diff --git "a/checkpoint-1182/trainer_state.json" "b/checkpoint-1182/trainer_state.json"
new file mode 100644--- /dev/null
+++ "b/checkpoint-1182/trainer_state.json"
@@ -0,0 +1,7152 @@
+{
+  "best_metric": null,
+  "best_model_checkpoint": null,
+  "epoch": 0.9995771670190274,
+  "eval_steps": 236,
+  "global_step": 1182,
+  "is_hyper_param_search": false,
+  "is_local_process_zero": true,
+  "is_world_process_zero": true,
+  "log_history": [
+    {
+      "epoch": 0.0,
+      "learning_rate": 2.8169014084507045e-08,
+      "loss": 1.3418,
+      "step": 1
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 5.633802816901409e-08,
+      "loss": 1.6801,
+      "step": 2
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 8.450704225352113e-08,
+      "loss": 1.3628,
+      "step": 3
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 1.1267605633802818e-07,
+      "loss": 1.3222,
+      "step": 4
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 1.4084507042253522e-07,
+      "loss": 1.519,
+      "step": 5
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 1.6901408450704225e-07,
+      "loss": 1.5033,
+      "step": 6
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 1.9718309859154932e-07,
+      "loss": 1.2227,
+      "step": 7
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 2.2535211267605636e-07,
+      "loss": 1.5744,
+      "step": 8
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 2.535211267605634e-07,
+      "loss": 1.1132,
+      "step": 9
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 2.8169014084507043e-07,
+      "loss": 1.2757,
+      "step": 10
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 3.0985915492957747e-07,
+      "loss": 1.6722,
+      "step": 11
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 3.380281690140845e-07,
+      "loss": 1.5187,
+      "step": 12
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 3.661971830985916e-07,
+      "loss": 1.2149,
+      "step": 13
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 3.9436619718309864e-07,
+      "loss": 1.9633,
+      "step": 14
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.225352112676057e-07,
+      "loss": 1.5469,
+      "step": 15
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.507042253521127e-07,
+      "loss": 1.2916,
+      "step": 16
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.788732394366198e-07,
+      "loss": 1.3984,
+      "step": 17
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 5.070422535211268e-07,
+      "loss": 1.3255,
+      "step": 18
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 5.352112676056338e-07,
+      "loss": 1.3907,
+      "step": 19
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 5.633802816901409e-07,
+      "loss": 1.3508,
+      "step": 20
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 5.915492957746479e-07,
+      "loss": 1.3875,
+      "step": 21
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 6.197183098591549e-07,
+      "loss": 1.2283,
+      "step": 22
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 6.47887323943662e-07,
+      "loss": 1.432,
+      "step": 23
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 6.76056338028169e-07,
+      "loss": 1.4333,
+      "step": 24
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 7.042253521126762e-07,
+      "loss": 1.4205,
+      "step": 25
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 7.323943661971832e-07,
+      "loss": 1.2151,
+      "step": 26
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 7.605633802816901e-07,
+      "loss": 1.3993,
+      "step": 27
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 7.887323943661973e-07,
+      "loss": 1.7116,
+      "step": 28
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 8.169014084507043e-07,
+      "loss": 1.6577,
+      "step": 29
+    },
+    {
+      "epoch": 0.03,
+      "learning_rate": 8.450704225352114e-07,
+      "loss": 1.439,
+      "step": 30
+    },
+    {
+      "epoch": 0.03,
+      "learning_rate": 8.732394366197183e-07,
+      "loss": 1.5013,
+      "step": 31
+    },
+    {
+      "epoch": 0.03,
+      "learning_rate": 9.014084507042254e-07,
+      "loss": 1.514,
+      "step": 32
+    },
+    {
+      "epoch": 0.03,
+      "learning_rate": 9.295774647887325e-07,
+      "loss": 1.5555,
+      "step": 33
+    },
+    {
+      "epoch": 0.03,
+      "learning_rate": 9.577464788732395e-07,
+      "loss": 1.6068,
+      "step": 34
+    },
+    {
+      "epoch": 0.03,
+      "learning_rate": 9.859154929577465e-07,
+      "loss": 1.3398,
+      "step": 35
+    },
+    {
+      "epoch": 0.03,
+      "learning_rate": 1.0140845070422536e-06,
+      "loss": 1.4168,
+      "step": 36
+    },
+    {
+      "epoch": 0.03,
+      "learning_rate": 1.0422535211267606e-06,
+      "loss": 1.5454,
+      "step": 37
+    },
+    {
+      "epoch": 0.03,
+      "learning_rate": 1.0704225352112677e-06,
+      "loss": 1.4684,
+      "step": 38
+    },
+    {
+      "epoch": 0.03,
+      "learning_rate": 1.098591549295775e-06,
+      "loss": 1.3311,
+      "step": 39
+    },
+    {
+      "epoch": 0.03,
+      "learning_rate": 1.1267605633802817e-06,
+      "loss": 1.2699,
+      "step": 40
+    },
+    {
+      "epoch": 0.03,
+      "learning_rate": 1.1549295774647888e-06,
+      "loss": 1.4445,
+      "step": 41
+    },
+    {
+      "epoch": 0.04,
+      "learning_rate": 1.1830985915492958e-06,
+      "loss": 1.221,
+      "step": 42
+    },
+    {
+      "epoch": 0.04,
+      "learning_rate": 1.211267605633803e-06,
+      "loss": 1.4705,
+      "step": 43
+    },
+    {
+      "epoch": 0.04,
+      "learning_rate": 1.2394366197183099e-06,
+      "loss": 1.4435,
+      "step": 44
+    },
+    {
+      "epoch": 0.04,
+      "learning_rate": 1.267605633802817e-06,
+      "loss": 1.3422,
+      "step": 45
+    },
+    {
+      "epoch": 0.04,
+      "learning_rate": 1.295774647887324e-06,
+      "loss": 1.1821,
+      "step": 46
+    },
+    {
+      "epoch": 0.04,
+      "learning_rate": 1.323943661971831e-06,
+      "loss": 1.2586,
+      "step": 47
+    },
+    {
+      "epoch": 0.04,
+      "learning_rate": 1.352112676056338e-06,
+      "loss": 1.5441,
+      "step": 48
+    },
+    {
+      "epoch": 0.04,
+      "learning_rate": 1.3802816901408453e-06,
+      "loss": 1.2419,
+      "step": 49
+    },
+    {
+      "epoch": 0.04,
+      "learning_rate": 1.4084507042253523e-06,
+      "loss": 1.3732,
+      "step": 50
+    },
+    {
+      "epoch": 0.04,
+      "learning_rate": 1.4366197183098594e-06,
+      "loss": 1.3104,
+      "step": 51
+    },
+    {
+      "epoch": 0.04,
+      "learning_rate": 1.4647887323943664e-06,
+      "loss": 1.2385,
+      "step": 52
+    },
+    {
+      "epoch": 0.04,
+      "learning_rate": 1.4929577464788732e-06,
+      "loss": 1.4739,
+      "step": 53
+    },
+    {
+      "epoch": 0.05,
+      "learning_rate": 1.5211267605633803e-06,
+      "loss": 1.1976,
+      "step": 54
+    },
+    {
+      "epoch": 0.05,
+      "learning_rate": 1.5492957746478873e-06,
+      "loss": 1.3504,
+      "step": 55
+    },
+    {
+      "epoch": 0.05,
+      "learning_rate": 1.5774647887323946e-06,
+      "loss": 1.6489,
+      "step": 56
+    },
+    {
+      "epoch": 0.05,
+      "learning_rate": 1.6056338028169016e-06,
+      "loss": 1.2824,
+      "step": 57
+    },
+    {
+      "epoch": 0.05,
+      "learning_rate": 1.6338028169014086e-06,
+      "loss": 1.3271,
+      "step": 58
+    },
+    {
+      "epoch": 0.05,
+      "learning_rate": 1.6619718309859157e-06,
+      "loss": 1.2428,
+      "step": 59
+    },
+    {
+      "epoch": 0.05,
+      "learning_rate": 1.6901408450704227e-06,
+      "loss": 1.2441,
+      "step": 60
+    },
+    {
+      "epoch": 0.05,
+      "learning_rate": 1.7183098591549297e-06,
+      "loss": 1.3945,
+      "step": 61
+    },
+    {
+      "epoch": 0.05,
+      "learning_rate": 1.7464788732394366e-06,
+      "loss": 1.4281,
+      "step": 62
+    },
+    {
+      "epoch": 0.05,
+      "learning_rate": 1.774647887323944e-06,
+      "loss": 1.3462,
+      "step": 63
+    },
+    {
+      "epoch": 0.05,
+      "learning_rate": 1.8028169014084509e-06,
+      "loss": 1.3179,
+      "step": 64
+    },
+    {
+      "epoch": 0.05,
+      "learning_rate": 1.8309859154929579e-06,
+      "loss": 1.3657,
+      "step": 65
+    },
+    {
+      "epoch": 0.06,
+      "learning_rate": 1.859154929577465e-06,
+      "loss": 1.4192,
+      "step": 66
+    },
+    {
+      "epoch": 0.06,
+      "learning_rate": 1.887323943661972e-06,
+      "loss": 1.258,
+      "step": 67
+    },
+    {
+      "epoch": 0.06,
+      "learning_rate": 1.915492957746479e-06,
+      "loss": 1.3094,
+      "step": 68
+    },
+    {
+      "epoch": 0.06,
+      "learning_rate": 1.943661971830986e-06,
+      "loss": 1.2587,
+      "step": 69
+    },
+    {
+      "epoch": 0.06,
+      "learning_rate": 1.971830985915493e-06,
+      "loss": 1.279,
+      "step": 70
+    },
+    {
+      "epoch": 0.06,
+      "learning_rate": 2.0000000000000003e-06,
+      "loss": 1.2552,
+      "step": 71
+    },
+    {
+      "epoch": 0.06,
+      "learning_rate": 2.028169014084507e-06,
+      "loss": 1.2978,
+      "step": 72
+    },
+    {
+      "epoch": 0.06,
+      "learning_rate": 2.0563380281690144e-06,
+      "loss": 1.333,
+      "step": 73
+    },
+    {
+      "epoch": 0.06,
+      "learning_rate": 2.0845070422535212e-06,
+      "loss": 1.2979,
+      "step": 74
+    },
+    {
+      "epoch": 0.06,
+      "learning_rate": 2.1126760563380285e-06,
+      "loss": 1.1694,
+      "step": 75
+    },
+    {
+      "epoch": 0.06,
+      "learning_rate": 2.1408450704225353e-06,
+      "loss": 1.487,
+      "step": 76
+    },
+    {
+      "epoch": 0.07,
+      "learning_rate": 2.169014084507042e-06,
+      "loss": 1.1773,
+      "step": 77
+    },
+    {
+      "epoch": 0.07,
+      "learning_rate": 2.19718309859155e-06,
+      "loss": 1.1597,
+      "step": 78
+    },
+    {
+      "epoch": 0.07,
+      "learning_rate": 2.2253521126760566e-06,
+      "loss": 1.2153,
+      "step": 79
+    },
+    {
+      "epoch": 0.07,
+      "learning_rate": 2.2535211267605635e-06,
+      "loss": 1.4597,
+      "step": 80
+    },
+    {
+      "epoch": 0.07,
+      "learning_rate": 2.2816901408450707e-06,
+      "loss": 1.2858,
+      "step": 81
+    },
+    {
+      "epoch": 0.07,
+      "learning_rate": 2.3098591549295775e-06,
+      "loss": 1.2605,
+      "step": 82
+    },
+    {
+      "epoch": 0.07,
+      "learning_rate": 2.338028169014085e-06,
+      "loss": 1.2524,
+      "step": 83
+    },
+    {
+      "epoch": 0.07,
+      "learning_rate": 2.3661971830985916e-06,
+      "loss": 1.2135,
+      "step": 84
+    },
+    {
+      "epoch": 0.07,
+      "learning_rate": 2.3943661971830984e-06,
+      "loss": 1.2342,
+      "step": 85
+    },
+    {
+      "epoch": 0.07,
+      "learning_rate": 2.422535211267606e-06,
+      "loss": 1.2146,
+      "step": 86
+    },
+    {
+      "epoch": 0.07,
+      "learning_rate": 2.450704225352113e-06,
+      "loss": 1.2575,
+      "step": 87
+    },
+    {
+      "epoch": 0.07,
+      "learning_rate": 2.4788732394366198e-06,
+      "loss": 1.3696,
+      "step": 88
+    },
+    {
+      "epoch": 0.08,
+      "learning_rate": 2.507042253521127e-06,
+      "loss": 1.221,
+      "step": 89
+    },
+    {
+      "epoch": 0.08,
+      "learning_rate": 2.535211267605634e-06,
+      "loss": 1.2187,
+      "step": 90
+    },
+    {
+      "epoch": 0.08,
+      "learning_rate": 2.563380281690141e-06,
+      "loss": 1.3197,
+      "step": 91
+    },
+    {
+      "epoch": 0.08,
+      "learning_rate": 2.591549295774648e-06,
+      "loss": 1.2436,
+      "step": 92
+    },
+    {
+      "epoch": 0.08,
+      "learning_rate": 2.619718309859155e-06,
+      "loss": 1.2079,
+      "step": 93
+    },
+    {
+      "epoch": 0.08,
+      "learning_rate": 2.647887323943662e-06,
+      "loss": 1.1906,
+      "step": 94
+    },
+    {
+      "epoch": 0.08,
+      "learning_rate": 2.676056338028169e-06,
+      "loss": 1.2254,
+      "step": 95
+    },
+    {
+      "epoch": 0.08,
+      "learning_rate": 2.704225352112676e-06,
+      "loss": 1.1684,
+      "step": 96
+    },
+    {
+      "epoch": 0.08,
+      "learning_rate": 2.7323943661971837e-06,
+      "loss": 1.2176,
+      "step": 97
+    },
+    {
+      "epoch": 0.08,
+      "learning_rate": 2.7605633802816906e-06,
+      "loss": 1.2144,
+      "step": 98
+    },
+    {
+      "epoch": 0.08,
+      "learning_rate": 2.7887323943661974e-06,
+      "loss": 1.1904,
+      "step": 99
+    },
+    {
+      "epoch": 0.08,
+      "learning_rate": 2.8169014084507046e-06,
+      "loss": 1.4412,
+      "step": 100
+    },
+    {
+      "epoch": 0.09,
+      "learning_rate": 2.8450704225352115e-06,
+      "loss": 1.403,
+      "step": 101
+    },
+    {
+      "epoch": 0.09,
+      "learning_rate": 2.8732394366197187e-06,
+      "loss": 1.2991,
+      "step": 102
+    },
+    {
+      "epoch": 0.09,
+      "learning_rate": 2.9014084507042255e-06,
+      "loss": 1.177,
+      "step": 103
+    },
+    {
+      "epoch": 0.09,
+      "learning_rate": 2.929577464788733e-06,
+      "loss": 1.2982,
+      "step": 104
+    },
+    {
+      "epoch": 0.09,
+      "learning_rate": 2.9577464788732396e-06,
+      "loss": 1.1565,
+      "step": 105
+    },
+    {
+      "epoch": 0.09,
+      "learning_rate": 2.9859154929577465e-06,
+      "loss": 1.3319,
+      "step": 106
+    },
+    {
+      "epoch": 0.09,
+      "learning_rate": 3.0140845070422537e-06,
+      "loss": 1.4427,
+      "step": 107
+    },
+    {
+      "epoch": 0.09,
+      "learning_rate": 3.0422535211267605e-06,
+      "loss": 1.2611,
+      "step": 108
+    },
+    {
+      "epoch": 0.09,
+      "learning_rate": 3.0704225352112678e-06,
+      "loss": 1.1248,
+      "step": 109
+    },
+    {
+      "epoch": 0.09,
+      "learning_rate": 3.0985915492957746e-06,
+      "loss": 1.2654,
+      "step": 110
+    },
+    {
+      "epoch": 0.09,
+      "learning_rate": 3.1267605633802823e-06,
+      "loss": 1.1998,
+      "step": 111
+    },
+    {
+      "epoch": 0.09,
+      "learning_rate": 3.154929577464789e-06,
+      "loss": 1.4276,
+      "step": 112
+    },
+    {
+      "epoch": 0.1,
+      "learning_rate": 3.1830985915492964e-06,
+      "loss": 1.2273,
+      "step": 113
+    },
+    {
+      "epoch": 0.1,
+      "learning_rate": 3.211267605633803e-06,
+      "loss": 1.1384,
+      "step": 114
+    },
+    {
+      "epoch": 0.1,
+      "learning_rate": 3.2394366197183104e-06,
+      "loss": 1.2032,
+      "step": 115
+    },
+    {
+      "epoch": 0.1,
+      "learning_rate": 3.2676056338028173e-06,
+      "loss": 1.0634,
+      "step": 116
+    },
+    {
+      "epoch": 0.1,
+      "learning_rate": 3.295774647887324e-06,
+      "loss": 1.2516,
+      "step": 117
+    },
+    {
+      "epoch": 0.1,
+      "learning_rate": 3.3239436619718313e-06,
+      "loss": 1.2097,
+      "step": 118
+    },
+    {
+      "epoch": 0.1,
+      "learning_rate": 3.352112676056338e-06,
+      "loss": 1.2057,
+      "step": 119
+    },
+    {
+      "epoch": 0.1,
+      "learning_rate": 3.3802816901408454e-06,
+      "loss": 1.0051,
+      "step": 120
+    },
+    {
+      "epoch": 0.1,
+      "learning_rate": 3.4084507042253522e-06,
+      "loss": 1.3015,
+      "step": 121
+    },
+    {
+      "epoch": 0.1,
+      "learning_rate": 3.4366197183098595e-06,
+      "loss": 1.1629,
+      "step": 122
+    },
+    {
+      "epoch": 0.1,
+      "learning_rate": 3.4647887323943663e-06,
+      "loss": 1.0463,
+      "step": 123
+    },
+    {
+      "epoch": 0.1,
+      "learning_rate": 3.492957746478873e-06,
+      "loss": 1.4034,
+      "step": 124
+    },
+    {
+      "epoch": 0.11,
+      "learning_rate": 3.5211267605633804e-06,
+      "loss": 1.3264,
+      "step": 125
+    },
+    {
+      "epoch": 0.11,
+      "learning_rate": 3.549295774647888e-06,
+      "loss": 1.129,
+      "step": 126
+    },
+    {
+      "epoch": 0.11,
+      "learning_rate": 3.577464788732395e-06,
+      "loss": 1.3082,
+      "step": 127
+    },
+    {
+      "epoch": 0.11,
+      "learning_rate": 3.6056338028169017e-06,
+      "loss": 1.2496,
+      "step": 128
+    },
+    {
+      "epoch": 0.11,
+      "learning_rate": 3.633802816901409e-06,
+      "loss": 1.35,
+      "step": 129
+    },
+    {
+      "epoch": 0.11,
+      "learning_rate": 3.6619718309859158e-06,
+      "loss": 1.1036,
+      "step": 130
+    },
+    {
+      "epoch": 0.11,
+      "learning_rate": 3.690140845070423e-06,
+      "loss": 1.1417,
+      "step": 131
+    },
+    {
+      "epoch": 0.11,
+      "learning_rate": 3.71830985915493e-06,
+      "loss": 1.2943,
+      "step": 132
+    },
+    {
+      "epoch": 0.11,
+      "learning_rate": 3.746478873239437e-06,
+      "loss": 1.3206,
+      "step": 133
+    },
+    {
+      "epoch": 0.11,
+      "learning_rate": 3.774647887323944e-06,
+      "loss": 1.1397,
+      "step": 134
+    },
+    {
+      "epoch": 0.11,
+      "learning_rate": 3.8028169014084508e-06,
+      "loss": 1.2328,
+      "step": 135
+    },
+    {
+      "epoch": 0.12,
+      "learning_rate": 3.830985915492958e-06,
+      "loss": 1.3558,
+      "step": 136
+    },
+    {
+      "epoch": 0.12,
+      "learning_rate": 3.859154929577465e-06,
+      "loss": 1.1015,
+      "step": 137
+    },
+    {
+      "epoch": 0.12,
+      "learning_rate": 3.887323943661972e-06,
+      "loss": 1.2189,
+      "step": 138
+    },
+    {
+      "epoch": 0.12,
+      "learning_rate": 3.915492957746479e-06,
+      "loss": 1.3142,
+      "step": 139
+    },
+    {
+      "epoch": 0.12,
+      "learning_rate": 3.943661971830986e-06,
+      "loss": 1.2039,
+      "step": 140
+    },
+    {
+      "epoch": 0.12,
+      "learning_rate": 3.971830985915493e-06,
+      "loss": 1.0501,
+      "step": 141
+    },
+    {
+      "epoch": 0.12,
+      "learning_rate": 4.000000000000001e-06,
+      "loss": 1.3026,
+      "step": 142
+    },
+    {
+      "epoch": 0.12,
+      "learning_rate": 4.028169014084508e-06,
+      "loss": 1.1499,
+      "step": 143
+    },
+    {
+      "epoch": 0.12,
+      "learning_rate": 4.056338028169014e-06,
+      "loss": 1.3819,
+      "step": 144
+    },
+    {
+      "epoch": 0.12,
+      "learning_rate": 4.0845070422535216e-06,
+      "loss": 1.2836,
+      "step": 145
+    },
+    {
+      "epoch": 0.12,
+      "learning_rate": 4.112676056338029e-06,
+      "loss": 1.1178,
+      "step": 146
+    },
+    {
+      "epoch": 0.12,
+      "learning_rate": 4.140845070422535e-06,
+      "loss": 1.1516,
+      "step": 147
+    },
+    {
+      "epoch": 0.13,
+      "learning_rate": 4.1690140845070425e-06,
