diff --git "a/checkpoint-2825/trainer_state.json" "b/checkpoint-2825/trainer_state.json"
new file mode 100644--- /dev/null
+++ "b/checkpoint-2825/trainer_state.json"
@@ -0,0 +1,18097 @@
+{
+  "best_metric": null,
+  "best_model_checkpoint": null,
+  "epoch": 0.9998230401698814,
+  "eval_steps": 20,
+  "global_step": 2825,
+  "is_hyper_param_search": false,
+  "is_local_process_zero": true,
+  "is_world_process_zero": true,
+  "log_history": [
+    {
+      "epoch": 0.0,
+      "learning_rate": 4.000000000000001e-06,
+      "loss": 3.367,
+      "step": 1
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 8.000000000000001e-06,
+      "loss": 5.6519,
+      "step": 2
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 1.2e-05,
+      "loss": 5.6154,
+      "step": 3
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 1.6000000000000003e-05,
+      "loss": 3.6124,
+      "step": 4
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 2e-05,
+      "loss": 3.614,
+      "step": 5
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 2.4e-05,
+      "loss": 2.482,
+      "step": 6
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 2.8000000000000003e-05,
+      "loss": 1.7146,
+      "step": 7
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 3.2000000000000005e-05,
+      "loss": 3.9876,
+      "step": 8
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 3.6e-05,
+      "loss": 2.0762,
+      "step": 9
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 4e-05,
+      "loss": 3.0303,
+      "step": 10
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 4.4000000000000006e-05,
+      "loss": 3.0686,
+      "step": 11
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 4.8e-05,
+      "loss": 1.1891,
+      "step": 12
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 5.2000000000000004e-05,
+      "loss": 2.1397,
+      "step": 13
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 5.6000000000000006e-05,
+      "loss": 1.9174,
+      "step": 14
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 6e-05,
+      "loss": 1.2838,
+      "step": 15
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 6.400000000000001e-05,
+      "loss": 0.9372,
+      "step": 16
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 6.800000000000001e-05,
+      "loss": 0.9274,
+      "step": 17
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 7.2e-05,
+      "loss": 0.862,
+      "step": 18
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 7.6e-05,
+      "loss": 0.8619,
+      "step": 19
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 8e-05,
+      "loss": 0.7438,
+      "step": 20
+    },
+    {
+      "epoch": 0.01,
+      "eval_loss": 0.6827173829078674,
+      "eval_runtime": 52.9454,
+      "eval_samples_per_second": 30.768,
+      "eval_steps_per_second": 30.768,
+      "step": 20
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 8.4e-05,
+      "loss": 0.6896,
+      "step": 21
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 8.800000000000001e-05,
+      "loss": 1.2293,
+      "step": 22
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 9.200000000000001e-05,
+      "loss": 0.7846,
+      "step": 23
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 9.6e-05,
+      "loss": 0.6529,
+      "step": 24
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 0.0001,
+      "loss": 0.7099,
+      "step": 25
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 0.00010400000000000001,
+      "loss": 0.6586,
+      "step": 26
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 0.00010800000000000001,
+      "loss": 0.8019,
+      "step": 27
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 0.00011200000000000001,
+      "loss": 0.8137,
+      "step": 28
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 0.000116,
+      "loss": 0.776,
+      "step": 29
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 0.00012,
+      "loss": 0.617,
+      "step": 30
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 0.000124,
+      "loss": 0.6589,
+      "step": 31
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 0.00012800000000000002,
+      "loss": 0.7027,
+      "step": 32
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 0.000132,
+      "loss": 0.6916,
+      "step": 33
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 0.00013600000000000003,
+      "loss": 0.5547,
+      "step": 34
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 0.00014,
+      "loss": 0.8393,
+      "step": 35
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 0.000144,
+      "loss": 0.6271,
+      "step": 36
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 0.000148,
+      "loss": 0.7236,
+      "step": 37
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 0.000152,
+      "loss": 0.7036,
+      "step": 38
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 0.00015600000000000002,
+      "loss": 0.892,
+      "step": 39
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 0.00016,
+      "loss": 0.8518,
+      "step": 40
+    },
+    {
+      "epoch": 0.01,
+      "eval_loss": 0.6707384586334229,
+      "eval_runtime": 52.9432,
+      "eval_samples_per_second": 30.769,
+      "eval_steps_per_second": 30.769,
+      "step": 40
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 0.000164,
+      "loss": 0.6173,
+      "step": 41
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 0.000168,
+      "loss": 0.8612,
+      "step": 42
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 0.000172,
+      "loss": 0.6524,
+      "step": 43
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 0.00017600000000000002,
+      "loss": 0.5965,
+      "step": 44
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 0.00018,
+      "loss": 0.6584,
+      "step": 45
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 0.00018400000000000003,
+      "loss": 0.9935,
+      "step": 46
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 0.000188,
+      "loss": 0.6044,
+      "step": 47
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 0.000192,
+      "loss": 0.6016,
+      "step": 48
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 0.000196,
+      "loss": 0.7533,
+      "step": 49
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 0.0002,
+      "loss": 0.6704,
+      "step": 50
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 0.00020400000000000003,
+      "loss": 0.6405,
+      "step": 51
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 0.00020800000000000001,
+      "loss": 0.5637,
+      "step": 52
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 0.00021200000000000003,
+      "loss": 0.5849,
+      "step": 53
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 0.00021600000000000002,
+      "loss": 0.5063,
+      "step": 54
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 0.00022000000000000003,
+      "loss": 0.6313,
+      "step": 55
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 0.00022400000000000002,
+      "loss": 0.6947,
+      "step": 56
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 0.00022799999999999999,
+      "loss": 0.8672,
+      "step": 57
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 0.000232,
+      "loss": 0.7494,
+      "step": 58
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 0.000236,
+      "loss": 0.8569,
+      "step": 59
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 0.00024,
+      "loss": 0.6798,
+      "step": 60
+    },
+    {
+      "epoch": 0.02,
+      "eval_loss": 0.65106201171875,
+      "eval_runtime": 52.9482,
+      "eval_samples_per_second": 30.766,
+      "eval_steps_per_second": 30.766,
+      "step": 60
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 0.000244,
+      "loss": 0.6531,
+      "step": 61
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 0.000248,
+      "loss": 0.7695,
+      "step": 62
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 0.000252,
+      "loss": 0.6125,
+      "step": 63
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 0.00025600000000000004,
+      "loss": 0.6594,
+      "step": 64
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 0.00026000000000000003,
+      "loss": 0.4425,
+      "step": 65
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 0.000264,
+      "loss": 0.8456,
+      "step": 66
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 0.000268,
+      "loss": 0.6297,
+      "step": 67
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 0.00027200000000000005,
+      "loss": 0.5551,
+      "step": 68
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 0.000276,
+      "loss": 0.7175,
+      "step": 69
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 0.00028,
+      "loss": 0.565,
+      "step": 70
+    },
+    {
+      "epoch": 0.03,
+      "learning_rate": 0.000284,
+      "loss": 0.6505,
+      "step": 71
+    },
+    {
+      "epoch": 0.03,
+      "learning_rate": 0.000288,
+      "loss": 0.8626,
+      "step": 72
+    },
+    {
+      "epoch": 0.03,
+      "learning_rate": 0.000292,
+      "loss": 0.83,
+      "step": 73
+    },
+    {
+      "epoch": 0.03,
+      "learning_rate": 0.000296,
+      "loss": 0.7591,
+      "step": 74
+    },
+    {
+      "epoch": 0.03,
+      "learning_rate": 0.00030000000000000003,
+      "loss": 0.4987,
+      "step": 75
+    },
+    {
+      "epoch": 0.03,
+      "learning_rate": 0.000304,
+      "loss": 0.9865,
+      "step": 76
+    },
+    {
+      "epoch": 0.03,
+      "learning_rate": 0.000308,
+      "loss": 0.7136,
+      "step": 77
+    },
+    {
+      "epoch": 0.03,
+      "learning_rate": 0.00031200000000000005,
+      "loss": 0.8114,
+      "step": 78
+    },
+    {
+      "epoch": 0.03,
+      "learning_rate": 0.00031600000000000004,
+      "loss": 0.6371,
+      "step": 79
+    },
+    {
+      "epoch": 0.03,
+      "learning_rate": 0.00032,
+      "loss": 0.7868,
+      "step": 80
+    },
+    {
+      "epoch": 0.03,
+      "eval_loss": 0.6231322884559631,
+      "eval_runtime": 52.9431,
+      "eval_samples_per_second": 30.769,
+      "eval_steps_per_second": 30.769,
+      "step": 80
+    },
+    {
+      "epoch": 0.03,
+      "learning_rate": 0.000324,
+      "loss": 0.515,
+      "step": 81
+    },
+    {
+      "epoch": 0.03,
+      "learning_rate": 0.000328,
+      "loss": 0.5867,
+      "step": 82
+    },
+    {
+      "epoch": 0.03,
+      "learning_rate": 0.000332,
+      "loss": 0.6319,
+      "step": 83
+    },
+    {
+      "epoch": 0.03,
+      "learning_rate": 0.000336,
+      "loss": 0.6759,
+      "step": 84
+    },
+    {
+      "epoch": 0.03,
+      "learning_rate": 0.00034,
+      "loss": 0.5591,
+      "step": 85
+    },
+    {
+      "epoch": 0.03,
+      "learning_rate": 0.000344,
+      "loss": 0.6928,
+      "step": 86
+    },
+    {
+      "epoch": 0.03,
+      "learning_rate": 0.000348,
+      "loss": 0.6269,
+      "step": 87
+    },
+    {
+      "epoch": 0.03,
+      "learning_rate": 0.00035200000000000005,
+      "loss": 1.1796,
+      "step": 88
+    },
+    {
+      "epoch": 0.03,
+      "learning_rate": 0.00035600000000000003,
+      "loss": 0.8427,
+      "step": 89
+    },
+    {
+      "epoch": 0.03,
+      "learning_rate": 0.00036,
+      "loss": 0.7287,
+      "step": 90
+    },
+    {
+      "epoch": 0.03,
+      "learning_rate": 0.000364,
+      "loss": 0.6944,
+      "step": 91
+    },
+    {
+      "epoch": 0.03,
+      "learning_rate": 0.00036800000000000005,
+      "loss": 0.8276,
+      "step": 92
+    },
+    {
+      "epoch": 0.03,
+      "learning_rate": 0.00037200000000000004,
+      "loss": 0.7115,
+      "step": 93
+    },
+    {
+      "epoch": 0.03,
+      "learning_rate": 0.000376,
+      "loss": 0.5991,
+      "step": 94
+    },
+    {
+      "epoch": 0.03,
+      "learning_rate": 0.00038,
+      "loss": 0.91,
+      "step": 95
+    },
+    {
+      "epoch": 0.03,
+      "learning_rate": 0.000384,
+      "loss": 0.5116,
+      "step": 96
+    },
+    {
+      "epoch": 0.03,
+      "learning_rate": 0.000388,
+      "loss": 0.5754,
+      "step": 97
+    },
+    {
+      "epoch": 0.03,
+      "learning_rate": 0.000392,
+      "loss": 0.7033,
+      "step": 98
+    },
+    {
+      "epoch": 0.04,
+      "learning_rate": 0.00039600000000000003,
+      "loss": 0.7283,
+      "step": 99
+    },
+    {
+      "epoch": 0.04,
+      "learning_rate": 0.0004,
+      "loss": 0.8218,
+      "step": 100
+    },
+    {
+      "epoch": 0.04,
+      "eval_loss": 0.6231823563575745,
+      "eval_runtime": 52.9454,
+      "eval_samples_per_second": 30.768,
+      "eval_steps_per_second": 30.768,
+      "step": 100
+    },
+    {
+      "epoch": 0.04,
+      "learning_rate": 0.00039999986708723773,
+      "loss": 0.6149,
+      "step": 101
+    },
+    {
+      "epoch": 0.04,
+      "learning_rate": 0.00039999946834912733,
+      "loss": 0.7602,
+      "step": 102
+    },
+    {
+      "epoch": 0.04,
+      "learning_rate": 0.000399998803786199,
+      "loss": 1.2686,
+      "step": 103
+    },
+    {
+      "epoch": 0.04,
+      "learning_rate": 0.00039999787339933585,
+      "loss": 0.8452,
+      "step": 104
+    },
+    {
+      "epoch": 0.04,
+      "learning_rate": 0.0003999966771897746,
+      "loss": 0.5212,
+      "step": 105
+    },
+    {
+      "epoch": 0.04,
+      "learning_rate": 0.00039999521515910515,
+      "loss": 0.9418,
+      "step": 106
+    },
+    {
+      "epoch": 0.04,
+      "learning_rate": 0.00039999348730927064,
+      "loss": 0.846,
+      "step": 107
+    },
+    {
+      "epoch": 0.04,
+      "learning_rate": 0.0003999914936425677,
+      "loss": 1.1,
+      "step": 108
+    },
+    {
+      "epoch": 0.04,
+      "learning_rate": 0.00039998923416164614,
+      "loss": 0.6772,
+      "step": 109
+    },
+    {
+      "epoch": 0.04,
+      "learning_rate": 0.00039998670886950906,
+      "loss": 0.9164,
+      "step": 110
+    },
+    {
+      "epoch": 0.04,
+      "learning_rate": 0.00039998391776951297,
+      "loss": 0.5778,
+      "step": 111
+    },
+    {
+      "epoch": 0.04,
+      "learning_rate": 0.0003999808608653675,
+      "loss": 0.4991,
+      "step": 112
+    },
+    {
+      "epoch": 0.04,
+      "learning_rate": 0.00039997753816113575,
+      "loss": 0.582,
+      "step": 113
+    },
+    {
+      "epoch": 0.04,
+      "learning_rate": 0.00039997394966123396,
+      "loss": 0.5575,
+      "step": 114
+    },
+    {
+      "epoch": 0.04,
+      "learning_rate": 0.00039997009537043175,
+      "loss": 0.5594,
+      "step": 115
+    },
+    {
+      "epoch": 0.04,
+      "learning_rate": 0.00039996597529385193,
+      "loss": 0.5737,
+      "step": 116
+    },
+    {
+      "epoch": 0.04,
+      "learning_rate": 0.0003999615894369706,
+      "loss": 0.5328,
+      "step": 117
+    },
+    {
+      "epoch": 0.04,
+      "learning_rate": 0.00039995693780561716,
+      "loss": 0.897,
+      "step": 118
+    },
+    {
+      "epoch": 0.04,
+      "learning_rate": 0.00039995202040597424,
+      "loss": 0.8001,
+      "step": 119
+    },
+    {
+      "epoch": 0.04,
+      "learning_rate": 0.0003999468372445777,
+      "loss": 0.6862,
+      "step": 120
+    },
+    {
+      "epoch": 0.04,
+      "eval_loss": 0.6324273347854614,
+      "eval_runtime": 52.9433,
+      "eval_samples_per_second": 30.769,
+      "eval_steps_per_second": 30.769,
+      "step": 120
+    },
+    {
+      "epoch": 0.04,
+      "learning_rate": 0.00039994138832831645,
+      "loss": 0.5601,
+      "step": 121
+    },
+    {
+      "epoch": 0.04,
+      "learning_rate": 0.000399935673664433,
+      "loss": 0.7061,
+      "step": 122
+    },
+    {
+      "epoch": 0.04,
+      "learning_rate": 0.00039992969326052287,
+      "loss": 0.6699,
+      "step": 123
+    },
+    {
+      "epoch": 0.04,
+      "learning_rate": 0.0003999234471245346,
+      "loss": 0.9431,
+      "step": 124
+    },
+    {
+      "epoch": 0.04,
+      "learning_rate": 0.00039991693526477025,
+      "loss": 0.6832,
+      "step": 125
+    },
+    {
+      "epoch": 0.04,
+      "learning_rate": 0.00039991015768988487,
+      "loss": 0.8264,
+      "step": 126
+    },
+    {
+      "epoch": 0.04,
+      "learning_rate": 0.00039990311440888667,
+      "loss": 0.5849,
+      "step": 127
+    },
+    {
+      "epoch": 0.05,
+      "learning_rate": 0.0003998958054311372,
+      "loss": 0.8529,
+      "step": 128
+    },
+    {
+      "epoch": 0.05,
+      "learning_rate": 0.0003998882307663509,
+      "loss": 0.7949,
+      "step": 129
+    },
+    {
+      "epoch": 0.05,
+      "learning_rate": 0.0003998803904245955,
+      "loss": 0.6135,
+      "step": 130
+    },
+    {
+      "epoch": 0.05,
+      "learning_rate": 0.00039987228441629185,
+      "loss": 0.6523,
+      "step": 131
+    },
+    {
+      "epoch": 0.05,
+      "learning_rate": 0.00039986391275221385,
+      "loss": 0.9872,
+      "step": 132
+    },
+    {
+      "epoch": 0.05,
+      "learning_rate": 0.00039985527544348856,
+      "loss": 0.7938,
+      "step": 133
+    },
+    {
+      "epoch": 0.05,
+      "learning_rate": 0.0003998463725015959,
+      "loss": 0.7713,
+      "step": 134
+    },
+    {
+      "epoch": 0.05,
+      "learning_rate": 0.0003998372039383692,
+      "loss": 0.7168,
+      "step": 135
+    },
+    {
+      "epoch": 0.05,
+      "learning_rate": 0.0003998277697659946,
+      "loss": 1.2103,
+      "step": 136
+    },
+    {
+      "epoch": 0.05,
+      "learning_rate": 0.0003998180699970112,
+      "loss": 0.7021,
+      "step": 137
+    },
+    {
+      "epoch": 0.05,
+      "learning_rate": 0.0003998081046443114,
+      "loss": 0.6784,
+      "step": 138
+    },
+    {
+      "epoch": 0.05,
+      "learning_rate": 0.0003997978737211403,
+      "loss": 0.8512,
+      "step": 139
+    },
+    {
+      "epoch": 0.05,
+      "learning_rate": 0.00039978737724109624,
+      "loss": 0.4989,
+      "step": 140
+    },
+    {
+      "epoch": 0.05,
+      "eval_loss": 0.6006720662117004,
+      "eval_runtime": 52.9292,
+      "eval_samples_per_second": 30.777,
+      "eval_steps_per_second": 30.777,
+      "step": 140
+    },
+    {
+      "epoch": 0.05,
+      "learning_rate": 0.00039977661521813025,
+      "loss": 0.5216,
+      "step": 141
+    },
+    {
+      "epoch": 0.05,
+      "learning_rate": 0.0003997655876665465,
+      "loss": 0.498,
+      "step": 142
+    },
+    {
+      "epoch": 0.05,
+      "learning_rate": 0.00039975429460100195,
+      "loss": 0.6807,
+      "step": 143
+    },
+    {
+      "epoch": 0.05,
+      "learning_rate": 0.0003997427360365066,
+      "loss": 0.5452,
+      "step": 144
+    },
+    {
+      "epoch": 0.05,
+      "learning_rate": 0.0003997309119884232,
+      "loss": 0.4559,
+      "step": 145
+    },
+    {
+      "epoch": 0.05,
+      "learning_rate": 0.0003997188224724675,
+      "loss": 0.5748,
+      "step": 146
+    },
+    {
+      "epoch": 0.05,
+      "learning_rate": 0.0003997064675047079,
+      "loss": 0.8329,
+      "step": 147
+    },
+    {
+      "epoch": 0.05,
+      "learning_rate": 0.00039969384710156577,
+      "loss": 0.5119,
+      "step": 148
+    },
+    {
+      "epoch": 0.05,
+      "learning_rate": 0.0003996809612798153,
+      "loss": 0.5973,
+      "step": 149
+    },
+    {
+      "epoch": 0.05,
+      "learning_rate": 0.0003996678100565833,
+      "loss": 0.5898,
+      "step": 150
+    },
+    {
+      "epoch": 0.05,
+      "learning_rate": 0.00039965439344934954,
+      "loss": 0.5919,
+      "step": 151
+    },
+    {
+      "epoch": 0.05,
+      "learning_rate": 0.00039964071147594623,
+      "loss": 0.4825,
+      "step": 152
+    },
+    {
+      "epoch": 0.05,
+      "learning_rate": 0.0003996267641545586,
+      "loss": 0.5227,
+      "step": 153
+    },
+    {
+      "epoch": 0.05,
+      "learning_rate": 0.00039961255150372445,
+      "loss": 0.6341,
+      "step": 154
+    },
+    {
+      "epoch": 0.05,
+      "learning_rate": 0.00039959807354233404,
+      "loss": 0.7057,
+      "step": 155
+    },
+    {
+      "epoch": 0.06,
+      "learning_rate": 0.00039958333028963057,
+      "loss": 0.4898,
+      "step": 156
+    },
+    {
+      "epoch": 0.06,
+      "learning_rate": 0.00039956832176520963,
+      "loss": 0.6788,
+      "step": 157
+    },
+    {
+      "epoch": 0.06,
+      "learning_rate": 0.00039955304798901947,
+      "loss": 0.6946,
+      "step": 158
+    },
+    {
+      "epoch": 0.06,
+      "learning_rate": 0.00039953750898136096,
+      "loss": 0.4716,
+      "step": 159
+    },
+    {
+      "epoch": 0.06,
+      "learning_rate": 0.0003995217047628874,
+      "loss": 0.6064,
+      "step": 160
+    },
+    {
+      "epoch": 0.06,
+      "eval_loss": 0.6126547455787659,
+      "eval_runtime": 52.946,
+      "eval_samples_per_second": 30.767,
+      "eval_steps_per_second": 30.767,
+      "step": 160
+    },
+    {
+      "epoch": 0.06,
+      "learning_rate": 0.0003995056353546045,
+      "loss": 0.6949,
+      "step": 161
+    },
+    {
+      "epoch": 0.06,
+      "learning_rate": 0.0003994893007778707,
+      "loss": 0.5829,
+      "step": 162
+    },
+    {
+      "epoch": 0.06,
+      "learning_rate": 0.0003994727010543967,
+      "loss": 0.7388,
+      "step": 163
+    },
+    {
+      "epoch": 0.06,
+      "learning_rate": 0.0003994558362062455,
+      "loss": 0.6947,
+      "step": 164
+    },
+    {
+      "epoch": 0.06,
+      "learning_rate": 0.0003994387062558329,
+      "loss": 0.5429,
+      "step": 165
+    },
+    {
+      "epoch": 0.06,
+      "learning_rate": 0.0003994213112259266,
+      "loss": 0.5463,
+      "step": 166
+    },
+    {
+      "epoch": 0.06,
+      "learning_rate": 0.0003994036511396469,
+      "loss": 0.5821,
+      "step": 167
+    },
+    {
+      "epoch": 0.06,
+      "learning_rate": 0.0003993857260204663,
+      "loss": 0.515,
+      "step": 168
+    },
+    {
+      "epoch": 0.06,
+      "learning_rate": 0.0003993675358922095,
+      "loss": 0.5928,
+      "step": 169
+    },
+    {
+      "epoch": 0.06,
+      "learning_rate": 0.00039934908077905355,
+      "loss": 0.475,
+      "step": 170
+    },
+    {
+      "epoch": 0.06,
+      "learning_rate": 0.0003993303607055277,
+      "loss": 0.5328,
+      "step": 171
+    },
+    {
+      "epoch": 0.06,
+      "learning_rate": 0.00039931137569651325,
+      "loss": 0.6802,
+      "step": 172
+    },
+    {
+      "epoch": 0.06,
+      "learning_rate": 0.00039929212577724375,
+      "loss": 0.5512,
+      "step": 173
+    },
+    {
+      "epoch": 0.06,
+      "learning_rate": 0.0003992726109733047,
+      "loss": 0.4935,
+      "step": 174
+    },
+    {
+      "epoch": 0.06,
+      "learning_rate": 0.00039925283131063383,
+      "loss": 0.5278,
+      "step": 175
+    },
+    {
+      "epoch": 0.06,
+      "learning_rate": 0.0003992327868155209,
+      "loss": 1.117,
+      "step": 176
+    },
+    {
+      "epoch": 0.06,
+      "learning_rate": 0.0003992124775146075,
+      "loss": 0.4618,
+      "step": 177
+    },
+    {
+      "epoch": 0.06,
+      "learning_rate": 0.00039919190343488736,
+      "loss": 0.5333,
+      "step": 178
+    },
+    {
+      "epoch": 0.06,
+      "learning_rate": 0.000399171064603706,
+      "loss": 0.6041,
+      "step": 179
+    },
+    {
+      "epoch": 0.06,
+      "learning_rate": 0.00039914996104876093,
+      "loss": 0.5355,
+      "step": 180
+    },
+    {
+      "epoch": 0.06,
+      "eval_loss": 0.6296773552894592,
+      "eval_runtime": 52.9469,
+      "eval_samples_per_second": 30.767,
+      "eval_steps_per_second": 30.767,
+      "step": 180
+    },
+    {
+      "epoch": 0.06,
+      "learning_rate": 0.0003991285927981014,
+      "loss": 0.7212,
+      "step": 181
+    },
+    {
+      "epoch": 0.06,
+      "learning_rate": 0.00039910695988012865,
+      "loss": 0.4435,
+      "step": 182
+    },
+    {
+      "epoch": 0.06,
+      "learning_rate": 0.0003990850623235955,
+      "loss": 0.693,
+      "step": 183
+    },
+    {
+      "epoch": 0.07,
+      "learning_rate": 0.0003990629001576066,
+      "loss": 0.6267,
+      "step": 184
+    },
+    {
+      "epoch": 0.07,
+      "learning_rate": 0.0003990404734116183,
+      "loss": 0.5952,
+      "step": 185
+    },
+    {
+      "epoch": 0.07,
+      "learning_rate": 0.0003990177821154387,
+      "loss": 0.5953,
+      "step": 186
+    },
+    {
+      "epoch": 0.07,
+      "learning_rate": 0.00039899482629922735,
+      "loss": 0.6198,
+      "step": 187
+    },
+    {
+      "epoch": 0.07,
+      "learning_rate": 0.0003989716059934955,
+      "loss": 0.6072,
+      "step": 188
+    },
+    {
+      "epoch": 0.07,
+      "learning_rate": 0.00039894812122910574,
+      "loss": 0.5169,
+      "step": 189
+    },
+    {
+      "epoch": 0.07,
+      "learning_rate": 0.0003989243720372725,
+      "loss": 0.6952,
+      "step": 190
+    },
+    {
+      "epoch": 0.07,
+      "learning_rate": 0.00039890035844956153,
+      "loss": 0.4809,
+      "step": 191
+    },
+    {
+      "epoch": 0.07,
+      "learning_rate": 0.0003988760804978897,
+      "loss": 0.6635,
+      "step": 192
+    },
+    {
+      "epoch": 0.07,
+      "learning_rate": 0.00039885153821452577,
+      "loss": 0.7432,
+      "step": 193
+    },
+    {
+      "epoch": 0.07,
+      "learning_rate": 0.00039882673163208946,
+      "loss": 0.9253,
+      "step": 194
+    },
+    {
+      "epoch": 0.07,
+      "learning_rate": 0.00039880166078355185,
+      "loss": 0.6323,
+      "step": 195
+    },
+    {
+      "epoch": 0.07,
+      "learning_rate": 0.0003987763257022353,
+      "loss": 0.4035,
+      "step": 196
+    },
+    {
+      "epoch": 0.07,
+      "learning_rate": 0.00039875072642181345,
+      "loss": 0.5926,
+      "step": 197
+    },
+    {
+      "epoch": 0.07,
+      "learning_rate": 0.00039872486297631095,
+      "loss": 0.6379,
+      "step": 198
+    },
+    {
+      "epoch": 0.07,
+      "learning_rate": 0.0003986987354001036,
+      "loss": 0.473,
+      "step": 199
+    },
+    {
+      "epoch": 0.07,
+      "learning_rate": 0.00039867234372791834,
+      "loss": 0.6274,
+      "step": 200
+    },
+    {
+      "epoch": 0.07,
+      "eval_loss": 0.6210886836051941,
+      "eval_runtime": 52.9457,
+      "eval_samples_per_second": 30.767,
+      "eval_steps_per_second": 30.767,
+      "step": 200
+    },
+    {
+      "epoch": 0.07,
+      "learning_rate": 0.000398645687994833,
+      "loss": 0.6188,
+      "step": 201
+    },
+    {
+      "epoch": 0.07,
+      "learning_rate": 0.00039861876823627657,
+      "loss": 0.618,
+      "step": 202
+    },
+    {
+      "epoch": 0.07,
+      "learning_rate": 0.00039859158448802873,
+      "loss": 0.457,
+      "step": 203
+    },
+    {
+      "epoch": 0.07,
+      "learning_rate": 0.00039856413678622017,
+      "loss": 0.5856,
+      "step": 204
+    },
+    {
+      "epoch": 0.07,
+      "learning_rate": 0.00039853642516733245,
+      "loss": 0.4512,
+      "step": 205
+    },
+    {
+      "epoch": 0.07,
+      "learning_rate": 0.0003985084496681977,
+      "loss": 0.5065,
+      "step": 206
+    },
+    {
+      "epoch": 0.07,
+      "learning_rate": 0.0003984802103259991,
+      "loss": 0.4398,
+      "step": 207
+    },
+    {
+      "epoch": 0.07,
+      "learning_rate": 0.0003984517071782703,
+      "loss": 0.5944,
+      "step": 208
+    },
+    {
+      "epoch": 0.07,
+      "learning_rate": 0.0003984229402628956,
+      "loss": 0.5898,
+      "step": 209
+    },
+    {
+      "epoch": 0.07,
+      "learning_rate": 0.0003983939096181098,
+      "loss": 0.8349,
+      "step": 210
+    },
+    {
+      "epoch": 0.07,
+      "learning_rate": 0.00039836461528249857,
+      "loss": 0.652,
+      "step": 211
+    },
+    {
+      "epoch": 0.08,
+      "learning_rate": 0.00039833505729499756,
+      "loss": 0.8563,
+      "step": 212
+    },
+    {
+      "epoch": 0.08,
+      "learning_rate": 0.00039830523569489324,
+      "loss": 0.5888,
+      "step": 213
+    },
+    {
+      "epoch": 0.08,
+      "learning_rate": 0.0003982751505218224,
+      "loss": 0.9032,
+      "step": 214
+    },
+    {
+      "epoch": 0.08,
+      "learning_rate": 0.0003982448018157719,
+      "loss": 1.0715,
+      "step": 215
+    },
+    {
+      "epoch": 0.08,
+      "learning_rate": 0.0003982141896170791,
+      "loss": 0.4641,
+      "step": 216
+    },
+    {
+      "epoch": 0.08,
+      "learning_rate": 0.00039818331396643166,
+      "loss": 0.8314,
+      "step": 217
+    },
+    {
+      "epoch": 0.08,
+      "learning_rate": 0.00039815217490486704,
+      "loss": 0.4499,
+      "step": 218
+    },
+    {
+      "epoch": 0.08,
+      "learning_rate": 0.00039812077247377317,
+      "loss": 0.6566,
+      "step": 219
+    },
+    {
+      "epoch": 0.08,
+      "learning_rate": 0.00039808910671488787,
+      "loss": 0.5512,
+      "step": 220
+    },
+    {
+      "epoch": 0.08,
+      "eval_loss": 0.6289973855018616,
+      "eval_runtime": 52.9443,
+      "eval_samples_per_second": 30.768,
+      "eval_steps_per_second": 30.768,
+      "step": 220
+    },
+    {
+      "epoch": 0.08,
+      "learning_rate": 0.00039805717767029894,
+      "loss": 0.4534,
+      "step": 221
+    },
+    {
+      "epoch": 0.08,
+      "learning_rate": 0.0003980249853824442,
+      "loss": 0.4703,
+      "step": 222
+    },
+    {
+      "epoch": 0.08,
+      "learning_rate": 0.00039799252989411124,
+      "loss": 0.4365,
+      "step": 223
+    },
+    {
+      "epoch": 0.08,
+      "learning_rate": 0.0003979598112484376,
+      "loss": 0.6108,
+      "step": 224
+    },
+    {
+      "epoch": 0.08,
+      "learning_rate": 0.0003979268294889105,
+      "loss": 0.4883,
+      "step": 225
+    },
+    {
+      "epoch": 0.08,
+      "learning_rate": 0.00039789358465936695,
+      "loss": 0.5375,
+      "step": 226
+    },
+    {
+      "epoch": 0.08,
+      "learning_rate": 0.00039786007680399355,
+      "loss": 0.5314,
+      "step": 227
+    },
+    {
+      "epoch": 0.08,
+      "learning_rate": 0.0003978263059673265,
+      "loss": 0.4699,
+      "step": 228
+    },
+    {
+      "epoch": 0.08,
+      "learning_rate": 0.0003977922721942516,
+      "loss": 0.6062,
+      "step": 229
+    },
+    {
+      "epoch": 0.08,
+      "learning_rate": 0.00039775797553000405,
+      "loss": 0.4805,
+      "step": 230
+    },
+    {
+      "epoch": 0.08,
+      "learning_rate": 0.00039772341602016846,
+      "loss": 0.6413,
+      "step": 231
+    },
+    {
+      "epoch": 0.08,
+      "learning_rate": 0.0003976885937106789,
+      "loss": 0.9074,
+      "step": 232
+    },
+    {
+      "epoch": 0.08,
+      "learning_rate": 0.00039765350864781856,
+      "loss": 0.5849,
+      "step": 233
+    },
+    {
+      "epoch": 0.08,
+      "learning_rate": 0.00039761816087822006,
+      "loss": 0.703,
+      "step": 234
+    },
+    {
+      "epoch": 0.08,
+      "learning_rate": 0.0003975825504488651,
+      "loss": 0.4657,
+      "step": 235
+    },
+    {
+      "epoch": 0.08,
+      "learning_rate": 0.0003975466774070844,
+      "loss": 0.6242,
+      "step": 236
+    },
+    {
+      "epoch": 0.08,
+      "learning_rate": 0.0003975105418005579,
+      "loss": 0.5329,
+      "step": 237
+    },
+    {
+      "epoch": 0.08,
+      "learning_rate": 0.0003974741436773144,
+      "loss": 0.5923,
+      "step": 238
+    },
+    {
+      "epoch": 0.08,
+      "learning_rate": 0.0003974374830857317,
+      "loss": 0.7399,
+      "step": 239
+    },
+    {
+      "epoch": 0.08,
+      "learning_rate": 0.0003974005600745362,
+      "loss": 0.727,
+      "step": 240
+    },
+    {
+      "epoch": 0.08,
+      "eval_loss": 0.6027610301971436,
+      "eval_runtime": 52.9795,
+      "eval_samples_per_second": 30.748,
+      "eval_steps_per_second": 30.748,
+      "step": 240
+    },
+    {
+      "epoch": 0.09,
+      "learning_rate": 0.00039736337469280356,
+      "loss": 0.4906,
+      "step": 241
+    },
+    {
+      "epoch": 0.09,
+      "learning_rate": 0.00039732592698995774,
+      "loss": 0.475,
+      "step": 242
+    },
+    {
+      "epoch": 0.09,
+      "learning_rate": 0.0003972882170157716,
+      "loss": 0.3886,
+      "step": 243
+    },
+    {
+      "epoch": 0.09,
+      "learning_rate": 0.00039725024482036643,
+      "loss": 0.6375,
+      "step": 244
+    },
+    {
+      "epoch": 0.09,
+      "learning_rate": 0.0003972120104542122,
+      "loss": 0.5932,
+      "step": 245
+    },
+    {
+      "epoch": 0.09,
+      "learning_rate": 0.0003971735139681272,
+      "loss": 0.7613,
+      "step": 246
+    },
+    {
+      "epoch": 0.09,
+      "learning_rate": 0.0003971347554132782,
+      "loss": 0.6992,
+      "step": 247
+    },
+    {
+      "epoch": 0.09,
+      "learning_rate": 0.00039709573484118023,
+      "loss": 0.5015,
+      "step": 248
+    },
+    {
+      "epoch": 0.09,
+      "learning_rate": 0.0003970564523036967,
+      "loss": 0.4739,
+      "step": 249
+    },
+    {
+      "epoch": 0.09,
+      "learning_rate": 0.000397016907853039,
+      "loss": 0.7145,
+      "step": 250
+    },
+    {
+      "epoch": 0.09,
+      "learning_rate": 0.0003969771015417669,
+      "loss": 0.4026,
+      "step": 251
+    },
+    {
+      "epoch": 0.09,
+      "learning_rate": 0.0003969370334227879,
+      "loss": 0.4941,
+      "step": 252
+    },
+    {
+      "epoch": 0.09,
+      "learning_rate": 0.0003968967035493577,
+      "loss": 0.6773,
+      "step": 253
+    },
+    {
+      "epoch": 0.09,
+      "learning_rate": 0.00039685611197507993,
+      "loss": 0.467,
+      "step": 254
+    },
+    {
+      "epoch": 0.09,
+      "learning_rate": 0.0003968152587539059,
+      "loss": 0.9371,
+      "step": 255
+    },
+    {
+      "epoch": 0.09,
+      "learning_rate": 0.0003967741439401348,
+      "loss": 0.6994,
+      "step": 256
+    },
+    {
+      "epoch": 0.09,
+      "learning_rate": 0.00039673276758841336,
+      "loss": 1.0367,
+      "step": 257
+    },
+    {
+      "epoch": 0.09,
+      "learning_rate": 0.00039669112975373614,
+      "loss": 0.7425,
+      "step": 258
+    },
+    {
+      "epoch": 0.09,
+      "learning_rate": 0.00039664923049144516,
+      "loss": 0.6727,
+      "step": 259
+    },
+    {
+      "epoch": 0.09,
+      "learning_rate": 0.0003966070698572298,
+      "loss": 0.5253,
+      "step": 260
+    },
+    {
+      "epoch": 0.09,
+      "eval_loss": 0.597068727016449,
+      "eval_runtime": 52.9468,
+      "eval_samples_per_second": 30.767,
+      "eval_steps_per_second": 30.767,
+      "step": 260
+    },
+    {
+      "epoch": 0.09,
+      "learning_rate": 0.00039656464790712695,
+      "loss": 0.9258,
+      "step": 261
+    },
+    {
+      "epoch": 0.09,
+      "learning_rate": 0.00039652196469752075,
+      "loss": 0.5787,
+      "step": 262
+    },
+    {
+      "epoch": 0.09,
+      "learning_rate": 0.00039647902028514274,
+      "loss": 0.5374,
+      "step": 263
+    },
+    {
+      "epoch": 0.09,
+      "learning_rate": 0.00039643581472707145,
+      "loss": 0.5907,
+      "step": 264
+    },
+    {
+      "epoch": 0.09,
+      "learning_rate": 0.00039639234808073255,
+      "loss": 0.5771,
+      "step": 265
+    },
+    {
+      "epoch": 0.09,
+      "learning_rate": 0.0003963486204038988,
+      "loss": 0.7109,
+      "step": 266
+    },
+    {
+      "epoch": 0.09,
+      "learning_rate": 0.00039630463175468995,
+      "loss": 0.5905,
+      "step": 267
+    },
+    {
+      "epoch": 0.09,
+      "learning_rate": 0.00039626038219157245,
+      "loss": 0.6573,
+      "step": 268
+    },
+    {
+      "epoch": 0.1,
+      "learning_rate": 0.00039621587177335956,
+      "loss": 0.6822,
+      "step": 269
+    },
+    {
+      "epoch": 0.1,
+      "learning_rate": 0.00039617110055921136,
+      "loss": 1.1458,
+      "step": 270
+    },
+    {
+      "epoch": 0.1,
+      "learning_rate": 0.00039612606860863457,
+      "loss": 0.4737,
+      "step": 271
+    },
+    {
+      "epoch": 0.1,
+      "learning_rate": 0.0003960807759814823,
+      "loss": 0.5358,
+      "step": 272
+    },
+    {
+      "epoch": 0.1,
+      "learning_rate": 0.00039603522273795426,
+      "loss": 0.6107,
+      "step": 273
+    },
+    {
+      "epoch": 0.1,
+      "learning_rate": 0.00039598940893859655,
+      "loss": 0.5075,
+      "step": 274
+    },
+    {
+      "epoch": 0.1,
+      "learning_rate": 0.00039594333464430153,
+      "loss": 0.5263,
+      "step": 275
+    },
+    {
+      "epoch": 0.1,
+      "learning_rate": 0.00039589699991630787,
+      "loss": 0.4556,
+      "step": 276
+    },
+    {
+      "epoch": 0.1,
+      "learning_rate": 0.0003958504048162002,
+      "loss": 0.6868,
+      "step": 277
+    },
+    {
+      "epoch": 0.1,
+      "learning_rate": 0.0003958035494059095,
+      "loss": 0.5381,
+      "step": 278
+    },
+    {
+      "epoch": 0.1,
+      "learning_rate": 0.0003957564337477126,
+      "loss": 0.3986,
+      "step": 279
+    },
+    {
+      "epoch": 0.1,
+      "learning_rate": 0.0003957090579042321,
+      "loss": 0.7679,
+      "step": 280
+    },
+    {
+      "epoch": 0.1,
+      "eval_loss": 0.5907992124557495,
+      "eval_runtime": 52.9445,
+      "eval_samples_per_second": 30.768,
+      "eval_steps_per_second": 30.768,
+      "step": 280
+    },
+    {
+      "epoch": 0.1,
+      "learning_rate": 0.0003956614219384366,
+      "loss": 0.7359,
+      "step": 281
+    },
+    {
+      "epoch": 0.1,
+      "learning_rate": 0.0003956135259136404,
+      "loss": 0.2891,
+      "step": 282
+    },
+    {
+      "epoch": 0.1,
+      "learning_rate": 0.0003955653698935034,
+      "loss": 0.4199,
+      "step": 283
+    },
+    {
+      "epoch": 0.1,
+      "learning_rate": 0.00039551695394203116,
+      "loss": 0.4537,
+      "step": 284
+    },
+    {
+      "epoch": 0.1,
+      "learning_rate": 0.0003954682781235746,
+      "loss": 0.5001,
+      "step": 285
+    },
+    {
+      "epoch": 0.1,
+      "learning_rate": 0.00039541934250283,
+      "loss": 0.5351,
+      "step": 286
+    },
+    {
+      "epoch": 0.1,
+      "learning_rate": 0.0003953701471448392,
+      "loss": 0.4536,
+      "step": 287
+    },
+    {
+      "epoch": 0.1,
+      "learning_rate": 0.00039532069211498913,
+      "loss": 0.5583,
+      "step": 288
+    },
+    {
+      "epoch": 0.1,
+      "learning_rate": 0.0003952709774790117,
+      "loss": 0.5538,
+      "step": 289
+    },
+    {
+      "epoch": 0.1,
+      "learning_rate": 0.00039522100330298415,
+      "loss": 0.4124,
+      "step": 290
+    },
+    {
+      "epoch": 0.1,
+      "learning_rate": 0.00039517076965332835,
+      "loss": 0.5856,
+      "step": 291
+    },
+    {
+      "epoch": 0.1,
+      "learning_rate": 0.00039512027659681137,
+      "loss": 0.6725,
+      "step": 292
+    },
+    {
+      "epoch": 0.1,
+      "learning_rate": 0.0003950695242005449,
+      "loss": 0.7254,
+      "step": 293
+    },
+    {
+      "epoch": 0.1,
+      "learning_rate": 0.00039501851253198535,
+      "loss": 0.6214,
+      "step": 294
+    },
+    {
+      "epoch": 0.1,
+      "learning_rate": 0.0003949672416589338,
+      "loss": 0.4711,
+      "step": 295
+    },
+    {
+      "epoch": 0.1,
+      "learning_rate": 0.0003949157116495356,
+      "loss": 0.549,
+      "step": 296
+    },
+    {
+      "epoch": 0.11,
+      "learning_rate": 0.00039486392257228085,
+      "loss": 0.4409,
+      "step": 297
+    },
+    {
+      "epoch": 0.11,
+      "learning_rate": 0.0003948118744960039,
+      "loss": 0.4832,
+      "step": 298
+    },
+    {
+      "epoch": 0.11,
+      "learning_rate": 0.00039475956748988316,
+      "loss": 0.8131,
+      "step": 299
+    },
+    {
+      "epoch": 0.11,
+      "learning_rate": 0.00039470700162344135,
+      "loss": 0.4804,
+      "step": 300
+    },
+    {
+      "epoch": 0.11,
+      "eval_loss": 0.6154400110244751,
+      "eval_runtime": 52.9456,
+      "eval_samples_per_second": 30.767,
+      "eval_steps_per_second": 30.767,
+      "step": 300
+    },
+    {
+      "epoch": 0.11,
+      "learning_rate": 0.00039465417696654524,
+      "loss": 0.573,
+      "step": 301
+    },
+    {
+      "epoch": 0.11,
+      "learning_rate": 0.0003946010935894055,
+      "loss": 0.4948,
+      "step": 302
+    },
+    {
+      "epoch": 0.11,
+      "learning_rate": 0.00039454775156257685,
+      "loss": 0.4017,
+      "step": 303
+    },
+    {
+      "epoch": 0.11,
+      "learning_rate": 0.00039449415095695745,
+      "loss": 0.5392,
+      "step": 304
+    },
+    {
+      "epoch": 0.11,
+      "learning_rate": 0.0003944402918437894,
+      "loss": 0.3521,
+      "step": 305
+    },
+    {
+      "epoch": 0.11,
+      "learning_rate": 0.00039438617429465845,
+      "loss": 0.5038,
+      "step": 306
+    },
+    {
+      "epoch": 0.11,
+      "learning_rate": 0.0003943317983814936,
+      "loss": 0.4347,
+      "step": 307
+    },
+    {
+      "epoch": 0.11,
+      "learning_rate": 0.00039427716417656753,
+      "loss": 0.5357,
+      "step": 308
+    },
+    {
+      "epoch": 0.11,
+      "learning_rate": 0.0003942222717524959,
+      "loss": 0.4866,
+      "step": 309
+    },
+    {
+      "epoch": 0.11,
+      "learning_rate": 0.0003941671211822378,
+      "loss": 0.39,
+      "step": 310
+    },
+    {
+      "epoch": 0.11,
+      "learning_rate": 0.0003941117125390954,
+      "loss": 0.5636,
+      "step": 311
+    },
+    {
+      "epoch": 0.11,
+      "learning_rate": 0.0003940560458967139,
+      "loss": 0.371,
+      "step": 312
+    },
+    {
+      "epoch": 0.11,
+      "learning_rate": 0.00039400012132908126,
+      "loss": 0.8025,
+      "step": 313
+    },
+    {
+      "epoch": 0.11,
+      "learning_rate": 0.00039394393891052854,
+      "loss": 0.826,
+      "step": 314
+    },
+    {
+      "epoch": 0.11,
+      "learning_rate": 0.00039388749871572913,
+      "loss": 0.5511,
+      "step": 315
+    },
+    {
+      "epoch": 0.11,
+      "learning_rate": 0.0003938308008196994,
+      "loss": 0.9309,
+      "step": 316
+    },
+    {
+      "epoch": 0.11,
+      "learning_rate": 0.00039377384529779805,
+      "loss": 0.4726,
+      "step": 317
+    },
+    {
+      "epoch": 0.11,
+      "learning_rate": 0.0003937166322257262,
+      "loss": 0.9408,
+      "step": 318
+    },
+    {
+      "epoch": 0.11,
+      "learning_rate": 0.00039365916167952744,
+      "loss": 0.617,
+      "step": 319
+    },
+    {
+      "epoch": 0.11,
+      "learning_rate": 0.0003936014337355873,
+      "loss": 0.5801,
+      "step": 320
+    },
+    {
+      "epoch": 0.11,
+      "eval_loss": 0.5967584252357483,
+      "eval_runtime": 52.9412,
+      "eval_samples_per_second": 30.77,
+      "eval_steps_per_second": 30.77,
+      "step": 320
+    },
+    {
+      "epoch": 0.11,
+      "learning_rate": 0.0003935434484706337,
+      "loss": 0.6666,
+      "step": 321
+    },
+    {
+      "epoch": 0.11,
+      "learning_rate": 0.0003934852059617365,
+      "loss": 0.6969,
+      "step": 322
+    },
+    {
+      "epoch": 0.11,
+      "learning_rate": 0.00039342670628630727,
+      "loss": 0.5649,
+      "step": 323
+    },
+    {
+      "epoch": 0.11,
+      "learning_rate": 0.00039336794952209966,
+      "loss": 0.8076,
+      "step": 324
+    },
+    {
+      "epoch": 0.12,
+      "learning_rate": 0.0003933089357472088,
+      "loss": 0.6254,
+      "step": 325
+    },
+    {
+      "epoch": 0.12,
+      "learning_rate": 0.0003932496650400717,
+      "loss": 0.5633,
+      "step": 326
+    },
+    {
+      "epoch": 0.12,
+      "learning_rate": 0.0003931901374794666,
+      "loss": 0.8027,
+      "step": 327
+    },
+    {
+      "epoch": 0.12,
+      "learning_rate": 0.00039313035314451314,
+      "loss": 0.7186,
+      "step": 328
+    },
+    {
+      "epoch": 0.12,
+      "learning_rate": 0.0003930703121146725,
+      "loss": 0.4136,
+      "step": 329
+    },
+    {
+      "epoch": 0.12,
+      "learning_rate": 0.0003930100144697467,
+      "loss": 0.5206,
+      "step": 330
+    },
+    {
+      "epoch": 0.12,
+      "learning_rate": 0.00039294946028987915,
+      "loss": 0.6067,
+      "step": 331
+    },
+    {
+      "epoch": 0.12,
+      "learning_rate": 0.00039288864965555405,
+      "loss": 0.9602,
+      "step": 332
+    },
+    {
+      "epoch": 0.12,
+      "learning_rate": 0.0003928275826475964,
+      "loss": 0.5413,
+      "step": 333
+    },
+    {
+      "epoch": 0.12,
+      "learning_rate": 0.00039276625934717217,
+      "loss": 0.4494,
+      "step": 334
+    },
+    {
+      "epoch": 0.12,
+      "learning_rate": 0.0003927046798357878,
+      "loss": 0.8186,
+      "step": 335
+    },
+    {
+      "epoch": 0.12,
+      "learning_rate": 0.00039264284419529024,
+      "loss": 0.6006,
+      "step": 336
+    },
+    {
+      "epoch": 0.12,
+      "learning_rate": 0.000392580752507867,
+      "loss": 0.7373,
+      "step": 337
+    },
+    {
+      "epoch": 0.12,
+      "learning_rate": 0.000392518404856046,
+      "loss": 0.5813,
+      "step": 338
+    },
+    {
+      "epoch": 0.12,
+      "learning_rate": 0.00039245580132269504,
+      "loss": 0.6654,
+      "step": 339
+    },
+    {
+      "epoch": 0.12,
+      "learning_rate": 0.00039239294199102233,
+      "loss": 0.3603,
+      "step": 340
+    },
+    {
+      "epoch": 0.12,
+      "eval_loss": 0.7126690149307251,
+      "eval_runtime": 52.9485,
+      "eval_samples_per_second": 30.766,
+      "eval_steps_per_second": 30.766,
+      "step": 340
+    },
+    {
+      "epoch": 0.12,
+      "learning_rate": 0.0003923298269445758,
+      "loss": 1.0371,
+      "step": 341
+    },
+    {
+      "epoch": 0.12,
+      "learning_rate": 0.0003922664562672436,
+      "loss": 0.7051,
+      "step": 342
+    },
+    {
+      "epoch": 0.12,
+      "learning_rate": 0.00039220283004325323,
+      "loss": 0.765,
+      "step": 343
+    },
+    {
+      "epoch": 0.12,
+      "learning_rate": 0.00039213894835717224,
+      "loss": 0.5714,
+      "step": 344
+    },
+    {
+      "epoch": 0.12,
+      "learning_rate": 0.00039207481129390745,
+      "loss": 0.8573,
+      "step": 345
+    },
+    {
+      "epoch": 0.12,
+      "learning_rate": 0.0003920104189387052,
+      "loss": 0.6183,
+      "step": 346
+    },
+    {
+      "epoch": 0.12,
+      "learning_rate": 0.0003919457713771512,
+      "loss": 0.5814,
+      "step": 347
+    },
+    {
+      "epoch": 0.12,
+      "learning_rate": 0.0003918808686951703,
+      "loss": 0.5289,
+      "step": 348
+    },
+    {
+      "epoch": 0.12,
+      "learning_rate": 0.0003918157109790264,
+      "loss": 0.5919,
+      "step": 349
+    },
+    {
+      "epoch": 0.12,
+      "learning_rate": 0.0003917502983153225,
+      "loss": 0.6336,
+      "step": 350
+    },
+    {
+      "epoch": 0.12,
+      "learning_rate": 0.00039168463079100027,
+      "loss": 0.5065,
+      "step": 351
+    },
+    {
+      "epoch": 0.12,
+      "learning_rate": 0.0003916187084933403,
+      "loss": 0.7715,
+      "step": 352
+    },
+    {
+      "epoch": 0.12,
+      "learning_rate": 0.00039155253150996173,
+      "loss": 0.6034,
+      "step": 353
+    },
+    {
+      "epoch": 0.13,
+      "learning_rate": 0.0003914860999288222,
+      "loss": 0.5623,
+      "step": 354
+    },
+    {
+      "epoch": 0.13,
+      "learning_rate": 0.0003914194138382178,
+      "loss": 0.6798,
+      "step": 355
+    },
+    {
+      "epoch": 0.13,
+      "learning_rate": 0.0003913524733267828,
+      "loss": 0.5554,
+      "step": 356
+    },
+    {
+      "epoch": 0.13,
+      "learning_rate": 0.00039128527848348975,
+      "loss": 0.4795,
+      "step": 357
+    },
+    {
+      "epoch": 0.13,
+      "learning_rate": 0.0003912178293976491,
+      "loss": 0.5324,
+      "step": 358
+    },
+    {
+      "epoch": 0.13,
+      "learning_rate": 0.0003911501261589094,
+      "loss": 0.5364,
+      "step": 359
+    },
+    {
+      "epoch": 0.13,
+      "learning_rate": 0.00039108216885725674,
+      "loss": 0.5948,
+      "step": 360
+    },
+    {
+      "epoch": 0.13,
+      "eval_loss": 0.5911298990249634,
+      "eval_runtime": 52.944,
+      "eval_samples_per_second": 30.768,
+      "eval_steps_per_second": 30.768,
+      "step": 360
+    },
+    {
+      "epoch": 0.13,
+      "learning_rate": 0.0003910139575830151,
+      "loss": 0.8344,
+      "step": 361
+    },
+    {
+      "epoch": 0.13,
