diff --git "a/trainer_state.json" "b/trainer_state.json"
new file mode 100644--- /dev/null
+++ "b/trainer_state.json"
@@ -0,0 +1,81817 @@
+{
+  "best_metric": null,
+  "best_model_checkpoint": null,
+  "epoch": 1.9879760807595859,
+  "eval_steps": 700,
+  "global_step": 11664,
+  "is_hyper_param_search": false,
+  "is_local_process_zero": true,
+  "is_world_process_zero": true,
+  "log_history": [
+    {
+      "epoch": 0.0001714640889898622,
+      "grad_norm": 223.0,
+      "learning_rate": 2.0000000000000002e-07,
+      "loss": 1.7472,
+      "step": 1
+    },
+    {
+      "epoch": 0.0001714640889898622,
+      "eval_loss": 2.7519192695617676,
+      "eval_runtime": 836.2413,
+      "eval_samples_per_second": 2.988,
+      "eval_steps_per_second": 2.988,
+      "step": 1
+    },
+    {
+      "epoch": 0.0003429281779797244,
+      "grad_norm": 127.0,
+      "learning_rate": 4.0000000000000003e-07,
+      "loss": 1.6937,
+      "step": 2
+    },
+    {
+      "epoch": 0.0005143922669695865,
+      "grad_norm": 237.0,
+      "learning_rate": 6.000000000000001e-07,
+      "loss": 1.661,
+      "step": 3
+    },
+    {
+      "epoch": 0.0006858563559594488,
+      "grad_norm": 118.5,
+      "learning_rate": 8.000000000000001e-07,
+      "loss": 1.5902,
+      "step": 4
+    },
+    {
+      "epoch": 0.000857320444949311,
+      "grad_norm": 156.0,
+      "learning_rate": 1.0000000000000002e-06,
+      "loss": 1.6023,
+      "step": 5
+    },
+    {
+      "epoch": 0.001028784533939173,
+      "grad_norm": 155.0,
+      "learning_rate": 1.2000000000000002e-06,
+      "loss": 1.7616,
+      "step": 6
+    },
+    {
+      "epoch": 0.0012002486229290353,
+      "grad_norm": 94.5,
+      "learning_rate": 1.4000000000000001e-06,
+      "loss": 1.7018,
+      "step": 7
+    },
+    {
+      "epoch": 0.0013717127119188975,
+      "grad_norm": 96.5,
+      "learning_rate": 1.6000000000000001e-06,
+      "loss": 1.6484,
+      "step": 8
+    },
+    {
+      "epoch": 0.0015431768009087597,
+      "grad_norm": 290.0,
+      "learning_rate": 1.8000000000000001e-06,
+      "loss": 1.5879,
+      "step": 9
+    },
+    {
+      "epoch": 0.001714640889898622,
+      "grad_norm": 154.0,
+      "learning_rate": 2.0000000000000003e-06,
+      "loss": 1.5615,
+      "step": 10
+    },
+    {
+      "epoch": 0.001886104978888484,
+      "grad_norm": 68.0,
+      "learning_rate": 2.2e-06,
+      "loss": 1.5076,
+      "step": 11
+    },
+    {
+      "epoch": 0.002057569067878346,
+      "grad_norm": 70.5,
+      "learning_rate": 2.4000000000000003e-06,
+      "loss": 1.5792,
+      "step": 12
+    },
+    {
+      "epoch": 0.0022290331568682086,
+      "grad_norm": 112.5,
+      "learning_rate": 2.6e-06,
+      "loss": 1.4464,
+      "step": 13
+    },
+    {
+      "epoch": 0.0024004972458580706,
+      "grad_norm": 47.5,
+      "learning_rate": 2.8000000000000003e-06,
+      "loss": 1.4436,
+      "step": 14
+    },
+    {
+      "epoch": 0.0025719613348479326,
+      "grad_norm": 33.75,
+      "learning_rate": 3e-06,
+      "loss": 1.4214,
+      "step": 15
+    },
+    {
+      "epoch": 0.002743425423837795,
+      "grad_norm": 46.0,
+      "learning_rate": 3.2000000000000003e-06,
+      "loss": 1.3857,
+      "step": 16
+    },
+    {
+      "epoch": 0.002914889512827657,
+      "grad_norm": 59.75,
+      "learning_rate": 3.4000000000000005e-06,
+      "loss": 1.477,
+      "step": 17
+    },
+    {
+      "epoch": 0.0030863536018175194,
+      "grad_norm": 23.75,
+      "learning_rate": 3.6000000000000003e-06,
+      "loss": 1.4501,
+      "step": 18
+    },
+    {
+      "epoch": 0.0032578176908073814,
+      "grad_norm": 19.625,
+      "learning_rate": 3.8000000000000005e-06,
+      "loss": 1.5123,
+      "step": 19
+    },
+    {
+      "epoch": 0.003429281779797244,
+      "grad_norm": 12.75,
+      "learning_rate": 4.000000000000001e-06,
+      "loss": 1.3971,
+      "step": 20
+    },
+    {
+      "epoch": 0.003600745868787106,
+      "grad_norm": 31.625,
+      "learning_rate": 4.2000000000000004e-06,
+      "loss": 1.2835,
+      "step": 21
+    },
+    {
+      "epoch": 0.003772209957776968,
+      "grad_norm": 56.75,
+      "learning_rate": 4.4e-06,
+      "loss": 1.3467,
+      "step": 22
+    },
+    {
+      "epoch": 0.00394367404676683,
+      "grad_norm": 5.09375,
+      "learning_rate": 4.600000000000001e-06,
+      "loss": 1.2356,
+      "step": 23
+    },
+    {
+      "epoch": 0.004115138135756692,
+      "grad_norm": 83.5,
+      "learning_rate": 4.800000000000001e-06,
+      "loss": 1.3952,
+      "step": 24
+    },
+    {
+      "epoch": 0.004286602224746554,
+      "grad_norm": 1004.0,
+      "learning_rate": 5e-06,
+      "loss": 1.3461,
+      "step": 25
+    },
+    {
+      "epoch": 0.004458066313736417,
+      "grad_norm": 18.875,
+      "learning_rate": 5.2e-06,
+      "loss": 1.4225,
+      "step": 26
+    },
+    {
+      "epoch": 0.004629530402726279,
+      "grad_norm": 430.0,
+      "learning_rate": 5.400000000000001e-06,
+      "loss": 1.356,
+      "step": 27
+    },
+    {
+      "epoch": 0.004800994491716141,
+      "grad_norm": 62.25,
+      "learning_rate": 5.600000000000001e-06,
+      "loss": 1.4085,
+      "step": 28
+    },
+    {
+      "epoch": 0.004972458580706003,
+      "grad_norm": 8.8125,
+      "learning_rate": 5.8e-06,
+      "loss": 1.2939,
+      "step": 29
+    },
+    {
+      "epoch": 0.005143922669695865,
+      "grad_norm": 131.0,
+      "learning_rate": 6e-06,
+      "loss": 1.3616,
+      "step": 30
+    },
+    {
+      "epoch": 0.005315386758685728,
+      "grad_norm": 43.0,
+      "learning_rate": 6.200000000000001e-06,
+      "loss": 1.4088,
+      "step": 31
+    },
+    {
+      "epoch": 0.00548685084767559,
+      "grad_norm": 28.0,
+      "learning_rate": 6.4000000000000006e-06,
+      "loss": 1.3492,
+      "step": 32
+    },
+    {
+      "epoch": 0.005658314936665452,
+      "grad_norm": 18.5,
+      "learning_rate": 6.600000000000001e-06,
+      "loss": 1.3562,
+      "step": 33
+    },
+    {
+      "epoch": 0.005829779025655314,
+      "grad_norm": 19.375,
+      "learning_rate": 6.800000000000001e-06,
+      "loss": 1.2559,
+      "step": 34
+    },
+    {
+      "epoch": 0.006001243114645177,
+      "grad_norm": 23.0,
+      "learning_rate": 7e-06,
+      "loss": 1.3713,
+      "step": 35
+    },
+    {
+      "epoch": 0.006172707203635039,
+      "grad_norm": 211.0,
+      "learning_rate": 7.2000000000000005e-06,
+      "loss": 1.2363,
+      "step": 36
+    },
+    {
+      "epoch": 0.006344171292624901,
+      "grad_norm": 34.5,
+      "learning_rate": 7.4e-06,
+      "loss": 1.371,
+      "step": 37
+    },
+    {
+      "epoch": 0.006515635381614763,
+      "grad_norm": 7.78125,
+      "learning_rate": 7.600000000000001e-06,
+      "loss": 1.3426,
+      "step": 38
+    },
+    {
+      "epoch": 0.006687099470604625,
+      "grad_norm": 21.0,
+      "learning_rate": 7.800000000000002e-06,
+      "loss": 1.2491,
+      "step": 39
+    },
+    {
+      "epoch": 0.006858563559594488,
+      "grad_norm": 2.859375,
+      "learning_rate": 8.000000000000001e-06,
+      "loss": 1.2641,
+      "step": 40
+    },
+    {
+      "epoch": 0.00703002764858435,
+      "grad_norm": 11.9375,
+      "learning_rate": 8.2e-06,
+      "loss": 1.3252,
+      "step": 41
+    },
+    {
+      "epoch": 0.007201491737574212,
+      "grad_norm": 6.75,
+      "learning_rate": 8.400000000000001e-06,
+      "loss": 1.242,
+      "step": 42
+    },
+    {
+      "epoch": 0.007372955826564074,
+      "grad_norm": 14.3125,
+      "learning_rate": 8.6e-06,
+      "loss": 1.3359,
+      "step": 43
+    },
+    {
+      "epoch": 0.007544419915553936,
+      "grad_norm": 23.75,
+      "learning_rate": 8.8e-06,
+      "loss": 1.3308,
+      "step": 44
+    },
+    {
+      "epoch": 0.007715884004543799,
+      "grad_norm": 17.25,
+      "learning_rate": 9e-06,
+      "loss": 1.2513,
+      "step": 45
+    },
+    {
+      "epoch": 0.00788734809353366,
+      "grad_norm": 4.65625,
+      "learning_rate": 9.200000000000002e-06,
+      "loss": 1.2823,
+      "step": 46
+    },
+    {
+      "epoch": 0.008058812182523523,
+      "grad_norm": 4.46875,
+      "learning_rate": 9.4e-06,
+      "loss": 1.3031,
+      "step": 47
+    },
+    {
+      "epoch": 0.008230276271513385,
+      "grad_norm": 4.25,
+      "learning_rate": 9.600000000000001e-06,
+      "loss": 1.2793,
+      "step": 48
+    },
+    {
+      "epoch": 0.008401740360503247,
+      "grad_norm": 4.3125,
+      "learning_rate": 9.800000000000001e-06,
+      "loss": 1.3314,
+      "step": 49
+    },
+    {
+      "epoch": 0.008573204449493109,
+      "grad_norm": 9.1875,
+      "learning_rate": 1e-05,
+      "loss": 1.3205,
+      "step": 50
+    },
+    {
+      "epoch": 0.008744668538482972,
+      "grad_norm": 8.625,
+      "learning_rate": 1.02e-05,
+      "loss": 1.2554,
+      "step": 51
+    },
+    {
+      "epoch": 0.008916132627472834,
+      "grad_norm": 2.734375,
+      "learning_rate": 1.04e-05,
+      "loss": 1.2897,
+      "step": 52
+    },
+    {
+      "epoch": 0.009087596716462696,
+      "grad_norm": 3.609375,
+      "learning_rate": 1.0600000000000002e-05,
+      "loss": 1.3327,
+      "step": 53
+    },
+    {
+      "epoch": 0.009259060805452558,
+      "grad_norm": 8.9375,
+      "learning_rate": 1.0800000000000002e-05,
+      "loss": 1.2707,
+      "step": 54
+    },
+    {
+      "epoch": 0.00943052489444242,
+      "grad_norm": 9.4375,
+      "learning_rate": 1.1000000000000001e-05,
+      "loss": 1.2724,
+      "step": 55
+    },
+    {
+      "epoch": 0.009601988983432282,
+      "grad_norm": 3.890625,
+      "learning_rate": 1.1200000000000001e-05,
+      "loss": 1.2567,
+      "step": 56
+    },
+    {
+      "epoch": 0.009773453072422144,
+      "grad_norm": 3.796875,
+      "learning_rate": 1.14e-05,
+      "loss": 1.3487,
+      "step": 57
+    },
+    {
+      "epoch": 0.009944917161412006,
+      "grad_norm": 2.65625,
+      "learning_rate": 1.16e-05,
+      "loss": 1.2577,
+      "step": 58
+    },
+    {
+      "epoch": 0.010116381250401868,
+      "grad_norm": 2.71875,
+      "learning_rate": 1.18e-05,
+      "loss": 1.2358,
+      "step": 59
+    },
+    {
+      "epoch": 0.01028784533939173,
+      "grad_norm": 2.546875,
+      "learning_rate": 1.2e-05,
+      "loss": 1.2995,
+      "step": 60
+    },
+    {
+      "epoch": 0.010459309428381594,
+      "grad_norm": 2.984375,
+      "learning_rate": 1.22e-05,
+      "loss": 1.2554,
+      "step": 61
+    },
+    {
+      "epoch": 0.010630773517371456,
+      "grad_norm": 11.75,
+      "learning_rate": 1.2400000000000002e-05,
+      "loss": 1.2732,
+      "step": 62
+    },
+    {
+      "epoch": 0.010802237606361318,
+      "grad_norm": 2.671875,
+      "learning_rate": 1.2600000000000001e-05,
+      "loss": 1.3157,
+      "step": 63
+    },
+    {
+      "epoch": 0.01097370169535118,
+      "grad_norm": 5.78125,
+      "learning_rate": 1.2800000000000001e-05,
+      "loss": 1.1681,
+      "step": 64
+    },
+    {
+      "epoch": 0.011145165784341042,
+      "grad_norm": 2.71875,
+      "learning_rate": 1.3000000000000001e-05,
+      "loss": 1.3121,
+      "step": 65
+    },
+    {
+      "epoch": 0.011316629873330904,
+      "grad_norm": 2.4375,
+      "learning_rate": 1.3200000000000002e-05,
+      "loss": 1.2333,
+      "step": 66
+    },
+    {
+      "epoch": 0.011488093962320766,
+      "grad_norm": 2.296875,
+      "learning_rate": 1.3400000000000002e-05,
+      "loss": 1.1963,
+      "step": 67
+    },
+    {
+      "epoch": 0.011659558051310628,
+      "grad_norm": 2.6875,
+      "learning_rate": 1.3600000000000002e-05,
+      "loss": 1.1645,
+      "step": 68
+    },
+    {
+      "epoch": 0.01183102214030049,
+      "grad_norm": 3.125,
+      "learning_rate": 1.38e-05,
+      "loss": 1.183,
+      "step": 69
+    },
+    {
+      "epoch": 0.012002486229290354,
+      "grad_norm": 2.265625,
+      "learning_rate": 1.4e-05,
+      "loss": 1.2707,
+      "step": 70
+    },
+    {
+      "epoch": 0.012173950318280216,
+      "grad_norm": 2.953125,
+      "learning_rate": 1.4200000000000001e-05,
+      "loss": 1.2276,
+      "step": 71
+    },
+    {
+      "epoch": 0.012345414407270078,
+      "grad_norm": 2.78125,
+      "learning_rate": 1.4400000000000001e-05,
+      "loss": 1.2386,
+      "step": 72
+    },
+    {
+      "epoch": 0.01251687849625994,
+      "grad_norm": 2.421875,
+      "learning_rate": 1.46e-05,
+      "loss": 1.1404,
+      "step": 73
+    },
+    {
+      "epoch": 0.012688342585249802,
+      "grad_norm": 3.796875,
+      "learning_rate": 1.48e-05,
+      "loss": 1.2231,
+      "step": 74
+    },
+    {
+      "epoch": 0.012859806674239664,
+      "grad_norm": 2.46875,
+      "learning_rate": 1.5000000000000002e-05,
+      "loss": 1.1835,
+      "step": 75
+    },
+    {
+      "epoch": 0.013031270763229526,
+      "grad_norm": 2.140625,
+      "learning_rate": 1.5200000000000002e-05,
+      "loss": 1.2235,
+      "step": 76
+    },
+    {
+      "epoch": 0.013202734852219388,
+      "grad_norm": 2.203125,
+      "learning_rate": 1.54e-05,
+      "loss": 1.1566,
+      "step": 77
+    },
+    {
+      "epoch": 0.01337419894120925,
+      "grad_norm": 2.0,
+      "learning_rate": 1.5600000000000003e-05,
+      "loss": 1.1565,
+      "step": 78
+    },
+    {
+      "epoch": 0.013545663030199113,
+      "grad_norm": 2.296875,
+      "learning_rate": 1.58e-05,
+      "loss": 1.2313,
+      "step": 79
+    },
+    {
+      "epoch": 0.013717127119188975,
+      "grad_norm": 2.046875,
+      "learning_rate": 1.6000000000000003e-05,
+      "loss": 1.1824,
+      "step": 80
+    },
+    {
+      "epoch": 0.013888591208178837,
+      "grad_norm": 1.96875,
+      "learning_rate": 1.62e-05,
+      "loss": 1.1746,
+      "step": 81
+    },
+    {
+      "epoch": 0.0140600552971687,
+      "grad_norm": 2.359375,
+      "learning_rate": 1.64e-05,
+      "loss": 1.2601,
+      "step": 82
+    },
+    {
+      "epoch": 0.014231519386158561,
+      "grad_norm": 2.359375,
+      "learning_rate": 1.66e-05,
+      "loss": 1.1281,
+      "step": 83
+    },
+    {
+      "epoch": 0.014402983475148423,
+      "grad_norm": 2.0625,
+      "learning_rate": 1.6800000000000002e-05,
+      "loss": 1.1808,
+      "step": 84
+    },
+    {
+      "epoch": 0.014574447564138285,
+      "grad_norm": 2.953125,
+      "learning_rate": 1.7e-05,
+      "loss": 1.1618,
+      "step": 85
+    },
+    {
+      "epoch": 0.014745911653128147,
+      "grad_norm": 2.25,
+      "learning_rate": 1.72e-05,
+      "loss": 1.1581,
+      "step": 86
+    },
+    {
+      "epoch": 0.01491737574211801,
+      "grad_norm": 2.265625,
+      "learning_rate": 1.7400000000000003e-05,
+      "loss": 1.2232,
+      "step": 87
+    },
+    {
+      "epoch": 0.015088839831107871,
+      "grad_norm": 3.390625,
+      "learning_rate": 1.76e-05,
+      "loss": 1.2678,
+      "step": 88
+    },
+    {
+      "epoch": 0.015260303920097735,
+      "grad_norm": 2.3125,
+      "learning_rate": 1.7800000000000002e-05,
+      "loss": 1.0981,
+      "step": 89
+    },
+    {
+      "epoch": 0.015431768009087597,
+      "grad_norm": 2.171875,
+      "learning_rate": 1.8e-05,
+      "loss": 1.1916,
+      "step": 90
+    },
+    {
+      "epoch": 0.01560323209807746,
+      "grad_norm": 2.015625,
+      "learning_rate": 1.8200000000000002e-05,
+      "loss": 1.1261,
+      "step": 91
+    },
+    {
+      "epoch": 0.01577469618706732,
+      "grad_norm": 2.09375,
+      "learning_rate": 1.8400000000000003e-05,
+      "loss": 1.1955,
+      "step": 92
+    },
+    {
+      "epoch": 0.015946160276057185,
+      "grad_norm": 10.6875,
+      "learning_rate": 1.86e-05,
+      "loss": 1.1729,
+      "step": 93
+    },
+    {
+      "epoch": 0.016117624365047045,
+      "grad_norm": 2.15625,
+      "learning_rate": 1.88e-05,
+      "loss": 1.1203,
+      "step": 94
+    },
+    {
+      "epoch": 0.01628908845403691,
+      "grad_norm": 2.171875,
+      "learning_rate": 1.9e-05,
+      "loss": 1.2245,
+      "step": 95
+    },
+    {
+      "epoch": 0.01646055254302677,
+      "grad_norm": 2.046875,
+      "learning_rate": 1.9200000000000003e-05,
+      "loss": 1.191,
+      "step": 96
+    },
+    {
+      "epoch": 0.016632016632016633,
+      "grad_norm": 2.34375,
+      "learning_rate": 1.94e-05,
+      "loss": 1.276,
+      "step": 97
+    },
+    {
+      "epoch": 0.016803480721006493,
+      "grad_norm": 2.453125,
+      "learning_rate": 1.9600000000000002e-05,
+      "loss": 1.2306,
+      "step": 98
+    },
+    {
+      "epoch": 0.016974944809996357,
+      "grad_norm": 2.0,
+      "learning_rate": 1.98e-05,
+      "loss": 1.2018,
+      "step": 99
+    },
+    {
+      "epoch": 0.017146408898986217,
+      "grad_norm": 2.875,
+      "learning_rate": 2e-05,
+      "loss": 1.2462,
+      "step": 100
+    },
+    {
+      "epoch": 0.01731787298797608,
+      "grad_norm": 2.0625,
+      "learning_rate": 1.9999999836931174e-05,
+      "loss": 1.1722,
+      "step": 101
+    },
+    {
+      "epoch": 0.017489337076965945,
+      "grad_norm": 1.8046875,
+      "learning_rate": 1.9999999347724693e-05,
+      "loss": 1.1661,
+      "step": 102
+    },
+    {
+      "epoch": 0.017660801165955805,
+      "grad_norm": 1.984375,
+      "learning_rate": 1.999999853238058e-05,
+      "loss": 1.1579,
+      "step": 103
+    },
+    {
+      "epoch": 0.01783226525494567,
+      "grad_norm": 2.578125,
+      "learning_rate": 1.9999997390898854e-05,
+      "loss": 1.244,
+      "step": 104
+    },
+    {
+      "epoch": 0.01800372934393553,
+      "grad_norm": 2.328125,
+      "learning_rate": 1.999999592327956e-05,
+      "loss": 1.1997,
+      "step": 105
+    },
+    {
+      "epoch": 0.018175193432925393,
+      "grad_norm": 2.28125,
+      "learning_rate": 1.999999412952274e-05,
+      "loss": 1.1714,
+      "step": 106
+    },
+    {
+      "epoch": 0.018346657521915253,
+      "grad_norm": 2.03125,
+      "learning_rate": 1.999999200962846e-05,
+      "loss": 1.1926,
+      "step": 107
+    },
+    {
+      "epoch": 0.018518121610905117,
+      "grad_norm": 6.46875,
+      "learning_rate": 1.999998956359678e-05,
+      "loss": 1.1984,
+      "step": 108
+    },
+    {
+      "epoch": 0.018689585699894977,
+      "grad_norm": 13.5625,
+      "learning_rate": 1.9999986791427782e-05,
+      "loss": 1.1753,
+      "step": 109
+    },
+    {
+      "epoch": 0.01886104978888484,
+      "grad_norm": 2.125,
+      "learning_rate": 1.9999983693121564e-05,
+      "loss": 1.1055,
+      "step": 110
+    },
+    {
+      "epoch": 0.0190325138778747,
+      "grad_norm": 2.046875,
+      "learning_rate": 1.9999980268678218e-05,
+      "loss": 1.1672,
+      "step": 111
+    },
+    {
+      "epoch": 0.019203977966864565,
+      "grad_norm": 2.09375,
+      "learning_rate": 1.999997651809786e-05,
+      "loss": 1.2352,
+      "step": 112
+    },
+    {
+      "epoch": 0.01937544205585443,
+      "grad_norm": 2.234375,
+      "learning_rate": 1.999997244138061e-05,
+      "loss": 1.095,
+      "step": 113
+    },
+    {
+      "epoch": 0.01954690614484429,
+      "grad_norm": 2.0625,
+      "learning_rate": 1.99999680385266e-05,
+      "loss": 1.2065,
+      "step": 114
+    },
+    {
+      "epoch": 0.019718370233834152,
+      "grad_norm": 2.296875,
+      "learning_rate": 1.999996330953598e-05,
+      "loss": 1.1259,
+      "step": 115
+    },
+    {
+      "epoch": 0.019889834322824013,
+      "grad_norm": 3.796875,
+      "learning_rate": 1.9999958254408897e-05,
+      "loss": 1.2381,
+      "step": 116
+    },
+    {
+      "epoch": 0.020061298411813876,
+      "grad_norm": 3.203125,
+      "learning_rate": 1.9999952873145523e-05,
+      "loss": 1.234,
+      "step": 117
+    },
+    {
+      "epoch": 0.020232762500803737,
+      "grad_norm": 2.390625,
+      "learning_rate": 1.9999947165746028e-05,
+      "loss": 1.2862,
+      "step": 118
+    },
+    {
+      "epoch": 0.0204042265897936,
+      "grad_norm": 2.25,
+      "learning_rate": 1.9999941132210598e-05,
+      "loss": 1.1938,
+      "step": 119
+    },
+    {
+      "epoch": 0.02057569067878346,
+      "grad_norm": 1.9765625,
+      "learning_rate": 1.999993477253943e-05,
+      "loss": 1.2301,
+      "step": 120
+    },
+    {
+      "epoch": 0.020747154767773324,
+      "grad_norm": 2.109375,
+      "learning_rate": 1.9999928086732736e-05,
+      "loss": 1.2761,
+      "step": 121
+    },
+    {
+      "epoch": 0.020918618856763188,
+      "grad_norm": 1.828125,
+      "learning_rate": 1.999992107479073e-05,
+      "loss": 1.1392,
+      "step": 122
+    },
+    {
+      "epoch": 0.02109008294575305,
+      "grad_norm": 1.859375,
+      "learning_rate": 1.9999913736713642e-05,
+      "loss": 1.0334,
+      "step": 123
+    },
+    {
+      "epoch": 0.021261547034742912,
+      "grad_norm": 1.84375,
+      "learning_rate": 1.9999906072501712e-05,
+      "loss": 1.1569,
+      "step": 124
+    },
+    {
+      "epoch": 0.021433011123732772,
+      "grad_norm": 1.8984375,
+      "learning_rate": 1.9999898082155185e-05,
+      "loss": 1.1254,
+      "step": 125
+    },
+    {
+      "epoch": 0.021604475212722636,
+      "grad_norm": 2.40625,
+      "learning_rate": 1.9999889765674326e-05,
+      "loss": 1.1018,
+      "step": 126
+    },
+    {
+      "epoch": 0.021775939301712496,
+      "grad_norm": 1.8125,
+      "learning_rate": 1.999988112305941e-05,
+      "loss": 1.1592,
+      "step": 127
+    },
+    {
+      "epoch": 0.02194740339070236,
+      "grad_norm": 1.9296875,
+      "learning_rate": 1.999987215431071e-05,
+      "loss": 1.1072,
+      "step": 128
+    },
+    {
+      "epoch": 0.02211886747969222,
+      "grad_norm": 2.21875,
+      "learning_rate": 1.9999862859428526e-05,
+      "loss": 1.1306,
+      "step": 129
+    },
+    {
+      "epoch": 0.022290331568682084,
+      "grad_norm": 2.125,
+      "learning_rate": 1.9999853238413154e-05,
+      "loss": 1.2421,
+      "step": 130
+    },
+    {
+      "epoch": 0.022461795657671948,
+      "grad_norm": 6.09375,
+      "learning_rate": 1.9999843291264915e-05,
+      "loss": 1.2552,
+      "step": 131
+    },
+    {
+      "epoch": 0.022633259746661808,
+      "grad_norm": 5.21875,
+      "learning_rate": 1.999983301798413e-05,
+      "loss": 1.2062,
+      "step": 132
+    },
+    {
+      "epoch": 0.022804723835651672,
+      "grad_norm": 2.03125,
+      "learning_rate": 1.9999822418571134e-05,
+      "loss": 1.1919,
+      "step": 133
+    },
+    {
+      "epoch": 0.022976187924641532,
+      "grad_norm": 1.9296875,
+      "learning_rate": 1.9999811493026275e-05,
+      "loss": 1.1638,
+      "step": 134
+    },
+    {
+      "epoch": 0.023147652013631396,
+      "grad_norm": 1.9609375,
+      "learning_rate": 1.9999800241349903e-05,
+      "loss": 1.2391,
+      "step": 135
+    },
+    {
+      "epoch": 0.023319116102621256,
+      "grad_norm": 1.84375,
+      "learning_rate": 1.9999788663542397e-05,
+      "loss": 1.1979,
+      "step": 136
+    },
+    {
+      "epoch": 0.02349058019161112,
+      "grad_norm": 2.0,
+      "learning_rate": 1.9999776759604123e-05,
+      "loss": 1.1564,
+      "step": 137
+    },
+    {
+      "epoch": 0.02366204428060098,
+      "grad_norm": 2.109375,
+      "learning_rate": 1.999976452953547e-05,
+      "loss": 1.1106,
+      "step": 138
+    },
+    {
+      "epoch": 0.023833508369590844,
+      "grad_norm": 2.5625,
+      "learning_rate": 1.9999751973336843e-05,
+      "loss": 1.1715,
+      "step": 139
+    },
+    {
+      "epoch": 0.024004972458580708,
+      "grad_norm": 1.90625,
+      "learning_rate": 1.9999739091008646e-05,
+      "loss": 1.0636,
+      "step": 140
+    },
+    {
+      "epoch": 0.024176436547570568,
+      "grad_norm": 1.8671875,
+      "learning_rate": 1.9999725882551305e-05,
+      "loss": 1.1546,
+      "step": 141
+    },
+    {
+      "epoch": 0.02434790063656043,
+      "grad_norm": 1.90625,
+      "learning_rate": 1.9999712347965245e-05,
+      "loss": 1.248,
+      "step": 142
+    },
+    {
+      "epoch": 0.024519364725550292,
+      "grad_norm": 1.8671875,
+      "learning_rate": 1.9999698487250914e-05,
+      "loss": 1.1635,
+      "step": 143
+    },
+    {
+      "epoch": 0.024690828814540156,
+      "grad_norm": 1.9140625,
+      "learning_rate": 1.9999684300408756e-05,
+      "loss": 1.2565,
+      "step": 144
+    },
+    {
+      "epoch": 0.024862292903530016,
+      "grad_norm": 1.9453125,
+      "learning_rate": 1.999966978743924e-05,
+      "loss": 1.1027,
+      "step": 145
+    },
+    {
+      "epoch": 0.02503375699251988,
+      "grad_norm": 2.09375,
+      "learning_rate": 1.9999654948342836e-05,
+      "loss": 1.1535,
+      "step": 146
+    },
+    {
+      "epoch": 0.02520522108150974,
+      "grad_norm": 2.265625,
+      "learning_rate": 1.999963978312003e-05,
+      "loss": 1.17,
+      "step": 147
+    },
+    {
+      "epoch": 0.025376685170499604,
+      "grad_norm": 1.9609375,
+      "learning_rate": 1.999962429177131e-05,
+      "loss": 1.1554,
+      "step": 148
+    },
+    {
+      "epoch": 0.025548149259489467,
+      "grad_norm": 2.046875,
+      "learning_rate": 1.9999608474297192e-05,
+      "loss": 1.2368,
+      "step": 149
+    },
+    {
+      "epoch": 0.025719613348479328,
+      "grad_norm": 2.28125,
+      "learning_rate": 1.9999592330698185e-05,
+      "loss": 1.1703,
+      "step": 150
+    },
+    {
+      "epoch": 0.02589107743746919,
+      "grad_norm": 1.875,
+      "learning_rate": 1.9999575860974817e-05,
+      "loss": 1.2139,
+      "step": 151
+    },
+    {
+      "epoch": 0.02606254152645905,
+      "grad_norm": 2.0625,
+      "learning_rate": 1.9999559065127627e-05,
+      "loss": 1.208,
+      "step": 152
+    },
+    {
+      "epoch": 0.026234005615448915,
+      "grad_norm": 1.9765625,
+      "learning_rate": 1.999954194315716e-05,
+      "loss": 1.1737,
+      "step": 153
+    },
+    {
+      "epoch": 0.026405469704438776,
+      "grad_norm": 1.890625,
+      "learning_rate": 1.9999524495063974e-05,
+      "loss": 1.1485,
+      "step": 154
+    },
+    {
+      "epoch": 0.02657693379342864,
+      "grad_norm": 1.9921875,
+      "learning_rate": 1.999950672084864e-05,
+      "loss": 1.1551,
+      "step": 155
+    },
+    {
+      "epoch": 0.0267483978824185,
+      "grad_norm": 1.8984375,
+      "learning_rate": 1.999948862051174e-05,
+      "loss": 1.1953,
+      "step": 156
+    },
+    {
+      "epoch": 0.026919861971408363,
+      "grad_norm": 1.90625,
+      "learning_rate": 1.999947019405386e-05,
+      "loss": 1.1708,
+      "step": 157
+    },
+    {
+      "epoch": 0.027091326060398227,
+      "grad_norm": 1.859375,
+      "learning_rate": 1.99994514414756e-05,
+      "loss": 1.2106,
+      "step": 158
+    },
+    {
+      "epoch": 0.027262790149388087,
+      "grad_norm": 2.28125,
+      "learning_rate": 1.999943236277758e-05,
+      "loss": 1.2831,
+      "step": 159
+    },
+    {
+      "epoch": 0.02743425423837795,
+      "grad_norm": 2.125,
+      "learning_rate": 1.999941295796041e-05,
+      "loss": 1.2245,
+      "step": 160
+    },
+    {
+      "epoch": 0.02760571832736781,
+      "grad_norm": 1.953125,
+      "learning_rate": 1.9999393227024733e-05,
+      "loss": 1.0303,
+      "step": 161
+    },
+    {
+      "epoch": 0.027777182416357675,
+      "grad_norm": 3.296875,
+      "learning_rate": 1.9999373169971184e-05,
+      "loss": 1.1286,
+      "step": 162
+    },
+    {
+      "epoch": 0.027948646505347535,
+      "grad_norm": 2.90625,
+      "learning_rate": 1.9999352786800427e-05,
+      "loss": 1.1177,
+      "step": 163
+    },
+    {
+      "epoch": 0.0281201105943374,
+      "grad_norm": 2.109375,
+      "learning_rate": 1.9999332077513118e-05,
+      "loss": 1.1579,
+      "step": 164
+    },
+    {
+      "epoch": 0.02829157468332726,
+      "grad_norm": 6.3125,
+      "learning_rate": 1.9999311042109938e-05,
+      "loss": 1.1761,
+      "step": 165
+    },
+    {
+      "epoch": 0.028463038772317123,
+      "grad_norm": 4.5,
+      "learning_rate": 1.9999289680591573e-05,
+      "loss": 1.1855,
+      "step": 166
+    },
+    {
+      "epoch": 0.028634502861306987,
+      "grad_norm": 3.03125,
+      "learning_rate": 1.9999267992958713e-05,
+      "loss": 1.1755,
+      "step": 167
+    },
+    {
+      "epoch": 0.028805966950296847,
+      "grad_norm": 2.453125,
+      "learning_rate": 1.999924597921207e-05,
+      "loss": 1.1019,
+      "step": 168
+    },
+    {
+      "epoch": 0.02897743103928671,
+      "grad_norm": 1.8515625,
+      "learning_rate": 1.9999223639352364e-05,
+      "loss": 1.1054,
+      "step": 169
+    },
+    {
+      "epoch": 0.02914889512827657,
+      "grad_norm": 1.9609375,
+      "learning_rate": 1.9999200973380325e-05,
+      "loss": 1.0905,
+      "step": 170
+    },
+    {
+      "epoch": 0.029320359217266435,
+      "grad_norm": 2.046875,
+      "learning_rate": 1.9999177981296682e-05,
+      "loss": 1.1277,
+      "step": 171
+    },
+    {
+      "epoch": 0.029491823306256295,
+      "grad_norm": 1.8046875,
+      "learning_rate": 1.9999154663102196e-05,
+      "loss": 1.1312,
+      "step": 172
+    },
+    {
+      "epoch": 0.02966328739524616,
+      "grad_norm": 2.015625,
+      "learning_rate": 1.999913101879762e-05,
+      "loss": 1.1578,
+      "step": 173
+    },
+    {
+      "epoch": 0.02983475148423602,
+      "grad_norm": 2.0,
+      "learning_rate": 1.999910704838373e-05,
+      "loss": 1.1742,
+      "step": 174
+    },
+    {
+      "epoch": 0.030006215573225883,
+      "grad_norm": 1.8671875,
+      "learning_rate": 1.9999082751861308e-05,
+      "loss": 1.1161,
+      "step": 175
+    },
+    {
+      "epoch": 0.030177679662215743,
+      "grad_norm": 1.9609375,
+      "learning_rate": 1.9999058129231144e-05,
+      "loss": 1.1534,
+      "step": 176
+    },
+    {
+      "epoch": 0.030349143751205607,
+      "grad_norm": 1.96875,
+      "learning_rate": 1.9999033180494037e-05,
+      "loss": 1.2287,
+      "step": 177
+    },
+    {
+      "epoch": 0.03052060784019547,
+      "grad_norm": 1.9453125,
+      "learning_rate": 1.999900790565081e-05,
+      "loss": 1.1124,
+      "step": 178
+    },
+    {
+      "epoch": 0.03069207192918533,
+      "grad_norm": 1.75,
+      "learning_rate": 1.9998982304702278e-05,
+      "loss": 1.1163,
+      "step": 179
+    },
+    {
+      "epoch": 0.030863536018175194,
+      "grad_norm": 2.140625,
+      "learning_rate": 1.9998956377649286e-05,
+      "loss": 1.1342,
+      "step": 180
+    },
+    {
+      "epoch": 0.031035000107165055,
+      "grad_norm": 1.9609375,
+      "learning_rate": 1.999893012449267e-05,
+      "loss": 1.2135,
+      "step": 181
+    },
+    {
+      "epoch": 0.03120646419615492,
+      "grad_norm": 1.7265625,
+      "learning_rate": 1.9998903545233293e-05,
+      "loss": 1.053,
+      "step": 182
+    },
+    {
+      "epoch": 0.03137792828514478,
+      "grad_norm": 1.9765625,
+      "learning_rate": 1.9998876639872016e-05,
+      "loss": 1.1699,
+      "step": 183
+    },
+    {
+      "epoch": 0.03154939237413464,
+      "grad_norm": 1.9921875,
+      "learning_rate": 1.999884940840972e-05,
+      "loss": 1.1576,
+      "step": 184
+    },
+    {
+      "epoch": 0.0317208564631245,
+      "grad_norm": 1.8828125,
+      "learning_rate": 1.9998821850847296e-05,
+      "loss": 1.1212,
+      "step": 185
+    },
+    {
+      "epoch": 0.03189232055211437,
+      "grad_norm": 2.046875,
+      "learning_rate": 1.9998793967185635e-05,
+      "loss": 1.2487,
+      "step": 186
+    },
+    {
+      "epoch": 0.03206378464110423,
+      "grad_norm": 1.953125,
+      "learning_rate": 1.999876575742565e-05,
+      "loss": 1.1613,
+      "step": 187
+    },
+    {
+      "epoch": 0.03223524873009409,
+      "grad_norm": 1.84375,
+      "learning_rate": 1.9998737221568264e-05,
+      "loss": 1.1357,
+      "step": 188
+    },
+    {
+      "epoch": 0.03240671281908395,
+      "grad_norm": 1.875,
+      "learning_rate": 1.9998708359614408e-05,
+      "loss": 1.0308,
+      "step": 189
+    },
+    {
+      "epoch": 0.03257817690807382,
+      "grad_norm": 1.9375,
+      "learning_rate": 1.9998679171565017e-05,
+      "loss": 1.1871,
+      "step": 190
+    },
+    {
+      "epoch": 0.03274964099706368,
+      "grad_norm": 1.9765625,
+      "learning_rate": 1.9998649657421047e-05,
+      "loss": 1.1016,
+      "step": 191
+    },
+    {
+      "epoch": 0.03292110508605354,
+      "grad_norm": 1.84375,
+      "learning_rate": 1.999861981718346e-05,
+      "loss": 1.1855,
+      "step": 192
+    },
+    {
+      "epoch": 0.0330925691750434,
+      "grad_norm": 2.765625,
+      "learning_rate": 1.999858965085323e-05,
+      "loss": 1.2059,
+      "step": 193
+    },
+    {
+      "epoch": 0.033264033264033266,
+      "grad_norm": 2.0625,
+      "learning_rate": 1.999855915843134e-05,
+      "loss": 1.2393,
+      "step": 194
+    },
+    {
+      "epoch": 0.033435497353023126,
+      "grad_norm": 1.90625,
+      "learning_rate": 1.999852833991879e-05,
+      "loss": 1.2064,
+      "step": 195
+    },
+    {
+      "epoch": 0.033606961442012986,
+      "grad_norm": 1.84375,
+      "learning_rate": 1.9998497195316572e-05,
+      "loss": 1.0907,
+      "step": 196
+    },
+    {
+      "epoch": 0.033778425531002854,
+      "grad_norm": 1.921875,
+      "learning_rate": 1.9998465724625715e-05,
+      "loss": 1.2311,
+      "step": 197
+    },
+    {
+      "epoch": 0.033949889619992714,
+      "grad_norm": 1.8125,
+      "learning_rate": 1.9998433927847238e-05,
+      "loss": 1.1369,
+      "step": 198
+    },
+    {
+      "epoch": 0.034121353708982574,
+      "grad_norm": 1.78125,
+      "learning_rate": 1.999840180498218e-05,
+      "loss": 1.1819,
+      "step": 199
+    },
+    {
+      "epoch": 0.034292817797972434,
+      "grad_norm": 1.8046875,
+      "learning_rate": 1.9998369356031587e-05,
+      "loss": 1.1348,
+      "step": 200
+    },
+    {
+      "epoch": 0.0344642818869623,
+      "grad_norm": 1.75,
+      "learning_rate": 1.9998336580996524e-05,
+      "loss": 1.141,
+      "step": 201
+    },
+    {
+      "epoch": 0.03463574597595216,
+      "grad_norm": 1.8359375,
+      "learning_rate": 1.999830347987805e-05,
+      "loss": 1.1466,
+      "step": 202
+    },
+    {
+      "epoch": 0.03480721006494202,
+      "grad_norm": 1.796875,
+      "learning_rate": 1.999827005267725e-05,
+      "loss": 1.2022,
+      "step": 203
+    },
+    {
+      "epoch": 0.03497867415393189,
+      "grad_norm": 1.8125,
+      "learning_rate": 1.9998236299395216e-05,
+      "loss": 1.2304,
+      "step": 204
+    },
+    {
+      "epoch": 0.03515013824292175,
+      "grad_norm": 1.875,
+      "learning_rate": 1.9998202220033044e-05,
+      "loss": 1.1918,
+      "step": 205
+    },
+    {
+      "epoch": 0.03532160233191161,
+      "grad_norm": 1.8203125,
+      "learning_rate": 1.9998167814591847e-05,
+      "loss": 1.1763,
+      "step": 206
+    },
+    {
+      "epoch": 0.03549306642090147,
+      "grad_norm": 1.8828125,
+      "learning_rate": 1.999813308307275e-05,
+      "loss": 1.1952,
+      "step": 207
+    },
+    {
+      "epoch": 0.03566453050989134,
+      "grad_norm": 1.9375,
+      "learning_rate": 1.9998098025476883e-05,
+      "loss": 1.0583,
+      "step": 208
+    },
+    {
+      "epoch": 0.0358359945988812,
+      "grad_norm": 4.21875,
+      "learning_rate": 1.9998062641805392e-05,
+      "loss": 1.1122,
+      "step": 209
+    },
+    {
+      "epoch": 0.03600745868787106,
+      "grad_norm": 1.890625,
+      "learning_rate": 1.9998026932059427e-05,
+      "loss": 1.0917,
+      "step": 210
+    },
+    {
+      "epoch": 0.03617892277686092,
+      "grad_norm": 1.8125,
+      "learning_rate": 1.9997990896240154e-05,
+      "loss": 1.1909,
+      "step": 211
+    },
+    {
+      "epoch": 0.036350386865850785,
+      "grad_norm": 1.796875,
+      "learning_rate": 1.999795453434875e-05,
+      "loss": 1.1432,
+      "step": 212
+    },
+    {
+      "epoch": 0.036521850954840646,
+      "grad_norm": 1.78125,
+      "learning_rate": 1.99979178463864e-05,
+      "loss": 1.0841,
+      "step": 213
+    },
+    {
+      "epoch": 0.036693315043830506,
+      "grad_norm": 1.796875,
+      "learning_rate": 1.9997880832354302e-05,
+      "loss": 1.2121,
+      "step": 214
+    },
+    {
+      "epoch": 0.03686477913282037,
+      "grad_norm": 1.8671875,
+      "learning_rate": 1.999784349225366e-05,
+      "loss": 1.0489,
+      "step": 215
+    },
+    {
+      "epoch": 0.03703624322181023,
+      "grad_norm": 1.8046875,
+      "learning_rate": 1.999780582608569e-05,
+      "loss": 1.1155,
+      "step": 216
+    },
+    {
+      "epoch": 0.037207707310800094,
+      "grad_norm": 2.234375,
+      "learning_rate": 1.999776783385163e-05,
+      "loss": 1.1808,
+      "step": 217
+    },
+    {
+      "epoch": 0.037379171399789954,
+      "grad_norm": 2.125,
+      "learning_rate": 1.9997729515552708e-05,
+      "loss": 1.2913,
+      "step": 218
+    },
+    {
+      "epoch": 0.03755063548877982,
+      "grad_norm": 1.875,
+      "learning_rate": 1.999769087119018e-05,
+      "loss": 1.1422,
+      "step": 219
+    },
+    {
+      "epoch": 0.03772209957776968,
+      "grad_norm": 1.9765625,
+      "learning_rate": 1.9997651900765308e-05,
+      "loss": 1.1079,
+      "step": 220
+    },
+    {
+      "epoch": 0.03789356366675954,
+      "grad_norm": 1.7890625,
+      "learning_rate": 1.999761260427935e-05,
+      "loss": 1.1709,
+      "step": 221
+    },
+    {
+      "epoch": 0.0380650277557494,
+      "grad_norm": 2.1875,
+      "learning_rate": 1.999757298173361e-05,
+      "loss": 1.2323,
+      "step": 222
+    },
+    {
+      "epoch": 0.03823649184473927,
+      "grad_norm": 1.953125,
+      "learning_rate": 1.999753303312936e-05,
+      "loss": 1.1526,
+      "step": 223
+    },
+    {
+      "epoch": 0.03840795593372913,
+      "grad_norm": 1.921875,
+      "learning_rate": 1.9997492758467915e-05,
+      "loss": 1.0609,
+      "step": 224
+    },
+    {
+      "epoch": 0.03857942002271899,
+      "grad_norm": 1.7890625,
+      "learning_rate": 1.9997452157750577e-05,
+      "loss": 1.1155,
+      "step": 225
+    },
+    {
+      "epoch": 0.03875088411170886,
+      "grad_norm": 1.8125,
+      "learning_rate": 1.9997411230978684e-05,
+      "loss": 1.1335,
+      "step": 226
+    },
+    {
+      "epoch": 0.03892234820069872,
+      "grad_norm": 1.9296875,
+      "learning_rate": 1.9997369978153564e-05,
+      "loss": 1.1145,
+      "step": 227
+    },
+    {
+      "epoch": 0.03909381228968858,
+      "grad_norm": 1.96875,
+      "learning_rate": 1.9997328399276558e-05,
+      "loss": 1.0612,
+      "step": 228
+    },
+    {
+      "epoch": 0.03926527637867844,
+      "grad_norm": 1.7890625,
+      "learning_rate": 1.9997286494349032e-05,
+      "loss": 1.1189,
+      "step": 229
+    },
+    {
+      "epoch": 0.039436740467668305,
+      "grad_norm": 1.84375,
+      "learning_rate": 1.999724426337234e-05,
+      "loss": 1.1008,
+      "step": 230
+    },
+    {
+      "epoch": 0.039608204556658165,
+      "grad_norm": 2.359375,
+      "learning_rate": 1.9997201706347875e-05,
+      "loss": 1.1649,
+      "step": 231
+    },
+    {
+      "epoch": 0.039779668645648025,
+      "grad_norm": 2.015625,
+      "learning_rate": 1.999715882327701e-05,
+      "loss": 1.1094,
+      "step": 232
+    },
+    {
+      "epoch": 0.03995113273463789,
+      "grad_norm": 1.84375,
+      "learning_rate": 1.999711561416115e-05,
+      "loss": 1.1228,
+      "step": 233
+    },
+    {
+      "epoch": 0.04012259682362775,
+      "grad_norm": 1.8125,
+      "learning_rate": 1.9997072079001705e-05,
+      "loss": 1.1056,
+      "step": 234
+    },
+    {
+      "epoch": 0.04029406091261761,
+      "grad_norm": 1.9296875,
+      "learning_rate": 1.9997028217800097e-05,
+      "loss": 1.2507,
+      "step": 235
+    },
+    {
+      "epoch": 0.04046552500160747,
+      "grad_norm": 1.796875,
+      "learning_rate": 1.999698403055775e-05,
+      "loss": 1.0986,
+      "step": 236
+    },
+    {
+      "epoch": 0.04063698909059734,
+      "grad_norm": 2.59375,
+      "learning_rate": 1.9996939517276107e-05,
+      "loss": 1.1712,
+      "step": 237
+    },
+    {
+      "epoch": 0.0408084531795872,
+      "grad_norm": 1.765625,
+      "learning_rate": 1.9996894677956628e-05,
+      "loss": 1.197,
+      "step": 238
+    },
+    {
+      "epoch": 0.04097991726857706,
+      "grad_norm": 1.953125,
+      "learning_rate": 1.9996849512600764e-05,
+      "loss": 1.0771,
+      "step": 239
+    },
+    {
+      "epoch": 0.04115138135756692,
+      "grad_norm": 1.8359375,
+      "learning_rate": 1.9996804021209995e-05,
+      "loss": 1.191,
+      "step": 240
+    },
+    {
+      "epoch": 0.04132284544655679,
+      "grad_norm": 1.8515625,
+      "learning_rate": 1.9996758203785797e-05,
+      "loss": 1.1842,
+      "step": 241
+    },
+    {
+      "epoch": 0.04149430953554665,
+      "grad_norm": 1.8359375,
+      "learning_rate": 1.9996712060329675e-05,
+      "loss": 1.1033,
+      "step": 242
+    },
+    {
+      "epoch": 0.04166577362453651,
+      "grad_norm": 1.734375,
+      "learning_rate": 1.999666559084313e-05,
+      "loss": 1.0938,
+      "step": 243
+    },
+    {
+      "epoch": 0.041837237713526376,
+      "grad_norm": 1.859375,
+      "learning_rate": 1.999661879532767e-05,
+      "loss": 1.2002,
+      "step": 244
+    },
+    {
+      "epoch": 0.042008701802516236,
+      "grad_norm": 1.7578125,
+      "learning_rate": 1.999657167378483e-05,
+      "loss": 1.195,
+      "step": 245
+    },
+    {
+      "epoch": 0.0421801658915061,
+      "grad_norm": 1.75,
+      "learning_rate": 1.9996524226216147e-05,
+      "loss": 1.1571,
+      "step": 246
+    },
+    {
+      "epoch": 0.04235162998049596,
+      "grad_norm": 1.921875,
+      "learning_rate": 1.9996476452623163e-05,
+      "loss": 1.0698,
+      "step": 247
+    },
+    {
+      "epoch": 0.042523094069485824,
+      "grad_norm": 1.8203125,
+      "learning_rate": 1.999642835300744e-05,
+      "loss": 1.1737,
+      "step": 248
+    },
+    {
+      "epoch": 0.042694558158475684,
+      "grad_norm": 1.90625,
+      "learning_rate": 1.9996379927370542e-05,
+      "loss": 1.1447,
+      "step": 249
+    },
+    {
+      "epoch": 0.042866022247465545,
+      "grad_norm": 1.8515625,
+      "learning_rate": 1.9996331175714056e-05,
+      "loss": 1.2027,
+      "step": 250
+    },
+    {
+      "epoch": 0.04303748633645541,
+      "grad_norm": 1.8515625,
+      "learning_rate": 1.9996282098039565e-05,
+      "loss": 1.109,
+      "step": 251
+    },
+    {
+      "epoch": 0.04320895042544527,
+      "grad_norm": 1.8203125,
+      "learning_rate": 1.9996232694348673e-05,
+      "loss": 1.1268,
+      "step": 252
+    },
+    {
+      "epoch": 0.04338041451443513,
+      "grad_norm": 1.7578125,
+      "learning_rate": 1.9996182964642992e-05,
+      "loss": 1.119,
+      "step": 253
+    },
+    {
+      "epoch": 0.04355187860342499,
+      "grad_norm": 1.7109375,
+      "learning_rate": 1.999613290892414e-05,
+      "loss": 1.139,
+      "step": 254
+    },
+    {
+      "epoch": 0.04372334269241486,
+      "grad_norm": 1.7421875,
+      "learning_rate": 1.999608252719375e-05,
+      "loss": 1.1325,
+      "step": 255
+    },
+    {
+      "epoch": 0.04389480678140472,
+      "grad_norm": 1.84375,
+      "learning_rate": 1.9996031819453474e-05,
+      "loss": 1.1633,
+      "step": 256
+    },
+    {
+      "epoch": 0.04406627087039458,
+      "grad_norm": 1.8828125,
+      "learning_rate": 1.9995980785704955e-05,
+      "loss": 1.2083,
+      "step": 257
+    },
+    {
+      "epoch": 0.04423773495938444,
+      "grad_norm": 1.71875,
+      "learning_rate": 1.999592942594986e-05,
+      "loss": 1.043,
+      "step": 258
+    },
+    {
+      "epoch": 0.04440919904837431,
+      "grad_norm": 1.90625,
+      "learning_rate": 1.9995877740189868e-05,
+      "loss": 1.153,
+      "step": 259
+    },
+    {
+      "epoch": 0.04458066313736417,
+      "grad_norm": 2.015625,
+      "learning_rate": 1.9995825728426655e-05,
+      "loss": 1.1146,
+      "step": 260
+    },
+    {
+      "epoch": 0.04475212722635403,
+      "grad_norm": 1.96875,
+      "learning_rate": 1.999577339066193e-05,
+      "loss": 1.1924,
+      "step": 261
+    },
+    {
+      "epoch": 0.044923591315343896,
+      "grad_norm": 1.859375,
+      "learning_rate": 1.9995720726897394e-05,
+      "loss": 1.1448,
+      "step": 262
+    },
+    {
+      "epoch": 0.045095055404333756,
+      "grad_norm": 1.9375,
+      "learning_rate": 1.9995667737134765e-05,
+      "loss": 1.1947,
+      "step": 263
+    },
+    {
+      "epoch": 0.045266519493323616,
+      "grad_norm": 1.8359375,
+      "learning_rate": 1.9995614421375768e-05,
+      "loss": 1.1596,
+      "step": 264
+    },
+    {
+      "epoch": 0.045437983582313476,
+      "grad_norm": 1.765625,
+      "learning_rate": 1.9995560779622148e-05,
+      "loss": 1.104,
+      "step": 265
+    },
+    {
+      "epoch": 0.045609447671303344,
+      "grad_norm": 1.9375,
+      "learning_rate": 1.999550681187565e-05,
+      "loss": 1.1444,
+      "step": 266
+    },
+    {
+      "epoch": 0.045780911760293204,
+      "grad_norm": 2.0,
+      "learning_rate": 1.9995452518138037e-05,
+      "loss": 1.1601,
+      "step": 267
+    },
+    {
+      "epoch": 0.045952375849283064,
+      "grad_norm": 1.875,
+      "learning_rate": 1.9995397898411073e-05,
+      "loss": 1.1729,
+      "step": 268
+    },
+    {
+      "epoch": 0.04612383993827293,
+      "grad_norm": 1.7734375,
+      "learning_rate": 1.999534295269655e-05,
+      "loss": 1.0968,
+      "step": 269
+    },
+    {
+      "epoch": 0.04629530402726279,
+      "grad_norm": 1.75,
+      "learning_rate": 1.999528768099625e-05,
+      "loss": 1.1298,
+      "step": 270
+    },
+    {
+      "epoch": 0.04646676811625265,
+      "grad_norm": 2.0,
+      "learning_rate": 1.999523208331198e-05,
+      "loss": 1.0809,
+      "step": 271
+    },
+    {
+      "epoch": 0.04663823220524251,
+      "grad_norm": 1.765625,
+      "learning_rate": 1.9995176159645557e-05,
+      "loss": 1.1636,
+      "step": 272
+    },
+    {
+      "epoch": 0.04680969629423238,
+      "grad_norm": 1.796875,
+      "learning_rate": 1.9995119909998798e-05,
+      "loss": 1.1,
+      "step": 273
+    },
+    {
+      "epoch": 0.04698116038322224,
+      "grad_norm": 4.6875,
+      "learning_rate": 1.9995063334373544e-05,
+      "loss": 1.1314,
+      "step": 274
+    },
+    {
+      "epoch": 0.0471526244722121,
+      "grad_norm": 1.859375,
+      "learning_rate": 1.9995006432771634e-05,
+      "loss": 1.1535,
+      "step": 275
+    },
+    {
+      "epoch": 0.04732408856120196,
+      "grad_norm": 1.734375,
+      "learning_rate": 1.9994949205194925e-05,
+      "loss": 1.0962,
+      "step": 276
+    },
+    {
+      "epoch": 0.04749555265019183,
+      "grad_norm": 1.875,
+      "learning_rate": 1.999489165164529e-05,
+      "loss": 1.2084,
+      "step": 277
+    },
+    {
+      "epoch": 0.04766701673918169,
+      "grad_norm": 1.765625,
+      "learning_rate": 1.9994833772124597e-05,
+      "loss": 1.1022,
+      "step": 278
+    },
+    {
+      "epoch": 0.04783848082817155,
+      "grad_norm": 1.8359375,
+      "learning_rate": 1.9994775566634737e-05,
+      "loss": 1.1553,
+      "step": 279
+    },
+    {
+      "epoch": 0.048009944917161415,
+      "grad_norm": 1.796875,
+      "learning_rate": 1.999471703517761e-05,
+      "loss": 1.1098,
+      "step": 280
+    },
+    {
+      "epoch": 0.048181409006151275,
+      "grad_norm": 1.8046875,
+      "learning_rate": 1.9994658177755124e-05,
+      "loss": 1.1211,
+      "step": 281
+    },
+    {
+      "epoch": 0.048352873095141136,
+      "grad_norm": 1.9296875,
+      "learning_rate": 1.99945989943692e-05,
+      "loss": 1.1674,
+      "step": 282
+    },
+    {
+      "epoch": 0.048524337184130996,
+      "grad_norm": 1.734375,
+      "learning_rate": 1.9994539485021765e-05,
+      "loss": 1.086,
+      "step": 283
+    },
+    {
+      "epoch": 0.04869580127312086,
+      "grad_norm": 1.921875,
+      "learning_rate": 1.999447964971476e-05,
+      "loss": 1.2183,
+      "step": 284
+    },
+    {
+      "epoch": 0.04886726536211072,
+      "grad_norm": 1.796875,
+      "learning_rate": 1.999441948845014e-05,
+      "loss": 1.1132,
+      "step": 285
+    },
+    {
+      "epoch": 0.049038729451100584,
+      "grad_norm": 1.765625,
+      "learning_rate": 1.9994359001229865e-05,
+      "loss": 1.0739,
+      "step": 286
+    },
+    {
+      "epoch": 0.049210193540090444,
+      "grad_norm": 1.8671875,
+      "learning_rate": 1.9994298188055907e-05,
+      "loss": 1.1649,
+      "step": 287
+    },
+    {
+      "epoch": 0.04938165762908031,
+      "grad_norm": 1.8828125,
+      "learning_rate": 1.999423704893025e-05,
+      "loss": 1.1438,
+      "step": 288
+    },
+    {
+      "epoch": 0.04955312171807017,
+      "grad_norm": 2.015625,
+      "learning_rate": 1.999417558385489e-05,
+      "loss": 1.166,
+      "step": 289
+    },
+    {
+      "epoch": 0.04972458580706003,
+      "grad_norm": 2.015625,
+      "learning_rate": 1.9994113792831825e-05,
+      "loss": 1.14,
+      "step": 290
+    },
+    {
+      "epoch": 0.0498960498960499,
+      "grad_norm": 1.8984375,
+      "learning_rate": 1.999405167586308e-05,
+      "loss": 1.186,
+      "step": 291
+    },
+    {
+      "epoch": 0.05006751398503976,
+      "grad_norm": 1.96875,
+      "learning_rate": 1.999398923295067e-05,
+      "loss": 1.1538,
+      "step": 292
+    },
+    {
+      "epoch": 0.05023897807402962,
+      "grad_norm": 1.84375,
+      "learning_rate": 1.9993926464096646e-05,
+      "loss": 1.1432,
+      "step": 293
+    },
+    {
+      "epoch": 0.05041044216301948,
+      "grad_norm": 1.78125,
+      "learning_rate": 1.999386336930304e-05,
+      "loss": 1.0595,
+      "step": 294
+    },
+    {
+      "epoch": 0.05058190625200935,
+      "grad_norm": 1.8203125,
+      "learning_rate": 1.999379994857192e-05,
+      "loss": 1.128,
+      "step": 295
+    },
+    {
+      "epoch": 0.05075337034099921,
+      "grad_norm": 1.78125,
+      "learning_rate": 1.9993736201905343e-05,
+      "loss": 1.1186,
+      "step": 296
+    },
+    {
+      "epoch": 0.05092483442998907,
+      "grad_norm": 1.71875,
+      "learning_rate": 1.9993672129305398e-05,
+      "loss": 1.086,
+      "step": 297
+    },
+    {
+      "epoch": 0.051096298518978935,
+      "grad_norm": 1.8125,
+      "learning_rate": 1.9993607730774176e-05,
+      "loss": 1.1423,
+      "step": 298
+    },
+    {
+      "epoch": 0.051267762607968795,
+      "grad_norm": 1.75,
+      "learning_rate": 1.9993543006313772e-05,
+      "loss": 1.1484,
+      "step": 299
+    },
+    {
+      "epoch": 0.051439226696958655,
+      "grad_norm": 1.7421875,
+      "learning_rate": 1.9993477955926298e-05,
+      "loss": 1.1604,
+      "step": 300
+    },
+    {
+      "epoch": 0.051610690785948515,
+      "grad_norm": 1.9375,
+      "learning_rate": 1.999341257961388e-05,
+      "loss": 1.1192,
+      "step": 301
+    },
+    {
+      "epoch": 0.05178215487493838,
+      "grad_norm": 1.90625,
+      "learning_rate": 1.9993346877378637e-05,
+      "loss": 1.0958,
+      "step": 302
+    },
+    {
+      "epoch": 0.05195361896392824,
+      "grad_norm": 1.828125,
+      "learning_rate": 1.9993280849222726e-05,
+      "loss": 1.0935,
+      "step": 303
+    },
+    {
+      "epoch": 0.0521250830529181,
+      "grad_norm": 1.8125,
+      "learning_rate": 1.9993214495148297e-05,
+      "loss": 1.1584,
+      "step": 304
+    },
+    {
+      "epoch": 0.05229654714190796,
+      "grad_norm": 1.7578125,
+      "learning_rate": 1.999314781515751e-05,
+      "loss": 1.0625,
+      "step": 305
+    },
+    {
+      "epoch": 0.05246801123089783,
+      "grad_norm": 1.75,
+      "learning_rate": 1.9993080809252542e-05,
+      "loss": 1.1468,
+      "step": 306
+    },
+    {
+      "epoch": 0.05263947531988769,
+      "grad_norm": 1.7421875,
+      "learning_rate": 1.999301347743558e-05,
+      "loss": 1.1285,
+      "step": 307
+    },
+    {
+      "epoch": 0.05281093940887755,
+      "grad_norm": 2.0,
+      "learning_rate": 1.999294581970882e-05,
+      "loss": 1.2316,
+      "step": 308
+    },
+    {
+      "epoch": 0.05298240349786742,
+      "grad_norm": 1.8515625,
+      "learning_rate": 1.9992877836074465e-05,
+      "loss": 1.1048,
+      "step": 309
+    },
+    {
+      "epoch": 0.05315386758685728,
+      "grad_norm": 1.8046875,
+      "learning_rate": 1.9992809526534732e-05,
+      "loss": 1.0039,
+      "step": 310
+    },
+    {
+      "epoch": 0.05332533167584714,
+      "grad_norm": 1.734375,
+      "learning_rate": 1.9992740891091852e-05,
+      "loss": 1.0773,
+      "step": 311
+    },
+    {
+      "epoch": 0.053496795764837,
+      "grad_norm": 1.7421875,
+      "learning_rate": 1.9992671929748062e-05,
+      "loss": 1.1283,
+      "step": 312
+    },
+    {
+      "epoch": 0.053668259853826866,
+      "grad_norm": 1.8671875,
+      "learning_rate": 1.999260264250561e-05,
+      "loss": 1.1078,
+      "step": 313
+    },
+    {
+      "epoch": 0.053839723942816727,
+      "grad_norm": 1.8046875,
+      "learning_rate": 1.9992533029366763e-05,
+      "loss": 1.1397,
+      "step": 314
+    },
+    {
+      "epoch": 0.05401118803180659,
+      "grad_norm": 1.6640625,
+      "learning_rate": 1.999246309033378e-05,
+      "loss": 1.0898,
+      "step": 315
+    },
+    {
+      "epoch": 0.054182652120796454,
+      "grad_norm": 1.6796875,
+      "learning_rate": 1.999239282540895e-05,
+      "loss": 1.066,
+      "step": 316
+    },
+    {
+      "epoch": 0.054354116209786314,
+      "grad_norm": 1.8203125,
+      "learning_rate": 1.9992322234594562e-05,
+      "loss": 1.2013,
+      "step": 317
+    },
+    {
+      "epoch": 0.054525580298776175,
+      "grad_norm": 1.8046875,
+      "learning_rate": 1.9992251317892916e-05,
+      "loss": 1.1571,
+      "step": 318
+    },
+    {
+      "epoch": 0.054697044387766035,
+      "grad_norm": 1.6953125,
+      "learning_rate": 1.999218007530633e-05,
+      "loss": 1.1173,
+      "step": 319
+    },
+    {
+      "epoch": 0.0548685084767559,
+      "grad_norm": 1.7109375,
+      "learning_rate": 1.9992108506837122e-05,
+      "loss": 1.099,
+      "step": 320
+    },
+    {
+      "epoch": 0.05503997256574576,
+      "grad_norm": 1.78125,
+      "learning_rate": 1.999203661248763e-05,
+      "loss": 1.1084,
+      "step": 321
+    },
+    {
+      "epoch": 0.05521143665473562,
+      "grad_norm": 1.765625,
+      "learning_rate": 1.9991964392260198e-05,
+      "loss": 1.0651,
+      "step": 322
+    },
+    {
+      "epoch": 0.05538290074372548,
+      "grad_norm": 1.7890625,
+      "learning_rate": 1.9991891846157182e-05,
+      "loss": 1.1868,
+      "step": 323
+    },
+    {
+      "epoch": 0.05555436483271535,
+      "grad_norm": 3.6875,
+      "learning_rate": 1.9991818974180944e-05,
+      "loss": 1.189,
+      "step": 324
+    },
+    {
+      "epoch": 0.05572582892170521,
+      "grad_norm": 1.734375,
+      "learning_rate": 1.9991745776333865e-05,
+      "loss": 1.1314,
+      "step": 325
+    },
+    {
+      "epoch": 0.05589729301069507,
+      "grad_norm": 2.234375,
+      "learning_rate": 1.9991672252618334e-05,
+      "loss": 1.1119,
+      "step": 326
+    },
+    {
+      "epoch": 0.05606875709968494,
+      "grad_norm": 2.015625,
+      "learning_rate": 1.999159840303674e-05,
+      "loss": 1.103,
+      "step": 327
+    },
+    {
+      "epoch": 0.0562402211886748,
+      "grad_norm": 1.7109375,
+      "learning_rate": 1.99915242275915e-05,
+      "loss": 1.1243,
+      "step": 328
+    },
+    {
+      "epoch": 0.05641168527766466,
+      "grad_norm": 1.96875,
+      "learning_rate": 1.9991449726285033e-05,
+      "loss": 1.1458,
+      "step": 329
+    },
+    {
+      "epoch": 0.05658314936665452,
+      "grad_norm": 1.8359375,
+      "learning_rate": 1.9991374899119763e-05,
+      "loss": 1.1669,
+      "step": 330
+    },
+    {
+      "epoch": 0.056754613455644386,
+      "grad_norm": 1.7265625,
+      "learning_rate": 1.9991299746098133e-05,
+      "loss": 1.1497,
+      "step": 331
+    },
+    {
+      "epoch": 0.056926077544634246,
+      "grad_norm": 1.828125,
+      "learning_rate": 1.9991224267222596e-05,
+      "loss": 1.0866,
+      "step": 332
+    },
+    {
+      "epoch": 0.057097541633624106,
+      "grad_norm": 1.890625,
+      "learning_rate": 1.9991148462495612e-05,
+      "loss": 1.1331,
+      "step": 333
+    },
+    {
+      "epoch": 0.05726900572261397,
+      "grad_norm": 1.7578125,
+      "learning_rate": 1.999107233191965e-05,
+      "loss": 1.1717,
+      "step": 334
+    },
+    {
+      "epoch": 0.057440469811603834,
+      "grad_norm": 1.734375,
+      "learning_rate": 1.9990995875497203e-05,
+      "loss": 1.1216,
+      "step": 335
+    },
+    {
+      "epoch": 0.057611933900593694,
+      "grad_norm": 1.671875,
+      "learning_rate": 1.9990919093230752e-05,
+      "loss": 1.0493,
+      "step": 336
+    },
+    {
+      "epoch": 0.057783397989583554,
+      "grad_norm": 1.8828125,
+      "learning_rate": 1.999084198512281e-05,
+      "loss": 1.0627,
+      "step": 337
+    },
+    {
+      "epoch": 0.05795486207857342,
+      "grad_norm": 1.9453125,
+      "learning_rate": 1.999076455117589e-05,
+      "loss": 1.0835,
+      "step": 338
+    },
+    {
+      "epoch": 0.05812632616756328,
+      "grad_norm": 1.765625,
+      "learning_rate": 1.999068679139251e-05,
+      "loss": 1.1594,
+      "step": 339
+    },
+    {
+      "epoch": 0.05829779025655314,
+      "grad_norm": 1.75,
+      "learning_rate": 1.9990608705775217e-05,
+      "loss": 1.1199,
+      "step": 340
+    },
+    {
+      "epoch": 0.058469254345543,
+      "grad_norm": 1.7578125,
+      "learning_rate": 1.9990530294326554e-05,
+      "loss": 1.1978,
+      "step": 341
+    },
+    {
+      "epoch": 0.05864071843453287,
+      "grad_norm": 1.8671875,
+      "learning_rate": 1.9990451557049077e-05,
+      "loss": 1.1584,
+      "step": 342
+    },
+    {
+      "epoch": 0.05881218252352273,
+      "grad_norm": 1.7578125,
+      "learning_rate": 1.9990372493945353e-05,
+      "loss": 1.1209,
+      "step": 343
+    },
+    {
+      "epoch": 0.05898364661251259,
+      "grad_norm": 1.796875,
+      "learning_rate": 1.9990293105017962e-05,
+      "loss": 1.1281,
+      "step": 344
+    },
+    {
+      "epoch": 0.05915511070150246,
+      "grad_norm": 1.734375,
+      "learning_rate": 1.999021339026949e-05,
+      "loss": 1.0771,
+      "step": 345
+    },
+    {
+      "epoch": 0.05932657479049232,
+      "grad_norm": 1.6796875,
+      "learning_rate": 1.9990133349702544e-05,
+      "loss": 1.1337,
+      "step": 346
+    },
+    {
+      "epoch": 0.05949803887948218,
+      "grad_norm": 1.8203125,
+      "learning_rate": 1.999005298331973e-05,
+      "loss": 1.1041,
+      "step": 347
+    },
+    {
+      "epoch": 0.05966950296847204,
+      "grad_norm": 1.8203125,
+      "learning_rate": 1.9989972291123666e-05,
+      "loss": 1.1493,
+      "step": 348
+    },
+    {
+      "epoch": 0.059840967057461905,
+      "grad_norm": 1.984375,
+      "learning_rate": 1.998989127311699e-05,
+      "loss": 1.1732,
+      "step": 349
+    },
+    {
+      "epoch": 0.060012431146451765,
+      "grad_norm": 1.8046875,
+      "learning_rate": 1.998980992930234e-05,
+      "loss": 1.0674,
+      "step": 350
+    },
+    {
+      "epoch": 0.060183895235441626,
+      "grad_norm": 3.609375,
+      "learning_rate": 1.9989728259682368e-05,
+      "loss": 1.23,
+      "step": 351
+    },
+    {
+      "epoch": 0.060355359324431486,
+      "grad_norm": 4.4375,
+      "learning_rate": 1.9989646264259743e-05,
+      "loss": 1.1521,
+      "step": 352
+    },
+    {
+      "epoch": 0.06052682341342135,
+      "grad_norm": 1.7734375,
+      "learning_rate": 1.9989563943037133e-05,
+      "loss": 1.1788,
+      "step": 353
+    },
+    {
+      "epoch": 0.06069828750241121,
+      "grad_norm": 1.7265625,
+      "learning_rate": 1.998948129601723e-05,
+      "loss": 1.1802,
+      "step": 354
+    },
+    {
+      "epoch": 0.060869751591401074,
+      "grad_norm": 1.6953125,
+      "learning_rate": 1.998939832320272e-05,
+      "loss": 1.073,
+      "step": 355
+    },
+    {
+      "epoch": 0.06104121568039094,
+      "grad_norm": 1.8515625,
+      "learning_rate": 1.9989315024596315e-05,
+      "loss": 1.1932,
+      "step": 356
+    },
+    {
+      "epoch": 0.0612126797693808,
+      "grad_norm": 1.8203125,
+      "learning_rate": 1.998923140020073e-05,
+      "loss": 1.0352,
+      "step": 357
+    },
+    {
+      "epoch": 0.06138414385837066,
+      "grad_norm": 1.7890625,
+      "learning_rate": 1.9989147450018698e-05,
+      "loss": 1.1364,
+      "step": 358
+    },
+    {
+      "epoch": 0.06155560794736052,
+      "grad_norm": 1.6328125,
+      "learning_rate": 1.9989063174052948e-05,
+      "loss": 1.0597,
+      "step": 359
+    },
+    {
+      "epoch": 0.06172707203635039,
+      "grad_norm": 1.78125,
+      "learning_rate": 1.998897857230623e-05,
+      "loss": 1.1841,
+      "step": 360
+    },
+    {
+      "epoch": 0.06189853612534025,
+      "grad_norm": 1.7265625,
+      "learning_rate": 1.9988893644781312e-05,
+      "loss": 1.0917,
+      "step": 361
+    },
+    {
+      "epoch": 0.06207000021433011,
+      "grad_norm": 1.65625,
+      "learning_rate": 1.9988808391480955e-05,
+      "loss": 1.1106,
+      "step": 362
+    },
+    {
+      "epoch": 0.06224146430331998,
+      "grad_norm": 1.7734375,
+      "learning_rate": 1.998872281240794e-05,
+      "loss": 1.0291,
+      "step": 363
+    },
+    {
+      "epoch": 0.06241292839230984,
+      "grad_norm": 1.6328125,
+      "learning_rate": 1.998863690756506e-05,
+      "loss": 1.1161,
+      "step": 364
+    },
+    {
+      "epoch": 0.0625843924812997,
+      "grad_norm": 1.671875,
+      "learning_rate": 1.998855067695512e-05,
+      "loss": 1.1159,
+      "step": 365
+    },
+    {
+      "epoch": 0.06275585657028956,
+      "grad_norm": 1.859375,
+      "learning_rate": 1.9988464120580925e-05,
+      "loss": 1.098,
+      "step": 366
+    },
+    {
+      "epoch": 0.06292732065927942,
+      "grad_norm": 1.7734375,
+      "learning_rate": 1.9988377238445308e-05,
+      "loss": 1.1378,
+      "step": 367
+    },
+    {
+      "epoch": 0.06309878474826928,
+      "grad_norm": 1.6953125,
+      "learning_rate": 1.9988290030551088e-05,
+      "loss": 1.0319,
+      "step": 368
+    },
+    {
+      "epoch": 0.06327024883725915,
+      "grad_norm": 1.734375,
+      "learning_rate": 1.9988202496901126e-05,
+      "loss": 1.0653,
+      "step": 369
+    },
+    {
+      "epoch": 0.063441712926249,
+      "grad_norm": 1.703125,
+      "learning_rate": 1.9988114637498264e-05,
+      "loss": 1.0302,
+      "step": 370
+    },
+    {
+      "epoch": 0.06361317701523887,
+      "grad_norm": 1.7421875,
+      "learning_rate": 1.9988026452345376e-05,
+      "loss": 1.0393,
+      "step": 371
+    },
+    {
+      "epoch": 0.06378464110422874,
+      "grad_norm": 1.7109375,
+      "learning_rate": 1.9987937941445327e-05,
+      "loss": 1.0318,
+      "step": 372
+    },
+    {
+      "epoch": 0.0639561051932186,
+      "grad_norm": 1.7265625,
+      "learning_rate": 1.9987849104801018e-05,
+      "loss": 1.1677,
+      "step": 373
+    },
+    {
+      "epoch": 0.06412756928220846,
+      "grad_norm": 1.765625,
+      "learning_rate": 1.9987759942415335e-05,
+      "loss": 1.1635,
+      "step": 374
+    },
+    {
+      "epoch": 0.06429903337119831,
+      "grad_norm": 1.7421875,
+      "learning_rate": 1.998767045429119e-05,
+      "loss": 1.1291,
+      "step": 375
+    },
+    {
+      "epoch": 0.06447049746018818,
+      "grad_norm": 1.78125,
+      "learning_rate": 1.99875806404315e-05,
+      "loss": 1.1369,
+      "step": 376
+    },
+    {
+      "epoch": 0.06464196154917805,
+      "grad_norm": 1.90625,
+      "learning_rate": 1.9987490500839198e-05,
+      "loss": 1.1804,
+      "step": 377
+    },
+    {
+      "epoch": 0.0648134256381679,
+      "grad_norm": 1.71875,
+      "learning_rate": 1.998740003551722e-05,
+      "loss": 1.1491,
+      "step": 378
+    },
+    {
+      "epoch": 0.06498488972715777,
+      "grad_norm": 1.7109375,
+      "learning_rate": 1.9987309244468517e-05,
+      "loss": 1.147,
+      "step": 379
+    },
+    {
+      "epoch": 0.06515635381614764,
+      "grad_norm": 1.8125,
+      "learning_rate": 1.9987218127696054e-05,
+      "loss": 1.0988,
+      "step": 380
+    },
+    {
+      "epoch": 0.06532781790513749,
+      "grad_norm": 2.078125,
+      "learning_rate": 1.9987126685202795e-05,
+      "loss": 1.0034,
+      "step": 381
+    },
+    {
+      "epoch": 0.06549928199412736,
+      "grad_norm": 1.7421875,
+      "learning_rate": 1.998703491699173e-05,
+      "loss": 1.1178,
+      "step": 382
+    },
+    {
+      "epoch": 0.06567074608311722,
+      "grad_norm": 1.671875,
+      "learning_rate": 1.998694282306585e-05,
+      "loss": 1.0409,
+      "step": 383
+    },
+    {
+      "epoch": 0.06584221017210708,
+      "grad_norm": 1.7421875,
+      "learning_rate": 1.9986850403428155e-05,
+      "loss": 1.1258,
+      "step": 384
+    },
+    {
+      "epoch": 0.06601367426109694,
+      "grad_norm": 1.75,
+      "learning_rate": 1.998675765808166e-05,
+      "loss": 1.1269,
+      "step": 385
+    },
+    {
+      "epoch": 0.0661851383500868,
+      "grad_norm": 1.8203125,
+      "learning_rate": 1.9986664587029395e-05,
+      "loss": 1.142,
+      "step": 386
+    },
+    {
+      "epoch": 0.06635660243907666,
+      "grad_norm": 1.609375,
+      "learning_rate": 1.9986571190274388e-05,
+      "loss": 1.1271,
+      "step": 387
+    },
+    {
+      "epoch": 0.06652806652806653,
+      "grad_norm": 1.78125,
+      "learning_rate": 1.9986477467819688e-05,
+      "loss": 1.0659,
+      "step": 388
+    },
+    {
+      "epoch": 0.06669953061705639,
+      "grad_norm": 1.7578125,
+      "learning_rate": 1.9986383419668355e-05,
+      "loss": 1.0759,
+      "step": 389
+    },
+    {
+      "epoch": 0.06687099470604625,
+      "grad_norm": 1.75,
+      "learning_rate": 1.998628904582345e-05,
+      "loss": 1.0522,
+      "step": 390
+    },
+    {
+      "epoch": 0.06704245879503612,
+      "grad_norm": 1.671875,
+      "learning_rate": 1.998619434628806e-05,
+      "loss": 1.1373,
+      "step": 391
+    },
+    {
+      "epoch": 0.06721392288402597,
+      "grad_norm": 1.6796875,
+      "learning_rate": 1.9986099321065266e-05,
+      "loss": 1.1222,
+      "step": 392
+    },
+    {
+      "epoch": 0.06738538697301584,
+      "grad_norm": 1.65625,
+      "learning_rate": 1.9986003970158164e-05,
+      "loss": 1.1167,
+      "step": 393
+    },
+    {
+      "epoch": 0.06755685106200571,
+      "grad_norm": 1.734375,
+      "learning_rate": 1.9985908293569873e-05,
+      "loss": 1.1531,
+      "step": 394
+    },
+    {
+      "epoch": 0.06772831515099556,
+      "grad_norm": 1.890625,
+      "learning_rate": 1.998581229130351e-05,
+      "loss": 1.1829,
+      "step": 395
+    },
+    {
+      "epoch": 0.06789977923998543,
+      "grad_norm": 1.765625,
+      "learning_rate": 1.99857159633622e-05,
+      "loss": 1.201,
+      "step": 396
+    },
+    {
+      "epoch": 0.06807124332897528,
+      "grad_norm": 1.7109375,
+      "learning_rate": 1.9985619309749096e-05,
+      "loss": 1.1298,
+      "step": 397
+    },
+    {
+      "epoch": 0.06824270741796515,
+      "grad_norm": 1.71875,
+      "learning_rate": 1.9985522330467343e-05,
+      "loss": 1.07,
+      "step": 398
+    },
+    {
+      "epoch": 0.06841417150695502,
+      "grad_norm": 1.6171875,
+      "learning_rate": 1.9985425025520098e-05,
+      "loss": 1.0821,
+      "step": 399
+    },
+    {
+      "epoch": 0.06858563559594487,
+      "grad_norm": 1.7578125,
+      "learning_rate": 1.998532739491055e-05,
+      "loss": 1.1798,
+      "step": 400
+    },
+    {
+      "epoch": 0.06875709968493474,
+      "grad_norm": 1.75,
+      "learning_rate": 1.998522943864187e-05,
+      "loss": 1.1651,
+      "step": 401
+    },
+    {
+      "epoch": 0.0689285637739246,
+      "grad_norm": 1.640625,
+      "learning_rate": 1.9985131156717258e-05,
+      "loss": 1.0264,
+      "step": 402
+    },
+    {
+      "epoch": 0.06910002786291446,
+      "grad_norm": 1.7109375,
+      "learning_rate": 1.9985032549139917e-05,
+      "loss": 1.153,
+      "step": 403
+    },
+    {
+      "epoch": 0.06927149195190432,
+      "grad_norm": 1.703125,
+      "learning_rate": 1.9984933615913068e-05,
+      "loss": 1.0631,
+      "step": 404
+    },
+    {
+      "epoch": 0.06944295604089419,
+      "grad_norm": 1.6796875,
+      "learning_rate": 1.9984834357039927e-05,
+      "loss": 1.1154,
+      "step": 405
+    },
+    {
+      "epoch": 0.06961442012988404,
+      "grad_norm": 1.75,
+      "learning_rate": 1.9984734772523747e-05,
+      "loss": 1.1756,
+      "step": 406
+    },
+    {
+      "epoch": 0.06978588421887391,
+      "grad_norm": 1.7890625,
+      "learning_rate": 1.998463486236776e-05,
+      "loss": 1.126,
+      "step": 407
+    },
+    {
+      "epoch": 0.06995734830786378,
+      "grad_norm": 1.6953125,
+      "learning_rate": 1.9984534626575236e-05,
+      "loss": 1.1288,
+      "step": 408
+    },
+    {
+      "epoch": 0.07012881239685363,
+      "grad_norm": 1.734375,
+      "learning_rate": 1.9984434065149435e-05,
+      "loss": 1.1455,
+      "step": 409
+    },
+    {
+      "epoch": 0.0703002764858435,
+      "grad_norm": 1.734375,
+      "learning_rate": 1.9984333178093646e-05,
+      "loss": 1.0953,
+      "step": 410
+    },
+    {
+      "epoch": 0.07047174057483335,
+      "grad_norm": 1.7890625,
+      "learning_rate": 1.9984231965411154e-05,
+      "loss": 1.1731,
+      "step": 411
+    },
+    {
+      "epoch": 0.07064320466382322,
+      "grad_norm": 1.828125,
+      "learning_rate": 1.9984130427105257e-05,
+      "loss": 1.0778,
+      "step": 412
+    },
+    {
+      "epoch": 0.07081466875281309,
+      "grad_norm": 1.7890625,
+      "learning_rate": 1.9984028563179274e-05,
+      "loss": 1.1916,
+      "step": 413
+    },
+    {
+      "epoch": 0.07098613284180294,
+      "grad_norm": 1.8359375,
+      "learning_rate": 1.998392637363652e-05,
+      "loss": 1.1299,
+      "step": 414
+    },
+    {
+      "epoch": 0.07115759693079281,
+      "grad_norm": 1.6953125,
+      "learning_rate": 1.9983823858480333e-05,
+      "loss": 1.0373,
+      "step": 415
+    },
+    {
+      "epoch": 0.07132906101978267,
+      "grad_norm": 1.90625,
+      "learning_rate": 1.9983721017714055e-05,
+      "loss": 1.1585,
+      "step": 416
+    },
+    {
+      "epoch": 0.07150052510877253,
+      "grad_norm": 1.71875,
+      "learning_rate": 1.9983617851341038e-05,
+      "loss": 1.1074,
+      "step": 417
+    },
+    {
+      "epoch": 0.0716719891977624,
+      "grad_norm": 1.9375,
+      "learning_rate": 1.998351435936465e-05,
+      "loss": 1.2491,
+      "step": 418
+    },
+    {
+      "epoch": 0.07184345328675226,
+      "grad_norm": 1.671875,
+      "learning_rate": 1.998341054178826e-05,
+      "loss": 1.006,
+      "step": 419
+    },
+    {
+      "epoch": 0.07201491737574212,
+      "grad_norm": 1.8046875,
+      "learning_rate": 1.998330639861526e-05,
+      "loss": 1.1695,
+      "step": 420
+    },
+    {
+      "epoch": 0.07218638146473198,
+      "grad_norm": 1.6953125,
+      "learning_rate": 1.9983201929849044e-05,
+      "loss": 1.1085,
+      "step": 421
+    },
+    {
+      "epoch": 0.07235784555372184,
+      "grad_norm": 1.671875,
+      "learning_rate": 1.9983097135493024e-05,
+      "loss": 1.1592,
+      "step": 422
+    },
+    {
+      "epoch": 0.0725293096427117,
+      "grad_norm": 1.78125,
+      "learning_rate": 1.998299201555061e-05,
+      "loss": 1.0885,
+      "step": 423
+    },
+    {
+      "epoch": 0.07270077373170157,
+      "grad_norm": 1.6953125,
+      "learning_rate": 1.998288657002523e-05,
+      "loss": 1.078,
+      "step": 424
+    },
+    {
+      "epoch": 0.07287223782069142,
+      "grad_norm": 1.75,
+      "learning_rate": 1.9982780798920337e-05,
+      "loss": 1.0714,
+      "step": 425
+    },
+    {
+      "epoch": 0.07304370190968129,
+      "grad_norm": 1.71875,
+      "learning_rate": 1.9982674702239363e-05,
+      "loss": 1.1147,
+      "step": 426
+    },
+    {
+      "epoch": 0.07321516599867116,
+      "grad_norm": 1.71875,
+      "learning_rate": 1.998256827998578e-05,
+      "loss": 1.1031,
+      "step": 427
+    },
+    {
+      "epoch": 0.07338663008766101,
+      "grad_norm": 1.765625,
+      "learning_rate": 1.9982461532163052e-05,
+      "loss": 1.0766,
+      "step": 428
+    },
+    {
+      "epoch": 0.07355809417665088,
+      "grad_norm": 1.90625,
+      "learning_rate": 1.9982354458774658e-05,
+      "loss": 1.0823,
+      "step": 429
+    },
+    {
+      "epoch": 0.07372955826564075,
+      "grad_norm": 1.7109375,
+      "learning_rate": 1.9982247059824103e-05,
+      "loss": 1.0892,
+      "step": 430
+    },
+    {
+      "epoch": 0.0739010223546306,
+      "grad_norm": 1.7578125,
+      "learning_rate": 1.9982139335314878e-05,
+      "loss": 1.081,
+      "step": 431
+    },
+    {
+      "epoch": 0.07407248644362047,
+      "grad_norm": 1.65625,
+      "learning_rate": 1.99820312852505e-05,
+      "loss": 1.0797,
+      "step": 432
+    },
+    {
+      "epoch": 0.07424395053261032,
+      "grad_norm": 1.9921875,
+      "learning_rate": 1.9981922909634495e-05,
+      "loss": 1.1905,
+      "step": 433
+    },
+    {
+      "epoch": 0.07441541462160019,
+      "grad_norm": 1.6796875,
+      "learning_rate": 1.9981814208470394e-05,
+      "loss": 1.0763,
+      "step": 434
+    },
+    {
+      "epoch": 0.07458687871059005,
+      "grad_norm": 1.7109375,
+      "learning_rate": 1.9981705181761747e-05,
+      "loss": 1.1211,
+      "step": 435
+    },
+    {
+      "epoch": 0.07475834279957991,
+      "grad_norm": 1.71875,
+      "learning_rate": 1.99815958295121e-05,
+      "loss": 1.1659,
+      "step": 436
+    },
+    {
+      "epoch": 0.07492980688856977,
+      "grad_norm": 1.7265625,
+      "learning_rate": 1.9981486151725027e-05,
+      "loss": 1.075,
+      "step": 437
+    },
+    {
+      "epoch": 0.07510127097755964,
+      "grad_norm": 1.6796875,
+      "learning_rate": 1.9981376148404112e-05,
+      "loss": 1.0265,
+      "step": 438
+    },
+    {
+      "epoch": 0.0752727350665495,
+      "grad_norm": 1.796875,
+      "learning_rate": 1.9981265819552927e-05,
+      "loss": 1.1912,
+      "step": 439
+    },
+    {
+      "epoch": 0.07544419915553936,
+      "grad_norm": 1.7578125,
+      "learning_rate": 1.998115516517508e-05,
+      "loss": 1.1765,
+      "step": 440
+    },
+    {
+      "epoch": 0.07561566324452923,
+      "grad_norm": 1.734375,
+      "learning_rate": 1.9981044185274174e-05,
+      "loss": 1.0593,
+      "step": 441
+    },
+    {
+      "epoch": 0.07578712733351908,
+      "grad_norm": 1.796875,
+      "learning_rate": 1.9980932879853835e-05,
+      "loss": 0.98,
+      "step": 442
+    },
+    {
+      "epoch": 0.07595859142250895,
+      "grad_norm": 1.75,
+      "learning_rate": 1.998082124891769e-05,
+      "loss": 1.1933,
+      "step": 443
+    },
+    {
+      "epoch": 0.0761300555114988,
+      "grad_norm": 1.828125,
+      "learning_rate": 1.998070929246938e-05,
+      "loss": 1.1674,
+      "step": 444
+    },
+    {
+      "epoch": 0.07630151960048867,
+      "grad_norm": 1.734375,
+      "learning_rate": 1.998059701051255e-05,
+      "loss": 1.1136,
+      "step": 445
+    },
+    {
+      "epoch": 0.07647298368947854,
+      "grad_norm": 1.8046875,
+      "learning_rate": 1.9980484403050876e-05,
+      "loss": 1.1286,
+      "step": 446
+    },
+    {
+      "epoch": 0.07664444777846839,
+      "grad_norm": 1.8515625,
+      "learning_rate": 1.998037147008802e-05,
+      "loss": 1.1737,
+      "step": 447
+    },
+    {
+      "epoch": 0.07681591186745826,
+      "grad_norm": 1.7421875,
+      "learning_rate": 1.998025821162767e-05,
+      "loss": 1.1155,
+      "step": 448
+    },
+    {
+      "epoch": 0.07698737595644813,
+      "grad_norm": 1.6171875,
+      "learning_rate": 1.9980144627673514e-05,
+      "loss": 1.1302,
+      "step": 449
+    },
+    {
+      "epoch": 0.07715884004543798,
+      "grad_norm": 1.828125,
+      "learning_rate": 1.9980030718229262e-05,
+      "loss": 1.1465,
+      "step": 450
+    },
+    {
+      "epoch": 0.07733030413442785,
+      "grad_norm": 1.640625,
+      "learning_rate": 1.9979916483298625e-05,
+      "loss": 1.0567,
+      "step": 451
+    },
+    {
+      "epoch": 0.07750176822341771,
+      "grad_norm": 3.21875,
+      "learning_rate": 1.997980192288533e-05,
+      "loss": 1.1799,
+      "step": 452
+    },
+    {
+      "epoch": 0.07767323231240757,
+      "grad_norm": 1.75,
+      "learning_rate": 1.9979687036993116e-05,
+      "loss": 1.0321,
+      "step": 453
+    },
+    {
+      "epoch": 0.07784469640139743,
+      "grad_norm": 1.6484375,
+      "learning_rate": 1.9979571825625726e-05,
+      "loss": 1.0242,
+      "step": 454
+    },
+    {
+      "epoch": 0.0780161604903873,
+      "grad_norm": 1.6640625,
+      "learning_rate": 1.9979456288786926e-05,
+      "loss": 1.0746,
+      "step": 455
+    },
+    {
+      "epoch": 0.07818762457937715,
+      "grad_norm": 1.7265625,
+      "learning_rate": 1.997934042648047e-05,
+      "loss": 1.0803,
+      "step": 456
+    },
+    {
+      "epoch": 0.07835908866836702,
+      "grad_norm": 2.234375,
+      "learning_rate": 1.9979224238710143e-05,
+      "loss": 1.094,
+      "step": 457
+    },
+    {
+      "epoch": 0.07853055275735688,
+      "grad_norm": 1.953125,
+      "learning_rate": 1.997910772547974e-05,
+      "loss": 1.1616,
+      "step": 458
+    },
+    {
+      "epoch": 0.07870201684634674,
+      "grad_norm": 6.75,
+      "learning_rate": 1.9978990886793056e-05,
+      "loss": 1.0693,
+      "step": 459
+    },
+    {
+      "epoch": 0.07887348093533661,
+      "grad_norm": 5.75,
+      "learning_rate": 1.9978873722653896e-05,
+      "loss": 1.1452,
+      "step": 460
+    },
+    {
+      "epoch": 0.07904494502432646,
+      "grad_norm": 3.59375,
+      "learning_rate": 1.997875623306609e-05,
+      "loss": 1.1542,
+      "step": 461
+    },
+    {
+      "epoch": 0.07921640911331633,
+      "grad_norm": 1.7265625,
+      "learning_rate": 1.9978638418033473e-05,
+      "loss": 1.1215,
+      "step": 462
+    },
+    {
+      "epoch": 0.0793878732023062,
+      "grad_norm": 1.7109375,
+      "learning_rate": 1.9978520277559873e-05,
+      "loss": 1.0942,
+      "step": 463
+    },
+    {
+      "epoch": 0.07955933729129605,
+      "grad_norm": 1.7265625,
+      "learning_rate": 1.9978401811649157e-05,
+      "loss": 1.0332,
+      "step": 464
+    },
+    {
+      "epoch": 0.07973080138028592,
+      "grad_norm": 1.8828125,
+      "learning_rate": 1.9978283020305177e-05,
+      "loss": 1.1336,
+      "step": 465
+    },
+    {
+      "epoch": 0.07990226546927578,
+      "grad_norm": 1.8359375,
+      "learning_rate": 1.9978163903531817e-05,
+      "loss": 1.0954,
+      "step": 466
+    },
+    {
+      "epoch": 0.08007372955826564,
+      "grad_norm": 1.6796875,
+      "learning_rate": 1.997804446133296e-05,
+      "loss": 1.0685,
+      "step": 467
+    },
+    {
+      "epoch": 0.0802451936472555,
+      "grad_norm": 1.59375,
+      "learning_rate": 1.9977924693712492e-05,
+      "loss": 1.1116,
+      "step": 468
+    },
+    {
+      "epoch": 0.08041665773624536,
+      "grad_norm": 1.7265625,
+      "learning_rate": 1.9977804600674327e-05,
+      "loss": 1.082,
+      "step": 469
+    },
+    {
+      "epoch": 0.08058812182523523,
+      "grad_norm": 1.6796875,
+      "learning_rate": 1.9977684182222387e-05,
+      "loss": 1.035,
+      "step": 470
+    },
+    {
+      "epoch": 0.0807595859142251,
+      "grad_norm": 1.734375,
+      "learning_rate": 1.9977563438360593e-05,
+      "loss": 1.1146,
+      "step": 471
+    },
+    {
+      "epoch": 0.08093105000321495,
+      "grad_norm": 1.78125,
+      "learning_rate": 1.997744236909288e-05,
+      "loss": 1.04,
+      "step": 472
+    },
+    {
+      "epoch": 0.08110251409220481,
+      "grad_norm": 1.7734375,
+      "learning_rate": 1.9977320974423196e-05,
+      "loss": 1.1159,
+      "step": 473
+    },
+    {
+      "epoch": 0.08127397818119468,
+      "grad_norm": 1.7734375,
+      "learning_rate": 1.997719925435551e-05,
+      "loss": 1.1449,
+      "step": 474
+    },
+    {
+      "epoch": 0.08144544227018453,
+      "grad_norm": 1.78125,
+      "learning_rate": 1.9977077208893788e-05,
+      "loss": 1.1145,
+      "step": 475
+    },
+    {
+      "epoch": 0.0816169063591744,
+      "grad_norm": 1.765625,
+      "learning_rate": 1.9976954838042e-05,
+      "loss": 1.1232,
+      "step": 476
+    },
+    {
+      "epoch": 0.08178837044816427,
+      "grad_norm": 1.8515625,
+      "learning_rate": 1.9976832141804153e-05,
+      "loss": 1.2183,
+      "step": 477
+    },
+    {
+      "epoch": 0.08195983453715412,
+      "grad_norm": 1.65625,
+      "learning_rate": 1.9976709120184234e-05,
+      "loss": 1.0607,
+      "step": 478
+    },
+    {
+      "epoch": 0.08213129862614399,
+      "grad_norm": 1.8046875,
+      "learning_rate": 1.9976585773186267e-05,
+      "loss": 1.1443,
+      "step": 479
+    },
+    {
+      "epoch": 0.08230276271513384,
+      "grad_norm": 1.8359375,
+      "learning_rate": 1.9976462100814268e-05,
+      "loss": 1.0947,
+      "step": 480
+    },
+    {
+      "epoch": 0.08247422680412371,
+      "grad_norm": 1.8046875,
+      "learning_rate": 1.997633810307227e-05,
+      "loss": 1.065,
+      "step": 481
+    },
+    {
+      "epoch": 0.08264569089311358,
+      "grad_norm": 1.7578125,
+      "learning_rate": 1.997621377996432e-05,
+      "loss": 1.0127,
+      "step": 482
+    },
+    {
+      "epoch": 0.08281715498210343,
+      "grad_norm": 1.953125,
+      "learning_rate": 1.997608913149447e-05,
+      "loss": 1.1413,
+      "step": 483
+    },
+    {
+      "epoch": 0.0829886190710933,
+      "grad_norm": 1.640625,
+      "learning_rate": 1.997596415766679e-05,
+      "loss": 1.1008,
+      "step": 484
+    },
+    {
+      "epoch": 0.08316008316008316,
+      "grad_norm": 1.6875,
+      "learning_rate": 1.997583885848535e-05,
+      "loss": 1.1314,
+      "step": 485
+    },
+    {
+      "epoch": 0.08333154724907302,
+      "grad_norm": 1.7265625,
+      "learning_rate": 1.9975713233954242e-05,
+      "loss": 1.1378,
+      "step": 486
+    },
+    {
+      "epoch": 0.08350301133806289,
+      "grad_norm": 1.890625,
+      "learning_rate": 1.9975587284077558e-05,
+      "loss": 1.1202,
+      "step": 487
+    },
+    {
+      "epoch": 0.08367447542705275,
+      "grad_norm": 1.765625,
+      "learning_rate": 1.9975461008859408e-05,
+      "loss": 1.1505,
+      "step": 488
+    },
+    {
+      "epoch": 0.0838459395160426,
+      "grad_norm": 1.6171875,
+      "learning_rate": 1.9975334408303916e-05,
+      "loss": 1.072,
+      "step": 489
+    },
+    {
+      "epoch": 0.08401740360503247,
+      "grad_norm": 1.8515625,
+      "learning_rate": 1.9975207482415198e-05,
+      "loss": 1.1842,
+      "step": 490
+    },
+    {
+      "epoch": 0.08418886769402233,
+      "grad_norm": 1.7109375,
+      "learning_rate": 1.9975080231197406e-05,
+      "loss": 1.1735,
+      "step": 491
+    },
+    {
+      "epoch": 0.0843603317830122,
+      "grad_norm": 1.6953125,
+      "learning_rate": 1.997495265465468e-05,
+      "loss": 1.0909,
+      "step": 492
+    },
+    {
+      "epoch": 0.08453179587200206,
+      "grad_norm": 1.7421875,
+      "learning_rate": 1.997482475279119e-05,
+      "loss": 1.0719,
+      "step": 493
+    },
+    {
+      "epoch": 0.08470325996099191,
+      "grad_norm": 1.8046875,
+      "learning_rate": 1.9974696525611102e-05,
+      "loss": 1.0718,
+      "step": 494
+    },
+    {
+      "epoch": 0.08487472404998178,
+      "grad_norm": 1.6796875,
+      "learning_rate": 1.99745679731186e-05,
+      "loss": 1.1493,
+      "step": 495
+    },
+    {
+      "epoch": 0.08504618813897165,
+      "grad_norm": 2.515625,
+      "learning_rate": 1.9974439095317874e-05,
+      "loss": 1.2006,
+      "step": 496
+    },
+    {
+      "epoch": 0.0852176522279615,
+      "grad_norm": 1.7109375,
+      "learning_rate": 1.997430989221313e-05,
+      "loss": 1.1398,
+      "step": 497
+    },
+    {
+      "epoch": 0.08538911631695137,
+      "grad_norm": 1.9140625,
+      "learning_rate": 1.9974180363808577e-05,
+      "loss": 1.1164,
+      "step": 498
+    },
+    {
+      "epoch": 0.08556058040594124,
+      "grad_norm": 1.6640625,
+      "learning_rate": 1.9974050510108447e-05,
+      "loss": 1.0891,
+      "step": 499
+    },
+    {
+      "epoch": 0.08573204449493109,
+      "grad_norm": 1.765625,
+      "learning_rate": 1.9973920331116973e-05,
+      "loss": 1.1879,
+      "step": 500
+    },
+    {
+      "epoch": 0.08590350858392096,
+      "grad_norm": 1.796875,
+      "learning_rate": 1.9973789826838393e-05,
+      "loss": 1.1017,
+      "step": 501
+    },
+    {
+      "epoch": 0.08607497267291082,
+      "grad_norm": 1.75,
+      "learning_rate": 1.997365899727697e-05,
+      "loss": 1.1061,
+      "step": 502
+    },
+    {
+      "epoch": 0.08624643676190068,
+      "grad_norm": 1.6796875,
+      "learning_rate": 1.9973527842436975e-05,
+      "loss": 1.1116,
+      "step": 503
+    },
+    {
+      "epoch": 0.08641790085089054,
+      "grad_norm": 1.9375,
+      "learning_rate": 1.9973396362322677e-05,
+      "loss": 1.1648,
+      "step": 504
+    },
+    {
+      "epoch": 0.0865893649398804,
+      "grad_norm": 1.7890625,
+      "learning_rate": 1.9973264556938365e-05,
+      "loss": 1.1449,
+      "step": 505
+    },
+    {
+      "epoch": 0.08676082902887026,
+      "grad_norm": 1.796875,
+      "learning_rate": 1.9973132426288343e-05,
+      "loss": 1.1397,
+      "step": 506
+    },
+    {
+      "epoch": 0.08693229311786013,
+      "grad_norm": 1.7265625,
+      "learning_rate": 1.9972999970376917e-05,
+      "loss": 1.0692,
+      "step": 507
+    },
+    {
+      "epoch": 0.08710375720684999,
+      "grad_norm": 1.8125,
+      "learning_rate": 1.9972867189208403e-05,
+      "loss": 1.0507,
+      "step": 508
+    },
+    {
+      "epoch": 0.08727522129583985,
+      "grad_norm": 1.8359375,
+      "learning_rate": 1.997273408278714e-05,
+      "loss": 1.1395,
+      "step": 509
+    },
+    {
+      "epoch": 0.08744668538482972,
+      "grad_norm": 1.734375,
+      "learning_rate": 1.9972600651117465e-05,
+      "loss": 1.0305,
+      "step": 510
+    },
+    {
+      "epoch": 0.08761814947381957,
+      "grad_norm": 1.7578125,
+      "learning_rate": 1.9972466894203727e-05,
+      "loss": 1.1262,
+      "step": 511
+    },
+    {
+      "epoch": 0.08778961356280944,
+      "grad_norm": 1.796875,
+      "learning_rate": 1.9972332812050293e-05,
+      "loss": 1.1948,
+      "step": 512
+    },
+    {
+      "epoch": 0.08796107765179931,
+      "grad_norm": 1.6953125,
+      "learning_rate": 1.997219840466153e-05,
+      "loss": 0.9924,
+      "step": 513
+    },
+    {
+      "epoch": 0.08813254174078916,
+      "grad_norm": 1.78125,
+      "learning_rate": 1.997206367204183e-05,
+      "loss": 0.992,
+      "step": 514
+    },
+    {
+      "epoch": 0.08830400582977903,
+      "grad_norm": 1.71875,
+      "learning_rate": 1.9971928614195578e-05,
+      "loss": 1.1893,
+      "step": 515
+    },
+    {
+      "epoch": 0.08847546991876888,
+      "grad_norm": 1.6796875,
+      "learning_rate": 1.9971793231127185e-05,
+      "loss": 1.1084,
+      "step": 516
+    },
+    {
+      "epoch": 0.08864693400775875,
+      "grad_norm": 1.8125,
+      "learning_rate": 1.9971657522841064e-05,
+      "loss": 1.0878,
+      "step": 517
+    },
+    {
+      "epoch": 0.08881839809674862,
+      "grad_norm": 1.671875,
+      "learning_rate": 1.9971521489341644e-05,
+      "loss": 1.0084,
+      "step": 518
+    },
+    {
+      "epoch": 0.08898986218573847,
+      "grad_norm": 1.6953125,
+      "learning_rate": 1.9971385130633357e-05,
+      "loss": 1.0803,
+      "step": 519
+    },
+    {
+      "epoch": 0.08916132627472834,
+      "grad_norm": 1.609375,
+      "learning_rate": 1.9971248446720654e-05,
+      "loss": 1.1219,
+      "step": 520
+    },
+    {
+      "epoch": 0.0893327903637182,
+      "grad_norm": 1.6484375,
+      "learning_rate": 1.9971111437607988e-05,
+      "loss": 1.0364,
+      "step": 521
+    },
+    {
+      "epoch": 0.08950425445270806,
+      "grad_norm": 1.6484375,
+      "learning_rate": 1.9970974103299833e-05,
+      "loss": 1.1015,
+      "step": 522
+    },
+    {
+      "epoch": 0.08967571854169792,
+      "grad_norm": 1.7421875,
+      "learning_rate": 1.997083644380066e-05,
+      "loss": 1.1518,
+      "step": 523
+    },
+    {
+      "epoch": 0.08984718263068779,
+      "grad_norm": 1.734375,
+      "learning_rate": 1.9970698459114967e-05,
+      "loss": 1.1541,
+      "step": 524
+    },
+    {
+      "epoch": 0.09001864671967764,
+      "grad_norm": 1.671875,
+      "learning_rate": 1.9970560149247255e-05,
+      "loss": 0.9705,
+      "step": 525
+    },
+    {
+      "epoch": 0.09019011080866751,
+      "grad_norm": 1.703125,
+      "learning_rate": 1.9970421514202025e-05,
+      "loss": 1.1515,
+      "step": 526
+    },
+    {
+      "epoch": 0.09036157489765737,
+      "grad_norm": 1.71875,
+      "learning_rate": 1.9970282553983805e-05,
+      "loss": 1.1664,
+      "step": 527
+    },
+    {
+      "epoch": 0.09053303898664723,
+      "grad_norm": 1.671875,
+      "learning_rate": 1.9970143268597127e-05,
+      "loss": 1.08,
+      "step": 528
+    },
+    {
+      "epoch": 0.0907045030756371,
+      "grad_norm": 1.65625,
+      "learning_rate": 1.9970003658046536e-05,
+      "loss": 1.129,
+      "step": 529
+    },
+    {
+      "epoch": 0.09087596716462695,
+      "grad_norm": 1.6328125,
+      "learning_rate": 1.996986372233658e-05,
+      "loss": 1.0173,
+      "step": 530
+    },
+    {
+      "epoch": 0.09104743125361682,
+      "grad_norm": 1.59375,
+      "learning_rate": 1.9969723461471826e-05,
+      "loss": 1.0457,
+      "step": 531
+    },
+    {
+      "epoch": 0.09121889534260669,
+      "grad_norm": 1.6796875,
+      "learning_rate": 1.9969582875456846e-05,
+      "loss": 1.1081,
+      "step": 532
+    },
+    {
+      "epoch": 0.09139035943159654,
+      "grad_norm": 1.734375,
+      "learning_rate": 1.9969441964296227e-05,
+      "loss": 1.0597,
+      "step": 533
+    },
+    {
+      "epoch": 0.09156182352058641,
+      "grad_norm": 1.640625,
+      "learning_rate": 1.9969300727994564e-05,
+      "loss": 1.0248,
+      "step": 534
+    },
+    {
+      "epoch": 0.09173328760957628,
+      "grad_norm": 1.7734375,
+      "learning_rate": 1.9969159166556462e-05,
+      "loss": 1.1426,
+      "step": 535
+    },
+    {
+      "epoch": 0.09190475169856613,
+      "grad_norm": 1.7265625,
+      "learning_rate": 1.9969017279986542e-05,
+      "loss": 1.1681,
+      "step": 536
+    },
+    {
+      "epoch": 0.092076215787556,
+      "grad_norm": 1.8203125,
+      "learning_rate": 1.9968875068289425e-05,
+      "loss": 1.09,
+      "step": 537
+    },
+    {
+      "epoch": 0.09224767987654586,
+      "grad_norm": 1.828125,
+      "learning_rate": 1.9968732531469757e-05,
+      "loss": 1.1262,
+      "step": 538
+    },
+    {
+      "epoch": 0.09241914396553572,
+      "grad_norm": 1.65625,
+      "learning_rate": 1.9968589669532183e-05,
+      "loss": 1.0702,
+      "step": 539
+    },
+    {
+      "epoch": 0.09259060805452558,
+      "grad_norm": 1.6875,
+      "learning_rate": 1.9968446482481357e-05,
+      "loss": 1.108,
+      "step": 540
+    },
+    {
+      "epoch": 0.09276207214351544,
+      "grad_norm": 1.7265625,
+      "learning_rate": 1.9968302970321956e-05,
+      "loss": 1.1241,
+      "step": 541
+    },
+    {
+      "epoch": 0.0929335362325053,
+      "grad_norm": 1.7734375,
+      "learning_rate": 1.9968159133058655e-05,
+      "loss": 1.1355,
+      "step": 542
+    },
+    {
+      "epoch": 0.09310500032149517,
+      "grad_norm": 1.7265625,
+      "learning_rate": 1.996801497069615e-05,
+      "loss": 1.1341,
+      "step": 543
+    },
+    {
+      "epoch": 0.09327646441048502,
+      "grad_norm": 1.78125,
+      "learning_rate": 1.996787048323914e-05,
+      "loss": 1.1449,
+      "step": 544
+    },
+    {
+      "epoch": 0.09344792849947489,
+      "grad_norm": 1.671875,
+      "learning_rate": 1.9967725670692343e-05,
+      "loss": 1.084,
+      "step": 545
+    },
+    {
+      "epoch": 0.09361939258846476,
+      "grad_norm": 1.890625,
+      "learning_rate": 1.996758053306047e-05,
+      "loss": 1.1289,
+      "step": 546
+    },
+    {
+      "epoch": 0.09379085667745461,
+      "grad_norm": 1.6171875,
+      "learning_rate": 1.9967435070348265e-05,
+      "loss": 1.0199,
+      "step": 547
+    },
+    {
+      "epoch": 0.09396232076644448,
+      "grad_norm": 1.7578125,
+      "learning_rate": 1.9967289282560468e-05,
+      "loss": 1.1458,
+      "step": 548
+    },
+    {
+      "epoch": 0.09413378485543435,
+      "grad_norm": 1.609375,
+      "learning_rate": 1.9967143169701836e-05,
+      "loss": 0.9749,
+      "step": 549
+    },
+    {
+      "epoch": 0.0943052489444242,
+      "grad_norm": 1.8125,
+      "learning_rate": 1.996699673177713e-05,
+      "loss": 1.0688,
+      "step": 550
+    },
+    {
+      "epoch": 0.09447671303341407,
+      "grad_norm": 1.703125,
+      "learning_rate": 1.996684996879113e-05,
+      "loss": 1.0636,
+      "step": 551
+    },
+    {
+      "epoch": 0.09464817712240392,
+      "grad_norm": 1.7109375,
+      "learning_rate": 1.9966702880748618e-05,
+      "loss": 1.1019,
+      "step": 552
+    },
+    {
+      "epoch": 0.09481964121139379,
+      "grad_norm": 1.6875,
+      "learning_rate": 1.9966555467654395e-05,
+      "loss": 1.1146,
+      "step": 553
+    },
+    {
+      "epoch": 0.09499110530038365,
+      "grad_norm": 1.890625,
+      "learning_rate": 1.996640772951327e-05,
+      "loss": 1.0833,
+      "step": 554
+    },
+    {
+      "epoch": 0.09516256938937351,
+      "grad_norm": 1.6640625,
+      "learning_rate": 1.9966259666330055e-05,
+      "loss": 1.0641,
+      "step": 555
+    },
+    {
+      "epoch": 0.09533403347836338,
+      "grad_norm": 1.65625,
+      "learning_rate": 1.9966111278109586e-05,
+      "loss": 1.016,
+      "step": 556
+    },
+    {
+      "epoch": 0.09550549756735324,
+      "grad_norm": 1.7890625,
+      "learning_rate": 1.99659625648567e-05,
+      "loss": 1.1166,
+      "step": 557
+    },
+    {
+      "epoch": 0.0956769616563431,
+      "grad_norm": 1.671875,
+      "learning_rate": 1.9965813526576246e-05,
+      "loss": 1.0452,
+      "step": 558
+    },
+    {
+      "epoch": 0.09584842574533296,
+      "grad_norm": 1.7734375,
+      "learning_rate": 1.996566416327308e-05,
+      "loss": 1.0721,
+      "step": 559
+    },
+    {
+      "epoch": 0.09601988983432283,
+      "grad_norm": 1.65625,
+      "learning_rate": 1.9965514474952084e-05,
+      "loss": 1.0086,
+      "step": 560
+    },
+    {
+      "epoch": 0.09619135392331268,
+      "grad_norm": 1.71875,
+      "learning_rate": 1.996536446161813e-05,
+      "loss": 1.0513,
+      "step": 561
+    },
+    {
+      "epoch": 0.09636281801230255,
+      "grad_norm": 1.828125,
+      "learning_rate": 1.996521412327612e-05,
+      "loss": 1.1228,
+      "step": 562
+    },
+    {
+      "epoch": 0.0965342821012924,
+      "grad_norm": 1.6796875,
+      "learning_rate": 1.996506345993095e-05,
+      "loss": 1.1084,
+      "step": 563
+    },
+    {
+      "epoch": 0.09670574619028227,
+      "grad_norm": 1.8984375,
+      "learning_rate": 1.9964912471587533e-05,
+      "loss": 1.1142,
+      "step": 564
+    },
+    {
+      "epoch": 0.09687721027927214,
+      "grad_norm": 1.6875,
+      "learning_rate": 1.99647611582508e-05,
+      "loss": 1.05,
+      "step": 565
+    },
+    {
+      "epoch": 0.09704867436826199,
+      "grad_norm": 1.8671875,
+      "learning_rate": 1.996460951992568e-05,
+      "loss": 1.1963,
+      "step": 566
+    },
+    {
+      "epoch": 0.09722013845725186,
+      "grad_norm": 1.6484375,
+      "learning_rate": 1.996445755661712e-05,
+      "loss": 1.0478,
+      "step": 567
+    },
+    {
+      "epoch": 0.09739160254624173,
+      "grad_norm": 1.75,
+      "learning_rate": 1.9964305268330075e-05,
+      "loss": 1.0692,
+      "step": 568
+    },
+    {
+      "epoch": 0.09756306663523158,
+      "grad_norm": 1.7421875,
+      "learning_rate": 1.9964152655069514e-05,
+      "loss": 1.0776,
+      "step": 569
+    },
+    {
+      "epoch": 0.09773453072422145,
+      "grad_norm": 1.8125,
+      "learning_rate": 1.9963999716840417e-05,
+      "loss": 1.0419,
+      "step": 570
+    },
+    {
+      "epoch": 0.09790599481321131,
+      "grad_norm": 1.6953125,
+      "learning_rate": 1.9963846453647763e-05,
+      "loss": 1.1413,
+      "step": 571
+    },
+    {
+      "epoch": 0.09807745890220117,
+      "grad_norm": 1.8203125,
+      "learning_rate": 1.996369286549656e-05,
+      "loss": 1.073,
+      "step": 572
+    },
+    {
+      "epoch": 0.09824892299119103,
+      "grad_norm": 1.7421875,
+      "learning_rate": 1.9963538952391808e-05,
+      "loss": 1.1669,
+      "step": 573
+    },
+    {
+      "epoch": 0.09842038708018089,
+      "grad_norm": 1.7578125,
+      "learning_rate": 1.9963384714338537e-05,
+      "loss": 1.1771,
+      "step": 574
+    },
+    {
+      "epoch": 0.09859185116917075,
+      "grad_norm": 1.859375,
+      "learning_rate": 1.996323015134177e-05,
+      "loss": 1.1514,
+      "step": 575
+    },
+    {
+      "epoch": 0.09876331525816062,
+      "grad_norm": 1.7734375,
+      "learning_rate": 1.9963075263406545e-05,
+      "loss": 1.163,
+      "step": 576
+    },
+    {
+      "epoch": 0.09893477934715048,
+      "grad_norm": 1.6875,
+      "learning_rate": 1.9962920050537922e-05,
+      "loss": 1.1313,
+      "step": 577
+    },
+    {
+      "epoch": 0.09910624343614034,
+      "grad_norm": 1.9453125,
+      "learning_rate": 1.996276451274096e-05,
+      "loss": 1.0454,
+      "step": 578
+    },
+    {
+      "epoch": 0.09927770752513021,
+      "grad_norm": 1.75,
+      "learning_rate": 1.996260865002073e-05,
+      "loss": 1.2105,
+      "step": 579
+    },
+    {
+      "epoch": 0.09944917161412006,
+      "grad_norm": 1.6484375,
+      "learning_rate": 1.996245246238232e-05,
+      "loss": 1.1686,
+      "step": 580
+    },
+    {
+      "epoch": 0.09962063570310993,
+      "grad_norm": 1.6953125,
+      "learning_rate": 1.9962295949830814e-05,
+      "loss": 1.1004,
+      "step": 581
+    },
+    {
+      "epoch": 0.0997920997920998,
+      "grad_norm": 1.765625,
+      "learning_rate": 1.9962139112371324e-05,
+      "loss": 1.2155,
+      "step": 582
+    },
+    {
+      "epoch": 0.09996356388108965,
+      "grad_norm": 1.703125,
+      "learning_rate": 1.9961981950008966e-05,
+      "loss": 1.1185,
+      "step": 583
+    },
+    {
+      "epoch": 0.10013502797007952,
+      "grad_norm": 1.65625,
+      "learning_rate": 1.9961824462748863e-05,
+      "loss": 1.0739,
+      "step": 584
+    },
+    {
+      "epoch": 0.10030649205906939,
+      "grad_norm": 1.65625,
+      "learning_rate": 1.996166665059615e-05,
+      "loss": 1.1047,
+      "step": 585
+    },
+    {
+      "epoch": 0.10047795614805924,
+      "grad_norm": 1.6484375,
+      "learning_rate": 1.9961508513555975e-05,
+      "loss": 1.1484,
+      "step": 586
+    },
+    {
+      "epoch": 0.1006494202370491,
+      "grad_norm": 1.671875,
+      "learning_rate": 1.9961350051633497e-05,
+      "loss": 1.0129,
+      "step": 587
+    },
+    {
+      "epoch": 0.10082088432603896,
+      "grad_norm": 1.734375,
+      "learning_rate": 1.9961191264833884e-05,
+      "loss": 1.0761,
+      "step": 588
+    },
+    {
+      "epoch": 0.10099234841502883,
+      "grad_norm": 1.859375,
+      "learning_rate": 1.9961032153162312e-05,
+      "loss": 1.1838,
+      "step": 589
+    },
+    {
+      "epoch": 0.1011638125040187,
+      "grad_norm": 1.6953125,
+      "learning_rate": 1.9960872716623972e-05,
+      "loss": 1.0953,
+      "step": 590
+    },
+    {
+      "epoch": 0.10133527659300855,
+      "grad_norm": 1.71875,
+      "learning_rate": 1.9960712955224065e-05,
+      "loss": 1.0843,
+      "step": 591
+    },
+    {
+      "epoch": 0.10150674068199841,
+      "grad_norm": 1.6953125,
+      "learning_rate": 1.9960552868967796e-05,
+      "loss": 1.2278,
+      "step": 592
+    },
+    {
+      "epoch": 0.10167820477098828,
+      "grad_norm": 1.6953125,
+      "learning_rate": 1.9960392457860394e-05,
+      "loss": 1.1326,
+      "step": 593
+    },
+    {
+      "epoch": 0.10184966885997813,
+      "grad_norm": 1.96875,
+      "learning_rate": 1.9960231721907083e-05,
+      "loss": 1.1644,
+      "step": 594
+    },
+    {
+      "epoch": 0.102021132948968,
+      "grad_norm": 1.6875,
+      "learning_rate": 1.996007066111311e-05,
+      "loss": 1.1278,
+      "step": 595
+    },
+    {
+      "epoch": 0.10219259703795787,
+      "grad_norm": 1.7890625,
+      "learning_rate": 1.995990927548373e-05,
+      "loss": 1.1009,
+      "step": 596
+    },
+    {
+      "epoch": 0.10236406112694772,
+      "grad_norm": 1.7421875,
+      "learning_rate": 1.99597475650242e-05,
+      "loss": 1.1035,
+      "step": 597
+    },
+    {
+      "epoch": 0.10253552521593759,
+      "grad_norm": 1.6953125,
+      "learning_rate": 1.9959585529739796e-05,
+      "loss": 1.1012,
+      "step": 598
+    },
+    {
+      "epoch": 0.10270698930492744,
+      "grad_norm": 1.65625,
+      "learning_rate": 1.9959423169635804e-05,
+      "loss": 1.1065,
+      "step": 599
+    },
+    {
+      "epoch": 0.10287845339391731,
+      "grad_norm": 1.6875,
+      "learning_rate": 1.995926048471752e-05,
+      "loss": 1.1969,
+      "step": 600
+    },
+    {
+      "epoch": 0.10304991748290718,
+      "grad_norm": 1.7734375,
+      "learning_rate": 1.9959097474990248e-05,
+      "loss": 1.163,
+      "step": 601
+    },
+    {
+      "epoch": 0.10322138157189703,
+      "grad_norm": 1.8046875,
+      "learning_rate": 1.9958934140459302e-05,
+      "loss": 1.1103,
+      "step": 602
+    },
+    {
+      "epoch": 0.1033928456608869,
+      "grad_norm": 1.6953125,
+      "learning_rate": 1.9958770481130016e-05,
+      "loss": 1.1603,
+      "step": 603
+    },
+    {
+      "epoch": 0.10356430974987677,
+      "grad_norm": 1.765625,
+      "learning_rate": 1.995860649700772e-05,
+      "loss": 1.1266,
+      "step": 604
+    },
+    {
+      "epoch": 0.10373577383886662,
+      "grad_norm": 1.8515625,
+      "learning_rate": 1.995844218809777e-05,
+      "loss": 1.1011,
+      "step": 605
+    },
+    {
+      "epoch": 0.10390723792785649,
+      "grad_norm": 1.78125,
+      "learning_rate": 1.9958277554405516e-05,
+      "loss": 1.0317,
+      "step": 606
+    },
+    {
+      "epoch": 0.10407870201684635,
+      "grad_norm": 1.8359375,
+      "learning_rate": 1.9958112595936333e-05,
+      "loss": 1.0968,
+      "step": 607
+    },
+    {
+      "epoch": 0.1042501661058362,
+      "grad_norm": 1.6640625,
+      "learning_rate": 1.9957947312695602e-05,
+      "loss": 1.0747,
+      "step": 608
+    },
+    {
+      "epoch": 0.10442163019482607,
+      "grad_norm": 1.7109375,
+      "learning_rate": 1.9957781704688707e-05,
+      "loss": 1.0624,
+      "step": 609
+    },
+    {
+      "epoch": 0.10459309428381593,
+      "grad_norm": 1.8125,
+      "learning_rate": 1.9957615771921055e-05,
+      "loss": 1.0525,
+      "step": 610
+    },
+    {
+      "epoch": 0.1047645583728058,
+      "grad_norm": 2.078125,
+      "learning_rate": 1.9957449514398054e-05,
+      "loss": 1.1324,
+      "step": 611
+    },
+    {
+      "epoch": 0.10493602246179566,
+      "grad_norm": 1.609375,
+      "learning_rate": 1.9957282932125132e-05,
+      "loss": 1.0306,
+      "step": 612
+    },
+    {
+      "epoch": 0.10510748655078551,
+      "grad_norm": 1.671875,
+      "learning_rate": 1.9957116025107713e-05,
+      "loss": 1.1239,
+      "step": 613
+    },
+    {
+      "epoch": 0.10527895063977538,
+      "grad_norm": 1.8828125,
+      "learning_rate": 1.9956948793351248e-05,
+      "loss": 1.0913,
+      "step": 614
+    },
+    {
+      "epoch": 0.10545041472876525,
+      "grad_norm": 1.7109375,
+      "learning_rate": 1.995678123686119e-05,
+      "loss": 1.0655,
+      "step": 615
+    },
+    {
+      "epoch": 0.1056218788177551,
+      "grad_norm": 1.6953125,
+      "learning_rate": 1.9956613355643e-05,
+      "loss": 1.0825,
+      "step": 616
+    },
+    {
+      "epoch": 0.10579334290674497,
+      "grad_norm": 1.90625,
+      "learning_rate": 1.9956445149702154e-05,
+      "loss": 1.1483,
+      "step": 617
+    },
+    {
+      "epoch": 0.10596480699573484,
+      "grad_norm": 2.34375,
+      "learning_rate": 1.9956276619044145e-05,
+      "loss": 1.1081,
+      "step": 618
+    },
+    {
+      "epoch": 0.10613627108472469,
+      "grad_norm": 1.734375,
+      "learning_rate": 1.9956107763674456e-05,
+      "loss": 1.1174,
+      "step": 619
+    },
+    {
+      "epoch": 0.10630773517371456,
+      "grad_norm": 1.6796875,
+      "learning_rate": 1.9955938583598606e-05,
+      "loss": 1.1238,
+      "step": 620
+    },
+    {
+      "epoch": 0.10647919926270442,
+      "grad_norm": 1.7890625,
+      "learning_rate": 1.9955769078822106e-05,
+      "loss": 1.1835,
+      "step": 621
+    },
+    {
+      "epoch": 0.10665066335169428,
+      "grad_norm": 1.8046875,
+      "learning_rate": 1.9955599249350487e-05,
+      "loss": 1.1702,
+      "step": 622
+    },
+    {
+      "epoch": 0.10682212744068414,
+      "grad_norm": 1.7421875,
+      "learning_rate": 1.995542909518929e-05,
+      "loss": 1.0848,
+      "step": 623
+    },
+    {
+      "epoch": 0.106993591529674,
+      "grad_norm": 1.7734375,
+      "learning_rate": 1.995525861634406e-05,
+      "loss": 1.106,
+      "step": 624
+    },
+    {
+      "epoch": 0.10716505561866387,
+      "grad_norm": 1.6875,
+      "learning_rate": 1.9955087812820357e-05,
+      "loss": 1.0085,
+      "step": 625
+    },
+    {
+      "epoch": 0.10733651970765373,
+      "grad_norm": 1.7578125,
+      "learning_rate": 1.9954916684623753e-05,
+      "loss": 1.083,
+      "step": 626
+    },
+    {
+      "epoch": 0.10750798379664359,
+      "grad_norm": 1.7578125,
+      "learning_rate": 1.995474523175983e-05,
+      "loss": 1.1413,
+      "step": 627
+    },
+    {
+      "epoch": 0.10767944788563345,
+      "grad_norm": 1.578125,
+      "learning_rate": 1.9954573454234177e-05,
+      "loss": 1.0141,
+      "step": 628
+    },
+    {
+      "epoch": 0.10785091197462332,
+      "grad_norm": 10.1875,
+      "learning_rate": 1.99544013520524e-05,
+      "loss": 1.1188,
+      "step": 629
+    },
+    {
+      "epoch": 0.10802237606361317,
+      "grad_norm": 1.8125,
+      "learning_rate": 1.9954228925220112e-05,
+      "loss": 1.0341,
+      "step": 630
+    },
+    {
+      "epoch": 0.10819384015260304,
+      "grad_norm": 1.78125,
+      "learning_rate": 1.9954056173742933e-05,
+      "loss": 1.1826,
+      "step": 631
+    },
+    {
+      "epoch": 0.10836530424159291,
+      "grad_norm": 1.609375,
+      "learning_rate": 1.99538830976265e-05,
+      "loss": 1.0859,
+      "step": 632
+    },
+    {
+      "epoch": 0.10853676833058276,
+      "grad_norm": 1.78125,
+      "learning_rate": 1.9953709696876452e-05,
+      "loss": 1.0958,
+      "step": 633
+    },
+    {
+      "epoch": 0.10870823241957263,
+      "grad_norm": 1.8125,
+      "learning_rate": 1.9953535971498452e-05,
+      "loss": 1.1274,
+      "step": 634
+    },
+    {
+      "epoch": 0.10887969650856248,
+      "grad_norm": 1.71875,
+      "learning_rate": 1.9953361921498163e-05,
+      "loss": 1.186,
+      "step": 635
+    },
+    {
+      "epoch": 0.10905116059755235,
+      "grad_norm": 1.734375,
+      "learning_rate": 1.9953187546881257e-05,
+      "loss": 1.0017,
+      "step": 636
+    },
+    {
+      "epoch": 0.10922262468654222,
+      "grad_norm": 1.71875,
+      "learning_rate": 1.995301284765343e-05,
+      "loss": 1.1105,
+      "step": 637
+    },
+    {
+      "epoch": 0.10939408877553207,
+      "grad_norm": 1.7109375,
+      "learning_rate": 1.9952837823820373e-05,
+      "loss": 1.0799,
+      "step": 638
+    },
+    {
+      "epoch": 0.10956555286452194,
+      "grad_norm": 1.734375,
+      "learning_rate": 1.9952662475387794e-05,
+      "loss": 1.0469,
+      "step": 639
+    },
+    {
+      "epoch": 0.1097370169535118,
+      "grad_norm": 1.625,
+      "learning_rate": 1.9952486802361416e-05,
+      "loss": 1.0985,
+      "step": 640
+    },
+    {
+      "epoch": 0.10990848104250166,
+      "grad_norm": 1.828125,
+      "learning_rate": 1.995231080474696e-05,
+      "loss": 1.1092,
+      "step": 641
+    },
+    {
+      "epoch": 0.11007994513149152,
+      "grad_norm": 1.640625,
+      "learning_rate": 1.9952134482550175e-05,
+      "loss": 1.0038,
+      "step": 642
+    },
+    {
+      "epoch": 0.11025140922048139,
+      "grad_norm": 1.703125,
+      "learning_rate": 1.9951957835776808e-05,
+      "loss": 1.1089,
+      "step": 643
+    },
+    {
+      "epoch": 0.11042287330947125,
+      "grad_norm": 1.6875,
+      "learning_rate": 1.9951780864432623e-05,
+      "loss": 1.0895,
+      "step": 644
+    },
+    {
+      "epoch": 0.11059433739846111,
+      "grad_norm": 1.65625,
+      "learning_rate": 1.9951603568523387e-05,
+      "loss": 1.0533,
+      "step": 645
+    },
+    {
+      "epoch": 0.11076580148745097,
+      "grad_norm": 1.8046875,
+      "learning_rate": 1.9951425948054886e-05,
+      "loss": 1.1723,
+      "step": 646
+    },
+    {
+      "epoch": 0.11093726557644083,
+      "grad_norm": 1.8203125,
+      "learning_rate": 1.995124800303291e-05,
+      "loss": 1.2228,
+      "step": 647
+    },
+    {
+      "epoch": 0.1111087296654307,
+      "grad_norm": 1.8359375,
+      "learning_rate": 1.9951069733463262e-05,
+      "loss": 1.1278,
+      "step": 648
+    },
+    {
+      "epoch": 0.11128019375442055,
+      "grad_norm": 1.8203125,
+      "learning_rate": 1.995089113935176e-05,
+      "loss": 1.1156,
+      "step": 649
+    },
+    {
+      "epoch": 0.11145165784341042,
+      "grad_norm": 1.7265625,
+      "learning_rate": 1.9950712220704224e-05,
+      "loss": 1.1134,
+      "step": 650
+    },
+    {
+      "epoch": 0.11162312193240029,
+      "grad_norm": 1.6953125,
+      "learning_rate": 1.9950532977526493e-05,
+      "loss": 1.1028,
+      "step": 651
+    },
+    {
+      "epoch": 0.11179458602139014,
+      "grad_norm": 1.828125,
+      "learning_rate": 1.9950353409824412e-05,
+      "loss": 1.0454,
+      "step": 652
+    },
+    {
+      "epoch": 0.11196605011038001,
+      "grad_norm": 1.859375,
+      "learning_rate": 1.995017351760384e-05,
+      "loss": 1.0748,
+      "step": 653
+    },
+    {
+      "epoch": 0.11213751419936988,
+      "grad_norm": 1.8359375,
+      "learning_rate": 1.9949993300870637e-05,
+      "loss": 1.1675,
+      "step": 654
+    },
+    {
+      "epoch": 0.11230897828835973,
+      "grad_norm": 1.7890625,
+      "learning_rate": 1.9949812759630683e-05,
+      "loss": 1.1367,
+      "step": 655
+    },
+    {
+      "epoch": 0.1124804423773496,
+      "grad_norm": 1.75,
+      "learning_rate": 1.9949631893889866e-05,
+      "loss": 1.0374,
+      "step": 656
+    },
+    {
+      "epoch": 0.11265190646633945,
+      "grad_norm": 1.8125,
+      "learning_rate": 1.994945070365409e-05,
+      "loss": 1.1405,
+      "step": 657
+    },
+    {
+      "epoch": 0.11282337055532932,
+      "grad_norm": 1.78125,
+      "learning_rate": 1.994926918892926e-05,
+      "loss": 1.1399,
+      "step": 658
+    },
+    {
+      "epoch": 0.11299483464431918,
+      "grad_norm": 2.0625,
+      "learning_rate": 1.9949087349721296e-05,
+      "loss": 1.1436,
+      "step": 659
+    },
+    {
+      "epoch": 0.11316629873330904,
+      "grad_norm": 1.7265625,
+      "learning_rate": 1.9948905186036128e-05,
+      "loss": 1.0613,
+      "step": 660
+    },
+    {
+      "epoch": 0.1133377628222989,
+      "grad_norm": 1.6484375,
+      "learning_rate": 1.9948722697879696e-05,
+      "loss": 1.1771,
+      "step": 661
+    },
+    {
+      "epoch": 0.11350922691128877,
+      "grad_norm": 1.8359375,
+      "learning_rate": 1.9948539885257956e-05,
+      "loss": 1.1123,
+      "step": 662
+    },
+    {
+      "epoch": 0.11368069100027862,
+      "grad_norm": 1.6953125,
+      "learning_rate": 1.9948356748176867e-05,
+      "loss": 1.0366,
+      "step": 663
+    },
+    {
+      "epoch": 0.11385215508926849,
+      "grad_norm": 1.6875,
+      "learning_rate": 1.9948173286642403e-05,
+      "loss": 1.0478,
+      "step": 664
+    },
+    {
+      "epoch": 0.11402361917825836,
+      "grad_norm": 1.734375,
+      "learning_rate": 1.9947989500660544e-05,
+      "loss": 1.0948,
+      "step": 665
+    },
+    {
+      "epoch": 0.11419508326724821,
+      "grad_norm": 1.6484375,
+      "learning_rate": 1.994780539023729e-05,
+      "loss": 1.0645,
+      "step": 666
+    },
+    {
+      "epoch": 0.11436654735623808,
+      "grad_norm": 1.7734375,
+      "learning_rate": 1.9947620955378642e-05,
+      "loss": 1.0534,
+      "step": 667
+    },
+    {
+      "epoch": 0.11453801144522795,
+      "grad_norm": 1.7421875,
+      "learning_rate": 1.9947436196090614e-05,
+      "loss": 1.077,
+      "step": 668
+    },
+    {
+      "epoch": 0.1147094755342178,
+      "grad_norm": 1.7890625,
+      "learning_rate": 1.9947251112379233e-05,
+      "loss": 1.0538,
+      "step": 669
+    },
+    {
+      "epoch": 0.11488093962320767,
+      "grad_norm": 1.6796875,
+      "learning_rate": 1.9947065704250533e-05,
+      "loss": 1.0747,
+      "step": 670
+    },
+    {
+      "epoch": 0.11505240371219752,
+      "grad_norm": 1.75,
+      "learning_rate": 1.994687997171057e-05,
+      "loss": 1.0738,
+      "step": 671
+    },
+    {
+      "epoch": 0.11522386780118739,
+      "grad_norm": 1.6953125,
+      "learning_rate": 1.994669391476539e-05,
+      "loss": 1.109,
+      "step": 672
+    },
+    {
+      "epoch": 0.11539533189017726,
+      "grad_norm": 1.5703125,
+      "learning_rate": 1.9946507533421067e-05,
+      "loss": 1.0449,
+      "step": 673
+    },
+    {
+      "epoch": 0.11556679597916711,
+      "grad_norm": 1.7109375,
+      "learning_rate": 1.9946320827683676e-05,
+      "loss": 1.046,
+      "step": 674
+    },
+    {
+      "epoch": 0.11573826006815698,
+      "grad_norm": 1.6796875,
+      "learning_rate": 1.994613379755931e-05,
+      "loss": 1.0857,
+      "step": 675
+    },
+    {
+      "epoch": 0.11590972415714684,
+      "grad_norm": 1.625,
+      "learning_rate": 1.994594644305407e-05,
+      "loss": 1.0279,
+      "step": 676
+    },
+    {
+      "epoch": 0.1160811882461367,
+      "grad_norm": 1.7421875,
+      "learning_rate": 1.9945758764174056e-05,
+      "loss": 1.1194,
+      "step": 677
+    },
+    {
+      "epoch": 0.11625265233512656,
+      "grad_norm": 1.7109375,
+      "learning_rate": 1.9945570760925403e-05,
+      "loss": 1.0465,
+      "step": 678
+    },
+    {
+      "epoch": 0.11642411642411643,
+      "grad_norm": 1.875,
+      "learning_rate": 1.9945382433314234e-05,
+      "loss": 1.042,
+      "step": 679
+    },
+    {
+      "epoch": 0.11659558051310628,
+      "grad_norm": 1.7734375,
+      "learning_rate": 1.9945193781346695e-05,
+      "loss": 1.0926,
+      "step": 680
+    },
+    {
+      "epoch": 0.11676704460209615,
+      "grad_norm": 1.7421875,
+      "learning_rate": 1.9945004805028933e-05,
+      "loss": 1.0987,
+      "step": 681
+    },
+    {
+      "epoch": 0.116938508691086,
+      "grad_norm": 1.7265625,
+      "learning_rate": 1.9944815504367117e-05,
+      "loss": 1.1226,
+      "step": 682
+    },
+    {
+      "epoch": 0.11710997278007587,
+      "grad_norm": 1.6953125,
+      "learning_rate": 1.9944625879367417e-05,
+      "loss": 1.0064,
+      "step": 683
+    },
+    {
+      "epoch": 0.11728143686906574,
+      "grad_norm": 1.7890625,
+      "learning_rate": 1.994443593003602e-05,
+      "loss": 1.1373,
+      "step": 684
+    },
+    {
+      "epoch": 0.11745290095805559,
+      "grad_norm": 1.734375,
+      "learning_rate": 1.9944245656379123e-05,
+      "loss": 1.1338,
+      "step": 685
+    },
+    {
+      "epoch": 0.11762436504704546,
+      "grad_norm": 1.7734375,
+      "learning_rate": 1.9944055058402923e-05,
+      "loss": 1.0695,
+      "step": 686
+    },
+    {
+      "epoch": 0.11779582913603533,
+      "grad_norm": 1.75,
+      "learning_rate": 1.9943864136113647e-05,
+      "loss": 1.1445,
+      "step": 687
+    },
+    {
+      "epoch": 0.11796729322502518,
+      "grad_norm": 1.765625,
+      "learning_rate": 1.9943672889517515e-05,
+      "loss": 1.1899,
+      "step": 688
+    },
+    {
+      "epoch": 0.11813875731401505,
+      "grad_norm": 1.71875,
+      "learning_rate": 1.9943481318620765e-05,
+      "loss": 1.0404,
+      "step": 689
+    },
+    {
+      "epoch": 0.11831022140300491,
+      "grad_norm": 1.8046875,
+      "learning_rate": 1.9943289423429645e-05,
+      "loss": 1.0936,
+      "step": 690
+    },
+    {
+      "epoch": 0.11848168549199477,
+      "grad_norm": 1.7578125,
+      "learning_rate": 1.9943097203950413e-05,
+      "loss": 1.0992,
+      "step": 691
+    },
+    {
+      "epoch": 0.11865314958098463,
+      "grad_norm": 1.7265625,
+      "learning_rate": 1.994290466018934e-05,
+      "loss": 1.0978,
+      "step": 692
+    },
+    {
+      "epoch": 0.11882461366997449,
+      "grad_norm": 1.7890625,
+      "learning_rate": 1.9942711792152708e-05,
+      "loss": 1.1341,
+      "step": 693
+    },
+    {
+      "epoch": 0.11899607775896436,
+      "grad_norm": 7.25,
+      "learning_rate": 1.99425185998468e-05,
+      "loss": 1.1517,
+      "step": 694
+    },
+    {
+      "epoch": 0.11916754184795422,
+      "grad_norm": 1.890625,
+      "learning_rate": 1.9942325083277917e-05,
+      "loss": 1.155,
+      "step": 695
+    },
+    {
+      "epoch": 0.11933900593694408,
+      "grad_norm": 1.6953125,
+      "learning_rate": 1.994213124245238e-05,
+      "loss": 1.0425,
+      "step": 696
+    },
+    {
+      "epoch": 0.11951047002593394,
+      "grad_norm": 1.7890625,
+      "learning_rate": 1.99419370773765e-05,
+      "loss": 1.0609,
+      "step": 697
+    },
+    {
+      "epoch": 0.11968193411492381,
+      "grad_norm": 1.6875,
+      "learning_rate": 1.9941742588056616e-05,
+      "loss": 1.0257,
+      "step": 698
+    },
+    {
+      "epoch": 0.11985339820391366,
+      "grad_norm": 1.7578125,
+      "learning_rate": 1.994154777449907e-05,
+      "loss": 1.1413,
+      "step": 699
+    },
+    {
+      "epoch": 0.12002486229290353,
+      "grad_norm": 1.7890625,
+      "learning_rate": 1.9941352636710215e-05,
+      "loss": 1.1103,
+      "step": 700
+    },
+    {
+      "epoch": 0.12002486229290353,
+      "eval_loss": 0.9447146058082581,
+      "eval_runtime": 837.1484,
+      "eval_samples_per_second": 2.985,
+      "eval_steps_per_second": 2.985,
+      "step": 700
+    },
+    {
+      "epoch": 0.1201963263818934,
+      "grad_norm": 1.6171875,
+      "learning_rate": 1.9941157174696412e-05,
+      "loss": 1.1096,
+      "step": 701
+    },
+    {
+      "epoch": 0.12036779047088325,
+      "grad_norm": 1.765625,
+      "learning_rate": 1.9940961388464042e-05,
+      "loss": 1.0937,
+      "step": 702
+    },
+    {
+      "epoch": 0.12053925455987312,
+      "grad_norm": 1.7109375,
+      "learning_rate": 1.9940765278019484e-05,
+      "loss": 1.0675,
+      "step": 703
+    },
+    {
+      "epoch": 0.12071071864886297,
+      "grad_norm": 1.75,
+      "learning_rate": 1.994056884336914e-05,
+      "loss": 1.0785,
+      "step": 704
+    },
+    {
+      "epoch": 0.12088218273785284,
+      "grad_norm": 1.7109375,
+      "learning_rate": 1.994037208451941e-05,
+      "loss": 1.1388,
+      "step": 705
+    },
+    {
+      "epoch": 0.1210536468268427,
+      "grad_norm": 1.7421875,
+      "learning_rate": 1.9940175001476715e-05,
+      "loss": 1.1303,
+      "step": 706
+    },
+    {
+      "epoch": 0.12122511091583256,
+      "grad_norm": 1.6484375,
+      "learning_rate": 1.9939977594247483e-05,
+      "loss": 1.0755,
+      "step": 707
+    },
+    {
+      "epoch": 0.12139657500482243,
+      "grad_norm": 1.6328125,
+      "learning_rate": 1.9939779862838153e-05,
+      "loss": 1.1185,
+      "step": 708
+    },
+    {
+      "epoch": 0.1215680390938123,
+      "grad_norm": 1.8828125,
+      "learning_rate": 1.9939581807255168e-05,
+      "loss": 1.1227,
+      "step": 709
+    },
+    {
+      "epoch": 0.12173950318280215,
+      "grad_norm": 1.828125,
+      "learning_rate": 1.9939383427504994e-05,
+      "loss": 1.0672,
+      "step": 710
+    },
+    {
+      "epoch": 0.12191096727179201,
+      "grad_norm": 1.6875,
+      "learning_rate": 1.9939184723594097e-05,
+      "loss": 1.076,
+      "step": 711
+    },
+    {
+      "epoch": 0.12208243136078188,
+      "grad_norm": 1.78125,
+      "learning_rate": 1.9938985695528957e-05,
+      "loss": 1.137,
+      "step": 712
+    },
+    {
+      "epoch": 0.12225389544977174,
+      "grad_norm": 1.78125,
+      "learning_rate": 1.993878634331607e-05,
+      "loss": 1.1944,
+      "step": 713
+    },
+    {
+      "epoch": 0.1224253595387616,
+      "grad_norm": 1.7578125,
+      "learning_rate": 1.9938586666961933e-05,
+      "loss": 1.1928,
+      "step": 714
+    },
+    {
+      "epoch": 0.12259682362775147,
+      "grad_norm": 1.6953125,
+      "learning_rate": 1.993838666647306e-05,
+      "loss": 1.0469,
+      "step": 715
+    },
+    {
+      "epoch": 0.12276828771674132,
+      "grad_norm": 1.734375,
+      "learning_rate": 1.9938186341855972e-05,
+      "loss": 1.0558,
+      "step": 716
+    },
+    {
+      "epoch": 0.12293975180573119,
+      "grad_norm": 1.7734375,
+      "learning_rate": 1.9937985693117205e-05,
+      "loss": 1.101,
+      "step": 717
+    },
+    {
+      "epoch": 0.12311121589472104,
+      "grad_norm": 1.921875,
+      "learning_rate": 1.99377847202633e-05,
+      "loss": 1.1684,
+      "step": 718
+    },
+    {
+      "epoch": 0.12328267998371091,
+      "grad_norm": 1.65625,
+      "learning_rate": 1.9937583423300816e-05,
+      "loss": 0.9685,
+      "step": 719
+    },
+    {
+      "epoch": 0.12345414407270078,
+      "grad_norm": 1.734375,
+      "learning_rate": 1.993738180223631e-05,
+      "loss": 1.1437,
+      "step": 720
+    },
+    {
+      "epoch": 0.12362560816169063,
+      "grad_norm": 1.71875,
+      "learning_rate": 1.9937179857076368e-05,
+      "loss": 1.0985,
+      "step": 721
+    },
+    {
+      "epoch": 0.1237970722506805,
+      "grad_norm": 1.6953125,
+      "learning_rate": 1.993697758782757e-05,
+      "loss": 1.028,
+      "step": 722
+    },
+    {
+      "epoch": 0.12396853633967037,
+      "grad_norm": 1.7265625,
+      "learning_rate": 1.9936774994496512e-05,
+      "loss": 1.1075,
+      "step": 723
+    },
+    {
+      "epoch": 0.12414000042866022,
+      "grad_norm": 1.828125,
+      "learning_rate": 1.99365720770898e-05,
+      "loss": 1.0201,
+      "step": 724
+    },
+    {
+      "epoch": 0.12431146451765009,
+      "grad_norm": 1.7578125,
+      "learning_rate": 1.993636883561406e-05,
+      "loss": 1.1393,
+      "step": 725
+    },
+    {
+      "epoch": 0.12448292860663995,
+      "grad_norm": 1.7265625,
+      "learning_rate": 1.9936165270075915e-05,
+      "loss": 1.0818,
+      "step": 726
+    },
+    {
+      "epoch": 0.1246543926956298,
+      "grad_norm": 1.7421875,
+      "learning_rate": 1.9935961380482003e-05,
+      "loss": 1.0204,
+      "step": 727
+    },
+    {
+      "epoch": 0.12482585678461967,
+      "grad_norm": 1.65625,
+      "learning_rate": 1.9935757166838974e-05,
+      "loss": 1.1329,
+      "step": 728
+    },
+    {
+      "epoch": 0.12499732087360953,
+      "grad_norm": 1.7421875,
+      "learning_rate": 1.993555262915349e-05,
+      "loss": 1.158,
+      "step": 729
+    },
+    {
+      "epoch": 0.1251687849625994,
+      "grad_norm": 1.671875,
+      "learning_rate": 1.9935347767432224e-05,
+      "loss": 1.151,
+      "step": 730
+    },
+    {
+      "epoch": 0.12534024905158925,
+      "grad_norm": 1.6328125,
+      "learning_rate": 1.993514258168185e-05,
+      "loss": 1.1295,
+      "step": 731
+    },
+    {
+      "epoch": 0.12551171314057913,
+      "grad_norm": 1.6328125,
+      "learning_rate": 1.9934937071909064e-05,
+      "loss": 1.1672,
+      "step": 732
+    },
+    {
+      "epoch": 0.12568317722956898,
+      "grad_norm": 1.7578125,
+      "learning_rate": 1.993473123812057e-05,
+      "loss": 1.0816,
+      "step": 733
+    },
+    {
+      "epoch": 0.12585464131855884,
+      "grad_norm": 1.71875,
+      "learning_rate": 1.993452508032308e-05,
+      "loss": 1.0609,
+      "step": 734
+    },
+    {
+      "epoch": 0.12602610540754872,
+      "grad_norm": 1.6015625,
+      "learning_rate": 1.9934318598523315e-05,
+      "loss": 1.0721,
+      "step": 735
+    },
+    {
+      "epoch": 0.12619756949653857,
+      "grad_norm": 1.7265625,
+      "learning_rate": 1.9934111792728015e-05,
+      "loss": 1.1286,
+      "step": 736
+    },
+    {
+      "epoch": 0.12636903358552842,
+      "grad_norm": 1.78125,
+      "learning_rate": 1.993390466294392e-05,
+      "loss": 1.134,
+      "step": 737
+    },
+    {
+      "epoch": 0.1265404976745183,
+      "grad_norm": 1.6875,
+      "learning_rate": 1.9933697209177785e-05,
+      "loss": 1.1024,
+      "step": 738
+    },
+    {
+      "epoch": 0.12671196176350816,
+      "grad_norm": 1.6171875,
+      "learning_rate": 1.9933489431436375e-05,
+      "loss": 1.0655,
+      "step": 739
+    },
+    {
+      "epoch": 0.126883425852498,
+      "grad_norm": 1.703125,
+      "learning_rate": 1.9933281329726473e-05,
+      "loss": 1.0384,
+      "step": 740
+    },
+    {
+      "epoch": 0.1270548899414879,
+      "grad_norm": 1.6328125,
+      "learning_rate": 1.993307290405486e-05,
+      "loss": 1.0531,
+      "step": 741
+    },
+    {
+      "epoch": 0.12722635403047775,
+      "grad_norm": 1.703125,
+      "learning_rate": 1.9932864154428335e-05,
+      "loss": 1.0878,
+      "step": 742
+    },
+    {
+      "epoch": 0.1273978181194676,
+      "grad_norm": 1.71875,
+      "learning_rate": 1.9932655080853706e-05,
+      "loss": 1.1547,
+      "step": 743
+    },
+    {
+      "epoch": 0.12756928220845748,
+      "grad_norm": 1.8046875,
+      "learning_rate": 1.9932445683337795e-05,
+      "loss": 1.1422,
+      "step": 744
+    },
+    {
+      "epoch": 0.12774074629744733,
+      "grad_norm": 1.765625,
+      "learning_rate": 1.9932235961887424e-05,
+      "loss": 1.1325,
+      "step": 745
+    },
+    {
+      "epoch": 0.1279122103864372,
+      "grad_norm": 1.78125,
+      "learning_rate": 1.9932025916509436e-05,
+      "loss": 1.122,
+      "step": 746
+    },
+    {
+      "epoch": 0.12808367447542704,
+      "grad_norm": 1.6640625,
+      "learning_rate": 1.9931815547210686e-05,
+      "loss": 1.1181,
+      "step": 747
+    },
+    {
+      "epoch": 0.12825513856441692,
+      "grad_norm": 1.640625,
+      "learning_rate": 1.993160485399803e-05,
+      "loss": 1.0483,
+      "step": 748
+    },
+    {
+      "epoch": 0.12842660265340677,
+      "grad_norm": 1.796875,
+      "learning_rate": 1.9931393836878338e-05,
+      "loss": 1.1126,
+      "step": 749
+    },
+    {
+      "epoch": 0.12859806674239663,
+      "grad_norm": 1.6484375,
+      "learning_rate": 1.99311824958585e-05,
+      "loss": 1.0719,
+      "step": 750
+    },
+    {
+      "epoch": 0.1287695308313865,
+      "grad_norm": 1.7578125,
+      "learning_rate": 1.99309708309454e-05,
+      "loss": 1.1756,
+      "step": 751
+    },
+    {
+      "epoch": 0.12894099492037636,
+      "grad_norm": 1.7265625,
+      "learning_rate": 1.9930758842145947e-05,
+      "loss": 1.1677,
+      "step": 752
+    },
+    {
+      "epoch": 0.12911245900936622,
+      "grad_norm": 1.6328125,
+      "learning_rate": 1.993054652946705e-05,
+      "loss": 1.1622,
+      "step": 753
+    },
+    {
+      "epoch": 0.1292839230983561,
+      "grad_norm": 1.71875,
+      "learning_rate": 1.9930333892915636e-05,
+      "loss": 1.0601,
+      "step": 754
+    },
+    {
+      "epoch": 0.12945538718734595,
+      "grad_norm": 1.5859375,
+      "learning_rate": 1.993012093249864e-05,
+      "loss": 1.022,
+      "step": 755
+    },
+    {
+      "epoch": 0.1296268512763358,
+      "grad_norm": 1.6796875,
+      "learning_rate": 1.9929907648223004e-05,
+      "loss": 1.122,
+      "step": 756
+    },
+    {
+      "epoch": 0.12979831536532568,
+      "grad_norm": 1.6171875,
+      "learning_rate": 1.9929694040095694e-05,
+      "loss": 1.0447,
+      "step": 757
+    },
+    {
+      "epoch": 0.12996977945431554,
+      "grad_norm": 1.6796875,
+      "learning_rate": 1.9929480108123666e-05,
+      "loss": 1.1163,
+      "step": 758
+    },
+    {
+      "epoch": 0.1301412435433054,
+      "grad_norm": 1.6875,
+      "learning_rate": 1.99292658523139e-05,
+      "loss": 1.136,
+      "step": 759
+    },
+    {
+      "epoch": 0.13031270763229527,
+      "grad_norm": 1.7265625,
+      "learning_rate": 1.9929051272673385e-05,
+      "loss": 1.0643,
+      "step": 760
+    },
+    {
+      "epoch": 0.13048417172128512,
+      "grad_norm": 1.7421875,
+      "learning_rate": 1.9928836369209115e-05,
+      "loss": 1.0715,
+      "step": 761
+    },
+    {
+      "epoch": 0.13065563581027498,
+      "grad_norm": 1.75,
+      "learning_rate": 1.9928621141928107e-05,
+      "loss": 1.0668,
+      "step": 762
+    },
+    {
+      "epoch": 0.13082709989926486,
+      "grad_norm": 1.796875,
+      "learning_rate": 1.9928405590837377e-05,
+      "loss": 1.2258,
+      "step": 763
+    },
+    {
+      "epoch": 0.1309985639882547,
+      "grad_norm": 1.7109375,
+      "learning_rate": 1.992818971594395e-05,
+      "loss": 1.0404,
+      "step": 764
+    },
+    {
+      "epoch": 0.13117002807724457,
+      "grad_norm": 1.6484375,
+      "learning_rate": 1.992797351725487e-05,
+      "loss": 1.0869,
+      "step": 765
+    },
+    {
+      "epoch": 0.13134149216623445,
+      "grad_norm": 1.5859375,
+      "learning_rate": 1.9927756994777193e-05,
+      "loss": 1.0574,
+      "step": 766
+    },
+    {
+      "epoch": 0.1315129562552243,
+      "grad_norm": 1.8515625,
+      "learning_rate": 1.9927540148517976e-05,
+      "loss": 1.1286,
+      "step": 767
+    },
+    {
+      "epoch": 0.13168442034421415,
+      "grad_norm": 1.734375,
+      "learning_rate": 1.9927322978484283e-05,
+      "loss": 1.129,
+      "step": 768
+    },
+    {
+      "epoch": 0.131855884433204,
+      "grad_norm": 1.703125,
+      "learning_rate": 1.9927105484683214e-05,
+      "loss": 1.0774,
+      "step": 769
+    },
+    {
+      "epoch": 0.1320273485221939,
+      "grad_norm": 1.7578125,
+      "learning_rate": 1.9926887667121847e-05,
+      "loss": 1.0588,
+      "step": 770
+    },
+    {
+      "epoch": 0.13219881261118374,
+      "grad_norm": 1.625,
+      "learning_rate": 1.9926669525807295e-05,
+      "loss": 1.1937,
+      "step": 771
+    },
+    {
+      "epoch": 0.1323702767001736,
+      "grad_norm": 1.6328125,
+      "learning_rate": 1.9926451060746668e-05,
+      "loss": 1.1166,
+      "step": 772
+    },
+    {
+      "epoch": 0.13254174078916348,
+      "grad_norm": 1.6796875,
+      "learning_rate": 1.9926232271947094e-05,
+      "loss": 1.1121,
+      "step": 773
+    },
+    {
+      "epoch": 0.13271320487815333,
+      "grad_norm": 1.796875,
+      "learning_rate": 1.9926013159415705e-05,
+      "loss": 1.1396,
+      "step": 774
+    },
+    {
+      "epoch": 0.13288466896714318,
+      "grad_norm": 1.7265625,
+      "learning_rate": 1.992579372315965e-05,
+      "loss": 1.0838,
+      "step": 775
+    },
+    {
+      "epoch": 0.13305613305613306,
+      "grad_norm": 1.7734375,
+      "learning_rate": 1.992557396318608e-05,
+      "loss": 1.1048,
+      "step": 776
+    },
+    {
+      "epoch": 0.13322759714512292,
+      "grad_norm": 1.6328125,
+      "learning_rate": 1.9925353879502175e-05,
+      "loss": 1.047,
+      "step": 777
+    },
+    {
+      "epoch": 0.13339906123411277,
+      "grad_norm": 1.6875,
+      "learning_rate": 1.9925133472115098e-05,
+      "loss": 1.0915,
+      "step": 778
+    },
+    {
+      "epoch": 0.13357052532310265,
+      "grad_norm": 1.796875,
+      "learning_rate": 1.9924912741032048e-05,
+      "loss": 1.1477,
+      "step": 779
+    },
+    {
+      "epoch": 0.1337419894120925,
+      "grad_norm": 1.59375,
+      "learning_rate": 1.9924691686260217e-05,
+      "loss": 1.0514,
+      "step": 780
+    },
+    {
+      "epoch": 0.13391345350108236,
+      "grad_norm": 1.5859375,
+      "learning_rate": 1.992447030780682e-05,
+      "loss": 1.0808,
+      "step": 781
+    },
+    {
+      "epoch": 0.13408491759007224,
+      "grad_norm": 1.75,
+      "learning_rate": 1.992424860567907e-05,
+      "loss": 1.165,
+      "step": 782
+    },
+    {
+      "epoch": 0.1342563816790621,
+      "grad_norm": 1.75,
+      "learning_rate": 1.9924026579884206e-05,
+      "loss": 1.0896,
+      "step": 783
+    },
+    {
+      "epoch": 0.13442784576805195,
+      "grad_norm": 1.671875,
+      "learning_rate": 1.9923804230429464e-05,
+      "loss": 1.1519,
+      "step": 784
+    },
+    {
+      "epoch": 0.13459930985704183,
+      "grad_norm": 1.6328125,
+      "learning_rate": 1.9923581557322094e-05,
+      "loss": 1.0796,
+      "step": 785
+    },
+    {
+      "epoch": 0.13477077394603168,
+      "grad_norm": 1.7265625,
+      "learning_rate": 1.9923358560569364e-05,
+      "loss": 1.1155,
+      "step": 786
+    },
+    {
+      "epoch": 0.13494223803502153,
+      "grad_norm": 1.6796875,
+      "learning_rate": 1.992313524017854e-05,
+      "loss": 1.0285,
+      "step": 787
+    },
+    {
+      "epoch": 0.13511370212401141,
+      "grad_norm": 1.7265625,
+      "learning_rate": 1.992291159615691e-05,
+      "loss": 1.1588,
+      "step": 788
+    },
+    {
+      "epoch": 0.13528516621300127,
+      "grad_norm": 1.7890625,
+      "learning_rate": 1.9922687628511768e-05,
+      "loss": 1.181,
+      "step": 789
+    },
+    {
+      "epoch": 0.13545663030199112,
+      "grad_norm": 1.703125,
+      "learning_rate": 1.9922463337250417e-05,
+      "loss": 1.0962,
+      "step": 790
+    },
+    {
+      "epoch": 0.135628094390981,
+      "grad_norm": 1.796875,
+      "learning_rate": 1.992223872238017e-05,
+      "loss": 1.1206,
+      "step": 791
+    },
+    {
+      "epoch": 0.13579955847997086,
+      "grad_norm": 1.703125,
+      "learning_rate": 1.9922013783908358e-05,
+      "loss": 1.117,
+      "step": 792
+    },
+    {
+      "epoch": 0.1359710225689607,
+      "grad_norm": 1.7265625,
+      "learning_rate": 1.9921788521842308e-05,
+      "loss": 1.0403,
+      "step": 793
+    },
+    {
+      "epoch": 0.13614248665795056,
+      "grad_norm": 1.8359375,
+      "learning_rate": 1.9921562936189377e-05,
+      "loss": 1.0875,
+      "step": 794
+    },
+    {
+      "epoch": 0.13631395074694044,
+      "grad_norm": 1.6796875,
+      "learning_rate": 1.9921337026956918e-05,
+      "loss": 1.1067,
+      "step": 795
+    },
+    {
+      "epoch": 0.1364854148359303,
+      "grad_norm": 1.5390625,
+      "learning_rate": 1.9921110794152296e-05,
+      "loss": 1.0446,
+      "step": 796
+    },
+    {
+      "epoch": 0.13665687892492015,
+      "grad_norm": 1.671875,
+      "learning_rate": 1.992088423778289e-05,
+      "loss": 0.9725,
+      "step": 797
+    },
+    {
+      "epoch": 0.13682834301391003,
+      "grad_norm": 1.8359375,
+      "learning_rate": 1.9920657357856092e-05,
+      "loss": 1.113,
+      "step": 798
+    },
+    {
+      "epoch": 0.13699980710289988,
+      "grad_norm": 1.9375,
+      "learning_rate": 1.9920430154379302e-05,
+      "loss": 1.1265,
+      "step": 799
+    },
+    {
+      "epoch": 0.13717127119188974,
+      "grad_norm": 1.6875,
+      "learning_rate": 1.9920202627359924e-05,
+      "loss": 1.0608,
+      "step": 800
+    },
+    {
+      "epoch": 0.13734273528087962,
+      "grad_norm": 1.734375,
+      "learning_rate": 1.9919974776805386e-05,
+      "loss": 1.0655,
+      "step": 801
+    },
+    {
+      "epoch": 0.13751419936986947,
+      "grad_norm": 1.703125,
+      "learning_rate": 1.9919746602723113e-05,
+      "loss": 1.1335,
+      "step": 802
+    },
+    {
+      "epoch": 0.13768566345885933,
+      "grad_norm": 1.7265625,
+      "learning_rate": 1.991951810512055e-05,
+      "loss": 1.1218,
+      "step": 803
+    },
+    {
+      "epoch": 0.1378571275478492,
+      "grad_norm": 1.6328125,
+      "learning_rate": 1.991928928400515e-05,
+      "loss": 1.0527,
+      "step": 804
+    },
+    {
+      "epoch": 0.13802859163683906,
+      "grad_norm": 1.609375,
+      "learning_rate": 1.9919060139384368e-05,
+      "loss": 1.0453,
+      "step": 805
+    },
+    {
+      "epoch": 0.1382000557258289,
+      "grad_norm": 1.75,
+      "learning_rate": 1.991883067126569e-05,
+      "loss": 1.1147,
+      "step": 806
+    },
+    {
+      "epoch": 0.1383715198148188,
+      "grad_norm": 1.75,
+      "learning_rate": 1.991860087965659e-05,
+      "loss": 1.1992,
+      "step": 807
+    },
+    {
+      "epoch": 0.13854298390380865,
+      "grad_norm": 1.6171875,
+      "learning_rate": 1.9918370764564563e-05,
+      "loss": 1.0232,
+      "step": 808
+    },
+    {
+      "epoch": 0.1387144479927985,
+      "grad_norm": 1.578125,
+      "learning_rate": 1.9918140325997117e-05,
+      "loss": 1.0054,
+      "step": 809
+    },
+    {
+      "epoch": 0.13888591208178838,
+      "grad_norm": 1.6953125,
+      "learning_rate": 1.991790956396177e-05,
+      "loss": 1.0625,
+      "step": 810
+    },
+    {
+      "epoch": 0.13905737617077824,
+      "grad_norm": 1.671875,
+      "learning_rate": 1.9917678478466043e-05,
+      "loss": 1.0785,
+      "step": 811
+    },
+    {
+      "epoch": 0.1392288402597681,
+      "grad_norm": 1.671875,
+      "learning_rate": 1.9917447069517473e-05,
+      "loss": 1.1155,
+      "step": 812
+    },
+    {
+      "epoch": 0.13940030434875797,
+      "grad_norm": 1.78125,
+      "learning_rate": 1.991721533712361e-05,
+      "loss": 1.0851,
+      "step": 813
+    },
+    {
+      "epoch": 0.13957176843774782,
+      "grad_norm": 1.8046875,
+      "learning_rate": 1.9916983281292008e-05,
+      "loss": 1.1355,
+      "step": 814
+    },
+    {
+      "epoch": 0.13974323252673768,
+      "grad_norm": 1.703125,
+      "learning_rate": 1.9916750902030243e-05,
+      "loss": 1.1464,
+      "step": 815
+    },
+    {
+      "epoch": 0.13991469661572756,
+      "grad_norm": 1.75,
+      "learning_rate": 1.991651819934588e-05,
+      "loss": 1.0878,
+      "step": 816
+    },
+    {
+      "epoch": 0.1400861607047174,
+      "grad_norm": 1.7578125,
+      "learning_rate": 1.9916285173246522e-05,
+      "loss": 1.0585,
+      "step": 817
+    },
+    {
+      "epoch": 0.14025762479370726,
+      "grad_norm": 1.6796875,
+      "learning_rate": 1.9916051823739758e-05,
+      "loss": 1.0578,
+      "step": 818
+    },
+    {
+      "epoch": 0.14042908888269712,
+      "grad_norm": 1.734375,
+      "learning_rate": 1.991581815083321e-05,
+      "loss": 1.1667,
+      "step": 819
+    },
+    {
+      "epoch": 0.140600552971687,
+      "grad_norm": 1.6796875,
+      "learning_rate": 1.9915584154534492e-05,
+      "loss": 0.9936,
+      "step": 820
+    },
+    {
+      "epoch": 0.14077201706067685,
+      "grad_norm": 1.59375,
+      "learning_rate": 1.991534983485123e-05,
+      "loss": 0.9491,
+      "step": 821
+    },
+    {
+      "epoch": 0.1409434811496667,
+      "grad_norm": 1.75,
+      "learning_rate": 1.9915115191791078e-05,
+      "loss": 1.15,
+      "step": 822
+    },
+    {
+      "epoch": 0.1411149452386566,
+      "grad_norm": 1.6328125,
+      "learning_rate": 1.991488022536168e-05,
+      "loss": 1.0197,
+      "step": 823
+    },
+    {
+      "epoch": 0.14128640932764644,
+      "grad_norm": 1.7265625,
+      "learning_rate": 1.99146449355707e-05,
+      "loss": 1.0575,
+      "step": 824
+    },
+    {
+      "epoch": 0.1414578734166363,
+      "grad_norm": 1.6953125,
+      "learning_rate": 1.9914409322425817e-05,
+      "loss": 1.0816,
+      "step": 825
+    },
+    {
+      "epoch": 0.14162933750562617,
+      "grad_norm": 1.7421875,
+      "learning_rate": 1.9914173385934715e-05,
+      "loss": 1.1971,
+      "step": 826
+    },
+    {
+      "epoch": 0.14180080159461603,
+      "grad_norm": 1.875,
+      "learning_rate": 1.991393712610508e-05,
+      "loss": 1.1115,
+      "step": 827
+    },
+    {
+      "epoch": 0.14197226568360588,
+      "grad_norm": 1.6875,
+      "learning_rate": 1.991370054294462e-05,
+      "loss": 1.0964,
+      "step": 828
+    },
+    {
+      "epoch": 0.14214372977259576,
+      "grad_norm": 1.65625,
+      "learning_rate": 1.9913463636461062e-05,
+      "loss": 1.0427,
+      "step": 829
+    },
+    {
+      "epoch": 0.14231519386158561,
+      "grad_norm": 1.7109375,
+      "learning_rate": 1.991322640666212e-05,
+      "loss": 1.1705,
+      "step": 830
+    },
+    {
+      "epoch": 0.14248665795057547,
+      "grad_norm": 1.6484375,
+      "learning_rate": 1.9912988853555536e-05,
+      "loss": 1.0714,
+      "step": 831
+    },
+    {
+      "epoch": 0.14265812203956535,
+      "grad_norm": 1.6328125,
+      "learning_rate": 1.9912750977149056e-05,
+      "loss": 1.0938,
+      "step": 832
+    },
+    {
+      "epoch": 0.1428295861285552,
+      "grad_norm": 1.625,
+      "learning_rate": 1.9912512777450436e-05,
+      "loss": 1.0464,
+      "step": 833
+    },
+    {
+      "epoch": 0.14300105021754506,
+      "grad_norm": 1.6015625,
+      "learning_rate": 1.9912274254467452e-05,
+      "loss": 1.0499,
+      "step": 834
+    },
+    {
+      "epoch": 0.14317251430653494,
+      "grad_norm": 1.796875,
+      "learning_rate": 1.9912035408207875e-05,
+      "loss": 1.1257,
+      "step": 835
+    },
+    {
+      "epoch": 0.1433439783955248,
+      "grad_norm": 1.640625,
+      "learning_rate": 1.9911796238679502e-05,
+      "loss": 1.0548,
+      "step": 836
+    },
+    {
+      "epoch": 0.14351544248451464,
+      "grad_norm": 1.765625,
+      "learning_rate": 1.9911556745890123e-05,
+      "loss": 1.0836,
+      "step": 837
+    },
+    {
+      "epoch": 0.14368690657350452,
+      "grad_norm": 1.671875,
+      "learning_rate": 1.9911316929847563e-05,
+      "loss": 1.1414,
+      "step": 838
+    },
+    {
+      "epoch": 0.14385837066249438,
+      "grad_norm": 1.6484375,
+      "learning_rate": 1.991107679055963e-05,
+      "loss": 1.0649,
+      "step": 839
+    },
+    {
+      "epoch": 0.14402983475148423,
+      "grad_norm": 2.0,
+      "learning_rate": 1.991083632803416e-05,
+      "loss": 1.0143,
+      "step": 840
+    },
+    {
+      "epoch": 0.14420129884047408,
+      "grad_norm": 1.7421875,
+      "learning_rate": 1.9910595542279002e-05,
+      "loss": 1.1205,
+      "step": 841
+    },
+    {
+      "epoch": 0.14437276292946397,
+      "grad_norm": 1.625,
+      "learning_rate": 1.9910354433302e-05,
+      "loss": 0.9626,
+      "step": 842
+    },
+    {
+      "epoch": 0.14454422701845382,
+      "grad_norm": 1.6796875,
+      "learning_rate": 1.991011300111102e-05,
+      "loss": 0.9339,
+      "step": 843
+    },
+    {
+      "epoch": 0.14471569110744367,
+      "grad_norm": 1.7109375,
+      "learning_rate": 1.990987124571394e-05,
+      "loss": 1.0628,
+      "step": 844
+    },
+    {
+      "epoch": 0.14488715519643355,
+      "grad_norm": 1.71875,
+      "learning_rate": 1.9909629167118638e-05,
+      "loss": 1.082,
+      "step": 845
+    },
+    {
+      "epoch": 0.1450586192854234,
+      "grad_norm": 1.9375,
+      "learning_rate": 1.9909386765333016e-05,
+      "loss": 1.1171,
+      "step": 846
+    },
+    {
+      "epoch": 0.14523008337441326,
+      "grad_norm": 1.71875,
+      "learning_rate": 1.9909144040364975e-05,
+      "loss": 1.0785,
+      "step": 847
+    },
+    {
+      "epoch": 0.14540154746340314,
+      "grad_norm": 3.125,
+      "learning_rate": 1.990890099222243e-05,
+      "loss": 1.2007,
+      "step": 848
+    },
+    {
+      "epoch": 0.145573011552393,
+      "grad_norm": 1.734375,
+      "learning_rate": 1.9908657620913315e-05,
+      "loss": 1.1942,
+      "step": 849
+    },
+    {
+      "epoch": 0.14574447564138285,
+      "grad_norm": 5.53125,
+      "learning_rate": 1.990841392644556e-05,
+      "loss": 1.1591,
+      "step": 850
+    },
+    {
+      "epoch": 0.14591593973037273,
+      "grad_norm": 1.7578125,
+      "learning_rate": 1.9908169908827113e-05,
+      "loss": 1.1318,
+      "step": 851
+    },
+    {
+      "epoch": 0.14608740381936258,
+      "grad_norm": 1.625,
+      "learning_rate": 1.9907925568065937e-05,
+      "loss": 0.9784,
+      "step": 852
+    },
+    {
+      "epoch": 0.14625886790835244,
+      "grad_norm": 1.6953125,
+      "learning_rate": 1.9907680904169996e-05,
+      "loss": 1.1982,
+      "step": 853
+    },
+    {
+      "epoch": 0.14643033199734232,
+      "grad_norm": 1.703125,
+      "learning_rate": 1.9907435917147276e-05,
+      "loss": 1.0986,
+      "step": 854
+    },
+    {
+      "epoch": 0.14660179608633217,
+      "grad_norm": 1.6484375,
+      "learning_rate": 1.9907190607005762e-05,
+      "loss": 1.0898,
+      "step": 855
+    },
+    {
+      "epoch": 0.14677326017532202,
+      "grad_norm": 1.7109375,
+      "learning_rate": 1.990694497375345e-05,
+      "loss": 1.0914,
+      "step": 856
+    },
+    {
+      "epoch": 0.1469447242643119,
+      "grad_norm": 1.9375,
+      "learning_rate": 1.9906699017398363e-05,
+      "loss": 1.1664,
+      "step": 857
+    },
+    {
+      "epoch": 0.14711618835330176,
+      "grad_norm": 1.7578125,
+      "learning_rate": 1.9906452737948512e-05,
+      "loss": 1.0166,
+      "step": 858
+    },
+    {
+      "epoch": 0.1472876524422916,
+      "grad_norm": 1.71875,
+      "learning_rate": 1.9906206135411934e-05,
+      "loss": 1.0844,
+      "step": 859
+    },
+    {
+      "epoch": 0.1474591165312815,
+      "grad_norm": 1.6484375,
+      "learning_rate": 1.990595920979667e-05,
+      "loss": 1.135,
+      "step": 860
+    },
+    {
+      "epoch": 0.14763058062027135,
+      "grad_norm": 1.78125,
+      "learning_rate": 1.9905711961110777e-05,
+      "loss": 1.0987,
+      "step": 861
+    },
+    {
+      "epoch": 0.1478020447092612,
+      "grad_norm": 1.65625,
+      "learning_rate": 1.9905464389362312e-05,
+      "loss": 1.0714,
+      "step": 862
+    },
+    {
+      "epoch": 0.14797350879825108,
+      "grad_norm": 1.7578125,
+      "learning_rate": 1.9905216494559354e-05,
+      "loss": 1.1026,
+      "step": 863
+    },
+    {
+      "epoch": 0.14814497288724093,
+      "grad_norm": 1.6640625,
+      "learning_rate": 1.9904968276709986e-05,
+      "loss": 1.0989,
+      "step": 864
+    },
+    {
+      "epoch": 0.1483164369762308,
+      "grad_norm": 1.65625,
+      "learning_rate": 1.9904719735822303e-05,
+      "loss": 1.0343,
+      "step": 865
+    },
+    {
+      "epoch": 0.14848790106522064,
+      "grad_norm": 1.625,
+      "learning_rate": 1.9904470871904413e-05,
+      "loss": 1.0357,
+      "step": 866
+    },
+    {
+      "epoch": 0.14865936515421052,
+      "grad_norm": 1.6796875,
+      "learning_rate": 1.9904221684964434e-05,
+      "loss": 1.0805,
+      "step": 867
+    },
+    {
+      "epoch": 0.14883082924320037,
+      "grad_norm": 1.65625,
+      "learning_rate": 1.9903972175010486e-05,
+      "loss": 1.1604,
+      "step": 868
+    },
+    {
+      "epoch": 0.14900229333219023,
+      "grad_norm": 1.765625,
+      "learning_rate": 1.9903722342050712e-05,
+      "loss": 1.0845,
+      "step": 869
+    },
+    {
+      "epoch": 0.1491737574211801,
+      "grad_norm": 1.609375,
+      "learning_rate": 1.9903472186093257e-05,
+      "loss": 1.022,
+      "step": 870
+    },
+    {
+      "epoch": 0.14934522151016996,
+      "grad_norm": 1.703125,
+      "learning_rate": 1.9903221707146283e-05,
+      "loss": 1.097,
+      "step": 871
+    },
+    {
+      "epoch": 0.14951668559915982,
+      "grad_norm": 1.796875,
+      "learning_rate": 1.9902970905217955e-05,
+      "loss": 1.1724,
+      "step": 872
+    },
+    {
+      "epoch": 0.1496881496881497,
+      "grad_norm": 1.5703125,
+      "learning_rate": 1.9902719780316455e-05,
+      "loss": 1.1537,
+      "step": 873
+    },
+    {
+      "epoch": 0.14985961377713955,
+      "grad_norm": 1.703125,
+      "learning_rate": 1.9902468332449973e-05,
+      "loss": 1.1606,
+      "step": 874
+    },
+    {
+      "epoch": 0.1500310778661294,
+      "grad_norm": 1.7734375,
+      "learning_rate": 1.990221656162671e-05,
+      "loss": 1.0925,
+      "step": 875
+    },
+    {
+      "epoch": 0.15020254195511928,
+      "grad_norm": 1.71875,
+      "learning_rate": 1.9901964467854876e-05,
+      "loss": 1.048,
+      "step": 876
+    },
+    {
+      "epoch": 0.15037400604410914,
+      "grad_norm": 1.65625,
+      "learning_rate": 1.9901712051142693e-05,
+      "loss": 1.1162,
+      "step": 877
+    },
+    {
+      "epoch": 0.150545470133099,
+      "grad_norm": 1.6015625,
+      "learning_rate": 1.9901459311498396e-05,
+      "loss": 1.0051,
+      "step": 878
+    },
+    {
+      "epoch": 0.15071693422208887,
+      "grad_norm": 1.7578125,
+      "learning_rate": 1.990120624893022e-05,
+      "loss": 1.1115,
+      "step": 879
+    },
+    {
+      "epoch": 0.15088839831107873,
+      "grad_norm": 1.6640625,
+      "learning_rate": 1.990095286344643e-05,
+      "loss": 1.0596,
+      "step": 880
+    },
+    {
+      "epoch": 0.15105986240006858,
+      "grad_norm": 1.734375,
+      "learning_rate": 1.990069915505528e-05,
+      "loss": 1.083,
+      "step": 881
+    },
+    {
+      "epoch": 0.15123132648905846,
+      "grad_norm": 1.7734375,
+      "learning_rate": 1.9900445123765052e-05,
+      "loss": 1.0373,
+      "step": 882
+    },
+    {
+      "epoch": 0.1514027905780483,
+      "grad_norm": 1.625,
+      "learning_rate": 1.990019076958402e-05,
+      "loss": 1.0076,
+      "step": 883
+    },
+    {
+      "epoch": 0.15157425466703817,
+      "grad_norm": 1.8359375,
+      "learning_rate": 1.9899936092520495e-05,
+      "loss": 1.1757,
+      "step": 884
+    },
+    {
+      "epoch": 0.15174571875602805,
+      "grad_norm": 1.703125,
+      "learning_rate": 1.9899681092582768e-05,
+      "loss": 1.1041,
+      "step": 885
+    },
+    {
+      "epoch": 0.1519171828450179,
+      "grad_norm": 1.6796875,
+      "learning_rate": 1.9899425769779165e-05,
+      "loss": 1.1192,
+      "step": 886
+    },
+    {
+      "epoch": 0.15208864693400775,
+      "grad_norm": 1.6328125,
+      "learning_rate": 1.9899170124118007e-05,
+      "loss": 1.1226,
+      "step": 887
+    },
+    {
+      "epoch": 0.1522601110229976,
+      "grad_norm": 1.65625,
+      "learning_rate": 1.9898914155607635e-05,
+      "loss": 1.0934,
+      "step": 888
+    },
+    {
+      "epoch": 0.1524315751119875,
+      "grad_norm": 1.7265625,
+      "learning_rate": 1.98986578642564e-05,
+      "loss": 1.1332,
+      "step": 889
+    },
+    {
+      "epoch": 0.15260303920097734,
+      "grad_norm": 1.6328125,
+      "learning_rate": 1.9898401250072653e-05,
+      "loss": 1.1179,
+      "step": 890
+    },
+    {
+      "epoch": 0.1527745032899672,
+      "grad_norm": 1.6640625,
+      "learning_rate": 1.9898144313064765e-05,
+      "loss": 1.1094,
+      "step": 891
+    },
+    {
+      "epoch": 0.15294596737895708,
+      "grad_norm": 1.671875,
+      "learning_rate": 1.9897887053241126e-05,
+      "loss": 1.0668,
+      "step": 892
+    },
+    {
+      "epoch": 0.15311743146794693,
+      "grad_norm": 1.59375,
+      "learning_rate": 1.9897629470610113e-05,
+      "loss": 1.104,
+      "step": 893
+    },
+    {
+      "epoch": 0.15328889555693678,
+      "grad_norm": 1.671875,
+      "learning_rate": 1.9897371565180133e-05,
+      "loss": 1.0576,
+      "step": 894
+    },
+    {
+      "epoch": 0.15346035964592666,
+      "grad_norm": 1.7109375,
+      "learning_rate": 1.9897113336959595e-05,
+      "loss": 1.1034,
+      "step": 895
+    },
+    {
+      "epoch": 0.15363182373491652,
+      "grad_norm": 1.8046875,
+      "learning_rate": 1.9896854785956922e-05,
+      "loss": 1.0618,
+      "step": 896
+    },
+    {
+      "epoch": 0.15380328782390637,
+      "grad_norm": 1.6640625,
+      "learning_rate": 1.9896595912180547e-05,
+      "loss": 1.1598,
+      "step": 897
+    },
+    {
+      "epoch": 0.15397475191289625,
+      "grad_norm": 1.59375,
+      "learning_rate": 1.9896336715638913e-05,
+      "loss": 1.0393,
+      "step": 898
+    },
+    {
+      "epoch": 0.1541462160018861,
+      "grad_norm": 1.6875,
+      "learning_rate": 1.989607719634047e-05,
+      "loss": 1.1174,
+      "step": 899
+    },
+    {
+      "epoch": 0.15431768009087596,
+      "grad_norm": 1.671875,
+      "learning_rate": 1.989581735429369e-05,
+      "loss": 1.1053,
+      "step": 900
+    },
+    {
+      "epoch": 0.15448914417986584,
+      "grad_norm": 1.6796875,
+      "learning_rate": 1.989555718950704e-05,
+      "loss": 1.072,
+      "step": 901
+    },
+    {
+      "epoch": 0.1546606082688557,
+      "grad_norm": 1.7734375,
+      "learning_rate": 1.9895296701989002e-05,
+      "loss": 1.095,
+      "step": 902
+    },
+    {
+      "epoch": 0.15483207235784555,
+      "grad_norm": 1.65625,
+      "learning_rate": 1.989503589174808e-05,
+      "loss": 1.1022,
+      "step": 903
+    },
+    {
+      "epoch": 0.15500353644683543,
+      "grad_norm": 1.65625,
+      "learning_rate": 1.9894774758792775e-05,
+      "loss": 0.9856,
+      "step": 904
+    },
+    {
+      "epoch": 0.15517500053582528,
+      "grad_norm": 1.7265625,
+      "learning_rate": 1.9894513303131607e-05,
+      "loss": 1.0835,
+      "step": 905
+    },
+    {
+      "epoch": 0.15534646462481513,
+      "grad_norm": 1.6796875,
+      "learning_rate": 1.98942515247731e-05,
+      "loss": 1.0362,
+      "step": 906
+    },
+    {
+      "epoch": 0.15551792871380501,
+      "grad_norm": 1.6875,
+      "learning_rate": 1.9893989423725792e-05,
+      "loss": 1.1525,
+      "step": 907
+    },
+    {
+      "epoch": 0.15568939280279487,
+      "grad_norm": 1.6171875,
+      "learning_rate": 1.989372699999823e-05,
+      "loss": 1.0246,
+      "step": 908
+    },
+    {
+      "epoch": 0.15586085689178472,
+      "grad_norm": 1.6953125,
+      "learning_rate": 1.9893464253598974e-05,
+      "loss": 1.0513,
+      "step": 909
+    },
+    {
+      "epoch": 0.1560323209807746,
+      "grad_norm": 1.6875,
+      "learning_rate": 1.9893201184536598e-05,
+      "loss": 1.055,
+      "step": 910
+    },
+    {
+      "epoch": 0.15620378506976446,
+      "grad_norm": 1.8046875,
+      "learning_rate": 1.9892937792819676e-05,
+      "loss": 1.182,
+      "step": 911
+    },
+    {
+      "epoch": 0.1563752491587543,
+      "grad_norm": 1.609375,
+      "learning_rate": 1.9892674078456795e-05,
+      "loss": 1.0651,
+      "step": 912
+    },
+    {
+      "epoch": 0.15654671324774416,
+      "grad_norm": 1.7578125,
+      "learning_rate": 1.989241004145656e-05,
+      "loss": 1.1454,
+      "step": 913
+    },
+    {
+      "epoch": 0.15671817733673404,
+      "grad_norm": 1.703125,
+      "learning_rate": 1.989214568182759e-05,
+      "loss": 1.1123,
+      "step": 914
+    },
+    {
+      "epoch": 0.1568896414257239,
+      "grad_norm": 1.625,
+      "learning_rate": 1.9891880999578492e-05,
+      "loss": 1.0728,
+      "step": 915
+    },
+    {
+      "epoch": 0.15706110551471375,
+      "grad_norm": 1.8515625,
+      "learning_rate": 1.9891615994717904e-05,
+      "loss": 1.0941,
+      "step": 916
+    },
+    {
+      "epoch": 0.15723256960370363,
+      "grad_norm": 1.6796875,
+      "learning_rate": 1.9891350667254474e-05,
+      "loss": 1.1077,
+      "step": 917
+    },
+    {
+      "epoch": 0.15740403369269348,
+      "grad_norm": 1.6171875,
+      "learning_rate": 1.9891085017196848e-05,
+      "loss": 1.0342,
+      "step": 918
+    },
+    {
+      "epoch": 0.15757549778168334,
+      "grad_norm": 1.7421875,
+      "learning_rate": 1.98908190445537e-05,
+      "loss": 1.0765,
+      "step": 919
+    },
+    {
+      "epoch": 0.15774696187067322,
+      "grad_norm": 1.65625,
+      "learning_rate": 1.989055274933369e-05,
+      "loss": 1.1752,
+      "step": 920
+    },
+    {
+      "epoch": 0.15791842595966307,
+      "grad_norm": 1.6953125,
+      "learning_rate": 1.9890286131545514e-05,
+      "loss": 1.1564,
+      "step": 921
+    },
+    {
+      "epoch": 0.15808989004865293,
+      "grad_norm": 1.6875,
+      "learning_rate": 1.9890019191197863e-05,
+      "loss": 1.0157,
+      "step": 922
+    },
+    {
+      "epoch": 0.1582613541376428,
+      "grad_norm": 1.6875,
+      "learning_rate": 1.9889751928299446e-05,
+      "loss": 0.9963,
+      "step": 923
+    },
+    {
+      "epoch": 0.15843281822663266,
+      "grad_norm": 1.6640625,
+      "learning_rate": 1.9889484342858974e-05,
+      "loss": 1.0637,
+      "step": 924
+    },
+    {
+      "epoch": 0.1586042823156225,
+      "grad_norm": 1.7421875,
+      "learning_rate": 1.988921643488518e-05,
+      "loss": 1.1039,
+      "step": 925
+    },
+    {
+      "epoch": 0.1587757464046124,
+      "grad_norm": 1.8046875,
+      "learning_rate": 1.9888948204386797e-05,
+      "loss": 1.1995,
+      "step": 926
+    },
+    {
+      "epoch": 0.15894721049360225,
+      "grad_norm": 1.640625,
+      "learning_rate": 1.9888679651372572e-05,
+      "loss": 1.0425,
+      "step": 927
+    },
+    {
+      "epoch": 0.1591186745825921,
+      "grad_norm": 1.75,
+      "learning_rate": 1.988841077585127e-05,
+      "loss": 1.0697,
+      "step": 928
+    },
+    {
+      "epoch": 0.15929013867158198,
+      "grad_norm": 1.6875,
+      "learning_rate": 1.9888141577831656e-05,
+      "loss": 1.0453,
+      "step": 929
+    },
+    {
+      "epoch": 0.15946160276057184,
+      "grad_norm": 1.7109375,
+      "learning_rate": 1.988787205732251e-05,
+      "loss": 1.1458,
+      "step": 930
+    },
+    {
+      "epoch": 0.1596330668495617,
+      "grad_norm": 1.6796875,
+      "learning_rate": 1.9887602214332622e-05,
+      "loss": 1.1275,
+      "step": 931
+    },
+    {
+      "epoch": 0.15980453093855157,
+      "grad_norm": 1.671875,
+      "learning_rate": 1.9887332048870792e-05,
+      "loss": 1.0878,
+      "step": 932
+    },
+    {
+      "epoch": 0.15997599502754142,
+      "grad_norm": 1.6484375,
+      "learning_rate": 1.988706156094583e-05,
+      "loss": 1.0844,
+      "step": 933
+    },
+    {
+      "epoch": 0.16014745911653128,
+      "grad_norm": 1.6875,
+      "learning_rate": 1.988679075056656e-05,
+      "loss": 1.0174,
+      "step": 934
+    },
+    {
+      "epoch": 0.16031892320552113,
+      "grad_norm": 1.6875,
+      "learning_rate": 1.988651961774182e-05,
+      "loss": 1.0657,
+      "step": 935
+    },
+    {
+      "epoch": 0.160490387294511,
+      "grad_norm": 1.640625,
+      "learning_rate": 1.9886248162480436e-05,
+      "loss": 1.0754,
+      "step": 936
+    },
+    {
+      "epoch": 0.16066185138350086,
+      "grad_norm": 1.734375,
+      "learning_rate": 1.9885976384791276e-05,
+      "loss": 1.1581,
+      "step": 937
+    },
+    {
+      "epoch": 0.16083331547249072,
+      "grad_norm": 1.6484375,
+      "learning_rate": 1.98857042846832e-05,
+      "loss": 1.0489,
+      "step": 938
+    },
+    {
+      "epoch": 0.1610047795614806,
+      "grad_norm": 1.6328125,
+      "learning_rate": 1.9885431862165078e-05,
+      "loss": 1.1013,
+      "step": 939
+    },
+    {
+      "epoch": 0.16117624365047045,
+      "grad_norm": 1.8046875,
+      "learning_rate": 1.98851591172458e-05,
+      "loss": 1.019,
+      "step": 940
+    },
+    {
+      "epoch": 0.1613477077394603,
+      "grad_norm": 1.5703125,
+      "learning_rate": 1.9884886049934258e-05,
+      "loss": 0.9853,
+      "step": 941
+    },
+    {
+      "epoch": 0.1615191718284502,
+      "grad_norm": 1.6875,
+      "learning_rate": 1.988461266023936e-05,
+      "loss": 1.0605,
+      "step": 942
+    },
+    {
+      "epoch": 0.16169063591744004,
+      "grad_norm": 1.7421875,
+      "learning_rate": 1.9884338948170022e-05,
+      "loss": 1.0944,
+      "step": 943
+    },
+    {
+      "epoch": 0.1618621000064299,
+      "grad_norm": 1.7109375,
+      "learning_rate": 1.9884064913735165e-05,
+      "loss": 1.0564,
+      "step": 944
+    },
+    {
+      "epoch": 0.16203356409541977,
+      "grad_norm": 1.7109375,
+      "learning_rate": 1.9883790556943736e-05,
+      "loss": 1.1367,
+      "step": 945
+    },
+    {
+      "epoch": 0.16220502818440963,
+      "grad_norm": 1.65625,
+      "learning_rate": 1.9883515877804676e-05,
+      "loss": 1.095,
+      "step": 946
+    },
+    {
+      "epoch": 0.16237649227339948,
+      "grad_norm": 1.65625,
+      "learning_rate": 1.9883240876326947e-05,
+      "loss": 1.0305,
+      "step": 947
+    },
+    {
+      "epoch": 0.16254795636238936,
+      "grad_norm": 1.6875,
+      "learning_rate": 1.9882965552519517e-05,
+      "loss": 1.0605,
+      "step": 948
+    },
+    {
+      "epoch": 0.16271942045137922,
+      "grad_norm": 1.6171875,
+      "learning_rate": 1.988268990639136e-05,
+      "loss": 1.0784,
+      "step": 949
+    },
+    {
+      "epoch": 0.16289088454036907,
+      "grad_norm": 1.6796875,
+      "learning_rate": 1.9882413937951475e-05,
+      "loss": 1.1448,
+      "step": 950
+    },
+    {
+      "epoch": 0.16306234862935895,
+      "grad_norm": 1.6875,
+      "learning_rate": 1.9882137647208858e-05,
+      "loss": 1.0633,
+      "step": 951
+    },
+    {
+      "epoch": 0.1632338127183488,
+      "grad_norm": 1.6875,
+      "learning_rate": 1.9881861034172514e-05,
+      "loss": 1.0318,
+      "step": 952
+    },
+    {
+      "epoch": 0.16340527680733866,
+      "grad_norm": 1.65625,
+      "learning_rate": 1.9881584098851475e-05,
+      "loss": 0.9796,
+      "step": 953
+    },
+    {
+      "epoch": 0.16357674089632854,
+      "grad_norm": 1.6796875,
+      "learning_rate": 1.9881306841254768e-05,
+      "loss": 1.1173,
+      "step": 954
+    },
+    {
+      "epoch": 0.1637482049853184,
+      "grad_norm": 1.7109375,
+      "learning_rate": 1.9881029261391434e-05,
+      "loss": 1.0393,
+      "step": 955
+    },
+    {
+      "epoch": 0.16391966907430824,
+      "grad_norm": 1.8125,
+      "learning_rate": 1.9880751359270526e-05,
+      "loss": 1.1808,
+      "step": 956
+    },
+    {
+      "epoch": 0.16409113316329813,
+      "grad_norm": 1.703125,
+      "learning_rate": 1.988047313490111e-05,
+      "loss": 1.1231,
+      "step": 957
+    },
+    {
+      "epoch": 0.16426259725228798,
+      "grad_norm": 1.6171875,
+      "learning_rate": 1.988019458829226e-05,
+      "loss": 1.0345,
+      "step": 958
+    },
+    {
+      "epoch": 0.16443406134127783,
+      "grad_norm": 1.7421875,
+      "learning_rate": 1.9879915719453055e-05,
+      "loss": 1.0765,
+      "step": 959
+    },
+    {
+      "epoch": 0.16460552543026769,
+      "grad_norm": 1.7578125,
+      "learning_rate": 1.9879636528392597e-05,
+      "loss": 1.1324,
+      "step": 960
+    },
+    {
+      "epoch": 0.16477698951925757,
+      "grad_norm": 1.609375,
+      "learning_rate": 1.9879357015119993e-05,
+      "loss": 1.0949,
+      "step": 961
+    },
+    {
+      "epoch": 0.16494845360824742,
+      "grad_norm": 1.53125,
+      "learning_rate": 1.9879077179644346e-05,
+      "loss": 1.0602,
+      "step": 962
+    },
+    {
+      "epoch": 0.16511991769723727,
+      "grad_norm": 1.640625,
+      "learning_rate": 1.9878797021974795e-05,
+      "loss": 1.1427,
+      "step": 963
+    },
+    {
+      "epoch": 0.16529138178622715,
+      "grad_norm": 1.765625,
+      "learning_rate": 1.9878516542120473e-05,
+      "loss": 1.0229,
+      "step": 964
+    },
+    {
+      "epoch": 0.165462845875217,
+      "grad_norm": 1.59375,
+      "learning_rate": 1.987823574009053e-05,
+      "loss": 1.0367,
+      "step": 965
+    },
+    {
+      "epoch": 0.16563430996420686,
+      "grad_norm": 1.7265625,
+      "learning_rate": 1.9877954615894118e-05,
+      "loss": 1.0842,
+      "step": 966
+    },
+    {
+      "epoch": 0.16580577405319674,
+      "grad_norm": 1.6796875,
+      "learning_rate": 1.9877673169540408e-05,
+      "loss": 1.0605,
+      "step": 967
+    },
+    {
+      "epoch": 0.1659772381421866,
+      "grad_norm": 1.6875,
+      "learning_rate": 1.9877391401038583e-05,
+      "loss": 1.1082,
+      "step": 968
+    },
+    {
+      "epoch": 0.16614870223117645,
+      "grad_norm": 1.7265625,
+      "learning_rate": 1.987710931039783e-05,
+      "loss": 1.171,
+      "step": 969
+    },
+    {
+      "epoch": 0.16632016632016633,
+      "grad_norm": 1.7109375,
+      "learning_rate": 1.9876826897627347e-05,
+      "loss": 1.0277,
+      "step": 970
+    },
+    {
+      "epoch": 0.16649163040915618,
+      "grad_norm": 1.75,
+      "learning_rate": 1.9876544162736343e-05,
+      "loss": 1.0761,
+      "step": 971
+    },
+    {
+      "epoch": 0.16666309449814604,
+      "grad_norm": 1.640625,
+      "learning_rate": 1.987626110573405e-05,
+      "loss": 1.0559,
+      "step": 972
+    },
+    {
+      "epoch": 0.16683455858713592,
+      "grad_norm": 1.609375,
+      "learning_rate": 1.9875977726629685e-05,
+      "loss": 1.0628,
+      "step": 973
+    },
+    {
+      "epoch": 0.16700602267612577,
+      "grad_norm": 1.65625,
+      "learning_rate": 1.98756940254325e-05,
+      "loss": 1.0941,
+      "step": 974
+    },
+    {
+      "epoch": 0.16717748676511562,
+      "grad_norm": 1.6953125,
+      "learning_rate": 1.9875410002151745e-05,
+      "loss": 1.0969,
+      "step": 975
+    },
+    {
+      "epoch": 0.1673489508541055,
+      "grad_norm": 1.640625,
+      "learning_rate": 1.987512565679668e-05,
+      "loss": 1.0515,
+      "step": 976
+    },
+    {
+      "epoch": 0.16752041494309536,
+      "grad_norm": 1.6953125,
+      "learning_rate": 1.987484098937658e-05,
+      "loss": 1.0269,
+      "step": 977
+    },
+    {
+      "epoch": 0.1676918790320852,
+      "grad_norm": 1.6015625,
+      "learning_rate": 1.9874555999900735e-05,
+      "loss": 1.0805,
+      "step": 978
+    },
+    {
+      "epoch": 0.1678633431210751,
+      "grad_norm": 1.6953125,
+      "learning_rate": 1.9874270688378433e-05,
+      "loss": 1.1372,
+      "step": 979
+    },
+    {
+      "epoch": 0.16803480721006495,
+      "grad_norm": 1.6953125,
+      "learning_rate": 1.987398505481898e-05,
+      "loss": 1.1989,
+      "step": 980
+    },
+    {
+      "epoch": 0.1682062712990548,
+      "grad_norm": 1.5859375,
+      "learning_rate": 1.987369909923169e-05,
+      "loss": 1.0114,
+      "step": 981
+    },
+    {
+      "epoch": 0.16837773538804465,
+      "grad_norm": 1.640625,
+      "learning_rate": 1.9873412821625894e-05,
+      "loss": 1.0541,
+      "step": 982
+    },
+    {
+      "epoch": 0.16854919947703453,
+      "grad_norm": 1.6171875,
+      "learning_rate": 1.987312622201093e-05,
+      "loss": 1.0933,
+      "step": 983
+    },
+    {
+      "epoch": 0.1687206635660244,
+      "grad_norm": 1.671875,
+      "learning_rate": 1.9872839300396132e-05,
+      "loss": 1.0502,
+      "step": 984
+    },
+    {
+      "epoch": 0.16889212765501424,
+      "grad_norm": 1.6640625,
+      "learning_rate": 1.987255205679087e-05,
+      "loss": 1.0388,
+      "step": 985
+    },
+    {
+      "epoch": 0.16906359174400412,
+      "grad_norm": 1.640625,
+      "learning_rate": 1.9872264491204514e-05,
+      "loss": 1.0497,
+      "step": 986
+    },
+    {
+      "epoch": 0.16923505583299397,
+      "grad_norm": 1.7421875,
+      "learning_rate": 1.9871976603646432e-05,
+      "loss": 1.0919,
+      "step": 987
+    },
+    {
+      "epoch": 0.16940651992198383,
+      "grad_norm": 1.734375,
+      "learning_rate": 1.987168839412602e-05,
+      "loss": 1.0666,
+      "step": 988
+    },
+    {
+      "epoch": 0.1695779840109737,
+      "grad_norm": 1.6953125,
+      "learning_rate": 1.9871399862652677e-05,
+      "loss": 1.0808,
+      "step": 989
+    },
+    {
+      "epoch": 0.16974944809996356,
+      "grad_norm": 1.59375,
+      "learning_rate": 1.9871111009235814e-05,
+      "loss": 1.0397,
+      "step": 990
+    },
+    {
+      "epoch": 0.16992091218895342,
+      "grad_norm": 1.578125,
+      "learning_rate": 1.9870821833884845e-05,
+      "loss": 0.9993,
+      "step": 991
+    },
+    {
+      "epoch": 0.1700923762779433,
+      "grad_norm": 1.75,
+      "learning_rate": 1.987053233660921e-05,
+      "loss": 1.2037,
+      "step": 992
+    },
+    {
+      "epoch": 0.17026384036693315,
+      "grad_norm": 1.578125,
+      "learning_rate": 1.9870242517418346e-05,
+      "loss": 1.0779,
+      "step": 993
+    },
+    {
+      "epoch": 0.170435304455923,
+      "grad_norm": 1.6953125,
+      "learning_rate": 1.9869952376321705e-05,
+      "loss": 1.1726,
+      "step": 994
+    },
+    {
+      "epoch": 0.17060676854491288,
+      "grad_norm": 1.671875,
+      "learning_rate": 1.986966191332875e-05,
+      "loss": 1.0898,
+      "step": 995
+    },
+    {
+      "epoch": 0.17077823263390274,
+      "grad_norm": 1.6796875,
+      "learning_rate": 1.9869371128448952e-05,
+      "loss": 1.0367,
+      "step": 996
+    },
+    {
+      "epoch": 0.1709496967228926,
+      "grad_norm": 1.6796875,
+      "learning_rate": 1.98690800216918e-05,
+      "loss": 1.063,
+      "step": 997
+    },
+    {
+      "epoch": 0.17112116081188247,
+      "grad_norm": 1.953125,
+      "learning_rate": 1.9868788593066783e-05,
+      "loss": 1.0421,
+      "step": 998
+    },
+    {
+      "epoch": 0.17129262490087233,
+      "grad_norm": 1.640625,
+      "learning_rate": 1.9868496842583412e-05,
+      "loss": 1.0655,
+      "step": 999
+    },
+    {
+      "epoch": 0.17146408898986218,
+      "grad_norm": 1.6875,
+      "learning_rate": 1.9868204770251194e-05,
+      "loss": 1.1434,
+      "step": 1000
+    },
+    {
+      "epoch": 0.17163555307885206,
+      "grad_norm": 1.6640625,
+      "learning_rate": 1.9867912376079657e-05,
+      "loss": 1.1156,
+      "step": 1001
+    },
+    {
+      "epoch": 0.1718070171678419,
+      "grad_norm": 1.6328125,
+      "learning_rate": 1.986761966007834e-05,
+      "loss": 1.0472,
+      "step": 1002
+    },
+    {
+      "epoch": 0.17197848125683177,
+      "grad_norm": 1.75,
+      "learning_rate": 1.986732662225679e-05,
+      "loss": 1.1131,
+      "step": 1003
+    },
+    {
+      "epoch": 0.17214994534582165,
+      "grad_norm": 1.640625,
+      "learning_rate": 1.986703326262456e-05,
+      "loss": 1.0208,
+      "step": 1004
+    },
+    {
+      "epoch": 0.1723214094348115,
+      "grad_norm": 1.609375,
+      "learning_rate": 1.9866739581191217e-05,
+      "loss": 1.0863,
+      "step": 1005
+    },
+    {
+      "epoch": 0.17249287352380135,
+      "grad_norm": 1.6796875,
+      "learning_rate": 1.9866445577966345e-05,
+      "loss": 1.095,
+      "step": 1006
+    },
+    {
+      "epoch": 0.1726643376127912,
+      "grad_norm": 1.7421875,
+      "learning_rate": 1.9866151252959524e-05,
+      "loss": 1.1355,
+      "step": 1007
+    },
+    {
+      "epoch": 0.1728358017017811,
+      "grad_norm": 1.6796875,
+      "learning_rate": 1.9865856606180364e-05,
+      "loss": 1.0531,
+      "step": 1008
+    },
+    {
+      "epoch": 0.17300726579077094,
+      "grad_norm": 1.703125,
+      "learning_rate": 1.9865561637638466e-05,
+      "loss": 1.0611,
+      "step": 1009
+    },
+    {
+      "epoch": 0.1731787298797608,
+      "grad_norm": 1.6953125,
+      "learning_rate": 1.9865266347343453e-05,
+      "loss": 1.0575,
+      "step": 1010
+    },
+    {
+      "epoch": 0.17335019396875068,
+      "grad_norm": 1.6171875,
+      "learning_rate": 1.9864970735304956e-05,
+      "loss": 0.9826,
+      "step": 1011
+    },
+    {
+      "epoch": 0.17352165805774053,
+      "grad_norm": 1.6171875,
+      "learning_rate": 1.9864674801532616e-05,
+      "loss": 1.0855,
+      "step": 1012
+    },
+    {
+      "epoch": 0.17369312214673038,
+      "grad_norm": 1.7109375,
+      "learning_rate": 1.9864378546036082e-05,
+      "loss": 1.1052,
+      "step": 1013
+    },
+    {
+      "epoch": 0.17386458623572026,
+      "grad_norm": 1.7734375,
+      "learning_rate": 1.986408196882502e-05,
+      "loss": 1.0988,
+      "step": 1014
+    },
+    {
+      "epoch": 0.17403605032471012,
+      "grad_norm": 1.6328125,
+      "learning_rate": 1.9863785069909095e-05,
+      "loss": 1.0497,
+      "step": 1015
+    },
+    {
+      "epoch": 0.17420751441369997,
+      "grad_norm": 1.71875,
+      "learning_rate": 1.9863487849298e-05,
+      "loss": 1.0612,
+      "step": 1016
+    },
+    {
+      "epoch": 0.17437897850268985,
+      "grad_norm": 1.78125,
+      "learning_rate": 1.9863190307001426e-05,
+      "loss": 1.0183,
+      "step": 1017
+    },
+    {
+      "epoch": 0.1745504425916797,
+      "grad_norm": 1.7734375,
+      "learning_rate": 1.986289244302907e-05,
+      "loss": 1.077,
+      "step": 1018
+    },
+    {
+      "epoch": 0.17472190668066956,
+      "grad_norm": 1.625,
+      "learning_rate": 1.986259425739065e-05,
+      "loss": 1.0255,
+      "step": 1019
+    },
+    {
+      "epoch": 0.17489337076965944,
+      "grad_norm": 1.6640625,
+      "learning_rate": 1.9862295750095896e-05,
+      "loss": 1.0715,
+      "step": 1020
+    },
+    {
+      "epoch": 0.1750648348586493,
+      "grad_norm": 1.6640625,
+      "learning_rate": 1.986199692115454e-05,
+      "loss": 1.1179,
+      "step": 1021
+    },
+    {
+      "epoch": 0.17523629894763915,
+      "grad_norm": 1.6171875,
+      "learning_rate": 1.9861697770576326e-05,
+      "loss": 1.0499,
+      "step": 1022
+    },
+    {
+      "epoch": 0.17540776303662903,
+      "grad_norm": 1.7265625,
+      "learning_rate": 1.9861398298371012e-05,
+      "loss": 1.1127,
+      "step": 1023
+    },
+    {
+      "epoch": 0.17557922712561888,
+      "grad_norm": 1.7578125,
+      "learning_rate": 1.9861098504548365e-05,
+      "loss": 1.0414,
+      "step": 1024
+    },
+    {
+      "epoch": 0.17575069121460873,
+      "grad_norm": 1.7265625,
+      "learning_rate": 1.9860798389118163e-05,
+      "loss": 1.0307,
+      "step": 1025
+    },
+    {
+      "epoch": 0.17592215530359862,
+      "grad_norm": 1.6328125,
+      "learning_rate": 1.9860497952090192e-05,
+      "loss": 1.0727,
+      "step": 1026
+    },
+    {
+      "epoch": 0.17609361939258847,
+      "grad_norm": 1.5859375,
+      "learning_rate": 1.9860197193474253e-05,
+      "loss": 1.0412,
+      "step": 1027
+    },
+    {
+      "epoch": 0.17626508348157832,
+      "grad_norm": 1.7109375,
+      "learning_rate": 1.9859896113280152e-05,
+      "loss": 1.0912,
+      "step": 1028
+    },
+    {
+      "epoch": 0.1764365475705682,
+      "grad_norm": 1.609375,
+      "learning_rate": 1.985959471151771e-05,
+      "loss": 1.1073,
+      "step": 1029
+    },
+    {
+      "epoch": 0.17660801165955806,
+      "grad_norm": 1.6171875,
+      "learning_rate": 1.9859292988196757e-05,
+      "loss": 1.1186,
+      "step": 1030
+    },
+    {
+      "epoch": 0.1767794757485479,
+      "grad_norm": 1.5703125,
+      "learning_rate": 1.9858990943327136e-05,
+      "loss": 1.0762,
+      "step": 1031
+    },
+    {
+      "epoch": 0.17695093983753776,
+      "grad_norm": 1.7890625,
+      "learning_rate": 1.985868857691869e-05,
+      "loss": 1.1374,
+      "step": 1032
+    },
+    {
+      "epoch": 0.17712240392652764,
+      "grad_norm": 1.703125,
+      "learning_rate": 1.985838588898129e-05,
+      "loss": 1.1087,
+      "step": 1033
+    },
+    {
+      "epoch": 0.1772938680155175,
+      "grad_norm": 1.6953125,
+      "learning_rate": 1.98580828795248e-05,
+      "loss": 1.0683,
+      "step": 1034
+    },
+    {
+      "epoch": 0.17746533210450735,
+      "grad_norm": 1.6484375,
+      "learning_rate": 1.9857779548559103e-05,
+      "loss": 1.1012,
+      "step": 1035
+    },
+    {
+      "epoch": 0.17763679619349723,
+      "grad_norm": 1.75,
+      "learning_rate": 1.9857475896094094e-05,
+      "loss": 1.1002,
+      "step": 1036
+    },
+    {
+      "epoch": 0.17780826028248709,
+      "grad_norm": 1.6875,
+      "learning_rate": 1.9857171922139678e-05,
+      "loss": 1.0632,
+      "step": 1037
+    },
+    {
+      "epoch": 0.17797972437147694,
+      "grad_norm": 1.5546875,
+      "learning_rate": 1.985686762670577e-05,
+      "loss": 0.9738,
+      "step": 1038
+    },
+    {
+      "epoch": 0.17815118846046682,
+      "grad_norm": 1.90625,
+      "learning_rate": 1.9856563009802286e-05,
+      "loss": 0.9626,
+      "step": 1039
+    },
+    {
+      "epoch": 0.17832265254945667,
+      "grad_norm": 1.65625,
+      "learning_rate": 1.985625807143917e-05,
+      "loss": 1.1021,
+      "step": 1040
+    },
+    {
+      "epoch": 0.17849411663844653,
+      "grad_norm": 1.71875,
+      "learning_rate": 1.985595281162636e-05,
+      "loss": 1.0798,
+      "step": 1041
+    },
+    {
+      "epoch": 0.1786655807274364,
+      "grad_norm": 1.7421875,
+      "learning_rate": 1.9855647230373817e-05,
+      "loss": 1.1084,
+      "step": 1042
+    },
+    {
+      "epoch": 0.17883704481642626,
+      "grad_norm": 1.7578125,
+      "learning_rate": 1.9855341327691506e-05,
+      "loss": 1.0181,
+      "step": 1043
+    },
+    {
+      "epoch": 0.1790085089054161,
+      "grad_norm": 1.65625,
+      "learning_rate": 1.9855035103589396e-05,
+      "loss": 1.0188,
+      "step": 1044
+    },
+    {
+      "epoch": 0.179179972994406,
+      "grad_norm": 1.640625,
+      "learning_rate": 1.9854728558077488e-05,
+      "loss": 1.1623,
+      "step": 1045
+    },
+    {
+      "epoch": 0.17935143708339585,
+      "grad_norm": 1.7109375,
+      "learning_rate": 1.9854421691165768e-05,
+      "loss": 1.0942,
+      "step": 1046
+    },
+    {
+      "epoch": 0.1795229011723857,
+      "grad_norm": 1.6015625,
+      "learning_rate": 1.9854114502864252e-05,
+      "loss": 0.9914,
+      "step": 1047
+    },
+    {
+      "epoch": 0.17969436526137558,
+      "grad_norm": 1.7421875,
+      "learning_rate": 1.9853806993182952e-05,
+      "loss": 1.1027,
+      "step": 1048
+    },
+    {
+      "epoch": 0.17986582935036544,
+      "grad_norm": 1.6640625,
+      "learning_rate": 1.98534991621319e-05,
+      "loss": 1.0728,
+      "step": 1049
+    },
+    {
+      "epoch": 0.1800372934393553,
+      "grad_norm": 1.6640625,
+      "learning_rate": 1.9853191009721137e-05,
+      "loss": 1.1005,
+      "step": 1050
+    },
+    {
+      "epoch": 0.18020875752834517,
+      "grad_norm": 1.59375,
+      "learning_rate": 1.9852882535960712e-05,
+      "loss": 1.0844,
+      "step": 1051
+    },
+    {
+      "epoch": 0.18038022161733502,
+      "grad_norm": 1.6953125,
+      "learning_rate": 1.985257374086068e-05,
+      "loss": 1.0699,
+      "step": 1052
+    },
+    {
+      "epoch": 0.18055168570632488,
+      "grad_norm": 1.6171875,
+      "learning_rate": 1.985226462443112e-05,
+      "loss": 1.0853,
+      "step": 1053
+    },
+    {
+      "epoch": 0.18072314979531473,
+      "grad_norm": 1.6171875,
+      "learning_rate": 1.9851955186682114e-05,
+      "loss": 1.0661,
+      "step": 1054
+    },
+    {
+      "epoch": 0.1808946138843046,
+      "grad_norm": 1.7265625,
+      "learning_rate": 1.9851645427623744e-05,
+      "loss": 1.11,
+      "step": 1055
+    },
+    {
+      "epoch": 0.18106607797329446,
+      "grad_norm": 1.7421875,
+      "learning_rate": 1.9851335347266124e-05,
+      "loss": 1.0096,
+      "step": 1056
+    },
+    {
+      "epoch": 0.18123754206228432,
+      "grad_norm": 1.7734375,
+      "learning_rate": 1.985102494561936e-05,
+      "loss": 1.0221,
+      "step": 1057
+    },
+    {
+      "epoch": 0.1814090061512742,
+      "grad_norm": 1.7578125,
+      "learning_rate": 1.9850714222693576e-05,
+      "loss": 1.0572,
+      "step": 1058
+    },
+    {
+      "epoch": 0.18158047024026405,
+      "grad_norm": 1.65625,
+      "learning_rate": 1.9850403178498906e-05,
+      "loss": 1.1359,
+      "step": 1059
+    },
+    {
+      "epoch": 0.1817519343292539,
+      "grad_norm": 1.6953125,
+      "learning_rate": 1.9850091813045496e-05,
+      "loss": 1.0709,
+      "step": 1060
+    },
+    {
+      "epoch": 0.1819233984182438,
+      "grad_norm": 1.875,
+      "learning_rate": 1.9849780126343502e-05,
+      "loss": 1.0282,
+      "step": 1061
+    },
+    {
+      "epoch": 0.18209486250723364,
+      "grad_norm": 1.765625,
+      "learning_rate": 1.984946811840309e-05,
+      "loss": 1.0842,
+      "step": 1062
+    },
+    {
+      "epoch": 0.1822663265962235,
+      "grad_norm": 1.71875,
+      "learning_rate": 1.9849155789234427e-05,
+      "loss": 1.0752,
+      "step": 1063
+    },
+    {
+      "epoch": 0.18243779068521337,
+      "grad_norm": 1.6875,
+      "learning_rate": 1.9848843138847704e-05,
+      "loss": 0.9987,
+      "step": 1064
+    },
+    {
+      "epoch": 0.18260925477420323,
+      "grad_norm": 1.671875,
+      "learning_rate": 1.984853016725312e-05,
+      "loss": 1.1689,
+      "step": 1065
+    },
+    {
+      "epoch": 0.18278071886319308,
+      "grad_norm": 1.640625,
+      "learning_rate": 1.9848216874460885e-05,
+      "loss": 1.0585,
+      "step": 1066
+    },
+    {
+      "epoch": 0.18295218295218296,
+      "grad_norm": 1.59375,
+      "learning_rate": 1.984790326048121e-05,
+      "loss": 1.1052,
+      "step": 1067
+    },
+    {
+      "epoch": 0.18312364704117282,
+      "grad_norm": 1.71875,
+      "learning_rate": 1.984758932532433e-05,
+      "loss": 1.0769,
+      "step": 1068
+    },
+    {
+      "epoch": 0.18329511113016267,
+      "grad_norm": 1.5625,
+      "learning_rate": 1.9847275069000473e-05,
+      "loss": 1.0894,
+      "step": 1069
+    },
+    {
+      "epoch": 0.18346657521915255,
+      "grad_norm": 1.7265625,
+      "learning_rate": 1.98469604915199e-05,
+      "loss": 1.1557,
+      "step": 1070
+    },
+    {
+      "epoch": 0.1836380393081424,
+      "grad_norm": 1.6484375,
+      "learning_rate": 1.9846645592892862e-05,
+      "loss": 1.0458,
+      "step": 1071
+    },
+    {
+      "epoch": 0.18380950339713226,
+      "grad_norm": 1.7421875,
+      "learning_rate": 1.9846330373129633e-05,
+      "loss": 1.1307,
+      "step": 1072
+    },
+    {
+      "epoch": 0.18398096748612214,
+      "grad_norm": 1.71875,
+      "learning_rate": 1.9846014832240494e-05,
+      "loss": 1.004,
+      "step": 1073
+    },
+    {
+      "epoch": 0.184152431575112,
+      "grad_norm": 1.7265625,
+      "learning_rate": 1.9845698970235733e-05,
+      "loss": 1.0424,
+      "step": 1074
+    },
+    {
+      "epoch": 0.18432389566410184,
+      "grad_norm": 1.7421875,
+      "learning_rate": 1.9845382787125653e-05,
+      "loss": 1.0541,
+      "step": 1075
+    },
+    {
+      "epoch": 0.18449535975309173,
+      "grad_norm": 1.65625,
+      "learning_rate": 1.984506628292057e-05,
+      "loss": 1.0967,
+      "step": 1076
+    },
+    {
+      "epoch": 0.18466682384208158,
+      "grad_norm": 1.6328125,
+      "learning_rate": 1.9844749457630797e-05,
+      "loss": 1.1174,
+      "step": 1077
+    },
+    {
+      "epoch": 0.18483828793107143,
+      "grad_norm": 1.65625,
+      "learning_rate": 1.9844432311266675e-05,
+      "loss": 1.0977,
+      "step": 1078
+    },
+    {
+      "epoch": 0.18500975202006129,
+      "grad_norm": 1.640625,
+      "learning_rate": 1.9844114843838542e-05,
+      "loss": 1.1425,
+      "step": 1079
+    },
+    {
+      "epoch": 0.18518121610905117,
+      "grad_norm": 1.75,
+      "learning_rate": 1.984379705535676e-05,
+      "loss": 1.0752,
+      "step": 1080
+    },
+    {
+      "epoch": 0.18535268019804102,
+      "grad_norm": 1.8046875,
+      "learning_rate": 1.9843478945831684e-05,
+      "loss": 1.1111,
+      "step": 1081
+    },
+    {
+      "epoch": 0.18552414428703087,
+      "grad_norm": 1.625,
+      "learning_rate": 1.984316051527369e-05,
+      "loss": 1.0563,
+      "step": 1082
+    },
+    {
+      "epoch": 0.18569560837602075,
+      "grad_norm": 1.765625,
+      "learning_rate": 1.9842841763693167e-05,
+      "loss": 1.0428,
+      "step": 1083
+    },
+    {
+      "epoch": 0.1858670724650106,
+      "grad_norm": 1.5703125,
+      "learning_rate": 1.9842522691100513e-05,
+      "loss": 1.0658,
+      "step": 1084
+    },
+    {
+      "epoch": 0.18603853655400046,
+      "grad_norm": 1.7421875,
+      "learning_rate": 1.984220329750613e-05,
+      "loss": 1.1422,
+      "step": 1085
+    },
+    {
+      "epoch": 0.18621000064299034,
+      "grad_norm": 1.6328125,
+      "learning_rate": 1.984188358292043e-05,
+      "loss": 1.1716,
+      "step": 1086
+    },
+    {
+      "epoch": 0.1863814647319802,
+      "grad_norm": 1.53125,
+      "learning_rate": 1.984156354735385e-05,
+      "loss": 1.0068,
+      "step": 1087
+    },
+    {
+      "epoch": 0.18655292882097005,
+      "grad_norm": 1.6796875,
+      "learning_rate": 1.984124319081682e-05,
+      "loss": 1.0618,
+      "step": 1088
+    },
+    {
+      "epoch": 0.18672439290995993,
+      "grad_norm": 1.6484375,
+      "learning_rate": 1.9840922513319793e-05,
+      "loss": 0.9207,
+      "step": 1089
+    },
+    {
+      "epoch": 0.18689585699894978,
+      "grad_norm": 1.796875,
+      "learning_rate": 1.9840601514873226e-05,
+      "loss": 1.0818,
+      "step": 1090
+    },
+    {
+      "epoch": 0.18706732108793964,
+      "grad_norm": 1.6875,
+      "learning_rate": 1.9840280195487587e-05,
+      "loss": 1.0598,
+      "step": 1091
+    },
+    {
+      "epoch": 0.18723878517692952,
+      "grad_norm": 1.7265625,
+      "learning_rate": 1.9839958555173354e-05,
+      "loss": 1.1416,
+      "step": 1092
+    },
+    {
+      "epoch": 0.18741024926591937,
+      "grad_norm": 1.5859375,
+      "learning_rate": 1.9839636593941018e-05,
+      "loss": 1.0442,
+      "step": 1093
+    },
+    {
+      "epoch": 0.18758171335490922,
+      "grad_norm": 1.7109375,
+      "learning_rate": 1.9839314311801085e-05,
+      "loss": 1.0928,
+      "step": 1094
+    },
+    {
+      "epoch": 0.1877531774438991,
+      "grad_norm": 1.71875,
+      "learning_rate": 1.9838991708764054e-05,
+      "loss": 1.0887,
+      "step": 1095
+    },
+    {
+      "epoch": 0.18792464153288896,
+      "grad_norm": 1.5703125,
+      "learning_rate": 1.9838668784840457e-05,
+      "loss": 1.0573,
+      "step": 1096
+    },
+    {
+      "epoch": 0.1880961056218788,
+      "grad_norm": 1.6015625,
+      "learning_rate": 1.9838345540040823e-05,
+      "loss": 1.0727,
+      "step": 1097
+    },
+    {
+      "epoch": 0.1882675697108687,
+      "grad_norm": 1.703125,
+      "learning_rate": 1.983802197437569e-05,
+      "loss": 1.0522,
+      "step": 1098
+    },
+    {
+      "epoch": 0.18843903379985855,
+      "grad_norm": 1.6328125,
+      "learning_rate": 1.9837698087855615e-05,
+      "loss": 1.0368,
+      "step": 1099
+    },
+    {
+      "epoch": 0.1886104978888484,
+      "grad_norm": 1.640625,
+      "learning_rate": 1.9837373880491162e-05,
+      "loss": 1.1018,
+      "step": 1100
+    },
+    {
+      "epoch": 0.18878196197783825,
+      "grad_norm": 1.6015625,
+      "learning_rate": 1.98370493522929e-05,
+      "loss": 1.0202,
+      "step": 1101
+    },
+    {
+      "epoch": 0.18895342606682813,
+      "grad_norm": 1.6484375,
+      "learning_rate": 1.9836724503271417e-05,
+      "loss": 1.1532,
+      "step": 1102
+    },
+    {
+      "epoch": 0.189124890155818,
+      "grad_norm": 1.7109375,
+      "learning_rate": 1.9836399333437307e-05,
+      "loss": 1.0213,
+      "step": 1103
+    },
+    {
+      "epoch": 0.18929635424480784,
+      "grad_norm": 1.703125,
+      "learning_rate": 1.9836073842801175e-05,
+      "loss": 1.0724,
+      "step": 1104
+    },
+    {
+      "epoch": 0.18946781833379772,
+      "grad_norm": 1.7265625,
+      "learning_rate": 1.983574803137363e-05,
+      "loss": 1.1454,
+      "step": 1105
+    },
+    {
+      "epoch": 0.18963928242278758,
+      "grad_norm": 1.65625,
+      "learning_rate": 1.9835421899165312e-05,
+      "loss": 1.0038,
+      "step": 1106
+    },
+    {
+      "epoch": 0.18981074651177743,
+      "grad_norm": 1.6875,
+      "learning_rate": 1.9835095446186842e-05,
+      "loss": 1.0624,
+      "step": 1107
+    },
+    {
+      "epoch": 0.1899822106007673,
+      "grad_norm": 1.625,
+      "learning_rate": 1.9834768672448877e-05,
+      "loss": 1.1086,
+      "step": 1108
+    },
+    {
+      "epoch": 0.19015367468975716,
+      "grad_norm": 1.75,
+      "learning_rate": 1.9834441577962072e-05,
+      "loss": 1.1179,
+      "step": 1109
+    },
+    {
+      "epoch": 0.19032513877874702,
+      "grad_norm": 1.765625,
+      "learning_rate": 1.983411416273709e-05,
+      "loss": 1.2127,
+      "step": 1110
+    },
+    {
+      "epoch": 0.1904966028677369,
+      "grad_norm": 1.734375,
+      "learning_rate": 1.983378642678462e-05,
+      "loss": 1.0469,
+      "step": 1111
+    },
+    {
+      "epoch": 0.19066806695672675,
+      "grad_norm": 1.703125,
+      "learning_rate": 1.983345837011534e-05,
+      "loss": 1.1232,
+      "step": 1112
+    },
+    {
+      "epoch": 0.1908395310457166,
+      "grad_norm": 1.6640625,
+      "learning_rate": 1.9833129992739956e-05,
+      "loss": 1.1014,
+      "step": 1113
+    },
+    {
+      "epoch": 0.19101099513470648,
+      "grad_norm": 1.671875,
+      "learning_rate": 1.983280129466917e-05,
+      "loss": 1.0703,
+      "step": 1114
+    },
+    {
+      "epoch": 0.19118245922369634,
+      "grad_norm": 1.6484375,
+      "learning_rate": 1.983247227591371e-05,
+      "loss": 1.0507,
+      "step": 1115
+    },
+    {
+      "epoch": 0.1913539233126862,
+      "grad_norm": 1.671875,
+      "learning_rate": 1.9832142936484303e-05,
+      "loss": 1.0769,
+      "step": 1116
+    },
+    {
+      "epoch": 0.19152538740167607,
+      "grad_norm": 1.7890625,
+      "learning_rate": 1.983181327639169e-05,
+      "loss": 1.071,
+      "step": 1117
+    },
+    {
+      "epoch": 0.19169685149066593,
+      "grad_norm": 1.6484375,
+      "learning_rate": 1.9831483295646623e-05,
+      "loss": 1.1299,
+      "step": 1118
+    },
+    {
+      "epoch": 0.19186831557965578,
+      "grad_norm": 1.640625,
+      "learning_rate": 1.9831152994259863e-05,
+      "loss": 0.9908,
+      "step": 1119
+    },
+    {
+      "epoch": 0.19203977966864566,
+      "grad_norm": 1.5859375,
+      "learning_rate": 1.9830822372242185e-05,
+      "loss": 1.1028,
+      "step": 1120
+    },
+    {
+      "epoch": 0.1922112437576355,
+      "grad_norm": 1.5703125,
+      "learning_rate": 1.983049142960437e-05,
+      "loss": 0.9642,
+      "step": 1121
+    },
+    {
+      "epoch": 0.19238270784662537,
+      "grad_norm": 1.7265625,
+      "learning_rate": 1.983016016635721e-05,
+      "loss": 1.0726,
+      "step": 1122
+    },
+    {
+      "epoch": 0.19255417193561525,
+      "grad_norm": 1.71875,
+      "learning_rate": 1.982982858251151e-05,
+      "loss": 1.0486,
+      "step": 1123
+    },
+    {
+      "epoch": 0.1927256360246051,
+      "grad_norm": 1.6953125,
+      "learning_rate": 1.9829496678078083e-05,
+      "loss": 1.0696,
+      "step": 1124
+    },
+    {
+      "epoch": 0.19289710011359495,
+      "grad_norm": 1.6171875,
+      "learning_rate": 1.9829164453067754e-05,
+      "loss": 1.0225,
+      "step": 1125
+    },
+    {
+      "epoch": 0.1930685642025848,
+      "grad_norm": 1.7265625,
+      "learning_rate": 1.9828831907491364e-05,
+      "loss": 1.1109,
+      "step": 1126
+    },
+    {
+      "epoch": 0.1932400282915747,
+      "grad_norm": 1.6796875,
+      "learning_rate": 1.982849904135975e-05,
+      "loss": 1.0883,
+      "step": 1127
+    },
+    {
+      "epoch": 0.19341149238056454,
+      "grad_norm": 1.7421875,
+      "learning_rate": 1.9828165854683774e-05,
+      "loss": 1.1209,
+      "step": 1128
+    },
+    {
+      "epoch": 0.1935829564695544,
+      "grad_norm": 1.6796875,
+      "learning_rate": 1.98278323474743e-05,
+      "loss": 1.1511,
+      "step": 1129
+    },
+    {
+      "epoch": 0.19375442055854428,
+      "grad_norm": 1.6484375,
+      "learning_rate": 1.9827498519742202e-05,
+      "loss": 1.1082,
+      "step": 1130
+    },
+    {
+      "epoch": 0.19392588464753413,
+      "grad_norm": 1.6640625,
+      "learning_rate": 1.982716437149837e-05,
+      "loss": 1.0247,
+      "step": 1131
+    },
+    {
+      "epoch": 0.19409734873652398,
+      "grad_norm": 1.7734375,
+      "learning_rate": 1.9826829902753706e-05,
+      "loss": 1.1547,
+      "step": 1132
+    },
+    {
+      "epoch": 0.19426881282551386,
+      "grad_norm": 1.671875,
+      "learning_rate": 1.9826495113519112e-05,
+      "loss": 1.0645,
+      "step": 1133
+    },
+    {
+      "epoch": 0.19444027691450372,
+      "grad_norm": 1.6171875,
+      "learning_rate": 1.982616000380551e-05,
+      "loss": 1.0384,
+      "step": 1134
+    },
+    {
+      "epoch": 0.19461174100349357,
+      "grad_norm": 1.71875,
+      "learning_rate": 1.9825824573623825e-05,
+      "loss": 1.086,
+      "step": 1135
+    },
+    {
+      "epoch": 0.19478320509248345,
+      "grad_norm": 1.7109375,
+      "learning_rate": 1.9825488822985007e-05,
+      "loss": 1.1152,
+      "step": 1136
+    },
+    {
+      "epoch": 0.1949546691814733,
+      "grad_norm": 1.5625,
+      "learning_rate": 1.9825152751899993e-05,
+      "loss": 0.9781,
+      "step": 1137
+    },
+    {
+      "epoch": 0.19512613327046316,
+      "grad_norm": 1.5859375,
+      "learning_rate": 1.9824816360379753e-05,
+      "loss": 1.0205,
+      "step": 1138
+    },
+    {
+      "epoch": 0.19529759735945304,
+      "grad_norm": 1.640625,
+      "learning_rate": 1.9824479648435253e-05,
+      "loss": 1.0696,
+      "step": 1139
+    },
+    {
+      "epoch": 0.1954690614484429,
+      "grad_norm": 1.65625,
+      "learning_rate": 1.9824142616077475e-05,
+      "loss": 1.0341,
+      "step": 1140
+    },
+    {
+      "epoch": 0.19564052553743275,
+      "grad_norm": 1.6484375,
+      "learning_rate": 1.9823805263317415e-05,
+      "loss": 0.989,
+      "step": 1141
+    },
+    {
+      "epoch": 0.19581198962642263,
+      "grad_norm": 1.6484375,
+      "learning_rate": 1.9823467590166073e-05,
+      "loss": 1.0009,
+      "step": 1142
+    },
+    {
+      "epoch": 0.19598345371541248,
+      "grad_norm": 1.7734375,
+      "learning_rate": 1.9823129596634456e-05,
+      "loss": 1.1526,
+      "step": 1143
+    },
+    {
+      "epoch": 0.19615491780440233,
+      "grad_norm": 1.84375,
+      "learning_rate": 1.98227912827336e-05,
+      "loss": 1.047,
+      "step": 1144
+    },
+    {
+      "epoch": 0.19632638189339222,
+      "grad_norm": 1.7265625,
+      "learning_rate": 1.9822452648474524e-05,
+      "loss": 1.0929,
+      "step": 1145
+    },
+    {
+      "epoch": 0.19649784598238207,
+      "grad_norm": 1.734375,
+      "learning_rate": 1.9822113693868282e-05,
+      "loss": 1.0357,
+      "step": 1146
+    },
+    {
+      "epoch": 0.19666931007137192,
+      "grad_norm": 1.6640625,
+      "learning_rate": 1.9821774418925924e-05,
+      "loss": 1.0461,
+      "step": 1147
+    },
+    {
+      "epoch": 0.19684077416036178,
+      "grad_norm": 1.6171875,
+      "learning_rate": 1.982143482365852e-05,
+      "loss": 1.0513,
+      "step": 1148
+    },
+    {
+      "epoch": 0.19701223824935166,
+      "grad_norm": 1.7734375,
+      "learning_rate": 1.982109490807714e-05,
+      "loss": 1.05,
+      "step": 1149
+    },
+    {
+      "epoch": 0.1971837023383415,
+      "grad_norm": 1.6640625,
+      "learning_rate": 1.982075467219287e-05,
+      "loss": 1.1052,
+      "step": 1150
+    },
+    {
+      "epoch": 0.19735516642733136,
+      "grad_norm": 1.7265625,
+      "learning_rate": 1.9820414116016814e-05,
+      "loss": 1.1074,
+      "step": 1151
+    },
+    {
+      "epoch": 0.19752663051632124,
+      "grad_norm": 1.734375,
+      "learning_rate": 1.982007323956007e-05,
+      "loss": 1.0711,
+      "step": 1152
+    },
+    {
+      "epoch": 0.1976980946053111,
+      "grad_norm": 1.6796875,
+      "learning_rate": 1.9819732042833762e-05,
+      "loss": 1.0514,
+      "step": 1153
+    },
+    {
+      "epoch": 0.19786955869430095,
+      "grad_norm": 1.6953125,
+      "learning_rate": 1.981939052584901e-05,
+      "loss": 1.0599,
+      "step": 1154
+    },
+    {
+      "epoch": 0.19804102278329083,
+      "grad_norm": 1.703125,
+      "learning_rate": 1.981904868861696e-05,
+      "loss": 1.1223,
+      "step": 1155
+    },
+    {
+      "epoch": 0.19821248687228069,
+      "grad_norm": 1.7734375,
+      "learning_rate": 1.9818706531148754e-05,
+      "loss": 1.1717,
+      "step": 1156
+    },
+    {
+      "epoch": 0.19838395096127054,
+      "grad_norm": 1.765625,
+      "learning_rate": 1.9818364053455553e-05,
+      "loss": 1.1644,
+      "step": 1157
+    },
+    {
+      "epoch": 0.19855541505026042,
+      "grad_norm": 1.65625,
+      "learning_rate": 1.9818021255548536e-05,
+      "loss": 1.064,
+      "step": 1158
+    },
+    {
+      "epoch": 0.19872687913925027,
+      "grad_norm": 1.8359375,
+      "learning_rate": 1.9817678137438866e-05,
+      "loss": 1.1106,
+      "step": 1159
+    },
+    {
+      "epoch": 0.19889834322824013,
+      "grad_norm": 1.7265625,
+      "learning_rate": 1.9817334699137746e-05,
+      "loss": 1.0919,
+      "step": 1160
+    },
+    {
+      "epoch": 0.19906980731723,
+      "grad_norm": 1.5546875,
+      "learning_rate": 1.9816990940656373e-05,
+      "loss": 1.0914,
+      "step": 1161
+    },
+    {
+      "epoch": 0.19924127140621986,
+      "grad_norm": 1.734375,
+      "learning_rate": 1.9816646862005954e-05,
+      "loss": 1.1322,
+      "step": 1162
+    },
+    {
+      "epoch": 0.19941273549520971,
+      "grad_norm": 1.6328125,
+      "learning_rate": 1.9816302463197718e-05,
+      "loss": 1.0757,
+      "step": 1163
+    },
+    {
+      "epoch": 0.1995841995841996,
+      "grad_norm": 1.6328125,
+      "learning_rate": 1.9815957744242894e-05,
+      "loss": 1.0408,
+      "step": 1164
+    },
+    {
+      "epoch": 0.19975566367318945,
+      "grad_norm": 1.703125,
+      "learning_rate": 1.9815612705152726e-05,
+      "loss": 1.1011,
+      "step": 1165
+    },
+    {
+      "epoch": 0.1999271277621793,
+      "grad_norm": 1.65625,
+      "learning_rate": 1.9815267345938464e-05,
+      "loss": 1.052,
+      "step": 1166
+    },
+    {
+      "epoch": 0.20009859185116918,
+      "grad_norm": 1.6953125,
+      "learning_rate": 1.9814921666611372e-05,
+      "loss": 0.9697,
+      "step": 1167
+    },
+    {
+      "epoch": 0.20027005594015904,
+      "grad_norm": 1.5859375,
+      "learning_rate": 1.9814575667182723e-05,
+      "loss": 1.0396,
+      "step": 1168
+    },
+    {
+      "epoch": 0.2004415200291489,
+      "grad_norm": 1.640625,
+      "learning_rate": 1.9814229347663806e-05,
+      "loss": 1.1019,
+      "step": 1169
+    },
+    {
+      "epoch": 0.20061298411813877,
+      "grad_norm": 1.625,
+      "learning_rate": 1.9813882708065914e-05,
+      "loss": 1.03,
+      "step": 1170
+    },
+    {
+      "epoch": 0.20078444820712862,
+      "grad_norm": 1.671875,
+      "learning_rate": 1.981353574840035e-05,
+      "loss": 1.0963,
+      "step": 1171
+    },
+    {
+      "epoch": 0.20095591229611848,
+      "grad_norm": 1.6953125,
+      "learning_rate": 1.981318846867843e-05,
+      "loss": 1.0078,
+      "step": 1172
+    },
+    {
+      "epoch": 0.20112737638510833,
+      "grad_norm": 1.6875,
+      "learning_rate": 1.9812840868911484e-05,
+      "loss": 1.1007,
+      "step": 1173
+    },
+    {
+      "epoch": 0.2012988404740982,
+      "grad_norm": 1.7421875,
+      "learning_rate": 1.981249294911084e-05,
+      "loss": 1.1998,
+      "step": 1174
+    },
+    {
+      "epoch": 0.20147030456308807,
+      "grad_norm": 1.671875,
+      "learning_rate": 1.9812144709287856e-05,
+      "loss": 1.0579,
+      "step": 1175
+    },
+    {
+      "epoch": 0.20164176865207792,
+      "grad_norm": 1.7109375,
+      "learning_rate": 1.9811796149453883e-05,
+      "loss": 1.1642,
+      "step": 1176
+    },
+    {
+      "epoch": 0.2018132327410678,
+      "grad_norm": 1.6484375,
+      "learning_rate": 1.9811447269620284e-05,
+      "loss": 1.1081,
+      "step": 1177
+    },
+    {
+      "epoch": 0.20198469683005765,
+      "grad_norm": 1.640625,
+      "learning_rate": 1.9811098069798448e-05,
+      "loss": 1.024,
+      "step": 1178
+    },
+    {
+      "epoch": 0.2021561609190475,
+      "grad_norm": 1.5625,
+      "learning_rate": 1.981074854999976e-05,
+      "loss": 1.077,
+      "step": 1179
+    },
+    {
+      "epoch": 0.2023276250080374,
+      "grad_norm": 2.1875,
+      "learning_rate": 1.9810398710235616e-05,
+      "loss": 1.0823,
+      "step": 1180
+    },
+    {
+      "epoch": 0.20249908909702724,
+      "grad_norm": 1.7578125,
+      "learning_rate": 1.981004855051743e-05,
+      "loss": 1.0865,
+      "step": 1181
+    },
+    {
+      "epoch": 0.2026705531860171,
+      "grad_norm": 1.734375,
+      "learning_rate": 1.9809698070856616e-05,
+      "loss": 1.1196,
+      "step": 1182
+    },
+    {
+      "epoch": 0.20284201727500697,
+      "grad_norm": 1.7734375,
+      "learning_rate": 1.980934727126461e-05,
+      "loss": 1.1146,
+      "step": 1183
+    },
+    {
+      "epoch": 0.20301348136399683,
+      "grad_norm": 1.703125,
+      "learning_rate": 1.980899615175285e-05,
+      "loss": 1.082,
+      "step": 1184
+    },
+    {
+      "epoch": 0.20318494545298668,
+      "grad_norm": 1.6875,
+      "learning_rate": 1.9808644712332786e-05,
+      "loss": 1.08,
+      "step": 1185
+    },
+    {
+      "epoch": 0.20335640954197656,
+      "grad_norm": 1.6796875,
+      "learning_rate": 1.9808292953015888e-05,
+      "loss": 1.0736,
+      "step": 1186
+    },
+    {
+      "epoch": 0.20352787363096642,
+      "grad_norm": 1.703125,
+      "learning_rate": 1.980794087381362e-05,
+      "loss": 1.0743,
+      "step": 1187
+    },
+    {
+      "epoch": 0.20369933771995627,
+      "grad_norm": 1.6796875,
+      "learning_rate": 1.980758847473746e-05,
+      "loss": 1.1012,
+      "step": 1188
+    },
+    {
+      "epoch": 0.20387080180894615,
+      "grad_norm": 1.609375,
+      "learning_rate": 1.9807235755798918e-05,
+      "loss": 0.9476,
+      "step": 1189
+    },
+    {
+      "epoch": 0.204042265897936,
+      "grad_norm": 1.65625,
+      "learning_rate": 1.980688271700948e-05,
+      "loss": 1.003,
+      "step": 1190
+    },
+    {
+      "epoch": 0.20421372998692586,
+      "grad_norm": 1.734375,
+      "learning_rate": 1.980652935838067e-05,
+      "loss": 1.0686,
+      "step": 1191
+    },
+    {
+      "epoch": 0.20438519407591574,
+      "grad_norm": 39.5,
+      "learning_rate": 1.9806175679924008e-05,
+      "loss": 1.0877,
+      "step": 1192
+    },
+    {
+      "epoch": 0.2045566581649056,
+      "grad_norm": 1.6015625,
+      "learning_rate": 1.9805821681651035e-05,
+      "loss": 1.0505,
+      "step": 1193
+    },
+    {
+      "epoch": 0.20472812225389544,
+      "grad_norm": 1.859375,
+      "learning_rate": 1.980546736357329e-05,
+      "loss": 1.117,
+      "step": 1194
+    },
+    {
+      "epoch": 0.2048995863428853,
+      "grad_norm": 1.703125,
+      "learning_rate": 1.9805112725702326e-05,
+      "loss": 1.07,
+      "step": 1195
+    },
+    {
+      "epoch": 0.20507105043187518,
+      "grad_norm": 1.703125,
+      "learning_rate": 1.980475776804972e-05,
+      "loss": 1.0231,
+      "step": 1196
+    },
+    {
+      "epoch": 0.20524251452086503,
+      "grad_norm": 1.6796875,
+      "learning_rate": 1.9804402490627038e-05,
+      "loss": 1.0791,
+      "step": 1197
+    },
+    {
+      "epoch": 0.20541397860985489,
+      "grad_norm": 1.7421875,
+      "learning_rate": 1.9804046893445868e-05,
+      "loss": 1.1112,
+      "step": 1198
+    },
+    {
+      "epoch": 0.20558544269884477,
+      "grad_norm": 1.6875,
+      "learning_rate": 1.9803690976517814e-05,
+      "loss": 1.0347,
+      "step": 1199
+    },
+    {
+      "epoch": 0.20575690678783462,
+      "grad_norm": 1.6640625,
+      "learning_rate": 1.9803334739854477e-05,
+      "loss": 1.1235,
+      "step": 1200
+    },
+    {
+      "epoch": 0.20592837087682447,
+      "grad_norm": 1.5546875,
+      "learning_rate": 1.9802978183467482e-05,
+      "loss": 1.0379,
+      "step": 1201
+    },
+    {
+      "epoch": 0.20609983496581435,
+      "grad_norm": 1.6953125,
+      "learning_rate": 1.9802621307368453e-05,
+      "loss": 1.0426,
+      "step": 1202
+    },
+    {
+      "epoch": 0.2062712990548042,
+      "grad_norm": 1.6640625,
+      "learning_rate": 1.980226411156903e-05,