+      "loss": 1.1855,
+      "step": 148
+    },
+    {
+      "epoch": 0.13,
+      "learning_rate": 4.19718309859155e-06,
+      "loss": 1.1717,
+      "step": 149
+    },
+    {
+      "epoch": 0.13,
+      "learning_rate": 4.225352112676057e-06,
+      "loss": 1.3517,
+      "step": 150
+    },
+    {
+      "epoch": 0.13,
+      "learning_rate": 4.253521126760563e-06,
+      "loss": 1.1338,
+      "step": 151
+    },
+    {
+      "epoch": 0.13,
+      "learning_rate": 4.281690140845071e-06,
+      "loss": 1.1674,
+      "step": 152
+    },
+    {
+      "epoch": 0.13,
+      "learning_rate": 4.309859154929578e-06,
+      "loss": 1.3618,
+      "step": 153
+    },
+    {
+      "epoch": 0.13,
+      "learning_rate": 4.338028169014084e-06,
+      "loss": 1.1007,
+      "step": 154
+    },
+    {
+      "epoch": 0.13,
+      "learning_rate": 4.3661971830985915e-06,
+      "loss": 1.1097,
+      "step": 155
+    },
+    {
+      "epoch": 0.13,
+      "learning_rate": 4.3943661971831e-06,
+      "loss": 1.1588,
+      "step": 156
+    },
+    {
+      "epoch": 0.13,
+      "learning_rate": 4.422535211267606e-06,
+      "loss": 1.0512,
+      "step": 157
+    },
+    {
+      "epoch": 0.13,
+      "learning_rate": 4.450704225352113e-06,
+      "loss": 1.2985,
+      "step": 158
+    },
+    {
+      "epoch": 0.13,
+      "learning_rate": 4.4788732394366205e-06,
+      "loss": 1.2587,
+      "step": 159
+    },
+    {
+      "epoch": 0.14,
+      "learning_rate": 4.507042253521127e-06,
+      "loss": 1.4696,
+      "step": 160
+    },
+    {
+      "epoch": 0.14,
+      "learning_rate": 4.535211267605634e-06,
+      "loss": 1.2118,
+      "step": 161
+    },
+    {
+      "epoch": 0.14,
+      "learning_rate": 4.5633802816901414e-06,
+      "loss": 1.2436,
+      "step": 162
+    },
+    {
+      "epoch": 0.14,
+      "learning_rate": 4.591549295774648e-06,
+      "loss": 1.2545,
+      "step": 163
+    },
+    {
+      "epoch": 0.14,
+      "learning_rate": 4.619718309859155e-06,
+      "loss": 1.21,
+      "step": 164
+    },
+    {
+      "epoch": 0.14,
+      "learning_rate": 4.647887323943662e-06,
+      "loss": 1.0979,
+      "step": 165
+    },
+    {
+      "epoch": 0.14,
+      "learning_rate": 4.67605633802817e-06,
+      "loss": 1.2912,
+      "step": 166
+    },
+    {
+      "epoch": 0.14,
+      "learning_rate": 4.704225352112676e-06,
+      "loss": 1.1633,
+      "step": 167
+    },
+    {
+      "epoch": 0.14,
+      "learning_rate": 4.732394366197183e-06,
+      "loss": 1.3976,
+      "step": 168
+    },
+    {
+      "epoch": 0.14,
+      "learning_rate": 4.7605633802816905e-06,
+      "loss": 1.2334,
+      "step": 169
+    },
+    {
+      "epoch": 0.14,
+      "learning_rate": 4.788732394366197e-06,
+      "loss": 1.4072,
+      "step": 170
+    },
+    {
+      "epoch": 0.14,
+      "learning_rate": 4.816901408450705e-06,
+      "loss": 1.4666,
+      "step": 171
+    },
+    {
+      "epoch": 0.15,
+      "learning_rate": 4.845070422535212e-06,
+      "loss": 1.2881,
+      "step": 172
+    },
+    {
+      "epoch": 0.15,
+      "learning_rate": 4.873239436619719e-06,
+      "loss": 1.1987,
+      "step": 173
+    },
+    {
+      "epoch": 0.15,
+      "learning_rate": 4.901408450704226e-06,
+      "loss": 1.4005,
+      "step": 174
+    },
+    {
+      "epoch": 0.15,
+      "learning_rate": 4.929577464788733e-06,
+      "loss": 1.2795,
+      "step": 175
+    },
+    {
+      "epoch": 0.15,
+      "learning_rate": 4.9577464788732395e-06,
+      "loss": 1.4463,
+      "step": 176
+    },
+    {
+      "epoch": 0.15,
+      "learning_rate": 4.985915492957747e-06,
+      "loss": 1.2518,
+      "step": 177
+    },
+    {
+      "epoch": 0.15,
+      "learning_rate": 5.014084507042254e-06,
+      "loss": 1.2853,
+      "step": 178
+    },
+    {
+      "epoch": 0.15,
+      "learning_rate": 5.042253521126761e-06,
+      "loss": 1.2258,
+      "step": 179
+    },
+    {
+      "epoch": 0.15,
+      "learning_rate": 5.070422535211268e-06,
+      "loss": 1.1489,
+      "step": 180
+    },
+    {
+      "epoch": 0.15,
+      "learning_rate": 5.098591549295775e-06,
+      "loss": 1.229,
+      "step": 181
+    },
+    {
+      "epoch": 0.15,
+      "learning_rate": 5.126760563380282e-06,
+      "loss": 1.2392,
+      "step": 182
+    },
+    {
+      "epoch": 0.15,
+      "learning_rate": 5.154929577464789e-06,
+      "loss": 1.339,
+      "step": 183
+    },
+    {
+      "epoch": 0.16,
+      "learning_rate": 5.183098591549296e-06,
+      "loss": 1.1776,
+      "step": 184
+    },
+    {
+      "epoch": 0.16,
+      "learning_rate": 5.211267605633803e-06,
+      "loss": 1.2591,
+      "step": 185
+    },
+    {
+      "epoch": 0.16,
+      "learning_rate": 5.23943661971831e-06,
+      "loss": 1.144,
+      "step": 186
+    },
+    {
+      "epoch": 0.16,
+      "learning_rate": 5.267605633802817e-06,
+      "loss": 1.3478,
+      "step": 187
+    },
+    {
+      "epoch": 0.16,
+      "learning_rate": 5.295774647887324e-06,
+      "loss": 1.2036,
+      "step": 188
+    },
+    {
+      "epoch": 0.16,
+      "learning_rate": 5.323943661971831e-06,
+      "loss": 1.1938,
+      "step": 189
+    },
+    {
+      "epoch": 0.16,
+      "learning_rate": 5.352112676056338e-06,
+      "loss": 1.2662,
+      "step": 190
+    },
+    {
+      "epoch": 0.16,
+      "learning_rate": 5.380281690140845e-06,
+      "loss": 1.235,
+      "step": 191
+    },
+    {
+      "epoch": 0.16,
+      "learning_rate": 5.408450704225352e-06,
+      "loss": 1.1768,
+      "step": 192
+    },
+    {
+      "epoch": 0.16,
+      "learning_rate": 5.43661971830986e-06,
+      "loss": 1.1144,
+      "step": 193
+    },
+    {
+      "epoch": 0.16,
+      "learning_rate": 5.4647887323943675e-06,
+      "loss": 1.4197,
+      "step": 194
+    },
+    {
+      "epoch": 0.16,
+      "learning_rate": 5.492957746478874e-06,
+      "loss": 1.2688,
+      "step": 195
+    },
+    {
+      "epoch": 0.17,
+      "learning_rate": 5.521126760563381e-06,
+      "loss": 1.2287,
+      "step": 196
+    },
+    {
+      "epoch": 0.17,
+      "learning_rate": 5.549295774647888e-06,
+      "loss": 1.2163,
+      "step": 197
+    },
+    {
+      "epoch": 0.17,
+      "learning_rate": 5.577464788732395e-06,
+      "loss": 1.0609,
+      "step": 198
+    },
+    {
+      "epoch": 0.17,
+      "learning_rate": 5.605633802816902e-06,
+      "loss": 1.1017,
+      "step": 199
+    },
+    {
+      "epoch": 0.17,
+      "learning_rate": 5.633802816901409e-06,
+      "loss": 1.185,
+      "step": 200
+    },
+    {
+      "epoch": 0.17,
+      "learning_rate": 5.6619718309859165e-06,
+      "loss": 1.0727,
+      "step": 201
+    },
+    {
+      "epoch": 0.17,
+      "learning_rate": 5.690140845070423e-06,
+      "loss": 1.5206,
+      "step": 202
+    },
+    {
+      "epoch": 0.17,
+      "learning_rate": 5.71830985915493e-06,
+      "loss": 1.0901,
+      "step": 203
+    },
+    {
+      "epoch": 0.17,
+      "learning_rate": 5.7464788732394374e-06,
+      "loss": 1.3183,
+      "step": 204
+    },
+    {
+      "epoch": 0.17,
+      "learning_rate": 5.774647887323944e-06,
+      "loss": 1.3105,
+      "step": 205
+    },
+    {
+      "epoch": 0.17,
+      "learning_rate": 5.802816901408451e-06,
+      "loss": 1.3785,
+      "step": 206
+    },
+    {
+      "epoch": 0.18,
+      "learning_rate": 5.830985915492958e-06,
+      "loss": 1.1496,
+      "step": 207
+    },
+    {
+      "epoch": 0.18,
+      "learning_rate": 5.859154929577466e-06,
+      "loss": 1.167,
+      "step": 208
+    },
+    {
+      "epoch": 0.18,
+      "learning_rate": 5.887323943661972e-06,
+      "loss": 1.2458,
+      "step": 209
+    },
+    {
+      "epoch": 0.18,
+      "learning_rate": 5.915492957746479e-06,
+      "loss": 1.1596,
+      "step": 210
+    },
+    {
+      "epoch": 0.18,
+      "learning_rate": 5.9436619718309865e-06,
+      "loss": 1.1357,
+      "step": 211
+    },
+    {
+      "epoch": 0.18,
+      "learning_rate": 5.971830985915493e-06,
+      "loss": 0.9843,
+      "step": 212
+    },
+    {
+      "epoch": 0.18,
+      "learning_rate": 6e-06,
+      "loss": 1.2245,
+      "step": 213
+    },
+    {
+      "epoch": 0.18,
+      "learning_rate": 6.028169014084507e-06,
+      "loss": 1.0025,
+      "step": 214
+    },
+    {
+      "epoch": 0.18,
+      "learning_rate": 6.056338028169015e-06,
+      "loss": 1.1183,
+      "step": 215
+    },
+    {
+      "epoch": 0.18,
+      "learning_rate": 6.084507042253521e-06,
+      "loss": 1.5397,
+      "step": 216
+    },
+    {
+      "epoch": 0.18,
+      "learning_rate": 6.112676056338028e-06,
+      "loss": 1.0552,
+      "step": 217
+    },
+    {
+      "epoch": 0.18,
+      "learning_rate": 6.1408450704225356e-06,
+      "loss": 1.3601,
+      "step": 218
+    },
+    {
+      "epoch": 0.19,
+      "learning_rate": 6.169014084507042e-06,
+      "loss": 1.2845,
+      "step": 219
+    },
+    {
+      "epoch": 0.19,
+      "learning_rate": 6.197183098591549e-06,
+      "loss": 1.0935,
+      "step": 220
+    },
+    {
+      "epoch": 0.19,
+      "learning_rate": 6.2253521126760565e-06,
+      "loss": 1.146,
+      "step": 221
+    },
+    {
+      "epoch": 0.19,
+      "learning_rate": 6.2535211267605646e-06,
+      "loss": 1.1768,
+      "step": 222
+    },
+    {
+      "epoch": 0.19,
+      "learning_rate": 6.281690140845072e-06,
+      "loss": 1.1167,
+      "step": 223
+    },
+    {
+      "epoch": 0.19,
+      "learning_rate": 6.309859154929578e-06,
+      "loss": 1.3253,
+      "step": 224
+    },
+    {
+      "epoch": 0.19,
+      "learning_rate": 6.3380281690140855e-06,
+      "loss": 1.1376,
+      "step": 225
+    },
+    {
+      "epoch": 0.19,
+      "learning_rate": 6.366197183098593e-06,
+      "loss": 1.3858,
+      "step": 226
+    },
+    {
+      "epoch": 0.19,
+      "learning_rate": 6.394366197183099e-06,
+      "loss": 1.0459,
+      "step": 227
+    },
+    {
+      "epoch": 0.19,
+      "learning_rate": 6.422535211267606e-06,
+      "loss": 1.1447,
+      "step": 228
+    },
+    {
+      "epoch": 0.19,
+      "learning_rate": 6.450704225352114e-06,
+      "loss": 1.1114,
+      "step": 229
+    },
+    {
+      "epoch": 0.19,
+      "learning_rate": 6.478873239436621e-06,
+      "loss": 1.464,
+      "step": 230
+    },
+    {
+      "epoch": 0.2,
+      "learning_rate": 6.507042253521127e-06,
+      "loss": 1.2157,
+      "step": 231
+    },
+    {
+      "epoch": 0.2,
+      "learning_rate": 6.5352112676056345e-06,
+      "loss": 1.122,
+      "step": 232
+    },
+    {
+      "epoch": 0.2,
+      "learning_rate": 6.563380281690142e-06,
+      "loss": 1.1739,
+      "step": 233
+    },
+    {
+      "epoch": 0.2,
+      "learning_rate": 6.591549295774648e-06,
+      "loss": 1.3235,
+      "step": 234
+    },
+    {
+      "epoch": 0.2,
+      "learning_rate": 6.619718309859155e-06,
+      "loss": 1.2631,
+      "step": 235
+    },
+    {
+      "epoch": 0.2,
+      "learning_rate": 6.647887323943663e-06,
+      "loss": 1.2404,
+      "step": 236
+    },
+    {
+      "epoch": 0.2,
+      "eval_loss": 1.1165865659713745,
+      "eval_runtime": 269.3914,
+      "eval_samples_per_second": 3.708,
+      "eval_steps_per_second": 0.464,
+      "step": 236
+    },
+    {
+      "epoch": 0.2,
+      "learning_rate": 6.67605633802817e-06,
+      "loss": 1.1415,
+      "step": 237
+    },
+    {
+      "epoch": 0.2,
+      "learning_rate": 6.704225352112676e-06,
+      "loss": 1.2497,
+      "step": 238
+    },
+    {
+      "epoch": 0.2,
+      "learning_rate": 6.7323943661971836e-06,
+      "loss": 1.1512,
+      "step": 239
+    },
+    {
+      "epoch": 0.2,
+      "learning_rate": 6.760563380281691e-06,
+      "loss": 1.0767,
+      "step": 240
+    },
+    {
+      "epoch": 0.2,
+      "learning_rate": 6.788732394366197e-06,
+      "loss": 0.9569,
+      "step": 241
+    },
+    {
+      "epoch": 0.2,
+      "learning_rate": 6.8169014084507045e-06,
+      "loss": 1.35,
+      "step": 242
+    },
+    {
+      "epoch": 0.21,
+      "learning_rate": 6.845070422535212e-06,
+      "loss": 1.2382,
+      "step": 243
+    },
+    {
+      "epoch": 0.21,
+      "learning_rate": 6.873239436619719e-06,
+      "loss": 1.1095,
+      "step": 244
+    },
+    {
+      "epoch": 0.21,
+      "learning_rate": 6.901408450704225e-06,
+      "loss": 1.2218,
+      "step": 245
+    },
+    {
+      "epoch": 0.21,
+      "learning_rate": 6.929577464788733e-06,
+      "loss": 1.382,
+      "step": 246
+    },
+    {
+      "epoch": 0.21,
+      "learning_rate": 6.95774647887324e-06,
+      "loss": 1.369,
+      "step": 247
+    },
+    {
+      "epoch": 0.21,
+      "learning_rate": 6.985915492957746e-06,
+      "loss": 1.2502,
+      "step": 248
+    },
+    {
+      "epoch": 0.21,
+      "learning_rate": 7.0140845070422535e-06,
+      "loss": 1.0344,
+      "step": 249
+    },
+    {
+      "epoch": 0.21,
+      "learning_rate": 7.042253521126761e-06,
+      "loss": 1.2101,
+      "step": 250
+    },
+    {
+      "epoch": 0.21,
+      "learning_rate": 7.070422535211268e-06,
+      "loss": 1.2263,
+      "step": 251
+    },
+    {
+      "epoch": 0.21,
+      "learning_rate": 7.098591549295776e-06,
+      "loss": 1.1012,
+      "step": 252
+    },
+    {
+      "epoch": 0.21,
+      "learning_rate": 7.1267605633802825e-06,
+      "loss": 1.2301,
+      "step": 253
+    },
+    {
+      "epoch": 0.21,
+      "learning_rate": 7.15492957746479e-06,
+      "loss": 1.2226,
+      "step": 254
+    },
+    {
+      "epoch": 0.22,
+      "learning_rate": 7.183098591549297e-06,
+      "loss": 1.2694,
+      "step": 255
+    },
+    {
+      "epoch": 0.22,
+      "learning_rate": 7.211267605633803e-06,
+      "loss": 1.4335,
+      "step": 256
+    },
+    {
+      "epoch": 0.22,
+      "learning_rate": 7.239436619718311e-06,
+      "loss": 1.1342,
+      "step": 257
+    },
+    {
+      "epoch": 0.22,
+      "learning_rate": 7.267605633802818e-06,
+      "loss": 1.2146,
+      "step": 258
+    },
+    {
+      "epoch": 0.22,
+      "learning_rate": 7.295774647887325e-06,
+      "loss": 1.1116,
+      "step": 259
+    },
+    {
+      "epoch": 0.22,
+      "learning_rate": 7.3239436619718316e-06,
+      "loss": 1.2331,
+      "step": 260
+    },
+    {
+      "epoch": 0.22,
+      "learning_rate": 7.352112676056339e-06,
+      "loss": 1.0549,
+      "step": 261
+    },
+    {
+      "epoch": 0.22,
+      "learning_rate": 7.380281690140846e-06,
+      "loss": 1.222,
+      "step": 262
+    },
+    {
+      "epoch": 0.22,
+      "learning_rate": 7.4084507042253525e-06,
+      "loss": 1.1478,
+      "step": 263
+    },
+    {
+      "epoch": 0.22,
+      "learning_rate": 7.43661971830986e-06,
+      "loss": 1.2954,
+      "step": 264
+    },
+    {
+      "epoch": 0.22,
+      "learning_rate": 7.464788732394367e-06,
+      "loss": 1.196,
+      "step": 265
+    },
+    {
+      "epoch": 0.22,
+      "learning_rate": 7.492957746478874e-06,
+      "loss": 1.0139,
+      "step": 266
+    },
+    {
+      "epoch": 0.23,
+      "learning_rate": 7.521126760563381e-06,
+      "loss": 1.5008,
+      "step": 267
+    },
+    {
+      "epoch": 0.23,
+      "learning_rate": 7.549295774647888e-06,
+      "loss": 1.1822,
+      "step": 268
+    },
+    {
+      "epoch": 0.23,
+      "learning_rate": 7.577464788732395e-06,
+      "loss": 1.103,
+      "step": 269
+    },
+    {
+      "epoch": 0.23,
+      "learning_rate": 7.6056338028169015e-06,
+      "loss": 1.1599,
+      "step": 270
+    },
+    {
+      "epoch": 0.23,
+      "learning_rate": 7.633802816901409e-06,
+      "loss": 1.1041,
+      "step": 271
+    },
+    {
+      "epoch": 0.23,
+      "learning_rate": 7.661971830985916e-06,
+      "loss": 1.4274,
+      "step": 272
+    },
+    {
+      "epoch": 0.23,
+      "learning_rate": 7.690140845070423e-06,
+      "loss": 1.1177,
+      "step": 273
+    },
+    {
+      "epoch": 0.23,
+      "learning_rate": 7.71830985915493e-06,
+      "loss": 1.3761,
+      "step": 274
+    },
+    {
+      "epoch": 0.23,
+      "learning_rate": 7.746478873239436e-06,
+      "loss": 1.1913,
+      "step": 275
+    },
+    {
+      "epoch": 0.23,
+      "learning_rate": 7.774647887323943e-06,
+      "loss": 1.0957,
+      "step": 276
+    },
+    {
+      "epoch": 0.23,
+      "learning_rate": 7.80281690140845e-06,
+      "loss": 1.0218,
+      "step": 277
+    },
+    {
+      "epoch": 0.24,
+      "learning_rate": 7.830985915492958e-06,
+      "loss": 1.2873,
+      "step": 278
+    },
+    {
+      "epoch": 0.24,
+      "learning_rate": 7.859154929577465e-06,
+      "loss": 1.2246,
+      "step": 279
+    },
+    {
+      "epoch": 0.24,
+      "learning_rate": 7.887323943661972e-06,
+      "loss": 1.2641,
+      "step": 280
+    },
+    {
+      "epoch": 0.24,
+      "learning_rate": 7.91549295774648e-06,
+      "loss": 1.2165,
+      "step": 281
+    },
+    {
+      "epoch": 0.24,
+      "learning_rate": 7.943661971830987e-06,
+      "loss": 1.1298,
+      "step": 282
+    },
+    {
+      "epoch": 0.24,
+      "learning_rate": 7.971830985915494e-06,
+      "loss": 1.2814,
+      "step": 283
+    },
+    {
+      "epoch": 0.24,
+      "learning_rate": 8.000000000000001e-06,
+      "loss": 1.0617,
+      "step": 284
+    },
+    {
+      "epoch": 0.24,
+      "learning_rate": 8.028169014084509e-06,
+      "loss": 1.1663,
+      "step": 285
+    },
+    {
+      "epoch": 0.24,
+      "learning_rate": 8.056338028169016e-06,
+      "loss": 1.1186,
+      "step": 286
+    },
+    {
+      "epoch": 0.24,
+      "learning_rate": 8.084507042253521e-06,
+      "loss": 1.3063,
+      "step": 287
+    },
+    {
+      "epoch": 0.24,
+      "learning_rate": 8.112676056338029e-06,
+      "loss": 1.2076,
+      "step": 288
+    },
+    {
+      "epoch": 0.24,
+      "learning_rate": 8.140845070422536e-06,
+      "loss": 1.084,
+      "step": 289
+    },
+    {
+      "epoch": 0.25,
+      "learning_rate": 8.169014084507043e-06,
+      "loss": 1.3851,
+      "step": 290
+    },
+    {
+      "epoch": 0.25,
+      "learning_rate": 8.19718309859155e-06,
+      "loss": 1.3329,
+      "step": 291
+    },
+    {
+      "epoch": 0.25,
+      "learning_rate": 8.225352112676058e-06,
+      "loss": 1.1447,
+      "step": 292
+    },
+    {
+      "epoch": 0.25,
+      "learning_rate": 8.253521126760565e-06,
+      "loss": 1.1243,
+      "step": 293
+    },
+    {
+      "epoch": 0.25,
+      "learning_rate": 8.28169014084507e-06,
+      "loss": 1.1835,
+      "step": 294
+    },
+    {
+      "epoch": 0.25,
+      "learning_rate": 8.309859154929578e-06,
+      "loss": 1.0693,
+      "step": 295
+    },
+    {