+      "learning_rate": 0.00039094549242684606,
+      "loss": 0.3842,
+      "step": 362
+    },
+    {
+      "epoch": 0.13,
+      "learning_rate": 0.0003908767734797485,
+      "loss": 0.5415,
+      "step": 363
+    },
+    {
+      "epoch": 0.13,
+      "learning_rate": 0.0003908078008330586,
+      "loss": 0.4936,
+      "step": 364
+    },
+    {
+      "epoch": 0.13,
+      "learning_rate": 0.0003907385745784499,
+      "loss": 0.5146,
+      "step": 365
+    },
+    {
+      "epoch": 0.13,
+      "learning_rate": 0.0003906690948079329,
+      "loss": 0.6268,
+      "step": 366
+    },
+    {
+      "epoch": 0.13,
+      "learning_rate": 0.00039059936161385506,
+      "loss": 0.7149,
+      "step": 367
+    },
+    {
+      "epoch": 0.13,
+      "learning_rate": 0.0003905293750889007,
+      "loss": 1.0178,
+      "step": 368
+    },
+    {
+      "epoch": 0.13,
+      "learning_rate": 0.00039045913532609084,
+      "loss": 0.3541,
+      "step": 369
+    },
+    {
+      "epoch": 0.13,
+      "learning_rate": 0.0003903886424187832,
+      "loss": 0.7432,
+      "step": 370
+    },
+    {
+      "epoch": 0.13,
+      "learning_rate": 0.0003903178964606717,
+      "loss": 0.828,
+      "step": 371
+    },
+    {
+      "epoch": 0.13,
+      "learning_rate": 0.0003902468975457868,
+      "loss": 0.5313,
+      "step": 372
+    },
+    {
+      "epoch": 0.13,
+      "learning_rate": 0.0003901756457684951,
+      "loss": 0.4123,
+      "step": 373
+    },
+    {
+      "epoch": 0.13,
+      "learning_rate": 0.0003901041412234993,
+      "loss": 0.5049,
+      "step": 374
+    },
+    {
+      "epoch": 0.13,
+      "learning_rate": 0.00039003238400583815,
+      "loss": 0.3551,
+      "step": 375
+    },
+    {
+      "epoch": 0.13,
+      "learning_rate": 0.0003899603742108861,
+      "loss": 0.691,
+      "step": 376
+    },
+    {
+      "epoch": 0.13,
+      "learning_rate": 0.0003898881119343533,
+      "loss": 0.7923,
+      "step": 377
+    },
+    {
+      "epoch": 0.13,
+      "learning_rate": 0.0003898155972722856,
+      "loss": 0.559,
+      "step": 378
+    },
+    {
+      "epoch": 0.13,
+      "learning_rate": 0.0003897428303210643,
+      "loss": 0.6673,
+      "step": 379
+    },
+    {
+      "epoch": 0.13,
+      "learning_rate": 0.0003896698111774058,
+      "loss": 0.7988,
+      "step": 380
+    },
+    {
+      "epoch": 0.13,
+      "eval_loss": 0.6060499548912048,
+      "eval_runtime": 52.9427,
+      "eval_samples_per_second": 30.769,
+      "eval_steps_per_second": 30.769,
+      "step": 380
+    },
+    {
+      "epoch": 0.13,
+      "learning_rate": 0.000389596539938362,
+      "loss": 0.7621,
+      "step": 381
+    },
+    {
+      "epoch": 0.14,
+      "learning_rate": 0.00038952301670131965,
+      "loss": 0.5031,
+      "step": 382
+    },
+    {
+      "epoch": 0.14,
+      "learning_rate": 0.00038944924156400056,
+      "loss": 0.5851,
+      "step": 383
+    },
+    {
+      "epoch": 0.14,
+      "learning_rate": 0.00038937521462446134,
+      "loss": 0.4939,
+      "step": 384
+    },
+    {
+      "epoch": 0.14,
+      "learning_rate": 0.0003893009359810931,
+      "loss": 0.7234,
+      "step": 385
+    },
+    {
+      "epoch": 0.14,
+      "learning_rate": 0.00038922640573262177,
+      "loss": 0.8176,
+      "step": 386
+    },
+    {
+      "epoch": 0.14,
+      "learning_rate": 0.00038915162397810753,
+      "loss": 0.5511,
+      "step": 387
+    },
+    {
+      "epoch": 0.14,
+      "learning_rate": 0.0003890765908169449,
+      "loss": 0.7284,
+      "step": 388
+    },
+    {
+      "epoch": 0.14,
+      "learning_rate": 0.0003890013063488624,
+      "loss": 0.3144,
+      "step": 389
+    },
+    {
+      "epoch": 0.14,
+      "learning_rate": 0.0003889257706739229,
+      "loss": 1.0848,
+      "step": 390
+    },
+    {
+      "epoch": 0.14,
+      "learning_rate": 0.00038884998389252277,
+      "loss": 1.0601,
+      "step": 391
+    },
+    {
+      "epoch": 0.14,
+      "learning_rate": 0.00038877394610539235,
+      "loss": 0.8447,
+      "step": 392
+    },
+    {
+      "epoch": 0.14,
+      "learning_rate": 0.00038869765741359564,
+      "loss": 0.7155,
+      "step": 393
+    },
+    {
+      "epoch": 0.14,
+      "learning_rate": 0.00038862111791853,
+      "loss": 1.3054,
+      "step": 394
+    },
+    {
+      "epoch": 0.14,
+      "learning_rate": 0.00038854432772192626,
+      "loss": 0.8986,
+      "step": 395
+    },
+    {
+      "epoch": 0.14,
+      "learning_rate": 0.00038846728692584824,
+      "loss": 0.5498,
+      "step": 396
+    },
+    {
+      "epoch": 0.14,
+      "learning_rate": 0.0003883899956326931,
+      "loss": 0.5406,
+      "step": 397
+    },
+    {
+      "epoch": 0.14,
+      "learning_rate": 0.00038831245394519077,
+      "loss": 0.4259,
+      "step": 398
+    },
+    {
+      "epoch": 0.14,
+      "learning_rate": 0.0003882346619664041,
+      "loss": 0.4894,
+      "step": 399
+    },
+    {
+      "epoch": 0.14,
+      "learning_rate": 0.0003881566197997286,
+      "loss": 0.6002,
+      "step": 400
+    },
+    {
+      "epoch": 0.14,
+      "eval_loss": 0.6303256750106812,
+      "eval_runtime": 52.9445,
+      "eval_samples_per_second": 30.768,
+      "eval_steps_per_second": 30.768,
+      "step": 400
+    },
+    {
+      "epoch": 0.14,
+      "learning_rate": 0.0003880783275488922,
+      "loss": 0.5851,
+      "step": 401
+    },
+    {
+      "epoch": 0.14,
+      "learning_rate": 0.00038799978531795523,
+      "loss": 0.5632,
+      "step": 402
+    },
+    {
+      "epoch": 0.14,
+      "learning_rate": 0.00038792099321131044,
+      "loss": 0.4326,
+      "step": 403
+    },
+    {
+      "epoch": 0.14,
+      "learning_rate": 0.00038784195133368266,
+      "loss": 0.5461,
+      "step": 404
+    },
+    {
+      "epoch": 0.14,
+      "learning_rate": 0.00038776265979012845,
+      "loss": 0.6155,
+      "step": 405
+    },
+    {
+      "epoch": 0.14,
+      "learning_rate": 0.0003876831186860366,
+      "loss": 0.4994,
+      "step": 406
+    },
+    {
+      "epoch": 0.14,
+      "learning_rate": 0.0003876033281271271,
+      "loss": 0.4271,
+      "step": 407
+    },
+    {
+      "epoch": 0.14,
+      "learning_rate": 0.00038752328821945207,
+      "loss": 0.5723,
+      "step": 408
+    },
+    {
+      "epoch": 0.14,
+      "learning_rate": 0.0003874429990693947,
+      "loss": 0.513,
+      "step": 409
+    },
+    {
+      "epoch": 0.15,
+      "learning_rate": 0.00038736246078366934,
+      "loss": 0.5687,
+      "step": 410
+    },
+    {
+      "epoch": 0.15,
+      "learning_rate": 0.0003872816734693219,
+      "loss": 0.5654,
+      "step": 411
+    },
+    {
+      "epoch": 0.15,
+      "learning_rate": 0.0003872006372337288,
+      "loss": 0.5768,
+      "step": 412
+    },
+    {
+      "epoch": 0.15,
+      "learning_rate": 0.00038711935218459773,
+      "loss": 0.5915,
+      "step": 413
+    },
+    {
+      "epoch": 0.15,
+      "learning_rate": 0.00038703781842996677,
+      "loss": 0.6538,
+      "step": 414
+    },
+    {
+      "epoch": 0.15,
+      "learning_rate": 0.0003869560360782047,
+      "loss": 0.6132,
+      "step": 415
+    },
+    {
+      "epoch": 0.15,
+      "learning_rate": 0.0003868740052380108,
+      "loss": 0.6519,
+      "step": 416
+    },
+    {
+      "epoch": 0.15,
+      "learning_rate": 0.00038679172601841443,
+      "loss": 0.5302,
+      "step": 417
+    },
+    {
+      "epoch": 0.15,
+      "learning_rate": 0.00038670919852877523,
+      "loss": 0.5439,
+      "step": 418
+    },
+    {
+      "epoch": 0.15,
+      "learning_rate": 0.00038662642287878266,
+      "loss": 0.5885,
+      "step": 419
+    },
+    {
+      "epoch": 0.15,
+      "learning_rate": 0.0003865433991784563,
+      "loss": 0.5522,
+      "step": 420
+    },
+    {
+      "epoch": 0.15,
+      "eval_loss": 0.6123644113540649,
+      "eval_runtime": 52.9389,
+      "eval_samples_per_second": 30.771,
+      "eval_steps_per_second": 30.771,
+      "step": 420
+    },
+    {
+      "epoch": 0.15,
+      "learning_rate": 0.00038646012753814507,
+      "loss": 0.5668,
+      "step": 421
+    },
+    {
+      "epoch": 0.15,
+      "learning_rate": 0.00038637660806852773,
+      "loss": 0.4459,
+      "step": 422
+    },
+    {
+      "epoch": 0.15,
+      "learning_rate": 0.0003862928408806122,
+      "loss": 0.8117,
+      "step": 423
+    },
+    {
+      "epoch": 0.15,
+      "learning_rate": 0.0003862088260857359,
+      "loss": 0.5138,
+      "step": 424
+    },
+    {
+      "epoch": 0.15,
+      "learning_rate": 0.0003861245637955651,
+      "loss": 0.5618,
+      "step": 425
+    },
+    {
+      "epoch": 0.15,
+      "learning_rate": 0.0003860400541220952,
+      "loss": 0.6349,
+      "step": 426
+    },
+    {
+      "epoch": 0.15,
+      "learning_rate": 0.00038595529717765027,
+      "loss": 0.4995,
+      "step": 427
+    },
+    {
+      "epoch": 0.15,
+      "learning_rate": 0.0003858702930748832,
+      "loss": 0.5811,
+      "step": 428
+    },
+    {
+      "epoch": 0.15,
+      "learning_rate": 0.0003857850419267753,
+      "loss": 0.5972,
+      "step": 429
+    },
+    {
+      "epoch": 0.15,
+      "learning_rate": 0.0003856995438466361,
+      "loss": 0.3932,
+      "step": 430
+    },
+    {
+      "epoch": 0.15,
+      "learning_rate": 0.00038561379894810357,
+      "loss": 0.5192,
+      "step": 431
+    },
+    {
+      "epoch": 0.15,
+      "learning_rate": 0.0003855278073451436,
+      "loss": 0.7157,
+      "step": 432
+    },
+    {
+      "epoch": 0.15,
+      "learning_rate": 0.00038544156915204996,
+      "loss": 0.8539,
+      "step": 433
+    },
+    {
+      "epoch": 0.15,
+      "learning_rate": 0.00038535508448344426,
+      "loss": 0.5895,
+      "step": 434
+    },
+    {
+      "epoch": 0.15,
+      "learning_rate": 0.0003852683534542757,
+      "loss": 0.5459,
+      "step": 435
+    },
+    {
+      "epoch": 0.15,
+      "learning_rate": 0.00038518137617982076,
+      "loss": 1.3949,
+      "step": 436
+    },
+    {
+      "epoch": 0.15,
+      "learning_rate": 0.0003850941527756835,
+      "loss": 0.5322,
+      "step": 437
+    },
+    {
+      "epoch": 0.16,
+      "learning_rate": 0.00038500668335779474,
+      "loss": 0.8428,
+      "step": 438
+    },
+    {
+      "epoch": 0.16,
+      "learning_rate": 0.0003849189680424127,
+      "loss": 0.6171,
+      "step": 439
+    },
+    {
+      "epoch": 0.16,
+      "learning_rate": 0.00038483100694612215,
+      "loss": 0.558,
+      "step": 440
+    },
+    {
+      "epoch": 0.16,
+      "eval_loss": 0.6061158180236816,
+      "eval_runtime": 52.9443,
+      "eval_samples_per_second": 30.768,
+      "eval_steps_per_second": 30.768,
+      "step": 440
+    },
+    {
+      "epoch": 0.16,
+      "learning_rate": 0.0003847428001858346,
+      "loss": 0.4824,
+      "step": 441
+    },
+    {
+      "epoch": 0.16,
+      "learning_rate": 0.0003846543478787882,
+      "loss": 1.0439,
+      "step": 442
+    },
+    {
+      "epoch": 0.16,
+      "learning_rate": 0.0003845656501425472,
+      "loss": 0.4834,
+      "step": 443
+    },
+    {
+      "epoch": 0.16,
+      "learning_rate": 0.0003844767070950023,
+      "loss": 0.47,
+      "step": 444
+    },
+    {
+      "epoch": 0.16,
+      "learning_rate": 0.00038438751885437014,
+      "loss": 0.4933,
+      "step": 445
+    },
+    {
+      "epoch": 0.16,
+      "learning_rate": 0.0003842980855391933,
+      "loss": 0.6541,
+      "step": 446
+    },
+    {
+      "epoch": 0.16,
+      "learning_rate": 0.00038420840726833996,
+      "loss": 0.8945,
+      "step": 447
+    },
+    {
+      "epoch": 0.16,
+      "learning_rate": 0.0003841184841610042,
+      "loss": 0.5052,
+      "step": 448
+    },
+    {
+      "epoch": 0.16,
+      "learning_rate": 0.00038402831633670506,
+      "loss": 0.4845,
+      "step": 449
+    },
+    {
+      "epoch": 0.16,
+      "learning_rate": 0.00038393790391528723,
+      "loss": 0.4646,
+      "step": 450
+    },
+    {
+      "epoch": 0.16,
+      "learning_rate": 0.00038384724701692036,
+      "loss": 0.4534,
+      "step": 451
+    },
+    {
+      "epoch": 0.16,
+      "learning_rate": 0.000383756345762099,
+      "loss": 0.6813,
+      "step": 452
+    },
+    {
+      "epoch": 0.16,
+      "learning_rate": 0.0003836652002716425,
+      "loss": 0.4005,
+      "step": 453
+    },
+    {
+      "epoch": 0.16,
+      "learning_rate": 0.0003835738106666949,
+      "loss": 0.371,
+      "step": 454
+    },
+    {
+      "epoch": 0.16,
+      "learning_rate": 0.00038348217706872463,
+      "loss": 0.6874,
+      "step": 455
+    },
+    {
+      "epoch": 0.16,
+      "learning_rate": 0.00038339029959952444,
+      "loss": 0.5947,
+      "step": 456
+    },
+    {
+      "epoch": 0.16,
+      "learning_rate": 0.0003832981783812112,
+      "loss": 0.665,
+      "step": 457
+    },
+    {
+      "epoch": 0.16,
+      "learning_rate": 0.00038320581353622576,
+      "loss": 0.6997,
+      "step": 458
+    },
+    {
+      "epoch": 0.16,
+      "learning_rate": 0.00038311320518733276,
+      "loss": 0.6616,
+      "step": 459
+    },
+    {
+      "epoch": 0.16,
+      "learning_rate": 0.0003830203534576206,
+      "loss": 0.393,
+      "step": 460
+    },
+    {
+      "epoch": 0.16,
+      "eval_loss": 0.6050332188606262,
+      "eval_runtime": 52.9391,
+      "eval_samples_per_second": 30.771,
+      "eval_steps_per_second": 30.771,
+      "step": 460
+    },
+    {
+      "epoch": 0.16,
+      "learning_rate": 0.00038292725847050105,
+      "loss": 0.6396,
+      "step": 461
+    },
+    {
+      "epoch": 0.16,
+      "learning_rate": 0.00038283392034970916,
+      "loss": 0.5498,
+      "step": 462
+    },
+    {
+      "epoch": 0.16,
+      "learning_rate": 0.0003827403392193032,
+      "loss": 0.6023,
+      "step": 463
+    },
+    {
+      "epoch": 0.16,
+      "learning_rate": 0.0003826465152036646,
+      "loss": 0.4682,
+      "step": 464
+    },
+    {
+      "epoch": 0.16,
+      "learning_rate": 0.0003825524484274973,
+      "loss": 0.9681,
+      "step": 465
+    },
+    {
+      "epoch": 0.16,
+      "learning_rate": 0.00038245813901582805,
+      "loss": 0.554,
+      "step": 466
+    },
+    {
+      "epoch": 0.17,
+      "learning_rate": 0.0003823635870940062,
+      "loss": 0.6618,
+      "step": 467
+    },
+    {
+      "epoch": 0.17,
+      "learning_rate": 0.0003822687927877032,
+      "loss": 0.8569,
+      "step": 468
+    },
+    {
+      "epoch": 0.17,
+      "learning_rate": 0.0003821737562229129,
+      "loss": 0.4906,
+      "step": 469
+    },
+    {
+      "epoch": 0.17,
+      "learning_rate": 0.0003820784775259509,
+      "loss": 0.6342,
+      "step": 470
+    },
+    {
+      "epoch": 0.17,
+      "learning_rate": 0.0003819829568234548,
+      "loss": 0.7656,
+      "step": 471
+    },
+    {
+      "epoch": 0.17,
+      "learning_rate": 0.0003818871942423838,
+      "loss": 0.7658,
+      "step": 472
+    },
+    {
+      "epoch": 0.17,
+      "learning_rate": 0.0003817911899100187,
+      "loss": 0.3334,
+      "step": 473
+    },
+    {
+      "epoch": 0.17,
+      "learning_rate": 0.00038169494395396135,
+      "loss": 0.3501,
+      "step": 474
+    },
+    {
+      "epoch": 0.17,
+      "learning_rate": 0.000381598456502135,
+      "loss": 0.6247,
+      "step": 475
+    },
+    {
+      "epoch": 0.17,
+      "learning_rate": 0.00038150172768278373,
+      "loss": 0.7764,
+      "step": 476
+    },
+    {
+      "epoch": 0.17,
+      "learning_rate": 0.00038140475762447255,
+      "loss": 0.8353,
+      "step": 477
+    },
+    {
+      "epoch": 0.17,
+      "learning_rate": 0.0003813075464560869,
+      "loss": 0.5875,
+      "step": 478
+    },
+    {
+      "epoch": 0.17,
+      "learning_rate": 0.00038121009430683306,
+      "loss": 0.6582,
+      "step": 479
+    },
+    {
+      "epoch": 0.17,
+      "learning_rate": 0.00038111240130623723,
+      "loss": 0.739,
+      "step": 480
+    },
+    {
+      "epoch": 0.17,
+      "eval_loss": 0.5977264642715454,
+      "eval_runtime": 52.9458,
+      "eval_samples_per_second": 30.767,
+      "eval_steps_per_second": 30.767,
+      "step": 480
+    },
+    {
+      "epoch": 0.17,
+      "learning_rate": 0.0003810144675841459,
+      "loss": 0.8122,
+      "step": 481
+    },
+    {
+      "epoch": 0.17,
+      "learning_rate": 0.00038091629327072544,
+      "loss": 0.7668,
+      "step": 482
+    },
+    {
+      "epoch": 0.17,
+      "learning_rate": 0.00038081787849646214,
+      "loss": 0.4192,
+      "step": 483
+    },
+    {
+      "epoch": 0.17,
+      "learning_rate": 0.0003807192233921617,
+      "loss": 0.8932,
+      "step": 484
+    },
+    {
+      "epoch": 0.17,
+      "learning_rate": 0.0003806203280889494,
+      "loss": 0.3671,
+      "step": 485
+    },
+    {
+      "epoch": 0.17,
+      "learning_rate": 0.0003805211927182697,
+      "loss": 0.5117,
+      "step": 486
+    },
+    {
+      "epoch": 0.17,
+      "learning_rate": 0.0003804218174118862,
+      "loss": 0.6125,
+      "step": 487
+    },
+    {
+      "epoch": 0.17,
+      "learning_rate": 0.0003803222023018813,
+      "loss": 0.7346,
+      "step": 488
+    },
+    {
+      "epoch": 0.17,
+      "learning_rate": 0.0003802223475206562,
+      "loss": 0.5083,
+      "step": 489
+    },
+    {
+      "epoch": 0.17,
+      "learning_rate": 0.00038012225320093064,
+      "loss": 0.495,
+      "step": 490
+    },
+    {
+      "epoch": 0.17,
+      "learning_rate": 0.00038002191947574286,
+      "loss": 0.6463,
+      "step": 491
+    },
+    {
+      "epoch": 0.17,
+      "learning_rate": 0.00037992134647844904,
+      "loss": 0.5505,
+      "step": 492
+    },
+    {
+      "epoch": 0.17,
+      "learning_rate": 0.00037982053434272367,
+      "loss": 0.3609,
+      "step": 493
+    },
+    {
+      "epoch": 0.17,
+      "learning_rate": 0.00037971948320255874,
+      "loss": 0.4935,
+      "step": 494
+    },
+    {
+      "epoch": 0.18,
+      "learning_rate": 0.0003796181931922643,
+      "loss": 0.498,
+      "step": 495
+    },
+    {
+      "epoch": 0.18,
+      "learning_rate": 0.00037951666444646767,
+      "loss": 0.7353,
+      "step": 496
+    },
+    {
+      "epoch": 0.18,
+      "learning_rate": 0.00037941489710011344,
+      "loss": 0.6197,
+      "step": 497
+    },
+    {
+      "epoch": 0.18,
+      "learning_rate": 0.0003793128912884635,
+      "loss": 0.4291,
+      "step": 498
+    },
+    {
+      "epoch": 0.18,
+      "learning_rate": 0.00037921064714709643,
+      "loss": 0.5644,
+      "step": 499
+    },
+    {
+      "epoch": 0.18,
+      "learning_rate": 0.0003791081648119079,
+      "loss": 0.6462,
+      "step": 500
+    },
+    {
+      "epoch": 0.18,
+      "eval_loss": 0.5904210805892944,
+      "eval_runtime": 52.9386,
+      "eval_samples_per_second": 30.772,
+      "eval_steps_per_second": 30.772,
+      "step": 500
+    },
+    {
+      "epoch": 0.18,
+      "learning_rate": 0.00037900544441910993,
+      "loss": 0.7194,
+      "step": 501
+    },
+    {
+      "epoch": 0.18,
+      "learning_rate": 0.00037890248610523107,
+      "loss": 1.0172,
+      "step": 502
+    },
+    {
+      "epoch": 0.18,
+      "learning_rate": 0.0003787992900071161,
+      "loss": 0.4528,
+      "step": 503
+    },
+    {
+      "epoch": 0.18,
+      "learning_rate": 0.00037869585626192564,
+      "loss": 0.5512,
+      "step": 504
+    },
+    {
+      "epoch": 0.18,
+      "learning_rate": 0.00037859218500713653,
+      "loss": 0.8036,
+      "step": 505
+    },
+    {
+      "epoch": 0.18,
+      "learning_rate": 0.00037848827638054097,
+      "loss": 0.61,
+      "step": 506
+    },
+    {
+      "epoch": 0.18,
+      "learning_rate": 0.0003783841305202469,
+      "loss": 0.3719,
+      "step": 507
+    },
+    {
+      "epoch": 0.18,
+      "learning_rate": 0.00037827974756467734,
+      "loss": 0.5674,
+      "step": 508
+    },
+    {
+      "epoch": 0.18,
+      "learning_rate": 0.00037817512765257066,
+      "loss": 0.7276,
+      "step": 509
+    },
+    {
+      "epoch": 0.18,
+      "learning_rate": 0.00037807027092298,
+      "loss": 0.5813,
+      "step": 510
+    },
+    {
+      "epoch": 0.18,
+      "learning_rate": 0.0003779651775152734,
+      "loss": 0.4805,
+      "step": 511
+    },
+    {
+      "epoch": 0.18,
+      "learning_rate": 0.0003778598475691333,
+      "loss": 0.6919,
+      "step": 512
+    },
+    {
+      "epoch": 0.18,
+      "learning_rate": 0.0003777542812245567,
+      "loss": 0.5747,
+      "step": 513
+    },
+    {
+      "epoch": 0.18,
+      "learning_rate": 0.00037764847862185484,
+      "loss": 0.4123,
+      "step": 514
+    },
+    {
+      "epoch": 0.18,
+      "learning_rate": 0.0003775424399016528,
+      "loss": 0.4342,
+      "step": 515
+    },
+    {
+      "epoch": 0.18,
+      "learning_rate": 0.0003774361652048895,
+      "loss": 0.5426,
+      "step": 516
+    },
+    {
+      "epoch": 0.18,
+      "learning_rate": 0.0003773296546728178,
+      "loss": 0.6733,
+      "step": 517
+    },
+    {
+      "epoch": 0.18,
+      "learning_rate": 0.0003772229084470035,
+      "loss": 0.8745,
+      "step": 518
+    },
+    {
+      "epoch": 0.18,
+      "learning_rate": 0.00037711592666932615,
+      "loss": 0.4167,
+      "step": 519
+    },
+    {
+      "epoch": 0.18,
+      "learning_rate": 0.00037700870948197807,
+      "loss": 0.5305,
+      "step": 520
+    },
+    {
+      "epoch": 0.18,
+      "eval_loss": 0.5852071046829224,
+      "eval_runtime": 52.9368,
+      "eval_samples_per_second": 30.773,
+      "eval_steps_per_second": 30.773,
+      "step": 520
+    },
+    {
+      "epoch": 0.18,
+      "learning_rate": 0.0003769012570274647,
+      "loss": 0.6676,
+      "step": 521
+    },
+    {
+      "epoch": 0.18,
+      "learning_rate": 0.00037679356944860404,
+      "loss": 0.6775,
+      "step": 522
+    },
+    {
+      "epoch": 0.19,
+      "learning_rate": 0.00037668564688852655,
+      "loss": 0.6208,
+      "step": 523
+    },
+    {
+      "epoch": 0.19,
+      "learning_rate": 0.0003765774894906752,
+      "loss": 1.0673,
+      "step": 524
+    },
+    {
+      "epoch": 0.19,
+      "learning_rate": 0.00037646909739880483,
+      "loss": 0.5043,
+      "step": 525
+    },
+    {
+      "epoch": 0.19,
+      "learning_rate": 0.0003763604707569825,
+      "loss": 0.5757,
+      "step": 526
+    },
+    {
+      "epoch": 0.19,
+      "learning_rate": 0.00037625160970958683,
+      "loss": 0.5377,
+      "step": 527
+    },
+    {
+      "epoch": 0.19,
+      "learning_rate": 0.00037614251440130805,
+      "loss": 0.5109,
+      "step": 528
+    },
+    {
+      "epoch": 0.19,
+      "learning_rate": 0.0003760331849771477,
+      "loss": 0.5641,
+      "step": 529
+    },
+    {
+      "epoch": 0.19,
+      "learning_rate": 0.0003759236215824186,
+      "loss": 1.1455,
+      "step": 530
+    },
+    {
+      "epoch": 0.19,
+      "learning_rate": 0.0003758138243627444,
+      "loss": 0.3512,
+      "step": 531
+    },
+    {
+      "epoch": 0.19,
+      "learning_rate": 0.0003757037934640598,
+      "loss": 0.3812,
+      "step": 532
+    },
+    {
+      "epoch": 0.19,
+      "learning_rate": 0.0003755935290326097,
+      "loss": 0.7183,
+      "step": 533
+    },
+    {
+      "epoch": 0.19,
+      "learning_rate": 0.0003754830312149497,
+      "loss": 1.0743,
+      "step": 534
+    },
+    {
+      "epoch": 0.19,
+      "learning_rate": 0.0003753723001579455,
+      "loss": 0.4601,
+      "step": 535
+    },
+    {
+      "epoch": 0.19,
+      "learning_rate": 0.0003752613360087727,
+      "loss": 0.5121,
+      "step": 536
+    },
+    {
+      "epoch": 0.19,
+      "learning_rate": 0.000375150138914917,
+      "loss": 0.6665,
+      "step": 537
+    },
+    {
+      "epoch": 0.19,
+      "learning_rate": 0.0003750387090241735,
+      "loss": 0.523,
+      "step": 538
+    },
+    {
+      "epoch": 0.19,
+      "learning_rate": 0.0003749270464846466,
+      "loss": 0.3325,
+      "step": 539
+    },
+    {
+      "epoch": 0.19,
+      "learning_rate": 0.00037481515144475015,
+      "loss": 0.4749,
+      "step": 540
+    },
+    {
+      "epoch": 0.19,
+      "eval_loss": 0.5974301695823669,
+      "eval_runtime": 52.9459,
+      "eval_samples_per_second": 30.767,
+      "eval_steps_per_second": 30.767,
+      "step": 540
+    },
+    {
+      "epoch": 0.19,
+      "learning_rate": 0.00037470302405320696,
+      "loss": 0.5555,
+      "step": 541
+    },
+    {
+      "epoch": 0.19,
+      "learning_rate": 0.0003745906644590486,
+      "loss": 0.4634,
+      "step": 542
+    },
+    {
+      "epoch": 0.19,
+      "learning_rate": 0.0003744780728116154,
+      "loss": 0.3894,
+      "step": 543
+    },
+    {
+      "epoch": 0.19,
+      "learning_rate": 0.00037436524926055586,
+      "loss": 0.6156,
+      "step": 544
+    },
+    {
+      "epoch": 0.19,
+      "learning_rate": 0.00037425219395582703,
+      "loss": 0.6197,
+      "step": 545
+    },
+    {
+      "epoch": 0.19,
+      "learning_rate": 0.00037413890704769383,
+      "loss": 0.5037,
+      "step": 546
+    },
+    {
+      "epoch": 0.19,
+      "learning_rate": 0.0003740253886867289,
+      "loss": 0.6372,
+      "step": 547
+    },
+    {
+      "epoch": 0.19,
+      "learning_rate": 0.00037391163902381275,
+      "loss": 0.6363,
+      "step": 548
+    },
+    {
+      "epoch": 0.19,
+      "learning_rate": 0.00037379765821013314,
+      "loss": 0.6155,
+      "step": 549
+    },
+    {
+      "epoch": 0.19,
+      "learning_rate": 0.0003736834463971852,
+      "loss": 0.708,
+      "step": 550
+    },
+    {
+      "epoch": 0.2,
+      "learning_rate": 0.00037356900373677085,
+      "loss": 0.6129,
+      "step": 551
+    },
+    {
+      "epoch": 0.2,
+      "learning_rate": 0.00037345433038099915,
+      "loss": 0.7304,
+      "step": 552
+    },
+    {
+      "epoch": 0.2,
+      "learning_rate": 0.0003733394264822855,
+      "loss": 0.5257,
+      "step": 553
+    },
+    {
+      "epoch": 0.2,
+      "learning_rate": 0.00037322429219335194,
+      "loss": 0.5656,
+      "step": 554
+    },
+    {
+      "epoch": 0.2,
+      "learning_rate": 0.0003731089276672266,
+      "loss": 0.6027,
+      "step": 555
+    },
+    {
+      "epoch": 0.2,
+      "learning_rate": 0.00037299333305724363,
+      "loss": 0.4811,
+      "step": 556
+    },
+    {
+      "epoch": 0.2,
+      "learning_rate": 0.00037287750851704306,
+      "loss": 0.6806,
+      "step": 557
+    },
+    {
+      "epoch": 0.2,
+      "learning_rate": 0.00037276145420057043,
+      "loss": 0.7017,
+      "step": 558
+    },
+    {
+      "epoch": 0.2,
+      "learning_rate": 0.00037264517026207685,
+      "loss": 0.5552,
+      "step": 559
+    },
+    {
+      "epoch": 0.2,
+      "learning_rate": 0.0003725286568561184,
+      "loss": 0.832,
+      "step": 560
+    },
+    {
+      "epoch": 0.2,
+      "eval_loss": 0.5918436646461487,
+      "eval_runtime": 52.9388,
+      "eval_samples_per_second": 30.771,
+      "eval_steps_per_second": 30.771,
+      "step": 560
+    },
+    {
+      "epoch": 0.2,
+      "learning_rate": 0.0003724119141375563,
+      "loss": 0.6042,
+      "step": 561
+    },
+    {
+      "epoch": 0.2,
+      "learning_rate": 0.0003722949422615565,
+      "loss": 0.5423,
+      "step": 562
+    },
+    {
+      "epoch": 0.2,
+      "learning_rate": 0.0003721777413835897,
+      "loss": 0.5255,
+      "step": 563
+    },
+    {
+      "epoch": 0.2,
+      "learning_rate": 0.00037206031165943053,
+      "loss": 0.7474,
+      "step": 564
+    },
+    {
+      "epoch": 0.2,
+      "learning_rate": 0.0003719426532451584,
+      "loss": 0.4348,
+      "step": 565
+    },
+    {
+      "epoch": 0.2,
+      "learning_rate": 0.00037182476629715614,
+      "loss": 0.466,
+      "step": 566
+    },
+    {
+      "epoch": 0.2,
+      "learning_rate": 0.00037170665097211066,
+      "loss": 0.5615,
+      "step": 567
+    },
+    {
+      "epoch": 0.2,
+      "learning_rate": 0.00037158830742701224,
+      "loss": 0.4147,
+      "step": 568
+    },
+    {
+      "epoch": 0.2,
+      "learning_rate": 0.0003714697358191546,
+      "loss": 0.5537,
+      "step": 569
+    },
+    {
+      "epoch": 0.2,
+      "learning_rate": 0.0003713509363061345,
+      "loss": 0.3568,
+      "step": 570
+    },
+    {
+      "epoch": 0.2,
+      "learning_rate": 0.00037123190904585175,
+      "loss": 0.4747,
+      "step": 571
+    },
+    {
+      "epoch": 0.2,
+      "learning_rate": 0.00037111265419650854,
+      "loss": 0.503,
+      "step": 572
+    },
+    {
+      "epoch": 0.2,
+      "learning_rate": 0.0003709931719166101,
+      "loss": 0.5135,
+      "step": 573
+    },
+    {
+      "epoch": 0.2,
+      "learning_rate": 0.00037087346236496334,
+      "loss": 0.3508,
+      "step": 574
+    },
+    {
+      "epoch": 0.2,
+      "learning_rate": 0.00037075352570067774,
+      "loss": 0.5525,
+      "step": 575
+    },
+    {
+      "epoch": 0.2,
+      "learning_rate": 0.00037063336208316433,
+      "loss": 0.652,
+      "step": 576
+    },
+    {
+      "epoch": 0.2,
+      "learning_rate": 0.0003705129716721359,
+      "loss": 0.918,
+      "step": 577
+    },
+    {
+      "epoch": 0.2,
+      "learning_rate": 0.00037039235462760667,
+      "loss": 0.4336,
+      "step": 578
+    },
+    {
+      "epoch": 0.2,
+      "learning_rate": 0.0003702715111098922,
+      "loss": 0.6895,
+      "step": 579
+    },
+    {
+      "epoch": 0.21,
+      "learning_rate": 0.0003701504412796087,
+      "loss": 0.7155,
+      "step": 580
+    },
+    {
+      "epoch": 0.21,
+      "eval_loss": 0.5856972336769104,
+      "eval_runtime": 52.9414,
+      "eval_samples_per_second": 30.77,
+      "eval_steps_per_second": 30.77,
+      "step": 580
+    },
+    {
+      "epoch": 0.21,
+      "learning_rate": 0.00037002914529767365,
+      "loss": 0.3763,
+      "step": 581
+    },
+    {
+      "epoch": 0.21,
+      "learning_rate": 0.00036990762332530483,
+      "loss": 0.5343,
+      "step": 582
+    },
+    {
+      "epoch": 0.21,
+      "learning_rate": 0.0003697858755240204,
+      "loss": 0.5286,
+      "step": 583
+    },
+    {
+      "epoch": 0.21,
+      "learning_rate": 0.0003696639020556388,
+      "loss": 0.4859,
+      "step": 584
+    },
+    {
+      "epoch": 0.21,
+      "learning_rate": 0.00036954170308227825,
+      "loss": 0.7289,
+      "step": 585
+    },
+    {
+      "epoch": 0.21,
+      "learning_rate": 0.0003694192787663569,
+      "loss": 0.434,
+      "step": 586
+    },
+    {
+      "epoch": 0.21,
+      "learning_rate": 0.0003692966292705922,
+      "loss": 0.5217,
+      "step": 587
+    },
+    {
+      "epoch": 0.21,
+      "learning_rate": 0.000369173754758001,
+      "loss": 0.4239,
+      "step": 588
+    },
+    {
+      "epoch": 0.21,
+      "learning_rate": 0.0003690506553918991,
+      "loss": 0.417,
+      "step": 589
+    },
+    {
+      "epoch": 0.21,
+      "learning_rate": 0.0003689273313359015,
+      "loss": 0.4632,
+      "step": 590
+    },
+    {
+      "epoch": 0.21,
+      "learning_rate": 0.0003688037827539215,
+      "loss": 0.651,
+      "step": 591
+    },
+    {
+      "epoch": 0.21,
+      "learning_rate": 0.0003686800098101708,
+      "loss": 0.4963,
+      "step": 592
+    },
+    {
+      "epoch": 0.21,
+      "learning_rate": 0.0003685560126691597,
+      "loss": 0.5972,
+      "step": 593
+    },
+    {
+      "epoch": 0.21,
+      "learning_rate": 0.000368431791495696,
+      "loss": 0.4713,
+      "step": 594
+    },
+    {
+      "epoch": 0.21,
+      "learning_rate": 0.00036830734645488557,
+      "loss": 0.7135,
+      "step": 595
+    },
+    {
+      "epoch": 0.21,
+      "learning_rate": 0.0003681826777121318,
+      "loss": 0.424,
+      "step": 596
+    },
+    {
+      "epoch": 0.21,
+      "learning_rate": 0.0003680577854331353,
+      "loss": 0.461,
+      "step": 597
+    },
+    {
+      "epoch": 0.21,
+      "learning_rate": 0.00036793266978389383,
+      "loss": 0.7106,
+      "step": 598
+    },
+    {
+      "epoch": 0.21,
+      "learning_rate": 0.00036780733093070216,
+      "loss": 0.4159,
+      "step": 599
+    },
+    {
+      "epoch": 0.21,
+      "learning_rate": 0.00036768176904015157,
+      "loss": 0.7374,
+      "step": 600
+    },
+    {
+      "epoch": 0.21,
+      "eval_loss": 0.6092067956924438,
+      "eval_runtime": 52.94,
+      "eval_samples_per_second": 30.771,
+      "eval_steps_per_second": 30.771,
+      "step": 600
+    },
+    {
+      "epoch": 0.21,
+      "learning_rate": 0.0003675559842791297,
+      "loss": 0.8042,
+      "step": 601
+    },
+    {
+      "epoch": 0.21,
+      "learning_rate": 0.00036742997681482077,
+      "loss": 0.559,
+      "step": 602
+    },
+    {
+      "epoch": 0.21,
+      "learning_rate": 0.00036730374681470467,
+      "loss": 0.682,
+      "step": 603
+    },
+    {
+      "epoch": 0.21,
+      "learning_rate": 0.0003671772944465572,
+      "loss": 0.5164,
+      "step": 604
+    },
+    {
+      "epoch": 0.21,
+      "learning_rate": 0.0003670506198784497,
+      "loss": 0.9357,
+      "step": 605
+    },
+    {
+      "epoch": 0.21,
+      "learning_rate": 0.00036692372327874875,
+      "loss": 0.4122,
+      "step": 606
+    },
+    {
+      "epoch": 0.21,
+      "learning_rate": 0.00036679660481611626,
+      "loss": 0.4066,
+      "step": 607
+    },
+    {
+      "epoch": 0.22,
+      "learning_rate": 0.0003666692646595088,
+      "loss": 0.4681,
+      "step": 608
+    },
+    {
+      "epoch": 0.22,
+      "learning_rate": 0.0003665417029781778,
+      "loss": 0.671,
+      "step": 609
+    },
+    {
+      "epoch": 0.22,
+      "learning_rate": 0.00036641391994166885,
+      "loss": 0.4709,
+      "step": 610
+    },
+    {
+      "epoch": 0.22,
+      "learning_rate": 0.00036628591571982207,
+      "loss": 0.5316,
+      "step": 611
+    },
+    {
+      "epoch": 0.22,
+      "learning_rate": 0.00036615769048277126,
+      "loss": 0.7164,
+      "step": 612
+    },
+    {
+      "epoch": 0.22,
+      "learning_rate": 0.0003660292444009443,
+      "loss": 0.4452,
+      "step": 613
+    },
+    {
+      "epoch": 0.22,
+      "learning_rate": 0.00036590057764506227,
+      "loss": 0.4127,
+      "step": 614
+    },
+    {
+      "epoch": 0.22,
+      "learning_rate": 0.00036577169038613973,
+      "loss": 0.5771,
+      "step": 615
+    },
+    {
+      "epoch": 0.22,
+      "learning_rate": 0.0003656425827954844,
+      "loss": 0.5129,
+      "step": 616
+    },
+    {
+      "epoch": 0.22,
+      "learning_rate": 0.00036551325504469666,
+      "loss": 0.4295,
+      "step": 617
+    },
+    {
+      "epoch": 0.22,
+      "learning_rate": 0.0003653837073056696,
+      "loss": 0.7901,
+      "step": 618
+    },
+    {
+      "epoch": 0.22,
+      "learning_rate": 0.0003652539397505887,
+      "loss": 0.3488,
+      "step": 619
+    },
+    {
+      "epoch": 0.22,
+      "learning_rate": 0.00036512395255193167,
+      "loss": 1.2165,
+      "step": 620
+    },
+    {
+      "epoch": 0.22,
+      "eval_loss": 0.607138991355896,
+      "eval_runtime": 52.9441,
+      "eval_samples_per_second": 30.768,
+      "eval_steps_per_second": 30.768,
+      "step": 620
+    },
+    {
+      "epoch": 0.22,
+      "learning_rate": 0.0003649937458824679,
+      "loss": 0.4766,
+      "step": 621
+    },
+    {
+      "epoch": 0.22,
+      "learning_rate": 0.00036486331991525894,
+      "loss": 0.7226,
+      "step": 622
+    },
+    {
+      "epoch": 0.22,
+      "learning_rate": 0.0003647326748236574,
+      "loss": 0.6234,
+      "step": 623
+    },
+    {
+      "epoch": 0.22,
+      "learning_rate": 0.0003646018107813072,
+      "loss": 0.5031,
+      "step": 624
+    },
+    {
+      "epoch": 0.22,
+      "learning_rate": 0.0003644707279621435,
+      "loss": 0.3683,
+      "step": 625
+    },
+    {
+      "epoch": 0.22,
+      "learning_rate": 0.0003643394265403921,
+      "loss": 0.6933,
+      "step": 626
+    },
+    {
+      "epoch": 0.22,
+      "learning_rate": 0.00036420790669056916,
+      "loss": 0.4115,
+      "step": 627
+    },
+    {
+      "epoch": 0.22,
+      "learning_rate": 0.0003640761685874815,
+      "loss": 0.506,
+      "step": 628
+    },
+    {
+      "epoch": 0.22,
+      "learning_rate": 0.0003639442124062259,
+      "loss": 0.4891,
+      "step": 629
+    },
+    {
+      "epoch": 0.22,
+      "learning_rate": 0.0003638120383221889,
+      "loss": 0.5222,
+      "step": 630
+    },
+    {
+      "epoch": 0.22,
+      "learning_rate": 0.00036367964651104674,
+      "loss": 0.6308,
+      "step": 631
+    },
+    {
+      "epoch": 0.22,
+      "learning_rate": 0.00036354703714876495,
+      "loss": 0.5243,
+      "step": 632
+    },
+    {
+      "epoch": 0.22,
+      "learning_rate": 0.0003634142104115984,
+      "loss": 0.5689,
+      "step": 633
+    },
+    {
+      "epoch": 0.22,
+      "learning_rate": 0.0003632811664760908,
+      "loss": 0.8009,
+      "step": 634
+    },
+    {
+      "epoch": 0.22,
+      "learning_rate": 0.0003631479055190744,
+      "loss": 0.5033,
+      "step": 635
+    },
+    {
+      "epoch": 0.23,
+      "learning_rate": 0.00036301442771767014,
+      "loss": 0.6997,
+      "step": 636
+    },
+    {
+      "epoch": 0.23,
+      "learning_rate": 0.000362880733249287,
+      "loss": 0.5391,
+      "step": 637
+    },
+    {
+      "epoch": 0.23,
+      "learning_rate": 0.000362746822291622,
+      "loss": 0.598,
+      "step": 638
+    },
+    {
+      "epoch": 0.23,
+      "learning_rate": 0.0003626126950226598,
+      "loss": 0.5856,
+      "step": 639
+    },
+    {
+      "epoch": 0.23,
+      "learning_rate": 0.0003624783516206728,
+      "loss": 0.4901,
+      "step": 640
+    },
+    {
+      "epoch": 0.23,
+      "eval_loss": 0.5908446907997131,
+      "eval_runtime": 52.939,
+      "eval_samples_per_second": 30.771,
+      "eval_steps_per_second": 30.771,
+      "step": 640
+    },
+    {
+      "epoch": 0.23,
+      "learning_rate": 0.0003623437922642205,
+      "loss": 0.5252,
+      "step": 641
+    },
+    {
+      "epoch": 0.23,
+      "learning_rate": 0.00036220901713214943,
+      "loss": 0.6231,
+      "step": 642
+    },
+    {
+      "epoch": 0.23,
+      "learning_rate": 0.0003620740264035929,
+      "loss": 0.6047,
+      "step": 643
+    },
+    {
+      "epoch": 0.23,
+      "learning_rate": 0.0003619388202579709,
+      "loss": 0.9507,
+      "step": 644
+    },
+    {
+      "epoch": 0.23,
+      "learning_rate": 0.0003618033988749895,
+      "loss": 0.4727,
+      "step": 645
+    },
+    {
+      "epoch": 0.23,
+      "learning_rate": 0.00036166776243464117,
+      "loss": 0.4453,
+      "step": 646
+    },
+    {
+      "epoch": 0.23,
+      "learning_rate": 0.00036153191111720404,
+      "loss": 0.439,
+      "step": 647
+    },
+    {
+      "epoch": 0.23,
+      "learning_rate": 0.0003613958451032417,
+      "loss": 0.7302,
+      "step": 648
+    },
+    {
+      "epoch": 0.23,
+      "learning_rate": 0.0003612595645736034,
+      "loss": 0.4572,
+      "step": 649
+    },
+    {
+      "epoch": 0.23,
+      "learning_rate": 0.0003611230697094233,
+      "loss": 0.5579,
+      "step": 650
+    },
+    {
+      "epoch": 0.23,
+      "learning_rate": 0.0003609863606921205,
+      "loss": 0.4515,
+      "step": 651
+    },
+    {
+      "epoch": 0.23,
+      "learning_rate": 0.0003608494377033987,
+      "loss": 0.4541,
+      "step": 652
+    },
+    {
+      "epoch": 0.23,
+      "learning_rate": 0.0003607123009252461,
+      "loss": 0.5537,
+      "step": 653
+    },
+    {
+      "epoch": 0.23,
+      "learning_rate": 0.00036057495053993486,
+      "loss": 0.4392,
+      "step": 654
+    },
+    {
+      "epoch": 0.23,
+      "learning_rate": 0.0003604373867300213,
+      "loss": 0.8037,
+      "step": 655
+    },
+    {
+      "epoch": 0.23,
+      "learning_rate": 0.0003602996096783452,
+      "loss": 0.6216,
+      "step": 656
+    },
+    {
+      "epoch": 0.23,
+      "learning_rate": 0.00036016161956802987,
+      "loss": 0.4515,
+      "step": 657
+    },
+    {
+      "epoch": 0.23,
+      "learning_rate": 0.0003600234165824818,
+      "loss": 0.4909,
+      "step": 658
+    },
+    {
+      "epoch": 0.23,
+      "learning_rate": 0.00035988500090539035,
+      "loss": 0.5205,
+      "step": 659
+    },
+    {
+      "epoch": 0.23,
+      "learning_rate": 0.00035974637272072763,
+      "loss": 0.4585,
+      "step": 660
+    },
+    {
+      "epoch": 0.23,
+      "eval_loss": 0.6041231751441956,
+      "eval_runtime": 52.9418,
+      "eval_samples_per_second": 30.77,
+      "eval_steps_per_second": 30.77,
+      "step": 660
+    },
+    {
+      "epoch": 0.23,
+      "learning_rate": 0.0003596075322127482,
+      "loss": 0.5612,
+      "step": 661
+    },
+    {
+      "epoch": 0.23,
+      "learning_rate": 0.00035946847956598886,
+      "loss": 0.4686,
+      "step": 662
+    },
+    {
+      "epoch": 0.23,
+      "learning_rate": 0.0003593292149652682,
+      "loss": 0.4862,
+      "step": 663
+    },
+    {
+      "epoch": 0.24,
+      "learning_rate": 0.0003591897385956869,
+      "loss": 0.6356,
+      "step": 664
+    },
+    {
+      "epoch": 0.24,
+      "learning_rate": 0.0003590500506426265,
+      "loss": 0.7539,
+      "step": 665
+    },
+    {
+      "epoch": 0.24,
+      "learning_rate": 0.0003589101512917504,
+      "loss": 0.6791,
+      "step": 666
+    },
+    {
+      "epoch": 0.24,
+      "learning_rate": 0.00035877004072900254,
+      "loss": 0.6586,
+      "step": 667
+    },
+    {
+      "epoch": 0.24,
+      "learning_rate": 0.0003586297191406078,
+      "loss": 0.5208,
+      "step": 668
+    },
+    {
+      "epoch": 0.24,
+      "learning_rate": 0.00035848918671307147,
+      "loss": 0.515,
+      "step": 669
+    },
+    {
+      "epoch": 0.24,
+      "learning_rate": 0.00035834844363317905,
+      "loss": 0.4707,
+      "step": 670
+    },
+    {
+      "epoch": 0.24,
+      "learning_rate": 0.0003582074900879962,
+      "loss": 0.6446,
+      "step": 671
+    },
+    {
+      "epoch": 0.24,
+      "learning_rate": 0.0003580663262648679,
+      "loss": 0.4023,
+      "step": 672
+    },
+    {
+      "epoch": 0.24,
+      "learning_rate": 0.0003579249523514192,
+      "loss": 0.5749,
+      "step": 673
+    },
+    {
+      "epoch": 0.24,
+      "learning_rate": 0.0003577833685355539,
+      "loss": 0.5433,
+      "step": 674
+    },
+    {
+      "epoch": 0.24,
+      "learning_rate": 0.00035764157500545483,
+      "loss": 0.6595,
+      "step": 675
+    },
+    {
+      "epoch": 0.24,
+      "learning_rate": 0.00035749957194958394,
+      "loss": 0.4279,
+      "step": 676
+    },
+    {
+      "epoch": 0.24,
+      "learning_rate": 0.00035735735955668134,
+      "loss": 1.1711,
+      "step": 677
+    },
+    {
+      "epoch": 0.24,
+      "learning_rate": 0.00035721493801576533,
+      "loss": 0.7779,
+      "step": 678
+    },
+    {
+      "epoch": 0.24,
+      "learning_rate": 0.0003570723075161325,
+      "loss": 0.6433,
+      "step": 679
+    },
+    {
+      "epoch": 0.24,
+      "learning_rate": 0.0003569294682473569,
+      "loss": 0.6474,
+      "step": 680
+    },
+    {
+      "epoch": 0.24,
+      "eval_loss": 0.5983539819717407,
+      "eval_runtime": 52.9467,
+      "eval_samples_per_second": 30.767,
+      "eval_steps_per_second": 30.767,
+      "step": 680
+    },
+    {
+      "epoch": 0.24,
+      "learning_rate": 0.00035678642039929003,
+      "loss": 0.9273,
+      "step": 681
+    },
+    {
+      "epoch": 0.24,
+      "learning_rate": 0.0003566431641620609,
+      "loss": 0.427,
+      "step": 682
+    },
+    {
+      "epoch": 0.24,
+      "learning_rate": 0.0003564996997260753,
+      "loss": 0.4777,
+      "step": 683
+    },
+    {
+      "epoch": 0.24,
+      "learning_rate": 0.00035635602728201573,
+      "loss": 0.5371,
+      "step": 684
+    },
+    {
+      "epoch": 0.24,
+      "learning_rate": 0.0003562121470208412,
+      "loss": 0.5051,
+      "step": 685
+    },
+    {
+      "epoch": 0.24,
+      "learning_rate": 0.000356068059133787,
+      "loss": 0.5973,
+      "step": 686
+    },
+    {
+      "epoch": 0.24,
+      "learning_rate": 0.00035592376381236434,
+      "loss": 0.3762,
+      "step": 687
+    },
+    {
+      "epoch": 0.24,
+      "learning_rate": 0.00035577926124836,
+      "loss": 0.5945,
+      "step": 688
+    },
+    {
+      "epoch": 0.24,
+      "learning_rate": 0.0003556345516338364,
+      "loss": 0.6031,
+      "step": 689
+    },
+    {
+      "epoch": 0.24,
+      "learning_rate": 0.00035548963516113106,
+      "loss": 0.3624,
+      "step": 690
+    },
+    {
+      "epoch": 0.24,
+      "learning_rate": 0.0003553445120228565,
+      "loss": 0.5733,
+      "step": 691
+    },
+    {
+      "epoch": 0.24,
+      "learning_rate": 0.00035519918241189993,
+      "loss": 0.4916,
+      "step": 692
+    },
+    {
+      "epoch": 0.25,
+      "learning_rate": 0.00035505364652142285,
+      "loss": 0.5303,
+      "step": 693
+    },
+    {
+      "epoch": 0.25,
+      "learning_rate": 0.0003549079045448611,
+      "loss": 0.57,
+      "step": 694
+    },
+    {
+      "epoch": 0.25,
+      "learning_rate": 0.00035476195667592434,
+      "loss": 0.439,
+      "step": 695
+    },
+    {
+      "epoch": 0.25,
+      "learning_rate": 0.0003546158031085959,
+      "loss": 0.4642,
+      "step": 696
+    },
+    {
+      "epoch": 0.25,
+      "learning_rate": 0.0003544694440371326,
+      "loss": 0.8208,
+      "step": 697
+    },
+    {
+      "epoch": 0.25,
+      "learning_rate": 0.00035432287965606426,
+      "loss": 0.4812,
+      "step": 698
+    },
+    {
+      "epoch": 0.25,
+      "learning_rate": 0.0003541761101601936,
+      "loss": 0.4941,
+      "step": 699
+    },
+    {
+      "epoch": 0.25,
+      "learning_rate": 0.0003540291357445961,
+      "loss": 0.4136,
+      "step": 700
+    },
+    {
+      "epoch": 0.25,
+      "eval_loss": 0.5860272645950317,
+      "eval_runtime": 52.9392,
+      "eval_samples_per_second": 30.771,
+      "eval_steps_per_second": 30.771,
+      "step": 700
+    },
+    {
+      "epoch": 0.25,
+      "learning_rate": 0.00035388195660461953,
+      "loss": 0.5575,
+      "step": 701
+    },
+    {
+      "epoch": 0.25,
+      "learning_rate": 0.00035373457293588366,
+      "loss": 0.5227,
+      "step": 702
+    },
+    {
+      "epoch": 0.25,
+      "learning_rate": 0.00035358698493428027,
+      "loss": 0.4733,
+      "step": 703
+    },
+    {
+      "epoch": 0.25,
+      "learning_rate": 0.00035343919279597253,
+      "loss": 0.6915,
+      "step": 704
+    },
+    {
+      "epoch": 0.25,
+      "learning_rate": 0.00035329119671739524,
+      "loss": 0.3746,
+      "step": 705
+    },
+    {
+      "epoch": 0.25,
+      "learning_rate": 0.000353142996895254,
+      "loss": 0.5656,
+      "step": 706
+    },
+    {
+      "epoch": 0.25,
+      "learning_rate": 0.0003529945935265252,
+      "loss": 0.5363,
+      "step": 707
+    },
+    {
+      "epoch": 0.25,
+      "learning_rate": 0.000352845986808456,
+      "loss": 0.5349,
+      "step": 708
+    },
+    {
+      "epoch": 0.25,
+      "learning_rate": 0.0003526971769385635,
+      "loss": 0.494,
+      "step": 709
+    },
+    {
+      "epoch": 0.25,
+      "learning_rate": 0.00035254816411463524,
+      "loss": 0.6291,
+      "step": 710
+    },
+    {
+      "epoch": 0.25,
+      "learning_rate": 0.0003523989485347281,
+      "loss": 0.4023,
+      "step": 711
+    },
+    {
+      "epoch": 0.25,
+      "learning_rate": 0.00035224953039716874,
+      "loss": 1.1261,
+      "step": 712
+    },
+    {
+      "epoch": 0.25,
+      "learning_rate": 0.0003520999099005529,
+      "loss": 0.83,
+      "step": 713
+    },
+    {
+      "epoch": 0.25,
+      "learning_rate": 0.00035195008724374523,
+      "loss": 0.9172,
+      "step": 714
+    },
+    {
+      "epoch": 0.25,
+      "learning_rate": 0.00035180006262587935,
+      "loss": 0.4254,
+      "step": 715
+    },
+    {
+      "epoch": 0.25,
+      "learning_rate": 0.0003516498362463569,
+      "loss": 0.6173,
+      "step": 716
+    },
+    {
+      "epoch": 0.25,
+      "learning_rate": 0.00035149940830484807,
