diff --git "a/trainer_state.json" "b/trainer_state.json"
new file mode 100644--- /dev/null
+++ "b/trainer_state.json"
@@ -0,0 +1,13663 @@
+{
+  "best_metric": 0.8184141516685486,
+  "best_model_checkpoint": "/workspace/models/phi_wizard_vicuna_uncensored/checkpoint-2250",
+  "epoch": 0.25320729237002026,
+  "eval_steps": 125,
+  "global_step": 2250,
+  "is_hyper_param_search": false,
+  "is_local_process_zero": true,
+  "is_world_process_zero": true,
+  "log_history": [
+    {
+      "epoch": 0.0,
+      "learning_rate": 0.0,
+      "loss": 1.1144,
+      "step": 1
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 1.5051499783199057e-05,
+      "loss": 0.9181,
+      "step": 2
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 2.385606273598312e-05,
+      "loss": 0.832,
+      "step": 3
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 3.0102999566398115e-05,
+      "loss": 1.0356,
+      "step": 4
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 3.4948500216800935e-05,
+      "loss": 0.7333,
+      "step": 5
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 3.890756251918218e-05,
+      "loss": 0.8684,
+      "step": 6
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 4.2254902000712836e-05,
+      "loss": 0.94,
+      "step": 7
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 4.515449934959717e-05,
+      "loss": 0.8714,
+      "step": 8
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 4.771212547196624e-05,
+      "loss": 0.7988,
+      "step": 9
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 4.9999999999999996e-05,
+      "loss": 1.0084,
+      "step": 10
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 5e-05,
+      "loss": 0.8936,
+      "step": 11
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 4.999812368658061e-05,
+      "loss": 0.8718,
+      "step": 12
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 4.9996247373161215e-05,
+      "loss": 0.8417,
+      "step": 13
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 4.999437105974182e-05,
+      "loss": 0.7193,
+      "step": 14
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 4.999249474632243e-05,
+      "loss": 0.7377,
+      "step": 15
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 4.999061843290303e-05,
+      "loss": 0.8965,
+      "step": 16
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 4.998874211948364e-05,
+      "loss": 0.9234,
+      "step": 17
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 4.9986865806064246e-05,
+      "loss": 1.1066,
+      "step": 18
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 4.998498949264485e-05,
+      "loss": 0.9531,
+      "step": 19
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 4.998311317922546e-05,
+      "loss": 0.7659,
+      "step": 20
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 4.998123686580607e-05,
+      "loss": 0.8075,
+      "step": 21
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 4.997936055238667e-05,
+      "loss": 0.7079,
+      "step": 22
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 4.997748423896728e-05,
+      "loss": 0.8311,
+      "step": 23
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 4.997560792554788e-05,
+      "loss": 0.8106,
+      "step": 24
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 4.9973731612128495e-05,
+      "loss": 0.6618,
+      "step": 25
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 4.9971855298709095e-05,
+      "loss": 0.8473,
+      "step": 26
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 4.996997898528971e-05,
+      "loss": 0.8045,
+      "step": 27
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 4.996810267187031e-05,
+      "loss": 0.7867,
+      "step": 28
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 4.996622635845092e-05,
+      "loss": 0.8984,
+      "step": 29
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 4.9964350045031526e-05,
+      "loss": 0.774,
+      "step": 30
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 4.996247373161213e-05,
+      "loss": 0.8261,
+      "step": 31
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 4.996059741819274e-05,
+      "loss": 0.9101,
+      "step": 32
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 4.9958721104773345e-05,
+      "loss": 0.6528,
+      "step": 33
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 4.995684479135395e-05,
+      "loss": 1.0518,
+      "step": 34
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 4.995496847793456e-05,
+      "loss": 0.8767,
+      "step": 35
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 4.995309216451516e-05,
+      "loss": 0.8162,
+      "step": 36
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 4.995121585109577e-05,
+      "loss": 1.1871,
+      "step": 37
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 4.9949339537676376e-05,
+      "loss": 0.6064,
+      "step": 38
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 4.994746322425698e-05,
+      "loss": 0.8823,
+      "step": 39
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 4.994558691083759e-05,
+      "loss": 0.6771,
+      "step": 40
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 4.9943710597418194e-05,
+      "loss": 0.9261,
+      "step": 41
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 4.99418342839988e-05,
+      "loss": 1.0417,
+      "step": 42
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 4.9939957970579406e-05,
+      "loss": 0.9086,
+      "step": 43
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 4.993808165716001e-05,
+      "loss": 0.8122,
+      "step": 44
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.993620534374062e-05,
+      "loss": 0.8589,
+      "step": 45
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.9934329030321225e-05,
+      "loss": 0.7518,
+      "step": 46
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.993245271690183e-05,
+      "loss": 0.7702,
+      "step": 47
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.993057640348244e-05,
+      "loss": 0.8724,
+      "step": 48
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.992870009006304e-05,
+      "loss": 1.1665,
+      "step": 49
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.9926823776643656e-05,
+      "loss": 0.6424,
+      "step": 50
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.9924947463224256e-05,
+      "loss": 0.7967,
+      "step": 51
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.992307114980487e-05,
+      "loss": 0.5744,
+      "step": 52
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.992119483638547e-05,
+      "loss": 0.85,
+      "step": 53
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.991931852296608e-05,
+      "loss": 1.0129,
+      "step": 54
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.991744220954668e-05,
+      "loss": 0.9215,
+      "step": 55
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.991556589612729e-05,
+      "loss": 0.8173,
+      "step": 56
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.991368958270789e-05,
+      "loss": 0.8801,
+      "step": 57
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.9911813269288506e-05,
+      "loss": 0.9592,
+      "step": 58
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.990993695586911e-05,
+      "loss": 0.8643,
+      "step": 59
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.990806064244972e-05,
+      "loss": 0.7594,
+      "step": 60
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.9906184329030324e-05,
+      "loss": 0.7222,
+      "step": 61
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.990430801561093e-05,
+      "loss": 0.9104,
+      "step": 62
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.9902431702191536e-05,
+      "loss": 0.9196,
+      "step": 63
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.990055538877214e-05,
+      "loss": 0.6478,
+      "step": 64
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.989867907535275e-05,
+      "loss": 0.9611,
+      "step": 65
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.9896802761933355e-05,
+      "loss": 0.758,
+      "step": 66
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.989492644851396e-05,
+      "loss": 0.944,
+      "step": 67
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.9893050135094574e-05,
+      "loss": 0.7635,
+      "step": 68
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.989117382167517e-05,
+      "loss": 0.8438,
+      "step": 69
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.9889297508255786e-05,
+      "loss": 0.8421,
+      "step": 70
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.9887421194836386e-05,
+      "loss": 0.9179,
+      "step": 71
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.9885544881417e-05,
+      "loss": 0.9359,
+      "step": 72
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.98836685679976e-05,
+      "loss": 0.5753,
+      "step": 73
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.988179225457821e-05,
+      "loss": 0.8972,
+      "step": 74
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.987991594115881e-05,
+      "loss": 0.6883,
+      "step": 75
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.987803962773942e-05,
+      "loss": 0.8092,
+      "step": 76
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.987616331432003e-05,
+      "loss": 0.6714,
+      "step": 77
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.9874287000900636e-05,
+      "loss": 0.8575,
+      "step": 78
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.987241068748124e-05,
+      "loss": 1.022,
+      "step": 79
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.987053437406185e-05,
+      "loss": 0.8809,
+      "step": 80
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.9868658060642454e-05,
+      "loss": 0.7326,
+      "step": 81
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.986678174722306e-05,
+      "loss": 1.0286,
+      "step": 82
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.9864905433803666e-05,
+      "loss": 0.636,
+      "step": 83
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.986302912038427e-05,
+      "loss": 1.0537,
+      "step": 84
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.986115280696488e-05,
+      "loss": 0.7521,
+      "step": 85
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.9859276493545485e-05,
+      "loss": 0.7486,
+      "step": 86
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.985740018012609e-05,
+      "loss": 0.8204,
+      "step": 87
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.98555238667067e-05,
+      "loss": 0.8751,
+      "step": 88
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.98536475532873e-05,
+      "loss": 0.8645,
+      "step": 89
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.985177123986791e-05,
+      "loss": 0.6587,
+      "step": 90
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.9849894926448516e-05,
+      "loss": 0.7749,
+      "step": 91
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.984801861302912e-05,
+      "loss": 0.6874,
+      "step": 92
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.984614229960973e-05,
+      "loss": 0.9132,
+      "step": 93
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.9844265986190334e-05,
+      "loss": 0.7744,
+      "step": 94
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.984238967277094e-05,
+      "loss": 0.8725,
+      "step": 95
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.9840513359351546e-05,
+      "loss": 0.9358,
+      "step": 96
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.983863704593216e-05,
+      "loss": 0.936,
+      "step": 97
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.983676073251276e-05,
+      "loss": 0.7336,
+      "step": 98
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.983488441909337e-05,
+      "loss": 0.8092,
+      "step": 99
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.983300810567397e-05,
+      "loss": 0.6633,
+      "step": 100
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.9831131792254584e-05,
+      "loss": 1.0522,
+      "step": 101
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.9829255478835183e-05,
+      "loss": 0.7187,
+      "step": 102
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.9827379165415796e-05,
+      "loss": 0.8115,
+      "step": 103
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.9825502851996396e-05,
+      "loss": 0.6716,
+      "step": 104
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.982362653857701e-05,
+      "loss": 0.7005,
+      "step": 105
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.9821750225157615e-05,
+      "loss": 0.8373,
+      "step": 106
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.981987391173822e-05,
+      "loss": 0.9641,
+      "step": 107
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.981799759831883e-05,
+      "loss": 0.8482,
+      "step": 108
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.981612128489943e-05,
+      "loss": 0.6718,
+      "step": 109
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.981424497148004e-05,
+      "loss": 0.9597,
+      "step": 110
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.9812368658060646e-05,
+      "loss": 0.8335,
+      "step": 111
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.981049234464125e-05,
+      "loss": 0.8604,
+      "step": 112
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.980861603122186e-05,
+      "loss": 0.6548,
+      "step": 113
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.9806739717802464e-05,
+      "loss": 0.7998,
+      "step": 114
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.980486340438307e-05,
+      "loss": 0.9691,
+      "step": 115
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.9802987090963676e-05,
+      "loss": 0.8362,
+      "step": 116
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.980111077754428e-05,
+      "loss": 0.8146,
+      "step": 117
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.979923446412489e-05,
+      "loss": 0.9636,
+      "step": 118
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.9797358150705495e-05,
+      "loss": 0.9878,
+      "step": 119
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.97954818372861e-05,
+      "loss": 0.8681,
+      "step": 120
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.979360552386671e-05,
+      "loss": 0.7043,
+      "step": 121
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.9791729210447313e-05,
+      "loss": 0.8683,
+      "step": 122
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.978985289702792e-05,
+      "loss": 0.8893,
+      "step": 123
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.978797658360853e-05,
+      "loss": 0.9536,
+      "step": 124
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.978610027018913e-05,
+      "loss": 1.0656,
+      "step": 125
+    },
+    {
+      "epoch": 0.01,
+      "eval_loss": 0.8471789956092834,
+      "eval_runtime": 23.2996,
+      "eval_samples_per_second": 43.906,
+      "eval_steps_per_second": 21.975,
+      "step": 125
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.9784223956769745e-05,
+      "loss": 1.0107,
+      "step": 126
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.9782347643350344e-05,
+      "loss": 0.6956,
+      "step": 127
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.978047132993096e-05,
+      "loss": 0.6716,
+      "step": 128
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.9778595016511557e-05,
+      "loss": 0.7368,
+      "step": 129
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.977671870309217e-05,
+      "loss": 0.8081,
+      "step": 130
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.977484238967277e-05,
+      "loss": 0.9898,
+      "step": 131
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.977296607625338e-05,
+      "loss": 0.4479,
+      "step": 132
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.977108976283399e-05,
+      "loss": 0.8237,
+      "step": 133
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.9769213449414594e-05,
+      "loss": 0.6529,
+      "step": 134
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.97673371359952e-05,
+      "loss": 1.1196,
+      "step": 135
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.9765460822575806e-05,
+      "loss": 0.6718,
+      "step": 136
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.976358450915641e-05,
+      "loss": 1.0806,
+      "step": 137
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.976170819573702e-05,
+      "loss": 0.9096,
+      "step": 138
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.9759831882317625e-05,
+      "loss": 0.7177,
+      "step": 139
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.975795556889823e-05,
+      "loss": 0.5833,
+      "step": 140
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.975607925547884e-05,
+      "loss": 0.8781,
+      "step": 141
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.9754202942059443e-05,
+      "loss": 0.7248,
+      "step": 142
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.975232662864005e-05,
+      "loss": 1.0227,
+      "step": 143
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.975045031522066e-05,
+      "loss": 0.9977,
+      "step": 144
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.974857400180126e-05,
+      "loss": 0.9053,
+      "step": 145
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.974669768838187e-05,
+      "loss": 0.8915,
+      "step": 146
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.9744821374962474e-05,
+      "loss": 0.8665,
+      "step": 147
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.974294506154308e-05,
+      "loss": 0.6849,
+      "step": 148
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.9741068748123687e-05,
+      "loss": 0.6791,
+      "step": 149
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.973919243470429e-05,
+      "loss": 0.7313,
+      "step": 150
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.97373161212849e-05,
+      "loss": 0.5531,
+      "step": 151
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.9735439807865505e-05,
+      "loss": 0.6714,
+      "step": 152
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.973356349444612e-05,
+      "loss": 0.7428,
+      "step": 153
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.973168718102672e-05,
+      "loss": 0.6771,
+      "step": 154
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.972981086760733e-05,
+      "loss": 0.8889,
+      "step": 155
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.972793455418793e-05,
+      "loss": 0.7279,
+      "step": 156
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.972605824076854e-05,
+      "loss": 0.6068,
+      "step": 157
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.972418192734914e-05,
+      "loss": 1.1141,
+      "step": 158
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.9722305613929755e-05,
+      "loss": 0.7562,
+      "step": 159
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.9720429300510354e-05,
+      "loss": 0.6447,
+      "step": 160
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.971855298709097e-05,
+      "loss": 0.8645,
+      "step": 161
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.9716676673671573e-05,
+      "loss": 0.6226,
+      "step": 162
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.971480036025218e-05,
+      "loss": 0.5578,
+      "step": 163
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.9712924046832786e-05,
+      "loss": 0.7299,
+      "step": 164
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.971104773341339e-05,
+      "loss": 0.7121,
+      "step": 165
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.9709171419994e-05,
+      "loss": 0.6609,
+      "step": 166
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.9707295106574604e-05,
+      "loss": 0.8262,
+      "step": 167
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.970541879315521e-05,
+      "loss": 0.8155,
+      "step": 168
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.9703542479735817e-05,
+      "loss": 0.8445,
+      "step": 169
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.970166616631642e-05,
+      "loss": 0.6519,
+      "step": 170
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.9699789852897036e-05,
+      "loss": 0.9663,
+      "step": 171
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.9697913539477635e-05,
+      "loss": 0.7763,
+      "step": 172
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.969603722605825e-05,
+      "loss": 0.7084,
+      "step": 173
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.969416091263885e-05,
+      "loss": 0.812,
+      "step": 174
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.969228459921946e-05,
+      "loss": 0.9422,
+      "step": 175
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.969040828580006e-05,
+      "loss": 0.9289,
+      "step": 176
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.968853197238067e-05,
+      "loss": 0.8564,
+      "step": 177
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.968665565896127e-05,
+      "loss": 0.7434,
+      "step": 178
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.9684779345541885e-05,
+      "loss": 0.9149,
+      "step": 179
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.968290303212249e-05,
+      "loss": 0.8226,
+      "step": 180
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.96810267187031e-05,
+      "loss": 0.8694,
+      "step": 181
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.9679150405283703e-05,
+      "loss": 0.7168,
+      "step": 182
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.967727409186431e-05,
+      "loss": 0.7846,
+      "step": 183
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.9675397778444916e-05,
+      "loss": 0.7555,
+      "step": 184
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.967352146502552e-05,
+      "loss": 0.7564,
+      "step": 185
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.967164515160613e-05,
+      "loss": 0.8801,
+      "step": 186
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.9669768838186734e-05,
+      "loss": 0.8891,
+      "step": 187
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.966789252476734e-05,
+      "loss": 0.8547,
+      "step": 188
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.9666016211347947e-05,
+      "loss": 0.9151,
+      "step": 189
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.966413989792855e-05,
+      "loss": 0.7768,
+      "step": 190
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.966226358450916e-05,
+      "loss": 0.9352,
+      "step": 191
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.9660387271089765e-05,
+      "loss": 0.9051,
+      "step": 192
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.965851095767037e-05,
+      "loss": 0.819,
+      "step": 193
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.965663464425098e-05,
+      "loss": 0.9249,
+      "step": 194
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.9654758330831583e-05,
+      "loss": 0.7689,
+      "step": 195
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.965288201741219e-05,
+      "loss": 0.8923,
+      "step": 196
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.9651005703992796e-05,
+      "loss": 0.7129,
+      "step": 197
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.96491293905734e-05,
+      "loss": 0.635,
+      "step": 198
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.964725307715401e-05,
+      "loss": 1.0711,
+      "step": 199
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.964537676373462e-05,
+      "loss": 0.7779,
+      "step": 200
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.964350045031522e-05,
+      "loss": 0.8622,
+      "step": 201
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.9641624136895833e-05,
+      "loss": 0.9166,
+      "step": 202
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.963974782347643e-05,
+      "loss": 0.7636,
+      "step": 203
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.9637871510057046e-05,
+      "loss": 0.5954,
+      "step": 204
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.9635995196637645e-05,
+      "loss": 0.8256,
+      "step": 205
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.963411888321826e-05,
+      "loss": 0.9206,
+      "step": 206
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.963224256979886e-05,
+      "loss": 0.8101,
+      "step": 207
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.963036625637947e-05,
+      "loss": 0.7983,
+      "step": 208
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.9628489942960077e-05,
+      "loss": 0.993,
+      "step": 209
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.962661362954068e-05,
+      "loss": 1.1817,
+      "step": 210
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.962473731612129e-05,
+      "loss": 0.8136,
+      "step": 211
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.9622861002701895e-05,
+      "loss": 0.814,
+      "step": 212
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.96209846892825e-05,
+      "loss": 0.8072,
+      "step": 213
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.961910837586311e-05,
+      "loss": 1.1131,
+      "step": 214
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.9617232062443713e-05,
+      "loss": 0.701,
+      "step": 215
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.961535574902432e-05,
+      "loss": 0.6915,
+      "step": 216
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.9613479435604926e-05,
+      "loss": 0.7472,
+      "step": 217
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.961160312218553e-05,
+      "loss": 1.0076,
+      "step": 218
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.960972680876614e-05,
+      "loss": 0.7439,
+      "step": 219
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.9607850495346744e-05,
+      "loss": 0.8943,
+      "step": 220
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.960597418192735e-05,
+      "loss": 1.0693,
+      "step": 221
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.960409786850796e-05,
+      "loss": 0.7249,
+      "step": 222
+    },
+    {
+      "epoch": 0.03,
+      "learning_rate": 4.960222155508856e-05,
+      "loss": 0.7381,
+      "step": 223
+    },
+    {
+      "epoch": 0.03,
+      "learning_rate": 4.960034524166917e-05,
+      "loss": 0.7282,
+      "step": 224
+    },
+    {
+      "epoch": 0.03,
+      "learning_rate": 4.9598468928249775e-05,
+      "loss": 0.8062,
+      "step": 225
+    },
+    {
+      "epoch": 0.03,
+      "learning_rate": 4.959659261483038e-05,
+      "loss": 0.7619,
+      "step": 226
+    },
+    {
+      "epoch": 0.03,
+      "learning_rate": 4.9594716301410994e-05,
+      "loss": 0.7548,
+      "step": 227
+    },
+    {
+      "epoch": 0.03,
+      "learning_rate": 4.9592839987991594e-05,
+      "loss": 0.892,
+      "step": 228
+    },
+    {
+      "epoch": 0.03,
+      "learning_rate": 4.9590963674572207e-05,
+      "loss": 1.0792,
+      "step": 229
+    },
+    {
+      "epoch": 0.03,
+      "learning_rate": 4.9589087361152806e-05,
+      "loss": 0.9112,
+      "step": 230
+    },
+    {
+      "epoch": 0.03,
+      "learning_rate": 4.958721104773342e-05,
+      "loss": 0.6992,
+      "step": 231
+    },
+    {
+      "epoch": 0.03,
+      "learning_rate": 4.958533473431402e-05,
+      "loss": 0.6015,
+      "step": 232
+    },
+    {
+      "epoch": 0.03,
+      "learning_rate": 4.958345842089463e-05,
+      "loss": 0.54,
+      "step": 233
+    },
+    {
+      "epoch": 0.03,
+      "learning_rate": 4.958158210747523e-05,
+      "loss": 1.0341,
+      "step": 234
+    },
+    {
+      "epoch": 0.03,
+      "learning_rate": 4.9579705794055843e-05,
+      "loss": 0.6531,
+      "step": 235
+    },
+    {
+      "epoch": 0.03,
+      "learning_rate": 4.957782948063644e-05,
+      "loss": 0.7168,
+      "step": 236
+    },
+    {
+      "epoch": 0.03,
+      "learning_rate": 4.9575953167217056e-05,
+      "loss": 0.7474,
+      "step": 237
+    },
+    {
+      "epoch": 0.03,
+      "learning_rate": 4.957407685379766e-05,
+      "loss": 0.8498,
+      "step": 238
+    },
+    {
+      "epoch": 0.03,
+      "learning_rate": 4.957220054037827e-05,
+      "loss": 1.0063,
+      "step": 239
+    },
+    {
+      "epoch": 0.03,
+      "learning_rate": 4.9570324226958874e-05,
+      "loss": 1.0043,
+      "step": 240
+    },
+    {
+      "epoch": 0.03,
+      "learning_rate": 4.956844791353948e-05,
+      "loss": 0.7871,
+      "step": 241
+    },
+    {
+      "epoch": 0.03,
+      "learning_rate": 4.956657160012009e-05,
+      "loss": 1.2475,
+      "step": 242
+    },
+    {
+      "epoch": 0.03,
+      "learning_rate": 4.956469528670069e-05,
+      "loss": 0.9004,
+      "step": 243
+    },
+    {
+      "epoch": 0.03,
+      "learning_rate": 4.95628189732813e-05,
+      "loss": 0.594,
+      "step": 244
+    },
+    {
+      "epoch": 0.03,
+      "learning_rate": 4.9560942659861905e-05,
+      "loss": 0.6828,
+      "step": 245
+    },
+    {
+      "epoch": 0.03,
+      "learning_rate": 4.955906634644251e-05,
+      "loss": 0.8419,
+      "step": 246
+    },
+    {
+      "epoch": 0.03,
+      "learning_rate": 4.9557190033023124e-05,
+      "loss": 1.1047,
+      "step": 247
+    },
+    {
+      "epoch": 0.03,
+      "learning_rate": 4.9555313719603724e-05,
+      "loss": 0.8193,
+      "step": 248
+    },
+    {
+      "epoch": 0.03,
+      "learning_rate": 4.9553437406184337e-05,
+      "loss": 0.8485,
+      "step": 249
+    },
+    {
+      "epoch": 0.03,
+      "learning_rate": 4.9551561092764936e-05,
+      "loss": 0.9435,
+      "step": 250
+    },
+    {
+      "epoch": 0.03,
+      "eval_loss": 0.8420971035957336,
+      "eval_runtime": 23.3138,
+      "eval_samples_per_second": 43.88,
+      "eval_steps_per_second": 21.961,
+      "step": 250
+    },
+    {
+      "epoch": 0.03,
+      "learning_rate": 4.954968477934554e-05,
+      "loss": 0.8044,
+      "step": 251
+    },
+    {
+      "epoch": 0.03,
+      "learning_rate": 4.954780846592615e-05,
+      "loss": 0.8077,
+      "step": 252
+    },
+    {
+      "epoch": 0.03,
+      "learning_rate": 4.9545932152506754e-05,
+      "loss": 0.8519,
+      "step": 253
+    },
+    {
+      "epoch": 0.03,
+      "learning_rate": 4.954405583908736e-05,
+      "loss": 0.8508,
+      "step": 254
+    },
+    {
+      "epoch": 0.03,
+      "learning_rate": 4.954217952566797e-05,
+      "loss": 0.9062,
+      "step": 255
+    },
+    {
+      "epoch": 0.03,
+      "learning_rate": 4.954030321224858e-05,
+      "loss": 1.1664,
+      "step": 256
+    },
+    {
+      "epoch": 0.03,
+      "learning_rate": 4.953842689882918e-05,
+      "loss": 0.9158,
+      "step": 257
+    },
+    {
+      "epoch": 0.03,
+      "learning_rate": 4.953655058540979e-05,
+      "loss": 1.0224,
+      "step": 258
+    },
+    {
+      "epoch": 0.03,
+      "learning_rate": 4.953467427199039e-05,
+      "loss": 0.8326,
+      "step": 259
+    },
+    {
+      "epoch": 0.03,
+      "learning_rate": 4.9532797958571004e-05,
+      "loss": 0.8291,
+      "step": 260
+    },
+    {
+      "epoch": 0.03,
+      "learning_rate": 4.9530921645151604e-05,
+      "loss": 0.9005,
+      "step": 261
+    },
+    {
+      "epoch": 0.03,
+      "learning_rate": 4.952904533173222e-05,
+      "loss": 0.9437,
+      "step": 262
+    },
+    {
+      "epoch": 0.03,
+      "learning_rate": 4.9527169018312816e-05,
+      "loss": 0.777,
+      "step": 263
+    },
+    {
+      "epoch": 0.03,
+      "learning_rate": 4.952529270489343e-05,
+      "loss": 0.6552,
+      "step": 264
+    },
+    {
+      "epoch": 0.03,
+      "learning_rate": 4.9523416391474035e-05,
+      "loss": 0.9966,
+      "step": 265
+    },
+    {
+      "epoch": 0.03,
+      "learning_rate": 4.952154007805464e-05,
+      "loss": 0.3182,
+      "step": 266
+    },
+    {
+      "epoch": 0.03,
+      "learning_rate": 4.951966376463525e-05,
+      "loss": 0.878,
+      "step": 267
+    },
+    {
+      "epoch": 0.03,
+      "learning_rate": 4.9517787451215854e-05,
+      "loss": 1.0225,
+      "step": 268
+    },
+    {
+      "epoch": 0.03,
+      "learning_rate": 4.951591113779646e-05,
+      "loss": 0.6964,
+      "step": 269
+    },
+    {
+      "epoch": 0.03,
+      "learning_rate": 4.9514034824377066e-05,
+      "loss": 0.7952,
+      "step": 270
+    },
+    {
+      "epoch": 0.03,
+      "learning_rate": 4.951215851095767e-05,
+      "loss": 0.8069,
+      "step": 271
+    },
+    {
+      "epoch": 0.03,
+      "learning_rate": 4.951028219753828e-05,
+      "loss": 0.8743,
+      "step": 272
+    },
+    {
+      "epoch": 0.03,
+      "learning_rate": 4.9508405884118884e-05,
+      "loss": 0.7554,
+      "step": 273
+    },
+    {
+      "epoch": 0.03,
+      "learning_rate": 4.95065295706995e-05,
+      "loss": 0.7386,
+      "step": 274
+    },
+    {
+      "epoch": 0.03,
+      "learning_rate": 4.95046532572801e-05,
+      "loss": 0.6898,
+      "step": 275
+    },
+    {
+      "epoch": 0.03,
+      "learning_rate": 4.950277694386071e-05,
+      "loss": 1.1274,
+      "step": 276
+    },
+    {
+      "epoch": 0.03,
+      "learning_rate": 4.950090063044131e-05,
+      "loss": 0.8201,
+      "step": 277
+    },
+    {
+      "epoch": 0.03,
+      "learning_rate": 4.949902431702192e-05,
+      "loss": 0.8338,
+      "step": 278
+    },
+    {
+      "epoch": 0.03,
+      "learning_rate": 4.949714800360252e-05,
+      "loss": 0.8832,
+      "step": 279
+    },
+    {
+      "epoch": 0.03,
+      "learning_rate": 4.9495271690183134e-05,
+      "loss": 0.7047,
+      "step": 280
+    },
+    {
+      "epoch": 0.03,
+      "learning_rate": 4.9493395376763734e-05,
+      "loss": 0.8521,
+      "step": 281
+    },
+    {
+      "epoch": 0.03,
+      "learning_rate": 4.949151906334435e-05,
+      "loss": 1.0266,
+      "step": 282
+    },
+    {
+      "epoch": 0.03,
+      "learning_rate": 4.9489642749924946e-05,
+      "loss": 0.7552,
+      "step": 283
+    },
+    {
+      "epoch": 0.03,
+      "learning_rate": 4.948776643650556e-05,
+      "loss": 0.9556,
+      "step": 284
+    },
+    {
+      "epoch": 0.03,
+      "learning_rate": 4.9485890123086165e-05,
+      "loss": 0.4726,
+      "step": 285
+    },
+    {
+      "epoch": 0.03,
+      "learning_rate": 4.948401380966677e-05,
+      "loss": 1.3049,
+      "step": 286
+    },
+    {
+      "epoch": 0.03,
+      "learning_rate": 4.948213749624738e-05,
+      "loss": 0.801,
+      "step": 287
+    },
+    {
+      "epoch": 0.03,
+      "learning_rate": 4.9480261182827984e-05,
+      "loss": 0.9412,
+      "step": 288
+    },
+    {
+      "epoch": 0.03,
+      "learning_rate": 4.947838486940859e-05,
+      "loss": 0.8314,
+      "step": 289
+    },
+    {
+      "epoch": 0.03,
+      "learning_rate": 4.9476508555989196e-05,
+      "loss": 0.7477,
+      "step": 290
+    },
+    {
+      "epoch": 0.03,
+      "learning_rate": 4.94746322425698e-05,
+      "loss": 0.7984,
+      "step": 291
+    },
+    {
+      "epoch": 0.03,
+      "learning_rate": 4.947275592915041e-05,
+      "loss": 0.925,
+      "step": 292
+    },
+    {
+      "epoch": 0.03,
+      "learning_rate": 4.9470879615731014e-05,
+      "loss": 0.7771,
+      "step": 293
+    },
+    {
+      "epoch": 0.03,
+      "learning_rate": 4.946900330231162e-05,
+      "loss": 0.9383,
