017e-05,
+      "loss": 0.0146,
+      "step": 2788
+    },
+    {
+      "epoch": 199.21428571428572,
+      "grad_norm": 0.1806640625,
+      "learning_rate": 1.2168426338246924e-05,
+      "loss": 0.0141,
+      "step": 2789
+    },
+    {
+      "epoch": 199.28571428571428,
+      "grad_norm": 0.2236328125,
+      "learning_rate": 1.2159702844172807e-05,
+      "loss": 0.0129,
+      "step": 2790
+    },
+    {
+      "epoch": 199.35714285714286,
+      "grad_norm": 0.1337890625,
+      "learning_rate": 1.2150979515955392e-05,
+      "loss": 0.0162,
+      "step": 2791
+    },
+    {
+      "epoch": 199.42857142857142,
+      "grad_norm": 0.11767578125,
+      "learning_rate": 1.2142256357846318e-05,
+      "loss": 0.0128,
+      "step": 2792
+    },
+    {
+      "epoch": 199.5,
+      "grad_norm": 0.154296875,
+      "learning_rate": 1.2133533374097148e-05,
+      "loss": 0.0128,
+      "step": 2793
+    },
+    {
+      "epoch": 199.57142857142858,
+      "grad_norm": 0.123046875,
+      "learning_rate": 1.2124810568959361e-05,
+      "loss": 0.0134,
+      "step": 2794
+    },
+    {
+      "epoch": 199.64285714285714,
+      "grad_norm": 0.1572265625,
+      "learning_rate": 1.211608794668434e-05,
+      "loss": 0.0124,
+      "step": 2795
+    },
+    {
+      "epoch": 199.71428571428572,
+      "grad_norm": 0.12451171875,
+      "learning_rate": 1.2107365511523396e-05,
+      "loss": 0.0126,
+      "step": 2796
+    },
+    {
+      "epoch": 199.78571428571428,
+      "grad_norm": 0.1787109375,
+      "learning_rate": 1.2098643267727732e-05,
+      "loss": 0.0139,
+      "step": 2797
+    },
+    {
+      "epoch": 199.85714285714286,
+      "grad_norm": 0.1982421875,
+      "learning_rate": 1.208992121954846e-05,
+      "loss": 0.0135,
+      "step": 2798
+    },
+    {
+      "epoch": 199.92857142857142,
+      "grad_norm": 0.11181640625,
+      "learning_rate": 1.2081199371236607e-05,
+      "loss": 0.0138,
+      "step": 2799
+    },
+    {
+      "epoch": 200.0,
+      "grad_norm": 0.16015625,
+      "learning_rate": 1.2072477727043094e-05,
+      "loss": 0.0135,
+      "step": 2800
+    },
+    {
+      "epoch": 200.07142857142858,
+      "grad_norm": 0.2158203125,
+      "learning_rate": 1.2063756291218742e-05,
+      "loss": 0.0121,
+      "step": 2801
+    },
+    {
+      "epoch": 200.14285714285714,
+      "grad_norm": 0.119140625,
+      "learning_rate": 1.205503506801427e-05,
+      "loss": 0.0138,
+      "step": 2802
+    },
+    {
+      "epoch": 200.21428571428572,
+      "grad_norm": 0.1533203125,
+      "learning_rate": 1.2046314061680304e-05,
+      "loss": 0.014,
+      "step": 2803
+    },
+    {
+      "epoch": 200.28571428571428,
+      "grad_norm": 0.1162109375,
+      "learning_rate": 1.2037593276467353e-05,
+      "loss": 0.0164,
+      "step": 2804
+    },
+    {
+      "epoch": 200.35714285714286,
+      "grad_norm": 0.1943359375,
+      "learning_rate": 1.2028872716625817e-05,
+      "loss": 0.0125,
+      "step": 2805
+    },
+    {
+      "epoch": 200.42857142857142,
+      "grad_norm": 0.1650390625,
+      "learning_rate": 1.2020152386406001e-05,
+      "loss": 0.0124,
+      "step": 2806
+    },
+    {
+      "epoch": 200.5,
+      "grad_norm": 0.1904296875,
+      "learning_rate": 1.2011432290058082e-05,
+      "loss": 0.0132,
+      "step": 2807
+    },
+    {
+      "epoch": 200.57142857142858,
+      "grad_norm": 0.1455078125,
+      "learning_rate": 1.2002712431832127e-05,
+      "loss": 0.0113,
+      "step": 2808
+    },
+    {
+      "epoch": 200.64285714285714,
+      "grad_norm": 0.10400390625,
+      "learning_rate": 1.1993992815978096e-05,
+      "loss": 0.0139,
+      "step": 2809
+    },
+    {
+      "epoch": 200.71428571428572,
+      "grad_norm": 0.138671875,
+      "learning_rate": 1.1985273446745818e-05,
+      "loss": 0.0159,
+      "step": 2810
+    },
+    {
+      "epoch": 200.78571428571428,
+      "grad_norm": 0.123046875,
+      "learning_rate": 1.1976554328385008e-05,
+      "loss": 0.013,
+      "step": 2811
+    },
+    {
+      "epoch": 200.85714285714286,
+      "grad_norm": 0.1396484375,
+      "learning_rate": 1.1967835465145256e-05,
+      "loss": 0.0147,
+      "step": 2812
+    },
+    {
+      "epoch": 200.92857142857142,
+      "grad_norm": 0.11669921875,
+      "learning_rate": 1.195911686127604e-05,
+      "loss": 0.0115,
+      "step": 2813
+    },
+    {
+      "epoch": 201.0,
+      "grad_norm": 0.1064453125,
+      "learning_rate": 1.1950398521026691e-05,
+      "loss": 0.0116,
+      "step": 2814
+    },
+    {
+      "epoch": 201.07142857142858,
+      "grad_norm": 0.134765625,
+      "learning_rate": 1.1941680448646425e-05,
+      "loss": 0.0131,
+      "step": 2815
+    },
+    {
+      "epoch": 201.14285714285714,
+      "grad_norm": 0.11572265625,
+      "learning_rate": 1.193296264838433e-05,
+      "loss": 0.0123,
+      "step": 2816
+    },
+    {
+      "epoch": 201.21428571428572,
+      "grad_norm": 0.35546875,
+      "learning_rate": 1.1924245124489345e-05,
+      "loss": 0.0148,
+      "step": 2817
+    },
+    {
+      "epoch": 201.28571428571428,
+      "grad_norm": 0.11474609375,
+      "learning_rate": 1.1915527881210303e-05,
+      "loss": 0.0127,
+      "step": 2818
+    },
+    {
+      "epoch": 201.35714285714286,
+      "grad_norm": 0.37890625,
+      "learning_rate": 1.1906810922795863e-05,
+      "loss": 0.0159,
+      "step": 2819
+    },
+    {
+      "epoch": 201.42857142857142,
+      "grad_norm": 0.2158203125,
+      "learning_rate": 1.1898094253494576e-05,
+      "loss": 0.0146,
+      "step": 2820
+    },
+    {
+      "epoch": 201.5,
+      "grad_norm": 0.2333984375,
+      "learning_rate": 1.1889377877554835e-05,
+      "loss": 0.0158,
+      "step": 2821
+    },
+    {
+      "epoch": 201.57142857142858,
+      "grad_norm": 0.330078125,
+      "learning_rate": 1.1880661799224901e-05,
+      "loss": 0.0142,
+      "step": 2822
+    },
+    {
+      "epoch": 201.64285714285714,
+      "grad_norm": 0.162109375,
+      "learning_rate": 1.1871946022752882e-05,
+      "loss": 0.0138,
+      "step": 2823
+    },
+    {
+      "epoch": 201.71428571428572,
+      "grad_norm": 0.146484375,
+      "learning_rate": 1.186323055238674e-05,
+      "loss": 0.0127,
+      "step": 2824
+    },
+    {
+      "epoch": 201.78571428571428,
+      "grad_norm": 0.12109375,
+      "learning_rate": 1.1854515392374293e-05,
+      "loss": 0.0128,
+      "step": 2825
+    },
+    {
+      "epoch": 201.85714285714286,
+      "grad_norm": 0.2275390625,
+      "learning_rate": 1.1845800546963202e-05,
+      "loss": 0.0135,
+      "step": 2826
+    },
+    {
+      "epoch": 201.92857142857142,
+      "grad_norm": 0.126953125,
+      "learning_rate": 1.183708602040098e-05,
+      "loss": 0.0129,
+      "step": 2827
+    },
+    {
+      "epoch": 202.0,
+      "grad_norm": 0.1103515625,
+      "learning_rate": 1.1828371816934974e-05,
+      "loss": 0.0131,
+      "step": 2828
+    },
+    {
+      "epoch": 202.07142857142858,
+      "grad_norm": 0.2734375,
+      "learning_rate": 1.1819657940812388e-05,
+      "loss": 0.0119,
+      "step": 2829
+    },
+    {
+      "epoch": 202.14285714285714,
+      "grad_norm": 0.1923828125,
+      "learning_rate": 1.1810944396280257e-05,
+      "loss": 0.0161,
+      "step": 2830
+    },
+    {
+      "epoch": 202.21428571428572,
+      "grad_norm": 0.26953125,
+      "learning_rate": 1.1802231187585453e-05,
+      "loss": 0.0124,
+      "step": 2831
+    },
+    {
+      "epoch": 202.28571428571428,
+      "grad_norm": 0.150390625,
+      "learning_rate": 1.1793518318974694e-05,
+      "loss": 0.0138,
+      "step": 2832
+    },
+    {
+      "epoch": 202.35714285714286,
+      "grad_norm": 0.12060546875,
+      "learning_rate": 1.1784805794694524e-05,
+      "loss": 0.0124,
+      "step": 2833
+    },
+    {
+      "epoch": 202.42857142857142,
+      "grad_norm": 0.1416015625,
+      "learning_rate": 1.177609361899132e-05,
+      "loss": 0.0184,
+      "step": 2834
+    },
+    {
+      "epoch": 202.5,
+      "grad_norm": 0.1298828125,
+      "learning_rate": 1.1767381796111293e-05,
+      "loss": 0.0158,
+      "step": 2835
+    },
+    {
+      "epoch": 202.57142857142858,
+      "grad_norm": 0.109375,
+      "learning_rate": 1.1758670330300482e-05,
+      "loss": 0.0102,
+      "step": 2836
+    },
+    {
+      "epoch": 202.64285714285714,
+      "grad_norm": 0.11328125,
+      "learning_rate": 1.1749959225804746e-05,
+      "loss": 0.0118,
+      "step": 2837
+    },
+    {
+      "epoch": 202.71428571428572,
+      "grad_norm": 0.109375,
+      "learning_rate": 1.1741248486869764e-05,
+      "loss": 0.013,
+      "step": 2838
+    },
+    {
+      "epoch": 202.78571428571428,
+      "grad_norm": 0.1337890625,
+      "learning_rate": 1.1732538117741058e-05,
+      "loss": 0.015,
+      "step": 2839
+    },
+    {
+      "epoch": 202.85714285714286,
+      "grad_norm": 0.193359375,
+      "learning_rate": 1.172382812266395e-05,
+      "loss": 0.0148,
+      "step": 2840
+    },
+    {
+      "epoch": 202.92857142857142,
+      "grad_norm": 0.10693359375,
+      "learning_rate": 1.1715118505883584e-05,
+      "loss": 0.0125,
+      "step": 2841
+    },
+    {
+      "epoch": 203.0,
+      "grad_norm": 0.11181640625,
+      "learning_rate": 1.1706409271644925e-05,
+      "loss": 0.0132,
+      "step": 2842
+    },
+    {
+      "epoch": 203.07142857142858,
+      "grad_norm": 0.130859375,
+      "learning_rate": 1.1697700424192743e-05,
+      "loss": 0.0131,
+      "step": 2843
+    },
+    {
+      "epoch": 203.14285714285714,
+      "grad_norm": 0.1171875,
+      "learning_rate": 1.1688991967771632e-05,
+      "loss": 0.0119,
+      "step": 2844
+    },
+    {
+      "epoch": 203.21428571428572,
+      "grad_norm": 0.1259765625,
+      "learning_rate": 1.1680283906625985e-05,
+      "loss": 0.0129,
+      "step": 2845
+    },
+    {
+      "epoch": 203.28571428571428,
+      "grad_norm": 0.1416015625,
+      "learning_rate": 1.1671576245000002e-05,
+      "loss": 0.0126,
+      "step": 2846
+    },
+    {
+      "epoch": 203.35714285714286,
+      "grad_norm": 0.1533203125,
+      "learning_rate": 1.166286898713769e-05,
+      "loss": 0.0141,
+      "step": 2847
+    },
+    {
+      "epoch": 203.42857142857142,
+      "grad_norm": 0.14453125,
+      "learning_rate": 1.1654162137282867e-05,
+      "loss": 0.0133,
+      "step": 2848
+    },
+    {
+      "epoch": 203.5,
+      "grad_norm": 0.17578125,
+      "learning_rate": 1.1645455699679143e-05,
+      "loss": 0.0125,
+      "step": 2849
+    },
+    {
+      "epoch": 203.57142857142858,
+      "grad_norm": 0.140625,
+      "learning_rate": 1.1636749678569928e-05,
+      "loss": 0.0145,
+      "step": 2850
+    },
+    {
+      "epoch": 203.64285714285714,
+      "grad_norm": 0.28515625,
+      "learning_rate": 1.1628044078198434e-05,
+      "loss": 0.0153,
+      "step": 2851
+    },
+    {
+      "epoch": 203.71428571428572,
+      "grad_norm": 0.1298828125,
+      "learning_rate": 1.1619338902807664e-05,
+      "loss": 0.0154,
+      "step": 2852
+    },
+    {
+      "epoch": 203.78571428571428,
+      "grad_norm": 0.1142578125,
+      "learning_rate": 1.1610634156640419e-05,
+      "loss": 0.0128,
+      "step": 2853
+    },
+    {
+      "epoch": 203.85714285714286,
+      "grad_norm": 0.138671875,
+      "learning_rate": 1.1601929843939274e-05,
+      "loss": 0.0135,
+      "step": 2854
+    },
+    {
+      "epoch": 203.92857142857142,
+      "grad_norm": 0.23046875,
+      "learning_rate": 1.1593225968946617e-05,
+      "loss": 0.0142,
+      "step": 2855
+    },
+    {
+      "epoch": 204.0,
+      "grad_norm": 0.12109375,
+      "learning_rate": 1.1584522535904607e-05,
+      "loss": 0.0153,
+      "step": 2856
+    },
+    {
+      "epoch": 204.07142857142858,
+      "grad_norm": 0.13671875,
+      "learning_rate": 1.1575819549055188e-05,
+      "loss": 0.0114,
+      "step": 2857
+    },
+    {
+      "epoch": 204.14285714285714,
+      "grad_norm": 0.1015625,
+      "learning_rate": 1.1567117012640098e-05,
+      "loss": 0.011,
+      "step": 2858
+    },
+    {
+      "epoch": 204.21428571428572,
+      "grad_norm": 0.11962890625,
+      "learning_rate": 1.1558414930900843e-05,
+      "loss": 0.013,
+      "step": 2859
+    },
+    {
+      "epoch": 204.28571428571428,
+      "grad_norm": 0.201171875,
+      "learning_rate": 1.154971330807871e-05,
+      "loss": 0.0153,
+      "step": 2860
+    },
+    {
+      "epoch": 204.35714285714286,
+      "grad_norm": 0.1435546875,
+      "learning_rate": 1.1541012148414768e-05,
+      "loss": 0.0135,
+      "step": 2861
+    },
+    {
+      "epoch": 204.42857142857142,
+      "grad_norm": 0.1474609375,
+      "learning_rate": 1.1532311456149856e-05,
+      "loss": 0.0139,
+      "step": 2862
+    },
+    {
+      "epoch": 204.5,
+      "grad_norm": 0.220703125,
+      "learning_rate": 1.1523611235524585e-05,
+      "loss": 0.0118,
+      "step": 2863
+    },
+    {
+      "epoch": 204.57142857142858,
+      "grad_norm": 0.134765625,
+      "learning_rate": 1.1514911490779333e-05,
+      "loss": 0.0158,
+      "step": 2864
+    },
+    {
+      "epoch": 204.64285714285714,
+      "grad_norm": 0.1318359375,
+      "learning_rate": 1.1506212226154254e-05,
+      "loss": 0.0135,
+      "step": 2865
+    },
+    {
+      "epoch": 204.71428571428572,
+      "grad_norm": 0.2177734375,
+      "learning_rate": 1.1497513445889265e-05,
+      "loss": 0.0158,
+      "step": 2866
+    },
+    {
+      "epoch": 204.78571428571428,
+      "grad_norm": 0.2138671875,
+      "learning_rate": 1.1488815154224042e-05,
+      "loss": 0.0135,
+      "step": 2867
+    },
+    {
+      "epoch": 204.85714285714286,
+      "grad_norm": 0.216796875,
+      "learning_rate": 1.1480117355398032e-05,
+      "loss": 0.0152,
+      "step": 2868
+    },
+    {
+      "epoch": 204.92857142857142,
+      "grad_norm": 0.248046875,
+      "learning_rate": 1.147142005365043e-05,
+      "loss": 0.0146,
+      "step": 2869
+    },
+    {
+      "epoch": 205.0,
+      "grad_norm": 0.1884765625,
+      "learning_rate": 1.14627232532202e-05,
+      "loss": 0.0123,
+      "step": 2870
+    },
+    {
+      "epoch": 205.07142857142858,
+      "grad_norm": 0.123046875,
+      "learning_rate": 1.1454026958346058e-05,
+      "loss": 0.0109,
+      "step": 2871
+    },
+    {
+      "epoch": 205.14285714285714,
+      "grad_norm": 0.1435546875,
+      "learning_rate": 1.1445331173266468e-05,
+      "loss": 0.0142,
+      "step": 2872
+    },
+    {
+      "epoch": 205.21428571428572,
+      "grad_norm": 0.1513671875,
+      "learning_rate": 1.143663590221965e-05,
+      "loss": 0.0139,
+      "step": 2873
+    },
+    {
+      "epoch": 205.28571428571428,
+      "grad_norm": 0.10595703125,
+      "learning_rate": 1.1427941149443578e-05,
+      "loss": 0.014,
+      "step": 2874
+    },
+    {
+      "epoch": 205.35714285714286,
+      "grad_norm": 0.1279296875,
+      "learning_rate": 1.1419246919175966e-05,
+      "loss": 0.012,
+      "step": 2875
+    },
+    {
+      "epoch": 205.42857142857142,
+      "grad_norm": 0.11328125,
+      "learning_rate": 1.1410553215654272e-05,
+      "loss": 0.0135,
+      "step": 2876
+    },
+    {
+      "epoch": 205.5,
+      "grad_norm": 0.125,
+      "learning_rate": 1.140186004311571e-05,
+      "loss": 0.0123,
+      "step": 2877
+    },
+    {
+      "epoch": 205.57142857142858,
+      "grad_norm": 0.1552734375,
+      "learning_rate": 1.1393167405797223e-05,
+      "loss": 0.0145,
+      "step": 2878
+    },
+    {
+      "epoch": 205.64285714285714,
+      "grad_norm": 0.2041015625,
+      "learning_rate": 1.1384475307935493e-05,
+      "loss": 0.0151,
+      "step": 2879
+    },
+    {
+      "epoch": 205.71428571428572,
+      "grad_norm": 0.10546875,
+      "learning_rate": 1.1375783753766953e-05,
+      "loss": 0.0112,
+      "step": 2880
+    },
+    {
+      "epoch": 205.78571428571428,
+      "grad_norm": 0.146484375,
+      "learning_rate": 1.1367092747527752e-05,
+      "loss": 0.0149,
+      "step": 2881
+    },
+    {
+      "epoch": 205.85714285714286,
+      "grad_norm": 0.1494140625,
+      "learning_rate": 1.135840229345378e-05,
+      "loss": 0.0115,
+      "step": 2882
+    },
+    {
+      "epoch": 205.92857142857142,
+      "grad_norm": 0.158203125,
+      "learning_rate": 1.1349712395780663e-05,
+      "loss": 0.016,
+      "step": 2883
+    },
+    {
+      "epoch": 206.0,
+      "grad_norm": 0.1484375,
+      "learning_rate": 1.1341023058743749e-05,
+      "loss": 0.0139,
+      "step": 2884
+    },
+    {
+      "epoch": 206.07142857142858,
+      "grad_norm": 0.2001953125,
+      "learning_rate": 1.133233428657812e-05,
+      "loss": 0.0141,
+      "step": 2885
+    },
+    {
+      "epoch": 206.14285714285714,
+      "grad_norm": 0.1396484375,
+      "learning_rate": 1.1323646083518572e-05,
+      "loss": 0.0132,
+      "step": 2886
+    },
+    {
+      "epoch": 206.21428571428572,
+      "grad_norm": 0.2080078125,
+      "learning_rate": 1.1314958453799634e-05,
+      "loss": 0.0148,
+      "step": 2887
+    },
+    {
+      "epoch": 206.28571428571428,
+      "grad_norm": 0.11328125,
+      "learning_rate": 1.1306271401655547e-05,
+      "loss": 0.0125,
+      "step": 2888
+    },
+    {
+      "epoch": 206.35714285714286,
+      "grad_norm": 0.2197265625,
+      "learning_rate": 1.1297584931320284e-05,
+      "loss": 0.0134,
+      "step": 2889
+    },
+    {
+      "epoch": 206.42857142857142,
+      "grad_norm": 0.126953125,
+      "learning_rate": 1.1288899047027512e-05,
+      "loss": 0.0148,
+      "step": 2890
+    },
+    {
+      "epoch": 206.5,
+      "grad_norm": 0.13671875,
+      "learning_rate": 1.1280213753010633e-05,
+      "loss": 0.0126,
+      "step": 2891
+    },
+    {
+      "epoch": 206.57142857142858,
+      "grad_norm": 0.224609375,
+      "learning_rate": 1.1271529053502755e-05,
+      "loss": 0.015,
+      "step": 2892
+    },
+    {
+      "epoch": 206.64285714285714,
+      "grad_norm": 0.31640625,
+      "learning_rate": 1.1262844952736688e-05,
+      "loss": 0.0101,
+      "step": 2893
+    },
+    {
+      "epoch": 206.71428571428572,
+      "grad_norm": 0.1953125,
+      "learning_rate": 1.1254161454944965e-05,
+      "loss": 0.0161,
+      "step": 2894
+    },
+    {
+      "epoch": 206.78571428571428,
+      "grad_norm": 0.1630859375,
+      "learning_rate": 1.1245478564359812e-05,
+      "loss": 0.0132,
+      "step": 2895
+    },
+    {
+      "epoch": 206.85714285714286,
+      "grad_norm": 0.234375,
+      "learning_rate": 1.123679628521317e-05,
+      "loss": 0.012,
+      "step": 2896
+    },
+    {
+      "epoch": 206.92857142857142,
+      "grad_norm": 0.169921875,
+      "learning_rate": 1.1228114621736673e-05,
+      "loss": 0.0115,
+      "step": 2897
+    },
+    {
+      "epoch": 207.0,
+      "grad_norm": 0.171875,
+      "learning_rate": 1.121943357816166e-05,
+      "loss": 0.0127,
+      "step": 2898
+    },
+    {
+      "epoch": 207.07142857142858,
+      "grad_norm": 0.1572265625,
+      "learning_rate": 1.1210753158719162e-05,
+      "loss": 0.012,
+      "step": 2899
+    },
+    {
+      "epoch": 207.14285714285714,
+      "grad_norm": 0.12060546875,
+      "learning_rate": 1.1202073367639911e-05,
+      "loss": 0.012,
+      "step": 2900
+    },
+    {
+      "epoch": 207.21428571428572,
+      "grad_norm": 0.15234375,
+      "learning_rate": 1.1193394209154334e-05,
+      "loss": 0.0141,
+      "step": 2901
+    },
+    {
+      "epoch": 207.28571428571428,
+      "grad_norm": 0.11572265625,
+      "learning_rate": 1.1184715687492544e-05,
+      "loss": 0.0146,
+      "step": 2902
+    },
+    {
+      "epoch": 207.35714285714286,
+      "grad_norm": 0.1416015625,
+      "learning_rate": 1.1176037806884354e-05,
+      "loss": 0.0148,
+      "step": 2903
+    },
+    {
+      "epoch": 207.42857142857142,
+      "grad_norm": 0.380859375,
+      "learning_rate": 1.1167360571559252e-05,
+      "loss": 0.0155,
+      "step": 2904
+    },
+    {
+      "epoch": 207.5,
+      "grad_norm": 0.1513671875,
+      "learning_rate": 1.1158683985746418e-05,
+      "loss": 0.0122,
+      "step": 2905
+    },
+    {
+      "epoch": 207.57142857142858,
+      "grad_norm": 0.169921875,
+      "learning_rate": 1.1150008053674715e-05,
+      "loss": 0.0146,
+      "step": 2906
+    },
+    {
+      "epoch": 207.64285714285714,
+      "grad_norm": 0.1806640625,
+      "learning_rate": 1.1141332779572689e-05,
+      "loss": 0.0131,
+      "step": 2907
+    },
+    {
+      "epoch": 207.71428571428572,
+      "grad_norm": 0.208984375,
+      "learning_rate": 1.113265816766856e-05,
+      "loss": 0.011,
+      "step": 2908
+    },
+    {
+      "epoch": 207.78571428571428,
+      "grad_norm": 0.185546875,
+      "learning_rate": 1.1123984222190225e-05,
+      "loss": 0.0152,
+      "step": 2909
+    },
+    {
+      "epoch": 207.85714285714286,
+      "grad_norm": 0.1513671875,
+      "learning_rate": 1.1115310947365268e-05,
+      "loss": 0.0135,
+      "step": 2910
+    },
+    {
+      "epoch": 207.92857142857142,
+      "grad_norm": 0.12890625,
+      "learning_rate": 1.1106638347420933e-05,
+      "loss": 0.0135,
+      "step": 2911
+    },
+    {
+      "epoch": 208.0,
+      "grad_norm": 0.1435546875,
+      "learning_rate": 1.1097966426584139e-05,
+      "loss": 0.0128,
+      "step": 2912
+    },
+    {
+      "epoch": 208.07142857142858,
+      "grad_norm": 0.1669921875,
+      "learning_rate": 1.108929518908148e-05,
+      "loss": 0.0171,
+      "step": 2913
+    },
+    {
+      "epoch": 208.14285714285714,
+      "grad_norm": 0.1240234375,
+      "learning_rate": 1.1080624639139203e-05,
+      "loss": 0.0129,
+      "step": 2914
+    },
+    {
+      "epoch": 208.21428571428572,
+      "grad_norm": 0.33203125,
+      "learning_rate": 1.1071954780983245e-05,
+      "loss": 0.0118,
+      "step": 2915
+    },
+    {
+      "epoch": 208.28571428571428,
+      "grad_norm": 0.158203125,
+      "learning_rate": 1.106328561883917e-05,
+      "loss": 0.0135,
+      "step": 2916
+    },
+    {
+      "epoch": 208.35714285714286,
+      "grad_norm": 0.17578125,
+      "learning_rate": 1.1054617156932233e-05,
+      "loss": 0.0149,
+      "step": 2917
+    },
+    {
+      "epoch": 208.42857142857142,
+      "grad_norm": 0.1279296875,
+      "learning_rate": 1.1045949399487334e-05,
+      "loss": 0.0147,
+      "step": 2918
+    },
+    {
+      "epoch": 208.5,
+      "grad_norm": 0.1640625,
+      "learning_rate": 1.1037282350729033e-05,
+      "loss": 0.0163,
+      "step": 2919
+    },
+    {
+      "epoch": 208.57142857142858,
+      "grad_norm": 0.1484375,
+      "learning_rate": 1.1028616014881546e-05,
+      "loss": 0.0145,
+      "step": 2920
+    },
+    {
+      "epoch": 208.64285714285714,
+      "grad_norm": 0.1474609375,
+      "learning_rate": 1.1019950396168737e-05,
+      "loss": 0.0129,
+      "step": 2921
+    },
+    {
+      "epoch": 208.71428571428572,
+      "grad_norm": 0.1669921875,
+      "learning_rate": 1.1011285498814128e-05,
+      "loss": 0.0102,
+      "step": 2922
+    },
+    {
+      "epoch": 208.78571428571428,
+      "grad_norm": 0.181640625,
+      "learning_rate": 1.100262132704088e-05,
+      "loss": 0.0152,
+      "step": 2923
+    },
+    {
+      "epoch": 208.85714285714286,
+      "grad_norm": 0.10693359375,
+      "learning_rate": 1.0993957885071812e-05,
+      "loss": 0.0126,
+      "step": 2924
+    },
+    {
+      "epoch": 208.92857142857142,
+      "grad_norm": 0.12158203125,
+      "learning_rate": 1.0985295177129373e-05,
+      "loss": 0.014,
+      "step": 2925
+    },
+    {
+      "epoch": 209.0,
+      "grad_norm": 0.1513671875,
+      "learning_rate": 1.097663320743566e-05,
+      "loss": 0.0122,
+      "step": 2926
+    },
+    {
+      "epoch": 209.07142857142858,
+      "grad_norm": 0.10888671875,
+      "learning_rate": 1.0967971980212417e-05,
+      "loss": 0.0109,
+      "step": 2927
+    },
+    {
+      "epoch": 209.14285714285714,
+      "grad_norm": 0.1455078125,
+      "learning_rate": 1.0959311499681018e-05,
+      "loss": 0.0133,
+      "step": 2928
+    },
+    {
+      "epoch": 209.21428571428572,
+      "grad_norm": 0.162109375,
+      "learning_rate": 1.0950651770062478e-05,
+      "loss": 0.0157,
+      "step": 2929
+    },
+    {
+      "epoch": 209.28571428571428,
+      "grad_norm": 0.134765625,
+      "learning_rate": 1.0941992795577447e-05,
+      "loss": 0.0117,
+      "step": 2930
+    },
+    {
+      "epoch": 209.35714285714286,
+      "grad_norm": 0.1591796875,
+      "learning_rate": 1.0933334580446195e-05,
+      "loss": 0.013,
+      "step": 2931
+    },
+    {
+      "epoch": 209.42857142857142,
+      "grad_norm": 0.1435546875,
+      "learning_rate": 1.092467712888864e-05,
+      "loss": 0.0138,
+      "step": 2932
+    },
+    {
+      "epoch": 209.5,
+      "grad_norm": 0.23046875,
+      "learning_rate": 1.091602044512432e-05,
+      "loss": 0.0094,
+      "step": 2933
+    },
+    {
+      "epoch": 209.57142857142858,
+      "grad_norm": 0.125,
+      "learning_rate": 1.0907364533372388e-05,
+      "loss": 0.0147,
+      "step": 2934
+    },
+    {
+      "epoch": 209.64285714285714,
+      "grad_norm": 0.181640625,
+      "learning_rate": 1.0898709397851634e-05,
+      "loss": 0.0141,
+      "step": 2935
+    },
+    {
+      "epoch": 209.71428571428572,
+      "grad_norm": 0.1611328125,
+      "learning_rate": 1.0890055042780467e-05,
+      "loss": 0.0112,
+      "step": 2936
+    },
+    {
+      "epoch": 209.78571428571428,
+      "grad_norm": 0.1201171875,
+      "learning_rate": 1.0881401472376915e-05,
+      "loss": 0.0137,
+      "step": 2937
+    },
+    {
+      "epoch": 209.85714285714286,
+      "grad_norm": 0.150390625,
+      "learning_rate": 1.087274869085862e-05,
+      "loss": 0.0141,
+      "step": 2938
+    },
+    {
+      "epoch": 209.92857142857142,
+      "grad_norm": 0.224609375,
+      "learning_rate": 1.0864096702442848e-05,
+      "loss": 0.0151,
+      "step": 2939
+    },
+    {
+      "epoch": 210.0,
+      "grad_norm": 0.111328125,
+      "learning_rate": 1.085544551134647e-05,
+      "loss": 0.0133,
+      "step": 2940
+    },
+    {
+      "epoch": 210.07142857142858,
+      "grad_norm": 0.1181640625,
+      "learning_rate": 1.0846795121785969e-05,
+      "loss": 0.0127,
+      "step": 2941
+    },
+    {
+      "epoch": 210.14285714285714,
+      "grad_norm": 0.11962890625,
+      "learning_rate": 1.0838145537977447e-05,
+      "loss": 0.0136,
+      "step": 2942
+    },
+    {
+      "epoch": 210.21428571428572,
+      "grad_norm": 0.1630859375,
+      "learning_rate": 1.0829496764136598e-05,
+      "loss": 0.0128,
+      "step": 2943
+    },
+    {
+      "epoch": 210.28571428571428,
+      "grad_norm": 0.1279296875,
+      "learning_rate": 1.0820848804478733e-05,
+      "loss": 0.0122,
+      "step": 2944
+    },
+    {
+      "epoch": 210.35714285714286,
+      "grad_norm": 0.171875,
+      "learning_rate": 1.0812201663218762e-05,
+      "loss": 0.0122,
+      "step": 2945
+    },
+    {
+      "epoch": 210.42857142857142,
+      "grad_norm": 0.119140625,
+      "learning_rate": 1.0803555344571197e-05,
+      "loss": 0.0131,
+      "step": 2946
+    },
+    {
+      "epoch": 210.5,
+      "grad_norm": 0.1923828125,
+      "learning_rate": 1.0794909852750145e-05,
+      "loss": 0.0141,
+      "step": 2947
+    },
+    {
+      "epoch": 210.57142857142858,
+      "grad_norm": 0.125,
+      "learning_rate": 1.0786265191969322e-05,
+      "loss": 0.0127,
+      "step": 2948
+    },
+    {
+      "epoch": 210.64285714285714,
+      "grad_norm": 0.251953125,
+      "learning_rate": 1.0777621366442026e-05,
+      "loss": 0.0144,
+      "step": 2949
+    },
+    {
+      "epoch": 210.71428571428572,
+      "grad_norm": 0.17578125,
+      "learning_rate": 1.0768978380381148e-05,
+      "loss": 0.0139,
+      "step": 2950
+    },
+    {
+      "epoch": 210.78571428571428,
+      "grad_norm": 0.1142578125,
+      "learning_rate": 1.0760336237999187e-05,
+      "loss": 0.0124,
+      "step": 2951
+    },
+    {
+      "epoch": 210.85714285714286,
+      "grad_norm": 0.126953125,
+      "learning_rate": 1.0751694943508202e-05,
+      "loss": 0.0122,
+      "step": 2952
+    },
+    {
+      "epoch": 210.92857142857142,
+      "grad_norm": 0.2236328125,
+      "learning_rate": 1.0743054501119865e-05,
+      "loss": 0.0154,
+      "step": 2953
+    },
+    {
+      "epoch": 211.0,
+      "grad_norm": 0.140625,
+      "learning_rate": 1.0734414915045419e-05,
+      "loss": 0.0156,
+      "step": 2954
+    },
+    {
+      "epoch": 211.07142857142858,
+      "grad_norm": 0.1513671875,
+      "learning_rate": 1.0725776189495696e-05,
+      "loss": 0.0133,
+      "step": 2955
+    },
+    {
+      "epoch": 211.14285714285714,
+      "grad_norm": 0.11962890625,
+      "learning_rate": 1.0717138328681106e-05,
+      "loss": 0.0116,
+      "step": 2956
+    },
+    {
+      "epoch": 211.21428571428572,
+      "grad_norm": 0.1376953125,
+      "learning_rate": 1.0708501336811632e-05,
+      "loss": 0.0123,
+      "step": 2957
+    },
+    {
+      "epoch": 211.28571428571428,
+      "grad_norm": 0.1640625,
+      "learning_rate": 1.0699865218096849e-05,
+      "loss": 0.0148,
+      "step": 2958
+    },
+    {
+      "epoch": 211.35714285714286,
+      "grad_norm": 0.1865234375,
+      "learning_rate": 1.0691229976745887e-05,
+      "loss": 0.0128,
+      "step": 2959
+    },
+    {
+      "epoch": 211.42857142857142,
+      "grad_norm": 0.2265625,
+      "learning_rate": 1.0682595616967465e-05,
+      "loss": 0.0136,
+      "step": 2960
+    },
+    {
+      "epoch": 211.5,
+      "grad_norm": 0.2001953125,
+      "learning_rate": 1.0673962142969857e-05,
+      "loss": 0.0157,
+      "step": 2961
+    },
+    {
+      "epoch": 211.57142857142858,
+      "grad_norm": 0.134765625,
+      "learning_rate": 1.0665329558960918e-05,
+      "loss": 0.0126,
+      "step": 2962
+    },
+    {
+      "epoch": 211.64285714285714,
+      "grad_norm": 0.2216796875,
+      "learning_rate": 1.0656697869148067e-05,
+      "loss": 0.0115,
+      "step": 2963
+    },
+    {
+      "epoch": 211.71428571428572,
+      "grad_norm": 0.1611328125,
+      "learning_rate": 1.0648067077738275e-05,
+      "loss": 0.0119,
+      "step": 2964
+    },
+    {
+      "epoch": 211.78571428571428,
+      "grad_norm": 0.1572265625,
+      "learning_rate": 1.0639437188938098e-05,
+      "loss": 0.0109,
+      "step": 2965
+    },
+    {
+      "epoch": 211.85714285714286,
+      "grad_norm": 0.220703125,
+      "learning_rate": 1.0630808206953632e-05,
+      "loss": 0.0118,
+      "step": 2966
+    },
+    {
+      "epoch": 211.92857142857142,
+      "grad_norm": 0.1240234375,
+      "learning_rate": 1.0622180135990537e-05,
+      "loss": 0.0129,
+      "step": 2967
+    },
+    {
+      "epoch": 212.0,
+      "grad_norm": 0.1435546875,
+      "learning_rate": 1.0613552980254032e-05,
+      "loss": 0.0134,
+      "step": 2968
+    },
+    {
+      "epoch": 212.07142857142858,
+      "grad_norm": 0.1796875,
+      "learning_rate": 1.0604926743948892e-05,
+      "loss": 0.0139,
+      "step": 2969
+    },
+    {
+      "epoch": 212.14285714285714,
+      "grad_norm": 0.154296875,
+      "learning_rate": 1.0596301431279432e-05,
+      "loss": 0.0124,
+      "step": 2970
+    },
+    {
+      "epoch": 212.21428571428572,
+      "grad_norm": 0.1123046875,
+      "learning_rate": 1.0587677046449526e-05,
+      "loss": 0.0121,
+      "step": 2971
+    },
+    {
+      "epoch": 212.28571428571428,
+      "grad_norm": 0.1416015625,
+      "learning_rate": 1.0579053593662599e-05,
+      "loss": 0.0138,
+      "step": 2972
+    },
+    {
+      "epoch": 212.35714285714286,
+      "grad_norm": 0.177734375,
+      "learning_rate": 1.057043107712161e-05,
+      "loss": 0.0112,
+      "step": 2973
+    },
+    {
+      "epoch": 212.42857142857142,
+      "grad_norm": 0.123046875,
+      "learning_rate": 1.0561809501029078e-05,
+      "loss": 0.0138,
+      "step": 2974
+    },
+    {
+      "epoch": 212.5,
+      "grad_norm": 0.1904296875,
+      "learning_rate": 1.0553188869587049e-05,
+      "loss": 0.0149,
+      "step": 2975
+    },
+    {
+      "epoch": 212.57142857142858,
+      "grad_norm": 0.12890625,
+      "learning_rate": 1.0544569186997112e-05,
+      "loss": 0.012,
+      "step": 2976
+    },
+    {
+      "epoch": 212.64285714285714,
+      "grad_norm": 0.259765625,
+      "learning_rate": 1.0535950457460408e-05,
+      "loss": 0.0143,
+      "step": 2977
+    },
+    {
+      "epoch": 212.71428571428572,
+      "grad_norm": 0.1298828125,
+      "learning_rate": 1.0527332685177583e-05,
+      "loss": 0.0145,
+      "step": 2978
+    },
+    {
+      "epoch": 212.78571428571428,
+      "grad_norm": 0.1396484375,
+      "learning_rate": 1.0518715874348844e-05,
+      "loss": 0.0142,
+      "step": 2979
+    },
+    {
+      "epoch": 212.85714285714286,
+      "grad_norm": 0.255859375,
+      "learning_rate": 1.0510100029173917e-05,
+      "loss": 0.0126,
+      "step": 2980
+    },
+    {
+      "epoch": 212.92857142857142,
+      "grad_norm": 0.1298828125,
+      "learning_rate": 1.0501485153852066e-05,
+      "loss": 0.0127,
+      "step": 2981
+    },
+    {
+      "epoch": 213.0,
+      "grad_norm": 0.15625,
+      "learning_rate": 1.0492871252582072e-05,
+      "loss": 0.013,
+      "step": 2982
+    },
+    {
+      "epoch": 213.07142857142858,
+      "grad_norm": 0.150390625,
+      "learning_rate": 1.0484258329562243e-05,
+      "loss": 0.0129,
+      "step": 2983
+    },
+    {
+      "epoch": 213.14285714285714,
+      "grad_norm": 0.1455078125,
+      "learning_rate": 1.047564638899042e-05,
+      "loss": 0.0143,
+      "step": 2984
+    },
+    {
+      "epoch": 213.21428571428572,
+      "grad_norm": 0.1611328125,
+      "learning_rate": 1.0467035435063956e-05,
+      "loss": 0.0114,
+      "step": 2985
+    },
+    {
+      "epoch": 213.28571428571428,
+      "grad_norm": 0.09326171875,
+      "learning_rate": 1.0458425471979726e-05,
+      "loss": 0.0115,
+      "step": 2986
+    },
+    {
+      "epoch": 213.35714285714286,
+      "grad_norm": 0.18359375,
+      "learning_rate": 1.0449816503934114e-05,
+      "loss": 0.0135,
+      "step": 2987
+    },
+    {
+      "epoch": 213.42857142857142,
+      "grad_norm": 0.1298828125,
+      "learning_rate": 1.0441208535123035e-05,
+      "loss": 0.0136,
+      "step": 2988
+    },
+    {
+      "epoch": 213.5,
+      "grad_norm": 0.12060546875,
+      "learning_rate": 1.0432601569741906e-05,
+      "loss": 0.013,
+      "step": 2989
+    },
+    {
+      "epoch": 213.57142857142858,
+      "grad_norm": 0.1328125,
+      "learning_rate": 1.0423995611985653e-05,
+      "loss": 0.0117,
+      "step": 2990
+    },
+    {
+      "epoch": 213.64285714285714,
+      "grad_norm": 0.16015625,
+      "learning_rate": 1.0415390666048723e-05,
+      "loss": 0.0114,
+      "step": 2991
+    },
+    {
+      "epoch": 213.71428571428572,
+      "grad_norm": 0.1435546875,
+      "learning_rate": 1.0406786736125057e-05,
+      "loss": 0.0127,
+      "step": 2992
+    },
+    {
+      "epoch": 213.78571428571428,
+      "grad_norm": 0.12451171875,
+      "learning_rate": 1.0398183826408107e-05,
+      "loss": 0.0139,
+      "step": 2993
+    },
+    {
+      "epoch": 213.85714285714286,
+      "grad_norm": 0.2060546875,
+      "learning_rate": 1.0389581941090827e-05,
+      "loss": 0.0121,
+      "step": 2994
+    },
+    {
+      "epoch": 213.92857142857142,
+      "grad_norm": 0.2265625,
+      "learning_rate": 1.0380981084365675e-05,
+      "loss": 0.0151,
+      "step": 2995
+    },
+    {
+      "epoch": 214.0,
+      "grad_norm": 0.107421875,
+      "learning_rate": 1.0372381260424597e-05,
+      "loss": 0.0121,
+      "step": 2996
+    },
+    {
+      "epoch": 214.07142857142858,
+      "grad_norm": 0.1982421875,
+      "learning_rate": 1.0363782473459044e-05,
+      "loss": 0.0137,
+      "step": 2997
+    },
+    {
+      "epoch": 214.14285714285714,
+      "grad_norm": 0.1396484375,
+      "learning_rate": 1.0355184727659964e-05,
+      "loss": 0.014,
+      "step": 2998
+    },
+    {
+      "epoch": 214.21428571428572,
+      "grad_norm": 0.2119140625,
+      "learning_rate": 1.0346588027217793e-05,
+      "loss": 0.0123,
+      "step": 2999
+    },
+    {
+      "epoch": 214.28571428571428,
+      "grad_norm": 0.169921875,
+      "learning_rate": 1.0337992376322456e-05,
+      "loss": 0.012,
+      "step": 3000
+    },
+    {
+      "epoch": 214.35714285714286,
+      "grad_norm": 0.224609375,
+      "learning_rate": 1.0329397779163372e-05,
+      "loss": 0.0143,
+      "step": 3001
+    },
+    {
+      "epoch": 214.42857142857142,
+      "grad_norm": 0.119140625,
+      "learning_rate": 1.0320804239929438e-05,
+      "loss": 0.0137,
+      "step": 3002
+    },
+    {
+      "epoch": 214.5,
+      "grad_norm": 0.1259765625,
+      "learning_rate": 1.0312211762809049e-05,
+      "loss": 0.0136,
+      "step": 3003
+    },
+    {
+      "epoch": 214.57142857142858,
+      "grad_norm": 0.255859375,
+      "learning_rate": 1.0303620351990071e-05,
+      "loss": 0.0144,
+      "step": 3004
+    },
+    {
+      "epoch": 214.64285714285714,
+      "grad_norm": 0.14453125,
+      "learning_rate": 1.0295030011659848e-05,
+      "loss": 0.0151,
+      "step": 3005
+    },
+    {
+      "epoch": 214.71428571428572,
+      "grad_norm": 0.11962890625,
+      "learning_rate": 1.028644074600521e-05,
+      "loss": 0.0144,
+      "step": 3006
+    },
+    {
+      "epoch": 214.78571428571428,
+      "grad_norm": 0.189453125,
+      "learning_rate": 1.0277852559212464e-05,
+      "loss": 0.0127,
+      "step": 3007
+    },
+    {
+      "epoch": 214.85714285714286,
+      "grad_norm": 0.1298828125,
+      "learning_rate": 1.0269265455467384e-05,
+      "loss": 0.0128,
+      "step": 3008
+    },
+    {
+      "epoch": 214.92857142857142,
+      "grad_norm": 0.1357421875,
+      "learning_rate": 1.0260679438955222e-05,
+      "loss": 0.013,
+      "step": 3009
+    },
+    {
+      "epoch": 215.0,
+      "grad_norm": 0.294921875,
+      "learning_rate": 1.0252094513860698e-05,
+      "loss": 0.0158,
+      "step": 3010
+    },
+    {
+      "epoch": 215.07142857142858,
+      "grad_norm": 0.19140625,
+      "learning_rate": 1.0243510684368002e-05,
+      "loss": 0.0114,
+      "step": 3011
+    },
+    {
+      "epoch": 215.14285714285714,
+      "grad_norm": 0.111328125,
+      "learning_rate": 1.0234927954660782e-05,
+      "loss": 0.0112,
+      "step": 3012
+    },
+    {
+      "epoch": 215.21428571428572,
+      "grad_norm": 0.216796875,
+      "learning_rate": 1.0226346328922167e-05,
+      "loss": 0.0186,
+      "step": 3013
+    },
+    {
+      "epoch": 215.28571428571428,
+      "grad_norm": 0.140625,
+      "learning_rate": 1.0217765811334727e-05,
+      "loss": 0.0126,
+      "step": 3014
+    },
+    {
+      "epoch": 215.35714285714286,
+      "grad_norm": 0.248046875,
+      "learning_rate": 1.0209186406080506e-05,
+      "loss": 0.0151,
+      "step": 3015
+    },
+    {
+      "epoch": 215.42857142857142,
+      "grad_norm": 0.11962890625,
+      "learning_rate": 1.0200608117340998e-05,
+      "loss": 0.0131,
+      "step": 3016
+    },
+    {
+      "epoch": 215.5,
+      "grad_norm": 0.150390625,
+      "learning_rate": 1.0192030949297165e-05,
+      "loss": 0.0135,
+      "step": 3017
+    },
+    {
+      "epoch": 215.57142857142858,
+      "grad_norm": 0.1259765625,
+      "learning_rate": 1.0183454906129409e-05,
+      "loss": 0.0132,
+      "step": 3018
+    },
+    {
+      "epoch": 215.64285714285714,
+      "grad_norm": 0.1328125,
+      "learning_rate": 1.0174879992017585e-05,
+      "loss": 0.0139,
+      "step": 3019
+    },
+    {
+      "epoch": 215.71428571428572,
+      "grad_norm": 0.205078125,
+      "learning_rate": 1.0166306211141013e-05,
+      "loss": 0.0106,
+      "step": 3020
+    },
+    {
+      "epoch": 215.78571428571428,
+      "grad_norm": 0.1669921875,
+      "learning_rate": 1.015773356767844e-05,
+      "loss": 0.012,
+      "step": 3021
+    },
+    {
+      "epoch": 215.85714285714286,
+      "grad_norm": 0.11376953125,
+      "learning_rate": 1.0149162065808079e-05,
+      "loss": 0.0133,
+      "step": 3022
+    },
+    {
+      "epoch": 215.92857142857142,
+      "grad_norm": 0.1494140625,
+      "learning_rate": 1.0140591709707562e-05,
+      "loss": 0.0132,
+      "step": 3023
+    },
+    {
+      "epoch": 216.0,
+      "grad_norm": 0.1328125,
+      "learning_rate": 1.0132022503553986e-05,
+      "loss": 0.013,
+      "step": 3024
+    },
+    {
+      "epoch": 216.07142857142858,
+      "grad_norm": 0.1923828125,
+      "learning_rate": 1.0123454451523875e-05,
+      "loss": 0.0133,
+      "step": 3025
+    },
+    {
+      "epoch": 216.14285714285714,
+      "grad_norm": 0.27734375,
+      "learning_rate": 1.011488755779319e-05,
+      "loss": 0.0106,
+      "step": 3026
+    },
+    {
+      "epoch": 216.21428571428572,
+      "grad_norm": 0.12255859375,
+      "learning_rate": 1.010632182653734e-05,
+      "loss": 0.0121,
+      "step": 3027
+    },
+    {
+      "epoch": 216.28571428571428,
+      "grad_norm": 0.19140625,
+      "learning_rate": 1.0097757261931147e-05,
+      "loss": 0.0115,
+      "step": 3028
+    },
+    {
+      "epoch": 216.35714285714286,
+      "grad_norm": 0.125,
+      "learning_rate": 1.0089193868148886e-05,
+      "loss": 0.0136,
+      "step": 3029
+    },
+    {
+      "epoch": 216.42857142857142,
+      "grad_norm": 0.11865234375,
+      "learning_rate": 1.0080631649364247e-05,
+      "loss": 0.0105,
+      "step": 3030
+    },
+    {
+      "epoch": 216.5,
+      "grad_norm": 0.123046875,
+      "learning_rate": 1.0072070609750353e-05,
+      "loss": 0.0123,
+      "step": 3031
+    },
+    {
+      "epoch": 216.57142857142858,
+      "grad_norm": 0.1337890625,
+      "learning_rate": 1.0063510753479744e-05,
+      "loss": 0.0104,
+      "step": 3032
+    },
+    {
+      "epoch": 216.64285714285714,
+      "grad_norm": 0.1611328125,
+      "learning_rate": 1.00549520847244e-05,
+      "loss": 0.0155,
+      "step": 3033
+    },
+    {
+      "epoch": 216.71428571428572,
+      "grad_norm": 0.185546875,
+      "learning_rate": 1.0046394607655705e-05,
+      "loss": 0.0131,
+      "step": 3034
+    },
+    {
+      "epoch": 216.78571428571428,
+      "grad_norm": 0.154296875,
+      "learning_rate": 1.0037838326444468e-05,
+      "loss": 0.0133,
+      "step": 3035
+    },
+    {
+      "epoch": 216.85714285714286,
+      "grad_norm": 0.1357421875,
+      "learning_rate": 1.0029283245260923e-05,
+      "loss": 0.0129,
+      "step": 3036
+    },
+    {
+      "epoch": 216.92857142857142,
+      "grad_norm": 0.1279296875,
+      "learning_rate": 1.002072936827471e-05,
+      "loss": 0.0109,
+      "step": 3037
+    },
+    {
+      "epoch": 217.0,
+      "grad_norm": 0.1337890625,
+      "learning_rate": 1.0012176699654883e-05,
+      "loss": 0.0131,
+      "step": 3038
+    },
+    {
+      "epoch": 217.07142857142858,
+      "grad_norm": 0.2080078125,
+      "learning_rate": 1.0003625243569915e-05,
+      "loss": 0.016,
+      "step": 3039
+    },
+    {
+      "epoch": 217.14285714285714,
+      "grad_norm": 0.1396484375,
+      "learning_rate": 9.995075004187673e-06,
+      "loss": 0.0116,
+      "step": 3040
+    },
+    {
+      "epoch": 217.21428571428572,
+      "grad_norm": 0.1591796875,
+      "learning_rate": 9.986525985675444e-06,
+      "loss": 0.0143,
+      "step": 3041
+    },
+    {
+      "epoch": 217.28571428571428,
+      "grad_norm": 0.11474609375,
+      "learning_rate": 9.977978192199914e-06,
+      "loss": 0.0107,
+      "step": 3042
+    },
+    {
+      "epoch": 217.35714285714286,
+      "grad_norm": 0.1708984375,
+      "learning_rate": 9.969431627927179e-06,
+      "loss": 0.0137,
+      "step": 3043
+    },
+    {
+      "epoch": 217.42857142857142,
+      "grad_norm": 0.1328125,
+      "learning_rate": 9.960886297022725e-06,
+      "loss": 0.013,
+      "step": 3044
+    },
+    {
+      "epoch": 217.5,
+      "grad_norm": 0.298828125,
+      "learning_rate": 9.952342203651441e-06,
+      "loss": 0.013,
+      "step": 3045
+    },
+    {
+      "epoch": 217.57142857142858,
+      "grad_norm": 0.1953125,
+      "learning_rate": 9.943799351977621e-06,
+      "loss": 0.0156,
+      "step": 3046
+    },
+    {
+      "epoch": 217.64285714285714,
+      "grad_norm": 0.12353515625,
+      "learning_rate": 9.935257746164942e-06,
+      "loss": 0.0136,
+      "step": 3047
+    },
+    {
+      "epoch": 217.71428571428572,
+      "grad_norm": 0.15234375,
+      "learning_rate": 9.926717390376487e-06,
+      "loss": 0.0114,
+      "step": 3048
+    },
+    {
+      "epoch": 217.78571428571428,
+      "grad_norm": 0.1357421875,
+      "learning_rate": 9.91817828877471e-06,
+      "loss": 0.0116,
+      "step": 3049
+    },
+    {
+      "epoch": 217.85714285714286,
+      "grad_norm": 0.123046875,
+      "learning_rate": 9.90964044552147e-06,
+      "loss": 0.014,
+      "step": 3050
+    },
+    {
+      "epoch": 217.92857142857142,
+      "grad_norm": 0.10302734375,
+      "learning_rate": 9.90110386477801e-06,
+      "loss": 0.0126,
+      "step": 3051
+    },
+    {
+      "epoch": 218.0,
+      "grad_norm": 0.21484375,
+      "learning_rate": 9.892568550704954e-06,
+      "loss": 0.0145,
+      "step": 3052
+    },
+    {
+      "epoch": 218.07142857142858,
+      "grad_norm": 0.138671875,
+      "learning_rate": 9.884034507462311e-06,
+      "loss": 0.014,
+      "step": 3053
+    },
+    {
+      "epoch": 218.14285714285714,
+      "grad_norm": 0.11865234375,
+      "learning_rate": 9.87550173920947e-06,
+      "loss": 0.0138,
+      "step": 3054
+    },
+    {
+      "epoch": 218.21428571428572,
+      "grad_norm": 0.119140625,
+      "learning_rate": 9.866970250105199e-06,
+      "loss": 0.0151,
+      "step": 3055
+    },
+    {
+      "epoch": 218.28571428571428,
+      "grad_norm": 0.208984375,
+      "learning_rate": 9.858440044307643e-06,
+      "loss": 0.0115,
+      "step": 3056
+    },
+    {
+      "epoch": 218.35714285714286,
+      "grad_norm": 0.11181640625,
+      "learning_rate": 9.84991112597432e-06,
+      "loss": 0.0114,
+      "step": 3057
+    },
+    {
+      "epoch": 218.42857142857142,
+      "grad_norm": 0.1376953125,
+      "learning_rate": 9.84138349926212e-06,
+      "loss": 0.0135,
+      "step": 3058
+    },
+    {
+      "epoch": 218.5,
+      "grad_norm": 0.1708984375,
+      "learning_rate": 9.832857168327301e-06,
+      "loss": 0.0128,
+      "step": 3059
+    },
+    {
+      "epoch": 218.57142857142858,
+      "grad_norm": 0.2275390625,
+      "learning_rate": 9.8243321373255e-06,
+      "loss": 0.0119,
+      "step": 3060
+    },
+    {
+      "epoch": 218.64285714285714,
+      "grad_norm": 0.1396484375,
+      "learning_rate": 9.81580841041171e-06,
+      "loss": 0.0126,
+      "step": 3061
+    },
+    {
+      "epoch": 218.71428571428572,
+      "grad_norm": 0.15234375,
+      "learning_rate": 9.807285991740296e-06,
+      "loss": 0.0113,
+      "step": 3062
+    },
+    {
+      "epoch": 218.78571428571428,
+      "grad_norm": 0.1484375,
+      "learning_rate": 9.798764885464975e-06,
+      "loss": 0.0131,
+      "step": 3063
+    },
+    {
+      "epoch": 218.85714285714286,
+      "grad_norm": 0.138671875,