+      "loss": 0.4568,
+      "step": 717
+    },
+    {
+      "epoch": 0.25,
+      "learning_rate": 0.0003513487790012908,
+      "loss": 0.3868,
+      "step": 718
+    },
+    {
+      "epoch": 0.25,
+      "learning_rate": 0.00035119794853589054,
+      "loss": 0.3407,
+      "step": 719
+    },
+    {
+      "epoch": 0.25,
+      "learning_rate": 0.0003510469171091203,
+      "loss": 0.602,
+      "step": 720
+    },
+    {
+      "epoch": 0.25,
+      "eval_loss": 0.6251004934310913,
+      "eval_runtime": 52.9409,
+      "eval_samples_per_second": 30.77,
+      "eval_steps_per_second": 30.77,
+      "step": 720
+    },
+    {
+      "epoch": 0.26,
+      "learning_rate": 0.00035089568492172014,
+      "loss": 0.4283,
+      "step": 721
+    },
+    {
+      "epoch": 0.26,
+      "learning_rate": 0.00035074425217469687,
+      "loss": 0.4738,
+      "step": 722
+    },
+    {
+      "epoch": 0.26,
+      "learning_rate": 0.000350592619069324,
+      "loss": 0.683,
+      "step": 723
+    },
+    {
+      "epoch": 0.26,
+      "learning_rate": 0.0003504407858071413,
+      "loss": 0.5574,
+      "step": 724
+    },
+    {
+      "epoch": 0.26,
+      "learning_rate": 0.0003502887525899544,
+      "loss": 0.4278,
+      "step": 725
+    },
+    {
+      "epoch": 0.26,
+      "learning_rate": 0.000350136519619835,
+      "loss": 0.4711,
+      "step": 726
+    },
+    {
+      "epoch": 0.26,
+      "learning_rate": 0.0003499840870991202,
+      "loss": 0.3959,
+      "step": 727
+    },
+    {
+      "epoch": 0.26,
+      "learning_rate": 0.00034983145523041206,
+      "loss": 0.5402,
+      "step": 728
+    },
+    {
+      "epoch": 0.26,
+      "learning_rate": 0.000349678624216578,
+      "loss": 0.7476,
+      "step": 729
+    },
+    {
+      "epoch": 0.26,
+      "learning_rate": 0.0003495255942607498,
+      "loss": 0.5137,
+      "step": 730
+    },
+    {
+      "epoch": 0.26,
+      "learning_rate": 0.00034937236556632393,
+      "loss": 0.4995,
+      "step": 731
+    },
+    {
+      "epoch": 0.26,
+      "learning_rate": 0.00034921893833696086,
+      "loss": 0.4191,
+      "step": 732
+    },
+    {
+      "epoch": 0.26,
+      "learning_rate": 0.00034906531277658487,
+      "loss": 0.3612,
+      "step": 733
+    },
+    {
+      "epoch": 0.26,
+      "learning_rate": 0.000348911489089384,
+      "loss": 0.8089,
+      "step": 734
+    },
+    {
+      "epoch": 0.26,
+      "learning_rate": 0.0003487574674798095,
+      "loss": 0.5864,
+      "step": 735
+    },
+    {
+      "epoch": 0.26,
+      "learning_rate": 0.00034860324815257586,
+      "loss": 0.4205,
+      "step": 736
+    },
+    {
+      "epoch": 0.26,
+      "learning_rate": 0.0003484488313126602,
+      "loss": 0.5107,
+      "step": 737
+    },
+    {
+      "epoch": 0.26,
+      "learning_rate": 0.00034829421716530215,
+      "loss": 0.454,
+      "step": 738
+    },
+    {
+      "epoch": 0.26,
+      "learning_rate": 0.00034813940591600374,
+      "loss": 0.6111,
+      "step": 739
+    },
+    {
+      "epoch": 0.26,
+      "learning_rate": 0.0003479843977705287,
+      "loss": 0.405,
+      "step": 740
+    },
+    {
+      "epoch": 0.26,
+      "eval_loss": 0.6182191371917725,
+      "eval_runtime": 52.9436,
+      "eval_samples_per_second": 30.769,
+      "eval_steps_per_second": 30.769,
+      "step": 740
+    },
+    {
+      "epoch": 0.26,
+      "learning_rate": 0.0003478291929349029,
+      "loss": 0.8168,
+      "step": 741
+    },
+    {
+      "epoch": 0.26,
+      "learning_rate": 0.0003476737916154132,
+      "loss": 0.7726,
+      "step": 742
+    },
+    {
+      "epoch": 0.26,
+      "learning_rate": 0.0003475181940186078,
+      "loss": 0.7087,
+      "step": 743
+    },
+    {
+      "epoch": 0.26,
+      "learning_rate": 0.00034736240035129583,
+      "loss": 0.4038,
+      "step": 744
+    },
+    {
+      "epoch": 0.26,
+      "learning_rate": 0.0003472064108205468,
+      "loss": 0.4883,
+      "step": 745
+    },
+    {
+      "epoch": 0.26,
+      "learning_rate": 0.00034705022563369094,
+      "loss": 0.3914,
+      "step": 746
+    },
+    {
+      "epoch": 0.26,
+      "learning_rate": 0.00034689384499831814,
+      "loss": 0.4279,
+      "step": 747
+    },
+    {
+      "epoch": 0.26,
+      "learning_rate": 0.00034673726912227824,
+      "loss": 1.0718,
+      "step": 748
+    },
+    {
+      "epoch": 0.27,
+      "learning_rate": 0.00034658049821368063,
+      "loss": 0.6347,
+      "step": 749
+    },
+    {
+      "epoch": 0.27,
+      "learning_rate": 0.00034642353248089374,
+      "loss": 0.4742,
+      "step": 750
+    },
+    {
+      "epoch": 0.27,
+      "learning_rate": 0.0003462663721325452,
+      "loss": 0.2997,
+      "step": 751
+    },
+    {
+      "epoch": 0.27,
+      "learning_rate": 0.000346109017377521,
+      "loss": 0.4605,
+      "step": 752
+    },
+    {
+      "epoch": 0.27,
+      "learning_rate": 0.00034595146842496586,
+      "loss": 0.4842,
+      "step": 753
+    },
+    {
+      "epoch": 0.27,
+      "learning_rate": 0.00034579372548428235,
+      "loss": 0.4732,
+      "step": 754
+    },
+    {
+      "epoch": 0.27,
+      "learning_rate": 0.00034563578876513096,
+      "loss": 0.6647,
+      "step": 755
+    },
+    {
+      "epoch": 0.27,
+      "learning_rate": 0.0003454776584774298,
+      "loss": 0.621,
+      "step": 756
+    },
+    {
+      "epoch": 0.27,
+      "learning_rate": 0.0003453193348313542,
+      "loss": 0.3721,
+      "step": 757
+    },
+    {
+      "epoch": 0.27,
+      "learning_rate": 0.00034516081803733645,
+      "loss": 0.694,
+      "step": 758
+    },
+    {
+      "epoch": 0.27,
+      "learning_rate": 0.00034500210830606563,
+      "loss": 0.5436,
+      "step": 759
+    },
+    {
+      "epoch": 0.27,
+      "learning_rate": 0.00034484320584848724,
+      "loss": 0.5059,
+      "step": 760
+    },
+    {
+      "epoch": 0.27,
+      "eval_loss": 0.5893740057945251,
+      "eval_runtime": 52.9418,
+      "eval_samples_per_second": 30.77,
+      "eval_steps_per_second": 30.77,
+      "step": 760
+    },
+    {
+      "epoch": 0.27,
+      "learning_rate": 0.00034468411087580284,
+      "loss": 0.743,
+      "step": 761
+    },
+    {
+      "epoch": 0.27,
+      "learning_rate": 0.00034452482359947006,
+      "loss": 0.499,
+      "step": 762
+    },
+    {
+      "epoch": 0.27,
+      "learning_rate": 0.00034436534423120204,
+      "loss": 0.4899,
+      "step": 763
+    },
+    {
+      "epoch": 0.27,
+      "learning_rate": 0.0003442056729829671,
+      "loss": 0.4482,
+      "step": 764
+    },
+    {
+      "epoch": 0.27,
+      "learning_rate": 0.0003440458100669887,
+      "loss": 0.6109,
+      "step": 765
+    },
+    {
+      "epoch": 0.27,
+      "learning_rate": 0.00034388575569574523,
+      "loss": 0.713,
+      "step": 766
+    },
+    {
+      "epoch": 0.27,
+      "learning_rate": 0.0003437255100819692,
+      "loss": 0.6203,
+      "step": 767
+    },
+    {
+      "epoch": 0.27,
+      "learning_rate": 0.0003435650734386475,
+      "loss": 0.5148,
+      "step": 768
+    },
+    {
+      "epoch": 0.27,
+      "learning_rate": 0.0003434044459790211,
+      "loss": 0.7719,
+      "step": 769
+    },
+    {
+      "epoch": 0.27,
+      "learning_rate": 0.0003432436279165841,
+      "loss": 0.6339,
+      "step": 770
+    },
+    {
+      "epoch": 0.27,
+      "learning_rate": 0.00034308261946508435,
+      "loss": 0.8842,
+      "step": 771
+    },
+    {
+      "epoch": 0.27,
+      "learning_rate": 0.00034292142083852277,
+      "loss": 0.5877,
+      "step": 772
+    },
+    {
+      "epoch": 0.27,
+      "learning_rate": 0.0003427600322511527,
+      "loss": 0.7573,
+      "step": 773
+    },
+    {
+      "epoch": 0.27,
+      "learning_rate": 0.0003425984539174803,
+      "loss": 0.3674,
+      "step": 774
+    },
+    {
+      "epoch": 0.27,
+      "learning_rate": 0.0003424366860522637,
+      "loss": 0.6808,
+      "step": 775
+    },
+    {
+      "epoch": 0.27,
+      "learning_rate": 0.0003422747288705132,
+      "loss": 0.4831,
+      "step": 776
+    },
+    {
+      "epoch": 0.27,
+      "learning_rate": 0.0003421125825874904,
+      "loss": 0.4888,
+      "step": 777
+    },
+    {
+      "epoch": 0.28,
+      "learning_rate": 0.00034195024741870845,
+      "loss": 0.3946,
+      "step": 778
+    },
+    {
+      "epoch": 0.28,
+      "learning_rate": 0.0003417877235799315,
+      "loss": 0.6767,
+      "step": 779
+    },
+    {
+      "epoch": 0.28,
+      "learning_rate": 0.00034162501128717456,
+      "loss": 0.4249,
+      "step": 780
+    },
+    {
+      "epoch": 0.28,
+      "eval_loss": 0.579978883266449,
+      "eval_runtime": 52.9796,
+      "eval_samples_per_second": 30.748,
+      "eval_steps_per_second": 30.748,
+      "step": 780
+    },
+    {
+      "epoch": 0.28,
+      "learning_rate": 0.0003414621107567029,
+      "loss": 0.4724,
+      "step": 781
+    },
+    {
+      "epoch": 0.28,
+      "learning_rate": 0.0003412990222050322,
+      "loss": 0.5866,
+      "step": 782
+    },
+    {
+      "epoch": 0.28,
+      "learning_rate": 0.0003411357458489279,
+      "loss": 0.5316,
+      "step": 783
+    },
+    {
+      "epoch": 0.28,
+      "learning_rate": 0.00034097228190540526,
+      "loss": 0.7415,
+      "step": 784
+    },
+    {
+      "epoch": 0.28,
+      "learning_rate": 0.0003408086305917285,
+      "loss": 0.8415,
+      "step": 785
+    },
+    {
+      "epoch": 0.28,
+      "learning_rate": 0.00034064479212541135,
+      "loss": 0.5674,
+      "step": 786
+    },
+    {
+      "epoch": 0.28,
+      "learning_rate": 0.00034048076672421585,
+      "loss": 0.4012,
+      "step": 787
+    },
+    {
+      "epoch": 0.28,
+      "learning_rate": 0.00034031655460615274,
+      "loss": 0.4711,
+      "step": 788
+    },
+    {
+      "epoch": 0.28,
+      "learning_rate": 0.00034015215598948086,
+      "loss": 0.7712,
+      "step": 789
+    },
+    {
+      "epoch": 0.28,
+      "learning_rate": 0.000339987571092707,
+      "loss": 0.4876,
+      "step": 790
+    },
+    {
+      "epoch": 0.28,
+      "learning_rate": 0.00033982280013458553,
+      "loss": 0.4908,
+      "step": 791
+    },
+    {
+      "epoch": 0.28,
+      "learning_rate": 0.00033965784333411797,
+      "loss": 0.689,
+      "step": 792
+    },
+    {
+      "epoch": 0.28,
+      "learning_rate": 0.000339492700910553,
+      "loss": 1.2844,
+      "step": 793
+    },
+    {
+      "epoch": 0.28,
+      "learning_rate": 0.00033932737308338607,
+      "loss": 0.4104,
+      "step": 794
+    },
+    {
+      "epoch": 0.28,
+      "learning_rate": 0.00033916186007235886,
+      "loss": 0.5245,
+      "step": 795
+    },
+    {
+      "epoch": 0.28,
+      "learning_rate": 0.0003389961620974593,
+      "loss": 0.3759,
+      "step": 796
+    },
+    {
+      "epoch": 0.28,
+      "learning_rate": 0.00033883027937892124,
+      "loss": 0.3775,
+      "step": 797
+    },
+    {
+      "epoch": 0.28,
+      "learning_rate": 0.00033866421213722377,
+      "loss": 0.5029,
+      "step": 798
+    },
+    {
+      "epoch": 0.28,
+      "learning_rate": 0.0003384979605930917,
+      "loss": 0.5544,
+      "step": 799
+    },
+    {
+      "epoch": 0.28,
+      "learning_rate": 0.00033833152496749437,
+      "loss": 0.4847,
+      "step": 800
+    },
+    {
+      "epoch": 0.28,
+      "eval_loss": 0.5804653763771057,
+      "eval_runtime": 52.9435,
+      "eval_samples_per_second": 30.769,
+      "eval_steps_per_second": 30.769,
+      "step": 800
+    },
+    {
+      "epoch": 0.28,
+      "learning_rate": 0.00033816490548164605,
+      "loss": 0.4636,
+      "step": 801
+    },
+    {
+      "epoch": 0.28,
+      "learning_rate": 0.0003379981023570053,
+      "loss": 0.5453,
+      "step": 802
+    },
+    {
+      "epoch": 0.28,
+      "learning_rate": 0.0003378311158152747,
+      "loss": 0.5462,
+      "step": 803
+    },
+    {
+      "epoch": 0.28,
+      "learning_rate": 0.00033766394607840077,
+      "loss": 0.5114,
+      "step": 804
+    },
+    {
+      "epoch": 0.28,
+      "learning_rate": 0.0003374965933685733,
+      "loss": 0.717,
+      "step": 805
+    },
+    {
+      "epoch": 0.29,
+      "learning_rate": 0.00033732905790822556,
+      "loss": 0.5221,
+      "step": 806
+    },
+    {
+      "epoch": 0.29,
+      "learning_rate": 0.0003371613399200334,
+      "loss": 0.9044,
+      "step": 807
+    },
+    {
+      "epoch": 0.29,
+      "learning_rate": 0.0003369934396269155,
+      "loss": 0.6634,
+      "step": 808
+    },
+    {
+      "epoch": 0.29,
+      "learning_rate": 0.0003368253572520328,
+      "loss": 0.5681,
+      "step": 809
+    },
+    {
+      "epoch": 0.29,
+      "learning_rate": 0.0003366570930187882,
+      "loss": 0.4512,
+      "step": 810
+    },
+    {
+      "epoch": 0.29,
+      "learning_rate": 0.0003364886471508264,
+      "loss": 0.4719,
+      "step": 811
+    },
+    {
+      "epoch": 0.29,
+      "learning_rate": 0.0003363200198720333,
+      "loss": 0.5558,
+      "step": 812
+    },
+    {
+      "epoch": 0.29,
+      "learning_rate": 0.0003361512114065362,
+      "loss": 0.8677,
+      "step": 813
+    },
+    {
+      "epoch": 0.29,
+      "learning_rate": 0.0003359822219787032,
+      "loss": 0.5319,
+      "step": 814
+    },
+    {
+      "epoch": 0.29,
+      "learning_rate": 0.0003358130518131425,
+      "loss": 0.5503,
+      "step": 815
+    },
+    {
+      "epoch": 0.29,
+      "learning_rate": 0.00033564370113470314,
+      "loss": 0.5876,
+      "step": 816
+    },
+    {
+      "epoch": 0.29,
+      "learning_rate": 0.00033547417016847365,
+      "loss": 0.4266,
+      "step": 817
+    },
+    {
+      "epoch": 0.29,
+      "learning_rate": 0.0003353044591397823,
+      "loss": 0.5786,
+      "step": 818
+    },
+    {
+      "epoch": 0.29,
+      "learning_rate": 0.0003351345682741968,
+      "loss": 0.7438,
+      "step": 819
+    },
+    {
+      "epoch": 0.29,
+      "learning_rate": 0.00033496449779752373,
+      "loss": 0.4709,
+      "step": 820
+    },
+    {
+      "epoch": 0.29,
+      "eval_loss": 0.6139721870422363,
+      "eval_runtime": 52.9441,
+      "eval_samples_per_second": 30.768,
+      "eval_steps_per_second": 30.768,
+      "step": 820
+    },
+    {
+      "epoch": 0.29,
+      "learning_rate": 0.00033479424793580846,
+      "loss": 0.3748,
+      "step": 821
+    },
+    {
+      "epoch": 0.29,
+      "learning_rate": 0.00033462381891533483,
+      "loss": 0.588,
+      "step": 822
+    },
+    {
+      "epoch": 0.29,
+      "learning_rate": 0.0003344532109626247,
+      "loss": 0.6014,
+      "step": 823
+    },
+    {
+      "epoch": 0.29,
+      "learning_rate": 0.0003342824243044378,
+      "loss": 0.7484,
+      "step": 824
+    },
+    {
+      "epoch": 0.29,
+      "learning_rate": 0.0003341114591677715,
+      "loss": 0.4751,
+      "step": 825
+    },
+    {
+      "epoch": 0.29,
+      "learning_rate": 0.0003339403157798602,
+      "loss": 0.4915,
+      "step": 826
+    },
+    {
+      "epoch": 0.29,
+      "learning_rate": 0.0003337689943681753,
+      "loss": 0.4768,
+      "step": 827
+    },
+    {
+      "epoch": 0.29,
+      "learning_rate": 0.0003335974951604248,
+      "loss": 0.5708,
+      "step": 828
+    },
+    {
+      "epoch": 0.29,
+      "learning_rate": 0.0003334258183845531,
+      "loss": 0.6791,
+      "step": 829
+    },
+    {
+      "epoch": 0.29,
+      "learning_rate": 0.00033325396426874053,
+      "loss": 1.0993,
+      "step": 830
+    },
+    {
+      "epoch": 0.29,
+      "learning_rate": 0.00033308193304140314,
+      "loss": 0.4316,
+      "step": 831
+    },
+    {
+      "epoch": 0.29,
+      "learning_rate": 0.00033290972493119243,
+      "loss": 0.7035,
+      "step": 832
+    },
+    {
+      "epoch": 0.29,
+      "learning_rate": 0.0003327373401669948,
+      "loss": 0.5042,
+      "step": 833
+    },
+    {
+      "epoch": 0.3,
+      "learning_rate": 0.0003325647789779318,
+      "loss": 0.4783,
+      "step": 834
+    },
+    {
+      "epoch": 0.3,
+      "learning_rate": 0.0003323920415933591,
+      "loss": 0.532,
+      "step": 835
+    },
+    {
+      "epoch": 0.3,
+      "learning_rate": 0.00033221912824286687,
+      "loss": 0.5168,
+      "step": 836
+    },
+    {
+      "epoch": 0.3,
+      "learning_rate": 0.0003320460391562789,
+      "loss": 0.6661,
+      "step": 837
+    },
+    {
+      "epoch": 0.3,
+      "learning_rate": 0.00033187277456365284,
+      "loss": 0.4811,
+      "step": 838
+    },
+    {
+      "epoch": 0.3,
+      "learning_rate": 0.0003316993346952793,
+      "loss": 0.7792,
+      "step": 839
+    },
+    {
+      "epoch": 0.3,
+      "learning_rate": 0.000331525719781682,
+      "loss": 0.4279,
+      "step": 840
+    },
+    {
+      "epoch": 0.3,
+      "eval_loss": 0.5877469182014465,
+      "eval_runtime": 52.9463,
+      "eval_samples_per_second": 30.767,
+      "eval_steps_per_second": 30.767,
+      "step": 840
+    },
+    {
+      "epoch": 0.3,
+      "learning_rate": 0.0003313519300536174,
+      "loss": 0.6581,
+      "step": 841
+    },
+    {
+      "epoch": 0.3,
+      "learning_rate": 0.00033117796574207414,
+      "loss": 0.6532,
+      "step": 842
+    },
+    {
+      "epoch": 0.3,
+      "learning_rate": 0.0003310038270782731,
+      "loss": 0.5394,
+      "step": 843
+    },
+    {
+      "epoch": 0.3,
+      "learning_rate": 0.0003308295142936667,
+      "loss": 0.7435,
+      "step": 844
+    },
+    {
+      "epoch": 0.3,
+      "learning_rate": 0.00033065502761993895,
+      "loss": 0.4245,
+      "step": 845
+    },
+    {
+      "epoch": 0.3,
+      "learning_rate": 0.00033048036728900484,
+      "loss": 0.4947,
+      "step": 846
+    },
+    {
+      "epoch": 0.3,
+      "learning_rate": 0.00033030553353301023,
+      "loss": 0.4603,
+      "step": 847
+    },
+    {
+      "epoch": 0.3,
+      "learning_rate": 0.0003301305265843315,
+      "loss": 0.5088,
+      "step": 848
+    },
+    {
+      "epoch": 0.3,
+      "learning_rate": 0.0003299553466755754,
+      "loss": 0.7071,
+      "step": 849
+    },
+    {
+      "epoch": 0.3,
+      "learning_rate": 0.00032977999403957807,
+      "loss": 0.4315,
+      "step": 850
+    },
+    {
+      "epoch": 0.3,
+      "learning_rate": 0.00032960446890940577,
+      "loss": 0.5741,
+      "step": 851
+    },
+    {
+      "epoch": 0.3,
+      "learning_rate": 0.00032942877151835375,
+      "loss": 0.412,
+      "step": 852
+    },
+    {
+      "epoch": 0.3,
+      "learning_rate": 0.0003292529020999462,
+      "loss": 0.4505,
+      "step": 853
+    },
+    {
+      "epoch": 0.3,
+      "learning_rate": 0.0003290768608879362,
+      "loss": 0.4506,
+      "step": 854
+    },
+    {
+      "epoch": 0.3,
+      "learning_rate": 0.0003289006481163048,
+      "loss": 0.4813,
+      "step": 855
+    },
+    {
+      "epoch": 0.3,
+      "learning_rate": 0.0003287242640192613,
+      "loss": 0.5475,
+      "step": 856
+    },
+    {
+      "epoch": 0.3,
+      "learning_rate": 0.00032854770883124267,
+      "loss": 0.6268,
+      "step": 857
+    },
+    {
+      "epoch": 0.3,
+      "learning_rate": 0.00032837098278691335,
+      "loss": 0.5765,
+      "step": 858
+    },
+    {
+      "epoch": 0.3,
+      "learning_rate": 0.0003281940861211648,
+      "loss": 0.346,
+      "step": 859
+    },
+    {
+      "epoch": 0.3,
+      "learning_rate": 0.0003280170190691152,
+      "loss": 0.7142,
+      "step": 860
+    },
+    {
+      "epoch": 0.3,
+      "eval_loss": 0.5801124572753906,
+      "eval_runtime": 52.9379,
+      "eval_samples_per_second": 30.772,
+      "eval_steps_per_second": 30.772,
+      "step": 860
+    },
+    {
+      "epoch": 0.3,
+      "learning_rate": 0.0003278397818661093,
+      "loss": 0.9087,
+      "step": 861
+    },
+    {
+      "epoch": 0.31,
+      "learning_rate": 0.00032766237474771804,
+      "loss": 0.5387,
+      "step": 862
+    },
+    {
+      "epoch": 0.31,
+      "learning_rate": 0.000327484797949738,
+      "loss": 0.6743,
+      "step": 863
+    },
+    {
+      "epoch": 0.31,
+      "learning_rate": 0.00032730705170819143,
+      "loss": 0.4575,
+      "step": 864
+    },
+    {
+      "epoch": 0.31,
+      "learning_rate": 0.00032712913625932577,
+      "loss": 0.4863,
+      "step": 865
+    },
+    {
+      "epoch": 0.31,
+      "learning_rate": 0.00032695105183961346,
+      "loss": 0.44,
+      "step": 866
+    },
+    {
+      "epoch": 0.31,
+      "learning_rate": 0.00032677279868575124,
+      "loss": 0.4757,
+      "step": 867
+    },
+    {
+      "epoch": 0.31,
+      "learning_rate": 0.0003265943770346605,
+      "loss": 0.4846,
+      "step": 868
+    },
+    {
+      "epoch": 0.31,
+      "learning_rate": 0.00032641578712348617,
+      "loss": 0.4258,
+      "step": 869
+    },
+    {
+      "epoch": 0.31,
+      "learning_rate": 0.00032623702918959725,
+      "loss": 0.4437,
+      "step": 870
+    },
+    {
+      "epoch": 0.31,
+      "learning_rate": 0.0003260581034705857,
+      "loss": 0.6164,
+      "step": 871
+    },
+    {
+      "epoch": 0.31,
+      "learning_rate": 0.0003258790102042667,
+      "loss": 0.8449,
+      "step": 872
+    },
+    {
+      "epoch": 0.31,
+      "learning_rate": 0.000325699749628678,
+      "loss": 0.3281,
+      "step": 873
+    },
+    {
+      "epoch": 0.31,
+      "learning_rate": 0.0003255203219820799,
+      "loss": 0.5365,
+      "step": 874
+    },
+    {
+      "epoch": 0.31,
+      "learning_rate": 0.0003253407275029545,
+      "loss": 0.5681,
+      "step": 875
+    },
+    {
+      "epoch": 0.31,
+      "learning_rate": 0.0003251609664300059,
+      "loss": 0.4682,
+      "step": 876
+    },
+    {
+      "epoch": 0.31,
+      "learning_rate": 0.00032498103900215937,
+      "loss": 0.3016,
+      "step": 877
+    },
+    {
+      "epoch": 0.31,
+      "learning_rate": 0.00032480094545856163,
+      "loss": 0.9157,
+      "step": 878
+    },
+    {
+      "epoch": 0.31,
+      "learning_rate": 0.0003246206860385798,
+      "loss": 0.4768,
+      "step": 879
+    },
+    {
+      "epoch": 0.31,
+      "learning_rate": 0.0003244402609818017,
+      "loss": 1.0536,
+      "step": 880
+    },
+    {
+      "epoch": 0.31,
+      "eval_loss": 0.6101779937744141,
+      "eval_runtime": 52.945,
+      "eval_samples_per_second": 30.768,
+      "eval_steps_per_second": 30.768,
+      "step": 880
+    },
+    {
+      "epoch": 0.31,
+      "learning_rate": 0.0003242596705280353,
+      "loss": 0.5903,
+      "step": 881
+    },
+    {
+      "epoch": 0.31,
+      "learning_rate": 0.00032407891491730836,
+      "loss": 0.5885,
+      "step": 882
+    },
+    {
+      "epoch": 0.31,
+      "learning_rate": 0.0003238979943898681,
+      "loss": 0.415,
+      "step": 883
+    },
+    {
+      "epoch": 0.31,
+      "learning_rate": 0.00032371690918618103,
+      "loss": 0.4641,
+      "step": 884
+    },
+    {
+      "epoch": 0.31,
+      "learning_rate": 0.0003235356595469325,
+      "loss": 0.5085,
+      "step": 885
+    },
+    {
+      "epoch": 0.31,
+      "learning_rate": 0.00032335424571302646,
+      "loss": 0.5616,
+      "step": 886
+    },
+    {
+      "epoch": 0.31,
+      "learning_rate": 0.0003231726679255849,
+      "loss": 0.6703,
+      "step": 887
+    },
+    {
+      "epoch": 0.31,
+      "learning_rate": 0.00032299092642594794,
+      "loss": 0.343,
+      "step": 888
+    },
+    {
+      "epoch": 0.31,
+      "learning_rate": 0.0003228090214556733,
+      "loss": 0.459,
+      "step": 889
+    },
+    {
+      "epoch": 0.31,
+      "learning_rate": 0.00032262695325653584,
+      "loss": 0.3299,
+      "step": 890
+    },
+    {
+      "epoch": 0.32,
+      "learning_rate": 0.0003224447220705274,
+      "loss": 0.3559,
+      "step": 891
+    },
+    {
+      "epoch": 0.32,
+      "learning_rate": 0.0003222623281398566,
+      "loss": 0.7606,
+      "step": 892
+    },
+    {
+      "epoch": 0.32,
+      "learning_rate": 0.00032207977170694814,
+      "loss": 0.3042,
+      "step": 893
+    },
+    {
+      "epoch": 0.32,
+      "learning_rate": 0.00032189705301444285,
+      "loss": 0.2309,
+      "step": 894
+    },
+    {
+      "epoch": 0.32,
+      "learning_rate": 0.00032171417230519725,
+      "loss": 0.4136,
+      "step": 895
+    },
+    {
+      "epoch": 0.32,
+      "learning_rate": 0.00032153112982228303,
+      "loss": 0.5464,
+      "step": 896
+    },
+    {
+      "epoch": 0.32,
+      "learning_rate": 0.0003213479258089871,
+      "loss": 0.5482,
+      "step": 897
+    },
+    {
+      "epoch": 0.32,
+      "learning_rate": 0.00032116456050881095,
+      "loss": 0.5362,
+      "step": 898
+    },
+    {
+      "epoch": 0.32,
+      "learning_rate": 0.0003209810341654705,
+      "loss": 0.4796,
+      "step": 899
+    },
+    {
+      "epoch": 0.32,
+      "learning_rate": 0.00032079734702289555,
+      "loss": 0.694,
+      "step": 900
+    },
+    {
+      "epoch": 0.32,
+      "eval_loss": 0.5811580419540405,
+      "eval_runtime": 52.9399,
+      "eval_samples_per_second": 30.771,
+      "eval_steps_per_second": 30.771,
+      "step": 900
+    },
+    {
+      "epoch": 0.32,
+      "learning_rate": 0.0003206134993252299,
+      "loss": 0.5679,
+      "step": 901
+    },
+    {
+      "epoch": 0.32,
+      "learning_rate": 0.0003204294913168305,
+      "loss": 0.8856,
+      "step": 902
+    },
+    {
+      "epoch": 0.32,
+      "learning_rate": 0.00032024532324226757,
+      "loss": 0.5797,
+      "step": 903
+    },
+    {
+      "epoch": 0.32,
+      "learning_rate": 0.0003200609953463239,
+      "loss": 0.5059,
+      "step": 904
+    },
+    {
+      "epoch": 0.32,
+      "learning_rate": 0.0003198765078739949,
+      "loss": 0.3525,
+      "step": 905
+    },
+    {
+      "epoch": 0.32,
+      "learning_rate": 0.0003196918610704879,
+      "loss": 0.4511,
+      "step": 906
+    },
+    {
+      "epoch": 0.32,
+      "learning_rate": 0.000319507055181222,
+      "loss": 0.548,
+      "step": 907
+    },
+    {
+      "epoch": 0.32,
+      "learning_rate": 0.0003193220904518279,
+      "loss": 0.4018,
+      "step": 908
+    },
+    {
+      "epoch": 0.32,
+      "learning_rate": 0.0003191369671281473,
+      "loss": 0.842,
+      "step": 909
+    },
+    {
+      "epoch": 0.32,
+      "learning_rate": 0.0003189516854562328,
+      "loss": 0.6802,
+      "step": 910
+    },
+    {
+      "epoch": 0.32,
+      "learning_rate": 0.0003187662456823473,
+      "loss": 0.5293,
+      "step": 911
+    },
+    {
+      "epoch": 0.32,
+      "learning_rate": 0.0003185806480529639,
+      "loss": 0.5778,
+      "step": 912
+    },
+    {
+      "epoch": 0.32,
+      "learning_rate": 0.0003183948928147656,
+      "loss": 0.6594,
+      "step": 913
+    },
+    {
+      "epoch": 0.32,
+      "learning_rate": 0.00031820898021464475,
+      "loss": 1.1996,
+      "step": 914
+    },
+    {
+      "epoch": 0.32,
+      "learning_rate": 0.00031802291049970303,
+      "loss": 0.6401,
+      "step": 915
+    },
+    {
+      "epoch": 0.32,
+      "learning_rate": 0.0003178366839172508,
+      "loss": 0.5293,
+      "step": 916
+    },
+    {
+      "epoch": 0.32,
+      "learning_rate": 0.0003176503007148069,
+      "loss": 0.6464,
+      "step": 917
+    },
+    {
+      "epoch": 0.32,
+      "learning_rate": 0.0003174637611400984,
+      "loss": 0.6673,
+      "step": 918
+    },
+    {
+      "epoch": 0.33,
+      "learning_rate": 0.0003172770654410603,
+      "loss": 0.4422,
+      "step": 919
+    },
+    {
+      "epoch": 0.33,
+      "learning_rate": 0.0003170902138658349,
+      "loss": 0.5034,
+      "step": 920
+    },
+    {
+      "epoch": 0.33,
+      "eval_loss": 0.5833107233047485,
+      "eval_runtime": 52.9444,
+      "eval_samples_per_second": 30.768,
+      "eval_steps_per_second": 30.768,
+      "step": 920
+    },
+    {
+      "epoch": 0.33,
+      "learning_rate": 0.0003169032066627718,
+      "loss": 0.5463,
+      "step": 921
+    },
+    {
+      "epoch": 0.33,
+      "learning_rate": 0.0003167160440804275,
+      "loss": 0.6004,
+      "step": 922
+    },
+    {
+      "epoch": 0.33,
+      "learning_rate": 0.00031652872636756495,
+      "loss": 0.6518,
+      "step": 923
+    },
+    {
+      "epoch": 0.33,
+      "learning_rate": 0.00031634125377315325,
+      "loss": 0.5229,
+      "step": 924
+    },
+    {
+      "epoch": 0.33,
+      "learning_rate": 0.0003161536265463674,
+      "loss": 0.3916,
+      "step": 925
+    },
+    {
+      "epoch": 0.33,
+      "learning_rate": 0.000315965844936588,
+      "loss": 0.6403,
+      "step": 926
+    },
+    {
+      "epoch": 0.33,
+      "learning_rate": 0.0003157779091934007,
+      "loss": 0.4317,
+      "step": 927
+    },
+    {
+      "epoch": 0.33,
+      "learning_rate": 0.00031558981956659614,
+      "loss": 0.5661,
+      "step": 928
+    },
+    {
+      "epoch": 0.33,
+      "learning_rate": 0.00031540157630616944,
+      "loss": 0.2793,
+      "step": 929
+    },
+    {
+      "epoch": 0.33,
+      "learning_rate": 0.00031521317966231986,
+      "loss": 0.4087,
+      "step": 930
+    },
+    {
+      "epoch": 0.33,
+      "learning_rate": 0.00031502462988545067,
+      "loss": 0.4256,
+      "step": 931
+    },
+    {
+      "epoch": 0.33,
+      "learning_rate": 0.0003148359272261685,
+      "loss": 0.4178,
+      "step": 932
+    },
+    {
+      "epoch": 0.33,
+      "learning_rate": 0.00031464707193528334,
+      "loss": 0.424,
+      "step": 933
+    },
+    {
+      "epoch": 0.33,
+      "learning_rate": 0.00031445806426380786,
+      "loss": 0.3553,
+      "step": 934
+    },
+    {
+      "epoch": 0.33,
+      "learning_rate": 0.0003142689044629575,
+      "loss": 0.4308,
+      "step": 935
+    },
+    {
+      "epoch": 0.33,
+      "learning_rate": 0.00031407959278414977,
+      "loss": 0.3873,
+      "step": 936
+    },
+    {
+      "epoch": 0.33,
+      "learning_rate": 0.00031389012947900395,
+      "loss": 0.3957,
+      "step": 937
+    },
+    {
+      "epoch": 0.33,
+      "learning_rate": 0.0003137005147993411,
+      "loss": 0.4488,
+      "step": 938
+    },
+    {
+      "epoch": 0.33,
+      "learning_rate": 0.00031351074899718315,
+      "loss": 0.6238,
+      "step": 939
+    },
+    {
+      "epoch": 0.33,
+      "learning_rate": 0.0003133208323247532,
+      "loss": 0.4208,
+      "step": 940
+    },
+    {
+      "epoch": 0.33,
+      "eval_loss": 0.5803355574607849,
+      "eval_runtime": 52.9407,
+      "eval_samples_per_second": 30.77,
+      "eval_steps_per_second": 30.77,
+      "step": 940
+    },
+    {
+      "epoch": 0.33,
+      "learning_rate": 0.0003131307650344747,
+      "loss": 0.5983,
+      "step": 941
+    },
+    {
+      "epoch": 0.33,
+      "learning_rate": 0.00031294054737897136,
+      "loss": 0.5357,
+      "step": 942
+    },
+    {
+      "epoch": 0.33,
+      "learning_rate": 0.0003127501796110667,
+      "loss": 0.4599,
+      "step": 943
+    },
+    {
+      "epoch": 0.33,
+      "learning_rate": 0.0003125596619837838,
+      "loss": 0.8019,
+      "step": 944
+    },
+    {
+      "epoch": 0.33,
+      "learning_rate": 0.00031236899475034486,
+      "loss": 0.6881,
+      "step": 945
+    },
+    {
+      "epoch": 0.33,
+      "learning_rate": 0.000312178178164171,
+      "loss": 0.4076,
+      "step": 946
+    },
+    {
+      "epoch": 0.34,
+      "learning_rate": 0.0003119872124788818,
+      "loss": 0.4997,
+      "step": 947
+    },
+    {
+      "epoch": 0.34,
+      "learning_rate": 0.00031179609794829507,
+      "loss": 0.3114,
+      "step": 948
+    },
+    {
+      "epoch": 0.34,
+      "learning_rate": 0.00031160483482642635,
+      "loss": 0.5566,
+      "step": 949
+    },
+    {
+      "epoch": 0.34,
+      "learning_rate": 0.00031141342336748877,
+      "loss": 0.5543,
+      "step": 950
+    },
+    {
+      "epoch": 0.34,
+      "learning_rate": 0.0003112218638258926,
+      "loss": 0.8279,
+      "step": 951
+    },
+    {
+      "epoch": 0.34,
+      "learning_rate": 0.00031103015645624483,
+      "loss": 0.5271,
+      "step": 952
+    },
+    {
+      "epoch": 0.34,
+      "learning_rate": 0.0003108383015133492,
+      "loss": 0.1997,
+      "step": 953
+    },
+    {
+      "epoch": 0.34,
+      "learning_rate": 0.0003106462992522053,
+      "loss": 0.4099,
+      "step": 954
+    },
+    {
+      "epoch": 0.34,
+      "learning_rate": 0.00031045414992800853,
+      "loss": 0.3807,
+      "step": 955
+    },
+    {
+      "epoch": 0.34,
+      "learning_rate": 0.0003102618537961501,
+      "loss": 0.4854,
+      "step": 956
+    },
+    {
+      "epoch": 0.34,
+      "learning_rate": 0.0003100694111122159,
+      "loss": 0.3855,
+      "step": 957
+    },
+    {
+      "epoch": 0.34,
+      "learning_rate": 0.000309876822131987,
+      "loss": 0.6301,
+      "step": 958
+    },
+    {
+      "epoch": 0.34,
+      "learning_rate": 0.00030968408711143857,
+      "loss": 1.0127,
+      "step": 959
+    },
+    {
+      "epoch": 0.34,
+      "learning_rate": 0.0003094912063067401,
+      "loss": 0.5917,
+      "step": 960
+    },
+    {
+      "epoch": 0.34,
+      "eval_loss": 0.5756197571754456,
+      "eval_runtime": 52.937,
+      "eval_samples_per_second": 30.772,
+      "eval_steps_per_second": 30.772,
+      "step": 960
+    },
+    {
+      "epoch": 0.34,
+      "learning_rate": 0.00030929817997425485,
+      "loss": 0.5026,
+      "step": 961
+    },
+    {
+      "epoch": 0.34,
+      "learning_rate": 0.0003091050083705395,
+      "loss": 0.4194,
+      "step": 962
+    },
+    {
+      "epoch": 0.34,
+      "learning_rate": 0.0003089116917523436,
+      "loss": 0.5897,
+      "step": 963
+    },
+    {
+      "epoch": 0.34,
+      "learning_rate": 0.0003087182303766097,
+      "loss": 0.6967,
+      "step": 964
+    },
+    {
+      "epoch": 0.34,
+      "learning_rate": 0.0003085246245004726,
+      "loss": 0.404,
+      "step": 965
+    },
+    {
+      "epoch": 0.34,
+      "learning_rate": 0.00030833087438125934,
+      "loss": 0.5646,
+      "step": 966
+    },
+    {
+      "epoch": 0.34,
+      "learning_rate": 0.00030813698027648843,
+      "loss": 0.3935,
+      "step": 967
+    },
+    {
+      "epoch": 0.34,
+      "learning_rate": 0.00030794294244387,
+      "loss": 0.4531,
+      "step": 968
+    },
+    {
+      "epoch": 0.34,
+      "learning_rate": 0.000307748761141305,
+      "loss": 0.5795,
+      "step": 969
+    },
+    {
+      "epoch": 0.34,
+      "learning_rate": 0.0003075544366268851,
+      "loss": 0.5793,
+      "step": 970
+    },
+    {
+      "epoch": 0.34,
+      "learning_rate": 0.00030735996915889257,
+      "loss": 0.667,
+      "step": 971
+    },
+    {
+      "epoch": 0.34,
+      "learning_rate": 0.0003071653589957993,
+      "loss": 0.4965,
+      "step": 972
+    },
+    {
+      "epoch": 0.34,
+      "learning_rate": 0.0003069706063962673,
+      "loss": 0.422,
+      "step": 973
+    },
+    {
+      "epoch": 0.34,
+      "learning_rate": 0.0003067757116191473,
+      "loss": 0.3609,
+      "step": 974
+    },
+    {
+      "epoch": 0.35,
+      "learning_rate": 0.0003065806749234795,
+      "loss": 0.4492,
+      "step": 975
+    },
+    {
+      "epoch": 0.35,
+      "learning_rate": 0.0003063854965684926,
+      "loss": 0.5979,
+      "step": 976
+    },
+    {
+      "epoch": 0.35,
+      "learning_rate": 0.00030619017681360344,
+      "loss": 0.6323,
+      "step": 977
+    },
+    {
+      "epoch": 0.35,
+      "learning_rate": 0.000305994715918417,
+      "loss": 0.4933,
+      "step": 978
+    },
+    {
+      "epoch": 0.35,
+      "learning_rate": 0.00030579911414272564,
+      "loss": 0.5725,
+      "step": 979
+    },
+    {
+      "epoch": 0.35,
+      "learning_rate": 0.0003056033717465092,
+      "loss": 0.4655,
+      "step": 980
+    },
+    {
+      "epoch": 0.35,
+      "eval_loss": 0.5705834031105042,
+      "eval_runtime": 52.9443,
+      "eval_samples_per_second": 30.768,
+      "eval_steps_per_second": 30.768,
+      "step": 980
+    },
+    {
+      "epoch": 0.35,
+      "learning_rate": 0.00030540748898993426,
+      "loss": 0.4179,
+      "step": 981
+    },
+    {
+      "epoch": 0.35,
+      "learning_rate": 0.000305211466133354,
+      "loss": 0.5273,
+      "step": 982
+    },
+    {
+      "epoch": 0.35,
+      "learning_rate": 0.00030501530343730784,
+      "loss": 0.2396,
+      "step": 983
+    },
+    {
+      "epoch": 0.35,
+      "learning_rate": 0.00030481900116252094,
+      "loss": 0.5257,
+      "step": 984
+    },
+    {
+      "epoch": 0.35,
+      "learning_rate": 0.00030462255956990414,
+      "loss": 0.5143,
+      "step": 985
+    },
+    {
+      "epoch": 0.35,
+      "learning_rate": 0.0003044259789205535,
+      "loss": 0.5708,
+      "step": 986
+    },
+    {
+      "epoch": 0.35,
+      "learning_rate": 0.0003042292594757496,
+      "loss": 0.4783,
+      "step": 987
+    },
+    {
+      "epoch": 0.35,
+      "learning_rate": 0.0003040324014969578,
+      "loss": 0.4965,
+      "step": 988
+    },
+    {
+      "epoch": 0.35,
+      "learning_rate": 0.00030383540524582737,
+      "loss": 0.4603,
+      "step": 989
+    },
+    {
+      "epoch": 0.35,
+      "learning_rate": 0.00030363827098419164,
+      "loss": 0.3645,
+      "step": 990
+    },
+    {
+      "epoch": 0.35,
+      "learning_rate": 0.0003034409989740671,
+      "loss": 0.9198,
+      "step": 991
+    },
+    {
+      "epoch": 0.35,
+      "learning_rate": 0.00030324358947765335,
+      "loss": 0.43,
+      "step": 992
+    },
+    {
+      "epoch": 0.35,
+      "learning_rate": 0.00030304604275733306,
+      "loss": 0.4155,
+      "step": 993
+    },
+    {
+      "epoch": 0.35,
+      "learning_rate": 0.0003028483590756707,
+      "loss": 0.6906,
+      "step": 994
+    },
+    {
+      "epoch": 0.35,
+      "learning_rate": 0.0003026505386954134,
+      "loss": 0.4546,
+      "step": 995
+    },
+    {
+      "epoch": 0.35,
+      "learning_rate": 0.0003024525818794895,
+      "loss": 0.5577,
+      "step": 996
+    },
+    {
+      "epoch": 0.35,
+      "learning_rate": 0.00030225448889100896,
+      "loss": 0.4469,
+      "step": 997
+    },
+    {
+      "epoch": 0.35,
+      "learning_rate": 0.00030205625999326274,
+      "loss": 0.474,
+      "step": 998
+    },
+    {
+      "epoch": 0.35,
+      "learning_rate": 0.00030185789544972207,
+      "loss": 0.8963,
+      "step": 999
+    },
+    {
+      "epoch": 0.35,
+      "learning_rate": 0.00030165939552403894,
+      "loss": 0.4274,
+      "step": 1000
+    },
+    {
+      "epoch": 0.35,
+      "eval_loss": 0.5674954652786255,
+      "eval_runtime": 52.9704,
+      "eval_samples_per_second": 30.753,
+      "eval_steps_per_second": 30.753,
+      "step": 1000
+    },
+    {
+      "epoch": 0.35,
+      "learning_rate": 0.00030146076048004507,
+      "loss": 0.3301,
+      "step": 1001
+    },
+    {
+      "epoch": 0.35,
+      "learning_rate": 0.00030126199058175176,
+      "loss": 0.9746,
+      "step": 1002
+    },
+    {
+      "epoch": 0.35,
+      "learning_rate": 0.0003010630860933496,
+      "loss": 0.2435,
+      "step": 1003
+    },
+    {
+      "epoch": 0.36,
+      "learning_rate": 0.000300864047279208,
+      "loss": 0.3932,
+      "step": 1004
+    },
+    {
+      "epoch": 0.36,
+      "learning_rate": 0.0003006648744038749,
+      "loss": 0.6331,
+      "step": 1005
+    },
+    {
+      "epoch": 0.36,
+      "learning_rate": 0.00030046556773207666,
+      "loss": 0.6568,
+      "step": 1006
+    },
+    {
+      "epoch": 0.36,
+      "learning_rate": 0.000300266127528717,
+      "loss": 0.5433,
+      "step": 1007
+    },
+    {
+      "epoch": 0.36,
+      "learning_rate": 0.0003000665540588777,
+      "loss": 1.1451,
+      "step": 1008
+    },
+    {
+      "epoch": 0.36,
+      "learning_rate": 0.00029986684758781713,
+      "loss": 0.3636,
+      "step": 1009
+    },
+    {
+      "epoch": 0.36,
+      "learning_rate": 0.0002996670083809709,
+      "loss": 0.4939,
+      "step": 1010
+    },
+    {
+      "epoch": 0.36,
+      "learning_rate": 0.00029946703670395057,
+      "loss": 0.5951,
+      "step": 1011
+    },
+    {
+      "epoch": 0.36,
+      "learning_rate": 0.00029926693282254416,
+      "loss": 0.7069,
+      "step": 1012
+    },
+    {
+      "epoch": 0.36,
+      "learning_rate": 0.0002990666970027153,
+      "loss": 0.5341,
+      "step": 1013
+    },
+    {
+      "epoch": 0.36,
+      "learning_rate": 0.0002988663295106029,
+      "loss": 0.6788,
+      "step": 1014
+    },
+    {
+      "epoch": 0.36,
+      "learning_rate": 0.0002986658306125209,
+      "loss": 0.6551,
+      "step": 1015
+    },
+    {
+      "epoch": 0.36,
+      "learning_rate": 0.000298465200574958,
+      "loss": 0.4747,
+      "step": 1016
+    },
+    {
+      "epoch": 0.36,
+      "learning_rate": 0.000298264439664577,
+      "loss": 0.2433,
+      "step": 1017
+    },
+    {
+      "epoch": 0.36,
+      "learning_rate": 0.00029806354814821496,
+      "loss": 0.4391,
+      "step": 1018
+    },
+    {
+      "epoch": 0.36,
+      "learning_rate": 0.00029786252629288213,
+      "loss": 0.8988,
+      "step": 1019
+    },
+    {
+      "epoch": 0.36,
+      "learning_rate": 0.0002976613743657624,
+      "loss": 0.3711,
+      "step": 1020
+    },
+    {
+      "epoch": 0.36,
+      "eval_loss": 0.5790918469429016,
+      "eval_runtime": 52.9396,
+      "eval_samples_per_second": 30.771,
+      "eval_steps_per_second": 30.771,
+      "step": 1020
+    },
+    {
+      "epoch": 0.36,
+      "learning_rate": 0.00029746009263421225,
+      "loss": 0.3897,
+      "step": 1021
+    },
+    {
+      "epoch": 0.36,
+      "learning_rate": 0.0002972586813657608,
+      "loss": 0.607,
+      "step": 1022
+    },
+    {
+      "epoch": 0.36,
+      "learning_rate": 0.0002970571408281093,
+      "loss": 0.6208,
+      "step": 1023
+    },
+    {
+      "epoch": 0.36,
+      "learning_rate": 0.0002968554712891309,
+      "loss": 0.4218,
+      "step": 1024
+    },
+    {
+      "epoch": 0.36,
+      "learning_rate": 0.0002966536730168702,
+      "loss": 0.5899,
+      "step": 1025
+    },
+    {
+      "epoch": 0.36,
+      "learning_rate": 0.0002964517462795427,
+      "loss": 0.388,
+      "step": 1026
+    },
+    {
+      "epoch": 0.36,
+      "learning_rate": 0.000296249691345535,
+      "loss": 0.487,
+      "step": 1027
+    },
+    {
+      "epoch": 0.36,
+      "learning_rate": 0.00029604750848340373,
+      "loss": 0.492,
+      "step": 1028
+    },
+    {
+      "epoch": 0.36,
+      "learning_rate": 0.00029584519796187577,
+      "loss": 0.4474,
+      "step": 1029
+    },
+    {
+      "epoch": 0.36,
+      "learning_rate": 0.0002956427600498477,
+      "loss": 0.4696,
+      "step": 1030
+    },
+    {
+      "epoch": 0.36,
+      "learning_rate": 0.00029544019501638526,
+      "loss": 0.6235,
+      "step": 1031
+    },
+    {
+      "epoch": 0.37,
+      "learning_rate": 0.00029523750313072323,
+      "loss": 0.58,
+      "step": 1032
+    },
+    {
+      "epoch": 0.37,
+      "learning_rate": 0.00029503468466226504,
+      "loss": 0.5844,
+      "step": 1033
+    },
+    {
+      "epoch": 0.37,
+      "learning_rate": 0.00029483173988058227,
+      "loss": 0.4359,
+      "step": 1034
+    },
+    {
+      "epoch": 0.37,
+      "learning_rate": 0.00029462866905541444,
+      "loss": 0.873,
+      "step": 1035
+    },
+    {
+      "epoch": 0.37,
+      "learning_rate": 0.00029442547245666864,
+      "loss": 0.3551,
+      "step": 1036
+    },
+    {
+      "epoch": 0.37,
+      "learning_rate": 0.00029422215035441904,
+      "loss": 0.4442,
+      "step": 1037
+    },
+    {
+      "epoch": 0.37,
+      "learning_rate": 0.0002940187030189067,
+      "loss": 0.5035,
+      "step": 1038
+    },
+    {
+      "epoch": 0.37,
+      "learning_rate": 0.000293815130720539,
+      "loss": 0.4747,
+      "step": 1039
+    },
+    {
+      "epoch": 0.37,
+      "learning_rate": 0.0002936114337298896,
+      "loss": 0.6792,
+      "step": 1040
+    },
+    {
+      "epoch": 0.37,
+      "eval_loss": 0.5772663354873657,
+      "eval_runtime": 52.9492,
+      "eval_samples_per_second": 30.765,
+      "eval_steps_per_second": 30.765,
+      "step": 1040
+    },
+    {
+      "epoch": 0.37,
+      "learning_rate": 0.0002934076123176978,
+      "loss": 0.4803,
+      "step": 1041
+    },
+    {
+      "epoch": 0.37,
+      "learning_rate": 0.0002932036667548683,
+      "loss": 0.406,
+      "step": 1042
+    },
+    {
+      "epoch": 0.37,
+      "learning_rate": 0.0002929995973124706,
+      "loss": 0.4428,
+      "step": 1043
+    },
+    {
+      "epoch": 0.37,
+      "learning_rate": 0.0002927954042617392,
+      "loss": 0.5016,
+      "step": 1044
+    },
+    {
+      "epoch": 0.37,
+      "learning_rate": 0.0002925910878740727,
+      "loss": 0.505,
+      "step": 1045
+    },
+    {
+      "epoch": 0.37,
+      "learning_rate": 0.0002923866484210336,
+      "loss": 0.6247,
+      "step": 1046
+    },
+    {
+      "epoch": 0.37,
+      "learning_rate": 0.0002921820861743481,
+      "loss": 0.3394,
+      "step": 1047
+    },
+    {
+      "epoch": 0.37,
+      "learning_rate": 0.0002919774014059055,
+      "loss": 0.4262,
+      "step": 1048
+    },
+    {
+      "epoch": 0.37,
+      "learning_rate": 0.00029177259438775786,
+      "loss": 0.9232,
+      "step": 1049
+    },
+    {
+      "epoch": 0.37,
+      "learning_rate": 0.0002915676653921201,
+      "loss": 0.4547,
+      "step": 1050
+    },
+    {
+      "epoch": 0.37,
+      "learning_rate": 0.0002913626146913688,
+      "loss": 0.6957,
+      "step": 1051
+    },
+    {
+      "epoch": 0.37,
+      "learning_rate": 0.00029115744255804257,
+      "loss": 0.8388,
+      "step": 1052
+    },
+    {
+      "epoch": 0.37,
+      "learning_rate": 0.00029095214926484135,
+      "loss": 0.657,
+      "step": 1053
+    },
+    {
+      "epoch": 0.37,
+      "learning_rate": 0.00029074673508462613,
+      "loss": 0.488,
+      "step": 1054
+    },
+    {
+      "epoch": 0.37,
+      "learning_rate": 0.00029054120029041865,
+      "loss": 0.3164,
+      "step": 1055
+    },
+    {
+      "epoch": 0.37,
+      "learning_rate": 0.00029033554515540077,
+      "loss": 0.6609,
+      "step": 1056
+    },
+    {
+      "epoch": 0.37,
+      "learning_rate": 0.00029012976995291444,
+      "loss": 0.5037,
+      "step": 1057
+    },
+    {
+      "epoch": 0.37,
+      "learning_rate": 0.0002899238749564612,
+      "loss": 0.4864,
+      "step": 1058
+    },
+    {
+      "epoch": 0.37,
+      "learning_rate": 0.0002897178604397018,
+      "loss": 0.6515,
+      "step": 1059
+    },
+    {
+      "epoch": 0.38,
+      "learning_rate": 0.0002895117266764557,
+      "loss": 0.3607,
+      "step": 1060
+    },
+    {
+      "epoch": 0.38,
+      "eval_loss": 0.5837768316268921,
+      "eval_runtime": 52.9451,
+      "eval_samples_per_second": 30.768,
+      "eval_steps_per_second": 30.768,
+      "step": 1060
+    },
+    {
+      "epoch": 0.38,
+      "learning_rate": 0.0002893054739407011,
+      "loss": 0.4446,
+      "step": 1061
+    },
+    {
+      "epoch": 0.38,
+      "learning_rate": 0.0002890991025065742,
+      "loss": 0.5464,
+      "step": 1062
+    },
+    {
+      "epoch": 0.38,
+      "learning_rate": 0.00028889261264836895,
+      "loss": 0.3211,
+      "step": 1063
+    },
+    {
+      "epoch": 0.38,
+      "learning_rate": 0.00028868600464053673,
+      "loss": 0.8575,
+      "step": 1064
+    },
+    {
+      "epoch": 0.38,
+      "learning_rate": 0.00028847927875768584,
+      "loss": 0.3476,
+      "step": 1065
+    },
+    {
+      "epoch": 0.38,
+      "learning_rate": 0.0002882724352745815,
+      "loss": 0.5673,
+      "step": 1066
+    },
+    {
+      "epoch": 0.38,
+      "learning_rate": 0.00028806547446614514,
+      "loss": 0.5144,
+      "step": 1067
+    },
+    {
+      "epoch": 0.38,
+      "learning_rate": 0.00028785839660745396,
+      "loss": 0.3791,
+      "step": 1068
+    },
+    {
+      "epoch": 0.38,
+      "learning_rate": 0.0002876512019737409,
+      "loss": 0.6135,
+      "step": 1069
+    },
+    {
+      "epoch": 0.38,
+      "learning_rate": 0.0002874438908403941,
