diff --git "a/trainer_state.json" "b/trainer_state.json"
new file mode 100644--- /dev/null
+++ "b/trainer_state.json"
@@ -0,0 +1,18628 @@
+{
+  "best_metric": 2.280937433242798,
+  "best_model_checkpoint": "./logs/checkpoint-1620",
+  "epoch": 5.47926267281106,
+  "global_step": 2970,
+  "is_hyper_param_search": false,
+  "is_local_process_zero": true,
+  "is_world_process_zero": true,
+  "log_history": [
+    {
+      "epoch": 0.0,
+      "learning_rate": 1.5996800000000002e-05,
+      "loss": 2.4657,
+      "step": 1
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 1.5993600000000002e-05,
+      "loss": 2.4023,
+      "step": 2
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 1.5990400000000002e-05,
+      "loss": 2.353,
+      "step": 3
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 1.5987200000000002e-05,
+      "loss": 2.3504,
+      "step": 4
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 1.5984e-05,
+      "loss": 2.4465,
+      "step": 5
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 1.5980800000000005e-05,
+      "loss": 2.2956,
+      "step": 6
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 1.5977600000000004e-05,
+      "loss": 2.3307,
+      "step": 7
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 1.59744e-05,
+      "loss": 2.3577,
+      "step": 8
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 1.5971200000000004e-05,
+      "loss": 2.3677,
+      "step": 9
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 1.5968000000000004e-05,
+      "loss": 2.3508,
+      "step": 10
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 1.5964800000000003e-05,
+      "loss": 2.3942,
+      "step": 11
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 1.5961600000000003e-05,
+      "loss": 2.3779,
+      "step": 12
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 1.5958400000000003e-05,
+      "loss": 2.3948,
+      "step": 13
+    },
+    {
+      "epoch": 0.03,
+      "learning_rate": 1.5955200000000003e-05,
+      "loss": 2.3949,
+      "step": 14
+    },
+    {
+      "epoch": 0.03,
+      "learning_rate": 1.5952000000000002e-05,
+      "loss": 2.3624,
+      "step": 15
+    },
+    {
+      "epoch": 0.03,
+      "learning_rate": 1.5948800000000002e-05,
+      "loss": 2.2899,
+      "step": 16
+    },
+    {
+      "epoch": 0.03,
+      "learning_rate": 1.5945600000000002e-05,
+      "loss": 2.3462,
+      "step": 17
+    },
+    {
+      "epoch": 0.03,
+      "learning_rate": 1.5942400000000002e-05,
+      "loss": 2.4089,
+      "step": 18
+    },
+    {
+      "epoch": 0.04,
+      "learning_rate": 1.59392e-05,
+      "loss": 2.3981,
+      "step": 19
+    },
+    {
+      "epoch": 0.04,
+      "learning_rate": 1.5936e-05,
+      "loss": 2.3784,
+      "step": 20
+    },
+    {
+      "epoch": 0.04,
+      "learning_rate": 1.5932800000000004e-05,
+      "loss": 2.3188,
+      "step": 21
+    },
+    {
+      "epoch": 0.04,
+      "learning_rate": 1.5929600000000004e-05,
+      "loss": 2.315,
+      "step": 22
+    },
+    {
+      "epoch": 0.04,
+      "learning_rate": 1.59264e-05,
+      "loss": 2.3314,
+      "step": 23
+    },
+    {
+      "epoch": 0.04,
+      "learning_rate": 1.5923200000000004e-05,
+      "loss": 2.2696,
+      "step": 24
+    },
+    {
+      "epoch": 0.05,
+      "learning_rate": 1.5920000000000003e-05,
+      "loss": 2.3486,
+      "step": 25
+    },
+    {
+      "epoch": 0.05,
+      "learning_rate": 1.5916800000000003e-05,
+      "loss": 2.3916,
+      "step": 26
+    },
+    {
+      "epoch": 0.05,
+      "learning_rate": 1.5913600000000003e-05,
+      "loss": 2.3193,
+      "step": 27
+    },
+    {
+      "epoch": 0.05,
+      "learning_rate": 1.5910400000000003e-05,
+      "loss": 2.3301,
+      "step": 28
+    },
+    {
+      "epoch": 0.05,
+      "learning_rate": 1.5907200000000002e-05,
+      "loss": 2.3993,
+      "step": 29
+    },
+    {
+      "epoch": 0.06,
+      "learning_rate": 1.5904000000000002e-05,
+      "loss": 2.2665,
+      "step": 30
+    },
+    {
+      "epoch": 0.06,
+      "eval_loss": 2.339500904083252,
+      "eval_runtime": 233.6969,
+      "eval_samples_per_second": 37.112,
+      "eval_steps_per_second": 0.582,
+      "step": 30
+    },
+    {
+      "epoch": 0.06,
+      "learning_rate": 1.5900800000000002e-05,
+      "loss": 2.2837,
+      "step": 31
+    },
+    {
+      "epoch": 0.06,
+      "learning_rate": 1.5897600000000002e-05,
+      "loss": 2.3697,
+      "step": 32
+    },
+    {
+      "epoch": 0.06,
+      "learning_rate": 1.58944e-05,
+      "loss": 2.3859,
+      "step": 33
+    },
+    {
+      "epoch": 0.06,
+      "learning_rate": 1.58912e-05,
+      "loss": 2.2994,
+      "step": 34
+    },
+    {
+      "epoch": 0.06,
+      "learning_rate": 1.5888e-05,
+      "loss": 2.3668,
+      "step": 35
+    },
+    {
+      "epoch": 0.07,
+      "learning_rate": 1.5884800000000004e-05,
+      "loss": 2.3253,
+      "step": 36
+    },
+    {
+      "epoch": 0.07,
+      "learning_rate": 1.5881600000000004e-05,
+      "loss": 2.4003,
+      "step": 37
+    },
+    {
+      "epoch": 0.07,
+      "learning_rate": 1.58784e-05,
+      "loss": 2.2534,
+      "step": 38
+    },
+    {
+      "epoch": 0.07,
+      "learning_rate": 1.5875200000000003e-05,
+      "loss": 2.3365,
+      "step": 39
+    },
+    {
+      "epoch": 0.07,
+      "learning_rate": 1.5872000000000003e-05,
+      "loss": 2.289,
+      "step": 40
+    },
+    {
+      "epoch": 0.08,
+      "learning_rate": 1.5868800000000003e-05,
+      "loss": 2.3769,
+      "step": 41
+    },
+    {
+      "epoch": 0.08,
+      "learning_rate": 1.5865600000000003e-05,
+      "loss": 2.2561,
+      "step": 42
+    },
+    {
+      "epoch": 0.08,
+      "learning_rate": 1.5862400000000002e-05,
+      "loss": 2.3124,
+      "step": 43
+    },
+    {
+      "epoch": 0.08,
+      "learning_rate": 1.5859200000000002e-05,
+      "loss": 2.257,
+      "step": 44
+    },
+    {
+      "epoch": 0.08,
+      "learning_rate": 1.5856000000000002e-05,
+      "loss": 2.3677,
+      "step": 45
+    },
+    {
+      "epoch": 0.08,
+      "learning_rate": 1.5852800000000002e-05,
+      "loss": 2.2666,
+      "step": 46
+    },
+    {
+      "epoch": 0.09,
+      "learning_rate": 1.58496e-05,
+      "loss": 2.3594,
+      "step": 47
+    },
+    {
+      "epoch": 0.09,
+      "learning_rate": 1.58464e-05,
+      "loss": 2.3171,
+      "step": 48
+    },
+    {
+      "epoch": 0.09,
+      "learning_rate": 1.58432e-05,
+      "loss": 2.2857,
+      "step": 49
+    },
+    {
+      "epoch": 0.09,
+      "learning_rate": 1.584e-05,
+      "loss": 2.3583,
+      "step": 50
+    },
+    {
+      "epoch": 0.09,
+      "learning_rate": 1.5836800000000004e-05,
+      "loss": 2.3365,
+      "step": 51
+    },
+    {
+      "epoch": 0.1,
+      "learning_rate": 1.5833600000000004e-05,
+      "loss": 2.3166,
+      "step": 52
+    },
+    {
+      "epoch": 0.1,
+      "learning_rate": 1.58304e-05,
+      "loss": 2.3404,
+      "step": 53
+    },
+    {
+      "epoch": 0.1,
+      "learning_rate": 1.5827200000000003e-05,
+      "loss": 2.463,
+      "step": 54
+    },
+    {
+      "epoch": 0.1,
+      "learning_rate": 1.5824000000000003e-05,
+      "loss": 2.3386,
+      "step": 55
+    },
+    {
+      "epoch": 0.1,
+      "learning_rate": 1.5820800000000003e-05,
+      "loss": 2.3495,
+      "step": 56
+    },
+    {
+      "epoch": 0.11,
+      "learning_rate": 1.5817600000000003e-05,
+      "loss": 2.2839,
+      "step": 57
+    },
+    {
+      "epoch": 0.11,
+      "learning_rate": 1.5814400000000002e-05,
+      "loss": 2.3823,
+      "step": 58
+    },
+    {
+      "epoch": 0.11,
+      "learning_rate": 1.5811200000000002e-05,
+      "loss": 2.3593,
+      "step": 59
+    },
+    {
+      "epoch": 0.11,
+      "learning_rate": 1.5808000000000002e-05,
+      "loss": 2.3533,
+      "step": 60
+    },
+    {
+      "epoch": 0.11,
+      "eval_loss": 2.332291841506958,
+      "eval_runtime": 234.0714,
+      "eval_samples_per_second": 37.053,
+      "eval_steps_per_second": 0.581,
+      "step": 60
+    },
+    {
+      "epoch": 0.11,
+      "learning_rate": 1.58048e-05,
+      "loss": 2.3827,
+      "step": 61
+    },
+    {
+      "epoch": 0.11,
+      "learning_rate": 1.5801600000000005e-05,
+      "loss": 2.2793,
+      "step": 62
+    },
+    {
+      "epoch": 0.12,
+      "learning_rate": 1.5798400000000004e-05,
+      "loss": 2.3328,
+      "step": 63
+    },
+    {
+      "epoch": 0.12,
+      "learning_rate": 1.57952e-05,
+      "loss": 2.3716,
+      "step": 64
+    },
+    {
+      "epoch": 0.12,
+      "learning_rate": 1.5792000000000004e-05,
+      "loss": 2.3238,
+      "step": 65
+    },
+    {
+      "epoch": 0.12,
+      "learning_rate": 1.5788800000000004e-05,
+      "loss": 2.3709,
+      "step": 66
+    },
+    {
+      "epoch": 0.12,
+      "learning_rate": 1.5785600000000003e-05,
+      "loss": 2.362,
+      "step": 67
+    },
+    {
+      "epoch": 0.13,
+      "learning_rate": 1.5782400000000003e-05,
+      "loss": 2.3093,
+      "step": 68
+    },
+    {
+      "epoch": 0.13,
+      "learning_rate": 1.5779200000000003e-05,
+      "loss": 2.2793,
+      "step": 69
+    },
+    {
+      "epoch": 0.13,
+      "learning_rate": 1.5776000000000003e-05,
+      "loss": 2.4214,
+      "step": 70
+    },
+    {
+      "epoch": 0.13,
+      "learning_rate": 1.5772800000000003e-05,
+      "loss": 2.3996,
+      "step": 71
+    },
+    {
+      "epoch": 0.13,
+      "learning_rate": 1.5769600000000002e-05,
+      "loss": 2.3376,
+      "step": 72
+    },
+    {
+      "epoch": 0.13,
+      "learning_rate": 1.5766400000000002e-05,
+      "loss": 2.3733,
+      "step": 73
+    },
+    {
+      "epoch": 0.14,
+      "learning_rate": 1.5763200000000002e-05,
+      "loss": 2.3671,
+      "step": 74
+    },
+    {
+      "epoch": 0.14,
+      "learning_rate": 1.576e-05,
+      "loss": 2.3607,
+      "step": 75
+    },
+    {
+      "epoch": 0.14,
+      "learning_rate": 1.57568e-05,
+      "loss": 2.402,
+      "step": 76
+    },
+    {
+      "epoch": 0.14,
+      "learning_rate": 1.5753600000000004e-05,
+      "loss": 2.2811,
+      "step": 77
+    },
+    {
+      "epoch": 0.14,
+      "learning_rate": 1.5750400000000004e-05,
+      "loss": 2.3009,
+      "step": 78
+    },
+    {
+      "epoch": 0.15,
+      "learning_rate": 1.57472e-05,
+      "loss": 2.2561,
+      "step": 79
+    },
+    {
+      "epoch": 0.15,
+      "learning_rate": 1.5744000000000004e-05,
+      "loss": 2.2153,
+      "step": 80
+    },
+    {
+      "epoch": 0.15,
+      "learning_rate": 1.5740800000000004e-05,
+      "loss": 2.3283,
+      "step": 81
+    },
+    {
+      "epoch": 0.15,
+      "learning_rate": 1.5737600000000003e-05,
+      "loss": 2.3149,
+      "step": 82
+    },
+    {
+      "epoch": 0.15,
+      "learning_rate": 1.5734400000000003e-05,
+      "loss": 2.2491,
+      "step": 83
+    },
+    {
+      "epoch": 0.15,
+      "learning_rate": 1.5731200000000003e-05,
+      "loss": 2.2849,
+      "step": 84
+    },
+    {
+      "epoch": 0.16,
+      "learning_rate": 1.5728000000000003e-05,
+      "loss": 2.4359,
+      "step": 85
+    },
+    {
+      "epoch": 0.16,
+      "learning_rate": 1.5724800000000002e-05,
+      "loss": 2.4064,
+      "step": 86
+    },
+    {
+      "epoch": 0.16,
+      "learning_rate": 1.5721600000000002e-05,
+      "loss": 2.3433,
+      "step": 87
+    },
+    {
+      "epoch": 0.16,
+      "learning_rate": 1.5718400000000002e-05,
+      "loss": 2.3072,
+      "step": 88
+    },
+    {
+      "epoch": 0.16,
+      "learning_rate": 1.57152e-05,
+      "loss": 2.3837,
+      "step": 89
+    },
+    {
+      "epoch": 0.17,
+      "learning_rate": 1.5712e-05,
+      "loss": 2.3325,
+      "step": 90
+    },
+    {
+      "epoch": 0.17,
+      "eval_loss": 2.3291423320770264,
+      "eval_runtime": 233.8538,
+      "eval_samples_per_second": 37.087,
+      "eval_steps_per_second": 0.582,
+      "step": 90
+    },
+    {
+      "epoch": 0.17,
+      "learning_rate": 1.57088e-05,
+      "loss": 2.371,
+      "step": 91
+    },
+    {
+      "epoch": 0.17,
+      "learning_rate": 1.5705600000000004e-05,
+      "loss": 2.341,
+      "step": 92
+    },
+    {
+      "epoch": 0.17,
+      "learning_rate": 1.5702400000000004e-05,
+      "loss": 2.2919,
+      "step": 93
+    },
+    {
+      "epoch": 0.17,
+      "learning_rate": 1.56992e-05,
+      "loss": 2.2594,
+      "step": 94
+    },
+    {
+      "epoch": 0.18,
+      "learning_rate": 1.5696000000000004e-05,
+      "loss": 2.3543,
+      "step": 95
+    },
+    {
+      "epoch": 0.18,
+      "learning_rate": 1.5692800000000003e-05,
+      "loss": 2.3065,
+      "step": 96
+    },
+    {
+      "epoch": 0.18,
+      "learning_rate": 1.5689600000000003e-05,
+      "loss": 2.2748,
+      "step": 97
+    },
+    {
+      "epoch": 0.18,
+      "learning_rate": 1.5686400000000003e-05,
+      "loss": 2.2879,
+      "step": 98
+    },
+    {
+      "epoch": 0.18,
+      "learning_rate": 1.5683200000000003e-05,
+      "loss": 2.3583,
+      "step": 99
+    },
+    {
+      "epoch": 0.18,
+      "learning_rate": 1.5680000000000002e-05,
+      "loss": 2.3628,
+      "step": 100
+    },
+    {
+      "epoch": 0.19,
+      "learning_rate": 1.5676800000000002e-05,
+      "loss": 2.3217,
+      "step": 101
+    },
+    {
+      "epoch": 0.19,
+      "learning_rate": 1.5673600000000002e-05,
+      "loss": 2.3143,
+      "step": 102
+    },
+    {
+      "epoch": 0.19,
+      "learning_rate": 1.5670400000000005e-05,
+      "loss": 2.3035,
+      "step": 103
+    },
+    {
+      "epoch": 0.19,
+      "learning_rate": 1.56672e-05,
+      "loss": 2.3254,
+      "step": 104
+    },
+    {
+      "epoch": 0.19,
+      "learning_rate": 1.5664e-05,
+      "loss": 2.4054,
+      "step": 105
+    },
+    {
+      "epoch": 0.2,
+      "learning_rate": 1.5660800000000004e-05,
+      "loss": 2.2705,
+      "step": 106
+    },
+    {
+      "epoch": 0.2,
+      "learning_rate": 1.5657600000000004e-05,
+      "loss": 2.3153,
+      "step": 107
+    },
+    {
+      "epoch": 0.2,
+      "learning_rate": 1.5654400000000004e-05,
+      "loss": 2.3928,
+      "step": 108
+    },
+    {
+      "epoch": 0.2,
+      "learning_rate": 1.5651200000000004e-05,
+      "loss": 2.369,
+      "step": 109
+    },
+    {
+      "epoch": 0.2,
+      "learning_rate": 1.5648000000000003e-05,
+      "loss": 2.3748,
+      "step": 110
+    },
+    {
+      "epoch": 0.2,
+      "learning_rate": 1.5644800000000003e-05,
+      "loss": 2.3598,
+      "step": 111
+    },
+    {
+      "epoch": 0.21,
+      "learning_rate": 1.5641600000000003e-05,
+      "loss": 2.3051,
+      "step": 112
+    },
+    {
+      "epoch": 0.21,
+      "learning_rate": 1.5638400000000003e-05,
+      "loss": 2.4088,
+      "step": 113
+    },
+    {
+      "epoch": 0.21,
+      "learning_rate": 1.5635200000000002e-05,
+      "loss": 2.3883,
+      "step": 114
+    },
+    {
+      "epoch": 0.21,
+      "learning_rate": 1.5632000000000002e-05,
+      "loss": 2.3914,
+      "step": 115
+    },
+    {
+      "epoch": 0.21,
+      "learning_rate": 1.5628800000000002e-05,
+      "loss": 2.3157,
+      "step": 116
+    },
+    {
+      "epoch": 0.22,
+      "learning_rate": 1.56256e-05,
+      "loss": 2.3533,
+      "step": 117
+    },
+    {
+      "epoch": 0.22,
+      "learning_rate": 1.5622400000000005e-05,
+      "loss": 2.3582,
+      "step": 118
+    },
+    {
+      "epoch": 0.22,
+      "learning_rate": 1.56192e-05,
+      "loss": 2.3096,
+      "step": 119
+    },
+    {
+      "epoch": 0.22,
+      "learning_rate": 1.5616e-05,
+      "loss": 2.3247,
+      "step": 120
+    },
+    {
+      "epoch": 0.22,
+      "eval_loss": 2.3282039165496826,
+      "eval_runtime": 231.4498,
+      "eval_samples_per_second": 37.472,
+      "eval_steps_per_second": 0.588,
+      "step": 120
+    },
+    {
+      "epoch": 0.22,
+      "learning_rate": 1.5612800000000004e-05,
+      "loss": 2.3074,
+      "step": 121
+    },
+    {
+      "epoch": 0.22,
+      "learning_rate": 1.5609600000000004e-05,
+      "loss": 2.3591,
+      "step": 122
+    },
+    {
+      "epoch": 0.23,
+      "learning_rate": 1.5606400000000004e-05,
+      "loss": 2.3771,
+      "step": 123
+    },
+    {
+      "epoch": 0.23,
+      "learning_rate": 1.5603200000000003e-05,
+      "loss": 2.2886,
+      "step": 124
+    },
+    {
+      "epoch": 0.23,
+      "learning_rate": 1.5600000000000003e-05,
+      "loss": 2.3164,
+      "step": 125
+    },
+    {
+      "epoch": 0.23,
+      "learning_rate": 1.5596800000000003e-05,
+      "loss": 2.33,
+      "step": 126
+    },
+    {
+      "epoch": 0.23,
+      "learning_rate": 1.5593600000000003e-05,
+      "loss": 2.3227,
+      "step": 127
+    },
+    {
+      "epoch": 0.24,
+      "learning_rate": 1.5590400000000002e-05,
+      "loss": 2.3051,
+      "step": 128
+    },
+    {
+      "epoch": 0.24,
+      "learning_rate": 1.5587200000000002e-05,
+      "loss": 2.4185,
+      "step": 129
+    },
+    {
+      "epoch": 0.24,
+      "learning_rate": 1.5584000000000002e-05,
+      "loss": 2.3094,
+      "step": 130
+    },
+    {
+      "epoch": 0.24,
+      "learning_rate": 1.55808e-05,
+      "loss": 2.3212,
+      "step": 131
+    },
+    {
+      "epoch": 0.24,
+      "learning_rate": 1.55776e-05,
+      "loss": 2.2747,
+      "step": 132
+    },
+    {
+      "epoch": 0.25,
+      "learning_rate": 1.5574400000000005e-05,
+      "loss": 2.4178,
+      "step": 133
+    },
+    {
+      "epoch": 0.25,
+      "learning_rate": 1.55712e-05,
+      "loss": 2.3459,
+      "step": 134
+    },
+    {
+      "epoch": 0.25,
+      "learning_rate": 1.5568e-05,
+      "loss": 2.2963,
+      "step": 135
+    },
+    {
+      "epoch": 0.25,
+      "learning_rate": 1.5564800000000004e-05,
+      "loss": 2.3195,
+      "step": 136
+    },
+    {
+      "epoch": 0.25,
+      "learning_rate": 1.5561600000000004e-05,
+      "loss": 2.3169,
+      "step": 137
+    },
+    {
+      "epoch": 0.25,
+      "learning_rate": 1.5558400000000003e-05,
+      "loss": 2.1906,
+      "step": 138
+    },
+    {
+      "epoch": 0.26,
+      "learning_rate": 1.5555200000000003e-05,
+      "loss": 2.3047,
+      "step": 139
+    },
+    {
+      "epoch": 0.26,
+      "learning_rate": 1.5552000000000003e-05,
+      "loss": 2.2747,
+      "step": 140
+    },
+    {
+      "epoch": 0.26,
+      "learning_rate": 1.5548800000000003e-05,
+      "loss": 2.2692,
+      "step": 141
+    },
+    {
+      "epoch": 0.26,
+      "learning_rate": 1.5545600000000002e-05,
+      "loss": 2.2777,
+      "step": 142
+    },
+    {
+      "epoch": 0.26,
+      "learning_rate": 1.5542400000000002e-05,
+      "loss": 2.3925,
+      "step": 143
+    },
+    {
+      "epoch": 0.27,
+      "learning_rate": 1.5539200000000002e-05,
+      "loss": 2.3083,
+      "step": 144
+    },
+    {
+      "epoch": 0.27,
+      "learning_rate": 1.5536e-05,
+      "loss": 2.2786,
+      "step": 145
+    },
+    {
+      "epoch": 0.27,
+      "learning_rate": 1.55328e-05,
+      "loss": 2.3158,
+      "step": 146
+    },
+    {
+      "epoch": 0.27,
+      "learning_rate": 1.55296e-05,
+      "loss": 2.3759,
+      "step": 147
+    },
+    {
+      "epoch": 0.27,
+      "learning_rate": 1.5526400000000004e-05,
+      "loss": 2.2707,
+      "step": 148
+    },
+    {
+      "epoch": 0.27,
+      "learning_rate": 1.55232e-05,
+      "loss": 2.3843,
+      "step": 149
+    },
+    {
+      "epoch": 0.28,
+      "learning_rate": 1.552e-05,
+      "loss": 2.3289,
+      "step": 150
+    },
+    {
+      "epoch": 0.28,
+      "eval_loss": 2.323575496673584,
+      "eval_runtime": 233.8046,
+      "eval_samples_per_second": 37.095,
+      "eval_steps_per_second": 0.582,
+      "step": 150
+    },
+    {
+      "epoch": 0.28,
+      "learning_rate": 1.5516800000000004e-05,
+      "loss": 2.4541,
+      "step": 151
+    },
+    {
+      "epoch": 0.28,
+      "learning_rate": 1.5513600000000003e-05,
+      "loss": 2.216,
+      "step": 152
+    },
+    {
+      "epoch": 0.28,
+      "learning_rate": 1.5510400000000003e-05,
+      "loss": 2.2142,
+      "step": 153
+    },
+    {
+      "epoch": 0.28,
+      "learning_rate": 1.5507200000000003e-05,
+      "loss": 2.3681,
+      "step": 154
+    },
+    {
+      "epoch": 0.29,
+      "learning_rate": 1.5504000000000003e-05,
+      "loss": 2.3715,
+      "step": 155
+    },
+    {
+      "epoch": 0.29,
+      "learning_rate": 1.5500800000000002e-05,
+      "loss": 2.3233,
+      "step": 156
+    },
+    {
+      "epoch": 0.29,
+      "learning_rate": 1.5497600000000002e-05,
+      "loss": 2.3549,
+      "step": 157
+    },
+    {
+      "epoch": 0.29,
+      "learning_rate": 1.5494400000000002e-05,
+      "loss": 2.4069,
+      "step": 158
+    },
+    {
+      "epoch": 0.29,
+      "learning_rate": 1.5491200000000002e-05,
+      "loss": 2.3082,
+      "step": 159
+    },
+    {
+      "epoch": 0.29,
+      "learning_rate": 1.5488e-05,
+      "loss": 2.3604,
+      "step": 160
+    },
+    {
+      "epoch": 0.3,
+      "learning_rate": 1.54848e-05,
+      "loss": 2.2714,
+      "step": 161
+    },
+    {
+      "epoch": 0.3,
+      "learning_rate": 1.5481600000000004e-05,
+      "loss": 2.2962,
+      "step": 162
+    },
+    {
+      "epoch": 0.3,
+      "learning_rate": 1.5478400000000004e-05,
+      "loss": 2.3027,
+      "step": 163
+    },
+    {
+      "epoch": 0.3,
+      "learning_rate": 1.54752e-05,
+      "loss": 2.3564,
+      "step": 164
+    },
+    {
+      "epoch": 0.3,
+      "learning_rate": 1.5472000000000004e-05,
+      "loss": 2.3075,
+      "step": 165
+    },
+    {
+      "epoch": 0.31,
+      "learning_rate": 1.5468800000000003e-05,
+      "loss": 2.3335,
+      "step": 166
+    },
+    {
+      "epoch": 0.31,
+      "learning_rate": 1.5465600000000003e-05,
+      "loss": 2.2894,
+      "step": 167
+    },
+    {
+      "epoch": 0.31,
+      "learning_rate": 1.5462400000000003e-05,
+      "loss": 2.2566,
+      "step": 168
+    },
+    {
+      "epoch": 0.31,
+      "learning_rate": 1.5459200000000003e-05,
+      "loss": 2.2905,
+      "step": 169
+    },
+    {
+      "epoch": 0.31,
+      "learning_rate": 1.5456000000000002e-05,
+      "loss": 2.2901,
+      "step": 170
+    },
+    {
+      "epoch": 0.32,
+      "learning_rate": 1.5452800000000002e-05,
+      "loss": 2.4244,
+      "step": 171
+    },
+    {
+      "epoch": 0.32,
+      "learning_rate": 1.5449600000000002e-05,
+      "loss": 2.2878,
+      "step": 172
+    },
+    {
+      "epoch": 0.32,
+      "learning_rate": 1.5446400000000002e-05,
+      "loss": 2.283,
+      "step": 173
+    },
+    {
+      "epoch": 0.32,
+      "learning_rate": 1.54432e-05,
+      "loss": 2.309,
+      "step": 174
+    },
+    {
+      "epoch": 0.32,
+      "learning_rate": 1.544e-05,
+      "loss": 2.3126,
+      "step": 175
+    },
+    {
+      "epoch": 0.32,
+      "learning_rate": 1.54368e-05,
+      "loss": 2.3671,
+      "step": 176
+    },
+    {
+      "epoch": 0.33,
+      "learning_rate": 1.5433600000000004e-05,
+      "loss": 2.3665,
+      "step": 177
+    },
+    {
+      "epoch": 0.33,
+      "learning_rate": 1.5430400000000004e-05,
+      "loss": 2.3252,
+      "step": 178
+    },
+    {
+      "epoch": 0.33,
+      "learning_rate": 1.54272e-05,
+      "loss": 2.3467,
+      "step": 179
+    },
+    {
+      "epoch": 0.33,
+      "learning_rate": 1.5424000000000003e-05,
+      "loss": 2.294,
+      "step": 180
+    },
+    {
+      "epoch": 0.33,
+      "eval_loss": 2.3187031745910645,
+      "eval_runtime": 232.1635,
+      "eval_samples_per_second": 37.357,
+      "eval_steps_per_second": 0.586,
+      "step": 180
+    },
+    {
+      "epoch": 0.33,
+      "learning_rate": 1.5420800000000003e-05,
+      "loss": 2.2453,
+      "step": 181
+    },
+    {
+      "epoch": 0.34,
+      "learning_rate": 1.5417600000000003e-05,
+      "loss": 2.3094,
+      "step": 182
+    },
+    {
+      "epoch": 0.34,
+      "learning_rate": 1.5414400000000003e-05,
+      "loss": 2.2994,
+      "step": 183
+    },
+    {
+      "epoch": 0.34,
+      "learning_rate": 1.5411200000000002e-05,
+      "loss": 2.3401,
+      "step": 184
+    },
+    {
+      "epoch": 0.34,
+      "learning_rate": 1.5408000000000002e-05,
+      "loss": 2.3376,
+      "step": 185
+    },
+    {
+      "epoch": 0.34,
+      "learning_rate": 1.5404800000000002e-05,
+      "loss": 2.3302,
+      "step": 186
+    },
+    {
+      "epoch": 0.34,
+      "learning_rate": 1.5401600000000002e-05,
+      "loss": 2.293,
+      "step": 187
+    },
+    {
+      "epoch": 0.35,
+      "learning_rate": 1.53984e-05,
+      "loss": 2.2901,
+      "step": 188
+    },
+    {
+      "epoch": 0.35,
+      "learning_rate": 1.5395200000000005e-05,
+      "loss": 2.3544,
+      "step": 189
+    },
+    {
+      "epoch": 0.35,
+      "learning_rate": 1.5392e-05,
+      "loss": 2.3329,
+      "step": 190
+    },
+    {
+      "epoch": 0.35,
+      "learning_rate": 1.53888e-05,
+      "loss": 2.2087,
+      "step": 191
+    },
+    {
+      "epoch": 0.35,
+      "learning_rate": 1.5385600000000004e-05,
+      "loss": 2.2379,
+      "step": 192
+    },
+    {
+      "epoch": 0.36,
+      "learning_rate": 1.5382400000000004e-05,
+      "loss": 2.2551,
+      "step": 193
+    },
+    {
+      "epoch": 0.36,
+      "learning_rate": 1.5379200000000003e-05,
+      "loss": 2.3064,
+      "step": 194
+    },
+    {
+      "epoch": 0.36,
+      "learning_rate": 1.5376000000000003e-05,
+      "loss": 2.38,
+      "step": 195
+    },
+    {
+      "epoch": 0.36,
+      "learning_rate": 1.5372800000000003e-05,
+      "loss": 2.264,
+      "step": 196
+    },
+    {
+      "epoch": 0.36,
+      "learning_rate": 1.5369600000000003e-05,
+      "loss": 2.3506,
+      "step": 197
+    },
+    {
+      "epoch": 0.36,
+      "learning_rate": 1.5366400000000002e-05,
+      "loss": 2.2571,
+      "step": 198
+    },
+    {
+      "epoch": 0.37,
+      "learning_rate": 1.5363200000000002e-05,
+      "loss": 2.2957,
+      "step": 199
+    },
+    {
+      "epoch": 0.37,
+      "learning_rate": 1.5360000000000002e-05,
+      "loss": 2.4273,
+      "step": 200
+    },
+    {
+      "epoch": 0.37,
+      "learning_rate": 1.5356800000000002e-05,
+      "loss": 2.3405,
+      "step": 201
+    },
+    {
+      "epoch": 0.37,
+      "learning_rate": 1.53536e-05,
+      "loss": 2.2805,
+      "step": 202
+    },
+    {
+      "epoch": 0.37,
+      "learning_rate": 1.5350400000000005e-05,
+      "loss": 2.3121,
+      "step": 203
+    },
+    {
+      "epoch": 0.38,
+      "learning_rate": 1.5347200000000004e-05,
+      "loss": 2.3283,
+      "step": 204
+    },
+    {
+      "epoch": 0.38,
+      "learning_rate": 1.5344e-05,
+      "loss": 2.3331,
+      "step": 205
+    },
+    {
+      "epoch": 0.38,
+      "learning_rate": 1.5340800000000004e-05,
+      "loss": 2.3161,
+      "step": 206
+    },
+    {
+      "epoch": 0.38,
+      "learning_rate": 1.5337600000000004e-05,
+      "loss": 2.2898,
+      "step": 207
+    },
+    {
+      "epoch": 0.38,
+      "learning_rate": 1.5334400000000003e-05,
+      "loss": 2.3712,
+      "step": 208
+    },
+    {
+      "epoch": 0.39,
+      "learning_rate": 1.5331200000000003e-05,
+      "loss": 2.3312,
+      "step": 209
+    },
+    {
+      "epoch": 0.39,
+      "learning_rate": 1.5328000000000003e-05,
+      "loss": 2.3217,
+      "step": 210
+    },
+    {
+      "epoch": 0.39,
+      "eval_loss": 2.316190004348755,
+      "eval_runtime": 232.5338,
+      "eval_samples_per_second": 37.298,
+      "eval_steps_per_second": 0.585,
+      "step": 210
+    },
+    {
+      "epoch": 0.39,
+      "learning_rate": 1.5324800000000003e-05,
+      "loss": 2.3321,
+      "step": 211
+    },
+    {
+      "epoch": 0.39,
+      "learning_rate": 1.5321600000000003e-05,
+      "loss": 2.3265,
+      "step": 212
+    },
+    {
+      "epoch": 0.39,
+      "learning_rate": 1.5318400000000002e-05,
+      "loss": 2.3551,
+      "step": 213
+    },
+    {
+      "epoch": 0.39,
+      "learning_rate": 1.5315200000000002e-05,
+      "loss": 2.4064,
+      "step": 214
+    },
+    {
+      "epoch": 0.4,
+      "learning_rate": 1.5312000000000002e-05,
+      "loss": 2.3714,
+      "step": 215
+    },
+    {
+      "epoch": 0.4,
+      "learning_rate": 1.53088e-05,
+      "loss": 2.2595,
+      "step": 216
+    },
+    {
+      "epoch": 0.4,
+      "learning_rate": 1.53056e-05,
+      "loss": 2.3556,
+      "step": 217
+    },
+    {
+      "epoch": 0.4,
+      "learning_rate": 1.5302400000000004e-05,
+      "loss": 2.2615,
+      "step": 218
+    },
+    {
+      "epoch": 0.4,
+      "learning_rate": 1.5299200000000004e-05,
+      "loss": 2.2678,
+      "step": 219
+    },
+    {
+      "epoch": 0.41,
+      "learning_rate": 1.5296e-05,
+      "loss": 2.3762,
+      "step": 220
+    },
+    {
+      "epoch": 0.41,
+      "learning_rate": 1.5292800000000004e-05,
+      "loss": 2.3305,
+      "step": 221
+    },
+    {
+      "epoch": 0.41,
+      "learning_rate": 1.5289600000000004e-05,
+      "loss": 2.3241,
+      "step": 222
+    },
+    {
+      "epoch": 0.41,
+      "learning_rate": 1.5286400000000003e-05,
+      "loss": 2.3708,
+      "step": 223
+    },
+    {
+      "epoch": 0.41,
+      "learning_rate": 1.5283200000000003e-05,
+      "loss": 2.3161,
+      "step": 224
+    },
+    {
+      "epoch": 0.41,
+      "learning_rate": 1.5280000000000003e-05,
+      "loss": 2.3123,
+      "step": 225
+    },
+    {
+      "epoch": 0.42,
+      "learning_rate": 1.5276800000000003e-05,
+      "loss": 2.3579,
+      "step": 226
+    },
+    {
+      "epoch": 0.42,
+      "learning_rate": 1.5273600000000002e-05,
+      "loss": 2.3288,
+      "step": 227
+    },
+    {
+      "epoch": 0.42,
+      "learning_rate": 1.5270400000000002e-05,
+      "loss": 2.344,
+      "step": 228
+    },
+    {
+      "epoch": 0.42,
+      "learning_rate": 1.5267200000000002e-05,
+      "loss": 2.3227,
+      "step": 229
+    },
+    {
+      "epoch": 0.42,
+      "learning_rate": 1.5264e-05,
+      "loss": 2.4174,
+      "step": 230
+    },
+    {
+      "epoch": 0.43,
+      "learning_rate": 1.52608e-05,
+      "loss": 2.2473,
+      "step": 231
+    },
+    {
+      "epoch": 0.43,
+      "learning_rate": 1.5257600000000003e-05,
+      "loss": 2.3326,
+      "step": 232
+    },
+    {
+      "epoch": 0.43,
+      "learning_rate": 1.5254400000000003e-05,
+      "loss": 2.3097,
+      "step": 233
+    },
+    {
+      "epoch": 0.43,
+      "learning_rate": 1.5251200000000004e-05,
+      "loss": 2.326,
+      "step": 234
+    },
+    {
+      "epoch": 0.43,
+      "learning_rate": 1.5248000000000002e-05,
+      "loss": 2.2891,
+      "step": 235
+    },
+    {
+      "epoch": 0.44,
+      "learning_rate": 1.5244800000000002e-05,
+      "loss": 2.361,
+      "step": 236
+    },
+    {
+      "epoch": 0.44,
+      "learning_rate": 1.5241600000000003e-05,
+      "loss": 2.2936,
+      "step": 237
+    },
+    {
+      "epoch": 0.44,
+      "learning_rate": 1.5238400000000003e-05,
+      "loss": 2.272,
+      "step": 238
+    },
+    {
+      "epoch": 0.44,
+      "learning_rate": 1.5235200000000003e-05,
+      "loss": 2.3161,
+      "step": 239
+    },
+    {
+      "epoch": 0.44,
+      "learning_rate": 1.5232000000000003e-05,
+      "loss": 2.2418,
+      "step": 240
+    },
+    {
+      "epoch": 0.44,
+      "eval_loss": 2.313016653060913,
+      "eval_runtime": 232.7214,
+      "eval_samples_per_second": 37.268,
+      "eval_steps_per_second": 0.584,
+      "step": 240
+    },
+    {
+      "epoch": 0.44,
+      "learning_rate": 1.5228800000000002e-05,
+      "loss": 2.3578,
+      "step": 241
+    },
+    {
+      "epoch": 0.45,
+      "learning_rate": 1.5225600000000002e-05,
+      "loss": 2.3686,
+      "step": 242
+    },
+    {
+      "epoch": 0.45,
+      "learning_rate": 1.5222400000000004e-05,
+      "loss": 2.2552,
+      "step": 243
+    },
+    {
+      "epoch": 0.45,
+      "learning_rate": 1.5219200000000003e-05,
+      "loss": 2.3495,
+      "step": 244
+    },
+    {
+      "epoch": 0.45,
+      "learning_rate": 1.5216000000000001e-05,
+      "loss": 2.2742,
+      "step": 245
+    },
+    {
+      "epoch": 0.45,
+      "learning_rate": 1.5212800000000003e-05,
+      "loss": 2.314,
+      "step": 246
+    },
+    {
+      "epoch": 0.46,
+      "learning_rate": 1.5209600000000003e-05,
+      "loss": 2.3639,
+      "step": 247
+    },
+    {
+      "epoch": 0.46,
+      "learning_rate": 1.5206400000000002e-05,
+      "loss": 2.4206,
+      "step": 248
+    },
+    {
+      "epoch": 0.46,
+      "learning_rate": 1.5203200000000004e-05,
+      "loss": 2.2844,
+      "step": 249
+    },
+    {
+      "epoch": 0.46,
+      "learning_rate": 1.5200000000000002e-05,
+      "loss": 2.2733,
+      "step": 250
+    },
+    {
+      "epoch": 0.46,
+      "learning_rate": 1.5196800000000002e-05,
+      "loss": 2.2922,
+      "step": 251
+    },
+    {
+      "epoch": 0.46,
+      "learning_rate": 1.5193600000000003e-05,
+      "loss": 2.3206,
+      "step": 252
+    },
+    {
+      "epoch": 0.47,
+      "learning_rate": 1.5190400000000003e-05,
+      "loss": 2.3254,
+      "step": 253
+    },
+    {
+      "epoch": 0.47,
+      "learning_rate": 1.5187200000000003e-05,
+      "loss": 2.4029,
+      "step": 254
+    },
+    {
+      "epoch": 0.47,
+      "learning_rate": 1.5184000000000002e-05,
+      "loss": 2.3158,
+      "step": 255
+    },
+    {
+      "epoch": 0.47,
+      "learning_rate": 1.5180800000000002e-05,
+      "loss": 2.3265,
+      "step": 256
+    },
+    {
+      "epoch": 0.47,
+      "learning_rate": 1.5177600000000002e-05,
+      "loss": 2.3196,
+      "step": 257
+    },
+    {
+      "epoch": 0.48,
+      "learning_rate": 1.5174400000000003e-05,
+      "loss": 2.3007,
+      "step": 258
+    },
+    {
+      "epoch": 0.48,
+      "learning_rate": 1.5171200000000003e-05,
+      "loss": 2.3006,
+      "step": 259
+    },
+    {
+      "epoch": 0.48,
+      "learning_rate": 1.5168000000000001e-05,
+      "loss": 2.28,
+      "step": 260
+    },
+    {
+      "epoch": 0.48,
+      "learning_rate": 1.5164800000000003e-05,
+      "loss": 2.3373,
+      "step": 261
+    },
+    {
+      "epoch": 0.48,
+      "learning_rate": 1.5161600000000002e-05,
+      "loss": 2.3462,
+      "step": 262
+    },
+    {
+      "epoch": 0.48,
+      "learning_rate": 1.5158400000000002e-05,
+      "loss": 2.3536,
+      "step": 263
+    },
+    {
+      "epoch": 0.49,
+      "learning_rate": 1.5155200000000004e-05,
+      "loss": 2.3202,
+      "step": 264
+    },
+    {
+      "epoch": 0.49,
+      "learning_rate": 1.5152000000000002e-05,
+      "loss": 2.3453,
+      "step": 265
+    },
+    {
+      "epoch": 0.49,
+      "learning_rate": 1.5148800000000001e-05,
+      "loss": 2.2496,
+      "step": 266
+    },
+    {
+      "epoch": 0.49,
+      "learning_rate": 1.5145600000000003e-05,
+      "loss": 2.2931,
+      "step": 267
+    },
+    {
+      "epoch": 0.49,
+      "learning_rate": 1.5142400000000003e-05,
+      "loss": 2.3647,
+      "step": 268
+    },
+    {
+      "epoch": 0.5,
+      "learning_rate": 1.5139200000000002e-05,
+      "loss": 2.328,
+      "step": 269
+    },
+    {
+      "epoch": 0.5,
+      "learning_rate": 1.5136000000000002e-05,
+      "loss": 2.2888,
+      "step": 270
+    },
+    {
+      "epoch": 0.5,
+      "eval_loss": 2.310141086578369,
+      "eval_runtime": 233.6977,
+      "eval_samples_per_second": 37.112,
+      "eval_steps_per_second": 0.582,
+      "step": 270
+    },
+    {
+      "epoch": 0.5,
+      "learning_rate": 1.5132800000000002e-05,
+      "loss": 2.2486,
+      "step": 271
+    },
+    {
+      "epoch": 0.5,
+      "learning_rate": 1.5129600000000002e-05,
+      "loss": 2.2706,
+      "step": 272
+    },
+    {
+      "epoch": 0.5,
+      "learning_rate": 1.5126400000000003e-05,
+      "loss": 2.3892,
+      "step": 273
+    },
+    {
+      "epoch": 0.51,
+      "learning_rate": 1.5123200000000003e-05,
+      "loss": 2.2992,
+      "step": 274
+    },
+    {
+      "epoch": 0.51,
+      "learning_rate": 1.5120000000000001e-05,
+      "loss": 2.3328,
+      "step": 275
+    },
+    {
+      "epoch": 0.51,
+      "learning_rate": 1.5116800000000002e-05,
+      "loss": 2.3367,
+      "step": 276
+    },
+    {
+      "epoch": 0.51,
+      "learning_rate": 1.5113600000000002e-05,
+      "loss": 2.3019,
+      "step": 277
+    },
+    {
+      "epoch": 0.51,
+      "learning_rate": 1.5110400000000004e-05,
+      "loss": 2.3007,
+      "step": 278
+    },
+    {
+      "epoch": 0.51,
+      "learning_rate": 1.5107200000000003e-05,
+      "loss": 2.3109,
+      "step": 279
+    },
+    {
+      "epoch": 0.52,
+      "learning_rate": 1.5104000000000001e-05,
+      "loss": 2.3243,
+      "step": 280
+    },
+    {
+      "epoch": 0.52,
+      "learning_rate": 1.5100800000000003e-05,
+      "loss": 2.3073,
+      "step": 281
+    },
+    {
+      "epoch": 0.52,
+      "learning_rate": 1.5097600000000003e-05,
+      "loss": 2.2874,
+      "step": 282
+    },
+    {
+      "epoch": 0.52,
+      "learning_rate": 1.5094400000000002e-05,
+      "loss": 2.2545,
+      "step": 283
+    },
+    {
+      "epoch": 0.52,
+      "learning_rate": 1.5091200000000004e-05,
+      "loss": 2.3699,
+      "step": 284
+    },
+    {
+      "epoch": 0.53,
+      "learning_rate": 1.5088000000000002e-05,
+      "loss": 2.3508,
+      "step": 285
+    },
+    {
+      "epoch": 0.53,
+      "learning_rate": 1.5084800000000002e-05,
+      "loss": 2.3006,
+      "step": 286
+    },
+    {
+      "epoch": 0.53,
+      "learning_rate": 1.5081600000000003e-05,
+      "loss": 2.3384,
+      "step": 287
+    },
+    {
+      "epoch": 0.53,
+      "learning_rate": 1.5078400000000003e-05,
+      "loss": 2.3202,
+      "step": 288
+    },
+    {
+      "epoch": 0.53,
+      "learning_rate": 1.5075200000000003e-05,
+      "loss": 2.3304,
+      "step": 289
+    },
+    {
+      "epoch": 0.53,
+      "learning_rate": 1.5072000000000002e-05,
+      "loss": 2.3224,
+      "step": 290
+    },
+    {
+      "epoch": 0.54,
+      "learning_rate": 1.5068800000000002e-05,
+      "loss": 2.3924,
+      "step": 291
+    },
+    {
+      "epoch": 0.54,
+      "learning_rate": 1.5065600000000002e-05,
+      "loss": 2.3047,
+      "step": 292
+    },
+    {
+      "epoch": 0.54,
+      "learning_rate": 1.5062400000000003e-05,
+      "loss": 2.3332,
+      "step": 293
+    },
+    {
+      "epoch": 0.54,
+      "learning_rate": 1.5059200000000003e-05,
+      "loss": 2.2173,
+      "step": 294
+    },
+    {
+      "epoch": 0.54,
+      "learning_rate": 1.5056000000000001e-05,
+      "loss": 2.4019,
+      "step": 295
+    },
+    {
+      "epoch": 0.55,
+      "learning_rate": 1.5052800000000003e-05,
+      "loss": 2.3265,
+      "step": 296
+    },
+    {
+      "epoch": 0.55,
+      "learning_rate": 1.5049600000000002e-05,
+      "loss": 2.2619,
+      "step": 297
+    },
+    {
+      "epoch": 0.55,
+      "learning_rate": 1.5046400000000002e-05,
+      "loss": 2.3385,
+      "step": 298
+    },
+    {
+      "epoch": 0.55,
+      "learning_rate": 1.5043200000000004e-05,
+      "loss": 2.3074,
+      "step": 299
+    },
+    {
+      "epoch": 0.55,
+      "learning_rate": 1.5040000000000002e-05,
+      "loss": 2.3238,
+      "step": 300
+    },
+    {
+      "epoch": 0.55,
+      "eval_loss": 2.311220169067383,
+      "eval_runtime": 232.933,
+      "eval_samples_per_second": 37.234,
+      "eval_steps_per_second": 0.584,
+      "step": 300
+    },
+    {
+      "epoch": 0.55,
+      "learning_rate": 1.5036800000000001e-05,
+      "loss": 2.3395,
+      "step": 301
+    },
+    {
+      "epoch": 0.56,
+      "learning_rate": 1.5033600000000003e-05,
+      "loss": 2.2715,
+      "step": 302
+    },
+    {
+      "epoch": 0.56,
+      "learning_rate": 1.5030400000000003e-05,
+      "loss": 2.3071,
+      "step": 303
+    },
+    {
+      "epoch": 0.56,
+      "learning_rate": 1.5027200000000002e-05,
+      "loss": 2.4088,
+      "step": 304
+    },
+    {
+      "epoch": 0.56,
+      "learning_rate": 1.5024000000000002e-05,
+      "loss": 2.2675,
+      "step": 305
+    },
+    {
+      "epoch": 0.56,
+      "learning_rate": 1.5020800000000002e-05,
+      "loss": 2.3524,
+      "step": 306
+    },
+    {
+      "epoch": 0.57,
+      "learning_rate": 1.5017600000000002e-05,
+      "loss": 2.3838,
+      "step": 307
+    },
+    {
+      "epoch": 0.57,
+      "learning_rate": 1.5014400000000003e-05,
+      "loss": 2.3926,
+      "step": 308
+    },
+    {
+      "epoch": 0.57,
+      "learning_rate": 1.5011200000000003e-05,
+      "loss": 2.3051,
+      "step": 309
+    },
+    {
+      "epoch": 0.57,
+      "learning_rate": 1.5008000000000001e-05,
+      "loss": 2.267,
+      "step": 310
+    },
+    {
+      "epoch": 0.57,
+      "learning_rate": 1.5004800000000002e-05,
+      "loss": 2.3642,
+      "step": 311
+    },
+    {
+      "epoch": 0.58,
+      "learning_rate": 1.5001600000000002e-05,
+      "loss": 2.2667,
+      "step": 312
+    },
+    {
+      "epoch": 0.58,
+      "learning_rate": 1.4998400000000002e-05,
+      "loss": 2.2641,
+      "step": 313
+    },
+    {
+      "epoch": 0.58,
+      "learning_rate": 1.4995200000000003e-05,
+      "loss": 2.315,
+      "step": 314
+    },
+    {
+      "epoch": 0.58,
+      "learning_rate": 1.4992000000000003e-05,
+      "loss": 2.2675,
+      "step": 315
+    },
+    {
+      "epoch": 0.58,
+      "learning_rate": 1.4988800000000001e-05,
+      "loss": 2.2665,
+      "step": 316
+    },
+    {
+      "epoch": 0.58,
+      "learning_rate": 1.4985600000000003e-05,
+      "loss": 2.2197,
+      "step": 317
+    },
+    {
+      "epoch": 0.59,
+      "learning_rate": 1.4982400000000002e-05,
+      "loss": 2.3284,
+      "step": 318
+    },
+    {
+      "epoch": 0.59,
+      "learning_rate": 1.4979200000000002e-05,
+      "loss": 2.2286,
+      "step": 319
+    },
+    {
+      "epoch": 0.59,
+      "learning_rate": 1.4976000000000004e-05,
+      "loss": 2.3843,
+      "step": 320
+    },
+    {
+      "epoch": 0.59,
+      "learning_rate": 1.4972800000000002e-05,
+      "loss": 2.2413,
+      "step": 321
+    },
+    {
+      "epoch": 0.59,
+      "learning_rate": 1.4969600000000001e-05,
+      "loss": 2.2844,
+      "step": 322
+    },
+    {
+      "epoch": 0.6,
+      "learning_rate": 1.4966400000000003e-05,
+      "loss": 2.3443,
+      "step": 323
+    },
+    {
+      "epoch": 0.6,
+      "learning_rate": 1.4963200000000003e-05,
+      "loss": 2.3216,
+      "step": 324
+    },
+    {
+      "epoch": 0.6,
+      "learning_rate": 1.4960000000000004e-05,
+      "loss": 2.2981,
+      "step": 325
+    },
+    {
+      "epoch": 0.6,
+      "learning_rate": 1.4956800000000002e-05,
+      "loss": 2.2741,
+      "step": 326
+    },
+    {
+      "epoch": 0.6,
+      "learning_rate": 1.4953600000000002e-05,
+      "loss": 2.236,
+      "step": 327
+    },
+    {
+      "epoch": 0.6,
+      "learning_rate": 1.4950400000000003e-05,
+      "loss": 2.2546,
+      "step": 328
+    },
+    {
+      "epoch": 0.61,
+      "learning_rate": 1.4947200000000003e-05,
+      "loss": 2.3405,
+      "step": 329
+    },
+    {
+      "epoch": 0.61,
+      "learning_rate": 1.4944000000000003e-05,
+      "loss": 2.377,
+      "step": 330
+    },
+    {
+      "epoch": 0.61,
+      "eval_loss": 2.3087046146392822,
+      "eval_runtime": 236.2566,
+      "eval_samples_per_second": 36.71,
+      "eval_steps_per_second": 0.576,
+      "step": 330
+    },
+    {
+      "epoch": 0.61,
+      "learning_rate": 1.4940800000000003e-05,
+      "loss": 2.3051,
+      "step": 331
+    },
+    {
+      "epoch": 0.61,
+      "learning_rate": 1.4937600000000002e-05,
+      "loss": 2.2366,
+      "step": 332
+    },
+    {
+      "epoch": 0.61,
+      "learning_rate": 1.4934400000000002e-05,
+      "loss": 2.315,
+      "step": 333
+    },
+    {
+      "epoch": 0.62,
+      "learning_rate": 1.4931200000000004e-05,
+      "loss": 2.2576,
+      "step": 334
+    },
+    {
+      "epoch": 0.62,
+      "learning_rate": 1.4928000000000003e-05,
+      "loss": 2.3574,
+      "step": 335
+    },
+    {
+      "epoch": 0.62,
+      "learning_rate": 1.4924800000000002e-05,
+      "loss": 2.323,
+      "step": 336
+    },
+    {
+      "epoch": 0.62,
+      "learning_rate": 1.4921600000000003e-05,
+      "loss": 2.3033,
+      "step": 337
+    },
+    {
+      "epoch": 0.62,
+      "learning_rate": 1.4918400000000003e-05,
+      "loss": 2.3013,
+      "step": 338
+    },
+    {
+      "epoch": 0.62,
+      "learning_rate": 1.4915200000000002e-05,
+      "loss": 2.3549,
+      "step": 339
+    },
+    {
+      "epoch": 0.63,
+      "learning_rate": 1.4912000000000004e-05,
+      "loss": 2.2643,
+      "step": 340
+    },
+    {
+      "epoch": 0.63,
+      "learning_rate": 1.4908800000000002e-05,
+      "loss": 2.2902,
+      "step": 341
+    },
+    {
+      "epoch": 0.63,
+      "learning_rate": 1.4905600000000002e-05,
+      "loss": 2.287,
+      "step": 342
+    },
+    {
+      "epoch": 0.63,
+      "learning_rate": 1.4902400000000003e-05,
+      "loss": 2.3526,
+      "step": 343
+    },
+    {
+      "epoch": 0.63,
+      "learning_rate": 1.4899200000000003e-05,
+      "loss": 2.3976,
+      "step": 344
+    },
+    {
+      "epoch": 0.64,
+      "learning_rate": 1.4896000000000003e-05,
+      "loss": 2.3484,
+      "step": 345
+    },
+    {
+      "epoch": 0.64,
+      "learning_rate": 1.4892800000000002e-05,
+      "loss": 2.4194,
+      "step": 346
+    },
+    {
+      "epoch": 0.64,
+      "learning_rate": 1.4889600000000002e-05,
+      "loss": 2.3177,
+      "step": 347
+    },
+    {
+      "epoch": 0.64,
+      "learning_rate": 1.4886400000000002e-05,
+      "loss": 2.278,
+      "step": 348
+    },
+    {
+      "epoch": 0.64,
+      "learning_rate": 1.4883200000000003e-05,
+      "loss": 2.3392,
+      "step": 349
+    },
+    {
+      "epoch": 0.65,
+      "learning_rate": 1.4880000000000003e-05,
+      "loss": 2.3189,
+      "step": 350
+    },
+    {
+      "epoch": 0.65,
+      "learning_rate": 1.4876800000000001e-05,
+      "loss": 2.315,
+      "step": 351
+    },
+    {
+      "epoch": 0.65,
+      "learning_rate": 1.4873600000000003e-05,
+      "loss": 2.2351,
+      "step": 352
+    },
+    {
+      "epoch": 0.65,
+      "learning_rate": 1.4870400000000003e-05,
+      "loss": 2.3228,
+      "step": 353
+    },
+    {
+      "epoch": 0.65,
+      "learning_rate": 1.4867200000000002e-05,
+      "loss": 2.3167,
+      "step": 354
+    },
+    {
+      "epoch": 0.65,
+      "learning_rate": 1.4864000000000004e-05,
+      "loss": 2.2967,
+      "step": 355
+    },
+    {
+      "epoch": 0.66,
+      "learning_rate": 1.4860800000000002e-05,
+      "loss": 2.3022,
+      "step": 356
+    },
+    {
+      "epoch": 0.66,
+      "learning_rate": 1.4857600000000002e-05,
+      "loss": 2.3298,
+      "step": 357
+    },
+    {
+      "epoch": 0.66,
+      "learning_rate": 1.4854400000000003e-05,
+      "loss": 2.3471,
+      "step": 358
+    },
+    {
+      "epoch": 0.66,
+      "learning_rate": 1.4851200000000003e-05,
+      "loss": 2.2998,
+      "step": 359
+    },
+    {
+      "epoch": 0.66,
+      "learning_rate": 1.4848000000000003e-05,
+      "loss": 2.2358,
+      "step": 360
+    },
+    {
+      "epoch": 0.66,
+      "eval_loss": 2.3044841289520264,
+      "eval_runtime": 233.6248,
+      "eval_samples_per_second": 37.124,
+      "eval_steps_per_second": 0.582,
+      "step": 360
+    },
+    {
+      "epoch": 0.67,
+      "learning_rate": 1.4844800000000002e-05,
+      "loss": 2.3446,
+      "step": 361
+    },
+    {
+      "epoch": 0.67,
+      "learning_rate": 1.4841600000000002e-05,
+      "loss": 2.3578,
+      "step": 362
+    },
+    {
+      "epoch": 0.67,
+      "learning_rate": 1.4838400000000002e-05,
+      "loss": 2.2833,
+      "step": 363
+    },
+    {
+      "epoch": 0.67,
+      "learning_rate": 1.4835200000000003e-05,
+      "loss": 2.3079,
+      "step": 364
+    },
+    {
+      "epoch": 0.67,
+      "learning_rate": 1.4832000000000003e-05,
+      "loss": 2.3692,
+      "step": 365
+    },
+    {
+      "epoch": 0.67,
+      "learning_rate": 1.4828800000000001e-05,
+      "loss": 2.2472,
+      "step": 366
+    },
+    {
+      "epoch": 0.68,
+      "learning_rate": 1.4825600000000003e-05,
+      "loss": 2.3266,
+      "step": 367
+    },
+    {
+      "epoch": 0.68,
+      "learning_rate": 1.4822400000000002e-05,
+      "loss": 2.3253,
+      "step": 368
+    },
+    {
+      "epoch": 0.68,
+      "learning_rate": 1.4819200000000002e-05,
+      "loss": 2.4031,
+      "step": 369
+    },
+    {
+      "epoch": 0.68,
+      "learning_rate": 1.4816000000000003e-05,
+      "loss": 2.3271,
+      "step": 370
+    },
+    {
+      "epoch": 0.68,
+      "learning_rate": 1.4812800000000002e-05,
+      "loss": 2.3289,
+      "step": 371
+    },
+    {
+      "epoch": 0.69,
+      "learning_rate": 1.4809600000000001e-05,
+      "loss": 2.3968,
+      "step": 372
+    },
+    {
+      "epoch": 0.69,
+      "learning_rate": 1.4806400000000003e-05,
+      "loss": 2.269,
+      "step": 373
+    },