+      "learning_rate": 9.790245095738834e-06,
+      "loss": 0.0131,
+      "step": 3064
+    },
+    {
+      "epoch": 218.92857142857142,
+      "grad_norm": 0.142578125,
+      "learning_rate": 9.781726626714315e-06,
+      "loss": 0.0144,
+      "step": 3065
+    },
+    {
+      "epoch": 219.0,
+      "grad_norm": 0.193359375,
+      "learning_rate": 9.773209482543221e-06,
+      "loss": 0.0129,
+      "step": 3066
+    },
+    {
+      "epoch": 219.07142857142858,
+      "grad_norm": 0.1474609375,
+      "learning_rate": 9.764693667376697e-06,
+      "loss": 0.0128,
+      "step": 3067
+    },
+    {
+      "epoch": 219.14285714285714,
+      "grad_norm": 0.181640625,
+      "learning_rate": 9.756179185365247e-06,
+      "loss": 0.0126,
+      "step": 3068
+    },
+    {
+      "epoch": 219.21428571428572,
+      "grad_norm": 0.09423828125,
+      "learning_rate": 9.747666040658732e-06,
+      "loss": 0.0122,
+      "step": 3069
+    },
+    {
+      "epoch": 219.28571428571428,
+      "grad_norm": 0.130859375,
+      "learning_rate": 9.739154237406355e-06,
+      "loss": 0.0132,
+      "step": 3070
+    },
+    {
+      "epoch": 219.35714285714286,
+      "grad_norm": 0.1513671875,
+      "learning_rate": 9.73064377975666e-06,
+      "loss": 0.0148,
+      "step": 3071
+    },
+    {
+      "epoch": 219.42857142857142,
+      "grad_norm": 0.1328125,
+      "learning_rate": 9.722134671857546e-06,
+      "loss": 0.0133,
+      "step": 3072
+    },
+    {
+      "epoch": 219.5,
+      "grad_norm": 0.1298828125,
+      "learning_rate": 9.713626917856244e-06,
+      "loss": 0.0147,
+      "step": 3073
+    },
+    {
+      "epoch": 219.57142857142858,
+      "grad_norm": 0.154296875,
+      "learning_rate": 9.705120521899338e-06,
+      "loss": 0.0132,
+      "step": 3074
+    },
+    {
+      "epoch": 219.64285714285714,
+      "grad_norm": 0.216796875,
+      "learning_rate": 9.696615488132737e-06,
+      "loss": 0.0137,
+      "step": 3075
+    },
+    {
+      "epoch": 219.71428571428572,
+      "grad_norm": 0.341796875,
+      "learning_rate": 9.68811182070169e-06,
+      "loss": 0.0176,
+      "step": 3076
+    },
+    {
+      "epoch": 219.78571428571428,
+      "grad_norm": 0.248046875,
+      "learning_rate": 9.679609523750782e-06,
+      "loss": 0.0127,
+      "step": 3077
+    },
+    {
+      "epoch": 219.85714285714286,
+      "grad_norm": 0.2080078125,
+      "learning_rate": 9.671108601423928e-06,
+      "loss": 0.0145,
+      "step": 3078
+    },
+    {
+      "epoch": 219.92857142857142,
+      "grad_norm": 0.126953125,
+      "learning_rate": 9.662609057864377e-06,
+      "loss": 0.0129,
+      "step": 3079
+    },
+    {
+      "epoch": 220.0,
+      "grad_norm": 0.1669921875,
+      "learning_rate": 9.654110897214701e-06,
+      "loss": 0.0139,
+      "step": 3080
+    },
+    {
+      "epoch": 220.07142857142858,
+      "grad_norm": 0.140625,
+      "learning_rate": 9.645614123616803e-06,
+      "loss": 0.0137,
+      "step": 3081
+    },
+    {
+      "epoch": 220.14285714285714,
+      "grad_norm": 0.1328125,
+      "learning_rate": 9.637118741211907e-06,
+      "loss": 0.0144,
+      "step": 3082
+    },
+    {
+      "epoch": 220.21428571428572,
+      "grad_norm": 0.11669921875,
+      "learning_rate": 9.628624754140554e-06,
+      "loss": 0.0105,
+      "step": 3083
+    },
+    {
+      "epoch": 220.28571428571428,
+      "grad_norm": 0.2080078125,
+      "learning_rate": 9.620132166542622e-06,
+      "loss": 0.0137,
+      "step": 3084
+    },
+    {
+      "epoch": 220.35714285714286,
+      "grad_norm": 0.12255859375,
+      "learning_rate": 9.61164098255728e-06,
+      "loss": 0.0123,
+      "step": 3085
+    },
+    {
+      "epoch": 220.42857142857142,
+      "grad_norm": 0.263671875,
+      "learning_rate": 9.603151206323036e-06,
+      "loss": 0.012,
+      "step": 3086
+    },
+    {
+      "epoch": 220.5,
+      "grad_norm": 0.1484375,
+      "learning_rate": 9.5946628419777e-06,
+      "loss": 0.012,
+      "step": 3087
+    },
+    {
+      "epoch": 220.57142857142858,
+      "grad_norm": 0.13671875,
+      "learning_rate": 9.586175893658406e-06,
+      "loss": 0.011,
+      "step": 3088
+    },
+    {
+      "epoch": 220.64285714285714,
+      "grad_norm": 0.1396484375,
+      "learning_rate": 9.57769036550158e-06,
+      "loss": 0.0136,
+      "step": 3089
+    },
+    {
+      "epoch": 220.71428571428572,
+      "grad_norm": 0.1630859375,
+      "learning_rate": 9.569206261642968e-06,
+      "loss": 0.0123,
+      "step": 3090
+    },
+    {
+      "epoch": 220.78571428571428,
+      "grad_norm": 0.197265625,
+      "learning_rate": 9.560723586217624e-06,
+      "loss": 0.0137,
+      "step": 3091
+    },
+    {
+      "epoch": 220.85714285714286,
+      "grad_norm": 0.09521484375,
+      "learning_rate": 9.552242343359897e-06,
+      "loss": 0.0137,
+      "step": 3092
+    },
+    {
+      "epoch": 220.92857142857142,
+      "grad_norm": 0.12109375,
+      "learning_rate": 9.543762537203445e-06,
+      "loss": 0.0144,
+      "step": 3093
+    },
+    {
+      "epoch": 221.0,
+      "grad_norm": 0.1533203125,
+      "learning_rate": 9.535284171881214e-06,
+      "loss": 0.0119,
+      "step": 3094
+    },
+    {
+      "epoch": 221.07142857142858,
+      "grad_norm": 0.30078125,
+      "learning_rate": 9.526807251525464e-06,
+      "loss": 0.0118,
+      "step": 3095
+    },
+    {
+      "epoch": 221.14285714285714,
+      "grad_norm": 0.1357421875,
+      "learning_rate": 9.518331780267742e-06,
+      "loss": 0.0106,
+      "step": 3096
+    },
+    {
+      "epoch": 221.21428571428572,
+      "grad_norm": 0.1435546875,
+      "learning_rate": 9.509857762238884e-06,
+      "loss": 0.0131,
+      "step": 3097
+    },
+    {
+      "epoch": 221.28571428571428,
+      "grad_norm": 0.10791015625,
+      "learning_rate": 9.501385201569033e-06,
+      "loss": 0.0114,
+      "step": 3098
+    },
+    {
+      "epoch": 221.35714285714286,
+      "grad_norm": 0.16796875,
+      "learning_rate": 9.492914102387603e-06,
+      "loss": 0.0119,
+      "step": 3099
+    },
+    {
+      "epoch": 221.42857142857142,
+      "grad_norm": 0.2353515625,
+      "learning_rate": 9.484444468823311e-06,
+      "loss": 0.0165,
+      "step": 3100
+    },
+    {
+      "epoch": 221.5,
+      "grad_norm": 0.35546875,
+      "learning_rate": 9.475976305004155e-06,
+      "loss": 0.0156,
+      "step": 3101
+    },
+    {
+      "epoch": 221.57142857142858,
+      "grad_norm": 0.10302734375,
+      "learning_rate": 9.46750961505741e-06,
+      "loss": 0.014,
+      "step": 3102
+    },
+    {
+      "epoch": 221.64285714285714,
+      "grad_norm": 0.11279296875,
+      "learning_rate": 9.45904440310964e-06,
+      "loss": 0.0135,
+      "step": 3103
+    },
+    {
+      "epoch": 221.71428571428572,
+      "grad_norm": 0.11669921875,
+      "learning_rate": 9.45058067328668e-06,
+      "loss": 0.0123,
+      "step": 3104
+    },
+    {
+      "epoch": 221.78571428571428,
+      "grad_norm": 0.1005859375,
+      "learning_rate": 9.44211842971366e-06,
+      "loss": 0.0147,
+      "step": 3105
+    },
+    {
+      "epoch": 221.85714285714286,
+      "grad_norm": 0.1533203125,
+      "learning_rate": 9.433657676514967e-06,
+      "loss": 0.0139,
+      "step": 3106
+    },
+    {
+      "epoch": 221.92857142857142,
+      "grad_norm": 0.259765625,
+      "learning_rate": 9.425198417814275e-06,
+      "loss": 0.0142,
+      "step": 3107
+    },
+    {
+      "epoch": 222.0,
+      "grad_norm": 0.2138671875,
+      "learning_rate": 9.416740657734521e-06,
+      "loss": 0.0129,
+      "step": 3108
+    },
+    {
+      "epoch": 222.07142857142858,
+      "grad_norm": 0.21875,
+      "learning_rate": 9.408284400397911e-06,
+      "loss": 0.0107,
+      "step": 3109
+    },
+    {
+      "epoch": 222.14285714285714,
+      "grad_norm": 0.1513671875,
+      "learning_rate": 9.399829649925935e-06,
+      "loss": 0.013,
+      "step": 3110
+    },
+    {
+      "epoch": 222.21428571428572,
+      "grad_norm": 0.130859375,
+      "learning_rate": 9.391376410439317e-06,
+      "loss": 0.014,
+      "step": 3111
+    },
+    {
+      "epoch": 222.28571428571428,
+      "grad_norm": 0.1396484375,
+      "learning_rate": 9.382924686058076e-06,
+      "loss": 0.0135,
+      "step": 3112
+    },
+    {
+      "epoch": 222.35714285714286,
+      "grad_norm": 0.11181640625,
+      "learning_rate": 9.374474480901474e-06,
+      "loss": 0.012,
+      "step": 3113
+    },
+    {
+      "epoch": 222.42857142857142,
+      "grad_norm": 0.216796875,
+      "learning_rate": 9.366025799088043e-06,
+      "loss": 0.0144,
+      "step": 3114
+    },
+    {
+      "epoch": 222.5,
+      "grad_norm": 0.1435546875,
+      "learning_rate": 9.357578644735565e-06,
+      "loss": 0.0139,
+      "step": 3115
+    },
+    {
+      "epoch": 222.57142857142858,
+      "grad_norm": 0.1416015625,
+      "learning_rate": 9.34913302196108e-06,
+      "loss": 0.0128,
+      "step": 3116
+    },
+    {
+      "epoch": 222.64285714285714,
+      "grad_norm": 0.10888671875,
+      "learning_rate": 9.340688934880885e-06,
+      "loss": 0.0118,
+      "step": 3117
+    },
+    {
+      "epoch": 222.71428571428572,
+      "grad_norm": 0.1142578125,
+      "learning_rate": 9.332246387610521e-06,
+      "loss": 0.0126,
+      "step": 3118
+    },
+    {
+      "epoch": 222.78571428571428,
+      "grad_norm": 0.1376953125,
+      "learning_rate": 9.32380538426479e-06,
+      "loss": 0.0149,
+      "step": 3119
+    },
+    {
+      "epoch": 222.85714285714286,
+      "grad_norm": 0.1591796875,
+      "learning_rate": 9.315365928957725e-06,
+      "loss": 0.0131,
+      "step": 3120
+    },
+    {
+      "epoch": 222.92857142857142,
+      "grad_norm": 0.146484375,
+      "learning_rate": 9.306928025802621e-06,
+      "loss": 0.0156,
+      "step": 3121
+    },
+    {
+      "epoch": 223.0,
+      "grad_norm": 0.1337890625,
+      "learning_rate": 9.298491678912006e-06,
+      "loss": 0.013,
+      "step": 3122
+    },
+    {
+      "epoch": 223.07142857142858,
+      "grad_norm": 0.2138671875,
+      "learning_rate": 9.290056892397652e-06,
+      "loss": 0.0135,
+      "step": 3123
+    },
+    {
+      "epoch": 223.14285714285714,
+      "grad_norm": 0.16796875,
+      "learning_rate": 9.28162367037058e-06,
+      "loss": 0.0109,
+      "step": 3124
+    },
+    {
+      "epoch": 223.21428571428572,
+      "grad_norm": 0.142578125,
+      "learning_rate": 9.273192016941031e-06,
+      "loss": 0.0119,
+      "step": 3125
+    },
+    {
+      "epoch": 223.28571428571428,
+      "grad_norm": 0.154296875,
+      "learning_rate": 9.26476193621849e-06,
+      "loss": 0.0125,
+      "step": 3126
+    },
+    {
+      "epoch": 223.35714285714286,
+      "grad_norm": 0.1416015625,
+      "learning_rate": 9.256333432311683e-06,
+      "loss": 0.0131,
+      "step": 3127
+    },
+    {
+      "epoch": 223.42857142857142,
+      "grad_norm": 0.166015625,
+      "learning_rate": 9.247906509328558e-06,
+      "loss": 0.0147,
+      "step": 3128
+    },
+    {
+      "epoch": 223.5,
+      "grad_norm": 0.12109375,
+      "learning_rate": 9.239481171376292e-06,
+      "loss": 0.0147,
+      "step": 3129
+    },
+    {
+      "epoch": 223.57142857142858,
+      "grad_norm": 0.1298828125,
+      "learning_rate": 9.23105742256129e-06,
+      "loss": 0.0138,
+      "step": 3130
+    },
+    {
+      "epoch": 223.64285714285714,
+      "grad_norm": 0.1279296875,
+      "learning_rate": 9.222635266989194e-06,
+      "loss": 0.0124,
+      "step": 3131
+    },
+    {
+      "epoch": 223.71428571428572,
+      "grad_norm": 0.1396484375,
+      "learning_rate": 9.21421470876485e-06,
+      "loss": 0.0135,
+      "step": 3132
+    },
+    {
+      "epoch": 223.78571428571428,
+      "grad_norm": 0.1416015625,
+      "learning_rate": 9.205795751992344e-06,
+      "loss": 0.0124,
+      "step": 3133
+    },
+    {
+      "epoch": 223.85714285714286,
+      "grad_norm": 0.21484375,
+      "learning_rate": 9.19737840077497e-06,
+      "loss": 0.0112,
+      "step": 3134
+    },
+    {
+      "epoch": 223.92857142857142,
+      "grad_norm": 0.1181640625,
+      "learning_rate": 9.188962659215238e-06,
+      "loss": 0.0114,
+      "step": 3135
+    },
+    {
+      "epoch": 224.0,
+      "grad_norm": 0.1474609375,
+      "learning_rate": 9.180548531414889e-06,
+      "loss": 0.0129,
+      "step": 3136
+    },
+    {
+      "epoch": 224.07142857142858,
+      "grad_norm": 0.134765625,
+      "learning_rate": 9.172136021474862e-06,
+      "loss": 0.0156,
+      "step": 3137
+    },
+    {
+      "epoch": 224.14285714285714,
+      "grad_norm": 0.28515625,
+      "learning_rate": 9.163725133495308e-06,
+      "loss": 0.0149,
+      "step": 3138
+    },
+    {
+      "epoch": 224.21428571428572,
+      "grad_norm": 0.103515625,
+      "learning_rate": 9.155315871575593e-06,
+      "loss": 0.011,
+      "step": 3139
+    },
+    {
+      "epoch": 224.28571428571428,
+      "grad_norm": 0.10888671875,
+      "learning_rate": 9.146908239814291e-06,
+      "loss": 0.0116,
+      "step": 3140
+    },
+    {
+      "epoch": 224.35714285714286,
+      "grad_norm": 0.1201171875,
+      "learning_rate": 9.13850224230918e-06,
+      "loss": 0.0103,
+      "step": 3141
+    },
+    {
+      "epoch": 224.42857142857142,
+      "grad_norm": 0.1279296875,
+      "learning_rate": 9.130097883157238e-06,
+      "loss": 0.0125,
+      "step": 3142
+    },
+    {
+      "epoch": 224.5,
+      "grad_norm": 0.1376953125,
+      "learning_rate": 9.121695166454652e-06,
+      "loss": 0.0107,
+      "step": 3143
+    },
+    {
+      "epoch": 224.57142857142858,
+      "grad_norm": 0.1572265625,
+      "learning_rate": 9.113294096296802e-06,
+      "loss": 0.0155,
+      "step": 3144
+    },
+    {
+      "epoch": 224.64285714285714,
+      "grad_norm": 0.181640625,
+      "learning_rate": 9.104894676778264e-06,
+      "loss": 0.0121,
+      "step": 3145
+    },
+    {
+      "epoch": 224.71428571428572,
+      "grad_norm": 0.1435546875,
+      "learning_rate": 9.096496911992821e-06,
+      "loss": 0.0124,
+      "step": 3146
+    },
+    {
+      "epoch": 224.78571428571428,
+      "grad_norm": 0.17578125,
+      "learning_rate": 9.088100806033436e-06,
+      "loss": 0.0129,
+      "step": 3147
+    },
+    {
+      "epoch": 224.85714285714286,
+      "grad_norm": 0.119140625,
+      "learning_rate": 9.079706362992272e-06,
+      "loss": 0.0126,
+      "step": 3148
+    },
+    {
+      "epoch": 224.92857142857142,
+      "grad_norm": 0.15625,
+      "learning_rate": 9.071313586960669e-06,
+      "loss": 0.0134,
+      "step": 3149
+    },
+    {
+      "epoch": 225.0,
+      "grad_norm": 0.255859375,
+      "learning_rate": 9.062922482029177e-06,
+      "loss": 0.0159,
+      "step": 3150
+    },
+    {
+      "epoch": 225.07142857142858,
+      "grad_norm": 0.18359375,
+      "learning_rate": 9.054533052287511e-06,
+      "loss": 0.0144,
+      "step": 3151
+    },
+    {
+      "epoch": 225.14285714285714,
+      "grad_norm": 0.2412109375,
+      "learning_rate": 9.04614530182458e-06,
+      "loss": 0.0135,
+      "step": 3152
+    },
+    {
+      "epoch": 225.21428571428572,
+      "grad_norm": 0.1748046875,
+      "learning_rate": 9.03775923472847e-06,
+      "loss": 0.0126,
+      "step": 3153
+    },
+    {
+      "epoch": 225.28571428571428,
+      "grad_norm": 0.12158203125,
+      "learning_rate": 9.02937485508645e-06,
+      "loss": 0.0116,
+      "step": 3154
+    },
+    {
+      "epoch": 225.35714285714286,
+      "grad_norm": 0.126953125,
+      "learning_rate": 9.02099216698497e-06,
+      "loss": 0.014,
+      "step": 3155
+    },
+    {
+      "epoch": 225.42857142857142,
+      "grad_norm": 0.1611328125,
+      "learning_rate": 9.012611174509635e-06,
+      "loss": 0.0108,
+      "step": 3156
+    },
+    {
+      "epoch": 225.5,
+      "grad_norm": 0.26171875,
+      "learning_rate": 9.004231881745254e-06,
+      "loss": 0.012,
+      "step": 3157
+    },
+    {
+      "epoch": 225.57142857142858,
+      "grad_norm": 0.314453125,
+      "learning_rate": 8.995854292775782e-06,
+      "loss": 0.0145,
+      "step": 3158
+    },
+    {
+      "epoch": 225.64285714285714,
+      "grad_norm": 0.267578125,
+      "learning_rate": 8.987478411684363e-06,
+      "loss": 0.0125,
+      "step": 3159
+    },
+    {
+      "epoch": 225.71428571428572,
+      "grad_norm": 0.203125,
+      "learning_rate": 8.979104242553293e-06,
+      "loss": 0.0119,
+      "step": 3160
+    },
+    {
+      "epoch": 225.78571428571428,
+      "grad_norm": 0.142578125,
+      "learning_rate": 8.970731789464043e-06,
+      "loss": 0.0136,
+      "step": 3161
+    },
+    {
+      "epoch": 225.85714285714286,
+      "grad_norm": 0.1474609375,
+      "learning_rate": 8.962361056497247e-06,
+      "loss": 0.0149,
+      "step": 3162
+    },
+    {
+      "epoch": 225.92857142857142,
+      "grad_norm": 0.130859375,
+      "learning_rate": 8.953992047732698e-06,
+      "loss": 0.0139,
+      "step": 3163
+    },
+    {
+      "epoch": 226.0,
+      "grad_norm": 0.146484375,
+      "learning_rate": 8.945624767249345e-06,
+      "loss": 0.0141,
+      "step": 3164
+    },
+    {
+      "epoch": 226.07142857142858,
+      "grad_norm": 0.1474609375,
+      "learning_rate": 8.937259219125299e-06,
+      "loss": 0.0132,
+      "step": 3165
+    },
+    {
+      "epoch": 226.14285714285714,
+      "grad_norm": 0.1474609375,
+      "learning_rate": 8.928895407437831e-06,
+      "loss": 0.0137,
+      "step": 3166
+    },
+    {
+      "epoch": 226.21428571428572,
+      "grad_norm": 0.11962890625,
+      "learning_rate": 8.920533336263361e-06,
+      "loss": 0.0154,
+      "step": 3167
+    },
+    {
+      "epoch": 226.28571428571428,
+      "grad_norm": 0.1298828125,
+      "learning_rate": 8.912173009677455e-06,
+      "loss": 0.0112,
+      "step": 3168
+    },
+    {
+      "epoch": 226.35714285714286,
+      "grad_norm": 0.12451171875,
+      "learning_rate": 8.903814431754842e-06,
+      "loss": 0.0152,
+      "step": 3169
+    },
+    {
+      "epoch": 226.42857142857142,
+      "grad_norm": 0.189453125,
+      "learning_rate": 8.895457606569388e-06,
+      "loss": 0.0125,
+      "step": 3170
+    },
+    {
+      "epoch": 226.5,
+      "grad_norm": 0.16796875,
+      "learning_rate": 8.887102538194103e-06,
+      "loss": 0.0135,
+      "step": 3171
+    },
+    {
+      "epoch": 226.57142857142858,
+      "grad_norm": 0.2578125,
+      "learning_rate": 8.878749230701159e-06,
+      "loss": 0.0122,
+      "step": 3172
+    },
+    {
+      "epoch": 226.64285714285714,
+      "grad_norm": 0.1318359375,
+      "learning_rate": 8.870397688161846e-06,
+      "loss": 0.0175,
+      "step": 3173
+    },
+    {
+      "epoch": 226.71428571428572,
+      "grad_norm": 0.173828125,
+      "learning_rate": 8.862047914646605e-06,
+      "loss": 0.0119,
+      "step": 3174
+    },
+    {
+      "epoch": 226.78571428571428,
+      "grad_norm": 0.1640625,
+      "learning_rate": 8.853699914225013e-06,
+      "loss": 0.0142,
+      "step": 3175
+    },
+    {
+      "epoch": 226.85714285714286,
+      "grad_norm": 0.1669921875,
+      "learning_rate": 8.845353690965792e-06,
+      "loss": 0.0105,
+      "step": 3176
+    },
+    {
+      "epoch": 226.92857142857142,
+      "grad_norm": 0.1171875,
+      "learning_rate": 8.837009248936784e-06,
+      "loss": 0.0159,
+      "step": 3177
+    },
+    {
+      "epoch": 227.0,
+      "grad_norm": 0.2021484375,
+      "learning_rate": 8.828666592204968e-06,
+      "loss": 0.0124,
+      "step": 3178
+    },
+    {
+      "epoch": 227.07142857142858,
+      "grad_norm": 0.1220703125,
+      "learning_rate": 8.820325724836457e-06,
+      "loss": 0.0131,
+      "step": 3179
+    },
+    {
+      "epoch": 227.14285714285714,
+      "grad_norm": 0.2412109375,
+      "learning_rate": 8.811986650896483e-06,
+      "loss": 0.0132,
+      "step": 3180
+    },
+    {
+      "epoch": 227.21428571428572,
+      "grad_norm": 0.11083984375,
+      "learning_rate": 8.803649374449421e-06,
+      "loss": 0.0108,
+      "step": 3181
+    },
+    {
+      "epoch": 227.28571428571428,
+      "grad_norm": 0.2333984375,
+      "learning_rate": 8.79531389955874e-06,
+      "loss": 0.015,
+      "step": 3182
+    },
+    {
+      "epoch": 227.35714285714286,
+      "grad_norm": 0.126953125,
+      "learning_rate": 8.786980230287067e-06,
+      "loss": 0.0126,
+      "step": 3183
+    },
+    {
+      "epoch": 227.42857142857142,
+      "grad_norm": 0.115234375,
+      "learning_rate": 8.77864837069612e-06,
+      "loss": 0.0133,
+      "step": 3184
+    },
+    {
+      "epoch": 227.5,
+      "grad_norm": 0.1923828125,
+      "learning_rate": 8.770318324846748e-06,
+      "loss": 0.011,
+      "step": 3185
+    },
+    {
+      "epoch": 227.57142857142858,
+      "grad_norm": 0.203125,
+      "learning_rate": 8.761990096798918e-06,
+      "loss": 0.0165,
+      "step": 3186
+    },
+    {
+      "epoch": 227.64285714285714,
+      "grad_norm": 0.12109375,
+      "learning_rate": 8.753663690611703e-06,
+      "loss": 0.0149,
+      "step": 3187
+    },
+    {
+      "epoch": 227.71428571428572,
+      "grad_norm": 0.111328125,
+      "learning_rate": 8.745339110343295e-06,
+      "loss": 0.0117,
+      "step": 3188
+    },
+    {
+      "epoch": 227.78571428571428,
+      "grad_norm": 0.1259765625,
+      "learning_rate": 8.737016360050995e-06,
+      "loss": 0.0138,
+      "step": 3189
+    },
+    {
+      "epoch": 227.85714285714286,
+      "grad_norm": 0.181640625,
+      "learning_rate": 8.72869544379121e-06,
+      "loss": 0.0135,
+      "step": 3190
+    },
+    {
+      "epoch": 227.92857142857142,
+      "grad_norm": 0.1708984375,
+      "learning_rate": 8.720376365619445e-06,
+      "loss": 0.0162,
+      "step": 3191
+    },
+    {
+      "epoch": 228.0,
+      "grad_norm": 0.1328125,
+      "learning_rate": 8.712059129590327e-06,
+      "loss": 0.0126,
+      "step": 3192
+    },
+    {
+      "epoch": 228.07142857142858,
+      "grad_norm": 0.1376953125,
+      "learning_rate": 8.703743739757573e-06,
+      "loss": 0.0154,
+      "step": 3193
+    },
+    {
+      "epoch": 228.14285714285714,
+      "grad_norm": 0.1865234375,
+      "learning_rate": 8.695430200174001e-06,
+      "loss": 0.0128,
+      "step": 3194
+    },
+    {
+      "epoch": 228.21428571428572,
+      "grad_norm": 0.1689453125,
+      "learning_rate": 8.687118514891531e-06,
+      "loss": 0.0142,
+      "step": 3195
+    },
+    {
+      "epoch": 228.28571428571428,
+      "grad_norm": 0.1533203125,
+      "learning_rate": 8.67880868796118e-06,
+      "loss": 0.0141,
+      "step": 3196
+    },
+    {
+      "epoch": 228.35714285714286,
+      "grad_norm": 0.1591796875,
+      "learning_rate": 8.67050072343305e-06,
+      "loss": 0.0139,
+      "step": 3197
+    },
+    {
+      "epoch": 228.42857142857142,
+      "grad_norm": 0.1533203125,
+      "learning_rate": 8.662194625356351e-06,
+      "loss": 0.013,
+      "step": 3198
+    },
+    {
+      "epoch": 228.5,
+      "grad_norm": 0.12353515625,
+      "learning_rate": 8.65389039777937e-06,
+      "loss": 0.0101,
+      "step": 3199
+    },
+    {
+      "epoch": 228.57142857142858,
+      "grad_norm": 0.1337890625,
+      "learning_rate": 8.645588044749484e-06,
+      "loss": 0.0111,
+      "step": 3200
+    },
+    {
+      "epoch": 228.64285714285714,
+      "grad_norm": 0.203125,
+      "learning_rate": 8.637287570313159e-06,
+      "loss": 0.0153,
+      "step": 3201
+    },
+    {
+      "epoch": 228.71428571428572,
+      "grad_norm": 0.1494140625,
+      "learning_rate": 8.62898897851595e-06,
+      "loss": 0.0151,
+      "step": 3202
+    },
+    {
+      "epoch": 228.78571428571428,
+      "grad_norm": 0.12451171875,
+      "learning_rate": 8.620692273402486e-06,
+      "loss": 0.0127,
+      "step": 3203
+    },
+    {
+      "epoch": 228.85714285714286,
+      "grad_norm": 0.1484375,
+      "learning_rate": 8.61239745901648e-06,
+      "loss": 0.0156,
+      "step": 3204
+    },
+    {
+      "epoch": 228.92857142857142,
+      "grad_norm": 0.14453125,
+      "learning_rate": 8.604104539400727e-06,
+      "loss": 0.0117,
+      "step": 3205
+    },
+    {
+      "epoch": 229.0,
+      "grad_norm": 0.1328125,
+      "learning_rate": 8.595813518597093e-06,
+      "loss": 0.0117,
+      "step": 3206
+    },
+    {
+      "epoch": 229.07142857142858,
+      "grad_norm": 0.12890625,
+      "learning_rate": 8.58752440064652e-06,
+      "loss": 0.0155,
+      "step": 3207
+    },
+    {
+      "epoch": 229.14285714285714,
+      "grad_norm": 0.12890625,
+      "learning_rate": 8.579237189589032e-06,
+      "loss": 0.0126,
+      "step": 3208
+    },
+    {
+      "epoch": 229.21428571428572,
+      "grad_norm": 0.134765625,
+      "learning_rate": 8.570951889463703e-06,
+      "loss": 0.0141,
+      "step": 3209
+    },
+    {
+      "epoch": 229.28571428571428,
+      "grad_norm": 0.1328125,
+      "learning_rate": 8.562668504308693e-06,
+      "loss": 0.013,
+      "step": 3210
+    },
+    {
+      "epoch": 229.35714285714286,
+      "grad_norm": 0.146484375,
+      "learning_rate": 8.554387038161221e-06,
+      "loss": 0.0145,
+      "step": 3211
+    },
+    {
+      "epoch": 229.42857142857142,
+      "grad_norm": 0.26953125,
+      "learning_rate": 8.546107495057576e-06,
+      "loss": 0.0144,
+      "step": 3212
+    },
+    {
+      "epoch": 229.5,
+      "grad_norm": 0.150390625,
+      "learning_rate": 8.537829879033104e-06,
+      "loss": 0.0147,
+      "step": 3213
+    },
+    {
+      "epoch": 229.57142857142858,
+      "grad_norm": 0.19140625,
+      "learning_rate": 8.529554194122216e-06,
+      "loss": 0.0124,
+      "step": 3214
+    },
+    {
+      "epoch": 229.64285714285714,
+      "grad_norm": 0.162109375,
+      "learning_rate": 8.521280444358383e-06,
+      "loss": 0.0151,
+      "step": 3215
+    },
+    {
+      "epoch": 229.71428571428572,
+      "grad_norm": 0.3828125,
+      "learning_rate": 8.513008633774125e-06,
+      "loss": 0.016,
+      "step": 3216
+    },
+    {
+      "epoch": 229.78571428571428,
+      "grad_norm": 0.146484375,
+      "learning_rate": 8.504738766401027e-06,
+      "loss": 0.0134,
+      "step": 3217
+    },
+    {
+      "epoch": 229.85714285714286,
+      "grad_norm": 0.1474609375,
+      "learning_rate": 8.496470846269718e-06,
+      "loss": 0.0115,
+      "step": 3218
+    },
+    {
+      "epoch": 229.92857142857142,
+      "grad_norm": 0.248046875,
+      "learning_rate": 8.488204877409884e-06,
+      "loss": 0.0154,
+      "step": 3219
+    },
+    {
+      "epoch": 230.0,
+      "grad_norm": 0.1982421875,
+      "learning_rate": 8.479940863850254e-06,
+      "loss": 0.0139,
+      "step": 3220
+    },
+    {
+      "epoch": 230.07142857142858,
+      "grad_norm": 0.1279296875,
+      "learning_rate": 8.471678809618613e-06,
+      "loss": 0.0118,
+      "step": 3221
+    },
+    {
+      "epoch": 230.14285714285714,
+      "grad_norm": 0.189453125,
+      "learning_rate": 8.463418718741783e-06,
+      "loss": 0.0137,
+      "step": 3222
+    },
+    {
+      "epoch": 230.21428571428572,
+      "grad_norm": 0.146484375,
+      "learning_rate": 8.455160595245631e-06,
+      "loss": 0.013,
+      "step": 3223
+    },
+    {
+      "epoch": 230.28571428571428,
+      "grad_norm": 0.1796875,
+      "learning_rate": 8.44690444315507e-06,
+      "loss": 0.0135,
+      "step": 3224
+    },
+    {
+      "epoch": 230.35714285714286,
+      "grad_norm": 0.1689453125,
+      "learning_rate": 8.438650266494047e-06,
+      "loss": 0.0128,
+      "step": 3225
+    },
+    {
+      "epoch": 230.42857142857142,
+      "grad_norm": 0.1494140625,
+      "learning_rate": 8.430398069285544e-06,
+      "loss": 0.0126,
+      "step": 3226
+    },
+    {
+      "epoch": 230.5,
+      "grad_norm": 0.1552734375,
+      "learning_rate": 8.422147855551583e-06,
+      "loss": 0.0147,
+      "step": 3227
+    },
+    {
+      "epoch": 230.57142857142858,
+      "grad_norm": 0.171875,
+      "learning_rate": 8.413899629313221e-06,
+      "loss": 0.0134,
+      "step": 3228
+    },
+    {
+      "epoch": 230.64285714285714,
+      "grad_norm": 0.1484375,
+      "learning_rate": 8.40565339459054e-06,
+      "loss": 0.0134,
+      "step": 3229
+    },
+    {
+      "epoch": 230.71428571428572,
+      "grad_norm": 0.11669921875,
+      "learning_rate": 8.397409155402655e-06,
+      "loss": 0.0151,
+      "step": 3230
+    },
+    {
+      "epoch": 230.78571428571428,
+      "grad_norm": 0.1474609375,
+      "learning_rate": 8.38916691576771e-06,
+      "loss": 0.0152,
+      "step": 3231
+    },
+    {
+      "epoch": 230.85714285714286,
+      "grad_norm": 0.1240234375,
+      "learning_rate": 8.380926679702868e-06,
+      "loss": 0.0135,
+      "step": 3232
+    },
+    {
+      "epoch": 230.92857142857142,
+      "grad_norm": 0.1875,
+      "learning_rate": 8.372688451224328e-06,
+      "loss": 0.013,
+      "step": 3233
+    },
+    {
+      "epoch": 231.0,
+      "grad_norm": 0.1591796875,
+      "learning_rate": 8.3644522343473e-06,
+      "loss": 0.0134,
+      "step": 3234
+    },
+    {
+      "epoch": 231.07142857142858,
+      "grad_norm": 0.10205078125,
+      "learning_rate": 8.35621803308601e-06,
+      "loss": 0.0108,
+      "step": 3235
+    },
+    {
+      "epoch": 231.14285714285714,
+      "grad_norm": 0.1376953125,
+      "learning_rate": 8.347985851453713e-06,
+      "loss": 0.0135,
+      "step": 3236
+    },
+    {
+      "epoch": 231.21428571428572,
+      "grad_norm": 0.193359375,
+      "learning_rate": 8.339755693462668e-06,
+      "loss": 0.0148,
+      "step": 3237
+    },
+    {
+      "epoch": 231.28571428571428,
+      "grad_norm": 0.1259765625,
+      "learning_rate": 8.331527563124163e-06,
+      "loss": 0.0154,
+      "step": 3238
+    },
+    {
+      "epoch": 231.35714285714286,
+      "grad_norm": 0.1748046875,
+      "learning_rate": 8.323301464448482e-06,
+      "loss": 0.0158,
+      "step": 3239
+    },
+    {
+      "epoch": 231.42857142857142,
+      "grad_norm": 0.11279296875,
+      "learning_rate": 8.31507740144493e-06,
+      "loss": 0.0117,
+      "step": 3240
+    },
+    {
+      "epoch": 231.5,
+      "grad_norm": 0.134765625,
+      "learning_rate": 8.306855378121813e-06,
+      "loss": 0.0135,
+      "step": 3241
+    },
+    {
+      "epoch": 231.57142857142858,
+      "grad_norm": 0.138671875,
+      "learning_rate": 8.298635398486445e-06,
+      "loss": 0.0109,
+      "step": 3242
+    },
+    {
+      "epoch": 231.64285714285714,
+      "grad_norm": 0.169921875,
+      "learning_rate": 8.29041746654515e-06,
+      "loss": 0.0138,
+      "step": 3243
+    },
+    {
+      "epoch": 231.71428571428572,
+      "grad_norm": 0.16015625,
+      "learning_rate": 8.282201586303237e-06,
+      "loss": 0.0138,
+      "step": 3244
+    },
+    {
+      "epoch": 231.78571428571428,
+      "grad_norm": 0.21484375,
+      "learning_rate": 8.273987761765037e-06,
+      "loss": 0.0148,
+      "step": 3245
+    },
+    {
+      "epoch": 231.85714285714286,
+      "grad_norm": 0.126953125,
+      "learning_rate": 8.265775996933858e-06,
+      "loss": 0.0121,
+      "step": 3246
+    },
+    {
+      "epoch": 231.92857142857142,
+      "grad_norm": 0.1376953125,
+      "learning_rate": 8.257566295812024e-06,
+      "loss": 0.0158,
+      "step": 3247
+    },
+    {
+      "epoch": 232.0,
+      "grad_norm": 0.1201171875,
+      "learning_rate": 8.249358662400842e-06,
+      "loss": 0.0145,
+      "step": 3248
+    },
+    {
+      "epoch": 232.07142857142858,
+      "grad_norm": 0.1298828125,
+      "learning_rate": 8.241153100700604e-06,
+      "loss": 0.0131,
+      "step": 3249
+    },
+    {
+      "epoch": 232.14285714285714,
+      "grad_norm": 0.166015625,
+      "learning_rate": 8.232949614710612e-06,
+      "loss": 0.0135,
+      "step": 3250
+    },
+    {
+      "epoch": 232.21428571428572,
+      "grad_norm": 0.16796875,
+      "learning_rate": 8.224748208429142e-06,
+      "loss": 0.0139,
+      "step": 3251
+    },
+    {
+      "epoch": 232.28571428571428,
+      "grad_norm": 0.1865234375,
+      "learning_rate": 8.21654888585346e-06,
+      "loss": 0.0133,
+      "step": 3252
+    },
+    {
+      "epoch": 232.35714285714286,
+      "grad_norm": 0.12451171875,
+      "learning_rate": 8.208351650979811e-06,
+      "loss": 0.0114,
+      "step": 3253
+    },
+    {
+      "epoch": 232.42857142857142,
+      "grad_norm": 0.189453125,
+      "learning_rate": 8.200156507803433e-06,
+      "loss": 0.0126,
+      "step": 3254
+    },
+    {
+      "epoch": 232.5,
+      "grad_norm": 0.134765625,
+      "learning_rate": 8.191963460318538e-06,
+      "loss": 0.0129,
+      "step": 3255
+    },
+    {
+      "epoch": 232.57142857142858,
+      "grad_norm": 0.1416015625,
+      "learning_rate": 8.183772512518318e-06,
+      "loss": 0.0133,
+      "step": 3256
+    },
+    {
+      "epoch": 232.64285714285714,
+      "grad_norm": 0.3828125,
+      "learning_rate": 8.17558366839494e-06,
+      "loss": 0.0115,
+      "step": 3257
+    },
+    {
+      "epoch": 232.71428571428572,
+      "grad_norm": 0.1318359375,
+      "learning_rate": 8.16739693193955e-06,
+      "loss": 0.0127,
+      "step": 3258
+    },
+    {
+      "epoch": 232.78571428571428,
+      "grad_norm": 0.1943359375,
+      "learning_rate": 8.159212307142262e-06,
+      "loss": 0.0152,
+      "step": 3259
+    },
+    {
+      "epoch": 232.85714285714286,
+      "grad_norm": 0.2197265625,
+      "learning_rate": 8.151029797992164e-06,
+      "loss": 0.0135,
+      "step": 3260
+    },
+    {
+      "epoch": 232.92857142857142,
+      "grad_norm": 0.1396484375,
+      "learning_rate": 8.142849408477312e-06,
+      "loss": 0.0134,
+      "step": 3261
+    },
+    {
+      "epoch": 233.0,
+      "grad_norm": 0.2119140625,
+      "learning_rate": 8.134671142584726e-06,
+      "loss": 0.0121,
+      "step": 3262
+    },
+    {
+      "epoch": 233.07142857142858,
+      "grad_norm": 0.15234375,
+      "learning_rate": 8.126495004300392e-06,
+      "loss": 0.0125,
+      "step": 3263
+    },
+    {
+      "epoch": 233.14285714285714,
+      "grad_norm": 0.1484375,
+      "learning_rate": 8.118320997609265e-06,
+      "loss": 0.0123,
+      "step": 3264
+    },
+    {
+      "epoch": 233.21428571428572,
+      "grad_norm": 0.166015625,
+      "learning_rate": 8.110149126495251e-06,
+      "loss": 0.0139,
+      "step": 3265
+    },
+    {
+      "epoch": 233.28571428571428,
+      "grad_norm": 0.17578125,
+      "learning_rate": 8.101979394941228e-06,
+      "loss": 0.0153,
+      "step": 3266
+    },
+    {
+      "epoch": 233.35714285714286,
+      "grad_norm": 0.140625,
+      "learning_rate": 8.093811806929019e-06,
+      "loss": 0.0147,
+      "step": 3267
+    },
+    {
+      "epoch": 233.42857142857142,
+      "grad_norm": 0.11083984375,
+      "learning_rate": 8.085646366439406e-06,
+      "loss": 0.0127,
+      "step": 3268
+    },
+    {
+      "epoch": 233.5,
+      "grad_norm": 0.2265625,
+      "learning_rate": 8.077483077452132e-06,
+      "loss": 0.0147,
+      "step": 3269
+    },
+    {
+      "epoch": 233.57142857142858,
+      "grad_norm": 0.1572265625,
+      "learning_rate": 8.069321943945881e-06,
+      "loss": 0.0135,
+      "step": 3270
+    },
+    {
+      "epoch": 233.64285714285714,
+      "grad_norm": 0.12890625,
+      "learning_rate": 8.06116296989829e-06,
+      "loss": 0.0138,
+      "step": 3271
+    },
+    {
+      "epoch": 233.71428571428572,
+      "grad_norm": 0.16015625,
+      "learning_rate": 8.053006159285939e-06,
+      "loss": 0.0139,
+      "step": 3272
+    },
+    {
+      "epoch": 233.78571428571428,
+      "grad_norm": 0.2060546875,
+      "learning_rate": 8.044851516084367e-06,
+      "loss": 0.0171,
+      "step": 3273
+    },
+    {
+      "epoch": 233.85714285714286,
+      "grad_norm": 0.1298828125,
+      "learning_rate": 8.036699044268045e-06,
+      "loss": 0.0123,
+      "step": 3274
+    },
+    {
+      "epoch": 233.92857142857142,
+      "grad_norm": 0.2021484375,
+      "learning_rate": 8.028548747810384e-06,
+      "loss": 0.0106,
+      "step": 3275
+    },
+    {
+      "epoch": 234.0,
+      "grad_norm": 0.12158203125,
+      "learning_rate": 8.020400630683746e-06,
+      "loss": 0.0116,
+      "step": 3276
+    },
+    {
+      "epoch": 234.07142857142858,
+      "grad_norm": 0.1337890625,
+      "learning_rate": 8.012254696859422e-06,
+      "loss": 0.0138,
+      "step": 3277
+    },
+    {
+      "epoch": 234.14285714285714,
+      "grad_norm": 0.1572265625,
+      "learning_rate": 8.00411095030764e-06,
+      "loss": 0.0113,
+      "step": 3278
+    },
+    {
+      "epoch": 234.21428571428572,
+      "grad_norm": 0.1494140625,
+      "learning_rate": 7.995969394997568e-06,
+      "loss": 0.0136,
+      "step": 3279
+    },
+    {
+      "epoch": 234.28571428571428,
+      "grad_norm": 0.169921875,
+      "learning_rate": 7.987830034897295e-06,
+      "loss": 0.0111,
+      "step": 3280
+    },
+    {
+      "epoch": 234.35714285714286,
+      "grad_norm": 0.11376953125,
+      "learning_rate": 7.979692873973849e-06,
+      "loss": 0.0119,
+      "step": 3281
+    },
+    {
+      "epoch": 234.42857142857142,
+      "grad_norm": 0.11376953125,
+      "learning_rate": 7.97155791619318e-06,
+      "loss": 0.0117,
+      "step": 3282
+    },
+    {
+      "epoch": 234.5,
+      "grad_norm": 0.302734375,
+      "learning_rate": 7.963425165520173e-06,
+      "loss": 0.0169,
+      "step": 3283
+    },
+    {
+      "epoch": 234.57142857142858,
+      "grad_norm": 0.123046875,
+      "learning_rate": 7.955294625918629e-06,
+      "loss": 0.0124,
+      "step": 3284
+    },
+    {
+      "epoch": 234.64285714285714,
+      "grad_norm": 0.12353515625,
+      "learning_rate": 7.947166301351273e-06,
+      "loss": 0.0123,
+      "step": 3285
+    },
+    {
+      "epoch": 234.71428571428572,
+      "grad_norm": 0.1220703125,
+      "learning_rate": 7.939040195779754e-06,
+      "loss": 0.0129,
+      "step": 3286
+    },
+    {
+      "epoch": 234.78571428571428,
+      "grad_norm": 0.12890625,
+      "learning_rate": 7.930916313164641e-06,
+      "loss": 0.0141,
+      "step": 3287
+    },
+    {
+      "epoch": 234.85714285714286,
+      "grad_norm": 0.33203125,
+      "learning_rate": 7.922794657465407e-06,
+      "loss": 0.0093,
+      "step": 3288
+    },
+    {
+      "epoch": 234.92857142857142,
+      "grad_norm": 0.1787109375,
+      "learning_rate": 7.91467523264045e-06,
+      "loss": 0.0142,
+      "step": 3289
+    },
+    {
+      "epoch": 235.0,
+      "grad_norm": 0.138671875,
+      "learning_rate": 7.906558042647086e-06,
+      "loss": 0.013,
+      "step": 3290
+    },
+    {
+      "epoch": 235.07142857142858,
+      "grad_norm": 0.1240234375,
+      "learning_rate": 7.898443091441526e-06,
+      "loss": 0.0142,
+      "step": 3291
+    },
+    {
+      "epoch": 235.14285714285714,
+      "grad_norm": 0.2470703125,
+      "learning_rate": 7.890330382978905e-06,
+      "loss": 0.0145,
+      "step": 3292
+    },
+    {
+      "epoch": 235.21428571428572,
+      "grad_norm": 0.1328125,
+      "learning_rate": 7.882219921213258e-06,
+      "loss": 0.0141,
+      "step": 3293
+    },
+    {
+      "epoch": 235.28571428571428,
+      "grad_norm": 0.130859375,
+      "learning_rate": 7.87411171009752e-06,
+      "loss": 0.0152,
+      "step": 3294
+    },
+    {
+      "epoch": 235.35714285714286,
+      "grad_norm": 0.1748046875,
+      "learning_rate": 7.866005753583542e-06,
+      "loss": 0.0149,
+      "step": 3295
+    },
+    {
+      "epoch": 235.42857142857142,
+      "grad_norm": 0.2021484375,
+      "learning_rate": 7.857902055622069e-06,
+      "loss": 0.013,
+      "step": 3296
+    },
+    {
+      "epoch": 235.5,
+      "grad_norm": 0.126953125,
+      "learning_rate": 7.849800620162737e-06,
+      "loss": 0.0123,
+      "step": 3297
+    },
+    {
+      "epoch": 235.57142857142858,
+      "grad_norm": 0.1474609375,
+      "learning_rate": 7.841701451154087e-06,
+      "loss": 0.0119,
+      "step": 3298
+    },
+    {
+      "epoch": 235.64285714285714,
+      "grad_norm": 0.31640625,
+      "learning_rate": 7.833604552543566e-06,
+      "loss": 0.0148,
+      "step": 3299
+    },
+    {
+      "epoch": 235.71428571428572,
+      "grad_norm": 0.1318359375,
+      "learning_rate": 7.825509928277498e-06,
+      "loss": 0.014,
+      "step": 3300
+    },
+    {
+      "epoch": 235.78571428571428,
+      "grad_norm": 0.162109375,
+      "learning_rate": 7.8174175823011e-06,
+      "loss": 0.0144,
+      "step": 3301
+    },
+    {
+      "epoch": 235.85714285714286,
+      "grad_norm": 0.134765625,
+      "learning_rate": 7.80932751855849e-06,
+      "loss": 0.0147,
+      "step": 3302
+    },
+    {
+      "epoch": 235.92857142857142,
+      "grad_norm": 0.11083984375,
+      "learning_rate": 7.801239740992666e-06,
+      "loss": 0.0111,
+      "step": 3303
+    },
+    {
+      "epoch": 236.0,
+      "grad_norm": 0.1328125,
+      "learning_rate": 7.793154253545509e-06,
+      "loss": 0.0136,
+      "step": 3304
+    },
+    {
+      "epoch": 236.07142857142858,
+      "grad_norm": 0.1435546875,
+      "learning_rate": 7.785071060157794e-06,
+      "loss": 0.0139,
+      "step": 3305
+    },
+    {
+      "epoch": 236.14285714285714,
+      "grad_norm": 0.1748046875,
+      "learning_rate": 7.776990164769163e-06,
+      "loss": 0.0128,
+      "step": 3306
+    },
+    {
+      "epoch": 236.21428571428572,
+      "grad_norm": 0.158203125,
+      "learning_rate": 7.768911571318152e-06,
+      "loss": 0.013,
+      "step": 3307
+    },
+    {
+      "epoch": 236.28571428571428,
+      "grad_norm": 0.1259765625,
+      "learning_rate": 7.760835283742162e-06,
+      "loss": 0.0132,
+      "step": 3308
+    },
+    {
+      "epoch": 236.35714285714286,
+      "grad_norm": 0.1435546875,
+      "learning_rate": 7.752761305977487e-06,
+      "loss": 0.0136,
+      "step": 3309
+    },
+    {
+      "epoch": 236.42857142857142,
+      "grad_norm": 0.11865234375,
+      "learning_rate": 7.744689641959283e-06,
+      "loss": 0.0124,
+      "step": 3310
+    },
+    {
+      "epoch": 236.5,
+      "grad_norm": 0.1484375,
+      "learning_rate": 7.736620295621574e-06,
+      "loss": 0.0156,
+      "step": 3311
+    },
+    {
+      "epoch": 236.57142857142858,
+      "grad_norm": 0.12255859375,
+      "learning_rate": 7.728553270897271e-06,
+      "loss": 0.0117,
+      "step": 3312
+    },
+    {
+      "epoch": 236.64285714285714,
+      "grad_norm": 0.134765625,
+      "learning_rate": 7.720488571718134e-06,
+      "loss": 0.0108,
+      "step": 3313
+    },
+    {
+      "epoch": 236.71428571428572,
+      "grad_norm": 0.1728515625,
+      "learning_rate": 7.712426202014817e-06,
+      "loss": 0.0117,
+      "step": 3314
+    },
+    {
+      "epoch": 236.78571428571428,
+      "grad_norm": 0.259765625,
+      "learning_rate": 7.704366165716798e-06,
+      "loss": 0.0113,
+      "step": 3315
+    },
+    {
+      "epoch": 236.85714285714286,
+      "grad_norm": 0.12890625,
+      "learning_rate": 7.696308466752455e-06,
+      "loss": 0.0114,
+      "step": 3316
+    },
+    {
+      "epoch": 236.92857142857142,
+      "grad_norm": 0.14453125,
+      "learning_rate": 7.688253109049007e-06,
+      "loss": 0.0118,
+      "step": 3317
+    },
+    {
+      "epoch": 237.0,
+      "grad_norm": 0.1337890625,
+      "learning_rate": 7.680200096532545e-06,
+      "loss": 0.0138,
+      "step": 3318
+    },
+    {
+      "epoch": 237.07142857142858,
+      "grad_norm": 0.259765625,
+      "learning_rate": 7.672149433128e-06,
+      "loss": 0.014,
+      "step": 3319
+    },
+    {
+      "epoch": 237.14285714285714,
+      "grad_norm": 0.1787109375,
+      "learning_rate": 7.664101122759173e-06,
+      "loss": 0.0131,
+      "step": 3320
+    },
+    {
+      "epoch": 237.21428571428572,
+      "grad_norm": 0.265625,
+      "learning_rate": 7.656055169348713e-06,
+      "loss": 0.0124,
+      "step": 3321
+    },
+    {
+      "epoch": 237.28571428571428,
+      "grad_norm": 0.283203125,
+      "learning_rate": 7.648011576818118e-06,
+      "loss": 0.0156,
+      "step": 3322
+    },
+    {
+      "epoch": 237.35714285714286,
+      "grad_norm": 0.1796875,
+      "learning_rate": 7.63997034908774e-06,
+      "loss": 0.0107,
+      "step": 3323
+    },
+    {
+      "epoch": 237.42857142857142,
+      "grad_norm": 0.171875,
+      "learning_rate": 7.63193149007677e-06,
+      "loss": 0.0133,
+      "step": 3324
+    },
+    {
+      "epoch": 237.5,
+      "grad_norm": 0.1337890625,
+      "learning_rate": 7.623895003703255e-06,
+      "loss": 0.0139,
+      "step": 3325
+    },
+    {
+      "epoch": 237.57142857142858,
+      "grad_norm": 0.2412109375,
+      "learning_rate": 7.61586089388408e-06,
+      "loss": 0.0144,
+      "step": 3326
+    },
+    {
+      "epoch": 237.64285714285714,
+      "grad_norm": 0.1748046875,
+      "learning_rate": 7.60782916453497e-06,
+      "loss": 0.0156,
+      "step": 3327
+    },
+    {
+      "epoch": 237.71428571428572,
+      "grad_norm": 0.1865234375,
+      "learning_rate": 7.599799819570497e-06,
+      "loss": 0.0128,
+      "step": 3328
+    },
+    {
+      "epoch": 237.78571428571428,
+      "grad_norm": 0.1328125,
+      "learning_rate": 7.591772862904064e-06,
+      "loss": 0.0124,
+      "step": 3329
+    },
+    {
+      "epoch": 237.85714285714286,
+      "grad_norm": 0.177734375,
+      "learning_rate": 7.583748298447907e-06,
+      "loss": 0.0131,
+      "step": 3330
+    },
+    {
+      "epoch": 237.92857142857142,
+      "grad_norm": 0.1845703125,
+      "learning_rate": 7.575726130113112e-06,
+      "loss": 0.0128,
+      "step": 3331
+    },
+    {
+      "epoch": 238.0,
+      "grad_norm": 0.146484375,
+      "learning_rate": 7.567706361809579e-06,
+      "loss": 0.0139,
+      "step": 3332
+    },
+    {
+      "epoch": 238.07142857142858,
+      "grad_norm": 0.13671875,
+      "learning_rate": 7.559688997446047e-06,
+      "loss": 0.0118,
+      "step": 3333
+    },
+    {
+      "epoch": 238.14285714285714,
+      "grad_norm": 0.1220703125,
+      "learning_rate": 7.551674040930079e-06,
+      "loss": 0.0131,
+      "step": 3334
+    },
+    {
+      "epoch": 238.21428571428572,
+      "grad_norm": 0.15234375,
+      "learning_rate": 7.543661496168074e-06,
+      "loss": 0.013,
+      "step": 3335
+    },
+    {
+      "epoch": 238.28571428571428,
+      "grad_norm": 0.11181640625,
+      "learning_rate": 7.535651367065244e-06,
+      "loss": 0.0122,
+      "step": 3336
+    },
+    {
+      "epoch": 238.35714285714286,
+      "grad_norm": 0.25390625,
+      "learning_rate": 7.52764365752563e-06,
+      "loss": 0.0124,
+      "step": 3337
+    },
+    {
+      "epoch": 238.42857142857142,
+      "grad_norm": 0.197265625,
+      "learning_rate": 7.519638371452092e-06,
+      "loss": 0.0147,
+      "step": 3338
+    },
+    {
+      "epoch": 238.5,
+      "grad_norm": 0.19921875,
+      "learning_rate": 7.511635512746309e-06,
+      "loss": 0.0139,
+      "step": 3339
+    },
+    {
+      "epoch": 238.57142857142858,
+      "grad_norm": 0.2109375,
+      "learning_rate": 7.50363508530878e-06,
+      "loss": 0.0112,
+      "step": 3340
+    },
+    {
+      "epoch": 238.64285714285714,
+      "grad_norm": 0.1279296875,
+      "learning_rate": 7.495637093038816e-06,