+      "epoch": 0.25,
+      "learning_rate": 8.338028169014085e-06,
+      "loss": 1.0549,
+      "step": 296
+    },
+    {
+      "epoch": 0.25,
+      "learning_rate": 8.366197183098592e-06,
+      "loss": 1.0551,
+      "step": 297
+    },
+    {
+      "epoch": 0.25,
+      "learning_rate": 8.3943661971831e-06,
+      "loss": 1.3071,
+      "step": 298
+    },
+    {
+      "epoch": 0.25,
+      "learning_rate": 8.422535211267607e-06,
+      "loss": 1.2693,
+      "step": 299
+    },
+    {
+      "epoch": 0.25,
+      "learning_rate": 8.450704225352114e-06,
+      "loss": 1.3027,
+      "step": 300
+    },
+    {
+      "epoch": 0.25,
+      "learning_rate": 8.47887323943662e-06,
+      "loss": 1.0142,
+      "step": 301
+    },
+    {
+      "epoch": 0.26,
+      "learning_rate": 8.507042253521127e-06,
+      "loss": 1.3342,
+      "step": 302
+    },
+    {
+      "epoch": 0.26,
+      "learning_rate": 8.535211267605634e-06,
+      "loss": 1.1116,
+      "step": 303
+    },
+    {
+      "epoch": 0.26,
+      "learning_rate": 8.563380281690141e-06,
+      "loss": 1.1299,
+      "step": 304
+    },
+    {
+      "epoch": 0.26,
+      "learning_rate": 8.591549295774648e-06,
+      "loss": 1.2015,
+      "step": 305
+    },
+    {
+      "epoch": 0.26,
+      "learning_rate": 8.619718309859156e-06,
+      "loss": 1.2035,
+      "step": 306
+    },
+    {
+      "epoch": 0.26,
+      "learning_rate": 8.647887323943663e-06,
+      "loss": 1.1743,
+      "step": 307
+    },
+    {
+      "epoch": 0.26,
+      "learning_rate": 8.676056338028169e-06,
+      "loss": 1.1202,
+      "step": 308
+    },
+    {
+      "epoch": 0.26,
+      "learning_rate": 8.704225352112676e-06,
+      "loss": 1.4374,
+      "step": 309
+    },
+    {
+      "epoch": 0.26,
+      "learning_rate": 8.732394366197183e-06,
+      "loss": 1.3139,
+      "step": 310
+    },
+    {
+      "epoch": 0.26,
+      "learning_rate": 8.760563380281692e-06,
+      "loss": 1.1208,
+      "step": 311
+    },
+    {
+      "epoch": 0.26,
+      "learning_rate": 8.7887323943662e-06,
+      "loss": 1.2294,
+      "step": 312
+    },
+    {
+      "epoch": 0.26,
+      "learning_rate": 8.816901408450705e-06,
+      "loss": 1.2346,
+      "step": 313
+    },
+    {
+      "epoch": 0.27,
+      "learning_rate": 8.845070422535212e-06,
+      "loss": 1.2432,
+      "step": 314
+    },
+    {
+      "epoch": 0.27,
+      "learning_rate": 8.87323943661972e-06,
+      "loss": 1.1142,
+      "step": 315
+    },
+    {
+      "epoch": 0.27,
+      "learning_rate": 8.901408450704227e-06,
+      "loss": 1.1793,
+      "step": 316
+    },
+    {
+      "epoch": 0.27,
+      "learning_rate": 8.929577464788734e-06,
+      "loss": 1.2635,
+      "step": 317
+    },
+    {
+      "epoch": 0.27,
+      "learning_rate": 8.957746478873241e-06,
+      "loss": 1.2721,
+      "step": 318
+    },
+    {
+      "epoch": 0.27,
+      "learning_rate": 8.985915492957748e-06,
+      "loss": 1.1499,
+      "step": 319
+    },
+    {
+      "epoch": 0.27,
+      "learning_rate": 9.014084507042254e-06,
+      "loss": 1.414,
+      "step": 320
+    },
+    {
+      "epoch": 0.27,
+      "learning_rate": 9.042253521126761e-06,
+      "loss": 1.3243,
+      "step": 321
+    },
+    {
+      "epoch": 0.27,
+      "learning_rate": 9.070422535211268e-06,
+      "loss": 1.2385,
+      "step": 322
+    },
+    {
+      "epoch": 0.27,
+      "learning_rate": 9.098591549295776e-06,
+      "loss": 1.4159,
+      "step": 323
+    },
+    {
+      "epoch": 0.27,
+      "learning_rate": 9.126760563380283e-06,
+      "loss": 1.1442,
+      "step": 324
+    },
+    {
+      "epoch": 0.27,
+      "learning_rate": 9.15492957746479e-06,
+      "loss": 1.1827,
+      "step": 325
+    },
+    {
+      "epoch": 0.28,
+      "learning_rate": 9.183098591549296e-06,
+      "loss": 1.0153,
+      "step": 326
+    },
+    {
+      "epoch": 0.28,
+      "learning_rate": 9.211267605633803e-06,
+      "loss": 1.1522,
+      "step": 327
+    },
+    {
+      "epoch": 0.28,
+      "learning_rate": 9.23943661971831e-06,
+      "loss": 1.0683,
+      "step": 328
+    },
+    {
+      "epoch": 0.28,
+      "learning_rate": 9.267605633802817e-06,
+      "loss": 1.0879,
+      "step": 329
+    },
+    {
+      "epoch": 0.28,
+      "learning_rate": 9.295774647887325e-06,
+      "loss": 1.0873,
+      "step": 330
+    },
+    {
+      "epoch": 0.28,
+      "learning_rate": 9.323943661971832e-06,
+      "loss": 1.3289,
+      "step": 331
+    },
+    {
+      "epoch": 0.28,
+      "learning_rate": 9.35211267605634e-06,
+      "loss": 1.224,
+      "step": 332
+    },
+    {
+      "epoch": 0.28,
+      "learning_rate": 9.380281690140845e-06,
+      "loss": 1.2185,
+      "step": 333
+    },
+    {
+      "epoch": 0.28,
+      "learning_rate": 9.408450704225352e-06,
+      "loss": 1.1307,
+      "step": 334
+    },
+    {
+      "epoch": 0.28,
+      "learning_rate": 9.43661971830986e-06,
+      "loss": 1.1897,
+      "step": 335
+    },
+    {
+      "epoch": 0.28,
+      "learning_rate": 9.464788732394366e-06,
+      "loss": 1.3596,
+      "step": 336
+    },
+    {
+      "epoch": 0.28,
+      "learning_rate": 9.492957746478874e-06,
+      "loss": 1.055,
+      "step": 337
+    },
+    {
+      "epoch": 0.29,
+      "learning_rate": 9.521126760563381e-06,
+      "loss": 1.2561,
+      "step": 338
+    },
+    {
+      "epoch": 0.29,
+      "learning_rate": 9.549295774647888e-06,
+      "loss": 1.0939,
+      "step": 339
+    },
+    {
+      "epoch": 0.29,
+      "learning_rate": 9.577464788732394e-06,
+      "loss": 1.2203,
+      "step": 340
+    },
+    {
+      "epoch": 0.29,
+      "learning_rate": 9.605633802816903e-06,
+      "loss": 1.2119,
+      "step": 341
+    },
+    {
+      "epoch": 0.29,
+      "learning_rate": 9.63380281690141e-06,
+      "loss": 1.1143,
+      "step": 342
+    },
+    {
+      "epoch": 0.29,
+      "learning_rate": 9.661971830985917e-06,
+      "loss": 1.1352,
+      "step": 343
+    },
+    {
+      "epoch": 0.29,
+      "learning_rate": 9.690140845070424e-06,
+      "loss": 0.9916,
+      "step": 344
+    },
+    {
+      "epoch": 0.29,
+      "learning_rate": 9.71830985915493e-06,
+      "loss": 1.131,
+      "step": 345
+    },
+    {
+      "epoch": 0.29,
+      "learning_rate": 9.746478873239437e-06,
+      "loss": 1.2705,
+      "step": 346
+    },
+    {
+      "epoch": 0.29,
+      "learning_rate": 9.774647887323945e-06,
+      "loss": 1.0525,
+      "step": 347
+    },
+    {
+      "epoch": 0.29,
+      "learning_rate": 9.802816901408452e-06,
+      "loss": 1.147,
+      "step": 348
+    },
+    {
+      "epoch": 0.3,
+      "learning_rate": 9.830985915492959e-06,
+      "loss": 1.2475,
+      "step": 349
+    },
+    {
+      "epoch": 0.3,
+      "learning_rate": 9.859154929577466e-06,
+      "loss": 1.3519,
+      "step": 350
+    },
+    {
+      "epoch": 0.3,
+      "learning_rate": 9.887323943661974e-06,
+      "loss": 1.3495,
+      "step": 351
+    },
+    {
+      "epoch": 0.3,
+      "learning_rate": 9.915492957746479e-06,
+      "loss": 1.088,
+      "step": 352
+    },
+    {
+      "epoch": 0.3,
+      "learning_rate": 9.943661971830986e-06,
+      "loss": 1.2713,
+      "step": 353
+    },
+    {
+      "epoch": 0.3,
+      "learning_rate": 9.971830985915494e-06,
+      "loss": 1.305,
+      "step": 354
+    },
+    {
+      "epoch": 0.3,
+      "learning_rate": 1e-05,
+      "loss": 1.2236,
+      "step": 355
+    },
+    {
+      "epoch": 0.3,
+      "learning_rate": 9.999997576817574e-06,
+      "loss": 1.3538,
+      "step": 356
+    },
+    {
+      "epoch": 0.3,
+      "learning_rate": 9.999990307272647e-06,
+      "loss": 1.1192,
+      "step": 357
+    },
+    {
+      "epoch": 0.3,
+      "learning_rate": 9.999978191372262e-06,
+      "loss": 1.6531,
+      "step": 358
+    },
+    {
+      "epoch": 0.3,
+      "learning_rate": 9.999961229128163e-06,
+      "loss": 1.2085,
+      "step": 359
+    },
+    {
+      "epoch": 0.3,
+      "learning_rate": 9.999939420556794e-06,
+      "loss": 1.1289,
+      "step": 360
+    },
+    {
+      "epoch": 0.31,
+      "learning_rate": 9.99991276567929e-06,
+      "loss": 1.2806,
+      "step": 361
+    },
+    {
+      "epoch": 0.31,
+      "learning_rate": 9.99988126452149e-06,
+      "loss": 1.2833,
+      "step": 362
+    },
+    {
+      "epoch": 0.31,
+      "learning_rate": 9.999844917113923e-06,
+      "loss": 1.2652,
+      "step": 363
+    },
+    {
+      "epoch": 0.31,
+      "learning_rate": 9.999803723491823e-06,
+      "loss": 1.2422,
+      "step": 364
+    },
+    {
+      "epoch": 0.31,
+      "learning_rate": 9.999757683695118e-06,
+      "loss": 1.1305,
+      "step": 365
+    },
+    {
+      "epoch": 0.31,
+      "learning_rate": 9.999706797768432e-06,
+      "loss": 1.0564,
+      "step": 366
+    },
+    {
+      "epoch": 0.31,
+      "learning_rate": 9.999651065761086e-06,
+      "loss": 1.3385,
+      "step": 367
+    },
+    {
+      "epoch": 0.31,
+      "learning_rate": 9.999590487727103e-06,
+      "loss": 1.1687,
+      "step": 368
+    },
+    {
+      "epoch": 0.31,
+      "learning_rate": 9.999525063725197e-06,
+      "loss": 1.1614,
+      "step": 369
+    },
+    {
+      "epoch": 0.31,
+      "learning_rate": 9.999454793818782e-06,
+      "loss": 1.0145,
+      "step": 370
+    },
+    {
+      "epoch": 0.31,
+      "learning_rate": 9.99937967807597e-06,
+      "loss": 1.2096,
+      "step": 371
+    },
+    {
+      "epoch": 0.31,
+      "learning_rate": 9.999299716569567e-06,
+      "loss": 1.1507,
+      "step": 372
+    },
+    {
+      "epoch": 0.32,
+      "learning_rate": 9.999214909377078e-06,
+      "loss": 1.1685,
+      "step": 373
+    },
+    {
+      "epoch": 0.32,
+      "learning_rate": 9.999125256580707e-06,
+      "loss": 1.0961,
+      "step": 374
+    },
+    {
+      "epoch": 0.32,
+      "learning_rate": 9.999030758267348e-06,
+      "loss": 1.2691,
+      "step": 375
+    },
+    {
+      "epoch": 0.32,
+      "learning_rate": 9.998931414528597e-06,
+      "loss": 1.1904,
+      "step": 376
+    },
+    {
+      "epoch": 0.32,
+      "learning_rate": 9.998827225460745e-06,
+      "loss": 1.0868,
+      "step": 377
+    },
+    {
+      "epoch": 0.32,
+      "learning_rate": 9.99871819116478e-06,
+      "loss": 0.9782,
+      "step": 378
+    },
+    {
+      "epoch": 0.32,
+      "learning_rate": 9.998604311746387e-06,
+      "loss": 1.0768,
+      "step": 379
+    },
+    {
+      "epoch": 0.32,
+      "learning_rate": 9.998485587315945e-06,
+      "loss": 1.2106,
+      "step": 380
+    },
+    {
+      "epoch": 0.32,
+      "learning_rate": 9.998362017988532e-06,
+      "loss": 1.1554,
+      "step": 381
+    },
+    {
+      "epoch": 0.32,
+      "learning_rate": 9.998233603883917e-06,
+      "loss": 1.1672,
+      "step": 382
+    },
+    {
+      "epoch": 0.32,
+      "learning_rate": 9.998100345126573e-06,
+      "loss": 1.3399,
+      "step": 383
+    },
+    {
+      "epoch": 0.32,
+      "learning_rate": 9.99796224184566e-06,
+      "loss": 1.2215,
+      "step": 384
+    },
+    {
+      "epoch": 0.33,
+      "learning_rate": 9.99781929417504e-06,
+      "loss": 1.1036,
+      "step": 385
+    },
+    {
+      "epoch": 0.33,
+      "learning_rate": 9.997671502253268e-06,
+      "loss": 1.1551,
+      "step": 386
+    },
+    {
+      "epoch": 0.33,
+      "learning_rate": 9.997518866223593e-06,
+      "loss": 1.2181,
+      "step": 387
+    },
+    {
+      "epoch": 0.33,
+      "learning_rate": 9.997361386233963e-06,
+      "loss": 1.2378,
+      "step": 388
+    },
+    {
+      "epoch": 0.33,
+      "learning_rate": 9.99719906243702e-06,
+      "loss": 1.2212,
+      "step": 389
+    },
+    {
+      "epoch": 0.33,
+      "learning_rate": 9.997031894990097e-06,
+      "loss": 1.1687,
+      "step": 390
+    },
+    {
+      "epoch": 0.33,
+      "learning_rate": 9.996859884055226e-06,
+      "loss": 1.1656,
+      "step": 391
+    },
+    {
+      "epoch": 0.33,
+      "learning_rate": 9.996683029799135e-06,
+      "loss": 1.266,
+      "step": 392
+    },
+    {
+      "epoch": 0.33,
+      "learning_rate": 9.996501332393241e-06,
+      "loss": 1.1556,
+      "step": 393
+    },
+    {
+      "epoch": 0.33,
+      "learning_rate": 9.996314792013659e-06,
+      "loss": 1.1053,
+      "step": 394
+    },
+    {
+      "epoch": 0.33,
+      "learning_rate": 9.996123408841198e-06,
+      "loss": 1.077,
+      "step": 395
+    },
+    {
+      "epoch": 0.33,
+      "learning_rate": 9.995927183061361e-06,
+      "loss": 1.0333,
+      "step": 396
+    },
+    {
+      "epoch": 0.34,
+      "learning_rate": 9.995726114864343e-06,
+      "loss": 1.4151,
+      "step": 397
+    },
+    {
+      "epoch": 0.34,
+      "learning_rate": 9.995520204445036e-06,
+      "loss": 1.171,
+      "step": 398
+    },
+    {
+      "epoch": 0.34,
+      "learning_rate": 9.995309452003022e-06,
+      "loss": 1.1679,
+      "step": 399
+    },
+    {
+      "epoch": 0.34,
+      "learning_rate": 9.995093857742579e-06,
+      "loss": 1.1301,
+      "step": 400
+    },
+    {
+      "epoch": 0.34,
+      "learning_rate": 9.994873421872673e-06,
+      "loss": 1.1813,
+      "step": 401
+    },
+    {
+      "epoch": 0.34,
+      "learning_rate": 9.994648144606971e-06,
+      "loss": 1.1935,
+      "step": 402
+    },
+    {
+      "epoch": 0.34,
+      "learning_rate": 9.994418026163826e-06,
+      "loss": 1.1786,
+      "step": 403
+    },
+    {
+      "epoch": 0.34,
+      "learning_rate": 9.994183066766285e-06,
+      "loss": 1.3293,
+      "step": 404
+    },
+    {
+      "epoch": 0.34,
+      "learning_rate": 9.99394326664209e-06,
+      "loss": 1.008,
+      "step": 405
+    },
+    {
+      "epoch": 0.34,
+      "learning_rate": 9.99369862602367e-06,
+      "loss": 1.1245,
+      "step": 406
+    },
+    {
+      "epoch": 0.34,
+      "learning_rate": 9.993449145148151e-06,
+      "loss": 1.1093,
+      "step": 407
+    },
+    {
+      "epoch": 0.35,
+      "learning_rate": 9.993194824257346e-06,
+      "loss": 1.192,
+      "step": 408
+    },
+    {
+      "epoch": 0.35,
+      "learning_rate": 9.992935663597764e-06,
+      "loss": 1.1796,
+      "step": 409
+    },
+    {
+      "epoch": 0.35,
+      "learning_rate": 9.9926716634206e-06,
+      "loss": 1.35,
+      "step": 410
+    },
+    {
+      "epoch": 0.35,
+      "learning_rate": 9.992402823981743e-06,
+      "loss": 1.1076,
+      "step": 411
+    },
+    {
+      "epoch": 0.35,
+      "learning_rate": 9.992129145541771e-06,
+      "loss": 1.2868,
+      "step": 412
+    },
+    {
+      "epoch": 0.35,
+      "learning_rate": 9.991850628365956e-06,
+      "loss": 1.034,
+      "step": 413
+    },
+    {
+      "epoch": 0.35,
+      "learning_rate": 9.991567272724254e-06,
+      "loss": 1.171,
+      "step": 414
+    },
+    {
+      "epoch": 0.35,
+      "learning_rate": 9.991279078891313e-06,
+      "loss": 1.1568,
+      "step": 415
+    },
+    {
+      "epoch": 0.35,
+      "learning_rate": 9.990986047146477e-06,
+      "loss": 1.3956,
+      "step": 416
+    },
+    {
+      "epoch": 0.35,
+      "learning_rate": 9.990688177773769e-06,
+      "loss": 1.0784,
+      "step": 417
+    },
+    {
+      "epoch": 0.35,
+      "learning_rate": 9.990385471061908e-06,
+      "loss": 1.1396,
+      "step": 418
+    },
+    {
+      "epoch": 0.35,
+      "learning_rate": 9.990077927304296e-06,
+      "loss": 1.0309,
+      "step": 419
+    },
+    {
+      "epoch": 0.36,
+      "learning_rate": 9.989765546799032e-06,
+      "loss": 1.1868,
+      "step": 420
+    },
+    {
+      "epoch": 0.36,
+      "learning_rate": 9.989448329848895e-06,
+      "loss": 1.1238,
+      "step": 421
+    },
+    {
+      "epoch": 0.36,
+      "learning_rate": 9.989126276761355e-06,
+      "loss": 1.3391,
+      "step": 422
+    },
+    {
+      "epoch": 0.36,
+      "learning_rate": 9.988799387848567e-06,
+      "loss": 1.2333,
+      "step": 423
+    },
+    {
+      "epoch": 0.36,
+      "learning_rate": 9.988467663427381e-06,
+      "loss": 1.3088,
+      "step": 424
+    },
+    {
+      "epoch": 0.36,
+      "learning_rate": 9.988131103819326e-06,
+      "loss": 1.2754,
+      "step": 425
+    },
+    {
+      "epoch": 0.36,
+      "learning_rate": 9.987789709350618e-06,
+      "loss": 1.05,
+      "step": 426
+    },
+    {
+      "epoch": 0.36,
+      "learning_rate": 9.987443480352164e-06,
+      "loss": 1.098,
+      "step": 427
+    },
+    {
+      "epoch": 0.36,
+      "learning_rate": 9.987092417159552e-06,
+      "loss": 1.1603,
+      "step": 428
+    },
+    {
+      "epoch": 0.36,
+      "learning_rate": 9.986736520113062e-06,
+      "loss": 1.2035,
+      "step": 429
+    },
+    {
+      "epoch": 0.36,
+      "learning_rate": 9.986375789557652e-06,
+      "loss": 1.2774,
+      "step": 430
+    },
+    {
+      "epoch": 0.36,
+      "learning_rate": 9.98601022584297e-06,
+      "loss": 1.1506,
+      "step": 431
+    },
+    {
+      "epoch": 0.37,
+      "learning_rate": 9.985639829323345e-06,
+      "loss": 1.1668,
+      "step": 432
+    },
+    {
+      "epoch": 0.37,
+      "learning_rate": 9.985264600357795e-06,
+      "loss": 1.2493,
+      "step": 433
+    },
+    {
+      "epoch": 0.37,
+      "learning_rate": 9.984884539310017e-06,
+      "loss": 1.3186,
+      "step": 434
+    },
+    {
+      "epoch": 0.37,
+      "learning_rate": 9.984499646548395e-06,
+      "loss": 1.0414,
+      "step": 435
+    },
+    {
+      "epoch": 0.37,
+      "learning_rate": 9.984109922445998e-06,
+      "loss": 1.3101,
+      "step": 436
+    },
+    {
+      "epoch": 0.37,
+      "learning_rate": 9.983715367380568e-06,
+      "loss": 1.1813,
+      "step": 437
+    },
+    {
+      "epoch": 0.37,
+      "learning_rate": 9.983315981734544e-06,
+      "loss": 1.112,
+      "step": 438
+    },
+    {
+      "epoch": 0.37,
+      "learning_rate": 9.982911765895035e-06,
+      "loss": 1.2157,
+      "step": 439
+    },
+    {
+      "epoch": 0.37,
+      "learning_rate": 9.982502720253838e-06,
+      "loss": 1.3154,
+      "step": 440
+    },
+    {
+      "epoch": 0.37,
+      "learning_rate": 9.98208884520743e-06,
+      "loss": 1.279,
+      "step": 441
+    },
+    {
+      "epoch": 0.37,
+      "learning_rate": 9.98167014115697e-06,
+      "loss": 1.2873,
+      "step": 442
+    },
+    {
+      "epoch": 0.37,
+      "learning_rate": 9.981246608508293e-06,
+      "loss": 0.9583,
+      "step": 443
+    },
+    {
+      "epoch": 0.38,
+      "learning_rate": 9.98081824767192e-06,
+      "loss": 1.1099,
+      "step": 444
+    },
+    {
+      "epoch": 0.38,
+      "learning_rate": 9.980385059063051e-06,