+    {
+      "epoch": 0.69,
+      "learning_rate": 1.4803200000000003e-05,
+      "loss": 2.2668,
+      "step": 374
+    },
+    {
+      "epoch": 0.69,
+      "learning_rate": 1.4800000000000004e-05,
+      "loss": 2.2337,
+      "step": 375
+    },
+    {
+      "epoch": 0.69,
+      "learning_rate": 1.4796800000000002e-05,
+      "loss": 2.2487,
+      "step": 376
+    },
+    {
+      "epoch": 0.69,
+      "learning_rate": 1.4793600000000002e-05,
+      "loss": 2.283,
+      "step": 377
+    },
+    {
+      "epoch": 0.7,
+      "learning_rate": 1.4790400000000003e-05,
+      "loss": 2.3126,
+      "step": 378
+    },
+    {
+      "epoch": 0.7,
+      "learning_rate": 1.4787200000000003e-05,
+      "loss": 2.3735,
+      "step": 379
+    },
+    {
+      "epoch": 0.7,
+      "learning_rate": 1.4784000000000003e-05,
+      "loss": 2.2532,
+      "step": 380
+    },
+    {
+      "epoch": 0.7,
+      "learning_rate": 1.4780800000000003e-05,
+      "loss": 2.2739,
+      "step": 381
+    },
+    {
+      "epoch": 0.7,
+      "learning_rate": 1.4777600000000002e-05,
+      "loss": 2.3643,
+      "step": 382
+    },
+    {
+      "epoch": 0.71,
+      "learning_rate": 1.4774400000000002e-05,
+      "loss": 2.2826,
+      "step": 383
+    },
+    {
+      "epoch": 0.71,
+      "learning_rate": 1.4771200000000004e-05,
+      "loss": 2.3355,
+      "step": 384
+    },
+    {
+      "epoch": 0.71,
+      "learning_rate": 1.4768000000000003e-05,
+      "loss": 2.2951,
+      "step": 385
+    },
+    {
+      "epoch": 0.71,
+      "learning_rate": 1.4764800000000001e-05,
+      "loss": 2.2563,
+      "step": 386
+    },
+    {
+      "epoch": 0.71,
+      "learning_rate": 1.4761600000000003e-05,
+      "loss": 2.3538,
+      "step": 387
+    },
+    {
+      "epoch": 0.72,
+      "learning_rate": 1.4758400000000003e-05,
+      "loss": 2.1856,
+      "step": 388
+    },
+    {
+      "epoch": 0.72,
+      "learning_rate": 1.4755200000000002e-05,
+      "loss": 2.2418,
+      "step": 389
+    },
+    {
+      "epoch": 0.72,
+      "learning_rate": 1.4752000000000004e-05,
+      "loss": 2.3026,
+      "step": 390
+    },
+    {
+      "epoch": 0.72,
+      "eval_loss": 2.3032002449035645,
+      "eval_runtime": 235.445,
+      "eval_samples_per_second": 36.837,
+      "eval_steps_per_second": 0.578,
+      "step": 390
+    },
+    {
+      "epoch": 0.72,
+      "learning_rate": 1.4748800000000002e-05,
+      "loss": 2.2479,
+      "step": 391
+    },
+    {
+      "epoch": 0.72,
+      "learning_rate": 1.4745600000000002e-05,
+      "loss": 2.2516,
+      "step": 392
+    },
+    {
+      "epoch": 0.72,
+      "learning_rate": 1.4742400000000003e-05,
+      "loss": 2.3399,
+      "step": 393
+    },
+    {
+      "epoch": 0.73,
+      "learning_rate": 1.4739200000000003e-05,
+      "loss": 2.3146,
+      "step": 394
+    },
+    {
+      "epoch": 0.73,
+      "learning_rate": 1.4736000000000003e-05,
+      "loss": 2.321,
+      "step": 395
+    },
+    {
+      "epoch": 0.73,
+      "learning_rate": 1.4732800000000002e-05,
+      "loss": 2.3183,
+      "step": 396
+    },
+    {
+      "epoch": 0.73,
+      "learning_rate": 1.4729600000000002e-05,
+      "loss": 2.3521,
+      "step": 397
+    },
+    {
+      "epoch": 0.73,
+      "learning_rate": 1.4726400000000002e-05,
+      "loss": 2.2776,
+      "step": 398
+    },
+    {
+      "epoch": 0.74,
+      "learning_rate": 1.4723200000000003e-05,
+      "loss": 2.3651,
+      "step": 399
+    },
+    {
+      "epoch": 0.74,
+      "learning_rate": 1.4720000000000003e-05,
+      "loss": 2.2907,
+      "step": 400
+    },
+    {
+      "epoch": 0.74,
+      "learning_rate": 1.4716800000000001e-05,
+      "loss": 2.2708,
+      "step": 401
+    },
+    {
+      "epoch": 0.74,
+      "learning_rate": 1.4713600000000003e-05,
+      "loss": 2.3394,
+      "step": 402
+    },
+    {
+      "epoch": 0.74,
+      "learning_rate": 1.4710400000000002e-05,
+      "loss": 2.2168,
+      "step": 403
+    },
+    {
+      "epoch": 0.74,
+      "learning_rate": 1.4707200000000002e-05,
+      "loss": 2.2749,
+      "step": 404
+    },
+    {
+      "epoch": 0.75,
+      "learning_rate": 1.4704000000000004e-05,
+      "loss": 2.4253,
+      "step": 405
+    },
+    {
+      "epoch": 0.75,
+      "learning_rate": 1.4700800000000002e-05,
+      "loss": 2.2722,
+      "step": 406
+    },
+    {
+      "epoch": 0.75,
+      "learning_rate": 1.4697600000000001e-05,
+      "loss": 2.3289,
+      "step": 407
+    },
+    {
+      "epoch": 0.75,
+      "learning_rate": 1.4694400000000003e-05,
+      "loss": 2.3695,
+      "step": 408
+    },
+    {
+      "epoch": 0.75,
+      "learning_rate": 1.4691200000000003e-05,
+      "loss": 2.3748,
+      "step": 409
+    },
+    {
+      "epoch": 0.76,
+      "learning_rate": 1.4688000000000002e-05,
+      "loss": 2.3178,
+      "step": 410
+    },
+    {
+      "epoch": 0.76,
+      "learning_rate": 1.4684800000000002e-05,
+      "loss": 2.294,
+      "step": 411
+    },
+    {
+      "epoch": 0.76,
+      "learning_rate": 1.4681600000000002e-05,
+      "loss": 2.2224,
+      "step": 412
+    },
+    {
+      "epoch": 0.76,
+      "learning_rate": 1.4678400000000002e-05,
+      "loss": 2.2724,
+      "step": 413
+    },
+    {
+      "epoch": 0.76,
+      "learning_rate": 1.4675200000000003e-05,
+      "loss": 2.3189,
+      "step": 414
+    },
+    {
+      "epoch": 0.76,
+      "learning_rate": 1.4672000000000003e-05,
+      "loss": 2.3091,
+      "step": 415
+    },
+    {
+      "epoch": 0.77,
+      "learning_rate": 1.4668800000000001e-05,
+      "loss": 2.3082,
+      "step": 416
+    },
+    {
+      "epoch": 0.77,
+      "learning_rate": 1.4665600000000002e-05,
+      "loss": 2.2514,
+      "step": 417
+    },
+    {
+      "epoch": 0.77,
+      "learning_rate": 1.4662400000000002e-05,
+      "loss": 2.3146,
+      "step": 418
+    },
+    {
+      "epoch": 0.77,
+      "learning_rate": 1.4659200000000002e-05,
+      "loss": 2.2698,
+      "step": 419
+    },
+    {
+      "epoch": 0.77,
+      "learning_rate": 1.4656000000000003e-05,
+      "loss": 2.248,
+      "step": 420
+    },
+    {
+      "epoch": 0.77,
+      "eval_loss": 2.3027491569519043,
+      "eval_runtime": 235.5268,
+      "eval_samples_per_second": 36.824,
+      "eval_steps_per_second": 0.577,
+      "step": 420
+    },
+    {
+      "epoch": 0.78,
+      "learning_rate": 1.4652800000000001e-05,
+      "loss": 2.2855,
+      "step": 421
+    },
+    {
+      "epoch": 0.78,
+      "learning_rate": 1.4649600000000001e-05,
+      "loss": 2.3107,
+      "step": 422
+    },
+    {
+      "epoch": 0.78,
+      "learning_rate": 1.4646400000000003e-05,
+      "loss": 2.3275,
+      "step": 423
+    },
+    {
+      "epoch": 0.78,
+      "learning_rate": 1.4643200000000002e-05,
+      "loss": 2.3411,
+      "step": 424
+    },
+    {
+      "epoch": 0.78,
+      "learning_rate": 1.4640000000000004e-05,
+      "loss": 2.3309,
+      "step": 425
+    },
+    {
+      "epoch": 0.79,
+      "learning_rate": 1.4636800000000002e-05,
+      "loss": 2.2853,
+      "step": 426
+    },
+    {
+      "epoch": 0.79,
+      "learning_rate": 1.4633600000000002e-05,
+      "loss": 2.2874,
+      "step": 427
+    },
+    {
+      "epoch": 0.79,
+      "learning_rate": 1.4630400000000003e-05,
+      "loss": 2.2757,
+      "step": 428
+    },
+    {
+      "epoch": 0.79,
+      "learning_rate": 1.4627200000000003e-05,
+      "loss": 2.3485,
+      "step": 429
+    },
+    {
+      "epoch": 0.79,
+      "learning_rate": 1.4624000000000003e-05,
+      "loss": 2.3152,
+      "step": 430
+    },
+    {
+      "epoch": 0.79,
+      "learning_rate": 1.4620800000000002e-05,
+      "loss": 2.3061,
+      "step": 431
+    },
+    {
+      "epoch": 0.8,
+      "learning_rate": 1.4617600000000002e-05,
+      "loss": 2.2332,
+      "step": 432
+    },
+    {
+      "epoch": 0.8,
+      "learning_rate": 1.4614400000000002e-05,
+      "loss": 2.2273,
+      "step": 433
+    },
+    {
+      "epoch": 0.8,
+      "learning_rate": 1.4611200000000003e-05,
+      "loss": 2.2821,
+      "step": 434
+    },
+    {
+      "epoch": 0.8,
+      "learning_rate": 1.4608000000000003e-05,
+      "loss": 2.2428,
+      "step": 435
+    },
+    {
+      "epoch": 0.8,
+      "learning_rate": 1.4604800000000001e-05,
+      "loss": 2.1994,
+      "step": 436
+    },
+    {
+      "epoch": 0.81,
+      "learning_rate": 1.4601600000000003e-05,
+      "loss": 2.399,
+      "step": 437
+    },
+    {
+      "epoch": 0.81,
+      "learning_rate": 1.4598400000000002e-05,
+      "loss": 2.2995,
+      "step": 438
+    },
+    {
+      "epoch": 0.81,
+      "learning_rate": 1.4595200000000002e-05,
+      "loss": 2.1998,
+      "step": 439
+    },
+    {
+      "epoch": 0.81,
+      "learning_rate": 1.4592000000000004e-05,
+      "loss": 2.3414,
+      "step": 440
+    },
+    {
+      "epoch": 0.81,
+      "learning_rate": 1.4588800000000003e-05,
+      "loss": 2.3244,
+      "step": 441
+    },
+    {
+      "epoch": 0.81,
+      "learning_rate": 1.4585600000000001e-05,
+      "loss": 2.214,
+      "step": 442
+    },
+    {
+      "epoch": 0.82,
+      "learning_rate": 1.4582400000000003e-05,
+      "loss": 2.2491,
+      "step": 443
+    },
+    {
+      "epoch": 0.82,
+      "learning_rate": 1.4579200000000003e-05,
+      "loss": 2.2686,
+      "step": 444
+    },
+    {
+      "epoch": 0.82,
+      "learning_rate": 1.4576000000000002e-05,
+      "loss": 2.2724,
+      "step": 445
+    },
+    {
+      "epoch": 0.82,
+      "learning_rate": 1.4572800000000004e-05,
+      "loss": 2.2836,
+      "step": 446
+    },
+    {
+      "epoch": 0.82,
+      "learning_rate": 1.4569600000000002e-05,
+      "loss": 2.2421,
+      "step": 447
+    },
+    {
+      "epoch": 0.83,
+      "learning_rate": 1.4566400000000002e-05,
+      "loss": 2.331,
+      "step": 448
+    },
+    {
+      "epoch": 0.83,
+      "learning_rate": 1.4563200000000003e-05,
+      "loss": 2.2311,
+      "step": 449
+    },
+    {
+      "epoch": 0.83,
+      "learning_rate": 1.4560000000000003e-05,
+      "loss": 2.3675,
+      "step": 450
+    },
+    {
+      "epoch": 0.83,
+      "eval_loss": 2.3037469387054443,
+      "eval_runtime": 233.2792,
+      "eval_samples_per_second": 37.179,
+      "eval_steps_per_second": 0.583,
+      "step": 450
+    },
+    {
+      "epoch": 0.83,
+      "learning_rate": 1.4556800000000003e-05,
+      "loss": 2.2116,
+      "step": 451
+    },
+    {
+      "epoch": 0.83,
+      "learning_rate": 1.4553600000000002e-05,
+      "loss": 2.2952,
+      "step": 452
+    },
+    {
+      "epoch": 0.84,
+      "learning_rate": 1.4550400000000002e-05,
+      "loss": 2.3275,
+      "step": 453
+    },
+    {
+      "epoch": 0.84,
+      "learning_rate": 1.4547200000000002e-05,
+      "loss": 2.2883,
+      "step": 454
+    },
+    {
+      "epoch": 0.84,
+      "learning_rate": 1.4544000000000003e-05,
+      "loss": 2.2209,
+      "step": 455
+    },
+    {
+      "epoch": 0.84,
+      "learning_rate": 1.4540800000000003e-05,
+      "loss": 2.3054,
+      "step": 456
+    },
+    {
+      "epoch": 0.84,
+      "learning_rate": 1.4537600000000001e-05,
+      "loss": 2.3142,
+      "step": 457
+    },
+    {
+      "epoch": 0.84,
+      "learning_rate": 1.4534400000000003e-05,
+      "loss": 2.2407,
+      "step": 458
+    },
+    {
+      "epoch": 0.85,
+      "learning_rate": 1.4531200000000002e-05,
+      "loss": 2.3158,
+      "step": 459
+    },
+    {
+      "epoch": 0.85,
+      "learning_rate": 1.4528000000000002e-05,
+      "loss": 2.3749,
+      "step": 460
+    },
+    {
+      "epoch": 0.85,
+      "learning_rate": 1.4524800000000004e-05,
+      "loss": 2.1941,
+      "step": 461
+    },
+    {
+      "epoch": 0.85,
+      "learning_rate": 1.4521600000000002e-05,
+      "loss": 2.3356,
+      "step": 462
+    },
+    {
+      "epoch": 0.85,
+      "learning_rate": 1.4518400000000001e-05,
+      "loss": 2.3059,
+      "step": 463
+    },
+    {
+      "epoch": 0.86,
+      "learning_rate": 1.4515200000000003e-05,
+      "loss": 2.2561,
+      "step": 464
+    },
+    {
+      "epoch": 0.86,
+      "learning_rate": 1.4512000000000003e-05,
+      "loss": 2.3465,
+      "step": 465
+    },
+    {
+      "epoch": 0.86,
+      "learning_rate": 1.4508800000000002e-05,
+      "loss": 2.2718,
+      "step": 466
+    },
+    {
+      "epoch": 0.86,
+      "learning_rate": 1.4505600000000002e-05,
+      "loss": 2.2663,
+      "step": 467
+    },
+    {
+      "epoch": 0.86,
+      "learning_rate": 1.4502400000000002e-05,
+      "loss": 2.2953,
+      "step": 468
+    },
+    {
+      "epoch": 0.86,
+      "learning_rate": 1.4499200000000002e-05,
+      "loss": 2.2959,
+      "step": 469
+    },
+    {
+      "epoch": 0.87,
+      "learning_rate": 1.4496000000000003e-05,
+      "loss": 2.278,
+      "step": 470
+    },
+    {
+      "epoch": 0.87,
+      "learning_rate": 1.4492800000000003e-05,
+      "loss": 2.3824,
+      "step": 471
+    },
+    {
+      "epoch": 0.87,
+      "learning_rate": 1.4489600000000001e-05,
+      "loss": 2.353,
+      "step": 472
+    },
+    {
+      "epoch": 0.87,
+      "learning_rate": 1.4486400000000002e-05,
+      "loss": 2.3246,
+      "step": 473
+    },
+    {
+      "epoch": 0.87,
+      "learning_rate": 1.4483200000000002e-05,
+      "loss": 2.2943,
+      "step": 474
+    },
+    {
+      "epoch": 0.88,
+      "learning_rate": 1.4480000000000004e-05,
+      "loss": 2.2964,
+      "step": 475
+    },
+    {
+      "epoch": 0.88,
+      "learning_rate": 1.4476800000000003e-05,
+      "loss": 2.3019,
+      "step": 476
+    },
+    {
+      "epoch": 0.88,
+      "learning_rate": 1.4473600000000001e-05,
+      "loss": 2.3156,
+      "step": 477
+    },
+    {
+      "epoch": 0.88,
+      "learning_rate": 1.4470400000000003e-05,
+      "loss": 2.2881,
+      "step": 478
+    },
+    {
+      "epoch": 0.88,
+      "learning_rate": 1.4467200000000003e-05,
+      "loss": 2.3454,
+      "step": 479
+    },
+    {
+      "epoch": 0.88,
+      "learning_rate": 1.4464000000000002e-05,
+      "loss": 2.2718,
+      "step": 480
+    },
+    {
+      "epoch": 0.88,
+      "eval_loss": 2.2999613285064697,
+      "eval_runtime": 232.9734,
+      "eval_samples_per_second": 37.227,
+      "eval_steps_per_second": 0.584,
+      "step": 480
+    },
+    {
+      "epoch": 0.89,
+      "learning_rate": 1.4460800000000004e-05,
+      "loss": 2.3102,
+      "step": 481
+    },
+    {
+      "epoch": 0.89,
+      "learning_rate": 1.4457600000000002e-05,
+      "loss": 2.2679,
+      "step": 482
+    },
+    {
+      "epoch": 0.89,
+      "learning_rate": 1.4454400000000002e-05,
+      "loss": 2.3396,
+      "step": 483
+    },
+    {
+      "epoch": 0.89,
+      "learning_rate": 1.4451200000000003e-05,
+      "loss": 2.2578,
+      "step": 484
+    },
+    {
+      "epoch": 0.89,
+      "learning_rate": 1.4448000000000003e-05,
+      "loss": 2.2606,
+      "step": 485
+    },
+    {
+      "epoch": 0.9,
+      "learning_rate": 1.4444800000000003e-05,
+      "loss": 2.3177,
+      "step": 486
+    },
+    {
+      "epoch": 0.9,
+      "learning_rate": 1.4441600000000002e-05,
+      "loss": 2.2287,
+      "step": 487
+    },
+    {
+      "epoch": 0.9,
+      "learning_rate": 1.4438400000000002e-05,
+      "loss": 2.2568,
+      "step": 488
+    },
+    {
+      "epoch": 0.9,
+      "learning_rate": 1.4435200000000002e-05,
+      "loss": 2.3595,
+      "step": 489
+    },
+    {
+      "epoch": 0.9,
+      "learning_rate": 1.4432000000000003e-05,
+      "loss": 2.331,
+      "step": 490
+    },
+    {
+      "epoch": 0.91,
+      "learning_rate": 1.4428800000000003e-05,
+      "loss": 2.268,
+      "step": 491
+    },
+    {
+      "epoch": 0.91,
+      "learning_rate": 1.4425600000000001e-05,
+      "loss": 2.2901,
+      "step": 492
+    },
+    {
+      "epoch": 0.91,
+      "learning_rate": 1.4422400000000003e-05,
+      "loss": 2.3807,
+      "step": 493
+    },
+    {
+      "epoch": 0.91,
+      "learning_rate": 1.4419200000000002e-05,
+      "loss": 2.3226,
+      "step": 494
+    },
+    {
+      "epoch": 0.91,
+      "learning_rate": 1.4416000000000002e-05,
+      "loss": 2.3805,
+      "step": 495
+    },
+    {
+      "epoch": 0.91,
+      "learning_rate": 1.4412800000000004e-05,
+      "loss": 2.3722,
+      "step": 496
+    },
+    {
+      "epoch": 0.92,
+      "learning_rate": 1.4409600000000002e-05,
+      "loss": 2.3298,
+      "step": 497
+    },
+    {
+      "epoch": 0.92,
+      "learning_rate": 1.4406400000000002e-05,
+      "loss": 2.3486,
+      "step": 498
+    },
+    {
+      "epoch": 0.92,
+      "learning_rate": 1.4403200000000003e-05,
+      "loss": 2.2935,
+      "step": 499
+    },
+    {
+      "epoch": 0.92,
+      "learning_rate": 1.4400000000000003e-05,
+      "loss": 2.3931,
+      "step": 500
+    },
+    {
+      "epoch": 0.92,
+      "learning_rate": 1.4396800000000002e-05,
+      "loss": 2.4184,
+      "step": 501
+    },
+    {
+      "epoch": 0.93,
+      "learning_rate": 1.4393600000000002e-05,
+      "loss": 2.3092,
+      "step": 502
+    },
+    {
+      "epoch": 0.93,
+      "learning_rate": 1.4390400000000002e-05,
+      "loss": 2.3223,
+      "step": 503
+    },
+    {
+      "epoch": 0.93,
+      "learning_rate": 1.4387200000000002e-05,
+      "loss": 2.1292,
+      "step": 504
+    },
+    {
+      "epoch": 0.93,
+      "learning_rate": 1.4384000000000003e-05,
+      "loss": 2.3141,
+      "step": 505
+    },
+    {
+      "epoch": 0.93,
+      "learning_rate": 1.4380800000000003e-05,
+      "loss": 2.3354,
+      "step": 506
+    },
+    {
+      "epoch": 0.93,
+      "learning_rate": 1.4377600000000001e-05,
+      "loss": 2.3592,
+      "step": 507
+    },
+    {
+      "epoch": 0.94,
+      "learning_rate": 1.4374400000000003e-05,
+      "loss": 2.2995,
+      "step": 508
+    },
+    {
+      "epoch": 0.94,
+      "learning_rate": 1.4371200000000002e-05,
+      "loss": 2.2825,
+      "step": 509
+    },
+    {
+      "epoch": 0.94,
+      "learning_rate": 1.4368000000000002e-05,
+      "loss": 2.2413,
+      "step": 510
+    },
+    {
+      "epoch": 0.94,
+      "eval_loss": 2.2967512607574463,
+      "eval_runtime": 232.8893,
+      "eval_samples_per_second": 37.241,
+      "eval_steps_per_second": 0.584,
+      "step": 510
+    },
+    {
+      "epoch": 0.94,
+      "learning_rate": 1.4364800000000003e-05,
+      "loss": 2.3555,
+      "step": 511
+    },
+    {
+      "epoch": 0.94,
+      "learning_rate": 1.4361600000000002e-05,
+      "loss": 2.3591,
+      "step": 512
+    },
+    {
+      "epoch": 0.95,
+      "learning_rate": 1.4358400000000001e-05,
+      "loss": 2.3544,
+      "step": 513
+    },
+    {
+      "epoch": 0.95,
+      "learning_rate": 1.4355200000000003e-05,
+      "loss": 2.3425,
+      "step": 514
+    },
+    {
+      "epoch": 0.95,
+      "learning_rate": 1.4352000000000003e-05,
+      "loss": 2.3037,
+      "step": 515
+    },
+    {
+      "epoch": 0.95,
+      "learning_rate": 1.4348800000000002e-05,
+      "loss": 2.2782,
+      "step": 516
+    },
+    {
+      "epoch": 0.95,
+      "learning_rate": 1.4345600000000002e-05,
+      "loss": 2.2897,
+      "step": 517
+    },
+    {
+      "epoch": 0.95,
+      "learning_rate": 1.4342400000000002e-05,
+      "loss": 2.3196,
+      "step": 518
+    },
+    {
+      "epoch": 0.96,
+      "learning_rate": 1.4339200000000002e-05,
+      "loss": 2.384,
+      "step": 519
+    },
+    {
+      "epoch": 0.96,
+      "learning_rate": 1.4336000000000003e-05,
+      "loss": 2.2617,
+      "step": 520
+    },
+    {
+      "epoch": 0.96,
+      "learning_rate": 1.4332800000000003e-05,
+      "loss": 2.2735,
+      "step": 521
+    },
+    {
+      "epoch": 0.96,
+      "learning_rate": 1.43296e-05,
+      "loss": 2.2238,
+      "step": 522
+    },
+    {
+      "epoch": 0.96,
+      "learning_rate": 1.4326400000000002e-05,
+      "loss": 2.2869,
+      "step": 523
+    },
+    {
+      "epoch": 0.97,
+      "learning_rate": 1.4323200000000002e-05,
+      "loss": 2.3476,
+      "step": 524
+    },
+    {
+      "epoch": 0.97,
+      "learning_rate": 1.4320000000000004e-05,
+      "loss": 2.2396,
+      "step": 525
+    },
+    {
+      "epoch": 0.97,
+      "learning_rate": 1.4316800000000003e-05,
+      "loss": 2.2643,
+      "step": 526
+    },
+    {
+      "epoch": 0.97,
+      "learning_rate": 1.4313600000000001e-05,
+      "loss": 2.2967,
+      "step": 527
+    },
+    {
+      "epoch": 0.97,
+      "learning_rate": 1.4310400000000003e-05,
+      "loss": 2.2945,
+      "step": 528
+    },
+    {
+      "epoch": 0.98,
+      "learning_rate": 1.4307200000000003e-05,
+      "loss": 2.3221,
+      "step": 529
+    },
+    {
+      "epoch": 0.98,
+      "learning_rate": 1.4304000000000002e-05,
+      "loss": 2.3217,
+      "step": 530
+    },
+    {
+      "epoch": 0.98,
+      "learning_rate": 1.4300800000000004e-05,
+      "loss": 2.2233,
+      "step": 531
+    },
+    {
+      "epoch": 0.98,
+      "learning_rate": 1.4297600000000002e-05,
+      "loss": 2.2762,
+      "step": 532
+    },
+    {
+      "epoch": 0.98,
+      "learning_rate": 1.4294400000000002e-05,
+      "loss": 2.3077,
+      "step": 533
+    },
+    {
+      "epoch": 0.98,
+      "learning_rate": 1.4291200000000003e-05,
+      "loss": 2.3276,
+      "step": 534
+    },
+    {
+      "epoch": 0.99,
+      "learning_rate": 1.4288000000000003e-05,
+      "loss": 2.2875,
+      "step": 535
+    },
+    {
+      "epoch": 0.99,
+      "learning_rate": 1.4284800000000003e-05,
+      "loss": 2.306,
+      "step": 536
+    },
+    {
+      "epoch": 0.99,
+      "learning_rate": 1.4281600000000002e-05,
+      "loss": 2.2854,
+      "step": 537
+    },
+    {
+      "epoch": 0.99,
+      "learning_rate": 1.4278400000000002e-05,
+      "loss": 2.3559,
+      "step": 538
+    },
+    {
+      "epoch": 0.99,
+      "learning_rate": 1.4275200000000002e-05,
+      "loss": 2.3465,
+      "step": 539
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 1.4272000000000003e-05,
+      "loss": 2.1996,
+      "step": 540
+    },
+    {
+      "epoch": 1.0,
+      "eval_loss": 2.295131206512451,
+      "eval_runtime": 231.1933,
+      "eval_samples_per_second": 37.514,
+      "eval_steps_per_second": 0.588,
+      "step": 540
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 1.4268800000000003e-05,
+      "loss": 2.3097,
+      "step": 541
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 1.4265600000000001e-05,
+      "loss": 2.3382,
+      "step": 542
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 1.4262400000000003e-05,
+      "loss": 3.3778,
+      "step": 543
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 1.4259200000000002e-05,
+      "loss": 2.1679,
+      "step": 544
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 1.4256000000000002e-05,
+      "loss": 2.2124,
+      "step": 545
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 1.4252800000000004e-05,
+      "loss": 2.2842,
+      "step": 546
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 1.4249600000000002e-05,
+      "loss": 2.2283,
+      "step": 547
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 1.4246400000000001e-05,
+      "loss": 2.2798,
+      "step": 548
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 1.4243200000000003e-05,
+      "loss": 2.1303,
+      "step": 549
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 1.4240000000000003e-05,
+      "loss": 2.2269,
+      "step": 550
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 1.4236800000000002e-05,
+      "loss": 2.1186,
+      "step": 551
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 1.4233600000000002e-05,
+      "loss": 2.253,
+      "step": 552
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 1.4230400000000002e-05,
+      "loss": 2.2734,
+      "step": 553
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 1.4227200000000002e-05,
+      "loss": 2.2395,
+      "step": 554
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 1.4224000000000003e-05,
+      "loss": 2.131,
+      "step": 555
+    },
+    {
+      "epoch": 1.03,
+      "learning_rate": 1.4220800000000003e-05,
+      "loss": 2.1922,
+      "step": 556
+    },
+    {
+      "epoch": 1.03,
+      "learning_rate": 1.4217600000000001e-05,
+      "loss": 2.2008,
+      "step": 557
+    },
+    {
+      "epoch": 1.03,
+      "learning_rate": 1.4214400000000002e-05,
+      "loss": 2.1084,
+      "step": 558
+    },
+    {
+      "epoch": 1.03,
+      "learning_rate": 1.4211200000000002e-05,
+      "loss": 2.2636,
+      "step": 559
+    },
+    {
+      "epoch": 1.03,
+      "learning_rate": 1.4208000000000002e-05,
+      "loss": 2.1628,
+      "step": 560
+    },
+    {
+      "epoch": 1.04,
+      "learning_rate": 1.4204800000000003e-05,
+      "loss": 2.2075,
+      "step": 561
+    },
+    {
+      "epoch": 1.04,
+      "learning_rate": 1.4201600000000001e-05,
+      "loss": 2.2908,
+      "step": 562
+    },
+    {
+      "epoch": 1.04,
+      "learning_rate": 1.4198400000000001e-05,
+      "loss": 2.2523,
+      "step": 563
+    },
+    {
+      "epoch": 1.04,
+      "learning_rate": 1.4195200000000003e-05,
+      "loss": 2.2526,
+      "step": 564
+    },
+    {
+      "epoch": 1.04,
+      "learning_rate": 1.4192000000000002e-05,
+      "loss": 2.2685,
+      "step": 565
+    },
+    {
+      "epoch": 1.04,
+      "learning_rate": 1.4188800000000002e-05,
+      "loss": 2.2141,
+      "step": 566
+    },
+    {
+      "epoch": 1.05,
+      "learning_rate": 1.4185600000000004e-05,
+      "loss": 2.1154,
+      "step": 567
+    },
+    {
+      "epoch": 1.05,
+      "learning_rate": 1.4182400000000002e-05,
+      "loss": 2.2448,
+      "step": 568
+    },
+    {
+      "epoch": 1.05,
+      "learning_rate": 1.4179200000000001e-05,
+      "loss": 2.2607,
+      "step": 569
+    },
+    {
+      "epoch": 1.05,
+      "learning_rate": 1.4176000000000003e-05,
+      "loss": 2.1543,
+      "step": 570
+    },
+    {
+      "epoch": 1.05,
+      "eval_loss": 2.2973735332489014,
+      "eval_runtime": 232.1815,
+      "eval_samples_per_second": 37.354,
+      "eval_steps_per_second": 0.586,
+      "step": 570
+    },
+    {
+      "epoch": 1.05,
+      "learning_rate": 1.4172800000000003e-05,
+      "loss": 2.2029,
+      "step": 571
+    },
+    {
+      "epoch": 1.06,
+      "learning_rate": 1.4169600000000004e-05,
+      "loss": 2.2319,
+      "step": 572
+    },
+    {
+      "epoch": 1.06,
+      "learning_rate": 1.4166400000000002e-05,
+      "loss": 2.2642,
+      "step": 573
+    },
+    {
+      "epoch": 1.06,
+      "learning_rate": 1.4163200000000002e-05,
+      "loss": 2.1439,
+      "step": 574
+    },
+    {
+      "epoch": 1.06,
+      "learning_rate": 1.4160000000000003e-05,
+      "loss": 2.2537,
+      "step": 575
+    },
+    {
+      "epoch": 1.06,
+      "learning_rate": 1.4156800000000003e-05,
+      "loss": 2.1039,
+      "step": 576
+    },
+    {
+      "epoch": 1.06,
+      "learning_rate": 1.4153600000000003e-05,
+      "loss": 2.1077,
+      "step": 577
+    },
+    {
+      "epoch": 1.07,
+      "learning_rate": 1.4150400000000003e-05,
+      "loss": 2.2256,
+      "step": 578
+    },
+    {
+      "epoch": 1.07,
+      "learning_rate": 1.4147200000000002e-05,
+      "loss": 2.1035,
+      "step": 579
+    },
+    {
+      "epoch": 1.07,
+      "learning_rate": 1.4144000000000002e-05,
+      "loss": 2.3031,
+      "step": 580
+    },
+    {
+      "epoch": 1.07,
+      "learning_rate": 1.4140800000000004e-05,
+      "loss": 2.1495,
+      "step": 581
+    },
+    {
+      "epoch": 1.07,
+      "learning_rate": 1.4137600000000003e-05,
+      "loss": 2.2456,
+      "step": 582
+    },
+    {
+      "epoch": 1.08,
+      "learning_rate": 1.4134400000000001e-05,
+      "loss": 2.24,
+      "step": 583
+    },
+    {
+      "epoch": 1.08,
+      "learning_rate": 1.4131200000000003e-05,
+      "loss": 2.2041,
+      "step": 584
+    },
+    {
+      "epoch": 1.08,
+      "learning_rate": 1.4128000000000003e-05,
+      "loss": 2.313,
+      "step": 585
+    },
+    {
+      "epoch": 1.08,
+      "learning_rate": 1.4124800000000002e-05,
+      "loss": 2.1506,
+      "step": 586
+    },
+    {
+      "epoch": 1.08,
+      "learning_rate": 1.4121600000000004e-05,
+      "loss": 2.2807,
+      "step": 587
+    },
+    {
+      "epoch": 1.08,
+      "learning_rate": 1.4118400000000002e-05,
+      "loss": 2.2544,
+      "step": 588
+    },
+    {
+      "epoch": 1.09,
+      "learning_rate": 1.4115200000000002e-05,
+      "loss": 2.2138,
+      "step": 589
+    },
+    {
+      "epoch": 1.09,
+      "learning_rate": 1.4112000000000003e-05,
+      "loss": 2.274,
+      "step": 590
+    },
+    {
+      "epoch": 1.09,
+      "learning_rate": 1.4108800000000003e-05,
+      "loss": 2.2029,
+      "step": 591
+    },
+    {
+      "epoch": 1.09,
+      "learning_rate": 1.4105600000000003e-05,
+      "loss": 2.145,
+      "step": 592
+    },
+    {
+      "epoch": 1.09,
+      "learning_rate": 1.4102400000000002e-05,
+      "loss": 2.1291,
+      "step": 593
+    },
+    {
+      "epoch": 1.1,
+      "learning_rate": 1.4099200000000002e-05,
+      "loss": 2.2327,
+      "step": 594
+    },
+    {
+      "epoch": 1.1,
+      "learning_rate": 1.4096000000000002e-05,
+      "loss": 2.2409,
+      "step": 595
+    },
+    {
+      "epoch": 1.1,
+      "learning_rate": 1.4092800000000003e-05,
+      "loss": 2.1623,
+      "step": 596
+    },
+    {
+      "epoch": 1.1,
+      "learning_rate": 1.4089600000000003e-05,
+      "loss": 2.2818,
+      "step": 597
+    },
+    {
+      "epoch": 1.1,
+      "learning_rate": 1.4086400000000001e-05,
+      "loss": 2.2654,
+      "step": 598
+    },
+    {
+      "epoch": 1.11,
+      "learning_rate": 1.4083200000000003e-05,
+      "loss": 2.2198,
+      "step": 599
+    },
+    {
+      "epoch": 1.11,
+      "learning_rate": 1.4080000000000002e-05,
+      "loss": 2.2933,
+      "step": 600
+    },
+    {
+      "epoch": 1.11,
+      "eval_loss": 2.2985422611236572,
+      "eval_runtime": 232.03,
+      "eval_samples_per_second": 37.379,
+      "eval_steps_per_second": 0.586,
+      "step": 600
+    },
+    {
+      "epoch": 1.11,
+      "learning_rate": 1.4076800000000002e-05,
+      "loss": 2.1996,
+      "step": 601
+    },
+    {
+      "epoch": 1.11,
+      "learning_rate": 1.4073600000000004e-05,
+      "loss": 2.1538,
+      "step": 602
+    },
+    {
+      "epoch": 1.11,
+      "learning_rate": 1.4070400000000002e-05,
+      "loss": 2.2071,
+      "step": 603
+    },
+    {
+      "epoch": 1.11,
+      "learning_rate": 1.4067200000000001e-05,
+      "loss": 2.1324,
+      "step": 604
+    },
+    {
+      "epoch": 1.12,
+      "learning_rate": 1.4064000000000003e-05,
+      "loss": 2.1835,
+      "step": 605
+    },
+    {
+      "epoch": 1.12,
+      "learning_rate": 1.4060800000000003e-05,
+      "loss": 2.2351,
+      "step": 606
+    },
+    {
+      "epoch": 1.12,
+      "learning_rate": 1.4057600000000002e-05,
+      "loss": 2.1257,
+      "step": 607
+    },
+    {
+      "epoch": 1.12,
+      "learning_rate": 1.4054400000000002e-05,
+      "loss": 2.2888,
+      "step": 608
+    },
+    {
+      "epoch": 1.12,
+      "learning_rate": 1.4051200000000002e-05,
+      "loss": 2.1535,
+      "step": 609
+    },
+    {
+      "epoch": 1.13,
+      "learning_rate": 1.4048000000000002e-05,
+      "loss": 2.2084,
+      "step": 610
+    },
+    {
+      "epoch": 1.13,
+      "learning_rate": 1.4044800000000003e-05,
+      "loss": 2.3088,
+      "step": 611
+    },
+    {
+      "epoch": 1.13,
+      "learning_rate": 1.4041600000000003e-05,
+      "loss": 2.2073,
+      "step": 612
+    },
+    {
+      "epoch": 1.13,
+      "learning_rate": 1.4038400000000001e-05,
+      "loss": 2.1326,
+      "step": 613
+    },
+    {
+      "epoch": 1.13,
+      "learning_rate": 1.4035200000000002e-05,
+      "loss": 2.2388,
+      "step": 614
+    },
+    {
+      "epoch": 1.13,
+      "learning_rate": 1.4032000000000002e-05,
+      "loss": 2.2349,
+      "step": 615
+    },
+    {
+      "epoch": 1.14,
+      "learning_rate": 1.4028800000000002e-05,
+      "loss": 2.2218,
+      "step": 616
+    },
+    {
+      "epoch": 1.14,
+      "learning_rate": 1.4025600000000003e-05,
+      "loss": 2.2542,
+      "step": 617
+    },
+    {
+      "epoch": 1.14,
+      "learning_rate": 1.4022400000000001e-05,
+      "loss": 2.2869,
+      "step": 618
+    },
+    {
+      "epoch": 1.14,
+      "learning_rate": 1.4019200000000001e-05,
+      "loss": 2.2315,
+      "step": 619
+    },
+    {
+      "epoch": 1.14,
+      "learning_rate": 1.4016000000000003e-05,
+      "loss": 2.1503,
+      "step": 620
+    },
+    {
+      "epoch": 1.15,
+      "learning_rate": 1.4012800000000002e-05,
+      "loss": 2.2251,
+      "step": 621
+    },
+    {
+      "epoch": 1.15,
+      "learning_rate": 1.4009600000000002e-05,
+      "loss": 2.2728,
+      "step": 622
+    },
+    {
+      "epoch": 1.15,
+      "learning_rate": 1.4006400000000002e-05,
+      "loss": 2.3338,
+      "step": 623
+    },
+    {
+      "epoch": 1.15,
+      "learning_rate": 1.4003200000000002e-05,
+      "loss": 2.3199,
+      "step": 624
+    },
+    {
+      "epoch": 1.15,
+      "learning_rate": 1.4000000000000001e-05,
+      "loss": 2.2945,
+      "step": 625
+    },
+    {
+      "epoch": 1.15,
+      "learning_rate": 1.3996800000000003e-05,
+      "loss": 2.2805,
+      "step": 626
+    },
+    {
+      "epoch": 1.16,
+      "learning_rate": 1.3993600000000003e-05,
+      "loss": 2.2963,
+      "step": 627
+    },
+    {
+      "epoch": 1.16,
+      "learning_rate": 1.3990400000000002e-05,
+      "loss": 2.2197,
+      "step": 628
+    },
+    {
+      "epoch": 1.16,
+      "learning_rate": 1.3987200000000002e-05,
+      "loss": 2.2485,
+      "step": 629
+    },
+    {
+      "epoch": 1.16,
+      "learning_rate": 1.3984000000000002e-05,
+      "loss": 2.1228,
+      "step": 630
+    },
+    {
+      "epoch": 1.16,
+      "eval_loss": 2.293792963027954,
+      "eval_runtime": 232.7454,
+      "eval_samples_per_second": 37.264,
+      "eval_steps_per_second": 0.584,
+      "step": 630
+    },
+    {
+      "epoch": 1.16,
+      "learning_rate": 1.3980800000000003e-05,
+      "loss": 2.2468,
+      "step": 631
+    },
+    {
+      "epoch": 1.17,
+      "learning_rate": 1.3977600000000003e-05,
+      "loss": 2.2749,
+      "step": 632
+    },
+    {
+      "epoch": 1.17,
+      "learning_rate": 1.3974400000000001e-05,
+      "loss": 2.2396,
+      "step": 633
+    },
+    {
+      "epoch": 1.17,
+      "learning_rate": 1.3971200000000003e-05,
+      "loss": 2.1842,
+      "step": 634
+    },
+    {
+      "epoch": 1.17,
+      "learning_rate": 1.3968000000000002e-05,
+      "loss": 2.2384,
+      "step": 635
+    },
+    {
+      "epoch": 1.17,
+      "learning_rate": 1.3964800000000002e-05,
+      "loss": 2.2866,
+      "step": 636
+    },
+    {
+      "epoch": 1.18,
+      "learning_rate": 1.3961600000000004e-05,
+      "loss": 2.2424,
+      "step": 637
+    },
+    {
+      "epoch": 1.18,
+      "learning_rate": 1.3958400000000002e-05,
+      "loss": 2.1828,
+      "step": 638
+    },
+    {
+      "epoch": 1.18,
+      "learning_rate": 1.3955200000000002e-05,
+      "loss": 2.2638,
+      "step": 639
+    },
+    {
+      "epoch": 1.18,
+      "learning_rate": 1.3952000000000003e-05,
+      "loss": 2.2135,
+      "step": 640
+    },
+    {
+      "epoch": 1.18,
+      "learning_rate": 1.3948800000000003e-05,
+      "loss": 2.1659,
+      "step": 641
+    },
+    {
+      "epoch": 1.18,
+      "learning_rate": 1.3945600000000002e-05,
+      "loss": 2.1449,
+      "step": 642
+    },
+    {
+      "epoch": 1.19,
+      "learning_rate": 1.3942400000000002e-05,
+      "loss": 2.2163,
+      "step": 643
+    },
+    {
+      "epoch": 1.19,
+      "learning_rate": 1.3939200000000002e-05,
+      "loss": 2.275,
+      "step": 644
+    },
+    {
+      "epoch": 1.19,
+      "learning_rate": 1.3936000000000002e-05,
+      "loss": 2.2825,
+      "step": 645
+    },
+    {
+      "epoch": 1.19,
+      "learning_rate": 1.3932800000000003e-05,
+      "loss": 2.2113,
+      "step": 646
+    },
+    {
+      "epoch": 1.19,
+      "learning_rate": 1.3929600000000003e-05,
+      "loss": 2.2232,
+      "step": 647
+    },
+    {
+      "epoch": 1.2,
+      "learning_rate": 1.3926400000000001e-05,
+      "loss": 2.2154,
+      "step": 648
+    },
+    {
+      "epoch": 1.2,
+      "learning_rate": 1.3923200000000002e-05,
+      "loss": 2.1691,
+      "step": 649
+    },
+    {
+      "epoch": 1.2,
+      "learning_rate": 1.3920000000000002e-05,
+      "loss": 2.2094,
+      "step": 650
+    },
+    {
+      "epoch": 1.2,
+      "learning_rate": 1.3916800000000002e-05,
+      "loss": 2.209,
+      "step": 651
+    },
+    {
+      "epoch": 1.2,
+      "learning_rate": 1.3913600000000003e-05,
+      "loss": 2.1444,
+      "step": 652
+    },
+    {
+      "epoch": 1.2,
+      "learning_rate": 1.3910400000000002e-05,
+      "loss": 2.2303,
+      "step": 653
+    },
+    {
+      "epoch": 1.21,
+      "learning_rate": 1.3907200000000001e-05,
+      "loss": 2.2567,
+      "step": 654
+    },
+    {
+      "epoch": 1.21,
+      "learning_rate": 1.3904000000000003e-05,
+      "loss": 2.3305,
+      "step": 655
+    },
+    {
+      "epoch": 1.21,
+      "learning_rate": 1.3900800000000003e-05,
+      "loss": 2.2562,
+      "step": 656
+    },
+    {
+      "epoch": 1.21,
+      "learning_rate": 1.3897600000000002e-05,
+      "loss": 2.2525,
+      "step": 657
+    },
+    {
+      "epoch": 1.21,
+      "learning_rate": 1.3894400000000002e-05,
+      "loss": 2.11,
+      "step": 658
+    },
+    {
+      "epoch": 1.22,
+      "learning_rate": 1.3891200000000002e-05,
+      "loss": 2.2009,
+      "step": 659
+    },
+    {
+      "epoch": 1.22,
+      "learning_rate": 1.3888000000000002e-05,
+      "loss": 2.2248,
+      "step": 660
+    },
+    {
+      "epoch": 1.22,
+      "eval_loss": 2.2939279079437256,
+      "eval_runtime": 232.3351,
+      "eval_samples_per_second": 37.33,
+      "eval_steps_per_second": 0.585,
+      "step": 660
+    },
+    {
+      "epoch": 1.22,
+      "learning_rate": 1.3884800000000003e-05,
+      "loss": 2.2822,
+      "step": 661
+    },
+    {
+      "epoch": 1.22,
+      "learning_rate": 1.3881600000000003e-05,
+      "loss": 2.1374,
+      "step": 662
+    },
+    {
+      "epoch": 1.22,
+      "learning_rate": 1.38784e-05,
+      "loss": 2.245,
+      "step": 663
+    },
+    {
+      "epoch": 1.22,
+      "learning_rate": 1.3875200000000002e-05,
+      "loss": 2.2539,
+      "step": 664
+    },
+    {
+      "epoch": 1.23,
+      "learning_rate": 1.3872000000000002e-05,
+      "loss": 2.1447,
+      "step": 665
+    },
+    {
+      "epoch": 1.23,
+      "learning_rate": 1.3868800000000002e-05,
+      "loss": 2.1927,
+      "step": 666
+    },
+    {
+      "epoch": 1.23,
+      "learning_rate": 1.3865600000000003e-05,
+      "loss": 2.2261,
+      "step": 667
+    },
+    {
+      "epoch": 1.23,
+      "learning_rate": 1.3862400000000001e-05,
+      "loss": 2.2315,
+      "step": 668
+    },
+    {
+      "epoch": 1.23,
+      "learning_rate": 1.3859200000000001e-05,
+      "loss": 2.1924,
+      "step": 669
+    },
+    {
+      "epoch": 1.24,
+      "learning_rate": 1.3856000000000003e-05,
+      "loss": 2.2366,
+      "step": 670
+    },
+    {
+      "epoch": 1.24,
+      "learning_rate": 1.3852800000000002e-05,
+      "loss": 2.2156,
+      "step": 671
+    },
+    {
+      "epoch": 1.24,
+      "learning_rate": 1.3849600000000002e-05,
+      "loss": 2.2028,
+      "step": 672
+    },
+    {
+      "epoch": 1.24,
+      "learning_rate": 1.3846400000000002e-05,
+      "loss": 2.2324,
+      "step": 673
+    },
+    {
+      "epoch": 1.24,
+      "learning_rate": 1.3843200000000002e-05,
+      "loss": 2.1853,
+      "step": 674
+    },
+    {
+      "epoch": 1.25,
+      "learning_rate": 1.3840000000000001e-05,
+      "loss": 2.1982,
+      "step": 675
+    },
+    {
+      "epoch": 1.25,
+      "learning_rate": 1.3836800000000003e-05,
+      "loss": 2.1962,
+      "step": 676
+    },
+    {
+      "epoch": 1.25,
+      "learning_rate": 1.3833600000000003e-05,
+      "loss": 2.2537,
+      "step": 677
+    },
+    {
+      "epoch": 1.25,
+      "learning_rate": 1.38304e-05,
+      "loss": 2.2566,
+      "step": 678
+    },
+    {
+      "epoch": 1.25,
+      "learning_rate": 1.3827200000000002e-05,
+      "loss": 2.0898,
+      "step": 679
+    },
+    {
+      "epoch": 1.25,
+      "learning_rate": 1.3824000000000002e-05,
+      "loss": 2.2167,
+      "step": 680
+    },
+    {
+      "epoch": 1.26,
+      "learning_rate": 1.3820800000000003e-05,
+      "loss": 2.2329,
+      "step": 681
+    },
+    {
+      "epoch": 1.26,
+      "learning_rate": 1.3817600000000003e-05,
+      "loss": 2.1995,
+      "step": 682
+    },
+    {
+      "epoch": 1.26,
+      "learning_rate": 1.3814400000000001e-05,
+      "loss": 2.233,
+      "step": 683
+    },
+    {
+      "epoch": 1.26,
+      "learning_rate": 1.3811200000000003e-05,
+      "loss": 2.2242,
+      "step": 684
+    },
+    {
+      "epoch": 1.26,
+      "learning_rate": 1.3808000000000002e-05,
+      "loss": 2.1919,
+      "step": 685
+    },
+    {
+      "epoch": 1.27,
+      "learning_rate": 1.3804800000000002e-05,
+      "loss": 2.1683,
+      "step": 686
+    },
+    {
+      "epoch": 1.27,
+      "learning_rate": 1.3801600000000004e-05,
+      "loss": 2.201,
+      "step": 687
+    },
+    {
+      "epoch": 1.27,
+      "learning_rate": 1.3798400000000003e-05,
+      "loss": 2.1782,
+      "step": 688
+    },
+    {
+      "epoch": 1.27,
+      "learning_rate": 1.3795200000000001e-05,
+      "loss": 2.226,
+      "step": 689
+    },
+    {
+      "epoch": 1.27,
+      "learning_rate": 1.3792000000000003e-05,
+      "loss": 2.247,
+      "step": 690
+    },
+    {
+      "epoch": 1.27,
+      "eval_loss": 2.295135021209717,
+      "eval_runtime": 231.2544,
+      "eval_samples_per_second": 37.504,
+      "eval_steps_per_second": 0.588,
+      "step": 690
+    },
+    {
+      "epoch": 1.27,
+      "learning_rate": 1.3788800000000003e-05,
+      "loss": 2.2268,
+      "step": 691
+    },
+    {
+      "epoch": 1.28,
+      "learning_rate": 1.3785600000000002e-05,
+      "loss": 2.1237,
+      "step": 692
+    },
+    {
+      "epoch": 1.28,
+      "learning_rate": 1.3782400000000004e-05,
+      "loss": 2.0994,
+      "step": 693
+    },
+    {
+      "epoch": 1.28,
+      "learning_rate": 1.3779200000000002e-05,
+      "loss": 2.1116,
+      "step": 694
+    },
+    {
+      "epoch": 1.28,
+      "learning_rate": 1.3776000000000002e-05,
+      "loss": 2.259,
+      "step": 695
+    },
+    {
+      "epoch": 1.28,
+      "learning_rate": 1.3772800000000003e-05,
+      "loss": 2.2245,
+      "step": 696
+    },
+    {
+      "epoch": 1.29,
+      "learning_rate": 1.3769600000000003e-05,
+      "loss": 2.2041,
+      "step": 697
+    },
+    {
+      "epoch": 1.29,
+      "learning_rate": 1.3766400000000003e-05,
+      "loss": 2.3184,
+      "step": 698
+    },
+    {
+      "epoch": 1.29,
+      "learning_rate": 1.3763200000000002e-05,
+      "loss": 2.1559,
+      "step": 699
+    },
+    {
+      "epoch": 1.29,
+      "learning_rate": 1.3760000000000002e-05,
+      "loss": 2.211,
+      "step": 700
+    },
+    {
+      "epoch": 1.29,
+      "learning_rate": 1.3756800000000002e-05,
+      "loss": 2.2464,
+      "step": 701
+    },
+    {
+      "epoch": 1.29,
+      "learning_rate": 1.3753600000000003e-05,
+      "loss": 2.2338,
+      "step": 702
+    },
+    {
+      "epoch": 1.3,
+      "learning_rate": 1.3750400000000003e-05,
+      "loss": 2.2037,
+      "step": 703
+    },
+    {
+      "epoch": 1.3,
+      "learning_rate": 1.3747200000000001e-05,
+      "loss": 2.2029,
+      "step": 704
+    },
+    {
+      "epoch": 1.3,
+      "learning_rate": 1.3744000000000003e-05,
+      "loss": 2.2065,
+      "step": 705
+    },
+    {
+      "epoch": 1.3,
+      "learning_rate": 1.3740800000000002e-05,
+      "loss": 2.1854,
+      "step": 706
+    },
+    {
+      "epoch": 1.3,
+      "learning_rate": 1.3737600000000002e-05,
+      "loss": 2.3513,
+      "step": 707
+    },
+    {
+      "epoch": 1.31,
+      "learning_rate": 1.3734400000000004e-05,
+      "loss": 2.2111,
+      "step": 708
+    },
+    {
+      "epoch": 1.31,
+      "learning_rate": 1.3731200000000002e-05,
+      "loss": 2.1945,
+      "step": 709
+    },
+    {
+      "epoch": 1.31,
+      "learning_rate": 1.3728000000000001e-05,
+      "loss": 2.1719,
+      "step": 710
+    },
+    {
+      "epoch": 1.31,
+      "learning_rate": 1.3724800000000003e-05,
+      "loss": 2.1693,
+      "step": 711
+    },
+    {
+      "epoch": 1.31,
+      "learning_rate": 1.3721600000000003e-05,
+      "loss": 2.2497,
+      "step": 712
+    },
+    {
+      "epoch": 1.32,
+      "learning_rate": 1.3718400000000002e-05,
+      "loss": 2.3514,
+      "step": 713
+    },
+    {
+      "epoch": 1.32,
+      "learning_rate": 1.3715200000000002e-05,
+      "loss": 2.1543,
+      "step": 714
+    },
+    {
+      "epoch": 1.32,
+      "learning_rate": 1.3712000000000002e-05,
+      "loss": 2.2005,
+      "step": 715
+    },
+    {
+      "epoch": 1.32,
+      "learning_rate": 1.3708800000000002e-05,
+      "loss": 2.2186,
+      "step": 716
+    },
+    {
+      "epoch": 1.32,
+      "learning_rate": 1.3705600000000003e-05,
+      "loss": 2.1782,
+      "step": 717
+    },
+    {
+      "epoch": 1.32,
+      "learning_rate": 1.3702400000000003e-05,
+      "loss": 2.2157,
+      "step": 718
+    },
+    {
+      "epoch": 1.33,
+      "learning_rate": 1.3699200000000001e-05,
+      "loss": 2.2445,
+      "step": 719
+    },
+    {
+      "epoch": 1.33,
+      "learning_rate": 1.3696000000000002e-05,
+      "loss": 2.2194,
+      "step": 720
+    },
+    {
+      "epoch": 1.33,
+      "eval_loss": 2.291602373123169,
+      "eval_runtime": 231.161,
+      "eval_samples_per_second": 37.519,
+      "eval_steps_per_second": 0.588,
+      "step": 720
+    },
+    {
+      "epoch": 1.33,
+      "learning_rate": 1.3692800000000002e-05,
+      "loss": 2.1957,
+      "step": 721
+    },
+    {
+      "epoch": 1.33,
+      "learning_rate": 1.3689600000000002e-05,
+      "loss": 2.2394,
+      "step": 722
+    },
+    {
+      "epoch": 1.33,
+      "learning_rate": 1.3686400000000003e-05,
+      "loss": 2.2511,
+      "step": 723
+    },
+    {
+      "epoch": 1.34,
+      "learning_rate": 1.3683200000000001e-05,
+      "loss": 2.3226,
+      "step": 724
+    },
+    {
+      "epoch": 1.34,
+      "learning_rate": 1.3680000000000001e-05,
+      "loss": 2.1522,
+      "step": 725
+    },
+    {
+      "epoch": 1.34,
+      "learning_rate": 1.3676800000000003e-05,
+      "loss": 2.2468,
+      "step": 726
+    },
+    {
+      "epoch": 1.34,
+      "learning_rate": 1.3673600000000002e-05,
+      "loss": 2.2184,
+      "step": 727
+    },
+    {
+      "epoch": 1.34,
+      "learning_rate": 1.3670400000000004e-05,
+      "loss": 2.2568,
+      "step": 728
+    },
+    {
+      "epoch": 1.34,
+      "learning_rate": 1.3667200000000002e-05,
+      "loss": 2.2545,
+      "step": 729
+    },
+    {
+      "epoch": 1.35,
+      "learning_rate": 1.3664000000000002e-05,
+      "loss": 2.2391,
+      "step": 730
+    },
+    {
+      "epoch": 1.35,
+      "learning_rate": 1.3660800000000003e-05,
+      "loss": 2.1568,
+      "step": 731
+    },
+    {
+      "epoch": 1.35,
+      "learning_rate": 1.3657600000000003e-05,
+      "loss": 2.2407,
+      "step": 732
+    },
+    {
+      "epoch": 1.35,
+      "learning_rate": 1.3654400000000003e-05,
+      "loss": 2.1564,
+      "step": 733
+    },
+    {
+      "epoch": 1.35,
+      "learning_rate": 1.3651200000000002e-05,
+      "loss": 2.2525,
+      "step": 734
+    },
+    {
+      "epoch": 1.36,
+      "learning_rate": 1.3648000000000002e-05,
+      "loss": 2.2041,
+      "step": 735
+    },
+    {
+      "epoch": 1.36,
+      "learning_rate": 1.3644800000000002e-05,
+      "loss": 2.2408,
+      "step": 736
+    },
+    {
+      "epoch": 1.36,
+      "learning_rate": 1.3641600000000003e-05,
+      "loss": 2.1354,
+      "step": 737
+    },
+    {
+      "epoch": 1.36,
+      "learning_rate": 1.3638400000000003e-05,
+      "loss": 2.2037,
+      "step": 738
+    },
+    {
+      "epoch": 1.36,
+      "learning_rate": 1.3635200000000001e-05,
+      "loss": 2.2109,
+      "step": 739
+    },
+    {
+      "epoch": 1.36,
+      "learning_rate": 1.3632000000000003e-05,
+      "loss": 2.1843,
+      "step": 740
+    },
+    {
+      "epoch": 1.37,
+      "learning_rate": 1.3628800000000002e-05,
+      "loss": 2.1899,
+      "step": 741
+    },
+    {
+      "epoch": 1.37,
+      "learning_rate": 1.3625600000000002e-05,
+      "loss": 2.2673,
+      "step": 742
+    },
+    {
+      "epoch": 1.37,
+      "learning_rate": 1.3622400000000004e-05,
+      "loss": 2.1802,
+      "step": 743
+    },
+    {
+      "epoch": 1.37,
+      "learning_rate": 1.3619200000000002e-05,
+      "loss": 2.1244,
+      "step": 744