+      "loss": 0.0133,
+      "step": 3341
+    },
+    {
+      "epoch": 238.71428571428572,
+      "grad_norm": 0.12060546875,
+      "learning_rate": 7.487641539834536e-06,
+      "loss": 0.0121,
+      "step": 3342
+    },
+    {
+      "epoch": 238.78571428571428,
+      "grad_norm": 0.2578125,
+      "learning_rate": 7.479648429592877e-06,
+      "loss": 0.0108,
+      "step": 3343
+    },
+    {
+      "epoch": 238.85714285714286,
+      "grad_norm": 0.1318359375,
+      "learning_rate": 7.4716577662095884e-06,
+      "loss": 0.0128,
+      "step": 3344
+    },
+    {
+      "epoch": 238.92857142857142,
+      "grad_norm": 0.11669921875,
+      "learning_rate": 7.463669553579218e-06,
+      "loss": 0.0119,
+      "step": 3345
+    },
+    {
+      "epoch": 239.0,
+      "grad_norm": 0.2578125,
+      "learning_rate": 7.455683795595124e-06,
+      "loss": 0.0156,
+      "step": 3346
+    },
+    {
+      "epoch": 239.07142857142858,
+      "grad_norm": 0.146484375,
+      "learning_rate": 7.447700496149471e-06,
+      "loss": 0.0131,
+      "step": 3347
+    },
+    {
+      "epoch": 239.14285714285714,
+      "grad_norm": 0.2412109375,
+      "learning_rate": 7.439719659133221e-06,
+      "loss": 0.0119,
+      "step": 3348
+    },
+    {
+      "epoch": 239.21428571428572,
+      "grad_norm": 0.11572265625,
+      "learning_rate": 7.4317412884361375e-06,
+      "loss": 0.0117,
+      "step": 3349
+    },
+    {
+      "epoch": 239.28571428571428,
+      "grad_norm": 0.14453125,
+      "learning_rate": 7.423765387946777e-06,
+      "loss": 0.0143,
+      "step": 3350
+    },
+    {
+      "epoch": 239.35714285714286,
+      "grad_norm": 0.1318359375,
+      "learning_rate": 7.4157919615524994e-06,
+      "loss": 0.0154,
+      "step": 3351
+    },
+    {
+      "epoch": 239.42857142857142,
+      "grad_norm": 0.11669921875,
+      "learning_rate": 7.407821013139457e-06,
+      "loss": 0.0112,
+      "step": 3352
+    },
+    {
+      "epoch": 239.5,
+      "grad_norm": 0.197265625,
+      "learning_rate": 7.399852546592591e-06,
+      "loss": 0.0158,
+      "step": 3353
+    },
+    {
+      "epoch": 239.57142857142858,
+      "grad_norm": 0.2099609375,
+      "learning_rate": 7.391886565795639e-06,
+      "loss": 0.0102,
+      "step": 3354
+    },
+    {
+      "epoch": 239.64285714285714,
+      "grad_norm": 0.1279296875,
+      "learning_rate": 7.3839230746311185e-06,
+      "loss": 0.0129,
+      "step": 3355
+    },
+    {
+      "epoch": 239.71428571428572,
+      "grad_norm": 0.1611328125,
+      "learning_rate": 7.37596207698034e-06,
+      "loss": 0.0128,
+      "step": 3356
+    },
+    {
+      "epoch": 239.78571428571428,
+      "grad_norm": 0.265625,
+      "learning_rate": 7.368003576723401e-06,
+      "loss": 0.0153,
+      "step": 3357
+    },
+    {
+      "epoch": 239.85714285714286,
+      "grad_norm": 0.134765625,
+      "learning_rate": 7.360047577739175e-06,
+      "loss": 0.012,
+      "step": 3358
+    },
+    {
+      "epoch": 239.92857142857142,
+      "grad_norm": 0.169921875,
+      "learning_rate": 7.352094083905317e-06,
+      "loss": 0.0119,
+      "step": 3359
+    },
+    {
+      "epoch": 240.0,
+      "grad_norm": 0.19140625,
+      "learning_rate": 7.344143099098262e-06,
+      "loss": 0.0133,
+      "step": 3360
+    },
+    {
+      "epoch": 240.07142857142858,
+      "grad_norm": 0.1455078125,
+      "learning_rate": 7.33619462719323e-06,
+      "loss": 0.0148,
+      "step": 3361
+    },
+    {
+      "epoch": 240.14285714285714,
+      "grad_norm": 0.1787109375,
+      "learning_rate": 7.328248672064206e-06,
+      "loss": 0.0134,
+      "step": 3362
+    },
+    {
+      "epoch": 240.21428571428572,
+      "grad_norm": 0.12060546875,
+      "learning_rate": 7.320305237583949e-06,
+      "loss": 0.0123,
+      "step": 3363
+    },
+    {
+      "epoch": 240.28571428571428,
+      "grad_norm": 0.349609375,
+      "learning_rate": 7.312364327623998e-06,
+      "loss": 0.0151,
+      "step": 3364
+    },
+    {
+      "epoch": 240.35714285714286,
+      "grad_norm": 0.2265625,
+      "learning_rate": 7.304425946054652e-06,
+      "loss": 0.0118,
+      "step": 3365
+    },
+    {
+      "epoch": 240.42857142857142,
+      "grad_norm": 0.134765625,
+      "learning_rate": 7.2964900967449875e-06,
+      "loss": 0.0148,
+      "step": 3366
+    },
+    {
+      "epoch": 240.5,
+      "grad_norm": 0.15234375,
+      "learning_rate": 7.288556783562834e-06,
+      "loss": 0.0131,
+      "step": 3367
+    },
+    {
+      "epoch": 240.57142857142858,
+      "grad_norm": 0.1982421875,
+      "learning_rate": 7.2806260103747955e-06,
+      "loss": 0.0122,
+      "step": 3368
+    },
+    {
+      "epoch": 240.64285714285714,
+      "grad_norm": 0.140625,
+      "learning_rate": 7.272697781046234e-06,
+      "loss": 0.0131,
+      "step": 3369
+    },
+    {
+      "epoch": 240.71428571428572,
+      "grad_norm": 0.248046875,
+      "learning_rate": 7.264772099441272e-06,
+      "loss": 0.0115,
+      "step": 3370
+    },
+    {
+      "epoch": 240.78571428571428,
+      "grad_norm": 0.14453125,
+      "learning_rate": 7.256848969422786e-06,
+      "loss": 0.0139,
+      "step": 3371
+    },
+    {
+      "epoch": 240.85714285714286,
+      "grad_norm": 0.1669921875,
+      "learning_rate": 7.248928394852422e-06,
+      "loss": 0.0119,
+      "step": 3372
+    },
+    {
+      "epoch": 240.92857142857142,
+      "grad_norm": 0.1142578125,
+      "learning_rate": 7.241010379590568e-06,
+      "loss": 0.0124,
+      "step": 3373
+    },
+    {
+      "epoch": 241.0,
+      "grad_norm": 0.2138671875,
+      "learning_rate": 7.233094927496367e-06,
+      "loss": 0.0132,
+      "step": 3374
+    },
+    {
+      "epoch": 241.07142857142858,
+      "grad_norm": 0.19921875,
+      "learning_rate": 7.225182042427717e-06,
+      "loss": 0.0128,
+      "step": 3375
+    },
+    {
+      "epoch": 241.14285714285714,
+      "grad_norm": 0.13671875,
+      "learning_rate": 7.21727172824126e-06,
+      "loss": 0.0122,
+      "step": 3376
+    },
+    {
+      "epoch": 241.21428571428572,
+      "grad_norm": 0.1640625,
+      "learning_rate": 7.209363988792387e-06,
+      "loss": 0.0139,
+      "step": 3377
+    },
+    {
+      "epoch": 241.28571428571428,
+      "grad_norm": 0.1455078125,
+      "learning_rate": 7.201458827935233e-06,
+      "loss": 0.0114,
+      "step": 3378
+    },
+    {
+      "epoch": 241.35714285714286,
+      "grad_norm": 0.1728515625,
+      "learning_rate": 7.19355624952268e-06,
+      "loss": 0.0155,
+      "step": 3379
+    },
+    {
+      "epoch": 241.42857142857142,
+      "grad_norm": 0.1787109375,
+      "learning_rate": 7.1856562574063426e-06,
+      "loss": 0.0129,
+      "step": 3380
+    },
+    {
+      "epoch": 241.5,
+      "grad_norm": 0.130859375,
+      "learning_rate": 7.177758855436592e-06,
+      "loss": 0.0149,
+      "step": 3381
+    },
+    {
+      "epoch": 241.57142857142858,
+      "grad_norm": 0.134765625,
+      "learning_rate": 7.169864047462518e-06,
+      "loss": 0.0147,
+      "step": 3382
+    },
+    {
+      "epoch": 241.64285714285714,
+      "grad_norm": 0.126953125,
+      "learning_rate": 7.161971837331957e-06,
+      "loss": 0.0125,
+      "step": 3383
+    },
+    {
+      "epoch": 241.71428571428572,
+      "grad_norm": 0.181640625,
+      "learning_rate": 7.154082228891476e-06,
+      "loss": 0.012,
+      "step": 3384
+    },
+    {
+      "epoch": 241.78571428571428,
+      "grad_norm": 0.1982421875,
+      "learning_rate": 7.146195225986373e-06,
+      "loss": 0.0148,
+      "step": 3385
+    },
+    {
+      "epoch": 241.85714285714286,
+      "grad_norm": 0.1298828125,
+      "learning_rate": 7.138310832460679e-06,
+      "loss": 0.0141,
+      "step": 3386
+    },
+    {
+      "epoch": 241.92857142857142,
+      "grad_norm": 0.1396484375,
+      "learning_rate": 7.130429052157149e-06,
+      "loss": 0.0122,
+      "step": 3387
+    },
+    {
+      "epoch": 242.0,
+      "grad_norm": 0.126953125,
+      "learning_rate": 7.1225498889172695e-06,
+      "loss": 0.015,
+      "step": 3388
+    },
+    {
+      "epoch": 242.07142857142858,
+      "grad_norm": 0.126953125,
+      "learning_rate": 7.11467334658125e-06,
+      "loss": 0.0127,
+      "step": 3389
+    },
+    {
+      "epoch": 242.14285714285714,
+      "grad_norm": 0.1875,
+      "learning_rate": 7.106799428988012e-06,
+      "loss": 0.0134,
+      "step": 3390
+    },
+    {
+      "epoch": 242.21428571428572,
+      "grad_norm": 0.134765625,
+      "learning_rate": 7.098928139975219e-06,
+      "loss": 0.0136,
+      "step": 3391
+    },
+    {
+      "epoch": 242.28571428571428,
+      "grad_norm": 0.1591796875,
+      "learning_rate": 7.091059483379241e-06,
+      "loss": 0.0148,
+      "step": 3392
+    },
+    {
+      "epoch": 242.35714285714286,
+      "grad_norm": 0.1923828125,
+      "learning_rate": 7.083193463035161e-06,
+      "loss": 0.0144,
+      "step": 3393
+    },
+    {
+      "epoch": 242.42857142857142,
+      "grad_norm": 0.126953125,
+      "learning_rate": 7.075330082776782e-06,
+      "loss": 0.0109,
+      "step": 3394
+    },
+    {
+      "epoch": 242.5,
+      "grad_norm": 0.15234375,
+      "learning_rate": 7.067469346436623e-06,
+      "loss": 0.0119,
+      "step": 3395
+    },
+    {
+      "epoch": 242.57142857142858,
+      "grad_norm": 0.11767578125,
+      "learning_rate": 7.059611257845908e-06,
+      "loss": 0.012,
+      "step": 3396
+    },
+    {
+      "epoch": 242.64285714285714,
+      "grad_norm": 0.1669921875,
+      "learning_rate": 7.051755820834576e-06,
+      "loss": 0.0119,
+      "step": 3397
+    },
+    {
+      "epoch": 242.71428571428572,
+      "grad_norm": 0.13671875,
+      "learning_rate": 7.04390303923127e-06,
+      "loss": 0.0137,
+      "step": 3398
+    },
+    {
+      "epoch": 242.78571428571428,
+      "grad_norm": 0.1953125,
+      "learning_rate": 7.036052916863342e-06,
+      "loss": 0.0146,
+      "step": 3399
+    },
+    {
+      "epoch": 242.85714285714286,
+      "grad_norm": 0.1552734375,
+      "learning_rate": 7.02820545755684e-06,
+      "loss": 0.0152,
+      "step": 3400
+    },
+    {
+      "epoch": 242.92857142857142,
+      "grad_norm": 0.1552734375,
+      "learning_rate": 7.020360665136531e-06,
+      "loss": 0.0131,
+      "step": 3401
+    },
+    {
+      "epoch": 243.0,
+      "grad_norm": 0.18359375,
+      "learning_rate": 7.012518543425866e-06,
+      "loss": 0.0109,
+      "step": 3402
+    },
+    {
+      "epoch": 243.07142857142858,
+      "grad_norm": 0.12353515625,
+      "learning_rate": 7.004679096247008e-06,
+      "loss": 0.0136,
+      "step": 3403
+    },
+    {
+      "epoch": 243.14285714285714,
+      "grad_norm": 0.1708984375,
+      "learning_rate": 6.996842327420787e-06,
+      "loss": 0.0134,
+      "step": 3404
+    },
+    {
+      "epoch": 243.21428571428572,
+      "grad_norm": 0.37890625,
+      "learning_rate": 6.98900824076677e-06,
+      "loss": 0.0136,
+      "step": 3405
+    },
+    {
+      "epoch": 243.28571428571428,
+      "grad_norm": 0.193359375,
+      "learning_rate": 6.981176840103186e-06,
+      "loss": 0.0106,
+      "step": 3406
+    },
+    {
+      "epoch": 243.35714285714286,
+      "grad_norm": 0.24609375,
+      "learning_rate": 6.973348129246968e-06,
+      "loss": 0.0149,
+      "step": 3407
+    },
+    {
+      "epoch": 243.42857142857142,
+      "grad_norm": 0.146484375,
+      "learning_rate": 6.965522112013734e-06,
+      "loss": 0.0123,
+      "step": 3408
+    },
+    {
+      "epoch": 243.5,
+      "grad_norm": 0.1533203125,
+      "learning_rate": 6.957698792217782e-06,
+      "loss": 0.013,
+      "step": 3409
+    },
+    {
+      "epoch": 243.57142857142858,
+      "grad_norm": 0.14453125,
+      "learning_rate": 6.949878173672117e-06,
+      "loss": 0.0133,
+      "step": 3410
+    },
+    {
+      "epoch": 243.64285714285714,
+      "grad_norm": 0.169921875,
+      "learning_rate": 6.94206026018841e-06,
+      "loss": 0.0144,
+      "step": 3411
+    },
+    {
+      "epoch": 243.71428571428572,
+      "grad_norm": 0.12353515625,
+      "learning_rate": 6.93424505557701e-06,
+      "loss": 0.0125,
+      "step": 3412
+    },
+    {
+      "epoch": 243.78571428571428,
+      "grad_norm": 0.169921875,
+      "learning_rate": 6.926432563646951e-06,
+      "loss": 0.012,
+      "step": 3413
+    },
+    {
+      "epoch": 243.85714285714286,
+      "grad_norm": 0.197265625,
+      "learning_rate": 6.9186227882059554e-06,
+      "loss": 0.0107,
+      "step": 3414
+    },
+    {
+      "epoch": 243.92857142857142,
+      "grad_norm": 0.13671875,
+      "learning_rate": 6.9108157330604106e-06,
+      "loss": 0.0137,
+      "step": 3415
+    },
+    {
+      "epoch": 244.0,
+      "grad_norm": 0.1142578125,
+      "learning_rate": 6.90301140201538e-06,
+      "loss": 0.0111,
+      "step": 3416
+    },
+    {
+      "epoch": 244.07142857142858,
+      "grad_norm": 0.1240234375,
+      "learning_rate": 6.8952097988745995e-06,
+      "loss": 0.013,
+      "step": 3417
+    },
+    {
+      "epoch": 244.14285714285714,
+      "grad_norm": 0.109375,
+      "learning_rate": 6.887410927440474e-06,
+      "loss": 0.0119,
+      "step": 3418
+    },
+    {
+      "epoch": 244.21428571428572,
+      "grad_norm": 0.119140625,
+      "learning_rate": 6.879614791514075e-06,
+      "loss": 0.0118,
+      "step": 3419
+    },
+    {
+      "epoch": 244.28571428571428,
+      "grad_norm": 0.1337890625,
+      "learning_rate": 6.871821394895159e-06,
+      "loss": 0.0109,
+      "step": 3420
+    },
+    {
+      "epoch": 244.35714285714286,
+      "grad_norm": 0.1962890625,
+      "learning_rate": 6.864030741382119e-06,
+      "loss": 0.013,
+      "step": 3421
+    },
+    {
+      "epoch": 244.42857142857142,
+      "grad_norm": 0.12353515625,
+      "learning_rate": 6.85624283477203e-06,
+      "loss": 0.014,
+      "step": 3422
+    },
+    {
+      "epoch": 244.5,
+      "grad_norm": 0.15625,
+      "learning_rate": 6.848457678860615e-06,
+      "loss": 0.0104,
+      "step": 3423
+    },
+    {
+      "epoch": 244.57142857142858,
+      "grad_norm": 0.1259765625,
+      "learning_rate": 6.8406752774422795e-06,
+      "loss": 0.0141,
+      "step": 3424
+    },
+    {
+      "epoch": 244.64285714285714,
+      "grad_norm": 0.1787109375,
+      "learning_rate": 6.83289563431006e-06,
+      "loss": 0.0149,
+      "step": 3425
+    },
+    {
+      "epoch": 244.71428571428572,
+      "grad_norm": 0.1728515625,
+      "learning_rate": 6.825118753255666e-06,
+      "loss": 0.0151,
+      "step": 3426
+    },
+    {
+      "epoch": 244.78571428571428,
+      "grad_norm": 0.2255859375,
+      "learning_rate": 6.817344638069452e-06,
+      "loss": 0.0143,
+      "step": 3427
+    },
+    {
+      "epoch": 244.85714285714286,
+      "grad_norm": 0.154296875,
+      "learning_rate": 6.809573292540422e-06,
+      "loss": 0.0138,
+      "step": 3428
+    },
+    {
+      "epoch": 244.92857142857142,
+      "grad_norm": 0.1533203125,
+      "learning_rate": 6.8018047204562535e-06,
+      "loss": 0.0134,
+      "step": 3429
+    },
+    {
+      "epoch": 245.0,
+      "grad_norm": 0.142578125,
+      "learning_rate": 6.794038925603234e-06,
+      "loss": 0.0129,
+      "step": 3430
+    },
+    {
+      "epoch": 245.07142857142858,
+      "grad_norm": 0.177734375,
+      "learning_rate": 6.786275911766327e-06,
+      "loss": 0.0132,
+      "step": 3431
+    },
+    {
+      "epoch": 245.14285714285714,
+      "grad_norm": 0.2109375,
+      "learning_rate": 6.778515682729128e-06,
+      "loss": 0.0129,
+      "step": 3432
+    },
+    {
+      "epoch": 245.21428571428572,
+      "grad_norm": 0.1279296875,
+      "learning_rate": 6.770758242273877e-06,
+      "loss": 0.0126,
+      "step": 3433
+    },
+    {
+      "epoch": 245.28571428571428,
+      "grad_norm": 0.1435546875,
+      "learning_rate": 6.763003594181462e-06,
+      "loss": 0.0129,
+      "step": 3434
+    },
+    {
+      "epoch": 245.35714285714286,
+      "grad_norm": 0.1083984375,
+      "learning_rate": 6.755251742231403e-06,
+      "loss": 0.0118,
+      "step": 3435
+    },
+    {
+      "epoch": 245.42857142857142,
+      "grad_norm": 0.162109375,
+      "learning_rate": 6.747502690201855e-06,
+      "loss": 0.0126,
+      "step": 3436
+    },
+    {
+      "epoch": 245.5,
+      "grad_norm": 0.251953125,
+      "learning_rate": 6.739756441869616e-06,
+      "loss": 0.0161,
+      "step": 3437
+    },
+    {
+      "epoch": 245.57142857142858,
+      "grad_norm": 0.15234375,
+      "learning_rate": 6.7320130010101095e-06,
+      "loss": 0.0119,
+      "step": 3438
+    },
+    {
+      "epoch": 245.64285714285714,
+      "grad_norm": 0.11865234375,
+      "learning_rate": 6.724272371397396e-06,
+      "loss": 0.0124,
+      "step": 3439
+    },
+    {
+      "epoch": 245.71428571428572,
+      "grad_norm": 0.162109375,
+      "learning_rate": 6.716534556804165e-06,
+      "loss": 0.015,
+      "step": 3440
+    },
+    {
+      "epoch": 245.78571428571428,
+      "grad_norm": 0.185546875,
+      "learning_rate": 6.708799561001732e-06,
+      "loss": 0.0136,
+      "step": 3441
+    },
+    {
+      "epoch": 245.85714285714286,
+      "grad_norm": 0.1318359375,
+      "learning_rate": 6.701067387760035e-06,
+      "loss": 0.015,
+      "step": 3442
+    },
+    {
+      "epoch": 245.92857142857142,
+      "grad_norm": 0.318359375,
+      "learning_rate": 6.6933380408476545e-06,
+      "loss": 0.0165,
+      "step": 3443
+    },
+    {
+      "epoch": 246.0,
+      "grad_norm": 0.3046875,
+      "learning_rate": 6.68561152403177e-06,
+      "loss": 0.0135,
+      "step": 3444
+    },
+    {
+      "epoch": 246.07142857142858,
+      "grad_norm": 0.1083984375,
+      "learning_rate": 6.677887841078197e-06,
+      "loss": 0.0127,
+      "step": 3445
+    },
+    {
+      "epoch": 246.14285714285714,
+      "grad_norm": 0.1435546875,
+      "learning_rate": 6.670166995751362e-06,
+      "loss": 0.0125,
+      "step": 3446
+    },
+    {
+      "epoch": 246.21428571428572,
+      "grad_norm": 0.1455078125,
+      "learning_rate": 6.66244899181431e-06,
+      "loss": 0.0163,
+      "step": 3447
+    },
+    {
+      "epoch": 246.28571428571428,
+      "grad_norm": 0.14453125,
+      "learning_rate": 6.6547338330287045e-06,
+      "loss": 0.0116,
+      "step": 3448
+    },
+    {
+      "epoch": 246.35714285714286,
+      "grad_norm": 0.1435546875,
+      "learning_rate": 6.64702152315482e-06,
+      "loss": 0.0132,
+      "step": 3449
+    },
+    {
+      "epoch": 246.42857142857142,
+      "grad_norm": 0.1533203125,
+      "learning_rate": 6.639312065951543e-06,
+      "loss": 0.0128,
+      "step": 3450
+    },
+    {
+      "epoch": 246.5,
+      "grad_norm": 0.1748046875,
+      "learning_rate": 6.631605465176368e-06,
+      "loss": 0.0137,
+      "step": 3451
+    },
+    {
+      "epoch": 246.57142857142858,
+      "grad_norm": 0.1337890625,
+      "learning_rate": 6.623901724585395e-06,
+      "loss": 0.0112,
+      "step": 3452
+    },
+    {
+      "epoch": 246.64285714285714,
+      "grad_norm": 0.1513671875,
+      "learning_rate": 6.6162008479333436e-06,
+      "loss": 0.0112,
+      "step": 3453
+    },
+    {
+      "epoch": 246.71428571428572,
+      "grad_norm": 0.11865234375,
+      "learning_rate": 6.6085028389735185e-06,
+      "loss": 0.0101,
+      "step": 3454
+    },
+    {
+      "epoch": 246.78571428571428,
+      "grad_norm": 0.10302734375,
+      "learning_rate": 6.600807701457842e-06,
+      "loss": 0.0132,
+      "step": 3455
+    },
+    {
+      "epoch": 246.85714285714286,
+      "grad_norm": 0.1708984375,
+      "learning_rate": 6.5931154391368266e-06,
+      "loss": 0.0134,
+      "step": 3456
+    },
+    {
+      "epoch": 246.92857142857142,
+      "grad_norm": 0.1171875,
+      "learning_rate": 6.585426055759588e-06,
+      "loss": 0.0117,
+      "step": 3457
+    },
+    {
+      "epoch": 247.0,
+      "grad_norm": 0.1767578125,
+      "learning_rate": 6.577739555073839e-06,
+      "loss": 0.0181,
+      "step": 3458
+    },
+    {
+      "epoch": 247.07142857142858,
+      "grad_norm": 0.21484375,
+      "learning_rate": 6.570055940825887e-06,
+      "loss": 0.0161,
+      "step": 3459
+    },
+    {
+      "epoch": 247.14285714285714,
+      "grad_norm": 0.1865234375,
+      "learning_rate": 6.5623752167606304e-06,
+      "loss": 0.0124,
+      "step": 3460
+    },
+    {
+      "epoch": 247.21428571428572,
+      "grad_norm": 0.171875,
+      "learning_rate": 6.5546973866215584e-06,
+      "loss": 0.0141,
+      "step": 3461
+    },
+    {
+      "epoch": 247.28571428571428,
+      "grad_norm": 0.1689453125,
+      "learning_rate": 6.54702245415076e-06,
+      "loss": 0.0138,
+      "step": 3462
+    },
+    {
+      "epoch": 247.35714285714286,
+      "grad_norm": 0.169921875,
+      "learning_rate": 6.539350423088899e-06,
+      "loss": 0.0138,
+      "step": 3463
+    },
+    {
+      "epoch": 247.42857142857142,
+      "grad_norm": 0.1513671875,
+      "learning_rate": 6.531681297175232e-06,
+      "loss": 0.0127,
+      "step": 3464
+    },
+    {
+      "epoch": 247.5,
+      "grad_norm": 0.12109375,
+      "learning_rate": 6.524015080147598e-06,
+      "loss": 0.0133,
+      "step": 3465
+    },
+    {
+      "epoch": 247.57142857142858,
+      "grad_norm": 0.130859375,
+      "learning_rate": 6.516351775742416e-06,
+      "loss": 0.0119,
+      "step": 3466
+    },
+    {
+      "epoch": 247.64285714285714,
+      "grad_norm": 0.138671875,
+      "learning_rate": 6.5086913876946875e-06,
+      "loss": 0.0131,
+      "step": 3467
+    },
+    {
+      "epoch": 247.71428571428572,
+      "grad_norm": 0.328125,
+      "learning_rate": 6.501033919737995e-06,
+      "loss": 0.0155,
+      "step": 3468
+    },
+    {
+      "epoch": 247.78571428571428,
+      "grad_norm": 0.16015625,
+      "learning_rate": 6.493379375604492e-06,
+      "loss": 0.0149,
+      "step": 3469
+    },
+    {
+      "epoch": 247.85714285714286,
+      "grad_norm": 0.39453125,
+      "learning_rate": 6.485727759024914e-06,
+      "loss": 0.011,
+      "step": 3470
+    },
+    {
+      "epoch": 247.92857142857142,
+      "grad_norm": 0.171875,
+      "learning_rate": 6.478079073728557e-06,
+      "loss": 0.0103,
+      "step": 3471
+    },
+    {
+      "epoch": 248.0,
+      "grad_norm": 0.134765625,
+      "learning_rate": 6.4704333234433106e-06,
+      "loss": 0.0139,
+      "step": 3472
+    },
+    {
+      "epoch": 248.07142857142858,
+      "grad_norm": 0.25,
+      "learning_rate": 6.46279051189562e-06,
+      "loss": 0.0141,
+      "step": 3473
+    },
+    {
+      "epoch": 248.14285714285714,
+      "grad_norm": 0.11328125,
+      "learning_rate": 6.455150642810487e-06,
+      "loss": 0.0133,
+      "step": 3474
+    },
+    {
+      "epoch": 248.21428571428572,
+      "grad_norm": 0.189453125,
+      "learning_rate": 6.447513719911491e-06,
+      "loss": 0.0107,
+      "step": 3475
+    },
+    {
+      "epoch": 248.28571428571428,
+      "grad_norm": 0.12109375,
+      "learning_rate": 6.4398797469207875e-06,
+      "loss": 0.0136,
+      "step": 3476
+    },
+    {
+      "epoch": 248.35714285714286,
+      "grad_norm": 0.1240234375,
+      "learning_rate": 6.432248727559076e-06,
+      "loss": 0.0128,
+      "step": 3477
+    },
+    {
+      "epoch": 248.42857142857142,
+      "grad_norm": 0.10107421875,
+      "learning_rate": 6.424620665545623e-06,
+      "loss": 0.0111,
+      "step": 3478
+    },
+    {
+      "epoch": 248.5,
+      "grad_norm": 0.2216796875,
+      "learning_rate": 6.416995564598254e-06,
+      "loss": 0.0139,
+      "step": 3479
+    },
+    {
+      "epoch": 248.57142857142858,
+      "grad_norm": 0.1435546875,
+      "learning_rate": 6.4093734284333435e-06,
+      "loss": 0.0153,
+      "step": 3480
+    },
+    {
+      "epoch": 248.64285714285714,
+      "grad_norm": 0.11767578125,
+      "learning_rate": 6.40175426076584e-06,
+      "loss": 0.0144,
+      "step": 3481
+    },
+    {
+      "epoch": 248.71428571428572,
+      "grad_norm": 0.1376953125,
+      "learning_rate": 6.394138065309233e-06,
+      "loss": 0.0151,
+      "step": 3482
+    },
+    {
+      "epoch": 248.78571428571428,
+      "grad_norm": 0.09619140625,
+      "learning_rate": 6.386524845775552e-06,
+      "loss": 0.0105,
+      "step": 3483
+    },
+    {
+      "epoch": 248.85714285714286,
+      "grad_norm": 0.1552734375,
+      "learning_rate": 6.378914605875393e-06,
+      "loss": 0.0138,
+      "step": 3484
+    },
+    {
+      "epoch": 248.92857142857142,
+      "grad_norm": 0.166015625,
+      "learning_rate": 6.371307349317891e-06,
+      "loss": 0.0143,
+      "step": 3485
+    },
+    {
+      "epoch": 249.0,
+      "grad_norm": 0.1533203125,
+      "learning_rate": 6.36370307981074e-06,
+      "loss": 0.0114,
+      "step": 3486
+    },
+    {
+      "epoch": 249.07142857142858,
+      "grad_norm": 0.1318359375,
+      "learning_rate": 6.356101801060163e-06,
+      "loss": 0.0126,
+      "step": 3487
+    },
+    {
+      "epoch": 249.14285714285714,
+      "grad_norm": 0.1474609375,
+      "learning_rate": 6.348503516770932e-06,
+      "loss": 0.0141,
+      "step": 3488
+    },
+    {
+      "epoch": 249.21428571428572,
+      "grad_norm": 0.146484375,
+      "learning_rate": 6.3409082306463574e-06,
+      "loss": 0.0123,
+      "step": 3489
+    },
+    {
+      "epoch": 249.28571428571428,
+      "grad_norm": 0.1142578125,
+      "learning_rate": 6.333315946388285e-06,
+      "loss": 0.015,
+      "step": 3490
+    },
+    {
+      "epoch": 249.35714285714286,
+      "grad_norm": 0.25,
+      "learning_rate": 6.3257266676971154e-06,
+      "loss": 0.0148,
+      "step": 3491
+    },
+    {
+      "epoch": 249.42857142857142,
+      "grad_norm": 0.134765625,
+      "learning_rate": 6.318140398271758e-06,
+      "loss": 0.0116,
+      "step": 3492
+    },
+    {
+      "epoch": 249.5,
+      "grad_norm": 0.1650390625,
+      "learning_rate": 6.3105571418096725e-06,
+      "loss": 0.0135,
+      "step": 3493
+    },
+    {
+      "epoch": 249.57142857142858,
+      "grad_norm": 0.21875,
+      "learning_rate": 6.3029769020068405e-06,
+      "loss": 0.0145,
+      "step": 3494
+    },
+    {
+      "epoch": 249.64285714285714,
+      "grad_norm": 0.1337890625,
+      "learning_rate": 6.29539968255779e-06,
+      "loss": 0.0134,
+      "step": 3495
+    },
+    {
+      "epoch": 249.71428571428572,
+      "grad_norm": 0.2890625,
+      "learning_rate": 6.287825487155559e-06,
+      "loss": 0.0156,
+      "step": 3496
+    },
+    {
+      "epoch": 249.78571428571428,
+      "grad_norm": 0.2451171875,
+      "learning_rate": 6.2802543194917185e-06,
+      "loss": 0.0125,
+      "step": 3497
+    },
+    {
+      "epoch": 249.85714285714286,
+      "grad_norm": 0.119140625,
+      "learning_rate": 6.272686183256365e-06,
+      "loss": 0.0139,
+      "step": 3498
+    },
+    {
+      "epoch": 249.92857142857142,
+      "grad_norm": 0.158203125,
+      "learning_rate": 6.265121082138118e-06,
+      "loss": 0.0121,
+      "step": 3499
+    },
+    {
+      "epoch": 250.0,
+      "grad_norm": 0.11865234375,
+      "learning_rate": 6.257559019824111e-06,
+      "loss": 0.0135,
+      "step": 3500
+    },
+    {
+      "epoch": 250.07142857142858,
+      "grad_norm": 0.130859375,
+      "learning_rate": 6.250000000000003e-06,
+      "loss": 0.0123,
+      "step": 3501
+    },
+    {
+      "epoch": 250.14285714285714,
+      "grad_norm": 0.12890625,
+      "learning_rate": 6.2424440263499694e-06,
+      "loss": 0.0141,
+      "step": 3502
+    },
+    {
+      "epoch": 250.21428571428572,
+      "grad_norm": 0.17578125,
+      "learning_rate": 6.2348911025567e-06,
+      "loss": 0.0121,
+      "step": 3503
+    },
+    {
+      "epoch": 250.28571428571428,
+      "grad_norm": 0.2138671875,
+      "learning_rate": 6.227341232301394e-06,
+      "loss": 0.0156,
+      "step": 3504
+    },
+    {
+      "epoch": 250.35714285714286,
+      "grad_norm": 0.1923828125,
+      "learning_rate": 6.219794419263775e-06,
+      "loss": 0.0136,
+      "step": 3505
+    },
+    {
+      "epoch": 250.42857142857142,
+      "grad_norm": 0.177734375,
+      "learning_rate": 6.212250667122061e-06,
+      "loss": 0.0116,
+      "step": 3506
+    },
+    {
+      "epoch": 250.5,
+      "grad_norm": 0.1455078125,
+      "learning_rate": 6.2047099795529905e-06,
+      "loss": 0.0144,
+      "step": 3507
+    },
+    {
+      "epoch": 250.57142857142858,
+      "grad_norm": 0.2333984375,
+      "learning_rate": 6.1971723602317984e-06,
+      "loss": 0.0139,
+      "step": 3508
+    },
+    {
+      "epoch": 250.64285714285714,
+      "grad_norm": 0.1416015625,
+      "learning_rate": 6.1896378128322306e-06,
+      "loss": 0.0153,
+      "step": 3509
+    },
+    {
+      "epoch": 250.71428571428572,
+      "grad_norm": 0.1142578125,
+      "learning_rate": 6.182106341026534e-06,
+      "loss": 0.0135,
+      "step": 3510
+    },
+    {
+      "epoch": 250.78571428571428,
+      "grad_norm": 0.1533203125,
+      "learning_rate": 6.174577948485455e-06,
+      "loss": 0.0112,
+      "step": 3511
+    },
+    {
+      "epoch": 250.85714285714286,
+      "grad_norm": 0.1982421875,
+      "learning_rate": 6.167052638878243e-06,
+      "loss": 0.0119,
+      "step": 3512
+    },
+    {
+      "epoch": 250.92857142857142,
+      "grad_norm": 0.271484375,
+      "learning_rate": 6.159530415872636e-06,
+      "loss": 0.012,
+      "step": 3513
+    },
+    {
+      "epoch": 251.0,
+      "grad_norm": 0.1376953125,
+      "learning_rate": 6.152011283134884e-06,
+      "loss": 0.0117,
+      "step": 3514
+    },
+    {
+      "epoch": 251.07142857142858,
+      "grad_norm": 0.1240234375,
+      "learning_rate": 6.144495244329716e-06,
+      "loss": 0.0164,
+      "step": 3515
+    },
+    {
+      "epoch": 251.14285714285714,
+      "grad_norm": 0.1455078125,
+      "learning_rate": 6.136982303120359e-06,
+      "loss": 0.0112,
+      "step": 3516
+    },
+    {
+      "epoch": 251.21428571428572,
+      "grad_norm": 0.205078125,
+      "learning_rate": 6.129472463168528e-06,
+      "loss": 0.0144,
+      "step": 3517
+    },
+    {
+      "epoch": 251.28571428571428,
+      "grad_norm": 0.11669921875,
+      "learning_rate": 6.121965728134429e-06,
+      "loss": 0.0129,
+      "step": 3518
+    },
+    {
+      "epoch": 251.35714285714286,
+      "grad_norm": 0.12255859375,
+      "learning_rate": 6.114462101676753e-06,
+      "loss": 0.0119,
+      "step": 3519
+    },
+    {
+      "epoch": 251.42857142857142,
+      "grad_norm": 0.2099609375,
+      "learning_rate": 6.106961587452677e-06,
+      "loss": 0.0139,
+      "step": 3520
+    },
+    {
+      "epoch": 251.5,
+      "grad_norm": 0.23828125,
+      "learning_rate": 6.099464189117858e-06,
+      "loss": 0.011,
+      "step": 3521
+    },
+    {
+      "epoch": 251.57142857142858,
+      "grad_norm": 0.1494140625,
+      "learning_rate": 6.0919699103264395e-06,
+      "loss": 0.014,
+      "step": 3522
+    },
+    {
+      "epoch": 251.64285714285714,
+      "grad_norm": 0.169921875,
+      "learning_rate": 6.084478754731037e-06,
+      "loss": 0.0125,
+      "step": 3523
+    },
+    {
+      "epoch": 251.71428571428572,
+      "grad_norm": 0.1513671875,
+      "learning_rate": 6.076990725982756e-06,
+      "loss": 0.0129,
+      "step": 3524
+    },
+    {
+      "epoch": 251.78571428571428,
+      "grad_norm": 0.205078125,
+      "learning_rate": 6.069505827731168e-06,
+      "loss": 0.0157,
+      "step": 3525
+    },
+    {
+      "epoch": 251.85714285714286,
+      "grad_norm": 0.2021484375,
+      "learning_rate": 6.062024063624322e-06,
+      "loss": 0.0144,
+      "step": 3526
+    },
+    {
+      "epoch": 251.92857142857142,
+      "grad_norm": 0.1474609375,
+      "learning_rate": 6.054545437308738e-06,
+      "loss": 0.014,
+      "step": 3527
+    },
+    {
+      "epoch": 252.0,
+      "grad_norm": 0.1337890625,
+      "learning_rate": 6.047069952429407e-06,
+      "loss": 0.0121,
+      "step": 3528
+    },
+    {
+      "epoch": 252.07142857142858,
+      "grad_norm": 0.140625,
+      "learning_rate": 6.039597612629789e-06,
+      "loss": 0.0137,
+      "step": 3529
+    },
+    {
+      "epoch": 252.14285714285714,
+      "grad_norm": 0.18359375,
+      "learning_rate": 6.032128421551815e-06,
+      "loss": 0.0164,
+      "step": 3530
+    },
+    {
+      "epoch": 252.21428571428572,
+      "grad_norm": 0.1953125,
+      "learning_rate": 6.024662382835874e-06,
+      "loss": 0.013,
+      "step": 3531
+    },
+    {
+      "epoch": 252.28571428571428,
+      "grad_norm": 0.10791015625,
+      "learning_rate": 6.017199500120827e-06,
+      "loss": 0.0128,
+      "step": 3532
+    },
+    {
+      "epoch": 252.35714285714286,
+      "grad_norm": 0.19140625,
+      "learning_rate": 6.009739777043982e-06,
+      "loss": 0.0114,
+      "step": 3533
+    },
+    {
+      "epoch": 252.42857142857142,
+      "grad_norm": 0.125,
+      "learning_rate": 6.0022832172411305e-06,
+      "loss": 0.0128,
+      "step": 3534
+    },
+    {
+      "epoch": 252.5,
+      "grad_norm": 0.1962890625,
+      "learning_rate": 5.994829824346508e-06,
+      "loss": 0.0129,
+      "step": 3535
+    },
+    {
+      "epoch": 252.57142857142858,
+      "grad_norm": 0.1845703125,
+      "learning_rate": 5.9873796019928e-06,
+      "loss": 0.0151,
+      "step": 3536
+    },
+    {
+      "epoch": 252.64285714285714,
+      "grad_norm": 0.1806640625,
+      "learning_rate": 5.979932553811153e-06,
+      "loss": 0.0154,
+      "step": 3537
+    },
+    {
+      "epoch": 252.71428571428572,
+      "grad_norm": 0.1298828125,
+      "learning_rate": 5.972488683431177e-06,
+      "loss": 0.0127,
+      "step": 3538
+    },
+    {
+      "epoch": 252.78571428571428,
+      "grad_norm": 0.26171875,
+      "learning_rate": 5.965047994480918e-06,
+      "loss": 0.015,
+      "step": 3539
+    },
+    {
+      "epoch": 252.85714285714286,
+      "grad_norm": 0.236328125,
+      "learning_rate": 5.957610490586882e-06,
+      "loss": 0.015,
+      "step": 3540
+    },
+    {
+      "epoch": 252.92857142857142,
+      "grad_norm": 0.11083984375,
+      "learning_rate": 5.950176175374013e-06,
+      "loss": 0.0122,
+      "step": 3541
+    },
+    {
+      "epoch": 253.0,
+      "grad_norm": 0.1787109375,
+      "learning_rate": 5.942745052465703e-06,
+      "loss": 0.015,
+      "step": 3542
+    },
+    {
+      "epoch": 253.07142857142858,
+      "grad_norm": 0.1748046875,
+      "learning_rate": 5.935317125483802e-06,
+      "loss": 0.015,
+      "step": 3543
+    },
+    {
+      "epoch": 253.14285714285714,
+      "grad_norm": 0.1376953125,
+      "learning_rate": 5.92789239804859e-06,
+      "loss": 0.0155,
+      "step": 3544
+    },
+    {
+      "epoch": 253.21428571428572,
+      "grad_norm": 0.138671875,
+      "learning_rate": 5.920470873778779e-06,
+      "loss": 0.0134,
+      "step": 3545
+    },
+    {
+      "epoch": 253.28571428571428,
+      "grad_norm": 0.1162109375,
+      "learning_rate": 5.9130525562915306e-06,
+      "loss": 0.0126,
+      "step": 3546
+    },
+    {
+      "epoch": 253.35714285714286,
+      "grad_norm": 0.1796875,
+      "learning_rate": 5.9056374492024515e-06,
+      "loss": 0.0133,
+      "step": 3547
+    },
+    {
+      "epoch": 253.42857142857142,
+      "grad_norm": 0.125,
+      "learning_rate": 5.898225556125571e-06,
+      "loss": 0.0145,
+      "step": 3548
+    },
+    {
+      "epoch": 253.5,
+      "grad_norm": 0.2109375,
+      "learning_rate": 5.890816880673354e-06,
+      "loss": 0.0132,
+      "step": 3549
+    },
+    {
+      "epoch": 253.57142857142858,
+      "grad_norm": 0.1123046875,
+      "learning_rate": 5.8834114264567015e-06,
+      "loss": 0.0134,
+      "step": 3550
+    },
+    {
+      "epoch": 253.64285714285714,
+      "grad_norm": 0.294921875,
+      "learning_rate": 5.87600919708494e-06,
+      "loss": 0.0122,
+      "step": 3551
+    },
+    {
+      "epoch": 253.71428571428572,
+      "grad_norm": 0.1318359375,
+      "learning_rate": 5.868610196165822e-06,
+      "loss": 0.0142,
+      "step": 3552
+    },
+    {
+      "epoch": 253.78571428571428,
+      "grad_norm": 0.1494140625,
+      "learning_rate": 5.861214427305545e-06,
+      "loss": 0.0157,
+      "step": 3553
+    },
+    {
+      "epoch": 253.85714285714286,
+      "grad_norm": 0.12890625,
+      "learning_rate": 5.853821894108702e-06,
+      "loss": 0.0137,
+      "step": 3554
+    },
+    {
+      "epoch": 253.92857142857142,
+      "grad_norm": 0.0908203125,
+      "learning_rate": 5.846432600178329e-06,
+      "loss": 0.0121,
+      "step": 3555
+    },
+    {
+      "epoch": 254.0,
+      "grad_norm": 0.138671875,
+      "learning_rate": 5.839046549115875e-06,
+      "loss": 0.0117,
+      "step": 3556
+    },
+    {
+      "epoch": 254.07142857142858,
+      "grad_norm": 0.203125,
+      "learning_rate": 5.831663744521218e-06,
+      "loss": 0.0111,
+      "step": 3557
+    },
+    {
+      "epoch": 254.14285714285714,
+      "grad_norm": 0.1220703125,
+      "learning_rate": 5.824284189992644e-06,
+      "loss": 0.0142,
+      "step": 3558
+    },
+    {
+      "epoch": 254.21428571428572,
+      "grad_norm": 0.17578125,
+      "learning_rate": 5.81690788912686e-06,
+      "loss": 0.0132,
+      "step": 3559
+    },
+    {
+      "epoch": 254.28571428571428,
+      "grad_norm": 0.1474609375,
+      "learning_rate": 5.809534845518982e-06,
+      "loss": 0.0127,
+      "step": 3560
+    },
+    {
+      "epoch": 254.35714285714286,
+      "grad_norm": 0.1337890625,
+      "learning_rate": 5.8021650627625395e-06,
+      "loss": 0.0122,
+      "step": 3561
+    },
+    {
+      "epoch": 254.42857142857142,
+      "grad_norm": 0.15625,
+      "learning_rate": 5.794798544449488e-06,
+      "loss": 0.0145,
+      "step": 3562
+    },
+    {
+      "epoch": 254.5,
+      "grad_norm": 0.173828125,
+      "learning_rate": 5.7874352941701666e-06,
+      "loss": 0.0152,
+      "step": 3563
+    },
+    {
+      "epoch": 254.57142857142858,
+      "grad_norm": 0.36328125,
+      "learning_rate": 5.780075315513338e-06,
+      "loss": 0.0151,
+      "step": 3564
+    },
+    {
+      "epoch": 254.64285714285714,
+      "grad_norm": 0.14453125,
+      "learning_rate": 5.772718612066159e-06,
+      "loss": 0.0157,
+      "step": 3565
+    },
+    {
+      "epoch": 254.71428571428572,
+      "grad_norm": 0.162109375,
+      "learning_rate": 5.765365187414211e-06,
+      "loss": 0.0128,
+      "step": 3566
+    },
+    {
+      "epoch": 254.78571428571428,
+      "grad_norm": 0.140625,
+      "learning_rate": 5.758015045141457e-06,
+      "loss": 0.0137,
+      "step": 3567
+    },
+    {
+      "epoch": 254.85714285714286,
+      "grad_norm": 0.1357421875,
+      "learning_rate": 5.7506681888302685e-06,
+      "loss": 0.0134,
+      "step": 3568
+    },
+    {
+      "epoch": 254.92857142857142,
+      "grad_norm": 0.1181640625,
+      "learning_rate": 5.743324622061412e-06,
+      "loss": 0.0139,
+      "step": 3569
+    },
+    {
+      "epoch": 255.0,
+      "grad_norm": 0.2578125,
+      "learning_rate": 5.735984348414053e-06,
+      "loss": 0.012,
+      "step": 3570
+    },
+    {
+      "epoch": 255.07142857142858,
+      "grad_norm": 0.1904296875,
+      "learning_rate": 5.7286473714657534e-06,
+      "loss": 0.0129,
+      "step": 3571
+    },
+    {
+      "epoch": 255.14285714285714,
+      "grad_norm": 0.10888671875,
+      "learning_rate": 5.721313694792466e-06,
+      "loss": 0.0133,
+      "step": 3572
+    },
+    {
+      "epoch": 255.21428571428572,
+      "grad_norm": 0.1455078125,
+      "learning_rate": 5.7139833219685345e-06,
+      "loss": 0.0146,
+      "step": 3573
+    },
+    {
+      "epoch": 255.28571428571428,
+      "grad_norm": 0.203125,
+      "learning_rate": 5.706656256566695e-06,
+      "loss": 0.013,
+      "step": 3574
+    },
+    {
+      "epoch": 255.35714285714286,
+      "grad_norm": 0.10302734375,
+      "learning_rate": 5.699332502158064e-06,
+      "loss": 0.0125,
+      "step": 3575
+    },
+    {
+      "epoch": 255.42857142857142,
+      "grad_norm": 0.12353515625,
+      "learning_rate": 5.692012062312162e-06,
+      "loss": 0.0144,
+      "step": 3576
+    },
+    {
+      "epoch": 255.5,
+      "grad_norm": 0.11962890625,
+      "learning_rate": 5.684694940596874e-06,
+      "loss": 0.0136,
+      "step": 3577
+    },
+    {
+      "epoch": 255.57142857142858,
+      "grad_norm": 0.125,
+      "learning_rate": 5.67738114057848e-06,
+      "loss": 0.0114,
+      "step": 3578
+    },
+    {
+      "epoch": 255.64285714285714,
+      "grad_norm": 0.1943359375,
+      "learning_rate": 5.670070665821638e-06,
+      "loss": 0.0139,
+      "step": 3579
+    },
+    {
+      "epoch": 255.71428571428572,
+      "grad_norm": 0.1435546875,
+      "learning_rate": 5.66276351988938e-06,
+      "loss": 0.0125,
+      "step": 3580
+    },
+    {
+      "epoch": 255.78571428571428,
+      "grad_norm": 0.12109375,
+      "learning_rate": 5.655459706343123e-06,
+      "loss": 0.0122,
+      "step": 3581
+    },
+    {
+      "epoch": 255.85714285714286,
+      "grad_norm": 0.1357421875,
+      "learning_rate": 5.6481592287426585e-06,
+      "loss": 0.0134,
+      "step": 3582
+    },
+    {
+      "epoch": 255.92857142857142,
+      "grad_norm": 0.1279296875,
+      "learning_rate": 5.640862090646147e-06,
+      "loss": 0.0114,
+      "step": 3583
+    },
+    {
+      "epoch": 256.0,
+      "grad_norm": 0.1552734375,
+      "learning_rate": 5.6335682956101284e-06,
+      "loss": 0.0137,
+      "step": 3584
+    },
+    {
+      "epoch": 256.07142857142856,
+      "grad_norm": 0.10888671875,
+      "learning_rate": 5.626277847189503e-06,
+      "loss": 0.0134,
+      "step": 3585
+    },
+    {
+      "epoch": 256.14285714285717,
+      "grad_norm": 0.150390625,
+      "learning_rate": 5.618990748937555e-06,
+      "loss": 0.0115,
+      "step": 3586
+    },
+    {
+      "epoch": 256.2142857142857,
+      "grad_norm": 0.197265625,
+      "learning_rate": 5.611707004405925e-06,
+      "loss": 0.0126,
+      "step": 3587
+    },
+    {
+      "epoch": 256.2857142857143,
+      "grad_norm": 0.140625,
+      "learning_rate": 5.604426617144623e-06,
+      "loss": 0.0149,
+      "step": 3588
+    },
+    {
+      "epoch": 256.35714285714283,
+      "grad_norm": 0.125,
+      "learning_rate": 5.597149590702018e-06,
+      "loss": 0.0142,
+      "step": 3589
+    },
+    {
+      "epoch": 256.42857142857144,
+      "grad_norm": 0.12451171875,
+      "learning_rate": 5.589875928624845e-06,
+      "loss": 0.0124,
+      "step": 3590
+    },
+    {
+      "epoch": 256.5,
+      "grad_norm": 0.12890625,
+      "learning_rate": 5.5826056344582e-06,
+      "loss": 0.013,
+      "step": 3591
+    },
+    {
+      "epoch": 256.57142857142856,
+      "grad_norm": 0.1474609375,
+      "learning_rate": 5.5753387117455365e-06,
+      "loss": 0.015,
+      "step": 3592
+    },
+    {
+      "epoch": 256.64285714285717,
+      "grad_norm": 0.1298828125,
+      "learning_rate": 5.568075164028662e-06,
+      "loss": 0.0122,
+      "step": 3593
+    },
+    {
+      "epoch": 256.7142857142857,
+      "grad_norm": 0.1171875,
+      "learning_rate": 5.560814994847738e-06,
+      "loss": 0.014,
+      "step": 3594
+    },
+    {
+      "epoch": 256.7857142857143,
+      "grad_norm": 0.1279296875,
+      "learning_rate": 5.553558207741293e-06,
+      "loss": 0.012,
+      "step": 3595
+    },
+    {
+      "epoch": 256.85714285714283,
+      "grad_norm": 0.2119140625,
+      "learning_rate": 5.54630480624619e-06,
+      "loss": 0.0153,
+      "step": 3596
+    },
+    {
+      "epoch": 256.92857142857144,
+      "grad_norm": 0.126953125,
+      "learning_rate": 5.539054793897655e-06,
+      "loss": 0.0135,
+      "step": 3597
+    },
+    {
+      "epoch": 257.0,
+      "grad_norm": 0.380859375,
+      "learning_rate": 5.531808174229241e-06,
+      "loss": 0.016,
+      "step": 3598
+    },
+    {
+      "epoch": 257.07142857142856,
+      "grad_norm": 0.1474609375,
+      "learning_rate": 5.5245649507728775e-06,
+      "loss": 0.0133,
+      "step": 3599
+    },
+    {
+      "epoch": 257.14285714285717,
+      "grad_norm": 0.130859375,
+      "learning_rate": 5.517325127058819e-06,
+      "loss": 0.0126,
+      "step": 3600
+    },
+    {
+      "epoch": 257.2142857142857,
+      "grad_norm": 0.130859375,
+      "learning_rate": 5.510088706615667e-06,
+      "loss": 0.0154,
+      "step": 3601
+    },
+    {
+      "epoch": 257.2857142857143,
+      "grad_norm": 0.2265625,
+      "learning_rate": 5.502855692970364e-06,
+      "loss": 0.0139,
+      "step": 3602
+    },
+    {
+      "epoch": 257.35714285714283,
+      "grad_norm": 0.271484375,
+      "learning_rate": 5.495626089648194e-06,
+      "loss": 0.0108,
+      "step": 3603
+    },
+    {
+      "epoch": 257.42857142857144,
+      "grad_norm": 0.11181640625,
+      "learning_rate": 5.488399900172775e-06,
+      "loss": 0.0119,
+      "step": 3604
+    },
+    {
+      "epoch": 257.5,
+      "grad_norm": 0.1572265625,
+      "learning_rate": 5.4811771280660704e-06,
+      "loss": 0.0109,
+      "step": 3605
+    },
+    {
+      "epoch": 257.57142857142856,
+      "grad_norm": 0.11572265625,
+      "learning_rate": 5.473957776848371e-06,
+      "loss": 0.0124,
+      "step": 3606
+    },
+    {
+      "epoch": 257.64285714285717,
+      "grad_norm": 0.12158203125,
+      "learning_rate": 5.466741850038296e-06,
+      "loss": 0.013,
+      "step": 3607
+    },
+    {
+      "epoch": 257.7142857142857,
+      "grad_norm": 0.1259765625,
+      "learning_rate": 5.459529351152795e-06,
+      "loss": 0.0123,
+      "step": 3608
+    },
+    {
+      "epoch": 257.7857142857143,
+      "grad_norm": 0.150390625,
+      "learning_rate": 5.452320283707166e-06,
+      "loss": 0.0148,
+      "step": 3609
+    },
+    {
+      "epoch": 257.85714285714283,
+      "grad_norm": 0.1591796875,
+      "learning_rate": 5.445114651215014e-06,
+      "loss": 0.0149,
+      "step": 3610
+    },
+    {
+      "epoch": 257.92857142857144,
+      "grad_norm": 0.1484375,
+      "learning_rate": 5.437912457188277e-06,
+      "loss": 0.0121,
+      "step": 3611
+    },
+    {
+      "epoch": 258.0,
+      "grad_norm": 0.1396484375,
+      "learning_rate": 5.430713705137218e-06,
+      "loss": 0.0149,
+      "step": 3612
+    },
+    {
+      "epoch": 258.07142857142856,
+      "grad_norm": 0.2265625,
+      "learning_rate": 5.4235183985704156e-06,
+      "loss": 0.011,
+      "step": 3613
+    },
+    {
+      "epoch": 258.14285714285717,
+      "grad_norm": 0.162109375,
+      "learning_rate": 5.416326540994783e-06,
+      "loss": 0.0131,
+      "step": 3614
+    },
+    {
+      "epoch": 258.2142857142857,
+      "grad_norm": 0.25390625,
+      "learning_rate": 5.4091381359155464e-06,
+      "loss": 0.0119,
+      "step": 3615
+    },
+    {
+      "epoch": 258.2857142857143,
+      "grad_norm": 0.185546875,
+      "learning_rate": 5.401953186836238e-06,
+      "loss": 0.0134,
+      "step": 3616
+    },
+    {
+      "epoch": 258.35714285714283,
+      "grad_norm": 0.26171875,
+      "learning_rate": 5.3947716972587175e-06,
+      "loss": 0.0101,
+      "step": 3617
+    },
+    {
+      "epoch": 258.42857142857144,
+      "grad_norm": 0.130859375,
+      "learning_rate": 5.3875936706831516e-06,
+      "loss": 0.0134,