+      "loss": 1.1002,
+      "step": 1203
+    },
+    {
+      "epoch": 0.20644276314379406,
+      "grad_norm": 1.53125,
+      "learning_rate": 1.980190659608086e-05,
+      "loss": 1.0196,
+      "step": 1204
+    },
+    {
+      "epoch": 0.20661422723278394,
+      "grad_norm": 1.7109375,
+      "learning_rate": 1.9801548760915607e-05,
+      "loss": 1.0715,
+      "step": 1205
+    },
+    {
+      "epoch": 0.2067856913217738,
+      "grad_norm": 1.671875,
+      "learning_rate": 1.980119060608494e-05,
+      "loss": 0.9744,
+      "step": 1206
+    },
+    {
+      "epoch": 0.20695715541076365,
+      "grad_norm": 1.6875,
+      "learning_rate": 1.9800832131600537e-05,
+      "loss": 1.1004,
+      "step": 1207
+    },
+    {
+      "epoch": 0.20712861949975353,
+      "grad_norm": 1.6484375,
+      "learning_rate": 1.98004733374741e-05,
+      "loss": 1.0266,
+      "step": 1208
+    },
+    {
+      "epoch": 0.20730008358874338,
+      "grad_norm": 1.671875,
+      "learning_rate": 1.9800114223717314e-05,
+      "loss": 0.9948,
+      "step": 1209
+    },
+    {
+      "epoch": 0.20747154767773324,
+      "grad_norm": 1.625,
+      "learning_rate": 1.97997547903419e-05,
+      "loss": 1.0021,
+      "step": 1210
+    },
+    {
+      "epoch": 0.20764301176672312,
+      "grad_norm": 1.640625,
+      "learning_rate": 1.9799395037359583e-05,
+      "loss": 1.0087,
+      "step": 1211
+    },
+    {
+      "epoch": 0.20781447585571297,
+      "grad_norm": 1.6953125,
+      "learning_rate": 1.9799034964782088e-05,
+      "loss": 1.0903,
+      "step": 1212
+    },
+    {
+      "epoch": 0.20798593994470282,
+      "grad_norm": 1.703125,
+      "learning_rate": 1.9798674572621167e-05,
+      "loss": 1.1095,
+      "step": 1213
+    },
+    {
+      "epoch": 0.2081574040336927,
+      "grad_norm": 1.671875,
+      "learning_rate": 1.979831386088857e-05,
+      "loss": 1.077,
+      "step": 1214
+    },
+    {
+      "epoch": 0.20832886812268256,
+      "grad_norm": 1.6640625,
+      "learning_rate": 1.979795282959606e-05,
+      "loss": 1.008,
+      "step": 1215
+    },
+    {
+      "epoch": 0.2085003322116724,
+      "grad_norm": 1.6875,
+      "learning_rate": 1.9797591478755413e-05,
+      "loss": 1.0306,
+      "step": 1216
+    },
+    {
+      "epoch": 0.2086717963006623,
+      "grad_norm": 1.578125,
+      "learning_rate": 1.9797229808378412e-05,
+      "loss": 1.0404,
+      "step": 1217
+    },
+    {
+      "epoch": 0.20884326038965215,
+      "grad_norm": 1.5546875,
+      "learning_rate": 1.9796867818476854e-05,
+      "loss": 1.049,
+      "step": 1218
+    },
+    {
+      "epoch": 0.209014724478642,
+      "grad_norm": 1.6484375,
+      "learning_rate": 1.9796505509062545e-05,
+      "loss": 0.9834,
+      "step": 1219
+    },
+    {
+      "epoch": 0.20918618856763185,
+      "grad_norm": 1.6640625,
+      "learning_rate": 1.97961428801473e-05,
+      "loss": 1.0725,
+      "step": 1220
+    },
+    {
+      "epoch": 0.20935765265662173,
+      "grad_norm": 1.75,
+      "learning_rate": 1.979577993174295e-05,
+      "loss": 1.0489,
+      "step": 1221
+    },
+    {
+      "epoch": 0.2095291167456116,
+      "grad_norm": 1.6796875,
+      "learning_rate": 1.9795416663861328e-05,
+      "loss": 1.0873,
+      "step": 1222
+    },
+    {
+      "epoch": 0.20970058083460144,
+      "grad_norm": 1.6328125,
+      "learning_rate": 1.9795053076514283e-05,
+      "loss": 1.008,
+      "step": 1223
+    },
+    {
+      "epoch": 0.20987204492359132,
+      "grad_norm": 3.4375,
+      "learning_rate": 1.979468916971367e-05,
+      "loss": 1.0834,
+      "step": 1224
+    },
+    {
+      "epoch": 0.21004350901258118,
+      "grad_norm": 1.7265625,
+      "learning_rate": 1.9794324943471365e-05,
+      "loss": 1.0316,
+      "step": 1225
+    },
+    {
+      "epoch": 0.21021497310157103,
+      "grad_norm": 1.7578125,
+      "learning_rate": 1.9793960397799233e-05,
+      "loss": 1.0937,
+      "step": 1226
+    },
+    {
+      "epoch": 0.2103864371905609,
+      "grad_norm": 1.65625,
+      "learning_rate": 1.979359553270918e-05,
+      "loss": 1.0939,
+      "step": 1227
+    },
+    {
+      "epoch": 0.21055790127955076,
+      "grad_norm": 1.671875,
+      "learning_rate": 1.9793230348213094e-05,
+      "loss": 1.0984,
+      "step": 1228
+    },
+    {
+      "epoch": 0.21072936536854062,
+      "grad_norm": 1.640625,
+      "learning_rate": 1.979286484432289e-05,
+      "loss": 1.0379,
+      "step": 1229
+    },
+    {
+      "epoch": 0.2109008294575305,
+      "grad_norm": 1.6953125,
+      "learning_rate": 1.9792499021050484e-05,
+      "loss": 1.1062,
+      "step": 1230
+    },
+    {
+      "epoch": 0.21107229354652035,
+      "grad_norm": 1.7109375,
+      "learning_rate": 1.979213287840781e-05,
+      "loss": 0.9739,
+      "step": 1231
+    },
+    {
+      "epoch": 0.2112437576355102,
+      "grad_norm": 1.71875,
+      "learning_rate": 1.979176641640681e-05,
+      "loss": 1.1277,
+      "step": 1232
+    },
+    {
+      "epoch": 0.21141522172450009,
+      "grad_norm": 1.7265625,
+      "learning_rate": 1.9791399635059438e-05,
+      "loss": 0.9814,
+      "step": 1233
+    },
+    {
+      "epoch": 0.21158668581348994,
+      "grad_norm": 1.6953125,
+      "learning_rate": 1.979103253437765e-05,
+      "loss": 1.0308,
+      "step": 1234
+    },
+    {
+      "epoch": 0.2117581499024798,
+      "grad_norm": 1.71875,
+      "learning_rate": 1.979066511437342e-05,
+      "loss": 1.0897,
+      "step": 1235
+    },
+    {
+      "epoch": 0.21192961399146967,
+      "grad_norm": 1.7890625,
+      "learning_rate": 1.9790297375058733e-05,
+      "loss": 1.0596,
+      "step": 1236
+    },
+    {
+      "epoch": 0.21210107808045953,
+      "grad_norm": 1.65625,
+      "learning_rate": 1.9789929316445584e-05,
+      "loss": 1.0471,
+      "step": 1237
+    },
+    {
+      "epoch": 0.21227254216944938,
+      "grad_norm": 1.609375,
+      "learning_rate": 1.9789560938545972e-05,
+      "loss": 1.0457,
+      "step": 1238
+    },
+    {
+      "epoch": 0.21244400625843926,
+      "grad_norm": 1.6953125,
+      "learning_rate": 1.9789192241371916e-05,
+      "loss": 1.0937,
+      "step": 1239
+    },
+    {
+      "epoch": 0.21261547034742911,
+      "grad_norm": 1.703125,
+      "learning_rate": 1.9788823224935436e-05,
+      "loss": 1.082,
+      "step": 1240
+    },
+    {
+      "epoch": 0.21278693443641897,
+      "grad_norm": 1.671875,
+      "learning_rate": 1.978845388924857e-05,
+      "loss": 1.0213,
+      "step": 1241
+    },
+    {
+      "epoch": 0.21295839852540885,
+      "grad_norm": 1.6796875,
+      "learning_rate": 1.9788084234323365e-05,
+      "loss": 1.0233,
+      "step": 1242
+    },
+    {
+      "epoch": 0.2131298626143987,
+      "grad_norm": 1.6484375,
+      "learning_rate": 1.978771426017187e-05,
+      "loss": 1.0576,
+      "step": 1243
+    },
+    {
+      "epoch": 0.21330132670338856,
+      "grad_norm": 1.6875,
+      "learning_rate": 1.978734396680616e-05,
+      "loss": 1.0684,
+      "step": 1244
+    },
+    {
+      "epoch": 0.2134727907923784,
+      "grad_norm": 1.7734375,
+      "learning_rate": 1.9786973354238304e-05,
+      "loss": 1.0791,
+      "step": 1245
+    },
+    {
+      "epoch": 0.2136442548813683,
+      "grad_norm": 1.6171875,
+      "learning_rate": 1.9786602422480397e-05,
+      "loss": 1.1244,
+      "step": 1246
+    },
+    {
+      "epoch": 0.21381571897035814,
+      "grad_norm": 1.671875,
+      "learning_rate": 1.9786231171544533e-05,
+      "loss": 1.0968,
+      "step": 1247
+    },
+    {
+      "epoch": 0.213987183059348,
+      "grad_norm": 1.6484375,
+      "learning_rate": 1.9785859601442813e-05,
+      "loss": 1.12,
+      "step": 1248
+    },
+    {
+      "epoch": 0.21415864714833788,
+      "grad_norm": 1.59375,
+      "learning_rate": 1.9785487712187365e-05,
+      "loss": 1.0423,
+      "step": 1249
+    },
+    {
+      "epoch": 0.21433011123732773,
+      "grad_norm": 1.625,
+      "learning_rate": 1.9785115503790314e-05,
+      "loss": 1.0428,
+      "step": 1250
+    },
+    {
+      "epoch": 0.21450157532631758,
+      "grad_norm": 1.671875,
+      "learning_rate": 1.9784742976263797e-05,
+      "loss": 1.134,
+      "step": 1251
+    },
+    {
+      "epoch": 0.21467303941530747,
+      "grad_norm": 1.75,
+      "learning_rate": 1.9784370129619966e-05,
+      "loss": 1.1417,
+      "step": 1252
+    },
+    {
+      "epoch": 0.21484450350429732,
+      "grad_norm": 1.7734375,
+      "learning_rate": 1.9783996963870983e-05,
+      "loss": 1.1021,
+      "step": 1253
+    },
+    {
+      "epoch": 0.21501596759328717,
+      "grad_norm": 1.7890625,
+      "learning_rate": 1.978362347902901e-05,
+      "loss": 1.0492,
+      "step": 1254
+    },
+    {
+      "epoch": 0.21518743168227705,
+      "grad_norm": 1.7265625,
+      "learning_rate": 1.978324967510624e-05,
+      "loss": 1.1161,
+      "step": 1255
+    },
+    {
+      "epoch": 0.2153588957712669,
+      "grad_norm": 1.7109375,
+      "learning_rate": 1.9782875552114857e-05,
+      "loss": 1.159,
+      "step": 1256
+    },
+    {
+      "epoch": 0.21553035986025676,
+      "grad_norm": 1.796875,
+      "learning_rate": 1.978250111006706e-05,
+      "loss": 1.1507,
+      "step": 1257
+    },
+    {
+      "epoch": 0.21570182394924664,
+      "grad_norm": 1.90625,
+      "learning_rate": 1.978212634897507e-05,
+      "loss": 1.0697,
+      "step": 1258
+    },
+    {
+      "epoch": 0.2158732880382365,
+      "grad_norm": 1.6875,
+      "learning_rate": 1.97817512688511e-05,
+      "loss": 1.0872,
+      "step": 1259
+    },
+    {
+      "epoch": 0.21604475212722635,
+      "grad_norm": 1.5859375,
+      "learning_rate": 1.9781375869707384e-05,
+      "loss": 1.0489,
+      "step": 1260
+    },
+    {
+      "epoch": 0.21621621621621623,
+      "grad_norm": 1.7265625,
+      "learning_rate": 1.978100015155617e-05,
+      "loss": 1.0415,
+      "step": 1261
+    },
+    {
+      "epoch": 0.21638768030520608,
+      "grad_norm": 1.7890625,
+      "learning_rate": 1.978062411440971e-05,
+      "loss": 1.0646,
+      "step": 1262
+    },
+    {
+      "epoch": 0.21655914439419593,
+      "grad_norm": 1.7109375,
+      "learning_rate": 1.9780247758280265e-05,
+      "loss": 1.0606,
+      "step": 1263
+    },
+    {
+      "epoch": 0.21673060848318582,
+      "grad_norm": 1.6484375,
+      "learning_rate": 1.9779871083180113e-05,
+      "loss": 1.0649,
+      "step": 1264
+    },
+    {
+      "epoch": 0.21690207257217567,
+      "grad_norm": 1.546875,
+      "learning_rate": 1.9779494089121537e-05,
+      "loss": 1.0249,
+      "step": 1265
+    },
+    {
+      "epoch": 0.21707353666116552,
+      "grad_norm": 1.6484375,
+      "learning_rate": 1.9779116776116837e-05,
+      "loss": 1.0685,
+      "step": 1266
+    },
+    {
+      "epoch": 0.21724500075015538,
+      "grad_norm": 1.671875,
+      "learning_rate": 1.9778739144178307e-05,
+      "loss": 1.011,
+      "step": 1267
+    },
+    {
+      "epoch": 0.21741646483914526,
+      "grad_norm": 1.78125,
+      "learning_rate": 1.9778361193318276e-05,
+      "loss": 1.118,
+      "step": 1268
+    },
+    {
+      "epoch": 0.2175879289281351,
+      "grad_norm": 1.7109375,
+      "learning_rate": 1.9777982923549062e-05,
+      "loss": 1.1274,
+      "step": 1269
+    },
+    {
+      "epoch": 0.21775939301712496,
+      "grad_norm": 1.5703125,
+      "learning_rate": 1.9777604334883003e-05,
+      "loss": 0.9581,
+      "step": 1270
+    },
+    {
+      "epoch": 0.21793085710611484,
+      "grad_norm": 1.640625,
+      "learning_rate": 1.9777225427332448e-05,
+      "loss": 1.049,
+      "step": 1271
+    },
+    {
+      "epoch": 0.2181023211951047,
+      "grad_norm": 1.6328125,
+      "learning_rate": 1.9776846200909757e-05,
+      "loss": 1.038,
+      "step": 1272
+    },
+    {
+      "epoch": 0.21827378528409455,
+      "grad_norm": 1.625,
+      "learning_rate": 1.9776466655627292e-05,
+      "loss": 1.004,
+      "step": 1273
+    },
+    {
+      "epoch": 0.21844524937308443,
+      "grad_norm": 1.7109375,
+      "learning_rate": 1.977608679149744e-05,
+      "loss": 1.0698,
+      "step": 1274
+    },
+    {
+      "epoch": 0.21861671346207429,
+      "grad_norm": 1.6015625,
+      "learning_rate": 1.9775706608532577e-05,
+      "loss": 1.025,
+      "step": 1275
+    },
+    {
+      "epoch": 0.21878817755106414,
+      "grad_norm": 1.609375,
+      "learning_rate": 1.9775326106745113e-05,
+      "loss": 0.9535,
+      "step": 1276
+    },
+    {
+      "epoch": 0.21895964164005402,
+      "grad_norm": 1.625,
+      "learning_rate": 1.9774945286147454e-05,
+      "loss": 1.0276,
+      "step": 1277
+    },
+    {
+      "epoch": 0.21913110572904387,
+      "grad_norm": 1.71875,
+      "learning_rate": 1.977456414675202e-05,
+      "loss": 1.0779,
+      "step": 1278
+    },
+    {
+      "epoch": 0.21930256981803373,
+      "grad_norm": 1.7265625,
+      "learning_rate": 1.9774182688571245e-05,
+      "loss": 1.1006,
+      "step": 1279
+    },
+    {
+      "epoch": 0.2194740339070236,
+      "grad_norm": 1.625,
+      "learning_rate": 1.9773800911617563e-05,
+      "loss": 1.1209,
+      "step": 1280
+    },
+    {
+      "epoch": 0.21964549799601346,
+      "grad_norm": 1.703125,
+      "learning_rate": 1.9773418815903428e-05,
+      "loss": 0.9508,
+      "step": 1281
+    },
+    {
+      "epoch": 0.21981696208500331,
+      "grad_norm": 1.6875,
+      "learning_rate": 1.97730364014413e-05,
+      "loss": 1.0266,
+      "step": 1282
+    },
+    {
+      "epoch": 0.2199884261739932,
+      "grad_norm": 1.6875,
+      "learning_rate": 1.9772653668243658e-05,
+      "loss": 1.0921,
+      "step": 1283
+    },
+    {
+      "epoch": 0.22015989026298305,
+      "grad_norm": 1.6484375,
+      "learning_rate": 1.9772270616322977e-05,
+      "loss": 1.0855,
+      "step": 1284
+    },
+    {
+      "epoch": 0.2203313543519729,
+      "grad_norm": 1.6484375,
+      "learning_rate": 1.9771887245691753e-05,
+      "loss": 1.025,
+      "step": 1285
+    },
+    {
+      "epoch": 0.22050281844096278,
+      "grad_norm": 1.6796875,
+      "learning_rate": 1.977150355636249e-05,
+      "loss": 1.0946,
+      "step": 1286
+    },
+    {
+      "epoch": 0.22067428252995264,
+      "grad_norm": 1.609375,
+      "learning_rate": 1.9771119548347693e-05,
+      "loss": 1.0725,
+      "step": 1287
+    },
+    {
+      "epoch": 0.2208457466189425,
+      "grad_norm": 1.734375,
+      "learning_rate": 1.97707352216599e-05,
+      "loss": 1.0588,
+      "step": 1288
+    },
+    {
+      "epoch": 0.22101721070793237,
+      "grad_norm": 1.6328125,
+      "learning_rate": 1.977035057631164e-05,
+      "loss": 0.9982,
+      "step": 1289
+    },
+    {
+      "epoch": 0.22118867479692222,
+      "grad_norm": 1.7421875,
+      "learning_rate": 1.976996561231545e-05,
+      "loss": 1.0788,
+      "step": 1290
+    },
+    {
+      "epoch": 0.22136013888591208,
+      "grad_norm": 1.7109375,
+      "learning_rate": 1.976958032968389e-05,
+      "loss": 1.0912,
+      "step": 1291
+    },
+    {
+      "epoch": 0.22153160297490193,
+      "grad_norm": 1.734375,
+      "learning_rate": 1.976919472842953e-05,
+      "loss": 1.1632,
+      "step": 1292
+    },
+    {
+      "epoch": 0.2217030670638918,
+      "grad_norm": 1.6953125,
+      "learning_rate": 1.9768808808564944e-05,
+      "loss": 1.0549,
+      "step": 1293
+    },
+    {
+      "epoch": 0.22187453115288167,
+      "grad_norm": 1.7578125,
+      "learning_rate": 1.9768422570102717e-05,
+      "loss": 1.0129,
+      "step": 1294
+    },
+    {
+      "epoch": 0.22204599524187152,
+      "grad_norm": 1.7890625,
+      "learning_rate": 1.9768036013055444e-05,
+      "loss": 1.0461,
+      "step": 1295
+    },
+    {
+      "epoch": 0.2222174593308614,
+      "grad_norm": 1.578125,
+      "learning_rate": 1.976764913743573e-05,
+      "loss": 1.0885,
+      "step": 1296
+    },
+    {
+      "epoch": 0.22238892341985125,
+      "grad_norm": 1.578125,
+      "learning_rate": 1.97672619432562e-05,
+      "loss": 1.068,
+      "step": 1297
+    },
+    {
+      "epoch": 0.2225603875088411,
+      "grad_norm": 1.609375,
+      "learning_rate": 1.976687443052948e-05,
+      "loss": 1.1145,
+      "step": 1298
+    },
+    {
+      "epoch": 0.222731851597831,
+      "grad_norm": 1.5390625,
+      "learning_rate": 1.9766486599268203e-05,
+      "loss": 1.0323,
+      "step": 1299
+    },
+    {
+      "epoch": 0.22290331568682084,
+      "grad_norm": 1.640625,
+      "learning_rate": 1.976609844948502e-05,
+      "loss": 1.0156,
+      "step": 1300
+    },
+    {
+      "epoch": 0.2230747797758107,
+      "grad_norm": 1.65625,
+      "learning_rate": 1.9765709981192592e-05,
+      "loss": 1.0087,
+      "step": 1301
+    },
+    {
+      "epoch": 0.22324624386480058,
+      "grad_norm": 1.65625,
+      "learning_rate": 1.9765321194403588e-05,
+      "loss": 0.998,
+      "step": 1302
+    },
+    {
+      "epoch": 0.22341770795379043,
+      "grad_norm": 1.6796875,
+      "learning_rate": 1.9764932089130687e-05,
+      "loss": 1.0416,
+      "step": 1303
+    },
+    {
+      "epoch": 0.22358917204278028,
+      "grad_norm": 1.6796875,
+      "learning_rate": 1.9764542665386578e-05,
+      "loss": 1.0588,
+      "step": 1304
+    },
+    {
+      "epoch": 0.22376063613177016,
+      "grad_norm": 1.75,
+      "learning_rate": 1.9764152923183965e-05,
+      "loss": 0.9994,
+      "step": 1305
+    },
+    {
+      "epoch": 0.22393210022076002,
+      "grad_norm": 1.640625,
+      "learning_rate": 1.9763762862535556e-05,
+      "loss": 1.1422,
+      "step": 1306
+    },
+    {
+      "epoch": 0.22410356430974987,
+      "grad_norm": 1.7890625,
+      "learning_rate": 1.976337248345407e-05,
+      "loss": 1.0702,
+      "step": 1307
+    },
+    {
+      "epoch": 0.22427502839873975,
+      "grad_norm": 1.5703125,
+      "learning_rate": 1.9762981785952248e-05,
+      "loss": 1.0187,
+      "step": 1308
+    },
+    {
+      "epoch": 0.2244464924877296,
+      "grad_norm": 1.6796875,
+      "learning_rate": 1.976259077004282e-05,
+      "loss": 1.0987,
+      "step": 1309
+    },
+    {
+      "epoch": 0.22461795657671946,
+      "grad_norm": 1.65625,
+      "learning_rate": 1.976219943573855e-05,
+      "loss": 1.0949,
+      "step": 1310
+    },
+    {
+      "epoch": 0.22478942066570934,
+      "grad_norm": 1.7265625,
+      "learning_rate": 1.976180778305219e-05,
+      "loss": 1.0813,
+      "step": 1311
+    },
+    {
+      "epoch": 0.2249608847546992,
+      "grad_norm": 1.671875,
+      "learning_rate": 1.976141581199652e-05,
+      "loss": 1.1172,
+      "step": 1312
+    },
+    {
+      "epoch": 0.22513234884368905,
+      "grad_norm": 1.65625,
+      "learning_rate": 1.9761023522584325e-05,
+      "loss": 1.125,
+      "step": 1313
+    },
+    {
+      "epoch": 0.2253038129326789,
+      "grad_norm": 1.7421875,
+      "learning_rate": 1.976063091482839e-05,
+      "loss": 1.0168,
+      "step": 1314
+    },
+    {
+      "epoch": 0.22547527702166878,
+      "grad_norm": 1.6875,
+      "learning_rate": 1.9760237988741534e-05,
+      "loss": 1.0785,
+      "step": 1315
+    },
+    {
+      "epoch": 0.22564674111065863,
+      "grad_norm": 1.640625,
+      "learning_rate": 1.975984474433656e-05,
+      "loss": 1.0187,
+      "step": 1316
+    },
+    {
+      "epoch": 0.2258182051996485,
+      "grad_norm": 1.6953125,
+      "learning_rate": 1.9759451181626295e-05,
+      "loss": 1.0427,
+      "step": 1317
+    },
+    {
+      "epoch": 0.22598966928863837,
+      "grad_norm": 1.7890625,
+      "learning_rate": 1.9759057300623578e-05,
+      "loss": 1.1146,
+      "step": 1318
+    },
+    {
+      "epoch": 0.22616113337762822,
+      "grad_norm": 1.6796875,
+      "learning_rate": 1.975866310134125e-05,
+      "loss": 1.0689,
+      "step": 1319
+    },
+    {
+      "epoch": 0.22633259746661807,
+      "grad_norm": 1.65625,
+      "learning_rate": 1.9758268583792176e-05,
+      "loss": 1.0353,
+      "step": 1320
+    },
+    {
+      "epoch": 0.22650406155560796,
+      "grad_norm": 1.6171875,
+      "learning_rate": 1.9757873747989214e-05,
+      "loss": 0.9958,
+      "step": 1321
+    },
+    {
+      "epoch": 0.2266755256445978,
+      "grad_norm": 1.6015625,
+      "learning_rate": 1.9757478593945244e-05,
+      "loss": 1.0269,
+      "step": 1322
+    },
+    {
+      "epoch": 0.22684698973358766,
+      "grad_norm": 1.7734375,
+      "learning_rate": 1.9757083121673157e-05,
+      "loss": 1.1019,
+      "step": 1323
+    },
+    {
+      "epoch": 0.22701845382257754,
+      "grad_norm": 1.6953125,
+      "learning_rate": 1.9756687331185845e-05,
+      "loss": 1.0571,
+      "step": 1324
+    },
+    {
+      "epoch": 0.2271899179115674,
+      "grad_norm": 1.84375,
+      "learning_rate": 1.9756291222496217e-05,
+      "loss": 1.163,
+      "step": 1325
+    },
+    {
+      "epoch": 0.22736138200055725,
+      "grad_norm": 1.59375,
+      "learning_rate": 1.9755894795617196e-05,
+      "loss": 1.0597,
+      "step": 1326
+    },
+    {
+      "epoch": 0.22753284608954713,
+      "grad_norm": 1.609375,
+      "learning_rate": 1.975549805056171e-05,
+      "loss": 0.9859,
+      "step": 1327
+    },
+    {
+      "epoch": 0.22770431017853698,
+      "grad_norm": 1.6015625,
+      "learning_rate": 1.9755100987342695e-05,
+      "loss": 0.979,
+      "step": 1328
+    },
+    {
+      "epoch": 0.22787577426752684,
+      "grad_norm": 1.6875,
+      "learning_rate": 1.9754703605973104e-05,
+      "loss": 1.0554,
+      "step": 1329
+    },
+    {
+      "epoch": 0.22804723835651672,
+      "grad_norm": 1.6640625,
+      "learning_rate": 1.975430590646589e-05,
+      "loss": 1.0303,
+      "step": 1330
+    },
+    {
+      "epoch": 0.22821870244550657,
+      "grad_norm": 1.65625,
+      "learning_rate": 1.9753907888834037e-05,
+      "loss": 1.1333,
+      "step": 1331
+    },
+    {
+      "epoch": 0.22839016653449642,
+      "grad_norm": 1.6484375,
+      "learning_rate": 1.9753509553090513e-05,
+      "loss": 1.0818,
+      "step": 1332
+    },
+    {
+      "epoch": 0.2285616306234863,
+      "grad_norm": 1.6796875,
+      "learning_rate": 1.9753110899248313e-05,
+      "loss": 1.1285,
+      "step": 1333
+    },
+    {
+      "epoch": 0.22873309471247616,
+      "grad_norm": 1.65625,
+      "learning_rate": 1.9752711927320444e-05,
+      "loss": 1.1243,
+      "step": 1334
+    },
+    {
+      "epoch": 0.228904558801466,
+      "grad_norm": 1.625,
+      "learning_rate": 1.975231263731991e-05,
+      "loss": 1.0094,
+      "step": 1335
+    },
+    {
+      "epoch": 0.2290760228904559,
+      "grad_norm": 1.6796875,
+      "learning_rate": 1.975191302925974e-05,
+      "loss": 1.1451,
+      "step": 1336
+    },
+    {
+      "epoch": 0.22924748697944575,
+      "grad_norm": 1.5703125,
+      "learning_rate": 1.975151310315296e-05,
+      "loss": 1.0273,
+      "step": 1337
+    },
+    {
+      "epoch": 0.2294189510684356,
+      "grad_norm": 1.5703125,
+      "learning_rate": 1.975111285901262e-05,
+      "loss": 1.1098,
+      "step": 1338
+    },
+    {
+      "epoch": 0.22959041515742545,
+      "grad_norm": 1.65625,
+      "learning_rate": 1.975071229685177e-05,
+      "loss": 1.0838,
+      "step": 1339
+    },
+    {
+      "epoch": 0.22976187924641533,
+      "grad_norm": 1.640625,
+      "learning_rate": 1.9750311416683475e-05,
+      "loss": 1.0626,
+      "step": 1340
+    },
+    {
+      "epoch": 0.2299333433354052,
+      "grad_norm": 1.6328125,
+      "learning_rate": 1.974991021852081e-05,
+      "loss": 0.9985,
+      "step": 1341
+    },
+    {
+      "epoch": 0.23010480742439504,
+      "grad_norm": 1.640625,
+      "learning_rate": 1.9749508702376853e-05,
+      "loss": 1.0117,
+      "step": 1342
+    },
+    {
+      "epoch": 0.23027627151338492,
+      "grad_norm": 1.734375,
+      "learning_rate": 1.9749106868264706e-05,
+      "loss": 1.1324,
+      "step": 1343
+    },
+    {
+      "epoch": 0.23044773560237478,
+      "grad_norm": 1.6484375,
+      "learning_rate": 1.9748704716197474e-05,
+      "loss": 1.0518,
+      "step": 1344
+    },
+    {
+      "epoch": 0.23061919969136463,
+      "grad_norm": 1.609375,
+      "learning_rate": 1.9748302246188267e-05,
+      "loss": 1.0752,
+      "step": 1345
+    },
+    {
+      "epoch": 0.2307906637803545,
+      "grad_norm": 1.671875,
+      "learning_rate": 1.974789945825022e-05,
+      "loss": 1.0666,
+      "step": 1346
+    },
+    {
+      "epoch": 0.23096212786934436,
+      "grad_norm": 1.703125,
+      "learning_rate": 1.974749635239646e-05,
+      "loss": 1.1034,
+      "step": 1347
+    },
+    {
+      "epoch": 0.23113359195833422,
+      "grad_norm": 1.6875,
+      "learning_rate": 1.974709292864014e-05,
+      "loss": 1.0834,
+      "step": 1348
+    },
+    {
+      "epoch": 0.2313050560473241,
+      "grad_norm": 1.7109375,
+      "learning_rate": 1.9746689186994417e-05,
+      "loss": 1.1175,
+      "step": 1349
+    },
+    {
+      "epoch": 0.23147652013631395,
+      "grad_norm": 1.703125,
+      "learning_rate": 1.9746285127472458e-05,
+      "loss": 1.0594,
+      "step": 1350
+    },
+    {
+      "epoch": 0.2316479842253038,
+      "grad_norm": 1.671875,
+      "learning_rate": 1.9745880750087437e-05,
+      "loss": 1.083,
+      "step": 1351
+    },
+    {
+      "epoch": 0.23181944831429369,
+      "grad_norm": 1.625,
+      "learning_rate": 1.9745476054852543e-05,
+      "loss": 1.0639,
+      "step": 1352
+    },
+    {
+      "epoch": 0.23199091240328354,
+      "grad_norm": 1.65625,
+      "learning_rate": 1.9745071041780983e-05,
+      "loss": 1.0399,
+      "step": 1353
+    },
+    {
+      "epoch": 0.2321623764922734,
+      "grad_norm": 1.625,
+      "learning_rate": 1.9744665710885955e-05,
+      "loss": 1.1361,
+      "step": 1354
+    },
+    {
+      "epoch": 0.23233384058126327,
+      "grad_norm": 1.625,
+      "learning_rate": 1.9744260062180686e-05,
+      "loss": 0.987,
+      "step": 1355
+    },
+    {
+      "epoch": 0.23250530467025313,
+      "grad_norm": 1.640625,
+      "learning_rate": 1.9743854095678402e-05,
+      "loss": 1.2077,
+      "step": 1356
+    },
+    {
+      "epoch": 0.23267676875924298,
+      "grad_norm": 1.671875,
+      "learning_rate": 1.9743447811392342e-05,
+      "loss": 1.1149,
+      "step": 1357
+    },
+    {
+      "epoch": 0.23284823284823286,
+      "grad_norm": 1.703125,
+      "learning_rate": 1.974304120933576e-05,
+      "loss": 1.0483,
+      "step": 1358
+    },
+    {
+      "epoch": 0.23301969693722271,
+      "grad_norm": 1.625,
+      "learning_rate": 1.9742634289521915e-05,
+      "loss": 1.032,
+      "step": 1359
+    },
+    {
+      "epoch": 0.23319116102621257,
+      "grad_norm": 1.671875,
+      "learning_rate": 1.9742227051964078e-05,
+      "loss": 1.0213,
+      "step": 1360
+    },
+    {
+      "epoch": 0.23336262511520242,
+      "grad_norm": 1.8125,
+      "learning_rate": 1.9741819496675533e-05,
+      "loss": 1.0281,
+      "step": 1361
+    },
+    {
+      "epoch": 0.2335340892041923,
+      "grad_norm": 1.6015625,
+      "learning_rate": 1.974141162366957e-05,
+      "loss": 0.9483,
+      "step": 1362
+    },
+    {
+      "epoch": 0.23370555329318216,
+      "grad_norm": 1.765625,
+      "learning_rate": 1.9741003432959486e-05,
+      "loss": 1.1563,
+      "step": 1363
+    },
+    {
+      "epoch": 0.233877017382172,
+      "grad_norm": 1.6484375,
+      "learning_rate": 1.9740594924558606e-05,
+      "loss": 1.1034,
+      "step": 1364
+    },
+    {
+      "epoch": 0.2340484814711619,
+      "grad_norm": 1.53125,
+      "learning_rate": 1.974018609848024e-05,
+      "loss": 0.9837,
+      "step": 1365
+    },
+    {
+      "epoch": 0.23421994556015174,
+      "grad_norm": 1.59375,
+      "learning_rate": 1.973977695473773e-05,
+      "loss": 1.0679,
+      "step": 1366
+    },
+    {
+      "epoch": 0.2343914096491416,
+      "grad_norm": 1.8125,
+      "learning_rate": 1.9739367493344415e-05,
+      "loss": 1.1194,
+      "step": 1367
+    },
+    {
+      "epoch": 0.23456287373813148,
+      "grad_norm": 1.640625,
+      "learning_rate": 1.9738957714313653e-05,
+      "loss": 1.0915,
+      "step": 1368
+    },
+    {
+      "epoch": 0.23473433782712133,
+      "grad_norm": 1.703125,
+      "learning_rate": 1.9738547617658806e-05,
+      "loss": 1.076,
+      "step": 1369
+    },
+    {
+      "epoch": 0.23490580191611118,
+      "grad_norm": 1.6640625,
+      "learning_rate": 1.973813720339325e-05,
+      "loss": 0.9973,
+      "step": 1370
+    },
+    {
+      "epoch": 0.23507726600510107,
+      "grad_norm": 1.671875,
+      "learning_rate": 1.973772647153037e-05,
+      "loss": 1.224,
+      "step": 1371
+    },
+    {
+      "epoch": 0.23524873009409092,
+      "grad_norm": 1.5859375,
+      "learning_rate": 1.9737315422083557e-05,
+      "loss": 1.1084,
+      "step": 1372
+    },
+    {
+      "epoch": 0.23542019418308077,
+      "grad_norm": 1.703125,
+      "learning_rate": 1.9736904055066227e-05,
+      "loss": 1.1346,
+      "step": 1373
+    },
+    {
+      "epoch": 0.23559165827207065,
+      "grad_norm": 1.7109375,
+      "learning_rate": 1.9736492370491786e-05,
+      "loss": 1.0141,
+      "step": 1374
+    },
+    {
+      "epoch": 0.2357631223610605,
+      "grad_norm": 1.6171875,
+      "learning_rate": 1.973608036837366e-05,
+      "loss": 1.0714,
+      "step": 1375
+    },
+    {
+      "epoch": 0.23593458645005036,
+      "grad_norm": 1.671875,
+      "learning_rate": 1.9735668048725296e-05,
+      "loss": 1.131,
+      "step": 1376
+    },
+    {
+      "epoch": 0.23610605053904024,
+      "grad_norm": 1.5546875,
+      "learning_rate": 1.9735255411560137e-05,
+      "loss": 1.0559,
+      "step": 1377
+    },
+    {
+      "epoch": 0.2362775146280301,
+      "grad_norm": 1.6484375,
+      "learning_rate": 1.973484245689164e-05,
+      "loss": 1.0569,
+      "step": 1378
+    },
+    {
+      "epoch": 0.23644897871701995,
+      "grad_norm": 1.671875,
+      "learning_rate": 1.973442918473327e-05,
+      "loss": 1.0454,
+      "step": 1379
+    },
+    {
+      "epoch": 0.23662044280600983,
+      "grad_norm": 1.71875,
+      "learning_rate": 1.9734015595098507e-05,
+      "loss": 0.9595,
+      "step": 1380
+    },
+    {
+      "epoch": 0.23679190689499968,
+      "grad_norm": 1.6328125,
+      "learning_rate": 1.9733601688000837e-05,
+      "loss": 1.1188,
+      "step": 1381
+    },
+    {
+      "epoch": 0.23696337098398954,
+      "grad_norm": 1.6015625,
+      "learning_rate": 1.973318746345377e-05,
+      "loss": 0.972,
+      "step": 1382
+    },
+    {
+      "epoch": 0.23713483507297942,
+      "grad_norm": 1.6484375,
+      "learning_rate": 1.9732772921470804e-05,
+      "loss": 0.9974,
+      "step": 1383
+    },
+    {
+      "epoch": 0.23730629916196927,
+      "grad_norm": 1.7265625,
+      "learning_rate": 1.9732358062065465e-05,
+      "loss": 1.0464,
+      "step": 1384
+    },
+    {
+      "epoch": 0.23747776325095912,
+      "grad_norm": 1.640625,
+      "learning_rate": 1.973194288525128e-05,
+      "loss": 1.1304,
+      "step": 1385
+    },
+    {
+      "epoch": 0.23764922733994898,
+      "grad_norm": 1.6328125,
+      "learning_rate": 1.9731527391041786e-05,
+      "loss": 1.1416,
+      "step": 1386
+    },
+    {
+      "epoch": 0.23782069142893886,
+      "grad_norm": 1.734375,
+      "learning_rate": 1.9731111579450545e-05,
+      "loss": 1.1684,
+      "step": 1387
+    },
+    {
+      "epoch": 0.2379921555179287,
+      "grad_norm": 1.640625,
+      "learning_rate": 1.9730695450491106e-05,
+      "loss": 1.0167,
+      "step": 1388
+    },
+    {
+      "epoch": 0.23816361960691856,
+      "grad_norm": 1.546875,
+      "learning_rate": 1.973027900417705e-05,
+      "loss": 0.9737,
+      "step": 1389
+    },
+    {
+      "epoch": 0.23833508369590845,
+      "grad_norm": 1.6953125,
+      "learning_rate": 1.9729862240521953e-05,
+      "loss": 1.0621,
+      "step": 1390
+    },
+    {
+      "epoch": 0.2385065477848983,
+      "grad_norm": 1.78125,
+      "learning_rate": 1.9729445159539412e-05,
+      "loss": 1.0712,
+      "step": 1391
+    },
+    {
+      "epoch": 0.23867801187388815,
+      "grad_norm": 1.703125,
+      "learning_rate": 1.9729027761243022e-05,
+      "loss": 1.0961,
+      "step": 1392
+    },
+    {
+      "epoch": 0.23884947596287803,
+      "grad_norm": 1.546875,
+      "learning_rate": 1.9728610045646402e-05,
+      "loss": 1.0431,
+      "step": 1393
+    },
+    {
+      "epoch": 0.2390209400518679,
+      "grad_norm": 1.5625,
+      "learning_rate": 1.9728192012763174e-05,
+      "loss": 1.0822,
+      "step": 1394
+    },
+    {
+      "epoch": 0.23919240414085774,
+      "grad_norm": 1.6875,
+      "learning_rate": 1.9727773662606973e-05,
+      "loss": 1.1067,
+      "step": 1395
+    },
+    {
+      "epoch": 0.23936386822984762,
+      "grad_norm": 1.7109375,
+      "learning_rate": 1.9727354995191442e-05,
+      "loss": 1.099,
+      "step": 1396
+    },
+    {
+      "epoch": 0.23953533231883747,
+      "grad_norm": 1.640625,
+      "learning_rate": 1.972693601053023e-05,
+      "loss": 1.1294,
+      "step": 1397
+    },
+    {
+      "epoch": 0.23970679640782733,
+      "grad_norm": 1.7578125,
+      "learning_rate": 1.9726516708637012e-05,
+      "loss": 1.1383,
+      "step": 1398
+    },
+    {
+      "epoch": 0.2398782604968172,
+      "grad_norm": 1.578125,
+      "learning_rate": 1.9726097089525456e-05,
+      "loss": 1.0686,
+      "step": 1399
+    },
+    {
+      "epoch": 0.24004972458580706,
+      "grad_norm": 1.671875,
+      "learning_rate": 1.9725677153209246e-05,
+      "loss": 1.1246,
+      "step": 1400
+    },
+    {
+      "epoch": 0.24004972458580706,
+      "eval_loss": 0.9177221655845642,
+      "eval_runtime": 837.168,
+      "eval_samples_per_second": 2.985,
+      "eval_steps_per_second": 2.985,
+      "step": 1400
+    },
+    {
+      "epoch": 0.24022118867479691,
+      "grad_norm": 1.7578125,
+      "learning_rate": 1.9725256899702085e-05,
+      "loss": 1.0207,
+      "step": 1401
+    },
+    {
+      "epoch": 0.2403926527637868,
+      "grad_norm": 1.6015625,
+      "learning_rate": 1.9724836329017673e-05,
+      "loss": 1.0681,
+      "step": 1402
+    },
+    {
+      "epoch": 0.24056411685277665,
+      "grad_norm": 1.6875,
+      "learning_rate": 1.9724415441169723e-05,
+      "loss": 1.0937,
+      "step": 1403
+    },
+    {
+      "epoch": 0.2407355809417665,
+      "grad_norm": 1.5625,
+      "learning_rate": 1.9723994236171972e-05,
+      "loss": 1.1032,
+      "step": 1404
+    },
+    {
+      "epoch": 0.24090704503075638,
+      "grad_norm": 1.6484375,
+      "learning_rate": 1.972357271403815e-05,
+      "loss": 1.0918,
+      "step": 1405
+    },
+    {
+      "epoch": 0.24107850911974624,
+      "grad_norm": 1.8828125,
+      "learning_rate": 1.972315087478201e-05,
+      "loss": 1.1631,
+      "step": 1406
+    },
+    {
+      "epoch": 0.2412499732087361,
+      "grad_norm": 1.6484375,
+      "learning_rate": 1.97227287184173e-05,
+      "loss": 1.0473,
+      "step": 1407
+    },
+    {
+      "epoch": 0.24142143729772594,
+      "grad_norm": 1.625,
+      "learning_rate": 1.9722306244957797e-05,
+      "loss": 1.0959,
+      "step": 1408
+    },
+    {
+      "epoch": 0.24159290138671582,
+      "grad_norm": 1.578125,
+      "learning_rate": 1.9721883454417278e-05,
+      "loss": 0.9669,
+      "step": 1409
+    },
+    {
+      "epoch": 0.24176436547570568,
+      "grad_norm": 1.5546875,
+      "learning_rate": 1.972146034680953e-05,
+      "loss": 1.083,
+      "step": 1410
+    },
+    {
+      "epoch": 0.24193582956469553,
+      "grad_norm": 1.609375,
+      "learning_rate": 1.9721036922148352e-05,
+      "loss": 1.0486,
+      "step": 1411
+    },
+    {
+      "epoch": 0.2421072936536854,
+      "grad_norm": 1.625,
+      "learning_rate": 1.9720613180447553e-05,
+      "loss": 1.0244,
+      "step": 1412
+    },
+    {
+      "epoch": 0.24227875774267527,
+      "grad_norm": 1.8203125,
+      "learning_rate": 1.9720189121720953e-05,
+      "loss": 1.1736,
+      "step": 1413
+    },
+    {
+      "epoch": 0.24245022183166512,
+      "grad_norm": 1.765625,
+      "learning_rate": 1.9719764745982383e-05,
+      "loss": 1.0677,
+      "step": 1414
+    },
+    {
+      "epoch": 0.242621685920655,
+      "grad_norm": 1.65625,
+      "learning_rate": 1.9719340053245682e-05,
+      "loss": 1.0646,
+      "step": 1415
+    },
+    {
+      "epoch": 0.24279315000964485,
+      "grad_norm": 1.6796875,
+      "learning_rate": 1.9718915043524703e-05,
+      "loss": 1.0764,
+      "step": 1416
+    },
+    {
+      "epoch": 0.2429646140986347,
+      "grad_norm": 1.6328125,
+      "learning_rate": 1.9718489716833308e-05,
+      "loss": 1.0921,
+      "step": 1417
+    },
+    {
+      "epoch": 0.2431360781876246,
+      "grad_norm": 1.59375,
+      "learning_rate": 1.971806407318537e-05,
+      "loss": 1.0519,
+      "step": 1418
+    },
+    {
+      "epoch": 0.24330754227661444,
+      "grad_norm": 1.703125,
+      "learning_rate": 1.9717638112594765e-05,
+      "loss": 1.0683,
+      "step": 1419
+    },
+    {
+      "epoch": 0.2434790063656043,
+      "grad_norm": 1.640625,
+      "learning_rate": 1.9717211835075388e-05,
+      "loss": 1.0414,
+      "step": 1420
+    },
+    {
+      "epoch": 0.24365047045459418,
+      "grad_norm": 1.734375,
+      "learning_rate": 1.971678524064114e-05,
+      "loss": 1.0836,
+      "step": 1421
+    },
+    {
+      "epoch": 0.24382193454358403,
+      "grad_norm": 1.71875,
+      "learning_rate": 1.9716358329305937e-05,
+      "loss": 1.1452,
+      "step": 1422
+    },
+    {
+      "epoch": 0.24399339863257388,
+      "grad_norm": 1.6015625,
+      "learning_rate": 1.9715931101083696e-05,
+      "loss": 1.0282,
+      "step": 1423
+    },
+    {
+      "epoch": 0.24416486272156376,
+      "grad_norm": 1.7421875,
+      "learning_rate": 1.971550355598836e-05,
+      "loss": 1.0765,
+      "step": 1424
+    },
+    {
+      "epoch": 0.24433632681055362,
+      "grad_norm": 1.7109375,
+      "learning_rate": 1.9715075694033866e-05,
+      "loss": 1.0627,
+      "step": 1425
+    },
+    {
+      "epoch": 0.24450779089954347,
+      "grad_norm": 1.671875,
+      "learning_rate": 1.971464751523417e-05,
+      "loss": 1.0239,
+      "step": 1426
+    },
+    {
+      "epoch": 0.24467925498853335,
+      "grad_norm": 1.6953125,
+      "learning_rate": 1.9714219019603236e-05,
+      "loss": 1.1714,
+      "step": 1427
+    },
+    {
+      "epoch": 0.2448507190775232,
+      "grad_norm": 1.5703125,
+      "learning_rate": 1.971379020715504e-05,
+      "loss": 1.015,
+      "step": 1428
+    },
+    {
+      "epoch": 0.24502218316651306,
+      "grad_norm": 1.75,
+      "learning_rate": 1.971336107790357e-05,
+      "loss": 1.0271,
+      "step": 1429
+    },
+    {
+      "epoch": 0.24519364725550294,
+      "grad_norm": 1.59375,
+      "learning_rate": 1.971293163186281e-05,
+      "loss": 1.103,
+      "step": 1430
+    },
+    {
+      "epoch": 0.2453651113444928,
+      "grad_norm": 1.71875,
+      "learning_rate": 1.9712501869046782e-05,
+      "loss": 1.0114,
+      "step": 1431
+    },
+    {
+      "epoch": 0.24553657543348265,
+      "grad_norm": 1.8125,
+      "learning_rate": 1.971207178946949e-05,
+      "loss": 1.1274,
+      "step": 1432
+    },
+    {
+      "epoch": 0.2457080395224725,
+      "grad_norm": 1.671875,
+      "learning_rate": 1.9711641393144967e-05,
+      "loss": 1.0402,
+      "step": 1433
+    },
+    {
+      "epoch": 0.24587950361146238,
+      "grad_norm": 1.59375,
+      "learning_rate": 1.971121068008725e-05,
+      "loss": 1.1112,
+      "step": 1434
+    },
+    {
+      "epoch": 0.24605096770045223,
+      "grad_norm": 1.7265625,
+      "learning_rate": 1.9710779650310376e-05,
+      "loss": 1.1078,
+      "step": 1435
+    },
+    {
+      "epoch": 0.2462224317894421,
+      "grad_norm": 1.6171875,
+      "learning_rate": 1.9710348303828412e-05,
+      "loss": 1.0754,
+      "step": 1436
+    },
+    {
+      "epoch": 0.24639389587843197,
+      "grad_norm": 1.671875,
+      "learning_rate": 1.970991664065543e-05,
+      "loss": 0.9915,
+      "step": 1437
+    },
+    {
+      "epoch": 0.24656535996742182,
+      "grad_norm": 1.703125,
+      "learning_rate": 1.9709484660805498e-05,
+      "loss": 1.0486,
+      "step": 1438
+    },
+    {
+      "epoch": 0.24673682405641167,
+      "grad_norm": 1.6875,
+      "learning_rate": 1.9709052364292705e-05,
+      "loss": 1.1207,
+      "step": 1439
+    },
+    {
+      "epoch": 0.24690828814540156,
+      "grad_norm": 1.5390625,
+      "learning_rate": 1.970861975113116e-05,
+      "loss": 0.9616,
+      "step": 1440
+    },
+    {
+      "epoch": 0.2470797522343914,
+      "grad_norm": 1.5234375,
+      "learning_rate": 1.9708186821334964e-05,
+      "loss": 1.0366,
+      "step": 1441
+    },
+    {
+      "epoch": 0.24725121632338126,
+      "grad_norm": 1.65625,
+      "learning_rate": 1.9707753574918235e-05,
+      "loss": 1.0376,
+      "step": 1442
+    },
+    {
+      "epoch": 0.24742268041237114,
+      "grad_norm": 1.640625,
+      "learning_rate": 1.970732001189511e-05,
+      "loss": 1.0982,
+      "step": 1443
+    },
+    {
+      "epoch": 0.247594144501361,
+      "grad_norm": 1.734375,
+      "learning_rate": 1.9706886132279724e-05,
+      "loss": 1.1314,
+      "step": 1444
+    },
+    {
+      "epoch": 0.24776560859035085,
+      "grad_norm": 2.546875,
+      "learning_rate": 1.9706451936086228e-05,
+      "loss": 0.9716,
+      "step": 1445
+    },
+    {
+      "epoch": 0.24793707267934073,
+      "grad_norm": 1.578125,
+      "learning_rate": 1.9706017423328782e-05,
+      "loss": 0.9978,
+      "step": 1446
+    },
+    {
+      "epoch": 0.24810853676833058,
+      "grad_norm": 1.703125,
+      "learning_rate": 1.970558259402156e-05,
+      "loss": 1.0639,
+      "step": 1447
+    },
+    {
+      "epoch": 0.24828000085732044,
+      "grad_norm": 1.7890625,
+      "learning_rate": 1.970514744817874e-05,
+      "loss": 1.0706,
+      "step": 1448
+    },
+    {
+      "epoch": 0.24845146494631032,
+      "grad_norm": 1.703125,
+      "learning_rate": 1.9704711985814516e-05,
+      "loss": 1.1416,
+      "step": 1449
+    },
+    {
+      "epoch": 0.24862292903530017,
+      "grad_norm": 1.59375,
+      "learning_rate": 1.970427620694309e-05,
+      "loss": 1.155,
+      "step": 1450
+    },
+    {
+      "epoch": 0.24879439312429003,
+      "grad_norm": 1.5625,
+      "learning_rate": 1.9703840111578673e-05,
+      "loss": 1.0571,
+      "step": 1451
+    },
+    {
+      "epoch": 0.2489658572132799,
+      "grad_norm": 1.65625,
+      "learning_rate": 1.970340369973549e-05,
+      "loss": 1.0132,
+      "step": 1452
+    },
+    {
+      "epoch": 0.24913732130226976,
+      "grad_norm": 1.6328125,
+      "learning_rate": 1.9702966971427773e-05,
+      "loss": 1.0421,
+      "step": 1453
+    },
+    {
+      "epoch": 0.2493087853912596,
+      "grad_norm": 1.5625,
+      "learning_rate": 1.9702529926669768e-05,
+      "loss": 1.0326,
+      "step": 1454
+    },
+    {
+      "epoch": 0.2494802494802495,
+      "grad_norm": 1.6796875,
+      "learning_rate": 1.970209256547572e-05,
+      "loss": 1.0629,
+      "step": 1455
+    },
+    {
+      "epoch": 0.24965171356923935,
+      "grad_norm": 1.6875,
+      "learning_rate": 1.9701654887859904e-05,
+      "loss": 1.0527,
+      "step": 1456
+    },
+    {
+      "epoch": 0.2498231776582292,
+      "grad_norm": 1.6875,
+      "learning_rate": 1.9701216893836584e-05,
+      "loss": 1.0612,
+      "step": 1457
+    },
+    {
+      "epoch": 0.24999464174721905,
+      "grad_norm": 1.578125,
+      "learning_rate": 1.970077858342005e-05,
+      "loss": 0.9602,
+      "step": 1458
+    },
+    {
+      "epoch": 0.2501661058362089,
+      "grad_norm": 1.59375,
+      "learning_rate": 1.9700339956624603e-05,
+      "loss": 0.9985,
+      "step": 1459
+    },
+    {
+      "epoch": 0.2503375699251988,
+      "grad_norm": 1.5625,
+      "learning_rate": 1.9699901013464534e-05,
+      "loss": 0.9584,
+      "step": 1460
+    },
+    {
+      "epoch": 0.25050903401418867,
+      "grad_norm": 1.625,
+      "learning_rate": 1.969946175395417e-05,
+      "loss": 0.9766,
+      "step": 1461
+    },
+    {
+      "epoch": 0.2506804981031785,
+      "grad_norm": 1.6875,
+      "learning_rate": 1.9699022178107833e-05,
+      "loss": 1.0457,
+      "step": 1462
+    },
+    {
+      "epoch": 0.2508519621921684,
+      "grad_norm": 1.625,
+      "learning_rate": 1.969858228593986e-05,
+      "loss": 1.0036,
+      "step": 1463
+    },
+    {
+      "epoch": 0.25102342628115826,
+      "grad_norm": 1.6484375,
+      "learning_rate": 1.9698142077464597e-05,
+      "loss": 1.0776,
+      "step": 1464
+    },
+    {
+      "epoch": 0.2511948903701481,
+      "grad_norm": 1.6484375,
+      "learning_rate": 1.96977015526964e-05,
+      "loss": 1.1579,
+      "step": 1465
+    },
+    {
+      "epoch": 0.25136635445913796,
+      "grad_norm": 1.7109375,
+      "learning_rate": 1.969726071164964e-05,
+      "loss": 1.0759,
+      "step": 1466
+    },
+    {
+      "epoch": 0.25153781854812785,
+      "grad_norm": 1.6015625,
+      "learning_rate": 1.9696819554338693e-05,
+      "loss": 1.0797,
+      "step": 1467
+    },
+    {
+      "epoch": 0.25170928263711767,
+      "grad_norm": 1.5859375,
+      "learning_rate": 1.9696378080777937e-05,
+      "loss": 1.0332,
+      "step": 1468
+    },
+    {
+      "epoch": 0.25188074672610755,
+      "grad_norm": 1.6328125,
+      "learning_rate": 1.9695936290981788e-05,
+      "loss": 1.0922,
+      "step": 1469
+    },
+    {
+      "epoch": 0.25205221081509743,
+      "grad_norm": 1.703125,
+      "learning_rate": 1.969549418496464e-05,
+      "loss": 1.0826,
+      "step": 1470
+    },
+    {
+      "epoch": 0.25222367490408726,
+      "grad_norm": 1.7421875,
+      "learning_rate": 1.9695051762740917e-05,
+      "loss": 1.1504,
+      "step": 1471
+    },
+    {
+      "epoch": 0.25239513899307714,
+      "grad_norm": 1.6796875,
+      "learning_rate": 1.969460902432505e-05,
+      "loss": 1.048,
+      "step": 1472
+    },
+    {
+      "epoch": 0.252566603082067,
+      "grad_norm": 1.625,
+      "learning_rate": 1.9694165969731473e-05,
+      "loss": 1.0838,
+      "step": 1473
+    },
+    {
+      "epoch": 0.25273806717105685,
+      "grad_norm": 1.671875,
+      "learning_rate": 1.969372259897464e-05,
+      "loss": 1.0886,
+      "step": 1474
+    },
+    {
+      "epoch": 0.2529095312600467,
+      "grad_norm": 1.6796875,
+      "learning_rate": 1.969327891206901e-05,
+      "loss": 1.0252,
+      "step": 1475
+    },
+    {
+      "epoch": 0.2530809953490366,
+      "grad_norm": 1.6796875,
+      "learning_rate": 1.9692834909029056e-05,
+      "loss": 1.0893,
+      "step": 1476
+    },
+    {
+      "epoch": 0.25325245943802643,
+      "grad_norm": 1.8671875,
+      "learning_rate": 1.9692390589869256e-05,
+      "loss": 1.0419,
+      "step": 1477
+    },
+    {
+      "epoch": 0.2534239235270163,
+      "grad_norm": 1.6796875,
+      "learning_rate": 1.9691945954604095e-05,
+      "loss": 1.0386,
+      "step": 1478
+    },
+    {
+      "epoch": 0.2535953876160062,
+      "grad_norm": 1.6171875,
+      "learning_rate": 1.9691501003248086e-05,
+      "loss": 0.976,
+      "step": 1479
+    },
+    {
+      "epoch": 0.253766851704996,
+      "grad_norm": 1.7109375,
+      "learning_rate": 1.969105573581573e-05,
+      "loss": 1.0139,
+      "step": 1480
+    },
+    {
+      "epoch": 0.2539383157939859,
+      "grad_norm": 1.6640625,
+      "learning_rate": 1.9690610152321557e-05,
+      "loss": 0.9874,
+      "step": 1481
+    },
+    {
+      "epoch": 0.2541097798829758,
+      "grad_norm": 1.6640625,
+      "learning_rate": 1.969016425278009e-05,
+      "loss": 1.0118,
+      "step": 1482
+    },
+    {
+      "epoch": 0.2542812439719656,
+      "grad_norm": 1.640625,
+      "learning_rate": 1.9689718037205883e-05,
+      "loss": 1.0416,
+      "step": 1483
+    },
+    {
+      "epoch": 0.2544527080609555,
+      "grad_norm": 1.65625,
+      "learning_rate": 1.9689271505613477e-05,
+      "loss": 1.0712,
+      "step": 1484
+    },
+    {
+      "epoch": 0.25462417214994537,
+      "grad_norm": 1.65625,
+      "learning_rate": 1.9688824658017446e-05,
+      "loss": 1.0663,
+      "step": 1485
+    },
+    {
+      "epoch": 0.2547956362389352,
+      "grad_norm": 1.734375,
+      "learning_rate": 1.968837749443236e-05,
+      "loss": 1.1637,
+      "step": 1486
+    },
+    {
+      "epoch": 0.2549671003279251,
+      "grad_norm": 1.6640625,
+      "learning_rate": 1.9687930014872794e-05,
+      "loss": 1.0739,
+      "step": 1487
+    },
+    {
+      "epoch": 0.25513856441691496,
+      "grad_norm": 1.765625,
+      "learning_rate": 1.9687482219353352e-05,
+      "loss": 1.0067,
+      "step": 1488
+    },
+    {
+      "epoch": 0.2553100285059048,
+      "grad_norm": 1.578125,
+      "learning_rate": 1.9687034107888636e-05,
+      "loss": 1.0853,
+      "step": 1489
+    },
+    {
+      "epoch": 0.25548149259489467,
+      "grad_norm": 1.6484375,
+      "learning_rate": 1.9686585680493257e-05,
+      "loss": 1.1097,
+      "step": 1490
+    },
+    {
+      "epoch": 0.2556529566838845,
+      "grad_norm": 1.609375,
+      "learning_rate": 1.9686136937181847e-05,
+      "loss": 1.0814,
+      "step": 1491
+    },
+    {
+      "epoch": 0.2558244207728744,
+      "grad_norm": 1.6171875,
+      "learning_rate": 1.9685687877969035e-05,
+      "loss": 1.0519,
+      "step": 1492
+    },
+    {
+      "epoch": 0.25599588486186425,
+      "grad_norm": 1.609375,
+      "learning_rate": 1.9685238502869468e-05,
+      "loss": 1.1121,
+      "step": 1493
+    },
+    {
+      "epoch": 0.2561673489508541,
+      "grad_norm": 1.609375,
+      "learning_rate": 1.96847888118978e-05,
+      "loss": 1.0927,
+      "step": 1494
+    },
+    {
+      "epoch": 0.25633881303984396,
+      "grad_norm": 1.6796875,
+      "learning_rate": 1.96843388050687e-05,
+      "loss": 1.0372,
+      "step": 1495
+    },
+    {
+      "epoch": 0.25651027712883384,
+      "grad_norm": 1.6328125,
+      "learning_rate": 1.9683888482396844e-05,
+      "loss": 1.0362,
+      "step": 1496
+    },
+    {
+      "epoch": 0.25668174121782367,
+      "grad_norm": 1.578125,
+      "learning_rate": 1.968343784389692e-05,
+      "loss": 1.0311,
+      "step": 1497
+    },
+    {
+      "epoch": 0.25685320530681355,
+      "grad_norm": 1.5859375,
+      "learning_rate": 1.9682986889583623e-05,
+      "loss": 1.1184,
+      "step": 1498
+    },
+    {
+      "epoch": 0.25702466939580343,
+      "grad_norm": 1.71875,
+      "learning_rate": 1.968253561947166e-05,
+      "loss": 1.1175,
+      "step": 1499
+    },
+    {
+      "epoch": 0.25719613348479325,
+      "grad_norm": 1.6640625,
+      "learning_rate": 1.968208403357575e-05,
+      "loss": 1.0388,
+      "step": 1500
+    },
+    {
+      "epoch": 0.25736759757378314,
+      "grad_norm": 1.5,
+      "learning_rate": 1.968163213191062e-05,
+      "loss": 0.9964,
+      "step": 1501
+    },
+    {
+      "epoch": 0.257539061662773,
+      "grad_norm": 1.65625,
+      "learning_rate": 1.968117991449101e-05,
+      "loss": 1.1443,
+      "step": 1502
+    },
+    {
+      "epoch": 0.25771052575176284,
+      "grad_norm": 1.671875,
+      "learning_rate": 1.9680727381331665e-05,
+      "loss": 1.04,
+      "step": 1503
+    },
+    {
+      "epoch": 0.2578819898407527,
+      "grad_norm": 1.6328125,
+      "learning_rate": 1.9680274532447344e-05,
+      "loss": 1.0192,
+      "step": 1504
+    },
+    {
+      "epoch": 0.2580534539297426,
+      "grad_norm": 1.609375,
+      "learning_rate": 1.9679821367852824e-05,
+      "loss": 1.1258,
+      "step": 1505
+    },
+    {
+      "epoch": 0.25822491801873243,
+      "grad_norm": 1.5859375,
+      "learning_rate": 1.9679367887562874e-05,
+      "loss": 1.0358,
+      "step": 1506
+    },
+    {
+      "epoch": 0.2583963821077223,
+      "grad_norm": 1.7109375,
+      "learning_rate": 1.967891409159229e-05,
+      "loss": 1.1577,
+      "step": 1507
+    },
+    {
+      "epoch": 0.2585678461967122,
+      "grad_norm": 1.65625,
+      "learning_rate": 1.967845997995587e-05,
+      "loss": 0.9771,
+      "step": 1508
+    },
+    {
+      "epoch": 0.258739310285702,
+      "grad_norm": 1.8984375,
+      "learning_rate": 1.9678005552668423e-05,
+      "loss": 1.1342,
+      "step": 1509
+    },
+    {
+      "epoch": 0.2589107743746919,
+      "grad_norm": 1.671875,
+      "learning_rate": 1.967755080974477e-05,
+      "loss": 1.0732,
+      "step": 1510
+    },
+    {
+      "epoch": 0.2590822384636818,
+      "grad_norm": 1.734375,
+      "learning_rate": 1.9677095751199746e-05,
+      "loss": 1.1263,
+      "step": 1511
+    },
+    {
+      "epoch": 0.2592537025526716,
+      "grad_norm": 1.7109375,
+      "learning_rate": 1.9676640377048185e-05,
+      "loss": 1.0274,
+      "step": 1512
+    },
+    {
+      "epoch": 0.2594251666416615,
+      "grad_norm": 1.6328125,
+      "learning_rate": 1.9676184687304945e-05,
+      "loss": 0.998,
+      "step": 1513
+    },
+    {
+      "epoch": 0.25959663073065137,
+      "grad_norm": 1.578125,
+      "learning_rate": 1.9675728681984886e-05,
+      "loss": 1.0682,
+      "step": 1514
+    },
+    {
+      "epoch": 0.2597680948196412,
+      "grad_norm": 1.703125,
+      "learning_rate": 1.9675272361102876e-05,
+      "loss": 1.1237,
+      "step": 1515
+    },
+    {
+      "epoch": 0.2599395589086311,
+      "grad_norm": 1.7578125,
+      "learning_rate": 1.9674815724673804e-05,
+      "loss": 1.0046,
+      "step": 1516
+    },
+    {
+      "epoch": 0.26011102299762096,
+      "grad_norm": 1.6875,
+      "learning_rate": 1.967435877271256e-05,
+      "loss": 1.142,
+      "step": 1517
+    },
+    {
+      "epoch": 0.2602824870866108,
+      "grad_norm": 1.5859375,
+      "learning_rate": 1.9673901505234042e-05,
+      "loss": 0.9946,
+      "step": 1518
+    },
+    {
+      "epoch": 0.26045395117560066,
+      "grad_norm": 1.671875,
+      "learning_rate": 1.967344392225317e-05,
+      "loss": 1.0523,
+      "step": 1519
+    },
+    {
+      "epoch": 0.26062541526459054,
+      "grad_norm": 1.765625,
+      "learning_rate": 1.9672986023784863e-05,
+      "loss": 1.1062,
+      "step": 1520
+    },
+    {
+      "epoch": 0.26079687935358037,
+      "grad_norm": 1.59375,
+      "learning_rate": 1.967252780984406e-05,
+      "loss": 1.051,
+      "step": 1521
+    },
+    {
+      "epoch": 0.26096834344257025,
+      "grad_norm": 1.6953125,
+      "learning_rate": 1.9672069280445696e-05,
+      "loss": 1.0351,
+      "step": 1522
+    },
+    {
+      "epoch": 0.26113980753156013,
+      "grad_norm": 1.7578125,
+      "learning_rate": 1.967161043560474e-05,
+      "loss": 1.0894,
+      "step": 1523
+    },
+    {
+      "epoch": 0.26131127162054996,
+      "grad_norm": 1.6484375,
+      "learning_rate": 1.967115127533614e-05,
+      "loss": 1.0919,
+      "step": 1524
+    },
+    {
+      "epoch": 0.26148273570953984,
+      "grad_norm": 1.7421875,
+      "learning_rate": 1.967069179965488e-05,
+      "loss": 1.0772,
+      "step": 1525
+    },
+    {
+      "epoch": 0.2616541997985297,
+      "grad_norm": 1.609375,
+      "learning_rate": 1.9670232008575945e-05,
+      "loss": 0.9477,
+      "step": 1526
+    },
+    {
+      "epoch": 0.26182566388751954,
+      "grad_norm": 1.6484375,
+      "learning_rate": 1.966977190211433e-05,
+      "loss": 1.0974,
+      "step": 1527
+    },
+    {
+      "epoch": 0.2619971279765094,
+      "grad_norm": 1.625,
+      "learning_rate": 1.966931148028504e-05,
+      "loss": 1.0565,
+      "step": 1528
+    },
+    {
+      "epoch": 0.2621685920654993,
+      "grad_norm": 1.671875,
+      "learning_rate": 1.9668850743103093e-05,
+      "loss": 1.0793,
+      "step": 1529
+    },
+    {
+      "epoch": 0.26234005615448913,
+      "grad_norm": 1.8359375,
+      "learning_rate": 1.9668389690583512e-05,
+      "loss": 1.0805,
+      "step": 1530
+    },
+    {
+      "epoch": 0.262511520243479,
+      "grad_norm": 1.546875,
+      "learning_rate": 1.9667928322741337e-05,
+      "loss": 1.0825,
+      "step": 1531
+    },
+    {
+      "epoch": 0.2626829843324689,
+      "grad_norm": 1.59375,
+      "learning_rate": 1.9667466639591612e-05,
+      "loss": 1.052,
+      "step": 1532
+    },
+    {
+      "epoch": 0.2628544484214587,
+      "grad_norm": 1.6328125,
+      "learning_rate": 1.9667004641149393e-05,
+      "loss": 1.095,
+      "step": 1533
+    },
+    {
+      "epoch": 0.2630259125104486,
+      "grad_norm": 1.671875,
+      "learning_rate": 1.9666542327429754e-05,
+      "loss": 0.9686,
+      "step": 1534
+    },
+    {
+      "epoch": 0.2631973765994385,
+      "grad_norm": 1.6171875,
+      "learning_rate": 1.966607969844777e-05,
+      "loss": 1.0208,
+      "step": 1535
+    },
+    {
+      "epoch": 0.2633688406884283,
+      "grad_norm": 1.671875,
+      "learning_rate": 1.9665616754218523e-05,
+      "loss": 1.0695,
+      "step": 1536
+    },
+    {
+      "epoch": 0.2635403047774182,
+      "grad_norm": 1.5625,
+      "learning_rate": 1.966515349475712e-05,
+      "loss": 1.0443,
+      "step": 1537
+    },
+    {
+      "epoch": 0.263711768866408,
+      "grad_norm": 1.59375,
+      "learning_rate": 1.9664689920078665e-05,
+      "loss": 1.0577,
+      "step": 1538
+    },
+    {
+      "epoch": 0.2638832329553979,
+      "grad_norm": 1.640625,
+      "learning_rate": 1.9664226030198278e-05,
+      "loss": 1.0162,
+      "step": 1539
+    },
+    {
+      "epoch": 0.2640546970443878,
+      "grad_norm": 1.6875,
+      "learning_rate": 1.9663761825131087e-05,
+      "loss": 1.0553,
+      "step": 1540
+    },
+    {
+      "epoch": 0.2642261611333776,
+      "grad_norm": 1.6171875,
+      "learning_rate": 1.966329730489223e-05,
+      "loss": 1.0822,
+      "step": 1541
+    },
+    {
+      "epoch": 0.2643976252223675,
+      "grad_norm": 1.6328125,
+      "learning_rate": 1.9662832469496863e-05,
+      "loss": 0.9951,
+      "step": 1542
+    },
+    {
+      "epoch": 0.26456908931135736,
+      "grad_norm": 1.609375,
+      "learning_rate": 1.9662367318960143e-05,
+      "loss": 1.0579,
+      "step": 1543
+    },
+    {
+      "epoch": 0.2647405534003472,
+      "grad_norm": 1.6015625,
+      "learning_rate": 1.966190185329724e-05,
+      "loss": 1.0021,
+      "step": 1544
+    },
+    {
+      "epoch": 0.26491201748933707,
+      "grad_norm": 1.75,
+      "learning_rate": 1.9661436072523333e-05,
+      "loss": 1.0497,
+      "step": 1545
+    },
+    {
+      "epoch": 0.26508348157832695,
+      "grad_norm": 1.640625,
+      "learning_rate": 1.966096997665361e-05,
+      "loss": 1.0591,
+      "step": 1546
+    },
+    {
+      "epoch": 0.2652549456673168,
+      "grad_norm": 1.796875,
+      "learning_rate": 1.966050356570328e-05,
+      "loss": 1.128,
+      "step": 1547
+    },
+    {
+      "epoch": 0.26542640975630666,
+      "grad_norm": 1.96875,
+      "learning_rate": 1.9660036839687552e-05,
+      "loss": 0.9565,
+      "step": 1548
+    },
+    {
+      "epoch": 0.26559787384529654,
+      "grad_norm": 1.65625,
+      "learning_rate": 1.9659569798621642e-05,
+      "loss": 0.9964,
+      "step": 1549
+    },
+    {
+      "epoch": 0.26576933793428636,
+      "grad_norm": 14.0625,
+      "learning_rate": 1.965910244252079e-05,
+      "loss": 1.0035,
+      "step": 1550
+    },
+    {
+      "epoch": 0.26594080202327625,
+      "grad_norm": 1.921875,
+      "learning_rate": 1.9658634771400235e-05,
+      "loss": 1.0745,
+      "step": 1551
+    },
+    {
+      "epoch": 0.2661122661122661,
+      "grad_norm": 1.8984375,
+      "learning_rate": 1.9658166785275227e-05,
+      "loss": 1.1087,
+      "step": 1552
+    },
+    {
+      "epoch": 0.26628373020125595,
+      "grad_norm": 1.5859375,
+      "learning_rate": 1.9657698484161032e-05,
+      "loss": 0.9657,
+      "step": 1553
+    },
+    {
+      "epoch": 0.26645519429024583,
+      "grad_norm": 1.6484375,
+      "learning_rate": 1.965722986807292e-05,
+      "loss": 1.0533,
+      "step": 1554
+    },
+    {
+      "epoch": 0.2666266583792357,
+      "grad_norm": 1.75,
+      "learning_rate": 1.965676093702618e-05,
+      "loss": 1.0809,
+      "step": 1555
+    },
+    {
+      "epoch": 0.26679812246822554,
+      "grad_norm": 1.734375,
+      "learning_rate": 1.9656291691036098e-05,
+      "loss": 1.0149,
+      "step": 1556
+    },
+    {
+      "epoch": 0.2669695865572154,
+      "grad_norm": 1.640625,
+      "learning_rate": 1.9655822130117985e-05,
+      "loss": 1.1726,
+      "step": 1557
+    },
+    {
+      "epoch": 0.2671410506462053,
+      "grad_norm": 1.703125,
+      "learning_rate": 1.965535225428715e-05,
+      "loss": 1.0171,
+      "step": 1558
+    },
+    {
+      "epoch": 0.26731251473519513,
+      "grad_norm": 1.6015625,
+      "learning_rate": 1.9654882063558918e-05,
+      "loss": 1.0369,
+      "step": 1559
+    },
+    {
+      "epoch": 0.267483978824185,
+      "grad_norm": 1.625,
+      "learning_rate": 1.965441155794863e-05,
+      "loss": 1.0901,
+      "step": 1560
+    },
+    {
+      "epoch": 0.2676554429131749,
+      "grad_norm": 1.578125,
+      "learning_rate": 1.965394073747162e-05,
+      "loss": 1.0468,
+      "step": 1561
+    },
+    {
+      "epoch": 0.2678269070021647,
+      "grad_norm": 1.625,
+      "learning_rate": 1.965346960214325e-05,
+      "loss": 1.0866,
+      "step": 1562
+    },
+    {
+      "epoch": 0.2679983710911546,
+      "grad_norm": 1.6640625,
+      "learning_rate": 1.9652998151978887e-05,
+      "loss": 1.0789,
+      "step": 1563
+    },
+    {
+      "epoch": 0.2681698351801445,
+      "grad_norm": 1.703125,
+      "learning_rate": 1.9652526386993903e-05,
+      "loss": 1.0895,
+      "step": 1564
+    },
+    {
+      "epoch": 0.2683412992691343,
+      "grad_norm": 1.6640625,
+      "learning_rate": 1.9652054307203687e-05,
+      "loss": 1.0682,
+      "step": 1565
+    },
+    {
+      "epoch": 0.2685127633581242,
+      "grad_norm": 1.6796875,
+      "learning_rate": 1.9651581912623633e-05,
+      "loss": 1.0731,
+      "step": 1566
+    },
+    {
+      "epoch": 0.26868422744711407,
+      "grad_norm": 1.796875,
+      "learning_rate": 1.9651109203269147e-05,
+      "loss": 1.0308,
+      "step": 1567
+    },
+    {
+      "epoch": 0.2688556915361039,
+      "grad_norm": 1.6171875,
+      "learning_rate": 1.965063617915565e-05,
+      "loss": 0.9955,
+      "step": 1568
+    },
+    {
+      "epoch": 0.2690271556250938,
+      "grad_norm": 1.625,
+      "learning_rate": 1.9650162840298564e-05,
+      "loss": 1.0101,
+      "step": 1569
+    },
+    {
+      "epoch": 0.26919861971408365,
+      "grad_norm": 1.7890625,
+      "learning_rate": 1.964968918671333e-05,
+      "loss": 0.9961,
+      "step": 1570
+    },
+    {
+      "epoch": 0.2693700838030735,
+      "grad_norm": 1.6875,
+      "learning_rate": 1.9649215218415393e-05,
+      "loss": 1.0691,
+      "step": 1571
+    },
+    {
+      "epoch": 0.26954154789206336,
+      "grad_norm": 1.6171875,
+      "learning_rate": 1.9648740935420212e-05,
+      "loss": 1.0502,
+      "step": 1572
+    },
+    {
+      "epoch": 0.26971301198105324,
+      "grad_norm": 1.6015625,
+      "learning_rate": 1.9648266337743254e-05,
+      "loss": 1.0375,
+      "step": 1573
+    },
+    {
+      "epoch": 0.26988447607004307,
+      "grad_norm": 1.5625,
+      "learning_rate": 1.96477914254e-05,
+      "loss": 1.0687,
+      "step": 1574
+    },
+    {
+      "epoch": 0.27005594015903295,
+      "grad_norm": 1.7421875,
+      "learning_rate": 1.9647316198405943e-05,
+      "loss": 1.1145,
+      "step": 1575
+    },
+    {
+      "epoch": 0.27022740424802283,
+      "grad_norm": 1.59375,
+      "learning_rate": 1.964684065677657e-05,
+      "loss": 0.9817,
+      "step": 1576
+    },
+    {
+      "epoch": 0.27039886833701265,
+      "grad_norm": 1.7890625,
+      "learning_rate": 1.9646364800527396e-05,
+      "loss": 1.0822,
+      "step": 1577
+    },
+    {
+      "epoch": 0.27057033242600254,
+      "grad_norm": 1.6328125,
+      "learning_rate": 1.9645888629673944e-05,
+      "loss": 1.0437,
+      "step": 1578
+    },
+    {
+      "epoch": 0.2707417965149924,
+      "grad_norm": 1.6171875,
+      "learning_rate": 1.964541214423174e-05,
+      "loss": 1.09,
+      "step": 1579
+    },
+    {
+      "epoch": 0.27091326060398224,
+      "grad_norm": 1.75,
+      "learning_rate": 1.9644935344216325e-05,
+      "loss": 1.0711,
+      "step": 1580
+    },
+    {
+      "epoch": 0.2710847246929721,
+      "grad_norm": 1.7265625,
+      "learning_rate": 1.964445822964325e-05,
+      "loss": 1.0824,
+      "step": 1581
+    },
+    {
+      "epoch": 0.271256188781962,
+      "grad_norm": 1.75,
+      "learning_rate": 1.964398080052807e-05,
+      "loss": 1.0594,
+      "step": 1582
+    },
+    {
+      "epoch": 0.27142765287095183,
+      "grad_norm": 1.6796875,
+      "learning_rate": 1.9643503056886364e-05,
+      "loss": 1.1174,
+      "step": 1583
+    },
+    {
+      "epoch": 0.2715991169599417,
+      "grad_norm": 1.6015625,
+      "learning_rate": 1.964302499873371e-05,
+      "loss": 1.0093,
+      "step": 1584
+    },
+    {
+      "epoch": 0.2717705810489316,
+      "grad_norm": 1.6953125,
+      "learning_rate": 1.9642546626085693e-05,
+      "loss": 1.0025,
+      "step": 1585
+    },
+    {
+      "epoch": 0.2719420451379214,
+      "grad_norm": 1.6171875,
+      "learning_rate": 1.9642067938957926e-05,
+      "loss": 1.1346,
+      "step": 1586
+    },
+    {
+      "epoch": 0.2721135092269113,
+      "grad_norm": 1.640625,
+      "learning_rate": 1.964158893736601e-05,
+      "loss": 1.0351,
+      "step": 1587
+    },
+    {
+      "epoch": 0.2722849733159011,
+      "grad_norm": 1.6171875,
+      "learning_rate": 1.9641109621325577e-05,
+      "loss": 1.1297,
+      "step": 1588
+    },
+    {
+      "epoch": 0.272456437404891,
+      "grad_norm": 1.6640625,
+      "learning_rate": 1.9640629990852253e-05,
+      "loss": 1.028,
+      "step": 1589
+    },
+    {
+      "epoch": 0.2726279014938809,
+      "grad_norm": 1.59375,
+      "learning_rate": 1.964015004596168e-05,
+      "loss": 1.0212,
+      "step": 1590
+    },
+    {
+      "epoch": 0.2727993655828707,
+      "grad_norm": 1.7421875,
+      "learning_rate": 1.9639669786669513e-05,
+      "loss": 1.1078,
+      "step": 1591
+    },
+    {
+      "epoch": 0.2729708296718606,
+      "grad_norm": 1.6640625,
+      "learning_rate": 1.9639189212991415e-05,
+      "loss": 1.0638,
+      "step": 1592
+    },
+    {
+      "epoch": 0.2731422937608505,
+      "grad_norm": 1.6953125,
+      "learning_rate": 1.9638708324943056e-05,
+      "loss": 1.0655,
+      "step": 1593
+    },
+    {
+      "epoch": 0.2733137578498403,
+      "grad_norm": 1.6484375,
+      "learning_rate": 1.963822712254013e-05,
+      "loss": 1.0881,
+      "step": 1594
+    },
+    {
+      "epoch": 0.2734852219388302,
+      "grad_norm": 1.65625,
+      "learning_rate": 1.9637745605798316e-05,
+      "loss": 1.2106,
+      "step": 1595
+    },
+    {
+      "epoch": 0.27365668602782006,
+      "grad_norm": 1.5703125,
+      "learning_rate": 1.9637263774733326e-05,
+      "loss": 1.0209,
+      "step": 1596
+    },
+    {
+      "epoch": 0.2738281501168099,
+      "grad_norm": 1.7109375,
+      "learning_rate": 1.963678162936088e-05,
+      "loss": 1.1191,
+      "step": 1597
+    },
+    {
+      "epoch": 0.27399961420579977,
+      "grad_norm": 1.7578125,
+      "learning_rate": 1.963629916969669e-05,
+      "loss": 1.0532,
+      "step": 1598
+    },
+    {
+      "epoch": 0.27417107829478965,
+      "grad_norm": 1.5625,
+      "learning_rate": 1.96358163957565e-05,
+      "loss": 1.0672,
+      "step": 1599
+    },
+    {
+      "epoch": 0.2743425423837795,
+      "grad_norm": 1.703125,
+      "learning_rate": 1.963533330755605e-05,
+      "loss": 1.0651,
+      "step": 1600
+    },
+    {
+      "epoch": 0.27451400647276936,
+      "grad_norm": 1.6328125,
+      "learning_rate": 1.96348499051111e-05,
+      "loss": 1.0525,
+      "step": 1601
+    },
+    {
+      "epoch": 0.27468547056175924,
+      "grad_norm": 1.6796875,
+      "learning_rate": 1.963436618843741e-05,
+      "loss": 0.9884,
+      "step": 1602
+    },
+    {
+      "epoch": 0.27485693465074906,
+      "grad_norm": 1.6796875,
+      "learning_rate": 1.963388215755076e-05,
+      "loss": 1.1011,
+      "step": 1603
+    },
+    {
+      "epoch": 0.27502839873973894,
+      "grad_norm": 1.6953125,
+      "learning_rate": 1.9633397812466938e-05,
+      "loss": 1.076,
+      "step": 1604
+    },
+    {
+      "epoch": 0.2751998628287288,
+      "grad_norm": 1.6875,
+      "learning_rate": 1.9632913153201733e-05,
+      "loss": 1.0313,
+      "step": 1605
+    },
+    {
+      "epoch": 0.27537132691771865,
+      "grad_norm": 1.6328125,
+      "learning_rate": 1.9632428179770958e-05,
+      "loss": 1.0222,
+      "step": 1606
+    },
+    {
+      "epoch": 0.27554279100670853,
+      "grad_norm": 1.90625,
+      "learning_rate": 1.9631942892190428e-05,
+      "loss": 1.0423,
+      "step": 1607
+    },
+    {
+      "epoch": 0.2757142550956984,
+      "grad_norm": 1.6484375,
+      "learning_rate": 1.963145729047597e-05,
+      "loss": 1.1188,
+      "step": 1608
+    },
+    {
+      "epoch": 0.27588571918468824,
+      "grad_norm": 1.59375,
+      "learning_rate": 1.963097137464342e-05,
+      "loss": 0.9623,
+      "step": 1609
+    },
+    {
+      "epoch": 0.2760571832736781,
+      "grad_norm": 1.6875,
+      "learning_rate": 1.9630485144708627e-05,
+      "loss": 1.0379,
+      "step": 1610
+    },
+    {
+      "epoch": 0.276228647362668,
+      "grad_norm": 1.6796875,
+      "learning_rate": 1.962999860068745e-05,
+      "loss": 1.1151,
+      "step": 1611
+    },
+    {
+      "epoch": 0.2764001114516578,
+      "grad_norm": 1.6328125,
+      "learning_rate": 1.9629511742595752e-05,
+      "loss": 1.0464,
+      "step": 1612
+    },
+    {
+      "epoch": 0.2765715755406477,
+      "grad_norm": 1.6953125,
+      "learning_rate": 1.962902457044942e-05,
+      "loss": 1.1254,
+      "step": 1613
+    },
+    {
+      "epoch": 0.2767430396296376,
+      "grad_norm": 1.8203125,
+      "learning_rate": 1.9628537084264333e-05,
+      "loss": 1.1636,
+      "step": 1614
+    },
+    {
+      "epoch": 0.2769145037186274,
+      "grad_norm": 1.6875,
+      "learning_rate": 1.9628049284056395e-05,
+      "loss": 1.053,
+      "step": 1615
+    },
+    {
+      "epoch": 0.2770859678076173,
+      "grad_norm": 1.5859375,
+      "learning_rate": 1.9627561169841512e-05,
+      "loss": 1.0359,
+      "step": 1616
+    },
+    {
+      "epoch": 0.2772574318966072,
+      "grad_norm": 1.6796875,
+      "learning_rate": 1.9627072741635608e-05,
+      "loss": 0.9806,
+      "step": 1617
+    },
+    {
+      "epoch": 0.277428895985597,
+      "grad_norm": 1.515625,
+      "learning_rate": 1.9626583999454608e-05,
+      "loss": 1.066,
+      "step": 1618
+    },
+    {
+      "epoch": 0.2776003600745869,
+      "grad_norm": 1.6953125,
+      "learning_rate": 1.9626094943314452e-05,
+      "loss": 1.0277,
+      "step": 1619
+    },
+    {
+      "epoch": 0.27777182416357676,
+      "grad_norm": 1.6484375,
+      "learning_rate": 1.9625605573231093e-05,
+      "loss": 1.1298,
+      "step": 1620
+    },
+    {
+      "epoch": 0.2779432882525666,
+      "grad_norm": 1.671875,
+      "learning_rate": 1.962511588922049e-05,
+      "loss": 1.0838,
+      "step": 1621
+    },
+    {
+      "epoch": 0.27811475234155647,
+      "grad_norm": 1.578125,
+      "learning_rate": 1.9624625891298615e-05,
+      "loss": 1.0209,
+      "step": 1622
+    },
+    {
+      "epoch": 0.27828621643054635,
+      "grad_norm": 1.7421875,
+      "learning_rate": 1.9624135579481446e-05,
+      "loss": 1.09,
+      "step": 1623
+    },
+    {
+      "epoch": 0.2784576805195362,
+      "grad_norm": 1.609375,
+      "learning_rate": 1.9623644953784974e-05,
+      "loss": 1.0581,
+      "step": 1624
+    },
+    {
+      "epoch": 0.27862914460852606,
+      "grad_norm": 1.515625,
+      "learning_rate": 1.96231540142252e-05,
+      "loss": 1.0417,
+      "step": 1625
+    },
+    {
+      "epoch": 0.27880060869751594,
+      "grad_norm": 1.5703125,
+      "learning_rate": 1.9622662760818136e-05,
+      "loss": 1.0268,
+      "step": 1626
+    },
+    {
+      "epoch": 0.27897207278650576,
+      "grad_norm": 1.578125,
+      "learning_rate": 1.9622171193579806e-05,
+      "loss": 1.0325,
+      "step": 1627
+    },
+    {
+      "epoch": 0.27914353687549565,
+      "grad_norm": 1.6328125,
+      "learning_rate": 1.9621679312526234e-05,
+      "loss": 1.0768,
+      "step": 1628
+    },
+    {
+      "epoch": 0.2793150009644855,
+      "grad_norm": 1.6640625,
+      "learning_rate": 1.962118711767347e-05,
+      "loss": 0.9856,
+      "step": 1629
+    },
+    {
+      "epoch": 0.27948646505347535,
+      "grad_norm": 1.65625,
+      "learning_rate": 1.9620694609037567e-05,
+      "loss": 1.0344,
+      "step": 1630
+    },
+    {
+      "epoch": 0.27965792914246523,
+      "grad_norm": 1.6015625,
+      "learning_rate": 1.9620201786634584e-05,
+      "loss": 1.0485,
+      "step": 1631
+    },
+    {
+      "epoch": 0.2798293932314551,
+      "grad_norm": 1.5390625,
+      "learning_rate": 1.9619708650480595e-05,
+      "loss": 1.0014,
+      "step": 1632
+    },
+    {
+      "epoch": 0.28000085732044494,
+      "grad_norm": 1.6328125,
+      "learning_rate": 1.961921520059168e-05,
+      "loss": 1.0522,
+      "step": 1633
+    },
+    {
+      "epoch": 0.2801723214094348,
+      "grad_norm": 1.640625,
+      "learning_rate": 1.9618721436983935e-05,
+      "loss": 1.036,
+      "step": 1634
+    },
+    {
+      "epoch": 0.28034378549842465,
+      "grad_norm": 1.625,
+      "learning_rate": 1.9618227359673464e-05,
+      "loss": 1.181,
+      "step": 1635
+    },
+    {
+      "epoch": 0.28051524958741453,
+      "grad_norm": 1.6640625,
+      "learning_rate": 1.961773296867638e-05,
+      "loss": 1.013,
+      "step": 1636
+    },
+    {
+      "epoch": 0.2806867136764044,
+      "grad_norm": 1.578125,
+      "learning_rate": 1.9617238264008806e-05,
+      "loss": 1.0042,
+      "step": 1637
+    },
+    {
+      "epoch": 0.28085817776539423,
+      "grad_norm": 1.5859375,
+      "learning_rate": 1.961674324568688e-05,
+      "loss": 0.9712,
+      "step": 1638
+    },
+    {
+      "epoch": 0.2810296418543841,
+      "grad_norm": 1.6875,
+      "learning_rate": 1.9616247913726738e-05,
+      "loss": 1.0074,
+      "step": 1639
+    },
+    {
+      "epoch": 0.281201105943374,
+      "grad_norm": 1.6640625,
+      "learning_rate": 1.9615752268144544e-05,
+      "loss": 1.1121,
+      "step": 1640
+    },
+    {
+      "epoch": 0.2813725700323638,
+      "grad_norm": 1.71875,
+      "learning_rate": 1.9615256308956458e-05,
+      "loss": 1.0317,
+      "step": 1641
+    },
+    {
+      "epoch": 0.2815440341213537,
+      "grad_norm": 1.5546875,
+      "learning_rate": 1.961476003617866e-05,
+      "loss": 1.0479,
+      "step": 1642
+    },
+    {
+      "epoch": 0.2817154982103436,
+      "grad_norm": 1.578125,
+      "learning_rate": 1.9614263449827324e-05,
+      "loss": 1.0782,
+      "step": 1643
+    },
+    {
+      "epoch": 0.2818869622993334,
+      "grad_norm": 1.671875,
+      "learning_rate": 1.9613766549918657e-05,
+      "loss": 1.0909,
+      "step": 1644
+    },
+    {
+      "epoch": 0.2820584263883233,
+      "grad_norm": 1.7109375,
+      "learning_rate": 1.9613269336468862e-05,
+      "loss": 1.0773,
+      "step": 1645
+    },
+    {
+      "epoch": 0.2822298904773132,
+      "grad_norm": 1.6328125,
+      "learning_rate": 1.961277180949415e-05,
+      "loss": 1.0504,
+      "step": 1646
+    },
+    {
+      "epoch": 0.282401354566303,
+      "grad_norm": 1.59375,
+      "learning_rate": 1.9612273969010755e-05,
+      "loss": 1.0147,
+      "step": 1647
+    },
+    {
+      "epoch": 0.2825728186552929,
+      "grad_norm": 1.71875,
+      "learning_rate": 1.9611775815034905e-05,
+      "loss": 1.0605,
+      "step": 1648
+    },
+    {
+      "epoch": 0.28274428274428276,
+      "grad_norm": 1.578125,
+      "learning_rate": 1.9611277347582853e-05,
+      "loss": 1.0812,
+      "step": 1649
+    },
+    {
+      "epoch": 0.2829157468332726,
+      "grad_norm": 1.7421875,
+      "learning_rate": 1.9610778566670858e-05,
+      "loss": 1.0974,
+      "step": 1650
+    },
+    {
+      "epoch": 0.28308721092226247,
+      "grad_norm": 1.6015625,
+      "learning_rate": 1.9610279472315177e-05,
+      "loss": 0.9812,
+      "step": 1651
+    },
+    {
+      "epoch": 0.28325867501125235,
+      "grad_norm": 1.6953125,
+      "learning_rate": 1.9609780064532095e-05,
+      "loss": 1.0965,
+      "step": 1652
+    },
+    {
+      "epoch": 0.2834301391002422,
+      "grad_norm": 1.6484375,
+      "learning_rate": 1.96092803433379e-05,
+      "loss": 1.0054,
+      "step": 1653
+    },
+    {
+      "epoch": 0.28360160318923205,
+      "grad_norm": 1.6875,
+      "learning_rate": 1.9608780308748886e-05,
+      "loss": 1.0366,
+      "step": 1654
+    },
+    {
+      "epoch": 0.28377306727822194,
+      "grad_norm": 1.6328125,
+      "learning_rate": 1.9608279960781363e-05,
+      "loss": 1.0656,
+      "step": 1655
+    },
+    {
+      "epoch": 0.28394453136721176,
+      "grad_norm": 1.6015625,
+      "learning_rate": 1.960777929945165e-05,
+      "loss": 1.0081,
+      "step": 1656
+    },
+    {
+      "epoch": 0.28411599545620164,
+      "grad_norm": 1.703125,
+      "learning_rate": 1.9607278324776072e-05,
+      "loss": 1.1173,
+      "step": 1657
+    },
+    {
+      "epoch": 0.2842874595451915,
+      "grad_norm": 1.640625,
+      "learning_rate": 1.9606777036770978e-05,
+      "loss": 1.0612,
+      "step": 1658
+    },
+    {
+      "epoch": 0.28445892363418135,
+      "grad_norm": 1.640625,
+      "learning_rate": 1.96062754354527e-05,
+      "loss": 0.9736,
+      "step": 1659
+    },
+    {
+      "epoch": 0.28463038772317123,
+      "grad_norm": 1.6484375,
+      "learning_rate": 1.960577352083761e-05,
+      "loss": 1.1171,
+      "step": 1660
+    },
+    {
+      "epoch": 0.2848018518121611,
+      "grad_norm": 1.6796875,
+      "learning_rate": 1.9605271292942073e-05,
+      "loss": 1.0382,
+      "step": 1661
+    },
+    {
+      "epoch": 0.28497331590115094,
+      "grad_norm": 1.640625,
+      "learning_rate": 1.9604768751782468e-05,
+      "loss": 1.0854,
+      "step": 1662
+    },
+    {
+      "epoch": 0.2851447799901408,
+      "grad_norm": 1.640625,
+      "learning_rate": 1.9604265897375187e-05,
+      "loss": 1.0552,
+      "step": 1663
+    },
+    {
+      "epoch": 0.2853162440791307,
+      "grad_norm": 1.625,
+      "learning_rate": 1.960376272973663e-05,
+      "loss": 1.076,
+      "step": 1664
+    },
+    {
+      "epoch": 0.2854877081681205,
+      "grad_norm": 1.53125,
+      "learning_rate": 1.9603259248883203e-05,
+      "loss": 0.9763,
+      "step": 1665
+    },
+    {
+      "epoch": 0.2856591722571104,
+      "grad_norm": 1.6640625,
+      "learning_rate": 1.9602755454831334e-05,
+      "loss": 1.0594,
+      "step": 1666
+    },
+    {
+      "epoch": 0.2858306363461003,
+      "grad_norm": 1.6796875,
+      "learning_rate": 1.9602251347597442e-05,
+      "loss": 0.9935,
+      "step": 1667
+    },
+    {
+      "epoch": 0.2860021004350901,
+      "grad_norm": 1.71875,
+      "learning_rate": 1.960174692719798e-05,
+      "loss": 1.0511,
+      "step": 1668
+    },
+    {
+      "epoch": 0.28617356452408,
+      "grad_norm": 1.640625,
+      "learning_rate": 1.9601242193649394e-05,
+      "loss": 1.0227,
+      "step": 1669
+    },
+    {
+      "epoch": 0.2863450286130699,
+      "grad_norm": 1.65625,
+      "learning_rate": 1.9600737146968143e-05,
+      "loss": 1.0802,
+      "step": 1670
+    },
+    {
+      "epoch": 0.2865164927020597,
+      "grad_norm": 1.609375,
+      "learning_rate": 1.9600231787170704e-05,
+      "loss": 1.0771,
+      "step": 1671
+    },
+    {
+      "epoch": 0.2866879567910496,
+      "grad_norm": 1.6171875,
+      "learning_rate": 1.959972611427355e-05,
+      "loss": 1.0267,
+      "step": 1672
+    },
+    {
+      "epoch": 0.28685942088003946,
+      "grad_norm": 1.6640625,
+      "learning_rate": 1.959922012829318e-05,
+      "loss": 1.0999,
+      "step": 1673
+    },
+    {
+      "epoch": 0.2870308849690293,
+      "grad_norm": 1.6484375,
+      "learning_rate": 1.9598713829246097e-05,
+      "loss": 0.9944,
+      "step": 1674
+    },
+    {
+      "epoch": 0.28720234905801917,
+      "grad_norm": 1.65625,
+      "learning_rate": 1.9598207217148806e-05,
+      "loss": 1.0756,
+      "step": 1675
+    },
+    {
+      "epoch": 0.28737381314700905,
+      "grad_norm": 1.6953125,
+      "learning_rate": 1.9597700292017838e-05,
+      "loss": 1.032,
+      "step": 1676
+    },
+    {
+      "epoch": 0.2875452772359989,
+      "grad_norm": 1.6484375,
+      "learning_rate": 1.959719305386972e-05,
+      "loss": 1.0701,
+      "step": 1677
+    },
+    {
+      "epoch": 0.28771674132498876,
+      "grad_norm": 1.5390625,
+      "learning_rate": 1.9596685502720997e-05,
+      "loss": 1.0368,
+      "step": 1678
+    },
+    {
+      "epoch": 0.28788820541397864,
+      "grad_norm": 1.609375,
+      "learning_rate": 1.9596177638588223e-05,
+      "loss": 1.0462,
+      "step": 1679
+    },
+    {
+      "epoch": 0.28805966950296846,
+      "grad_norm": 1.625,
+      "learning_rate": 1.959566946148796e-05,
+      "loss": 1.0831,
+      "step": 1680
+    },
+    {
+      "epoch": 0.28823113359195834,
+      "grad_norm": 1.75,
+      "learning_rate": 1.9595160971436784e-05,
+      "loss": 1.2239,
+      "step": 1681
+    },
+    {
+      "epoch": 0.28840259768094817,
+      "grad_norm": 1.6875,
+      "learning_rate": 1.9594652168451274e-05,
+      "loss": 1.0844,
+      "step": 1682
+    },
+    {
+      "epoch": 0.28857406176993805,
+      "grad_norm": 1.7109375,
+      "learning_rate": 1.9594143052548027e-05,
+      "loss": 1.1673,
+      "step": 1683
+    },
+    {
+      "epoch": 0.28874552585892793,
+      "grad_norm": 1.6171875,
+      "learning_rate": 1.9593633623743646e-05,
+      "loss": 0.9826,
+      "step": 1684
+    },
+    {
+      "epoch": 0.28891698994791776,
+      "grad_norm": 1.6328125,
+      "learning_rate": 1.9593123882054748e-05,
+      "loss": 0.9722,
+      "step": 1685
+    },
+    {
+      "epoch": 0.28908845403690764,
+      "grad_norm": 1.71875,
+      "learning_rate": 1.9592613827497953e-05,
+      "loss": 1.0444,
+      "step": 1686
+    },
+    {
+      "epoch": 0.2892599181258975,
+      "grad_norm": 1.6328125,
+      "learning_rate": 1.9592103460089903e-05,
+      "loss": 1.1105,
+      "step": 1687
+    },
+    {
+      "epoch": 0.28943138221488734,
+      "grad_norm": 1.7265625,
+      "learning_rate": 1.9591592779847234e-05,
+      "loss": 1.0662,
+      "step": 1688
+    },
+    {
+      "epoch": 0.2896028463038772,
+      "grad_norm": 1.6328125,
+      "learning_rate": 1.9591081786786608e-05,
+      "loss": 1.1187,
+      "step": 1689
+    },
+    {
+      "epoch": 0.2897743103928671,
+      "grad_norm": 1.5625,
+      "learning_rate": 1.959057048092469e-05,
+      "loss": 1.0104,
+      "step": 1690
+    },
+    {
+      "epoch": 0.28994577448185693,
+      "grad_norm": 1.6328125,
+      "learning_rate": 1.9590058862278154e-05,
+      "loss": 1.0834,
+      "step": 1691
+    },
+    {
+      "epoch": 0.2901172385708468,
+      "grad_norm": 1.6796875,
+      "learning_rate": 1.9589546930863685e-05,
+      "loss": 1.0711,
+      "step": 1692
+    },
+    {
+      "epoch": 0.2902887026598367,
+      "grad_norm": 1.5625,
+      "learning_rate": 1.9589034686697977e-05,
+      "loss": 1.0707,
+      "step": 1693
+    },
+    {
+      "epoch": 0.2904601667488265,
+      "grad_norm": 1.5390625,
+      "learning_rate": 1.9588522129797744e-05,
+      "loss": 1.0679,
+      "step": 1694
+    },
+    {
+      "epoch": 0.2906316308378164,
+      "grad_norm": 1.7890625,
+      "learning_rate": 1.9588009260179693e-05,
+      "loss": 1.1389,
+      "step": 1695
+    },
+    {
+      "epoch": 0.2908030949268063,
+      "grad_norm": 1.59375,
+      "learning_rate": 1.9587496077860553e-05,
+      "loss": 0.9303,
+      "step": 1696
+    },
+    {
+      "epoch": 0.2909745590157961,
+      "grad_norm": 1.609375,
+      "learning_rate": 1.9586982582857067e-05,
+      "loss": 1.0341,
+      "step": 1697
+    },
+    {
+      "epoch": 0.291146023104786,
+      "grad_norm": 1.6640625,
+      "learning_rate": 1.9586468775185975e-05,
+      "loss": 1.0564,
+      "step": 1698
+    },
+    {
+      "epoch": 0.29131748719377587,
+      "grad_norm": 1.5703125,
+      "learning_rate": 1.958595465486404e-05,
+      "loss": 1.0828,
+      "step": 1699
+    },
+    {
+      "epoch": 0.2914889512827657,
+      "grad_norm": 1.703125,
+      "learning_rate": 1.9585440221908026e-05,
+      "loss": 1.141,
+      "step": 1700
+    },
+    {
+      "epoch": 0.2916604153717556,
+      "grad_norm": 1.625,
+      "learning_rate": 1.958492547633471e-05,
+      "loss": 1.0198,
+      "step": 1701
+    },
+    {
+      "epoch": 0.29183187946074546,
+      "grad_norm": 1.78125,
+      "learning_rate": 1.9584410418160876e-05,
+      "loss": 1.1171,
+      "step": 1702
+    },
+    {
+      "epoch": 0.2920033435497353,
+      "grad_norm": 1.6171875,
+      "learning_rate": 1.9583895047403335e-05,
+      "loss": 1.064,
+      "step": 1703
+    },
+    {
+      "epoch": 0.29217480763872516,
+      "grad_norm": 1.8203125,
+      "learning_rate": 1.958337936407888e-05,
+      "loss": 1.0585,
+      "step": 1704
+    },
+    {
+      "epoch": 0.29234627172771505,
+      "grad_norm": 1.640625,
+      "learning_rate": 1.9582863368204342e-05,
+      "loss": 1.055,
+      "step": 1705
+    },
+    {
+      "epoch": 0.29251773581670487,
+      "grad_norm": 1.671875,
+      "learning_rate": 1.958234705979654e-05,
+      "loss": 1.0747,
+      "step": 1706
+    },
+    {
+      "epoch": 0.29268919990569475,
+      "grad_norm": 1.578125,
+      "learning_rate": 1.958183043887232e-05,
+      "loss": 1.0259,
+      "step": 1707
+    },
+    {
+      "epoch": 0.29286066399468463,
+      "grad_norm": 1.5859375,
+      "learning_rate": 1.958131350544852e-05,
+      "loss": 1.0217,
+      "step": 1708
+    },
+    {
+      "epoch": 0.29303212808367446,
+      "grad_norm": 1.6171875,
+      "learning_rate": 1.9580796259542018e-05,
+      "loss": 1.0129,
+      "step": 1709
+    },
+    {
+      "epoch": 0.29320359217266434,
+      "grad_norm": 1.5546875,
+      "learning_rate": 1.958027870116966e-05,
+      "loss": 0.9592,
+      "step": 1710
+    },
+    {
+      "epoch": 0.2933750562616542,
+      "grad_norm": 1.5859375,
+      "learning_rate": 1.9579760830348345e-05,
+      "loss": 1.1031,
+      "step": 1711
+    },
+    {
+      "epoch": 0.29354652035064405,
+      "grad_norm": 1.7109375,
+      "learning_rate": 1.9579242647094956e-05,
+      "loss": 1.117,
+      "step": 1712
+    },
+    {
+      "epoch": 0.29371798443963393,
+      "grad_norm": 1.796875,
+      "learning_rate": 1.957872415142639e-05,
+      "loss": 1.029,
+      "step": 1713
+    },
+    {
+      "epoch": 0.2938894485286238,
+      "grad_norm": 1.6171875,
+      "learning_rate": 1.957820534335956e-05,
+      "loss": 1.0084,
+      "step": 1714
+    },
+    {
+      "epoch": 0.29406091261761363,
+      "grad_norm": 1.640625,
+      "learning_rate": 1.9577686222911386e-05,
+      "loss": 1.0399,
+      "step": 1715
+    },
+    {
+      "epoch": 0.2942323767066035,
+      "grad_norm": 1.578125,
+      "learning_rate": 1.9577166790098797e-05,
+      "loss": 1.0133,
+      "step": 1716
+    },
+    {
+      "epoch": 0.2944038407955934,
+      "grad_norm": 1.6875,
+      "learning_rate": 1.9576647044938733e-05,
+      "loss": 1.1458,
+      "step": 1717
+    },
+    {
+      "epoch": 0.2945753048845832,
+      "grad_norm": 1.703125,
+      "learning_rate": 1.957612698744815e-05,
+      "loss": 1.1054,
+      "step": 1718
+    },
+    {
+      "epoch": 0.2947467689735731,
+      "grad_norm": 1.6640625,
+      "learning_rate": 1.9575606617644e-05,
+      "loss": 1.0095,
+      "step": 1719
+    },
+    {
+      "epoch": 0.294918233062563,
+      "grad_norm": 1.6796875,
+      "learning_rate": 1.9575085935543266e-05,
+      "loss": 1.0516,
+      "step": 1720
+    },
+    {
+      "epoch": 0.2950896971515528,
+      "grad_norm": 1.65625,
+      "learning_rate": 1.9574564941162918e-05,
+      "loss": 1.0511,
+      "step": 1721
+    },
+    {
+      "epoch": 0.2952611612405427,
+      "grad_norm": 1.671875,
+      "learning_rate": 1.9574043634519957e-05,
+      "loss": 1.071,
+      "step": 1722
+    },
+    {
+      "epoch": 0.29543262532953257,
+      "grad_norm": 1.6796875,
+      "learning_rate": 1.9573522015631378e-05,
+      "loss": 1.1122,
+      "step": 1723
+    },
+    {
+      "epoch": 0.2956040894185224,
+      "grad_norm": 1.6328125,
+      "learning_rate": 1.9573000084514197e-05,
+      "loss": 1.0394,
+      "step": 1724
+    },
+    {
+      "epoch": 0.2957755535075123,
+      "grad_norm": 1.703125,
+      "learning_rate": 1.9572477841185435e-05,
+      "loss": 1.1326,
+      "step": 1725
+    },
+    {
+      "epoch": 0.29594701759650216,
+      "grad_norm": 1.5859375,
+      "learning_rate": 1.9571955285662124e-05,
+      "loss": 1.0425,
+      "step": 1726
+    },
+    {
+      "epoch": 0.296118481685492,
+      "grad_norm": 1.6484375,
+      "learning_rate": 1.9571432417961308e-05,
+      "loss": 1.0776,
+      "step": 1727
+    },
+    {
+      "epoch": 0.29628994577448187,
+      "grad_norm": 1.546875,
+      "learning_rate": 1.9570909238100034e-05,
+      "loss": 1.0177,
+      "step": 1728
+    },
+    {
+      "epoch": 0.2964614098634717,
+      "grad_norm": 1.6328125,
+      "learning_rate": 1.9570385746095372e-05,
+      "loss": 1.0884,
+      "step": 1729
+    },
+    {
+      "epoch": 0.2966328739524616,
+      "grad_norm": 1.640625,
+      "learning_rate": 1.956986194196439e-05,
+      "loss": 1.069,
+      "step": 1730
+    },
+    {
+      "epoch": 0.29680433804145145,
+      "grad_norm": 1.640625,
+      "learning_rate": 1.9569337825724174e-05,
+      "loss": 1.0551,
+      "step": 1731
+    },
+    {
+      "epoch": 0.2969758021304413,
+      "grad_norm": 1.578125,
+      "learning_rate": 1.9568813397391816e-05,
+      "loss": 1.0309,
+      "step": 1732
+    },
+    {
+      "epoch": 0.29714726621943116,
+      "grad_norm": 1.8359375,
+      "learning_rate": 1.9568288656984423e-05,
+      "loss": 1.1874,
+      "step": 1733
+    },
+    {
+      "epoch": 0.29731873030842104,
+      "grad_norm": 1.578125,
+      "learning_rate": 1.9567763604519105e-05,
+      "loss": 1.0737,
+      "step": 1734
+    },
+    {
+      "epoch": 0.29749019439741087,
+      "grad_norm": 1.609375,
+      "learning_rate": 1.9567238240012988e-05,
+      "loss": 1.0948,
+      "step": 1735
+    },
+    {
+      "epoch": 0.29766165848640075,
+      "grad_norm": 1.5703125,
+      "learning_rate": 1.9566712563483203e-05,
+      "loss": 1.0346,
+      "step": 1736
+    },
+    {
+      "epoch": 0.29783312257539063,
+      "grad_norm": 1.609375,
+      "learning_rate": 1.9566186574946894e-05,
+      "loss": 1.0707,
+      "step": 1737
+    },
+    {
+      "epoch": 0.29800458666438046,
+      "grad_norm": 1.75,
+      "learning_rate": 1.956566027442122e-05,
+      "loss": 1.1326,
+      "step": 1738
+    },
+    {
+      "epoch": 0.29817605075337034,
+      "grad_norm": 1.734375,
+      "learning_rate": 1.9565133661923346e-05,
+      "loss": 1.1118,
+      "step": 1739
+    },
+    {
+      "epoch": 0.2983475148423602,
+      "grad_norm": 1.6640625,
+      "learning_rate": 1.9564606737470444e-05,
+      "loss": 1.056,
+      "step": 1740
+    },
+    {
+      "epoch": 0.29851897893135004,
+      "grad_norm": 1.546875,
+      "learning_rate": 1.9564079501079698e-05,
+      "loss": 1.0614,
+      "step": 1741
+    },
+    {
+      "epoch": 0.2986904430203399,
+      "grad_norm": 1.5546875,
+      "learning_rate": 1.9563551952768303e-05,
+      "loss": 1.0242,
+      "step": 1742
+    },
+    {
+      "epoch": 0.2988619071093298,
+      "grad_norm": 1.6484375,
+      "learning_rate": 1.9563024092553465e-05,
+      "loss": 0.999,
+      "step": 1743
+    },
+    {
+      "epoch": 0.29903337119831963,
+      "grad_norm": 1.6640625,
+      "learning_rate": 1.9562495920452405e-05,
+      "loss": 1.0743,
+      "step": 1744
+    },
+    {
+      "epoch": 0.2992048352873095,
+      "grad_norm": 1.6796875,
+      "learning_rate": 1.9561967436482342e-05,
+      "loss": 1.1008,
+      "step": 1745
+    },
+    {
+      "epoch": 0.2993762993762994,
+      "grad_norm": 1.7734375,
+      "learning_rate": 1.956143864066051e-05,
+      "loss": 1.0411,
+      "step": 1746
+    },
+    {
+      "epoch": 0.2995477634652892,
+      "grad_norm": 1.609375,
+      "learning_rate": 1.9560909533004168e-05,
+      "loss": 1.0089,
+      "step": 1747
+    },
+    {
+      "epoch": 0.2997192275542791,
+      "grad_norm": 1.6171875,
+      "learning_rate": 1.9560380113530555e-05,
+      "loss": 1.083,
+      "step": 1748
+    },
+    {
+      "epoch": 0.299890691643269,
+      "grad_norm": 1.609375,
+      "learning_rate": 1.955985038225695e-05,
+      "loss": 1.0368,
+      "step": 1749
+    },
+    {
+      "epoch": 0.3000621557322588,
+      "grad_norm": 1.6640625,
+      "learning_rate": 1.955932033920062e-05,
+      "loss": 1.1021,
+      "step": 1750
+    },
+    {
+      "epoch": 0.3002336198212487,
+      "grad_norm": 1.6796875,
+      "learning_rate": 1.955878998437886e-05,
+      "loss": 1.0446,
+      "step": 1751
+    },
+    {
+      "epoch": 0.30040508391023857,
+      "grad_norm": 1.6171875,
+      "learning_rate": 1.9558259317808964e-05,
+      "loss": 1.1077,
+      "step": 1752
+    },
+    {
+      "epoch": 0.3005765479992284,
+      "grad_norm": 1.59375,
+      "learning_rate": 1.9557728339508238e-05,
+      "loss": 1.0784,
+      "step": 1753
+    },
+    {
+      "epoch": 0.3007480120882183,
+      "grad_norm": 1.5234375,
+      "learning_rate": 1.9557197049493997e-05,
+      "loss": 0.9533,
+      "step": 1754
+    },
+    {
+      "epoch": 0.30091947617720816,
+      "grad_norm": 1.6328125,
+      "learning_rate": 1.9556665447783577e-05,
+      "loss": 1.051,
+      "step": 1755
+    },
+    {
+      "epoch": 0.301090940266198,
+      "grad_norm": 1.8125,
+      "learning_rate": 1.9556133534394304e-05,
+      "loss": 1.1247,
+      "step": 1756
+    },
+    {
+      "epoch": 0.30126240435518786,
+      "grad_norm": 1.6328125,
+      "learning_rate": 1.9555601309343536e-05,
+      "loss": 1.1179,
+      "step": 1757
+    },
+    {
+      "epoch": 0.30143386844417774,
+      "grad_norm": 1.609375,
+      "learning_rate": 1.955506877264862e-05,
+      "loss": 1.0374,
+      "step": 1758
+    },
+    {
+      "epoch": 0.30160533253316757,
+      "grad_norm": 1.7578125,
+      "learning_rate": 1.9554535924326936e-05,
+      "loss": 1.1062,
+      "step": 1759
+    },
+    {
+      "epoch": 0.30177679662215745,
+      "grad_norm": 1.5703125,
+      "learning_rate": 1.9554002764395856e-05,
+      "loss": 0.9826,
+      "step": 1760
+    },
+    {
+      "epoch": 0.30194826071114733,
+      "grad_norm": 1.5546875,
+      "learning_rate": 1.9553469292872765e-05,
+      "loss": 0.9971,
+      "step": 1761
+    },
+    {
+      "epoch": 0.30211972480013716,
+      "grad_norm": 1.65625,
+      "learning_rate": 1.955293550977507e-05,
+      "loss": 1.0656,
+      "step": 1762
+    },
+    {
+      "epoch": 0.30229118888912704,
+      "grad_norm": 1.5625,
+      "learning_rate": 1.955240141512017e-05,
+      "loss": 1.0908,
+      "step": 1763
+    },
+    {
+      "epoch": 0.3024626529781169,
+      "grad_norm": 1.6015625,
+      "learning_rate": 1.9551867008925492e-05,
+      "loss": 1.0144,
+      "step": 1764
+    },
+    {
+      "epoch": 0.30263411706710674,
+      "grad_norm": 1.6875,
+      "learning_rate": 1.955133229120846e-05,
+      "loss": 1.0528,
+      "step": 1765
+    },
+    {
+      "epoch": 0.3028055811560966,
+      "grad_norm": 1.671875,
+      "learning_rate": 1.9550797261986516e-05,
+      "loss": 1.0964,
+      "step": 1766
+    },
+    {
+      "epoch": 0.3029770452450865,
+      "grad_norm": 1.546875,
+      "learning_rate": 1.9550261921277108e-05,
+      "loss": 1.0305,
+      "step": 1767
+    },
+    {
+      "epoch": 0.30314850933407633,
+      "grad_norm": 1.625,
+      "learning_rate": 1.9549726269097696e-05,
+      "loss": 1.0677,
+      "step": 1768
+    },
+    {
+      "epoch": 0.3033199734230662,
+      "grad_norm": 1.703125,
+      "learning_rate": 1.9549190305465754e-05,
+      "loss": 1.0878,
+      "step": 1769
+    },
+    {
+      "epoch": 0.3034914375120561,
+      "grad_norm": 1.625,
+      "learning_rate": 1.9548654030398754e-05,
+      "loss": 1.0571,
+      "step": 1770
+    },
+    {
+      "epoch": 0.3036629016010459,
+      "grad_norm": 1.5078125,
+      "learning_rate": 1.954811744391419e-05,
+      "loss": 0.9636,
+      "step": 1771
+    },
+    {
+      "epoch": 0.3038343656900358,
+      "grad_norm": 1.609375,
+      "learning_rate": 1.9547580546029555e-05,
+      "loss": 1.0493,
+      "step": 1772
+    },
+    {
+      "epoch": 0.3040058297790257,
+      "grad_norm": 1.640625,
+      "learning_rate": 1.9547043336762372e-05,
+      "loss": 1.0085,
+      "step": 1773
+    },
+    {
+      "epoch": 0.3041772938680155,
+      "grad_norm": 1.6171875,
+      "learning_rate": 1.9546505816130153e-05,
+      "loss": 1.1213,
+      "step": 1774
+    },
+    {
+      "epoch": 0.3043487579570054,
+      "grad_norm": 1.671875,
+      "learning_rate": 1.954596798415043e-05,
+      "loss": 1.0836,
+      "step": 1775
+    },
+    {
+      "epoch": 0.3045202220459952,
+      "grad_norm": 1.65625,
+      "learning_rate": 1.9545429840840744e-05,
+      "loss": 1.0237,
+      "step": 1776
+    },
+    {
+      "epoch": 0.3046916861349851,
+      "grad_norm": 1.6875,
+      "learning_rate": 1.9544891386218647e-05,
+      "loss": 1.0164,
+      "step": 1777
+    },
+    {
+      "epoch": 0.304863150223975,
+      "grad_norm": 1.6484375,
+      "learning_rate": 1.95443526203017e-05,
+      "loss": 1.037,
+      "step": 1778
+    },
+    {
+      "epoch": 0.3050346143129648,
+      "grad_norm": 1.640625,
+      "learning_rate": 1.9543813543107473e-05,
+      "loss": 1.038,
+      "step": 1779
+    },
+    {
+      "epoch": 0.3052060784019547,
+      "grad_norm": 1.6171875,
+      "learning_rate": 1.9543274154653544e-05,
+      "loss": 1.0393,
+      "step": 1780
+    },
+    {
+      "epoch": 0.30537754249094456,
+      "grad_norm": 1.5859375,
+      "learning_rate": 1.954273445495751e-05,
+      "loss": 1.0726,
+      "step": 1781
+    },
+    {
+      "epoch": 0.3055490065799344,
+      "grad_norm": 1.7734375,
+      "learning_rate": 1.9542194444036976e-05,
+      "loss": 1.157,
+      "step": 1782
+    },
+    {
+      "epoch": 0.30572047066892427,
+      "grad_norm": 1.5859375,
+      "learning_rate": 1.9541654121909543e-05,
+      "loss": 1.0675,
+      "step": 1783
+    },
+    {
+      "epoch": 0.30589193475791415,
+      "grad_norm": 1.625,
+      "learning_rate": 1.954111348859284e-05,
+      "loss": 1.0843,
+      "step": 1784
+    },
+    {
+      "epoch": 0.306063398846904,
+      "grad_norm": 1.7421875,
+      "learning_rate": 1.95405725441045e-05,
+      "loss": 1.1394,
+      "step": 1785
+    },
+    {
+      "epoch": 0.30623486293589386,
+      "grad_norm": 1.6953125,
+      "learning_rate": 1.9540031288462157e-05,
+      "loss": 1.0638,
+      "step": 1786
+    },
+    {
+      "epoch": 0.30640632702488374,
+      "grad_norm": 1.6015625,
+      "learning_rate": 1.9539489721683477e-05,
+      "loss": 1.0697,
+      "step": 1787
+    },
+    {
+      "epoch": 0.30657779111387357,
+      "grad_norm": 1.6484375,
+      "learning_rate": 1.953894784378611e-05,
+      "loss": 1.1335,
+      "step": 1788
+    },
+    {
+      "epoch": 0.30674925520286345,
+      "grad_norm": 1.703125,
+      "learning_rate": 1.9538405654787733e-05,
+      "loss": 1.1439,
+      "step": 1789
+    },
+    {
+      "epoch": 0.30692071929185333,
+      "grad_norm": 1.609375,
+      "learning_rate": 1.953786315470603e-05,
+      "loss": 1.1355,
+      "step": 1790
+    },
+    {
+      "epoch": 0.30709218338084315,
+      "grad_norm": 1.625,
+      "learning_rate": 1.9537320343558696e-05,
+      "loss": 1.0443,
+      "step": 1791
+    },
+    {
+      "epoch": 0.30726364746983303,
+      "grad_norm": 1.6640625,
+      "learning_rate": 1.9536777221363427e-05,
+      "loss": 1.0282,
+      "step": 1792
+    },
+    {
+      "epoch": 0.3074351115588229,
+      "grad_norm": 1.5390625,
+      "learning_rate": 1.9536233788137945e-05,
+      "loss": 1.0803,
+      "step": 1793
+    },
+    {
+      "epoch": 0.30760657564781274,
+      "grad_norm": 1.609375,
+      "learning_rate": 1.9535690043899965e-05,
+      "loss": 1.0731,
+      "step": 1794
+    },
+    {
+      "epoch": 0.3077780397368026,
+      "grad_norm": 1.6484375,
+      "learning_rate": 1.953514598866723e-05,
+      "loss": 1.1285,
+      "step": 1795
+    },
+    {
+      "epoch": 0.3079495038257925,
+      "grad_norm": 1.625,
+      "learning_rate": 1.9534601622457473e-05,
+      "loss": 1.0675,
+      "step": 1796
+    },
+    {
+      "epoch": 0.30812096791478233,
+      "grad_norm": 1.5234375,
+      "learning_rate": 1.9534056945288454e-05,
+      "loss": 0.9412,
+      "step": 1797
+    },
+    {
+      "epoch": 0.3082924320037722,
+      "grad_norm": 1.6171875,
+      "learning_rate": 1.953351195717794e-05,
+      "loss": 1.0785,
+      "step": 1798
+    },
+    {
+      "epoch": 0.3084638960927621,
+      "grad_norm": 1.640625,
+      "learning_rate": 1.9532966658143697e-05,
+      "loss": 1.1037,
+      "step": 1799
+    },
+    {
+      "epoch": 0.3086353601817519,
+      "grad_norm": 1.7265625,
+      "learning_rate": 1.953242104820351e-05,
+      "loss": 1.1528,
+      "step": 1800
+    },
+    {
+      "epoch": 0.3088068242707418,
+      "grad_norm": 1.578125,
+      "learning_rate": 1.9531875127375185e-05,
+      "loss": 1.004,
+      "step": 1801
+    },
+    {
+      "epoch": 0.3089782883597317,
+      "grad_norm": 1.6171875,
+      "learning_rate": 1.9531328895676515e-05,
+      "loss": 0.9947,
+      "step": 1802
+    },
+    {
+      "epoch": 0.3091497524487215,
+      "grad_norm": 1.6796875,
+      "learning_rate": 1.9530782353125315e-05,
+      "loss": 0.9227,
+      "step": 1803
+    },
+    {
+      "epoch": 0.3093212165377114,
+      "grad_norm": 1.71875,
+      "learning_rate": 1.9530235499739417e-05,
+      "loss": 0.987,
+      "step": 1804
+    },
+    {
+      "epoch": 0.30949268062670127,
+      "grad_norm": 1.7265625,
+      "learning_rate": 1.952968833553665e-05,
+      "loss": 1.1483,
+      "step": 1805
+    },
+    {
+      "epoch": 0.3096641447156911,
+      "grad_norm": 1.75,
+      "learning_rate": 1.952914086053486e-05,
+      "loss": 1.0115,
+      "step": 1806
+    },
+    {
+      "epoch": 0.309835608804681,
+      "grad_norm": 1.7265625,
+      "learning_rate": 1.9528593074751903e-05,
+      "loss": 1.1096,
+      "step": 1807
+    },
+    {
+      "epoch": 0.31000707289367085,
+      "grad_norm": 1.578125,
+      "learning_rate": 1.952804497820565e-05,
+      "loss": 1.0568,
+      "step": 1808
+    },
+    {
+      "epoch": 0.3101785369826607,
+      "grad_norm": 1.5859375,
+      "learning_rate": 1.9527496570913964e-05,
+      "loss": 1.1535,
+      "step": 1809
+    },
+    {
+      "epoch": 0.31035000107165056,
+      "grad_norm": 1.6171875,
+      "learning_rate": 1.9526947852894743e-05,
+      "loss": 1.0408,
+      "step": 1810
+    },
+    {
+      "epoch": 0.31052146516064044,
+      "grad_norm": 1.59375,
+      "learning_rate": 1.9526398824165874e-05,
+      "loss": 1.0993,
+      "step": 1811
+    },
+    {
+      "epoch": 0.31069292924963027,
+      "grad_norm": 1.6484375,
+      "learning_rate": 1.9525849484745266e-05,
+      "loss": 1.0322,
+      "step": 1812
+    },
+    {
+      "epoch": 0.31086439333862015,
+      "grad_norm": 1.6171875,
+      "learning_rate": 1.9525299834650838e-05,
+      "loss": 1.0349,
+      "step": 1813
+    },
+    {
+      "epoch": 0.31103585742761003,
+      "grad_norm": 1.6796875,
+      "learning_rate": 1.9524749873900514e-05,
+      "loss": 1.108,
+      "step": 1814
+    },
+    {
+      "epoch": 0.31120732151659986,
+      "grad_norm": 1.703125,
+      "learning_rate": 1.9524199602512227e-05,
+      "loss": 1.1155,
+      "step": 1815
+    },
+    {
+      "epoch": 0.31137878560558974,
+      "grad_norm": 1.65625,
+      "learning_rate": 1.9523649020503925e-05,
+      "loss": 0.9738,
+      "step": 1816
+    },
+    {
+      "epoch": 0.3115502496945796,
+      "grad_norm": 1.65625,
+      "learning_rate": 1.9523098127893566e-05,
+      "loss": 1.0731,
+      "step": 1817
+    },
+    {
+      "epoch": 0.31172171378356944,
+      "grad_norm": 1.671875,
+      "learning_rate": 1.9522546924699117e-05,
+      "loss": 0.9524,
+      "step": 1818
+    },
+    {
+      "epoch": 0.3118931778725593,
+      "grad_norm": 1.7421875,
+      "learning_rate": 1.9521995410938556e-05,
+      "loss": 1.0931,
+      "step": 1819
+    },
+    {
+      "epoch": 0.3120646419615492,
+      "grad_norm": 1.7109375,
+      "learning_rate": 1.9521443586629866e-05,
+      "loss": 1.0354,
+      "step": 1820
+    },
+    {
+      "epoch": 0.31223610605053903,
+      "grad_norm": 1.6171875,
+      "learning_rate": 1.952089145179105e-05,
+      "loss": 1.0416,
+      "step": 1821
+    },
+    {
+      "epoch": 0.3124075701395289,
+      "grad_norm": 1.6015625,
+      "learning_rate": 1.9520339006440107e-05,
+      "loss": 1.0128,
+      "step": 1822
+    },
+    {
+      "epoch": 0.31257903422851874,
+      "grad_norm": 1.578125,
+      "learning_rate": 1.951978625059506e-05,
+      "loss": 1.0744,
+      "step": 1823
+    },
+    {
+      "epoch": 0.3127504983175086,
+      "grad_norm": 1.640625,
+      "learning_rate": 1.9519233184273937e-05,
+      "loss": 1.0984,
+      "step": 1824
+    },
+    {
+      "epoch": 0.3129219624064985,
+      "grad_norm": 1.7265625,
+      "learning_rate": 1.951867980749477e-05,
+      "loss": 1.1042,
+      "step": 1825
+    },
+    {
+      "epoch": 0.3130934264954883,
+      "grad_norm": 1.625,
+      "learning_rate": 1.9518126120275615e-05,
+      "loss": 1.0636,
+      "step": 1826
+    },
+    {
+      "epoch": 0.3132648905844782,
+      "grad_norm": 1.640625,
+      "learning_rate": 1.9517572122634522e-05,
+      "loss": 1.0009,
+      "step": 1827
+    },
+    {
+      "epoch": 0.3134363546734681,
+      "grad_norm": 1.6015625,
+      "learning_rate": 1.9517017814589562e-05,
+      "loss": 1.0826,
+      "step": 1828
+    },
+    {
+      "epoch": 0.3136078187624579,
+      "grad_norm": 1.6328125,
+      "learning_rate": 1.9516463196158818e-05,
+      "loss": 1.0033,
+      "step": 1829
+    },
+    {
+      "epoch": 0.3137792828514478,
+      "grad_norm": 1.546875,
+      "learning_rate": 1.951590826736037e-05,
+      "loss": 0.9962,
+      "step": 1830
+    },
+    {
+      "epoch": 0.3139507469404377,
+      "grad_norm": 1.6796875,
+      "learning_rate": 1.9515353028212317e-05,
+      "loss": 1.0731,
+      "step": 1831
+    },
+    {
+      "epoch": 0.3141222110294275,
+      "grad_norm": 1.6328125,
+      "learning_rate": 1.9514797478732773e-05,
+      "loss": 1.0536,
+      "step": 1832
+    },
+    {
+      "epoch": 0.3142936751184174,
+      "grad_norm": 1.5703125,
+      "learning_rate": 1.9514241618939855e-05,
+      "loss": 0.9774,
+      "step": 1833
+    },
+    {
+      "epoch": 0.31446513920740726,
+      "grad_norm": 1.6640625,
+      "learning_rate": 1.9513685448851688e-05,
+      "loss": 1.0468,
+      "step": 1834
+    },
+    {
+      "epoch": 0.3146366032963971,
+      "grad_norm": 1.5625,
+      "learning_rate": 1.9513128968486414e-05,
+      "loss": 1.0279,
+      "step": 1835
+    },
+    {
+      "epoch": 0.31480806738538697,
+      "grad_norm": 1.5390625,
+      "learning_rate": 1.9512572177862184e-05,
+      "loss": 0.9927,
+      "step": 1836
+    },
+    {
+      "epoch": 0.31497953147437685,
+      "grad_norm": 1.546875,
+      "learning_rate": 1.951201507699715e-05,
+      "loss": 1.0225,
+      "step": 1837
+    },
+    {
+      "epoch": 0.3151509955633667,
+      "grad_norm": 1.6015625,
+      "learning_rate": 1.951145766590949e-05,
+      "loss": 1.0497,
+      "step": 1838
+    },
+    {
+      "epoch": 0.31532245965235656,
+      "grad_norm": 1.59375,
+      "learning_rate": 1.9510899944617377e-05,
+      "loss": 0.9978,
+      "step": 1839
+    },
+    {
+      "epoch": 0.31549392374134644,
+      "grad_norm": 1.671875,
+      "learning_rate": 1.9510341913139e-05,
+      "loss": 1.0166,
+      "step": 1840
+    },
+    {
+      "epoch": 0.31566538783033626,
+      "grad_norm": 1.7265625,
+      "learning_rate": 1.950978357149256e-05,
+      "loss": 1.1047,
+      "step": 1841
+    },
+    {
+      "epoch": 0.31583685191932614,
+      "grad_norm": 1.6328125,
+      "learning_rate": 1.9509224919696274e-05,
+      "loss": 0.9729,
+      "step": 1842
+    },
+    {
+      "epoch": 0.316008316008316,
+      "grad_norm": 1.5859375,
+      "learning_rate": 1.9508665957768345e-05,
+      "loss": 1.0282,
+      "step": 1843
+    },
+    {
+      "epoch": 0.31617978009730585,
+      "grad_norm": 1.609375,
+      "learning_rate": 1.950810668572702e-05,
+      "loss": 1.0229,
+      "step": 1844
+    },
+    {
+      "epoch": 0.31635124418629573,
+      "grad_norm": 1.6953125,
+      "learning_rate": 1.9507547103590528e-05,
+      "loss": 1.1291,
+      "step": 1845
+    },
+    {
+      "epoch": 0.3165227082752856,
+      "grad_norm": 1.609375,
+      "learning_rate": 1.9506987211377125e-05,
+      "loss": 1.1058,
+      "step": 1846
+    },
+    {
+      "epoch": 0.31669417236427544,
+      "grad_norm": 1.6171875,
+      "learning_rate": 1.950642700910507e-05,
+      "loss": 1.0973,
+      "step": 1847
+    },
+    {
+      "epoch": 0.3168656364532653,
+      "grad_norm": 42.25,
+      "learning_rate": 1.950586649679263e-05,
+      "loss": 1.2741,
+      "step": 1848
+    },
+    {
+      "epoch": 0.3170371005422552,
+      "grad_norm": 1.703125,
+      "learning_rate": 1.9505305674458087e-05,
+      "loss": 1.0999,
+      "step": 1849
+    },
+    {
+      "epoch": 0.317208564631245,
+      "grad_norm": 1.609375,
+      "learning_rate": 1.9504744542119735e-05,
+      "loss": 1.004,
+      "step": 1850
+    },
+    {
+      "epoch": 0.3173800287202349,
+      "grad_norm": 1.6171875,
+      "learning_rate": 1.950418309979587e-05,
+      "loss": 1.0514,
+      "step": 1851
+    },
+    {
+      "epoch": 0.3175514928092248,
+      "grad_norm": 1.671875,
+      "learning_rate": 1.9503621347504806e-05,
+      "loss": 1.0184,
+      "step": 1852
+    },
+    {
+      "epoch": 0.3177229568982146,
+      "grad_norm": 1.7265625,
+      "learning_rate": 1.950305928526486e-05,
+      "loss": 1.0488,
+      "step": 1853
+    },
+    {
+      "epoch": 0.3178944209872045,
+      "grad_norm": 1.609375,
+      "learning_rate": 1.950249691309437e-05,
+      "loss": 0.9829,
+      "step": 1854
+    },
+    {
+      "epoch": 0.3180658850761944,
+      "grad_norm": 1.546875,
+      "learning_rate": 1.950193423101167e-05,
+      "loss": 0.9648,
+      "step": 1855
+    },
+    {
+      "epoch": 0.3182373491651842,
+      "grad_norm": 1.6015625,
+      "learning_rate": 1.9501371239035113e-05,
+      "loss": 0.9902,
+      "step": 1856
+    },
+    {
+      "epoch": 0.3184088132541741,
+      "grad_norm": 1.5859375,
+      "learning_rate": 1.950080793718306e-05,
+      "loss": 1.0648,
+      "step": 1857
+    },
+    {
+      "epoch": 0.31858027734316396,
+      "grad_norm": 1.578125,
+      "learning_rate": 1.9500244325473888e-05,
+      "loss": 1.0418,
+      "step": 1858
+    },
+    {
+      "epoch": 0.3187517414321538,
+      "grad_norm": 1.65625,
+      "learning_rate": 1.949968040392597e-05,
+      "loss": 1.0923,
+      "step": 1859
+    },
+    {
+      "epoch": 0.31892320552114367,
+      "grad_norm": 1.671875,
+      "learning_rate": 1.9499116172557703e-05,
+      "loss": 1.1025,
+      "step": 1860
+    },
+    {
+      "epoch": 0.31909466961013355,
+      "grad_norm": 1.6328125,
+      "learning_rate": 1.949855163138749e-05,
+      "loss": 1.0503,
+      "step": 1861
+    },
+    {
+      "epoch": 0.3192661336991234,
+      "grad_norm": 1.6171875,
+      "learning_rate": 1.9497986780433735e-05,
+      "loss": 1.0073,
+      "step": 1862
+    },
+    {
+      "epoch": 0.31943759778811326,
+      "grad_norm": 1.640625,
+      "learning_rate": 1.9497421619714866e-05,
+      "loss": 1.0719,
+      "step": 1863
+    },
+    {
+      "epoch": 0.31960906187710314,
+      "grad_norm": 2.03125,
+      "learning_rate": 1.9496856149249315e-05,
+      "loss": 1.0481,
+      "step": 1864
+    },
+    {
+      "epoch": 0.31978052596609297,
+      "grad_norm": 1.59375,
+      "learning_rate": 1.9496290369055522e-05,
+      "loss": 1.0105,
+      "step": 1865
+    },
+    {
+      "epoch": 0.31995199005508285,
+      "grad_norm": 1.59375,
+      "learning_rate": 1.9495724279151945e-05,
+      "loss": 1.0392,
+      "step": 1866
+    },
+    {
+      "epoch": 0.3201234541440727,
+      "grad_norm": 1.53125,
+      "learning_rate": 1.949515787955704e-05,
+      "loss": 1.0413,
+      "step": 1867
+    },
+    {
+      "epoch": 0.32029491823306255,
+      "grad_norm": 1.703125,
+      "learning_rate": 1.9494591170289276e-05,
+      "loss": 1.0007,
+      "step": 1868
+    },
+    {
+      "epoch": 0.32046638232205243,
+      "grad_norm": 1.59375,
+      "learning_rate": 1.9494024151367145e-05,
+      "loss": 1.0438,
+      "step": 1869
+    },
+    {
+      "epoch": 0.32063784641104226,
+      "grad_norm": 1.796875,
+      "learning_rate": 1.9493456822809135e-05,
+      "loss": 1.0791,
+      "step": 1870
+    },
+    {
+      "epoch": 0.32080931050003214,
+      "grad_norm": 1.65625,
+      "learning_rate": 1.949288918463375e-05,
+      "loss": 1.0887,
+      "step": 1871
+    },
+    {
+      "epoch": 0.320980774589022,
+      "grad_norm": 1.9140625,
+      "learning_rate": 1.9492321236859496e-05,
+      "loss": 1.0292,
+      "step": 1872
+    },
+    {
+      "epoch": 0.32115223867801185,
+      "grad_norm": 1.6171875,
+      "learning_rate": 1.9491752979504906e-05,
+      "loss": 1.0404,
+      "step": 1873
+    },
+    {
+      "epoch": 0.32132370276700173,
+      "grad_norm": 1.671875,
+      "learning_rate": 1.949118441258851e-05,
+      "loss": 1.0434,
+      "step": 1874
+    },
+    {
+      "epoch": 0.3214951668559916,
+      "grad_norm": 1.59375,
+      "learning_rate": 1.9490615536128847e-05,
+      "loss": 0.9624,
+      "step": 1875
+    },
+    {
+      "epoch": 0.32166663094498144,
+      "grad_norm": 1.6796875,
+      "learning_rate": 1.9490046350144475e-05,
+      "loss": 1.1028,
+      "step": 1876
+    },
+    {
+      "epoch": 0.3218380950339713,
+      "grad_norm": 1.6640625,
+      "learning_rate": 1.948947685465395e-05,
+      "loss": 1.1315,
+      "step": 1877
+    },
+    {
+      "epoch": 0.3220095591229612,
+      "grad_norm": 1.5625,
+      "learning_rate": 1.9488907049675858e-05,
+      "loss": 1.0452,
+      "step": 1878
+    },
+    {
+      "epoch": 0.322181023211951,
+      "grad_norm": 1.6640625,
+      "learning_rate": 1.9488336935228772e-05,
+      "loss": 1.1242,
+      "step": 1879
+    },
+    {
+      "epoch": 0.3223524873009409,
+      "grad_norm": 1.578125,
+      "learning_rate": 1.948776651133129e-05,
+      "loss": 1.0056,
+      "step": 1880
+    },
+    {
+      "epoch": 0.3225239513899308,
+      "grad_norm": 1.765625,
+      "learning_rate": 1.9487195778002013e-05,
+      "loss": 1.1483,
+      "step": 1881
+    },
+    {
+      "epoch": 0.3226954154789206,
+      "grad_norm": 1.7109375,
+      "learning_rate": 1.9486624735259557e-05,
+      "loss": 1.0993,
+      "step": 1882
+    },
+    {
+      "epoch": 0.3228668795679105,
+      "grad_norm": 1.5625,
+      "learning_rate": 1.9486053383122544e-05,
+      "loss": 1.0002,
+      "step": 1883
+    },
+    {
+      "epoch": 0.3230383436569004,
+      "grad_norm": 1.7578125,
+      "learning_rate": 1.948548172160961e-05,
+      "loss": 1.0762,
+      "step": 1884
+    },
+    {
+      "epoch": 0.3232098077458902,
+      "grad_norm": 1.5703125,
+      "learning_rate": 1.94849097507394e-05,
+      "loss": 1.0094,
+      "step": 1885
+    },
+    {
+      "epoch": 0.3233812718348801,
+      "grad_norm": 1.65625,
+      "learning_rate": 1.9484337470530563e-05,
+      "loss": 1.0485,
+      "step": 1886
+    },
+    {
+      "epoch": 0.32355273592386996,
+      "grad_norm": 1.75,
+      "learning_rate": 1.9483764881001767e-05,
+      "loss": 1.102,
+      "step": 1887
+    },
+    {
+      "epoch": 0.3237242000128598,
+      "grad_norm": 1.6875,
+      "learning_rate": 1.9483191982171686e-05,
+      "loss": 1.0389,
+      "step": 1888
+    },
+    {
+      "epoch": 0.32389566410184967,
+      "grad_norm": 1.5078125,
+      "learning_rate": 1.9482618774059007e-05,
+      "loss": 0.9699,
+      "step": 1889
+    },
+    {
+      "epoch": 0.32406712819083955,
+      "grad_norm": 1.7109375,
+      "learning_rate": 1.9482045256682415e-05,
+      "loss": 1.1713,
+      "step": 1890
+    },
+    {
+      "epoch": 0.3242385922798294,
+      "grad_norm": 1.734375,
+      "learning_rate": 1.9481471430060627e-05,
+      "loss": 1.074,
+      "step": 1891
+    },
+    {
+      "epoch": 0.32441005636881926,
+      "grad_norm": 1.7421875,
+      "learning_rate": 1.9480897294212348e-05,
+      "loss": 1.0589,
+      "step": 1892
+    },
+    {
+      "epoch": 0.32458152045780914,
+      "grad_norm": 1.6875,
+      "learning_rate": 1.9480322849156307e-05,
+      "loss": 1.0421,
+      "step": 1893
+    },
+    {
+      "epoch": 0.32475298454679896,
+      "grad_norm": 1.6015625,
+      "learning_rate": 1.947974809491124e-05,
+      "loss": 1.0185,
+      "step": 1894
+    },
+    {
+      "epoch": 0.32492444863578884,
+      "grad_norm": 3.296875,
+      "learning_rate": 1.947917303149589e-05,
+      "loss": 1.1065,
+      "step": 1895
+    },
+    {
+      "epoch": 0.3250959127247787,
+      "grad_norm": 1.6171875,
+      "learning_rate": 1.9478597658929012e-05,
+      "loss": 1.0147,
+      "step": 1896
+    },
+    {
+      "epoch": 0.32526737681376855,
+      "grad_norm": 2.28125,
+      "learning_rate": 1.947802197722937e-05,
+      "loss": 1.055,
+      "step": 1897
+    },
+    {
+      "epoch": 0.32543884090275843,
+      "grad_norm": 1.84375,
+      "learning_rate": 1.9477445986415743e-05,
+      "loss": 1.0291,
+      "step": 1898
+    },
+    {
+      "epoch": 0.3256103049917483,
+      "grad_norm": 1.6328125,
+      "learning_rate": 1.947686968650691e-05,
+      "loss": 1.0111,
+      "step": 1899
+    },
+    {
+      "epoch": 0.32578176908073814,
+      "grad_norm": 1.59375,
+      "learning_rate": 1.9476293077521674e-05,
+      "loss": 1.0271,
+      "step": 1900
+    },
+    {
+      "epoch": 0.325953233169728,
+      "grad_norm": 1.640625,
+      "learning_rate": 1.9475716159478836e-05,
+      "loss": 1.0097,
+      "step": 1901
+    },
+    {
+      "epoch": 0.3261246972587179,
+      "grad_norm": 1.7734375,
+      "learning_rate": 1.947513893239721e-05,
+      "loss": 1.0723,
+      "step": 1902
+    },
+    {
+      "epoch": 0.3262961613477077,
+      "grad_norm": 1.703125,
+      "learning_rate": 1.947456139629562e-05,
+      "loss": 1.092,
+      "step": 1903
+    },
+    {
+      "epoch": 0.3264676254366976,
+      "grad_norm": 1.71875,
+      "learning_rate": 1.947398355119291e-05,
+      "loss": 1.0427,
+      "step": 1904
+    },
+    {
+      "epoch": 0.3266390895256875,
+      "grad_norm": 1.453125,
+      "learning_rate": 1.9473405397107917e-05,
+      "loss": 0.9551,
+      "step": 1905
+    },
+    {
+      "epoch": 0.3268105536146773,
+      "grad_norm": 1.6328125,
+      "learning_rate": 1.9472826934059506e-05,
+      "loss": 1.0296,
+      "step": 1906
+    },
+    {
+      "epoch": 0.3269820177036672,
+      "grad_norm": 1.5859375,
+      "learning_rate": 1.947224816206653e-05,
+      "loss": 1.072,
+      "step": 1907
+    },
+    {
+      "epoch": 0.3271534817926571,
+      "grad_norm": 1.6796875,
+      "learning_rate": 1.9471669081147878e-05,
+      "loss": 0.9822,
+      "step": 1908
+    },
+    {
+      "epoch": 0.3273249458816469,
+      "grad_norm": 1.640625,
+      "learning_rate": 1.947108969132243e-05,
+      "loss": 1.0869,
+      "step": 1909
+    },
+    {
+      "epoch": 0.3274964099706368,
+      "grad_norm": 1.6015625,
+      "learning_rate": 1.947050999260908e-05,
+      "loss": 1.0469,
+      "step": 1910
+    },
+    {
+      "epoch": 0.32766787405962666,
+      "grad_norm": 1.671875,
+      "learning_rate": 1.9469929985026738e-05,
+      "loss": 1.1537,
+      "step": 1911
+    },
+    {
+      "epoch": 0.3278393381486165,
+      "grad_norm": 1.59375,
+      "learning_rate": 1.9469349668594314e-05,
+      "loss": 1.093,
+      "step": 1912
+    },
+    {
+      "epoch": 0.32801080223760637,
+      "grad_norm": 1.6171875,
+      "learning_rate": 1.9468769043330744e-05,
+      "loss": 1.078,
+      "step": 1913
+    },
+    {
+      "epoch": 0.32818226632659625,
+      "grad_norm": 1.640625,
+      "learning_rate": 1.946818810925496e-05,
+      "loss": 1.0673,
+      "step": 1914
+    },
+    {
+      "epoch": 0.3283537304155861,
+      "grad_norm": 1.671875,
+      "learning_rate": 1.9467606866385905e-05,
+      "loss": 1.105,
+      "step": 1915
+    },
+    {
+      "epoch": 0.32852519450457596,
+      "grad_norm": 1.7109375,
+      "learning_rate": 1.946702531474254e-05,
+      "loss": 1.0411,
+      "step": 1916
+    },
+    {
+      "epoch": 0.3286966585935658,
+      "grad_norm": 1.578125,
+      "learning_rate": 1.946644345434383e-05,
+      "loss": 0.9658,
+      "step": 1917
+    },
+    {
+      "epoch": 0.32886812268255566,
+      "grad_norm": 1.671875,
+      "learning_rate": 1.9465861285208752e-05,
+      "loss": 1.0598,
+      "step": 1918
+    },
+    {
+      "epoch": 0.32903958677154554,
+      "grad_norm": 1.6328125,
+      "learning_rate": 1.9465278807356292e-05,
+      "loss": 1.104,
+      "step": 1919
+    },
+    {
+      "epoch": 0.32921105086053537,
+      "grad_norm": 1.65625,
+      "learning_rate": 1.9464696020805446e-05,
+      "loss": 1.1023,
+      "step": 1920
+    },
+    {
+      "epoch": 0.32938251494952525,
+      "grad_norm": 1.6484375,
+      "learning_rate": 1.9464112925575224e-05,
+      "loss": 1.1455,
+      "step": 1921
+    },
+    {
+      "epoch": 0.32955397903851513,
+      "grad_norm": 1.5859375,
+      "learning_rate": 1.9463529521684638e-05,
+      "loss": 1.0643,
+      "step": 1922
+    },
+    {
+      "epoch": 0.32972544312750496,
+      "grad_norm": 1.65625,
+      "learning_rate": 1.9462945809152722e-05,
+      "loss": 1.0266,
+      "step": 1923
+    },
+    {
+      "epoch": 0.32989690721649484,
+      "grad_norm": 1.6484375,
+      "learning_rate": 1.9462361787998503e-05,
+      "loss": 0.9947,
+      "step": 1924
+    },
+    {
+      "epoch": 0.3300683713054847,
+      "grad_norm": 1.625,
+      "learning_rate": 1.9461777458241038e-05,
+      "loss": 1.1242,
+      "step": 1925
+    },
+    {
+      "epoch": 0.33023983539447455,
+      "grad_norm": 1.6484375,
+      "learning_rate": 1.9461192819899383e-05,
+      "loss": 1.04,
+      "step": 1926
+    },
+    {
+      "epoch": 0.3304112994834644,
+      "grad_norm": 1.6171875,
+      "learning_rate": 1.94606078729926e-05,
+      "loss": 0.9667,
+      "step": 1927
+    },
+    {
+      "epoch": 0.3305827635724543,
+      "grad_norm": 1.6328125,
+      "learning_rate": 1.9460022617539765e-05,
+      "loss": 1.0544,
+      "step": 1928
+    },
+    {
+      "epoch": 0.33075422766144413,
+      "grad_norm": 1.6328125,
+      "learning_rate": 1.9459437053559974e-05,
+      "loss": 1.0008,
+      "step": 1929
+    },
+    {
+      "epoch": 0.330925691750434,
+      "grad_norm": 1.6328125,
+      "learning_rate": 1.9458851181072317e-05,
+      "loss": 1.121,
+      "step": 1930
+    },
+    {
+      "epoch": 0.3310971558394239,
+      "grad_norm": 1.6484375,
+      "learning_rate": 1.9458265000095905e-05,
+      "loss": 1.0576,
+      "step": 1931
+    },
+    {
+      "epoch": 0.3312686199284137,
+      "grad_norm": 1.6640625,
+      "learning_rate": 1.9457678510649857e-05,
+      "loss": 1.1327,
+      "step": 1932
+    },
+    {
+      "epoch": 0.3314400840174036,
+      "grad_norm": 1.53125,
+      "learning_rate": 1.9457091712753298e-05,
+      "loss": 1.0161,
+      "step": 1933
+    },
+    {
+      "epoch": 0.3316115481063935,
+      "grad_norm": 1.6328125,
+      "learning_rate": 1.9456504606425365e-05,
+      "loss": 0.9686,
+      "step": 1934
+    },
+    {
+      "epoch": 0.3317830121953833,
+      "grad_norm": 1.6015625,
+      "learning_rate": 1.9455917191685207e-05,
+      "loss": 0.9736,
+      "step": 1935
+    },
+    {
+      "epoch": 0.3319544762843732,
+      "grad_norm": 1.546875,
+      "learning_rate": 1.945532946855198e-05,
+      "loss": 1.0819,
+      "step": 1936
+    },
+    {
+      "epoch": 0.33212594037336307,
+      "grad_norm": 1.609375,
+      "learning_rate": 1.9454741437044858e-05,
+      "loss": 1.1065,
+      "step": 1937
+    },
+    {
+      "epoch": 0.3322974044623529,
+      "grad_norm": 1.5859375,
+      "learning_rate": 1.945415309718301e-05,
+      "loss": 0.9961,
+      "step": 1938
+    },
+    {
+      "epoch": 0.3324688685513428,
+      "grad_norm": 1.6875,
+      "learning_rate": 1.945356444898563e-05,
+      "loss": 1.1321,
+      "step": 1939
+    },
+    {
+      "epoch": 0.33264033264033266,
+      "grad_norm": 1.6484375,
+      "learning_rate": 1.9452975492471915e-05,
+      "loss": 1.0079,
+      "step": 1940
+    },
+    {
+      "epoch": 0.3328117967293225,
+      "grad_norm": 1.5390625,
+      "learning_rate": 1.9452386227661076e-05,
+      "loss": 1.0916,
+      "step": 1941
+    },
+    {
+      "epoch": 0.33298326081831237,
+      "grad_norm": 1.6875,
+      "learning_rate": 1.945179665457232e-05,
+      "loss": 1.0289,
+      "step": 1942
+    },
+    {
+      "epoch": 0.33315472490730225,
+      "grad_norm": 1.7265625,
+      "learning_rate": 1.945120677322489e-05,
+      "loss": 1.1086,
+      "step": 1943
+    },
+    {
+      "epoch": 0.33332618899629207,
+      "grad_norm": 1.625,
+      "learning_rate": 1.9450616583638013e-05,
+      "loss": 0.9818,
+      "step": 1944
+    },
+    {
+      "epoch": 0.33349765308528195,
+      "grad_norm": 1.5703125,
+      "learning_rate": 1.9450026085830946e-05,
+      "loss": 1.0208,
+      "step": 1945
+    },
+    {
+      "epoch": 0.33366911717427183,
+      "grad_norm": 1.6484375,
+      "learning_rate": 1.9449435279822934e-05,
+      "loss": 1.0519,
+      "step": 1946
+    },
+    {
+      "epoch": 0.33384058126326166,
+      "grad_norm": 1.609375,
+      "learning_rate": 1.9448844165633264e-05,
+      "loss": 1.0612,
+      "step": 1947
+    },
+    {
+      "epoch": 0.33401204535225154,
+      "grad_norm": 1.6328125,
+      "learning_rate": 1.9448252743281196e-05,
+      "loss": 1.0525,
+      "step": 1948
+    },
+    {
+      "epoch": 0.3341835094412414,
+      "grad_norm": 1.6875,
+      "learning_rate": 1.9447661012786034e-05,
+      "loss": 1.0689,
+      "step": 1949
+    },
+    {
+      "epoch": 0.33435497353023125,
+      "grad_norm": 1.6640625,
+      "learning_rate": 1.9447068974167068e-05,
+      "loss": 1.1055,
+      "step": 1950
+    },
+    {
+      "epoch": 0.33452643761922113,
+      "grad_norm": 1.6328125,
+      "learning_rate": 1.9446476627443608e-05,
+      "loss": 0.9484,
+      "step": 1951
+    },
+    {
+      "epoch": 0.334697901708211,
+      "grad_norm": 1.5546875,
+      "learning_rate": 1.9445883972634973e-05,
+      "loss": 1.0176,
+      "step": 1952
+    },
+    {
+      "epoch": 0.33486936579720084,
+      "grad_norm": 1.5078125,
+      "learning_rate": 1.9445291009760493e-05,
+      "loss": 0.9543,
+      "step": 1953
+    },
+    {
+      "epoch": 0.3350408298861907,
+      "grad_norm": 1.609375,
+      "learning_rate": 1.9444697738839503e-05,
+      "loss": 1.065,
+      "step": 1954
+    },
+    {
+      "epoch": 0.3352122939751806,
+      "grad_norm": 1.7734375,
+      "learning_rate": 1.9444104159891353e-05,
+      "loss": 1.0328,
+      "step": 1955
+    },
+    {
+      "epoch": 0.3353837580641704,
+      "grad_norm": 1.703125,
+      "learning_rate": 1.9443510272935407e-05,
+      "loss": 1.175,
+      "step": 1956
+    },
+    {
+      "epoch": 0.3355552221531603,
+      "grad_norm": 1.6640625,
+      "learning_rate": 1.944291607799103e-05,
+      "loss": 0.9973,
+      "step": 1957
+    },
+    {
+      "epoch": 0.3357266862421502,
+      "grad_norm": 1.5625,
+      "learning_rate": 1.94423215750776e-05,
+      "loss": 1.0261,
+      "step": 1958
+    },
+    {
+      "epoch": 0.33589815033114,
+      "grad_norm": 1.578125,
+      "learning_rate": 1.9441726764214506e-05,
+      "loss": 1.0332,
+      "step": 1959
+    },
+    {
+      "epoch": 0.3360696144201299,
+      "grad_norm": 1.7109375,
+      "learning_rate": 1.9441131645421146e-05,
+      "loss": 1.091,
+      "step": 1960
+    },
+    {
+      "epoch": 0.3362410785091198,
+      "grad_norm": 1.6796875,
+      "learning_rate": 1.9440536218716934e-05,
+      "loss": 1.0468,
+      "step": 1961
+    },
+    {
+      "epoch": 0.3364125425981096,
+      "grad_norm": 1.671875,
+      "learning_rate": 1.9439940484121287e-05,
+      "loss": 1.0747,
+      "step": 1962
+    },
+    {
+      "epoch": 0.3365840066870995,
+      "grad_norm": 1.59375,
+      "learning_rate": 1.943934444165363e-05,
+      "loss": 1.0025,
+      "step": 1963
+    },
+    {
+      "epoch": 0.3367554707760893,
+      "grad_norm": 1.625,
+      "learning_rate": 1.943874809133341e-05,
+      "loss": 1.0103,
+      "step": 1964
+    },
+    {
+      "epoch": 0.3369269348650792,
+      "grad_norm": 1.5859375,
+      "learning_rate": 1.9438151433180068e-05,
+      "loss": 1.0693,
+      "step": 1965
+    },
+    {
+      "epoch": 0.33709839895406907,
+      "grad_norm": 1.546875,
+      "learning_rate": 1.9437554467213066e-05,
+      "loss": 1.0221,
+      "step": 1966
+    },
+    {
+      "epoch": 0.3372698630430589,
+      "grad_norm": 1.5390625,
+      "learning_rate": 1.943695719345188e-05,
+      "loss": 1.1032,
+      "step": 1967
+    },
+    {
+      "epoch": 0.3374413271320488,
+      "grad_norm": 1.578125,
+      "learning_rate": 1.9436359611915978e-05,
+      "loss": 1.0418,
+      "step": 1968
+    },
+    {
+      "epoch": 0.33761279122103866,
+      "grad_norm": 1.6640625,
+      "learning_rate": 1.943576172262486e-05,
+      "loss": 1.0459,
+      "step": 1969
+    },
+    {
+      "epoch": 0.3377842553100285,
+      "grad_norm": 1.6484375,
+      "learning_rate": 1.9435163525598017e-05,
+      "loss": 0.9038,
+      "step": 1970
+    },
+    {
+      "epoch": 0.33795571939901836,
+      "grad_norm": 1.6953125,
+      "learning_rate": 1.943456502085496e-05,
+      "loss": 1.061,
+      "step": 1971
+    },
+    {
+      "epoch": 0.33812718348800824,
+      "grad_norm": 1.625,
+      "learning_rate": 1.9433966208415212e-05,
+      "loss": 1.0721,
+      "step": 1972
+    },
+    {
+      "epoch": 0.33829864757699807,
+      "grad_norm": 1.5703125,
+      "learning_rate": 1.94333670882983e-05,
+      "loss": 1.0833,
+      "step": 1973
+    },
+    {
+      "epoch": 0.33847011166598795,
+      "grad_norm": 1.4921875,
+      "learning_rate": 1.9432767660523768e-05,
+      "loss": 0.9957,
+      "step": 1974
+    },
+    {
+      "epoch": 0.33864157575497783,
+      "grad_norm": 1.609375,
+      "learning_rate": 1.943216792511116e-05,
+      "loss": 0.9921,
+      "step": 1975
+    },
+    {
+      "epoch": 0.33881303984396766,
+      "grad_norm": 1.6640625,
+      "learning_rate": 1.9431567882080042e-05,
+      "loss": 1.002,
+      "step": 1976
+    },
+    {
+      "epoch": 0.33898450393295754,
+      "grad_norm": 1.6953125,
+      "learning_rate": 1.9430967531449973e-05,
+      "loss": 1.1183,
+      "step": 1977
+    },
+    {
+      "epoch": 0.3391559680219474,
+      "grad_norm": 1.65625,
+      "learning_rate": 1.943036687324054e-05,
+      "loss": 1.0156,
+      "step": 1978
+    },
+    {
+      "epoch": 0.33932743211093724,
+      "grad_norm": 1.671875,
+      "learning_rate": 1.9429765907471336e-05,
+      "loss": 1.1207,
+      "step": 1979
+    },
+    {
+      "epoch": 0.3394988961999271,
+      "grad_norm": 1.7265625,
+      "learning_rate": 1.942916463416195e-05,
+      "loss": 1.07,
+      "step": 1980
+    },
+    {
+      "epoch": 0.339670360288917,
+      "grad_norm": 1.6015625,
+      "learning_rate": 1.9428563053332004e-05,
+      "loss": 0.9844,
+      "step": 1981
+    },
+    {
+      "epoch": 0.33984182437790683,
+      "grad_norm": 1.59375,
+      "learning_rate": 1.942796116500111e-05,
+      "loss": 1.0781,
+      "step": 1982
+    },
+    {
+      "epoch": 0.3400132884668967,
+      "grad_norm": 1.6484375,
+      "learning_rate": 1.94273589691889e-05,
+      "loss": 1.078,
+      "step": 1983
+    },
+    {
+      "epoch": 0.3401847525558866,
+      "grad_norm": 1.6875,
+      "learning_rate": 1.9426756465915014e-05,
+      "loss": 1.1393,
+      "step": 1984
+    },
+    {
+      "epoch": 0.3403562166448764,
+      "grad_norm": 1.6015625,
+      "learning_rate": 1.94261536551991e-05,
+      "loss": 1.0316,
+      "step": 1985
+    },
+    {
+      "epoch": 0.3405276807338663,
+      "grad_norm": 1.578125,
+      "learning_rate": 1.9425550537060826e-05,
+      "loss": 1.1531,
+      "step": 1986
+    },
+    {
+      "epoch": 0.3406991448228562,
+      "grad_norm": 1.5546875,
+      "learning_rate": 1.942494711151985e-05,
+      "loss": 1.0325,
+      "step": 1987
+    },
+    {
+      "epoch": 0.340870608911846,
+      "grad_norm": 1.640625,
+      "learning_rate": 1.9424343378595857e-05,
+      "loss": 1.0903,
+      "step": 1988
+    },
+    {
+      "epoch": 0.3410420730008359,
+      "grad_norm": 1.65625,
+      "learning_rate": 1.942373933830854e-05,
+      "loss": 1.092,
+      "step": 1989
+    },
+    {
+      "epoch": 0.34121353708982577,
+      "grad_norm": 1.5625,
+      "learning_rate": 1.9423134990677596e-05,
+      "loss": 1.032,
+      "step": 1990
+    },
+    {
+      "epoch": 0.3413850011788156,
+      "grad_norm": 1.625,
+      "learning_rate": 1.9422530335722736e-05,
+      "loss": 1.0583,
+      "step": 1991
+    },
+    {
+      "epoch": 0.3415564652678055,
+      "grad_norm": 1.6015625,
+      "learning_rate": 1.942192537346368e-05,
+      "loss": 1.0195,
+      "step": 1992
+    },
+    {
+      "epoch": 0.34172792935679536,
+      "grad_norm": 1.6640625,
+      "learning_rate": 1.942132010392016e-05,
+      "loss": 1.1263,
+      "step": 1993
+    },
+    {
+      "epoch": 0.3418993934457852,
+      "grad_norm": 1.6171875,
+      "learning_rate": 1.9420714527111907e-05,
+      "loss": 1.1151,
+      "step": 1994
+    },
+    {
+      "epoch": 0.34207085753477506,
+      "grad_norm": 1.640625,
+      "learning_rate": 1.9420108643058685e-05,
+      "loss": 1.0326,
+      "step": 1995
+    },
+    {
+      "epoch": 0.34224232162376494,
+      "grad_norm": 1.578125,
+      "learning_rate": 1.9419502451780243e-05,
+      "loss": 0.9469,
+      "step": 1996
+    },
+    {
+      "epoch": 0.34241378571275477,
+      "grad_norm": 1.546875,
+      "learning_rate": 1.941889595329636e-05,
+      "loss": 1.0316,
+      "step": 1997
+    },
+    {
+      "epoch": 0.34258524980174465,
+      "grad_norm": 1.6171875,
+      "learning_rate": 1.9418289147626804e-05,
+      "loss": 1.069,
+      "step": 1998
+    },
+    {
+      "epoch": 0.34275671389073453,
+      "grad_norm": 1.59375,
+      "learning_rate": 1.941768203479138e-05,
+      "loss": 1.049,
+      "step": 1999
+    },
+    {
+      "epoch": 0.34292817797972436,
+      "grad_norm": 1.6171875,
+      "learning_rate": 1.941707461480988e-05,
+      "loss": 1.0452,
+      "step": 2000
+    },
+    {
+      "epoch": 0.34309964206871424,
+      "grad_norm": 1.640625,
+      "learning_rate": 1.9416466887702116e-05,
+      "loss": 1.0781,
+      "step": 2001
+    },
+    {
+      "epoch": 0.3432711061577041,
+      "grad_norm": 1.671875,
+      "learning_rate": 1.9415858853487904e-05,
+      "loss": 0.991,
+      "step": 2002
+    },
+    {
+      "epoch": 0.34344257024669395,
+      "grad_norm": 1.5703125,
+      "learning_rate": 1.941525051218708e-05,
+      "loss": 1.0131,
+      "step": 2003
+    },
+    {
+      "epoch": 0.3436140343356838,
+      "grad_norm": 2.25,
+      "learning_rate": 1.9414641863819484e-05,
+      "loss": 1.1457,
+      "step": 2004
+    },
+    {
+      "epoch": 0.3437854984246737,
+      "grad_norm": 1.6171875,
+      "learning_rate": 1.9414032908404962e-05,
+      "loss": 1.1031,
+      "step": 2005
+    },
+    {
+      "epoch": 0.34395696251366353,
+      "grad_norm": 1.671875,
+      "learning_rate": 1.9413423645963378e-05,
+      "loss": 1.1239,
+      "step": 2006
+    },
+    {
+      "epoch": 0.3441284266026534,
+      "grad_norm": 1.578125,
+      "learning_rate": 1.94128140765146e-05,
+      "loss": 1.0762,
+      "step": 2007
+    },
+    {
+      "epoch": 0.3442998906916433,
+      "grad_norm": 1.5625,
+      "learning_rate": 1.9412204200078514e-05,
+      "loss": 1.0155,
+      "step": 2008
+    },
+    {
+      "epoch": 0.3444713547806331,
+      "grad_norm": 1.6328125,
+      "learning_rate": 1.9411594016675e-05,
+      "loss": 1.1015,
+      "step": 2009
+    },
+    {
+      "epoch": 0.344642818869623,
+      "grad_norm": 1.703125,
+      "learning_rate": 1.9410983526323964e-05,
+      "loss": 1.1553,
+      "step": 2010
+    },
+    {
+      "epoch": 0.3448142829586129,
+      "grad_norm": 1.6171875,
+      "learning_rate": 1.941037272904532e-05,
+      "loss": 1.1046,
+      "step": 2011
+    },
+    {
+      "epoch": 0.3449857470476027,
+      "grad_norm": 1.5234375,
+      "learning_rate": 1.940976162485899e-05,
+      "loss": 1.0013,
+      "step": 2012
+    },
+    {
+      "epoch": 0.3451572111365926,
+      "grad_norm": 1.6875,
+      "learning_rate": 1.9409150213784892e-05,
+      "loss": 1.0267,
+      "step": 2013
+    },
+    {
+      "epoch": 0.3453286752255824,
+      "grad_norm": 1.546875,
+      "learning_rate": 1.9408538495842973e-05,
+      "loss": 1.0041,
+      "step": 2014
+    },
+    {
+      "epoch": 0.3455001393145723,
+      "grad_norm": 1.6015625,
+      "learning_rate": 1.9407926471053188e-05,
+      "loss": 1.02,
+      "step": 2015
+    },
+    {
+      "epoch": 0.3456716034035622,
+      "grad_norm": 1.6171875,
+      "learning_rate": 1.9407314139435495e-05,
+      "loss": 1.0167,
+      "step": 2016
+    },
+    {
+      "epoch": 0.345843067492552,
+      "grad_norm": 1.7109375,
+      "learning_rate": 1.9406701501009862e-05,
+      "loss": 1.0526,
+      "step": 2017
+    },
+    {
+      "epoch": 0.3460145315815419,
+      "grad_norm": 1.6328125,
+      "learning_rate": 1.9406088555796268e-05,
+      "loss": 1.0379,
+      "step": 2018
+    },
+    {
+      "epoch": 0.34618599567053177,
+      "grad_norm": 1.625,
+      "learning_rate": 1.940547530381471e-05,
+      "loss": 1.0715,
+      "step": 2019
+    },
+    {
+      "epoch": 0.3463574597595216,
+      "grad_norm": 1.6328125,
+      "learning_rate": 1.9404861745085184e-05,
+      "loss": 1.0622,
+      "step": 2020
+    },
+    {
+      "epoch": 0.34652892384851147,
+      "grad_norm": 1.5859375,
+      "learning_rate": 1.94042478796277e-05,
+      "loss": 0.9886,
+      "step": 2021
+    },
+    {
+      "epoch": 0.34670038793750135,
+      "grad_norm": 1.515625,
+      "learning_rate": 1.9403633707462282e-05,
+      "loss": 1.0259,
+      "step": 2022
+    },
+    {
+      "epoch": 0.3468718520264912,
+      "grad_norm": 1.703125,
+      "learning_rate": 1.9403019228608953e-05,
+      "loss": 0.9734,
+      "step": 2023
+    },
+    {
+      "epoch": 0.34704331611548106,
+      "grad_norm": 1.59375,
+      "learning_rate": 1.9402404443087763e-05,
+      "loss": 1.0332,
+      "step": 2024
+    },
+    {
+      "epoch": 0.34721478020447094,
+      "grad_norm": 1.625,
+      "learning_rate": 1.9401789350918755e-05,
+      "loss": 1.0274,
+      "step": 2025
+    },
+    {
+      "epoch": 0.34738624429346077,
+      "grad_norm": 1.7265625,
+      "learning_rate": 1.9401173952121993e-05,
+      "loss": 1.1486,
+      "step": 2026
+    },
+    {
+      "epoch": 0.34755770838245065,
+      "grad_norm": 1.5859375,
+      "learning_rate": 1.9400558246717547e-05,
+      "loss": 1.0557,
+      "step": 2027
+    },
+    {
+      "epoch": 0.34772917247144053,
+      "grad_norm": 1.578125,
+      "learning_rate": 1.93999422347255e-05,
+      "loss": 0.9885,
+      "step": 2028
+    },
+    {
+      "epoch": 0.34790063656043035,
+      "grad_norm": 1.5625,
+      "learning_rate": 1.9399325916165937e-05,
+      "loss": 1.1045,
+      "step": 2029
+    },
+    {
+      "epoch": 0.34807210064942024,
+      "grad_norm": 1.53125,
+      "learning_rate": 1.939870929105896e-05,
+      "loss": 0.9841,
+      "step": 2030
+    },
+    {
+      "epoch": 0.3482435647384101,
+      "grad_norm": 1.5859375,
+      "learning_rate": 1.9398092359424683e-05,
+      "loss": 0.9914,
+      "step": 2031
+    },
+    {
+      "epoch": 0.34841502882739994,
+      "grad_norm": 1.5703125,
+      "learning_rate": 1.9397475121283226e-05,
+      "loss": 1.0192,
+      "step": 2032
+    },
+    {
+      "epoch": 0.3485864929163898,
+      "grad_norm": 1.6015625,
+      "learning_rate": 1.9396857576654714e-05,
+      "loss": 1.0977,
+      "step": 2033
+    },
+    {
+      "epoch": 0.3487579570053797,
+      "grad_norm": 1.5703125,
+      "learning_rate": 1.9396239725559294e-05,
+      "loss": 1.0574,
+      "step": 2034
+    },
+    {
+      "epoch": 0.34892942109436953,
+      "grad_norm": 1.5859375,
+      "learning_rate": 1.939562156801711e-05,
+      "loss": 1.0004,
+      "step": 2035
+    },
+    {
+      "epoch": 0.3491008851833594,
+      "grad_norm": 1.6171875,
+      "learning_rate": 1.939500310404833e-05,
+      "loss": 1.0306,
+      "step": 2036
+    },
+    {
+      "epoch": 0.3492723492723493,
+      "grad_norm": 1.65625,
+      "learning_rate": 1.9394384333673117e-05,
+      "loss": 1.0474,
+      "step": 2037
+    },
+    {
+      "epoch": 0.3494438133613391,
+      "grad_norm": 1.6640625,
+      "learning_rate": 1.9393765256911657e-05,
+      "loss": 1.0378,
+      "step": 2038
+    },
+    {
+      "epoch": 0.349615277450329,
+      "grad_norm": 1.671875,
+      "learning_rate": 1.9393145873784137e-05,
+      "loss": 1.0387,
+      "step": 2039
+    },
+    {
+      "epoch": 0.3497867415393189,
+      "grad_norm": 1.6953125,
+      "learning_rate": 1.939252618431076e-05,
+      "loss": 1.0389,
+      "step": 2040
+    },
+    {
+      "epoch": 0.3499582056283087,
+      "grad_norm": 1.7578125,
+      "learning_rate": 1.9391906188511736e-05,
+      "loss": 1.041,
+      "step": 2041
+    },
+    {
+      "epoch": 0.3501296697172986,
+      "grad_norm": 1.6171875,
+      "learning_rate": 1.939128588640728e-05,
+      "loss": 1.1094,
+      "step": 2042
+    },
+    {
+      "epoch": 0.35030113380628847,
+      "grad_norm": 1.5859375,
+      "learning_rate": 1.9390665278017635e-05,
+      "loss": 1.0563,
+      "step": 2043
+    },
+    {
+      "epoch": 0.3504725978952783,
+      "grad_norm": 1.5625,
+      "learning_rate": 1.939004436336303e-05,
+      "loss": 1.0217,
+      "step": 2044
+    },
+    {
+      "epoch": 0.3506440619842682,
+      "grad_norm": 1.6328125,
+      "learning_rate": 1.9389423142463715e-05,
+      "loss": 1.0005,
+      "step": 2045
+    },
+    {
+      "epoch": 0.35081552607325805,
+      "grad_norm": 1.7265625,
+      "learning_rate": 1.9388801615339955e-05,
+      "loss": 1.0865,
+      "step": 2046
+    },
+    {
+      "epoch": 0.3509869901622479,
+      "grad_norm": 1.6171875,
+      "learning_rate": 1.9388179782012023e-05,
+      "loss": 1.0456,
+      "step": 2047
+    },
+    {
+      "epoch": 0.35115845425123776,
+      "grad_norm": 1.5625,
+      "learning_rate": 1.9387557642500192e-05,
+      "loss": 1.0332,
+      "step": 2048
+    },
+    {
+      "epoch": 0.35132991834022764,
+      "grad_norm": 1.6875,
+      "learning_rate": 1.9386935196824756e-05,
+      "loss": 1.057,
+      "step": 2049
+    },
+    {
+      "epoch": 0.35150138242921747,
+      "grad_norm": 1.71875,
+      "learning_rate": 1.938631244500602e-05,
+      "loss": 1.1701,
+      "step": 2050
+    },
+    {
+      "epoch": 0.35167284651820735,
+      "grad_norm": 1.640625,
+      "learning_rate": 1.9385689387064285e-05,
+      "loss": 0.9727,
+      "step": 2051
+    },
+    {
+      "epoch": 0.35184431060719723,
+      "grad_norm": 1.6640625,
+      "learning_rate": 1.9385066023019878e-05,
+      "loss": 1.0281,
+      "step": 2052
+    },
+    {
+      "epoch": 0.35201577469618706,
+      "grad_norm": 1.6015625,
+      "learning_rate": 1.9384442352893125e-05,
+      "loss": 1.0045,
+      "step": 2053
+    },
+    {
+      "epoch": 0.35218723878517694,
+      "grad_norm": 1.5859375,
+      "learning_rate": 1.938381837670437e-05,
+      "loss": 1.0382,
+      "step": 2054
+    },
+    {
+      "epoch": 0.3523587028741668,
+      "grad_norm": 1.6484375,
+      "learning_rate": 1.938319409447396e-05,
+      "loss": 1.0064,
+      "step": 2055
+    },
+    {
+      "epoch": 0.35253016696315664,
+      "grad_norm": 1.6171875,
+      "learning_rate": 1.938256950622226e-05,
+      "loss": 1.0506,
+      "step": 2056
+    },
+    {
+      "epoch": 0.3527016310521465,
+      "grad_norm": 1.5703125,
+      "learning_rate": 1.938194461196963e-05,
+      "loss": 1.0538,
+      "step": 2057
+    },
+    {
+      "epoch": 0.3528730951411364,
+      "grad_norm": 1.6015625,
+      "learning_rate": 1.9381319411736467e-05,
+      "loss": 1.0444,
+      "step": 2058
+    },
+    {
+      "epoch": 0.35304455923012623,
+      "grad_norm": 1.5390625,
+      "learning_rate": 1.9380693905543144e-05,
+      "loss": 1.0035,
+      "step": 2059
+    },
+    {
+      "epoch": 0.3532160233191161,
+      "grad_norm": 1.7578125,
+      "learning_rate": 1.938006809341007e-05,
+      "loss": 1.1376,
+      "step": 2060
+    },
+    {
+      "epoch": 0.35338748740810594,
+      "grad_norm": 1.671875,
+      "learning_rate": 1.9379441975357654e-05,
+      "loss": 1.0271,
+      "step": 2061
+    },
+    {
+      "epoch": 0.3535589514970958,
+      "grad_norm": 1.609375,
+      "learning_rate": 1.9378815551406315e-05,
+      "loss": 1.0318,
+      "step": 2062
+    },
+    {
+      "epoch": 0.3537304155860857,
+      "grad_norm": 1.6015625,
+      "learning_rate": 1.9378188821576484e-05,
+      "loss": 1.1232,
+      "step": 2063
+    },
+    {
+      "epoch": 0.3539018796750755,
+      "grad_norm": 1.6328125,
+      "learning_rate": 1.9377561785888602e-05,
+      "loss": 0.9949,
+      "step": 2064
+    },
+    {
+      "epoch": 0.3540733437640654,
+      "grad_norm": 1.6328125,
+      "learning_rate": 1.9376934444363114e-05,
+      "loss": 1.1025,
+      "step": 2065
+    },
+    {
+      "epoch": 0.3542448078530553,
+      "grad_norm": 1.6484375,
+      "learning_rate": 1.9376306797020486e-05,
+      "loss": 1.1526,
+      "step": 2066
+    },
+    {
+      "epoch": 0.3544162719420451,
+      "grad_norm": 20.375,
+      "learning_rate": 1.9375678843881186e-05,
+      "loss": 1.1018,
+      "step": 2067
+    },
+    {
+      "epoch": 0.354587736031035,
+      "grad_norm": 1.53125,
+      "learning_rate": 1.937505058496569e-05,
+      "loss": 1.0327,
+      "step": 2068
+    },
+    {
+      "epoch": 0.3547592001200249,
+      "grad_norm": 1.5625,
+      "learning_rate": 1.937442202029449e-05,
+      "loss": 1.0052,
+      "step": 2069
+    },
+    {
+      "epoch": 0.3549306642090147,
+      "grad_norm": 1.6484375,
+      "learning_rate": 1.9373793149888092e-05,
+      "loss": 1.0494,
+      "step": 2070
+    },
+    {
+      "epoch": 0.3551021282980046,
+      "grad_norm": 1.7109375,
+      "learning_rate": 1.9373163973767e-05,
+      "loss": 1.0583,
+      "step": 2071
+    },
+    {
+      "epoch": 0.35527359238699446,
+      "grad_norm": 1.8203125,
+      "learning_rate": 1.937253449195174e-05,
+      "loss": 1.0328,
+      "step": 2072
+    },
+    {
+      "epoch": 0.3554450564759843,
+      "grad_norm": 1.5703125,
+      "learning_rate": 1.9371904704462826e-05,
+      "loss": 1.0797,
+      "step": 2073
+    },
+    {
+      "epoch": 0.35561652056497417,
+      "grad_norm": 1.5625,
+      "learning_rate": 1.9371274611320813e-05,
+      "loss": 1.0437,
+      "step": 2074
+    },
+    {
+      "epoch": 0.35578798465396405,
+      "grad_norm": 1.8203125,
+      "learning_rate": 1.9370644212546246e-05,
+      "loss": 1.0953,
+      "step": 2075
+    },
+    {
+      "epoch": 0.3559594487429539,
+      "grad_norm": 1.953125,
+      "learning_rate": 1.9370013508159685e-05,
+      "loss": 1.1619,
+      "step": 2076
+    },
+    {
+      "epoch": 0.35613091283194376,
+      "grad_norm": 1.7421875,
+      "learning_rate": 1.9369382498181698e-05,
+      "loss": 1.0185,
+      "step": 2077
+    },
+    {
+      "epoch": 0.35630237692093364,
+      "grad_norm": 1.75,
+      "learning_rate": 1.936875118263287e-05,
+      "loss": 1.112,
+      "step": 2078
+    },
+    {
+      "epoch": 0.35647384100992346,
+      "grad_norm": 1.7109375,
+      "learning_rate": 1.936811956153378e-05,
+      "loss": 1.0548,
+      "step": 2079
+    },
+    {
+      "epoch": 0.35664530509891335,
+      "grad_norm": 1.6484375,
+      "learning_rate": 1.936748763490504e-05,
+      "loss": 1.0316,
+      "step": 2080
+    },
+    {
+      "epoch": 0.3568167691879032,
+      "grad_norm": 1.6953125,
+      "learning_rate": 1.936685540276725e-05,
+      "loss": 0.9346,
+      "step": 2081
+    },
+    {
+      "epoch": 0.35698823327689305,
+      "grad_norm": 1.6953125,
+      "learning_rate": 1.9366222865141032e-05,
+      "loss": 1.0405,
+      "step": 2082
+    },
+    {
+      "epoch": 0.35715969736588293,
+      "grad_norm": 1.6015625,
+      "learning_rate": 1.936559002204702e-05,
+      "loss": 0.9518,
+      "step": 2083
+    },
+    {
+      "epoch": 0.3573311614548728,
+      "grad_norm": 1.703125,
+      "learning_rate": 1.936495687350585e-05,
+      "loss": 1.0378,
+      "step": 2084
+    },
+    {
+      "epoch": 0.35750262554386264,
+      "grad_norm": 1.6015625,
+      "learning_rate": 1.9364323419538166e-05,
+      "loss": 0.9886,
+      "step": 2085
+    },
+    {
+      "epoch": 0.3576740896328525,
+      "grad_norm": 1.7578125,
+      "learning_rate": 1.9363689660164637e-05,
+      "loss": 1.0009,
+      "step": 2086
+    },
+    {
+      "epoch": 0.3578455537218424,
+      "grad_norm": 1.6015625,
+      "learning_rate": 1.9363055595405928e-05,
+      "loss": 0.9917,
+      "step": 2087
+    },
+    {
+      "epoch": 0.3580170178108322,
+      "grad_norm": 1.609375,
+      "learning_rate": 1.9362421225282717e-05,
+      "loss": 1.0459,
+      "step": 2088
+    },
+    {
+      "epoch": 0.3581884818998221,
+      "grad_norm": 1.65625,
+      "learning_rate": 1.936178654981569e-05,
+      "loss": 1.1376,
+      "step": 2089
+    },
+    {
+      "epoch": 0.358359945988812,
+      "grad_norm": 1.53125,
+      "learning_rate": 1.9361151569025556e-05,
+      "loss": 1.0498,
+      "step": 2090
+    },
+    {
+      "epoch": 0.3585314100778018,
+      "grad_norm": 1.6171875,
+      "learning_rate": 1.936051628293302e-05,
+      "loss": 1.0736,
+      "step": 2091
+    },
+    {
+      "epoch": 0.3587028741667917,
+      "grad_norm": 1.6171875,
+      "learning_rate": 1.9359880691558796e-05,
+      "loss": 1.1096,
+      "step": 2092
+    },
+    {
+      "epoch": 0.3588743382557816,
+      "grad_norm": 1.6640625,
+      "learning_rate": 1.9359244794923615e-05,
+      "loss": 1.001,
+      "step": 2093
+    },
+    {
+      "epoch": 0.3590458023447714,
+      "grad_norm": 1.7265625,
+      "learning_rate": 1.935860859304822e-05,
+      "loss": 1.0068,
+      "step": 2094
+    },
+    {
+      "epoch": 0.3592172664337613,
+      "grad_norm": 1.6953125,
+      "learning_rate": 1.9357972085953356e-05,
+      "loss": 1.0759,
+      "step": 2095
+    },
+    {
+      "epoch": 0.35938873052275117,
+      "grad_norm": 1.5859375,
+      "learning_rate": 1.9357335273659785e-05,
+      "loss": 1.0363,
+      "step": 2096
+    },
+    {
+      "epoch": 0.359560194611741,
+      "grad_norm": 1.546875,
+      "learning_rate": 1.9356698156188273e-05,
+      "loss": 0.9934,
+      "step": 2097
+    },
+    {
+      "epoch": 0.35973165870073087,
+      "grad_norm": 1.7578125,
+      "learning_rate": 1.9356060733559602e-05,
+      "loss": 1.1027,
+      "step": 2098
+    },
+    {
+      "epoch": 0.35990312278972075,
+      "grad_norm": 1.734375,
+      "learning_rate": 1.9355423005794556e-05,
+      "loss": 1.0325,
+      "step": 2099
+    },
+    {
+      "epoch": 0.3600745868787106,
+      "grad_norm": 1.6796875,
+      "learning_rate": 1.9354784972913938e-05,
+      "loss": 1.116,
+      "step": 2100
+    },
+    {
+      "epoch": 0.3600745868787106,
+      "eval_loss": 0.8986930847167969,
+      "eval_runtime": 837.0926,
+      "eval_samples_per_second": 2.985,
+      "eval_steps_per_second": 2.985,
+      "step": 2100
+    },
+    {
+      "epoch": 0.36024605096770046,
+      "grad_norm": 1.625,
+      "learning_rate": 1.9354146634938558e-05,
+      "loss": 1.1254,
+      "step": 2101
+    },
+    {
+      "epoch": 0.36041751505669034,
+      "grad_norm": 1.6640625,
+      "learning_rate": 1.9353507991889228e-05,
+      "loss": 1.0357,
+      "step": 2102
+    },
+    {
+      "epoch": 0.36058897914568017,
+      "grad_norm": 1.6328125,
+      "learning_rate": 1.935286904378678e-05,
+      "loss": 1.0318,
+      "step": 2103
+    },
+    {
+      "epoch": 0.36076044323467005,
+      "grad_norm": 1.640625,
+      "learning_rate": 1.9352229790652056e-05,
+      "loss": 1.0231,
+      "step": 2104
+    },
+    {
+      "epoch": 0.36093190732365993,
+      "grad_norm": 1.765625,
+      "learning_rate": 1.9351590232505898e-05,
+      "loss": 1.0675,
+      "step": 2105
+    },
+    {
+      "epoch": 0.36110337141264975,
+      "grad_norm": 1.65625,
+      "learning_rate": 1.935095036936917e-05,
+      "loss": 1.0321,
+      "step": 2106
+    },
+    {
+      "epoch": 0.36127483550163964,
+      "grad_norm": 1.6796875,
+      "learning_rate": 1.9350310201262737e-05,
+      "loss": 1.0532,
+      "step": 2107
+    },
+    {
+      "epoch": 0.36144629959062946,
+      "grad_norm": 1.625,
+      "learning_rate": 1.934966972820748e-05,
+      "loss": 1.0822,
+      "step": 2108
+    },
+    {
+      "epoch": 0.36161776367961934,
+      "grad_norm": 1.5859375,
+      "learning_rate": 1.934902895022428e-05,
+      "loss": 1.0376,
+      "step": 2109
+    },
+    {
+      "epoch": 0.3617892277686092,
+      "grad_norm": 1.6484375,
+      "learning_rate": 1.9348387867334047e-05,
+      "loss": 1.0499,
+      "step": 2110
+    },
+    {
+      "epoch": 0.36196069185759905,
+      "grad_norm": 1.6171875,
+      "learning_rate": 1.934774647955768e-05,
+      "loss": 1.0162,
+      "step": 2111
+    },
+    {
+      "epoch": 0.36213215594658893,
+      "grad_norm": 1.734375,
+      "learning_rate": 1.9347104786916103e-05,
+      "loss": 1.042,
+      "step": 2112
+    },
+    {
+      "epoch": 0.3623036200355788,
+      "grad_norm": 1.671875,
+      "learning_rate": 1.934646278943024e-05,
+      "loss": 1.057,
+      "step": 2113
+    },
+    {
+      "epoch": 0.36247508412456864,
+      "grad_norm": 1.703125,
+      "learning_rate": 1.9345820487121026e-05,
+      "loss": 0.9956,
+      "step": 2114
+    },
+    {
+      "epoch": 0.3626465482135585,
+      "grad_norm": 1.5546875,
+      "learning_rate": 1.9345177880009417e-05,
+      "loss": 0.9508,
+      "step": 2115
+    },
+    {
+      "epoch": 0.3628180123025484,
+      "grad_norm": 1.8359375,
+      "learning_rate": 1.9344534968116365e-05,
+      "loss": 1.1172,
+      "step": 2116
+    },
+    {
+      "epoch": 0.3629894763915382,
+      "grad_norm": 1.6875,
+      "learning_rate": 1.9343891751462838e-05,
+      "loss": 1.1314,
+      "step": 2117
+    },
+    {
+      "epoch": 0.3631609404805281,
+      "grad_norm": 1.5703125,
+      "learning_rate": 1.9343248230069815e-05,
+      "loss": 1.0684,
+      "step": 2118
+    },
+    {
+      "epoch": 0.363332404569518,
+      "grad_norm": 1.625,
+      "learning_rate": 1.9342604403958287e-05,
+      "loss": 0.9849,
+      "step": 2119
+    },
+    {
+      "epoch": 0.3635038686585078,
+      "grad_norm": 1.5390625,
+      "learning_rate": 1.9341960273149246e-05,
+      "loss": 1.0186,
+      "step": 2120
+    },
+    {
+      "epoch": 0.3636753327474977,
+      "grad_norm": 1.6640625,
+      "learning_rate": 1.9341315837663704e-05,
+      "loss": 1.0682,
+      "step": 2121
+    },
+    {
+      "epoch": 0.3638467968364876,
+      "grad_norm": 1.671875,
+      "learning_rate": 1.9340671097522678e-05,
+      "loss": 1.0753,
+      "step": 2122
+    },
+    {
+      "epoch": 0.3640182609254774,
+      "grad_norm": 1.6484375,
+      "learning_rate": 1.934002605274719e-05,
+      "loss": 1.1382,
+      "step": 2123
+    },
+    {
+      "epoch": 0.3641897250144673,
+      "grad_norm": 1.5625,
+      "learning_rate": 1.933938070335828e-05,
+      "loss": 1.0476,
+      "step": 2124
+    },
+    {
+      "epoch": 0.36436118910345716,
+      "grad_norm": 1.5859375,
+      "learning_rate": 1.9338735049377e-05,
+      "loss": 1.0797,
+      "step": 2125
+    },
+    {
+      "epoch": 0.364532653192447,
+      "grad_norm": 1.53125,
+      "learning_rate": 1.9338089090824402e-05,
+      "loss": 1.0387,
+      "step": 2126
+    },
+    {
+      "epoch": 0.36470411728143687,
+      "grad_norm": 1.5859375,
+      "learning_rate": 1.9337442827721555e-05,
+      "loss": 1.0245,
+      "step": 2127
+    },
+    {
+      "epoch": 0.36487558137042675,
+      "grad_norm": 1.6015625,
+      "learning_rate": 1.9336796260089534e-05,
+      "loss": 1.0181,
+      "step": 2128
+    },
+    {
+      "epoch": 0.3650470454594166,
+      "grad_norm": 1.65625,
+      "learning_rate": 1.933614938794943e-05,
+      "loss": 1.0161,
+      "step": 2129
+    },
+    {
+      "epoch": 0.36521850954840646,
+      "grad_norm": 1.75,
+      "learning_rate": 1.933550221132234e-05,
+      "loss": 1.1429,
+      "step": 2130
+    },
+    {
+      "epoch": 0.36538997363739634,
+      "grad_norm": 1.75,
+      "learning_rate": 1.933485473022936e-05,
+      "loss": 1.0651,
+      "step": 2131
+    },
+    {
+      "epoch": 0.36556143772638616,
+      "grad_norm": 1.5703125,
+      "learning_rate": 1.9334206944691626e-05,
+      "loss": 1.0549,
+      "step": 2132
+    },
+    {
+      "epoch": 0.36573290181537604,
+      "grad_norm": 1.6796875,
+      "learning_rate": 1.933355885473025e-05,
+      "loss": 1.0231,
+      "step": 2133
+    },
+    {
+      "epoch": 0.3659043659043659,
+      "grad_norm": 1.640625,
+      "learning_rate": 1.933291046036637e-05,
+      "loss": 1.0812,
+      "step": 2134
+    },
+    {
+      "epoch": 0.36607582999335575,
+      "grad_norm": 1.6953125,
+      "learning_rate": 1.9332261761621138e-05,
+      "loss": 1.1038,
+      "step": 2135
+    },
+    {
+      "epoch": 0.36624729408234563,
+      "grad_norm": 1.6328125,
+      "learning_rate": 1.9331612758515706e-05,
+      "loss": 1.0191,
+      "step": 2136
+    },
+    {
+      "epoch": 0.3664187581713355,
+      "grad_norm": 1.640625,
+      "learning_rate": 1.933096345107125e-05,
+      "loss": 1.0076,
+      "step": 2137
+    },
+    {
+      "epoch": 0.36659022226032534,
+      "grad_norm": 1.53125,
+      "learning_rate": 1.933031383930893e-05,
+      "loss": 1.016,
+      "step": 2138
+    },
+    {
+      "epoch": 0.3667616863493152,
+      "grad_norm": 1.8984375,
+      "learning_rate": 1.9329663923249945e-05,
+      "loss": 1.0134,
+      "step": 2139
+    },
+    {
+      "epoch": 0.3669331504383051,
+      "grad_norm": 1.8046875,
+      "learning_rate": 1.9329013702915485e-05,
+      "loss": 1.2297,
+      "step": 2140
+    },
+    {
+      "epoch": 0.3671046145272949,
+      "grad_norm": 1.609375,
+      "learning_rate": 1.932836317832676e-05,
+      "loss": 0.9963,
+      "step": 2141
+    },
+    {
+      "epoch": 0.3672760786162848,
+      "grad_norm": 1.6328125,
+      "learning_rate": 1.9327712349504982e-05,
+      "loss": 1.0519,
+      "step": 2142
+    },
+    {
+      "epoch": 0.3674475427052747,
+      "grad_norm": 1.734375,
+      "learning_rate": 1.9327061216471382e-05,
+      "loss": 1.0634,
+      "step": 2143
+    },
+    {
+      "epoch": 0.3676190067942645,
+      "grad_norm": 1.609375,
+      "learning_rate": 1.932640977924719e-05,
+      "loss": 1.117,
+      "step": 2144
+    },
+    {
+      "epoch": 0.3677904708832544,
+      "grad_norm": 1.609375,
+      "learning_rate": 1.932575803785366e-05,
+      "loss": 1.0166,
+      "step": 2145
+    },
+    {
+      "epoch": 0.3679619349722443,
+      "grad_norm": 1.7265625,
+      "learning_rate": 1.932510599231204e-05,
+      "loss": 1.1124,
+      "step": 2146
+    },
+    {
+      "epoch": 0.3681333990612341,
+      "grad_norm": 1.6015625,
+      "learning_rate": 1.9324453642643598e-05,
+      "loss": 0.9047,
+      "step": 2147
+    },
+    {
+      "epoch": 0.368304863150224,
+      "grad_norm": 1.578125,
+      "learning_rate": 1.932380098886961e-05,
+      "loss": 1.0224,
+      "step": 2148
+    },
+    {
+      "epoch": 0.36847632723921386,
+      "grad_norm": 1.734375,
+      "learning_rate": 1.9323148031011363e-05,
+      "loss": 1.0019,
+      "step": 2149
+    },
+    {
+      "epoch": 0.3686477913282037,
+      "grad_norm": 1.671875,
+      "learning_rate": 1.932249476909015e-05,
+      "loss": 1.079,
+      "step": 2150
+    },
+    {
+      "epoch": 0.36881925541719357,
+      "grad_norm": 1.7109375,
+      "learning_rate": 1.9321841203127277e-05,
+      "loss": 1.0723,
+      "step": 2151
+    },
+    {
+      "epoch": 0.36899071950618345,
+      "grad_norm": 1.65625,
+      "learning_rate": 1.9321187333144064e-05,
+      "loss": 0.9337,
+      "step": 2152
+    },
+    {
+      "epoch": 0.3691621835951733,
+      "grad_norm": 1.6484375,
+      "learning_rate": 1.9320533159161824e-05,
+      "loss": 1.0433,
+      "step": 2153
+    },
+    {
+      "epoch": 0.36933364768416316,
+      "grad_norm": 1.6484375,
+      "learning_rate": 1.9319878681201905e-05,
+      "loss": 1.0609,
+      "step": 2154
+    },
+    {
+      "epoch": 0.369505111773153,
+      "grad_norm": 1.703125,
+      "learning_rate": 1.9319223899285643e-05,
+      "loss": 1.0004,
+      "step": 2155
+    },
+    {
+      "epoch": 0.36967657586214286,
+      "grad_norm": 1.640625,
+      "learning_rate": 1.93185688134344e-05,
+      "loss": 1.0564,
+      "step": 2156
+    },
+    {
+      "epoch": 0.36984803995113275,
+      "grad_norm": 1.8125,
+      "learning_rate": 1.9317913423669535e-05,
+      "loss": 1.1132,
+      "step": 2157
+    },
+    {
+      "epoch": 0.37001950404012257,
+      "grad_norm": 1.59375,
+      "learning_rate": 1.931725773001243e-05,
+      "loss": 0.9931,
+      "step": 2158
+    },
+    {
+      "epoch": 0.37019096812911245,
+      "grad_norm": 1.6953125,
+      "learning_rate": 1.931660173248446e-05,
+      "loss": 1.121,
+      "step": 2159
+    },
+    {
+      "epoch": 0.37036243221810233,
+      "grad_norm": 1.5546875,
+      "learning_rate": 1.931594543110703e-05,
+      "loss": 1.0856,
+      "step": 2160
+    },
+    {
+      "epoch": 0.37053389630709216,
+      "grad_norm": 1.6640625,
+      "learning_rate": 1.9315288825901534e-05,
+      "loss": 1.1187,
+      "step": 2161
+    },
+    {
+      "epoch": 0.37070536039608204,
+      "grad_norm": 1.7734375,
+      "learning_rate": 1.931463191688939e-05,
+      "loss": 1.0063,
+      "step": 2162
+    },
+    {
+      "epoch": 0.3708768244850719,
+      "grad_norm": 1.703125,
+      "learning_rate": 1.9313974704092028e-05,
+      "loss": 1.0472,
+      "step": 2163
+    },
+    {
+      "epoch": 0.37104828857406175,
+      "grad_norm": 1.65625,
+      "learning_rate": 1.9313317187530875e-05,
+      "loss": 1.1154,
+      "step": 2164
+    },
+    {
+      "epoch": 0.3712197526630516,
+      "grad_norm": 1.7109375,
+      "learning_rate": 1.931265936722738e-05,
+      "loss": 1.0989,
+      "step": 2165
+    },
+    {
+      "epoch": 0.3713912167520415,
+      "grad_norm": 1.71875,
+      "learning_rate": 1.931200124320299e-05,
+      "loss": 1.0552,
+      "step": 2166
+    },
+    {
+      "epoch": 0.37156268084103133,
+      "grad_norm": 1.6484375,
+      "learning_rate": 1.931134281547918e-05,
+      "loss": 1.1136,
+      "step": 2167
+    },
+    {
+      "epoch": 0.3717341449300212,
+      "grad_norm": 1.6953125,
+      "learning_rate": 1.9310684084077413e-05,
+      "loss": 0.9891,
+      "step": 2168
+    },
+    {
+      "epoch": 0.3719056090190111,
+      "grad_norm": 1.734375,
+      "learning_rate": 1.931002504901918e-05,
+      "loss": 1.0897,
+      "step": 2169
+    },
+    {
+      "epoch": 0.3720770731080009,
+      "grad_norm": 1.6484375,
+      "learning_rate": 1.930936571032597e-05,
+      "loss": 1.0432,
+      "step": 2170
+    },
+    {
+      "epoch": 0.3722485371969908,
+      "grad_norm": 1.6484375,
+      "learning_rate": 1.930870606801929e-05,
+      "loss": 0.9868,
+      "step": 2171
+    },
+    {
+      "epoch": 0.3724200012859807,
+      "grad_norm": 1.6328125,
+      "learning_rate": 1.9308046122120648e-05,
+      "loss": 1.037,
+      "step": 2172
+    },
+    {
+      "epoch": 0.3725914653749705,
+      "grad_norm": 1.59375,
+      "learning_rate": 1.9307385872651574e-05,
+      "loss": 1.0493,
+      "step": 2173
+    },
+    {
+      "epoch": 0.3727629294639604,
+      "grad_norm": 1.65625,
+      "learning_rate": 1.9306725319633603e-05,
+      "loss": 1.011,
+      "step": 2174
+    },
+    {
+      "epoch": 0.37293439355295027,
+      "grad_norm": 1.578125,
+      "learning_rate": 1.9306064463088267e-05,
+      "loss": 1.0288,
+      "step": 2175
+    },
+    {
+      "epoch": 0.3731058576419401,
+      "grad_norm": 1.671875,
+      "learning_rate": 1.9305403303037128e-05,
+      "loss": 1.0481,
+      "step": 2176
+    },
+    {
+      "epoch": 0.37327732173093,
+      "grad_norm": 1.6796875,
+      "learning_rate": 1.930474183950174e-05,
+      "loss": 0.994,
+      "step": 2177
+    },
+    {
+      "epoch": 0.37344878581991986,
+      "grad_norm": 1.6328125,
+      "learning_rate": 1.930408007250369e-05,
+      "loss": 1.0024,
+      "step": 2178
+    },
+    {
+      "epoch": 0.3736202499089097,
+      "grad_norm": 1.609375,
+      "learning_rate": 1.9303418002064553e-05,
+      "loss": 1.0486,
+      "step": 2179
+    },
+    {
+      "epoch": 0.37379171399789957,
+      "grad_norm": 1.578125,
+      "learning_rate": 1.9302755628205917e-05,
+      "loss": 1.0745,
+      "step": 2180
+    },
+    {
+      "epoch": 0.37396317808688945,
+      "grad_norm": 1.5859375,
+      "learning_rate": 1.9302092950949393e-05,
+      "loss": 1.0946,
+      "step": 2181
+    },
+    {
+      "epoch": 0.3741346421758793,
+      "grad_norm": 1.625,
+      "learning_rate": 1.9301429970316588e-05,
+      "loss": 1.0594,
+      "step": 2182
+    },
+    {
+      "epoch": 0.37430610626486915,
+      "grad_norm": 1.5234375,
+      "learning_rate": 1.9300766686329124e-05,
+      "loss": 1.0049,
+      "step": 2183
+    },
+    {
+      "epoch": 0.37447757035385904,
+      "grad_norm": 1.671875,
+      "learning_rate": 1.9300103099008634e-05,
+      "loss": 1.0796,
+      "step": 2184
+    },
+    {
+      "epoch": 0.37464903444284886,
+      "grad_norm": 1.671875,
+      "learning_rate": 1.9299439208376763e-05,
+      "loss": 1.0713,
+      "step": 2185
+    },
+    {
+      "epoch": 0.37482049853183874,
+      "grad_norm": 1.59375,
+      "learning_rate": 1.9298775014455163e-05,
+      "loss": 1.1084,
+      "step": 2186
+    },
+    {
+      "epoch": 0.3749919626208286,
+      "grad_norm": 1.65625,
+      "learning_rate": 1.9298110517265492e-05,
+      "loss": 0.9409,
+      "step": 2187
+    },
+    {
+      "epoch": 0.37516342670981845,
+      "grad_norm": 1.6015625,
+      "learning_rate": 1.9297445716829423e-05,
+      "loss": 0.9509,
+      "step": 2188
+    },
+    {
+      "epoch": 0.37533489079880833,
+      "grad_norm": 1.609375,
+      "learning_rate": 1.9296780613168638e-05,
+      "loss": 1.0959,
+      "step": 2189
+    },
+    {
+      "epoch": 0.3755063548877982,
+      "grad_norm": 1.59375,
+      "learning_rate": 1.929611520630483e-05,
+      "loss": 1.0904,
+      "step": 2190
+    },
+    {
+      "epoch": 0.37567781897678804,
+      "grad_norm": 1.6953125,
+      "learning_rate": 1.92954494962597e-05,
+      "loss": 1.0285,
+      "step": 2191
+    },
+    {
+      "epoch": 0.3758492830657779,
+      "grad_norm": 1.5703125,
+      "learning_rate": 1.929478348305496e-05,
+      "loss": 1.0246,
+      "step": 2192
+    },
+    {
+      "epoch": 0.3760207471547678,
+      "grad_norm": 1.7578125,
+      "learning_rate": 1.9294117166712322e-05,
+      "loss": 1.1087,
+      "step": 2193
+    },
+    {
+      "epoch": 0.3761922112437576,
+      "grad_norm": 1.5859375,
+      "learning_rate": 1.929345054725353e-05,
+      "loss": 1.0125,
+      "step": 2194
+    },
+    {
+      "epoch": 0.3763636753327475,
+      "grad_norm": 1.734375,
+      "learning_rate": 1.929278362470032e-05,
+      "loss": 1.0662,
+      "step": 2195
+    },
+    {
+      "epoch": 0.3765351394217374,
+      "grad_norm": 1.6171875,
+      "learning_rate": 1.9292116399074442e-05,
+      "loss": 0.9725,
+      "step": 2196
+    },
+    {
+      "epoch": 0.3767066035107272,
+      "grad_norm": 1.59375,
+      "learning_rate": 1.9291448870397658e-05,
+      "loss": 0.9811,
+      "step": 2197
+    },
+    {
+      "epoch": 0.3768780675997171,
+      "grad_norm": 1.65625,
+      "learning_rate": 1.929078103869174e-05,
+      "loss": 1.0886,
+      "step": 2198
+    },
+    {
+      "epoch": 0.377049531688707,
+      "grad_norm": 1.578125,
+      "learning_rate": 1.929011290397846e-05,
+      "loss": 1.0572,
+      "step": 2199
+    },
+    {
+      "epoch": 0.3772209957776968,
+      "grad_norm": 1.4765625,
+      "learning_rate": 1.9289444466279622e-05,
+      "loss": 0.9523,
+      "step": 2200
+    },
+    {
+      "epoch": 0.3773924598666867,
+      "grad_norm": 1.5390625,
+      "learning_rate": 1.9288775725617014e-05,
+      "loss": 1.0021,
+      "step": 2201
+    },
+    {
+      "epoch": 0.3775639239556765,
+      "grad_norm": 1.5546875,
+      "learning_rate": 1.9288106682012456e-05,
+      "loss": 1.1431,
+      "step": 2202
+    },
+    {
+      "epoch": 0.3777353880446664,
+      "grad_norm": 1.7578125,
+      "learning_rate": 1.928743733548776e-05,
+      "loss": 1.1734,
+      "step": 2203
+    },
+    {
+      "epoch": 0.37790685213365627,
+      "grad_norm": 1.625,
+      "learning_rate": 1.9286767686064757e-05,
+      "loss": 1.0438,
+      "step": 2204
+    },
+    {
+      "epoch": 0.3780783162226461,
+      "grad_norm": 1.671875,
+      "learning_rate": 1.9286097733765296e-05,
+      "loss": 1.0226,
+      "step": 2205
+    },
+    {
+      "epoch": 0.378249780311636,
+      "grad_norm": 1.7734375,
+      "learning_rate": 1.9285427478611214e-05,
+      "loss": 1.0586,
+      "step": 2206
+    },
+    {
+      "epoch": 0.37842124440062586,
+      "grad_norm": 1.609375,
+      "learning_rate": 1.9284756920624378e-05,
+      "loss": 1.0845,
+      "step": 2207
+    },
+    {
+      "epoch": 0.3785927084896157,
+      "grad_norm": 1.578125,
+      "learning_rate": 1.9284086059826654e-05,
+      "loss": 1.048,
+      "step": 2208
+    },
+    {
+      "epoch": 0.37876417257860556,
+      "grad_norm": 2.296875,
+      "learning_rate": 1.9283414896239924e-05,
+      "loss": 0.9909,
+      "step": 2209
+    },
+    {
+      "epoch": 0.37893563666759544,
+      "grad_norm": 1.578125,
+      "learning_rate": 1.928274342988608e-05,
+      "loss": 1.0743,
+      "step": 2210
+    },
+    {
+      "epoch": 0.37910710075658527,
+      "grad_norm": 2.671875,
+      "learning_rate": 1.928207166078701e-05,
+      "loss": 1.0515,
+      "step": 2211
+    },
+    {
+      "epoch": 0.37927856484557515,
+      "grad_norm": 1.5625,
+      "learning_rate": 1.9281399588964636e-05,
+      "loss": 0.998,
+      "step": 2212
+    },
+    {
+      "epoch": 0.37945002893456503,
+      "grad_norm": 1.6640625,
+      "learning_rate": 1.9280727214440865e-05,
+      "loss": 1.0134,
+      "step": 2213
+    },
+    {
+      "epoch": 0.37962149302355486,
+      "grad_norm": 1.640625,
+      "learning_rate": 1.928005453723764e-05,
+      "loss": 0.9262,
+      "step": 2214
+    },
+    {
+      "epoch": 0.37979295711254474,
+      "grad_norm": 1.7265625,
+      "learning_rate": 1.9279381557376883e-05,
+      "loss": 1.1135,
+      "step": 2215
+    },
+    {
+      "epoch": 0.3799644212015346,
+      "grad_norm": 1.59375,
+      "learning_rate": 1.9278708274880556e-05,
+      "loss": 1.1122,
+      "step": 2216
+    },
+    {
+      "epoch": 0.38013588529052444,
+      "grad_norm": 1.53125,
+      "learning_rate": 1.927803468977061e-05,
+      "loss": 1.0032,
+      "step": 2217
+    },
+    {
+      "epoch": 0.3803073493795143,
+      "grad_norm": 1.6015625,
+      "learning_rate": 1.9277360802069013e-05,
+      "loss": 1.0734,
+      "step": 2218
+    },
+    {
+      "epoch": 0.3804788134685042,
+      "grad_norm": 1.6328125,
+      "learning_rate": 1.927668661179775e-05,
+      "loss": 1.0122,
+      "step": 2219
+    },
+    {
+      "epoch": 0.38065027755749403,
+      "grad_norm": 1.5,
+      "learning_rate": 1.9276012118978794e-05,
+      "loss": 0.9827,
+      "step": 2220
+    },
+    {
+      "epoch": 0.3808217416464839,
+      "grad_norm": 1.6640625,
+      "learning_rate": 1.927533732363416e-05,
+      "loss": 1.1203,
+      "step": 2221
+    },
+    {
+      "epoch": 0.3809932057354738,
+      "grad_norm": 1.6171875,
+      "learning_rate": 1.927466222578585e-05,
+      "loss": 1.0789,
+      "step": 2222
+    },
+    {
+      "epoch": 0.3811646698244636,
+      "grad_norm": 1.65625,
+      "learning_rate": 1.9273986825455874e-05,
+      "loss": 1.067,
+      "step": 2223
+    },
+    {
+      "epoch": 0.3813361339134535,
+      "grad_norm": 1.6171875,
+      "learning_rate": 1.927331112266627e-05,
+      "loss": 1.0309,
+      "step": 2224
+    },
+    {
+      "epoch": 0.3815075980024434,
+      "grad_norm": 1.5625,
+      "learning_rate": 1.9272635117439066e-05,
+      "loss": 1.0483,
+      "step": 2225
+    },
+    {
+      "epoch": 0.3816790620914332,
+      "grad_norm": 1.5078125,
+      "learning_rate": 1.9271958809796315e-05,
+      "loss": 1.0048,
+      "step": 2226
+    },
+    {
+      "epoch": 0.3818505261804231,
+      "grad_norm": 1.625,
+      "learning_rate": 1.9271282199760076e-05,
+      "loss": 1.0654,
+      "step": 2227
+    },
+    {
+      "epoch": 0.38202199026941297,
+      "grad_norm": 1.5546875,
+      "learning_rate": 1.9270605287352406e-05,
+      "loss": 1.0612,
+      "step": 2228
+    },
+    {
+      "epoch": 0.3821934543584028,
+      "grad_norm": 1.6015625,
+      "learning_rate": 1.9269928072595392e-05,
+      "loss": 1.0226,
+      "step": 2229
+    },
+    {
+      "epoch": 0.3823649184473927,
+      "grad_norm": 1.5703125,
+      "learning_rate": 1.926925055551111e-05,
+      "loss": 0.9787,
+      "step": 2230
+    },
+    {
+      "epoch": 0.38253638253638256,
+      "grad_norm": 1.53125,
+      "learning_rate": 1.9268572736121668e-05,
+      "loss": 1.0347,
+      "step": 2231
+    },
+    {
+      "epoch": 0.3827078466253724,
+      "grad_norm": 1.6015625,
+      "learning_rate": 1.9267894614449168e-05,
+      "loss": 1.1223,
+      "step": 2232
+    },
+    {
+      "epoch": 0.38287931071436226,
+      "grad_norm": 1.6015625,
+      "learning_rate": 1.9267216190515726e-05,
+      "loss": 1.0425,
+      "step": 2233
+    },
+    {
+      "epoch": 0.38305077480335215,
+      "grad_norm": 1.6484375,
+      "learning_rate": 1.926653746434346e-05,
+      "loss": 1.0454,
+      "step": 2234
+    },
+    {
+      "epoch": 0.38322223889234197,
+      "grad_norm": 1.6484375,
+      "learning_rate": 1.9265858435954515e-05,
+      "loss": 0.9754,
+      "step": 2235
+    },
+    {
+      "epoch": 0.38339370298133185,
+      "grad_norm": 1.75,
+      "learning_rate": 1.926517910537104e-05,
+      "loss": 1.148,
+      "step": 2236
+    },
+    {
+      "epoch": 0.38356516707032173,
+      "grad_norm": 1.7421875,
+      "learning_rate": 1.926449947261518e-05,
+      "loss": 1.041,
+      "step": 2237
+    },
+    {
+      "epoch": 0.38373663115931156,
+      "grad_norm": 1.703125,
+      "learning_rate": 1.9263819537709104e-05,
+      "loss": 1.0121,
+      "step": 2238
+    },
+    {
+      "epoch": 0.38390809524830144,
+      "grad_norm": 1.703125,
+      "learning_rate": 1.9263139300674994e-05,
+      "loss": 1.0594,
+      "step": 2239
+    },
+    {
+      "epoch": 0.3840795593372913,
+      "grad_norm": 1.7265625,
+      "learning_rate": 1.9262458761535026e-05,
+      "loss": 1.0862,
+      "step": 2240
+    },
+    {
+      "epoch": 0.38425102342628115,
+      "grad_norm": 1.546875,
+      "learning_rate": 1.92617779203114e-05,
+      "loss": 1.0087,
+      "step": 2241
+    },
+    {
+      "epoch": 0.384422487515271,
+      "grad_norm": 1.734375,
+      "learning_rate": 1.9261096777026318e-05,
+      "loss": 1.0964,
+      "step": 2242
+    },
+    {
+      "epoch": 0.3845939516042609,
+      "grad_norm": 1.546875,
+      "learning_rate": 1.9260415331701996e-05,
+      "loss": 0.9698,
+      "step": 2243
+    },
+    {
+      "epoch": 0.38476541569325073,
+      "grad_norm": 1.609375,
+      "learning_rate": 1.925973358436066e-05,
+      "loss": 0.9941,
+      "step": 2244
+    },
+    {
+      "epoch": 0.3849368797822406,
+      "grad_norm": 1.6484375,
+      "learning_rate": 1.9259051535024542e-05,
+      "loss": 1.0449,
+      "step": 2245
+    },
+    {
+      "epoch": 0.3851083438712305,
+      "grad_norm": 1.71875,
+      "learning_rate": 1.9258369183715887e-05,
+      "loss": 1.1611,
+      "step": 2246
+    },
+    {
+      "epoch": 0.3852798079602203,
+      "grad_norm": 1.6171875,
+      "learning_rate": 1.9257686530456947e-05,
+      "loss": 1.0714,
+      "step": 2247
+    },
+    {
+      "epoch": 0.3854512720492102,
+      "grad_norm": 1.5234375,
+      "learning_rate": 1.9257003575269992e-05,
+      "loss": 0.9581,
+      "step": 2248
+    },
+    {
+      "epoch": 0.38562273613820003,
+      "grad_norm": 1.578125,
+      "learning_rate": 1.925632031817729e-05,
+      "loss": 1.0036,
+      "step": 2249
+    },
+    {
+      "epoch": 0.3857942002271899,
+      "grad_norm": 1.5859375,
+      "learning_rate": 1.9255636759201127e-05,
+      "loss": 1.0602,
+      "step": 2250
+    },
+    {
+      "epoch": 0.3859656643161798,
+      "grad_norm": 1.703125,
+      "learning_rate": 1.925495289836379e-05,
+      "loss": 1.1031,
+      "step": 2251
+    },
+    {
+      "epoch": 0.3861371284051696,
+      "grad_norm": 1.65625,
+      "learning_rate": 1.9254268735687595e-05,
+      "loss": 1.1401,
+      "step": 2252
+    },
+    {
+      "epoch": 0.3863085924941595,
+      "grad_norm": 1.59375,
+      "learning_rate": 1.9253584271194848e-05,
+      "loss": 1.0456,
+      "step": 2253
+    },
+    {
+      "epoch": 0.3864800565831494,
+      "grad_norm": 1.6875,
+      "learning_rate": 1.925289950490787e-05,
+      "loss": 1.1156,
+      "step": 2254
+    },
+    {
+      "epoch": 0.3866515206721392,
+      "grad_norm": 1.75,
+      "learning_rate": 1.9252214436848996e-05,
+      "loss": 1.0134,
+      "step": 2255
+    },
+    {
+      "epoch": 0.3868229847611291,
+      "grad_norm": 1.5859375,
+      "learning_rate": 1.925152906704057e-05,
+      "loss": 1.0402,
+      "step": 2256
+    },
+    {
+      "epoch": 0.38699444885011897,
+      "grad_norm": 1.6953125,
+      "learning_rate": 1.925084339550494e-05,
+      "loss": 1.0452,
+      "step": 2257
+    },
+    {
+      "epoch": 0.3871659129391088,
+      "grad_norm": 1.671875,
+      "learning_rate": 1.9250157422264477e-05,
+      "loss": 1.0607,
+      "step": 2258
+    },
+    {
+      "epoch": 0.3873373770280987,
+      "grad_norm": 1.578125,
+      "learning_rate": 1.9249471147341544e-05,
+      "loss": 1.0349,
+      "step": 2259
+    },
+    {
+      "epoch": 0.38750884111708855,
+      "grad_norm": 1.6640625,
+      "learning_rate": 1.9248784570758526e-05,
+      "loss": 0.984,
+      "step": 2260
+    },
+    {
+      "epoch": 0.3876803052060784,
+      "grad_norm": 1.671875,
+      "learning_rate": 1.9248097692537815e-05,
+      "loss": 1.0693,
+      "step": 2261
+    },
+    {
+      "epoch": 0.38785176929506826,
+      "grad_norm": 1.65625,
+      "learning_rate": 1.9247410512701815e-05,
+      "loss": 1.0794,
+      "step": 2262
+    },
+    {
+      "epoch": 0.38802323338405814,
+      "grad_norm": 1.578125,
+      "learning_rate": 1.9246723031272934e-05,
+      "loss": 0.9385,
+      "step": 2263
+    },
+    {
+      "epoch": 0.38819469747304797,
+      "grad_norm": 1.578125,
+      "learning_rate": 1.9246035248273598e-05,
+      "loss": 1.04,
+      "step": 2264
+    },
+    {
+      "epoch": 0.38836616156203785,
+      "grad_norm": 1.703125,
+      "learning_rate": 1.9245347163726233e-05,
+      "loss": 1.0896,
+      "step": 2265
+    },
+    {
+      "epoch": 0.38853762565102773,
+      "grad_norm": 1.625,
+      "learning_rate": 1.9244658777653282e-05,
+      "loss": 1.0744,
+      "step": 2266
+    },
+    {
+      "epoch": 0.38870908974001755,
+      "grad_norm": 1.640625,
+      "learning_rate": 1.9243970090077196e-05,
+      "loss": 1.0505,
+      "step": 2267
+    },
+    {
+      "epoch": 0.38888055382900744,
+      "grad_norm": 1.5625,
+      "learning_rate": 1.9243281101020436e-05,
+      "loss": 1.0416,
+      "step": 2268
+    },
+    {
+      "epoch": 0.3890520179179973,
+      "grad_norm": 1.609375,
+      "learning_rate": 1.9242591810505474e-05,
+      "loss": 0.9873,
+      "step": 2269
+    },
+    {
+      "epoch": 0.38922348200698714,
+      "grad_norm": 1.6875,
+      "learning_rate": 1.9241902218554787e-05,
+      "loss": 1.0796,
+      "step": 2270
+    },
+    {
+      "epoch": 0.389394946095977,
+      "grad_norm": 1.6640625,
+      "learning_rate": 1.9241212325190867e-05,
+      "loss": 1.1094,
+      "step": 2271
+    },
+    {
+      "epoch": 0.3895664101849669,
+      "grad_norm": 1.6484375,
+      "learning_rate": 1.9240522130436213e-05,
+      "loss": 1.0207,
+      "step": 2272
+    },
+    {
+      "epoch": 0.38973787427395673,
+      "grad_norm": 1.734375,
+      "learning_rate": 1.9239831634313338e-05,
+      "loss": 0.9764,
+      "step": 2273
+    },
+    {
+      "epoch": 0.3899093383629466,
+      "grad_norm": 1.6484375,
+      "learning_rate": 1.923914083684476e-05,
+      "loss": 1.0445,
+      "step": 2274
+    },
+    {
+      "epoch": 0.3900808024519365,
+      "grad_norm": 1.6171875,
+      "learning_rate": 1.9238449738053003e-05,
+      "loss": 1.0329,
+      "step": 2275
+    },
+    {
+      "epoch": 0.3902522665409263,
+      "grad_norm": 1.671875,
+      "learning_rate": 1.9237758337960616e-05,
+      "loss": 1.0006,
+      "step": 2276
+    },
+    {
+      "epoch": 0.3904237306299162,
+      "grad_norm": 1.5625,
+      "learning_rate": 1.9237066636590142e-05,
+      "loss": 0.9341,
+      "step": 2277
+    },
+    {
+      "epoch": 0.3905951947189061,
+      "grad_norm": 1.6875,
+      "learning_rate": 1.923637463396414e-05,
+      "loss": 1.0451,
+      "step": 2278
+    },
+    {
+      "epoch": 0.3907666588078959,
+      "grad_norm": 1.578125,
+      "learning_rate": 1.9235682330105182e-05,
+      "loss": 1.0136,
+      "step": 2279
+    },
+    {
+      "epoch": 0.3909381228968858,
+      "grad_norm": 1.703125,
+      "learning_rate": 1.9234989725035843e-05,
+      "loss": 1.0312,
+      "step": 2280
+    },
+    {
+      "epoch": 0.39110958698587567,
+      "grad_norm": 1.6640625,
+      "learning_rate": 1.923429681877872e-05,
+      "loss": 0.9722,
+      "step": 2281
+    },
+    {
+      "epoch": 0.3912810510748655,
+      "grad_norm": 1.65625,
+      "learning_rate": 1.9233603611356394e-05,
+      "loss": 1.036,
+      "step": 2282
+    },
+    {
+      "epoch": 0.3914525151638554,
+      "grad_norm": 1.671875,
+      "learning_rate": 1.923291010279149e-05,
+      "loss": 1.0469,
+      "step": 2283
+    },
+    {
+      "epoch": 0.39162397925284526,
+      "grad_norm": 1.5703125,
+      "learning_rate": 1.923221629310662e-05,
+      "loss": 1.0114,
+      "step": 2284
+    },
+    {
+      "epoch": 0.3917954433418351,
+      "grad_norm": 1.6328125,
+      "learning_rate": 1.923152218232441e-05,
+      "loss": 0.9944,
+      "step": 2285
+    },
+    {
+      "epoch": 0.39196690743082496,
+      "grad_norm": 1.65625,
+      "learning_rate": 1.9230827770467497e-05,
+      "loss": 0.985,
+      "step": 2286
+    },
+    {
+      "epoch": 0.39213837151981484,
+      "grad_norm": 1.6015625,
+      "learning_rate": 1.9230133057558533e-05,
+      "loss": 0.9982,
+      "step": 2287
+    },
+    {
+      "epoch": 0.39230983560880467,
+      "grad_norm": 1.671875,
+      "learning_rate": 1.922943804362017e-05,
+      "loss": 1.0644,
+      "step": 2288
+    },
+    {
+      "epoch": 0.39248129969779455,
+      "grad_norm": 1.6015625,
+      "learning_rate": 1.922874272867508e-05,
+      "loss": 1.0168,
+      "step": 2289
+    },
+    {
+      "epoch": 0.39265276378678443,
+      "grad_norm": 1.5703125,
+      "learning_rate": 1.9228047112745938e-05,
+      "loss": 1.0118,
+      "step": 2290
+    },
+    {
+      "epoch": 0.39282422787577426,
+      "grad_norm": 1.53125,
+      "learning_rate": 1.9227351195855428e-05,
+      "loss": 1.0358,
+      "step": 2291
+    },
+    {
+      "epoch": 0.39299569196476414,
+      "grad_norm": 1.546875,
+      "learning_rate": 1.9226654978026248e-05,
+      "loss": 0.9959,
+      "step": 2292
+    },
+    {
+      "epoch": 0.393167156053754,
+      "grad_norm": 1.578125,
+      "learning_rate": 1.9225958459281105e-05,
+      "loss": 1.0322,
+      "step": 2293
+    },
+    {
+      "epoch": 0.39333862014274384,
+      "grad_norm": 1.5703125,
+      "learning_rate": 1.9225261639642717e-05,
+      "loss": 0.9162,
+      "step": 2294
+    },
+    {
+      "epoch": 0.3935100842317337,
+      "grad_norm": 1.546875,
+      "learning_rate": 1.922456451913381e-05,
+      "loss": 0.9757,
+      "step": 2295
+    },
+    {
+      "epoch": 0.39368154832072355,
+      "grad_norm": 1.6953125,
+      "learning_rate": 1.9223867097777113e-05,
+      "loss": 1.0772,
+      "step": 2296
+    },
+    {
+      "epoch": 0.39385301240971343,
+      "grad_norm": 1.5546875,
+      "learning_rate": 1.9223169375595376e-05,
+      "loss": 0.9295,
+      "step": 2297
+    },
+    {
+      "epoch": 0.3940244764987033,
+      "grad_norm": 1.6171875,
+      "learning_rate": 1.9222471352611357e-05,
+      "loss": 1.0406,
+      "step": 2298
+    },
+    {
+      "epoch": 0.39419594058769314,
+      "grad_norm": 1.578125,
+      "learning_rate": 1.9221773028847817e-05,
+      "loss": 0.9843,
+      "step": 2299
+    },
+    {
+      "epoch": 0.394367404676683,
+      "grad_norm": 1.5859375,
+      "learning_rate": 1.9221074404327532e-05,
+      "loss": 1.0198,
+      "step": 2300
+    },
+    {
+      "epoch": 0.3945388687656729,
+      "grad_norm": 1.5703125,
+      "learning_rate": 1.922037547907329e-05,
+      "loss": 1.0301,
+      "step": 2301
+    },
+    {
+      "epoch": 0.3947103328546627,
+      "grad_norm": 1.9375,
+      "learning_rate": 1.921967625310788e-05,
+      "loss": 0.9537,
+      "step": 2302
+    },
+    {
+      "epoch": 0.3948817969436526,
+      "grad_norm": 1.5546875,
+      "learning_rate": 1.9218976726454112e-05,
+      "loss": 0.9385,
+      "step": 2303
+    },
+    {
+      "epoch": 0.3950532610326425,
+      "grad_norm": 1.59375,
+      "learning_rate": 1.9218276899134794e-05,
+      "loss": 1.0448,
+      "step": 2304
+    },
+    {
+      "epoch": 0.3952247251216323,
+      "grad_norm": 8.9375,
+      "learning_rate": 1.9217576771172756e-05,
+      "loss": 1.1329,
+      "step": 2305
+    },
+    {
+      "epoch": 0.3953961892106222,
+      "grad_norm": 1.59375,
+      "learning_rate": 1.9216876342590827e-05,
+      "loss": 0.9946,
+      "step": 2306
+    },
+    {
+      "epoch": 0.3955676532996121,
+      "grad_norm": 1.609375,
+      "learning_rate": 1.9216175613411857e-05,
+      "loss": 1.1019,
+      "step": 2307
+    },
+    {
+      "epoch": 0.3957391173886019,
+      "grad_norm": 1.59375,
+      "learning_rate": 1.9215474583658693e-05,
+      "loss": 0.9478,
+      "step": 2308
+    },
+    {
+      "epoch": 0.3959105814775918,
+      "grad_norm": 1.6875,
+      "learning_rate": 1.92147732533542e-05,
+      "loss": 1.1428,
+      "step": 2309
+    },
+    {
+      "epoch": 0.39608204556658166,
+      "grad_norm": 1.5390625,
+      "learning_rate": 1.921407162252125e-05,
+      "loss": 1.1084,
+      "step": 2310
+    },
+    {
+      "epoch": 0.3962535096555715,
+      "grad_norm": 1.6796875,
+      "learning_rate": 1.921336969118273e-05,
+      "loss": 1.0258,
+      "step": 2311
+    },
+    {
+      "epoch": 0.39642497374456137,
+      "grad_norm": 1.53125,
+      "learning_rate": 1.9212667459361532e-05,
+      "loss": 0.9763,
+      "step": 2312
+    },
+    {
+      "epoch": 0.39659643783355125,
+      "grad_norm": 1.6015625,
+      "learning_rate": 1.9211964927080552e-05,
+      "loss": 1.0425,
+      "step": 2313
+    },
+    {
+      "epoch": 0.3967679019225411,
+      "grad_norm": 1.640625,
+      "learning_rate": 1.921126209436271e-05,
+      "loss": 1.0274,
+      "step": 2314
+    },
+    {
+      "epoch": 0.39693936601153096,
+      "grad_norm": 1.640625,
+      "learning_rate": 1.9210558961230922e-05,
+      "loss": 0.9467,
+      "step": 2315
+    },
+    {
+      "epoch": 0.39711083010052084,
+      "grad_norm": 1.65625,
+      "learning_rate": 1.9209855527708127e-05,
+      "loss": 1.0797,
+      "step": 2316
+    },
+    {
+      "epoch": 0.39728229418951067,
+      "grad_norm": 1.625,
+      "learning_rate": 1.9209151793817257e-05,
+      "loss": 1.0166,
+      "step": 2317
+    },
+    {
+      "epoch": 0.39745375827850055,
+      "grad_norm": 1.6796875,
+      "learning_rate": 1.920844775958127e-05,
+      "loss": 1.0113,
+      "step": 2318
+    },
+    {
+      "epoch": 0.3976252223674904,
+      "grad_norm": 1.609375,
+      "learning_rate": 1.9207743425023125e-05,
+      "loss": 1.0353,
+      "step": 2319
+    },
+    {
+      "epoch": 0.39779668645648025,
+      "grad_norm": 1.6015625,
+      "learning_rate": 1.9207038790165796e-05,
+      "loss": 1.0908,
+      "step": 2320
+    },
+    {
+      "epoch": 0.39796815054547013,
+      "grad_norm": 1.625,
+      "learning_rate": 1.9206333855032262e-05,
+      "loss": 0.9886,
+      "step": 2321
+    },
+    {
+      "epoch": 0.39813961463446,
+      "grad_norm": 1.6015625,
+      "learning_rate": 1.9205628619645514e-05,
+      "loss": 1.0451,
+      "step": 2322
+    },
+    {
+      "epoch": 0.39831107872344984,
+      "grad_norm": 1.7421875,
+      "learning_rate": 1.9204923084028548e-05,
+      "loss": 1.0668,
+      "step": 2323
+    },
+    {
+      "epoch": 0.3984825428124397,
+      "grad_norm": 1.7578125,
+      "learning_rate": 1.920421724820438e-05,
+      "loss": 1.1112,
+      "step": 2324
+    },
+    {
+      "epoch": 0.3986540069014296,
+      "grad_norm": 1.59375,
+      "learning_rate": 1.9203511112196026e-05,
+      "loss": 1.0489,
+      "step": 2325
+    },
+    {
+      "epoch": 0.39882547099041943,
+      "grad_norm": 1.5390625,
+      "learning_rate": 1.920280467602652e-05,
+      "loss": 0.9801,
+      "step": 2326
+    },
+    {
+      "epoch": 0.3989969350794093,
+      "grad_norm": 1.6875,
+      "learning_rate": 1.9202097939718896e-05,
+      "loss": 1.1454,
+      "step": 2327
+    },
+    {
+      "epoch": 0.3991683991683992,
+      "grad_norm": 1.59375,
+      "learning_rate": 1.9201390903296208e-05,
+      "loss": 0.9652,
+      "step": 2328
+    },
+    {
+      "epoch": 0.399339863257389,
+      "grad_norm": 1.53125,
+      "learning_rate": 1.9200683566781513e-05,
+      "loss": 0.9748,
+      "step": 2329
+    },
+    {
+      "epoch": 0.3995113273463789,
+      "grad_norm": 1.578125,
+      "learning_rate": 1.919997593019788e-05,
+      "loss": 1.0532,
+      "step": 2330
+    },
+    {
+      "epoch": 0.3996827914353688,
+      "grad_norm": 1.5859375,
+      "learning_rate": 1.9199267993568387e-05,
+      "loss": 0.9792,
+      "step": 2331
+    },
+    {
+      "epoch": 0.3998542555243586,
+      "grad_norm": 1.6328125,
+      "learning_rate": 1.9198559756916123e-05,
+      "loss": 1.0428,
+      "step": 2332
+    },
+    {
+      "epoch": 0.4000257196133485,
+      "grad_norm": 1.703125,
+      "learning_rate": 1.919785122026419e-05,
+      "loss": 1.038,
+      "step": 2333
+    },
+    {
+      "epoch": 0.40019718370233837,
+      "grad_norm": 1.625,
+      "learning_rate": 1.919714238363569e-05,
+      "loss": 1.0486,
+      "step": 2334
+    },
+    {
+      "epoch": 0.4003686477913282,
+      "grad_norm": 1.7578125,
+      "learning_rate": 1.9196433247053746e-05,
+      "loss": 1.0839,
+      "step": 2335
+    },
+    {
+      "epoch": 0.4005401118803181,
+      "grad_norm": 1.75,
+      "learning_rate": 1.919572381054148e-05,
+      "loss": 1.1026,
+      "step": 2336
+    },
+    {
+      "epoch": 0.40071157596930795,
+      "grad_norm": 1.546875,
+      "learning_rate": 1.9195014074122037e-05,
+      "loss": 1.0809,
+      "step": 2337
+    },
+    {
+      "epoch": 0.4008830400582978,
+      "grad_norm": 1.703125,
+      "learning_rate": 1.9194304037818555e-05,
+      "loss": 1.1521,
+      "step": 2338
+    },
+    {
+      "epoch": 0.40105450414728766,
+      "grad_norm": 1.5859375,
+      "learning_rate": 1.9193593701654202e-05,
+      "loss": 1.0516,
+      "step": 2339
+    },
+    {
+      "epoch": 0.40122596823627754,
+      "grad_norm": 1.5546875,
+      "learning_rate": 1.9192883065652132e-05,
+      "loss": 1.058,
+      "step": 2340
+    },
+    {
+      "epoch": 0.40139743232526737,
+      "grad_norm": 1.578125,
+      "learning_rate": 1.9192172129835533e-05,
+      "loss": 1.053,
+      "step": 2341
+    },
+    {
+      "epoch": 0.40156889641425725,
+      "grad_norm": 1.59375,
+      "learning_rate": 1.919146089422758e-05,
+      "loss": 1.0059,
+      "step": 2342
+    },
+    {
+      "epoch": 0.4017403605032471,
+      "grad_norm": 1.546875,
+      "learning_rate": 1.9190749358851485e-05,
+      "loss": 1.0745,
+      "step": 2343
+    },
+    {
+      "epoch": 0.40191182459223695,
+      "grad_norm": 1.59375,
+      "learning_rate": 1.9190037523730435e-05,
+      "loss": 0.991,
+      "step": 2344
+    },
+    {
+      "epoch": 0.40208328868122684,
+      "grad_norm": 1.6875,
+      "learning_rate": 1.918932538888766e-05,
+      "loss": 0.9776,
+      "step": 2345
+    },
+    {
+      "epoch": 0.40225475277021666,
+      "grad_norm": 1.6796875,
+      "learning_rate": 1.918861295434638e-05,
+      "loss": 1.0356,
+      "step": 2346
+    },
+    {
+      "epoch": 0.40242621685920654,
+      "grad_norm": 1.6328125,
+      "learning_rate": 1.918790022012983e-05,
+      "loss": 0.9826,
+      "step": 2347
+    },
+    {
+      "epoch": 0.4025976809481964,
+      "grad_norm": 1.59375,
+      "learning_rate": 1.9187187186261254e-05,
+      "loss": 1.0368,
+      "step": 2348
+    },
+    {
+      "epoch": 0.40276914503718625,
+      "grad_norm": 1.625,
+      "learning_rate": 1.918647385276391e-05,
+      "loss": 1.0812,
+      "step": 2349
+    },
+    {
+      "epoch": 0.40294060912617613,
+      "grad_norm": 1.6875,
+      "learning_rate": 1.9185760219661057e-05,
+      "loss": 1.0314,
+      "step": 2350
+    },
+    {
+      "epoch": 0.403112073215166,
+      "grad_norm": 1.5390625,
+      "learning_rate": 1.9185046286975978e-05,
+      "loss": 1.0582,
+      "step": 2351
+    },
+    {
+      "epoch": 0.40328353730415584,
+      "grad_norm": 1.703125,
+      "learning_rate": 1.9184332054731948e-05,
+      "loss": 1.0616,
+      "step": 2352
+    },
+    {
+      "epoch": 0.4034550013931457,
+      "grad_norm": 1.546875,
+      "learning_rate": 1.9183617522952267e-05,
+      "loss": 1.0125,
+      "step": 2353
+    },
+    {
+      "epoch": 0.4036264654821356,
+      "grad_norm": 1.7578125,
+      "learning_rate": 1.9182902691660235e-05,
+      "loss": 1.0078,
+      "step": 2354
+    },
+    {
+      "epoch": 0.4037979295711254,
+      "grad_norm": 1.671875,
+      "learning_rate": 1.9182187560879165e-05,
+      "loss": 1.0332,
+      "step": 2355
+    },
+    {
+      "epoch": 0.4039693936601153,
+      "grad_norm": 1.7109375,
+      "learning_rate": 1.9181472130632385e-05,
+      "loss": 1.0976,
+      "step": 2356
+    },
+    {
+      "epoch": 0.4041408577491052,
+      "grad_norm": 1.6796875,
+      "learning_rate": 1.9180756400943222e-05,
+      "loss": 1.0831,
+      "step": 2357
+    },
+    {
+      "epoch": 0.404312321838095,
+      "grad_norm": 1.5390625,
+      "learning_rate": 1.9180040371835025e-05,
+      "loss": 1.0052,
+      "step": 2358
+    },
+    {
+      "epoch": 0.4044837859270849,
+      "grad_norm": 1.6640625,
+      "learning_rate": 1.917932404333114e-05,
+      "loss": 1.1313,
+      "step": 2359
+    },
+    {
+      "epoch": 0.4046552500160748,
+      "grad_norm": 1.65625,
+      "learning_rate": 1.917860741545493e-05,
+      "loss": 1.087,
+      "step": 2360
+    },
+    {
+      "epoch": 0.4048267141050646,
+      "grad_norm": 1.6328125,
+      "learning_rate": 1.9177890488229775e-05,
+      "loss": 1.1294,
+      "step": 2361
+    },
+    {
+      "epoch": 0.4049981781940545,
+      "grad_norm": 1.5,
+      "learning_rate": 1.9177173261679045e-05,
+      "loss": 0.8989,
+      "step": 2362
+    },
+    {
+      "epoch": 0.40516964228304436,
+      "grad_norm": 1.5859375,
+      "learning_rate": 1.9176455735826136e-05,
+      "loss": 1.0715,
+      "step": 2363
+    },
+    {
+      "epoch": 0.4053411063720342,
+      "grad_norm": 1.7109375,
+      "learning_rate": 1.917573791069445e-05,
+      "loss": 1.042,
+      "step": 2364
+    },
+    {
+      "epoch": 0.40551257046102407,
+      "grad_norm": 1.640625,
+      "learning_rate": 1.9175019786307404e-05,
+      "loss": 1.0684,
+      "step": 2365
+    },
+    {
+      "epoch": 0.40568403455001395,
+      "grad_norm": 1.6875,
+      "learning_rate": 1.9174301362688408e-05,
+      "loss": 1.0898,
+      "step": 2366
+    },
+    {
+      "epoch": 0.4058554986390038,
+      "grad_norm": 1.6171875,
+      "learning_rate": 1.9173582639860895e-05,
+      "loss": 1.0967,
+      "step": 2367
+    },
+    {
+      "epoch": 0.40602696272799366,
+      "grad_norm": 1.7109375,
+      "learning_rate": 1.9172863617848315e-05,
+      "loss": 1.0429,
+      "step": 2368
+    },
+    {
+      "epoch": 0.40619842681698354,
+      "grad_norm": 1.6796875,
+      "learning_rate": 1.9172144296674106e-05,
+      "loss": 1.0442,
+      "step": 2369
+    },
+    {
+      "epoch": 0.40636989090597336,
+      "grad_norm": 1.625,
+      "learning_rate": 1.917142467636173e-05,
+      "loss": 1.1039,
+      "step": 2370
+    },
+    {
+      "epoch": 0.40654135499496324,
+      "grad_norm": 1.625,
+      "learning_rate": 1.9170704756934663e-05,
+      "loss": 0.9351,
+      "step": 2371
+    },
+    {
+      "epoch": 0.4067128190839531,
+      "grad_norm": 1.6640625,
+      "learning_rate": 1.9169984538416382e-05,
+      "loss": 1.0478,
+      "step": 2372
+    },
+    {
+      "epoch": 0.40688428317294295,
+      "grad_norm": 1.53125,
+      "learning_rate": 1.916926402083037e-05,
+      "loss": 1.0912,
+      "step": 2373
+    },
+    {
+      "epoch": 0.40705574726193283,
+      "grad_norm": 1.71875,
+      "learning_rate": 1.916854320420013e-05,
+      "loss": 1.0144,
+      "step": 2374
+    },
+    {
+      "epoch": 0.4072272113509227,
+      "grad_norm": 1.640625,
+      "learning_rate": 1.9167822088549177e-05,
+      "loss": 1.1344,
+      "step": 2375
+    },
+    {
+      "epoch": 0.40739867543991254,
+      "grad_norm": 1.546875,
+      "learning_rate": 1.9167100673901014e-05,
+      "loss": 0.9543,
+      "step": 2376
+    },
+    {
+      "epoch": 0.4075701395289024,
+      "grad_norm": 1.6484375,
+      "learning_rate": 1.9166378960279185e-05,
+      "loss": 1.0708,
+      "step": 2377
+    },
+    {
+      "epoch": 0.4077416036178923,
+      "grad_norm": 1.609375,
+      "learning_rate": 1.9165656947707216e-05,
+      "loss": 1.021,
+      "step": 2378
+    },
+    {
+      "epoch": 0.4079130677068821,
+      "grad_norm": 1.640625,
+      "learning_rate": 1.916493463620866e-05,
+      "loss": 0.9727,
+      "step": 2379
+    },
+    {
+      "epoch": 0.408084531795872,
+      "grad_norm": 1.6796875,
+      "learning_rate": 1.9164212025807073e-05,
+      "loss": 1.0842,
+      "step": 2380
+    },
+    {
+      "epoch": 0.4082559958848619,
+      "grad_norm": 1.703125,
+      "learning_rate": 1.9163489116526025e-05,
+      "loss": 1.0958,
+      "step": 2381
+    },
+    {
+      "epoch": 0.4084274599738517,
+      "grad_norm": 1.5390625,
+      "learning_rate": 1.916276590838909e-05,
+      "loss": 1.0029,
+      "step": 2382
+    },
+    {
+      "epoch": 0.4085989240628416,
+      "grad_norm": 1.8515625,
+      "learning_rate": 1.9162042401419853e-05,
+      "loss": 1.0555,
+      "step": 2383
+    },
+    {
+      "epoch": 0.4087703881518315,
+      "grad_norm": 1.5703125,
+      "learning_rate": 1.9161318595641915e-05,
+      "loss": 1.0398,
+      "step": 2384
+    },
+    {
+      "epoch": 0.4089418522408213,
+      "grad_norm": 1.765625,
+      "learning_rate": 1.9160594491078875e-05,
+      "loss": 1.0446,
+      "step": 2385
+    },
+    {
+      "epoch": 0.4091133163298112,
+      "grad_norm": 1.625,
+      "learning_rate": 1.915987008775436e-05,
+      "loss": 1.0675,
+      "step": 2386
+    },
+    {
+      "epoch": 0.40928478041880106,
+      "grad_norm": 1.625,
+      "learning_rate": 1.915914538569198e-05,
+      "loss": 1.1024,
+      "step": 2387
+    },
+    {
+      "epoch": 0.4094562445077909,
+      "grad_norm": 1.625,
+      "learning_rate": 1.9158420384915387e-05,
+      "loss": 1.106,
+      "step": 2388
+    },
+    {
+      "epoch": 0.40962770859678077,
+      "grad_norm": 1.6015625,
+      "learning_rate": 1.915769508544821e-05,
+      "loss": 1.0341,
+      "step": 2389
+    },
+    {
+      "epoch": 0.4097991726857706,
+      "grad_norm": 1.6875,
+      "learning_rate": 1.915696948731412e-05,
+      "loss": 1.062,
+      "step": 2390
+    },
+    {
+      "epoch": 0.4099706367747605,
+      "grad_norm": 1.6015625,
+      "learning_rate": 1.9156243590536764e-05,
+      "loss": 1.1496,
+      "step": 2391
+    },
+    {
+      "epoch": 0.41014210086375036,
+      "grad_norm": 1.640625,
+      "learning_rate": 1.9155517395139833e-05,
+      "loss": 1.0598,
+      "step": 2392
+    },
+    {
+      "epoch": 0.4103135649527402,
+      "grad_norm": 1.65625,
+      "learning_rate": 1.9154790901147e-05,
+      "loss": 0.9867,
+      "step": 2393
+    },
+    {
+      "epoch": 0.41048502904173007,
+      "grad_norm": 1.6015625,
+      "learning_rate": 1.915406410858196e-05,
+      "loss": 1.0782,
+      "step": 2394
+    },
+    {
+      "epoch": 0.41065649313071995,
+      "grad_norm": 1.65625,
+      "learning_rate": 1.9153337017468424e-05,
+      "loss": 1.0514,
+      "step": 2395
+    },
+    {
+      "epoch": 0.41082795721970977,
+      "grad_norm": 1.6484375,
+      "learning_rate": 1.9152609627830095e-05,
+      "loss": 1.0171,
+      "step": 2396
+    },
+    {
+      "epoch": 0.41099942130869965,
+      "grad_norm": 1.6484375,
+      "learning_rate": 1.91518819396907e-05,
+      "loss": 1.0764,
+      "step": 2397
+    },
+    {
+      "epoch": 0.41117088539768953,
+      "grad_norm": 1.8984375,
+      "learning_rate": 1.9151153953073976e-05,
+      "loss": 1.0801,
+      "step": 2398
+    },
+    {
+      "epoch": 0.41134234948667936,
+      "grad_norm": 1.5625,
+      "learning_rate": 1.9150425668003657e-05,
+      "loss": 1.0627,
+      "step": 2399
+    },
+    {
+      "epoch": 0.41151381357566924,
+      "grad_norm": 1.59375,
+      "learning_rate": 1.9149697084503502e-05,
+      "loss": 1.0157,