+    },
+    {
+      "epoch": 1.37,
+      "learning_rate": 1.3616000000000001e-05,
+      "loss": 2.1923,
+      "step": 745
+    },
+    {
+      "epoch": 1.38,
+      "learning_rate": 1.3612800000000003e-05,
+      "loss": 2.3065,
+      "step": 746
+    },
+    {
+      "epoch": 1.38,
+      "learning_rate": 1.3609600000000003e-05,
+      "loss": 2.1695,
+      "step": 747
+    },
+    {
+      "epoch": 1.38,
+      "learning_rate": 1.3606400000000002e-05,
+      "loss": 2.1747,
+      "step": 748
+    },
+    {
+      "epoch": 1.38,
+      "learning_rate": 1.3603200000000002e-05,
+      "loss": 2.2013,
+      "step": 749
+    },
+    {
+      "epoch": 1.38,
+      "learning_rate": 1.3600000000000002e-05,
+      "loss": 2.2766,
+      "step": 750
+    },
+    {
+      "epoch": 1.38,
+      "eval_loss": 2.2927091121673584,
+      "eval_runtime": 232.4744,
+      "eval_samples_per_second": 37.307,
+      "eval_steps_per_second": 0.585,
+      "step": 750
+    },
+    {
+      "epoch": 1.39,
+      "learning_rate": 1.3596800000000002e-05,
+      "loss": 2.2106,
+      "step": 751
+    },
+    {
+      "epoch": 1.39,
+      "learning_rate": 1.3593600000000003e-05,
+      "loss": 2.2175,
+      "step": 752
+    },
+    {
+      "epoch": 1.39,
+      "learning_rate": 1.3590400000000003e-05,
+      "loss": 2.2404,
+      "step": 753
+    },
+    {
+      "epoch": 1.39,
+      "learning_rate": 1.3587200000000001e-05,
+      "loss": 2.3086,
+      "step": 754
+    },
+    {
+      "epoch": 1.39,
+      "learning_rate": 1.3584000000000002e-05,
+      "loss": 2.2018,
+      "step": 755
+    },
+    {
+      "epoch": 1.39,
+      "learning_rate": 1.3580800000000002e-05,
+      "loss": 2.104,
+      "step": 756
+    },
+    {
+      "epoch": 1.4,
+      "learning_rate": 1.3577600000000002e-05,
+      "loss": 2.257,
+      "step": 757
+    },
+    {
+      "epoch": 1.4,
+      "learning_rate": 1.3574400000000003e-05,
+      "loss": 2.1499,
+      "step": 758
+    },
+    {
+      "epoch": 1.4,
+      "learning_rate": 1.3571200000000001e-05,
+      "loss": 2.2061,
+      "step": 759
+    },
+    {
+      "epoch": 1.4,
+      "learning_rate": 1.3568000000000001e-05,
+      "loss": 2.285,
+      "step": 760
+    },
+    {
+      "epoch": 1.4,
+      "learning_rate": 1.3564800000000003e-05,
+      "loss": 2.1288,
+      "step": 761
+    },
+    {
+      "epoch": 1.41,
+      "learning_rate": 1.3561600000000002e-05,
+      "loss": 2.2469,
+      "step": 762
+    },
+    {
+      "epoch": 1.41,
+      "learning_rate": 1.3558400000000002e-05,
+      "loss": 2.175,
+      "step": 763
+    },
+    {
+      "epoch": 1.41,
+      "learning_rate": 1.3555200000000002e-05,
+      "loss": 2.1991,
+      "step": 764
+    },
+    {
+      "epoch": 1.41,
+      "learning_rate": 1.3552000000000002e-05,
+      "loss": 2.2712,
+      "step": 765
+    },
+    {
+      "epoch": 1.41,
+      "learning_rate": 1.3548800000000001e-05,
+      "loss": 2.246,
+      "step": 766
+    },
+    {
+      "epoch": 1.41,
+      "learning_rate": 1.3545600000000003e-05,
+      "loss": 2.2732,
+      "step": 767
+    },
+    {
+      "epoch": 1.42,
+      "learning_rate": 1.3542400000000003e-05,
+      "loss": 2.2268,
+      "step": 768
+    },
+    {
+      "epoch": 1.42,
+      "learning_rate": 1.35392e-05,
+      "loss": 2.1733,
+      "step": 769
+    },
+    {
+      "epoch": 1.42,
+      "learning_rate": 1.3536000000000002e-05,
+      "loss": 2.1803,
+      "step": 770
+    },
+    {
+      "epoch": 1.42,
+      "learning_rate": 1.3532800000000002e-05,
+      "loss": 2.1591,
+      "step": 771
+    },
+    {
+      "epoch": 1.42,
+      "learning_rate": 1.3529600000000002e-05,
+      "loss": 2.229,
+      "step": 772
+    },
+    {
+      "epoch": 1.43,
+      "learning_rate": 1.3526400000000003e-05,
+      "loss": 2.2009,
+      "step": 773
+    },
+    {
+      "epoch": 1.43,
+      "learning_rate": 1.3523200000000001e-05,
+      "loss": 2.1682,
+      "step": 774
+    },
+    {
+      "epoch": 1.43,
+      "learning_rate": 1.3520000000000001e-05,
+      "loss": 2.2017,
+      "step": 775
+    },
+    {
+      "epoch": 1.43,
+      "learning_rate": 1.3516800000000002e-05,
+      "loss": 2.1668,
+      "step": 776
+    },
+    {
+      "epoch": 1.43,
+      "learning_rate": 1.3513600000000002e-05,
+      "loss": 2.2036,
+      "step": 777
+    },
+    {
+      "epoch": 1.44,
+      "learning_rate": 1.3510400000000004e-05,
+      "loss": 2.197,
+      "step": 778
+    },
+    {
+      "epoch": 1.44,
+      "learning_rate": 1.3507200000000002e-05,
+      "loss": 2.2089,
+      "step": 779
+    },
+    {
+      "epoch": 1.44,
+      "learning_rate": 1.3504000000000001e-05,
+      "loss": 2.3042,
+      "step": 780
+    },
+    {
+      "epoch": 1.44,
+      "eval_loss": 2.290692090988159,
+      "eval_runtime": 232.3543,
+      "eval_samples_per_second": 37.327,
+      "eval_steps_per_second": 0.585,
+      "step": 780
+    },
+    {
+      "epoch": 1.44,
+      "learning_rate": 1.3500800000000003e-05,
+      "loss": 2.0559,
+      "step": 781
+    },
+    {
+      "epoch": 1.44,
+      "learning_rate": 1.3497600000000003e-05,
+      "loss": 2.352,
+      "step": 782
+    },
+    {
+      "epoch": 1.44,
+      "learning_rate": 1.3494400000000002e-05,
+      "loss": 2.2146,
+      "step": 783
+    },
+    {
+      "epoch": 1.45,
+      "learning_rate": 1.3491200000000002e-05,
+      "loss": 2.2421,
+      "step": 784
+    },
+    {
+      "epoch": 1.45,
+      "learning_rate": 1.3488000000000002e-05,
+      "loss": 2.258,
+      "step": 785
+    },
+    {
+      "epoch": 1.45,
+      "learning_rate": 1.3484800000000002e-05,
+      "loss": 2.1495,
+      "step": 786
+    },
+    {
+      "epoch": 1.45,
+      "learning_rate": 1.3481600000000003e-05,
+      "loss": 2.205,
+      "step": 787
+    },
+    {
+      "epoch": 1.45,
+      "learning_rate": 1.3478400000000003e-05,
+      "loss": 2.2532,
+      "step": 788
+    },
+    {
+      "epoch": 1.46,
+      "learning_rate": 1.3475200000000001e-05,
+      "loss": 2.1342,
+      "step": 789
+    },
+    {
+      "epoch": 1.46,
+      "learning_rate": 1.3472000000000002e-05,
+      "loss": 2.2042,
+      "step": 790
+    },
+    {
+      "epoch": 1.46,
+      "learning_rate": 1.3468800000000002e-05,
+      "loss": 2.1637,
+      "step": 791
+    },
+    {
+      "epoch": 1.46,
+      "learning_rate": 1.3465600000000002e-05,
+      "loss": 2.2893,
+      "step": 792
+    },
+    {
+      "epoch": 1.46,
+      "learning_rate": 1.3462400000000003e-05,
+      "loss": 2.1728,
+      "step": 793
+    },
+    {
+      "epoch": 1.46,
+      "learning_rate": 1.3459200000000002e-05,
+      "loss": 2.3226,
+      "step": 794
+    },
+    {
+      "epoch": 1.47,
+      "learning_rate": 1.3456000000000001e-05,
+      "loss": 2.1695,
+      "step": 795
+    },
+    {
+      "epoch": 1.47,
+      "learning_rate": 1.3452800000000003e-05,
+      "loss": 2.1702,
+      "step": 796
+    },
+    {
+      "epoch": 1.47,
+      "learning_rate": 1.3449600000000002e-05,
+      "loss": 2.2193,
+      "step": 797
+    },
+    {
+      "epoch": 1.47,
+      "learning_rate": 1.3446400000000002e-05,
+      "loss": 2.1648,
+      "step": 798
+    },
+    {
+      "epoch": 1.47,
+      "learning_rate": 1.3443200000000002e-05,
+      "loss": 2.2333,
+      "step": 799
+    },
+    {
+      "epoch": 1.48,
+      "learning_rate": 1.3440000000000002e-05,
+      "loss": 2.2221,
+      "step": 800
+    },
+    {
+      "epoch": 1.48,
+      "learning_rate": 1.3436800000000002e-05,
+      "loss": 2.2616,
+      "step": 801
+    },
+    {
+      "epoch": 1.48,
+      "learning_rate": 1.3433600000000003e-05,
+      "loss": 2.2275,
+      "step": 802
+    },
+    {
+      "epoch": 1.48,
+      "learning_rate": 1.3430400000000003e-05,
+      "loss": 2.2246,
+      "step": 803
+    },
+    {
+      "epoch": 1.48,
+      "learning_rate": 1.34272e-05,
+      "loss": 2.23,
+      "step": 804
+    },
+    {
+      "epoch": 1.48,
+      "learning_rate": 1.3424000000000002e-05,
+      "loss": 2.2456,
+      "step": 805
+    },
+    {
+      "epoch": 1.49,
+      "learning_rate": 1.3420800000000002e-05,
+      "loss": 2.2447,
+      "step": 806
+    },
+    {
+      "epoch": 1.49,
+      "learning_rate": 1.3417600000000002e-05,
+      "loss": 2.0901,
+      "step": 807
+    },
+    {
+      "epoch": 1.49,
+      "learning_rate": 1.3414400000000003e-05,
+      "loss": 2.2727,
+      "step": 808
+    },
+    {
+      "epoch": 1.49,
+      "learning_rate": 1.3411200000000001e-05,
+      "loss": 2.1999,
+      "step": 809
+    },
+    {
+      "epoch": 1.49,
+      "learning_rate": 1.3408000000000001e-05,
+      "loss": 2.2332,
+      "step": 810
+    },
+    {
+      "epoch": 1.49,
+      "eval_loss": 2.2890512943267822,
+      "eval_runtime": 232.7031,
+      "eval_samples_per_second": 37.271,
+      "eval_steps_per_second": 0.584,
+      "step": 810
+    },
+    {
+      "epoch": 1.5,
+      "learning_rate": 1.3404800000000003e-05,
+      "loss": 2.1639,
+      "step": 811
+    },
+    {
+      "epoch": 1.5,
+      "learning_rate": 1.3401600000000002e-05,
+      "loss": 2.3147,
+      "step": 812
+    },
+    {
+      "epoch": 1.5,
+      "learning_rate": 1.3398400000000002e-05,
+      "loss": 2.2364,
+      "step": 813
+    },
+    {
+      "epoch": 1.5,
+      "learning_rate": 1.3395200000000003e-05,
+      "loss": 2.0979,
+      "step": 814
+    },
+    {
+      "epoch": 1.5,
+      "learning_rate": 1.3392000000000002e-05,
+      "loss": 2.2811,
+      "step": 815
+    },
+    {
+      "epoch": 1.51,
+      "learning_rate": 1.3388800000000001e-05,
+      "loss": 2.1574,
+      "step": 816
+    },
+    {
+      "epoch": 1.51,
+      "learning_rate": 1.3385600000000003e-05,
+      "loss": 2.2406,
+      "step": 817
+    },
+    {
+      "epoch": 1.51,
+      "learning_rate": 1.3382400000000003e-05,
+      "loss": 2.2284,
+      "step": 818
+    },
+    {
+      "epoch": 1.51,
+      "learning_rate": 1.3379200000000002e-05,
+      "loss": 2.2907,
+      "step": 819
+    },
+    {
+      "epoch": 1.51,
+      "learning_rate": 1.3376000000000002e-05,
+      "loss": 2.2372,
+      "step": 820
+    },
+    {
+      "epoch": 1.51,
+      "learning_rate": 1.3372800000000002e-05,
+      "loss": 2.1867,
+      "step": 821
+    },
+    {
+      "epoch": 1.52,
+      "learning_rate": 1.3369600000000002e-05,
+      "loss": 2.1514,
+      "step": 822
+    },
+    {
+      "epoch": 1.52,
+      "learning_rate": 1.3366400000000003e-05,
+      "loss": 2.1364,
+      "step": 823
+    },
+    {
+      "epoch": 1.52,
+      "learning_rate": 1.3363200000000003e-05,
+      "loss": 2.2233,
+      "step": 824
+    },
+    {
+      "epoch": 1.52,
+      "learning_rate": 1.336e-05,
+      "loss": 2.1635,
+      "step": 825
+    },
+    {
+      "epoch": 1.52,
+      "learning_rate": 1.3356800000000002e-05,
+      "loss": 2.1504,
+      "step": 826
+    },
+    {
+      "epoch": 1.53,
+      "learning_rate": 1.3353600000000002e-05,
+      "loss": 2.2704,
+      "step": 827
+    },
+    {
+      "epoch": 1.53,
+      "learning_rate": 1.3350400000000004e-05,
+      "loss": 2.2571,
+      "step": 828
+    },
+    {
+      "epoch": 1.53,
+      "learning_rate": 1.3347200000000003e-05,
+      "loss": 2.173,
+      "step": 829
+    },
+    {
+      "epoch": 1.53,
+      "learning_rate": 1.3344000000000001e-05,
+      "loss": 2.2287,
+      "step": 830
+    },
+    {
+      "epoch": 1.53,
+      "learning_rate": 1.3340800000000003e-05,
+      "loss": 2.2252,
+      "step": 831
+    },
+    {
+      "epoch": 1.53,
+      "learning_rate": 1.3337600000000003e-05,
+      "loss": 2.1803,
+      "step": 832
+    },
+    {
+      "epoch": 1.54,
+      "learning_rate": 1.3334400000000002e-05,
+      "loss": 2.2247,
+      "step": 833
+    },
+    {
+      "epoch": 1.54,
+      "learning_rate": 1.3331200000000004e-05,
+      "loss": 2.1865,
+      "step": 834
+    },
+    {
+      "epoch": 1.54,
+      "learning_rate": 1.3328000000000002e-05,
+      "loss": 2.2003,
+      "step": 835
+    },
+    {
+      "epoch": 1.54,
+      "learning_rate": 1.3324800000000002e-05,
+      "loss": 2.2139,
+      "step": 836
+    },
+    {
+      "epoch": 1.54,
+      "learning_rate": 1.3321600000000003e-05,
+      "loss": 2.2272,
+      "step": 837
+    },
+    {
+      "epoch": 1.55,
+      "learning_rate": 1.3318400000000003e-05,
+      "loss": 2.2433,
+      "step": 838
+    },
+    {
+      "epoch": 1.55,
+      "learning_rate": 1.3315200000000003e-05,
+      "loss": 2.2217,
+      "step": 839
+    },
+    {
+      "epoch": 1.55,
+      "learning_rate": 1.3312000000000002e-05,
+      "loss": 2.2517,
+      "step": 840
+    },
+    {
+      "epoch": 1.55,
+      "eval_loss": 2.290472984313965,
+      "eval_runtime": 233.1729,
+      "eval_samples_per_second": 37.196,
+      "eval_steps_per_second": 0.583,
+      "step": 840
+    },
+    {
+      "epoch": 1.55,
+      "learning_rate": 1.3308800000000002e-05,
+      "loss": 2.2016,
+      "step": 841
+    },
+    {
+      "epoch": 1.55,
+      "learning_rate": 1.3305600000000002e-05,
+      "loss": 2.1796,
+      "step": 842
+    },
+    {
+      "epoch": 1.55,
+      "learning_rate": 1.3302400000000003e-05,
+      "loss": 2.1869,
+      "step": 843
+    },
+    {
+      "epoch": 1.56,
+      "learning_rate": 1.3299200000000003e-05,
+      "loss": 2.211,
+      "step": 844
+    },
+    {
+      "epoch": 1.56,
+      "learning_rate": 1.3296000000000001e-05,
+      "loss": 2.1856,
+      "step": 845
+    },
+    {
+      "epoch": 1.56,
+      "learning_rate": 1.3292800000000003e-05,
+      "loss": 2.1855,
+      "step": 846
+    },
+    {
+      "epoch": 1.56,
+      "learning_rate": 1.3289600000000002e-05,
+      "loss": 2.2168,
+      "step": 847
+    },
+    {
+      "epoch": 1.56,
+      "learning_rate": 1.3286400000000002e-05,
+      "loss": 2.2778,
+      "step": 848
+    },
+    {
+      "epoch": 1.57,
+      "learning_rate": 1.3283200000000004e-05,
+      "loss": 2.2143,
+      "step": 849
+    },
+    {
+      "epoch": 1.57,
+      "learning_rate": 1.3280000000000002e-05,
+      "loss": 2.1996,
+      "step": 850
+    },
+    {
+      "epoch": 1.57,
+      "learning_rate": 1.3276800000000001e-05,
+      "loss": 2.1432,
+      "step": 851
+    },
+    {
+      "epoch": 1.57,
+      "learning_rate": 1.3273600000000003e-05,
+      "loss": 2.1425,
+      "step": 852
+    },
+    {
+      "epoch": 1.57,
+      "learning_rate": 1.3270400000000003e-05,
+      "loss": 2.143,
+      "step": 853
+    },
+    {
+      "epoch": 1.58,
+      "learning_rate": 1.3267200000000002e-05,
+      "loss": 2.2994,
+      "step": 854
+    },
+    {
+      "epoch": 1.58,
+      "learning_rate": 1.3264000000000002e-05,
+      "loss": 2.3091,
+      "step": 855
+    },
+    {
+      "epoch": 1.58,
+      "learning_rate": 1.3260800000000002e-05,
+      "loss": 2.2643,
+      "step": 856
+    },
+    {
+      "epoch": 1.58,
+      "learning_rate": 1.3257600000000002e-05,
+      "loss": 2.2149,
+      "step": 857
+    },
+    {
+      "epoch": 1.58,
+      "learning_rate": 1.3254400000000003e-05,
+      "loss": 2.1398,
+      "step": 858
+    },
+    {
+      "epoch": 1.58,
+      "learning_rate": 1.3251200000000003e-05,
+      "loss": 2.1697,
+      "step": 859
+    },
+    {
+      "epoch": 1.59,
+      "learning_rate": 1.3248000000000001e-05,
+      "loss": 2.1524,
+      "step": 860
+    },
+    {
+      "epoch": 1.59,
+      "learning_rate": 1.3244800000000002e-05,
+      "loss": 2.1547,
+      "step": 861
+    },
+    {
+      "epoch": 1.59,
+      "learning_rate": 1.3241600000000002e-05,
+      "loss": 2.1862,
+      "step": 862
+    },
+    {
+      "epoch": 1.59,
+      "learning_rate": 1.3238400000000002e-05,
+      "loss": 2.1916,
+      "step": 863
+    },
+    {
+      "epoch": 1.59,
+      "learning_rate": 1.3235200000000003e-05,
+      "loss": 2.1848,
+      "step": 864
+    },
+    {
+      "epoch": 1.6,
+      "learning_rate": 1.3232000000000001e-05,
+      "loss": 2.3006,
+      "step": 865
+    },
+    {
+      "epoch": 1.6,
+      "learning_rate": 1.3228800000000001e-05,
+      "loss": 2.1699,
+      "step": 866
+    },
+    {
+      "epoch": 1.6,
+      "learning_rate": 1.3225600000000003e-05,
+      "loss": 2.275,
+      "step": 867
+    },
+    {
+      "epoch": 1.6,
+      "learning_rate": 1.3222400000000002e-05,
+      "loss": 2.2865,
+      "step": 868
+    },
+    {
+      "epoch": 1.6,
+      "learning_rate": 1.3219200000000002e-05,
+      "loss": 2.1369,
+      "step": 869
+    },
+    {
+      "epoch": 1.6,
+      "learning_rate": 1.3216000000000002e-05,
+      "loss": 2.3137,
+      "step": 870
+    },
+    {
+      "epoch": 1.6,
+      "eval_loss": 2.2871532440185547,
+      "eval_runtime": 231.787,
+      "eval_samples_per_second": 37.418,
+      "eval_steps_per_second": 0.587,
+      "step": 870
+    },
+    {
+      "epoch": 1.61,
+      "learning_rate": 1.3212800000000002e-05,
+      "loss": 2.1445,
+      "step": 871
+    },
+    {
+      "epoch": 1.61,
+      "learning_rate": 1.3209600000000001e-05,
+      "loss": 2.1793,
+      "step": 872
+    },
+    {
+      "epoch": 1.61,
+      "learning_rate": 1.3206400000000003e-05,
+      "loss": 2.1425,
+      "step": 873
+    },
+    {
+      "epoch": 1.61,
+      "learning_rate": 1.3203200000000003e-05,
+      "loss": 2.2903,
+      "step": 874
+    },
+    {
+      "epoch": 1.61,
+      "learning_rate": 1.32e-05,
+      "loss": 2.2041,
+      "step": 875
+    },
+    {
+      "epoch": 1.62,
+      "learning_rate": 1.3196800000000002e-05,
+      "loss": 2.2611,
+      "step": 876
+    },
+    {
+      "epoch": 1.62,
+      "learning_rate": 1.3193600000000002e-05,
+      "loss": 2.184,
+      "step": 877
+    },
+    {
+      "epoch": 1.62,
+      "learning_rate": 1.3190400000000003e-05,
+      "loss": 2.2165,
+      "step": 878
+    },
+    {
+      "epoch": 1.62,
+      "learning_rate": 1.3187200000000003e-05,
+      "loss": 2.2439,
+      "step": 879
+    },
+    {
+      "epoch": 1.62,
+      "learning_rate": 1.3184000000000001e-05,
+      "loss": 2.2086,
+      "step": 880
+    },
+    {
+      "epoch": 1.62,
+      "learning_rate": 1.3180800000000003e-05,
+      "loss": 2.1322,
+      "step": 881
+    },
+    {
+      "epoch": 1.63,
+      "learning_rate": 1.3177600000000002e-05,
+      "loss": 2.3432,
+      "step": 882
+    },
+    {
+      "epoch": 1.63,
+      "learning_rate": 1.3174400000000002e-05,
+      "loss": 2.2874,
+      "step": 883
+    },
+    {
+      "epoch": 1.63,
+      "learning_rate": 1.3171200000000004e-05,
+      "loss": 2.167,
+      "step": 884
+    },
+    {
+      "epoch": 1.63,
+      "learning_rate": 1.3168000000000002e-05,
+      "loss": 2.0649,
+      "step": 885
+    },
+    {
+      "epoch": 1.63,
+      "learning_rate": 1.3164800000000001e-05,
+      "loss": 2.2231,
+      "step": 886
+    },
+    {
+      "epoch": 1.64,
+      "learning_rate": 1.3161600000000003e-05,
+      "loss": 2.1965,
+      "step": 887
+    },
+    {
+      "epoch": 1.64,
+      "learning_rate": 1.3158400000000003e-05,
+      "loss": 2.2812,
+      "step": 888
+    },
+    {
+      "epoch": 1.64,
+      "learning_rate": 1.3155200000000002e-05,
+      "loss": 2.1843,
+      "step": 889
+    },
+    {
+      "epoch": 1.64,
+      "learning_rate": 1.3152000000000002e-05,
+      "loss": 2.2331,
+      "step": 890
+    },
+    {
+      "epoch": 1.64,
+      "learning_rate": 1.3148800000000002e-05,
+      "loss": 2.2006,
+      "step": 891
+    },
+    {
+      "epoch": 1.65,
+      "learning_rate": 1.3145600000000002e-05,
+      "loss": 2.1237,
+      "step": 892
+    },
+    {
+      "epoch": 1.65,
+      "learning_rate": 1.3142400000000003e-05,
+      "loss": 2.1617,
+      "step": 893
+    },
+    {
+      "epoch": 1.65,
+      "learning_rate": 1.3139200000000003e-05,
+      "loss": 2.209,
+      "step": 894
+    },
+    {
+      "epoch": 1.65,
+      "learning_rate": 1.3136000000000001e-05,
+      "loss": 2.1492,
+      "step": 895
+    },
+    {
+      "epoch": 1.65,
+      "learning_rate": 1.3132800000000002e-05,
+      "loss": 2.1822,
+      "step": 896
+    },
+    {
+      "epoch": 1.65,
+      "learning_rate": 1.3129600000000002e-05,
+      "loss": 2.0924,
+      "step": 897
+    },
+    {
+      "epoch": 1.66,
+      "learning_rate": 1.3126400000000002e-05,
+      "loss": 2.1911,
+      "step": 898
+    },
+    {
+      "epoch": 1.66,
+      "learning_rate": 1.3123200000000003e-05,
+      "loss": 2.2086,
+      "step": 899
+    },
+    {
+      "epoch": 1.66,
+      "learning_rate": 1.3120000000000001e-05,
+      "loss": 2.2103,
+      "step": 900
+    },
+    {
+      "epoch": 1.66,
+      "eval_loss": 2.2874975204467773,
+      "eval_runtime": 233.0765,
+      "eval_samples_per_second": 37.211,
+      "eval_steps_per_second": 0.583,
+      "step": 900
+    },
+    {
+      "epoch": 1.66,
+      "learning_rate": 1.3116800000000001e-05,
+      "loss": 2.2239,
+      "step": 901
+    },
+    {
+      "epoch": 1.66,
+      "learning_rate": 1.3113600000000003e-05,
+      "loss": 2.2331,
+      "step": 902
+    },
+    {
+      "epoch": 1.67,
+      "learning_rate": 1.3110400000000002e-05,
+      "loss": 2.2453,
+      "step": 903
+    },
+    {
+      "epoch": 1.67,
+      "learning_rate": 1.3107200000000002e-05,
+      "loss": 2.1865,
+      "step": 904
+    },
+    {
+      "epoch": 1.67,
+      "learning_rate": 1.3104000000000002e-05,
+      "loss": 2.2399,
+      "step": 905
+    },
+    {
+      "epoch": 1.67,
+      "learning_rate": 1.3100800000000002e-05,
+      "loss": 2.1638,
+      "step": 906
+    },
+    {
+      "epoch": 1.67,
+      "learning_rate": 1.3097600000000001e-05,
+      "loss": 2.2275,
+      "step": 907
+    },
+    {
+      "epoch": 1.67,
+      "learning_rate": 1.3094400000000003e-05,
+      "loss": 2.1852,
+      "step": 908
+    },
+    {
+      "epoch": 1.68,
+      "learning_rate": 1.3091200000000003e-05,
+      "loss": 2.152,
+      "step": 909
+    },
+    {
+      "epoch": 1.68,
+      "learning_rate": 1.3088e-05,
+      "loss": 2.2367,
+      "step": 910
+    },
+    {
+      "epoch": 1.68,
+      "learning_rate": 1.3084800000000002e-05,
+      "loss": 2.1437,
+      "step": 911
+    },
+    {
+      "epoch": 1.68,
+      "learning_rate": 1.3081600000000002e-05,
+      "loss": 2.2955,
+      "step": 912
+    },
+    {
+      "epoch": 1.68,
+      "learning_rate": 1.3078400000000002e-05,
+      "loss": 2.1651,
+      "step": 913
+    },
+    {
+      "epoch": 1.69,
+      "learning_rate": 1.3075200000000003e-05,
+      "loss": 2.193,
+      "step": 914
+    },
+    {
+      "epoch": 1.69,
+      "learning_rate": 1.3072000000000001e-05,
+      "loss": 2.2826,
+      "step": 915
+    },
+    {
+      "epoch": 1.69,
+      "learning_rate": 1.3068800000000001e-05,
+      "loss": 2.1961,
+      "step": 916
+    },
+    {
+      "epoch": 1.69,
+      "learning_rate": 1.3065600000000002e-05,
+      "loss": 2.2367,
+      "step": 917
+    },
+    {
+      "epoch": 1.69,
+      "learning_rate": 1.3062400000000002e-05,
+      "loss": 2.2198,
+      "step": 918
+    },
+    {
+      "epoch": 1.69,
+      "learning_rate": 1.3059200000000002e-05,
+      "loss": 2.0632,
+      "step": 919
+    },
+    {
+      "epoch": 1.7,
+      "learning_rate": 1.3056000000000002e-05,
+      "loss": 2.2379,
+      "step": 920
+    },
+    {
+      "epoch": 1.7,
+      "learning_rate": 1.3052800000000001e-05,
+      "loss": 2.2197,
+      "step": 921
+    },
+    {
+      "epoch": 1.7,
+      "learning_rate": 1.3049600000000001e-05,
+      "loss": 2.1991,
+      "step": 922
+    },
+    {
+      "epoch": 1.7,
+      "learning_rate": 1.3046400000000003e-05,
+      "loss": 2.2131,
+      "step": 923
+    },
+    {
+      "epoch": 1.7,
+      "learning_rate": 1.3043200000000002e-05,
+      "loss": 2.2076,
+      "step": 924
+    },
+    {
+      "epoch": 1.71,
+      "learning_rate": 1.304e-05,
+      "loss": 2.0854,
+      "step": 925
+    },
+    {
+      "epoch": 1.71,
+      "learning_rate": 1.3036800000000002e-05,
+      "loss": 2.2908,
+      "step": 926
+    },
+    {
+      "epoch": 1.71,
+      "learning_rate": 1.3033600000000002e-05,
+      "loss": 2.2069,
+      "step": 927
+    },
+    {
+      "epoch": 1.71,
+      "learning_rate": 1.3030400000000003e-05,
+      "loss": 2.2386,
+      "step": 928
+    },
+    {
+      "epoch": 1.71,
+      "learning_rate": 1.3027200000000003e-05,
+      "loss": 2.1883,
+      "step": 929
+    },
+    {
+      "epoch": 1.72,
+      "learning_rate": 1.3024000000000001e-05,
+      "loss": 2.1304,
+      "step": 930
+    },
+    {
+      "epoch": 1.72,
+      "eval_loss": 2.2889654636383057,
+      "eval_runtime": 231.3088,
+      "eval_samples_per_second": 37.495,
+      "eval_steps_per_second": 0.588,
+      "step": 930
+    },
+    {
+      "epoch": 1.72,
+      "learning_rate": 1.3020800000000002e-05,
+      "loss": 2.2113,
+      "step": 931
+    },
+    {
+      "epoch": 1.72,
+      "learning_rate": 1.3017600000000002e-05,
+      "loss": 2.2006,
+      "step": 932
+    },
+    {
+      "epoch": 1.72,
+      "learning_rate": 1.3014400000000002e-05,
+      "loss": 2.2494,
+      "step": 933
+    },
+    {
+      "epoch": 1.72,
+      "learning_rate": 1.3011200000000003e-05,
+      "loss": 2.2525,
+      "step": 934
+    },
+    {
+      "epoch": 1.72,
+      "learning_rate": 1.3008000000000001e-05,
+      "loss": 2.2487,
+      "step": 935
+    },
+    {
+      "epoch": 1.73,
+      "learning_rate": 1.3004800000000001e-05,
+      "loss": 2.1816,
+      "step": 936
+    },
+    {
+      "epoch": 1.73,
+      "learning_rate": 1.3001600000000003e-05,
+      "loss": 2.2001,
+      "step": 937
+    },
+    {
+      "epoch": 1.73,
+      "learning_rate": 1.2998400000000002e-05,
+      "loss": 2.175,
+      "step": 938
+    },
+    {
+      "epoch": 1.73,
+      "learning_rate": 1.2995200000000002e-05,
+      "loss": 2.2604,
+      "step": 939
+    },
+    {
+      "epoch": 1.73,
+      "learning_rate": 1.2992000000000004e-05,
+      "loss": 2.2139,
+      "step": 940
+    },
+    {
+      "epoch": 1.74,
+      "learning_rate": 1.2988800000000002e-05,
+      "loss": 2.1728,
+      "step": 941
+    },
+    {
+      "epoch": 1.74,
+      "learning_rate": 1.2985600000000002e-05,
+      "loss": 2.2463,
+      "step": 942
+    },
+    {
+      "epoch": 1.74,
+      "learning_rate": 1.2982400000000003e-05,
+      "loss": 2.2253,
+      "step": 943
+    },
+    {
+      "epoch": 1.74,
+      "learning_rate": 1.2979200000000003e-05,
+      "loss": 2.2221,
+      "step": 944
+    },
+    {
+      "epoch": 1.74,
+      "learning_rate": 1.2976000000000002e-05,
+      "loss": 2.1808,
+      "step": 945
+    },
+    {
+      "epoch": 1.74,
+      "learning_rate": 1.2972800000000002e-05,
+      "loss": 2.1562,
+      "step": 946
+    },
+    {
+      "epoch": 1.75,
+      "learning_rate": 1.2969600000000002e-05,
+      "loss": 2.2843,
+      "step": 947
+    },
+    {
+      "epoch": 1.75,
+      "learning_rate": 1.2966400000000002e-05,
+      "loss": 2.2758,
+      "step": 948
+    },
+    {
+      "epoch": 1.75,
+      "learning_rate": 1.2963200000000003e-05,
+      "loss": 2.2179,
+      "step": 949
+    },
+    {
+      "epoch": 1.75,
+      "learning_rate": 1.2960000000000003e-05,
+      "loss": 2.2914,
+      "step": 950
+    },
+    {
+      "epoch": 1.75,
+      "learning_rate": 1.2956800000000001e-05,
+      "loss": 2.1993,
+      "step": 951
+    },
+    {
+      "epoch": 1.76,
+      "learning_rate": 1.2953600000000002e-05,
+      "loss": 2.1964,
+      "step": 952
+    },
+    {
+      "epoch": 1.76,
+      "learning_rate": 1.2950400000000002e-05,
+      "loss": 2.1743,
+      "step": 953
+    },
+    {
+      "epoch": 1.76,
+      "learning_rate": 1.2947200000000002e-05,
+      "loss": 2.1836,
+      "step": 954
+    },
+    {
+      "epoch": 1.76,
+      "learning_rate": 1.2944000000000003e-05,
+      "loss": 2.2589,
+      "step": 955
+    },
+    {
+      "epoch": 1.76,
+      "learning_rate": 1.2940800000000002e-05,
+      "loss": 2.1612,
+      "step": 956
+    },
+    {
+      "epoch": 1.76,
+      "learning_rate": 1.2937600000000001e-05,
+      "loss": 2.2416,
+      "step": 957
+    },
+    {
+      "epoch": 1.77,
+      "learning_rate": 1.2934400000000003e-05,
+      "loss": 2.1623,
+      "step": 958
+    },
+    {
+      "epoch": 1.77,
+      "learning_rate": 1.2931200000000003e-05,
+      "loss": 2.2235,
+      "step": 959
+    },
+    {
+      "epoch": 1.77,
+      "learning_rate": 1.2928000000000002e-05,
+      "loss": 2.2354,
+      "step": 960
+    },
+    {
+      "epoch": 1.77,
+      "eval_loss": 2.283588171005249,
+      "eval_runtime": 232.8873,
+      "eval_samples_per_second": 37.241,
+      "eval_steps_per_second": 0.584,
+      "step": 960
+    },
+    {
+      "epoch": 1.77,
+      "learning_rate": 1.2924800000000002e-05,
+      "loss": 2.2574,
+      "step": 961
+    },
+    {
+      "epoch": 1.77,
+      "learning_rate": 1.2921600000000002e-05,
+      "loss": 2.1907,
+      "step": 962
+    },
+    {
+      "epoch": 1.78,
+      "learning_rate": 1.2918400000000002e-05,
+      "loss": 2.2573,
+      "step": 963
+    },
+    {
+      "epoch": 1.78,
+      "learning_rate": 1.2915200000000003e-05,
+      "loss": 2.228,
+      "step": 964
+    },
+    {
+      "epoch": 1.78,
+      "learning_rate": 1.2912000000000003e-05,
+      "loss": 2.1263,
+      "step": 965
+    },
+    {
+      "epoch": 1.78,
+      "learning_rate": 1.29088e-05,
+      "loss": 2.1663,
+      "step": 966
+    },
+    {
+      "epoch": 1.78,
+      "learning_rate": 1.2905600000000002e-05,
+      "loss": 2.2189,
+      "step": 967
+    },
+    {
+      "epoch": 1.79,
+      "learning_rate": 1.2902400000000002e-05,
+      "loss": 2.1123,
+      "step": 968
+    },
+    {
+      "epoch": 1.79,
+      "learning_rate": 1.2899200000000002e-05,
+      "loss": 2.2449,
+      "step": 969
+    },
+    {
+      "epoch": 1.79,
+      "learning_rate": 1.2896000000000003e-05,
+      "loss": 2.109,
+      "step": 970
+    },
+    {
+      "epoch": 1.79,
+      "learning_rate": 1.2892800000000001e-05,
+      "loss": 2.2236,
+      "step": 971
+    },
+    {
+      "epoch": 1.79,
+      "learning_rate": 1.2889600000000001e-05,
+      "loss": 2.1828,
+      "step": 972
+    },
+    {
+      "epoch": 1.79,
+      "learning_rate": 1.2886400000000003e-05,
+      "loss": 2.1916,
+      "step": 973
+    },
+    {
+      "epoch": 1.8,
+      "learning_rate": 1.2883200000000002e-05,
+      "loss": 2.1637,
+      "step": 974
+    },
+    {
+      "epoch": 1.8,
+      "learning_rate": 1.2880000000000004e-05,
+      "loss": 2.2734,
+      "step": 975
+    },
+    {
+      "epoch": 1.8,
+      "learning_rate": 1.2876800000000002e-05,
+      "loss": 2.1904,
+      "step": 976
+    },
+    {
+      "epoch": 1.8,
+      "learning_rate": 1.2873600000000002e-05,
+      "loss": 2.2465,
+      "step": 977
+    },
+    {
+      "epoch": 1.8,
+      "learning_rate": 1.2870400000000003e-05,
+      "loss": 2.1789,
+      "step": 978
+    },
+    {
+      "epoch": 1.81,
+      "learning_rate": 1.2867200000000003e-05,
+      "loss": 2.1634,
+      "step": 979
+    },
+    {
+      "epoch": 1.81,
+      "learning_rate": 1.2864000000000003e-05,
+      "loss": 2.2994,
+      "step": 980
+    },
+    {
+      "epoch": 1.81,
+      "learning_rate": 1.2860800000000002e-05,
+      "loss": 2.2669,
+      "step": 981
+    },
+    {
+      "epoch": 1.81,
+      "learning_rate": 1.2857600000000002e-05,
+      "loss": 2.2016,
+      "step": 982
+    },
+    {
+      "epoch": 1.81,
+      "learning_rate": 1.2854400000000002e-05,
+      "loss": 2.2722,
+      "step": 983
+    },
+    {
+      "epoch": 1.81,
+      "learning_rate": 1.2851200000000003e-05,
+      "loss": 2.2367,
+      "step": 984
+    },
+    {
+      "epoch": 1.82,
+      "learning_rate": 1.2848000000000003e-05,
+      "loss": 2.1425,
+      "step": 985
+    },
+    {
+      "epoch": 1.82,
+      "learning_rate": 1.2844800000000001e-05,
+      "loss": 2.1624,
+      "step": 986
+    },
+    {
+      "epoch": 1.82,
+      "learning_rate": 1.2841600000000003e-05,
+      "loss": 2.1253,
+      "step": 987
+    },
+    {
+      "epoch": 1.82,
+      "learning_rate": 1.2838400000000002e-05,
+      "loss": 2.2248,
+      "step": 988
+    },
+    {
+      "epoch": 1.82,
+      "learning_rate": 1.2835200000000002e-05,
+      "loss": 2.1892,
+      "step": 989
+    },
+    {
+      "epoch": 1.83,
+      "learning_rate": 1.2832000000000004e-05,
+      "loss": 2.2207,
+      "step": 990
+    },
+    {
+      "epoch": 1.83,
+      "eval_loss": 2.2836029529571533,
+      "eval_runtime": 232.816,
+      "eval_samples_per_second": 37.253,
+      "eval_steps_per_second": 0.584,
+      "step": 990
+    },
+    {
+      "epoch": 1.83,
+      "learning_rate": 1.2828800000000002e-05,
+      "loss": 2.1592,
+      "step": 991
+    },
+    {
+      "epoch": 1.83,
+      "learning_rate": 1.2825600000000001e-05,
+      "loss": 2.1805,
+      "step": 992
+    },
+    {
+      "epoch": 1.83,
+      "learning_rate": 1.2822400000000003e-05,
+      "loss": 2.2063,
+      "step": 993
+    },
+    {
+      "epoch": 1.83,
+      "learning_rate": 1.2819200000000003e-05,
+      "loss": 2.2105,
+      "step": 994
+    },
+    {
+      "epoch": 1.84,
+      "learning_rate": 1.2816000000000002e-05,
+      "loss": 2.2151,
+      "step": 995
+    },
+    {
+      "epoch": 1.84,
+      "learning_rate": 1.2812800000000002e-05,
+      "loss": 2.148,
+      "step": 996
+    },
+    {
+      "epoch": 1.84,
+      "learning_rate": 1.2809600000000002e-05,
+      "loss": 2.2129,
+      "step": 997
+    },
+    {
+      "epoch": 1.84,
+      "learning_rate": 1.2806400000000002e-05,
+      "loss": 2.1779,
+      "step": 998
+    },
+    {
+      "epoch": 1.84,
+      "learning_rate": 1.2803200000000003e-05,
+      "loss": 2.2705,
+      "step": 999
+    },
+    {
+      "epoch": 1.84,
+      "learning_rate": 1.2800000000000003e-05,
+      "loss": 2.243,
+      "step": 1000
+    },
+    {
+      "epoch": 1.85,
+      "learning_rate": 1.2796800000000001e-05,
+      "loss": 2.2974,
+      "step": 1001
+    },
+    {
+      "epoch": 1.85,
+      "learning_rate": 1.2793600000000002e-05,
+      "loss": 2.2768,
+      "step": 1002
+    },
+    {
+      "epoch": 1.85,
+      "learning_rate": 1.2790400000000002e-05,
+      "loss": 2.1879,
+      "step": 1003
+    },
+    {
+      "epoch": 1.85,
+      "learning_rate": 1.2787200000000002e-05,
+      "loss": 2.1339,
+      "step": 1004
+    },
+    {
+      "epoch": 1.85,
+      "learning_rate": 1.2784000000000003e-05,
+      "loss": 2.2129,
+      "step": 1005
+    },
+    {
+      "epoch": 1.86,
+      "learning_rate": 1.2780800000000001e-05,
+      "loss": 2.2315,
+      "step": 1006
+    },
+    {
+      "epoch": 1.86,
+      "learning_rate": 1.2777600000000001e-05,
+      "loss": 2.2686,
+      "step": 1007
+    },
+    {
+      "epoch": 1.86,
+      "learning_rate": 1.2774400000000003e-05,
+      "loss": 2.1447,
+      "step": 1008
+    },
+    {
+      "epoch": 1.86,
+      "learning_rate": 1.2771200000000002e-05,
+      "loss": 2.1977,
+      "step": 1009
+    },
+    {
+      "epoch": 1.86,
+      "learning_rate": 1.2768000000000002e-05,
+      "loss": 2.2865,
+      "step": 1010
+    },
+    {
+      "epoch": 1.86,
+      "learning_rate": 1.2764800000000002e-05,
+      "loss": 2.1787,
+      "step": 1011
+    },
+    {
+      "epoch": 1.87,
+      "learning_rate": 1.2761600000000002e-05,
+      "loss": 2.2252,
+      "step": 1012
+    },
+    {
+      "epoch": 1.87,
+      "learning_rate": 1.2758400000000001e-05,
+      "loss": 2.2721,
+      "step": 1013
+    },
+    {
+      "epoch": 1.87,
+      "learning_rate": 1.2755200000000003e-05,
+      "loss": 2.2256,
+      "step": 1014
+    },
+    {
+      "epoch": 1.87,
+      "learning_rate": 1.2752000000000003e-05,
+      "loss": 2.2379,
+      "step": 1015
+    },
+    {
+      "epoch": 1.87,
+      "learning_rate": 1.27488e-05,
+      "loss": 2.1722,
+      "step": 1016
+    },
+    {
+      "epoch": 1.88,
+      "learning_rate": 1.2745600000000002e-05,
+      "loss": 2.2238,
+      "step": 1017
+    },
+    {
+      "epoch": 1.88,
+      "learning_rate": 1.2742400000000002e-05,
+      "loss": 2.2077,
+      "step": 1018
+    },
+    {
+      "epoch": 1.88,
+      "learning_rate": 1.2739200000000002e-05,
+      "loss": 2.2267,
+      "step": 1019
+    },
+    {
+      "epoch": 1.88,
+      "learning_rate": 1.2736000000000003e-05,
+      "loss": 2.1589,
+      "step": 1020
+    },
+    {
+      "epoch": 1.88,
+      "eval_loss": 2.286294460296631,
+      "eval_runtime": 232.5526,
+      "eval_samples_per_second": 37.295,
+      "eval_steps_per_second": 0.585,
+      "step": 1020
+    },
+    {
+      "epoch": 1.88,
+      "learning_rate": 1.2732800000000001e-05,
+      "loss": 2.1544,
+      "step": 1021
+    },
+    {
+      "epoch": 1.88,
+      "learning_rate": 1.2729600000000001e-05,
+      "loss": 2.1864,
+      "step": 1022
+    },
+    {
+      "epoch": 1.89,
+      "learning_rate": 1.2726400000000002e-05,
+      "loss": 2.2916,
+      "step": 1023
+    },
+    {
+      "epoch": 1.89,
+      "learning_rate": 1.2723200000000002e-05,
+      "loss": 2.2087,
+      "step": 1024
+    },
+    {
+      "epoch": 1.89,
+      "learning_rate": 1.2720000000000004e-05,
+      "loss": 2.1755,
+      "step": 1025
+    },
+    {
+      "epoch": 1.89,
+      "learning_rate": 1.2716800000000002e-05,
+      "loss": 2.2404,
+      "step": 1026
+    },
+    {
+      "epoch": 1.89,
+      "learning_rate": 1.2713600000000001e-05,
+      "loss": 2.3272,
+      "step": 1027
+    },
+    {
+      "epoch": 1.9,
+      "learning_rate": 1.2710400000000003e-05,
+      "loss": 2.3319,
+      "step": 1028
+    },
+    {
+      "epoch": 1.9,
+      "learning_rate": 1.2707200000000003e-05,
+      "loss": 2.2221,
+      "step": 1029
+    },
+    {
+      "epoch": 1.9,
+      "learning_rate": 1.2704000000000002e-05,
+      "loss": 2.2061,
+      "step": 1030
+    },
+    {
+      "epoch": 1.9,
+      "learning_rate": 1.2700800000000002e-05,
+      "loss": 2.2794,
+      "step": 1031
+    },
+    {
+      "epoch": 1.9,
+      "learning_rate": 1.2697600000000002e-05,
+      "loss": 2.2148,
+      "step": 1032
+    },
+    {
+      "epoch": 1.91,
+      "learning_rate": 1.2694400000000002e-05,
+      "loss": 2.2313,
+      "step": 1033
+    },
+    {
+      "epoch": 1.91,
+      "learning_rate": 1.2691200000000003e-05,
+      "loss": 2.3196,
+      "step": 1034
+    },
+    {
+      "epoch": 1.91,
+      "learning_rate": 1.2688000000000003e-05,
+      "loss": 2.1571,
+      "step": 1035
+    },
+    {
+      "epoch": 1.91,
+      "learning_rate": 1.2684800000000001e-05,
+      "loss": 2.1385,
+      "step": 1036
+    },
+    {
+      "epoch": 1.91,
+      "learning_rate": 1.2681600000000002e-05,
+      "loss": 2.1592,
+      "step": 1037
+    },
+    {
+      "epoch": 1.91,
+      "learning_rate": 1.2678400000000002e-05,
+      "loss": 2.1921,
+      "step": 1038
+    },
+    {
+      "epoch": 1.92,
+      "learning_rate": 1.2675200000000002e-05,
+      "loss": 2.1597,
+      "step": 1039
+    },
+    {
+      "epoch": 1.92,
+      "learning_rate": 1.2672000000000003e-05,
+      "loss": 2.2587,
+      "step": 1040
+    },
+    {
+      "epoch": 1.92,
+      "learning_rate": 1.2668800000000001e-05,
+      "loss": 2.1847,
+      "step": 1041
+    },
+    {
+      "epoch": 1.92,
+      "learning_rate": 1.2665600000000001e-05,
+      "loss": 2.2028,
+      "step": 1042
+    },
+    {
+      "epoch": 1.92,
+      "learning_rate": 1.2662400000000003e-05,
+      "loss": 2.0868,
+      "step": 1043
+    },
+    {
+      "epoch": 1.93,
+      "learning_rate": 1.2659200000000002e-05,
+      "loss": 2.1181,
+      "step": 1044
+    },
+    {
+      "epoch": 1.93,
+      "learning_rate": 1.2656000000000002e-05,
+      "loss": 2.1707,
+      "step": 1045
+    },
+    {
+      "epoch": 1.93,
+      "learning_rate": 1.2652800000000002e-05,
+      "loss": 2.1841,
+      "step": 1046
+    },
+    {
+      "epoch": 1.93,
+      "learning_rate": 1.2649600000000002e-05,
+      "loss": 2.1988,
+      "step": 1047
+    },
+    {
+      "epoch": 1.93,
+      "learning_rate": 1.2646400000000001e-05,
+      "loss": 2.2667,
+      "step": 1048
+    },
+    {
+      "epoch": 1.93,
+      "learning_rate": 1.2643200000000003e-05,
+      "loss": 2.24,
+      "step": 1049
+    },
+    {
+      "epoch": 1.94,
+      "learning_rate": 1.2640000000000003e-05,
+      "loss": 2.3176,
+      "step": 1050
+    },
+    {
+      "epoch": 1.94,
+      "eval_loss": 2.282726764678955,
+      "eval_runtime": 232.6019,
+      "eval_samples_per_second": 37.287,
+      "eval_steps_per_second": 0.585,
+      "step": 1050
+    },
+    {
+      "epoch": 1.94,
+      "learning_rate": 1.26368e-05,
+      "loss": 2.2317,
+      "step": 1051
+    },
+    {
+      "epoch": 1.94,
+      "learning_rate": 1.2633600000000002e-05,
+      "loss": 2.2193,
+      "step": 1052
+    },
+    {
+      "epoch": 1.94,
+      "learning_rate": 1.2630400000000002e-05,
+      "loss": 2.3179,
+      "step": 1053
+    },
+    {
+      "epoch": 1.94,
+      "learning_rate": 1.2627200000000002e-05,
+      "loss": 2.1861,
+      "step": 1054
+    },
+    {
+      "epoch": 1.95,
+      "learning_rate": 1.2624000000000003e-05,
+      "loss": 2.1846,
+      "step": 1055
+    },
+    {
+      "epoch": 1.95,
+      "learning_rate": 1.2620800000000001e-05,
+      "loss": 2.1691,
+      "step": 1056
+    },
+    {
+      "epoch": 1.95,
+      "learning_rate": 1.2617600000000001e-05,
+      "loss": 2.1489,
+      "step": 1057
+    },
+    {
+      "epoch": 1.95,
+      "learning_rate": 1.2614400000000002e-05,
+      "loss": 2.2763,
+      "step": 1058
+    },
+    {
+      "epoch": 1.95,
+      "learning_rate": 1.2611200000000002e-05,
+      "loss": 2.1954,
+      "step": 1059
+    },
+    {
+      "epoch": 1.95,
+      "learning_rate": 1.2608000000000002e-05,
+      "loss": 2.263,
+      "step": 1060
+    },
+    {
+      "epoch": 1.96,
+      "learning_rate": 1.2604800000000002e-05,
+      "loss": 2.284,
+      "step": 1061
+    },
+    {
+      "epoch": 1.96,
+      "learning_rate": 1.2601600000000001e-05,
+      "loss": 2.179,
+      "step": 1062
+    },
+    {
+      "epoch": 1.96,
+      "learning_rate": 1.2598400000000001e-05,
+      "loss": 2.2496,
+      "step": 1063
+    },
+    {
+      "epoch": 1.96,
+      "learning_rate": 1.2595200000000003e-05,
+      "loss": 2.1274,
+      "step": 1064
+    },
+    {
+      "epoch": 1.96,
+      "learning_rate": 1.2592000000000002e-05,
+      "loss": 2.1877,
+      "step": 1065
+    },
+    {
+      "epoch": 1.97,
+      "learning_rate": 1.2588800000000002e-05,
+      "loss": 2.1722,
+      "step": 1066
+    },
+    {
+      "epoch": 1.97,
+      "learning_rate": 1.2585600000000002e-05,
+      "loss": 2.172,
+      "step": 1067
+    },
+    {
+      "epoch": 1.97,
+      "learning_rate": 1.2582400000000002e-05,
+      "loss": 2.1049,
+      "step": 1068
+    },
+    {
+      "epoch": 1.97,
+      "learning_rate": 1.2579200000000001e-05,
+      "loss": 2.2471,
+      "step": 1069
+    },
+    {
+      "epoch": 1.97,
+      "learning_rate": 1.2576000000000003e-05,
+      "loss": 2.2287,
+      "step": 1070
+    },
+    {
+      "epoch": 1.98,
+      "learning_rate": 1.2572800000000003e-05,
+      "loss": 2.1895,
+      "step": 1071
+    },
+    {
+      "epoch": 1.98,
+      "learning_rate": 1.25696e-05,
+      "loss": 2.0958,
+      "step": 1072
+    },
+    {
+      "epoch": 1.98,
+      "learning_rate": 1.2566400000000002e-05,
+      "loss": 2.2092,
+      "step": 1073
+    },
+    {
+      "epoch": 1.98,
+      "learning_rate": 1.2563200000000002e-05,
+      "loss": 2.2783,
+      "step": 1074
+    },
+    {
+      "epoch": 1.98,
+      "learning_rate": 1.2560000000000003e-05,
+      "loss": 2.2276,
+      "step": 1075
+    },
+    {
+      "epoch": 1.98,
+      "learning_rate": 1.2556800000000003e-05,
+      "loss": 2.136,
+      "step": 1076
+    },
+    {
+      "epoch": 1.99,
+      "learning_rate": 1.2553600000000001e-05,
+      "loss": 2.2688,
+      "step": 1077
+    },
+    {
+      "epoch": 1.99,
+      "learning_rate": 1.2550400000000003e-05,
+      "loss": 2.1981,
+      "step": 1078
+    },
+    {
+      "epoch": 1.99,
+      "learning_rate": 1.2547200000000002e-05,
+      "loss": 2.2485,
+      "step": 1079
+    },
+    {
+      "epoch": 1.99,
+      "learning_rate": 1.2544000000000002e-05,
+      "loss": 2.1302,
+      "step": 1080
+    },
+    {
+      "epoch": 1.99,
+      "eval_loss": 2.2810168266296387,
+      "eval_runtime": 231.821,
+      "eval_samples_per_second": 37.412,
+      "eval_steps_per_second": 0.587,
+      "step": 1080
+    },
+    {
+      "epoch": 1.99,
+      "learning_rate": 1.2540800000000004e-05,
+      "loss": 2.211,
+      "step": 1081
+    },
+    {
+      "epoch": 2.0,
+      "learning_rate": 1.2537600000000002e-05,
+      "loss": 2.1877,
+      "step": 1082
+    },
+    {
+      "epoch": 2.0,
+      "learning_rate": 1.2534400000000001e-05,
+      "loss": 2.2446,
+      "step": 1083
+    },
+    {
+      "epoch": 2.0,
+      "learning_rate": 1.2531200000000003e-05,
+      "loss": 2.1393,
+      "step": 1084
+    },
+    {
+      "epoch": 2.0,
+      "learning_rate": 1.2528000000000003e-05,
+      "loss": 3.2826,
+      "step": 1085
+    },
+    {
+      "epoch": 2.0,
+      "learning_rate": 1.2524800000000002e-05,
+      "loss": 2.1195,
+      "step": 1086
+    },
+    {
+      "epoch": 2.01,
+      "learning_rate": 1.2521600000000002e-05,
+      "loss": 2.1111,
+      "step": 1087
+    },
+    {
+      "epoch": 2.01,
+      "learning_rate": 1.2518400000000002e-05,
+      "loss": 2.0742,
+      "step": 1088
+    },
+    {
+      "epoch": 2.01,
+      "learning_rate": 1.2515200000000002e-05,
+      "loss": 2.1254,
+      "step": 1089
+    },
+    {
+      "epoch": 2.01,
+      "learning_rate": 1.2512000000000003e-05,
+      "loss": 2.0826,
+      "step": 1090
+    },
+    {
+      "epoch": 2.01,
+      "learning_rate": 1.2508800000000003e-05,
+      "loss": 2.1623,
+      "step": 1091
+    },
+    {
+      "epoch": 2.01,
+      "learning_rate": 1.2505600000000001e-05,
+      "loss": 2.1527,
+      "step": 1092
+    },
+    {
+      "epoch": 2.02,
+      "learning_rate": 1.2502400000000002e-05,
+      "loss": 2.1113,
+      "step": 1093
+    },
+    {
+      "epoch": 2.02,
+      "learning_rate": 1.2499200000000002e-05,
+      "loss": 2.0738,
+      "step": 1094
+    },
+    {
+      "epoch": 2.02,
+      "learning_rate": 1.2496000000000002e-05,
+      "loss": 2.0631,
+      "step": 1095
+    },
+    {
+      "epoch": 2.02,
+      "learning_rate": 1.2492800000000003e-05,
+      "loss": 2.1067,
+      "step": 1096
+    },
+    {
+      "epoch": 2.02,
+      "learning_rate": 1.2489600000000002e-05,
+      "loss": 2.1819,
+      "step": 1097
+    },
+    {
+      "epoch": 2.03,
+      "learning_rate": 1.2486400000000001e-05,
+      "loss": 2.1105,
+      "step": 1098
+    },
+    {
+      "epoch": 2.03,
+      "learning_rate": 1.2483200000000003e-05,
+      "loss": 2.2094,
+      "step": 1099
+    },
+    {
+      "epoch": 2.03,
+      "learning_rate": 1.2480000000000002e-05,
+      "loss": 2.1086,
+      "step": 1100
+    },
+    {
+      "epoch": 2.03,
+      "learning_rate": 1.2476800000000002e-05,
+      "loss": 2.1344,
+      "step": 1101
+    },
+    {
+      "epoch": 2.03,
+      "learning_rate": 1.2473600000000002e-05,
+      "loss": 2.0553,
+      "step": 1102
+    },
+    {
+      "epoch": 2.04,
+      "learning_rate": 1.2470400000000002e-05,
+      "loss": 2.1484,
+      "step": 1103
+    },
+    {
+      "epoch": 2.04,
+      "learning_rate": 1.2467200000000002e-05,
+      "loss": 2.1273,
+      "step": 1104
+    },
+    {
+      "epoch": 2.04,
+      "learning_rate": 1.2464000000000003e-05,
+      "loss": 2.0588,
+      "step": 1105
+    },
+    {
+      "epoch": 2.04,
+      "learning_rate": 1.2460800000000003e-05,
+      "loss": 2.1328,
+      "step": 1106
+    },
+    {
+      "epoch": 2.04,
+      "learning_rate": 1.24576e-05,
+      "loss": 2.151,
+      "step": 1107
+    },
+    {
+      "epoch": 2.04,
+      "learning_rate": 1.2454400000000002e-05,
+      "loss": 2.0756,
+      "step": 1108
+    },
+    {
+      "epoch": 2.05,
+      "learning_rate": 1.2451200000000002e-05,
+      "loss": 2.1864,
+      "step": 1109
+    },
+    {
+      "epoch": 2.05,
+      "learning_rate": 1.2448000000000002e-05,
+      "loss": 2.0048,
+      "step": 1110
+    },
+    {
+      "epoch": 2.05,
+      "eval_loss": 2.2935333251953125,
+      "eval_runtime": 231.4632,
+      "eval_samples_per_second": 37.47,
+      "eval_steps_per_second": 0.588,
+      "step": 1110
+    },
+    {
+      "epoch": 2.05,
+      "learning_rate": 1.2444800000000003e-05,
+      "loss": 2.1659,
+      "step": 1111
+    },
+    {
+      "epoch": 2.05,
+      "learning_rate": 1.2441600000000001e-05,
+      "loss": 2.0766,
+      "step": 1112
+    },
+    {
+      "epoch": 2.05,