+      "step": 3618
+    },
+    {
+      "epoch": 258.5,
+      "grad_norm": 0.1630859375,
+      "learning_rate": 5.380419110608033e-06,
+      "loss": 0.0165,
+      "step": 3619
+    },
+    {
+      "epoch": 258.57142857142856,
+      "grad_norm": 0.16015625,
+      "learning_rate": 5.37324802053015e-06,
+      "loss": 0.0132,
+      "step": 3620
+    },
+    {
+      "epoch": 258.64285714285717,
+      "grad_norm": 0.1318359375,
+      "learning_rate": 5.3660804039446056e-06,
+      "loss": 0.0143,
+      "step": 3621
+    },
+    {
+      "epoch": 258.7142857142857,
+      "grad_norm": 0.126953125,
+      "learning_rate": 5.358916264344809e-06,
+      "loss": 0.0128,
+      "step": 3622
+    },
+    {
+      "epoch": 258.7857142857143,
+      "grad_norm": 0.12060546875,
+      "learning_rate": 5.351755605222468e-06,
+      "loss": 0.0111,
+      "step": 3623
+    },
+    {
+      "epoch": 258.85714285714283,
+      "grad_norm": 0.14453125,
+      "learning_rate": 5.344598430067617e-06,
+      "loss": 0.0143,
+      "step": 3624
+    },
+    {
+      "epoch": 258.92857142857144,
+      "grad_norm": 0.1240234375,
+      "learning_rate": 5.3374447423685625e-06,
+      "loss": 0.0126,
+      "step": 3625
+    },
+    {
+      "epoch": 259.0,
+      "grad_norm": 0.1494140625,
+      "learning_rate": 5.3302945456119276e-06,
+      "loss": 0.0167,
+      "step": 3626
+    },
+    {
+      "epoch": 259.07142857142856,
+      "grad_norm": 0.126953125,
+      "learning_rate": 5.323147843282629e-06,
+      "loss": 0.0149,
+      "step": 3627
+    },
+    {
+      "epoch": 259.14285714285717,
+      "grad_norm": 0.166015625,
+      "learning_rate": 5.31600463886389e-06,
+      "loss": 0.0121,
+      "step": 3628
+    },
+    {
+      "epoch": 259.2142857142857,
+      "grad_norm": 0.1708984375,
+      "learning_rate": 5.308864935837218e-06,
+      "loss": 0.0137,
+      "step": 3629
+    },
+    {
+      "epoch": 259.2857142857143,
+      "grad_norm": 0.119140625,
+      "learning_rate": 5.301728737682418e-06,
+      "loss": 0.011,
+      "step": 3630
+    },
+    {
+      "epoch": 259.35714285714283,
+      "grad_norm": 0.140625,
+      "learning_rate": 5.294596047877585e-06,
+      "loss": 0.0112,
+      "step": 3631
+    },
+    {
+      "epoch": 259.42857142857144,
+      "grad_norm": 0.1318359375,
+      "learning_rate": 5.2874668698991075e-06,
+      "loss": 0.0108,
+      "step": 3632
+    },
+    {
+      "epoch": 259.5,
+      "grad_norm": 0.11767578125,
+      "learning_rate": 5.280341207221658e-06,
+      "loss": 0.0118,
+      "step": 3633
+    },
+    {
+      "epoch": 259.57142857142856,
+      "grad_norm": 0.12255859375,
+      "learning_rate": 5.273219063318198e-06,
+      "loss": 0.0118,
+      "step": 3634
+    },
+    {
+      "epoch": 259.64285714285717,
+      "grad_norm": 0.1328125,
+      "learning_rate": 5.266100441659974e-06,
+      "loss": 0.0113,
+      "step": 3635
+    },
+    {
+      "epoch": 259.7142857142857,
+      "grad_norm": 0.134765625,
+      "learning_rate": 5.2589853457165175e-06,
+      "loss": 0.0118,
+      "step": 3636
+    },
+    {
+      "epoch": 259.7857142857143,
+      "grad_norm": 0.140625,
+      "learning_rate": 5.251873778955635e-06,
+      "loss": 0.0108,
+      "step": 3637
+    },
+    {
+      "epoch": 259.85714285714283,
+      "grad_norm": 0.197265625,
+      "learning_rate": 5.244765744843424e-06,
+      "loss": 0.0174,
+      "step": 3638
+    },
+    {
+      "epoch": 259.92857142857144,
+      "grad_norm": 0.22265625,
+      "learning_rate": 5.237661246844252e-06,
+      "loss": 0.0141,
+      "step": 3639
+    },
+    {
+      "epoch": 260.0,
+      "grad_norm": 0.123046875,
+      "learning_rate": 5.2305602884207665e-06,
+      "loss": 0.0148,
+      "step": 3640
+    },
+    {
+      "epoch": 260.07142857142856,
+      "grad_norm": 0.1767578125,
+      "learning_rate": 5.223462873033884e-06,
+      "loss": 0.0137,
+      "step": 3641
+    },
+    {
+      "epoch": 260.14285714285717,
+      "grad_norm": 0.1416015625,
+      "learning_rate": 5.216369004142802e-06,
+      "loss": 0.0136,
+      "step": 3642
+    },
+    {
+      "epoch": 260.2142857142857,
+      "grad_norm": 0.162109375,
+      "learning_rate": 5.209278685204983e-06,
+      "loss": 0.0149,
+      "step": 3643
+    },
+    {
+      "epoch": 260.2857142857143,
+      "grad_norm": 0.1953125,
+      "learning_rate": 5.202191919676165e-06,
+      "loss": 0.0153,
+      "step": 3644
+    },
+    {
+      "epoch": 260.35714285714283,
+      "grad_norm": 0.125,
+      "learning_rate": 5.19510871101035e-06,
+      "loss": 0.0145,
+      "step": 3645
+    },
+    {
+      "epoch": 260.42857142857144,
+      "grad_norm": 0.162109375,
+      "learning_rate": 5.188029062659806e-06,
+      "loss": 0.0124,
+      "step": 3646
+    },
+    {
+      "epoch": 260.5,
+      "grad_norm": 0.11572265625,
+      "learning_rate": 5.180952978075074e-06,
+      "loss": 0.0139,
+      "step": 3647
+    },
+    {
+      "epoch": 260.57142857142856,
+      "grad_norm": 0.1513671875,
+      "learning_rate": 5.173880460704947e-06,
+      "loss": 0.0153,
+      "step": 3648
+    },
+    {
+      "epoch": 260.64285714285717,
+      "grad_norm": 0.1103515625,
+      "learning_rate": 5.166811513996485e-06,
+      "loss": 0.0116,
+      "step": 3649
+    },
+    {
+      "epoch": 260.7142857142857,
+      "grad_norm": 0.1201171875,
+      "learning_rate": 5.159746141395006e-06,
+      "loss": 0.0123,
+      "step": 3650
+    },
+    {
+      "epoch": 260.7857142857143,
+      "grad_norm": 0.11279296875,
+      "learning_rate": 5.152684346344087e-06,
+      "loss": 0.0116,
+      "step": 3651
+    },
+    {
+      "epoch": 260.85714285714283,
+      "grad_norm": 0.134765625,
+      "learning_rate": 5.145626132285563e-06,
+      "loss": 0.0132,
+      "step": 3652
+    },
+    {
+      "epoch": 260.92857142857144,
+      "grad_norm": 0.11572265625,
+      "learning_rate": 5.138571502659519e-06,
+      "loss": 0.0128,
+      "step": 3653
+    },
+    {
+      "epoch": 261.0,
+      "grad_norm": 0.1474609375,
+      "learning_rate": 5.131520460904296e-06,
+      "loss": 0.0161,
+      "step": 3654
+    },
+    {
+      "epoch": 261.07142857142856,
+      "grad_norm": 0.2353515625,
+      "learning_rate": 5.124473010456489e-06,
+      "loss": 0.0104,
+      "step": 3655
+    },
+    {
+      "epoch": 261.14285714285717,
+      "grad_norm": 0.142578125,
+      "learning_rate": 5.117429154750931e-06,
+      "loss": 0.0143,
+      "step": 3656
+    },
+    {
+      "epoch": 261.2142857142857,
+      "grad_norm": 0.12451171875,
+      "learning_rate": 5.110388897220725e-06,
+      "loss": 0.0155,
+      "step": 3657
+    },
+    {
+      "epoch": 261.2857142857143,
+      "grad_norm": 0.1513671875,
+      "learning_rate": 5.1033522412972e-06,
+      "loss": 0.0121,
+      "step": 3658
+    },
+    {
+      "epoch": 261.35714285714283,
+      "grad_norm": 0.12890625,
+      "learning_rate": 5.096319190409941e-06,
+      "loss": 0.0133,
+      "step": 3659
+    },
+    {
+      "epoch": 261.42857142857144,
+      "grad_norm": 0.2001953125,
+      "learning_rate": 5.08928974798676e-06,
+      "loss": 0.0105,
+      "step": 3660
+    },
+    {
+      "epoch": 261.5,
+      "grad_norm": 0.2294921875,
+      "learning_rate": 5.0822639174537334e-06,
+      "loss": 0.0142,
+      "step": 3661
+    },
+    {
+      "epoch": 261.57142857142856,
+      "grad_norm": 0.12353515625,
+      "learning_rate": 5.075241702235162e-06,
+      "loss": 0.0121,
+      "step": 3662
+    },
+    {
+      "epoch": 261.64285714285717,
+      "grad_norm": 0.16796875,
+      "learning_rate": 5.068223105753588e-06,
+      "loss": 0.0129,
+      "step": 3663
+    },
+    {
+      "epoch": 261.7142857142857,
+      "grad_norm": 0.1220703125,
+      "learning_rate": 5.061208131429789e-06,
+      "loss": 0.0116,
+      "step": 3664
+    },
+    {
+      "epoch": 261.7857142857143,
+      "grad_norm": 0.1298828125,
+      "learning_rate": 5.054196782682776e-06,
+      "loss": 0.0121,
+      "step": 3665
+    },
+    {
+      "epoch": 261.85714285714283,
+      "grad_norm": 0.1376953125,
+      "learning_rate": 5.047189062929801e-06,
+      "loss": 0.0164,
+      "step": 3666
+    },
+    {
+      "epoch": 261.92857142857144,
+      "grad_norm": 0.1572265625,
+      "learning_rate": 5.04018497558634e-06,
+      "loss": 0.0131,
+      "step": 3667
+    },
+    {
+      "epoch": 262.0,
+      "grad_norm": 0.1865234375,
+      "learning_rate": 5.033184524066104e-06,
+      "loss": 0.0164,
+      "step": 3668
+    },
+    {
+      "epoch": 262.07142857142856,
+      "grad_norm": 0.11767578125,
+      "learning_rate": 5.026187711781019e-06,
+      "loss": 0.0131,
+      "step": 3669
+    },
+    {
+      "epoch": 262.14285714285717,
+      "grad_norm": 0.158203125,
+      "learning_rate": 5.019194542141245e-06,
+      "loss": 0.0116,
+      "step": 3670
+    },
+    {
+      "epoch": 262.2142857142857,
+      "grad_norm": 0.140625,
+      "learning_rate": 5.012205018555179e-06,
+      "loss": 0.0143,
+      "step": 3671
+    },
+    {
+      "epoch": 262.2857142857143,
+      "grad_norm": 0.22265625,
+      "learning_rate": 5.005219144429423e-06,
+      "loss": 0.0138,
+      "step": 3672
+    },
+    {
+      "epoch": 262.35714285714283,
+      "grad_norm": 0.1416015625,
+      "learning_rate": 4.998236923168807e-06,
+      "loss": 0.0114,
+      "step": 3673
+    },
+    {
+      "epoch": 262.42857142857144,
+      "grad_norm": 0.271484375,
+      "learning_rate": 4.991258358176384e-06,
+      "loss": 0.015,
+      "step": 3674
+    },
+    {
+      "epoch": 262.5,
+      "grad_norm": 0.1396484375,
+      "learning_rate": 4.984283452853417e-06,
+      "loss": 0.013,
+      "step": 3675
+    },
+    {
+      "epoch": 262.57142857142856,
+      "grad_norm": 0.1083984375,
+      "learning_rate": 4.977312210599396e-06,
+      "loss": 0.0097,
+      "step": 3676
+    },
+    {
+      "epoch": 262.64285714285717,
+      "grad_norm": 0.1962890625,
+      "learning_rate": 4.97034463481202e-06,
+      "loss": 0.0138,
+      "step": 3677
+    },
+    {
+      "epoch": 262.7142857142857,
+      "grad_norm": 0.126953125,
+      "learning_rate": 4.963380728887194e-06,
+      "loss": 0.0122,
+      "step": 3678
+    },
+    {
+      "epoch": 262.7857142857143,
+      "grad_norm": 0.12890625,
+      "learning_rate": 4.956420496219043e-06,
+      "loss": 0.0129,
+      "step": 3679
+    },
+    {
+      "epoch": 262.85714285714283,
+      "grad_norm": 0.146484375,
+      "learning_rate": 4.949463940199903e-06,
+      "loss": 0.0149,
+      "step": 3680
+    },
+    {
+      "epoch": 262.92857142857144,
+      "grad_norm": 0.1630859375,
+      "learning_rate": 4.942511064220315e-06,
+      "loss": 0.013,
+      "step": 3681
+    },
+    {
+      "epoch": 263.0,
+      "grad_norm": 0.2421875,
+      "learning_rate": 4.935561871669024e-06,
+      "loss": 0.0142,
+      "step": 3682
+    },
+    {
+      "epoch": 263.07142857142856,
+      "grad_norm": 0.1240234375,
+      "learning_rate": 4.928616365932984e-06,
+      "loss": 0.0127,
+      "step": 3683
+    },
+    {
+      "epoch": 263.14285714285717,
+      "grad_norm": 0.1826171875,
+      "learning_rate": 4.921674550397348e-06,
+      "loss": 0.0131,
+      "step": 3684
+    },
+    {
+      "epoch": 263.2142857142857,
+      "grad_norm": 0.287109375,
+      "learning_rate": 4.914736428445471e-06,
+      "loss": 0.0111,
+      "step": 3685
+    },
+    {
+      "epoch": 263.2857142857143,
+      "grad_norm": 0.1474609375,
+      "learning_rate": 4.9078020034589195e-06,
+      "loss": 0.0126,
+      "step": 3686
+    },
+    {
+      "epoch": 263.35714285714283,
+      "grad_norm": 0.220703125,
+      "learning_rate": 4.900871278817436e-06,
+      "loss": 0.0149,
+      "step": 3687
+    },
+    {
+      "epoch": 263.42857142857144,
+      "grad_norm": 0.1396484375,
+      "learning_rate": 4.893944257898976e-06,
+      "loss": 0.0124,
+      "step": 3688
+    },
+    {
+      "epoch": 263.5,
+      "grad_norm": 0.12451171875,
+      "learning_rate": 4.887020944079682e-06,
+      "loss": 0.0136,
+      "step": 3689
+    },
+    {
+      "epoch": 263.57142857142856,
+      "grad_norm": 0.140625,
+      "learning_rate": 4.8801013407338984e-06,
+      "loss": 0.0115,
+      "step": 3690
+    },
+    {
+      "epoch": 263.64285714285717,
+      "grad_norm": 0.11865234375,
+      "learning_rate": 4.8731854512341555e-06,
+      "loss": 0.0138,
+      "step": 3691
+    },
+    {
+      "epoch": 263.7142857142857,
+      "grad_norm": 0.1318359375,
+      "learning_rate": 4.866273278951172e-06,
+      "loss": 0.0109,
+      "step": 3692
+    },
+    {
+      "epoch": 263.7857142857143,
+      "grad_norm": 0.154296875,
+      "learning_rate": 4.859364827253857e-06,
+      "loss": 0.0138,
+      "step": 3693
+    },
+    {
+      "epoch": 263.85714285714283,
+      "grad_norm": 0.1494140625,
+      "learning_rate": 4.8524600995093e-06,
+      "loss": 0.0138,
+      "step": 3694
+    },
+    {
+      "epoch": 263.92857142857144,
+      "grad_norm": 0.10546875,
+      "learning_rate": 4.845559099082797e-06,
+      "loss": 0.0109,
+      "step": 3695
+    },
+    {
+      "epoch": 264.0,
+      "grad_norm": 0.109375,
+      "learning_rate": 4.838661829337797e-06,
+      "loss": 0.0135,
+      "step": 3696
+    },
+    {
+      "epoch": 264.07142857142856,
+      "grad_norm": 0.142578125,
+      "learning_rate": 4.831768293635951e-06,
+      "loss": 0.0101,
+      "step": 3697
+    },
+    {
+      "epoch": 264.14285714285717,
+      "grad_norm": 0.140625,
+      "learning_rate": 4.824878495337079e-06,
+      "loss": 0.0108,
+      "step": 3698
+    },
+    {
+      "epoch": 264.2142857142857,
+      "grad_norm": 0.12451171875,
+      "learning_rate": 4.817992437799194e-06,
+      "loss": 0.0131,
+      "step": 3699
+    },
+    {
+      "epoch": 264.2857142857143,
+      "grad_norm": 0.1884765625,
+      "learning_rate": 4.811110124378473e-06,
+      "loss": 0.0139,
+      "step": 3700
+    },
+    {
+      "epoch": 264.35714285714283,
+      "grad_norm": 0.296875,
+      "learning_rate": 4.804231558429271e-06,
+      "loss": 0.0141,
+      "step": 3701
+    },
+    {
+      "epoch": 264.42857142857144,
+      "grad_norm": 0.15625,
+      "learning_rate": 4.79735674330412e-06,
+      "loss": 0.0131,
+      "step": 3702
+    },
+    {
+      "epoch": 264.5,
+      "grad_norm": 0.1923828125,
+      "learning_rate": 4.79048568235372e-06,
+      "loss": 0.0149,
+      "step": 3703
+    },
+    {
+      "epoch": 264.57142857142856,
+      "grad_norm": 0.2109375,
+      "learning_rate": 4.783618378926939e-06,
+      "loss": 0.0151,
+      "step": 3704
+    },
+    {
+      "epoch": 264.64285714285717,
+      "grad_norm": 0.146484375,
+      "learning_rate": 4.776754836370822e-06,
+      "loss": 0.0128,
+      "step": 3705
+    },
+    {
+      "epoch": 264.7142857142857,
+      "grad_norm": 0.1357421875,
+      "learning_rate": 4.769895058030576e-06,
+      "loss": 0.0126,
+      "step": 3706
+    },
+    {
+      "epoch": 264.7857142857143,
+      "grad_norm": 0.15625,
+      "learning_rate": 4.7630390472495705e-06,
+      "loss": 0.0128,
+      "step": 3707
+    },
+    {
+      "epoch": 264.85714285714283,
+      "grad_norm": 0.13671875,
+      "learning_rate": 4.756186807369341e-06,
+      "loss": 0.0144,
+      "step": 3708
+    },
+    {
+      "epoch": 264.92857142857144,
+      "grad_norm": 0.14453125,
+      "learning_rate": 4.749338341729592e-06,
+      "loss": 0.0121,
+      "step": 3709
+    },
+    {
+      "epoch": 265.0,
+      "grad_norm": 0.265625,
+      "learning_rate": 4.74249365366818e-06,
+      "loss": 0.0145,
+      "step": 3710
+    },
+    {
+      "epoch": 265.07142857142856,
+      "grad_norm": 0.1142578125,
+      "learning_rate": 4.7356527465211215e-06,
+      "loss": 0.0139,
+      "step": 3711
+    },
+    {
+      "epoch": 265.14285714285717,
+      "grad_norm": 0.208984375,
+      "learning_rate": 4.728815623622591e-06,
+      "loss": 0.0143,
+      "step": 3712
+    },
+    {
+      "epoch": 265.2142857142857,
+      "grad_norm": 0.1748046875,
+      "learning_rate": 4.721982288304922e-06,
+      "loss": 0.0143,
+      "step": 3713
+    },
+    {
+      "epoch": 265.2857142857143,
+      "grad_norm": 0.1240234375,
+      "learning_rate": 4.715152743898596e-06,
+      "loss": 0.0128,
+      "step": 3714
+    },
+    {
+      "epoch": 265.35714285714283,
+      "grad_norm": 0.10546875,
+      "learning_rate": 4.708326993732251e-06,
+      "loss": 0.0122,
+      "step": 3715
+    },
+    {
+      "epoch": 265.42857142857144,
+      "grad_norm": 0.1630859375,
+      "learning_rate": 4.701505041132676e-06,
+      "loss": 0.0142,
+      "step": 3716
+    },
+    {
+      "epoch": 265.5,
+      "grad_norm": 0.1669921875,
+      "learning_rate": 4.694686889424806e-06,
+      "loss": 0.0123,
+      "step": 3717
+    },
+    {
+      "epoch": 265.57142857142856,
+      "grad_norm": 0.134765625,
+      "learning_rate": 4.687872541931722e-06,
+      "loss": 0.013,
+      "step": 3718
+    },
+    {
+      "epoch": 265.64285714285717,
+      "grad_norm": 0.166015625,
+      "learning_rate": 4.681062001974662e-06,
+      "loss": 0.0151,
+      "step": 3719
+    },
+    {
+      "epoch": 265.7142857142857,
+      "grad_norm": 0.123046875,
+      "learning_rate": 4.674255272872998e-06,
+      "loss": 0.0148,
+      "step": 3720
+    },
+    {
+      "epoch": 265.7857142857143,
+      "grad_norm": 0.2265625,
+      "learning_rate": 4.667452357944247e-06,
+      "loss": 0.0124,
+      "step": 3721
+    },
+    {
+      "epoch": 265.85714285714283,
+      "grad_norm": 0.158203125,
+      "learning_rate": 4.660653260504061e-06,
+      "loss": 0.0134,
+      "step": 3722
+    },
+    {
+      "epoch": 265.92857142857144,
+      "grad_norm": 0.1259765625,
+      "learning_rate": 4.6538579838662455e-06,
+      "loss": 0.0125,
+      "step": 3723
+    },
+    {
+      "epoch": 266.0,
+      "grad_norm": 0.1591796875,
+      "learning_rate": 4.647066531342732e-06,
+      "loss": 0.0135,
+      "step": 3724
+    },
+    {
+      "epoch": 266.07142857142856,
+      "grad_norm": 0.1484375,
+      "learning_rate": 4.640278906243594e-06,
+      "loss": 0.0149,
+      "step": 3725
+    },
+    {
+      "epoch": 266.14285714285717,
+      "grad_norm": 0.1376953125,
+      "learning_rate": 4.633495111877034e-06,
+      "loss": 0.0139,
+      "step": 3726
+    },
+    {
+      "epoch": 266.2142857142857,
+      "grad_norm": 0.193359375,
+      "learning_rate": 4.62671515154939e-06,
+      "loss": 0.015,
+      "step": 3727
+    },
+    {
+      "epoch": 266.2857142857143,
+      "grad_norm": 0.2373046875,
+      "learning_rate": 4.61993902856514e-06,
+      "loss": 0.0119,
+      "step": 3728
+    },
+    {
+      "epoch": 266.35714285714283,
+      "grad_norm": 0.1396484375,
+      "learning_rate": 4.61316674622688e-06,
+      "loss": 0.0166,
+      "step": 3729
+    },
+    {
+      "epoch": 266.42857142857144,
+      "grad_norm": 0.126953125,
+      "learning_rate": 4.60639830783534e-06,
+      "loss": 0.0131,
+      "step": 3730
+    },
+    {
+      "epoch": 266.5,
+      "grad_norm": 0.154296875,
+      "learning_rate": 4.5996337166893636e-06,
+      "loss": 0.015,
+      "step": 3731
+    },
+    {
+      "epoch": 266.57142857142856,
+      "grad_norm": 0.203125,
+      "learning_rate": 4.592872976085945e-06,
+      "loss": 0.0133,
+      "step": 3732
+    },
+    {
+      "epoch": 266.64285714285717,
+      "grad_norm": 0.2294921875,
+      "learning_rate": 4.58611608932018e-06,
+      "loss": 0.0127,
+      "step": 3733
+    },
+    {
+      "epoch": 266.7142857142857,
+      "grad_norm": 0.142578125,
+      "learning_rate": 4.579363059685294e-06,
+      "loss": 0.0155,
+      "step": 3734
+    },
+    {
+      "epoch": 266.7857142857143,
+      "grad_norm": 0.1357421875,
+      "learning_rate": 4.572613890472632e-06,
+      "loss": 0.0124,
+      "step": 3735
+    },
+    {
+      "epoch": 266.85714285714283,
+      "grad_norm": 0.1484375,
+      "learning_rate": 4.565868584971657e-06,
+      "loss": 0.0121,
+      "step": 3736
+    },
+    {
+      "epoch": 266.92857142857144,
+      "grad_norm": 0.1591796875,
+      "learning_rate": 4.559127146469945e-06,
+      "loss": 0.0143,
+      "step": 3737
+    },
+    {
+      "epoch": 267.0,
+      "grad_norm": 0.1494140625,
+      "learning_rate": 4.552389578253198e-06,
+      "loss": 0.0139,
+      "step": 3738
+    },
+    {
+      "epoch": 267.07142857142856,
+      "grad_norm": 0.150390625,
+      "learning_rate": 4.545655883605227e-06,
+      "loss": 0.015,
+      "step": 3739
+    },
+    {
+      "epoch": 267.14285714285717,
+      "grad_norm": 0.1552734375,
+      "learning_rate": 4.538926065807944e-06,
+      "loss": 0.0134,
+      "step": 3740
+    },
+    {
+      "epoch": 267.2142857142857,
+      "grad_norm": 0.17578125,
+      "learning_rate": 4.532200128141378e-06,
+      "loss": 0.0132,
+      "step": 3741
+    },
+    {
+      "epoch": 267.2857142857143,
+      "grad_norm": 0.275390625,
+      "learning_rate": 4.525478073883682e-06,
+      "loss": 0.0111,
+      "step": 3742
+    },
+    {
+      "epoch": 267.35714285714283,
+      "grad_norm": 0.142578125,
+      "learning_rate": 4.518759906311097e-06,
+      "loss": 0.0126,
+      "step": 3743
+    },
+    {
+      "epoch": 267.42857142857144,
+      "grad_norm": 0.138671875,
+      "learning_rate": 4.512045628697975e-06,
+      "loss": 0.0128,
+      "step": 3744
+    },
+    {
+      "epoch": 267.5,
+      "grad_norm": 0.173828125,
+      "learning_rate": 4.505335244316776e-06,
+      "loss": 0.014,
+      "step": 3745
+    },
+    {
+      "epoch": 267.57142857142856,
+      "grad_norm": 0.224609375,
+      "learning_rate": 4.498628756438054e-06,
+      "loss": 0.0146,
+      "step": 3746
+    },
+    {
+      "epoch": 267.64285714285717,
+      "grad_norm": 0.171875,
+      "learning_rate": 4.491926168330479e-06,
+      "loss": 0.0148,
+      "step": 3747
+    },
+    {
+      "epoch": 267.7142857142857,
+      "grad_norm": 0.1279296875,
+      "learning_rate": 4.4852274832608105e-06,
+      "loss": 0.0155,
+      "step": 3748
+    },
+    {
+      "epoch": 267.7857142857143,
+      "grad_norm": 0.2119140625,
+      "learning_rate": 4.478532704493899e-06,
+      "loss": 0.0124,
+      "step": 3749
+    },
+    {
+      "epoch": 267.85714285714283,
+      "grad_norm": 0.1279296875,
+      "learning_rate": 4.471841835292695e-06,
+      "loss": 0.0159,
+      "step": 3750
+    },
+    {
+      "epoch": 267.92857142857144,
+      "grad_norm": 0.1787109375,
+      "learning_rate": 4.465154878918258e-06,
+      "loss": 0.0141,
+      "step": 3751
+    },
+    {
+      "epoch": 268.0,
+      "grad_norm": 0.1396484375,
+      "learning_rate": 4.458471838629725e-06,
+      "loss": 0.0152,
+      "step": 3752
+    },
+    {
+      "epoch": 268.07142857142856,
+      "grad_norm": 0.279296875,
+      "learning_rate": 4.451792717684329e-06,
+      "loss": 0.0128,
+      "step": 3753
+    },
+    {
+      "epoch": 268.14285714285717,
+      "grad_norm": 0.1650390625,
+      "learning_rate": 4.4451175193373916e-06,
+      "loss": 0.0148,
+      "step": 3754
+    },
+    {
+      "epoch": 268.2142857142857,
+      "grad_norm": 0.14453125,
+      "learning_rate": 4.438446246842322e-06,
+      "loss": 0.0114,
+      "step": 3755
+    },
+    {
+      "epoch": 268.2857142857143,
+      "grad_norm": 0.1240234375,
+      "learning_rate": 4.431778903450616e-06,
+      "loss": 0.0127,
+      "step": 3756
+    },
+    {
+      "epoch": 268.35714285714283,
+      "grad_norm": 0.298828125,
+      "learning_rate": 4.425115492411867e-06,
+      "loss": 0.0119,
+      "step": 3757
+    },
+    {
+      "epoch": 268.42857142857144,
+      "grad_norm": 0.126953125,
+      "learning_rate": 4.418456016973728e-06,
+      "loss": 0.0112,
+      "step": 3758
+    },
+    {
+      "epoch": 268.5,
+      "grad_norm": 0.1552734375,
+      "learning_rate": 4.41180048038195e-06,
+      "loss": 0.0155,
+      "step": 3759
+    },
+    {
+      "epoch": 268.57142857142856,
+      "grad_norm": 0.296875,
+      "learning_rate": 4.4051488858803575e-06,
+      "loss": 0.0129,
+      "step": 3760
+    },
+    {
+      "epoch": 268.64285714285717,
+      "grad_norm": 0.1630859375,
+      "learning_rate": 4.398501236710864e-06,
+      "loss": 0.0119,
+      "step": 3761
+    },
+    {
+      "epoch": 268.7142857142857,
+      "grad_norm": 0.1337890625,
+      "learning_rate": 4.39185753611345e-06,
+      "loss": 0.0178,
+      "step": 3762
+    },
+    {
+      "epoch": 268.7857142857143,
+      "grad_norm": 0.1328125,
+      "learning_rate": 4.385217787326175e-06,
+      "loss": 0.0147,
+      "step": 3763
+    },
+    {
+      "epoch": 268.85714285714283,
+      "grad_norm": 0.1748046875,
+      "learning_rate": 4.3785819935851694e-06,
+      "loss": 0.0104,
+      "step": 3764
+    },
+    {
+      "epoch": 268.92857142857144,
+      "grad_norm": 0.134765625,
+      "learning_rate": 4.371950158124636e-06,
+      "loss": 0.0145,
+      "step": 3765
+    },
+    {
+      "epoch": 269.0,
+      "grad_norm": 0.310546875,
+      "learning_rate": 4.365322284176865e-06,
+      "loss": 0.0123,
+      "step": 3766
+    },
+    {
+      "epoch": 269.07142857142856,
+      "grad_norm": 0.1865234375,
+      "learning_rate": 4.358698374972187e-06,
+      "loss": 0.0154,
+      "step": 3767
+    },
+    {
+      "epoch": 269.14285714285717,
+      "grad_norm": 0.12451171875,
+      "learning_rate": 4.352078433739018e-06,
+      "loss": 0.0148,
+      "step": 3768
+    },
+    {
+      "epoch": 269.2142857142857,
+      "grad_norm": 0.1455078125,
+      "learning_rate": 4.3454624637038395e-06,
+      "loss": 0.0131,
+      "step": 3769
+    },
+    {
+      "epoch": 269.2857142857143,
+      "grad_norm": 0.171875,
+      "learning_rate": 4.3388504680911894e-06,
+      "loss": 0.0157,
+      "step": 3770
+    },
+    {
+      "epoch": 269.35714285714283,
+      "grad_norm": 0.1396484375,
+      "learning_rate": 4.3322424501236815e-06,
+      "loss": 0.0134,
+      "step": 3771
+    },
+    {
+      "epoch": 269.42857142857144,
+      "grad_norm": 0.140625,
+      "learning_rate": 4.325638413021981e-06,
+      "loss": 0.0127,
+      "step": 3772
+    },
+    {
+      "epoch": 269.5,
+      "grad_norm": 0.203125,
+      "learning_rate": 4.319038360004815e-06,
+      "loss": 0.0126,
+      "step": 3773
+    },
+    {
+      "epoch": 269.57142857142856,
+      "grad_norm": 0.11376953125,
+      "learning_rate": 4.312442294288969e-06,
+      "loss": 0.0129,
+      "step": 3774
+    },
+    {
+      "epoch": 269.64285714285717,
+      "grad_norm": 0.130859375,
+      "learning_rate": 4.305850219089285e-06,
+      "loss": 0.0138,
+      "step": 3775
+    },
+    {
+      "epoch": 269.7142857142857,
+      "grad_norm": 0.1669921875,
+      "learning_rate": 4.299262137618661e-06,
+      "loss": 0.0151,
+      "step": 3776
+    },
+    {
+      "epoch": 269.7857142857143,
+      "grad_norm": 0.1376953125,
+      "learning_rate": 4.292678053088047e-06,
+      "loss": 0.0146,
+      "step": 3777
+    },
+    {
+      "epoch": 269.85714285714283,
+      "grad_norm": 0.15234375,
+      "learning_rate": 4.2860979687064456e-06,
+      "loss": 0.0122,
+      "step": 3778
+    },
+    {
+      "epoch": 269.92857142857144,
+      "grad_norm": 0.1162109375,
+      "learning_rate": 4.279521887680908e-06,
+      "loss": 0.0127,
+      "step": 3779
+    },
+    {
+      "epoch": 270.0,
+      "grad_norm": 0.1689453125,
+      "learning_rate": 4.272949813216543e-06,
+      "loss": 0.014,
+      "step": 3780
+    },
+    {
+      "epoch": 270.07142857142856,
+      "grad_norm": 0.13671875,
+      "learning_rate": 4.266381748516493e-06,
+      "loss": 0.0129,
+      "step": 3781
+    },
+    {
+      "epoch": 270.14285714285717,
+      "grad_norm": 0.1201171875,
+      "learning_rate": 4.259817696781958e-06,
+      "loss": 0.0141,
+      "step": 3782
+    },
+    {
+      "epoch": 270.2142857142857,
+      "grad_norm": 0.1943359375,
+      "learning_rate": 4.253257661212175e-06,
+      "loss": 0.0153,
+      "step": 3783
+    },
+    {
+      "epoch": 270.2857142857143,
+      "grad_norm": 0.1181640625,
+      "learning_rate": 4.2467016450044235e-06,
+      "loss": 0.0124,
+      "step": 3784
+    },
+    {
+      "epoch": 270.35714285714283,
+      "grad_norm": 0.1318359375,
+      "learning_rate": 4.240149651354029e-06,
+      "loss": 0.0128,
+      "step": 3785
+    },
+    {
+      "epoch": 270.42857142857144,
+      "grad_norm": 0.11962890625,
+      "learning_rate": 4.233601683454353e-06,
+      "loss": 0.0112,
+      "step": 3786
+    },
+    {
+      "epoch": 270.5,
+      "grad_norm": 0.1650390625,
+      "learning_rate": 4.227057744496796e-06,
+      "loss": 0.014,
+      "step": 3787
+    },
+    {
+      "epoch": 270.57142857142856,
+      "grad_norm": 0.2021484375,
+      "learning_rate": 4.2205178376707915e-06,
+      "loss": 0.0121,
+      "step": 3788
+    },
+    {
+      "epoch": 270.64285714285717,
+      "grad_norm": 0.1982421875,
+      "learning_rate": 4.21398196616381e-06,
+      "loss": 0.0144,
+      "step": 3789
+    },
+    {
+      "epoch": 270.7142857142857,
+      "grad_norm": 0.1630859375,
+      "learning_rate": 4.207450133161362e-06,
+      "loss": 0.0132,
+      "step": 3790
+    },
+    {
+      "epoch": 270.7857142857143,
+      "grad_norm": 0.1572265625,
+      "learning_rate": 4.20092234184698e-06,
+      "loss": 0.0136,
+      "step": 3791
+    },
+    {
+      "epoch": 270.85714285714283,
+      "grad_norm": 0.1396484375,
+      "learning_rate": 4.194398595402234e-06,
+      "loss": 0.0148,
+      "step": 3792
+    },
+    {
+      "epoch": 270.92857142857144,
+      "grad_norm": 0.169921875,
+      "learning_rate": 4.187878897006707e-06,
+      "loss": 0.0133,
+      "step": 3793
+    },
+    {
+      "epoch": 271.0,
+      "grad_norm": 0.1630859375,
+      "learning_rate": 4.181363249838034e-06,
+      "loss": 0.0149,
+      "step": 3794
+    },
+    {
+      "epoch": 271.07142857142856,
+      "grad_norm": 0.1396484375,
+      "learning_rate": 4.174851657071855e-06,
+      "loss": 0.0135,
+      "step": 3795
+    },
+    {
+      "epoch": 271.14285714285717,
+      "grad_norm": 0.1572265625,
+      "learning_rate": 4.168344121881845e-06,
+      "loss": 0.0119,
+      "step": 3796
+    },
+    {
+      "epoch": 271.2142857142857,
+      "grad_norm": 0.2099609375,
+      "learning_rate": 4.161840647439697e-06,
+      "loss": 0.012,
+      "step": 3797
+    },
+    {
+      "epoch": 271.2857142857143,
+      "grad_norm": 0.1845703125,
+      "learning_rate": 4.155341236915118e-06,
+      "loss": 0.0151,
+      "step": 3798
+    },
+    {
+      "epoch": 271.35714285714283,
+      "grad_norm": 0.20703125,
+      "learning_rate": 4.1488458934758524e-06,
+      "loss": 0.0106,
+      "step": 3799
+    },
+    {
+      "epoch": 271.42857142857144,
+      "grad_norm": 0.1123046875,
+      "learning_rate": 4.142354620287649e-06,
+      "loss": 0.016,
+      "step": 3800
+    },
+    {
+      "epoch": 271.5,
+      "grad_norm": 0.1357421875,
+      "learning_rate": 4.1358674205142765e-06,
+      "loss": 0.0138,
+      "step": 3801
+    },
+    {
+      "epoch": 271.57142857142856,
+      "grad_norm": 0.1728515625,
+      "learning_rate": 4.129384297317511e-06,
+      "loss": 0.0149,
+      "step": 3802
+    },
+    {
+      "epoch": 271.64285714285717,
+      "grad_norm": 0.1318359375,
+      "learning_rate": 4.122905253857147e-06,
+      "loss": 0.0124,
+      "step": 3803
+    },
+    {
+      "epoch": 271.7142857142857,
+      "grad_norm": 0.134765625,
+      "learning_rate": 4.1164302932910005e-06,
+      "loss": 0.0138,
+      "step": 3804
+    },
+    {
+      "epoch": 271.7857142857143,
+      "grad_norm": 0.177734375,
+      "learning_rate": 4.109959418774885e-06,
+      "loss": 0.0095,
+      "step": 3805
+    },
+    {
+      "epoch": 271.85714285714283,
+      "grad_norm": 0.1376953125,
+      "learning_rate": 4.1034926334626244e-06,
+      "loss": 0.014,
+      "step": 3806
+    },
+    {
+      "epoch": 271.92857142857144,
+      "grad_norm": 0.12255859375,
+      "learning_rate": 4.097029940506053e-06,
+      "loss": 0.0126,
+      "step": 3807
+    },
+    {
+      "epoch": 272.0,
+      "grad_norm": 0.25390625,
+      "learning_rate": 4.090571343055004e-06,
+      "loss": 0.0135,
+      "step": 3808
+    },
+    {
+      "epoch": 272.07142857142856,
+      "grad_norm": 0.1259765625,
+      "learning_rate": 4.084116844257327e-06,
+      "loss": 0.0144,
+      "step": 3809
+    },
+    {
+      "epoch": 272.14285714285717,
+      "grad_norm": 0.10693359375,
+      "learning_rate": 4.077666447258867e-06,
+      "loss": 0.0115,
+      "step": 3810
+    },
+    {
+      "epoch": 272.2142857142857,
+      "grad_norm": 0.16015625,
+      "learning_rate": 4.071220155203462e-06,
+      "loss": 0.0143,
+      "step": 3811
+    },
+    {
+      "epoch": 272.2857142857143,
+      "grad_norm": 0.11865234375,
+      "learning_rate": 4.064777971232954e-06,
+      "loss": 0.0118,
+      "step": 3812
+    },
+    {
+      "epoch": 272.35714285714283,
+      "grad_norm": 0.1396484375,
+      "learning_rate": 4.058339898487194e-06,
+      "loss": 0.0148,
+      "step": 3813
+    },
+    {
+      "epoch": 272.42857142857144,
+      "grad_norm": 0.126953125,
+      "learning_rate": 4.051905940104018e-06,
+      "loss": 0.0141,
+      "step": 3814
+    },
+    {
+      "epoch": 272.5,
+      "grad_norm": 0.1591796875,
+      "learning_rate": 4.04547609921926e-06,
+      "loss": 0.0146,
+      "step": 3815
+    },
+    {
+      "epoch": 272.57142857142856,
+      "grad_norm": 0.1591796875,
+      "learning_rate": 4.039050378966743e-06,
+      "loss": 0.0145,
+      "step": 3816
+    },
+    {
+      "epoch": 272.64285714285717,
+      "grad_norm": 0.1748046875,
+      "learning_rate": 4.032628782478285e-06,
+      "loss": 0.0122,
+      "step": 3817
+    },
+    {
+      "epoch": 272.7142857142857,
+      "grad_norm": 0.11328125,
+      "learning_rate": 4.0262113128836934e-06,
+      "loss": 0.0115,
+      "step": 3818
+    },
+    {
+      "epoch": 272.7857142857143,
+      "grad_norm": 0.134765625,
+      "learning_rate": 4.019797973310775e-06,
+      "loss": 0.0149,
+      "step": 3819
+    },
+    {
+      "epoch": 272.85714285714283,
+      "grad_norm": 0.298828125,
+      "learning_rate": 4.013388766885302e-06,
+      "loss": 0.0128,
+      "step": 3820
+    },
+    {
+      "epoch": 272.92857142857144,
+      "grad_norm": 0.2294921875,
+      "learning_rate": 4.006983696731047e-06,
+      "loss": 0.0148,
+      "step": 3821
+    },
+    {
+      "epoch": 273.0,
+      "grad_norm": 0.1513671875,
+      "learning_rate": 4.00058276596976e-06,
+      "loss": 0.0121,
+      "step": 3822
+    },
+    {
+      "epoch": 273.07142857142856,
+      "grad_norm": 0.1669921875,
+      "learning_rate": 3.994185977721185e-06,
+      "loss": 0.0129,
+      "step": 3823
+    },
+    {
+      "epoch": 273.14285714285717,
+      "grad_norm": 0.1611328125,
+      "learning_rate": 3.987793335103033e-06,
+      "loss": 0.0109,
+      "step": 3824
+    },
+    {
+      "epoch": 273.2142857142857,
+      "grad_norm": 0.12890625,
+      "learning_rate": 3.981404841231004e-06,
+      "loss": 0.013,
+      "step": 3825
+    },
+    {
+      "epoch": 273.2857142857143,
+      "grad_norm": 0.1904296875,
+      "learning_rate": 3.975020499218771e-06,
+      "loss": 0.0113,
+      "step": 3826
+    },
+    {
+      "epoch": 273.35714285714283,
+      "grad_norm": 0.154296875,
+      "learning_rate": 3.968640312177978e-06,
+      "loss": 0.0136,
+      "step": 3827
+    },
+    {
+      "epoch": 273.42857142857144,
+      "grad_norm": 0.166015625,
+      "learning_rate": 3.9622642832182654e-06,
+      "loss": 0.0105,
+      "step": 3828
+    },
+    {
+      "epoch": 273.5,
+      "grad_norm": 0.1318359375,
+      "learning_rate": 3.95589241544722e-06,
+      "loss": 0.0151,
+      "step": 3829
+    },
+    {
+      "epoch": 273.57142857142856,
+      "grad_norm": 0.19140625,
+      "learning_rate": 3.9495247119704164e-06,
+      "loss": 0.0121,
+      "step": 3830
+    },
+    {
+      "epoch": 273.64285714285717,
+      "grad_norm": 0.15234375,
+      "learning_rate": 3.943161175891391e-06,
+      "loss": 0.0121,
+      "step": 3831
+    },
+    {
+      "epoch": 273.7142857142857,
+      "grad_norm": 0.21875,
+      "learning_rate": 3.936801810311663e-06,
+      "loss": 0.0128,
+      "step": 3832
+    },
+    {
+      "epoch": 273.7857142857143,
+      "grad_norm": 0.1318359375,
+      "learning_rate": 3.930446618330706e-06,
+      "loss": 0.0136,
+      "step": 3833
+    },
+    {
+      "epoch": 273.85714285714283,
+      "grad_norm": 0.115234375,
+      "learning_rate": 3.924095603045963e-06,
+      "loss": 0.0094,
+      "step": 3834
+    },
+    {
+      "epoch": 273.92857142857144,
+      "grad_norm": 0.142578125,
+      "learning_rate": 3.9177487675528425e-06,
+      "loss": 0.0157,
+      "step": 3835
+    },
+    {
+      "epoch": 274.0,
+      "grad_norm": 0.1455078125,
+      "learning_rate": 3.911406114944714e-06,
+      "loss": 0.0123,
+      "step": 3836
+    },
+    {
+      "epoch": 274.07142857142856,
+      "grad_norm": 0.1494140625,
+      "learning_rate": 3.905067648312909e-06,
+      "loss": 0.0135,
+      "step": 3837
+    },
+    {
+      "epoch": 274.14285714285717,
+      "grad_norm": 0.1279296875,
+      "learning_rate": 3.898733370746721e-06,
+      "loss": 0.0126,
+      "step": 3838
+    },
+    {
+      "epoch": 274.2142857142857,
+      "grad_norm": 0.1923828125,
+      "learning_rate": 3.892403285333397e-06,
+      "loss": 0.0114,
+      "step": 3839
+    },
+    {
+      "epoch": 274.2857142857143,
+      "grad_norm": 0.208984375,
+      "learning_rate": 3.886077395158147e-06,
+      "loss": 0.0122,
+      "step": 3840
+    },
+    {
+      "epoch": 274.35714285714283,
+      "grad_norm": 0.1376953125,
+      "learning_rate": 3.879755703304128e-06,
+      "loss": 0.0121,
+      "step": 3841
+    },
+    {
+      "epoch": 274.42857142857144,
+      "grad_norm": 0.140625,
+      "learning_rate": 3.873438212852461e-06,
+      "loss": 0.0125,
+      "step": 3842
+    },
+    {
+      "epoch": 274.5,
+      "grad_norm": 0.212890625,
+      "learning_rate": 3.867124926882213e-06,
+      "loss": 0.0154,
+      "step": 3843
+    },
+    {
+      "epoch": 274.57142857142856,
+      "grad_norm": 0.12890625,
+      "learning_rate": 3.860815848470404e-06,
+      "loss": 0.0129,
+      "step": 3844
+    },
+    {
+      "epoch": 274.64285714285717,
+      "grad_norm": 0.1318359375,
+      "learning_rate": 3.854510980692e-06,
+      "loss": 0.012,
+      "step": 3845
+    },
+    {
+      "epoch": 274.7142857142857,
+      "grad_norm": 0.1640625,
+      "learning_rate": 3.8482103266199166e-06,
+      "loss": 0.0121,
+      "step": 3846
+    },
+    {
+      "epoch": 274.7857142857143,
+      "grad_norm": 0.09912109375,
+      "learning_rate": 3.841913889325016e-06,
+      "loss": 0.0111,
+      "step": 3847
+    },
+    {
+      "epoch": 274.85714285714283,
+      "grad_norm": 0.14453125,
+      "learning_rate": 3.835621671876108e-06,
+      "loss": 0.0145,
+      "step": 3848
+    },
+    {
+      "epoch": 274.92857142857144,
+      "grad_norm": 0.1240234375,
+      "learning_rate": 3.82933367733994e-06,
+      "loss": 0.0107,
+      "step": 3849
+    },
+    {
+      "epoch": 275.0,
+      "grad_norm": 0.1357421875,
+      "learning_rate": 3.823049908781202e-06,
+      "loss": 0.0138,
+      "step": 3850
+    },
+    {
+      "epoch": 275.07142857142856,
+      "grad_norm": 0.158203125,
+      "learning_rate": 3.816770369262533e-06,
+      "loss": 0.011,
+      "step": 3851
+    },
+    {
+      "epoch": 275.14285714285717,
+      "grad_norm": 0.162109375,
+      "learning_rate": 3.810495061844502e-06,
+      "loss": 0.014,
+      "step": 3852
+    },
+    {
+      "epoch": 275.2142857142857,
+      "grad_norm": 0.1943359375,
+      "learning_rate": 3.8042239895856145e-06,
+      "loss": 0.0138,
+      "step": 3853
+    },
+    {
+      "epoch": 275.2857142857143,
+      "grad_norm": 0.1552734375,
+      "learning_rate": 3.7979571555423226e-06,
+      "loss": 0.017,
+      "step": 3854
+    },
+    {
+      "epoch": 275.35714285714283,
+      "grad_norm": 0.1640625,
+      "learning_rate": 3.7916945627689907e-06,
+      "loss": 0.0127,
+      "step": 3855
+    },
+    {
+      "epoch": 275.42857142857144,
+      "grad_norm": 0.1416015625,
+      "learning_rate": 3.7854362143179445e-06,
+      "loss": 0.0134,
+      "step": 3856
+    },
+    {
+      "epoch": 275.5,
+      "grad_norm": 0.1318359375,
+      "learning_rate": 3.779182113239421e-06,
+      "loss": 0.013,
+      "step": 3857
+    },
+    {
+      "epoch": 275.57142857142856,
+      "grad_norm": 0.287109375,
+      "learning_rate": 3.772932262581594e-06,
+      "loss": 0.0147,
+      "step": 3858
+    },
+    {
+      "epoch": 275.64285714285717,
+      "grad_norm": 0.16015625,
+      "learning_rate": 3.766686665390563e-06,
+      "loss": 0.0144,
+      "step": 3859
+    },
+    {
+      "epoch": 275.7142857142857,
+      "grad_norm": 0.185546875,
+      "learning_rate": 3.760445324710356e-06,
+      "loss": 0.0109,
+      "step": 3860
+    },
+    {
+      "epoch": 275.7857142857143,
+      "grad_norm": 0.111328125,
+      "learning_rate": 3.754208243582931e-06,
+      "loss": 0.0099,
+      "step": 3861
+    },
+    {
+      "epoch": 275.85714285714283,
+      "grad_norm": 0.15234375,
+      "learning_rate": 3.747975425048164e-06,
+      "loss": 0.0149,
+      "step": 3862
+    },
+    {
+      "epoch": 275.92857142857144,
+      "grad_norm": 0.1953125,
+      "learning_rate": 3.7417468721438574e-06,
+      "loss": 0.0125,
+      "step": 3863
+    },
+    {
+      "epoch": 276.0,
+      "grad_norm": 0.1669921875,
+      "learning_rate": 3.735522587905721e-06,
+      "loss": 0.0145,
+      "step": 3864
+    },
+    {
+      "epoch": 276.07142857142856,
+      "grad_norm": 0.1396484375,
+      "learning_rate": 3.729302575367408e-06,
+      "loss": 0.0144,
+      "step": 3865
+    },
+    {
+      "epoch": 276.14285714285717,
+      "grad_norm": 0.203125,
+      "learning_rate": 3.7230868375604702e-06,
+      "loss": 0.0156,
+      "step": 3866
+    },
+    {
+      "epoch": 276.2142857142857,
+      "grad_norm": 0.1279296875,
+      "learning_rate": 3.7168753775143865e-06,
+      "loss": 0.0152,
+      "step": 3867
+    },
+    {
+      "epoch": 276.2857142857143,
+      "grad_norm": 0.45703125,
+      "learning_rate": 3.710668198256545e-06,
+      "loss": 0.0161,
+      "step": 3868
+    },
+    {
+      "epoch": 276.35714285714283,
+      "grad_norm": 0.154296875,
+      "learning_rate": 3.704465302812249e-06,
+      "loss": 0.012,
+      "step": 3869
+    },
+    {
+      "epoch": 276.42857142857144,
+      "grad_norm": 0.1806640625,
+      "learning_rate": 3.6982666942047104e-06,
+      "loss": 0.0121,
+      "step": 3870
+    },
+    {
+      "epoch": 276.5,
+      "grad_norm": 0.138671875,
+      "learning_rate": 3.6920723754550664e-06,
+      "loss": 0.0123,
+      "step": 3871
+    },
+    {
+      "epoch": 276.57142857142856,
+      "grad_norm": 0.142578125,
+      "learning_rate": 3.685882349582348e-06,
+      "loss": 0.0136,
+      "step": 3872
+    },
+    {
+      "epoch": 276.64285714285717,
+      "grad_norm": 0.1416015625,
+      "learning_rate": 3.6796966196034955e-06,
+      "loss": 0.0128,
+      "step": 3873
+    },
+    {
+      "epoch": 276.7142857142857,
+      "grad_norm": 0.1455078125,
+      "learning_rate": 3.673515188533355e-06,
+      "loss": 0.0147,
+      "step": 3874
+    },
+    {
+      "epoch": 276.7857142857143,
+      "grad_norm": 0.1357421875,
+      "learning_rate": 3.667338059384691e-06,
+      "loss": 0.0128,
+      "step": 3875
+    },
+    {
+      "epoch": 276.85714285714283,
+      "grad_norm": 0.126953125,
+      "learning_rate": 3.6611652351681568e-06,
+      "loss": 0.0116,
+      "step": 3876
+    },
+    {
+      "epoch": 276.92857142857144,
+      "grad_norm": 0.11865234375,
+      "learning_rate": 3.654996718892312e-06,
+      "loss": 0.0137,
+      "step": 3877
+    },
+    {
+      "epoch": 277.0,
+      "grad_norm": 0.1943359375,
+      "learning_rate": 3.6488325135636147e-06,
+      "loss": 0.015,
+      "step": 3878
+    },
+    {
+      "epoch": 277.07142857142856,
+      "grad_norm": 0.14453125,
+      "learning_rate": 3.6426726221864217e-06,
+      "loss": 0.0135,
+      "step": 3879
+    },
+    {
+      "epoch": 277.14285714285717,
+      "grad_norm": 0.10205078125,
+      "learning_rate": 3.6365170477629966e-06,
+      "loss": 0.0119,
+      "step": 3880
+    },
+    {
+      "epoch": 277.2142857142857,
+      "grad_norm": 0.1884765625,
+      "learning_rate": 3.6303657932934913e-06,
+      "loss": 0.0141,
+      "step": 3881
+    },
+    {
+      "epoch": 277.2857142857143,
+      "grad_norm": 0.150390625,
+      "learning_rate": 3.624218861775945e-06,
+      "loss": 0.014,
+      "step": 3882
+    },
+    {
+      "epoch": 277.35714285714283,
+      "grad_norm": 0.1708984375,
+      "learning_rate": 3.6180762562062974e-06,
+      "loss": 0.014,
+      "step": 3883
+    },
+    {
+      "epoch": 277.42857142857144,
+      "grad_norm": 0.125,
+      "learning_rate": 3.611937979578388e-06,
+      "loss": 0.013,
+      "step": 3884
+    },
+    {
+      "epoch": 277.5,
+      "grad_norm": 0.162109375,
+      "learning_rate": 3.6058040348839334e-06,
+      "loss": 0.0116,
+      "step": 3885
+    },
+    {
+      "epoch": 277.57142857142856,
+      "grad_norm": 0.166015625,
+      "learning_rate": 3.599674425112545e-06,
+      "loss": 0.013,
+      "step": 3886
+    },
+    {
+      "epoch": 277.64285714285717,
+      "grad_norm": 0.154296875,
+      "learning_rate": 3.5935491532517186e-06,
+      "loss": 0.011,
+      "step": 3887
+    },
+    {
+      "epoch": 277.7142857142857,
+      "grad_norm": 0.2109375,
+      "learning_rate": 3.5874282222868386e-06,
+      "loss": 0.0127,
+      "step": 3888
+    },
+    {
+      "epoch": 277.7857142857143,
+      "grad_norm": 0.177734375,
+      "learning_rate": 3.5813116352011677e-06,
+      "loss": 0.0134,
+      "step": 3889
+    },
+    {
+      "epoch": 277.85714285714283,
+      "grad_norm": 0.1103515625,
+      "learning_rate": 3.5751993949758686e-06,
+      "loss": 0.0102,
+      "step": 3890
+    },
+    {
+      "epoch": 277.92857142857144,
+      "grad_norm": 0.130859375,
+      "learning_rate": 3.569091504589961e-06,
+      "loss": 0.0131,
+      "step": 3891
+    },
+    {
+      "epoch": 278.0,
+      "grad_norm": 0.173828125,
+      "learning_rate": 3.5629879670203594e-06,
+      "loss": 0.0117,
+      "step": 3892
+    },
+    {
+      "epoch": 278.07142857142856,
+      "grad_norm": 0.1513671875,
+      "learning_rate": 3.5568887852418504e-06,
+      "loss": 0.0109,
+      "step": 3893
+    },
+    {
+      "epoch": 278.14285714285717,
+      "grad_norm": 0.185546875,
+      "learning_rate": 3.55079396222711e-06,
+      "loss": 0.0139,
+      "step": 3894
+    },
+    {
+      "epoch": 278.2142857142857,
+      "grad_norm": 0.1728515625,
+      "learning_rate": 3.5447035009466763e-06,
+      "loss": 0.0158,
+      "step": 3895
+    },
+    {