+      "step": 294
+    },
+    {
+      "epoch": 0.03,
+      "learning_rate": 4.946712698889223e-05,
+      "loss": 0.9632,
+      "step": 295
+    },
+    {
+      "epoch": 0.03,
+      "learning_rate": 4.946525067547283e-05,
+      "loss": 0.8613,
+      "step": 296
+    },
+    {
+      "epoch": 0.03,
+      "learning_rate": 4.946337436205344e-05,
+      "loss": 0.9518,
+      "step": 297
+    },
+    {
+      "epoch": 0.03,
+      "learning_rate": 4.9461498048634045e-05,
+      "loss": 0.9645,
+      "step": 298
+    },
+    {
+      "epoch": 0.03,
+      "learning_rate": 4.945962173521465e-05,
+      "loss": 0.8497,
+      "step": 299
+    },
+    {
+      "epoch": 0.03,
+      "learning_rate": 4.945774542179526e-05,
+      "loss": 0.8013,
+      "step": 300
+    },
+    {
+      "epoch": 0.03,
+      "learning_rate": 4.9455869108375864e-05,
+      "loss": 0.8923,
+      "step": 301
+    },
+    {
+      "epoch": 0.03,
+      "learning_rate": 4.945399279495647e-05,
+      "loss": 0.8431,
+      "step": 302
+    },
+    {
+      "epoch": 0.03,
+      "learning_rate": 4.945211648153708e-05,
+      "loss": 0.8376,
+      "step": 303
+    },
+    {
+      "epoch": 0.03,
+      "learning_rate": 4.945024016811768e-05,
+      "loss": 0.8235,
+      "step": 304
+    },
+    {
+      "epoch": 0.03,
+      "learning_rate": 4.9448363854698295e-05,
+      "loss": 0.8932,
+      "step": 305
+    },
+    {
+      "epoch": 0.03,
+      "learning_rate": 4.9446487541278894e-05,
+      "loss": 0.8825,
+      "step": 306
+    },
+    {
+      "epoch": 0.03,
+      "learning_rate": 4.944461122785951e-05,
+      "loss": 1.0481,
+      "step": 307
+    },
+    {
+      "epoch": 0.03,
+      "learning_rate": 4.944273491444011e-05,
+      "loss": 0.8196,
+      "step": 308
+    },
+    {
+      "epoch": 0.03,
+      "learning_rate": 4.944085860102072e-05,
+      "loss": 1.0425,
+      "step": 309
+    },
+    {
+      "epoch": 0.03,
+      "learning_rate": 4.943898228760132e-05,
+      "loss": 0.9625,
+      "step": 310
+    },
+    {
+      "epoch": 0.03,
+      "learning_rate": 4.943710597418193e-05,
+      "loss": 0.5855,
+      "step": 311
+    },
+    {
+      "epoch": 0.04,
+      "learning_rate": 4.943522966076254e-05,
+      "loss": 0.9185,
+      "step": 312
+    },
+    {
+      "epoch": 0.04,
+      "learning_rate": 4.9433353347343144e-05,
+      "loss": 0.6893,
+      "step": 313
+    },
+    {
+      "epoch": 0.04,
+      "learning_rate": 4.943147703392375e-05,
+      "loss": 0.6367,
+      "step": 314
+    },
+    {
+      "epoch": 0.04,
+      "learning_rate": 4.942960072050436e-05,
+      "loss": 0.8172,
+      "step": 315
+    },
+    {
+      "epoch": 0.04,
+      "learning_rate": 4.942772440708496e-05,
+      "loss": 0.4704,
+      "step": 316
+    },
+    {
+      "epoch": 0.04,
+      "learning_rate": 4.942584809366557e-05,
+      "loss": 0.9546,
+      "step": 317
+    },
+    {
+      "epoch": 0.04,
+      "learning_rate": 4.9423971780246175e-05,
+      "loss": 0.9931,
+      "step": 318
+    },
+    {
+      "epoch": 0.04,
+      "learning_rate": 4.942209546682678e-05,
+      "loss": 0.9429,
+      "step": 319
+    },
+    {
+      "epoch": 0.04,
+      "learning_rate": 4.942021915340739e-05,
+      "loss": 1.0111,
+      "step": 320
+    },
+    {
+      "epoch": 0.04,
+      "learning_rate": 4.9418342839987994e-05,
+      "loss": 0.6577,
+      "step": 321
+    },
+    {
+      "epoch": 0.04,
+      "learning_rate": 4.94164665265686e-05,
+      "loss": 0.9256,
+      "step": 322
+    },
+    {
+      "epoch": 0.04,
+      "learning_rate": 4.9414590213149206e-05,
+      "loss": 0.7493,
+      "step": 323
+    },
+    {
+      "epoch": 0.04,
+      "learning_rate": 4.941271389972981e-05,
+      "loss": 0.8698,
+      "step": 324
+    },
+    {
+      "epoch": 0.04,
+      "learning_rate": 4.941083758631042e-05,
+      "loss": 0.9573,
+      "step": 325
+    },
+    {
+      "epoch": 0.04,
+      "learning_rate": 4.9408961272891024e-05,
+      "loss": 0.6053,
+      "step": 326
+    },
+    {
+      "epoch": 0.04,
+      "learning_rate": 4.940708495947163e-05,
+      "loss": 0.8846,
+      "step": 327
+    },
+    {
+      "epoch": 0.04,
+      "learning_rate": 4.940520864605224e-05,
+      "loss": 0.9002,
+      "step": 328
+    },
+    {
+      "epoch": 0.04,
+      "learning_rate": 4.940333233263284e-05,
+      "loss": 0.7962,
+      "step": 329
+    },
+    {
+      "epoch": 0.04,
+      "learning_rate": 4.940145601921345e-05,
+      "loss": 0.8842,
+      "step": 330
+    },
+    {
+      "epoch": 0.04,
+      "learning_rate": 4.9399579705794055e-05,
+      "loss": 0.8584,
+      "step": 331
+    },
+    {
+      "epoch": 0.04,
+      "learning_rate": 4.939770339237467e-05,
+      "loss": 0.8218,
+      "step": 332
+    },
+    {
+      "epoch": 0.04,
+      "learning_rate": 4.939582707895527e-05,
+      "loss": 0.9477,
+      "step": 333
+    },
+    {
+      "epoch": 0.04,
+      "learning_rate": 4.939395076553588e-05,
+      "loss": 0.7983,
+      "step": 334
+    },
+    {
+      "epoch": 0.04,
+      "learning_rate": 4.939207445211648e-05,
+      "loss": 0.8255,
+      "step": 335
+    },
+    {
+      "epoch": 0.04,
+      "learning_rate": 4.939019813869709e-05,
+      "loss": 1.0005,
+      "step": 336
+    },
+    {
+      "epoch": 0.04,
+      "learning_rate": 4.938832182527769e-05,
+      "loss": 0.7475,
+      "step": 337
+    },
+    {
+      "epoch": 0.04,
+      "learning_rate": 4.9386445511858305e-05,
+      "loss": 0.7432,
+      "step": 338
+    },
+    {
+      "epoch": 0.04,
+      "learning_rate": 4.9384569198438905e-05,
+      "loss": 0.8014,
+      "step": 339
+    },
+    {
+      "epoch": 0.04,
+      "learning_rate": 4.938269288501952e-05,
+      "loss": 0.9513,
+      "step": 340
+    },
+    {
+      "epoch": 0.04,
+      "learning_rate": 4.9380816571600124e-05,
+      "loss": 1.0256,
+      "step": 341
+    },
+    {
+      "epoch": 0.04,
+      "learning_rate": 4.937894025818073e-05,
+      "loss": 0.7841,
+      "step": 342
+    },
+    {
+      "epoch": 0.04,
+      "learning_rate": 4.9377063944761336e-05,
+      "loss": 0.6828,
+      "step": 343
+    },
+    {
+      "epoch": 0.04,
+      "learning_rate": 4.937518763134194e-05,
+      "loss": 1.0057,
+      "step": 344
+    },
+    {
+      "epoch": 0.04,
+      "learning_rate": 4.937331131792255e-05,
+      "loss": 0.8517,
+      "step": 345
+    },
+    {
+      "epoch": 0.04,
+      "learning_rate": 4.9371435004503154e-05,
+      "loss": 0.7777,
+      "step": 346
+    },
+    {
+      "epoch": 0.04,
+      "learning_rate": 4.936955869108376e-05,
+      "loss": 0.8717,
+      "step": 347
+    },
+    {
+      "epoch": 0.04,
+      "learning_rate": 4.936768237766437e-05,
+      "loss": 0.726,
+      "step": 348
+    },
+    {
+      "epoch": 0.04,
+      "learning_rate": 4.936580606424497e-05,
+      "loss": 0.8907,
+      "step": 349
+    },
+    {
+      "epoch": 0.04,
+      "learning_rate": 4.9363929750825586e-05,
+      "loss": 0.8789,
+      "step": 350
+    },
+    {
+      "epoch": 0.04,
+      "learning_rate": 4.9362053437406185e-05,
+      "loss": 0.6603,
+      "step": 351
+    },
+    {
+      "epoch": 0.04,
+      "learning_rate": 4.93601771239868e-05,
+      "loss": 0.6822,
+      "step": 352
+    },
+    {
+      "epoch": 0.04,
+      "learning_rate": 4.93583008105674e-05,
+      "loss": 0.7407,
+      "step": 353
+    },
+    {
+      "epoch": 0.04,
+      "learning_rate": 4.935642449714801e-05,
+      "loss": 1.0425,
+      "step": 354
+    },
+    {
+      "epoch": 0.04,
+      "learning_rate": 4.935454818372861e-05,
+      "loss": 0.8305,
+      "step": 355
+    },
+    {
+      "epoch": 0.04,
+      "learning_rate": 4.9352671870309216e-05,
+      "loss": 0.9695,
+      "step": 356
+    },
+    {
+      "epoch": 0.04,
+      "learning_rate": 4.935079555688982e-05,
+      "loss": 0.7811,
+      "step": 357
+    },
+    {
+      "epoch": 0.04,
+      "learning_rate": 4.934891924347043e-05,
+      "loss": 0.7599,
+      "step": 358
+    },
+    {
+      "epoch": 0.04,
+      "learning_rate": 4.934704293005104e-05,
+      "loss": 0.7533,
+      "step": 359
+    },
+    {
+      "epoch": 0.04,
+      "learning_rate": 4.934516661663164e-05,
+      "loss": 0.9709,
+      "step": 360
+    },
+    {
+      "epoch": 0.04,
+      "learning_rate": 4.9343290303212254e-05,
+      "loss": 0.9105,
+      "step": 361
+    },
+    {
+      "epoch": 0.04,
+      "learning_rate": 4.934141398979285e-05,
+      "loss": 0.9561,
+      "step": 362
+    },
+    {
+      "epoch": 0.04,
+      "learning_rate": 4.9339537676373466e-05,
+      "loss": 1.0189,
+      "step": 363
+    },
+    {
+      "epoch": 0.04,
+      "learning_rate": 4.9337661362954065e-05,
+      "loss": 0.5703,
+      "step": 364
+    },
+    {
+      "epoch": 0.04,
+      "learning_rate": 4.933578504953468e-05,
+      "loss": 0.9627,
+      "step": 365
+    },
+    {
+      "epoch": 0.04,
+      "learning_rate": 4.933390873611528e-05,
+      "loss": 0.8807,
+      "step": 366
+    },
+    {
+      "epoch": 0.04,
+      "learning_rate": 4.933203242269589e-05,
+      "loss": 0.6822,
+      "step": 367
+    },
+    {
+      "epoch": 0.04,
+      "learning_rate": 4.93301561092765e-05,
+      "loss": 1.073,
+      "step": 368
+    },
+    {
+      "epoch": 0.04,
+      "learning_rate": 4.93282797958571e-05,
+      "loss": 0.7212,
+      "step": 369
+    },
+    {
+      "epoch": 0.04,
+      "learning_rate": 4.932640348243771e-05,
+      "loss": 0.9796,
+      "step": 370
+    },
+    {
+      "epoch": 0.04,
+      "learning_rate": 4.9324527169018315e-05,
+      "loss": 0.8576,
+      "step": 371
+    },
+    {
+      "epoch": 0.04,
+      "learning_rate": 4.932265085559892e-05,
+      "loss": 0.8855,
+      "step": 372
+    },
+    {
+      "epoch": 0.04,
+      "learning_rate": 4.932077454217953e-05,
+      "loss": 0.9225,
+      "step": 373
+    },
+    {
+      "epoch": 0.04,
+      "learning_rate": 4.9318898228760134e-05,
+      "loss": 0.959,
+      "step": 374
+    },
+    {
+      "epoch": 0.04,
+      "learning_rate": 4.931702191534074e-05,
+      "loss": 0.6961,
+      "step": 375
+    },
+    {
+      "epoch": 0.04,
+      "eval_loss": 0.8414385318756104,
+      "eval_runtime": 23.4672,
+      "eval_samples_per_second": 43.593,
+      "eval_steps_per_second": 21.818,
+      "step": 375
+    },
+    {
+      "epoch": 0.04,
+      "learning_rate": 4.9315145601921346e-05,
+      "loss": 0.7116,
+      "step": 376
+    },
+    {
+      "epoch": 0.04,
+      "learning_rate": 4.931326928850196e-05,
+      "loss": 0.7943,
+      "step": 377
+    },
+    {
+      "epoch": 0.04,
+      "learning_rate": 4.931139297508256e-05,
+      "loss": 1.13,
+      "step": 378
+    },
+    {
+      "epoch": 0.04,
+      "learning_rate": 4.930951666166317e-05,
+      "loss": 0.9919,
+      "step": 379
+    },
+    {
+      "epoch": 0.04,
+      "learning_rate": 4.930764034824377e-05,
+      "loss": 0.8689,
+      "step": 380
+    },
+    {
+      "epoch": 0.04,
+      "learning_rate": 4.9305764034824384e-05,
+      "loss": 0.7226,
+      "step": 381
+    },
+    {
+      "epoch": 0.04,
+      "learning_rate": 4.930388772140498e-05,
+      "loss": 0.8599,
+      "step": 382
+    },
+    {
+      "epoch": 0.04,
+      "learning_rate": 4.9302011407985596e-05,
+      "loss": 0.7759,
+      "step": 383
+    },
+    {
+      "epoch": 0.04,
+      "learning_rate": 4.9300135094566195e-05,
+      "loss": 0.8065,
+      "step": 384
+    },
+    {
+      "epoch": 0.04,
+      "learning_rate": 4.929825878114681e-05,
+      "loss": 0.5453,
+      "step": 385
+    },
+    {
+      "epoch": 0.04,
+      "learning_rate": 4.929638246772741e-05,
+      "loss": 1.0245,
+      "step": 386
+    },
+    {
+      "epoch": 0.04,
+      "learning_rate": 4.929450615430802e-05,
+      "loss": 0.6772,
+      "step": 387
+    },
+    {
+      "epoch": 0.04,
+      "learning_rate": 4.929262984088863e-05,
+      "loss": 0.7517,
+      "step": 388
+    },
+    {
+      "epoch": 0.04,
+      "learning_rate": 4.929075352746923e-05,
+      "loss": 0.9985,
+      "step": 389
+    },
+    {
+      "epoch": 0.04,
+      "learning_rate": 4.928887721404984e-05,
+      "loss": 0.8158,
+      "step": 390
+    },
+    {
+      "epoch": 0.04,
+      "learning_rate": 4.9287000900630445e-05,
+      "loss": 0.5471,
+      "step": 391
+    },
+    {
+      "epoch": 0.04,
+      "learning_rate": 4.928512458721105e-05,
+      "loss": 1.1669,
+      "step": 392
+    },
+    {
+      "epoch": 0.04,
+      "learning_rate": 4.928324827379166e-05,
+      "loss": 0.7587,
+      "step": 393
+    },
+    {
+      "epoch": 0.04,
+      "learning_rate": 4.9281371960372264e-05,
+      "loss": 0.8903,
+      "step": 394
+    },
+    {
+      "epoch": 0.04,
+      "learning_rate": 4.927949564695287e-05,
+      "loss": 0.8407,
+      "step": 395
+    },
+    {
+      "epoch": 0.04,
+      "learning_rate": 4.9277619333533476e-05,
+      "loss": 0.7875,
+      "step": 396
+    },
+    {
+      "epoch": 0.04,
+      "learning_rate": 4.927574302011408e-05,
+      "loss": 0.8788,
+      "step": 397
+    },
+    {
+      "epoch": 0.04,
+      "learning_rate": 4.927386670669469e-05,
+      "loss": 0.8168,
+      "step": 398
+    },
+    {
+      "epoch": 0.04,
+      "learning_rate": 4.9271990393275295e-05,
+      "loss": 0.8401,
+      "step": 399
+    },
+    {
+      "epoch": 0.05,
+      "learning_rate": 4.92701140798559e-05,
+      "loss": 0.9616,
+      "step": 400
+    },
+    {
+      "epoch": 0.05,
+      "learning_rate": 4.926823776643651e-05,
+      "loss": 0.8407,
+      "step": 401
+    },
+    {
+      "epoch": 0.05,
+      "learning_rate": 4.926636145301711e-05,
+      "loss": 0.8909,
+      "step": 402
+    },
+    {
+      "epoch": 0.05,
+      "learning_rate": 4.926448513959772e-05,
+      "loss": 1.0076,
+      "step": 403
+    },
+    {
+      "epoch": 0.05,
+      "learning_rate": 4.9262608826178325e-05,
+      "loss": 0.7332,
+      "step": 404
+    },
+    {
+      "epoch": 0.05,
+      "learning_rate": 4.926073251275893e-05,
+      "loss": 0.8679,
+      "step": 405
+    },
+    {
+      "epoch": 0.05,
+      "learning_rate": 4.9258856199339544e-05,
+      "loss": 0.6794,
+      "step": 406
+    },
+    {
+      "epoch": 0.05,
+      "learning_rate": 4.9256979885920144e-05,
+      "loss": 0.8106,
+      "step": 407
+    },
+    {
+      "epoch": 0.05,
+      "learning_rate": 4.925510357250076e-05,
+      "loss": 1.0265,
+      "step": 408
+    },
+    {
+      "epoch": 0.05,
+      "learning_rate": 4.9253227259081356e-05,
+      "loss": 0.8566,
+      "step": 409
+    },
+    {
+      "epoch": 0.05,
+      "learning_rate": 4.925135094566197e-05,
+      "loss": 0.7816,
+      "step": 410
+    },
+    {
+      "epoch": 0.05,
+      "learning_rate": 4.924947463224257e-05,
+      "loss": 0.7635,
+      "step": 411
+    },
+    {
+      "epoch": 0.05,
+      "learning_rate": 4.924759831882318e-05,
+      "loss": 0.718,
+      "step": 412
+    },
+    {
+      "epoch": 0.05,
+      "learning_rate": 4.924572200540378e-05,
+      "loss": 0.8633,
+      "step": 413
+    },
+    {
+      "epoch": 0.05,
+      "learning_rate": 4.9243845691984394e-05,
+      "loss": 0.6477,
+      "step": 414
+    },
+    {
+      "epoch": 0.05,
+      "learning_rate": 4.9241969378565e-05,
+      "loss": 0.7102,
+      "step": 415
+    },
+    {
+      "epoch": 0.05,
+      "learning_rate": 4.9240093065145606e-05,
+      "loss": 0.6861,
+      "step": 416
+    },
+    {
+      "epoch": 0.05,
+      "learning_rate": 4.923821675172621e-05,
+      "loss": 1.0511,
+      "step": 417
+    },
+    {
+      "epoch": 0.05,
+      "learning_rate": 4.923634043830682e-05,
+      "loss": 0.8916,
+      "step": 418
+    },
+    {
+      "epoch": 0.05,
+      "learning_rate": 4.9234464124887425e-05,
+      "loss": 0.9898,
+      "step": 419
+    },
+    {
+      "epoch": 0.05,
+      "learning_rate": 4.923258781146803e-05,
+      "loss": 0.7804,
+      "step": 420
+    },
+    {
+      "epoch": 0.05,
+      "learning_rate": 4.923071149804864e-05,
+      "loss": 0.7741,
+      "step": 421
+    },
+    {
+      "epoch": 0.05,
+      "learning_rate": 4.922883518462924e-05,
+      "loss": 0.9148,
+      "step": 422
+    },
+    {
+      "epoch": 0.05,
+      "learning_rate": 4.922695887120985e-05,
+      "loss": 0.606,
+      "step": 423
+    },
+    {
+      "epoch": 0.05,
+      "learning_rate": 4.9225082557790455e-05,
+      "loss": 0.6733,
+      "step": 424
+    },
+    {
+      "epoch": 0.05,
+      "learning_rate": 4.922320624437106e-05,
+      "loss": 0.9069,
+      "step": 425
+    },
+    {
+      "epoch": 0.05,
+      "learning_rate": 4.922132993095167e-05,
+      "loss": 0.8743,
+      "step": 426
+    },
+    {
+      "epoch": 0.05,
+      "learning_rate": 4.9219453617532274e-05,
+      "loss": 0.8305,
+      "step": 427
+    },
+    {
+      "epoch": 0.05,
+      "learning_rate": 4.921757730411288e-05,
+      "loss": 0.8767,
+      "step": 428
+    },
+    {
+      "epoch": 0.05,
+      "learning_rate": 4.9215700990693486e-05,
+      "loss": 0.7956,
+      "step": 429
+    },
+    {
+      "epoch": 0.05,
+      "learning_rate": 4.921382467727409e-05,
+      "loss": 0.8003,
+      "step": 430
+    },
+    {
+      "epoch": 0.05,
+      "learning_rate": 4.92119483638547e-05,
+      "loss": 0.9309,
+      "step": 431
+    },
+    {
+      "epoch": 0.05,
+      "learning_rate": 4.9210072050435305e-05,
+      "loss": 0.8599,
+      "step": 432
+    },
+    {
+      "epoch": 0.05,
+      "learning_rate": 4.920819573701591e-05,
+      "loss": 0.8613,
+      "step": 433
+    },
+    {
+      "epoch": 0.05,
+      "learning_rate": 4.920631942359652e-05,
+      "loss": 0.8426,
+      "step": 434
+    },
+    {
+      "epoch": 0.05,
+      "learning_rate": 4.920444311017713e-05,
+      "loss": 0.8631,
+      "step": 435
+    },
+    {
+      "epoch": 0.05,
+      "learning_rate": 4.920256679675773e-05,
+      "loss": 0.6423,
+      "step": 436
+    },
+    {
+      "epoch": 0.05,
+      "learning_rate": 4.920069048333834e-05,
+      "loss": 0.8344,
+      "step": 437
+    },
+    {
+      "epoch": 0.05,
+      "learning_rate": 4.919881416991894e-05,
+      "loss": 1.1724,
+      "step": 438
+    },
+    {
+      "epoch": 0.05,
+      "learning_rate": 4.9196937856499555e-05,
+      "loss": 0.9498,
+      "step": 439
+    },
+    {
+      "epoch": 0.05,
+      "learning_rate": 4.9195061543080154e-05,
+      "loss": 0.9625,
+      "step": 440
+    },
+    {
+      "epoch": 0.05,
+      "learning_rate": 4.919318522966077e-05,
+      "loss": 0.9749,
+      "step": 441
+    },
+    {
+      "epoch": 0.05,
+      "learning_rate": 4.9191308916241366e-05,
+      "loss": 0.6747,
+      "step": 442
+    },
+    {
+      "epoch": 0.05,
+      "learning_rate": 4.918943260282198e-05,
+      "loss": 0.8073,
+      "step": 443
+    },
+    {
+      "epoch": 0.05,
+      "learning_rate": 4.9187556289402585e-05,
+      "loss": 1.0963,
+      "step": 444
+    },
+    {
+      "epoch": 0.05,
+      "learning_rate": 4.918567997598319e-05,
+      "loss": 0.846,
+      "step": 445
+    },
+    {
+      "epoch": 0.05,
+      "learning_rate": 4.91838036625638e-05,
+      "loss": 1.1609,
+      "step": 446
+    },
+    {
+      "epoch": 0.05,
+      "learning_rate": 4.9181927349144404e-05,
+      "loss": 0.7232,
+      "step": 447
+    },
+    {
+      "epoch": 0.05,
+      "learning_rate": 4.918005103572501e-05,
+      "loss": 0.7901,
+      "step": 448
+    },
+    {
+      "epoch": 0.05,
+      "learning_rate": 4.9178174722305616e-05,
+      "loss": 0.7564,
+      "step": 449
+    },
+    {
+      "epoch": 0.05,
+      "learning_rate": 4.917629840888622e-05,
+      "loss": 0.7094,
+      "step": 450
+    },
+    {
+      "epoch": 0.05,
+      "learning_rate": 4.917442209546683e-05,
+      "loss": 0.8122,
+      "step": 451
+    },
+    {
+      "epoch": 0.05,
+      "learning_rate": 4.9172545782047435e-05,
+      "loss": 0.9637,
+      "step": 452
+    },
+    {
+      "epoch": 0.05,
+      "learning_rate": 4.917066946862805e-05,
+      "loss": 0.8195,
+      "step": 453
+    },
+    {
+      "epoch": 0.05,
+      "learning_rate": 4.916879315520865e-05,
+      "loss": 0.9316,
+      "step": 454
+    },
+    {
+      "epoch": 0.05,
+      "learning_rate": 4.916691684178926e-05,
+      "loss": 0.9048,
+      "step": 455
+    },
+    {
+      "epoch": 0.05,
+      "learning_rate": 4.916504052836986e-05,
+      "loss": 0.8733,
+      "step": 456
+    },
+    {
+      "epoch": 0.05,
+      "learning_rate": 4.916316421495047e-05,
+      "loss": 0.7249,
+      "step": 457
+    },
+    {
+      "epoch": 0.05,
+      "learning_rate": 4.916128790153107e-05,
+      "loss": 1.0532,
+      "step": 458
+    },
+    {
+      "epoch": 0.05,
+      "learning_rate": 4.9159411588111685e-05,
+      "loss": 0.6826,
+      "step": 459
+    },
+    {
+      "epoch": 0.05,
+      "learning_rate": 4.9157535274692284e-05,
+      "loss": 0.8379,
+      "step": 460
+    },
+    {
+      "epoch": 0.05,
+      "learning_rate": 4.915565896127289e-05,
+      "loss": 0.7115,
+      "step": 461
+    },
+    {
+      "epoch": 0.05,
+      "learning_rate": 4.91537826478535e-05,
+      "loss": 0.7518,
+      "step": 462
+    },
+    {
+      "epoch": 0.05,
+      "learning_rate": 4.91519063344341e-05,
+      "loss": 1.0653,
+      "step": 463
+    },
+    {
+      "epoch": 0.05,
+      "learning_rate": 4.9150030021014715e-05,
+      "loss": 0.6959,
+      "step": 464
+    },
+    {
+      "epoch": 0.05,
+      "learning_rate": 4.9148153707595315e-05,
+      "loss": 0.7376,
+      "step": 465
+    },
+    {
+      "epoch": 0.05,
+      "learning_rate": 4.914627739417593e-05,
+      "loss": 0.8145,
+      "step": 466
+    },
+    {
+      "epoch": 0.05,
+      "learning_rate": 4.914440108075653e-05,
+      "loss": 0.9262,
+      "step": 467
+    },
+    {
+      "epoch": 0.05,
+      "learning_rate": 4.914252476733714e-05,
+      "loss": 0.7456,
+      "step": 468
+    },
+    {
+      "epoch": 0.05,
+      "learning_rate": 4.914064845391774e-05,
+      "loss": 0.9614,
+      "step": 469
+    },
+    {
+      "epoch": 0.05,
+      "learning_rate": 4.913877214049835e-05,
+      "loss": 0.8804,
+      "step": 470
+    },
+    {
+      "epoch": 0.05,
+      "learning_rate": 4.913689582707896e-05,
+      "loss": 0.8761,
+      "step": 471
+    },
+    {
+      "epoch": 0.05,
+      "learning_rate": 4.9135019513659565e-05,
+      "loss": 0.829,
+      "step": 472
+    },
+    {
+      "epoch": 0.05,
+      "learning_rate": 4.913314320024017e-05,
+      "loss": 0.7397,
+      "step": 473
+    },
+    {
+      "epoch": 0.05,
+      "learning_rate": 4.913126688682078e-05,
+      "loss": 0.7856,
+      "step": 474
+    },
+    {
+      "epoch": 0.05,
+      "learning_rate": 4.912939057340138e-05,
+      "loss": 0.8022,
+      "step": 475
+    },
+    {
+      "epoch": 0.05,
+      "learning_rate": 4.912751425998199e-05,
+      "loss": 0.9543,
+      "step": 476
+    },
+    {
+      "epoch": 0.05,
+      "learning_rate": 4.9125637946562595e-05,
+      "loss": 0.6939,
+      "step": 477
+    },
+    {
+      "epoch": 0.05,
+      "learning_rate": 4.91237616331432e-05,
+      "loss": 0.9849,
+      "step": 478
+    },
+    {
+      "epoch": 0.05,
+      "learning_rate": 4.912188531972381e-05,
+      "loss": 0.8896,
+      "step": 479
+    },
+    {
+      "epoch": 0.05,
+      "learning_rate": 4.9120009006304414e-05,
+      "loss": 1.0209,
+      "step": 480
+    },
+    {
+      "epoch": 0.05,
+      "learning_rate": 4.911813269288502e-05,
+      "loss": 1.0513,
+      "step": 481
+    },
+    {
+      "epoch": 0.05,
+      "learning_rate": 4.911625637946563e-05,
+      "loss": 1.0601,
+      "step": 482
+    },
+    {
+      "epoch": 0.05,
+      "learning_rate": 4.911438006604623e-05,
+      "loss": 0.9107,
+      "step": 483
+    },
+    {
+      "epoch": 0.05,
+      "learning_rate": 4.9112503752626845e-05,
+      "loss": 0.6771,
+      "step": 484
+    },
+    {
+      "epoch": 0.05,
+      "learning_rate": 4.9110627439207445e-05,
+      "loss": 0.7614,
+      "step": 485
+    },
+    {
+      "epoch": 0.05,
+      "learning_rate": 4.910875112578806e-05,
+      "loss": 0.69,
+      "step": 486
+    },
+    {
+      "epoch": 0.05,
+      "learning_rate": 4.910687481236866e-05,
+      "loss": 0.955,
+      "step": 487
+    },
+    {
+      "epoch": 0.05,
+      "learning_rate": 4.910499849894927e-05,
+      "loss": 0.8767,
+      "step": 488
+    },
+    {
+      "epoch": 0.06,
+      "learning_rate": 4.910312218552987e-05,
+      "loss": 0.9607,
+      "step": 489
+    },
+    {
+      "epoch": 0.06,
+      "learning_rate": 4.910124587211048e-05,
+      "loss": 0.8458,
+      "step": 490
+    },
+    {
+      "epoch": 0.06,
+      "learning_rate": 4.909936955869109e-05,
+      "loss": 0.834,
+      "step": 491
+    },
+    {
+      "epoch": 0.06,
+      "learning_rate": 4.9097493245271695e-05,
+      "loss": 0.9523,
+      "step": 492
+    },
+    {
+      "epoch": 0.06,
+      "learning_rate": 4.90956169318523e-05,
+      "loss": 0.8541,
+      "step": 493
+    },
+    {
+      "epoch": 0.06,
+      "learning_rate": 4.909374061843291e-05,
+      "loss": 0.5629,
+      "step": 494
+    },
+    {
+      "epoch": 0.06,
+      "learning_rate": 4.909186430501351e-05,
+      "loss": 0.7361,
+      "step": 495
+    },
+    {
+      "epoch": 0.06,
+      "learning_rate": 4.908998799159412e-05,
+      "loss": 1.1667,
+      "step": 496
+    },
+    {
+      "epoch": 0.06,
+      "learning_rate": 4.9088111678174725e-05,
+      "loss": 0.7682,
+      "step": 497
+    },
+    {
+      "epoch": 0.06,
+      "learning_rate": 4.908623536475533e-05,
+      "loss": 0.9249,
+      "step": 498
+    },
+    {
+      "epoch": 0.06,
+      "learning_rate": 4.908435905133594e-05,
+      "loss": 0.861,
+      "step": 499
+    },
+    {
+      "epoch": 0.06,
+      "learning_rate": 4.9082482737916544e-05,
+      "loss": 0.6147,
+      "step": 500
+    },
+    {
+      "epoch": 0.06,
+      "eval_loss": 0.8420451283454895,
+      "eval_runtime": 23.296,
+      "eval_samples_per_second": 43.913,
+      "eval_steps_per_second": 21.978,
+      "step": 500
+    },
+    {
+      "epoch": 0.06,
+      "learning_rate": 4.908060642449715e-05,
+      "loss": 0.6636,
+      "step": 501
+    },
+    {
+      "epoch": 0.06,
+      "learning_rate": 4.9078730111077756e-05,
+      "loss": 0.5953,
+      "step": 502
+    },
+    {
+      "epoch": 0.06,
+      "learning_rate": 4.907685379765836e-05,
+      "loss": 0.8413,
+      "step": 503
+    },
+    {
+      "epoch": 0.06,
+      "learning_rate": 4.907497748423897e-05,
+      "loss": 1.015,
+      "step": 504
+    },
+    {
+      "epoch": 0.06,
+      "learning_rate": 4.9073101170819575e-05,
+      "loss": 1.0021,
+      "step": 505
+    },
+    {
+      "epoch": 0.06,
+      "learning_rate": 4.907122485740018e-05,
+      "loss": 1.0199,
+      "step": 506
+    },
+    {
+      "epoch": 0.06,
+      "learning_rate": 4.906934854398079e-05,
+      "loss": 0.779,
+      "step": 507
+    },
+    {
+      "epoch": 0.06,
+      "learning_rate": 4.906747223056139e-05,
+      "loss": 0.9439,
+      "step": 508
+    },
+    {
+      "epoch": 0.06,
+      "learning_rate": 4.9065595917142006e-05,
+      "loss": 0.8013,
+      "step": 509
+    },
+    {
+      "epoch": 0.06,
+      "learning_rate": 4.9063719603722606e-05,
+      "loss": 0.7901,
+      "step": 510
+    },
+    {
+      "epoch": 0.06,
+      "learning_rate": 4.906184329030322e-05,
+      "loss": 0.8875,
+      "step": 511
+    },
+    {
+      "epoch": 0.06,
+      "learning_rate": 4.905996697688382e-05,
+      "loss": 0.8983,
+      "step": 512
+    },
+    {
+      "epoch": 0.06,
+      "learning_rate": 4.905809066346443e-05,
+      "loss": 0.7042,
+      "step": 513
+    },
+    {
+      "epoch": 0.06,
+      "learning_rate": 4.905621435004503e-05,
+      "loss": 0.8269,
+      "step": 514
+    },
+    {
+      "epoch": 0.06,
+      "learning_rate": 4.905433803662564e-05,
+      "loss": 0.9909,
+      "step": 515
+    },
+    {
+      "epoch": 0.06,
+      "learning_rate": 4.905246172320624e-05,
+      "loss": 0.7224,
+      "step": 516
+    },
+    {
+      "epoch": 0.06,
+      "learning_rate": 4.9050585409786855e-05,
+      "loss": 0.8239,
+      "step": 517
+    },
+    {
+      "epoch": 0.06,
+      "learning_rate": 4.904870909636746e-05,
+      "loss": 0.6648,
+      "step": 518
+    },
+    {
+      "epoch": 0.06,
+      "learning_rate": 4.904683278294807e-05,
+      "loss": 0.9614,
+      "step": 519
+    },
+    {
+      "epoch": 0.06,
+      "learning_rate": 4.9044956469528674e-05,
+      "loss": 0.8268,
+      "step": 520
+    },
+    {
+      "epoch": 0.06,
+      "learning_rate": 4.904308015610928e-05,
+      "loss": 0.7072,
+      "step": 521
+    },
+    {
+      "epoch": 0.06,
+      "learning_rate": 4.9041203842689886e-05,
+      "loss": 0.6029,
+      "step": 522
+    },
+    {
+      "epoch": 0.06,
+      "learning_rate": 4.903932752927049e-05,
+      "loss": 0.7992,
+      "step": 523
+    },
+    {
+      "epoch": 0.06,
+      "learning_rate": 4.90374512158511e-05,
+      "loss": 0.7442,
+      "step": 524
+    },
+    {
+      "epoch": 0.06,
+      "learning_rate": 4.9035574902431705e-05,
+      "loss": 0.7612,
+      "step": 525
+    },
+    {
+      "epoch": 0.06,
+      "learning_rate": 4.903369858901231e-05,
+      "loss": 0.9306,
+      "step": 526
+    },
+    {
+      "epoch": 0.06,
+      "learning_rate": 4.903182227559292e-05,
+      "loss": 0.8378,
+      "step": 527
+    },
+    {
+      "epoch": 0.06,
+      "learning_rate": 4.902994596217352e-05,
+      "loss": 0.8455,
+      "step": 528
+    },
+    {
+      "epoch": 0.06,
+      "learning_rate": 4.902806964875413e-05,
+      "loss": 0.7875,
+      "step": 529
+    },
+    {
+      "epoch": 0.06,
+      "learning_rate": 4.9026193335334736e-05,
+      "loss": 0.9288,
+      "step": 530
+    },
+    {
+      "epoch": 0.06,
+      "learning_rate": 4.902431702191534e-05,
+      "loss": 0.7924,
+      "step": 531
+    },
+    {
+      "epoch": 0.06,
+      "learning_rate": 4.902244070849595e-05,
+      "loss": 0.8319,
+      "step": 532
+    },
+    {
+      "epoch": 0.06,
+      "learning_rate": 4.9020564395076554e-05,
+      "loss": 0.8822,
+      "step": 533
+    },
+    {
+      "epoch": 0.06,
+      "learning_rate": 4.901868808165716e-05,
+      "loss": 0.6843,
+      "step": 534
+    },
+    {
+      "epoch": 0.06,
+      "learning_rate": 4.9016811768237766e-05,
+      "loss": 0.9904,
+      "step": 535
+    },
+    {
+      "epoch": 0.06,
+      "learning_rate": 4.901493545481837e-05,
+      "loss": 0.6959,
+      "step": 536
+    },
+    {
+      "epoch": 0.06,
+      "learning_rate": 4.901305914139898e-05,
+      "loss": 0.8204,
+      "step": 537
+    },
+    {
+      "epoch": 0.06,
+      "learning_rate": 4.901118282797959e-05,
+      "loss": 0.8604,
+      "step": 538
+    },
+    {
+      "epoch": 0.06,
+      "learning_rate": 4.900930651456019e-05,
+      "loss": 0.7223,
+      "step": 539
+    },
+    {
+      "epoch": 0.06,
+      "learning_rate": 4.9007430201140804e-05,
+      "loss": 0.5317,
+      "step": 540
+    },
+    {
+      "epoch": 0.06,
+      "learning_rate": 4.90055538877214e-05,
+      "loss": 0.8043,
+      "step": 541
+    },
+    {
+      "epoch": 0.06,
+      "learning_rate": 4.9003677574302016e-05,
+      "loss": 0.9608,
+      "step": 542
+    },
+    {
+      "epoch": 0.06,
+      "learning_rate": 4.9001801260882616e-05,
+      "loss": 0.7304,
+      "step": 543
+    },
+    {
+      "epoch": 0.06,
+      "learning_rate": 4.899992494746323e-05,
+      "loss": 0.7027,
+      "step": 544
+    },
+    {
+      "epoch": 0.06,
+      "learning_rate": 4.899804863404383e-05,
+      "loss": 0.9547,
+      "step": 545
+    },
+    {
+      "epoch": 0.06,
+      "learning_rate": 4.899617232062444e-05,
+      "loss": 0.8029,
+      "step": 546
+    },
+    {
+      "epoch": 0.06,
+      "learning_rate": 4.899429600720505e-05,
+      "loss": 0.7581,
+      "step": 547
+    },
+    {
+      "epoch": 0.06,
+      "learning_rate": 4.899241969378565e-05,
+      "loss": 0.757,
+      "step": 548
+    },
+    {
+      "epoch": 0.06,
+      "learning_rate": 4.899054338036626e-05,
+      "loss": 1.0249,
+      "step": 549
+    },
+    {
+      "epoch": 0.06,
+      "learning_rate": 4.8988667066946866e-05,
+      "loss": 0.7971,
+      "step": 550
+    },
+    {
+      "epoch": 0.06,
+      "learning_rate": 4.898679075352747e-05,
+      "loss": 0.721,
+      "step": 551
+    },
+    {
+      "epoch": 0.06,
+      "learning_rate": 4.898491444010808e-05,
+      "loss": 0.7525,
+      "step": 552
+    },
+    {
+      "epoch": 0.06,
+      "learning_rate": 4.8983038126688684e-05,
+      "loss": 0.8551,
+      "step": 553
+    },
+    {
+      "epoch": 0.06,
+      "learning_rate": 4.898116181326929e-05,
+      "loss": 0.9186,
+      "step": 554
+    },
+    {
+      "epoch": 0.06,
+      "learning_rate": 4.8979285499849896e-05,
+      "loss": 0.8027,
+      "step": 555
+    },
+    {
+      "epoch": 0.06,
+      "learning_rate": 4.897740918643051e-05,
+      "loss": 0.6168,
+      "step": 556
+    },
+    {
+      "epoch": 0.06,
+      "learning_rate": 4.897553287301111e-05,
+      "loss": 1.0176,
+      "step": 557
+    },
+    {
+      "epoch": 0.06,
+      "learning_rate": 4.897365655959172e-05,
+      "loss": 0.6712,
+      "step": 558
+    },
+    {
+      "epoch": 0.06,
+      "learning_rate": 4.897178024617232e-05,
+      "loss": 0.8377,
+      "step": 559
+    },
+    {
+      "epoch": 0.06,
+      "learning_rate": 4.8969903932752934e-05,
+      "loss": 0.9344,
+      "step": 560
+    },
+    {
+      "epoch": 0.06,
+      "learning_rate": 4.896802761933353e-05,
+      "loss": 0.767,
+      "step": 561
+    },
+    {
+      "epoch": 0.06,
+      "learning_rate": 4.8966151305914146e-05,
+      "loss": 0.683,
+      "step": 562
+    },
+    {
+      "epoch": 0.06,
+      "learning_rate": 4.8964274992494746e-05,
+      "loss": 1.002,
+      "step": 563
+    },
+    {
+      "epoch": 0.06,
+      "learning_rate": 4.896239867907536e-05,
+      "loss": 0.8311,
+      "step": 564
+    },
+    {
+      "epoch": 0.06,
+      "learning_rate": 4.8960522365655965e-05,
+      "loss": 0.9039,
+      "step": 565
+    },
+    {
+      "epoch": 0.06,
+      "learning_rate": 4.8958646052236564e-05,
+      "loss": 0.8753,
+      "step": 566
+    },
+    {
+      "epoch": 0.06,
+      "learning_rate": 4.895676973881718e-05,
+      "loss": 1.0358,
+      "step": 567
+    },
+    {
+      "epoch": 0.06,
+      "learning_rate": 4.8954893425397776e-05,
+      "loss": 0.9354,
+      "step": 568
+    },
+    {
+      "epoch": 0.06,
+      "learning_rate": 4.895301711197839e-05,
+      "loss": 0.7163,
+      "step": 569
+    },
+    {
+      "epoch": 0.06,
+      "learning_rate": 4.895114079855899e-05,
+      "loss": 0.9954,
+      "step": 570
+    },
+    {
+      "epoch": 0.06,
+      "learning_rate": 4.89492644851396e-05,
+      "loss": 0.9467,
+      "step": 571
+    },
+    {
+      "epoch": 0.06,
+      "learning_rate": 4.89473881717202e-05,
+      "loss": 0.9482,
+      "step": 572
+    },
+    {
+      "epoch": 0.06,
+      "learning_rate": 4.8945511858300814e-05,
+      "loss": 0.9584,
+      "step": 573
+    },
+    {
+      "epoch": 0.06,
+      "learning_rate": 4.894363554488142e-05,
+      "loss": 0.9096,
+      "step": 574
+    },
+    {
+      "epoch": 0.06,
+      "learning_rate": 4.8941759231462026e-05,
+      "loss": 0.8681,
+      "step": 575
+    },
+    {