+      "learning_rate": 1.2438400000000001e-05,
+      "loss": 2.0925,
+      "step": 1113
+    },
+    {
+      "epoch": 2.06,
+      "learning_rate": 1.2435200000000003e-05,
+      "loss": 2.1421,
+      "step": 1114
+    },
+    {
+      "epoch": 2.06,
+      "learning_rate": 1.2432000000000002e-05,
+      "loss": 2.0722,
+      "step": 1115
+    },
+    {
+      "epoch": 2.06,
+      "learning_rate": 1.2428800000000002e-05,
+      "loss": 2.0592,
+      "step": 1116
+    },
+    {
+      "epoch": 2.06,
+      "learning_rate": 1.2425600000000002e-05,
+      "loss": 2.1999,
+      "step": 1117
+    },
+    {
+      "epoch": 2.06,
+      "learning_rate": 1.2422400000000002e-05,
+      "loss": 2.0484,
+      "step": 1118
+    },
+    {
+      "epoch": 2.06,
+      "learning_rate": 1.2419200000000001e-05,
+      "loss": 2.1227,
+      "step": 1119
+    },
+    {
+      "epoch": 2.07,
+      "learning_rate": 1.2416000000000003e-05,
+      "loss": 2.0428,
+      "step": 1120
+    },
+    {
+      "epoch": 2.07,
+      "learning_rate": 1.2412800000000003e-05,
+      "loss": 2.1421,
+      "step": 1121
+    },
+    {
+      "epoch": 2.07,
+      "learning_rate": 1.24096e-05,
+      "loss": 2.208,
+      "step": 1122
+    },
+    {
+      "epoch": 2.07,
+      "learning_rate": 1.2406400000000002e-05,
+      "loss": 2.0962,
+      "step": 1123
+    },
+    {
+      "epoch": 2.07,
+      "learning_rate": 1.2403200000000002e-05,
+      "loss": 2.047,
+      "step": 1124
+    },
+    {
+      "epoch": 2.08,
+      "learning_rate": 1.2400000000000003e-05,
+      "loss": 2.1502,
+      "step": 1125
+    },
+    {
+      "epoch": 2.08,
+      "learning_rate": 1.2396800000000003e-05,
+      "loss": 2.1842,
+      "step": 1126
+    },
+    {
+      "epoch": 2.08,
+      "learning_rate": 1.2393600000000001e-05,
+      "loss": 1.935,
+      "step": 1127
+    },
+    {
+      "epoch": 2.08,
+      "learning_rate": 1.2390400000000003e-05,
+      "loss": 2.1215,
+      "step": 1128
+    },
+    {
+      "epoch": 2.08,
+      "learning_rate": 1.2387200000000002e-05,
+      "loss": 2.0953,
+      "step": 1129
+    },
+    {
+      "epoch": 2.08,
+      "learning_rate": 1.2384000000000002e-05,
+      "loss": 2.1837,
+      "step": 1130
+    },
+    {
+      "epoch": 2.09,
+      "learning_rate": 1.2380800000000004e-05,
+      "loss": 2.1699,
+      "step": 1131
+    },
+    {
+      "epoch": 2.09,
+      "learning_rate": 1.2377600000000002e-05,
+      "loss": 2.162,
+      "step": 1132
+    },
+    {
+      "epoch": 2.09,
+      "learning_rate": 1.2374400000000001e-05,
+      "loss": 2.0172,
+      "step": 1133
+    },
+    {
+      "epoch": 2.09,
+      "learning_rate": 1.2371200000000003e-05,
+      "loss": 2.188,
+      "step": 1134
+    },
+    {
+      "epoch": 2.09,
+      "learning_rate": 1.2368000000000003e-05,
+      "loss": 2.0974,
+      "step": 1135
+    },
+    {
+      "epoch": 2.1,
+      "learning_rate": 1.2364800000000002e-05,
+      "loss": 2.0069,
+      "step": 1136
+    },
+    {
+      "epoch": 2.1,
+      "learning_rate": 1.2361600000000002e-05,
+      "loss": 2.1845,
+      "step": 1137
+    },
+    {
+      "epoch": 2.1,
+      "learning_rate": 1.2358400000000002e-05,
+      "loss": 2.2086,
+      "step": 1138
+    },
+    {
+      "epoch": 2.1,
+      "learning_rate": 1.2355200000000002e-05,
+      "loss": 2.1328,
+      "step": 1139
+    },
+    {
+      "epoch": 2.1,
+      "learning_rate": 1.2352000000000003e-05,
+      "loss": 2.0693,
+      "step": 1140
+    },
+    {
+      "epoch": 2.1,
+      "eval_loss": 2.2939326763153076,
+      "eval_runtime": 232.6241,
+      "eval_samples_per_second": 37.283,
+      "eval_steps_per_second": 0.585,
+      "step": 1140
+    },
+    {
+      "epoch": 2.11,
+      "learning_rate": 1.2348800000000003e-05,
+      "loss": 2.1484,
+      "step": 1141
+    },
+    {
+      "epoch": 2.11,
+      "learning_rate": 1.2345600000000001e-05,
+      "loss": 2.128,
+      "step": 1142
+    },
+    {
+      "epoch": 2.11,
+      "learning_rate": 1.2342400000000002e-05,
+      "loss": 2.1743,
+      "step": 1143
+    },
+    {
+      "epoch": 2.11,
+      "learning_rate": 1.2339200000000002e-05,
+      "loss": 2.1598,
+      "step": 1144
+    },
+    {
+      "epoch": 2.11,
+      "learning_rate": 1.2336000000000002e-05,
+      "loss": 2.0533,
+      "step": 1145
+    },
+    {
+      "epoch": 2.11,
+      "learning_rate": 1.2332800000000003e-05,
+      "loss": 2.1912,
+      "step": 1146
+    },
+    {
+      "epoch": 2.12,
+      "learning_rate": 1.2329600000000001e-05,
+      "loss": 2.103,
+      "step": 1147
+    },
+    {
+      "epoch": 2.12,
+      "learning_rate": 1.2326400000000001e-05,
+      "loss": 2.1246,
+      "step": 1148
+    },
+    {
+      "epoch": 2.12,
+      "learning_rate": 1.2323200000000003e-05,
+      "loss": 2.0688,
+      "step": 1149
+    },
+    {
+      "epoch": 2.12,
+      "learning_rate": 1.2320000000000002e-05,
+      "loss": 2.0288,
+      "step": 1150
+    },
+    {
+      "epoch": 2.12,
+      "learning_rate": 1.2316800000000002e-05,
+      "loss": 2.1982,
+      "step": 1151
+    },
+    {
+      "epoch": 2.13,
+      "learning_rate": 1.2313600000000002e-05,
+      "loss": 2.0647,
+      "step": 1152
+    },
+    {
+      "epoch": 2.13,
+      "learning_rate": 1.2310400000000002e-05,
+      "loss": 2.1112,
+      "step": 1153
+    },
+    {
+      "epoch": 2.13,
+      "learning_rate": 1.2307200000000001e-05,
+      "loss": 2.1848,
+      "step": 1154
+    },
+    {
+      "epoch": 2.13,
+      "learning_rate": 1.2304000000000003e-05,
+      "loss": 2.2162,
+      "step": 1155
+    },
+    {
+      "epoch": 2.13,
+      "learning_rate": 1.2300800000000003e-05,
+      "loss": 2.08,
+      "step": 1156
+    },
+    {
+      "epoch": 2.13,
+      "learning_rate": 1.22976e-05,
+      "loss": 2.1313,
+      "step": 1157
+    },
+    {
+      "epoch": 2.14,
+      "learning_rate": 1.2294400000000002e-05,
+      "loss": 2.2197,
+      "step": 1158
+    },
+    {
+      "epoch": 2.14,
+      "learning_rate": 1.2291200000000002e-05,
+      "loss": 2.185,
+      "step": 1159
+    },
+    {
+      "epoch": 2.14,
+      "learning_rate": 1.2288000000000002e-05,
+      "loss": 2.1129,
+      "step": 1160
+    },
+    {
+      "epoch": 2.14,
+      "learning_rate": 1.2284800000000003e-05,
+      "loss": 2.0588,
+      "step": 1161
+    },
+    {
+      "epoch": 2.14,
+      "learning_rate": 1.2281600000000001e-05,
+      "loss": 2.0993,
+      "step": 1162
+    },
+    {
+      "epoch": 2.15,
+      "learning_rate": 1.2278400000000001e-05,
+      "loss": 2.1601,
+      "step": 1163
+    },
+    {
+      "epoch": 2.15,
+      "learning_rate": 1.2275200000000002e-05,
+      "loss": 2.1246,
+      "step": 1164
+    },
+    {
+      "epoch": 2.15,
+      "learning_rate": 1.2272000000000002e-05,
+      "loss": 2.0078,
+      "step": 1165
+    },
+    {
+      "epoch": 2.15,
+      "learning_rate": 1.2268800000000002e-05,
+      "loss": 2.1736,
+      "step": 1166
+    },
+    {
+      "epoch": 2.15,
+      "learning_rate": 1.2265600000000002e-05,
+      "loss": 2.0693,
+      "step": 1167
+    },
+    {
+      "epoch": 2.15,
+      "learning_rate": 1.2262400000000001e-05,
+      "loss": 2.1263,
+      "step": 1168
+    },
+    {
+      "epoch": 2.16,
+      "learning_rate": 1.2259200000000001e-05,
+      "loss": 2.1752,
+      "step": 1169
+    },
+    {
+      "epoch": 2.16,
+      "learning_rate": 1.2256000000000003e-05,
+      "loss": 2.0542,
+      "step": 1170
+    },
+    {
+      "epoch": 2.16,
+      "eval_loss": 2.2919065952301025,
+      "eval_runtime": 230.5365,
+      "eval_samples_per_second": 37.621,
+      "eval_steps_per_second": 0.59,
+      "step": 1170
+    },
+    {
+      "epoch": 2.16,
+      "learning_rate": 1.2252800000000002e-05,
+      "loss": 2.1073,
+      "step": 1171
+    },
+    {
+      "epoch": 2.16,
+      "learning_rate": 1.22496e-05,
+      "loss": 2.0539,
+      "step": 1172
+    },
+    {
+      "epoch": 2.16,
+      "learning_rate": 1.2246400000000002e-05,
+      "loss": 2.0592,
+      "step": 1173
+    },
+    {
+      "epoch": 2.17,
+      "learning_rate": 1.2243200000000002e-05,
+      "loss": 2.0873,
+      "step": 1174
+    },
+    {
+      "epoch": 2.17,
+      "learning_rate": 1.2240000000000003e-05,
+      "loss": 2.1092,
+      "step": 1175
+    },
+    {
+      "epoch": 2.17,
+      "learning_rate": 1.2236800000000003e-05,
+      "loss": 2.0677,
+      "step": 1176
+    },
+    {
+      "epoch": 2.17,
+      "learning_rate": 1.2233600000000001e-05,
+      "loss": 2.0814,
+      "step": 1177
+    },
+    {
+      "epoch": 2.17,
+      "learning_rate": 1.2230400000000002e-05,
+      "loss": 2.0902,
+      "step": 1178
+    },
+    {
+      "epoch": 2.18,
+      "learning_rate": 1.2227200000000002e-05,
+      "loss": 2.1716,
+      "step": 1179
+    },
+    {
+      "epoch": 2.18,
+      "learning_rate": 1.2224000000000002e-05,
+      "loss": 2.1329,
+      "step": 1180
+    },
+    {
+      "epoch": 2.18,
+      "learning_rate": 1.2220800000000003e-05,
+      "loss": 2.1566,
+      "step": 1181
+    },
+    {
+      "epoch": 2.18,
+      "learning_rate": 1.2217600000000001e-05,
+      "loss": 2.2131,
+      "step": 1182
+    },
+    {
+      "epoch": 2.18,
+      "learning_rate": 1.2214400000000001e-05,
+      "loss": 2.1207,
+      "step": 1183
+    },
+    {
+      "epoch": 2.18,
+      "learning_rate": 1.2211200000000003e-05,
+      "loss": 2.0733,
+      "step": 1184
+    },
+    {
+      "epoch": 2.19,
+      "learning_rate": 1.2208000000000002e-05,
+      "loss": 2.1668,
+      "step": 1185
+    },
+    {
+      "epoch": 2.19,
+      "learning_rate": 1.2204800000000002e-05,
+      "loss": 2.0699,
+      "step": 1186
+    },
+    {
+      "epoch": 2.19,
+      "learning_rate": 1.2201600000000002e-05,
+      "loss": 2.0769,
+      "step": 1187
+    },
+    {
+      "epoch": 2.19,
+      "learning_rate": 1.2198400000000002e-05,
+      "loss": 2.1274,
+      "step": 1188
+    },
+    {
+      "epoch": 2.19,
+      "learning_rate": 1.2195200000000001e-05,
+      "loss": 2.055,
+      "step": 1189
+    },
+    {
+      "epoch": 2.2,
+      "learning_rate": 1.2192000000000003e-05,
+      "loss": 2.1401,
+      "step": 1190
+    },
+    {
+      "epoch": 2.2,
+      "learning_rate": 1.2188800000000003e-05,
+      "loss": 2.1099,
+      "step": 1191
+    },
+    {
+      "epoch": 2.2,
+      "learning_rate": 1.2185600000000002e-05,
+      "loss": 2.1202,
+      "step": 1192
+    },
+    {
+      "epoch": 2.2,
+      "learning_rate": 1.2182400000000002e-05,
+      "loss": 2.1038,
+      "step": 1193
+    },
+    {
+      "epoch": 2.2,
+      "learning_rate": 1.2179200000000002e-05,
+      "loss": 2.1403,
+      "step": 1194
+    },
+    {
+      "epoch": 2.2,
+      "learning_rate": 1.2176000000000002e-05,
+      "loss": 2.1947,
+      "step": 1195
+    },
+    {
+      "epoch": 2.21,
+      "learning_rate": 1.2172800000000003e-05,
+      "loss": 2.0347,
+      "step": 1196
+    },
+    {
+      "epoch": 2.21,
+      "learning_rate": 1.2169600000000003e-05,
+      "loss": 2.0245,
+      "step": 1197
+    },
+    {
+      "epoch": 2.21,
+      "learning_rate": 1.2166400000000001e-05,
+      "loss": 2.0972,
+      "step": 1198
+    },
+    {
+      "epoch": 2.21,
+      "learning_rate": 1.2163200000000002e-05,
+      "loss": 2.1851,
+      "step": 1199
+    },
+    {
+      "epoch": 2.21,
+      "learning_rate": 1.2160000000000002e-05,
+      "loss": 2.1046,
+      "step": 1200
+    },
+    {
+      "epoch": 2.21,
+      "eval_loss": 2.2890968322753906,
+      "eval_runtime": 232.5178,
+      "eval_samples_per_second": 37.3,
+      "eval_steps_per_second": 0.585,
+      "step": 1200
+    },
+    {
+      "epoch": 2.22,
+      "learning_rate": 1.2156800000000002e-05,
+      "loss": 2.2643,
+      "step": 1201
+    },
+    {
+      "epoch": 2.22,
+      "learning_rate": 1.2153600000000003e-05,
+      "loss": 2.1066,
+      "step": 1202
+    },
+    {
+      "epoch": 2.22,
+      "learning_rate": 1.2150400000000001e-05,
+      "loss": 2.1068,
+      "step": 1203
+    },
+    {
+      "epoch": 2.22,
+      "learning_rate": 1.2147200000000001e-05,
+      "loss": 2.1341,
+      "step": 1204
+    },
+    {
+      "epoch": 2.22,
+      "learning_rate": 1.2144000000000003e-05,
+      "loss": 2.1408,
+      "step": 1205
+    },
+    {
+      "epoch": 2.22,
+      "learning_rate": 1.2140800000000002e-05,
+      "loss": 2.162,
+      "step": 1206
+    },
+    {
+      "epoch": 2.23,
+      "learning_rate": 1.2137600000000002e-05,
+      "loss": 2.113,
+      "step": 1207
+    },
+    {
+      "epoch": 2.23,
+      "learning_rate": 1.2134400000000002e-05,
+      "loss": 2.1179,
+      "step": 1208
+    },
+    {
+      "epoch": 2.23,
+      "learning_rate": 1.2131200000000002e-05,
+      "loss": 2.0839,
+      "step": 1209
+    },
+    {
+      "epoch": 2.23,
+      "learning_rate": 1.2128000000000001e-05,
+      "loss": 2.0123,
+      "step": 1210
+    },
+    {
+      "epoch": 2.23,
+      "learning_rate": 1.2124800000000003e-05,
+      "loss": 2.0599,
+      "step": 1211
+    },
+    {
+      "epoch": 2.24,
+      "learning_rate": 1.2121600000000003e-05,
+      "loss": 2.1564,
+      "step": 1212
+    },
+    {
+      "epoch": 2.24,
+      "learning_rate": 1.21184e-05,
+      "loss": 2.1631,
+      "step": 1213
+    },
+    {
+      "epoch": 2.24,
+      "learning_rate": 1.2115200000000002e-05,
+      "loss": 2.1618,
+      "step": 1214
+    },
+    {
+      "epoch": 2.24,
+      "learning_rate": 1.2112000000000002e-05,
+      "loss": 2.1799,
+      "step": 1215
+    },
+    {
+      "epoch": 2.24,
+      "learning_rate": 1.2108800000000002e-05,
+      "loss": 2.2024,
+      "step": 1216
+    },
+    {
+      "epoch": 2.25,
+      "learning_rate": 1.2105600000000003e-05,
+      "loss": 2.1256,
+      "step": 1217
+    },
+    {
+      "epoch": 2.25,
+      "learning_rate": 1.2102400000000001e-05,
+      "loss": 2.1213,
+      "step": 1218
+    },
+    {
+      "epoch": 2.25,
+      "learning_rate": 1.2099200000000001e-05,
+      "loss": 2.117,
+      "step": 1219
+    },
+    {
+      "epoch": 2.25,
+      "learning_rate": 1.2096000000000002e-05,
+      "loss": 2.1936,
+      "step": 1220
+    },
+    {
+      "epoch": 2.25,
+      "learning_rate": 1.2092800000000002e-05,
+      "loss": 2.1933,
+      "step": 1221
+    },
+    {
+      "epoch": 2.25,
+      "learning_rate": 1.2089600000000004e-05,
+      "loss": 2.1849,
+      "step": 1222
+    },
+    {
+      "epoch": 2.26,
+      "learning_rate": 1.2086400000000002e-05,
+      "loss": 2.082,
+      "step": 1223
+    },
+    {
+      "epoch": 2.26,
+      "learning_rate": 1.2083200000000001e-05,
+      "loss": 2.0907,
+      "step": 1224
+    },
+    {
+      "epoch": 2.26,
+      "learning_rate": 1.2080000000000003e-05,
+      "loss": 2.1038,
+      "step": 1225
+    },
+    {
+      "epoch": 2.26,
+      "learning_rate": 1.2076800000000003e-05,
+      "loss": 2.1621,
+      "step": 1226
+    },
+    {
+      "epoch": 2.26,
+      "learning_rate": 1.2073600000000002e-05,
+      "loss": 2.0994,
+      "step": 1227
+    },
+    {
+      "epoch": 2.27,
+      "learning_rate": 1.2070400000000002e-05,
+      "loss": 2.0911,
+      "step": 1228
+    },
+    {
+      "epoch": 2.27,
+      "learning_rate": 1.2067200000000002e-05,
+      "loss": 2.1435,
+      "step": 1229
+    },
+    {
+      "epoch": 2.27,
+      "learning_rate": 1.2064000000000002e-05,
+      "loss": 2.0924,
+      "step": 1230
+    },
+    {
+      "epoch": 2.27,
+      "eval_loss": 2.2888803482055664,
+      "eval_runtime": 233.6442,
+      "eval_samples_per_second": 37.121,
+      "eval_steps_per_second": 0.582,
+      "step": 1230
+    },
+    {
+      "epoch": 2.27,
+      "learning_rate": 1.2060800000000003e-05,
+      "loss": 2.0851,
+      "step": 1231
+    },
+    {
+      "epoch": 2.27,
+      "learning_rate": 1.2057600000000003e-05,
+      "loss": 2.1002,
+      "step": 1232
+    },
+    {
+      "epoch": 2.27,
+      "learning_rate": 1.2054400000000001e-05,
+      "loss": 2.0955,
+      "step": 1233
+    },
+    {
+      "epoch": 2.28,
+      "learning_rate": 1.2051200000000002e-05,
+      "loss": 2.1041,
+      "step": 1234
+    },
+    {
+      "epoch": 2.28,
+      "learning_rate": 1.2048000000000002e-05,
+      "loss": 2.067,
+      "step": 1235
+    },
+    {
+      "epoch": 2.28,
+      "learning_rate": 1.2044800000000002e-05,
+      "loss": 2.085,
+      "step": 1236
+    },
+    {
+      "epoch": 2.28,
+      "learning_rate": 1.2041600000000003e-05,
+      "loss": 2.0101,
+      "step": 1237
+    },
+    {
+      "epoch": 2.28,
+      "learning_rate": 1.2038400000000001e-05,
+      "loss": 2.0721,
+      "step": 1238
+    },
+    {
+      "epoch": 2.29,
+      "learning_rate": 1.2035200000000001e-05,
+      "loss": 2.1312,
+      "step": 1239
+    },
+    {
+      "epoch": 2.29,
+      "learning_rate": 1.2032000000000003e-05,
+      "loss": 2.038,
+      "step": 1240
+    },
+    {
+      "epoch": 2.29,
+      "learning_rate": 1.2028800000000002e-05,
+      "loss": 2.0453,
+      "step": 1241
+    },
+    {
+      "epoch": 2.29,
+      "learning_rate": 1.2025600000000002e-05,
+      "loss": 2.2162,
+      "step": 1242
+    },
+    {
+      "epoch": 2.29,
+      "learning_rate": 1.2022400000000002e-05,
+      "loss": 2.1599,
+      "step": 1243
+    },
+    {
+      "epoch": 2.29,
+      "learning_rate": 1.2019200000000002e-05,
+      "loss": 2.1203,
+      "step": 1244
+    },
+    {
+      "epoch": 2.3,
+      "learning_rate": 1.2016000000000002e-05,
+      "loss": 2.0944,
+      "step": 1245
+    },
+    {
+      "epoch": 2.3,
+      "learning_rate": 1.2012800000000003e-05,
+      "loss": 2.2567,
+      "step": 1246
+    },
+    {
+      "epoch": 2.3,
+      "learning_rate": 1.2009600000000003e-05,
+      "loss": 2.2627,
+      "step": 1247
+    },
+    {
+      "epoch": 2.3,
+      "learning_rate": 1.20064e-05,
+      "loss": 2.1495,
+      "step": 1248
+    },
+    {
+      "epoch": 2.3,
+      "learning_rate": 1.2003200000000002e-05,
+      "loss": 2.0678,
+      "step": 1249
+    },
+    {
+      "epoch": 2.31,
+      "learning_rate": 1.2000000000000002e-05,
+      "loss": 2.1147,
+      "step": 1250
+    },
+    {
+      "epoch": 2.31,
+      "learning_rate": 1.1996800000000002e-05,
+      "loss": 2.1433,
+      "step": 1251
+    },
+    {
+      "epoch": 2.31,
+      "learning_rate": 1.1993600000000003e-05,
+      "loss": 2.1132,
+      "step": 1252
+    },
+    {
+      "epoch": 2.31,
+      "learning_rate": 1.1990400000000001e-05,
+      "loss": 2.268,
+      "step": 1253
+    },
+    {
+      "epoch": 2.31,
+      "learning_rate": 1.1987200000000001e-05,
+      "loss": 2.1387,
+      "step": 1254
+    },
+    {
+      "epoch": 2.32,
+      "learning_rate": 1.1984000000000002e-05,
+      "loss": 2.0616,
+      "step": 1255
+    },
+    {
+      "epoch": 2.32,
+      "learning_rate": 1.1980800000000002e-05,
+      "loss": 2.1842,
+      "step": 1256
+    },
+    {
+      "epoch": 2.32,
+      "learning_rate": 1.1977600000000002e-05,
+      "loss": 2.0923,
+      "step": 1257
+    },
+    {
+      "epoch": 2.32,
+      "learning_rate": 1.1974400000000002e-05,
+      "loss": 2.1285,
+      "step": 1258
+    },
+    {
+      "epoch": 2.32,
+      "learning_rate": 1.1971200000000002e-05,
+      "loss": 2.1212,
+      "step": 1259
+    },
+    {
+      "epoch": 2.32,
+      "learning_rate": 1.1968000000000001e-05,
+      "loss": 2.1304,
+      "step": 1260
+    },
+    {
+      "epoch": 2.32,
+      "eval_loss": 2.287320852279663,
+      "eval_runtime": 232.3251,
+      "eval_samples_per_second": 37.331,
+      "eval_steps_per_second": 0.585,
+      "step": 1260
+    },
+    {
+      "epoch": 2.33,
+      "learning_rate": 1.1964800000000003e-05,
+      "loss": 2.2062,
+      "step": 1261
+    },
+    {
+      "epoch": 2.33,
+      "learning_rate": 1.1961600000000003e-05,
+      "loss": 2.1961,
+      "step": 1262
+    },
+    {
+      "epoch": 2.33,
+      "learning_rate": 1.19584e-05,
+      "loss": 2.058,
+      "step": 1263
+    },
+    {
+      "epoch": 2.33,
+      "learning_rate": 1.1955200000000002e-05,
+      "loss": 2.0563,
+      "step": 1264
+    },
+    {
+      "epoch": 2.33,
+      "learning_rate": 1.1952000000000002e-05,
+      "loss": 2.0918,
+      "step": 1265
+    },
+    {
+      "epoch": 2.34,
+      "learning_rate": 1.1948800000000002e-05,
+      "loss": 2.1321,
+      "step": 1266
+    },
+    {
+      "epoch": 2.34,
+      "learning_rate": 1.1945600000000003e-05,
+      "loss": 2.0254,
+      "step": 1267
+    },
+    {
+      "epoch": 2.34,
+      "learning_rate": 1.1942400000000001e-05,
+      "loss": 2.0884,
+      "step": 1268
+    },
+    {
+      "epoch": 2.34,
+      "learning_rate": 1.19392e-05,
+      "loss": 2.0407,
+      "step": 1269
+    },
+    {
+      "epoch": 2.34,
+      "learning_rate": 1.1936000000000002e-05,
+      "loss": 2.0143,
+      "step": 1270
+    },
+    {
+      "epoch": 2.34,
+      "learning_rate": 1.1932800000000002e-05,
+      "loss": 2.0913,
+      "step": 1271
+    },
+    {
+      "epoch": 2.35,
+      "learning_rate": 1.1929600000000002e-05,
+      "loss": 2.0589,
+      "step": 1272
+    },
+    {
+      "epoch": 2.35,
+      "learning_rate": 1.1926400000000002e-05,
+      "loss": 2.1104,
+      "step": 1273
+    },
+    {
+      "epoch": 2.35,
+      "learning_rate": 1.1923200000000001e-05,
+      "loss": 2.0615,
+      "step": 1274
+    },
+    {
+      "epoch": 2.35,
+      "learning_rate": 1.1920000000000001e-05,
+      "loss": 2.1406,
+      "step": 1275
+    },
+    {
+      "epoch": 2.35,
+      "learning_rate": 1.1916800000000003e-05,
+      "loss": 2.1886,
+      "step": 1276
+    },
+    {
+      "epoch": 2.36,
+      "learning_rate": 1.1913600000000002e-05,
+      "loss": 2.0376,
+      "step": 1277
+    },
+    {
+      "epoch": 2.36,
+      "learning_rate": 1.19104e-05,
+      "loss": 2.1103,
+      "step": 1278
+    },
+    {
+      "epoch": 2.36,
+      "learning_rate": 1.1907200000000002e-05,
+      "loss": 2.1855,
+      "step": 1279
+    },
+    {
+      "epoch": 2.36,
+      "learning_rate": 1.1904000000000002e-05,
+      "loss": 2.0736,
+      "step": 1280
+    },
+    {
+      "epoch": 2.36,
+      "learning_rate": 1.1900800000000003e-05,
+      "loss": 2.1127,
+      "step": 1281
+    },
+    {
+      "epoch": 2.36,
+      "learning_rate": 1.1897600000000003e-05,
+      "loss": 2.2324,
+      "step": 1282
+    },
+    {
+      "epoch": 2.37,
+      "learning_rate": 1.18944e-05,
+      "loss": 2.1625,
+      "step": 1283
+    },
+    {
+      "epoch": 2.37,
+      "learning_rate": 1.1891200000000002e-05,
+      "loss": 2.1008,
+      "step": 1284
+    },
+    {
+      "epoch": 2.37,
+      "learning_rate": 1.1888000000000002e-05,
+      "loss": 2.0907,
+      "step": 1285
+    },
+    {
+      "epoch": 2.37,
+      "learning_rate": 1.1884800000000002e-05,
+      "loss": 2.0833,
+      "step": 1286
+    },
+    {
+      "epoch": 2.37,
+      "learning_rate": 1.1881600000000003e-05,
+      "loss": 2.1576,
+      "step": 1287
+    },
+    {
+      "epoch": 2.38,
+      "learning_rate": 1.1878400000000001e-05,
+      "loss": 2.2006,
+      "step": 1288
+    },
+    {
+      "epoch": 2.38,
+      "learning_rate": 1.1875200000000001e-05,
+      "loss": 2.216,
+      "step": 1289
+    },
+    {
+      "epoch": 2.38,
+      "learning_rate": 1.1872000000000003e-05,
+      "loss": 2.1042,
+      "step": 1290
+    },
+    {
+      "epoch": 2.38,
+      "eval_loss": 2.2952046394348145,
+      "eval_runtime": 232.5677,
+      "eval_samples_per_second": 37.292,
+      "eval_steps_per_second": 0.585,
+      "step": 1290
+    },
+    {
+      "epoch": 2.38,
+      "learning_rate": 1.1868800000000002e-05,
+      "loss": 2.1665,
+      "step": 1291
+    },
+    {
+      "epoch": 2.38,
+      "learning_rate": 1.1865600000000002e-05,
+      "loss": 2.1033,
+      "step": 1292
+    },
+    {
+      "epoch": 2.39,
+      "learning_rate": 1.1862400000000002e-05,
+      "loss": 2.0517,
+      "step": 1293
+    },
+    {
+      "epoch": 2.39,
+      "learning_rate": 1.1859200000000002e-05,
+      "loss": 2.1015,
+      "step": 1294
+    },
+    {
+      "epoch": 2.39,
+      "learning_rate": 1.1856000000000001e-05,
+      "loss": 2.0766,
+      "step": 1295
+    },
+    {
+      "epoch": 2.39,
+      "learning_rate": 1.1852800000000003e-05,
+      "loss": 2.1794,
+      "step": 1296
+    },
+    {
+      "epoch": 2.39,
+      "learning_rate": 1.1849600000000003e-05,
+      "loss": 2.1876,
+      "step": 1297
+    },
+    {
+      "epoch": 2.39,
+      "learning_rate": 1.18464e-05,
+      "loss": 2.1707,
+      "step": 1298
+    },
+    {
+      "epoch": 2.4,
+      "learning_rate": 1.1843200000000002e-05,
+      "loss": 2.219,
+      "step": 1299
+    },
+    {
+      "epoch": 2.4,
+      "learning_rate": 1.1840000000000002e-05,
+      "loss": 2.1,
+      "step": 1300
+    },
+    {
+      "epoch": 2.4,
+      "learning_rate": 1.1836800000000002e-05,
+      "loss": 2.1269,
+      "step": 1301
+    },
+    {
+      "epoch": 2.4,
+      "learning_rate": 1.1833600000000003e-05,
+      "loss": 2.132,
+      "step": 1302
+    },
+    {
+      "epoch": 2.4,
+      "learning_rate": 1.1830400000000001e-05,
+      "loss": 2.1576,
+      "step": 1303
+    },
+    {
+      "epoch": 2.41,
+      "learning_rate": 1.1827200000000001e-05,
+      "loss": 2.1939,
+      "step": 1304
+    },
+    {
+      "epoch": 2.41,
+      "learning_rate": 1.1824000000000002e-05,
+      "loss": 2.0796,
+      "step": 1305
+    },
+    {
+      "epoch": 2.41,
+      "learning_rate": 1.1820800000000002e-05,
+      "loss": 1.9778,
+      "step": 1306
+    },
+    {
+      "epoch": 2.41,
+      "learning_rate": 1.1817600000000002e-05,
+      "loss": 2.0916,
+      "step": 1307
+    },
+    {
+      "epoch": 2.41,
+      "learning_rate": 1.1814400000000002e-05,
+      "loss": 2.0394,
+      "step": 1308
+    },
+    {
+      "epoch": 2.41,
+      "learning_rate": 1.1811200000000001e-05,
+      "loss": 2.0952,
+      "step": 1309
+    },
+    {
+      "epoch": 2.42,
+      "learning_rate": 1.1808000000000001e-05,
+      "loss": 2.059,
+      "step": 1310
+    },
+    {
+      "epoch": 2.42,
+      "learning_rate": 1.1804800000000003e-05,
+      "loss": 2.1795,
+      "step": 1311
+    },
+    {
+      "epoch": 2.42,
+      "learning_rate": 1.1801600000000002e-05,
+      "loss": 2.1999,
+      "step": 1312
+    },
+    {
+      "epoch": 2.42,
+      "learning_rate": 1.1798400000000002e-05,
+      "loss": 1.9934,
+      "step": 1313
+    },
+    {
+      "epoch": 2.42,
+      "learning_rate": 1.1795200000000002e-05,
+      "loss": 2.0544,
+      "step": 1314
+    },
+    {
+      "epoch": 2.43,
+      "learning_rate": 1.1792000000000002e-05,
+      "loss": 2.0816,
+      "step": 1315
+    },
+    {
+      "epoch": 2.43,
+      "learning_rate": 1.1788800000000001e-05,
+      "loss": 2.2387,
+      "step": 1316
+    },
+    {
+      "epoch": 2.43,
+      "learning_rate": 1.1785600000000003e-05,
+      "loss": 2.0259,
+      "step": 1317
+    },
+    {
+      "epoch": 2.43,
+      "learning_rate": 1.1782400000000003e-05,
+      "loss": 2.1356,
+      "step": 1318
+    },
+    {
+      "epoch": 2.43,
+      "learning_rate": 1.17792e-05,
+      "loss": 2.1602,
+      "step": 1319
+    },
+    {
+      "epoch": 2.44,
+      "learning_rate": 1.1776000000000002e-05,
+      "loss": 2.1413,
+      "step": 1320
+    },
+    {
+      "epoch": 2.44,
+      "eval_loss": 2.287919521331787,
+      "eval_runtime": 231.6908,
+      "eval_samples_per_second": 37.434,
+      "eval_steps_per_second": 0.587,
+      "step": 1320
+    },
+    {
+      "epoch": 2.44,
+      "learning_rate": 1.1772800000000002e-05,
+      "loss": 2.1976,
+      "step": 1321
+    },
+    {
+      "epoch": 2.44,
+      "learning_rate": 1.1769600000000002e-05,
+      "loss": 2.0816,
+      "step": 1322
+    },
+    {
+      "epoch": 2.44,
+      "learning_rate": 1.1766400000000003e-05,
+      "loss": 2.1756,
+      "step": 1323
+    },
+    {
+      "epoch": 2.44,
+      "learning_rate": 1.1763200000000001e-05,
+      "loss": 2.1943,
+      "step": 1324
+    },
+    {
+      "epoch": 2.44,
+      "learning_rate": 1.1760000000000001e-05,
+      "loss": 2.0805,
+      "step": 1325
+    },
+    {
+      "epoch": 2.45,
+      "learning_rate": 1.1756800000000002e-05,
+      "loss": 2.0479,
+      "step": 1326
+    },
+    {
+      "epoch": 2.45,
+      "learning_rate": 1.1753600000000002e-05,
+      "loss": 2.11,
+      "step": 1327
+    },
+    {
+      "epoch": 2.45,
+      "learning_rate": 1.1750400000000004e-05,
+      "loss": 2.0939,
+      "step": 1328
+    },
+    {
+      "epoch": 2.45,
+      "learning_rate": 1.1747200000000002e-05,
+      "loss": 2.1593,
+      "step": 1329
+    },
+    {
+      "epoch": 2.45,
+      "learning_rate": 1.1744000000000001e-05,
+      "loss": 2.1429,
+      "step": 1330
+    },
+    {
+      "epoch": 2.46,
+      "learning_rate": 1.1740800000000003e-05,
+      "loss": 2.1672,
+      "step": 1331
+    },
+    {
+      "epoch": 2.46,
+      "learning_rate": 1.1737600000000003e-05,
+      "loss": 2.1366,
+      "step": 1332
+    },
+    {
+      "epoch": 2.46,
+      "learning_rate": 1.1734400000000002e-05,
+      "loss": 2.1274,
+      "step": 1333
+    },
+    {
+      "epoch": 2.46,
+      "learning_rate": 1.1731200000000002e-05,
+      "loss": 2.2228,
+      "step": 1334
+    },
+    {
+      "epoch": 2.46,
+      "learning_rate": 1.1728000000000002e-05,
+      "loss": 2.1355,
+      "step": 1335
+    },
+    {
+      "epoch": 2.46,
+      "learning_rate": 1.1724800000000002e-05,
+      "loss": 2.2056,
+      "step": 1336
+    },
+    {
+      "epoch": 2.47,
+      "learning_rate": 1.1721600000000003e-05,
+      "loss": 2.1521,
+      "step": 1337
+    },
+    {
+      "epoch": 2.47,
+      "learning_rate": 1.1718400000000003e-05,
+      "loss": 2.0508,
+      "step": 1338
+    },
+    {
+      "epoch": 2.47,
+      "learning_rate": 1.1715200000000001e-05,
+      "loss": 2.0271,
+      "step": 1339
+    },
+    {
+      "epoch": 2.47,
+      "learning_rate": 1.1712000000000002e-05,
+      "loss": 2.1329,
+      "step": 1340
+    },
+    {
+      "epoch": 2.47,
+      "learning_rate": 1.1708800000000002e-05,
+      "loss": 2.1559,
+      "step": 1341
+    },
+    {
+      "epoch": 2.48,
+      "learning_rate": 1.1705600000000002e-05,
+      "loss": 2.094,
+      "step": 1342
+    },
+    {
+      "epoch": 2.48,
+      "learning_rate": 1.1702400000000003e-05,
+      "loss": 2.0863,
+      "step": 1343
+    },
+    {
+      "epoch": 2.48,
+      "learning_rate": 1.1699200000000001e-05,
+      "loss": 2.2716,
+      "step": 1344
+    },
+    {
+      "epoch": 2.48,
+      "learning_rate": 1.1696000000000001e-05,
+      "loss": 2.055,
+      "step": 1345
+    },
+    {
+      "epoch": 2.48,
+      "learning_rate": 1.1692800000000003e-05,
+      "loss": 2.0354,
+      "step": 1346
+    },
+    {
+      "epoch": 2.48,
+      "learning_rate": 1.1689600000000002e-05,
+      "loss": 2.1294,
+      "step": 1347
+    },
+    {
+      "epoch": 2.49,
+      "learning_rate": 1.1686400000000002e-05,
+      "loss": 2.1627,
+      "step": 1348
+    },
+    {
+      "epoch": 2.49,
+      "learning_rate": 1.1683200000000002e-05,
+      "loss": 2.0735,
+      "step": 1349
+    },
+    {
+      "epoch": 2.49,
+      "learning_rate": 1.1680000000000002e-05,
+      "loss": 2.1578,
+      "step": 1350
+    },
+    {
+      "epoch": 2.49,
+      "eval_loss": 2.287665367126465,
+      "eval_runtime": 232.6805,
+      "eval_samples_per_second": 37.274,
+      "eval_steps_per_second": 0.584,
+      "step": 1350
+    },
+    {
+      "epoch": 2.49,
+      "learning_rate": 1.1676800000000001e-05,
+      "loss": 2.04,
+      "step": 1351
+    },
+    {
+      "epoch": 2.49,
+      "learning_rate": 1.1673600000000003e-05,
+      "loss": 2.0384,
+      "step": 1352
+    },
+    {
+      "epoch": 2.5,
+      "learning_rate": 1.1670400000000003e-05,
+      "loss": 2.1205,
+      "step": 1353
+    },
+    {
+      "epoch": 2.5,
+      "learning_rate": 1.16672e-05,
+      "loss": 2.2474,
+      "step": 1354
+    },
+    {
+      "epoch": 2.5,
+      "learning_rate": 1.1664000000000002e-05,
+      "loss": 2.0858,
+      "step": 1355
+    },
+    {
+      "epoch": 2.5,
+      "learning_rate": 1.1660800000000002e-05,
+      "loss": 2.1499,
+      "step": 1356
+    },
+    {
+      "epoch": 2.5,
+      "learning_rate": 1.1657600000000002e-05,
+      "loss": 2.0534,
+      "step": 1357
+    },
+    {
+      "epoch": 2.51,
+      "learning_rate": 1.1654400000000003e-05,
+      "loss": 2.1181,
+      "step": 1358
+    },
+    {
+      "epoch": 2.51,
+      "learning_rate": 1.1651200000000001e-05,
+      "loss": 2.1329,
+      "step": 1359
+    },
+    {
+      "epoch": 2.51,
+      "learning_rate": 1.1648000000000001e-05,
+      "loss": 2.0602,
+      "step": 1360
+    },
+    {
+      "epoch": 2.51,
+      "learning_rate": 1.1644800000000002e-05,
+      "loss": 2.1612,
+      "step": 1361
+    },
+    {
+      "epoch": 2.51,
+      "learning_rate": 1.1641600000000002e-05,
+      "loss": 2.2464,
+      "step": 1362
+    },
+    {
+      "epoch": 2.51,
+      "learning_rate": 1.1638400000000002e-05,
+      "loss": 2.0919,
+      "step": 1363
+    },
+    {
+      "epoch": 2.52,
+      "learning_rate": 1.1635200000000002e-05,
+      "loss": 2.0862,
+      "step": 1364
+    },
+    {
+      "epoch": 2.52,
+      "learning_rate": 1.1632000000000001e-05,
+      "loss": 2.1087,
+      "step": 1365
+    },
+    {
+      "epoch": 2.52,
+      "learning_rate": 1.1628800000000001e-05,
+      "loss": 2.0895,
+      "step": 1366
+    },
+    {
+      "epoch": 2.52,
+      "learning_rate": 1.1625600000000003e-05,
+      "loss": 2.0956,
+      "step": 1367
+    },
+    {
+      "epoch": 2.52,
+      "learning_rate": 1.1622400000000002e-05,
+      "loss": 2.1388,
+      "step": 1368
+    },
+    {
+      "epoch": 2.53,
+      "learning_rate": 1.16192e-05,
+      "loss": 2.1653,
+      "step": 1369
+    },
+    {
+      "epoch": 2.53,
+      "learning_rate": 1.1616000000000002e-05,
+      "loss": 2.0394,
+      "step": 1370
+    },
+    {
+      "epoch": 2.53,
+      "learning_rate": 1.1612800000000002e-05,
+      "loss": 2.1593,
+      "step": 1371
+    },
+    {
+      "epoch": 2.53,
+      "learning_rate": 1.1609600000000001e-05,
+      "loss": 2.1207,
+      "step": 1372
+    },
+    {
+      "epoch": 2.53,
+      "learning_rate": 1.1606400000000003e-05,
+      "loss": 2.1,
+      "step": 1373
+    },
+    {
+      "epoch": 2.53,
+      "learning_rate": 1.1603200000000001e-05,
+      "loss": 2.2152,
+      "step": 1374
+    },
+    {
+      "epoch": 2.54,
+      "learning_rate": 1.16e-05,
+      "loss": 2.1517,
+      "step": 1375
+    },
+    {
+      "epoch": 2.54,
+      "learning_rate": 1.1596800000000002e-05,
+      "loss": 2.1265,
+      "step": 1376
+    },
+    {
+      "epoch": 2.54,
+      "learning_rate": 1.1593600000000002e-05,
+      "loss": 2.1268,
+      "step": 1377
+    },
+    {
+      "epoch": 2.54,
+      "learning_rate": 1.1590400000000003e-05,
+      "loss": 2.1924,
+      "step": 1378
+    },
+    {
+      "epoch": 2.54,
+      "learning_rate": 1.1587200000000001e-05,
+      "loss": 2.0224,
+      "step": 1379
+    },
+    {
+      "epoch": 2.55,
+      "learning_rate": 1.1584000000000001e-05,
+      "loss": 2.0671,
+      "step": 1380
+    },
+    {
+      "epoch": 2.55,
+      "eval_loss": 2.288414478302002,
+      "eval_runtime": 230.9965,
+      "eval_samples_per_second": 37.546,
+      "eval_steps_per_second": 0.589,
+      "step": 1380
+    },
+    {
+      "epoch": 2.55,
+      "learning_rate": 1.1580800000000003e-05,
+      "loss": 2.3072,
+      "step": 1381
+    },
+    {
+      "epoch": 2.55,
+      "learning_rate": 1.1577600000000002e-05,
+      "loss": 2.0782,
+      "step": 1382
+    },
+    {
+      "epoch": 2.55,
+      "learning_rate": 1.1574400000000002e-05,
+      "loss": 2.1299,
+      "step": 1383
+    },
+    {
+      "epoch": 2.55,
+      "learning_rate": 1.1571200000000002e-05,
+      "loss": 2.0501,
+      "step": 1384
+    },
+    {
+      "epoch": 2.55,
+      "learning_rate": 1.1568000000000002e-05,
+      "loss": 2.169,
+      "step": 1385
+    },
+    {
+      "epoch": 2.56,
+      "learning_rate": 1.1564800000000001e-05,
+      "loss": 2.1563,
+      "step": 1386
+    },
+    {
+      "epoch": 2.56,
+      "learning_rate": 1.1561600000000003e-05,
+      "loss": 2.0898,
+      "step": 1387
+    },
+    {
+      "epoch": 2.56,
+      "learning_rate": 1.1558400000000003e-05,
+      "loss": 2.2091,
+      "step": 1388
+    },
+    {
+      "epoch": 2.56,
+      "learning_rate": 1.15552e-05,
+      "loss": 2.031,
+      "step": 1389
+    },
+    {
+      "epoch": 2.56,
+      "learning_rate": 1.1552000000000002e-05,
+      "loss": 2.1617,
+      "step": 1390
+    },
+    {
+      "epoch": 2.57,
+      "learning_rate": 1.1548800000000002e-05,
+      "loss": 2.1021,
+      "step": 1391
+    },
+    {
+      "epoch": 2.57,
+      "learning_rate": 1.1545600000000002e-05,
+      "loss": 2.1394,
+      "step": 1392
+    },
+    {
+      "epoch": 2.57,
+      "learning_rate": 1.1542400000000003e-05,
+      "loss": 1.9802,
+      "step": 1393
+    },
+    {
+      "epoch": 2.57,
+      "learning_rate": 1.1539200000000001e-05,
+      "loss": 2.0452,
+      "step": 1394
+    },
+    {
+      "epoch": 2.57,
+      "learning_rate": 1.1536000000000001e-05,
+      "loss": 2.1448,
+      "step": 1395
+    },
+    {
+      "epoch": 2.58,
+      "learning_rate": 1.1532800000000002e-05,
+      "loss": 2.0834,
+      "step": 1396
+    },
+    {
+      "epoch": 2.58,
+      "learning_rate": 1.1529600000000002e-05,
+      "loss": 2.0202,
+      "step": 1397
+    },
+    {
+      "epoch": 2.58,
+      "learning_rate": 1.1526400000000002e-05,
+      "loss": 1.9655,
+      "step": 1398
+    },
+    {
+      "epoch": 2.58,
+      "learning_rate": 1.1523200000000002e-05,
+      "loss": 2.1787,
+      "step": 1399
+    },
+    {
+      "epoch": 2.58,
+      "learning_rate": 1.1520000000000002e-05,
+      "loss": 2.1064,
+      "step": 1400
+    },
+    {
+      "epoch": 2.58,
+      "learning_rate": 1.1516800000000001e-05,
+      "loss": 2.1086,
+      "step": 1401
+    },
+    {
+      "epoch": 2.59,
+      "learning_rate": 1.1513600000000003e-05,
+      "loss": 2.1583,
+      "step": 1402
+    },
+    {
+      "epoch": 2.59,
+      "learning_rate": 1.1510400000000002e-05,
+      "loss": 2.1197,
+      "step": 1403
+    },
+    {
+      "epoch": 2.59,
+      "learning_rate": 1.15072e-05,
+      "loss": 2.0729,
+      "step": 1404
+    },
+    {
+      "epoch": 2.59,
+      "learning_rate": 1.1504000000000002e-05,
+      "loss": 2.1323,
+      "step": 1405
+    },
+    {
+      "epoch": 2.59,
+      "learning_rate": 1.1500800000000002e-05,
+      "loss": 2.1916,
+      "step": 1406
+    },
+    {
+      "epoch": 2.6,
+      "learning_rate": 1.1497600000000002e-05,
+      "loss": 2.1636,
+      "step": 1407
+    },
+    {
+      "epoch": 2.6,
+      "learning_rate": 1.1494400000000003e-05,
+      "loss": 2.0634,
+      "step": 1408
+    },
+    {
+      "epoch": 2.6,
+      "learning_rate": 1.1491200000000001e-05,
+      "loss": 2.1394,
+      "step": 1409
+    },
+    {
+      "epoch": 2.6,
+      "learning_rate": 1.1488e-05,
+      "loss": 2.1116,
+      "step": 1410
+    },
+    {
+      "epoch": 2.6,
+      "eval_loss": 2.2905826568603516,
+      "eval_runtime": 230.9961,
+      "eval_samples_per_second": 37.546,
+      "eval_steps_per_second": 0.589,
+      "step": 1410
+    },
+    {
+      "epoch": 2.6,
+      "learning_rate": 1.1484800000000002e-05,
+      "loss": 2.146,
+      "step": 1411
+    },
+    {
+      "epoch": 2.6,
+      "learning_rate": 1.1481600000000002e-05,
+      "loss": 1.9689,
+      "step": 1412
+    },
+    {
+      "epoch": 2.61,
+      "learning_rate": 1.1478400000000002e-05,
+      "loss": 2.1207,
+      "step": 1413
+    },
+    {
+      "epoch": 2.61,
+      "learning_rate": 1.1475200000000002e-05,
+      "loss": 2.0631,
+      "step": 1414
+    },
+    {
+      "epoch": 2.61,
+      "learning_rate": 1.1472000000000001e-05,
+      "loss": 2.1355,
+      "step": 1415
+    },
+    {
+      "epoch": 2.61,
+      "learning_rate": 1.1468800000000001e-05,
+      "loss": 2.1174,
+      "step": 1416
+    },
+    {
+      "epoch": 2.61,
+      "learning_rate": 1.1465600000000003e-05,
+      "loss": 1.9836,
+      "step": 1417
+    },
+    {
+      "epoch": 2.62,
+      "learning_rate": 1.1462400000000002e-05,
+      "loss": 2.1065,
+      "step": 1418
+    },
+    {
+      "epoch": 2.62,
+      "learning_rate": 1.14592e-05,
+      "loss": 2.0627,
+      "step": 1419
+    },
+    {
+      "epoch": 2.62,
+      "learning_rate": 1.1456000000000002e-05,
+      "loss": 2.1303,
+      "step": 1420
+    },
+    {
+      "epoch": 2.62,
+      "learning_rate": 1.1452800000000002e-05,
+      "loss": 2.1067,
+      "step": 1421
+    },
+    {
+      "epoch": 2.62,
+      "learning_rate": 1.1449600000000001e-05,
+      "loss": 2.1272,
+      "step": 1422
+    },
+    {
+      "epoch": 2.62,
+      "learning_rate": 1.1446400000000003e-05,
+      "loss": 2.0576,
+      "step": 1423
+    },
+    {
+      "epoch": 2.63,
+      "learning_rate": 1.14432e-05,
+      "loss": 2.096,
+      "step": 1424
+    },
+    {
+      "epoch": 2.63,
+      "learning_rate": 1.144e-05,
+      "loss": 2.0924,
+      "step": 1425
+    },
+    {
+      "epoch": 2.63,
+      "learning_rate": 1.1436800000000002e-05,
+      "loss": 2.0442,
+      "step": 1426
+    },
+    {
+      "epoch": 2.63,
+      "learning_rate": 1.1433600000000002e-05,
+      "loss": 2.2118,
+      "step": 1427
+    },
+    {
+      "epoch": 2.63,
+      "learning_rate": 1.1430400000000003e-05,
+      "loss": 2.1159,
+      "step": 1428
+    },
+    {
+      "epoch": 2.64,
+      "learning_rate": 1.1427200000000001e-05,
+      "loss": 1.9823,
+      "step": 1429
+    },
+    {
+      "epoch": 2.64,
+      "learning_rate": 1.1424000000000001e-05,
+      "loss": 2.1221,
+      "step": 1430
+    },
+    {
+      "epoch": 2.64,
+      "learning_rate": 1.1420800000000003e-05,
+      "loss": 2.2007,
+      "step": 1431
+    },
+    {
+      "epoch": 2.64,
+      "learning_rate": 1.1417600000000002e-05,
+      "loss": 2.1018,
+      "step": 1432
+    },
+    {
+      "epoch": 2.64,
+      "learning_rate": 1.1414400000000002e-05,
+      "loss": 2.0707,
+      "step": 1433
+    },
+    {
+      "epoch": 2.65,
+      "learning_rate": 1.1411200000000002e-05,
+      "loss": 2.1127,
+      "step": 1434
+    },
+    {
+      "epoch": 2.65,
+      "learning_rate": 1.1408000000000002e-05,
+      "loss": 2.0932,
+      "step": 1435
+    },
+    {
+      "epoch": 2.65,
+      "learning_rate": 1.1404800000000001e-05,
+      "loss": 2.0444,
+      "step": 1436
+    },
+    {
+      "epoch": 2.65,
+      "learning_rate": 1.1401600000000003e-05,
+      "loss": 2.1346,
+      "step": 1437
+    },
+    {
+      "epoch": 2.65,
+      "learning_rate": 1.1398400000000003e-05,
+      "loss": 2.0532,
+      "step": 1438
+    },
+    {
+      "epoch": 2.65,
+      "learning_rate": 1.1395200000000002e-05,
+      "loss": 2.2054,
+      "step": 1439
+    },
+    {
+      "epoch": 2.66,
+      "learning_rate": 1.1392000000000002e-05,
+      "loss": 2.0191,
+      "step": 1440
+    },
+    {
+      "epoch": 2.66,
+      "eval_loss": 2.2885921001434326,
+      "eval_runtime": 232.0523,
+      "eval_samples_per_second": 37.375,
+      "eval_steps_per_second": 0.586,
+      "step": 1440
+    },
+    {
+      "epoch": 2.66,
+      "learning_rate": 1.1388800000000002e-05,
+      "loss": 2.1125,
+      "step": 1441
+    },
+    {
+      "epoch": 2.66,
+      "learning_rate": 1.1385600000000002e-05,
+      "loss": 2.0603,
+      "step": 1442
+    },
+    {
+      "epoch": 2.66,
+      "learning_rate": 1.1382400000000003e-05,
+      "loss": 2.0588,
+      "step": 1443
+    },
+    {
+      "epoch": 2.66,
+      "learning_rate": 1.1379200000000003e-05,
+      "loss": 2.1747,
+      "step": 1444
+    },
+    {
+      "epoch": 2.67,
+      "learning_rate": 1.1376000000000001e-05,
+      "loss": 2.0796,
+      "step": 1445
+    },
+    {
+      "epoch": 2.67,
+      "learning_rate": 1.1372800000000002e-05,
+      "loss": 2.069,
+      "step": 1446
+    },
+    {
+      "epoch": 2.67,
+      "learning_rate": 1.1369600000000002e-05,
+      "loss": 2.1894,
+      "step": 1447
+    },
+    {
+      "epoch": 2.67,
+      "learning_rate": 1.1366400000000002e-05,
+      "loss": 2.0513,
+      "step": 1448
+    },
+    {
+      "epoch": 2.67,
+      "learning_rate": 1.1363200000000003e-05,
+      "loss": 2.2322,
+      "step": 1449
+    },
+    {
+      "epoch": 2.67,
+      "learning_rate": 1.1360000000000001e-05,
+      "loss": 2.1594,
+      "step": 1450
+    },
+    {
+      "epoch": 2.68,
+      "learning_rate": 1.1356800000000001e-05,
+      "loss": 2.1205,
+      "step": 1451
+    },
+    {
+      "epoch": 2.68,
+      "learning_rate": 1.1353600000000003e-05,
+      "loss": 2.1469,
+      "step": 1452
+    },
+    {
+      "epoch": 2.68,
+      "learning_rate": 1.1350400000000002e-05,
+      "loss": 2.1562,
+      "step": 1453
+    },
+    {
+      "epoch": 2.68,
+      "learning_rate": 1.1347200000000002e-05,
+      "loss": 2.108,
+      "step": 1454
+    },
+    {
+      "epoch": 2.68,
+      "learning_rate": 1.1344000000000002e-05,
+      "loss": 2.0328,
+      "step": 1455
+    },
+    {
+      "epoch": 2.69,
+      "learning_rate": 1.1340800000000002e-05,
+      "loss": 2.0459,
+      "step": 1456
+    },
+    {
+      "epoch": 2.69,
+      "learning_rate": 1.1337600000000001e-05,
+      "loss": 2.1024,
+      "step": 1457
+    },
+    {
+      "epoch": 2.69,
+      "learning_rate": 1.1334400000000003e-05,
+      "loss": 2.215,
+      "step": 1458
+    },
+    {
+      "epoch": 2.69,
+      "learning_rate": 1.1331200000000003e-05,
+      "loss": 2.1646,
+      "step": 1459
+    },
+    {
+      "epoch": 2.69,
+      "learning_rate": 1.1328e-05,
+      "loss": 2.1283,
+      "step": 1460
+    },
+    {
+      "epoch": 2.69,
+      "learning_rate": 1.1324800000000002e-05,
+      "loss": 2.1019,
+      "step": 1461
+    },
+    {
+      "epoch": 2.7,
+      "learning_rate": 1.1321600000000002e-05,
+      "loss": 2.0578,
+      "step": 1462
+    },
+    {
+      "epoch": 2.7,
+      "learning_rate": 1.1318400000000002e-05,
+      "loss": 2.0834,
+      "step": 1463
+    },
+    {
+      "epoch": 2.7,
+      "learning_rate": 1.1315200000000003e-05,
+      "loss": 2.1167,
+      "step": 1464
+    },
+    {
+      "epoch": 2.7,
+      "learning_rate": 1.1312000000000001e-05,
+      "loss": 2.0931,
+      "step": 1465
+    },
+    {
+      "epoch": 2.7,
+      "learning_rate": 1.1308800000000001e-05,
+      "loss": 2.1836,
+      "step": 1466
+    },
+    {
+      "epoch": 2.71,
+      "learning_rate": 1.1305600000000002e-05,
+      "loss": 2.1949,
+      "step": 1467
+    },
+    {
+      "epoch": 2.71,
+      "learning_rate": 1.1302400000000002e-05,
+      "loss": 2.0342,
+      "step": 1468
+    },
+    {
+      "epoch": 2.71,
+      "learning_rate": 1.1299200000000002e-05,
+      "loss": 2.0389,
+      "step": 1469
+    },
+    {
+      "epoch": 2.71,
+      "learning_rate": 1.1296000000000002e-05,
+      "loss": 2.1825,
+      "step": 1470
+    },
+    {
+      "epoch": 2.71,
+      "eval_loss": 2.2837231159210205,
+      "eval_runtime": 231.9729,
+      "eval_samples_per_second": 37.388,
+      "eval_steps_per_second": 0.586,
+      "step": 1470
+    },
+    {
+      "epoch": 2.71,
+      "learning_rate": 1.1292800000000001e-05,
+      "loss": 2.1768,
+      "step": 1471
+    },
+    {
+      "epoch": 2.72,
+      "learning_rate": 1.1289600000000001e-05,
+      "loss": 2.0645,
+      "step": 1472
+    },
+    {
+      "epoch": 2.72,
+      "learning_rate": 1.1286400000000003e-05,
+      "loss": 2.0639,
+      "step": 1473
+    },
+    {
+      "epoch": 2.72,
+      "learning_rate": 1.1283200000000002e-05,
+      "loss": 2.2031,
+      "step": 1474
+    },
+    {
+      "epoch": 2.72,
+      "learning_rate": 1.128e-05,
+      "loss": 2.0512,
+      "step": 1475
+    },
+    {
+      "epoch": 2.72,
+      "learning_rate": 1.1276800000000002e-05,
+      "loss": 2.181,
+      "step": 1476
+    },
+    {
+      "epoch": 2.72,
+      "learning_rate": 1.1273600000000002e-05,
+      "loss": 2.0967,
+      "step": 1477
+    },
+    {
+      "epoch": 2.73,
+      "learning_rate": 1.1270400000000003e-05,
+      "loss": 2.04,
+      "step": 1478
+    },
+    {
+      "epoch": 2.73,
+      "learning_rate": 1.1267200000000003e-05,
+      "loss": 2.1247,
+      "step": 1479
+    },
+    {
+      "epoch": 2.73,
+      "learning_rate": 1.1264000000000001e-05,
+      "loss": 2.1085,
+      "step": 1480
+    },
+    {
+      "epoch": 2.73,
+      "learning_rate": 1.1260800000000002e-05,
+      "loss": 2.2127,
+      "step": 1481
+    },
+    {