+      "epoch": 278.2857142857143,
+      "grad_norm": 0.1669921875,
+      "learning_rate": 3.5386174043689658e-06,
+      "loss": 0.0129,
+      "step": 3896
+    },
+    {
+      "epoch": 278.35714285714283,
+      "grad_norm": 0.15234375,
+      "learning_rate": 3.532535675460269e-06,
+      "loss": 0.0139,
+      "step": 3897
+    },
+    {
+      "epoch": 278.42857142857144,
+      "grad_norm": 0.1416015625,
+      "learning_rate": 3.5264583171847416e-06,
+      "loss": 0.0131,
+      "step": 3898
+    },
+    {
+      "epoch": 278.5,
+      "grad_norm": 0.2421875,
+      "learning_rate": 3.5203853325044263e-06,
+      "loss": 0.0125,
+      "step": 3899
+    },
+    {
+      "epoch": 278.57142857142856,
+      "grad_norm": 0.1728515625,
+      "learning_rate": 3.5143167243792104e-06,
+      "loss": 0.0132,
+      "step": 3900
+    },
+    {
+      "epoch": 278.64285714285717,
+      "grad_norm": 0.1416015625,
+      "learning_rate": 3.508252495766863e-06,
+      "loss": 0.0125,
+      "step": 3901
+    },
+    {
+      "epoch": 278.7142857142857,
+      "grad_norm": 0.1748046875,
+      "learning_rate": 3.5021926496230177e-06,
+      "loss": 0.0161,
+      "step": 3902
+    },
+    {
+      "epoch": 278.7857142857143,
+      "grad_norm": 0.126953125,
+      "learning_rate": 3.4961371889011635e-06,
+      "loss": 0.0121,
+      "step": 3903
+    },
+    {
+      "epoch": 278.85714285714283,
+      "grad_norm": 0.19140625,
+      "learning_rate": 3.490086116552667e-06,
+      "loss": 0.0133,
+      "step": 3904
+    },
+    {
+      "epoch": 278.92857142857144,
+      "grad_norm": 0.1220703125,
+      "learning_rate": 3.484039435526744e-06,
+      "loss": 0.0144,
+      "step": 3905
+    },
+    {
+      "epoch": 279.0,
+      "grad_norm": 0.1611328125,
+      "learning_rate": 3.4779971487704733e-06,
+      "loss": 0.015,
+      "step": 3906
+    },
+    {
+      "epoch": 279.07142857142856,
+      "grad_norm": 0.2109375,
+      "learning_rate": 3.471959259228792e-06,
+      "loss": 0.0164,
+      "step": 3907
+    },
+    {
+      "epoch": 279.14285714285717,
+      "grad_norm": 0.134765625,
+      "learning_rate": 3.4659257698444926e-06,
+      "loss": 0.0127,
+      "step": 3908
+    },
+    {
+      "epoch": 279.2142857142857,
+      "grad_norm": 0.1376953125,
+      "learning_rate": 3.4598966835582265e-06,
+      "loss": 0.0122,
+      "step": 3909
+    },
+    {
+      "epoch": 279.2857142857143,
+      "grad_norm": 0.16796875,
+      "learning_rate": 3.453872003308496e-06,
+      "loss": 0.0132,
+      "step": 3910
+    },
+    {
+      "epoch": 279.35714285714283,
+      "grad_norm": 0.15234375,
+      "learning_rate": 3.4478517320316594e-06,
+      "loss": 0.0146,
+      "step": 3911
+    },
+    {
+      "epoch": 279.42857142857144,
+      "grad_norm": 0.123046875,
+      "learning_rate": 3.4418358726619164e-06,
+      "loss": 0.0125,
+      "step": 3912
+    },
+    {
+      "epoch": 279.5,
+      "grad_norm": 0.1865234375,
+      "learning_rate": 3.435824428131333e-06,
+      "loss": 0.013,
+      "step": 3913
+    },
+    {
+      "epoch": 279.57142857142856,
+      "grad_norm": 0.1591796875,
+      "learning_rate": 3.42981740136981e-06,
+      "loss": 0.0153,
+      "step": 3914
+    },
+    {
+      "epoch": 279.64285714285717,
+      "grad_norm": 0.21484375,
+      "learning_rate": 3.423814795305101e-06,
+      "loss": 0.0157,
+      "step": 3915
+    },
+    {
+      "epoch": 279.7142857142857,
+      "grad_norm": 0.1669921875,
+      "learning_rate": 3.4178166128628024e-06,
+      "loss": 0.0131,
+      "step": 3916
+    },
+    {
+      "epoch": 279.7857142857143,
+      "grad_norm": 0.111328125,
+      "learning_rate": 3.411822856966354e-06,
+      "loss": 0.013,
+      "step": 3917
+    },
+    {
+      "epoch": 279.85714285714283,
+      "grad_norm": 0.296875,
+      "learning_rate": 3.4058335305370404e-06,
+      "loss": 0.0152,
+      "step": 3918
+    },
+    {
+      "epoch": 279.92857142857144,
+      "grad_norm": 0.1318359375,
+      "learning_rate": 3.3998486364939878e-06,
+      "loss": 0.012,
+      "step": 3919
+    },
+    {
+      "epoch": 280.0,
+      "grad_norm": 0.12890625,
+      "learning_rate": 3.3938681777541586e-06,
+      "loss": 0.0137,
+      "step": 3920
+    },
+    {
+      "epoch": 280.07142857142856,
+      "grad_norm": 0.125,
+      "learning_rate": 3.3878921572323585e-06,
+      "loss": 0.012,
+      "step": 3921
+    },
+    {
+      "epoch": 280.14285714285717,
+      "grad_norm": 0.11767578125,
+      "learning_rate": 3.3819205778412216e-06,
+      "loss": 0.0132,
+      "step": 3922
+    },
+    {
+      "epoch": 280.2142857142857,
+      "grad_norm": 0.12109375,
+      "learning_rate": 3.375953442491231e-06,
+      "loss": 0.0124,
+      "step": 3923
+    },
+    {
+      "epoch": 280.2857142857143,
+      "grad_norm": 0.1337890625,
+      "learning_rate": 3.369990754090692e-06,
+      "loss": 0.0124,
+      "step": 3924
+    },
+    {
+      "epoch": 280.35714285714283,
+      "grad_norm": 0.1376953125,
+      "learning_rate": 3.3640325155457507e-06,
+      "loss": 0.0138,
+      "step": 3925
+    },
+    {
+      "epoch": 280.42857142857144,
+      "grad_norm": 0.1640625,
+      "learning_rate": 3.3580787297603695e-06,
+      "loss": 0.014,
+      "step": 3926
+    },
+    {
+      "epoch": 280.5,
+      "grad_norm": 0.162109375,
+      "learning_rate": 3.352129399636361e-06,
+      "loss": 0.0117,
+      "step": 3927
+    },
+    {
+      "epoch": 280.57142857142856,
+      "grad_norm": 0.158203125,
+      "learning_rate": 3.346184528073354e-06,
+      "loss": 0.0147,
+      "step": 3928
+    },
+    {
+      "epoch": 280.64285714285717,
+      "grad_norm": 0.1337890625,
+      "learning_rate": 3.3402441179688063e-06,
+      "loss": 0.0133,
+      "step": 3929
+    },
+    {
+      "epoch": 280.7142857142857,
+      "grad_norm": 0.12451171875,
+      "learning_rate": 3.334308172218002e-06,
+      "loss": 0.0154,
+      "step": 3930
+    },
+    {
+      "epoch": 280.7857142857143,
+      "grad_norm": 0.1259765625,
+      "learning_rate": 3.3283766937140444e-06,
+      "loss": 0.0139,
+      "step": 3931
+    },
+    {
+      "epoch": 280.85714285714283,
+      "grad_norm": 0.13671875,
+      "learning_rate": 3.3224496853478712e-06,
+      "loss": 0.0164,
+      "step": 3932
+    },
+    {
+      "epoch": 280.92857142857144,
+      "grad_norm": 0.1435546875,
+      "learning_rate": 3.3165271500082323e-06,
+      "loss": 0.0143,
+      "step": 3933
+    },
+    {
+      "epoch": 281.0,
+      "grad_norm": 0.12060546875,
+      "learning_rate": 3.310609090581701e-06,
+      "loss": 0.013,
+      "step": 3934
+    },
+    {
+      "epoch": 281.07142857142856,
+      "grad_norm": 0.2041015625,
+      "learning_rate": 3.3046955099526584e-06,
+      "loss": 0.0112,
+      "step": 3935
+    },
+    {
+      "epoch": 281.14285714285717,
+      "grad_norm": 0.1337890625,
+      "learning_rate": 3.2987864110033212e-06,
+      "loss": 0.0119,
+      "step": 3936
+    },
+    {
+      "epoch": 281.2142857142857,
+      "grad_norm": 0.1875,
+      "learning_rate": 3.29288179661371e-06,
+      "loss": 0.0136,
+      "step": 3937
+    },
+    {
+      "epoch": 281.2857142857143,
+      "grad_norm": 0.1689453125,
+      "learning_rate": 3.2869816696616623e-06,
+      "loss": 0.0116,
+      "step": 3938
+    },
+    {
+      "epoch": 281.35714285714283,
+      "grad_norm": 0.181640625,
+      "learning_rate": 3.2810860330228274e-06,
+      "loss": 0.0107,
+      "step": 3939
+    },
+    {
+      "epoch": 281.42857142857144,
+      "grad_norm": 0.138671875,
+      "learning_rate": 3.275194889570668e-06,
+      "loss": 0.0106,
+      "step": 3940
+    },
+    {
+      "epoch": 281.5,
+      "grad_norm": 0.1435546875,
+      "learning_rate": 3.269308242176451e-06,
+      "loss": 0.0127,
+      "step": 3941
+    },
+    {
+      "epoch": 281.57142857142856,
+      "grad_norm": 0.326171875,
+      "learning_rate": 3.2634260937092652e-06,
+      "loss": 0.0152,
+      "step": 3942
+    },
+    {
+      "epoch": 281.64285714285717,
+      "grad_norm": 0.263671875,
+      "learning_rate": 3.2575484470359974e-06,
+      "loss": 0.0144,
+      "step": 3943
+    },
+    {
+      "epoch": 281.7142857142857,
+      "grad_norm": 0.2216796875,
+      "learning_rate": 3.2516753050213345e-06,
+      "loss": 0.0133,
+      "step": 3944
+    },
+    {
+      "epoch": 281.7857142857143,
+      "grad_norm": 0.2265625,
+      "learning_rate": 3.2458066705277733e-06,
+      "loss": 0.0127,
+      "step": 3945
+    },
+    {
+      "epoch": 281.85714285714283,
+      "grad_norm": 0.32421875,
+      "learning_rate": 3.239942546415625e-06,
+      "loss": 0.0112,
+      "step": 3946
+    },
+    {
+      "epoch": 281.92857142857144,
+      "grad_norm": 0.1748046875,
+      "learning_rate": 3.2340829355429855e-06,
+      "loss": 0.0158,
+      "step": 3947
+    },
+    {
+      "epoch": 282.0,
+      "grad_norm": 0.11181640625,
+      "learning_rate": 3.2282278407657615e-06,
+      "loss": 0.0111,
+      "step": 3948
+    },
+    {
+      "epoch": 282.07142857142856,
+      "grad_norm": 0.1201171875,
+      "learning_rate": 3.222377264937651e-06,
+      "loss": 0.0124,
+      "step": 3949
+    },
+    {
+      "epoch": 282.14285714285717,
+      "grad_norm": 0.1572265625,
+      "learning_rate": 3.216531210910154e-06,
+      "loss": 0.0122,
+      "step": 3950
+    },
+    {
+      "epoch": 282.2142857142857,
+      "grad_norm": 0.134765625,
+      "learning_rate": 3.210689681532571e-06,
+      "loss": 0.0118,
+      "step": 3951
+    },
+    {
+      "epoch": 282.2857142857143,
+      "grad_norm": 0.1943359375,
+      "learning_rate": 3.204852679651993e-06,
+      "loss": 0.0135,
+      "step": 3952
+    },
+    {
+      "epoch": 282.35714285714283,
+      "grad_norm": 0.142578125,
+      "learning_rate": 3.199020208113296e-06,
+      "loss": 0.0131,
+      "step": 3953
+    },
+    {
+      "epoch": 282.42857142857144,
+      "grad_norm": 0.1064453125,
+      "learning_rate": 3.19319226975916e-06,
+      "loss": 0.0121,
+      "step": 3954
+    },
+    {
+      "epoch": 282.5,
+      "grad_norm": 0.1943359375,
+      "learning_rate": 3.187368867430049e-06,
+      "loss": 0.0164,
+      "step": 3955
+    },
+    {
+      "epoch": 282.57142857142856,
+      "grad_norm": 0.150390625,
+      "learning_rate": 3.181550003964223e-06,
+      "loss": 0.0159,
+      "step": 3956
+    },
+    {
+      "epoch": 282.64285714285717,
+      "grad_norm": 0.142578125,
+      "learning_rate": 3.1757356821977237e-06,
+      "loss": 0.0125,
+      "step": 3957
+    },
+    {
+      "epoch": 282.7142857142857,
+      "grad_norm": 0.1962890625,
+      "learning_rate": 3.169925904964381e-06,
+      "loss": 0.0124,
+      "step": 3958
+    },
+    {
+      "epoch": 282.7857142857143,
+      "grad_norm": 0.17578125,
+      "learning_rate": 3.1641206750958092e-06,
+      "loss": 0.0127,
+      "step": 3959
+    },
+    {
+      "epoch": 282.85714285714283,
+      "grad_norm": 0.1875,
+      "learning_rate": 3.158319995421402e-06,
+      "loss": 0.0124,
+      "step": 3960
+    },
+    {
+      "epoch": 282.92857142857144,
+      "grad_norm": 0.1611328125,
+      "learning_rate": 3.1525238687683534e-06,
+      "loss": 0.0164,
+      "step": 3961
+    },
+    {
+      "epoch": 283.0,
+      "grad_norm": 0.140625,
+      "learning_rate": 3.1467322979616132e-06,
+      "loss": 0.0127,
+      "step": 3962
+    },
+    {
+      "epoch": 283.07142857142856,
+      "grad_norm": 0.1796875,
+      "learning_rate": 3.1409452858239264e-06,
+      "loss": 0.0141,
+      "step": 3963
+    },
+    {
+      "epoch": 283.14285714285717,
+      "grad_norm": 0.130859375,
+      "learning_rate": 3.1351628351758077e-06,
+      "loss": 0.0134,
+      "step": 3964
+    },
+    {
+      "epoch": 283.2142857142857,
+      "grad_norm": 0.12255859375,
+      "learning_rate": 3.129384948835562e-06,
+      "loss": 0.0131,
+      "step": 3965
+    },
+    {
+      "epoch": 283.2857142857143,
+      "grad_norm": 0.208984375,
+      "learning_rate": 3.1236116296192554e-06,
+      "loss": 0.0113,
+      "step": 3966
+    },
+    {
+      "epoch": 283.35714285714283,
+      "grad_norm": 0.12890625,
+      "learning_rate": 3.1178428803407365e-06,
+      "loss": 0.0127,
+      "step": 3967
+    },
+    {
+      "epoch": 283.42857142857144,
+      "grad_norm": 0.1328125,
+      "learning_rate": 3.1120787038116216e-06,
+      "loss": 0.0145,
+      "step": 3968
+    },
+    {
+      "epoch": 283.5,
+      "grad_norm": 0.1640625,
+      "learning_rate": 3.1063191028413e-06,
+      "loss": 0.0122,
+      "step": 3969
+    },
+    {
+      "epoch": 283.57142857142856,
+      "grad_norm": 0.1181640625,
+      "learning_rate": 3.1005640802369326e-06,
+      "loss": 0.0142,
+      "step": 3970
+    },
+    {
+      "epoch": 283.64285714285717,
+      "grad_norm": 0.1474609375,
+      "learning_rate": 3.094813638803447e-06,
+      "loss": 0.0139,
+      "step": 3971
+    },
+    {
+      "epoch": 283.7142857142857,
+      "grad_norm": 0.13671875,
+      "learning_rate": 3.089067781343538e-06,
+      "loss": 0.0122,
+      "step": 3972
+    },
+    {
+      "epoch": 283.7857142857143,
+      "grad_norm": 0.1806640625,
+      "learning_rate": 3.0833265106576677e-06,
+      "loss": 0.0159,
+      "step": 3973
+    },
+    {
+      "epoch": 283.85714285714283,
+      "grad_norm": 0.1279296875,
+      "learning_rate": 3.077589829544057e-06,
+      "loss": 0.0121,
+      "step": 3974
+    },
+    {
+      "epoch": 283.92857142857144,
+      "grad_norm": 0.150390625,
+      "learning_rate": 3.071857740798702e-06,
+      "loss": 0.0119,
+      "step": 3975
+    },
+    {
+      "epoch": 284.0,
+      "grad_norm": 0.1103515625,
+      "learning_rate": 3.06613024721535e-06,
+      "loss": 0.0139,
+      "step": 3976
+    },
+    {
+      "epoch": 284.07142857142856,
+      "grad_norm": 0.130859375,
+      "learning_rate": 3.060407351585512e-06,
+      "loss": 0.0132,
+      "step": 3977
+    },
+    {
+      "epoch": 284.14285714285717,
+      "grad_norm": 0.130859375,
+      "learning_rate": 3.0546890566984564e-06,
+      "loss": 0.0118,
+      "step": 3978
+    },
+    {
+      "epoch": 284.2142857142857,
+      "grad_norm": 0.1279296875,
+      "learning_rate": 3.0489753653412127e-06,
+      "loss": 0.0136,
+      "step": 3979
+    },
+    {
+      "epoch": 284.2857142857143,
+      "grad_norm": 0.154296875,
+      "learning_rate": 3.0432662802985624e-06,
+      "loss": 0.0117,
+      "step": 3980
+    },
+    {
+      "epoch": 284.35714285714283,
+      "grad_norm": 0.11328125,
+      "learning_rate": 3.0375618043530466e-06,
+      "loss": 0.0138,
+      "step": 3981
+    },
+    {
+      "epoch": 284.42857142857144,
+      "grad_norm": 0.11962890625,
+      "learning_rate": 3.031861940284955e-06,
+      "loss": 0.0148,
+      "step": 3982
+    },
+    {
+      "epoch": 284.5,
+      "grad_norm": 0.12255859375,
+      "learning_rate": 3.0261666908723313e-06,
+      "loss": 0.013,
+      "step": 3983
+    },
+    {
+      "epoch": 284.57142857142856,
+      "grad_norm": 0.1455078125,
+      "learning_rate": 3.0204760588909765e-06,
+      "loss": 0.015,
+      "step": 3984
+    },
+    {
+      "epoch": 284.64285714285717,
+      "grad_norm": 0.13671875,
+      "learning_rate": 3.014790047114431e-06,
+      "loss": 0.0139,
+      "step": 3985
+    },
+    {
+      "epoch": 284.7142857142857,
+      "grad_norm": 0.1796875,
+      "learning_rate": 3.00910865831399e-06,
+      "loss": 0.0124,
+      "step": 3986
+    },
+    {
+      "epoch": 284.7857142857143,
+      "grad_norm": 0.1328125,
+      "learning_rate": 3.003431895258696e-06,
+      "loss": 0.0122,
+      "step": 3987
+    },
+    {
+      "epoch": 284.85714285714283,
+      "grad_norm": 0.138671875,
+      "learning_rate": 2.9977597607153228e-06,
+      "loss": 0.0131,
+      "step": 3988
+    },
+    {
+      "epoch": 284.92857142857144,
+      "grad_norm": 0.1279296875,
+      "learning_rate": 2.9920922574484102e-06,
+      "loss": 0.0115,
+      "step": 3989
+    },
+    {
+      "epoch": 285.0,
+      "grad_norm": 0.26953125,
+      "learning_rate": 2.98642938822023e-06,
+      "loss": 0.0167,
+      "step": 3990
+    },
+    {
+      "epoch": 285.07142857142856,
+      "grad_norm": 0.138671875,
+      "learning_rate": 2.980771155790793e-06,
+      "loss": 0.0133,
+      "step": 3991
+    },
+    {
+      "epoch": 285.14285714285717,
+      "grad_norm": 0.1474609375,
+      "learning_rate": 2.9751175629178534e-06,
+      "loss": 0.0146,
+      "step": 3992
+    },
+    {
+      "epoch": 285.2142857142857,
+      "grad_norm": 0.158203125,
+      "learning_rate": 2.9694686123569e-06,
+      "loss": 0.0157,
+      "step": 3993
+    },
+    {
+      "epoch": 285.2857142857143,
+      "grad_norm": 0.1767578125,
+      "learning_rate": 2.9638243068611722e-06,
+      "loss": 0.0137,
+      "step": 3994
+    },
+    {
+      "epoch": 285.35714285714283,
+      "grad_norm": 0.166015625,
+      "learning_rate": 2.9581846491816293e-06,
+      "loss": 0.0159,
+      "step": 3995
+    },
+    {
+      "epoch": 285.42857142857144,
+      "grad_norm": 0.1513671875,
+      "learning_rate": 2.952549642066976e-06,
+      "loss": 0.0122,
+      "step": 3996
+    },
+    {
+      "epoch": 285.5,
+      "grad_norm": 0.2177734375,
+      "learning_rate": 2.9469192882636375e-06,
+      "loss": 0.0144,
+      "step": 3997
+    },
+    {
+      "epoch": 285.57142857142856,
+      "grad_norm": 0.1279296875,
+      "learning_rate": 2.9412935905157884e-06,
+      "loss": 0.0123,
+      "step": 3998
+    },
+    {
+      "epoch": 285.64285714285717,
+      "grad_norm": 0.130859375,
+      "learning_rate": 2.935672551565323e-06,
+      "loss": 0.0126,
+      "step": 3999
+    },
+    {
+      "epoch": 285.7142857142857,
+      "grad_norm": 0.1806640625,
+      "learning_rate": 2.9300561741518685e-06,
+      "loss": 0.0133,
+      "step": 4000
+    },
+    {
+      "epoch": 285.7857142857143,
+      "grad_norm": 0.2177734375,
+      "learning_rate": 2.9244444610127764e-06,
+      "loss": 0.0114,
+      "step": 4001
+    },
+    {
+      "epoch": 285.85714285714283,
+      "grad_norm": 0.138671875,
+      "learning_rate": 2.9188374148831283e-06,
+      "loss": 0.0121,
+      "step": 4002
+    },
+    {
+      "epoch": 285.92857142857144,
+      "grad_norm": 0.1748046875,
+      "learning_rate": 2.913235038495729e-06,
+      "loss": 0.014,
+      "step": 4003
+    },
+    {
+      "epoch": 286.0,
+      "grad_norm": 0.20703125,
+      "learning_rate": 2.9076373345811133e-06,
+      "loss": 0.0151,
+      "step": 4004
+    },
+    {
+      "epoch": 286.07142857142856,
+      "grad_norm": 0.1259765625,
+      "learning_rate": 2.902044305867535e-06,
+      "loss": 0.0127,
+      "step": 4005
+    },
+    {
+      "epoch": 286.14285714285717,
+      "grad_norm": 0.1806640625,
+      "learning_rate": 2.896455955080961e-06,
+      "loss": 0.0108,
+      "step": 4006
+    },
+    {
+      "epoch": 286.2142857142857,
+      "grad_norm": 0.1611328125,
+      "learning_rate": 2.8908722849450855e-06,
+      "loss": 0.0145,
+      "step": 4007
+    },
+    {
+      "epoch": 286.2857142857143,
+      "grad_norm": 0.1240234375,
+      "learning_rate": 2.8852932981813296e-06,
+      "loss": 0.0132,
+      "step": 4008
+    },
+    {
+      "epoch": 286.35714285714283,
+      "grad_norm": 0.189453125,
+      "learning_rate": 2.8797189975088194e-06,
+      "loss": 0.0125,
+      "step": 4009
+    },
+    {
+      "epoch": 286.42857142857144,
+      "grad_norm": 0.1640625,
+      "learning_rate": 2.8741493856444007e-06,
+      "loss": 0.0128,
+      "step": 4010
+    },
+    {
+      "epoch": 286.5,
+      "grad_norm": 0.111328125,
+      "learning_rate": 2.868584465302636e-06,
+      "loss": 0.0118,
+      "step": 4011
+    },
+    {
+      "epoch": 286.57142857142856,
+      "grad_norm": 0.169921875,
+      "learning_rate": 2.8630242391957954e-06,
+      "loss": 0.0133,
+      "step": 4012
+    },
+    {
+      "epoch": 286.64285714285717,
+      "grad_norm": 0.2001953125,
+      "learning_rate": 2.857468710033873e-06,
+      "loss": 0.0119,
+      "step": 4013
+    },
+    {
+      "epoch": 286.7142857142857,
+      "grad_norm": 0.181640625,
+      "learning_rate": 2.8519178805245656e-06,
+      "loss": 0.0148,
+      "step": 4014
+    },
+    {
+      "epoch": 286.7857142857143,
+      "grad_norm": 0.12060546875,
+      "learning_rate": 2.8463717533732738e-06,
+      "loss": 0.0136,
+      "step": 4015
+    },
+    {
+      "epoch": 286.85714285714283,
+      "grad_norm": 0.1171875,
+      "learning_rate": 2.840830331283112e-06,
+      "loss": 0.0123,
+      "step": 4016
+    },
+    {
+      "epoch": 286.92857142857144,
+      "grad_norm": 0.142578125,
+      "learning_rate": 2.8352936169549093e-06,
+      "loss": 0.016,
+      "step": 4017
+    },
+    {
+      "epoch": 287.0,
+      "grad_norm": 0.1923828125,
+      "learning_rate": 2.8297616130871897e-06,
+      "loss": 0.0125,
+      "step": 4018
+    },
+    {
+      "epoch": 287.07142857142856,
+      "grad_norm": 0.228515625,
+      "learning_rate": 2.824234322376185e-06,
+      "loss": 0.0168,
+      "step": 4019
+    },
+    {
+      "epoch": 287.14285714285717,
+      "grad_norm": 0.1806640625,
+      "learning_rate": 2.818711747515829e-06,
+      "loss": 0.0121,
+      "step": 4020
+    },
+    {
+      "epoch": 287.2142857142857,
+      "grad_norm": 0.330078125,
+      "learning_rate": 2.813193891197756e-06,
+      "loss": 0.0138,
+      "step": 4021
+    },
+    {
+      "epoch": 287.2857142857143,
+      "grad_norm": 0.171875,
+      "learning_rate": 2.8076807561113026e-06,
+      "loss": 0.0157,
+      "step": 4022
+    },
+    {
+      "epoch": 287.35714285714283,
+      "grad_norm": 0.2216796875,
+      "learning_rate": 2.80217234494351e-06,
+      "loss": 0.0138,
+      "step": 4023
+    },
+    {
+      "epoch": 287.42857142857144,
+      "grad_norm": 0.2001953125,
+      "learning_rate": 2.7966686603791036e-06,
+      "loss": 0.015,
+      "step": 4024
+    },
+    {
+      "epoch": 287.5,
+      "grad_norm": 0.154296875,
+      "learning_rate": 2.7911697051005146e-06,
+      "loss": 0.0103,
+      "step": 4025
+    },
+    {
+      "epoch": 287.57142857142856,
+      "grad_norm": 0.1904296875,
+      "learning_rate": 2.785675481787864e-06,
+      "loss": 0.013,
+      "step": 4026
+    },
+    {
+      "epoch": 287.64285714285717,
+      "grad_norm": 0.259765625,
+      "learning_rate": 2.780185993118975e-06,
+      "loss": 0.0122,
+      "step": 4027
+    },
+    {
+      "epoch": 287.7142857142857,
+      "grad_norm": 0.11865234375,
+      "learning_rate": 2.7747012417693563e-06,
+      "loss": 0.012,
+      "step": 4028
+    },
+    {
+      "epoch": 287.7857142857143,
+      "grad_norm": 0.11328125,
+      "learning_rate": 2.7692212304122084e-06,
+      "loss": 0.0132,
+      "step": 4029
+    },
+    {
+      "epoch": 287.85714285714283,
+      "grad_norm": 0.2177734375,
+      "learning_rate": 2.7637459617184212e-06,
+      "loss": 0.011,
+      "step": 4030
+    },
+    {
+      "epoch": 287.92857142857144,
+      "grad_norm": 0.1611328125,
+      "learning_rate": 2.7582754383565708e-06,
+      "loss": 0.0152,
+      "step": 4031
+    },
+    {
+      "epoch": 288.0,
+      "grad_norm": 0.1123046875,
+      "learning_rate": 2.7528096629929363e-06,
+      "loss": 0.0116,
+      "step": 4032
+    },
+    {
+      "epoch": 288.07142857142856,
+      "grad_norm": 0.140625,
+      "learning_rate": 2.747348638291458e-06,
+      "loss": 0.0124,
+      "step": 4033
+    },
+    {
+      "epoch": 288.14285714285717,
+      "grad_norm": 0.13671875,
+      "learning_rate": 2.7418923669137758e-06,
+      "loss": 0.0129,
+      "step": 4034
+    },
+    {
+      "epoch": 288.2142857142857,
+      "grad_norm": 0.1494140625,
+      "learning_rate": 2.736440851519209e-06,
+      "loss": 0.0127,
+      "step": 4035
+    },
+    {
+      "epoch": 288.2857142857143,
+      "grad_norm": 0.18359375,
+      "learning_rate": 2.7309940947647646e-06,
+      "loss": 0.0126,
+      "step": 4036
+    },
+    {
+      "epoch": 288.35714285714283,
+      "grad_norm": 0.1357421875,
+      "learning_rate": 2.7255520993051243e-06,
+      "loss": 0.0121,
+      "step": 4037
+    },
+    {
+      "epoch": 288.42857142857144,
+      "grad_norm": 0.111328125,
+      "learning_rate": 2.720114867792649e-06,
+      "loss": 0.0121,
+      "step": 4038
+    },
+    {
+      "epoch": 288.5,
+      "grad_norm": 0.1552734375,
+      "learning_rate": 2.714682402877379e-06,
+      "loss": 0.0144,
+      "step": 4039
+    },
+    {
+      "epoch": 288.57142857142856,
+      "grad_norm": 0.134765625,
+      "learning_rate": 2.7092547072070332e-06,
+      "loss": 0.013,
+      "step": 4040
+    },
+    {
+      "epoch": 288.64285714285717,
+      "grad_norm": 0.197265625,
+      "learning_rate": 2.7038317834270038e-06,
+      "loss": 0.0146,
+      "step": 4041
+    },
+    {
+      "epoch": 288.7142857142857,
+      "grad_norm": 0.140625,
+      "learning_rate": 2.698413634180358e-06,
+      "loss": 0.0115,
+      "step": 4042
+    },
+    {
+      "epoch": 288.7857142857143,
+      "grad_norm": 0.1396484375,
+      "learning_rate": 2.693000262107835e-06,
+      "loss": 0.0107,
+      "step": 4043
+    },
+    {
+      "epoch": 288.85714285714283,
+      "grad_norm": 0.2099609375,
+      "learning_rate": 2.6875916698478483e-06,
+      "loss": 0.0099,
+      "step": 4044
+    },
+    {
+      "epoch": 288.92857142857144,
+      "grad_norm": 0.212890625,
+      "learning_rate": 2.6821878600364766e-06,
+      "loss": 0.0127,
+      "step": 4045
+    },
+    {
+      "epoch": 289.0,
+      "grad_norm": 0.2412109375,
+      "learning_rate": 2.6767888353074753e-06,
+      "loss": 0.0156,
+      "step": 4046
+    },
+    {
+      "epoch": 289.07142857142856,
+      "grad_norm": 0.138671875,
+      "learning_rate": 2.6713945982922637e-06,
+      "loss": 0.0134,
+      "step": 4047
+    },
+    {
+      "epoch": 289.14285714285717,
+      "grad_norm": 0.1630859375,
+      "learning_rate": 2.6660051516199253e-06,
+      "loss": 0.0136,
+      "step": 4048
+    },
+    {
+      "epoch": 289.2142857142857,
+      "grad_norm": 0.15625,
+      "learning_rate": 2.66062049791721e-06,
+      "loss": 0.0142,
+      "step": 4049
+    },
+    {
+      "epoch": 289.2857142857143,
+      "grad_norm": 0.1435546875,
+      "learning_rate": 2.655240639808536e-06,
+      "loss": 0.0132,
+      "step": 4050
+    },
+    {
+      "epoch": 289.35714285714283,
+      "grad_norm": 0.1533203125,
+      "learning_rate": 2.6498655799159763e-06,
+      "loss": 0.0125,
+      "step": 4051
+    },
+    {
+      "epoch": 289.42857142857144,
+      "grad_norm": 0.12158203125,
+      "learning_rate": 2.6444953208592736e-06,
+      "loss": 0.0119,
+      "step": 4052
+    },
+    {
+      "epoch": 289.5,
+      "grad_norm": 0.197265625,
+      "learning_rate": 2.6391298652558252e-06,
+      "loss": 0.0126,
+      "step": 4053
+    },
+    {
+      "epoch": 289.57142857142856,
+      "grad_norm": 0.1552734375,
+      "learning_rate": 2.63376921572069e-06,
+      "loss": 0.0158,
+      "step": 4054
+    },
+    {
+      "epoch": 289.64285714285717,
+      "grad_norm": 0.232421875,
+      "learning_rate": 2.6284133748665784e-06,
+      "loss": 0.0135,
+      "step": 4055
+    },
+    {
+      "epoch": 289.7142857142857,
+      "grad_norm": 0.11962890625,
+      "learning_rate": 2.62306234530387e-06,
+      "loss": 0.0145,
+      "step": 4056
+    },
+    {
+      "epoch": 289.7857142857143,
+      "grad_norm": 0.1455078125,
+      "learning_rate": 2.6177161296405876e-06,
+      "loss": 0.0154,
+      "step": 4057
+    },
+    {
+      "epoch": 289.85714285714283,
+      "grad_norm": 0.138671875,
+      "learning_rate": 2.612374730482417e-06,
+      "loss": 0.0141,
+      "step": 4058
+    },
+    {
+      "epoch": 289.92857142857144,
+      "grad_norm": 0.1240234375,
+      "learning_rate": 2.6070381504326796e-06,
+      "loss": 0.0114,
+      "step": 4059
+    },
+    {
+      "epoch": 290.0,
+      "grad_norm": 0.185546875,
+      "learning_rate": 2.6017063920923693e-06,
+      "loss": 0.0129,
+      "step": 4060
+    },
+    {
+      "epoch": 290.07142857142856,
+      "grad_norm": 0.169921875,
+      "learning_rate": 2.596379458060118e-06,
+      "loss": 0.013,
+      "step": 4061
+    },
+    {
+      "epoch": 290.14285714285717,
+      "grad_norm": 0.140625,
+      "learning_rate": 2.5910573509322078e-06,
+      "loss": 0.0122,
+      "step": 4062
+    },
+    {
+      "epoch": 290.2142857142857,
+      "grad_norm": 0.146484375,
+      "learning_rate": 2.58574007330257e-06,
+      "loss": 0.0151,
+      "step": 4063
+    },
+    {
+      "epoch": 290.2857142857143,
+      "grad_norm": 0.140625,
+      "learning_rate": 2.580427627762777e-06,
+      "loss": 0.0144,
+      "step": 4064
+    },
+    {
+      "epoch": 290.35714285714283,
+      "grad_norm": 0.220703125,
+      "learning_rate": 2.575120016902058e-06,
+      "loss": 0.0147,
+      "step": 4065
+    },
+    {
+      "epoch": 290.42857142857144,
+      "grad_norm": 0.1376953125,
+      "learning_rate": 2.569817243307272e-06,
+      "loss": 0.0135,
+      "step": 4066
+    },
+    {
+      "epoch": 290.5,
+      "grad_norm": 0.126953125,
+      "learning_rate": 2.5645193095629337e-06,
+      "loss": 0.0109,
+      "step": 4067
+    },
+    {
+      "epoch": 290.57142857142856,
+      "grad_norm": 0.2041015625,
+      "learning_rate": 2.5592262182511804e-06,
+      "loss": 0.0126,
+      "step": 4068
+    },
+    {
+      "epoch": 290.64285714285717,
+      "grad_norm": 0.1396484375,
+      "learning_rate": 2.553937971951807e-06,
+      "loss": 0.0119,
+      "step": 4069
+    },
+    {
+      "epoch": 290.7142857142857,
+      "grad_norm": 0.1484375,
+      "learning_rate": 2.548654573242243e-06,
+      "loss": 0.0155,
+      "step": 4070
+    },
+    {
+      "epoch": 290.7857142857143,
+      "grad_norm": 0.15234375,
+      "learning_rate": 2.543376024697547e-06,
+      "loss": 0.0151,
+      "step": 4071
+    },
+    {
+      "epoch": 290.85714285714283,
+      "grad_norm": 0.193359375,
+      "learning_rate": 2.5381023288904238e-06,
+      "loss": 0.0129,
+      "step": 4072
+    },
+    {
+      "epoch": 290.92857142857144,
+      "grad_norm": 0.1650390625,
+      "learning_rate": 2.5328334883912065e-06,
+      "loss": 0.013,
+      "step": 4073
+    },
+    {
+      "epoch": 291.0,
+      "grad_norm": 0.185546875,
+      "learning_rate": 2.5275695057678616e-06,
+      "loss": 0.011,
+      "step": 4074
+    },
+    {
+      "epoch": 291.07142857142856,
+      "grad_norm": 0.150390625,
+      "learning_rate": 2.522310383585996e-06,
+      "loss": 0.0126,
+      "step": 4075
+    },
+    {
+      "epoch": 291.14285714285717,
+      "grad_norm": 0.1689453125,
+      "learning_rate": 2.5170561244088426e-06,
+      "loss": 0.0136,
+      "step": 4076
+    },
+    {
+      "epoch": 291.2142857142857,
+      "grad_norm": 0.10986328125,
+      "learning_rate": 2.511806730797256e-06,
+      "loss": 0.013,
+      "step": 4077
+    },
+    {
+      "epoch": 291.2857142857143,
+      "grad_norm": 0.11767578125,
+      "learning_rate": 2.506562205309729e-06,
+      "loss": 0.011,
+      "step": 4078
+    },
+    {
+      "epoch": 291.35714285714283,
+      "grad_norm": 0.12353515625,
+      "learning_rate": 2.501322550502383e-06,
+      "loss": 0.0115,
+      "step": 4079
+    },
+    {
+      "epoch": 291.42857142857144,
+      "grad_norm": 0.11669921875,
+      "learning_rate": 2.496087768928962e-06,
+      "loss": 0.013,
+      "step": 4080
+    },
+    {
+      "epoch": 291.5,
+      "grad_norm": 0.166015625,
+      "learning_rate": 2.490857863140833e-06,
+      "loss": 0.0105,
+      "step": 4081
+    },
+    {
+      "epoch": 291.57142857142856,
+      "grad_norm": 0.1318359375,
+      "learning_rate": 2.485632835686988e-06,
+      "loss": 0.0145,
+      "step": 4082
+    },
+    {
+      "epoch": 291.64285714285717,
+      "grad_norm": 0.150390625,
+      "learning_rate": 2.480412689114041e-06,
+      "loss": 0.0139,
+      "step": 4083
+    },
+    {
+      "epoch": 291.7142857142857,
+      "grad_norm": 0.130859375,
+      "learning_rate": 2.4751974259662333e-06,
+      "loss": 0.0142,
+      "step": 4084
+    },
+    {
+      "epoch": 291.7857142857143,
+      "grad_norm": 0.263671875,
+      "learning_rate": 2.4699870487854203e-06,
+      "loss": 0.0128,
+      "step": 4085
+    },
+    {
+      "epoch": 291.85714285714283,
+      "grad_norm": 0.18359375,
+      "learning_rate": 2.4647815601110707e-06,
+      "loss": 0.014,
+      "step": 4086
+    },
+    {
+      "epoch": 291.92857142857144,
+      "grad_norm": 0.16015625,
+      "learning_rate": 2.4595809624802803e-06,
+      "loss": 0.0135,
+      "step": 4087
+    },
+    {
+      "epoch": 292.0,
+      "grad_norm": 0.1826171875,
+      "learning_rate": 2.4543852584277523e-06,
+      "loss": 0.0127,
+      "step": 4088
+    },
+    {
+      "epoch": 292.07142857142856,
+      "grad_norm": 0.1416015625,
+      "learning_rate": 2.449194450485817e-06,
+      "loss": 0.0146,
+      "step": 4089
+    },
+    {
+      "epoch": 292.14285714285717,
+      "grad_norm": 0.1484375,
+      "learning_rate": 2.4440085411844073e-06,
+      "loss": 0.011,
+      "step": 4090
+    },
+    {
+      "epoch": 292.2142857142857,
+      "grad_norm": 0.140625,
+      "learning_rate": 2.4388275330510745e-06,
+      "loss": 0.012,
+      "step": 4091
+    },
+    {
+      "epoch": 292.2857142857143,
+      "grad_norm": 0.16796875,
+      "learning_rate": 2.4336514286109764e-06,
+      "loss": 0.0143,
+      "step": 4092
+    },
+    {
+      "epoch": 292.35714285714283,
+      "grad_norm": 0.15625,
+      "learning_rate": 2.428480230386883e-06,
+      "loss": 0.0129,
+      "step": 4093
+    },
+    {
+      "epoch": 292.42857142857144,
+      "grad_norm": 0.1640625,
+      "learning_rate": 2.42331394089918e-06,
+      "loss": 0.0175,
+      "step": 4094
+    },
+    {
+      "epoch": 292.5,
+      "grad_norm": 0.1416015625,
+      "learning_rate": 2.418152562665847e-06,
+      "loss": 0.0111,
+      "step": 4095
+    },
+    {
+      "epoch": 292.57142857142856,
+      "grad_norm": 0.138671875,
+      "learning_rate": 2.4129960982024786e-06,
+      "loss": 0.0151,
+      "step": 4096
+    },
+    {
+      "epoch": 292.64285714285717,
+      "grad_norm": 0.302734375,
+      "learning_rate": 2.4078445500222703e-06,
+      "loss": 0.0133,
+      "step": 4097
+    },
+    {
+      "epoch": 292.7142857142857,
+      "grad_norm": 0.146484375,
+      "learning_rate": 2.402697920636031e-06,
+      "loss": 0.0133,
+      "step": 4098
+    },
+    {
+      "epoch": 292.7857142857143,
+      "grad_norm": 0.28515625,
+      "learning_rate": 2.3975562125521614e-06,
+      "loss": 0.0147,
+      "step": 4099
+    },
+    {
+      "epoch": 292.85714285714283,
+      "grad_norm": 0.326171875,
+      "learning_rate": 2.3924194282766684e-06,
+      "loss": 0.0139,
+      "step": 4100
+    },
+    {
+      "epoch": 292.92857142857144,
+      "grad_norm": 0.216796875,
+      "learning_rate": 2.3872875703131583e-06,
+      "loss": 0.0135,
+      "step": 4101
+    },
+    {
+      "epoch": 293.0,
+      "grad_norm": 0.16796875,
+      "learning_rate": 2.3821606411628354e-06,
+      "loss": 0.0137,
+      "step": 4102
+    },
+    {
+      "epoch": 293.07142857142856,
+      "grad_norm": 0.1259765625,
+      "learning_rate": 2.377038643324503e-06,
+      "loss": 0.0123,
+      "step": 4103
+    },
+    {
+      "epoch": 293.14285714285717,
+      "grad_norm": 0.12890625,
+      "learning_rate": 2.37192157929456e-06,
+      "loss": 0.0134,
+      "step": 4104
+    },
+    {
+      "epoch": 293.2142857142857,
+      "grad_norm": 0.1474609375,
+      "learning_rate": 2.366809451567002e-06,
+      "loss": 0.0123,
+      "step": 4105
+    },
+    {
+      "epoch": 293.2857142857143,
+      "grad_norm": 0.12109375,
+      "learning_rate": 2.3617022626334194e-06,
+      "loss": 0.0113,
+      "step": 4106
+    },
+    {
+      "epoch": 293.35714285714283,
+      "grad_norm": 0.1484375,
+      "learning_rate": 2.356600014982989e-06,
+      "loss": 0.0133,
+      "step": 4107
+    },
+    {
+      "epoch": 293.42857142857144,
+      "grad_norm": 0.134765625,
+      "learning_rate": 2.3515027111024893e-06,
+      "loss": 0.0146,
+      "step": 4108
+    },
+    {
+      "epoch": 293.5,
+      "grad_norm": 0.1689453125,
+      "learning_rate": 2.346410353476285e-06,
+      "loss": 0.0149,
+      "step": 4109
+    },
+    {
+      "epoch": 293.57142857142856,
+      "grad_norm": 0.10498046875,
+      "learning_rate": 2.3413229445863257e-06,
+      "loss": 0.0124,
+      "step": 4110
+    },
+    {
+      "epoch": 293.64285714285717,
+      "grad_norm": 0.177734375,
+      "learning_rate": 2.3362404869121557e-06,
+      "loss": 0.0127,
+      "step": 4111
+    },
+    {
+      "epoch": 293.7142857142857,
+      "grad_norm": 0.1845703125,
+      "learning_rate": 2.3311629829308998e-06,
+      "loss": 0.0154,
+      "step": 4112
+    },
+    {
+      "epoch": 293.7857142857143,
+      "grad_norm": 0.1201171875,
+      "learning_rate": 2.3260904351172757e-06,
+      "loss": 0.0105,
+      "step": 4113
+    },
+    {
+      "epoch": 293.85714285714283,
+      "grad_norm": 0.1484375,
+      "learning_rate": 2.3210228459435783e-06,
+      "loss": 0.0123,
+      "step": 4114
+    },
+    {
+      "epoch": 293.92857142857144,
+      "grad_norm": 0.15234375,
+      "learning_rate": 2.31596021787969e-06,
+      "loss": 0.0136,
+      "step": 4115
+    },
+    {
+      "epoch": 294.0,
+      "grad_norm": 0.12890625,
+      "learning_rate": 2.310902553393071e-06,
+      "loss": 0.0142,
+      "step": 4116
+    },
+    {
+      "epoch": 294.07142857142856,
+      "grad_norm": 0.1181640625,
+      "learning_rate": 2.3058498549487733e-06,
+      "loss": 0.0144,
+      "step": 4117
+    },
+    {
+      "epoch": 294.14285714285717,
+      "grad_norm": 0.16015625,
+      "learning_rate": 2.3008021250094157e-06,
+      "loss": 0.0142,
+      "step": 4118
+    },
+    {
+      "epoch": 294.2142857142857,
+      "grad_norm": 0.228515625,
+      "learning_rate": 2.295759366035201e-06,
+      "loss": 0.0128,
+      "step": 4119
+    },
+    {
+      "epoch": 294.2857142857143,
+      "grad_norm": 0.1376953125,
+      "learning_rate": 2.290721580483908e-06,
+      "loss": 0.0113,
+      "step": 4120
+    },
+    {
+      "epoch": 294.35714285714283,
+      "grad_norm": 0.193359375,
+      "learning_rate": 2.285688770810894e-06,
+      "loss": 0.0127,
+      "step": 4121
+    },
+    {
+      "epoch": 294.42857142857144,
+      "grad_norm": 0.1865234375,
+      "learning_rate": 2.2806609394690884e-06,
+      "loss": 0.0124,
+      "step": 4122
+    },
+    {
+      "epoch": 294.5,
+      "grad_norm": 0.1552734375,
+      "learning_rate": 2.2756380889089946e-06,
+      "loss": 0.0159,
+      "step": 4123
+    },
+    {
+      "epoch": 294.57142857142856,
+      "grad_norm": 0.1201171875,
+      "learning_rate": 2.27062022157869e-06,
+      "loss": 0.0141,
+      "step": 4124
+    },
+    {
+      "epoch": 294.64285714285717,
+      "grad_norm": 0.12060546875,
+      "learning_rate": 2.26560733992382e-06,
+      "loss": 0.0141,
+      "step": 4125
+    },
+    {
+      "epoch": 294.7142857142857,
+      "grad_norm": 0.197265625,
+      "learning_rate": 2.2605994463876013e-06,
+      "loss": 0.0166,
+      "step": 4126
+    },
+    {
+      "epoch": 294.7857142857143,
+      "grad_norm": 0.1455078125,
+      "learning_rate": 2.255596543410826e-06,
+      "loss": 0.0124,
+      "step": 4127
+    },
+    {
+      "epoch": 294.85714285714283,
+      "grad_norm": 0.12890625,
+      "learning_rate": 2.2505986334318455e-06,
+      "loss": 0.0149,
+      "step": 4128
+    },
+    {
+      "epoch": 294.92857142857144,
+      "grad_norm": 0.3984375,
+      "learning_rate": 2.245605718886583e-06,
+      "loss": 0.0158,
+      "step": 4129
+    },
+    {
+      "epoch": 295.0,
+      "grad_norm": 0.181640625,
+      "learning_rate": 2.240617802208514e-06,
+      "loss": 0.0156,
+      "step": 4130
+    },
+    {
+      "epoch": 295.07142857142856,
+      "grad_norm": 0.1552734375,
+      "learning_rate": 2.2356348858287e-06,
+      "loss": 0.0129,
+      "step": 4131
+    },
+    {
+      "epoch": 295.14285714285717,
+      "grad_norm": 0.126953125,
+      "learning_rate": 2.23065697217575e-06,
+      "loss": 0.0139,
+      "step": 4132
+    },
+    {
+      "epoch": 295.2142857142857,
+      "grad_norm": 0.244140625,
+      "learning_rate": 2.225684063675837e-06,
+      "loss": 0.0107,
+      "step": 4133
+    },
+    {
+      "epoch": 295.2857142857143,
+      "grad_norm": 0.1240234375,
+      "learning_rate": 2.2207161627526995e-06,
+      "loss": 0.0139,
+      "step": 4134
+    },
+    {
+      "epoch": 295.35714285714283,
+      "grad_norm": 0.177734375,
+      "learning_rate": 2.2157532718276253e-06,
+      "loss": 0.0132,
+      "step": 4135
+    },
+    {
+      "epoch": 295.42857142857144,
+      "grad_norm": 0.267578125,
+      "learning_rate": 2.2107953933194756e-06,
+      "loss": 0.0146,
+      "step": 4136
+    },
+    {
+      "epoch": 295.5,
+      "grad_norm": 0.13671875,
+      "learning_rate": 2.205842529644657e-06,
+      "loss": 0.0135,
+      "step": 4137
+    },
+    {
+      "epoch": 295.57142857142856,
+      "grad_norm": 0.13671875,
+      "learning_rate": 2.2008946832171363e-06,
+      "loss": 0.0139,
+      "step": 4138
+    },
+    {
+      "epoch": 295.64285714285717,
+      "grad_norm": 0.126953125,
+      "learning_rate": 2.1959518564484303e-06,
+      "loss": 0.0136,
+      "step": 4139
+    },
+    {
+      "epoch": 295.7142857142857,
+      "grad_norm": 0.12158203125,
+      "learning_rate": 2.1910140517476113e-06,
+      "loss": 0.013,
+      "step": 4140
+    },
+    {
+      "epoch": 295.7857142857143,
+      "grad_norm": 0.15234375,
+      "learning_rate": 2.186081271521312e-06,
+      "loss": 0.0131,
+      "step": 4141
+    },
+    {
+      "epoch": 295.85714285714283,
+      "grad_norm": 0.251953125,
+      "learning_rate": 2.181153518173705e-06,
+      "loss": 0.011,
+      "step": 4142
+    },
+    {
+      "epoch": 295.92857142857144,
+      "grad_norm": 0.1767578125,
+      "learning_rate": 2.176230794106518e-06,
+      "loss": 0.0122,
+      "step": 4143
+    },
+    {
+      "epoch": 296.0,
+      "grad_norm": 0.203125,
+      "learning_rate": 2.1713131017190288e-06,
+      "loss": 0.0151,
+      "step": 4144
+    },
+    {
+      "epoch": 296.07142857142856,
+      "grad_norm": 0.2177734375,
+      "learning_rate": 2.1664004434080566e-06,
+      "loss": 0.0146,
+      "step": 4145
+    },
+    {
+      "epoch": 296.14285714285717,
+      "grad_norm": 0.1396484375,
+      "learning_rate": 2.161492821567976e-06,
+      "loss": 0.0136,
+      "step": 4146
+    },
+    {
+      "epoch": 296.2142857142857,
+      "grad_norm": 0.197265625,
+      "learning_rate": 2.156590238590704e-06,
+      "loss": 0.0141,
+      "step": 4147
+    },
+    {
+      "epoch": 296.2857142857143,
+      "grad_norm": 0.1796875,
+      "learning_rate": 2.1516926968656938e-06,
+      "loss": 0.0125,
+      "step": 4148
+    },
+    {
+      "epoch": 296.35714285714283,
+      "grad_norm": 0.11083984375,
+      "learning_rate": 2.1468001987799484e-06,
+      "loss": 0.0134,
+      "step": 4149
+    },
+    {
+      "epoch": 296.42857142857144,
+      "grad_norm": 0.197265625,
+      "learning_rate": 2.1419127467180155e-06,
+      "loss": 0.0159,
+      "step": 4150
+    },
+    {
+      "epoch": 296.5,
+      "grad_norm": 0.11865234375,
+      "learning_rate": 2.1370303430619796e-06,
+      "loss": 0.0134,
+      "step": 4151
+    },
+    {
+      "epoch": 296.57142857142856,
+      "grad_norm": 0.2080078125,
+      "learning_rate": 2.1321529901914634e-06,
+      "loss": 0.0119,
+      "step": 4152
+    },
+    {
+      "epoch": 296.64285714285717,
+      "grad_norm": 0.1357421875,
+      "learning_rate": 2.1272806904836295e-06,
+      "loss": 0.0157,
+      "step": 4153
+    },
+    {
+      "epoch": 296.7142857142857,
+      "grad_norm": 0.1142578125,
+      "learning_rate": 2.1224134463131774e-06,
+      "loss": 0.0121,
+      "step": 4154
+    },
+    {
+      "epoch": 296.7857142857143,
+      "grad_norm": 0.1875,
+      "learning_rate": 2.1175512600523404e-06,
+      "loss": 0.01,
+      "step": 4155
+    },
+    {
+      "epoch": 296.85714285714283,
+      "grad_norm": 0.1171875,
+      "learning_rate": 2.1126941340708986e-06,
+      "loss": 0.0123,
+      "step": 4156
+    },
+    {
+      "epoch": 296.92857142857144,
+      "grad_norm": 0.1435546875,
+      "learning_rate": 2.107842070736146e-06,
+      "loss": 0.0138,
+      "step": 4157
+    },
+    {
+      "epoch": 297.0,
+      "grad_norm": 0.1611328125,
+      "learning_rate": 2.102995072412922e-06,
+      "loss": 0.0132,
+      "step": 4158
+    },
+    {
+      "epoch": 297.07142857142856,
+      "grad_norm": 0.12890625,
+      "learning_rate": 2.0981531414635937e-06,
+      "loss": 0.0114,
+      "step": 4159
+    },
+    {
+      "epoch": 297.14285714285717,
+      "grad_norm": 0.13671875,
+      "learning_rate": 2.093316280248063e-06,
+      "loss": 0.0115,
+      "step": 4160
+    },
+    {
+      "epoch": 297.2142857142857,
+      "grad_norm": 0.11669921875,
+      "learning_rate": 2.088484491123757e-06,
+      "loss": 0.012,
+      "step": 4161
+    },
+    {
+      "epoch": 297.2857142857143,
+      "grad_norm": 0.15234375,
+      "learning_rate": 2.0836577764456285e-06,
+      "loss": 0.0138,
+      "step": 4162
+    },
+    {
+      "epoch": 297.35714285714283,
+      "grad_norm": 0.134765625,
+      "learning_rate": 2.0788361385661605e-06,
+      "loss": 0.0152,
+      "step": 4163
+    },
+    {
+      "epoch": 297.42857142857144,
+      "grad_norm": 0.287109375,
+      "learning_rate": 2.0740195798353608e-06,
+      "loss": 0.015,
+      "step": 4164
+    },
+    {
+      "epoch": 297.5,
+      "grad_norm": 0.16015625,
+      "learning_rate": 2.0692081026007628e-06,
+      "loss": 0.0113,
+      "step": 4165
+    },
+    {
+      "epoch": 297.57142857142856,
+      "grad_norm": 0.1806640625,
+      "learning_rate": 2.0644017092074196e-06,
+      "loss": 0.0119,
+      "step": 4166
+    },
+    {
+      "epoch": 297.64285714285717,
+      "grad_norm": 0.23046875,
+      "learning_rate": 2.0596004019979118e-06,
+      "loss": 0.011,
+      "step": 4167
+    },
+    {
+      "epoch": 297.7142857142857,
+      "grad_norm": 0.12255859375,
+      "learning_rate": 2.0548041833123326e-06,
+      "loss": 0.0126,
+      "step": 4168
+    },
+    {
+      "epoch": 297.7857142857143,
+      "grad_norm": 0.125,
+      "learning_rate": 2.0500130554883076e-06,
+      "loss": 0.0115,
+      "step": 4169
+    },
+    {
+      "epoch": 297.85714285714283,
+      "grad_norm": 0.140625,
+      "learning_rate": 2.045227020860971e-06,
+      "loss": 0.0127,
+      "step": 4170
+    },
+    {
+      "epoch": 297.92857142857144,
+      "grad_norm": 0.12060546875,
+      "learning_rate": 2.04044608176298e-06,
+      "loss": 0.0137,
+      "step": 4171
+    },
+    {
+      "epoch": 298.0,
+      "grad_norm": 0.1064453125,
+      "learning_rate": 2.035670240524504e-06,
+      "loss": 0.0136,
+      "step": 4172
+    },
+    {