+      "epoch": 0.06,
+      "learning_rate": 4.893988291804263e-05,
+      "loss": 0.7053,
+      "step": 576
+    },
+    {
+      "epoch": 0.06,
+      "learning_rate": 4.893800660462324e-05,
+      "loss": 0.8684,
+      "step": 577
+    },
+    {
+      "epoch": 0.07,
+      "learning_rate": 4.8936130291203845e-05,
+      "loss": 1.024,
+      "step": 578
+    },
+    {
+      "epoch": 0.07,
+      "learning_rate": 4.893425397778445e-05,
+      "loss": 1.0455,
+      "step": 579
+    },
+    {
+      "epoch": 0.07,
+      "learning_rate": 4.893237766436506e-05,
+      "loss": 0.8569,
+      "step": 580
+    },
+    {
+      "epoch": 0.07,
+      "learning_rate": 4.893050135094566e-05,
+      "loss": 0.8066,
+      "step": 581
+    },
+    {
+      "epoch": 0.07,
+      "learning_rate": 4.892862503752627e-05,
+      "loss": 0.955,
+      "step": 582
+    },
+    {
+      "epoch": 0.07,
+      "learning_rate": 4.8926748724106876e-05,
+      "loss": 0.8754,
+      "step": 583
+    },
+    {
+      "epoch": 0.07,
+      "learning_rate": 4.892487241068748e-05,
+      "loss": 0.7963,
+      "step": 584
+    },
+    {
+      "epoch": 0.07,
+      "learning_rate": 4.8922996097268095e-05,
+      "loss": 0.7765,
+      "step": 585
+    },
+    {
+      "epoch": 0.07,
+      "learning_rate": 4.8921119783848694e-05,
+      "loss": 0.8186,
+      "step": 586
+    },
+    {
+      "epoch": 0.07,
+      "learning_rate": 4.891924347042931e-05,
+      "loss": 0.7469,
+      "step": 587
+    },
+    {
+      "epoch": 0.07,
+      "learning_rate": 4.8917367157009906e-05,
+      "loss": 0.9223,
+      "step": 588
+    },
+    {
+      "epoch": 0.07,
+      "learning_rate": 4.891549084359052e-05,
+      "loss": 0.8197,
+      "step": 589
+    },
+    {
+      "epoch": 0.07,
+      "learning_rate": 4.891361453017112e-05,
+      "loss": 0.8213,
+      "step": 590
+    },
+    {
+      "epoch": 0.07,
+      "learning_rate": 4.891173821675173e-05,
+      "loss": 1.0843,
+      "step": 591
+    },
+    {
+      "epoch": 0.07,
+      "learning_rate": 4.890986190333233e-05,
+      "loss": 0.9244,
+      "step": 592
+    },
+    {
+      "epoch": 0.07,
+      "learning_rate": 4.8907985589912944e-05,
+      "loss": 1.0238,
+      "step": 593
+    },
+    {
+      "epoch": 0.07,
+      "learning_rate": 4.890610927649355e-05,
+      "loss": 0.843,
+      "step": 594
+    },
+    {
+      "epoch": 0.07,
+      "learning_rate": 4.8904232963074156e-05,
+      "loss": 0.6664,
+      "step": 595
+    },
+    {
+      "epoch": 0.07,
+      "learning_rate": 4.890235664965476e-05,
+      "loss": 0.9005,
+      "step": 596
+    },
+    {
+      "epoch": 0.07,
+      "learning_rate": 4.890048033623537e-05,
+      "loss": 1.0116,
+      "step": 597
+    },
+    {
+      "epoch": 0.07,
+      "learning_rate": 4.8898604022815975e-05,
+      "loss": 0.6152,
+      "step": 598
+    },
+    {
+      "epoch": 0.07,
+      "learning_rate": 4.889672770939658e-05,
+      "loss": 0.7771,
+      "step": 599
+    },
+    {
+      "epoch": 0.07,
+      "learning_rate": 4.889485139597719e-05,
+      "loss": 0.7942,
+      "step": 600
+    },
+    {
+      "epoch": 0.07,
+      "learning_rate": 4.889297508255779e-05,
+      "loss": 0.6617,
+      "step": 601
+    },
+    {
+      "epoch": 0.07,
+      "learning_rate": 4.88910987691384e-05,
+      "loss": 0.9462,
+      "step": 602
+    },
+    {
+      "epoch": 0.07,
+      "learning_rate": 4.8889222455719006e-05,
+      "loss": 0.754,
+      "step": 603
+    },
+    {
+      "epoch": 0.07,
+      "learning_rate": 4.888734614229961e-05,
+      "loss": 0.8901,
+      "step": 604
+    },
+    {
+      "epoch": 0.07,
+      "learning_rate": 4.888546982888022e-05,
+      "loss": 0.7943,
+      "step": 605
+    },
+    {
+      "epoch": 0.07,
+      "learning_rate": 4.8883593515460824e-05,
+      "loss": 0.7407,
+      "step": 606
+    },
+    {
+      "epoch": 0.07,
+      "learning_rate": 4.888171720204143e-05,
+      "loss": 0.9452,
+      "step": 607
+    },
+    {
+      "epoch": 0.07,
+      "learning_rate": 4.8879840888622036e-05,
+      "loss": 0.9279,
+      "step": 608
+    },
+    {
+      "epoch": 0.07,
+      "learning_rate": 4.887796457520264e-05,
+      "loss": 0.8962,
+      "step": 609
+    },
+    {
+      "epoch": 0.07,
+      "learning_rate": 4.887608826178325e-05,
+      "loss": 0.7001,
+      "step": 610
+    },
+    {
+      "epoch": 0.07,
+      "learning_rate": 4.8874211948363855e-05,
+      "loss": 1.0702,
+      "step": 611
+    },
+    {
+      "epoch": 0.07,
+      "learning_rate": 4.887233563494447e-05,
+      "loss": 0.9126,
+      "step": 612
+    },
+    {
+      "epoch": 0.07,
+      "learning_rate": 4.887045932152507e-05,
+      "loss": 0.7755,
+      "step": 613
+    },
+    {
+      "epoch": 0.07,
+      "learning_rate": 4.886858300810568e-05,
+      "loss": 0.9223,
+      "step": 614
+    },
+    {
+      "epoch": 0.07,
+      "learning_rate": 4.886670669468628e-05,
+      "loss": 0.9497,
+      "step": 615
+    },
+    {
+      "epoch": 0.07,
+      "learning_rate": 4.886483038126689e-05,
+      "loss": 0.9018,
+      "step": 616
+    },
+    {
+      "epoch": 0.07,
+      "learning_rate": 4.886295406784749e-05,
+      "loss": 0.949,
+      "step": 617
+    },
+    {
+      "epoch": 0.07,
+      "learning_rate": 4.8861077754428105e-05,
+      "loss": 0.6477,
+      "step": 618
+    },
+    {
+      "epoch": 0.07,
+      "learning_rate": 4.8859201441008704e-05,
+      "loss": 0.8033,
+      "step": 619
+    },
+    {
+      "epoch": 0.07,
+      "learning_rate": 4.885732512758932e-05,
+      "loss": 0.906,
+      "step": 620
+    },
+    {
+      "epoch": 0.07,
+      "learning_rate": 4.885544881416992e-05,
+      "loss": 0.7078,
+      "step": 621
+    },
+    {
+      "epoch": 0.07,
+      "learning_rate": 4.885357250075053e-05,
+      "loss": 0.9106,
+      "step": 622
+    },
+    {
+      "epoch": 0.07,
+      "learning_rate": 4.8851696187331136e-05,
+      "loss": 0.676,
+      "step": 623
+    },
+    {
+      "epoch": 0.07,
+      "learning_rate": 4.884981987391174e-05,
+      "loss": 1.3475,
+      "step": 624
+    },
+    {
+      "epoch": 0.07,
+      "learning_rate": 4.884794356049235e-05,
+      "loss": 1.0204,
+      "step": 625
+    },
+    {
+      "epoch": 0.07,
+      "eval_loss": 0.8386953473091125,
+      "eval_runtime": 23.3207,
+      "eval_samples_per_second": 43.867,
+      "eval_steps_per_second": 21.955,
+      "step": 625
+    },
+    {
+      "epoch": 0.07,
+      "learning_rate": 4.8846067247072954e-05,
+      "loss": 0.7268,
+      "step": 626
+    },
+    {
+      "epoch": 0.07,
+      "learning_rate": 4.884419093365356e-05,
+      "loss": 0.6566,
+      "step": 627
+    },
+    {
+      "epoch": 0.07,
+      "learning_rate": 4.8842314620234166e-05,
+      "loss": 0.8268,
+      "step": 628
+    },
+    {
+      "epoch": 0.07,
+      "learning_rate": 4.884043830681477e-05,
+      "loss": 0.893,
+      "step": 629
+    },
+    {
+      "epoch": 0.07,
+      "learning_rate": 4.883856199339538e-05,
+      "loss": 0.7164,
+      "step": 630
+    },
+    {
+      "epoch": 0.07,
+      "learning_rate": 4.8836685679975985e-05,
+      "loss": 0.583,
+      "step": 631
+    },
+    {
+      "epoch": 0.07,
+      "learning_rate": 4.883480936655659e-05,
+      "loss": 0.8529,
+      "step": 632
+    },
+    {
+      "epoch": 0.07,
+      "learning_rate": 4.88329330531372e-05,
+      "loss": 0.7085,
+      "step": 633
+    },
+    {
+      "epoch": 0.07,
+      "learning_rate": 4.8831056739717803e-05,
+      "loss": 0.7276,
+      "step": 634
+    },
+    {
+      "epoch": 0.07,
+      "learning_rate": 4.882918042629841e-05,
+      "loss": 0.8391,
+      "step": 635
+    },
+    {
+      "epoch": 0.07,
+      "learning_rate": 4.8827304112879016e-05,
+      "loss": 0.8692,
+      "step": 636
+    },
+    {
+      "epoch": 0.07,
+      "learning_rate": 4.882542779945962e-05,
+      "loss": 0.7463,
+      "step": 637
+    },
+    {
+      "epoch": 0.07,
+      "learning_rate": 4.882355148604023e-05,
+      "loss": 0.883,
+      "step": 638
+    },
+    {
+      "epoch": 0.07,
+      "learning_rate": 4.8821675172620834e-05,
+      "loss": 0.7533,
+      "step": 639
+    },
+    {
+      "epoch": 0.07,
+      "learning_rate": 4.881979885920144e-05,
+      "loss": 0.6626,
+      "step": 640
+    },
+    {
+      "epoch": 0.07,
+      "learning_rate": 4.881792254578205e-05,
+      "loss": 0.9475,
+      "step": 641
+    },
+    {
+      "epoch": 0.07,
+      "learning_rate": 4.881604623236265e-05,
+      "loss": 0.6184,
+      "step": 642
+    },
+    {
+      "epoch": 0.07,
+      "learning_rate": 4.8814169918943266e-05,
+      "loss": 0.924,
+      "step": 643
+    },
+    {
+      "epoch": 0.07,
+      "learning_rate": 4.8812293605523865e-05,
+      "loss": 0.6359,
+      "step": 644
+    },
+    {
+      "epoch": 0.07,
+      "learning_rate": 4.881041729210448e-05,
+      "loss": 0.794,
+      "step": 645
+    },
+    {
+      "epoch": 0.07,
+      "learning_rate": 4.880854097868508e-05,
+      "loss": 0.9821,
+      "step": 646
+    },
+    {
+      "epoch": 0.07,
+      "learning_rate": 4.880666466526569e-05,
+      "loss": 0.5388,
+      "step": 647
+    },
+    {
+      "epoch": 0.07,
+      "learning_rate": 4.880478835184629e-05,
+      "loss": 0.7905,
+      "step": 648
+    },
+    {
+      "epoch": 0.07,
+      "learning_rate": 4.88029120384269e-05,
+      "loss": 0.7254,
+      "step": 649
+    },
+    {
+      "epoch": 0.07,
+      "learning_rate": 4.880103572500751e-05,
+      "loss": 0.8294,
+      "step": 650
+    },
+    {
+      "epoch": 0.07,
+      "learning_rate": 4.8799159411588115e-05,
+      "loss": 0.7727,
+      "step": 651
+    },
+    {
+      "epoch": 0.07,
+      "learning_rate": 4.879728309816872e-05,
+      "loss": 1.0338,
+      "step": 652
+    },
+    {
+      "epoch": 0.07,
+      "learning_rate": 4.879540678474933e-05,
+      "loss": 0.9553,
+      "step": 653
+    },
+    {
+      "epoch": 0.07,
+      "learning_rate": 4.8793530471329933e-05,
+      "loss": 1.0052,
+      "step": 654
+    },
+    {
+      "epoch": 0.07,
+      "learning_rate": 4.879165415791054e-05,
+      "loss": 0.7954,
+      "step": 655
+    },
+    {
+      "epoch": 0.07,
+      "learning_rate": 4.8789777844491146e-05,
+      "loss": 0.706,
+      "step": 656
+    },
+    {
+      "epoch": 0.07,
+      "learning_rate": 4.878790153107175e-05,
+      "loss": 0.8157,
+      "step": 657
+    },
+    {
+      "epoch": 0.07,
+      "learning_rate": 4.878602521765236e-05,
+      "loss": 0.8343,
+      "step": 658
+    },
+    {
+      "epoch": 0.07,
+      "learning_rate": 4.878414890423297e-05,
+      "loss": 1.0681,
+      "step": 659
+    },
+    {
+      "epoch": 0.07,
+      "learning_rate": 4.878227259081357e-05,
+      "loss": 0.9625,
+      "step": 660
+    },
+    {
+      "epoch": 0.07,
+      "learning_rate": 4.878039627739418e-05,
+      "loss": 1.0421,
+      "step": 661
+    },
+    {
+      "epoch": 0.07,
+      "learning_rate": 4.877851996397478e-05,
+      "loss": 1.2838,
+      "step": 662
+    },
+    {
+      "epoch": 0.07,
+      "learning_rate": 4.8776643650555396e-05,
+      "loss": 0.9499,
+      "step": 663
+    },
+    {
+      "epoch": 0.07,
+      "learning_rate": 4.8774767337135995e-05,
+      "loss": 0.8515,
+      "step": 664
+    },
+    {
+      "epoch": 0.07,
+      "learning_rate": 4.877289102371661e-05,
+      "loss": 0.9643,
+      "step": 665
+    },
+    {
+      "epoch": 0.07,
+      "learning_rate": 4.877101471029721e-05,
+      "loss": 0.7203,
+      "step": 666
+    },
+    {
+      "epoch": 0.08,
+      "learning_rate": 4.876913839687782e-05,
+      "loss": 0.7585,
+      "step": 667
+    },
+    {
+      "epoch": 0.08,
+      "learning_rate": 4.8767262083458427e-05,
+      "loss": 0.9227,
+      "step": 668
+    },
+    {
+      "epoch": 0.08,
+      "learning_rate": 4.876538577003903e-05,
+      "loss": 0.6762,
+      "step": 669
+    },
+    {
+      "epoch": 0.08,
+      "learning_rate": 4.876350945661964e-05,
+      "loss": 0.5795,
+      "step": 670
+    },
+    {
+      "epoch": 0.08,
+      "learning_rate": 4.876163314320024e-05,
+      "loss": 0.8657,
+      "step": 671
+    },
+    {
+      "epoch": 0.08,
+      "learning_rate": 4.875975682978085e-05,
+      "loss": 0.9672,
+      "step": 672
+    },
+    {
+      "epoch": 0.08,
+      "learning_rate": 4.875788051636145e-05,
+      "loss": 0.9555,
+      "step": 673
+    },
+    {
+      "epoch": 0.08,
+      "learning_rate": 4.8756004202942063e-05,
+      "loss": 0.8997,
+      "step": 674
+    },
+    {
+      "epoch": 0.08,
+      "learning_rate": 4.875412788952266e-05,
+      "loss": 0.4852,
+      "step": 675
+    },
+    {
+      "epoch": 0.08,
+      "learning_rate": 4.8752251576103276e-05,
+      "loss": 0.8435,
+      "step": 676
+    },
+    {
+      "epoch": 0.08,
+      "learning_rate": 4.8750375262683875e-05,
+      "loss": 0.9512,
+      "step": 677
+    },
+    {
+      "epoch": 0.08,
+      "learning_rate": 4.874849894926449e-05,
+      "loss": 1.0042,
+      "step": 678
+    },
+    {
+      "epoch": 0.08,
+      "learning_rate": 4.8746622635845094e-05,
+      "loss": 0.9583,
+      "step": 679
+    },
+    {
+      "epoch": 0.08,
+      "learning_rate": 4.87447463224257e-05,
+      "loss": 0.8799,
+      "step": 680
+    },
+    {
+      "epoch": 0.08,
+      "learning_rate": 4.8742870009006307e-05,
+      "loss": 0.8588,
+      "step": 681
+    },
+    {
+      "epoch": 0.08,
+      "learning_rate": 4.874099369558691e-05,
+      "loss": 0.8326,
+      "step": 682
+    },
+    {
+      "epoch": 0.08,
+      "learning_rate": 4.873911738216752e-05,
+      "loss": 0.8868,
+      "step": 683
+    },
+    {
+      "epoch": 0.08,
+      "learning_rate": 4.8737241068748125e-05,
+      "loss": 0.7914,
+      "step": 684
+    },
+    {
+      "epoch": 0.08,
+      "learning_rate": 4.873536475532873e-05,
+      "loss": 0.8986,
+      "step": 685
+    },
+    {
+      "epoch": 0.08,
+      "learning_rate": 4.873348844190934e-05,
+      "loss": 0.9089,
+      "step": 686
+    },
+    {
+      "epoch": 0.08,
+      "learning_rate": 4.8731612128489944e-05,
+      "loss": 0.8578,
+      "step": 687
+    },
+    {
+      "epoch": 0.08,
+      "learning_rate": 4.8729735815070557e-05,
+      "loss": 0.7991,
+      "step": 688
+    },
+    {
+      "epoch": 0.08,
+      "learning_rate": 4.8727859501651156e-05,
+      "loss": 0.9245,
+      "step": 689
+    },
+    {
+      "epoch": 0.08,
+      "learning_rate": 4.872598318823177e-05,
+      "loss": 0.725,
+      "step": 690
+    },
+    {
+      "epoch": 0.08,
+      "learning_rate": 4.872410687481237e-05,
+      "loss": 0.5199,
+      "step": 691
+    },
+    {
+      "epoch": 0.08,
+      "learning_rate": 4.872223056139298e-05,
+      "loss": 0.6348,
+      "step": 692
+    },
+    {
+      "epoch": 0.08,
+      "learning_rate": 4.872035424797358e-05,
+      "loss": 0.9489,
+      "step": 693
+    },
+    {
+      "epoch": 0.08,
+      "learning_rate": 4.8718477934554193e-05,
+      "loss": 1.0167,
+      "step": 694
+    },
+    {
+      "epoch": 0.08,
+      "learning_rate": 4.871660162113479e-05,
+      "loss": 0.8921,
+      "step": 695
+    },
+    {
+      "epoch": 0.08,
+      "learning_rate": 4.8714725307715406e-05,
+      "loss": 0.8178,
+      "step": 696
+    },
+    {
+      "epoch": 0.08,
+      "learning_rate": 4.871284899429601e-05,
+      "loss": 0.7182,
+      "step": 697
+    },
+    {
+      "epoch": 0.08,
+      "learning_rate": 4.871097268087662e-05,
+      "loss": 0.9062,
+      "step": 698
+    },
+    {
+      "epoch": 0.08,
+      "learning_rate": 4.8709096367457224e-05,
+      "loss": 0.8742,
+      "step": 699
+    },
+    {
+      "epoch": 0.08,
+      "learning_rate": 4.870722005403783e-05,
+      "loss": 0.7977,
+      "step": 700
+    },
+    {
+      "epoch": 0.08,
+      "learning_rate": 4.8705343740618437e-05,
+      "loss": 0.7701,
+      "step": 701
+    },
+    {
+      "epoch": 0.08,
+      "learning_rate": 4.870346742719904e-05,
+      "loss": 1.033,
+      "step": 702
+    },
+    {
+      "epoch": 0.08,
+      "learning_rate": 4.870159111377965e-05,
+      "loss": 0.9158,
+      "step": 703
+    },
+    {
+      "epoch": 0.08,
+      "learning_rate": 4.8699714800360255e-05,
+      "loss": 0.8774,
+      "step": 704
+    },
+    {
+      "epoch": 0.08,
+      "learning_rate": 4.869783848694086e-05,
+      "loss": 0.9477,
+      "step": 705
+    },
+    {
+      "epoch": 0.08,
+      "learning_rate": 4.869596217352147e-05,
+      "loss": 0.7851,
+      "step": 706
+    },
+    {
+      "epoch": 0.08,
+      "learning_rate": 4.8694085860102074e-05,
+      "loss": 0.9203,
+      "step": 707
+    },
+    {
+      "epoch": 0.08,
+      "learning_rate": 4.869220954668268e-05,
+      "loss": 0.846,
+      "step": 708
+    },
+    {
+      "epoch": 0.08,
+      "learning_rate": 4.8690333233263286e-05,
+      "loss": 0.8313,
+      "step": 709
+    },
+    {
+      "epoch": 0.08,
+      "learning_rate": 4.868845691984389e-05,
+      "loss": 0.8412,
+      "step": 710
+    },
+    {
+      "epoch": 0.08,
+      "learning_rate": 4.86865806064245e-05,
+      "loss": 0.8852,
+      "step": 711
+    },
+    {
+      "epoch": 0.08,
+      "learning_rate": 4.8684704293005104e-05,
+      "loss": 0.8134,
+      "step": 712
+    },
+    {
+      "epoch": 0.08,
+      "learning_rate": 4.868282797958571e-05,
+      "loss": 0.7475,
+      "step": 713
+    },
+    {
+      "epoch": 0.08,
+      "learning_rate": 4.868095166616632e-05,
+      "loss": 1.0171,
+      "step": 714
+    },
+    {
+      "epoch": 0.08,
+      "learning_rate": 4.867907535274693e-05,
+      "loss": 1.1111,
+      "step": 715
+    },
+    {
+      "epoch": 0.08,
+      "learning_rate": 4.867719903932753e-05,
+      "loss": 0.7489,
+      "step": 716
+    },
+    {
+      "epoch": 0.08,
+      "learning_rate": 4.867532272590814e-05,
+      "loss": 0.7716,
+      "step": 717
+    },
+    {
+      "epoch": 0.08,
+      "learning_rate": 4.867344641248874e-05,
+      "loss": 0.655,
+      "step": 718
+    },
+    {
+      "epoch": 0.08,
+      "learning_rate": 4.8671570099069354e-05,
+      "loss": 0.7738,
+      "step": 719
+    },
+    {
+      "epoch": 0.08,
+      "learning_rate": 4.8669693785649954e-05,
+      "loss": 0.8358,
+      "step": 720
+    },
+    {
+      "epoch": 0.08,
+      "learning_rate": 4.8667817472230567e-05,
+      "loss": 0.7714,
+      "step": 721
+    },
+    {
+      "epoch": 0.08,
+      "learning_rate": 4.8665941158811166e-05,
+      "loss": 0.8882,
+      "step": 722
+    },
+    {
+      "epoch": 0.08,
+      "learning_rate": 4.866406484539178e-05,
+      "loss": 0.6847,
+      "step": 723
+    },
+    {
+      "epoch": 0.08,
+      "learning_rate": 4.866218853197238e-05,
+      "loss": 0.9932,
+      "step": 724
+    },
+    {
+      "epoch": 0.08,
+      "learning_rate": 4.866031221855299e-05,
+      "loss": 0.8282,
+      "step": 725
+    },
+    {
+      "epoch": 0.08,
+      "learning_rate": 4.86584359051336e-05,
+      "loss": 0.7714,
+      "step": 726
+    },
+    {
+      "epoch": 0.08,
+      "learning_rate": 4.8656559591714204e-05,
+      "loss": 0.4912,
+      "step": 727
+    },
+    {
+      "epoch": 0.08,
+      "learning_rate": 4.865468327829481e-05,
+      "loss": 0.6558,
+      "step": 728
+    },
+    {
+      "epoch": 0.08,
+      "learning_rate": 4.8652806964875416e-05,
+      "loss": 0.8594,
+      "step": 729
+    },
+    {
+      "epoch": 0.08,
+      "learning_rate": 4.865093065145602e-05,
+      "loss": 0.8172,
+      "step": 730
+    },
+    {
+      "epoch": 0.08,
+      "learning_rate": 4.864905433803663e-05,
+      "loss": 0.7169,
+      "step": 731
+    },
+    {
+      "epoch": 0.08,
+      "learning_rate": 4.8647178024617234e-05,
+      "loss": 1.1294,
+      "step": 732
+    },
+    {
+      "epoch": 0.08,
+      "learning_rate": 4.864530171119784e-05,
+      "loss": 0.6889,
+      "step": 733
+    },
+    {
+      "epoch": 0.08,
+      "learning_rate": 4.864342539777845e-05,
+      "loss": 0.8981,
+      "step": 734
+    },
+    {
+      "epoch": 0.08,
+      "learning_rate": 4.864154908435905e-05,
+      "loss": 0.8059,
+      "step": 735
+    },
+    {
+      "epoch": 0.08,
+      "learning_rate": 4.863967277093966e-05,
+      "loss": 0.7533,
+      "step": 736
+    },
+    {
+      "epoch": 0.08,
+      "learning_rate": 4.8637796457520265e-05,
+      "loss": 0.7436,
+      "step": 737
+    },
+    {
+      "epoch": 0.08,
+      "learning_rate": 4.863592014410087e-05,
+      "loss": 0.8249,
+      "step": 738
+    },
+    {
+      "epoch": 0.08,
+      "learning_rate": 4.863404383068148e-05,
+      "loss": 1.0112,
+      "step": 739
+    },
+    {
+      "epoch": 0.08,
+      "learning_rate": 4.8632167517262084e-05,
+      "loss": 0.896,
+      "step": 740
+    },
+    {
+      "epoch": 0.08,
+      "learning_rate": 4.863029120384269e-05,
+      "loss": 0.7508,
+      "step": 741
+    },
+    {
+      "epoch": 0.08,
+      "learning_rate": 4.8628414890423296e-05,
+      "loss": 0.7255,
+      "step": 742
+    },
+    {
+      "epoch": 0.08,
+      "learning_rate": 4.86265385770039e-05,
+      "loss": 0.6539,
+      "step": 743
+    },
+    {
+      "epoch": 0.08,
+      "learning_rate": 4.8624662263584515e-05,
+      "loss": 0.5139,
+      "step": 744
+    },
+    {
+      "epoch": 0.08,
+      "learning_rate": 4.8622785950165114e-05,
+      "loss": 0.7258,
+      "step": 745
+    },
+    {
+      "epoch": 0.08,
+      "learning_rate": 4.862090963674573e-05,
+      "loss": 0.7497,
+      "step": 746
+    },
+    {
+      "epoch": 0.08,
+      "learning_rate": 4.861903332332633e-05,
+      "loss": 0.8153,
+      "step": 747
+    },
+    {
+      "epoch": 0.08,
+      "learning_rate": 4.861715700990694e-05,
+      "loss": 0.9408,
+      "step": 748
+    },
+    {
+      "epoch": 0.08,
+      "learning_rate": 4.861528069648754e-05,
+      "loss": 0.678,
+      "step": 749
+    },
+    {
+      "epoch": 0.08,
+      "learning_rate": 4.861340438306815e-05,
+      "loss": 0.7238,
+      "step": 750
+    },
+    {
+      "epoch": 0.08,
+      "eval_loss": 0.8382143974304199,
+      "eval_runtime": 23.2739,
+      "eval_samples_per_second": 43.955,
+      "eval_steps_per_second": 21.999,
+      "step": 750
+    },
+    {
+      "epoch": 0.08,
+      "learning_rate": 4.861152806964875e-05,
+      "loss": 1.0734,
+      "step": 751
+    },
+    {
+      "epoch": 0.08,
+      "learning_rate": 4.8609651756229364e-05,
+      "loss": 0.7758,
+      "step": 752
+    },
+    {
+      "epoch": 0.08,
+      "learning_rate": 4.860777544280997e-05,
+      "loss": 0.747,
+      "step": 753
+    },
+    {
+      "epoch": 0.08,
+      "learning_rate": 4.860589912939058e-05,
+      "loss": 0.8734,
+      "step": 754
+    },
+    {
+      "epoch": 0.08,
+      "learning_rate": 4.860402281597118e-05,
+      "loss": 1.0152,
+      "step": 755
+    },
+    {
+      "epoch": 0.09,
+      "learning_rate": 4.860214650255179e-05,
+      "loss": 1.0585,
+      "step": 756
+    },
+    {
+      "epoch": 0.09,
+      "learning_rate": 4.8600270189132395e-05,
+      "loss": 0.922,
+      "step": 757
+    },
+    {
+      "epoch": 0.09,
+      "learning_rate": 4.8598393875713e-05,
+      "loss": 0.9073,
+      "step": 758
+    },
+    {
+      "epoch": 0.09,
+      "learning_rate": 4.859651756229361e-05,
+      "loss": 0.8672,
+      "step": 759
+    },
+    {
+      "epoch": 0.09,
+      "learning_rate": 4.8594641248874214e-05,
+      "loss": 0.6201,
+      "step": 760
+    },
+    {
+      "epoch": 0.09,
+      "learning_rate": 4.859276493545482e-05,
+      "loss": 0.8183,
+      "step": 761
+    },
+    {
+      "epoch": 0.09,
+      "learning_rate": 4.859088862203543e-05,
+      "loss": 0.8566,
+      "step": 762
+    },
+    {
+      "epoch": 0.09,
+      "learning_rate": 4.858901230861603e-05,
+      "loss": 0.7723,
+      "step": 763
+    },
+    {
+      "epoch": 0.09,
+      "learning_rate": 4.8587135995196645e-05,
+      "loss": 0.758,
+      "step": 764
+    },
+    {
+      "epoch": 0.09,
+      "learning_rate": 4.8585259681777244e-05,
+      "loss": 0.914,
+      "step": 765
+    },
+    {
+      "epoch": 0.09,
+      "learning_rate": 4.858338336835786e-05,
+      "loss": 0.7292,
+      "step": 766
+    },
+    {
+      "epoch": 0.09,
+      "learning_rate": 4.858150705493846e-05,
+      "loss": 0.8119,
+      "step": 767
+    },
+    {
+      "epoch": 0.09,
+      "learning_rate": 4.857963074151907e-05,
+      "loss": 0.9238,
+      "step": 768
+    },
+    {
+      "epoch": 0.09,
+      "learning_rate": 4.857775442809967e-05,
+      "loss": 0.8213,
+      "step": 769
+    },
+    {
+      "epoch": 0.09,
+      "learning_rate": 4.857587811468028e-05,
+      "loss": 0.9829,
+      "step": 770
+    },
+    {
+      "epoch": 0.09,
+      "learning_rate": 4.857400180126088e-05,
+      "loss": 0.8231,
+      "step": 771
+    },
+    {
+      "epoch": 0.09,
+      "learning_rate": 4.8572125487841494e-05,
+      "loss": 0.7742,
+      "step": 772
+    },
+    {
+      "epoch": 0.09,
+      "learning_rate": 4.85702491744221e-05,
+      "loss": 0.9536,
+      "step": 773
+    },
+    {
+      "epoch": 0.09,
+      "learning_rate": 4.856837286100271e-05,
+      "loss": 0.7402,
+      "step": 774
+    },
+    {
+      "epoch": 0.09,
+      "learning_rate": 4.856649654758331e-05,
+      "loss": 0.8287,
+      "step": 775
+    },
+    {
+      "epoch": 0.09,
+      "learning_rate": 4.856462023416391e-05,
+      "loss": 0.8759,
+      "step": 776
+    },
+    {
+      "epoch": 0.09,
+      "learning_rate": 4.8562743920744525e-05,
+      "loss": 0.8538,
+      "step": 777
+    },
+    {
+      "epoch": 0.09,
+      "learning_rate": 4.8560867607325125e-05,
+      "loss": 0.7663,
+      "step": 778
+    },
+    {
+      "epoch": 0.09,
+      "learning_rate": 4.855899129390574e-05,
+      "loss": 1.0591,
+      "step": 779
+    },
+    {
+      "epoch": 0.09,
+      "learning_rate": 4.855711498048634e-05,
+      "loss": 0.8529,
+      "step": 780
+    },
+    {
+      "epoch": 0.09,
+      "learning_rate": 4.855523866706695e-05,
+      "loss": 1.0588,
+      "step": 781
+    },
+    {
+      "epoch": 0.09,
+      "learning_rate": 4.8553362353647556e-05,
+      "loss": 0.6651,
+      "step": 782
+    },
+    {
+      "epoch": 0.09,
+      "learning_rate": 4.855148604022816e-05,
+      "loss": 0.8395,
+      "step": 783
+    },
+    {
+      "epoch": 0.09,
+      "learning_rate": 4.854960972680877e-05,
+      "loss": 0.9055,
+      "step": 784
+    },
+    {
+      "epoch": 0.09,
+      "learning_rate": 4.8547733413389374e-05,
+      "loss": 0.7884,
+      "step": 785
+    },
+    {
+      "epoch": 0.09,
+      "learning_rate": 4.854585709996998e-05,
+      "loss": 0.8046,
+      "step": 786
+    },
+    {
+      "epoch": 0.09,
+      "learning_rate": 4.854398078655059e-05,
+      "loss": 0.8955,
+      "step": 787
+    },
+    {
+      "epoch": 0.09,
+      "learning_rate": 4.854210447313119e-05,
+      "loss": 0.9037,
+      "step": 788
+    },
+    {
+      "epoch": 0.09,
+      "learning_rate": 4.85402281597118e-05,
+      "loss": 0.5423,
+      "step": 789
+    },
+    {
+      "epoch": 0.09,
+      "learning_rate": 4.8538351846292405e-05,
+      "loss": 0.7103,
+      "step": 790
+    },
+    {
+      "epoch": 0.09,
+      "learning_rate": 4.853647553287302e-05,
+      "loss": 1.0082,
+      "step": 791
+    },
+    {
+      "epoch": 0.09,
+      "learning_rate": 4.853459921945362e-05,
+      "loss": 0.9003,
+      "step": 792
+    },
+    {
+      "epoch": 0.09,
+      "learning_rate": 4.853272290603423e-05,
+      "loss": 0.9384,
+      "step": 793
+    },
+    {
+      "epoch": 0.09,
+      "learning_rate": 4.853084659261483e-05,
+      "loss": 0.9275,
+      "step": 794
+    },
+    {
+      "epoch": 0.09,
+      "learning_rate": 4.852897027919544e-05,
+      "loss": 0.6847,
+      "step": 795
+    },
+    {
+      "epoch": 0.09,
+      "learning_rate": 4.852709396577604e-05,
+      "loss": 0.5573,
+      "step": 796
+    },
+    {
+      "epoch": 0.09,
+      "learning_rate": 4.8525217652356655e-05,
+      "loss": 0.6081,
+      "step": 797
+    },
+    {
+      "epoch": 0.09,
+      "learning_rate": 4.8523341338937255e-05,
+      "loss": 0.8629,
+      "step": 798
+    },
+    {
+      "epoch": 0.09,
+      "learning_rate": 4.852146502551787e-05,
+      "loss": 0.5761,
+      "step": 799
+    },
+    {
+      "epoch": 0.09,
+      "learning_rate": 4.8519588712098474e-05,
+      "loss": 0.7099,
+      "step": 800
+    },
+    {
+      "epoch": 0.09,
+      "learning_rate": 4.851771239867908e-05,
+      "loss": 1.0826,
+      "step": 801
+    },
+    {
+      "epoch": 0.09,
+      "learning_rate": 4.8515836085259686e-05,
+      "loss": 0.9529,
+      "step": 802
+    },
+    {
+      "epoch": 0.09,
+      "learning_rate": 4.851395977184029e-05,
+      "loss": 0.6763,
+      "step": 803
+    },
+    {
+      "epoch": 0.09,
+      "learning_rate": 4.85120834584209e-05,
+      "loss": 0.8367,
+      "step": 804
+    },
+    {
+      "epoch": 0.09,
+      "learning_rate": 4.8510207145001504e-05,
+      "loss": 0.8484,
+      "step": 805
+    },
+    {
+      "epoch": 0.09,
+      "learning_rate": 4.850833083158211e-05,
+      "loss": 0.8922,
+      "step": 806
+    },
+    {
+      "epoch": 0.09,
+      "learning_rate": 4.850645451816272e-05,
+      "loss": 0.7552,
+      "step": 807
+    },
+    {
+      "epoch": 0.09,
+      "learning_rate": 4.850457820474332e-05,
+      "loss": 0.9922,
+      "step": 808
+    },
+    {
+      "epoch": 0.09,
+      "learning_rate": 4.850270189132393e-05,
+      "loss": 0.8483,
+      "step": 809
+    },
+    {
+      "epoch": 0.09,
+      "learning_rate": 4.8500825577904535e-05,
+      "loss": 0.9775,
+      "step": 810
+    },
+    {
+      "epoch": 0.09,
+      "learning_rate": 4.849894926448514e-05,
+      "loss": 0.8145,
+      "step": 811
+    },
+    {
+      "epoch": 0.09,
+      "learning_rate": 4.849707295106575e-05,
+      "loss": 0.7837,
+      "step": 812
+    },
+    {
+      "epoch": 0.09,
+      "learning_rate": 4.8495196637646354e-05,
+      "loss": 0.7336,
+      "step": 813
+    },
+    {
+      "epoch": 0.09,
+      "learning_rate": 4.849332032422696e-05,
+      "loss": 0.9331,
+      "step": 814
+    },
+    {
+      "epoch": 0.09,
+      "learning_rate": 4.8491444010807566e-05,
+      "loss": 0.7345,
+      "step": 815
+    },
+    {
+      "epoch": 0.09,
+      "learning_rate": 4.848956769738817e-05,
+      "loss": 0.7843,
+      "step": 816
+    },
+    {
+      "epoch": 0.09,
+      "learning_rate": 4.848769138396878e-05,
+      "loss": 0.733,
+      "step": 817
+    },
+    {
+      "epoch": 0.09,
+      "learning_rate": 4.848581507054939e-05,
+      "loss": 0.8204,
+      "step": 818
+    },
+    {
+      "epoch": 0.09,
+      "learning_rate": 4.848393875712999e-05,
+      "loss": 1.0632,
+      "step": 819
+    },
+    {
+      "epoch": 0.09,
+      "learning_rate": 4.8482062443710604e-05,
+      "loss": 0.6843,
+      "step": 820
+    },
+    {
+      "epoch": 0.09,
+      "learning_rate": 4.84801861302912e-05,
+      "loss": 1.0393,
+      "step": 821
+    },
+    {
+      "epoch": 0.09,
+      "learning_rate": 4.8478309816871816e-05,
+      "loss": 0.9722,
+      "step": 822
+    },
+    {
+      "epoch": 0.09,
+      "learning_rate": 4.8476433503452415e-05,
+      "loss": 0.8166,
+      "step": 823
+    },
+    {
+      "epoch": 0.09,
+      "learning_rate": 4.847455719003303e-05,
+      "loss": 0.8661,
+      "step": 824
+    },
+    {
+      "epoch": 0.09,
+      "learning_rate": 4.847268087661363e-05,
+      "loss": 0.7098,
+      "step": 825
+    },
+    {
+      "epoch": 0.09,
+      "learning_rate": 4.847080456319424e-05,
+      "loss": 0.8272,
+      "step": 826
+    },
+    {
+      "epoch": 0.09,
+      "learning_rate": 4.846892824977484e-05,
+      "loss": 0.8542,
+      "step": 827
+    },
+    {
+      "epoch": 0.09,
+      "learning_rate": 4.846705193635545e-05,
+      "loss": 0.6482,
+      "step": 828
+    },
+    {
+      "epoch": 0.09,
+      "learning_rate": 4.846517562293606e-05,
+      "loss": 0.7476,
+      "step": 829
+    },
+    {
+      "epoch": 0.09,
+      "learning_rate": 4.8463299309516665e-05,
+      "loss": 0.8089,
+      "step": 830
+    },
+    {
+      "epoch": 0.09,
+      "learning_rate": 4.846142299609727e-05,
+      "loss": 0.8943,
+      "step": 831
+    },
+    {
+      "epoch": 0.09,
+      "learning_rate": 4.845954668267788e-05,
+      "loss": 0.8547,
+      "step": 832
+    },
+    {
+      "epoch": 0.09,
+      "learning_rate": 4.8457670369258484e-05,
+      "loss": 0.9796,
+      "step": 833
+    },
+    {
+      "epoch": 0.09,
+      "learning_rate": 4.845579405583909e-05,
+      "loss": 1.0832,
+      "step": 834
+    },
+    {
+      "epoch": 0.09,
+      "learning_rate": 4.8453917742419696e-05,
+      "loss": 0.8018,
+      "step": 835
+    },
+    {
+      "epoch": 0.09,
+      "learning_rate": 4.84520414290003e-05,
+      "loss": 0.6061,
+      "step": 836
+    },
+    {
+      "epoch": 0.09,
+      "learning_rate": 4.845016511558091e-05,
+      "loss": 0.9206,
+      "step": 837
+    },
+    {
+      "epoch": 0.09,
+      "learning_rate": 4.8448288802161515e-05,
+      "loss": 0.7035,
+      "step": 838
+    },
+    {
+      "epoch": 0.09,
+      "learning_rate": 4.844641248874212e-05,
+      "loss": 0.5935,
+      "step": 839
+    },
+    {
+      "epoch": 0.09,
+      "learning_rate": 4.844453617532273e-05,
+      "loss": 0.7761,
+      "step": 840
+    },
+    {
+      "epoch": 0.09,
+      "learning_rate": 4.844265986190333e-05,
+      "loss": 0.86,
+      "step": 841
+    },
+    {
+      "epoch": 0.09,
+      "learning_rate": 4.844078354848394e-05,
+      "loss": 0.9211,
+      "step": 842
+    },
+    {
+      "epoch": 0.09,
+      "learning_rate": 4.8438907235064545e-05,
+      "loss": 1.0339,
+      "step": 843
+    },
+    {
+      "epoch": 0.09,
+      "learning_rate": 4.843703092164515e-05,
+      "loss": 0.6919,
+      "step": 844
+    },
+    {
+      "epoch": 0.1,
+      "learning_rate": 4.843515460822576e-05,
+      "loss": 0.6919,
+      "step": 845
+    },
+    {
+      "epoch": 0.1,
+      "learning_rate": 4.8433278294806364e-05,
+      "loss": 1.0005,
+      "step": 846
+    },
+    {
+      "epoch": 0.1,
+      "learning_rate": 4.843140198138698e-05,
+      "loss": 0.9554,
+      "step": 847
+    },
+    {
+      "epoch": 0.1,
+      "learning_rate": 4.8429525667967576e-05,
+      "loss": 0.8868,
+      "step": 848
+    },
+    {
+      "epoch": 0.1,
+      "learning_rate": 4.842764935454819e-05,
+      "loss": 0.9183,
+      "step": 849
+    },
+    {
+      "epoch": 0.1,
+      "learning_rate": 4.842577304112879e-05,
+      "loss": 0.9071,
+      "step": 850
+    },
+    {
+      "epoch": 0.1,
+      "learning_rate": 4.84238967277094e-05,
+      "loss": 1.0527,
+      "step": 851
+    },
+    {
+      "epoch": 0.1,
+      "learning_rate": 4.842202041429e-05,
+      "loss": 0.8835,
+      "step": 852
+    },
+    {
+      "epoch": 0.1,
+      "learning_rate": 4.8420144100870614e-05,
+      "loss": 1.0234,
+      "step": 853
+    },
+    {
+      "epoch": 0.1,
+      "learning_rate": 4.841826778745121e-05,
+      "loss": 0.7471,
+      "step": 854
+    },
+    {
+      "epoch": 0.1,
+      "learning_rate": 4.8416391474031826e-05,
+      "loss": 0.9399,
+      "step": 855
+    },
+    {
+      "epoch": 0.1,
+      "learning_rate": 4.841451516061243e-05,
+      "loss": 0.9079,
+      "step": 856
+    },
+    {
+      "epoch": 0.1,
+      "learning_rate": 4.841263884719304e-05,