+      "step": 2400
+    },
+    {
+      "epoch": 0.4116852776646591,
+      "grad_norm": 1.53125,
+      "learning_rate": 1.9148968202597272e-05,
+      "loss": 0.9235,
+      "step": 2401
+    },
+    {
+      "epoch": 0.41185674175364895,
+      "grad_norm": 1.6171875,
+      "learning_rate": 1.9148239022308734e-05,
+      "loss": 1.0204,
+      "step": 2402
+    },
+    {
+      "epoch": 0.41202820584263883,
+      "grad_norm": 1.671875,
+      "learning_rate": 1.9147509543661677e-05,
+      "loss": 1.0096,
+      "step": 2403
+    },
+    {
+      "epoch": 0.4121996699316287,
+      "grad_norm": 1.703125,
+      "learning_rate": 1.9146779766679885e-05,
+      "loss": 1.0784,
+      "step": 2404
+    },
+    {
+      "epoch": 0.41237113402061853,
+      "grad_norm": 1.6953125,
+      "learning_rate": 1.9146049691387157e-05,
+      "loss": 0.9982,
+      "step": 2405
+    },
+    {
+      "epoch": 0.4125425981096084,
+      "grad_norm": 1.609375,
+      "learning_rate": 1.9145319317807314e-05,
+      "loss": 1.0323,
+      "step": 2406
+    },
+    {
+      "epoch": 0.4127140621985983,
+      "grad_norm": 1.6875,
+      "learning_rate": 1.9144588645964165e-05,
+      "loss": 1.0299,
+      "step": 2407
+    },
+    {
+      "epoch": 0.4128855262875881,
+      "grad_norm": 1.6484375,
+      "learning_rate": 1.914385767588155e-05,
+      "loss": 1.052,
+      "step": 2408
+    },
+    {
+      "epoch": 0.413056990376578,
+      "grad_norm": 1.7109375,
+      "learning_rate": 1.91431264075833e-05,
+      "loss": 1.0281,
+      "step": 2409
+    },
+    {
+      "epoch": 0.4132284544655679,
+      "grad_norm": 1.7109375,
+      "learning_rate": 1.914239484109327e-05,
+      "loss": 1.0676,
+      "step": 2410
+    },
+    {
+      "epoch": 0.4133999185545577,
+      "grad_norm": 1.640625,
+      "learning_rate": 1.9141662976435315e-05,
+      "loss": 1.1174,
+      "step": 2411
+    },
+    {
+      "epoch": 0.4135713826435476,
+      "grad_norm": 1.6640625,
+      "learning_rate": 1.9140930813633307e-05,
+      "loss": 1.1627,
+      "step": 2412
+    },
+    {
+      "epoch": 0.4137428467325375,
+      "grad_norm": 1.8046875,
+      "learning_rate": 1.9140198352711124e-05,
+      "loss": 1.0565,
+      "step": 2413
+    },
+    {
+      "epoch": 0.4139143108215273,
+      "grad_norm": 1.6484375,
+      "learning_rate": 1.9139465593692653e-05,
+      "loss": 1.0109,
+      "step": 2414
+    },
+    {
+      "epoch": 0.4140857749105172,
+      "grad_norm": 1.625,
+      "learning_rate": 1.9138732536601794e-05,
+      "loss": 1.0394,
+      "step": 2415
+    },
+    {
+      "epoch": 0.41425723899950706,
+      "grad_norm": 1.625,
+      "learning_rate": 1.9137999181462453e-05,
+      "loss": 1.023,
+      "step": 2416
+    },
+    {
+      "epoch": 0.4144287030884969,
+      "grad_norm": 1.8203125,
+      "learning_rate": 1.913726552829855e-05,
+      "loss": 1.0992,
+      "step": 2417
+    },
+    {
+      "epoch": 0.41460016717748677,
+      "grad_norm": 1.6640625,
+      "learning_rate": 1.9136531577134007e-05,
+      "loss": 1.1668,
+      "step": 2418
+    },
+    {
+      "epoch": 0.41477163126647665,
+      "grad_norm": 1.6171875,
+      "learning_rate": 1.9135797327992766e-05,
+      "loss": 0.9833,
+      "step": 2419
+    },
+    {
+      "epoch": 0.4149430953554665,
+      "grad_norm": 1.6640625,
+      "learning_rate": 1.9135062780898775e-05,
+      "loss": 1.0893,
+      "step": 2420
+    },
+    {
+      "epoch": 0.41511455944445635,
+      "grad_norm": 1.640625,
+      "learning_rate": 1.913432793587598e-05,
+      "loss": 1.0264,
+      "step": 2421
+    },
+    {
+      "epoch": 0.41528602353344624,
+      "grad_norm": 1.640625,
+      "learning_rate": 1.913359279294836e-05,
+      "loss": 0.9673,
+      "step": 2422
+    },
+    {
+      "epoch": 0.41545748762243606,
+      "grad_norm": 1.6171875,
+      "learning_rate": 1.9132857352139884e-05,
+      "loss": 1.0679,
+      "step": 2423
+    },
+    {
+      "epoch": 0.41562895171142594,
+      "grad_norm": 1.546875,
+      "learning_rate": 1.913212161347454e-05,
+      "loss": 0.9749,
+      "step": 2424
+    },
+    {
+      "epoch": 0.4158004158004158,
+      "grad_norm": 1.6640625,
+      "learning_rate": 1.913138557697632e-05,
+      "loss": 1.1008,
+      "step": 2425
+    },
+    {
+      "epoch": 0.41597187988940565,
+      "grad_norm": 1.5703125,
+      "learning_rate": 1.913064924266923e-05,
+      "loss": 1.0338,
+      "step": 2426
+    },
+    {
+      "epoch": 0.41614334397839553,
+      "grad_norm": 1.7109375,
+      "learning_rate": 1.9129912610577287e-05,
+      "loss": 1.0881,
+      "step": 2427
+    },
+    {
+      "epoch": 0.4163148080673854,
+      "grad_norm": 1.5625,
+      "learning_rate": 1.912917568072451e-05,
+      "loss": 1.1241,
+      "step": 2428
+    },
+    {
+      "epoch": 0.41648627215637524,
+      "grad_norm": 1.640625,
+      "learning_rate": 1.912843845313494e-05,
+      "loss": 1.1109,
+      "step": 2429
+    },
+    {
+      "epoch": 0.4166577362453651,
+      "grad_norm": 1.59375,
+      "learning_rate": 1.9127700927832616e-05,
+      "loss": 0.9705,
+      "step": 2430
+    },
+    {
+      "epoch": 0.416829200334355,
+      "grad_norm": 1.59375,
+      "learning_rate": 1.9126963104841595e-05,
+      "loss": 1.0619,
+      "step": 2431
+    },
+    {
+      "epoch": 0.4170006644233448,
+      "grad_norm": 1.6015625,
+      "learning_rate": 1.912622498418594e-05,
+      "loss": 1.0899,
+      "step": 2432
+    },
+    {
+      "epoch": 0.4171721285123347,
+      "grad_norm": 1.6484375,
+      "learning_rate": 1.912548656588972e-05,
+      "loss": 1.0203,
+      "step": 2433
+    },
+    {
+      "epoch": 0.4173435926013246,
+      "grad_norm": 1.703125,
+      "learning_rate": 1.9124747849977016e-05,
+      "loss": 1.0402,
+      "step": 2434
+    },
+    {
+      "epoch": 0.4175150566903144,
+      "grad_norm": 1.6328125,
+      "learning_rate": 1.9124008836471927e-05,
+      "loss": 1.0895,
+      "step": 2435
+    },
+    {
+      "epoch": 0.4176865207793043,
+      "grad_norm": 1.625,
+      "learning_rate": 1.9123269525398554e-05,
+      "loss": 1.0637,
+      "step": 2436
+    },
+    {
+      "epoch": 0.4178579848682941,
+      "grad_norm": 1.6953125,
+      "learning_rate": 1.9122529916781002e-05,
+      "loss": 0.9888,
+      "step": 2437
+    },
+    {
+      "epoch": 0.418029448957284,
+      "grad_norm": 1.671875,
+      "learning_rate": 1.9121790010643402e-05,
+      "loss": 1.0156,
+      "step": 2438
+    },
+    {
+      "epoch": 0.4182009130462739,
+      "grad_norm": 1.6328125,
+      "learning_rate": 1.912104980700988e-05,
+      "loss": 1.1407,
+      "step": 2439
+    },
+    {
+      "epoch": 0.4183723771352637,
+      "grad_norm": 1.5,
+      "learning_rate": 1.9120309305904575e-05,
+      "loss": 1.0139,
+      "step": 2440
+    },
+    {
+      "epoch": 0.4185438412242536,
+      "grad_norm": 1.640625,
+      "learning_rate": 1.911956850735164e-05,
+      "loss": 1.015,
+      "step": 2441
+    },
+    {
+      "epoch": 0.41871530531324347,
+      "grad_norm": 1.65625,
+      "learning_rate": 1.9118827411375233e-05,
+      "loss": 1.0958,
+      "step": 2442
+    },
+    {
+      "epoch": 0.4188867694022333,
+      "grad_norm": 1.546875,
+      "learning_rate": 1.911808601799953e-05,
+      "loss": 0.9161,
+      "step": 2443
+    },
+    {
+      "epoch": 0.4190582334912232,
+      "grad_norm": 1.6640625,
+      "learning_rate": 1.9117344327248704e-05,
+      "loss": 1.0281,
+      "step": 2444
+    },
+    {
+      "epoch": 0.41922969758021306,
+      "grad_norm": 1.5625,
+      "learning_rate": 1.9116602339146945e-05,
+      "loss": 1.0395,
+      "step": 2445
+    },
+    {
+      "epoch": 0.4194011616692029,
+      "grad_norm": 1.5625,
+      "learning_rate": 1.9115860053718455e-05,
+      "loss": 1.0219,
+      "step": 2446
+    },
+    {
+      "epoch": 0.41957262575819276,
+      "grad_norm": 1.609375,
+      "learning_rate": 1.911511747098744e-05,
+      "loss": 1.0217,
+      "step": 2447
+    },
+    {
+      "epoch": 0.41974408984718264,
+      "grad_norm": 1.6171875,
+      "learning_rate": 1.9114374590978123e-05,
+      "loss": 0.9609,
+      "step": 2448
+    },
+    {
+      "epoch": 0.41991555393617247,
+      "grad_norm": 2.296875,
+      "learning_rate": 1.911363141371473e-05,
+      "loss": 1.0245,
+      "step": 2449
+    },
+    {
+      "epoch": 0.42008701802516235,
+      "grad_norm": 1.6015625,
+      "learning_rate": 1.9112887939221495e-05,
+      "loss": 1.0485,
+      "step": 2450
+    },
+    {
+      "epoch": 0.42025848211415223,
+      "grad_norm": 1.703125,
+      "learning_rate": 1.911214416752267e-05,
+      "loss": 1.1228,
+      "step": 2451
+    },
+    {
+      "epoch": 0.42042994620314206,
+      "grad_norm": 1.953125,
+      "learning_rate": 1.911140009864251e-05,
+      "loss": 1.0937,
+      "step": 2452
+    },
+    {
+      "epoch": 0.42060141029213194,
+      "grad_norm": 1.6015625,
+      "learning_rate": 1.911065573260528e-05,
+      "loss": 1.0646,
+      "step": 2453
+    },
+    {
+      "epoch": 0.4207728743811218,
+      "grad_norm": 1.609375,
+      "learning_rate": 1.9109911069435263e-05,
+      "loss": 1.0335,
+      "step": 2454
+    },
+    {
+      "epoch": 0.42094433847011165,
+      "grad_norm": 1.640625,
+      "learning_rate": 1.910916610915674e-05,
+      "loss": 1.0465,
+      "step": 2455
+    },
+    {
+      "epoch": 0.4211158025591015,
+      "grad_norm": 1.5546875,
+      "learning_rate": 1.9108420851794007e-05,
+      "loss": 0.9507,
+      "step": 2456
+    },
+    {
+      "epoch": 0.4212872666480914,
+      "grad_norm": 1.6328125,
+      "learning_rate": 1.9107675297371372e-05,
+      "loss": 0.9774,
+      "step": 2457
+    },
+    {
+      "epoch": 0.42145873073708123,
+      "grad_norm": 1.6328125,
+      "learning_rate": 1.9106929445913147e-05,
+      "loss": 1.1207,
+      "step": 2458
+    },
+    {
+      "epoch": 0.4216301948260711,
+      "grad_norm": 1.5078125,
+      "learning_rate": 1.910618329744366e-05,
+      "loss": 1.0304,
+      "step": 2459
+    },
+    {
+      "epoch": 0.421801658915061,
+      "grad_norm": 1.578125,
+      "learning_rate": 1.9105436851987247e-05,
+      "loss": 1.0384,
+      "step": 2460
+    },
+    {
+      "epoch": 0.4219731230040508,
+      "grad_norm": 1.6484375,
+      "learning_rate": 1.9104690109568248e-05,
+      "loss": 1.0574,
+      "step": 2461
+    },
+    {
+      "epoch": 0.4221445870930407,
+      "grad_norm": 1.609375,
+      "learning_rate": 1.910394307021102e-05,
+      "loss": 0.9608,
+      "step": 2462
+    },
+    {
+      "epoch": 0.4223160511820306,
+      "grad_norm": 1.5703125,
+      "learning_rate": 1.910319573393993e-05,
+      "loss": 1.0314,
+      "step": 2463
+    },
+    {
+      "epoch": 0.4224875152710204,
+      "grad_norm": 1.6328125,
+      "learning_rate": 1.9102448100779343e-05,
+      "loss": 1.0206,
+      "step": 2464
+    },
+    {
+      "epoch": 0.4226589793600103,
+      "grad_norm": 1.6015625,
+      "learning_rate": 1.910170017075365e-05,
+      "loss": 1.0368,
+      "step": 2465
+    },
+    {
+      "epoch": 0.42283044344900017,
+      "grad_norm": 1.6796875,
+      "learning_rate": 1.910095194388724e-05,
+      "loss": 1.0978,
+      "step": 2466
+    },
+    {
+      "epoch": 0.42300190753799,
+      "grad_norm": 1.640625,
+      "learning_rate": 1.9100203420204516e-05,
+      "loss": 1.1047,
+      "step": 2467
+    },
+    {
+      "epoch": 0.4231733716269799,
+      "grad_norm": 1.5859375,
+      "learning_rate": 1.9099454599729887e-05,
+      "loss": 1.028,
+      "step": 2468
+    },
+    {
+      "epoch": 0.42334483571596976,
+      "grad_norm": 1.6328125,
+      "learning_rate": 1.909870548248778e-05,
+      "loss": 1.0247,
+      "step": 2469
+    },
+    {
+      "epoch": 0.4235162998049596,
+      "grad_norm": 1.5703125,
+      "learning_rate": 1.9097956068502626e-05,
+      "loss": 1.0003,
+      "step": 2470
+    },
+    {
+      "epoch": 0.42368776389394947,
+      "grad_norm": 1.546875,
+      "learning_rate": 1.9097206357798864e-05,
+      "loss": 1.086,
+      "step": 2471
+    },
+    {
+      "epoch": 0.42385922798293935,
+      "grad_norm": 1.65625,
+      "learning_rate": 1.9096456350400948e-05,
+      "loss": 1.0451,
+      "step": 2472
+    },
+    {
+      "epoch": 0.42403069207192917,
+      "grad_norm": 1.6953125,
+      "learning_rate": 1.9095706046333333e-05,
+      "loss": 1.1459,
+      "step": 2473
+    },
+    {
+      "epoch": 0.42420215616091905,
+      "grad_norm": 1.6171875,
+      "learning_rate": 1.9094955445620492e-05,
+      "loss": 1.0954,
+      "step": 2474
+    },
+    {
+      "epoch": 0.42437362024990893,
+      "grad_norm": 1.6171875,
+      "learning_rate": 1.909420454828691e-05,
+      "loss": 0.9514,
+      "step": 2475
+    },
+    {
+      "epoch": 0.42454508433889876,
+      "grad_norm": 1.5703125,
+      "learning_rate": 1.9093453354357064e-05,
+      "loss": 1.0758,
+      "step": 2476
+    },
+    {
+      "epoch": 0.42471654842788864,
+      "grad_norm": 1.765625,
+      "learning_rate": 1.909270186385547e-05,
+      "loss": 1.0863,
+      "step": 2477
+    },
+    {
+      "epoch": 0.4248880125168785,
+      "grad_norm": 1.609375,
+      "learning_rate": 1.9091950076806626e-05,
+      "loss": 1.0663,
+      "step": 2478
+    },
+    {
+      "epoch": 0.42505947660586835,
+      "grad_norm": 1.5390625,
+      "learning_rate": 1.909119799323505e-05,
+      "loss": 0.9911,
+      "step": 2479
+    },
+    {
+      "epoch": 0.42523094069485823,
+      "grad_norm": 1.6328125,
+      "learning_rate": 1.9090445613165273e-05,
+      "loss": 1.0238,
+      "step": 2480
+    },
+    {
+      "epoch": 0.4254024047838481,
+      "grad_norm": 1.734375,
+      "learning_rate": 1.9089692936621836e-05,
+      "loss": 1.1141,
+      "step": 2481
+    },
+    {
+      "epoch": 0.42557386887283793,
+      "grad_norm": 1.4609375,
+      "learning_rate": 1.908893996362928e-05,
+      "loss": 0.9739,
+      "step": 2482
+    },
+    {
+      "epoch": 0.4257453329618278,
+      "grad_norm": 1.625,
+      "learning_rate": 1.9088186694212174e-05,
+      "loss": 1.0654,
+      "step": 2483
+    },
+    {
+      "epoch": 0.4259167970508177,
+      "grad_norm": 1.6484375,
+      "learning_rate": 1.9087433128395073e-05,
+      "loss": 1.0327,
+      "step": 2484
+    },
+    {
+      "epoch": 0.4260882611398075,
+      "grad_norm": 1.671875,
+      "learning_rate": 1.9086679266202554e-05,
+      "loss": 1.0202,
+      "step": 2485
+    },
+    {
+      "epoch": 0.4262597252287974,
+      "grad_norm": 1.5546875,
+      "learning_rate": 1.908592510765921e-05,
+      "loss": 1.1057,
+      "step": 2486
+    },
+    {
+      "epoch": 0.42643118931778723,
+      "grad_norm": 1.640625,
+      "learning_rate": 1.908517065278963e-05,
+      "loss": 1.0686,
+      "step": 2487
+    },
+    {
+      "epoch": 0.4266026534067771,
+      "grad_norm": 1.5390625,
+      "learning_rate": 1.9084415901618428e-05,
+      "loss": 0.9874,
+      "step": 2488
+    },
+    {
+      "epoch": 0.426774117495767,
+      "grad_norm": 1.7265625,
+      "learning_rate": 1.9083660854170212e-05,
+      "loss": 1.0316,
+      "step": 2489
+    },
+    {
+      "epoch": 0.4269455815847568,
+      "grad_norm": 1.625,
+      "learning_rate": 1.908290551046961e-05,
+      "loss": 1.0549,
+      "step": 2490
+    },
+    {
+      "epoch": 0.4271170456737467,
+      "grad_norm": 1.6484375,
+      "learning_rate": 1.9082149870541258e-05,
+      "loss": 1.039,
+      "step": 2491
+    },
+    {
+      "epoch": 0.4272885097627366,
+      "grad_norm": 1.59375,
+      "learning_rate": 1.9081393934409797e-05,
+      "loss": 1.0129,
+      "step": 2492
+    },
+    {
+      "epoch": 0.4274599738517264,
+      "grad_norm": 1.734375,
+      "learning_rate": 1.9080637702099883e-05,
+      "loss": 1.1138,
+      "step": 2493
+    },
+    {
+      "epoch": 0.4276314379407163,
+      "grad_norm": 1.59375,
+      "learning_rate": 1.9079881173636182e-05,
+      "loss": 1.013,
+      "step": 2494
+    },
+    {
+      "epoch": 0.42780290202970617,
+      "grad_norm": 1.640625,
+      "learning_rate": 1.907912434904336e-05,
+      "loss": 1.0412,
+      "step": 2495
+    },
+    {
+      "epoch": 0.427974366118696,
+      "grad_norm": 1.6328125,
+      "learning_rate": 1.9078367228346106e-05,
+      "loss": 1.0565,
+      "step": 2496
+    },
+    {
+      "epoch": 0.4281458302076859,
+      "grad_norm": 1.6171875,
+      "learning_rate": 1.9077609811569107e-05,
+      "loss": 1.0311,
+      "step": 2497
+    },
+    {
+      "epoch": 0.42831729429667575,
+      "grad_norm": 1.671875,
+      "learning_rate": 1.9076852098737074e-05,
+      "loss": 1.1392,
+      "step": 2498
+    },
+    {
+      "epoch": 0.4284887583856656,
+      "grad_norm": 1.640625,
+      "learning_rate": 1.907609408987471e-05,
+      "loss": 1.0764,
+      "step": 2499
+    },
+    {
+      "epoch": 0.42866022247465546,
+      "grad_norm": 1.703125,
+      "learning_rate": 1.9075335785006743e-05,
+      "loss": 1.063,
+      "step": 2500
+    },
+    {
+      "epoch": 0.42883168656364534,
+      "grad_norm": 1.6015625,
+      "learning_rate": 1.90745771841579e-05,
+      "loss": 1.1018,
+      "step": 2501
+    },
+    {
+      "epoch": 0.42900315065263517,
+      "grad_norm": 1.609375,
+      "learning_rate": 1.9073818287352925e-05,
+      "loss": 1.0077,
+      "step": 2502
+    },
+    {
+      "epoch": 0.42917461474162505,
+      "grad_norm": 1.6015625,
+      "learning_rate": 1.9073059094616565e-05,
+      "loss": 0.9708,
+      "step": 2503
+    },
+    {
+      "epoch": 0.42934607883061493,
+      "grad_norm": 1.625,
+      "learning_rate": 1.9072299605973582e-05,
+      "loss": 1.068,
+      "step": 2504
+    },
+    {
+      "epoch": 0.42951754291960476,
+      "grad_norm": 1.5625,
+      "learning_rate": 1.907153982144875e-05,
+      "loss": 1.0559,
+      "step": 2505
+    },
+    {
+      "epoch": 0.42968900700859464,
+      "grad_norm": 1.578125,
+      "learning_rate": 1.907077974106684e-05,
+      "loss": 1.1482,
+      "step": 2506
+    },
+    {
+      "epoch": 0.4298604710975845,
+      "grad_norm": 1.546875,
+      "learning_rate": 1.9070019364852646e-05,
+      "loss": 1.071,
+      "step": 2507
+    },
+    {
+      "epoch": 0.43003193518657434,
+      "grad_norm": 1.6484375,
+      "learning_rate": 1.9069258692830964e-05,
+      "loss": 1.0705,
+      "step": 2508
+    },
+    {
+      "epoch": 0.4302033992755642,
+      "grad_norm": 1.6171875,
+      "learning_rate": 1.9068497725026607e-05,
+      "loss": 1.0141,
+      "step": 2509
+    },
+    {
+      "epoch": 0.4303748633645541,
+      "grad_norm": 1.6484375,
+      "learning_rate": 1.906773646146439e-05,
+      "loss": 1.1033,
+      "step": 2510
+    },
+    {
+      "epoch": 0.43054632745354393,
+      "grad_norm": 1.71875,
+      "learning_rate": 1.9066974902169142e-05,
+      "loss": 0.9858,
+      "step": 2511
+    },
+    {
+      "epoch": 0.4307177915425338,
+      "grad_norm": 1.7109375,
+      "learning_rate": 1.9066213047165698e-05,
+      "loss": 1.0306,
+      "step": 2512
+    },
+    {
+      "epoch": 0.4308892556315237,
+      "grad_norm": 1.65625,
+      "learning_rate": 1.9065450896478904e-05,
+      "loss": 1.0003,
+      "step": 2513
+    },
+    {
+      "epoch": 0.4310607197205135,
+      "grad_norm": 1.5078125,
+      "learning_rate": 1.906468845013362e-05,
+      "loss": 0.9826,
+      "step": 2514
+    },
+    {
+      "epoch": 0.4312321838095034,
+      "grad_norm": 1.671875,
+      "learning_rate": 1.9063925708154713e-05,
+      "loss": 1.0373,
+      "step": 2515
+    },
+    {
+      "epoch": 0.4314036478984933,
+      "grad_norm": 1.5546875,
+      "learning_rate": 1.9063162670567057e-05,
+      "loss": 1.072,
+      "step": 2516
+    },
+    {
+      "epoch": 0.4315751119874831,
+      "grad_norm": 1.671875,
+      "learning_rate": 1.906239933739554e-05,
+      "loss": 1.0728,
+      "step": 2517
+    },
+    {
+      "epoch": 0.431746576076473,
+      "grad_norm": 1.5859375,
+      "learning_rate": 1.9061635708665053e-05,
+      "loss": 1.0666,
+      "step": 2518
+    },
+    {
+      "epoch": 0.43191804016546287,
+      "grad_norm": 1.5078125,
+      "learning_rate": 1.90608717844005e-05,
+      "loss": 1.0418,
+      "step": 2519
+    },
+    {
+      "epoch": 0.4320895042544527,
+      "grad_norm": 1.6171875,
+      "learning_rate": 1.90601075646268e-05,
+      "loss": 0.9639,
+      "step": 2520
+    },
+    {
+      "epoch": 0.4322609683434426,
+      "grad_norm": 1.5390625,
+      "learning_rate": 1.905934304936887e-05,
+      "loss": 0.9816,
+      "step": 2521
+    },
+    {
+      "epoch": 0.43243243243243246,
+      "grad_norm": 1.6875,
+      "learning_rate": 1.9058578238651655e-05,
+      "loss": 1.0473,
+      "step": 2522
+    },
+    {
+      "epoch": 0.4326038965214223,
+      "grad_norm": 1.609375,
+      "learning_rate": 1.9057813132500095e-05,
+      "loss": 1.105,
+      "step": 2523
+    },
+    {
+      "epoch": 0.43277536061041216,
+      "grad_norm": 1.6171875,
+      "learning_rate": 1.9057047730939132e-05,
+      "loss": 1.0386,
+      "step": 2524
+    },
+    {
+      "epoch": 0.43294682469940204,
+      "grad_norm": 1.6953125,
+      "learning_rate": 1.905628203399374e-05,
+      "loss": 1.0777,
+      "step": 2525
+    },
+    {
+      "epoch": 0.43311828878839187,
+      "grad_norm": 1.6328125,
+      "learning_rate": 1.905551604168889e-05,
+      "loss": 1.0285,
+      "step": 2526
+    },
+    {
+      "epoch": 0.43328975287738175,
+      "grad_norm": 1.7421875,
+      "learning_rate": 1.905474975404956e-05,
+      "loss": 1.0409,
+      "step": 2527
+    },
+    {
+      "epoch": 0.43346121696637163,
+      "grad_norm": 1.5859375,
+      "learning_rate": 1.9053983171100744e-05,
+      "loss": 1.0576,
+      "step": 2528
+    },
+    {
+      "epoch": 0.43363268105536146,
+      "grad_norm": 1.75,
+      "learning_rate": 1.9053216292867445e-05,
+      "loss": 1.1138,
+      "step": 2529
+    },
+    {
+      "epoch": 0.43380414514435134,
+      "grad_norm": 1.6171875,
+      "learning_rate": 1.905244911937467e-05,
+      "loss": 1.0694,
+      "step": 2530
+    },
+    {
+      "epoch": 0.4339756092333412,
+      "grad_norm": 1.640625,
+      "learning_rate": 1.9051681650647437e-05,
+      "loss": 0.955,
+      "step": 2531
+    },
+    {
+      "epoch": 0.43414707332233105,
+      "grad_norm": 1.546875,
+      "learning_rate": 1.9050913886710786e-05,
+      "loss": 1.021,
+      "step": 2532
+    },
+    {
+      "epoch": 0.4343185374113209,
+      "grad_norm": 1.703125,
+      "learning_rate": 1.9050145827589745e-05,
+      "loss": 1.098,
+      "step": 2533
+    },
+    {
+      "epoch": 0.43449000150031075,
+      "grad_norm": 1.609375,
+      "learning_rate": 1.904937747330937e-05,
+      "loss": 1.0156,
+      "step": 2534
+    },
+    {
+      "epoch": 0.43466146558930063,
+      "grad_norm": 1.6640625,
+      "learning_rate": 1.9048608823894722e-05,
+      "loss": 1.0958,
+      "step": 2535
+    },
+    {
+      "epoch": 0.4348329296782905,
+      "grad_norm": 1.71875,
+      "learning_rate": 1.9047839879370867e-05,
+      "loss": 1.0483,
+      "step": 2536
+    },
+    {
+      "epoch": 0.43500439376728034,
+      "grad_norm": 1.59375,
+      "learning_rate": 1.9047070639762878e-05,
+      "loss": 0.9646,
+      "step": 2537
+    },
+    {
+      "epoch": 0.4351758578562702,
+      "grad_norm": 1.5859375,
+      "learning_rate": 1.9046301105095847e-05,
+      "loss": 0.9332,
+      "step": 2538
+    },
+    {
+      "epoch": 0.4353473219452601,
+      "grad_norm": 1.671875,
+      "learning_rate": 1.9045531275394878e-05,
+      "loss": 1.0767,
+      "step": 2539
+    },
+    {
+      "epoch": 0.4355187860342499,
+      "grad_norm": 1.6171875,
+      "learning_rate": 1.904476115068507e-05,
+      "loss": 1.0839,
+      "step": 2540
+    },
+    {
+      "epoch": 0.4356902501232398,
+      "grad_norm": 1.5703125,
+      "learning_rate": 1.9043990730991536e-05,
+      "loss": 1.0746,
+      "step": 2541
+    },
+    {
+      "epoch": 0.4358617142122297,
+      "grad_norm": 1.5859375,
+      "learning_rate": 1.9043220016339414e-05,
+      "loss": 0.9487,
+      "step": 2542
+    },
+    {
+      "epoch": 0.4360331783012195,
+      "grad_norm": 1.6875,
+      "learning_rate": 1.9042449006753827e-05,
+      "loss": 1.0746,
+      "step": 2543
+    },
+    {
+      "epoch": 0.4362046423902094,
+      "grad_norm": 1.5546875,
+      "learning_rate": 1.904167770225993e-05,
+      "loss": 1.0214,
+      "step": 2544
+    },
+    {
+      "epoch": 0.4363761064791993,
+      "grad_norm": 1.5703125,
+      "learning_rate": 1.904090610288288e-05,
+      "loss": 1.0514,
+      "step": 2545
+    },
+    {
+      "epoch": 0.4365475705681891,
+      "grad_norm": 1.53125,
+      "learning_rate": 1.904013420864783e-05,
+      "loss": 1.0005,
+      "step": 2546
+    },
+    {
+      "epoch": 0.436719034657179,
+      "grad_norm": 1.5625,
+      "learning_rate": 1.9039362019579965e-05,
+      "loss": 0.9825,
+      "step": 2547
+    },
+    {
+      "epoch": 0.43689049874616886,
+      "grad_norm": 1.75,
+      "learning_rate": 1.9038589535704467e-05,
+      "loss": 1.0655,
+      "step": 2548
+    },
+    {
+      "epoch": 0.4370619628351587,
+      "grad_norm": 1.6015625,
+      "learning_rate": 1.9037816757046528e-05,
+      "loss": 0.9854,
+      "step": 2549
+    },
+    {
+      "epoch": 0.43723342692414857,
+      "grad_norm": 1.5546875,
+      "learning_rate": 1.903704368363135e-05,
+      "loss": 1.0154,
+      "step": 2550
+    },
+    {
+      "epoch": 0.43740489101313845,
+      "grad_norm": 1.578125,
+      "learning_rate": 1.9036270315484144e-05,
+      "loss": 1.0617,
+      "step": 2551
+    },
+    {
+      "epoch": 0.4375763551021283,
+      "grad_norm": 1.6640625,
+      "learning_rate": 1.9035496652630138e-05,
+      "loss": 1.0488,
+      "step": 2552
+    },
+    {
+      "epoch": 0.43774781919111816,
+      "grad_norm": 1.640625,
+      "learning_rate": 1.9034722695094562e-05,
+      "loss": 1.0497,
+      "step": 2553
+    },
+    {
+      "epoch": 0.43791928328010804,
+      "grad_norm": 1.671875,
+      "learning_rate": 1.9033948442902658e-05,
+      "loss": 1.0427,
+      "step": 2554
+    },
+    {
+      "epoch": 0.43809074736909787,
+      "grad_norm": 1.625,
+      "learning_rate": 1.9033173896079677e-05,
+      "loss": 1.0454,
+      "step": 2555
+    },
+    {
+      "epoch": 0.43826221145808775,
+      "grad_norm": 1.53125,
+      "learning_rate": 1.9032399054650876e-05,
+      "loss": 0.962,
+      "step": 2556
+    },
+    {
+      "epoch": 0.43843367554707763,
+      "grad_norm": 1.5625,
+      "learning_rate": 1.903162391864153e-05,
+      "loss": 1.0095,
+      "step": 2557
+    },
+    {
+      "epoch": 0.43860513963606745,
+      "grad_norm": 1.75,
+      "learning_rate": 1.9030848488076924e-05,
+      "loss": 1.1543,
+      "step": 2558
+    },
+    {
+      "epoch": 0.43877660372505733,
+      "grad_norm": 1.671875,
+      "learning_rate": 1.9030072762982335e-05,
+      "loss": 1.0546,
+      "step": 2559
+    },
+    {
+      "epoch": 0.4389480678140472,
+      "grad_norm": 1.671875,
+      "learning_rate": 1.9029296743383074e-05,
+      "loss": 1.1123,
+      "step": 2560
+    },
+    {
+      "epoch": 0.43911953190303704,
+      "grad_norm": 1.71875,
+      "learning_rate": 1.902852042930444e-05,
+      "loss": 1.0054,
+      "step": 2561
+    },
+    {
+      "epoch": 0.4392909959920269,
+      "grad_norm": 1.6328125,
+      "learning_rate": 1.9027743820771756e-05,
+      "loss": 1.0301,
+      "step": 2562
+    },
+    {
+      "epoch": 0.4394624600810168,
+      "grad_norm": 1.671875,
+      "learning_rate": 1.9026966917810356e-05,
+      "loss": 1.0997,
+      "step": 2563
+    },
+    {
+      "epoch": 0.43963392417000663,
+      "grad_norm": 1.5390625,
+      "learning_rate": 1.9026189720445568e-05,
+      "loss": 1.0412,
+      "step": 2564
+    },
+    {
+      "epoch": 0.4398053882589965,
+      "grad_norm": 1.5078125,
+      "learning_rate": 1.9025412228702747e-05,
+      "loss": 1.0091,
+      "step": 2565
+    },
+    {
+      "epoch": 0.4399768523479864,
+      "grad_norm": 1.640625,
+      "learning_rate": 1.9024634442607245e-05,
+      "loss": 1.0701,
+      "step": 2566
+    },
+    {
+      "epoch": 0.4401483164369762,
+      "grad_norm": 1.6640625,
+      "learning_rate": 1.9023856362184433e-05,
+      "loss": 0.974,
+      "step": 2567
+    },
+    {
+      "epoch": 0.4403197805259661,
+      "grad_norm": 1.7109375,
+      "learning_rate": 1.902307798745968e-05,
+      "loss": 1.1008,
+      "step": 2568
+    },
+    {
+      "epoch": 0.440491244614956,
+      "grad_norm": 1.6015625,
+      "learning_rate": 1.9022299318458377e-05,
+      "loss": 1.0568,
+      "step": 2569
+    },
+    {
+      "epoch": 0.4406627087039458,
+      "grad_norm": 1.7734375,
+      "learning_rate": 1.902152035520592e-05,
+      "loss": 1.1136,
+      "step": 2570
+    },
+    {
+      "epoch": 0.4408341727929357,
+      "grad_norm": 1.6328125,
+      "learning_rate": 1.902074109772771e-05,
+      "loss": 0.989,
+      "step": 2571
+    },
+    {
+      "epoch": 0.44100563688192557,
+      "grad_norm": 1.6328125,
+      "learning_rate": 1.9019961546049165e-05,
+      "loss": 0.9976,
+      "step": 2572
+    },
+    {
+      "epoch": 0.4411771009709154,
+      "grad_norm": 1.6484375,
+      "learning_rate": 1.9019181700195714e-05,
+      "loss": 1.0227,
+      "step": 2573
+    },
+    {
+      "epoch": 0.4413485650599053,
+      "grad_norm": 1.6484375,
+      "learning_rate": 1.9018401560192775e-05,
+      "loss": 1.0334,
+      "step": 2574
+    },
+    {
+      "epoch": 0.44152002914889515,
+      "grad_norm": 1.6640625,
+      "learning_rate": 1.9017621126065804e-05,
+      "loss": 1.0257,
+      "step": 2575
+    },
+    {
+      "epoch": 0.441691493237885,
+      "grad_norm": 1.546875,
+      "learning_rate": 1.9016840397840256e-05,
+      "loss": 1.0305,
+      "step": 2576
+    },
+    {
+      "epoch": 0.44186295732687486,
+      "grad_norm": 1.5859375,
+      "learning_rate": 1.901605937554158e-05,
+      "loss": 1.0111,
+      "step": 2577
+    },
+    {
+      "epoch": 0.44203442141586474,
+      "grad_norm": 1.671875,
+      "learning_rate": 1.901527805919526e-05,
+      "loss": 1.0746,
+      "step": 2578
+    },
+    {
+      "epoch": 0.44220588550485457,
+      "grad_norm": 1.6484375,
+      "learning_rate": 1.9014496448826775e-05,
+      "loss": 1.08,
+      "step": 2579
+    },
+    {
+      "epoch": 0.44237734959384445,
+      "grad_norm": 1.546875,
+      "learning_rate": 1.9013714544461617e-05,
+      "loss": 1.0462,
+      "step": 2580
+    },
+    {
+      "epoch": 0.4425488136828343,
+      "grad_norm": 1.625,
+      "learning_rate": 1.9012932346125282e-05,
+      "loss": 1.0524,
+      "step": 2581
+    },
+    {
+      "epoch": 0.44272027777182416,
+      "grad_norm": 1.578125,
+      "learning_rate": 1.9012149853843283e-05,
+      "loss": 0.9439,
+      "step": 2582
+    },
+    {
+      "epoch": 0.44289174186081404,
+      "grad_norm": 1.7578125,
+      "learning_rate": 1.9011367067641146e-05,
+      "loss": 1.0259,
+      "step": 2583
+    },
+    {
+      "epoch": 0.44306320594980386,
+      "grad_norm": 1.5625,
+      "learning_rate": 1.901058398754439e-05,
+      "loss": 1.0868,
+      "step": 2584
+    },
+    {
+      "epoch": 0.44323467003879374,
+      "grad_norm": 1.6875,
+      "learning_rate": 1.9009800613578563e-05,
+      "loss": 1.1054,
+      "step": 2585
+    },
+    {
+      "epoch": 0.4434061341277836,
+      "grad_norm": 1.578125,
+      "learning_rate": 1.9009016945769207e-05,
+      "loss": 0.9721,
+      "step": 2586
+    },
+    {
+      "epoch": 0.44357759821677345,
+      "grad_norm": 1.6875,
+      "learning_rate": 1.9008232984141885e-05,
+      "loss": 1.0474,
+      "step": 2587
+    },
+    {
+      "epoch": 0.44374906230576333,
+      "grad_norm": 1.640625,
+      "learning_rate": 1.9007448728722165e-05,
+      "loss": 1.0311,
+      "step": 2588
+    },
+    {
+      "epoch": 0.4439205263947532,
+      "grad_norm": 1.71875,
+      "learning_rate": 1.900666417953562e-05,
+      "loss": 1.021,
+      "step": 2589
+    },
+    {
+      "epoch": 0.44409199048374304,
+      "grad_norm": 1.65625,
+      "learning_rate": 1.9005879336607844e-05,
+      "loss": 1.0546,
+      "step": 2590
+    },
+    {
+      "epoch": 0.4442634545727329,
+      "grad_norm": 1.5390625,
+      "learning_rate": 1.9005094199964427e-05,
+      "loss": 1.0064,
+      "step": 2591
+    },
+    {
+      "epoch": 0.4444349186617228,
+      "grad_norm": 1.640625,
+      "learning_rate": 1.900430876963098e-05,
+      "loss": 0.9986,
+      "step": 2592
+    },
+    {
+      "epoch": 0.4446063827507126,
+      "grad_norm": 1.5703125,
+      "learning_rate": 1.9003523045633116e-05,
+      "loss": 1.0127,
+      "step": 2593
+    },
+    {
+      "epoch": 0.4447778468397025,
+      "grad_norm": 1.6875,
+      "learning_rate": 1.900273702799646e-05,
+      "loss": 1.1392,
+      "step": 2594
+    },
+    {
+      "epoch": 0.4449493109286924,
+      "grad_norm": 1.515625,
+      "learning_rate": 1.9001950716746648e-05,
+      "loss": 0.9631,
+      "step": 2595
+    },
+    {
+      "epoch": 0.4451207750176822,
+      "grad_norm": 1.609375,
+      "learning_rate": 1.9001164111909327e-05,
+      "loss": 0.9925,
+      "step": 2596
+    },
+    {
+      "epoch": 0.4452922391066721,
+      "grad_norm": 1.578125,
+      "learning_rate": 1.9000377213510147e-05,
+      "loss": 1.0275,
+      "step": 2597
+    },
+    {
+      "epoch": 0.445463703195662,
+      "grad_norm": 1.7578125,
+      "learning_rate": 1.8999590021574776e-05,
+      "loss": 1.0874,
+      "step": 2598
+    },
+    {
+      "epoch": 0.4456351672846518,
+      "grad_norm": 1.546875,
+      "learning_rate": 1.8998802536128885e-05,
+      "loss": 0.979,
+      "step": 2599
+    },
+    {
+      "epoch": 0.4458066313736417,
+      "grad_norm": 1.6015625,
+      "learning_rate": 1.8998014757198152e-05,
+      "loss": 1.0238,
+      "step": 2600
+    },
+    {
+      "epoch": 0.44597809546263156,
+      "grad_norm": 1.75,
+      "learning_rate": 1.8997226684808282e-05,
+      "loss": 1.1384,
+      "step": 2601
+    },
+    {
+      "epoch": 0.4461495595516214,
+      "grad_norm": 1.515625,
+      "learning_rate": 1.8996438318984968e-05,
+      "loss": 1.0545,
+      "step": 2602
+    },
+    {
+      "epoch": 0.44632102364061127,
+      "grad_norm": 1.625,
+      "learning_rate": 1.8995649659753917e-05,
+      "loss": 1.0358,
+      "step": 2603
+    },
+    {
+      "epoch": 0.44649248772960115,
+      "grad_norm": 1.6015625,
+      "learning_rate": 1.8994860707140862e-05,
+      "loss": 1.0149,
+      "step": 2604
+    },
+    {
+      "epoch": 0.446663951818591,
+      "grad_norm": 1.5703125,
+      "learning_rate": 1.8994071461171522e-05,
+      "loss": 0.9884,
+      "step": 2605
+    },
+    {
+      "epoch": 0.44683541590758086,
+      "grad_norm": 1.5390625,
+      "learning_rate": 1.899328192187165e-05,
+      "loss": 1.0315,
+      "step": 2606
+    },
+    {
+      "epoch": 0.44700687999657074,
+      "grad_norm": 1.5703125,
+      "learning_rate": 1.8992492089266986e-05,
+      "loss": 0.9365,
+      "step": 2607
+    },
+    {
+      "epoch": 0.44717834408556056,
+      "grad_norm": 1.5546875,
+      "learning_rate": 1.899170196338329e-05,
+      "loss": 1.0134,
+      "step": 2608
+    },
+    {
+      "epoch": 0.44734980817455045,
+      "grad_norm": 1.65625,
+      "learning_rate": 1.8990911544246338e-05,
+      "loss": 1.0205,
+      "step": 2609
+    },
+    {
+      "epoch": 0.4475212722635403,
+      "grad_norm": 1.6015625,
+      "learning_rate": 1.89901208318819e-05,
+      "loss": 1.0267,
+      "step": 2610
+    },
+    {
+      "epoch": 0.44769273635253015,
+      "grad_norm": 1.5859375,
+      "learning_rate": 1.898932982631577e-05,
+      "loss": 1.1054,
+      "step": 2611
+    },
+    {
+      "epoch": 0.44786420044152003,
+      "grad_norm": 1.8203125,
+      "learning_rate": 1.8988538527573743e-05,
+      "loss": 0.9982,
+      "step": 2612
+    },
+    {
+      "epoch": 0.4480356645305099,
+      "grad_norm": 1.6015625,
+      "learning_rate": 1.8987746935681627e-05,
+      "loss": 1.0921,
+      "step": 2613
+    },
+    {
+      "epoch": 0.44820712861949974,
+      "grad_norm": 1.6484375,
+      "learning_rate": 1.898695505066524e-05,
+      "loss": 1.105,
+      "step": 2614
+    },
+    {
+      "epoch": 0.4483785927084896,
+      "grad_norm": 1.84375,
+      "learning_rate": 1.8986162872550405e-05,
+      "loss": 1.1337,
+      "step": 2615
+    },
+    {
+      "epoch": 0.4485500567974795,
+      "grad_norm": 1.6484375,
+      "learning_rate": 1.898537040136296e-05,
+      "loss": 1.0623,
+      "step": 2616
+    },
+    {
+      "epoch": 0.4487215208864693,
+      "grad_norm": 1.6953125,
+      "learning_rate": 1.8984577637128755e-05,
+      "loss": 1.1215,
+      "step": 2617
+    },
+    {
+      "epoch": 0.4488929849754592,
+      "grad_norm": 1.5,
+      "learning_rate": 1.8983784579873635e-05,
+      "loss": 0.9338,
+      "step": 2618
+    },
+    {
+      "epoch": 0.4490644490644491,
+      "grad_norm": 1.6484375,
+      "learning_rate": 1.898299122962347e-05,
+      "loss": 1.0079,
+      "step": 2619
+    },
+    {
+      "epoch": 0.4492359131534389,
+      "grad_norm": 1.65625,
+      "learning_rate": 1.8982197586404136e-05,
+      "loss": 1.0642,
+      "step": 2620
+    },
+    {
+      "epoch": 0.4494073772424288,
+      "grad_norm": 1.625,
+      "learning_rate": 1.8981403650241517e-05,
+      "loss": 1.0029,
+      "step": 2621
+    },
+    {
+      "epoch": 0.4495788413314187,
+      "grad_norm": 1.8125,
+      "learning_rate": 1.89806094211615e-05,
+      "loss": 1.0326,
+      "step": 2622
+    },
+    {
+      "epoch": 0.4497503054204085,
+      "grad_norm": 1.78125,
+      "learning_rate": 1.8979814899189993e-05,
+      "loss": 1.0617,
+      "step": 2623
+    },
+    {
+      "epoch": 0.4499217695093984,
+      "grad_norm": 1.6640625,
+      "learning_rate": 1.897902008435291e-05,
+      "loss": 1.0087,
+      "step": 2624
+    },
+    {
+      "epoch": 0.45009323359838826,
+      "grad_norm": 1.609375,
+      "learning_rate": 1.8978224976676168e-05,
+      "loss": 1.0239,
+      "step": 2625
+    },
+    {
+      "epoch": 0.4502646976873781,
+      "grad_norm": 1.6484375,
+      "learning_rate": 1.89774295761857e-05,
+      "loss": 1.0068,
+      "step": 2626
+    },
+    {
+      "epoch": 0.45043616177636797,
+      "grad_norm": 1.59375,
+      "learning_rate": 1.897663388290745e-05,
+      "loss": 0.9916,
+      "step": 2627
+    },
+    {
+      "epoch": 0.4506076258653578,
+      "grad_norm": 1.546875,
+      "learning_rate": 1.8975837896867365e-05,
+      "loss": 1.0105,
+      "step": 2628
+    },
+    {
+      "epoch": 0.4507790899543477,
+      "grad_norm": 1.5,
+      "learning_rate": 1.8975041618091406e-05,
+      "loss": 0.964,
+      "step": 2629
+    },
+    {
+      "epoch": 0.45095055404333756,
+      "grad_norm": 1.5859375,
+      "learning_rate": 1.8974245046605544e-05,
+      "loss": 1.0457,
+      "step": 2630
+    },
+    {
+      "epoch": 0.4511220181323274,
+      "grad_norm": 1.59375,
+      "learning_rate": 1.8973448182435757e-05,
+      "loss": 1.0377,
+      "step": 2631
+    },
+    {
+      "epoch": 0.45129348222131727,
+      "grad_norm": 1.640625,
+      "learning_rate": 1.897265102560803e-05,
+      "loss": 1.0343,
+      "step": 2632
+    },
+    {
+      "epoch": 0.45146494631030715,
+      "grad_norm": 1.59375,
+      "learning_rate": 1.897185357614837e-05,
+      "loss": 0.9959,
+      "step": 2633
+    },
+    {
+      "epoch": 0.451636410399297,
+      "grad_norm": 1.6953125,
+      "learning_rate": 1.8971055834082778e-05,
+      "loss": 1.0973,
+      "step": 2634
+    },
+    {
+      "epoch": 0.45180787448828685,
+      "grad_norm": 1.515625,
+      "learning_rate": 1.8970257799437274e-05,
+      "loss": 0.951,
+      "step": 2635
+    },
+    {
+      "epoch": 0.45197933857727673,
+      "grad_norm": 1.578125,
+      "learning_rate": 1.8969459472237886e-05,
+      "loss": 1.0957,
+      "step": 2636
+    },
+    {
+      "epoch": 0.45215080266626656,
+      "grad_norm": 1.671875,
+      "learning_rate": 1.8968660852510646e-05,
+      "loss": 1.0603,
+      "step": 2637
+    },
+    {
+      "epoch": 0.45232226675525644,
+      "grad_norm": 1.5625,
+      "learning_rate": 1.8967861940281603e-05,
+      "loss": 0.9595,
+      "step": 2638
+    },
+    {
+      "epoch": 0.4524937308442463,
+      "grad_norm": 1.6640625,
+      "learning_rate": 1.896706273557681e-05,
+      "loss": 1.0628,
+      "step": 2639
+    },
+    {
+      "epoch": 0.45266519493323615,
+      "grad_norm": 1.65625,
+      "learning_rate": 1.896626323842234e-05,
+      "loss": 1.0335,
+      "step": 2640
+    },
+    {
+      "epoch": 0.45283665902222603,
+      "grad_norm": 1.546875,
+      "learning_rate": 1.896546344884426e-05,
+      "loss": 0.9677,
+      "step": 2641
+    },
+    {
+      "epoch": 0.4530081231112159,
+      "grad_norm": 1.5625,
+      "learning_rate": 1.8964663366868655e-05,
+      "loss": 0.9559,
+      "step": 2642
+    },
+    {
+      "epoch": 0.45317958720020574,
+      "grad_norm": 1.5703125,
+      "learning_rate": 1.896386299252162e-05,
+      "loss": 1.0433,
+      "step": 2643
+    },
+    {
+      "epoch": 0.4533510512891956,
+      "grad_norm": 1.609375,
+      "learning_rate": 1.8963062325829256e-05,
+      "loss": 1.0023,
+      "step": 2644
+    },
+    {
+      "epoch": 0.4535225153781855,
+      "grad_norm": 1.6484375,
+      "learning_rate": 1.8962261366817684e-05,
+      "loss": 1.1453,
+      "step": 2645
+    },
+    {
+      "epoch": 0.4536939794671753,
+      "grad_norm": 1.59375,
+      "learning_rate": 1.8961460115513012e-05,
+      "loss": 0.9839,
+      "step": 2646
+    },
+    {
+      "epoch": 0.4538654435561652,
+      "grad_norm": 1.6015625,
+      "learning_rate": 1.8960658571941388e-05,
+      "loss": 1.1139,
+      "step": 2647
+    },
+    {
+      "epoch": 0.4540369076451551,
+      "grad_norm": 1.6484375,
+      "learning_rate": 1.895985673612894e-05,
+      "loss": 1.073,
+      "step": 2648
+    },
+    {
+      "epoch": 0.4542083717341449,
+      "grad_norm": 1.5390625,
+      "learning_rate": 1.8959054608101823e-05,
+      "loss": 1.0102,
+      "step": 2649
+    },
+    {
+      "epoch": 0.4543798358231348,
+      "grad_norm": 1.5859375,
+      "learning_rate": 1.8958252187886202e-05,
+      "loss": 1.0139,
+      "step": 2650
+    },
+    {
+      "epoch": 0.4545512999121247,
+      "grad_norm": 1.5859375,
+      "learning_rate": 1.8957449475508243e-05,
+      "loss": 0.9918,
+      "step": 2651
+    },
+    {
+      "epoch": 0.4547227640011145,
+      "grad_norm": 1.6171875,
+      "learning_rate": 1.8956646470994124e-05,
+      "loss": 1.0911,
+      "step": 2652
+    },
+    {
+      "epoch": 0.4548942280901044,
+      "grad_norm": 1.6328125,
+      "learning_rate": 1.895584317437004e-05,
+      "loss": 1.1194,
+      "step": 2653
+    },
+    {
+      "epoch": 0.45506569217909426,
+      "grad_norm": 1.609375,
+      "learning_rate": 1.8955039585662182e-05,
+      "loss": 1.0762,
+      "step": 2654
+    },
+    {
+      "epoch": 0.4552371562680841,
+      "grad_norm": 1.5703125,
+      "learning_rate": 1.895423570489676e-05,
+      "loss": 1.0032,
+      "step": 2655
+    },
+    {
+      "epoch": 0.45540862035707397,
+      "grad_norm": 1.6171875,
+      "learning_rate": 1.8953431532099994e-05,
+      "loss": 0.9551,
+      "step": 2656
+    },
+    {
+      "epoch": 0.45558008444606385,
+      "grad_norm": 1.59375,
+      "learning_rate": 1.8952627067298115e-05,
+      "loss": 1.0617,
+      "step": 2657
+    },
+    {
+      "epoch": 0.4557515485350537,
+      "grad_norm": 1.6953125,
+      "learning_rate": 1.895182231051735e-05,
+      "loss": 1.0103,
+      "step": 2658
+    },
+    {
+      "epoch": 0.45592301262404356,
+      "grad_norm": 1.6328125,
+      "learning_rate": 1.8951017261783954e-05,
+      "loss": 1.1338,
+      "step": 2659
+    },
+    {
+      "epoch": 0.45609447671303344,
+      "grad_norm": 1.5625,
+      "learning_rate": 1.8950211921124177e-05,
+      "loss": 1.0476,
+      "step": 2660
+    },
+    {
+      "epoch": 0.45626594080202326,
+      "grad_norm": 1.5390625,
+      "learning_rate": 1.8949406288564282e-05,
+      "loss": 1.0679,
+      "step": 2661
+    },
+    {
+      "epoch": 0.45643740489101314,
+      "grad_norm": 1.609375,
+      "learning_rate": 1.894860036413055e-05,
+      "loss": 1.0398,
+      "step": 2662
+    },
+    {
+      "epoch": 0.456608868980003,
+      "grad_norm": 1.6015625,
+      "learning_rate": 1.8947794147849264e-05,
+      "loss": 1.0968,
+      "step": 2663
+    },
+    {
+      "epoch": 0.45678033306899285,
+      "grad_norm": 1.609375,
+      "learning_rate": 1.8946987639746717e-05,
+      "loss": 1.0471,
+      "step": 2664
+    },
+    {
+      "epoch": 0.45695179715798273,
+      "grad_norm": 1.5703125,
+      "learning_rate": 1.894618083984921e-05,
+      "loss": 1.0101,
+      "step": 2665
+    },
+    {
+      "epoch": 0.4571232612469726,
+      "grad_norm": 1.6328125,
+      "learning_rate": 1.894537374818306e-05,
+      "loss": 1.0929,
+      "step": 2666
+    },
+    {
+      "epoch": 0.45729472533596244,
+      "grad_norm": 1.578125,
+      "learning_rate": 1.8944566364774585e-05,
+      "loss": 1.0572,
+      "step": 2667
+    },
+    {
+      "epoch": 0.4574661894249523,
+      "grad_norm": 1.625,
+      "learning_rate": 1.8943758689650117e-05,
+      "loss": 1.0125,
+      "step": 2668
+    },
+    {
+      "epoch": 0.4576376535139422,
+      "grad_norm": 1.5703125,
+      "learning_rate": 1.8942950722836e-05,
+      "loss": 1.0197,
+      "step": 2669
+    },
+    {
+      "epoch": 0.457809117602932,
+      "grad_norm": 1.5859375,
+      "learning_rate": 1.8942142464358587e-05,
+      "loss": 1.0973,
+      "step": 2670
+    },
+    {
+      "epoch": 0.4579805816919219,
+      "grad_norm": 1.546875,
+      "learning_rate": 1.8941333914244234e-05,
+      "loss": 1.0297,
+      "step": 2671
+    },
+    {
+      "epoch": 0.4581520457809118,
+      "grad_norm": 1.6796875,
+      "learning_rate": 1.894052507251931e-05,
+      "loss": 1.0836,
+      "step": 2672
+    },
+    {
+      "epoch": 0.4583235098699016,
+      "grad_norm": 1.640625,
+      "learning_rate": 1.8939715939210202e-05,
+      "loss": 1.1439,
+      "step": 2673
+    },
+    {
+      "epoch": 0.4584949739588915,
+      "grad_norm": 1.6328125,
+      "learning_rate": 1.8938906514343287e-05,
+      "loss": 1.0319,
+      "step": 2674
+    },
+    {
+      "epoch": 0.4586664380478813,
+      "grad_norm": 1.6875,
+      "learning_rate": 1.8938096797944972e-05,
+      "loss": 1.0257,
+      "step": 2675
+    },
+    {
+      "epoch": 0.4588379021368712,
+      "grad_norm": 1.5546875,
+      "learning_rate": 1.893728679004166e-05,
+      "loss": 1.0049,
+      "step": 2676
+    },
+    {
+      "epoch": 0.4590093662258611,
+      "grad_norm": 1.4921875,
+      "learning_rate": 1.8936476490659778e-05,
+      "loss": 0.9021,
+      "step": 2677
+    },
+    {
+      "epoch": 0.4591808303148509,
+      "grad_norm": 1.6484375,
+      "learning_rate": 1.893566589982574e-05,
+      "loss": 0.9815,
+      "step": 2678
+    },
+    {
+      "epoch": 0.4593522944038408,
+      "grad_norm": 1.6015625,
+      "learning_rate": 1.893485501756599e-05,
+      "loss": 1.1038,
+      "step": 2679
+    },
+    {
+      "epoch": 0.45952375849283067,
+      "grad_norm": 1.5859375,
+      "learning_rate": 1.8934043843906975e-05,
+      "loss": 0.9767,
+      "step": 2680
+    },
+    {
+      "epoch": 0.4596952225818205,
+      "grad_norm": 1.625,
+      "learning_rate": 1.8933232378875145e-05,
+      "loss": 1.0821,
+      "step": 2681
+    },
+    {
+      "epoch": 0.4598666866708104,
+      "grad_norm": 1.53125,
+      "learning_rate": 1.8932420622496964e-05,
+      "loss": 0.9355,
+      "step": 2682
+    },
+    {
+      "epoch": 0.46003815075980026,
+      "grad_norm": 1.5,
+      "learning_rate": 1.8931608574798915e-05,
+      "loss": 0.9788,
+      "step": 2683
+    },
+    {
+      "epoch": 0.4602096148487901,
+      "grad_norm": 1.5625,
+      "learning_rate": 1.8930796235807478e-05,
+      "loss": 1.0064,
+      "step": 2684
+    },
+    {
+      "epoch": 0.46038107893777996,
+      "grad_norm": 1.578125,
+      "learning_rate": 1.8929983605549146e-05,
+      "loss": 1.1276,
+      "step": 2685
+    },
+    {
+      "epoch": 0.46055254302676985,
+      "grad_norm": 1.5625,
+      "learning_rate": 1.8929170684050414e-05,
+      "loss": 1.0831,
+      "step": 2686
+    },
+    {
+      "epoch": 0.46072400711575967,
+      "grad_norm": 1.625,
+      "learning_rate": 1.892835747133781e-05,
+      "loss": 0.9837,
+      "step": 2687
+    },
+    {
+      "epoch": 0.46089547120474955,
+      "grad_norm": 1.5,
+      "learning_rate": 1.8927543967437846e-05,
+      "loss": 0.9239,
+      "step": 2688
+    },
+    {
+      "epoch": 0.46106693529373943,
+      "grad_norm": 1.609375,
+      "learning_rate": 1.892673017237705e-05,
+      "loss": 1.0372,
+      "step": 2689
+    },
+    {
+      "epoch": 0.46123839938272926,
+      "grad_norm": 1.640625,
+      "learning_rate": 1.8925916086181975e-05,
+      "loss": 1.1304,
+      "step": 2690
+    },
+    {
+      "epoch": 0.46140986347171914,
+      "grad_norm": 1.6875,
+      "learning_rate": 1.8925101708879162e-05,
+      "loss": 1.0963,
+      "step": 2691
+    },
+    {
+      "epoch": 0.461581327560709,
+      "grad_norm": 1.5703125,
+      "learning_rate": 1.8924287040495174e-05,
+      "loss": 0.9772,
+      "step": 2692
+    },
+    {
+      "epoch": 0.46175279164969885,
+      "grad_norm": 1.5859375,
+      "learning_rate": 1.8923472081056577e-05,
+      "loss": 0.9871,
+      "step": 2693
+    },
+    {
+      "epoch": 0.4619242557386887,
+      "grad_norm": 1.734375,
+      "learning_rate": 1.8922656830589955e-05,
+      "loss": 1.1368,
+      "step": 2694
+    },
+    {
+      "epoch": 0.4620957198276786,
+      "grad_norm": 1.6484375,
+      "learning_rate": 1.8921841289121894e-05,
+      "loss": 1.1031,
+      "step": 2695
+    },
+    {
+      "epoch": 0.46226718391666843,
+      "grad_norm": 1.5859375,
+      "learning_rate": 1.8921025456678988e-05,
+      "loss": 0.9948,
+      "step": 2696
+    },
+    {
+      "epoch": 0.4624386480056583,
+      "grad_norm": 1.609375,
+      "learning_rate": 1.8920209333287854e-05,
+      "loss": 0.9736,
+      "step": 2697
+    },
+    {
+      "epoch": 0.4626101120946482,
+      "grad_norm": 1.609375,
+      "learning_rate": 1.89193929189751e-05,
+      "loss": 1.0552,
+      "step": 2698
+    },
+    {
+      "epoch": 0.462781576183638,
+      "grad_norm": 1.6328125,
+      "learning_rate": 1.8918576213767358e-05,
+      "loss": 1.095,
+      "step": 2699
+    },
+    {
+      "epoch": 0.4629530402726279,
+      "grad_norm": 1.6015625,
+      "learning_rate": 1.891775921769126e-05,
+      "loss": 0.9765,
+      "step": 2700
+    },
+    {
+      "epoch": 0.4631245043616178,
+      "grad_norm": 1.6484375,
+      "learning_rate": 1.8916941930773457e-05,
+      "loss": 1.0,
+      "step": 2701
+    },
+    {
+      "epoch": 0.4632959684506076,
+      "grad_norm": 1.6328125,
+      "learning_rate": 1.8916124353040594e-05,
+      "loss": 1.013,
+      "step": 2702
+    },
+    {
+      "epoch": 0.4634674325395975,
+      "grad_norm": 1.6328125,
+      "learning_rate": 1.8915306484519344e-05,
+      "loss": 1.1182,
+      "step": 2703
+    },
+    {
+      "epoch": 0.46363889662858737,
+      "grad_norm": 1.6328125,
+      "learning_rate": 1.8914488325236373e-05,
+      "loss": 1.0811,
+      "step": 2704
+    },
+    {
+      "epoch": 0.4638103607175772,
+      "grad_norm": 1.6640625,
+      "learning_rate": 1.8913669875218375e-05,
+      "loss": 1.0969,
+      "step": 2705
+    },
+    {
+      "epoch": 0.4639818248065671,
+      "grad_norm": 1.546875,
+      "learning_rate": 1.8912851134492033e-05,
+      "loss": 1.0535,
+      "step": 2706
+    },
+    {
+      "epoch": 0.46415328889555696,
+      "grad_norm": 1.5703125,
+      "learning_rate": 1.8912032103084054e-05,
+      "loss": 1.0302,
+      "step": 2707
+    },
+    {
+      "epoch": 0.4643247529845468,
+      "grad_norm": 1.609375,
+      "learning_rate": 1.8911212781021148e-05,
+      "loss": 1.0506,
+      "step": 2708
+    },
+    {
+      "epoch": 0.46449621707353667,
+      "grad_norm": 1.6015625,
+      "learning_rate": 1.8910393168330036e-05,
+      "loss": 1.0178,
+      "step": 2709
+    },
+    {
+      "epoch": 0.46466768116252655,
+      "grad_norm": 1.6015625,
+      "learning_rate": 1.8909573265037448e-05,
+      "loss": 1.0261,
+      "step": 2710
+    },
+    {
+      "epoch": 0.4648391452515164,
+      "grad_norm": 1.5625,
+      "learning_rate": 1.8908753071170126e-05,
+      "loss": 1.0288,
+      "step": 2711
+    },
+    {
+      "epoch": 0.46501060934050625,
+      "grad_norm": 1.7109375,
+      "learning_rate": 1.8907932586754823e-05,
+      "loss": 1.0356,
+      "step": 2712
+    },
+    {
+      "epoch": 0.46518207342949613,
+      "grad_norm": 1.671875,
+      "learning_rate": 1.8907111811818292e-05,
+      "loss": 1.0415,
+      "step": 2713
+    },
+    {
+      "epoch": 0.46535353751848596,
+      "grad_norm": 1.578125,
+      "learning_rate": 1.89062907463873e-05,
+      "loss": 0.9969,
+      "step": 2714
+    },
+    {
+      "epoch": 0.46552500160747584,
+      "grad_norm": 1.65625,
+      "learning_rate": 1.8905469390488635e-05,
+      "loss": 1.0581,
+      "step": 2715
+    },
+    {
+      "epoch": 0.4656964656964657,
+      "grad_norm": 1.609375,
+      "learning_rate": 1.8904647744149077e-05,
+      "loss": 0.9877,
+      "step": 2716
+    },
+    {
+      "epoch": 0.46586792978545555,
+      "grad_norm": 1.6875,
+      "learning_rate": 1.8903825807395422e-05,
+      "loss": 0.997,
+      "step": 2717
+    },
+    {
+      "epoch": 0.46603939387444543,
+      "grad_norm": 1.5390625,
+      "learning_rate": 1.890300358025448e-05,
+      "loss": 1.0222,
+      "step": 2718
+    },
+    {
+      "epoch": 0.4662108579634353,
+      "grad_norm": 1.59375,
+      "learning_rate": 1.8902181062753064e-05,
+      "loss": 1.0158,
+      "step": 2719
+    },
+    {
+      "epoch": 0.46638232205242514,
+      "grad_norm": 1.609375,
+      "learning_rate": 1.8901358254918005e-05,
+      "loss": 0.9048,
+      "step": 2720
+    },
+    {
+      "epoch": 0.466553786141415,
+      "grad_norm": 1.6953125,
+      "learning_rate": 1.8900535156776128e-05,
+      "loss": 1.176,
+      "step": 2721
+    },
+    {
+      "epoch": 0.46672525023040484,
+      "grad_norm": 1.5546875,
+      "learning_rate": 1.8899711768354288e-05,
+      "loss": 1.059,
+      "step": 2722
+    },
+    {
+      "epoch": 0.4668967143193947,
+      "grad_norm": 1.59375,
+      "learning_rate": 1.8898888089679332e-05,
+      "loss": 0.9411,
+      "step": 2723
+    },
+    {
+      "epoch": 0.4670681784083846,
+      "grad_norm": 1.6875,
+      "learning_rate": 1.8898064120778126e-05,
+      "loss": 0.944,
+      "step": 2724
+    },
+    {
+      "epoch": 0.46723964249737443,
+      "grad_norm": 1.5625,
+      "learning_rate": 1.889723986167754e-05,
+      "loss": 1.0461,
+      "step": 2725
+    },
+    {
+      "epoch": 0.4674111065863643,
+      "grad_norm": 1.59375,
+      "learning_rate": 1.889641531240446e-05,
+      "loss": 1.095,
+      "step": 2726
+    },
+    {
+      "epoch": 0.4675825706753542,
+      "grad_norm": 1.640625,
+      "learning_rate": 1.8895590472985775e-05,
+      "loss": 1.0206,
+      "step": 2727
+    },
+    {
+      "epoch": 0.467754034764344,
+      "grad_norm": 1.6015625,
+      "learning_rate": 1.889476534344839e-05,
+      "loss": 1.0456,
+      "step": 2728
+    },
+    {
+      "epoch": 0.4679254988533339,
+      "grad_norm": 1.5703125,
+      "learning_rate": 1.889393992381921e-05,
+      "loss": 1.0462,
+      "step": 2729
+    },
+    {
+      "epoch": 0.4680969629423238,
+      "grad_norm": 1.5546875,
+      "learning_rate": 1.8893114214125154e-05,
+      "loss": 1.0204,
+      "step": 2730
+    },
+    {
+      "epoch": 0.4682684270313136,
+      "grad_norm": 1.734375,
+      "learning_rate": 1.889228821439316e-05,
+      "loss": 1.047,
+      "step": 2731
+    },
+    {
+      "epoch": 0.4684398911203035,
+      "grad_norm": 1.5703125,
+      "learning_rate": 1.8891461924650165e-05,
+      "loss": 0.9852,
+      "step": 2732
+    },
+    {
+      "epoch": 0.46861135520929337,
+      "grad_norm": 1.6328125,
+      "learning_rate": 1.8890635344923106e-05,
+      "loss": 1.0183,
+      "step": 2733
+    },
+    {
+      "epoch": 0.4687828192982832,
+      "grad_norm": 1.875,
+      "learning_rate": 1.8889808475238957e-05,
+      "loss": 1.0666,
+      "step": 2734
+    },
+    {
+      "epoch": 0.4689542833872731,
+      "grad_norm": 1.578125,
+      "learning_rate": 1.8888981315624672e-05,
+      "loss": 0.9941,
+      "step": 2735
+    },
+    {
+      "epoch": 0.46912574747626296,
+      "grad_norm": 1.5859375,
+      "learning_rate": 1.8888153866107236e-05,
+      "loss": 1.0518,
+      "step": 2736
+    },
+    {
+      "epoch": 0.4692972115652528,
+      "grad_norm": 1.609375,
+      "learning_rate": 1.888732612671363e-05,
+      "loss": 1.0533,
+      "step": 2737
+    },
+    {
+      "epoch": 0.46946867565424266,
+      "grad_norm": 1.609375,
+      "learning_rate": 1.888649809747086e-05,
+      "loss": 1.0669,
+      "step": 2738
+    },
+    {
+      "epoch": 0.46964013974323254,
+      "grad_norm": 1.6484375,
+      "learning_rate": 1.8885669778405917e-05,
+      "loss": 1.0136,
+      "step": 2739
+    },
+    {
+      "epoch": 0.46981160383222237,
+      "grad_norm": 1.5390625,
+      "learning_rate": 1.8884841169545826e-05,
+      "loss": 1.0767,
+      "step": 2740
+    },
+    {
+      "epoch": 0.46998306792121225,
+      "grad_norm": 1.65625,
+      "learning_rate": 1.8884012270917603e-05,
+      "loss": 0.9982,
+      "step": 2741
+    },
+    {
+      "epoch": 0.47015453201020213,
+      "grad_norm": 1.5703125,
+      "learning_rate": 1.888318308254829e-05,
+      "loss": 1.0528,
+      "step": 2742
+    },
+    {
+      "epoch": 0.47032599609919196,
+      "grad_norm": 1.5859375,
+      "learning_rate": 1.8882353604464923e-05,
+      "loss": 1.0539,
+      "step": 2743
+    },
+    {
+      "epoch": 0.47049746018818184,
+      "grad_norm": 1.71875,
+      "learning_rate": 1.8881523836694556e-05,
+      "loss": 1.0387,
+      "step": 2744
+    },
+    {
+      "epoch": 0.4706689242771717,
+      "grad_norm": 1.5546875,
+      "learning_rate": 1.8880693779264255e-05,
+      "loss": 0.991,
+      "step": 2745
+    },
+    {
+      "epoch": 0.47084038836616154,
+      "grad_norm": 1.53125,
+      "learning_rate": 1.8879863432201086e-05,
+      "loss": 0.9775,
+      "step": 2746
+    },
+    {
+      "epoch": 0.4710118524551514,
+      "grad_norm": 1.5703125,
+      "learning_rate": 1.8879032795532132e-05,
+      "loss": 0.9837,
+      "step": 2747
+    },
+    {
+      "epoch": 0.4711833165441413,
+      "grad_norm": 1.578125,
+      "learning_rate": 1.8878201869284483e-05,
+      "loss": 0.9901,
+      "step": 2748
+    },
+    {
+      "epoch": 0.47135478063313113,
+      "grad_norm": 1.7265625,
+      "learning_rate": 1.8877370653485242e-05,
+      "loss": 1.1065,
+      "step": 2749
+    },
+    {
+      "epoch": 0.471526244722121,
+      "grad_norm": 1.5859375,
+      "learning_rate": 1.8876539148161514e-05,
+      "loss": 1.0289,
+      "step": 2750
+    },
+    {
+      "epoch": 0.4716977088111109,
+      "grad_norm": 1.5703125,
+      "learning_rate": 1.887570735334042e-05,
+      "loss": 1.0634,
+      "step": 2751
+    },
+    {
+      "epoch": 0.4718691729001007,
+      "grad_norm": 1.546875,
+      "learning_rate": 1.887487526904908e-05,
+      "loss": 0.9623,
+      "step": 2752
+    },
+    {
+      "epoch": 0.4720406369890906,
+      "grad_norm": 1.671875,
+      "learning_rate": 1.8874042895314643e-05,
+      "loss": 1.0526,
+      "step": 2753
+    },
+    {
+      "epoch": 0.4722121010780805,
+      "grad_norm": 1.6796875,
+      "learning_rate": 1.8873210232164248e-05,
+      "loss": 1.0761,
+      "step": 2754
+    },
+    {
+      "epoch": 0.4723835651670703,
+      "grad_norm": 1.6484375,
+      "learning_rate": 1.8872377279625057e-05,
+      "loss": 1.0276,
+      "step": 2755
+    },
+    {
+      "epoch": 0.4725550292560602,
+      "grad_norm": 1.578125,
+      "learning_rate": 1.887154403772423e-05,
+      "loss": 0.9716,
+      "step": 2756
+    },
+    {
+      "epoch": 0.47272649334505007,
+      "grad_norm": 1.5390625,
+      "learning_rate": 1.8870710506488944e-05,
+      "loss": 1.0427,
+      "step": 2757
+    },
+    {
+      "epoch": 0.4728979574340399,
+      "grad_norm": 1.578125,
+      "learning_rate": 1.8869876685946388e-05,
+      "loss": 1.0066,
+      "step": 2758
+    },
+    {
+      "epoch": 0.4730694215230298,
+      "grad_norm": 1.5625,
+      "learning_rate": 1.886904257612375e-05,
+      "loss": 1.0671,
+      "step": 2759
+    },
+    {
+      "epoch": 0.47324088561201966,
+      "grad_norm": 1.6015625,
+      "learning_rate": 1.8868208177048238e-05,
+      "loss": 0.9966,
+      "step": 2760
+    },
+    {
+      "epoch": 0.4734123497010095,
+      "grad_norm": 1.6484375,
+      "learning_rate": 1.8867373488747058e-05,
+      "loss": 1.091,
+      "step": 2761
+    },
+    {
+      "epoch": 0.47358381378999936,
+      "grad_norm": 1.5546875,
+      "learning_rate": 1.886653851124744e-05,
+      "loss": 1.0605,
+      "step": 2762
+    },
+    {
+      "epoch": 0.47375527787898924,
+      "grad_norm": 1.578125,
+      "learning_rate": 1.8865703244576613e-05,
+      "loss": 0.9792,
+      "step": 2763
+    },
+    {
+      "epoch": 0.47392674196797907,
+      "grad_norm": 1.6171875,
+      "learning_rate": 1.8864867688761818e-05,
+      "loss": 1.1191,
+      "step": 2764
+    },
+    {
+      "epoch": 0.47409820605696895,
+      "grad_norm": 1.59375,
+      "learning_rate": 1.8864031843830305e-05,
+      "loss": 1.0338,
+      "step": 2765
+    },
+    {
+      "epoch": 0.47426967014595883,
+      "grad_norm": 1.578125,
+      "learning_rate": 1.8863195709809336e-05,
+      "loss": 1.0051,
+      "step": 2766
+    },
+    {
+      "epoch": 0.47444113423494866,
+      "grad_norm": 1.6015625,
+      "learning_rate": 1.886235928672618e-05,
+      "loss": 0.9636,
+      "step": 2767
+    },
+    {
+      "epoch": 0.47461259832393854,
+      "grad_norm": 1.6484375,
+      "learning_rate": 1.8861522574608113e-05,
+      "loss": 1.0052,
+      "step": 2768
+    },
+    {
+      "epoch": 0.47478406241292836,
+      "grad_norm": 1.5859375,
+      "learning_rate": 1.8860685573482427e-05,
+      "loss": 1.0447,
+      "step": 2769
+    },
+    {
+      "epoch": 0.47495552650191825,
+      "grad_norm": 1.65625,
+      "learning_rate": 1.8859848283376417e-05,
+      "loss": 1.098,
+      "step": 2770
+    },
+    {
+      "epoch": 0.4751269905909081,
+      "grad_norm": 1.703125,
+      "learning_rate": 1.885901070431739e-05,
+      "loss": 1.0925,
+      "step": 2771
+    },
+    {
+      "epoch": 0.47529845467989795,
+      "grad_norm": 1.6015625,
+      "learning_rate": 1.8858172836332667e-05,
+      "loss": 0.9781,
+      "step": 2772
+    },
+    {
+      "epoch": 0.47546991876888783,
+      "grad_norm": 1.5859375,
+      "learning_rate": 1.885733467944957e-05,
+      "loss": 1.0624,
+      "step": 2773
+    },
+    {
+      "epoch": 0.4756413828578777,
+      "grad_norm": 1.6171875,
+      "learning_rate": 1.8856496233695435e-05,
+      "loss": 1.097,
+      "step": 2774
+    },
+    {
+      "epoch": 0.47581284694686754,
+      "grad_norm": 1.6484375,
+      "learning_rate": 1.885565749909761e-05,
+      "loss": 1.0228,
+      "step": 2775
+    },
+    {
+      "epoch": 0.4759843110358574,
+      "grad_norm": 1.59375,
+      "learning_rate": 1.885481847568344e-05,
+      "loss": 1.0667,
+      "step": 2776
+    },
+    {
+      "epoch": 0.4761557751248473,
+      "grad_norm": 1.5703125,
+      "learning_rate": 1.8853979163480302e-05,
+      "loss": 0.9551,
+      "step": 2777
+    },
+    {
+      "epoch": 0.47632723921383713,
+      "grad_norm": 1.6484375,
+      "learning_rate": 1.885313956251556e-05,
+      "loss": 1.0382,
+      "step": 2778
+    },
+    {
+      "epoch": 0.476498703302827,
+      "grad_norm": 1.6484375,
+      "learning_rate": 1.8852299672816596e-05,
+      "loss": 1.0474,
+      "step": 2779
+    },
+    {
+      "epoch": 0.4766701673918169,
+      "grad_norm": 1.6484375,
+      "learning_rate": 1.885145949441081e-05,
+      "loss": 0.9978,
+      "step": 2780
+    },
+    {
+      "epoch": 0.4768416314808067,
+      "grad_norm": 1.5546875,
+      "learning_rate": 1.8850619027325595e-05,
+      "loss": 0.9717,
+      "step": 2781
+    },
+    {
+      "epoch": 0.4770130955697966,
+      "grad_norm": 1.625,
+      "learning_rate": 1.884977827158837e-05,
+      "loss": 1.0869,
+      "step": 2782
+    },
+    {
+      "epoch": 0.4771845596587865,
+      "grad_norm": 1.578125,
+      "learning_rate": 1.8848937227226542e-05,
+      "loss": 1.065,
+      "step": 2783
+    },
+    {
+      "epoch": 0.4773560237477763,
+      "grad_norm": 1.765625,
+      "learning_rate": 1.8848095894267556e-05,
+      "loss": 1.1608,
+      "step": 2784
+    },
+    {
+      "epoch": 0.4775274878367662,
+      "grad_norm": 1.7265625,
+      "learning_rate": 1.884725427273884e-05,
+      "loss": 1.1015,
+      "step": 2785
+    },
+    {
+      "epoch": 0.47769895192575607,
+      "grad_norm": 1.640625,
+      "learning_rate": 1.884641236266785e-05,
+      "loss": 1.0337,
+      "step": 2786
+    },
+    {
+      "epoch": 0.4778704160147459,
+      "grad_norm": 1.515625,
+      "learning_rate": 1.8845570164082038e-05,
+      "loss": 0.9727,
+      "step": 2787
+    },
+    {
+      "epoch": 0.4780418801037358,
+      "grad_norm": 1.5859375,
+      "learning_rate": 1.8844727677008877e-05,
+      "loss": 1.0569,
+      "step": 2788
+    },
+    {
+      "epoch": 0.47821334419272565,
+      "grad_norm": 1.59375,
+      "learning_rate": 1.8843884901475835e-05,
+      "loss": 1.0511,
+      "step": 2789
+    },
+    {
+      "epoch": 0.4783848082817155,
+      "grad_norm": 1.6796875,
+      "learning_rate": 1.8843041837510408e-05,
+      "loss": 1.0175,
+      "step": 2790
+    },
+    {
+      "epoch": 0.47855627237070536,
+      "grad_norm": 1.7265625,
+      "learning_rate": 1.8842198485140084e-05,
+      "loss": 1.0046,
+      "step": 2791
+    },
+    {
+      "epoch": 0.47872773645969524,
+      "grad_norm": 1.703125,
+      "learning_rate": 1.884135484439237e-05,
+      "loss": 1.0533,
+      "step": 2792
+    },
+    {
+      "epoch": 0.47889920054868507,
+      "grad_norm": 1.7109375,
+      "learning_rate": 1.8840510915294785e-05,
+      "loss": 1.0449,
+      "step": 2793
+    },
+    {
+      "epoch": 0.47907066463767495,
+      "grad_norm": 1.53125,
+      "learning_rate": 1.8839666697874845e-05,
+      "loss": 0.9818,
+      "step": 2794
+    },
+    {
+      "epoch": 0.47924212872666483,
+      "grad_norm": 1.5625,
+      "learning_rate": 1.883882219216009e-05,
+      "loss": 1.0963,
+      "step": 2795
+    },
+    {
+      "epoch": 0.47941359281565465,
+      "grad_norm": 1.515625,
+      "learning_rate": 1.8837977398178058e-05,
+      "loss": 1.0716,
+      "step": 2796
+    },
+    {
+      "epoch": 0.47958505690464454,
+      "grad_norm": 1.6328125,
+      "learning_rate": 1.88371323159563e-05,
+      "loss": 0.9346,
+      "step": 2797
+    },
+    {
+      "epoch": 0.4797565209936344,
+      "grad_norm": 1.546875,
+      "learning_rate": 1.8836286945522384e-05,
+      "loss": 1.0463,
+      "step": 2798
+    },
+    {
+      "epoch": 0.47992798508262424,
+      "grad_norm": 1.53125,
+      "learning_rate": 1.8835441286903874e-05,
+      "loss": 1.0739,
+      "step": 2799
+    },
+    {
+      "epoch": 0.4800994491716141,
+      "grad_norm": 1.5078125,
+      "learning_rate": 1.883459534012835e-05,
+      "loss": 0.9795,
+      "step": 2800
+    },
+    {
+      "epoch": 0.4800994491716141,
+      "eval_loss": 0.8867102265357971,
+      "eval_runtime": 837.1871,
+      "eval_samples_per_second": 2.985,
+      "eval_steps_per_second": 2.985,
+      "step": 2800
+    },
+    {
+      "epoch": 0.480270913260604,
+      "grad_norm": 1.578125,
+      "learning_rate": 1.8833749105223406e-05,
+      "loss": 1.0323,
+      "step": 2801
+    },
+    {
+      "epoch": 0.48044237734959383,
+      "grad_norm": 1.6171875,
+      "learning_rate": 1.883290258221664e-05,
+      "loss": 1.1122,
+      "step": 2802
+    },
+    {
+      "epoch": 0.4806138414385837,
+      "grad_norm": 1.640625,
+      "learning_rate": 1.8832055771135657e-05,
+      "loss": 1.1051,
+      "step": 2803
+    },
+    {
+      "epoch": 0.4807853055275736,
+      "grad_norm": 1.546875,
+      "learning_rate": 1.8831208672008082e-05,
+      "loss": 1.0378,
+      "step": 2804
+    },
+    {
+      "epoch": 0.4809567696165634,
+      "grad_norm": 1.59375,
+      "learning_rate": 1.8830361284861532e-05,
+      "loss": 1.0658,
+      "step": 2805
+    },
+    {
+      "epoch": 0.4811282337055533,
+      "grad_norm": 1.640625,
+      "learning_rate": 1.882951360972365e-05,
+      "loss": 1.0673,
+      "step": 2806
+    },
+    {
+      "epoch": 0.4812996977945432,
+      "grad_norm": 1.5703125,
+      "learning_rate": 1.882866564662208e-05,
+      "loss": 0.9835,
+      "step": 2807
+    },
+    {
+      "epoch": 0.481471161883533,
+      "grad_norm": 1.5703125,
+      "learning_rate": 1.8827817395584474e-05,
+      "loss": 1.0338,
+      "step": 2808
+    },
+    {
+      "epoch": 0.4816426259725229,
+      "grad_norm": 1.6953125,
+      "learning_rate": 1.8826968856638504e-05,
+      "loss": 1.0203,
+      "step": 2809
+    },
+    {
+      "epoch": 0.48181409006151277,
+      "grad_norm": 1.7265625,
+      "learning_rate": 1.8826120029811844e-05,
+      "loss": 1.0846,
+      "step": 2810
+    },
+    {
+      "epoch": 0.4819855541505026,
+      "grad_norm": 1.609375,
+      "learning_rate": 1.882527091513217e-05,
+      "loss": 1.1038,
+      "step": 2811
+    },
+    {
+      "epoch": 0.4821570182394925,
+      "grad_norm": 1.46875,
+      "learning_rate": 1.882442151262718e-05,
+      "loss": 0.9386,
+      "step": 2812
+    },
+    {
+      "epoch": 0.48232848232848236,
+      "grad_norm": 1.546875,
+      "learning_rate": 1.882357182232457e-05,
+      "loss": 1.0621,
+      "step": 2813
+    },
+    {
+      "epoch": 0.4824999464174722,
+      "grad_norm": 1.59375,
+      "learning_rate": 1.882272184425206e-05,
+      "loss": 1.088,
+      "step": 2814
+    },
+    {
+      "epoch": 0.48267141050646206,
+      "grad_norm": 1.6640625,
+      "learning_rate": 1.8821871578437367e-05,
+      "loss": 1.0637,
+      "step": 2815
+    },
+    {
+      "epoch": 0.4828428745954519,
+      "grad_norm": 1.578125,
+      "learning_rate": 1.8821021024908223e-05,
+      "loss": 1.0955,
+      "step": 2816
+    },
+    {
+      "epoch": 0.48301433868444177,
+      "grad_norm": 1.5625,
+      "learning_rate": 1.8820170183692364e-05,
+      "loss": 0.9826,
+      "step": 2817
+    },
+    {
+      "epoch": 0.48318580277343165,
+      "grad_norm": 1.6015625,
+      "learning_rate": 1.8819319054817543e-05,
+      "loss": 1.0452,
+      "step": 2818
+    },
+    {
+      "epoch": 0.4833572668624215,
+      "grad_norm": 1.640625,
+      "learning_rate": 1.8818467638311516e-05,
+      "loss": 1.0884,
+      "step": 2819
+    },
+    {
+      "epoch": 0.48352873095141136,
+      "grad_norm": 1.6015625,
+      "learning_rate": 1.8817615934202055e-05,
+      "loss": 1.1022,
+      "step": 2820
+    },
+    {
+      "epoch": 0.48370019504040124,
+      "grad_norm": 1.5546875,
+      "learning_rate": 1.881676394251693e-05,
+      "loss": 1.0365,
+      "step": 2821
+    },
+    {
+      "epoch": 0.48387165912939106,
+      "grad_norm": 1.515625,
+      "learning_rate": 1.8815911663283936e-05,
+      "loss": 0.9715,
+      "step": 2822
+    },
+    {
+      "epoch": 0.48404312321838094,
+      "grad_norm": 1.578125,
+      "learning_rate": 1.8815059096530863e-05,
+      "loss": 1.056,
+      "step": 2823
+    },
+    {
+      "epoch": 0.4842145873073708,
+      "grad_norm": 1.5546875,
+      "learning_rate": 1.881420624228552e-05,
+      "loss": 1.0723,
+      "step": 2824
+    },
+    {
+      "epoch": 0.48438605139636065,
+      "grad_norm": 1.6484375,
+      "learning_rate": 1.8813353100575716e-05,
+      "loss": 0.942,
+      "step": 2825
+    },
+    {
+      "epoch": 0.48455751548535053,
+      "grad_norm": 1.5234375,
+      "learning_rate": 1.8812499671429286e-05,
+      "loss": 1.0086,
+      "step": 2826
+    },
+    {
+      "epoch": 0.4847289795743404,
+      "grad_norm": 1.6796875,
+      "learning_rate": 1.8811645954874053e-05,
+      "loss": 1.0392,
+      "step": 2827
+    },
+    {
+      "epoch": 0.48490044366333024,
+      "grad_norm": 1.4921875,
+      "learning_rate": 1.8810791950937864e-05,
+      "loss": 1.002,
+      "step": 2828
+    },
+    {
+      "epoch": 0.4850719077523201,
+      "grad_norm": 1.5546875,
+      "learning_rate": 1.880993765964857e-05,
+      "loss": 0.9703,
+      "step": 2829
+    },
+    {
+      "epoch": 0.48524337184131,
+      "grad_norm": 1.640625,
+      "learning_rate": 1.8809083081034035e-05,
+      "loss": 1.1082,
+      "step": 2830
+    },
+    {
+      "epoch": 0.4854148359302998,
+      "grad_norm": 1.640625,
+      "learning_rate": 1.8808228215122127e-05,
+      "loss": 0.9924,
+      "step": 2831
+    },
+    {
+      "epoch": 0.4855863000192897,
+      "grad_norm": 2.03125,
+      "learning_rate": 1.880737306194073e-05,
+      "loss": 1.0646,
+      "step": 2832
+    },
+    {
+      "epoch": 0.4857577641082796,
+      "grad_norm": 1.5625,
+      "learning_rate": 1.8806517621517733e-05,
+      "loss": 1.0148,
+      "step": 2833
+    },
+    {
+      "epoch": 0.4859292281972694,
+      "grad_norm": 1.6484375,
+      "learning_rate": 1.880566189388103e-05,
+      "loss": 0.9831,
+      "step": 2834
+    },
+    {
+      "epoch": 0.4861006922862593,
+      "grad_norm": 1.5078125,
+      "learning_rate": 1.8804805879058538e-05,
+      "loss": 0.9783,
+      "step": 2835
+    },
+    {
+      "epoch": 0.4862721563752492,
+      "grad_norm": 1.546875,
+      "learning_rate": 1.8803949577078172e-05,
+      "loss": 0.9806,
+      "step": 2836
+    },
+    {
+      "epoch": 0.486443620464239,
+      "grad_norm": 1.6171875,
+      "learning_rate": 1.8803092987967853e-05,
+      "loss": 1.0607,
+      "step": 2837
+    },
+    {
+      "epoch": 0.4866150845532289,
+      "grad_norm": 1.5546875,
+      "learning_rate": 1.8802236111755524e-05,
+      "loss": 0.9994,
+      "step": 2838
+    },
+    {
+      "epoch": 0.48678654864221876,
+      "grad_norm": 1.6328125,
+      "learning_rate": 1.880137894846913e-05,
+      "loss": 1.053,
+      "step": 2839
+    },
+    {
+      "epoch": 0.4869580127312086,
+      "grad_norm": 1.65625,
+      "learning_rate": 1.8800521498136622e-05,
+      "loss": 1.0909,
+      "step": 2840
+    },
+    {
+      "epoch": 0.48712947682019847,
+      "grad_norm": 1.6484375,
+      "learning_rate": 1.8799663760785973e-05,
+      "loss": 1.0357,
+      "step": 2841
+    },
+    {
+      "epoch": 0.48730094090918835,
+      "grad_norm": 1.609375,
+      "learning_rate": 1.8798805736445153e-05,
+      "loss": 1.0848,
+      "step": 2842
+    },
+    {
+      "epoch": 0.4874724049981782,
+      "grad_norm": 1.578125,
+      "learning_rate": 1.8797947425142137e-05,
+      "loss": 1.0419,
+      "step": 2843
+    },
+    {
+      "epoch": 0.48764386908716806,
+      "grad_norm": 1.6875,
+      "learning_rate": 1.879708882690493e-05,
+      "loss": 1.15,
+      "step": 2844
+    },
+    {
+      "epoch": 0.48781533317615794,
+      "grad_norm": 1.546875,
+      "learning_rate": 1.879622994176153e-05,
+      "loss": 1.1013,
+      "step": 2845
+    },
+    {
+      "epoch": 0.48798679726514776,
+      "grad_norm": 1.6875,
+      "learning_rate": 1.879537076973995e-05,
+      "loss": 1.0203,
+      "step": 2846
+    },
+    {
+      "epoch": 0.48815826135413765,
+      "grad_norm": 1.640625,
+      "learning_rate": 1.8794511310868208e-05,
+      "loss": 0.9994,
+      "step": 2847
+    },
+    {
+      "epoch": 0.4883297254431275,
+      "grad_norm": 1.6171875,
+      "learning_rate": 1.8793651565174333e-05,
+      "loss": 1.0928,
+      "step": 2848
+    },
+    {
+      "epoch": 0.48850118953211735,
+      "grad_norm": 1.671875,
+      "learning_rate": 1.879279153268637e-05,
+      "loss": 1.0635,
+      "step": 2849
+    },
+    {
+      "epoch": 0.48867265362110723,
+      "grad_norm": 1.65625,
+      "learning_rate": 1.879193121343236e-05,
+      "loss": 1.1594,
+      "step": 2850
+    },
+    {
+      "epoch": 0.4888441177100971,
+      "grad_norm": 1.6640625,
+      "learning_rate": 1.879107060744037e-05,
+      "loss": 0.9951,
+      "step": 2851
+    },
+    {
+      "epoch": 0.48901558179908694,
+      "grad_norm": 1.6328125,
+      "learning_rate": 1.8790209714738462e-05,
+      "loss": 1.0426,
+      "step": 2852
+    },
+    {
+      "epoch": 0.4891870458880768,
+      "grad_norm": 1.578125,
+      "learning_rate": 1.8789348535354714e-05,
+      "loss": 1.0581,
+      "step": 2853
+    },
+    {
+      "epoch": 0.4893585099770667,
+      "grad_norm": 1.6328125,
+      "learning_rate": 1.8788487069317208e-05,
+      "loss": 1.0374,
+      "step": 2854
+    },
+    {
+      "epoch": 0.48952997406605653,
+      "grad_norm": 1.640625,
+      "learning_rate": 1.878762531665405e-05,
+      "loss": 1.0132,
+      "step": 2855
+    },
+    {
+      "epoch": 0.4897014381550464,
+      "grad_norm": 1.625,
+      "learning_rate": 1.878676327739334e-05,
+      "loss": 1.0167,
+      "step": 2856
+    },
+    {
+      "epoch": 0.4898729022440363,
+      "grad_norm": 1.6640625,
+      "learning_rate": 1.878590095156319e-05,
+      "loss": 1.0232,
+      "step": 2857
+    },
+    {
+      "epoch": 0.4900443663330261,
+      "grad_norm": 1.546875,
+      "learning_rate": 1.8785038339191723e-05,
+      "loss": 1.0708,
+      "step": 2858
+    },
+    {
+      "epoch": 0.490215830422016,
+      "grad_norm": 1.6015625,
+      "learning_rate": 1.8784175440307076e-05,
+      "loss": 1.0704,
+      "step": 2859
+    },
+    {
+      "epoch": 0.4903872945110059,
+      "grad_norm": 1.640625,
+      "learning_rate": 1.878331225493739e-05,
+      "loss": 1.0848,
+      "step": 2860
+    },
+    {
+      "epoch": 0.4905587585999957,
+      "grad_norm": 1.734375,
+      "learning_rate": 1.878244878311082e-05,
+      "loss": 1.0802,
+      "step": 2861
+    },
+    {
+      "epoch": 0.4907302226889856,
+      "grad_norm": 1.5703125,
+      "learning_rate": 1.8781585024855517e-05,
+      "loss": 1.0438,
+      "step": 2862
+    },
+    {
+      "epoch": 0.4909016867779754,
+      "grad_norm": 1.640625,
+      "learning_rate": 1.878072098019966e-05,
+      "loss": 0.9577,
+      "step": 2863
+    },
+    {
+      "epoch": 0.4910731508669653,
+      "grad_norm": 1.6953125,
+      "learning_rate": 1.8779856649171427e-05,
+      "loss": 1.0269,
+      "step": 2864
+    },
+    {
+      "epoch": 0.49124461495595517,
+      "grad_norm": 1.65625,
+      "learning_rate": 1.877899203179901e-05,
+      "loss": 0.978,
+      "step": 2865
+    },
+    {
+      "epoch": 0.491416079044945,
+      "grad_norm": 1.7734375,
+      "learning_rate": 1.8778127128110602e-05,
+      "loss": 1.0305,
+      "step": 2866
+    },
+    {
+      "epoch": 0.4915875431339349,
+      "grad_norm": 1.5234375,
+      "learning_rate": 1.877726193813441e-05,
+      "loss": 1.0869,
+      "step": 2867
+    },
+    {
+      "epoch": 0.49175900722292476,
+      "grad_norm": 1.640625,
+      "learning_rate": 1.877639646189866e-05,
+      "loss": 1.0588,
+      "step": 2868
+    },
+    {
+      "epoch": 0.4919304713119146,
+      "grad_norm": 1.5703125,
+      "learning_rate": 1.8775530699431566e-05,
+      "loss": 1.0311,
+      "step": 2869
+    },
+    {
+      "epoch": 0.49210193540090447,
+      "grad_norm": 1.578125,
+      "learning_rate": 1.8774664650761373e-05,
+      "loss": 1.0879,
+      "step": 2870
+    },
+    {
+      "epoch": 0.49227339948989435,
+      "grad_norm": 1.625,
+      "learning_rate": 1.8773798315916324e-05,
+      "loss": 1.0146,
+      "step": 2871
+    },
+    {
+      "epoch": 0.4924448635788842,
+      "grad_norm": 1.609375,
+      "learning_rate": 1.8772931694924677e-05,
+      "loss": 1.0515,
+      "step": 2872
+    },
+    {
+      "epoch": 0.49261632766787405,
+      "grad_norm": 1.59375,
+      "learning_rate": 1.8772064787814686e-05,
+      "loss": 1.0498,
+      "step": 2873
+    },
+    {
+      "epoch": 0.49278779175686394,
+      "grad_norm": 1.6171875,
+      "learning_rate": 1.877119759461463e-05,
+      "loss": 0.9728,
+      "step": 2874
+    },
+    {
+      "epoch": 0.49295925584585376,
+      "grad_norm": 1.609375,
+      "learning_rate": 1.8770330115352797e-05,
+      "loss": 1.0578,
+      "step": 2875
+    },
+    {
+      "epoch": 0.49313071993484364,
+      "grad_norm": 1.6015625,
+      "learning_rate": 1.8769462350057467e-05,
+      "loss": 1.0311,
+      "step": 2876
+    },
+    {
+      "epoch": 0.4933021840238335,
+      "grad_norm": 1.6484375,
+      "learning_rate": 1.876859429875695e-05,
+      "loss": 1.0546,
+      "step": 2877
+    },
+    {
+      "epoch": 0.49347364811282335,
+      "grad_norm": 1.6875,
+      "learning_rate": 1.8767725961479548e-05,
+      "loss": 0.9694,
+      "step": 2878
+    },
+    {
+      "epoch": 0.49364511220181323,
+      "grad_norm": 1.5390625,
+      "learning_rate": 1.876685733825359e-05,
+      "loss": 0.9932,
+      "step": 2879
+    },
+    {
+      "epoch": 0.4938165762908031,
+      "grad_norm": 1.59375,
+      "learning_rate": 1.87659884291074e-05,
+      "loss": 1.0917,
+      "step": 2880
+    },
+    {
+      "epoch": 0.49398804037979294,
+      "grad_norm": 1.6953125,
+      "learning_rate": 1.876511923406932e-05,
+      "loss": 1.0881,
+      "step": 2881
+    },
+    {
+      "epoch": 0.4941595044687828,
+      "grad_norm": 1.609375,
+      "learning_rate": 1.8764249753167693e-05,
+      "loss": 1.0479,
+      "step": 2882
+    },
+    {
+      "epoch": 0.4943309685577727,
+      "grad_norm": 1.609375,
+      "learning_rate": 1.8763379986430884e-05,
+      "loss": 1.1382,
+      "step": 2883
+    },
+    {
+      "epoch": 0.4945024326467625,
+      "grad_norm": 1.6171875,
+      "learning_rate": 1.8762509933887248e-05,
+      "loss": 1.0526,
+      "step": 2884
+    },
+    {
+      "epoch": 0.4946738967357524,
+      "grad_norm": 1.5078125,
+      "learning_rate": 1.8761639595565166e-05,
+      "loss": 0.9602,
+      "step": 2885
+    },
+    {
+      "epoch": 0.4948453608247423,
+      "grad_norm": 1.7109375,
+      "learning_rate": 1.876076897149303e-05,
+      "loss": 0.9996,
+      "step": 2886
+    },
+    {
+      "epoch": 0.4950168249137321,
+      "grad_norm": 1.59375,
+      "learning_rate": 1.8759898061699223e-05,
+      "loss": 1.0105,
+      "step": 2887
+    },
+    {
+      "epoch": 0.495188289002722,
+      "grad_norm": 1.625,
+      "learning_rate": 1.875902686621215e-05,
+      "loss": 1.0099,
+      "step": 2888
+    },
+    {
+      "epoch": 0.4953597530917119,
+      "grad_norm": 1.578125,
+      "learning_rate": 1.8758155385060232e-05,
+      "loss": 1.0506,
+      "step": 2889
+    },
+    {
+      "epoch": 0.4955312171807017,
+      "grad_norm": 1.640625,
+      "learning_rate": 1.8757283618271887e-05,
+      "loss": 1.0378,
+      "step": 2890
+    },
+    {
+      "epoch": 0.4957026812696916,
+      "grad_norm": 1.6640625,
+      "learning_rate": 1.8756411565875545e-05,
+      "loss": 1.152,
+      "step": 2891
+    },
+    {
+      "epoch": 0.49587414535868146,
+      "grad_norm": 1.6328125,
+      "learning_rate": 1.875553922789965e-05,
+      "loss": 1.0665,
+      "step": 2892
+    },
+    {
+      "epoch": 0.4960456094476713,
+      "grad_norm": 1.5390625,
+      "learning_rate": 1.875466660437265e-05,
+      "loss": 1.0447,
+      "step": 2893
+    },
+    {
+      "epoch": 0.49621707353666117,
+      "grad_norm": 1.546875,
+      "learning_rate": 1.8753793695323e-05,
+      "loss": 0.9895,
+      "step": 2894
+    },
+    {
+      "epoch": 0.49638853762565105,
+      "grad_norm": 1.5859375,
+      "learning_rate": 1.875292050077918e-05,
+      "loss": 0.9806,
+      "step": 2895
+    },
+    {
+      "epoch": 0.4965600017146409,
+      "grad_norm": 1.65625,
+      "learning_rate": 1.8752047020769663e-05,
+      "loss": 1.0423,
+      "step": 2896
+    },
+    {
+      "epoch": 0.49673146580363076,
+      "grad_norm": 1.8203125,
+      "learning_rate": 1.875117325532293e-05,
+      "loss": 1.0898,
+      "step": 2897
+    },
+    {
+      "epoch": 0.49690292989262064,
+      "grad_norm": 1.625,
+      "learning_rate": 1.8750299204467485e-05,
+      "loss": 1.0848,
+      "step": 2898
+    },
+    {
+      "epoch": 0.49707439398161046,
+      "grad_norm": 1.53125,
+      "learning_rate": 1.8749424868231837e-05,
+      "loss": 0.9389,
+      "step": 2899
+    },
+    {
+      "epoch": 0.49724585807060034,
+      "grad_norm": 1.6953125,
+      "learning_rate": 1.874855024664449e-05,
+      "loss": 1.0633,
+      "step": 2900
+    },
+    {
+      "epoch": 0.4974173221595902,
+      "grad_norm": 1.578125,
+      "learning_rate": 1.874767533973398e-05,
+      "loss": 1.0126,
+      "step": 2901
+    },
+    {
+      "epoch": 0.49758878624858005,
+      "grad_norm": 1.6328125,
+      "learning_rate": 1.874680014752883e-05,
+      "loss": 1.0707,
+      "step": 2902
+    },
+    {
+      "epoch": 0.49776025033756993,
+      "grad_norm": 1.7578125,
+      "learning_rate": 1.87459246700576e-05,
+      "loss": 1.0622,
+      "step": 2903
+    },
+    {
+      "epoch": 0.4979317144265598,
+      "grad_norm": 1.6484375,
+      "learning_rate": 1.8745048907348824e-05,
+      "loss": 1.0176,
+      "step": 2904
+    },
+    {
+      "epoch": 0.49810317851554964,
+      "grad_norm": 1.5625,
+      "learning_rate": 1.874417285943108e-05,
+      "loss": 1.0011,
+      "step": 2905
+    },
+    {
+      "epoch": 0.4982746426045395,
+      "grad_norm": 1.6015625,
+      "learning_rate": 1.8743296526332924e-05,
+      "loss": 1.0018,
+      "step": 2906
+    },
+    {
+      "epoch": 0.4984461066935294,
+      "grad_norm": 1.59375,
+      "learning_rate": 1.8742419908082946e-05,
+      "loss": 1.0249,
+      "step": 2907
+    },
+    {
+      "epoch": 0.4986175707825192,
+      "grad_norm": 1.5,
+      "learning_rate": 1.8741543004709735e-05,
+      "loss": 0.851,
+      "step": 2908
+    },
+    {
+      "epoch": 0.4987890348715091,
+      "grad_norm": 1.6015625,
+      "learning_rate": 1.874066581624189e-05,
+      "loss": 1.0617,
+      "step": 2909
+    },
+    {
+      "epoch": 0.498960498960499,
+      "grad_norm": 1.6875,
+      "learning_rate": 1.8739788342708016e-05,
+      "loss": 0.9858,
+      "step": 2910
+    },
+    {
+      "epoch": 0.4991319630494888,
+      "grad_norm": 1.546875,
+      "learning_rate": 1.8738910584136735e-05,
+      "loss": 1.0821,
+      "step": 2911
+    },
+    {
+      "epoch": 0.4993034271384787,
+      "grad_norm": 1.5625,
+      "learning_rate": 1.873803254055667e-05,
+      "loss": 1.0519,
+      "step": 2912
+    },
+    {
+      "epoch": 0.4994748912274685,
+      "grad_norm": 1.515625,
+      "learning_rate": 1.873715421199646e-05,
+      "loss": 0.9833,
+      "step": 2913
+    },
+    {
+      "epoch": 0.4996463553164584,
+      "grad_norm": 1.609375,
+      "learning_rate": 1.8736275598484753e-05,
+      "loss": 1.0917,
+      "step": 2914
+    },
+    {
+      "epoch": 0.4998178194054483,
+      "grad_norm": 1.5546875,
+      "learning_rate": 1.8735396700050202e-05,
+      "loss": 0.992,
+      "step": 2915
+    },
+    {
+      "epoch": 0.4999892834944381,
+      "grad_norm": 1.546875,
+      "learning_rate": 1.8734517516721467e-05,
+      "loss": 0.9666,
+      "step": 2916
+    },
+    {
+      "epoch": 0.500160747583428,
+      "grad_norm": 1.5546875,
+      "learning_rate": 1.8733638048527223e-05,
+      "loss": 1.0,
+      "step": 2917
+    },
+    {
+      "epoch": 0.5003322116724178,
+      "grad_norm": 1.5546875,
+      "learning_rate": 1.8732758295496158e-05,
+      "loss": 0.9783,
+      "step": 2918
+    },
+    {
+      "epoch": 0.5005036757614078,
+      "grad_norm": 1.640625,
+      "learning_rate": 1.8731878257656956e-05,
+      "loss": 1.1255,
+      "step": 2919
+    },
+    {
+      "epoch": 0.5006751398503976,
+      "grad_norm": 1.4921875,
+      "learning_rate": 1.8730997935038328e-05,
+      "loss": 1.0204,
+      "step": 2920
+    },
+    {
+      "epoch": 0.5008466039393874,
+      "grad_norm": 1.5703125,
+      "learning_rate": 1.8730117327668975e-05,
+      "loss": 0.9479,
+      "step": 2921
+    },
+    {
+      "epoch": 0.5010180680283773,
+      "grad_norm": 1.625,
+      "learning_rate": 1.8729236435577625e-05,
+      "loss": 0.9952,
+      "step": 2922
+    },
+    {
+      "epoch": 0.5011895321173672,
+      "grad_norm": 1.609375,
+      "learning_rate": 1.8728355258793e-05,
+      "loss": 1.0075,
+      "step": 2923
+    },
+    {
+      "epoch": 0.501360996206357,
+      "grad_norm": 1.6796875,
+      "learning_rate": 1.8727473797343846e-05,
+      "loss": 1.0614,
+      "step": 2924
+    },
+    {
+      "epoch": 0.5015324602953469,
+      "grad_norm": 1.640625,
+      "learning_rate": 1.8726592051258906e-05,
+      "loss": 0.9874,
+      "step": 2925
+    },
+    {
+      "epoch": 0.5017039243843368,
+      "grad_norm": 1.625,
+      "learning_rate": 1.8725710020566936e-05,
+      "loss": 1.0229,
+      "step": 2926
+    },
+    {
+      "epoch": 0.5018753884733266,
+      "grad_norm": 1.5703125,
+      "learning_rate": 1.8724827705296706e-05,
+      "loss": 1.0575,
+      "step": 2927
+    },
+    {
+      "epoch": 0.5020468525623165,
+      "grad_norm": 1.6328125,
+      "learning_rate": 1.872394510547699e-05,
+      "loss": 1.0114,
+      "step": 2928
+    },
+    {
+      "epoch": 0.5022183166513063,
+      "grad_norm": 1.6796875,
+      "learning_rate": 1.872306222113657e-05,
+      "loss": 1.1254,
+      "step": 2929
+    },
+    {
+      "epoch": 0.5023897807402962,
+      "grad_norm": 1.59375,
+      "learning_rate": 1.8722179052304245e-05,
+      "loss": 1.0574,
+      "step": 2930
+    },
+    {
+      "epoch": 0.5025612448292861,
+      "grad_norm": 1.625,
+      "learning_rate": 1.8721295599008815e-05,
+      "loss": 1.0549,
+      "step": 2931
+    },
+    {
+      "epoch": 0.5027327089182759,
+      "grad_norm": 1.5078125,
+      "learning_rate": 1.8720411861279094e-05,
+      "loss": 0.9429,
+      "step": 2932
+    },
+    {
+      "epoch": 0.5029041730072658,
+      "grad_norm": 1.5546875,
+      "learning_rate": 1.8719527839143906e-05,
+      "loss": 0.9691,
+      "step": 2933
+    },
+    {
+      "epoch": 0.5030756370962557,
+      "grad_norm": 1.5859375,
+      "learning_rate": 1.8718643532632083e-05,
+      "loss": 1.0602,
+      "step": 2934
+    },
+    {
+      "epoch": 0.5032471011852455,
+      "grad_norm": 1.6484375,
+      "learning_rate": 1.8717758941772458e-05,
+      "loss": 1.0318,
+      "step": 2935
+    },
+    {
+      "epoch": 0.5034185652742353,
+      "grad_norm": 1.6484375,
+      "learning_rate": 1.8716874066593885e-05,
+      "loss": 0.9521,
+      "step": 2936
+    },
+    {
+      "epoch": 0.5035900293632253,
+      "grad_norm": 1.609375,
+      "learning_rate": 1.871598890712523e-05,
+      "loss": 1.0152,
+      "step": 2937
+    },
+    {
+      "epoch": 0.5037614934522151,
+      "grad_norm": 1.6171875,
+      "learning_rate": 1.8715103463395352e-05,
+      "loss": 1.0642,
+      "step": 2938
+    },
+    {
+      "epoch": 0.5039329575412049,
+      "grad_norm": 1.65625,
+      "learning_rate": 1.8714217735433132e-05,
+      "loss": 1.0613,
+      "step": 2939
+    },
+    {
+      "epoch": 0.5041044216301949,
+      "grad_norm": 1.703125,
+      "learning_rate": 1.871333172326746e-05,
+      "loss": 1.142,
+      "step": 2940
+    },
+    {
+      "epoch": 0.5042758857191847,
+      "grad_norm": 1.65625,
+      "learning_rate": 1.8712445426927225e-05,
+      "loss": 1.0557,
+      "step": 2941
+    },
+    {
+      "epoch": 0.5044473498081745,
+      "grad_norm": 1.5859375,
+      "learning_rate": 1.8711558846441336e-05,
+      "loss": 1.0868,
+      "step": 2942
+    },
+    {
+      "epoch": 0.5046188138971645,
+      "grad_norm": 1.671875,
+      "learning_rate": 1.871067198183871e-05,
+      "loss": 1.0208,
+      "step": 2943
+    },
+    {
+      "epoch": 0.5047902779861543,
+      "grad_norm": 1.65625,
+      "learning_rate": 1.870978483314827e-05,
+      "loss": 1.069,
+      "step": 2944
+    },
+    {
+      "epoch": 0.5049617420751441,
+      "grad_norm": 1.6875,
+      "learning_rate": 1.870889740039895e-05,
+      "loss": 1.0765,
+      "step": 2945
+    },
+    {
+      "epoch": 0.505133206164134,
+      "grad_norm": 1.5859375,
+      "learning_rate": 1.8708009683619684e-05,
+      "loss": 0.9706,
+      "step": 2946
+    },
+    {
+      "epoch": 0.5053046702531239,
+      "grad_norm": 1.65625,
+      "learning_rate": 1.870712168283944e-05,
+      "loss": 1.0681,
+      "step": 2947
+    },
+    {
+      "epoch": 0.5054761343421137,
+      "grad_norm": 1.5703125,
+      "learning_rate": 1.8706233398087166e-05,
+      "loss": 0.9911,
+      "step": 2948
+    },
+    {
+      "epoch": 0.5056475984311036,
+      "grad_norm": 1.7265625,
+      "learning_rate": 1.8705344829391835e-05,
+      "loss": 1.1497,
+      "step": 2949
+    },
+    {
+      "epoch": 0.5058190625200935,
+      "grad_norm": 1.6328125,
+      "learning_rate": 1.8704455976782427e-05,
+      "loss": 0.9842,
+      "step": 2950
+    },
+    {
+      "epoch": 0.5059905266090833,
+      "grad_norm": 1.6328125,
+      "learning_rate": 1.8703566840287934e-05,
+      "loss": 1.07,
+      "step": 2951
+    },
+    {
+      "epoch": 0.5061619906980732,
+      "grad_norm": 1.5625,
+      "learning_rate": 1.8702677419937353e-05,
+      "loss": 0.9317,
+      "step": 2952
+    },
+    {
+      "epoch": 0.506333454787063,
+      "grad_norm": 1.5859375,
+      "learning_rate": 1.870178771575969e-05,
+      "loss": 1.101,
+      "step": 2953
+    },
+    {
+      "epoch": 0.5065049188760529,
+      "grad_norm": 1.59375,
+      "learning_rate": 1.8700897727783957e-05,
+      "loss": 1.1308,
+      "step": 2954
+    },
+    {
+      "epoch": 0.5066763829650428,
+      "grad_norm": 1.5390625,
+      "learning_rate": 1.8700007456039188e-05,
+      "loss": 0.9863,
+      "step": 2955
+    },
+    {
+      "epoch": 0.5068478470540326,
+      "grad_norm": 1.6796875,
+      "learning_rate": 1.8699116900554414e-05,
+      "loss": 1.0991,
+      "step": 2956
+    },
+    {
+      "epoch": 0.5070193111430225,
+      "grad_norm": 1.53125,
+      "learning_rate": 1.8698226061358685e-05,
+      "loss": 1.126,
+      "step": 2957
+    },
+    {
+      "epoch": 0.5071907752320124,
+      "grad_norm": 1.5390625,
+      "learning_rate": 1.8697334938481044e-05,
+      "loss": 1.0232,
+      "step": 2958
+    },
+    {
+      "epoch": 0.5073622393210022,
+      "grad_norm": 1.65625,
+      "learning_rate": 1.869644353195056e-05,
+      "loss": 1.0008,
+      "step": 2959
+    },
+    {
+      "epoch": 0.507533703409992,
+      "grad_norm": 1.6171875,
+      "learning_rate": 1.8695551841796305e-05,
+      "loss": 1.014,
+      "step": 2960
+    },
+    {
+      "epoch": 0.507705167498982,
+      "grad_norm": 1.578125,
+      "learning_rate": 1.869465986804736e-05,
+      "loss": 0.9823,
+      "step": 2961
+    },
+    {
+      "epoch": 0.5078766315879718,
+      "grad_norm": 1.5546875,
+      "learning_rate": 1.8693767610732815e-05,
+      "loss": 0.9592,
+      "step": 2962
+    },
+    {
+      "epoch": 0.5080480956769616,
+      "grad_norm": 1.578125,
+      "learning_rate": 1.8692875069881773e-05,
+      "loss": 1.0477,
+      "step": 2963
+    },
+    {
+      "epoch": 0.5082195597659516,
+      "grad_norm": 1.671875,
+      "learning_rate": 1.8691982245523336e-05,
+      "loss": 1.0331,
+      "step": 2964
+    },
+    {
+      "epoch": 0.5083910238549414,
+      "grad_norm": 1.5859375,
+      "learning_rate": 1.8691089137686633e-05,
+      "loss": 0.9992,
+      "step": 2965
+    },
+    {
+      "epoch": 0.5085624879439312,
+      "grad_norm": 1.5546875,
+      "learning_rate": 1.869019574640078e-05,
+      "loss": 1.0193,
+      "step": 2966
+    },
+    {
+      "epoch": 0.5087339520329212,
+      "grad_norm": 1.5546875,
+      "learning_rate": 1.8689302071694925e-05,
+      "loss": 1.021,
+      "step": 2967
+    },
+    {
+      "epoch": 0.508905416121911,
+      "grad_norm": 1.6640625,
+      "learning_rate": 1.8688408113598205e-05,
+      "loss": 1.0477,
+      "step": 2968
+    },
+    {
+      "epoch": 0.5090768802109008,
+      "grad_norm": 1.546875,
+      "learning_rate": 1.868751387213978e-05,
+      "loss": 0.9475,
+      "step": 2969
+    },
+    {
+      "epoch": 0.5092483442998907,
+      "grad_norm": 1.6015625,
+      "learning_rate": 1.868661934734881e-05,
+      "loss": 1.0664,
+      "step": 2970
+    },
+    {
+      "epoch": 0.5094198083888806,
+      "grad_norm": 1.609375,
+      "learning_rate": 1.8685724539254478e-05,
+      "loss": 1.0631,
+      "step": 2971
+    },
+    {
+      "epoch": 0.5095912724778704,
+      "grad_norm": 1.578125,
+      "learning_rate": 1.8684829447885958e-05,
+      "loss": 0.9959,
+      "step": 2972
+    },
+    {
+      "epoch": 0.5097627365668603,
+      "grad_norm": 1.640625,
+      "learning_rate": 1.868393407327245e-05,
+      "loss": 1.026,
+      "step": 2973
+    },
+    {
+      "epoch": 0.5099342006558502,
+      "grad_norm": 1.6015625,
+      "learning_rate": 1.8683038415443143e-05,
+      "loss": 0.9877,
+      "step": 2974
+    },
+    {
+      "epoch": 0.51010566474484,
+      "grad_norm": 1.5703125,
+      "learning_rate": 1.8682142474427264e-05,
+      "loss": 1.0109,
+      "step": 2975
+    },
+    {
+      "epoch": 0.5102771288338299,
+      "grad_norm": 1.671875,
+      "learning_rate": 1.868124625025402e-05,
+      "loss": 1.0571,
+      "step": 2976
+    },
+    {
+      "epoch": 0.5104485929228197,
+      "grad_norm": 1.6328125,
+      "learning_rate": 1.8680349742952648e-05,
+      "loss": 1.0695,
+      "step": 2977
+    },
+    {
+      "epoch": 0.5106200570118096,
+      "grad_norm": 1.5546875,
+      "learning_rate": 1.867945295255238e-05,
+      "loss": 1.0911,
+      "step": 2978
+    },
+    {
+      "epoch": 0.5107915211007995,
+      "grad_norm": 1.53125,
+      "learning_rate": 1.8678555879082473e-05,
+      "loss": 1.0142,
+      "step": 2979
+    },
+    {
+      "epoch": 0.5109629851897893,
+      "grad_norm": 1.6015625,
+      "learning_rate": 1.8677658522572173e-05,
+      "loss": 1.0067,
+      "step": 2980
+    },
+    {
+      "epoch": 0.5111344492787792,
+      "grad_norm": 1.609375,
+      "learning_rate": 1.8676760883050754e-05,
+      "loss": 0.9649,
+      "step": 2981
+    },
+    {
+      "epoch": 0.511305913367769,
+      "grad_norm": 1.5390625,
+      "learning_rate": 1.867586296054749e-05,
+      "loss": 1.0444,
+      "step": 2982
+    },
+    {
+      "epoch": 0.5114773774567589,
+      "grad_norm": 1.578125,
+      "learning_rate": 1.8674964755091663e-05,
+      "loss": 0.9835,
+      "step": 2983
+    },
+    {
+      "epoch": 0.5116488415457487,
+      "grad_norm": 1.5546875,
+      "learning_rate": 1.8674066266712567e-05,
+      "loss": 1.0934,
+      "step": 2984
+    },
+    {
+      "epoch": 0.5118203056347386,
+      "grad_norm": 1.5703125,
+      "learning_rate": 1.8673167495439507e-05,
+      "loss": 1.061,
+      "step": 2985
+    },
+    {
+      "epoch": 0.5119917697237285,
+      "grad_norm": 1.59375,
+      "learning_rate": 1.8672268441301797e-05,
+      "loss": 1.0825,
+      "step": 2986
+    },
+    {
+      "epoch": 0.5121632338127183,
+      "grad_norm": 1.640625,
+      "learning_rate": 1.8671369104328757e-05,
+      "loss": 1.0551,
+      "step": 2987
+    },
+    {
+      "epoch": 0.5123346979017082,
+      "grad_norm": 1.5625,
+      "learning_rate": 1.8670469484549716e-05,
+      "loss": 1.0457,
+      "step": 2988
+    },
+    {
+      "epoch": 0.5125061619906981,
+      "grad_norm": 1.578125,
+      "learning_rate": 1.8669569581994014e-05,
+      "loss": 1.0619,
+      "step": 2989
+    },
+    {
+      "epoch": 0.5126776260796879,
+      "grad_norm": 1.640625,
+      "learning_rate": 1.8668669396691003e-05,
+      "loss": 1.0387,
+      "step": 2990
+    },
+    {
+      "epoch": 0.5128490901686777,
+      "grad_norm": 1.546875,
+      "learning_rate": 1.8667768928670037e-05,
+      "loss": 0.9911,
+      "step": 2991
+    },
+    {
+      "epoch": 0.5130205542576677,
+      "grad_norm": 1.6875,
+      "learning_rate": 1.8666868177960492e-05,
+      "loss": 1.0664,
+      "step": 2992
+    },
+    {
+      "epoch": 0.5131920183466575,
+      "grad_norm": 1.5859375,
+      "learning_rate": 1.8665967144591733e-05,
+      "loss": 1.076,
+      "step": 2993
+    },
+    {
+      "epoch": 0.5133634824356473,
+      "grad_norm": 1.609375,
+      "learning_rate": 1.8665065828593155e-05,
+      "loss": 1.0758,
+      "step": 2994
+    },
+    {
+      "epoch": 0.5135349465246373,
+      "grad_norm": 1.609375,
+      "learning_rate": 1.8664164229994153e-05,
+      "loss": 1.1051,
+      "step": 2995
+    },
+    {
+      "epoch": 0.5137064106136271,
+      "grad_norm": 1.6328125,
+      "learning_rate": 1.8663262348824127e-05,
+      "loss": 1.0632,
+      "step": 2996
+    },
+    {
+      "epoch": 0.5138778747026169,
+      "grad_norm": 1.609375,
+      "learning_rate": 1.8662360185112495e-05,
+      "loss": 1.0605,
+      "step": 2997
+    },
+    {
+      "epoch": 0.5140493387916069,
+      "grad_norm": 1.53125,
+      "learning_rate": 1.8661457738888673e-05,
+      "loss": 1.0795,
+      "step": 2998
+    },
+    {
+      "epoch": 0.5142208028805967,
+      "grad_norm": 1.6328125,
+      "learning_rate": 1.86605550101821e-05,
+      "loss": 0.9883,
+      "step": 2999
+    },
+    {
+      "epoch": 0.5143922669695865,
+      "grad_norm": 1.5546875,
+      "learning_rate": 1.8659651999022218e-05,
+      "loss": 1.006,
+      "step": 3000
+    },
+    {
+      "epoch": 0.5145637310585764,
+      "grad_norm": 1.59375,
+      "learning_rate": 1.8658748705438474e-05,
+      "loss": 1.0278,
+      "step": 3001
+    },
+    {
+      "epoch": 0.5147351951475663,
+      "grad_norm": 1.5546875,
+      "learning_rate": 1.865784512946033e-05,
+      "loss": 0.9796,
+      "step": 3002
+    },
+    {
+      "epoch": 0.5149066592365561,
+      "grad_norm": 1.5625,
+      "learning_rate": 1.8656941271117252e-05,
+      "loss": 0.9706,
+      "step": 3003
+    },
+    {
+      "epoch": 0.515078123325546,
+      "grad_norm": 1.5625,
+      "learning_rate": 1.865603713043872e-05,
+      "loss": 0.9346,
+      "step": 3004
+    },
+    {
+      "epoch": 0.5152495874145359,
+      "grad_norm": 1.609375,
+      "learning_rate": 1.865513270745422e-05,
+      "loss": 1.0387,
+      "step": 3005
+    },
+    {
+      "epoch": 0.5154210515035257,
+      "grad_norm": 1.7109375,
+      "learning_rate": 1.8654228002193255e-05,
+      "loss": 1.0278,
+      "step": 3006
+    },
+    {
+      "epoch": 0.5155925155925156,
+      "grad_norm": 1.609375,
+      "learning_rate": 1.865332301468532e-05,
+      "loss": 1.0926,
+      "step": 3007
+    },
+    {
+      "epoch": 0.5157639796815054,
+      "grad_norm": 1.59375,
+      "learning_rate": 1.8652417744959942e-05,
+      "loss": 1.0507,
+      "step": 3008
+    },
+    {
+      "epoch": 0.5159354437704953,
+      "grad_norm": 1.59375,
+      "learning_rate": 1.865151219304664e-05,
+      "loss": 1.0022,
+      "step": 3009
+    },
+    {
+      "epoch": 0.5161069078594852,
+      "grad_norm": 1.5390625,
+      "learning_rate": 1.8650606358974942e-05,
+      "loss": 1.0055,
+      "step": 3010
+    },
+    {
+      "epoch": 0.516278371948475,
+      "grad_norm": 1.625,
+      "learning_rate": 1.8649700242774397e-05,
+      "loss": 1.0461,
+      "step": 3011
+    },
+    {
+      "epoch": 0.5164498360374649,
+      "grad_norm": 1.5859375,
+      "learning_rate": 1.8648793844474556e-05,
+      "loss": 1.0554,
+      "step": 3012
+    },
+    {
+      "epoch": 0.5166213001264548,
+      "grad_norm": 1.6328125,
+      "learning_rate": 1.8647887164104976e-05,
+      "loss": 1.0602,
+      "step": 3013
+    },
+    {
+      "epoch": 0.5167927642154446,
+      "grad_norm": 1.6328125,
+      "learning_rate": 1.8646980201695236e-05,
+      "loss": 1.0584,
+      "step": 3014
+    },
+    {
+      "epoch": 0.5169642283044344,
+      "grad_norm": 1.625,
+      "learning_rate": 1.8646072957274906e-05,
+      "loss": 1.0742,
+      "step": 3015
+    },
+    {
+      "epoch": 0.5171356923934244,
+      "grad_norm": 1.59375,
+      "learning_rate": 1.8645165430873578e-05,
+      "loss": 1.0544,
+      "step": 3016
+    },
+    {
+      "epoch": 0.5173071564824142,
+      "grad_norm": 1.671875,
+      "learning_rate": 1.8644257622520857e-05,
+      "loss": 1.0175,
+      "step": 3017
+    },
+    {
+      "epoch": 0.517478620571404,
+      "grad_norm": 1.6640625,
+      "learning_rate": 1.864334953224634e-05,
+      "loss": 0.9962,
+      "step": 3018
+    },
+    {
+      "epoch": 0.517650084660394,
+      "grad_norm": 1.5703125,
+      "learning_rate": 1.8642441160079644e-05,
+      "loss": 0.9973,
+      "step": 3019
+    },
+    {
+      "epoch": 0.5178215487493838,
+      "grad_norm": 1.6328125,
+      "learning_rate": 1.86415325060504e-05,
+      "loss": 1.0568,
+      "step": 3020
+    },
+    {
+      "epoch": 0.5179930128383736,
+      "grad_norm": 1.5390625,
+      "learning_rate": 1.864062357018824e-05,
+      "loss": 0.9947,
+      "step": 3021
+    },
+    {
+      "epoch": 0.5181644769273636,
+      "grad_norm": 1.5703125,
+      "learning_rate": 1.863971435252281e-05,
+      "loss": 0.9692,
+      "step": 3022
+    },
+    {
+      "epoch": 0.5183359410163534,
+      "grad_norm": 1.59375,
+      "learning_rate": 1.8638804853083757e-05,
+      "loss": 1.0223,
+      "step": 3023
+    },
+    {
+      "epoch": 0.5185074051053432,
+      "grad_norm": 1.5234375,
+      "learning_rate": 1.863789507190075e-05,
+      "loss": 0.9897,
+      "step": 3024
+    },
+    {
+      "epoch": 0.5186788691943331,
+      "grad_norm": 1.59375,
+      "learning_rate": 1.8636985009003456e-05,
+      "loss": 1.0149,
+      "step": 3025
+    },
+    {
+      "epoch": 0.518850333283323,
+      "grad_norm": 1.5390625,
+      "learning_rate": 1.8636074664421556e-05,
+      "loss": 1.0892,
+      "step": 3026
+    },
+    {
+      "epoch": 0.5190217973723128,
+      "grad_norm": 1.5625,
+      "learning_rate": 1.8635164038184742e-05,
+      "loss": 1.1259,
+      "step": 3027
+    },
+    {
+      "epoch": 0.5191932614613027,
+      "grad_norm": 1.5546875,
+      "learning_rate": 1.8634253130322714e-05,
+      "loss": 1.0054,
+      "step": 3028
+    },
+    {
+      "epoch": 0.5193647255502926,
+      "grad_norm": 1.5546875,
+      "learning_rate": 1.8633341940865172e-05,
+      "loss": 1.0098,
+      "step": 3029
+    },
+    {
+      "epoch": 0.5195361896392824,
+      "grad_norm": 1.6953125,
+      "learning_rate": 1.8632430469841844e-05,
+      "loss": 1.0955,
+      "step": 3030
+    },
+    {
+      "epoch": 0.5197076537282723,
+      "grad_norm": 1.640625,
+      "learning_rate": 1.863151871728245e-05,
+      "loss": 1.0311,
+      "step": 3031
+    },
+    {
+      "epoch": 0.5198791178172621,
+      "grad_norm": 1.5,
+      "learning_rate": 1.863060668321673e-05,
+      "loss": 0.9735,
+      "step": 3032
+    },
+    {
+      "epoch": 0.520050581906252,
+      "grad_norm": 1.6953125,
+      "learning_rate": 1.862969436767442e-05,
+      "loss": 1.166,
+      "step": 3033
+    },
+    {
+      "epoch": 0.5202220459952419,
+      "grad_norm": 1.6015625,
+      "learning_rate": 1.8628781770685282e-05,
+      "loss": 0.9992,
+      "step": 3034
+    },
+    {
+      "epoch": 0.5203935100842317,
+      "grad_norm": 1.625,
+      "learning_rate": 1.8627868892279083e-05,
+      "loss": 0.9968,
+      "step": 3035
+    },
+    {
+      "epoch": 0.5205649741732216,
+      "grad_norm": 1.7734375,
+      "learning_rate": 1.8626955732485585e-05,
+      "loss": 1.084,
+      "step": 3036
+    },
+    {
+      "epoch": 0.5207364382622115,
+      "grad_norm": 1.5546875,
+      "learning_rate": 1.8626042291334572e-05,
+      "loss": 0.9351,
+      "step": 3037
+    },
+    {
+      "epoch": 0.5209079023512013,
+      "grad_norm": 1.7109375,
+      "learning_rate": 1.8625128568855844e-05,
+      "loss": 1.0052,
+      "step": 3038
+    },
+    {
+      "epoch": 0.5210793664401911,
+      "grad_norm": 1.6015625,
+      "learning_rate": 1.862421456507919e-05,
+      "loss": 1.0152,
+      "step": 3039
+    },
+    {
+      "epoch": 0.5212508305291811,
+      "grad_norm": 1.515625,
+      "learning_rate": 1.8623300280034424e-05,
+      "loss": 0.9943,
+      "step": 3040
+    },
+    {
+      "epoch": 0.5214222946181709,
+      "grad_norm": 1.6484375,
+      "learning_rate": 1.8622385713751363e-05,
+      "loss": 1.0114,
+      "step": 3041
+    },
+    {
+      "epoch": 0.5215937587071607,
+      "grad_norm": 1.640625,
+      "learning_rate": 1.8621470866259835e-05,
+      "loss": 1.1507,
+      "step": 3042
+    },
+    {
+      "epoch": 0.5217652227961507,
+      "grad_norm": 1.6171875,
+      "learning_rate": 1.862055573758968e-05,
+      "loss": 0.9616,
+      "step": 3043
+    },
+    {
+      "epoch": 0.5219366868851405,
+      "grad_norm": 1.5546875,
+      "learning_rate": 1.8619640327770735e-05,
+      "loss": 0.9324,
+      "step": 3044
+    },
+    {
+      "epoch": 0.5221081509741303,
+      "grad_norm": 1.625,
+      "learning_rate": 1.8618724636832866e-05,
+      "loss": 0.9974,
+      "step": 3045
+    },
+    {
+      "epoch": 0.5222796150631203,
+      "grad_norm": 1.609375,
+      "learning_rate": 1.8617808664805926e-05,
+      "loss": 1.0601,
+      "step": 3046
+    },
+    {
+      "epoch": 0.5224510791521101,
+      "grad_norm": 1.6640625,
+      "learning_rate": 1.8616892411719798e-05,
+      "loss": 0.9951,
+      "step": 3047
+    },
+    {
+      "epoch": 0.5226225432410999,
+      "grad_norm": 1.734375,
+      "learning_rate": 1.8615975877604356e-05,
+      "loss": 1.1189,
+      "step": 3048
+    },
+    {
+      "epoch": 0.5227940073300898,
+      "grad_norm": 1.5625,
+      "learning_rate": 1.86150590624895e-05,
+      "loss": 1.0502,
+      "step": 3049
+    },
+    {
+      "epoch": 0.5229654714190797,
+      "grad_norm": 1.515625,
+      "learning_rate": 1.8614141966405125e-05,
+      "loss": 1.0211,
+      "step": 3050
+    },
+    {
+      "epoch": 0.5231369355080695,
+      "grad_norm": 1.640625,
+      "learning_rate": 1.8613224589381143e-05,
+      "loss": 1.0333,
+      "step": 3051
+    },
+    {
+      "epoch": 0.5233083995970594,
+      "grad_norm": 1.5390625,
+      "learning_rate": 1.8612306931447473e-05,
+      "loss": 1.0176,
+      "step": 3052
+    },
+    {
+      "epoch": 0.5234798636860493,
+      "grad_norm": 1.578125,
+      "learning_rate": 1.861138899263404e-05,
+      "loss": 1.1212,
+      "step": 3053
+    },
+    {
+      "epoch": 0.5236513277750391,
+      "grad_norm": 1.5390625,
+      "learning_rate": 1.8610470772970787e-05,
+      "loss": 1.0371,
+      "step": 3054
+    },
+    {
+      "epoch": 0.523822791864029,
+      "grad_norm": 1.5625,
+      "learning_rate": 1.860955227248766e-05,
+      "loss": 1.0394,
+      "step": 3055
+    },
+    {
+      "epoch": 0.5239942559530189,
+      "grad_norm": 1.609375,
+      "learning_rate": 1.8608633491214613e-05,
+      "loss": 1.0387,
+      "step": 3056
+    },
+    {
+      "epoch": 0.5241657200420087,
+      "grad_norm": 1.6796875,
+      "learning_rate": 1.8607714429181606e-05,
+      "loss": 1.1197,
+      "step": 3057
+    },
+    {
+      "epoch": 0.5243371841309986,
+      "grad_norm": 1.65625,
+      "learning_rate": 1.860679508641862e-05,
+      "loss": 1.1022,
+      "step": 3058
+    },
+    {
+      "epoch": 0.5245086482199884,
+      "grad_norm": 1.6171875,
+      "learning_rate": 1.8605875462955638e-05,
+      "loss": 1.0624,
+      "step": 3059
+    },
+    {
+      "epoch": 0.5246801123089783,
+      "grad_norm": 1.546875,
+      "learning_rate": 1.8604955558822647e-05,
+      "loss": 1.0418,
+      "step": 3060
+    },
+    {
+      "epoch": 0.5248515763979682,
+      "grad_norm": 1.6171875,
+      "learning_rate": 1.8604035374049654e-05,
+      "loss": 1.0223,
+      "step": 3061
+    },
+    {
+      "epoch": 0.525023040486958,
+      "grad_norm": 1.6171875,
+      "learning_rate": 1.860311490866667e-05,
+      "loss": 1.0532,
+      "step": 3062
+    },
+    {
+      "epoch": 0.5251945045759479,
+      "grad_norm": 1.5234375,
+      "learning_rate": 1.860219416270371e-05,
+      "loss": 0.9409,
+      "step": 3063
+    },
+    {
+      "epoch": 0.5253659686649378,
+      "grad_norm": 1.6328125,
+      "learning_rate": 1.8601273136190806e-05,
+      "loss": 1.0213,
+      "step": 3064
+    },
+    {
+      "epoch": 0.5255374327539276,
+      "grad_norm": 1.546875,
+      "learning_rate": 1.8600351829157996e-05,
+      "loss": 1.0198,
+      "step": 3065
+    },
+    {
+      "epoch": 0.5257088968429174,
+      "grad_norm": 1.546875,
+      "learning_rate": 1.859943024163532e-05,
+      "loss": 1.0321,
+      "step": 3066
+    },
+    {
+      "epoch": 0.5258803609319074,
+      "grad_norm": 1.609375,
+      "learning_rate": 1.859850837365285e-05,
+      "loss": 1.0562,
+      "step": 3067
+    },
+    {
+      "epoch": 0.5260518250208972,
+      "grad_norm": 1.6015625,
+      "learning_rate": 1.859758622524064e-05,
+      "loss": 1.0236,
+      "step": 3068
+    },
+    {
+      "epoch": 0.526223289109887,
+      "grad_norm": 1.5703125,
+      "learning_rate": 1.8596663796428766e-05,
+      "loss": 1.0014,
+      "step": 3069
+    },
+    {
+      "epoch": 0.526394753198877,
+      "grad_norm": 1.5234375,
+      "learning_rate": 1.859574108724732e-05,
+      "loss": 1.032,
+      "step": 3070
+    },
+    {
+      "epoch": 0.5265662172878668,
+      "grad_norm": 1.65625,
+      "learning_rate": 1.8594818097726382e-05,
+      "loss": 1.0493,
+      "step": 3071
+    },
+    {
+      "epoch": 0.5267376813768566,
+      "grad_norm": 1.5859375,
+      "learning_rate": 1.8593894827896063e-05,
+      "loss": 1.0415,
+      "step": 3072
+    },
+    {
+      "epoch": 0.5269091454658466,
+      "grad_norm": 1.59375,
+      "learning_rate": 1.859297127778647e-05,
+      "loss": 0.9893,
+      "step": 3073
+    },
+    {
+      "epoch": 0.5270806095548364,
+      "grad_norm": 1.5859375,
+      "learning_rate": 1.859204744742773e-05,
+      "loss": 1.0089,
+      "step": 3074
+    },
+    {
+      "epoch": 0.5272520736438262,
+      "grad_norm": 1.625,
+      "learning_rate": 1.8591123336849962e-05,
+      "loss": 1.0668,
+      "step": 3075
+    },
+    {
+      "epoch": 0.527423537732816,
+      "grad_norm": 1.703125,
+      "learning_rate": 1.8590198946083315e-05,
+      "loss": 1.1142,
+      "step": 3076
+    },
+    {
+      "epoch": 0.527595001821806,
+      "grad_norm": 1.6171875,
+      "learning_rate": 1.8589274275157935e-05,
+      "loss": 1.0525,
+      "step": 3077
+    },
+    {
+      "epoch": 0.5277664659107958,
+      "grad_norm": 1.5859375,
+      "learning_rate": 1.8588349324103974e-05,
+      "loss": 1.0156,
+      "step": 3078
+    },
+    {
+      "epoch": 0.5279379299997856,
+      "grad_norm": 1.5078125,
+      "learning_rate": 1.85874240929516e-05,
+      "loss": 1.02,
+      "step": 3079
+    },
+    {
+      "epoch": 0.5281093940887756,
+      "grad_norm": 1.5390625,
+      "learning_rate": 1.8586498581730987e-05,
+      "loss": 1.0628,
+      "step": 3080
+    },
+    {
+      "epoch": 0.5282808581777654,
+      "grad_norm": 1.4765625,
+      "learning_rate": 1.8585572790472326e-05,
+      "loss": 1.092,
+      "step": 3081
+    },
+    {
+      "epoch": 0.5284523222667552,
+      "grad_norm": 1.609375,
+      "learning_rate": 1.8584646719205803e-05,
+      "loss": 1.0836,
+      "step": 3082
+    },
+    {
+      "epoch": 0.5286237863557451,
+      "grad_norm": 1.53125,
+      "learning_rate": 1.8583720367961623e-05,
+      "loss": 0.9917,
+      "step": 3083
+    },
+    {
+      "epoch": 0.528795250444735,
+      "grad_norm": 1.609375,
+      "learning_rate": 1.858279373677e-05,
+      "loss": 1.0639,
+      "step": 3084
+    },
+    {
+      "epoch": 0.5289667145337248,
+      "grad_norm": 1.71875,
+      "learning_rate": 1.8581866825661157e-05,
+      "loss": 1.0244,
+      "step": 3085
+    },
+    {
+      "epoch": 0.5291381786227147,
+      "grad_norm": 1.6171875,
+      "learning_rate": 1.8580939634665315e-05,
+      "loss": 1.0353,
+      "step": 3086
+    },
+    {
+      "epoch": 0.5293096427117046,
+      "grad_norm": 1.5390625,
+      "learning_rate": 1.8580012163812724e-05,
+      "loss": 0.9599,
+      "step": 3087
+    },
+    {
+      "epoch": 0.5294811068006944,
+      "grad_norm": 1.5546875,
+      "learning_rate": 1.857908441313362e-05,
+      "loss": 1.024,
+      "step": 3088
+    },
+    {
+      "epoch": 0.5296525708896843,
+      "grad_norm": 1.6484375,
+      "learning_rate": 1.8578156382658275e-05,
+      "loss": 1.0846,
+      "step": 3089
+    },
+    {
+      "epoch": 0.5298240349786741,
+      "grad_norm": 1.78125,
+      "learning_rate": 1.8577228072416945e-05,
+      "loss": 1.1081,
+      "step": 3090
+    },
+    {
+      "epoch": 0.529995499067664,
+      "grad_norm": 1.59375,
+      "learning_rate": 1.8576299482439905e-05,
+      "loss": 1.0505,
+      "step": 3091
+    },
+    {
+      "epoch": 0.5301669631566539,
+      "grad_norm": 1.59375,
+      "learning_rate": 1.8575370612757447e-05,
+      "loss": 1.0444,
+      "step": 3092
+    },
+    {
+      "epoch": 0.5303384272456437,
+      "grad_norm": 1.609375,
+      "learning_rate": 1.8574441463399862e-05,
+      "loss": 1.0852,
+      "step": 3093
+    },
+    {
+      "epoch": 0.5305098913346336,
+      "grad_norm": 1.6015625,
+      "learning_rate": 1.8573512034397453e-05,
+      "loss": 1.0687,
+      "step": 3094
+    },
+    {
+      "epoch": 0.5306813554236235,
+      "grad_norm": 1.6015625,
+      "learning_rate": 1.857258232578053e-05,
+      "loss": 0.9603,
+      "step": 3095
+    },
+    {
+      "epoch": 0.5308528195126133,
+      "grad_norm": 1.625,
+      "learning_rate": 1.8571652337579414e-05,
+      "loss": 0.9849,
+      "step": 3096
+    },
+    {
+      "epoch": 0.5310242836016031,
+      "grad_norm": 1.6015625,
+      "learning_rate": 1.857072206982444e-05,
+      "loss": 0.9789,
+      "step": 3097
+    },
+    {
+      "epoch": 0.5311957476905931,
+      "grad_norm": 1.65625,
+      "learning_rate": 1.8569791522545943e-05,
+      "loss": 1.0004,
+      "step": 3098
+    },
+    {
+      "epoch": 0.5313672117795829,
+      "grad_norm": 1.546875,
+      "learning_rate": 1.8568860695774278e-05,
+      "loss": 0.9379,
+      "step": 3099
+    },
+    {
+      "epoch": 0.5315386758685727,
+      "grad_norm": 1.6640625,
+      "learning_rate": 1.8567929589539795e-05,
+      "loss": 0.9668,
+      "step": 3100
+    },
+    {
+      "epoch": 0.5317101399575627,
+      "grad_norm": 1.59375,
+      "learning_rate": 1.8566998203872866e-05,
+      "loss": 0.9812,
+      "step": 3101
+    },
+    {
+      "epoch": 0.5318816040465525,
+      "grad_norm": 1.578125,
+      "learning_rate": 1.8566066538803863e-05,
+      "loss": 1.0347,
+      "step": 3102
+    },
+    {
+      "epoch": 0.5320530681355423,
+      "grad_norm": 1.59375,
+      "learning_rate": 1.8565134594363174e-05,
+      "loss": 1.0362,
+      "step": 3103
+    },
+    {
+      "epoch": 0.5322245322245323,
+      "grad_norm": 1.5859375,
+      "learning_rate": 1.8564202370581193e-05,
+      "loss": 0.993,
+      "step": 3104
+    },
+    {
+      "epoch": 0.5323959963135221,
+      "grad_norm": 1.609375,
+      "learning_rate": 1.8563269867488323e-05,
+      "loss": 1.0241,
+      "step": 3105
+    },
+    {
+      "epoch": 0.5325674604025119,
+      "grad_norm": 1.6953125,
+      "learning_rate": 1.8562337085114978e-05,
+      "loss": 1.0542,
+      "step": 3106
+    },
+    {
+      "epoch": 0.5327389244915018,
+      "grad_norm": 1.5,
+      "learning_rate": 1.8561404023491577e-05,
+      "loss": 0.9356,
+      "step": 3107
+    },
+    {
+      "epoch": 0.5329103885804917,
+      "grad_norm": 1.515625,
+      "learning_rate": 1.856047068264855e-05,
+      "loss": 0.9925,
+      "step": 3108
+    },
+    {
+      "epoch": 0.5330818526694815,
+      "grad_norm": 1.5390625,
+      "learning_rate": 1.8559537062616337e-05,
+      "loss": 0.9417,
+      "step": 3109
+    },
+    {
+      "epoch": 0.5332533167584714,
+      "grad_norm": 1.640625,
+      "learning_rate": 1.8558603163425392e-05,
+      "loss": 1.0574,
+      "step": 3110
+    },
+    {
+      "epoch": 0.5334247808474613,
+      "grad_norm": 1.6328125,
+      "learning_rate": 1.8557668985106164e-05,
+      "loss": 1.0717,
+      "step": 3111
+    },
+    {
+      "epoch": 0.5335962449364511,
+      "grad_norm": 1.640625,
+      "learning_rate": 1.855673452768913e-05,
+      "loss": 1.0645,
+      "step": 3112
+    },
+    {
+      "epoch": 0.533767709025441,
+      "grad_norm": 1.5859375,
+      "learning_rate": 1.8555799791204762e-05,
+      "loss": 0.9766,
+      "step": 3113
+    },
+    {
+      "epoch": 0.5339391731144308,
+      "grad_norm": 1.7734375,
+      "learning_rate": 1.855486477568354e-05,
+      "loss": 1.0897,
+      "step": 3114
+    },
+    {
+      "epoch": 0.5341106372034207,
+      "grad_norm": 1.625,
+      "learning_rate": 1.8553929481155966e-05,
+      "loss": 1.0915,
+      "step": 3115
+    },
+    {
+      "epoch": 0.5342821012924106,
+      "grad_norm": 1.421875,
+      "learning_rate": 1.8552993907652538e-05,
+      "loss": 0.907,
+      "step": 3116
+    },
+    {
+      "epoch": 0.5344535653814004,
+      "grad_norm": 1.6015625,
+      "learning_rate": 1.8552058055203773e-05,
+      "loss": 1.0162,
+      "step": 3117
+    },
+    {
+      "epoch": 0.5346250294703903,
+      "grad_norm": 1.5625,
+      "learning_rate": 1.855112192384019e-05,
+      "loss": 1.0393,
+      "step": 3118
+    },
+    {
+      "epoch": 0.5347964935593802,
+      "grad_norm": 1.546875,
+      "learning_rate": 1.8550185513592325e-05,
+      "loss": 0.9972,
+      "step": 3119
+    },
+    {
+      "epoch": 0.53496795764837,
+      "grad_norm": 1.5546875,
+      "learning_rate": 1.854924882449071e-05,
+      "loss": 1.0278,
+      "step": 3120
+    },
+    {
+      "epoch": 0.5351394217373598,
+      "grad_norm": 1.6171875,
+      "learning_rate": 1.8548311856565897e-05,
+      "loss": 0.98,
+      "step": 3121
+    },
+    {
+      "epoch": 0.5353108858263498,
+      "grad_norm": 1.609375,
+      "learning_rate": 1.8547374609848442e-05,
+      "loss": 1.089,
+      "step": 3122
+    },
+    {
+      "epoch": 0.5354823499153396,
+      "grad_norm": 1.671875,
+      "learning_rate": 1.8546437084368917e-05,
+      "loss": 1.009,
+      "step": 3123
+    },
+    {
+      "epoch": 0.5356538140043294,
+      "grad_norm": 1.609375,
+      "learning_rate": 1.8545499280157897e-05,
+      "loss": 0.9897,
+      "step": 3124
+    },
+    {
+      "epoch": 0.5358252780933194,
+      "grad_norm": 1.6171875,
+      "learning_rate": 1.8544561197245966e-05,
+      "loss": 0.9395,
+      "step": 3125
+    },
+    {
+      "epoch": 0.5359967421823092,
+      "grad_norm": 1.5859375,
+      "learning_rate": 1.8543622835663715e-05,
+      "loss": 1.027,
+      "step": 3126
+    },
+    {
+      "epoch": 0.536168206271299,
+      "grad_norm": 1.5859375,
+      "learning_rate": 1.8542684195441754e-05,
+      "loss": 1.1007,
+      "step": 3127
+    },
+    {
+      "epoch": 0.536339670360289,
+      "grad_norm": 1.6015625,
+      "learning_rate": 1.8541745276610693e-05,
+      "loss": 1.0017,
+      "step": 3128
+    },
+    {
+      "epoch": 0.5365111344492788,
+      "grad_norm": 1.453125,
+      "learning_rate": 1.8540806079201152e-05,
+      "loss": 1.008,
+      "step": 3129
+    },
+    {
+      "epoch": 0.5366825985382686,
+      "grad_norm": 1.5390625,
+      "learning_rate": 1.8539866603243762e-05,
+      "loss": 1.0767,
+      "step": 3130
+    },
+    {
+      "epoch": 0.5368540626272585,
+      "grad_norm": 1.6015625,
+      "learning_rate": 1.8538926848769166e-05,
+      "loss": 1.1019,
+      "step": 3131
+    },
+    {
+      "epoch": 0.5370255267162484,
+      "grad_norm": 1.65625,
+      "learning_rate": 1.853798681580801e-05,
+      "loss": 1.0994,
+      "step": 3132
+    },
+    {
+      "epoch": 0.5371969908052382,
+      "grad_norm": 1.59375,
+      "learning_rate": 1.853704650439095e-05,
+      "loss": 1.0356,
+      "step": 3133
+    },
+    {
+      "epoch": 0.5373684548942281,
+      "grad_norm": 1.6953125,
+      "learning_rate": 1.853610591454866e-05,
+      "loss": 1.0899,
+      "step": 3134
+    },
+    {
+      "epoch": 0.537539918983218,
+      "grad_norm": 1.65625,
+      "learning_rate": 1.8535165046311814e-05,
+      "loss": 1.0018,
+      "step": 3135
+    },
+    {
+      "epoch": 0.5377113830722078,
+      "grad_norm": 1.546875,
+      "learning_rate": 1.853422389971109e-05,
+      "loss": 1.0659,
+      "step": 3136
+    },
+    {
+      "epoch": 0.5378828471611977,
+      "grad_norm": 1.5390625,
+      "learning_rate": 1.853328247477719e-05,
+      "loss": 0.9889,
+      "step": 3137
+    },
+    {
+      "epoch": 0.5380543112501875,
+      "grad_norm": 1.6171875,
+      "learning_rate": 1.8532340771540815e-05,
+      "loss": 1.2005,
+      "step": 3138
+    },
+    {
+      "epoch": 0.5382257753391774,
+      "grad_norm": 1.5234375,
+      "learning_rate": 1.853139879003268e-05,
+      "loss": 0.9841,
+      "step": 3139
+    },
+    {
+      "epoch": 0.5383972394281673,
+      "grad_norm": 1.609375,
+      "learning_rate": 1.85304565302835e-05,
+      "loss": 0.9865,
+      "step": 3140
+    },
+    {
+      "epoch": 0.5385687035171571,
+      "grad_norm": 1.578125,
+      "learning_rate": 1.8529513992324012e-05,
+      "loss": 1.0674,
+      "step": 3141
+    },
+    {
+      "epoch": 0.538740167606147,
+      "grad_norm": 1.515625,
+      "learning_rate": 1.8528571176184952e-05,
+      "loss": 1.0455,
+      "step": 3142
+    },
+    {
+      "epoch": 0.5389116316951369,
+      "grad_norm": 1.59375,
+      "learning_rate": 1.8527628081897076e-05,
+      "loss": 1.1141,
+      "step": 3143
+    },
+    {
+      "epoch": 0.5390830957841267,
+      "grad_norm": 1.484375,
+      "learning_rate": 1.852668470949113e-05,
+      "loss": 0.9851,
+      "step": 3144
+    },
+    {
+      "epoch": 0.5392545598731165,
+      "grad_norm": 1.703125,
+      "learning_rate": 1.8525741058997888e-05,
+      "loss": 1.08,
+      "step": 3145
+    },
+    {
+      "epoch": 0.5394260239621065,
+      "grad_norm": 1.6171875,
+      "learning_rate": 1.8524797130448127e-05,
+      "loss": 0.9748,
+      "step": 3146
+    },
+    {
+      "epoch": 0.5395974880510963,
+      "grad_norm": 1.5703125,
+      "learning_rate": 1.8523852923872628e-05,
+      "loss": 0.9932,
+      "step": 3147
+    },
+    {
+      "epoch": 0.5397689521400861,
+      "grad_norm": 1.6171875,
+      "learning_rate": 1.8522908439302193e-05,
+      "loss": 1.0925,
+      "step": 3148
+    },
+    {
+      "epoch": 0.5399404162290761,
+      "grad_norm": 1.515625,
+      "learning_rate": 1.8521963676767614e-05,
+      "loss": 1.0554,
+      "step": 3149
+    },
+    {
+      "epoch": 0.5401118803180659,
+      "grad_norm": 1.6875,
+      "learning_rate": 1.852101863629971e-05,
+      "loss": 1.0495,
+      "step": 3150
+    },
+    {
+      "epoch": 0.5402833444070557,
+      "grad_norm": 1.484375,
+      "learning_rate": 1.85200733179293e-05,
+      "loss": 1.043,
+      "step": 3151
+    },
+    {
+      "epoch": 0.5404548084960457,
+      "grad_norm": 1.5078125,
+      "learning_rate": 1.8519127721687217e-05,
+      "loss": 1.0885,
+      "step": 3152
+    },
+    {
+      "epoch": 0.5406262725850355,
+      "grad_norm": 1.4765625,
+      "learning_rate": 1.8518181847604298e-05,
+      "loss": 0.8962,
+      "step": 3153
+    },
+    {
+      "epoch": 0.5407977366740253,
+      "grad_norm": 1.6484375,
+      "learning_rate": 1.851723569571139e-05,
+      "loss": 1.0868,
+      "step": 3154
+    },
+    {
+      "epoch": 0.5409692007630152,
+      "grad_norm": 1.53125,
+      "learning_rate": 1.8516289266039358e-05,
+      "loss": 1.0079,
+      "step": 3155
+    },
+    {
+      "epoch": 0.5411406648520051,
+      "grad_norm": 1.6640625,
+      "learning_rate": 1.851534255861906e-05,
+      "loss": 1.0173,
+      "step": 3156
+    },
+    {
+      "epoch": 0.5413121289409949,
+      "grad_norm": 1.5859375,
+      "learning_rate": 1.8514395573481377e-05,
+      "loss": 1.0323,
+      "step": 3157
+    },
+    {
+      "epoch": 0.5414835930299848,
+      "grad_norm": 1.5390625,
+      "learning_rate": 1.8513448310657193e-05,
+      "loss": 0.9832,
+      "step": 3158
+    },
+    {
+      "epoch": 0.5416550571189747,
+      "grad_norm": 1.59375,
+      "learning_rate": 1.85125007701774e-05,
+      "loss": 1.0065,
+      "step": 3159
+    },
+    {
+      "epoch": 0.5418265212079645,
+      "grad_norm": 1.5859375,
+      "learning_rate": 1.85115529520729e-05,
+      "loss": 1.1247,
+      "step": 3160
+    },
+    {
+      "epoch": 0.5419979852969544,
+      "grad_norm": 1.5625,
+      "learning_rate": 1.851060485637461e-05,
+      "loss": 1.0102,
+      "step": 3161
+    },
+    {
+      "epoch": 0.5421694493859442,
+      "grad_norm": 1.5859375,
+      "learning_rate": 1.8509656483113444e-05,
+      "loss": 1.0124,
+      "step": 3162
+    },
+    {
+      "epoch": 0.5423409134749341,
+      "grad_norm": 1.5234375,
+      "learning_rate": 1.8508707832320338e-05,
+      "loss": 1.0461,
+      "step": 3163
+    },
+    {
+      "epoch": 0.542512377563924,
+      "grad_norm": 1.5,
+      "learning_rate": 1.850775890402623e-05,
+      "loss": 1.027,
+      "step": 3164
+    },
+    {
+      "epoch": 0.5426838416529138,
+      "grad_norm": 1.515625,
+      "learning_rate": 1.8506809698262064e-05,
+      "loss": 0.899,
+      "step": 3165
+    },
+    {
+      "epoch": 0.5428553057419037,
+      "grad_norm": 1.5703125,
+      "learning_rate": 1.85058602150588e-05,
+      "loss": 0.9396,
+      "step": 3166
+    },
+    {
+      "epoch": 0.5430267698308936,
+      "grad_norm": 1.5859375,
+      "learning_rate": 1.8504910454447407e-05,
+      "loss": 1.0611,
+      "step": 3167
+    },
+    {
+      "epoch": 0.5431982339198834,
+      "grad_norm": 1.5390625,
+      "learning_rate": 1.8503960416458854e-05,
+      "loss": 1.0304,
+      "step": 3168
+    },
+    {
+      "epoch": 0.5433696980088732,
+      "grad_norm": 1.609375,
+      "learning_rate": 1.8503010101124132e-05,
+      "loss": 0.9509,
+      "step": 3169
+    },
+    {
+      "epoch": 0.5435411620978632,
+      "grad_norm": 1.53125,
+      "learning_rate": 1.8502059508474232e-05,
+      "loss": 0.9562,
+      "step": 3170
+    },
+    {
+      "epoch": 0.543712626186853,
+      "grad_norm": 1.5546875,
+      "learning_rate": 1.850110863854015e-05,
+      "loss": 0.9525,
+      "step": 3171
+    },
+    {
+      "epoch": 0.5438840902758428,
+      "grad_norm": 1.671875,
+      "learning_rate": 1.850015749135291e-05,
+      "loss": 0.9873,
+      "step": 3172
+    },
+    {
+      "epoch": 0.5440555543648327,
+      "grad_norm": 1.6015625,
+      "learning_rate": 1.849920606694352e-05,
+      "loss": 1.0424,
+      "step": 3173
+    },
+    {
+      "epoch": 0.5442270184538226,
+      "grad_norm": 1.6015625,
+      "learning_rate": 1.849825436534302e-05,
+      "loss": 1.0221,
+      "step": 3174
+    },
+    {
+      "epoch": 0.5443984825428124,
+      "grad_norm": 1.5859375,
+      "learning_rate": 1.849730238658244e-05,
+      "loss": 1.0602,
+      "step": 3175
+    },
+    {
+      "epoch": 0.5445699466318022,
+      "grad_norm": 1.625,
+      "learning_rate": 1.849635013069283e-05,
+      "loss": 1.0207,
+      "step": 3176
+    },
+    {
+      "epoch": 0.5447414107207922,
+      "grad_norm": 1.625,
+      "learning_rate": 1.8495397597705253e-05,
+      "loss": 1.1356,
+      "step": 3177
+    },
+    {
+      "epoch": 0.544912874809782,
+      "grad_norm": 1.578125,
+      "learning_rate": 1.8494444787650768e-05,
+      "loss": 0.9768,
+      "step": 3178
+    },
+    {
+      "epoch": 0.5450843388987718,
+      "grad_norm": 1.671875,
+      "learning_rate": 1.849349170056045e-05,
+      "loss": 1.0797,
+      "step": 3179
+    },
+    {
+      "epoch": 0.5452558029877618,
+      "grad_norm": 1.4765625,
+      "learning_rate": 1.8492538336465387e-05,
+      "loss": 0.9896,
+      "step": 3180
+    },
+    {
+      "epoch": 0.5454272670767516,
+      "grad_norm": 1.609375,
+      "learning_rate": 1.8491584695396666e-05,
+      "loss": 0.9773,
+      "step": 3181
+    },
+    {
+      "epoch": 0.5455987311657414,
+      "grad_norm": 1.5390625,
+      "learning_rate": 1.8490630777385393e-05,
+      "loss": 1.0454,
+      "step": 3182
+    },
+    {
+      "epoch": 0.5457701952547314,
+      "grad_norm": 1.640625,
+      "learning_rate": 1.8489676582462675e-05,
+      "loss": 1.0276,
+      "step": 3183
+    },
+    {
+      "epoch": 0.5459416593437212,
+      "grad_norm": 1.609375,
+      "learning_rate": 1.8488722110659635e-05,
+      "loss": 1.0258,
+      "step": 3184
+    },
+    {
+      "epoch": 0.546113123432711,
+      "grad_norm": 1.5078125,
+      "learning_rate": 1.8487767362007403e-05,
+      "loss": 0.9235,
+      "step": 3185
+    },
+    {
+      "epoch": 0.546284587521701,
+      "grad_norm": 1.578125,
+      "learning_rate": 1.8486812336537117e-05,
+      "loss": 1.0482,
+      "step": 3186
+    },
+    {
+      "epoch": 0.5464560516106908,
+      "grad_norm": 1.546875,
+      "learning_rate": 1.848585703427992e-05,
+      "loss": 0.9701,
+      "step": 3187
+    },
+    {
+      "epoch": 0.5466275156996806,
+      "grad_norm": 1.484375,
+      "learning_rate": 1.848490145526697e-05,
+      "loss": 0.9514,
+      "step": 3188
+    },
+    {
+      "epoch": 0.5467989797886705,
+      "grad_norm": 1.578125,
+      "learning_rate": 1.8483945599529436e-05,
+      "loss": 1.0126,
+      "step": 3189
+    },
+    {
+      "epoch": 0.5469704438776604,
+      "grad_norm": 1.59375,
+      "learning_rate": 1.8482989467098483e-05,
+      "loss": 1.0683,
+      "step": 3190
+    },
+    {
+      "epoch": 0.5471419079666502,
+      "grad_norm": 1.7109375,
+      "learning_rate": 1.8482033058005305e-05,
+      "loss": 1.0875,
+      "step": 3191
+    },
+    {
+      "epoch": 0.5473133720556401,
+      "grad_norm": 1.625,
+      "learning_rate": 1.8481076372281083e-05,
+      "loss": 0.9823,
+      "step": 3192
+    },
+    {
+      "epoch": 0.54748483614463,
+      "grad_norm": 1.6015625,
+      "learning_rate": 1.8480119409957026e-05,
+      "loss": 1.0217,
+      "step": 3193
+    },
+    {
+      "epoch": 0.5476563002336198,
+      "grad_norm": 1.640625,
+      "learning_rate": 1.8479162171064345e-05,
+      "loss": 1.0225,
+      "step": 3194
+    },
+    {
+      "epoch": 0.5478277643226097,
+      "grad_norm": 1.609375,
+      "learning_rate": 1.847820465563425e-05,
+      "loss": 1.0182,
+      "step": 3195
+    },
+    {
+      "epoch": 0.5479992284115995,
+      "grad_norm": 1.734375,
+      "learning_rate": 1.847724686369798e-05,
+      "loss": 1.0994,
+      "step": 3196
+    },
+    {
+      "epoch": 0.5481706925005894,
+      "grad_norm": 1.625,
+      "learning_rate": 1.8476288795286764e-05,
+      "loss": 1.125,
+      "step": 3197
+    },
+    {
+      "epoch": 0.5483421565895793,
+      "grad_norm": 1.5859375,
+      "learning_rate": 1.8475330450431858e-05,
+      "loss": 1.0885,
+      "step": 3198
+    },
+    {
+      "epoch": 0.5485136206785691,
+      "grad_norm": 1.7421875,
+      "learning_rate": 1.8474371829164504e-05,
+      "loss": 1.0607,
+      "step": 3199
+    },
+    {
+      "epoch": 0.548685084767559,
+      "grad_norm": 1.609375,
+      "learning_rate": 1.8473412931515974e-05,
+      "loss": 1.0336,
+      "step": 3200
+    },
+    {
+      "epoch": 0.5488565488565489,
+      "grad_norm": 1.609375,
+      "learning_rate": 1.8472453757517544e-05,
+      "loss": 1.0461,
+      "step": 3201
+    },
+    {
+      "epoch": 0.5490280129455387,
+      "grad_norm": 1.6484375,
+      "learning_rate": 1.847149430720049e-05,
+      "loss": 1.0338,
+      "step": 3202
+    },
+    {
+      "epoch": 0.5491994770345285,
+      "grad_norm": 1.6171875,
+      "learning_rate": 1.8470534580596106e-05,
+      "loss": 1.0947,
+      "step": 3203
+    },
+    {
+      "epoch": 0.5493709411235185,
+      "grad_norm": 1.6484375,
+      "learning_rate": 1.8469574577735694e-05,
+      "loss": 1.0113,
+      "step": 3204
+    },
+    {
+      "epoch": 0.5495424052125083,
+      "grad_norm": 1.71875,
+      "learning_rate": 1.8468614298650562e-05,
+      "loss": 1.0102,
+      "step": 3205
+    },
+    {
+      "epoch": 0.5497138693014981,
+      "grad_norm": 1.6328125,
+      "learning_rate": 1.8467653743372026e-05,
+      "loss": 1.0543,
+      "step": 3206
+    },
+    {
+      "epoch": 0.5498853333904881,
+      "grad_norm": 1.6484375,
+      "learning_rate": 1.8466692911931414e-05,
+      "loss": 1.0869,
+      "step": 3207
+    },
+    {
+      "epoch": 0.5500567974794779,
+      "grad_norm": 1.59375,
+      "learning_rate": 1.8465731804360064e-05,
+      "loss": 1.0763,
+      "step": 3208
+    },
+    {
+      "epoch": 0.5502282615684677,
+      "grad_norm": 1.5078125,
+      "learning_rate": 1.8464770420689323e-05,
+      "loss": 0.967,
+      "step": 3209
+    },
+    {
+      "epoch": 0.5503997256574577,
+      "grad_norm": 1.6015625,
+      "learning_rate": 1.8463808760950544e-05,
+      "loss": 0.9791,
+      "step": 3210
+    },
+    {
+      "epoch": 0.5505711897464475,
+      "grad_norm": 1.6015625,
+      "learning_rate": 1.8462846825175087e-05,
+      "loss": 0.9765,
+      "step": 3211
+    },
+    {
+      "epoch": 0.5507426538354373,
+      "grad_norm": 1.53125,
+      "learning_rate": 1.8461884613394328e-05,
+      "loss": 1.0778,
+      "step": 3212
+    },
+    {
+      "epoch": 0.5509141179244272,
+      "grad_norm": 1.5390625,
+      "learning_rate": 1.8460922125639643e-05,
+      "loss": 1.0572,
+      "step": 3213
+    },
+    {
+      "epoch": 0.5510855820134171,
+      "grad_norm": 1.6640625,
+      "learning_rate": 1.845995936194243e-05,
+      "loss": 1.0458,
+      "step": 3214
+    },
+    {
+      "epoch": 0.5512570461024069,
+      "grad_norm": 1.640625,
+      "learning_rate": 1.8458996322334085e-05,
+      "loss": 1.0645,
+      "step": 3215
+    },
+    {
+      "epoch": 0.5514285101913968,
+      "grad_norm": 1.5703125,
+      "learning_rate": 1.845803300684602e-05,
+      "loss": 0.9418,
+      "step": 3216
+    },
+    {
+      "epoch": 0.5515999742803867,
+      "grad_norm": 1.5625,
+      "learning_rate": 1.8457069415509642e-05,
+      "loss": 1.0402,
+      "step": 3217
+    },
+    {
+      "epoch": 0.5517714383693765,
+      "grad_norm": 1.828125,
+      "learning_rate": 1.8456105548356392e-05,
+      "loss": 0.9247,
+      "step": 3218
+    },
+    {
+      "epoch": 0.5519429024583664,
+      "grad_norm": 1.5625,
+      "learning_rate": 1.845514140541769e-05,
+      "loss": 1.0088,
+      "step": 3219
+    },
+    {
+      "epoch": 0.5521143665473562,
+      "grad_norm": 1.6171875,
+      "learning_rate": 1.845417698672499e-05,
+      "loss": 1.0644,
+      "step": 3220
+    },
+    {
+      "epoch": 0.5522858306363461,
+      "grad_norm": 1.6328125,
+      "learning_rate": 1.8453212292309743e-05,
+      "loss": 1.1305,
+      "step": 3221
+    },
+    {
+      "epoch": 0.552457294725336,
+      "grad_norm": 1.5625,
+      "learning_rate": 1.8452247322203413e-05,
+      "loss": 1.0003,
+      "step": 3222
+    },
+    {
+      "epoch": 0.5526287588143258,
+      "grad_norm": 1.671875,
+      "learning_rate": 1.8451282076437468e-05,
+      "loss": 0.9957,
+      "step": 3223
+    },
+    {
+      "epoch": 0.5528002229033157,
+      "grad_norm": 1.65625,
+      "learning_rate": 1.845031655504339e-05,
+      "loss": 1.0345,
+      "step": 3224
+    },
+    {
+      "epoch": 0.5529716869923056,
+      "grad_norm": 1.515625,
+      "learning_rate": 1.8449350758052668e-05,
+      "loss": 0.9799,
+      "step": 3225
+    },
+    {
+      "epoch": 0.5531431510812954,
+      "grad_norm": 1.5703125,
+      "learning_rate": 1.84483846854968e-05,
+      "loss": 1.0793,
+      "step": 3226
+    },
+    {
+      "epoch": 0.5533146151702852,
+      "grad_norm": 1.59375,
+      "learning_rate": 1.8447418337407295e-05,
+      "loss": 1.0526,
+      "step": 3227
+    },
+    {
+      "epoch": 0.5534860792592752,
+      "grad_norm": 1.7421875,
+      "learning_rate": 1.8446451713815667e-05,
+      "loss": 1.1667,
+      "step": 3228
+    },
+    {
+      "epoch": 0.553657543348265,
+      "grad_norm": 1.59375,
+      "learning_rate": 1.8445484814753444e-05,
+      "loss": 1.0139,
+      "step": 3229
+    },
+    {
+      "epoch": 0.5538290074372548,
+      "grad_norm": 1.578125,
+      "learning_rate": 1.8444517640252156e-05,
+      "loss": 1.0396,
+      "step": 3230
+    },
+    {
+      "epoch": 0.5540004715262448,
+      "grad_norm": 1.7421875,
+      "learning_rate": 1.844355019034335e-05,
+      "loss": 0.9346,
+      "step": 3231
+    },
+    {
+      "epoch": 0.5541719356152346,
+      "grad_norm": 1.578125,
+      "learning_rate": 1.8442582465058577e-05,
+      "loss": 0.9973,
+      "step": 3232
+    },
+    {
+      "epoch": 0.5543433997042244,
+      "grad_norm": 1.5859375,
+      "learning_rate": 1.8441614464429396e-05,
+      "loss": 1.0392,
+      "step": 3233
+    },
+    {
+      "epoch": 0.5545148637932144,
+      "grad_norm": 1.5234375,
+      "learning_rate": 1.8440646188487382e-05,
+      "loss": 0.9467,
+      "step": 3234
+    },
+    {
+      "epoch": 0.5546863278822042,
+      "grad_norm": 1.625,
+      "learning_rate": 1.843967763726411e-05,
+      "loss": 1.0302,
+      "step": 3235
+    },
+    {
+      "epoch": 0.554857791971194,
+      "grad_norm": 1.5234375,
+      "learning_rate": 1.8438708810791167e-05,
+      "loss": 1.0771,
+      "step": 3236
+    },
+    {
+      "epoch": 0.5550292560601839,
+      "grad_norm": 1.640625,
+      "learning_rate": 1.8437739709100155e-05,
+      "loss": 1.0512,
+      "step": 3237
+    },
+    {
+      "epoch": 0.5552007201491738,
+      "grad_norm": 1.546875,
+      "learning_rate": 1.8436770332222677e-05,
+      "loss": 1.0086,
+      "step": 3238
+    },
+    {
+      "epoch": 0.5553721842381636,
+      "grad_norm": 1.4921875,
+      "learning_rate": 1.8435800680190347e-05,
+      "loss": 0.9962,
+      "step": 3239
+    },
+    {
+      "epoch": 0.5555436483271535,
+      "grad_norm": 1.6015625,
+      "learning_rate": 1.8434830753034795e-05,
+      "loss": 0.9521,
+      "step": 3240
+    },
+    {
+      "epoch": 0.5557151124161434,
+      "grad_norm": 1.5859375,
+      "learning_rate": 1.8433860550787647e-05,
+      "loss": 0.9977,
+      "step": 3241
+    },
+    {
+      "epoch": 0.5558865765051332,
+      "grad_norm": 1.6796875,
+      "learning_rate": 1.8432890073480546e-05,
+      "loss": 1.011,
+      "step": 3242
+    },
+    {
+      "epoch": 0.5560580405941231,
+      "grad_norm": 1.5859375,
+      "learning_rate": 1.8431919321145146e-05,
+      "loss": 1.0422,
+      "step": 3243
+    },
+    {
+      "epoch": 0.5562295046831129,
+      "grad_norm": 1.4765625,
+      "learning_rate": 1.8430948293813105e-05,
+      "loss": 0.9491,
+      "step": 3244
+    },
+    {
+      "epoch": 0.5564009687721028,
+      "grad_norm": 1.65625,
+      "learning_rate": 1.8429976991516093e-05,
+      "loss": 1.0425,
+      "step": 3245
+    },
+    {
+      "epoch": 0.5565724328610927,
+      "grad_norm": 1.640625,
+      "learning_rate": 1.8429005414285783e-05,
+      "loss": 1.0996,
+      "step": 3246
+    },
+    {
+      "epoch": 0.5567438969500825,
+      "grad_norm": 1.59375,
+      "learning_rate": 1.842803356215387e-05,
+      "loss": 0.9526,
+      "step": 3247
+    },
+    {
+      "epoch": 0.5569153610390724,
+      "grad_norm": 1.6796875,
+      "learning_rate": 1.8427061435152045e-05,
+      "loss": 0.9868,
+      "step": 3248
+    },
+    {
+      "epoch": 0.5570868251280623,
+      "grad_norm": 1.546875,
+      "learning_rate": 1.8426089033312013e-05,
+      "loss": 0.9752,
+      "step": 3249
+    },
+    {
+      "epoch": 0.5572582892170521,
+      "grad_norm": 1.6328125,
+      "learning_rate": 1.8425116356665484e-05,
+      "loss": 1.0016,
+      "step": 3250
+    },
+    {
+      "epoch": 0.5574297533060419,
+      "grad_norm": 1.6328125,
+      "learning_rate": 1.842414340524419e-05,
+      "loss": 1.0322,
+      "step": 3251
+    },
+    {
+      "epoch": 0.5576012173950319,
+      "grad_norm": 1.6171875,
+      "learning_rate": 1.8423170179079856e-05,
+      "loss": 0.9072,
+      "step": 3252
+    },
+    {
+      "epoch": 0.5577726814840217,
+      "grad_norm": 1.6953125,
+      "learning_rate": 1.8422196678204224e-05,
+      "loss": 1.0137,
+      "step": 3253
+    },
+    {
+      "epoch": 0.5579441455730115,
+      "grad_norm": 1.5625,
+      "learning_rate": 1.8421222902649042e-05,
+      "loss": 1.0086,
+      "step": 3254
+    },
+    {
+      "epoch": 0.5581156096620015,
+      "grad_norm": 1.5703125,
+      "learning_rate": 1.842024885244607e-05,
+      "loss": 1.0747,
+      "step": 3255
+    },
+    {
+      "epoch": 0.5582870737509913,
+      "grad_norm": 1.640625,
+      "learning_rate": 1.8419274527627072e-05,
+      "loss": 0.9744,
+      "step": 3256
+    },
+    {
+      "epoch": 0.5584585378399811,
+      "grad_norm": 1.53125,
+      "learning_rate": 1.8418299928223834e-05,
+      "loss": 1.0025,
+      "step": 3257
+    },
+    {
+      "epoch": 0.558630001928971,
+      "grad_norm": 1.6328125,
+      "learning_rate": 1.841732505426813e-05,
+      "loss": 1.051,
+      "step": 3258
+    },
+    {
+      "epoch": 0.5588014660179609,
+      "grad_norm": 1.5703125,
+      "learning_rate": 1.841634990579176e-05,
+      "loss": 0.9937,
+      "step": 3259
+    },
+    {
+      "epoch": 0.5589729301069507,
+      "grad_norm": 1.546875,
+      "learning_rate": 1.8415374482826526e-05,
+      "loss": 0.9564,
+      "step": 3260
+    },
+    {
+      "epoch": 0.5591443941959406,
+      "grad_norm": 1.59375,
+      "learning_rate": 1.8414398785404245e-05,
+      "loss": 1.1329,
+      "step": 3261
+    },
+    {
+      "epoch": 0.5593158582849305,
+      "grad_norm": 1.6328125,
+      "learning_rate": 1.8413422813556727e-05,
+      "loss": 0.9782,
+      "step": 3262
+    },
+    {
+      "epoch": 0.5594873223739203,
+      "grad_norm": 1.6171875,
+      "learning_rate": 1.8412446567315816e-05,
+      "loss": 1.0287,
+      "step": 3263
+    },
+    {
+      "epoch": 0.5596587864629102,
+      "grad_norm": 1.625,
+      "learning_rate": 1.841147004671334e-05,
+      "loss": 1.0996,
+      "step": 3264
+    },
+    {
+      "epoch": 0.5598302505519,
+      "grad_norm": 1.6015625,
+      "learning_rate": 1.841049325178115e-05,
+      "loss": 1.0127,
+      "step": 3265
+    },
+    {
+      "epoch": 0.5600017146408899,
+      "grad_norm": 1.578125,
+      "learning_rate": 1.840951618255111e-05,
+      "loss": 0.9634,
+      "step": 3266
+    },
+    {
+      "epoch": 0.5601731787298797,
+      "grad_norm": 1.6171875,
+      "learning_rate": 1.8408538839055078e-05,
+      "loss": 1.0352,
+      "step": 3267
+    },
+    {
+      "epoch": 0.5603446428188696,
+      "grad_norm": 1.515625,
+      "learning_rate": 1.8407561221324925e-05,
+      "loss": 0.9594,
+      "step": 3268
+    },
+    {
+      "epoch": 0.5605161069078595,
+      "grad_norm": 1.609375,
+      "learning_rate": 1.8406583329392547e-05,
+      "loss": 1.0399,
+      "step": 3269
+    },
+    {
+      "epoch": 0.5606875709968493,
+      "grad_norm": 1.578125,
+      "learning_rate": 1.8405605163289828e-05,
+      "loss": 1.0761,
+      "step": 3270
+    },
+    {
+      "epoch": 0.5608590350858392,
+      "grad_norm": 1.6875,
+      "learning_rate": 1.840462672304867e-05,
+      "loss": 0.9835,
+      "step": 3271
+    },
+    {
+      "epoch": 0.5610304991748291,
+      "grad_norm": 1.546875,
+      "learning_rate": 1.840364800870099e-05,
+      "loss": 1.0436,
+      "step": 3272
+    },
+    {
+      "epoch": 0.5612019632638189,
+      "grad_norm": 1.6484375,
+      "learning_rate": 1.84026690202787e-05,
+      "loss": 1.0471,
+      "step": 3273
+    },
+    {
+      "epoch": 0.5613734273528088,
+      "grad_norm": 1.46875,
+      "learning_rate": 1.8401689757813734e-05,
+      "loss": 0.98,
+      "step": 3274
+    },
+    {
+      "epoch": 0.5615448914417986,
+      "grad_norm": 1.6328125,
+      "learning_rate": 1.8400710221338026e-05,
+      "loss": 1.0721,
+      "step": 3275
+    },
+    {
+      "epoch": 0.5617163555307885,
+      "grad_norm": 1.6484375,
+      "learning_rate": 1.8399730410883524e-05,
+      "loss": 1.0973,
+      "step": 3276
+    },
+    {
+      "epoch": 0.5618878196197784,
+      "grad_norm": 1.578125,
+      "learning_rate": 1.8398750326482182e-05,
+      "loss": 0.991,
+      "step": 3277
+    },
+    {
+      "epoch": 0.5620592837087682,
+      "grad_norm": 1.5546875,
+      "learning_rate": 1.8397769968165967e-05,
+      "loss": 1.0093,
+      "step": 3278
+    },
+    {
+      "epoch": 0.5622307477977581,
+      "grad_norm": 1.6171875,
+      "learning_rate": 1.8396789335966846e-05,
+      "loss": 1.0551,
+      "step": 3279
+    },
+    {
+      "epoch": 0.562402211886748,
+      "grad_norm": 1.65625,
+      "learning_rate": 1.839580842991681e-05,
+      "loss": 0.969,
+      "step": 3280
+    },
+    {
+      "epoch": 0.5625736759757378,
+      "grad_norm": 1.5546875,
+      "learning_rate": 1.8394827250047842e-05,
+      "loss": 0.929,
+      "step": 3281
+    },
+    {
+      "epoch": 0.5627451400647276,
+      "grad_norm": 1.625,
+      "learning_rate": 1.8393845796391947e-05,
+      "loss": 1.0078,
+      "step": 3282
+    },
+    {
+      "epoch": 0.5629166041537176,
+      "grad_norm": 1.5703125,
+      "learning_rate": 1.839286406898113e-05,
+      "loss": 1.012,
+      "step": 3283
+    },
+    {
+      "epoch": 0.5630880682427074,
+      "grad_norm": 1.59375,
+      "learning_rate": 1.8391882067847414e-05,
+      "loss": 0.9506,
+      "step": 3284
+    },
+    {
+      "epoch": 0.5632595323316972,
+      "grad_norm": 1.640625,
+      "learning_rate": 1.8390899793022824e-05,
+      "loss": 1.0795,
+      "step": 3285
+    },
+    {
+      "epoch": 0.5634309964206872,
+      "grad_norm": 1.6484375,
+      "learning_rate": 1.8389917244539392e-05,
+      "loss": 1.0464,
+      "step": 3286
+    },
+    {
+      "epoch": 0.563602460509677,
+      "grad_norm": 1.640625,
+      "learning_rate": 1.8388934422429167e-05,
+      "loss": 1.1251,
+      "step": 3287
+    },
+    {
+      "epoch": 0.5637739245986668,
+      "grad_norm": 1.5859375,
+      "learning_rate": 1.83879513267242e-05,
+      "loss": 0.9987,
+      "step": 3288
+    },
+    {
+      "epoch": 0.5639453886876568,
+      "grad_norm": 1.6484375,
+      "learning_rate": 1.8386967957456555e-05,
+      "loss": 1.0367,
+      "step": 3289
+    },
+    {
+      "epoch": 0.5641168527766466,
+      "grad_norm": 1.6015625,
+      "learning_rate": 1.83859843146583e-05,
+      "loss": 1.0065,
+      "step": 3290
+    },
+    {
+      "epoch": 0.5642883168656364,
+      "grad_norm": 1.546875,
+      "learning_rate": 1.8385000398361517e-05,
+      "loss": 0.9502,
+      "step": 3291
+    },
+    {
+      "epoch": 0.5644597809546263,
+      "grad_norm": 1.640625,
+      "learning_rate": 1.83840162085983e-05,
+      "loss": 0.9974,
+      "step": 3292
+    },
+    {
+      "epoch": 0.5646312450436162,
+      "grad_norm": 1.5703125,
+      "learning_rate": 1.838303174540074e-05,
+      "loss": 1.1014,
+      "step": 3293
+    },
+    {
+      "epoch": 0.564802709132606,
+      "grad_norm": 1.5703125,
+      "learning_rate": 1.8382047008800947e-05,
+      "loss": 0.9488,
+      "step": 3294
+    },
+    {
+      "epoch": 0.5649741732215959,
+      "grad_norm": 1.6015625,
+      "learning_rate": 1.838106199883104e-05,
+      "loss": 1.0501,
+      "step": 3295
+    },
+    {
+      "epoch": 0.5651456373105858,
+      "grad_norm": 1.5625,
+      "learning_rate": 1.838007671552314e-05,
+      "loss": 0.9358,
+      "step": 3296
+    },
+    {
+      "epoch": 0.5653171013995756,
+      "grad_norm": 1.5859375,
+      "learning_rate": 1.837909115890938e-05,
+      "loss": 1.0654,
+      "step": 3297
+    },
+    {
+      "epoch": 0.5654885654885655,
+      "grad_norm": 1.5,
+      "learning_rate": 1.8378105329021908e-05,
+      "loss": 0.9435,
+      "step": 3298
+    },
+    {
+      "epoch": 0.5656600295775553,
+      "grad_norm": 1.65625,
+      "learning_rate": 1.8377119225892868e-05,
+      "loss": 0.9904,
+      "step": 3299
+    },
+    {
+      "epoch": 0.5658314936665452,
+      "grad_norm": 1.6015625,
+      "learning_rate": 1.8376132849554428e-05,
+      "loss": 1.0634,
+      "step": 3300
+    },
+    {
+      "epoch": 0.5660029577555351,
+      "grad_norm": 1.6015625,
+      "learning_rate": 1.8375146200038752e-05,
+      "loss": 1.0184,
+      "step": 3301
+    },
+    {
+      "epoch": 0.5661744218445249,
+      "grad_norm": 1.65625,
+      "learning_rate": 1.8374159277378024e-05,
+      "loss": 1.0845,
+      "step": 3302
+    },
+    {
+      "epoch": 0.5663458859335148,
+      "grad_norm": 1.578125,
+      "learning_rate": 1.8373172081604424e-05,
+      "loss": 0.9841,
+      "step": 3303
+    },
+    {
+      "epoch": 0.5665173500225047,
+      "grad_norm": 1.5859375,
+      "learning_rate": 1.8372184612750152e-05,
+      "loss": 1.0651,
+      "step": 3304
+    },
+    {
+      "epoch": 0.5666888141114945,
+      "grad_norm": 1.7265625,
+      "learning_rate": 1.8371196870847418e-05,
+      "loss": 1.0452,
+      "step": 3305
+    },
+    {
+      "epoch": 0.5668602782004843,
+      "grad_norm": 1.59375,
+      "learning_rate": 1.8370208855928427e-05,
+      "loss": 1.0319,
+      "step": 3306
+    },
+    {
+      "epoch": 0.5670317422894743,
+      "grad_norm": 1.6953125,
+      "learning_rate": 1.8369220568025405e-05,
+      "loss": 1.0709,
+      "step": 3307
+    },
+    {
+      "epoch": 0.5672032063784641,
+      "grad_norm": 1.6015625,
+      "learning_rate": 1.8368232007170587e-05,
+      "loss": 1.0725,
+      "step": 3308
+    },
+    {
+      "epoch": 0.5673746704674539,
+      "grad_norm": 1.5390625,
+      "learning_rate": 1.836724317339621e-05,
+      "loss": 1.0185,
+      "step": 3309
+    },
+    {
+      "epoch": 0.5675461345564439,
+      "grad_norm": 1.640625,
+      "learning_rate": 1.8366254066734526e-05,
+      "loss": 1.073,
+      "step": 3310
+    },
+    {
+      "epoch": 0.5677175986454337,
+      "grad_norm": 1.625,
+      "learning_rate": 1.836526468721779e-05,
+      "loss": 1.0662,
+      "step": 3311
+    },
+    {
+      "epoch": 0.5678890627344235,
+      "grad_norm": 1.5078125,
+      "learning_rate": 1.8364275034878277e-05,
+      "loss": 1.0078,
+      "step": 3312
+    },
+    {
+      "epoch": 0.5680605268234135,
+      "grad_norm": 1.5390625,
+      "learning_rate": 1.8363285109748253e-05,
+      "loss": 1.0112,
+      "step": 3313
+    },
+    {
+      "epoch": 0.5682319909124033,
+      "grad_norm": 1.6171875,
+      "learning_rate": 1.836229491186001e-05,
+      "loss": 1.0086,
+      "step": 3314
+    },
+    {
+      "epoch": 0.5684034550013931,
+      "grad_norm": 1.5625,
+      "learning_rate": 1.8361304441245843e-05,
+      "loss": 1.0831,
+      "step": 3315
+    },
+    {
+      "epoch": 0.568574919090383,
+      "grad_norm": 1.59375,
+      "learning_rate": 1.836031369793805e-05,
+      "loss": 1.0802,
+      "step": 3316
+    },
+    {
+      "epoch": 0.5687463831793729,
+      "grad_norm": 1.53125,
+      "learning_rate": 1.8359322681968942e-05,
+      "loss": 1.0011,
+      "step": 3317
+    },
+    {
+      "epoch": 0.5689178472683627,
+      "grad_norm": 1.5234375,
+      "learning_rate": 1.8358331393370847e-05,
+      "loss": 1.0521,
+      "step": 3318
+    },
+    {
+      "epoch": 0.5690893113573526,
+      "grad_norm": 1.6484375,
+      "learning_rate": 1.8357339832176092e-05,
+      "loss": 1.0235,
+      "step": 3319
+    },
+    {
+      "epoch": 0.5692607754463425,
+      "grad_norm": 1.5859375,
+      "learning_rate": 1.835634799841701e-05,
+      "loss": 1.0332,
+      "step": 3320
+    },
+    {
+      "epoch": 0.5694322395353323,
+      "grad_norm": 1.578125,
+      "learning_rate": 1.835535589212596e-05,
+      "loss": 1.0653,
+      "step": 3321
+    },
+    {
+      "epoch": 0.5696037036243222,
+      "grad_norm": 1.6484375,
+      "learning_rate": 1.8354363513335282e-05,
+      "loss": 1.1728,
+      "step": 3322
+    },
+    {
+      "epoch": 0.569775167713312,
+      "grad_norm": 1.53125,
+      "learning_rate": 1.8353370862077358e-05,
+      "loss": 1.0377,
+      "step": 3323
+    },
+    {
+      "epoch": 0.5699466318023019,
+      "grad_norm": 1.65625,
+      "learning_rate": 1.835237793838455e-05,
+      "loss": 1.0604,
+      "step": 3324
+    },
+    {
+      "epoch": 0.5701180958912918,
+      "grad_norm": 1.609375,
+      "learning_rate": 1.8351384742289244e-05,
+      "loss": 1.0777,
+      "step": 3325
+    },
+    {
+      "epoch": 0.5702895599802816,
+      "grad_norm": 1.53125,
+      "learning_rate": 1.8350391273823836e-05,
+      "loss": 1.0222,
+      "step": 3326
+    },
+    {
+      "epoch": 0.5704610240692715,
+      "grad_norm": 1.6015625,
+      "learning_rate": 1.8349397533020723e-05,
+      "loss": 1.0333,
+      "step": 3327
+    },
+    {
+      "epoch": 0.5706324881582614,
+      "grad_norm": 1.578125,
+      "learning_rate": 1.8348403519912316e-05,
+      "loss": 1.0752,
+      "step": 3328
+    },
+    {
+      "epoch": 0.5708039522472512,
+      "grad_norm": 1.5625,
+      "learning_rate": 1.8347409234531034e-05,
+      "loss": 1.0549,
+      "step": 3329
+    },
+    {
+      "epoch": 0.570975416336241,
+      "grad_norm": 1.6015625,
+      "learning_rate": 1.83464146769093e-05,
+      "loss": 1.0061,
+      "step": 3330
+    },
+    {
+      "epoch": 0.571146880425231,
+      "grad_norm": 1.6953125,
+      "learning_rate": 1.8345419847079555e-05,
+      "loss": 0.9759,
+      "step": 3331
+    },
+    {
+      "epoch": 0.5713183445142208,
+      "grad_norm": 1.5390625,
+      "learning_rate": 1.8344424745074245e-05,
+      "loss": 1.0317,
+      "step": 3332
+    },
+    {
+      "epoch": 0.5714898086032106,
+      "grad_norm": 1.6015625,
+      "learning_rate": 1.834342937092582e-05,
+      "loss": 1.0236,
+      "step": 3333
+    },
+    {
+      "epoch": 0.5716612726922006,
+      "grad_norm": 1.5859375,
+      "learning_rate": 1.8342433724666745e-05,
+      "loss": 1.032,
+      "step": 3334
+    },
+    {
+      "epoch": 0.5718327367811904,
+      "grad_norm": 1.5703125,
+      "learning_rate": 1.8341437806329492e-05,
+      "loss": 1.0583,
+      "step": 3335
+    },
+    {
+      "epoch": 0.5720042008701802,
+      "grad_norm": 1.5703125,
+      "learning_rate": 1.8340441615946544e-05,
+      "loss": 1.0017,
+      "step": 3336
+    },
+    {
+      "epoch": 0.5721756649591702,
+      "grad_norm": 1.6875,
+      "learning_rate": 1.8339445153550382e-05,
+      "loss": 1.1914,
+      "step": 3337
+    },
+    {
+      "epoch": 0.57234712904816,
+      "grad_norm": 1.65625,
+      "learning_rate": 1.8338448419173514e-05,
+      "loss": 1.0421,
+      "step": 3338
+    },
+    {
+      "epoch": 0.5725185931371498,
+      "grad_norm": 1.578125,
+      "learning_rate": 1.833745141284844e-05,
+      "loss": 1.0538,
+      "step": 3339
+    },
+    {
+      "epoch": 0.5726900572261397,
+      "grad_norm": 1.5546875,
+      "learning_rate": 1.833645413460768e-05,
+      "loss": 1.019,
+      "step": 3340
+    },
+    {
+      "epoch": 0.5728615213151296,
+      "grad_norm": 1.6796875,
+      "learning_rate": 1.8335456584483763e-05,
+      "loss": 1.0864,
+      "step": 3341
+    },
+    {
+      "epoch": 0.5730329854041194,
+      "grad_norm": 1.5234375,
+      "learning_rate": 1.8334458762509214e-05,
+      "loss": 0.988,
+      "step": 3342
+    },
+    {
+      "epoch": 0.5732044494931093,
+      "grad_norm": 1.65625,
+      "learning_rate": 1.833346066871658e-05,
+      "loss": 1.0151,
+      "step": 3343
+    },
+    {
+      "epoch": 0.5733759135820992,
+      "grad_norm": 1.6953125,
+      "learning_rate": 1.8332462303138418e-05,
+      "loss": 1.0402,
+      "step": 3344
+    },
+    {
+      "epoch": 0.573547377671089,
+      "grad_norm": 1.9375,
+      "learning_rate": 1.8331463665807278e-05,
+      "loss": 1.0646,
+      "step": 3345
+    },
+    {
+      "epoch": 0.5737188417600789,
+      "grad_norm": 1.5390625,
+      "learning_rate": 1.8330464756755737e-05,
+      "loss": 0.9698,
+      "step": 3346
+    },
+    {
+      "epoch": 0.5738903058490687,
+      "grad_norm": 1.484375,
+      "learning_rate": 1.832946557601637e-05,
+      "loss": 0.9967,
+      "step": 3347
+    },
+    {
+      "epoch": 0.5740617699380586,
+      "grad_norm": 1.6015625,
+      "learning_rate": 1.832846612362176e-05,
+      "loss": 0.9096,
+      "step": 3348
+    },
+    {
+      "epoch": 0.5742332340270485,
+      "grad_norm": 1.5546875,
+      "learning_rate": 1.8327466399604517e-05,
+      "loss": 1.0488,
+      "step": 3349
+    },
+    {
+      "epoch": 0.5744046981160383,
+      "grad_norm": 1.578125,
+      "learning_rate": 1.8326466403997228e-05,
+      "loss": 1.0233,
+      "step": 3350
+    },
+    {
+      "epoch": 0.5745761622050282,
+      "grad_norm": 1.65625,
+      "learning_rate": 1.832546613683252e-05,
+      "loss": 0.9991,
+      "step": 3351
+    },
+    {
+      "epoch": 0.5747476262940181,
+      "grad_norm": 1.6640625,
+      "learning_rate": 1.832446559814301e-05,
+      "loss": 1.12,
+      "step": 3352
+    },
+    {
+      "epoch": 0.5749190903830079,
+      "grad_norm": 1.7109375,
+      "learning_rate": 1.8323464787961327e-05,
+      "loss": 1.0985,
+      "step": 3353
+    },
+    {
+      "epoch": 0.5750905544719978,
+      "grad_norm": 1.4609375,
+      "learning_rate": 1.8322463706320116e-05,
+      "loss": 0.9235,
+      "step": 3354
+    },
+    {
+      "epoch": 0.5752620185609877,
+      "grad_norm": 1.5078125,
+      "learning_rate": 1.8321462353252024e-05,
+      "loss": 0.9686,
+      "step": 3355
+    },
+    {
+      "epoch": 0.5754334826499775,
+      "grad_norm": 1.609375,
+      "learning_rate": 1.8320460728789705e-05,
+      "loss": 0.991,
+      "step": 3356
+    },
+    {
+      "epoch": 0.5756049467389673,
+      "grad_norm": 1.625,
+      "learning_rate": 1.8319458832965836e-05,
+      "loss": 1.0094,
+      "step": 3357
+    },
+    {
+      "epoch": 0.5757764108279573,
+      "grad_norm": 1.625,
+      "learning_rate": 1.831845666581308e-05,
+      "loss": 1.0868,
+      "step": 3358
+    },
+    {
+      "epoch": 0.5759478749169471,
+      "grad_norm": 1.5703125,
+      "learning_rate": 1.8317454227364134e-05,
+      "loss": 0.9983,
+      "step": 3359
+    },
+    {
+      "epoch": 0.5761193390059369,
+      "grad_norm": 1.5703125,
+      "learning_rate": 1.8316451517651683e-05,
+      "loss": 1.0399,
+      "step": 3360
+    },
+    {
+      "epoch": 0.5762908030949268,
+      "grad_norm": 1.484375,
+      "learning_rate": 1.831544853670843e-05,
+      "loss": 0.9766,
+      "step": 3361
+    },
+    {
+      "epoch": 0.5764622671839167,
+      "grad_norm": 1.5859375,
+      "learning_rate": 1.8314445284567087e-05,
+      "loss": 0.9866,
+      "step": 3362
+    },
+    {
+      "epoch": 0.5766337312729065,
+      "grad_norm": 1.5703125,
+      "learning_rate": 1.8313441761260374e-05,
+      "loss": 1.0635,
+      "step": 3363
+    },
+    {
+      "epoch": 0.5768051953618963,
+      "grad_norm": 1.6015625,
+      "learning_rate": 1.831243796682102e-05,
+      "loss": 0.9684,
+      "step": 3364
+    },
+    {
+      "epoch": 0.5769766594508863,
+      "grad_norm": 1.5625,
+      "learning_rate": 1.831143390128176e-05,
+      "loss": 1.0406,
+      "step": 3365
+    },
+    {
+      "epoch": 0.5771481235398761,
+      "grad_norm": 1.6015625,
+      "learning_rate": 1.8310429564675347e-05,
+      "loss": 0.943,
+      "step": 3366
+    },
+    {
+      "epoch": 0.5773195876288659,
+      "grad_norm": 1.625,
+      "learning_rate": 1.8309424957034528e-05,
+      "loss": 1.0568,
+      "step": 3367
+    },
+    {
+      "epoch": 0.5774910517178559,
+      "grad_norm": 1.5390625,
+      "learning_rate": 1.8308420078392073e-05,
+      "loss": 1.0687,
+      "step": 3368
+    },
+    {
+      "epoch": 0.5776625158068457,
+      "grad_norm": 1.71875,
+      "learning_rate": 1.8307414928780753e-05,
+      "loss": 1.098,
+      "step": 3369
+    },
+    {
+      "epoch": 0.5778339798958355,
+      "grad_norm": 1.7109375,
+      "learning_rate": 1.8306409508233347e-05,
+      "loss": 1.1567,
+      "step": 3370
+    },
+    {
+      "epoch": 0.5780054439848255,
+      "grad_norm": 1.59375,
+      "learning_rate": 1.830540381678265e-05,
+      "loss": 1.0479,
+      "step": 3371
+    },
+    {
+      "epoch": 0.5781769080738153,
+      "grad_norm": 1.5546875,
+      "learning_rate": 1.830439785446146e-05,
+      "loss": 1.0058,
+      "step": 3372
+    },
+    {
+      "epoch": 0.5783483721628051,
+      "grad_norm": 1.65625,
+      "learning_rate": 1.830339162130258e-05,
+      "loss": 1.0528,
+      "step": 3373
+    },
+    {
+      "epoch": 0.578519836251795,
+      "grad_norm": 1.5859375,
+      "learning_rate": 1.8302385117338836e-05,
+      "loss": 1.0247,
+      "step": 3374
+    },
+    {
+      "epoch": 0.5786913003407849,
+      "grad_norm": 1.46875,
+      "learning_rate": 1.8301378342603045e-05,
+      "loss": 1.0035,
+      "step": 3375
+    },
+    {
+      "epoch": 0.5788627644297747,
+      "grad_norm": 1.5390625,
+      "learning_rate": 1.830037129712805e-05,
+      "loss": 1.0463,
+      "step": 3376
+    },
+    {
+      "epoch": 0.5790342285187646,
+      "grad_norm": 1.5859375,
+      "learning_rate": 1.829936398094669e-05,
+      "loss": 1.0622,
+      "step": 3377
+    },
+    {
+      "epoch": 0.5792056926077545,
+      "grad_norm": 1.59375,
+      "learning_rate": 1.829835639409182e-05,
+      "loss": 1.0306,
+      "step": 3378
+    },
+    {
+      "epoch": 0.5793771566967443,
+      "grad_norm": 1.578125,
+      "learning_rate": 1.8297348536596297e-05,
+      "loss": 1.0617,
+      "step": 3379
+    },
+    {
+      "epoch": 0.5795486207857342,
+      "grad_norm": 1.5390625,
+      "learning_rate": 1.829634040849299e-05,
+      "loss": 1.0848,
+      "step": 3380
+    },
+    {
+      "epoch": 0.579720084874724,
+      "grad_norm": 1.5546875,
+      "learning_rate": 1.829533200981479e-05,
+      "loss": 0.9907,
+      "step": 3381
+    },
+    {
+      "epoch": 0.5798915489637139,
+      "grad_norm": 1.578125,
+      "learning_rate": 1.8294323340594565e-05,
+      "loss": 1.0193,
+      "step": 3382
+    },
+    {
+      "epoch": 0.5800630130527038,
+      "grad_norm": 1.5390625,
+      "learning_rate": 1.829331440086523e-05,
+      "loss": 0.9932,
+      "step": 3383
+    },
+    {
+      "epoch": 0.5802344771416936,
+      "grad_norm": 1.5390625,
+      "learning_rate": 1.829230519065968e-05,
+      "loss": 0.9571,
+      "step": 3384
+    },
+    {
+      "epoch": 0.5804059412306835,
+      "grad_norm": 1.4296875,
+      "learning_rate": 1.829129571001083e-05,
+      "loss": 0.969,
+      "step": 3385
+    },
+    {
+      "epoch": 0.5805774053196734,
+      "grad_norm": 1.5,
+      "learning_rate": 1.829028595895161e-05,
+      "loss": 0.963,
+      "step": 3386
+    },
+    {
+      "epoch": 0.5807488694086632,
+      "grad_norm": 1.5390625,
+      "learning_rate": 1.828927593751494e-05,
+      "loss": 0.9628,
+      "step": 3387
+    },
+    {
+      "epoch": 0.580920333497653,
+      "grad_norm": 1.5234375,
+      "learning_rate": 1.828826564573377e-05,
+      "loss": 0.9359,
+      "step": 3388
+    },
+    {
+      "epoch": 0.581091797586643,
+      "grad_norm": 1.546875,
+      "learning_rate": 1.8287255083641048e-05,
+      "loss": 1.0194,
+      "step": 3389
+    },
+    {
+      "epoch": 0.5812632616756328,
+      "grad_norm": 1.515625,
+      "learning_rate": 1.828624425126973e-05,
+      "loss": 1.0237,
+      "step": 3390
+    },
+    {
+      "epoch": 0.5814347257646226,
+      "grad_norm": 1.6328125,
+      "learning_rate": 1.8285233148652784e-05,
+      "loss": 1.0385,
+      "step": 3391
+    },
+    {
+      "epoch": 0.5816061898536126,
+      "grad_norm": 1.5078125,
+      "learning_rate": 1.8284221775823187e-05,
+      "loss": 1.0283,
+      "step": 3392
+    },
+    {
+      "epoch": 0.5817776539426024,
+      "grad_norm": 1.5859375,
+      "learning_rate": 1.828321013281392e-05,
+      "loss": 1.0483,
+      "step": 3393
+    },
+    {
+      "epoch": 0.5819491180315922,
+      "grad_norm": 1.609375,
+      "learning_rate": 1.828219821965798e-05,
+      "loss": 1.0926,
+      "step": 3394
+    },
+    {
+      "epoch": 0.5821205821205822,
+      "grad_norm": 1.609375,
+      "learning_rate": 1.8281186036388368e-05,
+      "loss": 1.0863,
+      "step": 3395
+    },
+    {
+      "epoch": 0.582292046209572,
+      "grad_norm": 1.5390625,
+      "learning_rate": 1.8280173583038095e-05,
+      "loss": 1.0118,
+      "step": 3396
+    },
+    {
+      "epoch": 0.5824635102985618,
+      "grad_norm": 1.5859375,
+      "learning_rate": 1.8279160859640182e-05,
+      "loss": 0.978,
+      "step": 3397
+    },
+    {
+      "epoch": 0.5826349743875517,
+      "grad_norm": 1.4609375,
+      "learning_rate": 1.8278147866227658e-05,
+      "loss": 1.029,
+      "step": 3398
+    },
+    {
+      "epoch": 0.5828064384765416,
+      "grad_norm": 1.53125,
+      "learning_rate": 1.8277134602833556e-05,
+      "loss": 1.0615,
+      "step": 3399
+    },
+    {
+      "epoch": 0.5829779025655314,
+      "grad_norm": 1.515625,
+      "learning_rate": 1.8276121069490925e-05,
+      "loss": 1.0119,
+      "step": 3400
+    },
+    {
+      "epoch": 0.5831493666545213,
+      "grad_norm": 1.6171875,
+      "learning_rate": 1.8275107266232826e-05,
+      "loss": 1.0311,
+      "step": 3401
+    },
+    {
+      "epoch": 0.5833208307435112,
+      "grad_norm": 1.578125,
+      "learning_rate": 1.8274093193092313e-05,
+      "loss": 1.0166,
+      "step": 3402
+    },
+    {
+      "epoch": 0.583492294832501,
+      "grad_norm": 1.6015625,
+      "learning_rate": 1.827307885010247e-05,
+      "loss": 1.0905,
+      "step": 3403
+    },
+    {
+      "epoch": 0.5836637589214909,
+      "grad_norm": 1.75,
+      "learning_rate": 1.8272064237296367e-05,
+      "loss": 1.0299,
+      "step": 3404
+    },
+    {
+      "epoch": 0.5838352230104807,
+      "grad_norm": 1.546875,
+      "learning_rate": 1.82710493547071e-05,
+      "loss": 0.9962,
+      "step": 3405
+    },
+    {
+      "epoch": 0.5840066870994706,
+      "grad_norm": 1.625,
+      "learning_rate": 1.827003420236777e-05,
+      "loss": 0.9978,
+      "step": 3406
+    },
+    {
+      "epoch": 0.5841781511884605,
+      "grad_norm": 1.6171875,
+      "learning_rate": 1.826901878031148e-05,
+      "loss": 1.0804,
+      "step": 3407
+    },
+    {
+      "epoch": 0.5843496152774503,
+      "grad_norm": 1.6171875,
+      "learning_rate": 1.8268003088571352e-05,
+      "loss": 1.0453,
+      "step": 3408
+    },
+    {
+      "epoch": 0.5845210793664402,
+      "grad_norm": 1.5,
+      "learning_rate": 1.8266987127180507e-05,
+      "loss": 0.9565,
+      "step": 3409
+    },
+    {
+      "epoch": 0.5846925434554301,
+      "grad_norm": 1.671875,
+      "learning_rate": 1.8265970896172082e-05,
+      "loss": 1.0409,
+      "step": 3410
+    },
+    {
+      "epoch": 0.5848640075444199,
+      "grad_norm": 1.65625,
+      "learning_rate": 1.8264954395579216e-05,
+      "loss": 1.0311,
+      "step": 3411
+    },
+    {
+      "epoch": 0.5850354716334097,
+      "grad_norm": 1.7265625,
+      "learning_rate": 1.826393762543507e-05,
+      "loss": 1.0295,
+      "step": 3412
+    },
+    {
+      "epoch": 0.5852069357223997,
+      "grad_norm": 1.546875,
+      "learning_rate": 1.8262920585772793e-05,
+      "loss": 0.9599,
+      "step": 3413
+    },
+    {
+      "epoch": 0.5853783998113895,
+      "grad_norm": 1.5703125,
+      "learning_rate": 1.8261903276625562e-05,
+      "loss": 0.9765,
+      "step": 3414
+    },
+    {
+      "epoch": 0.5855498639003793,
+      "grad_norm": 1.546875,
+      "learning_rate": 1.826088569802656e-05,
+      "loss": 1.0553,
+      "step": 3415
+    },
+    {
+      "epoch": 0.5857213279893693,
+      "grad_norm": 1.6328125,
+      "learning_rate": 1.825986785000896e-05,
+      "loss": 0.9898,
+      "step": 3416
+    },
+    {
+      "epoch": 0.5858927920783591,
+      "grad_norm": 1.640625,
+      "learning_rate": 1.825884973260597e-05,
+      "loss": 1.1863,
+      "step": 3417
+    },
+    {
+      "epoch": 0.5860642561673489,
+      "grad_norm": 1.5625,
+      "learning_rate": 1.8257831345850787e-05,
+      "loss": 0.9711,
+      "step": 3418
+    },
+    {
+      "epoch": 0.5862357202563389,
+      "grad_norm": 1.5859375,
+      "learning_rate": 1.825681268977663e-05,
+      "loss": 1.0413,
+      "step": 3419
+    },
+    {
+      "epoch": 0.5864071843453287,
+      "grad_norm": 1.5546875,
+      "learning_rate": 1.8255793764416718e-05,
+      "loss": 1.0048,
+      "step": 3420
+    },
+    {
+      "epoch": 0.5865786484343185,
+      "grad_norm": 1.5625,
+      "learning_rate": 1.8254774569804284e-05,
+      "loss": 1.0406,
+      "step": 3421
+    },
+    {
+      "epoch": 0.5867501125233084,
+      "grad_norm": 1.609375,
+      "learning_rate": 1.8253755105972565e-05,
+      "loss": 1.0598,
+      "step": 3422
+    },
+    {
+      "epoch": 0.5869215766122983,
+      "grad_norm": 1.6640625,
+      "learning_rate": 1.8252735372954812e-05,
+      "loss": 1.0493,
+      "step": 3423
+    },
+    {
+      "epoch": 0.5870930407012881,
+      "grad_norm": 1.71875,
+      "learning_rate": 1.8251715370784283e-05,
+      "loss": 1.0204,
+      "step": 3424
+    },
+    {
+      "epoch": 0.587264504790278,
+      "grad_norm": 1.5390625,
+      "learning_rate": 1.825069509949424e-05,
+      "loss": 1.0498,
+      "step": 3425
+    },
+    {
+      "epoch": 0.5874359688792679,
+      "grad_norm": 1.6875,
+      "learning_rate": 1.8249674559117963e-05,
+      "loss": 1.0432,
+      "step": 3426
+    },
+    {
+      "epoch": 0.5876074329682577,
+      "grad_norm": 1.5546875,
+      "learning_rate": 1.824865374968873e-05,
+      "loss": 1.0371,
+      "step": 3427
+    },
+    {
+      "epoch": 0.5877788970572476,
+      "grad_norm": 1.5859375,
+      "learning_rate": 1.824763267123984e-05,
+      "loss": 1.0738,
+      "step": 3428
+    },
+    {
+      "epoch": 0.5879503611462374,
+      "grad_norm": 1.4921875,
+      "learning_rate": 1.824661132380459e-05,
+      "loss": 0.9195,
+      "step": 3429
+    },
+    {
+      "epoch": 0.5881218252352273,
+      "grad_norm": 1.5703125,
+      "learning_rate": 1.824558970741629e-05,
+      "loss": 1.043,
+      "step": 3430
+    },
+    {
+      "epoch": 0.5882932893242172,
+      "grad_norm": 1.546875,
+      "learning_rate": 1.8244567822108262e-05,
+      "loss": 0.9982,
+      "step": 3431
+    },
+    {
+      "epoch": 0.588464753413207,
+      "grad_norm": 1.546875,
+      "learning_rate": 1.824354566791383e-05,
+      "loss": 1.0171,
+      "step": 3432
+    },
+    {
+      "epoch": 0.5886362175021969,
+      "grad_norm": 1.5703125,
+      "learning_rate": 1.824252324486633e-05,
+      "loss": 0.9613,
+      "step": 3433
+    },
+    {
+      "epoch": 0.5888076815911868,
+      "grad_norm": 1.65625,
+      "learning_rate": 1.824150055299911e-05,
+      "loss": 1.1034,
+      "step": 3434
+    },
+    {
+      "epoch": 0.5889791456801766,
+      "grad_norm": 1.5546875,
+      "learning_rate": 1.8240477592345522e-05,
+      "loss": 0.9503,
+      "step": 3435
+    },
+    {
+      "epoch": 0.5891506097691664,
+      "grad_norm": 1.5703125,
+      "learning_rate": 1.8239454362938925e-05,
+      "loss": 1.0519,
+      "step": 3436
+    },
+    {
+      "epoch": 0.5893220738581564,
+      "grad_norm": 1.59375,
+      "learning_rate": 1.82384308648127e-05,
+      "loss": 1.0554,
+      "step": 3437
+    },
+    {
+      "epoch": 0.5894935379471462,
+      "grad_norm": 1.578125,
+      "learning_rate": 1.823740709800022e-05,
+      "loss": 0.9954,
+      "step": 3438
+    },
+    {
+      "epoch": 0.589665002036136,
+      "grad_norm": 1.8359375,
+      "learning_rate": 1.8236383062534874e-05,
+      "loss": 0.961,
+      "step": 3439
+    },
+    {
+      "epoch": 0.589836466125126,
+      "grad_norm": 1.7265625,
+      "learning_rate": 1.8235358758450062e-05,
+      "loss": 1.0856,
+      "step": 3440
+    },
+    {
+      "epoch": 0.5900079302141158,
+      "grad_norm": 1.5546875,
+      "learning_rate": 1.8234334185779186e-05,
+      "loss": 1.0408,
+      "step": 3441
+    },
+    {
+      "epoch": 0.5901793943031056,
+      "grad_norm": 1.5859375,
+      "learning_rate": 1.823330934455567e-05,
+      "loss": 1.0096,
+      "step": 3442
+    },
+    {
+      "epoch": 0.5903508583920956,
+      "grad_norm": 1.640625,
+      "learning_rate": 1.823228423481293e-05,
+      "loss": 1.0386,
+      "step": 3443
+    },
+    {
+      "epoch": 0.5905223224810854,
+      "grad_norm": 1.6171875,
+      "learning_rate": 1.82312588565844e-05,
+      "loss": 1.0921,
+      "step": 3444
+    },
+    {
+      "epoch": 0.5906937865700752,
+      "grad_norm": 1.6484375,
+      "learning_rate": 1.8230233209903527e-05,
+      "loss": 1.108,
+      "step": 3445
+    },
+    {
+      "epoch": 0.5908652506590651,
+      "grad_norm": 1.5859375,
+      "learning_rate": 1.8229207294803753e-05,
+      "loss": 0.9812,
+      "step": 3446
+    },
+    {
+      "epoch": 0.591036714748055,
+      "grad_norm": 1.5390625,
+      "learning_rate": 1.8228181111318538e-05,
+      "loss": 1.0402,
+      "step": 3447
+    },
+    {
+      "epoch": 0.5912081788370448,
+      "grad_norm": 1.578125,
+      "learning_rate": 1.822715465948136e-05,
+      "loss": 0.9975,
+      "step": 3448
+    },
+    {
+      "epoch": 0.5913796429260347,
+      "grad_norm": 1.5859375,
+      "learning_rate": 1.8226127939325683e-05,
+      "loss": 1.0768,
+      "step": 3449
+    },
+    {
+      "epoch": 0.5915511070150246,
+      "grad_norm": 1.5,
+      "learning_rate": 1.8225100950884997e-05,
+      "loss": 1.0326,
+      "step": 3450
+    },
+    {
+      "epoch": 0.5917225711040144,
+      "grad_norm": 1.5625,
+      "learning_rate": 1.8224073694192798e-05,
+      "loss": 1.0055,
+      "step": 3451
+    },
+    {
+      "epoch": 0.5918940351930043,
+      "grad_norm": 1.71875,
+      "learning_rate": 1.8223046169282585e-05,
+      "loss": 1.0598,
+      "step": 3452
+    },
+    {
+      "epoch": 0.5920654992819941,
+      "grad_norm": 1.6171875,
+      "learning_rate": 1.8222018376187873e-05,
+      "loss": 1.0051,
+      "step": 3453
+    },
+    {
+      "epoch": 0.592236963370984,
+      "grad_norm": 1.4765625,
+      "learning_rate": 1.822099031494218e-05,
+      "loss": 0.937,
+      "step": 3454
+    },
+    {
+      "epoch": 0.5924084274599738,
+      "grad_norm": 1.515625,
+      "learning_rate": 1.8219961985579035e-05,
+      "loss": 0.9945,
+      "step": 3455
+    },
+    {
+      "epoch": 0.5925798915489637,
+      "grad_norm": 1.6328125,
+      "learning_rate": 1.821893338813198e-05,
+      "loss": 1.1318,
+      "step": 3456
+    },
+    {
+      "epoch": 0.5927513556379536,
+      "grad_norm": 1.6171875,
+      "learning_rate": 1.8217904522634553e-05,
+      "loss": 1.0298,
+      "step": 3457
+    },
+    {
+      "epoch": 0.5929228197269434,
+      "grad_norm": 1.5546875,
+      "learning_rate": 1.8216875389120316e-05,
+      "loss": 0.9916,
+      "step": 3458
+    },
+    {
+      "epoch": 0.5930942838159333,
+      "grad_norm": 1.5390625,
+      "learning_rate": 1.8215845987622828e-05,
+      "loss": 1.0565,
+      "step": 3459
+    },
+    {
+      "epoch": 0.5932657479049231,
+      "grad_norm": 1.546875,
+      "learning_rate": 1.8214816318175663e-05,
+      "loss": 0.9459,
+      "step": 3460
+    },
+    {
+      "epoch": 0.593437211993913,
+      "grad_norm": 1.6015625,
+      "learning_rate": 1.821378638081241e-05,
+      "loss": 1.0692,
+      "step": 3461
+    },
+    {
+      "epoch": 0.5936086760829029,
+      "grad_norm": 1.6015625,
+      "learning_rate": 1.8212756175566646e-05,
+      "loss": 1.0501,
+      "step": 3462
+    },
+    {
+      "epoch": 0.5937801401718927,
+      "grad_norm": 1.734375,
+      "learning_rate": 1.8211725702471984e-05,
+      "loss": 1.0388,
+      "step": 3463
+    },
+    {
+      "epoch": 0.5939516042608826,
+      "grad_norm": 1.5859375,
+      "learning_rate": 1.8210694961562015e-05,
+      "loss": 1.0025,
+      "step": 3464
+    },
+    {
+      "epoch": 0.5941230683498725,
+      "grad_norm": 1.6796875,
+      "learning_rate": 1.8209663952870373e-05,
+      "loss": 1.046,
+      "step": 3465
+    },
+    {
+      "epoch": 0.5942945324388623,
+      "grad_norm": 1.59375,
+      "learning_rate": 1.8208632676430674e-05,
+      "loss": 1.0592,
+      "step": 3466
+    },
+    {
+      "epoch": 0.5944659965278521,
+      "grad_norm": 1.515625,
+      "learning_rate": 1.820760113227655e-05,
+      "loss": 0.9688,
+      "step": 3467
+    },
+    {
+      "epoch": 0.5946374606168421,
+      "grad_norm": 1.65625,
+      "learning_rate": 1.8206569320441645e-05,
+      "loss": 0.9846,
+      "step": 3468
+    },
+    {
+      "epoch": 0.5948089247058319,
+      "grad_norm": 1.59375,
+      "learning_rate": 1.8205537240959615e-05,
+      "loss": 0.9957,
+      "step": 3469
+    },
+    {
+      "epoch": 0.5949803887948217,
+      "grad_norm": 1.640625,
+      "learning_rate": 1.820450489386411e-05,
+      "loss": 1.0043,
+      "step": 3470
+    },
+    {
+      "epoch": 0.5951518528838117,
+      "grad_norm": 1.5859375,
+      "learning_rate": 1.820347227918881e-05,
+      "loss": 1.0168,
+      "step": 3471
+    },
+    {
+      "epoch": 0.5953233169728015,
+      "grad_norm": 1.6484375,
+      "learning_rate": 1.820243939696739e-05,
+      "loss": 1.0769,
+      "step": 3472
+    },
+    {
+      "epoch": 0.5954947810617913,
+      "grad_norm": 1.5546875,
+      "learning_rate": 1.820140624723353e-05,
+      "loss": 1.0118,
+      "step": 3473
+    },
+    {
+      "epoch": 0.5956662451507813,
+      "grad_norm": 1.578125,
+      "learning_rate": 1.820037283002093e-05,
+      "loss": 1.0421,
+      "step": 3474
+    },
+    {
+      "epoch": 0.5958377092397711,
+      "grad_norm": 1.6015625,
+      "learning_rate": 1.8199339145363292e-05,
+      "loss": 1.061,
+      "step": 3475
+    },
+    {
+      "epoch": 0.5960091733287609,
+      "grad_norm": 1.625,
+      "learning_rate": 1.819830519329433e-05,
+      "loss": 1.0801,
+      "step": 3476
+    },
+    {
+      "epoch": 0.5961806374177508,
+      "grad_norm": 1.671875,
+      "learning_rate": 1.819727097384776e-05,
+      "loss": 1.0314,
+      "step": 3477
+    },
+    {
+      "epoch": 0.5963521015067407,
+      "grad_norm": 1.625,
+      "learning_rate": 1.819623648705732e-05,
+      "loss": 0.9982,
+      "step": 3478
+    },
+    {
+      "epoch": 0.5965235655957305,
+      "grad_norm": 1.515625,
+      "learning_rate": 1.819520173295674e-05,
+      "loss": 1.0165,
+      "step": 3479
+    },
+    {
+      "epoch": 0.5966950296847204,
+      "grad_norm": 1.6484375,
+      "learning_rate": 1.819416671157977e-05,
+      "loss": 1.0337,
+      "step": 3480
+    },
+    {
+      "epoch": 0.5968664937737103,
+      "grad_norm": 1.5703125,
+      "learning_rate": 1.819313142296017e-05,
+      "loss": 1.0332,
+      "step": 3481
+    },
+    {
+      "epoch": 0.5970379578627001,
+      "grad_norm": 1.625,
+      "learning_rate": 1.8192095867131705e-05,
+      "loss": 1.0734,
+      "step": 3482
+    },
+    {
+      "epoch": 0.59720942195169,
+      "grad_norm": 1.5859375,
+      "learning_rate": 1.819106004412814e-05,
+      "loss": 1.1012,
+      "step": 3483
+    },
+    {
+      "epoch": 0.5973808860406798,
+      "grad_norm": 1.625,
+      "learning_rate": 1.8190023953983264e-05,
+      "loss": 0.9893,
+      "step": 3484
+    },
+    {
+      "epoch": 0.5975523501296697,
+      "grad_norm": 1.6171875,
+      "learning_rate": 1.818898759673087e-05,
+      "loss": 1.0762,
+      "step": 3485
+    },
+    {
+      "epoch": 0.5977238142186596,
+      "grad_norm": 1.640625,
+      "learning_rate": 1.8187950972404746e-05,
+      "loss": 1.093,
+      "step": 3486
+    },
+    {
+      "epoch": 0.5978952783076494,
+      "grad_norm": 1.59375,
+      "learning_rate": 1.818691408103871e-05,
+      "loss": 0.9958,
+      "step": 3487
+    },
+    {
+      "epoch": 0.5980667423966393,
+      "grad_norm": 1.5859375,
+      "learning_rate": 1.818587692266658e-05,
+      "loss": 1.0037,
+      "step": 3488
+    },
+    {
+      "epoch": 0.5982382064856292,
+      "grad_norm": 1.625,
+      "learning_rate": 1.818483949732218e-05,
+      "loss": 1.0488,
+      "step": 3489
+    },
+    {
+      "epoch": 0.598409670574619,
+      "grad_norm": 1.5703125,
+      "learning_rate": 1.8183801805039337e-05,
+      "loss": 1.027,
+      "step": 3490
+    },
+    {
+      "epoch": 0.5985811346636088,
+      "grad_norm": 1.734375,
+      "learning_rate": 1.8182763845851902e-05,
+      "loss": 1.053,
+      "step": 3491
+    },
+    {
+      "epoch": 0.5987525987525988,
+      "grad_norm": 1.7109375,
+      "learning_rate": 1.8181725619793727e-05,
+      "loss": 1.0962,
+      "step": 3492
+    },
+    {
+      "epoch": 0.5989240628415886,
+      "grad_norm": 1.5390625,
+      "learning_rate": 1.8180687126898666e-05,
+      "loss": 0.9581,
+      "step": 3493
+    },
+    {
+      "epoch": 0.5990955269305784,
+      "grad_norm": 1.5703125,
+      "learning_rate": 1.8179648367200596e-05,
+      "loss": 0.9726,
+      "step": 3494
+    },
+    {
+      "epoch": 0.5992669910195684,
+      "grad_norm": 1.515625,
+      "learning_rate": 1.817860934073339e-05,
+      "loss": 1.0337,
+      "step": 3495
+    },
+    {
+      "epoch": 0.5994384551085582,
+      "grad_norm": 1.59375,
+      "learning_rate": 1.8177570047530933e-05,
+      "loss": 1.0834,
+      "step": 3496
+    },
+    {
+      "epoch": 0.599609919197548,
+      "grad_norm": 1.546875,
+      "learning_rate": 1.8176530487627123e-05,
+      "loss": 0.9813,
+      "step": 3497
+    },
+    {
+      "epoch": 0.599781383286538,
+      "grad_norm": 1.5390625,
+      "learning_rate": 1.8175490661055864e-05,
+      "loss": 0.9932,
+      "step": 3498
+    },
+    {
+      "epoch": 0.5999528473755278,
+      "grad_norm": 1.609375,
+      "learning_rate": 1.817445056785107e-05,
+      "loss": 1.056,
+      "step": 3499
+    },
+    {
+      "epoch": 0.6001243114645176,
+      "grad_norm": 1.5703125,
+      "learning_rate": 1.8173410208046666e-05,
+      "loss": 1.021,
+      "step": 3500
+    },
+    {
+      "epoch": 0.6001243114645176,
+      "eval_loss": 0.8751522302627563,
+      "eval_runtime": 837.1447,
+      "eval_samples_per_second": 2.985,
+      "eval_steps_per_second": 2.985,
+      "step": 3500
+    },
+    {
+      "epoch": 0.6002957755535075,
+      "grad_norm": 1.6875,
+      "learning_rate": 1.8172369581676567e-05,
+      "loss": 1.0605,
+      "step": 3501
+    },
+    {
+      "epoch": 0.6004672396424974,
+      "grad_norm": 1.5625,
+      "learning_rate": 1.817132868877473e-05,
+      "loss": 0.9984,
+      "step": 3502
+    },
+    {
+      "epoch": 0.6006387037314872,
+      "grad_norm": 1.5390625,
+      "learning_rate": 1.817028752937509e-05,
+      "loss": 0.9996,
+      "step": 3503
+    },
+    {
+      "epoch": 0.6008101678204771,
+      "grad_norm": 1.578125,
+      "learning_rate": 1.8169246103511605e-05,
+      "loss": 1.0918,
+      "step": 3504
+    },
+    {
+      "epoch": 0.600981631909467,
+      "grad_norm": 1.6171875,
+      "learning_rate": 1.8168204411218247e-05,
+      "loss": 1.115,
+      "step": 3505
+    },
+    {
+      "epoch": 0.6011530959984568,
+      "grad_norm": 1.625,
+      "learning_rate": 1.8167162452528982e-05,
+      "loss": 0.9894,
+      "step": 3506
+    },
+    {
+      "epoch": 0.6013245600874467,
+      "grad_norm": 1.6015625,
+      "learning_rate": 1.8166120227477796e-05,
+      "loss": 1.0847,
+      "step": 3507
+    },
+    {
+      "epoch": 0.6014960241764366,
+      "grad_norm": 1.515625,
+      "learning_rate": 1.816507773609868e-05,
+      "loss": 0.9875,
+      "step": 3508
+    },
+    {
+      "epoch": 0.6016674882654264,
+      "grad_norm": 1.6875,
+      "learning_rate": 1.8164034978425627e-05,
+      "loss": 1.0064,
+      "step": 3509
+    },
+    {
+      "epoch": 0.6018389523544163,
+      "grad_norm": 1.734375,
+      "learning_rate": 1.8162991954492654e-05,
+      "loss": 1.1553,
+      "step": 3510
+    },
+    {
+      "epoch": 0.6020104164434061,
+      "grad_norm": 1.515625,
+      "learning_rate": 1.8161948664333773e-05,
+      "loss": 0.9663,
+      "step": 3511
+    },
+    {
+      "epoch": 0.602181880532396,
+      "grad_norm": 1.5625,
+      "learning_rate": 1.816090510798301e-05,
+      "loss": 0.9988,
+      "step": 3512
+    },
+    {
+      "epoch": 0.6023533446213859,
+      "grad_norm": 1.5234375,
+      "learning_rate": 1.81598612854744e-05,
+      "loss": 0.9419,
+      "step": 3513
+    },
+    {
+      "epoch": 0.6025248087103757,
+      "grad_norm": 1.578125,
+      "learning_rate": 1.8158817196841993e-05,
+      "loss": 1.1279,
+      "step": 3514
+    },
+    {
+      "epoch": 0.6026962727993656,
+      "grad_norm": 1.59375,
+      "learning_rate": 1.815777284211983e-05,
+      "loss": 1.0023,
+      "step": 3515
+    },
+    {
+      "epoch": 0.6028677368883555,
+      "grad_norm": 1.6015625,
+      "learning_rate": 1.815672822134197e-05,
+      "loss": 1.1439,
+      "step": 3516
+    },
+    {
+      "epoch": 0.6030392009773453,
+      "grad_norm": 1.6953125,
+      "learning_rate": 1.8155683334542493e-05,
+      "loss": 1.0355,
+      "step": 3517
+    },
+    {
+      "epoch": 0.6032106650663351,
+      "grad_norm": 1.4921875,
+      "learning_rate": 1.8154638181755467e-05,
+      "loss": 1.0313,
+      "step": 3518
+    },
+    {
+      "epoch": 0.6033821291553251,
+      "grad_norm": 1.734375,
+      "learning_rate": 1.8153592763014987e-05,
+      "loss": 1.1347,
+      "step": 3519
+    },
+    {
+      "epoch": 0.6035535932443149,
+      "grad_norm": 1.4921875,
+      "learning_rate": 1.8152547078355143e-05,
+      "loss": 0.9597,
+      "step": 3520
+    },
+    {
+      "epoch": 0.6037250573333047,
+      "grad_norm": 1.546875,
+      "learning_rate": 1.8151501127810038e-05,
+      "loss": 1.0531,
+      "step": 3521
+    },
+    {
+      "epoch": 0.6038965214222947,
+      "grad_norm": 1.5078125,
+      "learning_rate": 1.8150454911413783e-05,
+      "loss": 0.9631,
+      "step": 3522
+    },
+    {
+      "epoch": 0.6040679855112845,
+      "grad_norm": 1.5390625,
+      "learning_rate": 1.8149408429200503e-05,
+      "loss": 1.0816,
+      "step": 3523
+    },
+    {
+      "epoch": 0.6042394496002743,
+      "grad_norm": 1.6015625,
+      "learning_rate": 1.814836168120433e-05,
+      "loss": 0.9617,
+      "step": 3524
+    },
+    {
+      "epoch": 0.6044109136892643,
+      "grad_norm": 1.53125,
+      "learning_rate": 1.8147314667459394e-05,
+      "loss": 0.9863,
+      "step": 3525
+    },
+    {
+      "epoch": 0.6045823777782541,
+      "grad_norm": 1.5859375,
+      "learning_rate": 1.814626738799985e-05,
+      "loss": 1.0777,
+      "step": 3526
+    },
+    {
+      "epoch": 0.6047538418672439,
+      "grad_norm": 1.5703125,
+      "learning_rate": 1.814521984285985e-05,
+      "loss": 1.0113,
+      "step": 3527
+    },
+    {
+      "epoch": 0.6049253059562338,
+      "grad_norm": 1.5859375,
+      "learning_rate": 1.8144172032073558e-05,
+      "loss": 0.9739,
+      "step": 3528
+    },
+    {
+      "epoch": 0.6050967700452237,
+      "grad_norm": 1.6015625,
+      "learning_rate": 1.8143123955675147e-05,
+      "loss": 1.0028,
+      "step": 3529
+    },
+    {
+      "epoch": 0.6052682341342135,
+      "grad_norm": 1.671875,
+      "learning_rate": 1.8142075613698798e-05,
+      "loss": 1.0616,
+      "step": 3530
+    },
+    {
+      "epoch": 0.6054396982232034,
+      "grad_norm": 1.578125,
+      "learning_rate": 1.8141027006178707e-05,
+      "loss": 1.0612,
+      "step": 3531
+    },
+    {
+      "epoch": 0.6056111623121933,
+      "grad_norm": 1.46875,
+      "learning_rate": 1.813997813314907e-05,
+      "loss": 0.9675,
+      "step": 3532
+    },
+    {
+      "epoch": 0.6057826264011831,
+      "grad_norm": 1.5546875,
+      "learning_rate": 1.8138928994644088e-05,
+      "loss": 1.0159,
+      "step": 3533
+    },
+    {
+      "epoch": 0.605954090490173,
+      "grad_norm": 1.6015625,
+      "learning_rate": 1.8137879590697986e-05,
+      "loss": 1.0209,
+      "step": 3534
+    },
+    {
+      "epoch": 0.6061255545791628,
+      "grad_norm": 1.53125,
+      "learning_rate": 1.8136829921344986e-05,
+      "loss": 1.0564,
+      "step": 3535
+    },
+    {
+      "epoch": 0.6062970186681527,
+      "grad_norm": 1.6484375,
+      "learning_rate": 1.8135779986619324e-05,
+      "loss": 1.0058,
+      "step": 3536
+    },
+    {
+      "epoch": 0.6064684827571426,
+      "grad_norm": 1.578125,
+      "learning_rate": 1.813472978655524e-05,
+      "loss": 1.0321,
+      "step": 3537
+    },
+    {
+      "epoch": 0.6066399468461324,
+      "grad_norm": 1.59375,
+      "learning_rate": 1.813367932118698e-05,
+      "loss": 1.0533,
+      "step": 3538
+    },
+    {
+      "epoch": 0.6068114109351223,
+      "grad_norm": 1.59375,
+      "learning_rate": 1.8132628590548813e-05,
+      "loss": 0.9243,
+      "step": 3539
+    },
+    {
+      "epoch": 0.6069828750241122,
+      "grad_norm": 1.609375,
+      "learning_rate": 1.8131577594675003e-05,
+      "loss": 1.053,
+      "step": 3540
+    },
+    {
+      "epoch": 0.607154339113102,
+      "grad_norm": 1.6484375,
+      "learning_rate": 1.8130526333599826e-05,
+      "loss": 1.0014,
+      "step": 3541
+    },
+    {
+      "epoch": 0.6073258032020918,
+      "grad_norm": 1.8359375,
+      "learning_rate": 1.8129474807357567e-05,
+      "loss": 0.9082,
+      "step": 3542
+    },
+    {
+      "epoch": 0.6074972672910818,
+      "grad_norm": 1.546875,
+      "learning_rate": 1.8128423015982526e-05,
+      "loss": 1.0552,
+      "step": 3543
+    },
+    {
+      "epoch": 0.6076687313800716,
+      "grad_norm": 1.625,
+      "learning_rate": 1.8127370959508995e-05,
+      "loss": 1.1296,
+      "step": 3544
+    },
+    {
+      "epoch": 0.6078401954690614,
+      "grad_norm": 1.5703125,
+      "learning_rate": 1.81263186379713e-05,
+      "loss": 0.9598,
+      "step": 3545
+    },
+    {
+      "epoch": 0.6080116595580514,
+      "grad_norm": 1.59375,
+      "learning_rate": 1.812526605140375e-05,
+      "loss": 1.0185,
+      "step": 3546
+    },
+    {
+      "epoch": 0.6081831236470412,
+      "grad_norm": 1.6015625,
+      "learning_rate": 1.8124213199840674e-05,
+      "loss": 0.9668,
+      "step": 3547
+    },
+    {
+      "epoch": 0.608354587736031,
+      "grad_norm": 1.546875,
+      "learning_rate": 1.8123160083316417e-05,
+      "loss": 1.0008,
+      "step": 3548
+    },
+    {
+      "epoch": 0.6085260518250208,
+      "grad_norm": 1.5859375,
+      "learning_rate": 1.8122106701865317e-05,
+      "loss": 0.9254,
+      "step": 3549
+    },
+    {
+      "epoch": 0.6086975159140108,
+      "grad_norm": 1.5859375,
+      "learning_rate": 1.8121053055521736e-05,
+      "loss": 1.021,
+      "step": 3550
+    },
+    {
+      "epoch": 0.6088689800030006,
+      "grad_norm": 1.59375,
+      "learning_rate": 1.811999914432003e-05,
+      "loss": 1.0135,
+      "step": 3551
+    },
+    {
+      "epoch": 0.6090404440919904,
+      "grad_norm": 1.515625,
+      "learning_rate": 1.8118944968294578e-05,
+      "loss": 0.9984,
+      "step": 3552
+    },
+    {
+      "epoch": 0.6092119081809804,
+      "grad_norm": 1.4609375,
+      "learning_rate": 1.8117890527479757e-05,
+      "loss": 0.8945,
+      "step": 3553
+    },
+    {
+      "epoch": 0.6093833722699702,
+      "grad_norm": 1.4921875,
+      "learning_rate": 1.811683582190996e-05,
+      "loss": 0.9852,
+      "step": 3554
+    },
+    {
+      "epoch": 0.60955483635896,
+      "grad_norm": 1.578125,
+      "learning_rate": 1.8115780851619575e-05,
+      "loss": 1.0574,
+      "step": 3555
+    },
+    {
+      "epoch": 0.60972630044795,
+      "grad_norm": 1.65625,
+      "learning_rate": 1.8114725616643018e-05,
+      "loss": 1.0476,
+      "step": 3556
+    },
+    {
+      "epoch": 0.6098977645369398,
+      "grad_norm": 1.546875,
+      "learning_rate": 1.8113670117014702e-05,
+      "loss": 1.0345,
+      "step": 3557
+    },
+    {
+      "epoch": 0.6100692286259296,
+      "grad_norm": 1.484375,
+      "learning_rate": 1.8112614352769048e-05,
+      "loss": 0.9556,
+      "step": 3558
+    },
+    {
+      "epoch": 0.6102406927149195,
+      "grad_norm": 1.5859375,
+      "learning_rate": 1.8111558323940493e-05,
+      "loss": 1.0069,
+      "step": 3559
+    },
+    {
+      "epoch": 0.6104121568039094,
+      "grad_norm": 1.640625,
+      "learning_rate": 1.8110502030563477e-05,
+      "loss": 1.0486,
+      "step": 3560
+    },
+    {
+      "epoch": 0.6105836208928992,
+      "grad_norm": 1.6171875,
+      "learning_rate": 1.8109445472672448e-05,
+      "loss": 0.9778,
+      "step": 3561
+    },
+    {
+      "epoch": 0.6107550849818891,
+      "grad_norm": 1.5859375,
+      "learning_rate": 1.8108388650301862e-05,
+      "loss": 1.0094,
+      "step": 3562
+    },
+    {
+      "epoch": 0.610926549070879,
+      "grad_norm": 1.5859375,
+      "learning_rate": 1.810733156348619e-05,
+      "loss": 0.9897,
+      "step": 3563
+    },
+    {
+      "epoch": 0.6110980131598688,
+      "grad_norm": 1.625,
+      "learning_rate": 1.8106274212259906e-05,
+      "loss": 1.0612,
+      "step": 3564
+    },
+    {
+      "epoch": 0.6112694772488587,
+      "grad_norm": 1.6015625,
+      "learning_rate": 1.8105216596657497e-05,
+      "loss": 0.9924,
+      "step": 3565
+    },
+    {
+      "epoch": 0.6114409413378485,
+      "grad_norm": 1.5546875,
+      "learning_rate": 1.810415871671345e-05,
+      "loss": 1.0056,
+      "step": 3566
+    },
+    {
+      "epoch": 0.6116124054268384,
+      "grad_norm": 1.671875,
+      "learning_rate": 1.810310057246227e-05,
+      "loss": 1.1138,
+      "step": 3567
+    },
+    {
+      "epoch": 0.6117838695158283,
+      "grad_norm": 1.6328125,
+      "learning_rate": 1.8102042163938472e-05,
+      "loss": 1.1251,
+      "step": 3568
+    },
+    {
+      "epoch": 0.6119553336048181,
+      "grad_norm": 1.5703125,
+      "learning_rate": 1.8100983491176563e-05,
+      "loss": 0.981,
+      "step": 3569
+    },
+    {
+      "epoch": 0.612126797693808,
+      "grad_norm": 1.53125,
+      "learning_rate": 1.809992455421108e-05,
+      "loss": 1.0471,
+      "step": 3570
+    },
+    {
+      "epoch": 0.6122982617827979,
+      "grad_norm": 1.6328125,
+      "learning_rate": 1.8098865353076554e-05,
+      "loss": 1.0886,
+      "step": 3571
+    },
+    {
+      "epoch": 0.6124697258717877,
+      "grad_norm": 1.7109375,
+      "learning_rate": 1.8097805887807536e-05,
+      "loss": 1.1298,
+      "step": 3572
+    },
+    {
+      "epoch": 0.6126411899607775,
+      "grad_norm": 1.625,
+      "learning_rate": 1.8096746158438568e-05,
+      "loss": 1.0555,
+      "step": 3573
+    },
+    {
+      "epoch": 0.6128126540497675,
+      "grad_norm": 1.6171875,
+      "learning_rate": 1.8095686165004222e-05,
+      "loss": 1.0145,
+      "step": 3574
+    },
+    {
+      "epoch": 0.6129841181387573,
+      "grad_norm": 1.5390625,
+      "learning_rate": 1.8094625907539065e-05,
+      "loss": 0.9968,
+      "step": 3575
+    },
+    {
+      "epoch": 0.6131555822277471,
+      "grad_norm": 1.5625,
+      "learning_rate": 1.8093565386077675e-05,
+      "loss": 0.9829,
+      "step": 3576
+    },
+    {
+      "epoch": 0.6133270463167371,
+      "grad_norm": 1.609375,
+      "learning_rate": 1.8092504600654642e-05,
+      "loss": 1.142,
+      "step": 3577
+    },
+    {
+      "epoch": 0.6134985104057269,
+      "grad_norm": 1.5234375,
+      "learning_rate": 1.809144355130456e-05,
+      "loss": 1.0224,
+      "step": 3578
+    },
+    {
+      "epoch": 0.6136699744947167,
+      "grad_norm": 1.6171875,
+      "learning_rate": 1.8090382238062028e-05,
+      "loss": 1.0181,
+      "step": 3579
+    },
+    {
+      "epoch": 0.6138414385837067,
+      "grad_norm": 1.5234375,
+      "learning_rate": 1.808932066096167e-05,
+      "loss": 1.0756,
+      "step": 3580
+    },
+    {
+      "epoch": 0.6140129026726965,
+      "grad_norm": 1.5546875,
+      "learning_rate": 1.8088258820038105e-05,
+      "loss": 1.0064,
+      "step": 3581
+    },
+    {
+      "epoch": 0.6141843667616863,
+      "grad_norm": 1.5078125,
+      "learning_rate": 1.8087196715325962e-05,
+      "loss": 0.9854,
+      "step": 3582
+    },
+    {
+      "epoch": 0.6143558308506762,
+      "grad_norm": 1.625,
+      "learning_rate": 1.808613434685988e-05,
+      "loss": 0.9552,
+      "step": 3583
+    },
+    {
+      "epoch": 0.6145272949396661,
+      "grad_norm": 1.515625,
+      "learning_rate": 1.8085071714674506e-05,
+      "loss": 0.979,
+      "step": 3584
+    },
+    {
+      "epoch": 0.6146987590286559,
+      "grad_norm": 1.5859375,
+      "learning_rate": 1.80840088188045e-05,
+      "loss": 1.1012,
+      "step": 3585
+    },
+    {
+      "epoch": 0.6148702231176458,
+      "grad_norm": 1.546875,
+      "learning_rate": 1.8082945659284522e-05,
+      "loss": 1.0108,
+      "step": 3586
+    },
+    {
+      "epoch": 0.6150416872066357,
+      "grad_norm": 1.5390625,
+      "learning_rate": 1.8081882236149253e-05,
+      "loss": 0.9735,
+      "step": 3587
+    },
+    {
+      "epoch": 0.6152131512956255,
+      "grad_norm": 1.59375,
+      "learning_rate": 1.8080818549433367e-05,
+      "loss": 1.0742,
+      "step": 3588
+    },
+    {
+      "epoch": 0.6153846153846154,
+      "grad_norm": 1.59375,
+      "learning_rate": 1.807975459917156e-05,
+      "loss": 1.1254,
+      "step": 3589
+    },
+    {
+      "epoch": 0.6155560794736052,
+      "grad_norm": 1.5625,
+      "learning_rate": 1.807869038539853e-05,
+      "loss": 1.0363,
+      "step": 3590
+    },
+    {
+      "epoch": 0.6157275435625951,
+      "grad_norm": 1.578125,
+      "learning_rate": 1.8077625908148983e-05,
+      "loss": 1.0375,
+      "step": 3591
+    },
+    {
+      "epoch": 0.615899007651585,
+      "grad_norm": 1.5,
+      "learning_rate": 1.807656116745764e-05,
+      "loss": 0.9396,
+      "step": 3592
+    },
+    {
+      "epoch": 0.6160704717405748,
+      "grad_norm": 1.6015625,
+      "learning_rate": 1.8075496163359224e-05,
+      "loss": 1.0285,
+      "step": 3593
+    },
+    {
+      "epoch": 0.6162419358295647,
+      "grad_norm": 1.5625,
+      "learning_rate": 1.8074430895888466e-05,
+      "loss": 0.9645,
+      "step": 3594
+    },
+    {
+      "epoch": 0.6164133999185546,
+      "grad_norm": 1.625,
+      "learning_rate": 1.8073365365080114e-05,
+      "loss": 1.091,
+      "step": 3595
+    },
+    {
+      "epoch": 0.6165848640075444,
+      "grad_norm": 1.6875,
+      "learning_rate": 1.8072299570968915e-05,
+      "loss": 1.0924,
+      "step": 3596
+    },
+    {
+      "epoch": 0.6167563280965342,
+      "grad_norm": 1.65625,
+      "learning_rate": 1.8071233513589626e-05,
+      "loss": 1.0404,
+      "step": 3597
+    },
+    {
+      "epoch": 0.6169277921855242,
+      "grad_norm": 1.59375,
+      "learning_rate": 1.8070167192977024e-05,
+      "loss": 1.0805,
+      "step": 3598
+    },
+    {
+      "epoch": 0.617099256274514,
+      "grad_norm": 1.5390625,
+      "learning_rate": 1.8069100609165878e-05,
+      "loss": 1.0391,
+      "step": 3599
+    },
+    {
+      "epoch": 0.6172707203635038,