+      "loss": 0.7343,
+      "step": 1070
+    },
+    {
+      "epoch": 0.38,
+      "learning_rate": 0.00028723646348295645,
+      "loss": 0.3369,
+      "step": 1071
+    },
+    {
+      "epoch": 0.38,
+      "learning_rate": 0.00028702892017712546,
+      "loss": 0.4379,
+      "step": 1072
+    },
+    {
+      "epoch": 0.38,
+      "learning_rate": 0.00028682126119875266,
+      "loss": 0.3205,
+      "step": 1073
+    },
+    {
+      "epoch": 0.38,
+      "learning_rate": 0.00028661348682384324,
+      "loss": 0.3,
+      "step": 1074
+    },
+    {
+      "epoch": 0.38,
+      "learning_rate": 0.0002864055973285559,
+      "loss": 0.4403,
+      "step": 1075
+    },
+    {
+      "epoch": 0.38,
+      "learning_rate": 0.00028619759298920247,
+      "loss": 0.5668,
+      "step": 1076
+    },
+    {
+      "epoch": 0.38,
+      "learning_rate": 0.0002859894740822471,
+      "loss": 0.6295,
+      "step": 1077
+    },
+    {
+      "epoch": 0.38,
+      "learning_rate": 0.00028578124088430635,
+      "loss": 0.697,
+      "step": 1078
+    },
+    {
+      "epoch": 0.38,
+      "learning_rate": 0.0002855728936721488,
+      "loss": 0.396,
+      "step": 1079
+    },
+    {
+      "epoch": 0.38,
+      "learning_rate": 0.00028536443272269446,
+      "loss": 0.5336,
+      "step": 1080
+    },
+    {
+      "epoch": 0.38,
+      "eval_loss": 0.574394941329956,
+      "eval_runtime": 52.9407,
+      "eval_samples_per_second": 30.77,
+      "eval_steps_per_second": 30.77,
+      "step": 1080
+    },
+    {
+      "epoch": 0.38,
+      "learning_rate": 0.00028515585831301456,
+      "loss": 0.6114,
+      "step": 1081
+    },
+    {
+      "epoch": 0.38,
+      "learning_rate": 0.00028494717072033107,
+      "loss": 0.5408,
+      "step": 1082
+    },
+    {
+      "epoch": 0.38,
+      "learning_rate": 0.00028473837022201644,
+      "loss": 0.6635,
+      "step": 1083
+    },
+    {
+      "epoch": 0.38,
+      "learning_rate": 0.00028452945709559325,
+      "loss": 0.4986,
+      "step": 1084
+    },
+    {
+      "epoch": 0.38,
+      "learning_rate": 0.0002843204316187336,
+      "loss": 0.884,
+      "step": 1085
+    },
+    {
+      "epoch": 0.38,
+      "learning_rate": 0.0002841112940692591,
+      "loss": 0.4564,
+      "step": 1086
+    },
+    {
+      "epoch": 0.38,
+      "learning_rate": 0.0002839020447251402,
+      "loss": 0.5174,
+      "step": 1087
+    },
+    {
+      "epoch": 0.39,
+      "learning_rate": 0.000283692683864496,
+      "loss": 0.7432,
+      "step": 1088
+    },
+    {
+      "epoch": 0.39,
+      "learning_rate": 0.00028348321176559384,
+      "loss": 0.3276,
+      "step": 1089
+    },
+    {
+      "epoch": 0.39,
+      "learning_rate": 0.0002832736287068488,
+      "loss": 0.3395,
+      "step": 1090
+    },
+    {
+      "epoch": 0.39,
+      "learning_rate": 0.00028306393496682354,
+      "loss": 0.5683,
+      "step": 1091
+    },
+    {
+      "epoch": 0.39,
+      "learning_rate": 0.00028285413082422783,
+      "loss": 0.5404,
+      "step": 1092
+    },
+    {
+      "epoch": 0.39,
+      "learning_rate": 0.00028264421655791816,
+      "loss": 0.492,
+      "step": 1093
+    },
+    {
+      "epoch": 0.39,
+      "learning_rate": 0.0002824341924468974,
+      "loss": 0.4472,
+      "step": 1094
+    },
+    {
+      "epoch": 0.39,
+      "learning_rate": 0.0002822240587703142,
+      "loss": 0.5723,
+      "step": 1095
+    },
+    {
+      "epoch": 0.39,
+      "learning_rate": 0.00028201381580746337,
+      "loss": 0.6974,
+      "step": 1096
+    },
+    {
+      "epoch": 0.39,
+      "learning_rate": 0.00028180346383778433,
+      "loss": 0.5703,
+      "step": 1097
+    },
+    {
+      "epoch": 0.39,
+      "learning_rate": 0.0002815930031408619,
+      "loss": 0.4147,
+      "step": 1098
+    },
+    {
+      "epoch": 0.39,
+      "learning_rate": 0.00028138243399642506,
+      "loss": 0.6062,
+      "step": 1099
+    },
+    {
+      "epoch": 0.39,
+      "learning_rate": 0.0002811717566843471,
+      "loss": 0.5509,
+      "step": 1100
+    },
+    {
+      "epoch": 0.39,
+      "eval_loss": 0.5854412913322449,
+      "eval_runtime": 52.9399,
+      "eval_samples_per_second": 30.771,
+      "eval_steps_per_second": 30.771,
+      "step": 1100
+    },
+    {
+      "epoch": 0.39,
+      "learning_rate": 0.0002809609714846452,
+      "loss": 0.4389,
+      "step": 1101
+    },
+    {
+      "epoch": 0.39,
+      "learning_rate": 0.00028075007867747965,
+      "loss": 0.5572,
+      "step": 1102
+    },
+    {
+      "epoch": 0.39,
+      "learning_rate": 0.000280539078543154,
+      "loss": 0.5082,
+      "step": 1103
+    },
+    {
+      "epoch": 0.39,
+      "learning_rate": 0.0002803279713621142,
+      "loss": 0.4413,
+      "step": 1104
+    },
+    {
+      "epoch": 0.39,
+      "learning_rate": 0.00028011675741494875,
+      "loss": 0.6788,
+      "step": 1105
+    },
+    {
+      "epoch": 0.39,
+      "learning_rate": 0.00027990543698238804,
+      "loss": 0.5179,
+      "step": 1106
+    },
+    {
+      "epoch": 0.39,
+      "learning_rate": 0.00027969401034530364,
+      "loss": 0.6231,
+      "step": 1107
+    },
+    {
+      "epoch": 0.39,
+      "learning_rate": 0.0002794824777847088,
+      "loss": 0.5877,
+      "step": 1108
+    },
+    {
+      "epoch": 0.39,
+      "learning_rate": 0.0002792708395817571,
+      "loss": 0.3782,
+      "step": 1109
+    },
+    {
+      "epoch": 0.39,
+      "learning_rate": 0.00027905909601774276,
+      "loss": 0.3692,
+      "step": 1110
+    },
+    {
+      "epoch": 0.39,
+      "learning_rate": 0.00027884724737410014,
+      "loss": 0.771,
+      "step": 1111
+    },
+    {
+      "epoch": 0.39,
+      "learning_rate": 0.0002786352939324029,
+      "loss": 0.5384,
+      "step": 1112
+    },
+    {
+      "epoch": 0.39,
+      "learning_rate": 0.0002784232359743643,
+      "loss": 0.2958,
+      "step": 1113
+    },
+    {
+      "epoch": 0.39,
+      "learning_rate": 0.0002782110737818366,
+      "loss": 0.6074,
+      "step": 1114
+    },
+    {
+      "epoch": 0.39,
+      "learning_rate": 0.0002779988076368102,
+      "loss": 0.707,
+      "step": 1115
+    },
+    {
+      "epoch": 0.39,
+      "learning_rate": 0.00027778643782141397,
+      "loss": 0.3925,
+      "step": 1116
+    },
+    {
+      "epoch": 0.4,
+      "learning_rate": 0.0002775739646179145,
+      "loss": 0.8481,
+      "step": 1117
+    },
+    {
+      "epoch": 0.4,
+      "learning_rate": 0.00027736138830871586,
+      "loss": 0.5856,
+      "step": 1118
+    },
+    {
+      "epoch": 0.4,
+      "learning_rate": 0.00027714870917635903,
+      "loss": 0.5646,
+      "step": 1119
+    },
+    {
+      "epoch": 0.4,
+      "learning_rate": 0.00027693592750352165,
+      "loss": 0.2759,
+      "step": 1120
+    },
+    {
+      "epoch": 0.4,
+      "eval_loss": 0.5674733519554138,
+      "eval_runtime": 52.9903,
+      "eval_samples_per_second": 30.741,
+      "eval_steps_per_second": 30.741,
+      "step": 1120
+    },
+    {
+      "epoch": 0.4,
+      "learning_rate": 0.00027672304357301796,
+      "loss": 0.3807,
+      "step": 1121
+    },
+    {
+      "epoch": 0.4,
+      "learning_rate": 0.0002765100576677976,
+      "loss": 0.4743,
+      "step": 1122
+    },
+    {
+      "epoch": 0.4,
+      "learning_rate": 0.00027629697007094614,
+      "loss": 0.5121,
+      "step": 1123
+    },
+    {
+      "epoch": 0.4,
+      "learning_rate": 0.00027608378106568424,
+      "loss": 0.4256,
+      "step": 1124
+    },
+    {
+      "epoch": 0.4,
+      "learning_rate": 0.0002758704909353671,
+      "loss": 0.4684,
+      "step": 1125
+    },
+    {
+      "epoch": 0.4,
+      "learning_rate": 0.0002756570999634848,
+      "loss": 0.4215,
+      "step": 1126
+    },
+    {
+      "epoch": 0.4,
+      "learning_rate": 0.000275443608433661,
+      "loss": 0.4497,
+      "step": 1127
+    },
+    {
+      "epoch": 0.4,
+      "learning_rate": 0.0002752300166296532,
+      "loss": 0.4526,
+      "step": 1128
+    },
+    {
+      "epoch": 0.4,
+      "learning_rate": 0.00027501632483535224,
+      "loss": 0.4336,
+      "step": 1129
+    },
+    {
+      "epoch": 0.4,
+      "learning_rate": 0.0002748025333347817,
+      "loss": 0.4733,
+      "step": 1130
+    },
+    {
+      "epoch": 0.4,
+      "learning_rate": 0.0002745886424120979,
+      "loss": 0.7019,
+      "step": 1131
+    },
+    {
+      "epoch": 0.4,
+      "learning_rate": 0.00027437465235158895,
+      "loss": 0.7965,
+      "step": 1132
+    },
+    {
+      "epoch": 0.4,
+      "learning_rate": 0.00027416056343767517,
+      "loss": 0.3885,
+      "step": 1133
+    },
+    {
+      "epoch": 0.4,
+      "learning_rate": 0.00027394637595490795,
+      "loss": 0.4072,
+      "step": 1134
+    },
+    {
+      "epoch": 0.4,
+      "learning_rate": 0.0002737320901879698,
+      "loss": 0.4354,
+      "step": 1135
+    },
+    {
+      "epoch": 0.4,
+      "learning_rate": 0.00027351770642167385,
+      "loss": 0.7217,
+      "step": 1136
+    },
+    {
+      "epoch": 0.4,
+      "learning_rate": 0.00027330322494096345,
+      "loss": 0.4505,
+      "step": 1137
+    },
+    {
+      "epoch": 0.4,
+      "learning_rate": 0.000273088646030912,
+      "loss": 0.4395,
+      "step": 1138
+    },
+    {
+      "epoch": 0.4,
+      "learning_rate": 0.00027287396997672207,
+      "loss": 0.5214,
+      "step": 1139
+    },
+    {
+      "epoch": 0.4,
+      "learning_rate": 0.0002726591970637256,
+      "loss": 0.5058,
+      "step": 1140
+    },
+    {
+      "epoch": 0.4,
+      "eval_loss": 0.5789819359779358,
+      "eval_runtime": 52.9438,
+      "eval_samples_per_second": 30.768,
+      "eval_steps_per_second": 30.768,
+      "step": 1140
+    },
+    {
+      "epoch": 0.4,
+      "learning_rate": 0.00027244432757738335,
+      "loss": 0.6211,
+      "step": 1141
+    },
+    {
+      "epoch": 0.4,
+      "learning_rate": 0.0002722293618032841,
+      "loss": 0.9251,
+      "step": 1142
+    },
+    {
+      "epoch": 0.4,
+      "learning_rate": 0.0002720143000271448,
+      "loss": 0.5805,
+      "step": 1143
+    },
+    {
+      "epoch": 0.4,
+      "learning_rate": 0.00027179914253481016,
+      "loss": 0.4064,
+      "step": 1144
+    },
+    {
+      "epoch": 0.41,
+      "learning_rate": 0.0002715838896122518,
+      "loss": 0.361,
+      "step": 1145
+    },
+    {
+      "epoch": 0.41,
+      "learning_rate": 0.0002713685415455684,
+      "loss": 0.4251,
+      "step": 1146
+    },
+    {
+      "epoch": 0.41,
+      "learning_rate": 0.0002711530986209849,
+      "loss": 0.6008,
+      "step": 1147
+    },
+    {
+      "epoch": 0.41,
+      "learning_rate": 0.00027093756112485265,
+      "loss": 0.5226,
+      "step": 1148
+    },
+    {
+      "epoch": 0.41,
+      "learning_rate": 0.00027072192934364834,
+      "loss": 0.768,
+      "step": 1149
+    },
+    {
+      "epoch": 0.41,
+      "learning_rate": 0.0002705062035639741,
+      "loss": 0.5068,
+      "step": 1150
+    },
+    {
+      "epoch": 0.41,
+      "learning_rate": 0.00027029038407255725,
+      "loss": 0.6461,
+      "step": 1151
+    },
+    {
+      "epoch": 0.41,
+      "learning_rate": 0.0002700744711562491,
+      "loss": 0.8946,
+      "step": 1152
+    },
+    {
+      "epoch": 0.41,
+      "learning_rate": 0.0002698584651020258,
+      "loss": 0.6626,
+      "step": 1153
+    },
+    {
+      "epoch": 0.41,
+      "learning_rate": 0.0002696423661969867,
+      "loss": 0.4081,
+      "step": 1154
+    },
+    {
+      "epoch": 0.41,
+      "learning_rate": 0.00026942617472835496,
+      "loss": 0.4145,
+      "step": 1155
+    },
+    {
+      "epoch": 0.41,
+      "learning_rate": 0.00026920989098347663,
+      "loss": 0.5784,
+      "step": 1156
+    },
+    {
+      "epoch": 0.41,
+      "learning_rate": 0.0002689935152498203,
+      "loss": 0.6766,
+      "step": 1157
+    },
+    {
+      "epoch": 0.41,
+      "learning_rate": 0.0002687770478149772,
+      "loss": 0.3437,
+      "step": 1158
+    },
+    {
+      "epoch": 0.41,
+      "learning_rate": 0.00026856048896665985,
+      "loss": 0.6037,
+      "step": 1159
+    },
+    {
+      "epoch": 0.41,
+      "learning_rate": 0.0002683438389927027,
+      "loss": 0.4446,
+      "step": 1160
+    },
+    {
+      "epoch": 0.41,
+      "eval_loss": 0.5892747044563293,
+      "eval_runtime": 52.9403,
+      "eval_samples_per_second": 30.771,
+      "eval_steps_per_second": 30.771,
+      "step": 1160
+    },
+    {
+      "epoch": 0.41,
+      "learning_rate": 0.00026812709818106143,
+      "loss": 0.4562,
+      "step": 1161
+    },
+    {
+      "epoch": 0.41,
+      "learning_rate": 0.00026791026681981196,
+      "loss": 0.7074,
+      "step": 1162
+    },
+    {
+      "epoch": 0.41,
+      "learning_rate": 0.00026769334519715103,
+      "loss": 0.4269,
+      "step": 1163
+    },
+    {
+      "epoch": 0.41,
+      "learning_rate": 0.000267476333601395,
+      "loss": 0.4477,
+      "step": 1164
+    },
+    {
+      "epoch": 0.41,
+      "learning_rate": 0.0002672592323209801,
+      "loss": 0.4074,
+      "step": 1165
+    },
+    {
+      "epoch": 0.41,
+      "learning_rate": 0.0002670420416444616,
+      "loss": 0.5448,
+      "step": 1166
+    },
+    {
+      "epoch": 0.41,
+      "learning_rate": 0.00026682476186051357,
+      "loss": 0.3901,
+      "step": 1167
+    },
+    {
+      "epoch": 0.41,
+      "learning_rate": 0.0002666073932579287,
+      "loss": 0.4501,
+      "step": 1168
+    },
+    {
+      "epoch": 0.41,
+      "learning_rate": 0.0002663899361256175,
+      "loss": 0.5224,
+      "step": 1169
+    },
+    {
+      "epoch": 0.41,
+      "learning_rate": 0.0002661723907526083,
+      "loss": 0.5364,
+      "step": 1170
+    },
+    {
+      "epoch": 0.41,
+      "learning_rate": 0.00026595475742804666,
+      "loss": 0.4335,
+      "step": 1171
+    },
+    {
+      "epoch": 0.41,
+      "learning_rate": 0.000265737036441195,
+      "loss": 0.9272,
+      "step": 1172
+    },
+    {
+      "epoch": 0.42,
+      "learning_rate": 0.0002655192280814324,
+      "loss": 0.398,
+      "step": 1173
+    },
+    {
+      "epoch": 0.42,
+      "learning_rate": 0.0002653013326382538,
+      "loss": 0.4426,
+      "step": 1174
+    },
+    {
+      "epoch": 0.42,
+      "learning_rate": 0.00026508335040127017,
+      "loss": 0.498,
+      "step": 1175
+    },
+    {
+      "epoch": 0.42,
+      "learning_rate": 0.0002648652816602078,
+      "loss": 0.398,
+      "step": 1176
+    },
+    {
+      "epoch": 0.42,
+      "learning_rate": 0.00026464712670490765,
+      "loss": 0.6886,
+      "step": 1177
+    },
+    {
+      "epoch": 0.42,
+      "learning_rate": 0.0002644288858253257,
+      "loss": 0.4945,
+      "step": 1178
+    },
+    {
+      "epoch": 0.42,
+      "learning_rate": 0.00026421055931153184,
+      "loss": 0.4028,
+      "step": 1179
+    },
+    {
+      "epoch": 0.42,
+      "learning_rate": 0.00026399214745370994,
+      "loss": 0.4757,
+      "step": 1180
+    },
+    {
+      "epoch": 0.42,
+      "eval_loss": 0.5763522982597351,
+      "eval_runtime": 52.9389,
+      "eval_samples_per_second": 30.771,
+      "eval_steps_per_second": 30.771,
+      "step": 1180
+    },
+    {
+      "epoch": 0.42,
+      "learning_rate": 0.0002637736505421571,
+      "loss": 0.4496,
+      "step": 1181
+    },
+    {
+      "epoch": 0.42,
+      "learning_rate": 0.00026355506886728374,
+      "loss": 0.522,
+      "step": 1182
+    },
+    {
+      "epoch": 0.42,
+      "learning_rate": 0.0002633364027196128,
+      "loss": 0.4787,
+      "step": 1183
+    },
+    {
+      "epoch": 0.42,
+      "learning_rate": 0.00026311765238977935,
+      "loss": 0.4387,
+      "step": 1184
+    },
+    {
+      "epoch": 0.42,
+      "learning_rate": 0.0002628988181685306,
+      "loss": 0.4766,
+      "step": 1185
+    },
+    {
+      "epoch": 0.42,
+      "learning_rate": 0.00026267990034672516,
+      "loss": 0.3896,
+      "step": 1186
+    },
+    {
+      "epoch": 0.42,
+      "learning_rate": 0.00026246089921533273,
+      "loss": 0.5358,
+      "step": 1187
+    },
+    {
+      "epoch": 0.42,
+      "learning_rate": 0.00026224181506543385,
+      "loss": 0.4327,
+      "step": 1188
+    },
+    {
+      "epoch": 0.42,
+      "learning_rate": 0.00026202264818821913,
+      "loss": 0.4354,
+      "step": 1189
+    },
+    {
+      "epoch": 0.42,
+      "learning_rate": 0.00026180339887498953,
+      "loss": 0.8578,
+      "step": 1190
+    },
+    {
+      "epoch": 0.42,
+      "learning_rate": 0.00026158406741715517,
+      "loss": 0.3804,
+      "step": 1191
+    },
+    {
+      "epoch": 0.42,
+      "learning_rate": 0.0002613646541062357,
+      "loss": 0.45,
+      "step": 1192
+    },
+    {
+      "epoch": 0.42,
+      "learning_rate": 0.0002611451592338594,
+      "loss": 0.4807,
+      "step": 1193
+    },
+    {
+      "epoch": 0.42,
+      "learning_rate": 0.00026092558309176293,
+      "loss": 0.7022,
+      "step": 1194
+    },
+    {
+      "epoch": 0.42,
+      "learning_rate": 0.00026070592597179096,
+      "loss": 0.4402,
+      "step": 1195
+    },
+    {
+      "epoch": 0.42,
+      "learning_rate": 0.0002604861881658959,
+      "loss": 0.9842,
+      "step": 1196
+    },
+    {
+      "epoch": 0.42,
+      "learning_rate": 0.00026026636996613725,
+      "loss": 0.4801,
+      "step": 1197
+    },
+    {
+      "epoch": 0.42,
+      "learning_rate": 0.00026004647166468165,
+      "loss": 0.3027,
+      "step": 1198
+    },
+    {
+      "epoch": 0.42,
+      "learning_rate": 0.0002598264935538018,
+      "loss": 0.3352,
+      "step": 1199
+    },
+    {
+      "epoch": 0.42,
+      "learning_rate": 0.00025960643592587673,
+      "loss": 0.4153,
+      "step": 1200
+    },
+    {
+      "epoch": 0.42,
+      "eval_loss": 0.5706815123558044,
+      "eval_runtime": 52.9757,
+      "eval_samples_per_second": 30.75,
+      "eval_steps_per_second": 30.75,
+      "step": 1200
+    },
+    {
+      "epoch": 0.43,
+      "learning_rate": 0.00025938629907339123,
+      "loss": 0.4167,
+      "step": 1201
+    },
+    {
+      "epoch": 0.43,
+      "learning_rate": 0.0002591660832889351,
+      "loss": 0.5684,
+      "step": 1202
+    },
+    {
+      "epoch": 0.43,
+      "learning_rate": 0.0002589457888652034,
+      "loss": 0.5264,
+      "step": 1203
+    },
+    {
+      "epoch": 0.43,
+      "learning_rate": 0.0002587254160949954,
+      "loss": 0.3073,
+      "step": 1204
+    },
+    {
+      "epoch": 0.43,
+      "learning_rate": 0.00025850496527121463,
+      "loss": 0.554,
+      "step": 1205
+    },
+    {
+      "epoch": 0.43,
+      "learning_rate": 0.0002582844366868685,
+      "loss": 0.3932,
+      "step": 1206
+    },
+    {
+      "epoch": 0.43,
+      "learning_rate": 0.00025806383063506745,
+      "loss": 0.4984,
+      "step": 1207
+    },
+    {
+      "epoch": 0.43,
+      "learning_rate": 0.00025784314740902526,
+      "loss": 0.5486,
+      "step": 1208
+    },
+    {
+      "epoch": 0.43,
+      "learning_rate": 0.000257622387302058,
+      "loss": 0.3672,
+      "step": 1209
+    },
+    {
+      "epoch": 0.43,
+      "learning_rate": 0.0002574015506075841,
+      "loss": 0.3456,
+      "step": 1210
+    },
+    {
+      "epoch": 0.43,
+      "learning_rate": 0.0002571806376191236,
+      "loss": 0.6226,
+      "step": 1211
+    },
+    {
+      "epoch": 0.43,
+      "learning_rate": 0.0002569596486302982,
+      "loss": 0.6254,
+      "step": 1212
+    },
+    {
+      "epoch": 0.43,
+      "learning_rate": 0.00025673858393483035,
+      "loss": 0.3546,
+      "step": 1213
+    },
+    {
+      "epoch": 0.43,
+      "learning_rate": 0.0002565174438265433,
+      "loss": 0.4516,
+      "step": 1214
+    },
+    {
+      "epoch": 0.43,
+      "learning_rate": 0.0002562962285993604,
+      "loss": 0.6422,
+      "step": 1215
+    },
+    {
+      "epoch": 0.43,
+      "learning_rate": 0.00025607493854730505,
+      "loss": 0.635,
+      "step": 1216
+    },
+    {
+      "epoch": 0.43,
+      "learning_rate": 0.00025585357396449985,
+      "loss": 1.2689,
+      "step": 1217
+    },
+    {
+      "epoch": 0.43,
+      "learning_rate": 0.0002556321351451667,
+      "loss": 0.5264,
+      "step": 1218
+    },
+    {
+      "epoch": 0.43,
+      "learning_rate": 0.0002554106223836259,
+      "loss": 0.7739,
+      "step": 1219
+    },
+    {
+      "epoch": 0.43,
+      "learning_rate": 0.0002551890359742963,
+      "loss": 0.5369,
+      "step": 1220
+    },
+    {
+      "epoch": 0.43,
+      "eval_loss": 0.5729343891143799,
+      "eval_runtime": 52.9429,
+      "eval_samples_per_second": 30.769,
+      "eval_steps_per_second": 30.769,
+      "step": 1220
+    },
+    {
+      "epoch": 0.43,
+      "learning_rate": 0.0002549673762116945,
+      "loss": 0.2869,
+      "step": 1221
+    },
+    {
+      "epoch": 0.43,
+      "learning_rate": 0.00025474564339043464,
+      "loss": 0.3938,
+      "step": 1222
+    },
+    {
+      "epoch": 0.43,
+      "learning_rate": 0.00025452383780522785,
+      "loss": 0.4535,
+      "step": 1223
+    },
+    {
+      "epoch": 0.43,
+      "learning_rate": 0.00025430195975088203,
+      "loss": 0.4389,
+      "step": 1224
+    },
+    {
+      "epoch": 0.43,
+      "learning_rate": 0.0002540800095223016,
+      "loss": 0.3289,
+      "step": 1225
+    },
+    {
+      "epoch": 0.43,
+      "learning_rate": 0.00025385798741448657,
+      "loss": 0.5119,
+      "step": 1226
+    },
+    {
+      "epoch": 0.43,
+      "learning_rate": 0.0002536358937225328,
+      "loss": 1.2492,
+      "step": 1227
+    },
+    {
+      "epoch": 0.43,
+      "learning_rate": 0.0002534137287416311,
+      "loss": 0.3587,
+      "step": 1228
+    },
+    {
+      "epoch": 0.43,
+      "learning_rate": 0.00025319149276706695,
+      "loss": 0.3761,
+      "step": 1229
+    },
+    {
+      "epoch": 0.44,
+      "learning_rate": 0.0002529691860942205,
+      "loss": 0.6944,
+      "step": 1230
+    },
+    {
+      "epoch": 0.44,
+      "learning_rate": 0.00025274680901856566,
+      "loss": 0.4059,
+      "step": 1231
+    },
+    {
+      "epoch": 0.44,
+      "learning_rate": 0.0002525243618356699,
+      "loss": 0.4745,
+      "step": 1232
+    },
+    {
+      "epoch": 0.44,
+      "learning_rate": 0.00025230184484119393,
+      "loss": 0.4983,
+      "step": 1233
+    },
+    {
+      "epoch": 0.44,
+      "learning_rate": 0.0002520792583308912,
+      "loss": 0.4564,
+      "step": 1234
+    },
+    {
+      "epoch": 0.44,
+      "learning_rate": 0.0002518566026006076,
+      "loss": 0.4986,
+      "step": 1235
+    },
+    {
+      "epoch": 0.44,
+      "learning_rate": 0.0002516338779462811,
+      "loss": 0.4337,
+      "step": 1236
+    },
+    {
+      "epoch": 0.44,
+      "learning_rate": 0.00025141108466394106,
+      "loss": 0.3605,
+      "step": 1237
+    },
+    {
+      "epoch": 0.44,
+      "learning_rate": 0.00025118822304970827,
+      "loss": 1.0547,
+      "step": 1238
+    },
+    {
+      "epoch": 0.44,
+      "learning_rate": 0.0002509652933997942,
+      "loss": 0.3894,
+      "step": 1239
+    },
+    {
+      "epoch": 0.44,
+      "learning_rate": 0.0002507422960105009,
+      "loss": 0.4785,
+      "step": 1240
+    },
+    {
+      "epoch": 0.44,
+      "eval_loss": 0.5735045671463013,
+      "eval_runtime": 52.9466,
+      "eval_samples_per_second": 30.767,
+      "eval_steps_per_second": 30.767,
+      "step": 1240
+    },
+    {
+      "epoch": 0.44,
+      "learning_rate": 0.00025051923117822033,
+      "loss": 0.3762,
+      "step": 1241
+    },
+    {
+      "epoch": 0.44,
+      "learning_rate": 0.00025029609919943404,
+      "loss": 0.4859,
+      "step": 1242
+    },
+    {
+      "epoch": 0.44,
+      "learning_rate": 0.00025007290037071294,
+      "loss": 0.4397,
+      "step": 1243
+    },
+    {
+      "epoch": 0.44,
+      "learning_rate": 0.0002498496349887168,
+      "loss": 0.692,
+      "step": 1244
+    },
+    {
+      "epoch": 0.44,
+      "learning_rate": 0.00024962630335019384,
+      "loss": 0.6797,
+      "step": 1245
+    },
+    {
+      "epoch": 0.44,
+      "learning_rate": 0.00024940290575198017,
+      "loss": 0.3737,
+      "step": 1246
+    },
+    {
+      "epoch": 0.44,
+      "learning_rate": 0.0002491794424909998,
+      "loss": 0.3617,
+      "step": 1247
+    },
+    {
+      "epoch": 0.44,
+      "learning_rate": 0.000248955913864264,
+      "loss": 0.5009,
+      "step": 1248
+    },
+    {
+      "epoch": 0.44,
+      "learning_rate": 0.0002487323201688707,
+      "loss": 0.414,
+      "step": 1249
+    },
+    {
+      "epoch": 0.44,
+      "learning_rate": 0.0002485086617020045,
+      "loss": 0.5138,
+      "step": 1250
+    },
+    {
+      "epoch": 0.44,
+      "learning_rate": 0.00024828493876093605,
+      "loss": 0.5573,
+      "step": 1251
+    },
+    {
+      "epoch": 0.44,
+      "learning_rate": 0.00024806115164302176,
+      "loss": 0.8266,
+      "step": 1252
+    },
+    {
+      "epoch": 0.44,
+      "learning_rate": 0.0002478373006457032,
+      "loss": 0.2755,
+      "step": 1253
+    },
+    {
+      "epoch": 0.44,
+      "learning_rate": 0.0002476133860665069,
+      "loss": 0.6646,
+      "step": 1254
+    },
+    {
+      "epoch": 0.44,
+      "learning_rate": 0.0002473894082030439,
+      "loss": 0.5479,
+      "step": 1255
+    },
+    {
+      "epoch": 0.44,
+      "learning_rate": 0.00024716536735300943,
+      "loss": 0.387,
+      "step": 1256
+    },
+    {
+      "epoch": 0.44,
+      "learning_rate": 0.0002469412638141824,
+      "loss": 0.531,
+      "step": 1257
+    },
+    {
+      "epoch": 0.45,
+      "learning_rate": 0.00024671709788442493,
+      "loss": 0.514,
+      "step": 1258
+    },
+    {
+      "epoch": 0.45,
+      "learning_rate": 0.0002464928698616822,
+      "loss": 0.3646,
+      "step": 1259
+    },
+    {
+      "epoch": 0.45,
+      "learning_rate": 0.00024626858004398177,
+      "loss": 0.4335,
+      "step": 1260
+    },
+    {
+      "epoch": 0.45,
+      "eval_loss": 0.582056999206543,
+      "eval_runtime": 52.9395,
+      "eval_samples_per_second": 30.771,
+      "eval_steps_per_second": 30.771,
+      "step": 1260
+    },
+    {
+      "epoch": 0.45,
+      "learning_rate": 0.0002460442287294336,
+      "loss": 0.6787,
+      "step": 1261
+    },
+    {
+      "epoch": 0.45,
+      "learning_rate": 0.00024581981621622914,
+      "loss": 0.7749,
+      "step": 1262
+    },
+    {
+      "epoch": 0.45,
+      "learning_rate": 0.0002455953428026412,
+      "loss": 0.7386,
+      "step": 1263
+    },
+    {
+      "epoch": 0.45,
+      "learning_rate": 0.00024537080878702365,
+      "loss": 0.9756,
+      "step": 1264
+    },
+    {
+      "epoch": 0.45,
+      "learning_rate": 0.0002451462144678108,
+      "loss": 0.4038,
+      "step": 1265
+    },
+    {
+      "epoch": 0.45,
+      "learning_rate": 0.0002449215601435173,
+      "loss": 0.4417,
+      "step": 1266
+    },
+    {
+      "epoch": 0.45,
+      "learning_rate": 0.0002446968461127373,
+      "loss": 0.4071,
+      "step": 1267
+    },
+    {
+      "epoch": 0.45,
+      "learning_rate": 0.00024447207267414445,
+      "loss": 0.3004,
+      "step": 1268
+    },
+    {
+      "epoch": 0.45,
+      "learning_rate": 0.0002442472401264914,
+      "loss": 0.6453,
+      "step": 1269
+    },
+    {
+      "epoch": 0.45,
+      "learning_rate": 0.00024402234876860912,
+      "loss": 0.4884,
+      "step": 1270
+    },
+    {
+      "epoch": 0.45,
+      "learning_rate": 0.00024379739889940713,
+      "loss": 0.6829,
+      "step": 1271
+    },
+    {
+      "epoch": 0.45,
+      "learning_rate": 0.00024357239081787243,
+      "loss": 0.5509,
+      "step": 1272
+    },
+    {
+      "epoch": 0.45,
+      "learning_rate": 0.0002433473248230695,
+      "loss": 0.4143,
+      "step": 1273
+    },
+    {
+      "epoch": 0.45,
+      "learning_rate": 0.00024312220121413975,
+      "loss": 0.3998,
+      "step": 1274
+    },
+    {
+      "epoch": 0.45,
+      "learning_rate": 0.0002428970202903011,
+      "loss": 0.7329,
+      "step": 1275
+    },
+    {
+      "epoch": 0.45,
+      "learning_rate": 0.0002426717823508479,
+      "loss": 0.4964,
+      "step": 1276
+    },
+    {
+      "epoch": 0.45,
+      "learning_rate": 0.00024244648769515,
+      "loss": 0.4965,
+      "step": 1277
+    },
+    {
+      "epoch": 0.45,
+      "learning_rate": 0.00024222113662265282,
+      "loss": 0.4078,
+      "step": 1278
+    },
+    {
+      "epoch": 0.45,
+      "learning_rate": 0.0002419957294328767,
+      "loss": 0.4518,
+      "step": 1279
+    },
+    {
+      "epoch": 0.45,
+      "learning_rate": 0.00024177026642541643,
+      "loss": 0.5452,
+      "step": 1280
+    },
+    {
+      "epoch": 0.45,
+      "eval_loss": 0.5620944499969482,
+      "eval_runtime": 52.9439,
+      "eval_samples_per_second": 30.768,
+      "eval_steps_per_second": 30.768,
+      "step": 1280
+    },
+    {
+      "epoch": 0.45,
+      "learning_rate": 0.0002415447478999413,
+      "loss": 0.2885,
+      "step": 1281
+    },
+    {
+      "epoch": 0.45,
+      "learning_rate": 0.00024131917415619412,
+      "loss": 0.4729,
+      "step": 1282
+    },
+    {
+      "epoch": 0.45,
+      "learning_rate": 0.00024109354549399118,
+      "loss": 0.8796,
+      "step": 1283
+    },
+    {
+      "epoch": 0.45,
+      "learning_rate": 0.00024086786221322185,
+      "loss": 0.4021,
+      "step": 1284
+    },
+    {
+      "epoch": 0.45,
+      "learning_rate": 0.00024064212461384786,
+      "loss": 0.3739,
+      "step": 1285
+    },
+    {
+      "epoch": 0.46,
+      "learning_rate": 0.0002404163329959034,
+      "loss": 0.2599,
+      "step": 1286
+    },
+    {
+      "epoch": 0.46,
+      "learning_rate": 0.00024019048765949435,
+      "loss": 1.0206,
+      "step": 1287
+    },
+    {
+      "epoch": 0.46,
+      "learning_rate": 0.00023996458890479793,
+      "loss": 0.5124,
+      "step": 1288
+    },
+    {
+      "epoch": 0.46,
+      "learning_rate": 0.00023973863703206246,
+      "loss": 0.5078,
+      "step": 1289
+    },
+    {
+      "epoch": 0.46,
+      "learning_rate": 0.00023951263234160678,
+      "loss": 0.3403,
+      "step": 1290
+    },
+    {
+      "epoch": 0.46,
+      "learning_rate": 0.00023928657513381998,
+      "loss": 0.6973,
+      "step": 1291
+    },
+    {
+      "epoch": 0.46,
+      "learning_rate": 0.00023906046570916094,
+      "loss": 0.4932,
+      "step": 1292
+    },
+    {
+      "epoch": 0.46,
+      "learning_rate": 0.00023883430436815798,
+      "loss": 0.4725,
+      "step": 1293
+    },
+    {
+      "epoch": 0.46,
+      "learning_rate": 0.00023860809141140835,
+      "loss": 0.6949,
+      "step": 1294
+    },
+    {
+      "epoch": 0.46,
+      "learning_rate": 0.00023838182713957785,
+      "loss": 0.4261,
+      "step": 1295
+    },
+    {
+      "epoch": 0.46,
+      "learning_rate": 0.00023815551185340082,
+      "loss": 0.3906,
+      "step": 1296
+    },
+    {
+      "epoch": 0.46,
+      "learning_rate": 0.00023792914585367888,
+      "loss": 0.5062,
+      "step": 1297
+    },
+    {
+      "epoch": 0.46,
+      "learning_rate": 0.0002377027294412815,
+      "loss": 0.4241,
+      "step": 1298
+    },
+    {
+      "epoch": 0.46,
+      "learning_rate": 0.00023747626291714498,
+      "loss": 0.8483,
+      "step": 1299
+    },
+    {
+      "epoch": 0.46,
+      "learning_rate": 0.0002372497465822721,
+      "loss": 0.3461,
+      "step": 1300
+    },
+    {
+      "epoch": 0.46,
+      "eval_loss": 0.5614920854568481,
+      "eval_runtime": 52.9419,
+      "eval_samples_per_second": 30.77,
+      "eval_steps_per_second": 30.77,
+      "step": 1300
+    },
+    {
+      "epoch": 0.46,
+      "learning_rate": 0.0002370231807377322,
+      "loss": 1.0309,
+      "step": 1301
+    },
+    {
+      "epoch": 0.46,
+      "learning_rate": 0.00023679656568465998,
+      "loss": 0.3814,
+      "step": 1302
+    },
+    {
+      "epoch": 0.46,
+      "learning_rate": 0.00023656990172425587,
+      "loss": 0.9046,
+      "step": 1303
+    },
+    {
+      "epoch": 0.46,
+      "learning_rate": 0.00023634318915778527,
+      "loss": 0.4657,
+      "step": 1304
+    },
+    {
+      "epoch": 0.46,
+      "learning_rate": 0.00023611642828657794,
+      "loss": 0.5943,
+      "step": 1305
+    },
+    {
+      "epoch": 0.46,
+      "learning_rate": 0.00023588961941202824,
+      "loss": 0.3648,
+      "step": 1306
+    },
+    {
+      "epoch": 0.46,
+      "learning_rate": 0.0002356627628355939,
+      "loss": 0.6308,
+      "step": 1307
+    },
+    {
+      "epoch": 0.46,
+      "learning_rate": 0.00023543585885879638,
+      "loss": 0.5088,
+      "step": 1308
+    },
+    {
+      "epoch": 0.46,
+      "learning_rate": 0.00023520890778322003,
+      "loss": 0.4637,
+      "step": 1309
+    },
+    {
+      "epoch": 0.46,
+      "learning_rate": 0.00023498190991051168,
+      "loss": 0.7995,
+      "step": 1310
+    },
+    {
+      "epoch": 0.46,
+      "learning_rate": 0.00023475486554238063,
+      "loss": 0.5402,
+      "step": 1311
+    },
+    {
+      "epoch": 0.46,
+      "learning_rate": 0.00023452777498059768,
+      "loss": 0.3724,
+      "step": 1312
+    },
+    {
+      "epoch": 0.46,
+      "learning_rate": 0.00023430063852699525,
+      "loss": 0.4261,
+      "step": 1313
+    },
+    {
+      "epoch": 0.47,
+      "learning_rate": 0.00023407345648346665,
+      "loss": 0.298,
+      "step": 1314
+    },
+    {
+      "epoch": 0.47,
+      "learning_rate": 0.0002338462291519658,
+      "loss": 0.8048,
+      "step": 1315
+    },
+    {
+      "epoch": 0.47,
+      "learning_rate": 0.00023361895683450694,
+      "loss": 0.6901,
+      "step": 1316
+    },
+    {
+      "epoch": 0.47,
+      "learning_rate": 0.00023339163983316375,
+      "loss": 0.4686,
+      "step": 1317
+    },
+    {
+      "epoch": 0.47,
+      "learning_rate": 0.00023316427845006973,
+      "loss": 0.3343,
+      "step": 1318
+    },
+    {
+      "epoch": 0.47,
+      "learning_rate": 0.00023293687298741716,
+      "loss": 0.6008,
+      "step": 1319
+    },
+    {
+      "epoch": 0.47,
+      "learning_rate": 0.0002327094237474568,
+      "loss": 0.5579,
+      "step": 1320
+    },
+    {
+      "epoch": 0.47,
+      "eval_loss": 0.5768979787826538,
+      "eval_runtime": 52.951,
+      "eval_samples_per_second": 30.764,
+      "eval_steps_per_second": 30.764,
+      "step": 1320
+    },
+    {
+      "epoch": 0.47,
+      "learning_rate": 0.00023248193103249785,
+      "loss": 0.465,
+      "step": 1321
+    },
+    {
+      "epoch": 0.47,
+      "learning_rate": 0.00023225439514490705,
+      "loss": 0.5398,
+      "step": 1322
+    },
+    {
+      "epoch": 0.47,
+      "learning_rate": 0.00023202681638710874,
+      "loss": 0.6041,
+      "step": 1323
+    },
+    {
+      "epoch": 0.47,
+      "learning_rate": 0.00023179919506158405,
+      "loss": 0.4252,
+      "step": 1324
+    },
+    {
+      "epoch": 0.47,
+      "learning_rate": 0.0002315715314708708,
+      "loss": 0.836,
+      "step": 1325
+    },
+    {
+      "epoch": 0.47,
+      "learning_rate": 0.000231343825917563,
+      "loss": 0.7176,
+      "step": 1326
+    },
+    {
+      "epoch": 0.47,
+      "learning_rate": 0.00023111607870431034,
+      "loss": 0.4449,
+      "step": 1327
+    },
+    {
+      "epoch": 0.47,
+      "learning_rate": 0.00023088829013381793,
+      "loss": 0.6199,
+      "step": 1328
+    },
+    {
+      "epoch": 0.47,
+      "learning_rate": 0.00023066046050884586,
+      "loss": 0.4793,
+      "step": 1329
+    },
+    {
+      "epoch": 0.47,
+      "learning_rate": 0.00023043259013220876,
+      "loss": 0.5359,
+      "step": 1330
+    },
+    {
+      "epoch": 0.47,
+      "learning_rate": 0.0002302046793067756,
+      "loss": 0.3961,
+      "step": 1331
+    },
+    {
+      "epoch": 0.47,
+      "learning_rate": 0.00022997672833546866,
+      "loss": 0.3897,
+      "step": 1332
+    },
+    {
+      "epoch": 0.47,
+      "learning_rate": 0.00022974873752126408,
+      "loss": 0.6098,
+      "step": 1333
+    },
+    {
+      "epoch": 0.47,
+      "learning_rate": 0.00022952070716719076,
+      "loss": 0.6298,
+      "step": 1334
+    },
+    {
+      "epoch": 0.47,
+      "learning_rate": 0.00022929263757633003,
+      "loss": 0.3884,
+      "step": 1335
+    },
+    {
+      "epoch": 0.47,
+      "learning_rate": 0.00022906452905181558,
+      "loss": 0.4333,
+      "step": 1336
+    },
+    {
+      "epoch": 0.47,
+      "learning_rate": 0.0002288363818968326,
+      "loss": 0.328,
+      "step": 1337
+    },
+    {
+      "epoch": 0.47,
+      "learning_rate": 0.000228608196414618,
+      "loss": 0.3453,
+      "step": 1338
+    },
+    {
+      "epoch": 0.47,
+      "learning_rate": 0.00022837997290845924,
+      "loss": 0.4327,
+      "step": 1339
+    },
+    {
+      "epoch": 0.47,
+      "learning_rate": 0.0002281517116816945,
+      "loss": 0.6048,
+      "step": 1340
+    },
+    {
+      "epoch": 0.47,
+      "eval_loss": 0.585534930229187,
+      "eval_runtime": 52.9452,
+      "eval_samples_per_second": 30.768,
+      "eval_steps_per_second": 30.768,
+      "step": 1340
+    },
+    {
+      "epoch": 0.47,
+      "learning_rate": 0.00022792341303771223,
+      "loss": 0.7095,
+      "step": 1341
+    },
+    {
+      "epoch": 0.47,
+      "learning_rate": 0.0002276950772799503,
+      "loss": 0.4545,
+      "step": 1342
+    },
+    {
+      "epoch": 0.48,
+      "learning_rate": 0.00022746670471189619,
+      "loss": 0.4571,
+      "step": 1343
+    },
+    {
+      "epoch": 0.48,
+      "learning_rate": 0.000227238295637086,
+      "loss": 0.4263,
+      "step": 1344
+    },
+    {
+      "epoch": 0.48,
+      "learning_rate": 0.00022700985035910466,
+      "loss": 0.464,
+      "step": 1345
+    },
+    {
+      "epoch": 0.48,
+      "learning_rate": 0.00022678136918158527,
+      "loss": 0.4303,
+      "step": 1346
+    },
+    {
+      "epoch": 0.48,
+      "learning_rate": 0.00022655285240820815,
+      "loss": 0.4865,
+      "step": 1347
+    },
+    {
+      "epoch": 0.48,
+      "learning_rate": 0.00022632430034270153,
+      "loss": 0.5177,
+      "step": 1348
+    },
+    {
+      "epoch": 0.48,
+      "learning_rate": 0.00022609571328884007,
+      "loss": 0.4943,
+      "step": 1349
+    },
+    {
+      "epoch": 0.48,
+      "learning_rate": 0.00022586709155044527,
+      "loss": 0.5965,
+      "step": 1350
+    },
+    {
+      "epoch": 0.48,
+      "learning_rate": 0.00022563843543138456,
+      "loss": 0.5235,
+      "step": 1351
+    },
+    {
+      "epoch": 0.48,
+      "learning_rate": 0.00022540974523557104,
+      "loss": 0.7819,
+      "step": 1352
+    },
+    {
+      "epoch": 0.48,
+      "learning_rate": 0.00022518102126696334,
+      "loss": 0.3831,
+      "step": 1353
+    },
+    {
+      "epoch": 0.48,
+      "learning_rate": 0.0002249522638295646,
+      "loss": 0.4437,
+      "step": 1354
+    },
+    {
+      "epoch": 0.48,
+      "learning_rate": 0.00022472347322742277,
+      "loss": 0.5153,
+      "step": 1355
+    },
+    {
+      "epoch": 0.48,
+      "learning_rate": 0.00022449464976462974,
+      "loss": 0.8301,
+      "step": 1356
+    },
+    {
+      "epoch": 0.48,
+      "learning_rate": 0.000224265793745321,
+      "loss": 0.8167,
+      "step": 1357
+    },
+    {
+      "epoch": 0.48,
+      "learning_rate": 0.00022403690547367563,
+      "loss": 0.4618,
+      "step": 1358
+    },
+    {
+      "epoch": 0.48,
+      "learning_rate": 0.00022380798525391512,
+      "loss": 0.469,
+      "step": 1359
+    },
+    {
+      "epoch": 0.48,
+      "learning_rate": 0.0002235790333903038,
+      "loss": 0.6253,
+      "step": 1360
+    },
+    {
+      "epoch": 0.48,
+      "eval_loss": 0.5590018630027771,
+      "eval_runtime": 52.9391,
+      "eval_samples_per_second": 30.771,
+      "eval_steps_per_second": 30.771,
+      "step": 1360
+    },
+    {
+      "epoch": 0.48,
+      "learning_rate": 0.00022335005018714785,
+      "loss": 0.3623,
+      "step": 1361
+    },
+    {
+      "epoch": 0.48,
+      "learning_rate": 0.00022312103594879514,
+      "loss": 0.4775,
+      "step": 1362
+    },
+    {
+      "epoch": 0.48,
+      "learning_rate": 0.00022289199097963499,
+      "loss": 0.4883,
+      "step": 1363
+    },
+    {
+      "epoch": 0.48,
+      "learning_rate": 0.00022266291558409716,
+      "loss": 0.3027,
+      "step": 1364
+    },
+    {
+      "epoch": 0.48,
+      "learning_rate": 0.00022243381006665228,
+      "loss": 1.1161,
+      "step": 1365
+    },
+    {
+      "epoch": 0.48,
+      "learning_rate": 0.00022220467473181067,
+      "loss": 0.2888,
+      "step": 1366
+    },
+    {
+      "epoch": 0.48,
+      "learning_rate": 0.00022197550988412251,
+      "loss": 0.5013,
+      "step": 1367
+    },
+    {
+      "epoch": 0.48,
+      "learning_rate": 0.00022174631582817726,
+      "loss": 0.7496,
+      "step": 1368
+    },
+    {
+      "epoch": 0.48,
+      "learning_rate": 0.0002215170928686028,
+      "loss": 0.3778,
+      "step": 1369
+    },
+    {
+      "epoch": 0.48,
+      "learning_rate": 0.00022128784131006585,
+      "loss": 0.629,
+      "step": 1370
+    },
+    {
+      "epoch": 0.49,
+      "learning_rate": 0.00022105856145727103,
+      "loss": 0.3661,
+      "step": 1371
+    },
+    {
+      "epoch": 0.49,
+      "learning_rate": 0.0002208292536149604,
+      "loss": 0.2388,
+      "step": 1372
+    },
+    {
+      "epoch": 0.49,
+      "learning_rate": 0.0002205999180879135,
+      "loss": 0.1968,
+      "step": 1373
+    },
+    {
+      "epoch": 0.49,
+      "learning_rate": 0.0002203705551809464,
+      "loss": 0.5576,
+      "step": 1374
+    },
+    {
+      "epoch": 0.49,
+      "learning_rate": 0.00022014116519891168,
+      "loss": 0.5475,
+      "step": 1375
+    },
+    {
+      "epoch": 0.49,
+      "learning_rate": 0.00021991174844669797,
+      "loss": 0.5335,
+      "step": 1376
+    },
+    {
+      "epoch": 0.49,
+      "learning_rate": 0.0002196823052292293,
+      "loss": 0.6106,
+      "step": 1377
+    },
+    {
+      "epoch": 0.49,
+      "learning_rate": 0.00021945283585146513,
+      "loss": 0.2603,
+      "step": 1378
+    },
+    {
+      "epoch": 0.49,
+      "learning_rate": 0.00021922334061839945,
+      "loss": 0.9632,
+      "step": 1379
+    },
+    {
+      "epoch": 0.49,
+      "learning_rate": 0.00021899381983506074,
+      "loss": 0.5084,
+      "step": 1380
+    },
+    {
+      "epoch": 0.49,
+      "eval_loss": 0.5821849703788757,
+      "eval_runtime": 52.9399,
+      "eval_samples_per_second": 30.771,
+      "eval_steps_per_second": 30.771,
+      "step": 1380
+    },
+    {
+      "epoch": 0.49,
+      "learning_rate": 0.00021876427380651143,
+      "loss": 0.427,
+      "step": 1381
+    },
+    {
+      "epoch": 0.49,
+      "learning_rate": 0.0002185347028378474,
+      "loss": 0.4211,
+      "step": 1382
+    },
+    {
+      "epoch": 0.49,
+      "learning_rate": 0.000218305107234198,
+      "loss": 0.2844,
+      "step": 1383
+    },
+    {
+      "epoch": 0.49,
+      "learning_rate": 0.00021807548730072478,
+      "loss": 0.5428,
+      "step": 1384
+    },
+    {
+      "epoch": 0.49,
+      "learning_rate": 0.00021784584334262212,
+      "loss": 0.5607,
+      "step": 1385
+    },
+    {
+      "epoch": 0.49,
+      "learning_rate": 0.00021761617566511615,
+      "loss": 0.6452,
+      "step": 1386
+    },
+    {
+      "epoch": 0.49,
+      "learning_rate": 0.0002173864845734644,
+      "loss": 0.4032,
+      "step": 1387
+    },
+    {
+      "epoch": 0.49,
+      "learning_rate": 0.00021715677037295586,
+      "loss": 0.6412,
+      "step": 1388
+    },
+    {
+      "epoch": 0.49,
+      "learning_rate": 0.00021692703336890977,
+      "loss": 0.4017,
+      "step": 1389
+    },
+    {
+      "epoch": 0.49,
+      "learning_rate": 0.0002166972738666761,
+      "loss": 0.3904,
+      "step": 1390
+    },
+    {
+      "epoch": 0.49,
+      "learning_rate": 0.00021646749217163453,
+      "loss": 0.4563,
+      "step": 1391
+    },
+    {
+      "epoch": 0.49,
+      "learning_rate": 0.00021623768858919414,
+      "loss": 0.4135,
+      "step": 1392
+    },
+    {
+      "epoch": 0.49,
+      "learning_rate": 0.00021600786342479343,
+      "loss": 0.5486,
+      "step": 1393
+    },
+    {
+      "epoch": 0.49,
+      "learning_rate": 0.00021577801698389917,
+      "loss": 0.5983,
+      "step": 1394
+    },
+    {
+      "epoch": 0.49,
+      "learning_rate": 0.0002155481495720067,
+      "loss": 0.6857,
+      "step": 1395
+    },
+    {
+      "epoch": 0.49,
+      "learning_rate": 0.00021531826149463913,
+      "loss": 0.408,
+      "step": 1396
+    },
+    {
+      "epoch": 0.49,
+      "learning_rate": 0.00021508835305734714,
+      "loss": 0.8218,
+      "step": 1397
+    },
+    {
+      "epoch": 0.49,
+      "learning_rate": 0.00021485842456570823,
+      "loss": 0.4784,
+      "step": 1398
+    },
+    {
+      "epoch": 0.5,
+      "learning_rate": 0.00021462847632532678,
+      "loss": 0.6064,
+      "step": 1399
+    },
+    {
+      "epoch": 0.5,
+      "learning_rate": 0.0002143985086418334,
+      "loss": 0.3838,
+      "step": 1400
+    },
+    {
+      "epoch": 0.5,
+      "eval_loss": 0.5604258179664612,
+      "eval_runtime": 52.9403,
+      "eval_samples_per_second": 30.77,
+      "eval_steps_per_second": 30.77,
+      "step": 1400
+    },
+    {
+      "epoch": 0.5,
+      "learning_rate": 0.00021416852182088448,
+      "loss": 0.7189,
+      "step": 1401
+    },
+    {
+      "epoch": 0.5,
+      "learning_rate": 0.00021393851616816184,
+      "loss": 0.4682,
+      "step": 1402
+    },
+    {
+      "epoch": 0.5,
+      "learning_rate": 0.0002137084919893723,
+      "loss": 0.4147,
+      "step": 1403
+    },
+    {
+      "epoch": 0.5,
+      "learning_rate": 0.00021347844959024732,
+      "loss": 0.5998,
+      "step": 1404
+    },
+    {
+      "epoch": 0.5,
+      "learning_rate": 0.00021324838927654272,
+      "loss": 0.9028,
+      "step": 1405
+    },
+    {
+      "epoch": 0.5,
+      "learning_rate": 0.00021301831135403797,
+      "loss": 0.3067,
+      "step": 1406
+    },
+    {
+      "epoch": 0.5,
+      "learning_rate": 0.000212788216128536,
+      "loss": 0.2926,
+      "step": 1407
+    },
+    {
+      "epoch": 0.5,
+      "learning_rate": 0.0002125581039058627,
+      "loss": 0.4061,
+      "step": 1408
+    },
+    {
+      "epoch": 0.5,
+      "learning_rate": 0.0002123279749918666,
+      "loss": 1.1498,
+      "step": 1409
+    },
+    {
+      "epoch": 0.5,
+      "learning_rate": 0.0002120978296924184,
+      "loss": 0.3295,
+      "step": 1410
+    },
+    {
+      "epoch": 0.5,
+      "learning_rate": 0.0002118676683134106,
+      "loss": 0.5981,
+      "step": 1411
+    },
+    {
+      "epoch": 0.5,
+      "learning_rate": 0.00021163749116075698,
+      "loss": 0.5044,
+      "step": 1412
+    },
+    {
+      "epoch": 0.5,
+      "learning_rate": 0.00021140729854039246,
+      "loss": 0.6656,
+      "step": 1413
+    },
+    {
+      "epoch": 0.5,
+      "learning_rate": 0.0002111770907582723,
+      "loss": 0.647,
+      "step": 1414
+    },
+    {
+      "epoch": 0.5,
+      "learning_rate": 0.00021094686812037203,
+      "loss": 0.5875,
+      "step": 1415
+    },
+    {
+      "epoch": 0.5,
+      "learning_rate": 0.000210716630932687,
+      "loss": 0.3679,
+      "step": 1416
+    },
+    {
+      "epoch": 0.5,
+      "learning_rate": 0.00021048637950123178,
+      "loss": 0.5177,
+      "step": 1417
+    },
+    {
+      "epoch": 0.5,
+      "learning_rate": 0.00021025611413203987,
+      "loss": 0.3204,
+      "step": 1418
+    },
+    {
+      "epoch": 0.5,
+      "learning_rate": 0.0002100258351311633,
+      "loss": 0.5331,
+      "step": 1419
+    },
+    {
+      "epoch": 0.5,
+      "learning_rate": 0.00020979554280467236,
+      "loss": 0.667,
+      "step": 1420
+    },
+    {
+      "epoch": 0.5,
+      "eval_loss": 0.5621675848960876,
+      "eval_runtime": 52.9427,
+      "eval_samples_per_second": 30.769,
+      "eval_steps_per_second": 30.769,
+      "step": 1420