+      "loss": 0.9864,
+      "step": 857
+    },
+    {
+      "epoch": 0.1,
+      "learning_rate": 4.8410762533773645e-05,
+      "loss": 0.8399,
+      "step": 858
+    },
+    {
+      "epoch": 0.1,
+      "learning_rate": 4.840888622035425e-05,
+      "loss": 1.0736,
+      "step": 859
+    },
+    {
+      "epoch": 0.1,
+      "learning_rate": 4.840700990693486e-05,
+      "loss": 0.6895,
+      "step": 860
+    },
+    {
+      "epoch": 0.1,
+      "learning_rate": 4.840513359351546e-05,
+      "loss": 1.1455,
+      "step": 861
+    },
+    {
+      "epoch": 0.1,
+      "learning_rate": 4.840325728009607e-05,
+      "loss": 0.5564,
+      "step": 862
+    },
+    {
+      "epoch": 0.1,
+      "learning_rate": 4.8401380966676675e-05,
+      "loss": 1.192,
+      "step": 863
+    },
+    {
+      "epoch": 0.1,
+      "learning_rate": 4.839950465325728e-05,
+      "loss": 0.801,
+      "step": 864
+    },
+    {
+      "epoch": 0.1,
+      "learning_rate": 4.8397628339837894e-05,
+      "loss": 0.6923,
+      "step": 865
+    },
+    {
+      "epoch": 0.1,
+      "learning_rate": 4.8395752026418494e-05,
+      "loss": 0.9024,
+      "step": 866
+    },
+    {
+      "epoch": 0.1,
+      "learning_rate": 4.839387571299911e-05,
+      "loss": 0.8536,
+      "step": 867
+    },
+    {
+      "epoch": 0.1,
+      "learning_rate": 4.8391999399579706e-05,
+      "loss": 0.9223,
+      "step": 868
+    },
+    {
+      "epoch": 0.1,
+      "learning_rate": 4.839012308616032e-05,
+      "loss": 0.7735,
+      "step": 869
+    },
+    {
+      "epoch": 0.1,
+      "learning_rate": 4.838824677274092e-05,
+      "loss": 1.2006,
+      "step": 870
+    },
+    {
+      "epoch": 0.1,
+      "learning_rate": 4.838637045932153e-05,
+      "loss": 0.7558,
+      "step": 871
+    },
+    {
+      "epoch": 0.1,
+      "learning_rate": 4.838449414590213e-05,
+      "loss": 1.0054,
+      "step": 872
+    },
+    {
+      "epoch": 0.1,
+      "learning_rate": 4.8382617832482744e-05,
+      "loss": 0.8767,
+      "step": 873
+    },
+    {
+      "epoch": 0.1,
+      "learning_rate": 4.838074151906334e-05,
+      "loss": 0.7963,
+      "step": 874
+    },
+    {
+      "epoch": 0.1,
+      "learning_rate": 4.8378865205643956e-05,
+      "loss": 0.8052,
+      "step": 875
+    },
+    {
+      "epoch": 0.1,
+      "eval_loss": 0.8410152792930603,
+      "eval_runtime": 23.2674,
+      "eval_samples_per_second": 43.967,
+      "eval_steps_per_second": 22.005,
+      "step": 875
+    },
+    {
+      "epoch": 0.1,
+      "learning_rate": 4.837698889222456e-05,
+      "loss": 0.9137,
+      "step": 876
+    },
+    {
+      "epoch": 0.1,
+      "learning_rate": 4.837511257880517e-05,
+      "loss": 0.8211,
+      "step": 877
+    },
+    {
+      "epoch": 0.1,
+      "learning_rate": 4.8373236265385775e-05,
+      "loss": 0.7872,
+      "step": 878
+    },
+    {
+      "epoch": 0.1,
+      "learning_rate": 4.837135995196638e-05,
+      "loss": 0.8205,
+      "step": 879
+    },
+    {
+      "epoch": 0.1,
+      "learning_rate": 4.836948363854699e-05,
+      "loss": 0.9647,
+      "step": 880
+    },
+    {
+      "epoch": 0.1,
+      "learning_rate": 4.8367607325127586e-05,
+      "loss": 0.6533,
+      "step": 881
+    },
+    {
+      "epoch": 0.1,
+      "learning_rate": 4.83657310117082e-05,
+      "loss": 0.809,
+      "step": 882
+    },
+    {
+      "epoch": 0.1,
+      "learning_rate": 4.83638546982888e-05,
+      "loss": 0.7082,
+      "step": 883
+    },
+    {
+      "epoch": 0.1,
+      "learning_rate": 4.836197838486941e-05,
+      "loss": 0.8901,
+      "step": 884
+    },
+    {
+      "epoch": 0.1,
+      "learning_rate": 4.836010207145002e-05,
+      "loss": 0.781,
+      "step": 885
+    },
+    {
+      "epoch": 0.1,
+      "learning_rate": 4.8358225758030624e-05,
+      "loss": 0.7063,
+      "step": 886
+    },
+    {
+      "epoch": 0.1,
+      "learning_rate": 4.835634944461123e-05,
+      "loss": 0.6902,
+      "step": 887
+    },
+    {
+      "epoch": 0.1,
+      "learning_rate": 4.8354473131191836e-05,
+      "loss": 1.0274,
+      "step": 888
+    },
+    {
+      "epoch": 0.1,
+      "learning_rate": 4.835259681777244e-05,
+      "loss": 0.9839,
+      "step": 889
+    },
+    {
+      "epoch": 0.1,
+      "learning_rate": 4.835072050435305e-05,
+      "loss": 0.7253,
+      "step": 890
+    },
+    {
+      "epoch": 0.1,
+      "learning_rate": 4.8348844190933655e-05,
+      "loss": 0.4341,
+      "step": 891
+    },
+    {
+      "epoch": 0.1,
+      "learning_rate": 4.834696787751426e-05,
+      "loss": 0.7394,
+      "step": 892
+    },
+    {
+      "epoch": 0.1,
+      "learning_rate": 4.834509156409487e-05,
+      "loss": 0.8443,
+      "step": 893
+    },
+    {
+      "epoch": 0.1,
+      "learning_rate": 4.834321525067548e-05,
+      "loss": 1.0265,
+      "step": 894
+    },
+    {
+      "epoch": 0.1,
+      "learning_rate": 4.834133893725608e-05,
+      "loss": 0.9531,
+      "step": 895
+    },
+    {
+      "epoch": 0.1,
+      "learning_rate": 4.833946262383669e-05,
+      "loss": 0.7891,
+      "step": 896
+    },
+    {
+      "epoch": 0.1,
+      "learning_rate": 4.833758631041729e-05,
+      "loss": 0.8041,
+      "step": 897
+    },
+    {
+      "epoch": 0.1,
+      "learning_rate": 4.8335709996997905e-05,
+      "loss": 0.8085,
+      "step": 898
+    },
+    {
+      "epoch": 0.1,
+      "learning_rate": 4.8333833683578504e-05,
+      "loss": 0.8777,
+      "step": 899
+    },
+    {
+      "epoch": 0.1,
+      "learning_rate": 4.833195737015912e-05,
+      "loss": 0.7744,
+      "step": 900
+    },
+    {
+      "epoch": 0.1,
+      "learning_rate": 4.8330081056739716e-05,
+      "loss": 0.9111,
+      "step": 901
+    },
+    {
+      "epoch": 0.1,
+      "learning_rate": 4.832820474332033e-05,
+      "loss": 0.7524,
+      "step": 902
+    },
+    {
+      "epoch": 0.1,
+      "learning_rate": 4.8326328429900935e-05,
+      "loss": 0.9277,
+      "step": 903
+    },
+    {
+      "epoch": 0.1,
+      "learning_rate": 4.832445211648154e-05,
+      "loss": 0.9986,
+      "step": 904
+    },
+    {
+      "epoch": 0.1,
+      "learning_rate": 4.832257580306215e-05,
+      "loss": 0.916,
+      "step": 905
+    },
+    {
+      "epoch": 0.1,
+      "learning_rate": 4.8320699489642754e-05,
+      "loss": 0.9307,
+      "step": 906
+    },
+    {
+      "epoch": 0.1,
+      "learning_rate": 4.831882317622336e-05,
+      "loss": 0.4747,
+      "step": 907
+    },
+    {
+      "epoch": 0.1,
+      "learning_rate": 4.8316946862803966e-05,
+      "loss": 0.6906,
+      "step": 908
+    },
+    {
+      "epoch": 0.1,
+      "learning_rate": 4.831507054938457e-05,
+      "loss": 0.7728,
+      "step": 909
+    },
+    {
+      "epoch": 0.1,
+      "learning_rate": 4.831319423596518e-05,
+      "loss": 0.9248,
+      "step": 910
+    },
+    {
+      "epoch": 0.1,
+      "learning_rate": 4.8311317922545785e-05,
+      "loss": 0.8372,
+      "step": 911
+    },
+    {
+      "epoch": 0.1,
+      "learning_rate": 4.830944160912639e-05,
+      "loss": 0.8889,
+      "step": 912
+    },
+    {
+      "epoch": 0.1,
+      "learning_rate": 4.8307565295707e-05,
+      "loss": 0.8678,
+      "step": 913
+    },
+    {
+      "epoch": 0.1,
+      "learning_rate": 4.83056889822876e-05,
+      "loss": 1.0076,
+      "step": 914
+    },
+    {
+      "epoch": 0.1,
+      "learning_rate": 4.830381266886821e-05,
+      "loss": 0.7158,
+      "step": 915
+    },
+    {
+      "epoch": 0.1,
+      "learning_rate": 4.8301936355448815e-05,
+      "loss": 0.7648,
+      "step": 916
+    },
+    {
+      "epoch": 0.1,
+      "learning_rate": 4.830006004202942e-05,
+      "loss": 0.9556,
+      "step": 917
+    },
+    {
+      "epoch": 0.1,
+      "learning_rate": 4.829818372861003e-05,
+      "loss": 0.8648,
+      "step": 918
+    },
+    {
+      "epoch": 0.1,
+      "learning_rate": 4.8296307415190634e-05,
+      "loss": 0.8995,
+      "step": 919
+    },
+    {
+      "epoch": 0.1,
+      "learning_rate": 4.829443110177124e-05,
+      "loss": 0.9504,
+      "step": 920
+    },
+    {
+      "epoch": 0.1,
+      "learning_rate": 4.8292554788351846e-05,
+      "loss": 0.7443,
+      "step": 921
+    },
+    {
+      "epoch": 0.1,
+      "learning_rate": 4.829067847493245e-05,
+      "loss": 0.7777,
+      "step": 922
+    },
+    {
+      "epoch": 0.1,
+      "learning_rate": 4.8288802161513065e-05,
+      "loss": 0.7964,
+      "step": 923
+    },
+    {
+      "epoch": 0.1,
+      "learning_rate": 4.8286925848093665e-05,
+      "loss": 0.7275,
+      "step": 924
+    },
+    {
+      "epoch": 0.1,
+      "learning_rate": 4.828504953467428e-05,
+      "loss": 0.7551,
+      "step": 925
+    },
+    {
+      "epoch": 0.1,
+      "learning_rate": 4.828317322125488e-05,
+      "loss": 0.9491,
+      "step": 926
+    },
+    {
+      "epoch": 0.1,
+      "learning_rate": 4.828129690783549e-05,
+      "loss": 0.7844,
+      "step": 927
+    },
+    {
+      "epoch": 0.1,
+      "learning_rate": 4.827942059441609e-05,
+      "loss": 0.7079,
+      "step": 928
+    },
+    {
+      "epoch": 0.1,
+      "learning_rate": 4.82775442809967e-05,
+      "loss": 0.9315,
+      "step": 929
+    },
+    {
+      "epoch": 0.1,
+      "learning_rate": 4.82756679675773e-05,
+      "loss": 0.6758,
+      "step": 930
+    },
+    {
+      "epoch": 0.1,
+      "learning_rate": 4.8273791654157915e-05,
+      "loss": 1.423,
+      "step": 931
+    },
+    {
+      "epoch": 0.1,
+      "learning_rate": 4.827191534073852e-05,
+      "loss": 0.7734,
+      "step": 932
+    },
+    {
+      "epoch": 0.1,
+      "learning_rate": 4.827003902731913e-05,
+      "loss": 0.5214,
+      "step": 933
+    },
+    {
+      "epoch": 0.11,
+      "learning_rate": 4.826816271389973e-05,
+      "loss": 0.7521,
+      "step": 934
+    },
+    {
+      "epoch": 0.11,
+      "learning_rate": 4.826628640048034e-05,
+      "loss": 0.9078,
+      "step": 935
+    },
+    {
+      "epoch": 0.11,
+      "learning_rate": 4.8264410087060945e-05,
+      "loss": 0.8248,
+      "step": 936
+    },
+    {
+      "epoch": 0.11,
+      "learning_rate": 4.826253377364155e-05,
+      "loss": 0.8258,
+      "step": 937
+    },
+    {
+      "epoch": 0.11,
+      "learning_rate": 4.826065746022216e-05,
+      "loss": 0.7295,
+      "step": 938
+    },
+    {
+      "epoch": 0.11,
+      "learning_rate": 4.8258781146802764e-05,
+      "loss": 0.829,
+      "step": 939
+    },
+    {
+      "epoch": 0.11,
+      "learning_rate": 4.825690483338337e-05,
+      "loss": 0.6932,
+      "step": 940
+    },
+    {
+      "epoch": 0.11,
+      "learning_rate": 4.8255028519963976e-05,
+      "loss": 0.8445,
+      "step": 941
+    },
+    {
+      "epoch": 0.11,
+      "learning_rate": 4.825315220654458e-05,
+      "loss": 0.7253,
+      "step": 942
+    },
+    {
+      "epoch": 0.11,
+      "learning_rate": 4.825127589312519e-05,
+      "loss": 0.7039,
+      "step": 943
+    },
+    {
+      "epoch": 0.11,
+      "learning_rate": 4.8249399579705795e-05,
+      "loss": 1.027,
+      "step": 944
+    },
+    {
+      "epoch": 0.11,
+      "learning_rate": 4.82475232662864e-05,
+      "loss": 0.8019,
+      "step": 945
+    },
+    {
+      "epoch": 0.11,
+      "learning_rate": 4.824564695286701e-05,
+      "loss": 0.921,
+      "step": 946
+    },
+    {
+      "epoch": 0.11,
+      "learning_rate": 4.824377063944761e-05,
+      "loss": 0.9166,
+      "step": 947
+    },
+    {
+      "epoch": 0.11,
+      "learning_rate": 4.824189432602822e-05,
+      "loss": 1.0117,
+      "step": 948
+    },
+    {
+      "epoch": 0.11,
+      "learning_rate": 4.8240018012608826e-05,
+      "loss": 0.8809,
+      "step": 949
+    },
+    {
+      "epoch": 0.11,
+      "learning_rate": 4.823814169918944e-05,
+      "loss": 0.9183,
+      "step": 950
+    },
+    {
+      "epoch": 0.11,
+      "learning_rate": 4.823626538577004e-05,
+      "loss": 0.7974,
+      "step": 951
+    },
+    {
+      "epoch": 0.11,
+      "learning_rate": 4.823438907235065e-05,
+      "loss": 0.7634,
+      "step": 952
+    },
+    {
+      "epoch": 0.11,
+      "learning_rate": 4.823251275893125e-05,
+      "loss": 1.1069,
+      "step": 953
+    },
+    {
+      "epoch": 0.11,
+      "learning_rate": 4.823063644551186e-05,
+      "loss": 0.7772,
+      "step": 954
+    },
+    {
+      "epoch": 0.11,
+      "learning_rate": 4.822876013209246e-05,
+      "loss": 0.8435,
+      "step": 955
+    },
+    {
+      "epoch": 0.11,
+      "learning_rate": 4.8226883818673075e-05,
+      "loss": 0.5424,
+      "step": 956
+    },
+    {
+      "epoch": 0.11,
+      "learning_rate": 4.8225007505253675e-05,
+      "loss": 1.0266,
+      "step": 957
+    },
+    {
+      "epoch": 0.11,
+      "learning_rate": 4.822313119183429e-05,
+      "loss": 0.8429,
+      "step": 958
+    },
+    {
+      "epoch": 0.11,
+      "learning_rate": 4.8221254878414894e-05,
+      "loss": 0.9112,
+      "step": 959
+    },
+    {
+      "epoch": 0.11,
+      "learning_rate": 4.82193785649955e-05,
+      "loss": 0.8436,
+      "step": 960
+    },
+    {
+      "epoch": 0.11,
+      "learning_rate": 4.8217502251576106e-05,
+      "loss": 0.8641,
+      "step": 961
+    },
+    {
+      "epoch": 0.11,
+      "learning_rate": 4.821562593815671e-05,
+      "loss": 0.9221,
+      "step": 962
+    },
+    {
+      "epoch": 0.11,
+      "learning_rate": 4.821374962473732e-05,
+      "loss": 0.7731,
+      "step": 963
+    },
+    {
+      "epoch": 0.11,
+      "learning_rate": 4.8211873311317925e-05,
+      "loss": 0.8339,
+      "step": 964
+    },
+    {
+      "epoch": 0.11,
+      "learning_rate": 4.820999699789853e-05,
+      "loss": 1.0349,
+      "step": 965
+    },
+    {
+      "epoch": 0.11,
+      "learning_rate": 4.820812068447914e-05,
+      "loss": 0.9599,
+      "step": 966
+    },
+    {
+      "epoch": 0.11,
+      "learning_rate": 4.820624437105974e-05,
+      "loss": 0.7103,
+      "step": 967
+    },
+    {
+      "epoch": 0.11,
+      "learning_rate": 4.820436805764035e-05,
+      "loss": 0.9432,
+      "step": 968
+    },
+    {
+      "epoch": 0.11,
+      "learning_rate": 4.8202491744220956e-05,
+      "loss": 0.6391,
+      "step": 969
+    },
+    {
+      "epoch": 0.11,
+      "learning_rate": 4.820061543080157e-05,
+      "loss": 0.9445,
+      "step": 970
+    },
+    {
+      "epoch": 0.11,
+      "learning_rate": 4.819873911738217e-05,
+      "loss": 0.7888,
+      "step": 971
+    },
+    {
+      "epoch": 0.11,
+      "learning_rate": 4.819686280396278e-05,
+      "loss": 0.7551,
+      "step": 972
+    },
+    {
+      "epoch": 0.11,
+      "learning_rate": 4.819498649054338e-05,
+      "loss": 1.1008,
+      "step": 973
+    },
+    {
+      "epoch": 0.11,
+      "learning_rate": 4.819311017712399e-05,
+      "loss": 1.0997,
+      "step": 974
+    },
+    {
+      "epoch": 0.11,
+      "learning_rate": 4.819123386370459e-05,
+      "loss": 0.759,
+      "step": 975
+    },
+    {
+      "epoch": 0.11,
+      "learning_rate": 4.8189357550285205e-05,
+      "loss": 0.82,
+      "step": 976
+    },
+    {
+      "epoch": 0.11,
+      "learning_rate": 4.8187481236865805e-05,
+      "loss": 0.7447,
+      "step": 977
+    },
+    {
+      "epoch": 0.11,
+      "learning_rate": 4.818560492344642e-05,
+      "loss": 0.8375,
+      "step": 978
+    },
+    {
+      "epoch": 0.11,
+      "learning_rate": 4.8183728610027024e-05,
+      "loss": 0.8366,
+      "step": 979
+    },
+    {
+      "epoch": 0.11,
+      "learning_rate": 4.818185229660763e-05,
+      "loss": 0.7735,
+      "step": 980
+    },
+    {
+      "epoch": 0.11,
+      "learning_rate": 4.8179975983188236e-05,
+      "loss": 0.9932,
+      "step": 981
+    },
+    {
+      "epoch": 0.11,
+      "learning_rate": 4.817809966976884e-05,
+      "loss": 1.0104,
+      "step": 982
+    },
+    {
+      "epoch": 0.11,
+      "learning_rate": 4.817622335634945e-05,
+      "loss": 0.9309,
+      "step": 983
+    },
+    {
+      "epoch": 0.11,
+      "learning_rate": 4.8174347042930055e-05,
+      "loss": 0.8173,
+      "step": 984
+    },
+    {
+      "epoch": 0.11,
+      "learning_rate": 4.817247072951066e-05,
+      "loss": 0.8058,
+      "step": 985
+    },
+    {
+      "epoch": 0.11,
+      "learning_rate": 4.817059441609126e-05,
+      "loss": 0.6469,
+      "step": 986
+    },
+    {
+      "epoch": 0.11,
+      "learning_rate": 4.816871810267187e-05,
+      "loss": 0.9007,
+      "step": 987
+    },
+    {
+      "epoch": 0.11,
+      "learning_rate": 4.816684178925248e-05,
+      "loss": 0.7098,
+      "step": 988
+    },
+    {
+      "epoch": 0.11,
+      "learning_rate": 4.8164965475833086e-05,
+      "loss": 0.8838,
+      "step": 989
+    },
+    {
+      "epoch": 0.11,
+      "learning_rate": 4.816308916241369e-05,
+      "loss": 0.8293,
+      "step": 990
+    },
+    {
+      "epoch": 0.11,
+      "learning_rate": 4.81612128489943e-05,
+      "loss": 0.9018,
+      "step": 991
+    },
+    {
+      "epoch": 0.11,
+      "learning_rate": 4.8159336535574904e-05,
+      "loss": 0.8161,
+      "step": 992
+    },
+    {
+      "epoch": 0.11,
+      "learning_rate": 4.815746022215551e-05,
+      "loss": 0.8817,
+      "step": 993
+    },
+    {
+      "epoch": 0.11,
+      "learning_rate": 4.8155583908736116e-05,
+      "loss": 0.8087,
+      "step": 994
+    },
+    {
+      "epoch": 0.11,
+      "learning_rate": 4.815370759531672e-05,
+      "loss": 1.0042,
+      "step": 995
+    },
+    {
+      "epoch": 0.11,
+      "learning_rate": 4.815183128189733e-05,
+      "loss": 1.0677,
+      "step": 996
+    },
+    {
+      "epoch": 0.11,
+      "learning_rate": 4.814995496847794e-05,
+      "loss": 0.7129,
+      "step": 997
+    },
+    {
+      "epoch": 0.11,
+      "learning_rate": 4.814807865505854e-05,
+      "loss": 0.8895,
+      "step": 998
+    },
+    {
+      "epoch": 0.11,
+      "learning_rate": 4.8146202341639154e-05,
+      "loss": 0.7782,
+      "step": 999
+    },
+    {
+      "epoch": 0.11,
+      "learning_rate": 4.814432602821975e-05,
+      "loss": 0.6287,
+      "step": 1000
+    },
+    {
+      "epoch": 0.11,
+      "eval_loss": 0.8395036458969116,
+      "eval_runtime": 23.3244,
+      "eval_samples_per_second": 43.86,
+      "eval_steps_per_second": 21.951,
+      "step": 1000
+    },
+    {
+      "epoch": 0.11,
+      "learning_rate": 4.8142449714800366e-05,
+      "loss": 0.7234,
+      "step": 1001
+    },
+    {
+      "epoch": 0.11,
+      "learning_rate": 4.8140573401380966e-05,
+      "loss": 0.7681,
+      "step": 1002
+    },
+    {
+      "epoch": 0.11,
+      "learning_rate": 4.813869708796158e-05,
+      "loss": 0.8038,
+      "step": 1003
+    },
+    {
+      "epoch": 0.11,
+      "learning_rate": 4.813682077454218e-05,
+      "loss": 1.1349,
+      "step": 1004
+    },
+    {
+      "epoch": 0.11,
+      "learning_rate": 4.813494446112279e-05,
+      "loss": 0.8757,
+      "step": 1005
+    },
+    {
+      "epoch": 0.11,
+      "learning_rate": 4.81330681477034e-05,
+      "loss": 1.1285,
+      "step": 1006
+    },
+    {
+      "epoch": 0.11,
+      "learning_rate": 4.8131191834284e-05,
+      "loss": 1.0306,
+      "step": 1007
+    },
+    {
+      "epoch": 0.11,
+      "learning_rate": 4.812931552086461e-05,
+      "loss": 0.8722,
+      "step": 1008
+    },
+    {
+      "epoch": 0.11,
+      "learning_rate": 4.8127439207445216e-05,
+      "loss": 0.8727,
+      "step": 1009
+    },
+    {
+      "epoch": 0.11,
+      "learning_rate": 4.812556289402582e-05,
+      "loss": 0.9927,
+      "step": 1010
+    },
+    {
+      "epoch": 0.11,
+      "learning_rate": 4.812368658060643e-05,
+      "loss": 0.7226,
+      "step": 1011
+    },
+    {
+      "epoch": 0.11,
+      "learning_rate": 4.8121810267187034e-05,
+      "loss": 0.6983,
+      "step": 1012
+    },
+    {
+      "epoch": 0.11,
+      "learning_rate": 4.811993395376764e-05,
+      "loss": 1.093,
+      "step": 1013
+    },
+    {
+      "epoch": 0.11,
+      "learning_rate": 4.8118057640348246e-05,
+      "loss": 1.4359,
+      "step": 1014
+    },
+    {
+      "epoch": 0.11,
+      "learning_rate": 4.811618132692885e-05,
+      "loss": 0.7962,
+      "step": 1015
+    },
+    {
+      "epoch": 0.11,
+      "learning_rate": 4.811430501350946e-05,
+      "loss": 0.7805,
+      "step": 1016
+    },
+    {
+      "epoch": 0.11,
+      "learning_rate": 4.8112428700090065e-05,
+      "loss": 0.6476,
+      "step": 1017
+    },
+    {
+      "epoch": 0.11,
+      "learning_rate": 4.811055238667067e-05,
+      "loss": 0.7873,
+      "step": 1018
+    },
+    {
+      "epoch": 0.11,
+      "learning_rate": 4.810867607325128e-05,
+      "loss": 0.8419,
+      "step": 1019
+    },
+    {
+      "epoch": 0.11,
+      "learning_rate": 4.810679975983188e-05,
+      "loss": 0.64,
+      "step": 1020
+    },
+    {
+      "epoch": 0.11,
+      "learning_rate": 4.810492344641249e-05,
+      "loss": 1.0174,
+      "step": 1021
+    },
+    {
+      "epoch": 0.12,
+      "learning_rate": 4.8103047132993096e-05,
+      "loss": 0.8979,
+      "step": 1022
+    },
+    {
+      "epoch": 0.12,
+      "learning_rate": 4.81011708195737e-05,
+      "loss": 0.6844,
+      "step": 1023
+    },
+    {
+      "epoch": 0.12,
+      "learning_rate": 4.809929450615431e-05,
+      "loss": 0.9941,
+      "step": 1024
+    },
+    {
+      "epoch": 0.12,
+      "learning_rate": 4.8097418192734914e-05,
+      "loss": 0.7688,
+      "step": 1025
+    },
+    {
+      "epoch": 0.12,
+      "learning_rate": 4.809554187931553e-05,
+      "loss": 0.8199,
+      "step": 1026
+    },
+    {
+      "epoch": 0.12,
+      "learning_rate": 4.8093665565896126e-05,
+      "loss": 0.8422,
+      "step": 1027
+    },
+    {
+      "epoch": 0.12,
+      "learning_rate": 4.809178925247674e-05,
+      "loss": 0.7671,
+      "step": 1028
+    },
+    {
+      "epoch": 0.12,
+      "learning_rate": 4.808991293905734e-05,
+      "loss": 0.7163,
+      "step": 1029
+    },
+    {
+      "epoch": 0.12,
+      "learning_rate": 4.808803662563795e-05,
+      "loss": 0.5887,
+      "step": 1030
+    },
+    {
+      "epoch": 0.12,
+      "learning_rate": 4.808616031221855e-05,
+      "loss": 0.5771,
+      "step": 1031
+    },
+    {
+      "epoch": 0.12,
+      "learning_rate": 4.8084283998799164e-05,
+      "loss": 0.8844,
+      "step": 1032
+    },
+    {
+      "epoch": 0.12,
+      "learning_rate": 4.8082407685379763e-05,
+      "loss": 0.8933,
+      "step": 1033
+    },
+    {
+      "epoch": 0.12,
+      "learning_rate": 4.8080531371960376e-05,
+      "loss": 0.9045,
+      "step": 1034
+    },
+    {
+      "epoch": 0.12,
+      "learning_rate": 4.807865505854098e-05,
+      "loss": 1.0436,
+      "step": 1035
+    },
+    {
+      "epoch": 0.12,
+      "learning_rate": 4.807677874512159e-05,
+      "loss": 0.6091,
+      "step": 1036
+    },
+    {
+      "epoch": 0.12,
+      "learning_rate": 4.8074902431702195e-05,
+      "loss": 0.8878,
+      "step": 1037
+    },
+    {
+      "epoch": 0.12,
+      "learning_rate": 4.80730261182828e-05,
+      "loss": 0.7794,
+      "step": 1038
+    },
+    {
+      "epoch": 0.12,
+      "learning_rate": 4.807114980486341e-05,
+      "loss": 0.8561,
+      "step": 1039
+    },
+    {
+      "epoch": 0.12,
+      "learning_rate": 4.806927349144401e-05,
+      "loss": 1.0643,
+      "step": 1040
+    },
+    {
+      "epoch": 0.12,
+      "learning_rate": 4.806739717802462e-05,
+      "loss": 0.8065,
+      "step": 1041
+    },
+    {
+      "epoch": 0.12,
+      "learning_rate": 4.8065520864605226e-05,
+      "loss": 0.9921,
+      "step": 1042
+    },
+    {
+      "epoch": 0.12,
+      "learning_rate": 4.806364455118583e-05,
+      "loss": 0.8858,
+      "step": 1043
+    },
+    {
+      "epoch": 0.12,
+      "learning_rate": 4.806176823776644e-05,
+      "loss": 0.6771,
+      "step": 1044
+    },
+    {
+      "epoch": 0.12,
+      "learning_rate": 4.8059891924347044e-05,
+      "loss": 0.646,
+      "step": 1045
+    },
+    {
+      "epoch": 0.12,
+      "learning_rate": 4.805801561092765e-05,
+      "loss": 0.962,
+      "step": 1046
+    },
+    {
+      "epoch": 0.12,
+      "learning_rate": 4.8056139297508256e-05,
+      "loss": 0.7749,
+      "step": 1047
+    },
+    {
+      "epoch": 0.12,
+      "learning_rate": 4.805426298408886e-05,
+      "loss": 0.7601,
+      "step": 1048
+    },
+    {
+      "epoch": 0.12,
+      "learning_rate": 4.805238667066947e-05,
+      "loss": 0.8417,
+      "step": 1049
+    },
+    {
+      "epoch": 0.12,
+      "learning_rate": 4.8050510357250075e-05,
+      "loss": 1.1089,
+      "step": 1050
+    },
+    {
+      "epoch": 0.12,
+      "learning_rate": 4.804863404383068e-05,
+      "loss": 0.7967,
+      "step": 1051
+    },
+    {
+      "epoch": 0.12,
+      "learning_rate": 4.804675773041129e-05,
+      "loss": 0.7867,
+      "step": 1052
+    },
+    {
+      "epoch": 0.12,
+      "learning_rate": 4.80448814169919e-05,
+      "loss": 0.8045,
+      "step": 1053
+    },
+    {
+      "epoch": 0.12,
+      "learning_rate": 4.80430051035725e-05,
+      "loss": 0.6868,
+      "step": 1054
+    },
+    {
+      "epoch": 0.12,
+      "learning_rate": 4.804112879015311e-05,
+      "loss": 1.0149,
+      "step": 1055
+    },
+    {
+      "epoch": 0.12,
+      "learning_rate": 4.803925247673371e-05,
+      "loss": 1.082,
+      "step": 1056
+    },
+    {
+      "epoch": 0.12,
+      "learning_rate": 4.8037376163314325e-05,
+      "loss": 0.7378,
+      "step": 1057
+    },
+    {
+      "epoch": 0.12,
+      "learning_rate": 4.8035499849894924e-05,
+      "loss": 1.0112,
+      "step": 1058
+    },
+    {
+      "epoch": 0.12,
+      "learning_rate": 4.803362353647554e-05,
+      "loss": 0.8208,
+      "step": 1059
+    },
+    {
+      "epoch": 0.12,
+      "learning_rate": 4.8031747223056137e-05,
+      "loss": 0.6616,
+      "step": 1060
+    },
+    {
+      "epoch": 0.12,
+      "learning_rate": 4.802987090963675e-05,
+      "loss": 0.9016,
+      "step": 1061
+    },
+    {
+      "epoch": 0.12,
+      "learning_rate": 4.8027994596217356e-05,
+      "loss": 1.0897,
+      "step": 1062
+    },
+    {
+      "epoch": 0.12,
+      "learning_rate": 4.802611828279796e-05,
+      "loss": 1.0119,
+      "step": 1063
+    },
+    {
+      "epoch": 0.12,
+      "learning_rate": 4.802424196937857e-05,
+      "loss": 1.1127,
+      "step": 1064
+    },
+    {
+      "epoch": 0.12,
+      "learning_rate": 4.8022365655959174e-05,
+      "loss": 0.8491,
+      "step": 1065
+    },
+    {
+      "epoch": 0.12,
+      "learning_rate": 4.802048934253978e-05,
+      "loss": 0.9552,
+      "step": 1066
+    },
+    {
+      "epoch": 0.12,
+      "learning_rate": 4.8018613029120386e-05,
+      "loss": 0.8792,
+      "step": 1067
+    },
+    {
+      "epoch": 0.12,
+      "learning_rate": 4.801673671570099e-05,
+      "loss": 0.7711,
+      "step": 1068
+    },
+    {
+      "epoch": 0.12,
+      "learning_rate": 4.80148604022816e-05,
+      "loss": 0.5769,
+      "step": 1069
+    },
+    {
+      "epoch": 0.12,
+      "learning_rate": 4.8012984088862205e-05,
+      "loss": 1.003,
+      "step": 1070
+    },
+    {
+      "epoch": 0.12,
+      "learning_rate": 4.801110777544281e-05,
+      "loss": 0.7633,
+      "step": 1071
+    },
+    {
+      "epoch": 0.12,
+      "learning_rate": 4.800923146202342e-05,
+      "loss": 1.3069,
+      "step": 1072
+    },
+    {
+      "epoch": 0.12,
+      "learning_rate": 4.800735514860403e-05,
+      "loss": 0.6422,
+      "step": 1073
+    },
+    {
+      "epoch": 0.12,
+      "learning_rate": 4.800547883518463e-05,
+      "loss": 0.7508,
+      "step": 1074
+    },
+    {
+      "epoch": 0.12,
+      "learning_rate": 4.800360252176524e-05,
+      "loss": 0.7686,
+      "step": 1075
+    },
+    {
+      "epoch": 0.12,
+      "learning_rate": 4.800172620834584e-05,
+      "loss": 0.6608,
+      "step": 1076
+    },
+    {
+      "epoch": 0.12,
+      "learning_rate": 4.7999849894926455e-05,
+      "loss": 0.7277,
+      "step": 1077
+    },
+    {
+      "epoch": 0.12,
+      "learning_rate": 4.7997973581507054e-05,
+      "loss": 0.7048,
+      "step": 1078
+    },
+    {
+      "epoch": 0.12,
+      "learning_rate": 4.799609726808767e-05,
+      "loss": 0.8561,
+      "step": 1079
+    },
+    {
+      "epoch": 0.12,
+      "learning_rate": 4.7994220954668267e-05,
+      "loss": 0.5615,
+      "step": 1080
+    },
+    {
+      "epoch": 0.12,
+      "learning_rate": 4.799234464124888e-05,
+      "loss": 1.0399,
+      "step": 1081
+    },
+    {
+      "epoch": 0.12,
+      "learning_rate": 4.7990468327829486e-05,
+      "loss": 0.8226,
+      "step": 1082
+    },
+    {
+      "epoch": 0.12,
+      "learning_rate": 4.798859201441009e-05,
+      "loss": 1.1189,
+      "step": 1083
+    },
+    {
+      "epoch": 0.12,
+      "learning_rate": 4.79867157009907e-05,
+      "loss": 0.8086,
+      "step": 1084
+    },
+    {
+      "epoch": 0.12,
+      "learning_rate": 4.7984839387571304e-05,
+      "loss": 0.8765,
+      "step": 1085
+    },
+    {
+      "epoch": 0.12,
+      "learning_rate": 4.798296307415191e-05,
+      "loss": 0.8803,
+      "step": 1086
+    },
+    {
+      "epoch": 0.12,
+      "learning_rate": 4.7981086760732516e-05,
+      "loss": 0.7652,
+      "step": 1087
+    },
+    {
+      "epoch": 0.12,
+      "learning_rate": 4.797921044731312e-05,
+      "loss": 0.9554,
+      "step": 1088
+    },
+    {
+      "epoch": 0.12,
+      "learning_rate": 4.797733413389373e-05,
+      "loss": 0.628,
+      "step": 1089
+    },
+    {
+      "epoch": 0.12,
+      "learning_rate": 4.7975457820474335e-05,
+      "loss": 0.9607,
+      "step": 1090
+    },
+    {
+      "epoch": 0.12,
+      "learning_rate": 4.797358150705494e-05,
+      "loss": 0.7689,
+      "step": 1091
+    },
+    {
+      "epoch": 0.12,
+      "learning_rate": 4.797170519363555e-05,
+      "loss": 0.6322,
+      "step": 1092
+    },
+    {
+      "epoch": 0.12,
+      "learning_rate": 4.7969828880216153e-05,
+      "loss": 0.8569,
+      "step": 1093
+    },
+    {
+      "epoch": 0.12,
+      "learning_rate": 4.796795256679676e-05,
+      "loss": 0.6443,
+      "step": 1094
+    },
+    {
+      "epoch": 0.12,
+      "learning_rate": 4.7966076253377366e-05,
+      "loss": 0.7308,
+      "step": 1095
+    },
+    {
+      "epoch": 0.12,
+      "learning_rate": 4.796419993995797e-05,
+      "loss": 0.8549,
+      "step": 1096
+    },
+    {
+      "epoch": 0.12,
+      "learning_rate": 4.796232362653858e-05,
+      "loss": 0.9503,
+      "step": 1097
+    },
+    {
+      "epoch": 0.12,
+      "learning_rate": 4.7960447313119184e-05,
+      "loss": 0.6545,
+      "step": 1098
+    },
+    {
+      "epoch": 0.12,
+      "learning_rate": 4.795857099969979e-05,
+      "loss": 0.9333,
+      "step": 1099
+    },
+    {
+      "epoch": 0.12,
+      "learning_rate": 4.79566946862804e-05,
+      "loss": 0.7557,
+      "step": 1100
+    },
+    {
+      "epoch": 0.12,
+      "learning_rate": 4.7954818372861e-05,
+      "loss": 0.8639,
+      "step": 1101
+    },
+    {
+      "epoch": 0.12,
+      "learning_rate": 4.7952942059441616e-05,
+      "loss": 0.9165,
+      "step": 1102
+    },
+    {
+      "epoch": 0.12,
+      "learning_rate": 4.7951065746022215e-05,
+      "loss": 0.818,
+      "step": 1103
+    },
+    {
+      "epoch": 0.12,
+      "learning_rate": 4.794918943260283e-05,
+      "loss": 0.9549,
+      "step": 1104
+    },
+    {
+      "epoch": 0.12,
+      "learning_rate": 4.794731311918343e-05,
+      "loss": 1.1044,
+      "step": 1105
+    },
+    {
+      "epoch": 0.12,
+      "learning_rate": 4.794543680576404e-05,
+      "loss": 0.951,
+      "step": 1106
+    },
+    {
+      "epoch": 0.12,
+      "learning_rate": 4.794356049234464e-05,
+      "loss": 0.7397,
+      "step": 1107
+    },
+    {
+      "epoch": 0.12,
+      "learning_rate": 4.794168417892525e-05,
+      "loss": 0.9579,
+      "step": 1108
+    },
+    {
+      "epoch": 0.12,
+      "learning_rate": 4.793980786550586e-05,
+      "loss": 0.7469,
+      "step": 1109
+    },
+    {
+      "epoch": 0.12,
+      "learning_rate": 4.7937931552086465e-05,
+      "loss": 0.7588,
+      "step": 1110
+    },
+    {
+      "epoch": 0.13,
+      "learning_rate": 4.793605523866707e-05,
+      "loss": 0.7096,
+      "step": 1111
+    },
+    {
+      "epoch": 0.13,
+      "learning_rate": 4.793417892524768e-05,
+      "loss": 0.6289,
+      "step": 1112
+    },
+    {
+      "epoch": 0.13,
+      "learning_rate": 4.7932302611828283e-05,
+      "loss": 0.9324,
+      "step": 1113
+    },
+    {
+      "epoch": 0.13,
+      "learning_rate": 4.793042629840889e-05,
+      "loss": 0.6917,
+      "step": 1114
+    },
+    {
+      "epoch": 0.13,
+      "learning_rate": 4.7928549984989496e-05,
+      "loss": 1.3544,
+      "step": 1115
+    },
+    {
+      "epoch": 0.13,
+      "learning_rate": 4.79266736715701e-05,
+      "loss": 0.9192,
+      "step": 1116
+    },
+    {
+      "epoch": 0.13,
+      "learning_rate": 4.792479735815071e-05,
+      "loss": 0.8056,
+      "step": 1117
+    },
+    {
+      "epoch": 0.13,
+      "learning_rate": 4.7922921044731314e-05,
+      "loss": 0.6958,
+      "step": 1118
+    },
+    {
+      "epoch": 0.13,
+      "learning_rate": 4.792104473131192e-05,
+      "loss": 0.7443,
+      "step": 1119
+    },
+    {
+      "epoch": 0.13,
+      "learning_rate": 4.7919168417892527e-05,
+      "loss": 0.8597,
+      "step": 1120
+    },
+    {
+      "epoch": 0.13,
+      "learning_rate": 4.791729210447313e-05,
+      "loss": 0.8958,
+      "step": 1121
+    },
+    {
+      "epoch": 0.13,
+      "learning_rate": 4.791541579105374e-05,
+      "loss": 0.9189,
+      "step": 1122
+    },
+    {
+      "epoch": 0.13,
+      "learning_rate": 4.7913539477634345e-05,
+      "loss": 0.992,
+      "step": 1123
+    },
+    {
+      "epoch": 0.13,
+      "learning_rate": 4.791166316421495e-05,
+      "loss": 0.8947,
+      "step": 1124
+    },
+    {
+      "epoch": 0.13,
+      "learning_rate": 4.790978685079556e-05,
+      "loss": 0.9527,
+      "step": 1125
+    },
+    {
+      "epoch": 0.13,
+      "eval_loss": 0.8386918902397156,
+      "eval_runtime": 23.2801,
+      "eval_samples_per_second": 43.943,
+      "eval_steps_per_second": 21.993,
+      "step": 1125
+    },
+    {
+      "epoch": 0.13,
+      "learning_rate": 4.7907910537376164e-05,
+      "loss": 0.9707,
+      "step": 1126
+    },
+    {
+      "epoch": 0.13,
+      "learning_rate": 4.790603422395677e-05,
+      "loss": 0.7657,
+      "step": 1127
+    },
+    {
+      "epoch": 0.13,
+      "learning_rate": 4.7904157910537376e-05,
+      "loss": 0.858,
+      "step": 1128
+    },
+    {
+      "epoch": 0.13,
+      "learning_rate": 4.790228159711799e-05,
+      "loss": 0.7423,
+      "step": 1129
+    },
+    {
+      "epoch": 0.13,
+      "learning_rate": 4.790040528369859e-05,
+      "loss": 0.7857,
+      "step": 1130
+    },
+    {
+      "epoch": 0.13,
+      "learning_rate": 4.78985289702792e-05,
+      "loss": 0.8773,
+      "step": 1131
+    },
+    {
+      "epoch": 0.13,
+      "learning_rate": 4.78966526568598e-05,
+      "loss": 0.8446,
+      "step": 1132
+    },
+    {
+      "epoch": 0.13,
+      "learning_rate": 4.7894776343440413e-05,
+      "loss": 0.8454,
+      "step": 1133
+    },
+    {
+      "epoch": 0.13,
+      "learning_rate": 4.789290003002101e-05,
+      "loss": 0.7488,
+      "step": 1134
+    },
+    {
+      "epoch": 0.13,
+      "learning_rate": 4.7891023716601626e-05,
+      "loss": 0.6819,
+      "step": 1135
+    },
+    {
+      "epoch": 0.13,
+      "learning_rate": 4.7889147403182225e-05,
+      "loss": 0.7466,