+      "epoch": 298.07142857142856,
+      "grad_norm": 0.275390625,
+      "learning_rate": 2.0308994994732296e-06,
+      "loss": 0.0118,
+      "step": 4173
+    },
+    {
+      "epoch": 298.14285714285717,
+      "grad_norm": 0.1767578125,
+      "learning_rate": 2.0261338609343595e-06,
+      "loss": 0.0131,
+      "step": 4174
+    },
+    {
+      "epoch": 298.2142857142857,
+      "grad_norm": 0.1689453125,
+      "learning_rate": 2.021373327230608e-06,
+      "loss": 0.0145,
+      "step": 4175
+    },
+    {
+      "epoch": 298.2857142857143,
+      "grad_norm": 0.146484375,
+      "learning_rate": 2.0166179006822008e-06,
+      "loss": 0.011,
+      "step": 4176
+    },
+    {
+      "epoch": 298.35714285714283,
+      "grad_norm": 0.11962890625,
+      "learning_rate": 2.011867583606876e-06,
+      "loss": 0.01,
+      "step": 4177
+    },
+    {
+      "epoch": 298.42857142857144,
+      "grad_norm": 0.12109375,
+      "learning_rate": 2.0071223783198755e-06,
+      "loss": 0.0127,
+      "step": 4178
+    },
+    {
+      "epoch": 298.5,
+      "grad_norm": 0.224609375,
+      "learning_rate": 2.0023822871339637e-06,
+      "loss": 0.0138,
+      "step": 4179
+    },
+    {
+      "epoch": 298.57142857142856,
+      "grad_norm": 0.11572265625,
+      "learning_rate": 1.9976473123593976e-06,
+      "loss": 0.0139,
+      "step": 4180
+    },
+    {
+      "epoch": 298.64285714285717,
+      "grad_norm": 0.2099609375,
+      "learning_rate": 1.992917456303947e-06,
+      "loss": 0.0124,
+      "step": 4181
+    },
+    {
+      "epoch": 298.7142857142857,
+      "grad_norm": 0.1591796875,
+      "learning_rate": 1.988192721272887e-06,
+      "loss": 0.0121,
+      "step": 4182
+    },
+    {
+      "epoch": 298.7857142857143,
+      "grad_norm": 0.1416015625,
+      "learning_rate": 1.9834731095689964e-06,
+      "loss": 0.0151,
+      "step": 4183
+    },
+    {
+      "epoch": 298.85714285714283,
+      "grad_norm": 0.1279296875,
+      "learning_rate": 1.9787586234925547e-06,
+      "loss": 0.014,
+      "step": 4184
+    },
+    {
+      "epoch": 298.92857142857144,
+      "grad_norm": 0.14453125,
+      "learning_rate": 1.9740492653413475e-06,
+      "loss": 0.0111,
+      "step": 4185
+    },
+    {
+      "epoch": 299.0,
+      "grad_norm": 0.1748046875,
+      "learning_rate": 1.9693450374106565e-06,
+      "loss": 0.0134,
+      "step": 4186
+    },
+    {
+      "epoch": 299.07142857142856,
+      "grad_norm": 0.11962890625,
+      "learning_rate": 1.964645941993268e-06,
+      "loss": 0.0137,
+      "step": 4187
+    },
+    {
+      "epoch": 299.14285714285717,
+      "grad_norm": 0.1337890625,
+      "learning_rate": 1.9599519813794585e-06,
+      "loss": 0.0136,
+      "step": 4188
+    },
+    {
+      "epoch": 299.2142857142857,
+      "grad_norm": 0.1298828125,
+      "learning_rate": 1.955263157857015e-06,
+      "loss": 0.0147,
+      "step": 4189
+    },
+    {
+      "epoch": 299.2857142857143,
+      "grad_norm": 0.1298828125,
+      "learning_rate": 1.9505794737112087e-06,
+      "loss": 0.0118,
+      "step": 4190
+    },
+    {
+      "epoch": 299.35714285714283,
+      "grad_norm": 0.1630859375,
+      "learning_rate": 1.9459009312248144e-06,
+      "loss": 0.0131,
+      "step": 4191
+    },
+    {
+      "epoch": 299.42857142857144,
+      "grad_norm": 0.1630859375,
+      "learning_rate": 1.9412275326780872e-06,
+      "loss": 0.0133,
+      "step": 4192
+    },
+    {
+      "epoch": 299.5,
+      "grad_norm": 0.1328125,
+      "learning_rate": 1.9365592803487958e-06,
+      "loss": 0.0121,
+      "step": 4193
+    },
+    {
+      "epoch": 299.57142857142856,
+      "grad_norm": 0.158203125,
+      "learning_rate": 1.9318961765121837e-06,
+      "loss": 0.0098,
+      "step": 4194
+    },
+    {
+      "epoch": 299.64285714285717,
+      "grad_norm": 0.185546875,
+      "learning_rate": 1.9272382234409927e-06,
+      "loss": 0.0146,
+      "step": 4195
+    },
+    {
+      "epoch": 299.7142857142857,
+      "grad_norm": 0.1337890625,
+      "learning_rate": 1.922585423405451e-06,
+      "loss": 0.0133,
+      "step": 4196
+    },
+    {
+      "epoch": 299.7857142857143,
+      "grad_norm": 0.12109375,
+      "learning_rate": 1.917937778673276e-06,
+      "loss": 0.014,
+      "step": 4197
+    },
+    {
+      "epoch": 299.85714285714283,
+      "grad_norm": 0.10986328125,
+      "learning_rate": 1.9132952915096787e-06,
+      "loss": 0.011,
+      "step": 4198
+    },
+    {
+      "epoch": 299.92857142857144,
+      "grad_norm": 0.1474609375,
+      "learning_rate": 1.9086579641773475e-06,
+      "loss": 0.0119,
+      "step": 4199
+    },
+    {
+      "epoch": 300.0,
+      "grad_norm": 0.1533203125,
+      "learning_rate": 1.9040257989364624e-06,
+      "loss": 0.0119,
+      "step": 4200
+    },
+    {
+      "epoch": 300.07142857142856,
+      "grad_norm": 0.1962890625,
+      "learning_rate": 1.8993987980446755e-06,
+      "loss": 0.0132,
+      "step": 4201
+    },
+    {
+      "epoch": 300.14285714285717,
+      "grad_norm": 0.3828125,
+      "learning_rate": 1.8947769637571403e-06,
+      "loss": 0.0143,
+      "step": 4202
+    },
+    {
+      "epoch": 300.2142857142857,
+      "grad_norm": 0.1796875,
+      "learning_rate": 1.890160298326478e-06,
+      "loss": 0.0134,
+      "step": 4203
+    },
+    {
+      "epoch": 300.2857142857143,
+      "grad_norm": 0.134765625,
+      "learning_rate": 1.8855488040027971e-06,
+      "loss": 0.0115,
+      "step": 4204
+    },
+    {
+      "epoch": 300.35714285714283,
+      "grad_norm": 0.1201171875,
+      "learning_rate": 1.8809424830336847e-06,
+      "loss": 0.012,
+      "step": 4205
+    },
+    {
+      "epoch": 300.42857142857144,
+      "grad_norm": 0.1953125,
+      "learning_rate": 1.876341337664203e-06,
+      "loss": 0.014,
+      "step": 4206
+    },
+    {
+      "epoch": 300.5,
+      "grad_norm": 0.126953125,
+      "learning_rate": 1.8717453701368918e-06,
+      "loss": 0.0133,
+      "step": 4207
+    },
+    {
+      "epoch": 300.57142857142856,
+      "grad_norm": 0.126953125,
+      "learning_rate": 1.8671545826917783e-06,
+      "loss": 0.0149,
+      "step": 4208
+    },
+    {
+      "epoch": 300.64285714285717,
+      "grad_norm": 0.1240234375,
+      "learning_rate": 1.8625689775663535e-06,
+      "loss": 0.012,
+      "step": 4209
+    },
+    {
+      "epoch": 300.7142857142857,
+      "grad_norm": 0.138671875,
+      "learning_rate": 1.8579885569955824e-06,
+      "loss": 0.0123,
+      "step": 4210
+    },
+    {
+      "epoch": 300.7857142857143,
+      "grad_norm": 0.171875,
+      "learning_rate": 1.8534133232119042e-06,
+      "loss": 0.0135,
+      "step": 4211
+    },
+    {
+      "epoch": 300.85714285714283,
+      "grad_norm": 0.1943359375,
+      "learning_rate": 1.8488432784452406e-06,
+      "loss": 0.0124,
+      "step": 4212
+    },
+    {
+      "epoch": 300.92857142857144,
+      "grad_norm": 0.1552734375,
+      "learning_rate": 1.8442784249229696e-06,
+      "loss": 0.0129,
+      "step": 4213
+    },
+    {
+      "epoch": 301.0,
+      "grad_norm": 0.1806640625,
+      "learning_rate": 1.839718764869948e-06,
+      "loss": 0.0162,
+      "step": 4214
+    },
+    {
+      "epoch": 301.07142857142856,
+      "grad_norm": 0.25,
+      "learning_rate": 1.8351643005084975e-06,
+      "loss": 0.0136,
+      "step": 4215
+    },
+    {
+      "epoch": 301.14285714285717,
+      "grad_norm": 0.138671875,
+      "learning_rate": 1.8306150340584055e-06,
+      "loss": 0.0131,
+      "step": 4216
+    },
+    {
+      "epoch": 301.2142857142857,
+      "grad_norm": 0.15234375,
+      "learning_rate": 1.8260709677369369e-06,
+      "loss": 0.0141,
+      "step": 4217
+    },
+    {
+      "epoch": 301.2857142857143,
+      "grad_norm": 0.111328125,
+      "learning_rate": 1.821532103758812e-06,
+      "loss": 0.0114,
+      "step": 4218
+    },
+    {
+      "epoch": 301.35714285714283,
+      "grad_norm": 0.1552734375,
+      "learning_rate": 1.816998444336214e-06,
+      "loss": 0.0123,
+      "step": 4219
+    },
+    {
+      "epoch": 301.42857142857144,
+      "grad_norm": 0.1708984375,
+      "learning_rate": 1.8124699916787932e-06,
+      "loss": 0.0113,
+      "step": 4220
+    },
+    {
+      "epoch": 301.5,
+      "grad_norm": 0.1435546875,
+      "learning_rate": 1.8079467479936676e-06,
+      "loss": 0.0131,
+      "step": 4221
+    },
+    {
+      "epoch": 301.57142857142856,
+      "grad_norm": 0.12353515625,
+      "learning_rate": 1.803428715485407e-06,
+      "loss": 0.0141,
+      "step": 4222
+    },
+    {
+      "epoch": 301.64285714285717,
+      "grad_norm": 0.1376953125,
+      "learning_rate": 1.7989158963560487e-06,
+      "loss": 0.0129,
+      "step": 4223
+    },
+    {
+      "epoch": 301.7142857142857,
+      "grad_norm": 0.2021484375,
+      "learning_rate": 1.7944082928050846e-06,
+      "loss": 0.0124,
+      "step": 4224
+    },
+    {
+      "epoch": 301.7857142857143,
+      "grad_norm": 0.146484375,
+      "learning_rate": 1.7899059070294642e-06,
+      "loss": 0.0143,
+      "step": 4225
+    },
+    {
+      "epoch": 301.85714285714283,
+      "grad_norm": 0.2353515625,
+      "learning_rate": 1.7854087412235974e-06,
+      "loss": 0.0146,
+      "step": 4226
+    },
+    {
+      "epoch": 301.92857142857144,
+      "grad_norm": 0.337890625,
+      "learning_rate": 1.780916797579349e-06,
+      "loss": 0.0136,
+      "step": 4227
+    },
+    {
+      "epoch": 302.0,
+      "grad_norm": 0.19140625,
+      "learning_rate": 1.7764300782860357e-06,
+      "loss": 0.0145,
+      "step": 4228
+    },
+    {
+      "epoch": 302.07142857142856,
+      "grad_norm": 0.1748046875,
+      "learning_rate": 1.7719485855304308e-06,
+      "loss": 0.0129,
+      "step": 4229
+    },
+    {
+      "epoch": 302.14285714285717,
+      "grad_norm": 0.1416015625,
+      "learning_rate": 1.7674723214967561e-06,
+      "loss": 0.0139,
+      "step": 4230
+    },
+    {
+      "epoch": 302.2142857142857,
+      "grad_norm": 0.1435546875,
+      "learning_rate": 1.7630012883666947e-06,
+      "loss": 0.0101,
+      "step": 4231
+    },
+    {
+      "epoch": 302.2857142857143,
+      "grad_norm": 0.1875,
+      "learning_rate": 1.7585354883193712e-06,
+      "loss": 0.015,
+      "step": 4232
+    },
+    {
+      "epoch": 302.35714285714283,
+      "grad_norm": 0.125,
+      "learning_rate": 1.7540749235313625e-06,
+      "loss": 0.0116,
+      "step": 4233
+    },
+    {
+      "epoch": 302.42857142857144,
+      "grad_norm": 0.1162109375,
+      "learning_rate": 1.7496195961766922e-06,
+      "loss": 0.0133,
+      "step": 4234
+    },
+    {
+      "epoch": 302.5,
+      "grad_norm": 0.1728515625,
+      "learning_rate": 1.7451695084268332e-06,
+      "loss": 0.0119,
+      "step": 4235
+    },
+    {
+      "epoch": 302.57142857142856,
+      "grad_norm": 0.16796875,
+      "learning_rate": 1.7407246624507058e-06,
+      "loss": 0.0141,
+      "step": 4236
+    },
+    {
+      "epoch": 302.64285714285717,
+      "grad_norm": 0.1181640625,
+      "learning_rate": 1.7362850604146718e-06,
+      "loss": 0.0136,
+      "step": 4237
+    },
+    {
+      "epoch": 302.7142857142857,
+      "grad_norm": 0.2109375,
+      "learning_rate": 1.731850704482542e-06,
+      "loss": 0.0105,
+      "step": 4238
+    },
+    {
+      "epoch": 302.7857142857143,
+      "grad_norm": 0.1416015625,
+      "learning_rate": 1.7274215968155635e-06,
+      "loss": 0.0167,
+      "step": 4239
+    },
+    {
+      "epoch": 302.85714285714283,
+      "grad_norm": 0.166015625,
+      "learning_rate": 1.7229977395724308e-06,
+      "loss": 0.0129,
+      "step": 4240
+    },
+    {
+      "epoch": 302.92857142857144,
+      "grad_norm": 0.146484375,
+      "learning_rate": 1.7185791349092817e-06,
+      "loss": 0.0151,
+      "step": 4241
+    },
+    {
+      "epoch": 303.0,
+      "grad_norm": 0.294921875,
+      "learning_rate": 1.7141657849796876e-06,
+      "loss": 0.0134,
+      "step": 4242
+    },
+    {
+      "epoch": 303.07142857142856,
+      "grad_norm": 0.1328125,
+      "learning_rate": 1.709757691934663e-06,
+      "loss": 0.0146,
+      "step": 4243
+    },
+    {
+      "epoch": 303.14285714285717,
+      "grad_norm": 0.1474609375,
+      "learning_rate": 1.7053548579226572e-06,
+      "loss": 0.0135,
+      "step": 4244
+    },
+    {
+      "epoch": 303.2142857142857,
+      "grad_norm": 0.1337890625,
+      "learning_rate": 1.7009572850895592e-06,
+      "loss": 0.0146,
+      "step": 4245
+    },
+    {
+      "epoch": 303.2857142857143,
+      "grad_norm": 0.150390625,
+      "learning_rate": 1.696564975578692e-06,
+      "loss": 0.0145,
+      "step": 4246
+    },
+    {
+      "epoch": 303.35714285714283,
+      "grad_norm": 0.1435546875,
+      "learning_rate": 1.6921779315308154e-06,
+      "loss": 0.0131,
+      "step": 4247
+    },
+    {
+      "epoch": 303.42857142857144,
+      "grad_norm": 0.16015625,
+      "learning_rate": 1.6877961550841177e-06,
+      "loss": 0.0141,
+      "step": 4248
+    },
+    {
+      "epoch": 303.5,
+      "grad_norm": 0.1650390625,
+      "learning_rate": 1.683419648374225e-06,
+      "loss": 0.0139,
+      "step": 4249
+    },
+    {
+      "epoch": 303.57142857142856,
+      "grad_norm": 0.134765625,
+      "learning_rate": 1.679048413534197e-06,
+      "loss": 0.0131,
+      "step": 4250
+    },
+    {
+      "epoch": 303.64285714285717,
+      "grad_norm": 0.1875,
+      "learning_rate": 1.6746824526945163e-06,
+      "loss": 0.0127,
+      "step": 4251
+    },
+    {
+      "epoch": 303.7142857142857,
+      "grad_norm": 0.1318359375,
+      "learning_rate": 1.6703217679831018e-06,
+      "loss": 0.0138,
+      "step": 4252
+    },
+    {
+      "epoch": 303.7857142857143,
+      "grad_norm": 0.1328125,
+      "learning_rate": 1.665966361525298e-06,
+      "loss": 0.013,
+      "step": 4253
+    },
+    {
+      "epoch": 303.85714285714283,
+      "grad_norm": 0.2041015625,
+      "learning_rate": 1.6616162354438747e-06,
+      "loss": 0.0143,
+      "step": 4254
+    },
+    {
+      "epoch": 303.92857142857144,
+      "grad_norm": 0.1416015625,
+      "learning_rate": 1.6572713918590332e-06,
+      "loss": 0.0131,
+      "step": 4255
+    },
+    {
+      "epoch": 304.0,
+      "grad_norm": 0.1357421875,
+      "learning_rate": 1.6529318328883965e-06,
+      "loss": 0.0126,
+      "step": 4256
+    },
+    {
+      "epoch": 304.07142857142856,
+      "grad_norm": 0.125,
+      "learning_rate": 1.648597560647011e-06,
+      "loss": 0.0115,
+      "step": 4257
+    },
+    {
+      "epoch": 304.14285714285717,
+      "grad_norm": 0.134765625,
+      "learning_rate": 1.6442685772473515e-06,
+      "loss": 0.0116,
+      "step": 4258
+    },
+    {
+      "epoch": 304.2142857142857,
+      "grad_norm": 0.1796875,
+      "learning_rate": 1.6399448847993077e-06,
+      "loss": 0.0139,
+      "step": 4259
+    },
+    {
+      "epoch": 304.2857142857143,
+      "grad_norm": 0.1455078125,
+      "learning_rate": 1.6356264854102008e-06,
+      "loss": 0.0124,
+      "step": 4260
+    },
+    {
+      "epoch": 304.35714285714283,
+      "grad_norm": 0.177734375,
+      "learning_rate": 1.6313133811847626e-06,
+      "loss": 0.012,
+      "step": 4261
+    },
+    {
+      "epoch": 304.42857142857144,
+      "grad_norm": 0.10302734375,
+      "learning_rate": 1.6270055742251502e-06,
+      "loss": 0.011,
+      "step": 4262
+    },
+    {
+      "epoch": 304.5,
+      "grad_norm": 0.1533203125,
+      "learning_rate": 1.622703066630929e-06,
+      "loss": 0.0129,
+      "step": 4263
+    },
+    {
+      "epoch": 304.57142857142856,
+      "grad_norm": 0.1884765625,
+      "learning_rate": 1.6184058604990976e-06,
+      "loss": 0.013,
+      "step": 4264
+    },
+    {
+      "epoch": 304.64285714285717,
+      "grad_norm": 0.302734375,
+      "learning_rate": 1.6141139579240586e-06,
+      "loss": 0.0097,
+      "step": 4265
+    },
+    {
+      "epoch": 304.7142857142857,
+      "grad_norm": 0.1357421875,
+      "learning_rate": 1.609827360997633e-06,
+      "loss": 0.0131,
+      "step": 4266
+    },
+    {
+      "epoch": 304.7857142857143,
+      "grad_norm": 0.2216796875,
+      "learning_rate": 1.6055460718090567e-06,
+      "loss": 0.0133,
+      "step": 4267
+    },
+    {
+      "epoch": 304.85714285714283,
+      "grad_norm": 0.1337890625,
+      "learning_rate": 1.6012700924449745e-06,
+      "loss": 0.011,
+      "step": 4268
+    },
+    {
+      "epoch": 304.92857142857144,
+      "grad_norm": 0.2431640625,
+      "learning_rate": 1.5969994249894539e-06,
+      "loss": 0.0151,
+      "step": 4269
+    },
+    {
+      "epoch": 305.0,
+      "grad_norm": 0.1416015625,
+      "learning_rate": 1.5927340715239602e-06,
+      "loss": 0.0131,
+      "step": 4270
+    },
+    {
+      "epoch": 305.07142857142856,
+      "grad_norm": 0.1796875,
+      "learning_rate": 1.588474034127381e-06,
+      "loss": 0.013,
+      "step": 4271
+    },
+    {
+      "epoch": 305.14285714285717,
+      "grad_norm": 0.1474609375,
+      "learning_rate": 1.584219314876001e-06,
+      "loss": 0.0122,
+      "step": 4272
+    },
+    {
+      "epoch": 305.2142857142857,
+      "grad_norm": 0.1572265625,
+      "learning_rate": 1.5799699158435168e-06,
+      "loss": 0.0136,
+      "step": 4273
+    },
+    {
+      "epoch": 305.2857142857143,
+      "grad_norm": 0.12158203125,
+      "learning_rate": 1.57572583910104e-06,
+      "loss": 0.0139,
+      "step": 4274
+    },
+    {
+      "epoch": 305.35714285714283,
+      "grad_norm": 0.1533203125,
+      "learning_rate": 1.57148708671708e-06,
+      "loss": 0.0141,
+      "step": 4275
+    },
+    {
+      "epoch": 305.42857142857144,
+      "grad_norm": 0.130859375,
+      "learning_rate": 1.5672536607575533e-06,
+      "loss": 0.0128,
+      "step": 4276
+    },
+    {
+      "epoch": 305.5,
+      "grad_norm": 0.1708984375,
+      "learning_rate": 1.5630255632857804e-06,
+      "loss": 0.0133,
+      "step": 4277
+    },
+    {
+      "epoch": 305.57142857142856,
+      "grad_norm": 0.203125,
+      "learning_rate": 1.5588027963624797e-06,
+      "loss": 0.0127,
+      "step": 4278
+    },
+    {
+      "epoch": 305.64285714285717,
+      "grad_norm": 0.1123046875,
+      "learning_rate": 1.554585362045785e-06,
+      "loss": 0.0135,
+      "step": 4279
+    },
+    {
+      "epoch": 305.7142857142857,
+      "grad_norm": 0.212890625,
+      "learning_rate": 1.5503732623912206e-06,
+      "loss": 0.0128,
+      "step": 4280
+    },
+    {
+      "epoch": 305.7857142857143,
+      "grad_norm": 0.1396484375,
+      "learning_rate": 1.546166499451708e-06,
+      "loss": 0.0114,
+      "step": 4281
+    },
+    {
+      "epoch": 305.85714285714283,
+      "grad_norm": 0.1806640625,
+      "learning_rate": 1.541965075277571e-06,
+      "loss": 0.0148,
+      "step": 4282
+    },
+    {
+      "epoch": 305.92857142857144,
+      "grad_norm": 0.11279296875,
+      "learning_rate": 1.537768991916541e-06,
+      "loss": 0.0122,
+      "step": 4283
+    },
+    {
+      "epoch": 306.0,
+      "grad_norm": 0.14453125,
+      "learning_rate": 1.5335782514137322e-06,
+      "loss": 0.0137,
+      "step": 4284
+    },
+    {
+      "epoch": 306.07142857142856,
+      "grad_norm": 0.1484375,
+      "learning_rate": 1.5293928558116635e-06,
+      "loss": 0.013,
+      "step": 4285
+    },
+    {
+      "epoch": 306.14285714285717,
+      "grad_norm": 0.1416015625,
+      "learning_rate": 1.5252128071502433e-06,
+      "loss": 0.0133,
+      "step": 4286
+    },
+    {
+      "epoch": 306.2142857142857,
+      "grad_norm": 0.12158203125,
+      "learning_rate": 1.5210381074667782e-06,
+      "loss": 0.0132,
+      "step": 4287
+    },
+    {
+      "epoch": 306.2857142857143,
+      "grad_norm": 0.1767578125,
+      "learning_rate": 1.5168687587959653e-06,
+      "loss": 0.0113,
+      "step": 4288
+    },
+    {
+      "epoch": 306.35714285714283,
+      "grad_norm": 0.15234375,
+      "learning_rate": 1.5127047631698961e-06,
+      "loss": 0.011,
+      "step": 4289
+    },
+    {
+      "epoch": 306.42857142857144,
+      "grad_norm": 0.1923828125,
+      "learning_rate": 1.5085461226180487e-06,
+      "loss": 0.0121,
+      "step": 4290
+    },
+    {
+      "epoch": 306.5,
+      "grad_norm": 0.171875,
+      "learning_rate": 1.5043928391672957e-06,
+      "loss": 0.0127,
+      "step": 4291
+    },
+    {
+      "epoch": 306.57142857142856,
+      "grad_norm": 0.1796875,
+      "learning_rate": 1.5002449148418946e-06,
+      "loss": 0.0158,
+      "step": 4292
+    },
+    {
+      "epoch": 306.64285714285717,
+      "grad_norm": 0.11474609375,
+      "learning_rate": 1.4961023516634975e-06,
+      "loss": 0.0119,
+      "step": 4293
+    },
+    {
+      "epoch": 306.7142857142857,
+      "grad_norm": 0.158203125,
+      "learning_rate": 1.4919651516511381e-06,
+      "loss": 0.0165,
+      "step": 4294
+    },
+    {
+      "epoch": 306.7857142857143,
+      "grad_norm": 0.1904296875,
+      "learning_rate": 1.4878333168212354e-06,
+      "loss": 0.0144,
+      "step": 4295
+    },
+    {
+      "epoch": 306.85714285714283,
+      "grad_norm": 0.212890625,
+      "learning_rate": 1.483706849187598e-06,
+      "loss": 0.0144,
+      "step": 4296
+    },
+    {
+      "epoch": 306.92857142857144,
+      "grad_norm": 0.1123046875,
+      "learning_rate": 1.4795857507614142e-06,
+      "loss": 0.0136,
+      "step": 4297
+    },
+    {
+      "epoch": 307.0,
+      "grad_norm": 0.2255859375,
+      "learning_rate": 1.475470023551259e-06,
+      "loss": 0.0163,
+      "step": 4298
+    },
+    {
+      "epoch": 307.07142857142856,
+      "grad_norm": 0.1240234375,
+      "learning_rate": 1.4713596695630854e-06,
+      "loss": 0.0134,
+      "step": 4299
+    },
+    {
+      "epoch": 307.14285714285717,
+      "grad_norm": 0.1142578125,
+      "learning_rate": 1.4672546908002321e-06,
+      "loss": 0.0126,
+      "step": 4300
+    },
+    {
+      "epoch": 307.2142857142857,
+      "grad_norm": 0.13671875,
+      "learning_rate": 1.4631550892634127e-06,
+      "loss": 0.0126,
+      "step": 4301
+    },
+    {
+      "epoch": 307.2857142857143,
+      "grad_norm": 0.302734375,
+      "learning_rate": 1.459060866950729e-06,
+      "loss": 0.0135,
+      "step": 4302
+    },
+    {
+      "epoch": 307.35714285714283,
+      "grad_norm": 0.2177734375,
+      "learning_rate": 1.4549720258576496e-06,
+      "loss": 0.0125,
+      "step": 4303
+    },
+    {
+      "epoch": 307.42857142857144,
+      "grad_norm": 0.17578125,
+      "learning_rate": 1.450888567977031e-06,
+      "loss": 0.0106,
+      "step": 4304
+    },
+    {
+      "epoch": 307.5,
+      "grad_norm": 0.1708984375,
+      "learning_rate": 1.446810495299096e-06,
+      "loss": 0.0121,
+      "step": 4305
+    },
+    {
+      "epoch": 307.57142857142856,
+      "grad_norm": 0.1728515625,
+      "learning_rate": 1.4427378098114508e-06,
+      "loss": 0.014,
+      "step": 4306
+    },
+    {
+      "epoch": 307.64285714285717,
+      "grad_norm": 0.14453125,
+      "learning_rate": 1.4386705134990728e-06,
+      "loss": 0.0153,
+      "step": 4307
+    },
+    {
+      "epoch": 307.7142857142857,
+      "grad_norm": 0.1552734375,
+      "learning_rate": 1.4346086083443107e-06,
+      "loss": 0.0141,
+      "step": 4308
+    },
+    {
+      "epoch": 307.7857142857143,
+      "grad_norm": 0.26953125,
+      "learning_rate": 1.4305520963268897e-06,
+      "loss": 0.0151,
+      "step": 4309
+    },
+    {
+      "epoch": 307.85714285714283,
+      "grad_norm": 0.1669921875,
+      "learning_rate": 1.4265009794239045e-06,
+      "loss": 0.0133,
+      "step": 4310
+    },
+    {
+      "epoch": 307.92857142857144,
+      "grad_norm": 0.15234375,
+      "learning_rate": 1.4224552596098158e-06,
+      "loss": 0.0125,
+      "step": 4311
+    },
+    {
+      "epoch": 308.0,
+      "grad_norm": 0.1416015625,
+      "learning_rate": 1.4184149388564647e-06,
+      "loss": 0.0144,
+      "step": 4312
+    },
+    {
+      "epoch": 308.07142857142856,
+      "grad_norm": 0.2412109375,
+      "learning_rate": 1.414380019133052e-06,
+      "loss": 0.0155,
+      "step": 4313
+    },
+    {
+      "epoch": 308.14285714285717,
+      "grad_norm": 0.103515625,
+      "learning_rate": 1.4103505024061477e-06,
+      "loss": 0.013,
+      "step": 4314
+    },
+    {
+      "epoch": 308.2142857142857,
+      "grad_norm": 0.1494140625,
+      "learning_rate": 1.406326390639691e-06,
+      "loss": 0.0133,
+      "step": 4315
+    },
+    {
+      "epoch": 308.2857142857143,
+      "grad_norm": 0.142578125,
+      "learning_rate": 1.4023076857949846e-06,
+      "loss": 0.0147,
+      "step": 4316
+    },
+    {
+      "epoch": 308.35714285714283,
+      "grad_norm": 0.1396484375,
+      "learning_rate": 1.3982943898306942e-06,
+      "loss": 0.012,
+      "step": 4317
+    },
+    {
+      "epoch": 308.42857142857144,
+      "grad_norm": 0.11962890625,
+      "learning_rate": 1.3942865047028547e-06,
+      "loss": 0.0114,
+      "step": 4318
+    },
+    {
+      "epoch": 308.5,
+      "grad_norm": 0.15234375,
+      "learning_rate": 1.3902840323648581e-06,
+      "loss": 0.0129,
+      "step": 4319
+    },
+    {
+      "epoch": 308.57142857142856,
+      "grad_norm": 0.189453125,
+      "learning_rate": 1.386286974767459e-06,
+      "loss": 0.0121,
+      "step": 4320
+    },
+    {
+      "epoch": 308.64285714285717,
+      "grad_norm": 0.130859375,
+      "learning_rate": 1.382295333858781e-06,
+      "loss": 0.0149,
+      "step": 4321
+    },
+    {
+      "epoch": 308.7142857142857,
+      "grad_norm": 0.1337890625,
+      "learning_rate": 1.3783091115842972e-06,
+      "loss": 0.0113,
+      "step": 4322
+    },
+    {
+      "epoch": 308.7857142857143,
+      "grad_norm": 0.1572265625,
+      "learning_rate": 1.3743283098868437e-06,
+      "loss": 0.0129,
+      "step": 4323
+    },
+    {
+      "epoch": 308.85714285714283,
+      "grad_norm": 0.2392578125,
+      "learning_rate": 1.3703529307066198e-06,
+      "loss": 0.0145,
+      "step": 4324
+    },
+    {
+      "epoch": 308.92857142857144,
+      "grad_norm": 0.1474609375,
+      "learning_rate": 1.3663829759811675e-06,
+      "loss": 0.0136,
+      "step": 4325
+    },
+    {
+      "epoch": 309.0,
+      "grad_norm": 0.169921875,
+      "learning_rate": 1.3624184476454027e-06,
+      "loss": 0.0136,
+      "step": 4326
+    },
+    {
+      "epoch": 309.07142857142856,
+      "grad_norm": 0.1396484375,
+      "learning_rate": 1.3584593476315855e-06,
+      "loss": 0.0146,
+      "step": 4327
+    },
+    {
+      "epoch": 309.14285714285717,
+      "grad_norm": 0.1376953125,
+      "learning_rate": 1.3545056778693333e-06,
+      "loss": 0.0143,
+      "step": 4328
+    },
+    {
+      "epoch": 309.2142857142857,
+      "grad_norm": 0.1318359375,
+      "learning_rate": 1.3505574402856157e-06,
+      "loss": 0.0125,
+      "step": 4329
+    },
+    {
+      "epoch": 309.2857142857143,
+      "grad_norm": 0.162109375,
+      "learning_rate": 1.3466146368047564e-06,
+      "loss": 0.0111,
+      "step": 4330
+    },
+    {
+      "epoch": 309.35714285714283,
+      "grad_norm": 0.1357421875,
+      "learning_rate": 1.3426772693484303e-06,
+      "loss": 0.0145,
+      "step": 4331
+    },
+    {
+      "epoch": 309.42857142857144,
+      "grad_norm": 0.1611328125,
+      "learning_rate": 1.3387453398356647e-06,
+      "loss": 0.0149,
+      "step": 4332
+    },
+    {
+      "epoch": 309.5,
+      "grad_norm": 0.09423828125,
+      "learning_rate": 1.3348188501828326e-06,
+      "loss": 0.0113,
+      "step": 4333
+    },
+    {
+      "epoch": 309.57142857142856,
+      "grad_norm": 0.2109375,
+      "learning_rate": 1.3308978023036532e-06,
+      "loss": 0.0129,
+      "step": 4334
+    },
+    {
+      "epoch": 309.64285714285717,
+      "grad_norm": 0.1279296875,
+      "learning_rate": 1.3269821981092028e-06,
+      "loss": 0.0147,
+      "step": 4335
+    },
+    {
+      "epoch": 309.7142857142857,
+      "grad_norm": 0.11181640625,
+      "learning_rate": 1.3230720395078985e-06,
+      "loss": 0.0119,
+      "step": 4336
+    },
+    {
+      "epoch": 309.7857142857143,
+      "grad_norm": 0.150390625,
+      "learning_rate": 1.3191673284055031e-06,
+      "loss": 0.0126,
+      "step": 4337
+    },
+    {
+      "epoch": 309.85714285714283,
+      "grad_norm": 0.1455078125,
+      "learning_rate": 1.3152680667051248e-06,
+      "loss": 0.0137,
+      "step": 4338
+    },
+    {
+      "epoch": 309.92857142857144,
+      "grad_norm": 0.1240234375,
+      "learning_rate": 1.3113742563072185e-06,
+      "loss": 0.0141,
+      "step": 4339
+    },
+    {
+      "epoch": 310.0,
+      "grad_norm": 0.1826171875,
+      "learning_rate": 1.3074858991095743e-06,
+      "loss": 0.0154,
+      "step": 4340
+    },
+    {
+      "epoch": 310.07142857142856,
+      "grad_norm": 0.1826171875,
+      "learning_rate": 1.3036029970073388e-06,
+      "loss": 0.0147,
+      "step": 4341
+    },
+    {
+      "epoch": 310.14285714285717,
+      "grad_norm": 0.197265625,
+      "learning_rate": 1.2997255518929886e-06,
+      "loss": 0.0138,
+      "step": 4342
+    },
+    {
+      "epoch": 310.2142857142857,
+      "grad_norm": 0.228515625,
+      "learning_rate": 1.2958535656563408e-06,
+      "loss": 0.0141,
+      "step": 4343
+    },
+    {
+      "epoch": 310.2857142857143,
+      "grad_norm": 0.13671875,
+      "learning_rate": 1.2919870401845516e-06,
+      "loss": 0.0135,
+      "step": 4344
+    },
+    {
+      "epoch": 310.35714285714283,
+      "grad_norm": 0.1259765625,
+      "learning_rate": 1.2881259773621257e-06,
+      "loss": 0.0132,
+      "step": 4345
+    },
+    {
+      "epoch": 310.42857142857144,
+      "grad_norm": 0.1630859375,
+      "learning_rate": 1.2842703790708953e-06,
+      "loss": 0.0122,
+      "step": 4346
+    },
+    {
+      "epoch": 310.5,
+      "grad_norm": 0.1357421875,
+      "learning_rate": 1.2804202471900311e-06,
+      "loss": 0.0145,
+      "step": 4347
+    },
+    {
+      "epoch": 310.57142857142856,
+      "grad_norm": 0.12353515625,
+      "learning_rate": 1.276575583596043e-06,
+      "loss": 0.0148,
+      "step": 4348
+    },
+    {
+      "epoch": 310.64285714285717,
+      "grad_norm": 0.146484375,
+      "learning_rate": 1.2727363901627682e-06,
+      "loss": 0.0112,
+      "step": 4349
+    },
+    {
+      "epoch": 310.7142857142857,
+      "grad_norm": 0.1376953125,
+      "learning_rate": 1.2689026687613918e-06,
+      "loss": 0.0125,
+      "step": 4350
+    },
+    {
+      "epoch": 310.7857142857143,
+      "grad_norm": 0.1513671875,
+      "learning_rate": 1.2650744212604148e-06,
+      "loss": 0.0133,
+      "step": 4351
+    },
+    {
+      "epoch": 310.85714285714283,
+      "grad_norm": 0.1748046875,
+      "learning_rate": 1.2612516495256813e-06,
+      "loss": 0.0143,
+      "step": 4352
+    },
+    {
+      "epoch": 310.92857142857144,
+      "grad_norm": 0.130859375,
+      "learning_rate": 1.2574343554203615e-06,
+      "loss": 0.0109,
+      "step": 4353
+    },
+    {
+      "epoch": 311.0,
+      "grad_norm": 0.15625,
+      "learning_rate": 1.2536225408049633e-06,
+      "loss": 0.0132,
+      "step": 4354
+    },
+    {
+      "epoch": 311.07142857142856,
+      "grad_norm": 0.1103515625,
+      "learning_rate": 1.2498162075373174e-06,
+      "loss": 0.0118,
+      "step": 4355
+    },
+    {
+      "epoch": 311.14285714285717,
+      "grad_norm": 0.2578125,
+      "learning_rate": 1.246015357472581e-06,
+      "loss": 0.0144,
+      "step": 4356
+    },
+    {
+      "epoch": 311.2142857142857,
+      "grad_norm": 0.1435546875,
+      "learning_rate": 1.2422199924632464e-06,
+      "loss": 0.0136,
+      "step": 4357
+    },
+    {
+      "epoch": 311.2857142857143,
+      "grad_norm": 0.1494140625,
+      "learning_rate": 1.238430114359125e-06,
+      "loss": 0.0148,
+      "step": 4358
+    },
+    {
+      "epoch": 311.35714285714283,
+      "grad_norm": 0.1357421875,
+      "learning_rate": 1.2346457250073592e-06,
+      "loss": 0.0124,
+      "step": 4359
+    },
+    {
+      "epoch": 311.42857142857144,
+      "grad_norm": 0.19140625,
+      "learning_rate": 1.230866826252415e-06,
+      "loss": 0.0125,
+      "step": 4360
+    },
+    {
+      "epoch": 311.5,
+      "grad_norm": 0.1201171875,
+      "learning_rate": 1.227093419936079e-06,
+      "loss": 0.0128,
+      "step": 4361
+    },
+    {
+      "epoch": 311.57142857142856,
+      "grad_norm": 0.13671875,
+      "learning_rate": 1.223325507897466e-06,
+      "loss": 0.0153,
+      "step": 4362
+    },
+    {
+      "epoch": 311.64285714285717,
+      "grad_norm": 0.146484375,
+      "learning_rate": 1.2195630919730083e-06,
+      "loss": 0.0117,
+      "step": 4363
+    },
+    {
+      "epoch": 311.7142857142857,
+      "grad_norm": 0.1494140625,
+      "learning_rate": 1.2158061739964643e-06,
+      "loss": 0.0133,
+      "step": 4364
+    },
+    {
+      "epoch": 311.7857142857143,
+      "grad_norm": 0.13671875,
+      "learning_rate": 1.212054755798911e-06,
+      "loss": 0.0114,
+      "step": 4365
+    },
+    {
+      "epoch": 311.85714285714283,
+      "grad_norm": 0.279296875,
+      "learning_rate": 1.2083088392087398e-06,
+      "loss": 0.0145,
+      "step": 4366
+    },
+    {
+      "epoch": 311.92857142857144,
+      "grad_norm": 0.265625,
+      "learning_rate": 1.2045684260516693e-06,
+      "loss": 0.0143,
+      "step": 4367
+    },
+    {
+      "epoch": 312.0,
+      "grad_norm": 0.2109375,
+      "learning_rate": 1.2008335181507291e-06,
+      "loss": 0.0126,
+      "step": 4368
+    },
+    {
+      "epoch": 312.07142857142856,
+      "grad_norm": 0.12158203125,
+      "learning_rate": 1.197104117326267e-06,
+      "loss": 0.0132,
+      "step": 4369
+    },
+    {
+      "epoch": 312.14285714285717,
+      "grad_norm": 0.181640625,
+      "learning_rate": 1.19338022539595e-06,
+      "loss": 0.0139,
+      "step": 4370
+    },
+    {
+      "epoch": 312.2142857142857,
+      "grad_norm": 0.142578125,
+      "learning_rate": 1.189661844174758e-06,
+      "loss": 0.0127,
+      "step": 4371
+    },
+    {
+      "epoch": 312.2857142857143,
+      "grad_norm": 0.1552734375,
+      "learning_rate": 1.185948975474982e-06,
+      "loss": 0.0128,
+      "step": 4372
+    },
+    {
+      "epoch": 312.35714285714283,
+      "grad_norm": 0.1875,
+      "learning_rate": 1.182241621106228e-06,
+      "loss": 0.0172,
+      "step": 4373
+    },
+    {
+      "epoch": 312.42857142857144,
+      "grad_norm": 0.23828125,
+      "learning_rate": 1.1785397828754222e-06,
+      "loss": 0.0146,
+      "step": 4374
+    },
+    {
+      "epoch": 312.5,
+      "grad_norm": 0.181640625,
+      "learning_rate": 1.1748434625867902e-06,
+      "loss": 0.0134,
+      "step": 4375
+    },
+    {
+      "epoch": 312.57142857142856,
+      "grad_norm": 0.1328125,
+      "learning_rate": 1.1711526620418757e-06,
+      "loss": 0.01,
+      "step": 4376
+    },
+    {
+      "epoch": 312.64285714285717,
+      "grad_norm": 0.21875,
+      "learning_rate": 1.1674673830395291e-06,
+      "loss": 0.0148,
+      "step": 4377
+    },
+    {
+      "epoch": 312.7142857142857,
+      "grad_norm": 0.14453125,
+      "learning_rate": 1.1637876273759106e-06,
+      "loss": 0.0142,
+      "step": 4378
+    },
+    {
+      "epoch": 312.7857142857143,
+      "grad_norm": 0.15234375,
+      "learning_rate": 1.1601133968444893e-06,
+      "loss": 0.0148,
+      "step": 4379
+    },
+    {
+      "epoch": 312.85714285714283,
+      "grad_norm": 0.1337890625,
+      "learning_rate": 1.156444693236039e-06,
+      "loss": 0.0127,
+      "step": 4380
+    },
+    {
+      "epoch": 312.92857142857144,
+      "grad_norm": 0.1640625,
+      "learning_rate": 1.1527815183386428e-06,
+      "loss": 0.0116,
+      "step": 4381
+    },
+    {
+      "epoch": 313.0,
+      "grad_norm": 0.1884765625,
+      "learning_rate": 1.1491238739376852e-06,
+      "loss": 0.0139,
+      "step": 4382
+    },
+    {
+      "epoch": 313.07142857142856,
+      "grad_norm": 0.2138671875,
+      "learning_rate": 1.1454717618158611e-06,
+      "loss": 0.0122,
+      "step": 4383
+    },
+    {
+      "epoch": 313.14285714285717,
+      "grad_norm": 0.1533203125,
+      "learning_rate": 1.1418251837531633e-06,
+      "loss": 0.0114,
+      "step": 4384
+    },
+    {
+      "epoch": 313.2142857142857,
+      "grad_norm": 0.158203125,
+      "learning_rate": 1.1381841415268906e-06,
+      "loss": 0.0139,
+      "step": 4385
+    },
+    {
+      "epoch": 313.2857142857143,
+      "grad_norm": 0.140625,
+      "learning_rate": 1.1345486369116435e-06,
+      "loss": 0.0157,
+      "step": 4386
+    },
+    {
+      "epoch": 313.35714285714283,
+      "grad_norm": 0.134765625,
+      "learning_rate": 1.1309186716793208e-06,
+      "loss": 0.0154,
+      "step": 4387
+    },
+    {
+      "epoch": 313.42857142857144,
+      "grad_norm": 0.2158203125,
+      "learning_rate": 1.1272942475991242e-06,
+      "loss": 0.0156,
+      "step": 4388
+    },
+    {
+      "epoch": 313.5,
+      "grad_norm": 0.205078125,
+      "learning_rate": 1.1236753664375543e-06,
+      "loss": 0.0136,
+      "step": 4389
+    },
+    {
+      "epoch": 313.57142857142856,
+      "grad_norm": 0.189453125,
+      "learning_rate": 1.1200620299584102e-06,
+      "loss": 0.0145,
+      "step": 4390
+    },
+    {
+      "epoch": 313.64285714285717,
+      "grad_norm": 0.1572265625,
+      "learning_rate": 1.1164542399227873e-06,
+      "loss": 0.0149,
+      "step": 4391
+    },
+    {
+      "epoch": 313.7142857142857,
+      "grad_norm": 0.134765625,
+      "learning_rate": 1.1128519980890775e-06,
+      "loss": 0.0127,
+      "step": 4392
+    },
+    {
+      "epoch": 313.7857142857143,
+      "grad_norm": 0.24609375,
+      "learning_rate": 1.1092553062129724e-06,
+      "loss": 0.014,
+      "step": 4393
+    },
+    {
+      "epoch": 313.85714285714283,
+      "grad_norm": 0.162109375,
+      "learning_rate": 1.1056641660474543e-06,
+      "loss": 0.0103,
+      "step": 4394
+    },
+    {
+      "epoch": 313.92857142857144,
+      "grad_norm": 0.302734375,
+      "learning_rate": 1.1020785793428052e-06,
+      "loss": 0.0141,
+      "step": 4395
+    },
+    {
+      "epoch": 314.0,
+      "grad_norm": 0.2412109375,
+      "learning_rate": 1.0984985478465881e-06,
+      "loss": 0.0162,
+      "step": 4396
+    },
+    {
+      "epoch": 314.07142857142856,
+      "grad_norm": 0.1279296875,
+      "learning_rate": 1.094924073303674e-06,
+      "loss": 0.0134,
+      "step": 4397
+    },
+    {
+      "epoch": 314.14285714285717,
+      "grad_norm": 0.1923828125,
+      "learning_rate": 1.0913551574562157e-06,
+      "loss": 0.0147,
+      "step": 4398
+    },
+    {
+      "epoch": 314.2142857142857,
+      "grad_norm": 0.1416015625,
+      "learning_rate": 1.0877918020436598e-06,
+      "loss": 0.0154,
+      "step": 4399
+    },
+    {
+      "epoch": 314.2857142857143,
+      "grad_norm": 0.15625,
+      "learning_rate": 1.0842340088027425e-06,
+      "loss": 0.0149,
+      "step": 4400
+    },
+    {
+      "epoch": 314.35714285714283,
+      "grad_norm": 0.1259765625,
+      "learning_rate": 1.0806817794674878e-06,
+      "loss": 0.0109,
+      "step": 4401
+    },
+    {
+      "epoch": 314.42857142857144,
+      "grad_norm": 0.1875,
+      "learning_rate": 1.0771351157692123e-06,
+      "loss": 0.0162,
+      "step": 4402
+    },
+    {
+      "epoch": 314.5,
+      "grad_norm": 0.1513671875,
+      "learning_rate": 1.0735940194365148e-06,
+      "loss": 0.0132,
+      "step": 4403
+    },
+    {
+      "epoch": 314.57142857142856,
+      "grad_norm": 0.1455078125,
+      "learning_rate": 1.0700584921952846e-06,
+      "loss": 0.0134,
+      "step": 4404
+    },
+    {
+      "epoch": 314.64285714285717,
+      "grad_norm": 0.2236328125,
+      "learning_rate": 1.0665285357686876e-06,
+      "loss": 0.011,
+      "step": 4405
+    },
+    {
+      "epoch": 314.7142857142857,
+      "grad_norm": 0.11767578125,
+      "learning_rate": 1.0630041518771909e-06,
+      "loss": 0.0128,
+      "step": 4406
+    },
+    {
+      "epoch": 314.7857142857143,
+      "grad_norm": 0.115234375,
+      "learning_rate": 1.059485342238531e-06,
+      "loss": 0.0131,
+      "step": 4407
+    },
+    {
+      "epoch": 314.85714285714283,
+      "grad_norm": 0.126953125,
+      "learning_rate": 1.0559721085677348e-06,
+      "loss": 0.0138,
+      "step": 4408
+    },
+    {
+      "epoch": 314.92857142857144,
+      "grad_norm": 0.095703125,
+      "learning_rate": 1.0524644525771093e-06,
+      "loss": 0.011,
+      "step": 4409
+    },
+    {
+      "epoch": 315.0,
+      "grad_norm": 0.12890625,
+      "learning_rate": 1.0489623759762438e-06,
+      "loss": 0.0136,
+      "step": 4410
+    },
+    {
+      "epoch": 315.07142857142856,
+      "grad_norm": 0.208984375,
+      "learning_rate": 1.0454658804720048e-06,
+      "loss": 0.0115,
+      "step": 4411
+    },
+    {
+      "epoch": 315.14285714285717,
+      "grad_norm": 0.126953125,
+      "learning_rate": 1.0419749677685484e-06,
+      "loss": 0.0116,
+      "step": 4412
+    },
+    {
+      "epoch": 315.2142857142857,
+      "grad_norm": 0.1376953125,
+      "learning_rate": 1.0384896395672972e-06,
+      "loss": 0.0145,
+      "step": 4413
+    },
+    {
+      "epoch": 315.2857142857143,
+      "grad_norm": 0.19140625,
+      "learning_rate": 1.035009897566959e-06,
+      "loss": 0.0158,
+      "step": 4414
+    },
+    {
+      "epoch": 315.35714285714283,
+      "grad_norm": 0.1533203125,
+      "learning_rate": 1.0315357434635173e-06,
+      "loss": 0.0128,
+      "step": 4415
+    },
+    {
+      "epoch": 315.42857142857144,
+      "grad_norm": 0.1845703125,
+      "learning_rate": 1.0280671789502358e-06,
+      "loss": 0.0143,
+      "step": 4416
+    },
+    {
+      "epoch": 315.5,
+      "grad_norm": 0.220703125,
+      "learning_rate": 1.0246042057176483e-06,
+      "loss": 0.0149,
+      "step": 4417
+    },
+    {
+      "epoch": 315.57142857142856,
+      "grad_norm": 0.142578125,
+      "learning_rate": 1.0211468254535673e-06,
+      "loss": 0.0113,
+      "step": 4418
+    },
+    {
+      "epoch": 315.64285714285717,
+      "grad_norm": 0.1796875,
+      "learning_rate": 1.017695039843075e-06,
+      "loss": 0.0143,
+      "step": 4419
+    },
+    {
+      "epoch": 315.7142857142857,
+      "grad_norm": 0.1474609375,
+      "learning_rate": 1.0142488505685306e-06,
+      "loss": 0.0148,
+      "step": 4420
+    },
+    {
+      "epoch": 315.7857142857143,
+      "grad_norm": 0.1259765625,
+      "learning_rate": 1.0108082593095703e-06,
+      "loss": 0.0109,
+      "step": 4421
+    },
+    {
+      "epoch": 315.85714285714283,
+      "grad_norm": 0.1298828125,
+      "learning_rate": 1.0073732677430906e-06,
+      "loss": 0.0129,
+      "step": 4422
+    },
+    {
+      "epoch": 315.92857142857144,
+      "grad_norm": 0.1259765625,
+      "learning_rate": 1.003943877543266e-06,
+      "loss": 0.0126,
+      "step": 4423
+    },
+    {
+      "epoch": 316.0,
+      "grad_norm": 0.3515625,
+      "learning_rate": 1.000520090381539e-06,
+      "loss": 0.0153,
+      "step": 4424
+    },
+    {
+      "epoch": 316.07142857142856,
+      "grad_norm": 0.1259765625,
+      "learning_rate": 9.971019079266204e-07,
+      "loss": 0.0137,
+      "step": 4425
+    },
+    {
+      "epoch": 316.14285714285717,
+      "grad_norm": 0.1611328125,
+      "learning_rate": 9.936893318444953e-07,
+      "loss": 0.0113,
+      "step": 4426
+    },
+    {
+      "epoch": 316.2142857142857,
+      "grad_norm": 0.2109375,
+      "learning_rate": 9.902823637984102e-07,
+      "loss": 0.0114,
+      "step": 4427
+    },
+    {
+      "epoch": 316.2857142857143,
+      "grad_norm": 0.130859375,
+      "learning_rate": 9.868810054488794e-07,
+      "loss": 0.0126,
+      "step": 4428
+    },
+    {
+      "epoch": 316.35714285714283,
+      "grad_norm": 0.1669921875,
+      "learning_rate": 9.83485258453684e-07,
+      "loss": 0.011,
+      "step": 4429
+    },
+    {
+      "epoch": 316.42857142857144,
+      "grad_norm": 0.162109375,
+      "learning_rate": 9.800951244678682e-07,
+      "loss": 0.0123,
+      "step": 4430
+    },
+    {
+      "epoch": 316.5,
+      "grad_norm": 0.1240234375,
+      "learning_rate": 9.76710605143745e-07,
+      "loss": 0.015,
+      "step": 4431
+    },
+    {
+      "epoch": 316.57142857142856,
+      "grad_norm": 0.130859375,
+      "learning_rate": 9.733317021308849e-07,
+      "loss": 0.0151,
+      "step": 4432
+    },
+    {
+      "epoch": 316.64285714285717,
+      "grad_norm": 0.2333984375,
+      "learning_rate": 9.699584170761263e-07,
+      "loss": 0.0148,
+      "step": 4433
+    },
+    {
+      "epoch": 316.7142857142857,
+      "grad_norm": 0.15625,
+      "learning_rate": 9.665907516235638e-07,
+      "loss": 0.0129,
+      "step": 4434
+    },
+    {
+      "epoch": 316.7857142857143,
+      "grad_norm": 0.177734375,
+      "learning_rate": 9.632287074145625e-07,
+      "loss": 0.0163,
+      "step": 4435
+    },
+    {
+      "epoch": 316.85714285714283,
+      "grad_norm": 0.1748046875,
+      "learning_rate": 9.598722860877388e-07,
+      "loss": 0.0134,
+      "step": 4436
+    },
+    {
+      "epoch": 316.92857142857144,
+      "grad_norm": 0.24609375,
+      "learning_rate": 9.56521489278972e-07,
+      "loss": 0.0133,
+      "step": 4437
+    },
+    {
+      "epoch": 317.0,
+      "grad_norm": 0.154296875,
+      "learning_rate": 9.531763186213988e-07,
+      "loss": 0.0135,
+      "step": 4438
+    },
+    {
+      "epoch": 317.07142857142856,
+      "grad_norm": 0.2490234375,
+      "learning_rate": 9.498367757454163e-07,
+      "loss": 0.0117,
+      "step": 4439
+    },
+    {
+      "epoch": 317.14285714285717,
+      "grad_norm": 0.18359375,
+      "learning_rate": 9.465028622786754e-07,
+      "loss": 0.0156,
+      "step": 4440
+    },
+    {
+      "epoch": 317.2142857142857,
+      "grad_norm": 0.265625,
+      "learning_rate": 9.431745798460873e-07,
+      "loss": 0.0152,
+      "step": 4441
+    },
+    {
+      "epoch": 317.2857142857143,
+      "grad_norm": 0.1533203125,
+      "learning_rate": 9.398519300698142e-07,
+      "loss": 0.0147,
+      "step": 4442
+    },
+    {
+      "epoch": 317.35714285714283,
+      "grad_norm": 0.130859375,
+      "learning_rate": 9.365349145692759e-07,
+      "loss": 0.0103,
+      "step": 4443
+    },
+    {
+      "epoch": 317.42857142857144,
+      "grad_norm": 0.14453125,
+      "learning_rate": 9.332235349611445e-07,
+      "loss": 0.0145,
+      "step": 4444
+    },
+    {
+      "epoch": 317.5,
+      "grad_norm": 0.162109375,
+      "learning_rate": 9.299177928593486e-07,
+      "loss": 0.0112,
+      "step": 4445
+    },
+    {
+      "epoch": 317.57142857142856,
+      "grad_norm": 0.11962890625,
+      "learning_rate": 9.266176898750659e-07,
+      "loss": 0.0133,
+      "step": 4446
+    },
+    {
+      "epoch": 317.64285714285717,
+      "grad_norm": 0.17578125,
+      "learning_rate": 9.233232276167264e-07,
+      "loss": 0.0143,
+      "step": 4447
+    },
+    {
+      "epoch": 317.7142857142857,
+      "grad_norm": 0.2021484375,
+      "learning_rate": 9.200344076900097e-07,
+      "loss": 0.0139,
+      "step": 4448
+    },
+    {
+      "epoch": 317.7857142857143,
+      "grad_norm": 0.1318359375,