+    },
+    {
+      "epoch": 0.5,
+      "learning_rate": 0.00020956523745865492,
+      "loss": 0.3405,
+      "step": 1421
+    },
+    {
+      "epoch": 0.5,
+      "learning_rate": 0.00020933491939921614,
+      "loss": 0.4611,
+      "step": 1422
+    },
+    {
+      "epoch": 0.5,
+      "learning_rate": 0.00020910458893247817,
+      "loss": 0.3134,
+      "step": 1423
+    },
+    {
+      "epoch": 0.5,
+      "learning_rate": 0.00020887424636457946,
+      "loss": 0.2679,
+      "step": 1424
+    },
+    {
+      "epoch": 0.5,
+      "learning_rate": 0.0002086438920016748,
+      "loss": 0.4828,
+      "step": 1425
+    },
+    {
+      "epoch": 0.5,
+      "learning_rate": 0.00020841352614993453,
+      "loss": 0.5104,
+      "step": 1426
+    },
+    {
+      "epoch": 0.51,
+      "learning_rate": 0.00020818314911554424,
+      "loss": 1.0708,
+      "step": 1427
+    },
+    {
+      "epoch": 0.51,
+      "learning_rate": 0.0002079527612047044,
+      "loss": 0.5944,
+      "step": 1428
+    },
+    {
+      "epoch": 0.51,
+      "learning_rate": 0.00020772236272362994,
+      "loss": 0.4121,
+      "step": 1429
+    },
+    {
+      "epoch": 0.51,
+      "learning_rate": 0.0002074919539785499,
+      "loss": 0.4093,
+      "step": 1430
+    },
+    {
+      "epoch": 0.51,
+      "learning_rate": 0.00020726153527570688,
+      "loss": 0.5503,
+      "step": 1431
+    },
+    {
+      "epoch": 0.51,
+      "learning_rate": 0.00020703110692135671,
+      "loss": 0.4128,
+      "step": 1432
+    },
+    {
+      "epoch": 0.51,
+      "learning_rate": 0.00020680066922176816,
+      "loss": 0.3683,
+      "step": 1433
+    },
+    {
+      "epoch": 0.51,
+      "learning_rate": 0.00020657022248322222,
+      "loss": 0.6488,
+      "step": 1434
+    },
+    {
+      "epoch": 0.51,
+      "learning_rate": 0.00020633976701201214,
+      "loss": 0.5424,
+      "step": 1435
+    },
+    {
+      "epoch": 0.51,
+      "learning_rate": 0.0002061093031144426,
+      "loss": 0.655,
+      "step": 1436
+    },
+    {
+      "epoch": 0.51,
+      "learning_rate": 0.0002058788310968295,
+      "loss": 0.4379,
+      "step": 1437
+    },
+    {
+      "epoch": 0.51,
+      "learning_rate": 0.00020564835126549963,
+      "loss": 0.772,
+      "step": 1438
+    },
+    {
+      "epoch": 0.51,
+      "learning_rate": 0.00020541786392679,
+      "loss": 0.3853,
+      "step": 1439
+    },
+    {
+      "epoch": 0.51,
+      "learning_rate": 0.00020518736938704782,
+      "loss": 0.681,
+      "step": 1440
+    },
+    {
+      "epoch": 0.51,
+      "eval_loss": 0.5631526112556458,
+      "eval_runtime": 52.9361,
+      "eval_samples_per_second": 30.773,
+      "eval_steps_per_second": 30.773,
+      "step": 1440
+    },
+    {
+      "epoch": 0.51,
+      "learning_rate": 0.0002049568679526297,
+      "loss": 0.6577,
+      "step": 1441
+    },
+    {
+      "epoch": 0.51,
+      "learning_rate": 0.00020472635992990147,
+      "loss": 0.9441,
+      "step": 1442
+    },
+    {
+      "epoch": 0.51,
+      "learning_rate": 0.00020449584562523768,
+      "loss": 0.9719,
+      "step": 1443
+    },
+    {
+      "epoch": 0.51,
+      "learning_rate": 0.00020426532534502127,
+      "loss": 0.3703,
+      "step": 1444
+    },
+    {
+      "epoch": 0.51,
+      "learning_rate": 0.00020403479939564307,
+      "loss": 0.9871,
+      "step": 1445
+    },
+    {
+      "epoch": 0.51,
+      "learning_rate": 0.00020380426808350158,
+      "loss": 0.8238,
+      "step": 1446
+    },
+    {
+      "epoch": 0.51,
+      "learning_rate": 0.00020357373171500233,
+      "loss": 0.4112,
+      "step": 1447
+    },
+    {
+      "epoch": 0.51,
+      "learning_rate": 0.00020334319059655753,
+      "loss": 0.384,
+      "step": 1448
+    },
+    {
+      "epoch": 0.51,
+      "learning_rate": 0.00020311264503458575,
+      "loss": 0.4319,
+      "step": 1449
+    },
+    {
+      "epoch": 0.51,
+      "learning_rate": 0.00020288209533551144,
+      "loss": 0.4324,
+      "step": 1450
+    },
+    {
+      "epoch": 0.51,
+      "learning_rate": 0.00020265154180576462,
+      "loss": 0.5721,
+      "step": 1451
+    },
+    {
+      "epoch": 0.51,
+      "learning_rate": 0.0002024209847517804,
+      "loss": 0.3311,
+      "step": 1452
+    },
+    {
+      "epoch": 0.51,
+      "learning_rate": 0.00020219042447999842,
+      "loss": 0.4011,
+      "step": 1453
+    },
+    {
+      "epoch": 0.51,
+      "learning_rate": 0.00020195986129686278,
+      "loss": 0.5845,
+      "step": 1454
+    },
+    {
+      "epoch": 0.51,
+      "learning_rate": 0.0002017292955088213,
+      "loss": 0.392,
+      "step": 1455
+    },
+    {
+      "epoch": 0.52,
+      "learning_rate": 0.0002014987274223254,
+      "loss": 0.3069,
+      "step": 1456
+    },
+    {
+      "epoch": 0.52,
+      "learning_rate": 0.00020126815734382953,
+      "loss": 0.431,
+      "step": 1457
+    },
+    {
+      "epoch": 0.52,
+      "learning_rate": 0.0002010375855797907,
+      "loss": 0.6939,
+      "step": 1458
+    },
+    {
+      "epoch": 0.52,
+      "learning_rate": 0.0002008070124366682,
+      "loss": 0.3886,
+      "step": 1459
+    },
+    {
+      "epoch": 0.52,
+      "learning_rate": 0.00020057643822092317,
+      "loss": 0.3593,
+      "step": 1460
+    },
+    {
+      "epoch": 0.52,
+      "eval_loss": 0.5577579736709595,
+      "eval_runtime": 52.9412,
+      "eval_samples_per_second": 30.77,
+      "eval_steps_per_second": 30.77,
+      "step": 1460
+    },
+    {
+      "epoch": 0.52,
+      "learning_rate": 0.00020034586323901818,
+      "loss": 0.6216,
+      "step": 1461
+    },
+    {
+      "epoch": 0.52,
+      "learning_rate": 0.0002001152877974168,
+      "loss": 0.344,
+      "step": 1462
+    },
+    {
+      "epoch": 0.52,
+      "learning_rate": 0.0001998847122025832,
+      "loss": 0.6251,
+      "step": 1463
+    },
+    {
+      "epoch": 0.52,
+      "learning_rate": 0.0001996541367609819,
+      "loss": 0.3954,
+      "step": 1464
+    },
+    {
+      "epoch": 0.52,
+      "learning_rate": 0.0001994235617790769,
+      "loss": 0.2807,
+      "step": 1465
+    },
+    {
+      "epoch": 0.52,
+      "learning_rate": 0.00019919298756333184,
+      "loss": 0.4056,
+      "step": 1466
+    },
+    {
+      "epoch": 0.52,
+      "learning_rate": 0.00019896241442020934,
+      "loss": 0.3619,
+      "step": 1467
+    },
+    {
+      "epoch": 0.52,
+      "learning_rate": 0.00019873184265617051,
+      "loss": 0.4726,
+      "step": 1468
+    },
+    {
+      "epoch": 0.52,
+      "learning_rate": 0.0001985012725776746,
+      "loss": 0.5441,
+      "step": 1469
+    },
+    {
+      "epoch": 0.52,
+      "learning_rate": 0.0001982707044911787,
+      "loss": 0.3821,
+      "step": 1470
+    },
+    {
+      "epoch": 0.52,
+      "learning_rate": 0.00019804013870313727,
+      "loss": 0.3346,
+      "step": 1471
+    },
+    {
+      "epoch": 0.52,
+      "learning_rate": 0.00019780957552000168,
+      "loss": 0.6706,
+      "step": 1472
+    },
+    {
+      "epoch": 0.52,
+      "learning_rate": 0.00019757901524821962,
+      "loss": 0.5898,
+      "step": 1473
+    },
+    {
+      "epoch": 0.52,
+      "learning_rate": 0.0001973484581942354,
+      "loss": 0.4688,
+      "step": 1474
+    },
+    {
+      "epoch": 0.52,
+      "learning_rate": 0.0001971179046644886,
+      "loss": 0.5299,
+      "step": 1475
+    },
+    {
+      "epoch": 0.52,
+      "learning_rate": 0.0001968873549654143,
+      "loss": 0.3435,
+      "step": 1476
+    },
+    {
+      "epoch": 0.52,
+      "learning_rate": 0.0001966568094034425,
+      "loss": 0.6199,
+      "step": 1477
+    },
+    {
+      "epoch": 0.52,
+      "learning_rate": 0.0001964262682849977,
+      "loss": 0.6386,
+      "step": 1478
+    },
+    {
+      "epoch": 0.52,
+      "learning_rate": 0.00019619573191649844,
+      "loss": 0.7308,
+      "step": 1479
+    },
+    {
+      "epoch": 0.52,
+      "learning_rate": 0.00019596520060435692,
+      "loss": 0.4509,
+      "step": 1480
+    },
+    {
+      "epoch": 0.52,
+      "eval_loss": 0.5609290599822998,
+      "eval_runtime": 52.9389,
+      "eval_samples_per_second": 30.771,
+      "eval_steps_per_second": 30.771,
+      "step": 1480
+    },
+    {
+      "epoch": 0.52,
+      "learning_rate": 0.00019573467465497878,
+      "loss": 0.4007,
+      "step": 1481
+    },
+    {
+      "epoch": 0.52,
+      "learning_rate": 0.00019550415437476245,
+      "loss": 0.4716,
+      "step": 1482
+    },
+    {
+      "epoch": 0.52,
+      "learning_rate": 0.00019527364007009855,
+      "loss": 0.4021,
+      "step": 1483
+    },
+    {
+      "epoch": 0.53,
+      "learning_rate": 0.00019504313204737035,
+      "loss": 0.6118,
+      "step": 1484
+    },
+    {
+      "epoch": 0.53,
+      "learning_rate": 0.00019481263061295223,
+      "loss": 0.3569,
+      "step": 1485
+    },
+    {
+      "epoch": 0.53,
+      "learning_rate": 0.00019458213607321002,
+      "loss": 0.3796,
+      "step": 1486
+    },
+    {
+      "epoch": 0.53,
+      "learning_rate": 0.00019435164873450041,
+      "loss": 0.3684,
+      "step": 1487
+    },
+    {
+      "epoch": 0.53,
+      "learning_rate": 0.00019412116890317054,
+      "loss": 0.3912,
+      "step": 1488
+    },
+    {
+      "epoch": 0.53,
+      "learning_rate": 0.0001938906968855575,
+      "loss": 0.387,
+      "step": 1489
+    },
+    {
+      "epoch": 0.53,
+      "learning_rate": 0.00019366023298798788,
+      "loss": 0.439,
+      "step": 1490
+    },
+    {
+      "epoch": 0.53,
+      "learning_rate": 0.00019342977751677783,
+      "loss": 0.4072,
+      "step": 1491
+    },
+    {
+      "epoch": 0.53,
+      "learning_rate": 0.00019319933077823192,
+      "loss": 0.483,
+      "step": 1492
+    },
+    {
+      "epoch": 0.53,
+      "learning_rate": 0.0001929688930786433,
+      "loss": 0.9915,
+      "step": 1493
+    },
+    {
+      "epoch": 0.53,
+      "learning_rate": 0.00019273846472429314,
+      "loss": 0.3509,
+      "step": 1494
+    },
+    {
+      "epoch": 0.53,
+      "learning_rate": 0.00019250804602145014,
+      "loss": 1.0152,
+      "step": 1495
+    },
+    {
+      "epoch": 0.53,
+      "learning_rate": 0.00019227763727637005,
+      "loss": 0.2443,
+      "step": 1496
+    },
+    {
+      "epoch": 0.53,
+      "learning_rate": 0.00019204723879529562,
+      "loss": 0.4028,
+      "step": 1497
+    },
+    {
+      "epoch": 0.53,
+      "learning_rate": 0.0001918168508844558,
+      "loss": 0.5126,
+      "step": 1498
+    },
+    {
+      "epoch": 0.53,
+      "learning_rate": 0.00019158647385006557,
+      "loss": 0.4291,
+      "step": 1499
+    },
+    {
+      "epoch": 0.53,
+      "learning_rate": 0.00019135610799832518,
+      "loss": 0.4752,
+      "step": 1500
+    },
+    {
+      "epoch": 0.53,
+      "eval_loss": 0.5493614077568054,
+      "eval_runtime": 52.9357,
+      "eval_samples_per_second": 30.773,
+      "eval_steps_per_second": 30.773,
+      "step": 1500
+    },
+    {
+      "epoch": 0.53,
+      "learning_rate": 0.00019112575363542061,
+      "loss": 0.6607,
+      "step": 1501
+    },
+    {
+      "epoch": 0.53,
+      "learning_rate": 0.00019089541106752193,
+      "loss": 0.4385,
+      "step": 1502
+    },
+    {
+      "epoch": 0.53,
+      "learning_rate": 0.00019066508060078388,
+      "loss": 0.3693,
+      "step": 1503
+    },
+    {
+      "epoch": 0.53,
+      "learning_rate": 0.0001904347625413451,
+      "loss": 0.3227,
+      "step": 1504
+    },
+    {
+      "epoch": 0.53,
+      "learning_rate": 0.00019020445719532765,
+      "loss": 0.202,
+      "step": 1505
+    },
+    {
+      "epoch": 0.53,
+      "learning_rate": 0.0001899741648688367,
+      "loss": 0.9358,
+      "step": 1506
+    },
+    {
+      "epoch": 0.53,
+      "learning_rate": 0.00018974388586796018,
+      "loss": 0.3132,
+      "step": 1507
+    },
+    {
+      "epoch": 0.53,
+      "learning_rate": 0.0001895136204987683,
+      "loss": 0.6641,
+      "step": 1508
+    },
+    {
+      "epoch": 0.53,
+      "learning_rate": 0.0001892833690673131,
+      "loss": 0.6373,
+      "step": 1509
+    },
+    {
+      "epoch": 0.53,
+      "learning_rate": 0.00018905313187962797,
+      "loss": 0.6215,
+      "step": 1510
+    },
+    {
+      "epoch": 0.53,
+      "learning_rate": 0.00018882290924172778,
+      "loss": 0.5158,
+      "step": 1511
+    },
+    {
+      "epoch": 0.54,
+      "learning_rate": 0.00018859270145960764,
+      "loss": 0.3519,
+      "step": 1512
+    },
+    {
+      "epoch": 0.54,
+      "learning_rate": 0.00018836250883924304,
+      "loss": 0.5169,
+      "step": 1513
+    },
+    {
+      "epoch": 0.54,
+      "learning_rate": 0.00018813233168658944,
+      "loss": 0.7764,
+      "step": 1514
+    },
+    {
+      "epoch": 0.54,
+      "learning_rate": 0.00018790217030758164,
+      "loss": 0.4301,
+      "step": 1515
+    },
+    {
+      "epoch": 0.54,
+      "learning_rate": 0.00018767202500813342,
+      "loss": 0.4883,
+      "step": 1516
+    },
+    {
+      "epoch": 0.54,
+      "learning_rate": 0.00018744189609413734,
+      "loss": 0.39,
+      "step": 1517
+    },
+    {
+      "epoch": 0.54,
+      "learning_rate": 0.00018721178387146406,
+      "loss": 0.5836,
+      "step": 1518
+    },
+    {
+      "epoch": 0.54,
+      "learning_rate": 0.00018698168864596213,
+      "loss": 0.6658,
+      "step": 1519
+    },
+    {
+      "epoch": 0.54,
+      "learning_rate": 0.0001867516107234573,
+      "loss": 0.3152,
+      "step": 1520
+    },
+    {
+      "epoch": 0.54,
+      "eval_loss": 0.5541138648986816,
+      "eval_runtime": 52.9374,
+      "eval_samples_per_second": 30.772,
+      "eval_steps_per_second": 30.772,
+      "step": 1520
+    },
+    {
+      "epoch": 0.54,
+      "learning_rate": 0.00018652155040975275,
+      "loss": 0.8281,
+      "step": 1521
+    },
+    {
+      "epoch": 0.54,
+      "learning_rate": 0.00018629150801062784,
+      "loss": 0.4905,
+      "step": 1522
+    },
+    {
+      "epoch": 0.54,
+      "learning_rate": 0.00018606148383183823,
+      "loss": 0.4728,
+      "step": 1523
+    },
+    {
+      "epoch": 0.54,
+      "learning_rate": 0.00018583147817911554,
+      "loss": 0.397,
+      "step": 1524
+    },
+    {
+      "epoch": 0.54,
+      "learning_rate": 0.0001856014913581666,
+      "loss": 0.3386,
+      "step": 1525
+    },
+    {
+      "epoch": 0.54,
+      "learning_rate": 0.0001853715236746732,
+      "loss": 0.8133,
+      "step": 1526
+    },
+    {
+      "epoch": 0.54,
+      "learning_rate": 0.0001851415754342918,
+      "loss": 0.4954,
+      "step": 1527
+    },
+    {
+      "epoch": 0.54,
+      "learning_rate": 0.00018491164694265294,
+      "loss": 0.7294,
+      "step": 1528
+    },
+    {
+      "epoch": 0.54,
+      "learning_rate": 0.0001846817385053609,
+      "loss": 0.3751,
+      "step": 1529
+    },
+    {
+      "epoch": 0.54,
+      "learning_rate": 0.0001844518504279933,
+      "loss": 1.0028,
+      "step": 1530
+    },
+    {
+      "epoch": 0.54,
+      "learning_rate": 0.0001842219830161009,
+      "loss": 0.4282,
+      "step": 1531
+    },
+    {
+      "epoch": 0.54,
+      "learning_rate": 0.00018399213657520667,
+      "loss": 0.4271,
+      "step": 1532
+    },
+    {
+      "epoch": 0.54,
+      "learning_rate": 0.00018376231141080585,
+      "loss": 0.556,
+      "step": 1533
+    },
+    {
+      "epoch": 0.54,
+      "learning_rate": 0.00018353250782836552,
+      "loss": 1.2679,
+      "step": 1534
+    },
+    {
+      "epoch": 0.54,
+      "learning_rate": 0.00018330272613332392,
+      "loss": 0.5013,
+      "step": 1535
+    },
+    {
+      "epoch": 0.54,
+      "learning_rate": 0.00018307296663109022,
+      "loss": 0.6231,
+      "step": 1536
+    },
+    {
+      "epoch": 0.54,
+      "learning_rate": 0.00018284322962704418,
+      "loss": 0.7399,
+      "step": 1537
+    },
+    {
+      "epoch": 0.54,
+      "learning_rate": 0.00018261351542653563,
+      "loss": 0.4458,
+      "step": 1538
+    },
+    {
+      "epoch": 0.54,
+      "learning_rate": 0.00018238382433488395,
+      "loss": 0.313,
+      "step": 1539
+    },
+    {
+      "epoch": 0.55,
+      "learning_rate": 0.00018215415665737792,
+      "loss": 0.3699,
+      "step": 1540
+    },
+    {
+      "epoch": 0.55,
+      "eval_loss": 0.5448759198188782,
+      "eval_runtime": 52.9444,
+      "eval_samples_per_second": 30.768,
+      "eval_steps_per_second": 30.768,
+      "step": 1540
+    },
+    {
+      "epoch": 0.55,
+      "learning_rate": 0.0001819245126992753,
+      "loss": 0.4121,
+      "step": 1541
+    },
+    {
+      "epoch": 0.55,
+      "learning_rate": 0.0001816948927658021,
+      "loss": 0.7484,
+      "step": 1542
+    },
+    {
+      "epoch": 0.55,
+      "learning_rate": 0.00018146529716215256,
+      "loss": 0.4702,
+      "step": 1543
+    },
+    {
+      "epoch": 0.55,
+      "learning_rate": 0.0001812357261934886,
+      "loss": 0.2747,
+      "step": 1544
+    },
+    {
+      "epoch": 0.55,
+      "learning_rate": 0.00018100618016493928,
+      "loss": 0.5758,
+      "step": 1545
+    },
+    {
+      "epoch": 0.55,
+      "learning_rate": 0.00018077665938160057,
+      "loss": 0.4439,
+      "step": 1546
+    },
+    {
+      "epoch": 0.55,
+      "learning_rate": 0.00018054716414853488,
+      "loss": 0.257,
+      "step": 1547
+    },
+    {
+      "epoch": 0.55,
+      "learning_rate": 0.00018031769477077074,
+      "loss": 0.4212,
+      "step": 1548
+    },
+    {
+      "epoch": 0.55,
+      "learning_rate": 0.00018008825155330213,
+      "loss": 0.4067,
+      "step": 1549
+    },
+    {
+      "epoch": 0.55,
+      "learning_rate": 0.00017985883480108836,
+      "loss": 0.6826,
+      "step": 1550
+    },
+    {
+      "epoch": 0.55,
+      "learning_rate": 0.00017962944481905366,
+      "loss": 0.3538,
+      "step": 1551
+    },
+    {
+      "epoch": 0.55,
+      "learning_rate": 0.00017940008191208657,
+      "loss": 0.5107,
+      "step": 1552
+    },
+    {
+      "epoch": 0.55,
+      "learning_rate": 0.00017917074638503958,
+      "loss": 0.4253,
+      "step": 1553
+    },
+    {
+      "epoch": 0.55,
+      "learning_rate": 0.000178941438542729,
+      "loss": 0.449,
+      "step": 1554
+    },
+    {
+      "epoch": 0.55,
+      "learning_rate": 0.00017871215868993416,
+      "loss": 0.389,
+      "step": 1555
+    },
+    {
+      "epoch": 0.55,
+      "learning_rate": 0.00017848290713139722,
+      "loss": 0.4729,
+      "step": 1556
+    },
+    {
+      "epoch": 0.55,
+      "learning_rate": 0.0001782536841718228,
+      "loss": 0.4447,
+      "step": 1557
+    },
+    {
+      "epoch": 0.55,
+      "learning_rate": 0.0001780244901158775,
+      "loss": 0.6234,
+      "step": 1558
+    },
+    {
+      "epoch": 0.55,
+      "learning_rate": 0.00017779532526818938,
+      "loss": 0.506,
+      "step": 1559
+    },
+    {
+      "epoch": 0.55,
+      "learning_rate": 0.0001775661899333478,
+      "loss": 0.3009,
+      "step": 1560
+    },
+    {
+      "epoch": 0.55,
+      "eval_loss": 0.5655881762504578,
+      "eval_runtime": 52.9391,
+      "eval_samples_per_second": 30.771,
+      "eval_steps_per_second": 30.771,
+      "step": 1560
+    },
+    {
+      "epoch": 0.55,
+      "learning_rate": 0.00017733708441590288,
+      "loss": 0.3759,
+      "step": 1561
+    },
+    {
+      "epoch": 0.55,
+      "learning_rate": 0.00017710800902036512,
+      "loss": 0.3225,
+      "step": 1562
+    },
+    {
+      "epoch": 0.55,
+      "learning_rate": 0.00017687896405120486,
+      "loss": 0.3991,
+      "step": 1563
+    },
+    {
+      "epoch": 0.55,
+      "learning_rate": 0.0001766499498128522,
+      "loss": 0.4062,
+      "step": 1564
+    },
+    {
+      "epoch": 0.55,
+      "learning_rate": 0.00017642096660969628,
+      "loss": 1.0127,
+      "step": 1565
+    },
+    {
+      "epoch": 0.55,
+      "learning_rate": 0.0001761920147460849,
+      "loss": 0.3806,
+      "step": 1566
+    },
+    {
+      "epoch": 0.55,
+      "learning_rate": 0.0001759630945263244,
+      "loss": 0.7003,
+      "step": 1567
+    },
+    {
+      "epoch": 0.55,
+      "learning_rate": 0.00017573420625467903,
+      "loss": 0.4204,
+      "step": 1568
+    },
+    {
+      "epoch": 0.56,
+      "learning_rate": 0.00017550535023537036,
+      "loss": 0.7833,
+      "step": 1569
+    },
+    {
+      "epoch": 0.56,
+      "learning_rate": 0.00017527652677257728,
+      "loss": 0.449,
+      "step": 1570
+    },
+    {
+      "epoch": 0.56,
+      "learning_rate": 0.00017504773617043546,
+      "loss": 0.7725,
+      "step": 1571
+    },
+    {
+      "epoch": 0.56,
+      "learning_rate": 0.00017481897873303673,
+      "loss": 0.5256,
+      "step": 1572
+    },
+    {
+      "epoch": 0.56,
+      "learning_rate": 0.00017459025476442895,
+      "loss": 0.4515,
+      "step": 1573
+    },
+    {
+      "epoch": 0.56,
+      "learning_rate": 0.00017436156456861548,
+      "loss": 0.3007,
+      "step": 1574
+    },
+    {
+      "epoch": 0.56,
+      "learning_rate": 0.00017413290844955478,
+      "loss": 0.2267,
+      "step": 1575
+    },
+    {
+      "epoch": 0.56,
+      "learning_rate": 0.00017390428671115992,
+      "loss": 0.3397,
+      "step": 1576
+    },
+    {
+      "epoch": 0.56,
+      "learning_rate": 0.0001736756996572985,
+      "loss": 0.8373,
+      "step": 1577
+    },
+    {
+      "epoch": 0.56,
+      "learning_rate": 0.0001734471475917919,
+      "loss": 0.3768,
+      "step": 1578
+    },
+    {
+      "epoch": 0.56,
+      "learning_rate": 0.00017321863081841486,
+      "loss": 0.5963,
+      "step": 1579
+    },
+    {
+      "epoch": 0.56,
+      "learning_rate": 0.00017299014964089533,
+      "loss": 0.2867,
+      "step": 1580
+    },
+    {
+      "epoch": 0.56,
+      "eval_loss": 0.5499367713928223,
+      "eval_runtime": 52.94,
+      "eval_samples_per_second": 30.771,
+      "eval_steps_per_second": 30.771,
+      "step": 1580
+    },
+    {
+      "epoch": 0.56,
+      "learning_rate": 0.00017276170436291406,
+      "loss": 0.7401,
+      "step": 1581
+    },
+    {
+      "epoch": 0.56,
+      "learning_rate": 0.00017253329528810394,
+      "loss": 0.3955,
+      "step": 1582
+    },
+    {
+      "epoch": 0.56,
+      "learning_rate": 0.00017230492272004972,
+      "loss": 0.4978,
+      "step": 1583
+    },
+    {
+      "epoch": 0.56,
+      "learning_rate": 0.00017207658696228779,
+      "loss": 0.4739,
+      "step": 1584
+    },
+    {
+      "epoch": 0.56,
+      "learning_rate": 0.0001718482883183055,
+      "loss": 0.7508,
+      "step": 1585
+    },
+    {
+      "epoch": 0.56,
+      "learning_rate": 0.00017162002709154075,
+      "loss": 0.9164,
+      "step": 1586
+    },
+    {
+      "epoch": 0.56,
+      "learning_rate": 0.00017139180358538206,
+      "loss": 0.3839,
+      "step": 1587
+    },
+    {
+      "epoch": 0.56,
+      "learning_rate": 0.00017116361810316742,
+      "loss": 0.5636,
+      "step": 1588
+    },
+    {
+      "epoch": 0.56,
+      "learning_rate": 0.00017093547094818452,
+      "loss": 0.6955,
+      "step": 1589
+    },
+    {
+      "epoch": 0.56,
+      "learning_rate": 0.00017070736242367,
+      "loss": 0.6553,
+      "step": 1590
+    },
+    {
+      "epoch": 0.56,
+      "learning_rate": 0.0001704792928328093,
+      "loss": 0.5366,
+      "step": 1591
+    },
+    {
+      "epoch": 0.56,
+      "learning_rate": 0.00017025126247873594,
+      "loss": 0.3131,
+      "step": 1592
+    },
+    {
+      "epoch": 0.56,
+      "learning_rate": 0.00017002327166453133,
+      "loss": 0.6914,
+      "step": 1593
+    },
+    {
+      "epoch": 0.56,
+      "learning_rate": 0.00016979532069322448,
+      "loss": 0.4003,
+      "step": 1594
+    },
+    {
+      "epoch": 0.56,
+      "learning_rate": 0.00016956740986779126,
+      "loss": 1.2095,
+      "step": 1595
+    },
+    {
+      "epoch": 0.56,
+      "learning_rate": 0.0001693395394911541,
+      "loss": 0.7205,
+      "step": 1596
+    },
+    {
+      "epoch": 0.57,
+      "learning_rate": 0.0001691117098661821,
+      "loss": 0.3489,
+      "step": 1597
+    },
+    {
+      "epoch": 0.57,
+      "learning_rate": 0.00016888392129568974,
+      "loss": 0.6098,
+      "step": 1598
+    },
+    {
+      "epoch": 0.57,
+      "learning_rate": 0.00016865617408243707,
+      "loss": 0.6616,
+      "step": 1599
+    },
+    {
+      "epoch": 0.57,
+      "learning_rate": 0.0001684284685291292,
+      "loss": 0.7261,
+      "step": 1600
+    },
+    {
+      "epoch": 0.57,
+      "eval_loss": 0.5490433573722839,
+      "eval_runtime": 52.9351,
+      "eval_samples_per_second": 30.774,
+      "eval_steps_per_second": 30.774,
+      "step": 1600
+    },
+    {
+      "epoch": 0.57,
+      "learning_rate": 0.00016820080493841597,
+      "loss": 0.986,
+      "step": 1601
+    },
+    {
+      "epoch": 0.57,
+      "learning_rate": 0.00016797318361289133,
+      "loss": 0.3943,
+      "step": 1602
+    },
+    {
+      "epoch": 0.57,
+      "learning_rate": 0.00016774560485509295,
+      "loss": 0.4264,
+      "step": 1603
+    },
+    {
+      "epoch": 0.57,
+      "learning_rate": 0.00016751806896750216,
+      "loss": 0.3746,
+      "step": 1604
+    },
+    {
+      "epoch": 0.57,
+      "learning_rate": 0.00016729057625254327,
+      "loss": 0.3294,
+      "step": 1605
+    },
+    {
+      "epoch": 0.57,
+      "learning_rate": 0.00016706312701258286,
+      "loss": 1.2316,
+      "step": 1606
+    },
+    {
+      "epoch": 0.57,
+      "learning_rate": 0.0001668357215499303,
+      "loss": 0.4212,
+      "step": 1607
+    },
+    {
+      "epoch": 0.57,
+      "learning_rate": 0.00016660836016683627,
+      "loss": 0.482,
+      "step": 1608
+    },
+    {
+      "epoch": 0.57,
+      "learning_rate": 0.00016638104316549316,
+      "loss": 0.559,
+      "step": 1609
+    },
+    {
+      "epoch": 0.57,
+      "learning_rate": 0.0001661537708480342,
+      "loss": 0.6198,
+      "step": 1610
+    },
+    {
+      "epoch": 0.57,
+      "learning_rate": 0.00016592654351653337,
+      "loss": 0.3006,
+      "step": 1611
+    },
+    {
+      "epoch": 0.57,
+      "learning_rate": 0.0001656993614730048,
+      "loss": 0.6537,
+      "step": 1612
+    },
+    {
+      "epoch": 0.57,
+      "learning_rate": 0.00016547222501940234,
+      "loss": 0.4003,
+      "step": 1613
+    },
+    {
+      "epoch": 0.57,
+      "learning_rate": 0.00016524513445761942,
+      "loss": 0.6035,
+      "step": 1614
+    },
+    {
+      "epoch": 0.57,
+      "learning_rate": 0.0001650180900894884,
+      "loss": 0.4058,
+      "step": 1615
+    },
+    {
+      "epoch": 0.57,
+      "learning_rate": 0.00016479109221678,
+      "loss": 0.4622,
+      "step": 1616
+    },
+    {
+      "epoch": 0.57,
+      "learning_rate": 0.00016456414114120367,
+      "loss": 0.4073,
+      "step": 1617
+    },
+    {
+      "epoch": 0.57,
+      "learning_rate": 0.00016433723716440614,
+      "loss": 0.3906,
+      "step": 1618
+    },
+    {
+      "epoch": 0.57,
+      "learning_rate": 0.00016411038058797184,
+      "loss": 0.4646,
+      "step": 1619
+    },
+    {
+      "epoch": 0.57,
+      "learning_rate": 0.00016388357171342202,
+      "loss": 0.5149,
+      "step": 1620
+    },
+    {
+      "epoch": 0.57,
+      "eval_loss": 0.5565251111984253,
+      "eval_runtime": 52.9453,
+      "eval_samples_per_second": 30.768,
+      "eval_steps_per_second": 30.768,
+      "step": 1620
+    },
+    {
+      "epoch": 0.57,
+      "learning_rate": 0.00016365681084221478,
+      "loss": 0.6541,
+      "step": 1621
+    },
+    {
+      "epoch": 0.57,
+      "learning_rate": 0.00016343009827574415,
+      "loss": 0.4784,
+      "step": 1622
+    },
+    {
+      "epoch": 0.57,
+      "learning_rate": 0.00016320343431534002,
+      "loss": 0.5526,
+      "step": 1623
+    },
+    {
+      "epoch": 0.57,
+      "learning_rate": 0.00016297681926226784,
+      "loss": 0.4894,
+      "step": 1624
+    },
+    {
+      "epoch": 0.58,
+      "learning_rate": 0.00016275025341772794,
+      "loss": 0.3379,
+      "step": 1625
+    },
+    {
+      "epoch": 0.58,
+      "learning_rate": 0.00016252373708285504,
+      "loss": 0.5429,
+      "step": 1626
+    },
+    {
+      "epoch": 0.58,
+      "learning_rate": 0.00016229727055871854,
+      "loss": 0.9478,
+      "step": 1627
+    },
+    {
+      "epoch": 0.58,
+      "learning_rate": 0.00016207085414632116,
+      "loss": 0.4644,
+      "step": 1628
+    },
+    {
+      "epoch": 0.58,
+      "learning_rate": 0.00016184448814659928,
+      "loss": 0.4745,
+      "step": 1629
+    },
+    {
+      "epoch": 0.58,
+      "learning_rate": 0.00016161817286042214,
+      "loss": 0.3051,
+      "step": 1630
+    },
+    {
+      "epoch": 0.58,
+      "learning_rate": 0.00016139190858859172,
+      "loss": 0.3633,
+      "step": 1631
+    },
+    {
+      "epoch": 0.58,
+      "learning_rate": 0.00016116569563184212,
+      "loss": 0.3494,
+      "step": 1632
+    },
+    {
+      "epoch": 0.58,
+      "learning_rate": 0.00016093953429083905,
+      "loss": 0.4802,
+      "step": 1633
+    },
+    {
+      "epoch": 0.58,
+      "learning_rate": 0.0001607134248661801,
+      "loss": 0.5268,
+      "step": 1634
+    },
+    {
+      "epoch": 0.58,
+      "learning_rate": 0.0001604873676583933,
+      "loss": 0.674,
+      "step": 1635
+    },
+    {
+      "epoch": 0.58,
+      "learning_rate": 0.00016026136296793756,
+      "loss": 0.5817,
+      "step": 1636
+    },
+    {
+      "epoch": 0.58,
+      "learning_rate": 0.00016003541109520212,
+      "loss": 0.3799,
+      "step": 1637
+    },
+    {
+      "epoch": 0.58,
+      "learning_rate": 0.00015980951234050573,
+      "loss": 0.4465,
+      "step": 1638
+    },
+    {
+      "epoch": 0.58,
+      "learning_rate": 0.00015958366700409666,
+      "loss": 0.5499,
+      "step": 1639
+    },
+    {
+      "epoch": 0.58,
+      "learning_rate": 0.00015935787538615216,
+      "loss": 0.473,
+      "step": 1640
+    },
+    {
+      "epoch": 0.58,
+      "eval_loss": 0.5444967746734619,
+      "eval_runtime": 52.9348,
+      "eval_samples_per_second": 30.774,
+      "eval_steps_per_second": 30.774,
+      "step": 1640
+    },
+    {
+      "epoch": 0.58,
+      "learning_rate": 0.0001591321377867782,
+      "loss": 0.3187,
+      "step": 1641
+    },
+    {
+      "epoch": 0.58,
+      "learning_rate": 0.00015890645450600886,
+      "loss": 0.3908,
+      "step": 1642
+    },
+    {
+      "epoch": 0.58,
+      "learning_rate": 0.00015868082584380585,
+      "loss": 0.7436,
+      "step": 1643
+    },
+    {
+      "epoch": 0.58,
+      "learning_rate": 0.00015845525210005874,
+      "loss": 0.5456,
+      "step": 1644
+    },
+    {
+      "epoch": 0.58,
+      "learning_rate": 0.00015822973357458358,
+      "loss": 0.4146,
+      "step": 1645
+    },
+    {
+      "epoch": 0.58,
+      "learning_rate": 0.00015800427056712334,
+      "loss": 0.4188,
+      "step": 1646
+    },
+    {
+      "epoch": 0.58,
+      "learning_rate": 0.0001577788633773472,
+      "loss": 0.936,
+      "step": 1647
+    },
+    {
+      "epoch": 0.58,
+      "learning_rate": 0.00015755351230485002,
+      "loss": 0.373,
+      "step": 1648
+    },
+    {
+      "epoch": 0.58,
+      "learning_rate": 0.00015732821764915218,
+      "loss": 0.378,
+      "step": 1649
+    },
+    {
+      "epoch": 0.58,
+      "learning_rate": 0.00015710297970969892,
+      "loss": 0.5807,
+      "step": 1650
+    },
+    {
+      "epoch": 0.58,
+      "learning_rate": 0.00015687779878586033,
+      "loss": 0.4829,
+      "step": 1651
+    },
+    {
+      "epoch": 0.58,
+      "learning_rate": 0.0001566526751769306,
+      "loss": 0.901,
+      "step": 1652
+    },
+    {
+      "epoch": 0.59,
+      "learning_rate": 0.00015642760918212756,
+      "loss": 0.3496,
+      "step": 1653
+    },
+    {
+      "epoch": 0.59,
+      "learning_rate": 0.0001562026011005929,
+      "loss": 0.4739,
+      "step": 1654
+    },
+    {
+      "epoch": 0.59,
+      "learning_rate": 0.00015597765123139093,
+      "loss": 0.4666,
+      "step": 1655
+    },
+    {
+      "epoch": 0.59,
+      "learning_rate": 0.00015575275987350867,
+      "loss": 0.4569,
+      "step": 1656
+    },
+    {
+      "epoch": 0.59,
+      "learning_rate": 0.0001555279273258556,
+      "loss": 0.336,
+      "step": 1657
+    },
+    {
+      "epoch": 0.59,
+      "learning_rate": 0.00015530315388726274,
+      "loss": 0.5362,
+      "step": 1658
+    },
+    {
+      "epoch": 0.59,
+      "learning_rate": 0.00015507843985648276,
+      "loss": 0.8438,
+      "step": 1659
+    },
+    {
+      "epoch": 0.59,
+      "learning_rate": 0.0001548537855321892,
+      "loss": 0.9732,
+      "step": 1660
+    },
+    {
+      "epoch": 0.59,
+      "eval_loss": 0.5505257844924927,
+      "eval_runtime": 52.9427,
+      "eval_samples_per_second": 30.769,
+      "eval_steps_per_second": 30.769,
+      "step": 1660
+    },
+    {
+      "epoch": 0.59,
+      "learning_rate": 0.0001546291912129764,
+      "loss": 0.4887,
+      "step": 1661
+    },
+    {
+      "epoch": 0.59,
+      "learning_rate": 0.00015440465719735892,
+      "loss": 0.7585,
+      "step": 1662
+    },
+    {
+      "epoch": 0.59,
+      "learning_rate": 0.0001541801837837709,
+      "loss": 0.4984,
+      "step": 1663
+    },
+    {
+      "epoch": 0.59,
+      "learning_rate": 0.00015395577127056646,
+      "loss": 0.5058,
+      "step": 1664
+    },
+    {
+      "epoch": 0.59,
+      "learning_rate": 0.00015373141995601825,
+      "loss": 0.6153,
+      "step": 1665
+    },
+    {
+      "epoch": 0.59,
+      "learning_rate": 0.00015350713013831787,
+      "loss": 0.5976,
+      "step": 1666
+    },
+    {
+      "epoch": 0.59,
+      "learning_rate": 0.0001532829021155751,
+      "loss": 0.6651,
+      "step": 1667
+    },
+    {
+      "epoch": 0.59,
+      "learning_rate": 0.00015305873618581765,
+      "loss": 0.5251,
+      "step": 1668
+    },
+    {
+      "epoch": 0.59,
+      "learning_rate": 0.0001528346326469906,
+      "loss": 0.296,
+      "step": 1669
+    },
+    {
+      "epoch": 0.59,
+      "learning_rate": 0.0001526105917969561,
+      "loss": 1.1409,
+      "step": 1670
+    },
+    {
+      "epoch": 0.59,
+      "learning_rate": 0.00015238661393349317,
+      "loss": 0.3475,
+      "step": 1671
+    },
+    {
+      "epoch": 0.59,
+      "learning_rate": 0.00015216269935429692,
+      "loss": 0.4764,
+      "step": 1672
+    },
+    {
+      "epoch": 0.59,
+      "learning_rate": 0.00015193884835697826,
+      "loss": 0.3182,
+      "step": 1673
+    },
+    {
+      "epoch": 0.59,
+      "learning_rate": 0.00015171506123906397,
+      "loss": 0.4693,
+      "step": 1674
+    },
+    {
+      "epoch": 0.59,
+      "learning_rate": 0.0001514913382979956,
+      "loss": 0.5937,
+      "step": 1675
+    },
+    {
+      "epoch": 0.59,
+      "learning_rate": 0.00015126767983112934,
+      "loss": 0.5481,
+      "step": 1676
+    },
+    {
+      "epoch": 0.59,
+      "learning_rate": 0.00015104408613573608,
+      "loss": 0.9115,
+      "step": 1677
+    },
+    {
+      "epoch": 0.59,
+      "learning_rate": 0.00015082055750900022,
+      "loss": 0.5761,
+      "step": 1678
+    },
+    {
+      "epoch": 0.59,
+      "learning_rate": 0.0001505970942480199,
+      "loss": 0.4047,
+      "step": 1679
+    },
+    {
+      "epoch": 0.59,
+      "learning_rate": 0.0001503736966498062,
+      "loss": 0.9606,
+      "step": 1680
+    },
+    {
+      "epoch": 0.59,
+      "eval_loss": 0.5470832586288452,
+      "eval_runtime": 52.9431,
+      "eval_samples_per_second": 30.769,
+      "eval_steps_per_second": 30.769,
+      "step": 1680
+    },
+    {
+      "epoch": 0.59,
+      "learning_rate": 0.00015015036501128323,
+      "loss": 0.4405,
+      "step": 1681
+    },
+    {
+      "epoch": 0.6,
+      "learning_rate": 0.0001499270996292871,
+      "loss": 0.407,
+      "step": 1682
+    },
+    {
+      "epoch": 0.6,
+      "learning_rate": 0.000149703900800566,
+      "loss": 0.4156,
+      "step": 1683
+    },
+    {
+      "epoch": 0.6,
+      "learning_rate": 0.00014948076882177972,
+      "loss": 0.4558,
+      "step": 1684
+    },
+    {
+      "epoch": 0.6,
+      "learning_rate": 0.00014925770398949913,
+      "loss": 0.5148,
+      "step": 1685
+    },
+    {
+      "epoch": 0.6,
+      "learning_rate": 0.00014903470660020577,
+      "loss": 0.4339,
+      "step": 1686
+    },
+    {
+      "epoch": 0.6,
+      "learning_rate": 0.00014881177695029175,
+      "loss": 0.4863,
+      "step": 1687
+    },
+    {
+      "epoch": 0.6,
+      "learning_rate": 0.00014858891533605896,
+      "loss": 0.3916,
+      "step": 1688
+    },
+    {
+      "epoch": 0.6,
+      "learning_rate": 0.000148366122053719,
+      "loss": 0.5385,
+      "step": 1689
+    },
+    {
+      "epoch": 0.6,
+      "learning_rate": 0.00014814339739939239,
+      "loss": 0.3995,
+      "step": 1690
+    },
+    {
+      "epoch": 0.6,
+      "learning_rate": 0.00014792074166910887,
+      "loss": 0.3499,
+      "step": 1691
+    },
+    {
+      "epoch": 0.6,
+      "learning_rate": 0.0001476981551588062,
+      "loss": 0.4006,
+      "step": 1692
+    },
+    {
+      "epoch": 0.6,
+      "learning_rate": 0.00014747563816433013,
+      "loss": 0.6519,
+      "step": 1693
+    },
+    {
+      "epoch": 0.6,
+      "learning_rate": 0.00014725319098143438,
+      "loss": 0.4737,
+      "step": 1694
+    },
+    {
+      "epoch": 0.6,
+      "learning_rate": 0.00014703081390577953,
+      "loss": 0.381,
+      "step": 1695
+    },
+    {
+      "epoch": 0.6,
+      "learning_rate": 0.00014680850723293305,
+      "loss": 0.9777,
+      "step": 1696
+    },
+    {
+      "epoch": 0.6,
+      "learning_rate": 0.00014658627125836897,
+      "loss": 0.5883,
+      "step": 1697
+    },
+    {
+      "epoch": 0.6,
+      "learning_rate": 0.00014636410627746722,
+      "loss": 0.4787,
+      "step": 1698
+    },
+    {
+      "epoch": 0.6,
+      "learning_rate": 0.00014614201258551348,
+      "loss": 0.4814,
+      "step": 1699
+    },
+    {
+      "epoch": 0.6,
+      "learning_rate": 0.00014591999047769842,
+      "loss": 0.2714,
+      "step": 1700
+    },
+    {
+      "epoch": 0.6,
+      "eval_loss": 0.5651383996009827,
+      "eval_runtime": 52.9245,
+      "eval_samples_per_second": 30.78,
+      "eval_steps_per_second": 30.78,
+      "step": 1700
+    },
+    {
+      "epoch": 0.6,
+      "learning_rate": 0.00014569804024911802,
+      "loss": 0.4936,
+      "step": 1701
+    },
+    {
+      "epoch": 0.6,
+      "learning_rate": 0.00014547616219477228,
+      "loss": 0.5891,
+      "step": 1702
+    },
+    {
+      "epoch": 0.6,
+      "learning_rate": 0.00014525435660956543,
+      "loss": 0.4512,
+      "step": 1703
+    },
+    {
+      "epoch": 0.6,
+      "learning_rate": 0.0001450326237883055,
+      "loss": 0.4975,
+      "step": 1704
+    },
+    {
+      "epoch": 0.6,
+      "learning_rate": 0.0001448109640257037,
+      "loss": 0.373,
+      "step": 1705
+    },
+    {
+      "epoch": 0.6,
+      "learning_rate": 0.00014458937761637408,
+      "loss": 0.961,
+      "step": 1706
+    },
+    {
+      "epoch": 0.6,
+      "learning_rate": 0.00014436786485483335,
+      "loss": 0.3207,
+      "step": 1707
+    },
+    {
+      "epoch": 0.6,
+      "learning_rate": 0.00014414642603550014,
+      "loss": 0.6255,
+      "step": 1708
+    },
+    {
+      "epoch": 0.6,
+      "learning_rate": 0.00014392506145269503,
+      "loss": 0.4162,
+      "step": 1709
+    },
+    {
+      "epoch": 0.61,
+      "learning_rate": 0.00014370377140063956,
+      "loss": 0.4452,
+      "step": 1710
+    },
+    {
+      "epoch": 0.61,
+      "learning_rate": 0.0001434825561734568,
+      "loss": 0.304,
+      "step": 1711
+    },
+    {
+      "epoch": 0.61,
+      "learning_rate": 0.00014326141606516973,
+      "loss": 0.2823,
+      "step": 1712
+    },
+    {
+      "epoch": 0.61,
+      "learning_rate": 0.00014304035136970185,
+      "loss": 0.3219,
+      "step": 1713
+    },
+    {
+      "epoch": 0.61,
+      "learning_rate": 0.00014281936238087642,
+      "loss": 0.2895,
+      "step": 1714
+    },
+    {
+      "epoch": 0.61,
+      "learning_rate": 0.00014259844939241596,
+      "loss": 0.2757,
+      "step": 1715
+    },
+    {
+      "epoch": 0.61,
+      "learning_rate": 0.00014237761269794198,
+      "loss": 0.5828,
+      "step": 1716
+    },
+    {
+      "epoch": 0.61,
+      "learning_rate": 0.00014215685259097476,
+      "loss": 0.6543,
+      "step": 1717
+    },
+    {
+      "epoch": 0.61,
+      "learning_rate": 0.00014193616936493256,
+      "loss": 0.4037,
+      "step": 1718
+    },
+    {
+      "epoch": 0.61,
+      "learning_rate": 0.00014171556331313162,
+      "loss": 0.3212,
+      "step": 1719
+    },
+    {
+      "epoch": 0.61,
+      "learning_rate": 0.00014149503472878539,
+      "loss": 0.4927,
+      "step": 1720
+    },
+    {
+      "epoch": 0.61,
+      "eval_loss": 0.5527106523513794,
+      "eval_runtime": 52.9418,
+      "eval_samples_per_second": 30.77,
+      "eval_steps_per_second": 30.77,
+      "step": 1720
+    },
+    {
+      "epoch": 0.61,
+      "learning_rate": 0.00014127458390500469,
+      "loss": 0.3853,
+      "step": 1721
+    },
+    {
+      "epoch": 0.61,
+      "learning_rate": 0.00014105421113479668,
+      "loss": 0.3582,
+      "step": 1722
+    },
+    {
+      "epoch": 0.61,
+      "learning_rate": 0.0001408339167110649,
+      "loss": 0.3974,
+      "step": 1723
+    },
+    {
+      "epoch": 0.61,
+      "learning_rate": 0.00014061370092660882,
+      "loss": 0.5884,
+      "step": 1724
+    },
+    {
+      "epoch": 0.61,
+      "learning_rate": 0.00014039356407412326,
+      "loss": 0.5498,
+      "step": 1725
+    },
+    {
+      "epoch": 0.61,
+      "learning_rate": 0.0001401735064461982,
+      "loss": 0.35,
+      "step": 1726
+    },
+    {
+      "epoch": 0.61,
+      "learning_rate": 0.00013995352833531837,
+      "loss": 0.5158,
+      "step": 1727
+    },
+    {
+      "epoch": 0.61,
+      "learning_rate": 0.00013973363003386277,
+      "loss": 0.9577,
+      "step": 1728
+    },
+    {
+      "epoch": 0.61,
+      "learning_rate": 0.00013951381183410422,
+      "loss": 0.3639,
+      "step": 1729
+    },
+    {
+      "epoch": 0.61,
+      "learning_rate": 0.0001392940740282091,
+      "loss": 0.4004,
+      "step": 1730
+    },
+    {
+      "epoch": 0.61,
+      "learning_rate": 0.00013907441690823714,
+      "loss": 0.3886,
+      "step": 1731
+    },
+    {
+      "epoch": 0.61,
+      "learning_rate": 0.00013885484076614065,
+      "loss": 0.5985,
+      "step": 1732
+    },
+    {
+      "epoch": 0.61,
+      "learning_rate": 0.00013863534589376428,
+      "loss": 0.4303,
+      "step": 1733
+    },
+    {
+      "epoch": 0.61,
+      "learning_rate": 0.00013841593258284482,
+      "loss": 0.1511,
+      "step": 1734
+    },
+    {
+      "epoch": 0.61,
+      "learning_rate": 0.00013819660112501054,
+      "loss": 0.5006,
+      "step": 1735
+    },
+    {
+      "epoch": 0.61,
+      "learning_rate": 0.00013797735181178092,
+      "loss": 0.3277,
+      "step": 1736
+    },
+    {
+      "epoch": 0.61,
+      "learning_rate": 0.0001377581849345662,
+      "loss": 0.3718,
+      "step": 1737
+    },
+    {
+      "epoch": 0.62,
+      "learning_rate": 0.0001375391007846673,
+      "loss": 0.4065,
+      "step": 1738
+    },
+    {
+      "epoch": 0.62,
+      "learning_rate": 0.0001373200996532749,
+      "loss": 0.5566,
+      "step": 1739
+    },
+    {
+      "epoch": 0.62,
+      "learning_rate": 0.00013710118183146944,
+      "loss": 0.6928,
+      "step": 1740
+    },
+    {
+      "epoch": 0.62,
+      "eval_loss": 0.5433302521705627,
+      "eval_runtime": 52.9369,
+      "eval_samples_per_second": 30.773,
+      "eval_steps_per_second": 30.773,
+      "step": 1740
+    },
+    {
+      "epoch": 0.62,
+      "learning_rate": 0.00013688234761022073,
+      "loss": 0.4263,
+      "step": 1741
+    },
+    {
+      "epoch": 0.62,
+      "learning_rate": 0.00013666359728038732,
+      "loss": 0.5534,
+      "step": 1742
+    },
+    {
+      "epoch": 0.62,
+      "learning_rate": 0.00013644493113271628,
+      "loss": 0.8234,
+      "step": 1743
+    },
+    {
+      "epoch": 0.62,
+      "learning_rate": 0.0001362263494578429,
+      "loss": 0.4592,
+      "step": 1744
+    },
+    {
+      "epoch": 0.62,
+      "learning_rate": 0.0001360078525462901,
+      "loss": 0.5484,
+      "step": 1745
+    },
+    {
+      "epoch": 0.62,
+      "learning_rate": 0.00013578944068846823,
+      "loss": 0.556,
+      "step": 1746
+    },
+    {
+      "epoch": 0.62,
+      "learning_rate": 0.0001355711141746743,
+      "loss": 0.3097,
+      "step": 1747
+    },
+    {
+      "epoch": 0.62,
+      "learning_rate": 0.0001353528732950924,
+      "loss": 0.3167,
+      "step": 1748
+    },
+    {
+      "epoch": 0.62,
+      "learning_rate": 0.00013513471833979232,
+      "loss": 0.4893,
+      "step": 1749
+    },
+    {
+      "epoch": 0.62,
+      "learning_rate": 0.00013491664959872982,
+      "loss": 0.4819,
+      "step": 1750
+    },
+    {
+      "epoch": 0.62,
+      "learning_rate": 0.00013469866736174626,
+      "loss": 0.5817,
+      "step": 1751
+    },
+    {
+      "epoch": 0.62,
+      "learning_rate": 0.0001344807719185677,
+      "loss": 0.5153,
+      "step": 1752
+    },
+    {
+      "epoch": 0.62,
+      "learning_rate": 0.00013426296355880502,
+      "loss": 0.2236,
+      "step": 1753
+    },
+    {
+      "epoch": 0.62,
+      "learning_rate": 0.0001340452425719534,
+      "loss": 0.8153,
+      "step": 1754
+    },
+    {
+      "epoch": 0.62,
+      "learning_rate": 0.00013382760924739176,
+      "loss": 0.3157,
+      "step": 1755
+    },
+    {
+      "epoch": 0.62,
+      "learning_rate": 0.00013361006387438258,
+      "loss": 0.2784,
+      "step": 1756
+    },
+    {
+      "epoch": 0.62,
+      "learning_rate": 0.0001333926067420713,
+      "loss": 0.3701,
+      "step": 1757
+    },
+    {
+      "epoch": 0.62,
+      "learning_rate": 0.00013317523813948648,
+      "loss": 0.4396,
+      "step": 1758
+    },
+    {
+      "epoch": 0.62,
+      "learning_rate": 0.0001329579583555385,
+      "loss": 1.032,
+      "step": 1759
+    },
+    {
+      "epoch": 0.62,
+      "learning_rate": 0.00013274076767901994,
+      "loss": 0.3776,
+      "step": 1760
+    },
+    {
+      "epoch": 0.62,
+      "eval_loss": 0.5506594777107239,
+      "eval_runtime": 52.9375,
+      "eval_samples_per_second": 30.772,
+      "eval_steps_per_second": 30.772,
+      "step": 1760
+    },
+    {
+      "epoch": 0.62,
+      "learning_rate": 0.00013252366639860506,
+      "loss": 0.2526,
+      "step": 1761
+    },
+    {
+      "epoch": 0.62,
+      "learning_rate": 0.00013230665480284907,
+      "loss": 0.4091,
+      "step": 1762
+    },
+    {
+      "epoch": 0.62,
+      "learning_rate": 0.00013208973318018806,
+      "loss": 0.2905,
+      "step": 1763
+    },
+    {
+      "epoch": 0.62,
+      "learning_rate": 0.00013187290181893865,
+      "loss": 0.5923,
+      "step": 1764
+    },
+    {
+      "epoch": 0.62,
+      "learning_rate": 0.00013165616100729727,
+      "loss": 0.4891,
+      "step": 1765
+    },
+    {
+      "epoch": 0.63,
+      "learning_rate": 0.00013143951103334028,
+      "loss": 0.5123,
+      "step": 1766
+    },
+    {
+      "epoch": 0.63,
+      "learning_rate": 0.00013122295218502287,
+      "loss": 0.3186,
+      "step": 1767
+    },
+    {
+      "epoch": 0.63,
+      "learning_rate": 0.00013100648475017972,
+      "loss": 0.3733,
+      "step": 1768
+    },
+    {
+      "epoch": 0.63,
+      "learning_rate": 0.00013079010901652347,
+      "loss": 0.4298,
+      "step": 1769
+    },
+    {
+      "epoch": 0.63,
+      "learning_rate": 0.0001305738252716451,
+      "loss": 0.3822,
+      "step": 1770
+    },
+    {
+      "epoch": 0.63,
+      "learning_rate": 0.00013035763380301333,
+      "loss": 0.3157,
+      "step": 1771
+    },
+    {
+      "epoch": 0.63,