+      "loss": 1.1703,
+      "step": 445
+    },
+    {
+      "epoch": 0.38,
+      "learning_rate": 9.979947043101562e-06,
+      "loss": 1.2188,
+      "step": 446
+    },
+    {
+      "epoch": 0.38,
+      "learning_rate": 9.979504200212011e-06,
+      "loss": 1.2011,
+      "step": 447
+    },
+    {
+      "epoch": 0.38,
+      "learning_rate": 9.979056530823632e-06,
+      "loss": 1.1022,
+      "step": 448
+    },
+    {
+      "epoch": 0.38,
+      "learning_rate": 9.978604035370341e-06,
+      "loss": 1.1497,
+      "step": 449
+    },
+    {
+      "epoch": 0.38,
+      "learning_rate": 9.978146714290729e-06,
+      "loss": 1.2801,
+      "step": 450
+    },
+    {
+      "epoch": 0.38,
+      "learning_rate": 9.977684568028064e-06,
+      "loss": 1.0909,
+      "step": 451
+    },
+    {
+      "epoch": 0.38,
+      "learning_rate": 9.977217597030293e-06,
+      "loss": 1.144,
+      "step": 452
+    },
+    {
+      "epoch": 0.38,
+      "learning_rate": 9.976745801750038e-06,
+      "loss": 1.1655,
+      "step": 453
+    },
+    {
+      "epoch": 0.38,
+      "learning_rate": 9.976269182644596e-06,
+      "loss": 1.2935,
+      "step": 454
+    },
+    {
+      "epoch": 0.38,
+      "learning_rate": 9.975787740175944e-06,
+      "loss": 1.2087,
+      "step": 455
+    },
+    {
+      "epoch": 0.39,
+      "learning_rate": 9.975301474810727e-06,
+      "loss": 1.1307,
+      "step": 456
+    },
+    {
+      "epoch": 0.39,
+      "learning_rate": 9.974810387020276e-06,
+      "loss": 1.1211,
+      "step": 457
+    },
+    {
+      "epoch": 0.39,
+      "learning_rate": 9.97431447728058e-06,
+      "loss": 1.0913,
+      "step": 458
+    },
+    {
+      "epoch": 0.39,
+      "learning_rate": 9.973813746072319e-06,
+      "loss": 1.2474,
+      "step": 459
+    },
+    {
+      "epoch": 0.39,
+      "learning_rate": 9.97330819388083e-06,
+      "loss": 1.6276,
+      "step": 460
+    },
+    {
+      "epoch": 0.39,
+      "learning_rate": 9.972797821196139e-06,
+      "loss": 1.2004,
+      "step": 461
+    },
+    {
+      "epoch": 0.39,
+      "learning_rate": 9.972282628512933e-06,
+      "loss": 1.1322,
+      "step": 462
+    },
+    {
+      "epoch": 0.39,
+      "learning_rate": 9.971762616330574e-06,
+      "loss": 1.2291,
+      "step": 463
+    },
+    {
+      "epoch": 0.39,
+      "learning_rate": 9.971237785153096e-06,
+      "loss": 1.1639,
+      "step": 464
+    },
+    {
+      "epoch": 0.39,
+      "learning_rate": 9.970708135489204e-06,
+      "loss": 1.3482,
+      "step": 465
+    },
+    {
+      "epoch": 0.39,
+      "learning_rate": 9.970173667852275e-06,
+      "loss": 1.0859,
+      "step": 466
+    },
+    {
+      "epoch": 0.39,
+      "learning_rate": 9.96963438276035e-06,
+      "loss": 1.3802,
+      "step": 467
+    },
+    {
+      "epoch": 0.4,
+      "learning_rate": 9.969090280736145e-06,
+      "loss": 1.1761,
+      "step": 468
+    },
+    {
+      "epoch": 0.4,
+      "learning_rate": 9.968541362307045e-06,
+      "loss": 1.1965,
+      "step": 469
+    },
+    {
+      "epoch": 0.4,
+      "learning_rate": 9.9679876280051e-06,
+      "loss": 1.1966,
+      "step": 470
+    },
+    {
+      "epoch": 0.4,
+      "learning_rate": 9.96742907836703e-06,
+      "loss": 1.175,
+      "step": 471
+    },
+    {
+      "epoch": 0.4,
+      "learning_rate": 9.966865713934224e-06,
+      "loss": 1.2103,
+      "step": 472
+    },
+    {
+      "epoch": 0.4,
+      "eval_loss": 1.1101255416870117,
+      "eval_runtime": 268.0616,
+      "eval_samples_per_second": 3.727,
+      "eval_steps_per_second": 0.466,
+      "step": 472
+    },
+    {
+      "epoch": 0.4,
+      "learning_rate": 9.966297535252733e-06,
+      "loss": 1.056,
+      "step": 473
+    },
+    {
+      "epoch": 0.4,
+      "learning_rate": 9.96572454287328e-06,
+      "loss": 1.3044,
+      "step": 474
+    },
+    {
+      "epoch": 0.4,
+      "learning_rate": 9.965146737351249e-06,
+      "loss": 1.1039,
+      "step": 475
+    },
+    {
+      "epoch": 0.4,
+      "learning_rate": 9.964564119246692e-06,
+      "loss": 1.2216,
+      "step": 476
+    },
+    {
+      "epoch": 0.4,
+      "learning_rate": 9.963976689124324e-06,
+      "loss": 1.1563,
+      "step": 477
+    },
+    {
+      "epoch": 0.4,
+      "learning_rate": 9.963384447553527e-06,
+      "loss": 1.3368,
+      "step": 478
+    },
+    {
+      "epoch": 0.41,
+      "learning_rate": 9.962787395108344e-06,
+      "loss": 1.27,
+      "step": 479
+    },
+    {
+      "epoch": 0.41,
+      "learning_rate": 9.96218553236748e-06,
+      "loss": 1.2054,
+      "step": 480
+    },
+    {
+      "epoch": 0.41,
+      "learning_rate": 9.961578859914307e-06,
+      "loss": 1.2318,
+      "step": 481
+    },
+    {
+      "epoch": 0.41,
+      "learning_rate": 9.960967378336854e-06,
+      "loss": 0.975,
+      "step": 482
+    },
+    {
+      "epoch": 0.41,
+      "learning_rate": 9.960351088227817e-06,
+      "loss": 0.9899,
+      "step": 483
+    },
+    {
+      "epoch": 0.41,
+      "learning_rate": 9.959729990184546e-06,
+      "loss": 1.1608,
+      "step": 484
+    },
+    {
+      "epoch": 0.41,
+      "learning_rate": 9.959104084809055e-06,
+      "loss": 1.2532,
+      "step": 485
+    },
+    {
+      "epoch": 0.41,
+      "learning_rate": 9.958473372708018e-06,
+      "loss": 1.0833,
+      "step": 486
+    },
+    {
+      "epoch": 0.41,
+      "learning_rate": 9.957837854492766e-06,
+      "loss": 1.2435,
+      "step": 487
+    },
+    {
+      "epoch": 0.41,
+      "learning_rate": 9.95719753077929e-06,
+      "loss": 1.0572,
+      "step": 488
+    },
+    {
+      "epoch": 0.41,
+      "learning_rate": 9.95655240218824e-06,
+      "loss": 1.0929,
+      "step": 489
+    },
+    {
+      "epoch": 0.41,
+      "learning_rate": 9.955902469344923e-06,
+      "loss": 1.2556,
+      "step": 490
+    },
+    {
+      "epoch": 0.42,
+      "learning_rate": 9.955247732879296e-06,
+      "loss": 1.0618,
+      "step": 491
+    },
+    {
+      "epoch": 0.42,
+      "learning_rate": 9.95458819342598e-06,
+      "loss": 1.167,
+      "step": 492
+    },
+    {
+      "epoch": 0.42,
+      "learning_rate": 9.953923851624251e-06,
+      "loss": 1.081,
+      "step": 493
+    },
+    {
+      "epoch": 0.42,
+      "learning_rate": 9.953254708118035e-06,
+      "loss": 0.9734,
+      "step": 494
+    },
+    {
+      "epoch": 0.42,
+      "learning_rate": 9.952580763555916e-06,
+      "loss": 1.1847,
+      "step": 495
+    },
+    {
+      "epoch": 0.42,
+      "learning_rate": 9.951902018591129e-06,
+      "loss": 1.2165,
+      "step": 496
+    },
+    {
+      "epoch": 0.42,
+      "learning_rate": 9.951218473881564e-06,
+      "loss": 1.2169,
+      "step": 497
+    },
+    {
+      "epoch": 0.42,
+      "learning_rate": 9.950530130089762e-06,
+      "loss": 1.212,
+      "step": 498
+    },
+    {
+      "epoch": 0.42,
+      "learning_rate": 9.949836987882916e-06,
+      "loss": 1.3074,
+      "step": 499
+    },
+    {
+      "epoch": 0.42,
+      "learning_rate": 9.949139047932872e-06,
+      "loss": 1.0401,
+      "step": 500
+    },
+    {
+      "epoch": 0.42,
+      "learning_rate": 9.948436310916121e-06,
+      "loss": 1.0383,
+      "step": 501
+    },
+    {
+      "epoch": 0.42,
+      "learning_rate": 9.947728777513809e-06,
+      "loss": 1.1037,
+      "step": 502
+    },
+    {
+      "epoch": 0.43,
+      "learning_rate": 9.947016448411728e-06,
+      "loss": 1.529,
+      "step": 503
+    },
+    {
+      "epoch": 0.43,
+      "learning_rate": 9.94629932430032e-06,
+      "loss": 0.9924,
+      "step": 504
+    },
+    {
+      "epoch": 0.43,
+      "learning_rate": 9.945577405874675e-06,
+      "loss": 1.1132,
+      "step": 505
+    },
+    {
+      "epoch": 0.43,
+      "learning_rate": 9.944850693834527e-06,
+      "loss": 1.1532,
+      "step": 506
+    },
+    {
+      "epoch": 0.43,
+      "learning_rate": 9.94411918888426e-06,
+      "loss": 1.0351,
+      "step": 507
+    },
+    {
+      "epoch": 0.43,
+      "learning_rate": 9.9433828917329e-06,
+      "loss": 1.295,
+      "step": 508
+    },
+    {
+      "epoch": 0.43,
+      "learning_rate": 9.942641803094122e-06,
+      "loss": 1.2403,
+      "step": 509
+    },
+    {
+      "epoch": 0.43,
+      "learning_rate": 9.941895923686243e-06,
+      "loss": 1.1403,
+      "step": 510
+    },
+    {
+      "epoch": 0.43,
+      "learning_rate": 9.941145254232223e-06,
+      "loss": 1.2117,
+      "step": 511
+    },
+    {
+      "epoch": 0.43,
+      "learning_rate": 9.940389795459665e-06,
+      "loss": 1.293,
+      "step": 512
+    },
+    {
+      "epoch": 0.43,
+      "learning_rate": 9.939629548100815e-06,
+      "loss": 1.0857,
+      "step": 513
+    },
+    {
+      "epoch": 0.43,
+      "learning_rate": 9.938864512892562e-06,
+      "loss": 1.144,
+      "step": 514
+    },
+    {
+      "epoch": 0.44,
+      "learning_rate": 9.938094690576432e-06,
+      "loss": 1.1896,
+      "step": 515
+    },
+    {
+      "epoch": 0.44,
+      "learning_rate": 9.937320081898594e-06,
+      "loss": 1.2558,
+      "step": 516
+    },
+    {
+      "epoch": 0.44,
+      "learning_rate": 9.936540687609854e-06,
+      "loss": 1.0021,
+      "step": 517
+    },
+    {
+      "epoch": 0.44,
+      "learning_rate": 9.935756508465659e-06,
+      "loss": 1.2673,
+      "step": 518
+    },
+    {
+      "epoch": 0.44,
+      "learning_rate": 9.934967545226093e-06,
+      "loss": 1.325,
+      "step": 519
+    },
+    {
+      "epoch": 0.44,
+      "learning_rate": 9.934173798655877e-06,
+      "loss": 1.2052,
+      "step": 520
+    },
+    {
+      "epoch": 0.44,
+      "learning_rate": 9.933375269524366e-06,
+      "loss": 1.1101,
+      "step": 521
+    },
+    {
+      "epoch": 0.44,
+      "learning_rate": 9.932571958605555e-06,
+      "loss": 1.2369,
+      "step": 522
+    },
+    {
+      "epoch": 0.44,
+      "learning_rate": 9.931763866678068e-06,
+      "loss": 1.1627,
+      "step": 523
+    },
+    {
+      "epoch": 0.44,
+      "learning_rate": 9.93095099452517e-06,
+      "loss": 1.1519,
+      "step": 524
+    },
+    {
+      "epoch": 0.44,
+      "learning_rate": 9.930133342934757e-06,
+      "loss": 0.973,
+      "step": 525
+    },
+    {
+      "epoch": 0.44,
+      "learning_rate": 9.929310912699353e-06,
+      "loss": 1.328,
+      "step": 526
+    },
+    {
+      "epoch": 0.45,
+      "learning_rate": 9.928483704616119e-06,
+      "loss": 1.3404,
+      "step": 527
+    },
+    {
+      "epoch": 0.45,
+      "learning_rate": 9.927651719486847e-06,
+      "loss": 1.136,
+      "step": 528
+    },
+    {
+      "epoch": 0.45,
+      "learning_rate": 9.926814958117955e-06,
+      "loss": 1.0129,
+      "step": 529
+    },
+    {
+      "epoch": 0.45,
+      "learning_rate": 9.925973421320496e-06,
+      "loss": 1.0048,
+      "step": 530
+    },
+    {
+      "epoch": 0.45,
+      "learning_rate": 9.925127109910145e-06,
+      "loss": 1.2779,
+      "step": 531
+    },
+    {
+      "epoch": 0.45,
+      "learning_rate": 9.924276024707213e-06,
+      "loss": 1.2642,
+      "step": 532
+    },
+    {
+      "epoch": 0.45,
+      "learning_rate": 9.92342016653663e-06,
+      "loss": 1.209,
+      "step": 533
+    },
+    {
+      "epoch": 0.45,
+      "learning_rate": 9.92255953622796e-06,
+      "loss": 1.1691,
+      "step": 534
+    },
+    {
+      "epoch": 0.45,
+      "learning_rate": 9.921694134615384e-06,
+      "loss": 1.1225,
+      "step": 535
+    },
+    {
+      "epoch": 0.45,
+      "learning_rate": 9.920823962537719e-06,
+      "loss": 1.1844,
+      "step": 536
+    },
+    {
+      "epoch": 0.45,
+      "learning_rate": 9.919949020838395e-06,
+      "loss": 1.2377,
+      "step": 537
+    },
+    {
+      "epoch": 0.45,
+      "learning_rate": 9.919069310365467e-06,
+      "loss": 1.1053,
+      "step": 538
+    },
+    {
+      "epoch": 0.46,
+      "learning_rate": 9.91818483197162e-06,
+      "loss": 1.0328,
+      "step": 539
+    },
+    {
+      "epoch": 0.46,
+      "learning_rate": 9.917295586514151e-06,
+      "loss": 1.2847,
+      "step": 540
+    },
+    {
+      "epoch": 0.46,
+      "learning_rate": 9.916401574854983e-06,
+      "loss": 1.1179,
+      "step": 541
+    },
+    {
+      "epoch": 0.46,
+      "learning_rate": 9.91550279786066e-06,
+      "loss": 1.3723,
+      "step": 542
+    },
+    {
+      "epoch": 0.46,
+      "learning_rate": 9.914599256402337e-06,
+      "loss": 1.345,
+      "step": 543
+    },
+    {
+      "epoch": 0.46,
+      "learning_rate": 9.913690951355794e-06,
+      "loss": 1.2219,
+      "step": 544
+    },
+    {
+      "epoch": 0.46,
+      "learning_rate": 9.91277788360143e-06,
+      "loss": 1.1517,
+      "step": 545
+    },
+    {
+      "epoch": 0.46,
+      "learning_rate": 9.911860054024254e-06,
+      "loss": 1.125,
+      "step": 546
+    },
+    {
+      "epoch": 0.46,
+      "learning_rate": 9.910937463513893e-06,
+      "loss": 1.3486,
+      "step": 547
+    },
+    {
+      "epoch": 0.46,
+      "learning_rate": 9.910010112964589e-06,
+      "loss": 1.3102,
+      "step": 548
+    },
+    {
+      "epoch": 0.46,
+      "learning_rate": 9.909078003275198e-06,
+      "loss": 1.081,
+      "step": 549
+    },
+    {
+      "epoch": 0.47,
+      "learning_rate": 9.908141135349192e-06,
+      "loss": 1.1321,
+      "step": 550
+    },
+    {
+      "epoch": 0.47,
+      "learning_rate": 9.907199510094647e-06,
+      "loss": 1.0972,
+      "step": 551
+    },
+    {
+      "epoch": 0.47,
+      "learning_rate": 9.906253128424257e-06,
+      "loss": 1.1546,
+      "step": 552
+    },
+    {
+      "epoch": 0.47,
+      "learning_rate": 9.905301991255325e-06,
+      "loss": 0.8754,
+      "step": 553
+    },
+    {
+      "epoch": 0.47,
+      "learning_rate": 9.904346099509763e-06,
+      "loss": 1.0774,
+      "step": 554
+    },
+    {
+      "epoch": 0.47,
+      "learning_rate": 9.903385454114088e-06,
+      "loss": 1.0208,
+      "step": 555
+    },
+    {
+      "epoch": 0.47,
+      "learning_rate": 9.90242005599943e-06,
+      "loss": 1.1422,
+      "step": 556
+    },
+    {
+      "epoch": 0.47,
+      "learning_rate": 9.901449906101522e-06,
+      "loss": 1.2779,
+      "step": 557
+    },
+    {
+      "epoch": 0.47,
+      "learning_rate": 9.900475005360705e-06,
+      "loss": 1.1482,
+      "step": 558
+    },
+    {
+      "epoch": 0.47,
+      "learning_rate": 9.899495354721923e-06,
+      "loss": 1.2687,
+      "step": 559
+    },
+    {
+      "epoch": 0.47,
+      "learning_rate": 9.898510955134728e-06,
+      "loss": 1.3523,
+      "step": 560
+    },
+    {
+      "epoch": 0.47,
+      "learning_rate": 9.897521807553267e-06,
+      "loss": 1.197,
+      "step": 561
+    },
+    {
+      "epoch": 0.48,
+      "learning_rate": 9.896527912936298e-06,
+      "loss": 1.3878,
+      "step": 562
+    },
+    {
+      "epoch": 0.48,
+      "learning_rate": 9.895529272247173e-06,
+      "loss": 1.1694,
+      "step": 563
+    },
+    {
+      "epoch": 0.48,
+      "learning_rate": 9.894525886453852e-06,
+      "loss": 1.0472,
+      "step": 564
+    },
+    {
+      "epoch": 0.48,
+      "learning_rate": 9.893517756528884e-06,
+      "loss": 1.1021,
+      "step": 565
+    },
+    {
+      "epoch": 0.48,
+      "learning_rate": 9.892504883449424e-06,
+      "loss": 1.1179,
+      "step": 566
+    },
+    {
+      "epoch": 0.48,
+      "learning_rate": 9.891487268197226e-06,
+      "loss": 1.1236,
+      "step": 567
+    },
+    {
+      "epoch": 0.48,
+      "learning_rate": 9.890464911758636e-06,
+      "loss": 1.2332,
+      "step": 568
+    },
+    {
+      "epoch": 0.48,
+      "learning_rate": 9.889437815124591e-06,
+      "loss": 1.1649,
+      "step": 569
+    },
+    {
+      "epoch": 0.48,
+      "learning_rate": 9.888405979290634e-06,
+      "loss": 1.0852,
+      "step": 570
+    },
+    {
+      "epoch": 0.48,
+      "learning_rate": 9.887369405256893e-06,
+      "loss": 1.1948,
+      "step": 571
+    },
+    {
+      "epoch": 0.48,
+      "learning_rate": 9.886328094028091e-06,
+      "loss": 1.1706,
+      "step": 572
+    },
+    {
+      "epoch": 0.48,
+      "learning_rate": 9.885282046613544e-06,
+      "loss": 1.2262,
+      "step": 573
+    },
+    {
+      "epoch": 0.49,
+      "learning_rate": 9.884231264027155e-06,
+      "loss": 1.0444,
+      "step": 574
+    },
+    {
+      "epoch": 0.49,
+      "learning_rate": 9.883175747287424e-06,
+      "loss": 1.1165,
+      "step": 575
+    },
+    {
+      "epoch": 0.49,
+      "learning_rate": 9.882115497417431e-06,
+      "loss": 1.2076,
+      "step": 576
+    },
+    {
+      "epoch": 0.49,
+      "learning_rate": 9.881050515444848e-06,
+      "loss": 1.2477,
+      "step": 577
+    },
+    {
+      "epoch": 0.49,
+      "learning_rate": 9.879980802401933e-06,
+      "loss": 1.3205,
+      "step": 578
+    },
+    {
+      "epoch": 0.49,
+      "learning_rate": 9.878906359325531e-06,
+      "loss": 1.164,
+      "step": 579
+    },
+    {
+      "epoch": 0.49,
+      "learning_rate": 9.877827187257071e-06,
+      "loss": 1.0069,
+      "step": 580
+    },
+    {
+      "epoch": 0.49,
+      "learning_rate": 9.876743287242565e-06,
+      "loss": 1.1968,
+      "step": 581
+    },
+    {
+      "epoch": 0.49,
+      "learning_rate": 9.875654660332608e-06,
+      "loss": 0.9908,
+      "step": 582
+    },
+    {
+      "epoch": 0.49,
+      "learning_rate": 9.874561307582375e-06,
+      "loss": 1.1746,
+      "step": 583
+    },
+    {
+      "epoch": 0.49,
+      "learning_rate": 9.873463230051624e-06,
+      "loss": 1.1264,
+      "step": 584
+    },
+    {
+      "epoch": 0.49,
+      "learning_rate": 9.872360428804693e-06,
+      "loss": 1.1619,
+      "step": 585
+    },
+    {
+      "epoch": 0.5,
+      "learning_rate": 9.871252904910497e-06,
+      "loss": 0.8161,
+      "step": 586
+    },
+    {
+      "epoch": 0.5,
+      "learning_rate": 9.870140659442529e-06,
+      "loss": 1.1386,
+      "step": 587
+    },
+    {
+      "epoch": 0.5,
+      "learning_rate": 9.869023693478857e-06,
+      "loss": 1.0719,
+      "step": 588
+    },
+    {
+      "epoch": 0.5,
+      "learning_rate": 9.867902008102129e-06,
+      "loss": 1.0873,
+      "step": 589
+    },
+    {
+      "epoch": 0.5,
+      "learning_rate": 9.866775604399562e-06,
+      "loss": 1.2557,
+      "step": 590
+    },
+    {
+      "epoch": 0.5,
+      "learning_rate": 9.865644483462948e-06,
+      "loss": 1.2357,
+      "step": 591
+    },
+    {
+      "epoch": 0.5,
+      "learning_rate": 9.864508646388654e-06,
+      "loss": 1.2229,
+      "step": 592
+    },
+    {
+      "epoch": 0.5,
