diff --git "a/trainer_state.json" "b/trainer_state.json"
new file mode 100644--- /dev/null
+++ "b/trainer_state.json"
@@ -0,0 +1,5931 @@
+{
+  "best_metric": null,
+  "best_model_checkpoint": null,
+  "epoch": 0.4967230079337703,
+  "eval_steps": 50,
+  "global_step": 960,
+  "is_hyper_param_search": false,
+  "is_local_process_zero": true,
+  "is_world_process_zero": true,
+  "log_history": [
+    {
+      "epoch": 0.0,
+      "learning_rate": 2.5000000000000004e-07,
+      "loss": 0.3488,
+      "step": 1
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 5.000000000000001e-07,
+      "loss": 0.3961,
+      "step": 2
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 7.5e-07,
+      "loss": 0.3929,
+      "step": 3
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 1.0000000000000002e-06,
+      "loss": 0.3716,
+      "step": 4
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 1.25e-06,
+      "loss": 0.3652,
+      "step": 5
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 1.5e-06,
+      "loss": 0.3658,
+      "step": 6
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 1.75e-06,
+      "loss": 0.3447,
+      "step": 7
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 2.0000000000000003e-06,
+      "loss": 0.3614,
+      "step": 8
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 2.25e-06,
+      "loss": 0.2988,
+      "step": 9
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 2.5e-06,
+      "loss": 0.3748,
+      "step": 10
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 2.7500000000000004e-06,
+      "loss": 0.3553,
+      "step": 11
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 3e-06,
+      "loss": 0.345,
+      "step": 12
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 3.2500000000000002e-06,
+      "loss": 0.3443,
+      "step": 13
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 3.5e-06,
+      "loss": 0.3097,
+      "step": 14
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 3.7500000000000005e-06,
+      "loss": 0.3065,
+      "step": 15
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.000000000000001e-06,
+      "loss": 0.2663,
+      "step": 16
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.25e-06,
+      "loss": 0.3287,
+      "step": 17
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.5e-06,
+      "loss": 0.277,
+      "step": 18
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.75e-06,
+      "loss": 0.3128,
+      "step": 19
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 5e-06,
+      "loss": 0.2824,
+      "step": 20
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.999999165083631e-06,
+      "loss": 0.2868,
+      "step": 21
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.999996660335078e-06,
+      "loss": 0.2892,
+      "step": 22
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.999992485756017e-06,
+      "loss": 0.2768,
+      "step": 23
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.999986641349234e-06,
+      "loss": 0.3163,
+      "step": 24
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.999979127118634e-06,
+      "loss": 0.2892,
+      "step": 25
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.999969943069236e-06,
+      "loss": 0.2936,
+      "step": 26
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.999959089207175e-06,
+      "loss": 0.3267,
+      "step": 27
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.9999465655396985e-06,
+      "loss": 0.2814,
+      "step": 28
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.999932372075173e-06,
+      "loss": 0.286,
+      "step": 29
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.999916508823078e-06,
+      "loss": 0.2761,
+      "step": 30
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.99989897579401e-06,
+      "loss": 0.28,
+      "step": 31
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.999879772999679e-06,
+      "loss": 0.2635,
+      "step": 32
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.999858900452913e-06,
+      "loss": 0.3012,
+      "step": 33
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.9998363581676515e-06,
+      "loss": 0.3157,
+      "step": 34
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.9998121461589514e-06,
+      "loss": 0.2489,
+      "step": 35
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.999786264442985e-06,
+      "loss": 0.2687,
+      "step": 36
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.9997587130370395e-06,
+      "loss": 0.2555,
+      "step": 37
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.999729491959518e-06,
+      "loss": 0.2497,
+      "step": 38
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.999698601229939e-06,
+      "loss": 0.2369,
+      "step": 39
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.999666040868932e-06,
+      "loss": 0.2705,
+      "step": 40
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.999631810898249e-06,
+      "loss": 0.2561,
+      "step": 41
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.999595911340751e-06,
+      "loss": 0.2834,
+      "step": 42
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.999558342220418e-06,
+      "loss": 0.2728,
+      "step": 43
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.999519103562342e-06,
+      "loss": 0.2629,
+      "step": 44
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.9994781953927335e-06,
+      "loss": 0.3022,
+      "step": 45
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.999435617738915e-06,
+      "loss": 0.2425,
+      "step": 46
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.999391370629327e-06,
+      "loss": 0.2988,
+      "step": 47
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.999345454093521e-06,
+      "loss": 0.2346,
+      "step": 48
+    },
+    {
+      "epoch": 0.03,
+      "learning_rate": 4.9992978681621696e-06,
+      "loss": 0.2479,
+      "step": 49
+    },
+    {
+      "epoch": 0.03,
+      "learning_rate": 4.999248612867054e-06,
+      "loss": 0.2482,
+      "step": 50
+    },
+    {
+      "epoch": 0.03,
+      "eval_loss": 0.3235985338687897,
+      "eval_runtime": 117.1064,
+      "eval_samples_per_second": 26.66,
+      "eval_steps_per_second": 8.889,
+      "step": 50
+    },
+    {
+      "epoch": 0.03,
+      "learning_rate": 4.999197688241076e-06,
+      "loss": 0.2523,
+      "step": 51
+    },
+    {
+      "epoch": 0.03,
+      "learning_rate": 4.999145094318249e-06,
+      "loss": 0.2898,
+      "step": 52
+    },
+    {
+      "epoch": 0.03,
+      "learning_rate": 4.999090831133701e-06,
+      "loss": 0.2385,
+      "step": 53
+    },
+    {
+      "epoch": 0.03,
+      "learning_rate": 4.999034898723677e-06,
+      "loss": 0.2673,
+      "step": 54
+    },
+    {
+      "epoch": 0.03,
+      "learning_rate": 4.998977297125536e-06,
+      "loss": 0.257,
+      "step": 55
+    },
+    {
+      "epoch": 0.03,
+      "learning_rate": 4.998918026377753e-06,
+      "loss": 0.2344,
+      "step": 56
+    },
+    {
+      "epoch": 0.03,
+      "learning_rate": 4.998857086519915e-06,
+      "loss": 0.2608,
+      "step": 57
+    },
+    {
+      "epoch": 0.03,
+      "learning_rate": 4.998794477592727e-06,
+      "loss": 0.2471,
+      "step": 58
+    },
+    {
+      "epoch": 0.03,
+      "learning_rate": 4.998730199638008e-06,
+      "loss": 0.2191,
+      "step": 59
+    },
+    {
+      "epoch": 0.03,
+      "learning_rate": 4.99866425269869e-06,
+      "loss": 0.2582,
+      "step": 60
+    },
+    {
+      "epoch": 0.03,
+      "learning_rate": 4.9985966368188225e-06,
+      "loss": 0.2645,
+      "step": 61
+    },
+    {
+      "epoch": 0.03,
+      "learning_rate": 4.998527352043567e-06,
+      "loss": 0.2342,
+      "step": 62
+    },
+    {
+      "epoch": 0.03,
+      "learning_rate": 4.998456398419203e-06,
+      "loss": 0.2615,
+      "step": 63
+    },
+    {
+      "epoch": 0.03,
+      "learning_rate": 4.99838377599312e-06,
+      "loss": 0.2342,
+      "step": 64
+    },
+    {
+      "epoch": 0.03,
+      "learning_rate": 4.998309484813828e-06,
+      "loss": 0.2626,
+      "step": 65
+    },
+    {
+      "epoch": 0.03,
+      "learning_rate": 4.998233524930946e-06,
+      "loss": 0.2303,
+      "step": 66
+    },
+    {
+      "epoch": 0.03,
+      "learning_rate": 4.998155896395212e-06,
+      "loss": 0.2463,
+      "step": 67
+    },
+    {
+      "epoch": 0.04,
+      "learning_rate": 4.9980765992584755e-06,
+      "loss": 0.2393,
+      "step": 68
+    },
+    {
+      "epoch": 0.04,
+      "learning_rate": 4.997995633573702e-06,
+      "loss": 0.2566,
+      "step": 69
+    },
+    {
+      "epoch": 0.04,
+      "learning_rate": 4.997912999394972e-06,
+      "loss": 0.2355,
+      "step": 70
+    },
+    {
+      "epoch": 0.04,
+      "learning_rate": 4.997828696777478e-06,
+      "loss": 0.2166,
+      "step": 71
+    },
+    {
+      "epoch": 0.04,
+      "learning_rate": 4.997742725777529e-06,
+      "loss": 0.2704,
+      "step": 72
+    },
+    {
+      "epoch": 0.04,
+      "learning_rate": 4.997655086452549e-06,
+      "loss": 0.2706,
+      "step": 73
+    },
+    {
+      "epoch": 0.04,
+      "learning_rate": 4.997565778861075e-06,
+      "loss": 0.2327,
+      "step": 74
+    },
+    {
+      "epoch": 0.04,
+      "learning_rate": 4.997474803062756e-06,
+      "loss": 0.2388,
+      "step": 75
+    },
+    {
+      "epoch": 0.04,
+      "learning_rate": 4.997382159118361e-06,
+      "loss": 0.2228,
+      "step": 76
+    },
+    {
+      "epoch": 0.04,
+      "learning_rate": 4.997287847089769e-06,
+      "loss": 0.2532,
+      "step": 77
+    },
+    {
+      "epoch": 0.04,
+      "learning_rate": 4.997191867039973e-06,
+      "loss": 0.2245,
+      "step": 78
+    },
+    {
+      "epoch": 0.04,
+      "learning_rate": 4.997094219033082e-06,
+      "loss": 0.2912,
+      "step": 79
+    },
+    {
+      "epoch": 0.04,
+      "learning_rate": 4.996994903134318e-06,
+      "loss": 0.2234,
+      "step": 80
+    },
+    {
+      "epoch": 0.04,
+      "learning_rate": 4.996893919410018e-06,
+      "loss": 0.2382,
+      "step": 81
+    },
+    {
+      "epoch": 0.04,
+      "learning_rate": 4.996791267927632e-06,
+      "loss": 0.2556,
+      "step": 82
+    },
+    {
+      "epoch": 0.04,
+      "learning_rate": 4.996686948755725e-06,
+      "loss": 0.2353,
+      "step": 83
+    },
+    {
+      "epoch": 0.04,
+      "learning_rate": 4.996580961963974e-06,
+      "loss": 0.24,
+      "step": 84
+    },
+    {
+      "epoch": 0.04,
+      "learning_rate": 4.996473307623172e-06,
+      "loss": 0.2214,
+      "step": 85
+    },
+    {
+      "epoch": 0.04,
+      "learning_rate": 4.9963639858052245e-06,
+      "loss": 0.2514,
+      "step": 86
+    },
+    {
+      "epoch": 0.05,
+      "learning_rate": 4.996252996583151e-06,
+      "loss": 0.2285,
+      "step": 87
+    },
+    {
+      "epoch": 0.05,
+      "learning_rate": 4.996140340031086e-06,
+      "loss": 0.271,
+      "step": 88
+    },
+    {
+      "epoch": 0.05,
+      "learning_rate": 4.9960260162242745e-06,
+      "loss": 0.2343,
+      "step": 89
+    },
+    {
+      "epoch": 0.05,
+      "learning_rate": 4.9959100252390794e-06,
+      "loss": 0.2588,
+      "step": 90
+    },
+    {
+      "epoch": 0.05,
+      "learning_rate": 4.995792367152974e-06,
+      "loss": 0.2639,
+      "step": 91
+    },
+    {
+      "epoch": 0.05,
+      "learning_rate": 4.995673042044544e-06,
+      "loss": 0.2155,
+      "step": 92
+    },
+    {
+      "epoch": 0.05,
+      "learning_rate": 4.995552049993494e-06,
+      "loss": 0.2256,
+      "step": 93
+    },
+    {
+      "epoch": 0.05,
+      "learning_rate": 4.995429391080638e-06,
+      "loss": 0.2099,
+      "step": 94
+    },
+    {
+      "epoch": 0.05,
+      "learning_rate": 4.995305065387901e-06,
+      "loss": 0.2211,
+      "step": 95
+    },
+    {
+      "epoch": 0.05,
+      "learning_rate": 4.995179072998328e-06,
+      "loss": 0.2201,
+      "step": 96
+    },
+    {
+      "epoch": 0.05,
+      "learning_rate": 4.99505141399607e-06,
+      "loss": 0.225,
+      "step": 97
+    },
+    {
+      "epoch": 0.05,
+      "learning_rate": 4.994922088466397e-06,
+      "loss": 0.2187,
+      "step": 98
+    },
+    {
+      "epoch": 0.05,
+      "learning_rate": 4.9947910964956895e-06,
+      "loss": 0.2434,
+      "step": 99
+    },
+    {
+      "epoch": 0.05,
+      "learning_rate": 4.994658438171441e-06,
+      "loss": 0.2511,
+      "step": 100
+    },
+    {
+      "epoch": 0.05,
+      "eval_loss": 0.3000856041908264,
+      "eval_runtime": 116.9645,
+      "eval_samples_per_second": 26.692,
+      "eval_steps_per_second": 8.9,
+      "step": 100
+    },
+    {
+      "epoch": 0.05,
+      "learning_rate": 4.994524113582259e-06,
+      "loss": 0.2097,
+      "step": 101
+    },
+    {
+      "epoch": 0.05,
+      "learning_rate": 4.994388122817862e-06,
+      "loss": 0.2432,
+      "step": 102
+    },
+    {
+      "epoch": 0.05,
+      "learning_rate": 4.994250465969084e-06,
+      "loss": 0.2268,
+      "step": 103
+    },
+    {
+      "epoch": 0.05,
+      "learning_rate": 4.99411114312787e-06,
+      "loss": 0.2487,
+      "step": 104
+    },
+    {
+      "epoch": 0.05,
+      "learning_rate": 4.993970154387278e-06,
+      "loss": 0.2358,
+      "step": 105
+    },
+    {
+      "epoch": 0.05,
+      "learning_rate": 4.99382749984148e-06,
+      "loss": 0.2331,
+      "step": 106
+    },
+    {
+      "epoch": 0.06,
+      "learning_rate": 4.993683179585759e-06,
+      "loss": 0.2435,
+      "step": 107
+    },
+    {
+      "epoch": 0.06,
+      "learning_rate": 4.993537193716512e-06,
+      "loss": 0.1995,
+      "step": 108
+    },
+    {
+      "epoch": 0.06,
+      "learning_rate": 4.9933895423312465e-06,
+      "loss": 0.2153,
+      "step": 109
+    },
+    {
+      "epoch": 0.06,
+      "learning_rate": 4.9932402255285854e-06,
+      "loss": 0.242,
+      "step": 110
+    },
+    {
+      "epoch": 0.06,
+      "learning_rate": 4.99308924340826e-06,
+      "loss": 0.2495,
+      "step": 111
+    },
+    {
+      "epoch": 0.06,
+      "learning_rate": 4.99293659607112e-06,
+      "loss": 0.2303,
+      "step": 112
+    },
+    {
+      "epoch": 0.06,
+      "learning_rate": 4.9927822836191185e-06,
+      "loss": 0.2162,
+      "step": 113
+    },
+    {
+      "epoch": 0.06,
+      "learning_rate": 4.99262630615533e-06,
+      "loss": 0.217,
+      "step": 114
+    },
+    {
+      "epoch": 0.06,
+      "learning_rate": 4.992468663783936e-06,
+      "loss": 0.2485,
+      "step": 115
+    },
+    {
+      "epoch": 0.06,
+      "learning_rate": 4.992309356610231e-06,
+      "loss": 0.2618,
+      "step": 116
+    },
+    {
+      "epoch": 0.06,
+      "learning_rate": 4.99214838474062e-06,
+      "loss": 0.2237,
+      "step": 117
+    },
+    {
+      "epoch": 0.06,
+      "learning_rate": 4.991985748282625e-06,
+      "loss": 0.2327,
+      "step": 118
+    },
+    {
+      "epoch": 0.06,
+      "learning_rate": 4.991821447344873e-06,
+      "loss": 0.2544,
+      "step": 119
+    },
+    {
+      "epoch": 0.06,
+      "learning_rate": 4.991655482037105e-06,
+      "loss": 0.2337,
+      "step": 120
+    },
+    {
+      "epoch": 0.06,
+      "learning_rate": 4.9914878524701795e-06,
+      "loss": 0.2346,
+      "step": 121
+    },
+    {
+      "epoch": 0.06,
+      "learning_rate": 4.991318558756057e-06,
+      "loss": 0.2369,
+      "step": 122
+    },
+    {
+      "epoch": 0.06,
+      "learning_rate": 4.991147601007818e-06,