+      "grad_norm": 1.6015625,
+      "learning_rate": 1.8068033762190977e-05,
+      "loss": 1.083,
+      "step": 3600
+    },
+    {
+      "epoch": 0.6174421844524938,
+      "grad_norm": 1.6328125,
+      "learning_rate": 1.8066966652087114e-05,
+      "loss": 0.9962,
+      "step": 3601
+    },
+    {
+      "epoch": 0.6176136485414836,
+      "grad_norm": 1.75,
+      "learning_rate": 1.8065899278889088e-05,
+      "loss": 1.0419,
+      "step": 3602
+    },
+    {
+      "epoch": 0.6177851126304734,
+      "grad_norm": 1.53125,
+      "learning_rate": 1.8064831642631712e-05,
+      "loss": 1.1506,
+      "step": 3603
+    },
+    {
+      "epoch": 0.6179565767194634,
+      "grad_norm": 1.640625,
+      "learning_rate": 1.8063763743349813e-05,
+      "loss": 1.063,
+      "step": 3604
+    },
+    {
+      "epoch": 0.6181280408084532,
+      "grad_norm": 1.515625,
+      "learning_rate": 1.806269558107821e-05,
+      "loss": 0.9929,
+      "step": 3605
+    },
+    {
+      "epoch": 0.618299504897443,
+      "grad_norm": 1.53125,
+      "learning_rate": 1.806162715585174e-05,
+      "loss": 0.9639,
+      "step": 3606
+    },
+    {
+      "epoch": 0.618470968986433,
+      "grad_norm": 1.5859375,
+      "learning_rate": 1.8060558467705254e-05,
+      "loss": 0.9289,
+      "step": 3607
+    },
+    {
+      "epoch": 0.6186424330754228,
+      "grad_norm": 1.5703125,
+      "learning_rate": 1.8059489516673603e-05,
+      "loss": 0.9864,
+      "step": 3608
+    },
+    {
+      "epoch": 0.6188138971644126,
+      "grad_norm": 1.6328125,
+      "learning_rate": 1.8058420302791647e-05,
+      "loss": 1.0511,
+      "step": 3609
+    },
+    {
+      "epoch": 0.6189853612534025,
+      "grad_norm": 1.5859375,
+      "learning_rate": 1.8057350826094263e-05,
+      "loss": 0.9789,
+      "step": 3610
+    },
+    {
+      "epoch": 0.6191568253423924,
+      "grad_norm": 1.515625,
+      "learning_rate": 1.805628108661633e-05,
+      "loss": 1.0056,
+      "step": 3611
+    },
+    {
+      "epoch": 0.6193282894313822,
+      "grad_norm": 1.625,
+      "learning_rate": 1.8055211084392728e-05,
+      "loss": 0.9905,
+      "step": 3612
+    },
+    {
+      "epoch": 0.6194997535203721,
+      "grad_norm": 1.6015625,
+      "learning_rate": 1.8054140819458362e-05,
+      "loss": 0.9457,
+      "step": 3613
+    },
+    {
+      "epoch": 0.619671217609362,
+      "grad_norm": 2.078125,
+      "learning_rate": 1.8053070291848132e-05,
+      "loss": 1.0677,
+      "step": 3614
+    },
+    {
+      "epoch": 0.6198426816983518,
+      "grad_norm": 1.453125,
+      "learning_rate": 1.805199950159696e-05,
+      "loss": 0.9021,
+      "step": 3615
+    },
+    {
+      "epoch": 0.6200141457873417,
+      "grad_norm": 1.6484375,
+      "learning_rate": 1.805092844873976e-05,
+      "loss": 1.0903,
+      "step": 3616
+    },
+    {
+      "epoch": 0.6201856098763315,
+      "grad_norm": 1.59375,
+      "learning_rate": 1.8049857133311467e-05,
+      "loss": 1.01,
+      "step": 3617
+    },
+    {
+      "epoch": 0.6203570739653214,
+      "grad_norm": 1.640625,
+      "learning_rate": 1.804878555534702e-05,
+      "loss": 0.9704,
+      "step": 3618
+    },
+    {
+      "epoch": 0.6205285380543113,
+      "grad_norm": 1.609375,
+      "learning_rate": 1.8047713714881366e-05,
+      "loss": 0.9934,
+      "step": 3619
+    },
+    {
+      "epoch": 0.6207000021433011,
+      "grad_norm": 1.6015625,
+      "learning_rate": 1.8046641611949463e-05,
+      "loss": 1.0609,
+      "step": 3620
+    },
+    {
+      "epoch": 0.620871466232291,
+      "grad_norm": 2.3125,
+      "learning_rate": 1.8045569246586274e-05,
+      "loss": 1.0085,
+      "step": 3621
+    },
+    {
+      "epoch": 0.6210429303212809,
+      "grad_norm": 1.609375,
+      "learning_rate": 1.804449661882678e-05,
+      "loss": 1.0796,
+      "step": 3622
+    },
+    {
+      "epoch": 0.6212143944102707,
+      "grad_norm": 1.6484375,
+      "learning_rate": 1.8043423728705955e-05,
+      "loss": 1.0437,
+      "step": 3623
+    },
+    {
+      "epoch": 0.6213858584992605,
+      "grad_norm": 1.53125,
+      "learning_rate": 1.8042350576258796e-05,
+      "loss": 1.0536,
+      "step": 3624
+    },
+    {
+      "epoch": 0.6215573225882505,
+      "grad_norm": 1.5546875,
+      "learning_rate": 1.8041277161520296e-05,
+      "loss": 0.9373,
+      "step": 3625
+    },
+    {
+      "epoch": 0.6217287866772403,
+      "grad_norm": 2.59375,
+      "learning_rate": 1.804020348452547e-05,
+      "loss": 1.0301,
+      "step": 3626
+    },
+    {
+      "epoch": 0.6219002507662301,
+      "grad_norm": 1.5703125,
+      "learning_rate": 1.8039129545309333e-05,
+      "loss": 1.0713,
+      "step": 3627
+    },
+    {
+      "epoch": 0.6220717148552201,
+      "grad_norm": 1.546875,
+      "learning_rate": 1.8038055343906906e-05,
+      "loss": 0.9714,
+      "step": 3628
+    },
+    {
+      "epoch": 0.6222431789442099,
+      "grad_norm": 1.515625,
+      "learning_rate": 1.803698088035323e-05,
+      "loss": 1.069,
+      "step": 3629
+    },
+    {
+      "epoch": 0.6224146430331997,
+      "grad_norm": 1.6875,
+      "learning_rate": 1.8035906154683342e-05,
+      "loss": 1.0507,
+      "step": 3630
+    },
+    {
+      "epoch": 0.6225861071221896,
+      "grad_norm": 1.8359375,
+      "learning_rate": 1.8034831166932288e-05,
+      "loss": 1.0842,
+      "step": 3631
+    },
+    {
+      "epoch": 0.6227575712111795,
+      "grad_norm": 1.609375,
+      "learning_rate": 1.803375591713514e-05,
+      "loss": 0.9876,
+      "step": 3632
+    },
+    {
+      "epoch": 0.6229290353001693,
+      "grad_norm": 1.5390625,
+      "learning_rate": 1.8032680405326956e-05,
+      "loss": 0.9642,
+      "step": 3633
+    },
+    {
+      "epoch": 0.6231004993891592,
+      "grad_norm": 1.7421875,
+      "learning_rate": 1.8031604631542815e-05,
+      "loss": 1.0226,
+      "step": 3634
+    },
+    {
+      "epoch": 0.6232719634781491,
+      "grad_norm": 1.5859375,
+      "learning_rate": 1.8030528595817804e-05,
+      "loss": 1.0634,
+      "step": 3635
+    },
+    {
+      "epoch": 0.6234434275671389,
+      "grad_norm": 1.4453125,
+      "learning_rate": 1.8029452298187015e-05,
+      "loss": 0.9581,
+      "step": 3636
+    },
+    {
+      "epoch": 0.6236148916561288,
+      "grad_norm": 1.6953125,
+      "learning_rate": 1.802837573868555e-05,
+      "loss": 1.0147,
+      "step": 3637
+    },
+    {
+      "epoch": 0.6237863557451186,
+      "grad_norm": 1.7109375,
+      "learning_rate": 1.802729891734852e-05,
+      "loss": 0.9945,
+      "step": 3638
+    },
+    {
+      "epoch": 0.6239578198341085,
+      "grad_norm": 1.578125,
+      "learning_rate": 1.8026221834211045e-05,
+      "loss": 0.9873,
+      "step": 3639
+    },
+    {
+      "epoch": 0.6241292839230984,
+      "grad_norm": 2.296875,
+      "learning_rate": 1.8025144489308255e-05,
+      "loss": 1.0331,
+      "step": 3640
+    },
+    {
+      "epoch": 0.6243007480120882,
+      "grad_norm": 1.5625,
+      "learning_rate": 1.802406688267528e-05,
+      "loss": 0.9741,
+      "step": 3641
+    },
+    {
+      "epoch": 0.6244722121010781,
+      "grad_norm": 1.5703125,
+      "learning_rate": 1.802298901434727e-05,
+      "loss": 1.0698,
+      "step": 3642
+    },
+    {
+      "epoch": 0.624643676190068,
+      "grad_norm": 1.71875,
+      "learning_rate": 1.8021910884359373e-05,
+      "loss": 1.0824,
+      "step": 3643
+    },
+    {
+      "epoch": 0.6248151402790578,
+      "grad_norm": 1.546875,
+      "learning_rate": 1.802083249274676e-05,
+      "loss": 0.9868,
+      "step": 3644
+    },
+    {
+      "epoch": 0.6249866043680476,
+      "grad_norm": 1.90625,
+      "learning_rate": 1.8019753839544588e-05,
+      "loss": 0.9885,
+      "step": 3645
+    },
+    {
+      "epoch": 0.6251580684570375,
+      "grad_norm": 1.5546875,
+      "learning_rate": 1.801867492478805e-05,
+      "loss": 0.9759,
+      "step": 3646
+    },
+    {
+      "epoch": 0.6253295325460274,
+      "grad_norm": 1.53125,
+      "learning_rate": 1.8017595748512327e-05,
+      "loss": 0.9293,
+      "step": 3647
+    },
+    {
+      "epoch": 0.6255009966350172,
+      "grad_norm": 1.625,
+      "learning_rate": 1.8016516310752614e-05,
+      "loss": 1.0215,
+      "step": 3648
+    },
+    {
+      "epoch": 0.6256724607240071,
+      "grad_norm": 1.5390625,
+      "learning_rate": 1.8015436611544112e-05,
+      "loss": 1.0435,
+      "step": 3649
+    },
+    {
+      "epoch": 0.625843924812997,
+      "grad_norm": 1.5859375,
+      "learning_rate": 1.8014356650922043e-05,
+      "loss": 1.016,
+      "step": 3650
+    },
+    {
+      "epoch": 0.6260153889019868,
+      "grad_norm": 1.6171875,
+      "learning_rate": 1.8013276428921625e-05,
+      "loss": 1.0187,
+      "step": 3651
+    },
+    {
+      "epoch": 0.6261868529909767,
+      "grad_norm": 1.6484375,
+      "learning_rate": 1.8012195945578085e-05,
+      "loss": 1.0304,
+      "step": 3652
+    },
+    {
+      "epoch": 0.6263583170799666,
+      "grad_norm": 1.6328125,
+      "learning_rate": 1.8011115200926666e-05,
+      "loss": 1.1196,
+      "step": 3653
+    },
+    {
+      "epoch": 0.6265297811689564,
+      "grad_norm": 1.5390625,
+      "learning_rate": 1.8010034195002616e-05,
+      "loss": 0.9716,
+      "step": 3654
+    },
+    {
+      "epoch": 0.6267012452579462,
+      "grad_norm": 1.5703125,
+      "learning_rate": 1.800895292784118e-05,
+      "loss": 1.0112,
+      "step": 3655
+    },
+    {
+      "epoch": 0.6268727093469362,
+      "grad_norm": 1.6640625,
+      "learning_rate": 1.800787139947764e-05,
+      "loss": 1.0518,
+      "step": 3656
+    },
+    {
+      "epoch": 0.627044173435926,
+      "grad_norm": 1.5546875,
+      "learning_rate": 1.800678960994725e-05,
+      "loss": 0.9209,
+      "step": 3657
+    },
+    {
+      "epoch": 0.6272156375249158,
+      "grad_norm": 1.640625,
+      "learning_rate": 1.8005707559285306e-05,
+      "loss": 1.0306,
+      "step": 3658
+    },
+    {
+      "epoch": 0.6273871016139058,
+      "grad_norm": 1.5625,
+      "learning_rate": 1.800462524752709e-05,
+      "loss": 1.0427,
+      "step": 3659
+    },
+    {
+      "epoch": 0.6275585657028956,
+      "grad_norm": 1.609375,
+      "learning_rate": 1.80035426747079e-05,
+      "loss": 1.0499,
+      "step": 3660
+    },
+    {
+      "epoch": 0.6277300297918854,
+      "grad_norm": 1.5234375,
+      "learning_rate": 1.8002459840863044e-05,
+      "loss": 0.9191,
+      "step": 3661
+    },
+    {
+      "epoch": 0.6279014938808753,
+      "grad_norm": 1.6875,
+      "learning_rate": 1.8001376746027844e-05,
+      "loss": 1.1145,
+      "step": 3662
+    },
+    {
+      "epoch": 0.6280729579698652,
+      "grad_norm": 1.5390625,
+      "learning_rate": 1.800029339023761e-05,
+      "loss": 0.9926,
+      "step": 3663
+    },
+    {
+      "epoch": 0.628244422058855,
+      "grad_norm": 1.609375,
+      "learning_rate": 1.799920977352769e-05,
+      "loss": 1.0446,
+      "step": 3664
+    },
+    {
+      "epoch": 0.6284158861478449,
+      "grad_norm": 1.6328125,
+      "learning_rate": 1.799812589593341e-05,
+      "loss": 1.0364,
+      "step": 3665
+    },
+    {
+      "epoch": 0.6285873502368348,
+      "grad_norm": 1.4765625,
+      "learning_rate": 1.7997041757490133e-05,
+      "loss": 0.954,
+      "step": 3666
+    },
+    {
+      "epoch": 0.6287588143258246,
+      "grad_norm": 1.671875,
+      "learning_rate": 1.7995957358233206e-05,
+      "loss": 1.0657,
+      "step": 3667
+    },
+    {
+      "epoch": 0.6289302784148145,
+      "grad_norm": 1.6015625,
+      "learning_rate": 1.7994872698198e-05,
+      "loss": 1.0128,
+      "step": 3668
+    },
+    {
+      "epoch": 0.6291017425038044,
+      "grad_norm": 1.5234375,
+      "learning_rate": 1.799378777741989e-05,
+      "loss": 1.011,
+      "step": 3669
+    },
+    {
+      "epoch": 0.6292732065927942,
+      "grad_norm": 1.625,
+      "learning_rate": 1.7992702595934262e-05,
+      "loss": 1.1297,
+      "step": 3670
+    },
+    {
+      "epoch": 0.6294446706817841,
+      "grad_norm": 1.578125,
+      "learning_rate": 1.79916171537765e-05,
+      "loss": 1.0064,
+      "step": 3671
+    },
+    {
+      "epoch": 0.6296161347707739,
+      "grad_norm": 1.5703125,
+      "learning_rate": 1.7990531450982013e-05,
+      "loss": 0.9921,
+      "step": 3672
+    },
+    {
+      "epoch": 0.6297875988597638,
+      "grad_norm": 1.4765625,
+      "learning_rate": 1.7989445487586202e-05,
+      "loss": 1.0384,
+      "step": 3673
+    },
+    {
+      "epoch": 0.6299590629487537,
+      "grad_norm": 1.5546875,
+      "learning_rate": 1.798835926362449e-05,
+      "loss": 1.1058,
+      "step": 3674
+    },
+    {
+      "epoch": 0.6301305270377435,
+      "grad_norm": 1.515625,
+      "learning_rate": 1.79872727791323e-05,
+      "loss": 0.9876,
+      "step": 3675
+    },
+    {
+      "epoch": 0.6303019911267334,
+      "grad_norm": 1.6171875,
+      "learning_rate": 1.798618603414507e-05,
+      "loss": 1.0731,
+      "step": 3676
+    },
+    {
+      "epoch": 0.6304734552157233,
+      "grad_norm": 1.6953125,
+      "learning_rate": 1.7985099028698237e-05,
+      "loss": 1.0111,
+      "step": 3677
+    },
+    {
+      "epoch": 0.6306449193047131,
+      "grad_norm": 1.5625,
+      "learning_rate": 1.7984011762827257e-05,
+      "loss": 0.9985,
+      "step": 3678
+    },
+    {
+      "epoch": 0.6308163833937029,
+      "grad_norm": 1.609375,
+      "learning_rate": 1.7982924236567587e-05,
+      "loss": 1.0882,
+      "step": 3679
+    },
+    {
+      "epoch": 0.6309878474826929,
+      "grad_norm": 1.609375,
+      "learning_rate": 1.79818364499547e-05,
+      "loss": 1.0011,
+      "step": 3680
+    },
+    {
+      "epoch": 0.6311593115716827,
+      "grad_norm": 1.6875,
+      "learning_rate": 1.7980748403024067e-05,
+      "loss": 1.0398,
+      "step": 3681
+    },
+    {
+      "epoch": 0.6313307756606725,
+      "grad_norm": 1.59375,
+      "learning_rate": 1.7979660095811177e-05,
+      "loss": 1.0896,
+      "step": 3682
+    },
+    {
+      "epoch": 0.6315022397496625,
+      "grad_norm": 1.7109375,
+      "learning_rate": 1.7978571528351517e-05,
+      "loss": 1.0367,
+      "step": 3683
+    },
+    {
+      "epoch": 0.6316737038386523,
+      "grad_norm": 1.6015625,
+      "learning_rate": 1.7977482700680605e-05,
+      "loss": 1.0867,
+      "step": 3684
+    },
+    {
+      "epoch": 0.6318451679276421,
+      "grad_norm": 1.640625,
+      "learning_rate": 1.7976393612833935e-05,
+      "loss": 1.028,
+      "step": 3685
+    },
+    {
+      "epoch": 0.632016632016632,
+      "grad_norm": 1.6328125,
+      "learning_rate": 1.7975304264847035e-05,
+      "loss": 1.0476,
+      "step": 3686
+    },
+    {
+      "epoch": 0.6321880961056219,
+      "grad_norm": 1.578125,
+      "learning_rate": 1.7974214656755432e-05,
+      "loss": 0.9904,
+      "step": 3687
+    },
+    {
+      "epoch": 0.6323595601946117,
+      "grad_norm": 1.6484375,
+      "learning_rate": 1.7973124788594658e-05,
+      "loss": 1.0644,
+      "step": 3688
+    },
+    {
+      "epoch": 0.6325310242836016,
+      "grad_norm": 1.5234375,
+      "learning_rate": 1.797203466040026e-05,
+      "loss": 1.0607,
+      "step": 3689
+    },
+    {
+      "epoch": 0.6327024883725915,
+      "grad_norm": 1.53125,
+      "learning_rate": 1.7970944272207795e-05,
+      "loss": 0.9432,
+      "step": 3690
+    },
+    {
+      "epoch": 0.6328739524615813,
+      "grad_norm": 1.53125,
+      "learning_rate": 1.796985362405282e-05,
+      "loss": 1.0377,
+      "step": 3691
+    },
+    {
+      "epoch": 0.6330454165505712,
+      "grad_norm": 1.7265625,
+      "learning_rate": 1.7968762715970905e-05,
+      "loss": 1.1736,
+      "step": 3692
+    },
+    {
+      "epoch": 0.633216880639561,
+      "grad_norm": 1.578125,
+      "learning_rate": 1.7967671547997633e-05,
+      "loss": 1.0287,
+      "step": 3693
+    },
+    {
+      "epoch": 0.6333883447285509,
+      "grad_norm": 1.5390625,
+      "learning_rate": 1.7966580120168588e-05,
+      "loss": 1.0058,
+      "step": 3694
+    },
+    {
+      "epoch": 0.6335598088175408,
+      "grad_norm": 1.7109375,
+      "learning_rate": 1.7965488432519363e-05,
+      "loss": 1.0454,
+      "step": 3695
+    },
+    {
+      "epoch": 0.6337312729065306,
+      "grad_norm": 1.640625,
+      "learning_rate": 1.796439648508557e-05,
+      "loss": 0.9996,
+      "step": 3696
+    },
+    {
+      "epoch": 0.6339027369955205,
+      "grad_norm": 1.5625,
+      "learning_rate": 1.7963304277902815e-05,
+      "loss": 0.9916,
+      "step": 3697
+    },
+    {
+      "epoch": 0.6340742010845104,
+      "grad_norm": 1.5390625,
+      "learning_rate": 1.7962211811006717e-05,
+      "loss": 1.0748,
+      "step": 3698
+    },
+    {
+      "epoch": 0.6342456651735002,
+      "grad_norm": 1.578125,
+      "learning_rate": 1.7961119084432914e-05,
+      "loss": 0.9708,
+      "step": 3699
+    },
+    {
+      "epoch": 0.63441712926249,
+      "grad_norm": 1.4921875,
+      "learning_rate": 1.7960026098217033e-05,
+      "loss": 1.0117,
+      "step": 3700
+    },
+    {
+      "epoch": 0.63458859335148,
+      "grad_norm": 1.625,
+      "learning_rate": 1.795893285239473e-05,
+      "loss": 1.0854,
+      "step": 3701
+    },
+    {
+      "epoch": 0.6347600574404698,
+      "grad_norm": 1.5625,
+      "learning_rate": 1.7957839347001656e-05,
+      "loss": 1.0339,
+      "step": 3702
+    },
+    {
+      "epoch": 0.6349315215294596,
+      "grad_norm": 1.578125,
+      "learning_rate": 1.7956745582073472e-05,
+      "loss": 1.0549,
+      "step": 3703
+    },
+    {
+      "epoch": 0.6351029856184496,
+      "grad_norm": 1.65625,
+      "learning_rate": 1.7955651557645856e-05,
+      "loss": 1.0122,
+      "step": 3704
+    },
+    {
+      "epoch": 0.6352744497074394,
+      "grad_norm": 1.5234375,
+      "learning_rate": 1.795455727375448e-05,
+      "loss": 0.9592,
+      "step": 3705
+    },
+    {
+      "epoch": 0.6354459137964292,
+      "grad_norm": 1.6484375,
+      "learning_rate": 1.7953462730435043e-05,
+      "loss": 1.1369,
+      "step": 3706
+    },
+    {
+      "epoch": 0.6356173778854192,
+      "grad_norm": 1.53125,
+      "learning_rate": 1.795236792772323e-05,
+      "loss": 0.9327,
+      "step": 3707
+    },
+    {
+      "epoch": 0.635788841974409,
+      "grad_norm": 1.546875,
+      "learning_rate": 1.795127286565476e-05,
+      "loss": 1.0152,
+      "step": 3708
+    },
+    {
+      "epoch": 0.6359603060633988,
+      "grad_norm": 1.6015625,
+      "learning_rate": 1.7950177544265333e-05,
+      "loss": 1.049,
+      "step": 3709
+    },
+    {
+      "epoch": 0.6361317701523888,
+      "grad_norm": 1.5390625,
+      "learning_rate": 1.7949081963590683e-05,
+      "loss": 1.0378,
+      "step": 3710
+    },
+    {
+      "epoch": 0.6363032342413786,
+      "grad_norm": 1.609375,
+      "learning_rate": 1.7947986123666536e-05,
+      "loss": 1.0444,
+      "step": 3711
+    },
+    {
+      "epoch": 0.6364746983303684,
+      "grad_norm": 1.5546875,
+      "learning_rate": 1.7946890024528633e-05,
+      "loss": 1.0106,
+      "step": 3712
+    },
+    {
+      "epoch": 0.6366461624193583,
+      "grad_norm": 1.5625,
+      "learning_rate": 1.794579366621272e-05,
+      "loss": 1.0274,
+      "step": 3713
+    },
+    {
+      "epoch": 0.6368176265083482,
+      "grad_norm": 1.5546875,
+      "learning_rate": 1.7944697048754552e-05,
+      "loss": 0.9483,
+      "step": 3714
+    },
+    {
+      "epoch": 0.636989090597338,
+      "grad_norm": 1.5859375,
+      "learning_rate": 1.79436001721899e-05,
+      "loss": 1.096,
+      "step": 3715
+    },
+    {
+      "epoch": 0.6371605546863279,
+      "grad_norm": 1.5234375,
+      "learning_rate": 1.794250303655453e-05,
+      "loss": 1.0592,
+      "step": 3716
+    },
+    {
+      "epoch": 0.6373320187753178,
+      "grad_norm": 1.5703125,
+      "learning_rate": 1.794140564188423e-05,
+      "loss": 1.0463,
+      "step": 3717
+    },
+    {
+      "epoch": 0.6375034828643076,
+      "grad_norm": 1.546875,
+      "learning_rate": 1.794030798821479e-05,
+      "loss": 0.9117,
+      "step": 3718
+    },
+    {
+      "epoch": 0.6376749469532975,
+      "grad_norm": 1.671875,
+      "learning_rate": 1.7939210075582003e-05,
+      "loss": 1.051,
+      "step": 3719
+    },
+    {
+      "epoch": 0.6378464110422873,
+      "grad_norm": 1.5703125,
+      "learning_rate": 1.7938111904021677e-05,
+      "loss": 1.0476,
+      "step": 3720
+    },
+    {
+      "epoch": 0.6380178751312772,
+      "grad_norm": 1.609375,
+      "learning_rate": 1.7937013473569633e-05,
+      "loss": 0.9729,
+      "step": 3721
+    },
+    {
+      "epoch": 0.6381893392202671,
+      "grad_norm": 1.625,
+      "learning_rate": 1.793591478426169e-05,
+      "loss": 1.0366,
+      "step": 3722
+    },
+    {
+      "epoch": 0.6383608033092569,
+      "grad_norm": 1.5625,
+      "learning_rate": 1.7934815836133684e-05,
+      "loss": 1.0052,
+      "step": 3723
+    },
+    {
+      "epoch": 0.6385322673982468,
+      "grad_norm": 1.6328125,
+      "learning_rate": 1.7933716629221455e-05,
+      "loss": 1.0623,
+      "step": 3724
+    },
+    {
+      "epoch": 0.6387037314872367,
+      "grad_norm": 1.5546875,
+      "learning_rate": 1.7932617163560847e-05,
+      "loss": 0.9928,
+      "step": 3725
+    },
+    {
+      "epoch": 0.6388751955762265,
+      "grad_norm": 1.5625,
+      "learning_rate": 1.793151743918772e-05,
+      "loss": 1.0026,
+      "step": 3726
+    },
+    {
+      "epoch": 0.6390466596652163,
+      "grad_norm": 1.5234375,
+      "learning_rate": 1.793041745613795e-05,
+      "loss": 1.0157,
+      "step": 3727
+    },
+    {
+      "epoch": 0.6392181237542063,
+      "grad_norm": 1.46875,
+      "learning_rate": 1.7929317214447395e-05,
+      "loss": 1.0197,
+      "step": 3728
+    },
+    {
+      "epoch": 0.6393895878431961,
+      "grad_norm": 1.515625,
+      "learning_rate": 1.7928216714151953e-05,
+      "loss": 0.9981,
+      "step": 3729
+    },
+    {
+      "epoch": 0.6395610519321859,
+      "grad_norm": 1.5546875,
+      "learning_rate": 1.7927115955287506e-05,
+      "loss": 1.009,
+      "step": 3730
+    },
+    {
+      "epoch": 0.6397325160211759,
+      "grad_norm": 1.6171875,
+      "learning_rate": 1.7926014937889956e-05,
+      "loss": 1.0699,
+      "step": 3731
+    },
+    {
+      "epoch": 0.6399039801101657,
+      "grad_norm": 1.609375,
+      "learning_rate": 1.7924913661995212e-05,
+      "loss": 1.0051,
+      "step": 3732
+    },
+    {
+      "epoch": 0.6400754441991555,
+      "grad_norm": 1.7265625,
+      "learning_rate": 1.7923812127639194e-05,
+      "loss": 1.063,
+      "step": 3733
+    },
+    {
+      "epoch": 0.6402469082881455,
+      "grad_norm": 1.6484375,
+      "learning_rate": 1.7922710334857824e-05,
+      "loss": 0.9803,
+      "step": 3734
+    },
+    {
+      "epoch": 0.6404183723771353,
+      "grad_norm": 1.609375,
+      "learning_rate": 1.7921608283687035e-05,
+      "loss": 1.0446,
+      "step": 3735
+    },
+    {
+      "epoch": 0.6405898364661251,
+      "grad_norm": 1.640625,
+      "learning_rate": 1.7920505974162766e-05,
+      "loss": 1.02,
+      "step": 3736
+    },
+    {
+      "epoch": 0.640761300555115,
+      "grad_norm": 1.6015625,
+      "learning_rate": 1.7919403406320978e-05,
+      "loss": 1.007,
+      "step": 3737
+    },
+    {
+      "epoch": 0.6409327646441049,
+      "grad_norm": 1.609375,
+      "learning_rate": 1.7918300580197617e-05,
+      "loss": 1.0234,
+      "step": 3738
+    },
+    {
+      "epoch": 0.6411042287330947,
+      "grad_norm": 1.578125,
+      "learning_rate": 1.7917197495828663e-05,
+      "loss": 0.9882,
+      "step": 3739
+    },
+    {
+      "epoch": 0.6412756928220845,
+      "grad_norm": 1.640625,
+      "learning_rate": 1.791609415325008e-05,
+      "loss": 1.0269,
+      "step": 3740
+    },
+    {
+      "epoch": 0.6414471569110745,
+      "grad_norm": 1.6015625,
+      "learning_rate": 1.7914990552497863e-05,
+      "loss": 1.018,
+      "step": 3741
+    },
+    {
+      "epoch": 0.6416186210000643,
+      "grad_norm": 1.5625,
+      "learning_rate": 1.7913886693607992e-05,
+      "loss": 1.0189,
+      "step": 3742
+    },
+    {
+      "epoch": 0.6417900850890541,
+      "grad_norm": 1.5234375,
+      "learning_rate": 1.791278257661648e-05,
+      "loss": 1.0041,
+      "step": 3743
+    },
+    {
+      "epoch": 0.641961549178044,
+      "grad_norm": 1.5703125,
+      "learning_rate": 1.791167820155933e-05,
+      "loss": 1.0526,
+      "step": 3744
+    },
+    {
+      "epoch": 0.6421330132670339,
+      "grad_norm": 1.515625,
+      "learning_rate": 1.7910573568472556e-05,
+      "loss": 1.0612,
+      "step": 3745
+    },
+    {
+      "epoch": 0.6423044773560237,
+      "grad_norm": 1.5390625,
+      "learning_rate": 1.7909468677392198e-05,
+      "loss": 1.0807,
+      "step": 3746
+    },
+    {
+      "epoch": 0.6424759414450136,
+      "grad_norm": 1.5390625,
+      "learning_rate": 1.790836352835428e-05,
+      "loss": 1.0416,
+      "step": 3747
+    },
+    {
+      "epoch": 0.6426474055340035,
+      "grad_norm": 1.5234375,
+      "learning_rate": 1.7907258121394843e-05,
+      "loss": 0.9954,
+      "step": 3748
+    },
+    {
+      "epoch": 0.6428188696229933,
+      "grad_norm": 1.515625,
+      "learning_rate": 1.7906152456549946e-05,
+      "loss": 0.9858,
+      "step": 3749
+    },
+    {
+      "epoch": 0.6429903337119832,
+      "grad_norm": 1.546875,
+      "learning_rate": 1.7905046533855642e-05,
+      "loss": 1.0425,
+      "step": 3750
+    },
+    {
+      "epoch": 0.643161797800973,
+      "grad_norm": 1.5234375,
+      "learning_rate": 1.790394035334801e-05,
+      "loss": 0.9905,
+      "step": 3751
+    },
+    {
+      "epoch": 0.6433332618899629,
+      "grad_norm": 1.609375,
+      "learning_rate": 1.790283391506311e-05,
+      "loss": 0.9942,
+      "step": 3752
+    },
+    {
+      "epoch": 0.6435047259789528,
+      "grad_norm": 1.59375,
+      "learning_rate": 1.7901727219037047e-05,
+      "loss": 1.0426,
+      "step": 3753
+    },
+    {
+      "epoch": 0.6436761900679426,
+      "grad_norm": 1.5546875,
+      "learning_rate": 1.7900620265305895e-05,
+      "loss": 1.1176,
+      "step": 3754
+    },
+    {
+      "epoch": 0.6438476541569325,
+      "grad_norm": 1.6171875,
+      "learning_rate": 1.7899513053905772e-05,
+      "loss": 1.0297,
+      "step": 3755
+    },
+    {
+      "epoch": 0.6440191182459224,
+      "grad_norm": 1.6484375,
+      "learning_rate": 1.789840558487278e-05,
+      "loss": 1.0504,
+      "step": 3756
+    },
+    {
+      "epoch": 0.6441905823349122,
+      "grad_norm": 1.609375,
+      "learning_rate": 1.7897297858243036e-05,
+      "loss": 1.021,
+      "step": 3757
+    },
+    {
+      "epoch": 0.644362046423902,
+      "grad_norm": 1.640625,
+      "learning_rate": 1.7896189874052675e-05,
+      "loss": 1.0672,
+      "step": 3758
+    },
+    {
+      "epoch": 0.644533510512892,
+      "grad_norm": 1.5234375,
+      "learning_rate": 1.7895081632337826e-05,
+      "loss": 0.9723,
+      "step": 3759
+    },
+    {
+      "epoch": 0.6447049746018818,
+      "grad_norm": 1.5234375,
+      "learning_rate": 1.7893973133134634e-05,
+      "loss": 0.946,
+      "step": 3760
+    },
+    {
+      "epoch": 0.6448764386908716,
+      "grad_norm": 1.6484375,
+      "learning_rate": 1.7892864376479254e-05,
+      "loss": 0.978,
+      "step": 3761
+    },
+    {
+      "epoch": 0.6450479027798616,
+      "grad_norm": 1.5859375,
+      "learning_rate": 1.789175536240784e-05,
+      "loss": 1.0083,
+      "step": 3762
+    },
+    {
+      "epoch": 0.6452193668688514,
+      "grad_norm": 1.5078125,
+      "learning_rate": 1.789064609095657e-05,
+      "loss": 1.0333,
+      "step": 3763
+    },
+    {
+      "epoch": 0.6453908309578412,
+      "grad_norm": 1.6328125,
+      "learning_rate": 1.788953656216162e-05,
+      "loss": 1.0297,
+      "step": 3764
+    },
+    {
+      "epoch": 0.6455622950468312,
+      "grad_norm": 1.5625,
+      "learning_rate": 1.7888426776059166e-05,
+      "loss": 0.9605,
+      "step": 3765
+    },
+    {
+      "epoch": 0.645733759135821,
+      "grad_norm": 1.6640625,
+      "learning_rate": 1.7887316732685415e-05,
+      "loss": 1.0634,
+      "step": 3766
+    },
+    {
+      "epoch": 0.6459052232248108,
+      "grad_norm": 1.65625,
+      "learning_rate": 1.7886206432076566e-05,
+      "loss": 1.0777,
+      "step": 3767
+    },
+    {
+      "epoch": 0.6460766873138007,
+      "grad_norm": 1.6328125,
+      "learning_rate": 1.7885095874268826e-05,
+      "loss": 1.0468,
+      "step": 3768
+    },
+    {
+      "epoch": 0.6462481514027906,
+      "grad_norm": 1.546875,
+      "learning_rate": 1.7883985059298418e-05,
+      "loss": 0.9256,
+      "step": 3769
+    },
+    {
+      "epoch": 0.6464196154917804,
+      "grad_norm": 1.671875,
+      "learning_rate": 1.7882873987201566e-05,
+      "loss": 1.0545,
+      "step": 3770
+    },
+    {
+      "epoch": 0.6465910795807703,
+      "grad_norm": 1.546875,
+      "learning_rate": 1.788176265801451e-05,
+      "loss": 0.9954,
+      "step": 3771
+    },
+    {
+      "epoch": 0.6467625436697602,
+      "grad_norm": 1.5390625,
+      "learning_rate": 1.7880651071773495e-05,
+      "loss": 0.9356,
+      "step": 3772
+    },
+    {
+      "epoch": 0.64693400775875,
+      "grad_norm": 1.6875,
+      "learning_rate": 1.7879539228514775e-05,
+      "loss": 1.0697,
+      "step": 3773
+    },
+    {
+      "epoch": 0.6471054718477399,
+      "grad_norm": 1.59375,
+      "learning_rate": 1.7878427128274607e-05,
+      "loss": 0.9816,
+      "step": 3774
+    },
+    {
+      "epoch": 0.6472769359367297,
+      "grad_norm": 1.5234375,
+      "learning_rate": 1.787731477108926e-05,
+      "loss": 0.9684,
+      "step": 3775
+    },
+    {
+      "epoch": 0.6474484000257196,
+      "grad_norm": 1.609375,
+      "learning_rate": 1.7876202156995018e-05,
+      "loss": 1.05,
+      "step": 3776
+    },
+    {
+      "epoch": 0.6476198641147095,
+      "grad_norm": 1.5859375,
+      "learning_rate": 1.7875089286028167e-05,
+      "loss": 0.9508,
+      "step": 3777
+    },
+    {
+      "epoch": 0.6477913282036993,
+      "grad_norm": 1.5859375,
+      "learning_rate": 1.7873976158225e-05,
+      "loss": 0.9991,
+      "step": 3778
+    },
+    {
+      "epoch": 0.6479627922926892,
+      "grad_norm": 1.5,
+      "learning_rate": 1.7872862773621814e-05,
+      "loss": 0.9117,
+      "step": 3779
+    },
+    {
+      "epoch": 0.6481342563816791,
+      "grad_norm": 1.6484375,
+      "learning_rate": 1.787174913225493e-05,
+      "loss": 0.9993,
+      "step": 3780
+    },
+    {
+      "epoch": 0.6483057204706689,
+      "grad_norm": 45.5,
+      "learning_rate": 1.7870635234160663e-05,
+      "loss": 1.1466,
+      "step": 3781
+    },
+    {
+      "epoch": 0.6484771845596587,
+      "grad_norm": 1.65625,
+      "learning_rate": 1.7869521079375345e-05,
+      "loss": 0.9891,
+      "step": 3782
+    },
+    {
+      "epoch": 0.6486486486486487,
+      "grad_norm": 1.5390625,
+      "learning_rate": 1.786840666793531e-05,
+      "loss": 0.9753,
+      "step": 3783
+    },
+    {
+      "epoch": 0.6488201127376385,
+      "grad_norm": 1.6171875,
+      "learning_rate": 1.7867291999876905e-05,
+      "loss": 1.0144,
+      "step": 3784
+    },
+    {
+      "epoch": 0.6489915768266283,
+      "grad_norm": 1.625,
+      "learning_rate": 1.786617707523648e-05,
+      "loss": 1.1283,
+      "step": 3785
+    },
+    {
+      "epoch": 0.6491630409156183,
+      "grad_norm": 1.6171875,
+      "learning_rate": 1.78650618940504e-05,
+      "loss": 0.996,
+      "step": 3786
+    },
+    {
+      "epoch": 0.6493345050046081,
+      "grad_norm": 1.6328125,
+      "learning_rate": 1.7863946456355036e-05,
+      "loss": 1.0807,
+      "step": 3787
+    },
+    {
+      "epoch": 0.6495059690935979,
+      "grad_norm": 1.5390625,
+      "learning_rate": 1.786283076218676e-05,
+      "loss": 1.0865,
+      "step": 3788
+    },
+    {
+      "epoch": 0.6496774331825879,
+      "grad_norm": 1.484375,
+      "learning_rate": 1.786171481158197e-05,
+      "loss": 1.019,
+      "step": 3789
+    },
+    {
+      "epoch": 0.6498488972715777,
+      "grad_norm": 1.4921875,
+      "learning_rate": 1.786059860457705e-05,
+      "loss": 0.8982,
+      "step": 3790
+    },
+    {
+      "epoch": 0.6500203613605675,
+      "grad_norm": 1.5234375,
+      "learning_rate": 1.7859482141208413e-05,
+      "loss": 0.9797,
+      "step": 3791
+    },
+    {
+      "epoch": 0.6501918254495574,
+      "grad_norm": 1.5234375,
+      "learning_rate": 1.7858365421512467e-05,
+      "loss": 0.9876,
+      "step": 3792
+    },
+    {
+      "epoch": 0.6503632895385473,
+      "grad_norm": 1.546875,
+      "learning_rate": 1.7857248445525636e-05,
+      "loss": 1.0003,
+      "step": 3793
+    },
+    {
+      "epoch": 0.6505347536275371,
+      "grad_norm": 1.6015625,
+      "learning_rate": 1.785613121328434e-05,
+      "loss": 0.9599,
+      "step": 3794
+    },
+    {
+      "epoch": 0.650706217716527,
+      "grad_norm": 1.59375,
+      "learning_rate": 1.785501372482503e-05,
+      "loss": 1.0637,
+      "step": 3795
+    },
+    {
+      "epoch": 0.6508776818055169,
+      "grad_norm": 1.546875,
+      "learning_rate": 1.7853895980184136e-05,
+      "loss": 0.9595,
+      "step": 3796
+    },
+    {
+      "epoch": 0.6510491458945067,
+      "grad_norm": 1.546875,
+      "learning_rate": 1.785277797939812e-05,
+      "loss": 0.9918,
+      "step": 3797
+    },
+    {
+      "epoch": 0.6512206099834966,
+      "grad_norm": 1.5859375,
+      "learning_rate": 1.7851659722503442e-05,
+      "loss": 0.9836,
+      "step": 3798
+    },
+    {
+      "epoch": 0.6513920740724864,
+      "grad_norm": 1.5859375,
+      "learning_rate": 1.7850541209536575e-05,
+      "loss": 1.044,
+      "step": 3799
+    },
+    {
+      "epoch": 0.6515635381614763,
+      "grad_norm": 1.5234375,
+      "learning_rate": 1.7849422440533998e-05,
+      "loss": 0.9397,
+      "step": 3800
+    },
+    {
+      "epoch": 0.6517350022504662,
+      "grad_norm": 1.59375,
+      "learning_rate": 1.7848303415532197e-05,
+      "loss": 0.9828,
+      "step": 3801
+    },
+    {
+      "epoch": 0.651906466339456,
+      "grad_norm": 1.640625,
+      "learning_rate": 1.7847184134567664e-05,
+      "loss": 1.0753,
+      "step": 3802
+    },
+    {
+      "epoch": 0.6520779304284459,
+      "grad_norm": 1.5859375,
+      "learning_rate": 1.7846064597676913e-05,
+      "loss": 1.0568,
+      "step": 3803
+    },
+    {
+      "epoch": 0.6522493945174358,
+      "grad_norm": 1.7265625,
+      "learning_rate": 1.7844944804896446e-05,
+      "loss": 0.978,
+      "step": 3804
+    },
+    {
+      "epoch": 0.6524208586064256,
+      "grad_norm": 1.53125,
+      "learning_rate": 1.784382475626279e-05,
+      "loss": 1.0221,
+      "step": 3805
+    },
+    {
+      "epoch": 0.6525923226954154,
+      "grad_norm": 1.6015625,
+      "learning_rate": 1.784270445181247e-05,
+      "loss": 1.0098,
+      "step": 3806
+    },
+    {
+      "epoch": 0.6527637867844054,
+      "grad_norm": 1.6015625,
+      "learning_rate": 1.784158389158202e-05,
+      "loss": 1.0093,
+      "step": 3807
+    },
+    {
+      "epoch": 0.6529352508733952,
+      "grad_norm": 1.6796875,
+      "learning_rate": 1.7840463075607996e-05,
+      "loss": 1.0862,
+      "step": 3808
+    },
+    {
+      "epoch": 0.653106714962385,
+      "grad_norm": 1.53125,
+      "learning_rate": 1.7839342003926945e-05,
+      "loss": 1.0648,
+      "step": 3809
+    },
+    {
+      "epoch": 0.653278179051375,
+      "grad_norm": 1.5234375,
+      "learning_rate": 1.7838220676575433e-05,
+      "loss": 1.0024,
+      "step": 3810
+    },
+    {
+      "epoch": 0.6534496431403648,
+      "grad_norm": 1.5859375,
+      "learning_rate": 1.783709909359003e-05,
+      "loss": 1.1038,
+      "step": 3811
+    },
+    {
+      "epoch": 0.6536211072293546,
+      "grad_norm": 1.5546875,
+      "learning_rate": 1.783597725500731e-05,
+      "loss": 0.9757,
+      "step": 3812
+    },
+    {
+      "epoch": 0.6537925713183446,
+      "grad_norm": 1.6796875,
+      "learning_rate": 1.7834855160863864e-05,
+      "loss": 1.0116,
+      "step": 3813
+    },
+    {
+      "epoch": 0.6539640354073344,
+      "grad_norm": 1.625,
+      "learning_rate": 1.783373281119629e-05,
+      "loss": 1.0175,
+      "step": 3814
+    },
+    {
+      "epoch": 0.6541354994963242,
+      "grad_norm": 1.625,
+      "learning_rate": 1.783261020604119e-05,
+      "loss": 0.991,
+      "step": 3815
+    },
+    {
+      "epoch": 0.6543069635853141,
+      "grad_norm": 1.7421875,
+      "learning_rate": 1.7831487345435175e-05,
+      "loss": 1.0568,
+      "step": 3816
+    },
+    {
+      "epoch": 0.654478427674304,
+      "grad_norm": 1.5625,
+      "learning_rate": 1.783036422941487e-05,
+      "loss": 0.976,
+      "step": 3817
+    },
+    {
+      "epoch": 0.6546498917632938,
+      "grad_norm": 1.5859375,
+      "learning_rate": 1.7829240858016896e-05,
+      "loss": 0.9601,
+      "step": 3818
+    },
+    {
+      "epoch": 0.6548213558522837,
+      "grad_norm": 1.6015625,
+      "learning_rate": 1.78281172312779e-05,
+      "loss": 1.0032,
+      "step": 3819
+    },
+    {
+      "epoch": 0.6549928199412736,
+      "grad_norm": 1.59375,
+      "learning_rate": 1.782699334923452e-05,
+      "loss": 1.0149,
+      "step": 3820
+    },
+    {
+      "epoch": 0.6551642840302634,
+      "grad_norm": 1.5234375,
+      "learning_rate": 1.7825869211923415e-05,
+      "loss": 0.9881,
+      "step": 3821
+    },
+    {
+      "epoch": 0.6553357481192533,
+      "grad_norm": 1.6640625,
+      "learning_rate": 1.7824744819381244e-05,
+      "loss": 1.096,
+      "step": 3822
+    },
+    {
+      "epoch": 0.6555072122082432,
+      "grad_norm": 1.6328125,
+      "learning_rate": 1.782362017164468e-05,
+      "loss": 1.0724,
+      "step": 3823
+    },
+    {
+      "epoch": 0.655678676297233,
+      "grad_norm": 1.6328125,
+      "learning_rate": 1.7822495268750402e-05,
+      "loss": 0.9867,
+      "step": 3824
+    },
+    {
+      "epoch": 0.6558501403862229,
+      "grad_norm": 1.640625,
+      "learning_rate": 1.7821370110735094e-05,
+      "loss": 1.0498,
+      "step": 3825
+    },
+    {
+      "epoch": 0.6560216044752127,
+      "grad_norm": 1.5625,
+      "learning_rate": 1.7820244697635458e-05,
+      "loss": 1.0758,
+      "step": 3826
+    },
+    {
+      "epoch": 0.6561930685642026,
+      "grad_norm": 1.6171875,
+      "learning_rate": 1.781911902948819e-05,
+      "loss": 1.0126,
+      "step": 3827
+    },
+    {
+      "epoch": 0.6563645326531925,
+      "grad_norm": 1.609375,
+      "learning_rate": 1.781799310633001e-05,
+      "loss": 1.1322,
+      "step": 3828
+    },
+    {
+      "epoch": 0.6565359967421823,
+      "grad_norm": 1.578125,
+      "learning_rate": 1.7816866928197632e-05,
+      "loss": 1.1316,
+      "step": 3829
+    },
+    {
+      "epoch": 0.6567074608311722,
+      "grad_norm": 1.5625,
+      "learning_rate": 1.781574049512779e-05,
+      "loss": 0.9376,
+      "step": 3830
+    },
+    {
+      "epoch": 0.6568789249201621,
+      "grad_norm": 1.609375,
+      "learning_rate": 1.781461380715722e-05,
+      "loss": 0.9902,
+      "step": 3831
+    },
+    {
+      "epoch": 0.6570503890091519,
+      "grad_norm": 1.609375,
+      "learning_rate": 1.781348686432266e-05,
+      "loss": 1.0493,
+      "step": 3832
+    },
+    {
+      "epoch": 0.6572218530981417,
+      "grad_norm": 1.515625,
+      "learning_rate": 1.781235966666088e-05,
+      "loss": 0.9787,
+      "step": 3833
+    },
+    {
+      "epoch": 0.6573933171871316,
+      "grad_norm": 1.6328125,
+      "learning_rate": 1.7811232214208626e-05,
+      "loss": 0.9926,
+      "step": 3834
+    },
+    {
+      "epoch": 0.6575647812761215,
+      "grad_norm": 1.65625,
+      "learning_rate": 1.7810104507002675e-05,
+      "loss": 1.0555,
+      "step": 3835
+    },
+    {
+      "epoch": 0.6577362453651113,
+      "grad_norm": 1.640625,
+      "learning_rate": 1.780897654507981e-05,
+      "loss": 0.9789,
+      "step": 3836
+    },
+    {
+      "epoch": 0.6579077094541012,
+      "grad_norm": 1.7109375,
+      "learning_rate": 1.7807848328476813e-05,
+      "loss": 1.0648,
+      "step": 3837
+    },
+    {
+      "epoch": 0.6580791735430911,
+      "grad_norm": 1.5859375,
+      "learning_rate": 1.780671985723048e-05,
+      "loss": 0.8998,
+      "step": 3838
+    },
+    {
+      "epoch": 0.6582506376320809,
+      "grad_norm": 3.859375,
+      "learning_rate": 1.7805591131377612e-05,
+      "loss": 1.0997,
+      "step": 3839
+    },
+    {
+      "epoch": 0.6584221017210707,
+      "grad_norm": 1.6875,
+      "learning_rate": 1.780446215095503e-05,
+      "loss": 1.0947,
+      "step": 3840
+    },
+    {
+      "epoch": 0.6585935658100607,
+      "grad_norm": 1.5390625,
+      "learning_rate": 1.7803332915999542e-05,
+      "loss": 1.0082,
+      "step": 3841
+    },
+    {
+      "epoch": 0.6587650298990505,
+      "grad_norm": 1.640625,
+      "learning_rate": 1.780220342654799e-05,
+      "loss": 1.0477,
+      "step": 3842
+    },
+    {
+      "epoch": 0.6589364939880403,
+      "grad_norm": 1.625,
+      "learning_rate": 1.78010736826372e-05,
+      "loss": 1.1199,
+      "step": 3843
+    },
+    {
+      "epoch": 0.6591079580770303,
+      "grad_norm": 1.515625,
+      "learning_rate": 1.7799943684304016e-05,
+      "loss": 0.9562,
+      "step": 3844
+    },
+    {
+      "epoch": 0.6592794221660201,
+      "grad_norm": 1.7421875,
+      "learning_rate": 1.77988134315853e-05,
+      "loss": 1.0783,
+      "step": 3845
+    },
+    {
+      "epoch": 0.6594508862550099,
+      "grad_norm": 1.4765625,
+      "learning_rate": 1.7797682924517917e-05,
+      "loss": 0.8665,
+      "step": 3846
+    },
+    {
+      "epoch": 0.6596223503439999,
+      "grad_norm": 1.6015625,
+      "learning_rate": 1.7796552163138722e-05,
+      "loss": 1.1352,
+      "step": 3847
+    },
+    {
+      "epoch": 0.6597938144329897,
+      "grad_norm": 1.5078125,
+      "learning_rate": 1.7795421147484608e-05,
+      "loss": 1.0046,
+      "step": 3848
+    },
+    {
+      "epoch": 0.6599652785219795,
+      "grad_norm": 1.515625,
+      "learning_rate": 1.7794289877592453e-05,
+      "loss": 1.0031,
+      "step": 3849
+    },
+    {
+      "epoch": 0.6601367426109694,
+      "grad_norm": 1.4609375,
+      "learning_rate": 1.779315835349915e-05,
+      "loss": 0.9268,
+      "step": 3850
+    },
+    {
+      "epoch": 0.6603082066999593,
+      "grad_norm": 1.625,
+      "learning_rate": 1.7792026575241616e-05,
+      "loss": 0.9385,
+      "step": 3851
+    },
+    {
+      "epoch": 0.6604796707889491,
+      "grad_norm": 1.6640625,
+      "learning_rate": 1.7790894542856748e-05,
+      "loss": 1.0262,
+      "step": 3852
+    },
+    {
+      "epoch": 0.660651134877939,
+      "grad_norm": 1.609375,
+      "learning_rate": 1.778976225638147e-05,
+      "loss": 1.0465,
+      "step": 3853
+    },
+    {
+      "epoch": 0.6608225989669289,
+      "grad_norm": 1.7265625,
+      "learning_rate": 1.778862971585271e-05,
+      "loss": 1.0516,
+      "step": 3854
+    },
+    {
+      "epoch": 0.6609940630559187,
+      "grad_norm": 1.6171875,
+      "learning_rate": 1.778749692130741e-05,
+      "loss": 0.9965,
+      "step": 3855
+    },
+    {
+      "epoch": 0.6611655271449086,
+      "grad_norm": 1.65625,
+      "learning_rate": 1.778636387278251e-05,
+      "loss": 1.0474,
+      "step": 3856
+    },
+    {
+      "epoch": 0.6613369912338984,
+      "grad_norm": 1.65625,
+      "learning_rate": 1.7785230570314963e-05,
+      "loss": 1.0767,
+      "step": 3857
+    },
+    {
+      "epoch": 0.6615084553228883,
+      "grad_norm": 1.5625,
+      "learning_rate": 1.7784097013941732e-05,
+      "loss": 1.0504,
+      "step": 3858
+    },
+    {
+      "epoch": 0.6616799194118782,
+      "grad_norm": 1.578125,
+      "learning_rate": 1.778296320369978e-05,
+      "loss": 1.009,
+      "step": 3859
+    },
+    {
+      "epoch": 0.661851383500868,
+      "grad_norm": 1.59375,
+      "learning_rate": 1.7781829139626096e-05,
+      "loss": 1.0706,
+      "step": 3860
+    },
+    {
+      "epoch": 0.6620228475898579,
+      "grad_norm": 1.65625,
+      "learning_rate": 1.7780694821757654e-05,
+      "loss": 0.9951,
+      "step": 3861
+    },
+    {
+      "epoch": 0.6621943116788478,
+      "grad_norm": 1.6171875,
+      "learning_rate": 1.7779560250131457e-05,
+      "loss": 1.0454,
+      "step": 3862
+    },
+    {
+      "epoch": 0.6623657757678376,
+      "grad_norm": 1.609375,
+      "learning_rate": 1.777842542478451e-05,
+      "loss": 1.1059,
+      "step": 3863
+    },
+    {
+      "epoch": 0.6625372398568274,
+      "grad_norm": 1.5390625,
+      "learning_rate": 1.7777290345753814e-05,
+      "loss": 1.0174,
+      "step": 3864
+    },
+    {
+      "epoch": 0.6627087039458174,
+      "grad_norm": 1.609375,
+      "learning_rate": 1.7776155013076394e-05,
+      "loss": 1.0829,
+      "step": 3865
+    },
+    {
+      "epoch": 0.6628801680348072,
+      "grad_norm": 1.515625,
+      "learning_rate": 1.7775019426789274e-05,
+      "loss": 0.997,
+      "step": 3866
+    },
+    {
+      "epoch": 0.663051632123797,
+      "grad_norm": 1.546875,
+      "learning_rate": 1.7773883586929498e-05,
+      "loss": 0.9961,
+      "step": 3867
+    },
+    {
+      "epoch": 0.663223096212787,
+      "grad_norm": 1.4765625,
+      "learning_rate": 1.77727474935341e-05,
+      "loss": 0.9763,
+      "step": 3868
+    },
+    {
+      "epoch": 0.6633945603017768,
+      "grad_norm": 1.65625,
+      "learning_rate": 1.777161114664014e-05,
+      "loss": 1.0266,
+      "step": 3869
+    },
+    {
+      "epoch": 0.6635660243907666,
+      "grad_norm": 1.6171875,
+      "learning_rate": 1.7770474546284674e-05,
+      "loss": 1.0637,
+      "step": 3870
+    },
+    {
+      "epoch": 0.6637374884797566,
+      "grad_norm": 1.6171875,
+      "learning_rate": 1.776933769250477e-05,
+      "loss": 1.0638,
+      "step": 3871
+    },
+    {
+      "epoch": 0.6639089525687464,
+      "grad_norm": 1.5625,
+      "learning_rate": 1.776820058533751e-05,
+      "loss": 1.0038,
+      "step": 3872
+    },
+    {
+      "epoch": 0.6640804166577362,
+      "grad_norm": 1.6171875,
+      "learning_rate": 1.7767063224819976e-05,
+      "loss": 1.0904,
+      "step": 3873
+    },
+    {
+      "epoch": 0.6642518807467261,
+      "grad_norm": 1.6640625,
+      "learning_rate": 1.7765925610989263e-05,
+      "loss": 0.968,
+      "step": 3874
+    },
+    {
+      "epoch": 0.664423344835716,
+      "grad_norm": 1.6953125,
+      "learning_rate": 1.776478774388247e-05,
+      "loss": 1.1252,
+      "step": 3875
+    },
+    {
+      "epoch": 0.6645948089247058,
+      "grad_norm": 1.5546875,
+      "learning_rate": 1.7763649623536712e-05,
+      "loss": 1.0071,
+      "step": 3876
+    },
+    {
+      "epoch": 0.6647662730136957,
+      "grad_norm": 1.6640625,
+      "learning_rate": 1.7762511249989104e-05,
+      "loss": 1.214,
+      "step": 3877
+    },
+    {
+      "epoch": 0.6649377371026856,
+      "grad_norm": 1.65625,
+      "learning_rate": 1.776137262327677e-05,
+      "loss": 1.0238,
+      "step": 3878
+    },
+    {
+      "epoch": 0.6651092011916754,
+      "grad_norm": 1.5390625,
+      "learning_rate": 1.776023374343685e-05,
+      "loss": 0.9687,
+      "step": 3879
+    },
+    {
+      "epoch": 0.6652806652806653,
+      "grad_norm": 1.5859375,
+      "learning_rate": 1.7759094610506486e-05,
+      "loss": 0.9881,
+      "step": 3880
+    },
+    {
+      "epoch": 0.6654521293696551,
+      "grad_norm": 1.46875,
+      "learning_rate": 1.775795522452283e-05,
+      "loss": 1.0175,
+      "step": 3881
+    },
+    {
+      "epoch": 0.665623593458645,
+      "grad_norm": 1.578125,
+      "learning_rate": 1.7756815585523038e-05,
+      "loss": 1.0085,
+      "step": 3882
+    },
+    {
+      "epoch": 0.6657950575476349,
+      "grad_norm": 1.578125,
+      "learning_rate": 1.7755675693544277e-05,
+      "loss": 0.9782,
+      "step": 3883
+    },
+    {
+      "epoch": 0.6659665216366247,
+      "grad_norm": 1.578125,
+      "learning_rate": 1.775453554862373e-05,
+      "loss": 0.9616,
+      "step": 3884
+    },
+    {
+      "epoch": 0.6661379857256146,
+      "grad_norm": 1.6796875,
+      "learning_rate": 1.775339515079858e-05,
+      "loss": 1.0437,
+      "step": 3885
+    },
+    {
+      "epoch": 0.6663094498146045,
+      "grad_norm": 1.65625,
+      "learning_rate": 1.7752254500106016e-05,
+      "loss": 1.146,
+      "step": 3886
+    },
+    {
+      "epoch": 0.6664809139035943,
+      "grad_norm": 1.7265625,
+      "learning_rate": 1.775111359658324e-05,
+      "loss": 1.0646,
+      "step": 3887
+    },
+    {
+      "epoch": 0.6666523779925841,
+      "grad_norm": 1.640625,
+      "learning_rate": 1.7749972440267463e-05,
+      "loss": 1.0783,
+      "step": 3888
+    },
+    {
+      "epoch": 0.6668238420815741,
+      "grad_norm": 1.609375,
+      "learning_rate": 1.7748831031195898e-05,
+      "loss": 1.0681,
+      "step": 3889
+    },
+    {
+      "epoch": 0.6669953061705639,
+      "grad_norm": 1.609375,
+      "learning_rate": 1.774768936940578e-05,
+      "loss": 1.0427,
+      "step": 3890
+    },
+    {
+      "epoch": 0.6671667702595537,
+      "grad_norm": 1.609375,
+      "learning_rate": 1.7746547454934327e-05,
+      "loss": 1.064,
+      "step": 3891
+    },
+    {
+      "epoch": 0.6673382343485437,
+      "grad_norm": 1.578125,
+      "learning_rate": 1.77454052878188e-05,
+      "loss": 0.9948,
+      "step": 3892
+    },
+    {
+      "epoch": 0.6675096984375335,
+      "grad_norm": 1.6640625,
+      "learning_rate": 1.7744262868096432e-05,
+      "loss": 1.0956,
+      "step": 3893
+    },
+    {
+      "epoch": 0.6676811625265233,
+      "grad_norm": 1.609375,
+      "learning_rate": 1.7743120195804498e-05,
+      "loss": 0.9624,
+      "step": 3894
+    },
+    {
+      "epoch": 0.6678526266155133,
+      "grad_norm": 1.7890625,
+      "learning_rate": 1.774197727098025e-05,
+      "loss": 1.0827,
+      "step": 3895
+    },
+    {
+      "epoch": 0.6680240907045031,
+      "grad_norm": 1.5859375,
+      "learning_rate": 1.7740834093660974e-05,
+      "loss": 0.8746,
+      "step": 3896
+    },
+    {
+      "epoch": 0.6681955547934929,
+      "grad_norm": 1.5703125,
+      "learning_rate": 1.7739690663883948e-05,
+      "loss": 1.0361,
+      "step": 3897
+    },
+    {
+      "epoch": 0.6683670188824828,
+      "grad_norm": 1.6484375,
+      "learning_rate": 1.7738546981686458e-05,
+      "loss": 0.9436,
+      "step": 3898
+    },
+    {
+      "epoch": 0.6685384829714727,
+      "grad_norm": 1.46875,
+      "learning_rate": 1.773740304710582e-05,
+      "loss": 0.9772,
+      "step": 3899
+    },
+    {
+      "epoch": 0.6687099470604625,
+      "grad_norm": 1.546875,
+      "learning_rate": 1.7736258860179326e-05,
+      "loss": 1.015,
+      "step": 3900
+    },
+    {
+      "epoch": 0.6688814111494524,
+      "grad_norm": 1.59375,
+      "learning_rate": 1.77351144209443e-05,
+      "loss": 0.9838,
+      "step": 3901
+    },
+    {
+      "epoch": 0.6690528752384423,
+      "grad_norm": 1.65625,
+      "learning_rate": 1.7733969729438064e-05,
+      "loss": 1.0158,
+      "step": 3902
+    },
+    {
+      "epoch": 0.6692243393274321,
+      "grad_norm": 1.5234375,
+      "learning_rate": 1.7732824785697956e-05,
+      "loss": 0.9799,
+      "step": 3903
+    },
+    {
+      "epoch": 0.669395803416422,
+      "grad_norm": 1.625,
+      "learning_rate": 1.7731679589761307e-05,
+      "loss": 0.9674,
+      "step": 3904
+    },
+    {
+      "epoch": 0.6695672675054118,
+      "grad_norm": 1.625,
+      "learning_rate": 1.7730534141665473e-05,
+      "loss": 1.0857,
+      "step": 3905
+    },
+    {
+      "epoch": 0.6697387315944017,
+      "grad_norm": 1.5546875,
+      "learning_rate": 1.7729388441447813e-05,
+      "loss": 1.0258,
+      "step": 3906
+    },
+    {
+      "epoch": 0.6699101956833916,
+      "grad_norm": 1.5625,
+      "learning_rate": 1.7728242489145687e-05,
+      "loss": 1.0592,
+      "step": 3907
+    },
+    {
+      "epoch": 0.6700816597723814,
+      "grad_norm": 1.5546875,
+      "learning_rate": 1.7727096284796476e-05,
+      "loss": 1.0646,
+      "step": 3908
+    },
+    {
+      "epoch": 0.6702531238613713,
+      "grad_norm": 1.5390625,
+      "learning_rate": 1.7725949828437552e-05,
+      "loss": 0.9314,
+      "step": 3909
+    },
+    {
+      "epoch": 0.6704245879503612,
+      "grad_norm": 1.5078125,
+      "learning_rate": 1.7724803120106312e-05,
+      "loss": 1.0248,
+      "step": 3910
+    },
+    {
+      "epoch": 0.670596052039351,
+      "grad_norm": 1.515625,
+      "learning_rate": 1.7723656159840156e-05,
+      "loss": 0.9268,
+      "step": 3911
+    },
+    {
+      "epoch": 0.6707675161283408,
+      "grad_norm": 1.5,
+      "learning_rate": 1.7722508947676488e-05,
+      "loss": 0.9512,
+      "step": 3912
+    },
+    {
+      "epoch": 0.6709389802173308,
+      "grad_norm": 1.5859375,
+      "learning_rate": 1.772136148365272e-05,
+      "loss": 0.9944,
+      "step": 3913
+    },
+    {
+      "epoch": 0.6711104443063206,
+      "grad_norm": 1.5390625,
+      "learning_rate": 1.772021376780628e-05,
+      "loss": 1.0312,
+      "step": 3914
+    },
+    {
+      "epoch": 0.6712819083953104,
+      "grad_norm": 1.6875,
+      "learning_rate": 1.7719065800174595e-05,
+      "loss": 1.0387,
+      "step": 3915
+    },
+    {
+      "epoch": 0.6714533724843004,
+      "grad_norm": 1.578125,
+      "learning_rate": 1.7717917580795108e-05,
+      "loss": 1.0055,
+      "step": 3916
+    },
+    {
+      "epoch": 0.6716248365732902,
+      "grad_norm": 1.515625,
+      "learning_rate": 1.771676910970527e-05,
+      "loss": 1.0512,
+      "step": 3917
+    },
+    {
+      "epoch": 0.67179630066228,
+      "grad_norm": 1.5390625,
+      "learning_rate": 1.7715620386942528e-05,
+      "loss": 1.0302,
+      "step": 3918
+    },
+    {
+      "epoch": 0.67196776475127,
+      "grad_norm": 1.578125,
+      "learning_rate": 1.7714471412544353e-05,
+      "loss": 1.0381,
+      "step": 3919
+    },
+    {
+      "epoch": 0.6721392288402598,
+      "grad_norm": 1.59375,
+      "learning_rate": 1.771332218654821e-05,
+      "loss": 1.0399,
+      "step": 3920
+    },
+    {
+      "epoch": 0.6723106929292496,
+      "grad_norm": 1.4921875,
+      "learning_rate": 1.7712172708991594e-05,
+      "loss": 1.0186,
+      "step": 3921
+    },
+    {
+      "epoch": 0.6724821570182395,
+      "grad_norm": 1.5546875,
+      "learning_rate": 1.7711022979911977e-05,
+      "loss": 1.0577,
+      "step": 3922
+    },
+    {
+      "epoch": 0.6726536211072294,
+      "grad_norm": 1.578125,
+      "learning_rate": 1.7709872999346867e-05,
+      "loss": 1.1428,
+      "step": 3923
+    },
+    {
+      "epoch": 0.6728250851962192,
+      "grad_norm": 1.625,
+      "learning_rate": 1.7708722767333766e-05,
+      "loss": 0.9536,
+      "step": 3924
+    },
+    {
+      "epoch": 0.6729965492852091,
+      "grad_norm": 1.6953125,
+      "learning_rate": 1.7707572283910184e-05,
+      "loss": 1.112,
+      "step": 3925
+    },
+    {
+      "epoch": 0.673168013374199,
+      "grad_norm": 1.4921875,
+      "learning_rate": 1.770642154911365e-05,
+      "loss": 0.9273,
+      "step": 3926
+    },
+    {
+      "epoch": 0.6733394774631888,
+      "grad_norm": 1.5,
+      "learning_rate": 1.7705270562981688e-05,
+      "loss": 1.0,
+      "step": 3927
+    },
+    {
+      "epoch": 0.6735109415521786,
+      "grad_norm": 1.5859375,
+      "learning_rate": 1.7704119325551835e-05,
+      "loss": 0.9893,
+      "step": 3928
+    },
+    {
+      "epoch": 0.6736824056411685,
+      "grad_norm": 1.609375,
+      "learning_rate": 1.7702967836861643e-05,
+      "loss": 1.0211,
+      "step": 3929
+    },
+    {
+      "epoch": 0.6738538697301584,
+      "grad_norm": 1.5625,
+      "learning_rate": 1.7701816096948665e-05,
+      "loss": 1.0401,
+      "step": 3930
+    },
+    {
+      "epoch": 0.6740253338191482,
+      "grad_norm": 1.546875,
+      "learning_rate": 1.7700664105850454e-05,
+      "loss": 1.0329,
+      "step": 3931
+    },
+    {
+      "epoch": 0.6741967979081381,
+      "grad_norm": 1.625,
+      "learning_rate": 1.7699511863604597e-05,
+      "loss": 1.0191,
+      "step": 3932
+    },
+    {
+      "epoch": 0.674368261997128,
+      "grad_norm": 1.671875,
+      "learning_rate": 1.769835937024866e-05,
+      "loss": 0.9552,
+      "step": 3933
+    },
+    {
+      "epoch": 0.6745397260861178,
+      "grad_norm": 1.59375,
+      "learning_rate": 1.7697206625820237e-05,
+      "loss": 1.0417,
+      "step": 3934
+    },
+    {
+      "epoch": 0.6747111901751077,
+      "grad_norm": 1.5703125,
+      "learning_rate": 1.7696053630356918e-05,
+      "loss": 1.0664,
+      "step": 3935
+    },
+    {
+      "epoch": 0.6748826542640975,
+      "grad_norm": 1.640625,
+      "learning_rate": 1.769490038389631e-05,
+      "loss": 1.1068,
+      "step": 3936
+    },
+    {
+      "epoch": 0.6750541183530874,
+      "grad_norm": 1.5234375,
+      "learning_rate": 1.7693746886476028e-05,
+      "loss": 1.0836,
+      "step": 3937
+    },
+    {
+      "epoch": 0.6752255824420773,
+      "grad_norm": 1.671875,
+      "learning_rate": 1.7692593138133684e-05,
+      "loss": 0.974,
+      "step": 3938
+    },
+    {
+      "epoch": 0.6753970465310671,
+      "grad_norm": 1.59375,
+      "learning_rate": 1.769143913890691e-05,
+      "loss": 1.0171,
+      "step": 3939
+    },
+    {
+      "epoch": 0.675568510620057,
+      "grad_norm": 1.5546875,
+      "learning_rate": 1.7690284888833344e-05,
+      "loss": 1.1185,
+      "step": 3940
+    },
+    {
+      "epoch": 0.6757399747090469,
+      "grad_norm": 1.515625,
+      "learning_rate": 1.768913038795063e-05,
+      "loss": 1.0596,
+      "step": 3941
+    },
+    {
+      "epoch": 0.6759114387980367,
+      "grad_norm": 1.4765625,
+      "learning_rate": 1.7687975636296414e-05,
+      "loss": 1.0241,
+      "step": 3942
+    },
+    {
+      "epoch": 0.6760829028870265,
+      "grad_norm": 1.625,
+      "learning_rate": 1.7686820633908368e-05,
+      "loss": 1.1086,
+      "step": 3943
+    },
+    {
+      "epoch": 0.6762543669760165,
+      "grad_norm": 1.53125,
+      "learning_rate": 1.7685665380824156e-05,
+      "loss": 0.9996,
+      "step": 3944
+    },
+    {
+      "epoch": 0.6764258310650063,
+      "grad_norm": 1.5234375,
+      "learning_rate": 1.768450987708145e-05,
+      "loss": 0.9955,
+      "step": 3945
+    },
+    {
+      "epoch": 0.6765972951539961,
+      "grad_norm": 1.4375,
+      "learning_rate": 1.7683354122717942e-05,
+      "loss": 0.937,
+      "step": 3946
+    },
+    {
+      "epoch": 0.6767687592429861,
+      "grad_norm": 1.546875,
+      "learning_rate": 1.768219811777132e-05,
+      "loss": 1.054,
+      "step": 3947
+    },
+    {
+      "epoch": 0.6769402233319759,
+      "grad_norm": 1.625,
+      "learning_rate": 1.7681041862279294e-05,
+      "loss": 0.9462,
+      "step": 3948
+    },
+    {
+      "epoch": 0.6771116874209657,
+      "grad_norm": 1.546875,
+      "learning_rate": 1.7679885356279566e-05,
+      "loss": 0.97,
+      "step": 3949
+    },
+    {
+      "epoch": 0.6772831515099557,
+      "grad_norm": 1.5390625,
+      "learning_rate": 1.7678728599809858e-05,
+      "loss": 0.9661,
+      "step": 3950
+    },
+    {
+      "epoch": 0.6774546155989455,
+      "grad_norm": 1.5703125,
+      "learning_rate": 1.7677571592907893e-05,
+      "loss": 1.0832,
+      "step": 3951
+    },
+    {
+      "epoch": 0.6776260796879353,
+      "grad_norm": 1.546875,
+      "learning_rate": 1.767641433561141e-05,
+      "loss": 1.0797,
+      "step": 3952
+    },
+    {
+      "epoch": 0.6777975437769252,
+      "grad_norm": 1.6484375,
+      "learning_rate": 1.7675256827958148e-05,
+      "loss": 1.0648,
+      "step": 3953
+    },
+    {
+      "epoch": 0.6779690078659151,
+      "grad_norm": 1.5,
+      "learning_rate": 1.7674099069985855e-05,
+      "loss": 0.9913,
+      "step": 3954
+    },
+    {
+      "epoch": 0.6781404719549049,
+      "grad_norm": 1.5703125,
+      "learning_rate": 1.7672941061732297e-05,
+      "loss": 1.0111,
+      "step": 3955
+    },
+    {
+      "epoch": 0.6783119360438948,
+      "grad_norm": 1.5546875,
+      "learning_rate": 1.7671782803235235e-05,
+      "loss": 0.9879,
+      "step": 3956
+    },
+    {
+      "epoch": 0.6784834001328847,
+      "grad_norm": 1.625,
+      "learning_rate": 1.767062429453245e-05,
+      "loss": 1.0341,
+      "step": 3957
+    },
+    {
+      "epoch": 0.6786548642218745,
+      "grad_norm": 1.5703125,
+      "learning_rate": 1.766946553566172e-05,
+      "loss": 1.0252,
+      "step": 3958
+    },
+    {
+      "epoch": 0.6788263283108644,
+      "grad_norm": 1.5703125,
+      "learning_rate": 1.7668306526660836e-05,
+      "loss": 1.0166,
+      "step": 3959
+    },
+    {
+      "epoch": 0.6789977923998542,
+      "grad_norm": 1.53125,
+      "learning_rate": 1.7667147267567602e-05,
+      "loss": 0.9378,
+      "step": 3960
+    },
+    {
+      "epoch": 0.6791692564888441,
+      "grad_norm": 1.5859375,
+      "learning_rate": 1.7665987758419824e-05,
+      "loss": 1.0179,
+      "step": 3961
+    },
+    {
+      "epoch": 0.679340720577834,
+      "grad_norm": 1.578125,
+      "learning_rate": 1.766482799925532e-05,
+      "loss": 1.0277,
+      "step": 3962
+    },
+    {
+      "epoch": 0.6795121846668238,
+      "grad_norm": 1.5234375,
+      "learning_rate": 1.7663667990111908e-05,
+      "loss": 1.0449,
+      "step": 3963
+    },
+    {
+      "epoch": 0.6796836487558137,
+      "grad_norm": 1.7109375,
+      "learning_rate": 1.7662507731027426e-05,
+      "loss": 1.0957,
+      "step": 3964
+    },
+    {
+      "epoch": 0.6798551128448036,
+      "grad_norm": 1.6953125,
+      "learning_rate": 1.7661347222039714e-05,
+      "loss": 1.0475,
+      "step": 3965
+    },
+    {
+      "epoch": 0.6800265769337934,
+      "grad_norm": 1.5703125,
+      "learning_rate": 1.7660186463186617e-05,
+      "loss": 0.8924,
+      "step": 3966
+    },
+    {
+      "epoch": 0.6801980410227833,
+      "grad_norm": 1.5859375,
+      "learning_rate": 1.7659025454505994e-05,
+      "loss": 1.0355,
+      "step": 3967
+    },
+    {
+      "epoch": 0.6803695051117732,
+      "grad_norm": 1.5234375,
+      "learning_rate": 1.765786419603571e-05,
+      "loss": 0.9575,
+      "step": 3968
+    },
+    {
+      "epoch": 0.680540969200763,
+      "grad_norm": 1.7109375,
+      "learning_rate": 1.765670268781364e-05,
+      "loss": 1.0357,
+      "step": 3969
+    },
+    {
+      "epoch": 0.6807124332897528,
+      "grad_norm": 1.5703125,
+      "learning_rate": 1.765554092987766e-05,
+      "loss": 0.9695,
+      "step": 3970
+    },
+    {
+      "epoch": 0.6808838973787428,
+      "grad_norm": 1.5546875,
+      "learning_rate": 1.765437892226566e-05,
+      "loss": 1.1055,
+      "step": 3971
+    },
+    {
+      "epoch": 0.6810553614677326,
+      "grad_norm": 1.5546875,
+      "learning_rate": 1.765321666501554e-05,
+      "loss": 0.951,
+      "step": 3972
+    },
+    {
+      "epoch": 0.6812268255567224,
+      "grad_norm": 1.515625,
+      "learning_rate": 1.765205415816521e-05,
+      "loss": 0.9672,
+      "step": 3973
+    },
+    {
+      "epoch": 0.6813982896457124,
+      "grad_norm": 1.6015625,
+      "learning_rate": 1.7650891401752578e-05,
+      "loss": 1.0185,
+      "step": 3974
+    },
+    {
+      "epoch": 0.6815697537347022,
+      "grad_norm": 1.6953125,
+      "learning_rate": 1.7649728395815567e-05,
+      "loss": 1.0638,
+      "step": 3975
+    },
+    {
+      "epoch": 0.681741217823692,
+      "grad_norm": 1.4609375,
+      "learning_rate": 1.7648565140392103e-05,
+      "loss": 0.9564,
+      "step": 3976
+    },
+    {
+      "epoch": 0.681912681912682,
+      "grad_norm": 1.515625,
+      "learning_rate": 1.764740163552013e-05,
+      "loss": 1.0024,
+      "step": 3977
+    },
+    {
+      "epoch": 0.6820841460016718,
+      "grad_norm": 1.53125,
+      "learning_rate": 1.764623788123759e-05,
+      "loss": 1.0701,
+      "step": 3978
+    },
+    {
+      "epoch": 0.6822556100906616,
+      "grad_norm": 1.515625,
+      "learning_rate": 1.7645073877582445e-05,
+      "loss": 1.0215,
+      "step": 3979
+    },
+    {
+      "epoch": 0.6824270741796515,
+      "grad_norm": 1.5703125,
+      "learning_rate": 1.764390962459265e-05,
+      "loss": 1.0535,
+      "step": 3980
+    },
+    {
+      "epoch": 0.6825985382686414,
+      "grad_norm": 1.5859375,
+      "learning_rate": 1.7642745122306177e-05,
+      "loss": 0.9549,
+      "step": 3981
+    },
+    {
+      "epoch": 0.6827700023576312,
+      "grad_norm": 1.5234375,
+      "learning_rate": 1.7641580370761e-05,
+      "loss": 0.9616,
+      "step": 3982
+    },
+    {
+      "epoch": 0.6829414664466211,
+      "grad_norm": 1.6171875,
+      "learning_rate": 1.764041536999512e-05,
+      "loss": 1.0469,
+      "step": 3983
+    },
+    {
+      "epoch": 0.683112930535611,
+      "grad_norm": 1.5703125,
+      "learning_rate": 1.763925012004652e-05,
+      "loss": 1.0718,
+      "step": 3984
+    },
+    {
+      "epoch": 0.6832843946246008,
+      "grad_norm": 1.5859375,
+      "learning_rate": 1.763808462095321e-05,
+      "loss": 1.0762,
+      "step": 3985
+    },
+    {
+      "epoch": 0.6834558587135907,
+      "grad_norm": 1.484375,
+      "learning_rate": 1.7636918872753194e-05,
+      "loss": 0.9343,
+      "step": 3986
+    },
+    {
+      "epoch": 0.6836273228025805,
+      "grad_norm": 1.6171875,
+      "learning_rate": 1.76357528754845e-05,
+      "loss": 1.0311,
+      "step": 3987
+    },
+    {
+      "epoch": 0.6837987868915704,
+      "grad_norm": 1.59375,
+      "learning_rate": 1.7634586629185147e-05,
+      "loss": 1.0325,
+      "step": 3988
+    },
+    {
+      "epoch": 0.6839702509805603,
+      "grad_norm": 1.53125,
+      "learning_rate": 1.7633420133893176e-05,
+      "loss": 0.9296,
+      "step": 3989
+    },
+    {
+      "epoch": 0.6841417150695501,
+      "grad_norm": 1.6171875,
+      "learning_rate": 1.763225338964663e-05,
+      "loss": 1.0297,
+      "step": 3990
+    },
+    {
+      "epoch": 0.68431317915854,
+      "grad_norm": 1.8046875,
+      "learning_rate": 1.7631086396483562e-05,
+      "loss": 1.0326,
+      "step": 3991
+    },
+    {
+      "epoch": 0.6844846432475299,
+      "grad_norm": 1.5703125,
+      "learning_rate": 1.762991915444203e-05,
+      "loss": 0.9755,
+      "step": 3992
+    },
+    {
+      "epoch": 0.6846561073365197,
+      "grad_norm": 1.515625,
+      "learning_rate": 1.76287516635601e-05,
+      "loss": 0.9867,
+      "step": 3993
+    },
+    {
+      "epoch": 0.6848275714255095,
+      "grad_norm": 1.59375,
+      "learning_rate": 1.762758392387585e-05,
+      "loss": 1.0919,
+      "step": 3994
+    },
+    {
+      "epoch": 0.6849990355144995,
+      "grad_norm": 1.484375,
+      "learning_rate": 1.7626415935427373e-05,
+      "loss": 0.9688,
+      "step": 3995
+    },
+    {
+      "epoch": 0.6851704996034893,
+      "grad_norm": 1.5625,
+      "learning_rate": 1.762524769825275e-05,
+      "loss": 1.0599,
+      "step": 3996
+    },
+    {
+      "epoch": 0.6853419636924791,
+      "grad_norm": 1.6328125,
+      "learning_rate": 1.762407921239008e-05,
+      "loss": 1.0304,
+      "step": 3997
+    },
+    {
+      "epoch": 0.6855134277814691,
+      "grad_norm": 1.6484375,
+      "learning_rate": 1.7622910477877484e-05,
+      "loss": 1.0556,
+      "step": 3998
+    },
+    {
+      "epoch": 0.6856848918704589,
+      "grad_norm": 1.6015625,
+      "learning_rate": 1.762174149475307e-05,
+      "loss": 1.007,
+      "step": 3999
+    },
+    {
+      "epoch": 0.6858563559594487,
+      "grad_norm": 1.578125,
+      "learning_rate": 1.7620572263054964e-05,
+      "loss": 0.9835,
+      "step": 4000
+    },
+    {
+      "epoch": 0.6860278200484387,
+      "grad_norm": 1.609375,
+      "learning_rate": 1.7619402782821306e-05,
+      "loss": 0.9957,
+      "step": 4001
+    },
+    {
+      "epoch": 0.6861992841374285,
+      "grad_norm": 1.59375,
+      "learning_rate": 1.7618233054090223e-05,
+      "loss": 1.0368,
+      "step": 4002
+    },
+    {
+      "epoch": 0.6863707482264183,
+      "grad_norm": 1.5546875,
+      "learning_rate": 1.7617063076899875e-05,
+      "loss": 1.0038,
+      "step": 4003
+    },
+    {
+      "epoch": 0.6865422123154082,
+      "grad_norm": 1.6484375,
+      "learning_rate": 1.7615892851288417e-05,
+      "loss": 1.0597,
+      "step": 4004
+    },
+    {
+      "epoch": 0.6867136764043981,
+      "grad_norm": 1.59375,
+      "learning_rate": 1.7614722377294017e-05,
+      "loss": 0.9894,
+      "step": 4005
+    },
+    {
+      "epoch": 0.6868851404933879,
+      "grad_norm": 1.5234375,
+      "learning_rate": 1.7613551654954846e-05,
+      "loss": 1.0048,
+      "step": 4006
+    },
+    {
+      "epoch": 0.6870566045823778,
+      "grad_norm": 1.7109375,
+      "learning_rate": 1.761238068430908e-05,
+      "loss": 1.1029,
+      "step": 4007
+    },
+    {
+      "epoch": 0.6872280686713677,
+      "grad_norm": 1.734375,
+      "learning_rate": 1.761120946539492e-05,
+      "loss": 1.1666,
+      "step": 4008
+    },
+    {
+      "epoch": 0.6873995327603575,
+      "grad_norm": 1.5625,
+      "learning_rate": 1.7610037998250555e-05,
+      "loss": 1.0329,
+      "step": 4009
+    },
+    {
+      "epoch": 0.6875709968493474,
+      "grad_norm": 1.5703125,
+      "learning_rate": 1.7608866282914195e-05,
+      "loss": 0.9206,
+      "step": 4010
+    },
+    {
+      "epoch": 0.6877424609383372,
+      "grad_norm": 1.484375,
+      "learning_rate": 1.7607694319424054e-05,
+      "loss": 0.9154,
+      "step": 4011
+    },
+    {
+      "epoch": 0.6879139250273271,
+      "grad_norm": 1.53125,
+      "learning_rate": 1.7606522107818355e-05,
+      "loss": 1.0722,
+      "step": 4012
+    },
+    {
+      "epoch": 0.688085389116317,
+      "grad_norm": 1.6328125,
+      "learning_rate": 1.760534964813532e-05,
+      "loss": 1.0418,
+      "step": 4013
+    },
+    {
+      "epoch": 0.6882568532053068,
+      "grad_norm": 1.5859375,
+      "learning_rate": 1.76041769404132e-05,
+      "loss": 0.9458,
+      "step": 4014
+    },
+    {
+      "epoch": 0.6884283172942967,
+      "grad_norm": 1.6953125,
+      "learning_rate": 1.7603003984690234e-05,
+      "loss": 1.0555,
+      "step": 4015
+    },
+    {
+      "epoch": 0.6885997813832866,
+      "grad_norm": 1.546875,
+      "learning_rate": 1.7601830781004676e-05,
+      "loss": 1.0789,
+      "step": 4016
+    },
+    {
+      "epoch": 0.6887712454722764,
+      "grad_norm": 1.5703125,
+      "learning_rate": 1.7600657329394794e-05,
+      "loss": 0.9962,
+      "step": 4017
+    },
+    {
+      "epoch": 0.6889427095612662,
+      "grad_norm": 1.6328125,
+      "learning_rate": 1.759948362989885e-05,
+      "loss": 1.0139,
+      "step": 4018
+    },
+    {
+      "epoch": 0.6891141736502562,
+      "grad_norm": 1.6171875,
+      "learning_rate": 1.7598309682555133e-05,
+      "loss": 1.0394,
+      "step": 4019
+    },
+    {
+      "epoch": 0.689285637739246,
+      "grad_norm": 1.515625,
+      "learning_rate": 1.759713548740192e-05,
+      "loss": 0.9697,
+      "step": 4020
+    },
+    {
+      "epoch": 0.6894571018282358,
+      "grad_norm": 1.5859375,
+      "learning_rate": 1.7595961044477516e-05,
+      "loss": 1.0688,
+      "step": 4021
+    },
+    {
+      "epoch": 0.6896285659172258,
+      "grad_norm": 1.6171875,
+      "learning_rate": 1.7594786353820215e-05,
+      "loss": 0.9847,
+      "step": 4022
+    },
+    {
+      "epoch": 0.6898000300062156,
+      "grad_norm": 1.6875,
+      "learning_rate": 1.759361141546833e-05,
+      "loss": 1.0528,
+      "step": 4023
+    },
+    {
+      "epoch": 0.6899714940952054,
+      "grad_norm": 1.578125,
+      "learning_rate": 1.7592436229460185e-05,
+      "loss": 1.0234,
+      "step": 4024
+    },
+    {
+      "epoch": 0.6901429581841952,
+      "grad_norm": 1.6171875,
+      "learning_rate": 1.7591260795834104e-05,
+      "loss": 1.1078,
+      "step": 4025
+    },
+    {
+      "epoch": 0.6903144222731852,
+      "grad_norm": 1.6796875,
+      "learning_rate": 1.7590085114628422e-05,
+      "loss": 1.1082,
+      "step": 4026
+    },
+    {
+      "epoch": 0.690485886362175,
+      "grad_norm": 1.53125,
+      "learning_rate": 1.758890918588148e-05,
+      "loss": 1.0392,
+      "step": 4027
+    },
+    {
+      "epoch": 0.6906573504511648,
+      "grad_norm": 1.5859375,
+      "learning_rate": 1.7587733009631637e-05,
+      "loss": 0.9816,
+      "step": 4028
+    },
+    {
+      "epoch": 0.6908288145401548,
+      "grad_norm": 1.5390625,
+      "learning_rate": 1.758655658591724e-05,
+      "loss": 1.0778,
+      "step": 4029
+    },
+    {
+      "epoch": 0.6910002786291446,
+      "grad_norm": 1.5078125,
+      "learning_rate": 1.7585379914776672e-05,
+      "loss": 1.0351,
+      "step": 4030
+    },
+    {
+      "epoch": 0.6911717427181344,
+      "grad_norm": 1.609375,
+      "learning_rate": 1.75842029962483e-05,
+      "loss": 1.0163,
+      "step": 4031
+    },
+    {
+      "epoch": 0.6913432068071244,
+      "grad_norm": 1.46875,
+      "learning_rate": 1.7583025830370507e-05,
+      "loss": 0.9156,
+      "step": 4032
+    },
+    {
+      "epoch": 0.6915146708961142,
+      "grad_norm": 1.8203125,
+      "learning_rate": 1.7581848417181687e-05,
+      "loss": 1.0033,
+      "step": 4033
+    },
+    {
+      "epoch": 0.691686134985104,
+      "grad_norm": 1.59375,
+      "learning_rate": 1.758067075672024e-05,
+      "loss": 1.0246,
+      "step": 4034
+    },
+    {
+      "epoch": 0.6918575990740939,
+      "grad_norm": 1.5234375,
+      "learning_rate": 1.7579492849024574e-05,
+      "loss": 0.9565,
+      "step": 4035
+    },
+    {
+      "epoch": 0.6920290631630838,
+      "grad_norm": 1.546875,
+      "learning_rate": 1.7578314694133105e-05,
+      "loss": 1.0043,
+      "step": 4036
+    },
+    {
+      "epoch": 0.6922005272520736,
+      "grad_norm": 1.5390625,
+      "learning_rate": 1.7577136292084255e-05,
+      "loss": 0.931,
+      "step": 4037
+    },
+    {
+      "epoch": 0.6923719913410635,
+      "grad_norm": 1.6015625,
+      "learning_rate": 1.7575957642916453e-05,
+      "loss": 1.0022,
+      "step": 4038
+    },
+    {
+      "epoch": 0.6925434554300534,
+      "grad_norm": 1.6015625,
+      "learning_rate": 1.7574778746668152e-05,
+      "loss": 0.9138,
+      "step": 4039
+    },
+    {
+      "epoch": 0.6927149195190432,
+      "grad_norm": 1.5390625,
+      "learning_rate": 1.757359960337779e-05,
+      "loss": 0.957,
+      "step": 4040
+    },
+    {
+      "epoch": 0.6928863836080331,
+      "grad_norm": 1.578125,
+      "learning_rate": 1.7572420213083822e-05,
+      "loss": 1.0356,
+      "step": 4041
+    },
+    {
+      "epoch": 0.6930578476970229,
+      "grad_norm": 1.6015625,
+      "learning_rate": 1.757124057582472e-05,
+      "loss": 1.0043,
+      "step": 4042
+    },
+    {
+      "epoch": 0.6932293117860128,
+      "grad_norm": 1.671875,
+      "learning_rate": 1.757006069163895e-05,
+      "loss": 1.0444,
+      "step": 4043
+    },
+    {
+      "epoch": 0.6934007758750027,
+      "grad_norm": 1.5625,
+      "learning_rate": 1.7568880560564994e-05,
+      "loss": 1.0236,
+      "step": 4044
+    },
+    {
+      "epoch": 0.6935722399639925,
+      "grad_norm": 1.5390625,
+      "learning_rate": 1.756770018264134e-05,
+      "loss": 0.9874,
+      "step": 4045
+    },
+    {
+      "epoch": 0.6937437040529824,
+      "grad_norm": 1.46875,
+      "learning_rate": 1.7566519557906488e-05,
+      "loss": 1.0284,
+      "step": 4046
+    },
+    {
+      "epoch": 0.6939151681419723,
+      "grad_norm": 1.671875,
+      "learning_rate": 1.756533868639894e-05,
+      "loss": 1.1392,
+      "step": 4047
+    },
+    {
+      "epoch": 0.6940866322309621,
+      "grad_norm": 1.5859375,
+      "learning_rate": 1.7564157568157208e-05,
+      "loss": 0.9929,
+      "step": 4048
+    },
+    {
+      "epoch": 0.694258096319952,
+      "grad_norm": 1.578125,
+      "learning_rate": 1.7562976203219815e-05,
+      "loss": 1.0021,
+      "step": 4049
+    },
+    {
+      "epoch": 0.6944295604089419,
+      "grad_norm": 1.5703125,
+      "learning_rate": 1.756179459162529e-05,
+      "loss": 0.9882,
+      "step": 4050
+    },
+    {
+      "epoch": 0.6946010244979317,
+      "grad_norm": 1.53125,
+      "learning_rate": 1.7560612733412167e-05,
+      "loss": 1.0035,
+      "step": 4051
+    },
+    {
+      "epoch": 0.6947724885869215,
+      "grad_norm": 1.609375,
+      "learning_rate": 1.755943062861899e-05,
+      "loss": 1.0026,
+      "step": 4052
+    },
+    {
+      "epoch": 0.6949439526759115,
+      "grad_norm": 1.609375,
+      "learning_rate": 1.7558248277284318e-05,
+      "loss": 0.9078,
+      "step": 4053
+    },
+    {
+      "epoch": 0.6951154167649013,
+      "grad_norm": 1.546875,
+      "learning_rate": 1.7557065679446705e-05,
+      "loss": 1.0097,
+      "step": 4054
+    },
+    {
+      "epoch": 0.6952868808538911,
+      "grad_norm": 1.5703125,
+      "learning_rate": 1.755588283514472e-05,
+      "loss": 1.1547,
+      "step": 4055
+    },
+    {
+      "epoch": 0.6954583449428811,
+      "grad_norm": 1.5625,
+      "learning_rate": 1.755469974441695e-05,
+      "loss": 0.9205,
+      "step": 4056
+    },
+    {
+      "epoch": 0.6956298090318709,
+      "grad_norm": 1.703125,
+      "learning_rate": 1.7553516407301967e-05,
+      "loss": 1.0198,
+      "step": 4057
+    },
+    {
+      "epoch": 0.6958012731208607,
+      "grad_norm": 1.5,
+      "learning_rate": 1.7552332823838375e-05,
+      "loss": 0.9808,
+      "step": 4058
+    },
+    {
+      "epoch": 0.6959727372098506,
+      "grad_norm": 1.5546875,
+      "learning_rate": 1.7551148994064765e-05,
+      "loss": 0.9738,
+      "step": 4059
+    },
+    {
+      "epoch": 0.6961442012988405,
+      "grad_norm": 1.4375,
+      "learning_rate": 1.7549964918019754e-05,
+      "loss": 0.9615,
+      "step": 4060
+    },
+    {
+      "epoch": 0.6963156653878303,
+      "grad_norm": 1.6171875,
+      "learning_rate": 1.7548780595741957e-05,
+      "loss": 1.0502,
+      "step": 4061
+    },
+    {
+      "epoch": 0.6964871294768202,
+      "grad_norm": 1.5390625,
+      "learning_rate": 1.754759602727e-05,
+      "loss": 1.0078,
+      "step": 4062
+    },
+    {
+      "epoch": 0.6966585935658101,
+      "grad_norm": 1.5390625,
+      "learning_rate": 1.754641121264251e-05,
+      "loss": 0.9771,
+      "step": 4063
+    },
+    {
+      "epoch": 0.6968300576547999,
+      "grad_norm": 1.6328125,
+      "learning_rate": 1.7545226151898134e-05,
+      "loss": 1.1078,
+      "step": 4064
+    },
+    {
+      "epoch": 0.6970015217437898,
+      "grad_norm": 1.671875,
+      "learning_rate": 1.7544040845075528e-05,
+      "loss": 1.0964,
+      "step": 4065
+    },
+    {
+      "epoch": 0.6971729858327796,
+      "grad_norm": 1.5546875,
+      "learning_rate": 1.7542855292213334e-05,
+      "loss": 1.0188,
+      "step": 4066
+    },
+    {
+      "epoch": 0.6973444499217695,
+      "grad_norm": 1.515625,
+      "learning_rate": 1.7541669493350227e-05,
+      "loss": 0.9677,
+      "step": 4067
+    },
+    {
+      "epoch": 0.6975159140107594,
+      "grad_norm": 1.640625,
+      "learning_rate": 1.754048344852488e-05,
+      "loss": 1.0634,
+      "step": 4068
+    },
+    {
+      "epoch": 0.6976873780997492,
+      "grad_norm": 1.6015625,
+      "learning_rate": 1.753929715777597e-05,
+      "loss": 1.0472,
+      "step": 4069
+    },
+    {
+      "epoch": 0.6978588421887391,
+      "grad_norm": 1.6796875,
+      "learning_rate": 1.7538110621142194e-05,
+      "loss": 1.0953,
+      "step": 4070
+    },
+    {
+      "epoch": 0.698030306277729,
+      "grad_norm": 1.609375,
+      "learning_rate": 1.7536923838662243e-05,
+      "loss": 1.0702,
+      "step": 4071
+    },
+    {
+      "epoch": 0.6982017703667188,
+      "grad_norm": 1.4921875,
+      "learning_rate": 1.7535736810374822e-05,
+      "loss": 1.0504,
+      "step": 4072
+    },
+    {
+      "epoch": 0.6983732344557086,
+      "grad_norm": 1.5078125,
+      "learning_rate": 1.7534549536318647e-05,
+      "loss": 0.9602,
+      "step": 4073
+    },
+    {
+      "epoch": 0.6985446985446986,
+      "grad_norm": 1.5859375,
+      "learning_rate": 1.753336201653244e-05,
+      "loss": 1.083,
+      "step": 4074
+    },
+    {
+      "epoch": 0.6987161626336884,
+      "grad_norm": 1.5859375,
+      "learning_rate": 1.753217425105493e-05,
+      "loss": 0.9854,
+      "step": 4075
+    },
+    {
+      "epoch": 0.6988876267226782,
+      "grad_norm": 1.5625,
+      "learning_rate": 1.7530986239924858e-05,
+      "loss": 0.9757,
+      "step": 4076
+    },
+    {
+      "epoch": 0.6990590908116682,
+      "grad_norm": 1.5703125,
+      "learning_rate": 1.7529797983180962e-05,
+      "loss": 1.0032,
+      "step": 4077
+    },
+    {
+      "epoch": 0.699230554900658,
+      "grad_norm": 1.6015625,
+      "learning_rate": 1.7528609480861995e-05,
+      "loss": 0.9975,
+      "step": 4078
+    },
+    {
+      "epoch": 0.6994020189896478,
+      "grad_norm": 1.5859375,
+      "learning_rate": 1.752742073300673e-05,
+      "loss": 1.0549,
+      "step": 4079
+    },
+    {
+      "epoch": 0.6995734830786378,
+      "grad_norm": 1.5625,
+      "learning_rate": 1.7526231739653923e-05,
+      "loss": 1.051,
+      "step": 4080
+    },
+    {
+      "epoch": 0.6997449471676276,
+      "grad_norm": 1.5234375,
+      "learning_rate": 1.7525042500842363e-05,
+      "loss": 1.0099,
+      "step": 4081
+    },
+    {
+      "epoch": 0.6999164112566174,
+      "grad_norm": 1.5546875,
+      "learning_rate": 1.7523853016610826e-05,
+      "loss": 1.0453,
+      "step": 4082
+    },
+    {
+      "epoch": 0.7000878753456073,
+      "grad_norm": 1.7421875,
+      "learning_rate": 1.7522663286998113e-05,
+      "loss": 1.1041,
+      "step": 4083
+    },
+    {
+      "epoch": 0.7002593394345972,
+      "grad_norm": 1.546875,
+      "learning_rate": 1.752147331204302e-05,
+      "loss": 0.9947,
+      "step": 4084
+    },
+    {
+      "epoch": 0.700430803523587,
+      "grad_norm": 1.53125,
+      "learning_rate": 1.752028309178436e-05,
+      "loss": 1.012,
+      "step": 4085
+    },
+    {
+      "epoch": 0.7006022676125769,
+      "grad_norm": 1.5234375,
+      "learning_rate": 1.751909262626095e-05,
+      "loss": 1.0268,
+      "step": 4086
+    },
+    {
+      "epoch": 0.7007737317015668,
+      "grad_norm": 1.59375,
+      "learning_rate": 1.7517901915511614e-05,
+      "loss": 0.9548,
+      "step": 4087
+    },
+    {
+      "epoch": 0.7009451957905566,
+      "grad_norm": 1.6484375,
+      "learning_rate": 1.7516710959575188e-05,
+      "loss": 1.0622,
+      "step": 4088
+    },
+    {
+      "epoch": 0.7011166598795465,
+      "grad_norm": 1.5234375,
+      "learning_rate": 1.751551975849051e-05,
+      "loss": 0.9667,
+      "step": 4089
+    },
+    {
+      "epoch": 0.7012881239685363,
+      "grad_norm": 1.625,
+      "learning_rate": 1.751432831229644e-05,
+      "loss": 1.0628,
+      "step": 4090
+    },
+    {
+      "epoch": 0.7014595880575262,
+      "grad_norm": 1.7578125,
+      "learning_rate": 1.751313662103182e-05,
+      "loss": 1.0634,
+      "step": 4091
+    },
+    {
+      "epoch": 0.7016310521465161,
+      "grad_norm": 1.515625,
+      "learning_rate": 1.7511944684735523e-05,
+      "loss": 1.0062,
+      "step": 4092
+    },
+    {
+      "epoch": 0.7018025162355059,
+      "grad_norm": 1.609375,
+      "learning_rate": 1.7510752503446423e-05,
+      "loss": 1.0688,
+      "step": 4093
+    },
+    {
+      "epoch": 0.7019739803244958,
+      "grad_norm": 1.6015625,
+      "learning_rate": 1.7509560077203404e-05,
+      "loss": 0.9527,
+      "step": 4094
+    },
+    {
+      "epoch": 0.7021454444134857,
+      "grad_norm": 1.578125,
+      "learning_rate": 1.7508367406045348e-05,
+      "loss": 1.0027,
+      "step": 4095
+    },
+    {
+      "epoch": 0.7023169085024755,
+      "grad_norm": 1.46875,
+      "learning_rate": 1.750717449001116e-05,
+      "loss": 0.9678,
+      "step": 4096
+    },
+    {
+      "epoch": 0.7024883725914653,
+      "grad_norm": 1.484375,
+      "learning_rate": 1.750598132913974e-05,
+      "loss": 0.9813,
+      "step": 4097
+    },
+    {
+      "epoch": 0.7026598366804553,
+      "grad_norm": 1.59375,
+      "learning_rate": 1.7504787923470007e-05,
+      "loss": 1.0437,
+      "step": 4098
+    },
+    {
+      "epoch": 0.7028313007694451,
+      "grad_norm": 1.46875,
+      "learning_rate": 1.7503594273040877e-05,
+      "loss": 0.981,
+      "step": 4099
+    },
+    {
+      "epoch": 0.7030027648584349,
+      "grad_norm": 1.53125,
+      "learning_rate": 1.7502400377891285e-05,
+      "loss": 0.9728,
+      "step": 4100
+    },
+    {
+      "epoch": 0.7031742289474249,
+      "grad_norm": 1.6484375,
+      "learning_rate": 1.7501206238060162e-05,
+      "loss": 1.0596,
+      "step": 4101
+    },
+    {
+      "epoch": 0.7033456930364147,
+      "grad_norm": 1.4921875,
+      "learning_rate": 1.7500011853586457e-05,
+      "loss": 0.9799,
+      "step": 4102
+    },
+    {
+      "epoch": 0.7035171571254045,
+      "grad_norm": 1.5859375,
+      "learning_rate": 1.7498817224509124e-05,
+      "loss": 0.9499,
+      "step": 4103
+    },
+    {
+      "epoch": 0.7036886212143945,
+      "grad_norm": 1.6171875,
+      "learning_rate": 1.7497622350867125e-05,
+      "loss": 1.1101,
+      "step": 4104
+    },
+    {
+      "epoch": 0.7038600853033843,
+      "grad_norm": 1.578125,
+      "learning_rate": 1.7496427232699423e-05,
+      "loss": 1.0013,
+      "step": 4105
+    },
+    {
+      "epoch": 0.7040315493923741,
+      "grad_norm": 1.53125,
+      "learning_rate": 1.7495231870045002e-05,
+      "loss": 0.9436,
+      "step": 4106
+    },
+    {
+      "epoch": 0.704203013481364,
+      "grad_norm": 1.6875,
+      "learning_rate": 1.749403626294285e-05,
+      "loss": 1.1108,
+      "step": 4107
+    },
+    {
+      "epoch": 0.7043744775703539,
+      "grad_norm": 1.578125,
+      "learning_rate": 1.7492840411431952e-05,
+      "loss": 0.9605,
+      "step": 4108
+    },
+    {
+      "epoch": 0.7045459416593437,
+      "grad_norm": 1.578125,
+      "learning_rate": 1.7491644315551314e-05,
+      "loss": 1.0315,
+      "step": 4109
+    },
+    {
+      "epoch": 0.7047174057483336,
+      "grad_norm": 1.5,
+      "learning_rate": 1.7490447975339943e-05,
+      "loss": 0.9563,
+      "step": 4110
+    },
+    {
+      "epoch": 0.7048888698373235,
+      "grad_norm": 1.546875,
+      "learning_rate": 1.7489251390836853e-05,
+      "loss": 0.9553,
+      "step": 4111
+    },
+    {
+      "epoch": 0.7050603339263133,
+      "grad_norm": 1.609375,
+      "learning_rate": 1.748805456208108e-05,
+      "loss": 1.0408,
+      "step": 4112
+    },
+    {
+      "epoch": 0.7052317980153032,
+      "grad_norm": 1.5390625,
+      "learning_rate": 1.7486857489111643e-05,
+      "loss": 1.0166,
+      "step": 4113
+    },
+    {
+      "epoch": 0.705403262104293,
+      "grad_norm": 1.4921875,
+      "learning_rate": 1.7485660171967595e-05,
+      "loss": 0.9863,
+      "step": 4114
+    },
+    {
+      "epoch": 0.7055747261932829,
+      "grad_norm": 1.5625,
+      "learning_rate": 1.748446261068798e-05,
+      "loss": 0.9943,
+      "step": 4115
+    },
+    {
+      "epoch": 0.7057461902822728,
+      "grad_norm": 1.5546875,
+      "learning_rate": 1.7483264805311856e-05,
+      "loss": 0.9912,
+      "step": 4116
+    },
+    {
+      "epoch": 0.7059176543712626,
+      "grad_norm": 1.5,
+      "learning_rate": 1.7482066755878287e-05,
+      "loss": 1.0156,
+      "step": 4117
+    },
+    {
+      "epoch": 0.7060891184602525,
+      "grad_norm": 1.6015625,
+      "learning_rate": 1.7480868462426345e-05,
+      "loss": 1.0942,
+      "step": 4118
+    },
+    {
+      "epoch": 0.7062605825492423,
+      "grad_norm": 1.4921875,
+      "learning_rate": 1.7479669924995117e-05,
+      "loss": 0.9926,
+      "step": 4119
+    },
+    {
+      "epoch": 0.7064320466382322,
+      "grad_norm": 1.5390625,
+      "learning_rate": 1.747847114362368e-05,
+      "loss": 1.0409,
+      "step": 4120
+    },
+    {
+      "epoch": 0.706603510727222,
+      "grad_norm": 1.5859375,
+      "learning_rate": 1.747727211835114e-05,
+      "loss": 1.0482,
+      "step": 4121
+    },
+    {
+      "epoch": 0.7067749748162119,
+      "grad_norm": 1.5625,
+      "learning_rate": 1.7476072849216602e-05,
+      "loss": 1.0351,
+      "step": 4122
+    },
+    {
+      "epoch": 0.7069464389052018,
+      "grad_norm": 1.6171875,
+      "learning_rate": 1.7474873336259172e-05,
+      "loss": 1.0221,
+      "step": 4123
+    },
+    {
+      "epoch": 0.7071179029941916,
+      "grad_norm": 1.59375,
+      "learning_rate": 1.747367357951798e-05,
+      "loss": 0.9936,
+      "step": 4124
+    },
+    {
+      "epoch": 0.7072893670831815,
+      "grad_norm": 1.6640625,
+      "learning_rate": 1.747247357903214e-05,
+      "loss": 1.1066,
+      "step": 4125
+    },
+    {
+      "epoch": 0.7074608311721714,
+      "grad_norm": 1.546875,
+      "learning_rate": 1.7471273334840807e-05,
+      "loss": 1.0051,
+      "step": 4126
+    },
+    {
+      "epoch": 0.7076322952611612,
+      "grad_norm": 1.546875,
+      "learning_rate": 1.7470072846983112e-05,
+      "loss": 1.006,
+      "step": 4127
+    },
+    {
+      "epoch": 0.707803759350151,
+      "grad_norm": 1.6171875,
+      "learning_rate": 1.7468872115498216e-05,
+      "loss": 0.9908,
+      "step": 4128
+    },
+    {
+      "epoch": 0.707975223439141,
+      "grad_norm": 1.5625,
+      "learning_rate": 1.7467671140425272e-05,
+      "loss": 0.9295,
+      "step": 4129
+    },
+    {
+      "epoch": 0.7081466875281308,
+      "grad_norm": 1.5234375,
+      "learning_rate": 1.746646992180345e-05,
+      "loss": 0.9709,
+      "step": 4130
+    },
+    {
+      "epoch": 0.7083181516171206,
+      "grad_norm": 1.5234375,
+      "learning_rate": 1.7465268459671932e-05,
+      "loss": 1.0124,
+      "step": 4131
+    },
+    {
+      "epoch": 0.7084896157061106,
+      "grad_norm": 1.5390625,
+      "learning_rate": 1.7464066754069893e-05,
+      "loss": 0.966,
+      "step": 4132
+    },
+    {
+      "epoch": 0.7086610797951004,
+      "grad_norm": 1.53125,
+      "learning_rate": 1.7462864805036535e-05,
+      "loss": 0.9522,
+      "step": 4133
+    },
+    {
+      "epoch": 0.7088325438840902,
+      "grad_norm": 1.4921875,
+      "learning_rate": 1.746166261261105e-05,
+      "loss": 0.9749,
+      "step": 4134
+    },
+    {
+      "epoch": 0.7090040079730802,
+      "grad_norm": 1.5625,
+      "learning_rate": 1.746046017683265e-05,
+      "loss": 0.9992,
+      "step": 4135
+    },
+    {
+      "epoch": 0.70917547206207,
+      "grad_norm": 1.6015625,
+      "learning_rate": 1.7459257497740548e-05,
+      "loss": 1.0595,
+      "step": 4136
+    },
+    {
+      "epoch": 0.7093469361510598,
+      "grad_norm": 1.5390625,
+      "learning_rate": 1.7458054575373973e-05,
+      "loss": 1.0378,
+      "step": 4137
+    },
+    {
+      "epoch": 0.7095184002400498,
+      "grad_norm": 1.4375,
+      "learning_rate": 1.7456851409772153e-05,
+      "loss": 0.9598,
+      "step": 4138
+    },
+    {
+      "epoch": 0.7096898643290396,
+      "grad_norm": 1.609375,
+      "learning_rate": 1.745564800097433e-05,
+      "loss": 1.0255,
+      "step": 4139
+    },
+    {
+      "epoch": 0.7098613284180294,
+      "grad_norm": 1.5390625,
+      "learning_rate": 1.7454444349019747e-05,
+      "loss": 1.005,
+      "step": 4140
+    },
+    {
+      "epoch": 0.7100327925070193,
+      "grad_norm": 1.6953125,
+      "learning_rate": 1.7453240453947664e-05,
+      "loss": 1.0376,
+      "step": 4141
+    },
+    {
+      "epoch": 0.7102042565960092,
+      "grad_norm": 1.53125,
+      "learning_rate": 1.7452036315797346e-05,
+      "loss": 0.9216,
+      "step": 4142
+    },
+    {
+      "epoch": 0.710375720684999,
+      "grad_norm": 1.703125,
+      "learning_rate": 1.745083193460806e-05,
+      "loss": 1.0864,
+      "step": 4143
+    },
+    {
+      "epoch": 0.7105471847739889,
+      "grad_norm": 1.5546875,
+      "learning_rate": 1.7449627310419086e-05,
+      "loss": 0.9865,
+      "step": 4144
+    },
+    {
+      "epoch": 0.7107186488629788,
+      "grad_norm": 1.640625,
+      "learning_rate": 1.7448422443269716e-05,
+      "loss": 1.0528,
+      "step": 4145
+    },
+    {
+      "epoch": 0.7108901129519686,
+      "grad_norm": 1.640625,
+      "learning_rate": 1.744721733319924e-05,
+      "loss": 1.0079,
+      "step": 4146
+    },
+    {
+      "epoch": 0.7110615770409585,
+      "grad_norm": 1.5078125,
+      "learning_rate": 1.7446011980246963e-05,
+      "loss": 0.9296,
+      "step": 4147
+    },
+    {
+      "epoch": 0.7112330411299483,
+      "grad_norm": 1.578125,
+      "learning_rate": 1.7444806384452198e-05,
+      "loss": 1.0596,
+      "step": 4148
+    },
+    {
+      "epoch": 0.7114045052189382,
+      "grad_norm": 1.5859375,
+      "learning_rate": 1.744360054585426e-05,
+      "loss": 0.9702,
+      "step": 4149
+    },
+    {
+      "epoch": 0.7115759693079281,
+      "grad_norm": 1.5078125,
+      "learning_rate": 1.744239446449248e-05,
+      "loss": 0.933,
+      "step": 4150
+    },
+    {
+      "epoch": 0.7117474333969179,
+      "grad_norm": 1.5859375,
+      "learning_rate": 1.744118814040619e-05,
+      "loss": 1.1285,
+      "step": 4151
+    },
+    {
+      "epoch": 0.7119188974859078,
+      "grad_norm": 1.5625,
+      "learning_rate": 1.7439981573634734e-05,
+      "loss": 1.0722,
+      "step": 4152
+    },
+    {
+      "epoch": 0.7120903615748977,
+      "grad_norm": 1.609375,
+      "learning_rate": 1.743877476421746e-05,
+      "loss": 1.0448,
+      "step": 4153
+    },
+    {
+      "epoch": 0.7122618256638875,
+      "grad_norm": 1.5390625,
+      "learning_rate": 1.7437567712193732e-05,
+      "loss": 0.93,
+      "step": 4154
+    },
+    {
+      "epoch": 0.7124332897528773,
+      "grad_norm": 1.5625,
+      "learning_rate": 1.7436360417602915e-05,
+      "loss": 0.989,
+      "step": 4155
+    },
+    {
+      "epoch": 0.7126047538418673,
+      "grad_norm": 1.53125,
+      "learning_rate": 1.743515288048438e-05,
+      "loss": 1.0188,
+      "step": 4156
+    },
+    {
+      "epoch": 0.7127762179308571,
+      "grad_norm": 1.6171875,
+      "learning_rate": 1.7433945100877513e-05,
+      "loss": 1.0177,
+      "step": 4157
+    },
+    {
+      "epoch": 0.7129476820198469,
+      "grad_norm": 1.5234375,
+      "learning_rate": 1.74327370788217e-05,
+      "loss": 0.9703,
+      "step": 4158
+    },
+    {
+      "epoch": 0.7131191461088369,
+      "grad_norm": 1.6015625,
+      "learning_rate": 1.743152881435634e-05,
+      "loss": 1.0018,
+      "step": 4159
+    },
+    {
+      "epoch": 0.7132906101978267,
+      "grad_norm": 1.65625,
+      "learning_rate": 1.7430320307520844e-05,
+      "loss": 1.0156,
+      "step": 4160
+    },
+    {
+      "epoch": 0.7134620742868165,
+      "grad_norm": 1.6015625,
+      "learning_rate": 1.7429111558354624e-05,
+      "loss": 1.0338,
+      "step": 4161
+    },
+    {
+      "epoch": 0.7136335383758065,
+      "grad_norm": 1.6640625,
+      "learning_rate": 1.74279025668971e-05,
+      "loss": 1.0291,
+      "step": 4162
+    },
+    {
+      "epoch": 0.7138050024647963,
+      "grad_norm": 1.640625,
+      "learning_rate": 1.7426693333187702e-05,
+      "loss": 1.1897,
+      "step": 4163
+    },
+    {
+      "epoch": 0.7139764665537861,
+      "grad_norm": 1.6328125,
+      "learning_rate": 1.7425483857265865e-05,
+      "loss": 1.0438,
+      "step": 4164
+    },
+    {
+      "epoch": 0.714147930642776,
+      "grad_norm": 1.5390625,
+      "learning_rate": 1.742427413917104e-05,
+      "loss": 1.0547,
+      "step": 4165
+    },
+    {
+      "epoch": 0.7143193947317659,
+      "grad_norm": 1.546875,
+      "learning_rate": 1.742306417894268e-05,
+      "loss": 1.0058,
+      "step": 4166
+    },
+    {
+      "epoch": 0.7144908588207557,
+      "grad_norm": 1.6015625,
+      "learning_rate": 1.7421853976620245e-05,
+      "loss": 1.0136,
+      "step": 4167
+    },
+    {
+      "epoch": 0.7146623229097456,
+      "grad_norm": 1.515625,
+      "learning_rate": 1.74206435322432e-05,
+      "loss": 1.0062,
+      "step": 4168
+    },
+    {
+      "epoch": 0.7148337869987355,
+      "grad_norm": 1.6328125,
+      "learning_rate": 1.7419432845851027e-05,
+      "loss": 1.1097,
+      "step": 4169
+    },
+    {
+      "epoch": 0.7150052510877253,
+      "grad_norm": 1.6640625,
+      "learning_rate": 1.741822191748321e-05,
+      "loss": 1.017,
+      "step": 4170
+    },
+    {
+      "epoch": 0.7151767151767152,
+      "grad_norm": 1.609375,
+      "learning_rate": 1.7417010747179245e-05,
+      "loss": 0.9241,
+      "step": 4171
+    },
+    {
+      "epoch": 0.715348179265705,
+      "grad_norm": 1.5859375,
+      "learning_rate": 1.7415799334978624e-05,
+      "loss": 0.9328,
+      "step": 4172
+    },
+    {
+      "epoch": 0.7155196433546949,
+      "grad_norm": 1.5859375,
+      "learning_rate": 1.7414587680920864e-05,
+      "loss": 0.9638,
+      "step": 4173
+    },
+    {
+      "epoch": 0.7156911074436848,
+      "grad_norm": 1.5625,
+      "learning_rate": 1.741337578504548e-05,
+      "loss": 1.0168,
+      "step": 4174
+    },
+    {
+      "epoch": 0.7158625715326746,
+      "grad_norm": 1.609375,
+      "learning_rate": 1.7412163647391996e-05,
+      "loss": 1.0322,
+      "step": 4175
+    },
+    {
+      "epoch": 0.7160340356216645,
+      "grad_norm": 1.59375,
+      "learning_rate": 1.7410951267999943e-05,
+      "loss": 1.0177,
+      "step": 4176
+    },
+    {
+      "epoch": 0.7162054997106544,
+      "grad_norm": 1.5625,
+      "learning_rate": 1.740973864690886e-05,
+      "loss": 1.0237,
+      "step": 4177
+    },
+    {
+      "epoch": 0.7163769637996442,
+      "grad_norm": 1.5546875,
+      "learning_rate": 1.7408525784158298e-05,
+      "loss": 0.9636,
+      "step": 4178
+    },
+    {
+      "epoch": 0.716548427888634,
+      "grad_norm": 1.6171875,
+      "learning_rate": 1.7407312679787814e-05,
+      "loss": 0.9987,
+      "step": 4179
+    },
+    {
+      "epoch": 0.716719891977624,
+      "grad_norm": 1.5234375,
+      "learning_rate": 1.740609933383697e-05,
+      "loss": 0.9429,
+      "step": 4180
+    },
+    {
+      "epoch": 0.7168913560666138,
+      "grad_norm": 1.625,
+      "learning_rate": 1.7404885746345333e-05,
+      "loss": 1.0416,
+      "step": 4181
+    },
+    {
+      "epoch": 0.7170628201556036,
+      "grad_norm": 1.6171875,
+      "learning_rate": 1.7403671917352492e-05,
+      "loss": 1.0059,
+      "step": 4182
+    },
+    {
+      "epoch": 0.7172342842445936,
+      "grad_norm": 1.6796875,
+      "learning_rate": 1.7402457846898032e-05,
+      "loss": 1.0392,
+      "step": 4183
+    },
+    {
+      "epoch": 0.7174057483335834,
+      "grad_norm": 1.5390625,
+      "learning_rate": 1.7401243535021547e-05,
+      "loss": 1.034,
+      "step": 4184
+    },
+    {
+      "epoch": 0.7175772124225732,
+      "grad_norm": 1.5703125,
+      "learning_rate": 1.7400028981762634e-05,
+      "loss": 1.0569,
+      "step": 4185
+    },
+    {
+      "epoch": 0.7177486765115632,
+      "grad_norm": 1.53125,
+      "learning_rate": 1.7398814187160913e-05,
+      "loss": 0.996,
+      "step": 4186
+    },
+    {
+      "epoch": 0.717920140600553,
+      "grad_norm": 1.59375,
+      "learning_rate": 1.7397599151256003e-05,
+      "loss": 1.0013,
+      "step": 4187
+    },
+    {
+      "epoch": 0.7180916046895428,
+      "grad_norm": 1.578125,
+      "learning_rate": 1.739638387408752e-05,
+      "loss": 0.9945,
+      "step": 4188
+    },
+    {
+      "epoch": 0.7182630687785327,
+      "grad_norm": 1.578125,
+      "learning_rate": 1.7395168355695116e-05,
+      "loss": 1.0584,
+      "step": 4189
+    },
+    {
+      "epoch": 0.7184345328675226,
+      "grad_norm": 1.5390625,
+      "learning_rate": 1.739395259611842e-05,
+      "loss": 0.9664,
+      "step": 4190
+    },
+    {
+      "epoch": 0.7186059969565124,
+      "grad_norm": 1.578125,
+      "learning_rate": 1.7392736595397086e-05,
+      "loss": 1.0509,
+      "step": 4191
+    },
+    {
+      "epoch": 0.7187774610455023,
+      "grad_norm": 1.5546875,
+      "learning_rate": 1.7391520353570772e-05,
+      "loss": 0.9658,
+      "step": 4192
+    },
+    {
+      "epoch": 0.7189489251344922,
+      "grad_norm": 1.5546875,
+      "learning_rate": 1.739030387067915e-05,
+      "loss": 1.0161,
+      "step": 4193
+    },
+    {
+      "epoch": 0.719120389223482,
+      "grad_norm": 1.5625,
+      "learning_rate": 1.7389087146761886e-05,
+      "loss": 0.998,
+      "step": 4194
+    },
+    {
+      "epoch": 0.7192918533124719,
+      "grad_norm": 1.6640625,
+      "learning_rate": 1.7387870181858666e-05,
+      "loss": 1.0322,
+      "step": 4195
+    },
+    {
+      "epoch": 0.7194633174014617,
+      "grad_norm": 1.65625,
+      "learning_rate": 1.738665297600918e-05,
+      "loss": 1.0163,
+      "step": 4196
+    },
+    {
+      "epoch": 0.7196347814904516,
+      "grad_norm": 1.59375,
+      "learning_rate": 1.738543552925312e-05,
+      "loss": 0.9975,
+      "step": 4197
+    },
+    {
+      "epoch": 0.7198062455794415,
+      "grad_norm": 1.5,
+      "learning_rate": 1.7384217841630207e-05,
+      "loss": 1.0121,
+      "step": 4198
+    },
+    {
+      "epoch": 0.7199777096684313,
+      "grad_norm": 1.6171875,
+      "learning_rate": 1.7382999913180135e-05,
+      "loss": 1.0565,
+      "step": 4199
+    },
+    {
+      "epoch": 0.7201491737574212,
+      "grad_norm": 1.671875,
+      "learning_rate": 1.7381781743942636e-05,
+      "loss": 0.9933,
+      "step": 4200
+    },
+    {
+      "epoch": 0.7201491737574212,
+      "eval_loss": 0.8648467063903809,
+      "eval_runtime": 837.3272,
+      "eval_samples_per_second": 2.984,
+      "eval_steps_per_second": 2.984,
+      "step": 4200
+    },
+    {
+      "epoch": 0.7203206378464111,
+      "grad_norm": 1.5625,
+      "learning_rate": 1.738056333395744e-05,
+      "loss": 1.0291,
+      "step": 4201
+    },
+    {
+      "epoch": 0.7204921019354009,
+      "grad_norm": 1.5625,
+      "learning_rate": 1.7379344683264275e-05,
+      "loss": 1.0251,
+      "step": 4202
+    },
+    {
+      "epoch": 0.7206635660243907,
+      "grad_norm": 1.5,
+      "learning_rate": 1.7378125791902897e-05,
+      "loss": 0.9141,
+      "step": 4203
+    },
+    {
+      "epoch": 0.7208350301133807,
+      "grad_norm": 1.5625,
+      "learning_rate": 1.737690665991305e-05,
+      "loss": 0.9753,
+      "step": 4204
+    },
+    {
+      "epoch": 0.7210064942023705,
+      "grad_norm": 1.5703125,
+      "learning_rate": 1.7375687287334502e-05,
+      "loss": 1.0581,
+      "step": 4205
+    },
+    {
+      "epoch": 0.7211779582913603,
+      "grad_norm": 1.6640625,
+      "learning_rate": 1.7374467674207015e-05,
+      "loss": 1.0378,
+      "step": 4206
+    },
+    {
+      "epoch": 0.7213494223803503,
+      "grad_norm": 1.5859375,
+      "learning_rate": 1.7373247820570368e-05,
+      "loss": 1.0007,
+      "step": 4207
+    },
+    {
+      "epoch": 0.7215208864693401,
+      "grad_norm": 1.640625,
+      "learning_rate": 1.7372027726464346e-05,
+      "loss": 0.9991,
+      "step": 4208
+    },
+    {
+      "epoch": 0.7216923505583299,
+      "grad_norm": 1.515625,
+      "learning_rate": 1.7370807391928737e-05,
+      "loss": 0.9788,
+      "step": 4209
+    },
+    {
+      "epoch": 0.7218638146473199,
+      "grad_norm": 1.5625,
+      "learning_rate": 1.7369586817003345e-05,
+      "loss": 0.9624,
+      "step": 4210
+    },
+    {
+      "epoch": 0.7220352787363097,
+      "grad_norm": 1.6171875,
+      "learning_rate": 1.7368366001727973e-05,
+      "loss": 0.9637,
+      "step": 4211
+    },
+    {
+      "epoch": 0.7222067428252995,
+      "grad_norm": 1.65625,
+      "learning_rate": 1.736714494614244e-05,
+      "loss": 1.0159,
+      "step": 4212
+    },
+    {
+      "epoch": 0.7223782069142893,
+      "grad_norm": 1.640625,
+      "learning_rate": 1.736592365028657e-05,
+      "loss": 0.9424,
+      "step": 4213
+    },
+    {
+      "epoch": 0.7225496710032793,
+      "grad_norm": 1.59375,
+      "learning_rate": 1.736470211420019e-05,
+      "loss": 1.0695,
+      "step": 4214
+    },
+    {
+      "epoch": 0.7227211350922691,
+      "grad_norm": 1.5625,
+      "learning_rate": 1.7363480337923143e-05,
+      "loss": 0.9998,
+      "step": 4215
+    },
+    {
+      "epoch": 0.7228925991812589,
+      "grad_norm": 1.5546875,
+      "learning_rate": 1.736225832149527e-05,
+      "loss": 0.9964,
+      "step": 4216
+    },
+    {
+      "epoch": 0.7230640632702489,
+      "grad_norm": 1.640625,
+      "learning_rate": 1.7361036064956433e-05,
+      "loss": 0.9818,
+      "step": 4217
+    },
+    {
+      "epoch": 0.7232355273592387,
+      "grad_norm": 1.8046875,
+      "learning_rate": 1.7359813568346492e-05,
+      "loss": 1.0565,
+      "step": 4218
+    },
+    {
+      "epoch": 0.7234069914482285,
+      "grad_norm": 1.546875,
+      "learning_rate": 1.735859083170531e-05,
+      "loss": 0.9854,
+      "step": 4219
+    },
+    {
+      "epoch": 0.7235784555372184,
+      "grad_norm": 1.625,
+      "learning_rate": 1.7357367855072778e-05,
+      "loss": 1.0865,
+      "step": 4220
+    },
+    {
+      "epoch": 0.7237499196262083,
+      "grad_norm": 1.609375,
+      "learning_rate": 1.7356144638488772e-05,
+      "loss": 0.9946,
+      "step": 4221
+    },
+    {
+      "epoch": 0.7239213837151981,
+      "grad_norm": 1.5859375,
+      "learning_rate": 1.7354921181993187e-05,
+      "loss": 0.948,
+      "step": 4222
+    },
+    {
+      "epoch": 0.724092847804188,
+      "grad_norm": 1.5703125,
+      "learning_rate": 1.7353697485625928e-05,
+      "loss": 1.0151,
+      "step": 4223
+    },
+    {
+      "epoch": 0.7242643118931779,
+      "grad_norm": 1.5,
+      "learning_rate": 1.73524735494269e-05,
+      "loss": 0.953,
+      "step": 4224
+    },
+    {
+      "epoch": 0.7244357759821677,
+      "grad_norm": 1.734375,
+      "learning_rate": 1.7351249373436028e-05,
+      "loss": 1.007,
+      "step": 4225
+    },
+    {
+      "epoch": 0.7246072400711576,
+      "grad_norm": 1.71875,
+      "learning_rate": 1.7350024957693232e-05,
+      "loss": 1.072,
+      "step": 4226
+    },
+    {
+      "epoch": 0.7247787041601474,
+      "grad_norm": 1.6171875,
+      "learning_rate": 1.734880030223844e-05,
+      "loss": 1.1049,
+      "step": 4227
+    },
+    {
+      "epoch": 0.7249501682491373,
+      "grad_norm": 1.6015625,
+      "learning_rate": 1.7347575407111595e-05,
+      "loss": 1.0501,
+      "step": 4228
+    },
+    {
+      "epoch": 0.7251216323381272,
+      "grad_norm": 1.6796875,
+      "learning_rate": 1.7346350272352653e-05,
+      "loss": 1.1176,
+      "step": 4229
+    },
+    {
+      "epoch": 0.725293096427117,
+      "grad_norm": 1.59375,
+      "learning_rate": 1.7345124898001562e-05,
+      "loss": 1.0401,
+      "step": 4230
+    },
+    {
+      "epoch": 0.7254645605161069,
+      "grad_norm": 1.6015625,
+      "learning_rate": 1.7343899284098292e-05,
+      "loss": 1.073,
+      "step": 4231
+    },
+    {
+      "epoch": 0.7256360246050968,
+      "grad_norm": 1.671875,
+      "learning_rate": 1.734267343068281e-05,
+      "loss": 1.0733,
+      "step": 4232
+    },
+    {
+      "epoch": 0.7258074886940866,
+      "grad_norm": 1.609375,
+      "learning_rate": 1.7341447337795098e-05,
+      "loss": 1.0307,
+      "step": 4233
+    },
+    {
+      "epoch": 0.7259789527830764,
+      "grad_norm": 1.6328125,
+      "learning_rate": 1.734022100547514e-05,
+      "loss": 1.0375,
+      "step": 4234
+    },
+    {
+      "epoch": 0.7261504168720664,
+      "grad_norm": 1.46875,
+      "learning_rate": 1.7338994433762938e-05,
+      "loss": 0.8565,
+      "step": 4235
+    },
+    {
+      "epoch": 0.7263218809610562,
+      "grad_norm": 1.578125,
+      "learning_rate": 1.733776762269849e-05,
+      "loss": 1.0315,
+      "step": 4236
+    },
+    {
+      "epoch": 0.726493345050046,
+      "grad_norm": 1.6171875,
+      "learning_rate": 1.7336540572321807e-05,
+      "loss": 0.9393,
+      "step": 4237
+    },
+    {
+      "epoch": 0.726664809139036,
+      "grad_norm": 1.59375,
+      "learning_rate": 1.733531328267291e-05,
+      "loss": 1.0198,
+      "step": 4238
+    },
+    {
+      "epoch": 0.7268362732280258,
+      "grad_norm": 1.5078125,
+      "learning_rate": 1.7334085753791824e-05,
+      "loss": 0.9681,
+      "step": 4239
+    },
+    {
+      "epoch": 0.7270077373170156,
+      "grad_norm": 1.59375,
+      "learning_rate": 1.7332857985718584e-05,
+      "loss": 1.0287,
+      "step": 4240
+    },
+    {
+      "epoch": 0.7271792014060056,
+      "grad_norm": 1.640625,
+      "learning_rate": 1.7331629978493234e-05,
+      "loss": 1.0395,
+      "step": 4241
+    },
+    {
+      "epoch": 0.7273506654949954,
+      "grad_norm": 1.6953125,
+      "learning_rate": 1.733040173215582e-05,
+      "loss": 1.0084,
+      "step": 4242
+    },
+    {
+      "epoch": 0.7275221295839852,
+      "grad_norm": 1.5703125,
+      "learning_rate": 1.7329173246746406e-05,
+      "loss": 1.0966,
+      "step": 4243
+    },
+    {
+      "epoch": 0.7276935936729751,
+      "grad_norm": 1.6953125,
+      "learning_rate": 1.732794452230505e-05,
+      "loss": 1.1253,
+      "step": 4244
+    },
+    {
+      "epoch": 0.727865057761965,
+      "grad_norm": 1.5,
+      "learning_rate": 1.7326715558871826e-05,
+      "loss": 0.9047,
+      "step": 4245
+    },
+    {
+      "epoch": 0.7280365218509548,
+      "grad_norm": 1.6328125,
+      "learning_rate": 1.7325486356486823e-05,
+      "loss": 1.1244,
+      "step": 4246
+    },
+    {
+      "epoch": 0.7282079859399447,
+      "grad_norm": 1.6953125,
+      "learning_rate": 1.732425691519012e-05,
+      "loss": 1.0456,
+      "step": 4247
+    },
+    {
+      "epoch": 0.7283794500289346,
+      "grad_norm": 1.59375,
+      "learning_rate": 1.732302723502182e-05,
+      "loss": 0.9747,
+      "step": 4248
+    },
+    {
+      "epoch": 0.7285509141179244,
+      "grad_norm": 1.5,
+      "learning_rate": 1.732179731602203e-05,
+      "loss": 1.0131,
+      "step": 4249
+    },
+    {
+      "epoch": 0.7287223782069143,
+      "grad_norm": 1.46875,
+      "learning_rate": 1.7320567158230855e-05,
+      "loss": 0.988,
+      "step": 4250
+    },
+    {
+      "epoch": 0.7288938422959041,
+      "grad_norm": 1.609375,
+      "learning_rate": 1.7319336761688415e-05,
+      "loss": 1.0683,
+      "step": 4251
+    },
+    {
+      "epoch": 0.729065306384894,
+      "grad_norm": 1.5703125,
+      "learning_rate": 1.7318106126434845e-05,
+      "loss": 1.0612,
+      "step": 4252
+    },
+    {
+      "epoch": 0.7292367704738839,
+      "grad_norm": 1.4765625,
+      "learning_rate": 1.7316875252510274e-05,
+      "loss": 0.9791,
+      "step": 4253
+    },
+    {
+      "epoch": 0.7294082345628737,
+      "grad_norm": 1.4609375,
+      "learning_rate": 1.7315644139954852e-05,
+      "loss": 0.9713,
+      "step": 4254
+    },
+    {
+      "epoch": 0.7295796986518636,
+      "grad_norm": 1.5546875,
+      "learning_rate": 1.7314412788808727e-05,
+      "loss": 1.002,
+      "step": 4255
+    },
+    {
+      "epoch": 0.7297511627408535,
+      "grad_norm": 1.671875,
+      "learning_rate": 1.731318119911205e-05,
+      "loss": 1.0398,
+      "step": 4256
+    },
+    {
+      "epoch": 0.7299226268298433,
+      "grad_norm": 1.4921875,
+      "learning_rate": 1.7311949370905e-05,
+      "loss": 0.9147,
+      "step": 4257
+    },
+    {
+      "epoch": 0.7300940909188331,
+      "grad_norm": 1.4921875,
+      "learning_rate": 1.731071730422775e-05,
+      "loss": 1.0034,
+      "step": 4258
+    },
+    {
+      "epoch": 0.7302655550078231,
+      "grad_norm": 1.5,
+      "learning_rate": 1.7309484999120475e-05,
+      "loss": 0.9908,
+      "step": 4259
+    },
+    {
+      "epoch": 0.7304370190968129,
+      "grad_norm": 1.59375,
+      "learning_rate": 1.730825245562337e-05,
+      "loss": 1.0843,
+      "step": 4260
+    },
+    {
+      "epoch": 0.7306084831858027,
+      "grad_norm": 1.703125,
+      "learning_rate": 1.730701967377663e-05,
+      "loss": 1.0963,
+      "step": 4261
+    },
+    {
+      "epoch": 0.7307799472747927,
+      "grad_norm": 1.515625,
+      "learning_rate": 1.7305786653620466e-05,
+      "loss": 0.9635,
+      "step": 4262
+    },
+    {
+      "epoch": 0.7309514113637825,
+      "grad_norm": 1.515625,
+      "learning_rate": 1.730455339519509e-05,
+      "loss": 1.0101,
+      "step": 4263
+    },
+    {
+      "epoch": 0.7311228754527723,
+      "grad_norm": 1.640625,
+      "learning_rate": 1.730331989854072e-05,
+      "loss": 1.0204,
+      "step": 4264
+    },
+    {
+      "epoch": 0.7312943395417623,
+      "grad_norm": 1.609375,
+      "learning_rate": 1.730208616369758e-05,
+      "loss": 1.019,
+      "step": 4265
+    },
+    {
+      "epoch": 0.7314658036307521,
+      "grad_norm": 1.4609375,
+      "learning_rate": 1.730085219070592e-05,
+      "loss": 0.928,
+      "step": 4266
+    },
+    {
+      "epoch": 0.7316372677197419,
+      "grad_norm": 1.6015625,
+      "learning_rate": 1.7299617979605976e-05,
+      "loss": 1.0206,
+      "step": 4267
+    },
+    {
+      "epoch": 0.7318087318087318,
+      "grad_norm": 1.5546875,
+      "learning_rate": 1.7298383530438002e-05,
+      "loss": 0.9391,
+      "step": 4268
+    },
+    {
+      "epoch": 0.7319801958977217,
+      "grad_norm": 1.4921875,
+      "learning_rate": 1.7297148843242257e-05,
+      "loss": 0.9757,
+      "step": 4269
+    },
+    {
+      "epoch": 0.7321516599867115,
+      "grad_norm": 1.5546875,
+      "learning_rate": 1.7295913918059008e-05,
+      "loss": 0.993,
+      "step": 4270
+    },
+    {
+      "epoch": 0.7323231240757014,
+      "grad_norm": 1.59375,
+      "learning_rate": 1.7294678754928535e-05,
+      "loss": 0.9771,
+      "step": 4271
+    },
+    {
+      "epoch": 0.7324945881646913,
+      "grad_norm": 1.5625,
+      "learning_rate": 1.7293443353891118e-05,
+      "loss": 1.0116,
+      "step": 4272
+    },
+    {
+      "epoch": 0.7326660522536811,
+      "grad_norm": 1.4921875,
+      "learning_rate": 1.729220771498705e-05,
+      "loss": 0.9867,
+      "step": 4273
+    },
+    {
+      "epoch": 0.732837516342671,
+      "grad_norm": 1.59375,
+      "learning_rate": 1.7290971838256624e-05,
+      "loss": 1.0323,
+      "step": 4274
+    },
+    {
+      "epoch": 0.7330089804316609,
+      "grad_norm": 1.6171875,
+      "learning_rate": 1.7289735723740157e-05,
+      "loss": 0.9824,
+      "step": 4275
+    },
+    {
+      "epoch": 0.7331804445206507,
+      "grad_norm": 1.578125,
+      "learning_rate": 1.7288499371477954e-05,
+      "loss": 0.9785,
+      "step": 4276
+    },
+    {
+      "epoch": 0.7333519086096406,
+      "grad_norm": 1.625,
+      "learning_rate": 1.728726278151034e-05,
+      "loss": 1.0296,
+      "step": 4277
+    },
+    {
+      "epoch": 0.7335233726986304,
+      "grad_norm": 1.5078125,
+      "learning_rate": 1.7286025953877644e-05,
+      "loss": 1.1321,
+      "step": 4278
+    },
+    {
+      "epoch": 0.7336948367876203,
+      "grad_norm": 1.5703125,
+      "learning_rate": 1.728478888862021e-05,
+      "loss": 1.0588,
+      "step": 4279
+    },
+    {
+      "epoch": 0.7338663008766102,
+      "grad_norm": 1.609375,
+      "learning_rate": 1.7283551585778375e-05,
+      "loss": 1.0378,
+      "step": 4280
+    },
+    {
+      "epoch": 0.7340377649656,
+      "grad_norm": 1.875,
+      "learning_rate": 1.72823140453925e-05,
+      "loss": 0.9893,
+      "step": 4281
+    },
+    {
+      "epoch": 0.7342092290545899,
+      "grad_norm": 1.625,
+      "learning_rate": 1.7281076267502936e-05,
+      "loss": 1.0624,
+      "step": 4282
+    },
+    {
+      "epoch": 0.7343806931435798,
+      "grad_norm": 1.5859375,
+      "learning_rate": 1.7279838252150057e-05,
+      "loss": 1.083,
+      "step": 4283
+    },
+    {
+      "epoch": 0.7345521572325696,
+      "grad_norm": 1.546875,
+      "learning_rate": 1.727859999937424e-05,
+      "loss": 1.0401,
+      "step": 4284
+    },
+    {
+      "epoch": 0.7347236213215594,
+      "grad_norm": 1.546875,
+      "learning_rate": 1.727736150921587e-05,
+      "loss": 1.0169,
+      "step": 4285
+    },
+    {
+      "epoch": 0.7348950854105494,
+      "grad_norm": 1.546875,
+      "learning_rate": 1.7276122781715335e-05,
+      "loss": 1.0907,
+      "step": 4286
+    },
+    {
+      "epoch": 0.7350665494995392,
+      "grad_norm": 1.6015625,
+      "learning_rate": 1.727488381691304e-05,
+      "loss": 1.038,
+      "step": 4287
+    },
+    {
+      "epoch": 0.735238013588529,
+      "grad_norm": 1.53125,
+      "learning_rate": 1.727364461484939e-05,
+      "loss": 1.0173,
+      "step": 4288
+    },
+    {
+      "epoch": 0.735409477677519,
+      "grad_norm": 1.6328125,
+      "learning_rate": 1.727240517556479e-05,
+      "loss": 1.078,
+      "step": 4289
+    },
+    {
+      "epoch": 0.7355809417665088,
+      "grad_norm": 1.59375,
+      "learning_rate": 1.7271165499099682e-05,
+      "loss": 1.0485,
+      "step": 4290
+    },
+    {
+      "epoch": 0.7357524058554986,
+      "grad_norm": 1.5625,
+      "learning_rate": 1.7269925585494483e-05,
+      "loss": 0.9896,
+      "step": 4291
+    },
+    {
+      "epoch": 0.7359238699444886,
+      "grad_norm": 1.59375,
+      "learning_rate": 1.7268685434789634e-05,
+      "loss": 1.0272,
+      "step": 4292
+    },
+    {
+      "epoch": 0.7360953340334784,
+      "grad_norm": 1.546875,
+      "learning_rate": 1.7267445047025582e-05,
+      "loss": 0.9564,
+      "step": 4293
+    },
+    {
+      "epoch": 0.7362667981224682,
+      "grad_norm": 1.5390625,
+      "learning_rate": 1.726620442224278e-05,
+      "loss": 1.0263,
+      "step": 4294
+    },
+    {
+      "epoch": 0.7364382622114581,
+      "grad_norm": 1.59375,
+      "learning_rate": 1.7264963560481688e-05,
+      "loss": 1.0433,
+      "step": 4295
+    },
+    {
+      "epoch": 0.736609726300448,
+      "grad_norm": 1.65625,
+      "learning_rate": 1.7263722461782782e-05,
+      "loss": 0.9827,
+      "step": 4296
+    },
+    {
+      "epoch": 0.7367811903894378,
+      "grad_norm": 1.6015625,
+      "learning_rate": 1.726248112618653e-05,
+      "loss": 1.0504,
+      "step": 4297
+    },
+    {
+      "epoch": 0.7369526544784277,
+      "grad_norm": 1.5546875,
+      "learning_rate": 1.7261239553733424e-05,
+      "loss": 0.9999,
+      "step": 4298
+    },
+    {
+      "epoch": 0.7371241185674176,
+      "grad_norm": 1.5234375,
+      "learning_rate": 1.7259997744463948e-05,
+      "loss": 0.9925,
+      "step": 4299
+    },
+    {
+      "epoch": 0.7372955826564074,
+      "grad_norm": 1.59375,
+      "learning_rate": 1.7258755698418606e-05,
+      "loss": 0.9823,
+      "step": 4300
+    },
+    {
+      "epoch": 0.7374670467453973,
+      "grad_norm": 1.640625,
+      "learning_rate": 1.7257513415637913e-05,
+      "loss": 1.0207,
+      "step": 4301
+    },
+    {
+      "epoch": 0.7376385108343871,
+      "grad_norm": 1.671875,
+      "learning_rate": 1.7256270896162373e-05,
+      "loss": 0.9931,
+      "step": 4302
+    },
+    {
+      "epoch": 0.737809974923377,
+      "grad_norm": 1.5234375,
+      "learning_rate": 1.7255028140032517e-05,
+      "loss": 0.985,
+      "step": 4303
+    },
+    {
+      "epoch": 0.7379814390123669,
+      "grad_norm": 1.625,
+      "learning_rate": 1.7253785147288875e-05,
+      "loss": 1.0894,
+      "step": 4304
+    },
+    {
+      "epoch": 0.7381529031013567,
+      "grad_norm": 1.671875,
+      "learning_rate": 1.7252541917971984e-05,
+      "loss": 1.0508,
+      "step": 4305
+    },
+    {
+      "epoch": 0.7383243671903466,
+      "grad_norm": 1.6484375,
+      "learning_rate": 1.7251298452122385e-05,
+      "loss": 1.0104,
+      "step": 4306
+    },
+    {
+      "epoch": 0.7384958312793364,
+      "grad_norm": 1.625,
+      "learning_rate": 1.7250054749780643e-05,
+      "loss": 1.0856,
+      "step": 4307
+    },
+    {
+      "epoch": 0.7386672953683263,
+      "grad_norm": 1.609375,
+      "learning_rate": 1.7248810810987312e-05,
+      "loss": 0.968,
+      "step": 4308
+    },
+    {
+      "epoch": 0.7388387594573161,
+      "grad_norm": 1.59375,
+      "learning_rate": 1.7247566635782965e-05,
+      "loss": 1.0389,
+      "step": 4309
+    },
+    {
+      "epoch": 0.739010223546306,
+      "grad_norm": 1.578125,
+      "learning_rate": 1.724632222420818e-05,
+      "loss": 0.9872,
+      "step": 4310
+    },
+    {
+      "epoch": 0.7391816876352959,
+      "grad_norm": 1.59375,
+      "learning_rate": 1.7245077576303537e-05,
+      "loss": 1.0246,
+      "step": 4311
+    },
+    {
+      "epoch": 0.7393531517242857,
+      "grad_norm": 1.7421875,
+      "learning_rate": 1.7243832692109632e-05,
+      "loss": 0.9628,
+      "step": 4312
+    },
+    {
+      "epoch": 0.7395246158132756,
+      "grad_norm": 1.5390625,
+      "learning_rate": 1.7242587571667067e-05,
+      "loss": 1.0645,
+      "step": 4313
+    },
+    {
+      "epoch": 0.7396960799022655,
+      "grad_norm": 1.53125,
+      "learning_rate": 1.7241342215016446e-05,
+      "loss": 0.9326,
+      "step": 4314
+    },
+    {
+      "epoch": 0.7398675439912553,
+      "grad_norm": 1.65625,
+      "learning_rate": 1.724009662219839e-05,
+      "loss": 1.1156,
+      "step": 4315
+    },
+    {
+      "epoch": 0.7400390080802451,
+      "grad_norm": 1.5390625,
+      "learning_rate": 1.7238850793253516e-05,
+      "loss": 0.974,
+      "step": 4316
+    },
+    {
+      "epoch": 0.7402104721692351,
+      "grad_norm": 1.5234375,
+      "learning_rate": 1.723760472822246e-05,
+      "loss": 0.9032,
+      "step": 4317
+    },
+    {
+      "epoch": 0.7403819362582249,
+      "grad_norm": 1.8671875,
+      "learning_rate": 1.723635842714586e-05,
+      "loss": 0.9871,
+      "step": 4318
+    },
+    {
+      "epoch": 0.7405534003472147,
+      "grad_norm": 1.7734375,
+      "learning_rate": 1.7235111890064363e-05,
+      "loss": 1.0518,
+      "step": 4319
+    },
+    {
+      "epoch": 0.7407248644362047,
+      "grad_norm": 1.6875,
+      "learning_rate": 1.723386511701862e-05,
+      "loss": 0.998,
+      "step": 4320
+    },
+    {
+      "epoch": 0.7408963285251945,
+      "grad_norm": 3.078125,
+      "learning_rate": 1.72326181080493e-05,
+      "loss": 1.138,
+      "step": 4321
+    },
+    {
+      "epoch": 0.7410677926141843,
+      "grad_norm": 2.921875,
+      "learning_rate": 1.723137086319706e-05,
+      "loss": 0.8734,
+      "step": 4322
+    },
+    {
+      "epoch": 0.7412392567031743,
+      "grad_norm": 4.34375,
+      "learning_rate": 1.7230123382502592e-05,
+      "loss": 1.0155,
+      "step": 4323
+    },
+    {
+      "epoch": 0.7414107207921641,
+      "grad_norm": 2.140625,
+      "learning_rate": 1.7228875666006572e-05,
+      "loss": 1.0432,
+      "step": 4324
+    },
+    {
+      "epoch": 0.7415821848811539,
+      "grad_norm": 1.765625,
+      "learning_rate": 1.72276277137497e-05,
+      "loss": 1.0208,
+      "step": 4325
+    },
+    {
+      "epoch": 0.7417536489701438,
+      "grad_norm": 1.6796875,
+      "learning_rate": 1.7226379525772664e-05,
+      "loss": 1.0444,
+      "step": 4326
+    },
+    {
+      "epoch": 0.7419251130591337,
+      "grad_norm": 1.5625,
+      "learning_rate": 1.7225131102116185e-05,
+      "loss": 1.0232,
+      "step": 4327
+    },
+    {
+      "epoch": 0.7420965771481235,
+      "grad_norm": 1.5546875,
+      "learning_rate": 1.7223882442820968e-05,
+      "loss": 1.0519,
+      "step": 4328
+    },
+    {
+      "epoch": 0.7422680412371134,
+      "grad_norm": 1.671875,
+      "learning_rate": 1.7222633547927745e-05,
+      "loss": 1.0556,
+      "step": 4329
+    },
+    {
+      "epoch": 0.7424395053261033,
+      "grad_norm": 1.6796875,
+      "learning_rate": 1.7221384417477244e-05,
+      "loss": 1.0376,
+      "step": 4330
+    },
+    {
+      "epoch": 0.7426109694150931,
+      "grad_norm": 1.5859375,
+      "learning_rate": 1.7220135051510202e-05,
+      "loss": 1.0469,
+      "step": 4331
+    },
+    {
+      "epoch": 0.742782433504083,
+      "grad_norm": 1.53125,
+      "learning_rate": 1.721888545006737e-05,
+      "loss": 0.9343,
+      "step": 4332
+    },
+    {
+      "epoch": 0.7429538975930728,
+      "grad_norm": 2.390625,
+      "learning_rate": 1.72176356131895e-05,
+      "loss": 1.045,
+      "step": 4333
+    },
+    {
+      "epoch": 0.7431253616820627,
+      "grad_norm": 1.828125,
+      "learning_rate": 1.7216385540917353e-05,
+      "loss": 1.0854,
+      "step": 4334
+    },
+    {
+      "epoch": 0.7432968257710526,
+      "grad_norm": 1.765625,
+      "learning_rate": 1.7215135233291698e-05,
+      "loss": 1.0709,
+      "step": 4335
+    },
+    {
+      "epoch": 0.7434682898600424,
+      "grad_norm": 1.625,
+      "learning_rate": 1.7213884690353317e-05,
+      "loss": 1.0443,
+      "step": 4336
+    },
+    {
+      "epoch": 0.7436397539490323,
+      "grad_norm": 1.953125,
+      "learning_rate": 1.7212633912142986e-05,
+      "loss": 1.0119,
+      "step": 4337
+    },
+    {
+      "epoch": 0.7438112180380222,
+      "grad_norm": 1.4921875,
+      "learning_rate": 1.7211382898701506e-05,
+      "loss": 0.9788,
+      "step": 4338
+    },
+    {
+      "epoch": 0.743982682127012,
+      "grad_norm": 1.625,
+      "learning_rate": 1.7210131650069674e-05,
+      "loss": 0.9576,
+      "step": 4339
+    },
+    {
+      "epoch": 0.7441541462160018,
+      "grad_norm": 1.546875,
+      "learning_rate": 1.7208880166288294e-05,
+      "loss": 0.9501,
+      "step": 4340
+    },
+    {
+      "epoch": 0.7443256103049918,
+      "grad_norm": 1.671875,
+      "learning_rate": 1.7207628447398192e-05,
+      "loss": 1.085,
+      "step": 4341
+    },
+    {
+      "epoch": 0.7444970743939816,
+      "grad_norm": 1.625,
+      "learning_rate": 1.720637649344018e-05,
+      "loss": 0.9839,
+      "step": 4342
+    },
+    {
+      "epoch": 0.7446685384829714,
+      "grad_norm": 1.6875,
+      "learning_rate": 1.7205124304455098e-05,
+      "loss": 1.0275,
+      "step": 4343
+    },
+    {
+      "epoch": 0.7448400025719614,
+      "grad_norm": 1.5859375,
+      "learning_rate": 1.7203871880483776e-05,
+      "loss": 0.9456,
+      "step": 4344
+    },
+    {
+      "epoch": 0.7450114666609512,
+      "grad_norm": 1.515625,
+      "learning_rate": 1.720261922156707e-05,
+      "loss": 0.978,
+      "step": 4345
+    },
+    {
+      "epoch": 0.745182930749941,
+      "grad_norm": 1.59375,
+      "learning_rate": 1.7201366327745824e-05,
+      "loss": 0.9333,
+      "step": 4346
+    },
+    {
+      "epoch": 0.745354394838931,
+      "grad_norm": 1.515625,
+      "learning_rate": 1.7200113199060905e-05,
+      "loss": 1.0612,
+      "step": 4347
+    },
+    {
+      "epoch": 0.7455258589279208,
+      "grad_norm": 1.5859375,
+      "learning_rate": 1.7198859835553184e-05,
+      "loss": 1.0532,
+      "step": 4348
+    },
+    {
+      "epoch": 0.7456973230169106,
+      "grad_norm": 1.59375,
+      "learning_rate": 1.7197606237263533e-05,
+      "loss": 1.0471,
+      "step": 4349
+    },
+    {
+      "epoch": 0.7458687871059005,
+      "grad_norm": 1.5078125,
+      "learning_rate": 1.719635240423284e-05,
+      "loss": 0.9832,
+      "step": 4350
+    },
+    {
+      "epoch": 0.7460402511948904,
+      "grad_norm": 1.5546875,
+      "learning_rate": 1.7195098336501997e-05,
+      "loss": 0.962,
+      "step": 4351
+    },
+    {
+      "epoch": 0.7462117152838802,
+      "grad_norm": 1.640625,
+      "learning_rate": 1.71938440341119e-05,
+      "loss": 1.0084,
+      "step": 4352
+    },
+    {
+      "epoch": 0.7463831793728701,
+      "grad_norm": 1.546875,
+      "learning_rate": 1.7192589497103462e-05,
+      "loss": 1.0764,
+      "step": 4353
+    },
+    {
+      "epoch": 0.74655464346186,
+      "grad_norm": 1.5703125,
+      "learning_rate": 1.7191334725517593e-05,
+      "loss": 0.95,
+      "step": 4354
+    },
+    {
+      "epoch": 0.7467261075508498,
+      "grad_norm": 1.609375,
+      "learning_rate": 1.7190079719395222e-05,
+      "loss": 0.9804,
+      "step": 4355
+    },
+    {
+      "epoch": 0.7468975716398397,
+      "grad_norm": 1.5078125,
+      "learning_rate": 1.7188824478777275e-05,
+      "loss": 0.9459,
+      "step": 4356
+    },
+    {
+      "epoch": 0.7470690357288295,
+      "grad_norm": 1.4765625,
+      "learning_rate": 1.7187569003704688e-05,
+      "loss": 0.9771,
+      "step": 4357
+    },
+    {
+      "epoch": 0.7472404998178194,
+      "grad_norm": 1.5703125,
+      "learning_rate": 1.7186313294218416e-05,
+      "loss": 0.9737,
+      "step": 4358
+    },
+    {
+      "epoch": 0.7474119639068093,
+      "grad_norm": 1.6015625,
+      "learning_rate": 1.71850573503594e-05,
+      "loss": 1.0141,
+      "step": 4359
+    },
+    {
+      "epoch": 0.7475834279957991,
+      "grad_norm": 1.515625,
+      "learning_rate": 1.718380117216861e-05,
+      "loss": 0.9638,
+      "step": 4360
+    },
+    {
+      "epoch": 0.747754892084789,
+      "grad_norm": 1.5078125,
+      "learning_rate": 1.7182544759687015e-05,
+      "loss": 0.9517,
+      "step": 4361
+    },
+    {
+      "epoch": 0.7479263561737789,
+      "grad_norm": 1.5546875,
+      "learning_rate": 1.7181288112955586e-05,
+      "loss": 1.0122,
+      "step": 4362
+    },
+    {
+      "epoch": 0.7480978202627687,
+      "grad_norm": 1.4609375,
+      "learning_rate": 1.7180031232015307e-05,
+      "loss": 0.9521,
+      "step": 4363
+    },
+    {
+      "epoch": 0.7482692843517585,
+      "grad_norm": 1.65625,
+      "learning_rate": 1.7178774116907177e-05,
+      "loss": 0.994,
+      "step": 4364
+    },
+    {
+      "epoch": 0.7484407484407485,
+      "grad_norm": 1.5546875,
+      "learning_rate": 1.7177516767672187e-05,
+      "loss": 1.0372,
+      "step": 4365
+    },
+    {
+      "epoch": 0.7486122125297383,
+      "grad_norm": 1.5078125,
+      "learning_rate": 1.717625918435135e-05,
+      "loss": 0.9872,
+      "step": 4366
+    },
+    {
+      "epoch": 0.7487836766187281,
+      "grad_norm": 1.625,
+      "learning_rate": 1.717500136698567e-05,
+      "loss": 1.1055,
+      "step": 4367
+    },
+    {
+      "epoch": 0.7489551407077181,
+      "grad_norm": 1.6640625,
+      "learning_rate": 1.7173743315616184e-05,
+      "loss": 1.1213,
+      "step": 4368
+    },
+    {
+      "epoch": 0.7491266047967079,
+      "grad_norm": 1.6015625,
+      "learning_rate": 1.7172485030283912e-05,
+      "loss": 1.0286,
+      "step": 4369
+    },
+    {
+      "epoch": 0.7492980688856977,
+      "grad_norm": 1.609375,
+      "learning_rate": 1.7171226511029895e-05,
+      "loss": 1.0856,
+      "step": 4370
+    },
+    {
+      "epoch": 0.7494695329746877,
+      "grad_norm": 1.53125,
+      "learning_rate": 1.716996775789518e-05,
+      "loss": 1.0403,
+      "step": 4371
+    },
+    {
+      "epoch": 0.7496409970636775,
+      "grad_norm": 1.609375,
+      "learning_rate": 1.7168708770920815e-05,
+      "loss": 1.0312,
+      "step": 4372
+    },
+    {
+      "epoch": 0.7498124611526673,
+      "grad_norm": 1.484375,
+      "learning_rate": 1.716744955014786e-05,
+      "loss": 0.9761,
+      "step": 4373
+    },
+    {
+      "epoch": 0.7499839252416572,
+      "grad_norm": 1.5625,
+      "learning_rate": 1.7166190095617386e-05,
+      "loss": 1.0441,
+      "step": 4374
+    },
+    {
+      "epoch": 0.7501553893306471,
+      "grad_norm": 1.5390625,
+      "learning_rate": 1.7164930407370465e-05,
+      "loss": 0.9607,
+      "step": 4375
+    },
+    {
+      "epoch": 0.7503268534196369,
+      "grad_norm": 1.6796875,
+      "learning_rate": 1.716367048544819e-05,
+      "loss": 1.0862,
+      "step": 4376
+    },
+    {
+      "epoch": 0.7504983175086268,
+      "grad_norm": 1.53125,
+      "learning_rate": 1.7162410329891636e-05,
+      "loss": 1.0016,
+      "step": 4377
+    },
+    {
+      "epoch": 0.7506697815976167,
+      "grad_norm": 1.5078125,
+      "learning_rate": 1.7161149940741918e-05,
+      "loss": 1.0165,
+      "step": 4378
+    },
+    {
+      "epoch": 0.7508412456866065,
+      "grad_norm": 1.5390625,
+      "learning_rate": 1.7159889318040128e-05,
+      "loss": 1.0394,
+      "step": 4379
+    },
+    {
+      "epoch": 0.7510127097755964,
+      "grad_norm": 1.5859375,
+      "learning_rate": 1.7158628461827386e-05,
+      "loss": 0.9978,
+      "step": 4380
+    },
+    {
+      "epoch": 0.7511841738645862,
+      "grad_norm": 1.5,
+      "learning_rate": 1.715736737214482e-05,
+      "loss": 0.9515,
+      "step": 4381
+    },
+    {
+      "epoch": 0.7513556379535761,
+      "grad_norm": 1.671875,
+      "learning_rate": 1.7156106049033544e-05,
+      "loss": 1.0019,
+      "step": 4382
+    },
+    {
+      "epoch": 0.751527102042566,
+      "grad_norm": 1.7890625,
+      "learning_rate": 1.7154844492534704e-05,
+      "loss": 1.1323,
+      "step": 4383
+    },
+    {
+      "epoch": 0.7516985661315558,
+      "grad_norm": 1.5546875,
+      "learning_rate": 1.7153582702689445e-05,
+      "loss": 1.0557,
+      "step": 4384
+    },
+    {
+      "epoch": 0.7518700302205457,
+      "grad_norm": 1.5703125,
+      "learning_rate": 1.7152320679538914e-05,
+      "loss": 1.0072,
+      "step": 4385
+    },
+    {
+      "epoch": 0.7520414943095356,
+      "grad_norm": 1.4453125,
+      "learning_rate": 1.7151058423124275e-05,
+      "loss": 0.889,
+      "step": 4386
+    },
+    {
+      "epoch": 0.7522129583985254,
+      "grad_norm": 1.5703125,
+      "learning_rate": 1.714979593348669e-05,
+      "loss": 1.0677,
+      "step": 4387
+    },
+    {
+      "epoch": 0.7523844224875152,
+      "grad_norm": 1.609375,
+      "learning_rate": 1.7148533210667337e-05,
+      "loss": 1.0883,
+      "step": 4388
+    },
+    {
+      "epoch": 0.7525558865765052,
+      "grad_norm": 1.5,
+      "learning_rate": 1.7147270254707394e-05,
+      "loss": 0.9649,
+      "step": 4389
+    },
+    {
+      "epoch": 0.752727350665495,
+      "grad_norm": 1.5859375,
+      "learning_rate": 1.714600706564806e-05,
+      "loss": 1.0362,
+      "step": 4390
+    },
+    {
+      "epoch": 0.7528988147544848,
+      "grad_norm": 1.59375,
+      "learning_rate": 1.714474364353052e-05,
+      "loss": 1.0016,
+      "step": 4391
+    },
+    {
+      "epoch": 0.7530702788434748,
+      "grad_norm": 1.53125,
+      "learning_rate": 1.714347998839599e-05,
+      "loss": 0.9474,
+      "step": 4392
+    },
+    {
+      "epoch": 0.7532417429324646,
+      "grad_norm": 1.6171875,
+      "learning_rate": 1.7142216100285672e-05,
+      "loss": 1.0142,
+      "step": 4393
+    },
+    {
+      "epoch": 0.7534132070214544,
+      "grad_norm": 1.6796875,
+      "learning_rate": 1.7140951979240797e-05,
+      "loss": 1.0681,
+      "step": 4394
+    },
+    {
+      "epoch": 0.7535846711104444,
+      "grad_norm": 1.6328125,
+      "learning_rate": 1.7139687625302587e-05,
+      "loss": 1.041,
+      "step": 4395
+    },
+    {
+      "epoch": 0.7537561351994342,
+      "grad_norm": 1.6015625,
+      "learning_rate": 1.7138423038512275e-05,
+      "loss": 0.9334,
+      "step": 4396
+    },
+    {
+      "epoch": 0.753927599288424,
+      "grad_norm": 1.5390625,
+      "learning_rate": 1.713715821891111e-05,
+      "loss": 0.9708,
+      "step": 4397
+    },
+    {
+      "epoch": 0.754099063377414,
+      "grad_norm": 1.6015625,
+      "learning_rate": 1.713589316654034e-05,
+      "loss": 0.9977,
+      "step": 4398
+    },
+    {
+      "epoch": 0.7542705274664038,
+      "grad_norm": 1.53125,
+      "learning_rate": 1.713462788144122e-05,
+      "loss": 0.9526,
+      "step": 4399
+    },
+    {
+      "epoch": 0.7544419915553936,
+      "grad_norm": 1.5390625,
+      "learning_rate": 1.713336236365502e-05,
+      "loss": 1.0819,
+      "step": 4400
+    },
+    {
+      "epoch": 0.7546134556443834,
+      "grad_norm": 1.5078125,
+      "learning_rate": 1.713209661322301e-05,
+      "loss": 0.9711,
+      "step": 4401
+    },
+    {
+      "epoch": 0.7547849197333734,
+      "grad_norm": 1.640625,
+      "learning_rate": 1.7130830630186476e-05,
+      "loss": 1.0617,
+      "step": 4402
+    },
+    {
+      "epoch": 0.7549563838223632,
+      "grad_norm": 1.5546875,
+      "learning_rate": 1.7129564414586698e-05,
+      "loss": 1.0486,
+      "step": 4403
+    },
+    {
+      "epoch": 0.755127847911353,
+      "grad_norm": 1.671875,
+      "learning_rate": 1.712829796646498e-05,
+      "loss": 1.0216,
+      "step": 4404
+    },
+    {
+      "epoch": 0.755299312000343,
+      "grad_norm": 1.546875,
+      "learning_rate": 1.712703128586262e-05,
+      "loss": 0.9582,
+      "step": 4405
+    },
+    {
+      "epoch": 0.7554707760893328,
+      "grad_norm": 1.625,
+      "learning_rate": 1.712576437282093e-05,
+      "loss": 1.029,
+      "step": 4406
+    },
+    {
+      "epoch": 0.7556422401783226,
+      "grad_norm": 1.546875,
+      "learning_rate": 1.7124497227381238e-05,
+      "loss": 1.0258,
+      "step": 4407
+    },
+    {
+      "epoch": 0.7558137042673125,
+      "grad_norm": 1.4921875,
+      "learning_rate": 1.712322984958486e-05,
+      "loss": 0.9825,
+      "step": 4408
+    },
+    {
+      "epoch": 0.7559851683563024,
+      "grad_norm": 1.6875,
+      "learning_rate": 1.7121962239473134e-05,
+      "loss": 1.0222,
+      "step": 4409
+    },
+    {
+      "epoch": 0.7561566324452922,
+      "grad_norm": 1.5703125,
+      "learning_rate": 1.7120694397087396e-05,
+      "loss": 0.9514,
+      "step": 4410
+    },
+    {
+      "epoch": 0.7563280965342821,
+      "grad_norm": 1.546875,
+      "learning_rate": 1.7119426322469002e-05,
+      "loss": 0.9866,
+      "step": 4411
+    },
+    {
+      "epoch": 0.756499560623272,
+      "grad_norm": 1.5703125,
+      "learning_rate": 1.7118158015659308e-05,
+      "loss": 0.9585,
+      "step": 4412
+    },
+    {
+      "epoch": 0.7566710247122618,
+      "grad_norm": 1.5390625,
+      "learning_rate": 1.7116889476699675e-05,
+      "loss": 0.985,
+      "step": 4413
+    },
+    {
+      "epoch": 0.7568424888012517,
+      "grad_norm": 1.5625,
+      "learning_rate": 1.7115620705631477e-05,
+      "loss": 0.9986,
+      "step": 4414
+    },
+    {
+      "epoch": 0.7570139528902415,
+      "grad_norm": 1.5234375,
+      "learning_rate": 1.711435170249609e-05,
+      "loss": 0.9748,
+      "step": 4415
+    },
+    {
+      "epoch": 0.7571854169792314,
+      "grad_norm": 1.5703125,
+      "learning_rate": 1.711308246733491e-05,
+      "loss": 0.9782,
+      "step": 4416
+    },
+    {
+      "epoch": 0.7573568810682213,
+      "grad_norm": 1.625,
+      "learning_rate": 1.7111813000189322e-05,
+      "loss": 0.9702,
+      "step": 4417
+    },
+    {
+      "epoch": 0.7575283451572111,
+      "grad_norm": 1.5,
+      "learning_rate": 1.711054330110073e-05,
+      "loss": 0.9542,
+      "step": 4418
+    },
+    {
+      "epoch": 0.757699809246201,
+      "grad_norm": 1.7109375,
+      "learning_rate": 1.7109273370110547e-05,
+      "loss": 1.0391,
+      "step": 4419
+    },
+    {
+      "epoch": 0.7578712733351909,
+      "grad_norm": 1.5859375,
+      "learning_rate": 1.7108003207260188e-05,
+      "loss": 0.9986,
+      "step": 4420
+    },
+    {
+      "epoch": 0.7580427374241807,
+      "grad_norm": 1.6171875,
+      "learning_rate": 1.7106732812591077e-05,
+      "loss": 1.0372,
+      "step": 4421
+    },
+    {
+      "epoch": 0.7582142015131705,
+      "grad_norm": 1.578125,
+      "learning_rate": 1.7105462186144652e-05,
+      "loss": 1.0716,
+      "step": 4422
+    },
+    {
+      "epoch": 0.7583856656021605,
+      "grad_norm": 1.6328125,
+      "learning_rate": 1.7104191327962345e-05,
+      "loss": 1.0082,
+      "step": 4423
+    },
+    {
+      "epoch": 0.7585571296911503,
+      "grad_norm": 1.703125,
+      "learning_rate": 1.7102920238085606e-05,
+      "loss": 0.9464,
+      "step": 4424
+    },
+    {
+      "epoch": 0.7587285937801401,
+      "grad_norm": 1.6015625,
+      "learning_rate": 1.710164891655589e-05,
+      "loss": 1.0328,
+      "step": 4425
+    },
+    {
+      "epoch": 0.7589000578691301,
+      "grad_norm": 1.59375,
+      "learning_rate": 1.7100377363414665e-05,
+      "loss": 0.9934,
+      "step": 4426
+    },
+    {
+      "epoch": 0.7590715219581199,
+      "grad_norm": 1.5703125,
+      "learning_rate": 1.7099105578703393e-05,
+      "loss": 1.0205,
+      "step": 4427
+    },
+    {
+      "epoch": 0.7592429860471097,
+      "grad_norm": 1.53125,
+      "learning_rate": 1.7097833562463556e-05,
+      "loss": 0.9942,
+      "step": 4428
+    },
+    {
+      "epoch": 0.7594144501360997,
+      "grad_norm": 1.5546875,
+      "learning_rate": 1.7096561314736638e-05,
+      "loss": 0.9955,
+      "step": 4429
+    },
+    {
+      "epoch": 0.7595859142250895,
+      "grad_norm": 1.5078125,
+      "learning_rate": 1.709528883556413e-05,
+      "loss": 0.936,
+      "step": 4430
+    },
+    {
+      "epoch": 0.7597573783140793,
+      "grad_norm": 1.65625,
+      "learning_rate": 1.709401612498754e-05,
+      "loss": 0.9424,
+      "step": 4431
+    },
+    {
+      "epoch": 0.7599288424030692,
+      "grad_norm": 1.6875,
+      "learning_rate": 1.709274318304837e-05,
+      "loss": 1.1093,
+      "step": 4432
+    },
+    {
+      "epoch": 0.7601003064920591,
+      "grad_norm": 1.5546875,
+      "learning_rate": 1.7091470009788135e-05,
+      "loss": 0.9639,
+      "step": 4433
+    },
+    {
+      "epoch": 0.7602717705810489,
+      "grad_norm": 1.5625,
+      "learning_rate": 1.7090196605248358e-05,
+      "loss": 0.974,
+      "step": 4434
+    },
+    {
+      "epoch": 0.7604432346700388,
+      "grad_norm": 1.578125,
+      "learning_rate": 1.708892296947057e-05,
+      "loss": 1.0629,
+      "step": 4435
+    },
+    {
+      "epoch": 0.7606146987590287,
+      "grad_norm": 1.609375,
+      "learning_rate": 1.708764910249631e-05,
+      "loss": 1.1132,
+      "step": 4436
+    },
+    {
+      "epoch": 0.7607861628480185,
+      "grad_norm": 1.546875,
+      "learning_rate": 1.7086375004367124e-05,
+      "loss": 0.994,
+      "step": 4437
+    },
+    {
+      "epoch": 0.7609576269370084,
+      "grad_norm": 1.640625,
+      "learning_rate": 1.708510067512456e-05,
+      "loss": 1.0587,
+      "step": 4438
+    },
+    {
+      "epoch": 0.7611290910259982,
+      "grad_norm": 1.5078125,
+      "learning_rate": 1.708382611481019e-05,
+      "loss": 0.9997,
+      "step": 4439
+    },
+    {
+      "epoch": 0.7613005551149881,
+      "grad_norm": 1.5703125,
+      "learning_rate": 1.7082551323465573e-05,
+      "loss": 1.0589,
+      "step": 4440
+    },
+    {
+      "epoch": 0.761472019203978,
+      "grad_norm": 1.6171875,
+      "learning_rate": 1.7081276301132284e-05,
+      "loss": 1.0245,
+      "step": 4441
+    },
+    {
+      "epoch": 0.7616434832929678,
+      "grad_norm": 1.484375,
+      "learning_rate": 1.7080001047851912e-05,
+      "loss": 0.9763,
+      "step": 4442
+    },
+    {
+      "epoch": 0.7618149473819577,
+      "grad_norm": 1.5234375,
+      "learning_rate": 1.7078725563666045e-05,
+      "loss": 0.8622,
+      "step": 4443
+    },
+    {
+      "epoch": 0.7619864114709476,
+      "grad_norm": 1.6484375,
+      "learning_rate": 1.707744984861628e-05,
+      "loss": 1.0999,
+      "step": 4444
+    },
+    {
+      "epoch": 0.7621578755599374,
+      "grad_norm": 1.609375,
+      "learning_rate": 1.7076173902744226e-05,
+      "loss": 1.0719,
+      "step": 4445
+    },
+    {
+      "epoch": 0.7623293396489272,
+      "grad_norm": 1.6171875,
+      "learning_rate": 1.7074897726091492e-05,
+      "loss": 1.0189,
+      "step": 4446
+    },
+    {
+      "epoch": 0.7625008037379172,
+      "grad_norm": 1.5859375,
+      "learning_rate": 1.7073621318699703e-05,
+      "loss": 0.9854,
+      "step": 4447
+    },
+    {
+      "epoch": 0.762672267826907,
+      "grad_norm": 1.5859375,
+      "learning_rate": 1.7072344680610485e-05,
+      "loss": 1.007,
+      "step": 4448
+    },
+    {
+      "epoch": 0.7628437319158968,
+      "grad_norm": 1.6015625,
+      "learning_rate": 1.7071067811865477e-05,
+      "loss": 1.0101,
+      "step": 4449
+    },
+    {
+      "epoch": 0.7630151960048868,
+      "grad_norm": 1.625,
+      "learning_rate": 1.7069790712506317e-05,
+      "loss": 0.9622,
+      "step": 4450
+    },
+    {
+      "epoch": 0.7631866600938766,
+      "grad_norm": 1.546875,
+      "learning_rate": 1.7068513382574665e-05,
+      "loss": 0.9194,
+      "step": 4451
+    },
+    {
+      "epoch": 0.7633581241828664,
+      "grad_norm": 1.546875,
+      "learning_rate": 1.7067235822112168e-05,
+      "loss": 1.0135,
+      "step": 4452
+    },
+    {
+      "epoch": 0.7635295882718564,
+      "grad_norm": 1.671875,
+      "learning_rate": 1.7065958031160503e-05,
+      "loss": 0.9483,
+      "step": 4453
+    },
+    {
+      "epoch": 0.7637010523608462,
+      "grad_norm": 1.6875,
+      "learning_rate": 1.7064680009761338e-05,
+      "loss": 1.0235,
+      "step": 4454
+    },
+    {
+      "epoch": 0.763872516449836,
+      "grad_norm": 1.5859375,
+      "learning_rate": 1.7063401757956353e-05,
+      "loss": 1.05,
+      "step": 4455
+    },
+    {
+      "epoch": 0.7640439805388259,
+      "grad_norm": 1.546875,
+      "learning_rate": 1.7062123275787236e-05,
+      "loss": 0.9476,
+      "step": 4456
+    },
+    {
+      "epoch": 0.7642154446278158,
+      "grad_norm": 1.6171875,
+      "learning_rate": 1.706084456329569e-05,
+      "loss": 1.05,
+      "step": 4457
+    },
+    {
+      "epoch": 0.7643869087168056,
+      "grad_norm": 1.5234375,
+      "learning_rate": 1.7059565620523414e-05,
+      "loss": 1.0445,
+      "step": 4458
+    },
+    {
+      "epoch": 0.7645583728057955,
+      "grad_norm": 1.625,
+      "learning_rate": 1.7058286447512115e-05,
+      "loss": 1.0614,
+      "step": 4459
+    },
+    {
+      "epoch": 0.7647298368947854,
+      "grad_norm": 1.484375,
+      "learning_rate": 1.7057007044303518e-05,
+      "loss": 1.0182,
+      "step": 4460
+    },
+    {
+      "epoch": 0.7649013009837752,
+      "grad_norm": 1.578125,
+      "learning_rate": 1.705572741093935e-05,
+      "loss": 0.9285,
+      "step": 4461
+    },
+    {
+      "epoch": 0.7650727650727651,
+      "grad_norm": 1.5625,
+      "learning_rate": 1.7054447547461337e-05,
+      "loss": 0.9481,
+      "step": 4462
+    },
+    {
+      "epoch": 0.7652442291617549,
+      "grad_norm": 1.515625,
+      "learning_rate": 1.705316745391123e-05,
+      "loss": 1.0469,
+      "step": 4463
+    },
+    {
+      "epoch": 0.7654156932507448,
+      "grad_norm": 1.5703125,
+      "learning_rate": 1.7051887130330767e-05,
+      "loss": 1.0112,
+      "step": 4464
+    },
+    {
+      "epoch": 0.7655871573397347,
+      "grad_norm": 1.5234375,
+      "learning_rate": 1.7050606576761714e-05,
+      "loss": 0.9453,
+      "step": 4465
+    },
+    {
+      "epoch": 0.7657586214287245,
+      "grad_norm": 1.5234375,
+      "learning_rate": 1.704932579324583e-05,
+      "loss": 0.9819,
+      "step": 4466
+    },
+    {
+      "epoch": 0.7659300855177144,
+      "grad_norm": 1.5703125,
+      "learning_rate": 1.7048044779824885e-05,
+      "loss": 1.0449,
+      "step": 4467
+    },
+    {
+      "epoch": 0.7661015496067043,
+      "grad_norm": 1.640625,
+      "learning_rate": 1.7046763536540657e-05,
+      "loss": 0.9743,
+      "step": 4468
+    },
+    {
+      "epoch": 0.7662730136956941,
+      "grad_norm": 1.75,
+      "learning_rate": 1.704548206343494e-05,
+      "loss": 1.0509,
+      "step": 4469
+    },
+    {
+      "epoch": 0.7664444777846839,
+      "grad_norm": 1.6171875,
+      "learning_rate": 1.704420036054952e-05,
+      "loss": 0.9419,
+      "step": 4470
+    },
+    {
+      "epoch": 0.7666159418736739,
+      "grad_norm": 1.5625,
+      "learning_rate": 1.70429184279262e-05,
+      "loss": 0.9786,
+      "step": 4471
+    },
+    {
+      "epoch": 0.7667874059626637,
+      "grad_norm": 1.4609375,
+      "learning_rate": 1.7041636265606786e-05,
+      "loss": 0.9696,
+      "step": 4472
+    },
+    {
+      "epoch": 0.7669588700516535,
+      "grad_norm": 1.5625,
+      "learning_rate": 1.7040353873633097e-05,
+      "loss": 1.0663,
+      "step": 4473
+    },
+    {
+      "epoch": 0.7671303341406435,
+      "grad_norm": 1.5390625,
+      "learning_rate": 1.703907125204696e-05,
+      "loss": 0.975,
+      "step": 4474
+    },
+    {
+      "epoch": 0.7673017982296333,
+      "grad_norm": 1.5,
+      "learning_rate": 1.7037788400890206e-05,
+      "loss": 0.963,
+      "step": 4475
+    },
+    {
+      "epoch": 0.7674732623186231,
+      "grad_norm": 1.65625,
+      "learning_rate": 1.7036505320204664e-05,
+      "loss": 1.0571,
+      "step": 4476
+    },
+    {
+      "epoch": 0.767644726407613,
+      "grad_norm": 1.5703125,
+      "learning_rate": 1.703522201003219e-05,
+      "loss": 0.9805,
+      "step": 4477
+    },
+    {
+      "epoch": 0.7678161904966029,
+      "grad_norm": 1.5703125,
+      "learning_rate": 1.703393847041463e-05,
+      "loss": 1.014,
+      "step": 4478
+    },
+    {
+      "epoch": 0.7679876545855927,
+      "grad_norm": 1.546875,
+      "learning_rate": 1.7032654701393853e-05,
+      "loss": 1.036,
+      "step": 4479
+    },
+    {
+      "epoch": 0.7681591186745826,
+      "grad_norm": 1.5859375,
+      "learning_rate": 1.7031370703011724e-05,
+      "loss": 1.0157,
+      "step": 4480
+    },
+    {
+      "epoch": 0.7683305827635725,
+      "grad_norm": 1.5625,
+      "learning_rate": 1.7030086475310116e-05,
+      "loss": 0.9388,
+      "step": 4481
+    },
+    {
+      "epoch": 0.7685020468525623,
+      "grad_norm": 1.578125,
+      "learning_rate": 1.7028802018330915e-05,
+      "loss": 0.9365,
+      "step": 4482
+    },
+    {
+      "epoch": 0.7686735109415522,
+      "grad_norm": 1.6171875,
+      "learning_rate": 1.7027517332116014e-05,
+      "loss": 1.0533,
+      "step": 4483
+    },
+    {
+      "epoch": 0.768844975030542,
+      "grad_norm": 1.4921875,
+      "learning_rate": 1.7026232416707312e-05,
+      "loss": 0.9399,
+      "step": 4484
+    },
+    {
+      "epoch": 0.7690164391195319,
+      "grad_norm": 1.515625,
+      "learning_rate": 1.7024947272146704e-05,
+      "loss": 1.0067,
+      "step": 4485
+    },
+    {
+      "epoch": 0.7691879032085218,
+      "grad_norm": 1.546875,
+      "learning_rate": 1.702366189847612e-05,
+      "loss": 0.9739,
+      "step": 4486
+    },
+    {
+      "epoch": 0.7693593672975116,
+      "grad_norm": 1.578125,
+      "learning_rate": 1.7022376295737473e-05,
+      "loss": 1.0058,
+      "step": 4487
+    },
+    {
+      "epoch": 0.7695308313865015,
+      "grad_norm": 1.6875,
+      "learning_rate": 1.702109046397269e-05,
+      "loss": 0.9879,
+      "step": 4488
+    },
+    {
+      "epoch": 0.7697022954754914,
+      "grad_norm": 1.5546875,
+      "learning_rate": 1.7019804403223702e-05,
+      "loss": 0.9565,
+      "step": 4489
+    },
+    {
+      "epoch": 0.7698737595644812,
+      "grad_norm": 1.625,
+      "learning_rate": 1.7018518113532467e-05,
+      "loss": 1.0858,
+      "step": 4490
+    },
+    {
+      "epoch": 0.7700452236534711,
+      "grad_norm": 1.4921875,
+      "learning_rate": 1.701723159494092e-05,
+      "loss": 1.0049,
+      "step": 4491
+    },
+    {
+      "epoch": 0.770216687742461,
+      "grad_norm": 1.578125,
+      "learning_rate": 1.7015944847491027e-05,
+      "loss": 1.0759,
+      "step": 4492
+    },
+    {
+      "epoch": 0.7703881518314508,
+      "grad_norm": 1.5625,
+      "learning_rate": 1.7014657871224758e-05,
+      "loss": 0.9866,
+      "step": 4493
+    },
+    {
+      "epoch": 0.7705596159204406,
+      "grad_norm": 1.515625,
+      "learning_rate": 1.7013370666184078e-05,
+      "loss": 0.9232,
+      "step": 4494
+    },
+    {
+      "epoch": 0.7707310800094306,
+      "grad_norm": 1.5859375,
+      "learning_rate": 1.701208323241097e-05,
+      "loss": 0.9587,
+      "step": 4495
+    },
+    {
+      "epoch": 0.7709025440984204,
+      "grad_norm": 1.53125,
+      "learning_rate": 1.7010795569947426e-05,
+      "loss": 0.9278,
+      "step": 4496
+    },
+    {
+      "epoch": 0.7710740081874102,
+      "grad_norm": 1.5859375,
+      "learning_rate": 1.7009507678835432e-05,
+      "loss": 1.004,
+      "step": 4497
+    },
+    {
+      "epoch": 0.7712454722764001,
+      "grad_norm": 1.65625,
+      "learning_rate": 1.7008219559117002e-05,
+      "loss": 1.0233,
+      "step": 4498
+    },
+    {
+      "epoch": 0.77141693636539,
+      "grad_norm": 1.4296875,
+      "learning_rate": 1.7006931210834144e-05,
+      "loss": 0.929,
+      "step": 4499
+    },
+    {
+      "epoch": 0.7715884004543798,
+      "grad_norm": 1.5859375,
+      "learning_rate": 1.700564263402887e-05,
+      "loss": 0.9735,
+      "step": 4500
+    },
+    {
+      "epoch": 0.7717598645433696,
+      "grad_norm": 1.578125,
+      "learning_rate": 1.7004353828743206e-05,
+      "loss": 1.009,
+      "step": 4501
+    },
+    {
+      "epoch": 0.7719313286323596,
+      "grad_norm": 1.5625,
+      "learning_rate": 1.7003064795019193e-05,
+      "loss": 1.008,
+      "step": 4502
+    },
+    {
+      "epoch": 0.7721027927213494,
+      "grad_norm": 1.6015625,
+      "learning_rate": 1.7001775532898865e-05,
+      "loss": 0.9422,
+      "step": 4503
+    },
+    {
+      "epoch": 0.7722742568103392,
+      "grad_norm": 1.6171875,
+      "learning_rate": 1.7000486042424268e-05,
+      "loss": 1.0274,
+      "step": 4504
+    },
+    {
+      "epoch": 0.7724457208993292,
+      "grad_norm": 1.5859375,
+      "learning_rate": 1.6999196323637463e-05,
+      "loss": 0.9829,
+      "step": 4505
+    },
+    {
+      "epoch": 0.772617184988319,
+      "grad_norm": 1.59375,
+      "learning_rate": 1.6997906376580506e-05,
+      "loss": 1.1202,
+      "step": 4506
+    },
+    {
+      "epoch": 0.7727886490773088,
+      "grad_norm": 1.4765625,
+      "learning_rate": 1.6996616201295472e-05,
+      "loss": 0.9832,
+      "step": 4507
+    },
+    {
+      "epoch": 0.7729601131662988,
+      "grad_norm": 1.546875,
+      "learning_rate": 1.6995325797824434e-05,
+      "loss": 0.9579,
+      "step": 4508
+    },
+    {
+      "epoch": 0.7731315772552886,
+      "grad_norm": 1.59375,
+      "learning_rate": 1.6994035166209483e-05,
+      "loss": 1.0081,
+      "step": 4509
+    },
+    {
+      "epoch": 0.7733030413442784,
+      "grad_norm": 1.4765625,
+      "learning_rate": 1.699274430649271e-05,
+      "loss": 0.9441,
+      "step": 4510
+    },
+    {
+      "epoch": 0.7734745054332683,
+      "grad_norm": 1.609375,
+      "learning_rate": 1.6991453218716212e-05,
+      "loss": 0.9585,
+      "step": 4511
+    },
+    {
+      "epoch": 0.7736459695222582,
+      "grad_norm": 1.5234375,
+      "learning_rate": 1.6990161902922094e-05,
+      "loss": 0.9768,
+      "step": 4512
+    },
+    {
+      "epoch": 0.773817433611248,
+      "grad_norm": 1.6015625,
+      "learning_rate": 1.6988870359152473e-05,
+      "loss": 1.0389,
+      "step": 4513
+    },
+    {
+      "epoch": 0.7739888977002379,
+      "grad_norm": 1.6640625,
+      "learning_rate": 1.6987578587449478e-05,
+      "loss": 0.9938,
+      "step": 4514
+    },
+    {
+      "epoch": 0.7741603617892278,
+      "grad_norm": 1.515625,
+      "learning_rate": 1.6986286587855226e-05,
+      "loss": 0.9872,
+      "step": 4515
+    },
+    {
+      "epoch": 0.7743318258782176,
+      "grad_norm": 1.546875,
+      "learning_rate": 1.6984994360411866e-05,
+      "loss": 0.9156,
+      "step": 4516
+    },
+    {
+      "epoch": 0.7745032899672075,
+      "grad_norm": 1.5546875,
+      "learning_rate": 1.6983701905161533e-05,
+      "loss": 0.9448,
+      "step": 4517
+    },
+    {
+      "epoch": 0.7746747540561973,
+      "grad_norm": 1.5,
+      "learning_rate": 1.6982409222146384e-05,
+      "loss": 0.9568,
+      "step": 4518
+    },
+    {
+      "epoch": 0.7748462181451872,
+      "grad_norm": 1.5234375,
+      "learning_rate": 1.698111631140858e-05,
+      "loss": 1.0083,
+      "step": 4519
+    },
+    {
+      "epoch": 0.7750176822341771,
+      "grad_norm": 1.578125,
+      "learning_rate": 1.697982317299028e-05,
+      "loss": 1.0531,
+      "step": 4520
+    },
+    {
+      "epoch": 0.7751891463231669,
+      "grad_norm": 1.5703125,
+      "learning_rate": 1.6978529806933665e-05,
+      "loss": 0.9687,
+      "step": 4521
+    },
+    {
+      "epoch": 0.7753606104121568,
+      "grad_norm": 1.46875,
+      "learning_rate": 1.6977236213280913e-05,
+      "loss": 0.9364,
+      "step": 4522
+    },
+    {
+      "epoch": 0.7755320745011467,
+      "grad_norm": 1.609375,
+      "learning_rate": 1.6975942392074217e-05,
+      "loss": 1.0638,
+      "step": 4523
+    },
+    {
+      "epoch": 0.7757035385901365,
+      "grad_norm": 1.5703125,
+      "learning_rate": 1.6974648343355765e-05,
+      "loss": 1.0329,
+      "step": 4524
+    },
+    {
+      "epoch": 0.7758750026791263,
+      "grad_norm": 1.5703125,
+      "learning_rate": 1.697335406716777e-05,
+      "loss": 0.9509,
+      "step": 4525
+    },
+    {
+      "epoch": 0.7760464667681163,
+      "grad_norm": 1.515625,
+      "learning_rate": 1.697205956355244e-05,
+      "loss": 0.9671,
+      "step": 4526
+    },
+    {
+      "epoch": 0.7762179308571061,
+      "grad_norm": 1.65625,
+      "learning_rate": 1.6970764832551996e-05,
+      "loss": 1.0382,
+      "step": 4527
+    },
+    {
+      "epoch": 0.7763893949460959,
+      "grad_norm": 1.5625,
+      "learning_rate": 1.6969469874208657e-05,
+      "loss": 0.992,
+      "step": 4528
+    },
+    {
+      "epoch": 0.7765608590350859,
+      "grad_norm": 1.5078125,
+      "learning_rate": 1.696817468856466e-05,
+      "loss": 0.9358,
+      "step": 4529
+    },
+    {
+      "epoch": 0.7767323231240757,
+      "grad_norm": 1.625,
+      "learning_rate": 1.6966879275662252e-05,
+      "loss": 1.0602,
+      "step": 4530
+    },
+    {
+      "epoch": 0.7769037872130655,
+      "grad_norm": 1.65625,
+      "learning_rate": 1.6965583635543673e-05,
+      "loss": 1.0713,
+      "step": 4531
+    },
+    {
+      "epoch": 0.7770752513020555,
+      "grad_norm": 1.5390625,
+      "learning_rate": 1.696428776825118e-05,
+      "loss": 1.0382,
+      "step": 4532
+    },
+    {
+      "epoch": 0.7772467153910453,
+      "grad_norm": 1.6875,
+      "learning_rate": 1.6962991673827038e-05,
+      "loss": 1.0826,
+      "step": 4533
+    },
+    {
+      "epoch": 0.7774181794800351,
+      "grad_norm": 1.6484375,
+      "learning_rate": 1.696169535231352e-05,
+      "loss": 1.0855,
+      "step": 4534
+    },
+    {
+      "epoch": 0.777589643569025,
+      "grad_norm": 1.59375,
+      "learning_rate": 1.69603988037529e-05,
+      "loss": 0.9902,
+      "step": 4535
+    },
+    {
+      "epoch": 0.7777611076580149,
+      "grad_norm": 1.609375,
+      "learning_rate": 1.695910202818746e-05,
+      "loss": 1.0447,
+      "step": 4536
+    },
+    {
+      "epoch": 0.7779325717470047,
+      "grad_norm": 1.6484375,
+      "learning_rate": 1.6957805025659504e-05,
+      "loss": 1.1,
+      "step": 4537
+    },
+    {
+      "epoch": 0.7781040358359946,
+      "grad_norm": 1.5390625,
+      "learning_rate": 1.6956507796211325e-05,
+      "loss": 0.9557,
+      "step": 4538
+    },
+    {
+      "epoch": 0.7782754999249845,
+      "grad_norm": 1.53125,
+      "learning_rate": 1.695521033988523e-05,
+      "loss": 0.9928,
+      "step": 4539
+    },
+    {
+      "epoch": 0.7784469640139743,
+      "grad_norm": 1.59375,
+      "learning_rate": 1.695391265672353e-05,
+      "loss": 0.9859,
+      "step": 4540
+    },
+    {
+      "epoch": 0.7786184281029642,
+      "grad_norm": 1.5390625,
+      "learning_rate": 1.6952614746768555e-05,
+      "loss": 1.0547,
+      "step": 4541
+    },
+    {
+      "epoch": 0.778789892191954,
+      "grad_norm": 1.625,
+      "learning_rate": 1.6951316610062634e-05,
+      "loss": 1.0959,
+      "step": 4542
+    },
+    {
+      "epoch": 0.7789613562809439,
+      "grad_norm": 1.59375,
+      "learning_rate": 1.6950018246648105e-05,
+      "loss": 1.0866,
+      "step": 4543
+    },
+    {
+      "epoch": 0.7791328203699338,
+      "grad_norm": 1.5546875,
+      "learning_rate": 1.6948719656567304e-05,
+      "loss": 0.9895,
+      "step": 4544
+    },
+    {
+      "epoch": 0.7793042844589236,
+      "grad_norm": 1.59375,
+      "learning_rate": 1.694742083986259e-05,
+      "loss": 0.9399,
+      "step": 4545
+    },
+    {
+      "epoch": 0.7794757485479135,
+      "grad_norm": 1.59375,
+      "learning_rate": 1.6946121796576324e-05,
+      "loss": 1.1096,
+      "step": 4546
+    },
+    {
+      "epoch": 0.7796472126369034,
+      "grad_norm": 1.5078125,
+      "learning_rate": 1.6944822526750865e-05,
+      "loss": 0.9154,
+      "step": 4547
+    },
+    {
+      "epoch": 0.7798186767258932,
+      "grad_norm": 1.5234375,
+      "learning_rate": 1.6943523030428592e-05,
+      "loss": 0.9304,
+      "step": 4548
+    },
+    {
+      "epoch": 0.779990140814883,
+      "grad_norm": 1.671875,
+      "learning_rate": 1.694222330765189e-05,
+      "loss": 1.0091,
+      "step": 4549
+    },
+    {
+      "epoch": 0.780161604903873,
+      "grad_norm": 1.5,
+      "learning_rate": 1.694092335846314e-05,
+      "loss": 1.0471,
+      "step": 4550
+    },
+    {
+      "epoch": 0.7803330689928628,
+      "grad_norm": 1.6484375,
+      "learning_rate": 1.6939623182904746e-05,
+      "loss": 1.0572,
+      "step": 4551
+    },
+    {
+      "epoch": 0.7805045330818526,
+      "grad_norm": 1.5078125,
+      "learning_rate": 1.6938322781019106e-05,
+      "loss": 0.8963,
+      "step": 4552
+    },
+    {
+      "epoch": 0.7806759971708426,
+      "grad_norm": 1.65625,
+      "learning_rate": 1.693702215284863e-05,
+      "loss": 1.0282,
+      "step": 4553
+    },
+    {
+      "epoch": 0.7808474612598324,
+      "grad_norm": 1.6484375,
+      "learning_rate": 1.693572129843574e-05,
+      "loss": 0.9697,
+      "step": 4554
+    },
+    {
+      "epoch": 0.7810189253488222,
+      "grad_norm": 1.6796875,
+      "learning_rate": 1.6934420217822864e-05,
+      "loss": 1.0408,
+      "step": 4555
+    },
+    {
+      "epoch": 0.7811903894378122,
+      "grad_norm": 1.515625,
+      "learning_rate": 1.693311891105243e-05,
+      "loss": 0.9593,
+      "step": 4556
+    },
+    {
+      "epoch": 0.781361853526802,
+      "grad_norm": 1.5625,
+      "learning_rate": 1.6931817378166885e-05,
+      "loss": 0.9947,
+      "step": 4557
+    },
+    {
+      "epoch": 0.7815333176157918,
+      "grad_norm": 1.734375,
+      "learning_rate": 1.6930515619208665e-05,
+      "loss": 1.0501,
+      "step": 4558
+    },
+    {
+      "epoch": 0.7817047817047817,
+      "grad_norm": 1.6171875,
+      "learning_rate": 1.6929213634220235e-05,
+      "loss": 1.0618,
+      "step": 4559
+    },
+    {
+      "epoch": 0.7818762457937716,
+      "grad_norm": 1.59375,
+      "learning_rate": 1.692791142324406e-05,
+      "loss": 1.0571,
+      "step": 4560
+    },
+    {
+      "epoch": 0.7820477098827614,
+      "grad_norm": 1.5390625,
+      "learning_rate": 1.69266089863226e-05,
+      "loss": 0.9615,
+      "step": 4561
+    },
+    {
+      "epoch": 0.7822191739717513,
+      "grad_norm": 1.6640625,
+      "learning_rate": 1.692530632349834e-05,
+      "loss": 1.072,
+      "step": 4562
+    },
+    {
+      "epoch": 0.7823906380607412,
+      "grad_norm": 1.5546875,
+      "learning_rate": 1.6924003434813763e-05,
+      "loss": 0.9741,
+      "step": 4563
+    },
+    {
+      "epoch": 0.782562102149731,
+      "grad_norm": 1.6015625,
+      "learning_rate": 1.6922700320311357e-05,
+      "loss": 1.0453,
+      "step": 4564
+    },
+    {
+      "epoch": 0.7827335662387209,
+      "grad_norm": 1.5546875,
+      "learning_rate": 1.692139698003363e-05,
+      "loss": 1.0435,
+      "step": 4565
+    },
+    {
+      "epoch": 0.7829050303277107,
+      "grad_norm": 1.4765625,
+      "learning_rate": 1.6920093414023083e-05,
+      "loss": 1.017,
+      "step": 4566
+    },
+    {
+      "epoch": 0.7830764944167006,
+      "grad_norm": 1.578125,
+      "learning_rate": 1.691878962232223e-05,
+      "loss": 1.0075,
+      "step": 4567
+    },
+    {
+      "epoch": 0.7832479585056905,
+      "grad_norm": 1.6171875,
+      "learning_rate": 1.6917485604973595e-05,
+      "loss": 0.9461,
+      "step": 4568
+    },
+    {
+      "epoch": 0.7834194225946803,
+      "grad_norm": 1.640625,
+      "learning_rate": 1.6916181362019704e-05,
+      "loss": 1.0606,
+      "step": 4569
+    },
+    {
+      "epoch": 0.7835908866836702,
+      "grad_norm": 1.578125,
+      "learning_rate": 1.6914876893503093e-05,
+      "loss": 1.0641,
+      "step": 4570
+    },
+    {
+      "epoch": 0.7837623507726601,
+      "grad_norm": 1.5625,
+      "learning_rate": 1.6913572199466312e-05,
+      "loss": 0.9939,
+      "step": 4571
+    },
+    {
+      "epoch": 0.7839338148616499,
+      "grad_norm": 1.5703125,
+      "learning_rate": 1.6912267279951904e-05,
+      "loss": 0.9261,
+      "step": 4572
+    },
+    {
+      "epoch": 0.7841052789506398,
+      "grad_norm": 1.5546875,
+      "learning_rate": 1.6910962135002433e-05,
+      "loss": 0.9663,
+      "step": 4573
+    },
+    {
+      "epoch": 0.7842767430396297,
+      "grad_norm": 1.5703125,
+      "learning_rate": 1.690965676466046e-05,
+      "loss": 1.0686,
+      "step": 4574
+    },
+    {
+      "epoch": 0.7844482071286195,
+      "grad_norm": 1.53125,
+      "learning_rate": 1.6908351168968563e-05,
+      "loss": 1.02,
+      "step": 4575
+    },
+    {
+      "epoch": 0.7846196712176093,
+      "grad_norm": 1.6484375,
+      "learning_rate": 1.690704534796932e-05,
+      "loss": 1.0676,
+      "step": 4576
+    },
+    {
+      "epoch": 0.7847911353065993,
+      "grad_norm": 1.640625,
+      "learning_rate": 1.6905739301705316e-05,
+      "loss": 1.0048,
+      "step": 4577
+    },
+    {
+      "epoch": 0.7849625993955891,
+      "grad_norm": 1.609375,
+      "learning_rate": 1.690443303021915e-05,
+      "loss": 1.0309,
+      "step": 4578
+    },
+    {
+      "epoch": 0.7851340634845789,
+      "grad_norm": 1.46875,
+      "learning_rate": 1.6903126533553425e-05,
+      "loss": 1.0769,
+      "step": 4579
+    },
+    {
+      "epoch": 0.7853055275735689,
+      "grad_norm": 1.5,
+      "learning_rate": 1.690181981175075e-05,
+      "loss": 0.9724,
+      "step": 4580
+    },
+    {
+      "epoch": 0.7854769916625587,
+      "grad_norm": 1.453125,
+      "learning_rate": 1.690051286485374e-05,
+      "loss": 1.0019,
+      "step": 4581
+    },
+    {
+      "epoch": 0.7856484557515485,
+      "grad_norm": 1.609375,
+      "learning_rate": 1.6899205692905016e-05,
+      "loss": 0.9549,
+      "step": 4582
+    },
+    {
+      "epoch": 0.7858199198405385,
+      "grad_norm": 1.578125,
+      "learning_rate": 1.6897898295947217e-05,
+      "loss": 0.9591,
+      "step": 4583
+    },
+    {
+      "epoch": 0.7859913839295283,
+      "grad_norm": 1.5234375,
+      "learning_rate": 1.6896590674022977e-05,
+      "loss": 0.9805,
+      "step": 4584
+    },
+    {
+      "epoch": 0.7861628480185181,
+      "grad_norm": 1.546875,
+      "learning_rate": 1.6895282827174952e-05,
+      "loss": 0.9356,
+      "step": 4585
+    },
+    {
+      "epoch": 0.786334312107508,
+      "grad_norm": 1.5546875,
+      "learning_rate": 1.6893974755445785e-05,
+      "loss": 1.0637,
+      "step": 4586
+    },
+    {
+      "epoch": 0.7865057761964979,
+      "grad_norm": 1.515625,
+      "learning_rate": 1.689266645887814e-05,
+      "loss": 0.9466,
+      "step": 4587
+    },
+    {
+      "epoch": 0.7866772402854877,
+      "grad_norm": 1.5859375,
+      "learning_rate": 1.6891357937514685e-05,
+      "loss": 0.9378,
+      "step": 4588
+    },
+    {
+      "epoch": 0.7868487043744776,
+      "grad_norm": 1.5546875,
+      "learning_rate": 1.6890049191398102e-05,
+      "loss": 0.9389,
+      "step": 4589
+    },
+    {
+      "epoch": 0.7870201684634675,
+      "grad_norm": 1.59375,
+      "learning_rate": 1.688874022057107e-05,
+      "loss": 1.0351,
+      "step": 4590
+    },
+    {
+      "epoch": 0.7871916325524573,
+      "grad_norm": 1.6328125,
+      "learning_rate": 1.688743102507627e-05,
+      "loss": 1.0766,
+      "step": 4591
+    },
+    {
+      "epoch": 0.7873630966414471,
+      "grad_norm": 1.609375,
+      "learning_rate": 1.6886121604956415e-05,
+      "loss": 1.1055,
+      "step": 4592
+    },
+    {
+      "epoch": 0.787534560730437,
+      "grad_norm": 1.5390625,
+      "learning_rate": 1.6884811960254203e-05,
+      "loss": 1.0666,
+      "step": 4593
+    },
+    {
+      "epoch": 0.7877060248194269,
+      "grad_norm": 1.6328125,
+      "learning_rate": 1.6883502091012346e-05,
+      "loss": 1.036,
+      "step": 4594
+    },
+    {
+      "epoch": 0.7878774889084167,
+      "grad_norm": 1.5625,
+      "learning_rate": 1.6882191997273567e-05,
+      "loss": 1.0623,
+      "step": 4595
+    },
+    {
+      "epoch": 0.7880489529974066,
+      "grad_norm": 1.578125,
+      "learning_rate": 1.6880881679080592e-05,
+      "loss": 0.9519,
+      "step": 4596
+    },
+    {
+      "epoch": 0.7882204170863965,
+      "grad_norm": 1.5546875,
+      "learning_rate": 1.687957113647615e-05,
+      "loss": 1.1319,
+      "step": 4597
+    },
+    {
+      "epoch": 0.7883918811753863,
+      "grad_norm": 1.5078125,
+      "learning_rate": 1.6878260369502993e-05,
+      "loss": 0.984,
+      "step": 4598
+    },
+    {
+      "epoch": 0.7885633452643762,
+      "grad_norm": 1.5703125,
+      "learning_rate": 1.6876949378203858e-05,
+      "loss": 1.0901,
+      "step": 4599
+    },
+    {
+      "epoch": 0.788734809353366,
+      "grad_norm": 1.5703125,
+      "learning_rate": 1.6875638162621512e-05,
+      "loss": 0.9269,
+      "step": 4600
+    },
+    {
+      "epoch": 0.7889062734423559,
+      "grad_norm": 1.5234375,
+      "learning_rate": 1.6874326722798713e-05,
+      "loss": 1.0038,
+      "step": 4601
+    },
+    {
+      "epoch": 0.7890777375313458,
+      "grad_norm": 1.6171875,
+      "learning_rate": 1.6873015058778232e-05,
+      "loss": 1.0358,
+      "step": 4602
+    },
+    {
+      "epoch": 0.7892492016203356,
+      "grad_norm": 1.6484375,
+      "learning_rate": 1.687170317060285e-05,
+      "loss": 1.0957,
+      "step": 4603
+    },
+    {
+      "epoch": 0.7894206657093255,
+      "grad_norm": 1.515625,
+      "learning_rate": 1.687039105831535e-05,
+      "loss": 0.9912,
+      "step": 4604
+    },
+    {
+      "epoch": 0.7895921297983154,
+      "grad_norm": 1.453125,
+      "learning_rate": 1.6869078721958528e-05,
+      "loss": 0.982,
+      "step": 4605
+    },
+    {
+      "epoch": 0.7897635938873052,
+      "grad_norm": 1.546875,
+      "learning_rate": 1.686776616157518e-05,
+      "loss": 0.9703,
+      "step": 4606
+    },
+    {
+      "epoch": 0.789935057976295,
+      "grad_norm": 1.5234375,
+      "learning_rate": 1.6866453377208115e-05,
+      "loss": 1.0642,
+      "step": 4607
+    },
+    {
+      "epoch": 0.790106522065285,
+      "grad_norm": 1.65625,
+      "learning_rate": 1.6865140368900153e-05,
+      "loss": 1.0153,
+      "step": 4608
+    },
+    {
+      "epoch": 0.7902779861542748,
+      "grad_norm": 1.59375,
+      "learning_rate": 1.686382713669411e-05,
+      "loss": 0.9728,
+      "step": 4609
+    },
+    {
+      "epoch": 0.7904494502432646,
+      "grad_norm": 1.578125,
+      "learning_rate": 1.6862513680632815e-05,
+      "loss": 0.9923,
+      "step": 4610
+    },
+    {
+      "epoch": 0.7906209143322546,
+      "grad_norm": 1.53125,
+      "learning_rate": 1.6861200000759108e-05,
+      "loss": 1.0171,
+      "step": 4611
+    },
+    {
+      "epoch": 0.7907923784212444,
+      "grad_norm": 1.6953125,
+      "learning_rate": 1.6859886097115833e-05,
+      "loss": 1.075,
+      "step": 4612
+    },
+    {
+      "epoch": 0.7909638425102342,
+      "grad_norm": 1.671875,
+      "learning_rate": 1.685857196974584e-05,
+      "loss": 0.9838,
+      "step": 4613
+    },
+    {
+      "epoch": 0.7911353065992242,
+      "grad_norm": 1.6328125,
+      "learning_rate": 1.6857257618691992e-05,
+      "loss": 1.0337,
+      "step": 4614
+    },
+    {
+      "epoch": 0.791306770688214,
+      "grad_norm": 1.609375,
+      "learning_rate": 1.6855943043997144e-05,
+      "loss": 0.9971,
+      "step": 4615
+    },
+    {
+      "epoch": 0.7914782347772038,
+      "grad_norm": 1.5703125,
+      "learning_rate": 1.685462824570418e-05,
+      "loss": 1.028,
+      "step": 4616
+    },
+    {
+      "epoch": 0.7916496988661937,
+      "grad_norm": 1.6796875,
+      "learning_rate": 1.6853313223855977e-05,
+      "loss": 1.0616,
+      "step": 4617
+    },
+    {
+      "epoch": 0.7918211629551836,
+      "grad_norm": 1.53125,
+      "learning_rate": 1.685199797849542e-05,
+      "loss": 1.0246,
+      "step": 4618
+    },
+    {
+      "epoch": 0.7919926270441734,
+      "grad_norm": 1.578125,
+      "learning_rate": 1.685068250966541e-05,
+      "loss": 1.0009,
+      "step": 4619
+    },
+    {
+      "epoch": 0.7921640911331633,
+      "grad_norm": 1.6015625,
+      "learning_rate": 1.6849366817408846e-05,
+      "loss": 1.0326,
+      "step": 4620
+    },
+    {
+      "epoch": 0.7923355552221532,
+      "grad_norm": 1.5625,
+      "learning_rate": 1.684805090176864e-05,
+      "loss": 0.9854,
+      "step": 4621
+    },
+    {
+      "epoch": 0.792507019311143,
+      "grad_norm": 1.4921875,
+      "learning_rate": 1.68467347627877e-05,
+      "loss": 0.9683,
+      "step": 4622
+    },
+    {
+      "epoch": 0.7926784834001329,
+      "grad_norm": 1.515625,
+      "learning_rate": 1.684541840050896e-05,
+      "loss": 0.9217,
+      "step": 4623
+    },
+    {
+      "epoch": 0.7928499474891227,
+      "grad_norm": 1.53125,
+      "learning_rate": 1.6844101814975345e-05,
+      "loss": 0.9986,
+      "step": 4624
+    },
+    {
+      "epoch": 0.7930214115781126,
+      "grad_norm": 1.671875,
+      "learning_rate": 1.6842785006229805e-05,
+      "loss": 1.0345,
+      "step": 4625
+    },
+    {
+      "epoch": 0.7931928756671025,
+      "grad_norm": 1.609375,
+      "learning_rate": 1.6841467974315275e-05,
+      "loss": 1.059,
+      "step": 4626
+    },
+    {
+      "epoch": 0.7933643397560923,
+      "grad_norm": 1.6015625,
+      "learning_rate": 1.6840150719274712e-05,
+      "loss": 1.0887,
+      "step": 4627
+    },
+    {
+      "epoch": 0.7935358038450822,
+      "grad_norm": 1.609375,
+      "learning_rate": 1.6838833241151072e-05,
+      "loss": 0.9706,
+      "step": 4628
+    },
+    {
+      "epoch": 0.7937072679340721,
+      "grad_norm": 1.6328125,
+      "learning_rate": 1.6837515539987333e-05,
+      "loss": 1.0665,
+      "step": 4629
+    },
+    {
+      "epoch": 0.7938787320230619,
+      "grad_norm": 1.546875,
+      "learning_rate": 1.6836197615826463e-05,
+      "loss": 0.9952,
+      "step": 4630
+    },
+    {
+      "epoch": 0.7940501961120517,
+      "grad_norm": 1.5546875,
+      "learning_rate": 1.6834879468711444e-05,
+      "loss": 1.026,
+      "step": 4631
+    },
+    {
+      "epoch": 0.7942216602010417,
+      "grad_norm": 1.59375,
+      "learning_rate": 1.6833561098685272e-05,
+      "loss": 1.0303,
+      "step": 4632
+    },
+    {
+      "epoch": 0.7943931242900315,
+      "grad_norm": 1.5078125,
+      "learning_rate": 1.6832242505790938e-05,
+      "loss": 0.94,
+      "step": 4633
+    },
+    {
+      "epoch": 0.7945645883790213,
+      "grad_norm": 1.6328125,
+      "learning_rate": 1.6830923690071445e-05,
+      "loss": 1.047,
+      "step": 4634
+    },
+    {
+      "epoch": 0.7947360524680113,
+      "grad_norm": 1.6328125,
+      "learning_rate": 1.6829604651569805e-05,
+      "loss": 1.0085,
+      "step": 4635
+    },
+    {
+      "epoch": 0.7949075165570011,
+      "grad_norm": 1.640625,
+      "learning_rate": 1.6828285390329047e-05,
+      "loss": 0.9228,
+      "step": 4636
+    },
+    {
+      "epoch": 0.7950789806459909,
+      "grad_norm": 1.5546875,
+      "learning_rate": 1.6826965906392187e-05,
+      "loss": 1.0364,
+      "step": 4637
+    },
+    {
+      "epoch": 0.7952504447349809,
+      "grad_norm": 1.5078125,
+      "learning_rate": 1.682564619980226e-05,
+      "loss": 0.965,
+      "step": 4638
+    },
+    {
+      "epoch": 0.7954219088239707,
+      "grad_norm": 1.5546875,
+      "learning_rate": 1.682432627060231e-05,
+      "loss": 1.1089,
+      "step": 4639
+    },
+    {
+      "epoch": 0.7955933729129605,
+      "grad_norm": 1.578125,
+      "learning_rate": 1.682300611883538e-05,
+      "loss": 0.9678,
+      "step": 4640
+    },
+    {
+      "epoch": 0.7957648370019504,
+      "grad_norm": 1.59375,
+      "learning_rate": 1.6821685744544526e-05,
+      "loss": 0.9893,
+      "step": 4641
+    },
+    {
+      "epoch": 0.7959363010909403,
+      "grad_norm": 1.6640625,
+      "learning_rate": 1.6820365147772812e-05,
+      "loss": 1.008,
+      "step": 4642
+    },
+    {
+      "epoch": 0.7961077651799301,
+      "grad_norm": 1.6015625,
+      "learning_rate": 1.681904432856331e-05,
+      "loss": 1.0865,
+      "step": 4643
+    },
+    {
+      "epoch": 0.79627922926892,
+      "grad_norm": 1.5703125,
+      "learning_rate": 1.6817723286959092e-05,
+      "loss": 1.0863,
+      "step": 4644
+    },
+    {
+      "epoch": 0.7964506933579099,
+      "grad_norm": 1.5546875,
+      "learning_rate": 1.6816402023003246e-05,
+      "loss": 1.0579,
+      "step": 4645
+    },
+    {
+      "epoch": 0.7966221574468997,
+      "grad_norm": 1.609375,
+      "learning_rate": 1.6815080536738862e-05,
+      "loss": 1.0737,
+      "step": 4646
+    },
+    {
+      "epoch": 0.7967936215358896,
+      "grad_norm": 1.5078125,
+      "learning_rate": 1.6813758828209036e-05,
+      "loss": 1.0252,
+      "step": 4647
+    },
+    {
+      "epoch": 0.7969650856248794,
+      "grad_norm": 1.59375,
+      "learning_rate": 1.681243689745688e-05,
+      "loss": 1.0197,
+      "step": 4648
+    },
+    {
+      "epoch": 0.7971365497138693,
+      "grad_norm": 1.578125,
+      "learning_rate": 1.6811114744525502e-05,
+      "loss": 1.0342,
+      "step": 4649
+    },
+    {
+      "epoch": 0.7973080138028592,
+      "grad_norm": 1.5546875,
+      "learning_rate": 1.6809792369458022e-05,
+      "loss": 0.9819,
+      "step": 4650
+    },
+    {
+      "epoch": 0.797479477891849,
+      "grad_norm": 1.5859375,
+      "learning_rate": 1.6808469772297572e-05,
+      "loss": 1.0754,
+      "step": 4651
+    },
+    {
+      "epoch": 0.7976509419808389,
+      "grad_norm": 1.515625,
+      "learning_rate": 1.6807146953087282e-05,
+      "loss": 0.925,
+      "step": 4652
+    },
+    {
+      "epoch": 0.7978224060698288,
+      "grad_norm": 1.6015625,
+      "learning_rate": 1.6805823911870298e-05,
+      "loss": 1.0692,
+      "step": 4653
+    },
+    {
+      "epoch": 0.7979938701588186,
+      "grad_norm": 1.609375,
+      "learning_rate": 1.680450064868977e-05,
+      "loss": 1.0828,
+      "step": 4654
+    },
+    {
+      "epoch": 0.7981653342478084,
+      "grad_norm": 1.6015625,
+      "learning_rate": 1.6803177163588848e-05,
+      "loss": 1.0006,
+      "step": 4655
+    },
+    {
+      "epoch": 0.7983367983367984,
+      "grad_norm": 1.5234375,
+      "learning_rate": 1.6801853456610705e-05,
+      "loss": 0.9446,
+      "step": 4656
+    },
+    {
+      "epoch": 0.7985082624257882,
+      "grad_norm": 1.65625,
+      "learning_rate": 1.68005295277985e-05,
+      "loss": 0.9129,
+      "step": 4657
+    },
+    {
+      "epoch": 0.798679726514778,
+      "grad_norm": 1.578125,
+      "learning_rate": 1.6799205377195427e-05,
+      "loss": 0.971,
+      "step": 4658
+    },
+    {
+      "epoch": 0.798851190603768,
+      "grad_norm": 1.5546875,
+      "learning_rate": 1.6797881004844658e-05,
+      "loss": 0.9304,
+      "step": 4659
+    },
+    {
+      "epoch": 0.7990226546927578,
+      "grad_norm": 1.5859375,
+      "learning_rate": 1.6796556410789394e-05,
+      "loss": 1.0222,
+      "step": 4660
+    },
+    {
+      "epoch": 0.7991941187817476,
+      "grad_norm": 1.6640625,
+      "learning_rate": 1.6795231595072832e-05,
+      "loss": 0.969,
+      "step": 4661
+    },
+    {
+      "epoch": 0.7993655828707376,
+      "grad_norm": 1.59375,
+      "learning_rate": 1.6793906557738177e-05,
+      "loss": 0.9851,
+      "step": 4662
+    },
+    {
+      "epoch": 0.7995370469597274,
+      "grad_norm": 1.765625,
+      "learning_rate": 1.6792581298828644e-05,
+      "loss": 0.9575,
+      "step": 4663
+    },
+    {
+      "epoch": 0.7997085110487172,
+      "grad_norm": 1.625,
+      "learning_rate": 1.6791255818387462e-05,
+      "loss": 0.9878,
+      "step": 4664
+    },
+    {
+      "epoch": 0.7998799751377071,
+      "grad_norm": 1.5078125,
+      "learning_rate": 1.678993011645785e-05,
+      "loss": 0.9399,
+      "step": 4665
+    },
+    {
+      "epoch": 0.800051439226697,
+      "grad_norm": 1.6640625,
+      "learning_rate": 1.6788604193083052e-05,
+      "loss": 0.9622,
+      "step": 4666
+    },
+    {
+      "epoch": 0.8002229033156868,
+      "grad_norm": 1.5546875,
+      "learning_rate": 1.6787278048306307e-05,
+      "loss": 0.98,
+      "step": 4667
+    },
+    {
+      "epoch": 0.8003943674046767,
+      "grad_norm": 1.640625,
+      "learning_rate": 1.6785951682170863e-05,
+      "loss": 1.1163,
+      "step": 4668
+    },
+    {
+      "epoch": 0.8005658314936666,
+      "grad_norm": 1.546875,
+      "learning_rate": 1.678462509471998e-05,
+      "loss": 1.0644,
+      "step": 4669
+    },
+    {
+      "epoch": 0.8007372955826564,
+      "grad_norm": 1.53125,
+      "learning_rate": 1.6783298285996928e-05,
+      "loss": 0.9675,
+      "step": 4670
+    },
+    {
+      "epoch": 0.8009087596716463,
+      "grad_norm": 1.6640625,
+      "learning_rate": 1.6781971256044975e-05,
+      "loss": 0.9141,
+      "step": 4671
+    },
+    {
+      "epoch": 0.8010802237606361,
+      "grad_norm": 1.53125,
+      "learning_rate": 1.67806440049074e-05,
+      "loss": 0.971,
+      "step": 4672
+    },
+    {
+      "epoch": 0.801251687849626,
+      "grad_norm": 1.5625,
+      "learning_rate": 1.677931653262749e-05,
+      "loss": 0.9711,
+      "step": 4673
+    },
+    {
+      "epoch": 0.8014231519386159,
+      "grad_norm": 1.5703125,
+      "learning_rate": 1.677798883924854e-05,
+      "loss": 0.9369,
+      "step": 4674
+    },
+    {
+      "epoch": 0.8015946160276057,
+      "grad_norm": 1.5625,
+      "learning_rate": 1.6776660924813854e-05,
+      "loss": 0.9912,
+      "step": 4675
+    },
+    {
+      "epoch": 0.8017660801165956,
+      "grad_norm": 1.6328125,
+      "learning_rate": 1.677533278936673e-05,
+      "loss": 1.1182,
+      "step": 4676
+    },
+    {
+      "epoch": 0.8019375442055855,
+      "grad_norm": 1.6328125,
+      "learning_rate": 1.6774004432950496e-05,
+      "loss": 1.0211,
+      "step": 4677
+    },
+    {
+      "epoch": 0.8021090082945753,
+      "grad_norm": 1.671875,
+      "learning_rate": 1.6772675855608466e-05,
+      "loss": 0.9907,
+      "step": 4678
+    },
+    {
+      "epoch": 0.8022804723835651,
+      "grad_norm": 1.5625,
+      "learning_rate": 1.677134705738397e-05,
+      "loss": 0.9965,
+      "step": 4679
+    },
+    {
+      "epoch": 0.8024519364725551,
+      "grad_norm": 1.6171875,
+      "learning_rate": 1.6770018038320353e-05,
+      "loss": 1.0059,
+      "step": 4680
+    },
+    {
+      "epoch": 0.8026234005615449,
+      "grad_norm": 1.765625,
+      "learning_rate": 1.676868879846095e-05,
+      "loss": 1.092,
+      "step": 4681
+    },
+    {
+      "epoch": 0.8027948646505347,
+      "grad_norm": 1.625,
+      "learning_rate": 1.676735933784912e-05,
+      "loss": 1.0292,
+      "step": 4682
+    },
+    {
+      "epoch": 0.8029663287395247,
+      "grad_norm": 1.703125,
+      "learning_rate": 1.6766029656528218e-05,
+      "loss": 1.0185,
+      "step": 4683
+    },
+    {
+      "epoch": 0.8031377928285145,
+      "grad_norm": 2.484375,
+      "learning_rate": 1.6764699754541608e-05,
+      "loss": 1.0001,
+      "step": 4684
+    },
+    {
+      "epoch": 0.8033092569175043,
+      "grad_norm": 1.8125,
+      "learning_rate": 1.6763369631932665e-05,
+      "loss": 1.0169,
+      "step": 4685
+    },
+    {
+      "epoch": 0.8034807210064941,
+      "grad_norm": 1.5859375,
+      "learning_rate": 1.6762039288744777e-05,
+      "loss": 0.9109,
+      "step": 4686
+    },
+    {
+      "epoch": 0.8036521850954841,
+      "grad_norm": 3.171875,
+      "learning_rate": 1.6760708725021316e-05,
+      "loss": 0.9967,
+      "step": 4687
+    },
+    {
+      "epoch": 0.8038236491844739,
+      "grad_norm": 1.5,
+      "learning_rate": 1.675937794080569e-05,
+      "loss": 0.963,
+      "step": 4688
+    },
+    {
+      "epoch": 0.8039951132734637,
+      "grad_norm": 1.6015625,
+      "learning_rate": 1.6758046936141295e-05,
+      "loss": 1.0488,
+      "step": 4689
+    },
+    {
+      "epoch": 0.8041665773624537,
+      "grad_norm": 1.5,
+      "learning_rate": 1.675671571107154e-05,
+      "loss": 0.9804,
+      "step": 4690
+    },
+    {
+      "epoch": 0.8043380414514435,
+      "grad_norm": 1.671875,
+      "learning_rate": 1.675538426563984e-05,
+      "loss": 1.0073,
+      "step": 4691
+    },
+    {
+      "epoch": 0.8045095055404333,
+      "grad_norm": 2.015625,
+      "learning_rate": 1.6754052599889626e-05,
+      "loss": 1.0464,
+      "step": 4692
+    },
+    {
+      "epoch": 0.8046809696294233,
+      "grad_norm": 1.59375,
+      "learning_rate": 1.675272071386432e-05,
+      "loss": 1.0297,
+      "step": 4693
+    },
+    {
+      "epoch": 0.8048524337184131,
+      "grad_norm": 1.5859375,
+      "learning_rate": 1.6751388607607363e-05,
+      "loss": 1.1082,
+      "step": 4694
+    },
+    {
+      "epoch": 0.8050238978074029,
+      "grad_norm": 1.5859375,
+      "learning_rate": 1.6750056281162203e-05,
+      "loss": 0.9478,
+      "step": 4695
+    },
+    {
+      "epoch": 0.8051953618963928,
+      "grad_norm": 1.671875,
+      "learning_rate": 1.6748723734572285e-05,
+      "loss": 1.0545,
+      "step": 4696
+    },
+    {
+      "epoch": 0.8053668259853827,
+      "grad_norm": 1.5703125,
+      "learning_rate": 1.6747390967881073e-05,
+      "loss": 1.0008,
+      "step": 4697
+    },
+    {
+      "epoch": 0.8055382900743725,
+      "grad_norm": 1.515625,
+      "learning_rate": 1.6746057981132037e-05,
+      "loss": 0.9503,
+      "step": 4698
+    },
+    {
+      "epoch": 0.8057097541633624,
+      "grad_norm": 1.671875,
+      "learning_rate": 1.6744724774368645e-05,
+      "loss": 1.0142,
+      "step": 4699
+    },
+    {
+      "epoch": 0.8058812182523523,
+      "grad_norm": 1.546875,
+      "learning_rate": 1.674339134763438e-05,
+      "loss": 0.9319,
+      "step": 4700
+    },
+    {
+      "epoch": 0.8060526823413421,
+      "grad_norm": 1.5703125,
+      "learning_rate": 1.674205770097273e-05,
+      "loss": 1.079,
+      "step": 4701
+    },
+    {
+      "epoch": 0.806224146430332,
+      "grad_norm": 2.1875,
+      "learning_rate": 1.674072383442719e-05,
+      "loss": 1.0402,
+      "step": 4702
+    },
+    {
+      "epoch": 0.8063956105193218,
+      "grad_norm": 1.7578125,
+      "learning_rate": 1.673938974804126e-05,
+      "loss": 1.0822,
+      "step": 4703
+    },
+    {
+      "epoch": 0.8065670746083117,
+      "grad_norm": 1.6328125,
+      "learning_rate": 1.6738055441858456e-05,
+      "loss": 1.075,
+      "step": 4704
+    },
+    {
+      "epoch": 0.8067385386973016,
+      "grad_norm": 1.65625,
+      "learning_rate": 1.673672091592229e-05,
+      "loss": 1.077,
+      "step": 4705
+    },
+    {
+      "epoch": 0.8069100027862914,
+      "grad_norm": 1.5078125,
+      "learning_rate": 1.6735386170276283e-05,
+      "loss": 1.0157,
+      "step": 4706
+    },
+    {
+      "epoch": 0.8070814668752813,
+      "grad_norm": 1.6640625,
+      "learning_rate": 1.6734051204963974e-05,
+      "loss": 1.1408,
+      "step": 4707
+    },
+    {
+      "epoch": 0.8072529309642712,
+      "grad_norm": 1.609375,
+      "learning_rate": 1.6732716020028897e-05,
+      "loss": 0.9205,
+      "step": 4708
+    },
+    {
+      "epoch": 0.807424395053261,
+      "grad_norm": 1.5078125,
+      "learning_rate": 1.6731380615514593e-05,
+      "loss": 0.9743,
+      "step": 4709
+    },
+    {
+      "epoch": 0.8075958591422508,
+      "grad_norm": 1.5703125,
+      "learning_rate": 1.6730044991464626e-05,
+      "loss": 1.0179,
+      "step": 4710
+    },
+    {
+      "epoch": 0.8077673232312408,
+      "grad_norm": 1.59375,
+      "learning_rate": 1.6728709147922543e-05,
+      "loss": 0.9116,
+      "step": 4711
+    },
+    {
+      "epoch": 0.8079387873202306,
+      "grad_norm": 1.6171875,
+      "learning_rate": 1.672737308493192e-05,
+      "loss": 0.9609,
+      "step": 4712
+    },
+    {
+      "epoch": 0.8081102514092204,
+      "grad_norm": 1.5625,
+      "learning_rate": 1.672603680253633e-05,
+      "loss": 0.9789,
+      "step": 4713
+    },
+    {
+      "epoch": 0.8082817154982104,
+      "grad_norm": 1.6640625,
+      "learning_rate": 1.6724700300779346e-05,
+      "loss": 1.0335,
+      "step": 4714
+    },
+    {
+      "epoch": 0.8084531795872002,
+      "grad_norm": 1.625,
+      "learning_rate": 1.6723363579704567e-05,
+      "loss": 1.0538,
+      "step": 4715
+    },
+    {
+      "epoch": 0.80862464367619,
+      "grad_norm": 1.640625,
+      "learning_rate": 1.672202663935558e-05,
+      "loss": 1.092,
+      "step": 4716
+    },
+    {
+      "epoch": 0.80879610776518,
+      "grad_norm": 1.53125,
+      "learning_rate": 1.6720689479775996e-05,
+      "loss": 0.9346,
+      "step": 4717
+    },
+    {
+      "epoch": 0.8089675718541698,
+      "grad_norm": 1.5546875,
+      "learning_rate": 1.6719352101009417e-05,
+      "loss": 0.9112,
+      "step": 4718
+    },
+    {
+      "epoch": 0.8091390359431596,
+      "grad_norm": 1.484375,
+      "learning_rate": 1.671801450309946e-05,
+      "loss": 0.9348,
+      "step": 4719
+    },
+    {
+      "epoch": 0.8093105000321495,
+      "grad_norm": 1.5,
+      "learning_rate": 1.671667668608976e-05,
+      "loss": 0.9716,
+      "step": 4720
+    },
+    {
+      "epoch": 0.8094819641211394,
+      "grad_norm": 1.53125,
+      "learning_rate": 1.6715338650023936e-05,
+      "loss": 0.9984,
+      "step": 4721
+    },
+    {
+      "epoch": 0.8096534282101292,
+      "grad_norm": 1.5703125,
+      "learning_rate": 1.6714000394945632e-05,
+      "loss": 0.9771,
+      "step": 4722
+    },
+    {
+      "epoch": 0.8098248922991191,
+      "grad_norm": 1.5859375,
+      "learning_rate": 1.6712661920898493e-05,
+      "loss": 0.9918,
+      "step": 4723
+    },
+    {
+      "epoch": 0.809996356388109,
+      "grad_norm": 1.4921875,
+      "learning_rate": 1.671132322792617e-05,
+      "loss": 0.9857,
+      "step": 4724
+    },
+    {
+      "epoch": 0.8101678204770988,
+      "grad_norm": 1.4921875,
+      "learning_rate": 1.6709984316072327e-05,
+      "loss": 0.9775,
+      "step": 4725
+    },
+    {
+      "epoch": 0.8103392845660887,
+      "grad_norm": 1.6015625,
+      "learning_rate": 1.6708645185380622e-05,
+      "loss": 1.0013,
+      "step": 4726
+    },
+    {
+      "epoch": 0.8105107486550786,
+      "grad_norm": 1.5,
+      "learning_rate": 1.670730583589474e-05,
+      "loss": 0.9176,
+      "step": 4727
+    },
+    {
+      "epoch": 0.8106822127440684,
+      "grad_norm": 1.6328125,
+      "learning_rate": 1.6705966267658355e-05,
+      "loss": 0.948,
+      "step": 4728
+    },
+    {
+      "epoch": 0.8108536768330583,
+      "grad_norm": 1.5078125,
+      "learning_rate": 1.6704626480715162e-05,
+      "loss": 0.9791,
+      "step": 4729
+    },
+    {
+      "epoch": 0.8110251409220481,
+      "grad_norm": 1.875,
+      "learning_rate": 1.670328647510885e-05,
+      "loss": 1.0124,
+      "step": 4730
+    },
+    {
+      "epoch": 0.811196605011038,
+      "grad_norm": 1.6640625,
+      "learning_rate": 1.670194625088312e-05,
+      "loss": 1.1446,
+      "step": 4731
+    },
+    {
+      "epoch": 0.8113680691000279,
+      "grad_norm": 1.546875,
+      "learning_rate": 1.6700605808081688e-05,
+      "loss": 1.0095,
+      "step": 4732
+    },
+    {
+      "epoch": 0.8115395331890177,
+      "grad_norm": 1.546875,
+      "learning_rate": 1.669926514674827e-05,
+      "loss": 0.9893,
+      "step": 4733
+    },
+    {
+      "epoch": 0.8117109972780076,
+      "grad_norm": 1.484375,
+      "learning_rate": 1.6697924266926588e-05,
+      "loss": 1.0897,
+      "step": 4734
+    },
+    {
+      "epoch": 0.8118824613669975,
+      "grad_norm": 1.6171875,
+      "learning_rate": 1.6696583168660373e-05,
+      "loss": 0.9558,
+      "step": 4735
+    },
+    {
+      "epoch": 0.8120539254559873,
+      "grad_norm": 1.515625,
+      "learning_rate": 1.6695241851993363e-05,
+      "loss": 0.952,
+      "step": 4736
+    },
+    {
+      "epoch": 0.8122253895449771,
+      "grad_norm": 1.9765625,
+      "learning_rate": 1.6693900316969303e-05,
+      "loss": 1.0141,
+      "step": 4737
+    },
+    {
+      "epoch": 0.8123968536339671,
+      "grad_norm": 1.5,
+      "learning_rate": 1.669255856363195e-05,
+      "loss": 0.9915,
+      "step": 4738
+    },
+    {
+      "epoch": 0.8125683177229569,
+      "grad_norm": 1.8828125,
+      "learning_rate": 1.669121659202506e-05,
+      "loss": 1.0528,
+      "step": 4739
+    },
+    {
+      "epoch": 0.8127397818119467,
+      "grad_norm": 1.546875,
+      "learning_rate": 1.66898744021924e-05,
+      "loss": 1.0535,
+      "step": 4740
+    },
+    {
+      "epoch": 0.8129112459009367,
+      "grad_norm": 1.5859375,
+      "learning_rate": 1.6688531994177745e-05,
+      "loss": 0.9725,
+      "step": 4741
+    },
+    {
+      "epoch": 0.8130827099899265,
+      "grad_norm": 1.578125,
+      "learning_rate": 1.6687189368024874e-05,
+      "loss": 1.026,
+      "step": 4742
+    },
+    {
+      "epoch": 0.8132541740789163,
+      "grad_norm": 1.609375,
+      "learning_rate": 1.6685846523777573e-05,
+      "loss": 1.0988,
+      "step": 4743
+    },
+    {
+      "epoch": 0.8134256381679063,
+      "grad_norm": 1.4765625,
+      "learning_rate": 1.668450346147964e-05,
+      "loss": 0.8971,
+      "step": 4744
+    },
+    {
+      "epoch": 0.8135971022568961,
+      "grad_norm": 1.6328125,
+      "learning_rate": 1.6683160181174886e-05,
+      "loss": 1.0287,
+      "step": 4745
+    },
+    {
+      "epoch": 0.8137685663458859,
+      "grad_norm": 1.53125,
+      "learning_rate": 1.6681816682907104e-05,
+      "loss": 1.0021,
+      "step": 4746
+    },
+    {
+      "epoch": 0.8139400304348758,
+      "grad_norm": 1.515625,
+      "learning_rate": 1.668047296672012e-05,
+      "loss": 0.9952,
+      "step": 4747
+    },
+    {
+      "epoch": 0.8141114945238657,
+      "grad_norm": 1.59375,
+      "learning_rate": 1.667912903265776e-05,
+      "loss": 0.9552,
+      "step": 4748
+    },
+    {
+      "epoch": 0.8142829586128555,
+      "grad_norm": 1.6171875,
+      "learning_rate": 1.6677784880763847e-05,
+      "loss": 1.0447,
+      "step": 4749
+    },
+    {
+      "epoch": 0.8144544227018454,
+      "grad_norm": 1.5546875,
+      "learning_rate": 1.6676440511082223e-05,
+      "loss": 1.0072,
+      "step": 4750
+    },
+    {
+      "epoch": 0.8146258867908353,
+      "grad_norm": 1.6328125,
+      "learning_rate": 1.6675095923656736e-05,
+      "loss": 1.0373,
+      "step": 4751
+    },
+    {
+      "epoch": 0.8147973508798251,
+      "grad_norm": 1.5390625,
+      "learning_rate": 1.667375111853123e-05,
+      "loss": 1.0665,
+      "step": 4752
+    },
+    {
+      "epoch": 0.814968814968815,
+      "grad_norm": 1.6171875,
+      "learning_rate": 1.6672406095749573e-05,
+      "loss": 0.9888,
+      "step": 4753
+    },
+    {
+      "epoch": 0.8151402790578048,
+      "grad_norm": 1.5546875,
+      "learning_rate": 1.6671060855355626e-05,
+      "loss": 1.0541,
+      "step": 4754
+    },
+    {
+      "epoch": 0.8153117431467947,
+      "grad_norm": 1.5703125,
+      "learning_rate": 1.666971539739326e-05,
+      "loss": 1.0444,
+      "step": 4755
+    },
+    {
+      "epoch": 0.8154832072357846,
+      "grad_norm": 1.6171875,
+      "learning_rate": 1.6668369721906367e-05,
+      "loss": 1.0401,
+      "step": 4756
+    },
+    {
+      "epoch": 0.8156546713247744,
+      "grad_norm": 1.578125,
+      "learning_rate": 1.666702382893882e-05,
+      "loss": 1.0007,
+      "step": 4757
+    },
+    {
+      "epoch": 0.8158261354137643,
+      "grad_norm": 1.5703125,
+      "learning_rate": 1.6665677718534525e-05,
+      "loss": 1.074,
+      "step": 4758
+    },
+    {
+      "epoch": 0.8159975995027542,
+      "grad_norm": 1.4609375,
+      "learning_rate": 1.6664331390737373e-05,
+      "loss": 0.9096,
+      "step": 4759
+    },
+    {
+      "epoch": 0.816169063591744,
+      "grad_norm": 1.5625,
+      "learning_rate": 1.6662984845591283e-05,
+      "loss": 1.0082,
+      "step": 4760
+    },
+    {
+      "epoch": 0.8163405276807338,
+      "grad_norm": 1.515625,
+      "learning_rate": 1.6661638083140167e-05,
+      "loss": 0.9651,
+      "step": 4761
+    },
+    {
+      "epoch": 0.8165119917697238,
+      "grad_norm": 1.5234375,
+      "learning_rate": 1.6660291103427946e-05,
+      "loss": 0.9039,
+      "step": 4762
+    },
+    {
+      "epoch": 0.8166834558587136,
+      "grad_norm": 1.6015625,
+      "learning_rate": 1.665894390649855e-05,
+      "loss": 0.9503,
+      "step": 4763
+    },
+    {
+      "epoch": 0.8168549199477034,
+      "grad_norm": 1.5234375,
+      "learning_rate": 1.6657596492395926e-05,
+      "loss": 0.9711,
+      "step": 4764
+    },
+    {
+      "epoch": 0.8170263840366934,
+      "grad_norm": 1.5859375,
+      "learning_rate": 1.6656248861164e-05,
+      "loss": 1.0314,
+      "step": 4765
+    },
+    {
+      "epoch": 0.8171978481256832,
+      "grad_norm": 1.53125,
+      "learning_rate": 1.6654901012846737e-05,
+      "loss": 0.948,
+      "step": 4766
+    },
+    {
+      "epoch": 0.817369312214673,
+      "grad_norm": 1.515625,
+      "learning_rate": 1.6653552947488095e-05,
+      "loss": 0.9148,
+      "step": 4767
+    },
+    {
+      "epoch": 0.817540776303663,
+      "grad_norm": 1.671875,
+      "learning_rate": 1.6652204665132036e-05,
+      "loss": 1.0365,
+      "step": 4768
+    },
+    {
+      "epoch": 0.8177122403926528,
+      "grad_norm": 1.4921875,
+      "learning_rate": 1.665085616582253e-05,
+      "loss": 0.9708,
+      "step": 4769
+    },
+    {
+      "epoch": 0.8178837044816426,
+      "grad_norm": 1.484375,
+      "learning_rate": 1.664950744960356e-05,
+      "loss": 1.0439,
+      "step": 4770
+    },
+    {
+      "epoch": 0.8180551685706325,
+      "grad_norm": 1.546875,
+      "learning_rate": 1.6648158516519115e-05,
+      "loss": 0.9592,
+      "step": 4771
+    },
+    {
+      "epoch": 0.8182266326596224,
+      "grad_norm": 1.6328125,
+      "learning_rate": 1.6646809366613186e-05,
+      "loss": 1.0561,
+      "step": 4772
+    },
+    {
+      "epoch": 0.8183980967486122,
+      "grad_norm": 1.59375,
+      "learning_rate": 1.664545999992977e-05,
+      "loss": 1.0024,
+      "step": 4773
+    },
+    {
+      "epoch": 0.8185695608376021,
+      "grad_norm": 1.5546875,
+      "learning_rate": 1.6644110416512883e-05,
+      "loss": 1.0269,
+      "step": 4774
+    },
+    {
+      "epoch": 0.818741024926592,
+      "grad_norm": 1.640625,
+      "learning_rate": 1.6642760616406536e-05,
+      "loss": 1.0261,
+      "step": 4775
+    },
+    {
+      "epoch": 0.8189124890155818,
+      "grad_norm": 1.5078125,
+      "learning_rate": 1.664141059965475e-05,
+      "loss": 1.0186,
+      "step": 4776
+    },
+    {
+      "epoch": 0.8190839531045717,
+      "grad_norm": 1.59375,
+      "learning_rate": 1.664006036630155e-05,
+      "loss": 0.9797,
+      "step": 4777
+    },
+    {
+      "epoch": 0.8192554171935615,
+      "grad_norm": 1.5078125,
+      "learning_rate": 1.663870991639098e-05,
+      "loss": 0.9059,
+      "step": 4778
+    },
+    {
+      "epoch": 0.8194268812825514,
+      "grad_norm": 1.5625,
+      "learning_rate": 1.6637359249967085e-05,
+      "loss": 0.9974,
+      "step": 4779
+    },
+    {
+      "epoch": 0.8195983453715412,
+      "grad_norm": 1.546875,
+      "learning_rate": 1.6636008367073908e-05,
+      "loss": 0.9284,
+      "step": 4780
+    },
+    {
+      "epoch": 0.8197698094605311,
+      "grad_norm": 1.546875,
+      "learning_rate": 1.6634657267755508e-05,
+      "loss": 0.9575,
+      "step": 4781
+    },
+    {
+      "epoch": 0.819941273549521,
+      "grad_norm": 1.53125,
+      "learning_rate": 1.6633305952055955e-05,
+      "loss": 1.0132,
+      "step": 4782
+    },
+    {
+      "epoch": 0.8201127376385108,
+      "grad_norm": 1.625,
+      "learning_rate": 1.6631954420019312e-05,
+      "loss": 1.03,
+      "step": 4783
+    },
+    {
+      "epoch": 0.8202842017275007,
+      "grad_norm": 1.578125,
+      "learning_rate": 1.6630602671689665e-05,
+      "loss": 0.9498,
+      "step": 4784
+    },
+    {
+      "epoch": 0.8204556658164905,
+      "grad_norm": 1.5703125,
+      "learning_rate": 1.662925070711109e-05,
+      "loss": 0.9973,
+      "step": 4785
+    },
+    {
+      "epoch": 0.8206271299054804,
+      "grad_norm": 1.5234375,
+      "learning_rate": 1.662789852632769e-05,
+      "loss": 1.1094,
+      "step": 4786
+    },
+    {
+      "epoch": 0.8207985939944703,
+      "grad_norm": 1.546875,
+      "learning_rate": 1.6626546129383564e-05,
+      "loss": 0.9227,
+      "step": 4787
+    },
+    {
+      "epoch": 0.8209700580834601,
+      "grad_norm": 1.625,
+      "learning_rate": 1.6625193516322813e-05,
+      "loss": 1.1013,
+      "step": 4788
+    },
+    {
+      "epoch": 0.82114152217245,
+      "grad_norm": 1.53125,
+      "learning_rate": 1.6623840687189554e-05,
+      "loss": 0.9486,
+      "step": 4789
+    },
+    {
+      "epoch": 0.8213129862614399,
+      "grad_norm": 1.5390625,
+      "learning_rate": 1.6622487642027904e-05,
+      "loss": 0.9604,
+      "step": 4790
+    },
+    {
+      "epoch": 0.8214844503504297,
+      "grad_norm": 1.4765625,
+      "learning_rate": 1.6621134380881997e-05,
+      "loss": 0.955,
+      "step": 4791
+    },
+    {
+      "epoch": 0.8216559144394195,
+      "grad_norm": 1.625,
+      "learning_rate": 1.6619780903795964e-05,
+      "loss": 1.0258,
+      "step": 4792
+    },
+    {
+      "epoch": 0.8218273785284095,
+      "grad_norm": 1.6015625,
+      "learning_rate": 1.6618427210813948e-05,
+      "loss": 1.0394,
+      "step": 4793
+    },
+    {
+      "epoch": 0.8219988426173993,
+      "grad_norm": 1.5390625,
+      "learning_rate": 1.6617073301980097e-05,
+      "loss": 0.9275,
+      "step": 4794
+    },
+    {
+      "epoch": 0.8221703067063891,
+      "grad_norm": 1.5234375,
+      "learning_rate": 1.6615719177338568e-05,
+      "loss": 1.0212,
+      "step": 4795
+    },
+    {
+      "epoch": 0.8223417707953791,
+      "grad_norm": 1.4375,
+      "learning_rate": 1.6614364836933526e-05,
+      "loss": 0.9618,
+      "step": 4796
+    },
+    {
+      "epoch": 0.8225132348843689,
+      "grad_norm": 1.4765625,
+      "learning_rate": 1.6613010280809138e-05,
+      "loss": 1.0033,
+      "step": 4797
+    },
+    {
+      "epoch": 0.8226846989733587,
+      "grad_norm": 1.5234375,
+      "learning_rate": 1.661165550900958e-05,
+      "loss": 1.0089,
+      "step": 4798
+    },
+    {
+      "epoch": 0.8228561630623487,
+      "grad_norm": 1.515625,
+      "learning_rate": 1.6610300521579046e-05,
+      "loss": 1.0124,
+      "step": 4799
+    },
+    {
+      "epoch": 0.8230276271513385,
+      "grad_norm": 1.6171875,
+      "learning_rate": 1.6608945318561715e-05,
+      "loss": 0.9778,
+      "step": 4800
+    },
+    {
+      "epoch": 0.8231990912403283,
+      "grad_norm": 1.6015625,
+      "learning_rate": 1.6607589900001787e-05,
+      "loss": 0.9307,
+      "step": 4801
+    },
+    {
+      "epoch": 0.8233705553293182,
+      "grad_norm": 1.6484375,
+      "learning_rate": 1.6606234265943477e-05,
+      "loss": 0.9979,