+      "step": 1136
+    },
+    {
+      "epoch": 0.13,
+      "learning_rate": 4.788727108976284e-05,
+      "loss": 0.8829,
+      "step": 1137
+    },
+    {
+      "epoch": 0.13,
+      "learning_rate": 4.7885394776343444e-05,
+      "loss": 0.8278,
+      "step": 1138
+    },
+    {
+      "epoch": 0.13,
+      "learning_rate": 4.788351846292405e-05,
+      "loss": 0.9994,
+      "step": 1139
+    },
+    {
+      "epoch": 0.13,
+      "learning_rate": 4.7881642149504657e-05,
+      "loss": 0.8189,
+      "step": 1140
+    },
+    {
+      "epoch": 0.13,
+      "learning_rate": 4.787976583608526e-05,
+      "loss": 0.8508,
+      "step": 1141
+    },
+    {
+      "epoch": 0.13,
+      "learning_rate": 4.787788952266587e-05,
+      "loss": 0.8698,
+      "step": 1142
+    },
+    {
+      "epoch": 0.13,
+      "learning_rate": 4.7876013209246475e-05,
+      "loss": 0.8778,
+      "step": 1143
+    },
+    {
+      "epoch": 0.13,
+      "learning_rate": 4.787413689582708e-05,
+      "loss": 0.9228,
+      "step": 1144
+    },
+    {
+      "epoch": 0.13,
+      "learning_rate": 4.787226058240769e-05,
+      "loss": 0.6808,
+      "step": 1145
+    },
+    {
+      "epoch": 0.13,
+      "learning_rate": 4.7870384268988294e-05,
+      "loss": 0.7235,
+      "step": 1146
+    },
+    {
+      "epoch": 0.13,
+      "learning_rate": 4.7868507955568906e-05,
+      "loss": 0.8906,
+      "step": 1147
+    },
+    {
+      "epoch": 0.13,
+      "learning_rate": 4.7866631642149506e-05,
+      "loss": 0.8155,
+      "step": 1148
+    },
+    {
+      "epoch": 0.13,
+      "learning_rate": 4.786475532873011e-05,
+      "loss": 0.9334,
+      "step": 1149
+    },
+    {
+      "epoch": 0.13,
+      "learning_rate": 4.786287901531072e-05,
+      "loss": 0.8421,
+      "step": 1150
+    },
+    {
+      "epoch": 0.13,
+      "learning_rate": 4.7861002701891324e-05,
+      "loss": 0.7268,
+      "step": 1151
+    },
+    {
+      "epoch": 0.13,
+      "learning_rate": 4.785912638847193e-05,
+      "loss": 0.9638,
+      "step": 1152
+    },
+    {
+      "epoch": 0.13,
+      "learning_rate": 4.785725007505254e-05,
+      "loss": 0.9943,
+      "step": 1153
+    },
+    {
+      "epoch": 0.13,
+      "learning_rate": 4.785537376163314e-05,
+      "loss": 0.7114,
+      "step": 1154
+    },
+    {
+      "epoch": 0.13,
+      "learning_rate": 4.785349744821375e-05,
+      "loss": 0.848,
+      "step": 1155
+    },
+    {
+      "epoch": 0.13,
+      "learning_rate": 4.785162113479436e-05,
+      "loss": 0.7313,
+      "step": 1156
+    },
+    {
+      "epoch": 0.13,
+      "learning_rate": 4.784974482137496e-05,
+      "loss": 0.8828,
+      "step": 1157
+    },
+    {
+      "epoch": 0.13,
+      "learning_rate": 4.7847868507955574e-05,
+      "loss": 0.7315,
+      "step": 1158
+    },
+    {
+      "epoch": 0.13,
+      "learning_rate": 4.7845992194536174e-05,
+      "loss": 0.8636,
+      "step": 1159
+    },
+    {
+      "epoch": 0.13,
+      "learning_rate": 4.7844115881116787e-05,
+      "loss": 1.0256,
+      "step": 1160
+    },
+    {
+      "epoch": 0.13,
+      "learning_rate": 4.7842239567697386e-05,
+      "loss": 0.8942,
+      "step": 1161
+    },
+    {
+      "epoch": 0.13,
+      "learning_rate": 4.7840363254278e-05,
+      "loss": 0.7788,
+      "step": 1162
+    },
+    {
+      "epoch": 0.13,
+      "learning_rate": 4.78384869408586e-05,
+      "loss": 0.6433,
+      "step": 1163
+    },
+    {
+      "epoch": 0.13,
+      "learning_rate": 4.783661062743921e-05,
+      "loss": 0.8933,
+      "step": 1164
+    },
+    {
+      "epoch": 0.13,
+      "learning_rate": 4.783473431401981e-05,
+      "loss": 0.9185,
+      "step": 1165
+    },
+    {
+      "epoch": 0.13,
+      "learning_rate": 4.7832858000600424e-05,
+      "loss": 0.6516,
+      "step": 1166
+    },
+    {
+      "epoch": 0.13,
+      "learning_rate": 4.783098168718103e-05,
+      "loss": 0.819,
+      "step": 1167
+    },
+    {
+      "epoch": 0.13,
+      "learning_rate": 4.7829105373761636e-05,
+      "loss": 0.9679,
+      "step": 1168
+    },
+    {
+      "epoch": 0.13,
+      "learning_rate": 4.782722906034224e-05,
+      "loss": 0.7469,
+      "step": 1169
+    },
+    {
+      "epoch": 0.13,
+      "learning_rate": 4.782535274692285e-05,
+      "loss": 0.9467,
+      "step": 1170
+    },
+    {
+      "epoch": 0.13,
+      "learning_rate": 4.7823476433503454e-05,
+      "loss": 0.7951,
+      "step": 1171
+    },
+    {
+      "epoch": 0.13,
+      "learning_rate": 4.782160012008406e-05,
+      "loss": 0.7277,
+      "step": 1172
+    },
+    {
+      "epoch": 0.13,
+      "learning_rate": 4.781972380666467e-05,
+      "loss": 0.8946,
+      "step": 1173
+    },
+    {
+      "epoch": 0.13,
+      "learning_rate": 4.781784749324527e-05,
+      "loss": 0.6313,
+      "step": 1174
+    },
+    {
+      "epoch": 0.13,
+      "learning_rate": 4.781597117982588e-05,
+      "loss": 0.9029,
+      "step": 1175
+    },
+    {
+      "epoch": 0.13,
+      "learning_rate": 4.781409486640649e-05,
+      "loss": 0.5346,
+      "step": 1176
+    },
+    {
+      "epoch": 0.13,
+      "learning_rate": 4.781221855298709e-05,
+      "loss": 0.982,
+      "step": 1177
+    },
+    {
+      "epoch": 0.13,
+      "learning_rate": 4.7810342239567704e-05,
+      "loss": 0.8371,
+      "step": 1178
+    },
+    {
+      "epoch": 0.13,
+      "learning_rate": 4.7808465926148304e-05,
+      "loss": 0.7892,
+      "step": 1179
+    },
+    {
+      "epoch": 0.13,
+      "learning_rate": 4.7806589612728917e-05,
+      "loss": 0.9829,
+      "step": 1180
+    },
+    {
+      "epoch": 0.13,
+      "learning_rate": 4.7804713299309516e-05,
+      "loss": 0.6128,
+      "step": 1181
+    },
+    {
+      "epoch": 0.13,
+      "learning_rate": 4.780283698589013e-05,
+      "loss": 0.8113,
+      "step": 1182
+    },
+    {
+      "epoch": 0.13,
+      "learning_rate": 4.780096067247073e-05,
+      "loss": 0.9449,
+      "step": 1183
+    },
+    {
+      "epoch": 0.13,
+      "learning_rate": 4.779908435905134e-05,
+      "loss": 0.906,
+      "step": 1184
+    },
+    {
+      "epoch": 0.13,
+      "learning_rate": 4.779720804563195e-05,
+      "loss": 0.9223,
+      "step": 1185
+    },
+    {
+      "epoch": 0.13,
+      "learning_rate": 4.7795331732212554e-05,
+      "loss": 0.7941,
+      "step": 1186
+    },
+    {
+      "epoch": 0.13,
+      "learning_rate": 4.779345541879316e-05,
+      "loss": 0.8199,
+      "step": 1187
+    },
+    {
+      "epoch": 0.13,
+      "learning_rate": 4.7791579105373766e-05,
+      "loss": 1.064,
+      "step": 1188
+    },
+    {
+      "epoch": 0.13,
+      "learning_rate": 4.778970279195437e-05,
+      "loss": 0.7282,
+      "step": 1189
+    },
+    {
+      "epoch": 0.13,
+      "learning_rate": 4.778782647853498e-05,
+      "loss": 0.8164,
+      "step": 1190
+    },
+    {
+      "epoch": 0.13,
+      "learning_rate": 4.7785950165115584e-05,
+      "loss": 0.9722,
+      "step": 1191
+    },
+    {
+      "epoch": 0.13,
+      "learning_rate": 4.778407385169619e-05,
+      "loss": 0.8081,
+      "step": 1192
+    },
+    {
+      "epoch": 0.13,
+      "learning_rate": 4.77821975382768e-05,
+      "loss": 0.6439,
+      "step": 1193
+    },
+    {
+      "epoch": 0.13,
+      "learning_rate": 4.77803212248574e-05,
+      "loss": 0.7943,
+      "step": 1194
+    },
+    {
+      "epoch": 0.13,
+      "learning_rate": 4.777844491143801e-05,
+      "loss": 0.6779,
+      "step": 1195
+    },
+    {
+      "epoch": 0.13,
+      "learning_rate": 4.7776568598018615e-05,
+      "loss": 1.0185,
+      "step": 1196
+    },
+    {
+      "epoch": 0.13,
+      "learning_rate": 4.777469228459922e-05,
+      "loss": 0.9522,
+      "step": 1197
+    },
+    {
+      "epoch": 0.13,
+      "learning_rate": 4.777281597117983e-05,
+      "loss": 0.7837,
+      "step": 1198
+    },
+    {
+      "epoch": 0.13,
+      "learning_rate": 4.7770939657760434e-05,
+      "loss": 0.9505,
+      "step": 1199
+    },
+    {
+      "epoch": 0.14,
+      "learning_rate": 4.776906334434104e-05,
+      "loss": 0.7613,
+      "step": 1200
+    },
+    {
+      "epoch": 0.14,
+      "learning_rate": 4.7767187030921646e-05,
+      "loss": 0.7953,
+      "step": 1201
+    },
+    {
+      "epoch": 0.14,
+      "learning_rate": 4.776531071750225e-05,
+      "loss": 0.7998,
+      "step": 1202
+    },
+    {
+      "epoch": 0.14,
+      "learning_rate": 4.7763434404082865e-05,
+      "loss": 0.8031,
+      "step": 1203
+    },
+    {
+      "epoch": 0.14,
+      "learning_rate": 4.7761558090663464e-05,
+      "loss": 0.8669,
+      "step": 1204
+    },
+    {
+      "epoch": 0.14,
+      "learning_rate": 4.775968177724408e-05,
+      "loss": 0.7904,
+      "step": 1205
+    },
+    {
+      "epoch": 0.14,
+      "learning_rate": 4.775780546382468e-05,
+      "loss": 1.0999,
+      "step": 1206
+    },
+    {
+      "epoch": 0.14,
+      "learning_rate": 4.775592915040529e-05,
+      "loss": 0.7108,
+      "step": 1207
+    },
+    {
+      "epoch": 0.14,
+      "learning_rate": 4.775405283698589e-05,
+      "loss": 0.7221,
+      "step": 1208
+    },
+    {
+      "epoch": 0.14,
+      "learning_rate": 4.77521765235665e-05,
+      "loss": 0.8417,
+      "step": 1209
+    },
+    {
+      "epoch": 0.14,
+      "learning_rate": 4.77503002101471e-05,
+      "loss": 0.7641,
+      "step": 1210
+    },
+    {
+      "epoch": 0.14,
+      "learning_rate": 4.7748423896727714e-05,
+      "loss": 0.8399,
+      "step": 1211
+    },
+    {
+      "epoch": 0.14,
+      "learning_rate": 4.7746547583308314e-05,
+      "loss": 0.7773,
+      "step": 1212
+    },
+    {
+      "epoch": 0.14,
+      "learning_rate": 4.774467126988893e-05,
+      "loss": 0.8358,
+      "step": 1213
+    },
+    {
+      "epoch": 0.14,
+      "learning_rate": 4.774279495646953e-05,
+      "loss": 0.6636,
+      "step": 1214
+    },
+    {
+      "epoch": 0.14,
+      "learning_rate": 4.774091864305014e-05,
+      "loss": 0.6754,
+      "step": 1215
+    },
+    {
+      "epoch": 0.14,
+      "learning_rate": 4.7739042329630745e-05,
+      "loss": 0.6962,
+      "step": 1216
+    },
+    {
+      "epoch": 0.14,
+      "learning_rate": 4.773716601621135e-05,
+      "loss": 0.7416,
+      "step": 1217
+    },
+    {
+      "epoch": 0.14,
+      "learning_rate": 4.773528970279196e-05,
+      "loss": 0.8865,
+      "step": 1218
+    },
+    {
+      "epoch": 0.14,
+      "learning_rate": 4.7733413389372564e-05,
+      "loss": 0.7519,
+      "step": 1219
+    },
+    {
+      "epoch": 0.14,
+      "learning_rate": 4.773153707595317e-05,
+      "loss": 1.1135,
+      "step": 1220
+    },
+    {
+      "epoch": 0.14,
+      "learning_rate": 4.7729660762533776e-05,
+      "loss": 0.5116,
+      "step": 1221
+    },
+    {
+      "epoch": 0.14,
+      "learning_rate": 4.772778444911438e-05,
+      "loss": 0.7506,
+      "step": 1222
+    },
+    {
+      "epoch": 0.14,
+      "learning_rate": 4.772590813569499e-05,
+      "loss": 0.8692,
+      "step": 1223
+    },
+    {
+      "epoch": 0.14,
+      "learning_rate": 4.7724031822275594e-05,
+      "loss": 0.8402,
+      "step": 1224
+    },
+    {
+      "epoch": 0.14,
+      "learning_rate": 4.77221555088562e-05,
+      "loss": 0.9531,
+      "step": 1225
+    },
+    {
+      "epoch": 0.14,
+      "learning_rate": 4.772027919543681e-05,
+      "loss": 1.0211,
+      "step": 1226
+    },
+    {
+      "epoch": 0.14,
+      "learning_rate": 4.771840288201741e-05,
+      "loss": 0.8579,
+      "step": 1227
+    },
+    {
+      "epoch": 0.14,
+      "learning_rate": 4.771652656859802e-05,
+      "loss": 1.5264,
+      "step": 1228
+    },
+    {
+      "epoch": 0.14,
+      "learning_rate": 4.7714650255178625e-05,
+      "loss": 0.7947,
+      "step": 1229
+    },
+    {
+      "epoch": 0.14,
+      "learning_rate": 4.771277394175923e-05,
+      "loss": 0.7031,
+      "step": 1230
+    },
+    {
+      "epoch": 0.14,
+      "learning_rate": 4.771089762833984e-05,
+      "loss": 0.7751,
+      "step": 1231
+    },
+    {
+      "epoch": 0.14,
+      "learning_rate": 4.770902131492045e-05,
+      "loss": 0.7359,
+      "step": 1232
+    },
+    {
+      "epoch": 0.14,
+      "learning_rate": 4.770714500150105e-05,
+      "loss": 0.8261,
+      "step": 1233
+    },
+    {
+      "epoch": 0.14,
+      "learning_rate": 4.770526868808166e-05,
+      "loss": 0.925,
+      "step": 1234
+    },
+    {
+      "epoch": 0.14,
+      "learning_rate": 4.770339237466226e-05,
+      "loss": 0.8386,
+      "step": 1235
+    },
+    {
+      "epoch": 0.14,
+      "learning_rate": 4.7701516061242875e-05,
+      "loss": 0.8481,
+      "step": 1236
+    },
+    {
+      "epoch": 0.14,
+      "learning_rate": 4.7699639747823475e-05,
+      "loss": 0.9728,
+      "step": 1237
+    },
+    {
+      "epoch": 0.14,
+      "learning_rate": 4.769776343440409e-05,
+      "loss": 1.025,
+      "step": 1238
+    },
+    {
+      "epoch": 0.14,
+      "learning_rate": 4.769588712098469e-05,
+      "loss": 0.9787,
+      "step": 1239
+    },
+    {
+      "epoch": 0.14,
+      "learning_rate": 4.76940108075653e-05,
+      "loss": 1.2448,
+      "step": 1240
+    },
+    {
+      "epoch": 0.14,
+      "learning_rate": 4.7692134494145906e-05,
+      "loss": 0.9918,
+      "step": 1241
+    },
+    {
+      "epoch": 0.14,
+      "learning_rate": 4.769025818072651e-05,
+      "loss": 0.8757,
+      "step": 1242
+    },
+    {
+      "epoch": 0.14,
+      "learning_rate": 4.768838186730712e-05,
+      "loss": 0.6142,
+      "step": 1243
+    },
+    {
+      "epoch": 0.14,
+      "learning_rate": 4.7686505553887724e-05,
+      "loss": 0.4668,
+      "step": 1244
+    },
+    {
+      "epoch": 0.14,
+      "learning_rate": 4.768462924046833e-05,
+      "loss": 0.7959,
+      "step": 1245
+    },
+    {
+      "epoch": 0.14,
+      "learning_rate": 4.768275292704894e-05,
+      "loss": 1.0293,
+      "step": 1246
+    },
+    {
+      "epoch": 0.14,
+      "learning_rate": 4.768087661362954e-05,
+      "loss": 0.9403,
+      "step": 1247
+    },
+    {
+      "epoch": 0.14,
+      "learning_rate": 4.767900030021015e-05,
+      "loss": 0.8092,
+      "step": 1248
+    },
+    {
+      "epoch": 0.14,
+      "learning_rate": 4.7677123986790755e-05,
+      "loss": 0.7033,
+      "step": 1249
+    },
+    {
+      "epoch": 0.14,
+      "learning_rate": 4.767524767337137e-05,
+      "loss": 0.8358,
+      "step": 1250
+    },
+    {
+      "epoch": 0.14,
+      "eval_loss": 0.8454775810241699,
+      "eval_runtime": 23.264,
+      "eval_samples_per_second": 43.973,
+      "eval_steps_per_second": 22.008,
+      "step": 1250
+    },
+    {
+      "epoch": 0.14,
+      "learning_rate": 4.767337135995197e-05,
+      "loss": 0.8802,
+      "step": 1251
+    },
+    {
+      "epoch": 0.14,
+      "learning_rate": 4.767149504653258e-05,
+      "loss": 1.0382,
+      "step": 1252
+    },
+    {
+      "epoch": 0.14,
+      "learning_rate": 4.766961873311318e-05,
+      "loss": 0.8793,
+      "step": 1253
+    },
+    {
+      "epoch": 0.14,
+      "learning_rate": 4.7667742419693786e-05,
+      "loss": 0.8415,
+      "step": 1254
+    },
+    {
+      "epoch": 0.14,
+      "learning_rate": 4.766586610627439e-05,
+      "loss": 0.9093,
+      "step": 1255
+    },
+    {
+      "epoch": 0.14,
+      "learning_rate": 4.7663989792855e-05,
+      "loss": 0.6973,
+      "step": 1256
+    },
+    {
+      "epoch": 0.14,
+      "learning_rate": 4.7662113479435605e-05,
+      "loss": 0.7711,
+      "step": 1257
+    },
+    {
+      "epoch": 0.14,
+      "learning_rate": 4.766023716601621e-05,
+      "loss": 0.7227,
+      "step": 1258
+    },
+    {
+      "epoch": 0.14,
+      "learning_rate": 4.7658360852596824e-05,
+      "loss": 0.7722,
+      "step": 1259
+    },
+    {
+      "epoch": 0.14,
+      "learning_rate": 4.765648453917742e-05,
+      "loss": 0.7825,
+      "step": 1260
+    },
+    {
+      "epoch": 0.14,
+      "learning_rate": 4.7654608225758036e-05,
+      "loss": 1.433,
+      "step": 1261
+    },
+    {
+      "epoch": 0.14,
+      "learning_rate": 4.7652731912338635e-05,
+      "loss": 0.8276,
+      "step": 1262
+    },
+    {
+      "epoch": 0.14,
+      "learning_rate": 4.765085559891925e-05,
+      "loss": 0.7996,
+      "step": 1263
+    },
+    {
+      "epoch": 0.14,
+      "learning_rate": 4.764897928549985e-05,
+      "loss": 0.6995,
+      "step": 1264
+    },
+    {
+      "epoch": 0.14,
+      "learning_rate": 4.764710297208046e-05,
+      "loss": 0.7156,
+      "step": 1265
+    },
+    {
+      "epoch": 0.14,
+      "learning_rate": 4.764522665866106e-05,
+      "loss": 0.6824,
+      "step": 1266
+    },
+    {
+      "epoch": 0.14,
+      "learning_rate": 4.764335034524167e-05,
+      "loss": 0.9043,
+      "step": 1267
+    },
+    {
+      "epoch": 0.14,
+      "learning_rate": 4.764147403182227e-05,
+      "loss": 0.9692,
+      "step": 1268
+    },
+    {
+      "epoch": 0.14,
+      "learning_rate": 4.7639597718402885e-05,
+      "loss": 0.9589,
+      "step": 1269
+    },
+    {
+      "epoch": 0.14,
+      "learning_rate": 4.763772140498349e-05,
+      "loss": 0.8076,
+      "step": 1270
+    },
+    {
+      "epoch": 0.14,
+      "learning_rate": 4.76358450915641e-05,
+      "loss": 0.5774,
+      "step": 1271
+    },
+    {
+      "epoch": 0.14,
+      "learning_rate": 4.7633968778144704e-05,
+      "loss": 0.6714,
+      "step": 1272
+    },
+    {
+      "epoch": 0.14,
+      "learning_rate": 4.763209246472531e-05,
+      "loss": 0.9146,
+      "step": 1273
+    },
+    {
+      "epoch": 0.14,
+      "learning_rate": 4.7630216151305916e-05,
+      "loss": 0.9334,
+      "step": 1274
+    },
+    {
+      "epoch": 0.14,
+      "learning_rate": 4.762833983788652e-05,
+      "loss": 0.7841,
+      "step": 1275
+    },
+    {
+      "epoch": 0.14,
+      "learning_rate": 4.762646352446713e-05,
+      "loss": 0.8671,
+      "step": 1276
+    },
+    {
+      "epoch": 0.14,
+      "learning_rate": 4.7624587211047735e-05,
+      "loss": 0.8705,
+      "step": 1277
+    },
+    {
+      "epoch": 0.14,
+      "learning_rate": 4.762271089762834e-05,
+      "loss": 0.8903,
+      "step": 1278
+    },
+    {
+      "epoch": 0.14,
+      "learning_rate": 4.7620834584208954e-05,
+      "loss": 0.6884,
+      "step": 1279
+    },
+    {
+      "epoch": 0.14,
+      "learning_rate": 4.761895827078955e-05,
+      "loss": 0.8855,
+      "step": 1280
+    },
+    {
+      "epoch": 0.14,
+      "learning_rate": 4.7617081957370166e-05,
+      "loss": 1.1246,
+      "step": 1281
+    },
+    {
+      "epoch": 0.14,
+      "learning_rate": 4.7615205643950765e-05,
+      "loss": 1.0012,
+      "step": 1282
+    },
+    {
+      "epoch": 0.14,
+      "learning_rate": 4.761332933053138e-05,
+      "loss": 0.6517,
+      "step": 1283
+    },
+    {
+      "epoch": 0.14,
+      "learning_rate": 4.761145301711198e-05,
+      "loss": 0.8869,
+      "step": 1284
+    },
+    {
+      "epoch": 0.14,
+      "learning_rate": 4.760957670369259e-05,
+      "loss": 0.8043,
+      "step": 1285
+    },
+    {
+      "epoch": 0.14,
+      "learning_rate": 4.760770039027319e-05,
+      "loss": 1.0325,
+      "step": 1286
+    },
+    {
+      "epoch": 0.14,
+      "learning_rate": 4.76058240768538e-05,
+      "loss": 0.8939,
+      "step": 1287
+    },
+    {
+      "epoch": 0.14,
+      "learning_rate": 4.760394776343441e-05,
+      "loss": 0.7432,
+      "step": 1288
+    },
+    {
+      "epoch": 0.15,
+      "learning_rate": 4.7602071450015015e-05,
+      "loss": 1.0562,
+      "step": 1289
+    },
+    {
+      "epoch": 0.15,
+      "learning_rate": 4.760019513659562e-05,
+      "loss": 0.7166,
+      "step": 1290
+    },
+    {
+      "epoch": 0.15,
+      "learning_rate": 4.759831882317623e-05,
+      "loss": 0.9527,
+      "step": 1291
+    },
+    {
+      "epoch": 0.15,
+      "learning_rate": 4.7596442509756834e-05,
+      "loss": 0.5522,
+      "step": 1292
+    },
+    {
+      "epoch": 0.15,
+      "learning_rate": 4.759456619633744e-05,
+      "loss": 0.8278,
+      "step": 1293
+    },
+    {
+      "epoch": 0.15,
+      "learning_rate": 4.7592689882918046e-05,
+      "loss": 0.9573,
+      "step": 1294
+    },
+    {
+      "epoch": 0.15,
+      "learning_rate": 4.759081356949865e-05,
+      "loss": 0.7878,
+      "step": 1295
+    },
+    {
+      "epoch": 0.15,
+      "learning_rate": 4.758893725607926e-05,
+      "loss": 0.5653,
+      "step": 1296
+    },
+    {
+      "epoch": 0.15,
+      "learning_rate": 4.7587060942659865e-05,
+      "loss": 0.7137,
+      "step": 1297
+    },
+    {
+      "epoch": 0.15,
+      "learning_rate": 4.758518462924047e-05,
+      "loss": 0.9022,
+      "step": 1298
+    },
+    {
+      "epoch": 0.15,
+      "learning_rate": 4.758330831582108e-05,
+      "loss": 0.9431,
+      "step": 1299
+    },
+    {
+      "epoch": 0.15,
+      "learning_rate": 4.758143200240168e-05,
+      "loss": 0.8683,
+      "step": 1300
+    },
+    {
+      "epoch": 0.15,
+      "learning_rate": 4.757955568898229e-05,
+      "loss": 0.7182,
+      "step": 1301
+    },
+    {
+      "epoch": 0.15,
+      "learning_rate": 4.7577679375562895e-05,
+      "loss": 0.8616,
+      "step": 1302
+    },
+    {
+      "epoch": 0.15,
+      "learning_rate": 4.75758030621435e-05,
+      "loss": 0.6719,
+      "step": 1303
+    },
+    {
+      "epoch": 0.15,
+      "learning_rate": 4.757392674872411e-05,
+      "loss": 0.9155,
+      "step": 1304
+    },
+    {
+      "epoch": 0.15,
+      "learning_rate": 4.7572050435304714e-05,
+      "loss": 0.8044,
+      "step": 1305
+    },
+    {
+      "epoch": 0.15,
+      "learning_rate": 4.757017412188533e-05,
+      "loss": 0.8486,
+      "step": 1306
+    },
+    {
+      "epoch": 0.15,
+      "learning_rate": 4.7568297808465926e-05,
+      "loss": 0.9509,
+      "step": 1307
+    },
+    {
+      "epoch": 0.15,
+      "learning_rate": 4.756642149504654e-05,
+      "loss": 1.0364,
+      "step": 1308
+    },
+    {
+      "epoch": 0.15,
+      "learning_rate": 4.756454518162714e-05,
+      "loss": 0.8608,
+      "step": 1309
+    },
+    {
+      "epoch": 0.15,
+      "learning_rate": 4.756266886820775e-05,
+      "loss": 0.7159,
+      "step": 1310
+    },
+    {
+      "epoch": 0.15,
+      "learning_rate": 4.756079255478835e-05,
+      "loss": 0.8629,
+      "step": 1311
+    },
+    {
+      "epoch": 0.15,
+      "learning_rate": 4.7558916241368964e-05,
+      "loss": 0.8241,
+      "step": 1312
+    },
+    {
+      "epoch": 0.15,
+      "learning_rate": 4.755703992794956e-05,
+      "loss": 0.8133,
+      "step": 1313
+    },
+    {
+      "epoch": 0.15,
+      "learning_rate": 4.7555163614530176e-05,
+      "loss": 0.8969,
+      "step": 1314
+    },
+    {
+      "epoch": 0.15,
+      "learning_rate": 4.7553287301110775e-05,
+      "loss": 0.8663,
+      "step": 1315
+    },
+    {
+      "epoch": 0.15,
+      "learning_rate": 4.755141098769139e-05,
+      "loss": 1.0138,
+      "step": 1316
+    },
+    {
+      "epoch": 0.15,
+      "learning_rate": 4.7549534674271995e-05,
+      "loss": 0.905,
+      "step": 1317
+    },
+    {
+      "epoch": 0.15,
+      "learning_rate": 4.75476583608526e-05,
+      "loss": 0.8468,
+      "step": 1318
+    },
+    {
+      "epoch": 0.15,
+      "learning_rate": 4.754578204743321e-05,
+      "loss": 0.8261,
+      "step": 1319
+    },
+    {
+      "epoch": 0.15,
+      "learning_rate": 4.754390573401381e-05,
+      "loss": 0.7843,
+      "step": 1320
+    },
+    {
+      "epoch": 0.15,
+      "learning_rate": 4.754202942059442e-05,
+      "loss": 0.8756,
+      "step": 1321
+    },
+    {
+      "epoch": 0.15,
+      "learning_rate": 4.7540153107175025e-05,
+      "loss": 0.7636,
+      "step": 1322
+    },
+    {
+      "epoch": 0.15,
+      "learning_rate": 4.753827679375563e-05,
+      "loss": 0.8869,
+      "step": 1323
+    },
+    {
+      "epoch": 0.15,
+      "learning_rate": 4.753640048033624e-05,
+      "loss": 0.8134,
+      "step": 1324
+    },
+    {
+      "epoch": 0.15,
+      "learning_rate": 4.7534524166916844e-05,
+      "loss": 0.7398,
+      "step": 1325
+    },
+    {
+      "epoch": 0.15,
+      "learning_rate": 4.753264785349745e-05,
+      "loss": 1.2354,
+      "step": 1326
+    },
+    {
+      "epoch": 0.15,
+      "learning_rate": 4.7530771540078056e-05,
+      "loss": 0.9329,
+      "step": 1327
+    },
+    {
+      "epoch": 0.15,
+      "learning_rate": 4.752889522665866e-05,
+      "loss": 0.8967,
+      "step": 1328
+    },
+    {
+      "epoch": 0.15,
+      "learning_rate": 4.752701891323927e-05,
+      "loss": 0.8282,
+      "step": 1329
+    },
+    {
+      "epoch": 0.15,
+      "learning_rate": 4.7525142599819875e-05,
+      "loss": 0.8504,
+      "step": 1330
+    },
+    {
+      "epoch": 0.15,
+      "learning_rate": 4.752326628640048e-05,
+      "loss": 1.0708,
+      "step": 1331
+    },
+    {
+      "epoch": 0.15,
+      "learning_rate": 4.752138997298109e-05,
+      "loss": 0.9397,
+      "step": 1332
+    },
+    {
+      "epoch": 0.15,
+      "learning_rate": 4.751951365956169e-05,
+      "loss": 0.5868,
+      "step": 1333
+    },
+    {
+      "epoch": 0.15,
+      "learning_rate": 4.75176373461423e-05,
+      "loss": 0.763,
+      "step": 1334
+    },
+    {
+      "epoch": 0.15,
+      "learning_rate": 4.751576103272291e-05,
+      "loss": 0.9832,
+      "step": 1335
+    },
+    {
+      "epoch": 0.15,
+      "learning_rate": 4.751388471930351e-05,
+      "loss": 0.719,
+      "step": 1336
+    },
+    {
+      "epoch": 0.15,
+      "learning_rate": 4.7512008405884125e-05,
+      "loss": 1.0214,
+      "step": 1337
+    },
+    {
+      "epoch": 0.15,
+      "learning_rate": 4.7510132092464724e-05,
+      "loss": 0.6745,
+      "step": 1338
+    },
+    {
+      "epoch": 0.15,
+      "learning_rate": 4.750825577904534e-05,
+      "loss": 1.0588,
+      "step": 1339
+    },
+    {
+      "epoch": 0.15,
+      "learning_rate": 4.7506379465625936e-05,
+      "loss": 0.9462,
+      "step": 1340
+    },
+    {
+      "epoch": 0.15,
+      "learning_rate": 4.750450315220655e-05,
+      "loss": 0.8276,
+      "step": 1341
+    },
+    {
+      "epoch": 0.15,
+      "learning_rate": 4.750262683878715e-05,
+      "loss": 0.5151,
+      "step": 1342
+    },
+    {
+      "epoch": 0.15,
+      "learning_rate": 4.750075052536776e-05,
+      "loss": 0.9749,
+      "step": 1343
+    },
+    {
+      "epoch": 0.15,
+      "learning_rate": 4.749887421194837e-05,
+      "loss": 0.8208,
+      "step": 1344
+    },
+    {
+      "epoch": 0.15,
+      "learning_rate": 4.7496997898528974e-05,
+      "loss": 0.9076,
+      "step": 1345
+    },
+    {
+      "epoch": 0.15,
+      "learning_rate": 4.749512158510958e-05,
+      "loss": 0.7102,
+      "step": 1346
+    },
+    {
+      "epoch": 0.15,
+      "learning_rate": 4.7493245271690186e-05,
+      "loss": 0.9107,
+      "step": 1347
+    },
+    {
+      "epoch": 0.15,
+      "learning_rate": 4.749136895827079e-05,
+      "loss": 0.7494,
+      "step": 1348
+    },
+    {
+      "epoch": 0.15,
+      "learning_rate": 4.74894926448514e-05,
+      "loss": 0.8311,
+      "step": 1349
+    },
+    {
+      "epoch": 0.15,
+      "learning_rate": 4.7487616331432005e-05,
+      "loss": 0.9243,
+      "step": 1350
+    },
+    {
+      "epoch": 0.15,
+      "learning_rate": 4.748574001801261e-05,
+      "loss": 0.608,
+      "step": 1351
+    },
+    {
+      "epoch": 0.15,
+      "learning_rate": 4.748386370459322e-05,
+      "loss": 0.8557,
+      "step": 1352
+    },
+    {
+      "epoch": 0.15,
+      "learning_rate": 4.748198739117383e-05,
+      "loss": 0.7223,
+      "step": 1353
+    },
+    {
+      "epoch": 0.15,
+      "learning_rate": 4.748011107775443e-05,
+      "loss": 0.6461,
+      "step": 1354
+    },
+    {
+      "epoch": 0.15,
+      "learning_rate": 4.747823476433504e-05,
+      "loss": 0.878,
+      "step": 1355
+    },
+    {
+      "epoch": 0.15,
+      "learning_rate": 4.747635845091564e-05,
+      "loss": 0.7438,
+      "step": 1356
+    },
+    {
+      "epoch": 0.15,
+      "learning_rate": 4.7474482137496255e-05,
+      "loss": 0.9582,
+      "step": 1357
+    },
+    {
+      "epoch": 0.15,
+      "learning_rate": 4.7472605824076854e-05,
+      "loss": 0.9543,
+      "step": 1358
+    },
+    {
+      "epoch": 0.15,
+      "learning_rate": 4.747072951065746e-05,
+      "loss": 0.7127,
+      "step": 1359
+    },
+    {
+      "epoch": 0.15,
+      "learning_rate": 4.7468853197238066e-05,
+      "loss": 0.894,
+      "step": 1360
+    },
+    {
+      "epoch": 0.15,
+      "learning_rate": 4.746697688381867e-05,
+      "loss": 0.8096,
+      "step": 1361
+    },
+    {
+      "epoch": 0.15,
+      "learning_rate": 4.746510057039928e-05,
+      "loss": 0.8655,
+      "step": 1362
+    },
+    {
+      "epoch": 0.15,
+      "learning_rate": 4.7463224256979885e-05,
+      "loss": 0.6958,
+      "step": 1363
+    },
+    {
+      "epoch": 0.15,
+      "learning_rate": 4.74613479435605e-05,
+      "loss": 0.8198,
+      "step": 1364
+    },
+    {
+      "epoch": 0.15,
+      "learning_rate": 4.74594716301411e-05,
+      "loss": 0.5781,
+      "step": 1365
+    },
+    {
+      "epoch": 0.15,
+      "learning_rate": 4.745759531672171e-05,
+      "loss": 0.8596,
+      "step": 1366
+    },
+    {
+      "epoch": 0.15,
+      "learning_rate": 4.745571900330231e-05,
+      "loss": 0.7785,
+      "step": 1367
+    },
+    {
+      "epoch": 0.15,
+      "learning_rate": 4.745384268988292e-05,
+      "loss": 0.9106,
+      "step": 1368
+    },
+    {
+      "epoch": 0.15,
+      "learning_rate": 4.745196637646352e-05,
+      "loss": 0.8571,
+      "step": 1369
+    },
+    {
+      "epoch": 0.15,
+      "learning_rate": 4.7450090063044135e-05,
+      "loss": 0.7287,
+      "step": 1370
+    },
+    {
+      "epoch": 0.15,
+      "learning_rate": 4.7448213749624734e-05,
+      "loss": 1.0001,
+      "step": 1371
+    },
+    {
+      "epoch": 0.15,
+      "learning_rate": 4.744633743620535e-05,
+      "loss": 0.9245,
+      "step": 1372
+    },
+    {
+      "epoch": 0.15,
+      "learning_rate": 4.744446112278595e-05,
+      "loss": 0.7015,
+      "step": 1373
+    },
+    {
+      "epoch": 0.15,
+      "learning_rate": 4.744258480936656e-05,
+      "loss": 0.9246,
+      "step": 1374
+    },
+    {
+      "epoch": 0.15,
+      "learning_rate": 4.7440708495947165e-05,
+      "loss": 0.815,
+      "step": 1375
+    },
+    {
+      "epoch": 0.15,
+      "eval_loss": 0.8313619494438171,
+      "eval_runtime": 23.269,
+      "eval_samples_per_second": 43.964,
+      "eval_steps_per_second": 22.004,
+      "step": 1375
+    },
+    {
+      "epoch": 0.15,
+      "learning_rate": 4.743883218252777e-05,
+      "loss": 0.8317,
+      "step": 1376
+    },
+    {
+      "epoch": 0.15,
+      "learning_rate": 4.743695586910838e-05,
+      "loss": 0.9412,
+      "step": 1377
+    },
+    {
+      "epoch": 0.16,
+      "learning_rate": 4.7435079555688984e-05,
+      "loss": 0.775,
+      "step": 1378
+    },
+    {
+      "epoch": 0.16,
+      "learning_rate": 4.743320324226959e-05,
+      "loss": 0.6891,
+      "step": 1379
+    },
+    {
+      "epoch": 0.16,
+      "learning_rate": 4.7431326928850196e-05,
+      "loss": 0.8518,
+      "step": 1380
+    },
+    {
+      "epoch": 0.16,
+      "learning_rate": 4.74294506154308e-05,
+      "loss": 0.8542,
+      "step": 1381
+    },
+    {
+      "epoch": 0.16,
+      "learning_rate": 4.7427574302011415e-05,
+      "loss": 0.9218,
+      "step": 1382
+    },
+    {
+      "epoch": 0.16,
+      "learning_rate": 4.7425697988592015e-05,
+      "loss": 0.8477,
+      "step": 1383
+    },
+    {
+      "epoch": 0.16,
+      "learning_rate": 4.742382167517263e-05,
+      "loss": 0.8496,
+      "step": 1384
+    },
+    {
+      "epoch": 0.16,
+      "learning_rate": 4.742194536175323e-05,
+      "loss": 0.9437,
+      "step": 1385
+    },
+    {
+      "epoch": 0.16,
+      "learning_rate": 4.742006904833384e-05,
+      "loss": 0.6442,
+      "step": 1386
+    },
+    {
+      "epoch": 0.16,
+      "learning_rate": 4.741819273491444e-05,
+      "loss": 0.9634,
+      "step": 1387
+    },
+    {
+      "epoch": 0.16,
+      "learning_rate": 4.741631642149505e-05,
+      "loss": 0.7,
+      "step": 1388
+    },
+    {
+      "epoch": 0.16,
+      "learning_rate": 4.741444010807565e-05,
+      "loss": 0.8589,
+      "step": 1389
+    },
+    {
+      "epoch": 0.16,
+      "learning_rate": 4.7412563794656265e-05,
+      "loss": 0.7754,
+      "step": 1390
+    },
+    {
+      "epoch": 0.16,
+      "learning_rate": 4.741068748123687e-05,
+      "loss": 0.8202,
+      "step": 1391
+    },
+    {
+      "epoch": 0.16,
+      "learning_rate": 4.740881116781748e-05,
+      "loss": 1.1149,
+      "step": 1392
+    },
+    {
+      "epoch": 0.16,
+      "learning_rate": 4.740693485439808e-05,
+      "loss": 0.8032,
+      "step": 1393
+    },
+    {
+      "epoch": 0.16,
+      "learning_rate": 4.740505854097869e-05,
+      "loss": 0.9726,
+      "step": 1394
+    },
+    {
+      "epoch": 0.16,
+      "learning_rate": 4.7403182227559295e-05,
+      "loss": 0.876,
+      "step": 1395
+    },
+    {
+      "epoch": 0.16,
+      "learning_rate": 4.74013059141399e-05,
+      "loss": 0.9822,
+      "step": 1396
+    },
+    {
+      "epoch": 0.16,
+      "learning_rate": 4.739942960072051e-05,
+      "loss": 0.9338,
+      "step": 1397
+    },
+    {
+      "epoch": 0.16,
+      "learning_rate": 4.7397553287301114e-05,
+      "loss": 0.5906,
+      "step": 1398
+    },
+    {
+      "epoch": 0.16,
+      "learning_rate": 4.739567697388172e-05,
+      "loss": 1.0757,
+      "step": 1399
+    },
+    {
+      "epoch": 0.16,
+      "learning_rate": 4.7393800660462326e-05,
+      "loss": 0.9404,
+      "step": 1400
+    },
+    {
+      "epoch": 0.16,
+      "learning_rate": 4.739192434704293e-05,
+      "loss": 0.7623,
+      "step": 1401
+    },
+    {
+      "epoch": 0.16,
+      "learning_rate": 4.739004803362354e-05,
+      "loss": 1.0624,
+      "step": 1402
+    },
+    {
+      "epoch": 0.16,
+      "learning_rate": 4.7388171720204145e-05,
+      "loss": 0.9082,
+      "step": 1403
+    },
+    {
+      "epoch": 0.16,
+      "learning_rate": 4.738629540678475e-05,
+      "loss": 0.8471,
+      "step": 1404
+    },
+    {
+      "epoch": 0.16,
+      "learning_rate": 4.738441909336536e-05,
+      "loss": 0.9336,
+      "step": 1405
+    },
+    {
+      "epoch": 0.16,
+      "learning_rate": 4.738254277994596e-05,
+      "loss": 1.0463,
+      "step": 1406
+    },
+    {
+      "epoch": 0.16,
+      "learning_rate": 4.738066646652657e-05,
+      "loss": 0.7241,
+      "step": 1407
+    },
+    {
+      "epoch": 0.16,
+      "learning_rate": 4.7378790153107176e-05,
+      "loss": 0.989,
+      "step": 1408
+    },
+    {
+      "epoch": 0.16,
+      "learning_rate": 4.737691383968778e-05,
+      "loss": 0.8239,
+      "step": 1409
+    },
+    {
+      "epoch": 0.16,
+      "learning_rate": 4.737503752626839e-05,
+      "loss": 1.0675,
+      "step": 1410
+    },
+    {
+      "epoch": 0.16,
+      "learning_rate": 4.7373161212849e-05,
+      "loss": 0.6865,
+      "step": 1411
+    },
+    {
+      "epoch": 0.16,
+      "learning_rate": 4.73712848994296e-05,
+      "loss": 0.7318,
+      "step": 1412
+    },
+    {
+      "epoch": 0.16,
+      "learning_rate": 4.736940858601021e-05,
+      "loss": 0.8047,
+      "step": 1413
+    },
+    {
+      "epoch": 0.16,
+      "learning_rate": 4.736753227259081e-05,
+      "loss": 0.7187,
+      "step": 1414
+    },
+    {
+      "epoch": 0.16,
+      "learning_rate": 4.7365655959171425e-05,
+      "loss": 0.6413,
+      "step": 1415