+      "learning_rate": 0.0001301415348979743,
+      "loss": 0.6844,
+      "step": 1772
+    },
+    {
+      "epoch": 0.63,
+      "learning_rate": 0.00012992552884375086,
+      "loss": 0.5713,
+      "step": 1773
+    },
+    {
+      "epoch": 0.63,
+      "learning_rate": 0.00012970961592744283,
+      "loss": 0.4232,
+      "step": 1774
+    },
+    {
+      "epoch": 0.63,
+      "learning_rate": 0.0001294937964360259,
+      "loss": 0.4354,
+      "step": 1775
+    },
+    {
+      "epoch": 0.63,
+      "learning_rate": 0.00012927807065635173,
+      "loss": 0.2787,
+      "step": 1776
+    },
+    {
+      "epoch": 0.63,
+      "learning_rate": 0.0001290624388751474,
+      "loss": 0.4573,
+      "step": 1777
+    },
+    {
+      "epoch": 0.63,
+      "learning_rate": 0.0001288469013790151,
+      "loss": 0.5611,
+      "step": 1778
+    },
+    {
+      "epoch": 0.63,
+      "learning_rate": 0.0001286314584544317,
+      "loss": 0.3162,
+      "step": 1779
+    },
+    {
+      "epoch": 0.63,
+      "learning_rate": 0.00012841611038774824,
+      "loss": 0.4636,
+      "step": 1780
+    },
+    {
+      "epoch": 0.63,
+      "eval_loss": 0.5442768335342407,
+      "eval_runtime": 52.9414,
+      "eval_samples_per_second": 30.77,
+      "eval_steps_per_second": 30.77,
+      "step": 1780
+    },
+    {
+      "epoch": 0.63,
+      "learning_rate": 0.0001282008574651899,
+      "loss": 0.6031,
+      "step": 1781
+    },
+    {
+      "epoch": 0.63,
+      "learning_rate": 0.0001279856999728552,
+      "loss": 0.546,
+      "step": 1782
+    },
+    {
+      "epoch": 0.63,
+      "learning_rate": 0.00012777063819671594,
+      "loss": 0.7327,
+      "step": 1783
+    },
+    {
+      "epoch": 0.63,
+      "learning_rate": 0.00012755567242261667,
+      "loss": 0.3747,
+      "step": 1784
+    },
+    {
+      "epoch": 0.63,
+      "learning_rate": 0.00012734080293627442,
+      "loss": 0.3859,
+      "step": 1785
+    },
+    {
+      "epoch": 0.63,
+      "learning_rate": 0.000127126030023278,
+      "loss": 0.4273,
+      "step": 1786
+    },
+    {
+      "epoch": 0.63,
+      "learning_rate": 0.0001269113539690881,
+      "loss": 0.4171,
+      "step": 1787
+    },
+    {
+      "epoch": 0.63,
+      "learning_rate": 0.0001266967750590366,
+      "loss": 0.5345,
+      "step": 1788
+    },
+    {
+      "epoch": 0.63,
+      "learning_rate": 0.00012648229357832623,
+      "loss": 0.6726,
+      "step": 1789
+    },
+    {
+      "epoch": 0.63,
+      "learning_rate": 0.00012626790981203024,
+      "loss": 0.4437,
+      "step": 1790
+    },
+    {
+      "epoch": 0.63,
+      "learning_rate": 0.0001260536240450921,
+      "loss": 0.351,
+      "step": 1791
+    },
+    {
+      "epoch": 0.63,
+      "learning_rate": 0.00012583943656232488,
+      "loss": 0.4259,
+      "step": 1792
+    },
+    {
+      "epoch": 0.63,
+      "learning_rate": 0.00012562534764841104,
+      "loss": 0.8814,
+      "step": 1793
+    },
+    {
+      "epoch": 0.63,
+      "learning_rate": 0.00012541135758790216,
+      "loss": 0.378,
+      "step": 1794
+    },
+    {
+      "epoch": 0.64,
+      "learning_rate": 0.00012519746666521834,
+      "loss": 0.55,
+      "step": 1795
+    },
+    {
+      "epoch": 0.64,
+      "learning_rate": 0.00012498367516464783,
+      "loss": 0.2484,
+      "step": 1796
+    },
+    {
+      "epoch": 0.64,
+      "learning_rate": 0.00012476998337034682,
+      "loss": 0.376,
+      "step": 1797
+    },
+    {
+      "epoch": 0.64,
+      "learning_rate": 0.00012455639156633904,
+      "loss": 0.6253,
+      "step": 1798
+    },
+    {
+      "epoch": 0.64,
+      "learning_rate": 0.00012434290003651526,
+      "loss": 0.4784,
+      "step": 1799
+    },
+    {
+      "epoch": 0.64,
+      "learning_rate": 0.00012412950906463287,
+      "loss": 0.43,
+      "step": 1800
+    },
+    {
+      "epoch": 0.64,
+      "eval_loss": 0.5527411699295044,
+      "eval_runtime": 52.9394,
+      "eval_samples_per_second": 30.771,
+      "eval_steps_per_second": 30.771,
+      "step": 1800
+    },
+    {
+      "epoch": 0.64,
+      "learning_rate": 0.00012391621893431584,
+      "loss": 0.6789,
+      "step": 1801
+    },
+    {
+      "epoch": 0.64,
+      "learning_rate": 0.00012370302992905388,
+      "loss": 0.5521,
+      "step": 1802
+    },
+    {
+      "epoch": 0.64,
+      "learning_rate": 0.0001234899423322024,
+      "loss": 0.476,
+      "step": 1803
+    },
+    {
+      "epoch": 0.64,
+      "learning_rate": 0.0001232769564269821,
+      "loss": 1.1217,
+      "step": 1804
+    },
+    {
+      "epoch": 0.64,
+      "learning_rate": 0.00012306407249647834,
+      "loss": 0.4226,
+      "step": 1805
+    },
+    {
+      "epoch": 0.64,
+      "learning_rate": 0.00012285129082364104,
+      "loss": 0.6088,
+      "step": 1806
+    },
+    {
+      "epoch": 0.64,
+      "learning_rate": 0.0001226386116912842,
+      "loss": 0.9425,
+      "step": 1807
+    },
+    {
+      "epoch": 0.64,
+      "learning_rate": 0.00012242603538208552,
+      "loss": 0.4274,
+      "step": 1808
+    },
+    {
+      "epoch": 0.64,
+      "learning_rate": 0.00012221356217858608,
+      "loss": 0.7362,
+      "step": 1809
+    },
+    {
+      "epoch": 0.64,
+      "learning_rate": 0.00012200119236318986,
+      "loss": 0.6808,
+      "step": 1810
+    },
+    {
+      "epoch": 0.64,
+      "learning_rate": 0.00012178892621816347,
+      "loss": 0.6089,
+      "step": 1811
+    },
+    {
+      "epoch": 0.64,
+      "learning_rate": 0.00012157676402563569,
+      "loss": 0.4028,
+      "step": 1812
+    },
+    {
+      "epoch": 0.64,
+      "learning_rate": 0.00012136470606759713,
+      "loss": 0.9581,
+      "step": 1813
+    },
+    {
+      "epoch": 0.64,
+      "learning_rate": 0.00012115275262589994,
+      "loss": 0.286,
+      "step": 1814
+    },
+    {
+      "epoch": 0.64,
+      "learning_rate": 0.00012094090398225729,
+      "loss": 0.7448,
+      "step": 1815
+    },
+    {
+      "epoch": 0.64,
+      "learning_rate": 0.00012072916041824301,
+      "loss": 0.5804,
+      "step": 1816
+    },
+    {
+      "epoch": 0.64,
+      "learning_rate": 0.00012051752221529127,
+      "loss": 1.0029,
+      "step": 1817
+    },
+    {
+      "epoch": 0.64,
+      "learning_rate": 0.00012030598965469638,
+      "loss": 0.3921,
+      "step": 1818
+    },
+    {
+      "epoch": 0.64,
+      "learning_rate": 0.00012009456301761206,
+      "loss": 0.4126,
+      "step": 1819
+    },
+    {
+      "epoch": 0.64,
+      "learning_rate": 0.00011988324258505121,
+      "loss": 0.5656,
+      "step": 1820
+    },
+    {
+      "epoch": 0.64,
+      "eval_loss": 0.5478140115737915,
+      "eval_runtime": 52.943,
+      "eval_samples_per_second": 30.769,
+      "eval_steps_per_second": 30.769,
+      "step": 1820
+    },
+    {
+      "epoch": 0.64,
+      "learning_rate": 0.00011967202863788582,
+      "loss": 0.3112,
+      "step": 1821
+    },
+    {
+      "epoch": 0.64,
+      "learning_rate": 0.00011946092145684611,
+      "loss": 0.4343,
+      "step": 1822
+    },
+    {
+      "epoch": 0.65,
+      "learning_rate": 0.00011924992132252035,
+      "loss": 0.5407,
+      "step": 1823
+    },
+    {
+      "epoch": 0.65,
+      "learning_rate": 0.00011903902851535483,
+      "loss": 0.4383,
+      "step": 1824
+    },
+    {
+      "epoch": 0.65,
+      "learning_rate": 0.0001188282433156529,
+      "loss": 0.4131,
+      "step": 1825
+    },
+    {
+      "epoch": 0.65,
+      "learning_rate": 0.00011861756600357502,
+      "loss": 0.1918,
+      "step": 1826
+    },
+    {
+      "epoch": 0.65,
+      "learning_rate": 0.00011840699685913818,
+      "loss": 0.4841,
+      "step": 1827
+    },
+    {
+      "epoch": 0.65,
+      "learning_rate": 0.0001181965361622157,
+      "loss": 0.3759,
+      "step": 1828
+    },
+    {
+      "epoch": 0.65,
+      "learning_rate": 0.0001179861841925367,
+      "loss": 0.5492,
+      "step": 1829
+    },
+    {
+      "epoch": 0.65,
+      "learning_rate": 0.00011777594122968576,
+      "loss": 0.7391,
+      "step": 1830
+    },
+    {
+      "epoch": 0.65,
+      "learning_rate": 0.00011756580755310267,
+      "loss": 0.4276,
+      "step": 1831
+    },
+    {
+      "epoch": 0.65,
+      "learning_rate": 0.0001173557834420819,
+      "loss": 0.3919,
+      "step": 1832
+    },
+    {
+      "epoch": 0.65,
+      "learning_rate": 0.00011714586917577213,
+      "loss": 0.5649,
+      "step": 1833
+    },
+    {
+      "epoch": 0.65,
+      "learning_rate": 0.00011693606503317647,
+      "loss": 0.5007,
+      "step": 1834
+    },
+    {
+      "epoch": 0.65,
+      "learning_rate": 0.00011672637129315122,
+      "loss": 0.47,
+      "step": 1835
+    },
+    {
+      "epoch": 0.65,
+      "learning_rate": 0.00011651678823440629,
+      "loss": 0.3533,
+      "step": 1836
+    },
+    {
+      "epoch": 0.65,
+      "learning_rate": 0.00011630731613550403,
+      "loss": 0.5038,
+      "step": 1837
+    },
+    {
+      "epoch": 0.65,
+      "learning_rate": 0.00011609795527485985,
+      "loss": 0.5176,
+      "step": 1838
+    },
+    {
+      "epoch": 0.65,
+      "learning_rate": 0.00011588870593074097,
+      "loss": 0.4364,
+      "step": 1839
+    },
+    {
+      "epoch": 0.65,
+      "learning_rate": 0.00011567956838126645,
+      "loss": 0.729,
+      "step": 1840
+    },
+    {
+      "epoch": 0.65,
+      "eval_loss": 0.5541649460792542,
+      "eval_runtime": 52.9378,
+      "eval_samples_per_second": 30.772,
+      "eval_steps_per_second": 30.772,
+      "step": 1840
+    },
+    {
+      "epoch": 0.65,
+      "learning_rate": 0.00011547054290440682,
+      "loss": 0.6639,
+      "step": 1841
+    },
+    {
+      "epoch": 0.65,
+      "learning_rate": 0.00011526162977798361,
+      "loss": 0.4572,
+      "step": 1842
+    },
+    {
+      "epoch": 0.65,
+      "learning_rate": 0.00011505282927966891,
+      "loss": 0.639,
+      "step": 1843
+    },
+    {
+      "epoch": 0.65,
+      "learning_rate": 0.00011484414168698547,
+      "loss": 0.4153,
+      "step": 1844
+    },
+    {
+      "epoch": 0.65,
+      "learning_rate": 0.00011463556727730555,
+      "loss": 0.373,
+      "step": 1845
+    },
+    {
+      "epoch": 0.65,
+      "learning_rate": 0.00011442710632785132,
+      "loss": 0.3626,
+      "step": 1846
+    },
+    {
+      "epoch": 0.65,
+      "learning_rate": 0.00011421875911569371,
+      "loss": 0.3548,
+      "step": 1847
+    },
+    {
+      "epoch": 0.65,
+      "learning_rate": 0.00011401052591775299,
+      "loss": 0.4508,
+      "step": 1848
+    },
+    {
+      "epoch": 0.65,
+      "learning_rate": 0.0001138024070107976,
+      "loss": 0.8129,
+      "step": 1849
+    },
+    {
+      "epoch": 0.65,
+      "learning_rate": 0.00011359440267144412,
+      "loss": 0.2444,
+      "step": 1850
+    },
+    {
+      "epoch": 0.66,
+      "learning_rate": 0.00011338651317615684,
+      "loss": 0.4259,
+      "step": 1851
+    },
+    {
+      "epoch": 0.66,
+      "learning_rate": 0.00011317873880124745,
+      "loss": 0.5604,
+      "step": 1852
+    },
+    {
+      "epoch": 0.66,
+      "learning_rate": 0.00011297107982287456,
+      "loss": 0.8979,
+      "step": 1853
+    },
+    {
+      "epoch": 0.66,
+      "learning_rate": 0.00011276353651704354,
+      "loss": 0.2095,
+      "step": 1854
+    },
+    {
+      "epoch": 0.66,
+      "learning_rate": 0.00011255610915960595,
+      "loss": 0.5063,
+      "step": 1855
+    },
+    {
+      "epoch": 0.66,
+      "learning_rate": 0.0001123487980262592,
+      "loss": 0.5512,
+      "step": 1856
+    },
+    {
+      "epoch": 0.66,
+      "learning_rate": 0.00011214160339254607,
+      "loss": 0.3916,
+      "step": 1857
+    },
+    {
+      "epoch": 0.66,
+      "learning_rate": 0.00011193452553385491,
+      "loss": 0.5775,
+      "step": 1858
+    },
+    {
+      "epoch": 0.66,
+      "learning_rate": 0.0001117275647254185,
+      "loss": 0.33,
+      "step": 1859
+    },
+    {
+      "epoch": 0.66,
+      "learning_rate": 0.00011152072124231418,
+      "loss": 0.4355,
+      "step": 1860
+    },
+    {
+      "epoch": 0.66,
+      "eval_loss": 0.5411290526390076,
+      "eval_runtime": 52.9405,
+      "eval_samples_per_second": 30.77,
+      "eval_steps_per_second": 30.77,
+      "step": 1860
+    },
+    {
+      "epoch": 0.66,
+      "learning_rate": 0.00011131399535946339,
+      "loss": 0.3629,
+      "step": 1861
+    },
+    {
+      "epoch": 0.66,
+      "learning_rate": 0.00011110738735163114,
+      "loss": 0.4105,
+      "step": 1862
+    },
+    {
+      "epoch": 0.66,
+      "learning_rate": 0.00011090089749342578,
+      "loss": 0.7106,
+      "step": 1863
+    },
+    {
+      "epoch": 0.66,
+      "learning_rate": 0.00011069452605929886,
+      "loss": 0.839,
+      "step": 1864
+    },
+    {
+      "epoch": 0.66,
+      "learning_rate": 0.0001104882733235443,
+      "loss": 0.68,
+      "step": 1865
+    },
+    {
+      "epoch": 0.66,
+      "learning_rate": 0.00011028213956029833,
+      "loss": 0.1928,
+      "step": 1866
+    },
+    {
+      "epoch": 0.66,
+      "learning_rate": 0.00011007612504353882,
+      "loss": 0.5412,
+      "step": 1867
+    },
+    {
+      "epoch": 0.66,
+      "learning_rate": 0.00010987023004708559,
+      "loss": 0.2842,
+      "step": 1868
+    },
+    {
+      "epoch": 0.66,
+      "learning_rate": 0.00010966445484459928,
+      "loss": 0.4469,
+      "step": 1869
+    },
+    {
+      "epoch": 0.66,
+      "learning_rate": 0.00010945879970958141,
+      "loss": 0.4715,
+      "step": 1870
+    },
+    {
+      "epoch": 0.66,
+      "learning_rate": 0.0001092532649153739,
+      "loss": 0.3671,
+      "step": 1871
+    },
+    {
+      "epoch": 0.66,
+      "learning_rate": 0.00010904785073515871,
+      "loss": 0.3854,
+      "step": 1872
+    },
+    {
+      "epoch": 0.66,
+      "learning_rate": 0.00010884255744195746,
+      "loss": 0.3962,
+      "step": 1873
+    },
+    {
+      "epoch": 0.66,
+      "learning_rate": 0.00010863738530863123,
+      "loss": 0.2672,
+      "step": 1874
+    },
+    {
+      "epoch": 0.66,
+      "learning_rate": 0.00010843233460788,
+      "loss": 0.6842,
+      "step": 1875
+    },
+    {
+      "epoch": 0.66,
+      "learning_rate": 0.00010822740561224218,
+      "loss": 0.3248,
+      "step": 1876
+    },
+    {
+      "epoch": 0.66,
+      "learning_rate": 0.00010802259859409454,
+      "loss": 0.4767,
+      "step": 1877
+    },
+    {
+      "epoch": 0.66,
+      "learning_rate": 0.00010781791382565192,
+      "loss": 0.3836,
+      "step": 1878
+    },
+    {
+      "epoch": 0.67,
+      "learning_rate": 0.00010761335157896642,
+      "loss": 0.3545,
+      "step": 1879
+    },
+    {
+      "epoch": 0.67,
+      "learning_rate": 0.00010740891212592735,
+      "loss": 0.377,
+      "step": 1880
+    },
+    {
+      "epoch": 0.67,
+      "eval_loss": 0.5426122546195984,
+      "eval_runtime": 52.9795,
+      "eval_samples_per_second": 30.748,
+      "eval_steps_per_second": 30.748,
+      "step": 1880
+    },
+    {
+      "epoch": 0.67,
+      "learning_rate": 0.00010720459573826083,
+      "loss": 0.3347,
+      "step": 1881
+    },
+    {
+      "epoch": 0.67,
+      "learning_rate": 0.00010700040268752944,
+      "loss": 0.3557,
+      "step": 1882
+    },
+    {
+      "epoch": 0.67,
+      "learning_rate": 0.00010679633324513174,
+      "loss": 0.397,
+      "step": 1883
+    },
+    {
+      "epoch": 0.67,
+      "learning_rate": 0.00010659238768230215,
+      "loss": 0.5991,
+      "step": 1884
+    },
+    {
+      "epoch": 0.67,
+      "learning_rate": 0.00010638856627011045,
+      "loss": 0.6608,
+      "step": 1885
+    },
+    {
+      "epoch": 0.67,
+      "learning_rate": 0.00010618486927946107,
+      "loss": 0.5068,
+      "step": 1886
+    },
+    {
+      "epoch": 0.67,
+      "learning_rate": 0.00010598129698109334,
+      "loss": 0.4619,
+      "step": 1887
+    },
+    {
+      "epoch": 0.67,
+      "learning_rate": 0.000105777849645581,
+      "loss": 0.4624,
+      "step": 1888
+    },
+    {
+      "epoch": 0.67,
+      "learning_rate": 0.0001055745275433314,
+      "loss": 0.4252,
+      "step": 1889
+    },
+    {
+      "epoch": 0.67,
+      "learning_rate": 0.0001053713309445856,
+      "loss": 0.5567,
+      "step": 1890
+    },
+    {
+      "epoch": 0.67,
+      "learning_rate": 0.0001051682601194178,
+      "loss": 0.4753,
+      "step": 1891
+    },
+    {
+      "epoch": 0.67,
+      "learning_rate": 0.00010496531533773503,
+      "loss": 0.3621,
+      "step": 1892
+    },
+    {
+      "epoch": 0.67,
+      "learning_rate": 0.00010476249686927677,
+      "loss": 0.3746,
+      "step": 1893
+    },
+    {
+      "epoch": 0.67,
+      "learning_rate": 0.00010455980498361473,
+      "loss": 0.658,
+      "step": 1894
+    },
+    {
+      "epoch": 0.67,
+      "learning_rate": 0.00010435723995015236,
+      "loss": 0.4619,
+      "step": 1895
+    },
+    {
+      "epoch": 0.67,
+      "learning_rate": 0.00010415480203812429,
+      "loss": 0.4662,
+      "step": 1896
+    },
+    {
+      "epoch": 0.67,
+      "learning_rate": 0.0001039524915165963,
+      "loss": 0.7505,
+      "step": 1897
+    },
+    {
+      "epoch": 0.67,
+      "learning_rate": 0.00010375030865446507,
+      "loss": 0.5952,
+      "step": 1898
+    },
+    {
+      "epoch": 0.67,
+      "learning_rate": 0.00010354825372045731,
+      "loss": 1.2613,
+      "step": 1899
+    },
+    {
+      "epoch": 0.67,
+      "learning_rate": 0.0001033463269831299,
+      "loss": 0.5345,
+      "step": 1900
+    },
+    {
+      "epoch": 0.67,
+      "eval_loss": 0.543377161026001,
+      "eval_runtime": 52.9382,
+      "eval_samples_per_second": 30.772,
+      "eval_steps_per_second": 30.772,
+      "step": 1900
+    },
+    {
+      "epoch": 0.67,
+      "learning_rate": 0.00010314452871086913,
+      "loss": 0.3847,
+      "step": 1901
+    },
+    {
+      "epoch": 0.67,
+      "learning_rate": 0.00010294285917189076,
+      "loss": 0.4378,
+      "step": 1902
+    },
+    {
+      "epoch": 0.67,
+      "learning_rate": 0.00010274131863423923,
+      "loss": 0.4684,
+      "step": 1903
+    },
+    {
+      "epoch": 0.67,
+      "learning_rate": 0.00010253990736578777,
+      "loss": 0.3052,
+      "step": 1904
+    },
+    {
+      "epoch": 0.67,
+      "learning_rate": 0.0001023386256342377,
+      "loss": 0.3326,
+      "step": 1905
+    },
+    {
+      "epoch": 0.67,
+      "learning_rate": 0.00010213747370711795,
+      "loss": 0.469,
+      "step": 1906
+    },
+    {
+      "epoch": 0.67,
+      "learning_rate": 0.00010193645185178508,
+      "loss": 0.3558,
+      "step": 1907
+    },
+    {
+      "epoch": 0.68,
+      "learning_rate": 0.000101735560335423,
+      "loss": 0.5909,
+      "step": 1908
+    },
+    {
+      "epoch": 0.68,
+      "learning_rate": 0.00010153479942504205,
+      "loss": 0.8615,
+      "step": 1909
+    },
+    {
+      "epoch": 0.68,
+      "learning_rate": 0.00010133416938747914,
+      "loss": 0.9861,
+      "step": 1910
+    },
+    {
+      "epoch": 0.68,
+      "learning_rate": 0.00010113367048939717,
+      "loss": 0.6974,
+      "step": 1911
+    },
+    {
+      "epoch": 0.68,
+      "learning_rate": 0.00010093330299728476,
+      "loss": 0.4876,
+      "step": 1912
+    },
+    {
+      "epoch": 0.68,
+      "learning_rate": 0.0001007330671774558,
+      "loss": 0.2391,
+      "step": 1913
+    },
+    {
+      "epoch": 0.68,
+      "learning_rate": 0.00010053296329604945,
+      "loss": 0.3748,
+      "step": 1914
+    },
+    {
+      "epoch": 0.68,
+      "learning_rate": 0.00010033299161902925,
+      "loss": 0.5047,
+      "step": 1915
+    },
+    {
+      "epoch": 0.68,
+      "learning_rate": 0.00010013315241218293,
+      "loss": 0.7621,
+      "step": 1916
+    },
+    {
+      "epoch": 0.68,
+      "learning_rate": 9.993344594112233e-05,
+      "loss": 0.7228,
+      "step": 1917
+    },
+    {
+      "epoch": 0.68,
+      "learning_rate": 9.973387247128298e-05,
+      "loss": 0.7847,
+      "step": 1918
+    },
+    {
+      "epoch": 0.68,
+      "learning_rate": 9.95344322679234e-05,
+      "loss": 0.4665,
+      "step": 1919
+    },
+    {
+      "epoch": 0.68,
+      "learning_rate": 9.93351255961251e-05,
+      "loss": 0.4042,
+      "step": 1920
+    },
+    {
+      "epoch": 0.68,
+      "eval_loss": 0.5383409261703491,
+      "eval_runtime": 52.9352,
+      "eval_samples_per_second": 30.773,
+      "eval_steps_per_second": 30.773,
+      "step": 1920
+    },
+    {
+      "epoch": 0.68,
+      "learning_rate": 9.913595272079206e-05,
+      "loss": 0.4262,
+      "step": 1921
+    },
+    {
+      "epoch": 0.68,
+      "learning_rate": 9.893691390665047e-05,
+      "loss": 0.3771,
+      "step": 1922
+    },
+    {
+      "epoch": 0.68,
+      "learning_rate": 9.873800941824822e-05,
+      "loss": 0.6729,
+      "step": 1923
+    },
+    {
+      "epoch": 0.68,
+      "learning_rate": 9.853923951995499e-05,
+      "loss": 0.2703,
+      "step": 1924
+    },
+    {
+      "epoch": 0.68,
+      "learning_rate": 9.834060447596114e-05,
+      "loss": 0.4845,
+      "step": 1925
+    },
+    {
+      "epoch": 0.68,
+      "learning_rate": 9.814210455027803e-05,
+      "loss": 0.5307,
+      "step": 1926
+    },
+    {
+      "epoch": 0.68,
+      "learning_rate": 9.794374000673735e-05,
+      "loss": 0.4081,
+      "step": 1927
+    },
+    {
+      "epoch": 0.68,
+      "learning_rate": 9.774551110899103e-05,
+      "loss": 0.4366,
+      "step": 1928
+    },
+    {
+      "epoch": 0.68,
+      "learning_rate": 9.754741812051051e-05,
+      "loss": 0.3481,
+      "step": 1929
+    },
+    {
+      "epoch": 0.68,
+      "learning_rate": 9.734946130458664e-05,
+      "loss": 0.4935,
+      "step": 1930
+    },
+    {
+      "epoch": 0.68,
+      "learning_rate": 9.715164092432931e-05,
+      "loss": 0.4739,
+      "step": 1931
+    },
+    {
+      "epoch": 0.68,
+      "learning_rate": 9.695395724266703e-05,
+      "loss": 0.423,
+      "step": 1932
+    },
+    {
+      "epoch": 0.68,
+      "learning_rate": 9.675641052234659e-05,
+      "loss": 0.5048,
+      "step": 1933
+    },
+    {
+      "epoch": 0.68,
+      "learning_rate": 9.655900102593301e-05,
+      "loss": 0.4681,
+      "step": 1934
+    },
+    {
+      "epoch": 0.68,
+      "learning_rate": 9.636172901580845e-05,
+      "loss": 0.5774,
+      "step": 1935
+    },
+    {
+      "epoch": 0.69,
+      "learning_rate": 9.616459475417269e-05,
+      "loss": 0.4351,
+      "step": 1936
+    },
+    {
+      "epoch": 0.69,
+      "learning_rate": 9.596759850304226e-05,
+      "loss": 0.2977,
+      "step": 1937
+    },
+    {
+      "epoch": 0.69,
+      "learning_rate": 9.577074052425045e-05,
+      "loss": 0.3905,
+      "step": 1938
+    },
+    {
+      "epoch": 0.69,
+      "learning_rate": 9.557402107944657e-05,
+      "loss": 0.4845,
+      "step": 1939
+    },
+    {
+      "epoch": 0.69,
+      "learning_rate": 9.537744043009588e-05,
+      "loss": 0.3676,
+      "step": 1940
+    },
+    {
+      "epoch": 0.69,
+      "eval_loss": 0.5371923446655273,
+      "eval_runtime": 52.9438,
+      "eval_samples_per_second": 30.768,
+      "eval_steps_per_second": 30.768,
+      "step": 1940
+    },
+    {
+      "epoch": 0.69,
+      "learning_rate": 9.518099883747912e-05,
+      "loss": 0.8065,
+      "step": 1941
+    },
+    {
+      "epoch": 0.69,
+      "learning_rate": 9.498469656269226e-05,
+      "loss": 0.3934,
+      "step": 1942
+    },
+    {
+      "epoch": 0.69,
+      "learning_rate": 9.478853386664599e-05,
+      "loss": 0.6151,
+      "step": 1943
+    },
+    {
+      "epoch": 0.69,
+      "learning_rate": 9.459251101006581e-05,
+      "loss": 0.5769,
+      "step": 1944
+    },
+    {
+      "epoch": 0.69,
+      "learning_rate": 9.439662825349085e-05,
+      "loss": 0.4229,
+      "step": 1945
+    },
+    {
+      "epoch": 0.69,
+      "learning_rate": 9.420088585727443e-05,
+      "loss": 0.5796,
+      "step": 1946
+    },
+    {
+      "epoch": 0.69,
+      "learning_rate": 9.400528408158305e-05,
+      "loss": 1.5279,
+      "step": 1947
+    },
+    {
+      "epoch": 0.69,
+      "learning_rate": 9.380982318639657e-05,
+      "loss": 0.5703,
+      "step": 1948
+    },
+    {
+      "epoch": 0.69,
+      "learning_rate": 9.361450343150746e-05,
+      "loss": 0.5642,
+      "step": 1949
+    },
+    {
+      "epoch": 0.69,
+      "learning_rate": 9.341932507652053e-05,
+      "loss": 0.4181,
+      "step": 1950
+    },
+    {
+      "epoch": 0.69,
+      "learning_rate": 9.322428838085278e-05,
+      "loss": 0.316,
+      "step": 1951
+    },
+    {
+      "epoch": 0.69,
+      "learning_rate": 9.302939360373281e-05,
+      "loss": 0.5275,
+      "step": 1952
+    },
+    {
+      "epoch": 0.69,
+      "learning_rate": 9.283464100420063e-05,
+      "loss": 0.7782,
+      "step": 1953
+    },
+    {
+      "epoch": 0.69,
+      "learning_rate": 9.264003084110751e-05,
+      "loss": 0.2878,
+      "step": 1954
+    },
+    {
+      "epoch": 0.69,
+      "learning_rate": 9.244556337311496e-05,
+      "loss": 0.4412,
+      "step": 1955
+    },
+    {
+      "epoch": 0.69,
+      "learning_rate": 9.225123885869513e-05,
+      "loss": 0.3984,
+      "step": 1956
+    },
+    {
+      "epoch": 0.69,
+      "learning_rate": 9.205705755613003e-05,
+      "loss": 0.2531,
+      "step": 1957
+    },
+    {
+      "epoch": 0.69,
+      "learning_rate": 9.186301972351157e-05,
+      "loss": 0.5272,
+      "step": 1958
+    },
+    {
+      "epoch": 0.69,
+      "learning_rate": 9.16691256187407e-05,
+      "loss": 0.5925,
+      "step": 1959
+    },
+    {
+      "epoch": 0.69,
+      "learning_rate": 9.147537549952743e-05,
+      "loss": 0.4758,
+      "step": 1960
+    },
+    {
+      "epoch": 0.69,
+      "eval_loss": 0.5411304235458374,
+      "eval_runtime": 52.9384,
+      "eval_samples_per_second": 30.772,
+      "eval_steps_per_second": 30.772,
+      "step": 1960
+    },
+    {
+      "epoch": 0.69,
+      "learning_rate": 9.128176962339036e-05,
+      "loss": 0.4649,
+      "step": 1961
+    },
+    {
+      "epoch": 0.69,
+      "learning_rate": 9.108830824765648e-05,
+      "loss": 0.4571,
+      "step": 1962
+    },
+    {
+      "epoch": 0.69,
+      "learning_rate": 9.089499162946051e-05,
+      "loss": 0.575,
+      "step": 1963
+    },
+    {
+      "epoch": 0.7,
+      "learning_rate": 9.070182002574518e-05,
+      "loss": 0.4608,
+      "step": 1964
+    },
+    {
+      "epoch": 0.7,
+      "learning_rate": 9.050879369325995e-05,
+      "loss": 0.5249,
+      "step": 1965
+    },
+    {
+      "epoch": 0.7,
+      "learning_rate": 9.031591288856152e-05,
+      "loss": 0.5472,
+      "step": 1966
+    },
+    {
+      "epoch": 0.7,
+      "learning_rate": 9.012317786801305e-05,
+      "loss": 0.2289,
+      "step": 1967
+    },
+    {
+      "epoch": 0.7,
+      "learning_rate": 8.993058888778412e-05,
+      "loss": 0.4431,
+      "step": 1968
+    },
+    {
+      "epoch": 0.7,
+      "learning_rate": 8.973814620385004e-05,
+      "loss": 0.3248,
+      "step": 1969
+    },
+    {
+      "epoch": 0.7,
+      "learning_rate": 8.954585007199147e-05,
+      "loss": 0.3366,
+      "step": 1970
+    },
+    {
+      "epoch": 0.7,
+      "learning_rate": 8.935370074779479e-05,
+      "loss": 0.4847,
+      "step": 1971
+    },
+    {
+      "epoch": 0.7,
+      "learning_rate": 8.916169848665085e-05,
+      "loss": 0.9995,
+      "step": 1972
+    },
+    {
+      "epoch": 0.7,
+      "learning_rate": 8.89698435437552e-05,
+      "loss": 0.3843,
+      "step": 1973
+    },
+    {
+      "epoch": 0.7,
+      "learning_rate": 8.877813617410748e-05,
+      "loss": 0.4822,
+      "step": 1974
+    },
+    {
+      "epoch": 0.7,
+      "learning_rate": 8.858657663251131e-05,
+      "loss": 0.4256,
+      "step": 1975
+    },
+    {
+      "epoch": 0.7,
+      "learning_rate": 8.839516517357374e-05,
+      "loss": 0.3367,
+      "step": 1976
+    },
+    {
+      "epoch": 0.7,
+      "learning_rate": 8.820390205170495e-05,
+      "loss": 0.5951,
+      "step": 1977
+    },
+    {
+      "epoch": 0.7,
+      "learning_rate": 8.80127875211182e-05,
+      "loss": 0.3253,
+      "step": 1978
+    },
+    {
+      "epoch": 0.7,
+      "learning_rate": 8.78218218358291e-05,
+      "loss": 0.6211,
+      "step": 1979
+    },
+    {
+      "epoch": 0.7,
+      "learning_rate": 8.763100524965518e-05,
+      "loss": 0.4919,
+      "step": 1980
+    },
+    {
+      "epoch": 0.7,
+      "eval_loss": 0.5353367924690247,
+      "eval_runtime": 52.9401,
+      "eval_samples_per_second": 30.771,
+      "eval_steps_per_second": 30.771,
+      "step": 1980
+    },
+    {
+      "epoch": 0.7,
+      "learning_rate": 8.744033801621624e-05,
+      "loss": 0.5549,
+      "step": 1981
+    },
+    {
+      "epoch": 0.7,
+      "learning_rate": 8.724982038893335e-05,
+      "loss": 0.6219,
+      "step": 1982
+    },
+    {
+      "epoch": 0.7,
+      "learning_rate": 8.70594526210287e-05,
+      "loss": 0.5022,
+      "step": 1983
+    },
+    {
+      "epoch": 0.7,
+      "learning_rate": 8.686923496552535e-05,
+      "loss": 0.4203,
+      "step": 1984
+    },
+    {
+      "epoch": 0.7,
+      "learning_rate": 8.667916767524688e-05,
+      "loss": 0.4764,
+      "step": 1985
+    },
+    {
+      "epoch": 0.7,
+      "learning_rate": 8.648925100281692e-05,
+      "loss": 0.4204,
+      "step": 1986
+    },
+    {
+      "epoch": 0.7,
+      "learning_rate": 8.629948520065896e-05,
+      "loss": 0.2961,
+      "step": 1987
+    },
+    {
+      "epoch": 0.7,
+      "learning_rate": 8.610987052099607e-05,
+      "loss": 0.7589,
+      "step": 1988
+    },
+    {
+      "epoch": 0.7,
+      "learning_rate": 8.592040721585035e-05,
+      "loss": 0.4692,
+      "step": 1989
+    },
+    {
+      "epoch": 0.7,
+      "learning_rate": 8.57310955370425e-05,
+      "loss": 0.439,
+      "step": 1990
+    },
+    {
+      "epoch": 0.7,
+      "learning_rate": 8.554193573619216e-05,
+      "loss": 0.4118,
+      "step": 1991
+    },
+    {
+      "epoch": 0.71,
+      "learning_rate": 8.535292806471673e-05,
+      "loss": 0.5114,
+      "step": 1992
+    },
+    {
+      "epoch": 0.71,
+      "learning_rate": 8.516407277383156e-05,
+      "loss": 0.3324,
+      "step": 1993
+    },
+    {
+      "epoch": 0.71,
+      "learning_rate": 8.497537011454939e-05,
+      "loss": 0.4603,
+      "step": 1994
+    },
+    {
+      "epoch": 0.71,
+      "learning_rate": 8.478682033768017e-05,
+      "loss": 0.9736,
+      "step": 1995
+    },
+    {
+      "epoch": 0.71,
+      "learning_rate": 8.459842369383062e-05,
+      "loss": 0.4285,
+      "step": 1996
+    },
+    {
+      "epoch": 0.71,
+      "learning_rate": 8.441018043340385e-05,
+      "loss": 0.5847,
+      "step": 1997
+    },
+    {
+      "epoch": 0.71,
+      "learning_rate": 8.422209080659928e-05,
+      "loss": 0.4782,
+      "step": 1998
+    },
+    {
+      "epoch": 0.71,
+      "learning_rate": 8.40341550634121e-05,
+      "loss": 0.8181,
+      "step": 1999
+    },
+    {
+      "epoch": 0.71,
+      "learning_rate": 8.384637345363263e-05,
+      "loss": 0.2312,
+      "step": 2000
+    },
+    {
+      "epoch": 0.71,
+      "eval_loss": 0.5351304411888123,
+      "eval_runtime": 52.9375,
+      "eval_samples_per_second": 30.772,
+      "eval_steps_per_second": 30.772,
+      "step": 2000
+    },
+    {
+      "epoch": 0.71,
+      "learning_rate": 8.36587462268468e-05,
+      "loss": 0.5259,
+      "step": 2001
+    },
+    {
+      "epoch": 0.71,
+      "learning_rate": 8.347127363243513e-05,
+      "loss": 0.7685,
+      "step": 2002
+    },
+    {
+      "epoch": 0.71,
+      "learning_rate": 8.328395591957254e-05,
+      "loss": 0.3851,
+      "step": 2003
+    },
+    {
+      "epoch": 0.71,
+      "learning_rate": 8.309679333722826e-05,
+      "loss": 0.6608,
+      "step": 2004
+    },
+    {
+      "epoch": 0.71,
+      "learning_rate": 8.29097861341652e-05,
+      "loss": 0.8655,
+      "step": 2005
+    },
+    {
+      "epoch": 0.71,
+      "learning_rate": 8.27229345589398e-05,
+      "loss": 0.505,
+      "step": 2006
+    },
+    {
+      "epoch": 0.71,
+      "learning_rate": 8.253623885990161e-05,
+      "loss": 0.2962,
+      "step": 2007
+    },
+    {
+      "epoch": 0.71,
+      "learning_rate": 8.234969928519314e-05,
+      "loss": 0.7488,
+      "step": 2008
+    },
+    {
+      "epoch": 0.71,
+      "learning_rate": 8.216331608274932e-05,
+      "loss": 0.9103,
+      "step": 2009
+    },
+    {
+      "epoch": 0.71,
+      "learning_rate": 8.197708950029699e-05,
+      "loss": 0.7655,
+      "step": 2010
+    },
+    {
+      "epoch": 0.71,
+      "learning_rate": 8.179101978535526e-05,
+      "loss": 0.4715,
+      "step": 2011
+    },
+    {
+      "epoch": 0.71,
+      "learning_rate": 8.160510718523446e-05,
+      "loss": 0.5252,
+      "step": 2012
+    },
+    {
+      "epoch": 0.71,
+      "learning_rate": 8.141935194703616e-05,
+      "loss": 0.6784,
+      "step": 2013
+    },
+    {
+      "epoch": 0.71,
+      "learning_rate": 8.123375431765279e-05,
+      "loss": 0.6038,
+      "step": 2014
+    },
+    {
+      "epoch": 0.71,
+      "learning_rate": 8.104831454376727e-05,
+      "loss": 0.7442,
+      "step": 2015
+    },
+    {
+      "epoch": 0.71,
+      "learning_rate": 8.086303287185273e-05,
+      "loss": 0.3276,
+      "step": 2016
+    },
+    {
+      "epoch": 0.71,
+      "learning_rate": 8.067790954817208e-05,
+      "loss": 0.395,
+      "step": 2017
+    },
+    {
+      "epoch": 0.71,
+      "learning_rate": 8.049294481877807e-05,
+      "loss": 0.3629,
+      "step": 2018
+    },
+    {
+      "epoch": 0.71,
+      "learning_rate": 8.030813892951219e-05,
+      "loss": 0.6586,
+      "step": 2019
+    },
+    {
+      "epoch": 0.71,
+      "learning_rate": 8.01234921260051e-05,
+      "loss": 0.7224,
+      "step": 2020
+    },
+    {
+      "epoch": 0.71,
+      "eval_loss": 0.5363853573799133,
+      "eval_runtime": 52.9392,
+      "eval_samples_per_second": 30.771,
+      "eval_steps_per_second": 30.771,
+      "step": 2020
+    },
+    {
+      "epoch": 0.72,
+      "learning_rate": 7.993900465367608e-05,
+      "loss": 0.5167,
+      "step": 2021
+    },
+    {
+      "epoch": 0.72,
+      "learning_rate": 7.975467675773245e-05,
+      "loss": 0.5536,
+      "step": 2022
+    },
+    {
+      "epoch": 0.72,
+      "learning_rate": 7.957050868316953e-05,
+      "loss": 0.9368,
+      "step": 2023
+    },
+    {
+      "epoch": 0.72,
+      "learning_rate": 7.938650067477019e-05,
+      "loss": 0.3816,
+      "step": 2024
+    },
+    {
+      "epoch": 0.72,
+      "learning_rate": 7.920265297710452e-05,
+      "loss": 0.6563,
+      "step": 2025
+    },
+    {
+      "epoch": 0.72,
+      "learning_rate": 7.901896583452962e-05,
+      "loss": 0.4537,
+      "step": 2026
+    },
+    {
+      "epoch": 0.72,
+      "learning_rate": 7.883543949118906e-05,
+      "loss": 0.4082,
+      "step": 2027
+    },
+    {
+      "epoch": 0.72,
+      "learning_rate": 7.865207419101297e-05,
+      "loss": 0.5244,
+      "step": 2028
+    },
+    {
+      "epoch": 0.72,
+      "learning_rate": 7.846887017771703e-05,
+      "loss": 0.5391,
+      "step": 2029
+    },
+    {
+      "epoch": 0.72,
+      "learning_rate": 7.828582769480277e-05,
+      "loss": 0.5204,
+      "step": 2030
+    },
+    {
+      "epoch": 0.72,
+      "learning_rate": 7.810294698555713e-05,
+      "loss": 0.4968,
+      "step": 2031
+    },
+    {
+      "epoch": 0.72,
+      "learning_rate": 7.792022829305186e-05,
+      "loss": 0.2879,
+      "step": 2032
+    },
+    {
+      "epoch": 0.72,
+      "learning_rate": 7.773767186014342e-05,
+      "loss": 0.3429,
+      "step": 2033
+    },
+    {
+      "epoch": 0.72,
+      "learning_rate": 7.75552779294726e-05,
+      "loss": 0.7769,
+      "step": 2034
+    },
+    {
+      "epoch": 0.72,
+      "learning_rate": 7.737304674346422e-05,
+      "loss": 0.3696,
+      "step": 2035
+    },
+    {
+      "epoch": 0.72,
+      "learning_rate": 7.719097854432676e-05,
+      "loss": 0.719,
+      "step": 2036
+    },
+    {
+      "epoch": 0.72,
+      "learning_rate": 7.700907357405205e-05,
+      "loss": 0.2607,
+      "step": 2037
+    },
+    {
+      "epoch": 0.72,
+      "learning_rate": 7.682733207441521e-05,
+      "loss": 0.4511,
+      "step": 2038
+    },
+    {
+      "epoch": 0.72,
+      "learning_rate": 7.664575428697367e-05,
+      "loss": 0.5368,
+      "step": 2039
+    },
+    {
+      "epoch": 0.72,
+      "learning_rate": 7.646434045306752e-05,
+      "loss": 0.3617,
+      "step": 2040
+    },
+    {
+      "epoch": 0.72,
+      "eval_loss": 0.5357111096382141,
+      "eval_runtime": 52.9377,
+      "eval_samples_per_second": 30.772,
+      "eval_steps_per_second": 30.772,
+      "step": 2040
+    },
+    {
+      "epoch": 0.72,
+      "learning_rate": 7.628309081381896e-05,
+      "loss": 0.3812,
+      "step": 2041
+    },
+    {
+      "epoch": 0.72,
+      "learning_rate": 7.610200561013192e-05,
+      "loss": 0.4126,
+      "step": 2042
+    },
+    {
+      "epoch": 0.72,
+      "learning_rate": 7.592108508269167e-05,
+      "loss": 0.5468,
+      "step": 2043
+    },
+    {
+      "epoch": 0.72,
+      "learning_rate": 7.574032947196472e-05,
+      "loss": 0.4325,
+      "step": 2044
+    },
+    {
+      "epoch": 0.72,
+      "learning_rate": 7.555973901819833e-05,
+      "loss": 0.3803,
+      "step": 2045
+    },
+    {
+      "epoch": 0.72,
+      "learning_rate": 7.537931396142025e-05,
+      "loss": 0.3779,
+      "step": 2046
+    },
+    {
+      "epoch": 0.72,
+      "learning_rate": 7.519905454143836e-05,
+      "loss": 0.5036,
+      "step": 2047
+    },
+    {
+      "epoch": 0.72,
+      "learning_rate": 7.501896099784065e-05,
+      "loss": 0.2827,
+      "step": 2048
+    },
+    {
+      "epoch": 0.73,
+      "learning_rate": 7.48390335699942e-05,
+      "loss": 0.4214,
+      "step": 2049
+    },
+    {
+      "epoch": 0.73,
+      "learning_rate": 7.46592724970455e-05,
+      "loss": 0.3664,
+      "step": 2050
+    },
+    {
+      "epoch": 0.73,
+      "learning_rate": 7.447967801792013e-05,
+      "loss": 0.2106,
+      "step": 2051
+    },
+    {
+      "epoch": 0.73,
+      "learning_rate": 7.4300250371322e-05,
+      "loss": 0.9233,
+      "step": 2052
+    },
+    {
+      "epoch": 0.73,
+      "learning_rate": 7.412098979573335e-05,
+      "loss": 0.3071,
+      "step": 2053
+    },
+    {
+      "epoch": 0.73,
+      "learning_rate": 7.394189652941435e-05,
+      "loss": 0.3139,
+      "step": 2054
+    },
+    {
+      "epoch": 0.73,
+      "learning_rate": 7.376297081040281e-05,
+      "loss": 0.4044,
+      "step": 2055
+    },
+    {
+      "epoch": 0.73,
+      "learning_rate": 7.358421287651387e-05,
+      "loss": 0.2927,
+      "step": 2056
+    },
+    {
+      "epoch": 0.73,
+      "learning_rate": 7.340562296533954e-05,
+      "loss": 0.5994,
+      "step": 2057
+    },
+    {
+      "epoch": 0.73,
+      "learning_rate": 7.322720131424882e-05,
+      "loss": 0.2545,
+      "step": 2058
+    },
+    {
+      "epoch": 0.73,
+      "learning_rate": 7.304894816038668e-05,
+      "loss": 0.5272,
+      "step": 2059
+    },
+    {
+      "epoch": 0.73,
+      "learning_rate": 7.287086374067424e-05,
+      "loss": 0.8601,
+      "step": 2060
+    },
+    {
+      "epoch": 0.73,
+      "eval_loss": 0.5401872992515564,
+      "eval_runtime": 52.9385,
+      "eval_samples_per_second": 30.772,
+      "eval_steps_per_second": 30.772,
+      "step": 2060
+    },
+    {
+      "epoch": 0.73,
+      "learning_rate": 7.269294829180862e-05,
+      "loss": 0.4021,
+      "step": 2061
+    },
+    {
+      "epoch": 0.73,
+      "learning_rate": 7.251520205026205e-05,
+      "loss": 0.5128,
+      "step": 2062
+    },
+    {
+      "epoch": 0.73,
+      "learning_rate": 7.2337625252282e-05,
+      "loss": 0.5183,
+      "step": 2063
+    },
+    {
+      "epoch": 0.73,
+      "learning_rate": 7.216021813389071e-05,
+      "loss": 0.4734,
+      "step": 2064
+    },
+    {
+      "epoch": 0.73,
+      "learning_rate": 7.198298093088482e-05,
+      "loss": 0.6901,
+      "step": 2065
+    },
+    {
+      "epoch": 0.73,
+      "learning_rate": 7.180591387883526e-05,
+      "loss": 0.6101,
+      "step": 2066
+    },
+    {
+      "epoch": 0.73,
+      "learning_rate": 7.16290172130867e-05,
+      "loss": 0.4975,
+      "step": 2067
+    },
+    {
+      "epoch": 0.73,
+      "learning_rate": 7.145229116875738e-05,
+      "loss": 0.5818,
+      "step": 2068
+    },
+    {
+      "epoch": 0.73,
+      "learning_rate": 7.12757359807388e-05,
+      "loss": 0.4736,
+      "step": 2069
+    },
+    {
+      "epoch": 0.73,
+      "learning_rate": 7.109935188369523e-05,
+      "loss": 0.6994,
+      "step": 2070
+    },
+    {
+      "epoch": 0.73,
+      "learning_rate": 7.092313911206383e-05,
+      "loss": 0.3049,
+      "step": 2071
+    },
+    {
+      "epoch": 0.73,
+      "learning_rate": 7.074709790005376e-05,
+      "loss": 0.3348,
+      "step": 2072
+    },
+    {
+      "epoch": 0.73,
+      "learning_rate": 7.057122848164625e-05,
+      "loss": 0.4658,
+      "step": 2073
+    },
+    {
+      "epoch": 0.73,
+      "learning_rate": 7.039553109059425e-05,
+      "loss": 0.5748,
+      "step": 2074
+    },
+    {
+      "epoch": 0.73,
+      "learning_rate": 7.022000596042195e-05,
+      "loss": 0.3313,
+      "step": 2075
+    },
+    {
+      "epoch": 0.73,
+      "learning_rate": 7.004465332442469e-05,
+      "loss": 0.3836,
+      "step": 2076
+    },
+    {
+      "epoch": 0.74,
+      "learning_rate": 6.986947341566852e-05,
+      "loss": 0.3522,
+      "step": 2077
+    },
+    {
+      "epoch": 0.74,
+      "learning_rate": 6.969446646698983e-05,
+      "loss": 0.5366,
+      "step": 2078
+    },
+    {
+      "epoch": 0.74,
+      "learning_rate": 6.951963271099526e-05,
+      "loss": 0.4448,
+      "step": 2079
+    },
+    {
+      "epoch": 0.74,
+      "learning_rate": 6.934497238006108e-05,
+      "loss": 0.3218,
+      "step": 2080
+    },
+    {
+      "epoch": 0.74,
+      "eval_loss": 0.5309258699417114,
+      "eval_runtime": 52.9421,
+      "eval_samples_per_second": 30.769,
+      "eval_steps_per_second": 30.769,
+      "step": 2080
+    },
+    {
+      "epoch": 0.74,
+      "learning_rate": 6.917048570633331e-05,
+      "loss": 0.5885,
+      "step": 2081
+    },
+    {
+      "epoch": 0.74,
+      "learning_rate": 6.89961729217269e-05,
+      "loss": 0.5008,
+      "step": 2082
+    },
+    {
+      "epoch": 0.74,
+      "learning_rate": 6.882203425792585e-05,
+      "loss": 0.9902,
+      "step": 2083
+    },
+    {
+      "epoch": 0.74,
+      "learning_rate": 6.864806994638264e-05,
+      "loss": 1.0607,
+      "step": 2084
+    },
+    {
+      "epoch": 0.74,
+      "learning_rate": 6.847428021831803e-05,
+      "loss": 0.7099,
+      "step": 2085
+    },
+    {
+      "epoch": 0.74,
+      "learning_rate": 6.830066530472077e-05,
+      "loss": 0.2343,
+      "step": 2086
+    },
+    {
+      "epoch": 0.74,
+      "learning_rate": 6.812722543634722e-05,
+      "loss": 0.5244,
+      "step": 2087
+    },
+    {
+      "epoch": 0.74,
+      "learning_rate": 6.795396084372112e-05,
+      "loss": 0.3426,
+      "step": 2088
+    },
+    {
+      "epoch": 0.74,
+      "learning_rate": 6.778087175713319e-05,
+      "loss": 0.5699,
+      "step": 2089
+    },
+    {
+      "epoch": 0.74,
+      "learning_rate": 6.760795840664089e-05,
+      "loss": 0.37,
+      "step": 2090
+    },
+    {
+      "epoch": 0.74,
+      "learning_rate": 6.743522102206824e-05,
+      "loss": 0.5993,
+      "step": 2091
+    },
+    {
+      "epoch": 0.74,
+      "learning_rate": 6.726265983300519e-05,
+      "loss": 0.6894,
+      "step": 2092
+    },
+    {
+      "epoch": 0.74,
+      "learning_rate": 6.709027506880763e-05,
+      "loss": 0.5941,
+      "step": 2093
+    },
+    {
+      "epoch": 0.74,
+      "learning_rate": 6.691806695859688e-05,
+      "loss": 0.3062,
+      "step": 2094
+    },
+    {
+      "epoch": 0.74,
+      "learning_rate": 6.674603573125948e-05,
+      "loss": 0.2463,
+      "step": 2095
+    },
+    {
+      "epoch": 0.74,
+      "learning_rate": 6.657418161544693e-05,
+      "loss": 0.3182,
+      "step": 2096
+    },
+    {
+      "epoch": 0.74,
+      "learning_rate": 6.640250483957525e-05,
+      "loss": 0.5129,
+      "step": 2097
+    },
+    {
+      "epoch": 0.74,
+      "learning_rate": 6.623100563182481e-05,
+      "loss": 0.3639,
+      "step": 2098
+    },
+    {
+      "epoch": 0.74,
+      "learning_rate": 6.605968422013991e-05,
+      "loss": 0.368,
+      "step": 2099
+    },
+    {
+      "epoch": 0.74,
+      "learning_rate": 6.58885408322285e-05,
+      "loss": 0.3611,
+      "step": 2100
+    },
+    {
+      "epoch": 0.74,
+      "eval_loss": 0.541223406791687,
+      "eval_runtime": 52.9423,
+      "eval_samples_per_second": 30.769,
+      "eval_steps_per_second": 30.769,
+      "step": 2100
+    },
+    {
+      "epoch": 0.74,
+      "learning_rate": 6.571757569556218e-05,
+      "loss": 0.4968,
+      "step": 2101
+    },
+    {
+      "epoch": 0.74,
+      "learning_rate": 6.554678903737533e-05,
+      "loss": 0.5947,
+      "step": 2102
+    },
+    {
+      "epoch": 0.74,
+      "learning_rate": 6.537618108466519e-05,
+      "loss": 0.274,
+      "step": 2103
+    },
+    {
+      "epoch": 0.74,
+      "learning_rate": 6.520575206419153e-05,
+      "loss": 0.5487,
+      "step": 2104
+    },
+    {
+      "epoch": 0.75,
+      "learning_rate": 6.503550220247628e-05,
+      "loss": 0.3866,
+      "step": 2105
+    },
+    {
+      "epoch": 0.75,
+      "learning_rate": 6.486543172580322e-05,
+      "loss": 0.2653,
+      "step": 2106
+    },
+    {
+      "epoch": 0.75,
+      "learning_rate": 6.469554086021771e-05,
+      "loss": 0.5232,
+      "step": 2107
+    },
+    {
+      "epoch": 0.75,
+      "learning_rate": 6.45258298315264e-05,
+      "loss": 0.4115,
+      "step": 2108
+    },
+    {
+      "epoch": 0.75,
+      "learning_rate": 6.435629886529692e-05,
+      "loss": 0.7051,
+      "step": 2109
+    },
+    {
+      "epoch": 0.75,
+      "learning_rate": 6.418694818685747e-05,
+      "loss": 0.3161,
+      "step": 2110
+    },
+    {
+      "epoch": 0.75,
+      "learning_rate": 6.401777802129688e-05,
+      "loss": 0.6998,
+      "step": 2111
+    },
+    {
+      "epoch": 0.75,
+      "learning_rate": 6.384878859346383e-05,
+      "loss": 0.3928,
+      "step": 2112
+    },
+    {
+      "epoch": 0.75,
+      "learning_rate": 6.36799801279667e-05,
+      "loss": 0.5447,
+      "step": 2113
+    },
+    {
+      "epoch": 0.75,
+      "learning_rate": 6.351135284917366e-05,
+      "loss": 0.3641,
+      "step": 2114
+    },
+    {
+      "epoch": 0.75,
+      "learning_rate": 6.334290698121182e-05,
+      "loss": 0.3219,
+      "step": 2115
+    },
+    {
+      "epoch": 0.75,
+      "learning_rate": 6.317464274796722e-05,
+      "loss": 0.5217,
+      "step": 2116
+    },
+    {
+      "epoch": 0.75,
+      "learning_rate": 6.300656037308452e-05,
+      "loss": 0.3864,
+      "step": 2117
+    },
+    {
+      "epoch": 0.75,
+      "learning_rate": 6.283866007996666e-05,
+      "loss": 0.6832,
+      "step": 2118
+    },
+    {
+      "epoch": 0.75,
+      "learning_rate": 6.267094209177451e-05,
+      "loss": 0.3055,
+      "step": 2119
+    },
+    {
+      "epoch": 0.75,
+      "learning_rate": 6.250340663142666e-05,
+      "loss": 0.4466,
+      "step": 2120
+    },
+    {
+      "epoch": 0.75,
+      "eval_loss": 0.5431803464889526,
+      "eval_runtime": 52.9353,
+      "eval_samples_per_second": 30.773,
+      "eval_steps_per_second": 30.773,
+      "step": 2120
+    },
+    {
+      "epoch": 0.75,
+      "learning_rate": 6.233605392159925e-05,
+      "loss": 0.4938,
+      "step": 2121
+    },
+    {
+      "epoch": 0.75,
+      "learning_rate": 6.216888418472535e-05,
+      "loss": 0.2872,
+      "step": 2122
+    },
+    {
+      "epoch": 0.75,