+      "step": 4802
+    },
+    {
+      "epoch": 0.8235420194183081,
+      "grad_norm": 1.609375,
+      "learning_rate": 1.6604878416430986e-05,
+      "loss": 1.032,
+      "step": 4803
+    },
+    {
+      "epoch": 0.8237134835072979,
+      "grad_norm": 1.640625,
+      "learning_rate": 1.660352235150854e-05,
+      "loss": 0.9481,
+      "step": 4804
+    },
+    {
+      "epoch": 0.8238849475962878,
+      "grad_norm": 1.9921875,
+      "learning_rate": 1.6602166071220365e-05,
+      "loss": 0.9735,
+      "step": 4805
+    },
+    {
+      "epoch": 0.8240564116852777,
+      "grad_norm": 1.5625,
+      "learning_rate": 1.660080957561069e-05,
+      "loss": 1.0189,
+      "step": 4806
+    },
+    {
+      "epoch": 0.8242278757742675,
+      "grad_norm": 1.75,
+      "learning_rate": 1.659945286472376e-05,
+      "loss": 0.9874,
+      "step": 4807
+    },
+    {
+      "epoch": 0.8243993398632574,
+      "grad_norm": 1.7109375,
+      "learning_rate": 1.6598095938603818e-05,
+      "loss": 1.0096,
+      "step": 4808
+    },
+    {
+      "epoch": 0.8245708039522472,
+      "grad_norm": 1.6015625,
+      "learning_rate": 1.6596738797295126e-05,
+      "loss": 0.9903,
+      "step": 4809
+    },
+    {
+      "epoch": 0.8247422680412371,
+      "grad_norm": 1.5390625,
+      "learning_rate": 1.659538144084194e-05,
+      "loss": 1.0101,
+      "step": 4810
+    },
+    {
+      "epoch": 0.824913732130227,
+      "grad_norm": 1.65625,
+      "learning_rate": 1.6594023869288528e-05,
+      "loss": 0.928,
+      "step": 4811
+    },
+    {
+      "epoch": 0.8250851962192168,
+      "grad_norm": 1.6015625,
+      "learning_rate": 1.6592666082679165e-05,
+      "loss": 0.9943,
+      "step": 4812
+    },
+    {
+      "epoch": 0.8252566603082067,
+      "grad_norm": 1.671875,
+      "learning_rate": 1.6591308081058136e-05,
+      "loss": 1.0472,
+      "step": 4813
+    },
+    {
+      "epoch": 0.8254281243971966,
+      "grad_norm": 1.609375,
+      "learning_rate": 1.6589949864469733e-05,
+      "loss": 0.9844,
+      "step": 4814
+    },
+    {
+      "epoch": 0.8255995884861864,
+      "grad_norm": 1.6015625,
+      "learning_rate": 1.6588591432958246e-05,
+      "loss": 1.0661,
+      "step": 4815
+    },
+    {
+      "epoch": 0.8257710525751762,
+      "grad_norm": 1.5546875,
+      "learning_rate": 1.6587232786567985e-05,
+      "loss": 1.0042,
+      "step": 4816
+    },
+    {
+      "epoch": 0.8259425166641662,
+      "grad_norm": 1.5703125,
+      "learning_rate": 1.6585873925343257e-05,
+      "loss": 1.0843,
+      "step": 4817
+    },
+    {
+      "epoch": 0.826113980753156,
+      "grad_norm": 1.5625,
+      "learning_rate": 1.6584514849328378e-05,
+      "loss": 1.0532,
+      "step": 4818
+    },
+    {
+      "epoch": 0.8262854448421458,
+      "grad_norm": 2.015625,
+      "learning_rate": 1.658315555856768e-05,
+      "loss": 0.8471,
+      "step": 4819
+    },
+    {
+      "epoch": 0.8264569089311358,
+      "grad_norm": 1.5859375,
+      "learning_rate": 1.6581796053105483e-05,
+      "loss": 0.9945,
+      "step": 4820
+    },
+    {
+      "epoch": 0.8266283730201256,
+      "grad_norm": 1.578125,
+      "learning_rate": 1.6580436332986134e-05,
+      "loss": 0.9524,
+      "step": 4821
+    },
+    {
+      "epoch": 0.8267998371091154,
+      "grad_norm": 1.640625,
+      "learning_rate": 1.657907639825398e-05,
+      "loss": 1.1008,
+      "step": 4822
+    },
+    {
+      "epoch": 0.8269713011981054,
+      "grad_norm": 1.640625,
+      "learning_rate": 1.6577716248953364e-05,
+      "loss": 1.1223,
+      "step": 4823
+    },
+    {
+      "epoch": 0.8271427652870952,
+      "grad_norm": 1.6328125,
+      "learning_rate": 1.6576355885128657e-05,
+      "loss": 1.0223,
+      "step": 4824
+    },
+    {
+      "epoch": 0.827314229376085,
+      "grad_norm": 1.484375,
+      "learning_rate": 1.6574995306824217e-05,
+      "loss": 0.9231,
+      "step": 4825
+    },
+    {
+      "epoch": 0.827485693465075,
+      "grad_norm": 1.5390625,
+      "learning_rate": 1.657363451408442e-05,
+      "loss": 1.0,
+      "step": 4826
+    },
+    {
+      "epoch": 0.8276571575540648,
+      "grad_norm": 1.59375,
+      "learning_rate": 1.657227350695365e-05,
+      "loss": 1.0248,
+      "step": 4827
+    },
+    {
+      "epoch": 0.8278286216430546,
+      "grad_norm": 1.640625,
+      "learning_rate": 1.6570912285476294e-05,
+      "loss": 0.9989,
+      "step": 4828
+    },
+    {
+      "epoch": 0.8280000857320445,
+      "grad_norm": 1.515625,
+      "learning_rate": 1.6569550849696742e-05,
+      "loss": 1.0243,
+      "step": 4829
+    },
+    {
+      "epoch": 0.8281715498210344,
+      "grad_norm": 1.453125,
+      "learning_rate": 1.65681891996594e-05,
+      "loss": 0.9558,
+      "step": 4830
+    },
+    {
+      "epoch": 0.8283430139100242,
+      "grad_norm": 1.53125,
+      "learning_rate": 1.656682733540867e-05,
+      "loss": 1.0354,
+      "step": 4831
+    },
+    {
+      "epoch": 0.8285144779990141,
+      "grad_norm": 1.5546875,
+      "learning_rate": 1.6565465256988976e-05,
+      "loss": 1.0249,
+      "step": 4832
+    },
+    {
+      "epoch": 0.828685942088004,
+      "grad_norm": 1.484375,
+      "learning_rate": 1.6564102964444736e-05,
+      "loss": 0.8995,
+      "step": 4833
+    },
+    {
+      "epoch": 0.8288574061769938,
+      "grad_norm": 1.5625,
+      "learning_rate": 1.6562740457820382e-05,
+      "loss": 0.9886,
+      "step": 4834
+    },
+    {
+      "epoch": 0.8290288702659837,
+      "grad_norm": 1.59375,
+      "learning_rate": 1.6561377737160347e-05,
+      "loss": 1.084,
+      "step": 4835
+    },
+    {
+      "epoch": 0.8292003343549735,
+      "grad_norm": 1.6484375,
+      "learning_rate": 1.6560014802509076e-05,
+      "loss": 0.9952,
+      "step": 4836
+    },
+    {
+      "epoch": 0.8293717984439634,
+      "grad_norm": 1.5625,
+      "learning_rate": 1.6558651653911018e-05,
+      "loss": 0.981,
+      "step": 4837
+    },
+    {
+      "epoch": 0.8295432625329533,
+      "grad_norm": 1.6484375,
+      "learning_rate": 1.6557288291410632e-05,
+      "loss": 1.0717,
+      "step": 4838
+    },
+    {
+      "epoch": 0.8297147266219431,
+      "grad_norm": 1.546875,
+      "learning_rate": 1.6555924715052387e-05,
+      "loss": 1.0198,
+      "step": 4839
+    },
+    {
+      "epoch": 0.829886190710933,
+      "grad_norm": 1.6015625,
+      "learning_rate": 1.6554560924880747e-05,
+      "loss": 1.0254,
+      "step": 4840
+    },
+    {
+      "epoch": 0.8300576547999229,
+      "grad_norm": 1.734375,
+      "learning_rate": 1.6553196920940192e-05,
+      "loss": 1.043,
+      "step": 4841
+    },
+    {
+      "epoch": 0.8302291188889127,
+      "grad_norm": 1.6875,
+      "learning_rate": 1.655183270327521e-05,
+      "loss": 1.0715,
+      "step": 4842
+    },
+    {
+      "epoch": 0.8304005829779025,
+      "grad_norm": 1.6015625,
+      "learning_rate": 1.6550468271930288e-05,
+      "loss": 1.0748,
+      "step": 4843
+    },
+    {
+      "epoch": 0.8305720470668925,
+      "grad_norm": 1.6015625,
+      "learning_rate": 1.6549103626949934e-05,
+      "loss": 0.9705,
+      "step": 4844
+    },
+    {
+      "epoch": 0.8307435111558823,
+      "grad_norm": 1.5546875,
+      "learning_rate": 1.6547738768378648e-05,
+      "loss": 0.9233,
+      "step": 4845
+    },
+    {
+      "epoch": 0.8309149752448721,
+      "grad_norm": 1.5234375,
+      "learning_rate": 1.654637369626094e-05,
+      "loss": 1.0039,
+      "step": 4846
+    },
+    {
+      "epoch": 0.8310864393338621,
+      "grad_norm": 1.6796875,
+      "learning_rate": 1.654500841064134e-05,
+      "loss": 1.0685,
+      "step": 4847
+    },
+    {
+      "epoch": 0.8312579034228519,
+      "grad_norm": 1.5703125,
+      "learning_rate": 1.6543642911564366e-05,
+      "loss": 0.9914,
+      "step": 4848
+    },
+    {
+      "epoch": 0.8314293675118417,
+      "grad_norm": 1.5625,
+      "learning_rate": 1.654227719907456e-05,
+      "loss": 1.0263,
+      "step": 4849
+    },
+    {
+      "epoch": 0.8316008316008316,
+      "grad_norm": 1.640625,
+      "learning_rate": 1.654091127321645e-05,
+      "loss": 0.9904,
+      "step": 4850
+    },
+    {
+      "epoch": 0.8317722956898215,
+      "grad_norm": 1.6171875,
+      "learning_rate": 1.65395451340346e-05,
+      "loss": 0.9814,
+      "step": 4851
+    },
+    {
+      "epoch": 0.8319437597788113,
+      "grad_norm": 1.546875,
+      "learning_rate": 1.6538178781573553e-05,
+      "loss": 1.0059,
+      "step": 4852
+    },
+    {
+      "epoch": 0.8321152238678012,
+      "grad_norm": 1.5546875,
+      "learning_rate": 1.6536812215877877e-05,
+      "loss": 1.0505,
+      "step": 4853
+    },
+    {
+      "epoch": 0.8322866879567911,
+      "grad_norm": 1.5390625,
+      "learning_rate": 1.6535445436992142e-05,
+      "loss": 0.9706,
+      "step": 4854
+    },
+    {
+      "epoch": 0.8324581520457809,
+      "grad_norm": 1.53125,
+      "learning_rate": 1.653407844496092e-05,
+      "loss": 0.9743,
+      "step": 4855
+    },
+    {
+      "epoch": 0.8326296161347708,
+      "grad_norm": 1.53125,
+      "learning_rate": 1.6532711239828792e-05,
+      "loss": 1.0592,
+      "step": 4856
+    },
+    {
+      "epoch": 0.8328010802237606,
+      "grad_norm": 1.671875,
+      "learning_rate": 1.6531343821640353e-05,
+      "loss": 1.1352,
+      "step": 4857
+    },
+    {
+      "epoch": 0.8329725443127505,
+      "grad_norm": 1.6171875,
+      "learning_rate": 1.6529976190440197e-05,
+      "loss": 1.0237,
+      "step": 4858
+    },
+    {
+      "epoch": 0.8331440084017404,
+      "grad_norm": 1.625,
+      "learning_rate": 1.652860834627293e-05,
+      "loss": 1.0109,
+      "step": 4859
+    },
+    {
+      "epoch": 0.8333154724907302,
+      "grad_norm": 1.5859375,
+      "learning_rate": 1.6527240289183156e-05,
+      "loss": 0.9852,
+      "step": 4860
+    },
+    {
+      "epoch": 0.8334869365797201,
+      "grad_norm": 1.5625,
+      "learning_rate": 1.6525872019215495e-05,
+      "loss": 1.0334,
+      "step": 4861
+    },
+    {
+      "epoch": 0.83365840066871,
+      "grad_norm": 1.515625,
+      "learning_rate": 1.652450353641458e-05,
+      "loss": 1.0522,
+      "step": 4862
+    },
+    {
+      "epoch": 0.8338298647576998,
+      "grad_norm": 1.7578125,
+      "learning_rate": 1.652313484082503e-05,
+      "loss": 1.0455,
+      "step": 4863
+    },
+    {
+      "epoch": 0.8340013288466896,
+      "grad_norm": 1.6328125,
+      "learning_rate": 1.6521765932491493e-05,
+      "loss": 1.0399,
+      "step": 4864
+    },
+    {
+      "epoch": 0.8341727929356796,
+      "grad_norm": 1.5859375,
+      "learning_rate": 1.652039681145861e-05,
+      "loss": 0.9762,
+      "step": 4865
+    },
+    {
+      "epoch": 0.8343442570246694,
+      "grad_norm": 1.5859375,
+      "learning_rate": 1.6519027477771034e-05,
+      "loss": 0.9692,
+      "step": 4866
+    },
+    {
+      "epoch": 0.8345157211136592,
+      "grad_norm": 1.640625,
+      "learning_rate": 1.651765793147342e-05,
+      "loss": 1.0545,
+      "step": 4867
+    },
+    {
+      "epoch": 0.8346871852026492,
+      "grad_norm": 1.546875,
+      "learning_rate": 1.651628817261044e-05,
+      "loss": 1.0397,
+      "step": 4868
+    },
+    {
+      "epoch": 0.834858649291639,
+      "grad_norm": 1.609375,
+      "learning_rate": 1.651491820122676e-05,
+      "loss": 1.059,
+      "step": 4869
+    },
+    {
+      "epoch": 0.8350301133806288,
+      "grad_norm": 1.59375,
+      "learning_rate": 1.651354801736707e-05,
+      "loss": 1.0279,
+      "step": 4870
+    },
+    {
+      "epoch": 0.8352015774696188,
+      "grad_norm": 1.5078125,
+      "learning_rate": 1.651217762107605e-05,
+      "loss": 0.9188,
+      "step": 4871
+    },
+    {
+      "epoch": 0.8353730415586086,
+      "grad_norm": 1.5234375,
+      "learning_rate": 1.6510807012398394e-05,
+      "loss": 1.0087,
+      "step": 4872
+    },
+    {
+      "epoch": 0.8355445056475984,
+      "grad_norm": 1.625,
+      "learning_rate": 1.650943619137881e-05,
+      "loss": 1.0444,
+      "step": 4873
+    },
+    {
+      "epoch": 0.8357159697365882,
+      "grad_norm": 1.5078125,
+      "learning_rate": 1.6508065158061992e-05,
+      "loss": 1.052,
+      "step": 4874
+    },
+    {
+      "epoch": 0.8358874338255782,
+      "grad_norm": 1.53125,
+      "learning_rate": 1.6506693912492666e-05,
+      "loss": 1.0841,
+      "step": 4875
+    },
+    {
+      "epoch": 0.836058897914568,
+      "grad_norm": 1.515625,
+      "learning_rate": 1.650532245471555e-05,
+      "loss": 0.9297,
+      "step": 4876
+    },
+    {
+      "epoch": 0.8362303620035578,
+      "grad_norm": 1.5546875,
+      "learning_rate": 1.6503950784775368e-05,
+      "loss": 0.9995,
+      "step": 4877
+    },
+    {
+      "epoch": 0.8364018260925478,
+      "grad_norm": 1.703125,
+      "learning_rate": 1.6502578902716863e-05,
+      "loss": 0.9401,
+      "step": 4878
+    },
+    {
+      "epoch": 0.8365732901815376,
+      "grad_norm": 1.546875,
+      "learning_rate": 1.650120680858477e-05,
+      "loss": 1.0145,
+      "step": 4879
+    },
+    {
+      "epoch": 0.8367447542705274,
+      "grad_norm": 1.578125,
+      "learning_rate": 1.6499834502423848e-05,
+      "loss": 1.0129,
+      "step": 4880
+    },
+    {
+      "epoch": 0.8369162183595173,
+      "grad_norm": 1.6640625,
+      "learning_rate": 1.649846198427884e-05,
+      "loss": 1.1061,
+      "step": 4881
+    },
+    {
+      "epoch": 0.8370876824485072,
+      "grad_norm": 1.5234375,
+      "learning_rate": 1.6497089254194515e-05,
+      "loss": 1.017,
+      "step": 4882
+    },
+    {
+      "epoch": 0.837259146537497,
+      "grad_norm": 1.6328125,
+      "learning_rate": 1.649571631221565e-05,
+      "loss": 0.9935,
+      "step": 4883
+    },
+    {
+      "epoch": 0.8374306106264869,
+      "grad_norm": 1.6484375,
+      "learning_rate": 1.649434315838701e-05,
+      "loss": 1.0117,
+      "step": 4884
+    },
+    {
+      "epoch": 0.8376020747154768,
+      "grad_norm": 1.515625,
+      "learning_rate": 1.6492969792753387e-05,
+      "loss": 0.9968,
+      "step": 4885
+    },
+    {
+      "epoch": 0.8377735388044666,
+      "grad_norm": 1.5234375,
+      "learning_rate": 1.649159621535957e-05,
+      "loss": 1.0011,
+      "step": 4886
+    },
+    {
+      "epoch": 0.8379450028934565,
+      "grad_norm": 1.5546875,
+      "learning_rate": 1.6490222426250354e-05,
+      "loss": 1.019,
+      "step": 4887
+    },
+    {
+      "epoch": 0.8381164669824464,
+      "grad_norm": 1.578125,
+      "learning_rate": 1.6488848425470543e-05,
+      "loss": 1.0191,
+      "step": 4888
+    },
+    {
+      "epoch": 0.8382879310714362,
+      "grad_norm": 1.5703125,
+      "learning_rate": 1.648747421306495e-05,
+      "loss": 0.9839,
+      "step": 4889
+    },
+    {
+      "epoch": 0.8384593951604261,
+      "grad_norm": 1.53125,
+      "learning_rate": 1.64860997890784e-05,
+      "loss": 0.9773,
+      "step": 4890
+    },
+    {
+      "epoch": 0.8386308592494159,
+      "grad_norm": 1.546875,
+      "learning_rate": 1.6484725153555703e-05,
+      "loss": 1.0045,
+      "step": 4891
+    },
+    {
+      "epoch": 0.8388023233384058,
+      "grad_norm": 1.6015625,
+      "learning_rate": 1.6483350306541703e-05,
+      "loss": 1.0592,
+      "step": 4892
+    },
+    {
+      "epoch": 0.8389737874273957,
+      "grad_norm": 1.5625,
+      "learning_rate": 1.6481975248081233e-05,
+      "loss": 0.9609,
+      "step": 4893
+    },
+    {
+      "epoch": 0.8391452515163855,
+      "grad_norm": 1.5078125,
+      "learning_rate": 1.6480599978219143e-05,
+      "loss": 0.9875,
+      "step": 4894
+    },
+    {
+      "epoch": 0.8393167156053754,
+      "grad_norm": 1.5703125,
+      "learning_rate": 1.6479224497000287e-05,
+      "loss": 1.0036,
+      "step": 4895
+    },
+    {
+      "epoch": 0.8394881796943653,
+      "grad_norm": 1.5625,
+      "learning_rate": 1.647784880446952e-05,
+      "loss": 0.9743,
+      "step": 4896
+    },
+    {
+      "epoch": 0.8396596437833551,
+      "grad_norm": 1.46875,
+      "learning_rate": 1.647647290067171e-05,
+      "loss": 0.9982,
+      "step": 4897
+    },
+    {
+      "epoch": 0.8398311078723449,
+      "grad_norm": 1.5234375,
+      "learning_rate": 1.647509678565173e-05,
+      "loss": 1.1047,
+      "step": 4898
+    },
+    {
+      "epoch": 0.8400025719613349,
+      "grad_norm": 1.6015625,
+      "learning_rate": 1.647372045945446e-05,
+      "loss": 1.0122,
+      "step": 4899
+    },
+    {
+      "epoch": 0.8401740360503247,
+      "grad_norm": 1.5546875,
+      "learning_rate": 1.647234392212479e-05,
+      "loss": 0.9983,
+      "step": 4900
+    },
+    {
+      "epoch": 0.8401740360503247,
+      "eval_loss": 0.8566790223121643,
+      "eval_runtime": 836.9111,
+      "eval_samples_per_second": 2.986,
+      "eval_steps_per_second": 2.986,
+      "step": 4900
+    },
+    {
+      "epoch": 0.8403455001393145,
+      "grad_norm": 1.5859375,
+      "learning_rate": 1.647096717370761e-05,
+      "loss": 1.0025,
+      "step": 4901
+    },
+    {
+      "epoch": 0.8405169642283045,
+      "grad_norm": 1.5625,
+      "learning_rate": 1.6469590214247828e-05,
+      "loss": 1.0119,
+      "step": 4902
+    },
+    {
+      "epoch": 0.8406884283172943,
+      "grad_norm": 1.578125,
+      "learning_rate": 1.646821304379034e-05,
+      "loss": 1.0166,
+      "step": 4903
+    },
+    {
+      "epoch": 0.8408598924062841,
+      "grad_norm": 1.7109375,
+      "learning_rate": 1.6466835662380072e-05,
+      "loss": 1.0238,
+      "step": 4904
+    },
+    {
+      "epoch": 0.841031356495274,
+      "grad_norm": 1.5078125,
+      "learning_rate": 1.646545807006194e-05,
+      "loss": 0.9224,
+      "step": 4905
+    },
+    {
+      "epoch": 0.8412028205842639,
+      "grad_norm": 1.6015625,
+      "learning_rate": 1.646408026688087e-05,
+      "loss": 1.0507,
+      "step": 4906
+    },
+    {
+      "epoch": 0.8413742846732537,
+      "grad_norm": 1.5703125,
+      "learning_rate": 1.6462702252881805e-05,
+      "loss": 1.0593,
+      "step": 4907
+    },
+    {
+      "epoch": 0.8415457487622436,
+      "grad_norm": 1.46875,
+      "learning_rate": 1.6461324028109683e-05,
+      "loss": 0.9803,
+      "step": 4908
+    },
+    {
+      "epoch": 0.8417172128512335,
+      "grad_norm": 1.4765625,
+      "learning_rate": 1.6459945592609455e-05,
+      "loss": 0.9086,
+      "step": 4909
+    },
+    {
+      "epoch": 0.8418886769402233,
+      "grad_norm": 1.578125,
+      "learning_rate": 1.6458566946426072e-05,
+      "loss": 1.0014,
+      "step": 4910
+    },
+    {
+      "epoch": 0.8420601410292132,
+      "grad_norm": 1.609375,
+      "learning_rate": 1.6457188089604505e-05,
+      "loss": 0.9632,
+      "step": 4911
+    },
+    {
+      "epoch": 0.842231605118203,
+      "grad_norm": 1.640625,
+      "learning_rate": 1.6455809022189716e-05,
+      "loss": 1.0341,
+      "step": 4912
+    },
+    {
+      "epoch": 0.8424030692071929,
+      "grad_norm": 1.6328125,
+      "learning_rate": 1.645442974422668e-05,
+      "loss": 0.962,
+      "step": 4913
+    },
+    {
+      "epoch": 0.8425745332961828,
+      "grad_norm": 1.5625,
+      "learning_rate": 1.6453050255760392e-05,
+      "loss": 0.9729,
+      "step": 4914
+    },
+    {
+      "epoch": 0.8427459973851726,
+      "grad_norm": 1.671875,
+      "learning_rate": 1.645167055683583e-05,
+      "loss": 1.027,
+      "step": 4915
+    },
+    {
+      "epoch": 0.8429174614741625,
+      "grad_norm": 1.59375,
+      "learning_rate": 1.6450290647498e-05,
+      "loss": 1.1018,
+      "step": 4916
+    },
+    {
+      "epoch": 0.8430889255631524,
+      "grad_norm": 1.59375,
+      "learning_rate": 1.64489105277919e-05,
+      "loss": 0.9738,
+      "step": 4917
+    },
+    {
+      "epoch": 0.8432603896521422,
+      "grad_norm": 1.53125,
+      "learning_rate": 1.6447530197762545e-05,
+      "loss": 1.0227,
+      "step": 4918
+    },
+    {
+      "epoch": 0.843431853741132,
+      "grad_norm": 1.5390625,
+      "learning_rate": 1.644614965745495e-05,
+      "loss": 1.0394,
+      "step": 4919
+    },
+    {
+      "epoch": 0.843603317830122,
+      "grad_norm": 1.59375,
+      "learning_rate": 1.6444768906914143e-05,
+      "loss": 1.0459,
+      "step": 4920
+    },
+    {
+      "epoch": 0.8437747819191118,
+      "grad_norm": 1.5234375,
+      "learning_rate": 1.644338794618515e-05,
+      "loss": 0.9538,
+      "step": 4921
+    },
+    {
+      "epoch": 0.8439462460081016,
+      "grad_norm": 1.4296875,
+      "learning_rate": 1.6442006775313017e-05,
+      "loss": 0.9086,
+      "step": 4922
+    },
+    {
+      "epoch": 0.8441177100970916,
+      "grad_norm": 1.4609375,
+      "learning_rate": 1.6440625394342777e-05,
+      "loss": 0.9426,
+      "step": 4923
+    },
+    {
+      "epoch": 0.8442891741860814,
+      "grad_norm": 1.4921875,
+      "learning_rate": 1.6439243803319492e-05,
+      "loss": 0.908,
+      "step": 4924
+    },
+    {
+      "epoch": 0.8444606382750712,
+      "grad_norm": 1.5390625,
+      "learning_rate": 1.6437862002288226e-05,
+      "loss": 1.0451,
+      "step": 4925
+    },
+    {
+      "epoch": 0.8446321023640612,
+      "grad_norm": 1.5234375,
+      "learning_rate": 1.6436479991294032e-05,
+      "loss": 1.0059,
+      "step": 4926
+    },
+    {
+      "epoch": 0.844803566453051,
+      "grad_norm": 1.4921875,
+      "learning_rate": 1.6435097770381988e-05,
+      "loss": 1.0136,
+      "step": 4927
+    },
+    {
+      "epoch": 0.8449750305420408,
+      "grad_norm": 1.65625,
+      "learning_rate": 1.643371533959717e-05,
+      "loss": 1.049,
+      "step": 4928
+    },
+    {
+      "epoch": 0.8451464946310308,
+      "grad_norm": 1.5078125,
+      "learning_rate": 1.6432332698984672e-05,
+      "loss": 0.9991,
+      "step": 4929
+    },
+    {
+      "epoch": 0.8453179587200206,
+      "grad_norm": 1.5234375,
+      "learning_rate": 1.6430949848589583e-05,
+      "loss": 0.9708,
+      "step": 4930
+    },
+    {
+      "epoch": 0.8454894228090104,
+      "grad_norm": 1.5234375,
+      "learning_rate": 1.6429566788457007e-05,
+      "loss": 0.9411,
+      "step": 4931
+    },
+    {
+      "epoch": 0.8456608868980003,
+      "grad_norm": 1.7421875,
+      "learning_rate": 1.6428183518632036e-05,
+      "loss": 1.0671,
+      "step": 4932
+    },
+    {
+      "epoch": 0.8458323509869902,
+      "grad_norm": 1.5,
+      "learning_rate": 1.64268000391598e-05,
+      "loss": 0.9859,
+      "step": 4933
+    },
+    {
+      "epoch": 0.84600381507598,
+      "grad_norm": 1.46875,
+      "learning_rate": 1.6425416350085416e-05,
+      "loss": 0.9405,
+      "step": 4934
+    },
+    {
+      "epoch": 0.8461752791649699,
+      "grad_norm": 1.5,
+      "learning_rate": 1.6424032451454004e-05,
+      "loss": 0.9694,
+      "step": 4935
+    },
+    {
+      "epoch": 0.8463467432539598,
+      "grad_norm": 2.28125,
+      "learning_rate": 1.6422648343310706e-05,
+      "loss": 1.0205,
+      "step": 4936
+    },
+    {
+      "epoch": 0.8465182073429496,
+      "grad_norm": 1.546875,
+      "learning_rate": 1.642126402570066e-05,
+      "loss": 0.9597,
+      "step": 4937
+    },
+    {
+      "epoch": 0.8466896714319395,
+      "grad_norm": 1.484375,
+      "learning_rate": 1.6419879498669012e-05,
+      "loss": 1.0294,
+      "step": 4938
+    },
+    {
+      "epoch": 0.8468611355209293,
+      "grad_norm": 1.578125,
+      "learning_rate": 1.641849476226092e-05,
+      "loss": 1.0367,
+      "step": 4939
+    },
+    {
+      "epoch": 0.8470325996099192,
+      "grad_norm": 1.5546875,
+      "learning_rate": 1.641710981652154e-05,
+      "loss": 1.0519,
+      "step": 4940
+    },
+    {
+      "epoch": 0.8472040636989091,
+      "grad_norm": 1.6015625,
+      "learning_rate": 1.641572466149605e-05,
+      "loss": 0.9819,
+      "step": 4941
+    },
+    {
+      "epoch": 0.8473755277878989,
+      "grad_norm": 1.5625,
+      "learning_rate": 1.6414339297229614e-05,
+      "loss": 0.9086,
+      "step": 4942
+    },
+    {
+      "epoch": 0.8475469918768888,
+      "grad_norm": 1.625,
+      "learning_rate": 1.6412953723767426e-05,
+      "loss": 1.0298,
+      "step": 4943
+    },
+    {
+      "epoch": 0.8477184559658787,
+      "grad_norm": 1.578125,
+      "learning_rate": 1.6411567941154666e-05,
+      "loss": 0.9991,
+      "step": 4944
+    },
+    {
+      "epoch": 0.8478899200548685,
+      "grad_norm": 1.6328125,
+      "learning_rate": 1.641018194943653e-05,
+      "loss": 1.0165,
+      "step": 4945
+    },
+    {
+      "epoch": 0.8480613841438583,
+      "grad_norm": 2.640625,
+      "learning_rate": 1.6408795748658223e-05,
+      "loss": 1.0844,
+      "step": 4946
+    },
+    {
+      "epoch": 0.8482328482328483,
+      "grad_norm": 1.5,
+      "learning_rate": 1.640740933886495e-05,
+      "loss": 0.9461,
+      "step": 4947
+    },
+    {
+      "epoch": 0.8484043123218381,
+      "grad_norm": 1.5703125,
+      "learning_rate": 1.6406022720101935e-05,
+      "loss": 0.9793,
+      "step": 4948
+    },
+    {
+      "epoch": 0.8485757764108279,
+      "grad_norm": 1.5703125,
+      "learning_rate": 1.6404635892414392e-05,
+      "loss": 1.0438,
+      "step": 4949
+    },
+    {
+      "epoch": 0.8487472404998179,
+      "grad_norm": 1.5234375,
+      "learning_rate": 1.640324885584756e-05,
+      "loss": 0.9791,
+      "step": 4950
+    },
+    {
+      "epoch": 0.8489187045888077,
+      "grad_norm": 1.5546875,
+      "learning_rate": 1.6401861610446667e-05,
+      "loss": 0.9802,
+      "step": 4951
+    },
+    {
+      "epoch": 0.8490901686777975,
+      "grad_norm": 1.5546875,
+      "learning_rate": 1.640047415625696e-05,
+      "loss": 0.9295,
+      "step": 4952
+    },
+    {
+      "epoch": 0.8492616327667875,
+      "grad_norm": 1.65625,
+      "learning_rate": 1.639908649332369e-05,
+      "loss": 1.0223,
+      "step": 4953
+    },
+    {
+      "epoch": 0.8494330968557773,
+      "grad_norm": 1.5546875,
+      "learning_rate": 1.6397698621692112e-05,
+      "loss": 0.9678,
+      "step": 4954
+    },
+    {
+      "epoch": 0.8496045609447671,
+      "grad_norm": 1.6171875,
+      "learning_rate": 1.639631054140749e-05,
+      "loss": 1.0231,
+      "step": 4955
+    },
+    {
+      "epoch": 0.849776025033757,
+      "grad_norm": 1.5703125,
+      "learning_rate": 1.63949222525151e-05,
+      "loss": 1.0565,
+      "step": 4956
+    },
+    {
+      "epoch": 0.8499474891227469,
+      "grad_norm": 1.71875,
+      "learning_rate": 1.6393533755060204e-05,
+      "loss": 1.0848,
+      "step": 4957
+    },
+    {
+      "epoch": 0.8501189532117367,
+      "grad_norm": 1.703125,
+      "learning_rate": 1.6392145049088105e-05,
+      "loss": 1.1402,
+      "step": 4958
+    },
+    {
+      "epoch": 0.8502904173007266,
+      "grad_norm": 1.5546875,
+      "learning_rate": 1.639075613464408e-05,
+      "loss": 0.9722,
+      "step": 4959
+    },
+    {
+      "epoch": 0.8504618813897165,
+      "grad_norm": 1.515625,
+      "learning_rate": 1.6389367011773435e-05,
+      "loss": 1.0092,
+      "step": 4960
+    },
+    {
+      "epoch": 0.8506333454787063,
+      "grad_norm": 1.46875,
+      "learning_rate": 1.6387977680521472e-05,
+      "loss": 0.9036,
+      "step": 4961
+    },
+    {
+      "epoch": 0.8508048095676962,
+      "grad_norm": 1.5703125,
+      "learning_rate": 1.6386588140933503e-05,
+      "loss": 1.0088,
+      "step": 4962
+    },
+    {
+      "epoch": 0.850976273656686,
+      "grad_norm": 1.515625,
+      "learning_rate": 1.6385198393054843e-05,
+      "loss": 0.9637,
+      "step": 4963
+    },
+    {
+      "epoch": 0.8511477377456759,
+      "grad_norm": 1.546875,
+      "learning_rate": 1.638380843693082e-05,
+      "loss": 0.9548,
+      "step": 4964
+    },
+    {
+      "epoch": 0.8513192018346658,
+      "grad_norm": 1.5390625,
+      "learning_rate": 1.6382418272606763e-05,
+      "loss": 0.9837,
+      "step": 4965
+    },
+    {
+      "epoch": 0.8514906659236556,
+      "grad_norm": 1.578125,
+      "learning_rate": 1.6381027900128013e-05,
+      "loss": 1.1239,
+      "step": 4966
+    },
+    {
+      "epoch": 0.8516621300126455,
+      "grad_norm": 1.5390625,
+      "learning_rate": 1.6379637319539913e-05,
+      "loss": 0.9881,
+      "step": 4967
+    },
+    {
+      "epoch": 0.8518335941016354,
+      "grad_norm": 1.6015625,
+      "learning_rate": 1.637824653088782e-05,
+      "loss": 0.9896,
+      "step": 4968
+    },
+    {
+      "epoch": 0.8520050581906252,
+      "grad_norm": 1.6640625,
+      "learning_rate": 1.637685553421709e-05,
+      "loss": 1.053,
+      "step": 4969
+    },
+    {
+      "epoch": 0.852176522279615,
+      "grad_norm": 1.484375,
+      "learning_rate": 1.6375464329573084e-05,
+      "loss": 1.0277,
+      "step": 4970
+    },
+    {
+      "epoch": 0.8523479863686049,
+      "grad_norm": 1.6015625,
+      "learning_rate": 1.637407291700118e-05,
+      "loss": 0.9823,
+      "step": 4971
+    },
+    {
+      "epoch": 0.8525194504575948,
+      "grad_norm": 1.5859375,
+      "learning_rate": 1.6372681296546755e-05,
+      "loss": 0.9795,
+      "step": 4972
+    },
+    {
+      "epoch": 0.8526909145465846,
+      "grad_norm": 1.5703125,
+      "learning_rate": 1.6371289468255196e-05,
+      "loss": 1.042,
+      "step": 4973
+    },
+    {
+      "epoch": 0.8528623786355745,
+      "grad_norm": 1.5859375,
+      "learning_rate": 1.6369897432171896e-05,
+      "loss": 1.0476,
+      "step": 4974
+    },
+    {
+      "epoch": 0.8530338427245644,
+      "grad_norm": 1.671875,
+      "learning_rate": 1.6368505188342253e-05,
+      "loss": 1.0198,
+      "step": 4975
+    },
+    {
+      "epoch": 0.8532053068135542,
+      "grad_norm": 1.6796875,
+      "learning_rate": 1.636711273681167e-05,
+      "loss": 1.0091,
+      "step": 4976
+    },
+    {
+      "epoch": 0.853376770902544,
+      "grad_norm": 1.5234375,
+      "learning_rate": 1.6365720077625575e-05,
+      "loss": 1.0143,
+      "step": 4977
+    },
+    {
+      "epoch": 0.853548234991534,
+      "grad_norm": 1.5859375,
+      "learning_rate": 1.6364327210829373e-05,
+      "loss": 0.8933,
+      "step": 4978
+    },
+    {
+      "epoch": 0.8537196990805238,
+      "grad_norm": 1.5546875,
+      "learning_rate": 1.636293413646849e-05,
+      "loss": 0.9725,
+      "step": 4979
+    },
+    {
+      "epoch": 0.8538911631695136,
+      "grad_norm": 1.578125,
+      "learning_rate": 1.6361540854588364e-05,
+      "loss": 1.0688,
+      "step": 4980
+    },
+    {
+      "epoch": 0.8540626272585036,
+      "grad_norm": 1.6171875,
+      "learning_rate": 1.636014736523444e-05,
+      "loss": 1.0411,
+      "step": 4981
+    },
+    {
+      "epoch": 0.8542340913474934,
+      "grad_norm": 1.59375,
+      "learning_rate": 1.6358753668452162e-05,
+      "loss": 1.0243,
+      "step": 4982
+    },
+    {
+      "epoch": 0.8544055554364832,
+      "grad_norm": 1.6015625,
+      "learning_rate": 1.6357359764286978e-05,
+      "loss": 1.0091,
+      "step": 4983
+    },
+    {
+      "epoch": 0.8545770195254732,
+      "grad_norm": 1.5390625,
+      "learning_rate": 1.6355965652784353e-05,
+      "loss": 0.9835,
+      "step": 4984
+    },
+    {
+      "epoch": 0.854748483614463,
+      "grad_norm": 1.5859375,
+      "learning_rate": 1.635457133398975e-05,
+      "loss": 0.9444,
+      "step": 4985
+    },
+    {
+      "epoch": 0.8549199477034528,
+      "grad_norm": 1.5078125,
+      "learning_rate": 1.6353176807948654e-05,
+      "loss": 1.0489,
+      "step": 4986
+    },
+    {
+      "epoch": 0.8550914117924427,
+      "grad_norm": 1.4453125,
+      "learning_rate": 1.6351782074706536e-05,
+      "loss": 0.8774,
+      "step": 4987
+    },
+    {
+      "epoch": 0.8552628758814326,
+      "grad_norm": 1.5546875,
+      "learning_rate": 1.6350387134308884e-05,
+      "loss": 1.002,
+      "step": 4988
+    },
+    {
+      "epoch": 0.8554343399704224,
+      "grad_norm": 1.46875,
+      "learning_rate": 1.6348991986801197e-05,
+      "loss": 0.9584,
+      "step": 4989
+    },
+    {
+      "epoch": 0.8556058040594123,
+      "grad_norm": 1.484375,
+      "learning_rate": 1.634759663222897e-05,
+      "loss": 0.9562,
+      "step": 4990
+    },
+    {
+      "epoch": 0.8557772681484022,
+      "grad_norm": 1.7578125,
+      "learning_rate": 1.634620107063772e-05,
+      "loss": 0.9777,
+      "step": 4991
+    },
+    {
+      "epoch": 0.855948732237392,
+      "grad_norm": 1.5625,
+      "learning_rate": 1.634480530207295e-05,
+      "loss": 0.9741,
+      "step": 4992
+    },
+    {
+      "epoch": 0.8561201963263819,
+      "grad_norm": 1.59375,
+      "learning_rate": 1.6343409326580185e-05,
+      "loss": 1.0333,
+      "step": 4993
+    },
+    {
+      "epoch": 0.8562916604153717,
+      "grad_norm": 1.640625,
+      "learning_rate": 1.6342013144204956e-05,
+      "loss": 1.0337,
+      "step": 4994
+    },
+    {
+      "epoch": 0.8564631245043616,
+      "grad_norm": 1.625,
+      "learning_rate": 1.6340616754992803e-05,
+      "loss": 0.9508,
+      "step": 4995
+    },
+    {
+      "epoch": 0.8566345885933515,
+      "grad_norm": 1.4609375,
+      "learning_rate": 1.6339220158989256e-05,
+      "loss": 0.9595,
+      "step": 4996
+    },
+    {
+      "epoch": 0.8568060526823413,
+      "grad_norm": 1.5390625,
+      "learning_rate": 1.633782335623987e-05,
+      "loss": 0.9851,
+      "step": 4997
+    },
+    {
+      "epoch": 0.8569775167713312,
+      "grad_norm": 1.5703125,
+      "learning_rate": 1.6336426346790192e-05,
+      "loss": 1.0379,
+      "step": 4998
+    },
+    {
+      "epoch": 0.8571489808603211,
+      "grad_norm": 1.6015625,
+      "learning_rate": 1.6335029130685794e-05,
+      "loss": 1.1435,
+      "step": 4999
+    },
+    {
+      "epoch": 0.8573204449493109,
+      "grad_norm": 1.515625,
+      "learning_rate": 1.6333631707972245e-05,
+      "loss": 0.9131,
+      "step": 5000
+    },
+    {
+      "epoch": 0.8574919090383007,
+      "grad_norm": 1.4921875,
+      "learning_rate": 1.6332234078695113e-05,
+      "loss": 0.9825,
+      "step": 5001
+    },
+    {
+      "epoch": 0.8576633731272907,
+      "grad_norm": 1.625,
+      "learning_rate": 1.633083624289998e-05,
+      "loss": 1.008,
+      "step": 5002
+    },
+    {
+      "epoch": 0.8578348372162805,
+      "grad_norm": 1.671875,
+      "learning_rate": 1.6329438200632443e-05,
+      "loss": 1.1064,
+      "step": 5003
+    },
+    {
+      "epoch": 0.8580063013052703,
+      "grad_norm": 1.5078125,
+      "learning_rate": 1.6328039951938086e-05,
+      "loss": 0.9704,
+      "step": 5004
+    },
+    {
+      "epoch": 0.8581777653942603,
+      "grad_norm": 1.5234375,
+      "learning_rate": 1.632664149686252e-05,
+      "loss": 1.0046,
+      "step": 5005
+    },
+    {
+      "epoch": 0.8583492294832501,
+      "grad_norm": 1.4453125,
+      "learning_rate": 1.6325242835451353e-05,
+      "loss": 0.9914,
+      "step": 5006
+    },
+    {
+      "epoch": 0.8585206935722399,
+      "grad_norm": 1.5546875,
+      "learning_rate": 1.632384396775019e-05,
+      "loss": 1.0318,
+      "step": 5007
+    },
+    {
+      "epoch": 0.8586921576612299,
+      "grad_norm": 1.609375,
+      "learning_rate": 1.632244489380467e-05,
+      "loss": 1.0568,
+      "step": 5008
+    },
+    {
+      "epoch": 0.8588636217502197,
+      "grad_norm": 1.578125,
+      "learning_rate": 1.632104561366041e-05,
+      "loss": 1.0481,
+      "step": 5009
+    },
+    {
+      "epoch": 0.8590350858392095,
+      "grad_norm": 1.609375,
+      "learning_rate": 1.6319646127363053e-05,
+      "loss": 1.0812,
+      "step": 5010
+    },
+    {
+      "epoch": 0.8592065499281994,
+      "grad_norm": 1.5859375,
+      "learning_rate": 1.6318246434958234e-05,
+      "loss": 1.0274,
+      "step": 5011
+    },
+    {
+      "epoch": 0.8593780140171893,
+      "grad_norm": 1.5703125,
+      "learning_rate": 1.631684653649161e-05,
+      "loss": 0.9391,
+      "step": 5012
+    },
+    {
+      "epoch": 0.8595494781061791,
+      "grad_norm": 1.53125,
+      "learning_rate": 1.631544643200883e-05,
+      "loss": 0.9186,
+      "step": 5013
+    },
+    {
+      "epoch": 0.859720942195169,
+      "grad_norm": 1.671875,
+      "learning_rate": 1.6314046121555563e-05,
+      "loss": 0.9996,
+      "step": 5014
+    },
+    {
+      "epoch": 0.8598924062841589,
+      "grad_norm": 1.5078125,
+      "learning_rate": 1.631264560517748e-05,
+      "loss": 0.9689,
+      "step": 5015
+    },
+    {
+      "epoch": 0.8600638703731487,
+      "grad_norm": 1.5390625,
+      "learning_rate": 1.6311244882920243e-05,
+      "loss": 1.0294,
+      "step": 5016
+    },
+    {
+      "epoch": 0.8602353344621386,
+      "grad_norm": 1.46875,
+      "learning_rate": 1.6309843954829547e-05,
+      "loss": 0.9928,
+      "step": 5017
+    },
+    {
+      "epoch": 0.8604067985511284,
+      "grad_norm": 1.5390625,
+      "learning_rate": 1.630844282095108e-05,
+      "loss": 0.968,
+      "step": 5018
+    },
+    {
+      "epoch": 0.8605782626401183,
+      "grad_norm": 1.5,
+      "learning_rate": 1.6307041481330542e-05,
+      "loss": 1.0326,
+      "step": 5019
+    },
+    {
+      "epoch": 0.8607497267291082,
+      "grad_norm": 1.5234375,
+      "learning_rate": 1.6305639936013625e-05,
+      "loss": 1.0035,
+      "step": 5020
+    },
+    {
+      "epoch": 0.860921190818098,
+      "grad_norm": 1.515625,
+      "learning_rate": 1.6304238185046046e-05,
+      "loss": 0.9567,
+      "step": 5021
+    },
+    {
+      "epoch": 0.8610926549070879,
+      "grad_norm": 1.53125,
+      "learning_rate": 1.6302836228473518e-05,
+      "loss": 1.0195,
+      "step": 5022
+    },
+    {
+      "epoch": 0.8612641189960778,
+      "grad_norm": 1.5,
+      "learning_rate": 1.6301434066341774e-05,
+      "loss": 1.0022,
+      "step": 5023
+    },
+    {
+      "epoch": 0.8614355830850676,
+      "grad_norm": 1.5546875,
+      "learning_rate": 1.630003169869653e-05,
+      "loss": 1.0037,
+      "step": 5024
+    },
+    {
+      "epoch": 0.8616070471740574,
+      "grad_norm": 1.5703125,
+      "learning_rate": 1.629862912558353e-05,
+      "loss": 0.91,
+      "step": 5025
+    },
+    {
+      "epoch": 0.8617785112630474,
+      "grad_norm": 1.5234375,
+      "learning_rate": 1.6297226347048516e-05,
+      "loss": 0.9843,
+      "step": 5026
+    },
+    {
+      "epoch": 0.8619499753520372,
+      "grad_norm": 1.6171875,
+      "learning_rate": 1.6295823363137238e-05,
+      "loss": 0.9593,
+      "step": 5027
+    },
+    {
+      "epoch": 0.862121439441027,
+      "grad_norm": 1.5625,
+      "learning_rate": 1.629442017389545e-05,
+      "loss": 0.9987,
+      "step": 5028
+    },
+    {
+      "epoch": 0.862292903530017,
+      "grad_norm": 1.5703125,
+      "learning_rate": 1.629301677936892e-05,
+      "loss": 0.9576,
+      "step": 5029
+    },
+    {
+      "epoch": 0.8624643676190068,
+      "grad_norm": 1.4921875,
+      "learning_rate": 1.6291613179603417e-05,
+      "loss": 0.9212,
+      "step": 5030
+    },
+    {
+      "epoch": 0.8626358317079966,
+      "grad_norm": 1.609375,
+      "learning_rate": 1.6290209374644712e-05,
+      "loss": 1.0911,
+      "step": 5031
+    },
+    {
+      "epoch": 0.8628072957969866,
+      "grad_norm": 1.6171875,
+      "learning_rate": 1.6288805364538592e-05,
+      "loss": 1.0831,
+      "step": 5032
+    },
+    {
+      "epoch": 0.8629787598859764,
+      "grad_norm": 1.5546875,
+      "learning_rate": 1.6287401149330854e-05,
+      "loss": 0.9477,
+      "step": 5033
+    },
+    {
+      "epoch": 0.8631502239749662,
+      "grad_norm": 1.6171875,
+      "learning_rate": 1.6285996729067284e-05,
+      "loss": 0.993,
+      "step": 5034
+    },
+    {
+      "epoch": 0.8633216880639561,
+      "grad_norm": 1.6171875,
+      "learning_rate": 1.628459210379369e-05,
+      "loss": 0.9876,
+      "step": 5035
+    },
+    {
+      "epoch": 0.863493152152946,
+      "grad_norm": 1.5390625,
+      "learning_rate": 1.628318727355588e-05,
+      "loss": 0.8755,
+      "step": 5036
+    },
+    {
+      "epoch": 0.8636646162419358,
+      "grad_norm": 1.640625,
+      "learning_rate": 1.6281782238399677e-05,
+      "loss": 1.0774,
+      "step": 5037
+    },
+    {
+      "epoch": 0.8638360803309257,
+      "grad_norm": 1.5703125,
+      "learning_rate": 1.6280376998370896e-05,
+      "loss": 0.9763,
+      "step": 5038
+    },
+    {
+      "epoch": 0.8640075444199156,
+      "grad_norm": 1.578125,
+      "learning_rate": 1.6278971553515375e-05,
+      "loss": 1.0475,
+      "step": 5039
+    },
+    {
+      "epoch": 0.8641790085089054,
+      "grad_norm": 1.6484375,
+      "learning_rate": 1.6277565903878947e-05,
+      "loss": 1.0331,
+      "step": 5040
+    },
+    {
+      "epoch": 0.8643504725978953,
+      "grad_norm": 1.5390625,
+      "learning_rate": 1.6276160049507455e-05,
+      "loss": 1.018,
+      "step": 5041
+    },
+    {
+      "epoch": 0.8645219366868852,
+      "grad_norm": 1.5625,
+      "learning_rate": 1.627475399044675e-05,
+      "loss": 0.9472,
+      "step": 5042
+    },
+    {
+      "epoch": 0.864693400775875,
+      "grad_norm": 1.609375,
+      "learning_rate": 1.6273347726742685e-05,
+      "loss": 1.0766,
+      "step": 5043
+    },
+    {
+      "epoch": 0.8648648648648649,
+      "grad_norm": 1.6171875,
+      "learning_rate": 1.6271941258441134e-05,
+      "loss": 0.9634,
+      "step": 5044
+    },
+    {
+      "epoch": 0.8650363289538547,
+      "grad_norm": 1.625,
+      "learning_rate": 1.6270534585587957e-05,
+      "loss": 0.971,
+      "step": 5045
+    },
+    {
+      "epoch": 0.8652077930428446,
+      "grad_norm": 1.4453125,
+      "learning_rate": 1.6269127708229032e-05,
+      "loss": 0.9035,
+      "step": 5046
+    },
+    {
+      "epoch": 0.8653792571318345,
+      "grad_norm": 1.5390625,
+      "learning_rate": 1.6267720626410248e-05,
+      "loss": 1.032,
+      "step": 5047
+    },
+    {
+      "epoch": 0.8655507212208243,
+      "grad_norm": 1.5546875,
+      "learning_rate": 1.6266313340177494e-05,
+      "loss": 1.0,
+      "step": 5048
+    },
+    {
+      "epoch": 0.8657221853098142,
+      "grad_norm": 1.5390625,
+      "learning_rate": 1.6264905849576663e-05,
+      "loss": 1.0141,
+      "step": 5049
+    },
+    {
+      "epoch": 0.8658936493988041,
+      "grad_norm": 1.5703125,
+      "learning_rate": 1.626349815465366e-05,
+      "loss": 0.9323,
+      "step": 5050
+    },
+    {
+      "epoch": 0.8660651134877939,
+      "grad_norm": 1.8125,
+      "learning_rate": 1.62620902554544e-05,
+      "loss": 1.0563,
+      "step": 5051
+    },
+    {
+      "epoch": 0.8662365775767837,
+      "grad_norm": 1.5625,
+      "learning_rate": 1.626068215202479e-05,
+      "loss": 1.0467,
+      "step": 5052
+    },
+    {
+      "epoch": 0.8664080416657737,
+      "grad_norm": 1.5859375,
+      "learning_rate": 1.6259273844410762e-05,
+      "loss": 0.9192,
+      "step": 5053
+    },
+    {
+      "epoch": 0.8665795057547635,
+      "grad_norm": 1.59375,
+      "learning_rate": 1.6257865332658248e-05,
+      "loss": 0.9773,
+      "step": 5054
+    },
+    {
+      "epoch": 0.8667509698437533,
+      "grad_norm": 1.5703125,
+      "learning_rate": 1.6256456616813176e-05,
+      "loss": 0.9858,
+      "step": 5055
+    },
+    {
+      "epoch": 0.8669224339327433,
+      "grad_norm": 1.5859375,
+      "learning_rate": 1.6255047696921492e-05,
+      "loss": 0.9744,
+      "step": 5056
+    },
+    {
+      "epoch": 0.8670938980217331,
+      "grad_norm": 1.59375,
+      "learning_rate": 1.6253638573029153e-05,
+      "loss": 1.067,
+      "step": 5057
+    },
+    {
+      "epoch": 0.8672653621107229,
+      "grad_norm": 1.5625,
+      "learning_rate": 1.625222924518211e-05,
+      "loss": 1.0363,
+      "step": 5058
+    },
+    {
+      "epoch": 0.8674368261997129,
+      "grad_norm": 1.4921875,
+      "learning_rate": 1.6250819713426326e-05,
+      "loss": 0.9379,
+      "step": 5059
+    },
+    {
+      "epoch": 0.8676082902887027,
+      "grad_norm": 1.53125,
+      "learning_rate": 1.6249409977807772e-05,
+      "loss": 1.0446,
+      "step": 5060
+    },
+    {
+      "epoch": 0.8677797543776925,
+      "grad_norm": 1.5234375,
+      "learning_rate": 1.624800003837243e-05,
+      "loss": 1.0111,
+      "step": 5061
+    },
+    {
+      "epoch": 0.8679512184666824,
+      "grad_norm": 1.5546875,
+      "learning_rate": 1.6246589895166277e-05,
+      "loss": 0.9468,
+      "step": 5062
+    },
+    {
+      "epoch": 0.8681226825556723,
+      "grad_norm": 1.6796875,
+      "learning_rate": 1.6245179548235303e-05,
+      "loss": 0.9303,
+      "step": 5063
+    },
+    {
+      "epoch": 0.8682941466446621,
+      "grad_norm": 1.59375,
+      "learning_rate": 1.624376899762551e-05,
+      "loss": 0.983,
+      "step": 5064
+    },
+    {
+      "epoch": 0.8684656107336519,
+      "grad_norm": 1.5625,
+      "learning_rate": 1.6242358243382894e-05,
+      "loss": 0.9681,
+      "step": 5065
+    },
+    {
+      "epoch": 0.8686370748226419,
+      "grad_norm": 1.6015625,
+      "learning_rate": 1.6240947285553473e-05,
+      "loss": 1.0484,
+      "step": 5066
+    },
+    {
+      "epoch": 0.8688085389116317,
+      "grad_norm": 1.5234375,
+      "learning_rate": 1.6239536124183258e-05,
+      "loss": 1.0441,
+      "step": 5067
+    },
+    {
+      "epoch": 0.8689800030006215,
+      "grad_norm": 1.59375,
+      "learning_rate": 1.623812475931827e-05,
+      "loss": 0.9782,
+      "step": 5068
+    },
+    {
+      "epoch": 0.8691514670896114,
+      "grad_norm": 1.46875,
+      "learning_rate": 1.623671319100455e-05,
+      "loss": 0.9758,
+      "step": 5069
+    },
+    {
+      "epoch": 0.8693229311786013,
+      "grad_norm": 1.515625,
+      "learning_rate": 1.6235301419288125e-05,
+      "loss": 0.924,
+      "step": 5070
+    },
+    {
+      "epoch": 0.8694943952675911,
+      "grad_norm": 1.5390625,
+      "learning_rate": 1.6233889444215042e-05,
+      "loss": 0.9122,
+      "step": 5071
+    },
+    {
+      "epoch": 0.869665859356581,
+      "grad_norm": 1.515625,
+      "learning_rate": 1.6232477265831344e-05,
+      "loss": 1.0136,
+      "step": 5072
+    },
+    {
+      "epoch": 0.8698373234455709,
+      "grad_norm": 1.640625,
+      "learning_rate": 1.62310648841831e-05,
+      "loss": 1.0216,
+      "step": 5073
+    },
+    {
+      "epoch": 0.8700087875345607,
+      "grad_norm": 1.4765625,
+      "learning_rate": 1.622965229931636e-05,
+      "loss": 0.9828,
+      "step": 5074
+    },
+    {
+      "epoch": 0.8701802516235506,
+      "grad_norm": 1.671875,
+      "learning_rate": 1.6228239511277207e-05,
+      "loss": 0.9817,
+      "step": 5075
+    },
+    {
+      "epoch": 0.8703517157125404,
+      "grad_norm": 1.6328125,
+      "learning_rate": 1.6226826520111704e-05,
+      "loss": 1.0222,
+      "step": 5076
+    },
+    {
+      "epoch": 0.8705231798015303,
+      "grad_norm": 1.6171875,
+      "learning_rate": 1.622541332586594e-05,
+      "loss": 1.089,
+      "step": 5077
+    },
+    {
+      "epoch": 0.8706946438905202,
+      "grad_norm": 1.5859375,
+      "learning_rate": 1.622399992858601e-05,
+      "loss": 0.9925,
+      "step": 5078
+    },
+    {
+      "epoch": 0.87086610797951,
+      "grad_norm": 1.546875,
+      "learning_rate": 1.6222586328317998e-05,
+      "loss": 1.0387,
+      "step": 5079
+    },
+    {
+      "epoch": 0.8710375720684999,
+      "grad_norm": 1.640625,
+      "learning_rate": 1.622117252510802e-05,
+      "loss": 1.0318,
+      "step": 5080
+    },
+    {
+      "epoch": 0.8712090361574898,
+      "grad_norm": 1.65625,
+      "learning_rate": 1.6219758519002174e-05,
+      "loss": 1.0202,
+      "step": 5081
+    },
+    {
+      "epoch": 0.8713805002464796,
+      "grad_norm": 1.5625,
+      "learning_rate": 1.6218344310046585e-05,
+      "loss": 0.9705,
+      "step": 5082
+    },
+    {
+      "epoch": 0.8715519643354694,
+      "grad_norm": 1.5234375,
+      "learning_rate": 1.621692989828737e-05,
+      "loss": 1.0089,
+      "step": 5083
+    },
+    {
+      "epoch": 0.8717234284244594,
+      "grad_norm": 1.578125,
+      "learning_rate": 1.6215515283770666e-05,
+      "loss": 0.9512,
+      "step": 5084
+    },
+    {
+      "epoch": 0.8718948925134492,
+      "grad_norm": 1.515625,
+      "learning_rate": 1.6214100466542598e-05,
+      "loss": 1.0778,
+      "step": 5085
+    },
+    {
+      "epoch": 0.872066356602439,
+      "grad_norm": 1.5546875,
+      "learning_rate": 1.6212685446649313e-05,
+      "loss": 1.0466,
+      "step": 5086
+    },
+    {
+      "epoch": 0.872237820691429,
+      "grad_norm": 1.5078125,
+      "learning_rate": 1.6211270224136962e-05,
+      "loss": 0.9975,
+      "step": 5087
+    },
+    {
+      "epoch": 0.8724092847804188,
+      "grad_norm": 1.5,
+      "learning_rate": 1.6209854799051695e-05,
+      "loss": 0.9996,
+      "step": 5088
+    },
+    {
+      "epoch": 0.8725807488694086,
+      "grad_norm": 1.6484375,
+      "learning_rate": 1.6208439171439688e-05,
+      "loss": 1.0921,
+      "step": 5089
+    },
+    {
+      "epoch": 0.8727522129583986,
+      "grad_norm": 1.65625,
+      "learning_rate": 1.6207023341347094e-05,
+      "loss": 0.9706,
+      "step": 5090
+    },
+    {
+      "epoch": 0.8729236770473884,
+      "grad_norm": 1.5703125,
+      "learning_rate": 1.6205607308820097e-05,
+      "loss": 1.0387,
+      "step": 5091
+    },
+    {
+      "epoch": 0.8730951411363782,
+      "grad_norm": 1.6015625,
+      "learning_rate": 1.6204191073904877e-05,
+      "loss": 1.0514,
+      "step": 5092
+    },
+    {
+      "epoch": 0.8732666052253681,
+      "grad_norm": 1.6875,
+      "learning_rate": 1.6202774636647624e-05,
+      "loss": 1.0053,
+      "step": 5093
+    },
+    {
+      "epoch": 0.873438069314358,
+      "grad_norm": 1.5859375,
+      "learning_rate": 1.6201357997094534e-05,
+      "loss": 0.9727,
+      "step": 5094
+    },
+    {
+      "epoch": 0.8736095334033478,
+      "grad_norm": 1.515625,
+      "learning_rate": 1.6199941155291805e-05,
+      "loss": 1.0132,
+      "step": 5095
+    },
+    {
+      "epoch": 0.8737809974923377,
+      "grad_norm": 1.609375,
+      "learning_rate": 1.6198524111285648e-05,
+      "loss": 1.0101,
+      "step": 5096
+    },
+    {
+      "epoch": 0.8739524615813276,
+      "grad_norm": 1.5390625,
+      "learning_rate": 1.6197106865122282e-05,
+      "loss": 1.0265,
+      "step": 5097
+    },
+    {
+      "epoch": 0.8741239256703174,
+      "grad_norm": 1.546875,
+      "learning_rate": 1.619568941684792e-05,
+      "loss": 1.0285,
+      "step": 5098
+    },
+    {
+      "epoch": 0.8742953897593073,
+      "grad_norm": 1.4765625,
+      "learning_rate": 1.61942717665088e-05,
+      "loss": 0.9578,
+      "step": 5099
+    },
+    {
+      "epoch": 0.8744668538482971,
+      "grad_norm": 1.5078125,
+      "learning_rate": 1.6192853914151147e-05,
+      "loss": 1.0192,
+      "step": 5100
+    },
+    {
+      "epoch": 0.874638317937287,
+      "grad_norm": 1.5390625,
+      "learning_rate": 1.6191435859821215e-05,
+      "loss": 0.9872,
+      "step": 5101
+    },
+    {
+      "epoch": 0.8748097820262769,
+      "grad_norm": 1.59375,
+      "learning_rate": 1.6190017603565238e-05,
+      "loss": 1.0022,
+      "step": 5102
+    },
+    {
+      "epoch": 0.8749812461152667,
+      "grad_norm": 1.5234375,
+      "learning_rate": 1.618859914542948e-05,
+      "loss": 0.9653,
+      "step": 5103
+    },
+    {
+      "epoch": 0.8751527102042566,
+      "grad_norm": 1.578125,
+      "learning_rate": 1.61871804854602e-05,
+      "loss": 1.0534,
+      "step": 5104
+    },
+    {
+      "epoch": 0.8753241742932465,
+      "grad_norm": 1.6171875,
+      "learning_rate": 1.6185761623703668e-05,
+      "loss": 0.9797,
+      "step": 5105
+    },
+    {
+      "epoch": 0.8754956383822363,
+      "grad_norm": 1.6015625,
+      "learning_rate": 1.6184342560206153e-05,
+      "loss": 1.0583,
+      "step": 5106
+    },
+    {
+      "epoch": 0.8756671024712261,
+      "grad_norm": 1.5703125,
+      "learning_rate": 1.6182923295013935e-05,
+      "loss": 1.0972,
+      "step": 5107
+    },
+    {
+      "epoch": 0.8758385665602161,
+      "grad_norm": 1.5859375,
+      "learning_rate": 1.6181503828173314e-05,
+      "loss": 0.9856,
+      "step": 5108
+    },
+    {
+      "epoch": 0.8760100306492059,
+      "grad_norm": 1.53125,
+      "learning_rate": 1.618008415973057e-05,
+      "loss": 1.0702,
+      "step": 5109
+    },
+    {
+      "epoch": 0.8761814947381957,
+      "grad_norm": 1.640625,
+      "learning_rate": 1.6178664289732014e-05,
+      "loss": 1.0591,
+      "step": 5110
+    },
+    {
+      "epoch": 0.8763529588271857,
+      "grad_norm": 1.6328125,
+      "learning_rate": 1.6177244218223943e-05,
+      "loss": 0.9781,
+      "step": 5111
+    },
+    {
+      "epoch": 0.8765244229161755,
+      "grad_norm": 1.640625,
+      "learning_rate": 1.6175823945252684e-05,
+      "loss": 0.9926,
+      "step": 5112
+    },
+    {
+      "epoch": 0.8766958870051653,
+      "grad_norm": 1.46875,
+      "learning_rate": 1.6174403470864545e-05,
+      "loss": 0.951,
+      "step": 5113
+    },
+    {
+      "epoch": 0.8768673510941553,
+      "grad_norm": 1.4375,
+      "learning_rate": 1.6172982795105857e-05,
+      "loss": 0.9442,
+      "step": 5114
+    },
+    {
+      "epoch": 0.8770388151831451,
+      "grad_norm": 1.5703125,
+      "learning_rate": 1.6171561918022954e-05,
+      "loss": 0.9973,
+      "step": 5115
+    },
+    {
+      "epoch": 0.8772102792721349,
+      "grad_norm": 1.7109375,
+      "learning_rate": 1.6170140839662184e-05,
+      "loss": 0.9902,
+      "step": 5116
+    },
+    {
+      "epoch": 0.8773817433611248,
+      "grad_norm": 1.5859375,
+      "learning_rate": 1.6168719560069882e-05,
+      "loss": 1.1159,
+      "step": 5117
+    },
+    {
+      "epoch": 0.8775532074501147,
+      "grad_norm": 1.578125,
+      "learning_rate": 1.616729807929241e-05,
+      "loss": 0.9024,
+      "step": 5118
+    },
+    {
+      "epoch": 0.8777246715391045,
+      "grad_norm": 1.5234375,
+      "learning_rate": 1.6165876397376117e-05,
+      "loss": 0.9801,
+      "step": 5119
+    },
+    {
+      "epoch": 0.8778961356280944,
+      "grad_norm": 1.6953125,
+      "learning_rate": 1.616445451436738e-05,
+      "loss": 1.106,
+      "step": 5120
+    },
+    {
+      "epoch": 0.8780675997170843,
+      "grad_norm": 1.6875,
+      "learning_rate": 1.616303243031257e-05,
+      "loss": 0.9608,
+      "step": 5121
+    },
+    {
+      "epoch": 0.8782390638060741,
+      "grad_norm": 1.5390625,
+      "learning_rate": 1.6161610145258058e-05,
+      "loss": 1.0416,
+      "step": 5122
+    },
+    {
+      "epoch": 0.878410527895064,
+      "grad_norm": 1.6484375,
+      "learning_rate": 1.616018765925024e-05,
+      "loss": 1.0368,
+      "step": 5123
+    },
+    {
+      "epoch": 0.8785819919840538,
+      "grad_norm": 1.6640625,
+      "learning_rate": 1.6158764972335507e-05,
+      "loss": 0.9011,
+      "step": 5124
+    },
+    {
+      "epoch": 0.8787534560730437,
+      "grad_norm": 1.53125,
+      "learning_rate": 1.6157342084560258e-05,
+      "loss": 0.9417,
+      "step": 5125
+    },
+    {
+      "epoch": 0.8789249201620336,
+      "grad_norm": 1.5546875,
+      "learning_rate": 1.6155918995970892e-05,
+      "loss": 1.0066,
+      "step": 5126
+    },
+    {
+      "epoch": 0.8790963842510234,
+      "grad_norm": 1.5859375,
+      "learning_rate": 1.615449570661383e-05,
+      "loss": 1.0111,
+      "step": 5127
+    },
+    {
+      "epoch": 0.8792678483400133,
+      "grad_norm": 1.5546875,
+      "learning_rate": 1.6153072216535485e-05,
+      "loss": 1.0012,
+      "step": 5128
+    },
+    {
+      "epoch": 0.8794393124290032,
+      "grad_norm": 1.6640625,
+      "learning_rate": 1.6151648525782286e-05,
+      "loss": 1.0302,
+      "step": 5129
+    },
+    {
+      "epoch": 0.879610776517993,
+      "grad_norm": 1.6328125,
+      "learning_rate": 1.6150224634400666e-05,
+      "loss": 1.0521,
+      "step": 5130
+    },
+    {
+      "epoch": 0.8797822406069828,
+      "grad_norm": 1.625,
+      "learning_rate": 1.614880054243706e-05,
+      "loss": 1.0379,
+      "step": 5131
+    },
+    {
+      "epoch": 0.8799537046959728,
+      "grad_norm": 1.484375,
+      "learning_rate": 1.614737624993791e-05,
+      "loss": 0.9418,
+      "step": 5132
+    },
+    {
+      "epoch": 0.8801251687849626,
+      "grad_norm": 1.6796875,
+      "learning_rate": 1.6145951756949676e-05,
+      "loss": 1.0277,
+      "step": 5133
+    },
+    {
+      "epoch": 0.8802966328739524,
+      "grad_norm": 1.5859375,
+      "learning_rate": 1.6144527063518813e-05,
+      "loss": 0.9721,
+      "step": 5134
+    },
+    {
+      "epoch": 0.8804680969629424,
+      "grad_norm": 1.515625,
+      "learning_rate": 1.614310216969178e-05,
+      "loss": 1.0545,
+      "step": 5135
+    },
+    {
+      "epoch": 0.8806395610519322,
+      "grad_norm": 1.5859375,
+      "learning_rate": 1.6141677075515053e-05,
+      "loss": 1.0213,
+      "step": 5136
+    },
+    {
+      "epoch": 0.880811025140922,
+      "grad_norm": 1.5625,
+      "learning_rate": 1.614025178103511e-05,
+      "loss": 0.9924,
+      "step": 5137
+    },
+    {
+      "epoch": 0.880982489229912,
+      "grad_norm": 1.5546875,
+      "learning_rate": 1.6138826286298434e-05,
+      "loss": 1.0368,
+      "step": 5138
+    },
+    {
+      "epoch": 0.8811539533189018,
+      "grad_norm": 1.6328125,
+      "learning_rate": 1.6137400591351518e-05,
+      "loss": 1.032,
+      "step": 5139
+    },
+    {
+      "epoch": 0.8813254174078916,
+      "grad_norm": 1.53125,
+      "learning_rate": 1.6135974696240854e-05,
+      "loss": 1.0255,
+      "step": 5140
+    },
+    {
+      "epoch": 0.8814968814968815,
+      "grad_norm": 1.6015625,
+      "learning_rate": 1.6134548601012954e-05,
+      "loss": 0.9538,
+      "step": 5141
+    },
+    {
+      "epoch": 0.8816683455858714,
+      "grad_norm": 1.53125,
+      "learning_rate": 1.613312230571432e-05,
+      "loss": 0.9656,
+      "step": 5142
+    },
+    {
+      "epoch": 0.8818398096748612,
+      "grad_norm": 1.515625,
+      "learning_rate": 1.613169581039147e-05,
+      "loss": 1.0057,
+      "step": 5143
+    },
+    {
+      "epoch": 0.8820112737638511,
+      "grad_norm": 1.5859375,
+      "learning_rate": 1.6130269115090936e-05,
+      "loss": 1.0336,
+      "step": 5144
+    },
+    {
+      "epoch": 0.882182737852841,
+      "grad_norm": 1.6328125,
+      "learning_rate": 1.6128842219859238e-05,
+      "loss": 1.1094,
+      "step": 5145
+    },
+    {
+      "epoch": 0.8823542019418308,
+      "grad_norm": 1.609375,
+      "learning_rate": 1.6127415124742917e-05,
+      "loss": 1.0314,
+      "step": 5146
+    },
+    {
+      "epoch": 0.8825256660308207,
+      "grad_norm": 1.59375,
+      "learning_rate": 1.6125987829788514e-05,
+      "loss": 0.9689,
+      "step": 5147
+    },
+    {
+      "epoch": 0.8826971301198105,
+      "grad_norm": 1.5390625,
+      "learning_rate": 1.612456033504258e-05,
+      "loss": 0.9291,
+      "step": 5148
+    },
+    {
+      "epoch": 0.8828685942088004,
+      "grad_norm": 1.59375,
+      "learning_rate": 1.6123132640551674e-05,
+      "loss": 0.9894,
+      "step": 5149
+    },
+    {
+      "epoch": 0.8830400582977903,
+      "grad_norm": 1.5703125,
+      "learning_rate": 1.6121704746362352e-05,
+      "loss": 0.965,
+      "step": 5150
+    },
+    {
+      "epoch": 0.8832115223867801,
+      "grad_norm": 39.5,
+      "learning_rate": 1.6120276652521185e-05,
+      "loss": 1.0894,
+      "step": 5151
+    },
+    {
+      "epoch": 0.88338298647577,
+      "grad_norm": 1.5859375,
+      "learning_rate": 1.6118848359074753e-05,
+      "loss": 1.0246,
+      "step": 5152
+    },
+    {
+      "epoch": 0.8835544505647599,
+      "grad_norm": 1.5234375,
+      "learning_rate": 1.611741986606963e-05,
+      "loss": 0.974,
+      "step": 5153
+    },
+    {
+      "epoch": 0.8837259146537497,
+      "grad_norm": 1.65625,
+      "learning_rate": 1.611599117355241e-05,
+      "loss": 1.0874,
+      "step": 5154
+    },
+    {
+      "epoch": 0.8838973787427395,
+      "grad_norm": 1.640625,
+      "learning_rate": 1.611456228156969e-05,
+      "loss": 1.0439,
+      "step": 5155
+    },
+    {
+      "epoch": 0.8840688428317295,
+      "grad_norm": 1.65625,
+      "learning_rate": 1.611313319016807e-05,
+      "loss": 0.9967,
+      "step": 5156
+    },
+    {
+      "epoch": 0.8842403069207193,
+      "grad_norm": 1.5703125,
+      "learning_rate": 1.611170389939415e-05,
+      "loss": 0.9404,
+      "step": 5157
+    },
+    {
+      "epoch": 0.8844117710097091,
+      "grad_norm": 1.5703125,
+      "learning_rate": 1.6110274409294556e-05,
+      "loss": 0.9975,
+      "step": 5158
+    },
+    {
+      "epoch": 0.884583235098699,
+      "grad_norm": 1.625,
+      "learning_rate": 1.6108844719915905e-05,
+      "loss": 1.0937,
+      "step": 5159
+    },
+    {
+      "epoch": 0.8847546991876889,
+      "grad_norm": 1.53125,
+      "learning_rate": 1.610741483130482e-05,
+      "loss": 1.0137,
+      "step": 5160
+    },
+    {
+      "epoch": 0.8849261632766787,
+      "grad_norm": 1.59375,
+      "learning_rate": 1.6105984743507944e-05,
+      "loss": 1.0278,
+      "step": 5161
+    },
+    {
+      "epoch": 0.8850976273656685,
+      "grad_norm": 1.5078125,
+      "learning_rate": 1.610455445657191e-05,
+      "loss": 0.9889,
+      "step": 5162
+    },
+    {
+      "epoch": 0.8852690914546585,
+      "grad_norm": 1.6328125,
+      "learning_rate": 1.6103123970543366e-05,
+      "loss": 0.9304,
+      "step": 5163
+    },
+    {
+      "epoch": 0.8854405555436483,
+      "grad_norm": 1.6328125,
+      "learning_rate": 1.6101693285468968e-05,
+      "loss": 1.0058,
+      "step": 5164
+    },
+    {
+      "epoch": 0.8856120196326381,
+      "grad_norm": 1.4921875,
+      "learning_rate": 1.6100262401395376e-05,
+      "loss": 1.0408,
+      "step": 5165
+    },
+    {
+      "epoch": 0.8857834837216281,
+      "grad_norm": 1.5234375,
+      "learning_rate": 1.6098831318369253e-05,
+      "loss": 1.0108,
+      "step": 5166
+    },
+    {
+      "epoch": 0.8859549478106179,
+      "grad_norm": 1.4921875,
+      "learning_rate": 1.6097400036437276e-05,
+      "loss": 0.9301,
+      "step": 5167
+    },
+    {
+      "epoch": 0.8861264118996077,
+      "grad_norm": 1.53125,
+      "learning_rate": 1.6095968555646128e-05,
+      "loss": 0.9718,
+      "step": 5168
+    },
+    {
+      "epoch": 0.8862978759885977,
+      "grad_norm": 1.7421875,
+      "learning_rate": 1.6094536876042486e-05,
+      "loss": 1.1003,
+      "step": 5169
+    },
+    {
+      "epoch": 0.8864693400775875,
+      "grad_norm": 1.6015625,
+      "learning_rate": 1.6093104997673045e-05,
+      "loss": 1.0531,
+      "step": 5170
+    },
+    {
+      "epoch": 0.8866408041665773,
+      "grad_norm": 1.609375,
+      "learning_rate": 1.6091672920584508e-05,
+      "loss": 1.0346,
+      "step": 5171
+    },
+    {
+      "epoch": 0.8868122682555672,
+      "grad_norm": 1.5625,
+      "learning_rate": 1.6090240644823577e-05,
+      "loss": 0.9871,
+      "step": 5172
+    },
+    {
+      "epoch": 0.8869837323445571,
+      "grad_norm": 1.6015625,
+      "learning_rate": 1.6088808170436964e-05,
+      "loss": 0.9438,
+      "step": 5173
+    },
+    {
+      "epoch": 0.8871551964335469,
+      "grad_norm": 1.578125,
+      "learning_rate": 1.608737549747139e-05,
+      "loss": 1.0102,
+      "step": 5174
+    },
+    {
+      "epoch": 0.8873266605225368,
+      "grad_norm": 1.59375,
+      "learning_rate": 1.6085942625973577e-05,
+      "loss": 1.0018,
+      "step": 5175
+    },
+    {
+      "epoch": 0.8874981246115267,
+      "grad_norm": 1.4609375,
+      "learning_rate": 1.6084509555990258e-05,
+      "loss": 0.9204,
+      "step": 5176
+    },
+    {
+      "epoch": 0.8876695887005165,
+      "grad_norm": 1.59375,
+      "learning_rate": 1.608307628756817e-05,
+      "loss": 1.0178,
+      "step": 5177
+    },
+    {
+      "epoch": 0.8878410527895064,
+      "grad_norm": 1.5625,
+      "learning_rate": 1.608164282075406e-05,
+      "loss": 0.9534,
+      "step": 5178
+    },
+    {
+      "epoch": 0.8880125168784962,
+      "grad_norm": 1.4921875,
+      "learning_rate": 1.608020915559467e-05,
+      "loss": 1.0153,
+      "step": 5179
+    },
+    {
+      "epoch": 0.8881839809674861,
+      "grad_norm": 1.6953125,
+      "learning_rate": 1.607877529213677e-05,
+      "loss": 1.0366,
+      "step": 5180
+    },
+    {
+      "epoch": 0.888355445056476,
+      "grad_norm": 1.5703125,
+      "learning_rate": 1.607734123042711e-05,
+      "loss": 1.0342,
+      "step": 5181
+    },
+    {
+      "epoch": 0.8885269091454658,
+      "grad_norm": 1.484375,
+      "learning_rate": 1.6075906970512475e-05,
+      "loss": 0.9779,
+      "step": 5182
+    },
+    {
+      "epoch": 0.8886983732344557,
+      "grad_norm": 1.578125,
+      "learning_rate": 1.607447251243963e-05,
+      "loss": 0.994,
+      "step": 5183
+    },
+    {
+      "epoch": 0.8888698373234456,
+      "grad_norm": 1.59375,
+      "learning_rate": 1.6073037856255362e-05,
+      "loss": 0.9917,
+      "step": 5184
+    },
+    {
+      "epoch": 0.8890413014124354,
+      "grad_norm": 1.5703125,
+      "learning_rate": 1.607160300200646e-05,
+      "loss": 1.0062,
+      "step": 5185
+    },
+    {
+      "epoch": 0.8892127655014253,
+      "grad_norm": 1.609375,
+      "learning_rate": 1.6070167949739724e-05,
+      "loss": 1.039,
+      "step": 5186
+    },
+    {
+      "epoch": 0.8893842295904152,
+      "grad_norm": 1.5234375,
+      "learning_rate": 1.606873269950195e-05,
+      "loss": 0.9748,
+      "step": 5187
+    },
+    {
+      "epoch": 0.889555693679405,
+      "grad_norm": 1.59375,
+      "learning_rate": 1.606729725133995e-05,
+      "loss": 0.9604,
+      "step": 5188
+    },
+    {
+      "epoch": 0.8897271577683948,
+      "grad_norm": 1.5390625,
+      "learning_rate": 1.606586160530054e-05,
+      "loss": 0.9808,
+      "step": 5189
+    },
+    {
+      "epoch": 0.8898986218573848,
+      "grad_norm": 1.5703125,
+      "learning_rate": 1.606442576143054e-05,
+      "loss": 0.894,
+      "step": 5190
+    },
+    {
+      "epoch": 0.8900700859463746,
+      "grad_norm": 1.546875,
+      "learning_rate": 1.6062989719776782e-05,
+      "loss": 0.9603,
+      "step": 5191
+    },
+    {
+      "epoch": 0.8902415500353644,
+      "grad_norm": 1.5,
+      "learning_rate": 1.60615534803861e-05,
+      "loss": 0.9645,
+      "step": 5192
+    },
+    {
+      "epoch": 0.8904130141243544,
+      "grad_norm": 1.5,
+      "learning_rate": 1.6060117043305327e-05,
+      "loss": 0.9169,
+      "step": 5193
+    },
+    {
+      "epoch": 0.8905844782133442,
+      "grad_norm": 1.4765625,
+      "learning_rate": 1.605868040858132e-05,
+      "loss": 0.9474,
+      "step": 5194
+    },
+    {
+      "epoch": 0.890755942302334,
+      "grad_norm": 1.625,
+      "learning_rate": 1.605724357626093e-05,
+      "loss": 1.0531,
+      "step": 5195
+    },
+    {
+      "epoch": 0.890927406391324,
+      "grad_norm": 1.6328125,
+      "learning_rate": 1.605580654639102e-05,
+      "loss": 1.035,
+      "step": 5196
+    },
+    {
+      "epoch": 0.8910988704803138,
+      "grad_norm": 1.53125,
+      "learning_rate": 1.605436931901845e-05,
+      "loss": 1.0524,
+      "step": 5197
+    },
+    {
+      "epoch": 0.8912703345693036,
+      "grad_norm": 1.578125,
+      "learning_rate": 1.60529318941901e-05,
+      "loss": 1.1239,
+      "step": 5198
+    },
+    {
+      "epoch": 0.8914417986582935,
+      "grad_norm": 1.5078125,
+      "learning_rate": 1.6051494271952844e-05,
+      "loss": 1.0065,
+      "step": 5199
+    },
+    {
+      "epoch": 0.8916132627472834,
+      "grad_norm": 1.578125,
+      "learning_rate": 1.605005645235358e-05,
+      "loss": 0.9343,
+      "step": 5200
+    },
+    {
+      "epoch": 0.8917847268362732,
+      "grad_norm": 1.484375,
+      "learning_rate": 1.6048618435439184e-05,
+      "loss": 0.9976,
+      "step": 5201
+    },
+    {
+      "epoch": 0.8919561909252631,
+      "grad_norm": 1.53125,
+      "learning_rate": 1.604718022125657e-05,
+      "loss": 0.986,
+      "step": 5202
+    },
+    {
+      "epoch": 0.892127655014253,
+      "grad_norm": 1.4921875,
+      "learning_rate": 1.6045741809852636e-05,
+      "loss": 0.9586,
+      "step": 5203
+    },
+    {
+      "epoch": 0.8922991191032428,
+      "grad_norm": 1.625,
+      "learning_rate": 1.6044303201274293e-05,
+      "loss": 1.0265,
+      "step": 5204
+    },
+    {
+      "epoch": 0.8924705831922327,
+      "grad_norm": 1.515625,
+      "learning_rate": 1.6042864395568466e-05,
+      "loss": 0.9714,
+      "step": 5205
+    },
+    {
+      "epoch": 0.8926420472812225,
+      "grad_norm": 1.59375,
+      "learning_rate": 1.6041425392782073e-05,
+      "loss": 1.0083,
+      "step": 5206
+    },
+    {
+      "epoch": 0.8928135113702124,
+      "grad_norm": 1.5859375,
+      "learning_rate": 1.6039986192962048e-05,
+      "loss": 1.0321,
+      "step": 5207
+    },
+    {
+      "epoch": 0.8929849754592023,
+      "grad_norm": 1.6640625,
+      "learning_rate": 1.603854679615533e-05,
+      "loss": 1.0479,
+      "step": 5208
+    },
+    {
+      "epoch": 0.8931564395481921,
+      "grad_norm": 1.6875,
+      "learning_rate": 1.6037107202408862e-05,
+      "loss": 1.0948,
+      "step": 5209
+    },
+    {
+      "epoch": 0.893327903637182,
+      "grad_norm": 1.6015625,
+      "learning_rate": 1.6035667411769593e-05,
+      "loss": 1.0308,
+      "step": 5210
+    },
+    {
+      "epoch": 0.8934993677261719,
+      "grad_norm": 1.7578125,
+      "learning_rate": 1.6034227424284482e-05,
+      "loss": 1.0654,
+      "step": 5211
+    },
+    {
+      "epoch": 0.8936708318151617,
+      "grad_norm": 1.609375,
+      "learning_rate": 1.603278724000049e-05,
+      "loss": 0.976,
+      "step": 5212
+    },
+    {
+      "epoch": 0.8938422959041515,
+      "grad_norm": 1.6015625,
+      "learning_rate": 1.603134685896459e-05,
+      "loss": 1.0447,
+      "step": 5213
+    },
+    {
+      "epoch": 0.8940137599931415,
+      "grad_norm": 1.5390625,
+      "learning_rate": 1.602990628122376e-05,
+      "loss": 1.0081,
+      "step": 5214
+    },
+    {
+      "epoch": 0.8941852240821313,
+      "grad_norm": 1.6796875,
+      "learning_rate": 1.6028465506824978e-05,
+      "loss": 1.143,
+      "step": 5215
+    },
+    {
+      "epoch": 0.8943566881711211,
+      "grad_norm": 1.625,
+      "learning_rate": 1.6027024535815234e-05,
+      "loss": 1.0891,
+      "step": 5216
+    },
+    {
+      "epoch": 0.8945281522601111,
+      "grad_norm": 1.5078125,
+      "learning_rate": 1.6025583368241524e-05,
+      "loss": 0.9675,
+      "step": 5217
+    },
+    {
+      "epoch": 0.8946996163491009,
+      "grad_norm": 1.4609375,
+      "learning_rate": 1.602414200415085e-05,
+      "loss": 0.9697,
+      "step": 5218
+    },
+    {
+      "epoch": 0.8948710804380907,
+      "grad_norm": 1.5546875,
+      "learning_rate": 1.602270044359022e-05,
+      "loss": 1.0339,
+      "step": 5219
+    },
+    {
+      "epoch": 0.8950425445270807,
+      "grad_norm": 1.515625,
+      "learning_rate": 1.6021258686606652e-05,
+      "loss": 1.018,
+      "step": 5220
+    },
+    {
+      "epoch": 0.8952140086160705,
+      "grad_norm": 1.78125,
+      "learning_rate": 1.6019816733247158e-05,
+      "loss": 0.9383,
+      "step": 5221
+    },
+    {
+      "epoch": 0.8953854727050603,
+      "grad_norm": 1.53125,
+      "learning_rate": 1.601837458355878e-05,
+      "loss": 0.9684,
+      "step": 5222
+    },
+    {
+      "epoch": 0.8955569367940502,
+      "grad_norm": 1.453125,
+      "learning_rate": 1.6016932237588537e-05,
+      "loss": 1.0065,
+      "step": 5223
+    },
+    {
+      "epoch": 0.8957284008830401,
+      "grad_norm": 1.6171875,
+      "learning_rate": 1.601548969538348e-05,
+      "loss": 0.9799,
+      "step": 5224
+    },
+    {
+      "epoch": 0.8958998649720299,
+      "grad_norm": 1.5625,
+      "learning_rate": 1.6014046956990653e-05,
+      "loss": 1.0463,
+      "step": 5225
+    },
+    {
+      "epoch": 0.8960713290610198,
+      "grad_norm": 1.5390625,
+      "learning_rate": 1.6012604022457104e-05,
+      "loss": 1.0254,
+      "step": 5226
+    },
+    {
+      "epoch": 0.8962427931500097,
+      "grad_norm": 1.5703125,
+      "learning_rate": 1.6011160891829898e-05,
+      "loss": 1.0115,
+      "step": 5227
+    },
+    {
+      "epoch": 0.8964142572389995,
+      "grad_norm": 1.578125,
+      "learning_rate": 1.60097175651561e-05,
+      "loss": 0.982,
+      "step": 5228
+    },
+    {
+      "epoch": 0.8965857213279894,
+      "grad_norm": 1.5234375,
+      "learning_rate": 1.600827404248278e-05,
+      "loss": 0.9934,
+      "step": 5229
+    },
+    {
+      "epoch": 0.8967571854169792,
+      "grad_norm": 1.5859375,
+      "learning_rate": 1.600683032385702e-05,
+      "loss": 1.0717,
+      "step": 5230
+    },
+    {
+      "epoch": 0.8969286495059691,
+      "grad_norm": 1.578125,
+      "learning_rate": 1.6005386409325906e-05,
+      "loss": 1.0007,
+      "step": 5231
+    },
+    {
+      "epoch": 0.897100113594959,
+      "grad_norm": 1.53125,
+      "learning_rate": 1.6003942298936524e-05,
+      "loss": 0.9075,
+      "step": 5232
+    },
+    {
+      "epoch": 0.8972715776839488,
+      "grad_norm": 1.515625,
+      "learning_rate": 1.6002497992735973e-05,
+      "loss": 0.9794,
+      "step": 5233
+    },
+    {
+      "epoch": 0.8974430417729387,
+      "grad_norm": 1.4765625,
+      "learning_rate": 1.600105349077136e-05,
+      "loss": 0.9492,
+      "step": 5234
+    },
+    {
+      "epoch": 0.8976145058619286,
+      "grad_norm": 1.546875,
+      "learning_rate": 1.5999608793089797e-05,
+      "loss": 0.9647,
+      "step": 5235
+    },
+    {
+      "epoch": 0.8977859699509184,
+      "grad_norm": 1.640625,
+      "learning_rate": 1.5998163899738398e-05,
+      "loss": 1.0268,
+      "step": 5236
+    },
+    {
+      "epoch": 0.8979574340399082,
+      "grad_norm": 1.546875,
+      "learning_rate": 1.5996718810764285e-05,
+      "loss": 1.0403,
+      "step": 5237
+    },
+    {
+      "epoch": 0.8981288981288982,
+      "grad_norm": 1.6328125,
+      "learning_rate": 1.5995273526214596e-05,
+      "loss": 0.9826,
+      "step": 5238
+    },
+    {
+      "epoch": 0.898300362217888,
+      "grad_norm": 1.5234375,
+      "learning_rate": 1.5993828046136454e-05,
+      "loss": 0.9027,
+      "step": 5239
+    },
+    {
+      "epoch": 0.8984718263068778,
+      "grad_norm": 1.578125,
+      "learning_rate": 1.5992382370577013e-05,
+      "loss": 0.9631,
+      "step": 5240
+    },
+    {
+      "epoch": 0.8986432903958678,
+      "grad_norm": 1.578125,
+      "learning_rate": 1.5990936499583415e-05,
+      "loss": 1.037,
+      "step": 5241
+    },
+    {
+      "epoch": 0.8988147544848576,
+      "grad_norm": 1.53125,
+      "learning_rate": 1.598949043320282e-05,
+      "loss": 1.0115,
+      "step": 5242
+    },
+    {
+      "epoch": 0.8989862185738474,
+      "grad_norm": 1.578125,
+      "learning_rate": 1.598804417148239e-05,
+      "loss": 0.9891,
+      "step": 5243
+    },
+    {
+      "epoch": 0.8991576826628374,
+      "grad_norm": 1.4609375,
+      "learning_rate": 1.5986597714469288e-05,
+      "loss": 1.0014,
+      "step": 5244
+    },
+    {
+      "epoch": 0.8993291467518272,
+      "grad_norm": 1.53125,
+      "learning_rate": 1.5985151062210687e-05,
+      "loss": 0.9914,
+      "step": 5245
+    },
+    {
+      "epoch": 0.899500610840817,
+      "grad_norm": 1.625,
+      "learning_rate": 1.5983704214753777e-05,
+      "loss": 1.0197,
+      "step": 5246
+    },
+    {
+      "epoch": 0.8996720749298069,
+      "grad_norm": 1.5,
+      "learning_rate": 1.5982257172145738e-05,
+      "loss": 0.9477,
+      "step": 5247
+    },
+    {
+      "epoch": 0.8998435390187968,
+      "grad_norm": 1.5703125,
+      "learning_rate": 1.598080993443377e-05,
+      "loss": 0.9749,
+      "step": 5248
+    },
+    {
+      "epoch": 0.9000150031077866,
+      "grad_norm": 1.5234375,
+      "learning_rate": 1.5979362501665062e-05,
+      "loss": 0.9667,
+      "step": 5249
+    },
+    {
+      "epoch": 0.9001864671967765,
+      "grad_norm": 1.5078125,
+      "learning_rate": 1.597791487388683e-05,
+      "loss": 1.0182,
+      "step": 5250
+    },
+    {
+      "epoch": 0.9003579312857664,
+      "grad_norm": 1.5234375,
+      "learning_rate": 1.5976467051146284e-05,
+      "loss": 1.0425,
+      "step": 5251
+    },
+    {
+      "epoch": 0.9005293953747562,
+      "grad_norm": 1.546875,
+      "learning_rate": 1.597501903349064e-05,
+      "loss": 1.0029,
+      "step": 5252
+    },
+    {
+      "epoch": 0.900700859463746,
+      "grad_norm": 1.484375,
+      "learning_rate": 1.5973570820967125e-05,
+      "loss": 1.0055,
+      "step": 5253
+    },
+    {
+      "epoch": 0.9008723235527359,
+      "grad_norm": 1.640625,
+      "learning_rate": 1.5972122413622972e-05,
+      "loss": 1.0582,
+      "step": 5254
+    },
+    {
+      "epoch": 0.9010437876417258,
+      "grad_norm": 1.578125,
+      "learning_rate": 1.597067381150542e-05,
+      "loss": 1.0042,
+      "step": 5255
+    },
+    {
+      "epoch": 0.9012152517307156,
+      "grad_norm": 1.7421875,
+      "learning_rate": 1.5969225014661708e-05,
+      "loss": 1.101,
+      "step": 5256
+    },
+    {
+      "epoch": 0.9013867158197055,
+      "grad_norm": 1.484375,
+      "learning_rate": 1.5967776023139094e-05,
+      "loss": 0.9062,
+      "step": 5257
+    },
+    {
+      "epoch": 0.9015581799086954,
+      "grad_norm": 1.5859375,
+      "learning_rate": 1.596632683698483e-05,
+      "loss": 1.0323,
+      "step": 5258
+    },
+    {
+      "epoch": 0.9017296439976852,
+      "grad_norm": 1.6171875,
+      "learning_rate": 1.596487745624618e-05,
+      "loss": 1.0726,
+      "step": 5259
+    },
+    {
+      "epoch": 0.9019011080866751,
+      "grad_norm": 1.5859375,
+      "learning_rate": 1.5963427880970414e-05,
+      "loss": 1.0093,
+      "step": 5260
+    },
+    {
+      "epoch": 0.9020725721756649,
+      "grad_norm": 1.53125,
+      "learning_rate": 1.596197811120481e-05,
+      "loss": 0.9529,
+      "step": 5261
+    },
+    {
+      "epoch": 0.9022440362646548,
+      "grad_norm": 1.546875,
+      "learning_rate": 1.596052814699665e-05,
+      "loss": 0.9538,
+      "step": 5262
+    },
+    {
+      "epoch": 0.9024155003536447,
+      "grad_norm": 1.546875,
+      "learning_rate": 1.595907798839322e-05,
+      "loss": 1.0119,
+      "step": 5263
+    },
+    {
+      "epoch": 0.9025869644426345,
+      "grad_norm": 1.5390625,
+      "learning_rate": 1.5957627635441815e-05,
+      "loss": 0.9586,
+      "step": 5264
+    },
+    {
+      "epoch": 0.9027584285316244,
+      "grad_norm": 1.5625,
+      "learning_rate": 1.5956177088189742e-05,
+      "loss": 0.9865,
+      "step": 5265
+    },
+    {
+      "epoch": 0.9029298926206143,
+      "grad_norm": 1.6171875,
+      "learning_rate": 1.5954726346684303e-05,
+      "loss": 1.0195,
+      "step": 5266
+    },
+    {
+      "epoch": 0.9031013567096041,
+      "grad_norm": 1.5703125,
+      "learning_rate": 1.5953275410972817e-05,
+      "loss": 1.0341,
+      "step": 5267
+    },
+    {
+      "epoch": 0.903272820798594,
+      "grad_norm": 1.4765625,
+      "learning_rate": 1.59518242811026e-05,
+      "loss": 0.974,
+      "step": 5268
+    },
+    {
+      "epoch": 0.9034442848875839,
+      "grad_norm": 1.5,
+      "learning_rate": 1.595037295712098e-05,
+      "loss": 0.9357,
+      "step": 5269
+    },
+    {
+      "epoch": 0.9036157489765737,
+      "grad_norm": 1.546875,
+      "learning_rate": 1.5948921439075293e-05,
+      "loss": 1.0212,
+      "step": 5270
+    },
+    {
+      "epoch": 0.9037872130655635,
+      "grad_norm": 1.578125,
+      "learning_rate": 1.5947469727012876e-05,
+      "loss": 0.9645,
+      "step": 5271
+    },
+    {
+      "epoch": 0.9039586771545535,
+      "grad_norm": 1.5078125,
+      "learning_rate": 1.5946017820981073e-05,
+      "loss": 0.9818,
+      "step": 5272
+    },
+    {
+      "epoch": 0.9041301412435433,
+      "grad_norm": 1.5390625,
+      "learning_rate": 1.5944565721027243e-05,
+      "loss": 1.0271,
+      "step": 5273
+    },
+    {
+      "epoch": 0.9043016053325331,
+      "grad_norm": 1.5234375,
+      "learning_rate": 1.5943113427198735e-05,
+      "loss": 0.9747,
+      "step": 5274
+    },
+    {
+      "epoch": 0.9044730694215231,
+      "grad_norm": 1.546875,
+      "learning_rate": 1.594166093954292e-05,
+      "loss": 0.9688,
+      "step": 5275
+    },
+    {
+      "epoch": 0.9046445335105129,
+      "grad_norm": 1.625,
+      "learning_rate": 1.5940208258107164e-05,
+      "loss": 0.9406,
+      "step": 5276
+    },
+    {
+      "epoch": 0.9048159975995027,
+      "grad_norm": 1.5546875,
+      "learning_rate": 1.5938755382938853e-05,
+      "loss": 0.9435,
+      "step": 5277
+    },
+    {
+      "epoch": 0.9049874616884926,
+      "grad_norm": 1.5390625,
+      "learning_rate": 1.5937302314085363e-05,
+      "loss": 0.9882,
+      "step": 5278
+    },
+    {
+      "epoch": 0.9051589257774825,
+      "grad_norm": 1.5546875,
+      "learning_rate": 1.5935849051594086e-05,
+      "loss": 0.9744,
+      "step": 5279
+    },
+    {
+      "epoch": 0.9053303898664723,
+      "grad_norm": 1.515625,
+      "learning_rate": 1.5934395595512423e-05,
+      "loss": 0.9818,
+      "step": 5280
+    },
+    {
+      "epoch": 0.9055018539554622,
+      "grad_norm": 1.671875,
+      "learning_rate": 1.593294194588777e-05,
+      "loss": 1.0195,
+      "step": 5281
+    },
+    {
+      "epoch": 0.9056733180444521,
+      "grad_norm": 1.6015625,
+      "learning_rate": 1.593148810276754e-05,
+      "loss": 1.0049,
+      "step": 5282
+    },
+    {
+      "epoch": 0.9058447821334419,
+      "grad_norm": 1.71875,
+      "learning_rate": 1.5930034066199146e-05,
+      "loss": 1.0413,
+      "step": 5283
+    },
+    {
+      "epoch": 0.9060162462224318,
+      "grad_norm": 1.4921875,
+      "learning_rate": 1.5928579836230012e-05,
+      "loss": 0.9894,
+      "step": 5284
+    },
+    {
+      "epoch": 0.9061877103114216,
+      "grad_norm": 1.546875,
+      "learning_rate": 1.5927125412907563e-05,
+      "loss": 0.9478,
+      "step": 5285
+    },
+    {
+      "epoch": 0.9063591744004115,
+      "grad_norm": 1.5234375,
+      "learning_rate": 1.5925670796279235e-05,
+      "loss": 0.9541,
+      "step": 5286
+    },
+    {
+      "epoch": 0.9065306384894014,
+      "grad_norm": 1.5703125,
+      "learning_rate": 1.592421598639247e-05,
+      "loss": 0.969,
+      "step": 5287
+    },
+    {
+      "epoch": 0.9067021025783912,
+      "grad_norm": 1.5078125,
+      "learning_rate": 1.592276098329471e-05,
+      "loss": 0.9962,
+      "step": 5288
+    },
+    {
+      "epoch": 0.9068735666673811,
+      "grad_norm": 1.5390625,
+      "learning_rate": 1.5921305787033416e-05,
+      "loss": 1.0632,
+      "step": 5289
+    },
+    {
+      "epoch": 0.907045030756371,
+      "grad_norm": 1.609375,
+      "learning_rate": 1.5919850397656036e-05,
+      "loss": 0.9742,
+      "step": 5290
+    },
+    {
+      "epoch": 0.9072164948453608,
+      "grad_norm": 1.515625,
+      "learning_rate": 1.591839481521005e-05,
+      "loss": 1.0608,
+      "step": 5291
+    },
+    {
+      "epoch": 0.9073879589343506,
+      "grad_norm": 1.609375,
+      "learning_rate": 1.5916939039742915e-05,
+      "loss": 0.9504,
+      "step": 5292
+    },
+    {
+      "epoch": 0.9075594230233406,
+      "grad_norm": 1.609375,
+      "learning_rate": 1.591548307130212e-05,
+      "loss": 0.9848,
+      "step": 5293
+    },
+    {
+      "epoch": 0.9077308871123304,
+      "grad_norm": 1.578125,
+      "learning_rate": 1.591402690993515e-05,
+      "loss": 1.0109,
+      "step": 5294
+    },
+    {
+      "epoch": 0.9079023512013202,
+      "grad_norm": 1.5625,
+      "learning_rate": 1.591257055568949e-05,
+      "loss": 1.0148,
+      "step": 5295
+    },
+    {
+      "epoch": 0.9080738152903102,
+      "grad_norm": 1.5546875,
+      "learning_rate": 1.5911114008612637e-05,
+      "loss": 1.0152,
+      "step": 5296
+    },
+    {
+      "epoch": 0.9082452793793,
+      "grad_norm": 1.578125,
+      "learning_rate": 1.5909657268752096e-05,
+      "loss": 1.0008,
+      "step": 5297
+    },
+    {
+      "epoch": 0.9084167434682898,
+      "grad_norm": 1.65625,
+      "learning_rate": 1.590820033615538e-05,
+      "loss": 1.0524,
+      "step": 5298
+    },
+    {
+      "epoch": 0.9085882075572798,
+      "grad_norm": 1.546875,
+      "learning_rate": 1.5906743210870007e-05,
+      "loss": 0.9841,
+      "step": 5299
+    },
+    {
+      "epoch": 0.9087596716462696,
+      "grad_norm": 1.6171875,
+      "learning_rate": 1.590528589294349e-05,
+      "loss": 0.9474,
+      "step": 5300
+    },
+    {
+      "epoch": 0.9089311357352594,
+      "grad_norm": 1.8046875,
+      "learning_rate": 1.5903828382423366e-05,
+      "loss": 1.0609,
+      "step": 5301
+    },
+    {
+      "epoch": 0.9091025998242493,
+      "grad_norm": 1.546875,
+      "learning_rate": 1.5902370679357165e-05,
+      "loss": 0.9618,
+      "step": 5302
+    },
+    {
+      "epoch": 0.9092740639132392,
+      "grad_norm": 1.5,
+      "learning_rate": 1.5900912783792426e-05,
+      "loss": 0.9268,
+      "step": 5303
+    },
+    {
+      "epoch": 0.909445528002229,
+      "grad_norm": 1.578125,
+      "learning_rate": 1.5899454695776705e-05,
+      "loss": 1.0258,
+      "step": 5304
+    },
+    {
+      "epoch": 0.9096169920912189,
+      "grad_norm": 1.546875,
+      "learning_rate": 1.589799641535755e-05,
+      "loss": 0.9598,
+      "step": 5305
+    },
+    {
+      "epoch": 0.9097884561802088,
+      "grad_norm": 1.703125,
+      "learning_rate": 1.5896537942582523e-05,
+      "loss": 1.0599,
+      "step": 5306
+    },
+    {
+      "epoch": 0.9099599202691986,
+      "grad_norm": 1.5703125,
+      "learning_rate": 1.589507927749919e-05,
+      "loss": 0.9962,
+      "step": 5307
+    },
+    {
+      "epoch": 0.9101313843581885,
+      "grad_norm": 1.5703125,
+      "learning_rate": 1.589362042015512e-05,
+      "loss": 1.0258,
+      "step": 5308
+    },
+    {
+      "epoch": 0.9103028484471783,
+      "grad_norm": 1.5625,
+      "learning_rate": 1.5892161370597897e-05,
+      "loss": 0.9494,
+      "step": 5309
+    },
+    {
+      "epoch": 0.9104743125361682,
+      "grad_norm": 1.5625,
+      "learning_rate": 1.5890702128875103e-05,
+      "loss": 1.0038,
+      "step": 5310
+    },
+    {
+      "epoch": 0.9106457766251581,
+      "grad_norm": 1.6484375,
+      "learning_rate": 1.588924269503433e-05,
+      "loss": 0.9653,
+      "step": 5311
+    },
+    {
+      "epoch": 0.9108172407141479,
+      "grad_norm": 1.6640625,
+      "learning_rate": 1.5887783069123178e-05,
+      "loss": 1.0415,
+      "step": 5312
+    },
+    {
+      "epoch": 0.9109887048031378,
+      "grad_norm": 1.4921875,
+      "learning_rate": 1.5886323251189247e-05,
+      "loss": 0.9358,
+      "step": 5313
+    },
+    {
+      "epoch": 0.9111601688921277,
+      "grad_norm": 1.5546875,
+      "learning_rate": 1.5884863241280147e-05,
+      "loss": 0.9164,
+      "step": 5314
+    },
+    {
+      "epoch": 0.9113316329811175,
+      "grad_norm": 1.5234375,
+      "learning_rate": 1.5883403039443498e-05,
+      "loss": 0.9586,
+      "step": 5315
+    },
+    {
+      "epoch": 0.9115030970701073,
+      "grad_norm": 1.5546875,
+      "learning_rate": 1.5881942645726924e-05,
+      "loss": 1.0247,
+      "step": 5316
+    },
+    {
+      "epoch": 0.9116745611590973,
+      "grad_norm": 1.4765625,
+      "learning_rate": 1.5880482060178048e-05,
+      "loss": 0.9309,
+      "step": 5317
+    },
+    {
+      "epoch": 0.9118460252480871,
+      "grad_norm": 1.53125,
+      "learning_rate": 1.587902128284451e-05,
+      "loss": 0.9456,
+      "step": 5318
+    },
+    {
+      "epoch": 0.9120174893370769,
+      "grad_norm": 1.6328125,
+      "learning_rate": 1.5877560313773946e-05,
+      "loss": 0.9501,
+      "step": 5319
+    },
+    {
+      "epoch": 0.9121889534260669,
+      "grad_norm": 1.609375,
+      "learning_rate": 1.5876099153014008e-05,
+      "loss": 0.9963,
+      "step": 5320
+    },
+    {
+      "epoch": 0.9123604175150567,
+      "grad_norm": 1.53125,
+      "learning_rate": 1.587463780061235e-05,
+      "loss": 0.9477,
+      "step": 5321
+    },
+    {
+      "epoch": 0.9125318816040465,
+      "grad_norm": 1.5078125,
+      "learning_rate": 1.5873176256616634e-05,
+      "loss": 0.9406,
+      "step": 5322
+    },
+    {
+      "epoch": 0.9127033456930365,
+      "grad_norm": 1.5625,
+      "learning_rate": 1.587171452107452e-05,
+      "loss": 0.9296,
+      "step": 5323
+    },
+    {
+      "epoch": 0.9128748097820263,
+      "grad_norm": 1.578125,
+      "learning_rate": 1.5870252594033687e-05,
+      "loss": 1.0263,
+      "step": 5324
+    },
+    {
+      "epoch": 0.9130462738710161,
+      "grad_norm": 1.4921875,
+      "learning_rate": 1.5868790475541812e-05,
+      "loss": 0.934,
+      "step": 5325
+    },
+    {
+      "epoch": 0.913217737960006,
+      "grad_norm": 1.59375,
+      "learning_rate": 1.5867328165646583e-05,
+      "loss": 1.0729,
+      "step": 5326
+    },
+    {
+      "epoch": 0.9133892020489959,
+      "grad_norm": 1.546875,
+      "learning_rate": 1.5865865664395684e-05,
+      "loss": 0.9704,
+      "step": 5327
+    },
+    {
+      "epoch": 0.9135606661379857,
+      "grad_norm": 1.65625,
+      "learning_rate": 1.5864402971836813e-05,
+      "loss": 1.025,
+      "step": 5328
+    },
+    {
+      "epoch": 0.9137321302269756,
+      "grad_norm": 1.5546875,
+      "learning_rate": 1.5862940088017683e-05,
+      "loss": 1.0173,
+      "step": 5329
+    },
+    {
+      "epoch": 0.9139035943159655,
+      "grad_norm": 1.5703125,
+      "learning_rate": 1.5861477012986002e-05,
+      "loss": 1.0344,
+      "step": 5330
+    },
+    {
+      "epoch": 0.9140750584049553,
+      "grad_norm": 1.6171875,
+      "learning_rate": 1.5860013746789477e-05,
+      "loss": 0.9556,
+      "step": 5331
+    },
+    {
+      "epoch": 0.9142465224939452,
+      "grad_norm": 1.6171875,
+      "learning_rate": 1.585855028947584e-05,
+      "loss": 0.9806,
+      "step": 5332
+    },
+    {
+      "epoch": 0.914417986582935,
+      "grad_norm": 1.625,
+      "learning_rate": 1.5857086641092817e-05,
+      "loss": 0.9937,
+      "step": 5333
+    },
+    {
+      "epoch": 0.9145894506719249,
+      "grad_norm": 1.6171875,
+      "learning_rate": 1.585562280168814e-05,
+      "loss": 1.0304,
+      "step": 5334
+    },
+    {
+      "epoch": 0.9147609147609148,
+      "grad_norm": 1.578125,
+      "learning_rate": 1.5854158771309557e-05,
+      "loss": 1.0163,
+      "step": 5335
+    },
+    {
+      "epoch": 0.9149323788499046,
+      "grad_norm": 1.5703125,
+      "learning_rate": 1.585269455000481e-05,
+      "loss": 0.9672,
+      "step": 5336
+    },
+    {
+      "epoch": 0.9151038429388945,
+      "grad_norm": 1.5625,
+      "learning_rate": 1.5851230137821656e-05,
+      "loss": 1.0051,
+      "step": 5337
+    },
+    {
+      "epoch": 0.9152753070278844,
+      "grad_norm": 1.625,
+      "learning_rate": 1.584976553480785e-05,
+      "loss": 1.0918,
+      "step": 5338
+    },
+    {
+      "epoch": 0.9154467711168742,
+      "grad_norm": 1.5390625,
+      "learning_rate": 1.5848300741011163e-05,
+      "loss": 0.9945,
+      "step": 5339
+    },
+    {
+      "epoch": 0.915618235205864,
+      "grad_norm": 1.5703125,
+      "learning_rate": 1.5846835756479367e-05,
+      "loss": 0.9911,
+      "step": 5340
+    },
+    {
+      "epoch": 0.915789699294854,
+      "grad_norm": 1.5859375,
+      "learning_rate": 1.5845370581260243e-05,
+      "loss": 1.0362,
+      "step": 5341
+    },
+    {
+      "epoch": 0.9159611633838438,
+      "grad_norm": 1.703125,
+      "learning_rate": 1.5843905215401566e-05,
+      "loss": 1.0358,
+      "step": 5342
+    },
+    {
+      "epoch": 0.9161326274728336,
+      "grad_norm": 1.5234375,
+      "learning_rate": 1.5842439658951137e-05,
+      "loss": 1.0377,
+      "step": 5343
+    },
+    {
+      "epoch": 0.9163040915618236,
+      "grad_norm": 1.671875,
+      "learning_rate": 1.584097391195675e-05,
+      "loss": 1.104,
+      "step": 5344
+    },
+    {
+      "epoch": 0.9164755556508134,
+      "grad_norm": 1.515625,
+      "learning_rate": 1.583950797446621e-05,
+      "loss": 0.8953,
+      "step": 5345
+    },
+    {
+      "epoch": 0.9166470197398032,
+      "grad_norm": 1.6796875,
+      "learning_rate": 1.5838041846527325e-05,
+      "loss": 1.0442,
+      "step": 5346
+    },
+    {
+      "epoch": 0.9168184838287932,
+      "grad_norm": 1.5703125,
+      "learning_rate": 1.583657552818791e-05,
+      "loss": 0.9812,
+      "step": 5347
+    },
+    {
+      "epoch": 0.916989947917783,
+      "grad_norm": 1.640625,
+      "learning_rate": 1.583510901949579e-05,
+      "loss": 1.0907,
+      "step": 5348
+    },
+    {
+      "epoch": 0.9171614120067728,
+      "grad_norm": 1.6171875,
+      "learning_rate": 1.583364232049879e-05,
+      "loss": 0.9746,
+      "step": 5349
+    },
+    {
+      "epoch": 0.9173328760957626,
+      "grad_norm": 1.546875,
+      "learning_rate": 1.583217543124475e-05,
+      "loss": 1.0124,
+      "step": 5350
+    },
+    {
+      "epoch": 0.9175043401847526,
+      "grad_norm": 1.5546875,
+      "learning_rate": 1.5830708351781507e-05,
+      "loss": 1.044,
+      "step": 5351
+    },
+    {
+      "epoch": 0.9176758042737424,
+      "grad_norm": 1.5859375,
+      "learning_rate": 1.5829241082156907e-05,
+      "loss": 1.047,
+      "step": 5352
+    },
+    {
+      "epoch": 0.9178472683627322,
+      "grad_norm": 1.484375,
+      "learning_rate": 1.58277736224188e-05,
+      "loss": 0.946,
+      "step": 5353
+    },
+    {
+      "epoch": 0.9180187324517222,
+      "grad_norm": 1.6328125,
+      "learning_rate": 1.5826305972615053e-05,
+      "loss": 1.0134,
+      "step": 5354
+    },
+    {
+      "epoch": 0.918190196540712,
+      "grad_norm": 1.609375,
+      "learning_rate": 1.582483813279353e-05,
+      "loss": 1.0489,
+      "step": 5355
+    },
+    {
+      "epoch": 0.9183616606297018,
+      "grad_norm": 1.6796875,
+      "learning_rate": 1.58233701030021e-05,
+      "loss": 1.0232,
+      "step": 5356
+    },
+    {
+      "epoch": 0.9185331247186918,
+      "grad_norm": 1.5703125,
+      "learning_rate": 1.5821901883288642e-05,
+      "loss": 1.0153,
+      "step": 5357
+    },
+    {
+      "epoch": 0.9187045888076816,
+      "grad_norm": 1.625,
+      "learning_rate": 1.582043347370104e-05,
+      "loss": 1.0195,
+      "step": 5358
+    },
+    {
+      "epoch": 0.9188760528966714,
+      "grad_norm": 1.5859375,
+      "learning_rate": 1.5818964874287185e-05,
+      "loss": 1.0454,
+      "step": 5359
+    },
+    {
+      "epoch": 0.9190475169856613,
+      "grad_norm": 1.65625,
+      "learning_rate": 1.5817496085094974e-05,
+      "loss": 1.0176,
+      "step": 5360
+    },
+    {
+      "epoch": 0.9192189810746512,
+      "grad_norm": 1.6328125,
+      "learning_rate": 1.5816027106172307e-05,
+      "loss": 0.9667,
+      "step": 5361
+    },
+    {
+      "epoch": 0.919390445163641,
+      "grad_norm": 1.515625,
+      "learning_rate": 1.5814557937567097e-05,
+      "loss": 1.0203,
+      "step": 5362
+    },
+    {
+      "epoch": 0.9195619092526309,
+      "grad_norm": 1.59375,
+      "learning_rate": 1.5813088579327256e-05,
+      "loss": 1.0232,
+      "step": 5363
+    },
+    {
+      "epoch": 0.9197333733416208,
+      "grad_norm": 1.5234375,
+      "learning_rate": 1.5811619031500706e-05,
+      "loss": 0.9786,
+      "step": 5364
+    },
+    {
+      "epoch": 0.9199048374306106,
+      "grad_norm": 1.703125,
+      "learning_rate": 1.5810149294135376e-05,
+      "loss": 1.0583,
+      "step": 5365
+    },
+    {
+      "epoch": 0.9200763015196005,
+      "grad_norm": 1.625,
+      "learning_rate": 1.5808679367279197e-05,
+      "loss": 1.0685,
+      "step": 5366
+    },
+    {
+      "epoch": 0.9202477656085903,
+      "grad_norm": 1.6171875,
+      "learning_rate": 1.5807209250980112e-05,
+      "loss": 1.0277,
+      "step": 5367
+    },
+    {
+      "epoch": 0.9204192296975802,
+      "grad_norm": 1.5234375,
+      "learning_rate": 1.5805738945286066e-05,
+      "loss": 0.9964,
+      "step": 5368
+    },
+    {
+      "epoch": 0.9205906937865701,
+      "grad_norm": 1.640625,
+      "learning_rate": 1.580426845024501e-05,
+      "loss": 1.0182,
+      "step": 5369
+    },
+    {
+      "epoch": 0.9207621578755599,
+      "grad_norm": 1.6015625,
+      "learning_rate": 1.5802797765904903e-05,
+      "loss": 1.051,
+      "step": 5370
+    },
+    {
+      "epoch": 0.9209336219645498,
+      "grad_norm": 1.515625,
+      "learning_rate": 1.5801326892313707e-05,
+      "loss": 1.0433,
+      "step": 5371
+    },
+    {
+      "epoch": 0.9211050860535397,
+      "grad_norm": 1.5625,
+      "learning_rate": 1.57998558295194e-05,
+      "loss": 0.9663,
+      "step": 5372
+    },
+    {
+      "epoch": 0.9212765501425295,
+      "grad_norm": 1.65625,
+      "learning_rate": 1.579838457756995e-05,
+      "loss": 1.055,
+      "step": 5373
+    },
+    {
+      "epoch": 0.9214480142315193,
+      "grad_norm": 1.4921875,
+      "learning_rate": 1.579691313651335e-05,
+      "loss": 1.0098,
+      "step": 5374
+    },
+    {
+      "epoch": 0.9216194783205093,
+      "grad_norm": 1.46875,
+      "learning_rate": 1.5795441506397574e-05,
+      "loss": 0.9559,
+      "step": 5375
+    },
+    {
+      "epoch": 0.9217909424094991,
+      "grad_norm": 1.5390625,
+      "learning_rate": 1.5793969687270634e-05,
+      "loss": 0.9153,
+      "step": 5376
+    },
+    {
+      "epoch": 0.9219624064984889,
+      "grad_norm": 1.4921875,
+      "learning_rate": 1.5792497679180524e-05,
+      "loss": 0.9999,
+      "step": 5377
+    },
+    {
+      "epoch": 0.9221338705874789,
+      "grad_norm": 1.5078125,
+      "learning_rate": 1.5791025482175247e-05,
+      "loss": 0.9207,
+      "step": 5378
+    },
+    {
+      "epoch": 0.9223053346764687,
+      "grad_norm": 1.59375,
+      "learning_rate": 1.578955309630283e-05,
+      "loss": 0.9692,
+      "step": 5379
+    },
+    {
+      "epoch": 0.9224767987654585,
+      "grad_norm": 1.609375,
+      "learning_rate": 1.578808052161128e-05,
+      "loss": 1.0302,
+      "step": 5380
+    },
+    {
+      "epoch": 0.9226482628544485,
+      "grad_norm": 1.65625,
+      "learning_rate": 1.5786607758148628e-05,
+      "loss": 1.007,
+      "step": 5381
+    },
+    {
+      "epoch": 0.9228197269434383,
+      "grad_norm": 1.546875,
+      "learning_rate": 1.5785134805962907e-05,
+      "loss": 0.9924,
+      "step": 5382
+    },
+    {
+      "epoch": 0.9229911910324281,
+      "grad_norm": 1.5625,
+      "learning_rate": 1.578366166510216e-05,
+      "loss": 1.0786,
+      "step": 5383
+    },
+    {
+      "epoch": 0.923162655121418,
+      "grad_norm": 1.6875,
+      "learning_rate": 1.578218833561442e-05,
+      "loss": 1.1266,
+      "step": 5384
+    },
+    {
+      "epoch": 0.9233341192104079,
+      "grad_norm": 1.609375,
+      "learning_rate": 1.5780714817547745e-05,
+      "loss": 0.9992,
+      "step": 5385
+    },
+    {
+      "epoch": 0.9235055832993977,
+      "grad_norm": 1.6015625,
+      "learning_rate": 1.5779241110950195e-05,
+      "loss": 1.0309,
+      "step": 5386
+    },
+    {
+      "epoch": 0.9236770473883876,
+      "grad_norm": 1.5,
+      "learning_rate": 1.577776721586983e-05,
+      "loss": 0.9404,
+      "step": 5387
+    },
+    {
+      "epoch": 0.9238485114773775,
+      "grad_norm": 1.671875,
+      "learning_rate": 1.577629313235472e-05,
+      "loss": 0.9784,
+      "step": 5388
+    },
+    {
+      "epoch": 0.9240199755663673,
+      "grad_norm": 1.546875,
+      "learning_rate": 1.5774818860452933e-05,
+      "loss": 0.9775,
+      "step": 5389
+    },
+    {
+      "epoch": 0.9241914396553572,
+      "grad_norm": 1.6640625,
+      "learning_rate": 1.577334440021256e-05,
+      "loss": 0.9955,
+      "step": 5390
+    },
+    {
+      "epoch": 0.924362903744347,
+      "grad_norm": 1.65625,
+      "learning_rate": 1.577186975168169e-05,
+      "loss": 0.9753,
+      "step": 5391
+    },
+    {
+      "epoch": 0.9245343678333369,
+      "grad_norm": 1.546875,
+      "learning_rate": 1.577039491490841e-05,
+      "loss": 1.0302,
+      "step": 5392
+    },
+    {
+      "epoch": 0.9247058319223268,
+      "grad_norm": 1.5234375,
+      "learning_rate": 1.576891988994082e-05,
+      "loss": 0.9805,
+      "step": 5393
+    },
+    {
+      "epoch": 0.9248772960113166,
+      "grad_norm": 1.640625,
+      "learning_rate": 1.576744467682703e-05,
+      "loss": 1.0489,
+      "step": 5394
+    },
+    {
+      "epoch": 0.9250487601003065,
+      "grad_norm": 1.640625,
+      "learning_rate": 1.5765969275615153e-05,
+      "loss": 0.9113,
+      "step": 5395
+    },
+    {
+      "epoch": 0.9252202241892964,
+      "grad_norm": 1.484375,
+      "learning_rate": 1.5764493686353307e-05,
+      "loss": 1.0295,
+      "step": 5396
+    },
+    {
+      "epoch": 0.9253916882782862,
+      "grad_norm": 1.53125,
+      "learning_rate": 1.5763017909089608e-05,
+      "loss": 0.9964,
+      "step": 5397
+    },
+    {
+      "epoch": 0.925563152367276,
+      "grad_norm": 1.65625,
+      "learning_rate": 1.57615419438722e-05,
+      "loss": 0.958,
+      "step": 5398
+    },
+    {
+      "epoch": 0.925734616456266,
+      "grad_norm": 1.5859375,
+      "learning_rate": 1.576006579074921e-05,
+      "loss": 0.9351,
+      "step": 5399
+    },
+    {
+      "epoch": 0.9259060805452558,
+      "grad_norm": 1.5703125,
+      "learning_rate": 1.5758589449768784e-05,
+      "loss": 0.9385,
+      "step": 5400
+    },
+    {
+      "epoch": 0.9260775446342456,
+      "grad_norm": 1.578125,
+      "learning_rate": 1.575711292097907e-05,
+      "loss": 1.0104,
+      "step": 5401
+    },
+    {
+      "epoch": 0.9262490087232356,
+      "grad_norm": 1.71875,
+      "learning_rate": 1.5755636204428228e-05,
+      "loss": 1.0488,
+      "step": 5402
+    },
+    {
+      "epoch": 0.9264204728122254,
+      "grad_norm": 1.5390625,
+      "learning_rate": 1.5754159300164414e-05,
+      "loss": 0.9892,
+      "step": 5403
+    },
+    {
+      "epoch": 0.9265919369012152,
+      "grad_norm": 1.546875,
+      "learning_rate": 1.57526822082358e-05,
+      "loss": 1.0528,
+      "step": 5404
+    },
+    {
+      "epoch": 0.9267634009902052,
+      "grad_norm": 1.6640625,
+      "learning_rate": 1.5751204928690552e-05,
+      "loss": 1.0059,
+      "step": 5405
+    },
+    {
+      "epoch": 0.926934865079195,
+      "grad_norm": 1.4140625,
+      "learning_rate": 1.5749727461576852e-05,
+      "loss": 0.9006,
+      "step": 5406
+    },
+    {
+      "epoch": 0.9271063291681848,
+      "grad_norm": 1.578125,
+      "learning_rate": 1.5748249806942895e-05,
+      "loss": 1.0689,
+      "step": 5407
+    },
+    {
+      "epoch": 0.9272777932571747,
+      "grad_norm": 1.7734375,
+      "learning_rate": 1.5746771964836864e-05,
+      "loss": 0.9228,
+      "step": 5408
+    },
+    {
+      "epoch": 0.9274492573461646,
+      "grad_norm": 1.6328125,
+      "learning_rate": 1.5745293935306954e-05,
+      "loss": 1.0405,
+      "step": 5409
+    },
+    {
+      "epoch": 0.9276207214351544,
+      "grad_norm": 1.640625,
+      "learning_rate": 1.574381571840138e-05,
+      "loss": 1.0065,
+      "step": 5410
+    },
+    {
+      "epoch": 0.9277921855241443,
+      "grad_norm": 1.5546875,
+      "learning_rate": 1.5742337314168343e-05,
+      "loss": 1.0412,
+      "step": 5411
+    },
+    {
+      "epoch": 0.9279636496131342,
+      "grad_norm": 1.5234375,
+      "learning_rate": 1.574085872265606e-05,
+      "loss": 1.0683,
+      "step": 5412
+    },
+    {
+      "epoch": 0.928135113702124,
+      "grad_norm": 1.5625,
+      "learning_rate": 1.573937994391276e-05,
+      "loss": 1.0458,
+      "step": 5413
+    },
+    {
+      "epoch": 0.9283065777911139,
+      "grad_norm": 1.5390625,
+      "learning_rate": 1.5737900977986667e-05,
+      "loss": 0.9001,
+      "step": 5414
+    },
+    {
+      "epoch": 0.9284780418801037,
+      "grad_norm": 1.59375,
+      "learning_rate": 1.5736421824926016e-05,
+      "loss": 1.0702,
+      "step": 5415
+    },
+    {
+      "epoch": 0.9286495059690936,
+      "grad_norm": 1.46875,
+      "learning_rate": 1.573494248477905e-05,
+      "loss": 0.9452,
+      "step": 5416
+    },
+    {
+      "epoch": 0.9288209700580835,
+      "grad_norm": 1.5703125,
+      "learning_rate": 1.573346295759401e-05,
+      "loss": 0.9938,
+      "step": 5417
+    },
+    {
+      "epoch": 0.9289924341470733,
+      "grad_norm": 1.640625,
+      "learning_rate": 1.5731983243419154e-05,
+      "loss": 1.0848,
+      "step": 5418
+    },
+    {
+      "epoch": 0.9291638982360632,
+      "grad_norm": 1.5390625,
+      "learning_rate": 1.573050334230274e-05,
+      "loss": 1.0467,
+      "step": 5419
+    },
+    {
+      "epoch": 0.9293353623250531,
+      "grad_norm": 1.625,
+      "learning_rate": 1.5729023254293034e-05,
+      "loss": 0.9948,
+      "step": 5420
+    },
+    {
+      "epoch": 0.9295068264140429,
+      "grad_norm": 1.6171875,
+      "learning_rate": 1.5727542979438305e-05,
+      "loss": 0.968,
+      "step": 5421
+    },
+    {
+      "epoch": 0.9296782905030327,
+      "grad_norm": 1.5625,
+      "learning_rate": 1.5726062517786834e-05,
+      "loss": 1.0831,
+      "step": 5422
+    },
+    {
+      "epoch": 0.9298497545920227,
+      "grad_norm": 1.5234375,
+      "learning_rate": 1.57245818693869e-05,
+      "loss": 1.0289,
+      "step": 5423
+    },
+    {
+      "epoch": 0.9300212186810125,
+      "grad_norm": 1.5546875,
+      "learning_rate": 1.5723101034286794e-05,
+      "loss": 1.0607,
+      "step": 5424
+    },
+    {
+      "epoch": 0.9301926827700023,
+      "grad_norm": 1.5234375,
+      "learning_rate": 1.572162001253481e-05,
+      "loss": 1.0166,
+      "step": 5425
+    },
+    {
+      "epoch": 0.9303641468589923,
+      "grad_norm": 1.6640625,
+      "learning_rate": 1.5720138804179256e-05,
+      "loss": 1.0684,
+      "step": 5426
+    },
+    {
+      "epoch": 0.9305356109479821,
+      "grad_norm": 1.625,
+      "learning_rate": 1.5718657409268436e-05,
+      "loss": 1.0084,
+      "step": 5427
+    },
+    {
+      "epoch": 0.9307070750369719,
+      "grad_norm": 1.46875,
+      "learning_rate": 1.5717175827850657e-05,
+      "loss": 1.0336,
+      "step": 5428
+    },
+    {
+      "epoch": 0.9308785391259619,
+      "grad_norm": 1.5234375,
+      "learning_rate": 1.571569405997425e-05,
+      "loss": 1.0188,
+      "step": 5429
+    },
+    {
+      "epoch": 0.9310500032149517,
+      "grad_norm": 1.5,
+      "learning_rate": 1.5714212105687535e-05,
+      "loss": 0.9354,
+      "step": 5430
+    },
+    {
+      "epoch": 0.9312214673039415,
+      "grad_norm": 1.5390625,
+      "learning_rate": 1.571272996503885e-05,
+      "loss": 0.9608,
+      "step": 5431
+    },
+    {
+      "epoch": 0.9313929313929314,
+      "grad_norm": 1.5859375,
+      "learning_rate": 1.571124763807652e-05,
+      "loss": 1.0545,
+      "step": 5432
+    },
+    {
+      "epoch": 0.9315643954819213,
+      "grad_norm": 1.5546875,
+      "learning_rate": 1.5709765124848907e-05,
+      "loss": 0.9436,
+      "step": 5433
+    },
+    {
+      "epoch": 0.9317358595709111,
+      "grad_norm": 1.5390625,
+      "learning_rate": 1.5708282425404345e-05,
+      "loss": 0.9691,
+      "step": 5434
+    },
+    {
+      "epoch": 0.931907323659901,
+      "grad_norm": 1.515625,
+      "learning_rate": 1.5706799539791204e-05,
+      "loss": 0.9409,
+      "step": 5435
+    },
+    {
+      "epoch": 0.9320787877488909,
+      "grad_norm": 1.5859375,
+      "learning_rate": 1.5705316468057837e-05,
+      "loss": 1.0345,
+      "step": 5436
+    },
+    {
+      "epoch": 0.9322502518378807,
+      "grad_norm": 1.671875,
+      "learning_rate": 1.5703833210252613e-05,
+      "loss": 0.9729,
+      "step": 5437
+    },
+    {
+      "epoch": 0.9324217159268706,
+      "grad_norm": 1.4921875,
+      "learning_rate": 1.5702349766423913e-05,
+      "loss": 0.9774,
+      "step": 5438
+    },
+    {
+      "epoch": 0.9325931800158604,
+      "grad_norm": 1.5234375,
+      "learning_rate": 1.5700866136620114e-05,
+      "loss": 0.9663,
+      "step": 5439
+    },
+    {
+      "epoch": 0.9327646441048503,
+      "grad_norm": 1.5625,
+      "learning_rate": 1.56993823208896e-05,
+      "loss": 0.9804,
+      "step": 5440
+    },
+    {
+      "epoch": 0.9329361081938402,
+      "grad_norm": 1.5078125,
+      "learning_rate": 1.5697898319280767e-05,
+      "loss": 0.9341,
+      "step": 5441
+    },
+    {
+      "epoch": 0.93310757228283,
+      "grad_norm": 1.546875,
+      "learning_rate": 1.569641413184202e-05,
+      "loss": 1.0012,
+      "step": 5442
+    },
+    {
+      "epoch": 0.9332790363718199,
+      "grad_norm": 1.46875,
+      "learning_rate": 1.569492975862175e-05,
+      "loss": 1.0137,
+      "step": 5443
+    },
+    {
+      "epoch": 0.9334505004608097,
+      "grad_norm": 1.7109375,
+      "learning_rate": 1.569344519966838e-05,
+      "loss": 1.0661,
+      "step": 5444
+    },
+    {
+      "epoch": 0.9336219645497996,
+      "grad_norm": 1.4921875,
+      "learning_rate": 1.569196045503032e-05,
+      "loss": 0.9675,
+      "step": 5445
+    },
+    {
+      "epoch": 0.9337934286387894,
+      "grad_norm": 1.609375,
+      "learning_rate": 1.5690475524755996e-05,
+      "loss": 1.1199,
+      "step": 5446
+    },
+    {
+      "epoch": 0.9339648927277793,
+      "grad_norm": 1.5078125,
+      "learning_rate": 1.5688990408893837e-05,
+      "loss": 0.973,
+      "step": 5447
+    },
+    {
+      "epoch": 0.9341363568167692,
+      "grad_norm": 1.546875,
+      "learning_rate": 1.5687505107492278e-05,
+      "loss": 1.0056,
+      "step": 5448
+    },
+    {
+      "epoch": 0.934307820905759,
+      "grad_norm": 1.6171875,
+      "learning_rate": 1.568601962059976e-05,
+      "loss": 1.0917,
+      "step": 5449
+    },
+    {
+      "epoch": 0.9344792849947489,
+      "grad_norm": 1.5,
+      "learning_rate": 1.5684533948264734e-05,
+      "loss": 0.9369,
+      "step": 5450
+    },
+    {
+      "epoch": 0.9346507490837388,
+      "grad_norm": 1.5625,
+      "learning_rate": 1.5683048090535645e-05,
+      "loss": 1.0045,
+      "step": 5451
+    },
+    {
+      "epoch": 0.9348222131727286,
+      "grad_norm": 1.5546875,
+      "learning_rate": 1.568156204746096e-05,
+      "loss": 0.9499,
+      "step": 5452
+    },
+    {
+      "epoch": 0.9349936772617184,
+      "grad_norm": 1.6328125,
+      "learning_rate": 1.568007581908914e-05,
+      "loss": 1.0495,
+      "step": 5453
+    },
+    {
+      "epoch": 0.9351651413507084,
+      "grad_norm": 1.5625,
+      "learning_rate": 1.567858940546866e-05,
+      "loss": 1.0209,
+      "step": 5454
+    },
+    {
+      "epoch": 0.9353366054396982,
+      "grad_norm": 1.5234375,
+      "learning_rate": 1.5677102806647993e-05,
+      "loss": 1.0253,
+      "step": 5455
+    },
+    {
+      "epoch": 0.935508069528688,
+      "grad_norm": 1.625,
+      "learning_rate": 1.567561602267563e-05,
+      "loss": 1.0378,
+      "step": 5456
+    },
+    {
+      "epoch": 0.935679533617678,
+      "grad_norm": 1.5390625,
+      "learning_rate": 1.5674129053600054e-05,
+      "loss": 1.0162,
+      "step": 5457
+    },
+    {
+      "epoch": 0.9358509977066678,
+      "grad_norm": 1.59375,
+      "learning_rate": 1.5672641899469764e-05,
+      "loss": 0.9873,
+      "step": 5458
+    },
+    {
+      "epoch": 0.9360224617956576,
+      "grad_norm": 1.5703125,
+      "learning_rate": 1.5671154560333258e-05,
+      "loss": 1.0778,
+      "step": 5459
+    },
+    {
+      "epoch": 0.9361939258846476,
+      "grad_norm": 1.546875,
+      "learning_rate": 1.5669667036239047e-05,
+      "loss": 1.0814,
+      "step": 5460
+    },
+    {
+      "epoch": 0.9363653899736374,
+      "grad_norm": 1.5078125,
+      "learning_rate": 1.5668179327235644e-05,
+      "loss": 1.015,
+      "step": 5461
+    },
+    {
+      "epoch": 0.9365368540626272,
+      "grad_norm": 1.5546875,
+      "learning_rate": 1.566669143337157e-05,
+      "loss": 0.9743,
+      "step": 5462
+    },
+    {
+      "epoch": 0.9367083181516171,
+      "grad_norm": 1.5546875,
+      "learning_rate": 1.566520335469535e-05,
+      "loss": 0.9868,
+      "step": 5463
+    },
+    {
+      "epoch": 0.936879782240607,
+      "grad_norm": 1.625,
+      "learning_rate": 1.5663715091255513e-05,
+      "loss": 1.0205,
+      "step": 5464
+    },
+    {
+      "epoch": 0.9370512463295968,
+      "grad_norm": 1.4609375,
+      "learning_rate": 1.56622266431006e-05,
+      "loss": 0.9398,
+      "step": 5465
+    },
+    {
+      "epoch": 0.9372227104185867,
+      "grad_norm": 1.53125,
+      "learning_rate": 1.5660738010279156e-05,
+      "loss": 0.9776,
+      "step": 5466
+    },
+    {
+      "epoch": 0.9373941745075766,
+      "grad_norm": 1.6484375,
+      "learning_rate": 1.5659249192839724e-05,
+      "loss": 0.958,
+      "step": 5467
+    },
+    {
+      "epoch": 0.9375656385965664,
+      "grad_norm": 1.7265625,
+      "learning_rate": 1.565776019083087e-05,
+      "loss": 0.9683,
+      "step": 5468
+    },
+    {
+      "epoch": 0.9377371026855563,
+      "grad_norm": 1.703125,
+      "learning_rate": 1.565627100430115e-05,
+      "loss": 1.0691,
+      "step": 5469
+    },
+    {
+      "epoch": 0.9379085667745461,
+      "grad_norm": 1.546875,
+      "learning_rate": 1.5654781633299134e-05,
+      "loss": 0.9959,
+      "step": 5470
+    },
+    {
+      "epoch": 0.938080030863536,
+      "grad_norm": 1.640625,
+      "learning_rate": 1.5653292077873393e-05,
+      "loss": 1.0553,
+      "step": 5471
+    },
+    {
+      "epoch": 0.9382514949525259,
+      "grad_norm": 1.4765625,
+      "learning_rate": 1.565180233807251e-05,
+      "loss": 1.0085,
+      "step": 5472
+    },
+    {
+      "epoch": 0.9384229590415157,
+      "grad_norm": 1.609375,
+      "learning_rate": 1.565031241394507e-05,
+      "loss": 1.0366,
+      "step": 5473
+    },
+    {
+      "epoch": 0.9385944231305056,
+      "grad_norm": 1.4765625,
+      "learning_rate": 1.5648822305539667e-05,
+      "loss": 0.8648,
+      "step": 5474
+    },
+    {
+      "epoch": 0.9387658872194955,
+      "grad_norm": 1.5703125,
+      "learning_rate": 1.5647332012904892e-05,
+      "loss": 0.9774,
+      "step": 5475
+    },
+    {
+      "epoch": 0.9389373513084853,
+      "grad_norm": 1.5625,
+      "learning_rate": 1.564584153608936e-05,
+      "loss": 0.9816,
+      "step": 5476
+    },
+    {
+      "epoch": 0.9391088153974751,
+      "grad_norm": 1.609375,
+      "learning_rate": 1.5644350875141673e-05,
+      "loss": 1.0061,
+      "step": 5477
+    },
+    {
+      "epoch": 0.9392802794864651,
+      "grad_norm": 1.484375,
+      "learning_rate": 1.564286003011045e-05,
+      "loss": 0.9747,
+      "step": 5478
+    },
+    {
+      "epoch": 0.9394517435754549,
+      "grad_norm": 1.4921875,
+      "learning_rate": 1.5641369001044308e-05,
+      "loss": 0.8715,
+      "step": 5479
+    },
+    {
+      "epoch": 0.9396232076644447,
+      "grad_norm": 1.5,
+      "learning_rate": 1.5639877787991884e-05,
+      "loss": 0.9639,
+      "step": 5480
+    },
+    {
+      "epoch": 0.9397946717534347,
+      "grad_norm": 1.5390625,
+      "learning_rate": 1.5638386391001804e-05,
+      "loss": 1.0736,
+      "step": 5481
+    },
+    {
+      "epoch": 0.9399661358424245,
+      "grad_norm": 1.515625,
+      "learning_rate": 1.5636894810122717e-05,
+      "loss": 1.0084,
+      "step": 5482
+    },
+    {
+      "epoch": 0.9401375999314143,
+      "grad_norm": 1.609375,
+      "learning_rate": 1.5635403045403257e-05,
+      "loss": 1.0358,
+      "step": 5483
+    },
+    {
+      "epoch": 0.9403090640204043,
+      "grad_norm": 1.4609375,
+      "learning_rate": 1.5633911096892088e-05,
+      "loss": 0.9374,
+      "step": 5484
+    },
+    {
+      "epoch": 0.9404805281093941,
+      "grad_norm": 1.5859375,
+      "learning_rate": 1.563241896463786e-05,
+      "loss": 1.019,
+      "step": 5485
+    },
+    {
+      "epoch": 0.9406519921983839,
+      "grad_norm": 1.578125,
+      "learning_rate": 1.5630926648689243e-05,
+      "loss": 1.0141,
+      "step": 5486
+    },
+    {
+      "epoch": 0.9408234562873738,
+      "grad_norm": 1.546875,
+      "learning_rate": 1.5629434149094898e-05,
+      "loss": 1.0172,
+      "step": 5487
+    },
+    {
+      "epoch": 0.9409949203763637,
+      "grad_norm": 1.484375,
+      "learning_rate": 1.5627941465903512e-05,
+      "loss": 0.941,
+      "step": 5488
+    },
+    {
+      "epoch": 0.9411663844653535,
+      "grad_norm": 1.625,
+      "learning_rate": 1.562644859916376e-05,
+      "loss": 1.0477,
+      "step": 5489
+    },
+    {
+      "epoch": 0.9413378485543434,
+      "grad_norm": 1.59375,
+      "learning_rate": 1.5624955548924334e-05,
+      "loss": 0.9985,
+      "step": 5490
+    },
+    {
+      "epoch": 0.9415093126433333,
+      "grad_norm": 1.5,
+      "learning_rate": 1.5623462315233923e-05,
+      "loss": 0.9092,
+      "step": 5491
+    },
+    {
+      "epoch": 0.9416807767323231,
+      "grad_norm": 1.5859375,
+      "learning_rate": 1.562196889814123e-05,
+      "loss": 1.0168,
+      "step": 5492
+    },
+    {
+      "epoch": 0.941852240821313,
+      "grad_norm": 1.6796875,
+      "learning_rate": 1.5620475297694963e-05,
+      "loss": 1.0219,
+      "step": 5493
+    },
+    {
+      "epoch": 0.9420237049103029,
+      "grad_norm": 1.796875,
+      "learning_rate": 1.5618981513943833e-05,
+      "loss": 1.0164,
+      "step": 5494
+    },
+    {
+      "epoch": 0.9421951689992927,
+      "grad_norm": 1.640625,
+      "learning_rate": 1.5617487546936555e-05,
+      "loss": 1.0886,
+      "step": 5495
+    },
+    {
+      "epoch": 0.9423666330882826,
+      "grad_norm": 1.5078125,
+      "learning_rate": 1.5615993396721852e-05,
+      "loss": 0.9554,
+      "step": 5496
+    },
+    {
+      "epoch": 0.9425380971772724,
+      "grad_norm": 1.5859375,
+      "learning_rate": 1.5614499063348457e-05,
+      "loss": 1.0361,
+      "step": 5497
+    },
+    {
+      "epoch": 0.9427095612662623,
+      "grad_norm": 1.578125,
+      "learning_rate": 1.561300454686511e-05,
+      "loss": 1.0378,
+      "step": 5498
+    },
+    {
+      "epoch": 0.9428810253552522,
+      "grad_norm": 1.5390625,
+      "learning_rate": 1.561150984732054e-05,
+      "loss": 0.975,
+      "step": 5499
+    },
+    {
+      "epoch": 0.943052489444242,
+      "grad_norm": 1.546875,
+      "learning_rate": 1.561001496476351e-05,
+      "loss": 0.9535,
+      "step": 5500
+    },
+    {
+      "epoch": 0.9432239535332319,
+      "grad_norm": 1.53125,
+      "learning_rate": 1.5608519899242764e-05,
+      "loss": 1.0368,
+      "step": 5501
+    },
+    {
+      "epoch": 0.9433954176222218,
+      "grad_norm": 1.609375,
+      "learning_rate": 1.560702465080707e-05,
+      "loss": 0.9941,
+      "step": 5502
+    },
+    {
+      "epoch": 0.9435668817112116,
+      "grad_norm": 1.5234375,
+      "learning_rate": 1.560552921950518e-05,
+      "loss": 0.8931,
+      "step": 5503
+    },
+    {
+      "epoch": 0.9437383458002014,
+      "grad_norm": 1.5234375,
+      "learning_rate": 1.5604033605385874e-05,
+      "loss": 1.0116,
+      "step": 5504
+    },
+    {
+      "epoch": 0.9439098098891914,
+      "grad_norm": 1.3984375,
+      "learning_rate": 1.5602537808497935e-05,
+      "loss": 0.9978,
+      "step": 5505
+    },
+    {
+      "epoch": 0.9440812739781812,
+      "grad_norm": 1.5546875,
+      "learning_rate": 1.560104182889014e-05,
+      "loss": 0.941,
+      "step": 5506
+    },
+    {
+      "epoch": 0.944252738067171,
+      "grad_norm": 1.6875,
+      "learning_rate": 1.5599545666611272e-05,
+      "loss": 1.0056,
+      "step": 5507
+    },
+    {
+      "epoch": 0.944424202156161,
+      "grad_norm": 1.578125,
+      "learning_rate": 1.5598049321710143e-05,
+      "loss": 0.9915,
+      "step": 5508
+    },
+    {
+      "epoch": 0.9445956662451508,
+      "grad_norm": 1.546875,
+      "learning_rate": 1.559655279423554e-05,
+      "loss": 0.9526,
+      "step": 5509
+    },
+    {
+      "epoch": 0.9447671303341406,
+      "grad_norm": 1.671875,
+      "learning_rate": 1.5595056084236277e-05,
+      "loss": 1.0734,
+      "step": 5510
+    },
+    {
+      "epoch": 0.9449385944231306,
+      "grad_norm": 1.5390625,
+      "learning_rate": 1.5593559191761165e-05,
+      "loss": 0.9538,
+      "step": 5511
+    },
+    {
+      "epoch": 0.9451100585121204,
+      "grad_norm": 1.6015625,
+      "learning_rate": 1.5592062116859026e-05,
+      "loss": 1.0745,
+      "step": 5512
+    },
+    {
+      "epoch": 0.9452815226011102,
+      "grad_norm": 1.546875,
+      "learning_rate": 1.5590564859578682e-05,
+      "loss": 0.9885,
+      "step": 5513
+    },
+    {
+      "epoch": 0.9454529866901001,
+      "grad_norm": 1.5859375,
+      "learning_rate": 1.5589067419968967e-05,
+      "loss": 0.9589,
+      "step": 5514
+    },
+    {
+      "epoch": 0.94562445077909,
+      "grad_norm": 1.5390625,
+      "learning_rate": 1.558756979807872e-05,
+      "loss": 1.0317,
+      "step": 5515
+    },
+    {
+      "epoch": 0.9457959148680798,
+      "grad_norm": 1.4453125,
+      "learning_rate": 1.558607199395678e-05,
+      "loss": 0.959,
+      "step": 5516
+    },
+    {
+      "epoch": 0.9459673789570697,
+      "grad_norm": 1.5703125,
+      "learning_rate": 1.5584574007651993e-05,
+      "loss": 1.029,
+      "step": 5517
+    },
+    {
+      "epoch": 0.9461388430460596,
+      "grad_norm": 1.5546875,
+      "learning_rate": 1.558307583921322e-05,
+      "loss": 0.9865,
+      "step": 5518
+    },
+    {
+      "epoch": 0.9463103071350494,
+      "grad_norm": 1.5625,
+      "learning_rate": 1.558157748868932e-05,
+      "loss": 1.0517,
+      "step": 5519
+    },
+    {
+      "epoch": 0.9464817712240393,
+      "grad_norm": 1.5703125,
+      "learning_rate": 1.5580078956129158e-05,
+      "loss": 1.0287,
+      "step": 5520
+    },
+    {
+      "epoch": 0.9466532353130291,
+      "grad_norm": 1.5625,
+      "learning_rate": 1.5578580241581612e-05,
+      "loss": 0.966,
+      "step": 5521
+    },
+    {
+      "epoch": 0.946824699402019,
+      "grad_norm": 1.703125,
+      "learning_rate": 1.5577081345095556e-05,
+      "loss": 1.0418,
+      "step": 5522
+    },
+    {
+      "epoch": 0.9469961634910089,
+      "grad_norm": 1.6796875,
+      "learning_rate": 1.557558226671987e-05,
+      "loss": 0.9901,
+      "step": 5523
+    },
+    {
+      "epoch": 0.9471676275799987,
+      "grad_norm": 1.578125,
+      "learning_rate": 1.5574083006503462e-05,
+      "loss": 1.0459,
+      "step": 5524
+    },
+    {
+      "epoch": 0.9473390916689886,
+      "grad_norm": 1.46875,
+      "learning_rate": 1.5572583564495212e-05,
+      "loss": 0.9744,
+      "step": 5525
+    },
+    {
+      "epoch": 0.9475105557579785,
+      "grad_norm": 1.59375,
+      "learning_rate": 1.5571083940744025e-05,
+      "loss": 0.9907,
+      "step": 5526
+    },
+    {
+      "epoch": 0.9476820198469683,
+      "grad_norm": 1.4375,
+      "learning_rate": 1.5569584135298813e-05,
+      "loss": 0.8529,
+      "step": 5527
+    },
+    {
+      "epoch": 0.9478534839359581,
+      "grad_norm": 1.5,
+      "learning_rate": 1.5568084148208493e-05,
+      "loss": 0.9824,
+      "step": 5528
+    },
+    {
+      "epoch": 0.9480249480249481,
+      "grad_norm": 1.671875,
+      "learning_rate": 1.5566583979521976e-05,
+      "loss": 0.9589,
+      "step": 5529
+    },
+    {
+      "epoch": 0.9481964121139379,
+      "grad_norm": 1.546875,
+      "learning_rate": 1.5565083629288195e-05,
+      "loss": 0.9606,
+      "step": 5530
+    },
+    {
+      "epoch": 0.9483678762029277,
+      "grad_norm": 1.5703125,
+      "learning_rate": 1.5563583097556086e-05,
+      "loss": 1.0618,
+      "step": 5531
+    },
+    {
+      "epoch": 0.9485393402919177,
+      "grad_norm": 1.5859375,
+      "learning_rate": 1.556208238437458e-05,
+      "loss": 0.965,
+      "step": 5532
+    },
+    {
+      "epoch": 0.9487108043809075,
+      "grad_norm": 1.546875,
+      "learning_rate": 1.5560581489792617e-05,
+      "loss": 0.992,
+      "step": 5533
+    },
+    {
+      "epoch": 0.9488822684698973,
+      "grad_norm": 1.4453125,
+      "learning_rate": 1.5559080413859153e-05,
+      "loss": 0.9641,
+      "step": 5534
+    },
+    {
+      "epoch": 0.9490537325588873,
+      "grad_norm": 1.578125,
+      "learning_rate": 1.555757915662315e-05,
+      "loss": 0.9974,
+      "step": 5535
+    },
+    {
+      "epoch": 0.9492251966478771,
+      "grad_norm": 1.5078125,
+      "learning_rate": 1.5556077718133556e-05,
+      "loss": 0.9771,
+      "step": 5536
+    },
+    {
+      "epoch": 0.9493966607368669,
+      "grad_norm": 1.5703125,
+      "learning_rate": 1.555457609843935e-05,
+      "loss": 0.9597,
+      "step": 5537
+    },
+    {
+      "epoch": 0.9495681248258567,
+      "grad_norm": 1.6484375,
+      "learning_rate": 1.55530742975895e-05,
+      "loss": 1.0468,
+      "step": 5538
+    },
+    {
+      "epoch": 0.9497395889148467,
+      "grad_norm": 1.5859375,
+      "learning_rate": 1.5551572315632983e-05,
+      "loss": 1.0865,
+      "step": 5539
+    },
+    {
+      "epoch": 0.9499110530038365,
+      "grad_norm": 1.5703125,
+      "learning_rate": 1.555007015261879e-05,
+      "loss": 1.0402,
+      "step": 5540
+    },
+    {
+      "epoch": 0.9500825170928263,
+      "grad_norm": 1.5703125,
+      "learning_rate": 1.5548567808595905e-05,
+      "loss": 0.9962,
+      "step": 5541
+    },
+    {
+      "epoch": 0.9502539811818163,
+      "grad_norm": 1.515625,
+      "learning_rate": 1.5547065283613332e-05,
+      "loss": 1.0879,
+      "step": 5542
+    },
+    {
+      "epoch": 0.9504254452708061,
+      "grad_norm": 1.625,
+      "learning_rate": 1.5545562577720077e-05,
+      "loss": 1.0379,
+      "step": 5543
+    },
+    {
+      "epoch": 0.9505969093597959,
+      "grad_norm": 1.53125,
+      "learning_rate": 1.554405969096514e-05,
+      "loss": 1.0111,
+      "step": 5544
+    },
+    {
+      "epoch": 0.9507683734487858,
+      "grad_norm": 1.53125,
+      "learning_rate": 1.5542556623397542e-05,
+      "loss": 0.9819,
+      "step": 5545
+    },
+    {
+      "epoch": 0.9509398375377757,
+      "grad_norm": 1.5078125,
+      "learning_rate": 1.5541053375066298e-05,
+      "loss": 0.9977,
+      "step": 5546
+    },
+    {
+      "epoch": 0.9511113016267655,
+      "grad_norm": 1.5625,
+      "learning_rate": 1.553954994602044e-05,
+      "loss": 1.124,
+      "step": 5547
+    },
+    {
+      "epoch": 0.9512827657157554,
+      "grad_norm": 1.4765625,
+      "learning_rate": 1.5538046336308996e-05,
+      "loss": 0.9165,
+      "step": 5548
+    },
+    {
+      "epoch": 0.9514542298047453,
+      "grad_norm": 1.5625,
+      "learning_rate": 1.553654254598101e-05,
+      "loss": 0.9962,
+      "step": 5549
+    },
+    {
+      "epoch": 0.9516256938937351,
+      "grad_norm": 1.609375,
+      "learning_rate": 1.553503857508552e-05,
+      "loss": 1.0265,
+      "step": 5550
+    },
+    {
+      "epoch": 0.951797157982725,
+      "grad_norm": 1.625,
+      "learning_rate": 1.5533534423671578e-05,
+      "loss": 1.0501,
+      "step": 5551
+    },
+    {
+      "epoch": 0.9519686220717148,
+      "grad_norm": 1.484375,
+      "learning_rate": 1.5532030091788243e-05,
+      "loss": 0.8984,
+      "step": 5552
+    },
+    {
+      "epoch": 0.9521400861607047,
+      "grad_norm": 1.5703125,
+      "learning_rate": 1.5530525579484575e-05,
+      "loss": 1.0077,
+      "step": 5553
+    },
+    {
+      "epoch": 0.9523115502496946,
+      "grad_norm": 1.578125,
+      "learning_rate": 1.552902088680964e-05,
+      "loss": 1.0534,
+      "step": 5554
+    },
+    {
+      "epoch": 0.9524830143386844,
+      "grad_norm": 1.5390625,
+      "learning_rate": 1.552751601381252e-05,
+      "loss": 0.9774,
+      "step": 5555
+    },
+    {
+      "epoch": 0.9526544784276743,
+      "grad_norm": 1.5859375,
+      "learning_rate": 1.5526010960542278e-05,
+      "loss": 1.0769,
+      "step": 5556
+    },
+    {
+      "epoch": 0.9528259425166642,
+      "grad_norm": 1.6328125,
+      "learning_rate": 1.5524505727048017e-05,
+      "loss": 1.0191,
+      "step": 5557
+    },
+    {
+      "epoch": 0.952997406605654,
+      "grad_norm": 1.5078125,
+      "learning_rate": 1.5523000313378816e-05,
+      "loss": 0.9485,
+      "step": 5558
+    },
+    {
+      "epoch": 0.9531688706946438,
+      "grad_norm": 1.5546875,
+      "learning_rate": 1.552149471958378e-05,
+      "loss": 1.0232,
+      "step": 5559
+    },
+    {
+      "epoch": 0.9533403347836338,
+      "grad_norm": 1.6015625,
+      "learning_rate": 1.551998894571201e-05,
+      "loss": 1.0259,
+      "step": 5560
+    },
+    {
+      "epoch": 0.9535117988726236,
+      "grad_norm": 1.5859375,
+      "learning_rate": 1.5518482991812614e-05,
+      "loss": 0.9999,
+      "step": 5561
+    },
+    {
+      "epoch": 0.9536832629616134,
+      "grad_norm": 1.5,
+      "learning_rate": 1.5516976857934703e-05,
+      "loss": 1.0025,
+      "step": 5562
+    },
+    {
+      "epoch": 0.9538547270506034,
+      "grad_norm": 1.5390625,
+      "learning_rate": 1.5515470544127405e-05,
+      "loss": 1.0224,
+      "step": 5563
+    },
+    {
+      "epoch": 0.9540261911395932,
+      "grad_norm": 1.5078125,
+      "learning_rate": 1.5513964050439842e-05,
+      "loss": 1.0007,
+      "step": 5564
+    },
+    {
+      "epoch": 0.954197655228583,
+      "grad_norm": 1.5234375,
+      "learning_rate": 1.5512457376921147e-05,
+      "loss": 0.9864,
+      "step": 5565
+    },
+    {
+      "epoch": 0.954369119317573,
+      "grad_norm": 1.5234375,
+      "learning_rate": 1.5510950523620465e-05,
+      "loss": 1.0098,
+      "step": 5566
+    },
+    {
+      "epoch": 0.9545405834065628,
+      "grad_norm": 1.6328125,
+      "learning_rate": 1.550944349058693e-05,
+      "loss": 0.9862,
+      "step": 5567
+    },
+    {
+      "epoch": 0.9547120474955526,
+      "grad_norm": 1.5,
+      "learning_rate": 1.5507936277869693e-05,
+      "loss": 1.0668,
+      "step": 5568
+    },
+    {
+      "epoch": 0.9548835115845425,
+      "grad_norm": 1.5625,
+      "learning_rate": 1.5506428885517917e-05,
+      "loss": 0.9649,
+      "step": 5569
+    },
+    {
+      "epoch": 0.9550549756735324,
+      "grad_norm": 1.4921875,
+      "learning_rate": 1.5504921313580757e-05,
+      "loss": 0.9597,
+      "step": 5570
+    },
+    {
+      "epoch": 0.9552264397625222,
+      "grad_norm": 1.5234375,
+      "learning_rate": 1.5503413562107385e-05,
+      "loss": 0.9652,
+      "step": 5571
+    },
+    {
+      "epoch": 0.9553979038515121,
+      "grad_norm": 1.5,
+      "learning_rate": 1.5501905631146975e-05,
+      "loss": 0.9575,
+      "step": 5572
+    },
+    {
+      "epoch": 0.955569367940502,
+      "grad_norm": 1.515625,
+      "learning_rate": 1.55003975207487e-05,
+      "loss": 0.9843,
+      "step": 5573
+    },
+    {
+      "epoch": 0.9557408320294918,
+      "grad_norm": 1.6796875,
+      "learning_rate": 1.5498889230961753e-05,
+      "loss": 1.0935,
+      "step": 5574
+    },
+    {
+      "epoch": 0.9559122961184817,
+      "grad_norm": 1.5546875,
+      "learning_rate": 1.5497380761835318e-05,
+      "loss": 0.9589,
+      "step": 5575
+    },
+    {
+      "epoch": 0.9560837602074715,
+      "grad_norm": 1.609375,
+      "learning_rate": 1.5495872113418594e-05,
+      "loss": 0.9689,
+      "step": 5576
+    },
+    {
+      "epoch": 0.9562552242964614,
+      "grad_norm": 1.625,
+      "learning_rate": 1.549436328576079e-05,
+      "loss": 0.9995,
+      "step": 5577
+    },
+    {
+      "epoch": 0.9564266883854513,
+      "grad_norm": 1.53125,
+      "learning_rate": 1.5492854278911103e-05,
+      "loss": 0.9451,
+      "step": 5578
+    },
+    {
+      "epoch": 0.9565981524744411,
+      "grad_norm": 1.5546875,
+      "learning_rate": 1.549134509291876e-05,
+      "loss": 0.974,
+      "step": 5579
+    },
+    {
+      "epoch": 0.956769616563431,
+      "grad_norm": 1.578125,
+      "learning_rate": 1.5489835727832973e-05,
+      "loss": 0.9706,
+      "step": 5580
+    },
+    {
+      "epoch": 0.9569410806524209,
+      "grad_norm": 1.4921875,
+      "learning_rate": 1.5488326183702967e-05,
+      "loss": 0.9371,
+      "step": 5581
+    },
+    {
+      "epoch": 0.9571125447414107,
+      "grad_norm": 1.6875,
+      "learning_rate": 1.548681646057798e-05,
+      "loss": 1.0169,
+      "step": 5582
+    },
+    {
+      "epoch": 0.9572840088304005,
+      "grad_norm": 1.6015625,
+      "learning_rate": 1.548530655850725e-05,
+      "loss": 1.0346,
+      "step": 5583
+    },
+    {
+      "epoch": 0.9574554729193905,
+      "grad_norm": 1.578125,
+      "learning_rate": 1.548379647754001e-05,
+      "loss": 0.9309,
+      "step": 5584
+    },
+    {
+      "epoch": 0.9576269370083803,
+      "grad_norm": 1.59375,
+      "learning_rate": 1.548228621772552e-05,
+      "loss": 0.9644,
+      "step": 5585
+    },
+    {
+      "epoch": 0.9577984010973701,
+      "grad_norm": 1.546875,
+      "learning_rate": 1.5480775779113032e-05,
+      "loss": 0.9528,
+      "step": 5586
+    },
+    {
+      "epoch": 0.9579698651863601,
+      "grad_norm": 1.5,
+      "learning_rate": 1.5479265161751807e-05,
+      "loss": 0.9366,
+      "step": 5587
+    },
+    {
+      "epoch": 0.9581413292753499,
+      "grad_norm": 1.578125,
+      "learning_rate": 1.5477754365691113e-05,
+      "loss": 0.9291,
+      "step": 5588
+    },
+    {
+      "epoch": 0.9583127933643397,
+      "grad_norm": 1.53125,
+      "learning_rate": 1.5476243390980224e-05,
+      "loss": 1.0644,
+      "step": 5589
+    },
+    {
+      "epoch": 0.9584842574533297,
+      "grad_norm": 1.515625,
+      "learning_rate": 1.5474732237668413e-05,
+      "loss": 0.9302,
+      "step": 5590
+    },
+    {
+      "epoch": 0.9586557215423195,
+      "grad_norm": 1.5546875,
+      "learning_rate": 1.547322090580497e-05,
+      "loss": 0.956,
+      "step": 5591
+    },
+    {
+      "epoch": 0.9588271856313093,
+      "grad_norm": 1.59375,
+      "learning_rate": 1.5471709395439182e-05,
+      "loss": 1.0559,
+      "step": 5592
+    },
+    {
+      "epoch": 0.9589986497202992,
+      "grad_norm": 1.5234375,
+      "learning_rate": 1.5470197706620345e-05,
+      "loss": 0.9851,
+      "step": 5593
+    },
+    {
+      "epoch": 0.9591701138092891,
+      "grad_norm": 1.4296875,
+      "learning_rate": 1.5468685839397764e-05,
+      "loss": 0.9395,
+      "step": 5594
+    },
+    {
+      "epoch": 0.9593415778982789,
+      "grad_norm": 1.640625,
+      "learning_rate": 1.546717379382074e-05,
+      "loss": 1.025,
+      "step": 5595
+    },
+    {
+      "epoch": 0.9595130419872688,
+      "grad_norm": 1.59375,
+      "learning_rate": 1.5465661569938597e-05,
+      "loss": 1.0304,
+      "step": 5596
+    },
+    {
+      "epoch": 0.9596845060762587,
+      "grad_norm": 1.5546875,
+      "learning_rate": 1.5464149167800643e-05,
+      "loss": 0.9483,
+      "step": 5597
+    },
+    {
+      "epoch": 0.9598559701652485,
+      "grad_norm": 1.546875,
+      "learning_rate": 1.5462636587456216e-05,
+      "loss": 0.9549,
+      "step": 5598
+    },
+    {
+      "epoch": 0.9600274342542384,
+      "grad_norm": 1.546875,
+      "learning_rate": 1.5461123828954635e-05,
+      "loss": 0.9554,
+      "step": 5599
+    },
+    {
+      "epoch": 0.9601988983432282,
+      "grad_norm": 1.65625,
+      "learning_rate": 1.545961089234524e-05,
+      "loss": 1.1681,
+      "step": 5600
+    },
+    {
+      "epoch": 0.9601988983432282,
+      "eval_loss": 0.8496836423873901,
+      "eval_runtime": 836.7427,
+      "eval_samples_per_second": 2.987,
+      "eval_steps_per_second": 2.987,
+      "step": 5600
+    },
+    {
+      "epoch": 0.9603703624322181,
+      "grad_norm": 1.5078125,
+      "learning_rate": 1.5458097777677377e-05,
+      "loss": 1.0117,
+      "step": 5601
+    },
+    {
+      "epoch": 0.960541826521208,
+      "grad_norm": 1.6015625,
+      "learning_rate": 1.545658448500039e-05,
+      "loss": 1.0554,
+      "step": 5602
+    },
+    {
+      "epoch": 0.9607132906101978,
+      "grad_norm": 1.5859375,
+      "learning_rate": 1.5455071014363637e-05,
+      "loss": 0.9915,
+      "step": 5603
+    },
+    {
+      "epoch": 0.9608847546991877,
+      "grad_norm": 1.6328125,
+      "learning_rate": 1.5453557365816477e-05,
+      "loss": 1.0956,
+      "step": 5604
+    },
+    {
+      "epoch": 0.9610562187881776,
+      "grad_norm": 1.53125,
+      "learning_rate": 1.545204353940827e-05,
+      "loss": 0.9851,
+      "step": 5605
+    },
+    {
+      "epoch": 0.9612276828771674,
+      "grad_norm": 1.5234375,
+      "learning_rate": 1.54505295351884e-05,
+      "loss": 0.9669,
+      "step": 5606
+    },
+    {
+      "epoch": 0.9613991469661572,
+      "grad_norm": 1.515625,
+      "learning_rate": 1.5449015353206232e-05,
+      "loss": 0.9688,
+      "step": 5607
+    },
+    {
+      "epoch": 0.9615706110551472,
+      "grad_norm": 1.5625,
+      "learning_rate": 1.5447500993511155e-05,
+      "loss": 0.9418,
+      "step": 5608
+    },
+    {
+      "epoch": 0.961742075144137,
+      "grad_norm": 1.453125,
+      "learning_rate": 1.5445986456152557e-05,
+      "loss": 1.0225,
+      "step": 5609
+    },
+    {
+      "epoch": 0.9619135392331268,
+      "grad_norm": 1.5390625,
+      "learning_rate": 1.5444471741179838e-05,
+      "loss": 1.0441,
+      "step": 5610
+    },
+    {
+      "epoch": 0.9620850033221168,
+      "grad_norm": 1.6484375,
+      "learning_rate": 1.544295684864239e-05,
+      "loss": 1.0412,
+      "step": 5611
+    },
+    {
+      "epoch": 0.9622564674111066,
+      "grad_norm": 1.5859375,
+      "learning_rate": 1.5441441778589622e-05,
+      "loss": 1.0482,
+      "step": 5612
+    },
+    {
+      "epoch": 0.9624279315000964,
+      "grad_norm": 1.53125,
+      "learning_rate": 1.5439926531070944e-05,
+      "loss": 1.0165,
+      "step": 5613
+    },
+    {
+      "epoch": 0.9625993955890864,
+      "grad_norm": 1.5078125,
+      "learning_rate": 1.5438411106135784e-05,
+      "loss": 0.9765,
+      "step": 5614
+    },
+    {
+      "epoch": 0.9627708596780762,
+      "grad_norm": 1.6015625,
+      "learning_rate": 1.5436895503833555e-05,
+      "loss": 1.0007,
+      "step": 5615
+    },
+    {
+      "epoch": 0.962942323767066,
+      "grad_norm": 1.6640625,
+      "learning_rate": 1.543537972421369e-05,
+      "loss": 0.9622,
+      "step": 5616
+    },
+    {
+      "epoch": 0.963113787856056,
+      "grad_norm": 1.59375,
+      "learning_rate": 1.5433863767325626e-05,
+      "loss": 1.1012,
+      "step": 5617
+    },
+    {
+      "epoch": 0.9632852519450458,
+      "grad_norm": 1.609375,
+      "learning_rate": 1.5432347633218802e-05,
+      "loss": 1.0813,
+      "step": 5618
+    },
+    {
+      "epoch": 0.9634567160340356,
+      "grad_norm": 1.796875,
+      "learning_rate": 1.5430831321942664e-05,
+      "loss": 1.0686,
+      "step": 5619
+    },
+    {
+      "epoch": 0.9636281801230255,
+      "grad_norm": 1.625,
+      "learning_rate": 1.5429314833546665e-05,
+      "loss": 1.0027,
+      "step": 5620
+    },
+    {
+      "epoch": 0.9637996442120154,
+      "grad_norm": 1.6328125,
+      "learning_rate": 1.5427798168080267e-05,
+      "loss": 1.0444,
+      "step": 5621
+    },
+    {
+      "epoch": 0.9639711083010052,
+      "grad_norm": 1.5078125,
+      "learning_rate": 1.5426281325592932e-05,
+      "loss": 0.9338,
+      "step": 5622
+    },
+    {
+      "epoch": 0.9641425723899951,
+      "grad_norm": 1.640625,
+      "learning_rate": 1.542476430613413e-05,
+      "loss": 1.0205,
+      "step": 5623
+    },
+    {
+      "epoch": 0.964314036478985,
+      "grad_norm": 1.515625,
+      "learning_rate": 1.5423247109753332e-05,
+      "loss": 0.9084,
+      "step": 5624
+    },
+    {
+      "epoch": 0.9644855005679748,
+      "grad_norm": 1.609375,
+      "learning_rate": 1.5421729736500024e-05,
+      "loss": 1.0294,
+      "step": 5625
+    },
+    {
+      "epoch": 0.9646569646569647,
+      "grad_norm": 1.53125,
+      "learning_rate": 1.5420212186423696e-05,
+      "loss": 0.9791,
+      "step": 5626
+    },
+    {
+      "epoch": 0.9648284287459545,
+      "grad_norm": 1.59375,
+      "learning_rate": 1.541869445957384e-05,
+      "loss": 1.0413,
+      "step": 5627
+    },
+    {
+      "epoch": 0.9649998928349444,
+      "grad_norm": 1.5859375,
+      "learning_rate": 1.5417176555999948e-05,
+      "loss": 0.9642,
+      "step": 5628
+    },
+    {
+      "epoch": 0.9651713569239343,
+      "grad_norm": 1.6015625,
+      "learning_rate": 1.541565847575153e-05,
+      "loss": 0.9477,
+      "step": 5629
+    },
+    {
+      "epoch": 0.9653428210129241,
+      "grad_norm": 1.5078125,
+      "learning_rate": 1.5414140218878096e-05,
+      "loss": 0.9937,
+      "step": 5630
+    },
+    {
+      "epoch": 0.965514285101914,
+      "grad_norm": 1.7578125,
+      "learning_rate": 1.5412621785429162e-05,
+      "loss": 1.0819,
+      "step": 5631
+    },
+    {
+      "epoch": 0.9656857491909038,
+      "grad_norm": 1.515625,
+      "learning_rate": 1.541110317545425e-05,
+      "loss": 0.9953,
+      "step": 5632
+    },
+    {
+      "epoch": 0.9658572132798937,
+      "grad_norm": 1.5546875,
+      "learning_rate": 1.5409584389002885e-05,
+      "loss": 0.9715,
+      "step": 5633
+    },
+    {
+      "epoch": 0.9660286773688835,
+      "grad_norm": 1.484375,
+      "learning_rate": 1.5408065426124607e-05,
+      "loss": 0.916,
+      "step": 5634
+    },
+    {
+      "epoch": 0.9662001414578734,
+      "grad_norm": 1.578125,
+      "learning_rate": 1.5406546286868946e-05,
+      "loss": 0.9839,
+      "step": 5635
+    },
+    {
+      "epoch": 0.9663716055468633,
+      "grad_norm": 1.484375,
+      "learning_rate": 1.5405026971285454e-05,
+      "loss": 0.9272,
+      "step": 5636
+    },
+    {
+      "epoch": 0.9665430696358531,
+      "grad_norm": 1.546875,
+      "learning_rate": 1.540350747942368e-05,
+      "loss": 1.017,
+      "step": 5637
+    },
+    {
+      "epoch": 0.966714533724843,
+      "grad_norm": 1.6171875,
+      "learning_rate": 1.540198781133318e-05,
+      "loss": 1.0335,
+      "step": 5638
+    },
+    {
+      "epoch": 0.9668859978138329,
+      "grad_norm": 1.5,
+      "learning_rate": 1.5400467967063512e-05,
+      "loss": 1.0387,
+      "step": 5639
+    },
+    {
+      "epoch": 0.9670574619028227,
+      "grad_norm": 1.6015625,
+      "learning_rate": 1.5398947946664247e-05,
+      "loss": 1.0063,
+      "step": 5640
+    },
+    {
+      "epoch": 0.9672289259918125,
+      "grad_norm": 1.5390625,
+      "learning_rate": 1.5397427750184962e-05,
+      "loss": 0.9206,
+      "step": 5641
+    },
+    {
+      "epoch": 0.9674003900808025,
+      "grad_norm": 1.578125,
+      "learning_rate": 1.539590737767523e-05,
+      "loss": 1.0224,
+      "step": 5642
+    },
+    {
+      "epoch": 0.9675718541697923,
+      "grad_norm": 1.5625,
+      "learning_rate": 1.5394386829184643e-05,
+      "loss": 1.0046,
+      "step": 5643
+    },
+    {
+      "epoch": 0.9677433182587821,
+      "grad_norm": 1.578125,
+      "learning_rate": 1.5392866104762783e-05,
+      "loss": 1.0165,
+      "step": 5644
+    },
+    {
+      "epoch": 0.9679147823477721,
+      "grad_norm": 1.546875,
+      "learning_rate": 1.5391345204459255e-05,
+      "loss": 1.051,
+      "step": 5645
+    },
+    {
+      "epoch": 0.9680862464367619,
+      "grad_norm": 1.578125,
+      "learning_rate": 1.538982412832366e-05,
+      "loss": 0.9658,
+      "step": 5646
+    },
+    {
+      "epoch": 0.9682577105257517,
+      "grad_norm": 1.609375,
+      "learning_rate": 1.5388302876405602e-05,
+      "loss": 0.9937,
+      "step": 5647
+    },
+    {
+      "epoch": 0.9684291746147417,
+      "grad_norm": 1.65625,
+      "learning_rate": 1.5386781448754696e-05,
+      "loss": 0.9847,
+      "step": 5648
+    },
+    {
+      "epoch": 0.9686006387037315,
+      "grad_norm": 1.4765625,
+      "learning_rate": 1.538525984542056e-05,
+      "loss": 0.9305,
+      "step": 5649
+    },
+    {
+      "epoch": 0.9687721027927213,
+      "grad_norm": 1.4609375,
+      "learning_rate": 1.5383738066452825e-05,
+      "loss": 0.9352,
+      "step": 5650
+    },
+    {
+      "epoch": 0.9689435668817112,
+      "grad_norm": 1.7265625,
+      "learning_rate": 1.5382216111901116e-05,
+      "loss": 0.9158,
+      "step": 5651
+    },
+    {
+      "epoch": 0.9691150309707011,
+      "grad_norm": 1.6328125,
+      "learning_rate": 1.5380693981815077e-05,
+      "loss": 0.9998,
+      "step": 5652
+    },
+    {
+      "epoch": 0.9692864950596909,
+      "grad_norm": 1.6015625,
+      "learning_rate": 1.5379171676244343e-05,
+      "loss": 1.1242,
+      "step": 5653
+    },
+    {
+      "epoch": 0.9694579591486808,
+      "grad_norm": 1.4765625,
+      "learning_rate": 1.537764919523856e-05,
+      "loss": 0.9611,
+      "step": 5654
+    },
+    {
+      "epoch": 0.9696294232376707,
+      "grad_norm": 1.609375,
+      "learning_rate": 1.537612653884739e-05,
+      "loss": 0.9709,
+      "step": 5655
+    },
+    {
+      "epoch": 0.9698008873266605,
+      "grad_norm": 1.609375,
+      "learning_rate": 1.537460370712049e-05,
+      "loss": 1.0086,
+      "step": 5656
+    },
+    {
+      "epoch": 0.9699723514156504,
+      "grad_norm": 1.6015625,
+      "learning_rate": 1.5373080700107522e-05,
+      "loss": 0.9748,
+      "step": 5657
+    },
+    {
+      "epoch": 0.9701438155046402,
+      "grad_norm": 1.625,
+      "learning_rate": 1.5371557517858162e-05,
+      "loss": 0.9927,
+      "step": 5658
+    },
+    {
+      "epoch": 0.9703152795936301,
+      "grad_norm": 1.5078125,
+      "learning_rate": 1.5370034160422084e-05,
+      "loss": 1.0601,
+      "step": 5659
+    },
+    {
+      "epoch": 0.97048674368262,
+      "grad_norm": 1.6875,
+      "learning_rate": 1.5368510627848963e-05,
+      "loss": 1.0856,
+      "step": 5660
+    },
+    {
+      "epoch": 0.9706582077716098,
+      "grad_norm": 1.59375,
+      "learning_rate": 1.53669869201885e-05,
+      "loss": 1.0443,
+      "step": 5661
+    },
+    {
+      "epoch": 0.9708296718605997,
+      "grad_norm": 1.5,
+      "learning_rate": 1.5365463037490386e-05,
+      "loss": 0.9502,
+      "step": 5662
+    },
+    {
+      "epoch": 0.9710011359495896,
+      "grad_norm": 1.6171875,
+      "learning_rate": 1.5363938979804312e-05,
+      "loss": 0.9061,
+      "step": 5663
+    },
+    {
+      "epoch": 0.9711726000385794,
+      "grad_norm": 1.6953125,
+      "learning_rate": 1.5362414747179996e-05,
+      "loss": 0.9255,
+      "step": 5664
+    },
+    {
+      "epoch": 0.9713440641275692,
+      "grad_norm": 1.59375,
+      "learning_rate": 1.536089033966714e-05,
+      "loss": 1.0111,
+      "step": 5665
+    },
+    {
+      "epoch": 0.9715155282165592,
+      "grad_norm": 1.59375,
+      "learning_rate": 1.535936575731546e-05,
+      "loss": 0.9452,
+      "step": 5666
+    },
+    {
+      "epoch": 0.971686992305549,
+      "grad_norm": 1.4609375,
+      "learning_rate": 1.535784100017468e-05,
+      "loss": 0.9667,
+      "step": 5667
+    },
+    {
+      "epoch": 0.9718584563945388,
+      "grad_norm": 1.5859375,
+      "learning_rate": 1.5356316068294533e-05,
+      "loss": 1.0044,
+      "step": 5668
+    },
+    {
+      "epoch": 0.9720299204835288,
+      "grad_norm": 1.6171875,
+      "learning_rate": 1.535479096172475e-05,
+      "loss": 1.051,
+      "step": 5669
+    },
+    {
+      "epoch": 0.9722013845725186,
+      "grad_norm": 1.59375,
+      "learning_rate": 1.5353265680515067e-05,
+      "loss": 1.0812,
+      "step": 5670
+    },
+    {
+      "epoch": 0.9723728486615084,
+      "grad_norm": 1.578125,
+      "learning_rate": 1.535174022471523e-05,
+      "loss": 0.9584,
+      "step": 5671
+    },
+    {
+      "epoch": 0.9725443127504984,
+      "grad_norm": 1.6328125,
+      "learning_rate": 1.5350214594374995e-05,
+      "loss": 0.9947,
+      "step": 5672
+    },
+    {
+      "epoch": 0.9727157768394882,
+      "grad_norm": 1.5859375,
+      "learning_rate": 1.5348688789544114e-05,
+      "loss": 0.9427,
+      "step": 5673
+    },
+    {
+      "epoch": 0.972887240928478,
+      "grad_norm": 1.53125,
+      "learning_rate": 1.5347162810272348e-05,
+      "loss": 0.974,
+      "step": 5674
+    },
+    {
+      "epoch": 0.9730587050174679,
+      "grad_norm": 1.4921875,
+      "learning_rate": 1.534563665660947e-05,
+      "loss": 1.0238,
+      "step": 5675
+    },
+    {
+      "epoch": 0.9732301691064578,
+      "grad_norm": 1.4921875,
+      "learning_rate": 1.5344110328605248e-05,
+      "loss": 0.9584,
+      "step": 5676
+    },
+    {
+      "epoch": 0.9734016331954476,
+      "grad_norm": 1.578125,
+      "learning_rate": 1.5342583826309464e-05,
+      "loss": 1.0112,
+      "step": 5677
+    },
+    {
+      "epoch": 0.9735730972844375,
+      "grad_norm": 1.65625,
+      "learning_rate": 1.534105714977191e-05,
+      "loss": 1.0449,
+      "step": 5678
+    },
+    {
+      "epoch": 0.9737445613734274,
+      "grad_norm": 1.46875,
+      "learning_rate": 1.533953029904236e-05,
+      "loss": 0.9855,
+      "step": 5679
+    },
+    {
+      "epoch": 0.9739160254624172,
+      "grad_norm": 1.53125,
+      "learning_rate": 1.5338003274170626e-05,
+      "loss": 0.9667,
+      "step": 5680
+    },
+    {
+      "epoch": 0.9740874895514071,
+      "grad_norm": 1.59375,
+      "learning_rate": 1.5336476075206506e-05,
+      "loss": 1.02,
+      "step": 5681
+    },
+    {
+      "epoch": 0.9742589536403969,
+      "grad_norm": 1.46875,
+      "learning_rate": 1.5334948702199803e-05,
+      "loss": 1.0029,
+      "step": 5682
+    },
+    {
+      "epoch": 0.9744304177293868,
+      "grad_norm": 1.5546875,
+      "learning_rate": 1.533342115520033e-05,
+      "loss": 0.9316,
+      "step": 5683
+    },
+    {
+      "epoch": 0.9746018818183767,
+      "grad_norm": 1.5625,
+      "learning_rate": 1.533189343425791e-05,
+      "loss": 0.8841,
+      "step": 5684
+    },
+    {
+      "epoch": 0.9747733459073665,
+      "grad_norm": 1.546875,
+      "learning_rate": 1.533036553942237e-05,
+      "loss": 0.9879,
+      "step": 5685
+    },
+    {
+      "epoch": 0.9749448099963564,
+      "grad_norm": 1.5390625,
+      "learning_rate": 1.532883747074354e-05,
+      "loss": 1.0564,
+      "step": 5686
+    },
+    {
+      "epoch": 0.9751162740853463,
+      "grad_norm": 1.5546875,
+      "learning_rate": 1.532730922827125e-05,
+      "loss": 0.9616,
+      "step": 5687
+    },
+    {
+      "epoch": 0.9752877381743361,
+      "grad_norm": 1.5625,
+      "learning_rate": 1.5325780812055345e-05,
+      "loss": 0.9662,
+      "step": 5688
+    },
+    {
+      "epoch": 0.9754592022633259,
+      "grad_norm": 1.6953125,
+      "learning_rate": 1.5324252222145673e-05,
+      "loss": 0.9497,
+      "step": 5689
+    },
+    {
+      "epoch": 0.9756306663523159,
+      "grad_norm": 1.5234375,
+      "learning_rate": 1.532272345859209e-05,
+      "loss": 1.0239,
+      "step": 5690
+    },
+    {
+      "epoch": 0.9758021304413057,
+      "grad_norm": 1.578125,
+      "learning_rate": 1.5321194521444445e-05,
+      "loss": 0.9976,
+      "step": 5691
+    },
+    {
+      "epoch": 0.9759735945302955,
+      "grad_norm": 1.5546875,
+      "learning_rate": 1.5319665410752615e-05,
+      "loss": 1.005,
+      "step": 5692
+    },
+    {
+      "epoch": 0.9761450586192855,
+      "grad_norm": 1.515625,
+      "learning_rate": 1.531813612656646e-05,
+      "loss": 0.9203,
+      "step": 5693
+    },
+    {
+      "epoch": 0.9763165227082753,
+      "grad_norm": 1.6796875,
+      "learning_rate": 1.5316606668935862e-05,
+      "loss": 1.037,
+      "step": 5694
+    },
+    {
+      "epoch": 0.9764879867972651,
+      "grad_norm": 1.4921875,
+      "learning_rate": 1.53150770379107e-05,
+      "loss": 0.929,
+      "step": 5695
+    },
+    {
+      "epoch": 0.976659450886255,
+      "grad_norm": 1.5703125,
+      "learning_rate": 1.5313547233540858e-05,
+      "loss": 1.0949,
+      "step": 5696
+    },
+    {
+      "epoch": 0.9768309149752449,
+      "grad_norm": 1.5625,
+      "learning_rate": 1.5312017255876234e-05,
+      "loss": 0.9868,
+      "step": 5697
+    },
+    {
+      "epoch": 0.9770023790642347,
+      "grad_norm": 1.5078125,
+      "learning_rate": 1.5310487104966725e-05,
+      "loss": 0.933,
+      "step": 5698
+    },
+    {
+      "epoch": 0.9771738431532246,
+      "grad_norm": 1.5625,
+      "learning_rate": 1.5308956780862234e-05,
+      "loss": 0.9512,
+      "step": 5699
+    },
+    {
+      "epoch": 0.9773453072422145,
+      "grad_norm": 1.5234375,
+      "learning_rate": 1.5307426283612668e-05,
+      "loss": 0.9092,
+      "step": 5700
+    },
+    {
+      "epoch": 0.9775167713312043,
+      "grad_norm": 1.5703125,
+      "learning_rate": 1.530589561326795e-05,
+      "loss": 0.9629,
+      "step": 5701
+    },
+    {
+      "epoch": 0.9776882354201942,
+      "grad_norm": 1.6796875,
+      "learning_rate": 1.530436476987799e-05,
+      "loss": 1.0289,
+      "step": 5702
+    },
+    {
+      "epoch": 0.977859699509184,
+      "grad_norm": 1.5,
+      "learning_rate": 1.5302833753492726e-05,
+      "loss": 0.8838,
+      "step": 5703
+    },
+    {
+      "epoch": 0.9780311635981739,
+      "grad_norm": 1.5390625,
+      "learning_rate": 1.530130256416208e-05,
+      "loss": 0.9919,
+      "step": 5704
+    },
+    {
+      "epoch": 0.9782026276871638,
+      "grad_norm": 1.5390625,
+      "learning_rate": 1.5299771201935998e-05,
+      "loss": 0.9795,
+      "step": 5705
+    },
+    {
+      "epoch": 0.9783740917761536,
+      "grad_norm": 1.6328125,
+      "learning_rate": 1.5298239666864417e-05,
+      "loss": 0.9597,
+      "step": 5706
+    },
+    {
+      "epoch": 0.9785455558651435,
+      "grad_norm": 1.5859375,
+      "learning_rate": 1.529670795899729e-05,
+      "loss": 0.9669,
+      "step": 5707
+    },
+    {
+      "epoch": 0.9787170199541334,
+      "grad_norm": 1.546875,
+      "learning_rate": 1.529517607838457e-05,
+      "loss": 1.0354,
+      "step": 5708
+    },
+    {
+      "epoch": 0.9788884840431232,
+      "grad_norm": 1.5625,
+      "learning_rate": 1.5293644025076223e-05,
+      "loss": 1.0429,
+      "step": 5709
+    },
+    {
+      "epoch": 0.9790599481321131,
+      "grad_norm": 1.5625,
+      "learning_rate": 1.5292111799122208e-05,
+      "loss": 0.98,
+      "step": 5710
+    },
+    {
+      "epoch": 0.979231412221103,
+      "grad_norm": 1.65625,
+      "learning_rate": 1.5290579400572497e-05,
+      "loss": 1.0995,
+      "step": 5711
+    },
+    {
+      "epoch": 0.9794028763100928,
+      "grad_norm": 1.515625,
+      "learning_rate": 1.528904682947707e-05,
+      "loss": 0.9783,
+      "step": 5712
+    },
+    {
+      "epoch": 0.9795743403990826,
+      "grad_norm": 1.4921875,
+      "learning_rate": 1.528751408588591e-05,
+      "loss": 0.9386,
+      "step": 5713
+    },
+    {
+      "epoch": 0.9797458044880726,
+      "grad_norm": 1.578125,
+      "learning_rate": 1.5285981169849002e-05,
+      "loss": 1.0191,
+      "step": 5714
+    },
+    {
+      "epoch": 0.9799172685770624,
+      "grad_norm": 1.5,
+      "learning_rate": 1.5284448081416346e-05,
+      "loss": 0.8976,
+      "step": 5715
+    },
+    {
+      "epoch": 0.9800887326660522,
+      "grad_norm": 1.4921875,
+      "learning_rate": 1.5282914820637938e-05,
+      "loss": 0.8958,
+      "step": 5716
+    },
+    {
+      "epoch": 0.9802601967550422,
+      "grad_norm": 1.59375,
+      "learning_rate": 1.5281381387563785e-05,
+      "loss": 1.0732,
+      "step": 5717
+    },
+    {
+      "epoch": 0.980431660844032,
+      "grad_norm": 1.53125,
+      "learning_rate": 1.5279847782243896e-05,
+      "loss": 0.9982,
+      "step": 5718
+    },
+    {
+      "epoch": 0.9806031249330218,
+      "grad_norm": 1.625,
+      "learning_rate": 1.5278314004728288e-05,
+      "loss": 1.0392,
+      "step": 5719
+    },
+    {
+      "epoch": 0.9807745890220118,
+      "grad_norm": 1.5546875,
+      "learning_rate": 1.5276780055066985e-05,
+      "loss": 1.02,
+      "step": 5720
+    },
+    {
+      "epoch": 0.9809460531110016,
+      "grad_norm": 1.578125,
+      "learning_rate": 1.5275245933310014e-05,
+      "loss": 0.9646,
+      "step": 5721
+    },
+    {
+      "epoch": 0.9811175171999914,
+      "grad_norm": 1.546875,
+      "learning_rate": 1.5273711639507408e-05,
+      "loss": 0.9817,
+      "step": 5722
+    },
+    {
+      "epoch": 0.9812889812889813,
+      "grad_norm": 1.5625,
+      "learning_rate": 1.5272177173709205e-05,
+      "loss": 1.0012,
+      "step": 5723
+    },
+    {
+      "epoch": 0.9814604453779712,
+      "grad_norm": 1.7265625,
+      "learning_rate": 1.5270642535965455e-05,
+      "loss": 1.0409,
+      "step": 5724
+    },
+    {
+      "epoch": 0.981631909466961,
+      "grad_norm": 1.484375,
+      "learning_rate": 1.52691077263262e-05,
+      "loss": 0.8934,
+      "step": 5725
+    },
+    {
+      "epoch": 0.9818033735559508,
+      "grad_norm": 1.46875,
+      "learning_rate": 1.52675727448415e-05,
+      "loss": 0.9786,
+      "step": 5726
+    },
+    {
+      "epoch": 0.9819748376449408,
+      "grad_norm": 1.515625,
+      "learning_rate": 1.526603759156142e-05,
+      "loss": 1.0649,
+      "step": 5727
+    },
+    {
+      "epoch": 0.9821463017339306,
+      "grad_norm": 1.4765625,
+      "learning_rate": 1.526450226653603e-05,
+      "loss": 0.9111,
+      "step": 5728
+    },
+    {
+      "epoch": 0.9823177658229204,
+      "grad_norm": 1.6015625,
+      "learning_rate": 1.5262966769815387e-05,
+      "loss": 1.0624,
+      "step": 5729
+    },
+    {
+      "epoch": 0.9824892299119103,
+      "grad_norm": 1.46875,
+      "learning_rate": 1.5261431101449584e-05,
+      "loss": 0.9538,
+      "step": 5730
+    },
+    {
+      "epoch": 0.9826606940009002,
+      "grad_norm": 1.546875,
+      "learning_rate": 1.52598952614887e-05,
+      "loss": 1.0416,
+      "step": 5731
+    },
+    {
+      "epoch": 0.98283215808989,
+      "grad_norm": 1.4921875,
+      "learning_rate": 1.5258359249982818e-05,
+      "loss": 0.9478,
+      "step": 5732
+    },
+    {
+      "epoch": 0.9830036221788799,
+      "grad_norm": 1.5703125,
+      "learning_rate": 1.5256823066982046e-05,
+      "loss": 1.006,
+      "step": 5733
+    },
+    {
+      "epoch": 0.9831750862678698,
+      "grad_norm": 1.515625,
+      "learning_rate": 1.525528671253648e-05,
+      "loss": 1.0836,
+      "step": 5734
+    },
+    {
+      "epoch": 0.9833465503568596,
+      "grad_norm": 1.546875,
+      "learning_rate": 1.5253750186696219e-05,
+      "loss": 1.029,
+      "step": 5735
+    },
+    {
+      "epoch": 0.9835180144458495,
+      "grad_norm": 1.5390625,
+      "learning_rate": 1.5252213489511386e-05,
+      "loss": 1.0368,
+      "step": 5736
+    },
+    {
+      "epoch": 0.9836894785348393,
+      "grad_norm": 1.6171875,
+      "learning_rate": 1.5250676621032091e-05,
+      "loss": 0.9558,
+      "step": 5737
+    },
+    {
+      "epoch": 0.9838609426238292,
+      "grad_norm": 1.546875,
+      "learning_rate": 1.5249139581308457e-05,
+      "loss": 1.0056,
+      "step": 5738
+    },
+    {
+      "epoch": 0.9840324067128191,
+      "grad_norm": 1.5078125,
+      "learning_rate": 1.524760237039062e-05,
+      "loss": 0.8954,
+      "step": 5739
+    },
+    {
+      "epoch": 0.9842038708018089,
+      "grad_norm": 1.5859375,
+      "learning_rate": 1.5246064988328706e-05,
+      "loss": 1.0487,
+      "step": 5740
+    },
+    {
+      "epoch": 0.9843753348907988,
+      "grad_norm": 1.5234375,
+      "learning_rate": 1.5244527435172858e-05,
+      "loss": 0.9816,
+      "step": 5741
+    },
+    {
+      "epoch": 0.9845467989797887,
+      "grad_norm": 1.5625,
+      "learning_rate": 1.5242989710973218e-05,
+      "loss": 0.9851,
+      "step": 5742
+    },
+    {
+      "epoch": 0.9847182630687785,
+      "grad_norm": 1.5703125,
+      "learning_rate": 1.5241451815779944e-05,
+      "loss": 1.0613,
+      "step": 5743
+    },
+    {
+      "epoch": 0.9848897271577683,
+      "grad_norm": 1.6328125,
+      "learning_rate": 1.523991374964319e-05,
+      "loss": 1.0254,
+      "step": 5744
+    },
+    {
+      "epoch": 0.9850611912467583,
+      "grad_norm": 1.5390625,
+      "learning_rate": 1.5238375512613113e-05,
+      "loss": 0.9928,
+      "step": 5745
+    },
+    {
+      "epoch": 0.9852326553357481,
+      "grad_norm": 1.6015625,
+      "learning_rate": 1.5236837104739887e-05,
+      "loss": 1.0757,
+      "step": 5746
+    },
+    {
+      "epoch": 0.9854041194247379,
+      "grad_norm": 1.5234375,
+      "learning_rate": 1.5235298526073684e-05,
+      "loss": 0.9682,
+      "step": 5747
+    },
+    {
+      "epoch": 0.9855755835137279,
+      "grad_norm": 1.65625,
+      "learning_rate": 1.523375977666468e-05,
+      "loss": 1.0222,
+      "step": 5748
+    },
+    {
+      "epoch": 0.9857470476027177,
+      "grad_norm": 1.65625,
+      "learning_rate": 1.523222085656306e-05,
+      "loss": 1.0489,
+      "step": 5749
+    },
+    {
+      "epoch": 0.9859185116917075,
+      "grad_norm": 1.5703125,
+      "learning_rate": 1.523068176581902e-05,
+      "loss": 1.0337,
+      "step": 5750
+    },
+    {
+      "epoch": 0.9860899757806975,
+      "grad_norm": 1.5703125,
+      "learning_rate": 1.5229142504482743e-05,
+      "loss": 0.987,
+      "step": 5751
+    },
+    {
+      "epoch": 0.9862614398696873,
+      "grad_norm": 1.5625,
+      "learning_rate": 1.5227603072604442e-05,
+      "loss": 1.0087,
+      "step": 5752
+    },
+    {
+      "epoch": 0.9864329039586771,
+      "grad_norm": 1.453125,
+      "learning_rate": 1.522606347023432e-05,
+      "loss": 0.9602,
+      "step": 5753
+    },
+    {
+      "epoch": 0.986604368047667,
+      "grad_norm": 1.6015625,
+      "learning_rate": 1.5224523697422588e-05,
+      "loss": 1.0332,
+      "step": 5754
+    },
+    {
+      "epoch": 0.9867758321366569,
+      "grad_norm": 1.4765625,
+      "learning_rate": 1.5222983754219466e-05,
+      "loss": 0.9192,
+      "step": 5755
+    },
+    {
+      "epoch": 0.9869472962256467,
+      "grad_norm": 1.5390625,
+      "learning_rate": 1.5221443640675175e-05,
+      "loss": 0.9666,
+      "step": 5756
+    },
+    {
+      "epoch": 0.9871187603146366,
+      "grad_norm": 1.53125,
+      "learning_rate": 1.5219903356839947e-05,
+      "loss": 0.9095,
+      "step": 5757
+    },
+    {
+      "epoch": 0.9872902244036265,
+      "grad_norm": 1.53125,
+      "learning_rate": 1.521836290276401e-05,
+      "loss": 0.9543,
+      "step": 5758
+    },
+    {
+      "epoch": 0.9874616884926163,
+      "grad_norm": 1.5703125,
+      "learning_rate": 1.521682227849761e-05,
+      "loss": 1.002,
+      "step": 5759
+    },
+    {
+      "epoch": 0.9876331525816062,
+      "grad_norm": 1.5,
+      "learning_rate": 1.5215281484090989e-05,
+      "loss": 0.9872,
+      "step": 5760
+    },
+    {
+      "epoch": 0.987804616670596,
+      "grad_norm": 1.4765625,
+      "learning_rate": 1.5213740519594402e-05,
+      "loss": 0.9745,
+      "step": 5761
+    },
+    {
+      "epoch": 0.9879760807595859,
+      "grad_norm": 2.0,
+      "learning_rate": 1.5212199385058104e-05,
+      "loss": 0.9453,
+      "step": 5762
+    },
+    {
+      "epoch": 0.9881475448485758,
+      "grad_norm": 1.484375,
+      "learning_rate": 1.5210658080532354e-05,
+      "loss": 0.9648,
+      "step": 5763
+    },
+    {
+      "epoch": 0.9883190089375656,
+      "grad_norm": 1.640625,
+      "learning_rate": 1.5209116606067424e-05,
+      "loss": 1.0302,
+      "step": 5764
+    },
+    {
+      "epoch": 0.9884904730265555,
+      "grad_norm": 1.5703125,
+      "learning_rate": 1.5207574961713585e-05,
+      "loss": 1.0159,
+      "step": 5765
+    },
+    {
+      "epoch": 0.9886619371155454,
+      "grad_norm": 1.578125,
+      "learning_rate": 1.520603314752112e-05,
+      "loss": 0.9792,
+      "step": 5766
+    },
+    {
+      "epoch": 0.9888334012045352,
+      "grad_norm": 1.6015625,
+      "learning_rate": 1.5204491163540307e-05,
+      "loss": 0.9606,
+      "step": 5767
+    },
+    {
+      "epoch": 0.989004865293525,
+      "grad_norm": 1.6171875,
+      "learning_rate": 1.5202949009821439e-05,
+      "loss": 1.0483,
+      "step": 5768
+    },
+    {
+      "epoch": 0.989176329382515,
+      "grad_norm": 1.5390625,
+      "learning_rate": 1.5201406686414812e-05,
+      "loss": 1.0426,
+      "step": 5769
+    },
+    {
+      "epoch": 0.9893477934715048,
+      "grad_norm": 1.46875,
+      "learning_rate": 1.5199864193370725e-05,
+      "loss": 0.9885,
+      "step": 5770
+    },
+    {
+      "epoch": 0.9895192575604946,
+      "grad_norm": 1.5390625,
+      "learning_rate": 1.5198321530739487e-05,
+      "loss": 1.0593,
+      "step": 5771
+    },
+    {
+      "epoch": 0.9896907216494846,
+      "grad_norm": 1.5625,
+      "learning_rate": 1.5196778698571409e-05,
+      "loss": 1.0358,
+      "step": 5772
+    },
+    {
+      "epoch": 0.9898621857384744,
+      "grad_norm": 1.6640625,
+      "learning_rate": 1.5195235696916809e-05,
+      "loss": 0.9692,
+      "step": 5773
+    },
+    {
+      "epoch": 0.9900336498274642,
+      "grad_norm": 1.5703125,
+      "learning_rate": 1.5193692525826007e-05,
+      "loss": 0.8914,
+      "step": 5774
+    },
+    {
+      "epoch": 0.9902051139164542,
+      "grad_norm": 1.5625,
+      "learning_rate": 1.5192149185349334e-05,
+      "loss": 0.9372,
+      "step": 5775
+    },
+    {
+      "epoch": 0.990376578005444,
+      "grad_norm": 1.4921875,
+      "learning_rate": 1.5190605675537128e-05,
+      "loss": 0.8688,
+      "step": 5776
+    },
+    {
+      "epoch": 0.9905480420944338,
+      "grad_norm": 1.5,
+      "learning_rate": 1.518906199643972e-05,
+      "loss": 0.9517,
+      "step": 5777
+    },
+    {
+      "epoch": 0.9907195061834237,
+      "grad_norm": 1.46875,
+      "learning_rate": 1.5187518148107466e-05,
+      "loss": 0.8902,
+      "step": 5778
+    },
+    {
+      "epoch": 0.9908909702724136,
+      "grad_norm": 1.5703125,
+      "learning_rate": 1.5185974130590704e-05,
+      "loss": 1.0075,
+      "step": 5779
+    },
+    {
+      "epoch": 0.9910624343614034,
+      "grad_norm": 1.625,
+      "learning_rate": 1.51844299439398e-05,
+      "loss": 1.0397,
+      "step": 5780
+    },
+    {
+      "epoch": 0.9912338984503933,
+      "grad_norm": 1.5546875,
+      "learning_rate": 1.5182885588205114e-05,
+      "loss": 0.9898,
+      "step": 5781
+    },
+    {
+      "epoch": 0.9914053625393832,
+      "grad_norm": 1.53125,
+      "learning_rate": 1.5181341063437012e-05,
+      "loss": 0.9545,
+      "step": 5782
+    },
+    {
+      "epoch": 0.991576826628373,
+      "grad_norm": 1.65625,
+      "learning_rate": 1.5179796369685867e-05,
+      "loss": 1.1139,
+      "step": 5783
+    },
+    {
+      "epoch": 0.9917482907173629,
+      "grad_norm": 1.46875,
+      "learning_rate": 1.5178251507002056e-05,
+      "loss": 0.883,
+      "step": 5784
+    },
+    {
+      "epoch": 0.9919197548063527,
+      "grad_norm": 1.5234375,
+      "learning_rate": 1.5176706475435964e-05,
+      "loss": 0.9783,
+      "step": 5785
+    },
+    {
+      "epoch": 0.9920912188953426,
+      "grad_norm": 1.640625,
+      "learning_rate": 1.517516127503798e-05,
+      "loss": 1.1041,
+      "step": 5786
+    },
+    {
+      "epoch": 0.9922626829843325,
+      "grad_norm": 1.453125,
+      "learning_rate": 1.5173615905858499e-05,
+      "loss": 0.9326,
+      "step": 5787
+    },
+    {
+      "epoch": 0.9924341470733223,
+      "grad_norm": 1.640625,
+      "learning_rate": 1.5172070367947922e-05,
+      "loss": 1.0264,
+      "step": 5788
+    },
+    {
+      "epoch": 0.9926056111623122,
+      "grad_norm": 1.53125,
+      "learning_rate": 1.5170524661356654e-05,
+      "loss": 0.9666,
+      "step": 5789
+    },
+    {
+      "epoch": 0.9927770752513021,
+      "grad_norm": 1.5078125,
+      "learning_rate": 1.5168978786135102e-05,
+      "loss": 0.9852,
+      "step": 5790
+    },
+    {
+      "epoch": 0.9929485393402919,
+      "grad_norm": 1.6015625,
+      "learning_rate": 1.5167432742333694e-05,
+      "loss": 1.0273,
+      "step": 5791
+    },
+    {
+      "epoch": 0.9931200034292818,
+      "grad_norm": 1.6015625,
+      "learning_rate": 1.5165886530002842e-05,
+      "loss": 1.0046,
+      "step": 5792
+    },
+    {
+      "epoch": 0.9932914675182717,
+      "grad_norm": 1.578125,
+      "learning_rate": 1.516434014919298e-05,
+      "loss": 0.9921,
+      "step": 5793
+    },
+    {
+      "epoch": 0.9934629316072615,
+      "grad_norm": 1.59375,
+      "learning_rate": 1.5162793599954535e-05,
+      "loss": 1.0535,
+      "step": 5794
+    },
+    {
+      "epoch": 0.9936343956962513,
+      "grad_norm": 1.59375,
+      "learning_rate": 1.5161246882337952e-05,
+      "loss": 1.0442,
+      "step": 5795
+    },
+    {
+      "epoch": 0.9938058597852413,
+      "grad_norm": 1.453125,
+      "learning_rate": 1.5159699996393672e-05,
+      "loss": 0.94,
+      "step": 5796
+    },
+    {
+      "epoch": 0.9939773238742311,
+      "grad_norm": 1.59375,
+      "learning_rate": 1.5158152942172144e-05,
+      "loss": 1.0225,
+      "step": 5797
+    },
+    {
+      "epoch": 0.9941487879632209,
+      "grad_norm": 1.546875,
+      "learning_rate": 1.5156605719723824e-05,
+      "loss": 0.9897,
+      "step": 5798
+    },
+    {
+      "epoch": 0.9943202520522109,
+      "grad_norm": 1.5703125,
+      "learning_rate": 1.5155058329099176e-05,
+      "loss": 1.0447,
+      "step": 5799
+    },
+    {
+      "epoch": 0.9944917161412007,
+      "grad_norm": 1.578125,
+      "learning_rate": 1.5153510770348665e-05,
+      "loss": 1.0434,
+      "step": 5800
+    },
+    {
+      "epoch": 0.9946631802301905,
+      "grad_norm": 1.6015625,
+      "learning_rate": 1.5151963043522759e-05,
+      "loss": 0.934,
+      "step": 5801
+    },
+    {
+      "epoch": 0.9948346443191805,
+      "grad_norm": 1.46875,
+      "learning_rate": 1.515041514867194e-05,
+      "loss": 0.9631,
+      "step": 5802
+    },
+    {
+      "epoch": 0.9950061084081703,
+      "grad_norm": 1.546875,
+      "learning_rate": 1.5148867085846686e-05,
+      "loss": 0.9318,
+      "step": 5803
+    },
+    {
+      "epoch": 0.9951775724971601,
+      "grad_norm": 1.5,
+      "learning_rate": 1.5147318855097491e-05,
+      "loss": 0.954,
+      "step": 5804
+    },
+    {
+      "epoch": 0.99534903658615,
+      "grad_norm": 1.53125,
+      "learning_rate": 1.5145770456474842e-05,
+      "loss": 1.044,
+      "step": 5805
+    },
+    {
+      "epoch": 0.9955205006751399,
+      "grad_norm": 1.625,
+      "learning_rate": 1.514422189002924e-05,
+      "loss": 1.0305,
+      "step": 5806
+    },
+    {
+      "epoch": 0.9956919647641297,
+      "grad_norm": 1.59375,
+      "learning_rate": 1.5142673155811192e-05,
+      "loss": 1.0378,
+      "step": 5807
+    },
+    {
+      "epoch": 0.9958634288531196,
+      "grad_norm": 1.5625,
+      "learning_rate": 1.5141124253871206e-05,
+      "loss": 1.0143,
+      "step": 5808
+    },
+    {
+      "epoch": 0.9960348929421095,
+      "grad_norm": 1.7578125,
+      "learning_rate": 1.51395751842598e-05,
+      "loss": 1.0365,
+      "step": 5809
+    },
+    {
+      "epoch": 0.9962063570310993,
+      "grad_norm": 1.6171875,
+      "learning_rate": 1.5138025947027495e-05,
+      "loss": 0.9591,
+      "step": 5810
+    },
+    {
+      "epoch": 0.9963778211200892,
+      "grad_norm": 1.6171875,
+      "learning_rate": 1.5136476542224813e-05,
+      "loss": 1.0616,
+      "step": 5811
+    },
+    {
+      "epoch": 0.996549285209079,
+      "grad_norm": 1.5546875,
+      "learning_rate": 1.513492696990229e-05,
+      "loss": 0.9995,
+      "step": 5812
+    },
+    {
+      "epoch": 0.9967207492980689,
+      "grad_norm": 1.5234375,
+      "learning_rate": 1.5133377230110461e-05,
+      "loss": 0.9262,
+      "step": 5813
+    },
+    {
+      "epoch": 0.9968922133870588,
+      "grad_norm": 1.6171875,
+      "learning_rate": 1.5131827322899873e-05,
+      "loss": 1.0377,
+      "step": 5814
+    },
+    {
+      "epoch": 0.9970636774760486,
+      "grad_norm": 1.5859375,
+      "learning_rate": 1.5130277248321068e-05,
+      "loss": 1.0408,
+      "step": 5815
+    },
+    {
+      "epoch": 0.9972351415650385,
+      "grad_norm": 1.6953125,
+      "learning_rate": 1.5128727006424605e-05,
+      "loss": 0.9927,
+      "step": 5816
+    },
+    {
+      "epoch": 0.9974066056540284,
+      "grad_norm": 1.59375,
+      "learning_rate": 1.512717659726104e-05,
+      "loss": 0.9785,
+      "step": 5817
+    },
+    {
+      "epoch": 0.9975780697430182,
+      "grad_norm": 1.6015625,
+      "learning_rate": 1.5125626020880937e-05,
+      "loss": 1.0031,
+      "step": 5818
+    },
+    {
+      "epoch": 0.997749533832008,
+      "grad_norm": 1.6171875,
+      "learning_rate": 1.5124075277334873e-05,
+      "loss": 1.0179,
+      "step": 5819
+    },
+    {
+      "epoch": 0.997920997920998,
+      "grad_norm": 1.5,
+      "learning_rate": 1.5122524366673416e-05,
+      "loss": 1.0134,
+      "step": 5820
+    },
+    {
+      "epoch": 0.9980924620099878,
+      "grad_norm": 1.5390625,
+      "learning_rate": 1.5120973288947149e-05,
+      "loss": 1.0051,
+      "step": 5821
+    },
+    {
+      "epoch": 0.9982639260989776,
+      "grad_norm": 1.53125,
+      "learning_rate": 1.5119422044206661e-05,
+      "loss": 1.0071,
+      "step": 5822
+    },
+    {
+      "epoch": 0.9984353901879675,
+      "grad_norm": 1.578125,
+      "learning_rate": 1.5117870632502542e-05,
+      "loss": 0.9875,
+      "step": 5823
+    },
+    {
+      "epoch": 0.9986068542769574,
+      "grad_norm": 1.6171875,
+      "learning_rate": 1.5116319053885388e-05,
+      "loss": 1.0501,
+      "step": 5824
+    },
+    {
+      "epoch": 0.9987783183659472,
+      "grad_norm": 1.53125,
+      "learning_rate": 1.5114767308405807e-05,
+      "loss": 0.9723,
+      "step": 5825
+    },
+    {
+      "epoch": 0.998949782454937,
+      "grad_norm": 1.5703125,
+      "learning_rate": 1.51132153961144e-05,
+      "loss": 1.0194,
+      "step": 5826
+    },
+    {
+      "epoch": 0.999121246543927,
+      "grad_norm": 1.4296875,
+      "learning_rate": 1.5111663317061785e-05,
+      "loss": 1.0033,
+      "step": 5827
+    },
+    {
+      "epoch": 0.9992927106329168,
+      "grad_norm": 1.6328125,
+      "learning_rate": 1.5110111071298581e-05,
+      "loss": 0.9697,
+      "step": 5828
+    },
+    {
+      "epoch": 0.9994641747219066,
+      "grad_norm": 1.4765625,
+      "learning_rate": 1.5108558658875411e-05,
+      "loss": 0.9253,
+      "step": 5829
+    },
+    {
+      "epoch": 0.9996356388108966,
+      "grad_norm": 1.4765625,
+      "learning_rate": 1.510700607984291e-05,
+      "loss": 0.9752,
+      "step": 5830
+    },
+    {
+      "epoch": 0.9998071028998864,
+      "grad_norm": 1.6015625,
+      "learning_rate": 1.5105453334251708e-05,
+      "loss": 0.9819,
+      "step": 5831
+    },
+    {
+      "epoch": 0.9999785669888762,
+      "grad_norm": 1.5078125,
+      "learning_rate": 1.5103900422152445e-05,
+      "loss": 0.9919,
+      "step": 5832
+    },
+    {
+      "epoch": 1.0001500310778662,
+      "grad_norm": 1.5390625,
+      "learning_rate": 1.510234734359577e-05,
+      "loss": 0.9498,
+      "step": 5833
+    },
+    {
+      "epoch": 1.000321495166856,
+      "grad_norm": 1.5390625,
+      "learning_rate": 1.5100794098632336e-05,
+      "loss": 0.9506,
+      "step": 5834
+    },
+    {
+      "epoch": 1.0004929592558458,
+      "grad_norm": 1.5546875,
+      "learning_rate": 1.5099240687312797e-05,
+      "loss": 1.128,
+      "step": 5835
+    },
+    {
+      "epoch": 1.0006644233448356,
+      "grad_norm": 1.46875,
+      "learning_rate": 1.5097687109687818e-05,
+      "loss": 0.9603,
+      "step": 5836
+    },
+    {
+      "epoch": 1.0008358874338257,
+      "grad_norm": 1.625,
+      "learning_rate": 1.5096133365808068e-05,
+      "loss": 0.935,
+      "step": 5837
+    },
+    {
+      "epoch": 1.0010073515228155,
+      "grad_norm": 1.5625,
+      "learning_rate": 1.5094579455724216e-05,
+      "loss": 1.0625,
+      "step": 5838
+    },
+    {
+      "epoch": 1.0011788156118053,
+      "grad_norm": 1.5703125,
+      "learning_rate": 1.5093025379486945e-05,
+      "loss": 1.0261,
+      "step": 5839
+    },
+    {
+      "epoch": 1.0013502797007952,
+      "grad_norm": 1.65625,
+      "learning_rate": 1.5091471137146938e-05,
+      "loss": 1.0069,
+      "step": 5840
+    },
+    {
+      "epoch": 1.001521743789785,
+      "grad_norm": 1.5859375,
+      "learning_rate": 1.5089916728754882e-05,
+      "loss": 0.9623,
+      "step": 5841
+    },
+    {
+      "epoch": 1.0016932078787748,
+      "grad_norm": 1.5390625,
+      "learning_rate": 1.5088362154361477e-05,
+      "loss": 0.9904,
+      "step": 5842
+    },
+    {
+      "epoch": 1.0018646719677649,
+      "grad_norm": 1.4296875,
+      "learning_rate": 1.5086807414017424e-05,
+      "loss": 0.8465,
+      "step": 5843
+    },
+    {
+      "epoch": 1.0020361360567547,
+      "grad_norm": 1.6171875,
+      "learning_rate": 1.508525250777342e-05,
+      "loss": 1.0136,
+      "step": 5844
+    },
+    {
+      "epoch": 1.0022076001457445,
+      "grad_norm": 1.6015625,
+      "learning_rate": 1.5083697435680185e-05,
+      "loss": 1.001,
+      "step": 5845
+    },
+    {
+      "epoch": 1.0023790642347343,
+      "grad_norm": 1.5703125,
+      "learning_rate": 1.5082142197788435e-05,
+      "loss": 1.0401,
+      "step": 5846
+    },
+    {
+      "epoch": 1.0025505283237242,
+      "grad_norm": 1.5078125,
+      "learning_rate": 1.5080586794148887e-05,
+      "loss": 1.0249,
+      "step": 5847
+    },
+    {
+      "epoch": 1.002721992412714,
+      "grad_norm": 1.5859375,
+      "learning_rate": 1.5079031224812276e-05,
+      "loss": 0.9805,
+      "step": 5848
+    },
+    {
+      "epoch": 1.002893456501704,
+      "grad_norm": 1.484375,
+      "learning_rate": 1.5077475489829327e-05,
+      "loss": 0.9111,
+      "step": 5849
+    },
+    {
+      "epoch": 1.0030649205906939,
+      "grad_norm": 1.5078125,
+      "learning_rate": 1.5075919589250785e-05,
+      "loss": 0.9759,
+      "step": 5850
+    },
+    {
+      "epoch": 1.0032363846796837,
+      "grad_norm": 1.546875,
+      "learning_rate": 1.5074363523127389e-05,
+      "loss": 0.9997,
+      "step": 5851
+    },
+    {
+      "epoch": 1.0034078487686735,
+      "grad_norm": 1.5546875,
+      "learning_rate": 1.5072807291509892e-05,
+      "loss": 0.9388,
+      "step": 5852
+    },
+    {
+      "epoch": 1.0035793128576633,
+      "grad_norm": 1.453125,
+      "learning_rate": 1.5071250894449046e-05,
+      "loss": 0.9081,
+      "step": 5853
+    },
+    {
+      "epoch": 1.0037507769466532,
+      "grad_norm": 1.5,
+      "learning_rate": 1.506969433199561e-05,
+      "loss": 0.9728,
+      "step": 5854
+    },
+    {
+      "epoch": 1.0039222410356432,
+      "grad_norm": 1.5078125,
+      "learning_rate": 1.5068137604200353e-05,
+      "loss": 0.9584,
+      "step": 5855
+    },
+    {
+      "epoch": 1.004093705124633,
+      "grad_norm": 1.546875,
+      "learning_rate": 1.5066580711114045e-05,
+      "loss": 1.0177,
+      "step": 5856
+    },
+    {
+      "epoch": 1.0042651692136229,
+      "grad_norm": 1.5546875,
+      "learning_rate": 1.5065023652787459e-05,
+      "loss": 1.0031,
+      "step": 5857
+    },
+    {
+      "epoch": 1.0044366333026127,
+      "grad_norm": 1.6171875,
+      "learning_rate": 1.506346642927138e-05,
+      "loss": 0.9219,
+      "step": 5858
+    },
+    {
+      "epoch": 1.0046080973916025,
+      "grad_norm": 1.546875,
+      "learning_rate": 1.5061909040616592e-05,
+      "loss": 0.9689,
+      "step": 5859
+    },
+    {
+      "epoch": 1.0047795614805923,
+      "grad_norm": 1.6171875,
+      "learning_rate": 1.5060351486873893e-05,
+      "loss": 1.0111,
+      "step": 5860
+    },
+    {
+      "epoch": 1.0049510255695824,
+      "grad_norm": 1.6484375,
+      "learning_rate": 1.505879376809407e-05,
+      "loss": 0.9455,
+      "step": 5861
+    },
+    {
+      "epoch": 1.0051224896585722,
+      "grad_norm": 1.65625,
+      "learning_rate": 1.5057235884327938e-05,
+      "loss": 0.9665,
+      "step": 5862
+    },
+    {
+      "epoch": 1.005293953747562,
+      "grad_norm": 1.53125,
+      "learning_rate": 1.5055677835626296e-05,
+      "loss": 0.9387,
+      "step": 5863
+    },
+    {
+      "epoch": 1.0054654178365519,
+      "grad_norm": 1.5546875,
+      "learning_rate": 1.5054119622039962e-05,
+      "loss": 0.998,
+      "step": 5864
+    },
+    {
+      "epoch": 1.0056368819255417,
+      "grad_norm": 1.578125,
+      "learning_rate": 1.5052561243619756e-05,
+      "loss": 1.047,
+      "step": 5865
+    },
+    {
+      "epoch": 1.0058083460145315,
+      "grad_norm": 1.5390625,
+      "learning_rate": 1.5051002700416505e-05,
+      "loss": 0.9357,
+      "step": 5866
+    },
+    {
+      "epoch": 1.0059798101035216,
+      "grad_norm": 1.6171875,
+      "learning_rate": 1.5049443992481031e-05,
+      "loss": 1.1065,
+      "step": 5867
+    },
+    {
+      "epoch": 1.0061512741925114,
+      "grad_norm": 1.5390625,
+      "learning_rate": 1.5047885119864179e-05,
+      "loss": 1.032,
+      "step": 5868
+    },
+    {
+      "epoch": 1.0063227382815012,
+      "grad_norm": 2.53125,
+      "learning_rate": 1.504632608261678e-05,
+      "loss": 0.9967,
+      "step": 5869
+    },
+    {
+      "epoch": 1.006494202370491,
+      "grad_norm": 1.6015625,
+      "learning_rate": 1.5044766880789686e-05,
+      "loss": 0.9772,
+      "step": 5870
+    },
+    {
+      "epoch": 1.0066656664594809,
+      "grad_norm": 1.5234375,
+      "learning_rate": 1.504320751443375e-05,
+      "loss": 0.9998,
+      "step": 5871
+    },
+    {
+      "epoch": 1.0068371305484707,
+      "grad_norm": 1.46875,
+      "learning_rate": 1.504164798359982e-05,
+      "loss": 0.9344,
+      "step": 5872
+    },
+    {
+      "epoch": 1.0070085946374605,
+      "grad_norm": 1.578125,
+      "learning_rate": 1.504008828833877e-05,
+      "loss": 1.1216,
+      "step": 5873
+    },
+    {
+      "epoch": 1.0071800587264506,
+      "grad_norm": 1.59375,
+      "learning_rate": 1.503852842870146e-05,
+      "loss": 1.004,
+      "step": 5874
+    },
+    {
+      "epoch": 1.0073515228154404,
+      "grad_norm": 1.5078125,
+      "learning_rate": 1.503696840473876e-05,
+      "loss": 0.9362,
+      "step": 5875
+    },
+    {
+      "epoch": 1.0075229869044302,
+      "grad_norm": 1.46875,
+      "learning_rate": 1.5035408216501559e-05,
+      "loss": 0.9073,
+      "step": 5876
+    },
+    {
+      "epoch": 1.00769445099342,
+      "grad_norm": 1.4609375,
+      "learning_rate": 1.5033847864040734e-05,
+      "loss": 0.9681,
+      "step": 5877
+    },
+    {
+      "epoch": 1.0078659150824099,
+      "grad_norm": 1.5625,
+      "learning_rate": 1.5032287347407173e-05,
+      "loss": 0.9663,
+      "step": 5878
+    },
+    {
+      "epoch": 1.0080373791713997,
+      "grad_norm": 1.5234375,
+      "learning_rate": 1.5030726666651771e-05,
+      "loss": 0.9741,
+      "step": 5879
+    },
+    {
+      "epoch": 1.0082088432603897,
+      "grad_norm": 1.53125,
+      "learning_rate": 1.5029165821825429e-05,
+      "loss": 1.0372,
+      "step": 5880
+    },
+    {
+      "epoch": 1.0083803073493796,
+      "grad_norm": 1.5625,
+      "learning_rate": 1.502760481297905e-05,
+      "loss": 1.0015,
+      "step": 5881
+    },
+    {
+      "epoch": 1.0085517714383694,
+      "grad_norm": 1.6953125,
+      "learning_rate": 1.5026043640163548e-05,
+      "loss": 1.0011,
+      "step": 5882
+    },
+    {
+      "epoch": 1.0087232355273592,
+      "grad_norm": 1.6640625,
+      "learning_rate": 1.5024482303429837e-05,
+      "loss": 1.0835,
+      "step": 5883
+    },
+    {
+      "epoch": 1.008894699616349,
+      "grad_norm": 1.5390625,
+      "learning_rate": 1.5022920802828835e-05,
+      "loss": 0.9673,
+      "step": 5884
+    },
+    {
+      "epoch": 1.0090661637053389,
+      "grad_norm": 2.34375,
+      "learning_rate": 1.5021359138411473e-05,
+      "loss": 1.0423,
+      "step": 5885
+    },
+    {
+      "epoch": 1.009237627794329,
+      "grad_norm": 1.53125,
+      "learning_rate": 1.5019797310228681e-05,
+      "loss": 0.9883,
+      "step": 5886
+    },
+    {
+      "epoch": 1.0094090918833187,
+      "grad_norm": 1.6484375,
+      "learning_rate": 1.5018235318331394e-05,
+      "loss": 1.0063,
+      "step": 5887
+    },
+    {
+      "epoch": 1.0095805559723086,
+      "grad_norm": 1.53125,
+      "learning_rate": 1.5016673162770556e-05,
+      "loss": 1.0686,
+      "step": 5888
+    },
+    {
+      "epoch": 1.0097520200612984,
+      "grad_norm": 1.5625,
+      "learning_rate": 1.5015110843597121e-05,
+      "loss": 0.997,
+      "step": 5889
+    },
+    {
+      "epoch": 1.0099234841502882,
+      "grad_norm": 1.6328125,
+      "learning_rate": 1.5013548360862032e-05,
+      "loss": 0.9651,
+      "step": 5890
+    },
+    {
+      "epoch": 1.010094948239278,
+      "grad_norm": 1.796875,
+      "learning_rate": 1.5011985714616248e-05,
+      "loss": 1.0072,
+      "step": 5891
+    },
+    {
+      "epoch": 1.010266412328268,
+      "grad_norm": 1.6328125,
+      "learning_rate": 1.501042290491074e-05,
+      "loss": 1.0448,
+      "step": 5892
+    },
+    {
+      "epoch": 1.010437876417258,
+      "grad_norm": 1.640625,
+      "learning_rate": 1.5008859931796474e-05,
+      "loss": 1.0269,
+      "step": 5893
+    },
+    {
+      "epoch": 1.0106093405062477,
+      "grad_norm": 1.5078125,
+      "learning_rate": 1.5007296795324427e-05,
+      "loss": 1.0092,
+      "step": 5894
+    },
+    {
+      "epoch": 1.0107808045952376,
+      "grad_norm": 1.609375,
+      "learning_rate": 1.5005733495545572e-05,
+      "loss": 0.9801,
+      "step": 5895
+    },
+    {
+      "epoch": 1.0109522686842274,
+      "grad_norm": 1.53125,
+      "learning_rate": 1.5004170032510896e-05,
+      "loss": 0.9324,
+      "step": 5896
+    },
+    {
+      "epoch": 1.0111237327732172,
+      "grad_norm": 1.625,
+      "learning_rate": 1.5002606406271393e-05,
+      "loss": 1.0021,
+      "step": 5897
+    },
+    {
+      "epoch": 1.0112951968622073,
+      "grad_norm": 1.5078125,
+      "learning_rate": 1.5001042616878057e-05,
+      "loss": 0.9742,
+      "step": 5898
+    },
+    {
+      "epoch": 1.011466660951197,
+      "grad_norm": 1.5625,
+      "learning_rate": 1.499947866438189e-05,
+      "loss": 0.9153,
+      "step": 5899
+    },
+    {
+      "epoch": 1.011638125040187,
+      "grad_norm": 1.5625,
+      "learning_rate": 1.4997914548833898e-05,
+      "loss": 1.026,
+      "step": 5900
+    },
+    {
+      "epoch": 1.0118095891291767,
+      "grad_norm": 1.625,
+      "learning_rate": 1.4996350270285089e-05,
+      "loss": 0.9569,
+      "step": 5901
+    },
+    {
+      "epoch": 1.0119810532181666,
+      "grad_norm": 1.8359375,
+      "learning_rate": 1.4994785828786487e-05,
+      "loss": 0.9064,
+      "step": 5902
+    },
+    {
+      "epoch": 1.0001714640889898,
+      "grad_norm": 1.75,
+      "learning_rate": 1.4993221224389109e-05,
+      "loss": 0.9062,
+      "step": 5903
+    },
+    {
+      "epoch": 1.0003429281779797,
+      "grad_norm": 1.7109375,
+      "learning_rate": 1.4991656457143984e-05,
+      "loss": 0.8624,
+      "step": 5904
+    },
+    {
+      "epoch": 1.0005143922669695,
+      "grad_norm": 1.515625,
+      "learning_rate": 1.4990091527102145e-05,
+      "loss": 0.8074,
+      "step": 5905
+    },
+    {
+      "epoch": 1.0006858563559595,
+      "grad_norm": 1.625,
+      "learning_rate": 1.4988526434314633e-05,
+      "loss": 0.8797,
+      "step": 5906
+    },
+    {
+      "epoch": 1.0008573204449493,
+      "grad_norm": 1.5625,
+      "learning_rate": 1.4986961178832485e-05,
+      "loss": 0.9083,
+      "step": 5907
+    },
+    {
+      "epoch": 1.0010287845339392,
+      "grad_norm": 1.6171875,
+      "learning_rate": 1.4985395760706756e-05,
+      "loss": 0.8394,
+      "step": 5908
+    },
+    {
+      "epoch": 1.001200248622929,
+      "grad_norm": 1.7109375,
+      "learning_rate": 1.4983830179988497e-05,
+      "loss": 0.7831,
+      "step": 5909
+    },
+    {
+      "epoch": 1.0013717127119188,
+      "grad_norm": 1.8359375,
+      "learning_rate": 1.4982264436728768e-05,
+      "loss": 0.9046,
+      "step": 5910
+    },
+    {
+      "epoch": 1.0015431768009087,
+      "grad_norm": 1.84375,
+      "learning_rate": 1.4980698530978632e-05,
+      "loss": 0.8875,
+      "step": 5911
+    },
+    {
+      "epoch": 1.0017146408898987,
+      "grad_norm": 1.703125,
+      "learning_rate": 1.4979132462789168e-05,
+      "loss": 0.7999,
+      "step": 5912
+    },
+    {
+      "epoch": 1.0018861049788885,
+      "grad_norm": 1.6015625,
+      "learning_rate": 1.497756623221144e-05,
+      "loss": 0.8818,
+      "step": 5913
+    },
+    {
+      "epoch": 1.0020575690678784,
+      "grad_norm": 1.609375,
+      "learning_rate": 1.4975999839296534e-05,
+      "loss": 0.8981,
+      "step": 5914
+    },
+    {
+      "epoch": 1.0022290331568682,
+      "grad_norm": 1.71875,
+      "learning_rate": 1.4974433284095535e-05,
+      "loss": 0.9054,
+      "step": 5915
+    },
+    {
+      "epoch": 1.002400497245858,
+      "grad_norm": 1.6953125,
+      "learning_rate": 1.4972866566659537e-05,
+      "loss": 0.8543,
+      "step": 5916
+    },
+    {
+      "epoch": 1.0025719613348478,
+      "grad_norm": 2.0,
+      "learning_rate": 1.4971299687039634e-05,
+      "loss": 0.9116,
+      "step": 5917
+    },
+    {
+      "epoch": 1.0027434254238379,
+      "grad_norm": 1.6015625,
+      "learning_rate": 1.4969732645286925e-05,
+      "loss": 0.8365,
+      "step": 5918
+    },
+    {
+      "epoch": 1.0029148895128277,
+      "grad_norm": 1.71875,
+      "learning_rate": 1.496816544145252e-05,
+      "loss": 0.9019,
+      "step": 5919
+    },
+    {
+      "epoch": 1.0030863536018175,
+      "grad_norm": 1.609375,
+      "learning_rate": 1.4966598075587532e-05,
+      "loss": 0.8775,
+      "step": 5920
+    },
+    {
+      "epoch": 1.0032578176908074,
+      "grad_norm": 1.59375,
+      "learning_rate": 1.496503054774308e-05,
+      "loss": 0.7859,
+      "step": 5921
+    },
+    {
+      "epoch": 1.0034292817797972,
+      "grad_norm": 1.703125,
+      "learning_rate": 1.4963462857970285e-05,
+      "loss": 0.8872,
+      "step": 5922
+    },
+    {
+      "epoch": 1.003600745868787,
+      "grad_norm": 1.7109375,
+      "learning_rate": 1.4961895006320276e-05,
+      "loss": 0.9129,
+      "step": 5923
+    },
+    {
+      "epoch": 1.003772209957777,
+      "grad_norm": 1.5703125,
+      "learning_rate": 1.4960326992844188e-05,
+      "loss": 0.8566,
+      "step": 5924
+    },
+    {
+      "epoch": 1.0039436740467669,
+      "grad_norm": 1.7578125,
+      "learning_rate": 1.4958758817593153e-05,
+      "loss": 0.9195,
+      "step": 5925
+    },
+    {
+      "epoch": 1.0041151381357567,
+      "grad_norm": 1.75,
+      "learning_rate": 1.4957190480618323e-05,
+      "loss": 0.9396,
+      "step": 5926
+    },
+    {
+      "epoch": 1.0042866022247465,
+      "grad_norm": 1.6953125,
+      "learning_rate": 1.4955621981970844e-05,
+      "loss": 0.8806,
+      "step": 5927
+    },
+    {
+      "epoch": 1.0044580663137364,
+      "grad_norm": 1.578125,
+      "learning_rate": 1.4954053321701871e-05,
+      "loss": 0.8617,
+      "step": 5928
+    },
+    {
+      "epoch": 1.0046295304027262,
+      "grad_norm": 1.609375,
+      "learning_rate": 1.4952484499862563e-05,
+      "loss": 0.8603,
+      "step": 5929
+    },
+    {
+      "epoch": 1.0048009944917162,
+      "grad_norm": 1.6953125,
+      "learning_rate": 1.4950915516504088e-05,
+      "loss": 0.8906,
+      "step": 5930
+    },
+    {
+      "epoch": 1.004972458580706,
+      "grad_norm": 1.6953125,
+      "learning_rate": 1.4949346371677612e-05,
+      "loss": 0.9146,
+      "step": 5931
+    },
+    {
+      "epoch": 1.0051439226696959,
+      "grad_norm": 1.625,
+      "learning_rate": 1.4947777065434315e-05,
+      "loss": 0.9114,
+      "step": 5932
+    },
+    {
+      "epoch": 1.0053153867586857,
+      "grad_norm": 1.71875,
+      "learning_rate": 1.4946207597825375e-05,
+      "loss": 0.8995,
+      "step": 5933
+    },
+    {
+      "epoch": 1.0054868508476755,
+      "grad_norm": 1.671875,
+      "learning_rate": 1.494463796890198e-05,
+      "loss": 0.8308,
+      "step": 5934
+    },
+    {
+      "epoch": 1.0056583149366654,
+      "grad_norm": 1.65625,
+      "learning_rate": 1.4943068178715322e-05,
+      "loss": 0.8965,
+      "step": 5935
+    },
+    {
+      "epoch": 1.0058297790256554,
+      "grad_norm": 1.7109375,
+      "learning_rate": 1.49414982273166e-05,
+      "loss": 0.8661,
+      "step": 5936
+    },
+    {
+      "epoch": 1.0060012431146452,
+      "grad_norm": 1.671875,
+      "learning_rate": 1.4939928114757007e-05,
+      "loss": 0.863,
+      "step": 5937
+    },
+    {
+      "epoch": 1.006172707203635,
+      "grad_norm": 1.71875,
+      "learning_rate": 1.4938357841087757e-05,
+      "loss": 0.8777,
+      "step": 5938
+    },
+    {
+      "epoch": 1.0063441712926249,
+      "grad_norm": 1.6640625,
+      "learning_rate": 1.4936787406360062e-05,
+      "loss": 0.8211,
+      "step": 5939
+    },
+    {
+      "epoch": 1.0065156353816147,
+      "grad_norm": 1.6484375,
+      "learning_rate": 1.4935216810625141e-05,
+      "loss": 0.9032,
+      "step": 5940
+    },
+    {
+      "epoch": 1.0066870994706045,
+      "grad_norm": 1.59375,
+      "learning_rate": 1.4933646053934216e-05,
+      "loss": 0.8251,
+      "step": 5941
+    },
+    {
+      "epoch": 1.0068585635595946,
+      "grad_norm": 1.6796875,
+      "learning_rate": 1.4932075136338515e-05,
+      "loss": 0.8489,
+      "step": 5942
+    },
+    {
+      "epoch": 1.0070300276485844,
+      "grad_norm": 1.6015625,
+      "learning_rate": 1.493050405788927e-05,
+      "loss": 0.8285,
+      "step": 5943
+    },
+    {
+      "epoch": 1.0072014917375742,
+      "grad_norm": 1.4921875,
+      "learning_rate": 1.492893281863772e-05,
+      "loss": 0.8343,
+      "step": 5944
+    },
+    {
+      "epoch": 1.007372955826564,
+      "grad_norm": 1.5703125,
+      "learning_rate": 1.4927361418635111e-05,
+      "loss": 0.8195,
+      "step": 5945
+    },
+    {
+      "epoch": 1.0075444199155539,
+      "grad_norm": 1.609375,
+      "learning_rate": 1.4925789857932693e-05,
+      "loss": 0.8629,
+      "step": 5946
+    },
+    {
+      "epoch": 1.0077158840045437,
+      "grad_norm": 1.5859375,
+      "learning_rate": 1.492421813658172e-05,
+      "loss": 0.7921,
+      "step": 5947
+    },
+    {
+      "epoch": 1.0078873480935338,
+      "grad_norm": 1.53125,
+      "learning_rate": 1.4922646254633446e-05,
+      "loss": 0.8804,
+      "step": 5948
+    },
+    {
+      "epoch": 1.0080588121825236,
+      "grad_norm": 1.640625,
+      "learning_rate": 1.4921074212139141e-05,
+      "loss": 0.8934,
+      "step": 5949
+    },
+    {
+      "epoch": 1.0082302762715134,
+      "grad_norm": 1.6171875,
+      "learning_rate": 1.4919502009150077e-05,
+      "loss": 0.8651,
+      "step": 5950
+    },
+    {
+      "epoch": 1.0084017403605032,
+      "grad_norm": 1.5546875,
+      "learning_rate": 1.4917929645717528e-05,
+      "loss": 0.8723,
+      "step": 5951
+    },
+    {
+      "epoch": 1.008573204449493,
+      "grad_norm": 1.6328125,
+      "learning_rate": 1.4916357121892775e-05,
+      "loss": 0.8532,
+      "step": 5952
+    },
+    {
+      "epoch": 1.0087446685384829,
+      "grad_norm": 1.609375,
+      "learning_rate": 1.49147844377271e-05,
+      "loss": 0.8857,
+      "step": 5953
+    },
+    {
+      "epoch": 1.008916132627473,
+      "grad_norm": 1.65625,
+      "learning_rate": 1.4913211593271802e-05,
+      "loss": 0.8886,
+      "step": 5954
+    },
+    {
+      "epoch": 1.0090875967164628,
+      "grad_norm": 1.7734375,
+      "learning_rate": 1.4911638588578167e-05,
+      "loss": 0.8719,
+      "step": 5955
+    },
+    {
+      "epoch": 1.0092590608054526,
+      "grad_norm": 1.65625,
+      "learning_rate": 1.4910065423697504e-05,
+      "loss": 0.9048,
+      "step": 5956
+    },
+    {
+      "epoch": 1.0094305248944424,
+      "grad_norm": 1.5859375,
+      "learning_rate": 1.4908492098681118e-05,
+      "loss": 0.7576,
+      "step": 5957
+    },
+    {
+      "epoch": 1.0096019889834322,
+      "grad_norm": 1.671875,
+      "learning_rate": 1.4906918613580322e-05,
+      "loss": 0.8747,
+      "step": 5958
+    },
+    {
+      "epoch": 1.009773453072422,
+      "grad_norm": 1.546875,
+      "learning_rate": 1.490534496844643e-05,
+      "loss": 0.8154,
+      "step": 5959
+    },
+    {
+      "epoch": 1.009944917161412,
+      "grad_norm": 1.6640625,
+      "learning_rate": 1.4903771163330766e-05,
+      "loss": 0.7974,
+      "step": 5960
+    },
+    {
+      "epoch": 1.010116381250402,
+      "grad_norm": 1.7890625,
+      "learning_rate": 1.4902197198284663e-05,
+      "loss": 0.8984,
+      "step": 5961
+    },
+    {
+      "epoch": 1.0102878453393918,
+      "grad_norm": 1.5234375,
+      "learning_rate": 1.4900623073359445e-05,
+      "loss": 0.8233,
+      "step": 5962
+    },
+    {
+      "epoch": 1.0104593094283816,
+      "grad_norm": 1.5234375,
+      "learning_rate": 1.4899048788606456e-05,
+      "loss": 0.8013,
+      "step": 5963
+    },
+    {
+      "epoch": 1.0106307735173714,
+      "grad_norm": 1.6015625,
+      "learning_rate": 1.4897474344077041e-05,
+      "loss": 0.8674,
+      "step": 5964
+    },
+    {
+      "epoch": 1.0108022376063612,
+      "grad_norm": 1.6171875,
+      "learning_rate": 1.4895899739822542e-05,
+      "loss": 0.8388,
+      "step": 5965
+    },
+    {
+      "epoch": 1.0109737016953513,
+      "grad_norm": 1.5234375,
+      "learning_rate": 1.4894324975894314e-05,
+      "loss": 0.8599,
+      "step": 5966
+    },
+    {
+      "epoch": 1.011145165784341,
+      "grad_norm": 1.6796875,
+      "learning_rate": 1.4892750052343723e-05,
+      "loss": 0.8238,
+      "step": 5967
+    },
+    {
+      "epoch": 1.011316629873331,
+      "grad_norm": 1.5625,
+      "learning_rate": 1.4891174969222127e-05,
+      "loss": 0.8735,
+      "step": 5968
+    },
+    {
+      "epoch": 1.0114880939623208,
+      "grad_norm": 1.546875,
+      "learning_rate": 1.4889599726580898e-05,
+      "loss": 0.7963,
+      "step": 5969
+    },
+    {
+      "epoch": 1.0116595580513106,
+      "grad_norm": 1.5703125,
+      "learning_rate": 1.4888024324471406e-05,
+      "loss": 0.9024,
+      "step": 5970
+    },
+    {
+      "epoch": 1.0118310221403004,
+      "grad_norm": 1.546875,
+      "learning_rate": 1.4886448762945035e-05,
+      "loss": 0.8448,
+      "step": 5971
+    },
+    {
+      "epoch": 1.0120024862292905,
+      "grad_norm": 1.578125,
+      "learning_rate": 1.4884873042053171e-05,
+      "loss": 0.8464,
+      "step": 5972
+    },
+    {
+      "epoch": 1.0121739503182803,
+      "grad_norm": 1.5546875,
+      "learning_rate": 1.4883297161847204e-05,
+      "loss": 0.8066,
+      "step": 5973
+    },
+    {
+      "epoch": 1.01234541440727,
+      "grad_norm": 1.625,
+      "learning_rate": 1.4881721122378523e-05,
+      "loss": 0.8588,
+      "step": 5974
+    },
+    {
+      "epoch": 1.01251687849626,
+      "grad_norm": 1.640625,
+      "learning_rate": 1.4880144923698537e-05,
+      "loss": 0.8857,
+      "step": 5975
+    },
+    {
+      "epoch": 1.0126883425852498,
+      "grad_norm": 1.6171875,
+      "learning_rate": 1.4878568565858644e-05,
+      "loss": 0.8952,
+      "step": 5976
+    },
+    {
+      "epoch": 1.0128598066742396,
+      "grad_norm": 1.6875,
+      "learning_rate": 1.4876992048910264e-05,
+      "loss": 0.8991,
+      "step": 5977
+    },
+    {
+      "epoch": 1.0130312707632296,
+      "grad_norm": 1.6171875,
+      "learning_rate": 1.4875415372904804e-05,
+      "loss": 0.842,
+      "step": 5978
+    },
+    {
+      "epoch": 1.0132027348522195,
+      "grad_norm": 1.5625,
+      "learning_rate": 1.4873838537893693e-05,
+      "loss": 0.9149,
+      "step": 5979
+    },
+    {
+      "epoch": 1.0133741989412093,
+      "grad_norm": 1.609375,
+      "learning_rate": 1.4872261543928353e-05,
+      "loss": 0.8517,
+      "step": 5980
+    },
+    {
+      "epoch": 1.013545663030199,
+      "grad_norm": 1.6015625,
+      "learning_rate": 1.4870684391060217e-05,
+      "loss": 0.8273,
+      "step": 5981
+    },
+    {
+      "epoch": 1.013717127119189,
+      "grad_norm": 1.5703125,
+      "learning_rate": 1.4869107079340721e-05,
+      "loss": 0.8642,
+      "step": 5982
+    },
+    {
+      "epoch": 1.0138885912081788,
+      "grad_norm": 1.625,
+      "learning_rate": 1.486752960882131e-05,
+      "loss": 0.8797,
+      "step": 5983
+    },
+    {
+      "epoch": 1.0140600552971688,
+      "grad_norm": 1.6328125,
+      "learning_rate": 1.4865951979553426e-05,
+      "loss": 0.8991,
+      "step": 5984
+    },
+    {
+      "epoch": 1.0142315193861586,
+      "grad_norm": 1.65625,
+      "learning_rate": 1.4864374191588522e-05,
+      "loss": 0.8193,
+      "step": 5985
+    },
+    {
+      "epoch": 1.0144029834751485,
+      "grad_norm": 1.6640625,
+      "learning_rate": 1.4862796244978061e-05,
+      "loss": 0.8228,
+      "step": 5986
+    },
+    {
+      "epoch": 1.0145744475641383,
+      "grad_norm": 1.5546875,
+      "learning_rate": 1.4861218139773506e-05,
+      "loss": 0.7662,
+      "step": 5987
+    },
+    {
+      "epoch": 1.014745911653128,
+      "grad_norm": 1.609375,
+      "learning_rate": 1.485963987602632e-05,
+      "loss": 0.9135,
+      "step": 5988
+    },
+    {
+      "epoch": 1.014917375742118,
+      "grad_norm": 1.6015625,
+      "learning_rate": 1.4858061453787977e-05,
+      "loss": 0.8718,
+      "step": 5989
+    },
+    {
+      "epoch": 1.015088839831108,
+      "grad_norm": 1.640625,
+      "learning_rate": 1.4856482873109956e-05,
+      "loss": 0.8989,
+      "step": 5990
+    },
+    {
+      "epoch": 1.0152603039200978,
+      "grad_norm": 1.6484375,
+      "learning_rate": 1.4854904134043741e-05,
+      "loss": 0.7878,
+      "step": 5991
+    },
+    {
+      "epoch": 1.0154317680090876,
+      "grad_norm": 1.7265625,
+      "learning_rate": 1.4853325236640821e-05,
+      "loss": 0.886,
+      "step": 5992
+    },
+    {
+      "epoch": 1.0156032320980775,
+      "grad_norm": 1.6796875,
+      "learning_rate": 1.4851746180952687e-05,
+      "loss": 0.849,
+      "step": 5993
+    },
+    {
+      "epoch": 1.0157746961870673,
+      "grad_norm": 1.5859375,
+      "learning_rate": 1.4850166967030842e-05,
+      "loss": 0.8621,
+      "step": 5994
+    },
+    {
+      "epoch": 1.015946160276057,
+      "grad_norm": 1.6484375,
+      "learning_rate": 1.4848587594926787e-05,
+      "loss": 0.9694,
+      "step": 5995
+    },
+    {
+      "epoch": 1.0161176243650472,
+      "grad_norm": 1.6328125,
+      "learning_rate": 1.4847008064692035e-05,
+      "loss": 0.932,
+      "step": 5996
+    },
+    {
+      "epoch": 1.016289088454037,
+      "grad_norm": 1.6796875,
+      "learning_rate": 1.4845428376378094e-05,
+      "loss": 0.8344,
+      "step": 5997
+    },
+    {
+      "epoch": 1.0164605525430268,
+      "grad_norm": 1.59375,
+      "learning_rate": 1.484384853003649e-05,
+      "loss": 0.8891,
+      "step": 5998
+    },
+    {
+      "epoch": 1.0166320166320166,
+      "grad_norm": 1.6640625,
+      "learning_rate": 1.4842268525718747e-05,
+      "loss": 0.9256,
+      "step": 5999
+    },
+    {
+      "epoch": 1.0168034807210065,
+      "grad_norm": 1.6484375,
+      "learning_rate": 1.484068836347639e-05,
+      "loss": 0.822,
+      "step": 6000
+    },
+    {
+      "epoch": 1.0169749448099963,
+      "grad_norm": 1.5859375,
+      "learning_rate": 1.4839108043360958e-05,
+      "loss": 0.8274,
+      "step": 6001
+    },
+    {
+      "epoch": 1.017146408898986,
+      "grad_norm": 1.6328125,
+      "learning_rate": 1.483752756542399e-05,
+      "loss": 0.8475,
+      "step": 6002
+    },
+    {
+      "epoch": 1.0173178729879762,
+      "grad_norm": 1.546875,
+      "learning_rate": 1.4835946929717031e-05,
+      "loss": 0.797,
+      "step": 6003