+    },
+    {
+      "epoch": 0.16,
+      "learning_rate": 4.7363779645752025e-05,
+      "loss": 0.9672,
+      "step": 1416
+    },
+    {
+      "epoch": 0.16,
+      "learning_rate": 4.736190333233264e-05,
+      "loss": 0.908,
+      "step": 1417
+    },
+    {
+      "epoch": 0.16,
+      "learning_rate": 4.736002701891324e-05,
+      "loss": 0.8289,
+      "step": 1418
+    },
+    {
+      "epoch": 0.16,
+      "learning_rate": 4.735815070549385e-05,
+      "loss": 0.8563,
+      "step": 1419
+    },
+    {
+      "epoch": 0.16,
+      "learning_rate": 4.7356274392074456e-05,
+      "loss": 0.9645,
+      "step": 1420
+    },
+    {
+      "epoch": 0.16,
+      "learning_rate": 4.735439807865506e-05,
+      "loss": 0.6286,
+      "step": 1421
+    },
+    {
+      "epoch": 0.16,
+      "learning_rate": 4.735252176523567e-05,
+      "loss": 0.9367,
+      "step": 1422
+    },
+    {
+      "epoch": 0.16,
+      "learning_rate": 4.7350645451816275e-05,
+      "loss": 0.8589,
+      "step": 1423
+    },
+    {
+      "epoch": 0.16,
+      "learning_rate": 4.734876913839688e-05,
+      "loss": 0.6181,
+      "step": 1424
+    },
+    {
+      "epoch": 0.16,
+      "learning_rate": 4.734689282497749e-05,
+      "loss": 1.1623,
+      "step": 1425
+    },
+    {
+      "epoch": 0.16,
+      "learning_rate": 4.734501651155809e-05,
+      "loss": 0.6772,
+      "step": 1426
+    },
+    {
+      "epoch": 0.16,
+      "learning_rate": 4.73431401981387e-05,
+      "loss": 0.6302,
+      "step": 1427
+    },
+    {
+      "epoch": 0.16,
+      "learning_rate": 4.7341263884719306e-05,
+      "loss": 0.7673,
+      "step": 1428
+    },
+    {
+      "epoch": 0.16,
+      "learning_rate": 4.733938757129991e-05,
+      "loss": 1.0155,
+      "step": 1429
+    },
+    {
+      "epoch": 0.16,
+      "learning_rate": 4.733751125788052e-05,
+      "loss": 0.8304,
+      "step": 1430
+    },
+    {
+      "epoch": 0.16,
+      "learning_rate": 4.7335634944461124e-05,
+      "loss": 0.8735,
+      "step": 1431
+    },
+    {
+      "epoch": 0.16,
+      "learning_rate": 4.733375863104173e-05,
+      "loss": 0.9445,
+      "step": 1432
+    },
+    {
+      "epoch": 0.16,
+      "learning_rate": 4.7331882317622336e-05,
+      "loss": 0.6555,
+      "step": 1433
+    },
+    {
+      "epoch": 0.16,
+      "learning_rate": 4.733000600420294e-05,
+      "loss": 1.1519,
+      "step": 1434
+    },
+    {
+      "epoch": 0.16,
+      "learning_rate": 4.732812969078355e-05,
+      "loss": 1.0851,
+      "step": 1435
+    },
+    {
+      "epoch": 0.16,
+      "learning_rate": 4.7326253377364155e-05,
+      "loss": 0.9667,
+      "step": 1436
+    },
+    {
+      "epoch": 0.16,
+      "learning_rate": 4.732437706394476e-05,
+      "loss": 0.8334,
+      "step": 1437
+    },
+    {
+      "epoch": 0.16,
+      "learning_rate": 4.7322500750525374e-05,
+      "loss": 0.9675,
+      "step": 1438
+    },
+    {
+      "epoch": 0.16,
+      "learning_rate": 4.732062443710597e-05,
+      "loss": 0.5795,
+      "step": 1439
+    },
+    {
+      "epoch": 0.16,
+      "learning_rate": 4.7318748123686586e-05,
+      "loss": 0.7915,
+      "step": 1440
+    },
+    {
+      "epoch": 0.16,
+      "learning_rate": 4.7316871810267186e-05,
+      "loss": 1.1506,
+      "step": 1441
+    },
+    {
+      "epoch": 0.16,
+      "learning_rate": 4.73149954968478e-05,
+      "loss": 0.9093,
+      "step": 1442
+    },
+    {
+      "epoch": 0.16,
+      "learning_rate": 4.73131191834284e-05,
+      "loss": 0.9823,
+      "step": 1443
+    },
+    {
+      "epoch": 0.16,
+      "learning_rate": 4.731124287000901e-05,
+      "loss": 0.8551,
+      "step": 1444
+    },
+    {
+      "epoch": 0.16,
+      "learning_rate": 4.730936655658961e-05,
+      "loss": 0.828,
+      "step": 1445
+    },
+    {
+      "epoch": 0.16,
+      "learning_rate": 4.730749024317022e-05,
+      "loss": 0.8119,
+      "step": 1446
+    },
+    {
+      "epoch": 0.16,
+      "learning_rate": 4.730561392975083e-05,
+      "loss": 0.8782,
+      "step": 1447
+    },
+    {
+      "epoch": 0.16,
+      "learning_rate": 4.7303737616331436e-05,
+      "loss": 0.8275,
+      "step": 1448
+    },
+    {
+      "epoch": 0.16,
+      "learning_rate": 4.730186130291204e-05,
+      "loss": 0.8313,
+      "step": 1449
+    },
+    {
+      "epoch": 0.16,
+      "learning_rate": 4.729998498949265e-05,
+      "loss": 1.0695,
+      "step": 1450
+    },
+    {
+      "epoch": 0.16,
+      "learning_rate": 4.7298108676073254e-05,
+      "loss": 0.751,
+      "step": 1451
+    },
+    {
+      "epoch": 0.16,
+      "learning_rate": 4.729623236265386e-05,
+      "loss": 0.8184,
+      "step": 1452
+    },
+    {
+      "epoch": 0.16,
+      "learning_rate": 4.7294356049234466e-05,
+      "loss": 0.888,
+      "step": 1453
+    },
+    {
+      "epoch": 0.16,
+      "learning_rate": 4.729247973581507e-05,
+      "loss": 0.5361,
+      "step": 1454
+    },
+    {
+      "epoch": 0.16,
+      "learning_rate": 4.729060342239568e-05,
+      "loss": 0.8116,
+      "step": 1455
+    },
+    {
+      "epoch": 0.16,
+      "learning_rate": 4.728872710897629e-05,
+      "loss": 0.8194,
+      "step": 1456
+    },
+    {
+      "epoch": 0.16,
+      "learning_rate": 4.728685079555689e-05,
+      "loss": 0.8432,
+      "step": 1457
+    },
+    {
+      "epoch": 0.16,
+      "learning_rate": 4.7284974482137504e-05,
+      "loss": 0.9397,
+      "step": 1458
+    },
+    {
+      "epoch": 0.16,
+      "learning_rate": 4.72830981687181e-05,
+      "loss": 1.1865,
+      "step": 1459
+    },
+    {
+      "epoch": 0.16,
+      "learning_rate": 4.7281221855298716e-05,
+      "loss": 1.0974,
+      "step": 1460
+    },
+    {
+      "epoch": 0.16,
+      "learning_rate": 4.7279345541879316e-05,
+      "loss": 0.9583,
+      "step": 1461
+    },
+    {
+      "epoch": 0.16,
+      "learning_rate": 4.727746922845993e-05,
+      "loss": 0.8521,
+      "step": 1462
+    },
+    {
+      "epoch": 0.16,
+      "learning_rate": 4.727559291504053e-05,
+      "loss": 1.3013,
+      "step": 1463
+    },
+    {
+      "epoch": 0.16,
+      "learning_rate": 4.7273716601621134e-05,
+      "loss": 0.6553,
+      "step": 1464
+    },
+    {
+      "epoch": 0.16,
+      "learning_rate": 4.727184028820174e-05,
+      "loss": 0.8034,
+      "step": 1465
+    },
+    {
+      "epoch": 0.16,
+      "learning_rate": 4.7269963974782346e-05,
+      "loss": 0.9928,
+      "step": 1466
+    },
+    {
+      "epoch": 0.17,
+      "learning_rate": 4.726808766136296e-05,
+      "loss": 0.7968,
+      "step": 1467
+    },
+    {
+      "epoch": 0.17,
+      "learning_rate": 4.726621134794356e-05,
+      "loss": 0.5806,
+      "step": 1468
+    },
+    {
+      "epoch": 0.17,
+      "learning_rate": 4.726433503452417e-05,
+      "loss": 0.7799,
+      "step": 1469
+    },
+    {
+      "epoch": 0.17,
+      "learning_rate": 4.726245872110477e-05,
+      "loss": 0.8213,
+      "step": 1470
+    },
+    {
+      "epoch": 0.17,
+      "learning_rate": 4.7260582407685384e-05,
+      "loss": 0.6597,
+      "step": 1471
+    },
+    {
+      "epoch": 0.17,
+      "learning_rate": 4.7258706094265983e-05,
+      "loss": 0.7973,
+      "step": 1472
+    },
+    {
+      "epoch": 0.17,
+      "learning_rate": 4.7256829780846596e-05,
+      "loss": 0.7968,
+      "step": 1473
+    },
+    {
+      "epoch": 0.17,
+      "learning_rate": 4.7254953467427196e-05,
+      "loss": 0.7801,
+      "step": 1474
+    },
+    {
+      "epoch": 0.17,
+      "learning_rate": 4.725307715400781e-05,
+      "loss": 0.9039,
+      "step": 1475
+    },
+    {
+      "epoch": 0.17,
+      "learning_rate": 4.7251200840588415e-05,
+      "loss": 1.0253,
+      "step": 1476
+    },
+    {
+      "epoch": 0.17,
+      "learning_rate": 4.724932452716902e-05,
+      "loss": 0.9475,
+      "step": 1477
+    },
+    {
+      "epoch": 0.17,
+      "learning_rate": 4.724744821374963e-05,
+      "loss": 0.8988,
+      "step": 1478
+    },
+    {
+      "epoch": 0.17,
+      "learning_rate": 4.724557190033023e-05,
+      "loss": 0.9462,
+      "step": 1479
+    },
+    {
+      "epoch": 0.17,
+      "learning_rate": 4.724369558691084e-05,
+      "loss": 0.9633,
+      "step": 1480
+    },
+    {
+      "epoch": 0.17,
+      "learning_rate": 4.7241819273491446e-05,
+      "loss": 0.792,
+      "step": 1481
+    },
+    {
+      "epoch": 0.17,
+      "learning_rate": 4.723994296007205e-05,
+      "loss": 0.8284,
+      "step": 1482
+    },
+    {
+      "epoch": 0.17,
+      "learning_rate": 4.723806664665266e-05,
+      "loss": 0.8397,
+      "step": 1483
+    },
+    {
+      "epoch": 0.17,
+      "learning_rate": 4.7236190333233264e-05,
+      "loss": 1.0146,
+      "step": 1484
+    },
+    {
+      "epoch": 0.17,
+      "learning_rate": 4.723431401981388e-05,
+      "loss": 0.7875,
+      "step": 1485
+    },
+    {
+      "epoch": 0.17,
+      "learning_rate": 4.7232437706394476e-05,
+      "loss": 0.725,
+      "step": 1486
+    },
+    {
+      "epoch": 0.17,
+      "learning_rate": 4.723056139297509e-05,
+      "loss": 0.7469,
+      "step": 1487
+    },
+    {
+      "epoch": 0.17,
+      "learning_rate": 4.722868507955569e-05,
+      "loss": 0.6142,
+      "step": 1488
+    },
+    {
+      "epoch": 0.17,
+      "learning_rate": 4.72268087661363e-05,
+      "loss": 0.9941,
+      "step": 1489
+    },
+    {
+      "epoch": 0.17,
+      "learning_rate": 4.72249324527169e-05,
+      "loss": 0.8636,
+      "step": 1490
+    },
+    {
+      "epoch": 0.17,
+      "learning_rate": 4.7223056139297514e-05,
+      "loss": 0.8904,
+      "step": 1491
+    },
+    {
+      "epoch": 0.17,
+      "learning_rate": 4.7221179825878113e-05,
+      "loss": 0.771,
+      "step": 1492
+    },
+    {
+      "epoch": 0.17,
+      "learning_rate": 4.7219303512458726e-05,
+      "loss": 0.9497,
+      "step": 1493
+    },
+    {
+      "epoch": 0.17,
+      "learning_rate": 4.721742719903933e-05,
+      "loss": 0.8909,
+      "step": 1494
+    },
+    {
+      "epoch": 0.17,
+      "learning_rate": 4.721555088561994e-05,
+      "loss": 0.8934,
+      "step": 1495
+    },
+    {
+      "epoch": 0.17,
+      "learning_rate": 4.7213674572200545e-05,
+      "loss": 0.849,
+      "step": 1496
+    },
+    {
+      "epoch": 0.17,
+      "learning_rate": 4.721179825878115e-05,
+      "loss": 0.7978,
+      "step": 1497
+    },
+    {
+      "epoch": 0.17,
+      "learning_rate": 4.720992194536176e-05,
+      "loss": 1.0256,
+      "step": 1498
+    },
+    {
+      "epoch": 0.17,
+      "learning_rate": 4.720804563194236e-05,
+      "loss": 0.7477,
+      "step": 1499
+    },
+    {
+      "epoch": 0.17,
+      "learning_rate": 4.720616931852297e-05,
+      "loss": 1.0071,
+      "step": 1500
+    },
+    {
+      "epoch": 0.17,
+      "eval_loss": 0.8365153074264526,
+      "eval_runtime": 23.2756,
+      "eval_samples_per_second": 43.952,
+      "eval_steps_per_second": 21.997,
+      "step": 1500
+    },
+    {
+      "epoch": 0.17,
+      "learning_rate": 4.7204293005103576e-05,
+      "loss": 0.5559,
+      "step": 1501
+    },
+    {
+      "epoch": 0.17,
+      "learning_rate": 4.720241669168418e-05,
+      "loss": 0.873,
+      "step": 1502
+    },
+    {
+      "epoch": 0.17,
+      "learning_rate": 4.720054037826479e-05,
+      "loss": 0.787,
+      "step": 1503
+    },
+    {
+      "epoch": 0.17,
+      "learning_rate": 4.7198664064845394e-05,
+      "loss": 0.9837,
+      "step": 1504
+    },
+    {
+      "epoch": 0.17,
+      "learning_rate": 4.7196787751426e-05,
+      "loss": 0.8993,
+      "step": 1505
+    },
+    {
+      "epoch": 0.17,
+      "learning_rate": 4.7194911438006606e-05,
+      "loss": 0.8075,
+      "step": 1506
+    },
+    {
+      "epoch": 0.17,
+      "learning_rate": 4.719303512458721e-05,
+      "loss": 0.8087,
+      "step": 1507
+    },
+    {
+      "epoch": 0.17,
+      "learning_rate": 4.719115881116782e-05,
+      "loss": 0.607,
+      "step": 1508
+    },
+    {
+      "epoch": 0.17,
+      "learning_rate": 4.7189282497748425e-05,
+      "loss": 0.8514,
+      "step": 1509
+    },
+    {
+      "epoch": 0.17,
+      "learning_rate": 4.718740618432903e-05,
+      "loss": 0.8364,
+      "step": 1510
+    },
+    {
+      "epoch": 0.17,
+      "learning_rate": 4.718552987090964e-05,
+      "loss": 0.8592,
+      "step": 1511
+    },
+    {
+      "epoch": 0.17,
+      "learning_rate": 4.7183653557490243e-05,
+      "loss": 0.9718,
+      "step": 1512
+    },
+    {
+      "epoch": 0.17,
+      "learning_rate": 4.718177724407085e-05,
+      "loss": 1.024,
+      "step": 1513
+    },
+    {
+      "epoch": 0.17,
+      "learning_rate": 4.717990093065146e-05,
+      "loss": 0.8969,
+      "step": 1514
+    },
+    {
+      "epoch": 0.17,
+      "learning_rate": 4.717802461723206e-05,
+      "loss": 0.7329,
+      "step": 1515
+    },
+    {
+      "epoch": 0.17,
+      "learning_rate": 4.7176148303812675e-05,
+      "loss": 0.6818,
+      "step": 1516
+    },
+    {
+      "epoch": 0.17,
+      "learning_rate": 4.7174271990393274e-05,
+      "loss": 0.7506,
+      "step": 1517
+    },
+    {
+      "epoch": 0.17,
+      "learning_rate": 4.717239567697389e-05,
+      "loss": 0.8702,
+      "step": 1518
+    },
+    {
+      "epoch": 0.17,
+      "learning_rate": 4.7170519363554487e-05,
+      "loss": 0.9879,
+      "step": 1519
+    },
+    {
+      "epoch": 0.17,
+      "learning_rate": 4.71686430501351e-05,
+      "loss": 0.901,
+      "step": 1520
+    },
+    {
+      "epoch": 0.17,
+      "learning_rate": 4.71667667367157e-05,
+      "loss": 0.75,
+      "step": 1521
+    },
+    {
+      "epoch": 0.17,
+      "learning_rate": 4.716489042329631e-05,
+      "loss": 0.8801,
+      "step": 1522
+    },
+    {
+      "epoch": 0.17,
+      "learning_rate": 4.716301410987692e-05,
+      "loss": 0.6589,
+      "step": 1523
+    },
+    {
+      "epoch": 0.17,
+      "learning_rate": 4.7161137796457524e-05,
+      "loss": 0.7649,
+      "step": 1524
+    },
+    {
+      "epoch": 0.17,
+      "learning_rate": 4.715926148303813e-05,
+      "loss": 0.9687,
+      "step": 1525
+    },
+    {
+      "epoch": 0.17,
+      "learning_rate": 4.7157385169618736e-05,
+      "loss": 0.7149,
+      "step": 1526
+    },
+    {
+      "epoch": 0.17,
+      "learning_rate": 4.715550885619934e-05,
+      "loss": 0.7873,
+      "step": 1527
+    },
+    {
+      "epoch": 0.17,
+      "learning_rate": 4.715363254277995e-05,
+      "loss": 0.9057,
+      "step": 1528
+    },
+    {
+      "epoch": 0.17,
+      "learning_rate": 4.7151756229360555e-05,
+      "loss": 0.8022,
+      "step": 1529
+    },
+    {
+      "epoch": 0.17,
+      "learning_rate": 4.714987991594116e-05,
+      "loss": 0.7756,
+      "step": 1530
+    },
+    {
+      "epoch": 0.17,
+      "learning_rate": 4.714800360252177e-05,
+      "loss": 0.5657,
+      "step": 1531
+    },
+    {
+      "epoch": 0.17,
+      "learning_rate": 4.7146127289102373e-05,
+      "loss": 0.7243,
+      "step": 1532
+    },
+    {
+      "epoch": 0.17,
+      "learning_rate": 4.714425097568298e-05,
+      "loss": 0.8143,
+      "step": 1533
+    },
+    {
+      "epoch": 0.17,
+      "learning_rate": 4.7142374662263586e-05,
+      "loss": 0.9879,
+      "step": 1534
+    },
+    {
+      "epoch": 0.17,
+      "learning_rate": 4.714049834884419e-05,
+      "loss": 0.8703,
+      "step": 1535
+    },
+    {
+      "epoch": 0.17,
+      "learning_rate": 4.71386220354248e-05,
+      "loss": 0.8404,
+      "step": 1536
+    },
+    {
+      "epoch": 0.17,
+      "learning_rate": 4.7136745722005404e-05,
+      "loss": 0.8915,
+      "step": 1537
+    },
+    {
+      "epoch": 0.17,
+      "learning_rate": 4.713486940858601e-05,
+      "loss": 0.8352,
+      "step": 1538
+    },
+    {
+      "epoch": 0.17,
+      "learning_rate": 4.7132993095166617e-05,
+      "loss": 0.7034,
+      "step": 1539
+    },
+    {
+      "epoch": 0.17,
+      "learning_rate": 4.713111678174722e-05,
+      "loss": 0.7067,
+      "step": 1540
+    },
+    {
+      "epoch": 0.17,
+      "learning_rate": 4.7129240468327836e-05,
+      "loss": 0.7453,
+      "step": 1541
+    },
+    {
+      "epoch": 0.17,
+      "learning_rate": 4.7127364154908435e-05,
+      "loss": 0.7835,
+      "step": 1542
+    },
+    {
+      "epoch": 0.17,
+      "learning_rate": 4.712548784148905e-05,
+      "loss": 0.8758,
+      "step": 1543
+    },
+    {
+      "epoch": 0.17,
+      "learning_rate": 4.712361152806965e-05,
+      "loss": 1.0893,
+      "step": 1544
+    },
+    {
+      "epoch": 0.17,
+      "learning_rate": 4.712173521465026e-05,
+      "loss": 0.8165,
+      "step": 1545
+    },
+    {
+      "epoch": 0.17,
+      "learning_rate": 4.711985890123086e-05,
+      "loss": 0.9908,
+      "step": 1546
+    },
+    {
+      "epoch": 0.17,
+      "learning_rate": 4.711798258781147e-05,
+      "loss": 0.8689,
+      "step": 1547
+    },
+    {
+      "epoch": 0.17,
+      "learning_rate": 4.711610627439207e-05,
+      "loss": 0.7701,
+      "step": 1548
+    },
+    {
+      "epoch": 0.17,
+      "learning_rate": 4.7114229960972685e-05,
+      "loss": 0.6929,
+      "step": 1549
+    },
+    {
+      "epoch": 0.17,
+      "learning_rate": 4.711235364755329e-05,
+      "loss": 0.6899,
+      "step": 1550
+    },
+    {
+      "epoch": 0.17,
+      "learning_rate": 4.71104773341339e-05,
+      "loss": 0.8955,
+      "step": 1551
+    },
+    {
+      "epoch": 0.17,
+      "learning_rate": 4.7108601020714503e-05,
+      "loss": 0.872,
+      "step": 1552
+    },
+    {
+      "epoch": 0.17,
+      "learning_rate": 4.710672470729511e-05,
+      "loss": 0.942,
+      "step": 1553
+    },
+    {
+      "epoch": 0.17,
+      "learning_rate": 4.7104848393875716e-05,
+      "loss": 0.799,
+      "step": 1554
+    },
+    {
+      "epoch": 0.17,
+      "learning_rate": 4.710297208045632e-05,
+      "loss": 0.7573,
+      "step": 1555
+    },
+    {
+      "epoch": 0.18,
+      "learning_rate": 4.710109576703693e-05,
+      "loss": 0.8916,
+      "step": 1556
+    },
+    {
+      "epoch": 0.18,
+      "learning_rate": 4.7099219453617534e-05,
+      "loss": 0.8893,
+      "step": 1557
+    },
+    {
+      "epoch": 0.18,
+      "learning_rate": 4.709734314019814e-05,
+      "loss": 0.8596,
+      "step": 1558
+    },
+    {
+      "epoch": 0.18,
+      "learning_rate": 4.7095466826778747e-05,
+      "loss": 0.6168,
+      "step": 1559
+    },
+    {
+      "epoch": 0.18,
+      "learning_rate": 4.709359051335935e-05,
+      "loss": 0.6672,
+      "step": 1560
+    },
+    {
+      "epoch": 0.18,
+      "learning_rate": 4.7091714199939966e-05,
+      "loss": 0.9161,
+      "step": 1561
+    },
+    {
+      "epoch": 0.18,
+      "learning_rate": 4.7089837886520565e-05,
+      "loss": 1.1067,
+      "step": 1562
+    },
+    {
+      "epoch": 0.18,
+      "learning_rate": 4.708796157310118e-05,
+      "loss": 0.7982,
+      "step": 1563
+    },
+    {
+      "epoch": 0.18,
+      "learning_rate": 4.708608525968178e-05,
+      "loss": 0.7808,
+      "step": 1564
+    },
+    {
+      "epoch": 0.18,
+      "learning_rate": 4.708420894626239e-05,
+      "loss": 0.7662,
+      "step": 1565
+    },
+    {
+      "epoch": 0.18,
+      "learning_rate": 4.708233263284299e-05,
+      "loss": 0.8594,
+      "step": 1566
+    },
+    {
+      "epoch": 0.18,
+      "learning_rate": 4.70804563194236e-05,
+      "loss": 0.8614,
+      "step": 1567
+    },
+    {
+      "epoch": 0.18,
+      "learning_rate": 4.70785800060042e-05,
+      "loss": 0.7642,
+      "step": 1568
+    },
+    {
+      "epoch": 0.18,
+      "learning_rate": 4.707670369258481e-05,
+      "loss": 1.1507,
+      "step": 1569
+    },
+    {
+      "epoch": 0.18,
+      "learning_rate": 4.707482737916542e-05,
+      "loss": 0.9909,
+      "step": 1570
+    },
+    {
+      "epoch": 0.18,
+      "learning_rate": 4.707295106574602e-05,
+      "loss": 0.4516,
+      "step": 1571
+    },
+    {
+      "epoch": 0.18,
+      "learning_rate": 4.7071074752326633e-05,
+      "loss": 0.9312,
+      "step": 1572
+    },
+    {
+      "epoch": 0.18,
+      "learning_rate": 4.706919843890723e-05,
+      "loss": 0.9115,
+      "step": 1573
+    },
+    {
+      "epoch": 0.18,
+      "learning_rate": 4.7067322125487846e-05,
+      "loss": 0.8653,
+      "step": 1574
+    },
+    {
+      "epoch": 0.18,
+      "learning_rate": 4.7065445812068445e-05,
+      "loss": 0.7221,
+      "step": 1575
+    },
+    {
+      "epoch": 0.18,
+      "learning_rate": 4.706356949864906e-05,
+      "loss": 0.6456,
+      "step": 1576
+    },
+    {
+      "epoch": 0.18,
+      "learning_rate": 4.706169318522966e-05,
+      "loss": 0.774,
+      "step": 1577
+    },
+    {
+      "epoch": 0.18,
+      "learning_rate": 4.705981687181027e-05,
+      "loss": 1.0592,
+      "step": 1578
+    },
+    {
+      "epoch": 0.18,
+      "learning_rate": 4.7057940558390877e-05,
+      "loss": 0.7358,
+      "step": 1579
+    },
+    {
+      "epoch": 0.18,
+      "learning_rate": 4.705606424497148e-05,
+      "loss": 0.787,
+      "step": 1580
+    },
+    {
+      "epoch": 0.18,
+      "learning_rate": 4.705418793155209e-05,
+      "loss": 0.853,
+      "step": 1581
+    },
+    {
+      "epoch": 0.18,
+      "learning_rate": 4.7052311618132695e-05,
+      "loss": 0.6793,
+      "step": 1582
+    },
+    {
+      "epoch": 0.18,
+      "learning_rate": 4.70504353047133e-05,
+      "loss": 1.103,
+      "step": 1583
+    },
+    {
+      "epoch": 0.18,
+      "learning_rate": 4.704855899129391e-05,
+      "loss": 0.6504,
+      "step": 1584
+    },
+    {
+      "epoch": 0.18,
+      "learning_rate": 4.7046682677874514e-05,
+      "loss": 0.7795,
+      "step": 1585
+    },
+    {
+      "epoch": 0.18,
+      "learning_rate": 4.704480636445512e-05,
+      "loss": 0.9125,
+      "step": 1586
+    },
+    {
+      "epoch": 0.18,
+      "learning_rate": 4.7042930051035726e-05,
+      "loss": 0.6962,
+      "step": 1587
+    },
+    {
+      "epoch": 0.18,
+      "learning_rate": 4.704105373761634e-05,
+      "loss": 0.8114,
+      "step": 1588
+    },
+    {
+      "epoch": 0.18,
+      "learning_rate": 4.703917742419694e-05,
+      "loss": 0.9447,
+      "step": 1589
+    },
+    {
+      "epoch": 0.18,
+      "learning_rate": 4.703730111077755e-05,
+      "loss": 0.6463,
+      "step": 1590
+    },
+    {
+      "epoch": 0.18,
+      "learning_rate": 4.703542479735815e-05,
+      "loss": 1.0268,
+      "step": 1591
+    },
+    {
+      "epoch": 0.18,
+      "learning_rate": 4.7033548483938763e-05,
+      "loss": 0.7585,
+      "step": 1592
+    },
+    {
+      "epoch": 0.18,
+      "learning_rate": 4.703167217051936e-05,
+      "loss": 0.792,
+      "step": 1593
+    },
+    {
+      "epoch": 0.18,
+      "learning_rate": 4.7029795857099976e-05,
+      "loss": 0.5712,
+      "step": 1594
+    },
+    {
+      "epoch": 0.18,
+      "learning_rate": 4.7027919543680575e-05,
+      "loss": 0.8952,
+      "step": 1595
+    },
+    {
+      "epoch": 0.18,
+      "learning_rate": 4.702604323026119e-05,
+      "loss": 0.8175,
+      "step": 1596
+    },
+    {
+      "epoch": 0.18,
+      "learning_rate": 4.7024166916841794e-05,
+      "loss": 0.994,
+      "step": 1597
+    },
+    {
+      "epoch": 0.18,
+      "learning_rate": 4.70222906034224e-05,
+      "loss": 0.7758,
+      "step": 1598
+    },
+    {
+      "epoch": 0.18,
+      "learning_rate": 4.7020414290003007e-05,
+      "loss": 0.9766,
+      "step": 1599
+    },
+    {
+      "epoch": 0.18,
+      "learning_rate": 4.701853797658361e-05,
+      "loss": 0.9681,
+      "step": 1600
+    },
+    {
+      "epoch": 0.18,
+      "learning_rate": 4.701666166316422e-05,
+      "loss": 0.9473,
+      "step": 1601
+    },
+    {
+      "epoch": 0.18,
+      "learning_rate": 4.7014785349744825e-05,
+      "loss": 0.7894,
+      "step": 1602
+    },
+    {
+      "epoch": 0.18,
+      "learning_rate": 4.701290903632543e-05,
+      "loss": 0.8474,
+      "step": 1603
+    },
+    {
+      "epoch": 0.18,
+      "learning_rate": 4.701103272290604e-05,
+      "loss": 0.834,
+      "step": 1604
+    },
+    {
+      "epoch": 0.18,
+      "learning_rate": 4.7009156409486644e-05,
+      "loss": 0.7196,
+      "step": 1605
+    },
+    {
+      "epoch": 0.18,
+      "learning_rate": 4.700728009606725e-05,
+      "loss": 0.6325,
+      "step": 1606
+    },
+    {
+      "epoch": 0.18,
+      "learning_rate": 4.7005403782647856e-05,
+      "loss": 0.8261,
+      "step": 1607
+    },
+    {
+      "epoch": 0.18,
+      "learning_rate": 4.700352746922846e-05,
+      "loss": 0.7672,
+      "step": 1608
+    },
+    {
+      "epoch": 0.18,
+      "learning_rate": 4.700165115580907e-05,
+      "loss": 1.0163,
+      "step": 1609
+    },
+    {
+      "epoch": 0.18,
+      "learning_rate": 4.6999774842389674e-05,
+      "loss": 0.8303,
+      "step": 1610
+    },
+    {
+      "epoch": 0.18,
+      "learning_rate": 4.699789852897028e-05,
+      "loss": 0.9587,
+      "step": 1611
+    },
+    {
+      "epoch": 0.18,
+      "learning_rate": 4.699602221555089e-05,
+      "loss": 0.647,
+      "step": 1612
+    },
+    {
+      "epoch": 0.18,
+      "learning_rate": 4.699414590213149e-05,
+      "loss": 0.8768,
+      "step": 1613
+    },
+    {
+      "epoch": 0.18,
+      "learning_rate": 4.69922695887121e-05,
+      "loss": 0.8948,
+      "step": 1614
+    },
+    {
+      "epoch": 0.18,
+      "learning_rate": 4.6990393275292705e-05,
+      "loss": 0.7702,
+      "step": 1615
+    },
+    {
+      "epoch": 0.18,
+      "learning_rate": 4.698851696187331e-05,
+      "loss": 0.6073,
+      "step": 1616
+    },
+    {
+      "epoch": 0.18,
+      "learning_rate": 4.6986640648453924e-05,
+      "loss": 0.8142,
+      "step": 1617
+    },
+    {
+      "epoch": 0.18,
+      "learning_rate": 4.6984764335034524e-05,
+      "loss": 0.9531,
+      "step": 1618
+    },
+    {
+      "epoch": 0.18,
+      "learning_rate": 4.6982888021615137e-05,
+      "loss": 0.5311,
+      "step": 1619
+    },
+    {
+      "epoch": 0.18,
+      "learning_rate": 4.6981011708195736e-05,
+      "loss": 0.7324,
+      "step": 1620
+    },
+    {
+      "epoch": 0.18,
+      "learning_rate": 4.697913539477635e-05,
+      "loss": 0.7822,
+      "step": 1621
+    },
+    {
+      "epoch": 0.18,
+      "learning_rate": 4.697725908135695e-05,
+      "loss": 0.9993,
+      "step": 1622
+    },
+    {
+      "epoch": 0.18,
+      "learning_rate": 4.697538276793756e-05,
+      "loss": 0.8727,
+      "step": 1623
+    },
+    {
+      "epoch": 0.18,
+      "learning_rate": 4.697350645451816e-05,
+      "loss": 1.0126,
+      "step": 1624
+    },
+    {
+      "epoch": 0.18,
+      "learning_rate": 4.6971630141098774e-05,
+      "loss": 0.5977,
+      "step": 1625
+    },
+    {
+      "epoch": 0.18,
+      "eval_loss": 0.8300822377204895,
+      "eval_runtime": 23.2695,
+      "eval_samples_per_second": 43.963,
+      "eval_steps_per_second": 22.003,
+      "step": 1625
+    },
+    {
+      "epoch": 0.18,
+      "learning_rate": 4.696975382767938e-05,
+      "loss": 0.8251,
+      "step": 1626
+    },
+    {
+      "epoch": 0.18,
+      "learning_rate": 4.6967877514259986e-05,
+      "loss": 0.8218,
+      "step": 1627
+    },
+    {
+      "epoch": 0.18,
+      "learning_rate": 4.696600120084059e-05,
+      "loss": 0.7684,
+      "step": 1628
+    },
+    {
+      "epoch": 0.18,
+      "learning_rate": 4.69641248874212e-05,
+      "loss": 0.6328,
+      "step": 1629
+    },
+    {
+      "epoch": 0.18,
+      "learning_rate": 4.6962248574001804e-05,
+      "loss": 0.7648,
+      "step": 1630
+    },
+    {
+      "epoch": 0.18,
+      "learning_rate": 4.696037226058241e-05,
+      "loss": 0.7926,
+      "step": 1631
+    },
+    {
+      "epoch": 0.18,
+      "learning_rate": 4.695849594716302e-05,
+      "loss": 0.6411,
+      "step": 1632
+    },
+    {
+      "epoch": 0.18,
+      "learning_rate": 4.695661963374362e-05,
+      "loss": 0.7724,
+      "step": 1633
+    },
+    {
+      "epoch": 0.18,
+      "learning_rate": 4.695474332032423e-05,
+      "loss": 0.858,
+      "step": 1634
+    },
+    {
+      "epoch": 0.18,
+      "learning_rate": 4.6952867006904835e-05,
+      "loss": 0.6698,
+      "step": 1635
+    },
+    {
+      "epoch": 0.18,
+      "learning_rate": 4.695099069348544e-05,
+      "loss": 0.7484,
+      "step": 1636
+    },
+    {
+      "epoch": 0.18,
+      "learning_rate": 4.694911438006605e-05,
+      "loss": 0.7578,
+      "step": 1637
+    },
+    {
+      "epoch": 0.18,
+      "learning_rate": 4.6947238066646654e-05,
+      "loss": 1.0238,
+      "step": 1638
+    },
+    {
+      "epoch": 0.18,
+      "learning_rate": 4.694536175322726e-05,
+      "loss": 0.9051,
+      "step": 1639
+    },
+    {
+      "epoch": 0.18,
+      "learning_rate": 4.6943485439807866e-05,
+      "loss": 1.0099,
+      "step": 1640
+    },
+    {
+      "epoch": 0.18,
+      "learning_rate": 4.694160912638847e-05,
+      "loss": 0.7814,
+      "step": 1641
+    },
+    {
+      "epoch": 0.18,
+      "learning_rate": 4.693973281296908e-05,
+      "loss": 0.6164,
+      "step": 1642
+    },
+    {
+      "epoch": 0.18,
+      "learning_rate": 4.6937856499549684e-05,
+      "loss": 0.6994,
+      "step": 1643
+    },
+    {
+      "epoch": 0.19,
+      "learning_rate": 4.69359801861303e-05,
+      "loss": 0.8528,
+      "step": 1644
+    },
+    {
+      "epoch": 0.19,
+      "learning_rate": 4.69341038727109e-05,
+      "loss": 0.7722,
+      "step": 1645
+    },
+    {
+      "epoch": 0.19,
+      "learning_rate": 4.693222755929151e-05,
+      "loss": 0.6272,
+      "step": 1646
+    },
+    {
+      "epoch": 0.19,
+      "learning_rate": 4.693035124587211e-05,
+      "loss": 0.9646,
+      "step": 1647
+    },
+    {
+      "epoch": 0.19,
+      "learning_rate": 4.692847493245272e-05,
+      "loss": 0.793,
+      "step": 1648
+    },
+    {
+      "epoch": 0.19,
+      "learning_rate": 4.692659861903332e-05,
+      "loss": 0.6589,
+      "step": 1649
+    },
+    {
+      "epoch": 0.19,
+      "learning_rate": 4.6924722305613934e-05,
+      "loss": 0.8772,
+      "step": 1650
+    },
+    {
+      "epoch": 0.19,
+      "learning_rate": 4.6922845992194534e-05,
+      "loss": 0.779,
+      "step": 1651
+    },
+    {
+      "epoch": 0.19,
+      "learning_rate": 4.692096967877515e-05,
+      "loss": 0.6408,
+      "step": 1652
+    },
+    {
+      "epoch": 0.19,
+      "learning_rate": 4.6919093365355746e-05,
+      "loss": 0.7833,
+      "step": 1653
+    },
+    {
+      "epoch": 0.19,
+      "learning_rate": 4.691721705193636e-05,
+      "loss": 0.9819,
+      "step": 1654
+    },
+    {
+      "epoch": 0.19,
+      "learning_rate": 4.6915340738516965e-05,
+      "loss": 0.5844,
+      "step": 1655
+    },
+    {
+      "epoch": 0.19,
+      "learning_rate": 4.691346442509757e-05,
+      "loss": 0.6483,
+      "step": 1656
+    },
+    {
+      "epoch": 0.19,
+      "learning_rate": 4.691158811167818e-05,
+      "loss": 0.6537,
+      "step": 1657
+    },
+    {
+      "epoch": 0.19,
+      "learning_rate": 4.6909711798258784e-05,
+      "loss": 0.7625,
+      "step": 1658
+    },
+    {
+      "epoch": 0.19,
+      "learning_rate": 4.690783548483939e-05,
+      "loss": 0.6739,
+      "step": 1659
+    },
+    {
+      "epoch": 0.19,
+      "learning_rate": 4.6905959171419996e-05,
+      "loss": 1.1043,
+      "step": 1660
+    },
+    {
+      "epoch": 0.19,
+      "learning_rate": 4.69040828580006e-05,
+      "loss": 0.8615,
+      "step": 1661
+    },
+    {
+      "epoch": 0.19,
+      "learning_rate": 4.690220654458121e-05,
+      "loss": 1.0921,
+      "step": 1662
+    },
+    {
+      "epoch": 0.19,
+      "learning_rate": 4.6900330231161814e-05,
+      "loss": 0.7747,
+      "step": 1663
+    },
+    {
+      "epoch": 0.19,
+      "learning_rate": 4.689845391774243e-05,
+      "loss": 1.0102,
+      "step": 1664
+    },
+    {
+      "epoch": 0.19,
+      "learning_rate": 4.689657760432303e-05,
+      "loss": 0.7325,
+      "step": 1665
+    },
+    {
+      "epoch": 0.19,
+      "learning_rate": 4.689470129090364e-05,
+      "loss": 0.8095,
+      "step": 1666
+    },
+    {
+      "epoch": 0.19,
+      "learning_rate": 4.689282497748424e-05,
+      "loss": 0.7777,
+      "step": 1667
+    },
+    {
+      "epoch": 0.19,
+      "learning_rate": 4.689094866406485e-05,
+      "loss": 0.999,
+      "step": 1668
+    },
+    {
+      "epoch": 0.19,
+      "learning_rate": 4.688907235064545e-05,
+      "loss": 0.7767,
+      "step": 1669
+    },
+    {
+      "epoch": 0.19,
+      "learning_rate": 4.6887196037226064e-05,
+      "loss": 0.9132,
+      "step": 1670
+    },
+    {
+      "epoch": 0.19,
+      "learning_rate": 4.6885319723806664e-05,
+      "loss": 0.7775,
+      "step": 1671
+    },
+    {
+      "epoch": 0.19,
+      "learning_rate": 4.688344341038728e-05,
+      "loss": 1.0496,
+      "step": 1672
+    },
+    {
+      "epoch": 0.19,
+      "learning_rate": 4.688156709696788e-05,
+      "loss": 0.9859,
+      "step": 1673
+    },
+    {
+      "epoch": 0.19,
+      "learning_rate": 4.687969078354848e-05,
+      "loss": 0.7401,
+      "step": 1674
+    },
+    {
+      "epoch": 0.19,
+      "learning_rate": 4.6877814470129095e-05,
+      "loss": 0.9427,
+      "step": 1675
+    },
+    {
+      "epoch": 0.19,
+      "learning_rate": 4.6875938156709695e-05,
+      "loss": 0.633,
+      "step": 1676
+    },
+    {
+      "epoch": 0.19,
+      "learning_rate": 4.687406184329031e-05,
+      "loss": 1.0524,
+      "step": 1677
+    },
+    {
+      "epoch": 0.19,
+      "learning_rate": 4.687218552987091e-05,
+      "loss": 0.9047,
+      "step": 1678
+    },
+    {
+      "epoch": 0.19,
+      "learning_rate": 4.687030921645152e-05,
+      "loss": 0.7394,
+      "step": 1679
+    },
+    {
+      "epoch": 0.19,
+      "learning_rate": 4.686843290303212e-05,
+      "loss": 0.7733,
+      "step": 1680
+    },
+    {
+      "epoch": 0.19,
+      "learning_rate": 4.686655658961273e-05,
+      "loss": 0.8078,
+      "step": 1681
+    },
+    {
+      "epoch": 0.19,
+      "learning_rate": 4.686468027619334e-05,
+      "loss": 0.9703,
+      "step": 1682
+    },
+    {
+      "epoch": 0.19,
+      "learning_rate": 4.6862803962773944e-05,
+      "loss": 0.735,
+      "step": 1683
+    },
+    {
+      "epoch": 0.19,
+      "learning_rate": 4.686092764935455e-05,
+      "loss": 0.9072,
+      "step": 1684
+    },
+    {
+      "epoch": 0.19,
+      "learning_rate": 4.685905133593516e-05,
+      "loss": 1.0512,
+      "step": 1685
+    },
+    {
+      "epoch": 0.19,
+      "learning_rate": 4.685717502251576e-05,
+      "loss": 0.5705,
+      "step": 1686
+    },
+    {
+      "epoch": 0.19,
+      "learning_rate": 4.685529870909637e-05,
+      "loss": 0.8766,
+      "step": 1687
+    },
+    {
+      "epoch": 0.19,
+      "learning_rate": 4.6853422395676975e-05,
+      "loss": 0.8075,
+      "step": 1688
+    },
+    {
+      "epoch": 0.19,
+      "learning_rate": 4.685154608225758e-05,
+      "loss": 0.8635,
+      "step": 1689
+    },
+    {
+      "epoch": 0.19,
+      "learning_rate": 4.684966976883819e-05,
+      "loss": 1.0114,
+      "step": 1690
+    },
+    {
+      "epoch": 0.19,
+      "learning_rate": 4.68477934554188e-05,
+      "loss": 0.8562,
+      "step": 1691
+    },
+    {
+      "epoch": 0.19,
+      "learning_rate": 4.68459171419994e-05,
+      "loss": 0.9755,
+      "step": 1692
+    },
+    {
+      "epoch": 0.19,
+      "learning_rate": 4.684404082858001e-05,
+      "loss": 1.0923,
+      "step": 1693
+    },
+    {
+      "epoch": 0.19,
+      "learning_rate": 4.684216451516061e-05,
+      "loss": 0.912,
+      "step": 1694
+    },
+    {