+      "learning_rate": 9.863368094277615e-06,
+      "loss": 1.1834,
+      "step": 593
+    },
+    {
+      "epoch": 0.5,
+      "learning_rate": 9.862222828235337e-06,
+      "loss": 1.1732,
+      "step": 594
+    },
+    {
+      "epoch": 0.5,
+      "learning_rate": 9.861072849371898e-06,
+      "loss": 1.2159,
+      "step": 595
+    },
+    {
+      "epoch": 0.5,
+      "learning_rate": 9.859918158801938e-06,
+      "loss": 1.2349,
+      "step": 596
+    },
+    {
+      "epoch": 0.5,
+      "learning_rate": 9.858758757644669e-06,
+      "loss": 1.0245,
+      "step": 597
+    },
+    {
+      "epoch": 0.51,
+      "learning_rate": 9.857594647023866e-06,
+      "loss": 1.0835,
+      "step": 598
+    },
+    {
+      "epoch": 0.51,
+      "learning_rate": 9.856425828067872e-06,
+      "loss": 1.0209,
+      "step": 599
+    },
+    {
+      "epoch": 0.51,
+      "learning_rate": 9.85525230190959e-06,
+      "loss": 1.1577,
+      "step": 600
+    },
+    {
+      "epoch": 0.51,
+      "learning_rate": 9.854074069686488e-06,
+      "loss": 1.1819,
+      "step": 601
+    },
+    {
+      "epoch": 0.51,
+      "learning_rate": 9.852891132540594e-06,
+      "loss": 1.2076,
+      "step": 602
+    },
+    {
+      "epoch": 0.51,
+      "learning_rate": 9.851703491618498e-06,
+      "loss": 1.0536,
+      "step": 603
+    },
+    {
+      "epoch": 0.51,
+      "learning_rate": 9.850511148071349e-06,
+      "loss": 1.2979,
+      "step": 604
+    },
+    {
+      "epoch": 0.51,
+      "learning_rate": 9.84931410305485e-06,
+      "loss": 1.2886,
+      "step": 605
+    },
+    {
+      "epoch": 0.51,
+      "learning_rate": 9.848112357729267e-06,
+      "loss": 1.3502,
+      "step": 606
+    },
+    {
+      "epoch": 0.51,
+      "learning_rate": 9.846905913259419e-06,
+      "loss": 1.1714,
+      "step": 607
+    },
+    {
+      "epoch": 0.51,
+      "learning_rate": 9.84569477081468e-06,
+      "loss": 1.3495,
+      "step": 608
+    },
+    {
+      "epoch": 0.52,
+      "learning_rate": 9.844478931568977e-06,
+      "loss": 1.2452,
+      "step": 609
+    },
+    {
+      "epoch": 0.52,
+      "learning_rate": 9.84325839670079e-06,
+      "loss": 1.3248,
+      "step": 610
+    },
+    {
+      "epoch": 0.52,
+      "learning_rate": 9.842033167393152e-06,
+      "loss": 1.0893,
+      "step": 611
+    },
+    {
+      "epoch": 0.52,
+      "learning_rate": 9.84080324483364e-06,
+      "loss": 1.2009,
+      "step": 612
+    },
+    {
+      "epoch": 0.52,
+      "learning_rate": 9.839568630214394e-06,
+      "loss": 1.1399,
+      "step": 613
+    },
+    {
+      "epoch": 0.52,
+      "learning_rate": 9.838329324732082e-06,
+      "loss": 1.3179,
+      "step": 614
+    },
+    {
+      "epoch": 0.52,
+      "learning_rate": 9.837085329587936e-06,
+      "loss": 1.1108,
+      "step": 615
+    },
+    {
+      "epoch": 0.52,
+      "learning_rate": 9.835836645987726e-06,
+      "loss": 1.3108,
+      "step": 616
+    },
+    {
+      "epoch": 0.52,
+      "learning_rate": 9.834583275141764e-06,
+      "loss": 1.1305,
+      "step": 617
+    },
+    {
+      "epoch": 0.52,
+      "learning_rate": 9.833325218264913e-06,
+      "loss": 1.1946,
+      "step": 618
+    },
+    {
+      "epoch": 0.52,
+      "learning_rate": 9.832062476576569e-06,
+      "loss": 1.0692,
+      "step": 619
+    },
+    {
+      "epoch": 0.52,
+      "learning_rate": 9.830795051300678e-06,
+      "loss": 1.1116,
+      "step": 620
+    },
+    {
+      "epoch": 0.53,
+      "learning_rate": 9.829522943665717e-06,
+      "loss": 1.0437,
+      "step": 621
+    },
+    {
+      "epoch": 0.53,
+      "learning_rate": 9.828246154904709e-06,
+      "loss": 1.1046,
+      "step": 622
+    },
+    {
+      "epoch": 0.53,
+      "learning_rate": 9.826964686255207e-06,
+      "loss": 1.0374,
+      "step": 623
+    },
+    {
+      "epoch": 0.53,
+      "learning_rate": 9.825678538959307e-06,
+      "loss": 1.1186,
+      "step": 624
+    },
+    {
+      "epoch": 0.53,
+      "learning_rate": 9.824387714263636e-06,
+      "loss": 0.9855,
+      "step": 625
+    },
+    {
+      "epoch": 0.53,
+      "learning_rate": 9.823092213419354e-06,
+      "loss": 1.1694,
+      "step": 626
+    },
+    {
+      "epoch": 0.53,
+      "learning_rate": 9.82179203768216e-06,
+      "loss": 1.0869,
+      "step": 627
+    },
+    {
+      "epoch": 0.53,
+      "learning_rate": 9.820487188312271e-06,
+      "loss": 1.3015,
+      "step": 628
+    },
+    {
+      "epoch": 0.53,
+      "learning_rate": 9.81917766657445e-06,
+      "loss": 1.3485,
+      "step": 629
+    },
+    {
+      "epoch": 0.53,
+      "learning_rate": 9.817863473737974e-06,
+      "loss": 1.3147,
+      "step": 630
+    },
+    {
+      "epoch": 0.53,
+      "learning_rate": 9.816544611076662e-06,
+      "loss": 1.1984,
+      "step": 631
+    },
+    {
+      "epoch": 0.53,
+      "learning_rate": 9.815221079868847e-06,
+      "loss": 1.1805,
+      "step": 632
+    },
+    {
+      "epoch": 0.54,
+      "learning_rate": 9.813892881397392e-06,
+      "loss": 1.2298,
+      "step": 633
+    },
+    {
+      "epoch": 0.54,
+      "learning_rate": 9.812560016949685e-06,
+      "loss": 1.2681,
+      "step": 634
+    },
+    {
+      "epoch": 0.54,
+      "learning_rate": 9.811222487817638e-06,
+      "loss": 1.1365,
+      "step": 635
+    },
+    {
+      "epoch": 0.54,
+      "learning_rate": 9.809880295297677e-06,
+      "loss": 1.1779,
+      "step": 636
+    },
+    {
+      "epoch": 0.54,
+      "learning_rate": 9.808533440690757e-06,
+      "loss": 1.1847,
+      "step": 637
+    },
+    {
+      "epoch": 0.54,
+      "learning_rate": 9.807181925302347e-06,
+      "loss": 1.052,
+      "step": 638
+    },
+    {
+      "epoch": 0.54,
+      "learning_rate": 9.805825750442431e-06,
+      "loss": 1.171,
+      "step": 639
+    },
+    {
+      "epoch": 0.54,
+      "learning_rate": 9.804464917425518e-06,
+      "loss": 1.3062,
+      "step": 640
+    },
+    {
+      "epoch": 0.54,
+      "learning_rate": 9.803099427570622e-06,
+      "loss": 1.1465,
+      "step": 641
+    },
+    {
+      "epoch": 0.54,
+      "learning_rate": 9.801729282201277e-06,
+      "loss": 1.234,
+      "step": 642
+    },
+    {
+      "epoch": 0.54,
+      "learning_rate": 9.800354482645529e-06,
+      "loss": 1.0271,
+      "step": 643
+    },
+    {
+      "epoch": 0.54,
+      "learning_rate": 9.79897503023593e-06,
+      "loss": 1.181,
+      "step": 644
+    },
+    {
+      "epoch": 0.55,
+      "learning_rate": 9.797590926309553e-06,
+      "loss": 1.196,
+      "step": 645
+    },
+    {
+      "epoch": 0.55,
+      "learning_rate": 9.796202172207966e-06,
+      "loss": 1.1196,
+      "step": 646
+    },
+    {
+      "epoch": 0.55,
+      "learning_rate": 9.794808769277256e-06,
+      "loss": 1.1135,
+      "step": 647
+    },
+    {
+      "epoch": 0.55,
+      "learning_rate": 9.793410718868007e-06,
+      "loss": 1.1422,
+      "step": 648
+    },
+    {
+      "epoch": 0.55,
+      "learning_rate": 9.792008022335312e-06,
+      "loss": 1.1302,
+      "step": 649
+    },
+    {
+      "epoch": 0.55,
+      "learning_rate": 9.790600681038768e-06,
+      "loss": 1.377,
+      "step": 650
+    },
+    {
+      "epoch": 0.55,
+      "learning_rate": 9.789188696342472e-06,
+      "loss": 1.0816,
+      "step": 651
+    },
+    {
+      "epoch": 0.55,
+      "learning_rate": 9.787772069615024e-06,
+      "loss": 1.1914,
+      "step": 652
+    },
+    {
+      "epoch": 0.55,
+      "learning_rate": 9.786350802229519e-06,
+      "loss": 1.2652,
+      "step": 653
+    },
+    {
+      "epoch": 0.55,
+      "learning_rate": 9.784924895563557e-06,
+      "loss": 1.0834,
+      "step": 654
+    },
+    {
+      "epoch": 0.55,
+      "learning_rate": 9.783494350999227e-06,
+      "loss": 1.1317,
+      "step": 655
+    },
+    {
+      "epoch": 0.55,
+      "learning_rate": 9.782059169923122e-06,
+      "loss": 1.1723,
+      "step": 656
+    },
+    {
+      "epoch": 0.56,
+      "learning_rate": 9.780619353726318e-06,
+      "loss": 1.0223,
+      "step": 657
+    },
+    {
+      "epoch": 0.56,
+      "learning_rate": 9.779174903804394e-06,
+      "loss": 1.0725,
+      "step": 658
+    },
+    {
+      "epoch": 0.56,
+      "learning_rate": 9.777725821557416e-06,
+      "loss": 1.3925,
+      "step": 659
+    },
+    {
+      "epoch": 0.56,
+      "learning_rate": 9.77627210838994e-06,
+      "loss": 1.2944,
+      "step": 660
+    },
+    {
+      "epoch": 0.56,
+      "learning_rate": 9.77481376571101e-06,
+      "loss": 1.3887,
+      "step": 661
+    },
+    {
+      "epoch": 0.56,
+      "learning_rate": 9.773350794934158e-06,
+      "loss": 1.2,
+      "step": 662
+    },
+    {
+      "epoch": 0.56,
+      "learning_rate": 9.771883197477403e-06,
+      "loss": 1.0415,
+      "step": 663
+    },
+    {
+      "epoch": 0.56,
+      "learning_rate": 9.770410974763249e-06,
+      "loss": 1.3197,
+      "step": 664
+    },
+    {
+      "epoch": 0.56,
+      "learning_rate": 9.768934128218678e-06,
+      "loss": 1.1447,
+      "step": 665
+    },
+    {
+      "epoch": 0.56,
+      "learning_rate": 9.76745265927516e-06,
+      "loss": 1.0206,
+      "step": 666
+    },
+    {
+      "epoch": 0.56,
+      "learning_rate": 9.765966569368642e-06,
+      "loss": 1.0944,
+      "step": 667
+    },
+    {
+      "epoch": 0.56,
+      "learning_rate": 9.76447585993955e-06,
+      "loss": 1.0766,
+      "step": 668
+    },
+    {
+      "epoch": 0.57,
+      "learning_rate": 9.762980532432791e-06,
+      "loss": 1.0156,
+      "step": 669
+    },
+    {
+      "epoch": 0.57,
+      "learning_rate": 9.761480588297743e-06,
+      "loss": 1.3631,
+      "step": 670
+    },
+    {
+      "epoch": 0.57,
+      "learning_rate": 9.759976028988263e-06,
+      "loss": 1.1277,
+      "step": 671
+    },
+    {
+      "epoch": 0.57,
+      "learning_rate": 9.758466855962679e-06,
+      "loss": 1.1512,
+      "step": 672
+    },
+    {
+      "epoch": 0.57,
+      "learning_rate": 9.75695307068379e-06,
+      "loss": 1.1474,
+      "step": 673
+    },
+    {
+      "epoch": 0.57,
+      "learning_rate": 9.755434674618869e-06,
+      "loss": 0.9991,
+      "step": 674
+    },
+    {
+      "epoch": 0.57,
+      "learning_rate": 9.753911669239656e-06,
+      "loss": 1.2495,
+      "step": 675
+    },
+    {
+      "epoch": 0.57,
+      "learning_rate": 9.752384056022359e-06,
+      "loss": 1.3236,
+      "step": 676
+    },
+    {
+      "epoch": 0.57,
+      "learning_rate": 9.750851836447652e-06,
+      "loss": 0.9799,
+      "step": 677
+    },
+    {
+      "epoch": 0.57,
+      "learning_rate": 9.749315012000674e-06,
+      "loss": 1.1405,
+      "step": 678
+    },
+    {
+      "epoch": 0.57,
+      "learning_rate": 9.747773584171027e-06,
+      "loss": 1.3172,
+      "step": 679
+    },
+    {
+      "epoch": 0.58,
+      "learning_rate": 9.746227554452777e-06,
+      "loss": 1.1769,
+      "step": 680
+    },
+    {
+      "epoch": 0.58,
+      "learning_rate": 9.744676924344447e-06,
+      "loss": 1.1341,
+      "step": 681
+    },
+    {
+      "epoch": 0.58,
+      "learning_rate": 9.74312169534902e-06,
+      "loss": 1.151,
+      "step": 682
+    },
+    {
+      "epoch": 0.58,
+      "learning_rate": 9.741561868973942e-06,
+      "loss": 1.1281,
+      "step": 683
+    },
+    {
+      "epoch": 0.58,
+      "learning_rate": 9.739997446731104e-06,
+      "loss": 1.2655,
+      "step": 684
+    },
+    {
+      "epoch": 0.58,
+      "learning_rate": 9.738428430136864e-06,
+      "loss": 1.2719,
+      "step": 685
+    },
+    {
+      "epoch": 0.58,
+      "learning_rate": 9.736854820712023e-06,
+      "loss": 1.2859,
+      "step": 686
+    },
+    {
+      "epoch": 0.58,
+      "learning_rate": 9.735276619981841e-06,
+      "loss": 1.1492,
+      "step": 687
+    },
+    {
+      "epoch": 0.58,
+      "learning_rate": 9.733693829476025e-06,
+      "loss": 1.0478,
+      "step": 688
+    },
+    {
+      "epoch": 0.58,
+      "learning_rate": 9.73210645072873e-06,
+      "loss": 1.0658,
+      "step": 689
+    },
+    {
+      "epoch": 0.58,
+      "learning_rate": 9.73051448527856e-06,
+      "loss": 1.1349,
+      "step": 690
+    },
+    {
+      "epoch": 0.58,
+      "learning_rate": 9.728917934668562e-06,
+      "loss": 1.1609,
+      "step": 691
+    },
+    {
+      "epoch": 0.59,
+      "learning_rate": 9.727316800446233e-06,
+      "loss": 1.0521,
+      "step": 692
+    },
+    {
+      "epoch": 0.59,
+      "learning_rate": 9.725711084163507e-06,
+      "loss": 1.0287,
+      "step": 693
+    },
+    {
+      "epoch": 0.59,
+      "learning_rate": 9.724100787376761e-06,
+      "loss": 1.1164,
+      "step": 694
+    },
+    {
+      "epoch": 0.59,
+      "learning_rate": 9.722485911646812e-06,
+      "loss": 1.1328,
+      "step": 695
+    },
+    {
+      "epoch": 0.59,
+      "learning_rate": 9.720866458538918e-06,
+      "loss": 1.3098,
+      "step": 696
+    },
+    {
+      "epoch": 0.59,
+      "learning_rate": 9.71924242962277e-06,
+      "loss": 1.118,
+      "step": 697
+    },
+    {
+      "epoch": 0.59,
+      "learning_rate": 9.717613826472494e-06,
+      "loss": 1.2226,
+      "step": 698
+    },
+    {
+      "epoch": 0.59,
+      "learning_rate": 9.71598065066665e-06,
+      "loss": 1.3669,
+      "step": 699
+    },
+    {
+      "epoch": 0.59,
+      "learning_rate": 9.714342903788233e-06,
+      "loss": 1.0277,
+      "step": 700
+    },
+    {
+      "epoch": 0.59,
+      "learning_rate": 9.712700587424668e-06,
+      "loss": 0.9084,
+      "step": 701
+    },
+    {
+      "epoch": 0.59,
+      "learning_rate": 9.711053703167807e-06,
+      "loss": 1.2625,
+      "step": 702
+    },
+    {
+      "epoch": 0.59,
+      "learning_rate": 9.709402252613928e-06,
+      "loss": 1.2487,
+      "step": 703
+    },
+    {
+      "epoch": 0.6,
+      "learning_rate": 9.707746237363742e-06,
+      "loss": 1.1017,
+      "step": 704
+    },
+    {
+      "epoch": 0.6,
+      "learning_rate": 9.706085659022374e-06,
+      "loss": 1.1921,
+      "step": 705
+    },
+    {
+      "epoch": 0.6,
+      "learning_rate": 9.704420519199382e-06,
+      "loss": 0.9304,
+      "step": 706
+    },
+    {
+      "epoch": 0.6,
+      "learning_rate": 9.702750819508742e-06,
+      "loss": 1.0809,
+      "step": 707
+    },
+    {
+      "epoch": 0.6,
+      "learning_rate": 9.701076561568845e-06,
+      "loss": 1.0357,
+      "step": 708
+    },
+    {
+      "epoch": 0.6,
+      "eval_loss": 1.0738600492477417,
+      "eval_runtime": 268.3516,
+      "eval_samples_per_second": 3.723,
+      "eval_steps_per_second": 0.466,
+      "step": 708
+    },
+    {
+      "epoch": 0.6,
+      "learning_rate": 9.699397747002504e-06,
+      "loss": 1.2802,
+      "step": 709
+    },
+    {
+      "epoch": 0.6,
+      "learning_rate": 9.697714377436952e-06,
+      "loss": 1.114,
+      "step": 710
+    },
+    {
+      "epoch": 0.6,
+      "learning_rate": 9.69602645450383e-06,
+      "loss": 1.1755,
+      "step": 711
+    },
+    {
+      "epoch": 0.6,
+      "learning_rate": 9.694333979839199e-06,
+      "loss": 0.9908,
+      "step": 712
+    },
+    {
+      "epoch": 0.6,
+      "learning_rate": 9.692636955083527e-06,
+      "loss": 0.9812,
+      "step": 713
+    },
+    {
+      "epoch": 0.6,
+      "learning_rate": 9.690935381881697e-06,
+      "loss": 1.0437,
+      "step": 714
+    },
+    {
+      "epoch": 0.6,
+      "learning_rate": 9.689229261882993e-06,
+      "loss": 1.1983,
+      "step": 715
+    },
+    {
+      "epoch": 0.61,
+      "learning_rate": 9.687518596741118e-06,
+      "loss": 0.9838,
+      "step": 716
+    },
+    {
+      "epoch": 0.61,
+      "learning_rate": 9.685803388114165e-06,
+      "loss": 1.052,
+      "step": 717
+    },
+    {
+      "epoch": 0.61,
+      "learning_rate": 9.684083637664646e-06,
+      "loss": 1.0882,
+      "step": 718
+    },
+    {
+      "epoch": 0.61,
+      "learning_rate": 9.682359347059467e-06,
+      "loss": 1.4096,
+      "step": 719
+    },
+    {
+      "epoch": 0.61,
+      "learning_rate": 9.680630517969935e-06,
+      "loss": 1.1297,
+      "step": 720
+    },
+    {
+      "epoch": 0.61,
+      "learning_rate": 9.678897152071757e-06,
+      "loss": 1.1865,
+      "step": 721
+    },
+    {
+      "epoch": 0.61,
+      "learning_rate": 9.67715925104504e-06,
+      "loss": 1.1132,
+      "step": 722
+    },
+    {
+      "epoch": 0.61,
+      "learning_rate": 9.67541681657428e-06,
+      "loss": 1.1738,
+      "step": 723
+    },
+    {
+      "epoch": 0.61,
+      "learning_rate": 9.673669850348379e-06,
+      "loss": 0.9507,
+      "step": 724
+    },
+    {
+      "epoch": 0.61,
+      "learning_rate": 9.67191835406062e-06,
+      "loss": 1.2744,
+      "step": 725
+    },
+    {
+      "epoch": 0.61,
+      "learning_rate": 9.670162329408678e-06,
+      "loss": 1.1879,
+      "step": 726
+    },
+    {
+      "epoch": 0.61,
+      "learning_rate": 9.668401778094625e-06,
+      "loss": 1.1947,
+      "step": 727
+    },
+    {
+      "epoch": 0.62,
+      "learning_rate": 9.666636701824913e-06,
+      "loss": 1.2646,
+      "step": 728
+    },
+    {
+      "epoch": 0.62,
+      "learning_rate": 9.664867102310385e-06,
+      "loss": 1.1888,
+      "step": 729
+    },
+    {
+      "epoch": 0.62,
+      "learning_rate": 9.663092981266262e-06,
+      "loss": 1.2276,
+      "step": 730
+    },
+    {
+      "epoch": 0.62,
+      "learning_rate": 9.661314340412157e-06,
+      "loss": 1.1809,
+      "step": 731
+    },
+    {
+      "epoch": 0.62,
+      "learning_rate": 9.659531181472055e-06,
+      "loss": 1.1176,
+      "step": 732
+    },
+    {
+      "epoch": 0.62,
+      "learning_rate": 9.657743506174324e-06,
+      "loss": 1.0581,
+      "step": 733
+    },
+    {
+      "epoch": 0.62,
+      "learning_rate": 9.65595131625171e-06,
+      "loss": 1.3788,
+      "step": 734
+    },
+    {
+      "epoch": 0.62,
+      "learning_rate": 9.654154613441336e-06,
+      "loss": 1.1932,
+      "step": 735
+    },
+    {
+      "epoch": 0.62,
+      "learning_rate": 9.652353399484693e-06,
+      "loss": 1.2889,
+      "step": 736
+    },
+    {
+      "epoch": 0.62,
+      "learning_rate": 9.650547676127654e-06,
+      "loss": 1.3505,
+      "step": 737
+    },
+    {
+      "epoch": 0.62,
+      "learning_rate": 9.648737445120453e-06,
+      "loss": 1.1538,
+      "step": 738
+    },
+    {
+      "epoch": 0.62,
+      "learning_rate": 9.6469227082177e-06,
+      "loss": 1.1364,
+      "step": 739
+    },
+    {
+      "epoch": 0.63,
+      "learning_rate": 9.645103467178374e-06,
+      "loss": 1.1804,
+      "step": 740
+    },