+      "loss": 0.2299,
+      "step": 123
+    },
+    {
+      "epoch": 0.06,
+      "learning_rate": 4.990974979339649e-06,
+      "loss": 0.2344,
+      "step": 124
+    },
+    {
+      "epoch": 0.06,
+      "learning_rate": 4.99080069386685e-06,
+      "loss": 0.2508,
+      "step": 125
+    },
+    {
+      "epoch": 0.07,
+      "learning_rate": 4.990624744705832e-06,
+      "loss": 0.2102,
+      "step": 126
+    },
+    {
+      "epoch": 0.07,
+      "learning_rate": 4.990447131974116e-06,
+      "loss": 0.2381,
+      "step": 127
+    },
+    {
+      "epoch": 0.07,
+      "learning_rate": 4.990267855790339e-06,
+      "loss": 0.2171,
+      "step": 128
+    },
+    {
+      "epoch": 0.07,
+      "learning_rate": 4.9900869162742424e-06,
+      "loss": 0.2274,
+      "step": 129
+    },
+    {
+      "epoch": 0.07,
+      "learning_rate": 4.989904313546682e-06,
+      "loss": 0.2222,
+      "step": 130
+    },
+    {
+      "epoch": 0.07,
+      "learning_rate": 4.989720047729625e-06,
+      "loss": 0.24,
+      "step": 131
+    },
+    {
+      "epoch": 0.07,
+      "learning_rate": 4.989534118946149e-06,
+      "loss": 0.2312,
+      "step": 132
+    },
+    {
+      "epoch": 0.07,
+      "learning_rate": 4.989346527320441e-06,
+      "loss": 0.2501,
+      "step": 133
+    },
+    {
+      "epoch": 0.07,
+      "learning_rate": 4.9891572729778e-06,
+      "loss": 0.2186,
+      "step": 134
+    },
+    {
+      "epoch": 0.07,
+      "learning_rate": 4.988966356044635e-06,
+      "loss": 0.1928,
+      "step": 135
+    },
+    {
+      "epoch": 0.07,
+      "learning_rate": 4.9887737766484666e-06,
+      "loss": 0.2003,
+      "step": 136
+    },
+    {
+      "epoch": 0.07,
+      "learning_rate": 4.988579534917924e-06,
+      "loss": 0.2293,
+      "step": 137
+    },
+    {
+      "epoch": 0.07,
+      "learning_rate": 4.9883836309827485e-06,
+      "loss": 0.2479,
+      "step": 138
+    },
+    {
+      "epoch": 0.07,
+      "learning_rate": 4.988186064973789e-06,
+      "loss": 0.2275,
+      "step": 139
+    },
+    {
+      "epoch": 0.07,
+      "learning_rate": 4.987986837023009e-06,
+      "loss": 0.2336,
+      "step": 140
+    },
+    {
+      "epoch": 0.07,
+      "learning_rate": 4.987785947263478e-06,
+      "loss": 0.2337,
+      "step": 141
+    },
+    {
+      "epoch": 0.07,
+      "learning_rate": 4.987583395829378e-06,
+      "loss": 0.215,
+      "step": 142
+    },
+    {
+      "epoch": 0.07,
+      "learning_rate": 4.987379182855997e-06,
+      "loss": 0.2349,
+      "step": 143
+    },
+    {
+      "epoch": 0.07,
+      "learning_rate": 4.987173308479738e-06,
+      "loss": 0.2352,
+      "step": 144
+    },
+    {
+      "epoch": 0.08,
+      "learning_rate": 4.986965772838112e-06,
+      "loss": 0.2227,
+      "step": 145
+    },
+    {
+      "epoch": 0.08,
+      "learning_rate": 4.9867565760697365e-06,
+      "loss": 0.2324,
+      "step": 146
+    },
+    {
+      "epoch": 0.08,
+      "learning_rate": 4.986545718314343e-06,
+      "loss": 0.2121,
+      "step": 147
+    },
+    {
+      "epoch": 0.08,
+      "learning_rate": 4.986333199712769e-06,
+      "loss": 0.2546,
+      "step": 148
+    },
+    {
+      "epoch": 0.08,
+      "learning_rate": 4.986119020406963e-06,
+      "loss": 0.2686,
+      "step": 149
+    },
+    {
+      "epoch": 0.08,
+      "learning_rate": 4.985903180539984e-06,
+      "loss": 0.2241,
+      "step": 150
+    },
+    {
+      "epoch": 0.08,
+      "eval_loss": 0.2863710820674896,
+      "eval_runtime": 116.8972,
+      "eval_samples_per_second": 26.707,
+      "eval_steps_per_second": 8.905,
+      "step": 150
+    },
+    {
+      "epoch": 0.08,
+      "learning_rate": 4.985685680255996e-06,
+      "loss": 0.2098,
+      "step": 151
+    },
+    {
+      "epoch": 0.08,
+      "learning_rate": 4.985466519700277e-06,
+      "loss": 0.209,
+      "step": 152
+    },
+    {
+      "epoch": 0.08,
+      "learning_rate": 4.98524569901921e-06,
+      "loss": 0.2044,
+      "step": 153
+    },
+    {
+      "epoch": 0.08,
+      "learning_rate": 4.985023218360289e-06,
+      "loss": 0.2301,
+      "step": 154
+    },
+    {
+      "epoch": 0.08,
+      "learning_rate": 4.984799077872116e-06,
+      "loss": 0.212,
+      "step": 155
+    },
+    {
+      "epoch": 0.08,
+      "learning_rate": 4.984573277704403e-06,
+      "loss": 0.2578,
+      "step": 156
+    },
+    {
+      "epoch": 0.08,
+      "learning_rate": 4.984345818007967e-06,
+      "loss": 0.2233,
+      "step": 157
+    },
+    {
+      "epoch": 0.08,
+      "learning_rate": 4.984116698934738e-06,
+      "loss": 0.206,
+      "step": 158
+    },
+    {
+      "epoch": 0.08,
+      "learning_rate": 4.983885920637752e-06,
+      "loss": 0.2257,
+      "step": 159
+    },
+    {
+      "epoch": 0.08,
+      "learning_rate": 4.983653483271152e-06,
+      "loss": 0.2336,
+      "step": 160
+    },
+    {
+      "epoch": 0.08,
+      "learning_rate": 4.983419386990192e-06,
+      "loss": 0.2057,
+      "step": 161
+    },
+    {
+      "epoch": 0.08,
+      "learning_rate": 4.983183631951232e-06,
+      "loss": 0.2126,
+      "step": 162
+    },
+    {
+      "epoch": 0.08,
+      "learning_rate": 4.982946218311742e-06,
+      "loss": 0.2219,
+      "step": 163
+    },
+    {
+      "epoch": 0.08,
+      "learning_rate": 4.982707146230296e-06,
+      "loss": 0.2153,
+      "step": 164
+    },
+    {
+      "epoch": 0.09,
+      "learning_rate": 4.982466415866579e-06,
+      "loss": 0.2564,
+      "step": 165
+    },
+    {
+      "epoch": 0.09,
+      "learning_rate": 4.9822240273813846e-06,
+      "loss": 0.1964,
+      "step": 166
+    },
+    {
+      "epoch": 0.09,
+      "learning_rate": 4.98197998093661e-06,
+      "loss": 0.2361,
+      "step": 167
+    },
+    {
+      "epoch": 0.09,
+      "learning_rate": 4.981734276695263e-06,
+      "loss": 0.2403,
+      "step": 168
+    },
+    {
+      "epoch": 0.09,
+      "learning_rate": 4.981486914821456e-06,
+      "loss": 0.2576,
+      "step": 169
+    },
+    {
+      "epoch": 0.09,
+      "learning_rate": 4.981237895480412e-06,
+      "loss": 0.2026,
+      "step": 170
+    },
+    {
+      "epoch": 0.09,
+      "learning_rate": 4.980987218838459e-06,
+      "loss": 0.203,
+      "step": 171
+    },
+    {
+      "epoch": 0.09,
+      "learning_rate": 4.980734885063031e-06,
+      "loss": 0.2299,
+      "step": 172
+    },
+    {
+      "epoch": 0.09,
+      "learning_rate": 4.980480894322672e-06,
+      "loss": 0.2457,
+      "step": 173
+    },
+    {
+      "epoch": 0.09,
+      "learning_rate": 4.980225246787028e-06,
+      "loss": 0.2218,
+      "step": 174
+    },
+    {
+      "epoch": 0.09,
+      "learning_rate": 4.9799679426268575e-06,
+      "loss": 0.223,
+      "step": 175
+    },
+    {
+      "epoch": 0.09,
+      "learning_rate": 4.979708982014021e-06,
+      "loss": 0.2425,
+      "step": 176
+    },
+    {
+      "epoch": 0.09,
+      "learning_rate": 4.9794483651214875e-06,
+      "loss": 0.2131,
+      "step": 177
+    },
+    {
+      "epoch": 0.09,
+      "learning_rate": 4.97918609212333e-06,
+      "loss": 0.2241,
+      "step": 178
+    },
+    {
+      "epoch": 0.09,
+      "learning_rate": 4.978922163194732e-06,
+      "loss": 0.2258,
+      "step": 179
+    },
+    {
+      "epoch": 0.09,
+      "learning_rate": 4.978656578511978e-06,
+      "loss": 0.2201,
+      "step": 180
+    },
+    {
+      "epoch": 0.09,
+      "learning_rate": 4.978389338252461e-06,
+      "loss": 0.1853,
+      "step": 181
+    },
+    {
+      "epoch": 0.09,
+      "learning_rate": 4.978120442594681e-06,
+      "loss": 0.2359,
+      "step": 182
+    },
+    {
+      "epoch": 0.09,
+      "learning_rate": 4.977849891718241e-06,
+      "loss": 0.2272,
+      "step": 183
+    },
+    {
+      "epoch": 0.1,
+      "learning_rate": 4.977577685803852e-06,
+      "loss": 0.2686,
+      "step": 184
+    },
+    {
+      "epoch": 0.1,
+      "learning_rate": 4.977303825033329e-06,
+      "loss": 0.2422,
+      "step": 185
+    },
+    {
+      "epoch": 0.1,
+      "learning_rate": 4.9770283095895926e-06,
+      "loss": 0.207,
+      "step": 186
+    },
+    {
+      "epoch": 0.1,
+      "learning_rate": 4.976751139656668e-06,
+      "loss": 0.2236,
+      "step": 187
+    },
+    {
+      "epoch": 0.1,
+      "learning_rate": 4.9764723154196855e-06,
+      "loss": 0.2552,
+      "step": 188
+    },
+    {
+      "epoch": 0.1,
+      "learning_rate": 4.976191837064884e-06,
+      "loss": 0.2377,
+      "step": 189
+    },
+    {
+      "epoch": 0.1,
+      "learning_rate": 4.975909704779602e-06,
+      "loss": 0.2262,
+      "step": 190
+    },
+    {
+      "epoch": 0.1,
+      "learning_rate": 4.975625918752286e-06,
+      "loss": 0.2227,
+      "step": 191
+    },
+    {
+      "epoch": 0.1,
+      "learning_rate": 4.975340479172484e-06,
+      "loss": 0.2024,
+      "step": 192
+    },
+    {
+      "epoch": 0.1,
+      "learning_rate": 4.975053386230854e-06,
+      "loss": 0.2103,
+      "step": 193
+    },
+    {
+      "epoch": 0.1,
+      "learning_rate": 4.974764640119153e-06,
+      "loss": 0.1749,
+      "step": 194
+    },
+    {
+      "epoch": 0.1,
+      "learning_rate": 4.974474241030243e-06,
+      "loss": 0.2218,
+      "step": 195
+    },
+    {
+      "epoch": 0.1,
+      "learning_rate": 4.974182189158093e-06,
+      "loss": 0.2189,
+      "step": 196
+    },
+    {
+      "epoch": 0.1,
+      "learning_rate": 4.973888484697773e-06,
+      "loss": 0.2354,
+      "step": 197
+    },
+    {
+      "epoch": 0.1,
+      "learning_rate": 4.973593127845458e-06,
+      "loss": 0.2029,
+      "step": 198
+    },
+    {
+      "epoch": 0.1,
+      "learning_rate": 4.973296118798428e-06,
+      "loss": 0.2164,
+      "step": 199
+    },
+    {
+      "epoch": 0.1,
+      "learning_rate": 4.972997457755063e-06,
+      "loss": 0.2146,
+      "step": 200
+    },
+    {
+      "epoch": 0.1,
+      "eval_loss": 0.27695992588996887,
+      "eval_runtime": 117.0831,
+      "eval_samples_per_second": 26.665,
+      "eval_steps_per_second": 8.891,
+      "step": 200
+    },
+    {
+      "epoch": 0.1,
+      "learning_rate": 4.9726971449148494e-06,
+      "loss": 0.2072,
+      "step": 201
+    },
+    {
+      "epoch": 0.1,
+      "learning_rate": 4.972395180478378e-06,
+      "loss": 0.2469,
+      "step": 202
+    },
+    {
+      "epoch": 0.11,
+      "learning_rate": 4.972091564647338e-06,
+      "loss": 0.2104,
+      "step": 203
+    },
+    {
+      "epoch": 0.11,
+      "learning_rate": 4.971786297624526e-06,
+      "loss": 0.2325,
+      "step": 204
+    },
+    {
+      "epoch": 0.11,
+      "learning_rate": 4.971479379613839e-06,
+      "loss": 0.2345,
+      "step": 205
+    },
+    {
+      "epoch": 0.11,
+      "learning_rate": 4.971170810820279e-06,
+      "loss": 0.2032,
+      "step": 206
+    },
+    {
+      "epoch": 0.11,
+      "learning_rate": 4.970860591449949e-06,
+      "loss": 0.2041,
+      "step": 207
+    },
+    {
+      "epoch": 0.11,
+      "learning_rate": 4.9705487217100534e-06,
+      "loss": 0.2067,
+      "step": 208
+    },
+    {
+      "epoch": 0.11,
+      "learning_rate": 4.970235201808902e-06,
+      "loss": 0.2151,
+      "step": 209
+    },
+    {
+      "epoch": 0.11,
+      "learning_rate": 4.969920031955904e-06,
+      "loss": 0.2036,
+      "step": 210
+    },
+    {
+      "epoch": 0.11,
+      "learning_rate": 4.969603212361572e-06,
+      "loss": 0.2124,
+      "step": 211
+    },
+    {
+      "epoch": 0.11,
+      "learning_rate": 4.969284743237521e-06,
+      "loss": 0.2226,
+      "step": 212
+    },
+    {
+      "epoch": 0.11,
+      "learning_rate": 4.968964624796467e-06,
+      "loss": 0.2345,
+      "step": 213
+    },
+    {
+      "epoch": 0.11,
+      "learning_rate": 4.968642857252227e-06,
+      "loss": 0.2463,
+      "step": 214
+    },
+    {
+      "epoch": 0.11,
+      "learning_rate": 4.96831944081972e-06,
+      "loss": 0.2056,
+      "step": 215
+    },
+    {
+      "epoch": 0.11,
+      "learning_rate": 4.967994375714969e-06,
+      "loss": 0.2419,
+      "step": 216
+    },
+    {
+      "epoch": 0.11,
+      "learning_rate": 4.967667662155091e-06,
+      "loss": 0.2028,
+      "step": 217
+    },
+    {
+      "epoch": 0.11,
+      "learning_rate": 4.967339300358313e-06,
+      "loss": 0.206,
+      "step": 218
+    },
+    {
+      "epoch": 0.11,
+      "learning_rate": 4.9670092905439574e-06,
+      "loss": 0.2233,
+      "step": 219
+    },
+    {
+      "epoch": 0.11,
+      "learning_rate": 4.966677632932448e-06,
+      "loss": 0.2201,
+      "step": 220
+    },
+    {
+      "epoch": 0.11,
+      "learning_rate": 4.966344327745311e-06,
+      "loss": 0.2028,
+      "step": 221
+    },
+    {
+      "epoch": 0.11,
+      "learning_rate": 4.966009375205171e-06,
+      "loss": 0.2245,
+      "step": 222
+    },
+    {
+      "epoch": 0.12,
+      "learning_rate": 4.965672775535754e-06,
+      "loss": 0.269,
+      "step": 223
+    },
+    {
+      "epoch": 0.12,
+      "learning_rate": 4.965334528961887e-06,
+      "loss": 0.203,
+      "step": 224
+    },
+    {
+      "epoch": 0.12,
+      "learning_rate": 4.964994635709494e-06,
+      "loss": 0.2087,
+      "step": 225
+    },
+    {
+      "epoch": 0.12,
+      "learning_rate": 4.964653096005604e-06,
+      "loss": 0.2251,
+      "step": 226
+    },
+    {
+      "epoch": 0.12,
+      "learning_rate": 4.964309910078341e-06,
+      "loss": 0.2166,
+      "step": 227
+    },
+    {
+      "epoch": 0.12,
+      "learning_rate": 4.96396507815693e-06,
+      "loss": 0.2383,
+      "step": 228
+    },
+    {
+      "epoch": 0.12,
+      "learning_rate": 4.963618600471696e-06,
+      "loss": 0.2301,
+      "step": 229
+    },
+    {
+      "epoch": 0.12,
+      "learning_rate": 4.963270477254063e-06,
+      "loss": 0.2323,
+      "step": 230
+    },
+    {
+      "epoch": 0.12,
+      "learning_rate": 4.962920708736553e-06,
+      "loss": 0.2053,
+      "step": 231
+    },
+    {
+      "epoch": 0.12,
+      "learning_rate": 4.96256929515279e-06,
+      "loss": 0.1843,
+      "step": 232
+    },
+    {
+      "epoch": 0.12,
+      "learning_rate": 4.962216236737493e-06,
+      "loss": 0.2677,
+      "step": 233
+    },
+    {
+      "epoch": 0.12,
+      "learning_rate": 4.9618615337264815e-06,
+      "loss": 0.2152,
+      "step": 234
+    },
+    {
+      "epoch": 0.12,
+      "learning_rate": 4.961505186356675e-06,
+      "loss": 0.2521,
+      "step": 235
+    },
+    {
+      "epoch": 0.12,
+      "learning_rate": 4.961147194866088e-06,
+      "loss": 0.1904,
+      "step": 236
+    },
+    {
+      "epoch": 0.12,
+      "learning_rate": 4.960787559493836e-06,
+      "loss": 0.2214,
+      "step": 237
+    },
+    {
+      "epoch": 0.12,
+      "learning_rate": 4.960426280480131e-06,
+      "loss": 0.1983,
+      "step": 238
+    },
+    {
+      "epoch": 0.12,
+      "learning_rate": 4.960063358066282e-06,
+      "loss": 0.2036,
+      "step": 239
+    },
+    {
+      "epoch": 0.12,
+      "learning_rate": 4.9596987924946986e-06,
+      "loss": 0.1975,
+      "step": 240
+    },
+    {
+      "epoch": 0.12,
+      "learning_rate": 4.959332584008884e-06,
+      "loss": 0.2071,
+      "step": 241
+    },
+    {
+      "epoch": 0.13,
+      "learning_rate": 4.958964732853445e-06,
+      "loss": 0.2146,
+      "step": 242
+    },
+    {
+      "epoch": 0.13,
+      "learning_rate": 4.9585952392740775e-06,
+      "loss": 0.2242,
+      "step": 243
+    },
+    {
+      "epoch": 0.13,