+      "epoch": 0.19,
+      "learning_rate": 4.6840288201741225e-05,
+      "loss": 0.8952,
+      "step": 1695
+    },
+    {
+      "epoch": 0.19,
+      "learning_rate": 4.6838411888321825e-05,
+      "loss": 0.8594,
+      "step": 1696
+    },
+    {
+      "epoch": 0.19,
+      "learning_rate": 4.683653557490244e-05,
+      "loss": 0.736,
+      "step": 1697
+    },
+    {
+      "epoch": 0.19,
+      "learning_rate": 4.683465926148304e-05,
+      "loss": 0.777,
+      "step": 1698
+    },
+    {
+      "epoch": 0.19,
+      "learning_rate": 4.683278294806365e-05,
+      "loss": 0.6591,
+      "step": 1699
+    },
+    {
+      "epoch": 0.19,
+      "learning_rate": 4.6830906634644256e-05,
+      "loss": 0.8806,
+      "step": 1700
+    },
+    {
+      "epoch": 0.19,
+      "learning_rate": 4.682903032122486e-05,
+      "loss": 0.7342,
+      "step": 1701
+    },
+    {
+      "epoch": 0.19,
+      "learning_rate": 4.682715400780547e-05,
+      "loss": 0.8212,
+      "step": 1702
+    },
+    {
+      "epoch": 0.19,
+      "learning_rate": 4.6825277694386074e-05,
+      "loss": 0.9187,
+      "step": 1703
+    },
+    {
+      "epoch": 0.19,
+      "learning_rate": 4.682340138096668e-05,
+      "loss": 0.9016,
+      "step": 1704
+    },
+    {
+      "epoch": 0.19,
+      "learning_rate": 4.682152506754729e-05,
+      "loss": 0.8388,
+      "step": 1705
+    },
+    {
+      "epoch": 0.19,
+      "learning_rate": 4.681964875412789e-05,
+      "loss": 0.6481,
+      "step": 1706
+    },
+    {
+      "epoch": 0.19,
+      "learning_rate": 4.68177724407085e-05,
+      "loss": 0.6732,
+      "step": 1707
+    },
+    {
+      "epoch": 0.19,
+      "learning_rate": 4.6815896127289105e-05,
+      "loss": 0.7406,
+      "step": 1708
+    },
+    {
+      "epoch": 0.19,
+      "learning_rate": 4.681401981386971e-05,
+      "loss": 0.8431,
+      "step": 1709
+    },
+    {
+      "epoch": 0.19,
+      "learning_rate": 4.681214350045032e-05,
+      "loss": 0.6893,
+      "step": 1710
+    },
+    {
+      "epoch": 0.19,
+      "learning_rate": 4.6810267187030924e-05,
+      "loss": 0.9369,
+      "step": 1711
+    },
+    {
+      "epoch": 0.19,
+      "learning_rate": 4.680839087361153e-05,
+      "loss": 0.8064,
+      "step": 1712
+    },
+    {
+      "epoch": 0.19,
+      "learning_rate": 4.6806514560192136e-05,
+      "loss": 0.6934,
+      "step": 1713
+    },
+    {
+      "epoch": 0.19,
+      "learning_rate": 4.680463824677274e-05,
+      "loss": 0.8338,
+      "step": 1714
+    },
+    {
+      "epoch": 0.19,
+      "learning_rate": 4.680276193335335e-05,
+      "loss": 0.843,
+      "step": 1715
+    },
+    {
+      "epoch": 0.19,
+      "learning_rate": 4.6800885619933955e-05,
+      "loss": 0.7624,
+      "step": 1716
+    },
+    {
+      "epoch": 0.19,
+      "learning_rate": 4.679900930651456e-05,
+      "loss": 0.8815,
+      "step": 1717
+    },
+    {
+      "epoch": 0.19,
+      "learning_rate": 4.679713299309517e-05,
+      "loss": 0.9382,
+      "step": 1718
+    },
+    {
+      "epoch": 0.19,
+      "learning_rate": 4.679525667967577e-05,
+      "loss": 0.8639,
+      "step": 1719
+    },
+    {
+      "epoch": 0.19,
+      "learning_rate": 4.6793380366256386e-05,
+      "loss": 0.6454,
+      "step": 1720
+    },
+    {
+      "epoch": 0.19,
+      "learning_rate": 4.6791504052836985e-05,
+      "loss": 0.8154,
+      "step": 1721
+    },
+    {
+      "epoch": 0.19,
+      "learning_rate": 4.67896277394176e-05,
+      "loss": 0.9152,
+      "step": 1722
+    },
+    {
+      "epoch": 0.19,
+      "learning_rate": 4.67877514259982e-05,
+      "loss": 0.6474,
+      "step": 1723
+    },
+    {
+      "epoch": 0.19,
+      "learning_rate": 4.678587511257881e-05,
+      "loss": 0.6942,
+      "step": 1724
+    },
+    {
+      "epoch": 0.19,
+      "learning_rate": 4.678399879915941e-05,
+      "loss": 0.6988,
+      "step": 1725
+    },
+    {
+      "epoch": 0.19,
+      "learning_rate": 4.678212248574002e-05,
+      "loss": 0.7143,
+      "step": 1726
+    },
+    {
+      "epoch": 0.19,
+      "learning_rate": 4.678024617232062e-05,
+      "loss": 0.8829,
+      "step": 1727
+    },
+    {
+      "epoch": 0.19,
+      "learning_rate": 4.6778369858901235e-05,
+      "loss": 0.7721,
+      "step": 1728
+    },
+    {
+      "epoch": 0.19,
+      "learning_rate": 4.677649354548184e-05,
+      "loss": 0.7004,
+      "step": 1729
+    },
+    {
+      "epoch": 0.19,
+      "learning_rate": 4.677461723206245e-05,
+      "loss": 0.9397,
+      "step": 1730
+    },
+    {
+      "epoch": 0.19,
+      "learning_rate": 4.6772740918643054e-05,
+      "loss": 1.0509,
+      "step": 1731
+    },
+    {
+      "epoch": 0.19,
+      "learning_rate": 4.677086460522366e-05,
+      "loss": 0.7832,
+      "step": 1732
+    },
+    {
+      "epoch": 0.2,
+      "learning_rate": 4.6768988291804266e-05,
+      "loss": 0.6619,
+      "step": 1733
+    },
+    {
+      "epoch": 0.2,
+      "learning_rate": 4.676711197838487e-05,
+      "loss": 0.8088,
+      "step": 1734
+    },
+    {
+      "epoch": 0.2,
+      "learning_rate": 4.676523566496548e-05,
+      "loss": 0.9868,
+      "step": 1735
+    },
+    {
+      "epoch": 0.2,
+      "learning_rate": 4.6763359351546085e-05,
+      "loss": 1.0153,
+      "step": 1736
+    },
+    {
+      "epoch": 0.2,
+      "learning_rate": 4.676148303812669e-05,
+      "loss": 0.7701,
+      "step": 1737
+    },
+    {
+      "epoch": 0.2,
+      "learning_rate": 4.67596067247073e-05,
+      "loss": 0.833,
+      "step": 1738
+    },
+    {
+      "epoch": 0.2,
+      "learning_rate": 4.67577304112879e-05,
+      "loss": 0.7783,
+      "step": 1739
+    },
+    {
+      "epoch": 0.2,
+      "learning_rate": 4.675585409786851e-05,
+      "loss": 0.8394,
+      "step": 1740
+    },
+    {
+      "epoch": 0.2,
+      "learning_rate": 4.6753977784449115e-05,
+      "loss": 0.7896,
+      "step": 1741
+    },
+    {
+      "epoch": 0.2,
+      "learning_rate": 4.675210147102972e-05,
+      "loss": 0.953,
+      "step": 1742
+    },
+    {
+      "epoch": 0.2,
+      "learning_rate": 4.675022515761033e-05,
+      "loss": 0.904,
+      "step": 1743
+    },
+    {
+      "epoch": 0.2,
+      "learning_rate": 4.6748348844190934e-05,
+      "loss": 0.8268,
+      "step": 1744
+    },
+    {
+      "epoch": 0.2,
+      "learning_rate": 4.674647253077154e-05,
+      "loss": 0.779,
+      "step": 1745
+    },
+    {
+      "epoch": 0.2,
+      "learning_rate": 4.6744596217352146e-05,
+      "loss": 0.7501,
+      "step": 1746
+    },
+    {
+      "epoch": 0.2,
+      "learning_rate": 4.674271990393276e-05,
+      "loss": 0.8394,
+      "step": 1747
+    },
+    {
+      "epoch": 0.2,
+      "learning_rate": 4.674084359051336e-05,
+      "loss": 0.7463,
+      "step": 1748
+    },
+    {
+      "epoch": 0.2,
+      "learning_rate": 4.673896727709397e-05,
+      "loss": 0.9642,
+      "step": 1749
+    },
+    {
+      "epoch": 0.2,
+      "learning_rate": 4.673709096367457e-05,
+      "loss": 0.8634,
+      "step": 1750
+    },
+    {
+      "epoch": 0.2,
+      "eval_loss": 0.8240072131156921,
+      "eval_runtime": 23.2583,
+      "eval_samples_per_second": 43.984,
+      "eval_steps_per_second": 22.014,
+      "step": 1750
+    },
+    {
+      "epoch": 0.2,
+      "learning_rate": 4.6735214650255184e-05,
+      "loss": 0.7509,
+      "step": 1751
+    },
+    {
+      "epoch": 0.2,
+      "learning_rate": 4.673333833683578e-05,
+      "loss": 0.9436,
+      "step": 1752
+    },
+    {
+      "epoch": 0.2,
+      "learning_rate": 4.6731462023416396e-05,
+      "loss": 1.0338,
+      "step": 1753
+    },
+    {
+      "epoch": 0.2,
+      "learning_rate": 4.6729585709996995e-05,
+      "loss": 0.9217,
+      "step": 1754
+    },
+    {
+      "epoch": 0.2,
+      "learning_rate": 4.672770939657761e-05,
+      "loss": 0.6874,
+      "step": 1755
+    },
+    {
+      "epoch": 0.2,
+      "learning_rate": 4.672583308315821e-05,
+      "loss": 0.6236,
+      "step": 1756
+    },
+    {
+      "epoch": 0.2,
+      "learning_rate": 4.672395676973882e-05,
+      "loss": 0.8483,
+      "step": 1757
+    },
+    {
+      "epoch": 0.2,
+      "learning_rate": 4.672208045631943e-05,
+      "loss": 0.8632,
+      "step": 1758
+    },
+    {
+      "epoch": 0.2,
+      "learning_rate": 4.672020414290003e-05,
+      "loss": 0.6379,
+      "step": 1759
+    },
+    {
+      "epoch": 0.2,
+      "learning_rate": 4.671832782948064e-05,
+      "loss": 0.7474,
+      "step": 1760
+    },
+    {
+      "epoch": 0.2,
+      "learning_rate": 4.6716451516061245e-05,
+      "loss": 0.7942,
+      "step": 1761
+    },
+    {
+      "epoch": 0.2,
+      "learning_rate": 4.671457520264185e-05,
+      "loss": 0.8625,
+      "step": 1762
+    },
+    {
+      "epoch": 0.2,
+      "learning_rate": 4.671269888922246e-05,
+      "loss": 0.7562,
+      "step": 1763
+    },
+    {
+      "epoch": 0.2,
+      "learning_rate": 4.6710822575803064e-05,
+      "loss": 0.6956,
+      "step": 1764
+    },
+    {
+      "epoch": 0.2,
+      "learning_rate": 4.670894626238367e-05,
+      "loss": 0.7193,
+      "step": 1765
+    },
+    {
+      "epoch": 0.2,
+      "learning_rate": 4.6707069948964276e-05,
+      "loss": 0.6686,
+      "step": 1766
+    },
+    {
+      "epoch": 0.2,
+      "learning_rate": 4.670519363554489e-05,
+      "loss": 0.9965,
+      "step": 1767
+    },
+    {
+      "epoch": 0.2,
+      "learning_rate": 4.670331732212549e-05,
+      "loss": 0.8843,
+      "step": 1768
+    },
+    {
+      "epoch": 0.2,
+      "learning_rate": 4.67014410087061e-05,
+      "loss": 0.8249,
+      "step": 1769
+    },
+    {
+      "epoch": 0.2,
+      "learning_rate": 4.66995646952867e-05,
+      "loss": 0.8173,
+      "step": 1770
+    },
+    {
+      "epoch": 0.2,
+      "learning_rate": 4.6697688381867314e-05,
+      "loss": 0.8148,
+      "step": 1771
+    },
+    {
+      "epoch": 0.2,
+      "learning_rate": 4.669581206844791e-05,
+      "loss": 0.8845,
+      "step": 1772
+    },
+    {
+      "epoch": 0.2,
+      "learning_rate": 4.6693935755028526e-05,
+      "loss": 0.8485,
+      "step": 1773
+    },
+    {
+      "epoch": 0.2,
+      "learning_rate": 4.6692059441609125e-05,
+      "loss": 1.0421,
+      "step": 1774
+    },
+    {
+      "epoch": 0.2,
+      "learning_rate": 4.669018312818974e-05,
+      "loss": 0.9129,
+      "step": 1775
+    },
+    {
+      "epoch": 0.2,
+      "learning_rate": 4.6688306814770345e-05,
+      "loss": 1.2056,
+      "step": 1776
+    },
+    {
+      "epoch": 0.2,
+      "learning_rate": 4.668643050135095e-05,
+      "loss": 0.9106,
+      "step": 1777
+    },
+    {
+      "epoch": 0.2,
+      "learning_rate": 4.668455418793156e-05,
+      "loss": 0.8137,
+      "step": 1778
+    },
+    {
+      "epoch": 0.2,
+      "learning_rate": 4.6682677874512156e-05,
+      "loss": 1.1095,
+      "step": 1779
+    },
+    {
+      "epoch": 0.2,
+      "learning_rate": 4.668080156109277e-05,
+      "loss": 0.781,
+      "step": 1780
+    },
+    {
+      "epoch": 0.2,
+      "learning_rate": 4.667892524767337e-05,
+      "loss": 0.826,
+      "step": 1781
+    },
+    {
+      "epoch": 0.2,
+      "learning_rate": 4.667704893425398e-05,
+      "loss": 0.8443,
+      "step": 1782
+    },
+    {
+      "epoch": 0.2,
+      "learning_rate": 4.667517262083458e-05,
+      "loss": 0.8214,
+      "step": 1783
+    },
+    {
+      "epoch": 0.2,
+      "learning_rate": 4.6673296307415194e-05,
+      "loss": 0.7067,
+      "step": 1784
+    },
+    {
+      "epoch": 0.2,
+      "learning_rate": 4.66714199939958e-05,
+      "loss": 0.8733,
+      "step": 1785
+    },
+    {
+      "epoch": 0.2,
+      "learning_rate": 4.6669543680576406e-05,
+      "loss": 0.888,
+      "step": 1786
+    },
+    {
+      "epoch": 0.2,
+      "learning_rate": 4.666766736715701e-05,
+      "loss": 0.8421,
+      "step": 1787
+    },
+    {
+      "epoch": 0.2,
+      "learning_rate": 4.666579105373762e-05,
+      "loss": 0.5273,
+      "step": 1788
+    },
+    {
+      "epoch": 0.2,
+      "learning_rate": 4.6663914740318225e-05,
+      "loss": 0.9055,
+      "step": 1789
+    },
+    {
+      "epoch": 0.2,
+      "learning_rate": 4.666203842689883e-05,
+      "loss": 0.8279,
+      "step": 1790
+    },
+    {
+      "epoch": 0.2,
+      "learning_rate": 4.666016211347944e-05,
+      "loss": 0.9634,
+      "step": 1791
+    },
+    {
+      "epoch": 0.2,
+      "learning_rate": 4.665828580006004e-05,
+      "loss": 0.5038,
+      "step": 1792
+    },
+    {
+      "epoch": 0.2,
+      "learning_rate": 4.665640948664065e-05,
+      "loss": 0.9116,
+      "step": 1793
+    },
+    {
+      "epoch": 0.2,
+      "learning_rate": 4.665453317322126e-05,
+      "loss": 1.0829,
+      "step": 1794
+    },
+    {
+      "epoch": 0.2,
+      "learning_rate": 4.665265685980186e-05,
+      "loss": 0.8307,
+      "step": 1795
+    },
+    {
+      "epoch": 0.2,
+      "learning_rate": 4.6650780546382475e-05,
+      "loss": 0.692,
+      "step": 1796
+    },
+    {
+      "epoch": 0.2,
+      "learning_rate": 4.6648904232963074e-05,
+      "loss": 0.5896,
+      "step": 1797
+    },
+    {
+      "epoch": 0.2,
+      "learning_rate": 4.664702791954369e-05,
+      "loss": 0.8186,
+      "step": 1798
+    },
+    {
+      "epoch": 0.2,
+      "learning_rate": 4.6645151606124286e-05,
+      "loss": 0.6684,
+      "step": 1799
+    },
+    {
+      "epoch": 0.2,
+      "learning_rate": 4.66432752927049e-05,
+      "loss": 0.8075,
+      "step": 1800
+    },
+    {
+      "epoch": 0.2,
+      "learning_rate": 4.66413989792855e-05,
+      "loss": 0.9208,
+      "step": 1801
+    },
+    {
+      "epoch": 0.2,
+      "learning_rate": 4.663952266586611e-05,
+      "loss": 0.8946,
+      "step": 1802
+    },
+    {
+      "epoch": 0.2,
+      "learning_rate": 4.663764635244671e-05,
+      "loss": 0.8721,
+      "step": 1803
+    },
+    {
+      "epoch": 0.2,
+      "learning_rate": 4.6635770039027324e-05,
+      "loss": 0.7168,
+      "step": 1804
+    },
+    {
+      "epoch": 0.2,
+      "learning_rate": 4.663389372560793e-05,
+      "loss": 0.9797,
+      "step": 1805
+    },
+    {
+      "epoch": 0.2,
+      "learning_rate": 4.6632017412188536e-05,
+      "loss": 0.7276,
+      "step": 1806
+    },
+    {
+      "epoch": 0.2,
+      "learning_rate": 4.663014109876914e-05,
+      "loss": 0.9529,
+      "step": 1807
+    },
+    {
+      "epoch": 0.2,
+      "learning_rate": 4.662826478534975e-05,
+      "loss": 0.6039,
+      "step": 1808
+    },
+    {
+      "epoch": 0.2,
+      "learning_rate": 4.6626388471930355e-05,
+      "loss": 0.6843,
+      "step": 1809
+    },
+    {
+      "epoch": 0.2,
+      "learning_rate": 4.662451215851096e-05,
+      "loss": 0.8813,
+      "step": 1810
+    },
+    {
+      "epoch": 0.2,
+      "learning_rate": 4.662263584509157e-05,
+      "loss": 0.5797,
+      "step": 1811
+    },
+    {
+      "epoch": 0.2,
+      "learning_rate": 4.662075953167217e-05,
+      "loss": 0.9014,
+      "step": 1812
+    },
+    {
+      "epoch": 0.2,
+      "learning_rate": 4.661888321825278e-05,
+      "loss": 0.6663,
+      "step": 1813
+    },
+    {
+      "epoch": 0.2,
+      "learning_rate": 4.6617006904833385e-05,
+      "loss": 0.8453,
+      "step": 1814
+    },
+    {
+      "epoch": 0.2,
+      "learning_rate": 4.661513059141399e-05,
+      "loss": 0.6201,
+      "step": 1815
+    },
+    {
+      "epoch": 0.2,
+      "learning_rate": 4.66132542779946e-05,
+      "loss": 0.9721,
+      "step": 1816
+    },
+    {
+      "epoch": 0.2,
+      "learning_rate": 4.6611377964575204e-05,
+      "loss": 0.8493,
+      "step": 1817
+    },
+    {
+      "epoch": 0.2,
+      "learning_rate": 4.660950165115581e-05,
+      "loss": 0.8442,
+      "step": 1818
+    },
+    {
+      "epoch": 0.2,
+      "learning_rate": 4.6607625337736416e-05,
+      "loss": 0.7253,
+      "step": 1819
+    },
+    {
+      "epoch": 0.2,
+      "learning_rate": 4.660574902431702e-05,
+      "loss": 0.7373,
+      "step": 1820
+    },
+    {
+      "epoch": 0.2,
+      "learning_rate": 4.660387271089763e-05,
+      "loss": 0.9214,
+      "step": 1821
+    },
+    {
+      "epoch": 0.21,
+      "learning_rate": 4.6601996397478235e-05,
+      "loss": 0.5685,
+      "step": 1822
+    },
+    {
+      "epoch": 0.21,
+      "learning_rate": 4.660012008405885e-05,
+      "loss": 0.9336,
+      "step": 1823
+    },
+    {
+      "epoch": 0.21,
+      "learning_rate": 4.659824377063945e-05,
+      "loss": 0.8148,
+      "step": 1824
+    },
+    {
+      "epoch": 0.21,
+      "learning_rate": 4.659636745722006e-05,
+      "loss": 0.9981,
+      "step": 1825
+    },
+    {
+      "epoch": 0.21,
+      "learning_rate": 4.659449114380066e-05,
+      "loss": 1.0388,
+      "step": 1826
+    },
+    {
+      "epoch": 0.21,
+      "learning_rate": 4.659261483038127e-05,
+      "loss": 0.9462,
+      "step": 1827
+    },
+    {
+      "epoch": 0.21,
+      "learning_rate": 4.659073851696187e-05,
+      "loss": 0.6925,
+      "step": 1828
+    },
+    {
+      "epoch": 0.21,
+      "learning_rate": 4.6588862203542485e-05,
+      "loss": 0.6772,
+      "step": 1829
+    },
+    {
+      "epoch": 0.21,
+      "learning_rate": 4.6586985890123084e-05,
+      "loss": 0.9687,
+      "step": 1830
+    },
+    {
+      "epoch": 0.21,
+      "learning_rate": 4.65851095767037e-05,
+      "loss": 0.6729,
+      "step": 1831
+    },
+    {
+      "epoch": 0.21,
+      "learning_rate": 4.65832332632843e-05,
+      "loss": 0.8875,
+      "step": 1832
+    },
+    {
+      "epoch": 0.21,
+      "learning_rate": 4.658135694986491e-05,
+      "loss": 0.6476,
+      "step": 1833
+    },
+    {
+      "epoch": 0.21,
+      "learning_rate": 4.6579480636445515e-05,
+      "loss": 0.8849,
+      "step": 1834
+    },
+    {
+      "epoch": 0.21,
+      "learning_rate": 4.657760432302612e-05,
+      "loss": 0.7497,
+      "step": 1835
+    },
+    {
+      "epoch": 0.21,
+      "learning_rate": 4.657572800960673e-05,
+      "loss": 1.07,
+      "step": 1836
+    },
+    {
+      "epoch": 0.21,
+      "learning_rate": 4.6573851696187334e-05,
+      "loss": 0.5383,
+      "step": 1837
+    },
+    {
+      "epoch": 0.21,
+      "learning_rate": 4.657197538276794e-05,
+      "loss": 1.5891,
+      "step": 1838
+    },
+    {
+      "epoch": 0.21,
+      "learning_rate": 4.6570099069348546e-05,
+      "loss": 0.9832,
+      "step": 1839
+    },
+    {
+      "epoch": 0.21,
+      "learning_rate": 4.656822275592915e-05,
+      "loss": 0.7111,
+      "step": 1840
+    },
+    {
+      "epoch": 0.21,
+      "learning_rate": 4.656634644250976e-05,
+      "loss": 1.1197,
+      "step": 1841
+    },
+    {
+      "epoch": 0.21,
+      "learning_rate": 4.6564470129090365e-05,
+      "loss": 0.6505,
+      "step": 1842
+    },
+    {
+      "epoch": 0.21,
+      "learning_rate": 4.656259381567097e-05,
+      "loss": 0.7584,
+      "step": 1843
+    },
+    {
+      "epoch": 0.21,
+      "learning_rate": 4.656071750225158e-05,
+      "loss": 0.7956,
+      "step": 1844
+    },
+    {
+      "epoch": 0.21,
+      "learning_rate": 4.655884118883218e-05,
+      "loss": 0.6881,
+      "step": 1845
+    },
+    {
+      "epoch": 0.21,
+      "learning_rate": 4.655696487541279e-05,
+      "loss": 0.6798,
+      "step": 1846
+    },
+    {
+      "epoch": 0.21,
+      "learning_rate": 4.6555088561993396e-05,
+      "loss": 0.7302,
+      "step": 1847
+    },
+    {
+      "epoch": 0.21,
+      "learning_rate": 4.6553212248574e-05,
+      "loss": 0.9246,
+      "step": 1848
+    },
+    {
+      "epoch": 0.21,
+      "learning_rate": 4.655133593515461e-05,
+      "loss": 0.9639,
+      "step": 1849
+    },
+    {
+      "epoch": 0.21,
+      "learning_rate": 4.6549459621735214e-05,
+      "loss": 1.0424,
+      "step": 1850
+    },
+    {
+      "epoch": 0.21,
+      "learning_rate": 4.654758330831582e-05,
+      "loss": 0.6103,
+      "step": 1851
+    },
+    {
+      "epoch": 0.21,
+      "learning_rate": 4.654570699489643e-05,
+      "loss": 0.7235,
+      "step": 1852
+    },
+    {
+      "epoch": 0.21,
+      "learning_rate": 4.654383068147703e-05,
+      "loss": 0.8675,
+      "step": 1853
+    },
+    {
+      "epoch": 0.21,
+      "learning_rate": 4.6541954368057645e-05,
+      "loss": 0.7287,
+      "step": 1854
+    },
+    {
+      "epoch": 0.21,
+      "learning_rate": 4.6540078054638245e-05,
+      "loss": 0.8038,
+      "step": 1855
+    },
+    {
+      "epoch": 0.21,
+      "learning_rate": 4.653820174121886e-05,
+      "loss": 0.9624,
+      "step": 1856
+    },
+    {
+      "epoch": 0.21,
+      "learning_rate": 4.653632542779946e-05,
+      "loss": 0.8678,
+      "step": 1857
+    },
+    {
+      "epoch": 0.21,
+      "learning_rate": 4.653444911438007e-05,
+      "loss": 0.5334,
+      "step": 1858
+    },
+    {
+      "epoch": 0.21,
+      "learning_rate": 4.653257280096067e-05,
+      "loss": 1.1447,
+      "step": 1859
+    },
+    {
+      "epoch": 0.21,
+      "learning_rate": 4.653069648754128e-05,
+      "loss": 0.8687,
+      "step": 1860
+    },
+    {
+      "epoch": 0.21,
+      "learning_rate": 4.652882017412189e-05,
+      "loss": 1.1066,
+      "step": 1861
+    },
+    {
+      "epoch": 0.21,
+      "learning_rate": 4.6526943860702495e-05,
+      "loss": 0.8024,
+      "step": 1862
+    },
+    {
+      "epoch": 0.21,
+      "learning_rate": 4.65250675472831e-05,
+      "loss": 0.7126,
+      "step": 1863
+    },
+    {
+      "epoch": 0.21,
+      "learning_rate": 4.652319123386371e-05,
+      "loss": 0.8822,
+      "step": 1864
+    },
+    {
+      "epoch": 0.21,
+      "learning_rate": 4.652131492044431e-05,
+      "loss": 0.7024,
+      "step": 1865
+    },
+    {
+      "epoch": 0.21,
+      "learning_rate": 4.651943860702492e-05,
+      "loss": 0.851,
+      "step": 1866
+    },
+    {
+      "epoch": 0.21,
+      "learning_rate": 4.6517562293605526e-05,
+      "loss": 0.861,
+      "step": 1867
+    },
+    {
+      "epoch": 0.21,
+      "learning_rate": 4.651568598018613e-05,
+      "loss": 1.0385,
+      "step": 1868
+    },
+    {
+      "epoch": 0.21,
+      "learning_rate": 4.651380966676674e-05,
+      "loss": 0.8219,
+      "step": 1869
+    },
+    {
+      "epoch": 0.21,
+      "learning_rate": 4.651193335334735e-05,
+      "loss": 0.7862,
+      "step": 1870
+    },
+    {
+      "epoch": 0.21,
+      "learning_rate": 4.651005703992795e-05,
+      "loss": 0.7041,
+      "step": 1871
+    },
+    {
+      "epoch": 0.21,
+      "learning_rate": 4.650818072650856e-05,
+      "loss": 0.8006,
+      "step": 1872
+    },
+    {
+      "epoch": 0.21,
+      "learning_rate": 4.650630441308916e-05,
+      "loss": 0.6275,
+      "step": 1873
+    },
+    {
+      "epoch": 0.21,
+      "learning_rate": 4.6504428099669775e-05,
+      "loss": 0.8793,
+      "step": 1874
+    },
+    {
+      "epoch": 0.21,
+      "learning_rate": 4.6502551786250375e-05,
+      "loss": 0.8487,
+      "step": 1875
+    },
+    {
+      "epoch": 0.21,
+      "eval_loss": 0.8268378376960754,
+      "eval_runtime": 23.301,
+      "eval_samples_per_second": 43.904,
+      "eval_steps_per_second": 21.973,
+      "step": 1875
+    },
+    {
+      "epoch": 0.21,
+      "learning_rate": 4.650067547283099e-05,
+      "loss": 0.9243,
+      "step": 1876
+    },
+    {
+      "epoch": 0.21,
+      "learning_rate": 4.649879915941159e-05,
+      "loss": 0.8766,
+      "step": 1877
+    },
+    {
+      "epoch": 0.21,
+      "learning_rate": 4.64969228459922e-05,
+      "loss": 0.7194,
+      "step": 1878
+    },
+    {
+      "epoch": 0.21,
+      "learning_rate": 4.6495046532572806e-05,
+      "loss": 0.8982,
+      "step": 1879
+    },
+    {
+      "epoch": 0.21,
+      "learning_rate": 4.649317021915341e-05,
+      "loss": 0.7734,
+      "step": 1880
+    },
+    {
+      "epoch": 0.21,
+      "learning_rate": 4.649129390573402e-05,
+      "loss": 0.9306,
+      "step": 1881
+    },
+    {
+      "epoch": 0.21,
+      "learning_rate": 4.6489417592314625e-05,
+      "loss": 1.058,
+      "step": 1882
+    },
+    {
+      "epoch": 0.21,
+      "learning_rate": 4.648754127889523e-05,
+      "loss": 0.9668,
+      "step": 1883
+    },
+    {
+      "epoch": 0.21,
+      "learning_rate": 4.648566496547583e-05,
+      "loss": 0.7406,
+      "step": 1884
+    },
+    {
+      "epoch": 0.21,
+      "learning_rate": 4.648378865205644e-05,
+      "loss": 0.5407,
+      "step": 1885
+    },
+    {
+      "epoch": 0.21,
+      "learning_rate": 4.648191233863704e-05,
+      "loss": 0.6796,
+      "step": 1886
+    },
+    {
+      "epoch": 0.21,
+      "learning_rate": 4.6480036025217656e-05,
+      "loss": 0.9454,
+      "step": 1887
+    },
+    {
+      "epoch": 0.21,
+      "learning_rate": 4.647815971179826e-05,
+      "loss": 0.7325,
+      "step": 1888
+    },
+    {
+      "epoch": 0.21,
+      "learning_rate": 4.647628339837887e-05,
+      "loss": 0.7654,
+      "step": 1889
+    },
+    {
+      "epoch": 0.21,
+      "learning_rate": 4.6474407084959474e-05,
+      "loss": 0.7804,
+      "step": 1890
+    },
+    {
+      "epoch": 0.21,
+      "learning_rate": 4.647253077154008e-05,
+      "loss": 0.926,
+      "step": 1891
+    },
+    {
+      "epoch": 0.21,
+      "learning_rate": 4.6470654458120686e-05,
+      "loss": 1.0614,
+      "step": 1892
+    },
+    {
+      "epoch": 0.21,
+      "learning_rate": 4.646877814470129e-05,
+      "loss": 0.9422,
+      "step": 1893
+    },
+    {
+      "epoch": 0.21,
+      "learning_rate": 4.64669018312819e-05,
+      "loss": 1.0455,
+      "step": 1894
+    },
+    {
+      "epoch": 0.21,
+      "learning_rate": 4.6465025517862505e-05,
+      "loss": 0.7479,
+      "step": 1895
+    },
+    {
+      "epoch": 0.21,
+      "learning_rate": 4.646314920444311e-05,
+      "loss": 0.8461,
+      "step": 1896
+    },
+    {
+      "epoch": 0.21,
+      "learning_rate": 4.646127289102372e-05,
+      "loss": 0.5912,
+      "step": 1897
+    },
+    {
+      "epoch": 0.21,
+      "learning_rate": 4.645939657760432e-05,
+      "loss": 0.6915,
+      "step": 1898
+    },
+    {
+      "epoch": 0.21,
+      "learning_rate": 4.6457520264184936e-05,
+      "loss": 0.6906,
+      "step": 1899
+    },
+    {
+      "epoch": 0.21,
+      "learning_rate": 4.6455643950765536e-05,
+      "loss": 0.7078,
+      "step": 1900
+    },
+    {
+      "epoch": 0.21,
+      "learning_rate": 4.645376763734615e-05,
+      "loss": 0.9341,
+      "step": 1901
+    },
+    {
+      "epoch": 0.21,
+      "learning_rate": 4.645189132392675e-05,
+      "loss": 0.8112,
+      "step": 1902
+    },
+    {
+      "epoch": 0.21,
+      "learning_rate": 4.645001501050736e-05,
+      "loss": 0.8639,
+      "step": 1903
+    },
+    {
+      "epoch": 0.21,
+      "learning_rate": 4.644813869708796e-05,
+      "loss": 1.119,
+      "step": 1904
+    },
+    {
+      "epoch": 0.21,
+      "learning_rate": 4.644626238366857e-05,
+      "loss": 1.0528,
+      "step": 1905
+    },
+    {
+      "epoch": 0.21,
+      "learning_rate": 4.644438607024917e-05,
+      "loss": 0.9822,
+      "step": 1906
+    },
+    {
+      "epoch": 0.21,
+      "learning_rate": 4.6442509756829786e-05,
+      "loss": 1.038,
+      "step": 1907
+    },
+    {
+      "epoch": 0.21,
+      "learning_rate": 4.644063344341039e-05,
+      "loss": 0.8811,
+      "step": 1908
+    },
+    {
+      "epoch": 0.21,
+      "learning_rate": 4.6438757129991e-05,
+      "loss": 0.7731,
+      "step": 1909
+    },
+    {
+      "epoch": 0.21,
+      "learning_rate": 4.6436880816571604e-05,
+      "loss": 0.8913,
+      "step": 1910
+    },
+    {
+      "epoch": 0.22,
+      "learning_rate": 4.643500450315221e-05,
+      "loss": 1.2326,
+      "step": 1911
+    },
+    {
+      "epoch": 0.22,
+      "learning_rate": 4.6433128189732816e-05,
+      "loss": 0.7175,
+      "step": 1912
+    },
+    {
+      "epoch": 0.22,
+      "learning_rate": 4.643125187631342e-05,
+      "loss": 1.0018,
+      "step": 1913
+    },
+    {
+      "epoch": 0.22,
+      "learning_rate": 4.642937556289403e-05,
+      "loss": 0.7303,
+      "step": 1914
+    },
+    {
+      "epoch": 0.22,
+      "learning_rate": 4.6427499249474635e-05,
+      "loss": 0.983,
+      "step": 1915
+    },
+    {
+      "epoch": 0.22,
+      "learning_rate": 4.642562293605524e-05,
+      "loss": 0.6231,
+      "step": 1916
+    },
+    {
+      "epoch": 0.22,
+      "learning_rate": 4.642374662263585e-05,
+      "loss": 0.8752,
+      "step": 1917
+    },
+    {
+      "epoch": 0.22,
+      "learning_rate": 4.642187030921645e-05,
+      "loss": 0.8189,
+      "step": 1918
+    },
+    {
+      "epoch": 0.22,
+      "learning_rate": 4.641999399579706e-05,
+      "loss": 0.5936,
+      "step": 1919
+    },
+    {
+      "epoch": 0.22,
+      "learning_rate": 4.6418117682377666e-05,
+      "loss": 0.7311,
+      "step": 1920
+    },
+    {
+      "epoch": 0.22,
+      "learning_rate": 4.641624136895827e-05,
+      "loss": 0.7717,
+      "step": 1921
+    },
+    {
+      "epoch": 0.22,
+      "learning_rate": 4.641436505553888e-05,
+      "loss": 0.7604,
+      "step": 1922
+    },
+    {
+      "epoch": 0.22,
+      "learning_rate": 4.6412488742119484e-05,
+      "loss": 0.8956,
+      "step": 1923
+    },
+    {
+      "epoch": 0.22,
+      "learning_rate": 4.641061242870009e-05,
+      "loss": 0.8509,
+      "step": 1924
+    },
+    {
+      "epoch": 0.22,
+      "learning_rate": 4.6408736115280696e-05,
+      "loss": 0.6271,
+      "step": 1925
+    },
+    {
+      "epoch": 0.22,
+      "learning_rate": 4.640685980186131e-05,
+      "loss": 0.9268,
+      "step": 1926
+    },
+    {
+      "epoch": 0.22,
+      "learning_rate": 4.640498348844191e-05,
+      "loss": 1.1174,
+      "step": 1927
+    },
+    {
+      "epoch": 0.22,
+      "learning_rate": 4.640310717502252e-05,
+      "loss": 0.651,
+      "step": 1928
+    },
+    {
+      "epoch": 0.22,
+      "learning_rate": 4.640123086160312e-05,
+      "loss": 0.727,
+      "step": 1929
+    },
+    {
+      "epoch": 0.22,
+      "learning_rate": 4.6399354548183734e-05,
+      "loss": 1.09,
+      "step": 1930
+    },
+    {
+      "epoch": 0.22,
+      "learning_rate": 4.639747823476433e-05,
+      "loss": 0.808,
+      "step": 1931
+    },
+    {
+      "epoch": 0.22,
+      "learning_rate": 4.6395601921344946e-05,
+      "loss": 0.5748,
+      "step": 1932
+    },
+    {
+      "epoch": 0.22,
+      "learning_rate": 4.6393725607925546e-05,
+      "loss": 0.8565,
+      "step": 1933
+    },
+    {
+      "epoch": 0.22,
+      "learning_rate": 4.639184929450616e-05,
+      "loss": 0.9,
+      "step": 1934
+    },
+    {
+      "epoch": 0.22,
+      "learning_rate": 4.6389972981086765e-05,
+      "loss": 0.86,
+      "step": 1935
+    },
+    {
+      "epoch": 0.22,
+      "learning_rate": 4.638809666766737e-05,
+      "loss": 0.837,
+      "step": 1936
+    },
+    {
+      "epoch": 0.22,
+      "learning_rate": 4.638622035424798e-05,
+      "loss": 0.8581,
+      "step": 1937
+    },
+    {
+      "epoch": 0.22,
+      "learning_rate": 4.638434404082858e-05,
+      "loss": 0.6104,
+      "step": 1938
+    },
+    {
+      "epoch": 0.22,
+      "learning_rate": 4.638246772740919e-05,
+      "loss": 1.0226,
+      "step": 1939
+    },
+    {
+      "epoch": 0.22,
+      "learning_rate": 4.6380591413989796e-05,
+      "loss": 0.7262,
+      "step": 1940
+    },
+    {
+      "epoch": 0.22,
+      "learning_rate": 4.63787151005704e-05,
+      "loss": 0.6409,
+      "step": 1941
+    },
+    {
+      "epoch": 0.22,
+      "learning_rate": 4.637683878715101e-05,
+      "loss": 0.8026,
+      "step": 1942
+    },
+    {
+      "epoch": 0.22,
+      "learning_rate": 4.6374962473731614e-05,
+      "loss": 0.7446,
+      "step": 1943
+    },
+    {
+      "epoch": 0.22,
+      "learning_rate": 4.637308616031222e-05,
+      "loss": 0.7252,
+      "step": 1944
+    },
+    {
+      "epoch": 0.22,
+      "learning_rate": 4.6371209846892826e-05,
+      "loss": 0.6414,
+      "step": 1945
+    },
+    {
+      "epoch": 0.22,
+      "learning_rate": 4.636933353347343e-05,
+      "loss": 1.1155,
+      "step": 1946
+    },
+    {
+      "epoch": 0.22,
+      "learning_rate": 4.636745722005404e-05,
+      "loss": 0.8739,
+      "step": 1947
+    },
+    {
+      "epoch": 0.22,
+      "learning_rate": 4.6365580906634645e-05,
+      "loss": 1.1591,
+      "step": 1948
+    },
+    {
+      "epoch": 0.22,
+      "learning_rate": 4.636370459321525e-05,
+      "loss": 0.957,
+      "step": 1949
+    },
+    {
+      "epoch": 0.22,
+      "learning_rate": 4.636182827979586e-05,
+      "loss": 0.7736,
+      "step": 1950
+    },
+    {
+      "epoch": 0.22,
+      "learning_rate": 4.635995196637646e-05,
+      "loss": 0.6722,
+      "step": 1951
+    },
+    {
+      "epoch": 0.22,
+      "learning_rate": 4.635807565295707e-05,
+      "loss": 0.8452,
+      "step": 1952
+    },
+    {
+      "epoch": 0.22,
+      "learning_rate": 4.6356199339537676e-05,
+      "loss": 0.7393,
+      "step": 1953
+    },
+    {
+      "epoch": 0.22,
+      "learning_rate": 4.635432302611828e-05,
+      "loss": 0.7709,
+      "step": 1954
+    },
+    {
+      "epoch": 0.22,
+      "learning_rate": 4.6352446712698895e-05,
+      "loss": 0.8239,
+      "step": 1955
+    },
+    {
+      "epoch": 0.22,
+      "learning_rate": 4.6350570399279494e-05,
+      "loss": 0.7828,
+      "step": 1956
+    },
+    {
+      "epoch": 0.22,
+      "learning_rate": 4.634869408586011e-05,
+      "loss": 0.7354,
+      "step": 1957
+    },
+    {
+      "epoch": 0.22,
+      "learning_rate": 4.6346817772440707e-05,
+      "loss": 0.8331,
+      "step": 1958
+    },
+    {
+      "epoch": 0.22,
+      "learning_rate": 4.634494145902132e-05,
+      "loss": 0.7014,
+      "step": 1959
+    },
+    {
+      "epoch": 0.22,
+      "learning_rate": 4.634306514560192e-05,
+      "loss": 0.7559,
+      "step": 1960
+    },
+    {
+      "epoch": 0.22,
+      "learning_rate": 4.634118883218253e-05,
+      "loss": 1.0001,
+      "step": 1961
+    },
+    {
+      "epoch": 0.22,
+      "learning_rate": 4.633931251876313e-05,
+      "loss": 0.8047,
+      "step": 1962
+    },
+    {
+      "epoch": 0.22,
+      "learning_rate": 4.6337436205343744e-05,
+      "loss": 0.9574,
+      "step": 1963
+    },
+    {
+      "epoch": 0.22,
+      "learning_rate": 4.633555989192435e-05,
+      "loss": 1.3116,
+      "step": 1964
+    },
+    {
+      "epoch": 0.22,
+      "learning_rate": 4.6333683578504956e-05,
+      "loss": 0.7672,
+      "step": 1965
+    },
+    {
+      "epoch": 0.22,
+      "learning_rate": 4.633180726508556e-05,
+      "loss": 0.9874,
+      "step": 1966
+    },
+    {
+      "epoch": 0.22,
+      "learning_rate": 4.632993095166617e-05,
+      "loss": 0.867,
+      "step": 1967
+    },
+    {
+      "epoch": 0.22,
+      "learning_rate": 4.6328054638246775e-05,
+      "loss": 0.8476,
+      "step": 1968
+    },
+    {
+      "epoch": 0.22,
+      "learning_rate": 4.632617832482738e-05,
+      "loss": 0.8108,
+      "step": 1969
+    },
+    {
+      "epoch": 0.22,
+      "learning_rate": 4.632430201140799e-05,
+      "loss": 0.9072,
+      "step": 1970
+    },
+    {
+      "epoch": 0.22,
+      "learning_rate": 4.632242569798859e-05,
+      "loss": 0.6676,
+      "step": 1971
+    },
+    {
+      "epoch": 0.22,
+      "learning_rate": 4.63205493845692e-05,
+      "loss": 0.9107,
+      "step": 1972
+    },
+    {
+      "epoch": 0.22,
+      "learning_rate": 4.631867307114981e-05,
+      "loss": 0.7405,
+      "step": 1973
+    },
+    {
+      "epoch": 0.22,
+      "learning_rate": 4.631679675773041e-05,
+      "loss": 0.9837,
+      "step": 1974
+    },