+    },
+    {
+      "epoch": 1.017489337076966,
+      "grad_norm": 1.6640625,
+      "learning_rate": 1.4834366136291633e-05,
+      "loss": 0.8124,
+      "step": 6004
+    },
+    {
+      "epoch": 1.0176608011659558,
+      "grad_norm": 1.7265625,
+      "learning_rate": 1.483278518519935e-05,
+      "loss": 0.8464,
+      "step": 6005
+    },
+    {
+      "epoch": 1.0178322652549456,
+      "grad_norm": 1.6796875,
+      "learning_rate": 1.4831204076491744e-05,
+      "loss": 0.8662,
+      "step": 6006
+    },
+    {
+      "epoch": 1.0180037293439355,
+      "grad_norm": 1.6953125,
+      "learning_rate": 1.4829622810220383e-05,
+      "loss": 0.8657,
+      "step": 6007
+    },
+    {
+      "epoch": 1.0181751934329253,
+      "grad_norm": 1.625,
+      "learning_rate": 1.482804138643683e-05,
+      "loss": 0.8258,
+      "step": 6008
+    },
+    {
+      "epoch": 1.0183466575219153,
+      "grad_norm": 1.6875,
+      "learning_rate": 1.4826459805192673e-05,
+      "loss": 0.8241,
+      "step": 6009
+    },
+    {
+      "epoch": 1.0185181216109052,
+      "grad_norm": 1.703125,
+      "learning_rate": 1.4824878066539485e-05,
+      "loss": 0.8901,
+      "step": 6010
+    },
+    {
+      "epoch": 1.018689585699895,
+      "grad_norm": 1.7578125,
+      "learning_rate": 1.4823296170528854e-05,
+      "loss": 0.8794,
+      "step": 6011
+    },
+    {
+      "epoch": 1.0188610497888848,
+      "grad_norm": 1.6015625,
+      "learning_rate": 1.4821714117212369e-05,
+      "loss": 0.8573,
+      "step": 6012
+    },
+    {
+      "epoch": 1.0190325138778746,
+      "grad_norm": 1.625,
+      "learning_rate": 1.4820131906641631e-05,
+      "loss": 0.7691,
+      "step": 6013
+    },
+    {
+      "epoch": 1.0192039779668645,
+      "grad_norm": 1.8125,
+      "learning_rate": 1.481854953886824e-05,
+      "loss": 0.9168,
+      "step": 6014
+    },
+    {
+      "epoch": 1.0193754420558545,
+      "grad_norm": 1.65625,
+      "learning_rate": 1.4816967013943808e-05,
+      "loss": 0.8843,
+      "step": 6015
+    },
+    {
+      "epoch": 1.0195469061448443,
+      "grad_norm": 1.546875,
+      "learning_rate": 1.481538433191994e-05,
+      "loss": 0.8351,
+      "step": 6016
+    },
+    {
+      "epoch": 1.0197183702338342,
+      "grad_norm": 1.578125,
+      "learning_rate": 1.4813801492848256e-05,
+      "loss": 0.8354,
+      "step": 6017
+    },
+    {
+      "epoch": 1.019889834322824,
+      "grad_norm": 1.65625,
+      "learning_rate": 1.4812218496780378e-05,
+      "loss": 1.0101,
+      "step": 6018
+    },
+    {
+      "epoch": 1.0200612984118138,
+      "grad_norm": 1.6640625,
+      "learning_rate": 1.4810635343767935e-05,
+      "loss": 0.8163,
+      "step": 6019
+    },
+    {
+      "epoch": 1.0202327625008036,
+      "grad_norm": 1.5703125,
+      "learning_rate": 1.4809052033862555e-05,
+      "loss": 0.781,
+      "step": 6020
+    },
+    {
+      "epoch": 1.0204042265897937,
+      "grad_norm": 1.609375,
+      "learning_rate": 1.480746856711588e-05,
+      "loss": 0.8783,
+      "step": 6021
+    },
+    {
+      "epoch": 1.0205756906787835,
+      "grad_norm": 1.5859375,
+      "learning_rate": 1.4805884943579551e-05,
+      "loss": 0.8485,
+      "step": 6022
+    },
+    {
+      "epoch": 1.0207471547677733,
+      "grad_norm": 1.7265625,
+      "learning_rate": 1.4804301163305219e-05,
+      "loss": 0.9251,
+      "step": 6023
+    },
+    {
+      "epoch": 1.0209186188567632,
+      "grad_norm": 1.6953125,
+      "learning_rate": 1.4802717226344533e-05,
+      "loss": 0.8207,
+      "step": 6024
+    },
+    {
+      "epoch": 1.021090082945753,
+      "grad_norm": 1.6953125,
+      "learning_rate": 1.4801133132749157e-05,
+      "loss": 0.8953,
+      "step": 6025
+    },
+    {
+      "epoch": 1.0212615470347428,
+      "grad_norm": 2.15625,
+      "learning_rate": 1.4799548882570744e-05,
+      "loss": 0.8914,
+      "step": 6026
+    },
+    {
+      "epoch": 1.0214330111237329,
+      "grad_norm": 1.6796875,
+      "learning_rate": 1.4797964475860975e-05,
+      "loss": 0.8625,
+      "step": 6027
+    },
+    {
+      "epoch": 1.0216044752127227,
+      "grad_norm": 1.6328125,
+      "learning_rate": 1.4796379912671514e-05,
+      "loss": 0.8708,
+      "step": 6028
+    },
+    {
+      "epoch": 1.0217759393017125,
+      "grad_norm": 1.6875,
+      "learning_rate": 1.479479519305404e-05,
+      "loss": 0.8928,
+      "step": 6029
+    },
+    {
+      "epoch": 1.0219474033907023,
+      "grad_norm": 1.5078125,
+      "learning_rate": 1.4793210317060243e-05,
+      "loss": 0.8637,
+      "step": 6030
+    },
+    {
+      "epoch": 1.0221188674796922,
+      "grad_norm": 1.625,
+      "learning_rate": 1.4791625284741807e-05,
+      "loss": 0.8429,
+      "step": 6031
+    },
+    {
+      "epoch": 1.022290331568682,
+      "grad_norm": 1.53125,
+      "learning_rate": 1.4790040096150425e-05,
+      "loss": 0.7881,
+      "step": 6032
+    },
+    {
+      "epoch": 1.022461795657672,
+      "grad_norm": 1.6484375,
+      "learning_rate": 1.4788454751337799e-05,
+      "loss": 0.9203,
+      "step": 6033
+    },
+    {
+      "epoch": 1.0226332597466619,
+      "grad_norm": 1.6328125,
+      "learning_rate": 1.4786869250355632e-05,
+      "loss": 0.8146,
+      "step": 6034
+    },
+    {
+      "epoch": 1.0228047238356517,
+      "grad_norm": 1.5859375,
+      "learning_rate": 1.4785283593255633e-05,
+      "loss": 0.7602,
+      "step": 6035
+    },
+    {
+      "epoch": 1.0229761879246415,
+      "grad_norm": 1.6484375,
+      "learning_rate": 1.4783697780089517e-05,
+      "loss": 0.8346,
+      "step": 6036
+    },
+    {
+      "epoch": 1.0231476520136313,
+      "grad_norm": 1.671875,
+      "learning_rate": 1.4782111810909002e-05,
+      "loss": 0.9469,
+      "step": 6037
+    },
+    {
+      "epoch": 1.0233191161026212,
+      "grad_norm": 1.6640625,
+      "learning_rate": 1.4780525685765813e-05,
+      "loss": 0.9122,
+      "step": 6038
+    },
+    {
+      "epoch": 1.0234905801916112,
+      "grad_norm": 1.734375,
+      "learning_rate": 1.4778939404711682e-05,
+      "loss": 0.8712,
+      "step": 6039
+    },
+    {
+      "epoch": 1.023662044280601,
+      "grad_norm": 1.7734375,
+      "learning_rate": 1.4777352967798335e-05,
+      "loss": 0.9272,
+      "step": 6040
+    },
+    {
+      "epoch": 1.0238335083695909,
+      "grad_norm": 1.6796875,
+      "learning_rate": 1.4775766375077523e-05,
+      "loss": 0.885,
+      "step": 6041
+    },
+    {
+      "epoch": 1.0240049724585807,
+      "grad_norm": 1.6953125,
+      "learning_rate": 1.4774179626600984e-05,
+      "loss": 0.8927,
+      "step": 6042
+    },
+    {
+      "epoch": 1.0241764365475705,
+      "grad_norm": 1.578125,
+      "learning_rate": 1.4772592722420468e-05,
+      "loss": 0.7589,
+      "step": 6043
+    },
+    {
+      "epoch": 1.0243479006365603,
+      "grad_norm": 1.7890625,
+      "learning_rate": 1.4771005662587737e-05,
+      "loss": 0.7592,
+      "step": 6044
+    },
+    {
+      "epoch": 1.0245193647255504,
+      "grad_norm": 1.609375,
+      "learning_rate": 1.4769418447154541e-05,
+      "loss": 0.7909,
+      "step": 6045
+    },
+    {
+      "epoch": 1.0246908288145402,
+      "grad_norm": 1.609375,
+      "learning_rate": 1.4767831076172649e-05,
+      "loss": 0.8449,
+      "step": 6046
+    },
+    {
+      "epoch": 1.02486229290353,
+      "grad_norm": 1.5390625,
+      "learning_rate": 1.4766243549693836e-05,
+      "loss": 0.7959,
+      "step": 6047
+    },
+    {
+      "epoch": 1.0250337569925199,
+      "grad_norm": 1.6796875,
+      "learning_rate": 1.4764655867769869e-05,
+      "loss": 0.8639,
+      "step": 6048
+    },
+    {
+      "epoch": 1.0252052210815097,
+      "grad_norm": 1.6328125,
+      "learning_rate": 1.4763068030452532e-05,
+      "loss": 0.8417,
+      "step": 6049
+    },
+    {
+      "epoch": 1.0253766851704995,
+      "grad_norm": 1.546875,
+      "learning_rate": 1.476148003779361e-05,
+      "loss": 0.8112,
+      "step": 6050
+    },
+    {
+      "epoch": 1.0255481492594896,
+      "grad_norm": 1.6640625,
+      "learning_rate": 1.4759891889844895e-05,
+      "loss": 0.845,
+      "step": 6051
+    },
+    {
+      "epoch": 1.0257196133484794,
+      "grad_norm": 1.5703125,
+      "learning_rate": 1.4758303586658183e-05,
+      "loss": 0.7849,
+      "step": 6052
+    },
+    {
+      "epoch": 1.0258910774374692,
+      "grad_norm": 1.609375,
+      "learning_rate": 1.4756715128285271e-05,
+      "loss": 0.7499,
+      "step": 6053
+    },
+    {
+      "epoch": 1.026062541526459,
+      "grad_norm": 1.609375,
+      "learning_rate": 1.4755126514777965e-05,
+      "loss": 0.8702,
+      "step": 6054
+    },
+    {
+      "epoch": 1.0262340056154489,
+      "grad_norm": 1.7578125,
+      "learning_rate": 1.4753537746188081e-05,
+      "loss": 0.8347,
+      "step": 6055
+    },
+    {
+      "epoch": 1.0264054697044387,
+      "grad_norm": 1.59375,
+      "learning_rate": 1.475194882256743e-05,
+      "loss": 0.8415,
+      "step": 6056
+    },
+    {
+      "epoch": 1.0265769337934287,
+      "grad_norm": 1.6640625,
+      "learning_rate": 1.4750359743967833e-05,
+      "loss": 0.8358,
+      "step": 6057
+    },
+    {
+      "epoch": 1.0267483978824186,
+      "grad_norm": 1.6484375,
+      "learning_rate": 1.4748770510441115e-05,
+      "loss": 0.8432,
+      "step": 6058
+    },
+    {
+      "epoch": 1.0269198619714084,
+      "grad_norm": 1.6328125,
+      "learning_rate": 1.4747181122039109e-05,
+      "loss": 0.9186,
+      "step": 6059
+    },
+    {
+      "epoch": 1.0270913260603982,
+      "grad_norm": 1.671875,
+      "learning_rate": 1.4745591578813651e-05,
+      "loss": 0.8913,
+      "step": 6060
+    },
+    {
+      "epoch": 1.027262790149388,
+      "grad_norm": 1.671875,
+      "learning_rate": 1.4744001880816581e-05,
+      "loss": 0.8778,
+      "step": 6061
+    },
+    {
+      "epoch": 1.0274342542383779,
+      "grad_norm": 1.78125,
+      "learning_rate": 1.4742412028099748e-05,
+      "loss": 0.9405,
+      "step": 6062
+    },
+    {
+      "epoch": 1.027605718327368,
+      "grad_norm": 1.6953125,
+      "learning_rate": 1.4740822020714999e-05,
+      "loss": 0.8893,
+      "step": 6063
+    },
+    {
+      "epoch": 1.0277771824163577,
+      "grad_norm": 1.6328125,
+      "learning_rate": 1.473923185871419e-05,
+      "loss": 0.8459,
+      "step": 6064
+    },
+    {
+      "epoch": 1.0279486465053476,
+      "grad_norm": 1.6953125,
+      "learning_rate": 1.4737641542149184e-05,
+      "loss": 0.867,
+      "step": 6065
+    },
+    {
+      "epoch": 1.0281201105943374,
+      "grad_norm": 1.6875,
+      "learning_rate": 1.4736051071071848e-05,
+      "loss": 0.8569,
+      "step": 6066
+    },
+    {
+      "epoch": 1.0282915746833272,
+      "grad_norm": 1.6640625,
+      "learning_rate": 1.4734460445534053e-05,
+      "loss": 0.8905,
+      "step": 6067
+    },
+    {
+      "epoch": 1.028463038772317,
+      "grad_norm": 1.5546875,
+      "learning_rate": 1.4732869665587673e-05,
+      "loss": 0.8276,
+      "step": 6068
+    },
+    {
+      "epoch": 1.028634502861307,
+      "grad_norm": 1.7265625,
+      "learning_rate": 1.4731278731284591e-05,
+      "loss": 0.904,
+      "step": 6069
+    },
+    {
+      "epoch": 1.028805966950297,
+      "grad_norm": 1.6171875,
+      "learning_rate": 1.4729687642676693e-05,
+      "loss": 0.8826,
+      "step": 6070
+    },
+    {
+      "epoch": 1.0289774310392867,
+      "grad_norm": 1.6875,
+      "learning_rate": 1.4728096399815873e-05,
+      "loss": 0.8896,
+      "step": 6071
+    },
+    {
+      "epoch": 1.0291488951282766,
+      "grad_norm": 1.8046875,
+      "learning_rate": 1.4726505002754023e-05,
+      "loss": 0.8389,
+      "step": 6072
+    },
+    {
+      "epoch": 1.0293203592172664,
+      "grad_norm": 1.6171875,
+      "learning_rate": 1.4724913451543048e-05,
+      "loss": 0.8424,
+      "step": 6073
+    },
+    {
+      "epoch": 1.0294918233062562,
+      "grad_norm": 1.7421875,
+      "learning_rate": 1.4723321746234852e-05,
+      "loss": 0.9061,
+      "step": 6074
+    },
+    {
+      "epoch": 1.0296632873952463,
+      "grad_norm": 1.59375,
+      "learning_rate": 1.4721729886881346e-05,
+      "loss": 0.8773,
+      "step": 6075
+    },
+    {
+      "epoch": 1.029834751484236,
+      "grad_norm": 1.640625,
+      "learning_rate": 1.4720137873534451e-05,
+      "loss": 0.8438,
+      "step": 6076
+    },
+    {
+      "epoch": 1.030006215573226,
+      "grad_norm": 1.6640625,
+      "learning_rate": 1.4718545706246083e-05,
+      "loss": 0.8368,
+      "step": 6077
+    },
+    {
+      "epoch": 1.0301776796622157,
+      "grad_norm": 1.6171875,
+      "learning_rate": 1.471695338506817e-05,
+      "loss": 0.9028,
+      "step": 6078
+    },
+    {
+      "epoch": 1.0303491437512056,
+      "grad_norm": 1.625,
+      "learning_rate": 1.4715360910052648e-05,
+      "loss": 0.8155,
+      "step": 6079
+    },
+    {
+      "epoch": 1.0305206078401954,
+      "grad_norm": 1.6640625,
+      "learning_rate": 1.4713768281251448e-05,
+      "loss": 0.9443,
+      "step": 6080
+    },
+    {
+      "epoch": 1.0306920719291854,
+      "grad_norm": 1.609375,
+      "learning_rate": 1.4712175498716517e-05,
+      "loss": 0.8505,
+      "step": 6081
+    },
+    {
+      "epoch": 1.0308635360181753,
+      "grad_norm": 1.5625,
+      "learning_rate": 1.4710582562499796e-05,
+      "loss": 0.8387,
+      "step": 6082
+    },
+    {
+      "epoch": 1.031035000107165,
+      "grad_norm": 1.5703125,
+      "learning_rate": 1.4708989472653239e-05,
+      "loss": 0.7899,
+      "step": 6083
+    },
+    {
+      "epoch": 1.031206464196155,
+      "grad_norm": 1.546875,
+      "learning_rate": 1.4707396229228803e-05,
+      "loss": 0.8112,
+      "step": 6084
+    },
+    {
+      "epoch": 1.0313779282851447,
+      "grad_norm": 1.625,
+      "learning_rate": 1.4705802832278453e-05,
+      "loss": 0.9203,
+      "step": 6085
+    },
+    {
+      "epoch": 1.0315493923741346,
+      "grad_norm": 1.625,
+      "learning_rate": 1.4704209281854149e-05,
+      "loss": 0.8632,
+      "step": 6086
+    },
+    {
+      "epoch": 1.0317208564631244,
+      "grad_norm": 1.703125,
+      "learning_rate": 1.4702615578007863e-05,
+      "loss": 0.9001,
+      "step": 6087
+    },
+    {
+      "epoch": 1.0318923205521144,
+      "grad_norm": 1.625,
+      "learning_rate": 1.470102172079158e-05,
+      "loss": 0.7713,
+      "step": 6088
+    },
+    {
+      "epoch": 1.0320637846411043,
+      "grad_norm": 1.625,
+      "learning_rate": 1.4699427710257275e-05,
+      "loss": 0.8201,
+      "step": 6089
+    },
+    {
+      "epoch": 1.032235248730094,
+      "grad_norm": 1.671875,
+      "learning_rate": 1.4697833546456935e-05,
+      "loss": 0.915,
+      "step": 6090
+    },
+    {
+      "epoch": 1.032406712819084,
+      "grad_norm": 1.6015625,
+      "learning_rate": 1.4696239229442553e-05,
+      "loss": 0.8307,
+      "step": 6091
+    },
+    {
+      "epoch": 1.0325781769080737,
+      "grad_norm": 1.6171875,
+      "learning_rate": 1.4694644759266126e-05,
+      "loss": 0.8525,
+      "step": 6092
+    },
+    {
+      "epoch": 1.0327496409970638,
+      "grad_norm": 1.625,
+      "learning_rate": 1.4693050135979654e-05,
+      "loss": 0.784,
+      "step": 6093
+    },
+    {
+      "epoch": 1.0329211050860536,
+      "grad_norm": 1.578125,
+      "learning_rate": 1.4691455359635143e-05,
+      "loss": 0.8912,
+      "step": 6094
+    },
+    {
+      "epoch": 1.0330925691750434,
+      "grad_norm": 1.6796875,
+      "learning_rate": 1.4689860430284608e-05,
+      "loss": 0.8415,
+      "step": 6095
+    },
+    {
+      "epoch": 1.0332640332640333,
+      "grad_norm": 1.640625,
+      "learning_rate": 1.4688265347980065e-05,
+      "loss": 0.9884,
+      "step": 6096
+    },
+    {
+      "epoch": 1.033435497353023,
+      "grad_norm": 1.609375,
+      "learning_rate": 1.4686670112773534e-05,
+      "loss": 0.8707,
+      "step": 6097
+    },
+    {
+      "epoch": 1.033606961442013,
+      "grad_norm": 1.4765625,
+      "learning_rate": 1.4685074724717041e-05,
+      "loss": 0.7489,
+      "step": 6098
+    },
+    {
+      "epoch": 1.0337784255310027,
+      "grad_norm": 1.640625,
+      "learning_rate": 1.468347918386262e-05,
+      "loss": 0.866,
+      "step": 6099
+    },
+    {
+      "epoch": 1.0339498896199928,
+      "grad_norm": 1.6328125,
+      "learning_rate": 1.4681883490262308e-05,
+      "loss": 0.8207,
+      "step": 6100
+    },
+    {
+      "epoch": 1.0341213537089826,
+      "grad_norm": 1.609375,
+      "learning_rate": 1.4680287643968143e-05,
+      "loss": 0.7684,
+      "step": 6101
+    },
+    {
+      "epoch": 1.0342928177979724,
+      "grad_norm": 1.7890625,
+      "learning_rate": 1.4678691645032175e-05,
+      "loss": 0.9001,
+      "step": 6102
+    },
+    {
+      "epoch": 1.0344642818869623,
+      "grad_norm": 1.6484375,
+      "learning_rate": 1.4677095493506452e-05,
+      "loss": 0.9167,
+      "step": 6103
+    },
+    {
+      "epoch": 1.034635745975952,
+      "grad_norm": 1.640625,
+      "learning_rate": 1.4675499189443035e-05,
+      "loss": 0.8546,
+      "step": 6104
+    },
+    {
+      "epoch": 1.034807210064942,
+      "grad_norm": 1.6015625,
+      "learning_rate": 1.4673902732893981e-05,
+      "loss": 0.924,
+      "step": 6105
+    },
+    {
+      "epoch": 1.034978674153932,
+      "grad_norm": 1.703125,
+      "learning_rate": 1.4672306123911359e-05,
+      "loss": 0.8848,
+      "step": 6106
+    },
+    {
+      "epoch": 1.0351501382429218,
+      "grad_norm": 1.6640625,
+      "learning_rate": 1.467070936254724e-05,
+      "loss": 0.8747,
+      "step": 6107
+    },
+    {
+      "epoch": 1.0353216023319116,
+      "grad_norm": 1.65625,
+      "learning_rate": 1.4669112448853702e-05,
+      "loss": 0.8458,
+      "step": 6108
+    },
+    {
+      "epoch": 1.0354930664209014,
+      "grad_norm": 1.6015625,
+      "learning_rate": 1.4667515382882825e-05,
+      "loss": 0.899,
+      "step": 6109
+    },
+    {
+      "epoch": 1.0356645305098913,
+      "grad_norm": 1.5078125,
+      "learning_rate": 1.4665918164686692e-05,
+      "loss": 0.8748,
+      "step": 6110
+    },
+    {
+      "epoch": 1.035835994598881,
+      "grad_norm": 1.6875,
+      "learning_rate": 1.4664320794317399e-05,
+      "loss": 0.8575,
+      "step": 6111
+    },
+    {
+      "epoch": 1.0360074586878711,
+      "grad_norm": 1.6875,
+      "learning_rate": 1.4662723271827043e-05,
+      "loss": 0.8943,
+      "step": 6112
+    },
+    {
+      "epoch": 1.036178922776861,
+      "grad_norm": 1.640625,
+      "learning_rate": 1.466112559726772e-05,
+      "loss": 0.8982,
+      "step": 6113
+    },
+    {
+      "epoch": 1.0363503868658508,
+      "grad_norm": 1.703125,
+      "learning_rate": 1.4659527770691541e-05,
+      "loss": 0.9443,
+      "step": 6114
+    },
+    {
+      "epoch": 1.0365218509548406,
+      "grad_norm": 1.71875,
+      "learning_rate": 1.4657929792150611e-05,
+      "loss": 0.8187,
+      "step": 6115
+    },
+    {
+      "epoch": 1.0366933150438304,
+      "grad_norm": 1.6015625,
+      "learning_rate": 1.4656331661697055e-05,
+      "loss": 0.8568,
+      "step": 6116
+    },
+    {
+      "epoch": 1.0368647791328203,
+      "grad_norm": 1.5703125,
+      "learning_rate": 1.4654733379382986e-05,
+      "loss": 0.8306,
+      "step": 6117
+    },
+    {
+      "epoch": 1.0370362432218103,
+      "grad_norm": 1.6953125,
+      "learning_rate": 1.4653134945260536e-05,
+      "loss": 0.9314,
+      "step": 6118
+    },
+    {
+      "epoch": 1.0372077073108001,
+      "grad_norm": 1.71875,
+      "learning_rate": 1.4651536359381835e-05,
+      "loss": 0.8781,
+      "step": 6119
+    },
+    {
+      "epoch": 1.03737917139979,
+      "grad_norm": 1.7109375,
+      "learning_rate": 1.4649937621799016e-05,
+      "loss": 0.9768,
+      "step": 6120
+    },
+    {
+      "epoch": 1.0375506354887798,
+      "grad_norm": 1.59375,
+      "learning_rate": 1.464833873256422e-05,
+      "loss": 0.7535,
+      "step": 6121
+    },
+    {
+      "epoch": 1.0377220995777696,
+      "grad_norm": 1.546875,
+      "learning_rate": 1.4646739691729592e-05,
+      "loss": 0.7748,
+      "step": 6122
+    },
+    {
+      "epoch": 1.0378935636667594,
+      "grad_norm": 1.671875,
+      "learning_rate": 1.4645140499347287e-05,
+      "loss": 0.8655,
+      "step": 6123
+    },
+    {
+      "epoch": 1.0380650277557495,
+      "grad_norm": 1.5625,
+      "learning_rate": 1.4643541155469459e-05,
+      "loss": 0.8241,
+      "step": 6124
+    },
+    {
+      "epoch": 1.0382364918447393,
+      "grad_norm": 1.515625,
+      "learning_rate": 1.4641941660148267e-05,
+      "loss": 0.8163,
+      "step": 6125
+    },
+    {
+      "epoch": 1.0384079559337291,
+      "grad_norm": 1.71875,
+      "learning_rate": 1.4640342013435875e-05,
+      "loss": 0.9213,
+      "step": 6126
+    },
+    {
+      "epoch": 1.038579420022719,
+      "grad_norm": 1.625,
+      "learning_rate": 1.463874221538446e-05,
+      "loss": 0.7863,
+      "step": 6127
+    },
+    {
+      "epoch": 1.0387508841117088,
+      "grad_norm": 1.7890625,
+      "learning_rate": 1.4637142266046192e-05,
+      "loss": 0.8414,
+      "step": 6128
+    },
+    {
+      "epoch": 1.0389223482006986,
+      "grad_norm": 1.6640625,
+      "learning_rate": 1.4635542165473254e-05,
+      "loss": 0.9,
+      "step": 6129
+    },
+    {
+      "epoch": 1.0390938122896887,
+      "grad_norm": 1.703125,
+      "learning_rate": 1.4633941913717826e-05,
+      "loss": 0.8092,
+      "step": 6130
+    },
+    {
+      "epoch": 1.0392652763786785,
+      "grad_norm": 1.6484375,
+      "learning_rate": 1.4632341510832109e-05,
+      "loss": 0.9552,
+      "step": 6131
+    },
+    {
+      "epoch": 1.0394367404676683,
+      "grad_norm": 1.6015625,
+      "learning_rate": 1.4630740956868289e-05,
+      "loss": 0.7918,
+      "step": 6132
+    },
+    {
+      "epoch": 1.0396082045566581,
+      "grad_norm": 1.640625,
+      "learning_rate": 1.4629140251878567e-05,
+      "loss": 0.8943,
+      "step": 6133
+    },
+    {
+      "epoch": 1.039779668645648,
+      "grad_norm": 1.6640625,
+      "learning_rate": 1.4627539395915146e-05,
+      "loss": 0.8804,
+      "step": 6134
+    },
+    {
+      "epoch": 1.0399511327346378,
+      "grad_norm": 1.7578125,
+      "learning_rate": 1.4625938389030246e-05,
+      "loss": 0.8774,
+      "step": 6135
+    },
+    {
+      "epoch": 1.0401225968236278,
+      "grad_norm": 1.6171875,
+      "learning_rate": 1.4624337231276071e-05,
+      "loss": 0.7932,
+      "step": 6136
+    },
+    {
+      "epoch": 1.0402940609126177,
+      "grad_norm": 1.59375,
+      "learning_rate": 1.4622735922704849e-05,
+      "loss": 0.8117,
+      "step": 6137
+    },
+    {
+      "epoch": 1.0404655250016075,
+      "grad_norm": 1.703125,
+      "learning_rate": 1.46211344633688e-05,
+      "loss": 0.9195,
+      "step": 6138
+    },
+    {
+      "epoch": 1.0406369890905973,
+      "grad_norm": 1.734375,
+      "learning_rate": 1.4619532853320153e-05,
+      "loss": 0.8734,
+      "step": 6139
+    },
+    {
+      "epoch": 1.0408084531795871,
+      "grad_norm": 1.5859375,
+      "learning_rate": 1.4617931092611146e-05,
+      "loss": 0.7903,
+      "step": 6140
+    },
+    {
+      "epoch": 1.040979917268577,
+      "grad_norm": 1.8359375,
+      "learning_rate": 1.4616329181294016e-05,
+      "loss": 0.8192,
+      "step": 6141
+    },
+    {
+      "epoch": 1.041151381357567,
+      "grad_norm": 1.609375,
+      "learning_rate": 1.4614727119421007e-05,
+      "loss": 0.8648,
+      "step": 6142
+    },
+    {
+      "epoch": 1.0413228454465568,
+      "grad_norm": 1.7265625,
+      "learning_rate": 1.4613124907044369e-05,
+      "loss": 0.8881,
+      "step": 6143
+    },
+    {
+      "epoch": 1.0414943095355467,
+      "grad_norm": 1.71875,
+      "learning_rate": 1.4611522544216357e-05,
+      "loss": 0.8722,
+      "step": 6144
+    },
+    {
+      "epoch": 1.0416657736245365,
+      "grad_norm": 1.59375,
+      "learning_rate": 1.460992003098923e-05,
+      "loss": 0.81,
+      "step": 6145
+    },
+    {
+      "epoch": 1.0418372377135263,
+      "grad_norm": 1.6484375,
+      "learning_rate": 1.460831736741525e-05,
+      "loss": 0.8484,
+      "step": 6146
+    },
+    {
+      "epoch": 1.0420087018025161,
+      "grad_norm": 1.6953125,
+      "learning_rate": 1.4606714553546689e-05,
+      "loss": 0.8397,
+      "step": 6147
+    },
+    {
+      "epoch": 1.0421801658915062,
+      "grad_norm": 1.671875,
+      "learning_rate": 1.460511158943582e-05,
+      "loss": 0.8013,
+      "step": 6148
+    },
+    {
+      "epoch": 1.042351629980496,
+      "grad_norm": 1.734375,
+      "learning_rate": 1.460350847513492e-05,
+      "loss": 0.9249,
+      "step": 6149
+    },
+    {
+      "epoch": 1.0425230940694858,
+      "grad_norm": 1.5625,
+      "learning_rate": 1.4601905210696273e-05,
+      "loss": 0.843,
+      "step": 6150
+    },
+    {
+      "epoch": 1.0426945581584757,
+      "grad_norm": 1.8984375,
+      "learning_rate": 1.4600301796172167e-05,
+      "loss": 0.9169,
+      "step": 6151
+    },
+    {
+      "epoch": 1.0428660222474655,
+      "grad_norm": 1.6640625,
+      "learning_rate": 1.4598698231614896e-05,
+      "loss": 0.8661,
+      "step": 6152
+    },
+    {
+      "epoch": 1.0430374863364553,
+      "grad_norm": 1.625,
+      "learning_rate": 1.459709451707676e-05,
+      "loss": 0.8222,
+      "step": 6153
+    },
+    {
+      "epoch": 1.0432089504254454,
+      "grad_norm": 1.578125,
+      "learning_rate": 1.459549065261006e-05,
+      "loss": 0.8195,
+      "step": 6154
+    },
+    {
+      "epoch": 1.0433804145144352,
+      "grad_norm": 1.6328125,
+      "learning_rate": 1.4593886638267104e-05,
+      "loss": 0.9094,
+      "step": 6155
+    },
+    {
+      "epoch": 1.043551878603425,
+      "grad_norm": 1.5859375,
+      "learning_rate": 1.4592282474100207e-05,
+      "loss": 0.8695,
+      "step": 6156
+    },
+    {
+      "epoch": 1.0437233426924148,
+      "grad_norm": 1.609375,
+      "learning_rate": 1.4590678160161686e-05,
+      "loss": 0.8661,
+      "step": 6157
+    },
+    {
+      "epoch": 1.0438948067814047,
+      "grad_norm": 1.5625,
+      "learning_rate": 1.4589073696503864e-05,
+      "loss": 0.8281,
+      "step": 6158
+    },
+    {
+      "epoch": 1.0440662708703945,
+      "grad_norm": 1.75,
+      "learning_rate": 1.4587469083179065e-05,
+      "loss": 0.9914,
+      "step": 6159
+    },
+    {
+      "epoch": 1.0442377349593845,
+      "grad_norm": 1.6796875,
+      "learning_rate": 1.4585864320239629e-05,
+      "loss": 0.8431,
+      "step": 6160
+    },
+    {
+      "epoch": 1.0444091990483744,
+      "grad_norm": 1.6953125,
+      "learning_rate": 1.4584259407737884e-05,
+      "loss": 0.8542,
+      "step": 6161
+    },
+    {
+      "epoch": 1.0445806631373642,
+      "grad_norm": 1.609375,
+      "learning_rate": 1.4582654345726177e-05,
+      "loss": 0.8523,
+      "step": 6162
+    },
+    {
+      "epoch": 1.044752127226354,
+      "grad_norm": 1.625,
+      "learning_rate": 1.4581049134256857e-05,
+      "loss": 0.9015,
+      "step": 6163
+    },
+    {
+      "epoch": 1.0449235913153438,
+      "grad_norm": 1.640625,
+      "learning_rate": 1.4579443773382273e-05,
+      "loss": 0.8365,
+      "step": 6164
+    },
+    {
+      "epoch": 1.0450950554043337,
+      "grad_norm": 1.609375,
+      "learning_rate": 1.4577838263154787e-05,
+      "loss": 0.8479,
+      "step": 6165
+    },
+    {
+      "epoch": 1.0452665194933237,
+      "grad_norm": 1.65625,
+      "learning_rate": 1.4576232603626754e-05,
+      "loss": 0.8198,
+      "step": 6166
+    },
+    {
+      "epoch": 1.0454379835823135,
+      "grad_norm": 1.65625,
+      "learning_rate": 1.4574626794850541e-05,
+      "loss": 0.8248,
+      "step": 6167
+    },
+    {
+      "epoch": 1.0456094476713034,
+      "grad_norm": 1.625,
+      "learning_rate": 1.4573020836878524e-05,
+      "loss": 0.8216,
+      "step": 6168
+    },
+    {
+      "epoch": 1.0457809117602932,
+      "grad_norm": 1.609375,
+      "learning_rate": 1.4571414729763076e-05,
+      "loss": 0.8822,
+      "step": 6169
+    },
+    {
+      "epoch": 1.045952375849283,
+      "grad_norm": 1.6484375,
+      "learning_rate": 1.456980847355658e-05,
+      "loss": 0.9406,
+      "step": 6170
+    },
+    {
+      "epoch": 1.0461238399382728,
+      "grad_norm": 1.5234375,
+      "learning_rate": 1.4568202068311421e-05,
+      "loss": 0.7995,
+      "step": 6171
+    },
+    {
+      "epoch": 1.046295304027263,
+      "grad_norm": 1.6640625,
+      "learning_rate": 1.4566595514079991e-05,
+      "loss": 0.7696,
+      "step": 6172
+    },
+    {
+      "epoch": 1.0464667681162527,
+      "grad_norm": 1.515625,
+      "learning_rate": 1.4564988810914686e-05,
+      "loss": 0.8146,
+      "step": 6173
+    },
+    {
+      "epoch": 1.0466382322052425,
+      "grad_norm": 1.6171875,
+      "learning_rate": 1.4563381958867903e-05,
+      "loss": 0.8455,
+      "step": 6174
+    },
+    {
+      "epoch": 1.0468096962942324,
+      "grad_norm": 1.59375,
+      "learning_rate": 1.4561774957992052e-05,
+      "loss": 0.8198,
+      "step": 6175
+    },
+    {
+      "epoch": 1.0469811603832222,
+      "grad_norm": 1.609375,
+      "learning_rate": 1.456016780833954e-05,
+      "loss": 0.8871,
+      "step": 6176
+    },
+    {
+      "epoch": 1.047152624472212,
+      "grad_norm": 1.6640625,
+      "learning_rate": 1.455856050996279e-05,
+      "loss": 0.8185,
+      "step": 6177
+    },
+    {
+      "epoch": 1.047324088561202,
+      "grad_norm": 1.546875,
+      "learning_rate": 1.455695306291421e-05,
+      "loss": 0.8382,
+      "step": 6178
+    },
+    {
+      "epoch": 1.047495552650192,
+      "grad_norm": 1.625,
+      "learning_rate": 1.4555345467246231e-05,
+      "loss": 0.9098,
+      "step": 6179
+    },
+    {
+      "epoch": 1.0476670167391817,
+      "grad_norm": 1.609375,
+      "learning_rate": 1.4553737723011283e-05,
+      "loss": 0.89,
+      "step": 6180
+    },
+    {
+      "epoch": 1.0478384808281715,
+      "grad_norm": 1.671875,
+      "learning_rate": 1.4552129830261797e-05,
+      "loss": 0.8876,
+      "step": 6181
+    },
+    {
+      "epoch": 1.0480099449171614,
+      "grad_norm": 1.671875,
+      "learning_rate": 1.4550521789050218e-05,
+      "loss": 0.9873,
+      "step": 6182
+    },
+    {
+      "epoch": 1.0481814090061512,
+      "grad_norm": 1.6015625,
+      "learning_rate": 1.454891359942899e-05,
+      "loss": 0.8154,
+      "step": 6183
+    },
+    {
+      "epoch": 1.048352873095141,
+      "grad_norm": 1.59375,
+      "learning_rate": 1.4547305261450559e-05,
+      "loss": 0.9423,
+      "step": 6184
+    },
+    {
+      "epoch": 1.048524337184131,
+      "grad_norm": 1.671875,
+      "learning_rate": 1.454569677516738e-05,
+      "loss": 0.8704,
+      "step": 6185
+    },
+    {
+      "epoch": 1.048695801273121,
+      "grad_norm": 1.6796875,
+      "learning_rate": 1.4544088140631912e-05,
+      "loss": 0.8998,
+      "step": 6186
+    },
+    {
+      "epoch": 1.0488672653621107,
+      "grad_norm": 1.6328125,
+      "learning_rate": 1.4542479357896615e-05,
+      "loss": 0.9563,
+      "step": 6187
+    },
+    {
+      "epoch": 1.0490387294511005,
+      "grad_norm": 1.6015625,
+      "learning_rate": 1.4540870427013962e-05,
+      "loss": 0.8739,
+      "step": 6188
+    },
+    {
+      "epoch": 1.0492101935400904,
+      "grad_norm": 1.6953125,
+      "learning_rate": 1.4539261348036426e-05,
+      "loss": 0.8912,
+      "step": 6189
+    },
+    {
+      "epoch": 1.0493816576290804,
+      "grad_norm": 1.671875,
+      "learning_rate": 1.4537652121016485e-05,
+      "loss": 0.8597,
+      "step": 6190
+    },
+    {
+      "epoch": 1.0495531217180702,
+      "grad_norm": 1.671875,
+      "learning_rate": 1.453604274600662e-05,
+      "loss": 0.9169,
+      "step": 6191
+    },
+    {
+      "epoch": 1.04972458580706,
+      "grad_norm": 1.796875,
+      "learning_rate": 1.453443322305932e-05,
+      "loss": 0.8305,
+      "step": 6192
+    },
+    {
+      "epoch": 1.04989604989605,
+      "grad_norm": 1.6015625,
+      "learning_rate": 1.4532823552227078e-05,
+      "loss": 0.832,
+      "step": 6193
+    },
+    {
+      "epoch": 1.0500675139850397,
+      "grad_norm": 1.7890625,
+      "learning_rate": 1.4531213733562392e-05,
+      "loss": 0.9607,
+      "step": 6194
+    },
+    {
+      "epoch": 1.0502389780740295,
+      "grad_norm": 1.6640625,
+      "learning_rate": 1.4529603767117761e-05,
+      "loss": 0.8243,
+      "step": 6195
+    },
+    {
+      "epoch": 1.0504104421630194,
+      "grad_norm": 1.625,
+      "learning_rate": 1.4527993652945696e-05,
+      "loss": 0.8031,
+      "step": 6196
+    },
+    {
+      "epoch": 1.0505819062520094,
+      "grad_norm": 1.625,
+      "learning_rate": 1.4526383391098704e-05,
+      "loss": 0.9309,
+      "step": 6197
+    },
+    {
+      "epoch": 1.0507533703409992,
+      "grad_norm": 1.640625,
+      "learning_rate": 1.4524772981629309e-05,
+      "loss": 0.8646,
+      "step": 6198
+    },
+    {
+      "epoch": 1.050924834429989,
+      "grad_norm": 1.734375,
+      "learning_rate": 1.4523162424590025e-05,
+      "loss": 0.9192,
+      "step": 6199
+    },
+    {
+      "epoch": 1.051096298518979,
+      "grad_norm": 1.5,
+      "learning_rate": 1.4521551720033382e-05,
+      "loss": 0.8055,
+      "step": 6200
+    },
+    {
+      "epoch": 1.0512677626079687,
+      "grad_norm": 1.6640625,
+      "learning_rate": 1.4519940868011913e-05,
+      "loss": 0.8371,
+      "step": 6201
+    },
+    {
+      "epoch": 1.0514392266969586,
+      "grad_norm": 1.546875,
+      "learning_rate": 1.4518329868578149e-05,
+      "loss": 0.8334,
+      "step": 6202
+    },
+    {
+      "epoch": 1.0516106907859486,
+      "grad_norm": 1.671875,
+      "learning_rate": 1.4516718721784635e-05,
+      "loss": 0.8649,
+      "step": 6203
+    },
+    {
+      "epoch": 1.0517821548749384,
+      "grad_norm": 1.703125,
+      "learning_rate": 1.4515107427683917e-05,
+      "loss": 0.9058,
+      "step": 6204
+    },
+    {
+      "epoch": 1.0519536189639282,
+      "grad_norm": 1.6328125,
+      "learning_rate": 1.4513495986328541e-05,
+      "loss": 0.8435,
+      "step": 6205
+    },
+    {
+      "epoch": 1.052125083052918,
+      "grad_norm": 1.734375,
+      "learning_rate": 1.4511884397771065e-05,
+      "loss": 0.846,
+      "step": 6206
+    },
+    {
+      "epoch": 1.052296547141908,
+      "grad_norm": 1.6875,
+      "learning_rate": 1.451027266206405e-05,
+      "loss": 0.8909,
+      "step": 6207
+    },
+    {
+      "epoch": 1.0524680112308977,
+      "grad_norm": 1.6328125,
+      "learning_rate": 1.4508660779260057e-05,
+      "loss": 0.9091,
+      "step": 6208
+    },
+    {
+      "epoch": 1.0526394753198878,
+      "grad_norm": 1.5859375,
+      "learning_rate": 1.4507048749411658e-05,
+      "loss": 0.8144,
+      "step": 6209
+    },
+    {
+      "epoch": 1.0528109394088776,
+      "grad_norm": 1.71875,
+      "learning_rate": 1.4505436572571428e-05,
+      "loss": 0.8372,
+      "step": 6210
+    },
+    {
+      "epoch": 1.0529824034978674,
+      "grad_norm": 1.5703125,
+      "learning_rate": 1.4503824248791946e-05,
+      "loss": 0.8087,
+      "step": 6211
+    },
+    {
+      "epoch": 1.0531538675868573,
+      "grad_norm": 1.671875,
+      "learning_rate": 1.4502211778125799e-05,
+      "loss": 0.9126,
+      "step": 6212
+    },
+    {
+      "epoch": 1.053325331675847,
+      "grad_norm": 1.7421875,
+      "learning_rate": 1.4500599160625565e-05,
+      "loss": 0.8595,
+      "step": 6213
+    },
+    {
+      "epoch": 1.053496795764837,
+      "grad_norm": 1.578125,
+      "learning_rate": 1.4498986396343851e-05,
+      "loss": 0.8395,
+      "step": 6214
+    },
+    {
+      "epoch": 1.053668259853827,
+      "grad_norm": 1.6015625,
+      "learning_rate": 1.4497373485333245e-05,
+      "loss": 0.8365,
+      "step": 6215
+    },
+    {
+      "epoch": 1.0538397239428168,
+      "grad_norm": 1.703125,
+      "learning_rate": 1.4495760427646355e-05,
+      "loss": 0.9233,
+      "step": 6216
+    },
+    {
+      "epoch": 1.0540111880318066,
+      "grad_norm": 1.640625,
+      "learning_rate": 1.4494147223335789e-05,
+      "loss": 0.7995,
+      "step": 6217
+    },
+    {
+      "epoch": 1.0541826521207964,
+      "grad_norm": 1.8125,
+      "learning_rate": 1.4492533872454157e-05,
+      "loss": 0.9579,
+      "step": 6218
+    },
+    {
+      "epoch": 1.0543541162097863,
+      "grad_norm": 1.7734375,
+      "learning_rate": 1.449092037505408e-05,
+      "loss": 0.9061,
+      "step": 6219
+    },
+    {
+      "epoch": 1.054525580298776,
+      "grad_norm": 1.78125,
+      "learning_rate": 1.4489306731188178e-05,
+      "loss": 0.8286,
+      "step": 6220
+    },
+    {
+      "epoch": 1.0546970443877661,
+      "grad_norm": 1.734375,
+      "learning_rate": 1.4487692940909077e-05,
+      "loss": 0.9238,
+      "step": 6221
+    },
+    {
+      "epoch": 1.054868508476756,
+      "grad_norm": 1.7265625,
+      "learning_rate": 1.4486079004269411e-05,
+      "loss": 0.8596,
+      "step": 6222
+    },
+    {
+      "epoch": 1.0550399725657458,
+      "grad_norm": 1.6796875,
+      "learning_rate": 1.4484464921321818e-05,
+      "loss": 0.8859,
+      "step": 6223
+    },
+    {
+      "epoch": 1.0552114366547356,
+      "grad_norm": 1.640625,
+      "learning_rate": 1.4482850692118932e-05,
+      "loss": 0.8388,
+      "step": 6224
+    },
+    {
+      "epoch": 1.0553829007437254,
+      "grad_norm": 1.625,
+      "learning_rate": 1.4481236316713408e-05,
+      "loss": 0.8139,
+      "step": 6225
+    },
+    {
+      "epoch": 1.0555543648327153,
+      "grad_norm": 1.7265625,
+      "learning_rate": 1.447962179515789e-05,
+      "loss": 0.9119,
+      "step": 6226
+    },
+    {
+      "epoch": 1.0557258289217053,
+      "grad_norm": 1.6875,
+      "learning_rate": 1.4478007127505039e-05,
+      "loss": 0.8191,
+      "step": 6227
+    },
+    {
+      "epoch": 1.0558972930106951,
+      "grad_norm": 1.703125,
+      "learning_rate": 1.4476392313807508e-05,
+      "loss": 0.8268,
+      "step": 6228
+    },
+    {
+      "epoch": 1.056068757099685,
+      "grad_norm": 1.703125,
+      "learning_rate": 1.4474777354117974e-05,
+      "loss": 0.8519,
+      "step": 6229
+    },
+    {
+      "epoch": 1.0562402211886748,
+      "grad_norm": 1.6015625,
+      "learning_rate": 1.4473162248489097e-05,
+      "loss": 0.8713,
+      "step": 6230
+    },
+    {
+      "epoch": 1.0564116852776646,
+      "grad_norm": 1.703125,
+      "learning_rate": 1.4471546996973555e-05,
+      "loss": 0.8951,
+      "step": 6231
+    },
+    {
+      "epoch": 1.0565831493666544,
+      "grad_norm": 1.6171875,
+      "learning_rate": 1.4469931599624027e-05,
+      "loss": 0.806,
+      "step": 6232
+    },
+    {
+      "epoch": 1.0567546134556445,
+      "grad_norm": 1.6328125,
+      "learning_rate": 1.4468316056493197e-05,
+      "loss": 0.8072,
+      "step": 6233
+    },
+    {
+      "epoch": 1.0569260775446343,
+      "grad_norm": 1.6875,
+      "learning_rate": 1.4466700367633754e-05,
+      "loss": 0.8183,
+      "step": 6234
+    },
+    {
+      "epoch": 1.0570975416336241,
+      "grad_norm": 1.6328125,
+      "learning_rate": 1.4465084533098394e-05,
+      "loss": 0.7764,
+      "step": 6235
+    },
+    {
+      "epoch": 1.057269005722614,
+      "grad_norm": 1.6875,
+      "learning_rate": 1.446346855293981e-05,
+      "loss": 0.9423,
+      "step": 6236
+    },
+    {
+      "epoch": 1.0574404698116038,
+      "grad_norm": 1.5078125,
+      "learning_rate": 1.4461852427210711e-05,
+      "loss": 0.7811,
+      "step": 6237
+    },
+    {
+      "epoch": 1.0576119339005936,
+      "grad_norm": 1.6953125,
+      "learning_rate": 1.4460236155963803e-05,
+      "loss": 0.8399,
+      "step": 6238
+    },
+    {
+      "epoch": 1.0577833979895837,
+      "grad_norm": 1.6953125,
+      "learning_rate": 1.4458619739251795e-05,
+      "loss": 0.8339,
+      "step": 6239
+    },
+    {
+      "epoch": 1.0579548620785735,
+      "grad_norm": 1.59375,
+      "learning_rate": 1.4457003177127414e-05,
+      "loss": 0.8646,
+      "step": 6240
+    },
+    {
+      "epoch": 1.0581263261675633,
+      "grad_norm": 1.609375,
+      "learning_rate": 1.4455386469643371e-05,
+      "loss": 0.7481,
+      "step": 6241
+    },
+    {
+      "epoch": 1.0582977902565531,
+      "grad_norm": 1.6640625,
+      "learning_rate": 1.44537696168524e-05,
+      "loss": 0.8451,
+      "step": 6242
+    },
+    {
+      "epoch": 1.058469254345543,
+      "grad_norm": 1.609375,
+      "learning_rate": 1.4452152618807228e-05,
+      "loss": 0.8559,
+      "step": 6243
+    },
+    {
+      "epoch": 1.0586407184345328,
+      "grad_norm": 1.6171875,
+      "learning_rate": 1.4450535475560594e-05,
+      "loss": 0.8876,
+      "step": 6244
+    },
+    {
+      "epoch": 1.0588121825235228,
+      "grad_norm": 1.6171875,
+      "learning_rate": 1.4448918187165242e-05,
+      "loss": 0.8205,
+      "step": 6245
+    },
+    {
+      "epoch": 1.0589836466125127,
+      "grad_norm": 1.7109375,
+      "learning_rate": 1.4447300753673912e-05,
+      "loss": 0.8484,
+      "step": 6246
+    },
+    {
+      "epoch": 1.0591551107015025,
+      "grad_norm": 1.6015625,
+      "learning_rate": 1.4445683175139357e-05,
+      "loss": 0.8203,
+      "step": 6247
+    },
+    {
+      "epoch": 1.0593265747904923,
+      "grad_norm": 1.609375,
+      "learning_rate": 1.4444065451614336e-05,
+      "loss": 0.8436,
+      "step": 6248
+    },
+    {
+      "epoch": 1.0594980388794821,
+      "grad_norm": 1.65625,
+      "learning_rate": 1.4442447583151604e-05,
+      "loss": 0.8511,
+      "step": 6249
+    },
+    {
+      "epoch": 1.059669502968472,
+      "grad_norm": 1.6328125,
+      "learning_rate": 1.4440829569803927e-05,
+      "loss": 0.8101,
+      "step": 6250
+    },
+    {
+      "epoch": 1.059840967057462,
+      "grad_norm": 1.6328125,
+      "learning_rate": 1.4439211411624074e-05,
+      "loss": 0.8623,
+      "step": 6251
+    },
+    {
+      "epoch": 1.0600124311464518,
+      "grad_norm": 1.7421875,
+      "learning_rate": 1.4437593108664825e-05,
+      "loss": 0.8703,
+      "step": 6252
+    },
+    {
+      "epoch": 1.0601838952354417,
+      "grad_norm": 1.6328125,
+      "learning_rate": 1.4435974660978951e-05,
+      "loss": 0.8501,
+      "step": 6253
+    },
+    {
+      "epoch": 1.0603553593244315,
+      "grad_norm": 1.71875,
+      "learning_rate": 1.443435606861924e-05,
+      "loss": 0.9815,
+      "step": 6254
+    },
+    {
+      "epoch": 1.0605268234134213,
+      "grad_norm": 1.6015625,
+      "learning_rate": 1.4432737331638477e-05,
+      "loss": 0.8542,
+      "step": 6255
+    },
+    {
+      "epoch": 1.0606982875024111,
+      "grad_norm": 1.6171875,
+      "learning_rate": 1.4431118450089458e-05,
+      "loss": 0.8357,
+      "step": 6256
+    },
+    {
+      "epoch": 1.0608697515914012,
+      "grad_norm": 1.5546875,
+      "learning_rate": 1.4429499424024982e-05,
+      "loss": 0.8416,
+      "step": 6257
+    },
+    {
+      "epoch": 1.061041215680391,
+      "grad_norm": 1.625,
+      "learning_rate": 1.442788025349785e-05,
+      "loss": 0.8575,
+      "step": 6258
+    },
+    {
+      "epoch": 1.0612126797693808,
+      "grad_norm": 1.59375,
+      "learning_rate": 1.4426260938560868e-05,
+      "loss": 0.8339,
+      "step": 6259
+    },
+    {
+      "epoch": 1.0613841438583707,
+      "grad_norm": 1.75,
+      "learning_rate": 1.442464147926685e-05,
+      "loss": 0.8274,
+      "step": 6260
+    },
+    {
+      "epoch": 1.0615556079473605,
+      "grad_norm": 1.578125,
+      "learning_rate": 1.442302187566861e-05,
+      "loss": 0.796,
+      "step": 6261
+    },
+    {
+      "epoch": 1.0617270720363503,
+      "grad_norm": 1.6640625,
+      "learning_rate": 1.4421402127818974e-05,
+      "loss": 0.8236,
+      "step": 6262
+    },
+    {
+      "epoch": 1.0618985361253404,
+      "grad_norm": 1.6875,
+      "learning_rate": 1.4419782235770763e-05,
+      "loss": 0.8753,
+      "step": 6263
+    },
+    {
+      "epoch": 1.0620700002143302,
+      "grad_norm": 1.6953125,
+      "learning_rate": 1.4418162199576808e-05,
+      "loss": 0.9048,
+      "step": 6264
+    },
+    {
+      "epoch": 1.06224146430332,
+      "grad_norm": 1.703125,
+      "learning_rate": 1.441654201928995e-05,
+      "loss": 0.8329,
+      "step": 6265
+    },
+    {
+      "epoch": 1.0624129283923098,
+      "grad_norm": 1.6875,
+      "learning_rate": 1.4414921694963024e-05,
+      "loss": 0.8477,
+      "step": 6266
+    },
+    {
+      "epoch": 1.0625843924812997,
+      "grad_norm": 1.6171875,
+      "learning_rate": 1.4413301226648877e-05,
+      "loss": 0.8548,
+      "step": 6267
+    },
+    {
+      "epoch": 1.0627558565702895,
+      "grad_norm": 1.75,
+      "learning_rate": 1.4411680614400356e-05,
+      "loss": 0.861,
+      "step": 6268
+    },
+    {
+      "epoch": 1.0629273206592793,
+      "grad_norm": 1.625,
+      "learning_rate": 1.4410059858270322e-05,
+      "loss": 0.7642,
+      "step": 6269
+    },
+    {
+      "epoch": 1.0630987847482694,
+      "grad_norm": 1.53125,
+      "learning_rate": 1.4408438958311626e-05,
+      "loss": 0.7788,
+      "step": 6270
+    },
+    {
+      "epoch": 1.0632702488372592,
+      "grad_norm": 1.578125,
+      "learning_rate": 1.4406817914577135e-05,
+      "loss": 0.799,
+      "step": 6271
+    },
+    {
+      "epoch": 1.063441712926249,
+      "grad_norm": 1.6640625,
+      "learning_rate": 1.4405196727119717e-05,
+      "loss": 0.8563,
+      "step": 6272
+    },
+    {
+      "epoch": 1.0636131770152388,
+      "grad_norm": 1.578125,
+      "learning_rate": 1.4403575395992247e-05,
+      "loss": 0.7594,
+      "step": 6273
+    },
+    {
+      "epoch": 1.0637846411042287,
+      "grad_norm": 1.671875,
+      "learning_rate": 1.44019539212476e-05,
+      "loss": 0.8161,
+      "step": 6274
+    },
+    {
+      "epoch": 1.0639561051932187,
+      "grad_norm": 1.6640625,
+      "learning_rate": 1.4400332302938658e-05,
+      "loss": 0.8292,
+      "step": 6275
+    },
+    {
+      "epoch": 1.0641275692822085,
+      "grad_norm": 1.7109375,
+      "learning_rate": 1.439871054111831e-05,
+      "loss": 0.808,
+      "step": 6276
+    },
+    {
+      "epoch": 1.0642990333711984,
+      "grad_norm": 1.6171875,
+      "learning_rate": 1.439708863583945e-05,
+      "loss": 0.8312,
+      "step": 6277
+    },
+    {
+      "epoch": 1.0644704974601882,
+      "grad_norm": 1.6484375,
+      "learning_rate": 1.4395466587154969e-05,
+      "loss": 0.8997,
+      "step": 6278
+    },
+    {
+      "epoch": 1.064641961549178,
+      "grad_norm": 1.609375,
+      "learning_rate": 1.4393844395117771e-05,
+      "loss": 0.7794,
+      "step": 6279
+    },
+    {
+      "epoch": 1.0648134256381678,
+      "grad_norm": 1.6328125,
+      "learning_rate": 1.439222205978076e-05,
+      "loss": 0.9294,
+      "step": 6280
+    },
+    {
+      "epoch": 1.0649848897271577,
+      "grad_norm": 1.5625,
+      "learning_rate": 1.4390599581196854e-05,
+      "loss": 0.8536,
+      "step": 6281
+    },
+    {
+      "epoch": 1.0651563538161477,
+      "grad_norm": 1.7265625,
+      "learning_rate": 1.4388976959418956e-05,
+      "loss": 0.8836,
+      "step": 6282
+    },
+    {
+      "epoch": 1.0653278179051375,
+      "grad_norm": 1.6953125,
+      "learning_rate": 1.4387354194499994e-05,
+      "loss": 0.8734,
+      "step": 6283
+    },
+    {
+      "epoch": 1.0654992819941274,
+      "grad_norm": 1.5625,
+      "learning_rate": 1.438573128649289e-05,
+      "loss": 0.7896,
+      "step": 6284
+    },
+    {
+      "epoch": 1.0656707460831172,
+      "grad_norm": 1.734375,
+      "learning_rate": 1.4384108235450574e-05,
+      "loss": 0.9422,
+      "step": 6285
+    },
+    {
+      "epoch": 1.065842210172107,
+      "grad_norm": 1.6015625,
+      "learning_rate": 1.438248504142598e-05,
+      "loss": 0.8169,
+      "step": 6286
+    },
+    {
+      "epoch": 1.066013674261097,
+      "grad_norm": 1.7421875,
+      "learning_rate": 1.4380861704472047e-05,
+      "loss": 0.8534,
+      "step": 6287
+    },
+    {
+      "epoch": 1.0661851383500869,
+      "grad_norm": 1.6171875,
+      "learning_rate": 1.4379238224641716e-05,
+      "loss": 0.8926,
+      "step": 6288
+    },
+    {
+      "epoch": 1.0663566024390767,
+      "grad_norm": 1.5546875,
+      "learning_rate": 1.4377614601987934e-05,
+      "loss": 0.8646,
+      "step": 6289
+    },
+    {
+      "epoch": 1.0665280665280665,
+      "grad_norm": 1.546875,
+      "learning_rate": 1.4375990836563658e-05,
+      "loss": 0.8611,
+      "step": 6290
+    },
+    {
+      "epoch": 1.0666995306170564,
+      "grad_norm": 1.6875,
+      "learning_rate": 1.437436692842184e-05,
+      "loss": 0.7749,
+      "step": 6291
+    },
+    {
+      "epoch": 1.0668709947060462,
+      "grad_norm": 1.671875,
+      "learning_rate": 1.4372742877615447e-05,
+      "loss": 0.8207,
+      "step": 6292
+    },
+    {
+      "epoch": 1.067042458795036,
+      "grad_norm": 1.7265625,
+      "learning_rate": 1.437111868419744e-05,
+      "loss": 0.9227,
+      "step": 6293
+    },
+    {
+      "epoch": 1.067213922884026,
+      "grad_norm": 1.6015625,
+      "learning_rate": 1.4369494348220791e-05,
+      "loss": 0.8164,
+      "step": 6294
+    },
+    {
+      "epoch": 1.0673853869730159,
+      "grad_norm": 1.578125,
+      "learning_rate": 1.4367869869738482e-05,
+      "loss": 0.8078,
+      "step": 6295
+    },
+    {
+      "epoch": 1.0675568510620057,
+      "grad_norm": 1.703125,
+      "learning_rate": 1.4366245248803485e-05,
+      "loss": 0.8588,
+      "step": 6296
+    },
+    {
+      "epoch": 1.0677283151509955,
+      "grad_norm": 1.734375,
+      "learning_rate": 1.436462048546879e-05,
+      "loss": 0.9065,
+      "step": 6297
+    },
+    {
+      "epoch": 1.0678997792399854,
+      "grad_norm": 1.734375,
+      "learning_rate": 1.4362995579787389e-05,
+      "loss": 0.8279,
+      "step": 6298
+    },
+    {
+      "epoch": 1.0680712433289752,
+      "grad_norm": 1.6328125,
+      "learning_rate": 1.4361370531812266e-05,
+      "loss": 0.8336,
+      "step": 6299
+    },
+    {
+      "epoch": 1.0682427074179652,
+      "grad_norm": 1.7421875,
+      "learning_rate": 1.4359745341596431e-05,
+      "loss": 0.9098,
+      "step": 6300
+    },
+    {
+      "epoch": 1.0682427074179652,
+      "eval_loss": 0.8528460264205933,
+      "eval_runtime": 837.1005,
+      "eval_samples_per_second": 2.985,
+      "eval_steps_per_second": 2.985,
+      "step": 6300
+    },
+    {
+      "epoch": 1.068414171506955,
+      "grad_norm": 1.6640625,
+      "learning_rate": 1.4358120009192881e-05,
+      "loss": 0.8324,
+      "step": 6301
+    },
+    {
+      "epoch": 1.0685856355959449,
+      "grad_norm": 1.7109375,
+      "learning_rate": 1.4356494534654627e-05,
+      "loss": 0.9125,
+      "step": 6302
+    },
+    {
+      "epoch": 1.0687570996849347,
+      "grad_norm": 1.921875,
+      "learning_rate": 1.4354868918034679e-05,
+      "loss": 0.7998,
+      "step": 6303
+    },
+    {
+      "epoch": 1.0689285637739245,
+      "grad_norm": 1.7265625,
+      "learning_rate": 1.435324315938606e-05,
+      "loss": 0.9459,
+      "step": 6304
+    },
+    {
+      "epoch": 1.0691000278629144,
+      "grad_norm": 1.6953125,
+      "learning_rate": 1.4351617258761787e-05,
+      "loss": 0.874,
+      "step": 6305
+    },
+    {
+      "epoch": 1.0692714919519044,
+      "grad_norm": 1.65625,
+      "learning_rate": 1.434999121621489e-05,
+      "loss": 0.8602,
+      "step": 6306
+    },
+    {
+      "epoch": 1.0694429560408942,
+      "grad_norm": 1.6796875,
+      "learning_rate": 1.4348365031798398e-05,
+      "loss": 0.881,
+      "step": 6307
+    },
+    {
+      "epoch": 1.069614420129884,
+      "grad_norm": 1.734375,
+      "learning_rate": 1.4346738705565348e-05,
+      "loss": 0.8294,
+      "step": 6308
+    },
+    {
+      "epoch": 1.0697858842188739,
+      "grad_norm": 1.7265625,
+      "learning_rate": 1.4345112237568781e-05,
+      "loss": 0.8579,
+      "step": 6309
+    },
+    {
+      "epoch": 1.0699573483078637,
+      "grad_norm": 1.6171875,
+      "learning_rate": 1.4343485627861742e-05,
+      "loss": 0.8984,
+      "step": 6310
+    },
+    {
+      "epoch": 1.0701288123968535,
+      "grad_norm": 1.640625,
+      "learning_rate": 1.4341858876497279e-05,
+      "loss": 0.7913,
+      "step": 6311
+    },
+    {
+      "epoch": 1.0703002764858436,
+      "grad_norm": 1.703125,
+      "learning_rate": 1.4340231983528448e-05,
+      "loss": 0.8667,
+      "step": 6312
+    },
+    {
+      "epoch": 1.0704717405748334,
+      "grad_norm": 1.65625,
+      "learning_rate": 1.433860494900831e-05,
+      "loss": 0.8287,
+      "step": 6313
+    },
+    {
+      "epoch": 1.0706432046638232,
+      "grad_norm": 1.6484375,
+      "learning_rate": 1.4336977772989924e-05,
+      "loss": 0.8595,
+      "step": 6314
+    },
+    {
+      "epoch": 1.070814668752813,
+      "grad_norm": 1.640625,
+      "learning_rate": 1.4335350455526367e-05,
+      "loss": 0.8163,
+      "step": 6315
+    },
+    {
+      "epoch": 1.0709861328418029,
+      "grad_norm": 1.734375,
+      "learning_rate": 1.4333722996670702e-05,
+      "loss": 0.8464,
+      "step": 6316
+    },
+    {
+      "epoch": 1.0711575969307927,
+      "grad_norm": 1.7265625,
+      "learning_rate": 1.4332095396476012e-05,
+      "loss": 0.8942,
+      "step": 6317
+    },
+    {
+      "epoch": 1.0713290610197828,
+      "grad_norm": 1.640625,
+      "learning_rate": 1.4330467654995376e-05,
+      "loss": 0.8218,
+      "step": 6318
+    },
+    {
+      "epoch": 1.0715005251087726,
+      "grad_norm": 1.6953125,
+      "learning_rate": 1.4328839772281884e-05,
+      "loss": 0.8912,
+      "step": 6319
+    },
+    {
+      "epoch": 1.0716719891977624,
+      "grad_norm": 1.6328125,
+      "learning_rate": 1.4327211748388626e-05,
+      "loss": 0.8496,
+      "step": 6320
+    },
+    {
+      "epoch": 1.0718434532867522,
+      "grad_norm": 1.7734375,
+      "learning_rate": 1.4325583583368698e-05,
+      "loss": 0.9215,
+      "step": 6321
+    },
+    {
+      "epoch": 1.072014917375742,
+      "grad_norm": 1.59375,
+      "learning_rate": 1.4323955277275201e-05,
+      "loss": 0.824,
+      "step": 6322
+    },
+    {
+      "epoch": 1.0721863814647319,
+      "grad_norm": 1.6015625,
+      "learning_rate": 1.432232683016124e-05,
+      "loss": 0.903,
+      "step": 6323
+    },
+    {
+      "epoch": 1.072357845553722,
+      "grad_norm": 1.8046875,
+      "learning_rate": 1.4320698242079925e-05,
+      "loss": 0.8151,
+      "step": 6324
+    },
+    {
+      "epoch": 1.0725293096427118,
+      "grad_norm": 1.5234375,
+      "learning_rate": 1.431906951308437e-05,
+      "loss": 0.7505,
+      "step": 6325
+    },
+    {
+      "epoch": 1.0727007737317016,
+      "grad_norm": 1.59375,
+      "learning_rate": 1.4317440643227693e-05,
+      "loss": 0.84,
+      "step": 6326
+    },
+    {
+      "epoch": 1.0728722378206914,
+      "grad_norm": 1.6640625,
+      "learning_rate": 1.4315811632563022e-05,
+      "loss": 0.8941,
+      "step": 6327
+    },
+    {
+      "epoch": 1.0730437019096812,
+      "grad_norm": 1.6171875,
+      "learning_rate": 1.4314182481143478e-05,
+      "loss": 0.9537,
+      "step": 6328
+    },
+    {
+      "epoch": 1.073215165998671,
+      "grad_norm": 1.5546875,
+      "learning_rate": 1.4312553189022201e-05,
+      "loss": 0.8015,
+      "step": 6329
+    },
+    {
+      "epoch": 1.073386630087661,
+      "grad_norm": 1.7109375,
+      "learning_rate": 1.431092375625232e-05,
+      "loss": 0.8483,
+      "step": 6330
+    },
+    {
+      "epoch": 1.073558094176651,
+      "grad_norm": 1.7265625,
+      "learning_rate": 1.4309294182886984e-05,
+      "loss": 0.874,
+      "step": 6331
+    },
+    {
+      "epoch": 1.0737295582656408,
+      "grad_norm": 1.6171875,
+      "learning_rate": 1.4307664468979336e-05,
+      "loss": 0.9018,
+      "step": 6332
+    },
+    {
+      "epoch": 1.0739010223546306,
+      "grad_norm": 1.5625,
+      "learning_rate": 1.4306034614582534e-05,
+      "loss": 0.7619,
+      "step": 6333
+    },
+    {
+      "epoch": 1.0740724864436204,
+      "grad_norm": 1.6484375,
+      "learning_rate": 1.4304404619749724e-05,
+      "loss": 0.7982,
+      "step": 6334
+    },
+    {
+      "epoch": 1.0742439505326102,
+      "grad_norm": 1.578125,
+      "learning_rate": 1.4302774484534073e-05,
+      "loss": 0.8727,
+      "step": 6335
+    },
+    {
+      "epoch": 1.0744154146216003,
+      "grad_norm": 1.8125,
+      "learning_rate": 1.430114420898874e-05,
+      "loss": 0.8913,
+      "step": 6336
+    },
+    {
+      "epoch": 1.07458687871059,
+      "grad_norm": 1.6171875,
+      "learning_rate": 1.4299513793166896e-05,
+      "loss": 0.9844,
+      "step": 6337
+    },
+    {
+      "epoch": 1.07475834279958,
+      "grad_norm": 1.671875,
+      "learning_rate": 1.4297883237121721e-05,
+      "loss": 0.8231,
+      "step": 6338
+    },
+    {
+      "epoch": 1.0749298068885698,
+      "grad_norm": 1.640625,
+      "learning_rate": 1.429625254090639e-05,
+      "loss": 0.878,
+      "step": 6339
+    },
+    {
+      "epoch": 1.0751012709775596,
+      "grad_norm": 1.703125,
+      "learning_rate": 1.4294621704574084e-05,
+      "loss": 0.8635,
+      "step": 6340
+    },
+    {
+      "epoch": 1.0752727350665494,
+      "grad_norm": 1.734375,
+      "learning_rate": 1.4292990728177992e-05,
+      "loss": 0.8971,
+      "step": 6341
+    },
+    {
+      "epoch": 1.0754441991555395,
+      "grad_norm": 1.59375,
+      "learning_rate": 1.4291359611771304e-05,
+      "loss": 0.8481,
+      "step": 6342
+    },
+    {
+      "epoch": 1.0756156632445293,
+      "grad_norm": 1.6953125,
+      "learning_rate": 1.4289728355407221e-05,
+      "loss": 0.8761,
+      "step": 6343
+    },
+    {
+      "epoch": 1.075787127333519,
+      "grad_norm": 1.5078125,
+      "learning_rate": 1.4288096959138946e-05,
+      "loss": 0.7794,
+      "step": 6344
+    },
+    {
+      "epoch": 1.075958591422509,
+      "grad_norm": 1.78125,
+      "learning_rate": 1.4286465423019679e-05,
+      "loss": 0.9239,
+      "step": 6345
+    },
+    {
+      "epoch": 1.0761300555114988,
+      "grad_norm": 1.6484375,
+      "learning_rate": 1.4284833747102634e-05,
+      "loss": 0.8863,
+      "step": 6346
+    },
+    {
+      "epoch": 1.0763015196004886,
+      "grad_norm": 1.6953125,
+      "learning_rate": 1.4283201931441024e-05,
+      "loss": 0.9214,
+      "step": 6347
+    },
+    {
+      "epoch": 1.0764729836894786,
+      "grad_norm": 1.671875,
+      "learning_rate": 1.428156997608807e-05,
+      "loss": 0.8686,
+      "step": 6348
+    },
+    {
+      "epoch": 1.0766444477784685,
+      "grad_norm": 1.703125,
+      "learning_rate": 1.4279937881096997e-05,
+      "loss": 0.8189,
+      "step": 6349
+    },
+    {
+      "epoch": 1.0768159118674583,
+      "grad_norm": 1.65625,
+      "learning_rate": 1.4278305646521032e-05,
+      "loss": 0.8565,
+      "step": 6350
+    },
+    {
+      "epoch": 1.0769873759564481,
+      "grad_norm": 1.6015625,
+      "learning_rate": 1.427667327241341e-05,
+      "loss": 0.7881,
+      "step": 6351
+    },
+    {
+      "epoch": 1.077158840045438,
+      "grad_norm": 1.5390625,
+      "learning_rate": 1.4275040758827367e-05,
+      "loss": 0.8265,
+      "step": 6352
+    },
+    {
+      "epoch": 1.0773303041344278,
+      "grad_norm": 1.71875,
+      "learning_rate": 1.4273408105816148e-05,
+      "loss": 0.8923,
+      "step": 6353
+    },
+    {
+      "epoch": 1.0775017682234178,
+      "grad_norm": 1.6484375,
+      "learning_rate": 1.4271775313432998e-05,
+      "loss": 0.8301,
+      "step": 6354
+    },
+    {
+      "epoch": 1.0776732323124076,
+      "grad_norm": 1.6640625,
+      "learning_rate": 1.4270142381731168e-05,
+      "loss": 0.7962,
+      "step": 6355
+    },
+    {
+      "epoch": 1.0778446964013975,
+      "grad_norm": 1.65625,
+      "learning_rate": 1.426850931076392e-05,
+      "loss": 0.856,
+      "step": 6356
+    },
+    {
+      "epoch": 1.0780161604903873,
+      "grad_norm": 1.7109375,
+      "learning_rate": 1.4266876100584505e-05,
+      "loss": 0.9114,
+      "step": 6357
+    },
+    {
+      "epoch": 1.0781876245793771,
+      "grad_norm": 1.6953125,
+      "learning_rate": 1.4265242751246191e-05,
+      "loss": 0.8902,
+      "step": 6358
+    },
+    {
+      "epoch": 1.078359088668367,
+      "grad_norm": 1.6953125,
+      "learning_rate": 1.4263609262802255e-05,
+      "loss": 0.9172,
+      "step": 6359
+    },
+    {
+      "epoch": 1.078530552757357,
+      "grad_norm": 1.6484375,
+      "learning_rate": 1.426197563530596e-05,
+      "loss": 0.9066,
+      "step": 6360
+    },
+    {
+      "epoch": 1.0787020168463468,
+      "grad_norm": 1.6171875,
+      "learning_rate": 1.4260341868810591e-05,
+      "loss": 0.7365,
+      "step": 6361
+    },
+    {
+      "epoch": 1.0788734809353366,
+      "grad_norm": 1.7109375,
+      "learning_rate": 1.4258707963369438e-05,
+      "loss": 0.8886,
+      "step": 6362
+    },
+    {
+      "epoch": 1.0790449450243265,
+      "grad_norm": 1.640625,
+      "learning_rate": 1.4257073919035775e-05,
+      "loss": 0.8961,
+      "step": 6363
+    },
+    {
+      "epoch": 1.0792164091133163,
+      "grad_norm": 1.609375,
+      "learning_rate": 1.4255439735862901e-05,
+      "loss": 0.782,
+      "step": 6364
+    },
+    {
+      "epoch": 1.0793878732023061,
+      "grad_norm": 1.671875,
+      "learning_rate": 1.4253805413904114e-05,
+      "loss": 0.8904,
+      "step": 6365
+    },
+    {
+      "epoch": 1.079559337291296,
+      "grad_norm": 1.75,
+      "learning_rate": 1.4252170953212713e-05,
+      "loss": 0.8601,
+      "step": 6366
+    },
+    {
+      "epoch": 1.079730801380286,
+      "grad_norm": 1.78125,
+      "learning_rate": 1.4250536353842009e-05,
+      "loss": 0.8834,
+      "step": 6367
+    },
+    {
+      "epoch": 1.0799022654692758,
+      "grad_norm": 1.625,
+      "learning_rate": 1.4248901615845304e-05,
+      "loss": 0.8273,
+      "step": 6368
+    },
+    {
+      "epoch": 1.0800737295582656,
+      "grad_norm": 1.5859375,
+      "learning_rate": 1.4247266739275918e-05,
+      "loss": 0.8262,
+      "step": 6369
+    },
+    {
+      "epoch": 1.0802451936472555,
+      "grad_norm": 1.6328125,
+      "learning_rate": 1.4245631724187172e-05,
+      "loss": 0.8552,
+      "step": 6370
+    },
+    {
+      "epoch": 1.0804166577362453,
+      "grad_norm": 1.578125,
+      "learning_rate": 1.4243996570632385e-05,
+      "loss": 0.7704,
+      "step": 6371
+    },
+    {
+      "epoch": 1.0805881218252353,
+      "grad_norm": 1.59375,
+      "learning_rate": 1.4242361278664891e-05,
+      "loss": 0.8284,
+      "step": 6372
+    },
+    {
+      "epoch": 1.0807595859142252,
+      "grad_norm": 1.765625,
+      "learning_rate": 1.4240725848338023e-05,
+      "loss": 0.8723,
+      "step": 6373
+    },
+    {
+      "epoch": 1.080931050003215,
+      "grad_norm": 1.7265625,
+      "learning_rate": 1.423909027970511e-05,
+      "loss": 0.8555,
+      "step": 6374
+    },
+    {
+      "epoch": 1.0811025140922048,
+      "grad_norm": 1.6796875,
+      "learning_rate": 1.4237454572819505e-05,
+      "loss": 0.9163,
+      "step": 6375
+    },
+    {
+      "epoch": 1.0812739781811946,
+      "grad_norm": 1.6953125,
+      "learning_rate": 1.4235818727734544e-05,
+      "loss": 0.8108,
+      "step": 6376
+    },
+    {
+      "epoch": 1.0814454422701845,
+      "grad_norm": 1.6015625,
+      "learning_rate": 1.4234182744503586e-05,
+      "loss": 0.8376,
+      "step": 6377
+    },
+    {
+      "epoch": 1.0816169063591743,
+      "grad_norm": 1.640625,
+      "learning_rate": 1.4232546623179985e-05,
+      "loss": 0.7456,
+      "step": 6378
+    },
+    {
+      "epoch": 1.0817883704481643,
+      "grad_norm": 1.7265625,
+      "learning_rate": 1.4230910363817104e-05,
+      "loss": 0.8937,
+      "step": 6379
+    },
+    {
+      "epoch": 1.0819598345371542,
+      "grad_norm": 1.6484375,
+      "learning_rate": 1.4229273966468298e-05,
+      "loss": 0.8592,
+      "step": 6380
+    },
+    {
+      "epoch": 1.082131298626144,
+      "grad_norm": 1.6171875,
+      "learning_rate": 1.4227637431186945e-05,
+      "loss": 0.8733,
+      "step": 6381
+    },
+    {
+      "epoch": 1.0823027627151338,
+      "grad_norm": 1.7265625,
+      "learning_rate": 1.4226000758026414e-05,
+      "loss": 0.9073,
+      "step": 6382
+    },
+    {
+      "epoch": 1.0824742268041236,
+      "grad_norm": 1.59375,
+      "learning_rate": 1.4224363947040085e-05,
+      "loss": 0.7889,
+      "step": 6383
+    },
+    {
+      "epoch": 1.0826456908931137,
+      "grad_norm": 1.71875,
+      "learning_rate": 1.4222726998281342e-05,
+      "loss": 0.8612,
+      "step": 6384
+    },
+    {
+      "epoch": 1.0828171549821035,
+      "grad_norm": 1.6796875,
+      "learning_rate": 1.422108991180357e-05,
+      "loss": 0.8388,
+      "step": 6385
+    },
+    {
+      "epoch": 1.0829886190710933,
+      "grad_norm": 1.6875,
+      "learning_rate": 1.4219452687660158e-05,
+      "loss": 0.8867,
+      "step": 6386
+    },
+    {
+      "epoch": 1.0831600831600832,
+      "grad_norm": 1.640625,
+      "learning_rate": 1.4217815325904508e-05,
+      "loss": 0.8568,
+      "step": 6387
+    },
+    {
+      "epoch": 1.083331547249073,
+      "grad_norm": 1.71875,
+      "learning_rate": 1.4216177826590017e-05,
+      "loss": 0.9192,
+      "step": 6388
+    },
+    {
+      "epoch": 1.0835030113380628,
+      "grad_norm": 1.6875,
+      "learning_rate": 1.4214540189770087e-05,
+      "loss": 0.9198,
+      "step": 6389
+    },
+    {
+      "epoch": 1.0836744754270526,
+      "grad_norm": 1.6015625,
+      "learning_rate": 1.4212902415498136e-05,
+      "loss": 0.8886,
+      "step": 6390
+    },
+    {
+      "epoch": 1.0838459395160427,
+      "grad_norm": 1.640625,
+      "learning_rate": 1.4211264503827571e-05,
+      "loss": 0.8625,
+      "step": 6391
+    },
+    {
+      "epoch": 1.0840174036050325,
+      "grad_norm": 1.6484375,
+      "learning_rate": 1.4209626454811812e-05,
+      "loss": 0.7927,
+      "step": 6392
+    },
+    {
+      "epoch": 1.0841888676940223,
+      "grad_norm": 1.6640625,
+      "learning_rate": 1.420798826850428e-05,
+      "loss": 0.8718,
+      "step": 6393
+    },
+    {
+      "epoch": 1.0843603317830122,
+      "grad_norm": 1.75,
+      "learning_rate": 1.4206349944958407e-05,
+      "loss": 0.9271,
+      "step": 6394
+    },
+    {
+      "epoch": 1.084531795872002,
+      "grad_norm": 1.7265625,
+      "learning_rate": 1.4204711484227623e-05,
+      "loss": 0.8981,
+      "step": 6395
+    },
+    {
+      "epoch": 1.0847032599609918,
+      "grad_norm": 1.7109375,
+      "learning_rate": 1.4203072886365364e-05,
+      "loss": 0.8554,
+      "step": 6396
+    },
+    {
+      "epoch": 1.0848747240499819,
+      "grad_norm": 1.7265625,
+      "learning_rate": 1.4201434151425072e-05,
+      "loss": 0.9236,
+      "step": 6397
+    },
+    {
+      "epoch": 1.0850461881389717,
+      "grad_norm": 1.6953125,
+      "learning_rate": 1.4199795279460187e-05,
+      "loss": 0.8479,
+      "step": 6398
+    },
+    {
+      "epoch": 1.0852176522279615,
+      "grad_norm": 1.671875,
+      "learning_rate": 1.4198156270524167e-05,
+      "loss": 0.8149,
+      "step": 6399
+    },
+    {
+      "epoch": 1.0853891163169513,
+      "grad_norm": 1.65625,
+      "learning_rate": 1.419651712467046e-05,
+      "loss": 0.8815,
+      "step": 6400
+    },
+    {
+      "epoch": 1.0855605804059412,
+      "grad_norm": 1.7109375,
+      "learning_rate": 1.4194877841952526e-05,
+      "loss": 0.9174,
+      "step": 6401
+    },
+    {
+      "epoch": 1.085732044494931,
+      "grad_norm": 1.609375,
+      "learning_rate": 1.4193238422423834e-05,
+      "loss": 0.8742,
+      "step": 6402
+    },
+    {
+      "epoch": 1.085903508583921,
+      "grad_norm": 1.765625,
+      "learning_rate": 1.4191598866137844e-05,
+      "loss": 0.902,
+      "step": 6403
+    },
+    {
+      "epoch": 1.0860749726729109,
+      "grad_norm": 1.7265625,
+      "learning_rate": 1.418995917314803e-05,
+      "loss": 0.9654,
+      "step": 6404
+    },
+    {
+      "epoch": 1.0862464367619007,
+      "grad_norm": 1.75,
+      "learning_rate": 1.4188319343507865e-05,
+      "loss": 0.8984,
+      "step": 6405
+    },
+    {
+      "epoch": 1.0864179008508905,
+      "grad_norm": 1.6640625,
+      "learning_rate": 1.4186679377270841e-05,
+      "loss": 0.9152,
+      "step": 6406
+    },
+    {
+      "epoch": 1.0865893649398803,
+      "grad_norm": 1.59375,
+      "learning_rate": 1.4185039274490436e-05,
+      "loss": 0.8611,
+      "step": 6407
+    },
+    {
+      "epoch": 1.0867608290288702,
+      "grad_norm": 1.671875,
+      "learning_rate": 1.4183399035220143e-05,
+      "loss": 0.8521,
+      "step": 6408
+    },
+    {
+      "epoch": 1.0869322931178602,
+      "grad_norm": 1.765625,
+      "learning_rate": 1.418175865951345e-05,
+      "loss": 0.8818,
+      "step": 6409
+    },
+    {
+      "epoch": 1.08710375720685,
+      "grad_norm": 1.6015625,
+      "learning_rate": 1.4180118147423861e-05,
+      "loss": 0.8461,
+      "step": 6410
+    },
+    {
+      "epoch": 1.0872752212958399,
+      "grad_norm": 1.7109375,
+      "learning_rate": 1.4178477499004879e-05,
+      "loss": 0.7814,
+      "step": 6411
+    },
+    {
+      "epoch": 1.0874466853848297,
+      "grad_norm": 1.703125,
+      "learning_rate": 1.417683671431001e-05,
+      "loss": 0.9004,
+      "step": 6412
+    },
+    {
+      "epoch": 1.0876181494738195,
+      "grad_norm": 1.6484375,
+      "learning_rate": 1.4175195793392769e-05,
+      "loss": 0.851,
+      "step": 6413
+    },
+    {
+      "epoch": 1.0877896135628093,
+      "grad_norm": 1.65625,
+      "learning_rate": 1.417355473630667e-05,
+      "loss": 0.825,
+      "step": 6414
+    },
+    {
+      "epoch": 1.0879610776517994,
+      "grad_norm": 1.7109375,
+      "learning_rate": 1.4171913543105233e-05,
+      "loss": 0.932,
+      "step": 6415
+    },
+    {
+      "epoch": 1.0881325417407892,
+      "grad_norm": 1.7734375,
+      "learning_rate": 1.4170272213841988e-05,
+      "loss": 0.8789,
+      "step": 6416
+    },
+    {
+      "epoch": 1.088304005829779,
+      "grad_norm": 1.625,
+      "learning_rate": 1.4168630748570462e-05,
+      "loss": 0.8809,
+      "step": 6417
+    },
+    {
+      "epoch": 1.0884754699187689,
+      "grad_norm": 1.6875,
+      "learning_rate": 1.4166989147344188e-05,
+      "loss": 0.8888,
+      "step": 6418
+    },
+    {
+      "epoch": 1.0886469340077587,
+      "grad_norm": 1.6015625,
+      "learning_rate": 1.416534741021671e-05,
+      "loss": 0.8732,
+      "step": 6419
+    },
+    {
+      "epoch": 1.0888183980967485,
+      "grad_norm": 1.5703125,
+      "learning_rate": 1.4163705537241565e-05,
+      "loss": 0.7691,
+      "step": 6420
+    },
+    {
+      "epoch": 1.0889898621857386,
+      "grad_norm": 1.6328125,
+      "learning_rate": 1.4162063528472302e-05,
+      "loss": 0.864,
+      "step": 6421
+    },
+    {
+      "epoch": 1.0891613262747284,
+      "grad_norm": 1.6328125,
+      "learning_rate": 1.4160421383962478e-05,
+      "loss": 0.793,
+      "step": 6422
+    },
+    {
+      "epoch": 1.0893327903637182,
+      "grad_norm": 1.6015625,
+      "learning_rate": 1.4158779103765642e-05,
+      "loss": 0.8254,
+      "step": 6423
+    },
+    {
+      "epoch": 1.089504254452708,
+      "grad_norm": 1.7578125,
+      "learning_rate": 1.415713668793536e-05,
+      "loss": 0.8618,
+      "step": 6424
+    },
+    {
+      "epoch": 1.0896757185416979,
+      "grad_norm": 1.640625,
+      "learning_rate": 1.4155494136525198e-05,
+      "loss": 0.8551,
+      "step": 6425
+    },
+    {
+      "epoch": 1.0898471826306877,
+      "grad_norm": 1.7109375,
+      "learning_rate": 1.4153851449588725e-05,
+      "loss": 0.9366,
+      "step": 6426
+    },
+    {
+      "epoch": 1.0900186467196777,
+      "grad_norm": 1.640625,
+      "learning_rate": 1.4152208627179513e-05,
+      "loss": 0.8696,
+      "step": 6427
+    },
+    {
+      "epoch": 1.0901901108086676,
+      "grad_norm": 1.6015625,
+      "learning_rate": 1.4150565669351141e-05,
+      "loss": 0.889,
+      "step": 6428
+    },
+    {
+      "epoch": 1.0903615748976574,
+      "grad_norm": 1.7265625,
+      "learning_rate": 1.4148922576157194e-05,
+      "loss": 0.9755,
+      "step": 6429
+    },
+    {
+      "epoch": 1.0905330389866472,
+      "grad_norm": 1.703125,
+      "learning_rate": 1.4147279347651256e-05,
+      "loss": 0.9041,
+      "step": 6430
+    },
+    {
+      "epoch": 1.090704503075637,
+      "grad_norm": 1.546875,
+      "learning_rate": 1.4145635983886927e-05,
+      "loss": 0.7879,
+      "step": 6431
+    },
+    {
+      "epoch": 1.0908759671646269,
+      "grad_norm": 1.65625,
+      "learning_rate": 1.4143992484917792e-05,
+      "loss": 0.7683,
+      "step": 6432
+    },
+    {
+      "epoch": 1.091047431253617,
+      "grad_norm": 1.609375,
+      "learning_rate": 1.4142348850797458e-05,
+      "loss": 0.7695,
+      "step": 6433
+    },
+    {
+      "epoch": 1.0912188953426067,
+      "grad_norm": 1.6171875,
+      "learning_rate": 1.414070508157953e-05,
+      "loss": 0.8802,
+      "step": 6434
+    },
+    {
+      "epoch": 1.0913903594315966,
+      "grad_norm": 1.671875,
+      "learning_rate": 1.4139061177317616e-05,
+      "loss": 0.9043,
+      "step": 6435
+    },
+    {
+      "epoch": 1.0915618235205864,
+      "grad_norm": 1.6796875,
+      "learning_rate": 1.4137417138065333e-05,
+      "loss": 0.8847,
+      "step": 6436
+    },
+    {
+      "epoch": 1.0917332876095762,
+      "grad_norm": 1.640625,
+      "learning_rate": 1.4135772963876297e-05,
+      "loss": 0.8592,
+      "step": 6437
+    },
+    {
+      "epoch": 1.091904751698566,
+      "grad_norm": 1.625,
+      "learning_rate": 1.4134128654804131e-05,
+      "loss": 0.7798,
+      "step": 6438
+    },
+    {
+      "epoch": 1.092076215787556,
+      "grad_norm": 1.65625,
+      "learning_rate": 1.413248421090246e-05,
+      "loss": 0.8814,
+      "step": 6439
+    },
+    {
+      "epoch": 1.092247679876546,
+      "grad_norm": 1.5625,
+      "learning_rate": 1.4130839632224918e-05,
+      "loss": 0.8163,
+      "step": 6440
+    },
+    {
+      "epoch": 1.0924191439655357,
+      "grad_norm": 1.65625,
+      "learning_rate": 1.412919491882514e-05,
+      "loss": 0.9243,
+      "step": 6441
+    },
+    {
+      "epoch": 1.0925906080545256,
+      "grad_norm": 1.6171875,
+      "learning_rate": 1.4127550070756768e-05,
+      "loss": 0.8714,
+      "step": 6442
+    },
+    {
+      "epoch": 1.0927620721435154,
+      "grad_norm": 1.640625,
+      "learning_rate": 1.4125905088073442e-05,
+      "loss": 0.9572,
+      "step": 6443
+    },
+    {
+      "epoch": 1.0929335362325052,
+      "grad_norm": 1.703125,
+      "learning_rate": 1.4124259970828817e-05,
+      "loss": 0.8775,
+      "step": 6444
+    },
+    {
+      "epoch": 1.0931050003214953,
+      "grad_norm": 1.703125,
+      "learning_rate": 1.4122614719076544e-05,
+      "loss": 0.9437,
+      "step": 6445
+    },
+    {
+      "epoch": 1.093276464410485,
+      "grad_norm": 1.6875,
+      "learning_rate": 1.412096933287028e-05,
+      "loss": 0.907,
+      "step": 6446
+    },
+    {
+      "epoch": 1.093447928499475,
+      "grad_norm": 1.5703125,
+      "learning_rate": 1.4119323812263688e-05,
+      "loss": 0.807,
+      "step": 6447
+    },
+    {
+      "epoch": 1.0936193925884647,
+      "grad_norm": 1.5625,
+      "learning_rate": 1.4117678157310436e-05,
+      "loss": 0.8288,
+      "step": 6448
+    },
+    {
+      "epoch": 1.0937908566774546,
+      "grad_norm": 1.6875,
+      "learning_rate": 1.4116032368064192e-05,
+      "loss": 0.8559,
+      "step": 6449
+    },
+    {
+      "epoch": 1.0939623207664444,
+      "grad_norm": 1.6484375,
+      "learning_rate": 1.4114386444578632e-05,
+      "loss": 0.8325,
+      "step": 6450
+    },
+    {
+      "epoch": 1.0941337848554344,
+      "grad_norm": 1.6875,
+      "learning_rate": 1.4112740386907438e-05,
+      "loss": 0.8723,
+      "step": 6451
+    },
+    {
+      "epoch": 1.0943052489444243,
+      "grad_norm": 1.6640625,
+      "learning_rate": 1.4111094195104291e-05,
+      "loss": 0.8334,
+      "step": 6452
+    },
+    {
+      "epoch": 1.094476713033414,
+      "grad_norm": 1.7421875,
+      "learning_rate": 1.410944786922288e-05,
+      "loss": 1.0709,
+      "step": 6453
+    },
+    {
+      "epoch": 1.094648177122404,
+      "grad_norm": 1.84375,
+      "learning_rate": 1.4107801409316907e-05,
+      "loss": 0.907,
+      "step": 6454
+    },
+    {
+      "epoch": 1.0948196412113937,
+      "grad_norm": 1.65625,
+      "learning_rate": 1.4106154815440056e-05,
+      "loss": 0.8274,
+      "step": 6455
+    },
+    {
+      "epoch": 1.0949911053003836,
+      "grad_norm": 1.59375,
+      "learning_rate": 1.4104508087646033e-05,
+      "loss": 0.7924,
+      "step": 6456
+    },
+    {
+      "epoch": 1.0951625693893736,
+      "grad_norm": 1.578125,
+      "learning_rate": 1.410286122598855e-05,
+      "loss": 0.8505,
+      "step": 6457
+    },
+    {
+      "epoch": 1.0953340334783634,
+      "grad_norm": 1.6953125,
+      "learning_rate": 1.4101214230521307e-05,
+      "loss": 0.9114,
+      "step": 6458
+    },
+    {
+      "epoch": 1.0955054975673533,
+      "grad_norm": 1.734375,
+      "learning_rate": 1.4099567101298026e-05,
+      "loss": 0.8981,
+      "step": 6459
+    },
+    {
+      "epoch": 1.095676961656343,
+      "grad_norm": 1.7109375,
+      "learning_rate": 1.4097919838372425e-05,
+      "loss": 0.8765,
+      "step": 6460
+    },
+    {
+      "epoch": 1.095848425745333,
+      "grad_norm": 1.6953125,
+      "learning_rate": 1.4096272441798227e-05,
+      "loss": 0.9022,
+      "step": 6461
+    },
+    {
+      "epoch": 1.0960198898343227,
+      "grad_norm": 2.046875,
+      "learning_rate": 1.4094624911629158e-05,
+      "loss": 0.8844,
+      "step": 6462
+    },
+    {
+      "epoch": 1.0961913539233126,
+      "grad_norm": 1.546875,
+      "learning_rate": 1.4092977247918953e-05,
+      "loss": 0.7989,
+      "step": 6463
+    },
+    {
+      "epoch": 1.0963628180123026,
+      "grad_norm": 1.625,
+      "learning_rate": 1.4091329450721347e-05,
+      "loss": 0.8971,
+      "step": 6464
+    },
+    {
+      "epoch": 1.0965342821012924,
+      "grad_norm": 1.6875,
+      "learning_rate": 1.4089681520090084e-05,
+      "loss": 0.8169,
+      "step": 6465
+    },
+    {
+      "epoch": 1.0967057461902823,
+      "grad_norm": 1.6640625,
+      "learning_rate": 1.4088033456078902e-05,
+      "loss": 0.8382,
+      "step": 6466
+    },
+    {
+      "epoch": 1.096877210279272,
+      "grad_norm": 1.6640625,
+      "learning_rate": 1.4086385258741557e-05,
+      "loss": 0.8602,
+      "step": 6467
+    },
+    {
+      "epoch": 1.097048674368262,
+      "grad_norm": 1.6875,
+      "learning_rate": 1.40847369281318e-05,
+      "loss": 0.8289,
+      "step": 6468
+    },
+    {
+      "epoch": 1.097220138457252,
+      "grad_norm": 1.640625,
+      "learning_rate": 1.408308846430339e-05,
+      "loss": 0.868,
+      "step": 6469
+    },
+    {
+      "epoch": 1.0973916025462418,
+      "grad_norm": 1.609375,
+      "learning_rate": 1.408143986731009e-05,
+      "loss": 0.8289,
+      "step": 6470
+    },
+    {
+      "epoch": 1.0975630666352316,
+      "grad_norm": 1.6796875,
+      "learning_rate": 1.4079791137205665e-05,
+      "loss": 0.8554,
+      "step": 6471
+    },
+    {
+      "epoch": 1.0977345307242214,
+      "grad_norm": 1.703125,
+      "learning_rate": 1.407814227404389e-05,
+      "loss": 0.9221,
+      "step": 6472
+    },
+    {
+      "epoch": 1.0979059948132113,
+      "grad_norm": 1.7265625,
+      "learning_rate": 1.4076493277878537e-05,
+      "loss": 0.8511,
+      "step": 6473
+    },
+    {
+      "epoch": 1.098077458902201,
+      "grad_norm": 1.609375,
+      "learning_rate": 1.4074844148763388e-05,
+      "loss": 0.8883,
+      "step": 6474
+    },
+    {
+      "epoch": 1.098248922991191,
+      "grad_norm": 1.578125,
+      "learning_rate": 1.4073194886752228e-05,
+      "loss": 0.8379,
+      "step": 6475
+    },
+    {
+      "epoch": 1.098420387080181,
+      "grad_norm": 1.59375,
+      "learning_rate": 1.4071545491898843e-05,
+      "loss": 0.8347,
+      "step": 6476
+    },
+    {
+      "epoch": 1.0985918511691708,
+      "grad_norm": 1.59375,
+      "learning_rate": 1.4069895964257032e-05,
+      "loss": 0.9084,
+      "step": 6477
+    },
+    {
+      "epoch": 1.0987633152581606,
+      "grad_norm": 1.6953125,
+      "learning_rate": 1.4068246303880584e-05,
+      "loss": 0.9504,
+      "step": 6478
+    },
+    {
+      "epoch": 1.0989347793471504,
+      "grad_norm": 1.6015625,
+      "learning_rate": 1.4066596510823303e-05,
+      "loss": 0.8512,
+      "step": 6479
+    },
+    {
+      "epoch": 1.0991062434361403,
+      "grad_norm": 1.5859375,
+      "learning_rate": 1.4064946585138997e-05,
+      "loss": 0.8665,
+      "step": 6480
+    },
+    {
+      "epoch": 1.0992777075251303,
+      "grad_norm": 1.71875,
+      "learning_rate": 1.4063296526881477e-05,
+      "loss": 0.8951,
+      "step": 6481
+    },
+    {
+      "epoch": 1.0994491716141201,
+      "grad_norm": 1.71875,
+      "learning_rate": 1.4061646336104556e-05,
+      "loss": 0.8214,
+      "step": 6482
+    },
+    {
+      "epoch": 1.09962063570311,
+      "grad_norm": 1.6484375,
+      "learning_rate": 1.4059996012862055e-05,
+      "loss": 0.9158,
+      "step": 6483
+    },
+    {
+      "epoch": 1.0997920997920998,
+      "grad_norm": 1.640625,
+      "learning_rate": 1.4058345557207794e-05,
+      "loss": 0.8622,
+      "step": 6484
+    },
+    {
+      "epoch": 1.0999635638810896,
+      "grad_norm": 1.6640625,
+      "learning_rate": 1.40566949691956e-05,
+      "loss": 0.8542,
+      "step": 6485
+    },
+    {
+      "epoch": 1.1001350279700794,
+      "grad_norm": 1.671875,
+      "learning_rate": 1.405504424887931e-05,
+      "loss": 0.9208,
+      "step": 6486
+    },
+    {
+      "epoch": 1.1003064920590693,
+      "grad_norm": 1.7421875,
+      "learning_rate": 1.4053393396312756e-05,
+      "loss": 0.8273,
+      "step": 6487
+    },
+    {
+      "epoch": 1.1004779561480593,
+      "grad_norm": 1.640625,
+      "learning_rate": 1.405174241154978e-05,
+      "loss": 0.8275,
+      "step": 6488
+    },
+    {
+      "epoch": 1.1006494202370491,
+      "grad_norm": 1.6640625,
+      "learning_rate": 1.4050091294644226e-05,
+      "loss": 0.8867,
+      "step": 6489
+    },
+    {
+      "epoch": 1.100820884326039,
+      "grad_norm": 1.671875,
+      "learning_rate": 1.4048440045649943e-05,
+      "loss": 0.8546,
+      "step": 6490
+    },
+    {
+      "epoch": 1.1009923484150288,
+      "grad_norm": 1.7890625,
+      "learning_rate": 1.4046788664620785e-05,
+      "loss": 0.8363,
+      "step": 6491
+    },
+    {
+      "epoch": 1.1011638125040186,
+      "grad_norm": 1.703125,
+      "learning_rate": 1.4045137151610614e-05,
+      "loss": 0.9141,
+      "step": 6492
+    },
+    {
+      "epoch": 1.1013352765930084,
+      "grad_norm": 1.78125,
+      "learning_rate": 1.4043485506673282e-05,
+      "loss": 0.9679,
+      "step": 6493
+    },
+    {
+      "epoch": 1.1015067406819985,
+      "grad_norm": 1.640625,
+      "learning_rate": 1.4041833729862666e-05,
+      "loss": 0.8664,
+      "step": 6494
+    },
+    {
+      "epoch": 1.1016782047709883,
+      "grad_norm": 1.6953125,
+      "learning_rate": 1.404018182123263e-05,
+      "loss": 0.8076,
+      "step": 6495
+    },
+    {
+      "epoch": 1.1018496688599781,
+      "grad_norm": 1.65625,
+      "learning_rate": 1.4038529780837049e-05,
+      "loss": 0.9172,
+      "step": 6496
+    },
+    {
+      "epoch": 1.102021132948968,
+      "grad_norm": 1.6328125,
+      "learning_rate": 1.4036877608729806e-05,
+      "loss": 0.8456,
+      "step": 6497
+    },
+    {
+      "epoch": 1.1021925970379578,
+      "grad_norm": 1.71875,
+      "learning_rate": 1.4035225304964781e-05,
+      "loss": 0.8858,
+      "step": 6498
+    },
+    {
+      "epoch": 1.1023640611269476,
+      "grad_norm": 1.6640625,
+      "learning_rate": 1.4033572869595866e-05,
+      "loss": 0.8169,
+      "step": 6499
+    },
+    {
+      "epoch": 1.1025355252159377,
+      "grad_norm": 1.5859375,
+      "learning_rate": 1.4031920302676951e-05,
+      "loss": 0.7992,
+      "step": 6500
+    },
+    {
+      "epoch": 1.1027069893049275,
+      "grad_norm": 1.7265625,
+      "learning_rate": 1.403026760426193e-05,
+      "loss": 0.9074,
+      "step": 6501
+    },
+    {
+      "epoch": 1.1028784533939173,
+      "grad_norm": 1.65625,
+      "learning_rate": 1.4028614774404707e-05,
+      "loss": 0.8155,
+      "step": 6502
+    },
+    {
+      "epoch": 1.1030499174829071,
+      "grad_norm": 1.5703125,
+      "learning_rate": 1.4026961813159186e-05,
+      "loss": 0.8348,
+      "step": 6503
+    },
+    {
+      "epoch": 1.103221381571897,
+      "grad_norm": 1.6953125,
+      "learning_rate": 1.4025308720579276e-05,
+      "loss": 0.8808,
+      "step": 6504
+    },
+    {
+      "epoch": 1.1033928456608868,
+      "grad_norm": 1.640625,
+      "learning_rate": 1.4023655496718892e-05,
+      "loss": 0.9387,
+      "step": 6505
+    },
+    {
+      "epoch": 1.1035643097498768,
+      "grad_norm": 1.734375,
+      "learning_rate": 1.402200214163195e-05,
+      "loss": 0.9147,
+      "step": 6506
+    },
+    {
+      "epoch": 1.1037357738388667,
+      "grad_norm": 1.6796875,
+      "learning_rate": 1.402034865537237e-05,
+      "loss": 0.8228,
+      "step": 6507
+    },
+    {
+      "epoch": 1.1039072379278565,
+      "grad_norm": 1.7109375,
+      "learning_rate": 1.4018695037994081e-05,
+      "loss": 0.8541,
+      "step": 6508
+    },
+    {
+      "epoch": 1.1040787020168463,
+      "grad_norm": 1.6171875,
+      "learning_rate": 1.4017041289551017e-05,
+      "loss": 0.8786,
+      "step": 6509
+    },
+    {
+      "epoch": 1.1042501661058362,
+      "grad_norm": 1.609375,
+      "learning_rate": 1.4015387410097108e-05,
+      "loss": 0.734,
+      "step": 6510
+    },
+    {
+      "epoch": 1.104421630194826,
+      "grad_norm": 1.625,
+      "learning_rate": 1.4013733399686296e-05,
+      "loss": 0.8647,
+      "step": 6511
+    },
+    {
+      "epoch": 1.104593094283816,
+      "grad_norm": 1.6015625,
+      "learning_rate": 1.4012079258372523e-05,
+      "loss": 0.871,
+      "step": 6512
+    },
+    {
+      "epoch": 1.1047645583728058,
+      "grad_norm": 1.6875,
+      "learning_rate": 1.4010424986209738e-05,
+      "loss": 0.9104,
+      "step": 6513
+    },
+    {
+      "epoch": 1.1049360224617957,
+      "grad_norm": 1.71875,
+      "learning_rate": 1.4008770583251892e-05,
+      "loss": 0.8532,
+      "step": 6514
+    },
+    {
+      "epoch": 1.1051074865507855,
+      "grad_norm": 1.6171875,
+      "learning_rate": 1.4007116049552942e-05,
+      "loss": 0.8285,
+      "step": 6515
+    },
+    {
+      "epoch": 1.1052789506397753,
+      "grad_norm": 1.71875,
+      "learning_rate": 1.4005461385166847e-05,
+      "loss": 0.8479,
+      "step": 6516
+    },
+    {
+      "epoch": 1.1054504147287652,
+      "grad_norm": 1.609375,
+      "learning_rate": 1.4003806590147573e-05,
+      "loss": 0.8497,
+      "step": 6517
+    },
+    {
+      "epoch": 1.1056218788177552,
+      "grad_norm": 1.7109375,
+      "learning_rate": 1.4002151664549088e-05,
+      "loss": 0.8999,
+      "step": 6518
+    },
+    {
+      "epoch": 1.105793342906745,
+      "grad_norm": 1.59375,
+      "learning_rate": 1.4000496608425369e-05,
+      "loss": 0.9195,
+      "step": 6519
+    },
+    {
+      "epoch": 1.1059648069957349,
+      "grad_norm": 1.5859375,
+      "learning_rate": 1.3998841421830391e-05,
+      "loss": 0.7609,
+      "step": 6520
+    },
+    {
+      "epoch": 1.1061362710847247,
+      "grad_norm": 1.6484375,
+      "learning_rate": 1.3997186104818134e-05,
+      "loss": 0.8273,
+      "step": 6521
+    },
+    {
+      "epoch": 1.1063077351737145,
+      "grad_norm": 1.6015625,
+      "learning_rate": 1.3995530657442588e-05,
+      "loss": 0.8517,
+      "step": 6522
+    },
+    {
+      "epoch": 1.1064791992627043,
+      "grad_norm": 1.6796875,
+      "learning_rate": 1.3993875079757744e-05,
+      "loss": 0.8764,
+      "step": 6523
+    },
+    {
+      "epoch": 1.1066506633516944,
+      "grad_norm": 1.6640625,
+      "learning_rate": 1.399221937181759e-05,
+      "loss": 0.8312,
+      "step": 6524
+    },
+    {
+      "epoch": 1.1068221274406842,
+      "grad_norm": 1.5859375,
+      "learning_rate": 1.3990563533676129e-05,
+      "loss": 0.8393,
+      "step": 6525
+    },
+    {
+      "epoch": 1.106993591529674,
+      "grad_norm": 1.625,
+      "learning_rate": 1.3988907565387364e-05,
+      "loss": 0.9126,
+      "step": 6526
+    },
+    {
+      "epoch": 1.1071650556186639,
+      "grad_norm": 1.7578125,
+      "learning_rate": 1.3987251467005303e-05,
+      "loss": 0.8461,
+      "step": 6527
+    },
+    {
+      "epoch": 1.1073365197076537,
+      "grad_norm": 1.640625,
+      "learning_rate": 1.3985595238583958e-05,
+      "loss": 0.889,
+      "step": 6528
+    },
+    {
+      "epoch": 1.1075079837966435,
+      "grad_norm": 1.75,
+      "learning_rate": 1.3983938880177345e-05,
+      "loss": 0.8923,
+      "step": 6529
+    },
+    {
+      "epoch": 1.1076794478856336,
+      "grad_norm": 1.6796875,
+      "learning_rate": 1.3982282391839483e-05,
+      "loss": 0.8516,
+      "step": 6530
+    },
+    {
+      "epoch": 1.1078509119746234,
+      "grad_norm": 1.7578125,
+      "learning_rate": 1.3980625773624394e-05,
+      "loss": 0.9837,
+      "step": 6531
+    },
+    {
+      "epoch": 1.1080223760636132,
+      "grad_norm": 1.7734375,
+      "learning_rate": 1.3978969025586109e-05,
+      "loss": 0.8922,
+      "step": 6532
+    },
+    {
+      "epoch": 1.108193840152603,
+      "grad_norm": 1.7421875,
+      "learning_rate": 1.397731214777866e-05,
+      "loss": 0.8567,
+      "step": 6533
+    },
+    {
+      "epoch": 1.1083653042415929,
+      "grad_norm": 1.546875,
+      "learning_rate": 1.3975655140256089e-05,
+      "loss": 0.7654,
+      "step": 6534
+    },
+    {
+      "epoch": 1.1085367683305827,
+      "grad_norm": 1.640625,
+      "learning_rate": 1.3973998003072429e-05,
+      "loss": 0.8481,
+      "step": 6535
+    },
+    {
+      "epoch": 1.1087082324195727,
+      "grad_norm": 1.5625,
+      "learning_rate": 1.397234073628173e-05,
+      "loss": 0.8526,
+      "step": 6536
+    },
+    {
+      "epoch": 1.1088796965085626,
+      "grad_norm": 1.6640625,
+      "learning_rate": 1.3970683339938042e-05,
+      "loss": 0.886,
+      "step": 6537
+    },
+    {
+      "epoch": 1.1090511605975524,
+      "grad_norm": 1.6640625,
+      "learning_rate": 1.3969025814095419e-05,
+      "loss": 0.9093,
+      "step": 6538
+    },
+    {
+      "epoch": 1.1092226246865422,
+      "grad_norm": 1.640625,
+      "learning_rate": 1.3967368158807915e-05,
+      "loss": 0.8276,
+      "step": 6539
+    },
+    {
+      "epoch": 1.109394088775532,
+      "grad_norm": 1.6484375,
+      "learning_rate": 1.3965710374129599e-05,
+      "loss": 0.8418,
+      "step": 6540
+    },
+    {
+      "epoch": 1.1095655528645219,
+      "grad_norm": 1.625,
+      "learning_rate": 1.3964052460114532e-05,
+      "loss": 0.837,
+      "step": 6541
+    },
+    {
+      "epoch": 1.109737016953512,
+      "grad_norm": 1.609375,
+      "learning_rate": 1.396239441681679e-05,
+      "loss": 0.8559,
+      "step": 6542
+    },
+    {
+      "epoch": 1.1099084810425017,
+      "grad_norm": 1.703125,
+      "learning_rate": 1.396073624429044e-05,
+      "loss": 0.96,
+      "step": 6543
+    },
+    {
+      "epoch": 1.1100799451314916,
+      "grad_norm": 1.6796875,
+      "learning_rate": 1.395907794258957e-05,
+      "loss": 0.894,
+      "step": 6544
+    },
+    {
+      "epoch": 1.1102514092204814,
+      "grad_norm": 1.6015625,
+      "learning_rate": 1.3957419511768257e-05,
+      "loss": 0.8482,
+      "step": 6545
+    },
+    {
+      "epoch": 1.1104228733094712,
+      "grad_norm": 1.640625,
+      "learning_rate": 1.3955760951880595e-05,
+      "loss": 0.8397,
+      "step": 6546
+    },
+    {
+      "epoch": 1.110594337398461,
+      "grad_norm": 1.75,
+      "learning_rate": 1.3954102262980672e-05,
+      "loss": 0.9021,
+      "step": 6547
+    },
+    {
+      "epoch": 1.1107658014874509,
+      "grad_norm": 1.703125,
+      "learning_rate": 1.3952443445122583e-05,
+      "loss": 0.8697,
+      "step": 6548
+    },
+    {
+      "epoch": 1.110937265576441,
+      "grad_norm": 1.6328125,
+      "learning_rate": 1.395078449836043e-05,
+      "loss": 0.8267,
+      "step": 6549
+    },
+    {
+      "epoch": 1.1111087296654307,
+      "grad_norm": 1.6953125,
+      "learning_rate": 1.3949125422748315e-05,
+      "loss": 0.8296,
+      "step": 6550
+    },
+    {
+      "epoch": 1.1112801937544206,
+      "grad_norm": 1.671875,
+      "learning_rate": 1.394746621834035e-05,
+      "loss": 0.9149,
+      "step": 6551
+    },
+    {
+      "epoch": 1.1114516578434104,
+      "grad_norm": 1.703125,
+      "learning_rate": 1.3945806885190651e-05,
+      "loss": 0.7707,
+      "step": 6552
+    },
+    {
+      "epoch": 1.1116231219324002,
+      "grad_norm": 1.6640625,
+      "learning_rate": 1.394414742335333e-05,
+      "loss": 0.8369,
+      "step": 6553
+    },
+    {
+      "epoch": 1.1117945860213903,
+      "grad_norm": 1.578125,
+      "learning_rate": 1.3942487832882503e-05,
+      "loss": 0.8169,
+      "step": 6554
+    },
+    {
+      "epoch": 1.11196605011038,
+      "grad_norm": 1.65625,
+      "learning_rate": 1.3940828113832306e-05,
+      "loss": 0.8029,
+      "step": 6555
+    },
+    {
+      "epoch": 1.11213751419937,
+      "grad_norm": 1.7578125,
+      "learning_rate": 1.3939168266256865e-05,
+      "loss": 0.8772,
+      "step": 6556
+    },
+    {
+      "epoch": 1.1123089782883597,
+      "grad_norm": 1.6640625,
+      "learning_rate": 1.3937508290210313e-05,
+      "loss": 0.7688,
+      "step": 6557
+    },
+    {
+      "epoch": 1.1124804423773496,
+      "grad_norm": 1.6875,
+      "learning_rate": 1.393584818574679e-05,
+      "loss": 0.8588,
+      "step": 6558
+    },
+    {
+      "epoch": 1.1126519064663394,
+      "grad_norm": 1.65625,
+      "learning_rate": 1.3934187952920433e-05,
+      "loss": 0.8125,
+      "step": 6559
+    },
+    {
+      "epoch": 1.1128233705553292,
+      "grad_norm": 1.6484375,
+      "learning_rate": 1.3932527591785394e-05,
+      "loss": 0.9402,
+      "step": 6560
+    },
+    {
+      "epoch": 1.1129948346443193,
+      "grad_norm": 1.7109375,
+      "learning_rate": 1.3930867102395821e-05,
+      "loss": 0.891,
+      "step": 6561
+    },
+    {
+      "epoch": 1.113166298733309,
+      "grad_norm": 1.65625,
+      "learning_rate": 1.3929206484805869e-05,
+      "loss": 0.8461,
+      "step": 6562
+    },
+    {
+      "epoch": 1.113337762822299,
+      "grad_norm": 1.6171875,
+      "learning_rate": 1.3927545739069699e-05,
+      "loss": 0.7938,
+      "step": 6563
+    },
+    {
+      "epoch": 1.1135092269112887,
+      "grad_norm": 22.125,
+      "learning_rate": 1.3925884865241472e-05,
+      "loss": 1.3668,
+      "step": 6564
+    },
+    {
+      "epoch": 1.1136806910002786,
+      "grad_norm": 1.6953125,
+      "learning_rate": 1.3924223863375356e-05,
+      "loss": 0.8056,
+      "step": 6565
+    },
+    {
+      "epoch": 1.1138521550892686,
+      "grad_norm": 1.6171875,
+      "learning_rate": 1.3922562733525522e-05,
+      "loss": 0.8781,
+      "step": 6566
+    },
+    {
+      "epoch": 1.1140236191782584,
+      "grad_norm": 1.671875,
+      "learning_rate": 1.3920901475746147e-05,
+      "loss": 0.7755,
+      "step": 6567
+    },
+    {
+      "epoch": 1.1141950832672483,
+      "grad_norm": 1.7421875,
+      "learning_rate": 1.391924009009141e-05,
+      "loss": 0.8503,
+      "step": 6568
+    },
+    {
+      "epoch": 1.114366547356238,
+      "grad_norm": 1.671875,
+      "learning_rate": 1.3917578576615495e-05,
+      "loss": 0.8424,
+      "step": 6569
+    },
+    {
+      "epoch": 1.114538011445228,
+      "grad_norm": 1.6640625,
+      "learning_rate": 1.3915916935372594e-05,
+      "loss": 0.8289,
+      "step": 6570
+    },
+    {
+      "epoch": 1.1147094755342177,
+      "grad_norm": 1.703125,
+      "learning_rate": 1.391425516641689e-05,
+      "loss": 0.8466,
+      "step": 6571
+    },
+    {
+      "epoch": 1.1148809396232076,
+      "grad_norm": 1.734375,
+      "learning_rate": 1.3912593269802588e-05,
+      "loss": 0.9011,
+      "step": 6572
+    },
+    {
+      "epoch": 1.1150524037121976,
+      "grad_norm": 1.6640625,
+      "learning_rate": 1.3910931245583883e-05,
+      "loss": 0.8765,
+      "step": 6573
+    },
+    {
+      "epoch": 1.1152238678011874,
+      "grad_norm": 1.7109375,
+      "learning_rate": 1.3909269093814985e-05,
+      "loss": 0.8353,
+      "step": 6574
+    },
+    {
+      "epoch": 1.1153953318901773,
+      "grad_norm": 1.6953125,
+      "learning_rate": 1.3907606814550105e-05,
+      "loss": 0.9173,
+      "step": 6575
+    },
+    {
+      "epoch": 1.115566795979167,
+      "grad_norm": 1.546875,
+      "learning_rate": 1.3905944407843447e-05,
+      "loss": 0.8287,
+      "step": 6576
+    },
+    {
+      "epoch": 1.115738260068157,
+      "grad_norm": 1.65625,
+      "learning_rate": 1.3904281873749237e-05,
+      "loss": 0.857,
+      "step": 6577
+    },
+    {
+      "epoch": 1.115909724157147,
+      "grad_norm": 1.75,
+      "learning_rate": 1.3902619212321692e-05,
+      "loss": 0.8939,
+      "step": 6578
+    },
+    {
+      "epoch": 1.1160811882461368,
+      "grad_norm": 1.7734375,
+      "learning_rate": 1.3900956423615035e-05,
+      "loss": 0.8707,
+      "step": 6579
+    },
+    {
+      "epoch": 1.1162526523351266,
+      "grad_norm": 1.6796875,
+      "learning_rate": 1.3899293507683502e-05,
+      "loss": 0.7755,
+      "step": 6580
+    },
+    {
+      "epoch": 1.1164241164241164,
+      "grad_norm": 1.640625,
+      "learning_rate": 1.3897630464581325e-05,
+      "loss": 0.8604,
+      "step": 6581
+    },
+    {
+      "epoch": 1.1165955805131063,
+      "grad_norm": 1.6484375,
+      "learning_rate": 1.389596729436274e-05,
+      "loss": 0.8777,
+      "step": 6582
+    },
+    {
+      "epoch": 1.116767044602096,
+      "grad_norm": 1.625,
+      "learning_rate": 1.3894303997081991e-05,
+      "loss": 0.8515,
+      "step": 6583
+    },
+    {
+      "epoch": 1.116938508691086,
+      "grad_norm": 1.6171875,
+      "learning_rate": 1.3892640572793323e-05,
+      "loss": 0.9125,
+      "step": 6584
+    },
+    {
+      "epoch": 1.117109972780076,
+      "grad_norm": 1.703125,
+      "learning_rate": 1.3890977021550988e-05,
+      "loss": 0.8635,
+      "step": 6585
+    },
+    {
+      "epoch": 1.1172814368690658,
+      "grad_norm": 1.6953125,
+      "learning_rate": 1.3889313343409244e-05,
+      "loss": 0.8413,
+      "step": 6586
+    },
+    {
+      "epoch": 1.1174529009580556,
+      "grad_norm": 1.6171875,
+      "learning_rate": 1.3887649538422344e-05,
+      "loss": 0.8536,
+      "step": 6587
+    },
+    {
+      "epoch": 1.1176243650470454,
+      "grad_norm": 1.78125,
+      "learning_rate": 1.3885985606644552e-05,
+      "loss": 0.872,
+      "step": 6588
+    },
+    {
+      "epoch": 1.1177958291360353,
+      "grad_norm": 1.7109375,
+      "learning_rate": 1.3884321548130136e-05,
+      "loss": 0.8876,
+      "step": 6589
+    },
+    {
+      "epoch": 1.117967293225025,
+      "grad_norm": 1.625,
+      "learning_rate": 1.3882657362933368e-05,
+      "loss": 0.9244,
+      "step": 6590
+    },
+    {
+      "epoch": 1.1181387573140151,
+      "grad_norm": 1.625,
+      "learning_rate": 1.3880993051108522e-05,
+      "loss": 0.868,
+      "step": 6591
+    },
+    {
+      "epoch": 1.118310221403005,
+      "grad_norm": 1.7109375,
+      "learning_rate": 1.3879328612709879e-05,
+      "loss": 0.9156,
+      "step": 6592
+    },
+    {
+      "epoch": 1.1184816854919948,
+      "grad_norm": 1.765625,
+      "learning_rate": 1.387766404779172e-05,
+      "loss": 0.8918,
+      "step": 6593
+    },
+    {
+      "epoch": 1.1186531495809846,
+      "grad_norm": 1.671875,
+      "learning_rate": 1.3875999356408336e-05,
+      "loss": 0.9196,
+      "step": 6594
+    },
+    {
+      "epoch": 1.1188246136699744,
+      "grad_norm": 1.625,
+      "learning_rate": 1.3874334538614016e-05,
+      "loss": 0.8413,
+      "step": 6595
+    },
+    {
+      "epoch": 1.1189960777589643,
+      "grad_norm": 1.7734375,
+      "learning_rate": 1.3872669594463059e-05,
+      "loss": 0.8718,
+      "step": 6596
+    },
+    {
+      "epoch": 1.1191675418479543,
+      "grad_norm": 1.6796875,
+      "learning_rate": 1.3871004524009764e-05,
+      "loss": 0.8636,
+      "step": 6597
+    },
+    {
+      "epoch": 1.1193390059369441,
+      "grad_norm": 1.75,
+      "learning_rate": 1.3869339327308433e-05,
+      "loss": 0.9268,
+      "step": 6598
+    },
+    {
+      "epoch": 1.119510470025934,
+      "grad_norm": 1.6328125,
+      "learning_rate": 1.3867674004413379e-05,
+      "loss": 0.8645,
+      "step": 6599
+    },
+    {
+      "epoch": 1.1196819341149238,
+      "grad_norm": 1.671875,
+      "learning_rate": 1.3866008555378906e-05,
+      "loss": 0.8711,
+      "step": 6600
+    },
+    {
+      "epoch": 1.1198533982039136,
+      "grad_norm": 1.6015625,
+      "learning_rate": 1.3864342980259338e-05,
+      "loss": 0.7778,
+      "step": 6601
+    },
+    {
+      "epoch": 1.1200248622929034,
+      "grad_norm": 1.6953125,
+      "learning_rate": 1.3862677279108993e-05,
+      "loss": 0.9177,
+      "step": 6602
+    },
+    {
+      "epoch": 1.1201963263818935,
+      "grad_norm": 1.6484375,
+      "learning_rate": 1.3861011451982198e-05,
+      "loss": 0.8764,
+      "step": 6603
+    },
+    {
+      "epoch": 1.1203677904708833,
+      "grad_norm": 1.6484375,
+      "learning_rate": 1.3859345498933283e-05,
+      "loss": 0.7912,
+      "step": 6604
+    },
+    {
+      "epoch": 1.1205392545598731,
+      "grad_norm": 1.59375,
+      "learning_rate": 1.3857679420016575e-05,
+      "loss": 0.8659,
+      "step": 6605
+    },
+    {
+      "epoch": 1.120710718648863,
+      "grad_norm": 1.6484375,
+      "learning_rate": 1.3856013215286415e-05,
+      "loss": 0.832,
+      "step": 6606
+    },
+    {
+      "epoch": 1.1208821827378528,
+      "grad_norm": 1.6875,
+      "learning_rate": 1.3854346884797143e-05,
+      "loss": 0.8807,
+      "step": 6607
+    },
+    {
+      "epoch": 1.1210536468268426,
+      "grad_norm": 1.6953125,
+      "learning_rate": 1.3852680428603105e-05,
+      "loss": 0.8882,
+      "step": 6608
+    },
+    {
+      "epoch": 1.1212251109158327,
+      "grad_norm": 1.7109375,
+      "learning_rate": 1.3851013846758651e-05,
+      "loss": 0.9261,
+      "step": 6609
+    },
+    {
+      "epoch": 1.1213965750048225,
+      "grad_norm": 1.609375,
+      "learning_rate": 1.3849347139318134e-05,
+      "loss": 0.8339,
+      "step": 6610
+    },
+    {
+      "epoch": 1.1215680390938123,
+      "grad_norm": 1.6015625,
+      "learning_rate": 1.3847680306335911e-05,
+      "loss": 0.8992,
+      "step": 6611
+    },
+    {
+      "epoch": 1.1217395031828021,
+      "grad_norm": 1.6171875,
+      "learning_rate": 1.3846013347866345e-05,
+      "loss": 0.8608,
+      "step": 6612
+    },
+    {
+      "epoch": 1.121910967271792,
+      "grad_norm": 1.609375,
+      "learning_rate": 1.38443462639638e-05,
+      "loss": 0.922,
+      "step": 6613
+    },
+    {
+      "epoch": 1.1220824313607818,
+      "grad_norm": 1.65625,
+      "learning_rate": 1.3842679054682646e-05,
+      "loss": 0.7879,
+      "step": 6614
+    },
+    {
+      "epoch": 1.1222538954497718,
+      "grad_norm": 1.640625,
+      "learning_rate": 1.3841011720077262e-05,
+      "loss": 0.8357,
+      "step": 6615
+    },
+    {
+      "epoch": 1.1224253595387617,
+      "grad_norm": 1.6640625,
+      "learning_rate": 1.383934426020202e-05,
+      "loss": 0.814,
+      "step": 6616
+    },
+    {
+      "epoch": 1.1225968236277515,
+      "grad_norm": 1.625,
+      "learning_rate": 1.3837676675111303e-05,
+      "loss": 0.8158,
+      "step": 6617
+    },
+    {
+      "epoch": 1.1227682877167413,
+      "grad_norm": 1.65625,
+      "learning_rate": 1.3836008964859497e-05,
+      "loss": 0.8426,
+      "step": 6618
+    },
+    {
+      "epoch": 1.1229397518057311,
+      "grad_norm": 1.75,
+      "learning_rate": 1.3834341129500995e-05,
+      "loss": 0.9509,
+      "step": 6619
+    },
+    {
+      "epoch": 1.123111215894721,
+      "grad_norm": 1.6484375,
+      "learning_rate": 1.3832673169090187e-05,
+      "loss": 0.8784,
+      "step": 6620
+    },
+    {
+      "epoch": 1.123282679983711,
+      "grad_norm": 1.625,
+      "learning_rate": 1.3831005083681479e-05,
+      "loss": 0.8034,
+      "step": 6621
+    },
+    {
+      "epoch": 1.1234541440727008,
+      "grad_norm": 1.6484375,
+      "learning_rate": 1.3829336873329266e-05,
+      "loss": 0.9271,
+      "step": 6622
+    },
+    {
+      "epoch": 1.1236256081616907,
+      "grad_norm": 1.578125,
+      "learning_rate": 1.382766853808796e-05,
+      "loss": 0.8658,
+      "step": 6623
+    },
+    {
+      "epoch": 1.1237970722506805,
+      "grad_norm": 1.6640625,
+      "learning_rate": 1.3826000078011968e-05,
+      "loss": 0.8444,
+      "step": 6624
+    },
+    {
+      "epoch": 1.1239685363396703,
+      "grad_norm": 1.7265625,
+      "learning_rate": 1.3824331493155706e-05,
+      "loss": 0.9042,
+      "step": 6625
+    },
+    {
+      "epoch": 1.1241400004286601,
+      "grad_norm": 1.6640625,
+      "learning_rate": 1.3822662783573595e-05,
+      "loss": 0.7856,
+      "step": 6626
+    },
+    {
+      "epoch": 1.1243114645176502,
+      "grad_norm": 1.6171875,
+      "learning_rate": 1.3820993949320054e-05,
+      "loss": 0.8483,
+      "step": 6627
+    },
+    {
+      "epoch": 1.12448292860664,
+      "grad_norm": 1.59375,
+      "learning_rate": 1.381932499044951e-05,
+      "loss": 0.8497,
+      "step": 6628
+    },
+    {
+      "epoch": 1.1246543926956298,
+      "grad_norm": 1.6328125,
+      "learning_rate": 1.3817655907016399e-05,
+      "loss": 0.896,
+      "step": 6629
+    },
+    {
+      "epoch": 1.1248258567846197,
+      "grad_norm": 1.7265625,
+      "learning_rate": 1.381598669907515e-05,
+      "loss": 0.8452,
+      "step": 6630
+    },
+    {
+      "epoch": 1.1249973208736095,
+      "grad_norm": 1.6953125,
+      "learning_rate": 1.3814317366680206e-05,
+      "loss": 0.8716,
+      "step": 6631
+    },
+    {
+      "epoch": 1.1251687849625993,
+      "grad_norm": 1.7265625,
+      "learning_rate": 1.381264790988601e-05,
+      "loss": 0.882,
+      "step": 6632
+    },
+    {
+      "epoch": 1.1253402490515891,
+      "grad_norm": 1.6640625,
+      "learning_rate": 1.381097832874701e-05,
+      "loss": 0.8736,
+      "step": 6633
+    },
+    {
+      "epoch": 1.1255117131405792,
+      "grad_norm": 1.703125,
+      "learning_rate": 1.3809308623317655e-05,
+      "loss": 0.8683,
+      "step": 6634
+    },
+    {
+      "epoch": 1.125683177229569,
+      "grad_norm": 1.734375,
+      "learning_rate": 1.3807638793652401e-05,
+      "loss": 0.908,
+      "step": 6635
+    },
+    {
+      "epoch": 1.1258546413185588,
+      "grad_norm": 1.6484375,
+      "learning_rate": 1.3805968839805709e-05,
+      "loss": 0.8239,
+      "step": 6636
+    },
+    {
+      "epoch": 1.1260261054075487,
+      "grad_norm": 1.703125,
+      "learning_rate": 1.3804298761832038e-05,
+      "loss": 0.9048,
+      "step": 6637
+    },
+    {
+      "epoch": 1.1261975694965385,
+      "grad_norm": 1.703125,
+      "learning_rate": 1.380262855978586e-05,
+      "loss": 0.8182,
+      "step": 6638
+    },
+    {
+      "epoch": 1.1263690335855285,
+      "grad_norm": 1.6484375,
+      "learning_rate": 1.3800958233721644e-05,
+      "loss": 0.9245,
+      "step": 6639
+    },
+    {
+      "epoch": 1.1265404976745184,
+      "grad_norm": 1.65625,
+      "learning_rate": 1.379928778369387e-05,
+      "loss": 0.8484,
+      "step": 6640
+    },
+    {
+      "epoch": 1.1267119617635082,
+      "grad_norm": 1.6171875,
+      "learning_rate": 1.3797617209757013e-05,
+      "loss": 0.8341,
+      "step": 6641
+    },
+    {
+      "epoch": 1.126883425852498,
+      "grad_norm": 1.65625,
+      "learning_rate": 1.3795946511965557e-05,
+      "loss": 0.8392,
+      "step": 6642
+    },
+    {
+      "epoch": 1.1270548899414878,
+      "grad_norm": 1.6171875,
+      "learning_rate": 1.3794275690373993e-05,
+      "loss": 0.9658,
+      "step": 6643
+    },
+    {
+      "epoch": 1.1272263540304777,
+      "grad_norm": 1.609375,
+      "learning_rate": 1.379260474503681e-05,
+      "loss": 0.8144,
+      "step": 6644
+    },
+    {
+      "epoch": 1.1273978181194675,
+      "grad_norm": 1.7421875,
+      "learning_rate": 1.3790933676008505e-05,
+      "loss": 0.8777,
+      "step": 6645
+    },
+    {
+      "epoch": 1.1275692822084575,
+      "grad_norm": 1.6953125,
+      "learning_rate": 1.3789262483343578e-05,
+      "loss": 0.8284,
+      "step": 6646
+    },
+    {
+      "epoch": 1.1277407462974474,
+      "grad_norm": 2.03125,
+      "learning_rate": 1.3787591167096528e-05,
+      "loss": 0.846,
+      "step": 6647
+    },
+    {
+      "epoch": 1.1279122103864372,
+      "grad_norm": 1.609375,
+      "learning_rate": 1.3785919727321872e-05,
+      "loss": 0.8,
+      "step": 6648
+    },
+    {
+      "epoch": 1.128083674475427,
+      "grad_norm": 1.609375,
+      "learning_rate": 1.3784248164074116e-05,
+      "loss": 0.9209,
+      "step": 6649
+    },
+    {
+      "epoch": 1.1282551385644168,
+      "grad_norm": 1.625,
+      "learning_rate": 1.3782576477407778e-05,
+      "loss": 0.8437,
+      "step": 6650
+    },
+    {
+      "epoch": 1.1284266026534069,
+      "grad_norm": 1.7109375,
+      "learning_rate": 1.3780904667377377e-05,
+      "loss": 0.9282,
+      "step": 6651
+    },
+    {
+      "epoch": 1.1285980667423967,
+      "grad_norm": 1.6328125,
+      "learning_rate": 1.3779232734037437e-05,
+      "loss": 0.8401,
+      "step": 6652
+    },
+    {
+      "epoch": 1.1287695308313865,
+      "grad_norm": 1.7890625,
+      "learning_rate": 1.3777560677442485e-05,
+      "loss": 0.986,
+      "step": 6653
+    },
+    {
+      "epoch": 1.1289409949203764,
+      "grad_norm": 1.7265625,
+      "learning_rate": 1.3775888497647056e-05,
+      "loss": 0.8463,
+      "step": 6654
+    },
+    {
+      "epoch": 1.1291124590093662,
+      "grad_norm": 1.71875,
+      "learning_rate": 1.3774216194705683e-05,
+      "loss": 0.8735,
+      "step": 6655
+    },
+    {
+      "epoch": 1.129283923098356,
+      "grad_norm": 1.671875,
+      "learning_rate": 1.377254376867291e-05,
+      "loss": 0.8773,
+      "step": 6656
+    },
+    {
+      "epoch": 1.1294553871873458,
+      "grad_norm": 1.640625,
+      "learning_rate": 1.3770871219603276e-05,
+      "loss": 0.9427,
+      "step": 6657
+    },
+    {
+      "epoch": 1.1296268512763359,
+      "grad_norm": 1.640625,
+      "learning_rate": 1.3769198547551333e-05,
+      "loss": 0.8364,
+      "step": 6658
+    },
+    {
+      "epoch": 1.1297983153653257,
+      "grad_norm": 1.65625,
+      "learning_rate": 1.3767525752571631e-05,
+      "loss": 0.835,
+      "step": 6659
+    },
+    {
+      "epoch": 1.1299697794543155,
+      "grad_norm": 1.640625,
+      "learning_rate": 1.3765852834718727e-05,
+      "loss": 0.9108,
+      "step": 6660
+    },
+    {
+      "epoch": 1.1301412435433054,
+      "grad_norm": 1.609375,
+      "learning_rate": 1.3764179794047183e-05,
+      "loss": 0.8486,
+      "step": 6661
+    },
+    {
+      "epoch": 1.1303127076322952,
+      "grad_norm": 1.59375,
+      "learning_rate": 1.376250663061156e-05,
+      "loss": 0.8227,
+      "step": 6662
+    },
+    {
+      "epoch": 1.1304841717212852,
+      "grad_norm": 1.5859375,
+      "learning_rate": 1.3760833344466428e-05,
+      "loss": 0.8379,
+      "step": 6663
+    },
+    {
+      "epoch": 1.130655635810275,
+      "grad_norm": 1.7578125,
+      "learning_rate": 1.3759159935666358e-05,
+      "loss": 0.8692,
+      "step": 6664
+    },
+    {
+      "epoch": 1.1308270998992649,
+      "grad_norm": 1.671875,
+      "learning_rate": 1.3757486404265927e-05,
+      "loss": 0.8368,
+      "step": 6665
+    },
+    {
+      "epoch": 1.1309985639882547,
+      "grad_norm": 1.625,
+      "learning_rate": 1.3755812750319716e-05,
+      "loss": 0.793,
+      "step": 6666
+    },
+    {
+      "epoch": 1.1311700280772445,
+      "grad_norm": 1.65625,
+      "learning_rate": 1.3754138973882304e-05,
+      "loss": 0.8666,
+      "step": 6667
+    },
+    {
+      "epoch": 1.1313414921662344,
+      "grad_norm": 1.75,
+      "learning_rate": 1.3752465075008288e-05,
+      "loss": 0.909,
+      "step": 6668
+    },
+    {
+      "epoch": 1.1315129562552242,
+      "grad_norm": 1.703125,
+      "learning_rate": 1.3750791053752254e-05,
+      "loss": 0.9476,
+      "step": 6669
+    },
+    {
+      "epoch": 1.1316844203442142,
+      "grad_norm": 1.6171875,
+      "learning_rate": 1.3749116910168798e-05,
+      "loss": 0.8361,
+      "step": 6670
+    },
+    {
+      "epoch": 1.131855884433204,
+      "grad_norm": 1.6875,
+      "learning_rate": 1.3747442644312523e-05,
+      "loss": 0.9592,
+      "step": 6671
+    },
+    {
+      "epoch": 1.1320273485221939,
+      "grad_norm": 1.671875,
+      "learning_rate": 1.3745768256238033e-05,
+      "loss": 0.8749,
+      "step": 6672
+    },
+    {
+      "epoch": 1.1321988126111837,
+      "grad_norm": 1.6015625,
+      "learning_rate": 1.3744093745999935e-05,
+      "loss": 0.7972,
+      "step": 6673
+    },
+    {
+      "epoch": 1.1323702767001735,
+      "grad_norm": 1.6796875,
+      "learning_rate": 1.374241911365284e-05,
+      "loss": 0.8684,
+      "step": 6674
+    },
+    {
+      "epoch": 1.1325417407891636,
+      "grad_norm": 1.6640625,
+      "learning_rate": 1.3740744359251363e-05,
+      "loss": 0.7984,
+      "step": 6675
+    },
+    {
+      "epoch": 1.1327132048781534,
+      "grad_norm": 1.7421875,
+      "learning_rate": 1.3739069482850128e-05,
+      "loss": 0.8367,
+      "step": 6676
+    },
+    {
+      "epoch": 1.1328846689671432,
+      "grad_norm": 1.703125,
+      "learning_rate": 1.3737394484503756e-05,
+      "loss": 0.8666,
+      "step": 6677
+    },
+    {
+      "epoch": 1.133056133056133,
+      "grad_norm": 1.65625,
+      "learning_rate": 1.3735719364266878e-05,
+      "loss": 0.9206,
+      "step": 6678
+    },
+    {
+      "epoch": 1.133227597145123,
+      "grad_norm": 1.671875,
+      "learning_rate": 1.3734044122194126e-05,
+      "loss": 0.8831,
+      "step": 6679
+    },
+    {
+      "epoch": 1.1333990612341127,
+      "grad_norm": 1.6171875,
+      "learning_rate": 1.373236875834013e-05,
+      "loss": 0.8538,
+      "step": 6680
+    },
+    {
+      "epoch": 1.1335705253231025,
+      "grad_norm": 1.734375,
+      "learning_rate": 1.3730693272759537e-05,
+      "loss": 0.95,
+      "step": 6681
+    },
+    {
+      "epoch": 1.1337419894120926,
+      "grad_norm": 1.7265625,
+      "learning_rate": 1.3729017665506985e-05,
+      "loss": 0.8747,
+      "step": 6682
+    },
+    {
+      "epoch": 1.1339134535010824,
+      "grad_norm": 1.7421875,
+      "learning_rate": 1.3727341936637126e-05,
+      "loss": 0.8905,
+      "step": 6683
+    },
+    {
+      "epoch": 1.1340849175900722,
+      "grad_norm": 1.6796875,
+      "learning_rate": 1.372566608620461e-05,
+      "loss": 0.8614,
+      "step": 6684
+    },
+    {
+      "epoch": 1.134256381679062,
+      "grad_norm": 1.5546875,
+      "learning_rate": 1.3723990114264094e-05,
+      "loss": 0.8072,
+      "step": 6685
+    },
+    {
+      "epoch": 1.134427845768052,
+      "grad_norm": 1.6484375,
+      "learning_rate": 1.3722314020870237e-05,
+      "loss": 0.798,
+      "step": 6686
+    },
+    {
+      "epoch": 1.134599309857042,
+      "grad_norm": 1.7578125,
+      "learning_rate": 1.3720637806077703e-05,
+      "loss": 0.9451,
+      "step": 6687
+    },
+    {
+      "epoch": 1.1347707739460318,
+      "grad_norm": 1.71875,
+      "learning_rate": 1.3718961469941159e-05,
+      "loss": 0.8844,
+      "step": 6688
+    },
+    {
+      "epoch": 1.1349422380350216,
+      "grad_norm": 1.671875,
+      "learning_rate": 1.3717285012515278e-05,
+      "loss": 0.7849,
+      "step": 6689
+    },
+    {
+      "epoch": 1.1351137021240114,
+      "grad_norm": 1.65625,
+      "learning_rate": 1.3715608433854738e-05,
+      "loss": 0.9024,
+      "step": 6690
+    },
+    {
+      "epoch": 1.1352851662130012,
+      "grad_norm": 1.6875,
+      "learning_rate": 1.371393173401421e-05,
+      "loss": 0.8926,
+      "step": 6691
+    },
+    {
+      "epoch": 1.135456630301991,
+      "grad_norm": 1.7734375,
+      "learning_rate": 1.3712254913048384e-05,
+      "loss": 0.8341,
+      "step": 6692
+    },
+    {
+      "epoch": 1.135628094390981,
+      "grad_norm": 1.7109375,
+      "learning_rate": 1.3710577971011948e-05,
+      "loss": 0.934,
+      "step": 6693
+    },
+    {
+      "epoch": 1.135799558479971,
+      "grad_norm": 1.671875,
+      "learning_rate": 1.3708900907959585e-05,
+      "loss": 0.847,
+      "step": 6694
+    },
+    {
+      "epoch": 1.1359710225689608,
+      "grad_norm": 1.5703125,
+      "learning_rate": 1.3707223723946003e-05,
+      "loss": 0.827,
+      "step": 6695
+    },
+    {
+      "epoch": 1.1361424866579506,
+      "grad_norm": 1.65625,
+      "learning_rate": 1.3705546419025897e-05,
+      "loss": 0.8121,
+      "step": 6696
+    },
+    {
+      "epoch": 1.1363139507469404,
+      "grad_norm": 1.578125,
+      "learning_rate": 1.3703868993253967e-05,
+      "loss": 0.7692,
+      "step": 6697
+    },
+    {
+      "epoch": 1.1364854148359302,
+      "grad_norm": 1.6328125,
+      "learning_rate": 1.370219144668492e-05,
+      "loss": 0.8743,
+      "step": 6698
+    },
+    {
+      "epoch": 1.13665687892492,
+      "grad_norm": 1.6484375,
+      "learning_rate": 1.3700513779373467e-05,
+      "loss": 0.9635,
+      "step": 6699
+    },
+    {
+      "epoch": 1.1368283430139101,
+      "grad_norm": 1.7265625,
+      "learning_rate": 1.3698835991374328e-05,
+      "loss": 0.9609,
+      "step": 6700
+    },
+    {
+      "epoch": 1.1369998071029,
+      "grad_norm": 1.6640625,
+      "learning_rate": 1.3697158082742219e-05,
+      "loss": 0.8923,
+      "step": 6701
+    },
+    {
+      "epoch": 1.1371712711918898,
+      "grad_norm": 1.65625,
+      "learning_rate": 1.3695480053531864e-05,
+      "loss": 0.809,
+      "step": 6702
+    },
+    {
+      "epoch": 1.1373427352808796,
+      "grad_norm": 1.625,
+      "learning_rate": 1.3693801903797984e-05,
+      "loss": 0.8371,
+      "step": 6703
+    },
+    {
+      "epoch": 1.1375141993698694,
+      "grad_norm": 1.6875,
+      "learning_rate": 1.3692123633595315e-05,
+      "loss": 0.9109,
+      "step": 6704
+    },
+    {
+      "epoch": 1.1376856634588592,
+      "grad_norm": 1.7109375,
+      "learning_rate": 1.3690445242978594e-05,
+      "loss": 0.9534,
+      "step": 6705
+    },
+    {
+      "epoch": 1.1378571275478493,
+      "grad_norm": 1.640625,
+      "learning_rate": 1.3688766732002555e-05,
+      "loss": 0.8689,
+      "step": 6706
+    },
+    {
+      "epoch": 1.1380285916368391,
+      "grad_norm": 1.7109375,
+      "learning_rate": 1.3687088100721942e-05,
+      "loss": 0.8972,
+      "step": 6707
+    },
+    {
+      "epoch": 1.138200055725829,
+      "grad_norm": 1.71875,
+      "learning_rate": 1.3685409349191505e-05,
+      "loss": 0.8928,
+      "step": 6708
+    },
+    {
+      "epoch": 1.1383715198148188,
+      "grad_norm": 1.5859375,
+      "learning_rate": 1.3683730477465989e-05,
+      "loss": 0.882,
+      "step": 6709
+    },
+    {
+      "epoch": 1.1385429839038086,
+      "grad_norm": 1.75,
+      "learning_rate": 1.3682051485600148e-05,
+      "loss": 0.9091,
+      "step": 6710
+    },
+    {
+      "epoch": 1.1387144479927984,
+      "grad_norm": 1.6015625,
+      "learning_rate": 1.3680372373648744e-05,
+      "loss": 0.795,
+      "step": 6711
+    },
+    {
+      "epoch": 1.1388859120817885,
+      "grad_norm": 1.6953125,
+      "learning_rate": 1.3678693141666539e-05,
+      "loss": 0.9273,
+      "step": 6712
+    },
+    {
+      "epoch": 1.1390573761707783,
+      "grad_norm": 1.6640625,
+      "learning_rate": 1.3677013789708297e-05,
+      "loss": 0.9238,
+      "step": 6713
+    },
+    {
+      "epoch": 1.1392288402597681,
+      "grad_norm": 1.6328125,
+      "learning_rate": 1.367533431782879e-05,
+      "loss": 0.8532,
+      "step": 6714
+    },
+    {
+      "epoch": 1.139400304348758,
+      "grad_norm": 1.65625,
+      "learning_rate": 1.367365472608279e-05,
+      "loss": 0.8438,
+      "step": 6715
+    },
+    {
+      "epoch": 1.1395717684377478,
+      "grad_norm": 1.6875,
+      "learning_rate": 1.3671975014525076e-05,
+      "loss": 0.8949,
+      "step": 6716
+    },
+    {
+      "epoch": 1.1397432325267376,
+      "grad_norm": 1.7578125,
+      "learning_rate": 1.3670295183210431e-05,
+      "loss": 0.8876,
+      "step": 6717
+    },
+    {
+      "epoch": 1.1399146966157276,
+      "grad_norm": 1.6015625,
+      "learning_rate": 1.3668615232193636e-05,
+      "loss": 0.864,
+      "step": 6718
+    },
+    {
+      "epoch": 1.1400861607047175,
+      "grad_norm": 1.6328125,
+      "learning_rate": 1.3666935161529487e-05,
+      "loss": 0.8316,
+      "step": 6719
+    },
+    {
+      "epoch": 1.1402576247937073,
+      "grad_norm": 1.7578125,
+      "learning_rate": 1.3665254971272772e-05,
+      "loss": 0.9238,
+      "step": 6720
+    },
+    {
+      "epoch": 1.1404290888826971,
+      "grad_norm": 1.7890625,
+      "learning_rate": 1.3663574661478291e-05,
+      "loss": 0.9234,
+      "step": 6721
+    },
+    {
+      "epoch": 1.140600552971687,
+      "grad_norm": 1.796875,
+      "learning_rate": 1.366189423220084e-05,
+      "loss": 0.9048,
+      "step": 6722
+    },
+    {
+      "epoch": 1.1407720170606768,
+      "grad_norm": 1.7421875,
+      "learning_rate": 1.3660213683495233e-05,
+      "loss": 0.9594,
+      "step": 6723
+    },
+    {
+      "epoch": 1.1409434811496668,
+      "grad_norm": 1.6328125,
+      "learning_rate": 1.3658533015416276e-05,
+      "loss": 0.8342,
+      "step": 6724
+    },
+    {
+      "epoch": 1.1411149452386566,
+      "grad_norm": 1.6796875,
+      "learning_rate": 1.3656852228018779e-05,
+      "loss": 0.8728,
+      "step": 6725
+    },
+    {
+      "epoch": 1.1412864093276465,
+      "grad_norm": 1.640625,
+      "learning_rate": 1.3655171321357561e-05,
+      "loss": 0.8813,
+      "step": 6726
+    },
+    {
+      "epoch": 1.1414578734166363,
+      "grad_norm": 1.7109375,
+      "learning_rate": 1.3653490295487442e-05,
+      "loss": 0.8986,
+      "step": 6727
+    },
+    {
+      "epoch": 1.1416293375056261,
+      "grad_norm": 1.75,
+      "learning_rate": 1.3651809150463246e-05,
+      "loss": 0.9289,
+      "step": 6728
+    },
+    {
+      "epoch": 1.141800801594616,
+      "grad_norm": 1.6640625,
+      "learning_rate": 1.3650127886339801e-05,
+      "loss": 0.8557,
+      "step": 6729
+    },
+    {
+      "epoch": 1.1419722656836058,
+      "grad_norm": 1.640625,
+      "learning_rate": 1.3648446503171942e-05,
+      "loss": 0.9425,
+      "step": 6730
+    },
+    {
+      "epoch": 1.1421437297725958,
+      "grad_norm": 1.609375,
+      "learning_rate": 1.3646765001014504e-05,
+      "loss": 0.8632,
+      "step": 6731
+    },
+    {
+      "epoch": 1.1423151938615856,
+      "grad_norm": 1.625,
+      "learning_rate": 1.3645083379922327e-05,
+      "loss": 0.7903,
+      "step": 6732
+    },
+    {
+      "epoch": 1.1424866579505755,
+      "grad_norm": 1.71875,
+      "learning_rate": 1.3643401639950253e-05,
+      "loss": 0.9099,
+      "step": 6733
+    },
+    {
+      "epoch": 1.1426581220395653,
+      "grad_norm": 1.6953125,
+      "learning_rate": 1.3641719781153132e-05,
+      "loss": 0.8892,
+      "step": 6734
+    },
+    {
+      "epoch": 1.1428295861285551,
+      "grad_norm": 1.671875,
+      "learning_rate": 1.3640037803585818e-05,
+      "loss": 0.843,
+      "step": 6735
+    },
+    {
+      "epoch": 1.1430010502175452,
+      "grad_norm": 1.7265625,
+      "learning_rate": 1.3638355707303163e-05,
+      "loss": 0.8157,
+      "step": 6736
+    },
+    {
+      "epoch": 1.143172514306535,
+      "grad_norm": 1.5703125,
+      "learning_rate": 1.3636673492360029e-05,
+      "loss": 0.8234,
+      "step": 6737
+    },
+    {
+      "epoch": 1.1433439783955248,
+      "grad_norm": 1.6796875,
+      "learning_rate": 1.3634991158811276e-05,
+      "loss": 0.8297,
+      "step": 6738
+    },
+    {
+      "epoch": 1.1435154424845146,
+      "grad_norm": 1.7578125,
+      "learning_rate": 1.3633308706711772e-05,
+      "loss": 0.9657,
+      "step": 6739
+    },
+    {
+      "epoch": 1.1436869065735045,
+      "grad_norm": 1.5390625,
+      "learning_rate": 1.3631626136116392e-05,
+      "loss": 0.7461,
+      "step": 6740
+    },
+    {
+      "epoch": 1.1438583706624943,
+      "grad_norm": 1.6484375,
+      "learning_rate": 1.3629943447080003e-05,
+      "loss": 0.8086,
+      "step": 6741
+    },
+    {
+      "epoch": 1.1440298347514841,
+      "grad_norm": 1.7265625,
+      "learning_rate": 1.3628260639657496e-05,
+      "loss": 0.8388,
+      "step": 6742
+    },
+    {
+      "epoch": 1.1442012988404742,
+      "grad_norm": 1.6953125,
+      "learning_rate": 1.3626577713903744e-05,
+      "loss": 0.9211,
+      "step": 6743
+    },
+    {
+      "epoch": 1.144372762929464,
+      "grad_norm": 1.640625,
+      "learning_rate": 1.3624894669873636e-05,
+      "loss": 0.8763,
+      "step": 6744
+    },
+    {
+      "epoch": 1.1445442270184538,
+      "grad_norm": 1.84375,
+      "learning_rate": 1.3623211507622063e-05,
+      "loss": 0.9707,
+      "step": 6745
+    },
+    {
+      "epoch": 1.1447156911074436,
+      "grad_norm": 1.5859375,
+      "learning_rate": 1.3621528227203918e-05,
+      "loss": 0.841,
+      "step": 6746
+    },
+    {
+      "epoch": 1.1448871551964335,
+      "grad_norm": 1.609375,
+      "learning_rate": 1.3619844828674101e-05,
+      "loss": 0.8323,
+      "step": 6747
+    },
+    {
+      "epoch": 1.1450586192854235,
+      "grad_norm": 1.7109375,
+      "learning_rate": 1.3618161312087515e-05,
+      "loss": 0.838,
+      "step": 6748
+    },
+    {
+      "epoch": 1.1452300833744133,
+      "grad_norm": 1.6953125,
+      "learning_rate": 1.361647767749906e-05,
+      "loss": 0.8694,
+      "step": 6749
+    },
+    {
+      "epoch": 1.1454015474634032,
+      "grad_norm": 1.6953125,
+      "learning_rate": 1.3614793924963649e-05,
+      "loss": 0.8692,
+      "step": 6750
+    },
+    {
+      "epoch": 1.145573011552393,
+      "grad_norm": 1.6953125,
+      "learning_rate": 1.36131100545362e-05,
+      "loss": 0.8843,
+      "step": 6751
+    },
+    {
+      "epoch": 1.1457444756413828,
+      "grad_norm": 1.765625,
+      "learning_rate": 1.3611426066271625e-05,
+      "loss": 0.9358,
+      "step": 6752
+    },
+    {
+      "epoch": 1.1459159397303726,
+      "grad_norm": 1.625,
+      "learning_rate": 1.3609741960224847e-05,
+      "loss": 0.7755,
+      "step": 6753
+    },
+    {
+      "epoch": 1.1460874038193625,
+      "grad_norm": 1.6953125,
+      "learning_rate": 1.3608057736450792e-05,
+      "loss": 0.9001,
+      "step": 6754
+    },
+    {
+      "epoch": 1.1462588679083525,
+      "grad_norm": 1.671875,
+      "learning_rate": 1.3606373395004384e-05,
+      "loss": 0.9319,
+      "step": 6755
+    },
+    {
+      "epoch": 1.1464303319973423,
+      "grad_norm": 1.6171875,
+      "learning_rate": 1.3604688935940562e-05,
+      "loss": 0.8448,
+      "step": 6756
+    },
+    {
+      "epoch": 1.1466017960863322,
+      "grad_norm": 1.6484375,
+      "learning_rate": 1.3603004359314259e-05,
+      "loss": 0.8968,
+      "step": 6757
+    },
+    {
+      "epoch": 1.146773260175322,
+      "grad_norm": 1.703125,
+      "learning_rate": 1.3601319665180415e-05,
+      "loss": 0.8989,
+      "step": 6758
+    },
+    {
+      "epoch": 1.1469447242643118,
+      "grad_norm": 1.640625,
+      "learning_rate": 1.3599634853593977e-05,
+      "loss": 0.896,
+      "step": 6759
+    },
+    {
+      "epoch": 1.1471161883533019,
+      "grad_norm": 1.7734375,
+      "learning_rate": 1.359794992460989e-05,
+      "loss": 0.9144,
+      "step": 6760
+    },
+    {
+      "epoch": 1.1472876524422917,
+      "grad_norm": 1.8203125,
+      "learning_rate": 1.3596264878283107e-05,
+      "loss": 0.8946,
+      "step": 6761
+    },
+    {
+      "epoch": 1.1474591165312815,
+      "grad_norm": 1.640625,
+      "learning_rate": 1.3594579714668585e-05,
+      "loss": 0.8703,
+      "step": 6762
+    },
+    {
+      "epoch": 1.1476305806202713,
+      "grad_norm": 1.7265625,
+      "learning_rate": 1.3592894433821284e-05,
+      "loss": 0.9466,
+      "step": 6763
+    },
+    {
+      "epoch": 1.1478020447092612,
+      "grad_norm": 1.75,
+      "learning_rate": 1.3591209035796164e-05,
+      "loss": 0.8969,
+      "step": 6764
+    },
+    {
+      "epoch": 1.147973508798251,
+      "grad_norm": 1.640625,
+      "learning_rate": 1.3589523520648195e-05,
+      "loss": 0.9025,
+      "step": 6765
+    },
+    {
+      "epoch": 1.1481449728872408,
+      "grad_norm": 1.578125,
+      "learning_rate": 1.3587837888432348e-05,
+      "loss": 0.8256,
+      "step": 6766
+    },
+    {
+      "epoch": 1.1483164369762309,
+      "grad_norm": 1.6015625,
+      "learning_rate": 1.3586152139203596e-05,
+      "loss": 0.8648,
+      "step": 6767
+    },
+    {
+      "epoch": 1.1484879010652207,
+      "grad_norm": 1.703125,
+      "learning_rate": 1.3584466273016919e-05,
+      "loss": 0.8814,
+      "step": 6768
+    },
+    {
+      "epoch": 1.1486593651542105,
+      "grad_norm": 1.6796875,
+      "learning_rate": 1.3582780289927296e-05,
+      "loss": 0.9134,
+      "step": 6769
+    },
+    {
+      "epoch": 1.1488308292432003,
+      "grad_norm": 1.703125,
+      "learning_rate": 1.3581094189989718e-05,
+      "loss": 0.8339,
+      "step": 6770
+    },
+    {
+      "epoch": 1.1490022933321902,
+      "grad_norm": 1.71875,
+      "learning_rate": 1.3579407973259176e-05,
+      "loss": 0.8392,
+      "step": 6771
+    },
+    {
+      "epoch": 1.1491737574211802,
+      "grad_norm": 1.6171875,
+      "learning_rate": 1.3577721639790658e-05,
+      "loss": 0.8614,
+      "step": 6772
+    },
+    {
+      "epoch": 1.14934522151017,
+      "grad_norm": 1.625,
+      "learning_rate": 1.3576035189639165e-05,
+      "loss": 0.7583,
+      "step": 6773
+    },
+    {
+      "epoch": 1.1495166855991599,
+      "grad_norm": 1.6171875,
+      "learning_rate": 1.35743486228597e-05,
+      "loss": 0.8958,
+      "step": 6774
+    },
+    {
+      "epoch": 1.1496881496881497,
+      "grad_norm": 1.6640625,
+      "learning_rate": 1.3572661939507266e-05,
+      "loss": 0.8706,
+      "step": 6775
+    },
+    {
+      "epoch": 1.1498596137771395,
+      "grad_norm": 1.65625,
+      "learning_rate": 1.3570975139636872e-05,
+      "loss": 0.8905,
+      "step": 6776
+    },
+    {
+      "epoch": 1.1500310778661293,
+      "grad_norm": 1.6953125,
+      "learning_rate": 1.3569288223303531e-05,
+      "loss": 0.876,
+      "step": 6777
+    },
+    {
+      "epoch": 1.1502025419551192,
+      "grad_norm": 1.6953125,
+      "learning_rate": 1.3567601190562264e-05,
+      "loss": 0.818,
+      "step": 6778
+    },
+    {
+      "epoch": 1.1503740060441092,
+      "grad_norm": 1.7109375,
+      "learning_rate": 1.3565914041468085e-05,
+      "loss": 0.9179,
+      "step": 6779
+    },
+    {
+      "epoch": 1.150545470133099,
+      "grad_norm": 1.71875,
+      "learning_rate": 1.3564226776076022e-05,
+      "loss": 0.8875,
+      "step": 6780
+    },
+    {
+      "epoch": 1.1507169342220889,
+      "grad_norm": 1.578125,
+      "learning_rate": 1.3562539394441101e-05,
+      "loss": 0.875,
+      "step": 6781
+    },
+    {
+      "epoch": 1.1508883983110787,
+      "grad_norm": 1.6328125,
+      "learning_rate": 1.3560851896618354e-05,
+      "loss": 0.8554,
+      "step": 6782
+    },
+    {
+      "epoch": 1.1510598624000685,
+      "grad_norm": 1.78125,
+      "learning_rate": 1.3559164282662821e-05,
+      "loss": 0.8489,
+      "step": 6783
+    },
+    {
+      "epoch": 1.1512313264890586,
+      "grad_norm": 1.609375,
+      "learning_rate": 1.3557476552629538e-05,
+      "loss": 0.8179,
+      "step": 6784
+    },
+    {
+      "epoch": 1.1514027905780484,
+      "grad_norm": 1.640625,
+      "learning_rate": 1.3555788706573547e-05,
+      "loss": 0.9496,
+      "step": 6785
+    },
+    {
+      "epoch": 1.1515742546670382,
+      "grad_norm": 1.625,
+      "learning_rate": 1.3554100744549899e-05,
+      "loss": 0.8999,
+      "step": 6786
+    },
+    {
+      "epoch": 1.151745718756028,
+      "grad_norm": 1.6640625,
+      "learning_rate": 1.3552412666613637e-05,
+      "loss": 0.816,
+      "step": 6787
+    },
+    {
+      "epoch": 1.1519171828450179,
+      "grad_norm": 1.6796875,
+      "learning_rate": 1.3550724472819825e-05,
+      "loss": 0.8914,
+      "step": 6788
+    },
+    {
+      "epoch": 1.1520886469340077,
+      "grad_norm": 1.6796875,
+      "learning_rate": 1.3549036163223515e-05,
+      "loss": 0.8921,
+      "step": 6789
+    },
+    {
+      "epoch": 1.1522601110229975,
+      "grad_norm": 1.6171875,
+      "learning_rate": 1.3547347737879772e-05,
+      "loss": 0.8957,
+      "step": 6790
+    },
+    {
+      "epoch": 1.1524315751119876,
+      "grad_norm": 1.578125,
+      "learning_rate": 1.3545659196843661e-05,
+      "loss": 0.8737,
+      "step": 6791
+    },
+    {
+      "epoch": 1.1526030392009774,
+      "grad_norm": 1.671875,
+      "learning_rate": 1.3543970540170253e-05,
+      "loss": 0.8306,
+      "step": 6792
+    },
+    {
+      "epoch": 1.1527745032899672,
+      "grad_norm": 1.765625,
+      "learning_rate": 1.3542281767914617e-05,
+      "loss": 0.9172,
+      "step": 6793
+    },
+    {
+      "epoch": 1.152945967378957,
+      "grad_norm": 1.703125,
+      "learning_rate": 1.3540592880131839e-05,
+      "loss": 0.8589,
+      "step": 6794
+    },
+    {
+      "epoch": 1.1531174314679469,
+      "grad_norm": 1.625,
+      "learning_rate": 1.3538903876876993e-05,
+      "loss": 0.8532,
+      "step": 6795
+    },
+    {
+      "epoch": 1.1532888955569367,
+      "grad_norm": 1.78125,
+      "learning_rate": 1.353721475820516e-05,
+      "loss": 1.0322,
+      "step": 6796
+    },
+    {
+      "epoch": 1.1534603596459267,
+      "grad_norm": 1.6875,
+      "learning_rate": 1.3535525524171438e-05,
+      "loss": 0.8166,
+      "step": 6797
+    },
+    {
+      "epoch": 1.1536318237349166,
+      "grad_norm": 1.6328125,
+      "learning_rate": 1.3533836174830915e-05,
+      "loss": 0.8826,
+      "step": 6798
+    },
+    {
+      "epoch": 1.1538032878239064,
+      "grad_norm": 1.765625,
+      "learning_rate": 1.3532146710238684e-05,
+      "loss": 0.8866,
+      "step": 6799
+    },
+    {
+      "epoch": 1.1539747519128962,
+      "grad_norm": 1.609375,
+      "learning_rate": 1.3530457130449855e-05,
+      "loss": 0.9088,
+      "step": 6800
+    },
+    {
+      "epoch": 1.154146216001886,
+      "grad_norm": 1.6640625,
+      "learning_rate": 1.3528767435519521e-05,
+      "loss": 0.8938,
+      "step": 6801
+    },
+    {
+      "epoch": 1.1543176800908759,
+      "grad_norm": 1.6328125,
+      "learning_rate": 1.3527077625502791e-05,
+      "loss": 0.9035,
+      "step": 6802
+    },
+    {
+      "epoch": 1.154489144179866,
+      "grad_norm": 1.59375,
+      "learning_rate": 1.3525387700454779e-05,
+      "loss": 0.8444,
+      "step": 6803
+    },
+    {
+      "epoch": 1.1546606082688557,
+      "grad_norm": 1.625,
+      "learning_rate": 1.3523697660430599e-05,
+      "loss": 0.9083,
+      "step": 6804
+    },
+    {
+      "epoch": 1.1548320723578456,
+      "grad_norm": 1.734375,
+      "learning_rate": 1.3522007505485368e-05,
+      "loss": 0.8714,
+      "step": 6805
+    },
+    {
+      "epoch": 1.1550035364468354,
+      "grad_norm": 1.640625,
+      "learning_rate": 1.3520317235674212e-05,
+      "loss": 0.8453,
+      "step": 6806
+    },
+    {
+      "epoch": 1.1551750005358252,
+      "grad_norm": 1.7421875,
+      "learning_rate": 1.3518626851052251e-05,
+      "loss": 0.8262,
+      "step": 6807
+    },
+    {
+      "epoch": 1.155346464624815,
+      "grad_norm": 1.75,
+      "learning_rate": 1.3516936351674623e-05,
+      "loss": 0.8558,
+      "step": 6808
+    },
+    {
+      "epoch": 1.155517928713805,
+      "grad_norm": 1.6875,
+      "learning_rate": 1.3515245737596453e-05,
+      "loss": 0.8869,
+      "step": 6809
+    },
+    {
+      "epoch": 1.155689392802795,
+      "grad_norm": 1.6640625,
+      "learning_rate": 1.3513555008872884e-05,
+      "loss": 0.8404,
+      "step": 6810
+    },
+    {
+      "epoch": 1.1558608568917847,
+      "grad_norm": 1.7421875,
+      "learning_rate": 1.3511864165559056e-05,
+      "loss": 0.8585,
+      "step": 6811
+    },
+    {
+      "epoch": 1.1560323209807746,
+      "grad_norm": 1.6875,
+      "learning_rate": 1.3510173207710113e-05,
+      "loss": 0.8694,
+      "step": 6812
+    },
+    {
+      "epoch": 1.1562037850697644,
+      "grad_norm": 1.8046875,
+      "learning_rate": 1.3508482135381205e-05,
+      "loss": 0.9258,
+      "step": 6813
+    },
+    {
+      "epoch": 1.1563752491587542,
+      "grad_norm": 1.75,
+      "learning_rate": 1.350679094862748e-05,
+      "loss": 0.8493,
+      "step": 6814
+    },
+    {
+      "epoch": 1.156546713247744,
+      "grad_norm": 1.6640625,
+      "learning_rate": 1.3505099647504097e-05,
+      "loss": 0.8316,
+      "step": 6815
+    },
+    {
+      "epoch": 1.156718177336734,
+      "grad_norm": 1.609375,
+      "learning_rate": 1.3503408232066215e-05,
+      "loss": 0.8276,
+      "step": 6816
+    },
+    {
+      "epoch": 1.156889641425724,
+      "grad_norm": 1.671875,
+      "learning_rate": 1.3501716702369e-05,
+      "loss": 0.8804,
+      "step": 6817
+    },
+    {
+      "epoch": 1.1570611055147138,
+      "grad_norm": 1.6875,
+      "learning_rate": 1.3500025058467618e-05,
+      "loss": 0.8201,
+      "step": 6818
+    },
+    {
+      "epoch": 1.1572325696037036,
+      "grad_norm": 1.6953125,
+      "learning_rate": 1.3498333300417238e-05,
+      "loss": 0.8295,
+      "step": 6819
+    },
+    {
+      "epoch": 1.1574040336926934,
+      "grad_norm": 1.6484375,
+      "learning_rate": 1.3496641428273032e-05,
+      "loss": 0.907,
+      "step": 6820
+    },
+    {
+      "epoch": 1.1575754977816834,
+      "grad_norm": 1.640625,
+      "learning_rate": 1.3494949442090186e-05,
+      "loss": 0.8573,
+      "step": 6821
+    },
+    {
+      "epoch": 1.1577469618706733,
+      "grad_norm": 1.5546875,
+      "learning_rate": 1.3493257341923876e-05,
+      "loss": 0.8617,
+      "step": 6822
+    },
+    {
+      "epoch": 1.157918425959663,
+      "grad_norm": 1.78125,
+      "learning_rate": 1.3491565127829293e-05,
+      "loss": 0.9322,
+      "step": 6823
+    },
+    {
+      "epoch": 1.158089890048653,
+      "grad_norm": 1.6875,
+      "learning_rate": 1.3489872799861616e-05,
+      "loss": 0.8675,
+      "step": 6824
+    },
+    {
+      "epoch": 1.1582613541376428,
+      "grad_norm": 1.7421875,
+      "learning_rate": 1.348818035807605e-05,
+      "loss": 0.8458,
+      "step": 6825
+    },
+    {
+      "epoch": 1.1584328182266326,
+      "grad_norm": 1.6484375,
+      "learning_rate": 1.3486487802527788e-05,
+      "loss": 0.8289,
+      "step": 6826
+    },
+    {
+      "epoch": 1.1586042823156224,
+      "grad_norm": 1.7109375,
+      "learning_rate": 1.3484795133272028e-05,
+      "loss": 0.9288,
+      "step": 6827
+    },
+    {
+      "epoch": 1.1587757464046125,
+      "grad_norm": 1.7890625,
+      "learning_rate": 1.348310235036398e-05,
+      "loss": 0.917,
+      "step": 6828
+    },
+    {
+      "epoch": 1.1589472104936023,
+      "grad_norm": 1.6796875,
+      "learning_rate": 1.3481409453858846e-05,
+      "loss": 0.8495,
+      "step": 6829
+    },
+    {
+      "epoch": 1.159118674582592,
+      "grad_norm": 1.640625,
+      "learning_rate": 1.3479716443811838e-05,
+      "loss": 0.8831,
+      "step": 6830
+    },
+    {
+      "epoch": 1.159290138671582,
+      "grad_norm": 1.609375,
+      "learning_rate": 1.3478023320278175e-05,
+      "loss": 0.8112,
+      "step": 6831
+    },
+    {
+      "epoch": 1.1594616027605718,
+      "grad_norm": 1.7421875,
+      "learning_rate": 1.3476330083313074e-05,
+      "loss": 0.9331,
+      "step": 6832
+    },
+    {
+      "epoch": 1.1596330668495618,
+      "grad_norm": 1.6796875,
+      "learning_rate": 1.3474636732971758e-05,
+      "loss": 0.8153,
+      "step": 6833
+    },
+    {
+      "epoch": 1.1598045309385516,
+      "grad_norm": 1.75,
+      "learning_rate": 1.3472943269309455e-05,
+      "loss": 0.8215,
+      "step": 6834
+    },
+    {
+      "epoch": 1.1599759950275415,
+      "grad_norm": 1.625,
+      "learning_rate": 1.3471249692381394e-05,
+      "loss": 0.8797,
+      "step": 6835
+    },
+    {
+      "epoch": 1.1601474591165313,
+      "grad_norm": 1.8125,
+      "learning_rate": 1.3469556002242808e-05,
+      "loss": 0.846,
+      "step": 6836
+    },
+    {
+      "epoch": 1.160318923205521,
+      "grad_norm": 1.71875,
+      "learning_rate": 1.3467862198948935e-05,
+      "loss": 0.9078,
+      "step": 6837
+    },
+    {
+      "epoch": 1.160490387294511,
+      "grad_norm": 1.6015625,
+      "learning_rate": 1.3466168282555018e-05,
+      "loss": 0.772,
+      "step": 6838
+    },
+    {
+      "epoch": 1.1606618513835008,
+      "grad_norm": 1.703125,
+      "learning_rate": 1.3464474253116303e-05,
+      "loss": 0.8907,
+      "step": 6839
+    },
+    {
+      "epoch": 1.1608333154724908,
+      "grad_norm": 1.671875,
+      "learning_rate": 1.3462780110688036e-05,
+      "loss": 0.7626,
+      "step": 6840
+    },
+    {
+      "epoch": 1.1610047795614806,
+      "grad_norm": 1.734375,
+      "learning_rate": 1.3461085855325467e-05,
+      "loss": 0.8413,
+      "step": 6841
+    },
+    {
+      "epoch": 1.1611762436504705,
+      "grad_norm": 1.7890625,
+      "learning_rate": 1.3459391487083858e-05,
+      "loss": 0.8411,
+      "step": 6842
+    },
+    {
+      "epoch": 1.1613477077394603,
+      "grad_norm": 1.6171875,
+      "learning_rate": 1.3457697006018462e-05,
+      "loss": 0.8852,
+      "step": 6843
+    },
+    {
+      "epoch": 1.16151917182845,
+      "grad_norm": 1.625,
+      "learning_rate": 1.3456002412184548e-05,
+      "loss": 0.8167,
+      "step": 6844
+    },
+    {
+      "epoch": 1.1616906359174402,
+      "grad_norm": 1.6484375,
+      "learning_rate": 1.3454307705637382e-05,
+      "loss": 0.8421,
+      "step": 6845
+    },
+    {
+      "epoch": 1.16186210000643,
+      "grad_norm": 1.65625,
+      "learning_rate": 1.3452612886432234e-05,
+      "loss": 0.8503,
+      "step": 6846
+    },
+    {
+      "epoch": 1.1620335640954198,
+      "grad_norm": 1.625,
+      "learning_rate": 1.3450917954624378e-05,
+      "loss": 0.854,
+      "step": 6847
+    },
+    {
+      "epoch": 1.1622050281844096,
+      "grad_norm": 1.7578125,
+      "learning_rate": 1.3449222910269093e-05,
+      "loss": 0.8682,
+      "step": 6848
+    },
+    {
+      "epoch": 1.1623764922733995,
+      "grad_norm": 1.6328125,
+      "learning_rate": 1.3447527753421661e-05,
+      "loss": 0.976,
+      "step": 6849
+    },
+    {
+      "epoch": 1.1625479563623893,
+      "grad_norm": 1.640625,
+      "learning_rate": 1.3445832484137365e-05,
+      "loss": 0.8742,
+      "step": 6850
+    },
+    {
+      "epoch": 1.162719420451379,
+      "grad_norm": 1.6875,
+      "learning_rate": 1.3444137102471495e-05,
+      "loss": 0.8685,
+      "step": 6851
+    },
+    {
+      "epoch": 1.1628908845403692,
+      "grad_norm": 1.625,
+      "learning_rate": 1.3442441608479349e-05,
+      "loss": 0.7899,
+      "step": 6852
+    },
+    {
+      "epoch": 1.163062348629359,
+      "grad_norm": 1.734375,
+      "learning_rate": 1.3440746002216213e-05,
+      "loss": 0.8784,
+      "step": 6853
+    },
+    {
+      "epoch": 1.1632338127183488,
+      "grad_norm": 1.6875,
+      "learning_rate": 1.3439050283737399e-05,
+      "loss": 0.8723,
+      "step": 6854
+    },
+    {
+      "epoch": 1.1634052768073386,
+      "grad_norm": 1.6484375,
+      "learning_rate": 1.3437354453098202e-05,
+      "loss": 0.8222,
+      "step": 6855
+    },
+    {
+      "epoch": 1.1635767408963285,
+      "grad_norm": 1.6875,
+      "learning_rate": 1.3435658510353933e-05,
+      "loss": 0.8552,
+      "step": 6856
+    },
+    {
+      "epoch": 1.1637482049853185,
+      "grad_norm": 1.640625,
+      "learning_rate": 1.3433962455559901e-05,
+      "loss": 0.8358,
+      "step": 6857
+    },
+    {
+      "epoch": 1.1639196690743083,
+      "grad_norm": 1.7265625,
+      "learning_rate": 1.3432266288771427e-05,
+      "loss": 0.8489,
+      "step": 6858
+    },
+    {
+      "epoch": 1.1640911331632982,
+      "grad_norm": 1.578125,
+      "learning_rate": 1.3430570010043821e-05,
+      "loss": 0.7495,
+      "step": 6859
+    },
+    {
+      "epoch": 1.164262597252288,
+      "grad_norm": 1.6875,
+      "learning_rate": 1.342887361943241e-05,
+      "loss": 0.8586,
+      "step": 6860
+    },
+    {
+      "epoch": 1.1644340613412778,
+      "grad_norm": 1.6953125,
+      "learning_rate": 1.3427177116992515e-05,
+      "loss": 0.8377,
+      "step": 6861
+    },
+    {
+      "epoch": 1.1646055254302676,
+      "grad_norm": 1.6875,
+      "learning_rate": 1.3425480502779471e-05,
+      "loss": 0.7789,
+      "step": 6862
+    },
+    {
+      "epoch": 1.1647769895192575,
+      "grad_norm": 1.7109375,
+      "learning_rate": 1.3423783776848609e-05,
+      "loss": 0.8138,
+      "step": 6863
+    },
+    {
+      "epoch": 1.1649484536082475,
+      "grad_norm": 1.75,
+      "learning_rate": 1.3422086939255265e-05,
+      "loss": 0.9288,
+      "step": 6864
+    },
+    {
+      "epoch": 1.1651199176972373,
+      "grad_norm": 1.640625,
+      "learning_rate": 1.342038999005478e-05,
+      "loss": 0.8544,
+      "step": 6865
+    },
+    {
+      "epoch": 1.1652913817862272,
+      "grad_norm": 1.828125,
+      "learning_rate": 1.3418692929302497e-05,
+      "loss": 0.9111,
+      "step": 6866
+    },
+    {
+      "epoch": 1.165462845875217,
+      "grad_norm": 1.7109375,
+      "learning_rate": 1.3416995757053764e-05,
+      "loss": 0.9236,
+      "step": 6867
+    },
+    {
+      "epoch": 1.1656343099642068,
+      "grad_norm": 1.5859375,
+      "learning_rate": 1.3415298473363932e-05,
+      "loss": 0.8341,
+      "step": 6868
+    },
+    {
+      "epoch": 1.1658057740531969,
+      "grad_norm": 1.7109375,
+      "learning_rate": 1.3413601078288356e-05,
+      "loss": 0.8754,
+      "step": 6869
+    },
+    {
+      "epoch": 1.1659772381421867,
+      "grad_norm": 1.8125,
+      "learning_rate": 1.3411903571882395e-05,
+      "loss": 0.9529,
+      "step": 6870
+    },
+    {
+      "epoch": 1.1661487022311765,
+      "grad_norm": 1.6328125,
+      "learning_rate": 1.3410205954201407e-05,
+      "loss": 0.8552,
+      "step": 6871
+    },
+    {
+      "epoch": 1.1663201663201663,
+      "grad_norm": 1.6796875,
+      "learning_rate": 1.3408508225300765e-05,
+      "loss": 0.8721,
+      "step": 6872
+    },
+    {
+      "epoch": 1.1664916304091562,
+      "grad_norm": 2.015625,
+      "learning_rate": 1.3406810385235833e-05,
+      "loss": 0.8574,
+      "step": 6873
+    },
+    {
+      "epoch": 1.166663094498146,
+      "grad_norm": 1.78125,
+      "learning_rate": 1.3405112434061986e-05,
+      "loss": 0.893,
+      "step": 6874
+    },
+    {
+      "epoch": 1.1668345585871358,
+      "grad_norm": 1.7421875,
+      "learning_rate": 1.3403414371834602e-05,
+      "loss": 0.8774,
+      "step": 6875
+    },
+    {
+      "epoch": 1.1670060226761259,
+      "grad_norm": 1.6796875,
+      "learning_rate": 1.3401716198609056e-05,
+      "loss": 0.9101,
+      "step": 6876
+    },
+    {
+      "epoch": 1.1671774867651157,
+      "grad_norm": 1.703125,
+      "learning_rate": 1.3400017914440738e-05,
+      "loss": 0.883,
+      "step": 6877
+    },
+    {
+      "epoch": 1.1673489508541055,
+      "grad_norm": 1.65625,
+      "learning_rate": 1.339831951938503e-05,
+      "loss": 0.8582,
+      "step": 6878
+    },
+    {
+      "epoch": 1.1675204149430953,
+      "grad_norm": 1.6484375,
+      "learning_rate": 1.3396621013497327e-05,
+      "loss": 0.8527,
+      "step": 6879
+    },
+    {
+      "epoch": 1.1676918790320852,
+      "grad_norm": 1.7578125,
+      "learning_rate": 1.3394922396833021e-05,
+      "loss": 0.8982,
+      "step": 6880
+    },
+    {
+      "epoch": 1.1678633431210752,
+      "grad_norm": 1.703125,
+      "learning_rate": 1.3393223669447513e-05,
+      "loss": 0.9034,
+      "step": 6881
+    },
+    {
+      "epoch": 1.168034807210065,
+      "grad_norm": 1.671875,
+      "learning_rate": 1.3391524831396202e-05,
+      "loss": 0.945,
+      "step": 6882
+    },
+    {
+      "epoch": 1.1682062712990549,
+      "grad_norm": 1.6953125,
+      "learning_rate": 1.3389825882734495e-05,
+      "loss": 0.8772,
+      "step": 6883
+    },
+    {
+      "epoch": 1.1683777353880447,
+      "grad_norm": 1.6171875,
+      "learning_rate": 1.3388126823517802e-05,
+      "loss": 0.8102,
+      "step": 6884
+    },
+    {
+      "epoch": 1.1685491994770345,
+      "grad_norm": 1.6171875,
+      "learning_rate": 1.3386427653801535e-05,
+      "loss": 0.8419,
+      "step": 6885
+    },
+    {
+      "epoch": 1.1687206635660243,
+      "grad_norm": 1.703125,
+      "learning_rate": 1.3384728373641111e-05,
+      "loss": 0.9198,
+      "step": 6886
+    },
+    {
+      "epoch": 1.1688921276550142,
+      "grad_norm": 1.671875,
+      "learning_rate": 1.3383028983091948e-05,
+      "loss": 0.8829,
+      "step": 6887
+    },
+    {
+      "epoch": 1.1690635917440042,
+      "grad_norm": 1.625,
+      "learning_rate": 1.3381329482209471e-05,
+      "loss": 0.8461,
+      "step": 6888
+    },
+    {
+      "epoch": 1.169235055832994,
+      "grad_norm": 1.625,
+      "learning_rate": 1.3379629871049105e-05,
+      "loss": 0.8655,
+      "step": 6889
+    },
+    {
+      "epoch": 1.1694065199219839,
+      "grad_norm": 1.8203125,
+      "learning_rate": 1.337793014966628e-05,
+      "loss": 0.8972,
+      "step": 6890
+    },
+    {
+      "epoch": 1.1695779840109737,
+      "grad_norm": 1.6953125,
+      "learning_rate": 1.3376230318116437e-05,
+      "loss": 0.8264,
+      "step": 6891
+    },
+    {
+      "epoch": 1.1697494480999635,
+      "grad_norm": 1.65625,
+      "learning_rate": 1.337453037645501e-05,
+      "loss": 0.9174,
+      "step": 6892
+    },
+    {
+      "epoch": 1.1699209121889533,
+      "grad_norm": 1.609375,
+      "learning_rate": 1.3372830324737438e-05,
+      "loss": 0.7695,
+      "step": 6893
+    },
+    {
+      "epoch": 1.1700923762779434,
+      "grad_norm": 1.7421875,
+      "learning_rate": 1.3371130163019168e-05,
+      "loss": 0.8772,
+      "step": 6894
+    },
+    {
+      "epoch": 1.1702638403669332,
+      "grad_norm": 1.7109375,
+      "learning_rate": 1.3369429891355653e-05,
+      "loss": 0.9167,
+      "step": 6895
+    },
+    {
+      "epoch": 1.170435304455923,
+      "grad_norm": 1.640625,
+      "learning_rate": 1.3367729509802336e-05,
+      "loss": 0.8204,
+      "step": 6896
+    },
+    {
+      "epoch": 1.1706067685449129,
+      "grad_norm": 1.671875,
+      "learning_rate": 1.3366029018414679e-05,
+      "loss": 0.7964,
+      "step": 6897
+    },
+    {
+      "epoch": 1.1707782326339027,
+      "grad_norm": 1.640625,
+      "learning_rate": 1.3364328417248142e-05,
+      "loss": 0.8479,
+      "step": 6898
+    },
+    {
+      "epoch": 1.1709496967228925,
+      "grad_norm": 1.59375,
+      "learning_rate": 1.3362627706358187e-05,
+      "loss": 0.8342,
+      "step": 6899
+    },
+    {
+      "epoch": 1.1711211608118826,
+      "grad_norm": 1.7265625,
+      "learning_rate": 1.3360926885800279e-05,
+      "loss": 0.866,
+      "step": 6900
+    },
+    {
+      "epoch": 1.1712926249008724,
+      "grad_norm": 1.6953125,
+      "learning_rate": 1.3359225955629889e-05,
+      "loss": 0.9075,
+      "step": 6901
+    },
+    {
+      "epoch": 1.1714640889898622,
+      "grad_norm": 1.71875,
+      "learning_rate": 1.3357524915902488e-05,
+      "loss": 0.9485,
+      "step": 6902
+    },
+    {
+      "epoch": 1.171635553078852,
+      "grad_norm": 1.6640625,
+      "learning_rate": 1.335582376667356e-05,
+      "loss": 0.8911,
+      "step": 6903
+    },
+    {
+      "epoch": 1.1718070171678419,
+      "grad_norm": 1.625,
+      "learning_rate": 1.3354122507998584e-05,
+      "loss": 0.8715,
+      "step": 6904
+    },
+    {
+      "epoch": 1.1719784812568317,
+      "grad_norm": 1.6484375,
+      "learning_rate": 1.3352421139933038e-05,
+      "loss": 0.8291,
+      "step": 6905
+    },
+    {
+      "epoch": 1.1721499453458217,
+      "grad_norm": 1.625,
+      "learning_rate": 1.3350719662532416e-05,
+      "loss": 0.8443,
+      "step": 6906
+    },
+    {
+      "epoch": 1.1723214094348116,
+      "grad_norm": 1.6171875,
+      "learning_rate": 1.334901807585221e-05,
+      "loss": 0.8684,
+      "step": 6907
+    },
+    {
+      "epoch": 1.1724928735238014,
+      "grad_norm": 1.6796875,
+      "learning_rate": 1.3347316379947912e-05,
+      "loss": 0.8409,
+      "step": 6908
+    },
+    {
+      "epoch": 1.1726643376127912,
+      "grad_norm": 1.703125,
+      "learning_rate": 1.3345614574875022e-05,
+      "loss": 0.9154,
+      "step": 6909
+    },
+    {
+      "epoch": 1.172835801701781,
+      "grad_norm": 1.671875,
+      "learning_rate": 1.334391266068904e-05,
+      "loss": 0.8695,
+      "step": 6910
+    },
+    {
+      "epoch": 1.1730072657907709,
+      "grad_norm": 1.7109375,
+      "learning_rate": 1.3342210637445478e-05,
+      "loss": 0.8835,
+      "step": 6911
+    },
+    {
+      "epoch": 1.1731787298797607,
+      "grad_norm": 1.6875,
+      "learning_rate": 1.3340508505199839e-05,
+      "loss": 0.8492,
+      "step": 6912
+    },
+    {
+      "epoch": 1.1733501939687507,
+      "grad_norm": 1.6484375,
+      "learning_rate": 1.333880626400764e-05,
+      "loss": 0.793,
+      "step": 6913
+    },
+    {
+      "epoch": 1.1735216580577406,
+      "grad_norm": 1.59375,
+      "learning_rate": 1.3337103913924394e-05,
+      "loss": 0.8493,
+      "step": 6914
+    },
+    {
+      "epoch": 1.1736931221467304,
+      "grad_norm": 1.734375,
+      "learning_rate": 1.3335401455005625e-05,
+      "loss": 0.8716,
+      "step": 6915
+    },
+    {
+      "epoch": 1.1738645862357202,
+      "grad_norm": 1.6484375,
+      "learning_rate": 1.3333698887306854e-05,
+      "loss": 0.8844,
+      "step": 6916
+    },
+    {
+      "epoch": 1.17403605032471,
+      "grad_norm": 1.7734375,
+      "learning_rate": 1.3331996210883609e-05,
+      "loss": 0.9383,
+      "step": 6917
+    },
+    {
+      "epoch": 1.1742075144137,
+      "grad_norm": 1.6328125,
+      "learning_rate": 1.3330293425791419e-05,
+      "loss": 0.7867,
+      "step": 6918
+    },
+    {
+      "epoch": 1.17437897850269,
+      "grad_norm": 1.6171875,
+      "learning_rate": 1.3328590532085822e-05,
+      "loss": 0.8331,
+      "step": 6919
+    },
+    {
+      "epoch": 1.1745504425916797,
+      "grad_norm": 1.640625,
+      "learning_rate": 1.332688752982235e-05,
+      "loss": 0.8586,
+      "step": 6920
+    },
+    {
+      "epoch": 1.1747219066806696,
+      "grad_norm": 1.6171875,
+      "learning_rate": 1.3325184419056552e-05,
+      "loss": 0.8463,
+      "step": 6921
+    },
+    {
+      "epoch": 1.1748933707696594,
+      "grad_norm": 1.6953125,
+      "learning_rate": 1.3323481199843966e-05,
+      "loss": 0.8227,
+      "step": 6922
+    },
+    {
+      "epoch": 1.1750648348586492,
+      "grad_norm": 1.578125,
+      "learning_rate": 1.3321777872240142e-05,
+      "loss": 0.8337,
+      "step": 6923
+    },
+    {
+      "epoch": 1.175236298947639,
+      "grad_norm": 1.6328125,
+      "learning_rate": 1.3320074436300635e-05,
+      "loss": 0.9401,
+      "step": 6924
+    },
+    {
+      "epoch": 1.175407763036629,
+      "grad_norm": 1.609375,
+      "learning_rate": 1.3318370892080998e-05,
+      "loss": 0.8901,
+      "step": 6925
+    },
+    {
+      "epoch": 1.175579227125619,
+      "grad_norm": 1.5625,
+      "learning_rate": 1.3316667239636792e-05,
+      "loss": 0.7857,
+      "step": 6926
+    },
+    {
+      "epoch": 1.1757506912146087,
+      "grad_norm": 1.625,
+      "learning_rate": 1.3314963479023575e-05,
+      "loss": 0.8501,
+      "step": 6927
+    },
+    {
+      "epoch": 1.1759221553035986,
+      "grad_norm": 1.6640625,
+      "learning_rate": 1.3313259610296916e-05,
+      "loss": 0.9171,
+      "step": 6928
+    },
+    {
+      "epoch": 1.1760936193925884,
+      "grad_norm": 1.6796875,
+      "learning_rate": 1.3311555633512386e-05,
+      "loss": 0.8805,
+      "step": 6929
+    },
+    {
+      "epoch": 1.1762650834815784,
+      "grad_norm": 1.703125,
+      "learning_rate": 1.3309851548725553e-05,
+      "loss": 0.8211,
+      "step": 6930
+    },
+    {
+      "epoch": 1.1764365475705683,
+      "grad_norm": 1.6171875,
+      "learning_rate": 1.3308147355992002e-05,
+      "loss": 0.8142,
+      "step": 6931
+    },
+    {
+      "epoch": 1.176608011659558,
+      "grad_norm": 1.7421875,
+      "learning_rate": 1.3306443055367306e-05,
+      "loss": 0.9,
+      "step": 6932
+    },
+    {
+      "epoch": 1.176779475748548,
+      "grad_norm": 1.6796875,
+      "learning_rate": 1.3304738646907057e-05,
+      "loss": 0.8249,
+      "step": 6933
+    },
+    {
+      "epoch": 1.1769509398375377,
+      "grad_norm": 1.7421875,
+      "learning_rate": 1.330303413066683e-05,
+      "loss": 0.8355,
+      "step": 6934
+    },
+    {
+      "epoch": 1.1771224039265276,
+      "grad_norm": 1.78125,
+      "learning_rate": 1.3301329506702221e-05,
+      "loss": 0.8734,
+      "step": 6935
+    },
+    {
+      "epoch": 1.1772938680155174,
+      "grad_norm": 1.6640625,
+      "learning_rate": 1.3299624775068826e-05,
+      "loss": 0.9342,
+      "step": 6936
+    },
+    {
+      "epoch": 1.1774653321045074,
+      "grad_norm": 1.6484375,
+      "learning_rate": 1.3297919935822243e-05,
+      "loss": 0.8546,
+      "step": 6937
+    },
+    {
+      "epoch": 1.1776367961934973,
+      "grad_norm": 1.671875,
+      "learning_rate": 1.3296214989018075e-05,
+      "loss": 0.8482,
+      "step": 6938
+    },
+    {
+      "epoch": 1.177808260282487,
+      "grad_norm": 1.703125,
+      "learning_rate": 1.3294509934711919e-05,
+      "loss": 0.8412,
+      "step": 6939
+    },
+    {
+      "epoch": 1.177979724371477,
+      "grad_norm": 1.671875,
+      "learning_rate": 1.3292804772959391e-05,
+      "loss": 0.8311,
+      "step": 6940
+    },
+    {
+      "epoch": 1.1781511884604667,
+      "grad_norm": 1.6875,
+      "learning_rate": 1.3291099503816098e-05,
+      "loss": 0.8382,
+      "step": 6941
+    },
+    {
+      "epoch": 1.1783226525494568,
+      "grad_norm": 1.6171875,
+      "learning_rate": 1.3289394127337658e-05,
+      "loss": 0.8111,
+      "step": 6942
+    },
+    {
+      "epoch": 1.1784941166384466,
+      "grad_norm": 1.640625,
+      "learning_rate": 1.3287688643579688e-05,
+      "loss": 0.8764,
+      "step": 6943
+    },
+    {
+      "epoch": 1.1786655807274364,
+      "grad_norm": 1.6328125,
+      "learning_rate": 1.3285983052597813e-05,
+      "loss": 0.8151,
+      "step": 6944
+    },
+    {
+      "epoch": 1.1788370448164263,
+      "grad_norm": 1.6171875,
+      "learning_rate": 1.3284277354447655e-05,
+      "loss": 0.8074,
+      "step": 6945
+    },
+    {
+      "epoch": 1.179008508905416,
+      "grad_norm": 1.671875,
+      "learning_rate": 1.3282571549184844e-05,
+      "loss": 0.8439,
+      "step": 6946
+    },
+    {
+      "epoch": 1.179179972994406,
+      "grad_norm": 1.7265625,
+      "learning_rate": 1.3280865636865014e-05,
+      "loss": 0.9015,
+      "step": 6947
+    },
+    {
+      "epoch": 1.1793514370833957,
+      "grad_norm": 1.6015625,
+      "learning_rate": 1.3279159617543801e-05,
+      "loss": 0.8248,
+      "step": 6948
+    },
+    {
+      "epoch": 1.1795229011723858,
+      "grad_norm": 1.640625,
+      "learning_rate": 1.3277453491276849e-05,
+      "loss": 0.9074,
+      "step": 6949
+    },
+    {
+      "epoch": 1.1796943652613756,
+      "grad_norm": 1.6328125,
+      "learning_rate": 1.3275747258119793e-05,
+      "loss": 0.81,
+      "step": 6950
+    },
+    {
+      "epoch": 1.1798658293503654,
+      "grad_norm": 1.59375,
+      "learning_rate": 1.3274040918128286e-05,
+      "loss": 0.8182,
+      "step": 6951
+    },
+    {
+      "epoch": 1.1800372934393553,
+      "grad_norm": 1.625,
+      "learning_rate": 1.3272334471357975e-05,
+      "loss": 0.8618,
+      "step": 6952
+    },
+    {
+      "epoch": 1.180208757528345,
+      "grad_norm": 1.7265625,
+      "learning_rate": 1.327062791786451e-05,
+      "loss": 0.9241,
+      "step": 6953
+    },
+    {
+      "epoch": 1.1803802216173351,
+      "grad_norm": 1.671875,
+      "learning_rate": 1.3268921257703557e-05,
+      "loss": 0.7983,
+      "step": 6954
+    },
+    {
+      "epoch": 1.180551685706325,
+      "grad_norm": 1.671875,
+      "learning_rate": 1.3267214490930771e-05,
+      "loss": 0.8148,
+      "step": 6955
+    },
+    {
+      "epoch": 1.1807231497953148,
+      "grad_norm": 1.65625,
+      "learning_rate": 1.3265507617601818e-05,
+      "loss": 0.7872,
+      "step": 6956
+    },
+    {
+      "epoch": 1.1808946138843046,
+      "grad_norm": 1.71875,
+      "learning_rate": 1.3263800637772363e-05,
+      "loss": 0.8317,
+      "step": 6957
+    },
+    {
+      "epoch": 1.1810660779732944,
+      "grad_norm": 1.640625,
+      "learning_rate": 1.3262093551498084e-05,
+      "loss": 0.7933,
+      "step": 6958
+    },
+    {
+      "epoch": 1.1812375420622843,
+      "grad_norm": 1.71875,
+      "learning_rate": 1.3260386358834645e-05,
+      "loss": 0.8589,
+      "step": 6959
+    },
+    {
+      "epoch": 1.181409006151274,
+      "grad_norm": 1.703125,
+      "learning_rate": 1.3258679059837731e-05,
+      "loss": 0.8669,
+      "step": 6960
+    },
+    {
+      "epoch": 1.1815804702402641,
+      "grad_norm": 1.796875,
+      "learning_rate": 1.3256971654563022e-05,
+      "loss": 0.8813,
+      "step": 6961
+    },
+    {
+      "epoch": 1.181751934329254,
+      "grad_norm": 1.7109375,
+      "learning_rate": 1.3255264143066202e-05,
+      "loss": 0.8569,
+      "step": 6962
+    },
+    {
+      "epoch": 1.1819233984182438,
+      "grad_norm": 1.65625,
+      "learning_rate": 1.3253556525402963e-05,
+      "loss": 0.8348,
+      "step": 6963
+    },
+    {
+      "epoch": 1.1820948625072336,
+      "grad_norm": 1.65625,
+      "learning_rate": 1.325184880162899e-05,
+      "loss": 0.8558,
+      "step": 6964
+    },
+    {
+      "epoch": 1.1822663265962234,
+      "grad_norm": 1.6171875,
+      "learning_rate": 1.3250140971799984e-05,
+      "loss": 0.8327,
+      "step": 6965
+    },
+    {
+      "epoch": 1.1824377906852135,
+      "grad_norm": 1.703125,
+      "learning_rate": 1.324843303597164e-05,
+      "loss": 0.9319,
+      "step": 6966
+    },
+    {
+      "epoch": 1.1826092547742033,
+      "grad_norm": 1.6328125,
+      "learning_rate": 1.3246724994199664e-05,
+      "loss": 0.8491,
+      "step": 6967
+    },
+    {
+      "epoch": 1.1827807188631931,
+      "grad_norm": 1.734375,
+      "learning_rate": 1.324501684653976e-05,
+      "loss": 0.8954,
+      "step": 6968
+    },
+    {
+      "epoch": 1.182952182952183,
+      "grad_norm": 1.6875,
+      "learning_rate": 1.3243308593047637e-05,
+      "loss": 0.8383,
+      "step": 6969
+    },
+    {
+      "epoch": 1.1831236470411728,
+      "grad_norm": 1.734375,
+      "learning_rate": 1.3241600233779008e-05,
+      "loss": 0.8032,
+      "step": 6970
+    },
+    {
+      "epoch": 1.1832951111301626,
+      "grad_norm": 1.6328125,
+      "learning_rate": 1.3239891768789587e-05,
+      "loss": 0.8594,
+      "step": 6971
+    },
+    {
+      "epoch": 1.1834665752191524,
+      "grad_norm": 1.7265625,
+      "learning_rate": 1.3238183198135098e-05,
+      "loss": 0.8261,
+      "step": 6972
+    },
+    {
+      "epoch": 1.1836380393081425,
+      "grad_norm": 1.6875,
+      "learning_rate": 1.323647452187126e-05,
+      "loss": 0.9486,
+      "step": 6973
+    },
+    {
+      "epoch": 1.1838095033971323,
+      "grad_norm": 1.6328125,
+      "learning_rate": 1.3234765740053799e-05,
+      "loss": 0.8751,
+      "step": 6974
+    },
+    {
+      "epoch": 1.1839809674861221,
+      "grad_norm": 1.796875,
+      "learning_rate": 1.3233056852738446e-05,
+      "loss": 0.8476,
+      "step": 6975
+    },
+    {
+      "epoch": 1.184152431575112,
+      "grad_norm": 1.6796875,
+      "learning_rate": 1.3231347859980937e-05,
+      "loss": 0.88,
+      "step": 6976
+    },
+    {
+      "epoch": 1.1843238956641018,
+      "grad_norm": 1.7265625,
+      "learning_rate": 1.3229638761837003e-05,
+      "loss": 0.8816,
+      "step": 6977
+    },
+    {
+      "epoch": 1.1844953597530918,
+      "grad_norm": 1.6484375,
+      "learning_rate": 1.322792955836239e-05,
+      "loss": 0.8436,
+      "step": 6978
+    },
+    {
+      "epoch": 1.1846668238420817,
+      "grad_norm": 1.7265625,
+      "learning_rate": 1.3226220249612837e-05,
+      "loss": 0.8455,
+      "step": 6979
+    },
+    {
+      "epoch": 1.1848382879310715,
+      "grad_norm": 1.7109375,
+      "learning_rate": 1.3224510835644095e-05,
+      "loss": 0.833,
+      "step": 6980
+    },
+    {
+      "epoch": 1.1850097520200613,
+      "grad_norm": 1.6484375,
+      "learning_rate": 1.322280131651191e-05,
+      "loss": 0.885,
+      "step": 6981
+    },
+    {
+      "epoch": 1.1851812161090511,
+      "grad_norm": 1.6640625,
+      "learning_rate": 1.3221091692272042e-05,
+      "loss": 0.8174,
+      "step": 6982
+    },
+    {
+      "epoch": 1.185352680198041,
+      "grad_norm": 1.7109375,
+      "learning_rate": 1.321938196298024e-05,
+      "loss": 0.9002,
+      "step": 6983
+    },
+    {
+      "epoch": 1.1855241442870308,
+      "grad_norm": 1.71875,
+      "learning_rate": 1.321767212869227e-05,
+      "loss": 0.8802,
+      "step": 6984
+    },
+    {
+      "epoch": 1.1856956083760208,
+      "grad_norm": 1.5859375,
+      "learning_rate": 1.3215962189463896e-05,
+      "loss": 0.8047,
+      "step": 6985
+    },
+    {
+      "epoch": 1.1858670724650107,
+      "grad_norm": 1.5859375,
+      "learning_rate": 1.3214252145350883e-05,
+      "loss": 0.8862,
+      "step": 6986
+    },
+    {
+      "epoch": 1.1860385365540005,
+      "grad_norm": 1.625,
+      "learning_rate": 1.3212541996409005e-05,
+      "loss": 0.8768,
+      "step": 6987
+    },
+    {
+      "epoch": 1.1862100006429903,
+      "grad_norm": 1.6015625,
+      "learning_rate": 1.3210831742694037e-05,
+      "loss": 0.8169,
+      "step": 6988
+    },
+    {
+      "epoch": 1.1863814647319801,
+      "grad_norm": 1.6875,
+      "learning_rate": 1.3209121384261754e-05,
+      "loss": 0.8902,
+      "step": 6989
+    },
+    {
+      "epoch": 1.18655292882097,
+      "grad_norm": 1.796875,
+      "learning_rate": 1.3207410921167938e-05,
+      "loss": 0.8861,
+      "step": 6990
+    },
+    {
+      "epoch": 1.18672439290996,
+      "grad_norm": 1.7109375,
+      "learning_rate": 1.3205700353468373e-05,
+      "loss": 0.8325,
+      "step": 6991
+    },
+    {
+      "epoch": 1.1868958569989498,
+      "grad_norm": 1.6015625,
+      "learning_rate": 1.3203989681218847e-05,
+      "loss": 0.8863,
+      "step": 6992
+    },
+    {
+      "epoch": 1.1870673210879397,
+      "grad_norm": 1.703125,
+      "learning_rate": 1.3202278904475154e-05,
+      "loss": 0.8174,
+      "step": 6993
+    },
+    {
+      "epoch": 1.1872387851769295,
+      "grad_norm": 1.703125,
+      "learning_rate": 1.3200568023293085e-05,
+      "loss": 0.8864,
+      "step": 6994
+    },
+    {
+      "epoch": 1.1874102492659193,
+      "grad_norm": 1.65625,
+      "learning_rate": 1.3198857037728446e-05,
+      "loss": 0.8373,
+      "step": 6995
+    },
+    {
+      "epoch": 1.1875817133549091,
+      "grad_norm": 1.703125,
+      "learning_rate": 1.3197145947837031e-05,
+      "loss": 0.9123,
+      "step": 6996
+    },
+    {
+      "epoch": 1.1877531774438992,
+      "grad_norm": 1.6171875,
+      "learning_rate": 1.3195434753674645e-05,
+      "loss": 0.9354,
+      "step": 6997
+    },
+    {
+      "epoch": 1.187924641532889,
+      "grad_norm": 1.6171875,
+      "learning_rate": 1.31937234552971e-05,
+      "loss": 0.7199,
+      "step": 6998
+    },
+    {
+      "epoch": 1.1880961056218788,
+      "grad_norm": 1.578125,
+      "learning_rate": 1.3192012052760208e-05,
+      "loss": 0.7695,
+      "step": 6999
+    },
+    {
+      "epoch": 1.1882675697108687,
+      "grad_norm": 1.6640625,
+      "learning_rate": 1.3190300546119781e-05,
+      "loss": 0.8186,
+      "step": 7000
+    },
+    {
+      "epoch": 1.1882675697108687,
+      "eval_loss": 0.8492981195449829,
+      "eval_runtime": 836.9828,
+      "eval_samples_per_second": 2.986,
+      "eval_steps_per_second": 2.986,
+      "step": 7000
+    },
+    {
+      "epoch": 1.1884390337998585,
+      "grad_norm": 1.7421875,
+      "learning_rate": 1.3188588935431642e-05,
+      "loss": 0.9399,
+      "step": 7001
+    },
+    {
+      "epoch": 1.1886104978888483,
+      "grad_norm": 1.703125,
+      "learning_rate": 1.3186877220751605e-05,
+      "loss": 0.8776,
+      "step": 7002
+    },
+    {
+      "epoch": 1.1887819619778384,
+      "grad_norm": 11.1875,
+      "learning_rate": 1.3185165402135507e-05,
+      "loss": 0.8575,
+      "step": 7003
+    },
+    {
+      "epoch": 1.1889534260668282,
+      "grad_norm": 1.6875,
+      "learning_rate": 1.3183453479639167e-05,
+      "loss": 0.8264,
+      "step": 7004
+    },
+    {
+      "epoch": 1.189124890155818,
+      "grad_norm": 1.6171875,
+      "learning_rate": 1.3181741453318427e-05,
+      "loss": 0.8503,
+      "step": 7005
+    },
+    {
+      "epoch": 1.1892963542448078,
+      "grad_norm": 1.6875,
+      "learning_rate": 1.3180029323229111e-05,
+      "loss": 0.8506,
+      "step": 7006
+    },
+    {
+      "epoch": 1.1894678183337977,
+      "grad_norm": 1.78125,
+      "learning_rate": 1.3178317089427066e-05,
+      "loss": 0.9127,
+      "step": 7007
+    },
+    {
+      "epoch": 1.1896392824227875,
+      "grad_norm": 1.640625,
+      "learning_rate": 1.3176604751968133e-05,
+      "loss": 0.8574,
+      "step": 7008
+    },
+    {
+      "epoch": 1.1898107465117773,
+      "grad_norm": 1.6015625,
+      "learning_rate": 1.3174892310908158e-05,
+      "loss": 0.8767,
+      "step": 7009
+    },
+    {
+      "epoch": 1.1899822106007674,
+      "grad_norm": 1.7578125,
+      "learning_rate": 1.3173179766302988e-05,
+      "loss": 0.8028,
+      "step": 7010
+    },
+    {
+      "epoch": 1.1901536746897572,
+      "grad_norm": 1.671875,
+      "learning_rate": 1.3171467118208476e-05,
+      "loss": 0.8199,
+      "step": 7011
+    },
+    {
+      "epoch": 1.190325138778747,
+      "grad_norm": 1.6484375,
+      "learning_rate": 1.3169754366680476e-05,
+      "loss": 0.842,
+      "step": 7012
+    },
+    {
+      "epoch": 1.1904966028677368,
+      "grad_norm": 1.7109375,
+      "learning_rate": 1.3168041511774856e-05,
+      "loss": 0.9031,
+      "step": 7013
+    },
+    {
+      "epoch": 1.1906680669567267,
+      "grad_norm": 1.6796875,
+      "learning_rate": 1.3166328553547469e-05,
+      "loss": 0.7974,
+      "step": 7014
+    },
+    {
+      "epoch": 1.1908395310457167,
+      "grad_norm": 1.6875,
+      "learning_rate": 1.3164615492054184e-05,
+      "loss": 0.8576,
+      "step": 7015
+    },
+    {
+      "epoch": 1.1910109951347065,
+      "grad_norm": 1.71875,
+      "learning_rate": 1.3162902327350873e-05,
+      "loss": 0.9097,
+      "step": 7016
+    },
+    {
+      "epoch": 1.1911824592236964,
+      "grad_norm": 1.6328125,
+      "learning_rate": 1.3161189059493407e-05,
+      "loss": 0.8865,
+      "step": 7017
+    },
+    {
+      "epoch": 1.1913539233126862,
+      "grad_norm": 1.65625,
+      "learning_rate": 1.315947568853766e-05,
+      "loss": 0.8532,
+      "step": 7018
+    },
+    {
+      "epoch": 1.191525387401676,
+      "grad_norm": 1.7578125,
+      "learning_rate": 1.3157762214539516e-05,
+      "loss": 0.9153,
+      "step": 7019
+    },
+    {
+      "epoch": 1.1916968514906658,
+      "grad_norm": 1.6796875,
+      "learning_rate": 1.315604863755485e-05,
+      "loss": 0.7673,
+      "step": 7020
+    },
+    {
+      "epoch": 1.1918683155796557,
+      "grad_norm": 1.640625,
+      "learning_rate": 1.3154334957639557e-05,
+      "loss": 0.8389,
+      "step": 7021
+    },
+    {
+      "epoch": 1.1920397796686457,
+      "grad_norm": 1.625,
+      "learning_rate": 1.3152621174849522e-05,
+      "loss": 0.7504,
+      "step": 7022
+    },
+    {
+      "epoch": 1.1922112437576355,
+      "grad_norm": 1.71875,
+      "learning_rate": 1.3150907289240639e-05,
+      "loss": 0.8595,
+      "step": 7023
+    },
+    {
+      "epoch": 1.1923827078466254,
+      "grad_norm": 1.5625,
+      "learning_rate": 1.3149193300868803e-05,
+      "loss": 0.7952,
+      "step": 7024
+    },
+    {
+      "epoch": 1.1925541719356152,
+      "grad_norm": 1.6953125,
+      "learning_rate": 1.314747920978992e-05,
+      "loss": 0.8829,
+      "step": 7025
+    },
+    {
+      "epoch": 1.192725636024605,
+      "grad_norm": 1.703125,
+      "learning_rate": 1.3145765016059882e-05,
+      "loss": 0.9387,
+      "step": 7026
+    },
+    {
+      "epoch": 1.192897100113595,
+      "grad_norm": 1.953125,
+      "learning_rate": 1.3144050719734602e-05,
+      "loss": 0.8143,
+      "step": 7027
+    },
+    {
+      "epoch": 1.193068564202585,
+      "grad_norm": 1.71875,
+      "learning_rate": 1.314233632086999e-05,
+      "loss": 0.8377,
+      "step": 7028
+    },
+    {
+      "epoch": 1.1932400282915747,
+      "grad_norm": 1.7578125,
+      "learning_rate": 1.3140621819521957e-05,
+      "loss": 0.8073,
+      "step": 7029
+    },
+    {
+      "epoch": 1.1934114923805645,
+      "grad_norm": 1.7109375,
+      "learning_rate": 1.313890721574642e-05,
+      "loss": 0.8738,
+      "step": 7030
+    },
+    {
+      "epoch": 1.1935829564695544,
+      "grad_norm": 1.8203125,
+      "learning_rate": 1.3137192509599297e-05,
+      "loss": 0.8952,
+      "step": 7031
+    },
+    {
+      "epoch": 1.1937544205585442,
+      "grad_norm": 1.703125,
+      "learning_rate": 1.3135477701136515e-05,
+      "loss": 0.8685,
+      "step": 7032
+    },
+    {
+      "epoch": 1.193925884647534,
+      "grad_norm": 1.59375,
+      "learning_rate": 1.3133762790413998e-05,
+      "loss": 0.8827,
+      "step": 7033
+    },
+    {
+      "epoch": 1.194097348736524,
+      "grad_norm": 1.7109375,
+      "learning_rate": 1.3132047777487676e-05,
+      "loss": 0.8834,
+      "step": 7034
+    },
+    {
+      "epoch": 1.194268812825514,
+      "grad_norm": 1.609375,
+      "learning_rate": 1.3130332662413478e-05,
+      "loss": 0.8182,
+      "step": 7035
+    },
+    {
+      "epoch": 1.1944402769145037,
+      "grad_norm": 1.7265625,
+      "learning_rate": 1.312861744524735e-05,
+      "loss": 0.9155,
+      "step": 7036
+    },
+    {
+      "epoch": 1.1946117410034935,
+      "grad_norm": 1.765625,
+      "learning_rate": 1.3126902126045223e-05,
+      "loss": 0.9201,
+      "step": 7037
+    },
+    {
+      "epoch": 1.1947832050924834,
+      "grad_norm": 1.734375,
+      "learning_rate": 1.3125186704863042e-05,
+      "loss": 0.8421,
+      "step": 7038
+    },
+    {
+      "epoch": 1.1949546691814734,
+      "grad_norm": 1.6484375,
+      "learning_rate": 1.3123471181756753e-05,
+      "loss": 0.8264,
+      "step": 7039
+    },
+    {
+      "epoch": 1.1951261332704632,
+      "grad_norm": 1.703125,
+      "learning_rate": 1.3121755556782307e-05,
+      "loss": 0.8418,
+      "step": 7040
+    },
+    {
+      "epoch": 1.195297597359453,
+      "grad_norm": 1.6328125,
+      "learning_rate": 1.3120039829995659e-05,
+      "loss": 0.7933,
+      "step": 7041
+    },
+    {
+      "epoch": 1.195469061448443,
+      "grad_norm": 1.71875,
+      "learning_rate": 1.3118324001452765e-05,
+      "loss": 0.8414,
+      "step": 7042
+    },
+    {
+      "epoch": 1.1956405255374327,
+      "grad_norm": 1.65625,
+      "learning_rate": 1.311660807120958e-05,
+      "loss": 0.8003,
+      "step": 7043
+    },
+    {
+      "epoch": 1.1958119896264225,
+      "grad_norm": 1.6640625,
+      "learning_rate": 1.3114892039322072e-05,
+      "loss": 0.8051,
+      "step": 7044
+    },
+    {
+      "epoch": 1.1959834537154124,
+      "grad_norm": 1.6328125,
+      "learning_rate": 1.31131759058462e-05,
+      "loss": 0.8998,
+      "step": 7045
+    },
+    {
+      "epoch": 1.1961549178044024,
+      "grad_norm": 4.8125,
+      "learning_rate": 1.3111459670837942e-05,
+      "loss": 0.9592,
+      "step": 7046
+    },
+    {
+      "epoch": 1.1963263818933922,
+      "grad_norm": 1.671875,
+      "learning_rate": 1.310974333435327e-05,
+      "loss": 0.8949,
+      "step": 7047
+    },
+    {
+      "epoch": 1.196497845982382,
+      "grad_norm": 1.75,
+      "learning_rate": 1.3108026896448153e-05,
+      "loss": 0.8741,
+      "step": 7048
+    },
+    {
+      "epoch": 1.196669310071372,
+      "grad_norm": 1.8359375,
+      "learning_rate": 1.3106310357178575e-05,
+      "loss": 0.8458,
+      "step": 7049
+    },
+    {
+      "epoch": 1.1968407741603617,
+      "grad_norm": 1.609375,
+      "learning_rate": 1.3104593716600521e-05,
+      "loss": 0.8015,
+      "step": 7050
+    },
+    {
+      "epoch": 1.1970122382493518,
+      "grad_norm": 1.640625,
+      "learning_rate": 1.3102876974769975e-05,
+      "loss": 0.8168,
+      "step": 7051
+    },
+    {
+      "epoch": 1.1971837023383416,
+      "grad_norm": 1.7265625,
+      "learning_rate": 1.3101160131742927e-05,
+      "loss": 0.8824,
+      "step": 7052
+    },
+    {
+      "epoch": 1.1973551664273314,
+      "grad_norm": 1.65625,
+      "learning_rate": 1.309944318757537e-05,
+      "loss": 0.7913,
+      "step": 7053
+    },
+    {
+      "epoch": 1.1975266305163212,
+      "grad_norm": 1.6171875,
+      "learning_rate": 1.30977261423233e-05,
+      "loss": 0.792,
+      "step": 7054
+    },
+    {
+      "epoch": 1.197698094605311,
+      "grad_norm": 1.7734375,
+      "learning_rate": 1.3096008996042712e-05,
+      "loss": 0.805,
+      "step": 7055
+    },
+    {
+      "epoch": 1.197869558694301,
+      "grad_norm": 1.5625,
+      "learning_rate": 1.3094291748789614e-05,
+      "loss": 0.7919,
+      "step": 7056
+    },
+    {
+      "epoch": 1.1980410227832907,
+      "grad_norm": 1.6640625,
+      "learning_rate": 1.309257440062001e-05,
+      "loss": 0.8871,
+      "step": 7057
+    },
+    {
+      "epoch": 1.1982124868722808,
+      "grad_norm": 1.6171875,
+      "learning_rate": 1.309085695158991e-05,
+      "loss": 0.8488,
+      "step": 7058
+    },
+    {
+      "epoch": 1.1983839509612706,
+      "grad_norm": 1.6484375,
+      "learning_rate": 1.3089139401755325e-05,
+      "loss": 0.8912,
+      "step": 7059
+    },
+    {
+      "epoch": 1.1985554150502604,
+      "grad_norm": 1.8046875,
+      "learning_rate": 1.3087421751172269e-05,
+      "loss": 0.8928,
+      "step": 7060
+    },
+    {
+      "epoch": 1.1987268791392502,
+      "grad_norm": 1.6328125,
+      "learning_rate": 1.3085703999896765e-05,
+      "loss": 0.8323,
+      "step": 7061
+    },
+    {
+      "epoch": 1.19889834322824,
+      "grad_norm": 1.625,
+      "learning_rate": 1.3083986147984835e-05,
+      "loss": 0.8322,
+      "step": 7062
+    },
+    {
+      "epoch": 1.1990698073172301,
+      "grad_norm": 1.6171875,
+      "learning_rate": 1.3082268195492505e-05,
+      "loss": 0.8949,
+      "step": 7063
+    },
+    {
+      "epoch": 1.19924127140622,
+      "grad_norm": 1.71875,
+      "learning_rate": 1.3080550142475798e-05,
+      "loss": 0.9076,
+      "step": 7064
+    },
+    {
+      "epoch": 1.1994127354952098,
+      "grad_norm": 1.7890625,
+      "learning_rate": 1.3078831988990757e-05,
+      "loss": 0.9873,
+      "step": 7065
+    },
+    {
+      "epoch": 1.1995841995841996,
+      "grad_norm": 1.7421875,
+      "learning_rate": 1.3077113735093407e-05,
+      "loss": 0.8335,
+      "step": 7066
+    },
+    {
+      "epoch": 1.1997556636731894,
+      "grad_norm": 1.7578125,
+      "learning_rate": 1.3075395380839787e-05,
+      "loss": 0.9648,
+      "step": 7067
+    },
+    {
+      "epoch": 1.1999271277621792,
+      "grad_norm": 1.75,
+      "learning_rate": 1.3073676926285947e-05,
+      "loss": 0.8638,
+      "step": 7068
+    },
+    {
+      "epoch": 1.200098591851169,
+      "grad_norm": 1.609375,
+      "learning_rate": 1.3071958371487927e-05,
+      "loss": 0.7763,
+      "step": 7069
+    },
+    {
+      "epoch": 1.2002700559401591,
+      "grad_norm": 1.59375,
+      "learning_rate": 1.3070239716501778e-05,
+      "loss": 0.7944,
+      "step": 7070
+    },
+    {
+      "epoch": 1.200441520029149,
+      "grad_norm": 1.671875,
+      "learning_rate": 1.3068520961383552e-05,
+      "loss": 0.894,
+      "step": 7071
+    },
+    {
+      "epoch": 1.2006129841181388,
+      "grad_norm": 1.6796875,
+      "learning_rate": 1.30668021061893e-05,
+      "loss": 0.9565,
+      "step": 7072
+    },
+    {
+      "epoch": 1.2007844482071286,
+      "grad_norm": 1.703125,
+      "learning_rate": 1.3065083150975081e-05,
+      "loss": 0.8892,
+      "step": 7073
+    },
+    {
+      "epoch": 1.2009559122961184,
+      "grad_norm": 1.6953125,
+      "learning_rate": 1.3063364095796962e-05,
+      "loss": 0.7983,
+      "step": 7074
+    },
+    {
+      "epoch": 1.2011273763851082,
+      "grad_norm": 1.59375,
+      "learning_rate": 1.3061644940711002e-05,
+      "loss": 0.8658,
+      "step": 7075
+    },
+    {
+      "epoch": 1.2012988404740983,
+      "grad_norm": 1.703125,
+      "learning_rate": 1.3059925685773273e-05,
+      "loss": 0.8818,
+      "step": 7076
+    },
+    {
+      "epoch": 1.2014703045630881,
+      "grad_norm": 1.8203125,
+      "learning_rate": 1.3058206331039842e-05,
+      "loss": 0.875,
+      "step": 7077
+    },
+    {
+      "epoch": 1.201641768652078,
+      "grad_norm": 1.53125,
+      "learning_rate": 1.305648687656679e-05,
+      "loss": 0.792,
+      "step": 7078
+    },
+    {
+      "epoch": 1.2018132327410678,
+      "grad_norm": 1.6796875,
+      "learning_rate": 1.3054767322410188e-05,
+      "loss": 0.7946,
+      "step": 7079
+    },
+    {
+      "epoch": 1.2019846968300576,
+      "grad_norm": 1.6484375,
+      "learning_rate": 1.3053047668626122e-05,
+      "loss": 0.812,
+      "step": 7080
+    },
+    {
+      "epoch": 1.2021561609190474,
+      "grad_norm": 1.6875,
+      "learning_rate": 1.3051327915270676e-05,
+      "loss": 0.8986,
+      "step": 7081
+    },
+    {
+      "epoch": 1.2023276250080375,
+      "grad_norm": 1.703125,
+      "learning_rate": 1.3049608062399934e-05,
+      "loss": 0.8617,
+      "step": 7082
+    },
+    {
+      "epoch": 1.2024990890970273,
+      "grad_norm": 1.6875,
+      "learning_rate": 1.3047888110069993e-05,
+      "loss": 0.8427,
+      "step": 7083
+    },
+    {
+      "epoch": 1.2026705531860171,
+      "grad_norm": 1.703125,
+      "learning_rate": 1.3046168058336941e-05,
+      "loss": 0.8435,
+      "step": 7084
+    },
+    {
+      "epoch": 1.202842017275007,
+      "grad_norm": 1.609375,
+      "learning_rate": 1.3044447907256877e-05,
+      "loss": 0.8159,
+      "step": 7085
+    },
+    {
+      "epoch": 1.2030134813639968,
+      "grad_norm": 1.65625,
+      "learning_rate": 1.30427276568859e-05,
+      "loss": 0.8177,
+      "step": 7086
+    },
+    {
+      "epoch": 1.2031849454529866,
+      "grad_norm": 1.640625,
+      "learning_rate": 1.304100730728012e-05,
+      "loss": 0.8673,
+      "step": 7087
+    },
+    {
+      "epoch": 1.2033564095419766,
+      "grad_norm": 1.6640625,
+      "learning_rate": 1.3039286858495642e-05,
+      "loss": 0.7668,
+      "step": 7088
+    },
+    {
+      "epoch": 1.2035278736309665,
+      "grad_norm": 1.734375,
+      "learning_rate": 1.3037566310588571e-05,
+      "loss": 0.9049,
+      "step": 7089
+    },
+    {
+      "epoch": 1.2036993377199563,
+      "grad_norm": 1.5546875,
+      "learning_rate": 1.3035845663615025e-05,
+      "loss": 0.8266,
+      "step": 7090
+    },
+    {
+      "epoch": 1.2038708018089461,
+      "grad_norm": 1.6484375,
+      "learning_rate": 1.3034124917631124e-05,
+      "loss": 0.7289,
+      "step": 7091
+    },
+    {
+      "epoch": 1.204042265897936,
+      "grad_norm": 1.703125,
+      "learning_rate": 1.3032404072692979e-05,
+      "loss": 0.8224,
+      "step": 7092
+    },
+    {
+      "epoch": 1.2042137299869258,
+      "grad_norm": 1.59375,
+      "learning_rate": 1.3030683128856719e-05,
+      "loss": 0.879,
+      "step": 7093
+    },
+    {
+      "epoch": 1.2043851940759158,
+      "grad_norm": 1.640625,
+      "learning_rate": 1.3028962086178472e-05,
+      "loss": 0.7933,
+      "step": 7094
+    },
+    {
+      "epoch": 1.2045566581649056,
+      "grad_norm": 1.65625,
+      "learning_rate": 1.3027240944714361e-05,
+      "loss": 0.8349,
+      "step": 7095
+    },
+    {
+      "epoch": 1.2047281222538955,
+      "grad_norm": 1.7265625,
+      "learning_rate": 1.3025519704520528e-05,
+      "loss": 0.8045,
+      "step": 7096
+    },
+    {
+      "epoch": 1.2048995863428853,
+      "grad_norm": 1.7109375,
+      "learning_rate": 1.3023798365653102e-05,
+      "loss": 0.8647,
+      "step": 7097
+    },
+    {
+      "epoch": 1.2050710504318751,
+      "grad_norm": 1.7734375,
+      "learning_rate": 1.3022076928168224e-05,
+      "loss": 0.848,
+      "step": 7098
+    },
+    {
+      "epoch": 1.205242514520865,
+      "grad_norm": 1.7421875,
+      "learning_rate": 1.302035539212204e-05,
+      "loss": 0.9455,
+      "step": 7099
+    },
+    {
+      "epoch": 1.205413978609855,
+      "grad_norm": 1.703125,
+      "learning_rate": 1.301863375757069e-05,
+      "loss": 0.8987,
+      "step": 7100
+    },
+    {
+      "epoch": 1.2055854426988448,
+      "grad_norm": 1.640625,
+      "learning_rate": 1.3016912024570329e-05,
+      "loss": 0.8748,
+      "step": 7101
+    },
+    {
+      "epoch": 1.2057569067878346,
+      "grad_norm": 1.6015625,
+      "learning_rate": 1.3015190193177105e-05,
+      "loss": 0.8823,
+      "step": 7102
+    },
+    {
+      "epoch": 1.2059283708768245,
+      "grad_norm": 1.625,
+      "learning_rate": 1.3013468263447175e-05,
+      "loss": 0.8718,
+      "step": 7103
+    },
+    {
+      "epoch": 1.2060998349658143,
+      "grad_norm": 1.7109375,
+      "learning_rate": 1.3011746235436698e-05,
+      "loss": 0.9014,
+      "step": 7104
+    },
+    {
+      "epoch": 1.2062712990548041,
+      "grad_norm": 1.65625,
+      "learning_rate": 1.3010024109201834e-05,
+      "loss": 0.8978,
+      "step": 7105
+    },
+    {
+      "epoch": 1.206442763143794,
+      "grad_norm": 1.6171875,
+      "learning_rate": 1.3008301884798746e-05,
+      "loss": 0.8466,
+      "step": 7106
+    },
+    {
+      "epoch": 1.206614227232784,
+      "grad_norm": 1.671875,
+      "learning_rate": 1.3006579562283607e-05,
+      "loss": 0.9149,
+      "step": 7107
+    },
+    {
+      "epoch": 1.2067856913217738,
+      "grad_norm": 1.703125,
+      "learning_rate": 1.300485714171259e-05,
+      "loss": 0.8793,
+      "step": 7108
+    },
+    {
+      "epoch": 1.2069571554107636,
+      "grad_norm": 1.5625,
+      "learning_rate": 1.3003134623141864e-05,
+      "loss": 0.8049,
+      "step": 7109
+    },
+    {
+      "epoch": 1.2071286194997535,
+      "grad_norm": 1.625,
+      "learning_rate": 1.300141200662761e-05,
+      "loss": 0.8108,
+      "step": 7110
+    },
+    {
+      "epoch": 1.2073000835887433,
+      "grad_norm": 1.5703125,
+      "learning_rate": 1.299968929222601e-05,
+      "loss": 0.7865,
+      "step": 7111
+    },
+    {
+      "epoch": 1.2074715476777333,
+      "grad_norm": 1.6640625,
+      "learning_rate": 1.2997966479993243e-05,
+      "loss": 0.8582,
+      "step": 7112
+    },
+    {
+      "epoch": 1.2076430117667232,
+      "grad_norm": 1.65625,
+      "learning_rate": 1.2996243569985501e-05,
+      "loss": 0.8529,
+      "step": 7113
+    },
+    {
+      "epoch": 1.207814475855713,
+      "grad_norm": 1.71875,
+      "learning_rate": 1.299452056225897e-05,
+      "loss": 0.8457,
+      "step": 7114
+    },
+    {
+      "epoch": 1.2079859399447028,
+      "grad_norm": 1.671875,
+      "learning_rate": 1.299279745686985e-05,
+      "loss": 0.8492,
+      "step": 7115
+    },
+    {
+      "epoch": 1.2081574040336927,
+      "grad_norm": 1.578125,
+      "learning_rate": 1.2991074253874337e-05,
+      "loss": 0.7936,
+      "step": 7116
+    },
+    {
+      "epoch": 1.2083288681226825,
+      "grad_norm": 1.625,
+      "learning_rate": 1.2989350953328628e-05,
+      "loss": 0.861,
+      "step": 7117
+    },
+    {
+      "epoch": 1.2085003322116723,
+      "grad_norm": 1.6328125,
+      "learning_rate": 1.2987627555288928e-05,
+      "loss": 0.7935,
+      "step": 7118
+    },
+    {
+      "epoch": 1.2086717963006623,
+      "grad_norm": 1.6953125,
+      "learning_rate": 1.2985904059811442e-05,
+      "loss": 0.8859,
+      "step": 7119
+    },
+    {
+      "epoch": 1.2088432603896522,
+      "grad_norm": 1.6015625,
+      "learning_rate": 1.2984180466952381e-05,
+      "loss": 0.8689,
+      "step": 7120
+    },
+    {
+      "epoch": 1.209014724478642,
+      "grad_norm": 1.625,
+      "learning_rate": 1.2982456776767957e-05,
+      "loss": 0.8682,
+      "step": 7121
+    },
+    {
+      "epoch": 1.2091861885676318,
+      "grad_norm": 1.6796875,
+      "learning_rate": 1.298073298931439e-05,
+      "loss": 0.803,
+      "step": 7122
+    },
+    {
+      "epoch": 1.2093576526566217,
+      "grad_norm": 1.71875,
+      "learning_rate": 1.297900910464789e-05,
+      "loss": 0.7887,
+      "step": 7123
+    },
+    {
+      "epoch": 1.2095291167456117,
+      "grad_norm": 1.8046875,
+      "learning_rate": 1.2977285122824688e-05,
+      "loss": 0.8895,
+      "step": 7124
+    },
+    {
+      "epoch": 1.2097005808346015,
+      "grad_norm": 1.6796875,
+      "learning_rate": 1.2975561043901008e-05,
+      "loss": 0.8874,
+      "step": 7125
+    },
+    {
+      "epoch": 1.2098720449235913,
+      "grad_norm": 1.7890625,
+      "learning_rate": 1.2973836867933076e-05,
+      "loss": 0.885,
+      "step": 7126
+    },
+    {
+      "epoch": 1.2100435090125812,
+      "grad_norm": 1.6328125,
+      "learning_rate": 1.2972112594977127e-05,
+      "loss": 0.8098,
+      "step": 7127
+    },
+    {
+      "epoch": 1.210214973101571,
+      "grad_norm": 1.7578125,
+      "learning_rate": 1.2970388225089393e-05,
+      "loss": 0.8832,
+      "step": 7128
+    },
+    {
+      "epoch": 1.2103864371905608,
+      "grad_norm": 1.8359375,
+      "learning_rate": 1.2968663758326115e-05,
+      "loss": 0.9134,
+      "step": 7129
+    },
+    {
+      "epoch": 1.2105579012795507,
+      "grad_norm": 1.6328125,
+      "learning_rate": 1.2966939194743531e-05,
+      "loss": 0.8345,
+      "step": 7130
+    },
+    {
+      "epoch": 1.2107293653685407,
+      "grad_norm": 1.78125,
+      "learning_rate": 1.2965214534397888e-05,
+      "loss": 0.8522,
+      "step": 7131
+    },
+    {
+      "epoch": 1.2109008294575305,
+      "grad_norm": 1.6484375,
+      "learning_rate": 1.2963489777345433e-05,
+      "loss": 0.8498,
+      "step": 7132
+    },
+    {
+      "epoch": 1.2110722935465204,
+      "grad_norm": 1.7109375,
+      "learning_rate": 1.2961764923642415e-05,
+      "loss": 0.8472,
+      "step": 7133
+    },
+    {
+      "epoch": 1.2112437576355102,
+      "grad_norm": 1.703125,
+      "learning_rate": 1.2960039973345094e-05,
+      "loss": 0.9188,
+      "step": 7134
+    },
+    {
+      "epoch": 1.2114152217245,
+      "grad_norm": 1.625,
+      "learning_rate": 1.295831492650972e-05,
+      "loss": 0.8582,
+      "step": 7135
+    },
+    {
+      "epoch": 1.21158668581349,
+      "grad_norm": 1.78125,
+      "learning_rate": 1.2956589783192557e-05,
+      "loss": 0.902,
+      "step": 7136
+    },
+    {
+      "epoch": 1.2117581499024799,
+      "grad_norm": 1.671875,
+      "learning_rate": 1.2954864543449866e-05,
+      "loss": 0.8307,
+      "step": 7137
+    },
+    {
+      "epoch": 1.2119296139914697,
+      "grad_norm": 1.703125,
+      "learning_rate": 1.2953139207337917e-05,
+      "loss": 0.9602,
+      "step": 7138
+    },
+    {
+      "epoch": 1.2121010780804595,
+      "grad_norm": 1.6328125,
+      "learning_rate": 1.2951413774912977e-05,
+      "loss": 0.8885,
+      "step": 7139
+    },
+    {
+      "epoch": 1.2122725421694494,
+      "grad_norm": 1.7578125,
+      "learning_rate": 1.2949688246231324e-05,
+      "loss": 0.9143,
+      "step": 7140
+    },
+    {
+      "epoch": 1.2124440062584392,
+      "grad_norm": 1.6796875,
+      "learning_rate": 1.2947962621349224e-05,
+      "loss": 0.955,
+      "step": 7141
+    },
+    {
+      "epoch": 1.212615470347429,
+      "grad_norm": 1.7265625,
+      "learning_rate": 1.2946236900322965e-05,
+      "loss": 0.8418,
+      "step": 7142
+    },
+    {
+      "epoch": 1.212786934436419,
+      "grad_norm": 1.71875,
+      "learning_rate": 1.2944511083208827e-05,
+      "loss": 0.8636,
+      "step": 7143
+    },
+    {
+      "epoch": 1.2129583985254089,
+      "grad_norm": 1.578125,
+      "learning_rate": 1.294278517006309e-05,
+      "loss": 0.8267,
+      "step": 7144
+    },
+    {
+      "epoch": 1.2131298626143987,
+      "grad_norm": 1.6328125,
+      "learning_rate": 1.294105916094205e-05,
+      "loss": 0.8761,
+      "step": 7145
+    },
+    {
+      "epoch": 1.2133013267033885,
+      "grad_norm": 1.640625,
+      "learning_rate": 1.2939333055902e-05,
+      "loss": 0.7752,
+      "step": 7146
+    },
+    {
+      "epoch": 1.2134727907923784,
+      "grad_norm": 1.671875,
+      "learning_rate": 1.2937606854999225e-05,
+      "loss": 0.8522,
+      "step": 7147
+    },
+    {
+      "epoch": 1.2136442548813684,
+      "grad_norm": 1.65625,
+      "learning_rate": 1.293588055829003e-05,
+      "loss": 0.8549,
+      "step": 7148
+    },
+    {
+      "epoch": 1.2138157189703582,
+      "grad_norm": 1.6328125,
+      "learning_rate": 1.2934154165830714e-05,
+      "loss": 0.8497,
+      "step": 7149
+    },
+    {
+      "epoch": 1.213987183059348,
+      "grad_norm": 1.65625,
+      "learning_rate": 1.2932427677677582e-05,
+      "loss": 0.8336,
+      "step": 7150
+    },
+    {
+      "epoch": 1.2141586471483379,
+      "grad_norm": 1.6953125,
+      "learning_rate": 1.2930701093886942e-05,
+      "loss": 0.8674,
+      "step": 7151
+    },
+    {
+      "epoch": 1.2143301112373277,
+      "grad_norm": 1.6875,
+      "learning_rate": 1.29289744145151e-05,
+      "loss": 0.9201,
+      "step": 7152
+    },
+    {
+      "epoch": 1.2145015753263175,
+      "grad_norm": 1.6953125,
+      "learning_rate": 1.2927247639618376e-05,
+      "loss": 0.8189,
+      "step": 7153
+    },
+    {
+      "epoch": 1.2146730394153074,
+      "grad_norm": 1.6953125,
+      "learning_rate": 1.2925520769253085e-05,
+      "loss": 0.9168,
+      "step": 7154
+    },
+    {
+      "epoch": 1.2148445035042974,
+      "grad_norm": 1.6640625,
+      "learning_rate": 1.2923793803475542e-05,
+      "loss": 0.8362,
+      "step": 7155
+    },
+    {
+      "epoch": 1.2150159675932872,
+      "grad_norm": 1.7421875,
+      "learning_rate": 1.2922066742342074e-05,
+      "loss": 0.8896,
+      "step": 7156
+    },
+    {
+      "epoch": 1.215187431682277,
+      "grad_norm": 1.6640625,
+      "learning_rate": 1.2920339585909006e-05,
+      "loss": 0.8406,
+      "step": 7157
+    },
+    {
+      "epoch": 1.2153588957712669,
+      "grad_norm": 1.6171875,
+      "learning_rate": 1.291861233423267e-05,
+      "loss": 0.8135,
+      "step": 7158
+    },
+    {
+      "epoch": 1.2155303598602567,
+      "grad_norm": 1.6640625,
+      "learning_rate": 1.2916884987369391e-05,
+      "loss": 0.8631,
+      "step": 7159
+    },
+    {
+      "epoch": 1.2157018239492468,
+      "grad_norm": 1.6875,
+      "learning_rate": 1.291515754537551e-05,
+      "loss": 0.8561,
+      "step": 7160
+    },
+    {
+      "epoch": 1.2158732880382366,
+      "grad_norm": 1.703125,
+      "learning_rate": 1.2913430008307361e-05,
+      "loss": 0.8446,
+      "step": 7161
+    },
+    {
+      "epoch": 1.2160447521272264,
+      "grad_norm": 1.5859375,
+      "learning_rate": 1.2911702376221294e-05,
+      "loss": 0.8365,
+      "step": 7162
+    },
+    {
+      "epoch": 1.2162162162162162,
+      "grad_norm": 1.6875,
+      "learning_rate": 1.2909974649173646e-05,
+      "loss": 0.8221,
+      "step": 7163
+    },
+    {
+      "epoch": 1.216387680305206,
+      "grad_norm": 1.6640625,
+      "learning_rate": 1.2908246827220766e-05,
+      "loss": 0.8318,
+      "step": 7164
+    },
+    {
+      "epoch": 1.2165591443941959,
+      "grad_norm": 1.71875,
+      "learning_rate": 1.2906518910419006e-05,
+      "loss": 0.8276,
+      "step": 7165
+    },
+    {
+      "epoch": 1.2167306084831857,
+      "grad_norm": 1.671875,
+      "learning_rate": 1.2904790898824719e-05,
+      "loss": 0.9029,
+      "step": 7166
+    },
+    {
+      "epoch": 1.2169020725721758,
+      "grad_norm": 1.640625,
+      "learning_rate": 1.290306279249426e-05,
+      "loss": 0.7511,
+      "step": 7167
+    },
+    {
+      "epoch": 1.2170735366611656,
+      "grad_norm": 1.625,
+      "learning_rate": 1.2901334591483996e-05,
+      "loss": 0.8239,
+      "step": 7168
+    },
+    {
+      "epoch": 1.2172450007501554,
+      "grad_norm": 1.71875,
+      "learning_rate": 1.2899606295850284e-05,
+      "loss": 0.8241,
+      "step": 7169
+    },
+    {
+      "epoch": 1.2174164648391452,
+      "grad_norm": 1.625,
+      "learning_rate": 1.2897877905649492e-05,
+      "loss": 0.76,
+      "step": 7170
+    },
+    {
+      "epoch": 1.217587928928135,
+      "grad_norm": 1.6640625,
+      "learning_rate": 1.2896149420937988e-05,
+      "loss": 0.876,
+      "step": 7171
+    },
+    {
+      "epoch": 1.2177593930171249,
+      "grad_norm": 1.7109375,
+      "learning_rate": 1.2894420841772144e-05,
+      "loss": 0.911,
+      "step": 7172
+    },
+    {
+      "epoch": 1.217930857106115,
+      "grad_norm": 1.78125,
+      "learning_rate": 1.289269216820834e-05,
+      "loss": 0.9441,
+      "step": 7173
+    },
+    {
+      "epoch": 1.2181023211951048,
+      "grad_norm": 1.734375,
+      "learning_rate": 1.2890963400302949e-05,
+      "loss": 0.8377,
+      "step": 7174
+    },
+    {
+      "epoch": 1.2182737852840946,
+      "grad_norm": 1.7578125,
+      "learning_rate": 1.2889234538112359e-05,
+      "loss": 0.8068,
+      "step": 7175
+    },
+    {
+      "epoch": 1.2184452493730844,
+      "grad_norm": 1.7265625,
+      "learning_rate": 1.288750558169295e-05,
+      "loss": 0.8127,
+      "step": 7176
+    },
+    {
+      "epoch": 1.2186167134620742,
+      "grad_norm": 1.6640625,
+      "learning_rate": 1.2885776531101109e-05,
+      "loss": 0.8431,
+      "step": 7177
+    },
+    {
+      "epoch": 1.218788177551064,
+      "grad_norm": 1.6328125,
+      "learning_rate": 1.2884047386393228e-05,
+      "loss": 0.8766,
+      "step": 7178
+    },
+    {
+      "epoch": 1.218959641640054,
+      "grad_norm": 1.765625,
+      "learning_rate": 1.2882318147625701e-05,
+      "loss": 1.0066,
+      "step": 7179
+    },
+    {
+      "epoch": 1.219131105729044,
+      "grad_norm": 1.7734375,
+      "learning_rate": 1.2880588814854923e-05,
+      "loss": 0.8912,
+      "step": 7180
+    },
+    {
+      "epoch": 1.2193025698180338,
+      "grad_norm": 1.6953125,
+      "learning_rate": 1.2878859388137304e-05,
+      "loss": 0.9064,
+      "step": 7181
+    },
+    {
+      "epoch": 1.2194740339070236,
+      "grad_norm": 1.6953125,
+      "learning_rate": 1.2877129867529234e-05,
+      "loss": 0.9499,
+      "step": 7182
+    },
+    {
+      "epoch": 1.2196454979960134,
+      "grad_norm": 1.8125,
+      "learning_rate": 1.2875400253087126e-05,
+      "loss": 0.8005,
+      "step": 7183
+    },
+    {
+      "epoch": 1.2198169620850032,
+      "grad_norm": 1.59375,
+      "learning_rate": 1.2873670544867389e-05,
+      "loss": 0.8108,
+      "step": 7184
+    },
+    {
+      "epoch": 1.2199884261739933,
+      "grad_norm": 1.6328125,
+      "learning_rate": 1.2871940742926432e-05,
+      "loss": 0.8299,
+      "step": 7185
+    },
+    {
+      "epoch": 1.220159890262983,
+      "grad_norm": 1.6875,
+      "learning_rate": 1.2870210847320676e-05,
+      "loss": 0.8547,
+      "step": 7186
+    },
+    {
+      "epoch": 1.220331354351973,
+      "grad_norm": 1.7421875,
+      "learning_rate": 1.2868480858106535e-05,
+      "loss": 0.9192,
+      "step": 7187
+    },
+    {
+      "epoch": 1.2205028184409628,
+      "grad_norm": 1.640625,
+      "learning_rate": 1.286675077534043e-05,
+      "loss": 0.8247,
+      "step": 7188
+    },
+    {
+      "epoch": 1.2206742825299526,
+      "grad_norm": 1.7734375,
+      "learning_rate": 1.2865020599078786e-05,
+      "loss": 0.9666,
+      "step": 7189
+    },
+    {
+      "epoch": 1.2208457466189424,
+      "grad_norm": 1.6875,
+      "learning_rate": 1.2863290329378033e-05,
+      "loss": 0.8251,
+      "step": 7190
+    },
+    {
+      "epoch": 1.2210172107079325,
+      "grad_norm": 1.703125,
+      "learning_rate": 1.2861559966294602e-05,
+      "loss": 0.8475,
+      "step": 7191
+    },
+    {
+      "epoch": 1.2211886747969223,
+      "grad_norm": 1.6640625,
+      "learning_rate": 1.2859829509884924e-05,
+      "loss": 0.8498,
+      "step": 7192
+    },
+    {
+      "epoch": 1.221360138885912,
+      "grad_norm": 1.71875,
+      "learning_rate": 1.2858098960205437e-05,
+      "loss": 0.9615,
+      "step": 7193
+    },
+    {
+      "epoch": 1.221531602974902,
+      "grad_norm": 1.75,
+      "learning_rate": 1.285636831731258e-05,
+      "loss": 0.8848,
+      "step": 7194
+    },
+    {
+      "epoch": 1.2217030670638918,
+      "grad_norm": 1.5859375,
+      "learning_rate": 1.2854637581262794e-05,
+      "loss": 0.8595,
+      "step": 7195
+    },
+    {
+      "epoch": 1.2218745311528816,
+      "grad_norm": 1.59375,
+      "learning_rate": 1.2852906752112528e-05,
+      "loss": 0.8858,
+      "step": 7196
+    },
+    {
+      "epoch": 1.2220459952418716,
+      "grad_norm": 1.734375,
+      "learning_rate": 1.285117582991823e-05,
+      "loss": 0.9023,
+      "step": 7197
+    },
+    {
+      "epoch": 1.2222174593308615,
+      "grad_norm": 1.640625,
+      "learning_rate": 1.2849444814736351e-05,
+      "loss": 0.8564,
+      "step": 7198
+    },
+    {
+      "epoch": 1.2223889234198513,
+      "grad_norm": 1.625,
+      "learning_rate": 1.2847713706623348e-05,
+      "loss": 0.9097,
+      "step": 7199
+    },
+    {
+      "epoch": 1.222560387508841,
+      "grad_norm": 1.640625,
+      "learning_rate": 1.2845982505635677e-05,
+      "loss": 0.8507,
+      "step": 7200
+    },
+    {
+      "epoch": 1.222731851597831,
+      "grad_norm": 1.6953125,
+      "learning_rate": 1.2844251211829799e-05,
+      "loss": 0.7985,
+      "step": 7201
+    },
+    {
+      "epoch": 1.2229033156868208,
+      "grad_norm": 1.71875,
+      "learning_rate": 1.284251982526218e-05,
+      "loss": 0.863,
+      "step": 7202
+    },
+    {
+      "epoch": 1.2230747797758106,
+      "grad_norm": 1.6640625,
+      "learning_rate": 1.284078834598928e-05,
+      "loss": 0.8775,
+      "step": 7203
+    },
+    {
+      "epoch": 1.2232462438648006,
+      "grad_norm": 1.6953125,
+      "learning_rate": 1.2839056774067583e-05,