+      "learning_rate": 4.958224103517581e-06,
+      "loss": 0.2,
+      "step": 244
+    },
+    {
+      "epoch": 0.13,
+      "learning_rate": 4.957851325831848e-06,
+      "loss": 0.2162,
+      "step": 245
+    },
+    {
+      "epoch": 0.13,
+      "learning_rate": 4.957476906465869e-06,
+      "loss": 0.2308,
+      "step": 246
+    },
+    {
+      "epoch": 0.13,
+      "learning_rate": 4.957100845669732e-06,
+      "loss": 0.2388,
+      "step": 247
+    },
+    {
+      "epoch": 0.13,
+      "learning_rate": 4.9567231436946205e-06,
+      "loss": 0.2199,
+      "step": 248
+    },
+    {
+      "epoch": 0.13,
+      "learning_rate": 4.9563438007928135e-06,
+      "loss": 0.2159,
+      "step": 249
+    },
+    {
+      "epoch": 0.13,
+      "learning_rate": 4.955962817217686e-06,
+      "loss": 0.2195,
+      "step": 250
+    },
+    {
+      "epoch": 0.13,
+      "eval_loss": 0.26982513070106506,
+      "eval_runtime": 116.9388,
+      "eval_samples_per_second": 26.698,
+      "eval_steps_per_second": 8.902,
+      "step": 250
+    },
+    {
+      "epoch": 0.13,
+      "learning_rate": 4.955580193223712e-06,
+      "loss": 0.2099,
+      "step": 251
+    },
+    {
+      "epoch": 0.13,
+      "learning_rate": 4.955195929066455e-06,
+      "loss": 0.1712,
+      "step": 252
+    },
+    {
+      "epoch": 0.13,
+      "learning_rate": 4.954810025002581e-06,
+      "loss": 0.223,
+      "step": 253
+    },
+    {
+      "epoch": 0.13,
+      "learning_rate": 4.954422481289847e-06,
+      "loss": 0.2212,
+      "step": 254
+    },
+    {
+      "epoch": 0.13,
+      "learning_rate": 4.954033298187105e-06,
+      "loss": 0.2009,
+      "step": 255
+    },
+    {
+      "epoch": 0.13,
+      "learning_rate": 4.953642475954306e-06,
+      "loss": 0.2278,
+      "step": 256
+    },
+    {
+      "epoch": 0.13,
+      "learning_rate": 4.953250014852492e-06,
+      "loss": 0.201,
+      "step": 257
+    },
+    {
+      "epoch": 0.13,
+      "learning_rate": 4.952855915143799e-06,
+      "loss": 0.2165,
+      "step": 258
+    },
+    {
+      "epoch": 0.13,
+      "learning_rate": 4.952460177091461e-06,
+      "loss": 0.205,
+      "step": 259
+    },
+    {
+      "epoch": 0.13,
+      "learning_rate": 4.9520628009598045e-06,
+      "loss": 0.2069,
+      "step": 260
+    },
+    {
+      "epoch": 0.14,
+      "learning_rate": 4.951663787014249e-06,
+      "loss": 0.2349,
+      "step": 261
+    },
+    {
+      "epoch": 0.14,
+      "learning_rate": 4.951263135521311e-06,
+      "loss": 0.1846,
+      "step": 262
+    },
+    {
+      "epoch": 0.14,
+      "learning_rate": 4.9508608467485975e-06,
+      "loss": 0.1983,
+      "step": 263
+    },
+    {
+      "epoch": 0.14,
+      "learning_rate": 4.95045692096481e-06,
+      "loss": 0.1977,
+      "step": 264
+    },
+    {
+      "epoch": 0.14,
+      "learning_rate": 4.950051358439746e-06,
+      "loss": 0.2038,
+      "step": 265
+    },
+    {
+      "epoch": 0.14,
+      "learning_rate": 4.949644159444292e-06,
+      "loss": 0.2174,
+      "step": 266
+    },
+    {
+      "epoch": 0.14,
+      "learning_rate": 4.949235324250431e-06,
+      "loss": 0.2091,
+      "step": 267
+    },
+    {
+      "epoch": 0.14,
+      "learning_rate": 4.948824853131237e-06,
+      "loss": 0.1925,
+      "step": 268
+    },
+    {
+      "epoch": 0.14,
+      "learning_rate": 4.948412746360877e-06,
+      "loss": 0.2263,
+      "step": 269
+    },
+    {
+      "epoch": 0.14,
+      "learning_rate": 4.947999004214611e-06,
+      "loss": 0.2399,
+      "step": 270
+    },
+    {
+      "epoch": 0.14,
+      "learning_rate": 4.947583626968792e-06,
+      "loss": 0.2042,
+      "step": 271
+    },
+    {
+      "epoch": 0.14,
+      "learning_rate": 4.947166614900862e-06,
+      "loss": 0.217,
+      "step": 272
+    },
+    {
+      "epoch": 0.14,
+      "learning_rate": 4.946747968289359e-06,
+      "loss": 0.2031,
+      "step": 273
+    },
+    {
+      "epoch": 0.14,
+      "learning_rate": 4.946327687413911e-06,
+      "loss": 0.1871,
+      "step": 274
+    },
+    {
+      "epoch": 0.14,
+      "learning_rate": 4.9459057725552374e-06,
+      "loss": 0.2218,
+      "step": 275
+    },
+    {
+      "epoch": 0.14,
+      "learning_rate": 4.945482223995148e-06,
+      "loss": 0.1883,
+      "step": 276
+    },
+    {
+      "epoch": 0.14,
+      "learning_rate": 4.945057042016545e-06,
+      "loss": 0.1902,
+      "step": 277
+    },
+    {
+      "epoch": 0.14,
+      "learning_rate": 4.944630226903423e-06,
+      "loss": 0.1722,
+      "step": 278
+    },
+    {
+      "epoch": 0.14,
+      "learning_rate": 4.944201778940865e-06,
+      "loss": 0.209,
+      "step": 279
+    },
+    {
+      "epoch": 0.14,
+      "learning_rate": 4.943771698415045e-06,
+      "loss": 0.1753,
+      "step": 280
+    },
+    {
+      "epoch": 0.15,
+      "learning_rate": 4.943339985613228e-06,
+      "loss": 0.238,
+      "step": 281
+    },
+    {
+      "epoch": 0.15,
+      "learning_rate": 4.942906640823771e-06,
+      "loss": 0.196,
+      "step": 282
+    },
+    {
+      "epoch": 0.15,
+      "learning_rate": 4.942471664336118e-06,
+      "loss": 0.2061,
+      "step": 283
+    },
+    {
+      "epoch": 0.15,
+      "learning_rate": 4.942035056440804e-06,
+      "loss": 0.212,
+      "step": 284
+    },
+    {
+      "epoch": 0.15,
+      "learning_rate": 4.941596817429455e-06,
+      "loss": 0.1779,
+      "step": 285
+    },
+    {
+      "epoch": 0.15,
+      "learning_rate": 4.941156947594784e-06,
+      "loss": 0.1971,
+      "step": 286
+    },
+    {
+      "epoch": 0.15,
+      "learning_rate": 4.940715447230595e-06,
+      "loss": 0.1781,
+      "step": 287
+    },
+    {
+      "epoch": 0.15,
+      "learning_rate": 4.9402723166317814e-06,
+      "loss": 0.2442,
+      "step": 288
+    },
+    {
+      "epoch": 0.15,
+      "learning_rate": 4.939827556094324e-06,
+      "loss": 0.2088,
+      "step": 289
+    },
+    {
+      "epoch": 0.15,
+      "learning_rate": 4.939381165915294e-06,
+      "loss": 0.1878,
+      "step": 290
+    },
+    {
+      "epoch": 0.15,
+      "learning_rate": 4.938933146392849e-06,
+      "loss": 0.1964,
+      "step": 291
+    },
+    {
+      "epoch": 0.15,
+      "learning_rate": 4.938483497826237e-06,
+      "loss": 0.1979,
+      "step": 292
+    },
+    {
+      "epoch": 0.15,
+      "learning_rate": 4.938032220515793e-06,
+      "loss": 0.2117,
+      "step": 293
+    },
+    {
+      "epoch": 0.15,
+      "learning_rate": 4.937579314762939e-06,
+      "loss": 0.2216,
+      "step": 294
+    },
+    {
+      "epoch": 0.15,
+      "learning_rate": 4.937124780870187e-06,
+      "loss": 0.1968,
+      "step": 295
+    },
+    {
+      "epoch": 0.15,
+      "learning_rate": 4.936668619141135e-06,
+      "loss": 0.2281,
+      "step": 296
+    },
+    {
+      "epoch": 0.15,
+      "learning_rate": 4.936210829880468e-06,
+      "loss": 0.1981,
+      "step": 297
+    },
+    {
+      "epoch": 0.15,
+      "learning_rate": 4.93575141339396e-06,
+      "loss": 0.1928,
+      "step": 298
+    },
+    {
+      "epoch": 0.15,
+      "learning_rate": 4.935290369988468e-06,
+      "loss": 0.2009,
+      "step": 299
+    },
+    {
+      "epoch": 0.16,
+      "learning_rate": 4.93482769997194e-06,
+      "loss": 0.2103,
+      "step": 300
+    },
+    {
+      "epoch": 0.16,
+      "eval_loss": 0.2634771168231964,
+      "eval_runtime": 117.0691,
+      "eval_samples_per_second": 26.668,
+      "eval_steps_per_second": 8.892,
+      "step": 300
+    },
+    {
+      "epoch": 0.16,
+      "learning_rate": 4.934363403653408e-06,
+      "loss": 0.2057,
+      "step": 301
+    },
+    {
+      "epoch": 0.16,
+      "learning_rate": 4.9338974813429916e-06,
+      "loss": 0.1962,
+      "step": 302
+    },
+    {
+      "epoch": 0.16,
+      "learning_rate": 4.933429933351896e-06,
+      "loss": 0.2128,
+      "step": 303
+    },
+    {
+      "epoch": 0.16,
+      "learning_rate": 4.932960759992409e-06,
+      "loss": 0.1927,
+      "step": 304
+    },
+    {
+      "epoch": 0.16,
+      "learning_rate": 4.932489961577911e-06,
+      "loss": 0.2363,
+      "step": 305
+    },
+    {
+      "epoch": 0.16,
+      "learning_rate": 4.932017538422861e-06,
+      "loss": 0.2279,
+      "step": 306
+    },
+    {
+      "epoch": 0.16,
+      "learning_rate": 4.931543490842807e-06,
+      "loss": 0.2037,
+      "step": 307
+    },
+    {
+      "epoch": 0.16,
+      "learning_rate": 4.9310678191543814e-06,
+      "loss": 0.2031,
+      "step": 308
+    },
+    {
+      "epoch": 0.16,
+      "learning_rate": 4.9305905236753016e-06,
+      "loss": 0.2037,
+      "step": 309
+    },
+    {
+      "epoch": 0.16,
+      "learning_rate": 4.930111604724367e-06,
+      "loss": 0.1995,
+      "step": 310
+    },
+    {
+      "epoch": 0.16,
+      "learning_rate": 4.929631062621465e-06,
+      "loss": 0.2312,
+      "step": 311
+    },
+    {
+      "epoch": 0.16,
+      "learning_rate": 4.929148897687566e-06,
+      "loss": 0.2009,
+      "step": 312
+    },
+    {
+      "epoch": 0.16,
+      "learning_rate": 4.928665110244723e-06,
+      "loss": 0.2064,
+      "step": 313
+    },
+    {
+      "epoch": 0.16,
+      "learning_rate": 4.9281797006160735e-06,
+      "loss": 0.2177,
+      "step": 314
+    },
+    {
+      "epoch": 0.16,
+      "learning_rate": 4.9276926691258384e-06,
+      "loss": 0.1856,
+      "step": 315
+    },
+    {
+      "epoch": 0.16,
+      "learning_rate": 4.927204016099324e-06,
+      "loss": 0.1898,
+      "step": 316
+    },
+    {
+      "epoch": 0.16,
+      "learning_rate": 4.926713741862915e-06,
+      "loss": 0.2196,
+      "step": 317
+    },
+    {
+      "epoch": 0.16,
+      "learning_rate": 4.926221846744085e-06,
+      "loss": 0.201,
+      "step": 318
+    },
+    {
+      "epoch": 0.17,
+      "learning_rate": 4.925728331071384e-06,
+      "loss": 0.2074,
+      "step": 319
+    },
+    {
+      "epoch": 0.17,
+      "learning_rate": 4.92523319517445e-06,
+      "loss": 0.1953,
+      "step": 320
+    },
+    {
+      "epoch": 0.17,
+      "learning_rate": 4.924736439383998e-06,
+      "loss": 0.2067,
+      "step": 321
+    },
+    {
+      "epoch": 0.17,
+      "learning_rate": 4.924238064031831e-06,
+      "loss": 0.1851,
+      "step": 322
+    },
+    {
+      "epoch": 0.17,
+      "learning_rate": 4.923738069450827e-06,
+      "loss": 0.1941,
+      "step": 323
+    },
+    {
+      "epoch": 0.17,
+      "learning_rate": 4.9232364559749505e-06,
+      "loss": 0.2121,
+      "step": 324
+    },
+    {
+      "epoch": 0.17,
+      "learning_rate": 4.922733223939246e-06,
+      "loss": 0.2068,
+      "step": 325
+    },
+    {
+      "epoch": 0.17,
+      "learning_rate": 4.922228373679838e-06,
+      "loss": 0.1909,
+      "step": 326
+    },
+    {
+      "epoch": 0.17,
+      "learning_rate": 4.921721905533935e-06,
+      "loss": 0.1944,
+      "step": 327
+    },
+    {
+      "epoch": 0.17,
+      "learning_rate": 4.921213819839821e-06,
+      "loss": 0.1843,
+      "step": 328
+    },
+    {
+      "epoch": 0.17,
+      "learning_rate": 4.920704116936863e-06,
+      "loss": 0.2175,
+      "step": 329
+    },
+    {
+      "epoch": 0.17,
+      "learning_rate": 4.920192797165511e-06,
+      "loss": 0.1919,
+      "step": 330
+    },
+    {
+      "epoch": 0.17,
+      "learning_rate": 4.919679860867291e-06,
+      "loss": 0.2284,
+      "step": 331
+    },
+    {
+      "epoch": 0.17,
+      "learning_rate": 4.91916530838481e-06,
+      "loss": 0.1956,
+      "step": 332
+    },
+    {
+      "epoch": 0.17,
+      "learning_rate": 4.918649140061755e-06,
+      "loss": 0.1885,
+      "step": 333
+    },
+    {
+      "epoch": 0.17,
+      "learning_rate": 4.918131356242892e-06,
+      "loss": 0.1955,
+      "step": 334
+    },
+    {
+      "epoch": 0.17,
+      "learning_rate": 4.917611957274065e-06,
+      "loss": 0.1957,
+      "step": 335
+    },
+    {
+      "epoch": 0.17,
+      "learning_rate": 4.9170909435022e-06,
+      "loss": 0.2106,
+      "step": 336
+    },
+    {
+      "epoch": 0.17,
+      "learning_rate": 4.916568315275297e-06,
+      "loss": 0.1794,
+      "step": 337
+    },
+    {
+      "epoch": 0.17,
+      "learning_rate": 4.916044072942437e-06,
+      "loss": 0.1912,
+      "step": 338
+    },
+    {
+      "epoch": 0.18,
+      "learning_rate": 4.915518216853779e-06,
+      "loss": 0.172,
+      "step": 339
+    },
+    {
+      "epoch": 0.18,
+      "learning_rate": 4.9149907473605615e-06,
+      "loss": 0.1844,
+      "step": 340
+    },
+    {
+      "epoch": 0.18,
+      "learning_rate": 4.914461664815096e-06,
+      "loss": 0.1968,
+      "step": 341
+    },
+    {
+      "epoch": 0.18,
+      "learning_rate": 4.913930969570776e-06,
+      "loss": 0.1917,
+      "step": 342
+    },
+    {
+      "epoch": 0.18,
+      "learning_rate": 4.913398661982071e-06,
+      "loss": 0.2165,
+      "step": 343
+    },
+    {
+      "epoch": 0.18,
+      "learning_rate": 4.912864742404525e-06,
+      "loss": 0.1825,
+      "step": 344
+    },
+    {
+      "epoch": 0.18,
+      "learning_rate": 4.912329211194761e-06,
+      "loss": 0.1871,
+      "step": 345
+    },
+    {
+      "epoch": 0.18,
+      "learning_rate": 4.911792068710479e-06,
+      "loss": 0.1932,
+      "step": 346
+    },
+    {
+      "epoch": 0.18,
+      "learning_rate": 4.911253315310454e-06,
+      "loss": 0.2002,
+      "step": 347
+    },
+    {
+      "epoch": 0.18,
+      "learning_rate": 4.910712951354536e-06,
+      "loss": 0.2206,
+      "step": 348
+    },
+    {
+      "epoch": 0.18,
+      "learning_rate": 4.910170977203654e-06,
+      "loss": 0.2102,
+      "step": 349
+    },
+    {
+      "epoch": 0.18,
+      "learning_rate": 4.9096273932198086e-06,
+      "loss": 0.2127,
+      "step": 350
+    },
+    {
+      "epoch": 0.18,
+      "eval_loss": 0.2578897178173065,
+      "eval_runtime": 117.0059,
+      "eval_samples_per_second": 26.682,
+      "eval_steps_per_second": 8.897,
+      "step": 350
+    },
+    {
+      "epoch": 0.18,
+      "learning_rate": 4.909082199766079e-06,
+      "loss": 0.1925,
+      "step": 351
+    },
+    {
+      "epoch": 0.18,
+      "learning_rate": 4.908535397206616e-06,
+      "loss": 0.2034,
+      "step": 352
+    },
+    {
+      "epoch": 0.18,
+      "learning_rate": 4.90798698590665e-06,
+      "loss": 0.2146,
+      "step": 353
+    },
+    {
+      "epoch": 0.18,
+      "learning_rate": 4.9074369662324795e-06,
+      "loss": 0.2006,
+      "step": 354
+    },
+    {
+      "epoch": 0.18,
+      "learning_rate": 4.906885338551484e-06,
+      "loss": 0.1792,
+      "step": 355
+    },
+    {
+      "epoch": 0.18,
+      "learning_rate": 4.9063321032321125e-06,
+      "loss": 0.1941,
+      "step": 356
+    },
+    {
+      "epoch": 0.18,
+      "learning_rate": 4.905777260643889e-06,
+      "loss": 0.1696,
+      "step": 357
+    },
+    {
+      "epoch": 0.19,
+      "learning_rate": 4.905220811157412e-06,
+      "loss": 0.1858,
+      "step": 358
+    },
+    {
+      "epoch": 0.19,
+      "learning_rate": 4.904662755144352e-06,
+      "loss": 0.2128,
+      "step": 359
+    },
+    {
+      "epoch": 0.19,
+      "learning_rate": 4.904103092977452e-06,
+      "loss": 0.1789,
+      "step": 360
+    },
+    {
+      "epoch": 0.19,
+      "learning_rate": 4.9035418250305314e-06,
+      "loss": 0.1784,
+      "step": 361
+    },
+    {
+      "epoch": 0.19,
+      "learning_rate": 4.902978951678478e-06,
+      "loss": 0.2043,
+      "step": 362
+    },