+      "epoch": 2.73,
+      "learning_rate": 1.1257600000000002e-05,
+      "loss": 2.1068,
+      "step": 1482
+    },
+    {
+      "epoch": 2.74,
+      "learning_rate": 1.1254400000000002e-05,
+      "loss": 2.0824,
+      "step": 1483
+    },
+    {
+      "epoch": 2.74,
+      "learning_rate": 1.1251200000000003e-05,
+      "loss": 2.1354,
+      "step": 1484
+    },
+    {
+      "epoch": 2.74,
+      "learning_rate": 1.1248000000000001e-05,
+      "loss": 2.1369,
+      "step": 1485
+    },
+    {
+      "epoch": 2.74,
+      "learning_rate": 1.1244800000000001e-05,
+      "loss": 2.096,
+      "step": 1486
+    },
+    {
+      "epoch": 2.74,
+      "learning_rate": 1.1241600000000003e-05,
+      "loss": 2.0418,
+      "step": 1487
+    },
+    {
+      "epoch": 2.74,
+      "learning_rate": 1.1238400000000002e-05,
+      "loss": 2.1553,
+      "step": 1488
+    },
+    {
+      "epoch": 2.75,
+      "learning_rate": 1.1235200000000002e-05,
+      "loss": 2.1026,
+      "step": 1489
+    },
+    {
+      "epoch": 2.75,
+      "learning_rate": 1.1232000000000002e-05,
+      "loss": 2.0779,
+      "step": 1490
+    },
+    {
+      "epoch": 2.75,
+      "learning_rate": 1.1228800000000002e-05,
+      "loss": 2.1842,
+      "step": 1491
+    },
+    {
+      "epoch": 2.75,
+      "learning_rate": 1.1225600000000001e-05,
+      "loss": 2.1465,
+      "step": 1492
+    },
+    {
+      "epoch": 2.75,
+      "learning_rate": 1.1222400000000003e-05,
+      "loss": 2.1529,
+      "step": 1493
+    },
+    {
+      "epoch": 2.76,
+      "learning_rate": 1.1219200000000003e-05,
+      "loss": 2.1225,
+      "step": 1494
+    },
+    {
+      "epoch": 2.76,
+      "learning_rate": 1.1216e-05,
+      "loss": 2.1273,
+      "step": 1495
+    },
+    {
+      "epoch": 2.76,
+      "learning_rate": 1.1212800000000002e-05,
+      "loss": 2.2448,
+      "step": 1496
+    },
+    {
+      "epoch": 2.76,
+      "learning_rate": 1.1209600000000002e-05,
+      "loss": 2.2103,
+      "step": 1497
+    },
+    {
+      "epoch": 2.76,
+      "learning_rate": 1.1206400000000002e-05,
+      "loss": 2.1208,
+      "step": 1498
+    },
+    {
+      "epoch": 2.76,
+      "learning_rate": 1.1203200000000003e-05,
+      "loss": 2.0887,
+      "step": 1499
+    },
+    {
+      "epoch": 2.77,
+      "learning_rate": 1.1200000000000001e-05,
+      "loss": 2.033,
+      "step": 1500
+    },
+    {
+      "epoch": 2.77,
+      "eval_loss": 2.2855587005615234,
+      "eval_runtime": 230.7835,
+      "eval_samples_per_second": 37.581,
+      "eval_steps_per_second": 0.589,
+      "step": 1500
+    },
+    {
+      "epoch": 2.77,
+      "learning_rate": 1.1196800000000001e-05,
+      "loss": 2.0357,
+      "step": 1501
+    },
+    {
+      "epoch": 2.77,
+      "learning_rate": 1.1193600000000002e-05,
+      "loss": 2.1105,
+      "step": 1502
+    },
+    {
+      "epoch": 2.77,
+      "learning_rate": 1.1190400000000002e-05,
+      "loss": 2.0588,
+      "step": 1503
+    },
+    {
+      "epoch": 2.77,
+      "learning_rate": 1.1187200000000002e-05,
+      "loss": 2.1883,
+      "step": 1504
+    },
+    {
+      "epoch": 2.78,
+      "learning_rate": 1.1184000000000002e-05,
+      "loss": 2.0774,
+      "step": 1505
+    },
+    {
+      "epoch": 2.78,
+      "learning_rate": 1.1180800000000001e-05,
+      "loss": 1.9928,
+      "step": 1506
+    },
+    {
+      "epoch": 2.78,
+      "learning_rate": 1.1177600000000001e-05,
+      "loss": 2.1669,
+      "step": 1507
+    },
+    {
+      "epoch": 2.78,
+      "learning_rate": 1.1174400000000003e-05,
+      "loss": 2.1462,
+      "step": 1508
+    },
+    {
+      "epoch": 2.78,
+      "learning_rate": 1.1171200000000002e-05,
+      "loss": 2.1447,
+      "step": 1509
+    },
+    {
+      "epoch": 2.79,
+      "learning_rate": 1.1168e-05,
+      "loss": 2.0867,
+      "step": 1510
+    },
+    {
+      "epoch": 2.79,
+      "learning_rate": 1.1164800000000002e-05,
+      "loss": 2.043,
+      "step": 1511
+    },
+    {
+      "epoch": 2.79,
+      "learning_rate": 1.1161600000000002e-05,
+      "loss": 2.0973,
+      "step": 1512
+    },
+    {
+      "epoch": 2.79,
+      "learning_rate": 1.1158400000000001e-05,
+      "loss": 2.1355,
+      "step": 1513
+    },
+    {
+      "epoch": 2.79,
+      "learning_rate": 1.1155200000000003e-05,
+      "loss": 2.1634,
+      "step": 1514
+    },
+    {
+      "epoch": 2.79,
+      "learning_rate": 1.1152000000000001e-05,
+      "loss": 2.1084,
+      "step": 1515
+    },
+    {
+      "epoch": 2.8,
+      "learning_rate": 1.11488e-05,
+      "loss": 2.1147,
+      "step": 1516
+    },
+    {
+      "epoch": 2.8,
+      "learning_rate": 1.1145600000000002e-05,
+      "loss": 2.1876,
+      "step": 1517
+    },
+    {
+      "epoch": 2.8,
+      "learning_rate": 1.1142400000000002e-05,
+      "loss": 2.1241,
+      "step": 1518
+    },
+    {
+      "epoch": 2.8,
+      "learning_rate": 1.1139200000000002e-05,
+      "loss": 2.1135,
+      "step": 1519
+    },
+    {
+      "epoch": 2.8,
+      "learning_rate": 1.1136000000000001e-05,
+      "loss": 2.107,
+      "step": 1520
+    },
+    {
+      "epoch": 2.81,
+      "learning_rate": 1.1132800000000001e-05,
+      "loss": 2.0376,
+      "step": 1521
+    },
+    {
+      "epoch": 2.81,
+      "learning_rate": 1.1129600000000001e-05,
+      "loss": 2.1402,
+      "step": 1522
+    },
+    {
+      "epoch": 2.81,
+      "learning_rate": 1.1126400000000002e-05,
+      "loss": 2.0915,
+      "step": 1523
+    },
+    {
+      "epoch": 2.81,
+      "learning_rate": 1.1123200000000002e-05,
+      "loss": 2.0555,
+      "step": 1524
+    },
+    {
+      "epoch": 2.81,
+      "learning_rate": 1.112e-05,
+      "loss": 2.1219,
+      "step": 1525
+    },
+    {
+      "epoch": 2.81,
+      "learning_rate": 1.1116800000000002e-05,
+      "loss": 2.0563,
+      "step": 1526
+    },
+    {
+      "epoch": 2.82,
+      "learning_rate": 1.1113600000000001e-05,
+      "loss": 2.1427,
+      "step": 1527
+    },
+    {
+      "epoch": 2.82,
+      "learning_rate": 1.1110400000000003e-05,
+      "loss": 2.1689,
+      "step": 1528
+    },
+    {
+      "epoch": 2.82,
+      "learning_rate": 1.1107200000000003e-05,
+      "loss": 2.1329,
+      "step": 1529
+    },
+    {
+      "epoch": 2.82,
+      "learning_rate": 1.1104e-05,
+      "loss": 2.085,
+      "step": 1530
+    },
+    {
+      "epoch": 2.82,
+      "eval_loss": 2.2864410877227783,
+      "eval_runtime": 231.5794,
+      "eval_samples_per_second": 37.452,
+      "eval_steps_per_second": 0.587,
+      "step": 1530
+    },
+    {
+      "epoch": 2.82,
+      "learning_rate": 1.1100800000000002e-05,
+      "loss": 2.032,
+      "step": 1531
+    },
+    {
+      "epoch": 2.83,
+      "learning_rate": 1.1097600000000002e-05,
+      "loss": 2.0813,
+      "step": 1532
+    },
+    {
+      "epoch": 2.83,
+      "learning_rate": 1.1094400000000002e-05,
+      "loss": 2.0917,
+      "step": 1533
+    },
+    {
+      "epoch": 2.83,
+      "learning_rate": 1.1091200000000003e-05,
+      "loss": 2.1639,
+      "step": 1534
+    },
+    {
+      "epoch": 2.83,
+      "learning_rate": 1.1088000000000001e-05,
+      "loss": 2.186,
+      "step": 1535
+    },
+    {
+      "epoch": 2.83,
+      "learning_rate": 1.1084800000000001e-05,
+      "loss": 2.1059,
+      "step": 1536
+    },
+    {
+      "epoch": 2.84,
+      "learning_rate": 1.1081600000000002e-05,
+      "loss": 2.1105,
+      "step": 1537
+    },
+    {
+      "epoch": 2.84,
+      "learning_rate": 1.1078400000000002e-05,
+      "loss": 1.971,
+      "step": 1538
+    },
+    {
+      "epoch": 2.84,
+      "learning_rate": 1.1075200000000002e-05,
+      "loss": 2.0931,
+      "step": 1539
+    },
+    {
+      "epoch": 2.84,
+      "learning_rate": 1.1072000000000002e-05,
+      "loss": 2.139,
+      "step": 1540
+    },
+    {
+      "epoch": 2.84,
+      "learning_rate": 1.1068800000000001e-05,
+      "loss": 2.1476,
+      "step": 1541
+    },
+    {
+      "epoch": 2.84,
+      "learning_rate": 1.1065600000000001e-05,
+      "loss": 2.1717,
+      "step": 1542
+    },
+    {
+      "epoch": 2.85,
+      "learning_rate": 1.1062400000000003e-05,
+      "loss": 2.0623,
+      "step": 1543
+    },
+    {
+      "epoch": 2.85,
+      "learning_rate": 1.1059200000000002e-05,
+      "loss": 2.0833,
+      "step": 1544
+    },
+    {
+      "epoch": 2.85,
+      "learning_rate": 1.1056e-05,
+      "loss": 2.1215,
+      "step": 1545
+    },
+    {
+      "epoch": 2.85,
+      "learning_rate": 1.1052800000000002e-05,
+      "loss": 2.0889,
+      "step": 1546
+    },
+    {
+      "epoch": 2.85,
+      "learning_rate": 1.1049600000000002e-05,
+      "loss": 2.1099,
+      "step": 1547
+    },
+    {
+      "epoch": 2.86,
+      "learning_rate": 1.1046400000000002e-05,
+      "loss": 2.0909,
+      "step": 1548
+    },
+    {
+      "epoch": 2.86,
+      "learning_rate": 1.1043200000000003e-05,
+      "loss": 2.148,
+      "step": 1549
+    },
+    {
+      "epoch": 2.86,
+      "learning_rate": 1.1040000000000001e-05,
+      "loss": 2.1312,
+      "step": 1550
+    },
+    {
+      "epoch": 2.86,
+      "learning_rate": 1.10368e-05,
+      "loss": 2.0721,
+      "step": 1551
+    },
+    {
+      "epoch": 2.86,
+      "learning_rate": 1.1033600000000002e-05,
+      "loss": 2.1466,
+      "step": 1552
+    },
+    {
+      "epoch": 2.86,
+      "learning_rate": 1.1030400000000002e-05,
+      "loss": 1.9413,
+      "step": 1553
+    },
+    {
+      "epoch": 2.87,
+      "learning_rate": 1.1027200000000002e-05,
+      "loss": 2.117,
+      "step": 1554
+    },
+    {
+      "epoch": 2.87,
+      "learning_rate": 1.1024000000000002e-05,
+      "loss": 2.1132,
+      "step": 1555
+    },
+    {
+      "epoch": 2.87,
+      "learning_rate": 1.1020800000000001e-05,
+      "loss": 2.0938,
+      "step": 1556
+    },
+    {
+      "epoch": 2.87,
+      "learning_rate": 1.1017600000000001e-05,
+      "loss": 2.0845,
+      "step": 1557
+    },
+    {
+      "epoch": 2.87,
+      "learning_rate": 1.1014400000000002e-05,
+      "loss": 2.1581,
+      "step": 1558
+    },
+    {
+      "epoch": 2.88,
+      "learning_rate": 1.1011200000000002e-05,
+      "loss": 2.0453,
+      "step": 1559
+    },
+    {
+      "epoch": 2.88,
+      "learning_rate": 1.1008e-05,
+      "loss": 2.1198,
+      "step": 1560
+    },
+    {
+      "epoch": 2.88,
+      "eval_loss": 2.2824666500091553,
+      "eval_runtime": 231.547,
+      "eval_samples_per_second": 37.457,
+      "eval_steps_per_second": 0.587,
+      "step": 1560
+    },
+    {
+      "epoch": 2.88,
+      "learning_rate": 1.1004800000000002e-05,
+      "loss": 2.1508,
+      "step": 1561
+    },
+    {
+      "epoch": 2.88,
+      "learning_rate": 1.1001600000000002e-05,
+      "loss": 2.1049,
+      "step": 1562
+    },
+    {
+      "epoch": 2.88,
+      "learning_rate": 1.0998400000000001e-05,
+      "loss": 2.0975,
+      "step": 1563
+    },
+    {
+      "epoch": 2.88,
+      "learning_rate": 1.0995200000000003e-05,
+      "loss": 2.1673,
+      "step": 1564
+    },
+    {
+      "epoch": 2.89,
+      "learning_rate": 1.0992000000000003e-05,
+      "loss": 2.1656,
+      "step": 1565
+    },
+    {
+      "epoch": 2.89,
+      "learning_rate": 1.09888e-05,
+      "loss": 2.1481,
+      "step": 1566
+    },
+    {
+      "epoch": 2.89,
+      "learning_rate": 1.0985600000000002e-05,
+      "loss": 2.1297,
+      "step": 1567
+    },
+    {
+      "epoch": 2.89,
+      "learning_rate": 1.0982400000000002e-05,
+      "loss": 2.2329,
+      "step": 1568
+    },
+    {
+      "epoch": 2.89,
+      "learning_rate": 1.0979200000000002e-05,
+      "loss": 2.104,
+      "step": 1569
+    },
+    {
+      "epoch": 2.9,
+      "learning_rate": 1.0976000000000003e-05,
+      "loss": 2.1553,
+      "step": 1570
+    },
+    {
+      "epoch": 2.9,
+      "learning_rate": 1.0972800000000001e-05,
+      "loss": 2.0946,
+      "step": 1571
+    },
+    {
+      "epoch": 2.9,
+      "learning_rate": 1.09696e-05,
+      "loss": 2.1344,
+      "step": 1572
+    },
+    {
+      "epoch": 2.9,
+      "learning_rate": 1.0966400000000002e-05,
+      "loss": 2.063,
+      "step": 1573
+    },
+    {
+      "epoch": 2.9,
+      "learning_rate": 1.0963200000000002e-05,
+      "loss": 2.0724,
+      "step": 1574
+    },
+    {
+      "epoch": 2.91,
+      "learning_rate": 1.0960000000000003e-05,
+      "loss": 2.0107,
+      "step": 1575
+    },
+    {
+      "epoch": 2.91,
+      "learning_rate": 1.0956800000000002e-05,
+      "loss": 2.1189,
+      "step": 1576
+    },
+    {
+      "epoch": 2.91,
+      "learning_rate": 1.0953600000000001e-05,
+      "loss": 2.0696,
+      "step": 1577
+    },
+    {
+      "epoch": 2.91,
+      "learning_rate": 1.0950400000000003e-05,
+      "loss": 2.1333,
+      "step": 1578
+    },
+    {
+      "epoch": 2.91,
+      "learning_rate": 1.0947200000000003e-05,
+      "loss": 2.093,
+      "step": 1579
+    },
+    {
+      "epoch": 2.91,
+      "learning_rate": 1.0944000000000002e-05,
+      "loss": 2.1541,
+      "step": 1580
+    },
+    {
+      "epoch": 2.92,
+      "learning_rate": 1.0940800000000002e-05,
+      "loss": 2.1453,
+      "step": 1581
+    },
+    {
+      "epoch": 2.92,
+      "learning_rate": 1.0937600000000002e-05,
+      "loss": 2.1204,
+      "step": 1582
+    },
+    {
+      "epoch": 2.92,
+      "learning_rate": 1.0934400000000002e-05,
+      "loss": 2.1825,
+      "step": 1583
+    },
+    {
+      "epoch": 2.92,
+      "learning_rate": 1.0931200000000003e-05,
+      "loss": 2.1359,
+      "step": 1584
+    },
+    {
+      "epoch": 2.92,
+      "learning_rate": 1.0928000000000003e-05,
+      "loss": 2.0971,
+      "step": 1585
+    },
+    {
+      "epoch": 2.93,
+      "learning_rate": 1.09248e-05,
+      "loss": 2.0459,
+      "step": 1586
+    },
+    {
+      "epoch": 2.93,
+      "learning_rate": 1.0921600000000002e-05,
+      "loss": 2.1878,
+      "step": 1587
+    },
+    {
+      "epoch": 2.93,
+      "learning_rate": 1.0918400000000002e-05,
+      "loss": 2.0886,
+      "step": 1588
+    },
+    {
+      "epoch": 2.93,
+      "learning_rate": 1.0915200000000002e-05,
+      "loss": 2.142,
+      "step": 1589
+    },
+    {
+      "epoch": 2.93,
+      "learning_rate": 1.0912000000000003e-05,
+      "loss": 2.1462,
+      "step": 1590
+    },
+    {
+      "epoch": 2.93,
+      "eval_loss": 2.2834808826446533,
+      "eval_runtime": 231.1656,
+      "eval_samples_per_second": 37.519,
+      "eval_steps_per_second": 0.588,
+      "step": 1590
+    },
+    {
+      "epoch": 2.93,
+      "learning_rate": 1.0908800000000001e-05,
+      "loss": 2.1382,
+      "step": 1591
+    },
+    {
+      "epoch": 2.94,
+      "learning_rate": 1.0905600000000001e-05,
+      "loss": 2.0577,
+      "step": 1592
+    },
+    {
+      "epoch": 2.94,
+      "learning_rate": 1.0902400000000003e-05,
+      "loss": 2.1351,
+      "step": 1593
+    },
+    {
+      "epoch": 2.94,
+      "learning_rate": 1.0899200000000002e-05,
+      "loss": 2.1189,
+      "step": 1594
+    },
+    {
+      "epoch": 2.94,
+      "learning_rate": 1.0896000000000002e-05,
+      "loss": 2.1136,
+      "step": 1595
+    },
+    {
+      "epoch": 2.94,
+      "learning_rate": 1.0892800000000002e-05,
+      "loss": 2.1517,
+      "step": 1596
+    },
+    {
+      "epoch": 2.95,
+      "learning_rate": 1.0889600000000002e-05,
+      "loss": 2.1221,
+      "step": 1597
+    },
+    {
+      "epoch": 2.95,
+      "learning_rate": 1.0886400000000001e-05,
+      "loss": 2.0772,
+      "step": 1598
+    },
+    {
+      "epoch": 2.95,
+      "learning_rate": 1.0883200000000003e-05,
+      "loss": 2.0664,
+      "step": 1599
+    },
+    {
+      "epoch": 2.95,
+      "learning_rate": 1.0880000000000003e-05,
+      "loss": 2.0879,
+      "step": 1600
+    },
+    {
+      "epoch": 2.95,
+      "learning_rate": 1.08768e-05,
+      "loss": 2.0394,
+      "step": 1601
+    },
+    {
+      "epoch": 2.95,
+      "learning_rate": 1.0873600000000002e-05,
+      "loss": 2.1009,
+      "step": 1602
+    },
+    {
+      "epoch": 2.96,
+      "learning_rate": 1.0870400000000002e-05,
+      "loss": 2.1328,
+      "step": 1603
+    },
+    {
+      "epoch": 2.96,
+      "learning_rate": 1.0867200000000002e-05,
+      "loss": 2.1482,
+      "step": 1604
+    },
+    {
+      "epoch": 2.96,
+      "learning_rate": 1.0864000000000003e-05,
+      "loss": 2.1607,
+      "step": 1605
+    },
+    {
+      "epoch": 2.96,
+      "learning_rate": 1.0860800000000001e-05,
+      "loss": 1.989,
+      "step": 1606
+    },
+    {
+      "epoch": 2.96,
+      "learning_rate": 1.0857600000000001e-05,
+      "loss": 2.1071,
+      "step": 1607
+    },
+    {
+      "epoch": 2.97,
+      "learning_rate": 1.0854400000000002e-05,
+      "loss": 2.1953,
+      "step": 1608
+    },
+    {
+      "epoch": 2.97,
+      "learning_rate": 1.0851200000000002e-05,
+      "loss": 2.148,
+      "step": 1609
+    },
+    {
+      "epoch": 2.97,
+      "learning_rate": 1.0848000000000002e-05,
+      "loss": 2.1504,
+      "step": 1610
+    },
+    {
+      "epoch": 2.97,
+      "learning_rate": 1.0844800000000002e-05,
+      "loss": 2.1542,
+      "step": 1611
+    },
+    {
+      "epoch": 2.97,
+      "learning_rate": 1.0841600000000001e-05,
+      "loss": 1.976,
+      "step": 1612
+    },
+    {
+      "epoch": 2.98,
+      "learning_rate": 1.0838400000000001e-05,
+      "loss": 2.0151,
+      "step": 1613
+    },
+    {
+      "epoch": 2.98,
+      "learning_rate": 1.0835200000000003e-05,
+      "loss": 2.1586,
+      "step": 1614
+    },
+    {
+      "epoch": 2.98,
+      "learning_rate": 1.0832000000000002e-05,
+      "loss": 2.0913,
+      "step": 1615
+    },
+    {
+      "epoch": 2.98,
+      "learning_rate": 1.08288e-05,
+      "loss": 2.0083,
+      "step": 1616
+    },
+    {
+      "epoch": 2.98,
+      "learning_rate": 1.0825600000000002e-05,
+      "loss": 2.1194,
+      "step": 1617
+    },
+    {
+      "epoch": 2.98,
+      "learning_rate": 1.0822400000000002e-05,
+      "loss": 2.1226,
+      "step": 1618
+    },
+    {
+      "epoch": 2.99,
+      "learning_rate": 1.0819200000000001e-05,
+      "loss": 2.1064,
+      "step": 1619
+    },
+    {
+      "epoch": 2.99,
+      "learning_rate": 1.0816000000000003e-05,
+      "loss": 2.1564,
+      "step": 1620
+    },
+    {
+      "epoch": 2.99,
+      "eval_loss": 2.280937433242798,
+      "eval_runtime": 232.6554,
+      "eval_samples_per_second": 37.278,
+      "eval_steps_per_second": 0.585,
+      "step": 1620
+    },
+    {
+      "epoch": 2.99,
+      "learning_rate": 1.0812800000000001e-05,
+      "loss": 2.0937,
+      "step": 1621
+    },
+    {
+      "epoch": 2.99,
+      "learning_rate": 1.08096e-05,
+      "loss": 2.0156,
+      "step": 1622
+    },
+    {
+      "epoch": 2.99,
+      "learning_rate": 1.0806400000000002e-05,
+      "loss": 2.079,
+      "step": 1623
+    },
+    {
+      "epoch": 3.0,
+      "learning_rate": 1.0803200000000002e-05,
+      "loss": 2.1833,
+      "step": 1624
+    },
+    {
+      "epoch": 3.0,
+      "learning_rate": 1.0800000000000003e-05,
+      "loss": 2.2006,
+      "step": 1625
+    },
+    {
+      "epoch": 3.0,
+      "learning_rate": 1.0796800000000001e-05,
+      "loss": 2.1441,
+      "step": 1626
+    },
+    {
+      "epoch": 3.0,
+      "learning_rate": 1.0793600000000001e-05,
+      "loss": 3.06,
+      "step": 1627
+    },
+    {
+      "epoch": 3.0,
+      "learning_rate": 1.0790400000000003e-05,
+      "loss": 1.9771,
+      "step": 1628
+    },
+    {
+      "epoch": 3.01,
+      "learning_rate": 1.0787200000000002e-05,
+      "loss": 1.9959,
+      "step": 1629
+    },
+    {
+      "epoch": 3.01,
+      "learning_rate": 1.0784000000000002e-05,
+      "loss": 2.0505,
+      "step": 1630
+    },
+    {
+      "epoch": 3.01,
+      "learning_rate": 1.0780800000000002e-05,
+      "loss": 1.9922,
+      "step": 1631
+    },
+    {
+      "epoch": 3.01,
+      "learning_rate": 1.0777600000000002e-05,
+      "loss": 2.0331,
+      "step": 1632
+    },
+    {
+      "epoch": 3.01,
+      "learning_rate": 1.0774400000000001e-05,
+      "loss": 1.9856,
+      "step": 1633
+    },
+    {
+      "epoch": 3.01,
+      "learning_rate": 1.0771200000000003e-05,
+      "loss": 1.9203,
+      "step": 1634
+    },
+    {
+      "epoch": 3.02,
+      "learning_rate": 1.0768000000000003e-05,
+      "loss": 1.9155,
+      "step": 1635
+    },
+    {
+      "epoch": 3.02,
+      "learning_rate": 1.07648e-05,
+      "loss": 2.0055,
+      "step": 1636
+    },
+    {
+      "epoch": 3.02,
+      "learning_rate": 1.0761600000000002e-05,
+      "loss": 2.0483,
+      "step": 1637
+    },
+    {
+      "epoch": 3.02,
+      "learning_rate": 1.0758400000000002e-05,
+      "loss": 1.9202,
+      "step": 1638
+    },
+    {
+      "epoch": 3.02,
+      "learning_rate": 1.0755200000000002e-05,
+      "loss": 1.9218,
+      "step": 1639
+    },
+    {
+      "epoch": 3.03,
+      "learning_rate": 1.0752000000000003e-05,
+      "loss": 2.0421,
+      "step": 1640
+    },
+    {
+      "epoch": 3.03,
+      "learning_rate": 1.0748800000000001e-05,
+      "loss": 2.0984,
+      "step": 1641
+    },
+    {
+      "epoch": 3.03,
+      "learning_rate": 1.0745600000000001e-05,
+      "loss": 1.9973,
+      "step": 1642
+    },
+    {
+      "epoch": 3.03,
+      "learning_rate": 1.0742400000000002e-05,
+      "loss": 2.0266,
+      "step": 1643
+    },
+    {
+      "epoch": 3.03,
+      "learning_rate": 1.0739200000000002e-05,
+      "loss": 2.0385,
+      "step": 1644
+    },
+    {
+      "epoch": 3.04,
+      "learning_rate": 1.0736000000000002e-05,
+      "loss": 1.9821,
+      "step": 1645
+    },
+    {
+      "epoch": 3.04,
+      "learning_rate": 1.0732800000000002e-05,
+      "loss": 1.948,
+      "step": 1646
+    },
+    {
+      "epoch": 3.04,
+      "learning_rate": 1.0729600000000001e-05,
+      "loss": 2.0996,
+      "step": 1647
+    },
+    {
+      "epoch": 3.04,
+      "learning_rate": 1.0726400000000001e-05,
+      "loss": 2.1138,
+      "step": 1648
+    },
+    {
+      "epoch": 3.04,
+      "learning_rate": 1.0723200000000003e-05,
+      "loss": 2.0189,
+      "step": 1649
+    },
+    {
+      "epoch": 3.04,
+      "learning_rate": 1.0720000000000002e-05,
+      "loss": 2.0044,
+      "step": 1650
+    },
+    {
+      "epoch": 3.04,
+      "eval_loss": 2.301302671432495,
+      "eval_runtime": 233.8179,
+      "eval_samples_per_second": 37.093,
+      "eval_steps_per_second": 0.582,
+      "step": 1650
+    },
+    {
+      "epoch": 3.05,
+      "learning_rate": 1.07168e-05,
+      "loss": 2.0292,
+      "step": 1651
+    },
+    {
+      "epoch": 3.05,
+      "learning_rate": 1.0713600000000002e-05,
+      "loss": 2.0181,
+      "step": 1652
+    },
+    {
+      "epoch": 3.05,
+      "learning_rate": 1.0710400000000002e-05,
+      "loss": 2.0026,
+      "step": 1653
+    },
+    {
+      "epoch": 3.05,
+      "learning_rate": 1.0707200000000001e-05,
+      "loss": 1.8948,
+      "step": 1654
+    },
+    {
+      "epoch": 3.05,
+      "learning_rate": 1.0704000000000003e-05,
+      "loss": 2.0412,
+      "step": 1655
+    },
+    {
+      "epoch": 3.06,
+      "learning_rate": 1.0700800000000001e-05,
+      "loss": 1.9936,
+      "step": 1656
+    },
+    {
+      "epoch": 3.06,
+      "learning_rate": 1.06976e-05,
+      "loss": 2.1684,
+      "step": 1657
+    },
+    {
+      "epoch": 3.06,
+      "learning_rate": 1.0694400000000002e-05,
+      "loss": 2.0586,
+      "step": 1658
+    },
+    {
+      "epoch": 3.06,
+      "learning_rate": 1.0691200000000002e-05,
+      "loss": 1.9758,
+      "step": 1659
+    },
+    {
+      "epoch": 3.06,
+      "learning_rate": 1.0688000000000002e-05,
+      "loss": 1.974,
+      "step": 1660
+    },
+    {
+      "epoch": 3.06,
+      "learning_rate": 1.0684800000000001e-05,
+      "loss": 2.09,
+      "step": 1661
+    },
+    {
+      "epoch": 3.07,
+      "learning_rate": 1.0681600000000001e-05,
+      "loss": 2.0843,
+      "step": 1662
+    },
+    {
+      "epoch": 3.07,
+      "learning_rate": 1.0678400000000001e-05,
+      "loss": 1.8603,
+      "step": 1663
+    },
+    {
+      "epoch": 3.07,
+      "learning_rate": 1.0675200000000002e-05,
+      "loss": 1.9606,
+      "step": 1664
+    },
+    {
+      "epoch": 3.07,
+      "learning_rate": 1.0672000000000002e-05,
+      "loss": 1.9344,
+      "step": 1665
+    },
+    {
+      "epoch": 3.07,
+      "learning_rate": 1.06688e-05,
+      "loss": 1.8977,
+      "step": 1666
+    },
+    {
+      "epoch": 3.08,
+      "learning_rate": 1.0665600000000002e-05,
+      "loss": 2.0543,
+      "step": 1667
+    },
+    {
+      "epoch": 3.08,
+      "learning_rate": 1.0662400000000001e-05,
+      "loss": 1.9213,
+      "step": 1668
+    },
+    {
+      "epoch": 3.08,
+      "learning_rate": 1.0659200000000001e-05,
+      "loss": 1.9797,
+      "step": 1669
+    },
+    {
+      "epoch": 3.08,
+      "learning_rate": 1.0656000000000003e-05,
+      "loss": 2.0047,
+      "step": 1670
+    },
+    {
+      "epoch": 3.08,
+      "learning_rate": 1.06528e-05,
+      "loss": 2.053,
+      "step": 1671
+    },
+    {
+      "epoch": 3.08,
+      "learning_rate": 1.06496e-05,
+      "loss": 1.9343,
+      "step": 1672
+    },
+    {
+      "epoch": 3.09,
+      "learning_rate": 1.0646400000000002e-05,
+      "loss": 2.0762,
+      "step": 1673
+    },
+    {
+      "epoch": 3.09,
+      "learning_rate": 1.0643200000000002e-05,
+      "loss": 1.8935,
+      "step": 1674
+    },
+    {
+      "epoch": 3.09,
+      "learning_rate": 1.0640000000000003e-05,
+      "loss": 1.9431,
+      "step": 1675
+    },
+    {
+      "epoch": 3.09,
+      "learning_rate": 1.0636800000000001e-05,
+      "loss": 2.0579,
+      "step": 1676
+    },
+    {
+      "epoch": 3.09,
+      "learning_rate": 1.0633600000000001e-05,
+      "loss": 2.0625,
+      "step": 1677
+    },
+    {
+      "epoch": 3.1,
+      "learning_rate": 1.0630400000000002e-05,
+      "loss": 2.0415,
+      "step": 1678
+    },
+    {
+      "epoch": 3.1,
+      "learning_rate": 1.0627200000000002e-05,
+      "loss": 1.9964,
+      "step": 1679
+    },
+    {
+      "epoch": 3.1,
+      "learning_rate": 1.0624000000000002e-05,
+      "loss": 2.0382,
+      "step": 1680
+    },
+    {
+      "epoch": 3.1,
+      "eval_loss": 2.2976086139678955,
+      "eval_runtime": 233.2974,
+      "eval_samples_per_second": 37.176,
+      "eval_steps_per_second": 0.583,
+      "step": 1680
+    },
+    {
+      "epoch": 3.1,
+      "learning_rate": 1.0620800000000002e-05,
+      "loss": 1.9632,
+      "step": 1681
+    },
+    {
+      "epoch": 3.1,
+      "learning_rate": 1.0617600000000001e-05,
+      "loss": 2.0899,
+      "step": 1682
+    },
+    {
+      "epoch": 3.11,
+      "learning_rate": 1.0614400000000001e-05,
+      "loss": 1.9786,
+      "step": 1683
+    },
+    {
+      "epoch": 3.11,
+      "learning_rate": 1.0611200000000003e-05,
+      "loss": 2.0624,
+      "step": 1684
+    },
+    {
+      "epoch": 3.11,
+      "learning_rate": 1.0608000000000002e-05,
+      "loss": 2.0252,
+      "step": 1685
+    },
+    {
+      "epoch": 3.11,
+      "learning_rate": 1.06048e-05,
+      "loss": 2.0409,
+      "step": 1686
+    },
+    {
+      "epoch": 3.11,
+      "learning_rate": 1.0601600000000002e-05,
+      "loss": 1.9822,
+      "step": 1687
+    },
+    {
+      "epoch": 3.11,
+      "learning_rate": 1.0598400000000002e-05,
+      "loss": 1.9067,
+      "step": 1688
+    },
+    {
+      "epoch": 3.12,
+      "learning_rate": 1.0595200000000001e-05,
+      "loss": 2.0172,
+      "step": 1689
+    },
+    {
+      "epoch": 3.12,
+      "learning_rate": 1.0592000000000003e-05,
+      "loss": 2.0841,
+      "step": 1690
+    },
+    {
+      "epoch": 3.12,
+      "learning_rate": 1.0588800000000003e-05,
+      "loss": 2.0441,
+      "step": 1691
+    },
+    {
+      "epoch": 3.12,
+      "learning_rate": 1.05856e-05,
+      "loss": 1.8886,
+      "step": 1692
+    },
+    {
+      "epoch": 3.12,
+      "learning_rate": 1.0582400000000002e-05,
+      "loss": 2.0397,
+      "step": 1693
+    },
+    {
+      "epoch": 3.13,
+      "learning_rate": 1.0579200000000002e-05,
+      "loss": 2.0908,
+      "step": 1694
+    },
+    {
+      "epoch": 3.13,
+      "learning_rate": 1.0576000000000002e-05,
+      "loss": 2.0032,
+      "step": 1695
+    },
+    {
+      "epoch": 3.13,
+      "learning_rate": 1.0572800000000003e-05,
+      "loss": 2.008,
+      "step": 1696
+    },
+    {
+      "epoch": 3.13,
+      "learning_rate": 1.0569600000000001e-05,
+      "loss": 2.0664,
+      "step": 1697
+    },
+    {
+      "epoch": 3.13,
+      "learning_rate": 1.0566400000000001e-05,
+      "loss": 2.0795,
+      "step": 1698
+    },
+    {
+      "epoch": 3.13,
+      "learning_rate": 1.0563200000000002e-05,
+      "loss": 2.1053,
+      "step": 1699
+    },
+    {
+      "epoch": 3.14,
+      "learning_rate": 1.0560000000000002e-05,
+      "loss": 1.9984,
+      "step": 1700
+    },
+    {
+      "epoch": 3.14,
+      "learning_rate": 1.0556800000000002e-05,
+      "loss": 2.1,
+      "step": 1701
+    },
+    {
+      "epoch": 3.14,
+      "learning_rate": 1.0553600000000002e-05,
+      "loss": 1.9796,
+      "step": 1702
+    },
+    {
+      "epoch": 3.14,
+      "learning_rate": 1.0550400000000002e-05,
+      "loss": 1.9694,
+      "step": 1703
+    },
+    {
+      "epoch": 3.14,
+      "learning_rate": 1.0547200000000001e-05,
+      "loss": 2.0585,
+      "step": 1704
+    },
+    {
+      "epoch": 3.15,
+      "learning_rate": 1.0544000000000003e-05,
+      "loss": 2.032,
+      "step": 1705
+    },
+    {
+      "epoch": 3.15,
+      "learning_rate": 1.0540800000000002e-05,
+      "loss": 1.9497,
+      "step": 1706
+    },
+    {
+      "epoch": 3.15,
+      "learning_rate": 1.05376e-05,
+      "loss": 1.9583,
+      "step": 1707
+    },
+    {
+      "epoch": 3.15,
+      "learning_rate": 1.0534400000000002e-05,
+      "loss": 2.0815,
+      "step": 1708
+    },
+    {
+      "epoch": 3.15,
+      "learning_rate": 1.0531200000000002e-05,
+      "loss": 2.0166,
+      "step": 1709
+    },
+    {
+      "epoch": 3.15,
+      "learning_rate": 1.0528000000000002e-05,
+      "loss": 2.0007,
+      "step": 1710
+    },
+    {
+      "epoch": 3.15,
+      "eval_loss": 2.3039910793304443,
+      "eval_runtime": 231.8221,
+      "eval_samples_per_second": 37.412,
+      "eval_steps_per_second": 0.587,
+      "step": 1710
+    },
+    {
+      "epoch": 3.16,
+      "learning_rate": 1.0524800000000003e-05,
+      "loss": 2.0365,
+      "step": 1711
+    },
+    {
+      "epoch": 3.16,
+      "learning_rate": 1.0521600000000001e-05,
+      "loss": 1.9864,
+      "step": 1712
+    },
+    {
+      "epoch": 3.16,
+      "learning_rate": 1.05184e-05,
+      "loss": 1.9536,
+      "step": 1713
+    },
+    {
+      "epoch": 3.16,
+      "learning_rate": 1.0515200000000002e-05,
+      "loss": 1.9791,
+      "step": 1714
+    },
+    {
+      "epoch": 3.16,
+      "learning_rate": 1.0512000000000002e-05,
+      "loss": 1.9998,
+      "step": 1715
+    },
+    {
+      "epoch": 3.17,
+      "learning_rate": 1.0508800000000002e-05,
+      "loss": 1.8772,
+      "step": 1716
+    },
+    {
+      "epoch": 3.17,
+      "learning_rate": 1.0505600000000002e-05,
+      "loss": 2.058,
+      "step": 1717
+    },
+    {
+      "epoch": 3.17,
+      "learning_rate": 1.0502400000000001e-05,
+      "loss": 1.9569,
+      "step": 1718
+    },
+    {
+      "epoch": 3.17,
+      "learning_rate": 1.0499200000000001e-05,
+      "loss": 1.9535,
+      "step": 1719
+    },
+    {
+      "epoch": 3.17,
+      "learning_rate": 1.0496000000000003e-05,
+      "loss": 2.0236,
+      "step": 1720
+    },
+    {
+      "epoch": 3.18,
+      "learning_rate": 1.0492800000000002e-05,
+      "loss": 2.0177,
+      "step": 1721
+    },
+    {
+      "epoch": 3.18,
+      "learning_rate": 1.04896e-05,
+      "loss": 2.0502,
+      "step": 1722
+    },
+    {
+      "epoch": 3.18,
+      "learning_rate": 1.0486400000000002e-05,
+      "loss": 2.0548,
+      "step": 1723
+    },
+    {
+      "epoch": 3.18,
+      "learning_rate": 1.0483200000000002e-05,
+      "loss": 2.1707,
+      "step": 1724
+    },
+    {
+      "epoch": 3.18,
+      "learning_rate": 1.0480000000000003e-05,
+      "loss": 2.0731,
+      "step": 1725
+    },
+    {
+      "epoch": 3.18,
+      "learning_rate": 1.0476800000000003e-05,
+      "loss": 2.028,
+      "step": 1726
+    },
+    {
+      "epoch": 3.19,
+      "learning_rate": 1.04736e-05,
+      "loss": 2.0219,
+      "step": 1727
+    },
+    {
+      "epoch": 3.19,
+      "learning_rate": 1.0470400000000002e-05,
+      "loss": 1.9077,
+      "step": 1728
+    },
+    {
+      "epoch": 3.19,
+      "learning_rate": 1.0467200000000002e-05,
+      "loss": 1.9377,
+      "step": 1729
+    },
+    {
+      "epoch": 3.19,
+      "learning_rate": 1.0464000000000002e-05,
+      "loss": 1.9001,
+      "step": 1730
+    },
+    {
+      "epoch": 3.19,
+      "learning_rate": 1.0460800000000003e-05,
+      "loss": 2.0459,
+      "step": 1731
+    },
+    {
+      "epoch": 3.2,
+      "learning_rate": 1.0457600000000001e-05,
+      "loss": 2.0123,
+      "step": 1732
+    },
+    {
+      "epoch": 3.2,
+      "learning_rate": 1.0454400000000001e-05,
+      "loss": 1.9915,
+      "step": 1733
+    },
+    {
+      "epoch": 3.2,
+      "learning_rate": 1.0451200000000003e-05,
+      "loss": 1.9886,
+      "step": 1734
+    },
+    {
+      "epoch": 3.2,
+      "learning_rate": 1.0448000000000002e-05,
+      "loss": 1.9641,
+      "step": 1735
+    },
+    {
+      "epoch": 3.2,
+      "learning_rate": 1.0444800000000002e-05,
+      "loss": 2.0513,
+      "step": 1736
+    },
+    {
+      "epoch": 3.2,
+      "learning_rate": 1.0441600000000002e-05,
+      "loss": 1.9891,
+      "step": 1737
+    },
+    {
+      "epoch": 3.21,
+      "learning_rate": 1.0438400000000002e-05,
+      "loss": 1.9926,
+      "step": 1738
+    },
+    {
+      "epoch": 3.21,
+      "learning_rate": 1.0435200000000001e-05,
+      "loss": 2.0742,
+      "step": 1739
+    },
+    {
+      "epoch": 3.21,
+      "learning_rate": 1.0432000000000003e-05,
+      "loss": 1.9511,
+      "step": 1740
+    },
+    {
+      "epoch": 3.21,
+      "eval_loss": 2.3009254932403564,
+      "eval_runtime": 232.1916,
+      "eval_samples_per_second": 37.353,
+      "eval_steps_per_second": 0.586,
+      "step": 1740
+    },
+    {
+      "epoch": 3.21,
+      "learning_rate": 1.0428800000000003e-05,
+      "loss": 1.9928,
+      "step": 1741
+    },
+    {
+      "epoch": 3.21,
+      "learning_rate": 1.04256e-05,
+      "loss": 1.9674,
+      "step": 1742
+    },
+    {
+      "epoch": 3.22,
+      "learning_rate": 1.0422400000000002e-05,
+      "loss": 1.9836,
+      "step": 1743
+    },
+    {
+      "epoch": 3.22,
+      "learning_rate": 1.0419200000000002e-05,
+      "loss": 1.9797,
+      "step": 1744
+    },
+    {
+      "epoch": 3.22,
+      "learning_rate": 1.0416000000000002e-05,
+      "loss": 1.9437,
+      "step": 1745
+    },
+    {
+      "epoch": 3.22,
+      "learning_rate": 1.0412800000000003e-05,
+      "loss": 1.9615,
+      "step": 1746
+    },
+    {
+      "epoch": 3.22,
+      "learning_rate": 1.0409600000000001e-05,
+      "loss": 1.9801,
+      "step": 1747
+    },
+    {
+      "epoch": 3.22,
+      "learning_rate": 1.0406400000000001e-05,
+      "loss": 2.0428,
+      "step": 1748
+    },
+    {
+      "epoch": 3.23,
+      "learning_rate": 1.0403200000000002e-05,
+      "loss": 1.9746,
+      "step": 1749
+    },
+    {
+      "epoch": 3.23,
+      "learning_rate": 1.0400000000000002e-05,
+      "loss": 2.0161,
+      "step": 1750
+    },
+    {
+      "epoch": 3.23,
+      "learning_rate": 1.0396800000000002e-05,
+      "loss": 2.0454,
+      "step": 1751
+    },
+    {
+      "epoch": 3.23,
+      "learning_rate": 1.0393600000000002e-05,
+      "loss": 2.0519,
+      "step": 1752
+    },
+    {
+      "epoch": 3.23,
+      "learning_rate": 1.0390400000000001e-05,
+      "loss": 2.1508,
+      "step": 1753
+    },
+    {
+      "epoch": 3.24,
+      "learning_rate": 1.0387200000000001e-05,
+      "loss": 2.0628,
+      "step": 1754
+    },
+    {
+      "epoch": 3.24,
+      "learning_rate": 1.0384000000000003e-05,
+      "loss": 2.0049,
+      "step": 1755
+    },
+    {
+      "epoch": 3.24,
+      "learning_rate": 1.0380800000000002e-05,
+      "loss": 2.0246,
+      "step": 1756
+    },
+    {
+      "epoch": 3.24,
+      "learning_rate": 1.03776e-05,
+      "loss": 1.9649,
+      "step": 1757
+    },
+    {
+      "epoch": 3.24,
+      "learning_rate": 1.0374400000000002e-05,
+      "loss": 2.0322,
+      "step": 1758
+    },
+    {
+      "epoch": 3.25,
+      "learning_rate": 1.0371200000000002e-05,
+      "loss": 1.9875,
+      "step": 1759
+    },
+    {
+      "epoch": 3.25,
+      "learning_rate": 1.0368000000000001e-05,
+      "loss": 2.0085,
+      "step": 1760
+    },
+    {
+      "epoch": 3.25,
+      "learning_rate": 1.0364800000000003e-05,
+      "loss": 2.0715,
+      "step": 1761
+    },
+    {
+      "epoch": 3.25,
+      "learning_rate": 1.0361600000000001e-05,
+      "loss": 1.9981,
+      "step": 1762
+    },
+    {
+      "epoch": 3.25,
+      "learning_rate": 1.03584e-05,
+      "loss": 1.9662,
+      "step": 1763
+    },
+    {
+      "epoch": 3.25,
+      "learning_rate": 1.0355200000000002e-05,
+      "loss": 2.1093,
+      "step": 1764
+    },
+    {
+      "epoch": 3.26,
+      "learning_rate": 1.0352000000000002e-05,
+      "loss": 2.0128,
+      "step": 1765
+    },
+    {
+      "epoch": 3.26,
+      "learning_rate": 1.0348800000000002e-05,
+      "loss": 2.0179,
+      "step": 1766
+    },
+    {
+      "epoch": 3.26,
+      "learning_rate": 1.0345600000000001e-05,
+      "loss": 2.1388,
+      "step": 1767
+    },
+    {
+      "epoch": 3.26,
+      "learning_rate": 1.0342400000000001e-05,
+      "loss": 2.0409,
+      "step": 1768
+    },
+    {
+      "epoch": 3.26,
+      "learning_rate": 1.0339200000000001e-05,
+      "loss": 1.9415,
+      "step": 1769
+    },
+    {
+      "epoch": 3.27,
+      "learning_rate": 1.0336000000000002e-05,
+      "loss": 1.9658,
+      "step": 1770
+    },
+    {
+      "epoch": 3.27,
+      "eval_loss": 2.3059329986572266,
+      "eval_runtime": 232.7983,
+      "eval_samples_per_second": 37.255,
+      "eval_steps_per_second": 0.584,
+      "step": 1770
+    },
+    {
+      "epoch": 3.27,
+      "learning_rate": 1.0332800000000002e-05,
+      "loss": 2.0383,
+      "step": 1771
+    },
+    {
+      "epoch": 3.27,
+      "learning_rate": 1.03296e-05,
+      "loss": 2.0552,
+      "step": 1772
+    },
+    {
+      "epoch": 3.27,
+      "learning_rate": 1.0326400000000002e-05,
+      "loss": 2.0319,
+      "step": 1773
+    },
+    {
+      "epoch": 3.27,
+      "learning_rate": 1.0323200000000001e-05,
+      "loss": 2.0109,
+      "step": 1774
+    },
+    {
+      "epoch": 3.27,
+      "learning_rate": 1.0320000000000003e-05,
+      "loss": 1.8535,
+      "step": 1775
+    },
+    {
+      "epoch": 3.28,
+      "learning_rate": 1.0316800000000003e-05,
+      "loss": 1.9168,
+      "step": 1776
+    },
+    {
+      "epoch": 3.28,
+      "learning_rate": 1.03136e-05,
+      "loss": 1.9355,
+      "step": 1777
+    },
+    {
+      "epoch": 3.28,
+      "learning_rate": 1.0310400000000002e-05,
+      "loss": 1.9308,
+      "step": 1778
+    },
+    {
+      "epoch": 3.28,
+      "learning_rate": 1.0307200000000002e-05,
+      "loss": 2.015,
+      "step": 1779
+    },
+    {
+      "epoch": 3.28,
+      "learning_rate": 1.0304000000000002e-05,
+      "loss": 1.9882,
+      "step": 1780
+    },
+    {
+      "epoch": 3.29,
+      "learning_rate": 1.0300800000000003e-05,
+      "loss": 1.998,
+      "step": 1781
+    },
+    {
+      "epoch": 3.29,
+      "learning_rate": 1.0297600000000001e-05,
+      "loss": 2.0624,
+      "step": 1782
+    },
+    {
+      "epoch": 3.29,
+      "learning_rate": 1.0294400000000001e-05,
+      "loss": 1.9529,
+      "step": 1783
+    },
+    {
+      "epoch": 3.29,
+      "learning_rate": 1.0291200000000002e-05,
+      "loss": 2.0453,
+      "step": 1784
+    },
+    {
+      "epoch": 3.29,
+      "learning_rate": 1.0288000000000002e-05,
+      "loss": 2.0749,
+      "step": 1785
+    },
+    {
+      "epoch": 3.29,
+      "learning_rate": 1.0284800000000002e-05,
+      "loss": 1.9746,
+      "step": 1786
+    },
+    {
+      "epoch": 3.3,
+      "learning_rate": 1.0281600000000002e-05,
+      "loss": 2.0096,
+      "step": 1787
+    },
+    {
+      "epoch": 3.3,
+      "learning_rate": 1.0278400000000001e-05,
+      "loss": 2.0433,
+      "step": 1788
+    },
+    {
+      "epoch": 3.3,
+      "learning_rate": 1.0275200000000001e-05,
+      "loss": 1.9828,
+      "step": 1789
+    },
+    {
+      "epoch": 3.3,
+      "learning_rate": 1.0272000000000003e-05,
+      "loss": 2.0341,
+      "step": 1790
+    },
+    {
+      "epoch": 3.3,
+      "learning_rate": 1.0268800000000002e-05,
+      "loss": 2.0064,
+      "step": 1791
+    },
+    {
+      "epoch": 3.31,
+      "learning_rate": 1.02656e-05,
+      "loss": 2.0545,
+      "step": 1792
+    },
+    {
+      "epoch": 3.31,
+      "learning_rate": 1.0262400000000002e-05,
+      "loss": 2.057,
+      "step": 1793
+    },
+    {
+      "epoch": 3.31,
+      "learning_rate": 1.0259200000000002e-05,
+      "loss": 2.0018,
+      "step": 1794
+    },
+    {
+      "epoch": 3.31,
+      "learning_rate": 1.0256000000000001e-05,
+      "loss": 1.9688,
+      "step": 1795
+    },
+    {
+      "epoch": 3.31,
+      "learning_rate": 1.0252800000000003e-05,
+      "loss": 2.1464,
+      "step": 1796
+    },
+    {
+      "epoch": 3.32,
+      "learning_rate": 1.0249600000000001e-05,
+      "loss": 1.9954,
+      "step": 1797
+    },
+    {
+      "epoch": 3.32,
+      "learning_rate": 1.02464e-05,
+      "loss": 1.8981,
+      "step": 1798
+    },
+    {
+      "epoch": 3.32,
+      "learning_rate": 1.0243200000000002e-05,
+      "loss": 2.125,
+      "step": 1799
+    },
+    {
+      "epoch": 3.32,
+      "learning_rate": 1.0240000000000002e-05,
+      "loss": 1.9447,
+      "step": 1800
+    },
+    {
+      "epoch": 3.32,
+      "eval_loss": 2.3070602416992188,
+      "eval_runtime": 233.1327,
+      "eval_samples_per_second": 37.202,
+      "eval_steps_per_second": 0.583,
+      "step": 1800
+    },
+    {
+      "epoch": 3.32,
+      "learning_rate": 1.0236800000000002e-05,
+      "loss": 2.037,
+      "step": 1801
+    },
+    {
+      "epoch": 3.32,
+      "learning_rate": 1.0233600000000001e-05,
+      "loss": 2.097,
+      "step": 1802
+    },
+    {
+      "epoch": 3.33,
+      "learning_rate": 1.0230400000000001e-05,
+      "loss": 2.111,
+      "step": 1803
+    },
+    {
+      "epoch": 3.33,
+      "learning_rate": 1.0227200000000001e-05,
+      "loss": 2.0059,
+      "step": 1804
+    },
+    {
+      "epoch": 3.33,
+      "learning_rate": 1.0224000000000002e-05,
+      "loss": 1.9592,
+      "step": 1805
+    },
+    {
+      "epoch": 3.33,
+      "learning_rate": 1.0220800000000002e-05,
+      "loss": 2.1159,
+      "step": 1806
+    },
+    {
+      "epoch": 3.33,
+      "learning_rate": 1.02176e-05,
+      "loss": 1.993,
+      "step": 1807
+    },
+    {
+      "epoch": 3.34,
+      "learning_rate": 1.0214400000000002e-05,
+      "loss": 1.896,
+      "step": 1808
+    },
+    {
+      "epoch": 3.34,
+      "learning_rate": 1.0211200000000001e-05,
+      "loss": 2.0728,
+      "step": 1809
+    },
+    {
+      "epoch": 3.34,
+      "learning_rate": 1.0208000000000001e-05,
+      "loss": 2.006,
+      "step": 1810
+    },
+    {
+      "epoch": 3.34,
+      "learning_rate": 1.0204800000000003e-05,
+      "loss": 2.0284,
+      "step": 1811
+    },
+    {
+      "epoch": 3.34,
+      "learning_rate": 1.02016e-05,
+      "loss": 2.0378,
+      "step": 1812
+    },
+    {
+      "epoch": 3.34,
+      "learning_rate": 1.01984e-05,
+      "loss": 1.965,
+      "step": 1813
+    },
+    {
+      "epoch": 3.35,
+      "learning_rate": 1.0195200000000002e-05,
+      "loss": 1.9807,
+      "step": 1814
+    },
+    {
+      "epoch": 3.35,
+      "learning_rate": 1.0192000000000002e-05,
+      "loss": 2.015,
+      "step": 1815
+    },
+    {
+      "epoch": 3.35,
+      "learning_rate": 1.0188800000000001e-05,
+      "loss": 1.9717,
+      "step": 1816
+    },
+    {
+      "epoch": 3.35,
+      "learning_rate": 1.0185600000000003e-05,
+      "loss": 2.0454,
+      "step": 1817
+    },
+    {
+      "epoch": 3.35,
+      "learning_rate": 1.0182400000000001e-05,
+      "loss": 2.0485,
+      "step": 1818
+    },
+    {
+      "epoch": 3.36,
+      "learning_rate": 1.01792e-05,
+      "loss": 2.0221,
+      "step": 1819
+    },
+    {
+      "epoch": 3.36,
+      "learning_rate": 1.0176000000000002e-05,
+      "loss": 2.0217,
+      "step": 1820
+    },
+    {
+      "epoch": 3.36,
+      "learning_rate": 1.0172800000000002e-05,
+      "loss": 1.9151,
+      "step": 1821
+    },
+    {
+      "epoch": 3.36,
+      "learning_rate": 1.0169600000000003e-05,
+      "loss": 2.1617,
+      "step": 1822
+    },
+    {
+      "epoch": 3.36,
+      "learning_rate": 1.0166400000000001e-05,
+      "loss": 2.0554,
+      "step": 1823
+    },
+    {
+      "epoch": 3.36,
+      "learning_rate": 1.0163200000000001e-05,
+      "loss": 1.9888,
+      "step": 1824
+    },
+    {
+      "epoch": 3.37,
+      "learning_rate": 1.0160000000000003e-05,
+      "loss": 2.078,
+      "step": 1825
+    },
+    {
+      "epoch": 3.37,
+      "learning_rate": 1.0156800000000002e-05,
+      "loss": 2.1114,
+      "step": 1826
+    },
+    {
+      "epoch": 3.37,
+      "learning_rate": 1.0153600000000002e-05,
+      "loss": 2.0307,
+      "step": 1827
+    },
+    {
+      "epoch": 3.37,
+      "learning_rate": 1.0150400000000002e-05,
+      "loss": 2.0045,
+      "step": 1828
+    },
+    {
+      "epoch": 3.37,
+      "learning_rate": 1.0147200000000002e-05,
+      "loss": 1.9711,
+      "step": 1829
+    },
+    {
+      "epoch": 3.38,
+      "learning_rate": 1.0144000000000001e-05,
+      "loss": 1.9801,
+      "step": 1830
+    },
+    {
+      "epoch": 3.38,
+      "eval_loss": 2.300424575805664,
+      "eval_runtime": 231.6111,
+      "eval_samples_per_second": 37.446,
+      "eval_steps_per_second": 0.587,
+      "step": 1830
+    },
+    {
+      "epoch": 3.38,
+      "learning_rate": 1.0140800000000003e-05,
+      "loss": 2.0928,
+      "step": 1831
+    },
+    {
+      "epoch": 3.38,
+      "learning_rate": 1.0137600000000003e-05,
+      "loss": 2.0256,
+      "step": 1832
+    },
+    {
+      "epoch": 3.38,
+      "learning_rate": 1.01344e-05,
+      "loss": 2.0005,
+      "step": 1833
+    },
+    {
+      "epoch": 3.38,
+      "learning_rate": 1.0131200000000002e-05,
+      "loss": 2.0141,
+      "step": 1834
+    },
+    {
+      "epoch": 3.39,
+      "learning_rate": 1.0128000000000002e-05,
+      "loss": 2.0508,
+      "step": 1835
+    },
+    {
+      "epoch": 3.39,
+      "learning_rate": 1.0124800000000002e-05,
+      "loss": 2.0028,
+      "step": 1836
+    },
+    {
+      "epoch": 3.39,
+      "learning_rate": 1.0121600000000003e-05,
+      "loss": 1.9719,
+      "step": 1837
+    },
+    {
+      "epoch": 3.39,
+      "learning_rate": 1.0118400000000001e-05,
+      "loss": 2.0086,
+      "step": 1838
+    },
+    {
+      "epoch": 3.39,
+      "learning_rate": 1.0115200000000001e-05,
+      "loss": 2.0226,
+      "step": 1839
+    },
+    {
+      "epoch": 3.39,
+      "learning_rate": 1.0112000000000002e-05,
+      "loss": 2.1207,
+      "step": 1840
+    },
+    {
+      "epoch": 3.4,
+      "learning_rate": 1.0108800000000002e-05,
+      "loss": 2.0665,
+      "step": 1841
+    },
+    {
+      "epoch": 3.4,
+      "learning_rate": 1.0105600000000002e-05,
+      "loss": 2.0667,
+      "step": 1842
+    },
+    {
+      "epoch": 3.4,
+      "learning_rate": 1.0102400000000002e-05,
+      "loss": 2.072,
+      "step": 1843
+    },
+    {
+      "epoch": 3.4,
+      "learning_rate": 1.0099200000000001e-05,
+      "loss": 1.9796,
+      "step": 1844
+    },
+    {
+      "epoch": 3.4,
+      "learning_rate": 1.0096000000000001e-05,
+      "loss": 2.095,
+      "step": 1845
+    },
+    {
+      "epoch": 3.41,
+      "learning_rate": 1.0092800000000003e-05,
+      "loss": 2.0925,
+      "step": 1846
+    },
+    {
+      "epoch": 3.41,
+      "learning_rate": 1.0089600000000002e-05,
+      "loss": 2.0388,
+      "step": 1847
+    },
+    {
+      "epoch": 3.41,
+      "learning_rate": 1.00864e-05,
+      "loss": 1.9194,
+      "step": 1848
+    },
+    {
+      "epoch": 3.41,
+      "learning_rate": 1.0083200000000002e-05,
+      "loss": 2.0249,
+      "step": 1849
+    },
+    {
+      "epoch": 3.41,
+      "learning_rate": 1.0080000000000002e-05,
+      "loss": 1.981,
+      "step": 1850
+    },
+    {
+      "epoch": 3.41,
+      "learning_rate": 1.0076800000000002e-05,
+      "loss": 2.063,
+      "step": 1851
+    },
+    {
+      "epoch": 3.42,
+      "learning_rate": 1.0073600000000003e-05,