+      "learning_rate": 9.167512316978488e-07,
+      "loss": 0.0143,
+      "step": 4449
+    },
+    {
+      "epoch": 317.85714285714283,
+      "grad_norm": 0.11669921875,
+      "learning_rate": 9.134737012404221e-07,
+      "loss": 0.0105,
+      "step": 4450
+    },
+    {
+      "epoch": 317.92857142857144,
+      "grad_norm": 0.16796875,
+      "learning_rate": 9.102018179151587e-07,
+      "loss": 0.0127,
+      "step": 4451
+    },
+    {
+      "epoch": 318.0,
+      "grad_norm": 0.1767578125,
+      "learning_rate": 9.069355833167343e-07,
+      "loss": 0.0119,
+      "step": 4452
+    },
+    {
+      "epoch": 318.07142857142856,
+      "grad_norm": 0.1279296875,
+      "learning_rate": 9.036749990370713e-07,
+      "loss": 0.0142,
+      "step": 4453
+    },
+    {
+      "epoch": 318.14285714285717,
+      "grad_norm": 0.130859375,
+      "learning_rate": 9.004200666653418e-07,
+      "loss": 0.0147,
+      "step": 4454
+    },
+    {
+      "epoch": 318.2142857142857,
+      "grad_norm": 0.140625,
+      "learning_rate": 8.971707877879584e-07,
+      "loss": 0.0136,
+      "step": 4455
+    },
+    {
+      "epoch": 318.2857142857143,
+      "grad_norm": 0.255859375,
+      "learning_rate": 8.93927163988581e-07,
+      "loss": 0.0128,
+      "step": 4456
+    },
+    {
+      "epoch": 318.35714285714283,
+      "grad_norm": 0.1298828125,
+      "learning_rate": 8.906891968481129e-07,
+      "loss": 0.0138,
+      "step": 4457
+    },
+    {
+      "epoch": 318.42857142857144,
+      "grad_norm": 0.16796875,
+      "learning_rate": 8.874568879446945e-07,
+      "loss": 0.0141,
+      "step": 4458
+    },
+    {
+      "epoch": 318.5,
+      "grad_norm": 0.3359375,
+      "learning_rate": 8.842302388537213e-07,
+      "loss": 0.0147,
+      "step": 4459
+    },
+    {
+      "epoch": 318.57142857142856,
+      "grad_norm": 0.13671875,
+      "learning_rate": 8.810092511478185e-07,
+      "loss": 0.0127,
+      "step": 4460
+    },
+    {
+      "epoch": 318.64285714285717,
+      "grad_norm": 0.193359375,
+      "learning_rate": 8.777939263968582e-07,
+      "loss": 0.0122,
+      "step": 4461
+    },
+    {
+      "epoch": 318.7142857142857,
+      "grad_norm": 0.1494140625,
+      "learning_rate": 8.745842661679507e-07,
+      "loss": 0.0128,
+      "step": 4462
+    },
+    {
+      "epoch": 318.7857142857143,
+      "grad_norm": 0.1416015625,
+      "learning_rate": 8.713802720254418e-07,
+      "loss": 0.0149,
+      "step": 4463
+    },
+    {
+      "epoch": 318.85714285714283,
+      "grad_norm": 0.1689453125,
+      "learning_rate": 8.681819455309242e-07,
+      "loss": 0.0111,
+      "step": 4464
+    },
+    {
+      "epoch": 318.92857142857144,
+      "grad_norm": 0.134765625,
+      "learning_rate": 8.649892882432218e-07,
+      "loss": 0.0155,
+      "step": 4465
+    },
+    {
+      "epoch": 319.0,
+      "grad_norm": 0.12060546875,
+      "learning_rate": 8.618023017183979e-07,
+      "loss": 0.0126,
+      "step": 4466
+    },
+    {
+      "epoch": 319.07142857142856,
+      "grad_norm": 0.11083984375,
+      "learning_rate": 8.586209875097465e-07,
+      "loss": 0.0127,
+      "step": 4467
+    },
+    {
+      "epoch": 319.14285714285717,
+      "grad_norm": 0.216796875,
+      "learning_rate": 8.554453471678062e-07,
+      "loss": 0.0135,
+      "step": 4468
+    },
+    {
+      "epoch": 319.2142857142857,
+      "grad_norm": 0.11083984375,
+      "learning_rate": 8.522753822403446e-07,
+      "loss": 0.0127,
+      "step": 4469
+    },
+    {
+      "epoch": 319.2857142857143,
+      "grad_norm": 0.140625,
+      "learning_rate": 8.491110942723632e-07,
+      "loss": 0.0183,
+      "step": 4470
+    },
+    {
+      "epoch": 319.35714285714283,
+      "grad_norm": 0.1298828125,
+      "learning_rate": 8.459524848060979e-07,
+      "loss": 0.0135,
+      "step": 4471
+    },
+    {
+      "epoch": 319.42857142857144,
+      "grad_norm": 0.150390625,
+      "learning_rate": 8.427995553810172e-07,
+      "loss": 0.0144,
+      "step": 4472
+    },
+    {
+      "epoch": 319.5,
+      "grad_norm": 0.1513671875,
+      "learning_rate": 8.39652307533817e-07,
+      "loss": 0.0145,
+      "step": 4473
+    },
+    {
+      "epoch": 319.57142857142856,
+      "grad_norm": 0.12890625,
+      "learning_rate": 8.365107427984353e-07,
+      "loss": 0.013,
+      "step": 4474
+    },
+    {
+      "epoch": 319.64285714285717,
+      "grad_norm": 0.130859375,
+      "learning_rate": 8.333748627060254e-07,
+      "loss": 0.0131,
+      "step": 4475
+    },
+    {
+      "epoch": 319.7142857142857,
+      "grad_norm": 0.1611328125,
+      "learning_rate": 8.302446687849797e-07,
+      "loss": 0.014,
+      "step": 4476
+    },
+    {
+      "epoch": 319.7857142857143,
+      "grad_norm": 0.2216796875,
+      "learning_rate": 8.271201625609155e-07,
+      "loss": 0.0166,
+      "step": 4477
+    },
+    {
+      "epoch": 319.85714285714283,
+      "grad_norm": 0.1484375,
+      "learning_rate": 8.240013455566811e-07,
+      "loss": 0.0136,
+      "step": 4478
+    },
+    {
+      "epoch": 319.92857142857144,
+      "grad_norm": 0.3359375,
+      "learning_rate": 8.208882192923498e-07,
+      "loss": 0.0141,
+      "step": 4479
+    },
+    {
+      "epoch": 320.0,
+      "grad_norm": 0.1484375,
+      "learning_rate": 8.177807852852215e-07,
+      "loss": 0.0145,
+      "step": 4480
+    },
+    {
+      "epoch": 320.07142857142856,
+      "grad_norm": 0.1298828125,
+      "learning_rate": 8.146790450498195e-07,
+      "loss": 0.0114,
+      "step": 4481
+    },
+    {
+      "epoch": 320.14285714285717,
+      "grad_norm": 0.1865234375,
+      "learning_rate": 8.115830000978958e-07,
+      "loss": 0.0116,
+      "step": 4482
+    },
+    {
+      "epoch": 320.2142857142857,
+      "grad_norm": 0.146484375,
+      "learning_rate": 8.084926519384292e-07,
+      "loss": 0.014,
+      "step": 4483
+    },
+    {
+      "epoch": 320.2857142857143,
+      "grad_norm": 0.1396484375,
+      "learning_rate": 8.054080020776108e-07,
+      "loss": 0.0125,
+      "step": 4484
+    },
+    {
+      "epoch": 320.35714285714283,
+      "grad_norm": 0.1826171875,
+      "learning_rate": 8.023290520188659e-07,
+      "loss": 0.0152,
+      "step": 4485
+    },
+    {
+      "epoch": 320.42857142857144,
+      "grad_norm": 0.154296875,
+      "learning_rate": 7.992558032628342e-07,
+      "loss": 0.012,
+      "step": 4486
+    },
+    {
+      "epoch": 320.5,
+      "grad_norm": 0.2451171875,
+      "learning_rate": 7.961882573073828e-07,
+      "loss": 0.0138,
+      "step": 4487
+    },
+    {
+      "epoch": 320.57142857142856,
+      "grad_norm": 0.1953125,
+      "learning_rate": 7.931264156475965e-07,
+      "loss": 0.0132,
+      "step": 4488
+    },
+    {
+      "epoch": 320.64285714285717,
+      "grad_norm": 0.2177734375,
+      "learning_rate": 7.900702797757773e-07,
+      "loss": 0.0151,
+      "step": 4489
+    },
+    {
+      "epoch": 320.7142857142857,
+      "grad_norm": 0.1376953125,
+      "learning_rate": 7.870198511814503e-07,
+      "loss": 0.0138,
+      "step": 4490
+    },
+    {
+      "epoch": 320.7857142857143,
+      "grad_norm": 0.119140625,
+      "learning_rate": 7.839751313513568e-07,
+      "loss": 0.0156,
+      "step": 4491
+    },
+    {
+      "epoch": 320.85714285714283,
+      "grad_norm": 0.255859375,
+      "learning_rate": 7.809361217694572e-07,
+      "loss": 0.01,
+      "step": 4492
+    },
+    {
+      "epoch": 320.92857142857144,
+      "grad_norm": 0.2294921875,
+      "learning_rate": 7.779028239169262e-07,
+      "loss": 0.0119,
+      "step": 4493
+    },
+    {
+      "epoch": 321.0,
+      "grad_norm": 0.138671875,
+      "learning_rate": 7.748752392721576e-07,
+      "loss": 0.0129,
+      "step": 4494
+    },
+    {
+      "epoch": 321.07142857142856,
+      "grad_norm": 0.1279296875,
+      "learning_rate": 7.718533693107585e-07,
+      "loss": 0.0115,
+      "step": 4495
+    },
+    {
+      "epoch": 321.14285714285717,
+      "grad_norm": 0.287109375,
+      "learning_rate": 7.688372155055509e-07,
+      "loss": 0.0122,
+      "step": 4496
+    },
+    {
+      "epoch": 321.2142857142857,
+      "grad_norm": 0.19921875,
+      "learning_rate": 7.658267793265742e-07,
+      "loss": 0.0155,
+      "step": 4497
+    },
+    {
+      "epoch": 321.2857142857143,
+      "grad_norm": 0.1572265625,
+      "learning_rate": 7.628220622410767e-07,
+      "loss": 0.0123,
+      "step": 4498
+    },
+    {
+      "epoch": 321.35714285714283,
+      "grad_norm": 0.138671875,
+      "learning_rate": 7.59823065713522e-07,
+      "loss": 0.0136,
+      "step": 4499
+    },
+    {
+      "epoch": 321.42857142857144,
+      "grad_norm": 0.173828125,
+      "learning_rate": 7.568297912055822e-07,
+      "loss": 0.0116,
+      "step": 4500
+    },
+    {
+      "epoch": 321.5,
+      "grad_norm": 0.1337890625,
+      "learning_rate": 7.538422401761461e-07,
+      "loss": 0.0125,
+      "step": 4501
+    },
+    {
+      "epoch": 321.57142857142856,
+      "grad_norm": 0.3359375,
+      "learning_rate": 7.508604140813072e-07,
+      "loss": 0.0153,
+      "step": 4502
+    },
+    {
+      "epoch": 321.64285714285717,
+      "grad_norm": 0.1865234375,
+      "learning_rate": 7.478843143743725e-07,
+      "loss": 0.0126,
+      "step": 4503
+    },
+    {
+      "epoch": 321.7142857142857,
+      "grad_norm": 0.12109375,
+      "learning_rate": 7.449139425058579e-07,
+      "loss": 0.0137,
+      "step": 4504
+    },
+    {
+      "epoch": 321.7857142857143,
+      "grad_norm": 0.1640625,
+      "learning_rate": 7.41949299923482e-07,
+      "loss": 0.0101,
+      "step": 4505
+    },
+    {
+      "epoch": 321.85714285714283,
+      "grad_norm": 0.1650390625,
+      "learning_rate": 7.389903880721818e-07,
+      "loss": 0.0107,
+      "step": 4506
+    },
+    {
+      "epoch": 321.92857142857144,
+      "grad_norm": 0.1826171875,
+      "learning_rate": 7.360372083940914e-07,
+      "loss": 0.012,
+      "step": 4507
+    },
+    {
+      "epoch": 322.0,
+      "grad_norm": 0.10498046875,
+      "learning_rate": 7.330897623285551e-07,
+      "loss": 0.0115,
+      "step": 4508
+    },
+    {
+      "epoch": 322.07142857142856,
+      "grad_norm": 0.150390625,
+      "learning_rate": 7.301480513121239e-07,
+      "loss": 0.0136,
+      "step": 4509
+    },
+    {
+      "epoch": 322.14285714285717,
+      "grad_norm": 0.197265625,
+      "learning_rate": 7.272120767785496e-07,
+      "loss": 0.0151,
+      "step": 4510
+    },
+    {
+      "epoch": 322.2142857142857,
+      "grad_norm": 0.1845703125,
+      "learning_rate": 7.242818401587914e-07,
+      "loss": 0.0157,
+      "step": 4511
+    },
+    {
+      "epoch": 322.2857142857143,
+      "grad_norm": 0.1845703125,
+      "learning_rate": 7.213573428810114e-07,
+      "loss": 0.0156,
+      "step": 4512
+    },
+    {
+      "epoch": 322.35714285714283,
+      "grad_norm": 0.1259765625,
+      "learning_rate": 7.184385863705748e-07,
+      "loss": 0.0117,
+      "step": 4513
+    },
+    {
+      "epoch": 322.42857142857144,
+      "grad_norm": 0.166015625,
+      "learning_rate": 7.155255720500464e-07,
+      "loss": 0.0139,
+      "step": 4514
+    },
+    {
+      "epoch": 322.5,
+      "grad_norm": 0.10888671875,
+      "learning_rate": 7.126183013391935e-07,
+      "loss": 0.0129,
+      "step": 4515
+    },
+    {
+      "epoch": 322.57142857142856,
+      "grad_norm": 0.12890625,
+      "learning_rate": 7.097167756549881e-07,
+      "loss": 0.0098,
+      "step": 4516
+    },
+    {
+      "epoch": 322.64285714285717,
+      "grad_norm": 0.123046875,
+      "learning_rate": 7.068209964115946e-07,
+      "loss": 0.014,
+      "step": 4517
+    },
+    {
+      "epoch": 322.7142857142857,
+      "grad_norm": 0.12109375,
+      "learning_rate": 7.039309650203843e-07,
+      "loss": 0.0117,
+      "step": 4518
+    },
+    {
+      "epoch": 322.7857142857143,
+      "grad_norm": 0.109375,
+      "learning_rate": 7.010466828899193e-07,
+      "loss": 0.0112,
+      "step": 4519
+    },
+    {
+      "epoch": 322.85714285714283,
+      "grad_norm": 0.142578125,
+      "learning_rate": 6.981681514259666e-07,
+      "loss": 0.0145,
+      "step": 4520
+    },
+    {
+      "epoch": 322.92857142857144,
+      "grad_norm": 0.12109375,
+      "learning_rate": 6.952953720314876e-07,
+      "loss": 0.0123,
+      "step": 4521
+    },
+    {
+      "epoch": 323.0,
+      "grad_norm": 0.298828125,
+      "learning_rate": 6.924283461066383e-07,
+      "loss": 0.0136,
+      "step": 4522
+    },
+    {
+      "epoch": 323.07142857142856,
+      "grad_norm": 0.12158203125,
+      "learning_rate": 6.895670750487735e-07,
+      "loss": 0.016,
+      "step": 4523
+    },
+    {
+      "epoch": 323.14285714285717,
+      "grad_norm": 0.1728515625,
+      "learning_rate": 6.867115602524413e-07,
+      "loss": 0.0128,
+      "step": 4524
+    },
+    {
+      "epoch": 323.2142857142857,
+      "grad_norm": 0.140625,
+      "learning_rate": 6.838618031093827e-07,
+      "loss": 0.0151,
+      "step": 4525
+    },
+    {
+      "epoch": 323.2857142857143,
+      "grad_norm": 0.1845703125,
+      "learning_rate": 6.810178050085395e-07,
+      "loss": 0.0127,
+      "step": 4526
+    },
+    {
+      "epoch": 323.35714285714283,
+      "grad_norm": 0.1201171875,
+      "learning_rate": 6.781795673360402e-07,
+      "loss": 0.0123,
+      "step": 4527
+    },
+    {
+      "epoch": 323.42857142857144,
+      "grad_norm": 0.1591796875,
+      "learning_rate": 6.753470914752078e-07,
+      "loss": 0.0121,
+      "step": 4528
+    },
+    {
+      "epoch": 323.5,
+      "grad_norm": 0.1416015625,
+      "learning_rate": 6.725203788065515e-07,
+      "loss": 0.0135,
+      "step": 4529
+    },
+    {
+      "epoch": 323.57142857142856,
+      "grad_norm": 0.1533203125,
+      "learning_rate": 6.696994307077833e-07,
+      "loss": 0.0143,
+      "step": 4530
+    },
+    {
+      "epoch": 323.64285714285717,
+      "grad_norm": 0.11328125,
+      "learning_rate": 6.668842485537966e-07,
+      "loss": 0.0112,
+      "step": 4531
+    },
+    {
+      "epoch": 323.7142857142857,
+      "grad_norm": 0.2119140625,
+      "learning_rate": 6.640748337166755e-07,
+      "loss": 0.0144,
+      "step": 4532
+    },
+    {
+      "epoch": 323.7857142857143,
+      "grad_norm": 0.1455078125,
+      "learning_rate": 6.612711875656971e-07,
+      "loss": 0.0142,
+      "step": 4533
+    },
+    {
+      "epoch": 323.85714285714283,
+      "grad_norm": 0.1630859375,
+      "learning_rate": 6.584733114673211e-07,
+      "loss": 0.0126,
+      "step": 4534
+    },
+    {
+      "epoch": 323.92857142857144,
+      "grad_norm": 0.1376953125,
+      "learning_rate": 6.556812067852039e-07,
+      "loss": 0.0157,
+      "step": 4535
+    },
+    {
+      "epoch": 324.0,
+      "grad_norm": 0.1259765625,
+      "learning_rate": 6.528948748801805e-07,
+      "loss": 0.0134,
+      "step": 4536
+    },
+    {
+      "epoch": 324.07142857142856,
+      "grad_norm": 0.236328125,
+      "learning_rate": 6.501143171102731e-07,
+      "loss": 0.0152,
+      "step": 4537
+    },
+    {
+      "epoch": 324.14285714285717,
+      "grad_norm": 0.185546875,
+      "learning_rate": 6.473395348306919e-07,
+      "loss": 0.0152,
+      "step": 4538
+    },
+    {
+      "epoch": 324.2142857142857,
+      "grad_norm": 0.1484375,
+      "learning_rate": 6.445705293938356e-07,
+      "loss": 0.0155,
+      "step": 4539
+    },
+    {
+      "epoch": 324.2857142857143,
+      "grad_norm": 0.134765625,
+      "learning_rate": 6.418073021492818e-07,
+      "loss": 0.0128,
+      "step": 4540
+    },
+    {
+      "epoch": 324.35714285714283,
+      "grad_norm": 0.1572265625,
+      "learning_rate": 6.390498544437934e-07,
+      "loss": 0.0158,
+      "step": 4541
+    },
+    {
+      "epoch": 324.42857142857144,
+      "grad_norm": 0.265625,
+      "learning_rate": 6.362981876213189e-07,
+      "loss": 0.0165,
+      "step": 4542
+    },
+    {
+      "epoch": 324.5,
+      "grad_norm": 0.126953125,
+      "learning_rate": 6.335523030229843e-07,
+      "loss": 0.0134,
+      "step": 4543
+    },
+    {
+      "epoch": 324.57142857142856,
+      "grad_norm": 0.11572265625,
+      "learning_rate": 6.308122019871007e-07,
+      "loss": 0.0115,
+      "step": 4544
+    },
+    {
+      "epoch": 324.64285714285717,
+      "grad_norm": 0.138671875,
+      "learning_rate": 6.280778858491654e-07,
+      "loss": 0.0148,
+      "step": 4545
+    },
+    {
+      "epoch": 324.7142857142857,
+      "grad_norm": 0.1259765625,
+      "learning_rate": 6.253493559418455e-07,
+      "loss": 0.0108,
+      "step": 4546
+    },
+    {
+      "epoch": 324.7857142857143,
+      "grad_norm": 0.2353515625,
+      "learning_rate": 6.226266135949926e-07,
+      "loss": 0.0127,
+      "step": 4547
+    },
+    {
+      "epoch": 324.85714285714283,
+      "grad_norm": 0.189453125,
+      "learning_rate": 6.199096601356411e-07,
+      "loss": 0.0104,
+      "step": 4548
+    },
+    {
+      "epoch": 324.92857142857144,
+      "grad_norm": 0.1591796875,
+      "learning_rate": 6.171984968879995e-07,
+      "loss": 0.0141,
+      "step": 4549
+    },
+    {
+      "epoch": 325.0,
+      "grad_norm": 0.1650390625,
+      "learning_rate": 6.144931251734584e-07,
+      "loss": 0.0144,
+      "step": 4550
+    },
+    {
+      "epoch": 325.07142857142856,
+      "grad_norm": 0.166015625,
+      "learning_rate": 6.117935463105809e-07,
+      "loss": 0.0143,
+      "step": 4551
+    },
+    {
+      "epoch": 325.14285714285717,
+      "grad_norm": 0.21875,
+      "learning_rate": 6.09099761615109e-07,
+      "loss": 0.0131,
+      "step": 4552
+    },
+    {
+      "epoch": 325.2142857142857,
+      "grad_norm": 0.1572265625,
+      "learning_rate": 6.064117723999591e-07,
+      "loss": 0.0117,
+      "step": 4553
+    },
+    {
+      "epoch": 325.2857142857143,
+      "grad_norm": 0.1142578125,
+      "learning_rate": 6.037295799752304e-07,
+      "loss": 0.0114,
+      "step": 4554
+    },
+    {
+      "epoch": 325.35714285714283,
+      "grad_norm": 0.125,
+      "learning_rate": 6.010531856481858e-07,
+      "loss": 0.0119,
+      "step": 4555
+    },
+    {
+      "epoch": 325.42857142857144,
+      "grad_norm": 0.19140625,
+      "learning_rate": 5.983825907232677e-07,
+      "loss": 0.014,
+      "step": 4556
+    },
+    {
+      "epoch": 325.5,
+      "grad_norm": 0.2431640625,
+      "learning_rate": 5.957177965020918e-07,
+      "loss": 0.0166,
+      "step": 4557
+    },
+    {
+      "epoch": 325.57142857142856,
+      "grad_norm": 0.123046875,
+      "learning_rate": 5.930588042834459e-07,
+      "loss": 0.0117,
+      "step": 4558
+    },
+    {
+      "epoch": 325.64285714285717,
+      "grad_norm": 0.134765625,
+      "learning_rate": 5.904056153632945e-07,
+      "loss": 0.0133,
+      "step": 4559
+    },
+    {
+      "epoch": 325.7142857142857,
+      "grad_norm": 0.1259765625,
+      "learning_rate": 5.877582310347657e-07,
+      "loss": 0.0139,
+      "step": 4560
+    },
+    {
+      "epoch": 325.7857142857143,
+      "grad_norm": 0.263671875,
+      "learning_rate": 5.851166525881636e-07,
+      "loss": 0.0139,
+      "step": 4561
+    },
+    {
+      "epoch": 325.85714285714283,
+      "grad_norm": 0.1416015625,
+      "learning_rate": 5.824808813109609e-07,
+      "loss": 0.0149,
+      "step": 4562
+    },
+    {
+      "epoch": 325.92857142857144,
+      "grad_norm": 0.3203125,
+      "learning_rate": 5.79850918487801e-07,
+      "loss": 0.012,
+      "step": 4563
+    },
+    {
+      "epoch": 326.0,
+      "grad_norm": 0.16796875,
+      "learning_rate": 5.772267654004959e-07,
+      "loss": 0.0148,
+      "step": 4564
+    },
+    {
+      "epoch": 326.07142857142856,
+      "grad_norm": 0.1455078125,
+      "learning_rate": 5.746084233280269e-07,
+      "loss": 0.0116,
+      "step": 4565
+    },
+    {
+      "epoch": 326.14285714285717,
+      "grad_norm": 0.185546875,
+      "learning_rate": 5.719958935465397e-07,
+      "loss": 0.0119,
+      "step": 4566
+    },
+    {
+      "epoch": 326.2142857142857,
+      "grad_norm": 0.1494140625,
+      "learning_rate": 5.693891773293506e-07,
+      "loss": 0.0146,
+      "step": 4567
+    },
+    {
+      "epoch": 326.2857142857143,
+      "grad_norm": 0.169921875,
+      "learning_rate": 5.667882759469431e-07,
+      "loss": 0.0132,
+      "step": 4568
+    },
+    {
+      "epoch": 326.35714285714283,
+      "grad_norm": 0.294921875,
+      "learning_rate": 5.641931906669632e-07,
+      "loss": 0.0143,
+      "step": 4569
+    },
+    {
+      "epoch": 326.42857142857144,
+      "grad_norm": 0.1591796875,
+      "learning_rate": 5.616039227542239e-07,
+      "loss": 0.014,
+      "step": 4570
+    },
+    {
+      "epoch": 326.5,
+      "grad_norm": 0.146484375,
+      "learning_rate": 5.590204734707047e-07,
+      "loss": 0.0132,
+      "step": 4571
+    },
+    {
+      "epoch": 326.57142857142856,
+      "grad_norm": 0.1259765625,
+      "learning_rate": 5.564428440755457e-07,
+      "loss": 0.0123,
+      "step": 4572
+    },
+    {
+      "epoch": 326.64285714285717,
+      "grad_norm": 0.15234375,
+      "learning_rate": 5.538710358250516e-07,
+      "loss": 0.0124,
+      "step": 4573
+    },
+    {
+      "epoch": 326.7142857142857,
+      "grad_norm": 0.134765625,
+      "learning_rate": 5.513050499726918e-07,
+      "loss": 0.011,
+      "step": 4574
+    },
+    {
+      "epoch": 326.7857142857143,
+      "grad_norm": 0.2490234375,
+      "learning_rate": 5.487448877690967e-07,
+      "loss": 0.0161,
+      "step": 4575
+    },
+    {
+      "epoch": 326.85714285714283,
+      "grad_norm": 0.1298828125,
+      "learning_rate": 5.461905504620571e-07,
+      "loss": 0.0134,
+      "step": 4576
+    },
+    {
+      "epoch": 326.92857142857144,
+      "grad_norm": 0.12353515625,
+      "learning_rate": 5.436420392965255e-07,
+      "loss": 0.0143,
+      "step": 4577
+    },
+    {
+      "epoch": 327.0,
+      "grad_norm": 0.2451171875,
+      "learning_rate": 5.410993555146182e-07,
+      "loss": 0.013,
+      "step": 4578
+    },
+    {
+      "epoch": 327.07142857142856,
+      "grad_norm": 0.1328125,
+      "learning_rate": 5.385625003556078e-07,
+      "loss": 0.0137,
+      "step": 4579
+    },
+    {
+      "epoch": 327.14285714285717,
+      "grad_norm": 0.1181640625,
+      "learning_rate": 5.360314750559231e-07,
+      "loss": 0.0131,
+      "step": 4580
+    },
+    {
+      "epoch": 327.2142857142857,
+      "grad_norm": 0.2470703125,
+      "learning_rate": 5.335062808491595e-07,
+      "loss": 0.0151,
+      "step": 4581
+    },
+    {
+      "epoch": 327.2857142857143,
+      "grad_norm": 0.166015625,
+      "learning_rate": 5.30986918966063e-07,
+      "loss": 0.0107,
+      "step": 4582
+    },
+    {
+      "epoch": 327.35714285714283,
+      "grad_norm": 0.142578125,
+      "learning_rate": 5.284733906345415e-07,
+      "loss": 0.013,
+      "step": 4583
+    },
+    {
+      "epoch": 327.42857142857144,
+      "grad_norm": 0.13671875,
+      "learning_rate": 5.25965697079657e-07,
+      "loss": 0.0136,
+      "step": 4584
+    },
+    {
+      "epoch": 327.5,
+      "grad_norm": 0.1533203125,
+      "learning_rate": 5.234638395236302e-07,
+      "loss": 0.0117,
+      "step": 4585
+    },
+    {
+      "epoch": 327.57142857142856,
+      "grad_norm": 0.150390625,
+      "learning_rate": 5.209678191858344e-07,
+      "loss": 0.0123,
+      "step": 4586
+    },
+    {
+      "epoch": 327.64285714285717,
+      "grad_norm": 0.1513671875,
+      "learning_rate": 5.184776372828007e-07,
+      "loss": 0.0156,
+      "step": 4587
+    },
+    {
+      "epoch": 327.7142857142857,
+      "grad_norm": 0.26953125,
+      "learning_rate": 5.159932950282151e-07,
+      "loss": 0.009,
+      "step": 4588
+    },
+    {
+      "epoch": 327.7857142857143,
+      "grad_norm": 0.1181640625,
+      "learning_rate": 5.135147936329132e-07,
+      "loss": 0.0139,
+      "step": 4589
+    },
+    {
+      "epoch": 327.85714285714283,
+      "grad_norm": 0.1455078125,
+      "learning_rate": 5.110421343048871e-07,
+      "loss": 0.013,
+      "step": 4590
+    },
+    {
+      "epoch": 327.92857142857144,
+      "grad_norm": 0.14453125,
+      "learning_rate": 5.085753182492811e-07,
+      "loss": 0.0135,
+      "step": 4591
+    },
+    {
+      "epoch": 328.0,
+      "grad_norm": 0.185546875,
+      "learning_rate": 5.061143466683918e-07,
+      "loss": 0.0124,
+      "step": 4592
+    },
+    {
+      "epoch": 328.07142857142856,
+      "grad_norm": 0.173828125,
+      "learning_rate": 5.036592207616667e-07,
+      "loss": 0.0159,
+      "step": 4593
+    },
+    {
+      "epoch": 328.14285714285717,
+      "grad_norm": 0.2314453125,
+      "learning_rate": 5.012099417257041e-07,
+      "loss": 0.0123,
+      "step": 4594
+    },
+    {
+      "epoch": 328.2142857142857,
+      "grad_norm": 0.1796875,
+      "learning_rate": 4.987665107542533e-07,
+      "loss": 0.0153,
+      "step": 4595
+    },
+    {
+      "epoch": 328.2857142857143,
+      "grad_norm": 0.1572265625,
+      "learning_rate": 4.963289290382117e-07,
+      "loss": 0.0138,
+      "step": 4596
+    },
+    {
+      "epoch": 328.35714285714283,
+      "grad_norm": 0.2060546875,
+      "learning_rate": 4.938971977656276e-07,
+      "loss": 0.0118,
+      "step": 4597
+    },
+    {
+      "epoch": 328.42857142857144,
+      "grad_norm": 0.169921875,
+      "learning_rate": 4.914713181217001e-07,
+      "loss": 0.0143,
+      "step": 4598
+    },
+    {
+      "epoch": 328.5,
+      "grad_norm": 0.1357421875,
+      "learning_rate": 4.890512912887695e-07,
+      "loss": 0.016,
+      "step": 4599
+    },
+    {
+      "epoch": 328.57142857142856,
+      "grad_norm": 0.1796875,
+      "learning_rate": 4.866371184463273e-07,
+      "loss": 0.0147,
+      "step": 4600
+    },
+    {
+      "epoch": 328.64285714285717,
+      "grad_norm": 0.142578125,
+      "learning_rate": 4.842288007710139e-07,
+      "loss": 0.0135,
+      "step": 4601
+    },
+    {
+      "epoch": 328.7142857142857,
+      "grad_norm": 0.248046875,
+      "learning_rate": 4.818263394366154e-07,
+      "loss": 0.0138,
+      "step": 4602
+    },
+    {
+      "epoch": 328.7857142857143,
+      "grad_norm": 0.19140625,
+      "learning_rate": 4.794297356140593e-07,
+      "loss": 0.0112,
+      "step": 4603
+    },
+    {
+      "epoch": 328.85714285714283,
+      "grad_norm": 0.1513671875,
+      "learning_rate": 4.770389904714234e-07,
+      "loss": 0.0104,
+      "step": 4604
+    },
+    {
+      "epoch": 328.92857142857144,
+      "grad_norm": 0.1875,
+      "learning_rate": 4.746541051739259e-07,
+      "loss": 0.0111,
+      "step": 4605
+    },
+    {
+      "epoch": 329.0,
+      "grad_norm": 0.1591796875,
+      "learning_rate": 4.722750808839327e-07,
+      "loss": 0.0128,
+      "step": 4606
+    },
+    {
+      "epoch": 329.07142857142856,
+      "grad_norm": 0.1669921875,
+      "learning_rate": 4.699019187609524e-07,
+      "loss": 0.0135,
+      "step": 4607
+    },
+    {
+      "epoch": 329.14285714285717,
+      "grad_norm": 0.14453125,
+      "learning_rate": 4.6753461996163204e-07,
+      "loss": 0.0137,
+      "step": 4608
+    },
+    {
+      "epoch": 329.2142857142857,
+      "grad_norm": 0.15625,
+      "learning_rate": 4.6517318563976535e-07,
+      "loss": 0.0141,
+      "step": 4609
+    },
+    {
+      "epoch": 329.2857142857143,
+      "grad_norm": 0.12109375,
+      "learning_rate": 4.628176169462872e-07,
+      "loss": 0.0157,
+      "step": 4610
+    },
+    {
+      "epoch": 329.35714285714283,
+      "grad_norm": 0.138671875,
+      "learning_rate": 4.604679150292737e-07,
+      "loss": 0.0128,
+      "step": 4611
+    },
+    {
+      "epoch": 329.42857142857144,
+      "grad_norm": 0.25390625,
+      "learning_rate": 4.581240810339407e-07,
+      "loss": 0.0126,
+      "step": 4612
+    },
+    {
+      "epoch": 329.5,
+      "grad_norm": 0.1435546875,
+      "learning_rate": 4.557861161026439e-07,
+      "loss": 0.0147,
+      "step": 4613
+    },
+    {
+      "epoch": 329.57142857142856,
+      "grad_norm": 0.193359375,
+      "learning_rate": 4.534540213748786e-07,
+      "loss": 0.0119,
+      "step": 4614
+    },
+    {
+      "epoch": 329.64285714285717,
+      "grad_norm": 0.203125,
+      "learning_rate": 4.511277979872802e-07,
+      "loss": 0.0124,
+      "step": 4615
+    },
+    {
+      "epoch": 329.7142857142857,
+      "grad_norm": 0.14453125,
+      "learning_rate": 4.4880744707362507e-07,
+      "loss": 0.0132,
+      "step": 4616
+    },
+    {
+      "epoch": 329.7857142857143,
+      "grad_norm": 0.1455078125,
+      "learning_rate": 4.464929697648182e-07,
+      "loss": 0.0128,
+      "step": 4617
+    },
+    {
+      "epoch": 329.85714285714283,
+      "grad_norm": 0.138671875,
+      "learning_rate": 4.441843671889101e-07,
+      "loss": 0.0141,
+      "step": 4618
+    },
+    {
+      "epoch": 329.92857142857144,
+      "grad_norm": 0.154296875,
+      "learning_rate": 4.4188164047108405e-07,
+      "loss": 0.0113,
+      "step": 4619
+    },
+    {
+      "epoch": 330.0,
+      "grad_norm": 0.1416015625,
+      "learning_rate": 4.395847907336645e-07,
+      "loss": 0.0153,
+      "step": 4620
+    },
+    {
+      "epoch": 330.07142857142856,
+      "grad_norm": 0.125,
+      "learning_rate": 4.372938190961073e-07,
+      "loss": 0.0163,
+      "step": 4621
+    },
+    {
+      "epoch": 330.14285714285717,
+      "grad_norm": 0.1484375,
+      "learning_rate": 4.3500872667500274e-07,
+      "loss": 0.0131,
+      "step": 4622
+    },
+    {
+      "epoch": 330.2142857142857,
+      "grad_norm": 0.185546875,
+      "learning_rate": 4.327295145840765e-07,
+      "loss": 0.0129,
+      "step": 4623
+    },
+    {
+      "epoch": 330.2857142857143,
+      "grad_norm": 0.208984375,
+      "learning_rate": 4.304561839341914e-07,
+      "loss": 0.0135,
+      "step": 4624
+    },
+    {
+      "epoch": 330.35714285714283,
+      "grad_norm": 0.16796875,
+      "learning_rate": 4.281887358333389e-07,
+      "loss": 0.0158,
+      "step": 4625
+    },
+    {
+      "epoch": 330.42857142857144,
+      "grad_norm": 0.1748046875,
+      "learning_rate": 4.259271713866475e-07,
+      "loss": 0.0111,
+      "step": 4626
+    },
+    {
+      "epoch": 330.5,
+      "grad_norm": 0.11279296875,
+      "learning_rate": 4.2367149169637436e-07,
+      "loss": 0.0117,
+      "step": 4627
+    },
+    {
+      "epoch": 330.57142857142856,
+      "grad_norm": 0.103515625,
+      "learning_rate": 4.2142169786191234e-07,
+      "loss": 0.0123,
+      "step": 4628
+    },
+    {
+      "epoch": 330.64285714285717,
+      "grad_norm": 0.14453125,
+      "learning_rate": 4.1917779097978006e-07,
+      "loss": 0.0133,
+      "step": 4629
+    },
+    {
+      "epoch": 330.7142857142857,
+      "grad_norm": 0.2412109375,
+      "learning_rate": 4.1693977214363465e-07,
+      "loss": 0.016,
+      "step": 4630
+    },
+    {
+      "epoch": 330.7857142857143,
+      "grad_norm": 0.134765625,
+      "learning_rate": 4.1470764244425783e-07,
+      "loss": 0.0124,
+      "step": 4631
+    },
+    {
+      "epoch": 330.85714285714283,
+      "grad_norm": 0.158203125,
+      "learning_rate": 4.1248140296956256e-07,
+      "loss": 0.0126,
+      "step": 4632
+    },
+    {
+      "epoch": 330.92857142857144,
+      "grad_norm": 0.154296875,
+      "learning_rate": 4.102610548045893e-07,
+      "loss": 0.0159,
+      "step": 4633
+    },
+    {
+      "epoch": 331.0,
+      "grad_norm": 0.1318359375,
+      "learning_rate": 4.080465990315113e-07,
+      "loss": 0.0102,
+      "step": 4634
+    },
+    {
+      "epoch": 331.07142857142856,
+      "grad_norm": 0.173828125,
+      "learning_rate": 4.0583803672962635e-07,
+      "loss": 0.0152,
+      "step": 4635
+    },
+    {
+      "epoch": 331.14285714285717,
+      "grad_norm": 0.1845703125,
+      "learning_rate": 4.036353689753608e-07,
+      "loss": 0.0127,
+      "step": 4636
+    },
+    {
+      "epoch": 331.2142857142857,
+      "grad_norm": 0.189453125,
+      "learning_rate": 4.014385968422671e-07,
+      "loss": 0.0149,
+      "step": 4637
+    },
+    {
+      "epoch": 331.2857142857143,
+      "grad_norm": 0.140625,
+      "learning_rate": 3.9924772140102624e-07,
+      "loss": 0.0142,
+      "step": 4638
+    },
+    {
+      "epoch": 331.35714285714283,
+      "grad_norm": 0.138671875,
+      "learning_rate": 3.970627437194452e-07,
+      "loss": 0.0121,
+      "step": 4639
+    },
+    {
+      "epoch": 331.42857142857144,
+      "grad_norm": 0.140625,
+      "learning_rate": 3.9488366486245403e-07,
+      "loss": 0.0117,
+      "step": 4640
+    },
+    {
+      "epoch": 331.5,
+      "grad_norm": 0.2734375,
+      "learning_rate": 3.9271048589211155e-07,
+      "loss": 0.0154,
+      "step": 4641
+    },
+    {
+      "epoch": 331.57142857142856,
+      "grad_norm": 0.1123046875,
+      "learning_rate": 3.905432078675955e-07,
+      "loss": 0.0145,
+      "step": 4642
+    },
+    {
+      "epoch": 331.64285714285717,
+      "grad_norm": 0.1416015625,
+      "learning_rate": 3.883818318452137e-07,
+      "loss": 0.0123,
+      "step": 4643
+    },
+    {
+      "epoch": 331.7142857142857,
+      "grad_norm": 0.2099609375,
+      "learning_rate": 3.862263588783929e-07,
+      "loss": 0.0103,
+      "step": 4644
+    },
+    {
+      "epoch": 331.7857142857143,
+      "grad_norm": 0.294921875,
+      "learning_rate": 3.840767900176859e-07,
+      "loss": 0.0129,
+      "step": 4645
+    },
+    {
+      "epoch": 331.85714285714283,
+      "grad_norm": 0.1416015625,
+      "learning_rate": 3.819331263107642e-07,
+      "loss": 0.0141,
+      "step": 4646
+    },
+    {
+      "epoch": 331.92857142857144,
+      "grad_norm": 0.142578125,
+      "learning_rate": 3.797953688024256e-07,
+      "loss": 0.0141,
+      "step": 4647
+    },
+    {
+      "epoch": 332.0,
+      "grad_norm": 0.11474609375,
+      "learning_rate": 3.7766351853458506e-07,
+      "loss": 0.0118,
+      "step": 4648
+    },
+    {
+      "epoch": 332.07142857142856,
+      "grad_norm": 0.205078125,
+      "learning_rate": 3.755375765462837e-07,
+      "loss": 0.0156,
+      "step": 4649
+    },
+    {
+      "epoch": 332.14285714285717,
+      "grad_norm": 0.11328125,
+      "learning_rate": 3.7341754387367753e-07,
+      "loss": 0.0123,
+      "step": 4650
+    },
+    {
+      "epoch": 332.2142857142857,
+      "grad_norm": 0.2275390625,
+      "learning_rate": 3.713034215500441e-07,
+      "loss": 0.0137,
+      "step": 4651
+    },
+    {
+      "epoch": 332.2857142857143,
+      "grad_norm": 0.166015625,
+      "learning_rate": 3.6919521060578276e-07,
+      "loss": 0.0123,
+      "step": 4652
+    },
+    {
+      "epoch": 332.35714285714283,
+      "grad_norm": 0.2041015625,
+      "learning_rate": 3.6709291206840915e-07,
+      "loss": 0.0133,
+      "step": 4653
+    },
+    {
+      "epoch": 332.42857142857144,
+      "grad_norm": 0.11962890625,
+      "learning_rate": 3.649965269625577e-07,
+      "loss": 0.0129,
+      "step": 4654
+    },
+    {
+      "epoch": 332.5,
+      "grad_norm": 0.1328125,
+      "learning_rate": 3.6290605630998065e-07,
+      "loss": 0.0111,
+      "step": 4655
+    },
+    {
+      "epoch": 332.57142857142856,
+      "grad_norm": 0.1484375,
+      "learning_rate": 3.608215011295504e-07,
+      "loss": 0.0127,
+      "step": 4656
+    },
+    {
+      "epoch": 332.64285714285717,
+      "grad_norm": 0.140625,
+      "learning_rate": 3.587428624372502e-07,
+      "loss": 0.0141,
+      "step": 4657
+    },
+    {
+      "epoch": 332.7142857142857,
+      "grad_norm": 0.13671875,
+      "learning_rate": 3.566701412461848e-07,
+      "loss": 0.012,
+      "step": 4658
+    },
+    {
+      "epoch": 332.7857142857143,
+      "grad_norm": 0.1435546875,
+      "learning_rate": 3.5460333856657536e-07,
+      "loss": 0.0122,
+      "step": 4659
+    },
+    {
+      "epoch": 332.85714285714283,
+      "grad_norm": 0.158203125,
+      "learning_rate": 3.5254245540575633e-07,
+      "loss": 0.0148,
+      "step": 4660
+    },
+    {
+      "epoch": 332.92857142857144,
+      "grad_norm": 0.17578125,
+      "learning_rate": 3.504874927681728e-07,
+      "loss": 0.0137,
+      "step": 4661
+    },
+    {
+      "epoch": 333.0,
+      "grad_norm": 0.255859375,
+      "learning_rate": 3.484384516553918e-07,
+      "loss": 0.0132,
+      "step": 4662
+    },
+    {
+      "epoch": 333.07142857142856,
+      "grad_norm": 0.134765625,
+      "learning_rate": 3.4639533306609074e-07,
+      "loss": 0.0134,
+      "step": 4663
+    },
+    {
+      "epoch": 333.14285714285717,
+      "grad_norm": 0.357421875,
+      "learning_rate": 3.4435813799606057e-07,
+      "loss": 0.0157,
+      "step": 4664
+    },
+    {
+      "epoch": 333.2142857142857,
+      "grad_norm": 0.1728515625,
+      "learning_rate": 3.4232686743820576e-07,
+      "loss": 0.0138,
+      "step": 4665
+    },
+    {
+      "epoch": 333.2857142857143,
+      "grad_norm": 0.154296875,
+      "learning_rate": 3.4030152238254397e-07,
+      "loss": 0.0124,
+      "step": 4666
+    },
+    {
+      "epoch": 333.35714285714283,
+      "grad_norm": 0.1416015625,
+      "learning_rate": 3.3828210381619956e-07,
+      "loss": 0.0145,
+      "step": 4667
+    },
+    {
+      "epoch": 333.42857142857144,
+      "grad_norm": 0.1904296875,
+      "learning_rate": 3.3626861272341714e-07,
+      "loss": 0.0127,
+      "step": 4668
+    },
+    {
+      "epoch": 333.5,
+      "grad_norm": 0.12158203125,
+      "learning_rate": 3.342610500855478e-07,
+      "loss": 0.0122,
+      "step": 4669
+    },
+    {
+      "epoch": 333.57142857142856,
+      "grad_norm": 0.12060546875,
+      "learning_rate": 3.3225941688105057e-07,
+      "loss": 0.013,
+      "step": 4670
+    },
+    {
+      "epoch": 333.64285714285717,
+      "grad_norm": 0.259765625,
+      "learning_rate": 3.302637140854964e-07,
+      "loss": 0.0117,
+      "step": 4671
+    },
+    {
+      "epoch": 333.7142857142857,
+      "grad_norm": 0.1328125,
+      "learning_rate": 3.282739426715697e-07,
+      "loss": 0.0131,
+      "step": 4672
+    },
+    {
+      "epoch": 333.7857142857143,
+      "grad_norm": 0.150390625,
+      "learning_rate": 3.2629010360906015e-07,
+      "loss": 0.0114,
+      "step": 4673
+    },
+    {
+      "epoch": 333.85714285714283,
+      "grad_norm": 0.185546875,
+      "learning_rate": 3.24312197864865e-07,
+      "loss": 0.0132,
+      "step": 4674
+    },
+    {
+      "epoch": 333.92857142857144,
+      "grad_norm": 0.12060546875,
+      "learning_rate": 3.223402264029937e-07,
+      "loss": 0.0139,
+      "step": 4675
+    },
+    {
+      "epoch": 334.0,
+      "grad_norm": 0.16796875,
+      "learning_rate": 3.203741901845608e-07,
+      "loss": 0.0125,
+      "step": 4676
+    },
+    {
+      "epoch": 334.07142857142856,
+      "grad_norm": 0.17578125,
+      "learning_rate": 3.1841409016778454e-07,
+      "loss": 0.0133,
+      "step": 4677
+    },
+    {
+      "epoch": 334.14285714285717,
+      "grad_norm": 0.1875,
+      "learning_rate": 3.1645992730800073e-07,
+      "loss": 0.0168,
+      "step": 4678
+    },
+    {
+      "epoch": 334.2142857142857,
+      "grad_norm": 0.11083984375,
+      "learning_rate": 3.145117025576405e-07,
+      "loss": 0.0114,
+      "step": 4679
+    },
+    {
+      "epoch": 334.2857142857143,
+      "grad_norm": 0.19921875,
+      "learning_rate": 3.1256941686624434e-07,
+      "loss": 0.0141,
+      "step": 4680
+    },
+    {
+      "epoch": 334.35714285714283,
+      "grad_norm": 0.115234375,
+      "learning_rate": 3.106330711804592e-07,
+      "loss": 0.0112,
+      "step": 4681
+    },
+    {
+      "epoch": 334.42857142857144,
+      "grad_norm": 0.26953125,
+      "learning_rate": 3.0870266644403843e-07,
+      "loss": 0.0138,
+      "step": 4682
+    },
+    {
+      "epoch": 334.5,
+      "grad_norm": 0.154296875,
+      "learning_rate": 3.0677820359783643e-07,
+      "loss": 0.013,
+      "step": 4683
+    },
+    {
+      "epoch": 334.57142857142856,
+      "grad_norm": 0.1494140625,
+      "learning_rate": 3.048596835798126e-07,
+      "loss": 0.0131,
+      "step": 4684
+    },
+    {
+      "epoch": 334.64285714285717,
+      "grad_norm": 0.1201171875,
+      "learning_rate": 3.0294710732503146e-07,
+      "loss": 0.0134,
+      "step": 4685
+    },
+    {
+      "epoch": 334.7142857142857,
+      "grad_norm": 0.1640625,
+      "learning_rate": 3.010404757656571e-07,
+      "loss": 0.0147,
+      "step": 4686
+    },
+    {
+      "epoch": 334.7857142857143,
+      "grad_norm": 0.12890625,
+      "learning_rate": 2.991397898309628e-07,
+      "loss": 0.0125,
+      "step": 4687
+    },
+    {
+      "epoch": 334.85714285714283,
+      "grad_norm": 0.13671875,
+      "learning_rate": 2.972450504473159e-07,
+      "loss": 0.0141,
+      "step": 4688
+    },
+    {
+      "epoch": 334.92857142857144,
+      "grad_norm": 0.1494140625,
+      "learning_rate": 2.953562585381903e-07,
+      "loss": 0.0142,
+      "step": 4689
+    },
+    {
+      "epoch": 335.0,
+      "grad_norm": 0.2236328125,
+      "learning_rate": 2.9347341502415917e-07,
+      "loss": 0.0139,
+      "step": 4690
+    },
+    {
+      "epoch": 335.07142857142856,
+      "grad_norm": 0.2021484375,
+      "learning_rate": 2.91596520822901e-07,
+      "loss": 0.0137,
+      "step": 4691
+    },
+    {
+      "epoch": 335.14285714285717,
+      "grad_norm": 0.115234375,
+      "learning_rate": 2.897255768491897e-07,
+      "loss": 0.0162,
+      "step": 4692
+    },
+    {
+      "epoch": 335.2142857142857,
+      "grad_norm": 0.1474609375,
+      "learning_rate": 2.8786058401489845e-07,
+      "loss": 0.0141,
+      "step": 4693
+    },
+    {
+      "epoch": 335.2857142857143,
+      "grad_norm": 0.1669921875,
+      "learning_rate": 2.8600154322900587e-07,
+      "loss": 0.0147,
+      "step": 4694
+    },
+    {
+      "epoch": 335.35714285714283,
+      "grad_norm": 0.1318359375,
+      "learning_rate": 2.841484553975815e-07,
+      "loss": 0.0141,
+      "step": 4695
+    },
+    {
+      "epoch": 335.42857142857144,
+      "grad_norm": 0.1474609375,
+      "learning_rate": 2.8230132142380156e-07,
+      "loss": 0.0129,
+      "step": 4696
+    },
+    {
+      "epoch": 335.5,
+      "grad_norm": 0.125,
+      "learning_rate": 2.8046014220793616e-07,
+      "loss": 0.014,
+      "step": 4697
+    },
+    {
+      "epoch": 335.57142857142856,
+      "grad_norm": 0.130859375,
+      "learning_rate": 2.786249186473508e-07,
+      "loss": 0.0115,
+      "step": 4698
+    },
+    {
+      "epoch": 335.64285714285717,
+      "grad_norm": 0.25,
+      "learning_rate": 2.767956516365147e-07,
+      "loss": 0.0123,
+      "step": 4699
+    },
+    {
+      "epoch": 335.7142857142857,
+      "grad_norm": 0.181640625,
+      "learning_rate": 2.7497234206698697e-07,
+      "loss": 0.0163,
+      "step": 4700
+    },
+    {
+      "epoch": 335.7857142857143,
+      "grad_norm": 0.16796875,
+      "learning_rate": 2.7315499082742893e-07,
+      "loss": 0.0144,
+      "step": 4701
+    },
+    {
+      "epoch": 335.85714285714283,
+      "grad_norm": 0.1875,
+      "learning_rate": 2.713435988035959e-07,
+      "loss": 0.0116,
+      "step": 4702
+    },
+    {
+      "epoch": 335.92857142857144,
+      "grad_norm": 0.126953125,
+      "learning_rate": 2.6953816687833725e-07,
+      "loss": 0.0127,
+      "step": 4703
+    },
+    {
+      "epoch": 336.0,
+      "grad_norm": 0.1162109375,
+      "learning_rate": 2.6773869593159915e-07,
+      "loss": 0.0131,
+      "step": 4704
+    },
+    {
+      "epoch": 336.07142857142856,
+      "grad_norm": 0.1669921875,
+      "learning_rate": 2.659451868404203e-07,
+      "loss": 0.0138,
+      "step": 4705
+    },
+    {
+      "epoch": 336.14285714285717,
+      "grad_norm": 0.1552734375,
+      "learning_rate": 2.6415764047893754e-07,
+      "loss": 0.0124,
+      "step": 4706
+    },
+    {
+      "epoch": 336.2142857142857,
+      "grad_norm": 0.2275390625,
+      "learning_rate": 2.623760577183776e-07,
+      "loss": 0.0149,
+      "step": 4707
+    },
+    {
+      "epoch": 336.2857142857143,
+      "grad_norm": 0.17578125,
+      "learning_rate": 2.606004394270639e-07,
+      "loss": 0.0127,
+      "step": 4708
+    },
+    {
+      "epoch": 336.35714285714283,
+      "grad_norm": 0.146484375,
+      "learning_rate": 2.588307864704084e-07,
+      "loss": 0.0133,
+      "step": 4709
+    },
+    {
+      "epoch": 336.42857142857144,
+      "grad_norm": 0.2021484375,
+      "learning_rate": 2.570670997109198e-07,
+      "loss": 0.0114,
+      "step": 4710
+    },
+    {
+      "epoch": 336.5,
+      "grad_norm": 0.220703125,
+      "learning_rate": 2.5530938000819933e-07,
+      "loss": 0.0162,
+      "step": 4711
+    },
+    {
+      "epoch": 336.57142857142856,
+      "grad_norm": 0.1337890625,
+      "learning_rate": 2.535576282189367e-07,
+      "loss": 0.0127,
+      "step": 4712
+    },
+    {
+      "epoch": 336.64285714285717,
+      "grad_norm": 0.11376953125,
+      "learning_rate": 2.5181184519691425e-07,
+      "loss": 0.012,
+      "step": 4713
+    },
+    {
+      "epoch": 336.7142857142857,
+      "grad_norm": 0.1435546875,
+      "learning_rate": 2.5007203179300556e-07,
+      "loss": 0.0118,
+      "step": 4714
+    },
+    {
+      "epoch": 336.7857142857143,
+      "grad_norm": 0.16796875,
+      "learning_rate": 2.4833818885517527e-07,
+      "loss": 0.0128,
+      "step": 4715
+    },
+    {
+      "epoch": 336.85714285714283,
+      "grad_norm": 0.125,
+      "learning_rate": 2.466103172284781e-07,
+      "loss": 0.0123,
+      "step": 4716
+    },
+    {
+      "epoch": 336.92857142857144,
+      "grad_norm": 0.1220703125,
+      "learning_rate": 2.448884177550542e-07,
+      "loss": 0.0134,
+      "step": 4717
+    },
+    {
+      "epoch": 337.0,
+      "grad_norm": 0.1259765625,
+      "learning_rate": 2.4317249127414063e-07,
+      "loss": 0.0127,
+      "step": 4718
+    },
+    {
+      "epoch": 337.07142857142856,
+      "grad_norm": 0.201171875,
+      "learning_rate": 2.414625386220545e-07,
+      "loss": 0.0143,
+      "step": 4719
+    },
+    {
+      "epoch": 337.14285714285717,
+      "grad_norm": 0.1201171875,
+      "learning_rate": 2.3975856063221114e-07,
+      "loss": 0.0142,
+      "step": 4720
+    },
+    {
+      "epoch": 337.2142857142857,
+      "grad_norm": 0.1318359375,
+      "learning_rate": 2.3806055813510587e-07,
+      "loss": 0.0119,
+      "step": 4721
+    },
+    {
+      "epoch": 337.2857142857143,
+      "grad_norm": 0.1357421875,
+      "learning_rate": 2.363685319583253e-07,
+      "loss": 0.012,
+      "step": 4722
+    },
+    {
+      "epoch": 337.35714285714283,
+      "grad_norm": 0.1416015625,
+      "learning_rate": 2.346824829265404e-07,
+      "loss": 0.0123,
+      "step": 4723
+    },
+    {
+      "epoch": 337.42857142857144,
+      "grad_norm": 0.14453125,
+      "learning_rate": 2.3300241186151318e-07,
+      "loss": 0.0139,
+      "step": 4724
+    },
+    {
+      "epoch": 337.5,
+      "grad_norm": 0.11328125,
+      "learning_rate": 2.3132831958208868e-07,
+      "loss": 0.0114,
+      "step": 4725
+    },
+    {
+      "epoch": 337.57142857142856,
+      "grad_norm": 0.14453125,
+      "learning_rate": 2.2966020690420033e-07,
+      "loss": 0.0128,
+      "step": 4726
+    },
+    {