+    {
+      "epoch": 0.19,
+      "learning_rate": 4.902414473297254e-06,
+      "loss": 0.1821,
+      "step": 363
+    },
+    {
+      "epoch": 0.19,
+      "learning_rate": 4.901848390263892e-06,
+      "loss": 0.2001,
+      "step": 364
+    },
+    {
+      "epoch": 0.19,
+      "learning_rate": 4.9012807029565e-06,
+      "loss": 0.2046,
+      "step": 365
+    },
+    {
+      "epoch": 0.19,
+      "learning_rate": 4.9007114117542535e-06,
+      "loss": 0.1817,
+      "step": 366
+    },
+    {
+      "epoch": 0.19,
+      "learning_rate": 4.900140517037401e-06,
+      "loss": 0.2121,
+      "step": 367
+    },
+    {
+      "epoch": 0.19,
+      "learning_rate": 4.899568019187262e-06,
+      "loss": 0.1873,
+      "step": 368
+    },
+    {
+      "epoch": 0.19,
+      "learning_rate": 4.898993918586227e-06,
+      "loss": 0.1773,
+      "step": 369
+    },
+    {
+      "epoch": 0.19,
+      "learning_rate": 4.8984182156177565e-06,
+      "loss": 0.1785,
+      "step": 370
+    },
+    {
+      "epoch": 0.19,
+      "learning_rate": 4.8978409106663815e-06,
+      "loss": 0.1821,
+      "step": 371
+    },
+    {
+      "epoch": 0.19,
+      "learning_rate": 4.897262004117704e-06,
+      "loss": 0.1821,
+      "step": 372
+    },
+    {
+      "epoch": 0.19,
+      "learning_rate": 4.896681496358394e-06,
+      "loss": 0.1857,
+      "step": 373
+    },
+    {
+      "epoch": 0.19,
+      "learning_rate": 4.896099387776192e-06,
+      "loss": 0.2154,
+      "step": 374
+    },
+    {
+      "epoch": 0.19,
+      "learning_rate": 4.8955156787599074e-06,
+      "loss": 0.18,
+      "step": 375
+    },
+    {
+      "epoch": 0.19,
+      "learning_rate": 4.8949303696994195e-06,
+      "loss": 0.1929,
+      "step": 376
+    },
+    {
+      "epoch": 0.2,
+      "learning_rate": 4.894343460985676e-06,
+      "loss": 0.1914,
+      "step": 377
+    },
+    {
+      "epoch": 0.2,
+      "learning_rate": 4.8937549530106896e-06,
+      "loss": 0.178,
+      "step": 378
+    },
+    {
+      "epoch": 0.2,
+      "learning_rate": 4.893164846167549e-06,
+      "loss": 0.189,
+      "step": 379
+    },
+    {
+      "epoch": 0.2,
+      "learning_rate": 4.8925731408504015e-06,
+      "loss": 0.213,
+      "step": 380
+    },
+    {
+      "epoch": 0.2,
+      "learning_rate": 4.891979837454469e-06,
+      "loss": 0.1932,
+      "step": 381
+    },
+    {
+      "epoch": 0.2,
+      "learning_rate": 4.891384936376038e-06,
+      "loss": 0.2087,
+      "step": 382
+    },
+    {
+      "epoch": 0.2,
+      "learning_rate": 4.890788438012464e-06,
+      "loss": 0.1904,
+      "step": 383
+    },
+    {
+      "epoch": 0.2,
+      "learning_rate": 4.8901903427621654e-06,
+      "loss": 0.194,
+      "step": 384
+    },
+    {
+      "epoch": 0.2,
+      "learning_rate": 4.8895906510246315e-06,
+      "loss": 0.2125,
+      "step": 385
+    },
+    {
+      "epoch": 0.2,
+      "learning_rate": 4.888989363200416e-06,
+      "loss": 0.2141,
+      "step": 386
+    },
+    {
+      "epoch": 0.2,
+      "learning_rate": 4.888386479691139e-06,
+      "loss": 0.1937,
+      "step": 387
+    },
+    {
+      "epoch": 0.2,
+      "learning_rate": 4.887782000899486e-06,
+      "loss": 0.1912,
+      "step": 388
+    },
+    {
+      "epoch": 0.2,
+      "learning_rate": 4.887175927229209e-06,
+      "loss": 0.2249,
+      "step": 389
+    },
+    {
+      "epoch": 0.2,
+      "learning_rate": 4.8865682590851235e-06,
+      "loss": 0.1779,
+      "step": 390
+    },
+    {
+      "epoch": 0.2,
+      "learning_rate": 4.885958996873113e-06,
+      "loss": 0.2046,
+      "step": 391
+    },
+    {
+      "epoch": 0.2,
+      "learning_rate": 4.8853481410001225e-06,
+      "loss": 0.1748,
+      "step": 392
+    },
+    {
+      "epoch": 0.2,
+      "learning_rate": 4.8847356918741625e-06,
+      "loss": 0.1951,
+      "step": 393
+    },
+    {
+      "epoch": 0.2,
+      "learning_rate": 4.88412164990431e-06,
+      "loss": 0.2208,
+      "step": 394
+    },
+    {
+      "epoch": 0.2,
+      "learning_rate": 4.883506015500702e-06,
+      "loss": 0.1736,
+      "step": 395
+    },
+    {
+      "epoch": 0.2,
+      "learning_rate": 4.882888789074542e-06,
+      "loss": 0.1967,
+      "step": 396
+    },
+    {
+      "epoch": 0.21,
+      "learning_rate": 4.882269971038095e-06,
+      "loss": 0.2056,
+      "step": 397
+    },
+    {
+      "epoch": 0.21,
+      "learning_rate": 4.881649561804691e-06,
+      "loss": 0.1776,
+      "step": 398
+    },
+    {
+      "epoch": 0.21,
+      "learning_rate": 4.881027561788722e-06,
+      "loss": 0.2042,
+      "step": 399
+    },
+    {
+      "epoch": 0.21,
+      "learning_rate": 4.8804039714056415e-06,
+      "loss": 0.2175,
+      "step": 400
+    },
+    {
+      "epoch": 0.21,
+      "eval_loss": 0.25287875533103943,
+      "eval_runtime": 117.6076,
+      "eval_samples_per_second": 26.546,
+      "eval_steps_per_second": 8.851,
+      "step": 400
+    },
+    {
+      "epoch": 0.21,
+      "learning_rate": 4.879778791071967e-06,
+      "loss": 0.2075,
+      "step": 401
+    },
+    {
+      "epoch": 0.21,
+      "learning_rate": 4.879152021205278e-06,
+      "loss": 0.1878,
+      "step": 402
+    },
+    {
+      "epoch": 0.21,
+      "learning_rate": 4.878523662224212e-06,
+      "loss": 0.1799,
+      "step": 403
+    },
+    {
+      "epoch": 0.21,
+      "learning_rate": 4.8778937145484735e-06,
+      "loss": 0.1785,
+      "step": 404
+    },
+    {
+      "epoch": 0.21,
+      "learning_rate": 4.877262178598824e-06,
+      "loss": 0.1918,
+      "step": 405
+    },
+    {
+      "epoch": 0.21,
+      "learning_rate": 4.876629054797087e-06,
+      "loss": 0.1984,
+      "step": 406
+    },
+    {
+      "epoch": 0.21,
+      "learning_rate": 4.875994343566148e-06,
+      "loss": 0.1727,
+      "step": 407
+    },
+    {
+      "epoch": 0.21,
+      "learning_rate": 4.875358045329951e-06,
+      "loss": 0.2053,
+      "step": 408
+    },
+    {
+      "epoch": 0.21,
+      "learning_rate": 4.8747201605135006e-06,
+      "loss": 0.2315,
+      "step": 409
+    },
+    {
+      "epoch": 0.21,
+      "learning_rate": 4.874080689542862e-06,
+      "loss": 0.1865,
+      "step": 410
+    },
+    {
+      "epoch": 0.21,
+      "learning_rate": 4.873439632845157e-06,
+      "loss": 0.1975,
+      "step": 411
+    },
+    {
+      "epoch": 0.21,
+      "learning_rate": 4.872796990848571e-06,
+      "loss": 0.2101,
+      "step": 412
+    },
+    {
+      "epoch": 0.21,
+      "learning_rate": 4.872152763982344e-06,
+      "loss": 0.2033,
+      "step": 413
+    },
+    {
+      "epoch": 0.21,
+      "learning_rate": 4.871506952676778e-06,
+      "loss": 0.1901,
+      "step": 414
+    },
+    {
+      "epoch": 0.21,
+      "learning_rate": 4.87085955736323e-06,
+      "loss": 0.193,
+      "step": 415
+    },
+    {
+      "epoch": 0.22,
+      "learning_rate": 4.870210578474119e-06,
+      "loss": 0.1915,
+      "step": 416
+    },
+    {
+      "epoch": 0.22,
+      "learning_rate": 4.869560016442917e-06,
+      "loss": 0.1858,
+      "step": 417
+    },
+    {
+      "epoch": 0.22,
+      "learning_rate": 4.8689078717041575e-06,
+      "loss": 0.1779,
+      "step": 418
+    },
+    {
+      "epoch": 0.22,
+      "learning_rate": 4.86825414469343e-06,
+      "loss": 0.1726,
+      "step": 419
+    },
+    {
+      "epoch": 0.22,
+      "learning_rate": 4.8675988358473784e-06,
+      "loss": 0.2021,
+      "step": 420
+    },
+    {
+      "epoch": 0.22,
+      "learning_rate": 4.866941945603707e-06,
+      "loss": 0.2182,
+      "step": 421
+    },
+    {
+      "epoch": 0.22,
+      "learning_rate": 4.866283474401174e-06,
+      "loss": 0.2062,
+      "step": 422
+    },
+    {
+      "epoch": 0.22,
+      "learning_rate": 4.865623422679593e-06,
+      "loss": 0.2088,
+      "step": 423
+    },
+    {
+      "epoch": 0.22,
+      "learning_rate": 4.864961790879835e-06,
+      "loss": 0.1943,
+      "step": 424
+    },
+    {
+      "epoch": 0.22,
+      "learning_rate": 4.864298579443828e-06,
+      "loss": 0.2069,
+      "step": 425
+    },
+    {
+      "epoch": 0.22,
+      "learning_rate": 4.86363378881455e-06,
+      "loss": 0.172,
+      "step": 426
+    },
+    {
+      "epoch": 0.22,
+      "learning_rate": 4.862967419436038e-06,
+      "loss": 0.2052,
+      "step": 427
+    },
+    {
+      "epoch": 0.22,
+      "learning_rate": 4.862299471753381e-06,
+      "loss": 0.2261,
+      "step": 428
+    },
+    {
+      "epoch": 0.22,
+      "learning_rate": 4.861629946212725e-06,
+      "loss": 0.188,
+      "step": 429
+    },
+    {
+      "epoch": 0.22,
+      "learning_rate": 4.860958843261267e-06,
+      "loss": 0.1842,
+      "step": 430
+    },
+    {
+      "epoch": 0.22,
+      "learning_rate": 4.8602861633472595e-06,
+      "loss": 0.1784,
+      "step": 431
+    },
+    {
+      "epoch": 0.22,
+      "learning_rate": 4.859611906920008e-06,
+      "loss": 0.1678,
+      "step": 432
+    },
+    {
+      "epoch": 0.22,
+      "learning_rate": 4.85893607442987e-06,
+      "loss": 0.199,
+      "step": 433
+    },
+    {
+      "epoch": 0.22,
+      "learning_rate": 4.858258666328256e-06,
+      "loss": 0.1835,
+      "step": 434
+    },
+    {
+      "epoch": 0.23,
+      "learning_rate": 4.8575796830676305e-06,
+      "loss": 0.2551,
+      "step": 435
+    },
+    {
+      "epoch": 0.23,
+      "learning_rate": 4.856899125101508e-06,
+      "loss": 0.1783,
+      "step": 436
+    },
+    {
+      "epoch": 0.23,
+      "learning_rate": 4.856216992884456e-06,
+      "loss": 0.1776,
+      "step": 437
+    },
+    {
+      "epoch": 0.23,
+      "learning_rate": 4.855533286872093e-06,
+      "loss": 0.1648,
+      "step": 438
+    },
+    {
+      "epoch": 0.23,
+      "learning_rate": 4.854848007521089e-06,
+      "loss": 0.2,
+      "step": 439
+    },
+    {
+      "epoch": 0.23,
+      "learning_rate": 4.854161155289164e-06,
+      "loss": 0.189,
+      "step": 440
+    },
+    {
+      "epoch": 0.23,
+      "learning_rate": 4.853472730635092e-06,
+      "loss": 0.181,
+      "step": 441
+    },
+    {
+      "epoch": 0.23,
+      "learning_rate": 4.852782734018692e-06,
+      "loss": 0.2309,
+      "step": 442
+    },
+    {
+      "epoch": 0.23,
+      "learning_rate": 4.852091165900835e-06,
+      "loss": 0.2078,
+      "step": 443
+    },
+    {
+      "epoch": 0.23,
+      "learning_rate": 4.851398026743444e-06,
+      "loss": 0.2124,
+      "step": 444
+    },
+    {
+      "epoch": 0.23,
+      "learning_rate": 4.850703317009491e-06,
+      "loss": 0.2003,
+      "step": 445
+    },
+    {
+      "epoch": 0.23,
+      "learning_rate": 4.850007037162992e-06,
+      "loss": 0.2125,
+      "step": 446
+    },
+    {
+      "epoch": 0.23,
+      "learning_rate": 4.849309187669017e-06,
+      "loss": 0.1732,
+      "step": 447
+    },
+    {
+      "epoch": 0.23,
+      "learning_rate": 4.848609768993684e-06,
+      "loss": 0.1904,
+      "step": 448
+    },
+    {
+      "epoch": 0.23,
+      "learning_rate": 4.847908781604156e-06,
+      "loss": 0.1871,
+      "step": 449
+    },
+    {
+      "epoch": 0.23,
+      "learning_rate": 4.8472062259686465e-06,
+      "loss": 0.204,
+      "step": 450
+    },
+    {
+      "epoch": 0.23,
+      "eval_loss": 0.24905972182750702,
+      "eval_runtime": 117.6955,
+      "eval_samples_per_second": 26.526,
+      "eval_steps_per_second": 8.845,
+      "step": 450
+    },
+    {
+      "epoch": 0.23,
+      "learning_rate": 4.846502102556415e-06,
+      "loss": 0.206,
+      "step": 451
+    },
+    {
+      "epoch": 0.23,
+      "learning_rate": 4.845796411837771e-06,
+      "loss": 0.1906,
+      "step": 452
+    },
+    {
+      "epoch": 0.23,
+      "learning_rate": 4.845089154284066e-06,
+      "loss": 0.1859,
+      "step": 453
+    },
+    {
+      "epoch": 0.23,
+      "learning_rate": 4.844380330367701e-06,
+      "loss": 0.1841,
+      "step": 454
+    },
+    {
+      "epoch": 0.24,
+      "learning_rate": 4.8436699405621255e-06,
+      "loss": 0.1968,
+      "step": 455
+    },
+    {
+      "epoch": 0.24,
+      "learning_rate": 4.84295798534183e-06,
+      "loss": 0.1837,
+      "step": 456
+    },
+    {
+      "epoch": 0.24,
+      "learning_rate": 4.842244465182353e-06,
+      "loss": 0.1994,
+      "step": 457
+    },
+    {
+      "epoch": 0.24,
+      "learning_rate": 4.841529380560279e-06,
+      "loss": 0.2124,
+      "step": 458
+    },
+    {
+      "epoch": 0.24,
+      "learning_rate": 4.840812731953237e-06,
+      "loss": 0.2163,
+      "step": 459
+    },
+    {
+      "epoch": 0.24,
+      "learning_rate": 4.8400945198399e-06,
+      "loss": 0.1551,
+      "step": 460
+    },
+    {
+      "epoch": 0.24,
+      "learning_rate": 4.839374744699985e-06,
+      "loss": 0.2199,
+      "step": 461
+    },
+    {
+      "epoch": 0.24,
+      "learning_rate": 4.838653407014255e-06,
+      "loss": 0.1902,
+      "step": 462
+    },
+    {
+      "epoch": 0.24,
+      "learning_rate": 4.837930507264513e-06,
+      "loss": 0.1726,
+      "step": 463
+    },
+    {
+      "epoch": 0.24,
+      "learning_rate": 4.83720604593361e-06,
+      "loss": 0.1805,
+      "step": 464
+    },
+    {
+      "epoch": 0.24,
+      "learning_rate": 4.836480023505436e-06,
+      "loss": 0.1844,
+      "step": 465
+    },
+    {
+      "epoch": 0.24,
+      "learning_rate": 4.835752440464927e-06,
+      "loss": 0.1859,
+      "step": 466
+    },
+    {
+      "epoch": 0.24,
+      "learning_rate": 4.835023297298059e-06,
+      "loss": 0.1604,
+      "step": 467
+    },
+    {
+      "epoch": 0.24,
+      "learning_rate": 4.83429259449185e-06,
+      "loss": 0.2047,
+      "step": 468
+    },
+    {
+      "epoch": 0.24,
+      "learning_rate": 4.833560332534362e-06,
+      "loss": 0.1859,
+      "step": 469
+    },
+    {
+      "epoch": 0.24,
+      "learning_rate": 4.832826511914696e-06,
+      "loss": 0.2038,
+      "step": 470
+    },
+    {
+      "epoch": 0.24,
+      "learning_rate": 4.832091133122997e-06,
+      "loss": 0.1884,
+      "step": 471
+    },
+    {
+      "epoch": 0.24,
+      "learning_rate": 4.831354196650446e-06,
+      "loss": 0.1884,
+      "step": 472
+    },
+    {
+      "epoch": 0.24,
+      "learning_rate": 4.830615702989269e-06,
+      "loss": 0.1918,
+      "step": 473
+    },
+    {
+      "epoch": 0.25,
+      "learning_rate": 4.82987565263273e-06,
+      "loss": 0.1922,
+      "step": 474
+    },
+    {
+      "epoch": 0.25,
+      "learning_rate": 4.829134046075133e-06,
+      "loss": 0.191,
+      "step": 475
+    },
+    {
+      "epoch": 0.25,
+      "learning_rate": 4.828390883811821e-06,
+      "loss": 0.1665,
+      "step": 476
+    },
+    {
+      "epoch": 0.25,
+      "learning_rate": 4.827646166339178e-06,
+      "loss": 0.195,
+      "step": 477
+    },
+    {
+      "epoch": 0.25,
+      "learning_rate": 4.826899894154625e-06,
+      "loss": 0.1959,
+      "step": 478
+    },
+    {
+      "epoch": 0.25,
+      "learning_rate": 4.826152067756621e-06,
+      "loss": 0.1704,
+      "step": 479
+    },
+    {
+      "epoch": 0.25,
+      "learning_rate": 4.8254026876446655e-06,
+      "loss": 0.2036,
+      "step": 480
+    },
+    {
+      "epoch": 0.25,
+      "learning_rate": 4.824651754319293e-06,
+      "loss": 0.1723,
+      "step": 481
+    },
+    {
+      "epoch": 0.25,
+      "learning_rate": 4.823899268282076e-06,
+      "loss": 0.1782,
+      "step": 482
+    },
+    {
+      "epoch": 0.25,
+      "learning_rate": 4.823145230035628e-06,