+      "loss": 1.9559,
+      "step": 1852
+    },
+    {
+      "epoch": 3.42,
+      "learning_rate": 1.0070400000000001e-05,
+      "loss": 2.0505,
+      "step": 1853
+    },
+    {
+      "epoch": 3.42,
+      "learning_rate": 1.00672e-05,
+      "loss": 1.9894,
+      "step": 1854
+    },
+    {
+      "epoch": 3.42,
+      "learning_rate": 1.0064000000000002e-05,
+      "loss": 1.953,
+      "step": 1855
+    },
+    {
+      "epoch": 3.42,
+      "learning_rate": 1.0060800000000002e-05,
+      "loss": 2.0236,
+      "step": 1856
+    },
+    {
+      "epoch": 3.43,
+      "learning_rate": 1.0057600000000002e-05,
+      "loss": 2.1326,
+      "step": 1857
+    },
+    {
+      "epoch": 3.43,
+      "learning_rate": 1.0054400000000002e-05,
+      "loss": 2.0795,
+      "step": 1858
+    },
+    {
+      "epoch": 3.43,
+      "learning_rate": 1.0051200000000001e-05,
+      "loss": 2.0466,
+      "step": 1859
+    },
+    {
+      "epoch": 3.43,
+      "learning_rate": 1.0048000000000001e-05,
+      "loss": 2.0344,
+      "step": 1860
+    },
+    {
+      "epoch": 3.43,
+      "eval_loss": 2.3100414276123047,
+      "eval_runtime": 231.774,
+      "eval_samples_per_second": 37.42,
+      "eval_steps_per_second": 0.587,
+      "step": 1860
+    },
+    {
+      "epoch": 3.43,
+      "learning_rate": 1.0044800000000002e-05,
+      "loss": 2.0099,
+      "step": 1861
+    },
+    {
+      "epoch": 3.44,
+      "learning_rate": 1.0041600000000002e-05,
+      "loss": 1.9718,
+      "step": 1862
+    },
+    {
+      "epoch": 3.44,
+      "learning_rate": 1.00384e-05,
+      "loss": 2.102,
+      "step": 1863
+    },
+    {
+      "epoch": 3.44,
+      "learning_rate": 1.0035200000000002e-05,
+      "loss": 2.0208,
+      "step": 1864
+    },
+    {
+      "epoch": 3.44,
+      "learning_rate": 1.0032000000000002e-05,
+      "loss": 1.9554,
+      "step": 1865
+    },
+    {
+      "epoch": 3.44,
+      "learning_rate": 1.0028800000000001e-05,
+      "loss": 1.9648,
+      "step": 1866
+    },
+    {
+      "epoch": 3.44,
+      "learning_rate": 1.0025600000000003e-05,
+      "loss": 1.9051,
+      "step": 1867
+    },
+    {
+      "epoch": 3.45,
+      "learning_rate": 1.00224e-05,
+      "loss": 2.0713,
+      "step": 1868
+    },
+    {
+      "epoch": 3.45,
+      "learning_rate": 1.00192e-05,
+      "loss": 1.916,
+      "step": 1869
+    },
+    {
+      "epoch": 3.45,
+      "learning_rate": 1.0016000000000002e-05,
+      "loss": 2.0645,
+      "step": 1870
+    },
+    {
+      "epoch": 3.45,
+      "learning_rate": 1.0012800000000002e-05,
+      "loss": 2.0147,
+      "step": 1871
+    },
+    {
+      "epoch": 3.45,
+      "learning_rate": 1.0009600000000002e-05,
+      "loss": 1.9784,
+      "step": 1872
+    },
+    {
+      "epoch": 3.46,
+      "learning_rate": 1.0006400000000001e-05,
+      "loss": 2.0178,
+      "step": 1873
+    },
+    {
+      "epoch": 3.46,
+      "learning_rate": 1.0003200000000001e-05,
+      "loss": 1.9462,
+      "step": 1874
+    },
+    {
+      "epoch": 3.46,
+      "learning_rate": 1.0000000000000003e-05,
+      "loss": 2.114,
+      "step": 1875
+    },
+    {
+      "epoch": 3.46,
+      "learning_rate": 9.996800000000002e-06,
+      "loss": 2.0924,
+      "step": 1876
+    },
+    {
+      "epoch": 3.46,
+      "learning_rate": 9.993600000000002e-06,
+      "loss": 2.1227,
+      "step": 1877
+    },
+    {
+      "epoch": 3.46,
+      "learning_rate": 9.990400000000002e-06,
+      "loss": 2.068,
+      "step": 1878
+    },
+    {
+      "epoch": 3.47,
+      "learning_rate": 9.987200000000002e-06,
+      "loss": 2.1466,
+      "step": 1879
+    },
+    {
+      "epoch": 3.47,
+      "learning_rate": 9.984000000000001e-06,
+      "loss": 1.9498,
+      "step": 1880
+    },
+    {
+      "epoch": 3.47,
+      "learning_rate": 9.980800000000003e-06,
+      "loss": 1.9197,
+      "step": 1881
+    },
+    {
+      "epoch": 3.47,
+      "learning_rate": 9.977600000000003e-06,
+      "loss": 1.9376,
+      "step": 1882
+    },
+    {
+      "epoch": 3.47,
+      "learning_rate": 9.9744e-06,
+      "loss": 1.9923,
+      "step": 1883
+    },
+    {
+      "epoch": 3.48,
+      "learning_rate": 9.971200000000002e-06,
+      "loss": 2.026,
+      "step": 1884
+    },
+    {
+      "epoch": 3.48,
+      "learning_rate": 9.968000000000002e-06,
+      "loss": 2.0615,
+      "step": 1885
+    },
+    {
+      "epoch": 3.48,
+      "learning_rate": 9.964800000000002e-06,
+      "loss": 2.0883,
+      "step": 1886
+    },
+    {
+      "epoch": 3.48,
+      "learning_rate": 9.961600000000003e-06,
+      "loss": 2.0661,
+      "step": 1887
+    },
+    {
+      "epoch": 3.48,
+      "learning_rate": 9.958400000000001e-06,
+      "loss": 2.0151,
+      "step": 1888
+    },
+    {
+      "epoch": 3.48,
+      "learning_rate": 9.9552e-06,
+      "loss": 2.0385,
+      "step": 1889
+    },
+    {
+      "epoch": 3.49,
+      "learning_rate": 9.952000000000002e-06,
+      "loss": 2.0254,
+      "step": 1890
+    },
+    {
+      "epoch": 3.49,
+      "eval_loss": 2.3001210689544678,
+      "eval_runtime": 230.6869,
+      "eval_samples_per_second": 37.596,
+      "eval_steps_per_second": 0.59,
+      "step": 1890
+    },
+    {
+      "epoch": 3.49,
+      "learning_rate": 9.948800000000002e-06,
+      "loss": 2.0761,
+      "step": 1891
+    },
+    {
+      "epoch": 3.49,
+      "learning_rate": 9.945600000000002e-06,
+      "loss": 2.0318,
+      "step": 1892
+    },
+    {
+      "epoch": 3.49,
+      "learning_rate": 9.942400000000002e-06,
+      "loss": 1.948,
+      "step": 1893
+    },
+    {
+      "epoch": 3.49,
+      "learning_rate": 9.939200000000001e-06,
+      "loss": 2.0347,
+      "step": 1894
+    },
+    {
+      "epoch": 3.5,
+      "learning_rate": 9.936000000000001e-06,
+      "loss": 2.1169,
+      "step": 1895
+    },
+    {
+      "epoch": 3.5,
+      "learning_rate": 9.932800000000003e-06,
+      "loss": 1.979,
+      "step": 1896
+    },
+    {
+      "epoch": 3.5,
+      "learning_rate": 9.929600000000002e-06,
+      "loss": 1.9615,
+      "step": 1897
+    },
+    {
+      "epoch": 3.5,
+      "learning_rate": 9.9264e-06,
+      "loss": 2.0118,
+      "step": 1898
+    },
+    {
+      "epoch": 3.5,
+      "learning_rate": 9.923200000000002e-06,
+      "loss": 1.8853,
+      "step": 1899
+    },
+    {
+      "epoch": 3.51,
+      "learning_rate": 9.920000000000002e-06,
+      "loss": 1.9853,
+      "step": 1900
+    },
+    {
+      "epoch": 3.51,
+      "learning_rate": 9.916800000000001e-06,
+      "loss": 1.8975,
+      "step": 1901
+    },
+    {
+      "epoch": 3.51,
+      "learning_rate": 9.913600000000003e-06,
+      "loss": 2.084,
+      "step": 1902
+    },
+    {
+      "epoch": 3.51,
+      "learning_rate": 9.910400000000001e-06,
+      "loss": 2.1225,
+      "step": 1903
+    },
+    {
+      "epoch": 3.51,
+      "learning_rate": 9.9072e-06,
+      "loss": 1.9743,
+      "step": 1904
+    },
+    {
+      "epoch": 3.51,
+      "learning_rate": 9.904000000000002e-06,
+      "loss": 2.0148,
+      "step": 1905
+    },
+    {
+      "epoch": 3.52,
+      "learning_rate": 9.900800000000002e-06,
+      "loss": 2.1019,
+      "step": 1906
+    },
+    {
+      "epoch": 3.52,
+      "learning_rate": 9.897600000000002e-06,
+      "loss": 1.9885,
+      "step": 1907
+    },
+    {
+      "epoch": 3.52,
+      "learning_rate": 9.894400000000001e-06,
+      "loss": 2.0369,
+      "step": 1908
+    },
+    {
+      "epoch": 3.52,
+      "learning_rate": 9.891200000000001e-06,
+      "loss": 2.0306,
+      "step": 1909
+    },
+    {
+      "epoch": 3.52,
+      "learning_rate": 9.888000000000001e-06,
+      "loss": 1.9708,
+      "step": 1910
+    },
+    {
+      "epoch": 3.53,
+      "learning_rate": 9.884800000000002e-06,
+      "loss": 1.985,
+      "step": 1911
+    },
+    {
+      "epoch": 3.53,
+      "learning_rate": 9.881600000000002e-06,
+      "loss": 2.0791,
+      "step": 1912
+    },
+    {
+      "epoch": 3.53,
+      "learning_rate": 9.8784e-06,
+      "loss": 2.059,
+      "step": 1913
+    },
+    {
+      "epoch": 3.53,
+      "learning_rate": 9.875200000000002e-06,
+      "loss": 2.0259,
+      "step": 1914
+    },
+    {
+      "epoch": 3.53,
+      "learning_rate": 9.872000000000001e-06,
+      "loss": 1.977,
+      "step": 1915
+    },
+    {
+      "epoch": 3.53,
+      "learning_rate": 9.868800000000001e-06,
+      "loss": 2.0855,
+      "step": 1916
+    },
+    {
+      "epoch": 3.54,
+      "learning_rate": 9.865600000000003e-06,
+      "loss": 1.9539,
+      "step": 1917
+    },
+    {
+      "epoch": 3.54,
+      "learning_rate": 9.8624e-06,
+      "loss": 2.0435,
+      "step": 1918
+    },
+    {
+      "epoch": 3.54,
+      "learning_rate": 9.8592e-06,
+      "loss": 2.1338,
+      "step": 1919
+    },
+    {
+      "epoch": 3.54,
+      "learning_rate": 9.856000000000002e-06,
+      "loss": 1.9719,
+      "step": 1920
+    },
+    {
+      "epoch": 3.54,
+      "eval_loss": 2.3002588748931885,
+      "eval_runtime": 232.3342,
+      "eval_samples_per_second": 37.33,
+      "eval_steps_per_second": 0.585,
+      "step": 1920
+    },
+    {
+      "epoch": 3.54,
+      "learning_rate": 9.852800000000002e-06,
+      "loss": 2.1076,
+      "step": 1921
+    },
+    {
+      "epoch": 3.55,
+      "learning_rate": 9.849600000000001e-06,
+      "loss": 2.1559,
+      "step": 1922
+    },
+    {
+      "epoch": 3.55,
+      "learning_rate": 9.846400000000001e-06,
+      "loss": 1.9805,
+      "step": 1923
+    },
+    {
+      "epoch": 3.55,
+      "learning_rate": 9.843200000000001e-06,
+      "loss": 1.9651,
+      "step": 1924
+    },
+    {
+      "epoch": 3.55,
+      "learning_rate": 9.84e-06,
+      "loss": 2.0652,
+      "step": 1925
+    },
+    {
+      "epoch": 3.55,
+      "learning_rate": 9.836800000000002e-06,
+      "loss": 2.0221,
+      "step": 1926
+    },
+    {
+      "epoch": 3.55,
+      "learning_rate": 9.833600000000002e-06,
+      "loss": 1.948,
+      "step": 1927
+    },
+    {
+      "epoch": 3.56,
+      "learning_rate": 9.8304e-06,
+      "loss": 2.0203,
+      "step": 1928
+    },
+    {
+      "epoch": 3.56,
+      "learning_rate": 9.827200000000001e-06,
+      "loss": 1.8977,
+      "step": 1929
+    },
+    {
+      "epoch": 3.56,
+      "learning_rate": 9.824000000000001e-06,
+      "loss": 2.0157,
+      "step": 1930
+    },
+    {
+      "epoch": 3.56,
+      "learning_rate": 9.820800000000003e-06,
+      "loss": 2.1581,
+      "step": 1931
+    },
+    {
+      "epoch": 3.56,
+      "learning_rate": 9.817600000000002e-06,
+      "loss": 2.0923,
+      "step": 1932
+    },
+    {
+      "epoch": 3.57,
+      "learning_rate": 9.8144e-06,
+      "loss": 1.9152,
+      "step": 1933
+    },
+    {
+      "epoch": 3.57,
+      "learning_rate": 9.811200000000002e-06,
+      "loss": 2.0708,
+      "step": 1934
+    },
+    {
+      "epoch": 3.57,
+      "learning_rate": 9.808000000000002e-06,
+      "loss": 2.0313,
+      "step": 1935
+    },
+    {
+      "epoch": 3.57,
+      "learning_rate": 9.804800000000001e-06,
+      "loss": 2.0558,
+      "step": 1936
+    },
+    {
+      "epoch": 3.57,
+      "learning_rate": 9.801600000000003e-06,
+      "loss": 2.0661,
+      "step": 1937
+    },
+    {
+      "epoch": 3.58,
+      "learning_rate": 9.798400000000003e-06,
+      "loss": 2.0418,
+      "step": 1938
+    },
+    {
+      "epoch": 3.58,
+      "learning_rate": 9.7952e-06,
+      "loss": 2.054,
+      "step": 1939
+    },
+    {
+      "epoch": 3.58,
+      "learning_rate": 9.792000000000002e-06,
+      "loss": 2.0752,
+      "step": 1940
+    },
+    {
+      "epoch": 3.58,
+      "learning_rate": 9.788800000000002e-06,
+      "loss": 2.084,
+      "step": 1941
+    },
+    {
+      "epoch": 3.58,
+      "learning_rate": 9.785600000000002e-06,
+      "loss": 2.0736,
+      "step": 1942
+    },
+    {
+      "epoch": 3.58,
+      "learning_rate": 9.782400000000003e-06,
+      "loss": 2.0587,
+      "step": 1943
+    },
+    {
+      "epoch": 3.59,
+      "learning_rate": 9.779200000000001e-06,
+      "loss": 2.0264,
+      "step": 1944
+    },
+    {
+      "epoch": 3.59,
+      "learning_rate": 9.776000000000001e-06,
+      "loss": 1.9863,
+      "step": 1945
+    },
+    {
+      "epoch": 3.59,
+      "learning_rate": 9.772800000000002e-06,
+      "loss": 2.0989,
+      "step": 1946
+    },
+    {
+      "epoch": 3.59,
+      "learning_rate": 9.769600000000002e-06,
+      "loss": 2.1228,
+      "step": 1947
+    },
+    {
+      "epoch": 3.59,
+      "learning_rate": 9.766400000000002e-06,
+      "loss": 1.9614,
+      "step": 1948
+    },
+    {
+      "epoch": 3.6,
+      "learning_rate": 9.763200000000002e-06,
+      "loss": 2.0569,
+      "step": 1949
+    },
+    {
+      "epoch": 3.6,
+      "learning_rate": 9.760000000000001e-06,
+      "loss": 1.9952,
+      "step": 1950
+    },
+    {
+      "epoch": 3.6,
+      "eval_loss": 2.298185348510742,
+      "eval_runtime": 232.0334,
+      "eval_samples_per_second": 37.378,
+      "eval_steps_per_second": 0.586,
+      "step": 1950
+    },
+    {
+      "epoch": 3.6,
+      "learning_rate": 9.756800000000001e-06,
+      "loss": 1.9445,
+      "step": 1951
+    },
+    {
+      "epoch": 3.6,
+      "learning_rate": 9.753600000000003e-06,
+      "loss": 2.0083,
+      "step": 1952
+    },
+    {
+      "epoch": 3.6,
+      "learning_rate": 9.750400000000002e-06,
+      "loss": 1.9664,
+      "step": 1953
+    },
+    {
+      "epoch": 3.6,
+      "learning_rate": 9.7472e-06,
+      "loss": 1.9641,
+      "step": 1954
+    },
+    {
+      "epoch": 3.61,
+      "learning_rate": 9.744000000000002e-06,
+      "loss": 1.9301,
+      "step": 1955
+    },
+    {
+      "epoch": 3.61,
+      "learning_rate": 9.740800000000002e-06,
+      "loss": 2.0642,
+      "step": 1956
+    },
+    {
+      "epoch": 3.61,
+      "learning_rate": 9.737600000000001e-06,
+      "loss": 1.9078,
+      "step": 1957
+    },
+    {
+      "epoch": 3.61,
+      "learning_rate": 9.734400000000003e-06,
+      "loss": 2.0599,
+      "step": 1958
+    },
+    {
+      "epoch": 3.61,
+      "learning_rate": 9.731200000000001e-06,
+      "loss": 1.9509,
+      "step": 1959
+    },
+    {
+      "epoch": 3.62,
+      "learning_rate": 9.728e-06,
+      "loss": 2.0652,
+      "step": 1960
+    },
+    {
+      "epoch": 3.62,
+      "learning_rate": 9.724800000000002e-06,
+      "loss": 2.0378,
+      "step": 1961
+    },
+    {
+      "epoch": 3.62,
+      "learning_rate": 9.721600000000002e-06,
+      "loss": 2.0361,
+      "step": 1962
+    },
+    {
+      "epoch": 3.62,
+      "learning_rate": 9.718400000000002e-06,
+      "loss": 2.0629,
+      "step": 1963
+    },
+    {
+      "epoch": 3.62,
+      "learning_rate": 9.715200000000001e-06,
+      "loss": 2.0236,
+      "step": 1964
+    },
+    {
+      "epoch": 3.62,
+      "learning_rate": 9.712000000000001e-06,
+      "loss": 2.0492,
+      "step": 1965
+    },
+    {
+      "epoch": 3.63,
+      "learning_rate": 9.708800000000001e-06,
+      "loss": 2.0178,
+      "step": 1966
+    },
+    {
+      "epoch": 3.63,
+      "learning_rate": 9.705600000000002e-06,
+      "loss": 2.0596,
+      "step": 1967
+    },
+    {
+      "epoch": 3.63,
+      "learning_rate": 9.702400000000002e-06,
+      "loss": 2.0295,
+      "step": 1968
+    },
+    {
+      "epoch": 3.63,
+      "learning_rate": 9.6992e-06,
+      "loss": 1.9497,
+      "step": 1969
+    },
+    {
+      "epoch": 3.63,
+      "learning_rate": 9.696000000000002e-06,
+      "loss": 1.9882,
+      "step": 1970
+    },
+    {
+      "epoch": 3.64,
+      "learning_rate": 9.692800000000001e-06,
+      "loss": 2.1012,
+      "step": 1971
+    },
+    {
+      "epoch": 3.64,
+      "learning_rate": 9.689600000000001e-06,
+      "loss": 2.0541,
+      "step": 1972
+    },
+    {
+      "epoch": 3.64,
+      "learning_rate": 9.686400000000003e-06,
+      "loss": 2.0092,
+      "step": 1973
+    },
+    {
+      "epoch": 3.64,
+      "learning_rate": 9.6832e-06,
+      "loss": 2.0813,
+      "step": 1974
+    },
+    {
+      "epoch": 3.64,
+      "learning_rate": 9.68e-06,
+      "loss": 1.9697,
+      "step": 1975
+    },
+    {
+      "epoch": 3.65,
+      "learning_rate": 9.676800000000002e-06,
+      "loss": 1.9809,
+      "step": 1976
+    },
+    {
+      "epoch": 3.65,
+      "learning_rate": 9.673600000000002e-06,
+      "loss": 2.0808,
+      "step": 1977
+    },
+    {
+      "epoch": 3.65,
+      "learning_rate": 9.670400000000003e-06,
+      "loss": 2.0115,
+      "step": 1978
+    },
+    {
+      "epoch": 3.65,
+      "learning_rate": 9.667200000000001e-06,
+      "loss": 1.9972,
+      "step": 1979
+    },
+    {
+      "epoch": 3.65,
+      "learning_rate": 9.664000000000001e-06,
+      "loss": 2.0261,
+      "step": 1980
+    },
+    {
+      "epoch": 3.65,
+      "eval_loss": 2.306427001953125,
+      "eval_runtime": 232.5714,
+      "eval_samples_per_second": 37.292,
+      "eval_steps_per_second": 0.585,
+      "step": 1980
+    },
+    {
+      "epoch": 3.65,
+      "learning_rate": 9.660800000000002e-06,
+      "loss": 2.1576,
+      "step": 1981
+    },
+    {
+      "epoch": 3.66,
+      "learning_rate": 9.657600000000002e-06,
+      "loss": 1.9639,
+      "step": 1982
+    },
+    {
+      "epoch": 3.66,
+      "learning_rate": 9.654400000000002e-06,
+      "loss": 1.9921,
+      "step": 1983
+    },
+    {
+      "epoch": 3.66,
+      "learning_rate": 9.651200000000002e-06,
+      "loss": 2.0429,
+      "step": 1984
+    },
+    {
+      "epoch": 3.66,
+      "learning_rate": 9.648000000000001e-06,
+      "loss": 2.0803,
+      "step": 1985
+    },
+    {
+      "epoch": 3.66,
+      "learning_rate": 9.644800000000001e-06,
+      "loss": 2.0073,
+      "step": 1986
+    },
+    {
+      "epoch": 3.67,
+      "learning_rate": 9.641600000000003e-06,
+      "loss": 2.1166,
+      "step": 1987
+    },
+    {
+      "epoch": 3.67,
+      "learning_rate": 9.638400000000002e-06,
+      "loss": 2.0516,
+      "step": 1988
+    },
+    {
+      "epoch": 3.67,
+      "learning_rate": 9.6352e-06,
+      "loss": 1.9751,
+      "step": 1989
+    },
+    {
+      "epoch": 3.67,
+      "learning_rate": 9.632000000000002e-06,
+      "loss": 2.041,
+      "step": 1990
+    },
+    {
+      "epoch": 3.67,
+      "learning_rate": 9.628800000000002e-06,
+      "loss": 2.0529,
+      "step": 1991
+    },
+    {
+      "epoch": 3.67,
+      "learning_rate": 9.625600000000001e-06,
+      "loss": 2.0635,
+      "step": 1992
+    },
+    {
+      "epoch": 3.68,
+      "learning_rate": 9.622400000000003e-06,
+      "loss": 1.9942,
+      "step": 1993
+    },
+    {
+      "epoch": 3.68,
+      "learning_rate": 9.619200000000001e-06,
+      "loss": 1.9801,
+      "step": 1994
+    },
+    {
+      "epoch": 3.68,
+      "learning_rate": 9.616e-06,
+      "loss": 2.0477,
+      "step": 1995
+    },
+    {
+      "epoch": 3.68,
+      "learning_rate": 9.612800000000002e-06,
+      "loss": 2.0364,
+      "step": 1996
+    },
+    {
+      "epoch": 3.68,
+      "learning_rate": 9.609600000000002e-06,
+      "loss": 2.0551,
+      "step": 1997
+    },
+    {
+      "epoch": 3.69,
+      "learning_rate": 9.606400000000002e-06,
+      "loss": 1.9867,
+      "step": 1998
+    },
+    {
+      "epoch": 3.69,
+      "learning_rate": 9.603200000000002e-06,
+      "loss": 2.0515,
+      "step": 1999
+    },
+    {
+      "epoch": 3.69,
+      "learning_rate": 9.600000000000001e-06,
+      "loss": 2.0486,
+      "step": 2000
+    },
+    {
+      "epoch": 3.69,
+      "learning_rate": 9.596800000000001e-06,
+      "loss": 2.0562,
+      "step": 2001
+    },
+    {
+      "epoch": 3.69,
+      "learning_rate": 9.593600000000002e-06,
+      "loss": 2.0492,
+      "step": 2002
+    },
+    {
+      "epoch": 3.69,
+      "learning_rate": 9.590400000000002e-06,
+      "loss": 1.9445,
+      "step": 2003
+    },
+    {
+      "epoch": 3.7,
+      "learning_rate": 9.5872e-06,
+      "loss": 1.9764,
+      "step": 2004
+    },
+    {
+      "epoch": 3.7,
+      "learning_rate": 9.584000000000002e-06,
+      "loss": 1.951,
+      "step": 2005
+    },
+    {
+      "epoch": 3.7,
+      "learning_rate": 9.580800000000002e-06,
+      "loss": 2.0235,
+      "step": 2006
+    },
+    {
+      "epoch": 3.7,
+      "learning_rate": 9.577600000000001e-06,
+      "loss": 1.9376,
+      "step": 2007
+    },
+    {
+      "epoch": 3.7,
+      "learning_rate": 9.574400000000003e-06,
+      "loss": 2.0355,
+      "step": 2008
+    },
+    {
+      "epoch": 3.71,
+      "learning_rate": 9.5712e-06,
+      "loss": 2.0686,
+      "step": 2009
+    },
+    {
+      "epoch": 3.71,
+      "learning_rate": 9.568e-06,
+      "loss": 2.0919,
+      "step": 2010
+    },
+    {
+      "epoch": 3.71,
+      "eval_loss": 2.301406145095825,
+      "eval_runtime": 232.6485,
+      "eval_samples_per_second": 37.279,
+      "eval_steps_per_second": 0.585,
+      "step": 2010
+    },
+    {
+      "epoch": 3.71,
+      "learning_rate": 9.564800000000002e-06,
+      "loss": 2.0,
+      "step": 2011
+    },
+    {
+      "epoch": 3.71,
+      "learning_rate": 9.561600000000002e-06,
+      "loss": 2.0869,
+      "step": 2012
+    },
+    {
+      "epoch": 3.71,
+      "learning_rate": 9.558400000000002e-06,
+      "loss": 2.11,
+      "step": 2013
+    },
+    {
+      "epoch": 3.72,
+      "learning_rate": 9.555200000000001e-06,
+      "loss": 2.0304,
+      "step": 2014
+    },
+    {
+      "epoch": 3.72,
+      "learning_rate": 9.552000000000001e-06,
+      "loss": 2.1146,
+      "step": 2015
+    },
+    {
+      "epoch": 3.72,
+      "learning_rate": 9.5488e-06,
+      "loss": 2.0128,
+      "step": 2016
+    },
+    {
+      "epoch": 3.72,
+      "learning_rate": 9.545600000000002e-06,
+      "loss": 2.032,
+      "step": 2017
+    },
+    {
+      "epoch": 3.72,
+      "learning_rate": 9.542400000000002e-06,
+      "loss": 2.1155,
+      "step": 2018
+    },
+    {
+      "epoch": 3.72,
+      "learning_rate": 9.5392e-06,
+      "loss": 2.1136,
+      "step": 2019
+    },
+    {
+      "epoch": 3.73,
+      "learning_rate": 9.536000000000002e-06,
+      "loss": 2.0661,
+      "step": 2020
+    },
+    {
+      "epoch": 3.73,
+      "learning_rate": 9.532800000000001e-06,
+      "loss": 2.0361,
+      "step": 2021
+    },
+    {
+      "epoch": 3.73,
+      "learning_rate": 9.529600000000001e-06,
+      "loss": 1.9669,
+      "step": 2022
+    },
+    {
+      "epoch": 3.73,
+      "learning_rate": 9.526400000000003e-06,
+      "loss": 2.0161,
+      "step": 2023
+    },
+    {
+      "epoch": 3.73,
+      "learning_rate": 9.5232e-06,
+      "loss": 1.9443,
+      "step": 2024
+    },
+    {
+      "epoch": 3.74,
+      "learning_rate": 9.52e-06,
+      "loss": 2.0134,
+      "step": 2025
+    },
+    {
+      "epoch": 3.74,
+      "learning_rate": 9.516800000000002e-06,
+      "loss": 2.0322,
+      "step": 2026
+    },
+    {
+      "epoch": 3.74,
+      "learning_rate": 9.513600000000002e-06,
+      "loss": 2.0955,
+      "step": 2027
+    },
+    {
+      "epoch": 3.74,
+      "learning_rate": 9.510400000000003e-06,
+      "loss": 2.006,
+      "step": 2028
+    },
+    {
+      "epoch": 3.74,
+      "learning_rate": 9.507200000000001e-06,
+      "loss": 2.102,
+      "step": 2029
+    },
+    {
+      "epoch": 3.74,
+      "learning_rate": 9.504e-06,
+      "loss": 1.9389,
+      "step": 2030
+    },
+    {
+      "epoch": 3.75,
+      "learning_rate": 9.500800000000002e-06,
+      "loss": 2.0946,
+      "step": 2031
+    },
+    {
+      "epoch": 3.75,
+      "learning_rate": 9.497600000000002e-06,
+      "loss": 2.0141,
+      "step": 2032
+    },
+    {
+      "epoch": 3.75,
+      "learning_rate": 9.494400000000002e-06,
+      "loss": 2.0687,
+      "step": 2033
+    },
+    {
+      "epoch": 3.75,
+      "learning_rate": 9.491200000000002e-06,
+      "loss": 1.962,
+      "step": 2034
+    },
+    {
+      "epoch": 3.75,
+      "learning_rate": 9.488000000000001e-06,
+      "loss": 2.0208,
+      "step": 2035
+    },
+    {
+      "epoch": 3.76,
+      "learning_rate": 9.484800000000001e-06,
+      "loss": 1.8806,
+      "step": 2036
+    },
+    {
+      "epoch": 3.76,
+      "learning_rate": 9.481600000000003e-06,
+      "loss": 2.0968,
+      "step": 2037
+    },
+    {
+      "epoch": 3.76,
+      "learning_rate": 9.478400000000002e-06,
+      "loss": 1.9999,
+      "step": 2038
+    },
+    {
+      "epoch": 3.76,
+      "learning_rate": 9.4752e-06,
+      "loss": 2.1199,
+      "step": 2039
+    },
+    {
+      "epoch": 3.76,
+      "learning_rate": 9.472000000000002e-06,
+      "loss": 1.9706,
+      "step": 2040
+    },
+    {
+      "epoch": 3.76,
+      "eval_loss": 2.3070294857025146,
+      "eval_runtime": 232.2219,
+      "eval_samples_per_second": 37.348,
+      "eval_steps_per_second": 0.586,
+      "step": 2040
+    },
+    {
+      "epoch": 3.76,
+      "learning_rate": 9.468800000000002e-06,
+      "loss": 2.0887,
+      "step": 2041
+    },
+    {
+      "epoch": 3.77,
+      "learning_rate": 9.465600000000001e-06,
+      "loss": 2.1419,
+      "step": 2042
+    },
+    {
+      "epoch": 3.77,
+      "learning_rate": 9.462400000000003e-06,
+      "loss": 1.9527,
+      "step": 2043
+    },
+    {
+      "epoch": 3.77,
+      "learning_rate": 9.4592e-06,
+      "loss": 2.0024,
+      "step": 2044
+    },
+    {
+      "epoch": 3.77,
+      "learning_rate": 9.456e-06,
+      "loss": 1.9535,
+      "step": 2045
+    },
+    {
+      "epoch": 3.77,
+      "learning_rate": 9.452800000000002e-06,
+      "loss": 2.1175,
+      "step": 2046
+    },
+    {
+      "epoch": 3.78,
+      "learning_rate": 9.449600000000002e-06,
+      "loss": 1.9259,
+      "step": 2047
+    },
+    {
+      "epoch": 3.78,
+      "learning_rate": 9.446400000000002e-06,
+      "loss": 2.0051,
+      "step": 2048
+    },
+    {
+      "epoch": 3.78,
+      "learning_rate": 9.443200000000001e-06,
+      "loss": 2.051,
+      "step": 2049
+    },
+    {
+      "epoch": 3.78,
+      "learning_rate": 9.440000000000001e-06,
+      "loss": 1.9609,
+      "step": 2050
+    },
+    {
+      "epoch": 3.78,
+      "learning_rate": 9.436800000000001e-06,
+      "loss": 2.0081,
+      "step": 2051
+    },
+    {
+      "epoch": 3.79,
+      "learning_rate": 9.433600000000002e-06,
+      "loss": 2.0336,
+      "step": 2052
+    },
+    {
+      "epoch": 3.79,
+      "learning_rate": 9.430400000000002e-06,
+      "loss": 1.9311,
+      "step": 2053
+    },
+    {
+      "epoch": 3.79,
+      "learning_rate": 9.4272e-06,
+      "loss": 1.9634,
+      "step": 2054
+    },
+    {
+      "epoch": 3.79,
+      "learning_rate": 9.424000000000002e-06,
+      "loss": 2.0448,
+      "step": 2055
+    },
+    {
+      "epoch": 3.79,
+      "learning_rate": 9.420800000000001e-06,
+      "loss": 2.0461,
+      "step": 2056
+    },
+    {
+      "epoch": 3.79,
+      "learning_rate": 9.417600000000001e-06,
+      "loss": 2.0266,
+      "step": 2057
+    },
+    {
+      "epoch": 3.8,
+      "learning_rate": 9.414400000000003e-06,
+      "loss": 2.0294,
+      "step": 2058
+    },
+    {
+      "epoch": 3.8,
+      "learning_rate": 9.4112e-06,
+      "loss": 1.9713,
+      "step": 2059
+    },
+    {
+      "epoch": 3.8,
+      "learning_rate": 9.408e-06,
+      "loss": 2.0695,
+      "step": 2060
+    },
+    {
+      "epoch": 3.8,
+      "learning_rate": 9.404800000000002e-06,
+      "loss": 1.9872,
+      "step": 2061
+    },
+    {
+      "epoch": 3.8,
+      "learning_rate": 9.401600000000002e-06,
+      "loss": 1.926,
+      "step": 2062
+    },
+    {
+      "epoch": 3.81,
+      "learning_rate": 9.398400000000001e-06,
+      "loss": 2.0587,
+      "step": 2063
+    },
+    {
+      "epoch": 3.81,
+      "learning_rate": 9.395200000000003e-06,
+      "loss": 2.0685,
+      "step": 2064
+    },
+    {
+      "epoch": 3.81,
+      "learning_rate": 9.392000000000001e-06,
+      "loss": 2.0458,
+      "step": 2065
+    },
+    {
+      "epoch": 3.81,
+      "learning_rate": 9.3888e-06,
+      "loss": 2.0609,
+      "step": 2066
+    },
+    {
+      "epoch": 3.81,
+      "learning_rate": 9.385600000000002e-06,
+      "loss": 1.9377,
+      "step": 2067
+    },
+    {
+      "epoch": 3.81,
+      "learning_rate": 9.382400000000002e-06,
+      "loss": 2.0891,
+      "step": 2068
+    },
+    {
+      "epoch": 3.82,
+      "learning_rate": 9.379200000000002e-06,
+      "loss": 2.0746,
+      "step": 2069
+    },
+    {
+      "epoch": 3.82,
+      "learning_rate": 9.376000000000001e-06,
+      "loss": 1.9684,
+      "step": 2070
+    },
+    {
+      "epoch": 3.82,
+      "eval_loss": 2.2996225357055664,
+      "eval_runtime": 231.2235,
+      "eval_samples_per_second": 37.509,
+      "eval_steps_per_second": 0.588,
+      "step": 2070
+    },
+    {
+      "epoch": 3.82,
+      "learning_rate": 9.372800000000001e-06,
+      "loss": 1.9943,
+      "step": 2071
+    },
+    {
+      "epoch": 3.82,
+      "learning_rate": 9.369600000000001e-06,
+      "loss": 1.9695,
+      "step": 2072
+    },
+    {
+      "epoch": 3.82,
+      "learning_rate": 9.366400000000002e-06,
+      "loss": 2.0166,
+      "step": 2073
+    },
+    {
+      "epoch": 3.83,
+      "learning_rate": 9.363200000000002e-06,
+      "loss": 1.9823,
+      "step": 2074
+    },
+    {
+      "epoch": 3.83,
+      "learning_rate": 9.36e-06,
+      "loss": 2.0003,
+      "step": 2075
+    },
+    {
+      "epoch": 3.83,
+      "learning_rate": 9.356800000000002e-06,
+      "loss": 2.0669,
+      "step": 2076
+    },
+    {
+      "epoch": 3.83,
+      "learning_rate": 9.353600000000001e-06,
+      "loss": 1.9017,
+      "step": 2077
+    },
+    {
+      "epoch": 3.83,
+      "learning_rate": 9.350400000000003e-06,
+      "loss": 2.0468,
+      "step": 2078
+    },
+    {
+      "epoch": 3.84,
+      "learning_rate": 9.347200000000003e-06,
+      "loss": 2.0794,
+      "step": 2079
+    },
+    {
+      "epoch": 3.84,
+      "learning_rate": 9.344e-06,
+      "loss": 2.0067,
+      "step": 2080
+    },
+    {
+      "epoch": 3.84,
+      "learning_rate": 9.340800000000002e-06,
+      "loss": 1.9828,
+      "step": 2081
+    },
+    {
+      "epoch": 3.84,
+      "learning_rate": 9.337600000000002e-06,
+      "loss": 2.0576,
+      "step": 2082
+    },
+    {
+      "epoch": 3.84,
+      "learning_rate": 9.334400000000002e-06,
+      "loss": 1.9267,
+      "step": 2083
+    },
+    {
+      "epoch": 3.84,
+      "learning_rate": 9.331200000000003e-06,
+      "loss": 1.8724,
+      "step": 2084
+    },
+    {
+      "epoch": 3.85,
+      "learning_rate": 9.328000000000001e-06,
+      "loss": 2.0576,
+      "step": 2085
+    },
+    {
+      "epoch": 3.85,
+      "learning_rate": 9.324800000000001e-06,
+      "loss": 1.9575,
+      "step": 2086
+    },
+    {
+      "epoch": 3.85,
+      "learning_rate": 9.321600000000002e-06,
+      "loss": 2.0561,
+      "step": 2087
+    },
+    {
+      "epoch": 3.85,
+      "learning_rate": 9.318400000000002e-06,
+      "loss": 1.9718,
+      "step": 2088
+    },
+    {
+      "epoch": 3.85,
+      "learning_rate": 9.315200000000002e-06,
+      "loss": 2.0399,
+      "step": 2089
+    },
+    {
+      "epoch": 3.86,
+      "learning_rate": 9.312000000000002e-06,
+      "loss": 2.0415,
+      "step": 2090
+    },
+    {
+      "epoch": 3.86,
+      "learning_rate": 9.308800000000001e-06,
+      "loss": 2.0278,
+      "step": 2091
+    },
+    {
+      "epoch": 3.86,
+      "learning_rate": 9.305600000000001e-06,
+      "loss": 1.9753,
+      "step": 2092
+    },
+    {
+      "epoch": 3.86,
+      "learning_rate": 9.302400000000003e-06,
+      "loss": 2.016,
+      "step": 2093
+    },
+    {
+      "epoch": 3.86,
+      "learning_rate": 9.299200000000002e-06,
+      "loss": 2.0376,
+      "step": 2094
+    },
+    {
+      "epoch": 3.86,
+      "learning_rate": 9.296e-06,
+      "loss": 1.9788,
+      "step": 2095
+    },
+    {
+      "epoch": 3.87,
+      "learning_rate": 9.292800000000002e-06,
+      "loss": 2.1005,
+      "step": 2096
+    },
+    {
+      "epoch": 3.87,
+      "learning_rate": 9.289600000000002e-06,
+      "loss": 1.9394,
+      "step": 2097
+    },
+    {
+      "epoch": 3.87,
+      "learning_rate": 9.286400000000001e-06,
+      "loss": 2.0217,
+      "step": 2098
+    },
+    {
+      "epoch": 3.87,
+      "learning_rate": 9.283200000000003e-06,
+      "loss": 1.9995,
+      "step": 2099
+    },
+    {
+      "epoch": 3.87,
+      "learning_rate": 9.280000000000001e-06,
+      "loss": 2.0298,
+      "step": 2100
+    },
+    {
+      "epoch": 3.87,
+      "eval_loss": 2.2996666431427,
+      "eval_runtime": 231.4659,
+      "eval_samples_per_second": 37.47,
+      "eval_steps_per_second": 0.588,
+      "step": 2100
+    },
+    {
+      "epoch": 3.88,
+      "learning_rate": 9.2768e-06,
+      "loss": 1.9012,
+      "step": 2101
+    },
+    {
+      "epoch": 3.88,
+      "learning_rate": 9.273600000000002e-06,
+      "loss": 2.0099,
+      "step": 2102
+    },
+    {
+      "epoch": 3.88,
+      "learning_rate": 9.270400000000002e-06,
+      "loss": 2.041,
+      "step": 2103
+    },
+    {
+      "epoch": 3.88,
+      "learning_rate": 9.267200000000002e-06,
+      "loss": 1.9908,
+      "step": 2104
+    },
+    {
+      "epoch": 3.88,
+      "learning_rate": 9.264000000000001e-06,
+      "loss": 1.9513,
+      "step": 2105
+    },
+    {
+      "epoch": 3.88,
+      "learning_rate": 9.260800000000001e-06,
+      "loss": 1.996,
+      "step": 2106
+    },
+    {
+      "epoch": 3.89,
+      "learning_rate": 9.257600000000001e-06,
+      "loss": 2.196,
+      "step": 2107
+    },
+    {
+      "epoch": 3.89,
+      "learning_rate": 9.254400000000002e-06,
+      "loss": 1.9441,
+      "step": 2108
+    },
+    {
+      "epoch": 3.89,
+      "learning_rate": 9.251200000000002e-06,
+      "loss": 1.9529,
+      "step": 2109
+    },
+    {
+      "epoch": 3.89,
+      "learning_rate": 9.248e-06,
+      "loss": 2.1048,
+      "step": 2110
+    },
+    {
+      "epoch": 3.89,
+      "learning_rate": 9.244800000000002e-06,
+      "loss": 2.0327,
+      "step": 2111
+    },
+    {
+      "epoch": 3.9,
+      "learning_rate": 9.241600000000001e-06,
+      "loss": 1.9662,
+      "step": 2112
+    },
+    {
+      "epoch": 3.9,
+      "learning_rate": 9.238400000000001e-06,
+      "loss": 2.0379,
+      "step": 2113
+    },
+    {
+      "epoch": 3.9,
+      "learning_rate": 9.235200000000003e-06,
+      "loss": 1.9666,
+      "step": 2114
+    },
+    {
+      "epoch": 3.9,
+      "learning_rate": 9.232e-06,
+      "loss": 2.0036,
+      "step": 2115
+    },
+    {
+      "epoch": 3.9,
+      "learning_rate": 9.2288e-06,
+      "loss": 2.0032,
+      "step": 2116
+    },
+    {
+      "epoch": 3.91,
+      "learning_rate": 9.225600000000002e-06,
+      "loss": 1.8904,
+      "step": 2117
+    },
+    {
+      "epoch": 3.91,
+      "learning_rate": 9.222400000000002e-06,
+      "loss": 2.0743,
+      "step": 2118
+    },
+    {
+      "epoch": 3.91,
+      "learning_rate": 9.219200000000001e-06,
+      "loss": 2.1502,
+      "step": 2119
+    },
+    {
+      "epoch": 3.91,
+      "learning_rate": 9.216000000000001e-06,
+      "loss": 2.1354,
+      "step": 2120
+    },
+    {
+      "epoch": 3.91,
+      "learning_rate": 9.212800000000001e-06,
+      "loss": 1.9573,
+      "step": 2121
+    },
+    {
+      "epoch": 3.91,
+      "learning_rate": 9.2096e-06,
+      "loss": 1.9507,
+      "step": 2122
+    },
+    {
+      "epoch": 3.92,
+      "learning_rate": 9.206400000000002e-06,
+      "loss": 2.0346,
+      "step": 2123
+    },
+    {
+      "epoch": 3.92,
+      "learning_rate": 9.203200000000002e-06,
+      "loss": 1.9995,
+      "step": 2124
+    },
+    {
+      "epoch": 3.92,
+      "learning_rate": 9.2e-06,
+      "loss": 1.9842,
+      "step": 2125
+    },
+    {
+      "epoch": 3.92,
+      "learning_rate": 9.196800000000001e-06,
+      "loss": 2.0613,
+      "step": 2126
+    },
+    {
+      "epoch": 3.92,
+      "learning_rate": 9.193600000000001e-06,
+      "loss": 2.024,
+      "step": 2127
+    },
+    {
+      "epoch": 3.93,
+      "learning_rate": 9.190400000000003e-06,
+      "loss": 2.0027,
+      "step": 2128
+    },
+    {
+      "epoch": 3.93,
+      "learning_rate": 9.187200000000002e-06,
+      "loss": 1.9352,
+      "step": 2129
+    },
+    {
+      "epoch": 3.93,
+      "learning_rate": 9.184e-06,
+      "loss": 1.9924,
+      "step": 2130
+    },
+    {
+      "epoch": 3.93,
+      "eval_loss": 2.30216646194458,
+      "eval_runtime": 231.3321,
+      "eval_samples_per_second": 37.492,
+      "eval_steps_per_second": 0.588,
+      "step": 2130
+    },
+    {
+      "epoch": 3.93,
+      "learning_rate": 9.180800000000002e-06,
+      "loss": 1.9255,
+      "step": 2131
+    },
+    {
+      "epoch": 3.93,
+      "learning_rate": 9.177600000000002e-06,
+      "loss": 2.1242,
+      "step": 2132
+    },
+    {
+      "epoch": 3.93,
+      "learning_rate": 9.174400000000001e-06,
+      "loss": 1.8558,
+      "step": 2133
+    },
+    {
+      "epoch": 3.94,
+      "learning_rate": 9.171200000000003e-06,
+      "loss": 2.0522,
+      "step": 2134
+    },
+    {
+      "epoch": 3.94,
+      "learning_rate": 9.168000000000001e-06,
+      "loss": 2.0011,
+      "step": 2135
+    },
+    {
+      "epoch": 3.94,
+      "learning_rate": 9.1648e-06,
+      "loss": 2.0323,
+      "step": 2136
+    },
+    {
+      "epoch": 3.94,
+      "learning_rate": 9.161600000000002e-06,
+      "loss": 2.0911,
+      "step": 2137
+    },
+    {
+      "epoch": 3.94,
+      "learning_rate": 9.158400000000002e-06,
+      "loss": 1.9845,
+      "step": 2138
+    },
+    {
+      "epoch": 3.95,
+      "learning_rate": 9.155200000000002e-06,
+      "loss": 1.9319,
+      "step": 2139
+    },
+    {
+      "epoch": 3.95,
+      "learning_rate": 9.152000000000001e-06,
+      "loss": 1.9751,
+      "step": 2140
+    },
+    {
+      "epoch": 3.95,
+      "learning_rate": 9.148800000000001e-06,
+      "loss": 2.0537,
+      "step": 2141
+    },
+    {
+      "epoch": 3.95,
+      "learning_rate": 9.145600000000001e-06,
+      "loss": 1.987,
+      "step": 2142
+    },
+    {
+      "epoch": 3.95,
+      "learning_rate": 9.142400000000002e-06,
+      "loss": 2.1068,
+      "step": 2143
+    },
+    {
+      "epoch": 3.95,
+      "learning_rate": 9.139200000000002e-06,
+      "loss": 2.1309,
+      "step": 2144
+    },
+    {
+      "epoch": 3.96,
+      "learning_rate": 9.136e-06,
+      "loss": 2.1535,
+      "step": 2145
+    },
+    {
+      "epoch": 3.96,
+      "learning_rate": 9.132800000000002e-06,
+      "loss": 2.0569,
+      "step": 2146
+    },
+    {
+      "epoch": 3.96,
+      "learning_rate": 9.129600000000001e-06,
+      "loss": 2.0463,
+      "step": 2147
+    },
+    {
+      "epoch": 3.96,
+      "learning_rate": 9.126400000000001e-06,
+      "loss": 2.0805,
+      "step": 2148
+    },
+    {
+      "epoch": 3.96,
+      "learning_rate": 9.123200000000003e-06,
+      "loss": 2.0587,
+      "step": 2149
+    },
+    {
+      "epoch": 3.97,
+      "learning_rate": 9.12e-06,
+      "loss": 1.9781,
+      "step": 2150
+    },
+    {
+      "epoch": 3.97,
+      "learning_rate": 9.1168e-06,
+      "loss": 2.1186,
+      "step": 2151
+    },
+    {
+      "epoch": 3.97,
+      "learning_rate": 9.113600000000002e-06,
+      "loss": 2.0586,
+      "step": 2152
+    },
+    {
+      "epoch": 3.97,
+      "learning_rate": 9.110400000000002e-06,
+      "loss": 2.0175,
+      "step": 2153
+    },
+    {
+      "epoch": 3.97,
+      "learning_rate": 9.107200000000002e-06,
+      "loss": 2.1254,
+      "step": 2154
+    },
+    {
+      "epoch": 3.98,
+      "learning_rate": 9.104000000000001e-06,
+      "loss": 2.0402,
+      "step": 2155
+    },
+    {
+      "epoch": 3.98,
+      "learning_rate": 9.100800000000001e-06,
+      "loss": 2.0114,
+      "step": 2156
+    },
+    {
+      "epoch": 3.98,
+      "learning_rate": 9.0976e-06,
+      "loss": 2.0427,
+      "step": 2157
+    },
+    {
+      "epoch": 3.98,
+      "learning_rate": 9.094400000000002e-06,
+      "loss": 2.1336,
+      "step": 2158
+    },
+    {
+      "epoch": 3.98,
+      "learning_rate": 9.091200000000002e-06,
+      "loss": 2.0504,
+      "step": 2159
+    },
+    {
+      "epoch": 3.98,
+      "learning_rate": 9.088e-06,
+      "loss": 2.0192,
+      "step": 2160
+    },
+    {
+      "epoch": 3.98,
+      "eval_loss": 2.303264617919922,
+      "eval_runtime": 231.3266,
+      "eval_samples_per_second": 37.492,
+      "eval_steps_per_second": 0.588,
+      "step": 2160
+    },
+    {
+      "epoch": 3.99,
+      "learning_rate": 9.084800000000002e-06,
+      "loss": 2.0062,
+      "step": 2161
+    },
+    {
+      "epoch": 3.99,
+      "learning_rate": 9.081600000000001e-06,
+      "loss": 1.9704,
+      "step": 2162
+    },
+    {
+      "epoch": 3.99,
+      "learning_rate": 9.078400000000001e-06,
+      "loss": 2.0781,
+      "step": 2163
+    },
+    {
+      "epoch": 3.99,
+      "learning_rate": 9.075200000000002e-06,
+      "loss": 2.0219,
+      "step": 2164
+    },
+    {
+      "epoch": 3.99,
+      "learning_rate": 9.072e-06,
+      "loss": 2.0405,
+      "step": 2165
+    },
+    {
+      "epoch": 4.0,
+      "learning_rate": 9.0688e-06,
+      "loss": 2.0608,
+      "step": 2166
+    },
+    {
+      "epoch": 4.0,
+      "learning_rate": 9.065600000000002e-06,
+      "loss": 2.0081,
+      "step": 2167
+    },
+    {
+      "epoch": 4.0,
+      "learning_rate": 9.062400000000002e-06,
+      "loss": 2.0097,
+      "step": 2168
+    },
+    {
+      "epoch": 4.0,
+      "learning_rate": 9.059200000000001e-06,
+      "loss": 2.8819,
+      "step": 2169
+    },
+    {
+      "epoch": 4.0,
+      "learning_rate": 9.056000000000001e-06,
+      "loss": 2.0317,
+      "step": 2170
+    },
+    {
+      "epoch": 4.01,
+      "learning_rate": 9.0528e-06,
+      "loss": 1.7745,
+      "step": 2171
+    },
+    {
+      "epoch": 4.01,
+      "learning_rate": 9.0496e-06,
+      "loss": 1.9907,
+      "step": 2172
+    },
+    {
+      "epoch": 4.01,
+      "learning_rate": 9.046400000000002e-06,
+      "loss": 1.8727,
+      "step": 2173
+    },
+    {
+      "epoch": 4.01,
+      "learning_rate": 9.043200000000002e-06,
+      "loss": 1.9167,
+      "step": 2174
+    },
+    {
+      "epoch": 4.01,
+      "learning_rate": 9.04e-06,
+      "loss": 1.9502,
+      "step": 2175
+    },
+    {
+      "epoch": 4.01,
+      "learning_rate": 9.036800000000001e-06,
+      "loss": 1.8638,
+      "step": 2176
+    },
+    {
+      "epoch": 4.02,
+      "learning_rate": 9.033600000000001e-06,
+      "loss": 2.0148,
+      "step": 2177
+    },
+    {
+      "epoch": 4.02,
+      "learning_rate": 9.030400000000003e-06,
+      "loss": 1.7367,
+      "step": 2178
+    },
+    {
+      "epoch": 4.02,
+      "learning_rate": 9.027200000000002e-06,
+      "loss": 1.8874,
+      "step": 2179
+    },
+    {
+      "epoch": 4.02,
+      "learning_rate": 9.024e-06,
+      "loss": 1.8758,
+      "step": 2180
+    },
+    {
+      "epoch": 4.02,
+      "learning_rate": 9.020800000000002e-06,
+      "loss": 1.92,
+      "step": 2181
+    },
+    {
+      "epoch": 4.03,
+      "learning_rate": 9.017600000000002e-06,
+      "loss": 1.9892,
+      "step": 2182
+    },
+    {
+      "epoch": 4.03,
+      "learning_rate": 9.014400000000001e-06,
+      "loss": 1.8129,
+      "step": 2183
+    },
+    {
+      "epoch": 4.03,
+      "learning_rate": 9.011200000000003e-06,
+      "loss": 1.9367,
+      "step": 2184
+    },
+    {
+      "epoch": 4.03,
+      "learning_rate": 9.008e-06,
+      "loss": 1.871,
+      "step": 2185
+    },
+    {
+      "epoch": 4.03,
+      "learning_rate": 9.0048e-06,
+      "loss": 2.0168,
+      "step": 2186
+    },
+    {
+      "epoch": 4.04,
+      "learning_rate": 9.001600000000002e-06,
+      "loss": 1.9189,
+      "step": 2187
+    },
+    {
+      "epoch": 4.04,
+      "learning_rate": 8.998400000000002e-06,
+      "loss": 1.9572,
+      "step": 2188
+    },
+    {
+      "epoch": 4.04,
+      "learning_rate": 8.995200000000002e-06,
+      "loss": 1.9475,
+      "step": 2189
+    },
+    {
+      "epoch": 4.04,
+      "learning_rate": 8.992000000000003e-06,
+      "loss": 1.8397,
+      "step": 2190
+    },
+    {
+      "epoch": 4.04,
+      "eval_loss": 2.32212233543396,
+      "eval_runtime": 232.3862,
+      "eval_samples_per_second": 37.321,
+      "eval_steps_per_second": 0.585,
+      "step": 2190
+    },
+    {
+      "epoch": 4.04,
+      "learning_rate": 8.988800000000001e-06,
+      "loss": 1.9295,
+      "step": 2191
+    },
+    {
+      "epoch": 4.04,
+      "learning_rate": 8.9856e-06,
+      "loss": 1.9713,
+      "step": 2192
+    },
+    {
+      "epoch": 4.05,
+      "learning_rate": 8.982400000000002e-06,
+      "loss": 1.8761,
+      "step": 2193
+    },
+    {
+      "epoch": 4.05,
+      "learning_rate": 8.979200000000002e-06,
+      "loss": 1.8241,
+      "step": 2194
+    },
+    {
+      "epoch": 4.05,
+      "learning_rate": 8.976000000000002e-06,
+      "loss": 1.985,
+      "step": 2195
+    },
+    {
+      "epoch": 4.05,
+      "learning_rate": 8.972800000000002e-06,
+      "loss": 1.9131,
+      "step": 2196
+    },
+    {
+      "epoch": 4.05,
+      "learning_rate": 8.969600000000001e-06,
+      "loss": 1.9332,
+      "step": 2197
+    },
+    {
+      "epoch": 4.06,
+      "learning_rate": 8.966400000000001e-06,
+      "loss": 1.9253,
+      "step": 2198
+    },
+    {
+      "epoch": 4.06,
+      "learning_rate": 8.963200000000003e-06,
+      "loss": 1.8353,
+      "step": 2199
+    },
+    {
+      "epoch": 4.06,
+      "learning_rate": 8.960000000000002e-06,
+      "loss": 1.8929,
+      "step": 2200
+    },
+    {
+      "epoch": 4.06,
+      "learning_rate": 8.9568e-06,
+      "loss": 1.9084,
+      "step": 2201
+    },
+    {
+      "epoch": 4.06,
+      "learning_rate": 8.953600000000002e-06,
+      "loss": 2.0005,
+      "step": 2202
+    },
+    {
+      "epoch": 4.06,
+      "learning_rate": 8.950400000000002e-06,
+      "loss": 1.9175,
+      "step": 2203
+    },
+    {
+      "epoch": 4.07,
+      "learning_rate": 8.947200000000001e-06,
+      "loss": 1.921,
+      "step": 2204
+    },
+    {
+      "epoch": 4.07,
+      "learning_rate": 8.944000000000003e-06,
+      "loss": 1.957,
+      "step": 2205
+    },
+    {
+      "epoch": 4.07,
+      "learning_rate": 8.940800000000001e-06,
+      "loss": 1.7547,
+      "step": 2206
+    },
+    {
+      "epoch": 4.07,
+      "learning_rate": 8.9376e-06,
+      "loss": 1.9304,
+      "step": 2207
+    },
+    {
+      "epoch": 4.07,
+      "learning_rate": 8.934400000000002e-06,
+      "loss": 1.9347,
+      "step": 2208
+    },
+    {
+      "epoch": 4.08,
+      "learning_rate": 8.931200000000002e-06,
+      "loss": 2.0128,
+      "step": 2209
+    },
+    {
+      "epoch": 4.08,
+      "learning_rate": 8.928000000000002e-06,
+      "loss": 1.8628,
+      "step": 2210
+    },
+    {
+      "epoch": 4.08,
+      "learning_rate": 8.924800000000001e-06,
+      "loss": 1.9022,
+      "step": 2211
+    },
+    {
+      "epoch": 4.08,
+      "learning_rate": 8.921600000000001e-06,
+      "loss": 1.8613,
+      "step": 2212
+    },
+    {
+      "epoch": 4.08,
+      "learning_rate": 8.918400000000001e-06,
+      "loss": 1.7587,
+      "step": 2213
+    },
+    {
+      "epoch": 4.08,
+      "learning_rate": 8.915200000000002e-06,
+      "loss": 1.9873,
+      "step": 2214
+    },
+    {
+      "epoch": 4.09,
+      "learning_rate": 8.912000000000002e-06,
+      "loss": 1.8823,
+      "step": 2215
+    },
+    {
+      "epoch": 4.09,
+      "learning_rate": 8.9088e-06,
+      "loss": 1.9938,
+      "step": 2216
+    },
+    {
+      "epoch": 4.09,
+      "learning_rate": 8.905600000000002e-06,
+      "loss": 1.9524,
+      "step": 2217
+    },
+    {
+      "epoch": 4.09,
+      "learning_rate": 8.902400000000001e-06,
+      "loss": 1.95,
+      "step": 2218
+    },
+    {
+      "epoch": 4.09,
+      "learning_rate": 8.899200000000001e-06,
+      "loss": 1.9461,
+      "step": 2219
+    },
+    {
+      "epoch": 4.1,
+      "learning_rate": 8.896000000000003e-06,
+      "loss": 1.9863,
+      "step": 2220
+    },
+    {
+      "epoch": 4.1,
+      "eval_loss": 2.318183422088623,
+      "eval_runtime": 231.5665,
+      "eval_samples_per_second": 37.454,
+      "eval_steps_per_second": 0.587,
+      "step": 2220
+    },
+    {
+      "epoch": 4.1,
+      "learning_rate": 8.8928e-06,
+      "loss": 1.8885,
+      "step": 2221
+    },
+    {
+      "epoch": 4.1,
+      "learning_rate": 8.8896e-06,
+      "loss": 1.9055,
+      "step": 2222
+    },
+    {
+      "epoch": 4.1,
+      "learning_rate": 8.886400000000002e-06,
+      "loss": 1.9583,
+      "step": 2223
+    },
+    {
+      "epoch": 4.1,
+      "learning_rate": 8.883200000000002e-06,
+      "loss": 1.9552,
+      "step": 2224
+    },
+    {
+      "epoch": 4.11,
+      "learning_rate": 8.880000000000003e-06,
+      "loss": 1.999,
+      "step": 2225
+    },
+    {
+      "epoch": 4.11,
+      "learning_rate": 8.876800000000001e-06,