+    {
+      "epoch": 0.63,
+      "learning_rate": 9.64327972376581e-06,
+      "loss": 1.1698,
+      "step": 741
+    },
+    {
+      "epoch": 0.63,
+      "learning_rate": 9.641451479747718e-06,
+      "loss": 1.0781,
+      "step": 742
+    },
+    {
+      "epoch": 0.63,
+      "learning_rate": 9.639618736896163e-06,
+      "loss": 1.2016,
+      "step": 743
+    },
+    {
+      "epoch": 0.63,
+      "learning_rate": 9.637781496987575e-06,
+      "loss": 1.1431,
+      "step": 744
+    },
+    {
+      "epoch": 0.63,
+      "learning_rate": 9.635939761802739e-06,
+      "loss": 1.029,
+      "step": 745
+    },
+    {
+      "epoch": 0.63,
+      "learning_rate": 9.6340935331268e-06,
+      "loss": 1.1025,
+      "step": 746
+    },
+    {
+      "epoch": 0.63,
+      "learning_rate": 9.632242812749257e-06,
+      "loss": 1.2688,
+      "step": 747
+    },
+    {
+      "epoch": 0.63,
+      "learning_rate": 9.630387602463964e-06,
+      "loss": 1.0155,
+      "step": 748
+    },
+    {
+      "epoch": 0.63,
+      "learning_rate": 9.628527904069125e-06,
+      "loss": 1.1089,
+      "step": 749
+    },
+    {
+      "epoch": 0.63,
+      "learning_rate": 9.626663719367299e-06,
+      "loss": 1.0812,
+      "step": 750
+    },
+    {
+      "epoch": 0.64,
+      "learning_rate": 9.624795050165385e-06,
+      "loss": 0.9655,
+      "step": 751
+    },
+    {
+      "epoch": 0.64,
+      "learning_rate": 9.622921898274635e-06,
+      "loss": 1.2333,
+      "step": 752
+    },
+    {
+      "epoch": 0.64,
+      "learning_rate": 9.621044265510648e-06,
+      "loss": 1.2071,
+      "step": 753
+    },
+    {
+      "epoch": 0.64,
+      "learning_rate": 9.619162153693358e-06,
+      "loss": 0.924,
+      "step": 754
+    },
+    {
+      "epoch": 0.64,
+      "learning_rate": 9.617275564647047e-06,
+      "loss": 1.1042,
+      "step": 755
+    },
+    {
+      "epoch": 0.64,
+      "learning_rate": 9.615384500200336e-06,
+      "loss": 1.0014,
+      "step": 756
+    },
+    {
+      "epoch": 0.64,
+      "learning_rate": 9.61348896218618e-06,
+      "loss": 1.2376,
+      "step": 757
+    },
+    {
+      "epoch": 0.64,
+      "learning_rate": 9.611588952441874e-06,
+      "loss": 1.4647,
+      "step": 758
+    },
+    {
+      "epoch": 0.64,
+      "learning_rate": 9.609684472809046e-06,
+      "loss": 1.0991,
+      "step": 759
+    },
+    {
+      "epoch": 0.64,
+      "learning_rate": 9.607775525133657e-06,
+      "loss": 1.1941,
+      "step": 760
+    },
+    {
+      "epoch": 0.64,
+      "learning_rate": 9.605862111266e-06,
+      "loss": 1.1336,
+      "step": 761
+    },
+    {
+      "epoch": 0.64,
+      "learning_rate": 9.603944233060694e-06,
+      "loss": 1.4626,
+      "step": 762
+    },
+    {
+      "epoch": 0.65,
+      "learning_rate": 9.602021892376684e-06,
+      "loss": 1.2229,
+      "step": 763
+    },
+    {
+      "epoch": 0.65,
+      "learning_rate": 9.600095091077246e-06,
+      "loss": 1.2506,
+      "step": 764
+    },
+    {
+      "epoch": 0.65,
+      "learning_rate": 9.598163831029974e-06,
+      "loss": 1.2708,
+      "step": 765
+    },
+    {
+      "epoch": 0.65,
+      "learning_rate": 9.59622811410679e-06,
+      "loss": 1.0561,
+      "step": 766
+    },
+    {
+      "epoch": 0.65,
+      "learning_rate": 9.59428794218393e-06,
+      "loss": 1.0878,
+      "step": 767
+    },
+    {
+      "epoch": 0.65,
+      "learning_rate": 9.592343317141945e-06,
+      "loss": 1.1305,
+      "step": 768
+    },
+    {
+      "epoch": 0.65,
+      "learning_rate": 9.590394240865717e-06,
+      "loss": 1.1386,
+      "step": 769
+    },
+    {
+      "epoch": 0.65,
+      "learning_rate": 9.588440715244427e-06,
+      "loss": 1.05,
+      "step": 770
+    },
+    {
+      "epoch": 0.65,
+      "learning_rate": 9.586482742171576e-06,
+      "loss": 1.1249,
+      "step": 771
+    },
+    {
+      "epoch": 0.65,
+      "learning_rate": 9.584520323544974e-06,
+      "loss": 1.1213,
+      "step": 772
+    },
+    {
+      "epoch": 0.65,
+      "learning_rate": 9.58255346126674e-06,
+      "loss": 1.1652,
+      "step": 773
+    },
+    {
+      "epoch": 0.65,
+      "learning_rate": 9.580582157243302e-06,
+      "loss": 1.1854,
+      "step": 774
+    },
+    {
+      "epoch": 0.66,
+      "learning_rate": 9.57860641338539e-06,
+      "loss": 0.8986,
+      "step": 775
+    },
+    {
+      "epoch": 0.66,
+      "learning_rate": 9.57662623160804e-06,
+      "loss": 1.2885,
+      "step": 776
+    },
+    {
+      "epoch": 0.66,
+      "learning_rate": 9.57464161383059e-06,
+      "loss": 1.0786,
+      "step": 777
+    },
+    {
+      "epoch": 0.66,
+      "learning_rate": 9.572652561976673e-06,
+      "loss": 1.1266,
+      "step": 778
+    },
+    {
+      "epoch": 0.66,
+      "learning_rate": 9.570659077974227e-06,
+      "loss": 1.0,
+      "step": 779
+    },
+    {
+      "epoch": 0.66,
+      "learning_rate": 9.568661163755479e-06,
+      "loss": 1.2158,
+      "step": 780
+    },
+    {
+      "epoch": 0.66,
+      "learning_rate": 9.566658821256953e-06,
+      "loss": 1.1925,
+      "step": 781
+    },
+    {
+      "epoch": 0.66,
+      "learning_rate": 9.56465205241947e-06,
+      "loss": 1.1023,
+      "step": 782
+    },
+    {
+      "epoch": 0.66,
+      "learning_rate": 9.562640859188133e-06,
+      "loss": 1.171,
+      "step": 783
+    },
+    {
+      "epoch": 0.66,
+      "learning_rate": 9.560625243512336e-06,
+      "loss": 0.983,
+      "step": 784
+    },
+    {
+      "epoch": 0.66,
+      "learning_rate": 9.558605207345765e-06,
+      "loss": 0.977,
+      "step": 785
+    },
+    {
+      "epoch": 0.66,
+      "learning_rate": 9.556580752646383e-06,
+      "loss": 1.2945,
+      "step": 786
+    },
+    {
+      "epoch": 0.67,
+      "learning_rate": 9.554551881376441e-06,
+      "loss": 1.0642,
+      "step": 787
+    },
+    {
+      "epoch": 0.67,
+      "learning_rate": 9.55251859550247e-06,
+      "loss": 1.3097,
+      "step": 788
+    },
+    {
+      "epoch": 0.67,
+      "learning_rate": 9.550480896995275e-06,
+      "loss": 1.0686,
+      "step": 789
+    },
+    {
+      "epoch": 0.67,
+      "learning_rate": 9.548438787829946e-06,
+      "loss": 1.3291,
+      "step": 790
+    },
+    {
+      "epoch": 0.67,
+      "learning_rate": 9.546392269985842e-06,
+      "loss": 1.1261,
+      "step": 791
+    },
+    {
+      "epoch": 0.67,
+      "learning_rate": 9.544341345446599e-06,
+      "loss": 1.132,
+      "step": 792
+    },
+    {
+      "epoch": 0.67,
+      "learning_rate": 9.542286016200122e-06,
+      "loss": 1.3352,
+      "step": 793
+    },
+    {
+      "epoch": 0.67,
+      "learning_rate": 9.540226284238585e-06,
+      "loss": 1.2645,
+      "step": 794
+    },
+    {
+      "epoch": 0.67,
+      "learning_rate": 9.538162151558434e-06,
+      "loss": 1.0958,
+      "step": 795
+    },
+    {
+      "epoch": 0.67,
+      "learning_rate": 9.536093620160372e-06,
+      "loss": 1.1581,
+      "step": 796
+    },
+    {
+      "epoch": 0.67,
+      "learning_rate": 9.534020692049374e-06,
+      "loss": 1.2218,
+      "step": 797
+    },
+    {
+      "epoch": 0.67,
+      "learning_rate": 9.531943369234674e-06,
+      "loss": 1.1858,
+      "step": 798
+    },
+    {
+      "epoch": 0.68,
+      "learning_rate": 9.529861653729763e-06,
+      "loss": 1.3329,
+      "step": 799
+    },
+    {
+      "epoch": 0.68,
+      "learning_rate": 9.52777554755239e-06,
+      "loss": 1.2955,
+      "step": 800
+    },
+    {
+      "epoch": 0.68,
+      "learning_rate": 9.525685052724565e-06,
+      "loss": 1.2115,
+      "step": 801
+    },
+    {
+      "epoch": 0.68,
+      "learning_rate": 9.523590171272546e-06,
+      "loss": 1.0408,
+      "step": 802
+    },
+    {
+      "epoch": 0.68,
+      "learning_rate": 9.521490905226844e-06,
+      "loss": 1.2728,
+      "step": 803
+    },
+    {
+      "epoch": 0.68,
+      "learning_rate": 9.519387256622223e-06,
+      "loss": 0.8925,
+      "step": 804
+    },
+    {
+      "epoch": 0.68,
+      "learning_rate": 9.517279227497692e-06,
+      "loss": 1.021,
+      "step": 805
+    },
+    {
+      "epoch": 0.68,
+      "learning_rate": 9.515166819896506e-06,
+      "loss": 1.0705,
+      "step": 806
+    },
+    {
+      "epoch": 0.68,
+      "learning_rate": 9.513050035866165e-06,
+      "loss": 1.0632,
+      "step": 807
+    },
+    {
+      "epoch": 0.68,
+      "learning_rate": 9.510928877458412e-06,
+      "loss": 1.1912,
+      "step": 808
+    },
+    {
+      "epoch": 0.68,
+      "learning_rate": 9.508803346729226e-06,
+      "loss": 1.1943,
+      "step": 809
+    },
+    {
+      "epoch": 0.68,
+      "learning_rate": 9.506673445738828e-06,
+      "loss": 1.0524,
+      "step": 810
+    },
+    {
+      "epoch": 0.69,
+      "learning_rate": 9.504539176551672e-06,
+      "loss": 1.1134,
+      "step": 811
+    },
+    {
+      "epoch": 0.69,
+      "learning_rate": 9.502400541236451e-06,
+      "loss": 1.0985,
+      "step": 812
+    },
+    {
+      "epoch": 0.69,
+      "learning_rate": 9.500257541866082e-06,
+      "loss": 1.1162,
+      "step": 813
+    },
+    {
+      "epoch": 0.69,
+      "learning_rate": 9.498110180517719e-06,
+      "loss": 1.3487,
+      "step": 814
+    },
+    {
+      "epoch": 0.69,
+      "learning_rate": 9.49595845927274e-06,
+      "loss": 1.0569,
+      "step": 815
+    },
+    {
+      "epoch": 0.69,
+      "learning_rate": 9.49380238021675e-06,
+      "loss": 1.2644,
+      "step": 816
+    },
+    {
+      "epoch": 0.69,
+      "learning_rate": 9.49164194543958e-06,
+      "loss": 1.0979,
+      "step": 817
+    },
+    {
+      "epoch": 0.69,
+      "learning_rate": 9.48947715703528e-06,
+      "loss": 1.231,
+      "step": 818
+    },
+    {
+      "epoch": 0.69,
+      "learning_rate": 9.487308017102122e-06,
+      "loss": 1.1796,
+      "step": 819
+    },
+    {
+      "epoch": 0.69,
+      "learning_rate": 9.485134527742592e-06,
+      "loss": 1.2198,
+      "step": 820
+    },
+    {
+      "epoch": 0.69,
+      "learning_rate": 9.482956691063396e-06,
+      "loss": 1.1496,
+      "step": 821
+    },
+    {
+      "epoch": 0.7,
+      "learning_rate": 9.480774509175453e-06,
+      "loss": 1.4714,
+      "step": 822
+    },
+    {
+      "epoch": 0.7,
+      "learning_rate": 9.478587984193892e-06,
+      "loss": 1.1229,
+      "step": 823
+    },
+    {
+      "epoch": 0.7,
+      "learning_rate": 9.476397118238054e-06,
+      "loss": 1.1059,
+      "step": 824
+    },
+    {
+      "epoch": 0.7,
+      "learning_rate": 9.474201913431483e-06,
+      "loss": 1.2096,
+      "step": 825
+    },
+    {
+      "epoch": 0.7,
+      "learning_rate": 9.472002371901935e-06,
+      "loss": 1.1935,
+      "step": 826
+    },
+    {
+      "epoch": 0.7,
+      "learning_rate": 9.469798495781364e-06,
+      "loss": 1.0894,
+      "step": 827
+    },
+    {
+      "epoch": 0.7,
+      "learning_rate": 9.46759028720593e-06,
+      "loss": 1.2556,
+      "step": 828
+    },
+    {
+      "epoch": 0.7,
+      "learning_rate": 9.465377748315987e-06,
+      "loss": 1.2763,
+      "step": 829
+    },
+    {
+      "epoch": 0.7,
+      "learning_rate": 9.463160881256089e-06,
+      "loss": 1.0416,
+      "step": 830
+    },
+    {
+      "epoch": 0.7,
+      "learning_rate": 9.460939688174989e-06,
+      "loss": 1.0384,
+      "step": 831
+    },
+    {
+      "epoch": 0.7,
+      "learning_rate": 9.458714171225626e-06,
+      "loss": 1.1414,
+      "step": 832
+    },
+    {
+      "epoch": 0.7,
+      "learning_rate": 9.456484332565135e-06,
+      "loss": 1.3063,
+      "step": 833
+    },
+    {
+      "epoch": 0.71,
+      "learning_rate": 9.454250174354838e-06,
+      "loss": 1.0993,
+      "step": 834
+    },
+    {
+      "epoch": 0.71,
+      "learning_rate": 9.452011698760242e-06,
+      "loss": 0.9598,
+      "step": 835
+    },
+    {
+      "epoch": 0.71,
+      "learning_rate": 9.449768907951045e-06,
+      "loss": 1.186,
+      "step": 836
+    },
+    {
+      "epoch": 0.71,
+      "learning_rate": 9.447521804101121e-06,
+      "loss": 1.2067,
+      "step": 837
+    },
+    {
+      "epoch": 0.71,
+      "learning_rate": 9.445270389388527e-06,
+      "loss": 1.2154,
+      "step": 838
+    },
+    {
+      "epoch": 0.71,
+      "learning_rate": 9.4430146659955e-06,
+      "loss": 1.0513,
+      "step": 839
+    },
+    {
+      "epoch": 0.71,
+      "learning_rate": 9.44075463610845e-06,
+      "loss": 1.1081,
+      "step": 840
+    },
+    {
+      "epoch": 0.71,
+      "learning_rate": 9.438490301917964e-06,
+      "loss": 1.0729,
+      "step": 841
+    },
+    {
+      "epoch": 0.71,
+      "learning_rate": 9.436221665618799e-06,
+      "loss": 1.1742,
+      "step": 842
+    },
+    {
+      "epoch": 0.71,
+      "learning_rate": 9.433948729409882e-06,
+      "loss": 1.0925,
+      "step": 843
+    },
+    {
+      "epoch": 0.71,
+      "learning_rate": 9.431671495494312e-06,
+      "loss": 1.1372,
+      "step": 844
+    },
+    {
+      "epoch": 0.71,
+      "learning_rate": 9.429389966079347e-06,
+      "loss": 1.027,
+      "step": 845
+    },
+    {
+      "epoch": 0.72,
+      "learning_rate": 9.427104143376414e-06,
+      "loss": 1.0007,
+      "step": 846
+    },
+    {
+      "epoch": 0.72,
+      "learning_rate": 9.424814029601096e-06,
+      "loss": 1.1123,
+      "step": 847
+    },
+    {
+      "epoch": 0.72,
+      "learning_rate": 9.42251962697314e-06,
+      "loss": 1.1191,
+      "step": 848
+    },
+    {
+      "epoch": 0.72,
+      "learning_rate": 9.420220937716452e-06,
+      "loss": 1.063,
+      "step": 849
+    },
+    {
+      "epoch": 0.72,
+      "learning_rate": 9.417917964059085e-06,
+      "loss": 1.2922,
+      "step": 850
+    },
+    {
+      "epoch": 0.72,
+      "learning_rate": 9.415610708233252e-06,
+      "loss": 1.003,
+      "step": 851
+    },
+    {
+      "epoch": 0.72,
+      "learning_rate": 9.41329917247531e-06,
+      "loss": 1.1363,
+      "step": 852
+    },
+    {
+      "epoch": 0.72,
+      "learning_rate": 9.41098335902577e-06,
+      "loss": 1.2522,
+      "step": 853
+    },
+    {
+      "epoch": 0.72,
+      "learning_rate": 9.408663270129289e-06,
+      "loss": 1.2031,
+      "step": 854
+    },
+    {
+      "epoch": 0.72,
+      "learning_rate": 9.406338908034665e-06,
+      "loss": 0.9692,
+      "step": 855
+    },
+    {
+      "epoch": 0.72,
+      "learning_rate": 9.40401027499484e-06,
+      "loss": 1.101,
+      "step": 856
+    },
+    {
+      "epoch": 0.72,
+      "learning_rate": 9.401677373266892e-06,
+      "loss": 1.3211,
+      "step": 857
+    },
+    {
+      "epoch": 0.73,
+      "learning_rate": 9.399340205112045e-06,
+      "loss": 1.0759,
+      "step": 858
+    },
+    {
+      "epoch": 0.73,
+      "learning_rate": 9.396998772795647e-06,
+      "loss": 1.1563,
+      "step": 859
+    },
+    {
+      "epoch": 0.73,
+      "learning_rate": 9.39465307858719e-06,
+      "loss": 1.1472,
+      "step": 860
+    },
+    {
+      "epoch": 0.73,
+      "learning_rate": 9.39230312476029e-06,
+      "loss": 1.0359,
+      "step": 861
+    },
+    {
+      "epoch": 0.73,
+      "learning_rate": 9.389948913592694e-06,
+      "loss": 1.3914,
+      "step": 862
+    },
+    {
+      "epoch": 0.73,
+      "learning_rate": 9.387590447366274e-06,
+      "loss": 1.1398,
+      "step": 863
+    },
+    {
+      "epoch": 0.73,
+      "learning_rate": 9.38522772836703e-06,
+      "loss": 1.1828,
+      "step": 864
+    },
+    {
+      "epoch": 0.73,
+      "learning_rate": 9.38286075888508e-06,
+      "loss": 1.2239,
+      "step": 865
+    },
+    {
+      "epoch": 0.73,
+      "learning_rate": 9.380489541214664e-06,
+      "loss": 1.0,
+      "step": 866
+    },
+    {
+      "epoch": 0.73,
+      "learning_rate": 9.378114077654138e-06,
+      "loss": 1.0441,
+      "step": 867
+    },
+    {
+      "epoch": 0.73,
+      "learning_rate": 9.375734370505976e-06,
+      "loss": 1.149,
+      "step": 868
+    },
+    {
+      "epoch": 0.73,
+      "learning_rate": 9.373350422076764e-06,
+      "loss": 1.1879,
+      "step": 869
+    },
+    {
+      "epoch": 0.74,
+      "learning_rate": 9.370962234677198e-06,
+      "loss": 1.167,
+      "step": 870
+    },
+    {
+      "epoch": 0.74,
+      "learning_rate": 9.368569810622083e-06,
+      "loss": 1.4329,
+      "step": 871
+    },
+    {
+      "epoch": 0.74,
+      "learning_rate": 9.366173152230333e-06,
+      "loss": 1.1271,
+      "step": 872
+    },
+    {
+      "epoch": 0.74,
+      "learning_rate": 9.363772261824962e-06,
+      "loss": 1.1377,
+      "step": 873
+    },
+    {
+      "epoch": 0.74,
+      "learning_rate": 9.36136714173309e-06,
+      "loss": 1.3837,
+      "step": 874
+    },
+    {
+      "epoch": 0.74,
+      "learning_rate": 9.358957794285933e-06,
+      "loss": 1.0266,
+      "step": 875
+    },
+    {
+      "epoch": 0.74,
+      "learning_rate": 9.356544221818806e-06,
+      "loss": 1.187,
+      "step": 876
+    },
+    {
+      "epoch": 0.74,
+      "learning_rate": 9.354126426671124e-06,
+      "loss": 1.1631,
+      "step": 877
+    },
+    {
+      "epoch": 0.74,
+      "learning_rate": 9.351704411186384e-06,
+      "loss": 1.216,
+      "step": 878
+    },
+    {
+      "epoch": 0.74,
+      "learning_rate": 9.349278177712187e-06,
+      "loss": 1.0426,
+      "step": 879
+    },
+    {
+      "epoch": 0.74,
+      "learning_rate": 9.34684772860021e-06,
+      "loss": 1.1462,
+      "step": 880
+    },
+    {
+      "epoch": 0.75,
+      "learning_rate": 9.344413066206223e-06,
+      "loss": 1.216,
+      "step": 881
+    },
+    {
+      "epoch": 0.75,
+      "learning_rate": 9.34197419289008e-06,
+      "loss": 0.9109,
+      "step": 882
+    },
+    {
+      "epoch": 0.75,
+      "learning_rate": 9.339531111015713e-06,
+      "loss": 1.0075,
+      "step": 883
+    },
+    {
+      "epoch": 0.75,
+      "learning_rate": 9.337083822951139e-06,
+      "loss": 1.1929,
+      "step": 884
+    },
+    {
+      "epoch": 0.75,
+      "learning_rate": 9.334632331068445e-06,
+      "loss": 1.1048,
+      "step": 885
+    },
+    {
+      "epoch": 0.75,
+      "learning_rate": 9.332176637743796e-06,
+      "loss": 1.3165,
+      "step": 886
+    },
+    {
+      "epoch": 0.75,
+      "learning_rate": 9.329716745357427e-06,
+      "loss": 0.9306,
+      "step": 887
+    },
+    {
+      "epoch": 0.75,
+      "learning_rate": 9.32725265629365e-06,
+      "loss": 1.1577,
+      "step": 888
+    },
+    {
+      "epoch": 0.75,
+      "learning_rate": 9.324784372940836e-06,
+      "loss": 1.271,
+      "step": 889
+    },
+    {
+      "epoch": 0.75,
+      "learning_rate": 9.322311897691427e-06,