+      "loss": 0.2121,
+      "step": 483
+    },
+    {
+      "epoch": 0.25,
+      "learning_rate": 4.822389640083593e-06,
+      "loss": 0.1731,
+      "step": 484
+    },
+    {
+      "epoch": 0.25,
+      "learning_rate": 4.821632498930656e-06,
+      "loss": 0.2029,
+      "step": 485
+    },
+    {
+      "epoch": 0.25,
+      "learning_rate": 4.820873807082536e-06,
+      "loss": 0.1858,
+      "step": 486
+    },
+    {
+      "epoch": 0.25,
+      "learning_rate": 4.820113565045989e-06,
+      "loss": 0.2132,
+      "step": 487
+    },
+    {
+      "epoch": 0.25,
+      "learning_rate": 4.819351773328805e-06,
+      "loss": 0.1802,
+      "step": 488
+    },
+    {
+      "epoch": 0.25,
+      "learning_rate": 4.81858843243981e-06,
+      "loss": 0.1906,
+      "step": 489
+    },
+    {
+      "epoch": 0.25,
+      "learning_rate": 4.817823542888866e-06,
+      "loss": 0.1794,
+      "step": 490
+    },
+    {
+      "epoch": 0.25,
+      "learning_rate": 4.817057105186866e-06,
+      "loss": 0.2247,
+      "step": 491
+    },
+    {
+      "epoch": 0.25,
+      "learning_rate": 4.816289119845742e-06,
+      "loss": 0.1665,
+      "step": 492
+    },
+    {
+      "epoch": 0.26,
+      "learning_rate": 4.815519587378453e-06,
+      "loss": 0.1717,
+      "step": 493
+    },
+    {
+      "epoch": 0.26,
+      "learning_rate": 4.8147485082989984e-06,
+      "loss": 0.1809,
+      "step": 494
+    },
+    {
+      "epoch": 0.26,
+      "learning_rate": 4.813975883122406e-06,
+      "loss": 0.2045,
+      "step": 495
+    },
+    {
+      "epoch": 0.26,
+      "learning_rate": 4.813201712364738e-06,
+      "loss": 0.1711,
+      "step": 496
+    },
+    {
+      "epoch": 0.26,
+      "learning_rate": 4.812425996543088e-06,
+      "loss": 0.1918,
+      "step": 497
+    },
+    {
+      "epoch": 0.26,
+      "learning_rate": 4.8116487361755836e-06,
+      "loss": 0.1861,
+      "step": 498
+    },
+    {
+      "epoch": 0.26,
+      "learning_rate": 4.810869931781381e-06,
+      "loss": 0.2204,
+      "step": 499
+    },
+    {
+      "epoch": 0.26,
+      "learning_rate": 4.810089583880672e-06,
+      "loss": 0.1783,
+      "step": 500
+    },
+    {
+      "epoch": 0.26,
+      "eval_loss": 0.24596241116523743,
+      "eval_runtime": 117.801,
+      "eval_samples_per_second": 26.502,
+      "eval_steps_per_second": 8.837,
+      "step": 500
+    },
+    {
+      "epoch": 0.26,
+      "learning_rate": 4.809307692994674e-06,
+      "loss": 0.1955,
+      "step": 501
+    },
+    {
+      "epoch": 0.26,
+      "learning_rate": 4.80852425964564e-06,
+      "loss": 0.1945,
+      "step": 502
+    },
+    {
+      "epoch": 0.26,
+      "learning_rate": 4.807739284356849e-06,
+      "loss": 0.2011,
+      "step": 503
+    },
+    {
+      "epoch": 0.26,
+      "learning_rate": 4.806952767652614e-06,
+      "loss": 0.1991,
+      "step": 504
+    },
+    {
+      "epoch": 0.26,
+      "learning_rate": 4.806164710058274e-06,
+      "loss": 0.179,
+      "step": 505
+    },
+    {
+      "epoch": 0.26,
+      "learning_rate": 4.8053751121002e-06,
+      "loss": 0.1851,
+      "step": 506
+    },
+    {
+      "epoch": 0.26,
+      "learning_rate": 4.80458397430579e-06,
+      "loss": 0.1903,
+      "step": 507
+    },
+    {
+      "epoch": 0.26,
+      "learning_rate": 4.803791297203471e-06,
+      "loss": 0.1886,
+      "step": 508
+    },
+    {
+      "epoch": 0.26,
+      "learning_rate": 4.802997081322698e-06,
+      "loss": 0.1885,
+      "step": 509
+    },
+    {
+      "epoch": 0.26,
+      "learning_rate": 4.802201327193954e-06,
+      "loss": 0.1631,
+      "step": 510
+    },
+    {
+      "epoch": 0.26,
+      "learning_rate": 4.80140403534875e-06,
+      "loss": 0.172,
+      "step": 511
+    },
+    {
+      "epoch": 0.26,
+      "learning_rate": 4.800605206319624e-06,
+      "loss": 0.2098,
+      "step": 512
+    },
+    {
+      "epoch": 0.27,
+      "learning_rate": 4.79980484064014e-06,
+      "loss": 0.1893,
+      "step": 513
+    },
+    {
+      "epoch": 0.27,
+      "learning_rate": 4.799002938844889e-06,
+      "loss": 0.191,
+      "step": 514
+    },
+    {
+      "epoch": 0.27,
+      "learning_rate": 4.798199501469487e-06,
+      "loss": 0.1875,
+      "step": 515
+    },
+    {
+      "epoch": 0.27,
+      "learning_rate": 4.797394529050577e-06,
+      "loss": 0.1837,
+      "step": 516
+    },
+    {
+      "epoch": 0.27,
+      "learning_rate": 4.796588022125827e-06,
+      "loss": 0.1616,
+      "step": 517
+    },
+    {
+      "epoch": 0.27,
+      "learning_rate": 4.795779981233928e-06,
+      "loss": 0.179,
+      "step": 518
+    },
+    {
+      "epoch": 0.27,
+      "learning_rate": 4.7949704069145985e-06,
+      "loss": 0.2237,
+      "step": 519
+    },
+    {
+      "epoch": 0.27,
+      "learning_rate": 4.794159299708581e-06,
+      "loss": 0.1905,
+      "step": 520
+    },
+    {
+      "epoch": 0.27,
+      "learning_rate": 4.7933466601576385e-06,
+      "loss": 0.1673,
+      "step": 521
+    },
+    {
+      "epoch": 0.27,
+      "learning_rate": 4.792532488804562e-06,
+      "loss": 0.21,
+      "step": 522
+    },
+    {
+      "epoch": 0.27,
+      "learning_rate": 4.791716786193162e-06,
+      "loss": 0.176,
+      "step": 523
+    },
+    {
+      "epoch": 0.27,
+      "learning_rate": 4.790899552868274e-06,
+      "loss": 0.1935,
+      "step": 524
+    },
+    {
+      "epoch": 0.27,
+      "learning_rate": 4.790080789375756e-06,
+      "loss": 0.1917,
+      "step": 525
+    },
+    {
+      "epoch": 0.27,
+      "learning_rate": 4.789260496262485e-06,
+      "loss": 0.1915,
+      "step": 526
+    },
+    {
+      "epoch": 0.27,
+      "learning_rate": 4.788438674076364e-06,
+      "loss": 0.1932,
+      "step": 527
+    },
+    {
+      "epoch": 0.27,
+      "learning_rate": 4.787615323366315e-06,
+      "loss": 0.1679,
+      "step": 528
+    },
+    {
+      "epoch": 0.27,
+      "learning_rate": 4.78679044468228e-06,
+      "loss": 0.182,
+      "step": 529
+    },
+    {
+      "epoch": 0.27,
+      "learning_rate": 4.785964038575224e-06,
+      "loss": 0.1798,
+      "step": 530
+    },
+    {
+      "epoch": 0.27,
+      "learning_rate": 4.785136105597129e-06,
+      "loss": 0.1737,
+      "step": 531
+    },
+    {
+      "epoch": 0.28,
+      "learning_rate": 4.7843066463010015e-06,
+      "loss": 0.1874,
+      "step": 532
+    },
+    {
+      "epoch": 0.28,
+      "learning_rate": 4.783475661240863e-06,
+      "loss": 0.2011,
+      "step": 533
+    },
+    {
+      "epoch": 0.28,
+      "learning_rate": 4.782643150971757e-06,
+      "loss": 0.2158,
+      "step": 534
+    },
+    {
+      "epoch": 0.28,
+      "learning_rate": 4.781809116049744e-06,
+      "loss": 0.1784,
+      "step": 535
+    },
+    {
+      "epoch": 0.28,
+      "learning_rate": 4.780973557031903e-06,
+      "loss": 0.1662,
+      "step": 536
+    },
+    {
+      "epoch": 0.28,
+      "learning_rate": 4.7801364744763325e-06,
+      "loss": 0.196,
+      "step": 537
+    },
+    {
+      "epoch": 0.28,
+      "learning_rate": 4.7792978689421476e-06,
+      "loss": 0.1993,
+      "step": 538
+    },
+    {
+      "epoch": 0.28,
+      "learning_rate": 4.778457740989481e-06,
+      "loss": 0.1763,
+      "step": 539
+    },
+    {
+      "epoch": 0.28,
+      "learning_rate": 4.77761609117948e-06,
+      "loss": 0.1921,
+      "step": 540
+    },
+    {
+      "epoch": 0.28,
+      "learning_rate": 4.776772920074313e-06,
+      "loss": 0.1659,
+      "step": 541
+    },
+    {
+      "epoch": 0.28,
+      "learning_rate": 4.77592822823716e-06,
+      "loss": 0.1707,
+      "step": 542
+    },
+    {
+      "epoch": 0.28,
+      "learning_rate": 4.775082016232219e-06,
+      "loss": 0.1768,
+      "step": 543
+    },
+    {
+      "epoch": 0.28,
+      "learning_rate": 4.774234284624704e-06,
+      "loss": 0.2185,
+      "step": 544
+    },
+    {
+      "epoch": 0.28,
+      "learning_rate": 4.773385033980842e-06,
+      "loss": 0.1818,
+      "step": 545
+    },
+    {
+      "epoch": 0.28,
+      "learning_rate": 4.772534264867876e-06,
+      "loss": 0.203,
+      "step": 546
+    },
+    {
+      "epoch": 0.28,
+      "learning_rate": 4.771681977854062e-06,
+      "loss": 0.1771,
+      "step": 547
+    },
+    {
+      "epoch": 0.28,
+      "learning_rate": 4.770828173508673e-06,
+      "loss": 0.1985,
+      "step": 548
+    },
+    {
+      "epoch": 0.28,
+      "learning_rate": 4.76997285240199e-06,
+      "loss": 0.2099,
+      "step": 549
+    },
+    {
+      "epoch": 0.28,
+      "learning_rate": 4.769116015105313e-06,
+      "loss": 0.2257,
+      "step": 550
+    },
+    {
+      "epoch": 0.28,
+      "eval_loss": 0.24235348403453827,
+      "eval_runtime": 117.6205,
+      "eval_samples_per_second": 26.543,
+      "eval_steps_per_second": 8.85,
+      "step": 550
+    },
+    {
+      "epoch": 0.29,
+      "learning_rate": 4.7682576621909505e-06,
+      "loss": 0.1989,
+      "step": 551
+    },
+    {
+      "epoch": 0.29,
+      "learning_rate": 4.767397794232226e-06,
+      "loss": 0.209,
+      "step": 552
+    },
+    {
+      "epoch": 0.29,
+      "learning_rate": 4.766536411803472e-06,
+      "loss": 0.1762,
+      "step": 553
+    },
+    {
+      "epoch": 0.29,
+      "learning_rate": 4.7656735154800345e-06,
+      "loss": 0.1518,
+      "step": 554
+    },
+    {
+      "epoch": 0.29,
+      "learning_rate": 4.764809105838273e-06,
+      "loss": 0.1784,
+      "step": 555
+    },
+    {
+      "epoch": 0.29,
+      "learning_rate": 4.763943183455553e-06,
+      "loss": 0.1917,
+      "step": 556
+    },
+    {
+      "epoch": 0.29,
+      "learning_rate": 4.763075748910253e-06,
+      "loss": 0.1659,
+      "step": 557
+    },
+    {
+      "epoch": 0.29,
+      "learning_rate": 4.762206802781762e-06,
+      "loss": 0.2104,
+      "step": 558
+    },
+    {
+      "epoch": 0.29,
+      "learning_rate": 4.761336345650478e-06,
+      "loss": 0.2023,
+      "step": 559
+    },
+    {
+      "epoch": 0.29,
+      "learning_rate": 4.760464378097807e-06,
+      "loss": 0.1953,
+      "step": 560
+    },
+    {
+      "epoch": 0.29,
+      "learning_rate": 4.759590900706165e-06,
+      "loss": 0.1743,
+      "step": 561
+    },
+    {
+      "epoch": 0.29,
+      "learning_rate": 4.758715914058978e-06,
+      "loss": 0.1905,
+      "step": 562
+    },
+    {
+      "epoch": 0.29,
+      "learning_rate": 4.757839418740677e-06,
+      "loss": 0.1608,
+      "step": 563
+    },
+    {
+      "epoch": 0.29,
+      "learning_rate": 4.756961415336703e-06,
+      "loss": 0.1855,
+      "step": 564
+    },
+    {
+      "epoch": 0.29,
+      "learning_rate": 4.7560819044335035e-06,
+      "loss": 0.1795,
+      "step": 565
+    },
+    {
+      "epoch": 0.29,
+      "learning_rate": 4.755200886618532e-06,
+      "loss": 0.1872,
+      "step": 566
+    },
+    {
+      "epoch": 0.29,
+      "learning_rate": 4.7543183624802514e-06,
+      "loss": 0.1835,
+      "step": 567
+    },
+    {
+      "epoch": 0.29,
+      "learning_rate": 4.753434332608127e-06,
+      "loss": 0.177,
+      "step": 568
+    },
+    {
+      "epoch": 0.29,
+      "learning_rate": 4.752548797592632e-06,
+      "loss": 0.1673,
+      "step": 569
+    },
+    {
+      "epoch": 0.29,
+      "learning_rate": 4.751661758025245e-06,
+      "loss": 0.1697,
+      "step": 570
+    },
+    {
+      "epoch": 0.3,
+      "learning_rate": 4.75077321449845e-06,
+      "loss": 0.2244,
+      "step": 571
+    },
+    {
+      "epoch": 0.3,
+      "learning_rate": 4.749883167605732e-06,
+      "loss": 0.1857,
+      "step": 572
+    },
+    {
+      "epoch": 0.3,
+      "learning_rate": 4.748991617941585e-06,
+      "loss": 0.1842,
+      "step": 573
+    },
+    {
+      "epoch": 0.3,
+      "learning_rate": 4.748098566101503e-06,
+      "loss": 0.1861,
+      "step": 574
+    },
+    {
+      "epoch": 0.3,
+      "learning_rate": 4.747204012681987e-06,
+      "loss": 0.1831,
+      "step": 575
+    },
+    {
+      "epoch": 0.3,
+      "learning_rate": 4.746307958280536e-06,
+      "loss": 0.198,
+      "step": 576
+    },
+    {
+      "epoch": 0.3,
+      "learning_rate": 4.745410403495657e-06,
+      "loss": 0.1644,
+      "step": 577
+    },
+    {
+      "epoch": 0.3,
+      "learning_rate": 4.744511348926855e-06,
+      "loss": 0.1803,
+      "step": 578
+    },
+    {
+      "epoch": 0.3,
+      "learning_rate": 4.743610795174639e-06,
+      "loss": 0.1792,
+      "step": 579
+    },
+    {
+      "epoch": 0.3,
+      "learning_rate": 4.742708742840517e-06,
+      "loss": 0.1927,
+      "step": 580
+    },
+    {
+      "epoch": 0.3,
+      "learning_rate": 4.741805192527002e-06,
+      "loss": 0.1828,
+      "step": 581
+    },
+    {
+      "epoch": 0.3,
+      "learning_rate": 4.740900144837605e-06,
+      "loss": 0.1717,
+      "step": 582
+    },
+    {
+      "epoch": 0.3,
+      "learning_rate": 4.739993600376836e-06,
+      "loss": 0.1812,
+      "step": 583
+    },
+    {
+      "epoch": 0.3,
+      "learning_rate": 4.739085559750206e-06,
+      "loss": 0.1727,
+      "step": 584
+    },
+    {
+      "epoch": 0.3,
+      "learning_rate": 4.738176023564226e-06,
+      "loss": 0.1815,
+      "step": 585
+    },
+    {
+      "epoch": 0.3,
+      "learning_rate": 4.737264992426406e-06,
+      "loss": 0.1734,
+      "step": 586
+    },
+    {
+      "epoch": 0.3,
+      "learning_rate": 4.736352466945253e-06,
+      "loss": 0.1664,
+      "step": 587
+    },
+    {
+      "epoch": 0.3,
+      "learning_rate": 4.735438447730273e-06,
+      "loss": 0.1846,
+      "step": 588
+    },
+    {
+      "epoch": 0.3,
+      "learning_rate": 4.734522935391969e-06,
+      "loss": 0.1948,
+      "step": 589
+    },
+    {
+      "epoch": 0.31,
+      "learning_rate": 4.733605930541843e-06,
+      "loss": 0.1897,
+      "step": 590
+    },
+    {
+      "epoch": 0.31,
+      "learning_rate": 4.732687433792393e-06,
+      "loss": 0.1637,
+      "step": 591
+    },
+    {
+      "epoch": 0.31,
+      "learning_rate": 4.731767445757112e-06,
+      "loss": 0.171,
+      "step": 592
+    },
+    {
+      "epoch": 0.31,
+      "learning_rate": 4.7308459670504926e-06,
+      "loss": 0.1746,
+      "step": 593
+    },
+    {
+      "epoch": 0.31,
+      "learning_rate": 4.729922998288019e-06,
+      "loss": 0.1675,
+      "step": 594
+    },
+    {
+      "epoch": 0.31,
+      "learning_rate": 4.728998540086174e-06,
+      "loss": 0.1831,
+      "step": 595
+    },
+    {
+      "epoch": 0.31,
+      "learning_rate": 4.728072593062433e-06,
+      "loss": 0.1796,
+      "step": 596
+    },
+    {
+      "epoch": 0.31,
+      "learning_rate": 4.727145157835265e-06,
+      "loss": 0.2311,
+      "step": 597
+    },
+    {
+      "epoch": 0.31,
+      "learning_rate": 4.726216235024139e-06,
+      "loss": 0.2086,
+      "step": 598
+    },
+    {
+      "epoch": 0.31,
+      "learning_rate": 4.7252858252495105e-06,
+      "loss": 0.1648,
+      "step": 599
+    },
+    {
+      "epoch": 0.31,
+      "learning_rate": 4.724353929132831e-06,
+      "loss": 0.1785,
+      "step": 600
+    },
+    {
+      "epoch": 0.31,
+      "eval_loss": 0.23951774835586548,
+      "eval_runtime": 117.6786,
+      "eval_samples_per_second": 26.53,
+      "eval_steps_per_second": 8.846,
+      "step": 600
+    },
+    {
+      "epoch": 0.31,
+      "learning_rate": 4.723420547296544e-06,
+      "loss": 0.1814,
+      "step": 601
+    },
+    {
+      "epoch": 0.31,
+      "learning_rate": 4.7224856803640885e-06,