+      "loss": 1.8793,
+      "step": 2226
+    },
+    {
+      "epoch": 4.11,
+      "learning_rate": 8.873600000000001e-06,
+      "loss": 1.9279,
+      "step": 2227
+    },
+    {
+      "epoch": 4.11,
+      "learning_rate": 8.870400000000002e-06,
+      "loss": 1.9657,
+      "step": 2228
+    },
+    {
+      "epoch": 4.11,
+      "learning_rate": 8.867200000000002e-06,
+      "loss": 1.9833,
+      "step": 2229
+    },
+    {
+      "epoch": 4.11,
+      "learning_rate": 8.864000000000002e-06,
+      "loss": 1.9627,
+      "step": 2230
+    },
+    {
+      "epoch": 4.12,
+      "learning_rate": 8.860800000000002e-06,
+      "loss": 1.8369,
+      "step": 2231
+    },
+    {
+      "epoch": 4.12,
+      "learning_rate": 8.857600000000001e-06,
+      "loss": 1.9883,
+      "step": 2232
+    },
+    {
+      "epoch": 4.12,
+      "learning_rate": 8.854400000000001e-06,
+      "loss": 1.855,
+      "step": 2233
+    },
+    {
+      "epoch": 4.12,
+      "learning_rate": 8.851200000000003e-06,
+      "loss": 1.9154,
+      "step": 2234
+    },
+    {
+      "epoch": 4.12,
+      "learning_rate": 8.848000000000002e-06,
+      "loss": 1.7896,
+      "step": 2235
+    },
+    {
+      "epoch": 4.13,
+      "learning_rate": 8.8448e-06,
+      "loss": 1.931,
+      "step": 2236
+    },
+    {
+      "epoch": 4.13,
+      "learning_rate": 8.841600000000002e-06,
+      "loss": 1.8883,
+      "step": 2237
+    },
+    {
+      "epoch": 4.13,
+      "learning_rate": 8.838400000000002e-06,
+      "loss": 2.0366,
+      "step": 2238
+    },
+    {
+      "epoch": 4.13,
+      "learning_rate": 8.835200000000001e-06,
+      "loss": 1.9992,
+      "step": 2239
+    },
+    {
+      "epoch": 4.13,
+      "learning_rate": 8.832000000000003e-06,
+      "loss": 1.8383,
+      "step": 2240
+    },
+    {
+      "epoch": 4.13,
+      "learning_rate": 8.828800000000001e-06,
+      "loss": 1.9535,
+      "step": 2241
+    },
+    {
+      "epoch": 4.14,
+      "learning_rate": 8.8256e-06,
+      "loss": 1.9873,
+      "step": 2242
+    },
+    {
+      "epoch": 4.14,
+      "learning_rate": 8.822400000000002e-06,
+      "loss": 1.93,
+      "step": 2243
+    },
+    {
+      "epoch": 4.14,
+      "learning_rate": 8.819200000000002e-06,
+      "loss": 1.8848,
+      "step": 2244
+    },
+    {
+      "epoch": 4.14,
+      "learning_rate": 8.816000000000002e-06,
+      "loss": 1.9784,
+      "step": 2245
+    },
+    {
+      "epoch": 4.14,
+      "learning_rate": 8.812800000000001e-06,
+      "loss": 1.9654,
+      "step": 2246
+    },
+    {
+      "epoch": 4.15,
+      "learning_rate": 8.809600000000001e-06,
+      "loss": 1.8789,
+      "step": 2247
+    },
+    {
+      "epoch": 4.15,
+      "learning_rate": 8.806400000000001e-06,
+      "loss": 1.8982,
+      "step": 2248
+    },
+    {
+      "epoch": 4.15,
+      "learning_rate": 8.803200000000002e-06,
+      "loss": 1.844,
+      "step": 2249
+    },
+    {
+      "epoch": 4.15,
+      "learning_rate": 8.800000000000002e-06,
+      "loss": 1.953,
+      "step": 2250
+    },
+    {
+      "epoch": 4.15,
+      "eval_loss": 2.3303236961364746,
+      "eval_runtime": 233.1496,
+      "eval_samples_per_second": 37.199,
+      "eval_steps_per_second": 0.583,
+      "step": 2250
+    },
+    {
+      "epoch": 4.15,
+      "learning_rate": 8.7968e-06,
+      "loss": 1.849,
+      "step": 2251
+    },
+    {
+      "epoch": 4.15,
+      "learning_rate": 8.793600000000002e-06,
+      "loss": 1.9031,
+      "step": 2252
+    },
+    {
+      "epoch": 4.16,
+      "learning_rate": 8.790400000000001e-06,
+      "loss": 1.8023,
+      "step": 2253
+    },
+    {
+      "epoch": 4.16,
+      "learning_rate": 8.787200000000001e-06,
+      "loss": 1.9755,
+      "step": 2254
+    },
+    {
+      "epoch": 4.16,
+      "learning_rate": 8.784000000000003e-06,
+      "loss": 1.9287,
+      "step": 2255
+    },
+    {
+      "epoch": 4.16,
+      "learning_rate": 8.7808e-06,
+      "loss": 1.9073,
+      "step": 2256
+    },
+    {
+      "epoch": 4.16,
+      "learning_rate": 8.7776e-06,
+      "loss": 1.8806,
+      "step": 2257
+    },
+    {
+      "epoch": 4.17,
+      "learning_rate": 8.774400000000002e-06,
+      "loss": 1.8624,
+      "step": 2258
+    },
+    {
+      "epoch": 4.17,
+      "learning_rate": 8.771200000000002e-06,
+      "loss": 2.0144,
+      "step": 2259
+    },
+    {
+      "epoch": 4.17,
+      "learning_rate": 8.768000000000001e-06,
+      "loss": 1.9528,
+      "step": 2260
+    },
+    {
+      "epoch": 4.17,
+      "learning_rate": 8.764800000000001e-06,
+      "loss": 1.902,
+      "step": 2261
+    },
+    {
+      "epoch": 4.17,
+      "learning_rate": 8.761600000000001e-06,
+      "loss": 1.8326,
+      "step": 2262
+    },
+    {
+      "epoch": 4.18,
+      "learning_rate": 8.7584e-06,
+      "loss": 1.9848,
+      "step": 2263
+    },
+    {
+      "epoch": 4.18,
+      "learning_rate": 8.755200000000002e-06,
+      "loss": 1.991,
+      "step": 2264
+    },
+    {
+      "epoch": 4.18,
+      "learning_rate": 8.752000000000002e-06,
+      "loss": 1.9484,
+      "step": 2265
+    },
+    {
+      "epoch": 4.18,
+      "learning_rate": 8.7488e-06,
+      "loss": 1.9197,
+      "step": 2266
+    },
+    {
+      "epoch": 4.18,
+      "learning_rate": 8.745600000000001e-06,
+      "loss": 1.8264,
+      "step": 2267
+    },
+    {
+      "epoch": 4.18,
+      "learning_rate": 8.742400000000001e-06,
+      "loss": 1.9295,
+      "step": 2268
+    },
+    {
+      "epoch": 4.19,
+      "learning_rate": 8.739200000000001e-06,
+      "loss": 1.9701,
+      "step": 2269
+    },
+    {
+      "epoch": 4.19,
+      "learning_rate": 8.736000000000002e-06,
+      "loss": 1.9229,
+      "step": 2270
+    },
+    {
+      "epoch": 4.19,
+      "learning_rate": 8.7328e-06,
+      "loss": 1.9246,
+      "step": 2271
+    },
+    {
+      "epoch": 4.19,
+      "learning_rate": 8.7296e-06,
+      "loss": 1.8282,
+      "step": 2272
+    },
+    {
+      "epoch": 4.19,
+      "learning_rate": 8.726400000000002e-06,
+      "loss": 1.9954,
+      "step": 2273
+    },
+    {
+      "epoch": 4.2,
+      "learning_rate": 8.723200000000001e-06,
+      "loss": 1.8409,
+      "step": 2274
+    },
+    {
+      "epoch": 4.2,
+      "learning_rate": 8.720000000000003e-06,
+      "loss": 1.9021,
+      "step": 2275
+    },
+    {
+      "epoch": 4.2,
+      "learning_rate": 8.716800000000001e-06,
+      "loss": 1.8526,
+      "step": 2276
+    },
+    {
+      "epoch": 4.2,
+      "learning_rate": 8.7136e-06,
+      "loss": 1.9284,
+      "step": 2277
+    },
+    {
+      "epoch": 4.2,
+      "learning_rate": 8.710400000000002e-06,
+      "loss": 1.8828,
+      "step": 2278
+    },
+    {
+      "epoch": 4.2,
+      "learning_rate": 8.707200000000002e-06,
+      "loss": 1.9896,
+      "step": 2279
+    },
+    {
+      "epoch": 4.21,
+      "learning_rate": 8.704000000000002e-06,
+      "loss": 1.9576,
+      "step": 2280
+    },
+    {
+      "epoch": 4.21,
+      "eval_loss": 2.3367435932159424,
+      "eval_runtime": 233.2947,
+      "eval_samples_per_second": 37.176,
+      "eval_steps_per_second": 0.583,
+      "step": 2280
+    },
+    {
+      "epoch": 4.21,
+      "learning_rate": 8.700800000000001e-06,
+      "loss": 1.8308,
+      "step": 2281
+    },
+    {
+      "epoch": 4.21,
+      "learning_rate": 8.697600000000001e-06,
+      "loss": 1.9222,
+      "step": 2282
+    },
+    {
+      "epoch": 4.21,
+      "learning_rate": 8.694400000000001e-06,
+      "loss": 1.8744,
+      "step": 2283
+    },
+    {
+      "epoch": 4.21,
+      "learning_rate": 8.691200000000002e-06,
+      "loss": 1.8474,
+      "step": 2284
+    },
+    {
+      "epoch": 4.22,
+      "learning_rate": 8.688000000000002e-06,
+      "loss": 1.8551,
+      "step": 2285
+    },
+    {
+      "epoch": 4.22,
+      "learning_rate": 8.6848e-06,
+      "loss": 2.0191,
+      "step": 2286
+    },
+    {
+      "epoch": 4.22,
+      "learning_rate": 8.681600000000002e-06,
+      "loss": 1.8512,
+      "step": 2287
+    },
+    {
+      "epoch": 4.22,
+      "learning_rate": 8.678400000000001e-06,
+      "loss": 1.9083,
+      "step": 2288
+    },
+    {
+      "epoch": 4.22,
+      "learning_rate": 8.675200000000001e-06,
+      "loss": 1.8886,
+      "step": 2289
+    },
+    {
+      "epoch": 4.22,
+      "learning_rate": 8.672000000000003e-06,
+      "loss": 1.9638,
+      "step": 2290
+    },
+    {
+      "epoch": 4.23,
+      "learning_rate": 8.6688e-06,
+      "loss": 1.8377,
+      "step": 2291
+    },
+    {
+      "epoch": 4.23,
+      "learning_rate": 8.6656e-06,
+      "loss": 1.783,
+      "step": 2292
+    },
+    {
+      "epoch": 4.23,
+      "learning_rate": 8.662400000000002e-06,
+      "loss": 2.0039,
+      "step": 2293
+    },
+    {
+      "epoch": 4.23,
+      "learning_rate": 8.659200000000002e-06,
+      "loss": 1.966,
+      "step": 2294
+    },
+    {
+      "epoch": 4.23,
+      "learning_rate": 8.656000000000001e-06,
+      "loss": 1.9266,
+      "step": 2295
+    },
+    {
+      "epoch": 4.24,
+      "learning_rate": 8.652800000000001e-06,
+      "loss": 1.843,
+      "step": 2296
+    },
+    {
+      "epoch": 4.24,
+      "learning_rate": 8.649600000000001e-06,
+      "loss": 1.9155,
+      "step": 2297
+    },
+    {
+      "epoch": 4.24,
+      "learning_rate": 8.6464e-06,
+      "loss": 1.936,
+      "step": 2298
+    },
+    {
+      "epoch": 4.24,
+      "learning_rate": 8.643200000000002e-06,
+      "loss": 1.9508,
+      "step": 2299
+    },
+    {
+      "epoch": 4.24,
+      "learning_rate": 8.640000000000002e-06,
+      "loss": 2.0271,
+      "step": 2300
+    },
+    {
+      "epoch": 4.25,
+      "learning_rate": 8.6368e-06,
+      "loss": 1.9811,
+      "step": 2301
+    },
+    {
+      "epoch": 4.25,
+      "learning_rate": 8.633600000000001e-06,
+      "loss": 1.9356,
+      "step": 2302
+    },
+    {
+      "epoch": 4.25,
+      "learning_rate": 8.630400000000001e-06,
+      "loss": 2.0103,
+      "step": 2303
+    },
+    {
+      "epoch": 4.25,
+      "learning_rate": 8.627200000000001e-06,
+      "loss": 1.9568,
+      "step": 2304
+    },
+    {
+      "epoch": 4.25,
+      "learning_rate": 8.624000000000002e-06,
+      "loss": 1.9264,
+      "step": 2305
+    },
+    {
+      "epoch": 4.25,
+      "learning_rate": 8.6208e-06,
+      "loss": 1.9387,
+      "step": 2306
+    },
+    {
+      "epoch": 4.26,
+      "learning_rate": 8.6176e-06,
+      "loss": 1.9934,
+      "step": 2307
+    },
+    {
+      "epoch": 4.26,
+      "learning_rate": 8.614400000000002e-06,
+      "loss": 1.8476,
+      "step": 2308
+    },
+    {
+      "epoch": 4.26,
+      "learning_rate": 8.611200000000002e-06,
+      "loss": 1.8893,
+      "step": 2309
+    },
+    {
+      "epoch": 4.26,
+      "learning_rate": 8.608000000000001e-06,
+      "loss": 1.8693,
+      "step": 2310
+    },
+    {
+      "epoch": 4.26,
+      "eval_loss": 2.33791184425354,
+      "eval_runtime": 232.438,
+      "eval_samples_per_second": 37.313,
+      "eval_steps_per_second": 0.585,
+      "step": 2310
+    },
+    {
+      "epoch": 4.26,
+      "learning_rate": 8.604800000000001e-06,
+      "loss": 1.9703,
+      "step": 2311
+    },
+    {
+      "epoch": 4.27,
+      "learning_rate": 8.6016e-06,
+      "loss": 1.7882,
+      "step": 2312
+    },
+    {
+      "epoch": 4.27,
+      "learning_rate": 8.5984e-06,
+      "loss": 1.829,
+      "step": 2313
+    },
+    {
+      "epoch": 4.27,
+      "learning_rate": 8.595200000000002e-06,
+      "loss": 1.9951,
+      "step": 2314
+    },
+    {
+      "epoch": 4.27,
+      "learning_rate": 8.592000000000002e-06,
+      "loss": 1.9524,
+      "step": 2315
+    },
+    {
+      "epoch": 4.27,
+      "learning_rate": 8.588800000000002e-06,
+      "loss": 1.9723,
+      "step": 2316
+    },
+    {
+      "epoch": 4.27,
+      "learning_rate": 8.585600000000001e-06,
+      "loss": 1.9423,
+      "step": 2317
+    },
+    {
+      "epoch": 4.28,
+      "learning_rate": 8.582400000000001e-06,
+      "loss": 1.9169,
+      "step": 2318
+    },
+    {
+      "epoch": 4.28,
+      "learning_rate": 8.5792e-06,
+      "loss": 1.9064,
+      "step": 2319
+    },
+    {
+      "epoch": 4.28,
+      "learning_rate": 8.576000000000002e-06,
+      "loss": 2.0213,
+      "step": 2320
+    },
+    {
+      "epoch": 4.28,
+      "learning_rate": 8.572800000000002e-06,
+      "loss": 1.8791,
+      "step": 2321
+    },
+    {
+      "epoch": 4.28,
+      "learning_rate": 8.5696e-06,
+      "loss": 1.8607,
+      "step": 2322
+    },
+    {
+      "epoch": 4.29,
+      "learning_rate": 8.566400000000002e-06,
+      "loss": 1.8168,
+      "step": 2323
+    },
+    {
+      "epoch": 4.29,
+      "learning_rate": 8.563200000000001e-06,
+      "loss": 1.8464,
+      "step": 2324
+    },
+    {
+      "epoch": 4.29,
+      "learning_rate": 8.560000000000003e-06,
+      "loss": 1.9268,
+      "step": 2325
+    },
+    {
+      "epoch": 4.29,
+      "learning_rate": 8.556800000000003e-06,
+      "loss": 1.9425,
+      "step": 2326
+    },
+    {
+      "epoch": 4.29,
+      "learning_rate": 8.5536e-06,
+      "loss": 1.9002,
+      "step": 2327
+    },
+    {
+      "epoch": 4.29,
+      "learning_rate": 8.550400000000002e-06,
+      "loss": 1.8559,
+      "step": 2328
+    },
+    {
+      "epoch": 4.3,
+      "learning_rate": 8.547200000000002e-06,
+      "loss": 1.8366,
+      "step": 2329
+    },
+    {
+      "epoch": 4.3,
+      "learning_rate": 8.544000000000002e-06,
+      "loss": 1.9374,
+      "step": 2330
+    },
+    {
+      "epoch": 4.3,
+      "learning_rate": 8.540800000000003e-06,
+      "loss": 1.905,
+      "step": 2331
+    },
+    {
+      "epoch": 4.3,
+      "learning_rate": 8.537600000000001e-06,
+      "loss": 1.9251,
+      "step": 2332
+    },
+    {
+      "epoch": 4.3,
+      "learning_rate": 8.5344e-06,
+      "loss": 1.9702,
+      "step": 2333
+    },
+    {
+      "epoch": 4.31,
+      "learning_rate": 8.531200000000002e-06,
+      "loss": 1.8537,
+      "step": 2334
+    },
+    {
+      "epoch": 4.31,
+      "learning_rate": 8.528000000000002e-06,
+      "loss": 1.8827,
+      "step": 2335
+    },
+    {
+      "epoch": 4.31,
+      "learning_rate": 8.524800000000002e-06,
+      "loss": 2.0177,
+      "step": 2336
+    },
+    {
+      "epoch": 4.31,
+      "learning_rate": 8.521600000000002e-06,
+      "loss": 1.9354,
+      "step": 2337
+    },
+    {
+      "epoch": 4.31,
+      "learning_rate": 8.518400000000001e-06,
+      "loss": 1.9381,
+      "step": 2338
+    },
+    {
+      "epoch": 4.32,
+      "learning_rate": 8.515200000000001e-06,
+      "loss": 1.982,
+      "step": 2339
+    },
+    {
+      "epoch": 4.32,
+      "learning_rate": 8.512000000000003e-06,
+      "loss": 1.9253,
+      "step": 2340
+    },
+    {
+      "epoch": 4.32,
+      "eval_loss": 2.3310000896453857,
+      "eval_runtime": 233.0168,
+      "eval_samples_per_second": 37.22,
+      "eval_steps_per_second": 0.584,
+      "step": 2340
+    },
+    {
+      "epoch": 4.32,
+      "learning_rate": 8.508800000000002e-06,
+      "loss": 1.9586,
+      "step": 2341
+    },
+    {
+      "epoch": 4.32,
+      "learning_rate": 8.5056e-06,
+      "loss": 1.9322,
+      "step": 2342
+    },
+    {
+      "epoch": 4.32,
+      "learning_rate": 8.502400000000002e-06,
+      "loss": 1.806,
+      "step": 2343
+    },
+    {
+      "epoch": 4.32,
+      "learning_rate": 8.499200000000002e-06,
+      "loss": 1.8194,
+      "step": 2344
+    },
+    {
+      "epoch": 4.33,
+      "learning_rate": 8.496000000000001e-06,
+      "loss": 1.9236,
+      "step": 2345
+    },
+    {
+      "epoch": 4.33,
+      "learning_rate": 8.492800000000003e-06,
+      "loss": 1.975,
+      "step": 2346
+    },
+    {
+      "epoch": 4.33,
+      "learning_rate": 8.4896e-06,
+      "loss": 1.9255,
+      "step": 2347
+    },
+    {
+      "epoch": 4.33,
+      "learning_rate": 8.4864e-06,
+      "loss": 2.0614,
+      "step": 2348
+    },
+    {
+      "epoch": 4.33,
+      "learning_rate": 8.483200000000002e-06,
+      "loss": 1.9793,
+      "step": 2349
+    },
+    {
+      "epoch": 4.34,
+      "learning_rate": 8.480000000000002e-06,
+      "loss": 1.9878,
+      "step": 2350
+    },
+    {
+      "epoch": 4.34,
+      "learning_rate": 8.476800000000002e-06,
+      "loss": 1.9286,
+      "step": 2351
+    },
+    {
+      "epoch": 4.34,
+      "learning_rate": 8.473600000000001e-06,
+      "loss": 1.9,
+      "step": 2352
+    },
+    {
+      "epoch": 4.34,
+      "learning_rate": 8.470400000000001e-06,
+      "loss": 1.8288,
+      "step": 2353
+    },
+    {
+      "epoch": 4.34,
+      "learning_rate": 8.467200000000001e-06,
+      "loss": 1.8523,
+      "step": 2354
+    },
+    {
+      "epoch": 4.34,
+      "learning_rate": 8.464000000000002e-06,
+      "loss": 1.9871,
+      "step": 2355
+    },
+    {
+      "epoch": 4.35,
+      "learning_rate": 8.460800000000002e-06,
+      "loss": 2.0827,
+      "step": 2356
+    },
+    {
+      "epoch": 4.35,
+      "learning_rate": 8.4576e-06,
+      "loss": 1.8929,
+      "step": 2357
+    },
+    {
+      "epoch": 4.35,
+      "learning_rate": 8.454400000000002e-06,
+      "loss": 1.94,
+      "step": 2358
+    },
+    {
+      "epoch": 4.35,
+      "learning_rate": 8.451200000000001e-06,
+      "loss": 1.8771,
+      "step": 2359
+    },
+    {
+      "epoch": 4.35,
+      "learning_rate": 8.448000000000001e-06,
+      "loss": 1.8945,
+      "step": 2360
+    },
+    {
+      "epoch": 4.36,
+      "learning_rate": 8.444800000000003e-06,
+      "loss": 1.9332,
+      "step": 2361
+    },
+    {
+      "epoch": 4.36,
+      "learning_rate": 8.4416e-06,
+      "loss": 1.9066,
+      "step": 2362
+    },
+    {
+      "epoch": 4.36,
+      "learning_rate": 8.4384e-06,
+      "loss": 1.9741,
+      "step": 2363
+    },
+    {
+      "epoch": 4.36,
+      "learning_rate": 8.435200000000002e-06,
+      "loss": 2.0612,
+      "step": 2364
+    },
+    {
+      "epoch": 4.36,
+      "learning_rate": 8.432000000000002e-06,
+      "loss": 1.9377,
+      "step": 2365
+    },
+    {
+      "epoch": 4.36,
+      "learning_rate": 8.428800000000001e-06,
+      "loss": 1.9397,
+      "step": 2366
+    },
+    {
+      "epoch": 4.37,
+      "learning_rate": 8.425600000000001e-06,
+      "loss": 1.7891,
+      "step": 2367
+    },
+    {
+      "epoch": 4.37,
+      "learning_rate": 8.422400000000001e-06,
+      "loss": 1.9062,
+      "step": 2368
+    },
+    {
+      "epoch": 4.37,
+      "learning_rate": 8.4192e-06,
+      "loss": 1.896,
+      "step": 2369
+    },
+    {
+      "epoch": 4.37,
+      "learning_rate": 8.416000000000002e-06,
+      "loss": 1.9232,
+      "step": 2370
+    },
+    {
+      "epoch": 4.37,
+      "eval_loss": 2.3258056640625,
+      "eval_runtime": 232.7924,
+      "eval_samples_per_second": 37.256,
+      "eval_steps_per_second": 0.584,
+      "step": 2370
+    },
+    {
+      "epoch": 4.37,
+      "learning_rate": 8.412800000000002e-06,
+      "loss": 1.9433,
+      "step": 2371
+    },
+    {
+      "epoch": 4.38,
+      "learning_rate": 8.4096e-06,
+      "loss": 1.8917,
+      "step": 2372
+    },
+    {
+      "epoch": 4.38,
+      "learning_rate": 8.406400000000001e-06,
+      "loss": 1.9106,
+      "step": 2373
+    },
+    {
+      "epoch": 4.38,
+      "learning_rate": 8.403200000000001e-06,
+      "loss": 1.9128,
+      "step": 2374
+    },
+    {
+      "epoch": 4.38,
+      "learning_rate": 8.400000000000003e-06,
+      "loss": 1.9119,
+      "step": 2375
+    },
+    {
+      "epoch": 4.38,
+      "learning_rate": 8.396800000000002e-06,
+      "loss": 1.8807,
+      "step": 2376
+    },
+    {
+      "epoch": 4.39,
+      "learning_rate": 8.3936e-06,
+      "loss": 2.0262,
+      "step": 2377
+    },
+    {
+      "epoch": 4.39,
+      "learning_rate": 8.390400000000002e-06,
+      "loss": 1.9435,
+      "step": 2378
+    },
+    {
+      "epoch": 4.39,
+      "learning_rate": 8.387200000000002e-06,
+      "loss": 1.9769,
+      "step": 2379
+    },
+    {
+      "epoch": 4.39,
+      "learning_rate": 8.384000000000001e-06,
+      "loss": 1.8231,
+      "step": 2380
+    },
+    {
+      "epoch": 4.39,
+      "learning_rate": 8.380800000000003e-06,
+      "loss": 1.7439,
+      "step": 2381
+    },
+    {
+      "epoch": 4.39,
+      "learning_rate": 8.377600000000001e-06,
+      "loss": 1.9319,
+      "step": 2382
+    },
+    {
+      "epoch": 4.4,
+      "learning_rate": 8.3744e-06,
+      "loss": 1.9946,
+      "step": 2383
+    },
+    {
+      "epoch": 4.4,
+      "learning_rate": 8.371200000000002e-06,
+      "loss": 1.9596,
+      "step": 2384
+    },
+    {
+      "epoch": 4.4,
+      "learning_rate": 8.368000000000002e-06,
+      "loss": 2.0001,
+      "step": 2385
+    },
+    {
+      "epoch": 4.4,
+      "learning_rate": 8.364800000000002e-06,
+      "loss": 1.9111,
+      "step": 2386
+    },
+    {
+      "epoch": 4.4,
+      "learning_rate": 8.361600000000001e-06,
+      "loss": 1.8323,
+      "step": 2387
+    },
+    {
+      "epoch": 4.41,
+      "learning_rate": 8.358400000000001e-06,
+      "loss": 1.8428,
+      "step": 2388
+    },
+    {
+      "epoch": 4.41,
+      "learning_rate": 8.355200000000001e-06,
+      "loss": 1.9893,
+      "step": 2389
+    },
+    {
+      "epoch": 4.41,
+      "learning_rate": 8.352000000000002e-06,
+      "loss": 1.8122,
+      "step": 2390
+    },
+    {
+      "epoch": 4.41,
+      "learning_rate": 8.348800000000002e-06,
+      "loss": 1.8553,
+      "step": 2391
+    },
+    {
+      "epoch": 4.41,
+      "learning_rate": 8.3456e-06,
+      "loss": 1.8729,
+      "step": 2392
+    },
+    {
+      "epoch": 4.41,
+      "learning_rate": 8.342400000000002e-06,
+      "loss": 1.8413,
+      "step": 2393
+    },
+    {
+      "epoch": 4.42,
+      "learning_rate": 8.339200000000001e-06,
+      "loss": 1.9761,
+      "step": 2394
+    },
+    {
+      "epoch": 4.42,
+      "learning_rate": 8.336000000000001e-06,
+      "loss": 1.8736,
+      "step": 2395
+    },
+    {
+      "epoch": 4.42,
+      "learning_rate": 8.332800000000003e-06,
+      "loss": 1.8899,
+      "step": 2396
+    },
+    {
+      "epoch": 4.42,
+      "learning_rate": 8.3296e-06,
+      "loss": 1.8987,
+      "step": 2397
+    },
+    {
+      "epoch": 4.42,
+      "learning_rate": 8.3264e-06,
+      "loss": 1.9532,
+      "step": 2398
+    },
+    {
+      "epoch": 4.43,
+      "learning_rate": 8.323200000000002e-06,
+      "loss": 1.9233,
+      "step": 2399
+    },
+    {
+      "epoch": 4.43,
+      "learning_rate": 8.320000000000002e-06,
+      "loss": 1.976,
+      "step": 2400
+    },
+    {
+      "epoch": 4.43,
+      "eval_loss": 2.333469867706299,
+      "eval_runtime": 233.3194,
+      "eval_samples_per_second": 37.172,
+      "eval_steps_per_second": 0.583,
+      "step": 2400
+    },
+    {
+      "epoch": 4.43,
+      "learning_rate": 8.316800000000001e-06,
+      "loss": 2.0693,
+      "step": 2401
+    },
+    {
+      "epoch": 4.43,
+      "learning_rate": 8.313600000000001e-06,
+      "loss": 1.9067,
+      "step": 2402
+    },
+    {
+      "epoch": 4.43,
+      "learning_rate": 8.310400000000001e-06,
+      "loss": 1.8823,
+      "step": 2403
+    },
+    {
+      "epoch": 4.44,
+      "learning_rate": 8.3072e-06,
+      "loss": 1.9287,
+      "step": 2404
+    },
+    {
+      "epoch": 4.44,
+      "learning_rate": 8.304000000000002e-06,
+      "loss": 1.8302,
+      "step": 2405
+    },
+    {
+      "epoch": 4.44,
+      "learning_rate": 8.300800000000002e-06,
+      "loss": 2.0358,
+      "step": 2406
+    },
+    {
+      "epoch": 4.44,
+      "learning_rate": 8.2976e-06,
+      "loss": 1.9911,
+      "step": 2407
+    },
+    {
+      "epoch": 4.44,
+      "learning_rate": 8.294400000000001e-06,
+      "loss": 1.9111,
+      "step": 2408
+    },
+    {
+      "epoch": 4.44,
+      "learning_rate": 8.291200000000001e-06,
+      "loss": 1.9833,
+      "step": 2409
+    },
+    {
+      "epoch": 4.45,
+      "learning_rate": 8.288000000000001e-06,
+      "loss": 1.8335,
+      "step": 2410
+    },
+    {
+      "epoch": 4.45,
+      "learning_rate": 8.284800000000002e-06,
+      "loss": 1.9289,
+      "step": 2411
+    },
+    {
+      "epoch": 4.45,
+      "learning_rate": 8.2816e-06,
+      "loss": 1.9094,
+      "step": 2412
+    },
+    {
+      "epoch": 4.45,
+      "learning_rate": 8.2784e-06,
+      "loss": 1.8861,
+      "step": 2413
+    },
+    {
+      "epoch": 4.45,
+      "learning_rate": 8.275200000000002e-06,
+      "loss": 1.8781,
+      "step": 2414
+    },
+    {
+      "epoch": 4.46,
+      "learning_rate": 8.272000000000001e-06,
+      "loss": 1.8454,
+      "step": 2415
+    },
+    {
+      "epoch": 4.46,
+      "learning_rate": 8.268800000000001e-06,
+      "loss": 1.9469,
+      "step": 2416
+    },
+    {
+      "epoch": 4.46,
+      "learning_rate": 8.265600000000001e-06,
+      "loss": 1.8814,
+      "step": 2417
+    },
+    {
+      "epoch": 4.46,
+      "learning_rate": 8.2624e-06,
+      "loss": 2.0146,
+      "step": 2418
+    },
+    {
+      "epoch": 4.46,
+      "learning_rate": 8.2592e-06,
+      "loss": 1.9996,
+      "step": 2419
+    },
+    {
+      "epoch": 4.46,
+      "learning_rate": 8.256000000000002e-06,
+      "loss": 1.8804,
+      "step": 2420
+    },
+    {
+      "epoch": 4.47,
+      "learning_rate": 8.252800000000002e-06,
+      "loss": 1.9191,
+      "step": 2421
+    },
+    {
+      "epoch": 4.47,
+      "learning_rate": 8.2496e-06,
+      "loss": 2.021,
+      "step": 2422
+    },
+    {
+      "epoch": 4.47,
+      "learning_rate": 8.246400000000001e-06,
+      "loss": 1.9592,
+      "step": 2423
+    },
+    {
+      "epoch": 4.47,
+      "learning_rate": 8.243200000000001e-06,
+      "loss": 1.9281,
+      "step": 2424
+    },
+    {
+      "epoch": 4.47,
+      "learning_rate": 8.240000000000002e-06,
+      "loss": 1.9929,
+      "step": 2425
+    },
+    {
+      "epoch": 4.48,
+      "learning_rate": 8.236800000000002e-06,
+      "loss": 2.0001,
+      "step": 2426
+    },
+    {
+      "epoch": 4.48,
+      "learning_rate": 8.2336e-06,
+      "loss": 1.9867,
+      "step": 2427
+    },
+    {
+      "epoch": 4.48,
+      "learning_rate": 8.230400000000002e-06,
+      "loss": 2.065,
+      "step": 2428
+    },
+    {
+      "epoch": 4.48,
+      "learning_rate": 8.227200000000001e-06,
+      "loss": 1.7976,
+      "step": 2429
+    },
+    {
+      "epoch": 4.48,
+      "learning_rate": 8.224000000000001e-06,
+      "loss": 1.9256,
+      "step": 2430
+    },
+    {
+      "epoch": 4.48,
+      "eval_loss": 2.3247835636138916,
+      "eval_runtime": 232.402,
+      "eval_samples_per_second": 37.319,
+      "eval_steps_per_second": 0.585,
+      "step": 2430
+    },
+    {
+      "epoch": 4.48,
+      "learning_rate": 8.220800000000003e-06,
+      "loss": 1.9254,
+      "step": 2431
+    },
+    {
+      "epoch": 4.49,
+      "learning_rate": 8.2176e-06,
+      "loss": 1.919,
+      "step": 2432
+    },
+    {
+      "epoch": 4.49,
+      "learning_rate": 8.2144e-06,
+      "loss": 2.0171,
+      "step": 2433
+    },
+    {
+      "epoch": 4.49,
+      "learning_rate": 8.211200000000002e-06,
+      "loss": 1.9811,
+      "step": 2434
+    },
+    {
+      "epoch": 4.49,
+      "learning_rate": 8.208000000000002e-06,
+      "loss": 1.8611,
+      "step": 2435
+    },
+    {
+      "epoch": 4.49,
+      "learning_rate": 8.204800000000001e-06,
+      "loss": 1.9125,
+      "step": 2436
+    },
+    {
+      "epoch": 4.5,
+      "learning_rate": 8.201600000000001e-06,
+      "loss": 1.9969,
+      "step": 2437
+    },
+    {
+      "epoch": 4.5,
+      "learning_rate": 8.198400000000001e-06,
+      "loss": 1.8954,
+      "step": 2438
+    },
+    {
+      "epoch": 4.5,
+      "learning_rate": 8.1952e-06,
+      "loss": 1.8936,
+      "step": 2439
+    },
+    {
+      "epoch": 4.5,
+      "learning_rate": 8.192000000000002e-06,
+      "loss": 1.94,
+      "step": 2440
+    },
+    {
+      "epoch": 4.5,
+      "learning_rate": 8.188800000000002e-06,
+      "loss": 1.9099,
+      "step": 2441
+    },
+    {
+      "epoch": 4.51,
+      "learning_rate": 8.185600000000002e-06,
+      "loss": 1.8158,
+      "step": 2442
+    },
+    {
+      "epoch": 4.51,
+      "learning_rate": 8.182400000000001e-06,
+      "loss": 2.0431,
+      "step": 2443
+    },
+    {
+      "epoch": 4.51,
+      "learning_rate": 8.179200000000001e-06,
+      "loss": 1.9196,
+      "step": 2444
+    },
+    {
+      "epoch": 4.51,
+      "learning_rate": 8.176000000000001e-06,
+      "loss": 2.0142,
+      "step": 2445
+    },
+    {
+      "epoch": 4.51,
+      "learning_rate": 8.172800000000002e-06,
+      "loss": 1.847,
+      "step": 2446
+    },
+    {
+      "epoch": 4.51,
+      "learning_rate": 8.169600000000002e-06,
+      "loss": 1.9281,
+      "step": 2447
+    },
+    {
+      "epoch": 4.52,
+      "learning_rate": 8.1664e-06,
+      "loss": 1.9668,
+      "step": 2448
+    },
+    {
+      "epoch": 4.52,
+      "learning_rate": 8.163200000000002e-06,
+      "loss": 1.8736,
+      "step": 2449
+    },
+    {
+      "epoch": 4.52,
+      "learning_rate": 8.160000000000001e-06,
+      "loss": 1.9521,
+      "step": 2450
+    },
+    {
+      "epoch": 4.52,
+      "learning_rate": 8.156800000000001e-06,
+      "loss": 1.969,
+      "step": 2451
+    },
+    {
+      "epoch": 4.52,
+      "learning_rate": 8.153600000000003e-06,
+      "loss": 1.8936,
+      "step": 2452
+    },
+    {
+      "epoch": 4.53,
+      "learning_rate": 8.1504e-06,
+      "loss": 2.0122,
+      "step": 2453
+    },
+    {
+      "epoch": 4.53,
+      "learning_rate": 8.1472e-06,
+      "loss": 1.9069,
+      "step": 2454
+    },
+    {
+      "epoch": 4.53,
+      "learning_rate": 8.144000000000002e-06,
+      "loss": 1.9569,
+      "step": 2455
+    },
+    {
+      "epoch": 4.53,
+      "learning_rate": 8.140800000000002e-06,
+      "loss": 1.9934,
+      "step": 2456
+    },
+    {
+      "epoch": 4.53,
+      "learning_rate": 8.137600000000002e-06,
+      "loss": 1.9255,
+      "step": 2457
+    },
+    {
+      "epoch": 4.53,
+      "learning_rate": 8.134400000000001e-06,
+      "loss": 1.9475,
+      "step": 2458
+    },
+    {
+      "epoch": 4.54,
+      "learning_rate": 8.131200000000001e-06,
+      "loss": 1.901,
+      "step": 2459
+    },
+    {
+      "epoch": 4.54,
+      "learning_rate": 8.128e-06,
+      "loss": 2.0071,
+      "step": 2460
+    },
+    {
+      "epoch": 4.54,
+      "eval_loss": 2.323665142059326,
+      "eval_runtime": 233.4981,
+      "eval_samples_per_second": 37.144,
+      "eval_steps_per_second": 0.582,
+      "step": 2460
+    },
+    {
+      "epoch": 4.54,
+      "learning_rate": 8.124800000000002e-06,
+      "loss": 1.9504,
+      "step": 2461
+    },
+    {
+      "epoch": 4.54,
+      "learning_rate": 8.121600000000002e-06,
+      "loss": 1.9824,
+      "step": 2462
+    },
+    {
+      "epoch": 4.54,
+      "learning_rate": 8.1184e-06,
+      "loss": 1.9192,
+      "step": 2463
+    },
+    {
+      "epoch": 4.55,
+      "learning_rate": 8.115200000000002e-06,
+      "loss": 1.91,
+      "step": 2464
+    },
+    {
+      "epoch": 4.55,
+      "learning_rate": 8.112000000000001e-06,
+      "loss": 1.8872,
+      "step": 2465
+    },
+    {
+      "epoch": 4.55,
+      "learning_rate": 8.108800000000001e-06,
+      "loss": 1.9108,
+      "step": 2466
+    },
+    {
+      "epoch": 4.55,
+      "learning_rate": 8.105600000000002e-06,
+      "loss": 1.8816,
+      "step": 2467
+    },
+    {
+      "epoch": 4.55,
+      "learning_rate": 8.1024e-06,
+      "loss": 1.9847,
+      "step": 2468
+    },
+    {
+      "epoch": 4.55,
+      "learning_rate": 8.0992e-06,
+      "loss": 1.9992,
+      "step": 2469
+    },
+    {
+      "epoch": 4.56,
+      "learning_rate": 8.096000000000002e-06,
+      "loss": 2.0677,
+      "step": 2470
+    },
+    {
+      "epoch": 4.56,
+      "learning_rate": 8.092800000000002e-06,
+      "loss": 1.7419,
+      "step": 2471
+    },
+    {
+      "epoch": 4.56,
+      "learning_rate": 8.089600000000003e-06,
+      "loss": 1.8939,
+      "step": 2472
+    },
+    {
+      "epoch": 4.56,
+      "learning_rate": 8.086400000000001e-06,
+      "loss": 2.0715,
+      "step": 2473
+    },
+    {
+      "epoch": 4.56,
+      "learning_rate": 8.0832e-06,
+      "loss": 1.8917,
+      "step": 2474
+    },
+    {
+      "epoch": 4.57,
+      "learning_rate": 8.080000000000002e-06,
+      "loss": 2.0029,
+      "step": 2475
+    },
+    {
+      "epoch": 4.57,
+      "learning_rate": 8.076800000000002e-06,
+      "loss": 1.9604,
+      "step": 2476
+    },
+    {
+      "epoch": 4.57,
+      "learning_rate": 8.073600000000002e-06,
+      "loss": 1.8132,
+      "step": 2477
+    },
+    {
+      "epoch": 4.57,
+      "learning_rate": 8.070400000000002e-06,
+      "loss": 1.95,
+      "step": 2478
+    },
+    {
+      "epoch": 4.57,
+      "learning_rate": 8.067200000000001e-06,
+      "loss": 1.9338,
+      "step": 2479
+    },
+    {
+      "epoch": 4.58,
+      "learning_rate": 8.064000000000001e-06,
+      "loss": 1.932,
+      "step": 2480
+    },
+    {
+      "epoch": 4.58,
+      "learning_rate": 8.060800000000003e-06,
+      "loss": 1.8889,
+      "step": 2481
+    },
+    {
+      "epoch": 4.58,
+      "learning_rate": 8.057600000000002e-06,
+      "loss": 1.8513,
+      "step": 2482
+    },
+    {
+      "epoch": 4.58,
+      "learning_rate": 8.0544e-06,
+      "loss": 1.8729,
+      "step": 2483
+    },
+    {
+      "epoch": 4.58,
+      "learning_rate": 8.051200000000002e-06,
+      "loss": 1.9634,
+      "step": 2484
+    },
+    {
+      "epoch": 4.58,
+      "learning_rate": 8.048000000000002e-06,
+      "loss": 1.9367,
+      "step": 2485
+    },
+    {
+      "epoch": 4.59,
+      "learning_rate": 8.044800000000001e-06,
+      "loss": 2.0847,
+      "step": 2486
+    },
+    {
+      "epoch": 4.59,
+      "learning_rate": 8.041600000000003e-06,
+      "loss": 1.9105,
+      "step": 2487
+    },
+    {
+      "epoch": 4.59,
+      "learning_rate": 8.0384e-06,
+      "loss": 1.8886,
+      "step": 2488
+    },
+    {
+      "epoch": 4.59,
+      "learning_rate": 8.0352e-06,
+      "loss": 1.9678,
+      "step": 2489
+    },
+    {
+      "epoch": 4.59,
+      "learning_rate": 8.032000000000002e-06,
+      "loss": 1.9202,
+      "step": 2490
+    },
+    {
+      "epoch": 4.59,
+      "eval_loss": 2.3349480628967285,
+      "eval_runtime": 232.2978,
+      "eval_samples_per_second": 37.336,
+      "eval_steps_per_second": 0.585,
+      "step": 2490
+    },
+    {
+      "epoch": 4.6,
+      "learning_rate": 8.028800000000002e-06,
+      "loss": 1.8818,
+      "step": 2491
+    },
+    {
+      "epoch": 4.6,
+      "learning_rate": 8.025600000000002e-06,
+      "loss": 1.9333,
+      "step": 2492
+    },
+    {
+      "epoch": 4.6,
+      "learning_rate": 8.022400000000001e-06,
+      "loss": 1.9403,
+      "step": 2493
+    },
+    {
+      "epoch": 4.6,
+      "learning_rate": 8.019200000000001e-06,
+      "loss": 1.8588,
+      "step": 2494
+    },
+    {
+      "epoch": 4.6,
+      "learning_rate": 8.016e-06,
+      "loss": 1.8564,
+      "step": 2495
+    },
+    {
+      "epoch": 4.6,
+      "learning_rate": 8.012800000000002e-06,
+      "loss": 1.9575,
+      "step": 2496
+    },
+    {
+      "epoch": 4.61,
+      "learning_rate": 8.009600000000002e-06,
+      "loss": 1.9354,
+      "step": 2497
+    },
+    {
+      "epoch": 4.61,
+      "learning_rate": 8.0064e-06,
+      "loss": 1.8833,
+      "step": 2498
+    },
+    {
+      "epoch": 4.61,
+      "learning_rate": 8.003200000000002e-06,
+      "loss": 1.9407,
+      "step": 2499
+    },
+    {
+      "epoch": 4.61,
+      "learning_rate": 8.000000000000001e-06,
+      "loss": 1.9843,
+      "step": 2500
+    },
+    {
+      "epoch": 4.61,
+      "learning_rate": 7.996800000000001e-06,
+      "loss": 1.8626,
+      "step": 2501
+    },
+    {
+      "epoch": 4.62,
+      "learning_rate": 7.993600000000001e-06,
+      "loss": 1.9058,
+      "step": 2502
+    },
+    {
+      "epoch": 4.62,
+      "learning_rate": 7.990400000000002e-06,
+      "loss": 1.9534,
+      "step": 2503
+    },
+    {
+      "epoch": 4.62,
+      "learning_rate": 7.9872e-06,
+      "loss": 1.8841,
+      "step": 2504
+    },
+    {
+      "epoch": 4.62,
+      "learning_rate": 7.984000000000002e-06,
+      "loss": 1.9359,
+      "step": 2505
+    },
+    {
+      "epoch": 4.62,
+      "learning_rate": 7.980800000000002e-06,
+      "loss": 1.9591,
+      "step": 2506
+    },
+    {
+      "epoch": 4.62,
+      "learning_rate": 7.977600000000001e-06,
+      "loss": 1.9562,
+      "step": 2507
+    },
+    {
+      "epoch": 4.63,
+      "learning_rate": 7.974400000000001e-06,
+      "loss": 1.9081,
+      "step": 2508
+    },
+    {
+      "epoch": 4.63,
+      "learning_rate": 7.971200000000001e-06,
+      "loss": 1.935,
+      "step": 2509
+    },
+    {
+      "epoch": 4.63,
+      "learning_rate": 7.968e-06,
+      "loss": 1.8801,
+      "step": 2510
+    },
+    {
+      "epoch": 4.63,
+      "learning_rate": 7.964800000000002e-06,
+      "loss": 1.8351,
+      "step": 2511
+    },
+    {
+      "epoch": 4.63,
+      "learning_rate": 7.961600000000002e-06,
+      "loss": 1.8588,
+      "step": 2512
+    },
+    {
+      "epoch": 4.64,
+      "learning_rate": 7.958400000000002e-06,
+      "loss": 1.959,
+      "step": 2513
+    },
+    {
+      "epoch": 4.64,
+      "learning_rate": 7.955200000000001e-06,
+      "loss": 1.8872,
+      "step": 2514
+    },
+    {
+      "epoch": 4.64,
+      "learning_rate": 7.952000000000001e-06,
+      "loss": 1.9269,
+      "step": 2515
+    },
+    {
+      "epoch": 4.64,
+      "learning_rate": 7.948800000000001e-06,
+      "loss": 1.9122,
+      "step": 2516
+    },
+    {
+      "epoch": 4.64,
+      "learning_rate": 7.9456e-06,
+      "loss": 1.8921,
+      "step": 2517
+    },
+    {
+      "epoch": 4.65,
+      "learning_rate": 7.942400000000002e-06,
+      "loss": 2.0362,
+      "step": 2518
+    },
+    {
+      "epoch": 4.65,
+      "learning_rate": 7.9392e-06,
+      "loss": 1.8989,
+      "step": 2519
+    },
+    {
+      "epoch": 4.65,
+      "learning_rate": 7.936000000000002e-06,
+      "loss": 1.9799,
+      "step": 2520
+    },
+    {
+      "epoch": 4.65,
+      "eval_loss": 2.3289718627929688,
+      "eval_runtime": 231.0353,
+      "eval_samples_per_second": 37.54,
+      "eval_steps_per_second": 0.589,
+      "step": 2520
+    },
+    {
+      "epoch": 4.65,
+      "learning_rate": 7.932800000000001e-06,
+      "loss": 2.0154,
+      "step": 2521
+    },
+    {
+      "epoch": 4.65,
+      "learning_rate": 7.929600000000001e-06,
+      "loss": 1.9052,
+      "step": 2522
+    },
+    {
+      "epoch": 4.65,
+      "learning_rate": 7.926400000000001e-06,
+      "loss": 1.8425,
+      "step": 2523
+    },
+    {
+      "epoch": 4.66,
+      "learning_rate": 7.9232e-06,
+      "loss": 1.9118,
+      "step": 2524
+    },
+    {
+      "epoch": 4.66,
+      "learning_rate": 7.92e-06,
+      "loss": 1.9064,
+      "step": 2525
+    },
+    {
+      "epoch": 4.66,
+      "learning_rate": 7.916800000000002e-06,
+      "loss": 1.9375,
+      "step": 2526
+    },
+    {
+      "epoch": 4.66,
+      "learning_rate": 7.913600000000002e-06,
+      "loss": 1.9409,
+      "step": 2527
+    },
+    {
+      "epoch": 4.66,
+      "learning_rate": 7.910400000000001e-06,
+      "loss": 1.8784,
+      "step": 2528
+    },
+    {
+      "epoch": 4.67,
+      "learning_rate": 7.907200000000001e-06,
+      "loss": 1.9121,
+      "step": 2529
+    },
+    {
+      "epoch": 4.67,
+      "learning_rate": 7.904000000000001e-06,
+      "loss": 2.0343,
+      "step": 2530
+    },
+    {
+      "epoch": 4.67,
+      "learning_rate": 7.900800000000002e-06,
+      "loss": 2.0449,
+      "step": 2531
+    },
+    {
+      "epoch": 4.67,
+      "learning_rate": 7.8976e-06,
+      "loss": 1.8744,
+      "step": 2532
+    },
+    {
+      "epoch": 4.67,
+      "learning_rate": 7.894400000000002e-06,
+      "loss": 1.8817,
+      "step": 2533
+    },
+    {
+      "epoch": 4.67,
+      "learning_rate": 7.891200000000002e-06,
+      "loss": 1.7565,
+      "step": 2534
+    },
+    {
+      "epoch": 4.68,
+      "learning_rate": 7.888000000000001e-06,
+      "loss": 1.8801,
+      "step": 2535
+    },
+    {
+      "epoch": 4.68,
+      "learning_rate": 7.884800000000001e-06,
+      "loss": 1.8727,
+      "step": 2536
+    },
+    {
+      "epoch": 4.68,
+      "learning_rate": 7.881600000000001e-06,
+      "loss": 1.9227,
+      "step": 2537
+    },
+    {
+      "epoch": 4.68,
+      "learning_rate": 7.8784e-06,
+      "loss": 1.945,
+      "step": 2538
+    },
+    {
+      "epoch": 4.68,
+      "learning_rate": 7.875200000000002e-06,
+      "loss": 1.8098,
+      "step": 2539
+    },
+    {
+      "epoch": 4.69,
+      "learning_rate": 7.872000000000002e-06,
+      "loss": 1.9976,
+      "step": 2540
+    },
+    {
+      "epoch": 4.69,
+      "learning_rate": 7.868800000000002e-06,
+      "loss": 1.9871,
+      "step": 2541
+    },
+    {
+      "epoch": 4.69,
+      "learning_rate": 7.865600000000001e-06,
+      "loss": 1.9568,
+      "step": 2542
+    },
+    {
+      "epoch": 4.69,
+      "learning_rate": 7.862400000000001e-06,
+      "loss": 1.8938,
+      "step": 2543
+    },
+    {
+      "epoch": 4.69,
+      "learning_rate": 7.859200000000001e-06,
+      "loss": 2.0642,
+      "step": 2544
+    },
+    {
+      "epoch": 4.69,
+      "learning_rate": 7.856e-06,
+      "loss": 1.9479,
+      "step": 2545
+    },
+    {
+      "epoch": 4.7,
+      "learning_rate": 7.852800000000002e-06,
+      "loss": 1.9048,
+      "step": 2546
+    },
+    {
+      "epoch": 4.7,
+      "learning_rate": 7.8496e-06,
+      "loss": 1.9927,
+      "step": 2547
+    },
+    {
+      "epoch": 4.7,
+      "learning_rate": 7.846400000000002e-06,
+      "loss": 1.9195,
+      "step": 2548
+    },
+    {
+      "epoch": 4.7,
+      "learning_rate": 7.843200000000001e-06,
+      "loss": 1.8294,
+      "step": 2549
+    },
+    {
+      "epoch": 4.7,
+      "learning_rate": 7.840000000000001e-06,
+      "loss": 1.9392,
+      "step": 2550
+    },
+    {
+      "epoch": 4.7,
+      "eval_loss": 2.3357081413269043,
+      "eval_runtime": 231.0561,
+      "eval_samples_per_second": 37.536,
+      "eval_steps_per_second": 0.589,
+      "step": 2550
+    },
+    {
+      "epoch": 4.71,
+      "learning_rate": 7.836800000000001e-06,
+      "loss": 2.0408,
+      "step": 2551
+    },
+    {
+      "epoch": 4.71,
+      "learning_rate": 7.8336e-06,
+      "loss": 1.8745,
+      "step": 2552
+    },
+    {
+      "epoch": 4.71,
+      "learning_rate": 7.830400000000002e-06,
+      "loss": 1.949,
+      "step": 2553
+    },
+    {
+      "epoch": 4.71,
+      "learning_rate": 7.827200000000002e-06,
+      "loss": 1.7944,
+      "step": 2554
+    },
+    {
+      "epoch": 4.71,
+      "learning_rate": 7.824000000000002e-06,
+      "loss": 1.9048,
+      "step": 2555
+    },
+    {
+      "epoch": 4.72,
+      "learning_rate": 7.820800000000001e-06,
+      "loss": 1.7758,
+      "step": 2556
+    },
+    {
+      "epoch": 4.72,
+      "learning_rate": 7.817600000000001e-06,
+      "loss": 1.8961,
+      "step": 2557
+    },
+    {
+      "epoch": 4.72,
+      "learning_rate": 7.814400000000001e-06,
+      "loss": 1.9944,
+      "step": 2558
+    },
+    {
+      "epoch": 4.72,
+      "learning_rate": 7.811200000000002e-06,
+      "loss": 1.879,
+      "step": 2559
+    },
+    {
+      "epoch": 4.72,
+      "learning_rate": 7.808e-06,
+      "loss": 1.9827,
+      "step": 2560
+    },
+    {
+      "epoch": 4.72,
+      "learning_rate": 7.804800000000002e-06,
+      "loss": 1.8817,
+      "step": 2561
+    },
+    {
+      "epoch": 4.73,
+      "learning_rate": 7.801600000000002e-06,
+      "loss": 2.0028,
+      "step": 2562
+    },
+    {
+      "epoch": 4.73,
+      "learning_rate": 7.798400000000001e-06,
+      "loss": 1.8458,
+      "step": 2563
+    },
+    {
+      "epoch": 4.73,
+      "learning_rate": 7.795200000000001e-06,
+      "loss": 1.9572,
+      "step": 2564
+    },
+    {
+      "epoch": 4.73,
+      "learning_rate": 7.792000000000001e-06,
+      "loss": 1.8948,
+      "step": 2565
+    },
+    {
+      "epoch": 4.73,
+      "learning_rate": 7.7888e-06,
+      "loss": 1.8661,
+      "step": 2566
+    },
+    {
+      "epoch": 4.74,
+      "learning_rate": 7.7856e-06,
+      "loss": 1.9466,
+      "step": 2567
+    },
+    {
+      "epoch": 4.74,
+      "learning_rate": 7.782400000000002e-06,
+      "loss": 1.8045,
+      "step": 2568
+    },
+    {
+      "epoch": 4.74,
+      "learning_rate": 7.779200000000002e-06,
+      "loss": 1.9829,
+      "step": 2569
+    },
+    {
+      "epoch": 4.74,
+      "learning_rate": 7.776000000000001e-06,
+      "loss": 1.8898,
+      "step": 2570
+    },
+    {
+      "epoch": 4.74,
+      "learning_rate": 7.772800000000001e-06,
+      "loss": 1.9235,
+      "step": 2571
+    },
+    {
+      "epoch": 4.74,
+      "learning_rate": 7.769600000000001e-06,
+      "loss": 2.0424,
+      "step": 2572
+    },
+    {
+      "epoch": 4.75,
+      "learning_rate": 7.7664e-06,
+      "loss": 1.9636,
+      "step": 2573
+    },
+    {
+      "epoch": 4.75,
+      "learning_rate": 7.763200000000002e-06,
+      "loss": 1.9493,
+      "step": 2574
+    },
+    {
+      "epoch": 4.75,
+      "learning_rate": 7.76e-06,
+      "loss": 1.9251,
+      "step": 2575
+    },
+    {
+      "epoch": 4.75,
+      "learning_rate": 7.756800000000002e-06,
+      "loss": 1.9653,
+      "step": 2576
+    },
+    {
+      "epoch": 4.75,
+      "learning_rate": 7.753600000000001e-06,
+      "loss": 1.8906,
+      "step": 2577
+    },
+    {
+      "epoch": 4.76,
+      "learning_rate": 7.750400000000001e-06,
+      "loss": 1.9959,
+      "step": 2578
+    },
+    {
+      "epoch": 4.76,
+      "learning_rate": 7.747200000000001e-06,
+      "loss": 1.8662,
+      "step": 2579
+    },
+    {
+      "epoch": 4.76,
+      "learning_rate": 7.744e-06,
+      "loss": 1.9383,
+      "step": 2580
+    },
+    {
+      "epoch": 4.76,
+      "eval_loss": 2.331726312637329,
+      "eval_runtime": 232.689,
+      "eval_samples_per_second": 37.273,
+      "eval_steps_per_second": 0.584,
+      "step": 2580
+    },
+    {
+      "epoch": 4.76,
+      "learning_rate": 7.740800000000002e-06,
+      "loss": 2.0692,
+      "step": 2581
+    },
+    {
+      "epoch": 4.76,
+      "learning_rate": 7.7376e-06,
+      "loss": 1.7585,
+      "step": 2582
+    },
+    {
+      "epoch": 4.76,
+      "learning_rate": 7.734400000000002e-06,
+      "loss": 2.0359,
+      "step": 2583
+    },
+    {
+      "epoch": 4.77,
+      "learning_rate": 7.731200000000001e-06,
+      "loss": 1.9946,
+      "step": 2584
+    },
+    {
+      "epoch": 4.77,
+      "learning_rate": 7.728000000000001e-06,
+      "loss": 1.8792,
+      "step": 2585
+    },
+    {
+      "epoch": 4.77,
+      "learning_rate": 7.724800000000001e-06,
+      "loss": 1.9209,
+      "step": 2586
+    },
+    {
+      "epoch": 4.77,
+      "learning_rate": 7.7216e-06,
+      "loss": 1.86,
+      "step": 2587
+    },
+    {
+      "epoch": 4.77,
+      "learning_rate": 7.7184e-06,
+      "loss": 1.9539,
+      "step": 2588
+    },
+    {
+      "epoch": 4.78,
+      "learning_rate": 7.715200000000002e-06,
+      "loss": 2.0028,
+      "step": 2589
+    },
+    {
+      "epoch": 4.78,
+      "learning_rate": 7.712000000000002e-06,
+      "loss": 1.8852,
+      "step": 2590
+    },
+    {
+      "epoch": 4.78,
+      "learning_rate": 7.708800000000001e-06,
+      "loss": 1.9868,
+      "step": 2591
+    },
+    {
+      "epoch": 4.78,
+      "learning_rate": 7.705600000000001e-06,
+      "loss": 1.9016,
+      "step": 2592
+    },
+    {
+      "epoch": 4.78,
+      "learning_rate": 7.702400000000001e-06,
+      "loss": 1.9074,
+      "step": 2593
+    },
+    {
+      "epoch": 4.79,
+      "learning_rate": 7.6992e-06,
+      "loss": 1.9959,
+      "step": 2594
+    },
+    {
+      "epoch": 4.79,
+      "learning_rate": 7.696e-06,
+      "loss": 1.8709,
+      "step": 2595
+    },
+    {
+      "epoch": 4.79,
+      "learning_rate": 7.692800000000002e-06,
+      "loss": 1.8318,
+      "step": 2596
+    },
+    {
+      "epoch": 4.79,
+      "learning_rate": 7.689600000000002e-06,
+      "loss": 1.9542,
+      "step": 2597
+    },
+    {
+      "epoch": 4.79,
+      "learning_rate": 7.686400000000001e-06,
+      "loss": 1.9362,
+      "step": 2598
+    },
+    {
+      "epoch": 4.79,
+      "learning_rate": 7.683200000000001e-06,
+      "loss": 1.9319,
+      "step": 2599
+    },
+    {
+      "epoch": 4.8,
+      "learning_rate": 7.680000000000001e-06,
+      "loss": 1.9687,
+      "step": 2600
+    },
+    {
+      "epoch": 4.8,
+      "learning_rate": 7.6768e-06,
+      "loss": 1.8506,
+      "step": 2601
+    },
+    {
+      "epoch": 4.8,
+      "learning_rate": 7.673600000000002e-06,
+      "loss": 1.8887,
+      "step": 2602
+    },
+    {
+      "epoch": 4.8,
+      "learning_rate": 7.670400000000002e-06,
+      "loss": 1.9564,