+      "learning_rate": 6.200189764299472e-05,
+      "loss": 0.3578,
+      "step": 2123
+    },
+    {
+      "epoch": 0.75,
+      "learning_rate": 6.183509451835394e-05,
+      "loss": 0.3728,
+      "step": 2124
+    },
+    {
+      "epoch": 0.75,
+      "learning_rate": 6.166847503250563e-05,
+      "loss": 0.4717,
+      "step": 2125
+    },
+    {
+      "epoch": 0.75,
+      "learning_rate": 6.150203940690833e-05,
+      "loss": 0.437,
+      "step": 2126
+    },
+    {
+      "epoch": 0.75,
+      "learning_rate": 6.133578786277623e-05,
+      "loss": 0.6399,
+      "step": 2127
+    },
+    {
+      "epoch": 0.75,
+      "learning_rate": 6.116972062107884e-05,
+      "loss": 0.3396,
+      "step": 2128
+    },
+    {
+      "epoch": 0.75,
+      "learning_rate": 6.100383790254072e-05,
+      "loss": 0.4739,
+      "step": 2129
+    },
+    {
+      "epoch": 0.75,
+      "learning_rate": 6.083813992764111e-05,
+      "loss": 0.4777,
+      "step": 2130
+    },
+    {
+      "epoch": 0.75,
+      "learning_rate": 6.067262691661393e-05,
+      "loss": 0.2898,
+      "step": 2131
+    },
+    {
+      "epoch": 0.75,
+      "learning_rate": 6.050729908944703e-05,
+      "loss": 0.4033,
+      "step": 2132
+    },
+    {
+      "epoch": 0.75,
+      "learning_rate": 6.0342156665882056e-05,
+      "loss": 0.4797,
+      "step": 2133
+    },
+    {
+      "epoch": 0.76,
+      "learning_rate": 6.0177199865414525e-05,
+      "loss": 0.4101,
+      "step": 2134
+    },
+    {
+      "epoch": 0.76,
+      "learning_rate": 6.001242890729302e-05,
+      "loss": 1.0596,
+      "step": 2135
+    },
+    {
+      "epoch": 0.76,
+      "learning_rate": 5.9847844010519174e-05,
+      "loss": 0.3991,
+      "step": 2136
+    },
+    {
+      "epoch": 0.76,
+      "learning_rate": 5.968344539384734e-05,
+      "loss": 0.5784,
+      "step": 2137
+    },
+    {
+      "epoch": 0.76,
+      "learning_rate": 5.951923327578424e-05,
+      "loss": 0.3177,
+      "step": 2138
+    },
+    {
+      "epoch": 0.76,
+      "learning_rate": 5.935520787458873e-05,
+      "loss": 0.556,
+      "step": 2139
+    },
+    {
+      "epoch": 0.76,
+      "learning_rate": 5.919136940827144e-05,
+      "loss": 0.5551,
+      "step": 2140
+    },
+    {
+      "epoch": 0.76,
+      "eval_loss": 0.5344992876052856,
+      "eval_runtime": 52.9424,
+      "eval_samples_per_second": 30.769,
+      "eval_steps_per_second": 30.769,
+      "step": 2140
+    },
+    {
+      "epoch": 0.76,
+      "learning_rate": 5.902771809459473e-05,
+      "loss": 0.3109,
+      "step": 2141
+    },
+    {
+      "epoch": 0.76,
+      "learning_rate": 5.8864254151072106e-05,
+      "loss": 0.3933,
+      "step": 2142
+    },
+    {
+      "epoch": 0.76,
+      "learning_rate": 5.870097779496779e-05,
+      "loss": 0.6477,
+      "step": 2143
+    },
+    {
+      "epoch": 0.76,
+      "learning_rate": 5.8537889243297106e-05,
+      "loss": 0.262,
+      "step": 2144
+    },
+    {
+      "epoch": 0.76,
+      "learning_rate": 5.837498871282547e-05,
+      "loss": 0.4098,
+      "step": 2145
+    },
+    {
+      "epoch": 0.76,
+      "learning_rate": 5.82122764200685e-05,
+      "loss": 0.5022,
+      "step": 2146
+    },
+    {
+      "epoch": 0.76,
+      "learning_rate": 5.804975258129157e-05,
+      "loss": 0.2987,
+      "step": 2147
+    },
+    {
+      "epoch": 0.76,
+      "learning_rate": 5.788741741250965e-05,
+      "loss": 0.3666,
+      "step": 2148
+    },
+    {
+      "epoch": 0.76,
+      "learning_rate": 5.772527112948684e-05,
+      "loss": 0.2162,
+      "step": 2149
+    },
+    {
+      "epoch": 0.76,
+      "learning_rate": 5.756331394773624e-05,
+      "loss": 0.6933,
+      "step": 2150
+    },
+    {
+      "epoch": 0.76,
+      "learning_rate": 5.7401546082519686e-05,
+      "loss": 0.5159,
+      "step": 2151
+    },
+    {
+      "epoch": 0.76,
+      "learning_rate": 5.7239967748847345e-05,
+      "loss": 0.9534,
+      "step": 2152
+    },
+    {
+      "epoch": 0.76,
+      "learning_rate": 5.707857916147725e-05,
+      "loss": 0.9855,
+      "step": 2153
+    },
+    {
+      "epoch": 0.76,
+      "learning_rate": 5.691738053491562e-05,
+      "loss": 0.5289,
+      "step": 2154
+    },
+    {
+      "epoch": 0.76,
+      "learning_rate": 5.6756372083415955e-05,
+      "loss": 0.2614,
+      "step": 2155
+    },
+    {
+      "epoch": 0.76,
+      "learning_rate": 5.659555402097898e-05,
+      "loss": 0.279,
+      "step": 2156
+    },
+    {
+      "epoch": 0.76,
+      "learning_rate": 5.6434926561352494e-05,
+      "loss": 0.4141,
+      "step": 2157
+    },
+    {
+      "epoch": 0.76,
+      "learning_rate": 5.627448991803086e-05,
+      "loss": 0.5001,
+      "step": 2158
+    },
+    {
+      "epoch": 0.76,
+      "learning_rate": 5.6114244304254823e-05,
+      "loss": 1.0701,
+      "step": 2159
+    },
+    {
+      "epoch": 0.76,
+      "learning_rate": 5.595418993301125e-05,
+      "loss": 0.4047,
+      "step": 2160
+    },
+    {
+      "epoch": 0.76,
+      "eval_loss": 0.5321358442306519,
+      "eval_runtime": 52.9379,
+      "eval_samples_per_second": 30.772,
+      "eval_steps_per_second": 30.772,
+      "step": 2160
+    },
+    {
+      "epoch": 0.76,
+      "learning_rate": 5.5794327017032975e-05,
+      "loss": 0.6084,
+      "step": 2161
+    },
+    {
+      "epoch": 0.77,
+      "learning_rate": 5.5634655768798024e-05,
+      "loss": 0.4282,
+      "step": 2162
+    },
+    {
+      "epoch": 0.77,
+      "learning_rate": 5.5475176400529905e-05,
+      "loss": 0.8112,
+      "step": 2163
+    },
+    {
+      "epoch": 0.77,
+      "learning_rate": 5.531588912419714e-05,
+      "loss": 0.3541,
+      "step": 2164
+    },
+    {
+      "epoch": 0.77,
+      "learning_rate": 5.5156794151512805e-05,
+      "loss": 0.447,
+      "step": 2165
+    },
+    {
+      "epoch": 0.77,
+      "learning_rate": 5.499789169393439e-05,
+      "loss": 0.3828,
+      "step": 2166
+    },
+    {
+      "epoch": 0.77,
+      "learning_rate": 5.483918196266358e-05,
+      "loss": 0.3302,
+      "step": 2167
+    },
+    {
+      "epoch": 0.77,
+      "learning_rate": 5.468066516864585e-05,
+      "loss": 0.408,
+      "step": 2168
+    },
+    {
+      "epoch": 0.77,
+      "learning_rate": 5.452234152257025e-05,
+      "loss": 0.347,
+      "step": 2169
+    },
+    {
+      "epoch": 0.77,
+      "learning_rate": 5.4364211234869037e-05,
+      "loss": 0.4649,
+      "step": 2170
+    },
+    {
+      "epoch": 0.77,
+      "learning_rate": 5.4206274515717736e-05,
+      "loss": 0.4813,
+      "step": 2171
+    },
+    {
+      "epoch": 0.77,
+      "learning_rate": 5.404853157503422e-05,
+      "loss": 0.5809,
+      "step": 2172
+    },
+    {
+      "epoch": 0.77,
+      "learning_rate": 5.3890982622479004e-05,
+      "loss": 0.3522,
+      "step": 2173
+    },
+    {
+      "epoch": 0.77,
+      "learning_rate": 5.373362786745486e-05,
+      "loss": 0.3857,
+      "step": 2174
+    },
+    {
+      "epoch": 0.77,
+      "learning_rate": 5.357646751910628e-05,
+      "loss": 0.4032,
+      "step": 2175
+    },
+    {
+      "epoch": 0.77,
+      "learning_rate": 5.34195017863194e-05,
+      "loss": 1.074,
+      "step": 2176
+    },
+    {
+      "epoch": 0.77,
+      "learning_rate": 5.326273087772178e-05,
+      "loss": 0.4683,
+      "step": 2177
+    },
+    {
+      "epoch": 0.77,
+      "learning_rate": 5.3106155001681875e-05,
+      "loss": 0.4398,
+      "step": 2178
+    },
+    {
+      "epoch": 0.77,
+      "learning_rate": 5.294977436630908e-05,
+      "loss": 0.3122,
+      "step": 2179
+    },
+    {
+      "epoch": 0.77,
+      "learning_rate": 5.279358917945314e-05,
+      "loss": 0.4624,
+      "step": 2180
+    },
+    {
+      "epoch": 0.77,
+      "eval_loss": 0.5356577634811401,
+      "eval_runtime": 52.938,
+      "eval_samples_per_second": 30.772,
+      "eval_steps_per_second": 30.772,
+      "step": 2180
+    },
+    {
+      "epoch": 0.77,
+      "learning_rate": 5.263759964870425e-05,
+      "loss": 0.3998,
+      "step": 2181
+    },
+    {
+      "epoch": 0.77,
+      "learning_rate": 5.248180598139225e-05,
+      "loss": 0.6051,
+      "step": 2182
+    },
+    {
+      "epoch": 0.77,
+      "learning_rate": 5.2326208384586815e-05,
+      "loss": 0.3445,
+      "step": 2183
+    },
+    {
+      "epoch": 0.77,
+      "learning_rate": 5.2170807065097095e-05,
+      "loss": 0.2776,
+      "step": 2184
+    },
+    {
+      "epoch": 0.77,
+      "learning_rate": 5.2015602229471264e-05,
+      "loss": 0.4231,
+      "step": 2185
+    },
+    {
+      "epoch": 0.77,
+      "learning_rate": 5.1860594083996307e-05,
+      "loss": 0.2155,
+      "step": 2186
+    },
+    {
+      "epoch": 0.77,
+      "learning_rate": 5.170578283469787e-05,
+      "loss": 0.4701,
+      "step": 2187
+    },
+    {
+      "epoch": 0.77,
+      "learning_rate": 5.155116868733985e-05,
+      "loss": 0.4235,
+      "step": 2188
+    },
+    {
+      "epoch": 0.77,
+      "learning_rate": 5.139675184742416e-05,
+      "loss": 0.3434,
+      "step": 2189
+    },
+    {
+      "epoch": 0.78,
+      "learning_rate": 5.124253252019045e-05,
+      "loss": 0.3495,
+      "step": 2190
+    },
+    {
+      "epoch": 0.78,
+      "learning_rate": 5.108851091061606e-05,
+      "loss": 0.4046,
+      "step": 2191
+    },
+    {
+      "epoch": 0.78,
+      "learning_rate": 5.093468722341521e-05,
+      "loss": 0.475,
+      "step": 2192
+    },
+    {
+      "epoch": 0.78,
+      "learning_rate": 5.078106166303918e-05,
+      "loss": 0.5783,
+      "step": 2193
+    },
+    {
+      "epoch": 0.78,
+      "learning_rate": 5.062763443367606e-05,
+      "loss": 0.5158,
+      "step": 2194
+    },
+    {
+      "epoch": 0.78,
+      "learning_rate": 5.04744057392502e-05,
+      "loss": 0.4195,
+      "step": 2195
+    },
+    {
+      "epoch": 0.78,
+      "learning_rate": 5.032137578342204e-05,
+      "loss": 0.5004,
+      "step": 2196
+    },
+    {
+      "epoch": 0.78,
+      "learning_rate": 5.0168544769587965e-05,
+      "loss": 0.3999,
+      "step": 2197
+    },
+    {
+      "epoch": 0.78,
+      "learning_rate": 5.001591290087988e-05,
+      "loss": 0.3075,
+      "step": 2198
+    },
+    {
+      "epoch": 0.78,
+      "learning_rate": 4.9863480380164995e-05,
+      "loss": 0.4389,
+      "step": 2199
+    },
+    {
+      "epoch": 0.78,
+      "learning_rate": 4.971124741004558e-05,
+      "loss": 0.5704,
+      "step": 2200
+    },
+    {
+      "epoch": 0.78,
+      "eval_loss": 0.5325384736061096,
+      "eval_runtime": 52.9471,
+      "eval_samples_per_second": 30.767,
+      "eval_steps_per_second": 30.767,
+      "step": 2200
+    },
+    {
+      "epoch": 0.78,
+      "learning_rate": 4.955921419285881e-05,
+      "loss": 0.4032,
+      "step": 2201
+    },
+    {
+      "epoch": 0.78,
+      "learning_rate": 4.940738093067605e-05,
+      "loss": 0.2964,
+      "step": 2202
+    },
+    {
+      "epoch": 0.78,
+      "learning_rate": 4.9255747825303136e-05,
+      "loss": 0.7878,
+      "step": 2203
+    },
+    {
+      "epoch": 0.78,
+      "learning_rate": 4.910431507827988e-05,
+      "loss": 0.3689,
+      "step": 2204
+    },
+    {
+      "epoch": 0.78,
+      "learning_rate": 4.895308289087972e-05,
+      "loss": 0.5067,
+      "step": 2205
+    },
+    {
+      "epoch": 0.78,
+      "learning_rate": 4.880205146410954e-05,
+      "loss": 0.5042,
+      "step": 2206
+    },
+    {
+      "epoch": 0.78,
+      "learning_rate": 4.865122099870925e-05,
+      "loss": 0.2917,
+      "step": 2207
+    },
+    {
+      "epoch": 0.78,
+      "learning_rate": 4.8500591695151933e-05,
+      "loss": 0.7088,
+      "step": 2208
+    },
+    {
+      "epoch": 0.78,
+      "learning_rate": 4.835016375364312e-05,
+      "loss": 0.3908,
+      "step": 2209
+    },
+    {
+      "epoch": 0.78,
+      "learning_rate": 4.819993737412072e-05,
+      "loss": 0.5628,
+      "step": 2210
+    },
+    {
+      "epoch": 0.78,
+      "learning_rate": 4.804991275625479e-05,
+      "loss": 0.3073,
+      "step": 2211
+    },
+    {
+      "epoch": 0.78,
+      "learning_rate": 4.790009009944718e-05,
+      "loss": 0.8896,
+      "step": 2212
+    },
+    {
+      "epoch": 0.78,
+      "learning_rate": 4.7750469602831246e-05,
+      "loss": 0.7788,
+      "step": 2213
+    },
+    {
+      "epoch": 0.78,
+      "learning_rate": 4.76010514652719e-05,
+      "loss": 0.3803,
+      "step": 2214
+    },
+    {
+      "epoch": 0.78,
+      "learning_rate": 4.7451835885364794e-05,
+      "loss": 0.4751,
+      "step": 2215
+    },
+    {
+      "epoch": 0.78,
+      "learning_rate": 4.730282306143654e-05,
+      "loss": 0.5722,
+      "step": 2216
+    },
+    {
+      "epoch": 0.78,
+      "learning_rate": 4.7154013191544065e-05,
+      "loss": 0.4506,
+      "step": 2217
+    },
+    {
+      "epoch": 0.78,
+      "learning_rate": 4.700540647347482e-05,
+      "loss": 0.449,
+      "step": 2218
+    },
+    {
+      "epoch": 0.79,
+      "learning_rate": 4.6857003104746056e-05,
+      "loss": 0.4253,
+      "step": 2219
+    },
+    {
+      "epoch": 0.79,
+      "learning_rate": 4.670880328260479e-05,
+      "loss": 0.715,
+      "step": 2220
+    },
+    {
+      "epoch": 0.79,
+      "eval_loss": 0.5313291549682617,
+      "eval_runtime": 52.9346,
+      "eval_samples_per_second": 30.774,
+      "eval_steps_per_second": 30.774,
+      "step": 2220
+    },
+    {
+      "epoch": 0.79,
+      "learning_rate": 4.656080720402747e-05,
+      "loss": 0.3859,
+      "step": 2221
+    },
+    {
+      "epoch": 0.79,
+      "learning_rate": 4.64130150657198e-05,
+      "loss": 0.2703,
+      "step": 2222
+    },
+    {
+      "epoch": 0.79,
+      "learning_rate": 4.626542706411636e-05,
+      "loss": 0.625,
+      "step": 2223
+    },
+    {
+      "epoch": 0.79,
+      "learning_rate": 4.6118043395380505e-05,
+      "loss": 0.5131,
+      "step": 2224
+    },
+    {
+      "epoch": 0.79,
+      "learning_rate": 4.597086425540389e-05,
+      "loss": 0.8685,
+      "step": 2225
+    },
+    {
+      "epoch": 0.79,
+      "learning_rate": 4.582388983980648e-05,
+      "loss": 0.3595,
+      "step": 2226
+    },
+    {
+      "epoch": 0.79,
+      "learning_rate": 4.567712034393581e-05,
+      "loss": 0.2634,
+      "step": 2227
+    },
+    {
+      "epoch": 0.79,
+      "learning_rate": 4.553055596286744e-05,
+      "loss": 1.0479,
+      "step": 2228
+    },
+    {
+      "epoch": 0.79,
+      "learning_rate": 4.538419689140412e-05,
+      "loss": 0.2908,
+      "step": 2229
+    },
+    {
+      "epoch": 0.79,
+      "learning_rate": 4.523804332407571e-05,
+      "loss": 0.4188,
+      "step": 2230
+    },
+    {
+      "epoch": 0.79,
+      "learning_rate": 4.509209545513895e-05,
+      "loss": 0.3276,
+      "step": 2231
+    },
+    {
+      "epoch": 0.79,
+      "learning_rate": 4.4946353478577205e-05,
+      "loss": 0.3014,
+      "step": 2232
+    },
+    {
+      "epoch": 0.79,
+      "learning_rate": 4.4800817588100084e-05,
+      "loss": 0.3603,
+      "step": 2233
+    },
+    {
+      "epoch": 0.79,
+      "learning_rate": 4.465548797714347e-05,
+      "loss": 0.8648,
+      "step": 2234
+    },
+    {
+      "epoch": 0.79,
+      "learning_rate": 4.451036483886892e-05,
+      "loss": 0.5961,
+      "step": 2235
+    },
+    {
+      "epoch": 0.79,
+      "learning_rate": 4.4365448366163675e-05,
+      "loss": 0.4956,
+      "step": 2236
+    },
+    {
+      "epoch": 0.79,
+      "learning_rate": 4.4220738751640035e-05,
+      "loss": 0.2991,
+      "step": 2237
+    },
+    {
+      "epoch": 0.79,
+      "learning_rate": 4.40762361876357e-05,
+      "loss": 0.394,
+      "step": 2238
+    },
+    {
+      "epoch": 0.79,
+      "learning_rate": 4.393194086621299e-05,
+      "loss": 0.3386,
+      "step": 2239
+    },
+    {
+      "epoch": 0.79,
+      "learning_rate": 4.3787852979158794e-05,
+      "loss": 0.4913,
+      "step": 2240
+    },
+    {
+      "epoch": 0.79,
+      "eval_loss": 0.5299963355064392,
+      "eval_runtime": 52.9388,
+      "eval_samples_per_second": 30.771,
+      "eval_steps_per_second": 30.771,
+      "step": 2240
+    },
+    {
+      "epoch": 0.79,
+      "learning_rate": 4.364397271798431e-05,
+      "loss": 0.944,
+      "step": 2241
+    },
+    {
+      "epoch": 0.79,
+      "learning_rate": 4.350030027392476e-05,
+      "loss": 0.3061,
+      "step": 2242
+    },
+    {
+      "epoch": 0.79,
+      "learning_rate": 4.335683583793908e-05,
+      "loss": 1.0297,
+      "step": 2243
+    },
+    {
+      "epoch": 0.79,
+      "learning_rate": 4.321357960070995e-05,
+      "loss": 0.3663,
+      "step": 2244
+    },
+    {
+      "epoch": 0.79,
+      "learning_rate": 4.307053175264315e-05,
+      "loss": 0.4578,
+      "step": 2245
+    },
+    {
+      "epoch": 0.79,
+      "learning_rate": 4.292769248386757e-05,
+      "loss": 0.3208,
+      "step": 2246
+    },
+    {
+      "epoch": 0.8,
+      "learning_rate": 4.2785061984234644e-05,
+      "loss": 0.3434,
+      "step": 2247
+    },
+    {
+      "epoch": 0.8,
+      "learning_rate": 4.2642640443318695e-05,
+      "loss": 0.4696,
+      "step": 2248
+    },
+    {
+      "epoch": 0.8,
+      "learning_rate": 4.250042805041605e-05,
+      "loss": 0.3544,
+      "step": 2249
+    },
+    {
+      "epoch": 0.8,
+      "learning_rate": 4.235842499454517e-05,
+      "loss": 0.4487,
+      "step": 2250
+    },
+    {
+      "epoch": 0.8,
+      "learning_rate": 4.221663146444621e-05,
+      "loss": 0.6879,
+      "step": 2251
+    },
+    {
+      "epoch": 0.8,
+      "learning_rate": 4.2075047648580854e-05,
+      "loss": 0.4382,
+      "step": 2252
+    },
+    {
+      "epoch": 0.8,
+      "learning_rate": 4.193367373513204e-05,
+      "loss": 0.3326,
+      "step": 2253
+    },
+    {
+      "epoch": 0.8,
+      "learning_rate": 4.179250991200383e-05,
+      "loss": 0.5132,
+      "step": 2254
+    },
+    {
+      "epoch": 0.8,
+      "learning_rate": 4.1651556366820966e-05,
+      "loss": 0.6791,
+      "step": 2255
+    },
+    {
+      "epoch": 0.8,
+      "learning_rate": 4.151081328692859e-05,
+      "loss": 0.5043,
+      "step": 2256
+    },
+    {
+      "epoch": 0.8,
+      "learning_rate": 4.1370280859392226e-05,
+      "loss": 0.5805,
+      "step": 2257
+    },
+    {
+      "epoch": 0.8,
+      "learning_rate": 4.122995927099749e-05,
+      "loss": 0.645,
+      "step": 2258
+    },
+    {
+      "epoch": 0.8,
+      "learning_rate": 4.1089848708249654e-05,
+      "loss": 0.5171,
+      "step": 2259
+    },
+    {
+      "epoch": 0.8,
+      "learning_rate": 4.094994935737353e-05,
+      "loss": 0.3605,
+      "step": 2260
+    },
+    {
+      "epoch": 0.8,
+      "eval_loss": 0.5293885469436646,
+      "eval_runtime": 52.9373,
+      "eval_samples_per_second": 30.772,
+      "eval_steps_per_second": 30.772,
+      "step": 2260
+    },
+    {
+      "epoch": 0.8,
+      "learning_rate": 4.081026140431319e-05,
+      "loss": 0.8027,
+      "step": 2261
+    },
+    {
+      "epoch": 0.8,
+      "learning_rate": 4.067078503473178e-05,
+      "loss": 0.5978,
+      "step": 2262
+    },
+    {
+      "epoch": 0.8,
+      "learning_rate": 4.053152043401114e-05,
+      "loss": 0.3621,
+      "step": 2263
+    },
+    {
+      "epoch": 0.8,
+      "learning_rate": 4.039246778725177e-05,
+      "loss": 0.3037,
+      "step": 2264
+    },
+    {
+      "epoch": 0.8,
+      "learning_rate": 4.025362727927242e-05,
+      "loss": 0.5404,
+      "step": 2265
+    },
+    {
+      "epoch": 0.8,
+      "learning_rate": 4.011499909460972e-05,
+      "loss": 0.6448,
+      "step": 2266
+    },
+    {
+      "epoch": 0.8,
+      "learning_rate": 3.997658341751822e-05,
+      "loss": 0.3886,
+      "step": 2267
+    },
+    {
+      "epoch": 0.8,
+      "learning_rate": 3.983838043197015e-05,
+      "loss": 0.5002,
+      "step": 2268
+    },
+    {
+      "epoch": 0.8,
+      "learning_rate": 3.970039032165482e-05,
+      "loss": 0.3597,
+      "step": 2269
+    },
+    {
+      "epoch": 0.8,
+      "learning_rate": 3.9562613269978743e-05,
+      "loss": 0.7611,
+      "step": 2270
+    },
+    {
+      "epoch": 0.8,
+      "learning_rate": 3.9425049460065176e-05,
+      "loss": 0.3785,
+      "step": 2271
+    },
+    {
+      "epoch": 0.8,
+      "learning_rate": 3.928769907475398e-05,
+      "loss": 0.3976,
+      "step": 2272
+    },
+    {
+      "epoch": 0.8,
+      "learning_rate": 3.91505622966013e-05,
+      "loss": 0.2471,
+      "step": 2273
+    },
+    {
+      "epoch": 0.8,
+      "learning_rate": 3.901363930787951e-05,
+      "loss": 0.4863,
+      "step": 2274
+    },
+    {
+      "epoch": 0.81,
+      "learning_rate": 3.887693029057675e-05,
+      "loss": 0.634,
+      "step": 2275
+    },
+    {
+      "epoch": 0.81,
+      "learning_rate": 3.874043542639665e-05,
+      "loss": 0.5327,
+      "step": 2276
+    },
+    {
+      "epoch": 0.81,
+      "learning_rate": 3.860415489675831e-05,
+      "loss": 0.7522,
+      "step": 2277
+    },
+    {
+      "epoch": 0.81,
+      "learning_rate": 3.8468088882796026e-05,
+      "loss": 0.4392,
+      "step": 2278
+    },
+    {
+      "epoch": 0.81,
+      "learning_rate": 3.833223756535884e-05,
+      "loss": 0.5261,
+      "step": 2279
+    },
+    {
+      "epoch": 0.81,
+      "learning_rate": 3.819660112501053e-05,
+      "loss": 0.234,
+      "step": 2280
+    },
+    {
+      "epoch": 0.81,
+      "eval_loss": 0.5318276882171631,
+      "eval_runtime": 52.9453,
+      "eval_samples_per_second": 30.768,
+      "eval_steps_per_second": 30.768,
+      "step": 2280
+    },
+    {
+      "epoch": 0.81,
+      "learning_rate": 3.806117974202921e-05,
+      "loss": 0.511,
+      "step": 2281
+    },
+    {
+      "epoch": 0.81,
+      "learning_rate": 3.792597359640717e-05,
+      "loss": 0.351,
+      "step": 2282
+    },
+    {
+      "epoch": 0.81,
+      "learning_rate": 3.7790982867850654e-05,
+      "loss": 0.7619,
+      "step": 2283
+    },
+    {
+      "epoch": 0.81,
+      "learning_rate": 3.765620773577949e-05,
+      "loss": 0.3991,
+      "step": 2284
+    },
+    {
+      "epoch": 0.81,
+      "learning_rate": 3.7521648379327234e-05,
+      "loss": 0.5499,
+      "step": 2285
+    },
+    {
+      "epoch": 0.81,
+      "learning_rate": 3.738730497734024e-05,
+      "loss": 0.3447,
+      "step": 2286
+    },
+    {
+      "epoch": 0.81,
+      "learning_rate": 3.725317770837804e-05,
+      "loss": 0.574,
+      "step": 2287
+    },
+    {
+      "epoch": 0.81,
+      "learning_rate": 3.7119266750713025e-05,
+      "loss": 1.1332,
+      "step": 2288
+    },
+    {
+      "epoch": 0.81,
+      "learning_rate": 3.698557228232988e-05,
+      "loss": 0.3258,
+      "step": 2289
+    },
+    {
+      "epoch": 0.81,
+      "learning_rate": 3.685209448092561e-05,
+      "loss": 0.7222,
+      "step": 2290
+    },
+    {
+      "epoch": 0.81,
+      "learning_rate": 3.671883352390926e-05,
+      "loss": 0.27,
+      "step": 2291
+    },
+    {
+      "epoch": 0.81,
+      "learning_rate": 3.658578958840162e-05,
+      "loss": 0.239,
+      "step": 2292
+    },
+    {
+      "epoch": 0.81,
+      "learning_rate": 3.645296285123509e-05,
+      "loss": 0.4248,
+      "step": 2293
+    },
+    {
+      "epoch": 0.81,
+      "learning_rate": 3.632035348895331e-05,
+      "loss": 0.7101,
+      "step": 2294
+    },
+    {
+      "epoch": 0.81,
+      "learning_rate": 3.618796167781118e-05,
+      "loss": 0.545,
+      "step": 2295
+    },
+    {
+      "epoch": 0.81,
+      "learning_rate": 3.605578759377415e-05,
+      "loss": 0.3368,
+      "step": 2296
+    },
+    {
+      "epoch": 0.81,
+      "learning_rate": 3.592383141251847e-05,
+      "loss": 0.4287,
+      "step": 2297
+    },
+    {
+      "epoch": 0.81,
+      "learning_rate": 3.579209330943085e-05,
+      "loss": 0.294,
+      "step": 2298
+    },
+    {
+      "epoch": 0.81,
+      "learning_rate": 3.566057345960796e-05,
+      "loss": 0.2255,
+      "step": 2299
+    },
+    {
+      "epoch": 0.81,
+      "learning_rate": 3.55292720378565e-05,
+      "loss": 0.6128,
+      "step": 2300
+    },
+    {
+      "epoch": 0.81,
+      "eval_loss": 0.5364549160003662,
+      "eval_runtime": 52.9419,
+      "eval_samples_per_second": 30.77,
+      "eval_steps_per_second": 30.77,
+      "step": 2300
+    },
+    {
+      "epoch": 0.81,
+      "learning_rate": 3.53981892186928e-05,
+      "loss": 0.4691,
+      "step": 2301
+    },
+    {
+      "epoch": 0.81,
+      "learning_rate": 3.526732517634266e-05,
+      "loss": 0.39,
+      "step": 2302
+    },
+    {
+      "epoch": 0.82,
+      "learning_rate": 3.513668008474109e-05,
+      "loss": 0.5195,
+      "step": 2303
+    },
+    {
+      "epoch": 0.82,
+      "learning_rate": 3.500625411753209e-05,
+      "loss": 0.5023,
+      "step": 2304
+    },
+    {
+      "epoch": 0.82,
+      "learning_rate": 3.4876047448068426e-05,
+      "loss": 0.4777,
+      "step": 2305
+    },
+    {
+      "epoch": 0.82,
+      "learning_rate": 3.474606024941136e-05,
+      "loss": 0.3052,
+      "step": 2306
+    },
+    {
+      "epoch": 0.82,
+      "learning_rate": 3.4616292694330424e-05,
+      "loss": 0.6186,
+      "step": 2307
+    },
+    {
+      "epoch": 0.82,
+      "learning_rate": 3.448674495530337e-05,
+      "loss": 0.5685,
+      "step": 2308
+    },
+    {
+      "epoch": 0.82,
+      "learning_rate": 3.435741720451564e-05,
+      "loss": 0.3006,
+      "step": 2309
+    },
+    {
+      "epoch": 0.82,
+      "learning_rate": 3.422830961386028e-05,
+      "loss": 0.3993,
+      "step": 2310
+    },
+    {
+      "epoch": 0.82,
+      "learning_rate": 3.4099422354937794e-05,
+      "loss": 0.4321,
+      "step": 2311
+    },
+    {
+      "epoch": 0.82,
+      "learning_rate": 3.397075559905576e-05,
+      "loss": 0.3037,
+      "step": 2312
+    },
+    {
+      "epoch": 0.82,
+      "learning_rate": 3.3842309517228754e-05,
+      "loss": 0.4554,
+      "step": 2313
+    },
+    {
+      "epoch": 0.82,
+      "learning_rate": 3.371408428017799e-05,
+      "loss": 0.4206,
+      "step": 2314
+    },
+    {
+      "epoch": 0.82,
+      "learning_rate": 3.3586080058331194e-05,
+      "loss": 0.2972,
+      "step": 2315
+    },
+    {
+      "epoch": 0.82,
+      "learning_rate": 3.345829702182228e-05,
+      "loss": 0.382,
+      "step": 2316
+    },
+    {
+      "epoch": 0.82,
+      "learning_rate": 3.333073534049118e-05,
+      "loss": 0.4538,
+      "step": 2317
+    },
+    {
+      "epoch": 0.82,
+      "learning_rate": 3.320339518388376e-05,
+      "loss": 0.6326,
+      "step": 2318
+    },
+    {
+      "epoch": 0.82,
+      "learning_rate": 3.3076276721251267e-05,
+      "loss": 0.4349,
+      "step": 2319
+    },
+    {
+      "epoch": 0.82,
+      "learning_rate": 3.294938012155036e-05,
+      "loss": 0.236,
+      "step": 2320
+    },
+    {
+      "epoch": 0.82,
+      "eval_loss": 0.5341804623603821,
+      "eval_runtime": 52.9371,
+      "eval_samples_per_second": 30.772,
+      "eval_steps_per_second": 30.772,
+      "step": 2320
+    },
+    {
+      "epoch": 0.82,
+      "learning_rate": 3.2822705553442844e-05,
+      "loss": 0.427,
+      "step": 2321
+    },
+    {
+      "epoch": 0.82,
+      "learning_rate": 3.269625318529537e-05,
+      "loss": 0.3966,
+      "step": 2322
+    },
+    {
+      "epoch": 0.82,
+      "learning_rate": 3.2570023185179255e-05,
+      "loss": 0.3866,
+      "step": 2323
+    },
+    {
+      "epoch": 0.82,
+      "learning_rate": 3.24440157208703e-05,
+      "loss": 0.3568,
+      "step": 2324
+    },
+    {
+      "epoch": 0.82,
+      "learning_rate": 3.2318230959848514e-05,
+      "loss": 0.4621,
+      "step": 2325
+    },
+    {
+      "epoch": 0.82,
+      "learning_rate": 3.2192669069297896e-05,
+      "loss": 0.3675,
+      "step": 2326
+    },
+    {
+      "epoch": 0.82,
+      "learning_rate": 3.206733021610615e-05,
+      "loss": 0.3777,
+      "step": 2327
+    },
+    {
+      "epoch": 0.82,
+      "learning_rate": 3.1942214566864723e-05,
+      "loss": 0.4656,
+      "step": 2328
+    },
+    {
+      "epoch": 0.82,
+      "learning_rate": 3.1817322287868224e-05,
+      "loss": 0.3544,
+      "step": 2329
+    },
+    {
+      "epoch": 0.82,
+      "learning_rate": 3.169265354511446e-05,
+      "loss": 0.3866,
+      "step": 2330
+    },
+    {
+      "epoch": 0.82,
+      "learning_rate": 3.156820850430404e-05,
+      "loss": 0.4365,
+      "step": 2331
+    },
+    {
+      "epoch": 0.83,
+      "learning_rate": 3.144398733084037e-05,
+      "loss": 0.5055,
+      "step": 2332
+    },
+    {
+      "epoch": 0.83,
+      "learning_rate": 3.131999018982921e-05,
+      "loss": 0.4013,
+      "step": 2333
+    },
+    {
+      "epoch": 0.83,
+      "learning_rate": 3.119621724607859e-05,
+      "loss": 0.9756,
+      "step": 2334
+    },
+    {
+      "epoch": 0.83,
+      "learning_rate": 3.107266866409855e-05,
+      "loss": 0.5751,
+      "step": 2335
+    },
+    {
+      "epoch": 0.83,
+      "learning_rate": 3.09493446081009e-05,
+      "loss": 0.4667,
+      "step": 2336
+    },
+    {
+      "epoch": 0.83,
+      "learning_rate": 3.0826245241999064e-05,
+      "loss": 0.533,
+      "step": 2337
+    },
+    {
+      "epoch": 0.83,
+      "learning_rate": 3.070337072940783e-05,
+      "loss": 0.3222,
+      "step": 2338
+    },
+    {
+      "epoch": 0.83,
+      "learning_rate": 3.058072123364313e-05,
+      "loss": 0.2692,
+      "step": 2339
+    },
+    {
+      "epoch": 0.83,
+      "learning_rate": 3.0458296917721728e-05,
+      "loss": 0.3503,
+      "step": 2340
+    },
+    {
+      "epoch": 0.83,
+      "eval_loss": 0.53475022315979,
+      "eval_runtime": 52.9285,
+      "eval_samples_per_second": 30.777,
+      "eval_steps_per_second": 30.777,
+      "step": 2340
+    },
+    {
+      "epoch": 0.83,
+      "learning_rate": 3.0336097944361208e-05,
+      "loss": 0.7049,
+      "step": 2341
+    },
+    {
+      "epoch": 0.83,
+      "learning_rate": 3.02141244759796e-05,
+      "loss": 0.3237,
+      "step": 2342
+    },
+    {
+      "epoch": 0.83,
+      "learning_rate": 3.0092376674695198e-05,
+      "loss": 0.7397,
+      "step": 2343
+    },
+    {
+      "epoch": 0.83,
+      "learning_rate": 2.9970854702326366e-05,
+      "loss": 0.3226,
+      "step": 2344
+    },
+    {
+      "epoch": 0.83,
+      "learning_rate": 2.9849558720391325e-05,
+      "loss": 0.2001,
+      "step": 2345
+    },
+    {
+      "epoch": 0.83,
+      "learning_rate": 2.972848889010791e-05,
+      "loss": 0.3323,
+      "step": 2346
+    },
+    {
+      "epoch": 0.83,
+      "learning_rate": 2.960764537239331e-05,
+      "loss": 0.5685,
+      "step": 2347
+    },
+    {
+      "epoch": 0.83,
+      "learning_rate": 2.948702832786412e-05,
+      "loss": 0.3356,
+      "step": 2348
+    },
+    {
+      "epoch": 0.83,
+      "learning_rate": 2.936663791683574e-05,
+      "loss": 0.1694,
+      "step": 2349
+    },
+    {
+      "epoch": 0.83,
+      "learning_rate": 2.924647429932228e-05,
+      "loss": 0.2823,
+      "step": 2350
+    },
+    {
+      "epoch": 0.83,
+      "learning_rate": 2.9126537635036654e-05,
+      "loss": 0.4061,
+      "step": 2351
+    },
+    {
+      "epoch": 0.83,
+      "learning_rate": 2.9006828083389968e-05,
+      "loss": 0.4574,
+      "step": 2352
+    },
+    {
+      "epoch": 0.83,
+      "learning_rate": 2.8887345803491463e-05,
+      "loss": 0.3661,
+      "step": 2353
+    },
+    {
+      "epoch": 0.83,
+      "learning_rate": 2.876809095414834e-05,
+      "loss": 0.4093,
+      "step": 2354
+    },
+    {
+      "epoch": 0.83,
+      "learning_rate": 2.864906369386553e-05,
+      "loss": 0.8003,
+      "step": 2355
+    },
+    {
+      "epoch": 0.83,
+      "learning_rate": 2.8530264180845432e-05,
+      "loss": 0.5456,
+      "step": 2356
+    },
+    {
+      "epoch": 0.83,
+      "learning_rate": 2.8411692572987745e-05,
+      "loss": 0.5592,
+      "step": 2357
+    },
+    {
+      "epoch": 0.83,
+      "learning_rate": 2.829334902788936e-05,
+      "loss": 0.5634,
+      "step": 2358
+    },
+    {
+      "epoch": 0.83,
+      "learning_rate": 2.817523370284392e-05,
+      "loss": 1.0663,
+      "step": 2359
+    },
+    {
+      "epoch": 0.84,
+      "learning_rate": 2.805734675484164e-05,
+      "loss": 0.4874,
+      "step": 2360
+    },
+    {
+      "epoch": 0.84,
+      "eval_loss": 0.5313405394554138,
+      "eval_runtime": 52.9392,
+      "eval_samples_per_second": 30.771,
+      "eval_steps_per_second": 30.771,
+      "step": 2360
+    },
+    {
+      "epoch": 0.84,
+      "learning_rate": 2.7939688340569458e-05,
+      "loss": 0.3468,
+      "step": 2361
+    },
+    {
+      "epoch": 0.84,
+      "learning_rate": 2.782225861641039e-05,
+      "loss": 0.5032,
+      "step": 2362
+    },
+    {
+      "epoch": 0.84,
+      "learning_rate": 2.770505773844352e-05,
+      "loss": 0.334,
+      "step": 2363
+    },
+    {
+      "epoch": 0.84,
+      "learning_rate": 2.7588085862443746e-05,
+      "loss": 0.2274,
+      "step": 2364
+    },
+    {
+      "epoch": 0.84,
+      "learning_rate": 2.747134314388167e-05,
+      "loss": 0.8527,
+      "step": 2365
+    },
+    {
+      "epoch": 0.84,
+      "learning_rate": 2.7354829737923225e-05,
+      "loss": 0.3626,
+      "step": 2366
+    },
+    {
+      "epoch": 0.84,
+      "learning_rate": 2.723854579942955e-05,
+      "loss": 0.4072,
+      "step": 2367
+    },
+    {
+      "epoch": 0.84,
+      "learning_rate": 2.712249148295696e-05,
+      "loss": 0.5232,
+      "step": 2368
+    },
+    {
+      "epoch": 0.84,
+      "learning_rate": 2.700666694275642e-05,
+      "loss": 0.6428,
+      "step": 2369
+    },
+    {
+      "epoch": 0.84,
+      "learning_rate": 2.6891072332773415e-05,
+      "loss": 0.3952,
+      "step": 2370
+    },
+    {
+      "epoch": 0.84,
+      "learning_rate": 2.6775707806648064e-05,
+      "loss": 0.3903,
+      "step": 2371
+    },
+    {
+      "epoch": 0.84,
+      "learning_rate": 2.6660573517714516e-05,
+      "loss": 0.3047,
+      "step": 2372
+    },
+    {
+      "epoch": 0.84,
+      "learning_rate": 2.654566961900089e-05,
+      "loss": 0.3514,
+      "step": 2373
+    },
+    {
+      "epoch": 0.84,
+      "learning_rate": 2.6430996263229157e-05,
+      "loss": 0.6551,
+      "step": 2374
+    },
+    {
+      "epoch": 0.84,
+      "learning_rate": 2.6316553602814862e-05,
+      "loss": 0.3974,
+      "step": 2375
+    },
+    {
+      "epoch": 0.84,
+      "learning_rate": 2.620234178986687e-05,
+      "loss": 0.9819,
+      "step": 2376
+    },
+    {
+      "epoch": 0.84,
+      "learning_rate": 2.6088360976187233e-05,
+      "loss": 0.4284,
+      "step": 2377
+    },
+    {
+      "epoch": 0.84,
+      "learning_rate": 2.597461131327108e-05,
+      "loss": 0.8827,
+      "step": 2378
+    },
+    {
+      "epoch": 0.84,
+      "learning_rate": 2.5861092952306254e-05,
+      "loss": 0.3748,
+      "step": 2379
+    },
+    {
+      "epoch": 0.84,
+      "learning_rate": 2.5747806044172972e-05,
+      "loss": 0.5999,
+      "step": 2380
+    },
+    {
+      "epoch": 0.84,
+      "eval_loss": 0.5311623215675354,
+      "eval_runtime": 52.9417,
+      "eval_samples_per_second": 30.77,
+      "eval_steps_per_second": 30.77,
+      "step": 2380
+    },
+    {
+      "epoch": 0.84,
+      "learning_rate": 2.5634750739444148e-05,
+      "loss": 0.5711,
+      "step": 2381
+    },
+    {
+      "epoch": 0.84,
+      "learning_rate": 2.5521927188384665e-05,
+      "loss": 0.4514,
+      "step": 2382
+    },
+    {
+      "epoch": 0.84,
+      "learning_rate": 2.540933554095144e-05,
+      "loss": 0.9191,
+      "step": 2383
+    },
+    {
+      "epoch": 0.84,
+      "learning_rate": 2.52969759467931e-05,
+      "loss": 0.4108,
+      "step": 2384
+    },
+    {
+      "epoch": 0.84,
+      "learning_rate": 2.5184848555249895e-05,
+      "loss": 0.5009,
+      "step": 2385
+    },
+    {
+      "epoch": 0.84,
+      "learning_rate": 2.507295351535348e-05,
+      "loss": 0.2644,
+      "step": 2386
+    },
+    {
+      "epoch": 0.84,
+      "learning_rate": 2.4961290975826557e-05,
+      "loss": 0.5759,
+      "step": 2387
+    },
+    {
+      "epoch": 0.85,
+      "learning_rate": 2.4849861085082983e-05,
+      "loss": 0.4014,
+      "step": 2388
+    },
+    {
+      "epoch": 0.85,
+      "learning_rate": 2.473866399122733e-05,
+      "loss": 0.9743,
+      "step": 2389
+    },
+    {
+      "epoch": 0.85,
+      "learning_rate": 2.4627699842054554e-05,
+      "loss": 0.4819,
+      "step": 2390
+    },
+    {
+      "epoch": 0.85,
+      "learning_rate": 2.4516968785050344e-05,
+      "loss": 0.5736,
+      "step": 2391
+    },
+    {
+      "epoch": 0.85,
+      "learning_rate": 2.4406470967390348e-05,
+      "loss": 0.5788,
+      "step": 2392
+    },
+    {
+      "epoch": 0.85,
+      "learning_rate": 2.429620653594027e-05,
+      "loss": 0.8115,
+      "step": 2393
+    },
+    {
+      "epoch": 0.85,
+      "learning_rate": 2.4186175637255603e-05,
+      "loss": 0.3695,
+      "step": 2394
+    },
+    {
+      "epoch": 0.85,
+      "learning_rate": 2.4076378417581434e-05,
+      "loss": 0.6826,
+      "step": 2395
+    },
+    {
+      "epoch": 0.85,
+      "learning_rate": 2.396681502285234e-05,
+      "loss": 0.5183,
+      "step": 2396
+    },
+    {
+      "epoch": 0.85,
+      "learning_rate": 2.3857485598691963e-05,
+      "loss": 0.1947,
+      "step": 2397
+    },
+    {
+      "epoch": 0.85,
+      "learning_rate": 2.3748390290413202e-05,
+      "loss": 0.2864,
+      "step": 2398
+    },
+    {
+      "epoch": 0.85,
+      "learning_rate": 2.3639529243017512e-05,
+      "loss": 0.5579,
+      "step": 2399
+    },
+    {
+      "epoch": 0.85,
+      "learning_rate": 2.3530902601195147e-05,
+      "loss": 0.2757,
+      "step": 2400
+    },
+    {
+      "epoch": 0.85,
+      "eval_loss": 0.5292481184005737,
+      "eval_runtime": 52.9406,
+      "eval_samples_per_second": 30.77,
+      "eval_steps_per_second": 30.77,
+      "step": 2400
+    },
+    {
+      "epoch": 0.85,
+      "learning_rate": 2.3422510509324823e-05,
+      "loss": 0.6122,
+      "step": 2401
+    },
+    {
+      "epoch": 0.85,
+      "learning_rate": 2.3314353111473432e-05,
+      "loss": 0.5522,
+      "step": 2402
+    },
+    {
+      "epoch": 0.85,
+      "learning_rate": 2.3206430551395976e-05,
+      "loss": 0.2782,
+      "step": 2403
+    },
+    {
+      "epoch": 0.85,
+      "learning_rate": 2.3098742972535293e-05,
+      "loss": 0.3066,
+      "step": 2404
+    },
+    {
+      "epoch": 0.85,
+      "learning_rate": 2.2991290518021936e-05,
+      "loss": 0.447,
+      "step": 2405
+    },
+    {
+      "epoch": 0.85,
+      "learning_rate": 2.2884073330673907e-05,
+      "loss": 0.3061,
+      "step": 2406
+    },
+    {
+      "epoch": 0.85,
+      "learning_rate": 2.277709155299652e-05,
+      "loss": 0.3983,
+      "step": 2407
+    },
+    {
+      "epoch": 0.85,
+      "learning_rate": 2.2670345327182307e-05,
+      "loss": 0.3899,
+      "step": 2408
+    },
+    {
+      "epoch": 0.85,
+      "learning_rate": 2.2563834795110504e-05,
+      "loss": 0.6096,
+      "step": 2409
+    },
+    {
+      "epoch": 0.85,
+      "learning_rate": 2.245756009834723e-05,
+      "loss": 0.2354,
+      "step": 2410
+    },
+    {
+      "epoch": 0.85,
+      "learning_rate": 2.2351521378145156e-05,
+      "loss": 0.8637,
+      "step": 2411
+    },
+    {
+      "epoch": 0.85,
+      "learning_rate": 2.224571877544328e-05,
+      "loss": 0.5243,
+      "step": 2412
+    },
+    {
+      "epoch": 0.85,
+      "learning_rate": 2.2140152430866735e-05,
+      "loss": 0.4618,
+      "step": 2413
+    },
+    {
+      "epoch": 0.85,
+      "learning_rate": 2.203482248472668e-05,
+      "loss": 0.5446,
+      "step": 2414
+    },
+    {
+      "epoch": 0.85,
+      "learning_rate": 2.1929729077020046e-05,
+      "loss": 0.3771,
+      "step": 2415
+    },
+    {
+      "epoch": 0.86,
+      "learning_rate": 2.1824872347429383e-05,
+      "loss": 0.591,
+      "step": 2416
+    },
+    {
+      "epoch": 0.86,
+      "learning_rate": 2.172025243532265e-05,
+      "loss": 0.4789,
+      "step": 2417
+    },
+    {
+      "epoch": 0.86,
+      "learning_rate": 2.161586947975314e-05,
+      "loss": 0.3093,
+      "step": 2418
+    },
+    {
+      "epoch": 0.86,
+      "learning_rate": 2.1511723619459036e-05,
+      "loss": 0.4259,
+      "step": 2419
+    },
+    {
+      "epoch": 0.86,
+      "learning_rate": 2.1407814992863483e-05,
+      "loss": 0.322,
+      "step": 2420
+    },
+    {
+      "epoch": 0.86,
+      "eval_loss": 0.5298978090286255,
+      "eval_runtime": 52.9281,
+      "eval_samples_per_second": 30.778,
+      "eval_steps_per_second": 30.778,
+      "step": 2420
+    },
+    {
+      "epoch": 0.86,
+      "learning_rate": 2.1304143738074345e-05,
+      "loss": 0.495,
+      "step": 2421
+    },
+    {
+      "epoch": 0.86,
+      "learning_rate": 2.120070999288397e-05,
+      "loss": 0.5988,
+      "step": 2422
+    },
+    {
+      "epoch": 0.86,
+      "learning_rate": 2.109751389476895e-05,
+      "loss": 0.3414,
+      "step": 2423
+    },
+    {
+      "epoch": 0.86,
+      "learning_rate": 2.0994555580890097e-05,
+      "loss": 0.2744,
+      "step": 2424
+    },
+    {
+      "epoch": 0.86,
+      "learning_rate": 2.0891835188092144e-05,
+      "loss": 0.5426,
+      "step": 2425
+    },
+    {
+      "epoch": 0.86,
+      "learning_rate": 2.078935285290362e-05,
+      "loss": 0.4813,
+      "step": 2426
+    },
+    {
+      "epoch": 0.86,
+      "learning_rate": 2.0687108711536562e-05,
+      "loss": 0.3499,
+      "step": 2427
+    },
+    {
+      "epoch": 0.86,
+      "learning_rate": 2.05851028998866e-05,
+      "loss": 0.48,
+      "step": 2428
+    },
+    {
+      "epoch": 0.86,
+      "learning_rate": 2.048333555353239e-05,
+      "loss": 0.4013,
+      "step": 2429
+    },
+    {
+      "epoch": 0.86,
+      "learning_rate": 2.0381806807735692e-05,
+      "loss": 0.3352,
+      "step": 2430
+    },
+    {
+      "epoch": 0.86,
+      "learning_rate": 2.0280516797441252e-05,
+      "loss": 0.4557,
+      "step": 2431
+    },
+    {
+      "epoch": 0.86,
+      "learning_rate": 2.0179465657276398e-05,
+      "loss": 0.3334,
+      "step": 2432
+    },
+    {
+      "epoch": 0.86,
+      "learning_rate": 2.0078653521550982e-05,
+      "loss": 0.5975,
+      "step": 2433
+    },
+    {
+      "epoch": 0.86,
+      "learning_rate": 1.997808052425718e-05,
+      "loss": 0.5103,
+      "step": 2434
+    },
+    {
+      "epoch": 0.86,
+      "learning_rate": 1.9877746799069375e-05,
+      "loss": 0.3557,
+      "step": 2435
+    },
+    {
+      "epoch": 0.86,
+      "learning_rate": 1.9777652479343867e-05,
+      "loss": 1.1199,
+      "step": 2436
+    },
+    {
+      "epoch": 0.86,
+      "learning_rate": 1.9677797698118726e-05,
+      "loss": 0.2388,
+      "step": 2437
+    },
+    {
+      "epoch": 0.86,
+      "learning_rate": 1.957818258811388e-05,
+      "loss": 0.4032,
+      "step": 2438
+    },
+    {
+      "epoch": 0.86,
+      "learning_rate": 1.9478807281730327e-05,
+      "loss": 0.8172,
+      "step": 2439
+    },
+    {
+      "epoch": 0.86,
+      "learning_rate": 1.937967191105061e-05,
+      "loss": 0.4368,
+      "step": 2440
+    },
+    {
+      "epoch": 0.86,
+      "eval_loss": 0.5317589640617371,
+      "eval_runtime": 52.939,
+      "eval_samples_per_second": 30.771,
+      "eval_steps_per_second": 30.771,
+      "step": 2440
+    },
+    {
+      "epoch": 0.86,
+      "learning_rate": 1.9280776607838315e-05,
+      "loss": 0.5716,
+      "step": 2441
+    },
+    {
+      "epoch": 0.86,
+      "learning_rate": 1.9182121503537887e-05,
+      "loss": 0.3422,
+      "step": 2442
+    },
+    {
+      "epoch": 0.86,
+      "learning_rate": 1.9083706729274596e-05,
+      "loss": 0.582,
+      "step": 2443
+    },
+    {
+      "epoch": 0.86,
+      "learning_rate": 1.8985532415854123e-05,
+      "loss": 0.4689,
+      "step": 2444
+    },
+    {
+      "epoch": 0.87,
+      "learning_rate": 1.8887598693762776e-05,
+      "loss": 0.8763,
+      "step": 2445
+    },
+    {
+      "epoch": 0.87,
+      "learning_rate": 1.8789905693166944e-05,
+      "loss": 0.474,
+      "step": 2446
+    },
+    {
+      "epoch": 0.87,
+      "learning_rate": 1.8692453543913048e-05,
+      "loss": 0.4014,
+      "step": 2447
+    },
+    {
+      "epoch": 0.87,
+      "learning_rate": 1.8595242375527523e-05,
+      "loss": 0.2482,
+      "step": 2448
+    },
+    {
+      "epoch": 0.87,
+      "learning_rate": 1.8498272317216324e-05,
+      "loss": 0.5269,
+      "step": 2449
+    },
+    {
+      "epoch": 0.87,
+      "learning_rate": 1.8401543497865025e-05,
+      "loss": 0.4592,
+      "step": 2450
+    },
+    {
+      "epoch": 0.87,
+      "learning_rate": 1.8305056046038648e-05,
+      "loss": 0.5682,
+      "step": 2451
+    },
+    {
+      "epoch": 0.87,
+      "learning_rate": 1.820881008998132e-05,
+      "loss": 0.6269,
+      "step": 2452
+    },
+    {
+      "epoch": 0.87,
+      "learning_rate": 1.8112805757616203e-05,
+      "loss": 0.1605,
+      "step": 2453
+    },
+    {
+      "epoch": 0.87,
+      "learning_rate": 1.80170431765452e-05,
+      "loss": 0.421,
+      "step": 2454
+    },
+    {
+      "epoch": 0.87,
+      "learning_rate": 1.7921522474049146e-05,
+      "loss": 0.4538,
+      "step": 2455
+    },
+    {
+      "epoch": 0.87,
+      "learning_rate": 1.7826243777087148e-05,
+      "loss": 0.3548,
+      "step": 2456
+    },
+    {
+      "epoch": 0.87,
+      "learning_rate": 1.7731207212296817e-05,
+      "loss": 0.4007,
+      "step": 2457
+    },
+    {
+      "epoch": 0.87,
+      "learning_rate": 1.763641290599385e-05,
+      "loss": 0.3668,
+      "step": 2458
+    },
+    {
+      "epoch": 0.87,
+      "learning_rate": 1.7541860984171965e-05,
+      "loss": 0.5345,
+      "step": 2459
+    },
+    {
+      "epoch": 0.87,
+      "learning_rate": 1.7447551572502706e-05,
+      "loss": 0.318,
+      "step": 2460
+    },
+    {
+      "epoch": 0.87,
+      "eval_loss": 0.5318650007247925,
+      "eval_runtime": 52.9227,
+      "eval_samples_per_second": 30.781,
+      "eval_steps_per_second": 30.781,
+      "step": 2460
+    },
+    {
+      "epoch": 0.87,
+      "learning_rate": 1.7353484796335405e-05,
+      "loss": 0.317,
+      "step": 2461
+    },
+    {
+      "epoch": 0.87,
+      "learning_rate": 1.7259660780696763e-05,
+      "loss": 0.3965,
+      "step": 2462
+    },
+    {
+      "epoch": 0.87,
+      "learning_rate": 1.716607965029091e-05,
+      "loss": 0.2299,
+      "step": 2463
+    },
+    {
+      "epoch": 0.87,
+      "learning_rate": 1.7072741529499003e-05,
+      "loss": 0.6055,
+      "step": 2464
+    },
+    {
+      "epoch": 0.87,
+      "learning_rate": 1.6979646542379423e-05,
+      "loss": 0.5168,
+      "step": 2465
+    },
+    {
+      "epoch": 0.87,
+      "learning_rate": 1.688679481266724e-05,
+      "loss": 0.5985,
+      "step": 2466
+    },
+    {
+      "epoch": 0.87,
+      "learning_rate": 1.6794186463774287e-05,
+      "loss": 0.3809,
+      "step": 2467
+    },
+    {
+      "epoch": 0.87,
+      "learning_rate": 1.6701821618788838e-05,
+      "loss": 0.2401,
+      "step": 2468
+    },
+    {
+      "epoch": 0.87,
+      "learning_rate": 1.6609700400475603e-05,
+      "loss": 0.5972,
+      "step": 2469
+    },
+    {
+      "epoch": 0.87,
+      "learning_rate": 1.651782293127535e-05,
+      "loss": 0.3418,
+      "step": 2470
+    },
+    {
+      "epoch": 0.87,
+      "learning_rate": 1.6426189333305085e-05,
+      "loss": 1.0226,
+      "step": 2471
+    },
+    {
+      "epoch": 0.87,
+      "learning_rate": 1.6334799728357498e-05,
+      "loss": 0.3939,
+      "step": 2472
+    },
+    {
+      "epoch": 0.88,
+      "learning_rate": 1.624365423790104e-05,
+      "loss": 0.3762,
+      "step": 2473
+    },
+    {
+      "epoch": 0.88,
+      "learning_rate": 1.6152752983079657e-05,
+      "loss": 0.2775,
+      "step": 2474
+    },
+    {
+      "epoch": 0.88,
+      "learning_rate": 1.6062096084712784e-05,
+      "loss": 0.5808,