+      "loss": 0.1789,
+      "step": 602
+    },
+    {
+      "epoch": 0.31,
+      "learning_rate": 4.7215493289598914e-06,
+      "loss": 0.1818,
+      "step": 603
+    },
+    {
+      "epoch": 0.31,
+      "learning_rate": 4.7206114937093726e-06,
+      "loss": 0.1817,
+      "step": 604
+    },
+    {
+      "epoch": 0.31,
+      "learning_rate": 4.719672175238945e-06,
+      "loss": 0.1803,
+      "step": 605
+    },
+    {
+      "epoch": 0.31,
+      "learning_rate": 4.718731374176008e-06,
+      "loss": 0.1961,
+      "step": 606
+    },
+    {
+      "epoch": 0.31,
+      "learning_rate": 4.717789091148955e-06,
+      "loss": 0.2004,
+      "step": 607
+    },
+    {
+      "epoch": 0.31,
+      "learning_rate": 4.7168453267871695e-06,
+      "loss": 0.1732,
+      "step": 608
+    },
+    {
+      "epoch": 0.32,
+      "learning_rate": 4.715900081721021e-06,
+      "loss": 0.1821,
+      "step": 609
+    },
+    {
+      "epoch": 0.32,
+      "learning_rate": 4.7149533565818705e-06,
+      "loss": 0.1606,
+      "step": 610
+    },
+    {
+      "epoch": 0.32,
+      "learning_rate": 4.714005152002067e-06,
+      "loss": 0.1661,
+      "step": 611
+    },
+    {
+      "epoch": 0.32,
+      "learning_rate": 4.713055468614947e-06,
+      "loss": 0.1655,
+      "step": 612
+    },
+    {
+      "epoch": 0.32,
+      "learning_rate": 4.7121043070548376e-06,
+      "loss": 0.1843,
+      "step": 613
+    },
+    {
+      "epoch": 0.32,
+      "learning_rate": 4.711151667957049e-06,
+      "loss": 0.1625,
+      "step": 614
+    },
+    {
+      "epoch": 0.32,
+      "learning_rate": 4.710197551957882e-06,
+      "loss": 0.1728,
+      "step": 615
+    },
+    {
+      "epoch": 0.32,
+      "learning_rate": 4.70924195969462e-06,
+      "loss": 0.158,
+      "step": 616
+    },
+    {
+      "epoch": 0.32,
+      "learning_rate": 4.7082848918055375e-06,
+      "loss": 0.1937,
+      "step": 617
+    },
+    {
+      "epoch": 0.32,
+      "learning_rate": 4.707326348929889e-06,
+      "loss": 0.1736,
+      "step": 618
+    },
+    {
+      "epoch": 0.32,
+      "learning_rate": 4.70636633170792e-06,
+      "loss": 0.2109,
+      "step": 619
+    },
+    {
+      "epoch": 0.32,
+      "learning_rate": 4.705404840780855e-06,
+      "loss": 0.1767,
+      "step": 620
+    },
+    {
+      "epoch": 0.32,
+      "learning_rate": 4.704441876790907e-06,
+      "loss": 0.1792,
+      "step": 621
+    },
+    {
+      "epoch": 0.32,
+      "learning_rate": 4.703477440381272e-06,
+      "loss": 0.1906,
+      "step": 622
+    },
+    {
+      "epoch": 0.32,
+      "learning_rate": 4.702511532196128e-06,
+      "loss": 0.1643,
+      "step": 623
+    },
+    {
+      "epoch": 0.32,
+      "learning_rate": 4.701544152880637e-06,
+      "loss": 0.2475,
+      "step": 624
+    },
+    {
+      "epoch": 0.32,
+      "learning_rate": 4.700575303080944e-06,
+      "loss": 0.1845,
+      "step": 625
+    },
+    {
+      "epoch": 0.32,
+      "learning_rate": 4.699604983444176e-06,
+      "loss": 0.1939,
+      "step": 626
+    },
+    {
+      "epoch": 0.32,
+      "learning_rate": 4.698633194618442e-06,
+      "loss": 0.1664,
+      "step": 627
+    },
+    {
+      "epoch": 0.32,
+      "learning_rate": 4.697659937252831e-06,
+      "loss": 0.2106,
+      "step": 628
+    },
+    {
+      "epoch": 0.33,
+      "learning_rate": 4.696685211997414e-06,
+      "loss": 0.1881,
+      "step": 629
+    },
+    {
+      "epoch": 0.33,
+      "learning_rate": 4.695709019503242e-06,
+      "loss": 0.1486,
+      "step": 630
+    },
+    {
+      "epoch": 0.33,
+      "learning_rate": 4.694731360422348e-06,
+      "loss": 0.1873,
+      "step": 631
+    },
+    {
+      "epoch": 0.33,
+      "learning_rate": 4.69375223540774e-06,
+      "loss": 0.1664,
+      "step": 632
+    },
+    {
+      "epoch": 0.33,
+      "learning_rate": 4.692771645113411e-06,
+      "loss": 0.1567,
+      "step": 633
+    },
+    {
+      "epoch": 0.33,
+      "learning_rate": 4.691789590194327e-06,
+      "loss": 0.1966,
+      "step": 634
+    },
+    {
+      "epoch": 0.33,
+      "learning_rate": 4.690806071306438e-06,
+      "loss": 0.1699,
+      "step": 635
+    },
+    {
+      "epoch": 0.33,
+      "learning_rate": 4.689821089106666e-06,
+      "loss": 0.2049,
+      "step": 636
+    },
+    {
+      "epoch": 0.33,
+      "learning_rate": 4.6888346442529145e-06,
+      "loss": 0.187,
+      "step": 637
+    },
+    {
+      "epoch": 0.33,
+      "learning_rate": 4.6878467374040625e-06,
+      "loss": 0.1834,
+      "step": 638
+    },
+    {
+      "epoch": 0.33,
+      "learning_rate": 4.686857369219966e-06,
+      "loss": 0.1866,
+      "step": 639
+    },
+    {
+      "epoch": 0.33,
+      "learning_rate": 4.685866540361456e-06,
+      "loss": 0.1815,
+      "step": 640
+    },
+    {
+      "epoch": 0.33,
+      "learning_rate": 4.68487425149034e-06,
+      "loss": 0.153,
+      "step": 641
+    },
+    {
+      "epoch": 0.33,
+      "learning_rate": 4.683880503269402e-06,
+      "loss": 0.1891,
+      "step": 642
+    },
+    {
+      "epoch": 0.33,
+      "learning_rate": 4.682885296362397e-06,
+      "loss": 0.1736,
+      "step": 643
+    },
+    {
+      "epoch": 0.33,
+      "learning_rate": 4.681888631434058e-06,
+      "loss": 0.1838,
+      "step": 644
+    },
+    {
+      "epoch": 0.33,
+      "learning_rate": 4.68089050915009e-06,
+      "loss": 0.192,
+      "step": 645
+    },
+    {
+      "epoch": 0.33,
+      "learning_rate": 4.679890930177174e-06,
+      "loss": 0.1646,
+      "step": 646
+    },
+    {
+      "epoch": 0.33,
+      "learning_rate": 4.678889895182958e-06,
+      "loss": 0.1938,
+      "step": 647
+    },
+    {
+      "epoch": 0.34,
+      "learning_rate": 4.677887404836069e-06,
+      "loss": 0.1739,
+      "step": 648
+    },
+    {
+      "epoch": 0.34,
+      "learning_rate": 4.676883459806104e-06,
+      "loss": 0.2197,
+      "step": 649
+    },
+    {
+      "epoch": 0.34,
+      "learning_rate": 4.675878060763629e-06,
+      "loss": 0.1523,
+      "step": 650
+    },
+    {
+      "epoch": 0.34,
+      "eval_loss": 0.23699061572551727,
+      "eval_runtime": 117.3974,
+      "eval_samples_per_second": 26.593,
+      "eval_steps_per_second": 8.867,
+      "step": 650
+    },
+    {
+      "epoch": 0.34,
+      "learning_rate": 4.674871208380185e-06,
+      "loss": 0.1706,
+      "step": 651
+    },
+    {
+      "epoch": 0.34,
+      "learning_rate": 4.673862903328281e-06,
+      "loss": 0.1696,
+      "step": 652
+    },
+    {
+      "epoch": 0.34,
+      "learning_rate": 4.672853146281399e-06,
+      "loss": 0.1809,
+      "step": 653
+    },
+    {
+      "epoch": 0.34,
+      "learning_rate": 4.6718419379139865e-06,
+      "loss": 0.1834,
+      "step": 654
+    },
+    {
+      "epoch": 0.34,
+      "learning_rate": 4.670829278901465e-06,
+      "loss": 0.1824,
+      "step": 655
+    },
+    {
+      "epoch": 0.34,
+      "learning_rate": 4.669815169920222e-06,
+      "loss": 0.1843,
+      "step": 656
+    },
+    {
+      "epoch": 0.34,
+      "learning_rate": 4.668799611647616e-06,
+      "loss": 0.2005,
+      "step": 657
+    },
+    {
+      "epoch": 0.34,
+      "learning_rate": 4.667782604761969e-06,
+      "loss": 0.1572,
+      "step": 658
+    },
+    {
+      "epoch": 0.34,
+      "learning_rate": 4.666764149942577e-06,
+      "loss": 0.1679,
+      "step": 659
+    },
+    {
+      "epoch": 0.34,
+      "learning_rate": 4.665744247869697e-06,
+      "loss": 0.1704,
+      "step": 660
+    },
+    {
+      "epoch": 0.34,
+      "learning_rate": 4.664722899224557e-06,
+      "loss": 0.1838,
+      "step": 661
+    },
+    {
+      "epoch": 0.34,
+      "learning_rate": 4.663700104689348e-06,
+      "loss": 0.2192,
+      "step": 662
+    },
+    {
+      "epoch": 0.34,
+      "learning_rate": 4.66267586494723e-06,
+      "loss": 0.1848,
+      "step": 663
+    },
+    {
+      "epoch": 0.34,
+      "learning_rate": 4.6616501806823264e-06,
+      "loss": 0.2143,
+      "step": 664
+    },
+    {
+      "epoch": 0.34,
+      "learning_rate": 4.660623052579725e-06,
+      "loss": 0.1834,
+      "step": 665
+    },
+    {
+      "epoch": 0.34,
+      "learning_rate": 4.659594481325478e-06,
+      "loss": 0.1612,
+      "step": 666
+    },
+    {
+      "epoch": 0.35,
+      "learning_rate": 4.658564467606604e-06,
+      "loss": 0.1909,
+      "step": 667
+    },
+    {
+      "epoch": 0.35,
+      "learning_rate": 4.657533012111081e-06,
+      "loss": 0.1712,
+      "step": 668
+    },
+    {
+      "epoch": 0.35,
+      "learning_rate": 4.656500115527856e-06,
+      "loss": 0.1995,
+      "step": 669
+    },
+    {
+      "epoch": 0.35,
+      "learning_rate": 4.655465778546831e-06,
+      "loss": 0.1892,
+      "step": 670
+    },
+    {
+      "epoch": 0.35,
+      "learning_rate": 4.654430001858874e-06,
+      "loss": 0.1869,
+      "step": 671
+    },
+    {
+      "epoch": 0.35,
+      "learning_rate": 4.653392786155817e-06,
+      "loss": 0.1942,
+      "step": 672
+    },
+    {
+      "epoch": 0.35,
+      "learning_rate": 4.65235413213045e-06,
+      "loss": 0.1888,
+      "step": 673
+    },
+    {
+      "epoch": 0.35,
+      "learning_rate": 4.651314040476523e-06,
+      "loss": 0.1914,
+      "step": 674
+    },
+    {
+      "epoch": 0.35,
+      "learning_rate": 4.6502725118887486e-06,
+      "loss": 0.2008,
+      "step": 675
+    },
+    {
+      "epoch": 0.35,
+      "learning_rate": 4.649229547062798e-06,
+      "loss": 0.1597,
+      "step": 676
+    },
+    {
+      "epoch": 0.35,
+      "learning_rate": 4.648185146695303e-06,
+      "loss": 0.1646,
+      "step": 677
+    },
+    {
+      "epoch": 0.35,
+      "learning_rate": 4.647139311483851e-06,
+      "loss": 0.1888,
+      "step": 678
+    },
+    {
+      "epoch": 0.35,
+      "learning_rate": 4.6460920421269914e-06,
+      "loss": 0.2293,
+      "step": 679
+    },
+    {
+      "epoch": 0.35,
+      "learning_rate": 4.645043339324231e-06,
+      "loss": 0.1824,
+      "step": 680
+    },
+    {
+      "epoch": 0.35,
+      "learning_rate": 4.643993203776031e-06,
+      "loss": 0.1877,
+      "step": 681
+    },
+    {
+      "epoch": 0.35,
+      "learning_rate": 4.642941636183813e-06,
+      "loss": 0.1601,
+      "step": 682
+    },
+    {
+      "epoch": 0.35,
+      "learning_rate": 4.641888637249952e-06,
+      "loss": 0.1711,
+      "step": 683
+    },
+    {
+      "epoch": 0.35,
+      "learning_rate": 4.6408342076777845e-06,
+      "loss": 0.1947,
+      "step": 684
+    },
+    {
+      "epoch": 0.35,
+      "learning_rate": 4.639778348171595e-06,
+      "loss": 0.174,
+      "step": 685
+    },
+    {
+      "epoch": 0.35,
+      "learning_rate": 4.63872105943663e-06,
+      "loss": 0.1872,
+      "step": 686
+    },
+    {
+      "epoch": 0.36,
+      "learning_rate": 4.637662342179085e-06,
+      "loss": 0.186,
+      "step": 687
+    },
+    {
+      "epoch": 0.36,
+      "learning_rate": 4.636602197106115e-06,
+      "loss": 0.1751,
+      "step": 688
+    },
+    {
+      "epoch": 0.36,
+      "learning_rate": 4.635540624925824e-06,
+      "loss": 0.1964,
+      "step": 689
+    },
+    {
+      "epoch": 0.36,
+      "learning_rate": 4.634477626347271e-06,
+      "loss": 0.1778,
+      "step": 690
+    },
+    {
+      "epoch": 0.36,
+      "learning_rate": 4.6334132020804705e-06,
+      "loss": 0.1822,
+      "step": 691
+    },
+    {
+      "epoch": 0.36,
+      "learning_rate": 4.6323473528363835e-06,
+      "loss": 0.184,
+      "step": 692
+    },
+    {
+      "epoch": 0.36,
+      "learning_rate": 4.631280079326927e-06,
+      "loss": 0.1761,
+      "step": 693
+    },
+    {
+      "epoch": 0.36,
+      "learning_rate": 4.63021138226497e-06,
+      "loss": 0.1765,
+      "step": 694
+    },
+    {
+      "epoch": 0.36,
+      "learning_rate": 4.629141262364329e-06,
+      "loss": 0.1691,
+      "step": 695
+    },
+    {
+      "epoch": 0.36,
+      "learning_rate": 4.628069720339773e-06,
+      "loss": 0.1731,
+      "step": 696
+    },
+    {
+      "epoch": 0.36,
+      "learning_rate": 4.626996756907019e-06,
+      "loss": 0.2131,
+      "step": 697
+    },
+    {
+      "epoch": 0.36,
+      "learning_rate": 4.6259223727827375e-06,
+      "loss": 0.1864,
+      "step": 698
+    },
+    {
+      "epoch": 0.36,
+      "learning_rate": 4.624846568684544e-06,
+      "loss": 0.1818,
+      "step": 699
+    },
+    {
+      "epoch": 0.36,
+      "learning_rate": 4.623769345331002e-06,
+      "loss": 0.1762,
+      "step": 700
+    },
+    {
+      "epoch": 0.36,
+      "eval_loss": 0.23404967784881592,
+      "eval_runtime": 117.7277,
+      "eval_samples_per_second": 26.519,
+      "eval_steps_per_second": 8.842,
+      "step": 700
+    },
+    {
+      "epoch": 0.36,
+      "learning_rate": 4.6226907034416265e-06,
+      "loss": 0.1694,
+      "step": 701
+    },
+    {
+      "epoch": 0.36,
+      "learning_rate": 4.621610643736878e-06,
+      "loss": 0.1787,
+      "step": 702
+    },
+    {
+      "epoch": 0.36,
+      "learning_rate": 4.620529166938164e-06,
+      "loss": 0.1834,
+      "step": 703
+    },
+    {
+      "epoch": 0.36,
+      "learning_rate": 4.619446273767837e-06,
+      "loss": 0.1652,
+      "step": 704
+    },
+    {
+      "epoch": 0.36,
+      "learning_rate": 4.6183619649492e-06,
+      "loss": 0.1564,
+      "step": 705
+    },
+    {
+      "epoch": 0.37,
+      "learning_rate": 4.617276241206497e-06,
+      "loss": 0.1555,
+      "step": 706
+    },
+    {
+      "epoch": 0.37,
+      "learning_rate": 4.616189103264919e-06,
+      "loss": 0.214,
+      "step": 707
+    },
+    {
+      "epoch": 0.37,
+      "learning_rate": 4.615100551850601e-06,
+      "loss": 0.1686,
+      "step": 708
+    },
+    {
+      "epoch": 0.37,
+      "learning_rate": 4.614010587690624e-06,
+      "loss": 0.161,
+      "step": 709
+    },
+    {
+      "epoch": 0.37,
+      "learning_rate": 4.6129192115130096e-06,
+      "loss": 0.165,
+      "step": 710
+    },
+    {
+      "epoch": 0.37,
+      "learning_rate": 4.611826424046725e-06,
+      "loss": 0.1672,
+      "step": 711
+    },
+    {
+      "epoch": 0.37,
+      "learning_rate": 4.610732226021679e-06,
+      "loss": 0.1788,
+      "step": 712
+    },
+    {
+      "epoch": 0.37,
+      "learning_rate": 4.6096366181687225e-06,
+      "loss": 0.1726,
+      "step": 713
+    },
+    {
+      "epoch": 0.37,
+      "learning_rate": 4.608539601219648e-06,
+      "loss": 0.1742,
+      "step": 714
+    },
+    {
+      "epoch": 0.37,
+      "learning_rate": 4.607441175907189e-06,
+      "loss": 0.177,
+      "step": 715
+    },
+    {
+      "epoch": 0.37,
+      "learning_rate": 4.606341342965022e-06,
+      "loss": 0.1682,
+      "step": 716
+    },
+    {
+      "epoch": 0.37,
+      "learning_rate": 4.6052401031277595e-06,
+      "loss": 0.1827,
+      "step": 717
+    },
+    {
+      "epoch": 0.37,
+      "learning_rate": 4.604137457130958e-06,
+      "loss": 0.1739,
+      "step": 718
+    },
+    {
+      "epoch": 0.37,
+      "learning_rate": 4.60303340571111e-06,
+      "loss": 0.1738,
+      "step": 719
+    },
+    {
+      "epoch": 0.37,
+      "learning_rate": 4.601927949605649e-06,
+      "loss": 0.1664,
+      "step": 720
+    },
+    {
+      "epoch": 0.37,
+      "learning_rate": 4.600821089552945e-06,
+      "loss": 0.1673,
+      "step": 721
+    },
+    {
+      "epoch": 0.37,
+      "learning_rate": 4.599712826292305e-06,
+      "loss": 0.1814,
+      "step": 722
+    },