+      "loss": 0.8685,
+      "step": 7204
+    },
+    {
+      "epoch": 1.2234177079537905,
+      "grad_norm": 1.71875,
+      "learning_rate": 1.2837325109553549e-05,
+      "loss": 0.8441,
+      "step": 7205
+    },
+    {
+      "epoch": 1.2235891720427803,
+      "grad_norm": 1.796875,
+      "learning_rate": 1.283559335250366e-05,
+      "loss": 0.8704,
+      "step": 7206
+    },
+    {
+      "epoch": 1.22376063613177,
+      "grad_norm": 1.6328125,
+      "learning_rate": 1.2833861502974392e-05,
+      "loss": 0.901,
+      "step": 7207
+    },
+    {
+      "epoch": 1.22393210022076,
+      "grad_norm": 1.6484375,
+      "learning_rate": 1.2832129561022228e-05,
+      "loss": 0.7814,
+      "step": 7208
+    },
+    {
+      "epoch": 1.22410356430975,
+      "grad_norm": 1.578125,
+      "learning_rate": 1.2830397526703659e-05,
+      "loss": 0.789,
+      "step": 7209
+    },
+    {
+      "epoch": 1.2242750283987398,
+      "grad_norm": 1.6953125,
+      "learning_rate": 1.2828665400075166e-05,
+      "loss": 0.9164,
+      "step": 7210
+    },
+    {
+      "epoch": 1.2244464924877296,
+      "grad_norm": 1.7734375,
+      "learning_rate": 1.2826933181193244e-05,
+      "loss": 0.9192,
+      "step": 7211
+    },
+    {
+      "epoch": 1.2246179565767195,
+      "grad_norm": 1.6640625,
+      "learning_rate": 1.2825200870114382e-05,
+      "loss": 0.8172,
+      "step": 7212
+    },
+    {
+      "epoch": 1.2247894206657093,
+      "grad_norm": 1.625,
+      "learning_rate": 1.2823468466895083e-05,
+      "loss": 0.8955,
+      "step": 7213
+    },
+    {
+      "epoch": 1.224960884754699,
+      "grad_norm": 1.625,
+      "learning_rate": 1.2821735971591846e-05,
+      "loss": 0.8579,
+      "step": 7214
+    },
+    {
+      "epoch": 1.225132348843689,
+      "grad_norm": 1.5859375,
+      "learning_rate": 1.2820003384261175e-05,
+      "loss": 0.8119,
+      "step": 7215
+    },
+    {
+      "epoch": 1.225303812932679,
+      "grad_norm": 1.6953125,
+      "learning_rate": 1.281827070495957e-05,
+      "loss": 0.9551,
+      "step": 7216
+    },
+    {
+      "epoch": 1.2254752770216688,
+      "grad_norm": 1.8203125,
+      "learning_rate": 1.2816537933743547e-05,
+      "loss": 0.9452,
+      "step": 7217
+    },
+    {
+      "epoch": 1.2256467411106586,
+      "grad_norm": 1.7421875,
+      "learning_rate": 1.2814805070669616e-05,
+      "loss": 0.8825,
+      "step": 7218
+    },
+    {
+      "epoch": 1.2258182051996485,
+      "grad_norm": 1.6328125,
+      "learning_rate": 1.2813072115794292e-05,
+      "loss": 0.84,
+      "step": 7219
+    },
+    {
+      "epoch": 1.2259896692886383,
+      "grad_norm": 1.765625,
+      "learning_rate": 1.2811339069174091e-05,
+      "loss": 0.9508,
+      "step": 7220
+    },
+    {
+      "epoch": 1.2261611333776283,
+      "grad_norm": 1.640625,
+      "learning_rate": 1.280960593086554e-05,
+      "loss": 0.8388,
+      "step": 7221
+    },
+    {
+      "epoch": 1.2263325974666182,
+      "grad_norm": 1.8125,
+      "learning_rate": 1.280787270092516e-05,
+      "loss": 0.8698,
+      "step": 7222
+    },
+    {
+      "epoch": 1.226504061555608,
+      "grad_norm": 1.6953125,
+      "learning_rate": 1.2806139379409475e-05,
+      "loss": 0.8033,
+      "step": 7223
+    },
+    {
+      "epoch": 1.2266755256445978,
+      "grad_norm": 1.8203125,
+      "learning_rate": 1.2804405966375018e-05,
+      "loss": 0.9809,
+      "step": 7224
+    },
+    {
+      "epoch": 1.2268469897335876,
+      "grad_norm": 1.7265625,
+      "learning_rate": 1.2802672461878323e-05,
+      "loss": 0.857,
+      "step": 7225
+    },
+    {
+      "epoch": 1.2270184538225775,
+      "grad_norm": 1.546875,
+      "learning_rate": 1.2800938865975923e-05,
+      "loss": 0.7538,
+      "step": 7226
+    },
+    {
+      "epoch": 1.2271899179115673,
+      "grad_norm": 1.609375,
+      "learning_rate": 1.2799205178724362e-05,
+      "loss": 0.8331,
+      "step": 7227
+    },
+    {
+      "epoch": 1.2273613820005573,
+      "grad_norm": 1.7578125,
+      "learning_rate": 1.2797471400180177e-05,
+      "loss": 0.9054,
+      "step": 7228
+    },
+    {
+      "epoch": 1.2275328460895472,
+      "grad_norm": 1.703125,
+      "learning_rate": 1.2795737530399919e-05,
+      "loss": 0.8271,
+      "step": 7229
+    },
+    {
+      "epoch": 1.227704310178537,
+      "grad_norm": 1.6328125,
+      "learning_rate": 1.2794003569440128e-05,
+      "loss": 0.836,
+      "step": 7230
+    },
+    {
+      "epoch": 1.2278757742675268,
+      "grad_norm": 1.703125,
+      "learning_rate": 1.2792269517357361e-05,
+      "loss": 0.9147,
+      "step": 7231
+    },
+    {
+      "epoch": 1.2280472383565166,
+      "grad_norm": 1.7109375,
+      "learning_rate": 1.279053537420817e-05,
+      "loss": 0.8463,
+      "step": 7232
+    },
+    {
+      "epoch": 1.2282187024455067,
+      "grad_norm": 1.640625,
+      "learning_rate": 1.2788801140049117e-05,
+      "loss": 0.7947,
+      "step": 7233
+    },
+    {
+      "epoch": 1.2283901665344965,
+      "grad_norm": 1.6640625,
+      "learning_rate": 1.2787066814936753e-05,
+      "loss": 0.8187,
+      "step": 7234
+    },
+    {
+      "epoch": 1.2285616306234863,
+      "grad_norm": 1.703125,
+      "learning_rate": 1.2785332398927641e-05,
+      "loss": 0.8147,
+      "step": 7235
+    },
+    {
+      "epoch": 1.2287330947124762,
+      "grad_norm": 1.703125,
+      "learning_rate": 1.2783597892078357e-05,
+      "loss": 0.8814,
+      "step": 7236
+    },
+    {
+      "epoch": 1.228904558801466,
+      "grad_norm": 1.609375,
+      "learning_rate": 1.278186329444546e-05,
+      "loss": 0.8836,
+      "step": 7237
+    },
+    {
+      "epoch": 1.2290760228904558,
+      "grad_norm": 1.640625,
+      "learning_rate": 1.278012860608553e-05,
+      "loss": 0.8651,
+      "step": 7238
+    },
+    {
+      "epoch": 1.2292474869794456,
+      "grad_norm": 1.6640625,
+      "learning_rate": 1.2778393827055133e-05,
+      "loss": 0.8794,
+      "step": 7239
+    },
+    {
+      "epoch": 1.2294189510684357,
+      "grad_norm": 1.65625,
+      "learning_rate": 1.2776658957410852e-05,
+      "loss": 0.7917,
+      "step": 7240
+    },
+    {
+      "epoch": 1.2295904151574255,
+      "grad_norm": 1.609375,
+      "learning_rate": 1.2774923997209268e-05,
+      "loss": 0.8094,
+      "step": 7241
+    },
+    {
+      "epoch": 1.2297618792464153,
+      "grad_norm": 1.609375,
+      "learning_rate": 1.277318894650696e-05,
+      "loss": 0.867,
+      "step": 7242
+    },
+    {
+      "epoch": 1.2299333433354052,
+      "grad_norm": 1.703125,
+      "learning_rate": 1.2771453805360521e-05,
+      "loss": 0.8815,
+      "step": 7243
+    },
+    {
+      "epoch": 1.230104807424395,
+      "grad_norm": 1.7421875,
+      "learning_rate": 1.2769718573826536e-05,
+      "loss": 0.8418,
+      "step": 7244
+    },
+    {
+      "epoch": 1.230276271513385,
+      "grad_norm": 1.6953125,
+      "learning_rate": 1.2767983251961596e-05,
+      "loss": 0.9079,
+      "step": 7245
+    },
+    {
+      "epoch": 1.2304477356023749,
+      "grad_norm": 1.703125,
+      "learning_rate": 1.2766247839822302e-05,
+      "loss": 0.8975,
+      "step": 7246
+    },
+    {
+      "epoch": 1.2306191996913647,
+      "grad_norm": 1.640625,
+      "learning_rate": 1.2764512337465247e-05,
+      "loss": 0.9299,
+      "step": 7247
+    },
+    {
+      "epoch": 1.2307906637803545,
+      "grad_norm": 1.703125,
+      "learning_rate": 1.2762776744947034e-05,
+      "loss": 0.8715,
+      "step": 7248
+    },
+    {
+      "epoch": 1.2309621278693443,
+      "grad_norm": 1.6171875,
+      "learning_rate": 1.2761041062324268e-05,
+      "loss": 0.8434,
+      "step": 7249
+    },
+    {
+      "epoch": 1.2311335919583342,
+      "grad_norm": 1.6953125,
+      "learning_rate": 1.275930528965356e-05,
+      "loss": 0.9107,
+      "step": 7250
+    },
+    {
+      "epoch": 1.231305056047324,
+      "grad_norm": 1.7265625,
+      "learning_rate": 1.2757569426991514e-05,
+      "loss": 0.9146,
+      "step": 7251
+    },
+    {
+      "epoch": 1.231476520136314,
+      "grad_norm": 1.6328125,
+      "learning_rate": 1.2755833474394744e-05,
+      "loss": 0.8134,
+      "step": 7252
+    },
+    {
+      "epoch": 1.2316479842253039,
+      "grad_norm": 1.5390625,
+      "learning_rate": 1.2754097431919864e-05,
+      "loss": 0.8832,
+      "step": 7253
+    },
+    {
+      "epoch": 1.2318194483142937,
+      "grad_norm": 1.703125,
+      "learning_rate": 1.2752361299623499e-05,
+      "loss": 0.9229,
+      "step": 7254
+    },
+    {
+      "epoch": 1.2319909124032835,
+      "grad_norm": 1.7109375,
+      "learning_rate": 1.2750625077562266e-05,
+      "loss": 0.9005,
+      "step": 7255
+    },
+    {
+      "epoch": 1.2321623764922733,
+      "grad_norm": 1.6640625,
+      "learning_rate": 1.2748888765792792e-05,
+      "loss": 0.902,
+      "step": 7256
+    },
+    {
+      "epoch": 1.2323338405812634,
+      "grad_norm": 1.6171875,
+      "learning_rate": 1.2747152364371706e-05,
+      "loss": 0.8608,
+      "step": 7257
+    },
+    {
+      "epoch": 1.2325053046702532,
+      "grad_norm": 1.7421875,
+      "learning_rate": 1.2745415873355636e-05,
+      "loss": 0.7951,
+      "step": 7258
+    },
+    {
+      "epoch": 1.232676768759243,
+      "grad_norm": 1.6796875,
+      "learning_rate": 1.2743679292801215e-05,
+      "loss": 0.8875,
+      "step": 7259
+    },
+    {
+      "epoch": 1.2328482328482329,
+      "grad_norm": 1.6640625,
+      "learning_rate": 1.274194262276508e-05,
+      "loss": 0.8669,
+      "step": 7260
+    },
+    {
+      "epoch": 1.2330196969372227,
+      "grad_norm": 1.703125,
+      "learning_rate": 1.2740205863303873e-05,
+      "loss": 0.942,
+      "step": 7261
+    },
+    {
+      "epoch": 1.2331911610262125,
+      "grad_norm": 1.625,
+      "learning_rate": 1.2738469014474233e-05,
+      "loss": 0.8201,
+      "step": 7262
+    },
+    {
+      "epoch": 1.2333626251152023,
+      "grad_norm": 1.6484375,
+      "learning_rate": 1.2736732076332802e-05,
+      "loss": 0.8711,
+      "step": 7263
+    },
+    {
+      "epoch": 1.2335340892041924,
+      "grad_norm": 1.71875,
+      "learning_rate": 1.2734995048936236e-05,
+      "loss": 1.0015,
+      "step": 7264
+    },
+    {
+      "epoch": 1.2337055532931822,
+      "grad_norm": 1.5546875,
+      "learning_rate": 1.2733257932341182e-05,
+      "loss": 0.7611,
+      "step": 7265
+    },
+    {
+      "epoch": 1.233877017382172,
+      "grad_norm": 1.78125,
+      "learning_rate": 1.2731520726604298e-05,
+      "loss": 0.8811,
+      "step": 7266
+    },
+    {
+      "epoch": 1.2340484814711619,
+      "grad_norm": 1.6953125,
+      "learning_rate": 1.2729783431782234e-05,
+      "loss": 0.9193,
+      "step": 7267
+    },
+    {
+      "epoch": 1.2342199455601517,
+      "grad_norm": 1.6875,
+      "learning_rate": 1.2728046047931655e-05,
+      "loss": 0.8882,
+      "step": 7268
+    },
+    {
+      "epoch": 1.2343914096491415,
+      "grad_norm": 1.65625,
+      "learning_rate": 1.2726308575109219e-05,
+      "loss": 0.8562,
+      "step": 7269
+    },
+    {
+      "epoch": 1.2345628737381316,
+      "grad_norm": 1.59375,
+      "learning_rate": 1.2724571013371593e-05,
+      "loss": 0.865,
+      "step": 7270
+    },
+    {
+      "epoch": 1.2347343378271214,
+      "grad_norm": 1.6875,
+      "learning_rate": 1.2722833362775448e-05,
+      "loss": 0.8359,
+      "step": 7271
+    },
+    {
+      "epoch": 1.2349058019161112,
+      "grad_norm": 1.6796875,
+      "learning_rate": 1.2721095623377456e-05,
+      "loss": 0.9229,
+      "step": 7272
+    },
+    {
+      "epoch": 1.235077266005101,
+      "grad_norm": 1.8515625,
+      "learning_rate": 1.2719357795234287e-05,
+      "loss": 0.9621,
+      "step": 7273
+    },
+    {
+      "epoch": 1.2352487300940909,
+      "grad_norm": 1.6171875,
+      "learning_rate": 1.2717619878402618e-05,
+      "loss": 0.834,
+      "step": 7274
+    },
+    {
+      "epoch": 1.2354201941830807,
+      "grad_norm": 1.546875,
+      "learning_rate": 1.2715881872939133e-05,
+      "loss": 0.7711,
+      "step": 7275
+    },
+    {
+      "epoch": 1.2355916582720707,
+      "grad_norm": 1.6484375,
+      "learning_rate": 1.2714143778900514e-05,
+      "loss": 0.8419,
+      "step": 7276
+    },
+    {
+      "epoch": 1.2357631223610606,
+      "grad_norm": 1.6328125,
+      "learning_rate": 1.2712405596343444e-05,
+      "loss": 0.8519,
+      "step": 7277
+    },
+    {
+      "epoch": 1.2359345864500504,
+      "grad_norm": 1.7109375,
+      "learning_rate": 1.2710667325324618e-05,
+      "loss": 0.9297,
+      "step": 7278
+    },
+    {
+      "epoch": 1.2361060505390402,
+      "grad_norm": 1.7109375,
+      "learning_rate": 1.2708928965900719e-05,
+      "loss": 0.9045,
+      "step": 7279
+    },
+    {
+      "epoch": 1.23627751462803,
+      "grad_norm": 1.7265625,
+      "learning_rate": 1.2707190518128446e-05,
+      "loss": 0.9259,
+      "step": 7280
+    },
+    {
+      "epoch": 1.2364489787170199,
+      "grad_norm": 1.6171875,
+      "learning_rate": 1.2705451982064497e-05,
+      "loss": 0.8243,
+      "step": 7281
+    },
+    {
+      "epoch": 1.23662044280601,
+      "grad_norm": 1.6796875,
+      "learning_rate": 1.2703713357765569e-05,
+      "loss": 0.8922,
+      "step": 7282
+    },
+    {
+      "epoch": 1.2367919068949997,
+      "grad_norm": 1.828125,
+      "learning_rate": 1.2701974645288366e-05,
+      "loss": 0.8736,
+      "step": 7283
+    },
+    {
+      "epoch": 1.2369633709839896,
+      "grad_norm": 1.7265625,
+      "learning_rate": 1.2700235844689602e-05,
+      "loss": 0.9215,
+      "step": 7284
+    },
+    {
+      "epoch": 1.2371348350729794,
+      "grad_norm": 1.6953125,
+      "learning_rate": 1.2698496956025972e-05,
+      "loss": 0.9459,
+      "step": 7285
+    },
+    {
+      "epoch": 1.2373062991619692,
+      "grad_norm": 1.7265625,
+      "learning_rate": 1.2696757979354199e-05,
+      "loss": 0.9388,
+      "step": 7286
+    },
+    {
+      "epoch": 1.237477763250959,
+      "grad_norm": 1.640625,
+      "learning_rate": 1.269501891473099e-05,
+      "loss": 0.8158,
+      "step": 7287
+    },
+    {
+      "epoch": 1.2376492273399489,
+      "grad_norm": 1.6796875,
+      "learning_rate": 1.2693279762213067e-05,
+      "loss": 0.8949,
+      "step": 7288
+    },
+    {
+      "epoch": 1.237820691428939,
+      "grad_norm": 1.6484375,
+      "learning_rate": 1.269154052185715e-05,
+      "loss": 0.8572,
+      "step": 7289
+    },
+    {
+      "epoch": 1.2379921555179287,
+      "grad_norm": 1.7109375,
+      "learning_rate": 1.268980119371996e-05,
+      "loss": 0.7945,
+      "step": 7290
+    },
+    {
+      "epoch": 1.2381636196069186,
+      "grad_norm": 1.765625,
+      "learning_rate": 1.2688061777858223e-05,
+      "loss": 0.9051,
+      "step": 7291
+    },
+    {
+      "epoch": 1.2383350836959084,
+      "grad_norm": 1.84375,
+      "learning_rate": 1.268632227432867e-05,
+      "loss": 0.9626,
+      "step": 7292
+    },
+    {
+      "epoch": 1.2385065477848982,
+      "grad_norm": 1.765625,
+      "learning_rate": 1.2684582683188033e-05,
+      "loss": 0.8529,
+      "step": 7293
+    },
+    {
+      "epoch": 1.2386780118738883,
+      "grad_norm": 1.6875,
+      "learning_rate": 1.2682843004493047e-05,
+      "loss": 0.8847,
+      "step": 7294
+    },
+    {
+      "epoch": 1.238849475962878,
+      "grad_norm": 1.7265625,
+      "learning_rate": 1.2681103238300446e-05,
+      "loss": 0.8428,
+      "step": 7295
+    },
+    {
+      "epoch": 1.239020940051868,
+      "grad_norm": 1.6953125,
+      "learning_rate": 1.2679363384666972e-05,
+      "loss": 0.8807,
+      "step": 7296
+    },
+    {
+      "epoch": 1.2391924041408577,
+      "grad_norm": 1.7109375,
+      "learning_rate": 1.267762344364937e-05,
+      "loss": 0.8548,
+      "step": 7297
+    },
+    {
+      "epoch": 1.2393638682298476,
+      "grad_norm": 1.625,
+      "learning_rate": 1.2675883415304383e-05,
+      "loss": 0.8498,
+      "step": 7298
+    },
+    {
+      "epoch": 1.2395353323188374,
+      "grad_norm": 1.75,
+      "learning_rate": 1.2674143299688761e-05,
+      "loss": 0.8959,
+      "step": 7299
+    },
+    {
+      "epoch": 1.2397067964078272,
+      "grad_norm": 1.6484375,
+      "learning_rate": 1.2672403096859257e-05,
+      "loss": 0.8864,
+      "step": 7300
+    },
+    {
+      "epoch": 1.2398782604968173,
+      "grad_norm": 1.7421875,
+      "learning_rate": 1.2670662806872625e-05,
+      "loss": 0.8839,
+      "step": 7301
+    },
+    {
+      "epoch": 1.240049724585807,
+      "grad_norm": 1.6328125,
+      "learning_rate": 1.2668922429785619e-05,
+      "loss": 0.8581,
+      "step": 7302
+    },
+    {
+      "epoch": 1.240221188674797,
+      "grad_norm": 1.6484375,
+      "learning_rate": 1.2667181965655006e-05,
+      "loss": 0.8189,
+      "step": 7303
+    },
+    {
+      "epoch": 1.2403926527637867,
+      "grad_norm": 1.6640625,
+      "learning_rate": 1.2665441414537544e-05,
+      "loss": 0.8233,
+      "step": 7304
+    },
+    {
+      "epoch": 1.2405641168527766,
+      "grad_norm": 1.6328125,
+      "learning_rate": 1.266370077649e-05,
+      "loss": 0.8569,
+      "step": 7305
+    },
+    {
+      "epoch": 1.2407355809417666,
+      "grad_norm": 1.6796875,
+      "learning_rate": 1.2661960051569144e-05,
+      "loss": 0.8426,
+      "step": 7306
+    },
+    {
+      "epoch": 1.2409070450307564,
+      "grad_norm": 1.59375,
+      "learning_rate": 1.2660219239831748e-05,
+      "loss": 0.842,
+      "step": 7307
+    },
+    {
+      "epoch": 1.2410785091197463,
+      "grad_norm": 1.6953125,
+      "learning_rate": 1.2658478341334583e-05,
+      "loss": 0.943,
+      "step": 7308
+    },
+    {
+      "epoch": 1.241249973208736,
+      "grad_norm": 1.6328125,
+      "learning_rate": 1.265673735613443e-05,
+      "loss": 0.8756,
+      "step": 7309
+    },
+    {
+      "epoch": 1.241421437297726,
+      "grad_norm": 1.5546875,
+      "learning_rate": 1.2654996284288063e-05,
+      "loss": 0.7903,
+      "step": 7310
+    },
+    {
+      "epoch": 1.2415929013867157,
+      "grad_norm": 1.6796875,
+      "learning_rate": 1.2653255125852272e-05,
+      "loss": 0.8261,
+      "step": 7311
+    },
+    {
+      "epoch": 1.2417643654757056,
+      "grad_norm": 1.6484375,
+      "learning_rate": 1.2651513880883841e-05,
+      "loss": 0.8329,
+      "step": 7312
+    },
+    {
+      "epoch": 1.2419358295646956,
+      "grad_norm": 1.7578125,
+      "learning_rate": 1.264977254943956e-05,
+      "loss": 0.9497,
+      "step": 7313
+    },
+    {
+      "epoch": 1.2421072936536854,
+      "grad_norm": 1.71875,
+      "learning_rate": 1.2648031131576217e-05,
+      "loss": 0.9085,
+      "step": 7314
+    },
+    {
+      "epoch": 1.2422787577426753,
+      "grad_norm": 1.7421875,
+      "learning_rate": 1.2646289627350605e-05,
+      "loss": 0.8737,
+      "step": 7315
+    },
+    {
+      "epoch": 1.242450221831665,
+      "grad_norm": 1.6953125,
+      "learning_rate": 1.2644548036819525e-05,
+      "loss": 0.8706,
+      "step": 7316
+    },
+    {
+      "epoch": 1.242621685920655,
+      "grad_norm": 1.65625,
+      "learning_rate": 1.2642806360039774e-05,
+      "loss": 0.7986,
+      "step": 7317
+    },
+    {
+      "epoch": 1.242793150009645,
+      "grad_norm": 1.6484375,
+      "learning_rate": 1.2641064597068156e-05,
+      "loss": 0.9261,
+      "step": 7318
+    },
+    {
+      "epoch": 1.2429646140986348,
+      "grad_norm": 1.734375,
+      "learning_rate": 1.2639322747961476e-05,
+      "loss": 0.8666,
+      "step": 7319
+    },
+    {
+      "epoch": 1.2431360781876246,
+      "grad_norm": 1.640625,
+      "learning_rate": 1.2637580812776542e-05,
+      "loss": 0.8477,
+      "step": 7320
+    },
+    {
+      "epoch": 1.2433075422766144,
+      "grad_norm": 1.6953125,
+      "learning_rate": 1.2635838791570166e-05,
+      "loss": 0.8551,
+      "step": 7321
+    },
+    {
+      "epoch": 1.2434790063656043,
+      "grad_norm": 1.71875,
+      "learning_rate": 1.263409668439916e-05,
+      "loss": 0.954,
+      "step": 7322
+    },
+    {
+      "epoch": 1.243650470454594,
+      "grad_norm": 1.7421875,
+      "learning_rate": 1.2632354491320342e-05,
+      "loss": 0.8462,
+      "step": 7323
+    },
+    {
+      "epoch": 1.243821934543584,
+      "grad_norm": 1.734375,
+      "learning_rate": 1.2630612212390534e-05,
+      "loss": 0.8168,
+      "step": 7324
+    },
+    {
+      "epoch": 1.243993398632574,
+      "grad_norm": 1.578125,
+      "learning_rate": 1.2628869847666556e-05,
+      "loss": 0.7898,
+      "step": 7325
+    },
+    {
+      "epoch": 1.2441648627215638,
+      "grad_norm": 1.6328125,
+      "learning_rate": 1.2627127397205232e-05,
+      "loss": 0.803,
+      "step": 7326
+    },
+    {
+      "epoch": 1.2443363268105536,
+      "grad_norm": 1.7578125,
+      "learning_rate": 1.262538486106339e-05,
+      "loss": 0.8628,
+      "step": 7327
+    },
+    {
+      "epoch": 1.2445077908995434,
+      "grad_norm": 1.703125,
+      "learning_rate": 1.2623642239297862e-05,
+      "loss": 0.8727,
+      "step": 7328
+    },
+    {
+      "epoch": 1.2446792549885333,
+      "grad_norm": 1.7265625,
+      "learning_rate": 1.2621899531965477e-05,
+      "loss": 0.9508,
+      "step": 7329
+    },
+    {
+      "epoch": 1.2448507190775233,
+      "grad_norm": 1.6796875,
+      "learning_rate": 1.2620156739123079e-05,
+      "loss": 0.8475,
+      "step": 7330
+    },
+    {
+      "epoch": 1.2450221831665131,
+      "grad_norm": 1.6796875,
+      "learning_rate": 1.2618413860827506e-05,
+      "loss": 0.8502,
+      "step": 7331
+    },
+    {
+      "epoch": 1.245193647255503,
+      "grad_norm": 1.609375,
+      "learning_rate": 1.2616670897135592e-05,
+      "loss": 0.7271,
+      "step": 7332
+    },
+    {
+      "epoch": 1.2453651113444928,
+      "grad_norm": 1.6875,
+      "learning_rate": 1.2614927848104189e-05,
+      "loss": 0.851,
+      "step": 7333
+    },
+    {
+      "epoch": 1.2455365754334826,
+      "grad_norm": 1.59375,
+      "learning_rate": 1.261318471379014e-05,
+      "loss": 0.7724,
+      "step": 7334
+    },
+    {
+      "epoch": 1.2457080395224724,
+      "grad_norm": 1.71875,
+      "learning_rate": 1.2611441494250299e-05,
+      "loss": 0.8478,
+      "step": 7335
+    },
+    {
+      "epoch": 1.2458795036114623,
+      "grad_norm": 1.6640625,
+      "learning_rate": 1.260969818954152e-05,
+      "loss": 0.8933,
+      "step": 7336
+    },
+    {
+      "epoch": 1.2460509677004523,
+      "grad_norm": 1.609375,
+      "learning_rate": 1.2607954799720647e-05,
+      "loss": 0.8304,
+      "step": 7337
+    },
+    {
+      "epoch": 1.2462224317894421,
+      "grad_norm": 1.6796875,
+      "learning_rate": 1.2606211324844555e-05,
+      "loss": 0.8571,
+      "step": 7338
+    },
+    {
+      "epoch": 1.246393895878432,
+      "grad_norm": 1.6875,
+      "learning_rate": 1.2604467764970094e-05,
+      "loss": 0.8683,
+      "step": 7339
+    },
+    {
+      "epoch": 1.2465653599674218,
+      "grad_norm": 1.671875,
+      "learning_rate": 1.2602724120154134e-05,
+      "loss": 0.8936,
+      "step": 7340
+    },
+    {
+      "epoch": 1.2467368240564116,
+      "grad_norm": 1.7109375,
+      "learning_rate": 1.2600980390453537e-05,
+      "loss": 0.9597,
+      "step": 7341
+    },
+    {
+      "epoch": 1.2469082881454017,
+      "grad_norm": 1.71875,
+      "learning_rate": 1.2599236575925178e-05,
+      "loss": 0.8553,
+      "step": 7342
+    },
+    {
+      "epoch": 1.2470797522343915,
+      "grad_norm": 1.6328125,
+      "learning_rate": 1.2597492676625925e-05,
+      "loss": 0.806,
+      "step": 7343
+    },
+    {
+      "epoch": 1.2472512163233813,
+      "grad_norm": 1.8125,
+      "learning_rate": 1.2595748692612654e-05,
+      "loss": 0.81,
+      "step": 7344
+    },
+    {
+      "epoch": 1.2474226804123711,
+      "grad_norm": 1.6953125,
+      "learning_rate": 1.2594004623942244e-05,
+      "loss": 0.8586,
+      "step": 7345
+    },
+    {
+      "epoch": 1.247594144501361,
+      "grad_norm": 1.6640625,
+      "learning_rate": 1.2592260470671576e-05,
+      "loss": 0.8506,
+      "step": 7346
+    },
+    {
+      "epoch": 1.2477656085903508,
+      "grad_norm": 1.703125,
+      "learning_rate": 1.259051623285753e-05,
+      "loss": 0.9173,
+      "step": 7347
+    },
+    {
+      "epoch": 1.2479370726793406,
+      "grad_norm": 1.7265625,
+      "learning_rate": 1.2588771910556996e-05,
+      "loss": 0.8443,
+      "step": 7348
+    },
+    {
+      "epoch": 1.2481085367683307,
+      "grad_norm": 1.7109375,
+      "learning_rate": 1.2587027503826863e-05,
+      "loss": 0.8387,
+      "step": 7349
+    },
+    {
+      "epoch": 1.2482800008573205,
+      "grad_norm": 1.59375,
+      "learning_rate": 1.2585283012724019e-05,
+      "loss": 0.852,
+      "step": 7350
+    },
+    {
+      "epoch": 1.2484514649463103,
+      "grad_norm": 1.5625,
+      "learning_rate": 1.2583538437305363e-05,
+      "loss": 0.8271,
+      "step": 7351
+    },
+    {
+      "epoch": 1.2486229290353001,
+      "grad_norm": 1.609375,
+      "learning_rate": 1.2581793777627787e-05,
+      "loss": 0.9076,
+      "step": 7352
+    },
+    {
+      "epoch": 1.24879439312429,
+      "grad_norm": 1.609375,
+      "learning_rate": 1.2580049033748195e-05,
+      "loss": 0.9031,
+      "step": 7353
+    },
+    {
+      "epoch": 1.24896585721328,
+      "grad_norm": 1.703125,
+      "learning_rate": 1.257830420572349e-05,
+      "loss": 0.9212,
+      "step": 7354
+    },
+    {
+      "epoch": 1.2491373213022698,
+      "grad_norm": 1.671875,
+      "learning_rate": 1.2576559293610575e-05,
+      "loss": 0.8859,
+      "step": 7355
+    },
+    {
+      "epoch": 1.2493087853912597,
+      "grad_norm": 1.7109375,
+      "learning_rate": 1.257481429746636e-05,
+      "loss": 0.815,
+      "step": 7356
+    },
+    {
+      "epoch": 1.2494802494802495,
+      "grad_norm": 1.734375,
+      "learning_rate": 1.2573069217347751e-05,
+      "loss": 0.9199,
+      "step": 7357
+    },
+    {
+      "epoch": 1.2496517135692393,
+      "grad_norm": 1.6875,
+      "learning_rate": 1.257132405331167e-05,
+      "loss": 0.7968,
+      "step": 7358
+    },
+    {
+      "epoch": 1.2498231776582291,
+      "grad_norm": 1.6875,
+      "learning_rate": 1.2569578805415026e-05,
+      "loss": 0.8123,
+      "step": 7359
+    },
+    {
+      "epoch": 1.249994641747219,
+      "grad_norm": 1.5625,
+      "learning_rate": 1.2567833473714743e-05,
+      "loss": 0.8134,
+      "step": 7360
+    },
+    {
+      "epoch": 1.2501661058362088,
+      "grad_norm": 1.796875,
+      "learning_rate": 1.2566088058267737e-05,
+      "loss": 0.8872,
+      "step": 7361
+    },
+    {
+      "epoch": 1.2503375699251988,
+      "grad_norm": 1.671875,
+      "learning_rate": 1.256434255913094e-05,
+      "loss": 0.8235,
+      "step": 7362
+    },
+    {
+      "epoch": 1.2505090340141887,
+      "grad_norm": 1.59375,
+      "learning_rate": 1.2562596976361276e-05,
+      "loss": 0.8185,
+      "step": 7363
+    },
+    {
+      "epoch": 1.2506804981031785,
+      "grad_norm": 1.609375,
+      "learning_rate": 1.2560851310015674e-05,
+      "loss": 0.861,
+      "step": 7364
+    },
+    {
+      "epoch": 1.2508519621921683,
+      "grad_norm": 1.6953125,
+      "learning_rate": 1.2559105560151065e-05,
+      "loss": 0.8713,
+      "step": 7365
+    },
+    {
+      "epoch": 1.2510234262811584,
+      "grad_norm": 1.6640625,
+      "learning_rate": 1.2557359726824392e-05,
+      "loss": 0.8448,
+      "step": 7366
+    },
+    {
+      "epoch": 1.2511948903701482,
+      "grad_norm": 1.6875,
+      "learning_rate": 1.2555613810092585e-05,
+      "loss": 0.8949,
+      "step": 7367
+    },
+    {
+      "epoch": 1.251366354459138,
+      "grad_norm": 1.6328125,
+      "learning_rate": 1.2553867810012588e-05,
+      "loss": 0.8536,
+      "step": 7368
+    },
+    {
+      "epoch": 1.2515378185481278,
+      "grad_norm": 1.703125,
+      "learning_rate": 1.2552121726641344e-05,
+      "loss": 0.8656,
+      "step": 7369
+    },
+    {
+      "epoch": 1.2517092826371177,
+      "grad_norm": 1.7734375,
+      "learning_rate": 1.2550375560035804e-05,
+      "loss": 0.8059,
+      "step": 7370
+    },
+    {
+      "epoch": 1.2518807467261075,
+      "grad_norm": 1.6015625,
+      "learning_rate": 1.2548629310252912e-05,
+      "loss": 0.8736,
+      "step": 7371
+    },
+    {
+      "epoch": 1.2520522108150973,
+      "grad_norm": 1.7421875,
+      "learning_rate": 1.254688297734962e-05,
+      "loss": 0.9381,
+      "step": 7372
+    },
+    {
+      "epoch": 1.2522236749040871,
+      "grad_norm": 1.65625,
+      "learning_rate": 1.254513656138288e-05,
+      "loss": 0.8456,
+      "step": 7373
+    },
+    {
+      "epoch": 1.2523951389930772,
+      "grad_norm": 1.609375,
+      "learning_rate": 1.2543390062409655e-05,
+      "loss": 0.8379,
+      "step": 7374
+    },
+    {
+      "epoch": 1.252566603082067,
+      "grad_norm": 1.6796875,
+      "learning_rate": 1.2541643480486905e-05,
+      "loss": 0.868,
+      "step": 7375
+    },
+    {
+      "epoch": 1.2527380671710568,
+      "grad_norm": 1.71875,
+      "learning_rate": 1.2539896815671585e-05,
+      "loss": 0.8134,
+      "step": 7376
+    },
+    {
+      "epoch": 1.2529095312600467,
+      "grad_norm": 1.7578125,
+      "learning_rate": 1.2538150068020672e-05,
+      "loss": 0.8924,
+      "step": 7377
+    },
+    {
+      "epoch": 1.2530809953490367,
+      "grad_norm": 1.8046875,
+      "learning_rate": 1.2536403237591125e-05,
+      "loss": 0.9193,
+      "step": 7378
+    },
+    {
+      "epoch": 1.2532524594380265,
+      "grad_norm": 1.8046875,
+      "learning_rate": 1.2534656324439912e-05,
+      "loss": 0.8861,
+      "step": 7379
+    },
+    {
+      "epoch": 1.2534239235270164,
+      "grad_norm": 1.6328125,
+      "learning_rate": 1.253290932862402e-05,
+      "loss": 0.8459,
+      "step": 7380
+    },
+    {
+      "epoch": 1.2535953876160062,
+      "grad_norm": 1.7109375,
+      "learning_rate": 1.2531162250200411e-05,
+      "loss": 0.9036,
+      "step": 7381
+    },
+    {
+      "epoch": 1.253766851704996,
+      "grad_norm": 1.6875,
+      "learning_rate": 1.2529415089226073e-05,
+      "loss": 0.9044,
+      "step": 7382
+    },
+    {
+      "epoch": 1.2539383157939858,
+      "grad_norm": 1.7421875,
+      "learning_rate": 1.2527667845757983e-05,
+      "loss": 0.8205,
+      "step": 7383
+    },
+    {
+      "epoch": 1.2541097798829757,
+      "grad_norm": 1.7265625,
+      "learning_rate": 1.2525920519853123e-05,
+      "loss": 0.8355,
+      "step": 7384
+    },
+    {
+      "epoch": 1.2542812439719655,
+      "grad_norm": 1.625,
+      "learning_rate": 1.2524173111568485e-05,
+      "loss": 0.8957,
+      "step": 7385
+    },
+    {
+      "epoch": 1.2544527080609555,
+      "grad_norm": 1.6875,
+      "learning_rate": 1.2522425620961059e-05,
+      "loss": 0.8491,
+      "step": 7386
+    },
+    {
+      "epoch": 1.2546241721499454,
+      "grad_norm": 1.6640625,
+      "learning_rate": 1.252067804808783e-05,
+      "loss": 0.7969,
+      "step": 7387
+    },
+    {
+      "epoch": 1.2547956362389352,
+      "grad_norm": 1.7265625,
+      "learning_rate": 1.2518930393005807e-05,
+      "loss": 0.8935,
+      "step": 7388
+    },
+    {
+      "epoch": 1.254967100327925,
+      "grad_norm": 1.65625,
+      "learning_rate": 1.2517182655771971e-05,
+      "loss": 0.814,
+      "step": 7389
+    },
+    {
+      "epoch": 1.255138564416915,
+      "grad_norm": 1.6484375,
+      "learning_rate": 1.2515434836443331e-05,
+      "loss": 0.8395,
+      "step": 7390
+    },
+    {
+      "epoch": 1.255310028505905,
+      "grad_norm": 1.640625,
+      "learning_rate": 1.251368693507689e-05,
+      "loss": 0.8807,
+      "step": 7391
+    },
+    {
+      "epoch": 1.2554814925948947,
+      "grad_norm": 1.7109375,
+      "learning_rate": 1.251193895172965e-05,
+      "loss": 0.9083,
+      "step": 7392
+    },
+    {
+      "epoch": 1.2556529566838845,
+      "grad_norm": 1.6796875,
+      "learning_rate": 1.2510190886458625e-05,
+      "loss": 0.9081,
+      "step": 7393
+    },
+    {
+      "epoch": 1.2558244207728744,
+      "grad_norm": 1.6015625,
+      "learning_rate": 1.250844273932082e-05,
+      "loss": 0.8703,
+      "step": 7394
+    },
+    {
+      "epoch": 1.2559958848618642,
+      "grad_norm": 1.5625,
+      "learning_rate": 1.2506694510373252e-05,
+      "loss": 0.8511,
+      "step": 7395
+    },
+    {
+      "epoch": 1.256167348950854,
+      "grad_norm": 1.671875,
+      "learning_rate": 1.2504946199672935e-05,
+      "loss": 0.8408,
+      "step": 7396
+    },
+    {
+      "epoch": 1.2563388130398438,
+      "grad_norm": 1.75,
+      "learning_rate": 1.2503197807276894e-05,
+      "loss": 0.9484,
+      "step": 7397
+    },
+    {
+      "epoch": 1.256510277128834,
+      "grad_norm": 1.6171875,
+      "learning_rate": 1.2501449333242144e-05,
+      "loss": 0.8141,
+      "step": 7398
+    },
+    {
+      "epoch": 1.2566817412178237,
+      "grad_norm": 1.7109375,
+      "learning_rate": 1.2499700777625709e-05,
+      "loss": 0.9525,
+      "step": 7399
+    },
+    {
+      "epoch": 1.2568532053068135,
+      "grad_norm": 1.671875,
+      "learning_rate": 1.2497952140484624e-05,
+      "loss": 0.9133,
+      "step": 7400
+    },
+    {
+      "epoch": 1.2570246693958034,
+      "grad_norm": 1.71875,
+      "learning_rate": 1.2496203421875912e-05,
+      "loss": 0.8465,
+      "step": 7401
+    },
+    {
+      "epoch": 1.2571961334847932,
+      "grad_norm": 1.71875,
+      "learning_rate": 1.2494454621856604e-05,
+      "loss": 0.8892,
+      "step": 7402
+    },
+    {
+      "epoch": 1.2573675975737832,
+      "grad_norm": 1.6875,
+      "learning_rate": 1.249270574048374e-05,
+      "loss": 0.8771,
+      "step": 7403
+    },
+    {
+      "epoch": 1.257539061662773,
+      "grad_norm": 1.625,
+      "learning_rate": 1.249095677781435e-05,
+      "loss": 0.9029,
+      "step": 7404
+    },
+    {
+      "epoch": 1.257710525751763,
+      "grad_norm": 1.6640625,
+      "learning_rate": 1.2489207733905485e-05,
+      "loss": 0.8047,
+      "step": 7405
+    },
+    {
+      "epoch": 1.2578819898407527,
+      "grad_norm": 1.703125,
+      "learning_rate": 1.248745860881418e-05,
+      "loss": 0.8067,
+      "step": 7406
+    },
+    {
+      "epoch": 1.2580534539297425,
+      "grad_norm": 1.5859375,
+      "learning_rate": 1.2485709402597483e-05,
+      "loss": 0.8371,
+      "step": 7407
+    },
+    {
+      "epoch": 1.2582249180187324,
+      "grad_norm": 1.78125,
+      "learning_rate": 1.2483960115312443e-05,
+      "loss": 0.9358,
+      "step": 7408
+    },
+    {
+      "epoch": 1.2583963821077222,
+      "grad_norm": 1.6015625,
+      "learning_rate": 1.248221074701611e-05,
+      "loss": 0.8565,
+      "step": 7409
+    },
+    {
+      "epoch": 1.2585678461967122,
+      "grad_norm": 1.8359375,
+      "learning_rate": 1.2480461297765536e-05,
+      "loss": 0.8881,
+      "step": 7410
+    },
+    {
+      "epoch": 1.258739310285702,
+      "grad_norm": 1.671875,
+      "learning_rate": 1.2478711767617782e-05,
+      "loss": 0.8481,
+      "step": 7411
+    },
+    {
+      "epoch": 1.258910774374692,
+      "grad_norm": 1.796875,
+      "learning_rate": 1.2476962156629897e-05,
+      "loss": 0.8353,
+      "step": 7412
+    },
+    {
+      "epoch": 1.2590822384636817,
+      "grad_norm": 1.7421875,
+      "learning_rate": 1.2475212464858952e-05,
+      "loss": 0.851,
+      "step": 7413
+    },
+    {
+      "epoch": 1.2592537025526715,
+      "grad_norm": 1.7734375,
+      "learning_rate": 1.2473462692362008e-05,
+      "loss": 0.8653,
+      "step": 7414
+    },
+    {
+      "epoch": 1.2594251666416616,
+      "grad_norm": 1.6328125,
+      "learning_rate": 1.2471712839196133e-05,
+      "loss": 0.7982,
+      "step": 7415
+    },
+    {
+      "epoch": 1.2595966307306514,
+      "grad_norm": 1.6328125,
+      "learning_rate": 1.2469962905418393e-05,
+      "loss": 0.8258,
+      "step": 7416
+    },
+    {
+      "epoch": 1.2597680948196412,
+      "grad_norm": 1.640625,
+      "learning_rate": 1.2468212891085862e-05,
+      "loss": 0.8672,
+      "step": 7417
+    },
+    {
+      "epoch": 1.259939558908631,
+      "grad_norm": 1.671875,
+      "learning_rate": 1.2466462796255617e-05,
+      "loss": 0.7899,
+      "step": 7418
+    },
+    {
+      "epoch": 1.260111022997621,
+      "grad_norm": 1.6484375,
+      "learning_rate": 1.2464712620984727e-05,
+      "loss": 0.829,
+      "step": 7419
+    },
+    {
+      "epoch": 1.2602824870866107,
+      "grad_norm": 1.703125,
+      "learning_rate": 1.246296236533028e-05,
+      "loss": 0.8621,
+      "step": 7420
+    },
+    {
+      "epoch": 1.2604539511756006,
+      "grad_norm": 1.671875,
+      "learning_rate": 1.2461212029349354e-05,
+      "loss": 0.8306,
+      "step": 7421
+    },
+    {
+      "epoch": 1.2606254152645906,
+      "grad_norm": 1.7890625,
+      "learning_rate": 1.2459461613099037e-05,
+      "loss": 0.8217,
+      "step": 7422
+    },
+    {
+      "epoch": 1.2607968793535804,
+      "grad_norm": 1.6484375,
+      "learning_rate": 1.2457711116636416e-05,
+      "loss": 0.8073,
+      "step": 7423
+    },
+    {
+      "epoch": 1.2609683434425702,
+      "grad_norm": 1.671875,
+      "learning_rate": 1.2455960540018579e-05,
+      "loss": 0.8963,
+      "step": 7424
+    },
+    {
+      "epoch": 1.26113980753156,
+      "grad_norm": 1.6640625,
+      "learning_rate": 1.2454209883302622e-05,
+      "loss": 0.8579,
+      "step": 7425
+    },
+    {
+      "epoch": 1.26131127162055,
+      "grad_norm": 1.7265625,
+      "learning_rate": 1.2452459146545637e-05,
+      "loss": 0.8723,
+      "step": 7426
+    },
+    {
+      "epoch": 1.26148273570954,
+      "grad_norm": 1.7734375,
+      "learning_rate": 1.2450708329804724e-05,
+      "loss": 0.833,
+      "step": 7427
+    },
+    {
+      "epoch": 1.2616541997985298,
+      "grad_norm": 1.6328125,
+      "learning_rate": 1.2448957433136985e-05,
+      "loss": 0.8809,
+      "step": 7428
+    },
+    {
+      "epoch": 1.2618256638875196,
+      "grad_norm": 1.6796875,
+      "learning_rate": 1.2447206456599525e-05,
+      "loss": 0.8621,
+      "step": 7429
+    },
+    {
+      "epoch": 1.2619971279765094,
+      "grad_norm": 1.6796875,
+      "learning_rate": 1.2445455400249444e-05,
+      "loss": 0.8685,
+      "step": 7430
+    },
+    {
+      "epoch": 1.2621685920654993,
+      "grad_norm": 1.6875,
+      "learning_rate": 1.2443704264143852e-05,
+      "loss": 0.8792,
+      "step": 7431
+    },
+    {
+      "epoch": 1.262340056154489,
+      "grad_norm": 1.75,
+      "learning_rate": 1.2441953048339866e-05,
+      "loss": 0.8524,
+      "step": 7432
+    },
+    {
+      "epoch": 1.262511520243479,
+      "grad_norm": 1.703125,
+      "learning_rate": 1.2440201752894592e-05,
+      "loss": 0.832,
+      "step": 7433
+    },
+    {
+      "epoch": 1.262682984332469,
+      "grad_norm": 1.703125,
+      "learning_rate": 1.2438450377865153e-05,
+      "loss": 0.8459,
+      "step": 7434
+    },
+    {
+      "epoch": 1.2628544484214588,
+      "grad_norm": 1.5546875,
+      "learning_rate": 1.2436698923308662e-05,
+      "loss": 0.7745,
+      "step": 7435
+    },
+    {
+      "epoch": 1.2630259125104486,
+      "grad_norm": 1.703125,
+      "learning_rate": 1.2434947389282247e-05,
+      "loss": 0.8945,
+      "step": 7436
+    },
+    {
+      "epoch": 1.2631973765994384,
+      "grad_norm": 1.6875,
+      "learning_rate": 1.2433195775843026e-05,
+      "loss": 0.8573,
+      "step": 7437
+    },
+    {
+      "epoch": 1.2633688406884283,
+      "grad_norm": 1.6953125,
+      "learning_rate": 1.2431444083048128e-05,
+      "loss": 0.8573,
+      "step": 7438
+    },
+    {
+      "epoch": 1.2635403047774183,
+      "grad_norm": 1.765625,
+      "learning_rate": 1.2429692310954682e-05,
+      "loss": 0.8684,
+      "step": 7439
+    },
+    {
+      "epoch": 1.2637117688664081,
+      "grad_norm": 1.6484375,
+      "learning_rate": 1.2427940459619821e-05,
+      "loss": 0.8501,
+      "step": 7440
+    },
+    {
+      "epoch": 1.263883232955398,
+      "grad_norm": 1.6953125,
+      "learning_rate": 1.2426188529100677e-05,
+      "loss": 0.9016,
+      "step": 7441
+    },
+    {
+      "epoch": 1.2640546970443878,
+      "grad_norm": 1.7109375,
+      "learning_rate": 1.2424436519454393e-05,
+      "loss": 0.8824,
+      "step": 7442
+    },
+    {
+      "epoch": 1.2642261611333776,
+      "grad_norm": 1.734375,
+      "learning_rate": 1.24226844307381e-05,
+      "loss": 0.9148,
+      "step": 7443
+    },
+    {
+      "epoch": 1.2643976252223674,
+      "grad_norm": 1.71875,
+      "learning_rate": 1.2420932263008947e-05,
+      "loss": 0.8463,
+      "step": 7444
+    },
+    {
+      "epoch": 1.2645690893113573,
+      "grad_norm": 1.5859375,
+      "learning_rate": 1.2419180016324076e-05,
+      "loss": 0.8036,
+      "step": 7445
+    },
+    {
+      "epoch": 1.264740553400347,
+      "grad_norm": 1.703125,
+      "learning_rate": 1.241742769074064e-05,
+      "loss": 0.92,
+      "step": 7446
+    },
+    {
+      "epoch": 1.2649120174893371,
+      "grad_norm": 1.609375,
+      "learning_rate": 1.2415675286315778e-05,
+      "loss": 0.862,
+      "step": 7447
+    },
+    {
+      "epoch": 1.265083481578327,
+      "grad_norm": 1.5703125,
+      "learning_rate": 1.2413922803106649e-05,
+      "loss": 0.7859,
+      "step": 7448
+    },
+    {
+      "epoch": 1.2652549456673168,
+      "grad_norm": 1.6015625,
+      "learning_rate": 1.2412170241170406e-05,
+      "loss": 0.8329,
+      "step": 7449
+    },
+    {
+      "epoch": 1.2654264097563066,
+      "grad_norm": 1.6171875,
+      "learning_rate": 1.2410417600564212e-05,
+      "loss": 0.8083,
+      "step": 7450
+    },
+    {
+      "epoch": 1.2655978738452967,
+      "grad_norm": 1.734375,
+      "learning_rate": 1.2408664881345219e-05,
+      "loss": 0.9278,
+      "step": 7451
+    },
+    {
+      "epoch": 1.2657693379342865,
+      "grad_norm": 1.75,
+      "learning_rate": 1.2406912083570599e-05,
+      "loss": 0.8668,
+      "step": 7452
+    },
+    {
+      "epoch": 1.2659408020232763,
+      "grad_norm": 1.71875,
+      "learning_rate": 1.2405159207297508e-05,
+      "loss": 0.9055,
+      "step": 7453
+    },
+    {
+      "epoch": 1.2661122661122661,
+      "grad_norm": 1.6640625,
+      "learning_rate": 1.2403406252583121e-05,
+      "loss": 0.874,
+      "step": 7454
+    },
+    {
+      "epoch": 1.266283730201256,
+      "grad_norm": 1.6796875,
+      "learning_rate": 1.2401653219484604e-05,
+      "loss": 0.9812,
+      "step": 7455
+    },
+    {
+      "epoch": 1.2664551942902458,
+      "grad_norm": 1.65625,
+      "learning_rate": 1.2399900108059134e-05,
+      "loss": 0.86,
+      "step": 7456
+    },
+    {
+      "epoch": 1.2666266583792356,
+      "grad_norm": 1.5625,
+      "learning_rate": 1.2398146918363882e-05,
+      "loss": 0.7916,
+      "step": 7457
+    },
+    {
+      "epoch": 1.2667981224682254,
+      "grad_norm": 1.734375,
+      "learning_rate": 1.239639365045603e-05,
+      "loss": 0.9295,
+      "step": 7458
+    },
+    {
+      "epoch": 1.2669695865572155,
+      "grad_norm": 1.6875,
+      "learning_rate": 1.2394640304392753e-05,
+      "loss": 0.8695,
+      "step": 7459
+    },
+    {
+      "epoch": 1.2671410506462053,
+      "grad_norm": 1.6875,
+      "learning_rate": 1.2392886880231243e-05,
+      "loss": 0.8425,
+      "step": 7460
+    },
+    {
+      "epoch": 1.2673125147351951,
+      "grad_norm": 1.5703125,
+      "learning_rate": 1.239113337802868e-05,
+      "loss": 0.7969,
+      "step": 7461
+    },
+    {
+      "epoch": 1.267483978824185,
+      "grad_norm": 1.6796875,
+      "learning_rate": 1.2389379797842252e-05,
+      "loss": 0.8745,
+      "step": 7462
+    },
+    {
+      "epoch": 1.267655442913175,
+      "grad_norm": 1.7109375,
+      "learning_rate": 1.2387626139729155e-05,
+      "loss": 0.8732,
+      "step": 7463
+    },
+    {
+      "epoch": 1.2678269070021648,
+      "grad_norm": 1.7109375,
+      "learning_rate": 1.2385872403746575e-05,
+      "loss": 0.8747,
+      "step": 7464
+    },
+    {
+      "epoch": 1.2679983710911547,
+      "grad_norm": 1.6796875,
+      "learning_rate": 1.238411858995171e-05,
+      "loss": 0.8893,
+      "step": 7465
+    },
+    {
+      "epoch": 1.2681698351801445,
+      "grad_norm": 1.625,
+      "learning_rate": 1.2382364698401764e-05,
+      "loss": 0.7905,
+      "step": 7466
+    },
+    {
+      "epoch": 1.2683412992691343,
+      "grad_norm": 1.7734375,
+      "learning_rate": 1.238061072915393e-05,
+      "loss": 0.8982,
+      "step": 7467
+    },
+    {
+      "epoch": 1.2685127633581241,
+      "grad_norm": 1.671875,
+      "learning_rate": 1.2378856682265419e-05,
+      "loss": 0.7652,
+      "step": 7468
+    },
+    {
+      "epoch": 1.268684227447114,
+      "grad_norm": 1.7421875,
+      "learning_rate": 1.2377102557793433e-05,
+      "loss": 0.8618,
+      "step": 7469
+    },
+    {
+      "epoch": 1.2688556915361038,
+      "grad_norm": 1.671875,
+      "learning_rate": 1.2375348355795181e-05,
+      "loss": 0.8758,
+      "step": 7470
+    },
+    {
+      "epoch": 1.2690271556250938,
+      "grad_norm": 1.71875,
+      "learning_rate": 1.2373594076327873e-05,
+      "loss": 0.8996,
+      "step": 7471
+    },
+    {
+      "epoch": 1.2691986197140837,
+      "grad_norm": 1.6328125,
+      "learning_rate": 1.2371839719448725e-05,
+      "loss": 0.8519,
+      "step": 7472
+    },
+    {
+      "epoch": 1.2693700838030735,
+      "grad_norm": 1.6640625,
+      "learning_rate": 1.2370085285214953e-05,
+      "loss": 0.8879,
+      "step": 7473
+    },
+    {
+      "epoch": 1.2695415478920633,
+      "grad_norm": 1.6953125,
+      "learning_rate": 1.2368330773683774e-05,
+      "loss": 0.8703,
+      "step": 7474
+    },
+    {
+      "epoch": 1.2697130119810534,
+      "grad_norm": 1.6484375,
+      "learning_rate": 1.236657618491241e-05,
+      "loss": 0.8536,
+      "step": 7475
+    },
+    {
+      "epoch": 1.2698844760700432,
+      "grad_norm": 1.75,
+      "learning_rate": 1.2364821518958088e-05,
+      "loss": 0.8663,
+      "step": 7476
+    },
+    {
+      "epoch": 1.270055940159033,
+      "grad_norm": 1.546875,
+      "learning_rate": 1.2363066775878028e-05,
+      "loss": 0.8806,
+      "step": 7477
+    },
+    {
+      "epoch": 1.2702274042480228,
+      "grad_norm": 1.625,
+      "learning_rate": 1.236131195572946e-05,
+      "loss": 0.7836,
+      "step": 7478
+    },
+    {
+      "epoch": 1.2703988683370127,
+      "grad_norm": 1.6328125,
+      "learning_rate": 1.235955705856962e-05,
+      "loss": 0.7843,
+      "step": 7479
+    },
+    {
+      "epoch": 1.2705703324260025,
+      "grad_norm": 1.5703125,
+      "learning_rate": 1.2357802084455738e-05,
+      "loss": 0.8683,
+      "step": 7480
+    },
+    {
+      "epoch": 1.2707417965149923,
+      "grad_norm": 1.65625,
+      "learning_rate": 1.2356047033445055e-05,
+      "loss": 0.9125,
+      "step": 7481
+    },
+    {
+      "epoch": 1.2709132606039821,
+      "grad_norm": 1.6875,
+      "learning_rate": 1.2354291905594801e-05,
+      "loss": 0.8758,
+      "step": 7482
+    },
+    {
+      "epoch": 1.2710847246929722,
+      "grad_norm": 1.6171875,
+      "learning_rate": 1.2352536700962228e-05,
+      "loss": 0.8316,
+      "step": 7483
+    },
+    {
+      "epoch": 1.271256188781962,
+      "grad_norm": 1.609375,
+      "learning_rate": 1.2350781419604569e-05,
+      "loss": 0.8318,
+      "step": 7484
+    },
+    {
+      "epoch": 1.2714276528709518,
+      "grad_norm": 1.75,
+      "learning_rate": 1.234902606157908e-05,
+      "loss": 0.9431,
+      "step": 7485
+    },
+    {
+      "epoch": 1.2715991169599417,
+      "grad_norm": 1.7890625,
+      "learning_rate": 1.2347270626943002e-05,
+      "loss": 0.9024,
+      "step": 7486
+    },
+    {
+      "epoch": 1.2717705810489317,
+      "grad_norm": 1.7890625,
+      "learning_rate": 1.234551511575359e-05,
+      "loss": 0.945,
+      "step": 7487
+    },
+    {
+      "epoch": 1.2719420451379215,
+      "grad_norm": 1.6328125,
+      "learning_rate": 1.23437595280681e-05,
+      "loss": 0.8238,
+      "step": 7488
+    },
+    {
+      "epoch": 1.2721135092269114,
+      "grad_norm": 1.71875,
+      "learning_rate": 1.2342003863943785e-05,
+      "loss": 0.8964,
+      "step": 7489
+    },
+    {
+      "epoch": 1.2722849733159012,
+      "grad_norm": 1.640625,
+      "learning_rate": 1.2340248123437904e-05,
+      "loss": 0.8467,
+      "step": 7490
+    },
+    {
+      "epoch": 1.272456437404891,
+      "grad_norm": 1.7265625,
+      "learning_rate": 1.2338492306607721e-05,
+      "loss": 0.8657,
+      "step": 7491
+    },
+    {
+      "epoch": 1.2726279014938808,
+      "grad_norm": 1.75,
+      "learning_rate": 1.2336736413510497e-05,
+      "loss": 0.8956,
+      "step": 7492
+    },
+    {
+      "epoch": 1.2727993655828707,
+      "grad_norm": 1.609375,
+      "learning_rate": 1.23349804442035e-05,
+      "loss": 0.8019,
+      "step": 7493
+    },
+    {
+      "epoch": 1.2729708296718605,
+      "grad_norm": 1.6796875,
+      "learning_rate": 1.2333224398743997e-05,
+      "loss": 0.8398,
+      "step": 7494
+    },
+    {
+      "epoch": 1.2731422937608505,
+      "grad_norm": 1.65625,
+      "learning_rate": 1.233146827718926e-05,
+      "loss": 0.8467,
+      "step": 7495
+    },
+    {
+      "epoch": 1.2733137578498404,
+      "grad_norm": 1.6953125,
+      "learning_rate": 1.2329712079596563e-05,
+      "loss": 0.9317,
+      "step": 7496
+    },
+    {
+      "epoch": 1.2734852219388302,
+      "grad_norm": 1.71875,
+      "learning_rate": 1.2327955806023181e-05,
+      "loss": 0.8621,
+      "step": 7497
+    },
+    {
+      "epoch": 1.27365668602782,
+      "grad_norm": 1.625,
+      "learning_rate": 1.2326199456526397e-05,
+      "loss": 0.8108,
+      "step": 7498
+    },
+    {
+      "epoch": 1.2738281501168098,
+      "grad_norm": 1.6875,
+      "learning_rate": 1.2324443031163487e-05,
+      "loss": 0.8021,
+      "step": 7499
+    },
+    {
+      "epoch": 1.2739996142057999,
+      "grad_norm": 1.734375,
+      "learning_rate": 1.2322686529991739e-05,
+      "loss": 0.8839,
+      "step": 7500
+    },
+    {
+      "epoch": 1.2741710782947897,
+      "grad_norm": 1.7421875,
+      "learning_rate": 1.2320929953068435e-05,
+      "loss": 0.883,
+      "step": 7501
+    },
+    {
+      "epoch": 1.2743425423837795,
+      "grad_norm": 1.625,
+      "learning_rate": 1.2319173300450864e-05,
+      "loss": 0.8737,
+      "step": 7502
+    },
+    {
+      "epoch": 1.2745140064727694,
+      "grad_norm": 1.53125,
+      "learning_rate": 1.231741657219632e-05,
+      "loss": 0.797,
+      "step": 7503
+    },
+    {
+      "epoch": 1.2746854705617592,
+      "grad_norm": 1.6015625,
+      "learning_rate": 1.2315659768362097e-05,
+      "loss": 0.812,
+      "step": 7504
+    },
+    {
+      "epoch": 1.274856934650749,
+      "grad_norm": 1.65625,
+      "learning_rate": 1.2313902889005486e-05,
+      "loss": 0.9194,
+      "step": 7505
+    },
+    {
+      "epoch": 1.2750283987397388,
+      "grad_norm": 1.6640625,
+      "learning_rate": 1.2312145934183788e-05,
+      "loss": 0.9029,
+      "step": 7506
+    },
+    {
+      "epoch": 1.2751998628287289,
+      "grad_norm": 1.6640625,
+      "learning_rate": 1.2310388903954304e-05,
+      "loss": 0.8317,
+      "step": 7507
+    },
+    {
+      "epoch": 1.2753713269177187,
+      "grad_norm": 1.578125,
+      "learning_rate": 1.2308631798374339e-05,
+      "loss": 0.8674,
+      "step": 7508
+    },
+    {
+      "epoch": 1.2755427910067085,
+      "grad_norm": 1.625,
+      "learning_rate": 1.23068746175012e-05,
+      "loss": 0.7913,
+      "step": 7509
+    },
+    {
+      "epoch": 1.2757142550956984,
+      "grad_norm": 1.671875,
+      "learning_rate": 1.2305117361392192e-05,
+      "loss": 0.8466,
+      "step": 7510
+    },
+    {
+      "epoch": 1.2758857191846882,
+      "grad_norm": 1.625,
+      "learning_rate": 1.2303360030104625e-05,
+      "loss": 0.8082,
+      "step": 7511
+    },
+    {
+      "epoch": 1.2760571832736782,
+      "grad_norm": 1.8125,
+      "learning_rate": 1.2301602623695814e-05,
+      "loss": 0.9373,
+      "step": 7512
+    },
+    {
+      "epoch": 1.276228647362668,
+      "grad_norm": 1.6484375,
+      "learning_rate": 1.2299845142223075e-05,
+      "loss": 0.8069,
+      "step": 7513
+    },
+    {
+      "epoch": 1.2764001114516579,
+      "grad_norm": 1.65625,
+      "learning_rate": 1.2298087585743725e-05,
+      "loss": 0.8812,
+      "step": 7514
+    },
+    {
+      "epoch": 1.2765715755406477,
+      "grad_norm": 1.71875,
+      "learning_rate": 1.2296329954315085e-05,
+      "loss": 0.8012,
+      "step": 7515
+    },
+    {
+      "epoch": 1.2767430396296375,
+      "grad_norm": 1.6875,
+      "learning_rate": 1.2294572247994478e-05,
+      "loss": 0.8418,
+      "step": 7516
+    },
+    {
+      "epoch": 1.2769145037186274,
+      "grad_norm": 1.7265625,
+      "learning_rate": 1.229281446683923e-05,
+      "loss": 0.9418,
+      "step": 7517
+    },
+    {
+      "epoch": 1.2770859678076172,
+      "grad_norm": 1.71875,
+      "learning_rate": 1.2291056610906666e-05,
+      "loss": 0.9058,
+      "step": 7518
+    },
+    {
+      "epoch": 1.2772574318966072,
+      "grad_norm": 1.625,
+      "learning_rate": 1.228929868025412e-05,
+      "loss": 0.8718,
+      "step": 7519
+    },
+    {
+      "epoch": 1.277428895985597,
+      "grad_norm": 1.7421875,
+      "learning_rate": 1.2287540674938925e-05,
+      "loss": 0.8379,
+      "step": 7520
+    },
+    {
+      "epoch": 1.2776003600745869,
+      "grad_norm": 1.578125,
+      "learning_rate": 1.2285782595018417e-05,
+      "loss": 0.8615,
+      "step": 7521
+    },
+    {
+      "epoch": 1.2777718241635767,
+      "grad_norm": 1.7109375,
+      "learning_rate": 1.2284024440549924e-05,
+      "loss": 0.9499,
+      "step": 7522
+    },
+    {
+      "epoch": 1.2779432882525665,
+      "grad_norm": 1.59375,
+      "learning_rate": 1.2282266211590798e-05,
+      "loss": 0.8202,
+      "step": 7523
+    },
+    {
+      "epoch": 1.2781147523415566,
+      "grad_norm": 1.703125,
+      "learning_rate": 1.2280507908198375e-05,
+      "loss": 0.9373,
+      "step": 7524
+    },
+    {
+      "epoch": 1.2782862164305464,
+      "grad_norm": 1.6953125,
+      "learning_rate": 1.227874953043e-05,
+      "loss": 0.8337,
+      "step": 7525
+    },
+    {
+      "epoch": 1.2784576805195362,
+      "grad_norm": 1.65625,
+      "learning_rate": 1.2276991078343024e-05,
+      "loss": 0.8789,
+      "step": 7526
+    },
+    {
+      "epoch": 1.278629144608526,
+      "grad_norm": 1.625,
+      "learning_rate": 1.2275232551994795e-05,
+      "loss": 0.8343,
+      "step": 7527
+    },
+    {
+      "epoch": 1.2788006086975159,
+      "grad_norm": 1.6640625,
+      "learning_rate": 1.2273473951442663e-05,
+      "loss": 0.8961,
+      "step": 7528
+    },
+    {
+      "epoch": 1.2789720727865057,
+      "grad_norm": 1.7265625,
+      "learning_rate": 1.2271715276743984e-05,
+      "loss": 0.8509,
+      "step": 7529
+    },
+    {
+      "epoch": 1.2791435368754955,
+      "grad_norm": 1.625,
+      "learning_rate": 1.2269956527956118e-05,
+      "loss": 0.8277,
+      "step": 7530
+    },
+    {
+      "epoch": 1.2793150009644856,
+      "grad_norm": 1.7421875,
+      "learning_rate": 1.226819770513642e-05,
+      "loss": 0.8797,
+      "step": 7531
+    },
+    {
+      "epoch": 1.2794864650534754,
+      "grad_norm": 1.75,
+      "learning_rate": 1.2266438808342257e-05,
+      "loss": 0.8979,
+      "step": 7532
+    },
+    {
+      "epoch": 1.2796579291424652,
+      "grad_norm": 1.8359375,
+      "learning_rate": 1.2264679837630984e-05,
+      "loss": 0.8799,
+      "step": 7533
+    },
+    {
+      "epoch": 1.279829393231455,
+      "grad_norm": 1.640625,
+      "learning_rate": 1.2262920793059976e-05,
+      "loss": 0.8473,
+      "step": 7534
+    },
+    {
+      "epoch": 1.2800008573204449,
+      "grad_norm": 1.6640625,
+      "learning_rate": 1.22611616746866e-05,
+      "loss": 0.7805,
+      "step": 7535
+    },
+    {
+      "epoch": 1.280172321409435,
+      "grad_norm": 1.6328125,
+      "learning_rate": 1.2259402482568226e-05,
+      "loss": 0.8182,
+      "step": 7536
+    },
+    {
+      "epoch": 1.2803437854984248,
+      "grad_norm": 1.7421875,
+      "learning_rate": 1.225764321676223e-05,
+      "loss": 0.8051,
+      "step": 7537
+    },
+    {
+      "epoch": 1.2805152495874146,
+      "grad_norm": 1.765625,
+      "learning_rate": 1.2255883877325989e-05,
+      "loss": 0.7994,
+      "step": 7538
+    },
+    {
+      "epoch": 1.2806867136764044,
+      "grad_norm": 1.8125,
+      "learning_rate": 1.2254124464316876e-05,
+      "loss": 0.9064,
+      "step": 7539
+    },
+    {
+      "epoch": 1.2808581777653942,
+      "grad_norm": 1.7578125,
+      "learning_rate": 1.225236497779228e-05,
+      "loss": 0.9003,
+      "step": 7540
+    },
+    {
+      "epoch": 1.281029641854384,
+      "grad_norm": 1.765625,
+      "learning_rate": 1.2250605417809579e-05,
+      "loss": 0.9243,
+      "step": 7541
+    },
+    {
+      "epoch": 1.2812011059433739,
+      "grad_norm": 1.65625,
+      "learning_rate": 1.2248845784426157e-05,
+      "loss": 0.7572,
+      "step": 7542
+    },
+    {
+      "epoch": 1.2813725700323637,
+      "grad_norm": 1.640625,
+      "learning_rate": 1.2247086077699408e-05,
+      "loss": 0.9036,
+      "step": 7543
+    },
+    {
+      "epoch": 1.2815440341213538,
+      "grad_norm": 1.703125,
+      "learning_rate": 1.2245326297686721e-05,
+      "loss": 0.8296,
+      "step": 7544
+    },
+    {
+      "epoch": 1.2817154982103436,
+      "grad_norm": 1.71875,
+      "learning_rate": 1.2243566444445487e-05,
+      "loss": 0.8512,
+      "step": 7545
+    },
+    {
+      "epoch": 1.2818869622993334,
+      "grad_norm": 1.671875,
+      "learning_rate": 1.2241806518033104e-05,
+      "loss": 0.8552,
+      "step": 7546
+    },
+    {
+      "epoch": 1.2820584263883232,
+      "grad_norm": 1.734375,
+      "learning_rate": 1.2240046518506967e-05,
+      "loss": 0.8765,
+      "step": 7547
+    },
+    {
+      "epoch": 1.2822298904773133,
+      "grad_norm": 1.515625,
+      "learning_rate": 1.2238286445924479e-05,
+      "loss": 0.808,
+      "step": 7548
+    },
+    {
+      "epoch": 1.282401354566303,
+      "grad_norm": 1.71875,
+      "learning_rate": 1.2236526300343041e-05,
+      "loss": 0.8427,
+      "step": 7549
+    },
+    {
+      "epoch": 1.282572818655293,
+      "grad_norm": 1.6640625,
+      "learning_rate": 1.2234766081820063e-05,
+      "loss": 0.8815,
+      "step": 7550
+    },
+    {
+      "epoch": 1.2827442827442828,
+      "grad_norm": 1.6328125,
+      "learning_rate": 1.2233005790412942e-05,
+      "loss": 0.846,
+      "step": 7551
+    },
+    {
+      "epoch": 1.2829157468332726,
+      "grad_norm": 1.703125,
+      "learning_rate": 1.2231245426179095e-05,
+      "loss": 0.8415,
+      "step": 7552
+    },
+    {
+      "epoch": 1.2830872109222624,
+      "grad_norm": 1.6484375,
+      "learning_rate": 1.222948498917593e-05,
+      "loss": 0.8797,
+      "step": 7553
+    },
+    {
+      "epoch": 1.2832586750112522,
+      "grad_norm": 1.7578125,
+      "learning_rate": 1.2227724479460866e-05,
+      "loss": 0.8813,
+      "step": 7554
+    },
+    {
+      "epoch": 1.283430139100242,
+      "grad_norm": 1.6484375,
+      "learning_rate": 1.2225963897091317e-05,
+      "loss": 0.8223,
+      "step": 7555
+    },
+    {
+      "epoch": 1.283601603189232,
+      "grad_norm": 1.703125,
+      "learning_rate": 1.2224203242124707e-05,
+      "loss": 0.8077,
+      "step": 7556
+    },
+    {
+      "epoch": 1.283773067278222,
+      "grad_norm": 1.703125,
+      "learning_rate": 1.2222442514618452e-05,
+      "loss": 0.9232,
+      "step": 7557
+    },
+    {
+      "epoch": 1.2839445313672118,
+      "grad_norm": 1.6796875,
+      "learning_rate": 1.2220681714629976e-05,
+      "loss": 0.8709,
+      "step": 7558
+    },
+    {
+      "epoch": 1.2841159954562016,
+      "grad_norm": 1.71875,
+      "learning_rate": 1.2218920842216708e-05,
+      "loss": 0.877,
+      "step": 7559
+    },
+    {
+      "epoch": 1.2842874595451916,
+      "grad_norm": 1.609375,
+      "learning_rate": 1.2217159897436075e-05,
+      "loss": 0.9244,
+      "step": 7560
+    },
+    {
+      "epoch": 1.2844589236341815,
+      "grad_norm": 1.6796875,
+      "learning_rate": 1.2215398880345507e-05,
+      "loss": 0.9621,
+      "step": 7561
+    },
+    {
+      "epoch": 1.2846303877231713,
+      "grad_norm": 1.765625,
+      "learning_rate": 1.2213637791002443e-05,
+      "loss": 0.9206,
+      "step": 7562
+    },
+    {
+      "epoch": 1.284801851812161,
+      "grad_norm": 1.640625,
+      "learning_rate": 1.2211876629464314e-05,
+      "loss": 0.86,
+      "step": 7563
+    },
+    {
+      "epoch": 1.284973315901151,
+      "grad_norm": 1.6484375,
+      "learning_rate": 1.2210115395788558e-05,
+      "loss": 0.8369,
+      "step": 7564
+    },
+    {
+      "epoch": 1.2851447799901408,
+      "grad_norm": 1.625,
+      "learning_rate": 1.2208354090032615e-05,
+      "loss": 0.8893,
+      "step": 7565
+    },
+    {
+      "epoch": 1.2853162440791306,
+      "grad_norm": 1.75,
+      "learning_rate": 1.2206592712253931e-05,
+      "loss": 0.8371,
+      "step": 7566
+    },
+    {
+      "epoch": 1.2854877081681204,
+      "grad_norm": 1.71875,
+      "learning_rate": 1.220483126250995e-05,
+      "loss": 0.8877,
+      "step": 7567
+    },
+    {
+      "epoch": 1.2856591722571105,
+      "grad_norm": 1.7109375,
+      "learning_rate": 1.220306974085812e-05,
+      "loss": 0.8343,
+      "step": 7568
+    },
+    {
+      "epoch": 1.2858306363461003,
+      "grad_norm": 1.7421875,
+      "learning_rate": 1.2201308147355886e-05,
+      "loss": 0.927,
+      "step": 7569
+    },
+    {
+      "epoch": 1.2860021004350901,
+      "grad_norm": 1.734375,
+      "learning_rate": 1.2199546482060707e-05,
+      "loss": 0.8311,
+      "step": 7570
+    },
+    {
+      "epoch": 1.28617356452408,
+      "grad_norm": 1.6328125,
+      "learning_rate": 1.2197784745030033e-05,
+      "loss": 0.8742,
+      "step": 7571
+    },
+    {
+      "epoch": 1.28634502861307,
+      "grad_norm": 1.703125,
+      "learning_rate": 1.219602293632132e-05,
+      "loss": 0.8421,
+      "step": 7572
+    },
+    {
+      "epoch": 1.2865164927020598,
+      "grad_norm": 1.75,
+      "learning_rate": 1.2194261055992033e-05,
+      "loss": 0.9014,
+      "step": 7573
+    },
+    {
+      "epoch": 1.2866879567910496,
+      "grad_norm": 1.6953125,
+      "learning_rate": 1.219249910409963e-05,
+      "loss": 0.8843,
+      "step": 7574
+    },
+    {
+      "epoch": 1.2868594208800395,
+      "grad_norm": 1.625,
+      "learning_rate": 1.2190737080701574e-05,
+      "loss": 0.8677,
+      "step": 7575
+    },
+    {
+      "epoch": 1.2870308849690293,
+      "grad_norm": 1.6875,
+      "learning_rate": 1.2188974985855334e-05,
+      "loss": 0.8641,
+      "step": 7576
+    },
+    {
+      "epoch": 1.2872023490580191,
+      "grad_norm": 1.7421875,
+      "learning_rate": 1.2187212819618376e-05,
+      "loss": 0.8851,
+      "step": 7577
+    },
+    {
+      "epoch": 1.287373813147009,
+      "grad_norm": 1.8046875,
+      "learning_rate": 1.2185450582048168e-05,
+      "loss": 0.826,
+      "step": 7578
+    },
+    {
+      "epoch": 1.2875452772359988,
+      "grad_norm": 1.7265625,
+      "learning_rate": 1.2183688273202192e-05,
+      "loss": 0.9094,
+      "step": 7579
+    },
+    {
+      "epoch": 1.2877167413249888,
+      "grad_norm": 1.6796875,
+      "learning_rate": 1.2181925893137914e-05,
+      "loss": 0.853,
+      "step": 7580
+    },
+    {
+      "epoch": 1.2878882054139786,
+      "grad_norm": 1.75,
+      "learning_rate": 1.2180163441912817e-05,
+      "loss": 0.9173,
+      "step": 7581
+    },
+    {
+      "epoch": 1.2880596695029685,
+      "grad_norm": 1.75,
+      "learning_rate": 1.2178400919584384e-05,
+      "loss": 0.8648,
+      "step": 7582
+    },
+    {
+      "epoch": 1.2882311335919583,
+      "grad_norm": 1.6875,
+      "learning_rate": 1.2176638326210091e-05,
+      "loss": 0.8913,
+      "step": 7583
+    },
+    {
+      "epoch": 1.2884025976809481,
+      "grad_norm": 1.734375,
+      "learning_rate": 1.217487566184743e-05,
+      "loss": 0.8438,
+      "step": 7584
+    },
+    {
+      "epoch": 1.2885740617699382,
+      "grad_norm": 1.71875,
+      "learning_rate": 1.217311292655388e-05,
+      "loss": 0.8476,
+      "step": 7585
+    },
+    {
+      "epoch": 1.288745525858928,
+      "grad_norm": 1.75,
+      "learning_rate": 1.2171350120386933e-05,
+      "loss": 0.8477,
+      "step": 7586
+    },
+    {
+      "epoch": 1.2889169899479178,
+      "grad_norm": 1.765625,
+      "learning_rate": 1.2169587243404085e-05,
+      "loss": 0.8612,
+      "step": 7587
+    },
+    {
+      "epoch": 1.2890884540369076,
+      "grad_norm": 1.640625,
+      "learning_rate": 1.2167824295662825e-05,
+      "loss": 0.8484,
+      "step": 7588
+    },
+    {
+      "epoch": 1.2892599181258975,
+      "grad_norm": 1.7578125,
+      "learning_rate": 1.2166061277220653e-05,
+      "loss": 0.9121,
+      "step": 7589
+    },
+    {
+      "epoch": 1.2894313822148873,
+      "grad_norm": 1.671875,
+      "learning_rate": 1.2164298188135065e-05,
+      "loss": 0.8612,
+      "step": 7590
+    },
+    {
+      "epoch": 1.2896028463038771,
+      "grad_norm": 1.765625,
+      "learning_rate": 1.2162535028463563e-05,
+      "loss": 0.8444,
+      "step": 7591
+    },
+    {
+      "epoch": 1.2897743103928672,
+      "grad_norm": 1.6328125,
+      "learning_rate": 1.2160771798263651e-05,
+      "loss": 0.7786,
+      "step": 7592
+    },
+    {
+      "epoch": 1.289945774481857,
+      "grad_norm": 1.71875,
+      "learning_rate": 1.2159008497592834e-05,
+      "loss": 0.8248,
+      "step": 7593
+    },
+    {
+      "epoch": 1.2901172385708468,
+      "grad_norm": 1.6640625,
+      "learning_rate": 1.2157245126508619e-05,
+      "loss": 0.8333,
+      "step": 7594
+    },
+    {
+      "epoch": 1.2902887026598366,
+      "grad_norm": 1.7109375,
+      "learning_rate": 1.2155481685068517e-05,
+      "loss": 0.8895,
+      "step": 7595
+    },
+    {
+      "epoch": 1.2904601667488265,
+      "grad_norm": 1.7109375,
+      "learning_rate": 1.2153718173330045e-05,
+      "loss": 0.902,
+      "step": 7596
+    },
+    {
+      "epoch": 1.2906316308378165,
+      "grad_norm": 1.6796875,
+      "learning_rate": 1.2151954591350708e-05,
+      "loss": 0.8519,
+      "step": 7597
+    },
+    {
+      "epoch": 1.2908030949268063,
+      "grad_norm": 1.703125,
+      "learning_rate": 1.215019093918803e-05,
+      "loss": 0.8967,
+      "step": 7598
+    },
+    {
+      "epoch": 1.2909745590157962,
+      "grad_norm": 1.6796875,
+      "learning_rate": 1.2148427216899528e-05,
+      "loss": 0.8826,
+      "step": 7599
+    },
+    {
+      "epoch": 1.291146023104786,
+      "grad_norm": 1.5625,
+      "learning_rate": 1.2146663424542722e-05,
+      "loss": 0.8378,
+      "step": 7600
+    },
+    {
+      "epoch": 1.2913174871937758,
+      "grad_norm": 1.609375,
+      "learning_rate": 1.214489956217514e-05,
+      "loss": 0.8425,
+      "step": 7601
+    },
+    {
+      "epoch": 1.2914889512827656,
+      "grad_norm": 1.6171875,
+      "learning_rate": 1.2143135629854307e-05,
+      "loss": 0.8659,
+      "step": 7602
+    },
+    {
+      "epoch": 1.2916604153717555,
+      "grad_norm": 1.6640625,
+      "learning_rate": 1.2141371627637752e-05,
+      "loss": 0.8331,
+      "step": 7603
+    },
+    {
+      "epoch": 1.2918318794607455,
+      "grad_norm": 1.703125,
+      "learning_rate": 1.2139607555583004e-05,
+      "loss": 0.8546,
+      "step": 7604
+    },
+    {
+      "epoch": 1.2920033435497353,
+      "grad_norm": 1.71875,
+      "learning_rate": 1.2137843413747594e-05,
+      "loss": 0.9113,
+      "step": 7605
+    },
+    {
+      "epoch": 1.2921748076387252,
+      "grad_norm": 1.625,
+      "learning_rate": 1.2136079202189061e-05,
+      "loss": 0.8261,
+      "step": 7606
+    },
+    {
+      "epoch": 1.292346271727715,
+      "grad_norm": 1.6953125,
+      "learning_rate": 1.2134314920964947e-05,
+      "loss": 0.901,
+      "step": 7607
+    },
+    {
+      "epoch": 1.2925177358167048,
+      "grad_norm": 1.8125,
+      "learning_rate": 1.2132550570132779e-05,
+      "loss": 0.8526,
+      "step": 7608
+    },
+    {
+      "epoch": 1.2926891999056949,
+      "grad_norm": 1.6953125,
+      "learning_rate": 1.2130786149750108e-05,
+      "loss": 0.8381,
+      "step": 7609
+    },
+    {
+      "epoch": 1.2928606639946847,
+      "grad_norm": 1.7109375,
+      "learning_rate": 1.2129021659874478e-05,
+      "loss": 0.8988,
+      "step": 7610
+    },
+    {
+      "epoch": 1.2930321280836745,
+      "grad_norm": 1.703125,
+      "learning_rate": 1.2127257100563435e-05,
+      "loss": 0.866,
+      "step": 7611
+    },
+    {
+      "epoch": 1.2932035921726643,
+      "grad_norm": 1.6953125,
+      "learning_rate": 1.2125492471874526e-05,
+      "loss": 0.9133,
+      "step": 7612
+    },
+    {
+      "epoch": 1.2933750562616542,
+      "grad_norm": 1.8125,
+      "learning_rate": 1.2123727773865305e-05,
+      "loss": 0.8806,
+      "step": 7613
+    },
+    {
+      "epoch": 1.293546520350644,
+      "grad_norm": 1.734375,
+      "learning_rate": 1.2121963006593325e-05,
+      "loss": 0.9313,
+      "step": 7614
+    },
+    {
+      "epoch": 1.2937179844396338,
+      "grad_norm": 1.6953125,
+      "learning_rate": 1.212019817011614e-05,
+      "loss": 0.8593,
+      "step": 7615
+    },
+    {
+      "epoch": 1.2938894485286239,
+      "grad_norm": 1.6796875,
+      "learning_rate": 1.2118433264491307e-05,
+      "loss": 0.8664,
+      "step": 7616
+    },
+    {
+      "epoch": 1.2940609126176137,
+      "grad_norm": 1.7578125,
+      "learning_rate": 1.211666828977639e-05,
+      "loss": 0.8287,
+      "step": 7617
+    },
+    {
+      "epoch": 1.2942323767066035,
+      "grad_norm": 1.7265625,
+      "learning_rate": 1.2114903246028949e-05,
+      "loss": 0.9472,
+      "step": 7618
+    },
+    {
+      "epoch": 1.2944038407955933,
+      "grad_norm": 1.6953125,
+      "learning_rate": 1.2113138133306545e-05,
+      "loss": 0.807,
+      "step": 7619
+    },
+    {
+      "epoch": 1.2945753048845832,
+      "grad_norm": 1.7890625,
+      "learning_rate": 1.2111372951666755e-05,
+      "loss": 0.7895,
+      "step": 7620
+    },
+    {
+      "epoch": 1.2947467689735732,
+      "grad_norm": 1.6953125,
+      "learning_rate": 1.2109607701167139e-05,
+      "loss": 0.822,
+      "step": 7621
+    },
+    {
+      "epoch": 1.294918233062563,
+      "grad_norm": 1.75,
+      "learning_rate": 1.2107842381865273e-05,
+      "loss": 0.9415,
+      "step": 7622
+    },
+    {
+      "epoch": 1.2950896971515529,
+      "grad_norm": 1.671875,
+      "learning_rate": 1.210607699381873e-05,
+      "loss": 0.8165,
+      "step": 7623
+    },
+    {
+      "epoch": 1.2952611612405427,
+      "grad_norm": 1.671875,
+      "learning_rate": 1.2104311537085085e-05,
+      "loss": 0.793,
+      "step": 7624
+    },
+    {
+      "epoch": 1.2954326253295325,
+      "grad_norm": 1.6796875,
+      "learning_rate": 1.2102546011721918e-05,
+      "loss": 0.8712,
+      "step": 7625
+    },
+    {
+      "epoch": 1.2956040894185223,
+      "grad_norm": 1.6171875,
+      "learning_rate": 1.2100780417786805e-05,
+      "loss": 0.7509,
+      "step": 7626
+    },
+    {
+      "epoch": 1.2957755535075122,
+      "grad_norm": 1.7421875,
+      "learning_rate": 1.2099014755337333e-05,
+      "loss": 0.8279,
+      "step": 7627
+    },
+    {
+      "epoch": 1.2959470175965022,
+      "grad_norm": 1.7109375,
+      "learning_rate": 1.2097249024431086e-05,
+      "loss": 0.8896,
+      "step": 7628
+    },
+    {
+      "epoch": 1.296118481685492,
+      "grad_norm": 1.734375,
+      "learning_rate": 1.209548322512565e-05,
+      "loss": 0.8632,
+      "step": 7629
+    },
+    {
+      "epoch": 1.2962899457744819,
+      "grad_norm": 1.671875,
+      "learning_rate": 1.2093717357478618e-05,
+      "loss": 0.9059,
+      "step": 7630
+    },
+    {
+      "epoch": 1.2964614098634717,
+      "grad_norm": 1.75,
+      "learning_rate": 1.2091951421547579e-05,
+      "loss": 0.8138,
+      "step": 7631
+    },
+    {
+      "epoch": 1.2966328739524615,
+      "grad_norm": 1.65625,
+      "learning_rate": 1.2090185417390124e-05,
+      "loss": 0.8017,
+      "step": 7632
+    },
+    {
+      "epoch": 1.2968043380414516,
+      "grad_norm": 1.7109375,
+      "learning_rate": 1.2088419345063851e-05,
+      "loss": 0.8845,
+      "step": 7633
+    },
+    {
+      "epoch": 1.2969758021304414,
+      "grad_norm": 1.640625,
+      "learning_rate": 1.208665320462636e-05,
+      "loss": 0.8646,
+      "step": 7634
+    },
+    {
+      "epoch": 1.2971472662194312,
+      "grad_norm": 1.6875,
+      "learning_rate": 1.208488699613525e-05,
+      "loss": 0.8522,
+      "step": 7635
+    },
+    {
+      "epoch": 1.297318730308421,
+      "grad_norm": 1.71875,
+      "learning_rate": 1.2083120719648124e-05,
+      "loss": 0.8765,
+      "step": 7636
+    },
+    {
+      "epoch": 1.2974901943974109,
+      "grad_norm": 1.71875,
+      "learning_rate": 1.2081354375222589e-05,
+      "loss": 0.9103,
+      "step": 7637
+    },
+    {
+      "epoch": 1.2976616584864007,
+      "grad_norm": 1.6953125,
+      "learning_rate": 1.2079587962916248e-05,
+      "loss": 0.8577,
+      "step": 7638
+    },
+    {
+      "epoch": 1.2978331225753905,
+      "grad_norm": 1.75,
+      "learning_rate": 1.2077821482786714e-05,
+      "loss": 0.9572,
+      "step": 7639
+    },
+    {
+      "epoch": 1.2980045866643803,
+      "grad_norm": 1.546875,
+      "learning_rate": 1.2076054934891594e-05,
+      "loss": 0.8155,
+      "step": 7640
+    },
+    {
+      "epoch": 1.2981760507533704,
+      "grad_norm": 1.7265625,
+      "learning_rate": 1.2074288319288508e-05,
+      "loss": 0.8119,
+      "step": 7641
+    },
+    {
+      "epoch": 1.2983475148423602,
+      "grad_norm": 1.6875,
+      "learning_rate": 1.207252163603507e-05,
+      "loss": 0.8288,
+      "step": 7642
+    },
+    {
+      "epoch": 1.29851897893135,
+      "grad_norm": 1.7265625,
+      "learning_rate": 1.2070754885188895e-05,
+      "loss": 0.8586,
+      "step": 7643
+    },
+    {
+      "epoch": 1.2986904430203399,
+      "grad_norm": 1.8125,
+      "learning_rate": 1.2068988066807606e-05,
+      "loss": 0.9835,
+      "step": 7644
+    },
+    {
+      "epoch": 1.29886190710933,
+      "grad_norm": 1.734375,
+      "learning_rate": 1.2067221180948826e-05,
+      "loss": 0.8761,
+      "step": 7645
+    },
+    {
+      "epoch": 1.2990333711983197,
+      "grad_norm": 1.7421875,
+      "learning_rate": 1.2065454227670173e-05,
+      "loss": 0.8515,
+      "step": 7646
+    },
+    {
+      "epoch": 1.2992048352873096,
+      "grad_norm": 1.7265625,
+      "learning_rate": 1.2063687207029285e-05,
+      "loss": 0.9168,
+      "step": 7647
+    },
+    {
+      "epoch": 1.2993762993762994,
+      "grad_norm": 1.6015625,
+      "learning_rate": 1.2061920119083788e-05,
+      "loss": 0.7479,
+      "step": 7648
+    },
+    {
+      "epoch": 1.2995477634652892,
+      "grad_norm": 1.6875,
+      "learning_rate": 1.2060152963891307e-05,
+      "loss": 0.9024,
+      "step": 7649
+    },
+    {
+      "epoch": 1.299719227554279,
+      "grad_norm": 1.734375,
+      "learning_rate": 1.2058385741509481e-05,
+      "loss": 0.7964,
+      "step": 7650
+    },
+    {
+      "epoch": 1.2998906916432689,
+      "grad_norm": 1.5703125,
+      "learning_rate": 1.2056618451995947e-05,
+      "loss": 0.8361,
+      "step": 7651
+    },
+    {
+      "epoch": 1.3000621557322587,
+      "grad_norm": 1.7109375,
+      "learning_rate": 1.2054851095408339e-05,
+      "loss": 0.8796,
+      "step": 7652
+    },
+    {
+      "epoch": 1.3002336198212487,
+      "grad_norm": 1.640625,
+      "learning_rate": 1.20530836718043e-05,
+      "loss": 0.8677,
+      "step": 7653
+    },
+    {
+      "epoch": 1.3004050839102386,
+      "grad_norm": 1.6953125,
+      "learning_rate": 1.2051316181241472e-05,
+      "loss": 0.878,
+      "step": 7654
+    },
+    {
+      "epoch": 1.3005765479992284,
+      "grad_norm": 1.7734375,
+      "learning_rate": 1.2049548623777494e-05,
+      "loss": 0.9416,
+      "step": 7655
+    },
+    {
+      "epoch": 1.3007480120882182,
+      "grad_norm": 1.65625,
+      "learning_rate": 1.2047780999470023e-05,
+      "loss": 0.8483,
+      "step": 7656
+    },
+    {
+      "epoch": 1.3009194761772083,
+      "grad_norm": 1.71875,
+      "learning_rate": 1.2046013308376698e-05,
+      "loss": 0.9298,
+      "step": 7657
+    },
+    {
+      "epoch": 1.301090940266198,
+      "grad_norm": 1.625,
+      "learning_rate": 1.2044245550555179e-05,
+      "loss": 0.8033,
+      "step": 7658
+    },
+    {
+      "epoch": 1.301262404355188,
+      "grad_norm": 1.6328125,
+      "learning_rate": 1.2042477726063113e-05,
+      "loss": 0.8801,
+      "step": 7659
+    },
+    {
+      "epoch": 1.3014338684441777,
+      "grad_norm": 1.625,
+      "learning_rate": 1.2040709834958155e-05,
+      "loss": 0.8041,
+      "step": 7660
+    },
+    {
+      "epoch": 1.3016053325331676,
+      "grad_norm": 1.78125,
+      "learning_rate": 1.2038941877297968e-05,
+      "loss": 0.8988,
+      "step": 7661
+    },
+    {
+      "epoch": 1.3017767966221574,
+      "grad_norm": 1.75,
+      "learning_rate": 1.2037173853140206e-05,
+      "loss": 0.89,
+      "step": 7662
+    },
+    {
+      "epoch": 1.3019482607111472,
+      "grad_norm": 1.59375,
+      "learning_rate": 1.2035405762542534e-05,
+      "loss": 0.8521,
+      "step": 7663
+    },
+    {
+      "epoch": 1.302119724800137,
+      "grad_norm": 1.65625,
+      "learning_rate": 1.2033637605562616e-05,
+      "loss": 0.896,
+      "step": 7664
+    },
+    {
+      "epoch": 1.302291188889127,
+      "grad_norm": 1.8515625,
+      "learning_rate": 1.2031869382258116e-05,
+      "loss": 0.8838,
+      "step": 7665
+    },
+    {
+      "epoch": 1.302462652978117,
+      "grad_norm": 1.8359375,
+      "learning_rate": 1.2030101092686705e-05,
+      "loss": 0.9619,
+      "step": 7666
+    },
+    {
+      "epoch": 1.3026341170671067,
+      "grad_norm": 1.6875,
+      "learning_rate": 1.2028332736906052e-05,
+      "loss": 0.8727,
+      "step": 7667
+    },
+    {
+      "epoch": 1.3028055811560966,
+      "grad_norm": 1.59375,
+      "learning_rate": 1.2026564314973832e-05,
+      "loss": 0.8851,
+      "step": 7668
+    },
+    {
+      "epoch": 1.3029770452450866,
+      "grad_norm": 1.640625,
+      "learning_rate": 1.2024795826947718e-05,
+      "loss": 0.9059,
+      "step": 7669
+    },
+    {
+      "epoch": 1.3031485093340764,
+      "grad_norm": 1.7265625,
+      "learning_rate": 1.2023027272885388e-05,
+      "loss": 0.8688,
+      "step": 7670
+    },
+    {
+      "epoch": 1.3033199734230663,
+      "grad_norm": 1.8046875,
+      "learning_rate": 1.2021258652844521e-05,
+      "loss": 0.9054,
+      "step": 7671
+    },
+    {
+      "epoch": 1.303491437512056,
+      "grad_norm": 1.75,
+      "learning_rate": 1.2019489966882796e-05,
+      "loss": 0.8857,
+      "step": 7672
+    },
+    {
+      "epoch": 1.303662901601046,
+      "grad_norm": 1.6484375,
+      "learning_rate": 1.20177212150579e-05,
+      "loss": 0.8302,
+      "step": 7673
+    },
+    {
+      "epoch": 1.3038343656900357,
+      "grad_norm": 1.640625,
+      "learning_rate": 1.2015952397427513e-05,
+      "loss": 0.7771,
+      "step": 7674
+    },
+    {
+      "epoch": 1.3040058297790256,
+      "grad_norm": 1.71875,
+      "learning_rate": 1.2014183514049331e-05,
+      "loss": 0.8096,
+      "step": 7675
+    },
+    {
+      "epoch": 1.3041772938680154,
+      "grad_norm": 1.6796875,
+      "learning_rate": 1.2012414564981039e-05,
+      "loss": 0.8404,
+      "step": 7676
+    },
+    {
+      "epoch": 1.3043487579570054,
+      "grad_norm": 1.765625,
+      "learning_rate": 1.2010645550280332e-05,
+      "loss": 0.9386,
+      "step": 7677
+    },
+    {
+      "epoch": 1.3045202220459953,
+      "grad_norm": 1.625,
+      "learning_rate": 1.20088764700049e-05,
+      "loss": 0.8149,
+      "step": 7678
+    },
+    {
+      "epoch": 1.304691686134985,
+      "grad_norm": 1.65625,
+      "learning_rate": 1.2007107324212441e-05,
+      "loss": 0.8008,
+      "step": 7679
+    },
+    {
+      "epoch": 1.304863150223975,
+      "grad_norm": 1.609375,
+      "learning_rate": 1.2005338112960657e-05,
+      "loss": 0.8362,
+      "step": 7680
+    },
+    {
+      "epoch": 1.3050346143129647,
+      "grad_norm": 1.6796875,
+      "learning_rate": 1.2003568836307244e-05,
+      "loss": 0.838,
+      "step": 7681
+    },
+    {
+      "epoch": 1.3052060784019548,
+      "grad_norm": 1.6640625,
+      "learning_rate": 1.2001799494309905e-05,
+      "loss": 0.7765,
+      "step": 7682
+    },
+    {
+      "epoch": 1.3053775424909446,
+      "grad_norm": 1.625,
+      "learning_rate": 1.200003008702635e-05,
+      "loss": 0.8431,
+      "step": 7683
+    },
+    {
+      "epoch": 1.3055490065799344,
+      "grad_norm": 1.7734375,
+      "learning_rate": 1.1998260614514279e-05,
+      "loss": 0.8863,
+      "step": 7684
+    },
+    {
+      "epoch": 1.3057204706689243,
+      "grad_norm": 1.78125,
+      "learning_rate": 1.1996491076831404e-05,
+      "loss": 0.8998,
+      "step": 7685
+    },
+    {
+      "epoch": 1.305891934757914,
+      "grad_norm": 1.6953125,
+      "learning_rate": 1.199472147403544e-05,
+      "loss": 0.8527,
+      "step": 7686
+    },
+    {
+      "epoch": 1.306063398846904,
+      "grad_norm": 1.8125,
+      "learning_rate": 1.1992951806184095e-05,
+      "loss": 0.8908,
+      "step": 7687
+    },
+    {
+      "epoch": 1.3062348629358937,
+      "grad_norm": 1.578125,
+      "learning_rate": 1.199118207333509e-05,
+      "loss": 0.8322,
+      "step": 7688
+    },
+    {
+      "epoch": 1.3064063270248838,
+      "grad_norm": 1.6328125,
+      "learning_rate": 1.1989412275546137e-05,
+      "loss": 0.7763,
+      "step": 7689
+    },
+    {
+      "epoch": 1.3065777911138736,
+      "grad_norm": 1.6015625,
+      "learning_rate": 1.198764241287496e-05,
+      "loss": 0.8284,
+      "step": 7690
+    },
+    {
+      "epoch": 1.3067492552028634,
+      "grad_norm": 1.7109375,
+      "learning_rate": 1.1985872485379278e-05,
+      "loss": 0.8426,
+      "step": 7691
+    },
+    {
+      "epoch": 1.3069207192918533,
+      "grad_norm": 1.6328125,
+      "learning_rate": 1.1984102493116813e-05,
+      "loss": 0.8513,
+      "step": 7692
+    },
+    {
+      "epoch": 1.307092183380843,
+      "grad_norm": 1.6484375,
+      "learning_rate": 1.1982332436145296e-05,
+      "loss": 0.7398,
+      "step": 7693
+    },
+    {
+      "epoch": 1.3072636474698331,
+      "grad_norm": 1.78125,
+      "learning_rate": 1.1980562314522457e-05,
+      "loss": 0.8287,
+      "step": 7694
+    },
+    {
+      "epoch": 1.307435111558823,
+      "grad_norm": 1.703125,
+      "learning_rate": 1.1978792128306021e-05,
+      "loss": 0.82,
+      "step": 7695
+    },
+    {
+      "epoch": 1.3076065756478128,
+      "grad_norm": 1.7421875,
+      "learning_rate": 1.1977021877553721e-05,
+      "loss": 0.8936,
+      "step": 7696
+    },
+    {
+      "epoch": 1.3077780397368026,
+      "grad_norm": 1.7109375,
+      "learning_rate": 1.1975251562323293e-05,
+      "loss": 0.9517,
+      "step": 7697
+    },
+    {
+      "epoch": 1.3079495038257924,
+      "grad_norm": 1.671875,
+      "learning_rate": 1.1973481182672475e-05,
+      "loss": 0.8354,
+      "step": 7698
+    },
+    {
+      "epoch": 1.3081209679147823,
+      "grad_norm": 1.6328125,
+      "learning_rate": 1.1971710738659002e-05,
+      "loss": 0.7502,
+      "step": 7699
+    },
+    {
+      "epoch": 1.308292432003772,
+      "grad_norm": 1.640625,
+      "learning_rate": 1.196994023034062e-05,
+      "loss": 0.8179,
+      "step": 7700
+    },
+    {
+      "epoch": 1.308292432003772,
+      "eval_loss": 0.8456034660339355,
+      "eval_runtime": 836.8754,
+      "eval_samples_per_second": 2.986,
+      "eval_steps_per_second": 2.986,
+      "step": 7700
+    },
+    {
+      "epoch": 1.3084638960927621,
+      "grad_norm": 1.65625,
+      "learning_rate": 1.1968169657775069e-05,
+      "loss": 0.8484,
+      "step": 7701
+    },
+    {
+      "epoch": 1.308635360181752,
+      "grad_norm": 1.7265625,
+      "learning_rate": 1.196639902102009e-05,
+      "loss": 0.8856,
+      "step": 7702
+    },
+    {
+      "epoch": 1.3088068242707418,
+      "grad_norm": 1.7890625,
+      "learning_rate": 1.1964628320133436e-05,
+      "loss": 0.8913,
+      "step": 7703
+    },
+    {
+      "epoch": 1.3089782883597316,
+      "grad_norm": 1.671875,
+      "learning_rate": 1.1962857555172854e-05,
+      "loss": 0.8323,
+      "step": 7704
+    },
+    {
+      "epoch": 1.3091497524487214,
+      "grad_norm": 1.6640625,
+      "learning_rate": 1.1961086726196097e-05,
+      "loss": 0.7931,
+      "step": 7705
+    },
+    {
+      "epoch": 1.3093212165377115,
+      "grad_norm": 1.703125,
+      "learning_rate": 1.1959315833260918e-05,
+      "loss": 0.9209,
+      "step": 7706
+    },
+    {
+      "epoch": 1.3094926806267013,
+      "grad_norm": 1.6640625,
+      "learning_rate": 1.1957544876425069e-05,
+      "loss": 0.8677,
+      "step": 7707
+    },
+    {
+      "epoch": 1.3096641447156911,
+      "grad_norm": 1.6015625,
+      "learning_rate": 1.195577385574631e-05,
+      "loss": 0.7754,
+      "step": 7708
+    },
+    {
+      "epoch": 1.309835608804681,
+      "grad_norm": 1.78125,
+      "learning_rate": 1.1954002771282403e-05,
+      "loss": 0.887,
+      "step": 7709
+    },
+    {
+      "epoch": 1.3100070728936708,
+      "grad_norm": 1.7421875,
+      "learning_rate": 1.1952231623091106e-05,
+      "loss": 0.8784,
+      "step": 7710
+    },
+    {
+      "epoch": 1.3101785369826606,
+      "grad_norm": 1.75,
+      "learning_rate": 1.1950460411230185e-05,
+      "loss": 0.8745,
+      "step": 7711
+    },
+    {
+      "epoch": 1.3103500010716504,
+      "grad_norm": 1.6875,
+      "learning_rate": 1.1948689135757406e-05,
+      "loss": 0.8325,
+      "step": 7712
+    },
+    {
+      "epoch": 1.3105214651606405,
+      "grad_norm": 1.6796875,
+      "learning_rate": 1.1946917796730534e-05,
+      "loss": 0.8219,
+      "step": 7713
+    },
+    {
+      "epoch": 1.3106929292496303,
+      "grad_norm": 1.6953125,
+      "learning_rate": 1.1945146394207342e-05,
+      "loss": 0.8945,
+      "step": 7714
+    },
+    {
+      "epoch": 1.3108643933386201,
+      "grad_norm": 1.6484375,
+      "learning_rate": 1.1943374928245603e-05,
+      "loss": 0.8552,
+      "step": 7715
+    },
+    {
+      "epoch": 1.31103585742761,
+      "grad_norm": 1.609375,
+      "learning_rate": 1.1941603398903088e-05,
+      "loss": 0.8237,
+      "step": 7716
+    },
+    {
+      "epoch": 1.3112073215165998,
+      "grad_norm": 1.71875,
+      "learning_rate": 1.193983180623758e-05,
+      "loss": 0.8898,
+      "step": 7717
+    },
+    {
+      "epoch": 1.3113787856055898,
+      "grad_norm": 1.734375,
+      "learning_rate": 1.1938060150306844e-05,
+      "loss": 0.8595,
+      "step": 7718
+    },
+    {
+      "epoch": 1.3115502496945797,
+      "grad_norm": 1.7109375,
+      "learning_rate": 1.1936288431168673e-05,
+      "loss": 0.8116,
+      "step": 7719
+    },
+    {
+      "epoch": 1.3117217137835695,
+      "grad_norm": 1.6796875,
+      "learning_rate": 1.1934516648880842e-05,
+      "loss": 0.8697,
+      "step": 7720
+    },
+    {
+      "epoch": 1.3118931778725593,
+      "grad_norm": 1.703125,
+      "learning_rate": 1.193274480350114e-05,
+      "loss": 0.8571,
+      "step": 7721
+    },
+    {
+      "epoch": 1.3120646419615491,
+      "grad_norm": 1.7265625,
+      "learning_rate": 1.1930972895087349e-05,
+      "loss": 0.8991,
+      "step": 7722
+    },
+    {
+      "epoch": 1.312236106050539,
+      "grad_norm": 1.578125,
+      "learning_rate": 1.1929200923697267e-05,
+      "loss": 0.798,
+      "step": 7723
+    },
+    {
+      "epoch": 1.3124075701395288,
+      "grad_norm": 1.6953125,
+      "learning_rate": 1.1927428889388674e-05,
+      "loss": 0.8255,
+      "step": 7724
+    },
+    {
+      "epoch": 1.3125790342285186,
+      "grad_norm": 1.71875,
+      "learning_rate": 1.1925656792219366e-05,
+      "loss": 0.9045,
+      "step": 7725
+    },
+    {
+      "epoch": 1.3127504983175087,
+      "grad_norm": 1.6796875,
+      "learning_rate": 1.1923884632247143e-05,
+      "loss": 0.9382,
+      "step": 7726
+    },
+    {
+      "epoch": 1.3129219624064985,
+      "grad_norm": 1.65625,
+      "learning_rate": 1.1922112409529793e-05,
+      "loss": 0.86,
+      "step": 7727
+    },
+    {
+      "epoch": 1.3130934264954883,
+      "grad_norm": 1.6875,
+      "learning_rate": 1.192034012412512e-05,
+      "loss": 0.8747,
+      "step": 7728
+    },
+    {
+      "epoch": 1.3132648905844782,
+      "grad_norm": 1.6328125,
+      "learning_rate": 1.1918567776090925e-05,
+      "loss": 0.871,
+      "step": 7729
+    },
+    {
+      "epoch": 1.3134363546734682,
+      "grad_norm": 1.6484375,
+      "learning_rate": 1.191679536548501e-05,
+      "loss": 0.887,
+      "step": 7730
+    },
+    {
+      "epoch": 1.313607818762458,
+      "grad_norm": 1.5546875,
+      "learning_rate": 1.1915022892365182e-05,
+      "loss": 0.7853,
+      "step": 7731
+    },
+    {
+      "epoch": 1.3137792828514478,
+      "grad_norm": 1.7109375,
+      "learning_rate": 1.1913250356789244e-05,
+      "loss": 0.8789,
+      "step": 7732
+    },
+    {
+      "epoch": 1.3139507469404377,
+      "grad_norm": 1.6953125,
+      "learning_rate": 1.1911477758815006e-05,
+      "loss": 0.9076,
+      "step": 7733
+    },
+    {
+      "epoch": 1.3141222110294275,
+      "grad_norm": 1.6953125,
+      "learning_rate": 1.1909705098500287e-05,
+      "loss": 0.8747,
+      "step": 7734
+    },
+    {
+      "epoch": 1.3142936751184173,
+      "grad_norm": 1.7109375,
+      "learning_rate": 1.1907932375902886e-05,
+      "loss": 0.9129,
+      "step": 7735
+    },
+    {
+      "epoch": 1.3144651392074072,
+      "grad_norm": 1.6015625,
+      "learning_rate": 1.1906159591080628e-05,
+      "loss": 0.8726,
+      "step": 7736
+    },
+    {
+      "epoch": 1.314636603296397,
+      "grad_norm": 1.6328125,
+      "learning_rate": 1.190438674409133e-05,
+      "loss": 0.753,
+      "step": 7737
+    },
+    {
+      "epoch": 1.314808067385387,
+      "grad_norm": 1.6328125,
+      "learning_rate": 1.1902613834992807e-05,
+      "loss": 0.8014,
+      "step": 7738
+    },
+    {
+      "epoch": 1.3149795314743769,
+      "grad_norm": 1.71875,
+      "learning_rate": 1.1900840863842883e-05,
+      "loss": 0.8932,
+      "step": 7739
+    },
+    {
+      "epoch": 1.3151509955633667,
+      "grad_norm": 1.7421875,
+      "learning_rate": 1.1899067830699378e-05,
+      "loss": 0.8219,
+      "step": 7740
+    },
+    {
+      "epoch": 1.3153224596523565,
+      "grad_norm": 1.6796875,
+      "learning_rate": 1.1897294735620123e-05,
+      "loss": 0.9269,
+      "step": 7741
+    },
+    {
+      "epoch": 1.3154939237413465,
+      "grad_norm": 1.8046875,
+      "learning_rate": 1.1895521578662939e-05,
+      "loss": 0.9126,
+      "step": 7742
+    },
+    {
+      "epoch": 1.3156653878303364,
+      "grad_norm": 1.796875,
+      "learning_rate": 1.189374835988566e-05,
+      "loss": 0.8141,
+      "step": 7743
+    },
+    {
+      "epoch": 1.3158368519193262,
+      "grad_norm": 1.671875,
+      "learning_rate": 1.1891975079346117e-05,
+      "loss": 0.8801,
+      "step": 7744
+    },
+    {
+      "epoch": 1.316008316008316,
+      "grad_norm": 1.8125,
+      "learning_rate": 1.1890201737102141e-05,
+      "loss": 0.9298,
+      "step": 7745
+    },
+    {
+      "epoch": 1.3161797800973059,
+      "grad_norm": 1.8203125,
+      "learning_rate": 1.188842833321157e-05,
+      "loss": 0.9065,
+      "step": 7746
+    },
+    {
+      "epoch": 1.3163512441862957,
+      "grad_norm": 1.7109375,
+      "learning_rate": 1.1886654867732239e-05,
+      "loss": 0.8665,
+      "step": 7747
+    },
+    {
+      "epoch": 1.3165227082752855,
+      "grad_norm": 1.6875,
+      "learning_rate": 1.1884881340721985e-05,
+      "loss": 0.8517,
+      "step": 7748
+    },
+    {
+      "epoch": 1.3166941723642753,
+      "grad_norm": 1.765625,
+      "learning_rate": 1.1883107752238657e-05,
+      "loss": 0.8489,
+      "step": 7749
+    },
+    {
+      "epoch": 1.3168656364532654,
+      "grad_norm": 1.7421875,
+      "learning_rate": 1.188133410234009e-05,
+      "loss": 0.8792,
+      "step": 7750
+    },
+    {
+      "epoch": 1.3170371005422552,
+      "grad_norm": 1.703125,
+      "learning_rate": 1.1879560391084137e-05,
+      "loss": 0.8721,
+      "step": 7751
+    },
+    {
+      "epoch": 1.317208564631245,
+      "grad_norm": 1.6875,
+      "learning_rate": 1.1877786618528642e-05,
+      "loss": 0.874,
+      "step": 7752
+    },
+    {
+      "epoch": 1.3173800287202349,
+      "grad_norm": 1.734375,
+      "learning_rate": 1.1876012784731454e-05,
+      "loss": 0.8498,
+      "step": 7753
+    },
+    {
+      "epoch": 1.317551492809225,
+      "grad_norm": 1.6328125,
+      "learning_rate": 1.1874238889750423e-05,
+      "loss": 0.7978,
+      "step": 7754
+    },
+    {
+      "epoch": 1.3177229568982147,
+      "grad_norm": 1.6875,
+      "learning_rate": 1.1872464933643406e-05,
+      "loss": 0.834,
+      "step": 7755
+    },
+    {
+      "epoch": 1.3178944209872046,
+      "grad_norm": 1.6953125,
+      "learning_rate": 1.1870690916468255e-05,
+      "loss": 0.8322,
+      "step": 7756
+    },
+    {
+      "epoch": 1.3180658850761944,
+      "grad_norm": 1.7265625,
+      "learning_rate": 1.1868916838282829e-05,
+      "loss": 0.9117,
+      "step": 7757
+    },
+    {
+      "epoch": 1.3182373491651842,
+      "grad_norm": 1.6328125,
+      "learning_rate": 1.1867142699144986e-05,
+      "loss": 0.7738,
+      "step": 7758
+    },
+    {
+      "epoch": 1.318408813254174,
+      "grad_norm": 1.6875,
+      "learning_rate": 1.1865368499112591e-05,
+      "loss": 0.8178,
+      "step": 7759
+    },
+    {
+      "epoch": 1.3185802773431639,
+      "grad_norm": 1.6796875,
+      "learning_rate": 1.1863594238243503e-05,
+      "loss": 0.8666,
+      "step": 7760
+    },
+    {
+      "epoch": 1.3187517414321537,
+      "grad_norm": 1.640625,
+      "learning_rate": 1.1861819916595592e-05,
+      "loss": 0.8319,
+      "step": 7761
+    },
+    {
+      "epoch": 1.3189232055211437,
+      "grad_norm": 1.703125,
+      "learning_rate": 1.1860045534226719e-05,
+      "loss": 0.8273,
+      "step": 7762
+    },
+    {
+      "epoch": 1.3190946696101336,
+      "grad_norm": 1.671875,
+      "learning_rate": 1.1858271091194762e-05,
+      "loss": 0.8214,
+      "step": 7763
+    },
+    {
+      "epoch": 1.3192661336991234,
+      "grad_norm": 1.6796875,
+      "learning_rate": 1.1856496587557582e-05,
+      "loss": 0.8312,
+      "step": 7764
+    },
+    {
+      "epoch": 1.3194375977881132,
+      "grad_norm": 1.65625,
+      "learning_rate": 1.185472202337306e-05,
+      "loss": 0.8638,
+      "step": 7765
+    },
+    {
+      "epoch": 1.3196090618771033,
+      "grad_norm": 1.6484375,
+      "learning_rate": 1.1852947398699065e-05,
+      "loss": 0.8527,
+      "step": 7766
+    },
+    {
+      "epoch": 1.319780525966093,
+      "grad_norm": 1.671875,
+      "learning_rate": 1.1851172713593482e-05,
+      "loss": 0.8765,
+      "step": 7767
+    },
+    {
+      "epoch": 1.319951990055083,
+      "grad_norm": 6.34375,
+      "learning_rate": 1.1849397968114183e-05,
+      "loss": 0.9896,
+      "step": 7768
+    },
+    {
+      "epoch": 1.3201234541440727,
+      "grad_norm": 1.6484375,
+      "learning_rate": 1.1847623162319056e-05,
+      "loss": 0.7787,
+      "step": 7769
+    },
+    {
+      "epoch": 1.3202949182330626,
+      "grad_norm": 1.7890625,
+      "learning_rate": 1.1845848296265976e-05,
+      "loss": 0.8828,
+      "step": 7770
+    },
+    {
+      "epoch": 1.3204663823220524,
+      "grad_norm": 1.65625,
+      "learning_rate": 1.1844073370012835e-05,
+      "loss": 0.8136,
+      "step": 7771
+    },
+    {
+      "epoch": 1.3206378464110422,
+      "grad_norm": 1.7578125,
+      "learning_rate": 1.1842298383617515e-05,
+      "loss": 0.8852,
+      "step": 7772
+    },
+    {
+      "epoch": 1.320809310500032,
+      "grad_norm": 1.7734375,
+      "learning_rate": 1.1840523337137908e-05,
+      "loss": 0.8697,
+      "step": 7773
+    },
+    {
+      "epoch": 1.320980774589022,
+      "grad_norm": 1.7109375,
+      "learning_rate": 1.1838748230631905e-05,
+      "loss": 0.8943,
+      "step": 7774
+    },
+    {
+      "epoch": 1.321152238678012,
+      "grad_norm": 1.6796875,
+      "learning_rate": 1.1836973064157399e-05,
+      "loss": 0.8774,
+      "step": 7775
+    },
+    {
+      "epoch": 1.3213237027670017,
+      "grad_norm": 1.8125,
+      "learning_rate": 1.1835197837772282e-05,
+      "loss": 0.841,
+      "step": 7776
+    },
+    {
+      "epoch": 1.3214951668559916,
+      "grad_norm": 1.6953125,
+      "learning_rate": 1.1833422551534453e-05,
+      "loss": 0.9008,
+      "step": 7777
+    },
+    {
+      "epoch": 1.3216666309449814,
+      "grad_norm": 1.53125,
+      "learning_rate": 1.183164720550181e-05,
+      "loss": 0.8287,
+      "step": 7778
+    },
+    {
+      "epoch": 1.3218380950339714,
+      "grad_norm": 1.7109375,
+      "learning_rate": 1.1829871799732255e-05,
+      "loss": 0.8227,
+      "step": 7779
+    },
+    {
+      "epoch": 1.3220095591229613,
+      "grad_norm": 1.65625,
+      "learning_rate": 1.1828096334283692e-05,
+      "loss": 0.9078,
+      "step": 7780
+    },
+    {
+      "epoch": 1.322181023211951,
+      "grad_norm": 1.625,
+      "learning_rate": 1.1826320809214023e-05,
+      "loss": 0.8535,
+      "step": 7781
+    },
+    {
+      "epoch": 1.322352487300941,
+      "grad_norm": 1.71875,
+      "learning_rate": 1.1824545224581155e-05,
+      "loss": 0.8968,
+      "step": 7782
+    },
+    {
+      "epoch": 1.3225239513899307,
+      "grad_norm": 1.703125,
+      "learning_rate": 1.1822769580442995e-05,
+      "loss": 0.8102,
+      "step": 7783
+    },
+    {
+      "epoch": 1.3226954154789206,
+      "grad_norm": 1.7265625,
+      "learning_rate": 1.1820993876857453e-05,
+      "loss": 0.8553,
+      "step": 7784
+    },
+    {
+      "epoch": 1.3228668795679104,
+      "grad_norm": 1.6640625,
+      "learning_rate": 1.1819218113882448e-05,
+      "loss": 0.8806,
+      "step": 7785
+    },
+    {
+      "epoch": 1.3230383436569004,
+      "grad_norm": 1.609375,
+      "learning_rate": 1.1817442291575887e-05,
+      "loss": 0.7205,
+      "step": 7786
+    },
+    {
+      "epoch": 1.3232098077458903,
+      "grad_norm": 1.625,
+      "learning_rate": 1.181566640999569e-05,
+      "loss": 0.9008,
+      "step": 7787
+    },
+    {
+      "epoch": 1.32338127183488,
+      "grad_norm": 1.75,
+      "learning_rate": 1.1813890469199774e-05,
+      "loss": 0.8797,
+      "step": 7788
+    },
+    {
+      "epoch": 1.32355273592387,
+      "grad_norm": 1.7109375,
+      "learning_rate": 1.181211446924606e-05,
+      "loss": 0.9086,
+      "step": 7789
+    },
+    {
+      "epoch": 1.3237242000128597,
+      "grad_norm": 1.6015625,
+      "learning_rate": 1.1810338410192468e-05,
+      "loss": 0.8851,
+      "step": 7790
+    },
+    {
+      "epoch": 1.3238956641018498,
+      "grad_norm": 1.703125,
+      "learning_rate": 1.1808562292096923e-05,
+      "loss": 0.8203,
+      "step": 7791
+    },
+    {
+      "epoch": 1.3240671281908396,
+      "grad_norm": 1.7421875,
+      "learning_rate": 1.1806786115017354e-05,
+      "loss": 0.8045,
+      "step": 7792
+    },
+    {
+      "epoch": 1.3242385922798294,
+      "grad_norm": 1.671875,
+      "learning_rate": 1.1805009879011686e-05,
+      "loss": 0.8182,
+      "step": 7793
+    },
+    {
+      "epoch": 1.3244100563688193,
+      "grad_norm": 1.625,
+      "learning_rate": 1.1803233584137847e-05,
+      "loss": 0.9296,
+      "step": 7794
+    },
+    {
+      "epoch": 1.324581520457809,
+      "grad_norm": 1.734375,
+      "learning_rate": 1.1801457230453768e-05,
+      "loss": 0.9271,
+      "step": 7795
+    },
+    {
+      "epoch": 1.324752984546799,
+      "grad_norm": 1.765625,
+      "learning_rate": 1.1799680818017387e-05,
+      "loss": 0.8048,
+      "step": 7796
+    },
+    {
+      "epoch": 1.3249244486357887,
+      "grad_norm": 1.703125,
+      "learning_rate": 1.179790434688664e-05,
+      "loss": 0.8658,
+      "step": 7797
+    },
+    {
+      "epoch": 1.3250959127247788,
+      "grad_norm": 1.578125,
+      "learning_rate": 1.179612781711946e-05,
+      "loss": 0.8503,
+      "step": 7798
+    },
+    {
+      "epoch": 1.3252673768137686,
+      "grad_norm": 1.6484375,
+      "learning_rate": 1.179435122877379e-05,
+      "loss": 0.8722,
+      "step": 7799
+    },
+    {
+      "epoch": 1.3254388409027584,
+      "grad_norm": 1.5625,
+      "learning_rate": 1.1792574581907567e-05,
+      "loss": 0.7956,
+      "step": 7800
+    },
+    {
+      "epoch": 1.3256103049917483,
+      "grad_norm": 1.6875,
+      "learning_rate": 1.179079787657874e-05,
+      "loss": 0.8823,
+      "step": 7801
+    },
+    {
+      "epoch": 1.325781769080738,
+      "grad_norm": 1.75,
+      "learning_rate": 1.1789021112845247e-05,
+      "loss": 0.9627,
+      "step": 7802
+    },
+    {
+      "epoch": 1.3259532331697281,
+      "grad_norm": 1.59375,
+      "learning_rate": 1.178724429076504e-05,
+      "loss": 0.8166,
+      "step": 7803
+    },
+    {
+      "epoch": 1.326124697258718,
+      "grad_norm": 1.734375,
+      "learning_rate": 1.1785467410396066e-05,
+      "loss": 0.7876,
+      "step": 7804
+    },
+    {
+      "epoch": 1.3262961613477078,
+      "grad_norm": 1.7109375,
+      "learning_rate": 1.1783690471796276e-05,
+      "loss": 0.8619,
+      "step": 7805
+    },
+    {
+      "epoch": 1.3264676254366976,
+      "grad_norm": 1.6328125,
+      "learning_rate": 1.1781913475023626e-05,
+      "loss": 0.8014,
+      "step": 7806
+    },
+    {
+      "epoch": 1.3266390895256874,
+      "grad_norm": 1.6015625,
+      "learning_rate": 1.1780136420136063e-05,
+      "loss": 0.8473,
+      "step": 7807
+    },
+    {
+      "epoch": 1.3268105536146773,
+      "grad_norm": 1.6484375,
+      "learning_rate": 1.1778359307191552e-05,
+      "loss": 0.8229,
+      "step": 7808
+    },
+    {
+      "epoch": 1.326982017703667,
+      "grad_norm": 1.8203125,
+      "learning_rate": 1.1776582136248047e-05,
+      "loss": 0.8526,
+      "step": 7809
+    },
+    {
+      "epoch": 1.3271534817926571,
+      "grad_norm": 1.640625,
+      "learning_rate": 1.1774804907363508e-05,
+      "loss": 0.8695,
+      "step": 7810
+    },
+    {
+      "epoch": 1.327324945881647,
+      "grad_norm": 1.7109375,
+      "learning_rate": 1.1773027620595897e-05,
+      "loss": 0.8008,
+      "step": 7811
+    },
+    {
+      "epoch": 1.3274964099706368,
+      "grad_norm": 1.7421875,
+      "learning_rate": 1.1771250276003179e-05,
+      "loss": 0.9204,
+      "step": 7812
+    },
+    {
+      "epoch": 1.3276678740596266,
+      "grad_norm": 1.7109375,
+      "learning_rate": 1.1769472873643317e-05,
+      "loss": 0.8668,
+      "step": 7813
+    },
+    {
+      "epoch": 1.3278393381486164,
+      "grad_norm": 1.6953125,
+      "learning_rate": 1.1767695413574284e-05,
+      "loss": 0.8277,
+      "step": 7814
+    },
+    {
+      "epoch": 1.3280108022376065,
+      "grad_norm": 1.65625,
+      "learning_rate": 1.1765917895854046e-05,
+      "loss": 0.7791,
+      "step": 7815
+    },
+    {
+      "epoch": 1.3281822663265963,
+      "grad_norm": 1.7109375,
+      "learning_rate": 1.1764140320540578e-05,
+      "loss": 0.8521,
+      "step": 7816
+    },
+    {
+      "epoch": 1.3283537304155861,
+      "grad_norm": 1.734375,
+      "learning_rate": 1.1762362687691849e-05,
+      "loss": 0.9123,
+      "step": 7817
+    },
+    {
+      "epoch": 1.328525194504576,
+      "grad_norm": 1.546875,
+      "learning_rate": 1.1760584997365835e-05,
+      "loss": 0.8249,
+      "step": 7818
+    },
+    {
+      "epoch": 1.3286966585935658,
+      "grad_norm": 1.6796875,
+      "learning_rate": 1.1758807249620517e-05,
+      "loss": 0.7786,
+      "step": 7819
+    },
+    {
+      "epoch": 1.3288681226825556,
+      "grad_norm": 1.640625,
+      "learning_rate": 1.175702944451387e-05,
+      "loss": 0.8714,
+      "step": 7820
+    },
+    {
+      "epoch": 1.3290395867715454,
+      "grad_norm": 1.640625,
+      "learning_rate": 1.1755251582103879e-05,
+      "loss": 0.8321,
+      "step": 7821
+    },
+    {
+      "epoch": 1.3292110508605353,
+      "grad_norm": 1.6484375,
+      "learning_rate": 1.1753473662448523e-05,
+      "loss": 0.8579,
+      "step": 7822
+    },
+    {
+      "epoch": 1.3293825149495253,
+      "grad_norm": 1.6640625,
+      "learning_rate": 1.1751695685605784e-05,
+      "loss": 0.8547,
+      "step": 7823
+    },
+    {
+      "epoch": 1.3295539790385151,
+      "grad_norm": 1.7265625,
+      "learning_rate": 1.1749917651633655e-05,
+      "loss": 0.8092,
+      "step": 7824
+    },
+    {
+      "epoch": 1.329725443127505,
+      "grad_norm": 1.6953125,
+      "learning_rate": 1.1748139560590123e-05,
+      "loss": 0.8526,
+      "step": 7825
+    },
+    {
+      "epoch": 1.3298969072164948,
+      "grad_norm": 1.71875,
+      "learning_rate": 1.1746361412533175e-05,
+      "loss": 0.841,
+      "step": 7826
+    },
+    {
+      "epoch": 1.3300683713054848,
+      "grad_norm": 1.734375,
+      "learning_rate": 1.174458320752081e-05,
+      "loss": 0.8476,
+      "step": 7827
+    },
+    {
+      "epoch": 1.3302398353944747,
+      "grad_norm": 1.6484375,
+      "learning_rate": 1.1742804945611014e-05,
+      "loss": 0.932,
+      "step": 7828
+    },
+    {
+      "epoch": 1.3304112994834645,
+      "grad_norm": 1.5703125,
+      "learning_rate": 1.1741026626861784e-05,
+      "loss": 0.786,
+      "step": 7829
+    },
+    {
+      "epoch": 1.3305827635724543,
+      "grad_norm": 1.6875,
+      "learning_rate": 1.1739248251331124e-05,
+      "loss": 0.9036,
+      "step": 7830
+    },
+    {
+      "epoch": 1.3307542276614441,
+      "grad_norm": 1.7421875,
+      "learning_rate": 1.1737469819077026e-05,
+      "loss": 0.7656,
+      "step": 7831
+    },
+    {
+      "epoch": 1.330925691750434,
+      "grad_norm": 1.7265625,
+      "learning_rate": 1.1735691330157492e-05,
+      "loss": 0.9576,
+      "step": 7832
+    },
+    {
+      "epoch": 1.3310971558394238,
+      "grad_norm": 1.5703125,
+      "learning_rate": 1.1733912784630532e-05,
+      "loss": 0.7358,
+      "step": 7833
+    },
+    {
+      "epoch": 1.3312686199284136,
+      "grad_norm": 1.703125,
+      "learning_rate": 1.1732134182554144e-05,
+      "loss": 0.8518,
+      "step": 7834
+    },
+    {
+      "epoch": 1.3314400840174037,
+      "grad_norm": 1.734375,
+      "learning_rate": 1.173035552398634e-05,
+      "loss": 0.8398,
+      "step": 7835
+    },
+    {
+      "epoch": 1.3316115481063935,
+      "grad_norm": 1.6484375,
+      "learning_rate": 1.1728576808985126e-05,
+      "loss": 0.8388,
+      "step": 7836
+    },
+    {
+      "epoch": 1.3317830121953833,
+      "grad_norm": 1.7265625,
+      "learning_rate": 1.1726798037608514e-05,
+      "loss": 0.9044,
+      "step": 7837
+    },
+    {
+      "epoch": 1.3319544762843731,
+      "grad_norm": 1.7421875,
+      "learning_rate": 1.1725019209914514e-05,
+      "loss": 0.9309,
+      "step": 7838
+    },
+    {
+      "epoch": 1.3321259403733632,
+      "grad_norm": 1.65625,
+      "learning_rate": 1.1723240325961142e-05,
+      "loss": 0.8207,
+      "step": 7839
+    },
+    {
+      "epoch": 1.332297404462353,
+      "grad_norm": 1.671875,
+      "learning_rate": 1.1721461385806414e-05,
+      "loss": 0.846,
+      "step": 7840
+    },
+    {
+      "epoch": 1.3324688685513428,
+      "grad_norm": 1.6875,
+      "learning_rate": 1.1719682389508346e-05,
+      "loss": 0.8934,
+      "step": 7841
+    },
+    {
+      "epoch": 1.3326403326403327,
+      "grad_norm": 1.7578125,
+      "learning_rate": 1.171790333712496e-05,
+      "loss": 0.8431,
+      "step": 7842
+    },
+    {
+      "epoch": 1.3328117967293225,
+      "grad_norm": 1.7421875,
+      "learning_rate": 1.1716124228714278e-05,
+      "loss": 0.8962,
+      "step": 7843
+    },
+    {
+      "epoch": 1.3329832608183123,
+      "grad_norm": 1.71875,
+      "learning_rate": 1.1714345064334325e-05,
+      "loss": 0.923,
+      "step": 7844
+    },
+    {
+      "epoch": 1.3331547249073021,
+      "grad_norm": 1.734375,
+      "learning_rate": 1.171256584404312e-05,
+      "loss": 0.8543,
+      "step": 7845
+    },
+    {
+      "epoch": 1.333326188996292,
+      "grad_norm": 1.6328125,
+      "learning_rate": 1.1710786567898696e-05,
+      "loss": 0.8548,
+      "step": 7846
+    },
+    {
+      "epoch": 1.333497653085282,
+      "grad_norm": 1.796875,
+      "learning_rate": 1.1709007235959078e-05,
+      "loss": 0.8992,
+      "step": 7847
+    },
+    {
+      "epoch": 1.3336691171742718,
+      "grad_norm": 1.6796875,
+      "learning_rate": 1.1707227848282301e-05,
+      "loss": 0.8639,
+      "step": 7848
+    },
+    {
+      "epoch": 1.3338405812632617,
+      "grad_norm": 1.75,
+      "learning_rate": 1.1705448404926392e-05,
+      "loss": 0.8901,
+      "step": 7849
+    },
+    {
+      "epoch": 1.3340120453522515,
+      "grad_norm": 1.625,
+      "learning_rate": 1.1703668905949393e-05,
+      "loss": 0.8759,
+      "step": 7850
+    },
+    {
+      "epoch": 1.3341835094412415,
+      "grad_norm": 1.65625,
+      "learning_rate": 1.170188935140933e-05,
+      "loss": 0.8358,
+      "step": 7851
+    },
+    {
+      "epoch": 1.3343549735302314,
+      "grad_norm": 1.7578125,
+      "learning_rate": 1.170010974136425e-05,
+      "loss": 0.8509,
+      "step": 7852
+    },
+    {
+      "epoch": 1.3345264376192212,
+      "grad_norm": 1.7265625,
+      "learning_rate": 1.1698330075872188e-05,
+      "loss": 0.8816,
+      "step": 7853
+    },
+    {
+      "epoch": 1.334697901708211,
+      "grad_norm": 1.625,
+      "learning_rate": 1.1696550354991187e-05,
+      "loss": 0.8541,
+      "step": 7854
+    },
+    {
+      "epoch": 1.3348693657972008,
+      "grad_norm": 1.7578125,
+      "learning_rate": 1.1694770578779293e-05,
+      "loss": 0.887,
+      "step": 7855
+    },
+    {
+      "epoch": 1.3350408298861907,
+      "grad_norm": 1.703125,
+      "learning_rate": 1.1692990747294546e-05,
+      "loss": 0.8772,
+      "step": 7856
+    },
+    {
+      "epoch": 1.3352122939751805,
+      "grad_norm": 1.578125,
+      "learning_rate": 1.1691210860594997e-05,
+      "loss": 0.8723,
+      "step": 7857
+    },
+    {
+      "epoch": 1.3353837580641703,
+      "grad_norm": 1.6875,
+      "learning_rate": 1.1689430918738691e-05,
+      "loss": 0.901,
+      "step": 7858
+    },
+    {
+      "epoch": 1.3355552221531604,
+      "grad_norm": 1.640625,
+      "learning_rate": 1.1687650921783683e-05,
+      "loss": 0.8865,
+      "step": 7859
+    },
+    {
+      "epoch": 1.3357266862421502,
+      "grad_norm": 1.6796875,
+      "learning_rate": 1.168587086978802e-05,
+      "loss": 0.8233,
+      "step": 7860
+    },
+    {
+      "epoch": 1.33589815033114,
+      "grad_norm": 1.703125,
+      "learning_rate": 1.1684090762809762e-05,
+      "loss": 0.9395,
+      "step": 7861
+    },
+    {
+      "epoch": 1.3360696144201298,
+      "grad_norm": 1.640625,
+      "learning_rate": 1.1682310600906962e-05,
+      "loss": 0.8309,
+      "step": 7862
+    },
+    {
+      "epoch": 1.3362410785091199,
+      "grad_norm": 1.65625,
+      "learning_rate": 1.1680530384137681e-05,
+      "loss": 0.9013,
+      "step": 7863
+    },
+    {
+      "epoch": 1.3364125425981097,
+      "grad_norm": 1.6640625,
+      "learning_rate": 1.1678750112559971e-05,
+      "loss": 0.8431,
+      "step": 7864
+    },
+    {
+      "epoch": 1.3365840066870995,
+      "grad_norm": 1.78125,
+      "learning_rate": 1.1676969786231903e-05,
+      "loss": 0.9223,
+      "step": 7865
+    },
+    {
+      "epoch": 1.3367554707760894,
+      "grad_norm": 1.65625,
+      "learning_rate": 1.167518940521153e-05,
+      "loss": 0.7831,
+      "step": 7866
+    },
+    {
+      "epoch": 1.3369269348650792,
+      "grad_norm": 1.84375,
+      "learning_rate": 1.167340896955693e-05,
+      "loss": 0.8798,
+      "step": 7867
+    },
+    {
+      "epoch": 1.337098398954069,
+      "grad_norm": 1.7109375,
+      "learning_rate": 1.1671628479326157e-05,
+      "loss": 0.8683,
+      "step": 7868
+    },
+    {
+      "epoch": 1.3372698630430588,
+      "grad_norm": 1.6875,
+      "learning_rate": 1.1669847934577285e-05,
+      "loss": 0.8294,
+      "step": 7869
+    },
+    {
+      "epoch": 1.3374413271320487,
+      "grad_norm": 1.765625,
+      "learning_rate": 1.1668067335368383e-05,
+      "loss": 0.8769,
+      "step": 7870
+    },
+    {
+      "epoch": 1.3376127912210387,
+      "grad_norm": 1.6484375,
+      "learning_rate": 1.1666286681757524e-05,
+      "loss": 0.8865,
+      "step": 7871
+    },
+    {
+      "epoch": 1.3377842553100285,
+      "grad_norm": 1.6796875,
+      "learning_rate": 1.1664505973802782e-05,
+      "loss": 0.8907,
+      "step": 7872
+    },
+    {
+      "epoch": 1.3379557193990184,
+      "grad_norm": 1.65625,
+      "learning_rate": 1.1662725211562233e-05,
+      "loss": 0.8434,
+      "step": 7873
+    },
+    {
+      "epoch": 1.3381271834880082,
+      "grad_norm": 1.75,
+      "learning_rate": 1.1660944395093954e-05,
+      "loss": 0.8884,
+      "step": 7874
+    },
+    {
+      "epoch": 1.338298647576998,
+      "grad_norm": 1.6328125,
+      "learning_rate": 1.1659163524456023e-05,
+      "loss": 0.8571,
+      "step": 7875
+    },
+    {
+      "epoch": 1.338470111665988,
+      "grad_norm": 1.71875,
+      "learning_rate": 1.1657382599706522e-05,
+      "loss": 0.8605,
+      "step": 7876
+    },
+    {
+      "epoch": 1.3386415757549779,
+      "grad_norm": 1.671875,
+      "learning_rate": 1.1655601620903531e-05,
+      "loss": 0.9084,
+      "step": 7877
+    },
+    {
+      "epoch": 1.3388130398439677,
+      "grad_norm": 1.6953125,
+      "learning_rate": 1.165382058810514e-05,
+      "loss": 0.8482,
+      "step": 7878
+    },
+    {
+      "epoch": 1.3389845039329575,
+      "grad_norm": 1.7421875,
+      "learning_rate": 1.165203950136943e-05,
+      "loss": 0.8411,
+      "step": 7879
+    },
+    {
+      "epoch": 1.3391559680219474,
+      "grad_norm": 1.6953125,
+      "learning_rate": 1.1650258360754491e-05,
+      "loss": 0.8961,
+      "step": 7880
+    },
+    {
+      "epoch": 1.3393274321109372,
+      "grad_norm": 1.75,
+      "learning_rate": 1.1648477166318413e-05,
+      "loss": 0.8305,
+      "step": 7881
+    },
+    {
+      "epoch": 1.339498896199927,
+      "grad_norm": 1.6640625,
+      "learning_rate": 1.1646695918119289e-05,
+      "loss": 0.8248,
+      "step": 7882
+    },
+    {
+      "epoch": 1.339670360288917,
+      "grad_norm": 1.765625,
+      "learning_rate": 1.1644914616215206e-05,
+      "loss": 0.9079,
+      "step": 7883
+    },
+    {
+      "epoch": 1.3398418243779069,
+      "grad_norm": 1.7109375,
+      "learning_rate": 1.1643133260664268e-05,
+      "loss": 0.9124,
+      "step": 7884
+    },
+    {
+      "epoch": 1.3400132884668967,
+      "grad_norm": 1.7265625,
+      "learning_rate": 1.1641351851524564e-05,
+      "loss": 0.8732,
+      "step": 7885
+    },
+    {
+      "epoch": 1.3401847525558865,
+      "grad_norm": 1.796875,
+      "learning_rate": 1.1639570388854196e-05,
+      "loss": 0.8122,
+      "step": 7886
+    },
+    {
+      "epoch": 1.3403562166448764,
+      "grad_norm": 1.7734375,
+      "learning_rate": 1.1637788872711262e-05,
+      "loss": 0.8699,
+      "step": 7887
+    },
+    {
+      "epoch": 1.3405276807338664,
+      "grad_norm": 1.75,
+      "learning_rate": 1.1636007303153867e-05,
+      "loss": 0.9392,
+      "step": 7888
+    },
+    {
+      "epoch": 1.3406991448228562,
+      "grad_norm": 1.7265625,
+      "learning_rate": 1.163422568024011e-05,
+      "loss": 0.9003,
+      "step": 7889
+    },
+    {
+      "epoch": 1.340870608911846,
+      "grad_norm": 1.6875,
+      "learning_rate": 1.1632444004028105e-05,
+      "loss": 0.88,
+      "step": 7890
+    },
+    {
+      "epoch": 1.3410420730008359,
+      "grad_norm": 1.75,
+      "learning_rate": 1.1630662274575952e-05,
+      "loss": 0.8422,
+      "step": 7891
+    },
+    {
+      "epoch": 1.3412135370898257,
+      "grad_norm": 1.7421875,
+      "learning_rate": 1.162888049194176e-05,
+      "loss": 0.9335,
+      "step": 7892
+    },
+    {
+      "epoch": 1.3413850011788155,
+      "grad_norm": 1.703125,
+      "learning_rate": 1.1627098656183645e-05,
+      "loss": 0.8748,
+      "step": 7893
+    },
+    {
+      "epoch": 1.3415564652678054,
+      "grad_norm": 1.578125,
+      "learning_rate": 1.1625316767359711e-05,
+      "loss": 0.7939,
+      "step": 7894
+    },
+    {
+      "epoch": 1.3417279293567954,
+      "grad_norm": 1.7578125,
+      "learning_rate": 1.162353482552808e-05,
+      "loss": 0.8575,
+      "step": 7895
+    },
+    {
+      "epoch": 1.3418993934457852,
+      "grad_norm": 1.703125,
+      "learning_rate": 1.1621752830746863e-05,
+      "loss": 0.8959,
+      "step": 7896
+    },
+    {
+      "epoch": 1.342070857534775,
+      "grad_norm": 1.6328125,
+      "learning_rate": 1.1619970783074179e-05,
+      "loss": 0.8454,
+      "step": 7897
+    },
+    {
+      "epoch": 1.342242321623765,
+      "grad_norm": 1.7109375,
+      "learning_rate": 1.1618188682568148e-05,
+      "loss": 0.9552,
+      "step": 7898
+    },
+    {
+      "epoch": 1.3424137857127547,
+      "grad_norm": 1.71875,
+      "learning_rate": 1.1616406529286891e-05,
+      "loss": 0.858,
+      "step": 7899
+    },
+    {
+      "epoch": 1.3425852498017448,
+      "grad_norm": 1.625,
+      "learning_rate": 1.161462432328853e-05,
+      "loss": 0.9018,
+      "step": 7900
+    },
+    {
+      "epoch": 1.3427567138907346,
+      "grad_norm": 1.7109375,
+      "learning_rate": 1.161284206463119e-05,
+      "loss": 0.8469,
+      "step": 7901
+    },
+    {
+      "epoch": 1.3429281779797244,
+      "grad_norm": 1.6015625,
+      "learning_rate": 1.1611059753373e-05,
+      "loss": 0.8626,
+      "step": 7902
+    },
+    {
+      "epoch": 1.3430996420687142,
+      "grad_norm": 1.6953125,
+      "learning_rate": 1.1609277389572082e-05,
+      "loss": 0.8517,
+      "step": 7903
+    },
+    {
+      "epoch": 1.343271106157704,
+      "grad_norm": 1.828125,
+      "learning_rate": 1.1607494973286566e-05,
+      "loss": 0.888,
+      "step": 7904
+    },
+    {
+      "epoch": 1.343442570246694,
+      "grad_norm": 1.7578125,
+      "learning_rate": 1.1605712504574588e-05,
+      "loss": 0.993,
+      "step": 7905
+    },
+    {
+      "epoch": 1.3436140343356837,
+      "grad_norm": 1.8203125,
+      "learning_rate": 1.160392998349428e-05,
+      "loss": 0.9164,
+      "step": 7906
+    },
+    {
+      "epoch": 1.3437854984246738,
+      "grad_norm": 1.6640625,
+      "learning_rate": 1.1602147410103772e-05,
+      "loss": 0.7978,
+      "step": 7907
+    },
+    {
+      "epoch": 1.3439569625136636,
+      "grad_norm": 1.6171875,
+      "learning_rate": 1.1600364784461207e-05,
+      "loss": 0.8315,
+      "step": 7908
+    },
+    {
+      "epoch": 1.3441284266026534,
+      "grad_norm": 1.6328125,
+      "learning_rate": 1.1598582106624716e-05,
+      "loss": 0.8325,
+      "step": 7909
+    },
+    {
+      "epoch": 1.3442998906916432,
+      "grad_norm": 1.6796875,
+      "learning_rate": 1.1596799376652446e-05,
+      "loss": 0.8375,
+      "step": 7910
+    },
+    {
+      "epoch": 1.344471354780633,
+      "grad_norm": 1.8125,
+      "learning_rate": 1.1595016594602535e-05,
+      "loss": 0.8738,
+      "step": 7911
+    },
+    {
+      "epoch": 1.3446428188696231,
+      "grad_norm": 1.703125,
+      "learning_rate": 1.1593233760533127e-05,
+      "loss": 0.8572,
+      "step": 7912
+    },
+    {
+      "epoch": 1.344814282958613,
+      "grad_norm": 1.671875,
+      "learning_rate": 1.1591450874502365e-05,
+      "loss": 0.8272,
+      "step": 7913
+    },
+    {
+      "epoch": 1.3449857470476028,
+      "grad_norm": 1.6796875,
+      "learning_rate": 1.1589667936568399e-05,
+      "loss": 0.8233,
+      "step": 7914
+    },
+    {
+      "epoch": 1.3451572111365926,
+      "grad_norm": 1.7421875,
+      "learning_rate": 1.1587884946789374e-05,
+      "loss": 0.8449,
+      "step": 7915
+    },
+    {
+      "epoch": 1.3453286752255824,
+      "grad_norm": 1.6171875,
+      "learning_rate": 1.1586101905223442e-05,
+      "loss": 0.7674,
+      "step": 7916
+    },
+    {
+      "epoch": 1.3455001393145722,
+      "grad_norm": 1.7109375,
+      "learning_rate": 1.1584318811928752e-05,
+      "loss": 0.8283,
+      "step": 7917
+    },
+    {
+      "epoch": 1.345671603403562,
+      "grad_norm": 1.7421875,
+      "learning_rate": 1.1582535666963462e-05,
+      "loss": 0.8261,
+      "step": 7918
+    },
+    {
+      "epoch": 1.345843067492552,
+      "grad_norm": 1.625,
+      "learning_rate": 1.1580752470385725e-05,
+      "loss": 0.9092,
+      "step": 7919
+    },
+    {
+      "epoch": 1.346014531581542,
+      "grad_norm": 1.7734375,
+      "learning_rate": 1.1578969222253696e-05,
+      "loss": 0.8821,
+      "step": 7920
+    },
+    {
+      "epoch": 1.3461859956705318,
+      "grad_norm": 1.6953125,
+      "learning_rate": 1.1577185922625536e-05,
+      "loss": 0.8901,
+      "step": 7921
+    },
+    {
+      "epoch": 1.3463574597595216,
+      "grad_norm": 1.6640625,
+      "learning_rate": 1.1575402571559403e-05,
+      "loss": 0.8937,
+      "step": 7922
+    },
+    {
+      "epoch": 1.3465289238485114,
+      "grad_norm": 1.8046875,
+      "learning_rate": 1.157361916911346e-05,
+      "loss": 0.8723,
+      "step": 7923
+    },
+    {
+      "epoch": 1.3467003879375015,
+      "grad_norm": 1.6796875,
+      "learning_rate": 1.1571835715345869e-05,
+      "loss": 0.885,
+      "step": 7924
+    },
+    {
+      "epoch": 1.3468718520264913,
+      "grad_norm": 1.6640625,
+      "learning_rate": 1.1570052210314798e-05,
+      "loss": 0.8569,
+      "step": 7925
+    },
+    {
+      "epoch": 1.3470433161154811,
+      "grad_norm": 1.6953125,
+      "learning_rate": 1.1568268654078411e-05,
+      "loss": 0.9641,
+      "step": 7926
+    },
+    {
+      "epoch": 1.347214780204471,
+      "grad_norm": 1.65625,
+      "learning_rate": 1.1566485046694877e-05,
+      "loss": 0.8032,
+      "step": 7927
+    },
+    {
+      "epoch": 1.3473862442934608,
+      "grad_norm": 1.5859375,
+      "learning_rate": 1.156470138822237e-05,
+      "loss": 0.8226,
+      "step": 7928
+    },
+    {
+      "epoch": 1.3475577083824506,
+      "grad_norm": 1.640625,
+      "learning_rate": 1.1562917678719056e-05,
+      "loss": 0.8646,
+      "step": 7929
+    },
+    {
+      "epoch": 1.3477291724714404,
+      "grad_norm": 1.7109375,
+      "learning_rate": 1.1561133918243113e-05,
+      "loss": 0.8501,
+      "step": 7930
+    },
+    {
+      "epoch": 1.3479006365604302,
+      "grad_norm": 1.6640625,
+      "learning_rate": 1.1559350106852713e-05,
+      "loss": 0.735,
+      "step": 7931
+    },
+    {
+      "epoch": 1.3480721006494203,
+      "grad_norm": 1.640625,
+      "learning_rate": 1.1557566244606035e-05,
+      "loss": 0.9369,
+      "step": 7932
+    },
+    {
+      "epoch": 1.3482435647384101,
+      "grad_norm": 1.765625,
+      "learning_rate": 1.1555782331561253e-05,
+      "loss": 0.9249,
+      "step": 7933
+    },
+    {
+      "epoch": 1.3484150288274,
+      "grad_norm": 1.671875,
+      "learning_rate": 1.1553998367776555e-05,
+      "loss": 0.9296,
+      "step": 7934
+    },
+    {
+      "epoch": 1.3485864929163898,
+      "grad_norm": 1.65625,
+      "learning_rate": 1.1552214353310117e-05,
+      "loss": 0.8557,
+      "step": 7935
+    },
+    {
+      "epoch": 1.3487579570053798,
+      "grad_norm": 1.6015625,
+      "learning_rate": 1.1550430288220122e-05,
+      "loss": 0.8573,
+      "step": 7936
+    },
+    {
+      "epoch": 1.3489294210943696,
+      "grad_norm": 1.734375,
+      "learning_rate": 1.154864617256476e-05,
+      "loss": 0.832,
+      "step": 7937
+    },
+    {
+      "epoch": 1.3491008851833595,
+      "grad_norm": 1.703125,
+      "learning_rate": 1.1546862006402211e-05,
+      "loss": 0.8838,
+      "step": 7938
+    },
+    {
+      "epoch": 1.3492723492723493,
+      "grad_norm": 1.65625,
+      "learning_rate": 1.1545077789790668e-05,
+      "loss": 0.8245,
+      "step": 7939
+    },
+    {
+      "epoch": 1.3494438133613391,
+      "grad_norm": 1.7421875,
+      "learning_rate": 1.1543293522788321e-05,
+      "loss": 0.8864,
+      "step": 7940
+    },
+    {
+      "epoch": 1.349615277450329,
+      "grad_norm": 1.703125,
+      "learning_rate": 1.1541509205453363e-05,
+      "loss": 0.9193,
+      "step": 7941
+    },
+    {
+      "epoch": 1.3497867415393188,
+      "grad_norm": 1.75,
+      "learning_rate": 1.1539724837843984e-05,
+      "loss": 0.8673,
+      "step": 7942
+    },
+    {
+      "epoch": 1.3499582056283086,
+      "grad_norm": 1.6875,
+      "learning_rate": 1.1537940420018378e-05,
+      "loss": 0.8974,
+      "step": 7943
+    },
+    {
+      "epoch": 1.3501296697172986,
+      "grad_norm": 1.6640625,
+      "learning_rate": 1.1536155952034743e-05,
+      "loss": 0.8876,
+      "step": 7944
+    },
+    {
+      "epoch": 1.3503011338062885,
+      "grad_norm": 1.671875,
+      "learning_rate": 1.1534371433951279e-05,
+      "loss": 0.8667,
+      "step": 7945
+    },
+    {
+      "epoch": 1.3504725978952783,
+      "grad_norm": 1.7578125,
+      "learning_rate": 1.1532586865826185e-05,
+      "loss": 0.9604,
+      "step": 7946
+    },
+    {
+      "epoch": 1.3506440619842681,
+      "grad_norm": 1.546875,
+      "learning_rate": 1.153080224771766e-05,
+      "loss": 0.7985,
+      "step": 7947
+    },
+    {
+      "epoch": 1.3508155260732582,
+      "grad_norm": 1.671875,
+      "learning_rate": 1.1529017579683915e-05,
+      "loss": 0.8312,
+      "step": 7948
+    },
+    {
+      "epoch": 1.350986990162248,
+      "grad_norm": 1.6875,
+      "learning_rate": 1.1527232861783145e-05,
+      "loss": 0.9297,
+      "step": 7949
+    },
+    {
+      "epoch": 1.3511584542512378,
+      "grad_norm": 1.7265625,
+      "learning_rate": 1.1525448094073559e-05,
+      "loss": 0.8972,
+      "step": 7950
+    },
+    {
+      "epoch": 1.3513299183402276,
+      "grad_norm": 1.7109375,
+      "learning_rate": 1.1523663276613368e-05,
+      "loss": 0.8746,
+      "step": 7951
+    },
+    {
+      "epoch": 1.3515013824292175,
+      "grad_norm": 1.65625,
+      "learning_rate": 1.152187840946078e-05,
+      "loss": 0.8506,
+      "step": 7952
+    },
+    {
+      "epoch": 1.3516728465182073,
+      "grad_norm": 1.671875,
+      "learning_rate": 1.1520093492674004e-05,
+      "loss": 0.9189,
+      "step": 7953
+    },
+    {
+      "epoch": 1.3518443106071971,
+      "grad_norm": 1.71875,
+      "learning_rate": 1.1518308526311257e-05,
+      "loss": 0.829,
+      "step": 7954
+    },
+    {
+      "epoch": 1.352015774696187,
+      "grad_norm": 1.6875,
+      "learning_rate": 1.1516523510430753e-05,
+      "loss": 0.8295,
+      "step": 7955
+    },
+    {
+      "epoch": 1.352187238785177,
+      "grad_norm": 1.6796875,
+      "learning_rate": 1.1514738445090705e-05,
+      "loss": 0.8435,
+      "step": 7956
+    },
+    {
+      "epoch": 1.3523587028741668,
+      "grad_norm": 1.6015625,
+      "learning_rate": 1.1512953330349332e-05,
+      "loss": 0.8204,
+      "step": 7957
+    },
+    {
+      "epoch": 1.3525301669631566,
+      "grad_norm": 1.6953125,
+      "learning_rate": 1.1511168166264854e-05,
+      "loss": 0.8189,
+      "step": 7958
+    },
+    {
+      "epoch": 1.3527016310521465,
+      "grad_norm": 1.6640625,
+      "learning_rate": 1.1509382952895494e-05,
+      "loss": 0.794,
+      "step": 7959
+    },
+    {
+      "epoch": 1.3528730951411365,
+      "grad_norm": 1.6328125,
+      "learning_rate": 1.150759769029947e-05,
+      "loss": 0.7831,
+      "step": 7960
+    },
+    {
+      "epoch": 1.3530445592301263,
+      "grad_norm": 1.734375,
+      "learning_rate": 1.1505812378535008e-05,
+      "loss": 0.8546,
+      "step": 7961
+    },
+    {
+      "epoch": 1.3532160233191162,
+      "grad_norm": 1.671875,
+      "learning_rate": 1.1504027017660333e-05,
+      "loss": 0.9,
+      "step": 7962
+    },
+    {
+      "epoch": 1.353387487408106,
+      "grad_norm": 1.5703125,
+      "learning_rate": 1.1502241607733675e-05,
+      "loss": 0.8458,
+      "step": 7963
+    },
+    {
+      "epoch": 1.3535589514970958,
+      "grad_norm": 1.75,
+      "learning_rate": 1.1500456148813258e-05,
+      "loss": 0.8764,
+      "step": 7964
+    },
+    {
+      "epoch": 1.3537304155860856,
+      "grad_norm": 1.71875,
+      "learning_rate": 1.1498670640957319e-05,
+      "loss": 0.8472,
+      "step": 7965
+    },
+    {
+      "epoch": 1.3539018796750755,
+      "grad_norm": 1.6640625,
+      "learning_rate": 1.1496885084224088e-05,
+      "loss": 0.7855,
+      "step": 7966
+    },
+    {
+      "epoch": 1.3540733437640653,
+      "grad_norm": 1.765625,
+      "learning_rate": 1.1495099478671797e-05,
+      "loss": 0.962,
+      "step": 7967
+    },
+    {
+      "epoch": 1.3542448078530553,
+      "grad_norm": 1.625,
+      "learning_rate": 1.1493313824358678e-05,
+      "loss": 0.8217,
+      "step": 7968
+    },
+    {
+      "epoch": 1.3544162719420452,
+      "grad_norm": 1.6328125,
+      "learning_rate": 1.1491528121342977e-05,
+      "loss": 0.846,
+      "step": 7969
+    },
+    {
+      "epoch": 1.354587736031035,
+      "grad_norm": 1.6640625,
+      "learning_rate": 1.1489742369682923e-05,
+      "loss": 0.8412,
+      "step": 7970
+    },
+    {
+      "epoch": 1.3547592001200248,
+      "grad_norm": 1.671875,
+      "learning_rate": 1.1487956569436766e-05,
+      "loss": 0.9254,
+      "step": 7971
+    },
+    {
+      "epoch": 1.3549306642090146,
+      "grad_norm": 1.7734375,
+      "learning_rate": 1.1486170720662736e-05,
+      "loss": 0.8846,
+      "step": 7972
+    },
+    {
+      "epoch": 1.3551021282980047,
+      "grad_norm": 1.9296875,
+      "learning_rate": 1.1484384823419086e-05,
+      "loss": 0.8607,
+      "step": 7973
+    },
+    {
+      "epoch": 1.3552735923869945,
+      "grad_norm": 1.78125,
+      "learning_rate": 1.1482598877764055e-05,
+      "loss": 0.8201,
+      "step": 7974
+    },
+    {
+      "epoch": 1.3554450564759843,
+      "grad_norm": 1.6640625,
+      "learning_rate": 1.1480812883755894e-05,
+      "loss": 0.8109,
+      "step": 7975
+    },
+    {
+      "epoch": 1.3556165205649742,
+      "grad_norm": 1.640625,
+      "learning_rate": 1.1479026841452848e-05,
+      "loss": 0.9103,
+      "step": 7976
+    },
+    {
+      "epoch": 1.355787984653964,
+      "grad_norm": 1.7109375,
+      "learning_rate": 1.1477240750913168e-05,
+      "loss": 0.8906,
+      "step": 7977
+    },
+    {
+      "epoch": 1.3559594487429538,
+      "grad_norm": 1.734375,
+      "learning_rate": 1.1475454612195102e-05,
+      "loss": 0.8033,
+      "step": 7978
+    },
+    {
+      "epoch": 1.3561309128319436,
+      "grad_norm": 1.6953125,
+      "learning_rate": 1.1473668425356908e-05,
+      "loss": 0.8839,
+      "step": 7979
+    },
+    {
+      "epoch": 1.3563023769209337,
+      "grad_norm": 1.5625,
+      "learning_rate": 1.1471882190456836e-05,
+      "loss": 0.7843,
+      "step": 7980
+    },
+    {
+      "epoch": 1.3564738410099235,
+      "grad_norm": 1.7109375,
+      "learning_rate": 1.1470095907553143e-05,
+      "loss": 0.8963,
+      "step": 7981
+    },
+    {
+      "epoch": 1.3566453050989133,
+      "grad_norm": 1.734375,
+      "learning_rate": 1.1468309576704084e-05,
+      "loss": 0.9226,
+      "step": 7982
+    },
+    {
+      "epoch": 1.3568167691879032,
+      "grad_norm": 1.6484375,
+      "learning_rate": 1.1466523197967923e-05,
+      "loss": 0.8045,
+      "step": 7983
+    },
+    {
+      "epoch": 1.356988233276893,
+      "grad_norm": 1.6875,
+      "learning_rate": 1.1464736771402918e-05,
+      "loss": 0.8812,
+      "step": 7984
+    },
+    {
+      "epoch": 1.357159697365883,
+      "grad_norm": 1.6875,
+      "learning_rate": 1.1462950297067333e-05,
+      "loss": 0.8684,
+      "step": 7985
+    },
+    {
+      "epoch": 1.3573311614548729,
+      "grad_norm": 1.7890625,
+      "learning_rate": 1.1461163775019426e-05,
+      "loss": 0.9137,
+      "step": 7986
+    },
+    {
+      "epoch": 1.3575026255438627,
+      "grad_norm": 1.65625,
+      "learning_rate": 1.1459377205317467e-05,
+      "loss": 0.8363,
+      "step": 7987
+    },
+    {
+      "epoch": 1.3576740896328525,
+      "grad_norm": 1.703125,
+      "learning_rate": 1.1457590588019726e-05,
+      "loss": 0.8861,
+      "step": 7988
+    },
+    {
+      "epoch": 1.3578455537218423,
+      "grad_norm": 1.6015625,
+      "learning_rate": 1.1455803923184465e-05,
+      "loss": 0.8179,
+      "step": 7989
+    },
+    {
+      "epoch": 1.3580170178108322,
+      "grad_norm": 1.671875,
+      "learning_rate": 1.1454017210869954e-05,
+      "loss": 0.8174,
+      "step": 7990
+    },
+    {
+      "epoch": 1.358188481899822,
+      "grad_norm": 1.671875,
+      "learning_rate": 1.1452230451134465e-05,
+      "loss": 0.8978,
+      "step": 7991
+    },
+    {
+      "epoch": 1.358359945988812,
+      "grad_norm": 1.640625,
+      "learning_rate": 1.1450443644036276e-05,
+      "loss": 0.8766,
+      "step": 7992
+    },
+    {
+      "epoch": 1.3585314100778019,
+      "grad_norm": 1.734375,
+      "learning_rate": 1.1448656789633657e-05,
+      "loss": 0.9489,
+      "step": 7993
+    },
+    {
+      "epoch": 1.3587028741667917,
+      "grad_norm": 1.71875,
+      "learning_rate": 1.1446869887984887e-05,
+      "loss": 0.9773,
+      "step": 7994
+    },
+    {
+      "epoch": 1.3588743382557815,
+      "grad_norm": 1.734375,
+      "learning_rate": 1.144508293914824e-05,
+      "loss": 0.9206,
+      "step": 7995
+    },
+    {
+      "epoch": 1.3590458023447713,
+      "grad_norm": 1.859375,
+      "learning_rate": 1.1443295943181996e-05,
+      "loss": 0.8688,
+      "step": 7996
+    },
+    {
+      "epoch": 1.3592172664337614,
+      "grad_norm": 1.65625,
+      "learning_rate": 1.1441508900144436e-05,
+      "loss": 0.8146,
+      "step": 7997
+    },
+    {
+      "epoch": 1.3593887305227512,
+      "grad_norm": 1.6953125,
+      "learning_rate": 1.1439721810093843e-05,
+      "loss": 0.878,
+      "step": 7998
+    },
+    {
+      "epoch": 1.359560194611741,
+      "grad_norm": 1.7578125,
+      "learning_rate": 1.14379346730885e-05,
+      "loss": 0.8769,
+      "step": 7999
+    },
+    {
+      "epoch": 1.3597316587007309,
+      "grad_norm": 1.7421875,
+      "learning_rate": 1.1436147489186693e-05,
+      "loss": 0.8759,
+      "step": 8000
+    },
+    {
+      "epoch": 1.3599031227897207,
+      "grad_norm": 1.640625,
+      "learning_rate": 1.1434360258446705e-05,
+      "loss": 0.8362,
+      "step": 8001
+    },
+    {
+      "epoch": 1.3600745868787105,
+      "grad_norm": 1.7265625,
+      "learning_rate": 1.1432572980926833e-05,
+      "loss": 0.8737,
+      "step": 8002
+    },
+    {
+      "epoch": 1.3602460509677003,
+      "grad_norm": 1.6171875,
+      "learning_rate": 1.1430785656685358e-05,
+      "loss": 0.8076,
+      "step": 8003
+    },
+    {
+      "epoch": 1.3604175150566904,
+      "grad_norm": 1.609375,
+      "learning_rate": 1.1428998285780576e-05,
+      "loss": 0.8479,
+      "step": 8004
+    },
+    {
+      "epoch": 1.3605889791456802,
+      "grad_norm": 1.6796875,
+      "learning_rate": 1.142721086827078e-05,
+      "loss": 0.8055,
+      "step": 8005
+    },
+    {
+      "epoch": 1.36076044323467,
+      "grad_norm": 1.7421875,
+      "learning_rate": 1.1425423404214263e-05,
+      "loss": 0.847,
+      "step": 8006
+    },
+    {
+      "epoch": 1.3609319073236599,
+      "grad_norm": 1.765625,
+      "learning_rate": 1.142363589366932e-05,
+      "loss": 0.8638,
+      "step": 8007
+    },
+    {
+      "epoch": 1.3611033714126497,
+      "grad_norm": 1.6640625,
+      "learning_rate": 1.142184833669425e-05,
+      "loss": 0.859,
+      "step": 8008
+    },
+    {
+      "epoch": 1.3612748355016397,
+      "grad_norm": 1.703125,
+      "learning_rate": 1.1420060733347353e-05,
+      "loss": 0.7812,
+      "step": 8009
+    },
+    {
+      "epoch": 1.3614462995906296,
+      "grad_norm": 1.8359375,
+      "learning_rate": 1.1418273083686926e-05,
+      "loss": 0.8876,
+      "step": 8010
+    },
+    {
+      "epoch": 1.3616177636796194,
+      "grad_norm": 1.71875,
+      "learning_rate": 1.1416485387771274e-05,
+      "loss": 0.8667,
+      "step": 8011
+    },
+    {
+      "epoch": 1.3617892277686092,
+      "grad_norm": 1.6484375,
+      "learning_rate": 1.1414697645658699e-05,
+      "loss": 0.8258,
+      "step": 8012
+    },
+    {
+      "epoch": 1.361960691857599,
+      "grad_norm": 1.7578125,
+      "learning_rate": 1.1412909857407505e-05,
+      "loss": 0.8949,
+      "step": 8013
+    },
+    {
+      "epoch": 1.3621321559465889,
+      "grad_norm": 1.6640625,
+      "learning_rate": 1.1411122023076002e-05,
+      "loss": 0.8998,
+      "step": 8014
+    },
+    {
+      "epoch": 1.3623036200355787,
+      "grad_norm": 1.59375,
+      "learning_rate": 1.1409334142722494e-05,
+      "loss": 0.8753,
+      "step": 8015
+    },
+    {
+      "epoch": 1.3624750841245685,
+      "grad_norm": 1.671875,
+      "learning_rate": 1.1407546216405296e-05,
+      "loss": 0.8763,
+      "step": 8016
+    },
+    {
+      "epoch": 1.3626465482135586,
+      "grad_norm": 1.8671875,
+      "learning_rate": 1.1405758244182716e-05,
+      "loss": 0.9637,
+      "step": 8017
+    },
+    {
+      "epoch": 1.3628180123025484,
+      "grad_norm": 1.7734375,
+      "learning_rate": 1.1403970226113064e-05,
+      "loss": 0.9156,
+      "step": 8018
+    },
+    {
+      "epoch": 1.3629894763915382,
+      "grad_norm": 1.8125,
+      "learning_rate": 1.1402182162254653e-05,
+      "loss": 0.8782,
+      "step": 8019
+    },
+    {
+      "epoch": 1.363160940480528,
+      "grad_norm": 1.609375,
+      "learning_rate": 1.1400394052665804e-05,
+      "loss": 0.7742,
+      "step": 8020
+    },
+    {
+      "epoch": 1.363332404569518,
+      "grad_norm": 1.765625,
+      "learning_rate": 1.1398605897404833e-05,
+      "loss": 0.9132,
+      "step": 8021
+    },
+    {
+      "epoch": 1.363503868658508,
+      "grad_norm": 1.703125,
+      "learning_rate": 1.1396817696530055e-05,
+      "loss": 0.8553,
+      "step": 8022
+    },
+    {
+      "epoch": 1.3636753327474977,
+      "grad_norm": 1.640625,
+      "learning_rate": 1.1395029450099794e-05,
+      "loss": 0.8345,
+      "step": 8023
+    },
+    {
+      "epoch": 1.3638467968364876,
+      "grad_norm": 1.6953125,
+      "learning_rate": 1.139324115817237e-05,
+      "loss": 0.8422,
+      "step": 8024
+    },
+    {
+      "epoch": 1.3640182609254774,
+      "grad_norm": 1.7421875,
+      "learning_rate": 1.1391452820806103e-05,
+      "loss": 0.9355,
+      "step": 8025
+    },
+    {
+      "epoch": 1.3641897250144672,
+      "grad_norm": 1.7734375,
+      "learning_rate": 1.138966443805932e-05,
+      "loss": 0.9624,
+      "step": 8026
+    },
+    {
+      "epoch": 1.364361189103457,
+      "grad_norm": 1.7109375,
+      "learning_rate": 1.1387876009990348e-05,
+      "loss": 0.8817,
+      "step": 8027
+    },
+    {
+      "epoch": 1.3645326531924469,
+      "grad_norm": 1.7109375,
+      "learning_rate": 1.1386087536657513e-05,
+      "loss": 0.7908,
+      "step": 8028
+    },
+    {
+      "epoch": 1.364704117281437,
+      "grad_norm": 1.703125,
+      "learning_rate": 1.1384299018119142e-05,
+      "loss": 0.862,
+      "step": 8029
+    },
+    {
+      "epoch": 1.3648755813704267,
+      "grad_norm": 1.703125,
+      "learning_rate": 1.1382510454433568e-05,
+      "loss": 0.8278,
+      "step": 8030
+    },
+    {
+      "epoch": 1.3650470454594166,
+      "grad_norm": 1.734375,
+      "learning_rate": 1.1380721845659123e-05,
+      "loss": 0.8649,
+      "step": 8031
+    },
+    {
+      "epoch": 1.3652185095484064,
+      "grad_norm": 1.65625,
+      "learning_rate": 1.1378933191854137e-05,
+      "loss": 0.8856,
+      "step": 8032
+    },
+    {
+      "epoch": 1.3653899736373964,
+      "grad_norm": 1.6796875,
+      "learning_rate": 1.137714449307695e-05,
+      "loss": 0.9367,
+      "step": 8033
+    },
+    {
+      "epoch": 1.3655614377263863,
+      "grad_norm": 1.5703125,
+      "learning_rate": 1.1375355749385896e-05,
+      "loss": 0.7769,
+      "step": 8034
+    },
+    {
+      "epoch": 1.365732901815376,
+      "grad_norm": 1.578125,
+      "learning_rate": 1.137356696083931e-05,
+      "loss": 0.8368,
+      "step": 8035
+    },
+    {
+      "epoch": 1.365904365904366,
+      "grad_norm": 1.6875,
+      "learning_rate": 1.1371778127495535e-05,
+      "loss": 0.8367,
+      "step": 8036
+    },
+    {
+      "epoch": 1.3660758299933558,
+      "grad_norm": 1.6640625,
+      "learning_rate": 1.1369989249412905e-05,
+      "loss": 0.8095,
+      "step": 8037
+    },
+    {
+      "epoch": 1.3662472940823456,
+      "grad_norm": 1.625,
+      "learning_rate": 1.1368200326649768e-05,
+      "loss": 0.8846,
+      "step": 8038
+    },
+    {
+      "epoch": 1.3664187581713354,
+      "grad_norm": 1.7109375,
+      "learning_rate": 1.1366411359264466e-05,
+      "loss": 0.9167,
+      "step": 8039
+    },
+    {
+      "epoch": 1.3665902222603252,
+      "grad_norm": 1.6484375,
+      "learning_rate": 1.1364622347315348e-05,
+      "loss": 0.853,
+      "step": 8040
+    },
+    {
+      "epoch": 1.3667616863493153,
+      "grad_norm": 1.625,
+      "learning_rate": 1.1362833290860756e-05,
+      "loss": 0.8976,
+      "step": 8041
+    },
+    {
+      "epoch": 1.366933150438305,
+      "grad_norm": 1.6796875,
+      "learning_rate": 1.1361044189959035e-05,
+      "loss": 0.8774,
+      "step": 8042
+    },
+    {
+      "epoch": 1.367104614527295,
+      "grad_norm": 1.7265625,
+      "learning_rate": 1.135925504466854e-05,
+      "loss": 0.8337,
+      "step": 8043
+    },
+    {
+      "epoch": 1.3672760786162848,
+      "grad_norm": 1.7109375,
+      "learning_rate": 1.1357465855047617e-05,
+      "loss": 0.9059,
+      "step": 8044
+    },
+    {
+      "epoch": 1.3674475427052748,
+      "grad_norm": 1.8203125,
+      "learning_rate": 1.1355676621154624e-05,
+      "loss": 0.8321,
+      "step": 8045
+    },
+    {
+      "epoch": 1.3676190067942646,
+      "grad_norm": 1.7421875,
+      "learning_rate": 1.135388734304791e-05,
+      "loss": 0.8401,
+      "step": 8046
+    },
+    {
+      "epoch": 1.3677904708832545,
+      "grad_norm": 1.6484375,
+      "learning_rate": 1.1352098020785832e-05,
+      "loss": 0.8627,
+      "step": 8047
+    },
+    {
+      "epoch": 1.3679619349722443,
+      "grad_norm": 1.7265625,
+      "learning_rate": 1.1350308654426744e-05,
+      "loss": 0.8864,
+      "step": 8048
+    },
+    {
+      "epoch": 1.368133399061234,
+      "grad_norm": 1.796875,
+      "learning_rate": 1.1348519244029008e-05,
+      "loss": 0.8816,
+      "step": 8049
+    },
+    {
+      "epoch": 1.368304863150224,
+      "grad_norm": 1.734375,
+      "learning_rate": 1.1346729789650982e-05,
+      "loss": 0.8284,
+      "step": 8050
+    },
+    {
+      "epoch": 1.3684763272392138,
+      "grad_norm": 1.6875,
+      "learning_rate": 1.1344940291351022e-05,
+      "loss": 0.8998,
+      "step": 8051
+    },
+    {
+      "epoch": 1.3686477913282036,
+      "grad_norm": 1.59375,
+      "learning_rate": 1.1343150749187501e-05,
+      "loss": 0.8313,
+      "step": 8052
+    },
+    {
+      "epoch": 1.3688192554171936,
+      "grad_norm": 1.6953125,
+      "learning_rate": 1.1341361163218771e-05,
+      "loss": 0.9093,
+      "step": 8053
+    },
+    {
+      "epoch": 1.3689907195061835,
+      "grad_norm": 1.734375,
+      "learning_rate": 1.1339571533503207e-05,
+      "loss": 0.9283,
+      "step": 8054
+    },
+    {
+      "epoch": 1.3691621835951733,
+      "grad_norm": 1.7265625,
+      "learning_rate": 1.1337781860099165e-05,
+      "loss": 0.957,
+      "step": 8055
+    },
+    {
+      "epoch": 1.369333647684163,
+      "grad_norm": 1.6640625,
+      "learning_rate": 1.1335992143065023e-05,
+      "loss": 0.8692,
+      "step": 8056
+    },
+    {
+      "epoch": 1.369505111773153,
+      "grad_norm": 1.7109375,
+      "learning_rate": 1.1334202382459145e-05,
+      "loss": 0.8551,
+      "step": 8057
+    },
+    {
+      "epoch": 1.369676575862143,
+      "grad_norm": 1.703125,
+      "learning_rate": 1.1332412578339905e-05,
+      "loss": 0.8644,
+      "step": 8058
+    },
+    {
+      "epoch": 1.3698480399511328,
+      "grad_norm": 1.640625,
+      "learning_rate": 1.133062273076567e-05,
+      "loss": 0.848,
+      "step": 8059
+    },
+    {
+      "epoch": 1.3700195040401226,
+      "grad_norm": 1.75,
+      "learning_rate": 1.1328832839794819e-05,
+      "loss": 0.7959,
+      "step": 8060
+    },
+    {
+      "epoch": 1.3701909681291125,
+      "grad_norm": 1.6796875,
+      "learning_rate": 1.1327042905485725e-05,
+      "loss": 0.8229,
+      "step": 8061
+    },
+    {
+      "epoch": 1.3703624322181023,
+      "grad_norm": 1.7265625,
+      "learning_rate": 1.1325252927896765e-05,
+      "loss": 0.8541,
+      "step": 8062
+    },
+    {
+      "epoch": 1.370533896307092,
+      "grad_norm": 1.734375,
+      "learning_rate": 1.132346290708632e-05,
+      "loss": 0.8699,
+      "step": 8063
+    },
+    {
+      "epoch": 1.370705360396082,
+      "grad_norm": 1.734375,
+      "learning_rate": 1.132167284311276e-05,
+      "loss": 0.8418,
+      "step": 8064
+    },
+    {
+      "epoch": 1.370876824485072,
+      "grad_norm": 1.765625,
+      "learning_rate": 1.1319882736034476e-05,
+      "loss": 0.8203,
+      "step": 8065
+    },
+    {
+      "epoch": 1.3710482885740618,
+      "grad_norm": 1.5703125,
+      "learning_rate": 1.1318092585909841e-05,
+      "loss": 0.8273,
+      "step": 8066
+    },
+    {
+      "epoch": 1.3712197526630516,
+      "grad_norm": 1.640625,
+      "learning_rate": 1.1316302392797244e-05,
+      "loss": 0.8852,
+      "step": 8067
+    },
+    {
+      "epoch": 1.3713912167520415,
+      "grad_norm": 1.6953125,
+      "learning_rate": 1.1314512156755073e-05,
+      "loss": 0.8474,
+      "step": 8068
+    },
+    {
+      "epoch": 1.3715626808410313,
+      "grad_norm": 1.6640625,
+      "learning_rate": 1.131272187784171e-05,
+      "loss": 0.7935,
+      "step": 8069
+    },
+    {
+      "epoch": 1.3717341449300213,
+      "grad_norm": 1.765625,
+      "learning_rate": 1.1310931556115543e-05,
+      "loss": 0.9003,
+      "step": 8070
+    },
+    {
+      "epoch": 1.3719056090190112,
+      "grad_norm": 1.609375,
+      "learning_rate": 1.130914119163496e-05,
+      "loss": 0.8306,
+      "step": 8071
+    },
+    {
+      "epoch": 1.372077073108001,
+      "grad_norm": 1.703125,
+      "learning_rate": 1.1307350784458355e-05,
+      "loss": 0.9435,
+      "step": 8072
+    },
+    {
+      "epoch": 1.3722485371969908,
+      "grad_norm": 1.7421875,
+      "learning_rate": 1.1305560334644116e-05,
+      "loss": 0.9552,
+      "step": 8073
+    },
+    {
+      "epoch": 1.3724200012859806,
+      "grad_norm": 1.7109375,
+      "learning_rate": 1.1303769842250638e-05,
+      "loss": 0.8818,
+      "step": 8074
+    },
+    {
+      "epoch": 1.3725914653749705,
+      "grad_norm": 1.65625,
+      "learning_rate": 1.1301979307336317e-05,
+      "loss": 0.8798,
+      "step": 8075
+    },
+    {
+      "epoch": 1.3727629294639603,
+      "grad_norm": 1.71875,
+      "learning_rate": 1.1300188729959548e-05,
+      "loss": 0.8477,
+      "step": 8076
+    },
+    {
+      "epoch": 1.3729343935529503,
+      "grad_norm": 1.640625,
+      "learning_rate": 1.1298398110178728e-05,
+      "loss": 0.8337,
+      "step": 8077
+    },
+    {
+      "epoch": 1.3731058576419402,
+      "grad_norm": 1.6875,
+      "learning_rate": 1.1296607448052255e-05,
+      "loss": 0.865,
+      "step": 8078
+    },
+    {
+      "epoch": 1.37327732173093,
+      "grad_norm": 1.640625,
+      "learning_rate": 1.1294816743638533e-05,
+      "loss": 0.8704,
+      "step": 8079
+    },
+    {
+      "epoch": 1.3734487858199198,
+      "grad_norm": 1.65625,
+      "learning_rate": 1.1293025996995962e-05,
+      "loss": 0.8725,
+      "step": 8080
+    },
+    {
+      "epoch": 1.3736202499089096,
+      "grad_norm": 1.796875,
+      "learning_rate": 1.1291235208182942e-05,
+      "loss": 0.898,
+      "step": 8081
+    },
+    {
+      "epoch": 1.3737917139978997,
+      "grad_norm": 1.640625,
+      "learning_rate": 1.128944437725788e-05,
+      "loss": 0.814,
+      "step": 8082
+    },
+    {
+      "epoch": 1.3739631780868895,
+      "grad_norm": 1.6015625,
+      "learning_rate": 1.128765350427918e-05,
+      "loss": 0.9043,
+      "step": 8083
+    },
+    {
+      "epoch": 1.3741346421758793,
+      "grad_norm": 1.6171875,
+      "learning_rate": 1.1285862589305252e-05,
+      "loss": 0.8093,
+      "step": 8084
+    },
+    {
+      "epoch": 1.3743061062648692,
+      "grad_norm": 1.6796875,
+      "learning_rate": 1.12840716323945e-05,
+      "loss": 0.9416,
+      "step": 8085
+    },
+    {
+      "epoch": 1.374477570353859,
+      "grad_norm": 1.6484375,
+      "learning_rate": 1.128228063360534e-05,
+      "loss": 0.8444,
+      "step": 8086
+    },
+    {
+      "epoch": 1.3746490344428488,
+      "grad_norm": 1.7109375,
+      "learning_rate": 1.1280489592996177e-05,
+      "loss": 0.817,
+      "step": 8087
+    },
+    {
+      "epoch": 1.3748204985318386,
+      "grad_norm": 1.6953125,
+      "learning_rate": 1.127869851062543e-05,
+      "loss": 0.8625,
+      "step": 8088
+    },
+    {
+      "epoch": 1.3749919626208287,
+      "grad_norm": 1.7734375,
+      "learning_rate": 1.1276907386551508e-05,
+      "loss": 0.8986,
+      "step": 8089
+    },
+    {
+      "epoch": 1.3751634267098185,
+      "grad_norm": 1.6875,
+      "learning_rate": 1.1275116220832826e-05,
+      "loss": 0.8341,
+      "step": 8090
+    },
+    {
+      "epoch": 1.3753348907988083,
+      "grad_norm": 1.6171875,
+      "learning_rate": 1.1273325013527805e-05,
+      "loss": 0.8414,
+      "step": 8091
+    },
+    {
+      "epoch": 1.3755063548877982,
+      "grad_norm": 1.6796875,
+      "learning_rate": 1.1271533764694862e-05,
+      "loss": 0.8472,
+      "step": 8092
+    },
+    {
+      "epoch": 1.375677818976788,
+      "grad_norm": 1.6875,
+      "learning_rate": 1.126974247439241e-05,
+      "loss": 0.7282,
+      "step": 8093
+    },
+    {
+      "epoch": 1.375849283065778,
+      "grad_norm": 1.6484375,
+      "learning_rate": 1.1267951142678877e-05,
+      "loss": 0.8236,
+      "step": 8094
+    },
+    {
+      "epoch": 1.3760207471547679,
+      "grad_norm": 1.671875,
+      "learning_rate": 1.126615976961268e-05,
+      "loss": 0.8156,
+      "step": 8095
+    },
+    {
+      "epoch": 1.3761922112437577,
+      "grad_norm": 1.734375,
+      "learning_rate": 1.1264368355252248e-05,
+      "loss": 0.8662,
+      "step": 8096
+    },
+    {
+      "epoch": 1.3763636753327475,
+      "grad_norm": 1.71875,
+      "learning_rate": 1.1262576899656001e-05,
+      "loss": 0.8275,
+      "step": 8097
+    },
+    {
+      "epoch": 1.3765351394217373,
+      "grad_norm": 1.6640625,
+      "learning_rate": 1.1260785402882371e-05,
+      "loss": 0.9154,
+      "step": 8098
+    },
+    {
+      "epoch": 1.3767066035107272,
+      "grad_norm": 1.703125,
+      "learning_rate": 1.1258993864989777e-05,
+      "loss": 0.8768,
+      "step": 8099
+    },
+    {
+      "epoch": 1.376878067599717,
+      "grad_norm": 1.8671875,
+      "learning_rate": 1.1257202286036653e-05,
+      "loss": 0.9332,
+      "step": 8100
+    },
+    {
+      "epoch": 1.377049531688707,
+      "grad_norm": 1.6328125,
+      "learning_rate": 1.125541066608143e-05,
+      "loss": 0.8837,
+      "step": 8101
+    },
+    {
+      "epoch": 1.3772209957776969,
+      "grad_norm": 1.6953125,
+      "learning_rate": 1.1253619005182534e-05,
+      "loss": 0.8344,
+      "step": 8102
+    },
+    {
+      "epoch": 1.3773924598666867,
+      "grad_norm": 1.7734375,
+      "learning_rate": 1.1251827303398406e-05,
+      "loss": 0.9364,
+      "step": 8103
+    },
+    {
+      "epoch": 1.3775639239556765,
+      "grad_norm": 1.703125,
+      "learning_rate": 1.1250035560787473e-05,
+      "loss": 0.9229,
+      "step": 8104
+    },
+    {
+      "epoch": 1.3777353880446663,
+      "grad_norm": 1.6484375,
+      "learning_rate": 1.1248243777408172e-05,
+      "loss": 0.8675,
+      "step": 8105
+    },
+    {
+      "epoch": 1.3779068521336564,
+      "grad_norm": 1.6328125,
+      "learning_rate": 1.1246451953318944e-05,
+      "loss": 0.8326,
+      "step": 8106
+    },
+    {
+      "epoch": 1.3780783162226462,
+      "grad_norm": 1.6953125,
+      "learning_rate": 1.1244660088578223e-05,
+      "loss": 0.9618,
+      "step": 8107
+    },
+    {
+      "epoch": 1.378249780311636,
+      "grad_norm": 1.7265625,
+      "learning_rate": 1.124286818324445e-05,
+      "loss": 0.8605,
+      "step": 8108
+    },
+    {
+      "epoch": 1.3784212444006259,
+      "grad_norm": 1.6953125,
+      "learning_rate": 1.1241076237376065e-05,
+      "loss": 0.8746,
+      "step": 8109
+    },
+    {
+      "epoch": 1.3785927084896157,
+      "grad_norm": 1.78125,
+      "learning_rate": 1.1239284251031511e-05,
+      "loss": 0.904,
+      "step": 8110
+    },
+    {
+      "epoch": 1.3787641725786055,
+      "grad_norm": 1.6953125,
+      "learning_rate": 1.1237492224269229e-05,
+      "loss": 0.892,
+      "step": 8111
+    },
+    {
+      "epoch": 1.3789356366675953,
+      "grad_norm": 1.71875,
+      "learning_rate": 1.1235700157147666e-05,
+      "loss": 0.8849,
+      "step": 8112
+    },
+    {
+      "epoch": 1.3791071007565852,
+      "grad_norm": 1.78125,
+      "learning_rate": 1.1233908049725267e-05,
+      "loss": 0.8826,
+      "step": 8113
+    },
+    {
+      "epoch": 1.3792785648455752,
+      "grad_norm": 1.6484375,
+      "learning_rate": 1.1232115902060481e-05,
+      "loss": 0.835,
+      "step": 8114
+    },
+    {
+      "epoch": 1.379450028934565,
+      "grad_norm": 1.6484375,
+      "learning_rate": 1.1230323714211757e-05,
+      "loss": 0.837,
+      "step": 8115
+    },
+    {
+      "epoch": 1.3796214930235549,
+      "grad_norm": 1.546875,
+      "learning_rate": 1.1228531486237545e-05,
+      "loss": 0.8524,
+      "step": 8116
+    },
+    {
+      "epoch": 1.3797929571125447,
+      "grad_norm": 1.65625,
+      "learning_rate": 1.1226739218196289e-05,
+      "loss": 0.7318,
+      "step": 8117
+    },
+    {
+      "epoch": 1.3799644212015347,
+      "grad_norm": 1.703125,
+      "learning_rate": 1.1224946910146452e-05,
+      "loss": 0.8689,
+      "step": 8118
+    },
+    {
+      "epoch": 1.3801358852905246,
+      "grad_norm": 1.640625,
+      "learning_rate": 1.1223154562146484e-05,
+      "loss": 0.841,
+      "step": 8119
+    },
+    {
+      "epoch": 1.3803073493795144,
+      "grad_norm": 1.7109375,
+      "learning_rate": 1.1221362174254837e-05,
+      "loss": 0.8204,
+      "step": 8120
+    },
+    {
+      "epoch": 1.3804788134685042,
+      "grad_norm": 1.7890625,
+      "learning_rate": 1.1219569746529973e-05,
+      "loss": 0.9147,
+      "step": 8121
+    },
+    {
+      "epoch": 1.380650277557494,
+      "grad_norm": 1.5859375,
+      "learning_rate": 1.1217777279030346e-05,
+      "loss": 0.785,
+      "step": 8122
+    },
+    {
+      "epoch": 1.3808217416464839,
+      "grad_norm": 1.734375,
+      "learning_rate": 1.1215984771814416e-05,
+      "loss": 0.8118,
+      "step": 8123
+    },
+    {
+      "epoch": 1.3809932057354737,
+      "grad_norm": 1.640625,
+      "learning_rate": 1.1214192224940643e-05,
+      "loss": 0.8752,
+      "step": 8124
+    },
+    {
+      "epoch": 1.3811646698244635,
+      "grad_norm": 1.640625,
+      "learning_rate": 1.1212399638467491e-05,
+      "loss": 0.7907,
+      "step": 8125
+    },
+    {
+      "epoch": 1.3813361339134536,
+      "grad_norm": 1.671875,
+      "learning_rate": 1.121060701245342e-05,
+      "loss": 0.8421,
+      "step": 8126
+    },
+    {
+      "epoch": 1.3815075980024434,
+      "grad_norm": 1.7109375,
+      "learning_rate": 1.1208814346956898e-05,
+      "loss": 0.849,
+      "step": 8127
+    },
+    {
+      "epoch": 1.3816790620914332,
+      "grad_norm": 1.6484375,
+      "learning_rate": 1.1207021642036387e-05,
+      "loss": 0.8065,
+      "step": 8128
+    },
+    {
+      "epoch": 1.381850526180423,
+      "grad_norm": 1.671875,
+      "learning_rate": 1.1205228897750353e-05,
+      "loss": 0.8164,
+      "step": 8129
+    },
+    {
+      "epoch": 1.382021990269413,
+      "grad_norm": 1.6640625,
+      "learning_rate": 1.1203436114157267e-05,
+      "loss": 0.8265,
+      "step": 8130
+    },
+    {
+      "epoch": 1.382193454358403,
+      "grad_norm": 1.6875,
+      "learning_rate": 1.1201643291315599e-05,
+      "loss": 0.8398,
+      "step": 8131
+    },
+    {
+      "epoch": 1.3823649184473927,
+      "grad_norm": 1.8359375,
+      "learning_rate": 1.1199850429283817e-05,
+      "loss": 0.833,
+      "step": 8132
+    },
+    {
+      "epoch": 1.3825363825363826,
+      "grad_norm": 1.84375,
+      "learning_rate": 1.1198057528120394e-05,
+      "loss": 0.946,
+      "step": 8133
+    },
+    {
+      "epoch": 1.3827078466253724,
+      "grad_norm": 1.671875,
+      "learning_rate": 1.1196264587883803e-05,
+      "loss": 0.8525,
+      "step": 8134
+    },
+    {
+      "epoch": 1.3828793107143622,
+      "grad_norm": 1.8046875,
+      "learning_rate": 1.119447160863252e-05,
+      "loss": 0.9194,
+      "step": 8135
+    },
+    {
+      "epoch": 1.383050774803352,
+      "grad_norm": 1.640625,
+      "learning_rate": 1.1192678590425021e-05,
+      "loss": 0.8202,
+      "step": 8136
+    },
+    {
+      "epoch": 1.3832222388923419,
+      "grad_norm": 1.71875,
+      "learning_rate": 1.119088553331978e-05,
+      "loss": 0.9056,
+      "step": 8137
+    },
+    {
+      "epoch": 1.383393702981332,
+      "grad_norm": 1.7265625,
+      "learning_rate": 1.1189092437375282e-05,
+      "loss": 0.8426,
+      "step": 8138
+    },
+    {
+      "epoch": 1.3835651670703217,
+      "grad_norm": 1.6875,
+      "learning_rate": 1.1187299302649996e-05,
+      "loss": 0.9339,
+      "step": 8139
+    },
+    {
+      "epoch": 1.3837366311593116,
+      "grad_norm": 1.6953125,
+      "learning_rate": 1.118550612920241e-05,
+      "loss": 0.7956,
+      "step": 8140
+    },
+    {
+      "epoch": 1.3839080952483014,
+      "grad_norm": 1.6953125,
+      "learning_rate": 1.1183712917091006e-05,
+      "loss": 0.9118,
+      "step": 8141
+    },
+    {
+      "epoch": 1.3840795593372914,
+      "grad_norm": 1.703125,
+      "learning_rate": 1.1181919666374266e-05,
+      "loss": 0.8234,
+      "step": 8142
+    },
+    {
+      "epoch": 1.3842510234262813,
+      "grad_norm": 1.7109375,
+      "learning_rate": 1.1180126377110674e-05,
+      "loss": 0.8845,
+      "step": 8143
+    },
+    {
+      "epoch": 1.384422487515271,
+      "grad_norm": 1.703125,
+      "learning_rate": 1.117833304935872e-05,
+      "loss": 0.8255,
+      "step": 8144
+    },
+    {
+      "epoch": 1.384593951604261,
+      "grad_norm": 1.625,
+      "learning_rate": 1.1176539683176887e-05,
+      "loss": 0.8834,
+      "step": 8145
+    },
+    {
+      "epoch": 1.3847654156932507,
+      "grad_norm": 1.671875,
+      "learning_rate": 1.1174746278623662e-05,
+      "loss": 0.7859,
+      "step": 8146
+    },
+    {
+      "epoch": 1.3849368797822406,
+      "grad_norm": 1.65625,
+      "learning_rate": 1.117295283575754e-05,
+      "loss": 0.9149,
+      "step": 8147
+    },
+    {
+      "epoch": 1.3851083438712304,
+      "grad_norm": 1.6640625,
+      "learning_rate": 1.1171159354637008e-05,
+      "loss": 0.8333,
+      "step": 8148
+    },
+    {
+      "epoch": 1.3852798079602202,
+      "grad_norm": 1.734375,
+      "learning_rate": 1.116936583532056e-05,
+      "loss": 0.8805,
+      "step": 8149
+    },
+    {
+      "epoch": 1.3854512720492103,
+      "grad_norm": 1.7421875,
+      "learning_rate": 1.1167572277866688e-05,
+      "loss": 0.8682,
+      "step": 8150
+    },
+    {
+      "epoch": 1.3856227361382,
+      "grad_norm": 1.625,
+      "learning_rate": 1.1165778682333888e-05,
+      "loss": 0.8281,
+      "step": 8151
+    },
+    {
+      "epoch": 1.38579420022719,
+      "grad_norm": 1.59375,
+      "learning_rate": 1.1163985048780652e-05,
+      "loss": 0.8823,
+      "step": 8152
+    },
+    {
+      "epoch": 1.3859656643161797,
+      "grad_norm": 1.7421875,
+      "learning_rate": 1.1162191377265483e-05,
+      "loss": 0.8931,
+      "step": 8153
+    },
+    {
+      "epoch": 1.3861371284051696,
+      "grad_norm": 1.671875,
+      "learning_rate": 1.1160397667846875e-05,
+      "loss": 0.8454,
+      "step": 8154
+    },
+    {
+      "epoch": 1.3863085924941596,
+      "grad_norm": 1.6484375,
+      "learning_rate": 1.1158603920583336e-05,
+      "loss": 0.8052,
+      "step": 8155
+    },
+    {
+      "epoch": 1.3864800565831494,
+      "grad_norm": 1.7265625,
+      "learning_rate": 1.1156810135533354e-05,
+      "loss": 0.8256,
+      "step": 8156
+    },
+    {
+      "epoch": 1.3866515206721393,
+      "grad_norm": 1.7265625,
+      "learning_rate": 1.115501631275544e-05,
+      "loss": 0.8202,
+      "step": 8157
+    },
+    {
+      "epoch": 1.386822984761129,
+      "grad_norm": 1.5390625,
+      "learning_rate": 1.1153222452308093e-05,
+      "loss": 0.7897,
+      "step": 8158
+    },
+    {
+      "epoch": 1.386994448850119,
+      "grad_norm": 1.7890625,
+      "learning_rate": 1.1151428554249818e-05,
+      "loss": 0.931,
+      "step": 8159
+    },
+    {
+      "epoch": 1.3871659129391087,
+      "grad_norm": 1.609375,
+      "learning_rate": 1.1149634618639123e-05,
+      "loss": 0.7929,
+      "step": 8160
+    },
+    {
+      "epoch": 1.3873373770280986,
+      "grad_norm": 1.8125,
+      "learning_rate": 1.1147840645534515e-05,
+      "loss": 0.9386,
+      "step": 8161
+    },
+    {
+      "epoch": 1.3875088411170886,
+      "grad_norm": 1.6484375,
+      "learning_rate": 1.11460466349945e-05,
+      "loss": 0.8324,
+      "step": 8162
+    },
+    {
+      "epoch": 1.3876803052060784,
+      "grad_norm": 1.640625,
+      "learning_rate": 1.114425258707759e-05,
+      "loss": 0.8598,
+      "step": 8163
+    },
+    {
+      "epoch": 1.3878517692950683,
+      "grad_norm": 2.015625,
+      "learning_rate": 1.1142458501842292e-05,
+      "loss": 0.901,
+      "step": 8164
+    },
+    {
+      "epoch": 1.388023233384058,
+      "grad_norm": 1.84375,
+      "learning_rate": 1.1140664379347124e-05,
+      "loss": 0.8722,
+      "step": 8165
+    },
+    {
+      "epoch": 1.388194697473048,
+      "grad_norm": 1.6953125,
+      "learning_rate": 1.1138870219650592e-05,
+      "loss": 0.845,
+      "step": 8166
+    },
+    {
+      "epoch": 1.388366161562038,
+      "grad_norm": 2.015625,
+      "learning_rate": 1.1137076022811217e-05,
+      "loss": 0.8876,
+      "step": 8167
+    },
+    {
+      "epoch": 1.3885376256510278,
+      "grad_norm": 1.7109375,
+      "learning_rate": 1.1135281788887506e-05,
+      "loss": 0.8479,
+      "step": 8168
+    },
+    {
+      "epoch": 1.3887090897400176,
+      "grad_norm": 1.6171875,
+      "learning_rate": 1.1133487517937985e-05,
+      "loss": 0.8327,
+      "step": 8169
+    },
+    {
+      "epoch": 1.3888805538290074,
+      "grad_norm": 1.765625,
+      "learning_rate": 1.1131693210021163e-05,
+      "loss": 0.8874,
+      "step": 8170
+    },
+    {
+      "epoch": 1.3890520179179973,
+      "grad_norm": 1.703125,
+      "learning_rate": 1.1129898865195568e-05,
+      "loss": 0.8704,
+      "step": 8171
+    },
+    {
+      "epoch": 1.389223482006987,
+      "grad_norm": 1.59375,
+      "learning_rate": 1.1128104483519715e-05,
+      "loss": 0.8239,
+      "step": 8172
+    },
+    {
+      "epoch": 1.389394946095977,
+      "grad_norm": 1.6953125,
+      "learning_rate": 1.1126310065052128e-05,
+      "loss": 0.8521,
+      "step": 8173
+    },
+    {
+      "epoch": 1.389566410184967,
+      "grad_norm": 1.6484375,
+      "learning_rate": 1.1124515609851328e-05,
+      "loss": 0.8451,
+      "step": 8174
+    },
+    {
+      "epoch": 1.3897378742739568,
+      "grad_norm": 1.75,
+      "learning_rate": 1.1122721117975838e-05,
+      "loss": 0.8913,
+      "step": 8175
+    },
+    {
+      "epoch": 1.3899093383629466,
+      "grad_norm": 1.640625,
+      "learning_rate": 1.1120926589484187e-05,
+      "loss": 0.8485,
+      "step": 8176
+    },
+    {
+      "epoch": 1.3900808024519364,
+      "grad_norm": 1.6484375,
+      "learning_rate": 1.1119132024434896e-05,
+      "loss": 0.8612,
+      "step": 8177
+    },
+    {
+      "epoch": 1.3902522665409263,
+      "grad_norm": 1.609375,
+      "learning_rate": 1.1117337422886497e-05,
+      "loss": 0.8926,
+      "step": 8178
+    },
+    {
+      "epoch": 1.3904237306299163,
+      "grad_norm": 1.65625,
+      "learning_rate": 1.1115542784897518e-05,
+      "loss": 0.8277,
+      "step": 8179
+    },
+    {
+      "epoch": 1.3905951947189061,
+      "grad_norm": 1.7734375,
+      "learning_rate": 1.1113748110526486e-05,
+      "loss": 0.9561,
+      "step": 8180
+    },
+    {
+      "epoch": 1.390766658807896,
+      "grad_norm": 1.6484375,
+      "learning_rate": 1.1111953399831934e-05,
+      "loss": 0.7891,
+      "step": 8181
+    },
+    {
+      "epoch": 1.3909381228968858,
+      "grad_norm": 1.65625,
+      "learning_rate": 1.1110158652872395e-05,
+      "loss": 0.849,
+      "step": 8182
+    },
+    {
+      "epoch": 1.3911095869858756,
+      "grad_norm": 1.640625,
+      "learning_rate": 1.1108363869706402e-05,
+      "loss": 0.7991,
+      "step": 8183
+    },
+    {
+      "epoch": 1.3912810510748654,
+      "grad_norm": 1.671875,
+      "learning_rate": 1.1106569050392492e-05,
+      "loss": 0.8487,
+      "step": 8184
+    },
+    {
+      "epoch": 1.3914525151638553,
+      "grad_norm": 1.640625,
+      "learning_rate": 1.1104774194989197e-05,
+      "loss": 0.8098,
+      "step": 8185
+    },
+    {
+      "epoch": 1.3916239792528453,
+      "grad_norm": 1.75,
+      "learning_rate": 1.1102979303555052e-05,
+      "loss": 0.8735,
+      "step": 8186
+    },
+    {
+      "epoch": 1.3917954433418351,
+      "grad_norm": 1.6328125,
+      "learning_rate": 1.11011843761486e-05,
+      "loss": 0.7936,
+      "step": 8187
+    },
+    {
+      "epoch": 1.391966907430825,
+      "grad_norm": 1.6796875,
+      "learning_rate": 1.1099389412828379e-05,
+      "loss": 0.9163,
+      "step": 8188
+    },
+    {
+      "epoch": 1.3921383715198148,
+      "grad_norm": 1.7578125,
+      "learning_rate": 1.109759441365293e-05,
+      "loss": 0.9175,
+      "step": 8189
+    },
+    {
+      "epoch": 1.3923098356088046,
+      "grad_norm": 1.640625,
+      "learning_rate": 1.1095799378680796e-05,
+      "loss": 0.8568,
+      "step": 8190
+    },
+    {
+      "epoch": 1.3924812996977947,
+      "grad_norm": 1.703125,
+      "learning_rate": 1.1094004307970516e-05,
+      "loss": 0.8124,
+      "step": 8191
+    },
+    {
+      "epoch": 1.3926527637867845,
+      "grad_norm": 1.703125,
+      "learning_rate": 1.1092209201580634e-05,
+      "loss": 0.9517,
+      "step": 8192
+    },
+    {
+      "epoch": 1.3928242278757743,
+      "grad_norm": 1.7421875,
+      "learning_rate": 1.10904140595697e-05,
+      "loss": 0.83,
+      "step": 8193
+    },
+    {
+      "epoch": 1.3929956919647641,
+      "grad_norm": 1.6640625,
+      "learning_rate": 1.1088618881996257e-05,
+      "loss": 0.8657,
+      "step": 8194
+    },
+    {
+      "epoch": 1.393167156053754,
+      "grad_norm": 1.6484375,
+      "learning_rate": 1.108682366891885e-05,
+      "loss": 0.7925,
+      "step": 8195
+    },
+    {
+      "epoch": 1.3933386201427438,
+      "grad_norm": 1.578125,
+      "learning_rate": 1.1085028420396033e-05,
+      "loss": 0.6944,
+      "step": 8196
+    },
+    {
+      "epoch": 1.3935100842317336,
+      "grad_norm": 1.578125,
+      "learning_rate": 1.1083233136486354e-05,
+      "loss": 0.8304,
+      "step": 8197
+    },
+    {
+      "epoch": 1.3936815483207234,
+      "grad_norm": 1.7421875,
+      "learning_rate": 1.1081437817248362e-05,
+      "loss": 0.8687,
+      "step": 8198
+    },
+    {
+      "epoch": 1.3938530124097135,
+      "grad_norm": 1.5859375,
+      "learning_rate": 1.1079642462740612e-05,
+      "loss": 0.7692,
+      "step": 8199
+    },
+    {
+      "epoch": 1.3940244764987033,
+      "grad_norm": 1.671875,
+      "learning_rate": 1.1077847073021653e-05,
+      "loss": 0.8898,
+      "step": 8200
+    },
+    {
+      "epoch": 1.3941959405876931,
+      "grad_norm": 1.703125,
+      "learning_rate": 1.1076051648150043e-05,
+      "loss": 0.8604,
+      "step": 8201
+    },
+    {
+      "epoch": 1.394367404676683,
+      "grad_norm": 1.6328125,
+      "learning_rate": 1.107425618818434e-05,
+      "loss": 0.793,
+      "step": 8202
+    },
+    {
+      "epoch": 1.394538868765673,
+      "grad_norm": 1.7578125,
+      "learning_rate": 1.1072460693183095e-05,
+      "loss": 0.9108,
+      "step": 8203
+    },
+    {
+      "epoch": 1.3947103328546628,
+      "grad_norm": 1.7890625,
+      "learning_rate": 1.1070665163204868e-05,
+      "loss": 0.8878,
+      "step": 8204
+    },
+    {
+      "epoch": 1.3948817969436527,
+      "grad_norm": 1.6484375,
+      "learning_rate": 1.1068869598308218e-05,
+      "loss": 0.8829,
+      "step": 8205
+    },
+    {
+      "epoch": 1.3950532610326425,
+      "grad_norm": 1.7421875,
+      "learning_rate": 1.1067073998551706e-05,
+      "loss": 0.9368,
+      "step": 8206
+    },
+    {
+      "epoch": 1.3952247251216323,
+      "grad_norm": 1.625,
+      "learning_rate": 1.1065278363993892e-05,
+      "loss": 0.8212,
+      "step": 8207
+    },
+    {
+      "epoch": 1.3953961892106221,
+      "grad_norm": 1.734375,
+      "learning_rate": 1.1063482694693339e-05,
+      "loss": 0.892,
+      "step": 8208
+    },
+    {
+      "epoch": 1.395567653299612,
+      "grad_norm": 1.7265625,
+      "learning_rate": 1.1061686990708612e-05,
+      "loss": 0.8342,
+      "step": 8209
+    },
+    {
+      "epoch": 1.3957391173886018,
+      "grad_norm": 1.6875,
+      "learning_rate": 1.1059891252098273e-05,
+      "loss": 0.8443,
+      "step": 8210
+    },
+    {
+      "epoch": 1.3959105814775918,
+      "grad_norm": 1.6796875,
+      "learning_rate": 1.105809547892089e-05,
+      "loss": 0.8837,
+      "step": 8211
+    },
+    {
+      "epoch": 1.3960820455665817,
+      "grad_norm": 1.5859375,
+      "learning_rate": 1.1056299671235028e-05,
+      "loss": 0.8203,
+      "step": 8212
+    },
+    {
+      "epoch": 1.3962535096555715,
+      "grad_norm": 1.765625,
+      "learning_rate": 1.105450382909926e-05,
+      "loss": 0.8863,
+      "step": 8213
+    },
+    {
+      "epoch": 1.3964249737445613,
+      "grad_norm": 1.7421875,
+      "learning_rate": 1.1052707952572149e-05,
+      "loss": 0.8789,
+      "step": 8214
+    },
+    {
+      "epoch": 1.3965964378335514,
+      "grad_norm": 1.6875,
+      "learning_rate": 1.1050912041712264e-05,
+      "loss": 0.8143,
+      "step": 8215
+    },
+    {
+      "epoch": 1.3967679019225412,
+      "grad_norm": 1.734375,
+      "learning_rate": 1.1049116096578181e-05,
+      "loss": 0.8136,
+      "step": 8216
+    },
+    {
+      "epoch": 1.396939366011531,
+      "grad_norm": 1.8203125,
+      "learning_rate": 1.1047320117228472e-05,
+      "loss": 0.9049,
+      "step": 8217
+    },
+    {
+      "epoch": 1.3971108301005208,
+      "grad_norm": 1.6171875,
+      "learning_rate": 1.1045524103721712e-05,
+      "loss": 0.806,
+      "step": 8218
+    },
+    {
+      "epoch": 1.3972822941895107,
+      "grad_norm": 1.734375,
+      "learning_rate": 1.1043728056116474e-05,
+      "loss": 0.8712,
+      "step": 8219
+    },
+    {
+      "epoch": 1.3974537582785005,
+      "grad_norm": 1.5625,
+      "learning_rate": 1.1041931974471331e-05,
+      "loss": 0.8647,
+      "step": 8220
+    },
+    {
+      "epoch": 1.3976252223674903,
+      "grad_norm": 1.765625,
+      "learning_rate": 1.1040135858844864e-05,
+      "loss": 0.9376,
+      "step": 8221
+    },
+    {
+      "epoch": 1.3977966864564801,
+      "grad_norm": 1.6484375,
+      "learning_rate": 1.1038339709295652e-05,
+      "loss": 0.8598,
+      "step": 8222
+    },
+    {
+      "epoch": 1.3979681505454702,
+      "grad_norm": 1.71875,
+      "learning_rate": 1.103654352588227e-05,
+      "loss": 0.8858,
+      "step": 8223
+    },
+    {
+      "epoch": 1.39813961463446,
+      "grad_norm": 1.75,
+      "learning_rate": 1.10347473086633e-05,
+      "loss": 0.9499,
+      "step": 8224
+    },
+    {
+      "epoch": 1.3983110787234498,
+      "grad_norm": 1.765625,
+      "learning_rate": 1.1032951057697325e-05,
+      "loss": 0.8693,
+      "step": 8225
+    },
+    {
+      "epoch": 1.3984825428124397,
+      "grad_norm": 1.703125,
+      "learning_rate": 1.1031154773042925e-05,
+      "loss": 0.8588,
+      "step": 8226
+    },
+    {
+      "epoch": 1.3986540069014297,
+      "grad_norm": 1.65625,
+      "learning_rate": 1.1029358454758687e-05,
+      "loss": 0.8084,
+      "step": 8227
+    },
+    {
+      "epoch": 1.3988254709904195,
+      "grad_norm": 1.6640625,
+      "learning_rate": 1.102756210290319e-05,
+      "loss": 0.819,
+      "step": 8228
+    },
+    {
+      "epoch": 1.3989969350794094,
+      "grad_norm": 1.671875,
+      "learning_rate": 1.1025765717535026e-05,
+      "loss": 0.8563,
+      "step": 8229
+    },
+    {
+      "epoch": 1.3991683991683992,
+      "grad_norm": 1.7109375,
+      "learning_rate": 1.1023969298712783e-05,
+      "loss": 0.8326,
+      "step": 8230
+    },
+    {
+      "epoch": 1.399339863257389,
+      "grad_norm": 1.7890625,
+      "learning_rate": 1.102217284649504e-05,
+      "loss": 0.9333,
+      "step": 8231
+    },
+    {
+      "epoch": 1.3995113273463788,
+      "grad_norm": 1.734375,
+      "learning_rate": 1.1020376360940393e-05,
+      "loss": 0.8824,
+      "step": 8232
+    },
+    {
+      "epoch": 1.3996827914353687,
+      "grad_norm": 1.6953125,
+      "learning_rate": 1.101857984210743e-05,
+      "loss": 0.8249,
+      "step": 8233
+    },
+    {
+      "epoch": 1.3998542555243585,
+      "grad_norm": 1.8046875,
+      "learning_rate": 1.101678329005474e-05,
+      "loss": 0.8813,
+      "step": 8234
+    },
+    {
+      "epoch": 1.4000257196133485,
+      "grad_norm": 1.7734375,
+      "learning_rate": 1.101498670484092e-05,
+      "loss": 0.8005,
+      "step": 8235
+    },
+    {
+      "epoch": 1.4001971837023384,
+      "grad_norm": 1.7109375,
+      "learning_rate": 1.1013190086524564e-05,
+      "loss": 0.8701,
+      "step": 8236
+    },
+    {
+      "epoch": 1.4003686477913282,
+      "grad_norm": 1.6953125,
+      "learning_rate": 1.1011393435164261e-05,
+      "loss": 0.894,
+      "step": 8237
+    },
+    {
+      "epoch": 1.400540111880318,
+      "grad_norm": 1.6875,
+      "learning_rate": 1.1009596750818611e-05,
+      "loss": 0.843,
+      "step": 8238
+    },
+    {
+      "epoch": 1.400711575969308,
+      "grad_norm": 1.59375,
+      "learning_rate": 1.100780003354621e-05,
+      "loss": 0.844,
+      "step": 8239
+    },
+    {
+      "epoch": 1.400883040058298,
+      "grad_norm": 1.78125,
+      "learning_rate": 1.1006003283405653e-05,
+      "loss": 0.9017,
+      "step": 8240
+    },
+    {
+      "epoch": 1.4010545041472877,
+      "grad_norm": 1.703125,
+      "learning_rate": 1.1004206500455541e-05,
+      "loss": 0.8942,
+      "step": 8241
+    },
+    {
+      "epoch": 1.4012259682362775,
+      "grad_norm": 1.59375,
+      "learning_rate": 1.1002409684754473e-05,
+      "loss": 0.886,
+      "step": 8242
+    },
+    {
+      "epoch": 1.4013974323252674,
+      "grad_norm": 1.7578125,
+      "learning_rate": 1.1000612836361053e-05,
+      "loss": 0.9032,
+      "step": 8243
+    },
+    {
+      "epoch": 1.4015688964142572,
+      "grad_norm": 1.875,
+      "learning_rate": 1.0998815955333878e-05,
+      "loss": 0.8702,
+      "step": 8244
+    },
+    {
+      "epoch": 1.401740360503247,
+      "grad_norm": 1.6796875,
+      "learning_rate": 1.0997019041731553e-05,
+      "loss": 0.9045,
+      "step": 8245
+    },
+    {
+      "epoch": 1.4019118245922368,
+      "grad_norm": 1.6875,
+      "learning_rate": 1.0995222095612686e-05,
+      "loss": 0.942,
+      "step": 8246
+    },
+    {
+      "epoch": 1.402083288681227,
+      "grad_norm": 1.734375,
+      "learning_rate": 1.0993425117035876e-05,
+      "loss": 0.8191,
+      "step": 8247
+    },
+    {
+      "epoch": 1.4022547527702167,
+      "grad_norm": 1.7421875,
+      "learning_rate": 1.0991628106059734e-05,
+      "loss": 0.8259,
+      "step": 8248
+    },
+    {
+      "epoch": 1.4024262168592065,
+      "grad_norm": 1.6640625,
+      "learning_rate": 1.0989831062742867e-05,
+      "loss": 0.7985,
+      "step": 8249
+    },
+    {
+      "epoch": 1.4025976809481964,
+      "grad_norm": 1.6484375,
+      "learning_rate": 1.0988033987143876e-05,
+      "loss": 0.8193,
+      "step": 8250
+    },
+    {
+      "epoch": 1.4027691450371862,
+      "grad_norm": 1.7265625,
+      "learning_rate": 1.0986236879321379e-05,
+      "loss": 0.9145,
+      "step": 8251
+    },
+    {
+      "epoch": 1.4029406091261762,
+      "grad_norm": 1.6796875,
+      "learning_rate": 1.0984439739333983e-05,
+      "loss": 0.8936,
+      "step": 8252
+    },
+    {
+      "epoch": 1.403112073215166,
+      "grad_norm": 1.6328125,
+      "learning_rate": 1.09826425672403e-05,
+      "loss": 0.8913,
+      "step": 8253
+    },
+    {
+      "epoch": 1.403283537304156,
+      "grad_norm": 1.6875,
+      "learning_rate": 1.0980845363098945e-05,
+      "loss": 0.8355,
+      "step": 8254
+    },
+    {
+      "epoch": 1.4034550013931457,
+      "grad_norm": 1.671875,
+      "learning_rate": 1.0979048126968525e-05,
+      "loss": 0.8765,
+      "step": 8255
+    },
+    {
+      "epoch": 1.4036264654821355,
+      "grad_norm": 1.703125,
+      "learning_rate": 1.0977250858907662e-05,
+      "loss": 0.8471,
+      "step": 8256
+    },
+    {
+      "epoch": 1.4037979295711254,
+      "grad_norm": 1.6953125,
+      "learning_rate": 1.0975453558974968e-05,
+      "loss": 0.8599,
+      "step": 8257
+    },
+    {
+      "epoch": 1.4039693936601152,
+      "grad_norm": 1.6875,
+      "learning_rate": 1.0973656227229059e-05,
+      "loss": 0.8168,
+      "step": 8258
+    },
+    {
+      "epoch": 1.4041408577491052,
+      "grad_norm": 1.71875,
+      "learning_rate": 1.0971858863728557e-05,
+      "loss": 0.8604,
+      "step": 8259
+    },
+    {
+      "epoch": 1.404312321838095,
+      "grad_norm": 1.6484375,
+      "learning_rate": 1.0970061468532075e-05,
+      "loss": 0.836,
+      "step": 8260
+    },
+    {
+      "epoch": 1.404483785927085,
+      "grad_norm": 1.6875,
+      "learning_rate": 1.0968264041698237e-05,
+      "loss": 0.8421,
+      "step": 8261
+    },
+    {
+      "epoch": 1.4046552500160747,
+      "grad_norm": 1.703125,
+      "learning_rate": 1.096646658328566e-05,
+      "loss": 0.8652,
+      "step": 8262
+    },
+    {
+      "epoch": 1.4048267141050645,
+      "grad_norm": 1.7734375,
+      "learning_rate": 1.0964669093352969e-05,
+      "loss": 0.9318,
+      "step": 8263
+    },
+    {
+      "epoch": 1.4049981781940546,
+      "grad_norm": 1.6484375,
+      "learning_rate": 1.0962871571958788e-05,
+      "loss": 0.8129,
+      "step": 8264
+    },
+    {
+      "epoch": 1.4051696422830444,
+      "grad_norm": 1.671875,
+      "learning_rate": 1.096107401916174e-05,
+      "loss": 0.9011,
+      "step": 8265
+    },
+    {
+      "epoch": 1.4053411063720342,
+      "grad_norm": 1.78125,
+      "learning_rate": 1.0959276435020448e-05,
+      "loss": 0.9159,
+      "step": 8266
+    },
+    {
+      "epoch": 1.405512570461024,
+      "grad_norm": 1.6640625,
+      "learning_rate": 1.0957478819593538e-05,
+      "loss": 0.8199,
+      "step": 8267
+    },
+    {
+      "epoch": 1.405684034550014,
+      "grad_norm": 1.640625,
+      "learning_rate": 1.0955681172939639e-05,
+      "loss": 0.8743,
+      "step": 8268
+    },
+    {
+      "epoch": 1.4058554986390037,
+      "grad_norm": 1.6796875,
+      "learning_rate": 1.0953883495117379e-05,
+      "loss": 0.7695,
+      "step": 8269
+    },
+    {
+      "epoch": 1.4060269627279935,
+      "grad_norm": 1.640625,
+      "learning_rate": 1.0952085786185385e-05,
+      "loss": 0.8522,
+      "step": 8270
+    },
+    {
+      "epoch": 1.4061984268169836,
+      "grad_norm": 1.7421875,
+      "learning_rate": 1.095028804620229e-05,
+      "loss": 0.8782,
+      "step": 8271
+    },
+    {
+      "epoch": 1.4063698909059734,
+      "grad_norm": 1.75,
+      "learning_rate": 1.094849027522672e-05,
+      "loss": 1.0023,
+      "step": 8272
+    },
+    {
+      "epoch": 1.4065413549949632,
+      "grad_norm": 1.703125,
+      "learning_rate": 1.0946692473317313e-05,
+      "loss": 0.8444,
+      "step": 8273
+    },
+    {
+      "epoch": 1.406712819083953,
+      "grad_norm": 1.7578125,
+      "learning_rate": 1.0944894640532697e-05,
+      "loss": 0.825,
+      "step": 8274
+    },
+    {
+      "epoch": 1.406884283172943,
+      "grad_norm": 1.8125,
+      "learning_rate": 1.0943096776931512e-05,
+      "loss": 0.931,
+      "step": 8275
+    },
+    {
+      "epoch": 1.407055747261933,
+      "grad_norm": 1.7421875,
+      "learning_rate": 1.0941298882572387e-05,
+      "loss": 0.8311,
+      "step": 8276
+    },
+    {
+      "epoch": 1.4072272113509228,
+      "grad_norm": 1.7578125,
+      "learning_rate": 1.0939500957513964e-05,
+      "loss": 0.9291,
+      "step": 8277
+    },
+    {
+      "epoch": 1.4073986754399126,
+      "grad_norm": 1.6796875,
+      "learning_rate": 1.0937703001814877e-05,
+      "loss": 0.8732,
+      "step": 8278
+    },
+    {
+      "epoch": 1.4075701395289024,
+      "grad_norm": 1.6796875,
+      "learning_rate": 1.0935905015533761e-05,
+      "loss": 0.7694,
+      "step": 8279
+    },
+    {
+      "epoch": 1.4077416036178922,
+      "grad_norm": 1.6875,
+      "learning_rate": 1.093410699872926e-05,
+      "loss": 0.8506,
+      "step": 8280
+    },
+    {
+      "epoch": 1.407913067706882,
+      "grad_norm": 1.6171875,
+      "learning_rate": 1.093230895146001e-05,
+      "loss": 0.839,
+      "step": 8281
+    },
+    {
+      "epoch": 1.408084531795872,
+      "grad_norm": 1.6953125,
+      "learning_rate": 1.0930510873784657e-05,
+      "loss": 0.8417,
+      "step": 8282
+    },
+    {
+      "epoch": 1.408255995884862,
+      "grad_norm": 1.7734375,
+      "learning_rate": 1.0928712765761837e-05,
+      "loss": 0.8868,
+      "step": 8283
+    },
+    {
+      "epoch": 1.4084274599738518,
+      "grad_norm": 1.734375,
+      "learning_rate": 1.09269146274502e-05,
+      "loss": 0.8386,
+      "step": 8284
+    },
+    {
+      "epoch": 1.4085989240628416,
+      "grad_norm": 1.6171875,
+      "learning_rate": 1.0925116458908385e-05,
+      "loss": 0.7894,
+      "step": 8285
+    },
+    {
+      "epoch": 1.4087703881518314,
+      "grad_norm": 1.75,
+      "learning_rate": 1.092331826019504e-05,
+      "loss": 0.8861,
+      "step": 8286
+    },
+    {
+      "epoch": 1.4089418522408212,
+      "grad_norm": 1.765625,
+      "learning_rate": 1.0921520031368808e-05,
+      "loss": 0.8584,
+      "step": 8287
+    },
+    {
+      "epoch": 1.4091133163298113,
+      "grad_norm": 1.7578125,
+      "learning_rate": 1.0919721772488341e-05,
+      "loss": 0.869,
+      "step": 8288
+    },
+    {
+      "epoch": 1.4092847804188011,
+      "grad_norm": 1.5859375,
+      "learning_rate": 1.0917923483612278e-05,
+      "loss": 0.8477,
+      "step": 8289
+    },
+    {
+      "epoch": 1.409456244507791,
+      "grad_norm": 1.6484375,
+      "learning_rate": 1.0916125164799276e-05,
+      "loss": 0.8251,
+      "step": 8290
+    },
+    {
+      "epoch": 1.4096277085967808,
+      "grad_norm": 1.8046875,
+      "learning_rate": 1.0914326816107984e-05,
+      "loss": 0.8932,
+      "step": 8291
+    },
+    {
+      "epoch": 1.4097991726857706,
+      "grad_norm": 1.6796875,
+      "learning_rate": 1.091252843759705e-05,
+      "loss": 0.8542,
+      "step": 8292
+    },
+    {
+      "epoch": 1.4099706367747604,
+      "grad_norm": 1.71875,
+      "learning_rate": 1.0910730029325127e-05,
+      "loss": 0.9365,
+      "step": 8293
+    },
+    {
+      "epoch": 1.4101421008637502,
+      "grad_norm": 1.5546875,
+      "learning_rate": 1.0908931591350871e-05,
+      "loss": 0.873,
+      "step": 8294
+    },
+    {
+      "epoch": 1.41031356495274,
+      "grad_norm": 1.6953125,
+      "learning_rate": 1.0907133123732931e-05,
+      "loss": 0.849,
+      "step": 8295
+    },
+    {
+      "epoch": 1.4104850290417301,
+      "grad_norm": 1.6484375,
+      "learning_rate": 1.0905334626529964e-05,
+      "loss": 0.8907,
+      "step": 8296
+    },
+    {
+      "epoch": 1.41065649313072,
+      "grad_norm": 1.7109375,
+      "learning_rate": 1.0903536099800627e-05,
+      "loss": 0.8811,
+      "step": 8297
+    },
+    {
+      "epoch": 1.4108279572197098,
+      "grad_norm": 1.6953125,
+      "learning_rate": 1.0901737543603574e-05,
+      "loss": 0.8517,
+      "step": 8298
+    },
+    {
+      "epoch": 1.4109994213086996,
+      "grad_norm": 1.7265625,
+      "learning_rate": 1.0899938957997465e-05,
+      "loss": 0.8915,
+      "step": 8299
+    },
+    {
+      "epoch": 1.4111708853976896,
+      "grad_norm": 1.7421875,
+      "learning_rate": 1.0898140343040957e-05,
+      "loss": 0.8506,
+      "step": 8300
+    },
+    {
+      "epoch": 1.4113423494866795,
+      "grad_norm": 1.6875,
+      "learning_rate": 1.089634169879271e-05,
+      "loss": 0.8708,
+      "step": 8301
+    },
+    {
+      "epoch": 1.4115138135756693,
+      "grad_norm": 1.7734375,
+      "learning_rate": 1.0894543025311384e-05,
+      "loss": 0.9046,
+      "step": 8302
+    },
+    {
+      "epoch": 1.4116852776646591,
+      "grad_norm": 1.8046875,
+      "learning_rate": 1.0892744322655645e-05,
+      "loss": 0.8828,
+      "step": 8303
+    },
+    {
+      "epoch": 1.411856741753649,
+      "grad_norm": 1.625,
+      "learning_rate": 1.0890945590884147e-05,
+      "loss": 0.824,
+      "step": 8304
+    },
+    {
+      "epoch": 1.4120282058426388,
+      "grad_norm": 1.609375,
+      "learning_rate": 1.0889146830055564e-05,
+      "loss": 0.8713,
+      "step": 8305
+    },
+    {
+      "epoch": 1.4121996699316286,
+      "grad_norm": 1.703125,
+      "learning_rate": 1.0887348040228552e-05,
+      "loss": 0.8932,
+      "step": 8306
+    },
+    {
+      "epoch": 1.4123711340206184,
+      "grad_norm": 1.71875,
+      "learning_rate": 1.0885549221461777e-05,
+      "loss": 0.8144,
+      "step": 8307
+    },
+    {
+      "epoch": 1.4125425981096085,
+      "grad_norm": 1.6953125,
+      "learning_rate": 1.0883750373813909e-05,
+      "loss": 0.8233,
+      "step": 8308
+    },
+    {
+      "epoch": 1.4127140621985983,
+      "grad_norm": 1.703125,
+      "learning_rate": 1.088195149734361e-05,
+      "loss": 0.945,
+      "step": 8309
+    },
+    {
+      "epoch": 1.4128855262875881,
+      "grad_norm": 1.6640625,
+      "learning_rate": 1.0880152592109554e-05,
+      "loss": 0.8226,
+      "step": 8310
+    },
+    {
+      "epoch": 1.413056990376578,
+      "grad_norm": 1.6953125,
+      "learning_rate": 1.0878353658170411e-05,
+      "loss": 1.0318,
+      "step": 8311
+    },
+    {
+      "epoch": 1.413228454465568,
+      "grad_norm": 1.640625,
+      "learning_rate": 1.0876554695584845e-05,
+      "loss": 0.8666,
+      "step": 8312
+    },
+    {
+      "epoch": 1.4133999185545578,
+      "grad_norm": 1.6171875,
+      "learning_rate": 1.087475570441153e-05,
+      "loss": 0.8516,
+      "step": 8313
+    },
+    {
+      "epoch": 1.4135713826435476,
+      "grad_norm": 1.671875,
+      "learning_rate": 1.0872956684709138e-05,
+      "loss": 0.8655,
+      "step": 8314
+    },
+    {
+      "epoch": 1.4137428467325375,
+      "grad_norm": 1.7421875,
+      "learning_rate": 1.0871157636536338e-05,
+      "loss": 0.8286,
+      "step": 8315
+    },
+    {
+      "epoch": 1.4139143108215273,
+      "grad_norm": 1.6875,
+      "learning_rate": 1.086935855995181e-05,
+      "loss": 0.904,
+      "step": 8316
+    },
+    {
+      "epoch": 1.4140857749105171,
+      "grad_norm": 1.75,
+      "learning_rate": 1.0867559455014224e-05,
+      "loss": 0.9264,
+      "step": 8317
+    },
+    {
+      "epoch": 1.414257238999507,
+      "grad_norm": 1.671875,
+      "learning_rate": 1.086576032178226e-05,
+      "loss": 0.8874,
+      "step": 8318
+    },
+    {
+      "epoch": 1.4144287030884968,
+      "grad_norm": 1.6640625,
+      "learning_rate": 1.0863961160314588e-05,
+      "loss": 0.8748,
+      "step": 8319
+    },
+    {
+      "epoch": 1.4146001671774868,
+      "grad_norm": 1.7109375,
+      "learning_rate": 1.086216197066989e-05,
+      "loss": 0.8369,
+      "step": 8320
+    },
+    {
+      "epoch": 1.4147716312664766,
+      "grad_norm": 1.6875,
+      "learning_rate": 1.0860362752906846e-05,
+      "loss": 0.8304,
+      "step": 8321
+    },
+    {
+      "epoch": 1.4149430953554665,
+      "grad_norm": 1.7578125,
+      "learning_rate": 1.085856350708413e-05,
+      "loss": 0.8489,
+      "step": 8322
+    },
+    {
+      "epoch": 1.4151145594444563,
+      "grad_norm": 1.625,
+      "learning_rate": 1.0856764233260428e-05,
+      "loss": 0.7751,
+      "step": 8323
+    },
+    {
+      "epoch": 1.4152860235334463,
+      "grad_norm": 1.6640625,
+      "learning_rate": 1.0854964931494418e-05,
+      "loss": 0.8466,
+      "step": 8324
+    },
+    {
+      "epoch": 1.4154574876224362,
+      "grad_norm": 1.6484375,
+      "learning_rate": 1.0853165601844779e-05,
+      "loss": 0.8576,
+      "step": 8325
+    },
+    {
+      "epoch": 1.415628951711426,
+      "grad_norm": 1.7734375,
+      "learning_rate": 1.0851366244370199e-05,
+      "loss": 0.8965,
+      "step": 8326
+    },
+    {
+      "epoch": 1.4158004158004158,
+      "grad_norm": 1.765625,
+      "learning_rate": 1.084956685912936e-05,
+      "loss": 0.9559,
+      "step": 8327
+    },
+    {
+      "epoch": 1.4159718798894056,
+      "grad_norm": 1.59375,
+      "learning_rate": 1.0847767446180945e-05,
+      "loss": 0.8144,
+      "step": 8328
+    },
+    {
+      "epoch": 1.4161433439783955,
+      "grad_norm": 1.765625,
+      "learning_rate": 1.0845968005583643e-05,
+      "loss": 0.8764,
+      "step": 8329
+    },
+    {
+      "epoch": 1.4163148080673853,
+      "grad_norm": 1.6484375,
+      "learning_rate": 1.0844168537396138e-05,
+      "loss": 0.8322,
+      "step": 8330
+    },
+    {
+      "epoch": 1.4164862721563751,
+      "grad_norm": 1.6796875,
+      "learning_rate": 1.0842369041677116e-05,
+      "loss": 0.8158,
+      "step": 8331
+    },
+    {
+      "epoch": 1.4166577362453652,
+      "grad_norm": 1.796875,
+      "learning_rate": 1.0840569518485272e-05,
+      "loss": 0.9187,
+      "step": 8332
+    },
+    {
+      "epoch": 1.416829200334355,
+      "grad_norm": 1.65625,
+      "learning_rate": 1.0838769967879286e-05,
+      "loss": 0.7694,
+      "step": 8333
+    },
+    {
+      "epoch": 1.4170006644233448,
+      "grad_norm": 1.734375,
+      "learning_rate": 1.0836970389917857e-05,
+      "loss": 0.8279,
+      "step": 8334
+    },
+    {
+      "epoch": 1.4171721285123347,
+      "grad_norm": 1.6171875,
+      "learning_rate": 1.083517078465967e-05,
+      "loss": 0.814,
+      "step": 8335
+    },
+    {
+      "epoch": 1.4173435926013247,
+      "grad_norm": 1.65625,
+      "learning_rate": 1.0833371152163417e-05,
+      "loss": 0.7892,
+      "step": 8336
+    },
+    {
+      "epoch": 1.4175150566903145,
+      "grad_norm": 1.75,
+      "learning_rate": 1.0831571492487794e-05,
+      "loss": 0.8923,
+      "step": 8337
+    },
+    {
+      "epoch": 1.4176865207793043,
+      "grad_norm": 1.84375,
+      "learning_rate": 1.0829771805691493e-05,
+      "loss": 0.9074,
+      "step": 8338
+    },
+    {
+      "epoch": 1.4178579848682942,
+      "grad_norm": 1.6796875,
+      "learning_rate": 1.0827972091833207e-05,
+      "loss": 0.8024,
+      "step": 8339
+    },
+    {
+      "epoch": 1.418029448957284,
+      "grad_norm": 1.6953125,
+      "learning_rate": 1.082617235097164e-05,
+      "loss": 0.8634,
+      "step": 8340
+    },
+    {
+      "epoch": 1.4182009130462738,
+      "grad_norm": 1.6953125,
+      "learning_rate": 1.0824372583165477e-05,
+      "loss": 0.8486,
+      "step": 8341
+    },
+    {
+      "epoch": 1.4183723771352637,
+      "grad_norm": 1.6015625,
+      "learning_rate": 1.0822572788473418e-05,
+      "loss": 0.8686,
+      "step": 8342
+    },
+    {
+      "epoch": 1.4185438412242535,
+      "grad_norm": 1.6015625,
+      "learning_rate": 1.0820772966954164e-05,
+      "loss": 0.8334,
+      "step": 8343
+    },
+    {
+      "epoch": 1.4187153053132435,
+      "grad_norm": 1.7265625,
+      "learning_rate": 1.0818973118666414e-05,
+      "loss": 0.9122,
+      "step": 8344
+    },
+    {
+      "epoch": 1.4188867694022333,
+      "grad_norm": 1.796875,
+      "learning_rate": 1.0817173243668866e-05,
+      "loss": 0.888,
+      "step": 8345
+    },
+    {
+      "epoch": 1.4190582334912232,
+      "grad_norm": 1.6640625,
+      "learning_rate": 1.0815373342020221e-05,
+      "loss": 0.8482,
+      "step": 8346
+    },
+    {
+      "epoch": 1.419229697580213,
+      "grad_norm": 1.640625,
+      "learning_rate": 1.0813573413779181e-05,
+      "loss": 0.8716,
+      "step": 8347
+    },
+    {
+      "epoch": 1.4194011616692028,
+      "grad_norm": 1.625,
+      "learning_rate": 1.0811773459004449e-05,
+      "loss": 0.8647,
+      "step": 8348
+    },
+    {
+      "epoch": 1.4195726257581929,
+      "grad_norm": 1.5625,
+      "learning_rate": 1.0809973477754727e-05,
+      "loss": 0.7912,
+      "step": 8349
+    },
+    {
+      "epoch": 1.4197440898471827,
+      "grad_norm": 1.8203125,
+      "learning_rate": 1.0808173470088719e-05,
+      "loss": 0.9811,
+      "step": 8350
+    },
+    {
+      "epoch": 1.4199155539361725,
+      "grad_norm": 1.6640625,
+      "learning_rate": 1.0806373436065134e-05,
+      "loss": 0.8185,
+      "step": 8351
+    },
+    {
+      "epoch": 1.4200870180251624,
+      "grad_norm": 1.6953125,
+      "learning_rate": 1.0804573375742673e-05,
+      "loss": 0.9047,
+      "step": 8352
+    },
+    {
+      "epoch": 1.4202584821141522,
+      "grad_norm": 1.65625,
+      "learning_rate": 1.0802773289180044e-05,
+      "loss": 0.7983,
+      "step": 8353
+    },
+    {
+      "epoch": 1.420429946203142,
+      "grad_norm": 1.703125,
+      "learning_rate": 1.0800973176435953e-05,
+      "loss": 0.8938,
+      "step": 8354
+    },
+    {
+      "epoch": 1.4206014102921318,
+      "grad_norm": 1.75,
+      "learning_rate": 1.0799173037569116e-05,
+      "loss": 0.8996,
+      "step": 8355
+    },
+    {
+      "epoch": 1.4207728743811219,
+      "grad_norm": 1.7109375,
+      "learning_rate": 1.079737287263823e-05,
+      "loss": 0.86,
+      "step": 8356
+    },
+    {
+      "epoch": 1.4209443384701117,
+      "grad_norm": 1.78125,
+      "learning_rate": 1.0795572681702018e-05,
+      "loss": 0.8385,
+      "step": 8357
+    },
+    {
+      "epoch": 1.4211158025591015,
+      "grad_norm": 1.7421875,
+      "learning_rate": 1.079377246481918e-05,
+      "loss": 0.8656,
+      "step": 8358
+    },
+    {
+      "epoch": 1.4212872666480914,
+      "grad_norm": 1.7734375,
+      "learning_rate": 1.0791972222048436e-05,
+      "loss": 0.8572,
+      "step": 8359
+    },
+    {
+      "epoch": 1.4214587307370812,
+      "grad_norm": 1.7734375,
+      "learning_rate": 1.0790171953448496e-05,
+      "loss": 0.8884,
+      "step": 8360
+    },
+    {
+      "epoch": 1.4216301948260712,
+      "grad_norm": 1.625,
+      "learning_rate": 1.0788371659078072e-05,
+      "loss": 0.8412,
+      "step": 8361
+    },
+    {
+      "epoch": 1.421801658915061,
+      "grad_norm": 1.703125,
+      "learning_rate": 1.0786571338995879e-05,
+      "loss": 0.8863,
+      "step": 8362
+    },
+    {
+      "epoch": 1.4219731230040509,
+      "grad_norm": 1.7109375,
+      "learning_rate": 1.0784770993260634e-05,
+      "loss": 0.8877,
+      "step": 8363
+    },
+    {
+      "epoch": 1.4221445870930407,
+      "grad_norm": 1.6875,
+      "learning_rate": 1.0782970621931048e-05,
+      "loss": 0.8147,
+      "step": 8364
+    },
+    {
+      "epoch": 1.4223160511820305,
+      "grad_norm": 1.859375,
+      "learning_rate": 1.0781170225065845e-05,
+      "loss": 0.8803,
+      "step": 8365
+    },
+    {
+      "epoch": 1.4224875152710204,
+      "grad_norm": 1.7890625,
+      "learning_rate": 1.0779369802723738e-05,
+      "loss": 0.8713,
+      "step": 8366
+    },
+    {
+      "epoch": 1.4226589793600102,
+      "grad_norm": 1.65625,
+      "learning_rate": 1.0777569354963448e-05,
+      "loss": 0.8642,
+      "step": 8367
+    },
+    {
+      "epoch": 1.4228304434490002,
+      "grad_norm": 1.7109375,
+      "learning_rate": 1.0775768881843693e-05,
+      "loss": 0.8819,
+      "step": 8368
+    },
+    {
+      "epoch": 1.42300190753799,
+      "grad_norm": 1.703125,
+      "learning_rate": 1.0773968383423197e-05,
+      "loss": 0.8153,
+      "step": 8369
+    },
+    {
+      "epoch": 1.4231733716269799,
+      "grad_norm": 1.734375,
+      "learning_rate": 1.0772167859760674e-05,
+      "loss": 0.9,
+      "step": 8370
+    },
+    {
+      "epoch": 1.4233448357159697,
+      "grad_norm": 1.7265625,
+      "learning_rate": 1.0770367310914849e-05,
+      "loss": 0.8695,
+      "step": 8371
+    },
+    {
+      "epoch": 1.4235162998049595,
+      "grad_norm": 1.7734375,
+      "learning_rate": 1.0768566736944445e-05,
+      "loss": 0.9497,
+      "step": 8372
+    },
+    {
+      "epoch": 1.4236877638939496,
+      "grad_norm": 1.71875,
+      "learning_rate": 1.0766766137908187e-05,
+      "loss": 0.8198,
+      "step": 8373
+    },
+    {
+      "epoch": 1.4238592279829394,
+      "grad_norm": 1.7265625,
+      "learning_rate": 1.0764965513864796e-05,
+      "loss": 0.9214,
+      "step": 8374
+    },
+    {
+      "epoch": 1.4240306920719292,
+      "grad_norm": 1.6875,
+      "learning_rate": 1.0763164864873e-05,
+      "loss": 0.8708,
+      "step": 8375
+    },
+    {
+      "epoch": 1.424202156160919,
+      "grad_norm": 1.6796875,
+      "learning_rate": 1.0761364190991523e-05,
+      "loss": 0.8654,
+      "step": 8376
+    },
+    {
+      "epoch": 1.4243736202499089,
+      "grad_norm": 1.6171875,
+      "learning_rate": 1.0759563492279093e-05,
+      "loss": 0.8355,
+      "step": 8377
+    },
+    {
+      "epoch": 1.4245450843388987,
+      "grad_norm": 1.671875,
+      "learning_rate": 1.0757762768794438e-05,
+      "loss": 0.8843,
+      "step": 8378
+    },
+    {
+      "epoch": 1.4247165484278885,
+      "grad_norm": 1.6953125,
+      "learning_rate": 1.0755962020596285e-05,
+      "loss": 0.7986,
+      "step": 8379
+    },
+    {
+      "epoch": 1.4248880125168786,
+      "grad_norm": 1.6171875,
+      "learning_rate": 1.0754161247743367e-05,
+      "loss": 0.8253,
+      "step": 8380
+    },
+    {
+      "epoch": 1.4250594766058684,
+      "grad_norm": 1.6953125,
+      "learning_rate": 1.0752360450294408e-05,
+      "loss": 0.8199,
+      "step": 8381
+    },
+    {
+      "epoch": 1.4252309406948582,
+      "grad_norm": 1.6328125,
+      "learning_rate": 1.0750559628308142e-05,
+      "loss": 0.8519,
+      "step": 8382
+    },
+    {
+      "epoch": 1.425402404783848,
+      "grad_norm": 1.65625,
+      "learning_rate": 1.0748758781843296e-05,
+      "loss": 0.8915,
+      "step": 8383
+    },
+    {
+      "epoch": 1.4255738688728379,
+      "grad_norm": 1.703125,
+      "learning_rate": 1.0746957910958612e-05,
+      "loss": 0.8812,
+      "step": 8384
+    },
+    {
+      "epoch": 1.425745332961828,
+      "grad_norm": 1.7890625,
+      "learning_rate": 1.0745157015712814e-05,
+      "loss": 0.8603,
+      "step": 8385
+    },
+    {
+      "epoch": 1.4259167970508178,
+      "grad_norm": 1.78125,
+      "learning_rate": 1.0743356096164646e-05,
+      "loss": 0.8808,
+      "step": 8386
+    },
+    {
+      "epoch": 1.4260882611398076,
+      "grad_norm": 1.8046875,
+      "learning_rate": 1.074155515237283e-05,
+      "loss": 0.9764,
+      "step": 8387
+    },
+    {
+      "epoch": 1.4262597252287974,
+      "grad_norm": 1.640625,
+      "learning_rate": 1.0739754184396112e-05,
+      "loss": 0.8728,
+      "step": 8388
+    },
+    {
+      "epoch": 1.4264311893177872,
+      "grad_norm": 1.8671875,
+      "learning_rate": 1.0737953192293222e-05,
+      "loss": 0.8995,
+      "step": 8389
+    },
+    {
+      "epoch": 1.426602653406777,
+      "grad_norm": 1.671875,
+      "learning_rate": 1.0736152176122901e-05,
+      "loss": 0.7569,
+      "step": 8390
+    },
+    {
+      "epoch": 1.4267741174957669,
+      "grad_norm": 1.671875,
+      "learning_rate": 1.0734351135943883e-05,
+      "loss": 0.9227,
+      "step": 8391
+    },
+    {
+      "epoch": 1.4269455815847567,
+      "grad_norm": 1.65625,
+      "learning_rate": 1.0732550071814912e-05,
+      "loss": 0.8322,
+      "step": 8392
+    },
+    {
+      "epoch": 1.4271170456737468,
+      "grad_norm": 1.6796875,
+      "learning_rate": 1.0730748983794723e-05,
+      "loss": 0.884,
+      "step": 8393
+    },
+    {
+      "epoch": 1.4272885097627366,
+      "grad_norm": 1.6796875,
+      "learning_rate": 1.0728947871942057e-05,
+      "loss": 0.8518,
+      "step": 8394
+    },
+    {
+      "epoch": 1.4274599738517264,
+      "grad_norm": 1.6640625,
+      "learning_rate": 1.0727146736315656e-05,
+      "loss": 0.791,
+      "step": 8395
+    },
+    {
+      "epoch": 1.4276314379407162,
+      "grad_norm": 1.6328125,
+      "learning_rate": 1.0725345576974265e-05,
+      "loss": 0.7869,
+      "step": 8396
+    },
+    {
+      "epoch": 1.4278029020297063,
+      "grad_norm": 1.5859375,
+      "learning_rate": 1.0723544393976622e-05,
+      "loss": 0.812,
+      "step": 8397
+    },
+    {
+      "epoch": 1.427974366118696,
+      "grad_norm": 1.703125,
+      "learning_rate": 1.0721743187381473e-05,
+      "loss": 0.888,
+      "step": 8398
+    },
+    {
+      "epoch": 1.428145830207686,
+      "grad_norm": 1.75,
+      "learning_rate": 1.0719941957247557e-05,
+      "loss": 0.9114,
+      "step": 8399
+    },
+    {
+      "epoch": 1.4283172942966758,
+      "grad_norm": 1.6484375,
+      "learning_rate": 1.0718140703633626e-05,
+      "loss": 0.8384,
+      "step": 8400
+    },
+    {
+      "epoch": 1.4283172942966758,
+      "eval_loss": 0.8431193828582764,
+      "eval_runtime": 836.8951,
+      "eval_samples_per_second": 2.986,
+      "eval_steps_per_second": 2.986,
+      "step": 8400
+    },
+    {
+      "epoch": 1.4284887583856656,
+      "grad_norm": 1.6484375,
+      "learning_rate": 1.0716339426598421e-05,
+      "loss": 0.8651,
+      "step": 8401
+    },
+    {
+      "epoch": 1.4286602224746554,
+      "grad_norm": 1.765625,
+      "learning_rate": 1.0714538126200688e-05,
+      "loss": 0.8133,
+      "step": 8402
+    },
+    {
+      "epoch": 1.4288316865636452,
+      "grad_norm": 1.75,
+      "learning_rate": 1.0712736802499179e-05,
+      "loss": 0.8836,
+      "step": 8403
+    },
+    {
+      "epoch": 1.429003150652635,
+      "grad_norm": 1.6328125,
+      "learning_rate": 1.0710935455552637e-05,
+      "loss": 0.8841,
+      "step": 8404
+    },
+    {
+      "epoch": 1.429174614741625,
+      "grad_norm": 1.75,