+      "step": 2475
+    },
+    {
+      "epoch": 0.88,
+      "learning_rate": 1.597168366329498e-05,
+      "loss": 0.5958,
+      "step": 2476
+    },
+    {
+      "epoch": 0.88,
+      "learning_rate": 1.5881515838995887e-05,
+      "loss": 0.2538,
+      "step": 2477
+    },
+    {
+      "epoch": 0.88,
+      "learning_rate": 1.5791592731660044e-05,
+      "loss": 0.3754,
+      "step": 2478
+    },
+    {
+      "epoch": 0.88,
+      "learning_rate": 1.5701914460806755e-05,
+      "loss": 0.2358,
+      "step": 2479
+    },
+    {
+      "epoch": 0.88,
+      "learning_rate": 1.5612481145629854e-05,
+      "loss": 0.7192,
+      "step": 2480
+    },
+    {
+      "epoch": 0.88,
+      "eval_loss": 0.5306629538536072,
+      "eval_runtime": 52.943,
+      "eval_samples_per_second": 30.769,
+      "eval_steps_per_second": 30.769,
+      "step": 2480
+    },
+    {
+      "epoch": 0.88,
+      "learning_rate": 1.5523292904997698e-05,
+      "loss": 0.5778,
+      "step": 2481
+    },
+    {
+      "epoch": 0.88,
+      "learning_rate": 1.54343498574528e-05,
+      "loss": 0.3501,
+      "step": 2482
+    },
+    {
+      "epoch": 0.88,
+      "learning_rate": 1.534565212121186e-05,
+      "loss": 0.5881,
+      "step": 2483
+    },
+    {
+      "epoch": 0.88,
+      "learning_rate": 1.525719981416538e-05,
+      "loss": 0.6251,
+      "step": 2484
+    },
+    {
+      "epoch": 0.88,
+      "learning_rate": 1.5168993053877868e-05,
+      "loss": 0.3247,
+      "step": 2485
+    },
+    {
+      "epoch": 0.88,
+      "learning_rate": 1.5081031957587322e-05,
+      "loss": 0.4203,
+      "step": 2486
+    },
+    {
+      "epoch": 0.88,
+      "learning_rate": 1.4993316642205291e-05,
+      "loss": 0.5663,
+      "step": 2487
+    },
+    {
+      "epoch": 0.88,
+      "learning_rate": 1.4905847224316606e-05,
+      "loss": 0.6298,
+      "step": 2488
+    },
+    {
+      "epoch": 0.88,
+      "learning_rate": 1.481862382017929e-05,
+      "loss": 0.672,
+      "step": 2489
+    },
+    {
+      "epoch": 0.88,
+      "learning_rate": 1.4731646545724365e-05,
+      "loss": 0.2587,
+      "step": 2490
+    },
+    {
+      "epoch": 0.88,
+      "learning_rate": 1.4644915516555757e-05,
+      "loss": 0.578,
+      "step": 2491
+    },
+    {
+      "epoch": 0.88,
+      "learning_rate": 1.4558430847950099e-05,
+      "loss": 0.4526,
+      "step": 2492
+    },
+    {
+      "epoch": 0.88,
+      "learning_rate": 1.4472192654856464e-05,
+      "loss": 0.3769,
+      "step": 2493
+    },
+    {
+      "epoch": 0.88,
+      "learning_rate": 1.4386201051896452e-05,
+      "loss": 0.4564,
+      "step": 2494
+    },
+    {
+      "epoch": 0.88,
+      "learning_rate": 1.4300456153363905e-05,
+      "loss": 0.323,
+      "step": 2495
+    },
+    {
+      "epoch": 0.88,
+      "learning_rate": 1.4214958073224748e-05,
+      "loss": 0.593,
+      "step": 2496
+    },
+    {
+      "epoch": 0.88,
+      "learning_rate": 1.4129706925116792e-05,
+      "loss": 0.4349,
+      "step": 2497
+    },
+    {
+      "epoch": 0.88,
+      "learning_rate": 1.4044702822349731e-05,
+      "loss": 0.5818,
+      "step": 2498
+    },
+    {
+      "epoch": 0.88,
+      "learning_rate": 1.3959945877904857e-05,
+      "loss": 0.5079,
+      "step": 2499
+    },
+    {
+      "epoch": 0.88,
+      "learning_rate": 1.3875436204434922e-05,
+      "loss": 0.3465,
+      "step": 2500
+    },
+    {
+      "epoch": 0.88,
+      "eval_loss": 0.5315174460411072,
+      "eval_runtime": 52.9259,
+      "eval_samples_per_second": 30.779,
+      "eval_steps_per_second": 30.779,
+      "step": 2500
+    },
+    {
+      "epoch": 0.89,
+      "learning_rate": 1.3791173914264121e-05,
+      "loss": 0.4938,
+      "step": 2501
+    },
+    {
+      "epoch": 0.89,
+      "learning_rate": 1.3707159119387824e-05,
+      "loss": 0.3731,
+      "step": 2502
+    },
+    {
+      "epoch": 0.89,
+      "learning_rate": 1.3623391931472352e-05,
+      "loss": 0.4999,
+      "step": 2503
+    },
+    {
+      "epoch": 0.89,
+      "learning_rate": 1.3539872461854952e-05,
+      "loss": 0.7244,
+      "step": 2504
+    },
+    {
+      "epoch": 0.89,
+      "learning_rate": 1.3456600821543741e-05,
+      "loss": 0.229,
+      "step": 2505
+    },
+    {
+      "epoch": 0.89,
+      "learning_rate": 1.337357712121734e-05,
+      "loss": 0.5043,
+      "step": 2506
+    },
+    {
+      "epoch": 0.89,
+      "learning_rate": 1.3290801471224835e-05,
+      "loss": 0.3926,
+      "step": 2507
+    },
+    {
+      "epoch": 0.89,
+      "learning_rate": 1.3208273981585594e-05,
+      "loss": 0.406,
+      "step": 2508
+    },
+    {
+      "epoch": 0.89,
+      "learning_rate": 1.3125994761989235e-05,
+      "loss": 0.4126,
+      "step": 2509
+    },
+    {
+      "epoch": 0.89,
+      "learning_rate": 1.3043963921795277e-05,
+      "loss": 0.5851,
+      "step": 2510
+    },
+    {
+      "epoch": 0.89,
+      "learning_rate": 1.296218157003326e-05,
+      "loss": 0.506,
+      "step": 2511
+    },
+    {
+      "epoch": 0.89,
+      "learning_rate": 1.2880647815402347e-05,
+      "loss": 0.8374,
+      "step": 2512
+    },
+    {
+      "epoch": 0.89,
+      "learning_rate": 1.279936276627125e-05,
+      "loss": 0.2698,
+      "step": 2513
+    },
+    {
+      "epoch": 0.89,
+      "learning_rate": 1.271832653067817e-05,
+      "loss": 0.3671,
+      "step": 2514
+    },
+    {
+      "epoch": 0.89,
+      "learning_rate": 1.2637539216330664e-05,
+      "loss": 0.4274,
+      "step": 2515
+    },
+    {
+      "epoch": 0.89,
+      "learning_rate": 1.2557000930605367e-05,
+      "loss": 0.5943,
+      "step": 2516
+    },
+    {
+      "epoch": 0.89,
+      "learning_rate": 1.2476711780547944e-05,
+      "loss": 0.4514,
+      "step": 2517
+    },
+    {
+      "epoch": 0.89,
+      "learning_rate": 1.23966718728729e-05,
+      "loss": 1.139,
+      "step": 2518
+    },
+    {
+      "epoch": 0.89,
+      "learning_rate": 1.231688131396349e-05,
+      "loss": 0.5025,
+      "step": 2519
+    },
+    {
+      "epoch": 0.89,
+      "learning_rate": 1.2237340209871528e-05,
+      "loss": 0.4098,
+      "step": 2520
+    },
+    {
+      "epoch": 0.89,
+      "eval_loss": 0.5312186479568481,
+      "eval_runtime": 52.9427,
+      "eval_samples_per_second": 30.769,
+      "eval_steps_per_second": 30.769,
+      "step": 2520
+    },
+    {
+      "epoch": 0.89,
+      "learning_rate": 1.2158048666317356e-05,
+      "loss": 0.4276,
+      "step": 2521
+    },
+    {
+      "epoch": 0.89,
+      "learning_rate": 1.2079006788689562e-05,
+      "loss": 0.2868,
+      "step": 2522
+    },
+    {
+      "epoch": 0.89,
+      "learning_rate": 1.2000214682044797e-05,
+      "loss": 0.2387,
+      "step": 2523
+    },
+    {
+      "epoch": 0.89,
+      "learning_rate": 1.1921672451107847e-05,
+      "loss": 0.3691,
+      "step": 2524
+    },
+    {
+      "epoch": 0.89,
+      "learning_rate": 1.1843380200271426e-05,
+      "loss": 0.249,
+      "step": 2525
+    },
+    {
+      "epoch": 0.89,
+      "learning_rate": 1.1765338033595896e-05,
+      "loss": 0.2795,
+      "step": 2526
+    },
+    {
+      "epoch": 0.89,
+      "learning_rate": 1.1687546054809261e-05,
+      "loss": 0.4248,
+      "step": 2527
+    },
+    {
+      "epoch": 0.89,
+      "learning_rate": 1.1610004367306949e-05,
+      "loss": 0.5102,
+      "step": 2528
+    },
+    {
+      "epoch": 0.9,
+      "learning_rate": 1.1532713074151823e-05,
+      "loss": 0.4377,
+      "step": 2529
+    },
+    {
+      "epoch": 0.9,
+      "learning_rate": 1.1455672278073803e-05,
+      "loss": 0.6596,
+      "step": 2530
+    },
+    {
+      "epoch": 0.9,
+      "learning_rate": 1.137888208147e-05,
+      "loss": 0.4345,
+      "step": 2531
+    },
+    {
+      "epoch": 0.9,
+      "learning_rate": 1.130234258640439e-05,
+      "loss": 0.3382,
+      "step": 2532
+    },
+    {
+      "epoch": 0.9,
+      "learning_rate": 1.1226053894607692e-05,
+      "loss": 0.507,
+      "step": 2533
+    },
+    {
+      "epoch": 0.9,
+      "learning_rate": 1.1150016107477279e-05,
+      "loss": 0.7481,
+      "step": 2534
+    },
+    {
+      "epoch": 0.9,
+      "learning_rate": 1.1074229326077157e-05,
+      "loss": 0.3744,
+      "step": 2535
+    },
+    {
+      "epoch": 0.9,
+      "learning_rate": 1.0998693651137592e-05,
+      "loss": 0.4978,
+      "step": 2536
+    },
+    {
+      "epoch": 0.9,
+      "learning_rate": 1.0923409183055144e-05,
+      "loss": 0.665,
+      "step": 2537
+    },
+    {
+      "epoch": 0.9,
+      "learning_rate": 1.0848376021892482e-05,
+      "loss": 0.4645,
+      "step": 2538
+    },
+    {
+      "epoch": 0.9,
+      "learning_rate": 1.077359426737825e-05,
+      "loss": 1.1302,
+      "step": 2539
+    },
+    {
+      "epoch": 0.9,
+      "learning_rate": 1.0699064018906879e-05,
+      "loss": 0.682,
+      "step": 2540
+    },
+    {
+      "epoch": 0.9,
+      "eval_loss": 0.5308239459991455,
+      "eval_runtime": 52.9237,
+      "eval_samples_per_second": 30.78,
+      "eval_steps_per_second": 30.78,
+      "step": 2540
+    },
+    {
+      "epoch": 0.9,
+      "learning_rate": 1.0624785375538726e-05,
+      "loss": 0.4881,
+      "step": 2541
+    },
+    {
+      "epoch": 0.9,
+      "learning_rate": 1.0550758435999463e-05,
+      "loss": 0.4673,
+      "step": 2542
+    },
+    {
+      "epoch": 0.9,
+      "learning_rate": 1.0476983298680387e-05,
+      "loss": 0.4652,
+      "step": 2543
+    },
+    {
+      "epoch": 0.9,
+      "learning_rate": 1.0403460061638037e-05,
+      "loss": 0.3047,
+      "step": 2544
+    },
+    {
+      "epoch": 0.9,
+      "learning_rate": 1.0330188822594223e-05,
+      "loss": 0.4803,
+      "step": 2545
+    },
+    {
+      "epoch": 0.9,
+      "learning_rate": 1.0257169678935751e-05,
+      "loss": 0.3209,
+      "step": 2546
+    },
+    {
+      "epoch": 0.9,
+      "learning_rate": 1.0184402727714393e-05,
+      "loss": 0.5574,
+      "step": 2547
+    },
+    {
+      "epoch": 0.9,
+      "learning_rate": 1.0111888065646713e-05,
+      "loss": 0.3298,
+      "step": 2548
+    },
+    {
+      "epoch": 0.9,
+      "learning_rate": 1.0039625789113949e-05,
+      "loss": 0.3617,
+      "step": 2549
+    },
+    {
+      "epoch": 0.9,
+      "learning_rate": 9.96761599416185e-06,
+      "loss": 0.9207,
+      "step": 2550
+    },
+    {
+      "epoch": 0.9,
+      "learning_rate": 9.8958587765007e-06,
+      "loss": 0.2613,
+      "step": 2551
+    },
+    {
+      "epoch": 0.9,
+      "learning_rate": 9.824354231504961e-06,
+      "loss": 0.3195,
+      "step": 2552
+    },
+    {
+      "epoch": 0.9,
+      "learning_rate": 9.753102454213258e-06,
+      "loss": 0.5087,
+      "step": 2553
+    },
+    {
+      "epoch": 0.9,
+      "learning_rate": 9.682103539328346e-06,
+      "loss": 0.2584,
+      "step": 2554
+    },
+    {
+      "epoch": 0.9,
+      "learning_rate": 9.611357581216829e-06,
+      "loss": 0.4755,
+      "step": 2555
+    },
+    {
+      "epoch": 0.9,
+      "learning_rate": 9.540864673909134e-06,
+      "loss": 0.7704,
+      "step": 2556
+    },
+    {
+      "epoch": 0.9,
+      "learning_rate": 9.470624911099314e-06,
+      "loss": 0.4801,
+      "step": 2557
+    },
+    {
+      "epoch": 0.91,
+      "learning_rate": 9.400638386144977e-06,
+      "loss": 0.481,
+      "step": 2558
+    },
+    {
+      "epoch": 0.91,
+      "learning_rate": 9.330905192067141e-06,
+      "loss": 0.3408,
+      "step": 2559
+    },
+    {
+      "epoch": 0.91,
+      "learning_rate": 9.261425421550108e-06,
+      "loss": 1.4551,
+      "step": 2560
+    },
+    {
+      "epoch": 0.91,
+      "eval_loss": 0.5310149788856506,
+      "eval_runtime": 52.937,
+      "eval_samples_per_second": 30.772,
+      "eval_steps_per_second": 30.772,
+      "step": 2560
+    },
+    {
+      "epoch": 0.91,
+      "learning_rate": 9.19219916694143e-06,
+      "loss": 0.766,
+      "step": 2561
+    },
+    {
+      "epoch": 0.91,
+      "learning_rate": 9.12322652025157e-06,
+      "loss": 0.9035,
+      "step": 2562
+    },
+    {
+      "epoch": 0.91,
+      "learning_rate": 9.05450757315398e-06,
+      "loss": 0.5042,
+      "step": 2563
+    },
+    {
+      "epoch": 0.91,
+      "learning_rate": 8.986042416984886e-06,
+      "loss": 0.6808,
+      "step": 2564
+    },
+    {
+      "epoch": 0.91,
+      "learning_rate": 8.917831142743294e-06,
+      "loss": 0.2423,
+      "step": 2565
+    },
+    {
+      "epoch": 0.91,
+      "learning_rate": 8.849873841090661e-06,
+      "loss": 0.5144,
+      "step": 2566
+    },
+    {
+      "epoch": 0.91,
+      "learning_rate": 8.782170602350914e-06,
+      "loss": 0.3212,
+      "step": 2567
+    },
+    {
+      "epoch": 0.91,
+      "learning_rate": 8.714721516510272e-06,
+      "loss": 0.4517,
+      "step": 2568
+    },
+    {
+      "epoch": 0.91,
+      "learning_rate": 8.647526673217198e-06,
+      "loss": 0.4034,
+      "step": 2569
+    },
+    {
+      "epoch": 0.91,
+      "learning_rate": 8.580586161782189e-06,
+      "loss": 0.375,
+      "step": 2570
+    },
+    {
+      "epoch": 0.91,
+      "learning_rate": 8.513900071177805e-06,
+      "loss": 0.4475,
+      "step": 2571
+    },
+    {
+      "epoch": 0.91,
+      "learning_rate": 8.447468490038302e-06,
+      "loss": 0.3607,
+      "step": 2572
+    },
+    {
+      "epoch": 0.91,
+      "learning_rate": 8.381291506659738e-06,
+      "loss": 0.4224,
+      "step": 2573
+    },
+    {
+      "epoch": 0.91,
+      "learning_rate": 8.315369208999757e-06,
+      "loss": 0.4382,
+      "step": 2574
+    },
+    {
+      "epoch": 0.91,
+      "learning_rate": 8.249701684677557e-06,
+      "loss": 0.3184,
+      "step": 2575
+    },
+    {
+      "epoch": 0.91,
+      "learning_rate": 8.184289020973612e-06,
+      "loss": 0.5335,
+      "step": 2576
+    },
+    {
+      "epoch": 0.91,
+      "learning_rate": 8.11913130482973e-06,
+      "loss": 0.3504,
+      "step": 2577
+    },
+    {
+      "epoch": 0.91,
+      "learning_rate": 8.054228622848813e-06,
+      "loss": 0.3975,
+      "step": 2578
+    },
+    {
+      "epoch": 0.91,
+      "learning_rate": 7.989581061294814e-06,
+      "loss": 0.597,
+      "step": 2579
+    },
+    {
+      "epoch": 0.91,
+      "learning_rate": 7.925188706092557e-06,
+      "loss": 0.4396,
+      "step": 2580
+    },
+    {
+      "epoch": 0.91,
+      "eval_loss": 0.530106782913208,
+      "eval_runtime": 52.9241,
+      "eval_samples_per_second": 30.78,
+      "eval_steps_per_second": 30.78,
+      "step": 2580
+    },
+    {
+      "epoch": 0.91,
+      "learning_rate": 7.861051642827799e-06,
+      "loss": 0.2815,
+      "step": 2581
+    },
+    {
+      "epoch": 0.91,
+      "learning_rate": 7.797169956746774e-06,
+      "loss": 1.021,
+      "step": 2582
+    },
+    {
+      "epoch": 0.91,
+      "learning_rate": 7.733543732756453e-06,
+      "loss": 0.3395,
+      "step": 2583
+    },
+    {
+      "epoch": 0.91,
+      "learning_rate": 7.670173055424169e-06,
+      "loss": 0.2649,
+      "step": 2584
+    },
+    {
+      "epoch": 0.91,
+      "learning_rate": 7.607058008977697e-06,
+      "loss": 0.3965,
+      "step": 2585
+    },
+    {
+      "epoch": 0.92,
+      "learning_rate": 7.544198677304981e-06,
+      "loss": 0.4056,
+      "step": 2586
+    },
+    {
+      "epoch": 0.92,
+      "learning_rate": 7.481595143954012e-06,
+      "loss": 0.4364,
+      "step": 2587
+    },
+    {
+      "epoch": 0.92,
+      "learning_rate": 7.419247492132964e-06,
+      "loss": 0.3943,
+      "step": 2588
+    },
+    {
+      "epoch": 0.92,
+      "learning_rate": 7.357155804709814e-06,
+      "loss": 0.3045,
+      "step": 2589
+    },
+    {
+      "epoch": 0.92,
+      "learning_rate": 7.295320164212261e-06,
+      "loss": 0.3947,
+      "step": 2590
+    },
+    {
+      "epoch": 0.92,
+      "learning_rate": 7.233740652827892e-06,
+      "loss": 0.388,
+      "step": 2591
+    },
+    {
+      "epoch": 0.92,
+      "learning_rate": 7.172417352403638e-06,
+      "loss": 0.3172,
+      "step": 2592
+    },
+    {
+      "epoch": 0.92,
+      "learning_rate": 7.111350344446011e-06,
+      "loss": 0.6692,
+      "step": 2593
+    },
+    {
+      "epoch": 0.92,
+      "learning_rate": 7.05053971012084e-06,
+      "loss": 0.7039,
+      "step": 2594
+    },
+    {
+      "epoch": 0.92,
+      "learning_rate": 6.989985530253274e-06,
+      "loss": 0.3859,
+      "step": 2595
+    },
+    {
+      "epoch": 0.92,
+      "learning_rate": 6.929687885327552e-06,
+      "loss": 0.6253,
+      "step": 2596
+    },
+    {
+      "epoch": 0.92,
+      "learning_rate": 6.8696468554868556e-06,
+      "loss": 0.3792,
+      "step": 2597
+    },
+    {
+      "epoch": 0.92,
+      "learning_rate": 6.809862520533439e-06,
+      "loss": 0.4281,
+      "step": 2598
+    },
+    {
+      "epoch": 0.92,
+      "learning_rate": 6.750334959928295e-06,
+      "loss": 0.4691,
+      "step": 2599
+    },
+    {
+      "epoch": 0.92,
+      "learning_rate": 6.691064252791157e-06,
+      "loss": 0.5932,
+      "step": 2600
+    },
+    {
+      "epoch": 0.92,
+      "eval_loss": 0.5297734141349792,
+      "eval_runtime": 52.9463,
+      "eval_samples_per_second": 30.767,
+      "eval_steps_per_second": 30.767,
+      "step": 2600
+    },
+    {
+      "epoch": 0.92,
+      "learning_rate": 6.632050477900364e-06,
+      "loss": 0.5969,
+      "step": 2601
+    },
+    {
+      "epoch": 0.92,
+      "learning_rate": 6.573293713692752e-06,
+      "loss": 0.8658,
+      "step": 2602
+    },
+    {
+      "epoch": 0.92,
+      "learning_rate": 6.514794038263539e-06,
+      "loss": 0.4267,
+      "step": 2603
+    },
+    {
+      "epoch": 0.92,
+      "learning_rate": 6.456551529366261e-06,
+      "loss": 0.6129,
+      "step": 2604
+    },
+    {
+      "epoch": 0.92,
+      "learning_rate": 6.398566264412664e-06,
+      "loss": 0.5465,
+      "step": 2605
+    },
+    {
+      "epoch": 0.92,
+      "learning_rate": 6.340838320472586e-06,
+      "loss": 0.4516,
+      "step": 2606
+    },
+    {
+      "epoch": 0.92,
+      "learning_rate": 6.283367774273785e-06,
+      "loss": 1.0218,
+      "step": 2607
+    },
+    {
+      "epoch": 0.92,
+      "learning_rate": 6.226154702201958e-06,
+      "loss": 0.5935,
+      "step": 2608
+    },
+    {
+      "epoch": 0.92,
+      "learning_rate": 6.169199180300633e-06,
+      "loss": 0.4559,
+      "step": 2609
+    },
+    {
+      "epoch": 0.92,
+      "learning_rate": 6.112501284270899e-06,
+      "loss": 1.2199,
+      "step": 2610
+    },
+    {
+      "epoch": 0.92,
+      "learning_rate": 6.0560610894715205e-06,
+      "loss": 0.5747,
+      "step": 2611
+    },
+    {
+      "epoch": 0.92,
+      "learning_rate": 5.999878670918757e-06,
+      "loss": 0.479,
+      "step": 2612
+    },
+    {
+      "epoch": 0.92,
+      "learning_rate": 5.943954103286142e-06,
+      "loss": 0.5662,
+      "step": 2613
+    },
+    {
+      "epoch": 0.93,
+      "learning_rate": 5.8882874609045955e-06,
+      "loss": 0.5304,
+      "step": 2614
+    },
+    {
+      "epoch": 0.93,
+      "learning_rate": 5.8328788177622215e-06,
+      "loss": 0.3283,
+      "step": 2615
+    },
+    {
+      "epoch": 0.93,
+      "learning_rate": 5.777728247504177e-06,
+      "loss": 0.5404,
+      "step": 2616
+    },
+    {
+      "epoch": 0.93,
+      "learning_rate": 5.722835823432537e-06,
+      "loss": 0.2612,
+      "step": 2617
+    },
+    {
+      "epoch": 0.93,
+      "learning_rate": 5.668201618506386e-06,
+      "loss": 0.3828,
+      "step": 2618
+    },
+    {
+      "epoch": 0.93,
+      "learning_rate": 5.6138257053415685e-06,
+      "loss": 0.3331,
+      "step": 2619
+    },
+    {
+      "epoch": 0.93,
+      "learning_rate": 5.559708156210608e-06,
+      "loss": 0.3978,
+      "step": 2620
+    },
+    {
+      "epoch": 0.93,
+      "eval_loss": 0.5291677713394165,
+      "eval_runtime": 52.9358,
+      "eval_samples_per_second": 30.773,
+      "eval_steps_per_second": 30.773,
+      "step": 2620
+    },
+    {
+      "epoch": 0.93,
+      "learning_rate": 5.50584904304261e-06,
+      "loss": 0.2821,
+      "step": 2621
+    },
+    {
+      "epoch": 0.93,
+      "learning_rate": 5.452248437423224e-06,
+      "loss": 0.3297,
+      "step": 2622
+    },
+    {
+      "epoch": 0.93,
+      "learning_rate": 5.3989064105945064e-06,
+      "loss": 0.3155,
+      "step": 2623
+    },
+    {
+      "epoch": 0.93,
+      "learning_rate": 5.345823033454767e-06,
+      "loss": 0.3792,
+      "step": 2624
+    },
+    {
+      "epoch": 0.93,
+      "learning_rate": 5.2929983765586555e-06,
+      "loss": 0.4224,
+      "step": 2625
+    },
+    {
+      "epoch": 0.93,
+      "learning_rate": 5.240432510116877e-06,
+      "loss": 0.4178,
+      "step": 2626
+    },
+    {
+      "epoch": 0.93,
+      "learning_rate": 5.188125503996122e-06,
+      "loss": 0.4006,
+      "step": 2627
+    },
+    {
+      "epoch": 0.93,
+      "learning_rate": 5.136077427719133e-06,
+      "loss": 0.4323,
+      "step": 2628
+    },
+    {
+      "epoch": 0.93,
+      "learning_rate": 5.084288350464416e-06,
+      "loss": 0.6762,
+      "step": 2629
+    },
+    {
+      "epoch": 0.93,
+      "learning_rate": 5.032758341066268e-06,
+      "loss": 0.3197,
+      "step": 2630
+    },
+    {
+      "epoch": 0.93,
+      "learning_rate": 4.9814874680146566e-06,
+      "loss": 0.5223,
+      "step": 2631
+    },
+    {
+      "epoch": 0.93,
+      "learning_rate": 4.930475799455115e-06,
+      "loss": 0.4854,
+      "step": 2632
+    },
+    {
+      "epoch": 0.93,
+      "learning_rate": 4.8797234031886515e-06,
+      "loss": 0.3959,
+      "step": 2633
+    },
+    {
+      "epoch": 0.93,
+      "learning_rate": 4.829230346671665e-06,
+      "loss": 0.4698,
+      "step": 2634
+    },
+    {
+      "epoch": 0.93,
+      "learning_rate": 4.778996697015936e-06,
+      "loss": 0.3715,
+      "step": 2635
+    },
+    {
+      "epoch": 0.93,
+      "learning_rate": 4.729022520988324e-06,
+      "loss": 0.4384,
+      "step": 2636
+    },
+    {
+      "epoch": 0.93,
+      "learning_rate": 4.679307885010875e-06,
+      "loss": 0.3862,
+      "step": 2637
+    },
+    {
+      "epoch": 0.93,
+      "learning_rate": 4.629852855160777e-06,
+      "loss": 0.343,
+      "step": 2638
+    },
+    {
+      "epoch": 0.93,
+      "learning_rate": 4.580657497169982e-06,
+      "loss": 0.4257,
+      "step": 2639
+    },
+    {
+      "epoch": 0.93,
+      "learning_rate": 4.531721876425477e-06,
+      "loss": 0.2823,
+      "step": 2640
+    },
+    {
+      "epoch": 0.93,
+      "eval_loss": 0.5296382308006287,
+      "eval_runtime": 52.9422,
+      "eval_samples_per_second": 30.769,
+      "eval_steps_per_second": 30.769,
+      "step": 2640
+    },
+    {
+      "epoch": 0.93,
+      "learning_rate": 4.483046057968876e-06,
+      "loss": 0.2671,
+      "step": 2641
+    },
+    {
+      "epoch": 0.94,
+      "learning_rate": 4.4346301064966065e-06,
+      "loss": 0.3811,
+      "step": 2642
+    },
+    {
+      "epoch": 0.94,
+      "learning_rate": 4.386474086359615e-06,
+      "loss": 0.5297,
+      "step": 2643
+    },
+    {
+      "epoch": 0.94,
+      "learning_rate": 4.338578061563414e-06,
+      "loss": 0.7131,
+      "step": 2644
+    },
+    {
+      "epoch": 0.94,
+      "learning_rate": 4.290942095767947e-06,
+      "loss": 0.4564,
+      "step": 2645
+    },
+    {
+      "epoch": 0.94,
+      "learning_rate": 4.243566252287456e-06,
+      "loss": 0.5494,
+      "step": 2646
+    },
+    {
+      "epoch": 0.94,
+      "learning_rate": 4.196450594090506e-06,
+      "loss": 0.2698,
+      "step": 2647
+    },
+    {
+      "epoch": 0.94,
+      "learning_rate": 4.149595183799804e-06,
+      "loss": 0.7027,
+      "step": 2648
+    },
+    {
+      "epoch": 0.94,
+      "learning_rate": 4.103000083692199e-06,
+      "loss": 0.6326,
+      "step": 2649
+    },
+    {
+      "epoch": 0.94,
+      "learning_rate": 4.056665355698508e-06,
+      "loss": 0.3522,
+      "step": 2650
+    },
+    {
+      "epoch": 0.94,
+      "learning_rate": 4.01059106140349e-06,
+      "loss": 0.4714,
+      "step": 2651
+    },
+    {
+      "epoch": 0.94,
+      "learning_rate": 3.964777262045783e-06,
+      "loss": 0.7096,
+      "step": 2652
+    },
+    {
+      "epoch": 0.94,
+      "learning_rate": 3.919224018517742e-06,
+      "loss": 0.3801,
+      "step": 2653
+    },
+    {
+      "epoch": 0.94,
+      "learning_rate": 3.8739313913654705e-06,
+      "loss": 0.808,
+      "step": 2654
+    },
+    {
+      "epoch": 0.94,
+      "learning_rate": 3.828899440788658e-06,
+      "loss": 0.5405,
+      "step": 2655
+    },
+    {
+      "epoch": 0.94,
+      "learning_rate": 3.784128226640471e-06,
+      "loss": 0.3137,
+      "step": 2656
+    },
+    {
+      "epoch": 0.94,
+      "learning_rate": 3.739617808427598e-06,
+      "loss": 0.5865,
+      "step": 2657
+    },
+    {
+      "epoch": 0.94,
+      "learning_rate": 3.695368245310049e-06,
+      "loss": 0.6342,
+      "step": 2658
+    },
+    {
+      "epoch": 0.94,
+      "learning_rate": 3.651379596101179e-06,
+      "loss": 0.2108,
+      "step": 2659
+    },
+    {
+      "epoch": 0.94,
+      "learning_rate": 3.607651919267463e-06,
+      "loss": 0.4293,
+      "step": 2660
+    },
+    {
+      "epoch": 0.94,
+      "eval_loss": 0.5293708443641663,
+      "eval_runtime": 52.9453,
+      "eval_samples_per_second": 30.768,
+      "eval_steps_per_second": 30.768,
+      "step": 2660
+    },
+    {
+      "epoch": 0.94,
+      "learning_rate": 3.5641852729286107e-06,
+      "loss": 0.4289,
+      "step": 2661
+    },
+    {
+      "epoch": 0.94,
+      "learning_rate": 3.5209797148572974e-06,
+      "loss": 0.3348,
+      "step": 2662
+    },
+    {
+      "epoch": 0.94,
+      "learning_rate": 3.4780353024792766e-06,
+      "loss": 0.4778,
+      "step": 2663
+    },
+    {
+      "epoch": 0.94,
+      "learning_rate": 3.43535209287309e-06,
+      "loss": 0.2667,
+      "step": 2664
+    },
+    {
+      "epoch": 0.94,
+      "learning_rate": 3.392930142770245e-06,
+      "loss": 0.5274,
+      "step": 2665
+    },
+    {
+      "epoch": 0.94,
+      "learning_rate": 3.350769508554863e-06,
+      "loss": 0.8883,
+      "step": 2666
+    },
+    {
+      "epoch": 0.94,
+      "learning_rate": 3.308870246263829e-06,
+      "loss": 0.4515,
+      "step": 2667
+    },
+    {
+      "epoch": 0.94,
+      "learning_rate": 3.2672324115866184e-06,
+      "loss": 0.5645,
+      "step": 2668
+    },
+    {
+      "epoch": 0.94,
+      "learning_rate": 3.2258560598652288e-06,
+      "loss": 0.3581,
+      "step": 2669
+    },
+    {
+      "epoch": 0.94,
+      "learning_rate": 3.184741246094092e-06,
+      "loss": 0.3314,
+      "step": 2670
+    },
+    {
+      "epoch": 0.95,
+      "learning_rate": 3.143888024920072e-06,
+      "loss": 0.2684,
+      "step": 2671
+    },
+    {
+      "epoch": 0.95,
+      "learning_rate": 3.1032964506422903e-06,
+      "loss": 0.5782,
+      "step": 2672
+    },
+    {
+      "epoch": 0.95,
+      "learning_rate": 3.0629665772121453e-06,
+      "loss": 0.2067,
+      "step": 2673
+    },
+    {
+      "epoch": 0.95,
+      "learning_rate": 3.022898458233159e-06,
+      "loss": 0.8022,
+      "step": 2674
+    },
+    {
+      "epoch": 0.95,
+      "learning_rate": 2.9830921469610194e-06,
+      "loss": 0.3658,
+      "step": 2675
+    },
+    {
+      "epoch": 0.95,
+      "learning_rate": 2.9435476963033615e-06,
+      "loss": 0.3407,
+      "step": 2676
+    },
+    {
+      "epoch": 0.95,
+      "learning_rate": 2.9042651588197855e-06,
+      "loss": 0.2457,
+      "step": 2677
+    },
+    {
+      "epoch": 0.95,
+      "learning_rate": 2.865244586721838e-06,
+      "loss": 0.4167,
+      "step": 2678
+    },
+    {
+      "epoch": 0.95,
+      "learning_rate": 2.826486031872855e-06,
+      "loss": 0.2217,
+      "step": 2679
+    },
+    {
+      "epoch": 0.95,
+      "learning_rate": 2.7879895457878502e-06,
+      "loss": 0.4646,
+      "step": 2680
+    },
+    {
+      "epoch": 0.95,
+      "eval_loss": 0.5296375751495361,
+      "eval_runtime": 52.9457,
+      "eval_samples_per_second": 30.767,
+      "eval_steps_per_second": 30.767,
+      "step": 2680
+    },
+    {
+      "epoch": 0.95,
+      "learning_rate": 2.749755179633584e-06,
+      "loss": 0.3472,
+      "step": 2681
+    },
+    {
+      "epoch": 0.95,
+      "learning_rate": 2.711782984228428e-06,
+      "loss": 0.3629,
+      "step": 2682
+    },
+    {
+      "epoch": 0.95,
+      "learning_rate": 2.674073010042277e-06,
+      "loss": 0.4841,
+      "step": 2683
+    },
+    {
+      "epoch": 0.95,
+      "learning_rate": 2.6366253071964607e-06,
+      "loss": 0.409,
+      "step": 2684
+    },
+    {
+      "epoch": 0.95,
+      "learning_rate": 2.599439925463809e-06,
+      "loss": 0.23,
+      "step": 2685
+    },
+    {
+      "epoch": 0.95,
+      "learning_rate": 2.562516914268387e-06,
+      "loss": 0.5579,
+      "step": 2686
+    },
+    {
+      "epoch": 0.95,
+      "learning_rate": 2.5258563226856045e-06,
+      "loss": 0.5293,
+      "step": 2687
+    },
+    {
+      "epoch": 0.95,
+      "learning_rate": 2.4894581994420853e-06,
+      "loss": 0.415,
+      "step": 2688
+    },
+    {
+      "epoch": 0.95,
+      "learning_rate": 2.4533225929155965e-06,
+      "loss": 0.3163,
+      "step": 2689
+    },
+    {
+      "epoch": 0.95,
+      "learning_rate": 2.4174495511349203e-06,
+      "loss": 0.2383,
+      "step": 2690
+    },
+    {
+      "epoch": 0.95,
+      "learning_rate": 2.3818391217799383e-06,
+      "loss": 0.3771,
+      "step": 2691
+    },
+    {
+      "epoch": 0.95,
+      "learning_rate": 2.3464913521814347e-06,
+      "loss": 0.8086,
+      "step": 2692
+    },
+    {
+      "epoch": 0.95,
+      "learning_rate": 2.3114062893211384e-06,
+      "loss": 0.3455,
+      "step": 2693
+    },
+    {
+      "epoch": 0.95,
+      "learning_rate": 2.2765839798315703e-06,
+      "loss": 0.421,
+      "step": 2694
+    },
+    {
+      "epoch": 0.95,
+      "learning_rate": 2.242024469995996e-06,
+      "loss": 0.4701,
+      "step": 2695
+    },
+    {
+      "epoch": 0.95,
+      "learning_rate": 2.2077278057484273e-06,
+      "loss": 0.5162,
+      "step": 2696
+    },
+    {
+      "epoch": 0.95,
+      "learning_rate": 2.17369403267349e-06,
+      "loss": 0.3275,
+      "step": 2697
+    },
+    {
+      "epoch": 0.95,
+      "learning_rate": 2.1399231960064657e-06,
+      "loss": 0.4906,
+      "step": 2698
+    },
+    {
+      "epoch": 0.96,
+      "learning_rate": 2.106415340633072e-06,
+      "loss": 0.5806,
+      "step": 2699
+    },
+    {
+      "epoch": 0.96,
+      "learning_rate": 2.073170511089528e-06,
+      "loss": 0.8203,
+      "step": 2700
+    },
+    {
+      "epoch": 0.96,
+      "eval_loss": 0.529484212398529,
+      "eval_runtime": 52.9185,
+      "eval_samples_per_second": 30.783,
+      "eval_steps_per_second": 30.783,
+      "step": 2700
+    },
+    {
+      "epoch": 0.96,
+      "learning_rate": 2.040188751562444e-06,
+      "loss": 0.5264,
+      "step": 2701
+    },
+    {
+      "epoch": 0.96,
+      "learning_rate": 2.0074701058887977e-06,
+      "loss": 0.5487,
+      "step": 2702
+    },
+    {
+      "epoch": 0.96,
+      "learning_rate": 1.975014617555848e-06,
+      "loss": 0.4639,
+      "step": 2703
+    },
+    {
+      "epoch": 0.96,
+      "learning_rate": 1.942822329701088e-06,
+      "loss": 0.2501,
+      "step": 2704
+    },
+    {
+      "epoch": 0.96,
+      "learning_rate": 1.9108932851121585e-06,
+      "loss": 0.8735,
+      "step": 2705
+    },
+    {
+      "epoch": 0.96,
+      "learning_rate": 1.8792275262268232e-06,
+      "loss": 0.6326,
+      "step": 2706
+    },
+    {
+      "epoch": 0.96,
+      "learning_rate": 1.847825095132949e-06,
+      "loss": 0.4528,
+      "step": 2707
+    },
+    {
+      "epoch": 0.96,
+      "learning_rate": 1.8166860335683711e-06,
+      "loss": 0.6406,
+      "step": 2708
+    },
+    {
+      "epoch": 0.96,
+      "learning_rate": 1.7858103829208717e-06,
+      "loss": 0.5748,
+      "step": 2709
+    },
+    {
+      "epoch": 0.96,
+      "learning_rate": 1.7551981842281129e-06,
+      "loss": 0.3404,
+      "step": 2710
+    },
+    {
+      "epoch": 0.96,
+      "learning_rate": 1.7248494781776369e-06,
+      "loss": 0.5519,
+      "step": 2711
+    },
+    {
+      "epoch": 0.96,
+      "learning_rate": 1.694764305106733e-06,
+      "loss": 0.6573,
+      "step": 2712
+    },
+    {
+      "epoch": 0.96,
+      "learning_rate": 1.6649427050024369e-06,
+      "loss": 0.4334,
+      "step": 2713
+    },
+    {
+      "epoch": 0.96,
+      "learning_rate": 1.635384717501487e-06,
+      "loss": 0.4423,
+      "step": 2714
+    },
+    {
+      "epoch": 0.96,
+      "learning_rate": 1.6060903818901684e-06,
+      "loss": 0.4427,
+      "step": 2715
+    },
+    {
+      "epoch": 0.96,
+      "learning_rate": 1.577059737104447e-06,
+      "loss": 0.768,
+      "step": 2716
+    },
+    {
+      "epoch": 0.96,
+      "learning_rate": 1.5482928217297021e-06,
+      "loss": 0.6759,
+      "step": 2717
+    },
+    {
+      "epoch": 0.96,
+      "learning_rate": 1.519789674000882e-06,
+      "loss": 0.1994,
+      "step": 2718
+    },
+    {
+      "epoch": 0.96,
+      "learning_rate": 1.4915503318022827e-06,
+      "loss": 0.2351,
+      "step": 2719
+    },
+    {
+      "epoch": 0.96,
+      "learning_rate": 1.4635748326676136e-06,
+      "loss": 0.351,
+      "step": 2720
+    },
+    {
+      "epoch": 0.96,
+      "eval_loss": 0.52988201379776,
+      "eval_runtime": 52.9811,
+      "eval_samples_per_second": 30.747,
+      "eval_steps_per_second": 30.747,
+      "step": 2720
+    },
+    {
+      "epoch": 0.96,
+      "learning_rate": 1.4358632137798422e-06,
+      "loss": 0.4545,
+      "step": 2721
+    },
+    {
+      "epoch": 0.96,
+      "learning_rate": 1.408415511971306e-06,
+      "loss": 0.4132,
+      "step": 2722
+    },
+    {
+      "epoch": 0.96,
+      "learning_rate": 1.381231763723445e-06,
+      "loss": 0.5652,
+      "step": 2723
+    },
+    {
+      "epoch": 0.96,
+      "learning_rate": 1.35431200516698e-06,
+      "loss": 0.5608,
+      "step": 2724
+    },
+    {
+      "epoch": 0.96,
+      "learning_rate": 1.3276562720816677e-06,
+      "loss": 0.4555,
+      "step": 2725
+    },
+    {
+      "epoch": 0.96,
+      "learning_rate": 1.3012645998963901e-06,
+      "loss": 0.3761,
+      "step": 2726
+    },
+    {
+      "epoch": 0.97,
+      "learning_rate": 1.2751370236890658e-06,
+      "loss": 0.3411,
+      "step": 2727
+    },
+    {
+      "epoch": 0.97,
+      "learning_rate": 1.2492735781865606e-06,
+      "loss": 0.4472,
+      "step": 2728
+    },
+    {
+      "epoch": 0.97,
+      "learning_rate": 1.2236742977646877e-06,
+      "loss": 0.611,
+      "step": 2729
+    },
+    {
+      "epoch": 0.97,
+      "learning_rate": 1.1983392164481632e-06,
+      "loss": 0.4682,
+      "step": 2730
+    },
+    {
+      "epoch": 0.97,
+      "learning_rate": 1.1732683679105627e-06,
+      "loss": 0.3299,
+      "step": 2731
+    },
+    {
+      "epoch": 0.97,
+      "learning_rate": 1.1484617854742307e-06,
+      "loss": 0.4545,
+      "step": 2732
+    },
+    {
+      "epoch": 0.97,
+      "learning_rate": 1.1239195021102822e-06,
+      "loss": 0.5206,
+      "step": 2733
+    },
+    {
+      "epoch": 0.97,
+      "learning_rate": 1.099641550438535e-06,
+      "loss": 0.4135,
+      "step": 2734
+    },
+    {
+      "epoch": 0.97,
+      "learning_rate": 1.075627962727488e-06,
+      "loss": 0.6073,
+      "step": 2735
+    },
+    {
+      "epoch": 0.97,
+      "learning_rate": 1.0518787708942767e-06,
+      "loss": 0.3962,
+      "step": 2736
+    },
+    {
+      "epoch": 0.97,
+      "learning_rate": 1.028394006504585e-06,
+      "loss": 0.5752,
+      "step": 2737
+    },
+    {
+      "epoch": 0.97,
+      "learning_rate": 1.005173700772688e-06,
+      "loss": 0.3396,
+      "step": 2738
+    },
+    {
+      "epoch": 0.97,
+      "learning_rate": 9.822178845613207e-07,
+      "loss": 0.4057,
+      "step": 2739
+    },
+    {
+      "epoch": 0.97,
+      "learning_rate": 9.595265883816984e-07,
+      "loss": 0.328,
+      "step": 2740
+    },
+    {
+      "epoch": 0.97,
+      "eval_loss": 0.529229998588562,
+      "eval_runtime": 52.9382,
+      "eval_samples_per_second": 30.772,
+      "eval_steps_per_second": 30.772,
+      "step": 2740
+    },
+    {
+      "epoch": 0.97,
+      "learning_rate": 9.370998423934296e-07,
+      "loss": 0.4121,
+      "step": 2741
+    },
+    {
+      "epoch": 0.97,
+      "learning_rate": 9.149376764045591e-07,
+      "loss": 0.7769,
+      "step": 2742
+    },
+    {
+      "epoch": 0.97,
+      "learning_rate": 8.930401198713911e-07,
+      "loss": 0.3529,
+      "step": 2743
+    },
+    {
+      "epoch": 0.97,
+      "learning_rate": 8.714072018986219e-07,
+      "loss": 0.479,
+      "step": 2744
+    },
+    {
+      "epoch": 0.97,
+      "learning_rate": 8.500389512391182e-07,
+      "loss": 0.4133,
+      "step": 2745
+    },
+    {
+      "epoch": 0.97,
+      "learning_rate": 8.289353962940283e-07,
+      "loss": 0.3492,
+      "step": 2746
+    },
+    {
+      "epoch": 0.97,
+      "learning_rate": 8.080965651126704e-07,
+      "loss": 0.3388,
+      "step": 2747
+    },
+    {
+      "epoch": 0.97,
+      "learning_rate": 7.875224853925112e-07,
+      "loss": 0.6677,
+      "step": 2748
+    },
+    {
+      "epoch": 0.97,
+      "learning_rate": 7.67213184479143e-07,
+      "loss": 0.5792,
+      "step": 2749
+    },
+    {
+      "epoch": 0.97,
+      "learning_rate": 7.471686893661733e-07,
+      "loss": 0.8294,
+      "step": 2750
+    },
+    {
+      "epoch": 0.97,
+      "learning_rate": 7.273890266953354e-07,
+      "loss": 0.8618,
+      "step": 2751
+    },
+    {
+      "epoch": 0.97,
+      "learning_rate": 7.078742227563107e-07,
+      "loss": 0.614,
+      "step": 2752
+    },
+    {
+      "epoch": 0.97,
+      "learning_rate": 6.886243034867734e-07,
+      "loss": 0.5996,
+      "step": 2753
+    },
+    {
+      "epoch": 0.97,
+      "learning_rate": 6.69639294472324e-07,
+      "loss": 0.1425,
+      "step": 2754
+    },
+    {
+      "epoch": 0.98,
+      "learning_rate": 6.509192209464443e-07,
+      "loss": 0.3284,
+      "step": 2755
+    },
+    {
+      "epoch": 0.98,
+      "learning_rate": 6.324641077905202e-07,
+      "loss": 0.5701,
+      "step": 2756
+    },
+    {
+      "epoch": 0.98,
+      "learning_rate": 6.142739795337304e-07,
+      "loss": 0.7111,
+      "step": 2757
+    },
+    {
+      "epoch": 0.98,
+      "learning_rate": 5.963488603530909e-07,
+      "loss": 0.4338,
+      "step": 2758
+    },
+    {
+      "epoch": 0.98,
+      "learning_rate": 5.786887740733882e-07,
+      "loss": 0.3912,
+      "step": 2759
+    },
+    {
+      "epoch": 0.98,
+      "learning_rate": 5.612937441670907e-07,
+      "loss": 0.3347,
+      "step": 2760
+    },
+    {
+      "epoch": 0.98,
+      "eval_loss": 0.5301564931869507,
+      "eval_runtime": 52.9504,
+      "eval_samples_per_second": 30.765,
+      "eval_steps_per_second": 30.765,
+      "step": 2760
+    },
+    {
+      "epoch": 0.98,
+      "learning_rate": 5.441637937544597e-07,
+      "loss": 0.3179,
+      "step": 2761
+    },
+    {
+      "epoch": 0.98,
+      "learning_rate": 5.272989456033495e-07,
+      "loss": 0.3203,
+      "step": 2762
+    },
+    {
+      "epoch": 0.98,
+      "learning_rate": 5.106992221293183e-07,
+      "loss": 0.8962,
+      "step": 2763
+    },
+    {
+      "epoch": 0.98,
+      "learning_rate": 4.943646453954953e-07,
+      "loss": 0.799,
+      "step": 2764
+    },
+    {
+      "epoch": 0.98,
+      "learning_rate": 4.78295237112647e-07,
+      "loss": 0.5524,
+      "step": 2765
+    },
+    {
+      "epoch": 0.98,
+      "learning_rate": 4.624910186390441e-07,
+      "loss": 0.7879,
+      "step": 2766
+    },
+    {
+      "epoch": 0.98,
+      "learning_rate": 4.4695201098050586e-07,
+      "loss": 0.4025,
+      "step": 2767
+    },
+    {
+      "epoch": 0.98,
+      "learning_rate": 4.3167823479037805e-07,
+      "loss": 0.5326,
+      "step": 2768
+    },
+    {
+      "epoch": 0.98,
+      "learning_rate": 4.1666971036946613e-07,
+      "loss": 0.473,
+      "step": 2769
+    },
+    {
+      "epoch": 0.98,
+      "learning_rate": 4.019264576659687e-07,
+      "loss": 0.508,
+      "step": 2770
+    },
+    {
+      "epoch": 0.98,
+      "learning_rate": 3.8744849627558867e-07,
+      "loss": 0.5211,
+      "step": 2771
+    },
+    {
+      "epoch": 0.98,
+      "learning_rate": 3.732358454413776e-07,
+      "loss": 0.5891,
+      "step": 2772
+    },
+    {
+      "epoch": 0.98,
+      "learning_rate": 3.5928852405375803e-07,
+      "loss": 0.4128,
+      "step": 2773
+    },
+    {
+      "epoch": 0.98,
+      "learning_rate": 3.456065506505013e-07,
+      "loss": 0.7014,
+      "step": 2774
+    },
+    {
+      "epoch": 0.98,
+      "learning_rate": 3.321899434166831e-07,
+      "loss": 0.338,
+      "step": 2775
+    },
+    {
+      "epoch": 0.98,
+      "learning_rate": 3.190387201847056e-07,
+      "loss": 0.4293,
+      "step": 2776
+    },
+    {
+      "epoch": 0.98,
+      "learning_rate": 3.061528984342088e-07,
+      "loss": 0.6534,
+      "step": 2777
+    },
+    {
+      "epoch": 0.98,
+      "learning_rate": 2.935324952921148e-07,
+      "loss": 0.3362,
+      "step": 2778
+    },
+    {
+      "epoch": 0.98,
+      "learning_rate": 2.81177527532539e-07,
+      "loss": 0.3799,
+      "step": 2779
+    },
+    {
+      "epoch": 0.98,
+      "learning_rate": 2.690880115768124e-07,
+      "loss": 0.2608,
+      "step": 2780
+    },
+    {
+      "epoch": 0.98,
+      "eval_loss": 0.5295908451080322,
+      "eval_runtime": 52.9478,
+      "eval_samples_per_second": 30.766,
+      "eval_steps_per_second": 30.766,
+      "step": 2780
+    },
+    {
+      "epoch": 0.98,
+      "learning_rate": 2.572639634934371e-07,
+      "loss": 0.4232,
+      "step": 2781
+    },
+    {
+      "epoch": 0.98,
+      "learning_rate": 2.45705398998064e-07,
+      "loss": 0.7618,
+      "step": 2782
+    },
+    {
+      "epoch": 0.98,
+      "learning_rate": 2.3441233345353753e-07,
+      "loss": 1.0387,
+      "step": 2783
+    },
+    {
+      "epoch": 0.99,
+      "learning_rate": 2.233847818697843e-07,
+      "loss": 0.2859,
+      "step": 2784
+    },
+    {
+      "epoch": 0.99,
+      "learning_rate": 2.1262275890379102e-07,
+      "loss": 0.3277,
+      "step": 2785
+    },
+    {
+      "epoch": 0.99,
+      "learning_rate": 2.0212627885969339e-07,
+      "loss": 0.3719,
+      "step": 2786
+    },
+    {
+      "epoch": 0.99,
+      "learning_rate": 1.918953556886205e-07,
+      "loss": 0.2608,
+      "step": 2787
+    },
+    {
+      "epoch": 0.99,
+      "learning_rate": 1.8193000298880604e-07,
+      "loss": 0.3396,
+      "step": 2788
+    },
+    {
+      "epoch": 0.99,
+      "learning_rate": 1.722302340054549e-07,
+      "loss": 0.3869,
+      "step": 2789
+    },
+    {
+      "epoch": 0.99,
+      "learning_rate": 1.6279606163080995e-07,
+      "loss": 0.4162,
+      "step": 2790
+    },
+    {
+      "epoch": 0.99,
+      "learning_rate": 1.5362749840410752e-07,
+      "loss": 0.5021,
+      "step": 2791
+    },
+    {
+      "epoch": 0.99,
+      "learning_rate": 1.4472455651148852e-07,
+      "loss": 0.3643,
+      "step": 2792
+    },
+    {
+      "epoch": 0.99,
+      "learning_rate": 1.360872477861319e-07,
+      "loss": 0.512,
+      "step": 2793
+    },
+    {
+      "epoch": 0.99,
+      "learning_rate": 1.277155837081434e-07,
+      "loss": 0.4777,
+      "step": 2794
+    },
+    {
+      "epoch": 0.99,
+      "learning_rate": 1.1960957540448903e-07,
+      "loss": 0.5408,
+      "step": 2795
+    },
+    {
+      "epoch": 0.99,
+      "learning_rate": 1.1176923364910608e-07,
+      "loss": 0.7524,
+      "step": 2796
+    },
+    {
+      "epoch": 0.99,
+      "learning_rate": 1.0419456886281432e-07,
+      "loss": 0.3626,
+      "step": 2797
+    },
+    {
+      "epoch": 0.99,
+      "learning_rate": 9.688559111331597e-08,
+      "loss": 0.4167,
+      "step": 2798
+    },
+    {
+      "epoch": 0.99,
+      "learning_rate": 8.984231011515131e-08,
+      "loss": 0.6496,
+      "step": 2799
+    },
+    {
+      "epoch": 0.99,
+      "learning_rate": 8.306473522976532e-08,
+      "loss": 0.4274,
+      "step": 2800
+    },
+    {
+      "epoch": 0.99,
+      "eval_loss": 0.5293519496917725,
+      "eval_runtime": 52.953,
+      "eval_samples_per_second": 30.763,
+      "eval_steps_per_second": 30.763,
+      "step": 2800
+    },
+    {
+      "epoch": 0.99,
+      "learning_rate": 7.655287546541878e-08,
+      "loss": 0.4402,
+      "step": 2801
+    },
+    {
+      "epoch": 0.99,
+      "learning_rate": 7.030673947716615e-08,
+      "loss": 0.345,
+      "step": 2802
+    },
+    {
+      "epoch": 0.99,
+      "learning_rate": 6.432633556696655e-08,
+      "loss": 0.6122,
+      "step": 2803
+    },
+    {
+      "epoch": 0.99,
+      "learning_rate": 5.8611671683528325e-08,
+      "loss": 1.0764,
+      "step": 2804
+    },
+    {
+      "epoch": 0.99,
+      "learning_rate": 5.316275542235349e-08,
+      "loss": 0.3302,
+      "step": 2805
+    },
+    {
+      "epoch": 0.99,
+      "learning_rate": 4.797959402575991e-08,
+      "loss": 0.3881,
+      "step": 2806
+    },
+    {
+      "epoch": 0.99,
+      "learning_rate": 4.306219438283687e-08,
+      "loss": 0.3493,
+      "step": 2807
+    },
+    {
+      "epoch": 0.99,
+      "learning_rate": 3.8410563029400714e-08,
+      "loss": 0.2884,
+      "step": 2808
+    },
+    {
+      "epoch": 0.99,
+      "learning_rate": 3.402470614810582e-08,
+      "loss": 0.4118,
+      "step": 2809
+    },
+    {
+      "epoch": 0.99,
+      "learning_rate": 2.9904629568289214e-08,
+      "loss": 0.3168,
+      "step": 2810
+    },
+    {
+      "epoch": 0.99,
+      "learning_rate": 2.605033876608154e-08,
+      "loss": 0.4338,
+      "step": 2811
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 2.2461838864296092e-08,
+      "loss": 0.3457,
+      "step": 2812
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 1.91391346325398e-08,
+      "loss": 0.6276,
+      "step": 2813
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 1.6082230487080018e-08,
+      "loss": 0.4727,
+      "step": 2814
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 1.329113049095554e-08,
+      "loss": 0.5634,
+      "step": 2815
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 1.0765838353887781e-08,
+      "loss": 0.3701,
+      "step": 2816
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 8.506357432325196e-09,
+      "loss": 0.3308,
+      "step": 2817
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 6.5126907293766585e-09,
+      "loss": 0.6281,
+      "step": 2818
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 4.784840894900278e-09,
+      "loss": 0.3423,
+      "step": 2819
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 3.3228102254145855e-09,
+      "loss": 0.3349,
+      "step": 2820
+    },
+    {
+      "epoch": 1.0,
+      "eval_loss": 0.5301524996757507,
+      "eval_runtime": 52.9437,
+      "eval_samples_per_second": 30.769,
+      "eval_steps_per_second": 30.769,
+      "step": 2820
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 2.126600664165146e-09,
+      "loss": 0.4528,
+      "step": 2821
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 1.1962138010357392e-09,
+      "loss": 0.437,
+      "step": 2822
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 5.316508726815883e-10,
+      "loss": 0.2902,
+      "step": 2823
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 1.3291276232951788e-10,
+      "loss": 0.44,
+      "step": 2824
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 0.0,
+      "loss": 0.4575,
+      "step": 2825
+    }
+  ],
+  "logging_steps": 1,
+  "max_steps": 2825,
+  "num_train_epochs": 1,
+  "save_steps": 500,
+  "total_flos": 1.79325058940928e+18,
+  "trial_name": null,
+  "trial_params": null
+}