+    {
+      "epoch": 0.37,
+      "learning_rate": 4.598603160563978e-06,
+      "loss": 0.1648,
+      "step": 723
+    },
+    {
+      "epoch": 0.37,
+      "learning_rate": 4.597492093109144e-06,
+      "loss": 0.2179,
+      "step": 724
+    },
+    {
+      "epoch": 0.38,
+      "learning_rate": 4.596379624669923e-06,
+      "loss": 0.1678,
+      "step": 725
+    },
+    {
+      "epoch": 0.38,
+      "learning_rate": 4.595265755989368e-06,
+      "loss": 0.1721,
+      "step": 726
+    },
+    {
+      "epoch": 0.38,
+      "learning_rate": 4.594150487811471e-06,
+      "loss": 0.2122,
+      "step": 727
+    },
+    {
+      "epoch": 0.38,
+      "learning_rate": 4.593033820881153e-06,
+      "loss": 0.1799,
+      "step": 728
+    },
+    {
+      "epoch": 0.38,
+      "learning_rate": 4.591915755944277e-06,
+      "loss": 0.1738,
+      "step": 729
+    },
+    {
+      "epoch": 0.38,
+      "learning_rate": 4.590796293747634e-06,
+      "loss": 0.1818,
+      "step": 730
+    },
+    {
+      "epoch": 0.38,
+      "learning_rate": 4.589675435038948e-06,
+      "loss": 0.1917,
+      "step": 731
+    },
+    {
+      "epoch": 0.38,
+      "learning_rate": 4.588553180566879e-06,
+      "loss": 0.1781,
+      "step": 732
+    },
+    {
+      "epoch": 0.38,
+      "learning_rate": 4.587429531081019e-06,
+      "loss": 0.1642,
+      "step": 733
+    },
+    {
+      "epoch": 0.38,
+      "learning_rate": 4.586304487331889e-06,
+      "loss": 0.1731,
+      "step": 734
+    },
+    {
+      "epoch": 0.38,
+      "learning_rate": 4.5851780500709445e-06,
+      "loss": 0.2054,
+      "step": 735
+    },
+    {
+      "epoch": 0.38,
+      "learning_rate": 4.584050220050568e-06,
+      "loss": 0.1736,
+      "step": 736
+    },
+    {
+      "epoch": 0.38,
+      "learning_rate": 4.582920998024076e-06,
+      "loss": 0.1976,
+      "step": 737
+    },
+    {
+      "epoch": 0.38,
+      "learning_rate": 4.581790384745715e-06,
+      "loss": 0.1592,
+      "step": 738
+    },
+    {
+      "epoch": 0.38,
+      "learning_rate": 4.580658380970655e-06,
+      "loss": 0.1766,
+      "step": 739
+    },
+    {
+      "epoch": 0.38,
+      "learning_rate": 4.579524987455002e-06,
+      "loss": 0.1675,
+      "step": 740
+    },
+    {
+      "epoch": 0.38,
+      "learning_rate": 4.578390204955785e-06,
+      "loss": 0.172,
+      "step": 741
+    },
+    {
+      "epoch": 0.38,
+      "learning_rate": 4.577254034230965e-06,
+      "loss": 0.1811,
+      "step": 742
+    },
+    {
+      "epoch": 0.38,
+      "learning_rate": 4.576116476039426e-06,
+      "loss": 0.1938,
+      "step": 743
+    },
+    {
+      "epoch": 0.38,
+      "learning_rate": 4.574977531140982e-06,
+      "loss": 0.1736,
+      "step": 744
+    },
+    {
+      "epoch": 0.39,
+      "learning_rate": 4.5738372002963705e-06,
+      "loss": 0.1886,
+      "step": 745
+    },
+    {
+      "epoch": 0.39,
+      "learning_rate": 4.572695484267258e-06,
+      "loss": 0.1804,
+      "step": 746
+    },
+    {
+      "epoch": 0.39,
+      "learning_rate": 4.571552383816234e-06,
+      "loss": 0.1788,
+      "step": 747
+    },
+    {
+      "epoch": 0.39,
+      "learning_rate": 4.570407899706812e-06,
+      "loss": 0.1859,
+      "step": 748
+    },
+    {
+      "epoch": 0.39,
+      "learning_rate": 4.569262032703433e-06,
+      "loss": 0.1652,
+      "step": 749
+    },
+    {
+      "epoch": 0.39,
+      "learning_rate": 4.568114783571457e-06,
+      "loss": 0.1782,
+      "step": 750
+    },
+    {
+      "epoch": 0.39,
+      "eval_loss": 0.231343612074852,
+      "eval_runtime": 117.7778,
+      "eval_samples_per_second": 26.508,
+      "eval_steps_per_second": 8.839,
+      "step": 750
+    },
+    {
+      "epoch": 0.39,
+      "learning_rate": 4.566966153077171e-06,
+      "loss": 0.1585,
+      "step": 751
+    },
+    {
+      "epoch": 0.39,
+      "learning_rate": 4.565816141987783e-06,
+      "loss": 0.2166,
+      "step": 752
+    },
+    {
+      "epoch": 0.39,
+      "learning_rate": 4.564664751071423e-06,
+      "loss": 0.1624,
+      "step": 753
+    },
+    {
+      "epoch": 0.39,
+      "learning_rate": 4.563511981097144e-06,
+      "loss": 0.1897,
+      "step": 754
+    },
+    {
+      "epoch": 0.39,
+      "learning_rate": 4.56235783283492e-06,
+      "loss": 0.1821,
+      "step": 755
+    },
+    {
+      "epoch": 0.39,
+      "learning_rate": 4.561202307055642e-06,
+      "loss": 0.1737,
+      "step": 756
+    },
+    {
+      "epoch": 0.39,
+      "learning_rate": 4.560045404531127e-06,
+      "loss": 0.171,
+      "step": 757
+    },
+    {
+      "epoch": 0.39,
+      "learning_rate": 4.558887126034106e-06,
+      "loss": 0.1592,
+      "step": 758
+    },
+    {
+      "epoch": 0.39,
+      "learning_rate": 4.557727472338234e-06,
+      "loss": 0.1597,
+      "step": 759
+    },
+    {
+      "epoch": 0.39,
+      "learning_rate": 4.556566444218079e-06,
+      "loss": 0.1632,
+      "step": 760
+    },
+    {
+      "epoch": 0.39,
+      "learning_rate": 4.555404042449133e-06,
+      "loss": 0.1517,
+      "step": 761
+    },
+    {
+      "epoch": 0.39,
+      "learning_rate": 4.554240267807801e-06,
+      "loss": 0.1737,
+      "step": 762
+    },
+    {
+      "epoch": 0.39,
+      "learning_rate": 4.553075121071407e-06,
+      "loss": 0.1853,
+      "step": 763
+    },
+    {
+      "epoch": 0.4,
+      "learning_rate": 4.551908603018191e-06,
+      "loss": 0.177,
+      "step": 764
+    },
+    {
+      "epoch": 0.4,
+      "learning_rate": 4.55074071442731e-06,
+      "loss": 0.1741,
+      "step": 765
+    },
+    {
+      "epoch": 0.4,
+      "learning_rate": 4.549571456078834e-06,
+      "loss": 0.166,
+      "step": 766
+    },
+    {
+      "epoch": 0.4,
+      "learning_rate": 4.54840082875375e-06,
+      "loss": 0.1673,
+      "step": 767
+    },
+    {
+      "epoch": 0.4,
+      "learning_rate": 4.5472288332339574e-06,
+      "loss": 0.207,
+      "step": 768
+    },
+    {
+      "epoch": 0.4,
+      "learning_rate": 4.546055470302274e-06,
+      "loss": 0.1664,
+      "step": 769
+    },
+    {
+      "epoch": 0.4,
+      "learning_rate": 4.544880740742425e-06,
+      "loss": 0.1559,
+      "step": 770
+    },
+    {
+      "epoch": 0.4,
+      "learning_rate": 4.543704645339051e-06,
+      "loss": 0.1572,
+      "step": 771
+    },
+    {
+      "epoch": 0.4,
+      "learning_rate": 4.542527184877706e-06,
+      "loss": 0.1616,
+      "step": 772
+    },
+    {
+      "epoch": 0.4,
+      "learning_rate": 4.541348360144856e-06,
+      "loss": 0.1937,
+      "step": 773
+    },
+    {
+      "epoch": 0.4,
+      "learning_rate": 4.5401681719278755e-06,
+      "loss": 0.1825,
+      "step": 774
+    },
+    {
+      "epoch": 0.4,
+      "learning_rate": 4.5389866210150515e-06,
+      "loss": 0.1732,
+      "step": 775
+    },
+    {
+      "epoch": 0.4,
+      "learning_rate": 4.537803708195581e-06,
+      "loss": 0.1791,
+      "step": 776
+    },
+    {
+      "epoch": 0.4,
+      "learning_rate": 4.536619434259571e-06,
+      "loss": 0.1556,
+      "step": 777
+    },
+    {
+      "epoch": 0.4,
+      "learning_rate": 4.535433799998036e-06,
+      "loss": 0.1721,
+      "step": 778
+    },
+    {
+      "epoch": 0.4,
+      "learning_rate": 4.534246806202902e-06,
+      "loss": 0.1709,
+      "step": 779
+    },
+    {
+      "epoch": 0.4,
+      "learning_rate": 4.533058453667001e-06,
+      "loss": 0.1578,
+      "step": 780
+    },
+    {
+      "epoch": 0.4,
+      "learning_rate": 4.5318687431840725e-06,
+      "loss": 0.1576,
+      "step": 781
+    },
+    {
+      "epoch": 0.4,
+      "learning_rate": 4.530677675548764e-06,
+      "loss": 0.1791,
+      "step": 782
+    },
+    {
+      "epoch": 0.41,
+      "learning_rate": 4.529485251556628e-06,
+      "loss": 0.1801,
+      "step": 783
+    },
+    {
+      "epoch": 0.41,
+      "learning_rate": 4.528291472004125e-06,
+      "loss": 0.1837,
+      "step": 784
+    },
+    {
+      "epoch": 0.41,
+      "learning_rate": 4.52709633768862e-06,
+      "loss": 0.1667,
+      "step": 785
+    },
+    {
+      "epoch": 0.41,
+      "learning_rate": 4.525899849408382e-06,
+      "loss": 0.1746,
+      "step": 786
+    },
+    {
+      "epoch": 0.41,
+      "learning_rate": 4.524702007962585e-06,
+      "loss": 0.1659,
+      "step": 787
+    },
+    {
+      "epoch": 0.41,
+      "learning_rate": 4.523502814151308e-06,
+      "loss": 0.1635,
+      "step": 788
+    },
+    {
+      "epoch": 0.41,
+      "learning_rate": 4.5223022687755316e-06,
+      "loss": 0.1688,
+      "step": 789
+    },
+    {
+      "epoch": 0.41,
+      "learning_rate": 4.521100372637139e-06,
+      "loss": 0.1695,
+      "step": 790
+    },
+    {
+      "epoch": 0.41,
+      "learning_rate": 4.519897126538918e-06,
+      "loss": 0.1857,
+      "step": 791
+    },
+    {
+      "epoch": 0.41,
+      "learning_rate": 4.518692531284555e-06,
+      "loss": 0.1766,
+      "step": 792
+    },
+    {
+      "epoch": 0.41,
+      "learning_rate": 4.51748658767864e-06,
+      "loss": 0.1874,
+      "step": 793
+    },
+    {
+      "epoch": 0.41,
+      "learning_rate": 4.516279296526664e-06,
+      "loss": 0.1656,
+      "step": 794
+    },
+    {
+      "epoch": 0.41,
+      "learning_rate": 4.515070658635013e-06,
+      "loss": 0.1862,
+      "step": 795
+    },
+    {
+      "epoch": 0.41,
+      "learning_rate": 4.513860674810979e-06,
+      "loss": 0.2122,
+      "step": 796
+    },
+    {
+      "epoch": 0.41,
+      "learning_rate": 4.51264934586275e-06,
+      "loss": 0.1781,
+      "step": 797
+    },
+    {
+      "epoch": 0.41,
+      "learning_rate": 4.511436672599412e-06,
+      "loss": 0.1826,
+      "step": 798
+    },
+    {
+      "epoch": 0.41,
+      "learning_rate": 4.51022265583095e-06,
+      "loss": 0.166,
+      "step": 799
+    },
+    {
+      "epoch": 0.41,
+      "learning_rate": 4.509007296368246e-06,
+      "loss": 0.2039,
+      "step": 800
+    },
+    {
+      "epoch": 0.41,
+      "eval_loss": 0.2298305481672287,
+      "eval_runtime": 117.939,
+      "eval_samples_per_second": 26.471,
+      "eval_steps_per_second": 8.827,
+      "step": 800
+    },
+    {
+      "epoch": 0.41,
+      "learning_rate": 4.507790595023078e-06,
+      "loss": 0.1955,
+      "step": 801
+    },
+    {
+      "epoch": 0.41,
+      "learning_rate": 4.506572552608122e-06,
+      "loss": 0.16,
+      "step": 802
+    },
+    {
+      "epoch": 0.42,
+      "learning_rate": 4.505353169936949e-06,
+      "loss": 0.1939,
+      "step": 803
+    },
+    {
+      "epoch": 0.42,
+      "learning_rate": 4.5041324478240244e-06,
+      "loss": 0.1846,
+      "step": 804
+    },
+    {
+      "epoch": 0.42,
+      "learning_rate": 4.5029103870847094e-06,
+      "loss": 0.1774,
+      "step": 805
+    },
+    {
+      "epoch": 0.42,
+      "learning_rate": 4.501686988535259e-06,
+      "loss": 0.1619,
+      "step": 806
+    },
+    {
+      "epoch": 0.42,
+      "learning_rate": 4.50046225299282e-06,
+      "loss": 0.171,
+      "step": 807
+    },
+    {
+      "epoch": 0.42,
+      "learning_rate": 4.4992361812754355e-06,
+      "loss": 0.1763,
+      "step": 808
+    },
+    {
+      "epoch": 0.42,
+      "learning_rate": 4.498008774202039e-06,
+      "loss": 0.1701,
+      "step": 809
+    },
+    {
+      "epoch": 0.42,
+      "learning_rate": 4.496780032592457e-06,
+      "loss": 0.1715,
+      "step": 810
+    },
+    {
+      "epoch": 0.42,
+      "learning_rate": 4.495549957267404e-06,
+      "loss": 0.1631,
+      "step": 811
+    },
+    {
+      "epoch": 0.42,
+      "learning_rate": 4.4943185490484915e-06,
+      "loss": 0.1608,
+      "step": 812
+    },
+    {
+      "epoch": 0.42,
+      "learning_rate": 4.493085808758215e-06,
+      "loss": 0.1523,
+      "step": 813
+    },
+    {
+      "epoch": 0.42,
+      "learning_rate": 4.491851737219964e-06,
+      "loss": 0.1688,
+      "step": 814
+    },
+    {
+      "epoch": 0.42,
+      "learning_rate": 4.490616335258016e-06,
+      "loss": 0.1732,
+      "step": 815
+    },
+    {
+      "epoch": 0.42,
+      "learning_rate": 4.489379603697535e-06,
+      "loss": 0.1776,
+      "step": 816
+    },
+    {
+      "epoch": 0.42,
+      "learning_rate": 4.488141543364578e-06,
+      "loss": 0.1725,
+      "step": 817
+    },
+    {
+      "epoch": 0.42,
+      "learning_rate": 4.486902155086084e-06,
+      "loss": 0.1469,
+      "step": 818
+    },
+    {
+      "epoch": 0.42,
+      "learning_rate": 4.485661439689882e-06,
+      "loss": 0.1739,
+      "step": 819
+    },
+    {
+      "epoch": 0.42,
+      "learning_rate": 4.484419398004687e-06,
+      "loss": 0.1922,
+      "step": 820
+    },
+    {
+      "epoch": 0.42,
+      "learning_rate": 4.4831760308600995e-06,
+      "loss": 0.1653,
+      "step": 821
+    },
+    {
+      "epoch": 0.43,
+      "learning_rate": 4.481931339086606e-06,
+      "loss": 0.1723,
+      "step": 822
+    },
+    {
+      "epoch": 0.43,
+      "learning_rate": 4.4806853235155775e-06,
+      "loss": 0.1829,
+      "step": 823
+    },
+    {
+      "epoch": 0.43,
+      "learning_rate": 4.479437984979268e-06,
+      "loss": 0.1547,
+      "step": 824
+    },
+    {
+      "epoch": 0.43,
+      "learning_rate": 4.4781893243108175e-06,
+      "loss": 0.1989,
+      "step": 825
+    },
+    {
+      "epoch": 0.43,
+      "learning_rate": 4.476939342344246e-06,
+      "loss": 0.1706,
+      "step": 826
+    },
+    {
+      "epoch": 0.43,
+      "learning_rate": 4.475688039914461e-06,
+      "loss": 0.193,
+      "step": 827
+    },
+    {
+      "epoch": 0.43,
+      "learning_rate": 4.4744354178572456e-06,
+      "loss": 0.1943,
+      "step": 828
+    },
+    {
+      "epoch": 0.43,
+      "learning_rate": 4.473181477009269e-06,
+      "loss": 0.1568,
+      "step": 829
+    },
+    {
+      "epoch": 0.43,
+      "learning_rate": 4.471926218208079e-06,
+      "loss": 0.1628,
+      "step": 830
+    },
+    {
+      "epoch": 0.43,
+      "learning_rate": 4.470669642292106e-06,
+      "loss": 0.173,
+      "step": 831
+    },
+    {
+      "epoch": 0.43,
+      "learning_rate": 4.4694117501006576e-06,
+      "loss": 0.178,
+      "step": 832
+    },
+    {
+      "epoch": 0.43,
+      "learning_rate": 4.46815254247392e-06,
+      "loss": 0.18,
+      "step": 833
+    },
+    {
+      "epoch": 0.43,
+      "learning_rate": 4.466892020252963e-06,
+      "loss": 0.2128,
+      "step": 834
+    },
+    {
+      "epoch": 0.43,
+      "learning_rate": 4.465630184279729e-06,
+      "loss": 0.1716,
+      "step": 835
+    },
+    {
+      "epoch": 0.43,
+      "learning_rate": 4.464367035397041e-06,
+      "loss": 0.173,
+      "step": 836
+    },
+    {
+      "epoch": 0.43,
+      "learning_rate": 4.463102574448597e-06,
+      "loss": 0.1482,
+      "step": 837
+    },
+    {
+      "epoch": 0.43,
+      "learning_rate": 4.4618368022789725e-06,
+      "loss": 0.1769,
+      "step": 838
+    },
+    {
+      "epoch": 0.43,
+      "learning_rate": 4.460569719733619e-06,
+      "loss": 0.2125,
+      "step": 839
+    },
+    {
+      "epoch": 0.43,
+      "learning_rate": 4.459301327658863e-06,
+      "loss": 0.1719,
+      "step": 840
+    },
+    {
+      "epoch": 0.44,
+      "learning_rate": 4.458031626901904e-06,
+      "loss": 0.1849,
+      "step": 841
+    },
+    {
+      "epoch": 0.44,
+      "learning_rate": 4.45676061831082e-06,
+      "loss": 0.1638,
+      "step": 842
+    },
+    {
+      "epoch": 0.44,
+      "learning_rate": 4.455488302734557e-06,