+      "loss": 0.9726,
+      "step": 890
+    },
+    {
+      "epoch": 0.75,
+      "learning_rate": 9.319835232941929e-06,
+      "loss": 1.149,
+      "step": 891
+    },
+    {
+      "epoch": 0.75,
+      "learning_rate": 9.317354381092902e-06,
+      "loss": 1.315,
+      "step": 892
+    },
+    {
+      "epoch": 0.76,
+      "learning_rate": 9.314869344548969e-06,
+      "loss": 1.0514,
+      "step": 893
+    },
+    {
+      "epoch": 0.76,
+      "learning_rate": 9.312380125718811e-06,
+      "loss": 1.2799,
+      "step": 894
+    },
+    {
+      "epoch": 0.76,
+      "learning_rate": 9.30988672701516e-06,
+      "loss": 0.9576,
+      "step": 895
+    },
+    {
+      "epoch": 0.76,
+      "learning_rate": 9.3073891508548e-06,
+      "loss": 1.3859,
+      "step": 896
+    },
+    {
+      "epoch": 0.76,
+      "learning_rate": 9.304887399658562e-06,
+      "loss": 1.2188,
+      "step": 897
+    },
+    {
+      "epoch": 0.76,
+      "learning_rate": 9.302381475851328e-06,
+      "loss": 1.0086,
+      "step": 898
+    },
+    {
+      "epoch": 0.76,
+      "learning_rate": 9.29987138186202e-06,
+      "loss": 1.2477,
+      "step": 899
+    },
+    {
+      "epoch": 0.76,
+      "learning_rate": 9.297357120123606e-06,
+      "loss": 1.0726,
+      "step": 900
+    },
+    {
+      "epoch": 0.76,
+      "learning_rate": 9.294838693073093e-06,
+      "loss": 1.1705,
+      "step": 901
+    },
+    {
+      "epoch": 0.76,
+      "learning_rate": 9.292316103151524e-06,
+      "loss": 1.1084,
+      "step": 902
+    },
+    {
+      "epoch": 0.76,
+      "learning_rate": 9.289789352803975e-06,
+      "loss": 1.1051,
+      "step": 903
+    },
+    {
+      "epoch": 0.76,
+      "learning_rate": 9.287258444479558e-06,
+      "loss": 1.105,
+      "step": 904
+    },
+    {
+      "epoch": 0.77,
+      "learning_rate": 9.284723380631415e-06,
+      "loss": 1.2598,
+      "step": 905
+    },
+    {
+      "epoch": 0.77,
+      "learning_rate": 9.282184163716712e-06,
+      "loss": 0.9921,
+      "step": 906
+    },
+    {
+      "epoch": 0.77,
+      "learning_rate": 9.279640796196647e-06,
+      "loss": 1.0443,
+      "step": 907
+    },
+    {
+      "epoch": 0.77,
+      "learning_rate": 9.277093280536435e-06,
+      "loss": 1.293,
+      "step": 908
+    },
+    {
+      "epoch": 0.77,
+      "learning_rate": 9.274541619205314e-06,
+      "loss": 1.1358,
+      "step": 909
+    },
+    {
+      "epoch": 0.77,
+      "learning_rate": 9.271985814676542e-06,
+      "loss": 1.1665,
+      "step": 910
+    },
+    {
+      "epoch": 0.77,
+      "learning_rate": 9.26942586942739e-06,
+      "loss": 1.0294,
+      "step": 911
+    },
+    {
+      "epoch": 0.77,
+      "learning_rate": 9.266861785939145e-06,
+      "loss": 0.965,
+      "step": 912
+    },
+    {
+      "epoch": 0.77,
+      "learning_rate": 9.264293566697103e-06,
+      "loss": 1.0344,
+      "step": 913
+    },
+    {
+      "epoch": 0.77,
+      "learning_rate": 9.261721214190567e-06,
+      "loss": 1.2077,
+      "step": 914
+    },
+    {
+      "epoch": 0.77,
+      "learning_rate": 9.259144730912853e-06,
+      "loss": 0.9496,
+      "step": 915
+    },
+    {
+      "epoch": 0.77,
+      "learning_rate": 9.256564119361276e-06,
+      "loss": 1.1303,
+      "step": 916
+    },
+    {
+      "epoch": 0.78,
+      "learning_rate": 9.253979382037149e-06,
+      "loss": 1.2029,
+      "step": 917
+    },
+    {
+      "epoch": 0.78,
+      "learning_rate": 9.251390521445792e-06,
+      "loss": 1.1028,
+      "step": 918
+    },
+    {
+      "epoch": 0.78,
+      "learning_rate": 9.248797540096515e-06,
+      "loss": 1.4242,
+      "step": 919
+    },
+    {
+      "epoch": 0.78,
+      "learning_rate": 9.246200440502628e-06,
+      "loss": 1.2078,
+      "step": 920
+    },
+    {
+      "epoch": 0.78,
+      "learning_rate": 9.243599225181428e-06,
+      "loss": 1.1104,
+      "step": 921
+    },
+    {
+      "epoch": 0.78,
+      "learning_rate": 9.2409938966542e-06,
+      "loss": 1.1803,
+      "step": 922
+    },
+    {
+      "epoch": 0.78,
+      "learning_rate": 9.238384457446222e-06,
+      "loss": 1.1503,
+      "step": 923
+    },
+    {
+      "epoch": 0.78,
+      "learning_rate": 9.235770910086753e-06,
+      "loss": 1.0537,
+      "step": 924
+    },
+    {
+      "epoch": 0.78,
+      "learning_rate": 9.23315325710903e-06,
+      "loss": 1.2245,
+      "step": 925
+    },
+    {
+      "epoch": 0.78,
+      "learning_rate": 9.230531501050277e-06,
+      "loss": 1.0824,
+      "step": 926
+    },
+    {
+      "epoch": 0.78,
+      "learning_rate": 9.22790564445169e-06,
+      "loss": 1.1142,
+      "step": 927
+    },
+    {
+      "epoch": 0.78,
+      "learning_rate": 9.22527568985844e-06,
+      "loss": 0.9585,
+      "step": 928
+    },
+    {
+      "epoch": 0.79,
+      "learning_rate": 9.222641639819672e-06,
+      "loss": 1.1498,
+      "step": 929
+    },
+    {
+      "epoch": 0.79,
+      "learning_rate": 9.220003496888502e-06,
+      "loss": 1.1592,
+      "step": 930
+    },
+    {
+      "epoch": 0.79,
+      "learning_rate": 9.217361263622004e-06,
+      "loss": 1.0368,
+      "step": 931
+    },
+    {
+      "epoch": 0.79,
+      "learning_rate": 9.214714942581228e-06,
+      "loss": 1.0401,
+      "step": 932
+    },
+    {
+      "epoch": 0.79,
+      "learning_rate": 9.212064536331182e-06,
+      "loss": 0.9658,
+      "step": 933
+    },
+    {
+      "epoch": 0.79,
+      "learning_rate": 9.20941004744083e-06,
+      "loss": 1.1442,
+      "step": 934
+    },
+    {
+      "epoch": 0.79,
+      "learning_rate": 9.2067514784831e-06,
+      "loss": 1.2851,
+      "step": 935
+    },
+    {
+      "epoch": 0.79,
+      "learning_rate": 9.204088832034868e-06,
+      "loss": 0.9718,
+      "step": 936
+    },
+    {
+      "epoch": 0.79,
+      "learning_rate": 9.201422110676967e-06,
+      "loss": 1.1958,
+      "step": 937
+    },
+    {
+      "epoch": 0.79,
+      "learning_rate": 9.198751316994178e-06,
+      "loss": 0.9945,
+      "step": 938
+    },
+    {
+      "epoch": 0.79,
+      "learning_rate": 9.196076453575226e-06,
+      "loss": 1.2643,
+      "step": 939
+    },
+    {
+      "epoch": 0.79,
+      "learning_rate": 9.19339752301279e-06,
+      "loss": 1.1821,
+      "step": 940
+    },
+    {
+      "epoch": 0.8,
+      "learning_rate": 9.190714527903477e-06,
+      "loss": 1.1818,
+      "step": 941
+    },
+    {
+      "epoch": 0.8,
+      "learning_rate": 9.18802747084785e-06,
+      "loss": 1.0277,
+      "step": 942
+    },
+    {
+      "epoch": 0.8,
+      "learning_rate": 9.185336354450393e-06,
+      "loss": 1.1213,
+      "step": 943
+    },
+    {
+      "epoch": 0.8,
+      "learning_rate": 9.182641181319538e-06,
+      "loss": 1.27,
+      "step": 944
+    },
+    {
+      "epoch": 0.8,
+      "eval_loss": 1.0539518594741821,
+      "eval_runtime": 268.6045,
+      "eval_samples_per_second": 3.719,
+      "eval_steps_per_second": 0.465,
+      "step": 944
+    },
+    {
+      "epoch": 0.8,
+      "learning_rate": 9.179941954067641e-06,
+      "loss": 1.1262,
+      "step": 945
+    },
+    {
+      "epoch": 0.8,
+      "learning_rate": 9.177238675310991e-06,
+      "loss": 1.0239,
+      "step": 946
+    },
+    {
+      "epoch": 0.8,
+      "learning_rate": 9.174531347669801e-06,
+      "loss": 1.0628,
+      "step": 947
+    },
+    {
+      "epoch": 0.8,
+      "learning_rate": 9.171819973768213e-06,
+      "loss": 1.2047,
+      "step": 948
+    },
+    {
+      "epoch": 0.8,
+      "learning_rate": 9.169104556234288e-06,
+      "loss": 1.2009,
+      "step": 949
+    },
+    {
+      "epoch": 0.8,
+      "learning_rate": 9.166385097700005e-06,
+      "loss": 1.0352,
+      "step": 950
+    },
+    {
+      "epoch": 0.8,
+      "learning_rate": 9.163661600801264e-06,
+      "loss": 1.1939,
+      "step": 951
+    },
+    {
+      "epoch": 0.81,
+      "learning_rate": 9.160934068177875e-06,
+      "loss": 0.9893,
+      "step": 952
+    },
+    {
+      "epoch": 0.81,
+      "learning_rate": 9.158202502473563e-06,
+      "loss": 1.1192,
+      "step": 953
+    },
+    {
+      "epoch": 0.81,
+      "learning_rate": 9.155466906335961e-06,
+      "loss": 1.1379,
+      "step": 954
+    },
+    {
+      "epoch": 0.81,
+      "learning_rate": 9.152727282416607e-06,
+      "loss": 1.0479,
+      "step": 955
+    },
+    {
+      "epoch": 0.81,
+      "learning_rate": 9.149983633370945e-06,
+      "loss": 1.1274,
+      "step": 956
+    },
+    {
+      "epoch": 0.81,
+      "learning_rate": 9.147235961858321e-06,
+      "loss": 1.1024,
+      "step": 957
+    },
+    {
+      "epoch": 0.81,
+      "learning_rate": 9.144484270541976e-06,
+      "loss": 0.9003,
+      "step": 958
+    },
+    {
+      "epoch": 0.81,
+      "learning_rate": 9.141728562089054e-06,
+      "loss": 1.122,
+      "step": 959
+    },
+    {
+      "epoch": 0.81,
+      "learning_rate": 9.138968839170585e-06,
+      "loss": 1.0116,
+      "step": 960
+    },
+    {
+      "epoch": 0.81,
+      "learning_rate": 9.136205104461495e-06,
+      "loss": 1.4711,
+      "step": 961
+    },
+    {
+      "epoch": 0.81,
+      "learning_rate": 9.133437360640597e-06,
+      "loss": 1.2094,
+      "step": 962
+    },
+    {
+      "epoch": 0.81,
+      "learning_rate": 9.130665610390593e-06,
+      "loss": 1.163,
+      "step": 963
+    },
+    {
+      "epoch": 0.82,
+      "learning_rate": 9.127889856398063e-06,
+      "loss": 1.205,
+      "step": 964
+    },
+    {
+      "epoch": 0.82,
+      "learning_rate": 9.12511010135347e-06,
+      "loss": 1.0823,
+      "step": 965
+    },
+    {
+      "epoch": 0.82,
+      "learning_rate": 9.122326347951155e-06,
+      "loss": 1.0909,
+      "step": 966
+    },
+    {
+      "epoch": 0.82,
+      "learning_rate": 9.11953859888934e-06,
+      "loss": 0.9585,
+      "step": 967
+    },
+    {
+      "epoch": 0.82,
+      "learning_rate": 9.116746856870106e-06,
+      "loss": 1.2888,
+      "step": 968
+    },
+    {
+      "epoch": 0.82,
+      "learning_rate": 9.113951124599423e-06,
+      "loss": 1.1315,
+      "step": 969
+    },
+    {
+      "epoch": 0.82,
+      "learning_rate": 9.111151404787111e-06,
+      "loss": 1.2,
+      "step": 970
+    },
+    {
+      "epoch": 0.82,
+      "learning_rate": 9.108347700146866e-06,
+      "loss": 1.0591,
+      "step": 971
+    },
+    {
+      "epoch": 0.82,
+      "learning_rate": 9.105540013396243e-06,
+      "loss": 1.2939,
+      "step": 972
+    },
+    {
+      "epoch": 0.82,
+      "learning_rate": 9.102728347256656e-06,
+      "loss": 1.1627,
+      "step": 973
+    },
+    {
+      "epoch": 0.82,
+      "learning_rate": 9.099912704453377e-06,
+      "loss": 1.1061,
+      "step": 974
+    },
+    {
+      "epoch": 0.82,
+      "learning_rate": 9.097093087715534e-06,
+      "loss": 1.2272,
+      "step": 975
+    },
+    {
+      "epoch": 0.83,
+      "learning_rate": 9.094269499776106e-06,
+      "loss": 1.0806,
+      "step": 976
+    },
+    {
+      "epoch": 0.83,
+      "learning_rate": 9.091441943371914e-06,
+      "loss": 1.4281,
+      "step": 977
+    },
+    {
+      "epoch": 0.83,
+      "learning_rate": 9.08861042124364e-06,
+      "loss": 1.0855,
+      "step": 978
+    },
+    {
+      "epoch": 0.83,
+      "learning_rate": 9.085774936135798e-06,
+      "loss": 1.0476,
+      "step": 979
+    },
+    {
+      "epoch": 0.83,
+      "learning_rate": 9.08293549079675e-06,
+      "loss": 1.0993,
+      "step": 980
+    },
+    {
+      "epoch": 0.83,
+      "learning_rate": 9.080092087978688e-06,
+      "loss": 1.1712,
+      "step": 981
+    },
+    {
+      "epoch": 0.83,
+      "learning_rate": 9.07724473043765e-06,
+      "loss": 1.0634,
+      "step": 982
+    },
+    {
+      "epoch": 0.83,
+      "learning_rate": 9.0743934209335e-06,
+      "loss": 1.1836,
+      "step": 983
+    },
+    {
+      "epoch": 0.83,
+      "learning_rate": 9.071538162229939e-06,
+      "loss": 1.0579,
+      "step": 984
+    },
+    {
+      "epoch": 0.83,
+      "learning_rate": 9.068678957094487e-06,
+      "loss": 1.1601,
+      "step": 985
+    },
+    {
+      "epoch": 0.83,
+      "learning_rate": 9.0658158082985e-06,
+      "loss": 1.1144,
+      "step": 986
+    },
+    {
+      "epoch": 0.83,
+      "learning_rate": 9.062948718617147e-06,
+      "loss": 1.0275,
+      "step": 987
+    },
+    {
+      "epoch": 0.84,
+      "learning_rate": 9.060077690829421e-06,
+      "loss": 1.118,
+      "step": 988
+    },
+    {
+      "epoch": 0.84,
+      "learning_rate": 9.057202727718134e-06,
+      "loss": 1.1003,
+      "step": 989
+    },
+    {
+      "epoch": 0.84,
+      "learning_rate": 9.054323832069904e-06,
+      "loss": 1.3383,
+      "step": 990
+    },
+    {
+      "epoch": 0.84,
+      "learning_rate": 9.051441006675172e-06,
+      "loss": 1.1776,
+      "step": 991
+    },
+    {
+      "epoch": 0.84,
+      "learning_rate": 9.048554254328184e-06,
+      "loss": 1.0921,
+      "step": 992
+    },
+    {
+      "epoch": 0.84,
+      "learning_rate": 9.045663577826987e-06,
+      "loss": 1.1837,
+      "step": 993
+    },
+    {
+      "epoch": 0.84,
+      "learning_rate": 9.042768979973435e-06,
+      "loss": 0.9844,
+      "step": 994
+    },
+    {
+      "epoch": 0.84,
+      "learning_rate": 9.039870463573188e-06,
+      "loss": 0.9682,
+      "step": 995
+    },
+    {
+      "epoch": 0.84,
+      "learning_rate": 9.036968031435697e-06,
+      "loss": 0.9711,
+      "step": 996
+    },
+    {
+      "epoch": 0.84,
+      "learning_rate": 9.03406168637421e-06,
+      "loss": 1.1953,
+      "step": 997
+    },
+    {
+      "epoch": 0.84,
+      "learning_rate": 9.031151431205771e-06,
+      "loss": 1.0843,
+      "step": 998
+    },
+    {
+      "epoch": 0.84,
+      "learning_rate": 9.02823726875121e-06,
+      "loss": 1.0642,
+      "step": 999
+    },
+    {
+      "epoch": 0.85,
+      "learning_rate": 9.025319201835146e-06,
+      "loss": 1.0711,
+      "step": 1000
+    },
+    {
+      "epoch": 0.85,
+      "learning_rate": 9.022397233285983e-06,
+      "loss": 1.1597,
+      "step": 1001
+    },
+    {
+      "epoch": 0.85,
+      "learning_rate": 9.019471365935907e-06,
+      "loss": 1.2079,
+      "step": 1002
+    },
+    {
+      "epoch": 0.85,
+      "learning_rate": 9.01654160262088e-06,
+      "loss": 0.9484,
+      "step": 1003
+    },
+    {
+      "epoch": 0.85,
+      "learning_rate": 9.013607946180644e-06,
+      "loss": 0.9462,
+      "step": 1004
+    },
+    {
+      "epoch": 0.85,
+      "learning_rate": 9.010670399458711e-06,
+      "loss": 1.1408,
+      "step": 1005
+    },
+    {
+      "epoch": 0.85,
+      "learning_rate": 9.007728965302369e-06,
+      "loss": 1.0795,
+      "step": 1006
+    },
+    {
+      "epoch": 0.85,
+      "learning_rate": 9.004783646562669e-06,
+      "loss": 1.0688,
+      "step": 1007
+    },
+    {
+      "epoch": 0.85,
+      "learning_rate": 9.001834446094427e-06,
+      "loss": 1.0623,
+      "step": 1008
+    },
+    {
+      "epoch": 0.85,
+      "learning_rate": 8.998881366756224e-06,
+      "loss": 1.3446,
+      "step": 1009
+    },
+    {
+      "epoch": 0.85,
+      "learning_rate": 8.995924411410405e-06,
+      "loss": 1.0849,
+      "step": 1010
+    },
+    {
+      "epoch": 0.85,
+      "learning_rate": 8.992963582923057e-06,
+      "loss": 1.1452,
+      "step": 1011
+    },
+    {
+      "epoch": 0.86,
+      "learning_rate": 8.989998884164039e-06,
+      "loss": 1.1624,
+      "step": 1012
+    },
+    {
+      "epoch": 0.86,
+      "learning_rate": 8.987030318006952e-06,
+      "loss": 1.0914,
+      "step": 1013
+    },
+    {
+      "epoch": 0.86,
+      "learning_rate": 8.984057887329145e-06,
+      "loss": 1.1951,
+      "step": 1014
+    },
+    {
+      "epoch": 0.86,
+      "learning_rate": 8.981081595011717e-06,
+      "loss": 1.2496,
+      "step": 1015
+    },
+    {
+      "epoch": 0.86,
+      "learning_rate": 8.978101443939506e-06,
+      "loss": 1.0267,
+      "step": 1016
+    },
+    {
+      "epoch": 0.86,
+      "learning_rate": 8.97511743700109e-06,
+      "loss": 1.1496,
+      "step": 1017
+    },
+    {
+      "epoch": 0.86,
+      "learning_rate": 8.972129577088791e-06,
+      "loss": 1.1033,
+      "step": 1018
+    },
+    {
+      "epoch": 0.86,
+      "learning_rate": 8.969137867098658e-06,
+      "loss": 1.3013,
+      "step": 1019
+    },
+    {
+      "epoch": 0.86,
+      "learning_rate": 8.966142309930474e-06,
+      "loss": 1.0801,
+      "step": 1020
+    },
+    {
+      "epoch": 0.86,
+      "learning_rate": 8.963142908487755e-06,
+      "loss": 1.1514,
+      "step": 1021
+    },
+    {
+      "epoch": 0.86,
+      "learning_rate": 8.960139665677734e-06,
+      "loss": 1.116,
+      "step": 1022
+    },
+    {
+      "epoch": 0.87,
+      "learning_rate": 8.957132584411378e-06,
+      "loss": 1.1385,
+      "step": 1023
+    },
+    {
+      "epoch": 0.87,
+      "learning_rate": 8.954121667603368e-06,
+      "loss": 1.0869,
+      "step": 1024
+    },
+    {
+      "epoch": 0.87,
+      "learning_rate": 8.951106918172105e-06,
+      "loss": 1.1665,
+      "step": 1025
+    },
+    {
+      "epoch": 0.87,
+      "learning_rate": 8.948088339039703e-06,
+      "loss": 1.0617,
+      "step": 1026
+    },
+    {
+      "epoch": 0.87,
+      "learning_rate": 8.94506593313199e-06,
+      "loss": 0.9255,
+      "step": 1027
+    },
+    {
+      "epoch": 0.87,
+      "learning_rate": 8.942039703378502e-06,
+      "loss": 1.2122,
+      "step": 1028
+    },
+    {
+      "epoch": 0.87,
+      "learning_rate": 8.939009652712481e-06,
+      "loss": 1.1291,
+      "step": 1029
+    },
+    {
+      "epoch": 0.87,
+      "learning_rate": 8.935975784070875e-06,
+      "loss": 1.2508,
+      "step": 1030
+    },
+    {
+      "epoch": 0.87,
+      "learning_rate": 8.932938100394329e-06,
+      "loss": 1.1186,
+      "step": 1031
+    },
+    {
+      "epoch": 0.87,
+      "learning_rate": 8.929896604627186e-06,
+      "loss": 1.1314,
+      "step": 1032
+    },
+    {
+      "epoch": 0.87,
+      "learning_rate": 8.926851299717492e-06,
+      "loss": 1.0752,
+      "step": 1033
+    },
+    {
+      "epoch": 0.87,
+      "learning_rate": 8.923802188616973e-06,
+      "loss": 1.2764,
+      "step": 1034
+    },
+    {
+      "epoch": 0.88,
+      "learning_rate": 8.92074927428105e-06,
+      "loss": 1.2321,
+      "step": 1035
+    },
+    {
+      "epoch": 0.88,
+      "learning_rate": 8.917692559668835e-06,
+      "loss": 0.9983,
+      "step": 1036
+    },
+    {
+      "epoch": 0.88,
+      "learning_rate": 8.914632047743114e-06,
+      "loss": 1.0619,
+      "step": 1037