+    {
+      "epoch": 0.22,
+      "learning_rate": 4.6314920444311025e-05,
+      "loss": 0.7695,
+      "step": 1975
+    },
+    {
+      "epoch": 0.22,
+      "learning_rate": 4.6313044130891624e-05,
+      "loss": 0.8668,
+      "step": 1976
+    },
+    {
+      "epoch": 0.22,
+      "learning_rate": 4.631116781747224e-05,
+      "loss": 0.8589,
+      "step": 1977
+    },
+    {
+      "epoch": 0.22,
+      "learning_rate": 4.6309291504052837e-05,
+      "loss": 0.7893,
+      "step": 1978
+    },
+    {
+      "epoch": 0.22,
+      "learning_rate": 4.630741519063345e-05,
+      "loss": 0.8557,
+      "step": 1979
+    },
+    {
+      "epoch": 0.22,
+      "learning_rate": 4.630553887721405e-05,
+      "loss": 1.0667,
+      "step": 1980
+    },
+    {
+      "epoch": 0.22,
+      "learning_rate": 4.630366256379466e-05,
+      "loss": 0.9217,
+      "step": 1981
+    },
+    {
+      "epoch": 0.22,
+      "learning_rate": 4.630178625037527e-05,
+      "loss": 0.6812,
+      "step": 1982
+    },
+    {
+      "epoch": 0.22,
+      "learning_rate": 4.6299909936955874e-05,
+      "loss": 0.8155,
+      "step": 1983
+    },
+    {
+      "epoch": 0.22,
+      "learning_rate": 4.629803362353648e-05,
+      "loss": 0.5459,
+      "step": 1984
+    },
+    {
+      "epoch": 0.22,
+      "learning_rate": 4.6296157310117086e-05,
+      "loss": 0.7174,
+      "step": 1985
+    },
+    {
+      "epoch": 0.22,
+      "learning_rate": 4.629428099669769e-05,
+      "loss": 0.867,
+      "step": 1986
+    },
+    {
+      "epoch": 0.22,
+      "learning_rate": 4.62924046832783e-05,
+      "loss": 0.8662,
+      "step": 1987
+    },
+    {
+      "epoch": 0.22,
+      "learning_rate": 4.6290528369858905e-05,
+      "loss": 0.8378,
+      "step": 1988
+    },
+    {
+      "epoch": 0.22,
+      "learning_rate": 4.6288652056439504e-05,
+      "loss": 0.932,
+      "step": 1989
+    },
+    {
+      "epoch": 0.22,
+      "learning_rate": 4.628677574302012e-05,
+      "loss": 0.8364,
+      "step": 1990
+    },
+    {
+      "epoch": 0.22,
+      "learning_rate": 4.628489942960072e-05,
+      "loss": 0.7592,
+      "step": 1991
+    },
+    {
+      "epoch": 0.22,
+      "learning_rate": 4.628302311618133e-05,
+      "loss": 0.4678,
+      "step": 1992
+    },
+    {
+      "epoch": 0.22,
+      "learning_rate": 4.6281146802761936e-05,
+      "loss": 0.6394,
+      "step": 1993
+    },
+    {
+      "epoch": 0.22,
+      "learning_rate": 4.627927048934254e-05,
+      "loss": 0.7476,
+      "step": 1994
+    },
+    {
+      "epoch": 0.22,
+      "learning_rate": 4.627739417592315e-05,
+      "loss": 0.9214,
+      "step": 1995
+    },
+    {
+      "epoch": 0.22,
+      "learning_rate": 4.6275517862503754e-05,
+      "loss": 0.907,
+      "step": 1996
+    },
+    {
+      "epoch": 0.22,
+      "learning_rate": 4.627364154908436e-05,
+      "loss": 0.9182,
+      "step": 1997
+    },
+    {
+      "epoch": 0.22,
+      "learning_rate": 4.6271765235664967e-05,
+      "loss": 0.8543,
+      "step": 1998
+    },
+    {
+      "epoch": 0.22,
+      "learning_rate": 4.626988892224557e-05,
+      "loss": 0.8437,
+      "step": 1999
+    },
+    {
+      "epoch": 0.23,
+      "learning_rate": 4.626801260882618e-05,
+      "loss": 0.8062,
+      "step": 2000
+    },
+    {
+      "epoch": 0.23,
+      "eval_loss": 0.8281100988388062,
+      "eval_runtime": 23.2743,
+      "eval_samples_per_second": 43.954,
+      "eval_steps_per_second": 21.999,
+      "step": 2000
+    },
+    {
+      "epoch": 0.23,
+      "learning_rate": 4.6266136295406785e-05,
+      "loss": 0.6139,
+      "step": 2001
+    },
+    {
+      "epoch": 0.23,
+      "learning_rate": 4.62642599819874e-05,
+      "loss": 0.8488,
+      "step": 2002
+    },
+    {
+      "epoch": 0.23,
+      "learning_rate": 4.6262383668568e-05,
+      "loss": 0.7942,
+      "step": 2003
+    },
+    {
+      "epoch": 0.23,
+      "learning_rate": 4.626050735514861e-05,
+      "loss": 0.7327,
+      "step": 2004
+    },
+    {
+      "epoch": 0.23,
+      "learning_rate": 4.625863104172921e-05,
+      "loss": 0.7569,
+      "step": 2005
+    },
+    {
+      "epoch": 0.23,
+      "learning_rate": 4.625675472830982e-05,
+      "loss": 0.5993,
+      "step": 2006
+    },
+    {
+      "epoch": 0.23,
+      "learning_rate": 4.625487841489042e-05,
+      "loss": 0.6663,
+      "step": 2007
+    },
+    {
+      "epoch": 0.23,
+      "learning_rate": 4.6253002101471035e-05,
+      "loss": 0.7085,
+      "step": 2008
+    },
+    {
+      "epoch": 0.23,
+      "learning_rate": 4.6251125788051634e-05,
+      "loss": 0.8884,
+      "step": 2009
+    },
+    {
+      "epoch": 0.23,
+      "learning_rate": 4.624924947463225e-05,
+      "loss": 1.4176,
+      "step": 2010
+    },
+    {
+      "epoch": 0.23,
+      "learning_rate": 4.624737316121285e-05,
+      "loss": 0.874,
+      "step": 2011
+    },
+    {
+      "epoch": 0.23,
+      "learning_rate": 4.624549684779346e-05,
+      "loss": 0.6307,
+      "step": 2012
+    },
+    {
+      "epoch": 0.23,
+      "learning_rate": 4.6243620534374066e-05,
+      "loss": 0.8501,
+      "step": 2013
+    },
+    {
+      "epoch": 0.23,
+      "learning_rate": 4.624174422095467e-05,
+      "loss": 0.8217,
+      "step": 2014
+    },
+    {
+      "epoch": 0.23,
+      "learning_rate": 4.623986790753528e-05,
+      "loss": 0.7514,
+      "step": 2015
+    },
+    {
+      "epoch": 0.23,
+      "learning_rate": 4.6237991594115884e-05,
+      "loss": 0.9401,
+      "step": 2016
+    },
+    {
+      "epoch": 0.23,
+      "learning_rate": 4.623611528069649e-05,
+      "loss": 0.7676,
+      "step": 2017
+    },
+    {
+      "epoch": 0.23,
+      "learning_rate": 4.6234238967277097e-05,
+      "loss": 0.9181,
+      "step": 2018
+    },
+    {
+      "epoch": 0.23,
+      "learning_rate": 4.62323626538577e-05,
+      "loss": 0.8305,
+      "step": 2019
+    },
+    {
+      "epoch": 0.23,
+      "learning_rate": 4.623048634043831e-05,
+      "loss": 0.8189,
+      "step": 2020
+    },
+    {
+      "epoch": 0.23,
+      "learning_rate": 4.6228610027018915e-05,
+      "loss": 1.2006,
+      "step": 2021
+    },
+    {
+      "epoch": 0.23,
+      "learning_rate": 4.622673371359952e-05,
+      "loss": 0.8467,
+      "step": 2022
+    },
+    {
+      "epoch": 0.23,
+      "learning_rate": 4.622485740018013e-05,
+      "loss": 0.7731,
+      "step": 2023
+    },
+    {
+      "epoch": 0.23,
+      "learning_rate": 4.6222981086760733e-05,
+      "loss": 0.5928,
+      "step": 2024
+    },
+    {
+      "epoch": 0.23,
+      "learning_rate": 4.622110477334134e-05,
+      "loss": 0.7625,
+      "step": 2025
+    },
+    {
+      "epoch": 0.23,
+      "learning_rate": 4.6219228459921946e-05,
+      "loss": 0.6906,
+      "step": 2026
+    },
+    {
+      "epoch": 0.23,
+      "learning_rate": 4.621735214650255e-05,
+      "loss": 0.7122,
+      "step": 2027
+    },
+    {
+      "epoch": 0.23,
+      "learning_rate": 4.621547583308316e-05,
+      "loss": 0.7246,
+      "step": 2028
+    },
+    {
+      "epoch": 0.23,
+      "learning_rate": 4.621359951966377e-05,
+      "loss": 0.6637,
+      "step": 2029
+    },
+    {
+      "epoch": 0.23,
+      "learning_rate": 4.621172320624437e-05,
+      "loss": 0.6357,
+      "step": 2030
+    },
+    {
+      "epoch": 0.23,
+      "learning_rate": 4.620984689282498e-05,
+      "loss": 0.8028,
+      "step": 2031
+    },
+    {
+      "epoch": 0.23,
+      "learning_rate": 4.620797057940558e-05,
+      "loss": 0.619,
+      "step": 2032
+    },
+    {
+      "epoch": 0.23,
+      "learning_rate": 4.6206094265986196e-05,
+      "loss": 1.0655,
+      "step": 2033
+    },
+    {
+      "epoch": 0.23,
+      "learning_rate": 4.6204217952566795e-05,
+      "loss": 0.8729,
+      "step": 2034
+    },
+    {
+      "epoch": 0.23,
+      "learning_rate": 4.620234163914741e-05,
+      "loss": 1.0757,
+      "step": 2035
+    },
+    {
+      "epoch": 0.23,
+      "learning_rate": 4.620046532572801e-05,
+      "loss": 0.7954,
+      "step": 2036
+    },
+    {
+      "epoch": 0.23,
+      "learning_rate": 4.619858901230862e-05,
+      "loss": 0.9462,
+      "step": 2037
+    },
+    {
+      "epoch": 0.23,
+      "learning_rate": 4.6196712698889227e-05,
+      "loss": 0.8437,
+      "step": 2038
+    },
+    {
+      "epoch": 0.23,
+      "learning_rate": 4.619483638546983e-05,
+      "loss": 0.7878,
+      "step": 2039
+    },
+    {
+      "epoch": 0.23,
+      "learning_rate": 4.619296007205044e-05,
+      "loss": 0.6573,
+      "step": 2040
+    },
+    {
+      "epoch": 0.23,
+      "learning_rate": 4.6191083758631045e-05,
+      "loss": 1.0045,
+      "step": 2041
+    },
+    {
+      "epoch": 0.23,
+      "learning_rate": 4.618920744521165e-05,
+      "loss": 0.7829,
+      "step": 2042
+    },
+    {
+      "epoch": 0.23,
+      "learning_rate": 4.618733113179226e-05,
+      "loss": 0.7588,
+      "step": 2043
+    },
+    {
+      "epoch": 0.23,
+      "learning_rate": 4.6185454818372863e-05,
+      "loss": 0.8973,
+      "step": 2044
+    },
+    {
+      "epoch": 0.23,
+      "learning_rate": 4.618357850495347e-05,
+      "loss": 0.7733,
+      "step": 2045
+    },
+    {
+      "epoch": 0.23,
+      "learning_rate": 4.6181702191534076e-05,
+      "loss": 1.1289,
+      "step": 2046
+    },
+    {
+      "epoch": 0.23,
+      "learning_rate": 4.617982587811468e-05,
+      "loss": 0.8285,
+      "step": 2047
+    },
+    {
+      "epoch": 0.23,
+      "learning_rate": 4.617794956469529e-05,
+      "loss": 0.7812,
+      "step": 2048
+    },
+    {
+      "epoch": 0.23,
+      "learning_rate": 4.6176073251275894e-05,
+      "loss": 0.8579,
+      "step": 2049
+    },
+    {
+      "epoch": 0.23,
+      "learning_rate": 4.61741969378565e-05,
+      "loss": 0.6736,
+      "step": 2050
+    },
+    {
+      "epoch": 0.23,
+      "learning_rate": 4.6172320624437107e-05,
+      "loss": 0.7086,
+      "step": 2051
+    },
+    {
+      "epoch": 0.23,
+      "learning_rate": 4.617044431101771e-05,
+      "loss": 0.7578,
+      "step": 2052
+    },
+    {
+      "epoch": 0.23,
+      "learning_rate": 4.616856799759832e-05,
+      "loss": 0.8473,
+      "step": 2053
+    },
+    {
+      "epoch": 0.23,
+      "learning_rate": 4.6166691684178925e-05,
+      "loss": 0.8553,
+      "step": 2054
+    },
+    {
+      "epoch": 0.23,
+      "learning_rate": 4.616481537075953e-05,
+      "loss": 0.9135,
+      "step": 2055
+    },
+    {
+      "epoch": 0.23,
+      "learning_rate": 4.616293905734014e-05,
+      "loss": 0.5432,
+      "step": 2056
+    },
+    {
+      "epoch": 0.23,
+      "learning_rate": 4.6161062743920744e-05,
+      "loss": 0.9244,
+      "step": 2057
+    },
+    {
+      "epoch": 0.23,
+      "learning_rate": 4.6159186430501357e-05,
+      "loss": 0.7177,
+      "step": 2058
+    },
+    {
+      "epoch": 0.23,
+      "learning_rate": 4.6157310117081956e-05,
+      "loss": 1.0168,
+      "step": 2059
+    },
+    {
+      "epoch": 0.23,
+      "learning_rate": 4.615543380366257e-05,
+      "loss": 0.7755,
+      "step": 2060
+    },
+    {
+      "epoch": 0.23,
+      "learning_rate": 4.615355749024317e-05,
+      "loss": 0.7352,
+      "step": 2061
+    },
+    {
+      "epoch": 0.23,
+      "learning_rate": 4.615168117682378e-05,
+      "loss": 0.8613,
+      "step": 2062
+    },
+    {
+      "epoch": 0.23,
+      "learning_rate": 4.614980486340438e-05,
+      "loss": 0.7062,
+      "step": 2063
+    },
+    {
+      "epoch": 0.23,
+      "learning_rate": 4.6147928549984993e-05,
+      "loss": 0.9787,
+      "step": 2064
+    },
+    {
+      "epoch": 0.23,
+      "learning_rate": 4.614605223656559e-05,
+      "loss": 0.9314,
+      "step": 2065
+    },
+    {
+      "epoch": 0.23,
+      "learning_rate": 4.6144175923146206e-05,
+      "loss": 0.6936,
+      "step": 2066
+    },
+    {
+      "epoch": 0.23,
+      "learning_rate": 4.614229960972681e-05,
+      "loss": 0.6473,
+      "step": 2067
+    },
+    {
+      "epoch": 0.23,
+      "learning_rate": 4.614042329630742e-05,
+      "loss": 0.807,
+      "step": 2068
+    },
+    {
+      "epoch": 0.23,
+      "learning_rate": 4.6138546982888024e-05,
+      "loss": 0.9837,
+      "step": 2069
+    },
+    {
+      "epoch": 0.23,
+      "learning_rate": 4.613667066946863e-05,
+      "loss": 0.7832,
+      "step": 2070
+    },
+    {
+      "epoch": 0.23,
+      "learning_rate": 4.6134794356049237e-05,
+      "loss": 0.6913,
+      "step": 2071
+    },
+    {
+      "epoch": 0.23,
+      "learning_rate": 4.613291804262984e-05,
+      "loss": 0.6961,
+      "step": 2072
+    },
+    {
+      "epoch": 0.23,
+      "learning_rate": 4.613104172921045e-05,
+      "loss": 1.0639,
+      "step": 2073
+    },
+    {
+      "epoch": 0.23,
+      "learning_rate": 4.6129165415791055e-05,
+      "loss": 0.6812,
+      "step": 2074
+    },
+    {
+      "epoch": 0.23,
+      "learning_rate": 4.612728910237166e-05,
+      "loss": 1.6729,
+      "step": 2075
+    },
+    {
+      "epoch": 0.23,
+      "learning_rate": 4.6125412788952274e-05,
+      "loss": 0.776,
+      "step": 2076
+    },
+    {
+      "epoch": 0.23,
+      "learning_rate": 4.6123536475532874e-05,
+      "loss": 0.9295,
+      "step": 2077
+    },
+    {
+      "epoch": 0.23,
+      "learning_rate": 4.6121660162113487e-05,
+      "loss": 1.0518,
+      "step": 2078
+    },
+    {
+      "epoch": 0.23,
+      "learning_rate": 4.6119783848694086e-05,
+      "loss": 0.6596,
+      "step": 2079
+    },
+    {
+      "epoch": 0.23,
+      "learning_rate": 4.61179075352747e-05,
+      "loss": 0.9759,
+      "step": 2080
+    },
+    {
+      "epoch": 0.23,
+      "learning_rate": 4.61160312218553e-05,
+      "loss": 0.7469,
+      "step": 2081
+    },
+    {
+      "epoch": 0.23,
+      "learning_rate": 4.611415490843591e-05,
+      "loss": 0.7507,
+      "step": 2082
+    },
+    {
+      "epoch": 0.23,
+      "learning_rate": 4.611227859501651e-05,
+      "loss": 0.7314,
+      "step": 2083
+    },
+    {
+      "epoch": 0.23,
+      "learning_rate": 4.6110402281597123e-05,
+      "loss": 0.6094,
+      "step": 2084
+    },
+    {
+      "epoch": 0.23,
+      "learning_rate": 4.610852596817773e-05,
+      "loss": 1.078,
+      "step": 2085
+    },
+    {
+      "epoch": 0.23,
+      "learning_rate": 4.6106649654758336e-05,
+      "loss": 0.9751,
+      "step": 2086
+    },
+    {
+      "epoch": 0.23,
+      "learning_rate": 4.610477334133894e-05,
+      "loss": 0.9427,
+      "step": 2087
+    },
+    {
+      "epoch": 0.23,
+      "learning_rate": 4.610289702791955e-05,
+      "loss": 0.7971,
+      "step": 2088
+    },
+    {
+      "epoch": 0.24,
+      "learning_rate": 4.6101020714500154e-05,
+      "loss": 0.6541,
+      "step": 2089
+    },
+    {
+      "epoch": 0.24,
+      "learning_rate": 4.609914440108076e-05,
+      "loss": 0.703,
+      "step": 2090
+    },
+    {
+      "epoch": 0.24,
+      "learning_rate": 4.6097268087661367e-05,
+      "loss": 0.8014,
+      "step": 2091
+    },
+    {
+      "epoch": 0.24,
+      "learning_rate": 4.609539177424197e-05,
+      "loss": 0.7542,
+      "step": 2092
+    },
+    {
+      "epoch": 0.24,
+      "learning_rate": 4.609351546082258e-05,
+      "loss": 0.8642,
+      "step": 2093
+    },
+    {
+      "epoch": 0.24,
+      "learning_rate": 4.609163914740318e-05,
+      "loss": 0.8315,
+      "step": 2094
+    },
+    {
+      "epoch": 0.24,
+      "learning_rate": 4.608976283398379e-05,
+      "loss": 0.8837,
+      "step": 2095
+    },
+    {
+      "epoch": 0.24,
+      "learning_rate": 4.60878865205644e-05,
+      "loss": 0.6935,
+      "step": 2096
+    },
+    {
+      "epoch": 0.24,
+      "learning_rate": 4.6086010207145004e-05,
+      "loss": 0.8849,
+      "step": 2097
+    },
+    {
+      "epoch": 0.24,
+      "learning_rate": 4.608413389372561e-05,
+      "loss": 0.7817,
+      "step": 2098
+    },
+    {
+      "epoch": 0.24,
+      "learning_rate": 4.6082257580306216e-05,
+      "loss": 1.3052,
+      "step": 2099
+    },
+    {
+      "epoch": 0.24,
+      "learning_rate": 4.608038126688682e-05,
+      "loss": 0.7476,
+      "step": 2100
+    },
+    {
+      "epoch": 0.24,
+      "learning_rate": 4.607850495346743e-05,
+      "loss": 1.0755,
+      "step": 2101
+    },
+    {
+      "epoch": 0.24,
+      "learning_rate": 4.6076628640048034e-05,
+      "loss": 0.8254,
+      "step": 2102
+    },
+    {
+      "epoch": 0.24,
+      "learning_rate": 4.607475232662864e-05,
+      "loss": 0.9928,
+      "step": 2103
+    },
+    {
+      "epoch": 0.24,
+      "learning_rate": 4.607287601320925e-05,
+      "loss": 0.686,
+      "step": 2104
+    },
+    {
+      "epoch": 0.24,
+      "learning_rate": 4.607099969978986e-05,
+      "loss": 0.8907,
+      "step": 2105
+    },
+    {
+      "epoch": 0.24,
+      "learning_rate": 4.606912338637046e-05,
+      "loss": 0.6983,
+      "step": 2106
+    },
+    {
+      "epoch": 0.24,
+      "learning_rate": 4.606724707295107e-05,
+      "loss": 0.8782,
+      "step": 2107
+    },
+    {
+      "epoch": 0.24,
+      "learning_rate": 4.606537075953167e-05,
+      "loss": 0.9866,
+      "step": 2108
+    },
+    {
+      "epoch": 0.24,
+      "learning_rate": 4.6063494446112284e-05,
+      "loss": 0.6945,
+      "step": 2109
+    },
+    {
+      "epoch": 0.24,
+      "learning_rate": 4.6061618132692884e-05,
+      "loss": 0.7608,
+      "step": 2110
+    },
+    {
+      "epoch": 0.24,
+      "learning_rate": 4.6059741819273497e-05,
+      "loss": 0.8478,
+      "step": 2111
+    },
+    {
+      "epoch": 0.24,
+      "learning_rate": 4.6057865505854096e-05,
+      "loss": 0.8837,
+      "step": 2112
+    },
+    {
+      "epoch": 0.24,
+      "learning_rate": 4.605598919243471e-05,
+      "loss": 0.8421,
+      "step": 2113
+    },
+    {
+      "epoch": 0.24,
+      "learning_rate": 4.6054112879015315e-05,
+      "loss": 0.88,
+      "step": 2114
+    },
+    {
+      "epoch": 0.24,
+      "learning_rate": 4.605223656559592e-05,
+      "loss": 0.7255,
+      "step": 2115
+    },
+    {
+      "epoch": 0.24,
+      "learning_rate": 4.605036025217653e-05,
+      "loss": 0.7973,
+      "step": 2116
+    },
+    {
+      "epoch": 0.24,
+      "learning_rate": 4.6048483938757134e-05,
+      "loss": 0.8804,
+      "step": 2117
+    },
+    {
+      "epoch": 0.24,
+      "learning_rate": 4.604660762533774e-05,
+      "loss": 0.9517,
+      "step": 2118
+    },
+    {
+      "epoch": 0.24,
+      "learning_rate": 4.6044731311918346e-05,
+      "loss": 0.6882,
+      "step": 2119
+    },
+    {
+      "epoch": 0.24,
+      "learning_rate": 4.604285499849895e-05,
+      "loss": 0.9677,
+      "step": 2120
+    },
+    {
+      "epoch": 0.24,
+      "learning_rate": 4.604097868507956e-05,
+      "loss": 0.7759,
+      "step": 2121
+    },
+    {
+      "epoch": 0.24,
+      "learning_rate": 4.6039102371660164e-05,
+      "loss": 0.8101,
+      "step": 2122
+    },
+    {
+      "epoch": 0.24,
+      "learning_rate": 4.603722605824077e-05,
+      "loss": 0.9587,
+      "step": 2123
+    },
+    {
+      "epoch": 0.24,
+      "learning_rate": 4.603534974482138e-05,
+      "loss": 0.6244,
+      "step": 2124
+    },
+    {
+      "epoch": 0.24,
+      "learning_rate": 4.603347343140198e-05,
+      "loss": 0.6575,
+      "step": 2125
+    },
+    {
+      "epoch": 0.24,
+      "eval_loss": 0.8257102370262146,
+      "eval_runtime": 23.2762,
+      "eval_samples_per_second": 43.95,
+      "eval_steps_per_second": 21.997,
+      "step": 2125
+    },
+    {
+      "epoch": 0.24,
+      "learning_rate": 4.603159711798259e-05,
+      "loss": 1.0353,
+      "step": 2126
+    },
+    {
+      "epoch": 0.24,
+      "learning_rate": 4.6029720804563195e-05,
+      "loss": 0.8257,
+      "step": 2127
+    },
+    {
+      "epoch": 0.24,
+      "learning_rate": 4.60278444911438e-05,
+      "loss": 0.845,
+      "step": 2128
+    },
+    {
+      "epoch": 0.24,
+      "learning_rate": 4.602596817772441e-05,
+      "loss": 0.7706,
+      "step": 2129
+    },
+    {
+      "epoch": 0.24,
+      "learning_rate": 4.6024091864305014e-05,
+      "loss": 0.8716,
+      "step": 2130
+    },
+    {
+      "epoch": 0.24,
+      "learning_rate": 4.602221555088562e-05,
+      "loss": 0.7884,
+      "step": 2131
+    },
+    {
+      "epoch": 0.24,
+      "learning_rate": 4.602033923746623e-05,
+      "loss": 0.8251,
+      "step": 2132
+    },
+    {
+      "epoch": 0.24,
+      "learning_rate": 4.601846292404683e-05,
+      "loss": 0.6344,
+      "step": 2133
+    },
+    {
+      "epoch": 0.24,
+      "learning_rate": 4.6016586610627445e-05,
+      "loss": 0.7936,
+      "step": 2134
+    },
+    {
+      "epoch": 0.24,
+      "learning_rate": 4.6014710297208044e-05,
+      "loss": 0.945,
+      "step": 2135
+    },
+    {
+      "epoch": 0.24,
+      "learning_rate": 4.601283398378866e-05,
+      "loss": 0.9155,
+      "step": 2136
+    },
+    {
+      "epoch": 0.24,
+      "learning_rate": 4.601095767036926e-05,
+      "loss": 0.9735,
+      "step": 2137
+    },
+    {
+      "epoch": 0.24,
+      "learning_rate": 4.600908135694987e-05,
+      "loss": 0.7,
+      "step": 2138
+    },
+    {
+      "epoch": 0.24,
+      "learning_rate": 4.600720504353047e-05,
+      "loss": 0.5964,
+      "step": 2139
+    },
+    {
+      "epoch": 0.24,
+      "learning_rate": 4.600532873011108e-05,
+      "loss": 0.828,
+      "step": 2140
+    },
+    {
+      "epoch": 0.24,
+      "learning_rate": 4.600345241669169e-05,
+      "loss": 0.7008,
+      "step": 2141
+    },
+    {
+      "epoch": 0.24,
+      "learning_rate": 4.6001576103272294e-05,
+      "loss": 0.7105,
+      "step": 2142
+    },
+    {
+      "epoch": 0.24,
+      "learning_rate": 4.59996997898529e-05,
+      "loss": 0.7403,
+      "step": 2143
+    },
+    {
+      "epoch": 0.24,
+      "learning_rate": 4.599782347643351e-05,
+      "loss": 1.0066,
+      "step": 2144
+    },
+    {
+      "epoch": 0.24,
+      "learning_rate": 4.599594716301411e-05,
+      "loss": 0.787,
+      "step": 2145
+    },
+    {
+      "epoch": 0.24,
+      "learning_rate": 4.599407084959472e-05,
+      "loss": 1.1061,
+      "step": 2146
+    },
+    {
+      "epoch": 0.24,
+      "learning_rate": 4.5992194536175325e-05,
+      "loss": 0.7822,
+      "step": 2147
+    },
+    {
+      "epoch": 0.24,
+      "learning_rate": 4.599031822275593e-05,
+      "loss": 0.8096,
+      "step": 2148
+    },
+    {
+      "epoch": 0.24,
+      "learning_rate": 4.598844190933654e-05,
+      "loss": 0.8481,
+      "step": 2149
+    },
+    {
+      "epoch": 0.24,
+      "learning_rate": 4.5986565595917144e-05,
+      "loss": 0.8806,
+      "step": 2150
+    },
+    {
+      "epoch": 0.24,
+      "learning_rate": 4.598468928249775e-05,
+      "loss": 0.9633,
+      "step": 2151
+    },
+    {
+      "epoch": 0.24,
+      "learning_rate": 4.5982812969078356e-05,
+      "loss": 0.8224,
+      "step": 2152
+    },
+    {
+      "epoch": 0.24,
+      "learning_rate": 4.598093665565896e-05,
+      "loss": 0.693,
+      "step": 2153
+    },
+    {
+      "epoch": 0.24,
+      "learning_rate": 4.597906034223957e-05,
+      "loss": 0.6928,
+      "step": 2154
+    },
+    {
+      "epoch": 0.24,
+      "learning_rate": 4.5977184028820174e-05,
+      "loss": 0.7929,
+      "step": 2155
+    },
+    {
+      "epoch": 0.24,
+      "learning_rate": 4.597530771540078e-05,
+      "loss": 0.8872,
+      "step": 2156
+    },
+    {
+      "epoch": 0.24,
+      "learning_rate": 4.597343140198139e-05,
+      "loss": 0.8314,
+      "step": 2157
+    },
+    {
+      "epoch": 0.24,
+      "learning_rate": 4.597155508856199e-05,
+      "loss": 0.8042,
+      "step": 2158
+    },
+    {
+      "epoch": 0.24,
+      "learning_rate": 4.59696787751426e-05,
+      "loss": 0.7755,
+      "step": 2159
+    },
+    {
+      "epoch": 0.24,
+      "learning_rate": 4.5967802461723205e-05,
+      "loss": 0.7486,
+      "step": 2160
+    },
+    {
+      "epoch": 0.24,
+      "learning_rate": 4.596592614830382e-05,
+      "loss": 1.1765,
+      "step": 2161
+    },
+    {
+      "epoch": 0.24,
+      "learning_rate": 4.596404983488442e-05,
+      "loss": 0.6851,
+      "step": 2162
+    },
+    {
+      "epoch": 0.24,
+      "learning_rate": 4.596217352146503e-05,
+      "loss": 0.8084,
+      "step": 2163
+    },
+    {
+      "epoch": 0.24,
+      "learning_rate": 4.596029720804563e-05,
+      "loss": 0.932,
+      "step": 2164
+    },
+    {
+      "epoch": 0.24,
+      "learning_rate": 4.595842089462624e-05,
+      "loss": 0.8891,
+      "step": 2165
+    },
+    {
+      "epoch": 0.24,
+      "learning_rate": 4.595654458120684e-05,
+      "loss": 0.7191,
+      "step": 2166
+    },
+    {
+      "epoch": 0.24,
+      "learning_rate": 4.5954668267787455e-05,
+      "loss": 1.0465,
+      "step": 2167
+    },
+    {
+      "epoch": 0.24,
+      "learning_rate": 4.5952791954368055e-05,
+      "loss": 0.9019,
+      "step": 2168
+    },
+    {
+      "epoch": 0.24,
+      "learning_rate": 4.595091564094867e-05,
+      "loss": 0.9205,
+      "step": 2169
+    },
+    {
+      "epoch": 0.24,
+      "learning_rate": 4.5949039327529274e-05,
+      "loss": 0.8898,
+      "step": 2170
+    },
+    {
+      "epoch": 0.24,
+      "learning_rate": 4.594716301410988e-05,
+      "loss": 1.0656,
+      "step": 2171
+    },
+    {
+      "epoch": 0.24,
+      "learning_rate": 4.5945286700690486e-05,
+      "loss": 0.7261,
+      "step": 2172
+    },
+    {
+      "epoch": 0.24,
+      "learning_rate": 4.594341038727109e-05,
+      "loss": 0.9294,
+      "step": 2173
+    },
+    {
+      "epoch": 0.24,
+      "learning_rate": 4.59415340738517e-05,
+      "loss": 0.938,
+      "step": 2174
+    },
+    {
+      "epoch": 0.24,
+      "learning_rate": 4.5939657760432304e-05,
+      "loss": 1.0268,
+      "step": 2175
+    },
+    {
+      "epoch": 0.24,
+      "learning_rate": 4.593778144701291e-05,
+      "loss": 0.869,
+      "step": 2176
+    },
+    {
+      "epoch": 0.24,
+      "learning_rate": 4.593590513359352e-05,
+      "loss": 0.9373,
+      "step": 2177
+    },
+    {
+      "epoch": 0.25,
+      "learning_rate": 4.593402882017412e-05,
+      "loss": 0.8011,
+      "step": 2178
+    },
+    {
+      "epoch": 0.25,
+      "learning_rate": 4.5932152506754736e-05,
+      "loss": 1.0981,
+      "step": 2179
+    },
+    {
+      "epoch": 0.25,
+      "learning_rate": 4.5930276193335335e-05,
+      "loss": 0.7769,
+      "step": 2180
+    },
+    {
+      "epoch": 0.25,
+      "learning_rate": 4.592839987991595e-05,
+      "loss": 0.6895,
+      "step": 2181
+    },
+    {
+      "epoch": 0.25,
+      "learning_rate": 4.592652356649655e-05,
+      "loss": 0.8236,
+      "step": 2182
+    },
+    {
+      "epoch": 0.25,
+      "learning_rate": 4.592464725307716e-05,
+      "loss": 0.7826,
+      "step": 2183
+    },
+    {
+      "epoch": 0.25,
+      "learning_rate": 4.592277093965776e-05,
+      "loss": 0.7964,
+      "step": 2184
+    },
+    {
+      "epoch": 0.25,
+      "learning_rate": 4.592089462623837e-05,
+      "loss": 0.7076,
+      "step": 2185
+    },
+    {
+      "epoch": 0.25,
+      "learning_rate": 4.591901831281897e-05,
+      "loss": 0.6303,
+      "step": 2186
+    },
+    {
+      "epoch": 0.25,
+      "learning_rate": 4.5917141999399585e-05,
+      "loss": 0.6118,
+      "step": 2187
+    },
+    {
+      "epoch": 0.25,
+      "learning_rate": 4.591526568598019e-05,
+      "loss": 0.7234,
+      "step": 2188
+    },
+    {
+      "epoch": 0.25,
+      "learning_rate": 4.59133893725608e-05,
+      "loss": 0.7086,
+      "step": 2189
+    },
+    {
+      "epoch": 0.25,
+      "learning_rate": 4.5911513059141404e-05,
+      "loss": 0.6167,
+      "step": 2190
+    },
+    {
+      "epoch": 0.25,
+      "learning_rate": 4.590963674572201e-05,
+      "loss": 0.7695,
+      "step": 2191
+    },
+    {
+      "epoch": 0.25,
+      "learning_rate": 4.5907760432302616e-05,
+      "loss": 0.7136,
+      "step": 2192
+    },
+    {
+      "epoch": 0.25,
+      "learning_rate": 4.590588411888322e-05,
+      "loss": 0.908,
+      "step": 2193
+    },
+    {
+      "epoch": 0.25,
+      "learning_rate": 4.590400780546383e-05,
+      "loss": 0.8853,
+      "step": 2194
+    },
+    {
+      "epoch": 0.25,
+      "learning_rate": 4.5902131492044434e-05,
+      "loss": 1.0212,
+      "step": 2195
+    },
+    {
+      "epoch": 0.25,
+      "learning_rate": 4.590025517862504e-05,
+      "loss": 0.6,
+      "step": 2196
+    },
+    {
+      "epoch": 0.25,
+      "learning_rate": 4.589837886520565e-05,
+      "loss": 0.8309,
+      "step": 2197
+    },
+    {
+      "epoch": 0.25,
+      "learning_rate": 4.589650255178625e-05,
+      "loss": 0.9437,
+      "step": 2198
+    },
+    {
+      "epoch": 0.25,
+      "learning_rate": 4.589462623836686e-05,
+      "loss": 0.8774,
+      "step": 2199
+    },
+    {
+      "epoch": 0.25,
+      "learning_rate": 4.5892749924947465e-05,
+      "loss": 0.8541,
+      "step": 2200
+    },
+    {
+      "epoch": 0.25,
+      "learning_rate": 4.589087361152807e-05,
+      "loss": 0.8126,
+      "step": 2201
+    },
+    {
+      "epoch": 0.25,
+      "learning_rate": 4.588899729810868e-05,
+      "loss": 0.929,
+      "step": 2202
+    },
+    {
+      "epoch": 0.25,
+      "learning_rate": 4.5887120984689284e-05,
+      "loss": 0.755,
+      "step": 2203
+    },
+    {
+      "epoch": 0.25,
+      "learning_rate": 4.588524467126989e-05,
+      "loss": 0.9042,
+      "step": 2204
+    },
+    {
+      "epoch": 0.25,
+      "learning_rate": 4.5883368357850496e-05,
+      "loss": 0.7536,
+      "step": 2205
+    },
+    {
+      "epoch": 0.25,
+      "learning_rate": 4.58814920444311e-05,
+      "loss": 0.9156,
+      "step": 2206
+    },
+    {
+      "epoch": 0.25,
+      "learning_rate": 4.587961573101171e-05,
+      "loss": 0.8235,
+      "step": 2207
+    },
+    {
+      "epoch": 0.25,
+      "learning_rate": 4.587773941759232e-05,
+      "loss": 0.7286,
+      "step": 2208
+    },
+    {
+      "epoch": 0.25,
+      "learning_rate": 4.587586310417292e-05,
+      "loss": 0.5667,
+      "step": 2209
+    },
+    {
+      "epoch": 0.25,
+      "learning_rate": 4.5873986790753534e-05,
+      "loss": 0.8728,
+      "step": 2210
+    },
+    {
+      "epoch": 0.25,
+      "learning_rate": 4.587211047733413e-05,
+      "loss": 0.693,
+      "step": 2211
+    },
+    {
+      "epoch": 0.25,
+      "learning_rate": 4.5870234163914746e-05,
+      "loss": 0.762,
+      "step": 2212
+    },
+    {
+      "epoch": 0.25,
+      "learning_rate": 4.5868357850495345e-05,
+      "loss": 0.8834,
+      "step": 2213
+    },
+    {
+      "epoch": 0.25,
+      "learning_rate": 4.586648153707596e-05,
+      "loss": 0.7201,
+      "step": 2214
+    },
+    {
+      "epoch": 0.25,
+      "learning_rate": 4.586460522365656e-05,
+      "loss": 0.9333,
+      "step": 2215
+    },
+    {
+      "epoch": 0.25,
+      "learning_rate": 4.586272891023717e-05,
+      "loss": 0.5014,
+      "step": 2216
+    },
+    {
+      "epoch": 0.25,
+      "learning_rate": 4.586085259681778e-05,
+      "loss": 1.1135,
+      "step": 2217
+    },
+    {
+      "epoch": 0.25,
+      "learning_rate": 4.585897628339838e-05,
+      "loss": 0.7606,
+      "step": 2218
+    },
+    {
+      "epoch": 0.25,
+      "learning_rate": 4.585709996997899e-05,
+      "loss": 0.6818,
+      "step": 2219
+    },
+    {
+      "epoch": 0.25,
+      "learning_rate": 4.5855223656559595e-05,
+      "loss": 0.7406,
+      "step": 2220
+    },
+    {
+      "epoch": 0.25,
+      "learning_rate": 4.58533473431402e-05,
+      "loss": 0.9534,
+      "step": 2221
+    },
+    {
+      "epoch": 0.25,
+      "learning_rate": 4.585147102972081e-05,
+      "loss": 0.9177,
+      "step": 2222
+    },
+    {
+      "epoch": 0.25,
+      "learning_rate": 4.5849594716301414e-05,
+      "loss": 0.4775,
+      "step": 2223
+    },
+    {
+      "epoch": 0.25,
+      "learning_rate": 4.584771840288202e-05,
+      "loss": 0.7597,
+      "step": 2224
+    },
+    {
+      "epoch": 0.25,
+      "learning_rate": 4.5845842089462626e-05,
+      "loss": 0.7208,
+      "step": 2225
+    },
+    {
+      "epoch": 0.25,
+      "learning_rate": 4.584396577604323e-05,
+      "loss": 0.7349,
+      "step": 2226
+    },
+    {
+      "epoch": 0.25,
+      "learning_rate": 4.584208946262384e-05,
+      "loss": 0.7168,
+      "step": 2227
+    },
+    {
+      "epoch": 0.25,
+      "learning_rate": 4.5840213149204445e-05,
+      "loss": 0.9315,
+      "step": 2228
+    },
+    {
+      "epoch": 0.25,
+      "learning_rate": 4.583833683578505e-05,
+      "loss": 0.9055,
+      "step": 2229
+    },
+    {
+      "epoch": 0.25,
+      "learning_rate": 4.583646052236566e-05,
+      "loss": 0.6193,
+      "step": 2230
+    },
+    {
+      "epoch": 0.25,
+      "learning_rate": 4.583458420894626e-05,
+      "loss": 0.7789,
+      "step": 2231
+    },
+    {
+      "epoch": 0.25,
+      "learning_rate": 4.583270789552687e-05,
+      "loss": 0.7575,
+      "step": 2232
+    },
+    {
+      "epoch": 0.25,
+      "learning_rate": 4.5830831582107475e-05,
+      "loss": 0.7706,
+      "step": 2233
+    },
+    {
+      "epoch": 0.25,
+      "learning_rate": 4.582895526868808e-05,
+      "loss": 0.7217,
+      "step": 2234
+    },
+    {
+      "epoch": 0.25,
+      "learning_rate": 4.5827078955268694e-05,
+      "loss": 0.9682,
+      "step": 2235
+    },
+    {
+      "epoch": 0.25,
+      "learning_rate": 4.5825202641849294e-05,
+      "loss": 0.8069,
+      "step": 2236
+    },
+    {
+      "epoch": 0.25,
+      "learning_rate": 4.582332632842991e-05,
+      "loss": 0.6232,
+      "step": 2237
+    },
+    {
+      "epoch": 0.25,
+      "learning_rate": 4.5821450015010506e-05,
+      "loss": 0.6922,
+      "step": 2238
+    },
+    {
+      "epoch": 0.25,
+      "learning_rate": 4.581957370159112e-05,
+      "loss": 0.8322,
+      "step": 2239
+    },
+    {
+      "epoch": 0.25,
+      "learning_rate": 4.581769738817172e-05,
+      "loss": 0.8383,
+      "step": 2240
+    },
+    {
+      "epoch": 0.25,
+      "learning_rate": 4.581582107475233e-05,
+      "loss": 0.9319,
+      "step": 2241
+    },
+    {
+      "epoch": 0.25,
+      "learning_rate": 4.581394476133293e-05,
+      "loss": 0.9027,
+      "step": 2242
+    },
+    {
+      "epoch": 0.25,
+      "learning_rate": 4.5812068447913544e-05,
+      "loss": 0.9281,
+      "step": 2243
+    },
+    {
+      "epoch": 0.25,
+      "learning_rate": 4.581019213449414e-05,
+      "loss": 1.044,
+      "step": 2244
+    },
+    {
+      "epoch": 0.25,
+      "learning_rate": 4.5808315821074756e-05,
+      "loss": 0.9875,
+      "step": 2245
+    },
+    {
+      "epoch": 0.25,
+      "learning_rate": 4.580643950765536e-05,
+      "loss": 0.7692,
+      "step": 2246
+    },
+    {
+      "epoch": 0.25,
+      "learning_rate": 4.580456319423597e-05,
+      "loss": 0.8191,
+      "step": 2247
+    },
+    {
+      "epoch": 0.25,
+      "learning_rate": 4.5802686880816575e-05,
+      "loss": 0.8588,
+      "step": 2248
+    },
+    {
+      "epoch": 0.25,
+      "learning_rate": 4.580081056739718e-05,
+      "loss": 0.877,
+      "step": 2249
+    },
+    {
+      "epoch": 0.25,
+      "learning_rate": 4.579893425397779e-05,
+      "loss": 0.559,
+      "step": 2250
+    },
+    {
+      "epoch": 0.25,
+      "eval_loss": 0.8184141516685486,
+      "eval_runtime": 23.7293,
+      "eval_samples_per_second": 43.111,
+      "eval_steps_per_second": 21.577,
+      "step": 2250
+    }
+  ],
+  "logging_steps": 1,
+  "max_steps": 26658,
+  "num_train_epochs": 3,
+  "save_steps": 250,
+  "total_flos": 5.640881702541722e+16,
+  "trial_name": null,
+  "trial_params": null
+}