+      "loss": 0.1633,
+      "step": 843
+    },
+    {
+      "epoch": 0.44,
+      "learning_rate": 4.454214681022938e-06,
+      "loss": 0.1579,
+      "step": 844
+    },
+    {
+      "epoch": 0.44,
+      "learning_rate": 4.452939754026657e-06,
+      "loss": 0.1715,
+      "step": 845
+    },
+    {
+      "epoch": 0.44,
+      "learning_rate": 4.451663522597281e-06,
+      "loss": 0.1579,
+      "step": 846
+    },
+    {
+      "epoch": 0.44,
+      "learning_rate": 4.4503859875872444e-06,
+      "loss": 0.1867,
+      "step": 847
+    },
+    {
+      "epoch": 0.44,
+      "learning_rate": 4.449107149849857e-06,
+      "loss": 0.1696,
+      "step": 848
+    },
+    {
+      "epoch": 0.44,
+      "learning_rate": 4.447827010239297e-06,
+      "loss": 0.173,
+      "step": 849
+    },
+    {
+      "epoch": 0.44,
+      "learning_rate": 4.446545569610612e-06,
+      "loss": 0.1618,
+      "step": 850
+    },
+    {
+      "epoch": 0.44,
+      "eval_loss": 0.22647419571876526,
+      "eval_runtime": 118.2189,
+      "eval_samples_per_second": 26.409,
+      "eval_steps_per_second": 8.806,
+      "step": 850
+    },
+    {
+      "epoch": 0.44,
+      "learning_rate": 4.445262828819718e-06,
+      "loss": 0.2287,
+      "step": 851
+    },
+    {
+      "epoch": 0.44,
+      "learning_rate": 4.4439787887234e-06,
+      "loss": 0.168,
+      "step": 852
+    },
+    {
+      "epoch": 0.44,
+      "learning_rate": 4.442693450179311e-06,
+      "loss": 0.183,
+      "step": 853
+    },
+    {
+      "epoch": 0.44,
+      "learning_rate": 4.441406814045972e-06,
+      "loss": 0.1564,
+      "step": 854
+    },
+    {
+      "epoch": 0.44,
+      "learning_rate": 4.440118881182768e-06,
+      "loss": 0.1724,
+      "step": 855
+    },
+    {
+      "epoch": 0.44,
+      "learning_rate": 4.438829652449954e-06,
+      "loss": 0.1665,
+      "step": 856
+    },
+    {
+      "epoch": 0.44,
+      "learning_rate": 4.437539128708647e-06,
+      "loss": 0.1852,
+      "step": 857
+    },
+    {
+      "epoch": 0.44,
+      "learning_rate": 4.4362473108208315e-06,
+      "loss": 0.163,
+      "step": 858
+    },
+    {
+      "epoch": 0.44,
+      "learning_rate": 4.434954199649355e-06,
+      "loss": 0.1664,
+      "step": 859
+    },
+    {
+      "epoch": 0.44,
+      "learning_rate": 4.43365979605793e-06,
+      "loss": 0.1781,
+      "step": 860
+    },
+    {
+      "epoch": 0.45,
+      "learning_rate": 4.43236410091113e-06,
+      "loss": 0.1746,
+      "step": 861
+    },
+    {
+      "epoch": 0.45,
+      "learning_rate": 4.431067115074394e-06,
+      "loss": 0.158,
+      "step": 862
+    },
+    {
+      "epoch": 0.45,
+      "learning_rate": 4.429768839414022e-06,
+      "loss": 0.171,
+      "step": 863
+    },
+    {
+      "epoch": 0.45,
+      "learning_rate": 4.428469274797173e-06,
+      "loss": 0.1934,
+      "step": 864
+    },
+    {
+      "epoch": 0.45,
+      "learning_rate": 4.4271684220918715e-06,
+      "loss": 0.1963,
+      "step": 865
+    },
+    {
+      "epoch": 0.45,
+      "learning_rate": 4.425866282167e-06,
+      "loss": 0.1765,
+      "step": 866
+    },
+    {
+      "epoch": 0.45,
+      "learning_rate": 4.4245628558923e-06,
+      "loss": 0.1702,
+      "step": 867
+    },
+    {
+      "epoch": 0.45,
+      "learning_rate": 4.423258144138372e-06,
+      "loss": 0.1821,
+      "step": 868
+    },
+    {
+      "epoch": 0.45,
+      "learning_rate": 4.421952147776678e-06,
+      "loss": 0.1429,
+      "step": 869
+    },
+    {
+      "epoch": 0.45,
+      "learning_rate": 4.420644867679536e-06,
+      "loss": 0.1702,
+      "step": 870
+    },
+    {
+      "epoch": 0.45,
+      "learning_rate": 4.419336304720121e-06,
+      "loss": 0.1639,
+      "step": 871
+    },
+    {
+      "epoch": 0.45,
+      "learning_rate": 4.4180264597724655e-06,
+      "loss": 0.1807,
+      "step": 872
+    },
+    {
+      "epoch": 0.45,
+      "learning_rate": 4.4167153337114594e-06,
+      "loss": 0.1789,
+      "step": 873
+    },
+    {
+      "epoch": 0.45,
+      "learning_rate": 4.415402927412846e-06,
+      "loss": 0.1647,
+      "step": 874
+    },
+    {
+      "epoch": 0.45,
+      "learning_rate": 4.414089241753224e-06,
+      "loss": 0.1745,
+      "step": 875
+    },
+    {
+      "epoch": 0.45,
+      "learning_rate": 4.41277427761005e-06,
+      "loss": 0.1648,
+      "step": 876
+    },
+    {
+      "epoch": 0.45,
+      "learning_rate": 4.411458035861631e-06,
+      "loss": 0.1551,
+      "step": 877
+    },
+    {
+      "epoch": 0.45,
+      "learning_rate": 4.410140517387128e-06,
+      "loss": 0.16,
+      "step": 878
+    },
+    {
+      "epoch": 0.45,
+      "learning_rate": 4.408821723066555e-06,
+      "loss": 0.1865,
+      "step": 879
+    },
+    {
+      "epoch": 0.46,
+      "learning_rate": 4.407501653780779e-06,
+      "loss": 0.1626,
+      "step": 880
+    },
+    {
+      "epoch": 0.46,
+      "learning_rate": 4.4061803104115185e-06,
+      "loss": 0.1663,
+      "step": 881
+    },
+    {
+      "epoch": 0.46,
+      "learning_rate": 4.4048576938413414e-06,
+      "loss": 0.1671,
+      "step": 882
+    },
+    {
+      "epoch": 0.46,
+      "learning_rate": 4.403533804953668e-06,
+      "loss": 0.1821,
+      "step": 883
+    },
+    {
+      "epoch": 0.46,
+      "learning_rate": 4.402208644632766e-06,
+      "loss": 0.1643,
+      "step": 884
+    },
+    {
+      "epoch": 0.46,
+      "learning_rate": 4.4008822137637555e-06,
+      "loss": 0.1604,
+      "step": 885
+    },
+    {
+      "epoch": 0.46,
+      "learning_rate": 4.399554513232603e-06,
+      "loss": 0.1704,
+      "step": 886
+    },
+    {
+      "epoch": 0.46,
+      "learning_rate": 4.398225543926123e-06,
+      "loss": 0.148,
+      "step": 887
+    },
+    {
+      "epoch": 0.46,
+      "learning_rate": 4.396895306731978e-06,
+      "loss": 0.1588,
+      "step": 888
+    },
+    {
+      "epoch": 0.46,
+      "learning_rate": 4.395563802538678e-06,
+      "loss": 0.1635,
+      "step": 889
+    },
+    {
+      "epoch": 0.46,
+      "learning_rate": 4.3942310322355795e-06,
+      "loss": 0.1958,
+      "step": 890
+    },
+    {
+      "epoch": 0.46,
+      "learning_rate": 4.392896996712883e-06,
+      "loss": 0.1515,
+      "step": 891
+    },
+    {
+      "epoch": 0.46,
+      "learning_rate": 4.3915616968616345e-06,
+      "loss": 0.1832,
+      "step": 892
+    },
+    {
+      "epoch": 0.46,
+      "learning_rate": 4.390225133573727e-06,
+      "loss": 0.1559,
+      "step": 893
+    },
+    {
+      "epoch": 0.46,
+      "learning_rate": 4.388887307741893e-06,
+      "loss": 0.1712,
+      "step": 894
+    },
+    {
+      "epoch": 0.46,
+      "learning_rate": 4.387548220259711e-06,
+      "loss": 0.1682,
+      "step": 895
+    },
+    {
+      "epoch": 0.46,
+      "learning_rate": 4.386207872021603e-06,
+      "loss": 0.1564,
+      "step": 896
+    },
+    {
+      "epoch": 0.46,
+      "learning_rate": 4.384866263922832e-06,
+      "loss": 0.172,
+      "step": 897
+    },
+    {
+      "epoch": 0.46,
+      "learning_rate": 4.3835233968595015e-06,
+      "loss": 0.1749,
+      "step": 898
+    },
+    {
+      "epoch": 0.47,
+      "learning_rate": 4.382179271728557e-06,
+      "loss": 0.176,
+      "step": 899
+    },
+    {
+      "epoch": 0.47,
+      "learning_rate": 4.380833889427784e-06,
+      "loss": 0.1642,
+      "step": 900
+    },
+    {
+      "epoch": 0.47,
+      "eval_loss": 0.2244415283203125,
+      "eval_runtime": 118.085,
+      "eval_samples_per_second": 26.439,
+      "eval_steps_per_second": 8.816,
+      "step": 900
+    },
+    {
+      "epoch": 0.47,
+      "learning_rate": 4.3794872508558085e-06,
+      "loss": 0.1729,
+      "step": 901
+    },
+    {
+      "epoch": 0.47,
+      "learning_rate": 4.378139356912094e-06,
+      "loss": 0.1554,
+      "step": 902
+    },
+    {
+      "epoch": 0.47,
+      "learning_rate": 4.376790208496945e-06,
+      "loss": 0.1985,
+      "step": 903
+    },
+    {
+      "epoch": 0.47,
+      "learning_rate": 4.375439806511502e-06,
+      "loss": 0.1932,
+      "step": 904
+    },
+    {
+      "epoch": 0.47,
+      "learning_rate": 4.374088151857742e-06,
+      "loss": 0.1548,
+      "step": 905
+    },
+    {
+      "epoch": 0.47,
+      "learning_rate": 4.3727352454384805e-06,
+      "loss": 0.1765,
+      "step": 906
+    },
+    {
+      "epoch": 0.47,
+      "learning_rate": 4.37138108815737e-06,
+      "loss": 0.1806,
+      "step": 907
+    },
+    {
+      "epoch": 0.47,
+      "learning_rate": 4.370025680918893e-06,
+      "loss": 0.16,
+      "step": 908
+    },
+    {
+      "epoch": 0.47,
+      "learning_rate": 4.3686690246283744e-06,
+      "loss": 0.1593,
+      "step": 909
+    },
+    {
+      "epoch": 0.47,
+      "learning_rate": 4.367311120191969e-06,
+      "loss": 0.1788,
+      "step": 910
+    },
+    {
+      "epoch": 0.47,
+      "learning_rate": 4.365951968516666e-06,
+      "loss": 0.1736,
+      "step": 911
+    },
+    {
+      "epoch": 0.47,
+      "learning_rate": 4.3645915705102885e-06,
+      "loss": 0.1683,
+      "step": 912
+    },
+    {
+      "epoch": 0.47,
+      "learning_rate": 4.3632299270814895e-06,
+      "loss": 0.1534,
+      "step": 913
+    },
+    {
+      "epoch": 0.47,
+      "learning_rate": 4.361867039139756e-06,
+      "loss": 0.1737,
+      "step": 914
+    },
+    {
+      "epoch": 0.47,
+      "learning_rate": 4.360502907595409e-06,
+      "loss": 0.1542,
+      "step": 915
+    },
+    {
+      "epoch": 0.47,
+      "learning_rate": 4.3591375333595934e-06,
+      "loss": 0.1484,
+      "step": 916
+    },
+    {
+      "epoch": 0.47,
+      "learning_rate": 4.3577709173442895e-06,
+      "loss": 0.178,
+      "step": 917
+    },
+    {
+      "epoch": 0.47,
+      "learning_rate": 4.3564030604623055e-06,
+      "loss": 0.1565,
+      "step": 918
+    },
+    {
+      "epoch": 0.48,
+      "learning_rate": 4.355033963627277e-06,
+      "loss": 0.1672,
+      "step": 919
+    },
+    {
+      "epoch": 0.48,
+      "learning_rate": 4.353663627753671e-06,
+      "loss": 0.1636,
+      "step": 920
+    },
+    {
+      "epoch": 0.48,
+      "learning_rate": 4.3522920537567796e-06,
+      "loss": 0.1718,
+      "step": 921
+    },
+    {
+      "epoch": 0.48,
+      "learning_rate": 4.350919242552721e-06,
+      "loss": 0.168,
+      "step": 922
+    },
+    {
+      "epoch": 0.48,
+      "learning_rate": 4.349545195058442e-06,
+      "loss": 0.1834,
+      "step": 923
+    },
+    {
+      "epoch": 0.48,
+      "learning_rate": 4.348169912191715e-06,
+      "loss": 0.1542,
+      "step": 924
+    },
+    {
+      "epoch": 0.48,
+      "learning_rate": 4.346793394871138e-06,
+      "loss": 0.1642,
+      "step": 925
+    },
+    {
+      "epoch": 0.48,
+      "learning_rate": 4.3454156440161296e-06,
+      "loss": 0.1932,
+      "step": 926
+    },
+    {
+      "epoch": 0.48,
+      "learning_rate": 4.3440366605469375e-06,
+      "loss": 0.1593,
+      "step": 927
+    },
+    {
+      "epoch": 0.48,
+      "learning_rate": 4.3426564453846294e-06,
+      "loss": 0.1597,
+      "step": 928
+    },
+    {
+      "epoch": 0.48,
+      "learning_rate": 4.341274999451097e-06,
+      "loss": 0.1662,
+      "step": 929
+    },
+    {
+      "epoch": 0.48,
+      "learning_rate": 4.3398923236690535e-06,
+      "loss": 0.2078,
+      "step": 930
+    },
+    {
+      "epoch": 0.48,
+      "learning_rate": 4.3385084189620345e-06,
+      "loss": 0.1576,
+      "step": 931
+    },
+    {
+      "epoch": 0.48,
+      "learning_rate": 4.337123286254394e-06,
+      "loss": 0.1932,
+      "step": 932
+    },
+    {
+      "epoch": 0.48,
+      "learning_rate": 4.3357369264713105e-06,
+      "loss": 0.1646,
+      "step": 933
+    },
+    {
+      "epoch": 0.48,
+      "learning_rate": 4.334349340538777e-06,
+      "loss": 0.1654,
+      "step": 934
+    },
+    {
+      "epoch": 0.48,
+      "learning_rate": 4.332960529383611e-06,
+      "loss": 0.1528,
+      "step": 935
+    },
+    {
+      "epoch": 0.48,
+      "learning_rate": 4.331570493933443e-06,
+      "loss": 0.1651,
+      "step": 936
+    },
+    {
+      "epoch": 0.48,
+      "learning_rate": 4.3301792351167235e-06,
+      "loss": 0.1643,
+      "step": 937
+    },
+    {
+      "epoch": 0.49,
+      "learning_rate": 4.328786753862723e-06,
+      "loss": 0.1884,
+      "step": 938
+    },
+    {
+      "epoch": 0.49,
+      "learning_rate": 4.3273930511015226e-06,
+      "loss": 0.1391,
+      "step": 939
+    },
+    {
+      "epoch": 0.49,
+      "learning_rate": 4.325998127764025e-06,
+      "loss": 0.1769,
+      "step": 940
+    },
+    {
+      "epoch": 0.49,
+      "learning_rate": 4.324601984781944e-06,
+      "loss": 0.1534,
+      "step": 941
+    },
+    {
+      "epoch": 0.49,
+      "learning_rate": 4.323204623087811e-06,
+      "loss": 0.1587,
+      "step": 942
+    },
+    {
+      "epoch": 0.49,
+      "learning_rate": 4.321806043614969e-06,
+      "loss": 0.213,
+      "step": 943
+    },
+    {
+      "epoch": 0.49,
+      "learning_rate": 4.320406247297576e-06,
+      "loss": 0.158,
+      "step": 944
+    },
+    {
+      "epoch": 0.49,
+      "learning_rate": 4.319005235070603e-06,
+      "loss": 0.1981,
+      "step": 945
+    },
+    {
+      "epoch": 0.49,
+      "learning_rate": 4.317603007869831e-06,
+      "loss": 0.1884,
+      "step": 946
+    },
+    {
+      "epoch": 0.49,
+      "learning_rate": 4.316199566631855e-06,
+      "loss": 0.1595,
+      "step": 947
+    },
+    {
+      "epoch": 0.49,
+      "learning_rate": 4.314794912294079e-06,
+      "loss": 0.1675,
+      "step": 948
+    },
+    {
+      "epoch": 0.49,
+      "learning_rate": 4.313389045794719e-06,
+      "loss": 0.163,
+      "step": 949
+    },
+    {
+      "epoch": 0.49,
+      "learning_rate": 4.3119819680728e-06,
+      "loss": 0.1676,
+      "step": 950
+    },
+    {
+      "epoch": 0.49,
+      "eval_loss": 0.2226845920085907,
+      "eval_runtime": 118.493,
+      "eval_samples_per_second": 26.348,
+      "eval_steps_per_second": 8.785,
+      "step": 950
+    },
+    {
+      "epoch": 0.49,
+      "learning_rate": 4.310573680068155e-06,
+      "loss": 0.1477,
+      "step": 951
+    },
+    {
+      "epoch": 0.49,
+      "learning_rate": 4.309164182721426e-06,
+      "loss": 0.1766,
+      "step": 952
+    },
+    {
+      "epoch": 0.49,
+      "learning_rate": 4.307753476974064e-06,
+      "loss": 0.1676,
+      "step": 953
+    },
+    {
+      "epoch": 0.49,
+      "learning_rate": 4.306341563768325e-06,
+      "loss": 0.1455,
+      "step": 954
+    },
+    {
+      "epoch": 0.49,
+      "learning_rate": 4.304928444047272e-06,
+      "loss": 0.1865,
+      "step": 955
+    },
+    {
+      "epoch": 0.49,
+      "learning_rate": 4.303514118754776e-06,
+      "loss": 0.142,
+      "step": 956
+    },
+    {
+      "epoch": 0.5,
+      "learning_rate": 4.302098588835512e-06,
+      "loss": 0.1583,
+      "step": 957
+    },
+    {
+      "epoch": 0.5,
+      "learning_rate": 4.300681855234957e-06,
+      "loss": 0.1636,
+      "step": 958
+    },
+    {
+      "epoch": 0.5,
+      "learning_rate": 4.299263918899396e-06,
+      "loss": 0.1476,
+      "step": 959
+    },
+    {
+      "epoch": 0.5,
+      "learning_rate": 4.297844780775915e-06,
+      "loss": 0.1637,
+      "step": 960
+    }
+  ],
+  "logging_steps": 1,
+  "max_steps": 3864,
+  "num_train_epochs": 2,
+  "save_steps": 480,
+  "total_flos": 6.039421649713889e+18,
+  "trial_name": null,
+  "trial_params": null
+}