+      "epoch": 337.64285714285717,
+      "grad_norm": 0.18359375,
+      "learning_rate": 2.2799807464086582e-07,
+      "loss": 0.0106,
+      "step": 4727
+    },
+    {
+      "epoch": 337.7142857142857,
+      "grad_norm": 0.158203125,
+      "learning_rate": 2.2634192360218853e-07,
+      "loss": 0.0134,
+      "step": 4728
+    },
+    {
+      "epoch": 337.7857142857143,
+      "grad_norm": 0.1162109375,
+      "learning_rate": 2.2469175459535335e-07,
+      "loss": 0.0129,
+      "step": 4729
+    },
+    {
+      "epoch": 337.85714285714283,
+      "grad_norm": 0.1318359375,
+      "learning_rate": 2.2304756842463498e-07,
+      "loss": 0.0134,
+      "step": 4730
+    },
+    {
+      "epoch": 337.92857142857144,
+      "grad_norm": 0.1396484375,
+      "learning_rate": 2.2140936589139238e-07,
+      "loss": 0.0147,
+      "step": 4731
+    },
+    {
+      "epoch": 338.0,
+      "grad_norm": 0.2158203125,
+      "learning_rate": 2.197771477940619e-07,
+      "loss": 0.0169,
+      "step": 4732
+    },
+    {
+      "epoch": 338.07142857142856,
+      "grad_norm": 0.130859375,
+      "learning_rate": 2.1815091492816548e-07,
+      "loss": 0.0161,
+      "step": 4733
+    },
+    {
+      "epoch": 338.14285714285717,
+      "grad_norm": 0.185546875,
+      "learning_rate": 2.1653066808631357e-07,
+      "loss": 0.0121,
+      "step": 4734
+    },
+    {
+      "epoch": 338.2142857142857,
+      "grad_norm": 0.1611328125,
+      "learning_rate": 2.1491640805819113e-07,
+      "loss": 0.0125,
+      "step": 4735
+    },
+    {
+      "epoch": 338.2857142857143,
+      "grad_norm": 0.1376953125,
+      "learning_rate": 2.13308135630573e-07,
+      "loss": 0.0132,
+      "step": 4736
+    },
+    {
+      "epoch": 338.35714285714283,
+      "grad_norm": 0.2001953125,
+      "learning_rate": 2.1170585158730715e-07,
+      "loss": 0.0124,
+      "step": 4737
+    },
+    {
+      "epoch": 338.42857142857144,
+      "grad_norm": 0.23828125,
+      "learning_rate": 2.1010955670932863e-07,
+      "loss": 0.0147,
+      "step": 4738
+    },
+    {
+      "epoch": 338.5,
+      "grad_norm": 0.1748046875,
+      "learning_rate": 2.0851925177465536e-07,
+      "loss": 0.0137,
+      "step": 4739
+    },
+    {
+      "epoch": 338.57142857142856,
+      "grad_norm": 0.28125,
+      "learning_rate": 2.0693493755837984e-07,
+      "loss": 0.0106,
+      "step": 4740
+    },
+    {
+      "epoch": 338.64285714285717,
+      "grad_norm": 0.16015625,
+      "learning_rate": 2.0535661483267744e-07,
+      "loss": 0.0145,
+      "step": 4741
+    },
+    {
+      "epoch": 338.7142857142857,
+      "grad_norm": 0.1748046875,
+      "learning_rate": 2.0378428436680364e-07,
+      "loss": 0.0112,
+      "step": 4742
+    },
+    {
+      "epoch": 338.7857142857143,
+      "grad_norm": 0.1708984375,
+      "learning_rate": 2.0221794692709268e-07,
+      "loss": 0.012,
+      "step": 4743
+    },
+    {
+      "epoch": 338.85714285714283,
+      "grad_norm": 0.134765625,
+      "learning_rate": 2.0065760327696025e-07,
+      "loss": 0.0116,
+      "step": 4744
+    },
+    {
+      "epoch": 338.92857142857144,
+      "grad_norm": 0.1923828125,
+      "learning_rate": 1.9910325417689807e-07,
+      "loss": 0.0122,
+      "step": 4745
+    },
+    {
+      "epoch": 339.0,
+      "grad_norm": 0.1484375,
+      "learning_rate": 1.975549003844765e-07,
+      "loss": 0.0118,
+      "step": 4746
+    },
+    {
+      "epoch": 339.07142857142856,
+      "grad_norm": 0.16796875,
+      "learning_rate": 1.9601254265434465e-07,
+      "loss": 0.0114,
+      "step": 4747
+    },
+    {
+      "epoch": 339.14285714285717,
+      "grad_norm": 0.142578125,
+      "learning_rate": 1.9447618173822762e-07,
+      "loss": 0.014,
+      "step": 4748
+    },
+    {
+      "epoch": 339.2142857142857,
+      "grad_norm": 0.1103515625,
+      "learning_rate": 1.9294581838493337e-07,
+      "loss": 0.0133,
+      "step": 4749
+    },
+    {
+      "epoch": 339.2857142857143,
+      "grad_norm": 0.1357421875,
+      "learning_rate": 1.9142145334033745e-07,
+      "loss": 0.0121,
+      "step": 4750
+    },
+    {
+      "epoch": 339.35714285714283,
+      "grad_norm": 0.318359375,
+      "learning_rate": 1.8990308734739976e-07,
+      "loss": 0.0128,
+      "step": 4751
+    },
+    {
+      "epoch": 339.42857142857144,
+      "grad_norm": 0.10888671875,
+      "learning_rate": 1.883907211461519e-07,
+      "loss": 0.0118,
+      "step": 4752
+    },
+    {
+      "epoch": 339.5,
+      "grad_norm": 0.2255859375,
+      "learning_rate": 1.8688435547370292e-07,
+      "loss": 0.0125,
+      "step": 4753
+    },
+    {
+      "epoch": 339.57142857142856,
+      "grad_norm": 0.2158203125,
+      "learning_rate": 1.853839910642391e-07,
+      "loss": 0.0119,
+      "step": 4754
+    },
+    {
+      "epoch": 339.64285714285717,
+      "grad_norm": 0.1982421875,
+      "learning_rate": 1.8388962864901855e-07,
+      "loss": 0.0126,
+      "step": 4755
+    },
+    {
+      "epoch": 339.7142857142857,
+      "grad_norm": 0.2265625,
+      "learning_rate": 1.8240126895637393e-07,
+      "loss": 0.01,
+      "step": 4756
+    },
+    {
+      "epoch": 339.7857142857143,
+      "grad_norm": 0.1845703125,
+      "learning_rate": 1.8091891271171528e-07,
+      "loss": 0.0152,
+      "step": 4757
+    },
+    {
+      "epoch": 339.85714285714283,
+      "grad_norm": 0.18359375,
+      "learning_rate": 1.794425606375244e-07,
+      "loss": 0.0136,
+      "step": 4758
+    },
+    {
+      "epoch": 339.92857142857144,
+      "grad_norm": 0.234375,
+      "learning_rate": 1.77972213453359e-07,
+      "loss": 0.0154,
+      "step": 4759
+    },
+    {
+      "epoch": 340.0,
+      "grad_norm": 0.1240234375,
+      "learning_rate": 1.7650787187584587e-07,
+      "loss": 0.0126,
+      "step": 4760
+    },
+    {
+      "epoch": 340.07142857142856,
+      "grad_norm": 0.1669921875,
+      "learning_rate": 1.7504953661868912e-07,
+      "loss": 0.0111,
+      "step": 4761
+    },
+    {
+      "epoch": 340.14285714285717,
+      "grad_norm": 0.12890625,
+      "learning_rate": 1.735972083926618e-07,
+      "loss": 0.0128,
+      "step": 4762
+    },
+    {
+      "epoch": 340.2142857142857,
+      "grad_norm": 0.1259765625,
+      "learning_rate": 1.72150887905613e-07,
+      "loss": 0.0143,
+      "step": 4763
+    },
+    {
+      "epoch": 340.2857142857143,
+      "grad_norm": 0.1455078125,
+      "learning_rate": 1.707105758624622e-07,
+      "loss": 0.0132,
+      "step": 4764
+    },
+    {
+      "epoch": 340.35714285714283,
+      "grad_norm": 0.1669921875,
+      "learning_rate": 1.6927627296519782e-07,
+      "loss": 0.0144,
+      "step": 4765
+    },
+    {
+      "epoch": 340.42857142857144,
+      "grad_norm": 0.1484375,
+      "learning_rate": 1.6784797991288291e-07,
+      "loss": 0.0147,
+      "step": 4766
+    },
+    {
+      "epoch": 340.5,
+      "grad_norm": 0.1728515625,
+      "learning_rate": 1.6642569740164955e-07,
+      "loss": 0.0122,
+      "step": 4767
+    },
+    {
+      "epoch": 340.57142857142856,
+      "grad_norm": 0.2001953125,
+      "learning_rate": 1.650094261247015e-07,
+      "loss": 0.0109,
+      "step": 4768
+    },
+    {
+      "epoch": 340.64285714285717,
+      "grad_norm": 0.1494140625,
+      "learning_rate": 1.6359916677231162e-07,
+      "loss": 0.0122,
+      "step": 4769
+    },
+    {
+      "epoch": 340.7142857142857,
+      "grad_norm": 0.1630859375,
+      "learning_rate": 1.6219492003182457e-07,
+      "loss": 0.0132,
+      "step": 4770
+    },
+    {
+      "epoch": 340.7857142857143,
+      "grad_norm": 0.1318359375,
+      "learning_rate": 1.607966865876512e-07,
+      "loss": 0.0125,
+      "step": 4771
+    },
+    {
+      "epoch": 340.85714285714283,
+      "grad_norm": 0.1328125,
+      "learning_rate": 1.5940446712127412e-07,
+      "loss": 0.0156,
+      "step": 4772
+    },
+    {
+      "epoch": 340.92857142857144,
+      "grad_norm": 0.119140625,
+      "learning_rate": 1.58018262311245e-07,
+      "loss": 0.0143,
+      "step": 4773
+    },
+    {
+      "epoch": 341.0,
+      "grad_norm": 0.1806640625,
+      "learning_rate": 1.5663807283318316e-07,
+      "loss": 0.014,
+      "step": 4774
+    },
+    {
+      "epoch": 341.07142857142856,
+      "grad_norm": 0.1240234375,
+      "learning_rate": 1.5526389935977686e-07,
+      "loss": 0.0133,
+      "step": 4775
+    },
+    {
+      "epoch": 341.14285714285717,
+      "grad_norm": 0.162109375,
+      "learning_rate": 1.5389574256077926e-07,
+      "loss": 0.0144,
+      "step": 4776
+    },
+    {
+      "epoch": 341.2142857142857,
+      "grad_norm": 0.12353515625,
+      "learning_rate": 1.525336031030139e-07,
+      "loss": 0.0137,
+      "step": 4777
+    },
+    {
+      "epoch": 341.2857142857143,
+      "grad_norm": 0.29296875,
+      "learning_rate": 1.5117748165037194e-07,
+      "loss": 0.0156,
+      "step": 4778
+    },
+    {
+      "epoch": 341.35714285714283,
+      "grad_norm": 0.1533203125,
+      "learning_rate": 1.4982737886381077e-07,
+      "loss": 0.0134,
+      "step": 4779
+    },
+    {
+      "epoch": 341.42857142857144,
+      "grad_norm": 0.228515625,
+      "learning_rate": 1.4848329540135402e-07,
+      "loss": 0.0143,
+      "step": 4780
+    },
+    {
+      "epoch": 341.5,
+      "grad_norm": 0.1484375,
+      "learning_rate": 1.471452319180902e-07,
+      "loss": 0.0141,
+      "step": 4781
+    },
+    {
+      "epoch": 341.57142857142856,
+      "grad_norm": 0.177734375,
+      "learning_rate": 1.458131890661754e-07,
+      "loss": 0.0128,
+      "step": 4782
+    },
+    {
+      "epoch": 341.64285714285717,
+      "grad_norm": 0.1669921875,
+      "learning_rate": 1.4448716749483338e-07,
+      "loss": 0.0146,
+      "step": 4783
+    },
+    {
+      "epoch": 341.7142857142857,
+      "grad_norm": 0.130859375,
+      "learning_rate": 1.431671678503499e-07,
+      "loss": 0.0134,
+      "step": 4784
+    },
+    {
+      "epoch": 341.7857142857143,
+      "grad_norm": 0.28125,
+      "learning_rate": 1.4185319077607561e-07,
+      "loss": 0.0132,
+      "step": 4785
+    },
+    {
+      "epoch": 341.85714285714283,
+      "grad_norm": 0.28125,
+      "learning_rate": 1.4054523691242744e-07,
+      "loss": 0.0127,
+      "step": 4786
+    },
+    {
+      "epoch": 341.92857142857144,
+      "grad_norm": 0.2080078125,
+      "learning_rate": 1.392433068968857e-07,
+      "loss": 0.0157,
+      "step": 4787
+    },
+    {
+      "epoch": 342.0,
+      "grad_norm": 0.1259765625,
+      "learning_rate": 1.37947401363997e-07,
+      "loss": 0.0119,
+      "step": 4788
+    },
+    {
+      "epoch": 342.07142857142856,
+      "grad_norm": 0.189453125,
+      "learning_rate": 1.3665752094536994e-07,
+      "loss": 0.0123,
+      "step": 4789
+    },
+    {
+      "epoch": 342.14285714285717,
+      "grad_norm": 0.14453125,
+      "learning_rate": 1.3537366626967391e-07,
+      "loss": 0.0131,
+      "step": 4790
+    },
+    {
+      "epoch": 342.2142857142857,
+      "grad_norm": 0.2451171875,
+      "learning_rate": 1.3409583796264585e-07,
+      "loss": 0.0138,
+      "step": 4791
+    },
+    {
+      "epoch": 342.2857142857143,
+      "grad_norm": 0.1748046875,
+      "learning_rate": 1.328240366470848e-07,
+      "loss": 0.0133,
+      "step": 4792
+    },
+    {
+      "epoch": 342.35714285714283,
+      "grad_norm": 0.1396484375,
+      "learning_rate": 1.3155826294284912e-07,
+      "loss": 0.0105,
+      "step": 4793
+    },
+    {
+      "epoch": 342.42857142857144,
+      "grad_norm": 0.146484375,
+      "learning_rate": 1.3029851746686335e-07,
+      "loss": 0.0134,
+      "step": 4794
+    },
+    {
+      "epoch": 342.5,
+      "grad_norm": 0.1318359375,
+      "learning_rate": 1.2904480083310994e-07,
+      "loss": 0.0127,
+      "step": 4795
+    },
+    {
+      "epoch": 342.57142857142856,
+      "grad_norm": 0.11181640625,
+      "learning_rate": 1.2779711365263764e-07,
+      "loss": 0.012,
+      "step": 4796
+    },
+    {
+      "epoch": 342.64285714285717,
+      "grad_norm": 0.1318359375,
+      "learning_rate": 1.2655545653355304e-07,
+      "loss": 0.0141,
+      "step": 4797
+    },
+    {
+      "epoch": 342.7142857142857,
+      "grad_norm": 0.11279296875,
+      "learning_rate": 1.2531983008102487e-07,
+      "loss": 0.0112,
+      "step": 4798
+    },
+    {
+      "epoch": 342.7857142857143,
+      "grad_norm": 0.1318359375,
+      "learning_rate": 1.2409023489728249e-07,
+      "loss": 0.0135,
+      "step": 4799
+    },
+    {
+      "epoch": 342.85714285714283,
+      "grad_norm": 0.171875,
+      "learning_rate": 1.228666715816132e-07,
+      "loss": 0.0153,
+      "step": 4800
+    },
+    {
+      "epoch": 342.92857142857144,
+      "grad_norm": 0.1494140625,
+      "learning_rate": 1.2164914073037048e-07,
+      "loss": 0.0131,
+      "step": 4801
+    },
+    {
+      "epoch": 343.0,
+      "grad_norm": 0.1240234375,
+      "learning_rate": 1.204376429369644e-07,
+      "loss": 0.011,
+      "step": 4802
+    },
+    {
+      "epoch": 343.07142857142856,
+      "grad_norm": 0.1494140625,
+      "learning_rate": 1.1923217879186016e-07,
+      "loss": 0.0135,
+      "step": 4803
+    },
+    {
+      "epoch": 343.14285714285717,
+      "grad_norm": 0.1767578125,
+      "learning_rate": 1.1803274888258914e-07,
+      "loss": 0.0124,
+      "step": 4804
+    },
+    {
+      "epoch": 343.2142857142857,
+      "grad_norm": 0.177734375,
+      "learning_rate": 1.1683935379373928e-07,
+      "loss": 0.0131,
+      "step": 4805
+    },
+    {
+      "epoch": 343.2857142857143,
+      "grad_norm": 0.126953125,
+      "learning_rate": 1.1565199410695504e-07,
+      "loss": 0.0142,
+      "step": 4806
+    },
+    {
+      "epoch": 343.35714285714283,
+      "grad_norm": 0.15234375,
+      "learning_rate": 1.1447067040094295e-07,
+      "loss": 0.0148,
+      "step": 4807
+    },
+    {
+      "epoch": 343.42857142857144,
+      "grad_norm": 0.2890625,
+      "learning_rate": 1.1329538325146466e-07,
+      "loss": 0.0118,
+      "step": 4808
+    },
+    {
+      "epoch": 343.5,
+      "grad_norm": 0.22265625,
+      "learning_rate": 1.1212613323134114e-07,
+      "loss": 0.015,
+      "step": 4809
+    },
+    {
+      "epoch": 343.57142857142856,
+      "grad_norm": 0.1328125,
+      "learning_rate": 1.1096292091044985e-07,
+      "loss": 0.0127,
+      "step": 4810
+    },
+    {
+      "epoch": 343.64285714285717,
+      "grad_norm": 0.1455078125,
+      "learning_rate": 1.0980574685572897e-07,
+      "loss": 0.0126,
+      "step": 4811
+    },
+    {
+      "epoch": 343.7142857142857,
+      "grad_norm": 0.201171875,
+      "learning_rate": 1.0865461163116902e-07,
+      "loss": 0.0142,
+      "step": 4812
+    },
+    {
+      "epoch": 343.7857142857143,
+      "grad_norm": 0.146484375,
+      "learning_rate": 1.0750951579781843e-07,
+      "loss": 0.0126,
+      "step": 4813
+    },
+    {
+      "epoch": 343.85714285714283,
+      "grad_norm": 0.228515625,
+      "learning_rate": 1.0637045991378353e-07,
+      "loss": 0.0102,
+      "step": 4814
+    },
+    {
+      "epoch": 343.92857142857144,
+      "grad_norm": 0.1533203125,
+      "learning_rate": 1.0523744453422856e-07,
+      "loss": 0.0132,
+      "step": 4815
+    },
+    {
+      "epoch": 344.0,
+      "grad_norm": 0.146484375,
+      "learning_rate": 1.0411047021137016e-07,
+      "loss": 0.0132,
+      "step": 4816
+    },
+    {
+      "epoch": 344.07142857142856,
+      "grad_norm": 0.1552734375,
+      "learning_rate": 1.0298953749448003e-07,
+      "loss": 0.0155,
+      "step": 4817
+    },
+    {
+      "epoch": 344.14285714285717,
+      "grad_norm": 0.10546875,
+      "learning_rate": 1.0187464692989063e-07,
+      "loss": 0.0141,
+      "step": 4818
+    },
+    {
+      "epoch": 344.2142857142857,
+      "grad_norm": 0.12890625,
+      "learning_rate": 1.0076579906098255e-07,
+      "loss": 0.0143,
+      "step": 4819
+    },
+    {
+      "epoch": 344.2857142857143,
+      "grad_norm": 0.1640625,
+      "learning_rate": 9.96629944281971e-08,
+      "loss": 0.0132,
+      "step": 4820
+    },
+    {
+      "epoch": 344.35714285714283,
+      "grad_norm": 0.12890625,
+      "learning_rate": 9.856623356902794e-08,
+      "loss": 0.0155,
+      "step": 4821
+    },
+    {
+      "epoch": 344.42857142857144,
+      "grad_norm": 0.2333984375,
+      "learning_rate": 9.74755170180211e-08,
+      "loss": 0.0156,
+      "step": 4822
+    },
+    {
+      "epoch": 344.5,
+      "grad_norm": 0.1484375,
+      "learning_rate": 9.639084530677911e-08,
+      "loss": 0.0114,
+      "step": 4823
+    },
+    {
+      "epoch": 344.57142857142856,
+      "grad_norm": 0.1376953125,
+      "learning_rate": 9.531221896395965e-08,
+      "loss": 0.0131,
+      "step": 4824
+    },
+    {
+      "epoch": 344.64285714285717,
+      "grad_norm": 0.140625,
+      "learning_rate": 9.423963851526995e-08,
+      "loss": 0.0141,
+      "step": 4825
+    },
+    {
+      "epoch": 344.7142857142857,
+      "grad_norm": 0.15625,
+      "learning_rate": 9.317310448347521e-08,
+      "loss": 0.0134,
+      "step": 4826
+    },
+    {
+      "epoch": 344.7857142857143,
+      "grad_norm": 0.1396484375,
+      "learning_rate": 9.21126173883874e-08,
+      "loss": 0.0128,
+      "step": 4827
+    },
+    {
+      "epoch": 344.85714285714283,
+      "grad_norm": 0.12890625,
+      "learning_rate": 9.105817774687914e-08,
+      "loss": 0.0123,
+      "step": 4828
+    },
+    {
+      "epoch": 344.92857142857144,
+      "grad_norm": 0.1240234375,
+      "learning_rate": 9.000978607286853e-08,
+      "loss": 0.0129,
+      "step": 4829
+    },
+    {
+      "epoch": 345.0,
+      "grad_norm": 0.20703125,
+      "learning_rate": 8.896744287733017e-08,
+      "loss": 0.014,
+      "step": 4830
+    },
+    {
+      "epoch": 345.07142857142856,
+      "grad_norm": 0.1708984375,
+      "learning_rate": 8.793114866828822e-08,
+      "loss": 0.014,
+      "step": 4831
+    },
+    {
+      "epoch": 345.14285714285717,
+      "grad_norm": 0.2431640625,
+      "learning_rate": 8.690090395082201e-08,
+      "loss": 0.0145,
+      "step": 4832
+    },
+    {
+      "epoch": 345.2142857142857,
+      "grad_norm": 0.2451171875,
+      "learning_rate": 8.587670922705765e-08,
+      "loss": 0.0149,
+      "step": 4833
+    },
+    {
+      "epoch": 345.2857142857143,
+      "grad_norm": 0.1435546875,
+      "learning_rate": 8.485856499617778e-08,
+      "loss": 0.0122,
+      "step": 4834
+    },
+    {
+      "epoch": 345.35714285714283,
+      "grad_norm": 0.2294921875,
+      "learning_rate": 8.384647175441324e-08,
+      "loss": 0.0143,
+      "step": 4835
+    },
+    {
+      "epoch": 345.42857142857144,
+      "grad_norm": 0.1298828125,
+      "learning_rate": 8.284042999504444e-08,
+      "loss": 0.0118,
+      "step": 4836
+    },
+    {
+      "epoch": 345.5,
+      "grad_norm": 0.126953125,
+      "learning_rate": 8.184044020840554e-08,
+      "loss": 0.0127,
+      "step": 4837
+    },
+    {
+      "epoch": 345.57142857142856,
+      "grad_norm": 0.21484375,
+      "learning_rate": 8.08465028818775e-08,
+      "loss": 0.0159,
+      "step": 4838
+    },
+    {
+      "epoch": 345.64285714285717,
+      "grad_norm": 0.138671875,
+      "learning_rate": 7.985861849989507e-08,
+      "loss": 0.0127,
+      "step": 4839
+    },
+    {
+      "epoch": 345.7142857142857,
+      "grad_norm": 0.1904296875,
+      "learning_rate": 7.887678754393974e-08,
+      "loss": 0.0138,
+      "step": 4840
+    },
+    {
+      "epoch": 345.7857142857143,
+      "grad_norm": 0.2119140625,
+      "learning_rate": 7.790101049254538e-08,
+      "loss": 0.0134,
+      "step": 4841
+    },
+    {
+      "epoch": 345.85714285714283,
+      "grad_norm": 0.2197265625,
+      "learning_rate": 7.693128782129271e-08,
+      "loss": 0.0138,
+      "step": 4842
+    },
+    {
+      "epoch": 345.92857142857144,
+      "grad_norm": 0.234375,
+      "learning_rate": 7.596762000281337e-08,
+      "loss": 0.0108,
+      "step": 4843
+    },
+    {
+      "epoch": 346.0,
+      "grad_norm": 0.1669921875,
+      "learning_rate": 7.501000750678722e-08,
+      "loss": 0.0134,
+      "step": 4844
+    },
+    {
+      "epoch": 346.07142857142856,
+      "grad_norm": 0.2333984375,
+      "learning_rate": 7.405845079994227e-08,
+      "loss": 0.0109,
+      "step": 4845
+    },
+    {
+      "epoch": 346.14285714285717,
+      "grad_norm": 0.265625,
+      "learning_rate": 7.311295034605758e-08,
+      "loss": 0.0167,
+      "step": 4846
+    },
+    {
+      "epoch": 346.2142857142857,
+      "grad_norm": 0.1259765625,
+      "learning_rate": 7.217350660595618e-08,
+      "loss": 0.0102,
+      "step": 4847
+    },
+    {
+      "epoch": 346.2857142857143,
+      "grad_norm": 0.22265625,
+      "learning_rate": 7.124012003751345e-08,
+      "loss": 0.0132,
+      "step": 4848
+    },
+    {
+      "epoch": 346.35714285714283,
+      "grad_norm": 0.1416015625,
+      "learning_rate": 7.031279109565025e-08,
+      "loss": 0.0134,
+      "step": 4849
+    },
+    {
+      "epoch": 346.42857142857144,
+      "grad_norm": 0.138671875,
+      "learning_rate": 6.93915202323342e-08,
+      "loss": 0.0128,
+      "step": 4850
+    },
+    {
+      "epoch": 346.5,
+      "grad_norm": 0.12451171875,
+      "learning_rate": 6.847630789658388e-08,
+      "loss": 0.0136,
+      "step": 4851
+    },
+    {
+      "epoch": 346.57142857142856,
+      "grad_norm": 0.1279296875,
+      "learning_rate": 6.756715453445917e-08,
+      "loss": 0.0122,
+      "step": 4852
+    },
+    {
+      "epoch": 346.64285714285717,
+      "grad_norm": 0.2412109375,
+      "learning_rate": 6.666406058907366e-08,
+      "loss": 0.0138,
+      "step": 4853
+    },
+    {
+      "epoch": 346.7142857142857,
+      "grad_norm": 0.2158203125,
+      "learning_rate": 6.576702650058358e-08,
+      "loss": 0.0105,
+      "step": 4854
+    },
+    {
+      "epoch": 346.7857142857143,
+      "grad_norm": 0.138671875,
+      "learning_rate": 6.487605270619201e-08,
+      "loss": 0.0139,
+      "step": 4855
+    },
+    {
+      "epoch": 346.85714285714283,
+      "grad_norm": 0.1142578125,
+      "learning_rate": 6.399113964014741e-08,
+      "loss": 0.0124,
+      "step": 4856
+    },
+    {
+      "epoch": 346.92857142857144,
+      "grad_norm": 0.181640625,
+      "learning_rate": 6.311228773374783e-08,
+      "loss": 0.0126,
+      "step": 4857
+    },
+    {
+      "epoch": 347.0,
+      "grad_norm": 0.1240234375,
+      "learning_rate": 6.223949741533397e-08,
+      "loss": 0.0145,
+      "step": 4858
+    },
+    {
+      "epoch": 347.07142857142856,
+      "grad_norm": 0.1328125,
+      "learning_rate": 6.137276911029471e-08,
+      "loss": 0.0142,
+      "step": 4859
+    },
+    {
+      "epoch": 347.14285714285717,
+      "grad_norm": 0.2333984375,
+      "learning_rate": 6.051210324106021e-08,
+      "loss": 0.0126,
+      "step": 4860
+    },
+    {
+      "epoch": 347.2142857142857,
+      "grad_norm": 0.134765625,
+      "learning_rate": 5.96575002271102e-08,
+      "loss": 0.0134,
+      "step": 4861
+    },
+    {
+      "epoch": 347.2857142857143,
+      "grad_norm": 0.240234375,
+      "learning_rate": 5.8808960484967034e-08,
+      "loss": 0.0154,
+      "step": 4862
+    },
+    {
+      "epoch": 347.35714285714283,
+      "grad_norm": 0.142578125,
+      "learning_rate": 5.7966484428201294e-08,
+      "loss": 0.0162,
+      "step": 4863
+    },
+    {
+      "epoch": 347.42857142857144,
+      "grad_norm": 0.306640625,
+      "learning_rate": 5.713007246742202e-08,
+      "loss": 0.013,
+      "step": 4864
+    },
+    {
+      "epoch": 347.5,
+      "grad_norm": 0.162109375,
+      "learning_rate": 5.629972501028785e-08,
+      "loss": 0.0143,
+      "step": 4865
+    },
+    {
+      "epoch": 347.57142857142856,
+      "grad_norm": 0.1552734375,
+      "learning_rate": 5.547544246150005e-08,
+      "loss": 0.0149,
+      "step": 4866
+    },
+    {
+      "epoch": 347.64285714285717,
+      "grad_norm": 0.1279296875,
+      "learning_rate": 5.4657225222803924e-08,
+      "loss": 0.0157,
+      "step": 4867
+    },
+    {
+      "epoch": 347.7142857142857,
+      "grad_norm": 0.173828125,
+      "learning_rate": 5.38450736929888e-08,
+      "loss": 0.0149,
+      "step": 4868
+    },
+    {
+      "epoch": 347.7857142857143,
+      "grad_norm": 0.1787109375,
+      "learning_rate": 5.303898826788667e-08,
+      "loss": 0.0139,
+      "step": 4869
+    },
+    {
+      "epoch": 347.85714285714283,
+      "grad_norm": 0.134765625,
+      "learning_rate": 5.2238969340373535e-08,
+      "loss": 0.0129,
+      "step": 4870
+    },
+    {
+      "epoch": 347.92857142857144,
+      "grad_norm": 0.23046875,
+      "learning_rate": 5.1445017300370824e-08,
+      "loss": 0.0158,
+      "step": 4871
+    },
+    {
+      "epoch": 348.0,
+      "grad_norm": 0.15625,
+      "learning_rate": 5.0657132534838444e-08,
+      "loss": 0.0162,
+      "step": 4872
+    },
+    {
+      "epoch": 348.07142857142856,
+      "grad_norm": 0.1328125,
+      "learning_rate": 4.9875315427785896e-08,
+      "loss": 0.0118,
+      "step": 4873
+    },
+    {
+      "epoch": 348.14285714285717,
+      "grad_norm": 0.115234375,
+      "learning_rate": 4.909956636025698e-08,
+      "loss": 0.0117,
+      "step": 4874
+    },
+    {
+      "epoch": 348.2142857142857,
+      "grad_norm": 0.11279296875,
+      "learning_rate": 4.832988571034369e-08,
+      "loss": 0.012,
+      "step": 4875
+    },
+    {
+      "epoch": 348.2857142857143,
+      "grad_norm": 0.25390625,
+      "learning_rate": 4.756627385318069e-08,
+      "loss": 0.0135,
+      "step": 4876
+    },
+    {
+      "epoch": 348.35714285714283,
+      "grad_norm": 0.16015625,
+      "learning_rate": 4.6808731160942476e-08,
+      "loss": 0.0133,
+      "step": 4877
+    },
+    {
+      "epoch": 348.42857142857144,
+      "grad_norm": 0.19140625,
+      "learning_rate": 4.605725800284483e-08,
+      "loss": 0.015,
+      "step": 4878
+    },
+    {
+      "epoch": 348.5,
+      "grad_norm": 0.17578125,
+      "learning_rate": 4.5311854745146155e-08,
+      "loss": 0.0119,
+      "step": 4879
+    },
+    {
+      "epoch": 348.57142857142856,
+      "grad_norm": 0.119140625,
+      "learning_rate": 4.457252175115029e-08,
+      "loss": 0.0124,
+      "step": 4880
+    },
+    {
+      "epoch": 348.64285714285717,
+      "grad_norm": 0.1943359375,
+      "learning_rate": 4.383925938119537e-08,
+      "loss": 0.0126,
+      "step": 4881
+    },
+    {
+      "epoch": 348.7142857142857,
+      "grad_norm": 0.33984375,
+      "learning_rate": 4.3112067992667756e-08,
+      "loss": 0.012,
+      "step": 4882
+    },
+    {
+      "epoch": 348.7857142857143,
+      "grad_norm": 0.12451171875,
+      "learning_rate": 4.239094793998949e-08,
+      "loss": 0.0123,
+      "step": 4883
+    },
+    {
+      "epoch": 348.85714285714283,
+      "grad_norm": 0.12060546875,
+      "learning_rate": 4.167589957462664e-08,
+      "loss": 0.0115,
+      "step": 4884
+    },
+    {
+      "epoch": 348.92857142857144,
+      "grad_norm": 0.1171875,
+      "learning_rate": 4.096692324508516e-08,
+      "loss": 0.012,
+      "step": 4885
+    },
+    {
+      "epoch": 349.0,
+      "grad_norm": 0.21484375,
+      "learning_rate": 4.026401929691087e-08,
+      "loss": 0.0132,
+      "step": 4886
+    },
+    {
+      "epoch": 349.07142857142856,
+      "grad_norm": 0.1962890625,
+      "learning_rate": 3.956718807269083e-08,
+      "loss": 0.0128,
+      "step": 4887
+    },
+    {
+      "epoch": 349.14285714285717,
+      "grad_norm": 0.15234375,
+      "learning_rate": 3.887642991205198e-08,
+      "loss": 0.0164,
+      "step": 4888
+    },
+    {
+      "epoch": 349.2142857142857,
+      "grad_norm": 0.1552734375,
+      "learning_rate": 3.8191745151662517e-08,
+      "loss": 0.0106,
+      "step": 4889
+    },
+    {
+      "epoch": 349.2857142857143,
+      "grad_norm": 0.146484375,
+      "learning_rate": 3.751313412522772e-08,
+      "loss": 0.0144,
+      "step": 4890
+    },
+    {
+      "epoch": 349.35714285714283,
+      "grad_norm": 0.142578125,
+      "learning_rate": 3.684059716349691e-08,
+      "loss": 0.0136,
+      "step": 4891
+    },
+    {
+      "epoch": 349.42857142857144,
+      "grad_norm": 0.14453125,
+      "learning_rate": 3.6174134594253725e-08,
+      "loss": 0.0167,
+      "step": 4892
+    },
+    {
+      "epoch": 349.5,
+      "grad_norm": 0.171875,
+      "learning_rate": 3.5513746742325826e-08,
+      "loss": 0.014,
+      "step": 4893
+    },
+    {
+      "epoch": 349.57142857142856,
+      "grad_norm": 0.11767578125,
+      "learning_rate": 3.4859433929577966e-08,
+      "loss": 0.0138,
+      "step": 4894
+    },
+    {
+      "epoch": 349.64285714285717,
+      "grad_norm": 0.1240234375,
+      "learning_rate": 3.4211196474912e-08,
+      "loss": 0.0118,
+      "step": 4895
+    },
+    {
+      "epoch": 349.7142857142857,
+      "grad_norm": 0.1171875,
+      "learning_rate": 3.35690346942738e-08,
+      "loss": 0.0122,
+      "step": 4896
+    },
+    {
+      "epoch": 349.7857142857143,
+      "grad_norm": 0.125,
+      "learning_rate": 3.293294890064358e-08,
+      "loss": 0.0149,
+      "step": 4897
+    },
+    {
+      "epoch": 349.85714285714283,
+      "grad_norm": 0.13671875,
+      "learning_rate": 3.2302939404042774e-08,
+      "loss": 0.0132,
+      "step": 4898
+    },
+    {
+      "epoch": 349.92857142857144,
+      "grad_norm": 0.1943359375,
+      "learning_rate": 3.167900651153133e-08,
+      "loss": 0.0148,
+      "step": 4899
+    },
+    {
+      "epoch": 350.0,
+      "grad_norm": 0.14453125,
+      "learning_rate": 3.106115052720349e-08,
+      "loss": 0.0134,
+      "step": 4900
+    },
+    {
+      "epoch": 350.07142857142856,
+      "grad_norm": 0.1572265625,
+      "learning_rate": 3.044937175219753e-08,
+      "loss": 0.0131,
+      "step": 4901
+    },
+    {
+      "epoch": 350.14285714285717,
+      "grad_norm": 0.197265625,
+      "learning_rate": 2.984367048468467e-08,
+      "loss": 0.0147,
+      "step": 4902
+    },
+    {
+      "epoch": 350.2142857142857,
+      "grad_norm": 0.1708984375,
+      "learning_rate": 2.9244047019878764e-08,
+      "loss": 0.012,
+      "step": 4903
+    },
+    {
+      "epoch": 350.2857142857143,
+      "grad_norm": 0.1572265625,
+      "learning_rate": 2.8650501650027972e-08,
+      "loss": 0.0125,
+      "step": 4904
+    },
+    {
+      "epoch": 350.35714285714283,
+      "grad_norm": 0.173828125,
+      "learning_rate": 2.8063034664417566e-08,
+      "loss": 0.0138,
+      "step": 4905
+    },
+    {
+      "epoch": 350.42857142857144,
+      "grad_norm": 0.1982421875,
+      "learning_rate": 2.7481646349375445e-08,
+      "loss": 0.0146,
+      "step": 4906
+    },
+    {
+      "epoch": 350.5,
+      "grad_norm": 0.1259765625,
+      "learning_rate": 2.690633698825967e-08,
+      "loss": 0.0119,
+      "step": 4907
+    },
+    {
+      "epoch": 350.57142857142856,
+      "grad_norm": 0.177734375,
+      "learning_rate": 2.633710686147095e-08,
+      "loss": 0.0157,
+      "step": 4908
+    },
+    {
+      "epoch": 350.64285714285717,
+      "grad_norm": 0.123046875,
+      "learning_rate": 2.5773956246444298e-08,
+      "loss": 0.0112,
+      "step": 4909
+    },
+    {
+      "epoch": 350.7142857142857,
+      "grad_norm": 0.11083984375,
+      "learning_rate": 2.5216885417651826e-08,
+      "loss": 0.0144,
+      "step": 4910
+    },
+    {
+      "epoch": 350.7857142857143,
+      "grad_norm": 0.11865234375,
+      "learning_rate": 2.4665894646605513e-08,
+      "loss": 0.0105,
+      "step": 4911
+    },
+    {
+      "epoch": 350.85714285714283,
+      "grad_norm": 0.2001953125,
+      "learning_rate": 2.412098420185027e-08,
+      "loss": 0.0131,
+      "step": 4912
+    },
+    {
+      "epoch": 350.92857142857144,
+      "grad_norm": 0.267578125,
+      "learning_rate": 2.358215434896671e-08,
+      "loss": 0.0183,
+      "step": 4913
+    },
+    {
+      "epoch": 351.0,
+      "grad_norm": 0.1337890625,
+      "learning_rate": 2.304940535057809e-08,
+      "loss": 0.0143,
+      "step": 4914
+    },
+    {
+      "epoch": 351.07142857142856,
+      "grad_norm": 0.1611328125,
+      "learning_rate": 2.252273746633643e-08,
+      "loss": 0.012,
+      "step": 4915
+    },
+    {
+      "epoch": 351.14285714285717,
+      "grad_norm": 0.1279296875,
+      "learning_rate": 2.200215095293362e-08,
+      "loss": 0.012,
+      "step": 4916
+    },
+    {
+      "epoch": 351.2142857142857,
+      "grad_norm": 0.1591796875,
+      "learning_rate": 2.1487646064100032e-08,
+      "loss": 0.0157,
+      "step": 4917
+    },
+    {
+      "epoch": 351.2857142857143,
+      "grad_norm": 0.12451171875,
+      "learning_rate": 2.0979223050596187e-08,
+      "loss": 0.0107,
+      "step": 4918
+    },
+    {
+      "epoch": 351.35714285714283,
+      "grad_norm": 0.1474609375,
+      "learning_rate": 2.047688216022109e-08,
+      "loss": 0.0149,
+      "step": 4919
+    },
+    {
+      "epoch": 351.42857142857144,
+      "grad_norm": 0.12890625,
+      "learning_rate": 1.9980623637812224e-08,
+      "loss": 0.013,
+      "step": 4920
+    },
+    {
+      "epoch": 351.5,
+      "grad_norm": 0.123046875,
+      "learning_rate": 1.9490447725237227e-08,
+      "loss": 0.0122,
+      "step": 4921
+    },
+    {
+      "epoch": 351.57142857142856,
+      "grad_norm": 0.126953125,
+      "learning_rate": 1.90063546614036e-08,
+      "loss": 0.0134,
+      "step": 4922
+    },
+    {
+      "epoch": 351.64285714285717,
+      "grad_norm": 0.193359375,
+      "learning_rate": 1.8528344682251786e-08,
+      "loss": 0.0157,
+      "step": 4923
+    },
+    {
+      "epoch": 351.7142857142857,
+      "grad_norm": 0.2080078125,
+      "learning_rate": 1.8056418020759302e-08,
+      "loss": 0.0115,
+      "step": 4924
+    },
+    {
+      "epoch": 351.7857142857143,
+      "grad_norm": 0.248046875,
+      "learning_rate": 1.759057490693522e-08,
+      "loss": 0.0154,
+      "step": 4925
+    },
+    {
+      "epoch": 351.85714285714283,
+      "grad_norm": 0.1474609375,
+      "learning_rate": 1.7130815567827085e-08,
+      "loss": 0.016,
+      "step": 4926
+    },
+    {
+      "epoch": 351.92857142857144,
+      "grad_norm": 0.2060546875,
+      "learning_rate": 1.6677140227515375e-08,
+      "loss": 0.0136,
+      "step": 4927
+    },
+    {
+      "epoch": 352.0,
+      "grad_norm": 0.1513671875,
+      "learning_rate": 1.6229549107116272e-08,
+      "loss": 0.0126,
+      "step": 4928
+    },
+    {
+      "epoch": 352.07142857142856,
+      "grad_norm": 0.1904296875,
+      "learning_rate": 1.578804242478166e-08,
+      "loss": 0.013,
+      "step": 4929
+    },
+    {
+      "epoch": 352.14285714285717,
+      "grad_norm": 0.1640625,
+      "learning_rate": 1.5352620395693572e-08,
+      "loss": 0.0127,
+      "step": 4930
+    },
+    {
+      "epoch": 352.2142857142857,
+      "grad_norm": 0.1865234375,
+      "learning_rate": 1.4923283232075313e-08,
+      "loss": 0.0111,
+      "step": 4931
+    },
+    {
+      "epoch": 352.2857142857143,
+      "grad_norm": 0.197265625,
+      "learning_rate": 1.4500031143176162e-08,
+      "loss": 0.0135,
+      "step": 4932
+    },
+    {
+      "epoch": 352.35714285714283,
+      "grad_norm": 0.1494140625,
+      "learning_rate": 1.408286433528666e-08,
+      "loss": 0.0108,
+      "step": 4933
+    },
+    {
+      "epoch": 352.42857142857144,
+      "grad_norm": 0.1708984375,
+      "learning_rate": 1.3671783011730277e-08,
+      "loss": 0.015,
+      "step": 4934
+    },
+    {
+      "epoch": 352.5,
+      "grad_norm": 0.1435546875,
+      "learning_rate": 1.3266787372859247e-08,
+      "loss": 0.0138,
+      "step": 4935
+    },
+    {
+      "epoch": 352.57142857142856,
+      "grad_norm": 0.203125,
+      "learning_rate": 1.2867877616067059e-08,
+      "loss": 0.0126,
+      "step": 4936
+    },
+    {
+      "epoch": 352.64285714285717,
+      "grad_norm": 0.1142578125,
+      "learning_rate": 1.2475053935774584e-08,
+      "loss": 0.0116,
+      "step": 4937
+    },
+    {
+      "epoch": 352.7142857142857,
+      "grad_norm": 0.1845703125,
+      "learning_rate": 1.2088316523441167e-08,
+      "loss": 0.012,
+      "step": 4938
+    },
+    {
+      "epoch": 352.7857142857143,
+      "grad_norm": 0.12353515625,
+      "learning_rate": 1.1707665567557702e-08,
+      "loss": 0.0147,
+      "step": 4939
+    },
+    {
+      "epoch": 352.85714285714283,
+      "grad_norm": 0.1728515625,
+      "learning_rate": 1.1333101253648003e-08,
+      "loss": 0.0144,
+      "step": 4940
+    },
+    {
+      "epoch": 352.92857142857144,
+      "grad_norm": 0.1533203125,
+      "learning_rate": 1.096462376427021e-08,
+      "loss": 0.0124,
+      "step": 4941
+    },
+    {
+      "epoch": 353.0,
+      "grad_norm": 0.126953125,
+      "learning_rate": 1.060223327901816e-08,
+      "loss": 0.0126,
+      "step": 4942
+    },
+    {
+      "epoch": 353.07142857142856,
+      "grad_norm": 0.359375,
+      "learning_rate": 1.0245929974513069e-08,
+      "loss": 0.0115,
+      "step": 4943
+    },
+    {
+      "epoch": 353.14285714285717,
+      "grad_norm": 0.1318359375,
+      "learning_rate": 9.895714024414637e-09,
+      "loss": 0.0126,
+      "step": 4944
+    },
+    {
+      "epoch": 353.2142857142857,
+      "grad_norm": 0.19140625,
+      "learning_rate": 9.551585599414103e-09,
+      "loss": 0.0118,
+      "step": 4945
+    },
+    {
+      "epoch": 353.2857142857143,
+      "grad_norm": 0.169921875,
+      "learning_rate": 9.213544867235635e-09,
+      "loss": 0.0121,
+      "step": 4946
+    },
+    {
+      "epoch": 353.35714285714283,
+      "grad_norm": 0.134765625,
+      "learning_rate": 8.881591992634941e-09,
+      "loss": 0.0133,
+      "step": 4947
+    },
+    {
+      "epoch": 353.42857142857144,
+      "grad_norm": 0.322265625,
+      "learning_rate": 8.555727137402048e-09,
+      "loss": 0.0148,
+      "step": 4948
+    },
+    {
+      "epoch": 353.5,
+      "grad_norm": 0.12451171875,
+      "learning_rate": 8.23595046035991e-09,
+      "loss": 0.0122,
+      "step": 4949
+    },
+    {
+      "epoch": 353.57142857142856,
+      "grad_norm": 0.130859375,
+      "learning_rate": 7.92226211736441e-09,
+      "loss": 0.0131,
+      "step": 4950
+    },
+    {
+      "epoch": 353.64285714285717,
+      "grad_norm": 0.1318359375,
+      "learning_rate": 7.614662261302974e-09,
+      "loss": 0.0132,
+      "step": 4951
+    },
+    {
+      "epoch": 353.7142857142857,
+      "grad_norm": 0.142578125,
+      "learning_rate": 7.313151042095956e-09,
+      "loss": 0.0136,
+      "step": 4952
+    },
+    {
+      "epoch": 353.7857142857143,
+      "grad_norm": 0.26171875,
+      "learning_rate": 7.017728606696638e-09,
+      "loss": 0.0146,
+      "step": 4953
+    },
+    {
+      "epoch": 353.85714285714283,
+      "grad_norm": 0.11181640625,
+      "learning_rate": 6.728395099089846e-09,
+      "loss": 0.0128,
+      "step": 4954
+    },
+    {
+      "epoch": 353.92857142857144,
+      "grad_norm": 0.1357421875,
+      "learning_rate": 6.445150660291943e-09,
+      "loss": 0.0142,
+      "step": 4955
+    },
+    {
+      "epoch": 354.0,
+      "grad_norm": 0.30078125,
+      "learning_rate": 6.167995428355e-09,
+      "loss": 0.0173,
+      "step": 4956
+    },
+    {
+      "epoch": 354.07142857142856,
+      "grad_norm": 0.275390625,
+      "learning_rate": 5.89692953836124e-09,
+      "loss": 0.0106,
+      "step": 4957
+    },
+    {
+      "epoch": 354.14285714285717,
+      "grad_norm": 0.1064453125,
+      "learning_rate": 5.6319531224230395e-09,
+      "loss": 0.0123,
+      "step": 4958
+    },
+    {
+      "epoch": 354.2142857142857,
+      "grad_norm": 0.15234375,
+      "learning_rate": 5.3730663096870915e-09,
+      "loss": 0.014,
+      "step": 4959
+    },
+    {
+      "epoch": 354.2857142857143,
+      "grad_norm": 0.1484375,
+      "learning_rate": 5.120269226333019e-09,
+      "loss": 0.0141,
+      "step": 4960
+    },
+    {
+      "epoch": 354.35714285714283,
+      "grad_norm": 0.14453125,
+      "learning_rate": 4.873561995570597e-09,
+      "loss": 0.0133,
+      "step": 4961
+    },
+    {
+      "epoch": 354.42857142857144,
+      "grad_norm": 0.2890625,
+      "learning_rate": 4.632944737641143e-09,
+      "loss": 0.014,
+      "step": 4962
+    },
+    {
+      "epoch": 354.5,
+      "grad_norm": 0.1572265625,
+      "learning_rate": 4.398417569818902e-09,
+      "loss": 0.0154,
+      "step": 4963
+    },
+    {
+      "epoch": 354.57142857142856,
+      "grad_norm": 0.1572265625,
+      "learning_rate": 4.169980606409663e-09,
+      "loss": 0.0135,
+      "step": 4964
+    },
+    {
+      "epoch": 354.64285714285717,
+      "grad_norm": 0.1455078125,
+      "learning_rate": 3.947633958750752e-09,
+      "loss": 0.0123,
+      "step": 4965
+    },
+    {
+      "epoch": 354.7142857142857,
+      "grad_norm": 0.150390625,
+      "learning_rate": 3.731377735211039e-09,
+      "loss": 0.0136,
+      "step": 4966
+    },
+    {
+      "epoch": 354.7857142857143,
+      "grad_norm": 0.1201171875,
+      "learning_rate": 3.5212120411909354e-09,
+      "loss": 0.0115,
+      "step": 4967
+    },
+    {
+      "epoch": 354.85714285714283,
+      "grad_norm": 0.1220703125,
+      "learning_rate": 3.3171369791237806e-09,
+      "loss": 0.0137,
+      "step": 4968
+    },
+    {
+      "epoch": 354.92857142857144,
+      "grad_norm": 0.1357421875,
+      "learning_rate": 3.119152648473067e-09,
+      "loss": 0.0127,
+      "step": 4969
+    },
+    {
+      "epoch": 355.0,
+      "grad_norm": 0.158203125,
+      "learning_rate": 2.9272591457338295e-09,
+      "loss": 0.0148,
+      "step": 4970
+    },
+    {
+      "epoch": 355.07142857142856,
+      "grad_norm": 0.1435546875,
+      "learning_rate": 2.741456564431255e-09,
+      "loss": 0.0121,
+      "step": 4971
+    },
+    {
+      "epoch": 355.14285714285717,
+      "grad_norm": 0.150390625,
+      "learning_rate": 2.5617449951248483e-09,
+      "loss": 0.0136,
+      "step": 4972
+    },
+    {
+      "epoch": 355.2142857142857,
+      "grad_norm": 0.1201171875,
+      "learning_rate": 2.3881245254028796e-09,
+      "loss": 0.0139,
+      "step": 4973
+    },
+    {
+      "epoch": 355.2857142857143,
+      "grad_norm": 0.1572265625,
+      "learning_rate": 2.2205952398879347e-09,
+      "loss": 0.0116,
+      "step": 4974
+    },
+    {
+      "epoch": 355.35714285714283,
+      "grad_norm": 0.1416015625,
+      "learning_rate": 2.0591572202285915e-09,
+      "loss": 0.0145,
+      "step": 4975
+    },
+    {
+      "epoch": 355.42857142857144,
+      "grad_norm": 0.11669921875,
+      "learning_rate": 1.9038105451091305e-09,
+      "loss": 0.0131,
+      "step": 4976
+    },
+    {
+      "epoch": 355.5,
+      "grad_norm": 0.2578125,
+      "learning_rate": 1.7545552902453743e-09,
+      "loss": 0.0148,
+      "step": 4977
+    },
+    {
+      "epoch": 355.57142857142856,
+      "grad_norm": 0.1318359375,
+      "learning_rate": 1.6113915283791358e-09,
+      "loss": 0.0117,
+      "step": 4978
+    },
+    {
+      "epoch": 355.64285714285717,
+      "grad_norm": 0.1640625,
+      "learning_rate": 1.4743193292893198e-09,
+      "loss": 0.0156,
+      "step": 4979
+    },
+    {
+      "epoch": 355.7142857142857,
+      "grad_norm": 0.263671875,
+      "learning_rate": 1.3433387597822088e-09,
+      "loss": 0.016,
+      "step": 4980
+    },
+    {
+      "epoch": 355.7857142857143,
+      "grad_norm": 0.16015625,
+      "learning_rate": 1.2184498836970148e-09,
+      "loss": 0.0117,
+      "step": 4981
+    },
+    {
+      "epoch": 355.85714285714283,
+      "grad_norm": 0.2158203125,
+      "learning_rate": 1.0996527619031028e-09,
+      "loss": 0.0122,
+      "step": 4982
+    },
+    {
+      "epoch": 355.92857142857144,
+      "grad_norm": 0.125,
+      "learning_rate": 9.869474522986034e-10,
+      "loss": 0.0118,
+      "step": 4983
+    },
+    {
+      "epoch": 356.0,
+      "grad_norm": 0.1201171875,
+      "learning_rate": 8.803340098159641e-10,
+      "loss": 0.0144,
+      "step": 4984
+    },
+    {
+      "epoch": 356.07142857142856,
+      "grad_norm": 0.1923828125,
+      "learning_rate": 7.798124864191736e-10,
+      "loss": 0.0111,
+      "step": 4985
+    },
+    {
+      "epoch": 356.14285714285717,
+      "grad_norm": 0.16015625,
+      "learning_rate": 6.853829310982108e-10,
+      "loss": 0.0141,
+      "step": 4986
+    },
+    {
+      "epoch": 356.2142857142857,
+      "grad_norm": 0.236328125,
+      "learning_rate": 5.970453898787586e-10,
+      "loss": 0.0142,
+      "step": 4987
+    },
+    {
+      "epoch": 356.2857142857143,
+      "grad_norm": 0.162109375,
+      "learning_rate": 5.147999058152664e-10,
+      "loss": 0.0127,
+      "step": 4988
+    },
+    {
+      "epoch": 356.35714285714283,
+      "grad_norm": 0.1513671875,
+      "learning_rate": 4.386465189923361e-10,
+      "loss": 0.0135,
+      "step": 4989
+    },
+    {
+      "epoch": 356.42857142857144,
+      "grad_norm": 0.1806640625,
+      "learning_rate": 3.685852665261114e-10,
+      "loss": 0.0144,
+      "step": 4990
+    },
+    {
+      "epoch": 356.5,
+      "grad_norm": 0.1650390625,
+      "learning_rate": 3.046161825656646e-10,
+      "loss": 0.0142,
+      "step": 4991
+    },
+    {
+      "epoch": 356.57142857142856,
+      "grad_norm": 0.142578125,
+      "learning_rate": 2.4673929828605835e-10,
+      "loss": 0.0125,
+      "step": 4992
+    },
+    {
+      "epoch": 356.64285714285717,
+      "grad_norm": 0.1474609375,
+      "learning_rate": 1.949546418966719e-10,
+      "loss": 0.0124,
+      "step": 4993
+    },
+    {
+      "epoch": 356.7142857142857,
+      "grad_norm": 0.12109375,
+      "learning_rate": 1.492622386384257e-10,
+      "loss": 0.0151,
+      "step": 4994
+    },
+    {
+      "epoch": 356.7857142857143,
+      "grad_norm": 0.1376953125,
+      "learning_rate": 1.0966211077961808e-10,
+      "loss": 0.0123,
+      "step": 4995
+    },
+    {
+      "epoch": 356.85714285714283,
+      "grad_norm": 0.13671875,
+      "learning_rate": 7.615427762008854e-11,
+      "loss": 0.0149,
+      "step": 4996
+    },
+    {
+      "epoch": 356.92857142857144,
+      "grad_norm": 0.12890625,
+      "learning_rate": 4.8738755492605535e-11,
+      "loss": 0.0141,
+      "step": 4997
+    },
+    {
+      "epoch": 357.0,
+      "grad_norm": 0.1328125,
+      "learning_rate": 2.741555775870319e-11,
+      "loss": 0.0145,
+      "step": 4998
+    },
+    {
+      "epoch": 357.07142857142856,
+      "grad_norm": 0.12060546875,
+      "learning_rate": 1.2184694811456787e-11,
+      "loss": 0.0127,
+      "step": 4999
+    },
+    {
+      "epoch": 357.14285714285717,
+      "grad_norm": 0.1552734375,
+      "learning_rate": 3.0461740740950204e-12,
+      "loss": 0.0114,
+      "step": 5000
+    }
+  ],
+  "logging_steps": 1,
+  "max_steps": 5000,
+  "num_input_tokens_seen": 0,
+  "num_train_epochs": 358,
+  "save_steps": 200,
+  "stateful_callbacks": {
+    "TrainerControl": {
+      "args": {
+        "should_epoch_stop": false,
+        "should_evaluate": false,
+        "should_log": false,
+        "should_save": false,
+        "should_training_stop": false
+      },
+      "attributes": {}
+    }
+  },
+  "total_flos": 0.0,
+  "train_batch_size": 64,
+  "trial_name": null,
+  "trial_params": null
+}