+      "step": 2603
+    },
+    {
+      "epoch": 4.8,
+      "learning_rate": 7.667200000000002e-06,
+      "loss": 1.9905,
+      "step": 2604
+    },
+    {
+      "epoch": 4.81,
+      "learning_rate": 7.664000000000001e-06,
+      "loss": 2.025,
+      "step": 2605
+    },
+    {
+      "epoch": 4.81,
+      "learning_rate": 7.660800000000001e-06,
+      "loss": 1.9872,
+      "step": 2606
+    },
+    {
+      "epoch": 4.81,
+      "learning_rate": 7.657600000000001e-06,
+      "loss": 1.999,
+      "step": 2607
+    },
+    {
+      "epoch": 4.81,
+      "learning_rate": 7.6544e-06,
+      "loss": 1.9223,
+      "step": 2608
+    },
+    {
+      "epoch": 4.81,
+      "learning_rate": 7.651200000000002e-06,
+      "loss": 1.867,
+      "step": 2609
+    },
+    {
+      "epoch": 4.81,
+      "learning_rate": 7.648e-06,
+      "loss": 1.9189,
+      "step": 2610
+    },
+    {
+      "epoch": 4.81,
+      "eval_loss": 2.3364064693450928,
+      "eval_runtime": 231.8408,
+      "eval_samples_per_second": 37.409,
+      "eval_steps_per_second": 0.587,
+      "step": 2610
+    },
+    {
+      "epoch": 4.82,
+      "learning_rate": 7.644800000000002e-06,
+      "loss": 1.866,
+      "step": 2611
+    },
+    {
+      "epoch": 4.82,
+      "learning_rate": 7.641600000000002e-06,
+      "loss": 1.9361,
+      "step": 2612
+    },
+    {
+      "epoch": 4.82,
+      "learning_rate": 7.638400000000001e-06,
+      "loss": 1.9911,
+      "step": 2613
+    },
+    {
+      "epoch": 4.82,
+      "learning_rate": 7.635200000000001e-06,
+      "loss": 1.7862,
+      "step": 2614
+    },
+    {
+      "epoch": 4.82,
+      "learning_rate": 7.632e-06,
+      "loss": 1.8319,
+      "step": 2615
+    },
+    {
+      "epoch": 4.83,
+      "learning_rate": 7.628800000000001e-06,
+      "loss": 1.8625,
+      "step": 2616
+    },
+    {
+      "epoch": 4.83,
+      "learning_rate": 7.625600000000002e-06,
+      "loss": 2.001,
+      "step": 2617
+    },
+    {
+      "epoch": 4.83,
+      "learning_rate": 7.622400000000001e-06,
+      "loss": 1.811,
+      "step": 2618
+    },
+    {
+      "epoch": 4.83,
+      "learning_rate": 7.6192000000000015e-06,
+      "loss": 1.9155,
+      "step": 2619
+    },
+    {
+      "epoch": 4.83,
+      "learning_rate": 7.616000000000001e-06,
+      "loss": 1.9318,
+      "step": 2620
+    },
+    {
+      "epoch": 4.84,
+      "learning_rate": 7.612800000000001e-06,
+      "loss": 1.8423,
+      "step": 2621
+    },
+    {
+      "epoch": 4.84,
+      "learning_rate": 7.609600000000002e-06,
+      "loss": 1.9674,
+      "step": 2622
+    },
+    {
+      "epoch": 4.84,
+      "learning_rate": 7.606400000000001e-06,
+      "loss": 1.9374,
+      "step": 2623
+    },
+    {
+      "epoch": 4.84,
+      "learning_rate": 7.603200000000001e-06,
+      "loss": 1.9196,
+      "step": 2624
+    },
+    {
+      "epoch": 4.84,
+      "learning_rate": 7.600000000000001e-06,
+      "loss": 1.8532,
+      "step": 2625
+    },
+    {
+      "epoch": 4.84,
+      "learning_rate": 7.5968000000000015e-06,
+      "loss": 1.9386,
+      "step": 2626
+    },
+    {
+      "epoch": 4.85,
+      "learning_rate": 7.593600000000001e-06,
+      "loss": 1.8651,
+      "step": 2627
+    },
+    {
+      "epoch": 4.85,
+      "learning_rate": 7.590400000000001e-06,
+      "loss": 1.9193,
+      "step": 2628
+    },
+    {
+      "epoch": 4.85,
+      "learning_rate": 7.587200000000002e-06,
+      "loss": 2.0639,
+      "step": 2629
+    },
+    {
+      "epoch": 4.85,
+      "learning_rate": 7.5840000000000006e-06,
+      "loss": 1.8357,
+      "step": 2630
+    },
+    {
+      "epoch": 4.85,
+      "learning_rate": 7.580800000000001e-06,
+      "loss": 1.8824,
+      "step": 2631
+    },
+    {
+      "epoch": 4.86,
+      "learning_rate": 7.577600000000002e-06,
+      "loss": 1.9785,
+      "step": 2632
+    },
+    {
+      "epoch": 4.86,
+      "learning_rate": 7.574400000000001e-06,
+      "loss": 1.9543,
+      "step": 2633
+    },
+    {
+      "epoch": 4.86,
+      "learning_rate": 7.571200000000001e-06,
+      "loss": 1.8822,
+      "step": 2634
+    },
+    {
+      "epoch": 4.86,
+      "learning_rate": 7.568000000000001e-06,
+      "loss": 1.9292,
+      "step": 2635
+    },
+    {
+      "epoch": 4.86,
+      "learning_rate": 7.564800000000001e-06,
+      "loss": 1.8962,
+      "step": 2636
+    },
+    {
+      "epoch": 4.86,
+      "learning_rate": 7.5616000000000014e-06,
+      "loss": 1.8977,
+      "step": 2637
+    },
+    {
+      "epoch": 4.87,
+      "learning_rate": 7.558400000000001e-06,
+      "loss": 1.978,
+      "step": 2638
+    },
+    {
+      "epoch": 4.87,
+      "learning_rate": 7.555200000000002e-06,
+      "loss": 1.9722,
+      "step": 2639
+    },
+    {
+      "epoch": 4.87,
+      "learning_rate": 7.552000000000001e-06,
+      "loss": 2.0201,
+      "step": 2640
+    },
+    {
+      "epoch": 4.87,
+      "eval_loss": 2.328502893447876,
+      "eval_runtime": 232.9363,
+      "eval_samples_per_second": 37.233,
+      "eval_steps_per_second": 0.584,
+      "step": 2640
+    },
+    {
+      "epoch": 4.87,
+      "learning_rate": 7.548800000000001e-06,
+      "loss": 1.9562,
+      "step": 2641
+    },
+    {
+      "epoch": 4.87,
+      "learning_rate": 7.545600000000002e-06,
+      "loss": 1.8568,
+      "step": 2642
+    },
+    {
+      "epoch": 4.88,
+      "learning_rate": 7.542400000000001e-06,
+      "loss": 1.9314,
+      "step": 2643
+    },
+    {
+      "epoch": 4.88,
+      "learning_rate": 7.5392000000000014e-06,
+      "loss": 1.9394,
+      "step": 2644
+    },
+    {
+      "epoch": 4.88,
+      "learning_rate": 7.536000000000001e-06,
+      "loss": 1.8549,
+      "step": 2645
+    },
+    {
+      "epoch": 4.88,
+      "learning_rate": 7.532800000000001e-06,
+      "loss": 1.9995,
+      "step": 2646
+    },
+    {
+      "epoch": 4.88,
+      "learning_rate": 7.5296000000000016e-06,
+      "loss": 1.9574,
+      "step": 2647
+    },
+    {
+      "epoch": 4.88,
+      "learning_rate": 7.526400000000001e-06,
+      "loss": 2.0048,
+      "step": 2648
+    },
+    {
+      "epoch": 4.89,
+      "learning_rate": 7.523200000000001e-06,
+      "loss": 1.99,
+      "step": 2649
+    },
+    {
+      "epoch": 4.89,
+      "learning_rate": 7.520000000000001e-06,
+      "loss": 2.0299,
+      "step": 2650
+    },
+    {
+      "epoch": 4.89,
+      "learning_rate": 7.5168000000000015e-06,
+      "loss": 1.9355,
+      "step": 2651
+    },
+    {
+      "epoch": 4.89,
+      "learning_rate": 7.513600000000001e-06,
+      "loss": 1.8895,
+      "step": 2652
+    },
+    {
+      "epoch": 4.89,
+      "learning_rate": 7.510400000000001e-06,
+      "loss": 1.8538,
+      "step": 2653
+    },
+    {
+      "epoch": 4.9,
+      "learning_rate": 7.507200000000002e-06,
+      "loss": 1.7771,
+      "step": 2654
+    },
+    {
+      "epoch": 4.9,
+      "learning_rate": 7.5040000000000005e-06,
+      "loss": 2.0104,
+      "step": 2655
+    },
+    {
+      "epoch": 4.9,
+      "learning_rate": 7.500800000000001e-06,
+      "loss": 1.9733,
+      "step": 2656
+    },
+    {
+      "epoch": 4.9,
+      "learning_rate": 7.497600000000002e-06,
+      "loss": 2.0738,
+      "step": 2657
+    },
+    {
+      "epoch": 4.9,
+      "learning_rate": 7.494400000000001e-06,
+      "loss": 1.9057,
+      "step": 2658
+    },
+    {
+      "epoch": 4.91,
+      "learning_rate": 7.491200000000001e-06,
+      "loss": 2.0034,
+      "step": 2659
+    },
+    {
+      "epoch": 4.91,
+      "learning_rate": 7.488000000000002e-06,
+      "loss": 2.023,
+      "step": 2660
+    },
+    {
+      "epoch": 4.91,
+      "learning_rate": 7.484800000000001e-06,
+      "loss": 1.9705,
+      "step": 2661
+    },
+    {
+      "epoch": 4.91,
+      "learning_rate": 7.481600000000001e-06,
+      "loss": 1.892,
+      "step": 2662
+    },
+    {
+      "epoch": 4.91,
+      "learning_rate": 7.478400000000001e-06,
+      "loss": 1.9532,
+      "step": 2663
+    },
+    {
+      "epoch": 4.91,
+      "learning_rate": 7.475200000000002e-06,
+      "loss": 1.903,
+      "step": 2664
+    },
+    {
+      "epoch": 4.92,
+      "learning_rate": 7.4720000000000015e-06,
+      "loss": 1.9743,
+      "step": 2665
+    },
+    {
+      "epoch": 4.92,
+      "learning_rate": 7.468800000000001e-06,
+      "loss": 1.936,
+      "step": 2666
+    },
+    {
+      "epoch": 4.92,
+      "learning_rate": 7.465600000000002e-06,
+      "loss": 1.9907,
+      "step": 2667
+    },
+    {
+      "epoch": 4.92,
+      "learning_rate": 7.462400000000001e-06,
+      "loss": 1.9426,
+      "step": 2668
+    },
+    {
+      "epoch": 4.92,
+      "learning_rate": 7.459200000000001e-06,
+      "loss": 1.9318,
+      "step": 2669
+    },
+    {
+      "epoch": 4.93,
+      "learning_rate": 7.456000000000002e-06,
+      "loss": 1.8968,
+      "step": 2670
+    },
+    {
+      "epoch": 4.93,
+      "eval_loss": 2.332251787185669,
+      "eval_runtime": 231.7709,
+      "eval_samples_per_second": 37.421,
+      "eval_steps_per_second": 0.587,
+      "step": 2670
+    },
+    {
+      "epoch": 4.93,
+      "learning_rate": 7.452800000000001e-06,
+      "loss": 1.991,
+      "step": 2671
+    },
+    {
+      "epoch": 4.93,
+      "learning_rate": 7.4496000000000015e-06,
+      "loss": 2.0032,
+      "step": 2672
+    },
+    {
+      "epoch": 4.93,
+      "learning_rate": 7.446400000000001e-06,
+      "loss": 1.9601,
+      "step": 2673
+    },
+    {
+      "epoch": 4.93,
+      "learning_rate": 7.443200000000001e-06,
+      "loss": 1.9384,
+      "step": 2674
+    },
+    {
+      "epoch": 4.93,
+      "learning_rate": 7.440000000000002e-06,
+      "loss": 2.005,
+      "step": 2675
+    },
+    {
+      "epoch": 4.94,
+      "learning_rate": 7.436800000000001e-06,
+      "loss": 1.8592,
+      "step": 2676
+    },
+    {
+      "epoch": 4.94,
+      "learning_rate": 7.433600000000001e-06,
+      "loss": 1.8573,
+      "step": 2677
+    },
+    {
+      "epoch": 4.94,
+      "learning_rate": 7.430400000000001e-06,
+      "loss": 2.013,
+      "step": 2678
+    },
+    {
+      "epoch": 4.94,
+      "learning_rate": 7.4272000000000015e-06,
+      "loss": 1.8743,
+      "step": 2679
+    },
+    {
+      "epoch": 4.94,
+      "learning_rate": 7.424000000000001e-06,
+      "loss": 1.964,
+      "step": 2680
+    },
+    {
+      "epoch": 4.95,
+      "learning_rate": 7.420800000000001e-06,
+      "loss": 2.0032,
+      "step": 2681
+    },
+    {
+      "epoch": 4.95,
+      "learning_rate": 7.417600000000002e-06,
+      "loss": 1.9663,
+      "step": 2682
+    },
+    {
+      "epoch": 4.95,
+      "learning_rate": 7.4144000000000005e-06,
+      "loss": 1.9578,
+      "step": 2683
+    },
+    {
+      "epoch": 4.95,
+      "learning_rate": 7.411200000000001e-06,
+      "loss": 1.8873,
+      "step": 2684
+    },
+    {
+      "epoch": 4.95,
+      "learning_rate": 7.408000000000002e-06,
+      "loss": 2.0378,
+      "step": 2685
+    },
+    {
+      "epoch": 4.95,
+      "learning_rate": 7.404800000000001e-06,
+      "loss": 1.9195,
+      "step": 2686
+    },
+    {
+      "epoch": 4.96,
+      "learning_rate": 7.401600000000001e-06,
+      "loss": 1.9351,
+      "step": 2687
+    },
+    {
+      "epoch": 4.96,
+      "learning_rate": 7.398400000000001e-06,
+      "loss": 1.8651,
+      "step": 2688
+    },
+    {
+      "epoch": 4.96,
+      "learning_rate": 7.395200000000002e-06,
+      "loss": 1.8187,
+      "step": 2689
+    },
+    {
+      "epoch": 4.96,
+      "learning_rate": 7.392000000000001e-06,
+      "loss": 1.8903,
+      "step": 2690
+    },
+    {
+      "epoch": 4.96,
+      "learning_rate": 7.388800000000001e-06,
+      "loss": 2.0029,
+      "step": 2691
+    },
+    {
+      "epoch": 4.97,
+      "learning_rate": 7.385600000000002e-06,
+      "loss": 1.8551,
+      "step": 2692
+    },
+    {
+      "epoch": 4.97,
+      "learning_rate": 7.382400000000001e-06,
+      "loss": 1.9013,
+      "step": 2693
+    },
+    {
+      "epoch": 4.97,
+      "learning_rate": 7.379200000000001e-06,
+      "loss": 1.9249,
+      "step": 2694
+    },
+    {
+      "epoch": 4.97,
+      "learning_rate": 7.376000000000002e-06,
+      "loss": 1.7767,
+      "step": 2695
+    },
+    {
+      "epoch": 4.97,
+      "learning_rate": 7.372800000000001e-06,
+      "loss": 1.8548,
+      "step": 2696
+    },
+    {
+      "epoch": 4.98,
+      "learning_rate": 7.369600000000001e-06,
+      "loss": 1.8735,
+      "step": 2697
+    },
+    {
+      "epoch": 4.98,
+      "learning_rate": 7.366400000000001e-06,
+      "loss": 1.9597,
+      "step": 2698
+    },
+    {
+      "epoch": 4.98,
+      "learning_rate": 7.363200000000001e-06,
+      "loss": 1.9196,
+      "step": 2699
+    },
+    {
+      "epoch": 4.98,
+      "learning_rate": 7.3600000000000015e-06,
+      "loss": 1.9529,
+      "step": 2700
+    },
+    {
+      "epoch": 4.98,
+      "eval_loss": 2.328953742980957,
+      "eval_runtime": 234.1538,
+      "eval_samples_per_second": 37.04,
+      "eval_steps_per_second": 0.581,
+      "step": 2700
+    },
+    {
+      "epoch": 4.98,
+      "learning_rate": 7.356800000000001e-06,
+      "loss": 2.0365,
+      "step": 2701
+    },
+    {
+      "epoch": 4.98,
+      "learning_rate": 7.353600000000001e-06,
+      "loss": 1.9399,
+      "step": 2702
+    },
+    {
+      "epoch": 4.99,
+      "learning_rate": 7.350400000000001e-06,
+      "loss": 1.88,
+      "step": 2703
+    },
+    {
+      "epoch": 4.99,
+      "learning_rate": 7.347200000000001e-06,
+      "loss": 1.9955,
+      "step": 2704
+    },
+    {
+      "epoch": 4.99,
+      "learning_rate": 7.344000000000001e-06,
+      "loss": 2.0041,
+      "step": 2705
+    },
+    {
+      "epoch": 4.99,
+      "learning_rate": 7.340800000000001e-06,
+      "loss": 1.9545,
+      "step": 2706
+    },
+    {
+      "epoch": 4.99,
+      "learning_rate": 7.3376000000000015e-06,
+      "loss": 1.9446,
+      "step": 2707
+    },
+    {
+      "epoch": 5.0,
+      "learning_rate": 7.3344000000000005e-06,
+      "loss": 1.9548,
+      "step": 2708
+    },
+    {
+      "epoch": 5.0,
+      "learning_rate": 7.331200000000001e-06,
+      "loss": 1.9536,
+      "step": 2709
+    },
+    {
+      "epoch": 5.0,
+      "learning_rate": 7.328000000000002e-06,
+      "loss": 1.9908,
+      "step": 2710
+    },
+    {
+      "epoch": 5.0,
+      "learning_rate": 7.324800000000001e-06,
+      "loss": 2.7844,
+      "step": 2711
+    },
+    {
+      "epoch": 5.0,
+      "learning_rate": 7.321600000000001e-06,
+      "loss": 1.7113,
+      "step": 2712
+    },
+    {
+      "epoch": 5.01,
+      "learning_rate": 7.318400000000001e-06,
+      "loss": 1.9343,
+      "step": 2713
+    },
+    {
+      "epoch": 5.01,
+      "learning_rate": 7.3152000000000016e-06,
+      "loss": 1.8639,
+      "step": 2714
+    },
+    {
+      "epoch": 5.01,
+      "learning_rate": 7.312000000000001e-06,
+      "loss": 1.8697,
+      "step": 2715
+    },
+    {
+      "epoch": 5.01,
+      "learning_rate": 7.308800000000001e-06,
+      "loss": 1.9193,
+      "step": 2716
+    },
+    {
+      "epoch": 5.01,
+      "learning_rate": 7.305600000000002e-06,
+      "loss": 1.864,
+      "step": 2717
+    },
+    {
+      "epoch": 5.01,
+      "learning_rate": 7.302400000000001e-06,
+      "loss": 1.7946,
+      "step": 2718
+    },
+    {
+      "epoch": 5.02,
+      "learning_rate": 7.299200000000001e-06,
+      "loss": 1.9314,
+      "step": 2719
+    },
+    {
+      "epoch": 5.02,
+      "learning_rate": 7.296000000000002e-06,
+      "loss": 1.9329,
+      "step": 2720
+    },
+    {
+      "epoch": 5.02,
+      "learning_rate": 7.292800000000001e-06,
+      "loss": 1.9048,
+      "step": 2721
+    },
+    {
+      "epoch": 5.02,
+      "learning_rate": 7.289600000000001e-06,
+      "loss": 1.8942,
+      "step": 2722
+    },
+    {
+      "epoch": 5.02,
+      "learning_rate": 7.286400000000002e-06,
+      "loss": 1.8448,
+      "step": 2723
+    },
+    {
+      "epoch": 5.03,
+      "learning_rate": 7.283200000000001e-06,
+      "loss": 1.8172,
+      "step": 2724
+    },
+    {
+      "epoch": 5.03,
+      "learning_rate": 7.2800000000000014e-06,
+      "loss": 1.7414,
+      "step": 2725
+    },
+    {
+      "epoch": 5.03,
+      "learning_rate": 7.276800000000001e-06,
+      "loss": 1.7909,
+      "step": 2726
+    },
+    {
+      "epoch": 5.03,
+      "learning_rate": 7.273600000000001e-06,
+      "loss": 1.8731,
+      "step": 2727
+    },
+    {
+      "epoch": 5.03,
+      "learning_rate": 7.2704000000000016e-06,
+      "loss": 1.8621,
+      "step": 2728
+    },
+    {
+      "epoch": 5.04,
+      "learning_rate": 7.267200000000001e-06,
+      "loss": 1.9962,
+      "step": 2729
+    },
+    {
+      "epoch": 5.04,
+      "learning_rate": 7.264000000000001e-06,
+      "loss": 1.7698,
+      "step": 2730
+    },
+    {
+      "epoch": 5.04,
+      "eval_loss": 2.3660061359405518,
+      "eval_runtime": 231.0992,
+      "eval_samples_per_second": 37.529,
+      "eval_steps_per_second": 0.588,
+      "step": 2730
+    },
+    {
+      "epoch": 5.04,
+      "learning_rate": 7.260800000000001e-06,
+      "loss": 1.8858,
+      "step": 2731
+    },
+    {
+      "epoch": 5.04,
+      "learning_rate": 7.2576000000000015e-06,
+      "loss": 1.7748,
+      "step": 2732
+    },
+    {
+      "epoch": 5.04,
+      "learning_rate": 7.254400000000001e-06,
+      "loss": 1.8015,
+      "step": 2733
+    },
+    {
+      "epoch": 5.04,
+      "learning_rate": 7.251200000000001e-06,
+      "loss": 1.8494,
+      "step": 2734
+    },
+    {
+      "epoch": 5.05,
+      "learning_rate": 7.248000000000002e-06,
+      "loss": 1.8336,
+      "step": 2735
+    },
+    {
+      "epoch": 5.05,
+      "learning_rate": 7.2448000000000005e-06,
+      "loss": 1.8686,
+      "step": 2736
+    },
+    {
+      "epoch": 5.05,
+      "learning_rate": 7.241600000000001e-06,
+      "loss": 1.8021,
+      "step": 2737
+    },
+    {
+      "epoch": 5.05,
+      "learning_rate": 7.238400000000002e-06,
+      "loss": 1.9295,
+      "step": 2738
+    },
+    {
+      "epoch": 5.05,
+      "learning_rate": 7.2352000000000015e-06,
+      "loss": 1.6622,
+      "step": 2739
+    },
+    {
+      "epoch": 5.06,
+      "learning_rate": 7.232000000000001e-06,
+      "loss": 1.8784,
+      "step": 2740
+    },
+    {
+      "epoch": 5.06,
+      "learning_rate": 7.228800000000001e-06,
+      "loss": 1.8597,
+      "step": 2741
+    },
+    {
+      "epoch": 5.06,
+      "learning_rate": 7.225600000000002e-06,
+      "loss": 1.8988,
+      "step": 2742
+    },
+    {
+      "epoch": 5.06,
+      "learning_rate": 7.222400000000001e-06,
+      "loss": 1.8578,
+      "step": 2743
+    },
+    {
+      "epoch": 5.06,
+      "learning_rate": 7.219200000000001e-06,
+      "loss": 1.8354,
+      "step": 2744
+    },
+    {
+      "epoch": 5.06,
+      "learning_rate": 7.216000000000002e-06,
+      "loss": 1.8233,
+      "step": 2745
+    },
+    {
+      "epoch": 5.07,
+      "learning_rate": 7.212800000000001e-06,
+      "loss": 1.8446,
+      "step": 2746
+    },
+    {
+      "epoch": 5.07,
+      "learning_rate": 7.209600000000001e-06,
+      "loss": 1.799,
+      "step": 2747
+    },
+    {
+      "epoch": 5.07,
+      "learning_rate": 7.206400000000002e-06,
+      "loss": 1.8624,
+      "step": 2748
+    },
+    {
+      "epoch": 5.07,
+      "learning_rate": 7.203200000000001e-06,
+      "loss": 1.7947,
+      "step": 2749
+    },
+    {
+      "epoch": 5.07,
+      "learning_rate": 7.200000000000001e-06,
+      "loss": 1.8586,
+      "step": 2750
+    },
+    {
+      "epoch": 5.08,
+      "learning_rate": 7.196800000000001e-06,
+      "loss": 1.7553,
+      "step": 2751
+    },
+    {
+      "epoch": 5.08,
+      "learning_rate": 7.193600000000001e-06,
+      "loss": 1.7825,
+      "step": 2752
+    },
+    {
+      "epoch": 5.08,
+      "learning_rate": 7.1904000000000015e-06,
+      "loss": 1.9094,
+      "step": 2753
+    },
+    {
+      "epoch": 5.08,
+      "learning_rate": 7.187200000000001e-06,
+      "loss": 1.7661,
+      "step": 2754
+    },
+    {
+      "epoch": 5.08,
+      "learning_rate": 7.184000000000001e-06,
+      "loss": 1.8504,
+      "step": 2755
+    },
+    {
+      "epoch": 5.08,
+      "learning_rate": 7.180800000000001e-06,
+      "loss": 1.8532,
+      "step": 2756
+    },
+    {
+      "epoch": 5.09,
+      "learning_rate": 7.177600000000001e-06,
+      "loss": 1.84,
+      "step": 2757
+    },
+    {
+      "epoch": 5.09,
+      "learning_rate": 7.174400000000001e-06,
+      "loss": 1.724,
+      "step": 2758
+    },
+    {
+      "epoch": 5.09,
+      "learning_rate": 7.171200000000001e-06,
+      "loss": 1.8224,
+      "step": 2759
+    },
+    {
+      "epoch": 5.09,
+      "learning_rate": 7.1680000000000015e-06,
+      "loss": 1.8443,
+      "step": 2760
+    },
+    {
+      "epoch": 5.09,
+      "eval_loss": 2.3548214435577393,
+      "eval_runtime": 232.2922,
+      "eval_samples_per_second": 37.337,
+      "eval_steps_per_second": 0.585,
+      "step": 2760
+    },
+    {
+      "epoch": 5.09,
+      "learning_rate": 7.1648e-06,
+      "loss": 1.9022,
+      "step": 2761
+    },
+    {
+      "epoch": 5.1,
+      "learning_rate": 7.161600000000001e-06,
+      "loss": 1.95,
+      "step": 2762
+    },
+    {
+      "epoch": 5.1,
+      "learning_rate": 7.158400000000002e-06,
+      "loss": 1.8119,
+      "step": 2763
+    },
+    {
+      "epoch": 5.1,
+      "learning_rate": 7.155200000000001e-06,
+      "loss": 1.7476,
+      "step": 2764
+    },
+    {
+      "epoch": 5.1,
+      "learning_rate": 7.152000000000001e-06,
+      "loss": 1.9513,
+      "step": 2765
+    },
+    {
+      "epoch": 5.1,
+      "learning_rate": 7.148800000000001e-06,
+      "loss": 1.8619,
+      "step": 2766
+    },
+    {
+      "epoch": 5.11,
+      "learning_rate": 7.1456000000000015e-06,
+      "loss": 1.8817,
+      "step": 2767
+    },
+    {
+      "epoch": 5.11,
+      "learning_rate": 7.142400000000001e-06,
+      "loss": 1.9169,
+      "step": 2768
+    },
+    {
+      "epoch": 5.11,
+      "learning_rate": 7.139200000000001e-06,
+      "loss": 1.867,
+      "step": 2769
+    },
+    {
+      "epoch": 5.11,
+      "learning_rate": 7.136000000000002e-06,
+      "loss": 1.8255,
+      "step": 2770
+    },
+    {
+      "epoch": 5.11,
+      "learning_rate": 7.1328000000000006e-06,
+      "loss": 1.8181,
+      "step": 2771
+    },
+    {
+      "epoch": 5.11,
+      "learning_rate": 7.129600000000001e-06,
+      "loss": 1.8114,
+      "step": 2772
+    },
+    {
+      "epoch": 5.12,
+      "learning_rate": 7.126400000000002e-06,
+      "loss": 1.8054,
+      "step": 2773
+    },
+    {
+      "epoch": 5.12,
+      "learning_rate": 7.123200000000001e-06,
+      "loss": 1.782,
+      "step": 2774
+    },
+    {
+      "epoch": 5.12,
+      "learning_rate": 7.120000000000001e-06,
+      "loss": 1.8625,
+      "step": 2775
+    },
+    {
+      "epoch": 5.12,
+      "learning_rate": 7.116800000000001e-06,
+      "loss": 1.8975,
+      "step": 2776
+    },
+    {
+      "epoch": 5.12,
+      "learning_rate": 7.113600000000001e-06,
+      "loss": 1.847,
+      "step": 2777
+    },
+    {
+      "epoch": 5.13,
+      "learning_rate": 7.110400000000001e-06,
+      "loss": 1.7705,
+      "step": 2778
+    },
+    {
+      "epoch": 5.13,
+      "learning_rate": 7.107200000000001e-06,
+      "loss": 1.9381,
+      "step": 2779
+    },
+    {
+      "epoch": 5.13,
+      "learning_rate": 7.104000000000001e-06,
+      "loss": 1.8378,
+      "step": 2780
+    },
+    {
+      "epoch": 5.13,
+      "learning_rate": 7.100800000000001e-06,
+      "loss": 1.7894,
+      "step": 2781
+    },
+    {
+      "epoch": 5.13,
+      "learning_rate": 7.097600000000001e-06,
+      "loss": 1.8174,
+      "step": 2782
+    },
+    {
+      "epoch": 5.13,
+      "learning_rate": 7.094400000000001e-06,
+      "loss": 1.7869,
+      "step": 2783
+    },
+    {
+      "epoch": 5.14,
+      "learning_rate": 7.091200000000001e-06,
+      "loss": 1.7455,
+      "step": 2784
+    },
+    {
+      "epoch": 5.14,
+      "learning_rate": 7.0880000000000014e-06,
+      "loss": 1.8514,
+      "step": 2785
+    },
+    {
+      "epoch": 5.14,
+      "learning_rate": 7.084800000000002e-06,
+      "loss": 1.7468,
+      "step": 2786
+    },
+    {
+      "epoch": 5.14,
+      "learning_rate": 7.081600000000001e-06,
+      "loss": 1.8503,
+      "step": 2787
+    },
+    {
+      "epoch": 5.14,
+      "learning_rate": 7.0784000000000015e-06,
+      "loss": 1.891,
+      "step": 2788
+    },
+    {
+      "epoch": 5.15,
+      "learning_rate": 7.075200000000001e-06,
+      "loss": 1.7916,
+      "step": 2789
+    },
+    {
+      "epoch": 5.15,
+      "learning_rate": 7.072000000000001e-06,
+      "loss": 1.7369,
+      "step": 2790
+    },
+    {
+      "epoch": 5.15,
+      "eval_loss": 2.380058526992798,
+      "eval_runtime": 231.6845,
+      "eval_samples_per_second": 37.435,
+      "eval_steps_per_second": 0.587,
+      "step": 2790
+    },
+    {
+      "epoch": 5.15,
+      "learning_rate": 7.068800000000002e-06,
+      "loss": 1.8383,
+      "step": 2791
+    },
+    {
+      "epoch": 5.15,
+      "learning_rate": 7.0656000000000014e-06,
+      "loss": 1.8571,
+      "step": 2792
+    },
+    {
+      "epoch": 5.15,
+      "learning_rate": 7.062400000000001e-06,
+      "loss": 1.7824,
+      "step": 2793
+    },
+    {
+      "epoch": 5.15,
+      "learning_rate": 7.059200000000001e-06,
+      "loss": 1.9172,
+      "step": 2794
+    },
+    {
+      "epoch": 5.16,
+      "learning_rate": 7.0560000000000016e-06,
+      "loss": 1.8165,
+      "step": 2795
+    },
+    {
+      "epoch": 5.16,
+      "learning_rate": 7.052800000000001e-06,
+      "loss": 1.7883,
+      "step": 2796
+    },
+    {
+      "epoch": 5.16,
+      "learning_rate": 7.049600000000001e-06,
+      "loss": 1.8162,
+      "step": 2797
+    },
+    {
+      "epoch": 5.16,
+      "learning_rate": 7.046400000000002e-06,
+      "loss": 1.7881,
+      "step": 2798
+    },
+    {
+      "epoch": 5.16,
+      "learning_rate": 7.043200000000001e-06,
+      "loss": 1.7927,
+      "step": 2799
+    },
+    {
+      "epoch": 5.17,
+      "learning_rate": 7.040000000000001e-06,
+      "loss": 1.7914,
+      "step": 2800
+    },
+    {
+      "epoch": 5.17,
+      "learning_rate": 7.036800000000002e-06,
+      "loss": 1.7951,
+      "step": 2801
+    },
+    {
+      "epoch": 5.17,
+      "learning_rate": 7.033600000000001e-06,
+      "loss": 1.7242,
+      "step": 2802
+    },
+    {
+      "epoch": 5.17,
+      "learning_rate": 7.030400000000001e-06,
+      "loss": 1.8244,
+      "step": 2803
+    },
+    {
+      "epoch": 5.17,
+      "learning_rate": 7.027200000000001e-06,
+      "loss": 1.8932,
+      "step": 2804
+    },
+    {
+      "epoch": 5.18,
+      "learning_rate": 7.024000000000001e-06,
+      "loss": 1.9516,
+      "step": 2805
+    },
+    {
+      "epoch": 5.18,
+      "learning_rate": 7.0208000000000015e-06,
+      "loss": 1.8227,
+      "step": 2806
+    },
+    {
+      "epoch": 5.18,
+      "learning_rate": 7.017600000000001e-06,
+      "loss": 1.8471,
+      "step": 2807
+    },
+    {
+      "epoch": 5.18,
+      "learning_rate": 7.014400000000001e-06,
+      "loss": 1.8924,
+      "step": 2808
+    },
+    {
+      "epoch": 5.18,
+      "learning_rate": 7.011200000000001e-06,
+      "loss": 1.9147,
+      "step": 2809
+    },
+    {
+      "epoch": 5.18,
+      "learning_rate": 7.008000000000001e-06,
+      "loss": 1.8141,
+      "step": 2810
+    },
+    {
+      "epoch": 5.19,
+      "learning_rate": 7.004800000000001e-06,
+      "loss": 1.8605,
+      "step": 2811
+    },
+    {
+      "epoch": 5.19,
+      "learning_rate": 7.001600000000001e-06,
+      "loss": 1.9066,
+      "step": 2812
+    },
+    {
+      "epoch": 5.19,
+      "learning_rate": 6.9984000000000015e-06,
+      "loss": 1.7482,
+      "step": 2813
+    },
+    {
+      "epoch": 5.19,
+      "learning_rate": 6.995200000000001e-06,
+      "loss": 1.8231,
+      "step": 2814
+    },
+    {
+      "epoch": 5.19,
+      "learning_rate": 6.992000000000001e-06,
+      "loss": 1.8284,
+      "step": 2815
+    },
+    {
+      "epoch": 5.2,
+      "learning_rate": 6.988800000000002e-06,
+      "loss": 1.8165,
+      "step": 2816
+    },
+    {
+      "epoch": 5.2,
+      "learning_rate": 6.985600000000001e-06,
+      "loss": 1.8685,
+      "step": 2817
+    },
+    {
+      "epoch": 5.2,
+      "learning_rate": 6.982400000000001e-06,
+      "loss": 1.9147,
+      "step": 2818
+    },
+    {
+      "epoch": 5.2,
+      "learning_rate": 6.979200000000001e-06,
+      "loss": 1.8463,
+      "step": 2819
+    },
+    {
+      "epoch": 5.2,
+      "learning_rate": 6.9760000000000015e-06,
+      "loss": 1.8773,
+      "step": 2820
+    },
+    {
+      "epoch": 5.2,
+      "eval_loss": 2.3673200607299805,
+      "eval_runtime": 231.4962,
+      "eval_samples_per_second": 37.465,
+      "eval_steps_per_second": 0.587,
+      "step": 2820
+    },
+    {
+      "epoch": 5.2,
+      "learning_rate": 6.972800000000001e-06,
+      "loss": 1.7812,
+      "step": 2821
+    },
+    {
+      "epoch": 5.21,
+      "learning_rate": 6.969600000000001e-06,
+      "loss": 1.7275,
+      "step": 2822
+    },
+    {
+      "epoch": 5.21,
+      "learning_rate": 6.966400000000002e-06,
+      "loss": 1.8639,
+      "step": 2823
+    },
+    {
+      "epoch": 5.21,
+      "learning_rate": 6.9632000000000005e-06,
+      "loss": 1.796,
+      "step": 2824
+    },
+    {
+      "epoch": 5.21,
+      "learning_rate": 6.960000000000001e-06,
+      "loss": 1.8976,
+      "step": 2825
+    },
+    {
+      "epoch": 5.21,
+      "learning_rate": 6.956800000000002e-06,
+      "loss": 1.8258,
+      "step": 2826
+    },
+    {
+      "epoch": 5.22,
+      "learning_rate": 6.953600000000001e-06,
+      "loss": 1.7862,
+      "step": 2827
+    },
+    {
+      "epoch": 5.22,
+      "learning_rate": 6.950400000000001e-06,
+      "loss": 1.9464,
+      "step": 2828
+    },
+    {
+      "epoch": 5.22,
+      "learning_rate": 6.947200000000001e-06,
+      "loss": 1.6612,
+      "step": 2829
+    },
+    {
+      "epoch": 5.22,
+      "learning_rate": 6.944000000000001e-06,
+      "loss": 1.8003,
+      "step": 2830
+    },
+    {
+      "epoch": 5.22,
+      "learning_rate": 6.940800000000001e-06,
+      "loss": 1.7751,
+      "step": 2831
+    },
+    {
+      "epoch": 5.22,
+      "learning_rate": 6.937600000000001e-06,
+      "loss": 1.807,
+      "step": 2832
+    },
+    {
+      "epoch": 5.23,
+      "learning_rate": 6.934400000000001e-06,
+      "loss": 1.8316,
+      "step": 2833
+    },
+    {
+      "epoch": 5.23,
+      "learning_rate": 6.931200000000001e-06,
+      "loss": 1.9039,
+      "step": 2834
+    },
+    {
+      "epoch": 5.23,
+      "learning_rate": 6.928000000000001e-06,
+      "loss": 1.7871,
+      "step": 2835
+    },
+    {
+      "epoch": 5.23,
+      "learning_rate": 6.924800000000001e-06,
+      "loss": 1.7647,
+      "step": 2836
+    },
+    {
+      "epoch": 5.23,
+      "learning_rate": 6.921600000000001e-06,
+      "loss": 1.7434,
+      "step": 2837
+    },
+    {
+      "epoch": 5.24,
+      "learning_rate": 6.918400000000001e-06,
+      "loss": 1.9117,
+      "step": 2838
+    },
+    {
+      "epoch": 5.24,
+      "learning_rate": 6.9152e-06,
+      "loss": 1.8605,
+      "step": 2839
+    },
+    {
+      "epoch": 5.24,
+      "learning_rate": 6.912000000000001e-06,
+      "loss": 1.8344,
+      "step": 2840
+    },
+    {
+      "epoch": 5.24,
+      "learning_rate": 6.9088000000000015e-06,
+      "loss": 1.909,
+      "step": 2841
+    },
+    {
+      "epoch": 5.24,
+      "learning_rate": 6.905600000000001e-06,
+      "loss": 1.8055,
+      "step": 2842
+    },
+    {
+      "epoch": 5.25,
+      "learning_rate": 6.902400000000001e-06,
+      "loss": 1.7963,
+      "step": 2843
+    },
+    {
+      "epoch": 5.25,
+      "learning_rate": 6.899200000000002e-06,
+      "loss": 1.7781,
+      "step": 2844
+    },
+    {
+      "epoch": 5.25,
+      "learning_rate": 6.896000000000001e-06,
+      "loss": 1.8684,
+      "step": 2845
+    },
+    {
+      "epoch": 5.25,
+      "learning_rate": 6.892800000000001e-06,
+      "loss": 1.9724,
+      "step": 2846
+    },
+    {
+      "epoch": 5.25,
+      "learning_rate": 6.889600000000001e-06,
+      "loss": 1.8221,
+      "step": 2847
+    },
+    {
+      "epoch": 5.25,
+      "learning_rate": 6.8864000000000015e-06,
+      "loss": 1.8435,
+      "step": 2848
+    },
+    {
+      "epoch": 5.26,
+      "learning_rate": 6.883200000000001e-06,
+      "loss": 1.9209,
+      "step": 2849
+    },
+    {
+      "epoch": 5.26,
+      "learning_rate": 6.880000000000001e-06,
+      "loss": 1.7829,
+      "step": 2850
+    },
+    {
+      "epoch": 5.26,
+      "eval_loss": 2.3734424114227295,
+      "eval_runtime": 232.761,
+      "eval_samples_per_second": 37.261,
+      "eval_steps_per_second": 0.584,
+      "step": 2850
+    },
+    {
+      "epoch": 5.26,
+      "learning_rate": 6.876800000000002e-06,
+      "loss": 1.8336,
+      "step": 2851
+    },
+    {
+      "epoch": 5.26,
+      "learning_rate": 6.8736000000000006e-06,
+      "loss": 1.7559,
+      "step": 2852
+    },
+    {
+      "epoch": 5.26,
+      "learning_rate": 6.870400000000001e-06,
+      "loss": 1.9198,
+      "step": 2853
+    },
+    {
+      "epoch": 5.27,
+      "learning_rate": 6.867200000000002e-06,
+      "loss": 1.8121,
+      "step": 2854
+    },
+    {
+      "epoch": 5.27,
+      "learning_rate": 6.864000000000001e-06,
+      "loss": 1.821,
+      "step": 2855
+    },
+    {
+      "epoch": 5.27,
+      "learning_rate": 6.860800000000001e-06,
+      "loss": 1.8017,
+      "step": 2856
+    },
+    {
+      "epoch": 5.27,
+      "learning_rate": 6.857600000000001e-06,
+      "loss": 1.765,
+      "step": 2857
+    },
+    {
+      "epoch": 5.27,
+      "learning_rate": 6.854400000000001e-06,
+      "loss": 1.9255,
+      "step": 2858
+    },
+    {
+      "epoch": 5.27,
+      "learning_rate": 6.851200000000001e-06,
+      "loss": 1.884,
+      "step": 2859
+    },
+    {
+      "epoch": 5.28,
+      "learning_rate": 6.848000000000001e-06,
+      "loss": 1.7782,
+      "step": 2860
+    },
+    {
+      "epoch": 5.28,
+      "learning_rate": 6.844800000000001e-06,
+      "loss": 1.7807,
+      "step": 2861
+    },
+    {
+      "epoch": 5.28,
+      "learning_rate": 6.841600000000001e-06,
+      "loss": 1.7822,
+      "step": 2862
+    },
+    {
+      "epoch": 5.28,
+      "learning_rate": 6.838400000000001e-06,
+      "loss": 1.9451,
+      "step": 2863
+    },
+    {
+      "epoch": 5.28,
+      "learning_rate": 6.835200000000002e-06,
+      "loss": 1.8956,
+      "step": 2864
+    },
+    {
+      "epoch": 5.29,
+      "learning_rate": 6.832000000000001e-06,
+      "loss": 1.858,
+      "step": 2865
+    },
+    {
+      "epoch": 5.29,
+      "learning_rate": 6.8288000000000014e-06,
+      "loss": 1.8493,
+      "step": 2866
+    },
+    {
+      "epoch": 5.29,
+      "learning_rate": 6.825600000000001e-06,
+      "loss": 1.8363,
+      "step": 2867
+    },
+    {
+      "epoch": 5.29,
+      "learning_rate": 6.822400000000001e-06,
+      "loss": 1.8108,
+      "step": 2868
+    },
+    {
+      "epoch": 5.29,
+      "learning_rate": 6.8192000000000016e-06,
+      "loss": 1.8243,
+      "step": 2869
+    },
+    {
+      "epoch": 5.29,
+      "learning_rate": 6.816000000000001e-06,
+      "loss": 1.7456,
+      "step": 2870
+    },
+    {
+      "epoch": 5.3,
+      "learning_rate": 6.812800000000001e-06,
+      "loss": 1.846,
+      "step": 2871
+    },
+    {
+      "epoch": 5.3,
+      "learning_rate": 6.809600000000001e-06,
+      "loss": 1.9024,
+      "step": 2872
+    },
+    {
+      "epoch": 5.3,
+      "learning_rate": 6.8064000000000014e-06,
+      "loss": 1.8797,
+      "step": 2873
+    },
+    {
+      "epoch": 5.3,
+      "learning_rate": 6.803200000000001e-06,
+      "loss": 1.7383,
+      "step": 2874
+    },
+    {
+      "epoch": 5.3,
+      "learning_rate": 6.800000000000001e-06,
+      "loss": 1.785,
+      "step": 2875
+    },
+    {
+      "epoch": 5.31,
+      "learning_rate": 6.7968000000000016e-06,
+      "loss": 1.7989,
+      "step": 2876
+    },
+    {
+      "epoch": 5.31,
+      "learning_rate": 6.7936000000000005e-06,
+      "loss": 1.9349,
+      "step": 2877
+    },
+    {
+      "epoch": 5.31,
+      "learning_rate": 6.790400000000001e-06,
+      "loss": 1.9007,
+      "step": 2878
+    },
+    {
+      "epoch": 5.31,
+      "learning_rate": 6.787200000000002e-06,
+      "loss": 1.8467,
+      "step": 2879
+    },
+    {
+      "epoch": 5.31,
+      "learning_rate": 6.784000000000001e-06,
+      "loss": 1.7821,
+      "step": 2880
+    },
+    {
+      "epoch": 5.31,
+      "eval_loss": 2.3784682750701904,
+      "eval_runtime": 232.9619,
+      "eval_samples_per_second": 37.229,
+      "eval_steps_per_second": 0.584,
+      "step": 2880
+    },
+    {
+      "epoch": 5.32,
+      "learning_rate": 6.780800000000001e-06,
+      "loss": 1.8655,
+      "step": 2881
+    },
+    {
+      "epoch": 5.32,
+      "learning_rate": 6.777600000000001e-06,
+      "loss": 1.7835,
+      "step": 2882
+    },
+    {
+      "epoch": 5.32,
+      "learning_rate": 6.774400000000001e-06,
+      "loss": 1.888,
+      "step": 2883
+    },
+    {
+      "epoch": 5.32,
+      "learning_rate": 6.771200000000001e-06,
+      "loss": 1.6829,
+      "step": 2884
+    },
+    {
+      "epoch": 5.32,
+      "learning_rate": 6.768000000000001e-06,
+      "loss": 1.763,
+      "step": 2885
+    },
+    {
+      "epoch": 5.32,
+      "learning_rate": 6.764800000000001e-06,
+      "loss": 1.8716,
+      "step": 2886
+    },
+    {
+      "epoch": 5.33,
+      "learning_rate": 6.761600000000001e-06,
+      "loss": 1.7556,
+      "step": 2887
+    },
+    {
+      "epoch": 5.33,
+      "learning_rate": 6.758400000000001e-06,
+      "loss": 1.8822,
+      "step": 2888
+    },
+    {
+      "epoch": 5.33,
+      "learning_rate": 6.755200000000002e-06,
+      "loss": 1.9396,
+      "step": 2889
+    },
+    {
+      "epoch": 5.33,
+      "learning_rate": 6.752000000000001e-06,
+      "loss": 1.7706,
+      "step": 2890
+    },
+    {
+      "epoch": 5.33,
+      "learning_rate": 6.748800000000001e-06,
+      "loss": 1.8831,
+      "step": 2891
+    },
+    {
+      "epoch": 5.34,
+      "learning_rate": 6.745600000000001e-06,
+      "loss": 1.7318,
+      "step": 2892
+    },
+    {
+      "epoch": 5.34,
+      "learning_rate": 6.742400000000001e-06,
+      "loss": 1.9268,
+      "step": 2893
+    },
+    {
+      "epoch": 5.34,
+      "learning_rate": 6.7392000000000015e-06,
+      "loss": 1.7023,
+      "step": 2894
+    },
+    {
+      "epoch": 5.34,
+      "learning_rate": 6.736000000000001e-06,
+      "loss": 1.8401,
+      "step": 2895
+    },
+    {
+      "epoch": 5.34,
+      "learning_rate": 6.732800000000001e-06,
+      "loss": 1.8996,
+      "step": 2896
+    },
+    {
+      "epoch": 5.34,
+      "learning_rate": 6.729600000000001e-06,
+      "loss": 1.8096,
+      "step": 2897
+    },
+    {
+      "epoch": 5.35,
+      "learning_rate": 6.726400000000001e-06,
+      "loss": 1.9035,
+      "step": 2898
+    },
+    {
+      "epoch": 5.35,
+      "learning_rate": 6.723200000000001e-06,
+      "loss": 1.8658,
+      "step": 2899
+    },
+    {
+      "epoch": 5.35,
+      "learning_rate": 6.720000000000001e-06,
+      "loss": 1.9,
+      "step": 2900
+    },
+    {
+      "epoch": 5.35,
+      "learning_rate": 6.7168000000000015e-06,
+      "loss": 1.8286,
+      "step": 2901
+    },
+    {
+      "epoch": 5.35,
+      "learning_rate": 6.7136e-06,
+      "loss": 1.7465,
+      "step": 2902
+    },
+    {
+      "epoch": 5.36,
+      "learning_rate": 6.710400000000001e-06,
+      "loss": 1.9444,
+      "step": 2903
+    },
+    {
+      "epoch": 5.36,
+      "learning_rate": 6.707200000000002e-06,
+      "loss": 1.8672,
+      "step": 2904
+    },
+    {
+      "epoch": 5.36,
+      "learning_rate": 6.7040000000000005e-06,
+      "loss": 1.8351,
+      "step": 2905
+    },
+    {
+      "epoch": 5.36,
+      "learning_rate": 6.700800000000001e-06,
+      "loss": 1.8316,
+      "step": 2906
+    },
+    {
+      "epoch": 5.36,
+      "learning_rate": 6.697600000000002e-06,
+      "loss": 1.9979,
+      "step": 2907
+    },
+    {
+      "epoch": 5.36,
+      "learning_rate": 6.694400000000001e-06,
+      "loss": 1.8276,
+      "step": 2908
+    },
+    {
+      "epoch": 5.37,
+      "learning_rate": 6.691200000000001e-06,
+      "loss": 1.8336,
+      "step": 2909
+    },
+    {
+      "epoch": 5.37,
+      "learning_rate": 6.688000000000001e-06,
+      "loss": 1.758,
+      "step": 2910
+    },
+    {
+      "epoch": 5.37,
+      "eval_loss": 2.375408411026001,
+      "eval_runtime": 231.724,
+      "eval_samples_per_second": 37.428,
+      "eval_steps_per_second": 0.587,
+      "step": 2910
+    },
+    {
+      "epoch": 5.37,
+      "learning_rate": 6.684800000000001e-06,
+      "loss": 1.9416,
+      "step": 2911
+    },
+    {
+      "epoch": 5.37,
+      "learning_rate": 6.681600000000001e-06,
+      "loss": 1.7618,
+      "step": 2912
+    },
+    {
+      "epoch": 5.37,
+      "learning_rate": 6.678400000000001e-06,
+      "loss": 1.7976,
+      "step": 2913
+    },
+    {
+      "epoch": 5.38,
+      "learning_rate": 6.675200000000002e-06,
+      "loss": 1.7748,
+      "step": 2914
+    },
+    {
+      "epoch": 5.38,
+      "learning_rate": 6.672000000000001e-06,
+      "loss": 1.8782,
+      "step": 2915
+    },
+    {
+      "epoch": 5.38,
+      "learning_rate": 6.668800000000001e-06,
+      "loss": 1.8283,
+      "step": 2916
+    },
+    {
+      "epoch": 5.38,
+      "learning_rate": 6.665600000000002e-06,
+      "loss": 1.8363,
+      "step": 2917
+    },
+    {
+      "epoch": 5.38,
+      "learning_rate": 6.662400000000001e-06,
+      "loss": 1.8609,
+      "step": 2918
+    },
+    {
+      "epoch": 5.39,
+      "learning_rate": 6.659200000000001e-06,
+      "loss": 1.796,
+      "step": 2919
+    },
+    {
+      "epoch": 5.39,
+      "learning_rate": 6.656000000000001e-06,
+      "loss": 1.7557,
+      "step": 2920
+    },
+    {
+      "epoch": 5.39,
+      "learning_rate": 6.652800000000001e-06,
+      "loss": 1.8414,
+      "step": 2921
+    },
+    {
+      "epoch": 5.39,
+      "learning_rate": 6.6496000000000015e-06,
+      "loss": 1.8527,
+      "step": 2922
+    },
+    {
+      "epoch": 5.39,
+      "learning_rate": 6.646400000000001e-06,
+      "loss": 1.8348,
+      "step": 2923
+    },
+    {
+      "epoch": 5.39,
+      "learning_rate": 6.643200000000001e-06,
+      "loss": 1.7901,
+      "step": 2924
+    },
+    {
+      "epoch": 5.4,
+      "learning_rate": 6.640000000000001e-06,
+      "loss": 1.9481,
+      "step": 2925
+    },
+    {
+      "epoch": 5.4,
+      "learning_rate": 6.636800000000001e-06,
+      "loss": 1.8189,
+      "step": 2926
+    },
+    {
+      "epoch": 5.4,
+      "learning_rate": 6.633600000000001e-06,
+      "loss": 1.8921,
+      "step": 2927
+    },
+    {
+      "epoch": 5.4,
+      "learning_rate": 6.630400000000001e-06,
+      "loss": 1.8609,
+      "step": 2928
+    },
+    {
+      "epoch": 5.4,
+      "learning_rate": 6.6272000000000015e-06,
+      "loss": 1.8025,
+      "step": 2929
+    },
+    {
+      "epoch": 5.41,
+      "learning_rate": 6.6240000000000004e-06,
+      "loss": 1.7277,
+      "step": 2930
+    },
+    {
+      "epoch": 5.41,
+      "learning_rate": 6.620800000000001e-06,
+      "loss": 1.8821,
+      "step": 2931
+    },
+    {
+      "epoch": 5.41,
+      "learning_rate": 6.617600000000002e-06,
+      "loss": 1.9055,
+      "step": 2932
+    },
+    {
+      "epoch": 5.41,
+      "learning_rate": 6.6144000000000006e-06,
+      "loss": 1.8315,
+      "step": 2933
+    },
+    {
+      "epoch": 5.41,
+      "learning_rate": 6.611200000000001e-06,
+      "loss": 1.8693,
+      "step": 2934
+    },
+    {
+      "epoch": 5.41,
+      "learning_rate": 6.608000000000001e-06,
+      "loss": 1.8056,
+      "step": 2935
+    },
+    {
+      "epoch": 5.42,
+      "learning_rate": 6.604800000000001e-06,
+      "loss": 1.9654,
+      "step": 2936
+    },
+    {
+      "epoch": 5.42,
+      "learning_rate": 6.601600000000001e-06,
+      "loss": 1.6624,
+      "step": 2937
+    },
+    {
+      "epoch": 5.42,
+      "learning_rate": 6.598400000000001e-06,
+      "loss": 1.7861,
+      "step": 2938
+    },
+    {
+      "epoch": 5.42,
+      "learning_rate": 6.595200000000002e-06,
+      "loss": 1.8085,
+      "step": 2939
+    },
+    {
+      "epoch": 5.42,
+      "learning_rate": 6.592000000000001e-06,
+      "loss": 1.8502,
+      "step": 2940
+    },
+    {
+      "epoch": 5.42,
+      "eval_loss": 2.3807802200317383,
+      "eval_runtime": 232.2261,
+      "eval_samples_per_second": 37.347,
+      "eval_steps_per_second": 0.586,
+      "step": 2940
+    },
+    {
+      "epoch": 5.43,
+      "learning_rate": 6.588800000000001e-06,
+      "loss": 1.876,
+      "step": 2941
+    },
+    {
+      "epoch": 5.43,
+      "learning_rate": 6.585600000000002e-06,
+      "loss": 1.7999,
+      "step": 2942
+    },
+    {
+      "epoch": 5.43,
+      "learning_rate": 6.582400000000001e-06,
+      "loss": 1.894,
+      "step": 2943
+    },
+    {
+      "epoch": 5.43,
+      "learning_rate": 6.579200000000001e-06,
+      "loss": 1.8542,
+      "step": 2944
+    },
+    {
+      "epoch": 5.43,
+      "learning_rate": 6.576000000000001e-06,
+      "loss": 1.8653,
+      "step": 2945
+    },
+    {
+      "epoch": 5.44,
+      "learning_rate": 6.572800000000001e-06,
+      "loss": 1.806,
+      "step": 2946
+    },
+    {
+      "epoch": 5.44,
+      "learning_rate": 6.5696000000000014e-06,
+      "loss": 1.8023,
+      "step": 2947
+    },
+    {
+      "epoch": 5.44,
+      "learning_rate": 6.566400000000001e-06,
+      "loss": 1.857,
+      "step": 2948
+    },
+    {
+      "epoch": 5.44,
+      "learning_rate": 6.563200000000001e-06,
+      "loss": 1.8692,
+      "step": 2949
+    },
+    {
+      "epoch": 5.44,
+      "learning_rate": 6.560000000000001e-06,
+      "loss": 1.7441,
+      "step": 2950
+    },
+    {
+      "epoch": 5.44,
+      "learning_rate": 6.556800000000001e-06,
+      "loss": 2.0052,
+      "step": 2951
+    },
+    {
+      "epoch": 5.45,
+      "learning_rate": 6.553600000000001e-06,
+      "loss": 1.8395,
+      "step": 2952
+    },
+    {
+      "epoch": 5.45,
+      "learning_rate": 6.550400000000001e-06,
+      "loss": 1.779,
+      "step": 2953
+    },
+    {
+      "epoch": 5.45,
+      "learning_rate": 6.5472000000000015e-06,
+      "loss": 1.8024,
+      "step": 2954
+    },
+    {
+      "epoch": 5.45,
+      "learning_rate": 6.544e-06,
+      "loss": 1.9311,
+      "step": 2955
+    },
+    {
+      "epoch": 5.45,
+      "learning_rate": 6.540800000000001e-06,
+      "loss": 1.8333,
+      "step": 2956
+    },
+    {
+      "epoch": 5.46,
+      "learning_rate": 6.537600000000002e-06,
+      "loss": 1.7378,
+      "step": 2957
+    },
+    {
+      "epoch": 5.46,
+      "learning_rate": 6.5344000000000005e-06,
+      "loss": 1.8392,
+      "step": 2958
+    },
+    {
+      "epoch": 5.46,
+      "learning_rate": 6.531200000000001e-06,
+      "loss": 1.9705,
+      "step": 2959
+    },
+    {
+      "epoch": 5.46,
+      "learning_rate": 6.528000000000001e-06,
+      "loss": 1.8383,
+      "step": 2960
+    },
+    {
+      "epoch": 5.46,
+      "learning_rate": 6.524800000000001e-06,
+      "loss": 1.8056,
+      "step": 2961
+    },
+    {
+      "epoch": 5.46,
+      "learning_rate": 6.521600000000001e-06,
+      "loss": 1.894,
+      "step": 2962
+    },
+    {
+      "epoch": 5.47,
+      "learning_rate": 6.518400000000001e-06,
+      "loss": 1.7552,
+      "step": 2963
+    },
+    {
+      "epoch": 5.47,
+      "learning_rate": 6.515200000000002e-06,
+      "loss": 1.8638,
+      "step": 2964
+    },
+    {
+      "epoch": 5.47,
+      "learning_rate": 6.5120000000000005e-06,
+      "loss": 1.7236,
+      "step": 2965
+    },
+    {
+      "epoch": 5.47,
+      "learning_rate": 6.508800000000001e-06,
+      "loss": 1.7496,
+      "step": 2966
+    },
+    {
+      "epoch": 5.47,
+      "learning_rate": 6.505600000000002e-06,
+      "loss": 1.7965,
+      "step": 2967
+    },
+    {
+      "epoch": 5.48,
+      "learning_rate": 6.502400000000001e-06,
+      "loss": 1.9219,
+      "step": 2968
+    },
+    {
+      "epoch": 5.48,
+      "learning_rate": 6.499200000000001e-06,
+      "loss": 1.833,
+      "step": 2969
+    },
+    {
+      "epoch": 5.48,
+      "learning_rate": 6.496000000000002e-06,
+      "loss": 1.8251,
+      "step": 2970
+    },
+    {
+      "epoch": 5.48,
+      "eval_loss": 2.3719348907470703,
+      "eval_runtime": 232.8651,
+      "eval_samples_per_second": 37.245,
+      "eval_steps_per_second": 0.584,
+      "step": 2970
+    }
+  ],
+  "max_steps": 5000,
+  "num_train_epochs": 10,
+  "total_flos": 4.4665380463312896e+17,
+  "trial_name": null,
+  "trial_params": null
+}