+    },
+    {
+      "epoch": 0.88,
+      "learning_rate": 8.911567741470361e-06,
+      "loss": 1.0057,
+      "step": 1038
+    },
+    {
+      "epoch": 0.88,
+      "learning_rate": 8.908499643820726e-06,
+      "loss": 1.2063,
+      "step": 1039
+    },
+    {
+      "epoch": 0.88,
+      "learning_rate": 8.905427757768031e-06,
+      "loss": 1.157,
+      "step": 1040
+    },
+    {
+      "epoch": 0.88,
+      "learning_rate": 8.902352086289774e-06,
+      "loss": 0.9421,
+      "step": 1041
+    },
+    {
+      "epoch": 0.88,
+      "learning_rate": 8.899272632367117e-06,
+      "loss": 1.2596,
+      "step": 1042
+    },
+    {
+      "epoch": 0.88,
+      "learning_rate": 8.896189398984895e-06,
+      "loss": 1.2484,
+      "step": 1043
+    },
+    {
+      "epoch": 0.88,
+      "learning_rate": 8.8931023891316e-06,
+      "loss": 1.2569,
+      "step": 1044
+    },
+    {
+      "epoch": 0.88,
+      "learning_rate": 8.89001160579939e-06,
+      "loss": 1.0117,
+      "step": 1045
+    },
+    {
+      "epoch": 0.88,
+      "learning_rate": 8.886917051984077e-06,
+      "loss": 1.0436,
+      "step": 1046
+    },
+    {
+      "epoch": 0.89,
+      "learning_rate": 8.883818730685126e-06,
+      "loss": 1.2294,
+      "step": 1047
+    },
+    {
+      "epoch": 0.89,
+      "learning_rate": 8.880716644905658e-06,
+      "loss": 1.2142,
+      "step": 1048
+    },
+    {
+      "epoch": 0.89,
+      "learning_rate": 8.877610797652441e-06,
+      "loss": 0.9556,
+      "step": 1049
+    },
+    {
+      "epoch": 0.89,
+      "learning_rate": 8.87450119193589e-06,
+      "loss": 1.1449,
+      "step": 1050
+    },
+    {
+      "epoch": 0.89,
+      "learning_rate": 8.871387830770058e-06,
+      "loss": 1.1524,
+      "step": 1051
+    },
+    {
+      "epoch": 0.89,
+      "learning_rate": 8.868270717172645e-06,
+      "loss": 1.0689,
+      "step": 1052
+    },
+    {
+      "epoch": 0.89,
+      "learning_rate": 8.865149854164983e-06,
+      "loss": 0.964,
+      "step": 1053
+    },
+    {
+      "epoch": 0.89,
+      "learning_rate": 8.862025244772043e-06,
+      "loss": 1.3596,
+      "step": 1054
+    },
+    {
+      "epoch": 0.89,
+      "learning_rate": 8.858896892022421e-06,
+      "loss": 0.9729,
+      "step": 1055
+    },
+    {
+      "epoch": 0.89,
+      "learning_rate": 8.855764798948348e-06,
+      "loss": 1.112,
+      "step": 1056
+    },
+    {
+      "epoch": 0.89,
+      "learning_rate": 8.852628968585674e-06,
+      "loss": 1.2246,
+      "step": 1057
+    },
+    {
+      "epoch": 0.89,
+      "learning_rate": 8.849489403973879e-06,
+      "loss": 1.0653,
+      "step": 1058
+    },
+    {
+      "epoch": 0.9,
+      "learning_rate": 8.846346108156053e-06,
+      "loss": 1.1116,
+      "step": 1059
+    },
+    {
+      "epoch": 0.9,
+      "learning_rate": 8.84319908417891e-06,
+      "loss": 1.0794,
+      "step": 1060
+    },
+    {
+      "epoch": 0.9,
+      "learning_rate": 8.840048335092774e-06,
+      "loss": 1.0809,
+      "step": 1061
+    },
+    {
+      "epoch": 0.9,
+      "learning_rate": 8.836893863951584e-06,
+      "loss": 1.0488,
+      "step": 1062
+    },
+    {
+      "epoch": 0.9,
+      "learning_rate": 8.833735673812882e-06,
+      "loss": 1.0564,
+      "step": 1063
+    },
+    {
+      "epoch": 0.9,
+      "learning_rate": 8.830573767737815e-06,
+      "loss": 1.1996,
+      "step": 1064
+    },
+    {
+      "epoch": 0.9,
+      "learning_rate": 8.827408148791134e-06,
+      "loss": 1.0043,
+      "step": 1065
+    },
+    {
+      "epoch": 0.9,
+      "learning_rate": 8.824238820041189e-06,
+      "loss": 1.0391,
+      "step": 1066
+    },
+    {
+      "epoch": 0.9,
+      "learning_rate": 8.821065784559924e-06,
+      "loss": 1.0803,
+      "step": 1067
+    },
+    {
+      "epoch": 0.9,
+      "learning_rate": 8.817889045422876e-06,
+      "loss": 1.182,
+      "step": 1068
+    },
+    {
+      "epoch": 0.9,
+      "learning_rate": 8.814708605709172e-06,
+      "loss": 1.0138,
+      "step": 1069
+    },
+    {
+      "epoch": 0.9,
+      "learning_rate": 8.811524468501527e-06,
+      "loss": 1.1156,
+      "step": 1070
+    },
+    {
+      "epoch": 0.91,
+      "learning_rate": 8.808336636886238e-06,
+      "loss": 1.1226,
+      "step": 1071
+    },
+    {
+      "epoch": 0.91,
+      "learning_rate": 8.805145113953188e-06,
+      "loss": 1.0384,
+      "step": 1072
+    },
+    {
+      "epoch": 0.91,
+      "learning_rate": 8.80194990279583e-06,
+      "loss": 1.309,
+      "step": 1073
+    },
+    {
+      "epoch": 0.91,
+      "learning_rate": 8.798751006511195e-06,
+      "loss": 1.0248,
+      "step": 1074
+    },
+    {
+      "epoch": 0.91,
+      "learning_rate": 8.795548428199889e-06,
+      "loss": 1.2012,
+      "step": 1075
+    },
+    {
+      "epoch": 0.91,
+      "learning_rate": 8.792342170966086e-06,
+      "loss": 0.9943,
+      "step": 1076
+    },
+    {
+      "epoch": 0.91,
+      "learning_rate": 8.78913223791752e-06,
+      "loss": 1.2125,
+      "step": 1077
+    },
+    {
+      "epoch": 0.91,
+      "learning_rate": 8.785918632165497e-06,
+      "loss": 1.1148,
+      "step": 1078
+    },
+    {
+      "epoch": 0.91,
+      "learning_rate": 8.782701356824875e-06,
+      "loss": 1.1898,
+      "step": 1079
+    },
+    {
+      "epoch": 0.91,
+      "learning_rate": 8.779480415014074e-06,
+      "loss": 1.0791,
+      "step": 1080
+    },
+    {
+      "epoch": 0.91,
+      "learning_rate": 8.776255809855065e-06,
+      "loss": 0.9168,
+      "step": 1081
+    },
+    {
+      "epoch": 0.92,
+      "learning_rate": 8.773027544473369e-06,
+      "loss": 1.1382,
+      "step": 1082
+    },
+    {
+      "epoch": 0.92,
+      "learning_rate": 8.769795621998058e-06,
+      "loss": 1.3949,
+      "step": 1083
+    },
+    {
+      "epoch": 0.92,
+      "learning_rate": 8.766560045561749e-06,
+      "loss": 1.0926,
+      "step": 1084
+    },
+    {
+      "epoch": 0.92,
+      "learning_rate": 8.763320818300596e-06,
+      "loss": 1.0953,
+      "step": 1085
+    },
+    {
+      "epoch": 0.92,
+      "learning_rate": 8.760077943354295e-06,
+      "loss": 1.2448,
+      "step": 1086
+    },
+    {
+      "epoch": 0.92,
+      "learning_rate": 8.756831423866077e-06,
+      "loss": 1.1668,
+      "step": 1087
+    },
+    {
+      "epoch": 0.92,
+      "learning_rate": 8.753581262982707e-06,
+      "loss": 0.9305,
+      "step": 1088
+    },
+    {
+      "epoch": 0.92,
+      "learning_rate": 8.750327463854475e-06,
+      "loss": 1.2713,
+      "step": 1089
+    },
+    {
+      "epoch": 0.92,
+      "learning_rate": 8.747070029635204e-06,
+      "loss": 1.0204,
+      "step": 1090
+    },
+    {
+      "epoch": 0.92,
+      "learning_rate": 8.743808963482235e-06,
+      "loss": 1.1212,
+      "step": 1091
+    },
+    {
+      "epoch": 0.92,
+      "learning_rate": 8.740544268556432e-06,
+      "loss": 1.0726,
+      "step": 1092
+    },
+    {
+      "epoch": 0.92,
+      "learning_rate": 8.737275948022175e-06,
+      "loss": 1.1245,
+      "step": 1093
+    },
+    {
+      "epoch": 0.93,
+      "learning_rate": 8.734004005047358e-06,
+      "loss": 1.2051,
+      "step": 1094
+    },
+    {
+      "epoch": 0.93,
+      "learning_rate": 8.73072844280339e-06,
+      "loss": 0.9471,
+      "step": 1095
+    },
+    {
+      "epoch": 0.93,
+      "learning_rate": 8.727449264465182e-06,
+      "loss": 1.1714,
+      "step": 1096
+    },
+    {
+      "epoch": 0.93,
+      "learning_rate": 8.724166473211154e-06,
+      "loss": 0.8268,
+      "step": 1097
+    },
+    {
+      "epoch": 0.93,
+      "learning_rate": 8.720880072223225e-06,
+      "loss": 0.9204,
+      "step": 1098
+    },
+    {
+      "epoch": 0.93,
+      "learning_rate": 8.717590064686819e-06,
+      "loss": 1.1903,
+      "step": 1099
+    },
+    {
+      "epoch": 0.93,
+      "learning_rate": 8.714296453790845e-06,
+      "loss": 1.1292,
+      "step": 1100
+    },
+    {
+      "epoch": 0.93,
+      "learning_rate": 8.710999242727719e-06,
+      "loss": 1.0532,
+      "step": 1101
+    },
+    {
+      "epoch": 0.93,
+      "learning_rate": 8.70769843469333e-06,
+      "loss": 1.1348,
+      "step": 1102
+    },
+    {
+      "epoch": 0.93,
+      "learning_rate": 8.704394032887068e-06,
+      "loss": 1.1642,
+      "step": 1103
+    },
+    {
+      "epoch": 0.93,
+      "learning_rate": 8.7010860405118e-06,
+      "loss": 1.1413,
+      "step": 1104
+    },
+    {
+      "epoch": 0.93,
+      "learning_rate": 8.69777446077387e-06,
+      "loss": 1.0836,
+      "step": 1105
+    },
+    {
+      "epoch": 0.94,
+      "learning_rate": 8.694459296883107e-06,
+      "loss": 1.208,
+      "step": 1106
+    },
+    {
+      "epoch": 0.94,
+      "learning_rate": 8.691140552052807e-06,
+      "loss": 0.9945,
+      "step": 1107
+    },
+    {
+      "epoch": 0.94,
+      "learning_rate": 8.68781822949974e-06,
+      "loss": 1.2102,
+      "step": 1108
+    },
+    {
+      "epoch": 0.94,
+      "learning_rate": 8.684492332444143e-06,
+      "loss": 1.196,
+      "step": 1109
+    },
+    {
+      "epoch": 0.94,
+      "learning_rate": 8.68116286410972e-06,
+      "loss": 0.9378,
+      "step": 1110
+    },
+    {
+      "epoch": 0.94,
+      "learning_rate": 8.677829827723633e-06,
+      "loss": 1.1607,
+      "step": 1111
+    },
+    {
+      "epoch": 0.94,
+      "learning_rate": 8.674493226516506e-06,
+      "loss": 1.2279,
+      "step": 1112
+    },
+    {
+      "epoch": 0.94,
+      "learning_rate": 8.671153063722413e-06,
+      "loss": 1.1437,
+      "step": 1113
+    },
+    {
+      "epoch": 0.94,
+      "learning_rate": 8.667809342578886e-06,
+      "loss": 1.3294,
+      "step": 1114
+    },
+    {
+      "epoch": 0.94,
+      "learning_rate": 8.664462066326902e-06,
+      "loss": 1.0363,
+      "step": 1115
+    },
+    {
+      "epoch": 0.94,
+      "learning_rate": 8.661111238210889e-06,
+      "loss": 1.1603,
+      "step": 1116
+    },
+    {
+      "epoch": 0.94,
+      "learning_rate": 8.65775686147871e-06,
+      "loss": 1.0391,
+      "step": 1117
+    },
+    {
+      "epoch": 0.95,
+      "learning_rate": 8.654398939381672e-06,
+      "loss": 1.0092,
+      "step": 1118
+    },
+    {
+      "epoch": 0.95,
+      "learning_rate": 8.651037475174521e-06,
+      "loss": 1.1966,
+      "step": 1119
+    },
+    {
+      "epoch": 0.95,
+      "learning_rate": 8.647672472115433e-06,
+      "loss": 1.2529,
+      "step": 1120
+    },
+    {
+      "epoch": 0.95,
+      "learning_rate": 8.644303933466011e-06,
+      "loss": 1.0756,
+      "step": 1121
+    },
+    {
+      "epoch": 0.95,
+      "learning_rate": 8.640931862491294e-06,
+      "loss": 1.2182,
+      "step": 1122
+    },
+    {
+      "epoch": 0.95,
+      "learning_rate": 8.637556262459731e-06,
+      "loss": 1.2276,
+      "step": 1123
+    },
+    {
+      "epoch": 0.95,
+      "learning_rate": 8.634177136643208e-06,
+      "loss": 1.0768,
+      "step": 1124
+    },
+    {
+      "epoch": 0.95,
+      "learning_rate": 8.630794488317017e-06,
+      "loss": 1.1027,
+      "step": 1125
+    },
+    {
+      "epoch": 0.95,
+      "learning_rate": 8.627408320759868e-06,
+      "loss": 1.1422,
+      "step": 1126
+    },
+    {
+      "epoch": 0.95,
+      "learning_rate": 8.62401863725388e-06,
+      "loss": 0.8542,
+      "step": 1127
+    },
+    {
+      "epoch": 0.95,
+      "learning_rate": 8.620625441084584e-06,
+      "loss": 1.1349,
+      "step": 1128
+    },
+    {
+      "epoch": 0.95,
+      "learning_rate": 8.617228735540912e-06,
+      "loss": 1.1749,
+      "step": 1129
+    },
+    {
+      "epoch": 0.96,
+      "learning_rate": 8.6138285239152e-06,
+      "loss": 1.1928,
+      "step": 1130
+    },
+    {
+      "epoch": 0.96,
+      "learning_rate": 8.61042480950318e-06,
+      "loss": 1.0909,
+      "step": 1131
+    },
+    {
+      "epoch": 0.96,
+      "learning_rate": 8.607017595603979e-06,
+      "loss": 1.0666,
+      "step": 1132
+    },
+    {
+      "epoch": 0.96,
+      "learning_rate": 8.603606885520122e-06,
+      "loss": 1.0361,
+      "step": 1133
+    },
+    {
+      "epoch": 0.96,
+      "learning_rate": 8.600192682557515e-06,
+      "loss": 1.0654,
+      "step": 1134
+    },
+    {
+      "epoch": 0.96,
+      "learning_rate": 8.596774990025452e-06,
+      "loss": 1.0341,
+      "step": 1135
+    },
+    {
+      "epoch": 0.96,
+      "learning_rate": 8.593353811236611e-06,
+      "loss": 0.9078,
+      "step": 1136
+    },
+    {
+      "epoch": 0.96,
+      "learning_rate": 8.589929149507048e-06,
+      "loss": 1.0794,
+      "step": 1137
+    },
+    {
+      "epoch": 0.96,
+      "learning_rate": 8.586501008156195e-06,
+      "loss": 1.067,
+      "step": 1138
+    },
+    {
+      "epoch": 0.96,
+      "learning_rate": 8.583069390506857e-06,
+      "loss": 0.914,
+      "step": 1139
+    },
+    {
+      "epoch": 0.96,
+      "learning_rate": 8.579634299885208e-06,
+      "loss": 1.0398,
+      "step": 1140
+    },
+    {
+      "epoch": 0.96,
+      "learning_rate": 8.576195739620788e-06,
+      "loss": 1.0567,
+      "step": 1141
+    },
+    {
+      "epoch": 0.97,
+      "learning_rate": 8.572753713046503e-06,
+      "loss": 1.0994,
+      "step": 1142
+    },
+    {
+      "epoch": 0.97,
+      "learning_rate": 8.569308223498613e-06,
+      "loss": 0.9963,
+      "step": 1143
+    },
+    {
+      "epoch": 0.97,
+      "learning_rate": 8.56585927431674e-06,
+      "loss": 1.2382,
+      "step": 1144
+    },
+    {
+      "epoch": 0.97,
+      "learning_rate": 8.562406868843855e-06,
+      "loss": 1.0334,
+      "step": 1145
+    },
+    {
+      "epoch": 0.97,
+      "learning_rate": 8.558951010426284e-06,
+      "loss": 1.0346,
+      "step": 1146
+    },
+    {
+      "epoch": 0.97,
+      "learning_rate": 8.555491702413695e-06,
+      "loss": 1.0372,
+      "step": 1147
+    },
+    {
+      "epoch": 0.97,
+      "learning_rate": 8.552028948159103e-06,
+      "loss": 1.003,
+      "step": 1148
+    },
+    {
+      "epoch": 0.97,
+      "learning_rate": 8.548562751018863e-06,
+      "loss": 1.1725,
+      "step": 1149
+    },
+    {
+      "epoch": 0.97,
+      "learning_rate": 8.545093114352665e-06,
+      "loss": 1.2225,
+      "step": 1150
+    },
+    {
+      "epoch": 0.97,
+      "learning_rate": 8.541620041523533e-06,
+      "loss": 1.0574,
+      "step": 1151
+    },
+    {
+      "epoch": 0.97,
+      "learning_rate": 8.538143535897825e-06,
+      "loss": 1.1119,
+      "step": 1152
+    },
+    {
+      "epoch": 0.98,
+      "learning_rate": 8.534663600845222e-06,
+      "loss": 1.1181,
+      "step": 1153
+    },
+    {
+      "epoch": 0.98,
+      "learning_rate": 8.53118023973873e-06,
+      "loss": 1.1781,
+      "step": 1154
+    },
+    {
+      "epoch": 0.98,
+      "learning_rate": 8.52769345595468e-06,
+      "loss": 1.1527,
+      "step": 1155
+    },
+    {
+      "epoch": 0.98,
+      "learning_rate": 8.524203252872715e-06,
+      "loss": 0.9467,
+      "step": 1156
+    },
+    {
+      "epoch": 0.98,
+      "learning_rate": 8.520709633875796e-06,
+      "loss": 1.248,
+      "step": 1157
+    },
+    {
+      "epoch": 0.98,
+      "learning_rate": 8.51721260235019e-06,
+      "loss": 1.1148,
+      "step": 1158
+    },
+    {
+      "epoch": 0.98,
+      "learning_rate": 8.513712161685482e-06,
+      "loss": 1.0607,
+      "step": 1159
+    },
+    {
+      "epoch": 0.98,
+      "learning_rate": 8.510208315274545e-06,
+      "loss": 1.1193,
+      "step": 1160
+    },
+    {
+      "epoch": 0.98,
+      "learning_rate": 8.506701066513574e-06,
+      "loss": 1.1711,
+      "step": 1161
+    },
+    {
+      "epoch": 0.98,
+      "learning_rate": 8.50319041880204e-06,
+      "loss": 0.9829,
+      "step": 1162
+    },
+    {
+      "epoch": 0.98,
+      "learning_rate": 8.499676375542725e-06,
+      "loss": 1.1226,
+      "step": 1163
+    },
+    {
+      "epoch": 0.98,
+      "learning_rate": 8.496158940141694e-06,
+      "loss": 1.2635,
+      "step": 1164
+    },
+    {
+      "epoch": 0.99,
+      "learning_rate": 8.492638116008302e-06,
+      "loss": 1.0496,
+      "step": 1165
+    },
+    {
+      "epoch": 0.99,
+      "learning_rate": 8.48911390655519e-06,
+      "loss": 0.9152,
+      "step": 1166
+    },
+    {
+      "epoch": 0.99,
+      "learning_rate": 8.485586315198277e-06,
+      "loss": 0.8973,
+      "step": 1167
+    },
+    {
+      "epoch": 0.99,
+      "learning_rate": 8.482055345356764e-06,
+      "loss": 1.2606,
+      "step": 1168
+    },
+    {
+      "epoch": 0.99,
+      "learning_rate": 8.478521000453123e-06,
+      "loss": 1.0187,
+      "step": 1169
+    },
+    {
+      "epoch": 0.99,
+      "learning_rate": 8.4749832839131e-06,
+      "loss": 1.092,
+      "step": 1170
+    },
+    {
+      "epoch": 0.99,
+      "learning_rate": 8.471442199165708e-06,
+      "loss": 1.0946,
+      "step": 1171
+    },
+    {
+      "epoch": 0.99,
+      "learning_rate": 8.467897749643224e-06,
+      "loss": 0.9934,
+      "step": 1172
+    },
+    {
+      "epoch": 0.99,
+      "learning_rate": 8.464349938781186e-06,
+      "loss": 1.0903,
+      "step": 1173
+    },
+    {
+      "epoch": 0.99,
+      "learning_rate": 8.460798770018394e-06,
+      "loss": 1.2764,
+      "step": 1174
+    },
+    {
+      "epoch": 0.99,
+      "learning_rate": 8.457244246796898e-06,
+      "loss": 1.0176,
+      "step": 1175
+    },
+    {
+      "epoch": 0.99,
+      "learning_rate": 8.453686372562004e-06,
+      "loss": 1.142,
+      "step": 1176
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 8.45012515076226e-06,
+      "loss": 1.1121,
+      "step": 1177
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 8.446560584849462e-06,
+      "loss": 1.1098,
+      "step": 1178
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 8.44299267827865e-06,
+      "loss": 1.2095,
+      "step": 1179
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 8.439421434508097e-06,
+      "loss": 1.3557,
+      "step": 1180
+    },
+    {
+      "epoch": 1.0,
+      "eval_loss": 1.0330462455749512,
+      "eval_runtime": 268.4891,
+      "eval_samples_per_second": 3.721,
+      "eval_steps_per_second": 0.466,
+      "step": 1180
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 8.435846856999315e-06,
+      "loss": 1.0335,
+      "step": 1181
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 8.432268949217043e-06,
+      "loss": 1.1255,
+      "step": 1182
+    }
+  ],
+  "logging_steps": 1,
+  "max_steps": 3546,
+  "num_input_tokens_seen": 0,
+  "num_train_epochs": 3,
+  "save_steps": 500,
+  "total_flos": 1.4335795562164716e+18,
+  "trial_name": null,
+  "trial_params": null
+}