+      "learning_rate": 1.0709134085419814e-05,
+      "loss": 0.8592,
+      "step": 8405
+    },
+    {
+      "epoch": 1.429346078830615,
+      "grad_norm": 1.78125,
+      "learning_rate": 1.0707332692159459e-05,
+      "loss": 0.8613,
+      "step": 8406
+    },
+    {
+      "epoch": 1.4295175429196048,
+      "grad_norm": 1.75,
+      "learning_rate": 1.0705531275830319e-05,
+      "loss": 0.8449,
+      "step": 8407
+    },
+    {
+      "epoch": 1.4296890070085946,
+      "grad_norm": 1.6796875,
+      "learning_rate": 1.0703729836491148e-05,
+      "loss": 0.8013,
+      "step": 8408
+    },
+    {
+      "epoch": 1.4298604710975846,
+      "grad_norm": 1.65625,
+      "learning_rate": 1.0701928374200699e-05,
+      "loss": 0.8001,
+      "step": 8409
+    },
+    {
+      "epoch": 1.4300319351865745,
+      "grad_norm": 1.7109375,
+      "learning_rate": 1.070012688901772e-05,
+      "loss": 0.8349,
+      "step": 8410
+    },
+    {
+      "epoch": 1.4302033992755643,
+      "grad_norm": 1.6953125,
+      "learning_rate": 1.0698325381000965e-05,
+      "loss": 0.803,
+      "step": 8411
+    },
+    {
+      "epoch": 1.430374863364554,
+      "grad_norm": 1.7265625,
+      "learning_rate": 1.0696523850209193e-05,
+      "loss": 0.8824,
+      "step": 8412
+    },
+    {
+      "epoch": 1.430546327453544,
+      "grad_norm": 1.6796875,
+      "learning_rate": 1.0694722296701155e-05,
+      "loss": 0.8719,
+      "step": 8413
+    },
+    {
+      "epoch": 1.4307177915425338,
+      "grad_norm": 1.71875,
+      "learning_rate": 1.0692920720535604e-05,
+      "loss": 0.8876,
+      "step": 8414
+    },
+    {
+      "epoch": 1.4308892556315236,
+      "grad_norm": 1.6171875,
+      "learning_rate": 1.0691119121771305e-05,
+      "loss": 0.8705,
+      "step": 8415
+    },
+    {
+      "epoch": 1.4310607197205134,
+      "grad_norm": 1.7109375,
+      "learning_rate": 1.0689317500467006e-05,
+      "loss": 0.792,
+      "step": 8416
+    },
+    {
+      "epoch": 1.4312321838095035,
+      "grad_norm": 1.671875,
+      "learning_rate": 1.0687515856681466e-05,
+      "loss": 0.8669,
+      "step": 8417
+    },
+    {
+      "epoch": 1.4314036478984933,
+      "grad_norm": 1.59375,
+      "learning_rate": 1.0685714190473444e-05,
+      "loss": 0.8029,
+      "step": 8418
+    },
+    {
+      "epoch": 1.431575111987483,
+      "grad_norm": 1.625,
+      "learning_rate": 1.0683912501901703e-05,
+      "loss": 0.8535,
+      "step": 8419
+    },
+    {
+      "epoch": 1.431746576076473,
+      "grad_norm": 1.703125,
+      "learning_rate": 1.0682110791024997e-05,
+      "loss": 0.9202,
+      "step": 8420
+    },
+    {
+      "epoch": 1.431918040165463,
+      "grad_norm": 1.6171875,
+      "learning_rate": 1.0680309057902091e-05,
+      "loss": 0.92,
+      "step": 8421
+    },
+    {
+      "epoch": 1.4320895042544528,
+      "grad_norm": 1.625,
+      "learning_rate": 1.0678507302591748e-05,
+      "loss": 0.8744,
+      "step": 8422
+    },
+    {
+      "epoch": 1.4322609683434426,
+      "grad_norm": 1.59375,
+      "learning_rate": 1.0676705525152722e-05,
+      "loss": 0.7187,
+      "step": 8423
+    },
+    {
+      "epoch": 1.4324324324324325,
+      "grad_norm": 1.5859375,
+      "learning_rate": 1.0674903725643783e-05,
+      "loss": 0.8218,
+      "step": 8424
+    },
+    {
+      "epoch": 1.4326038965214223,
+      "grad_norm": 1.7265625,
+      "learning_rate": 1.067310190412369e-05,
+      "loss": 0.8625,
+      "step": 8425
+    },
+    {
+      "epoch": 1.432775360610412,
+      "grad_norm": 1.6484375,
+      "learning_rate": 1.0671300060651215e-05,
+      "loss": 0.8802,
+      "step": 8426
+    },
+    {
+      "epoch": 1.432946824699402,
+      "grad_norm": 1.640625,
+      "learning_rate": 1.0669498195285113e-05,
+      "loss": 0.9051,
+      "step": 8427
+    },
+    {
+      "epoch": 1.4331182887883918,
+      "grad_norm": 1.609375,
+      "learning_rate": 1.0667696308084152e-05,
+      "loss": 0.795,
+      "step": 8428
+    },
+    {
+      "epoch": 1.4332897528773818,
+      "grad_norm": 1.734375,
+      "learning_rate": 1.0665894399107103e-05,
+      "loss": 0.8164,
+      "step": 8429
+    },
+    {
+      "epoch": 1.4334612169663716,
+      "grad_norm": 1.7265625,
+      "learning_rate": 1.0664092468412726e-05,
+      "loss": 0.8222,
+      "step": 8430
+    },
+    {
+      "epoch": 1.4336326810553615,
+      "grad_norm": 1.6953125,
+      "learning_rate": 1.0662290516059797e-05,
+      "loss": 0.8728,
+      "step": 8431
+    },
+    {
+      "epoch": 1.4338041451443513,
+      "grad_norm": 1.8046875,
+      "learning_rate": 1.0660488542107081e-05,
+      "loss": 0.8853,
+      "step": 8432
+    },
+    {
+      "epoch": 1.4339756092333413,
+      "grad_norm": 1.65625,
+      "learning_rate": 1.0658686546613344e-05,
+      "loss": 0.9017,
+      "step": 8433
+    },
+    {
+      "epoch": 1.4341470733223312,
+      "grad_norm": 1.734375,
+      "learning_rate": 1.065688452963736e-05,
+      "loss": 0.8426,
+      "step": 8434
+    },
+    {
+      "epoch": 1.434318537411321,
+      "grad_norm": 1.7734375,
+      "learning_rate": 1.0655082491237896e-05,
+      "loss": 0.8915,
+      "step": 8435
+    },
+    {
+      "epoch": 1.4344900015003108,
+      "grad_norm": 1.65625,
+      "learning_rate": 1.0653280431473725e-05,
+      "loss": 0.8554,
+      "step": 8436
+    },
+    {
+      "epoch": 1.4346614655893006,
+      "grad_norm": 1.75,
+      "learning_rate": 1.0651478350403621e-05,
+      "loss": 0.9187,
+      "step": 8437
+    },
+    {
+      "epoch": 1.4348329296782905,
+      "grad_norm": 1.703125,
+      "learning_rate": 1.0649676248086353e-05,
+      "loss": 0.8988,
+      "step": 8438
+    },
+    {
+      "epoch": 1.4350043937672803,
+      "grad_norm": 1.65625,
+      "learning_rate": 1.0647874124580697e-05,
+      "loss": 0.9059,
+      "step": 8439
+    },
+    {
+      "epoch": 1.43517585785627,
+      "grad_norm": 1.65625,
+      "learning_rate": 1.0646071979945427e-05,
+      "loss": 0.883,
+      "step": 8440
+    },
+    {
+      "epoch": 1.4353473219452602,
+      "grad_norm": 1.8984375,
+      "learning_rate": 1.0644269814239314e-05,
+      "loss": 0.9435,
+      "step": 8441
+    },
+    {
+      "epoch": 1.43551878603425,
+      "grad_norm": 1.71875,
+      "learning_rate": 1.0642467627521138e-05,
+      "loss": 0.9301,
+      "step": 8442
+    },
+    {
+      "epoch": 1.4356902501232398,
+      "grad_norm": 1.6796875,
+      "learning_rate": 1.0640665419849674e-05,
+      "loss": 0.8366,
+      "step": 8443
+    },
+    {
+      "epoch": 1.4358617142122296,
+      "grad_norm": 1.6484375,
+      "learning_rate": 1.06388631912837e-05,
+      "loss": 0.8886,
+      "step": 8444
+    },
+    {
+      "epoch": 1.4360331783012195,
+      "grad_norm": 1.75,
+      "learning_rate": 1.063706094188199e-05,
+      "loss": 0.8579,
+      "step": 8445
+    },
+    {
+      "epoch": 1.4362046423902095,
+      "grad_norm": 1.6796875,
+      "learning_rate": 1.0635258671703324e-05,
+      "loss": 0.874,
+      "step": 8446
+    },
+    {
+      "epoch": 1.4363761064791993,
+      "grad_norm": 1.671875,
+      "learning_rate": 1.0633456380806478e-05,
+      "loss": 0.8619,
+      "step": 8447
+    },
+    {
+      "epoch": 1.4365475705681892,
+      "grad_norm": 1.7109375,
+      "learning_rate": 1.0631654069250235e-05,
+      "loss": 0.8869,
+      "step": 8448
+    },
+    {
+      "epoch": 1.436719034657179,
+      "grad_norm": 1.671875,
+      "learning_rate": 1.0629851737093375e-05,
+      "loss": 0.815,
+      "step": 8449
+    },
+    {
+      "epoch": 1.4368904987461688,
+      "grad_norm": 1.6328125,
+      "learning_rate": 1.062804938439468e-05,
+      "loss": 0.842,
+      "step": 8450
+    },
+    {
+      "epoch": 1.4370619628351586,
+      "grad_norm": 1.5625,
+      "learning_rate": 1.0626247011212924e-05,
+      "loss": 0.8771,
+      "step": 8451
+    },
+    {
+      "epoch": 1.4372334269241485,
+      "grad_norm": 1.6640625,
+      "learning_rate": 1.0624444617606899e-05,
+      "loss": 0.865,
+      "step": 8452
+    },
+    {
+      "epoch": 1.4374048910131385,
+      "grad_norm": 1.640625,
+      "learning_rate": 1.0622642203635383e-05,
+      "loss": 0.8332,
+      "step": 8453
+    },
+    {
+      "epoch": 1.4375763551021283,
+      "grad_norm": 1.78125,
+      "learning_rate": 1.0620839769357158e-05,
+      "loss": 0.904,
+      "step": 8454
+    },
+    {
+      "epoch": 1.4377478191911182,
+      "grad_norm": 1.6953125,
+      "learning_rate": 1.0619037314831015e-05,
+      "loss": 0.8609,
+      "step": 8455
+    },
+    {
+      "epoch": 1.437919283280108,
+      "grad_norm": 1.7265625,
+      "learning_rate": 1.0617234840115731e-05,
+      "loss": 0.9157,
+      "step": 8456
+    },
+    {
+      "epoch": 1.4380907473690978,
+      "grad_norm": 1.7421875,
+      "learning_rate": 1.0615432345270095e-05,
+      "loss": 0.7958,
+      "step": 8457
+    },
+    {
+      "epoch": 1.4382622114580879,
+      "grad_norm": 1.7265625,
+      "learning_rate": 1.0613629830352892e-05,
+      "loss": 0.9063,
+      "step": 8458
+    },
+    {
+      "epoch": 1.4384336755470777,
+      "grad_norm": 1.671875,
+      "learning_rate": 1.061182729542291e-05,
+      "loss": 0.8571,
+      "step": 8459
+    },
+    {
+      "epoch": 1.4386051396360675,
+      "grad_norm": 1.59375,
+      "learning_rate": 1.0610024740538933e-05,
+      "loss": 0.843,
+      "step": 8460
+    },
+    {
+      "epoch": 1.4387766037250573,
+      "grad_norm": 1.6328125,
+      "learning_rate": 1.0608222165759758e-05,
+      "loss": 0.7849,
+      "step": 8461
+    },
+    {
+      "epoch": 1.4389480678140472,
+      "grad_norm": 1.6328125,
+      "learning_rate": 1.0606419571144166e-05,
+      "loss": 0.7812,
+      "step": 8462
+    },
+    {
+      "epoch": 1.439119531903037,
+      "grad_norm": 1.6875,
+      "learning_rate": 1.0604616956750947e-05,
+      "loss": 0.7796,
+      "step": 8463
+    },
+    {
+      "epoch": 1.4392909959920268,
+      "grad_norm": 1.7265625,
+      "learning_rate": 1.0602814322638893e-05,
+      "loss": 0.8566,
+      "step": 8464
+    },
+    {
+      "epoch": 1.4394624600810169,
+      "grad_norm": 1.7890625,
+      "learning_rate": 1.0601011668866793e-05,
+      "loss": 0.8924,
+      "step": 8465
+    },
+    {
+      "epoch": 1.4396339241700067,
+      "grad_norm": 1.6015625,
+      "learning_rate": 1.0599208995493437e-05,
+      "loss": 0.7558,
+      "step": 8466
+    },
+    {
+      "epoch": 1.4398053882589965,
+      "grad_norm": 1.6875,
+      "learning_rate": 1.0597406302577622e-05,
+      "loss": 0.858,
+      "step": 8467
+    },
+    {
+      "epoch": 1.4399768523479863,
+      "grad_norm": 1.625,
+      "learning_rate": 1.0595603590178138e-05,
+      "loss": 0.8138,
+      "step": 8468
+    },
+    {
+      "epoch": 1.4401483164369762,
+      "grad_norm": 1.65625,
+      "learning_rate": 1.0593800858353778e-05,
+      "loss": 0.8355,
+      "step": 8469
+    },
+    {
+      "epoch": 1.4403197805259662,
+      "grad_norm": 1.6328125,
+      "learning_rate": 1.0591998107163335e-05,
+      "loss": 0.7906,
+      "step": 8470
+    },
+    {
+      "epoch": 1.440491244614956,
+      "grad_norm": 1.640625,
+      "learning_rate": 1.0590195336665605e-05,
+      "loss": 0.8069,
+      "step": 8471
+    },
+    {
+      "epoch": 1.4406627087039459,
+      "grad_norm": 1.6953125,
+      "learning_rate": 1.0588392546919384e-05,
+      "loss": 0.8495,
+      "step": 8472
+    },
+    {
+      "epoch": 1.4408341727929357,
+      "grad_norm": 1.6484375,
+      "learning_rate": 1.0586589737983465e-05,
+      "loss": 0.8981,
+      "step": 8473
+    },
+    {
+      "epoch": 1.4410056368819255,
+      "grad_norm": 1.6640625,
+      "learning_rate": 1.0584786909916647e-05,
+      "loss": 0.8765,
+      "step": 8474
+    },
+    {
+      "epoch": 1.4411771009709153,
+      "grad_norm": 1.7265625,
+      "learning_rate": 1.0582984062777726e-05,
+      "loss": 0.7787,
+      "step": 8475
+    },
+    {
+      "epoch": 1.4413485650599052,
+      "grad_norm": 1.7421875,
+      "learning_rate": 1.0581181196625496e-05,
+      "loss": 0.8043,
+      "step": 8476
+    },
+    {
+      "epoch": 1.4415200291488952,
+      "grad_norm": 1.6875,
+      "learning_rate": 1.0579378311518759e-05,
+      "loss": 0.8697,
+      "step": 8477
+    },
+    {
+      "epoch": 1.441691493237885,
+      "grad_norm": 1.65625,
+      "learning_rate": 1.0577575407516318e-05,
+      "loss": 0.9282,
+      "step": 8478
+    },
+    {
+      "epoch": 1.4418629573268749,
+      "grad_norm": 1.796875,
+      "learning_rate": 1.0575772484676965e-05,
+      "loss": 0.9028,
+      "step": 8479
+    },
+    {
+      "epoch": 1.4420344214158647,
+      "grad_norm": 1.8359375,
+      "learning_rate": 1.0573969543059503e-05,
+      "loss": 0.8979,
+      "step": 8480
+    },
+    {
+      "epoch": 1.4422058855048545,
+      "grad_norm": 1.7109375,
+      "learning_rate": 1.0572166582722734e-05,
+      "loss": 0.7764,
+      "step": 8481
+    },
+    {
+      "epoch": 1.4423773495938446,
+      "grad_norm": 1.6796875,
+      "learning_rate": 1.0570363603725456e-05,
+      "loss": 0.8835,
+      "step": 8482
+    },
+    {
+      "epoch": 1.4425488136828344,
+      "grad_norm": 1.6953125,
+      "learning_rate": 1.0568560606126475e-05,
+      "loss": 0.7674,
+      "step": 8483
+    },
+    {
+      "epoch": 1.4427202777718242,
+      "grad_norm": 1.7890625,
+      "learning_rate": 1.0566757589984593e-05,
+      "loss": 0.9045,
+      "step": 8484
+    },
+    {
+      "epoch": 1.442891741860814,
+      "grad_norm": 1.671875,
+      "learning_rate": 1.0564954555358606e-05,
+      "loss": 0.8809,
+      "step": 8485
+    },
+    {
+      "epoch": 1.4430632059498039,
+      "grad_norm": 1.7109375,
+      "learning_rate": 1.056315150230733e-05,
+      "loss": 0.8396,
+      "step": 8486
+    },
+    {
+      "epoch": 1.4432346700387937,
+      "grad_norm": 1.7578125,
+      "learning_rate": 1.056134843088956e-05,
+      "loss": 0.89,
+      "step": 8487
+    },
+    {
+      "epoch": 1.4434061341277835,
+      "grad_norm": 1.796875,
+      "learning_rate": 1.0559545341164103e-05,
+      "loss": 0.8926,
+      "step": 8488
+    },
+    {
+      "epoch": 1.4435775982167733,
+      "grad_norm": 1.75,
+      "learning_rate": 1.0557742233189767e-05,
+      "loss": 0.8657,
+      "step": 8489
+    },
+    {
+      "epoch": 1.4437490623057634,
+      "grad_norm": 1.75,
+      "learning_rate": 1.0555939107025357e-05,
+      "loss": 0.8467,
+      "step": 8490
+    },
+    {
+      "epoch": 1.4439205263947532,
+      "grad_norm": 1.65625,
+      "learning_rate": 1.055413596272968e-05,
+      "loss": 0.8783,
+      "step": 8491
+    },
+    {
+      "epoch": 1.444091990483743,
+      "grad_norm": 1.7734375,
+      "learning_rate": 1.055233280036154e-05,
+      "loss": 0.8765,
+      "step": 8492
+    },
+    {
+      "epoch": 1.4442634545727329,
+      "grad_norm": 1.7734375,
+      "learning_rate": 1.0550529619979748e-05,
+      "loss": 0.9313,
+      "step": 8493
+    },
+    {
+      "epoch": 1.444434918661723,
+      "grad_norm": 1.71875,
+      "learning_rate": 1.0548726421643112e-05,
+      "loss": 0.8334,
+      "step": 8494
+    },
+    {
+      "epoch": 1.4446063827507127,
+      "grad_norm": 1.765625,
+      "learning_rate": 1.054692320541044e-05,
+      "loss": 0.8327,
+      "step": 8495
+    },
+    {
+      "epoch": 1.4447778468397026,
+      "grad_norm": 1.7265625,
+      "learning_rate": 1.0545119971340545e-05,
+      "loss": 0.943,
+      "step": 8496
+    },
+    {
+      "epoch": 1.4449493109286924,
+      "grad_norm": 1.734375,
+      "learning_rate": 1.0543316719492233e-05,
+      "loss": 0.8807,
+      "step": 8497
+    },
+    {
+      "epoch": 1.4451207750176822,
+      "grad_norm": 1.703125,
+      "learning_rate": 1.054151344992432e-05,
+      "loss": 0.8186,
+      "step": 8498
+    },
+    {
+      "epoch": 1.445292239106672,
+      "grad_norm": 1.6484375,
+      "learning_rate": 1.0539710162695611e-05,
+      "loss": 0.943,
+      "step": 8499
+    },
+    {
+      "epoch": 1.4454637031956619,
+      "grad_norm": 1.796875,
+      "learning_rate": 1.053790685786492e-05,
+      "loss": 0.8957,
+      "step": 8500
+    },
+    {
+      "epoch": 1.4456351672846517,
+      "grad_norm": 1.6484375,
+      "learning_rate": 1.0536103535491067e-05,
+      "loss": 0.8395,
+      "step": 8501
+    },
+    {
+      "epoch": 1.4458066313736417,
+      "grad_norm": 1.7421875,
+      "learning_rate": 1.0534300195632856e-05,
+      "loss": 0.802,
+      "step": 8502
+    },
+    {
+      "epoch": 1.4459780954626316,
+      "grad_norm": 1.7421875,
+      "learning_rate": 1.0532496838349101e-05,
+      "loss": 0.8584,
+      "step": 8503
+    },
+    {
+      "epoch": 1.4461495595516214,
+      "grad_norm": 1.6484375,
+      "learning_rate": 1.0530693463698622e-05,
+      "loss": 0.8241,
+      "step": 8504
+    },
+    {
+      "epoch": 1.4463210236406112,
+      "grad_norm": 1.6015625,
+      "learning_rate": 1.0528890071740227e-05,
+      "loss": 0.8376,
+      "step": 8505
+    },
+    {
+      "epoch": 1.4464924877296013,
+      "grad_norm": 1.7578125,
+      "learning_rate": 1.0527086662532737e-05,
+      "loss": 0.9222,
+      "step": 8506
+    },
+    {
+      "epoch": 1.446663951818591,
+      "grad_norm": 1.8046875,
+      "learning_rate": 1.0525283236134968e-05,
+      "loss": 0.9086,
+      "step": 8507
+    },
+    {
+      "epoch": 1.446835415907581,
+      "grad_norm": 1.734375,
+      "learning_rate": 1.0523479792605733e-05,
+      "loss": 0.8853,
+      "step": 8508
+    },
+    {
+      "epoch": 1.4470068799965707,
+      "grad_norm": 1.6796875,
+      "learning_rate": 1.0521676332003851e-05,
+      "loss": 0.8754,
+      "step": 8509
+    },
+    {
+      "epoch": 1.4471783440855606,
+      "grad_norm": 1.828125,
+      "learning_rate": 1.051987285438814e-05,
+      "loss": 0.8787,
+      "step": 8510
+    },
+    {
+      "epoch": 1.4473498081745504,
+      "grad_norm": 1.6640625,
+      "learning_rate": 1.0518069359817418e-05,
+      "loss": 0.8779,
+      "step": 8511
+    },
+    {
+      "epoch": 1.4475212722635402,
+      "grad_norm": 1.7109375,
+      "learning_rate": 1.0516265848350502e-05,
+      "loss": 0.9028,
+      "step": 8512
+    },
+    {
+      "epoch": 1.44769273635253,
+      "grad_norm": 1.796875,
+      "learning_rate": 1.0514462320046212e-05,
+      "loss": 0.8451,
+      "step": 8513
+    },
+    {
+      "epoch": 1.44786420044152,
+      "grad_norm": 1.625,
+      "learning_rate": 1.051265877496337e-05,
+      "loss": 0.8443,
+      "step": 8514
+    },
+    {
+      "epoch": 1.44803566453051,
+      "grad_norm": 1.59375,
+      "learning_rate": 1.0510855213160793e-05,
+      "loss": 0.8919,
+      "step": 8515
+    },
+    {
+      "epoch": 1.4482071286194997,
+      "grad_norm": 1.7421875,
+      "learning_rate": 1.0509051634697305e-05,
+      "loss": 0.8251,
+      "step": 8516
+    },
+    {
+      "epoch": 1.4483785927084896,
+      "grad_norm": 1.7890625,
+      "learning_rate": 1.0507248039631726e-05,
+      "loss": 0.8602,
+      "step": 8517
+    },
+    {
+      "epoch": 1.4485500567974796,
+      "grad_norm": 1.71875,
+      "learning_rate": 1.050544442802288e-05,
+      "loss": 0.8148,
+      "step": 8518
+    },
+    {
+      "epoch": 1.4487215208864694,
+      "grad_norm": 1.796875,
+      "learning_rate": 1.0503640799929586e-05,
+      "loss": 0.8857,
+      "step": 8519
+    },
+    {
+      "epoch": 1.4488929849754593,
+      "grad_norm": 1.671875,
+      "learning_rate": 1.050183715541067e-05,
+      "loss": 0.8083,
+      "step": 8520
+    },
+    {
+      "epoch": 1.449064449064449,
+      "grad_norm": 1.6875,
+      "learning_rate": 1.0500033494524955e-05,
+      "loss": 0.9437,
+      "step": 8521
+    },
+    {
+      "epoch": 1.449235913153439,
+      "grad_norm": 1.6953125,
+      "learning_rate": 1.0498229817331262e-05,
+      "loss": 0.82,
+      "step": 8522
+    },
+    {
+      "epoch": 1.4494073772424287,
+      "grad_norm": 1.71875,
+      "learning_rate": 1.049642612388842e-05,
+      "loss": 0.9085,
+      "step": 8523
+    },
+    {
+      "epoch": 1.4495788413314186,
+      "grad_norm": 1.6015625,
+      "learning_rate": 1.049462241425525e-05,
+      "loss": 0.7999,
+      "step": 8524
+    },
+    {
+      "epoch": 1.4497503054204084,
+      "grad_norm": 1.6328125,
+      "learning_rate": 1.0492818688490587e-05,
+      "loss": 0.9326,
+      "step": 8525
+    },
+    {
+      "epoch": 1.4499217695093984,
+      "grad_norm": 1.640625,
+      "learning_rate": 1.0491014946653248e-05,
+      "loss": 0.8545,
+      "step": 8526
+    },
+    {
+      "epoch": 1.4500932335983883,
+      "grad_norm": 1.625,
+      "learning_rate": 1.0489211188802061e-05,
+      "loss": 0.9042,
+      "step": 8527
+    },
+    {
+      "epoch": 1.450264697687378,
+      "grad_norm": 1.7578125,
+      "learning_rate": 1.0487407414995856e-05,
+      "loss": 0.8962,
+      "step": 8528
+    },
+    {
+      "epoch": 1.450436161776368,
+      "grad_norm": 1.6640625,
+      "learning_rate": 1.048560362529346e-05,
+      "loss": 0.833,
+      "step": 8529
+    },
+    {
+      "epoch": 1.4506076258653577,
+      "grad_norm": 1.71875,
+      "learning_rate": 1.0483799819753704e-05,
+      "loss": 0.8709,
+      "step": 8530
+    },
+    {
+      "epoch": 1.4507790899543478,
+      "grad_norm": 1.6953125,
+      "learning_rate": 1.048199599843541e-05,
+      "loss": 0.8699,
+      "step": 8531
+    },
+    {
+      "epoch": 1.4509505540433376,
+      "grad_norm": 1.703125,
+      "learning_rate": 1.0480192161397408e-05,
+      "loss": 0.855,
+      "step": 8532
+    },
+    {
+      "epoch": 1.4511220181323274,
+      "grad_norm": 1.6640625,
+      "learning_rate": 1.0478388308698537e-05,
+      "loss": 0.9085,
+      "step": 8533
+    },
+    {
+      "epoch": 1.4512934822213173,
+      "grad_norm": 1.765625,
+      "learning_rate": 1.0476584440397622e-05,
+      "loss": 0.81,
+      "step": 8534
+    },
+    {
+      "epoch": 1.451464946310307,
+      "grad_norm": 1.6796875,
+      "learning_rate": 1.0474780556553491e-05,
+      "loss": 0.8483,
+      "step": 8535
+    },
+    {
+      "epoch": 1.451636410399297,
+      "grad_norm": 1.734375,
+      "learning_rate": 1.0472976657224982e-05,
+      "loss": 0.9003,
+      "step": 8536
+    },
+    {
+      "epoch": 1.4518078744882867,
+      "grad_norm": 1.7890625,
+      "learning_rate": 1.047117274247092e-05,
+      "loss": 0.848,
+      "step": 8537
+    },
+    {
+      "epoch": 1.4519793385772768,
+      "grad_norm": 1.765625,
+      "learning_rate": 1.046936881235014e-05,
+      "loss": 0.8042,
+      "step": 8538
+    },
+    {
+      "epoch": 1.4521508026662666,
+      "grad_norm": 1.609375,
+      "learning_rate": 1.0467564866921478e-05,
+      "loss": 0.8757,
+      "step": 8539
+    },
+    {
+      "epoch": 1.4523222667552564,
+      "grad_norm": 1.6875,
+      "learning_rate": 1.0465760906243762e-05,
+      "loss": 0.8383,
+      "step": 8540
+    },
+    {
+      "epoch": 1.4524937308442463,
+      "grad_norm": 1.71875,
+      "learning_rate": 1.0463956930375832e-05,
+      "loss": 0.7948,
+      "step": 8541
+    },
+    {
+      "epoch": 1.452665194933236,
+      "grad_norm": 1.6875,
+      "learning_rate": 1.0462152939376518e-05,
+      "loss": 0.9375,
+      "step": 8542
+    },
+    {
+      "epoch": 1.4528366590222261,
+      "grad_norm": 1.625,
+      "learning_rate": 1.0460348933304656e-05,
+      "loss": 0.8774,
+      "step": 8543
+    },
+    {
+      "epoch": 1.453008123111216,
+      "grad_norm": 1.7890625,
+      "learning_rate": 1.0458544912219082e-05,
+      "loss": 0.9465,
+      "step": 8544
+    },
+    {
+      "epoch": 1.4531795872002058,
+      "grad_norm": 1.7734375,
+      "learning_rate": 1.0456740876178633e-05,
+      "loss": 0.8724,
+      "step": 8545
+    },
+    {
+      "epoch": 1.4533510512891956,
+      "grad_norm": 1.78125,
+      "learning_rate": 1.0454936825242143e-05,
+      "loss": 0.9094,
+      "step": 8546
+    },
+    {
+      "epoch": 1.4535225153781854,
+      "grad_norm": 1.5859375,
+      "learning_rate": 1.0453132759468449e-05,
+      "loss": 0.9206,
+      "step": 8547
+    },
+    {
+      "epoch": 1.4536939794671753,
+      "grad_norm": 1.8046875,
+      "learning_rate": 1.0451328678916393e-05,
+      "loss": 0.9343,
+      "step": 8548
+    },
+    {
+      "epoch": 1.453865443556165,
+      "grad_norm": 1.6171875,
+      "learning_rate": 1.0449524583644806e-05,
+      "loss": 0.8568,
+      "step": 8549
+    },
+    {
+      "epoch": 1.4540369076451551,
+      "grad_norm": 1.7265625,
+      "learning_rate": 1.044772047371253e-05,
+      "loss": 0.8451,
+      "step": 8550
+    },
+    {
+      "epoch": 1.454208371734145,
+      "grad_norm": 1.765625,
+      "learning_rate": 1.0445916349178404e-05,
+      "loss": 0.8701,
+      "step": 8551
+    },
+    {
+      "epoch": 1.4543798358231348,
+      "grad_norm": 1.578125,
+      "learning_rate": 1.0444112210101264e-05,
+      "loss": 0.8086,
+      "step": 8552
+    },
+    {
+      "epoch": 1.4545512999121246,
+      "grad_norm": 1.6796875,
+      "learning_rate": 1.0442308056539956e-05,
+      "loss": 0.8005,
+      "step": 8553
+    },
+    {
+      "epoch": 1.4547227640011144,
+      "grad_norm": 1.640625,
+      "learning_rate": 1.0440503888553316e-05,
+      "loss": 0.8714,
+      "step": 8554
+    },
+    {
+      "epoch": 1.4548942280901045,
+      "grad_norm": 1.609375,
+      "learning_rate": 1.0438699706200184e-05,
+      "loss": 0.8887,
+      "step": 8555
+    },
+    {
+      "epoch": 1.4550656921790943,
+      "grad_norm": 1.796875,
+      "learning_rate": 1.0436895509539405e-05,
+      "loss": 0.9123,
+      "step": 8556
+    },
+    {
+      "epoch": 1.4552371562680841,
+      "grad_norm": 1.6484375,
+      "learning_rate": 1.0435091298629815e-05,
+      "loss": 0.8783,
+      "step": 8557
+    },
+    {
+      "epoch": 1.455408620357074,
+      "grad_norm": 1.71875,
+      "learning_rate": 1.0433287073530263e-05,
+      "loss": 0.897,
+      "step": 8558
+    },
+    {
+      "epoch": 1.4555800844460638,
+      "grad_norm": 1.78125,
+      "learning_rate": 1.0431482834299586e-05,
+      "loss": 0.8824,
+      "step": 8559
+    },
+    {
+      "epoch": 1.4557515485350536,
+      "grad_norm": 1.640625,
+      "learning_rate": 1.0429678580996627e-05,
+      "loss": 0.8112,
+      "step": 8560
+    },
+    {
+      "epoch": 1.4559230126240434,
+      "grad_norm": 1.75,
+      "learning_rate": 1.0427874313680234e-05,
+      "loss": 0.861,
+      "step": 8561
+    },
+    {
+      "epoch": 1.4560944767130335,
+      "grad_norm": 1.6953125,
+      "learning_rate": 1.0426070032409247e-05,
+      "loss": 0.863,
+      "step": 8562
+    },
+    {
+      "epoch": 1.4562659408020233,
+      "grad_norm": 1.703125,
+      "learning_rate": 1.0424265737242514e-05,
+      "loss": 0.8264,
+      "step": 8563
+    },
+    {
+      "epoch": 1.4564374048910131,
+      "grad_norm": 1.7734375,
+      "learning_rate": 1.0422461428238874e-05,
+      "loss": 0.8788,
+      "step": 8564
+    },
+    {
+      "epoch": 1.456608868980003,
+      "grad_norm": 1.71875,
+      "learning_rate": 1.0420657105457182e-05,
+      "loss": 0.8383,
+      "step": 8565
+    },
+    {
+      "epoch": 1.4567803330689928,
+      "grad_norm": 1.7109375,
+      "learning_rate": 1.0418852768956274e-05,
+      "loss": 0.9114,
+      "step": 8566
+    },
+    {
+      "epoch": 1.4569517971579828,
+      "grad_norm": 1.703125,
+      "learning_rate": 1.0417048418795e-05,
+      "loss": 0.8532,
+      "step": 8567
+    },
+    {
+      "epoch": 1.4571232612469727,
+      "grad_norm": 1.734375,
+      "learning_rate": 1.0415244055032205e-05,
+      "loss": 0.8919,
+      "step": 8568
+    },
+    {
+      "epoch": 1.4572947253359625,
+      "grad_norm": 1.625,
+      "learning_rate": 1.041343967772674e-05,
+      "loss": 0.9256,
+      "step": 8569
+    },
+    {
+      "epoch": 1.4574661894249523,
+      "grad_norm": 1.671875,
+      "learning_rate": 1.0411635286937448e-05,
+      "loss": 0.8848,
+      "step": 8570
+    },
+    {
+      "epoch": 1.4576376535139421,
+      "grad_norm": 1.625,
+      "learning_rate": 1.0409830882723182e-05,
+      "loss": 0.8562,
+      "step": 8571
+    },
+    {
+      "epoch": 1.457809117602932,
+      "grad_norm": 1.7578125,
+      "learning_rate": 1.0408026465142787e-05,
+      "loss": 0.938,
+      "step": 8572
+    },
+    {
+      "epoch": 1.4579805816919218,
+      "grad_norm": 1.671875,
+      "learning_rate": 1.040622203425511e-05,
+      "loss": 0.8254,
+      "step": 8573
+    },
+    {
+      "epoch": 1.4581520457809118,
+      "grad_norm": 1.6953125,
+      "learning_rate": 1.0404417590119004e-05,
+      "loss": 0.8174,
+      "step": 8574
+    },
+    {
+      "epoch": 1.4583235098699017,
+      "grad_norm": 1.7265625,
+      "learning_rate": 1.0402613132793317e-05,
+      "loss": 0.9119,
+      "step": 8575
+    },
+    {
+      "epoch": 1.4584949739588915,
+      "grad_norm": 1.6875,
+      "learning_rate": 1.0400808662336906e-05,
+      "loss": 0.8193,
+      "step": 8576
+    },
+    {
+      "epoch": 1.4586664380478813,
+      "grad_norm": 1.7265625,
+      "learning_rate": 1.0399004178808609e-05,
+      "loss": 0.8587,
+      "step": 8577
+    },
+    {
+      "epoch": 1.4588379021368711,
+      "grad_norm": 1.6015625,
+      "learning_rate": 1.0397199682267283e-05,
+      "loss": 0.8488,
+      "step": 8578
+    },
+    {
+      "epoch": 1.4590093662258612,
+      "grad_norm": 1.7421875,
+      "learning_rate": 1.0395395172771779e-05,
+      "loss": 0.8938,
+      "step": 8579
+    },
+    {
+      "epoch": 1.459180830314851,
+      "grad_norm": 1.6796875,
+      "learning_rate": 1.0393590650380951e-05,
+      "loss": 0.8643,
+      "step": 8580
+    },
+    {
+      "epoch": 1.4593522944038408,
+      "grad_norm": 1.5859375,
+      "learning_rate": 1.039178611515365e-05,
+      "loss": 0.7745,
+      "step": 8581
+    },
+    {
+      "epoch": 1.4595237584928307,
+      "grad_norm": 1.7265625,
+      "learning_rate": 1.038998156714873e-05,
+      "loss": 0.8998,
+      "step": 8582
+    },
+    {
+      "epoch": 1.4596952225818205,
+      "grad_norm": 1.734375,
+      "learning_rate": 1.0388177006425042e-05,
+      "loss": 0.9082,
+      "step": 8583
+    },
+    {
+      "epoch": 1.4598666866708103,
+      "grad_norm": 1.6484375,
+      "learning_rate": 1.0386372433041438e-05,
+      "loss": 0.8217,
+      "step": 8584
+    },
+    {
+      "epoch": 1.4600381507598001,
+      "grad_norm": 1.6953125,
+      "learning_rate": 1.0384567847056776e-05,
+      "loss": 0.8611,
+      "step": 8585
+    },
+    {
+      "epoch": 1.46020961484879,
+      "grad_norm": 1.6640625,
+      "learning_rate": 1.0382763248529908e-05,
+      "loss": 0.7299,
+      "step": 8586
+    },
+    {
+      "epoch": 1.46038107893778,
+      "grad_norm": 1.7734375,
+      "learning_rate": 1.0380958637519688e-05,
+      "loss": 0.9224,
+      "step": 8587
+    },
+    {
+      "epoch": 1.4605525430267698,
+      "grad_norm": 1.796875,
+      "learning_rate": 1.0379154014084973e-05,
+      "loss": 0.8753,
+      "step": 8588
+    },
+    {
+      "epoch": 1.4607240071157597,
+      "grad_norm": 1.7109375,
+      "learning_rate": 1.0377349378284618e-05,
+      "loss": 0.9347,
+      "step": 8589
+    },
+    {
+      "epoch": 1.4608954712047495,
+      "grad_norm": 1.609375,
+      "learning_rate": 1.0375544730177477e-05,
+      "loss": 0.8349,
+      "step": 8590
+    },
+    {
+      "epoch": 1.4610669352937395,
+      "grad_norm": 1.75,
+      "learning_rate": 1.0373740069822411e-05,
+      "loss": 0.8822,
+      "step": 8591
+    },
+    {
+      "epoch": 1.4612383993827294,
+      "grad_norm": 1.765625,
+      "learning_rate": 1.0371935397278272e-05,
+      "loss": 0.896,
+      "step": 8592
+    },
+    {
+      "epoch": 1.4614098634717192,
+      "grad_norm": 1.7265625,
+      "learning_rate": 1.0370130712603919e-05,
+      "loss": 0.8061,
+      "step": 8593
+    },
+    {
+      "epoch": 1.461581327560709,
+      "grad_norm": 1.71875,
+      "learning_rate": 1.0368326015858212e-05,
+      "loss": 0.8535,
+      "step": 8594
+    },
+    {
+      "epoch": 1.4617527916496988,
+      "grad_norm": 1.5703125,
+      "learning_rate": 1.0366521307100007e-05,
+      "loss": 0.773,
+      "step": 8595
+    },
+    {
+      "epoch": 1.4619242557386887,
+      "grad_norm": 1.6640625,
+      "learning_rate": 1.036471658638816e-05,
+      "loss": 0.8229,
+      "step": 8596
+    },
+    {
+      "epoch": 1.4620957198276785,
+      "grad_norm": 1.6953125,
+      "learning_rate": 1.036291185378153e-05,
+      "loss": 0.9088,
+      "step": 8597
+    },
+    {
+      "epoch": 1.4622671839166683,
+      "grad_norm": 1.7890625,
+      "learning_rate": 1.036110710933898e-05,
+      "loss": 0.8838,
+      "step": 8598
+    },
+    {
+      "epoch": 1.4624386480056584,
+      "grad_norm": 1.6953125,
+      "learning_rate": 1.0359302353119365e-05,
+      "loss": 0.8271,
+      "step": 8599
+    },
+    {
+      "epoch": 1.4626101120946482,
+      "grad_norm": 1.7109375,
+      "learning_rate": 1.035749758518155e-05,
+      "loss": 0.8883,
+      "step": 8600
+    },
+    {
+      "epoch": 1.462781576183638,
+      "grad_norm": 1.6953125,
+      "learning_rate": 1.0355692805584394e-05,
+      "loss": 0.8199,
+      "step": 8601
+    },
+    {
+      "epoch": 1.4629530402726278,
+      "grad_norm": 1.6328125,
+      "learning_rate": 1.0353888014386753e-05,
+      "loss": 0.8284,
+      "step": 8602
+    },
+    {
+      "epoch": 1.463124504361618,
+      "grad_norm": 1.65625,
+      "learning_rate": 1.0352083211647493e-05,
+      "loss": 0.8684,
+      "step": 8603
+    },
+    {
+      "epoch": 1.4632959684506077,
+      "grad_norm": 1.640625,
+      "learning_rate": 1.0350278397425473e-05,
+      "loss": 0.7954,
+      "step": 8604
+    },
+    {
+      "epoch": 1.4634674325395975,
+      "grad_norm": 1.6015625,
+      "learning_rate": 1.0348473571779556e-05,
+      "loss": 0.8125,
+      "step": 8605
+    },
+    {
+      "epoch": 1.4636388966285874,
+      "grad_norm": 1.6953125,
+      "learning_rate": 1.0346668734768603e-05,
+      "loss": 0.9514,
+      "step": 8606
+    },
+    {
+      "epoch": 1.4638103607175772,
+      "grad_norm": 1.6484375,
+      "learning_rate": 1.0344863886451476e-05,
+      "loss": 0.8339,
+      "step": 8607
+    },
+    {
+      "epoch": 1.463981824806567,
+      "grad_norm": 1.734375,
+      "learning_rate": 1.0343059026887039e-05,
+      "loss": 0.7786,
+      "step": 8608
+    },
+    {
+      "epoch": 1.4641532888955568,
+      "grad_norm": 1.6640625,
+      "learning_rate": 1.0341254156134157e-05,
+      "loss": 0.8554,
+      "step": 8609
+    },
+    {
+      "epoch": 1.4643247529845467,
+      "grad_norm": 1.5546875,
+      "learning_rate": 1.033944927425169e-05,
+      "loss": 0.8097,
+      "step": 8610
+    },
+    {
+      "epoch": 1.4644962170735367,
+      "grad_norm": 1.5625,
+      "learning_rate": 1.0337644381298508e-05,
+      "loss": 0.8734,
+      "step": 8611
+    },
+    {
+      "epoch": 1.4646676811625265,
+      "grad_norm": 1.65625,
+      "learning_rate": 1.033583947733347e-05,
+      "loss": 0.8432,
+      "step": 8612
+    },
+    {
+      "epoch": 1.4648391452515164,
+      "grad_norm": 1.6640625,
+      "learning_rate": 1.0334034562415438e-05,
+      "loss": 0.8445,
+      "step": 8613
+    },
+    {
+      "epoch": 1.4650106093405062,
+      "grad_norm": 1.6953125,
+      "learning_rate": 1.0332229636603283e-05,
+      "loss": 0.8689,
+      "step": 8614
+    },
+    {
+      "epoch": 1.4651820734294962,
+      "grad_norm": 1.671875,
+      "learning_rate": 1.033042469995587e-05,
+      "loss": 0.8642,
+      "step": 8615
+    },
+    {
+      "epoch": 1.465353537518486,
+      "grad_norm": 1.75,
+      "learning_rate": 1.0328619752532063e-05,
+      "loss": 0.989,
+      "step": 8616
+    },
+    {
+      "epoch": 1.465525001607476,
+      "grad_norm": 1.640625,
+      "learning_rate": 1.0326814794390728e-05,
+      "loss": 0.9084,
+      "step": 8617
+    },
+    {
+      "epoch": 1.4656964656964657,
+      "grad_norm": 1.6953125,
+      "learning_rate": 1.032500982559073e-05,
+      "loss": 0.8981,
+      "step": 8618
+    },
+    {
+      "epoch": 1.4658679297854555,
+      "grad_norm": 1.6328125,
+      "learning_rate": 1.032320484619094e-05,
+      "loss": 0.8218,
+      "step": 8619
+    },
+    {
+      "epoch": 1.4660393938744454,
+      "grad_norm": 1.671875,
+      "learning_rate": 1.032139985625022e-05,
+      "loss": 0.8495,
+      "step": 8620
+    },
+    {
+      "epoch": 1.4662108579634352,
+      "grad_norm": 1.6171875,
+      "learning_rate": 1.0319594855827444e-05,
+      "loss": 0.8132,
+      "step": 8621
+    },
+    {
+      "epoch": 1.466382322052425,
+      "grad_norm": 1.7421875,
+      "learning_rate": 1.0317789844981474e-05,
+      "loss": 0.8734,
+      "step": 8622
+    },
+    {
+      "epoch": 1.466553786141415,
+      "grad_norm": 1.703125,
+      "learning_rate": 1.0315984823771183e-05,
+      "loss": 0.7964,
+      "step": 8623
+    },
+    {
+      "epoch": 1.466725250230405,
+      "grad_norm": 1.671875,
+      "learning_rate": 1.0314179792255436e-05,
+      "loss": 0.9034,
+      "step": 8624
+    },
+    {
+      "epoch": 1.4668967143193947,
+      "grad_norm": 1.71875,
+      "learning_rate": 1.0312374750493101e-05,
+      "loss": 0.8649,
+      "step": 8625
+    },
+    {
+      "epoch": 1.4670681784083845,
+      "grad_norm": 1.625,
+      "learning_rate": 1.031056969854305e-05,
+      "loss": 0.8618,
+      "step": 8626
+    },
+    {
+      "epoch": 1.4672396424973744,
+      "grad_norm": 1.6015625,
+      "learning_rate": 1.030876463646415e-05,
+      "loss": 0.8017,
+      "step": 8627
+    },
+    {
+      "epoch": 1.4674111065863644,
+      "grad_norm": 1.75,
+      "learning_rate": 1.0306959564315278e-05,
+      "loss": 0.8205,
+      "step": 8628
+    },
+    {
+      "epoch": 1.4675825706753542,
+      "grad_norm": 1.6640625,
+      "learning_rate": 1.0305154482155296e-05,
+      "loss": 0.8564,
+      "step": 8629
+    },
+    {
+      "epoch": 1.467754034764344,
+      "grad_norm": 1.75,
+      "learning_rate": 1.0303349390043076e-05,
+      "loss": 0.8954,
+      "step": 8630
+    },
+    {
+      "epoch": 1.467925498853334,
+      "grad_norm": 1.609375,
+      "learning_rate": 1.030154428803749e-05,
+      "loss": 0.8172,
+      "step": 8631
+    },
+    {
+      "epoch": 1.4680969629423237,
+      "grad_norm": 1.640625,
+      "learning_rate": 1.0299739176197409e-05,
+      "loss": 0.8819,
+      "step": 8632
+    },
+    {
+      "epoch": 1.4682684270313135,
+      "grad_norm": 1.6171875,
+      "learning_rate": 1.0297934054581707e-05,
+      "loss": 0.7962,
+      "step": 8633
+    },
+    {
+      "epoch": 1.4684398911203034,
+      "grad_norm": 1.6328125,
+      "learning_rate": 1.0296128923249251e-05,
+      "loss": 0.757,
+      "step": 8634
+    },
+    {
+      "epoch": 1.4686113552092934,
+      "grad_norm": 1.703125,
+      "learning_rate": 1.0294323782258917e-05,
+      "loss": 0.8211,
+      "step": 8635
+    },
+    {
+      "epoch": 1.4687828192982832,
+      "grad_norm": 1.7890625,
+      "learning_rate": 1.0292518631669575e-05,
+      "loss": 0.8835,
+      "step": 8636
+    },
+    {
+      "epoch": 1.468954283387273,
+      "grad_norm": 1.6953125,
+      "learning_rate": 1.02907134715401e-05,
+      "loss": 0.8118,
+      "step": 8637
+    },
+    {
+      "epoch": 1.469125747476263,
+      "grad_norm": 1.6796875,
+      "learning_rate": 1.0288908301929364e-05,
+      "loss": 0.8418,
+      "step": 8638
+    },
+    {
+      "epoch": 1.4692972115652527,
+      "grad_norm": 1.7265625,
+      "learning_rate": 1.0287103122896237e-05,
+      "loss": 0.8533,
+      "step": 8639
+    },
+    {
+      "epoch": 1.4694686756542428,
+      "grad_norm": 1.7109375,
+      "learning_rate": 1.0285297934499604e-05,
+      "loss": 0.8903,
+      "step": 8640
+    },
+    {
+      "epoch": 1.4696401397432326,
+      "grad_norm": 1.7890625,
+      "learning_rate": 1.0283492736798327e-05,
+      "loss": 0.8225,
+      "step": 8641
+    },
+    {
+      "epoch": 1.4698116038322224,
+      "grad_norm": 1.65625,
+      "learning_rate": 1.028168752985128e-05,
+      "loss": 0.9088,
+      "step": 8642
+    },
+    {
+      "epoch": 1.4699830679212122,
+      "grad_norm": 1.6015625,
+      "learning_rate": 1.0279882313717346e-05,
+      "loss": 0.7967,
+      "step": 8643
+    },
+    {
+      "epoch": 1.470154532010202,
+      "grad_norm": 1.6875,
+      "learning_rate": 1.0278077088455396e-05,
+      "loss": 0.8248,
+      "step": 8644
+    },
+    {
+      "epoch": 1.470325996099192,
+      "grad_norm": 1.6640625,
+      "learning_rate": 1.0276271854124303e-05,
+      "loss": 0.8308,
+      "step": 8645
+    },
+    {
+      "epoch": 1.4704974601881817,
+      "grad_norm": 1.7265625,
+      "learning_rate": 1.0274466610782948e-05,
+      "loss": 0.8704,
+      "step": 8646
+    },
+    {
+      "epoch": 1.4706689242771718,
+      "grad_norm": 1.7109375,
+      "learning_rate": 1.0272661358490199e-05,
+      "loss": 0.934,
+      "step": 8647
+    },
+    {
+      "epoch": 1.4708403883661616,
+      "grad_norm": 1.875,
+      "learning_rate": 1.0270856097304939e-05,
+      "loss": 0.9382,
+      "step": 8648
+    },
+    {
+      "epoch": 1.4710118524551514,
+      "grad_norm": 1.78125,
+      "learning_rate": 1.026905082728604e-05,
+      "loss": 0.9287,
+      "step": 8649
+    },
+    {
+      "epoch": 1.4711833165441413,
+      "grad_norm": 1.671875,
+      "learning_rate": 1.0267245548492382e-05,
+      "loss": 0.8068,
+      "step": 8650
+    },
+    {
+      "epoch": 1.471354780633131,
+      "grad_norm": 1.6796875,
+      "learning_rate": 1.0265440260982841e-05,
+      "loss": 0.8128,
+      "step": 8651
+    },
+    {
+      "epoch": 1.4715262447221211,
+      "grad_norm": 1.6875,
+      "learning_rate": 1.0263634964816293e-05,
+      "loss": 0.9054,
+      "step": 8652
+    },
+    {
+      "epoch": 1.471697708811111,
+      "grad_norm": 1.6875,
+      "learning_rate": 1.0261829660051615e-05,
+      "loss": 0.8416,
+      "step": 8653
+    },
+    {
+      "epoch": 1.4718691729001008,
+      "grad_norm": 1.8125,
+      "learning_rate": 1.0260024346747683e-05,
+      "loss": 0.9562,
+      "step": 8654
+    },
+    {
+      "epoch": 1.4720406369890906,
+      "grad_norm": 1.6640625,
+      "learning_rate": 1.0258219024963382e-05,
+      "loss": 0.8248,
+      "step": 8655
+    },
+    {
+      "epoch": 1.4722121010780804,
+      "grad_norm": 1.6875,
+      "learning_rate": 1.0256413694757585e-05,
+      "loss": 0.9124,
+      "step": 8656
+    },
+    {
+      "epoch": 1.4723835651670703,
+      "grad_norm": 1.5859375,
+      "learning_rate": 1.0254608356189176e-05,
+      "loss": 0.8353,
+      "step": 8657
+    },
+    {
+      "epoch": 1.47255502925606,
+      "grad_norm": 1.75,
+      "learning_rate": 1.0252803009317025e-05,
+      "loss": 0.873,
+      "step": 8658
+    },
+    {
+      "epoch": 1.4727264933450501,
+      "grad_norm": 1.5625,
+      "learning_rate": 1.0250997654200014e-05,
+      "loss": 0.75,
+      "step": 8659
+    },
+    {
+      "epoch": 1.47289795743404,
+      "grad_norm": 1.671875,
+      "learning_rate": 1.0249192290897028e-05,
+      "loss": 0.9558,
+      "step": 8660
+    },
+    {
+      "epoch": 1.4730694215230298,
+      "grad_norm": 1.6796875,
+      "learning_rate": 1.0247386919466942e-05,
+      "loss": 0.8325,
+      "step": 8661
+    },
+    {
+      "epoch": 1.4732408856120196,
+      "grad_norm": 1.7578125,
+      "learning_rate": 1.0245581539968637e-05,
+      "loss": 0.9022,
+      "step": 8662
+    },
+    {
+      "epoch": 1.4734123497010094,
+      "grad_norm": 1.640625,
+      "learning_rate": 1.0243776152460993e-05,
+      "loss": 0.7822,
+      "step": 8663
+    },
+    {
+      "epoch": 1.4735838137899995,
+      "grad_norm": 1.625,
+      "learning_rate": 1.024197075700289e-05,
+      "loss": 0.8517,
+      "step": 8664
+    },
+    {
+      "epoch": 1.4737552778789893,
+      "grad_norm": 1.6796875,
+      "learning_rate": 1.0240165353653211e-05,
+      "loss": 0.859,
+      "step": 8665
+    },
+    {
+      "epoch": 1.4739267419679791,
+      "grad_norm": 1.6953125,
+      "learning_rate": 1.0238359942470834e-05,
+      "loss": 0.8532,
+      "step": 8666
+    },
+    {
+      "epoch": 1.474098206056969,
+      "grad_norm": 1.625,
+      "learning_rate": 1.0236554523514644e-05,
+      "loss": 0.7895,
+      "step": 8667
+    },
+    {
+      "epoch": 1.4742696701459588,
+      "grad_norm": 1.6875,
+      "learning_rate": 1.023474909684352e-05,
+      "loss": 0.8081,
+      "step": 8668
+    },
+    {
+      "epoch": 1.4744411342349486,
+      "grad_norm": 1.71875,
+      "learning_rate": 1.0232943662516343e-05,
+      "loss": 0.8918,
+      "step": 8669
+    },
+    {
+      "epoch": 1.4746125983239384,
+      "grad_norm": 1.703125,
+      "learning_rate": 1.0231138220591996e-05,
+      "loss": 0.9468,
+      "step": 8670
+    },
+    {
+      "epoch": 1.4747840624129283,
+      "grad_norm": 1.7734375,
+      "learning_rate": 1.022933277112936e-05,
+      "loss": 0.8936,
+      "step": 8671
+    },
+    {
+      "epoch": 1.4749555265019183,
+      "grad_norm": 1.8359375,
+      "learning_rate": 1.022752731418732e-05,
+      "loss": 1.0092,
+      "step": 8672
+    },
+    {
+      "epoch": 1.4751269905909081,
+      "grad_norm": 1.6953125,
+      "learning_rate": 1.0225721849824754e-05,
+      "loss": 0.8682,
+      "step": 8673
+    },
+    {
+      "epoch": 1.475298454679898,
+      "grad_norm": 1.6875,
+      "learning_rate": 1.0223916378100555e-05,
+      "loss": 0.8251,
+      "step": 8674
+    },
+    {
+      "epoch": 1.4754699187688878,
+      "grad_norm": 1.734375,
+      "learning_rate": 1.0222110899073597e-05,
+      "loss": 0.8766,
+      "step": 8675
+    },
+    {
+      "epoch": 1.4756413828578778,
+      "grad_norm": 1.6953125,
+      "learning_rate": 1.0220305412802765e-05,
+      "loss": 0.8304,
+      "step": 8676
+    },
+    {
+      "epoch": 1.4758128469468677,
+      "grad_norm": 1.671875,
+      "learning_rate": 1.0218499919346945e-05,
+      "loss": 0.8465,
+      "step": 8677
+    },
+    {
+      "epoch": 1.4759843110358575,
+      "grad_norm": 1.6640625,
+      "learning_rate": 1.0216694418765019e-05,
+      "loss": 0.8278,
+      "step": 8678
+    },
+    {
+      "epoch": 1.4761557751248473,
+      "grad_norm": 1.8203125,
+      "learning_rate": 1.0214888911115872e-05,
+      "loss": 0.9232,
+      "step": 8679
+    },
+    {
+      "epoch": 1.4763272392138371,
+      "grad_norm": 1.671875,
+      "learning_rate": 1.021308339645839e-05,
+      "loss": 0.864,
+      "step": 8680
+    },
+    {
+      "epoch": 1.476498703302827,
+      "grad_norm": 1.6328125,
+      "learning_rate": 1.0211277874851453e-05,
+      "loss": 0.7689,
+      "step": 8681
+    },
+    {
+      "epoch": 1.4766701673918168,
+      "grad_norm": 1.71875,
+      "learning_rate": 1.020947234635395e-05,
+      "loss": 0.9219,
+      "step": 8682
+    },
+    {
+      "epoch": 1.4768416314808066,
+      "grad_norm": 1.6328125,
+      "learning_rate": 1.0207666811024766e-05,
+      "loss": 0.8304,
+      "step": 8683
+    },
+    {
+      "epoch": 1.4770130955697967,
+      "grad_norm": 1.765625,
+      "learning_rate": 1.0205861268922786e-05,
+      "loss": 0.931,
+      "step": 8684
+    },
+    {
+      "epoch": 1.4771845596587865,
+      "grad_norm": 1.671875,
+      "learning_rate": 1.0204055720106891e-05,
+      "loss": 0.8371,
+      "step": 8685
+    },
+    {
+      "epoch": 1.4773560237477763,
+      "grad_norm": 1.6328125,
+      "learning_rate": 1.0202250164635977e-05,
+      "loss": 0.8339,
+      "step": 8686
+    },
+    {
+      "epoch": 1.4775274878367661,
+      "grad_norm": 1.71875,
+      "learning_rate": 1.020044460256892e-05,
+      "loss": 0.8499,
+      "step": 8687
+    },
+    {
+      "epoch": 1.4776989519257562,
+      "grad_norm": 1.703125,
+      "learning_rate": 1.0198639033964608e-05,
+      "loss": 0.8418,
+      "step": 8688
+    },
+    {
+      "epoch": 1.477870416014746,
+      "grad_norm": 1.640625,
+      "learning_rate": 1.0196833458881931e-05,
+      "loss": 0.8294,
+      "step": 8689
+    },
+    {
+      "epoch": 1.4780418801037358,
+      "grad_norm": 1.65625,
+      "learning_rate": 1.0195027877379773e-05,
+      "loss": 0.7643,
+      "step": 8690
+    },
+    {
+      "epoch": 1.4782133441927257,
+      "grad_norm": 1.7421875,
+      "learning_rate": 1.0193222289517021e-05,
+      "loss": 0.8708,
+      "step": 8691
+    },
+    {
+      "epoch": 1.4783848082817155,
+      "grad_norm": 1.7109375,
+      "learning_rate": 1.0191416695352564e-05,
+      "loss": 0.8079,
+      "step": 8692
+    },
+    {
+      "epoch": 1.4785562723707053,
+      "grad_norm": 1.671875,
+      "learning_rate": 1.0189611094945287e-05,
+      "loss": 0.8953,
+      "step": 8693
+    },
+    {
+      "epoch": 1.4787277364596951,
+      "grad_norm": 1.6953125,
+      "learning_rate": 1.0187805488354079e-05,
+      "loss": 0.8435,
+      "step": 8694
+    },
+    {
+      "epoch": 1.478899200548685,
+      "grad_norm": 1.609375,
+      "learning_rate": 1.0185999875637827e-05,
+      "loss": 0.7899,
+      "step": 8695
+    },
+    {
+      "epoch": 1.479070664637675,
+      "grad_norm": 1.671875,
+      "learning_rate": 1.0184194256855418e-05,
+      "loss": 0.8195,
+      "step": 8696
+    },
+    {
+      "epoch": 1.4792421287266648,
+      "grad_norm": 1.625,
+      "learning_rate": 1.0182388632065738e-05,
+      "loss": 0.8243,
+      "step": 8697
+    },
+    {
+      "epoch": 1.4794135928156547,
+      "grad_norm": 1.78125,
+      "learning_rate": 1.0180583001327684e-05,
+      "loss": 0.9301,
+      "step": 8698
+    },
+    {
+      "epoch": 1.4795850569046445,
+      "grad_norm": 1.71875,
+      "learning_rate": 1.0178777364700134e-05,
+      "loss": 0.8632,
+      "step": 8699
+    },
+    {
+      "epoch": 1.4797565209936345,
+      "grad_norm": 1.7265625,
+      "learning_rate": 1.0176971722241982e-05,
+      "loss": 0.8346,
+      "step": 8700
+    },
+    {
+      "epoch": 1.4799279850826244,
+      "grad_norm": 1.6640625,
+      "learning_rate": 1.0175166074012113e-05,
+      "loss": 0.8498,
+      "step": 8701
+    },
+    {
+      "epoch": 1.4800994491716142,
+      "grad_norm": 1.7734375,
+      "learning_rate": 1.0173360420069421e-05,
+      "loss": 0.8662,
+      "step": 8702
+    },
+    {
+      "epoch": 1.480270913260604,
+      "grad_norm": 1.703125,
+      "learning_rate": 1.0171554760472796e-05,
+      "loss": 0.811,
+      "step": 8703
+    },
+    {
+      "epoch": 1.4804423773495938,
+      "grad_norm": 1.7265625,
+      "learning_rate": 1.0169749095281118e-05,
+      "loss": 0.9006,
+      "step": 8704
+    },
+    {
+      "epoch": 1.4806138414385837,
+      "grad_norm": 1.8046875,
+      "learning_rate": 1.0167943424553288e-05,
+      "loss": 0.9017,
+      "step": 8705
+    },
+    {
+      "epoch": 1.4807853055275735,
+      "grad_norm": 1.6171875,
+      "learning_rate": 1.0166137748348188e-05,
+      "loss": 0.8064,
+      "step": 8706
+    },
+    {
+      "epoch": 1.4809567696165633,
+      "grad_norm": 1.6640625,
+      "learning_rate": 1.0164332066724712e-05,
+      "loss": 0.8161,
+      "step": 8707
+    },
+    {
+      "epoch": 1.4811282337055534,
+      "grad_norm": 1.6796875,
+      "learning_rate": 1.0162526379741746e-05,
+      "loss": 0.8809,
+      "step": 8708
+    },
+    {
+      "epoch": 1.4812996977945432,
+      "grad_norm": 1.6484375,
+      "learning_rate": 1.0160720687458183e-05,
+      "loss": 0.7899,
+      "step": 8709
+    },
+    {
+      "epoch": 1.481471161883533,
+      "grad_norm": 1.625,
+      "learning_rate": 1.0158914989932915e-05,
+      "loss": 0.8231,
+      "step": 8710
+    },
+    {
+      "epoch": 1.4816426259725228,
+      "grad_norm": 1.6171875,
+      "learning_rate": 1.015710928722483e-05,
+      "loss": 0.8379,
+      "step": 8711
+    },
+    {
+      "epoch": 1.4818140900615129,
+      "grad_norm": 1.671875,
+      "learning_rate": 1.0155303579392819e-05,
+      "loss": 0.8668,
+      "step": 8712
+    },
+    {
+      "epoch": 1.4819855541505027,
+      "grad_norm": 1.734375,
+      "learning_rate": 1.0153497866495774e-05,
+      "loss": 0.897,
+      "step": 8713
+    },
+    {
+      "epoch": 1.4821570182394925,
+      "grad_norm": 1.7109375,
+      "learning_rate": 1.0151692148592584e-05,
+      "loss": 0.8135,
+      "step": 8714
+    },
+    {
+      "epoch": 1.4823284823284824,
+      "grad_norm": 1.7109375,
+      "learning_rate": 1.0149886425742145e-05,
+      "loss": 0.827,
+      "step": 8715
+    },
+    {
+      "epoch": 1.4824999464174722,
+      "grad_norm": 1.734375,
+      "learning_rate": 1.0148080698003347e-05,
+      "loss": 0.8225,
+      "step": 8716
+    },
+    {
+      "epoch": 1.482671410506462,
+      "grad_norm": 1.71875,
+      "learning_rate": 1.0146274965435074e-05,
+      "loss": 0.9086,
+      "step": 8717
+    },
+    {
+      "epoch": 1.4828428745954518,
+      "grad_norm": 1.78125,
+      "learning_rate": 1.0144469228096228e-05,
+      "loss": 0.8494,
+      "step": 8718
+    },
+    {
+      "epoch": 1.4830143386844417,
+      "grad_norm": 1.6484375,
+      "learning_rate": 1.0142663486045692e-05,
+      "loss": 0.79,
+      "step": 8719
+    },
+    {
+      "epoch": 1.4831858027734317,
+      "grad_norm": 1.625,
+      "learning_rate": 1.0140857739342365e-05,
+      "loss": 0.8612,
+      "step": 8720
+    },
+    {
+      "epoch": 1.4833572668624215,
+      "grad_norm": 1.6328125,
+      "learning_rate": 1.0139051988045136e-05,
+      "loss": 0.8386,
+      "step": 8721
+    },
+    {
+      "epoch": 1.4835287309514114,
+      "grad_norm": 1.6796875,
+      "learning_rate": 1.0137246232212901e-05,
+      "loss": 0.8252,
+      "step": 8722
+    },
+    {
+      "epoch": 1.4837001950404012,
+      "grad_norm": 1.6640625,
+      "learning_rate": 1.0135440471904546e-05,
+      "loss": 0.7856,
+      "step": 8723
+    },
+    {
+      "epoch": 1.483871659129391,
+      "grad_norm": 1.7578125,
+      "learning_rate": 1.0133634707178968e-05,
+      "loss": 0.917,
+      "step": 8724
+    },
+    {
+      "epoch": 1.484043123218381,
+      "grad_norm": 1.7421875,
+      "learning_rate": 1.0131828938095059e-05,
+      "loss": 0.8445,
+      "step": 8725
+    },
+    {
+      "epoch": 1.4842145873073709,
+      "grad_norm": 1.6171875,
+      "learning_rate": 1.0130023164711715e-05,
+      "loss": 0.8146,
+      "step": 8726
+    },
+    {
+      "epoch": 1.4843860513963607,
+      "grad_norm": 1.6640625,
+      "learning_rate": 1.0128217387087825e-05,
+      "loss": 0.8103,
+      "step": 8727
+    },
+    {
+      "epoch": 1.4845575154853505,
+      "grad_norm": 1.609375,
+      "learning_rate": 1.0126411605282279e-05,
+      "loss": 0.8867,
+      "step": 8728
+    },
+    {
+      "epoch": 1.4847289795743404,
+      "grad_norm": 1.75,
+      "learning_rate": 1.0124605819353974e-05,
+      "loss": 0.96,
+      "step": 8729
+    },
+    {
+      "epoch": 1.4849004436633302,
+      "grad_norm": 1.703125,
+      "learning_rate": 1.0122800029361808e-05,
+      "loss": 0.9119,
+      "step": 8730
+    },
+    {
+      "epoch": 1.48507190775232,
+      "grad_norm": 1.703125,
+      "learning_rate": 1.012099423536467e-05,
+      "loss": 0.8755,
+      "step": 8731
+    },
+    {
+      "epoch": 1.48524337184131,
+      "grad_norm": 1.703125,
+      "learning_rate": 1.0119188437421456e-05,
+      "loss": 0.8813,
+      "step": 8732
+    },
+    {
+      "epoch": 1.4854148359302999,
+      "grad_norm": 1.7265625,
+      "learning_rate": 1.0117382635591055e-05,
+      "loss": 0.8764,
+      "step": 8733
+    },
+    {
+      "epoch": 1.4855863000192897,
+      "grad_norm": 1.7734375,
+      "learning_rate": 1.0115576829932365e-05,
+      "loss": 0.884,
+      "step": 8734
+    },
+    {
+      "epoch": 1.4857577641082795,
+      "grad_norm": 1.6640625,
+      "learning_rate": 1.0113771020504281e-05,
+      "loss": 0.8937,
+      "step": 8735
+    },
+    {
+      "epoch": 1.4859292281972694,
+      "grad_norm": 1.7734375,
+      "learning_rate": 1.0111965207365695e-05,
+      "loss": 0.8606,
+      "step": 8736
+    },
+    {
+      "epoch": 1.4861006922862594,
+      "grad_norm": 1.6171875,
+      "learning_rate": 1.0110159390575501e-05,
+      "loss": 0.8762,
+      "step": 8737
+    },
+    {
+      "epoch": 1.4862721563752492,
+      "grad_norm": 1.703125,
+      "learning_rate": 1.0108353570192595e-05,
+      "loss": 0.8707,
+      "step": 8738
+    },
+    {
+      "epoch": 1.486443620464239,
+      "grad_norm": 1.7578125,
+      "learning_rate": 1.010654774627587e-05,
+      "loss": 0.8746,
+      "step": 8739
+    },
+    {
+      "epoch": 1.4866150845532289,
+      "grad_norm": 1.640625,
+      "learning_rate": 1.0104741918884222e-05,
+      "loss": 0.8249,
+      "step": 8740
+    },
+    {
+      "epoch": 1.4867865486422187,
+      "grad_norm": 1.671875,
+      "learning_rate": 1.0102936088076547e-05,
+      "loss": 0.8829,
+      "step": 8741
+    },
+    {
+      "epoch": 1.4869580127312085,
+      "grad_norm": 1.6796875,
+      "learning_rate": 1.0101130253911737e-05,
+      "loss": 0.8185,
+      "step": 8742
+    },
+    {
+      "epoch": 1.4871294768201984,
+      "grad_norm": 1.6328125,
+      "learning_rate": 1.009932441644869e-05,
+      "loss": 0.9014,
+      "step": 8743
+    },
+    {
+      "epoch": 1.4873009409091884,
+      "grad_norm": 1.71875,
+      "learning_rate": 1.0097518575746301e-05,
+      "loss": 0.8529,
+      "step": 8744
+    },
+    {
+      "epoch": 1.4874724049981782,
+      "grad_norm": 1.6484375,
+      "learning_rate": 1.0095712731863463e-05,
+      "loss": 0.8076,
+      "step": 8745
+    },
+    {
+      "epoch": 1.487643869087168,
+      "grad_norm": 1.75,
+      "learning_rate": 1.009390688485907e-05,
+      "loss": 0.8656,
+      "step": 8746
+    },
+    {
+      "epoch": 1.4878153331761579,
+      "grad_norm": 1.671875,
+      "learning_rate": 1.009210103479202e-05,
+      "loss": 0.7715,
+      "step": 8747
+    },
+    {
+      "epoch": 1.4879867972651477,
+      "grad_norm": 1.7890625,
+      "learning_rate": 1.0090295181721209e-05,
+      "loss": 0.9446,
+      "step": 8748
+    },
+    {
+      "epoch": 1.4881582613541378,
+      "grad_norm": 1.640625,
+      "learning_rate": 1.0088489325705535e-05,
+      "loss": 0.8537,
+      "step": 8749
+    },
+    {
+      "epoch": 1.4883297254431276,
+      "grad_norm": 1.6875,
+      "learning_rate": 1.0086683466803889e-05,
+      "loss": 0.9564,
+      "step": 8750
+    },
+    {
+      "epoch": 1.4885011895321174,
+      "grad_norm": 1.734375,
+      "learning_rate": 1.0084877605075168e-05,
+      "loss": 0.9122,
+      "step": 8751
+    },
+    {
+      "epoch": 1.4886726536211072,
+      "grad_norm": 1.7421875,
+      "learning_rate": 1.0083071740578269e-05,
+      "loss": 0.8204,
+      "step": 8752
+    },
+    {
+      "epoch": 1.488844117710097,
+      "grad_norm": 1.7421875,
+      "learning_rate": 1.0081265873372088e-05,
+      "loss": 0.7796,
+      "step": 8753
+    },
+    {
+      "epoch": 1.4890155817990869,
+      "grad_norm": 1.7109375,
+      "learning_rate": 1.0079460003515522e-05,
+      "loss": 0.8749,
+      "step": 8754
+    },
+    {
+      "epoch": 1.4891870458880767,
+      "grad_norm": 1.65625,
+      "learning_rate": 1.0077654131067467e-05,
+      "loss": 0.8704,
+      "step": 8755
+    },
+    {
+      "epoch": 1.4893585099770668,
+      "grad_norm": 1.796875,
+      "learning_rate": 1.0075848256086814e-05,
+      "loss": 0.9248,
+      "step": 8756
+    },
+    {
+      "epoch": 1.4895299740660566,
+      "grad_norm": 1.65625,
+      "learning_rate": 1.0074042378632464e-05,
+      "loss": 0.8705,
+      "step": 8757
+    },
+    {
+      "epoch": 1.4897014381550464,
+      "grad_norm": 1.8046875,
+      "learning_rate": 1.0072236498763316e-05,
+      "loss": 0.8117,
+      "step": 8758
+    },
+    {
+      "epoch": 1.4898729022440362,
+      "grad_norm": 1.6953125,
+      "learning_rate": 1.0070430616538262e-05,
+      "loss": 0.8788,
+      "step": 8759
+    },
+    {
+      "epoch": 1.490044366333026,
+      "grad_norm": 1.7421875,
+      "learning_rate": 1.0068624732016201e-05,
+      "loss": 0.8653,
+      "step": 8760
+    },
+    {
+      "epoch": 1.490215830422016,
+      "grad_norm": 1.5859375,
+      "learning_rate": 1.0066818845256032e-05,
+      "loss": 0.7809,
+      "step": 8761
+    },
+    {
+      "epoch": 1.490387294511006,
+      "grad_norm": 1.6484375,
+      "learning_rate": 1.0065012956316647e-05,
+      "loss": 0.8333,
+      "step": 8762
+    },
+    {
+      "epoch": 1.4905587585999958,
+      "grad_norm": 1.6796875,
+      "learning_rate": 1.0063207065256944e-05,
+      "loss": 0.8556,
+      "step": 8763
+    },
+    {
+      "epoch": 1.4907302226889856,
+      "grad_norm": 1.71875,
+      "learning_rate": 1.006140117213582e-05,
+      "loss": 0.9508,
+      "step": 8764
+    },
+    {
+      "epoch": 1.4909016867779754,
+      "grad_norm": 1.6953125,
+      "learning_rate": 1.0059595277012173e-05,
+      "loss": 0.8847,
+      "step": 8765
+    },
+    {
+      "epoch": 1.4910731508669652,
+      "grad_norm": 1.8359375,
+      "learning_rate": 1.00577893799449e-05,
+      "loss": 0.9396,
+      "step": 8766
+    },
+    {
+      "epoch": 1.491244614955955,
+      "grad_norm": 1.6953125,
+      "learning_rate": 1.0055983480992897e-05,
+      "loss": 0.9099,
+      "step": 8767
+    },
+    {
+      "epoch": 1.4914160790449449,
+      "grad_norm": 1.71875,
+      "learning_rate": 1.0054177580215063e-05,
+      "loss": 0.8431,
+      "step": 8768
+    },
+    {
+      "epoch": 1.491587543133935,
+      "grad_norm": 1.7421875,
+      "learning_rate": 1.0052371677670291e-05,
+      "loss": 0.8217,
+      "step": 8769
+    },
+    {
+      "epoch": 1.4917590072229248,
+      "grad_norm": 1.640625,
+      "learning_rate": 1.0050565773417484e-05,
+      "loss": 0.8995,
+      "step": 8770
+    },
+    {
+      "epoch": 1.4919304713119146,
+      "grad_norm": 1.71875,
+      "learning_rate": 1.0048759867515538e-05,
+      "loss": 0.9212,
+      "step": 8771
+    },
+    {
+      "epoch": 1.4921019354009044,
+      "grad_norm": 1.671875,
+      "learning_rate": 1.0046953960023346e-05,
+      "loss": 0.8676,
+      "step": 8772
+    },
+    {
+      "epoch": 1.4922733994898945,
+      "grad_norm": 1.6796875,
+      "learning_rate": 1.0045148050999814e-05,
+      "loss": 0.8223,
+      "step": 8773
+    },
+    {
+      "epoch": 1.4924448635788843,
+      "grad_norm": 1.59375,
+      "learning_rate": 1.0043342140503829e-05,
+      "loss": 0.8115,
+      "step": 8774
+    },
+    {
+      "epoch": 1.492616327667874,
+      "grad_norm": 1.6328125,
+      "learning_rate": 1.0041536228594293e-05,
+      "loss": 0.8233,
+      "step": 8775
+    },
+    {
+      "epoch": 1.492787791756864,
+      "grad_norm": 1.625,
+      "learning_rate": 1.0039730315330105e-05,
+      "loss": 0.8075,
+      "step": 8776
+    },
+    {
+      "epoch": 1.4929592558458538,
+      "grad_norm": 1.6640625,
+      "learning_rate": 1.0037924400770163e-05,
+      "loss": 0.9386,
+      "step": 8777
+    },
+    {
+      "epoch": 1.4931307199348436,
+      "grad_norm": 1.6953125,
+      "learning_rate": 1.0036118484973365e-05,
+      "loss": 0.8208,
+      "step": 8778
+    },
+    {
+      "epoch": 1.4933021840238334,
+      "grad_norm": 1.625,
+      "learning_rate": 1.0034312567998605e-05,
+      "loss": 0.8615,
+      "step": 8779
+    },
+    {
+      "epoch": 1.4934736481128232,
+      "grad_norm": 1.7421875,
+      "learning_rate": 1.0032506649904783e-05,
+      "loss": 0.9546,
+      "step": 8780
+    },
+    {
+      "epoch": 1.4936451122018133,
+      "grad_norm": 1.65625,
+      "learning_rate": 1.0030700730750798e-05,
+      "loss": 0.8694,
+      "step": 8781
+    },
+    {
+      "epoch": 1.493816576290803,
+      "grad_norm": 1.5703125,
+      "learning_rate": 1.0028894810595547e-05,
+      "loss": 0.753,
+      "step": 8782
+    },
+    {
+      "epoch": 1.493988040379793,
+      "grad_norm": 1.6171875,
+      "learning_rate": 1.0027088889497924e-05,
+      "loss": 0.849,
+      "step": 8783
+    },
+    {
+      "epoch": 1.4941595044687828,
+      "grad_norm": 1.7421875,
+      "learning_rate": 1.0025282967516833e-05,
+      "loss": 0.8654,
+      "step": 8784
+    },
+    {
+      "epoch": 1.4943309685577728,
+      "grad_norm": 1.6484375,
+      "learning_rate": 1.0023477044711168e-05,
+      "loss": 0.8175,
+      "step": 8785
+    },
+    {
+      "epoch": 1.4945024326467626,
+      "grad_norm": 1.7265625,
+      "learning_rate": 1.0021671121139828e-05,
+      "loss": 0.8664,
+      "step": 8786
+    },
+    {
+      "epoch": 1.4946738967357525,
+      "grad_norm": 1.6953125,
+      "learning_rate": 1.0019865196861714e-05,
+      "loss": 0.8677,
+      "step": 8787
+    },
+    {
+      "epoch": 1.4948453608247423,
+      "grad_norm": 1.625,
+      "learning_rate": 1.0018059271935719e-05,
+      "loss": 0.8039,
+      "step": 8788
+    },
+    {
+      "epoch": 1.4950168249137321,
+      "grad_norm": 1.796875,
+      "learning_rate": 1.0016253346420742e-05,
+      "loss": 0.8663,
+      "step": 8789
+    },
+    {
+      "epoch": 1.495188289002722,
+      "grad_norm": 1.7109375,
+      "learning_rate": 1.0014447420375686e-05,
+      "loss": 0.8492,
+      "step": 8790
+    },
+    {
+      "epoch": 1.4953597530917118,
+      "grad_norm": 1.6796875,
+      "learning_rate": 1.0012641493859441e-05,
+      "loss": 0.8019,
+      "step": 8791
+    },
+    {
+      "epoch": 1.4955312171807016,
+      "grad_norm": 1.6640625,
+      "learning_rate": 1.0010835566930912e-05,
+      "loss": 0.8758,
+      "step": 8792
+    },
+    {
+      "epoch": 1.4957026812696916,
+      "grad_norm": 1.6953125,
+      "learning_rate": 1.0009029639648993e-05,
+      "loss": 0.8206,
+      "step": 8793
+    },
+    {
+      "epoch": 1.4958741453586815,
+      "grad_norm": 1.671875,
+      "learning_rate": 1.000722371207258e-05,
+      "loss": 0.895,
+      "step": 8794
+    },
+    {
+      "epoch": 1.4960456094476713,
+      "grad_norm": 1.640625,
+      "learning_rate": 1.0005417784260578e-05,
+      "loss": 0.8039,
+      "step": 8795
+    },
+    {
+      "epoch": 1.4962170735366611,
+      "grad_norm": 1.75,
+      "learning_rate": 1.0003611856271885e-05,
+      "loss": 0.8428,
+      "step": 8796
+    },
+    {
+      "epoch": 1.4963885376256512,
+      "grad_norm": 1.6953125,
+      "learning_rate": 1.0001805928165392e-05,
+      "loss": 0.8586,
+      "step": 8797
+    },
+    {
+      "epoch": 1.496560001714641,
+      "grad_norm": 1.671875,
+      "learning_rate": 1e-05,
+      "loss": 0.8142,
+      "step": 8798
+    },
+    {
+      "epoch": 1.4967314658036308,
+      "grad_norm": 1.6640625,
+      "learning_rate": 9.99819407183461e-06,
+      "loss": 0.835,
+      "step": 8799
+    },
+    {
+      "epoch": 1.4969029298926206,
+      "grad_norm": 1.7578125,
+      "learning_rate": 9.996388143728118e-06,
+      "loss": 0.8529,
+      "step": 8800
+    },
+    {
+      "epoch": 1.4970743939816105,
+      "grad_norm": 1.6953125,
+      "learning_rate": 9.994582215739422e-06,
+      "loss": 0.8341,
+      "step": 8801
+    },
+    {
+      "epoch": 1.4972458580706003,
+      "grad_norm": 1.7109375,
+      "learning_rate": 9.99277628792742e-06,
+      "loss": 0.8221,
+      "step": 8802
+    },
+    {
+      "epoch": 1.4974173221595901,
+      "grad_norm": 1.640625,
+      "learning_rate": 9.990970360351012e-06,
+      "loss": 0.8617,
+      "step": 8803
+    },
+    {
+      "epoch": 1.49758878624858,
+      "grad_norm": 1.7578125,
+      "learning_rate": 9.989164433069093e-06,
+      "loss": 0.8577,
+      "step": 8804
+    },
+    {
+      "epoch": 1.49776025033757,
+      "grad_norm": 1.859375,
+      "learning_rate": 9.987358506140564e-06,
+      "loss": 0.8556,
+      "step": 8805
+    },
+    {
+      "epoch": 1.4979317144265598,
+      "grad_norm": 1.8203125,
+      "learning_rate": 9.98555257962432e-06,
+      "loss": 0.8676,
+      "step": 8806
+    },
+    {
+      "epoch": 1.4981031785155496,
+      "grad_norm": 1.7109375,
+      "learning_rate": 9.983746653579262e-06,
+      "loss": 0.9136,
+      "step": 8807
+    },
+    {
+      "epoch": 1.4982746426045395,
+      "grad_norm": 1.6484375,
+      "learning_rate": 9.981940728064286e-06,
+      "loss": 0.9122,
+      "step": 8808
+    },
+    {
+      "epoch": 1.4984461066935295,
+      "grad_norm": 1.6796875,
+      "learning_rate": 9.980134803138291e-06,
+      "loss": 0.8316,
+      "step": 8809
+    },
+    {
+      "epoch": 1.4986175707825193,
+      "grad_norm": 1.71875,
+      "learning_rate": 9.978328878860174e-06,
+      "loss": 0.9408,
+      "step": 8810
+    },
+    {
+      "epoch": 1.4987890348715092,
+      "grad_norm": 1.6484375,
+      "learning_rate": 9.976522955288834e-06,
+      "loss": 0.7907,
+      "step": 8811
+    },
+    {
+      "epoch": 1.498960498960499,
+      "grad_norm": 1.6875,
+      "learning_rate": 9.974717032483169e-06,
+      "loss": 0.8882,
+      "step": 8812
+    },
+    {
+      "epoch": 1.4991319630494888,
+      "grad_norm": 1.6640625,
+      "learning_rate": 9.972911110502078e-06,
+      "loss": 0.787,
+      "step": 8813
+    },
+    {
+      "epoch": 1.4993034271384786,
+      "grad_norm": 1.71875,
+      "learning_rate": 9.971105189404455e-06,
+      "loss": 0.9062,
+      "step": 8814
+    },
+    {
+      "epoch": 1.4994748912274685,
+      "grad_norm": 1.7265625,
+      "learning_rate": 9.969299269249202e-06,
+      "loss": 0.8384,
+      "step": 8815
+    },
+    {
+      "epoch": 1.4996463553164583,
+      "grad_norm": 1.71875,
+      "learning_rate": 9.96749335009522e-06,
+      "loss": 0.8959,
+      "step": 8816
+    },
+    {
+      "epoch": 1.4998178194054483,
+      "grad_norm": 1.6328125,
+      "learning_rate": 9.9656874320014e-06,
+      "loss": 0.8492,
+      "step": 8817
+    },
+    {
+      "epoch": 1.4999892834944382,
+      "grad_norm": 1.6171875,
+      "learning_rate": 9.96388151502664e-06,
+      "loss": 0.7821,
+      "step": 8818
+    },
+    {
+      "epoch": 1.500160747583428,
+      "grad_norm": 1.609375,
+      "learning_rate": 9.96207559922984e-06,
+      "loss": 0.8641,
+      "step": 8819
+    },
+    {
+      "epoch": 1.5003322116724178,
+      "grad_norm": 1.7578125,
+      "learning_rate": 9.960269684669898e-06,
+      "loss": 0.8587,
+      "step": 8820
+    },
+    {
+      "epoch": 1.5005036757614079,
+      "grad_norm": 1.609375,
+      "learning_rate": 9.95846377140571e-06,
+      "loss": 0.8383,
+      "step": 8821
+    },
+    {
+      "epoch": 1.5006751398503977,
+      "grad_norm": 1.7109375,
+      "learning_rate": 9.956657859496175e-06,
+      "loss": 0.8629,
+      "step": 8822
+    },
+    {
+      "epoch": 1.5008466039393875,
+      "grad_norm": 1.671875,
+      "learning_rate": 9.954851949000189e-06,
+      "loss": 0.8301,
+      "step": 8823
+    },
+    {
+      "epoch": 1.5010180680283773,
+      "grad_norm": 1.640625,
+      "learning_rate": 9.953046039976654e-06,
+      "loss": 0.8623,
+      "step": 8824
+    },
+    {
+      "epoch": 1.5011895321173672,
+      "grad_norm": 1.65625,
+      "learning_rate": 9.951240132484464e-06,
+      "loss": 0.7772,
+      "step": 8825
+    },
+    {
+      "epoch": 1.501360996206357,
+      "grad_norm": 1.6484375,
+      "learning_rate": 9.949434226582516e-06,
+      "loss": 0.9161,
+      "step": 8826
+    },
+    {
+      "epoch": 1.5015324602953468,
+      "grad_norm": 1.734375,
+      "learning_rate": 9.947628322329714e-06,
+      "loss": 0.889,
+      "step": 8827
+    },
+    {
+      "epoch": 1.5017039243843366,
+      "grad_norm": 1.7421875,
+      "learning_rate": 9.945822419784944e-06,
+      "loss": 0.8393,
+      "step": 8828
+    },
+    {
+      "epoch": 1.5018753884733265,
+      "grad_norm": 1.7109375,
+      "learning_rate": 9.944016519007108e-06,
+      "loss": 0.8285,
+      "step": 8829
+    },
+    {
+      "epoch": 1.5020468525623165,
+      "grad_norm": 1.75,
+      "learning_rate": 9.942210620055102e-06,
+      "loss": 0.8412,
+      "step": 8830
+    },
+    {
+      "epoch": 1.5022183166513063,
+      "grad_norm": 1.78125,
+      "learning_rate": 9.94040472298783e-06,
+      "loss": 0.8117,
+      "step": 8831
+    },
+    {
+      "epoch": 1.5023897807402962,
+      "grad_norm": 1.75,
+      "learning_rate": 9.938598827864183e-06,
+      "loss": 0.8945,
+      "step": 8832
+    },
+    {
+      "epoch": 1.5025612448292862,
+      "grad_norm": 1.7265625,
+      "learning_rate": 9.936792934743059e-06,
+      "loss": 0.8701,
+      "step": 8833
+    },
+    {
+      "epoch": 1.502732708918276,
+      "grad_norm": 1.625,
+      "learning_rate": 9.934987043683356e-06,
+      "loss": 0.8044,
+      "step": 8834
+    },
+    {
+      "epoch": 1.5029041730072659,
+      "grad_norm": 1.765625,
+      "learning_rate": 9.933181154743971e-06,
+      "loss": 0.8532,
+      "step": 8835
+    },
+    {
+      "epoch": 1.5030756370962557,
+      "grad_norm": 1.625,
+      "learning_rate": 9.931375267983799e-06,
+      "loss": 0.8653,
+      "step": 8836
+    },
+    {
+      "epoch": 1.5032471011852455,
+      "grad_norm": 1.71875,
+      "learning_rate": 9.929569383461738e-06,
+      "loss": 0.8524,
+      "step": 8837
+    },
+    {
+      "epoch": 1.5034185652742353,
+      "grad_norm": 1.609375,
+      "learning_rate": 9.927763501236685e-06,
+      "loss": 0.8179,
+      "step": 8838
+    },
+    {
+      "epoch": 1.5035900293632252,
+      "grad_norm": 1.703125,
+      "learning_rate": 9.925957621367538e-06,
+      "loss": 0.841,
+      "step": 8839
+    },
+    {
+      "epoch": 1.503761493452215,
+      "grad_norm": 1.6796875,
+      "learning_rate": 9.92415174391319e-06,
+      "loss": 0.8189,
+      "step": 8840
+    },
+    {
+      "epoch": 1.5039329575412048,
+      "grad_norm": 1.6953125,
+      "learning_rate": 9.92234586893254e-06,
+      "loss": 0.7867,
+      "step": 8841
+    },
+    {
+      "epoch": 1.5041044216301949,
+      "grad_norm": 1.625,
+      "learning_rate": 9.920539996484482e-06,
+      "loss": 0.8511,
+      "step": 8842
+    },
+    {
+      "epoch": 1.5042758857191847,
+      "grad_norm": 1.640625,
+      "learning_rate": 9.918734126627914e-06,
+      "loss": 0.8939,
+      "step": 8843
+    },
+    {
+      "epoch": 1.5044473498081745,
+      "grad_norm": 1.7421875,
+      "learning_rate": 9.916928259421733e-06,
+      "loss": 0.8351,
+      "step": 8844
+    },
+    {
+      "epoch": 1.5046188138971646,
+      "grad_norm": 1.7109375,
+      "learning_rate": 9.915122394924834e-06,
+      "loss": 0.8679,
+      "step": 8845
+    },
+    {
+      "epoch": 1.5047902779861544,
+      "grad_norm": 1.7578125,
+      "learning_rate": 9.913316533196113e-06,
+      "loss": 0.8887,
+      "step": 8846
+    },
+    {
+      "epoch": 1.5049617420751442,
+      "grad_norm": 1.6875,
+      "learning_rate": 9.911510674294468e-06,
+      "loss": 0.8625,
+      "step": 8847
+    },
+    {
+      "epoch": 1.505133206164134,
+      "grad_norm": 1.65625,
+      "learning_rate": 9.909704818278791e-06,
+      "loss": 0.8188,
+      "step": 8848
+    },
+    {
+      "epoch": 1.5053046702531239,
+      "grad_norm": 1.703125,
+      "learning_rate": 9.907898965207982e-06,
+      "loss": 0.8792,
+      "step": 8849
+    },
+    {
+      "epoch": 1.5054761343421137,
+      "grad_norm": 1.7265625,
+      "learning_rate": 9.906093115140935e-06,
+      "loss": 0.8196,
+      "step": 8850
+    },
+    {
+      "epoch": 1.5056475984311035,
+      "grad_norm": 1.6640625,
+      "learning_rate": 9.904287268136544e-06,
+      "loss": 0.8907,
+      "step": 8851
+    },
+    {
+      "epoch": 1.5058190625200933,
+      "grad_norm": 1.640625,
+      "learning_rate": 9.902481424253704e-06,
+      "loss": 0.8197,
+      "step": 8852
+    },
+    {
+      "epoch": 1.5059905266090832,
+      "grad_norm": 1.6796875,
+      "learning_rate": 9.900675583551314e-06,
+      "loss": 0.9056,
+      "step": 8853
+    },
+    {
+      "epoch": 1.5061619906980732,
+      "grad_norm": 1.671875,
+      "learning_rate": 9.898869746088266e-06,
+      "loss": 0.8384,
+      "step": 8854
+    },
+    {
+      "epoch": 1.506333454787063,
+      "grad_norm": 1.6953125,
+      "learning_rate": 9.897063911923456e-06,
+      "loss": 0.9468,
+      "step": 8855
+    },
+    {
+      "epoch": 1.5065049188760529,
+      "grad_norm": 1.640625,
+      "learning_rate": 9.89525808111578e-06,
+      "loss": 0.828,
+      "step": 8856
+    },
+    {
+      "epoch": 1.506676382965043,
+      "grad_norm": 1.6171875,
+      "learning_rate": 9.893452253724133e-06,
+      "loss": 0.7832,
+      "step": 8857
+    },
+    {
+      "epoch": 1.5068478470540327,
+      "grad_norm": 1.640625,
+      "learning_rate": 9.891646429807409e-06,
+      "loss": 0.8636,
+      "step": 8858
+    },
+    {
+      "epoch": 1.5070193111430226,
+      "grad_norm": 1.71875,
+      "learning_rate": 9.889840609424502e-06,
+      "loss": 0.8303,
+      "step": 8859
+    },
+    {
+      "epoch": 1.5071907752320124,
+      "grad_norm": 1.703125,
+      "learning_rate": 9.888034792634308e-06,
+      "loss": 0.7927,
+      "step": 8860
+    },
+    {
+      "epoch": 1.5073622393210022,
+      "grad_norm": 1.7578125,
+      "learning_rate": 9.88622897949572e-06,
+      "loss": 0.8017,
+      "step": 8861
+    },
+    {
+      "epoch": 1.507533703409992,
+      "grad_norm": 1.796875,
+      "learning_rate": 9.884423170067639e-06,
+      "loss": 0.9169,
+      "step": 8862
+    },
+    {
+      "epoch": 1.5077051674989819,
+      "grad_norm": 1.6796875,
+      "learning_rate": 9.88261736440895e-06,
+      "loss": 0.8707,
+      "step": 8863
+    },
+    {
+      "epoch": 1.5078766315879717,
+      "grad_norm": 1.7265625,
+      "learning_rate": 9.880811562578549e-06,
+      "loss": 0.9191,
+      "step": 8864
+    },
+    {
+      "epoch": 1.5080480956769615,
+      "grad_norm": 1.734375,
+      "learning_rate": 9.879005764635334e-06,
+      "loss": 0.8502,
+      "step": 8865
+    },
+    {
+      "epoch": 1.5082195597659516,
+      "grad_norm": 1.6640625,
+      "learning_rate": 9.877199970638195e-06,
+      "loss": 0.8907,
+      "step": 8866
+    },
+    {
+      "epoch": 1.5083910238549414,
+      "grad_norm": 1.6640625,
+      "learning_rate": 9.875394180646028e-06,
+      "loss": 0.8445,
+      "step": 8867
+    },
+    {
+      "epoch": 1.5085624879439312,
+      "grad_norm": 1.7265625,
+      "learning_rate": 9.873588394717726e-06,
+      "loss": 0.87,
+      "step": 8868
+    },
+    {
+      "epoch": 1.5087339520329213,
+      "grad_norm": 1.703125,
+      "learning_rate": 9.87178261291218e-06,
+      "loss": 0.8434,
+      "step": 8869
+    },
+    {
+      "epoch": 1.508905416121911,
+      "grad_norm": 1.65625,
+      "learning_rate": 9.869976835288288e-06,
+      "loss": 0.7995,
+      "step": 8870
+    },
+    {
+      "epoch": 1.509076880210901,
+      "grad_norm": 1.640625,
+      "learning_rate": 9.868171061904941e-06,
+      "loss": 0.9007,
+      "step": 8871
+    },
+    {
+      "epoch": 1.5092483442998907,
+      "grad_norm": 1.828125,
+      "learning_rate": 9.866365292821032e-06,
+      "loss": 0.8768,
+      "step": 8872
+    },
+    {
+      "epoch": 1.5094198083888806,
+      "grad_norm": 1.7265625,
+      "learning_rate": 9.864559528095459e-06,
+      "loss": 0.8668,
+      "step": 8873
+    },
+    {
+      "epoch": 1.5095912724778704,
+      "grad_norm": 1.640625,
+      "learning_rate": 9.862753767787106e-06,
+      "loss": 0.9018,
+      "step": 8874
+    },
+    {
+      "epoch": 1.5097627365668602,
+      "grad_norm": 1.6953125,
+      "learning_rate": 9.860948011954868e-06,
+      "loss": 0.8425,
+      "step": 8875
+    },
+    {
+      "epoch": 1.50993420065585,
+      "grad_norm": 1.6875,
+      "learning_rate": 9.859142260657639e-06,
+      "loss": 0.7678,
+      "step": 8876
+    },
+    {
+      "epoch": 1.5101056647448399,
+      "grad_norm": 1.6875,
+      "learning_rate": 9.85733651395431e-06,
+      "loss": 0.8245,
+      "step": 8877
+    },
+    {
+      "epoch": 1.51027712883383,
+      "grad_norm": 1.6875,
+      "learning_rate": 9.855530771903776e-06,
+      "loss": 0.8217,
+      "step": 8878
+    },
+    {
+      "epoch": 1.5104485929228197,
+      "grad_norm": 1.6953125,
+      "learning_rate": 9.853725034564928e-06,
+      "loss": 0.8625,
+      "step": 8879
+    },
+    {
+      "epoch": 1.5106200570118096,
+      "grad_norm": 1.75,
+      "learning_rate": 9.851919301996658e-06,
+      "loss": 0.8898,
+      "step": 8880
+    },
+    {
+      "epoch": 1.5107915211007996,
+      "grad_norm": 1.6875,
+      "learning_rate": 9.850113574257857e-06,
+      "loss": 0.7756,
+      "step": 8881
+    },
+    {
+      "epoch": 1.5109629851897894,
+      "grad_norm": 1.7578125,
+      "learning_rate": 9.848307851407414e-06,
+      "loss": 0.8532,
+      "step": 8882
+    },
+    {
+      "epoch": 1.5111344492787793,
+      "grad_norm": 1.6796875,
+      "learning_rate": 9.846502133504227e-06,
+      "loss": 0.8205,
+      "step": 8883
+    },
+    {
+      "epoch": 1.511305913367769,
+      "grad_norm": 1.6796875,
+      "learning_rate": 9.844696420607181e-06,
+      "loss": 0.8913,
+      "step": 8884
+    },
+    {
+      "epoch": 1.511477377456759,
+      "grad_norm": 1.65625,
+      "learning_rate": 9.842890712775172e-06,
+      "loss": 0.8602,
+      "step": 8885
+    },
+    {
+      "epoch": 1.5116488415457487,
+      "grad_norm": 1.5859375,
+      "learning_rate": 9.841085010067089e-06,
+      "loss": 0.7896,
+      "step": 8886
+    },
+    {
+      "epoch": 1.5118203056347386,
+      "grad_norm": 1.65625,
+      "learning_rate": 9.83927931254182e-06,
+      "loss": 0.8216,
+      "step": 8887
+    },
+    {
+      "epoch": 1.5119917697237284,
+      "grad_norm": 1.734375,
+      "learning_rate": 9.837473620258258e-06,
+      "loss": 0.8148,
+      "step": 8888
+    },
+    {
+      "epoch": 1.5121632338127182,
+      "grad_norm": 1.6484375,
+      "learning_rate": 9.835667933275292e-06,
+      "loss": 0.7759,
+      "step": 8889
+    },
+    {
+      "epoch": 1.512334697901708,
+      "grad_norm": 1.828125,
+      "learning_rate": 9.833862251651815e-06,
+      "loss": 0.8444,
+      "step": 8890
+    },
+    {
+      "epoch": 1.512506161990698,
+      "grad_norm": 1.625,
+      "learning_rate": 9.832056575446715e-06,
+      "loss": 0.8576,
+      "step": 8891
+    },
+    {
+      "epoch": 1.512677626079688,
+      "grad_norm": 1.5703125,
+      "learning_rate": 9.830250904718884e-06,
+      "loss": 0.8579,
+      "step": 8892
+    },
+    {
+      "epoch": 1.5128490901686777,
+      "grad_norm": 1.8046875,
+      "learning_rate": 9.82844523952721e-06,
+      "loss": 0.9643,
+      "step": 8893
+    },
+    {
+      "epoch": 1.5130205542576678,
+      "grad_norm": 1.5703125,
+      "learning_rate": 9.82663957993058e-06,
+      "loss": 0.8405,
+      "step": 8894
+    },
+    {
+      "epoch": 1.5131920183466576,
+      "grad_norm": 1.8515625,
+      "learning_rate": 9.824833925987888e-06,
+      "loss": 0.9236,
+      "step": 8895
+    },
+    {
+      "epoch": 1.5133634824356474,
+      "grad_norm": 1.6328125,
+      "learning_rate": 9.823028277758021e-06,
+      "loss": 0.8099,
+      "step": 8896
+    },
+    {
+      "epoch": 1.5135349465246373,
+      "grad_norm": 1.6953125,
+      "learning_rate": 9.821222635299871e-06,
+      "loss": 0.8192,
+      "step": 8897
+    },
+    {
+      "epoch": 1.513706410613627,
+      "grad_norm": 1.65625,
+      "learning_rate": 9.819416998672323e-06,
+      "loss": 0.8839,
+      "step": 8898
+    },
+    {
+      "epoch": 1.513877874702617,
+      "grad_norm": 1.6015625,
+      "learning_rate": 9.817611367934264e-06,
+      "loss": 0.7973,
+      "step": 8899
+    },
+    {
+      "epoch": 1.5140493387916067,
+      "grad_norm": 1.703125,
+      "learning_rate": 9.815805743144587e-06,
+      "loss": 0.7841,
+      "step": 8900
+    },
+    {
+      "epoch": 1.5142208028805966,
+      "grad_norm": 1.8203125,
+      "learning_rate": 9.814000124362178e-06,
+      "loss": 0.8855,
+      "step": 8901
+    },
+    {
+      "epoch": 1.5143922669695864,
+      "grad_norm": 1.671875,
+      "learning_rate": 9.812194511645924e-06,
+      "loss": 0.7829,
+      "step": 8902
+    },
+    {
+      "epoch": 1.5145637310585764,
+      "grad_norm": 1.796875,
+      "learning_rate": 9.810388905054715e-06,
+      "loss": 0.8709,
+      "step": 8903
+    },
+    {
+      "epoch": 1.5147351951475663,
+      "grad_norm": 1.7109375,
+      "learning_rate": 9.808583304647439e-06,
+      "loss": 0.8896,
+      "step": 8904
+    },
+    {
+      "epoch": 1.514906659236556,
+      "grad_norm": 1.671875,
+      "learning_rate": 9.80677771048298e-06,
+      "loss": 0.8687,
+      "step": 8905
+    },
+    {
+      "epoch": 1.5150781233255461,
+      "grad_norm": 1.6484375,
+      "learning_rate": 9.80497212262023e-06,
+      "loss": 0.7913,
+      "step": 8906
+    },
+    {
+      "epoch": 1.515249587414536,
+      "grad_norm": 1.765625,
+      "learning_rate": 9.803166541118072e-06,
+      "loss": 0.81,
+      "step": 8907
+    },
+    {
+      "epoch": 1.5154210515035258,
+      "grad_norm": 1.6796875,
+      "learning_rate": 9.801360966035396e-06,
+      "loss": 0.807,
+      "step": 8908
+    },
+    {
+      "epoch": 1.5155925155925156,
+      "grad_norm": 1.8125,
+      "learning_rate": 9.799555397431086e-06,
+      "loss": 0.8345,
+      "step": 8909
+    },
+    {
+      "epoch": 1.5157639796815054,
+      "grad_norm": 1.71875,
+      "learning_rate": 9.797749835364028e-06,
+      "loss": 0.8034,
+      "step": 8910
+    },
+    {
+      "epoch": 1.5159354437704953,
+      "grad_norm": 1.765625,
+      "learning_rate": 9.79594427989311e-06,
+      "loss": 0.7871,
+      "step": 8911
+    },
+    {
+      "epoch": 1.516106907859485,
+      "grad_norm": 1.6953125,
+      "learning_rate": 9.794138731077219e-06,
+      "loss": 0.9255,
+      "step": 8912
+    },
+    {
+      "epoch": 1.516278371948475,
+      "grad_norm": 1.734375,
+      "learning_rate": 9.792333188975237e-06,
+      "loss": 0.9397,
+      "step": 8913
+    },
+    {
+      "epoch": 1.5164498360374647,
+      "grad_norm": 1.71875,
+      "learning_rate": 9.790527653646053e-06,
+      "loss": 0.8087,
+      "step": 8914
+    },
+    {
+      "epoch": 1.5166213001264548,
+      "grad_norm": 1.703125,
+      "learning_rate": 9.78872212514855e-06,
+      "loss": 0.8213,
+      "step": 8915
+    },
+    {
+      "epoch": 1.5167927642154446,
+      "grad_norm": 1.703125,
+      "learning_rate": 9.786916603541614e-06,
+      "loss": 0.8722,
+      "step": 8916
+    },
+    {
+      "epoch": 1.5169642283044344,
+      "grad_norm": 1.7421875,
+      "learning_rate": 9.78511108888413e-06,
+      "loss": 0.885,
+      "step": 8917
+    },
+    {
+      "epoch": 1.5171356923934245,
+      "grad_norm": 1.6640625,
+      "learning_rate": 9.783305581234983e-06,
+      "loss": 0.8474,
+      "step": 8918
+    },
+    {
+      "epoch": 1.5173071564824143,
+      "grad_norm": 1.6796875,
+      "learning_rate": 9.781500080653057e-06,
+      "loss": 0.7978,
+      "step": 8919
+    },
+    {
+      "epoch": 1.5174786205714041,
+      "grad_norm": 1.6171875,
+      "learning_rate": 9.77969458719724e-06,
+      "loss": 0.7995,
+      "step": 8920
+    },
+    {
+      "epoch": 1.517650084660394,
+      "grad_norm": 1.6015625,
+      "learning_rate": 9.777889100926408e-06,
+      "loss": 0.8285,
+      "step": 8921
+    },
+    {
+      "epoch": 1.5178215487493838,
+      "grad_norm": 1.59375,
+      "learning_rate": 9.77608362189945e-06,
+      "loss": 0.8133,
+      "step": 8922
+    },
+    {
+      "epoch": 1.5179930128383736,
+      "grad_norm": 1.6484375,
+      "learning_rate": 9.774278150175249e-06,
+      "loss": 0.8127,
+      "step": 8923
+    },
+    {
+      "epoch": 1.5181644769273634,
+      "grad_norm": 1.640625,
+      "learning_rate": 9.772472685812682e-06,
+      "loss": 0.8633,
+      "step": 8924
+    },
+    {
+      "epoch": 1.5183359410163533,
+      "grad_norm": 1.6796875,
+      "learning_rate": 9.770667228870643e-06,
+      "loss": 0.9239,
+      "step": 8925
+    },
+    {
+      "epoch": 1.518507405105343,
+      "grad_norm": 1.640625,
+      "learning_rate": 9.768861779408006e-06,
+      "loss": 0.942,
+      "step": 8926
+    },
+    {
+      "epoch": 1.5186788691943331,
+      "grad_norm": 1.671875,
+      "learning_rate": 9.767056337483659e-06,
+      "loss": 0.7934,
+      "step": 8927
+    },
+    {
+      "epoch": 1.518850333283323,
+      "grad_norm": 1.6953125,
+      "learning_rate": 9.765250903156482e-06,
+      "loss": 0.8728,
+      "step": 8928
+    },
+    {
+      "epoch": 1.5190217973723128,
+      "grad_norm": 1.625,
+      "learning_rate": 9.763445476485357e-06,
+      "loss": 0.8055,
+      "step": 8929
+    },
+    {
+      "epoch": 1.5191932614613028,
+      "grad_norm": 1.6875,
+      "learning_rate": 9.761640057529164e-06,
+      "loss": 0.8335,
+      "step": 8930
+    },
+    {
+      "epoch": 1.5193647255502927,
+      "grad_norm": 1.6875,
+      "learning_rate": 9.759834646346794e-06,
+      "loss": 0.9022,
+      "step": 8931
+    },
+    {
+      "epoch": 1.5195361896392825,
+      "grad_norm": 1.6875,
+      "learning_rate": 9.758029242997112e-06,
+      "loss": 0.8732,
+      "step": 8932
+    },
+    {
+      "epoch": 1.5197076537282723,
+      "grad_norm": 1.7890625,
+      "learning_rate": 9.75622384753901e-06,
+      "loss": 0.8624,
+      "step": 8933
+    },
+    {
+      "epoch": 1.5198791178172621,
+      "grad_norm": 1.734375,
+      "learning_rate": 9.754418460031365e-06,
+      "loss": 0.8726,
+      "step": 8934
+    },
+    {
+      "epoch": 1.520050581906252,
+      "grad_norm": 1.71875,
+      "learning_rate": 9.75261308053306e-06,
+      "loss": 0.8423,
+      "step": 8935
+    },
+    {
+      "epoch": 1.5202220459952418,
+      "grad_norm": 1.65625,
+      "learning_rate": 9.750807709102974e-06,
+      "loss": 0.8287,
+      "step": 8936
+    },
+    {
+      "epoch": 1.5203935100842316,
+      "grad_norm": 1.703125,
+      "learning_rate": 9.749002345799988e-06,
+      "loss": 0.8095,
+      "step": 8937
+    },
+    {
+      "epoch": 1.5205649741732215,
+      "grad_norm": 1.6796875,
+      "learning_rate": 9.747196990682979e-06,
+      "loss": 0.8468,
+      "step": 8938
+    },
+    {
+      "epoch": 1.5207364382622115,
+      "grad_norm": 1.71875,
+      "learning_rate": 9.74539164381083e-06,
+      "loss": 0.88,
+      "step": 8939
+    },
+    {
+      "epoch": 1.5209079023512013,
+      "grad_norm": 1.671875,
+      "learning_rate": 9.743586305242417e-06,
+      "loss": 0.8588,
+      "step": 8940
+    },
+    {
+      "epoch": 1.5210793664401911,
+      "grad_norm": 1.78125,
+      "learning_rate": 9.74178097503662e-06,
+      "loss": 0.8929,
+      "step": 8941
+    },
+    {
+      "epoch": 1.5212508305291812,
+      "grad_norm": 1.6640625,
+      "learning_rate": 9.739975653252317e-06,
+      "loss": 0.9824,
+      "step": 8942
+    },
+    {
+      "epoch": 1.521422294618171,
+      "grad_norm": 1.65625,
+      "learning_rate": 9.73817033994839e-06,
+      "loss": 0.8122,
+      "step": 8943
+    },
+    {
+      "epoch": 1.5215937587071608,
+      "grad_norm": 1.703125,
+      "learning_rate": 9.736365035183712e-06,
+      "loss": 0.824,
+      "step": 8944
+    },
+    {
+      "epoch": 1.5217652227961507,
+      "grad_norm": 1.703125,
+      "learning_rate": 9.734559739017162e-06,
+      "loss": 0.8004,
+      "step": 8945
+    },
+    {
+      "epoch": 1.5219366868851405,
+      "grad_norm": 1.671875,
+      "learning_rate": 9.732754451507621e-06,
+      "loss": 0.7771,
+      "step": 8946
+    },
+    {
+      "epoch": 1.5221081509741303,
+      "grad_norm": 1.7421875,
+      "learning_rate": 9.730949172713961e-06,
+      "loss": 0.9023,
+      "step": 8947
+    },
+    {
+      "epoch": 1.5222796150631202,
+      "grad_norm": 1.703125,
+      "learning_rate": 9.729143902695064e-06,
+      "loss": 0.8699,
+      "step": 8948
+    },
+    {
+      "epoch": 1.52245107915211,
+      "grad_norm": 1.7265625,
+      "learning_rate": 9.727338641509804e-06,
+      "loss": 0.9692,
+      "step": 8949
+    },
+    {
+      "epoch": 1.5226225432410998,
+      "grad_norm": 2.0,
+      "learning_rate": 9.725533389217056e-06,
+      "loss": 0.8999,
+      "step": 8950
+    },
+    {
+      "epoch": 1.5227940073300898,
+      "grad_norm": 1.7109375,
+      "learning_rate": 9.723728145875699e-06,
+      "loss": 0.8602,
+      "step": 8951
+    },
+    {
+      "epoch": 1.5229654714190797,
+      "grad_norm": 1.8671875,
+      "learning_rate": 9.721922911544607e-06,
+      "loss": 0.888,
+      "step": 8952
+    },
+    {
+      "epoch": 1.5231369355080695,
+      "grad_norm": 1.7421875,
+      "learning_rate": 9.720117686282657e-06,
+      "loss": 0.8557,
+      "step": 8953
+    },
+    {
+      "epoch": 1.5233083995970595,
+      "grad_norm": 1.6015625,
+      "learning_rate": 9.718312470148723e-06,
+      "loss": 0.8532,
+      "step": 8954
+    },
+    {
+      "epoch": 1.5234798636860494,
+      "grad_norm": 1.7578125,
+      "learning_rate": 9.71650726320168e-06,
+      "loss": 0.9081,
+      "step": 8955
+    },
+    {
+      "epoch": 1.5236513277750392,
+      "grad_norm": 1.7109375,
+      "learning_rate": 9.714702065500401e-06,
+      "loss": 0.8824,
+      "step": 8956
+    },
+    {
+      "epoch": 1.523822791864029,
+      "grad_norm": 1.703125,
+      "learning_rate": 9.712896877103764e-06,
+      "loss": 0.878,
+      "step": 8957
+    },
+    {
+      "epoch": 1.5239942559530189,
+      "grad_norm": 1.8125,
+      "learning_rate": 9.71109169807064e-06,
+      "loss": 0.9402,
+      "step": 8958
+    },
+    {
+      "epoch": 1.5241657200420087,
+      "grad_norm": 1.6328125,
+      "learning_rate": 9.709286528459904e-06,
+      "loss": 0.7948,
+      "step": 8959
+    },
+    {
+      "epoch": 1.5243371841309985,
+      "grad_norm": 1.59375,
+      "learning_rate": 9.707481368330428e-06,
+      "loss": 0.6961,
+      "step": 8960
+    },
+    {
+      "epoch": 1.5245086482199883,
+      "grad_norm": 1.6796875,
+      "learning_rate": 9.705676217741087e-06,
+      "loss": 0.8052,
+      "step": 8961
+    },
+    {
+      "epoch": 1.5246801123089782,
+      "grad_norm": 1.6796875,
+      "learning_rate": 9.703871076750752e-06,
+      "loss": 0.8027,
+      "step": 8962
+    },
+    {
+      "epoch": 1.5248515763979682,
+      "grad_norm": 1.7421875,
+      "learning_rate": 9.702065945418295e-06,
+      "loss": 0.9244,
+      "step": 8963
+    },
+    {
+      "epoch": 1.525023040486958,
+      "grad_norm": 1.7265625,
+      "learning_rate": 9.700260823802592e-06,
+      "loss": 0.8807,
+      "step": 8964
+    },
+    {
+      "epoch": 1.5251945045759479,
+      "grad_norm": 1.671875,
+      "learning_rate": 9.698455711962511e-06,
+      "loss": 0.9256,
+      "step": 8965
+    },
+    {
+      "epoch": 1.525365968664938,
+      "grad_norm": 1.796875,
+      "learning_rate": 9.696650609956931e-06,
+      "loss": 0.8056,
+      "step": 8966
+    },
+    {
+      "epoch": 1.5255374327539277,
+      "grad_norm": 1.71875,
+      "learning_rate": 9.69484551784471e-06,
+      "loss": 0.807,
+      "step": 8967
+    },
+    {
+      "epoch": 1.5257088968429176,
+      "grad_norm": 1.6875,
+      "learning_rate": 9.693040435684727e-06,
+      "loss": 0.9571,
+      "step": 8968
+    },
+    {
+      "epoch": 1.5258803609319074,
+      "grad_norm": 1.671875,
+      "learning_rate": 9.691235363535852e-06,
+      "loss": 0.8819,
+      "step": 8969
+    },
+    {
+      "epoch": 1.5260518250208972,
+      "grad_norm": 1.703125,
+      "learning_rate": 9.689430301456954e-06,
+      "loss": 0.828,
+      "step": 8970
+    },
+    {
+      "epoch": 1.526223289109887,
+      "grad_norm": 1.5859375,
+      "learning_rate": 9.6876252495069e-06,
+      "loss": 0.8439,
+      "step": 8971
+    },
+    {
+      "epoch": 1.5263947531988769,
+      "grad_norm": 1.71875,
+      "learning_rate": 9.685820207744565e-06,
+      "loss": 0.8378,
+      "step": 8972
+    },
+    {
+      "epoch": 1.5265662172878667,
+      "grad_norm": 1.6640625,
+      "learning_rate": 9.68401517622882e-06,
+      "loss": 0.9048,
+      "step": 8973
+    },
+    {
+      "epoch": 1.5267376813768565,
+      "grad_norm": 1.734375,
+      "learning_rate": 9.682210155018526e-06,
+      "loss": 0.8459,
+      "step": 8974
+    },
+    {
+      "epoch": 1.5269091454658466,
+      "grad_norm": 1.71875,
+      "learning_rate": 9.680405144172556e-06,
+      "loss": 0.79,
+      "step": 8975
+    },
+    {
+      "epoch": 1.5270806095548364,
+      "grad_norm": 1.6640625,
+      "learning_rate": 9.67860014374978e-06,
+      "loss": 0.8598,
+      "step": 8976
+    },
+    {
+      "epoch": 1.5272520736438262,
+      "grad_norm": 1.6640625,
+      "learning_rate": 9.676795153809065e-06,
+      "loss": 0.783,
+      "step": 8977
+    },
+    {
+      "epoch": 1.527423537732816,
+      "grad_norm": 1.71875,
+      "learning_rate": 9.674990174409275e-06,
+      "loss": 0.8452,
+      "step": 8978
+    },
+    {
+      "epoch": 1.527595001821806,
+      "grad_norm": 1.8359375,
+      "learning_rate": 9.673185205609277e-06,
+      "loss": 0.8368,
+      "step": 8979
+    },
+    {
+      "epoch": 1.527766465910796,
+      "grad_norm": 1.75,
+      "learning_rate": 9.67138024746794e-06,
+      "loss": 0.8541,
+      "step": 8980
+    },
+    {
+      "epoch": 1.5279379299997857,
+      "grad_norm": 1.6875,
+      "learning_rate": 9.669575300044133e-06,
+      "loss": 0.9475,
+      "step": 8981
+    },
+    {
+      "epoch": 1.5281093940887756,
+      "grad_norm": 1.7421875,
+      "learning_rate": 9.667770363396718e-06,
+      "loss": 0.8133,
+      "step": 8982
+    },
+    {
+      "epoch": 1.5282808581777654,
+      "grad_norm": 1.78125,
+      "learning_rate": 9.665965437584564e-06,
+      "loss": 0.8793,
+      "step": 8983
+    },
+    {
+      "epoch": 1.5284523222667552,
+      "grad_norm": 1.6796875,
+      "learning_rate": 9.664160522666534e-06,
+      "loss": 0.885,
+      "step": 8984
+    },
+    {
+      "epoch": 1.528623786355745,
+      "grad_norm": 1.6953125,
+      "learning_rate": 9.662355618701495e-06,
+      "loss": 0.8527,
+      "step": 8985
+    },
+    {
+      "epoch": 1.5287952504447349,
+      "grad_norm": 1.65625,
+      "learning_rate": 9.66055072574831e-06,
+      "loss": 0.8093,
+      "step": 8986
+    },
+    {
+      "epoch": 1.5289667145337247,
+      "grad_norm": 1.671875,
+      "learning_rate": 9.658745843865845e-06,
+      "loss": 0.8133,
+      "step": 8987
+    },
+    {
+      "epoch": 1.5291381786227147,
+      "grad_norm": 1.71875,
+      "learning_rate": 9.656940973112961e-06,
+      "loss": 0.8026,
+      "step": 8988
+    },
+    {
+      "epoch": 1.5293096427117046,
+      "grad_norm": 1.6875,
+      "learning_rate": 9.655136113548528e-06,
+      "loss": 0.7791,
+      "step": 8989
+    },
+    {
+      "epoch": 1.5294811068006944,
+      "grad_norm": 1.7734375,
+      "learning_rate": 9.6533312652314e-06,
+      "loss": 0.9006,
+      "step": 8990
+    },
+    {
+      "epoch": 1.5296525708896844,
+      "grad_norm": 1.6796875,
+      "learning_rate": 9.65152642822045e-06,
+      "loss": 0.9335,
+      "step": 8991
+    },
+    {
+      "epoch": 1.5298240349786743,
+      "grad_norm": 1.765625,
+      "learning_rate": 9.649721602574532e-06,
+      "loss": 0.8893,
+      "step": 8992
+    },
+    {
+      "epoch": 1.529995499067664,
+      "grad_norm": 1.6796875,
+      "learning_rate": 9.64791678835251e-06,
+      "loss": 0.8333,
+      "step": 8993
+    },
+    {
+      "epoch": 1.530166963156654,
+      "grad_norm": 1.6875,
+      "learning_rate": 9.64611198561325e-06,
+      "loss": 0.9223,
+      "step": 8994
+    },
+    {
+      "epoch": 1.5303384272456437,
+      "grad_norm": 1.640625,
+      "learning_rate": 9.64430719441561e-06,
+      "loss": 0.809,
+      "step": 8995
+    },
+    {
+      "epoch": 1.5305098913346336,
+      "grad_norm": 1.578125,
+      "learning_rate": 9.642502414818452e-06,
+      "loss": 0.805,
+      "step": 8996
+    },
+    {
+      "epoch": 1.5306813554236234,
+      "grad_norm": 1.84375,
+      "learning_rate": 9.640697646880636e-06,
+      "loss": 0.824,
+      "step": 8997
+    },
+    {
+      "epoch": 1.5308528195126132,
+      "grad_norm": 1.734375,
+      "learning_rate": 9.638892890661022e-06,
+      "loss": 0.8221,
+      "step": 8998
+    },
+    {
+      "epoch": 1.531024283601603,
+      "grad_norm": 1.59375,
+      "learning_rate": 9.637088146218471e-06,
+      "loss": 0.7839,
+      "step": 8999
+    },
+    {
+      "epoch": 1.531195747690593,
+      "grad_norm": 1.6484375,
+      "learning_rate": 9.635283413611846e-06,
+      "loss": 0.8316,
+      "step": 9000
+    },
+    {
+      "epoch": 1.531367211779583,
+      "grad_norm": 1.734375,
+      "learning_rate": 9.633478692899998e-06,
+      "loss": 0.8849,
+      "step": 9001
+    },
+    {
+      "epoch": 1.5315386758685727,
+      "grad_norm": 1.7421875,
+      "learning_rate": 9.631673984141792e-06,
+      "loss": 0.9323,
+      "step": 9002
+    },
+    {
+      "epoch": 1.5317101399575628,
+      "grad_norm": 1.78125,
+      "learning_rate": 9.629869287396083e-06,
+      "loss": 0.8271,
+      "step": 9003
+    },
+    {
+      "epoch": 1.5318816040465526,
+      "grad_norm": 1.7734375,
+      "learning_rate": 9.62806460272173e-06,
+      "loss": 0.9224,
+      "step": 9004
+    },
+    {
+      "epoch": 1.5320530681355424,
+      "grad_norm": 1.609375,
+      "learning_rate": 9.626259930177592e-06,
+      "loss": 0.8317,
+      "step": 9005
+    },
+    {
+      "epoch": 1.5322245322245323,
+      "grad_norm": 1.7421875,
+      "learning_rate": 9.624455269822526e-06,
+      "loss": 0.8575,
+      "step": 9006
+    },
+    {
+      "epoch": 1.532395996313522,
+      "grad_norm": 1.7578125,
+      "learning_rate": 9.622650621715385e-06,
+      "loss": 0.8243,
+      "step": 9007
+    },
+    {
+      "epoch": 1.532567460402512,
+      "grad_norm": 1.734375,
+      "learning_rate": 9.62084598591503e-06,
+      "loss": 0.8792,
+      "step": 9008
+    },
+    {
+      "epoch": 1.5327389244915017,
+      "grad_norm": 1.71875,
+      "learning_rate": 9.619041362480314e-06,
+      "loss": 0.9154,
+      "step": 9009
+    },
+    {
+      "epoch": 1.5329103885804916,
+      "grad_norm": 1.9140625,
+      "learning_rate": 9.617236751470094e-06,
+      "loss": 0.9161,
+      "step": 9010
+    },
+    {
+      "epoch": 1.5330818526694814,
+      "grad_norm": 1.6328125,
+      "learning_rate": 9.615432152943225e-06,
+      "loss": 0.8368,
+      "step": 9011
+    },
+    {
+      "epoch": 1.5332533167584714,
+      "grad_norm": 1.65625,
+      "learning_rate": 9.613627566958567e-06,
+      "loss": 0.828,
+      "step": 9012
+    },
+    {
+      "epoch": 1.5334247808474613,
+      "grad_norm": 1.703125,
+      "learning_rate": 9.611822993574963e-06,
+      "loss": 0.8458,
+      "step": 9013
+    },
+    {
+      "epoch": 1.533596244936451,
+      "grad_norm": 1.6171875,
+      "learning_rate": 9.610018432851275e-06,
+      "loss": 0.7959,
+      "step": 9014
+    },
+    {
+      "epoch": 1.5337677090254411,
+      "grad_norm": 1.734375,
+      "learning_rate": 9.608213884846353e-06,
+      "loss": 0.9074,
+      "step": 9015
+    },
+    {
+      "epoch": 1.533939173114431,
+      "grad_norm": 1.65625,
+      "learning_rate": 9.60640934961905e-06,
+      "loss": 0.7942,
+      "step": 9016
+    },
+    {
+      "epoch": 1.5341106372034208,
+      "grad_norm": 1.71875,
+      "learning_rate": 9.604604827228225e-06,
+      "loss": 0.9097,
+      "step": 9017
+    },
+    {
+      "epoch": 1.5342821012924106,
+      "grad_norm": 1.6640625,
+      "learning_rate": 9.602800317732718e-06,
+      "loss": 0.886,
+      "step": 9018
+    },
+    {
+      "epoch": 1.5344535653814004,
+      "grad_norm": 1.625,
+      "learning_rate": 9.600995821191395e-06,
+      "loss": 0.7985,
+      "step": 9019
+    },
+    {
+      "epoch": 1.5346250294703903,
+      "grad_norm": 1.7578125,
+      "learning_rate": 9.599191337663098e-06,
+      "loss": 0.82,
+      "step": 9020
+    },
+    {
+      "epoch": 1.53479649355938,
+      "grad_norm": 1.7421875,
+      "learning_rate": 9.597386867206683e-06,
+      "loss": 0.9522,
+      "step": 9021
+    },
+    {
+      "epoch": 1.53496795764837,
+      "grad_norm": 1.8515625,
+      "learning_rate": 9.595582409880996e-06,
+      "loss": 0.9203,
+      "step": 9022
+    },
+    {
+      "epoch": 1.5351394217373597,
+      "grad_norm": 1.7421875,
+      "learning_rate": 9.593777965744894e-06,
+      "loss": 0.8255,
+      "step": 9023
+    },
+    {
+      "epoch": 1.5353108858263498,
+      "grad_norm": 1.78125,
+      "learning_rate": 9.59197353485722e-06,
+      "loss": 0.8641,
+      "step": 9024
+    },
+    {
+      "epoch": 1.5354823499153396,
+      "grad_norm": 1.765625,
+      "learning_rate": 9.590169117276825e-06,
+      "loss": 0.8695,
+      "step": 9025
+    },
+    {
+      "epoch": 1.5356538140043294,
+      "grad_norm": 1.7421875,
+      "learning_rate": 9.588364713062555e-06,
+      "loss": 0.8724,
+      "step": 9026
+    },
+    {
+      "epoch": 1.5358252780933195,
+      "grad_norm": 1.828125,
+      "learning_rate": 9.586560322273264e-06,
+      "loss": 0.8957,
+      "step": 9027
+    },
+    {
+      "epoch": 1.5359967421823093,
+      "grad_norm": 1.640625,
+      "learning_rate": 9.584755944967798e-06,
+      "loss": 0.819,
+      "step": 9028
+    },
+    {
+      "epoch": 1.5361682062712991,
+      "grad_norm": 1.703125,
+      "learning_rate": 9.582951581205005e-06,
+      "loss": 0.8256,
+      "step": 9029
+    },
+    {
+      "epoch": 1.536339670360289,
+      "grad_norm": 1.8515625,
+      "learning_rate": 9.581147231043731e-06,
+      "loss": 0.8678,
+      "step": 9030
+    },
+    {
+      "epoch": 1.5365111344492788,
+      "grad_norm": 1.6953125,
+      "learning_rate": 9.579342894542822e-06,
+      "loss": 0.9207,
+      "step": 9031
+    },
+    {
+      "epoch": 1.5366825985382686,
+      "grad_norm": 1.7890625,
+      "learning_rate": 9.577538571761126e-06,
+      "loss": 0.7908,
+      "step": 9032
+    },
+    {
+      "epoch": 1.5368540626272584,
+      "grad_norm": 1.640625,
+      "learning_rate": 9.575734262757488e-06,
+      "loss": 0.8578,
+      "step": 9033
+    },
+    {
+      "epoch": 1.5370255267162483,
+      "grad_norm": 1.671875,
+      "learning_rate": 9.573929967590755e-06,
+      "loss": 0.8436,
+      "step": 9034
+    },
+    {
+      "epoch": 1.537196990805238,
+      "grad_norm": 1.734375,
+      "learning_rate": 9.57212568631977e-06,
+      "loss": 0.8757,
+      "step": 9035
+    },
+    {
+      "epoch": 1.5373684548942281,
+      "grad_norm": 1.671875,
+      "learning_rate": 9.570321419003375e-06,
+      "loss": 0.8262,
+      "step": 9036
+    },
+    {
+      "epoch": 1.537539918983218,
+      "grad_norm": 1.7265625,
+      "learning_rate": 9.568517165700419e-06,
+      "loss": 0.8632,
+      "step": 9037
+    },
+    {
+      "epoch": 1.5377113830722078,
+      "grad_norm": 1.65625,
+      "learning_rate": 9.56671292646974e-06,
+      "loss": 0.8484,
+      "step": 9038
+    },
+    {
+      "epoch": 1.5378828471611978,
+      "grad_norm": 1.765625,
+      "learning_rate": 9.564908701370186e-06,
+      "loss": 0.9305,
+      "step": 9039
+    },
+    {
+      "epoch": 1.5380543112501877,
+      "grad_norm": 1.640625,
+      "learning_rate": 9.563104490460599e-06,
+      "loss": 0.8929,
+      "step": 9040
+    },
+    {
+      "epoch": 1.5382257753391775,
+      "grad_norm": 1.6484375,
+      "learning_rate": 9.561300293799818e-06,
+      "loss": 0.8637,
+      "step": 9041
+    },
+    {
+      "epoch": 1.5383972394281673,
+      "grad_norm": 1.671875,
+      "learning_rate": 9.559496111446686e-06,
+      "loss": 0.7468,
+      "step": 9042
+    },
+    {
+      "epoch": 1.5385687035171571,
+      "grad_norm": 1.7265625,
+      "learning_rate": 9.557691943460046e-06,
+      "loss": 0.8259,
+      "step": 9043
+    },
+    {
+      "epoch": 1.538740167606147,
+      "grad_norm": 1.640625,
+      "learning_rate": 9.555887789898737e-06,
+      "loss": 0.8728,
+      "step": 9044
+    },
+    {
+      "epoch": 1.5389116316951368,
+      "grad_norm": 1.8125,
+      "learning_rate": 9.554083650821598e-06,
+      "loss": 0.858,
+      "step": 9045
+    },
+    {
+      "epoch": 1.5390830957841266,
+      "grad_norm": 1.7421875,
+      "learning_rate": 9.552279526287472e-06,
+      "loss": 0.8959,
+      "step": 9046
+    },
+    {
+      "epoch": 1.5392545598731164,
+      "grad_norm": 1.6796875,
+      "learning_rate": 9.550475416355199e-06,
+      "loss": 0.8063,
+      "step": 9047
+    },
+    {
+      "epoch": 1.5394260239621065,
+      "grad_norm": 1.65625,
+      "learning_rate": 9.548671321083612e-06,
+      "loss": 0.7583,
+      "step": 9048
+    },
+    {
+      "epoch": 1.5395974880510963,
+      "grad_norm": 1.65625,
+      "learning_rate": 9.546867240531553e-06,
+      "loss": 0.91,
+      "step": 9049
+    },
+    {
+      "epoch": 1.5397689521400861,
+      "grad_norm": 1.6953125,
+      "learning_rate": 9.54506317475786e-06,
+      "loss": 0.8993,
+      "step": 9050
+    },
+    {
+      "epoch": 1.5399404162290762,
+      "grad_norm": 1.828125,
+      "learning_rate": 9.54325912382137e-06,
+      "loss": 0.8975,
+      "step": 9051
+    },
+    {
+      "epoch": 1.540111880318066,
+      "grad_norm": 1.7109375,
+      "learning_rate": 9.54145508778092e-06,
+      "loss": 0.881,
+      "step": 9052
+    },
+    {
+      "epoch": 1.5402833444070558,
+      "grad_norm": 1.609375,
+      "learning_rate": 9.539651066695346e-06,
+      "loss": 0.8008,
+      "step": 9053
+    },
+    {
+      "epoch": 1.5404548084960457,
+      "grad_norm": 1.6796875,
+      "learning_rate": 9.537847060623484e-06,
+      "loss": 0.845,
+      "step": 9054
+    },
+    {
+      "epoch": 1.5406262725850355,
+      "grad_norm": 1.7109375,
+      "learning_rate": 9.53604306962417e-06,
+      "loss": 0.8601,
+      "step": 9055
+    },
+    {
+      "epoch": 1.5407977366740253,
+      "grad_norm": 1.7421875,
+      "learning_rate": 9.53423909375624e-06,
+      "loss": 0.8381,
+      "step": 9056
+    },
+    {
+      "epoch": 1.5409692007630151,
+      "grad_norm": 1.75,
+      "learning_rate": 9.532435133078523e-06,
+      "loss": 0.9294,
+      "step": 9057
+    },
+    {
+      "epoch": 1.541140664852005,
+      "grad_norm": 1.6796875,
+      "learning_rate": 9.530631187649864e-06,
+      "loss": 0.7982,
+      "step": 9058
+    },
+    {
+      "epoch": 1.5413121289409948,
+      "grad_norm": 1.7265625,
+      "learning_rate": 9.528827257529085e-06,
+      "loss": 0.8474,
+      "step": 9059
+    },
+    {
+      "epoch": 1.5414835930299848,
+      "grad_norm": 1.703125,
+      "learning_rate": 9.527023342775023e-06,
+      "loss": 0.8985,
+      "step": 9060
+    },
+    {
+      "epoch": 1.5416550571189747,
+      "grad_norm": 1.7109375,
+      "learning_rate": 9.525219443446512e-06,
+      "loss": 0.909,
+      "step": 9061
+    },
+    {
+      "epoch": 1.5418265212079645,
+      "grad_norm": 1.625,
+      "learning_rate": 9.523415559602383e-06,
+      "loss": 0.8313,
+      "step": 9062
+    },
+    {
+      "epoch": 1.5419979852969545,
+      "grad_norm": 1.6328125,
+      "learning_rate": 9.521611691301465e-06,
+      "loss": 0.8706,
+      "step": 9063
+    },
+    {
+      "epoch": 1.5421694493859444,
+      "grad_norm": 1.7109375,
+      "learning_rate": 9.519807838602593e-06,
+      "loss": 0.8734,
+      "step": 9064
+    },
+    {
+      "epoch": 1.5423409134749342,
+      "grad_norm": 1.7265625,
+      "learning_rate": 9.518004001564592e-06,
+      "loss": 0.9018,
+      "step": 9065
+    },
+    {
+      "epoch": 1.542512377563924,
+      "grad_norm": 1.7421875,
+      "learning_rate": 9.5162001802463e-06,
+      "loss": 0.8821,
+      "step": 9066
+    },
+    {
+      "epoch": 1.5426838416529138,
+      "grad_norm": 1.65625,
+      "learning_rate": 9.514396374706541e-06,
+      "loss": 0.8449,
+      "step": 9067
+    },
+    {
+      "epoch": 1.5428553057419037,
+      "grad_norm": 1.6875,
+      "learning_rate": 9.512592585004144e-06,
+      "loss": 0.9382,
+      "step": 9068
+    },
+    {
+      "epoch": 1.5430267698308935,
+      "grad_norm": 1.609375,
+      "learning_rate": 9.510788811197939e-06,
+      "loss": 0.7869,
+      "step": 9069
+    },
+    {
+      "epoch": 1.5431982339198833,
+      "grad_norm": 1.6328125,
+      "learning_rate": 9.508985053346757e-06,
+      "loss": 0.8338,
+      "step": 9070
+    },
+    {
+      "epoch": 1.5433696980088731,
+      "grad_norm": 1.6953125,
+      "learning_rate": 9.507181311509416e-06,
+      "loss": 0.7595,
+      "step": 9071
+    },
+    {
+      "epoch": 1.5435411620978632,
+      "grad_norm": 1.609375,
+      "learning_rate": 9.505377585744751e-06,
+      "loss": 0.9012,
+      "step": 9072
+    },
+    {
+      "epoch": 1.543712626186853,
+      "grad_norm": 1.6484375,
+      "learning_rate": 9.503573876111581e-06,
+      "loss": 0.8346,
+      "step": 9073
+    },
+    {
+      "epoch": 1.5438840902758428,
+      "grad_norm": 1.6796875,
+      "learning_rate": 9.501770182668739e-06,
+      "loss": 0.8816,
+      "step": 9074
+    },
+    {
+      "epoch": 1.5440555543648327,
+      "grad_norm": 1.546875,
+      "learning_rate": 9.499966505475049e-06,
+      "loss": 0.7961,
+      "step": 9075
+    },
+    {
+      "epoch": 1.5442270184538227,
+      "grad_norm": 1.78125,
+      "learning_rate": 9.498162844589333e-06,
+      "loss": 0.8428,
+      "step": 9076
+    },
+    {
+      "epoch": 1.5443984825428125,
+      "grad_norm": 1.7421875,
+      "learning_rate": 9.496359200070416e-06,
+      "loss": 0.9426,
+      "step": 9077
+    },
+    {
+      "epoch": 1.5445699466318024,
+      "grad_norm": 1.6953125,
+      "learning_rate": 9.494555571977122e-06,
+      "loss": 0.823,
+      "step": 9078
+    },
+    {
+      "epoch": 1.5447414107207922,
+      "grad_norm": 1.703125,
+      "learning_rate": 9.492751960368274e-06,
+      "loss": 0.8447,
+      "step": 9079
+    },
+    {
+      "epoch": 1.544912874809782,
+      "grad_norm": 1.7265625,
+      "learning_rate": 9.490948365302695e-06,
+      "loss": 0.9213,
+      "step": 9080
+    },
+    {
+      "epoch": 1.5450843388987718,
+      "grad_norm": 1.6015625,
+      "learning_rate": 9.489144786839209e-06,
+      "loss": 0.7861,
+      "step": 9081
+    },
+    {
+      "epoch": 1.5452558029877617,
+      "grad_norm": 1.7109375,
+      "learning_rate": 9.487341225036634e-06,
+      "loss": 0.8669,
+      "step": 9082
+    },
+    {
+      "epoch": 1.5454272670767515,
+      "grad_norm": 1.6875,
+      "learning_rate": 9.485537679953791e-06,
+      "loss": 0.8379,
+      "step": 9083
+    },
+    {
+      "epoch": 1.5455987311657413,
+      "grad_norm": 1.71875,
+      "learning_rate": 9.483734151649503e-06,
+      "loss": 0.9128,
+      "step": 9084
+    },
+    {
+      "epoch": 1.5457701952547314,
+      "grad_norm": 1.7265625,
+      "learning_rate": 9.481930640182586e-06,
+      "loss": 0.8098,
+      "step": 9085
+    },
+    {
+      "epoch": 1.5459416593437212,
+      "grad_norm": 1.6015625,
+      "learning_rate": 9.480127145611863e-06,
+      "loss": 0.8269,
+      "step": 9086
+    },
+    {
+      "epoch": 1.546113123432711,
+      "grad_norm": 1.75,
+      "learning_rate": 9.47832366799615e-06,
+      "loss": 0.9293,
+      "step": 9087
+    },
+    {
+      "epoch": 1.546284587521701,
+      "grad_norm": 1.734375,
+      "learning_rate": 9.476520207394268e-06,
+      "loss": 0.8146,
+      "step": 9088
+    },
+    {
+      "epoch": 1.5464560516106909,
+      "grad_norm": 1.6953125,
+      "learning_rate": 9.474716763865033e-06,
+      "loss": 0.8419,
+      "step": 9089
+    },
+    {
+      "epoch": 1.5466275156996807,
+      "grad_norm": 1.71875,
+      "learning_rate": 9.472913337467263e-06,
+      "loss": 0.7645,
+      "step": 9090
+    },
+    {
+      "epoch": 1.5467989797886705,
+      "grad_norm": 1.7578125,
+      "learning_rate": 9.471109928259774e-06,
+      "loss": 0.9052,
+      "step": 9091
+    },
+    {
+      "epoch": 1.5469704438776604,
+      "grad_norm": 1.7578125,
+      "learning_rate": 9.469306536301381e-06,
+      "loss": 0.9394,
+      "step": 9092
+    },
+    {
+      "epoch": 1.5471419079666502,
+      "grad_norm": 1.6953125,
+      "learning_rate": 9.467503161650902e-06,
+      "loss": 0.8726,
+      "step": 9093
+    },
+    {
+      "epoch": 1.54731337205564,
+      "grad_norm": 1.703125,
+      "learning_rate": 9.46569980436715e-06,
+      "loss": 0.833,
+      "step": 9094
+    },
+    {
+      "epoch": 1.5474848361446298,
+      "grad_norm": 1.59375,
+      "learning_rate": 9.463896464508938e-06,
+      "loss": 0.7878,
+      "step": 9095
+    },
+    {
+      "epoch": 1.5476563002336197,
+      "grad_norm": 1.625,
+      "learning_rate": 9.462093142135081e-06,
+      "loss": 0.7952,
+      "step": 9096
+    },
+    {
+      "epoch": 1.5478277643226097,
+      "grad_norm": 1.65625,
+      "learning_rate": 9.460289837304392e-06,
+      "loss": 0.8398,
+      "step": 9097
+    },
+    {
+      "epoch": 1.5479992284115995,
+      "grad_norm": 1.5859375,
+      "learning_rate": 9.458486550075684e-06,
+      "loss": 0.725,
+      "step": 9098
+    },
+    {
+      "epoch": 1.5481706925005894,
+      "grad_norm": 1.59375,
+      "learning_rate": 9.456683280507768e-06,
+      "loss": 0.8511,
+      "step": 9099
+    },
+    {
+      "epoch": 1.5483421565895794,
+      "grad_norm": 1.5859375,
+      "learning_rate": 9.454880028659458e-06,
+      "loss": 0.736,
+      "step": 9100
+    },
+    {
+      "epoch": 1.5483421565895794,
+      "eval_loss": 0.8392665386199951,
+      "eval_runtime": 835.9126,
+      "eval_samples_per_second": 2.99,
+      "eval_steps_per_second": 2.99,
+      "step": 9100
+    },
+    {
+      "epoch": 1.5485136206785692,
+      "grad_norm": 1.71875,
+      "learning_rate": 9.453076794589562e-06,
+      "loss": 0.8413,
+      "step": 9101
+    },
+    {
+      "epoch": 1.548685084767559,
+      "grad_norm": 1.734375,
+      "learning_rate": 9.451273578356892e-06,
+      "loss": 0.8524,
+      "step": 9102
+    },
+    {
+      "epoch": 1.5488565488565489,
+      "grad_norm": 1.6171875,
+      "learning_rate": 9.449470380020256e-06,
+      "loss": 0.7827,
+      "step": 9103
+    },
+    {
+      "epoch": 1.5490280129455387,
+      "grad_norm": 1.875,
+      "learning_rate": 9.447667199638467e-06,
+      "loss": 0.9264,
+      "step": 9104
+    },
+    {
+      "epoch": 1.5491994770345285,
+      "grad_norm": 1.71875,
+      "learning_rate": 9.445864037270328e-06,
+      "loss": 0.971,
+      "step": 9105
+    },
+    {
+      "epoch": 1.5493709411235184,
+      "grad_norm": 1.734375,
+      "learning_rate": 9.444060892974648e-06,
+      "loss": 0.864,
+      "step": 9106
+    },
+    {
+      "epoch": 1.5495424052125082,
+      "grad_norm": 1.8203125,
+      "learning_rate": 9.442257766810237e-06,
+      "loss": 0.8644,
+      "step": 9107
+    },
+    {
+      "epoch": 1.549713869301498,
+      "grad_norm": 1.59375,
+      "learning_rate": 9.4404546588359e-06,
+      "loss": 0.8537,
+      "step": 9108
+    },
+    {
+      "epoch": 1.549885333390488,
+      "grad_norm": 1.7578125,
+      "learning_rate": 9.438651569110444e-06,
+      "loss": 0.8993,
+      "step": 9109
+    },
+    {
+      "epoch": 1.5500567974794779,
+      "grad_norm": 1.6796875,
+      "learning_rate": 9.436848497692674e-06,
+      "loss": 0.8384,
+      "step": 9110
+    },
+    {
+      "epoch": 1.5502282615684677,
+      "grad_norm": 1.640625,
+      "learning_rate": 9.435045444641396e-06,
+      "loss": 0.8226,
+      "step": 9111
+    },
+    {
+      "epoch": 1.5503997256574578,
+      "grad_norm": 1.75,
+      "learning_rate": 9.43324241001541e-06,
+      "loss": 0.8532,
+      "step": 9112
+    },
+    {
+      "epoch": 1.5505711897464476,
+      "grad_norm": 1.6875,
+      "learning_rate": 9.431439393873527e-06,
+      "loss": 0.8572,
+      "step": 9113
+    },
+    {
+      "epoch": 1.5507426538354374,
+      "grad_norm": 1.65625,
+      "learning_rate": 9.429636396274544e-06,
+      "loss": 0.8009,
+      "step": 9114
+    },
+    {
+      "epoch": 1.5509141179244272,
+      "grad_norm": 1.7578125,
+      "learning_rate": 9.427833417277268e-06,
+      "loss": 0.8088,
+      "step": 9115
+    },
+    {
+      "epoch": 1.551085582013417,
+      "grad_norm": 1.65625,
+      "learning_rate": 9.426030456940502e-06,
+      "loss": 0.8109,
+      "step": 9116
+    },
+    {
+      "epoch": 1.551257046102407,
+      "grad_norm": 1.7578125,
+      "learning_rate": 9.42422751532304e-06,
+      "loss": 0.8577,
+      "step": 9117
+    },
+    {
+      "epoch": 1.5514285101913967,
+      "grad_norm": 1.7421875,
+      "learning_rate": 9.422424592483687e-06,
+      "loss": 0.9251,
+      "step": 9118
+    },
+    {
+      "epoch": 1.5515999742803865,
+      "grad_norm": 1.6953125,
+      "learning_rate": 9.420621688481244e-06,
+      "loss": 0.8698,
+      "step": 9119
+    },
+    {
+      "epoch": 1.5517714383693764,
+      "grad_norm": 1.65625,
+      "learning_rate": 9.418818803374506e-06,
+      "loss": 0.8414,
+      "step": 9120
+    },
+    {
+      "epoch": 1.5519429024583664,
+      "grad_norm": 1.71875,
+      "learning_rate": 9.417015937222279e-06,
+      "loss": 0.8404,
+      "step": 9121
+    },
+    {
+      "epoch": 1.5521143665473562,
+      "grad_norm": 1.6484375,
+      "learning_rate": 9.415213090083355e-06,
+      "loss": 0.8677,
+      "step": 9122
+    },
+    {
+      "epoch": 1.552285830636346,
+      "grad_norm": 1.7421875,
+      "learning_rate": 9.413410262016536e-06,
+      "loss": 0.8818,
+      "step": 9123
+    },
+    {
+      "epoch": 1.5524572947253361,
+      "grad_norm": 1.6640625,
+      "learning_rate": 9.411607453080616e-06,
+      "loss": 0.8228,
+      "step": 9124
+    },
+    {
+      "epoch": 1.552628758814326,
+      "grad_norm": 1.6328125,
+      "learning_rate": 9.409804663334395e-06,
+      "loss": 0.8459,
+      "step": 9125
+    },
+    {
+      "epoch": 1.5528002229033158,
+      "grad_norm": 1.703125,
+      "learning_rate": 9.408001892836665e-06,
+      "loss": 0.8643,
+      "step": 9126
+    },
+    {
+      "epoch": 1.5529716869923056,
+      "grad_norm": 1.671875,
+      "learning_rate": 9.406199141646227e-06,
+      "loss": 0.8558,
+      "step": 9127
+    },
+    {
+      "epoch": 1.5531431510812954,
+      "grad_norm": 1.6875,
+      "learning_rate": 9.404396409821863e-06,
+      "loss": 0.8688,
+      "step": 9128
+    },
+    {
+      "epoch": 1.5533146151702852,
+      "grad_norm": 1.6875,
+      "learning_rate": 9.40259369742238e-06,
+      "loss": 0.8741,
+      "step": 9129
+    },
+    {
+      "epoch": 1.553486079259275,
+      "grad_norm": 1.65625,
+      "learning_rate": 9.400791004506564e-06,
+      "loss": 0.826,
+      "step": 9130
+    },
+    {
+      "epoch": 1.553657543348265,
+      "grad_norm": 1.7109375,
+      "learning_rate": 9.39898833113321e-06,
+      "loss": 0.8623,
+      "step": 9131
+    },
+    {
+      "epoch": 1.5538290074372547,
+      "grad_norm": 1.7734375,
+      "learning_rate": 9.397185677361112e-06,
+      "loss": 0.925,
+      "step": 9132
+    },
+    {
+      "epoch": 1.5540004715262448,
+      "grad_norm": 30.375,
+      "learning_rate": 9.395383043249057e-06,
+      "loss": 1.2323,
+      "step": 9133
+    },
+    {
+      "epoch": 1.5541719356152346,
+      "grad_norm": 1.65625,
+      "learning_rate": 9.393580428855837e-06,
+      "loss": 0.8077,
+      "step": 9134
+    },
+    {
+      "epoch": 1.5543433997042244,
+      "grad_norm": 1.7890625,
+      "learning_rate": 9.391777834240245e-06,
+      "loss": 0.9564,
+      "step": 9135
+    },
+    {
+      "epoch": 1.5545148637932145,
+      "grad_norm": 1.734375,
+      "learning_rate": 9.389975259461065e-06,
+      "loss": 0.852,
+      "step": 9136
+    },
+    {
+      "epoch": 1.5546863278822043,
+      "grad_norm": 1.7109375,
+      "learning_rate": 9.388172704577092e-06,
+      "loss": 0.9216,
+      "step": 9137
+    },
+    {
+      "epoch": 1.5548577919711941,
+      "grad_norm": 1.75,
+      "learning_rate": 9.38637016964711e-06,
+      "loss": 0.895,
+      "step": 9138
+    },
+    {
+      "epoch": 1.555029256060184,
+      "grad_norm": 1.6640625,
+      "learning_rate": 9.38456765472991e-06,
+      "loss": 0.8907,
+      "step": 9139
+    },
+    {
+      "epoch": 1.5552007201491738,
+      "grad_norm": 1.6953125,
+      "learning_rate": 9.382765159884272e-06,
+      "loss": 0.8439,
+      "step": 9140
+    },
+    {
+      "epoch": 1.5553721842381636,
+      "grad_norm": 1.671875,
+      "learning_rate": 9.38096268516899e-06,
+      "loss": 0.8292,
+      "step": 9141
+    },
+    {
+      "epoch": 1.5555436483271534,
+      "grad_norm": 1.71875,
+      "learning_rate": 9.379160230642844e-06,
+      "loss": 0.8789,
+      "step": 9142
+    },
+    {
+      "epoch": 1.5557151124161432,
+      "grad_norm": 1.7265625,
+      "learning_rate": 9.37735779636462e-06,
+      "loss": 0.9112,
+      "step": 9143
+    },
+    {
+      "epoch": 1.555886576505133,
+      "grad_norm": 1.78125,
+      "learning_rate": 9.375555382393103e-06,
+      "loss": 0.991,
+      "step": 9144
+    },
+    {
+      "epoch": 1.5560580405941231,
+      "grad_norm": 1.7421875,
+      "learning_rate": 9.373752988787077e-06,
+      "loss": 0.856,
+      "step": 9145
+    },
+    {
+      "epoch": 1.556229504683113,
+      "grad_norm": 1.6015625,
+      "learning_rate": 9.371950615605326e-06,
+      "loss": 0.8673,
+      "step": 9146
+    },
+    {
+      "epoch": 1.5564009687721028,
+      "grad_norm": 1.703125,
+      "learning_rate": 9.370148262906628e-06,
+      "loss": 0.8611,
+      "step": 9147
+    },
+    {
+      "epoch": 1.5565724328610928,
+      "grad_norm": 1.65625,
+      "learning_rate": 9.368345930749767e-06,
+      "loss": 0.8286,
+      "step": 9148
+    },
+    {
+      "epoch": 1.5567438969500826,
+      "grad_norm": 1.6484375,
+      "learning_rate": 9.366543619193526e-06,
+      "loss": 0.8725,
+      "step": 9149
+    },
+    {
+      "epoch": 1.5569153610390725,
+      "grad_norm": 1.6875,
+      "learning_rate": 9.364741328296683e-06,
+      "loss": 0.8377,
+      "step": 9150
+    },
+    {
+      "epoch": 1.5570868251280623,
+      "grad_norm": 1.7734375,
+      "learning_rate": 9.362939058118015e-06,
+      "loss": 0.9226,
+      "step": 9151
+    },
+    {
+      "epoch": 1.5572582892170521,
+      "grad_norm": 1.671875,
+      "learning_rate": 9.361136808716305e-06,
+      "loss": 0.9356,
+      "step": 9152
+    },
+    {
+      "epoch": 1.557429753306042,
+      "grad_norm": 1.6171875,
+      "learning_rate": 9.359334580150328e-06,
+      "loss": 0.9033,
+      "step": 9153
+    },
+    {
+      "epoch": 1.5576012173950318,
+      "grad_norm": 1.6640625,
+      "learning_rate": 9.357532372478863e-06,
+      "loss": 0.8383,
+      "step": 9154
+    },
+    {
+      "epoch": 1.5577726814840216,
+      "grad_norm": 1.6875,
+      "learning_rate": 9.355730185760688e-06,
+      "loss": 0.9208,
+      "step": 9155
+    },
+    {
+      "epoch": 1.5579441455730114,
+      "grad_norm": 1.671875,
+      "learning_rate": 9.353928020054576e-06,
+      "loss": 0.8347,
+      "step": 9156
+    },
+    {
+      "epoch": 1.5581156096620015,
+      "grad_norm": 1.71875,
+      "learning_rate": 9.352125875419306e-06,
+      "loss": 0.8323,
+      "step": 9157
+    },
+    {
+      "epoch": 1.5582870737509913,
+      "grad_norm": 1.75,
+      "learning_rate": 9.35032375191365e-06,
+      "loss": 0.864,
+      "step": 9158
+    },
+    {
+      "epoch": 1.5584585378399811,
+      "grad_norm": 1.7109375,
+      "learning_rate": 9.34852164959638e-06,
+      "loss": 0.9107,
+      "step": 9159
+    },
+    {
+      "epoch": 1.5586300019289712,
+      "grad_norm": 1.8125,
+      "learning_rate": 9.346719568526275e-06,
+      "loss": 0.8692,
+      "step": 9160
+    },
+    {
+      "epoch": 1.558801466017961,
+      "grad_norm": 1.703125,
+      "learning_rate": 9.344917508762104e-06,
+      "loss": 0.9507,
+      "step": 9161
+    },
+    {
+      "epoch": 1.5589729301069508,
+      "grad_norm": 1.75,
+      "learning_rate": 9.343115470362645e-06,
+      "loss": 0.8296,
+      "step": 9162
+    },
+    {
+      "epoch": 1.5591443941959406,
+      "grad_norm": 1.765625,
+      "learning_rate": 9.34131345338666e-06,
+      "loss": 0.8738,
+      "step": 9163
+    },
+    {
+      "epoch": 1.5593158582849305,
+      "grad_norm": 1.6171875,
+      "learning_rate": 9.339511457892924e-06,
+      "loss": 0.8605,
+      "step": 9164
+    },
+    {
+      "epoch": 1.5594873223739203,
+      "grad_norm": 1.640625,
+      "learning_rate": 9.337709483940208e-06,
+      "loss": 0.8543,
+      "step": 9165
+    },
+    {
+      "epoch": 1.5596587864629101,
+      "grad_norm": 1.75,
+      "learning_rate": 9.335907531587275e-06,
+      "loss": 0.791,
+      "step": 9166
+    },
+    {
+      "epoch": 1.5598302505519,
+      "grad_norm": 1.7265625,
+      "learning_rate": 9.334105600892899e-06,
+      "loss": 0.8977,
+      "step": 9167
+    },
+    {
+      "epoch": 1.5600017146408898,
+      "grad_norm": 1.578125,
+      "learning_rate": 9.33230369191585e-06,
+      "loss": 0.8126,
+      "step": 9168
+    },
+    {
+      "epoch": 1.5601731787298796,
+      "grad_norm": 1.625,
+      "learning_rate": 9.33050180471489e-06,
+      "loss": 0.8863,
+      "step": 9169
+    },
+    {
+      "epoch": 1.5603446428188696,
+      "grad_norm": 1.6875,
+      "learning_rate": 9.328699939348788e-06,
+      "loss": 0.8737,
+      "step": 9170
+    },
+    {
+      "epoch": 1.5605161069078595,
+      "grad_norm": 1.65625,
+      "learning_rate": 9.326898095876308e-06,
+      "loss": 0.8067,
+      "step": 9171
+    },
+    {
+      "epoch": 1.5606875709968493,
+      "grad_norm": 1.6015625,
+      "learning_rate": 9.325096274356219e-06,
+      "loss": 0.8524,
+      "step": 9172
+    },
+    {
+      "epoch": 1.5608590350858393,
+      "grad_norm": 1.7109375,
+      "learning_rate": 9.323294474847281e-06,
+      "loss": 0.8329,
+      "step": 9173
+    },
+    {
+      "epoch": 1.5610304991748292,
+      "grad_norm": 1.625,
+      "learning_rate": 9.321492697408259e-06,
+      "loss": 0.7939,
+      "step": 9174
+    },
+    {
+      "epoch": 1.561201963263819,
+      "grad_norm": 1.7890625,
+      "learning_rate": 9.31969094209791e-06,
+      "loss": 0.8204,
+      "step": 9175
+    },
+    {
+      "epoch": 1.5613734273528088,
+      "grad_norm": 1.65625,
+      "learning_rate": 9.317889208975004e-06,
+      "loss": 0.8583,
+      "step": 9176
+    },
+    {
+      "epoch": 1.5615448914417986,
+      "grad_norm": 1.59375,
+      "learning_rate": 9.3160874980983e-06,
+      "loss": 0.8618,
+      "step": 9177
+    },
+    {
+      "epoch": 1.5617163555307885,
+      "grad_norm": 1.6171875,
+      "learning_rate": 9.314285809526558e-06,
+      "loss": 0.812,
+      "step": 9178
+    },
+    {
+      "epoch": 1.5618878196197783,
+      "grad_norm": 1.5859375,
+      "learning_rate": 9.312484143318538e-06,
+      "loss": 0.8151,
+      "step": 9179
+    },
+    {
+      "epoch": 1.5620592837087681,
+      "grad_norm": 2.03125,
+      "learning_rate": 9.310682499532998e-06,
+      "loss": 0.9445,
+      "step": 9180
+    },
+    {
+      "epoch": 1.562230747797758,
+      "grad_norm": 1.8046875,
+      "learning_rate": 9.308880878228698e-06,
+      "loss": 0.9197,
+      "step": 9181
+    },
+    {
+      "epoch": 1.562402211886748,
+      "grad_norm": 1.6796875,
+      "learning_rate": 9.307079279464396e-06,
+      "loss": 0.8719,
+      "step": 9182
+    },
+    {
+      "epoch": 1.5625736759757378,
+      "grad_norm": 1.7734375,
+      "learning_rate": 9.305277703298847e-06,
+      "loss": 0.8668,
+      "step": 9183
+    },
+    {
+      "epoch": 1.5627451400647276,
+      "grad_norm": 1.671875,
+      "learning_rate": 9.303476149790808e-06,
+      "loss": 0.797,
+      "step": 9184
+    },
+    {
+      "epoch": 1.5629166041537177,
+      "grad_norm": 1.7109375,
+      "learning_rate": 9.301674618999038e-06,
+      "loss": 0.8912,
+      "step": 9185
+    },
+    {
+      "epoch": 1.5630880682427075,
+      "grad_norm": 1.65625,
+      "learning_rate": 9.299873110982285e-06,
+      "loss": 0.8996,
+      "step": 9186
+    },
+    {
+      "epoch": 1.5632595323316973,
+      "grad_norm": 1.7265625,
+      "learning_rate": 9.298071625799306e-06,
+      "loss": 0.8605,
+      "step": 9187
+    },
+    {
+      "epoch": 1.5634309964206872,
+      "grad_norm": 1.6015625,
+      "learning_rate": 9.296270163508856e-06,
+      "loss": 0.8674,
+      "step": 9188
+    },
+    {
+      "epoch": 1.563602460509677,
+      "grad_norm": 1.7265625,
+      "learning_rate": 9.294468724169685e-06,
+      "loss": 0.886,
+      "step": 9189
+    },
+    {
+      "epoch": 1.5637739245986668,
+      "grad_norm": 1.703125,
+      "learning_rate": 9.292667307840546e-06,
+      "loss": 0.8608,
+      "step": 9190
+    },
+    {
+      "epoch": 1.5639453886876566,
+      "grad_norm": 1.703125,
+      "learning_rate": 9.29086591458019e-06,
+      "loss": 0.8972,
+      "step": 9191
+    },
+    {
+      "epoch": 1.5641168527766465,
+      "grad_norm": 1.65625,
+      "learning_rate": 9.289064544447366e-06,
+      "loss": 0.8055,
+      "step": 9192
+    },
+    {
+      "epoch": 1.5642883168656363,
+      "grad_norm": 1.71875,
+      "learning_rate": 9.287263197500825e-06,
+      "loss": 0.9147,
+      "step": 9193
+    },
+    {
+      "epoch": 1.5644597809546263,
+      "grad_norm": 1.6875,
+      "learning_rate": 9.285461873799315e-06,
+      "loss": 0.8236,
+      "step": 9194
+    },
+    {
+      "epoch": 1.5646312450436162,
+      "grad_norm": 1.6328125,
+      "learning_rate": 9.283660573401582e-06,
+      "loss": 0.827,
+      "step": 9195
+    },
+    {
+      "epoch": 1.564802709132606,
+      "grad_norm": 1.6484375,
+      "learning_rate": 9.28185929636638e-06,
+      "loss": 0.8257,
+      "step": 9196
+    },
+    {
+      "epoch": 1.564974173221596,
+      "grad_norm": 1.7421875,
+      "learning_rate": 9.280058042752446e-06,
+      "loss": 0.8898,
+      "step": 9197
+    },
+    {
+      "epoch": 1.5651456373105859,
+      "grad_norm": 1.6640625,
+      "learning_rate": 9.278256812618534e-06,
+      "loss": 0.8702,
+      "step": 9198
+    },
+    {
+      "epoch": 1.5653171013995757,
+      "grad_norm": 1.6328125,
+      "learning_rate": 9.276455606023382e-06,
+      "loss": 0.8156,
+      "step": 9199
+    },
+    {
+      "epoch": 1.5654885654885655,
+      "grad_norm": 1.71875,
+      "learning_rate": 9.274654423025739e-06,
+      "loss": 0.8452,
+      "step": 9200
+    },
+    {
+      "epoch": 1.5656600295775553,
+      "grad_norm": 1.7578125,
+      "learning_rate": 9.272853263684345e-06,
+      "loss": 0.8568,
+      "step": 9201
+    },
+    {
+      "epoch": 1.5658314936665452,
+      "grad_norm": 1.6953125,
+      "learning_rate": 9.271052128057946e-06,
+      "loss": 0.8569,
+      "step": 9202
+    },
+    {
+      "epoch": 1.566002957755535,
+      "grad_norm": 1.9375,
+      "learning_rate": 9.26925101620528e-06,
+      "loss": 0.8666,
+      "step": 9203
+    },
+    {
+      "epoch": 1.5661744218445248,
+      "grad_norm": 1.765625,
+      "learning_rate": 9.267449928185091e-06,
+      "loss": 0.8248,
+      "step": 9204
+    },
+    {
+      "epoch": 1.5663458859335146,
+      "grad_norm": 1.765625,
+      "learning_rate": 9.265648864056119e-06,
+      "loss": 0.8597,
+      "step": 9205
+    },
+    {
+      "epoch": 1.5665173500225047,
+      "grad_norm": 1.8671875,
+      "learning_rate": 9.2638478238771e-06,
+      "loss": 0.8764,
+      "step": 9206
+    },
+    {
+      "epoch": 1.5666888141114945,
+      "grad_norm": 1.65625,
+      "learning_rate": 9.26204680770678e-06,
+      "loss": 0.8861,
+      "step": 9207
+    },
+    {
+      "epoch": 1.5668602782004843,
+      "grad_norm": 1.65625,
+      "learning_rate": 9.260245815603894e-06,
+      "loss": 0.8443,
+      "step": 9208
+    },
+    {
+      "epoch": 1.5670317422894744,
+      "grad_norm": 1.7109375,
+      "learning_rate": 9.258444847627174e-06,
+      "loss": 0.859,
+      "step": 9209
+    },
+    {
+      "epoch": 1.5672032063784642,
+      "grad_norm": 1.640625,
+      "learning_rate": 9.256643903835361e-06,
+      "loss": 0.9118,
+      "step": 9210
+    },
+    {
+      "epoch": 1.567374670467454,
+      "grad_norm": 1.6484375,
+      "learning_rate": 9.254842984287187e-06,
+      "loss": 0.7796,
+      "step": 9211
+    },
+    {
+      "epoch": 1.5675461345564439,
+      "grad_norm": 1.5703125,
+      "learning_rate": 9.253042089041392e-06,
+      "loss": 0.7964,
+      "step": 9212
+    },
+    {
+      "epoch": 1.5677175986454337,
+      "grad_norm": 1.6015625,
+      "learning_rate": 9.251241218156706e-06,
+      "loss": 0.8026,
+      "step": 9213
+    },
+    {
+      "epoch": 1.5678890627344235,
+      "grad_norm": 1.65625,
+      "learning_rate": 9.249440371691863e-06,
+      "loss": 0.7103,
+      "step": 9214
+    },
+    {
+      "epoch": 1.5680605268234133,
+      "grad_norm": 1.7265625,
+      "learning_rate": 9.247639549705597e-06,
+      "loss": 0.8532,
+      "step": 9215
+    },
+    {
+      "epoch": 1.5682319909124032,
+      "grad_norm": 1.7421875,
+      "learning_rate": 9.245838752256636e-06,
+      "loss": 0.8715,
+      "step": 9216
+    },
+    {
+      "epoch": 1.568403455001393,
+      "grad_norm": 1.734375,
+      "learning_rate": 9.244037979403715e-06,
+      "loss": 0.9013,
+      "step": 9217
+    },
+    {
+      "epoch": 1.568574919090383,
+      "grad_norm": 1.8046875,
+      "learning_rate": 9.242237231205563e-06,
+      "loss": 0.8126,
+      "step": 9218
+    },
+    {
+      "epoch": 1.5687463831793729,
+      "grad_norm": 1.640625,
+      "learning_rate": 9.240436507720907e-06,
+      "loss": 0.816,
+      "step": 9219
+    },
+    {
+      "epoch": 1.5689178472683627,
+      "grad_norm": 1.671875,
+      "learning_rate": 9.238635809008482e-06,
+      "loss": 0.8114,
+      "step": 9220
+    },
+    {
+      "epoch": 1.5690893113573527,
+      "grad_norm": 1.796875,
+      "learning_rate": 9.236835135127005e-06,
+      "loss": 0.8607,
+      "step": 9221
+    },
+    {
+      "epoch": 1.5692607754463426,
+      "grad_norm": 1.703125,
+      "learning_rate": 9.235034486135207e-06,
+      "loss": 0.8988,
+      "step": 9222
+    },
+    {
+      "epoch": 1.5694322395353324,
+      "grad_norm": 1.71875,
+      "learning_rate": 9.233233862091818e-06,
+      "loss": 0.8391,
+      "step": 9223
+    },
+    {
+      "epoch": 1.5696037036243222,
+      "grad_norm": 1.65625,
+      "learning_rate": 9.231433263055558e-06,
+      "loss": 0.8328,
+      "step": 9224
+    },
+    {
+      "epoch": 1.569775167713312,
+      "grad_norm": 1.6484375,
+      "learning_rate": 9.229632689085155e-06,
+      "loss": 0.8381,
+      "step": 9225
+    },
+    {
+      "epoch": 1.5699466318023019,
+      "grad_norm": 1.65625,
+      "learning_rate": 9.22783214023933e-06,
+      "loss": 0.8908,
+      "step": 9226
+    },
+    {
+      "epoch": 1.5701180958912917,
+      "grad_norm": 1.7109375,
+      "learning_rate": 9.226031616576806e-06,
+      "loss": 0.8383,
+      "step": 9227
+    },
+    {
+      "epoch": 1.5702895599802815,
+      "grad_norm": 1.734375,
+      "learning_rate": 9.224231118156307e-06,
+      "loss": 0.9127,
+      "step": 9228
+    },
+    {
+      "epoch": 1.5704610240692713,
+      "grad_norm": 1.7578125,
+      "learning_rate": 9.222430645036552e-06,
+      "loss": 0.8773,
+      "step": 9229
+    },
+    {
+      "epoch": 1.5706324881582614,
+      "grad_norm": 1.640625,
+      "learning_rate": 9.220630197276262e-06,
+      "loss": 0.8129,
+      "step": 9230
+    },
+    {
+      "epoch": 1.5708039522472512,
+      "grad_norm": 1.6875,
+      "learning_rate": 9.218829774934157e-06,
+      "loss": 0.8681,
+      "step": 9231
+    },
+    {
+      "epoch": 1.570975416336241,
+      "grad_norm": 1.609375,
+      "learning_rate": 9.217029378068954e-06,
+      "loss": 0.8128,
+      "step": 9232
+    },
+    {
+      "epoch": 1.571146880425231,
+      "grad_norm": 1.609375,
+      "learning_rate": 9.21522900673937e-06,
+      "loss": 0.7568,
+      "step": 9233
+    },
+    {
+      "epoch": 1.571318344514221,
+      "grad_norm": 1.640625,
+      "learning_rate": 9.213428661004124e-06,
+      "loss": 0.7764,
+      "step": 9234
+    },
+    {
+      "epoch": 1.5714898086032107,
+      "grad_norm": 1.6796875,
+      "learning_rate": 9.211628340921932e-06,
+      "loss": 0.8369,
+      "step": 9235
+    },
+    {
+      "epoch": 1.5716612726922006,
+      "grad_norm": 1.734375,
+      "learning_rate": 9.209828046551507e-06,
+      "loss": 0.8511,
+      "step": 9236
+    },
+    {
+      "epoch": 1.5718327367811904,
+      "grad_norm": 1.6953125,
+      "learning_rate": 9.208027777951565e-06,
+      "loss": 0.8305,
+      "step": 9237
+    },
+    {
+      "epoch": 1.5720042008701802,
+      "grad_norm": 1.78125,
+      "learning_rate": 9.206227535180821e-06,
+      "loss": 0.8703,
+      "step": 9238
+    },
+    {
+      "epoch": 1.57217566495917,
+      "grad_norm": 1.703125,
+      "learning_rate": 9.204427318297987e-06,
+      "loss": 0.8569,
+      "step": 9239
+    },
+    {
+      "epoch": 1.5723471290481599,
+      "grad_norm": 1.7734375,
+      "learning_rate": 9.202627127361772e-06,
+      "loss": 0.8437,
+      "step": 9240
+    },
+    {
+      "epoch": 1.5725185931371497,
+      "grad_norm": 1.6796875,
+      "learning_rate": 9.20082696243089e-06,
+      "loss": 0.8106,
+      "step": 9241
+    },
+    {
+      "epoch": 1.5726900572261397,
+      "grad_norm": 1.6171875,
+      "learning_rate": 9.199026823564048e-06,
+      "loss": 0.7673,
+      "step": 9242
+    },
+    {
+      "epoch": 1.5728615213151296,
+      "grad_norm": 1.6796875,
+      "learning_rate": 9.197226710819961e-06,
+      "loss": 0.8538,
+      "step": 9243
+    },
+    {
+      "epoch": 1.5730329854041194,
+      "grad_norm": 1.734375,
+      "learning_rate": 9.195426624257332e-06,
+      "loss": 0.9001,
+      "step": 9244
+    },
+    {
+      "epoch": 1.5732044494931094,
+      "grad_norm": 1.734375,
+      "learning_rate": 9.19362656393487e-06,
+      "loss": 0.8341,
+      "step": 9245
+    },
+    {
+      "epoch": 1.5733759135820993,
+      "grad_norm": 1.640625,
+      "learning_rate": 9.191826529911283e-06,
+      "loss": 0.903,
+      "step": 9246
+    },
+    {
+      "epoch": 1.573547377671089,
+      "grad_norm": 1.6640625,
+      "learning_rate": 9.190026522245277e-06,
+      "loss": 0.8623,
+      "step": 9247
+    },
+    {
+      "epoch": 1.573718841760079,
+      "grad_norm": 1.6875,
+      "learning_rate": 9.188226540995555e-06,
+      "loss": 0.8689,
+      "step": 9248
+    },
+    {
+      "epoch": 1.5738903058490687,
+      "grad_norm": 1.6640625,
+      "learning_rate": 9.186426586220822e-06,
+      "loss": 0.8787,
+      "step": 9249
+    },
+    {
+      "epoch": 1.5740617699380586,
+      "grad_norm": 1.6953125,
+      "learning_rate": 9.184626657979782e-06,
+      "loss": 0.8006,
+      "step": 9250
+    },
+    {
+      "epoch": 1.5742332340270484,
+      "grad_norm": 1.640625,
+      "learning_rate": 9.182826756331137e-06,
+      "loss": 0.8874,
+      "step": 9251
+    },
+    {
+      "epoch": 1.5744046981160382,
+      "grad_norm": 1.703125,
+      "learning_rate": 9.181026881333589e-06,
+      "loss": 0.8036,
+      "step": 9252
+    },
+    {
+      "epoch": 1.574576162205028,
+      "grad_norm": 1.6484375,
+      "learning_rate": 9.179227033045836e-06,
+      "loss": 0.8686,
+      "step": 9253
+    },
+    {
+      "epoch": 1.574747626294018,
+      "grad_norm": 1.75,
+      "learning_rate": 9.177427211526585e-06,
+      "loss": 0.9301,
+      "step": 9254
+    },
+    {
+      "epoch": 1.574919090383008,
+      "grad_norm": 1.7265625,
+      "learning_rate": 9.17562741683453e-06,
+      "loss": 0.8932,
+      "step": 9255
+    },
+    {
+      "epoch": 1.5750905544719978,
+      "grad_norm": 1.640625,
+      "learning_rate": 9.173827649028366e-06,
+      "loss": 0.8502,
+      "step": 9256
+    },
+    {
+      "epoch": 1.5752620185609878,
+      "grad_norm": 1.6796875,
+      "learning_rate": 9.172027908166795e-06,
+      "loss": 0.8476,
+      "step": 9257
+    },
+    {
+      "epoch": 1.5754334826499776,
+      "grad_norm": 1.59375,
+      "learning_rate": 9.17022819430851e-06,
+      "loss": 0.8561,
+      "step": 9258
+    },
+    {
+      "epoch": 1.5756049467389674,
+      "grad_norm": 1.6484375,
+      "learning_rate": 9.16842850751221e-06,
+      "loss": 0.7854,
+      "step": 9259
+    },
+    {
+      "epoch": 1.5757764108279573,
+      "grad_norm": 1.7265625,
+      "learning_rate": 9.166628847836586e-06,
+      "loss": 0.7733,
+      "step": 9260
+    },
+    {
+      "epoch": 1.575947874916947,
+      "grad_norm": 1.8046875,
+      "learning_rate": 9.164829215340332e-06,
+      "loss": 0.8804,
+      "step": 9261
+    },
+    {
+      "epoch": 1.576119339005937,
+      "grad_norm": 1.6640625,
+      "learning_rate": 9.163029610082146e-06,
+      "loss": 0.7935,
+      "step": 9262
+    },
+    {
+      "epoch": 1.5762908030949268,
+      "grad_norm": 1.7265625,
+      "learning_rate": 9.161230032120714e-06,
+      "loss": 0.8089,
+      "step": 9263
+    },
+    {
+      "epoch": 1.5764622671839166,
+      "grad_norm": 1.59375,
+      "learning_rate": 9.15943048151473e-06,
+      "loss": 0.8143,
+      "step": 9264
+    },
+    {
+      "epoch": 1.5766337312729064,
+      "grad_norm": 1.703125,
+      "learning_rate": 9.157630958322884e-06,
+      "loss": 0.8527,
+      "step": 9265
+    },
+    {
+      "epoch": 1.5768051953618962,
+      "grad_norm": 1.71875,
+      "learning_rate": 9.155831462603867e-06,
+      "loss": 0.8017,
+      "step": 9266
+    },
+    {
+      "epoch": 1.5769766594508863,
+      "grad_norm": 1.6015625,
+      "learning_rate": 9.154031994416362e-06,
+      "loss": 0.8363,
+      "step": 9267
+    },
+    {
+      "epoch": 1.577148123539876,
+      "grad_norm": 1.7109375,
+      "learning_rate": 9.15223255381906e-06,
+      "loss": 0.8843,
+      "step": 9268
+    },
+    {
+      "epoch": 1.577319587628866,
+      "grad_norm": 1.796875,
+      "learning_rate": 9.150433140870642e-06,
+      "loss": 0.8803,
+      "step": 9269
+    },
+    {
+      "epoch": 1.577491051717856,
+      "grad_norm": 1.7421875,
+      "learning_rate": 9.148633755629803e-06,
+      "loss": 0.761,
+      "step": 9270
+    },
+    {
+      "epoch": 1.5776625158068458,
+      "grad_norm": 1.6796875,
+      "learning_rate": 9.146834398155223e-06,
+      "loss": 0.7771,
+      "step": 9271
+    },
+    {
+      "epoch": 1.5778339798958356,
+      "grad_norm": 1.5859375,
+      "learning_rate": 9.145035068505586e-06,
+      "loss": 0.7734,
+      "step": 9272
+    },
+    {
+      "epoch": 1.5780054439848255,
+      "grad_norm": 1.671875,
+      "learning_rate": 9.143235766739574e-06,
+      "loss": 0.8499,
+      "step": 9273
+    },
+    {
+      "epoch": 1.5781769080738153,
+      "grad_norm": 1.640625,
+      "learning_rate": 9.141436492915869e-06,
+      "loss": 0.8435,
+      "step": 9274
+    },
+    {
+      "epoch": 1.578348372162805,
+      "grad_norm": 1.65625,
+      "learning_rate": 9.139637247093156e-06,
+      "loss": 0.876,
+      "step": 9275
+    },
+    {
+      "epoch": 1.578519836251795,
+      "grad_norm": 1.7265625,
+      "learning_rate": 9.13783802933011e-06,
+      "loss": 0.8139,
+      "step": 9276
+    },
+    {
+      "epoch": 1.5786913003407848,
+      "grad_norm": 1.734375,
+      "learning_rate": 9.136038839685415e-06,
+      "loss": 0.935,
+      "step": 9277
+    },
+    {
+      "epoch": 1.5788627644297746,
+      "grad_norm": 1.6875,
+      "learning_rate": 9.134239678217744e-06,
+      "loss": 0.9228,
+      "step": 9278
+    },
+    {
+      "epoch": 1.5790342285187646,
+      "grad_norm": 1.734375,
+      "learning_rate": 9.13244054498578e-06,
+      "loss": 0.8317,
+      "step": 9279
+    },
+    {
+      "epoch": 1.5792056926077545,
+      "grad_norm": 1.7421875,
+      "learning_rate": 9.130641440048194e-06,
+      "loss": 0.8331,
+      "step": 9280
+    },
+    {
+      "epoch": 1.5793771566967443,
+      "grad_norm": 1.7109375,
+      "learning_rate": 9.128842363463663e-06,
+      "loss": 0.792,
+      "step": 9281
+    },
+    {
+      "epoch": 1.5795486207857343,
+      "grad_norm": 1.6953125,
+      "learning_rate": 9.127043315290867e-06,
+      "loss": 0.8437,
+      "step": 9282
+    },
+    {
+      "epoch": 1.5797200848747242,
+      "grad_norm": 1.59375,
+      "learning_rate": 9.125244295588473e-06,
+      "loss": 0.8124,
+      "step": 9283
+    },
+    {
+      "epoch": 1.579891548963714,
+      "grad_norm": 1.78125,
+      "learning_rate": 9.123445304415157e-06,
+      "loss": 0.8643,
+      "step": 9284
+    },
+    {
+      "epoch": 1.5800630130527038,
+      "grad_norm": 1.78125,
+      "learning_rate": 9.12164634182959e-06,
+      "loss": 0.9127,
+      "step": 9285
+    },
+    {
+      "epoch": 1.5802344771416936,
+      "grad_norm": 1.6640625,
+      "learning_rate": 9.119847407890446e-06,
+      "loss": 0.7783,
+      "step": 9286
+    },
+    {
+      "epoch": 1.5804059412306835,
+      "grad_norm": 1.8046875,
+      "learning_rate": 9.118048502656391e-06,
+      "loss": 0.8905,
+      "step": 9287
+    },
+    {
+      "epoch": 1.5805774053196733,
+      "grad_norm": 1.703125,
+      "learning_rate": 9.116249626186094e-06,
+      "loss": 0.8882,
+      "step": 9288
+    },
+    {
+      "epoch": 1.580748869408663,
+      "grad_norm": 1.671875,
+      "learning_rate": 9.114450778538228e-06,
+      "loss": 0.881,
+      "step": 9289
+    },
+    {
+      "epoch": 1.580920333497653,
+      "grad_norm": 1.6796875,
+      "learning_rate": 9.112651959771455e-06,
+      "loss": 0.863,
+      "step": 9290
+    },
+    {
+      "epoch": 1.581091797586643,
+      "grad_norm": 1.65625,
+      "learning_rate": 9.110853169944441e-06,
+      "loss": 0.8287,
+      "step": 9291
+    },
+    {
+      "epoch": 1.5812632616756328,
+      "grad_norm": 1.7109375,
+      "learning_rate": 9.109054409115854e-06,
+      "loss": 0.7872,
+      "step": 9292
+    },
+    {
+      "epoch": 1.5814347257646226,
+      "grad_norm": 1.7421875,
+      "learning_rate": 9.10725567734436e-06,
+      "loss": 0.9243,
+      "step": 9293
+    },
+    {
+      "epoch": 1.5816061898536127,
+      "grad_norm": 1.71875,
+      "learning_rate": 9.105456974688618e-06,
+      "loss": 0.905,
+      "step": 9294
+    },
+    {
+      "epoch": 1.5817776539426025,
+      "grad_norm": 1.75,
+      "learning_rate": 9.103658301207294e-06,
+      "loss": 0.899,
+      "step": 9295
+    },
+    {
+      "epoch": 1.5819491180315923,
+      "grad_norm": 1.8046875,
+      "learning_rate": 9.101859656959047e-06,
+      "loss": 0.8408,
+      "step": 9296
+    },
+    {
+      "epoch": 1.5821205821205822,
+      "grad_norm": 1.7578125,
+      "learning_rate": 9.100061042002539e-06,
+      "loss": 0.7829,
+      "step": 9297
+    },
+    {
+      "epoch": 1.582292046209572,
+      "grad_norm": 1.71875,
+      "learning_rate": 9.09826245639643e-06,
+      "loss": 0.9463,
+      "step": 9298
+    },
+    {
+      "epoch": 1.5824635102985618,
+      "grad_norm": 1.65625,
+      "learning_rate": 9.096463900199376e-06,
+      "loss": 0.8112,
+      "step": 9299
+    },
+    {
+      "epoch": 1.5826349743875516,
+      "grad_norm": 1.7265625,
+      "learning_rate": 9.094665373470039e-06,
+      "loss": 0.8755,
+      "step": 9300
+    },
+    {
+      "epoch": 1.5828064384765415,
+      "grad_norm": 1.78125,
+      "learning_rate": 9.092866876267074e-06,
+      "loss": 0.9219,
+      "step": 9301
+    },
+    {
+      "epoch": 1.5829779025655313,
+      "grad_norm": 1.6796875,
+      "learning_rate": 9.091068408649134e-06,
+      "loss": 0.8709,
+      "step": 9302
+    },
+    {
+      "epoch": 1.5831493666545213,
+      "grad_norm": 1.6484375,
+      "learning_rate": 9.089269970674875e-06,
+      "loss": 0.8616,
+      "step": 9303
+    },
+    {
+      "epoch": 1.5833208307435112,
+      "grad_norm": 1.6328125,
+      "learning_rate": 9.087471562402953e-06,
+      "loss": 0.8286,
+      "step": 9304
+    },
+    {
+      "epoch": 1.583492294832501,
+      "grad_norm": 1.6875,
+      "learning_rate": 9.08567318389202e-06,
+      "loss": 0.8534,
+      "step": 9305
+    },
+    {
+      "epoch": 1.583663758921491,
+      "grad_norm": 1.765625,
+      "learning_rate": 9.083874835200727e-06,
+      "loss": 0.9082,
+      "step": 9306
+    },
+    {
+      "epoch": 1.5838352230104809,
+      "grad_norm": 1.7265625,
+      "learning_rate": 9.082076516387726e-06,
+      "loss": 0.8264,
+      "step": 9307
+    },
+    {
+      "epoch": 1.5840066870994707,
+      "grad_norm": 1.8515625,
+      "learning_rate": 9.080278227511664e-06,
+      "loss": 0.9196,
+      "step": 9308
+    },
+    {
+      "epoch": 1.5841781511884605,
+      "grad_norm": 1.6328125,
+      "learning_rate": 9.078479968631193e-06,
+      "loss": 0.9158,
+      "step": 9309
+    },
+    {
+      "epoch": 1.5843496152774503,
+      "grad_norm": 1.65625,
+      "learning_rate": 9.076681739804961e-06,
+      "loss": 0.8469,
+      "step": 9310
+    },
+    {
+      "epoch": 1.5845210793664402,
+      "grad_norm": 1.703125,
+      "learning_rate": 9.074883541091616e-06,
+      "loss": 0.8779,
+      "step": 9311
+    },
+    {
+      "epoch": 1.58469254345543,
+      "grad_norm": 1.546875,
+      "learning_rate": 9.073085372549805e-06,
+      "loss": 0.7402,
+      "step": 9312
+    },
+    {
+      "epoch": 1.5848640075444198,
+      "grad_norm": 1.78125,
+      "learning_rate": 9.071287234238166e-06,
+      "loss": 0.8958,
+      "step": 9313
+    },
+    {
+      "epoch": 1.5850354716334096,
+      "grad_norm": 1.625,
+      "learning_rate": 9.06948912621535e-06,
+      "loss": 0.8287,
+      "step": 9314
+    },
+    {
+      "epoch": 1.5852069357223997,
+      "grad_norm": 11.0625,
+      "learning_rate": 9.067691048539996e-06,
+      "loss": 0.9041,
+      "step": 9315
+    },
+    {
+      "epoch": 1.5853783998113895,
+      "grad_norm": 1.71875,
+      "learning_rate": 9.065893001270744e-06,
+      "loss": 0.8723,
+      "step": 9316
+    },
+    {
+      "epoch": 1.5855498639003793,
+      "grad_norm": 1.609375,
+      "learning_rate": 9.064094984466242e-06,
+      "loss": 0.8329,
+      "step": 9317
+    },
+    {
+      "epoch": 1.5857213279893694,
+      "grad_norm": 1.59375,
+      "learning_rate": 9.062296998185127e-06,
+      "loss": 0.8169,
+      "step": 9318
+    },
+    {
+      "epoch": 1.5858927920783592,
+      "grad_norm": 1.734375,
+      "learning_rate": 9.060499042486039e-06,
+      "loss": 0.9146,
+      "step": 9319
+    },
+    {
+      "epoch": 1.586064256167349,
+      "grad_norm": 1.6015625,
+      "learning_rate": 9.058701117427613e-06,
+      "loss": 0.8901,
+      "step": 9320
+    },
+    {
+      "epoch": 1.5862357202563389,
+      "grad_norm": 1.78125,
+      "learning_rate": 9.05690322306849e-06,
+      "loss": 0.8549,
+      "step": 9321
+    },
+    {
+      "epoch": 1.5864071843453287,
+      "grad_norm": 1.640625,
+      "learning_rate": 9.055105359467303e-06,
+      "loss": 0.8646,
+      "step": 9322
+    },
+    {
+      "epoch": 1.5865786484343185,
+      "grad_norm": 1.6328125,
+      "learning_rate": 9.053307526682694e-06,
+      "loss": 0.8383,
+      "step": 9323
+    },
+    {
+      "epoch": 1.5867501125233083,
+      "grad_norm": 1.65625,
+      "learning_rate": 9.051509724773283e-06,
+      "loss": 0.835,
+      "step": 9324
+    },
+    {
+      "epoch": 1.5869215766122982,
+      "grad_norm": 1.65625,
+      "learning_rate": 9.049711953797716e-06,
+      "loss": 0.8508,
+      "step": 9325
+    },
+    {
+      "epoch": 1.587093040701288,
+      "grad_norm": 1.640625,
+      "learning_rate": 9.04791421381462e-06,
+      "loss": 0.8332,
+      "step": 9326
+    },
+    {
+      "epoch": 1.587264504790278,
+      "grad_norm": 1.671875,
+      "learning_rate": 9.046116504882625e-06,
+      "loss": 0.8394,
+      "step": 9327
+    },
+    {
+      "epoch": 1.5874359688792679,
+      "grad_norm": 1.7265625,
+      "learning_rate": 9.044318827060364e-06,
+      "loss": 0.8588,
+      "step": 9328
+    },
+    {
+      "epoch": 1.5876074329682577,
+      "grad_norm": 1.8125,
+      "learning_rate": 9.042521180406465e-06,
+      "loss": 0.9078,
+      "step": 9329
+    },
+    {
+      "epoch": 1.5877788970572477,
+      "grad_norm": 1.640625,
+      "learning_rate": 9.040723564979556e-06,
+      "loss": 0.8752,
+      "step": 9330
+    },
+    {
+      "epoch": 1.5879503611462376,
+      "grad_norm": 1.6015625,
+      "learning_rate": 9.038925980838262e-06,
+      "loss": 0.9109,
+      "step": 9331
+    },
+    {
+      "epoch": 1.5881218252352274,
+      "grad_norm": 1.703125,
+      "learning_rate": 9.037128428041214e-06,
+      "loss": 0.8279,
+      "step": 9332
+    },
+    {
+      "epoch": 1.5882932893242172,
+      "grad_norm": 1.6484375,
+      "learning_rate": 9.035330906647031e-06,
+      "loss": 0.8455,
+      "step": 9333
+    },
+    {
+      "epoch": 1.588464753413207,
+      "grad_norm": 1.6875,
+      "learning_rate": 9.033533416714341e-06,
+      "loss": 0.798,
+      "step": 9334
+    },
+    {
+      "epoch": 1.5886362175021969,
+      "grad_norm": 1.6875,
+      "learning_rate": 9.031735958301768e-06,
+      "loss": 0.9019,
+      "step": 9335
+    },
+    {
+      "epoch": 1.5888076815911867,
+      "grad_norm": 1.6328125,
+      "learning_rate": 9.02993853146793e-06,
+      "loss": 0.8193,
+      "step": 9336
+    },
+    {
+      "epoch": 1.5889791456801765,
+      "grad_norm": 1.6484375,
+      "learning_rate": 9.028141136271448e-06,
+      "loss": 0.8333,
+      "step": 9337
+    },
+    {
+      "epoch": 1.5891506097691663,
+      "grad_norm": 1.75,
+      "learning_rate": 9.026343772770944e-06,
+      "loss": 0.8714,
+      "step": 9338
+    },
+    {
+      "epoch": 1.5893220738581564,
+      "grad_norm": 1.7265625,
+      "learning_rate": 9.024546441025035e-06,
+      "loss": 0.8913,
+      "step": 9339
+    },
+    {
+      "epoch": 1.5894935379471462,
+      "grad_norm": 1.6953125,
+      "learning_rate": 9.022749141092341e-06,
+      "loss": 0.8662,
+      "step": 9340
+    },
+    {
+      "epoch": 1.589665002036136,
+      "grad_norm": 1.71875,
+      "learning_rate": 9.020951873031477e-06,
+      "loss": 0.9092,
+      "step": 9341
+    },
+    {
+      "epoch": 1.589836466125126,
+      "grad_norm": 1.75,
+      "learning_rate": 9.019154636901059e-06,
+      "loss": 0.8917,
+      "step": 9342
+    },
+    {
+      "epoch": 1.590007930214116,
+      "grad_norm": 1.640625,
+      "learning_rate": 9.017357432759702e-06,
+      "loss": 0.8658,
+      "step": 9343
+    },
+    {
+      "epoch": 1.5901793943031057,
+      "grad_norm": 1.625,
+      "learning_rate": 9.015560260666019e-06,
+      "loss": 0.8088,
+      "step": 9344
+    },
+    {
+      "epoch": 1.5903508583920956,
+      "grad_norm": 1.65625,
+      "learning_rate": 9.013763120678624e-06,
+      "loss": 0.9525,
+      "step": 9345
+    },
+    {
+      "epoch": 1.5905223224810854,
+      "grad_norm": 1.625,
+      "learning_rate": 9.011966012856129e-06,
+      "loss": 0.853,
+      "step": 9346
+    },
+    {
+      "epoch": 1.5906937865700752,
+      "grad_norm": 1.703125,
+      "learning_rate": 9.01016893725714e-06,
+      "loss": 0.8231,
+      "step": 9347
+    },
+    {
+      "epoch": 1.590865250659065,
+      "grad_norm": 1.75,
+      "learning_rate": 9.00837189394027e-06,
+      "loss": 0.8723,
+      "step": 9348
+    },
+    {
+      "epoch": 1.5910367147480549,
+      "grad_norm": 1.6796875,
+      "learning_rate": 9.006574882964127e-06,
+      "loss": 0.8768,
+      "step": 9349
+    },
+    {
+      "epoch": 1.5912081788370447,
+      "grad_norm": 1.6875,
+      "learning_rate": 9.004777904387318e-06,
+      "loss": 0.8245,
+      "step": 9350
+    },
+    {
+      "epoch": 1.5913796429260347,
+      "grad_norm": 1.703125,
+      "learning_rate": 9.002980958268448e-06,
+      "loss": 0.8931,
+      "step": 9351
+    },
+    {
+      "epoch": 1.5915511070150246,
+      "grad_norm": 1.703125,
+      "learning_rate": 9.001184044666125e-06,
+      "loss": 0.8787,
+      "step": 9352
+    },
+    {
+      "epoch": 1.5917225711040144,
+      "grad_norm": 1.625,
+      "learning_rate": 8.99938716363895e-06,
+      "loss": 0.8112,
+      "step": 9353
+    },
+    {
+      "epoch": 1.5918940351930044,
+      "grad_norm": 1.734375,
+      "learning_rate": 8.997590315245528e-06,
+      "loss": 0.9623,
+      "step": 9354
+    },
+    {
+      "epoch": 1.5920654992819943,
+      "grad_norm": 1.6484375,
+      "learning_rate": 8.995793499544459e-06,
+      "loss": 0.8516,
+      "step": 9355
+    },
+    {
+      "epoch": 1.592236963370984,
+      "grad_norm": 1.7265625,
+      "learning_rate": 8.993996716594347e-06,
+      "loss": 0.893,
+      "step": 9356
+    },
+    {
+      "epoch": 1.592408427459974,
+      "grad_norm": 1.671875,
+      "learning_rate": 8.99219996645379e-06,
+      "loss": 0.8418,
+      "step": 9357
+    },
+    {
+      "epoch": 1.5925798915489637,
+      "grad_norm": 1.71875,
+      "learning_rate": 8.990403249181394e-06,
+      "loss": 0.8901,
+      "step": 9358
+    },
+    {
+      "epoch": 1.5927513556379536,
+      "grad_norm": 1.59375,
+      "learning_rate": 8.988606564835742e-06,
+      "loss": 0.8342,
+      "step": 9359
+    },
+    {
+      "epoch": 1.5929228197269434,
+      "grad_norm": 1.71875,
+      "learning_rate": 8.986809913475441e-06,
+      "loss": 0.8132,
+      "step": 9360
+    },
+    {
+      "epoch": 1.5930942838159332,
+      "grad_norm": 1.7109375,
+      "learning_rate": 8.985013295159083e-06,
+      "loss": 0.8687,
+      "step": 9361
+    },
+    {
+      "epoch": 1.593265747904923,
+      "grad_norm": 1.734375,
+      "learning_rate": 8.983216709945264e-06,
+      "loss": 0.8738,
+      "step": 9362
+    },
+    {
+      "epoch": 1.5934372119939129,
+      "grad_norm": 1.765625,
+      "learning_rate": 8.981420157892574e-06,
+      "loss": 0.9278,
+      "step": 9363
+    },
+    {
+      "epoch": 1.593608676082903,
+      "grad_norm": 1.6796875,
+      "learning_rate": 8.97962363905961e-06,
+      "loss": 0.8692,
+      "step": 9364
+    },
+    {
+      "epoch": 1.5937801401718927,
+      "grad_norm": 1.75,
+      "learning_rate": 8.977827153504963e-06,
+      "loss": 0.8823,
+      "step": 9365
+    },
+    {
+      "epoch": 1.5939516042608826,
+      "grad_norm": 1.6640625,
+      "learning_rate": 8.97603070128722e-06,
+      "loss": 0.8109,
+      "step": 9366
+    },
+    {
+      "epoch": 1.5941230683498726,
+      "grad_norm": 1.8046875,
+      "learning_rate": 8.974234282464974e-06,
+      "loss": 0.8447,
+      "step": 9367
+    },
+    {
+      "epoch": 1.5942945324388624,
+      "grad_norm": 1.609375,
+      "learning_rate": 8.97243789709681e-06,
+      "loss": 0.917,
+      "step": 9368
+    },
+    {
+      "epoch": 1.5944659965278523,
+      "grad_norm": 1.625,
+      "learning_rate": 8.970641545241314e-06,
+      "loss": 0.8451,
+      "step": 9369
+    },
+    {
+      "epoch": 1.594637460616842,
+      "grad_norm": 1.59375,
+      "learning_rate": 8.968845226957079e-06,
+      "loss": 0.7834,
+      "step": 9370
+    },
+    {
+      "epoch": 1.594808924705832,
+      "grad_norm": 1.78125,
+      "learning_rate": 8.967048942302677e-06,
+      "loss": 0.8885,
+      "step": 9371
+    },
+    {
+      "epoch": 1.5949803887948217,
+      "grad_norm": 1.6328125,
+      "learning_rate": 8.965252691336701e-06,
+      "loss": 0.8003,
+      "step": 9372
+    },
+    {
+      "epoch": 1.5951518528838116,
+      "grad_norm": 1.65625,
+      "learning_rate": 8.963456474117732e-06,
+      "loss": 0.7934,
+      "step": 9373
+    },
+    {
+      "epoch": 1.5953233169728014,
+      "grad_norm": 1.6484375,
+      "learning_rate": 8.96166029070435e-06,
+      "loss": 0.8281,
+      "step": 9374
+    },
+    {
+      "epoch": 1.5954947810617912,
+      "grad_norm": 1.6796875,
+      "learning_rate": 8.959864141155137e-06,
+      "loss": 0.8975,
+      "step": 9375
+    },
+    {
+      "epoch": 1.5956662451507813,
+      "grad_norm": 1.609375,
+      "learning_rate": 8.95806802552867e-06,
+      "loss": 0.8455,
+      "step": 9376
+    },
+    {
+      "epoch": 1.595837709239771,
+      "grad_norm": 1.796875,
+      "learning_rate": 8.95627194388353e-06,
+      "loss": 0.8677,
+      "step": 9377
+    },
+    {
+      "epoch": 1.596009173328761,
+      "grad_norm": 1.6796875,
+      "learning_rate": 8.95447589627829e-06,
+      "loss": 0.7717,
+      "step": 9378
+    },
+    {
+      "epoch": 1.596180637417751,
+      "grad_norm": 1.765625,
+      "learning_rate": 8.95267988277153e-06,
+      "loss": 0.8799,
+      "step": 9379
+    },
+    {
+      "epoch": 1.5963521015067408,
+      "grad_norm": 1.7109375,
+      "learning_rate": 8.95088390342182e-06,
+      "loss": 0.8059,
+      "step": 9380
+    },
+    {
+      "epoch": 1.5965235655957306,
+      "grad_norm": 1.703125,
+      "learning_rate": 8.949087958287741e-06,
+      "loss": 0.855,
+      "step": 9381
+    },
+    {
+      "epoch": 1.5966950296847204,
+      "grad_norm": 1.65625,
+      "learning_rate": 8.947292047427858e-06,
+      "loss": 0.8773,
+      "step": 9382
+    },
+    {
+      "epoch": 1.5968664937737103,
+      "grad_norm": 1.7109375,
+      "learning_rate": 8.945496170900745e-06,
+      "loss": 0.8761,
+      "step": 9383
+    },
+    {
+      "epoch": 1.5970379578627,
+      "grad_norm": 1.7265625,
+      "learning_rate": 8.943700328764975e-06,
+      "loss": 0.9317,
+      "step": 9384
+    },
+    {
+      "epoch": 1.59720942195169,
+      "grad_norm": 1.7265625,
+      "learning_rate": 8.941904521079113e-06,
+      "loss": 0.8575,
+      "step": 9385
+    },
+    {
+      "epoch": 1.5973808860406797,
+      "grad_norm": 1.6875,
+      "learning_rate": 8.940108747901728e-06,
+      "loss": 0.845,
+      "step": 9386
+    },
+    {
+      "epoch": 1.5975523501296696,
+      "grad_norm": 1.6953125,
+      "learning_rate": 8.93831300929139e-06,
+      "loss": 0.8701,
+      "step": 9387
+    },
+    {
+      "epoch": 1.5977238142186596,
+      "grad_norm": 1.6328125,
+      "learning_rate": 8.936517305306663e-06,
+      "loss": 0.8437,
+      "step": 9388
+    },
+    {
+      "epoch": 1.5978952783076494,
+      "grad_norm": 1.671875,
+      "learning_rate": 8.93472163600611e-06,
+      "loss": 0.8959,
+      "step": 9389
+    },
+    {
+      "epoch": 1.5980667423966393,
+      "grad_norm": 1.6328125,
+      "learning_rate": 8.932926001448296e-06,
+      "loss": 0.7727,
+      "step": 9390
+    },
+    {
+      "epoch": 1.5982382064856293,
+      "grad_norm": 1.6796875,
+      "learning_rate": 8.931130401691784e-06,
+      "loss": 0.862,
+      "step": 9391
+    },
+    {
+      "epoch": 1.5984096705746191,
+      "grad_norm": 1.671875,
+      "learning_rate": 8.929334836795135e-06,
+      "loss": 0.8813,
+      "step": 9392
+    },
+    {
+      "epoch": 1.598581134663609,
+      "grad_norm": 1.6875,
+      "learning_rate": 8.92753930681691e-06,
+      "loss": 0.9302,
+      "step": 9393
+    },
+    {
+      "epoch": 1.5987525987525988,
+      "grad_norm": 1.71875,
+      "learning_rate": 8.925743811815665e-06,
+      "loss": 0.8973,
+      "step": 9394
+    },
+    {
+      "epoch": 1.5989240628415886,
+      "grad_norm": 1.6875,
+      "learning_rate": 8.923948351849959e-06,
+      "loss": 0.8164,
+      "step": 9395
+    },
+    {
+      "epoch": 1.5990955269305784,
+      "grad_norm": 1.671875,
+      "learning_rate": 8.922152926978349e-06,
+      "loss": 0.8623,
+      "step": 9396
+    },
+    {
+      "epoch": 1.5992669910195683,
+      "grad_norm": 1.7578125,
+      "learning_rate": 8.920357537259393e-06,
+      "loss": 0.7758,
+      "step": 9397
+    },
+    {
+      "epoch": 1.599438455108558,
+      "grad_norm": 1.71875,
+      "learning_rate": 8.918562182751642e-06,
+      "loss": 0.887,
+      "step": 9398
+    },
+    {
+      "epoch": 1.599609919197548,
+      "grad_norm": 1.640625,
+      "learning_rate": 8.916766863513648e-06,
+      "loss": 0.8277,
+      "step": 9399
+    },
+    {
+      "epoch": 1.599781383286538,
+      "grad_norm": 1.765625,
+      "learning_rate": 8.914971579603969e-06,
+      "loss": 0.8924,
+      "step": 9400
+    },
+    {
+      "epoch": 1.5999528473755278,
+      "grad_norm": 1.7109375,
+      "learning_rate": 8.913176331081151e-06,
+      "loss": 0.8944,
+      "step": 9401
+    },
+    {
+      "epoch": 1.6001243114645176,
+      "grad_norm": 1.609375,
+      "learning_rate": 8.911381118003743e-06,
+      "loss": 0.7726,
+      "step": 9402
+    },
+    {
+      "epoch": 1.6002957755535077,
+      "grad_norm": 1.6640625,
+      "learning_rate": 8.909585940430299e-06,
+      "loss": 0.8849,
+      "step": 9403
+    },
+    {
+      "epoch": 1.6004672396424975,
+      "grad_norm": 1.65625,
+      "learning_rate": 8.907790798419369e-06,
+      "loss": 0.8906,
+      "step": 9404
+    },
+    {
+      "epoch": 1.6006387037314873,
+      "grad_norm": 1.7265625,
+      "learning_rate": 8.90599569202949e-06,
+      "loss": 0.9278,
+      "step": 9405
+    },
+    {
+      "epoch": 1.6008101678204771,
+      "grad_norm": 1.6484375,
+      "learning_rate": 8.90420062131921e-06,
+      "loss": 0.8265,
+      "step": 9406
+    },
+    {
+      "epoch": 1.600981631909467,
+      "grad_norm": 1.7578125,
+      "learning_rate": 8.902405586347072e-06,
+      "loss": 0.9074,
+      "step": 9407
+    },
+    {
+      "epoch": 1.6011530959984568,
+      "grad_norm": 1.734375,
+      "learning_rate": 8.900610587171623e-06,
+      "loss": 0.9064,
+      "step": 9408
+    },
+    {
+      "epoch": 1.6013245600874466,
+      "grad_norm": 1.640625,
+      "learning_rate": 8.898815623851402e-06,
+      "loss": 0.8912,
+      "step": 9409
+    },
+    {
+      "epoch": 1.6014960241764364,
+      "grad_norm": 2.0625,
+      "learning_rate": 8.89702069644495e-06,
+      "loss": 0.828,
+      "step": 9410
+    },
+    {
+      "epoch": 1.6016674882654263,
+      "grad_norm": 1.625,
+      "learning_rate": 8.895225805010807e-06,
+      "loss": 0.7718,
+      "step": 9411
+    },
+    {
+      "epoch": 1.6018389523544163,
+      "grad_norm": 1.84375,
+      "learning_rate": 8.893430949607511e-06,
+      "loss": 0.9473,
+      "step": 9412
+    },
+    {
+      "epoch": 1.6020104164434061,
+      "grad_norm": 1.7265625,
+      "learning_rate": 8.891636130293598e-06,
+      "loss": 0.8774,
+      "step": 9413
+    },
+    {
+      "epoch": 1.602181880532396,
+      "grad_norm": 1.875,
+      "learning_rate": 8.889841347127607e-06,
+      "loss": 0.9145,
+      "step": 9414
+    },
+    {
+      "epoch": 1.602353344621386,
+      "grad_norm": 1.7890625,
+      "learning_rate": 8.888046600168067e-06,
+      "loss": 0.8802,
+      "step": 9415
+    },
+    {
+      "epoch": 1.6025248087103758,
+      "grad_norm": 1.671875,
+      "learning_rate": 8.886251889473519e-06,
+      "loss": 0.8245,
+      "step": 9416
+    },
+    {
+      "epoch": 1.6026962727993657,
+      "grad_norm": 1.6796875,
+      "learning_rate": 8.884457215102489e-06,
+      "loss": 0.8716,
+      "step": 9417
+    },
+    {
+      "epoch": 1.6028677368883555,
+      "grad_norm": 1.7421875,
+      "learning_rate": 8.882662577113505e-06,
+      "loss": 0.8579,
+      "step": 9418
+    },
+    {
+      "epoch": 1.6030392009773453,
+      "grad_norm": 1.765625,
+      "learning_rate": 8.880867975565106e-06,
+      "loss": 0.8044,
+      "step": 9419
+    },
+    {
+      "epoch": 1.6032106650663351,
+      "grad_norm": 1.6484375,
+      "learning_rate": 8.879073410515818e-06,
+      "loss": 0.7676,
+      "step": 9420
+    },
+    {
+      "epoch": 1.603382129155325,
+      "grad_norm": 1.6640625,
+      "learning_rate": 8.877278882024163e-06,
+      "loss": 0.8706,
+      "step": 9421
+    },
+    {
+      "epoch": 1.6035535932443148,
+      "grad_norm": 1.7890625,
+      "learning_rate": 8.875484390148674e-06,
+      "loss": 0.9197,
+      "step": 9422
+    },
+    {
+      "epoch": 1.6037250573333046,
+      "grad_norm": 1.6953125,
+      "learning_rate": 8.873689934947873e-06,
+      "loss": 0.9162,
+      "step": 9423
+    },
+    {
+      "epoch": 1.6038965214222947,
+      "grad_norm": 1.6953125,
+      "learning_rate": 8.871895516480286e-06,
+      "loss": 0.9041,
+      "step": 9424
+    },
+    {
+      "epoch": 1.6040679855112845,
+      "grad_norm": 1.6875,
+      "learning_rate": 8.870101134804432e-06,
+      "loss": 0.8439,
+      "step": 9425
+    },
+    {
+      "epoch": 1.6042394496002743,
+      "grad_norm": 1.703125,
+      "learning_rate": 8.868306789978837e-06,
+      "loss": 0.8439,
+      "step": 9426
+    },
+    {
+      "epoch": 1.6044109136892644,
+      "grad_norm": 1.703125,
+      "learning_rate": 8.866512482062019e-06,
+      "loss": 0.8968,
+      "step": 9427
+    },
+    {
+      "epoch": 1.6045823777782542,
+      "grad_norm": 1.6171875,
+      "learning_rate": 8.864718211112497e-06,
+      "loss": 0.7964,
+      "step": 9428
+    },
+    {
+      "epoch": 1.604753841867244,
+      "grad_norm": 1.7734375,
+      "learning_rate": 8.862923977188788e-06,
+      "loss": 0.8404,
+      "step": 9429
+    },
+    {
+      "epoch": 1.6049253059562338,
+      "grad_norm": 1.640625,
+      "learning_rate": 8.861129780349411e-06,
+      "loss": 0.8308,
+      "step": 9430
+    },
+    {
+      "epoch": 1.6050967700452237,
+      "grad_norm": 1.7109375,
+      "learning_rate": 8.85933562065288e-06,
+      "loss": 0.8702,
+      "step": 9431
+    },
+    {
+      "epoch": 1.6052682341342135,
+      "grad_norm": 1.6640625,
+      "learning_rate": 8.85754149815771e-06,
+      "loss": 0.8977,
+      "step": 9432
+    },
+    {
+      "epoch": 1.6054396982232033,
+      "grad_norm": 1.7109375,
+      "learning_rate": 8.855747412922413e-06,
+      "loss": 0.8239,
+      "step": 9433
+    },
+    {
+      "epoch": 1.6056111623121931,
+      "grad_norm": 1.609375,
+      "learning_rate": 8.853953365005503e-06,
+      "loss": 0.7898,
+      "step": 9434
+    },
+    {
+      "epoch": 1.605782626401183,
+      "grad_norm": 1.7890625,
+      "learning_rate": 8.852159354465486e-06,
+      "loss": 0.8764,
+      "step": 9435
+    },
+    {
+      "epoch": 1.605954090490173,
+      "grad_norm": 1.6953125,
+      "learning_rate": 8.850365381360879e-06,
+      "loss": 0.8465,
+      "step": 9436
+    },
+    {
+      "epoch": 1.6061255545791628,
+      "grad_norm": 1.7109375,
+      "learning_rate": 8.848571445750183e-06,
+      "loss": 0.8515,
+      "step": 9437
+    },
+    {
+      "epoch": 1.6062970186681527,
+      "grad_norm": 1.6640625,
+      "learning_rate": 8.84677754769191e-06,
+      "loss": 0.7793,
+      "step": 9438
+    },
+    {
+      "epoch": 1.6064684827571427,
+      "grad_norm": 1.7890625,
+      "learning_rate": 8.844983687244565e-06,
+      "loss": 0.8679,
+      "step": 9439
+    },
+    {
+      "epoch": 1.6066399468461325,
+      "grad_norm": 1.71875,
+      "learning_rate": 8.84318986446665e-06,
+      "loss": 0.8877,
+      "step": 9440
+    },
+    {
+      "epoch": 1.6068114109351224,
+      "grad_norm": 1.890625,
+      "learning_rate": 8.84139607941667e-06,
+      "loss": 0.9196,
+      "step": 9441
+    },
+    {
+      "epoch": 1.6069828750241122,
+      "grad_norm": 1.640625,
+      "learning_rate": 8.839602332153126e-06,
+      "loss": 0.7696,
+      "step": 9442
+    },
+    {
+      "epoch": 1.607154339113102,
+      "grad_norm": 1.6328125,
+      "learning_rate": 8.837808622734519e-06,
+      "loss": 0.8631,
+      "step": 9443
+    },
+    {
+      "epoch": 1.6073258032020918,
+      "grad_norm": 1.6953125,
+      "learning_rate": 8.83601495121935e-06,
+      "loss": 0.786,
+      "step": 9444
+    },
+    {
+      "epoch": 1.6074972672910817,
+      "grad_norm": 1.640625,
+      "learning_rate": 8.834221317666117e-06,
+      "loss": 0.8097,
+      "step": 9445
+    },
+    {
+      "epoch": 1.6076687313800715,
+      "grad_norm": 1.7265625,
+      "learning_rate": 8.832427722133315e-06,
+      "loss": 0.8723,
+      "step": 9446
+    },
+    {
+      "epoch": 1.6078401954690613,
+      "grad_norm": 1.796875,
+      "learning_rate": 8.830634164679444e-06,
+      "loss": 0.8215,
+      "step": 9447
+    },
+    {
+      "epoch": 1.6080116595580514,
+      "grad_norm": 1.6640625,
+      "learning_rate": 8.828840645362995e-06,
+      "loss": 0.8247,
+      "step": 9448
+    },
+    {
+      "epoch": 1.6081831236470412,
+      "grad_norm": 1.703125,
+      "learning_rate": 8.82704716424246e-06,
+      "loss": 0.8534,
+      "step": 9449
+    },
+    {
+      "epoch": 1.608354587736031,
+      "grad_norm": 1.734375,
+      "learning_rate": 8.825253721376341e-06,
+      "loss": 0.8736,
+      "step": 9450
+    },
+    {
+      "epoch": 1.6085260518250208,
+      "grad_norm": 1.7578125,
+      "learning_rate": 8.823460316823118e-06,
+      "loss": 0.9481,
+      "step": 9451
+    },
+    {
+      "epoch": 1.608697515914011,
+      "grad_norm": 1.6328125,
+      "learning_rate": 8.821666950641283e-06,
+      "loss": 0.8318,
+      "step": 9452
+    },
+    {
+      "epoch": 1.6088689800030007,
+      "grad_norm": 1.671875,
+      "learning_rate": 8.819873622889328e-06,
+      "loss": 0.9219,
+      "step": 9453
+    },
+    {
+      "epoch": 1.6090404440919905,
+      "grad_norm": 1.625,
+      "learning_rate": 8.818080333625737e-06,
+      "loss": 0.8387,
+      "step": 9454
+    },
+    {
+      "epoch": 1.6092119081809804,
+      "grad_norm": 1.5625,
+      "learning_rate": 8.816287082908998e-06,
+      "loss": 0.7747,
+      "step": 9455
+    },
+    {
+      "epoch": 1.6093833722699702,
+      "grad_norm": 1.65625,
+      "learning_rate": 8.814493870797594e-06,
+      "loss": 0.8218,
+      "step": 9456
+    },
+    {
+      "epoch": 1.60955483635896,
+      "grad_norm": 1.6171875,
+      "learning_rate": 8.812700697350006e-06,
+      "loss": 0.8,
+      "step": 9457
+    },
+    {
+      "epoch": 1.6097263004479498,
+      "grad_norm": 1.78125,
+      "learning_rate": 8.810907562624721e-06,
+      "loss": 0.8254,
+      "step": 9458
+    },
+    {
+      "epoch": 1.6098977645369397,
+      "grad_norm": 1.6875,
+      "learning_rate": 8.809114466680219e-06,
+      "loss": 0.8618,
+      "step": 9459
+    },
+    {
+      "epoch": 1.6100692286259295,
+      "grad_norm": 1.6875,
+      "learning_rate": 8.80732140957498e-06,
+      "loss": 0.8895,
+      "step": 9460
+    },
+    {
+      "epoch": 1.6102406927149195,
+      "grad_norm": 1.6796875,
+      "learning_rate": 8.80552839136748e-06,
+      "loss": 0.8411,
+      "step": 9461
+    },
+    {
+      "epoch": 1.6104121568039094,
+      "grad_norm": 1.6953125,
+      "learning_rate": 8.8037354121162e-06,
+      "loss": 0.8813,
+      "step": 9462
+    },
+    {
+      "epoch": 1.6105836208928992,
+      "grad_norm": 1.6484375,
+      "learning_rate": 8.801942471879611e-06,
+      "loss": 0.7896,
+      "step": 9463
+    },
+    {
+      "epoch": 1.6107550849818892,
+      "grad_norm": 1.8203125,
+      "learning_rate": 8.800149570716188e-06,
+      "loss": 0.8829,
+      "step": 9464
+    },
+    {
+      "epoch": 1.610926549070879,
+      "grad_norm": 1.640625,
+      "learning_rate": 8.798356708684405e-06,
+      "loss": 0.8807,
+      "step": 9465
+    },
+    {
+      "epoch": 1.611098013159869,
+      "grad_norm": 1.7578125,
+      "learning_rate": 8.796563885842737e-06,
+      "loss": 0.831,
+      "step": 9466
+    },
+    {
+      "epoch": 1.6112694772488587,
+      "grad_norm": 1.7734375,
+      "learning_rate": 8.79477110224965e-06,
+      "loss": 0.931,
+      "step": 9467
+    },
+    {
+      "epoch": 1.6114409413378485,
+      "grad_norm": 1.734375,
+      "learning_rate": 8.792978357963618e-06,
+      "loss": 0.8006,
+      "step": 9468
+    },
+    {
+      "epoch": 1.6116124054268384,
+      "grad_norm": 1.71875,
+      "learning_rate": 8.791185653043106e-06,
+      "loss": 0.8351,
+      "step": 9469
+    },
+    {
+      "epoch": 1.6117838695158282,
+      "grad_norm": 1.7734375,
+      "learning_rate": 8.789392987546581e-06,
+      "loss": 0.8104,
+      "step": 9470
+    },
+    {
+      "epoch": 1.611955333604818,
+      "grad_norm": 1.6796875,
+      "learning_rate": 8.78760036153251e-06,
+      "loss": 0.8881,
+      "step": 9471
+    },
+    {
+      "epoch": 1.6121267976938078,
+      "grad_norm": 1.6484375,
+      "learning_rate": 8.785807775059357e-06,
+      "loss": 0.8613,
+      "step": 9472
+    },
+    {
+      "epoch": 1.612298261782798,
+      "grad_norm": 1.6328125,
+      "learning_rate": 8.784015228185587e-06,
+      "loss": 0.8537,
+      "step": 9473
+    },
+    {
+      "epoch": 1.6124697258717877,
+      "grad_norm": 1.7578125,
+      "learning_rate": 8.782222720969658e-06,
+      "loss": 0.8582,
+      "step": 9474
+    },
+    {
+      "epoch": 1.6126411899607775,
+      "grad_norm": 1.65625,
+      "learning_rate": 8.78043025347003e-06,
+      "loss": 0.8746,
+      "step": 9475
+    },
+    {
+      "epoch": 1.6128126540497676,
+      "grad_norm": 1.703125,
+      "learning_rate": 8.778637825745165e-06,
+      "loss": 0.8207,
+      "step": 9476
+    },
+    {
+      "epoch": 1.6129841181387574,
+      "grad_norm": 1.6875,
+      "learning_rate": 8.77684543785352e-06,
+      "loss": 0.8755,
+      "step": 9477
+    },
+    {
+      "epoch": 1.6131555822277472,
+      "grad_norm": 1.609375,
+      "learning_rate": 8.77505308985355e-06,
+      "loss": 0.7851,
+      "step": 9478
+    },
+    {
+      "epoch": 1.613327046316737,
+      "grad_norm": 1.6171875,
+      "learning_rate": 8.773260781803713e-06,
+      "loss": 0.7995,
+      "step": 9479
+    },
+    {
+      "epoch": 1.613498510405727,
+      "grad_norm": 1.6796875,
+      "learning_rate": 8.77146851376246e-06,
+      "loss": 0.8759,
+      "step": 9480
+    },
+    {
+      "epoch": 1.6136699744947167,
+      "grad_norm": 1.6953125,
+      "learning_rate": 8.769676285788245e-06,
+      "loss": 0.8722,
+      "step": 9481
+    },
+    {
+      "epoch": 1.6138414385837065,
+      "grad_norm": 1.6953125,
+      "learning_rate": 8.76788409793952e-06,
+      "loss": 0.8435,
+      "step": 9482
+    },
+    {
+      "epoch": 1.6140129026726964,
+      "grad_norm": 1.6953125,
+      "learning_rate": 8.766091950274735e-06,
+      "loss": 0.8915,
+      "step": 9483
+    },
+    {
+      "epoch": 1.6141843667616862,
+      "grad_norm": 1.65625,
+      "learning_rate": 8.764299842852336e-06,
+      "loss": 0.839,
+      "step": 9484
+    },
+    {
+      "epoch": 1.6143558308506762,
+      "grad_norm": 1.65625,
+      "learning_rate": 8.762507775730776e-06,
+      "loss": 0.8317,
+      "step": 9485
+    },
+    {
+      "epoch": 1.614527294939666,
+      "grad_norm": 1.7421875,
+      "learning_rate": 8.760715748968494e-06,
+      "loss": 0.8901,
+      "step": 9486
+    },
+    {
+      "epoch": 1.614698759028656,
+      "grad_norm": 1.7734375,
+      "learning_rate": 8.75892376262394e-06,
+      "loss": 0.8669,
+      "step": 9487
+    },
+    {
+      "epoch": 1.614870223117646,
+      "grad_norm": 1.625,
+      "learning_rate": 8.757131816755554e-06,
+      "loss": 0.7872,
+      "step": 9488
+    },
+    {
+      "epoch": 1.6150416872066358,
+      "grad_norm": 1.6640625,
+      "learning_rate": 8.75533991142178e-06,
+      "loss": 0.7714,
+      "step": 9489
+    },
+    {
+      "epoch": 1.6152131512956256,
+      "grad_norm": 1.7109375,
+      "learning_rate": 8.75354804668106e-06,
+      "loss": 0.8928,
+      "step": 9490
+    },
+    {
+      "epoch": 1.6153846153846154,
+      "grad_norm": 1.609375,
+      "learning_rate": 8.75175622259183e-06,
+      "loss": 0.8651,
+      "step": 9491
+    },
+    {
+      "epoch": 1.6155560794736052,
+      "grad_norm": 1.5625,
+      "learning_rate": 8.74996443921253e-06,
+      "loss": 0.7452,
+      "step": 9492
+    },
+    {
+      "epoch": 1.615727543562595,
+      "grad_norm": 1.6171875,
+      "learning_rate": 8.748172696601597e-06,
+      "loss": 0.6888,
+      "step": 9493
+    },
+    {
+      "epoch": 1.615899007651585,
+      "grad_norm": 1.671875,
+      "learning_rate": 8.746380994817467e-06,
+      "loss": 0.7881,
+      "step": 9494
+    },
+    {
+      "epoch": 1.6160704717405747,
+      "grad_norm": 1.65625,
+      "learning_rate": 8.744589333918571e-06,
+      "loss": 0.8082,
+      "step": 9495
+    },
+    {
+      "epoch": 1.6162419358295645,
+      "grad_norm": 1.75,
+      "learning_rate": 8.742797713963352e-06,
+      "loss": 0.8459,
+      "step": 9496
+    },
+    {
+      "epoch": 1.6164133999185546,
+      "grad_norm": 1.7421875,
+      "learning_rate": 8.741006135010228e-06,
+      "loss": 0.8487,
+      "step": 9497
+    },
+    {
+      "epoch": 1.6165848640075444,
+      "grad_norm": 1.859375,
+      "learning_rate": 8.739214597117634e-06,
+      "loss": 0.8437,
+      "step": 9498
+    },
+    {
+      "epoch": 1.6167563280965342,
+      "grad_norm": 1.7578125,
+      "learning_rate": 8.737423100344002e-06,
+      "loss": 0.8587,
+      "step": 9499
+    },
+    {
+      "epoch": 1.6169277921855243,
+      "grad_norm": 1.7421875,
+      "learning_rate": 8.735631644747755e-06,
+      "loss": 0.8818,
+      "step": 9500
+    },
+    {
+      "epoch": 1.6170992562745141,
+      "grad_norm": 1.6484375,
+      "learning_rate": 8.733840230387322e-06,
+      "loss": 0.8194,
+      "step": 9501
+    },
+    {
+      "epoch": 1.617270720363504,
+      "grad_norm": 1.6328125,
+      "learning_rate": 8.732048857321126e-06,
+      "loss": 0.8206,
+      "step": 9502
+    },
+    {
+      "epoch": 1.6174421844524938,
+      "grad_norm": 1.6953125,
+      "learning_rate": 8.730257525607594e-06,
+      "loss": 0.8871,
+      "step": 9503
+    },
+    {
+      "epoch": 1.6176136485414836,
+      "grad_norm": 1.6796875,
+      "learning_rate": 8.728466235305143e-06,
+      "loss": 0.8642,
+      "step": 9504
+    },
+    {
+      "epoch": 1.6177851126304734,
+      "grad_norm": 1.7265625,
+      "learning_rate": 8.726674986472195e-06,
+      "loss": 0.8576,
+      "step": 9505
+    },
+    {
+      "epoch": 1.6179565767194632,
+      "grad_norm": 1.671875,
+      "learning_rate": 8.724883779167174e-06,
+      "loss": 0.8776,
+      "step": 9506
+    },
+    {
+      "epoch": 1.618128040808453,
+      "grad_norm": 1.640625,
+      "learning_rate": 8.723092613448494e-06,
+      "loss": 0.8235,
+      "step": 9507
+    },
+    {
+      "epoch": 1.618299504897443,
+      "grad_norm": 1.703125,
+      "learning_rate": 8.721301489374575e-06,
+      "loss": 0.8939,
+      "step": 9508
+    },
+    {
+      "epoch": 1.618470968986433,
+      "grad_norm": 1.703125,
+      "learning_rate": 8.719510407003826e-06,
+      "loss": 0.8792,
+      "step": 9509
+    },
+    {
+      "epoch": 1.6186424330754228,
+      "grad_norm": 1.671875,
+      "learning_rate": 8.717719366394665e-06,
+      "loss": 0.7704,
+      "step": 9510
+    },
+    {
+      "epoch": 1.6188138971644126,
+      "grad_norm": 1.7265625,
+      "learning_rate": 8.715928367605503e-06,
+      "loss": 0.8084,
+      "step": 9511
+    },
+    {
+      "epoch": 1.6189853612534026,
+      "grad_norm": 1.6796875,
+      "learning_rate": 8.714137410694753e-06,
+      "loss": 0.848,
+      "step": 9512
+    },
+    {
+      "epoch": 1.6191568253423925,
+      "grad_norm": 1.671875,
+      "learning_rate": 8.712346495720822e-06,
+      "loss": 0.8198,
+      "step": 9513
+    },
+    {
+      "epoch": 1.6193282894313823,
+      "grad_norm": 1.6171875,
+      "learning_rate": 8.710555622742124e-06,
+      "loss": 0.7781,
+      "step": 9514
+    },
+    {
+      "epoch": 1.6194997535203721,
+      "grad_norm": 1.75,
+      "learning_rate": 8.708764791817062e-06,
+      "loss": 0.9383,
+      "step": 9515
+    },
+    {
+      "epoch": 1.619671217609362,
+      "grad_norm": 1.7109375,
+      "learning_rate": 8.706974003004042e-06,
+      "loss": 0.8586,
+      "step": 9516
+    },
+    {
+      "epoch": 1.6198426816983518,
+      "grad_norm": 1.6484375,
+      "learning_rate": 8.705183256361468e-06,
+      "loss": 0.7714,
+      "step": 9517
+    },
+    {
+      "epoch": 1.6200141457873416,
+      "grad_norm": 1.609375,
+      "learning_rate": 8.703392551947744e-06,
+      "loss": 0.9317,
+      "step": 9518
+    },
+    {
+      "epoch": 1.6201856098763314,
+      "grad_norm": 1.6953125,
+      "learning_rate": 8.701601889821278e-06,
+      "loss": 0.8142,
+      "step": 9519
+    },
+    {
+      "epoch": 1.6203570739653212,
+      "grad_norm": 1.65625,
+      "learning_rate": 8.699811270040456e-06,
+      "loss": 0.8342,
+      "step": 9520
+    },
+    {
+      "epoch": 1.6205285380543113,
+      "grad_norm": 1.78125,
+      "learning_rate": 8.698020692663686e-06,
+      "loss": 0.9059,
+      "step": 9521
+    },
+    {
+      "epoch": 1.6207000021433011,
+      "grad_norm": 1.7421875,
+      "learning_rate": 8.696230157749365e-06,
+      "loss": 0.8225,
+      "step": 9522
+    },
+    {
+      "epoch": 1.620871466232291,
+      "grad_norm": 1.75,
+      "learning_rate": 8.694439665355887e-06,
+      "loss": 0.8743,
+      "step": 9523
+    },
+    {
+      "epoch": 1.621042930321281,
+      "grad_norm": 4.96875,
+      "learning_rate": 8.692649215541648e-06,
+      "loss": 0.8313,
+      "step": 9524
+    },
+    {
+      "epoch": 1.6212143944102708,
+      "grad_norm": 1.7578125,
+      "learning_rate": 8.690858808365041e-06,
+      "loss": 0.8802,
+      "step": 9525
+    },
+    {
+      "epoch": 1.6213858584992606,
+      "grad_norm": 1.6953125,
+      "learning_rate": 8.68906844388446e-06,
+      "loss": 0.7753,
+      "step": 9526
+    },
+    {
+      "epoch": 1.6215573225882505,
+      "grad_norm": 1.7734375,
+      "learning_rate": 8.68727812215829e-06,
+      "loss": 0.7983,
+      "step": 9527
+    },
+    {
+      "epoch": 1.6217287866772403,
+      "grad_norm": 1.8125,
+      "learning_rate": 8.685487843244927e-06,
+      "loss": 0.964,
+      "step": 9528
+    },
+    {
+      "epoch": 1.6219002507662301,
+      "grad_norm": 1.71875,
+      "learning_rate": 8.683697607202754e-06,
+      "loss": 0.7903,
+      "step": 9529
+    },
+    {
+      "epoch": 1.62207171485522,
+      "grad_norm": 1.7109375,
+      "learning_rate": 8.681907414090159e-06,
+      "loss": 0.8497,
+      "step": 9530
+    },
+    {
+      "epoch": 1.6222431789442098,
+      "grad_norm": 1.6015625,
+      "learning_rate": 8.68011726396553e-06,
+      "loss": 0.8279,
+      "step": 9531
+    },
+    {
+      "epoch": 1.6224146430331996,
+      "grad_norm": 1.796875,
+      "learning_rate": 8.678327156887243e-06,
+      "loss": 0.8786,
+      "step": 9532
+    },
+    {
+      "epoch": 1.6225861071221896,
+      "grad_norm": 1.78125,
+      "learning_rate": 8.676537092913685e-06,
+      "loss": 0.9326,
+      "step": 9533
+    },
+    {
+      "epoch": 1.6227575712111795,
+      "grad_norm": 1.875,
+      "learning_rate": 8.674747072103236e-06,
+      "loss": 0.7747,
+      "step": 9534
+    },
+    {
+      "epoch": 1.6229290353001693,
+      "grad_norm": 1.6953125,
+      "learning_rate": 8.672957094514278e-06,
+      "loss": 0.9462,
+      "step": 9535
+    },
+    {
+      "epoch": 1.6231004993891593,
+      "grad_norm": 1.703125,
+      "learning_rate": 8.671167160205183e-06,
+      "loss": 0.7966,
+      "step": 9536
+    },
+    {
+      "epoch": 1.6232719634781492,
+      "grad_norm": 1.6015625,
+      "learning_rate": 8.66937726923433e-06,
+      "loss": 0.8396,
+      "step": 9537
+    },
+    {
+      "epoch": 1.623443427567139,
+      "grad_norm": 1.8125,
+      "learning_rate": 8.667587421660099e-06,
+      "loss": 0.8396,
+      "step": 9538
+    },
+    {
+      "epoch": 1.6236148916561288,
+      "grad_norm": 1.6875,
+      "learning_rate": 8.665797617540857e-06,
+      "loss": 0.8863,
+      "step": 9539
+    },
+    {
+      "epoch": 1.6237863557451186,
+      "grad_norm": 1.5625,
+      "learning_rate": 8.664007856934979e-06,
+      "loss": 0.7412,
+      "step": 9540
+    },
+    {
+      "epoch": 1.6239578198341085,
+      "grad_norm": 1.71875,
+      "learning_rate": 8.662218139900836e-06,
+      "loss": 0.8568,
+      "step": 9541
+    },
+    {
+      "epoch": 1.6241292839230983,
+      "grad_norm": 1.59375,
+      "learning_rate": 8.660428466496795e-06,
+      "loss": 0.8518,
+      "step": 9542
+    },
+    {
+      "epoch": 1.6243007480120881,
+      "grad_norm": 1.7421875,
+      "learning_rate": 8.658638836781232e-06,
+      "loss": 0.9646,
+      "step": 9543
+    },
+    {
+      "epoch": 1.624472212101078,
+      "grad_norm": 1.71875,
+      "learning_rate": 8.656849250812504e-06,
+      "loss": 0.8214,
+      "step": 9544
+    },
+    {
+      "epoch": 1.624643676190068,
+      "grad_norm": 1.6953125,
+      "learning_rate": 8.65505970864898e-06,
+      "loss": 0.8817,
+      "step": 9545
+    },
+    {
+      "epoch": 1.6248151402790578,
+      "grad_norm": 1.625,
+      "learning_rate": 8.653270210349023e-06,
+      "loss": 0.7976,
+      "step": 9546
+    },
+    {
+      "epoch": 1.6249866043680476,
+      "grad_norm": 1.703125,
+      "learning_rate": 8.651480755970995e-06,
+      "loss": 0.892,
+      "step": 9547
+    },
+    {
+      "epoch": 1.6251580684570375,
+      "grad_norm": 1.6875,
+      "learning_rate": 8.649691345573259e-06,
+      "loss": 0.8196,
+      "step": 9548
+    },
+    {
+      "epoch": 1.6253295325460275,
+      "grad_norm": 1.765625,
+      "learning_rate": 8.647901979214173e-06,
+      "loss": 0.9425,
+      "step": 9549
+    },
+    {
+      "epoch": 1.6255009966350173,
+      "grad_norm": 1.7265625,
+      "learning_rate": 8.646112656952094e-06,
+      "loss": 0.8447,
+      "step": 9550
+    },
+    {
+      "epoch": 1.6256724607240072,
+      "grad_norm": 1.6328125,
+      "learning_rate": 8.644323378845378e-06,
+      "loss": 0.8281,
+      "step": 9551
+    },
+    {
+      "epoch": 1.625843924812997,
+      "grad_norm": 1.7265625,
+      "learning_rate": 8.642534144952383e-06,
+      "loss": 0.9119,
+      "step": 9552
+    },
+    {
+      "epoch": 1.6260153889019868,
+      "grad_norm": 1.7421875,
+      "learning_rate": 8.640744955331462e-06,
+      "loss": 0.8358,
+      "step": 9553
+    },
+    {
+      "epoch": 1.6261868529909767,
+      "grad_norm": 1.609375,
+      "learning_rate": 8.63895581004097e-06,
+      "loss": 0.8309,
+      "step": 9554
+    },
+    {
+      "epoch": 1.6263583170799665,
+      "grad_norm": 1.6640625,
+      "learning_rate": 8.637166709139251e-06,
+      "loss": 0.8626,
+      "step": 9555
+    },
+    {
+      "epoch": 1.6265297811689563,
+      "grad_norm": 1.6484375,
+      "learning_rate": 8.635377652684657e-06,
+      "loss": 0.7618,
+      "step": 9556
+    },
+    {
+      "epoch": 1.6267012452579461,
+      "grad_norm": 1.6796875,
+      "learning_rate": 8.633588640735536e-06,
+      "loss": 0.8348,
+      "step": 9557
+    },
+    {
+      "epoch": 1.6268727093469362,
+      "grad_norm": 1.6796875,
+      "learning_rate": 8.631799673350235e-06,
+      "loss": 0.9185,
+      "step": 9558
+    },
+    {
+      "epoch": 1.627044173435926,
+      "grad_norm": 1.7578125,
+      "learning_rate": 8.630010750587099e-06,
+      "loss": 0.9125,
+      "step": 9559
+    },
+    {
+      "epoch": 1.6272156375249158,
+      "grad_norm": 1.6875,
+      "learning_rate": 8.62822187250447e-06,
+      "loss": 0.9099,
+      "step": 9560
+    },
+    {
+      "epoch": 1.6273871016139059,
+      "grad_norm": 1.6171875,
+      "learning_rate": 8.626433039160693e-06,
+      "loss": 0.7726,
+      "step": 9561
+    },
+    {
+      "epoch": 1.6275585657028957,
+      "grad_norm": 1.9296875,
+      "learning_rate": 8.624644250614107e-06,
+      "loss": 0.863,
+      "step": 9562
+    },
+    {
+      "epoch": 1.6277300297918855,
+      "grad_norm": 1.703125,
+      "learning_rate": 8.622855506923051e-06,
+      "loss": 0.9398,
+      "step": 9563
+    },
+    {
+      "epoch": 1.6279014938808753,
+      "grad_norm": 1.7734375,
+      "learning_rate": 8.621066808145863e-06,
+      "loss": 0.8906,
+      "step": 9564
+    },
+    {
+      "epoch": 1.6280729579698652,
+      "grad_norm": 1.6640625,
+      "learning_rate": 8.619278154340877e-06,
+      "loss": 0.8521,
+      "step": 9565
+    },
+    {
+      "epoch": 1.628244422058855,
+      "grad_norm": 1.6796875,
+      "learning_rate": 8.617489545566437e-06,
+      "loss": 0.9032,
+      "step": 9566
+    },
+    {
+      "epoch": 1.6284158861478448,
+      "grad_norm": 1.671875,
+      "learning_rate": 8.615700981880861e-06,
+      "loss": 0.7931,
+      "step": 9567
+    },
+    {
+      "epoch": 1.6285873502368347,
+      "grad_norm": 1.6875,
+      "learning_rate": 8.61391246334249e-06,
+      "loss": 0.8021,
+      "step": 9568
+    },
+    {
+      "epoch": 1.6287588143258245,
+      "grad_norm": 1.734375,
+      "learning_rate": 8.612123990009655e-06,
+      "loss": 0.8412,
+      "step": 9569
+    },
+    {
+      "epoch": 1.6289302784148145,
+      "grad_norm": 1.6953125,
+      "learning_rate": 8.610335561940682e-06,
+      "loss": 0.9334,
+      "step": 9570
+    },
+    {
+      "epoch": 1.6291017425038044,
+      "grad_norm": 1.703125,
+      "learning_rate": 8.6085471791939e-06,
+      "loss": 0.8338,
+      "step": 9571
+    },
+    {
+      "epoch": 1.6292732065927942,
+      "grad_norm": 1.734375,
+      "learning_rate": 8.606758841827634e-06,
+      "loss": 0.8811,
+      "step": 9572
+    },
+    {
+      "epoch": 1.6294446706817842,
+      "grad_norm": 1.6796875,
+      "learning_rate": 8.604970549900208e-06,
+      "loss": 0.7772,
+      "step": 9573
+    },
+    {
+      "epoch": 1.629616134770774,
+      "grad_norm": 1.53125,
+      "learning_rate": 8.603182303469947e-06,
+      "loss": 0.7959,
+      "step": 9574
+    },
+    {
+      "epoch": 1.6297875988597639,
+      "grad_norm": 1.6328125,
+      "learning_rate": 8.601394102595169e-06,
+      "loss": 0.8504,
+      "step": 9575
+    },
+    {
+      "epoch": 1.6299590629487537,
+      "grad_norm": 1.640625,
+      "learning_rate": 8.599605947334196e-06,
+      "loss": 0.8329,
+      "step": 9576
+    },
+    {
+      "epoch": 1.6301305270377435,
+      "grad_norm": 1.578125,
+      "learning_rate": 8.59781783774535e-06,
+      "loss": 0.7754,
+      "step": 9577
+    },
+    {
+      "epoch": 1.6303019911267334,
+      "grad_norm": 1.71875,
+      "learning_rate": 8.59602977388694e-06,
+      "loss": 0.8126,
+      "step": 9578
+    },
+    {
+      "epoch": 1.6304734552157232,
+      "grad_norm": 1.6328125,
+      "learning_rate": 8.594241755817289e-06,
+      "loss": 0.7677,
+      "step": 9579
+    },
+    {
+      "epoch": 1.630644919304713,
+      "grad_norm": 1.5703125,
+      "learning_rate": 8.592453783594707e-06,
+      "loss": 0.8123,
+      "step": 9580
+    },
+    {
+      "epoch": 1.6308163833937028,
+      "grad_norm": 1.7421875,
+      "learning_rate": 8.590665857277507e-06,
+      "loss": 0.8539,
+      "step": 9581
+    },
+    {
+      "epoch": 1.6309878474826929,
+      "grad_norm": 1.65625,
+      "learning_rate": 8.588877976924e-06,
+      "loss": 0.7789,
+      "step": 9582
+    },
+    {
+      "epoch": 1.6311593115716827,
+      "grad_norm": 1.71875,
+      "learning_rate": 8.587090142592497e-06,
+      "loss": 0.808,
+      "step": 9583
+    },
+    {
+      "epoch": 1.6313307756606725,
+      "grad_norm": 1.8125,
+      "learning_rate": 8.585302354341305e-06,
+      "loss": 0.8538,
+      "step": 9584
+    },
+    {
+      "epoch": 1.6315022397496626,
+      "grad_norm": 1.6171875,
+      "learning_rate": 8.58351461222873e-06,
+      "loss": 0.7721,
+      "step": 9585
+    },
+    {
+      "epoch": 1.6316737038386524,
+      "grad_norm": 1.7109375,
+      "learning_rate": 8.581726916313077e-06,
+      "loss": 0.882,
+      "step": 9586
+    },
+    {
+      "epoch": 1.6318451679276422,
+      "grad_norm": 1.6640625,
+      "learning_rate": 8.57993926665265e-06,
+      "loss": 0.855,
+      "step": 9587
+    },
+    {
+      "epoch": 1.632016632016632,
+      "grad_norm": 1.7265625,
+      "learning_rate": 8.578151663305751e-06,
+      "loss": 0.8971,
+      "step": 9588
+    },
+    {
+      "epoch": 1.6321880961056219,
+      "grad_norm": 1.6796875,
+      "learning_rate": 8.576364106330683e-06,
+      "loss": 0.8524,
+      "step": 9589
+    },
+    {
+      "epoch": 1.6323595601946117,
+      "grad_norm": 1.75,
+      "learning_rate": 8.574576595785742e-06,
+      "loss": 0.8495,
+      "step": 9590
+    },
+    {
+      "epoch": 1.6325310242836015,
+      "grad_norm": 1.7109375,
+      "learning_rate": 8.572789131729224e-06,
+      "loss": 0.8608,
+      "step": 9591
+    },
+    {
+      "epoch": 1.6327024883725914,
+      "grad_norm": 1.6875,
+      "learning_rate": 8.571001714219425e-06,
+      "loss": 0.8293,
+      "step": 9592
+    },
+    {
+      "epoch": 1.6328739524615812,
+      "grad_norm": 1.6796875,
+      "learning_rate": 8.569214343314645e-06,
+      "loss": 0.8418,
+      "step": 9593
+    },
+    {
+      "epoch": 1.6330454165505712,
+      "grad_norm": 1.609375,
+      "learning_rate": 8.56742701907317e-06,
+      "loss": 0.8288,
+      "step": 9594
+    },
+    {
+      "epoch": 1.633216880639561,
+      "grad_norm": 1.7265625,
+      "learning_rate": 8.565639741553296e-06,
+      "loss": 0.8054,
+      "step": 9595
+    },
+    {
+      "epoch": 1.6333883447285509,
+      "grad_norm": 1.6640625,
+      "learning_rate": 8.56385251081331e-06,
+      "loss": 0.8483,
+      "step": 9596
+    },
+    {
+      "epoch": 1.633559808817541,
+      "grad_norm": 1.65625,
+      "learning_rate": 8.562065326911502e-06,
+      "loss": 0.856,
+      "step": 9597
+    },
+    {
+      "epoch": 1.6337312729065308,
+      "grad_norm": 1.6328125,
+      "learning_rate": 8.560278189906159e-06,
+      "loss": 0.7193,
+      "step": 9598
+    },
+    {
+      "epoch": 1.6339027369955206,
+      "grad_norm": 1.7109375,
+      "learning_rate": 8.558491099855565e-06,
+      "loss": 0.8671,
+      "step": 9599
+    },
+    {
+      "epoch": 1.6340742010845104,
+      "grad_norm": 1.65625,
+      "learning_rate": 8.556704056818011e-06,
+      "loss": 0.8556,
+      "step": 9600
+    },
+    {
+      "epoch": 1.6342456651735002,
+      "grad_norm": 1.7265625,
+      "learning_rate": 8.554917060851767e-06,
+      "loss": 0.8631,
+      "step": 9601
+    },
+    {
+      "epoch": 1.63441712926249,
+      "grad_norm": 1.75,
+      "learning_rate": 8.553130112015118e-06,
+      "loss": 0.8055,
+      "step": 9602
+    },
+    {
+      "epoch": 1.6345885933514799,
+      "grad_norm": 1.6875,
+      "learning_rate": 8.551343210366346e-06,
+      "loss": 0.8503,
+      "step": 9603
+    },
+    {
+      "epoch": 1.6347600574404697,
+      "grad_norm": 1.734375,
+      "learning_rate": 8.549556355963726e-06,
+      "loss": 0.903,
+      "step": 9604
+    },
+    {
+      "epoch": 1.6349315215294595,
+      "grad_norm": 1.6796875,
+      "learning_rate": 8.547769548865537e-06,
+      "loss": 0.8312,
+      "step": 9605
+    },
+    {
+      "epoch": 1.6351029856184496,
+      "grad_norm": 1.7265625,
+      "learning_rate": 8.545982789130048e-06,
+      "loss": 0.8706,
+      "step": 9606
+    },
+    {
+      "epoch": 1.6352744497074394,
+      "grad_norm": 1.71875,
+      "learning_rate": 8.544196076815539e-06,
+      "loss": 0.905,
+      "step": 9607
+    },
+    {
+      "epoch": 1.6354459137964292,
+      "grad_norm": 1.5546875,
+      "learning_rate": 8.542409411980276e-06,
+      "loss": 0.7822,
+      "step": 9608
+    },
+    {
+      "epoch": 1.6356173778854193,
+      "grad_norm": 1.6953125,
+      "learning_rate": 8.540622794682531e-06,
+      "loss": 0.9522,
+      "step": 9609
+    },
+    {
+      "epoch": 1.635788841974409,
+      "grad_norm": 1.609375,
+      "learning_rate": 8.538836224980574e-06,
+      "loss": 0.8145,
+      "step": 9610
+    },
+    {
+      "epoch": 1.635960306063399,
+      "grad_norm": 1.7734375,
+      "learning_rate": 8.537049702932669e-06,
+      "loss": 0.9594,
+      "step": 9611
+    },
+    {
+      "epoch": 1.6361317701523888,
+      "grad_norm": 1.6640625,
+      "learning_rate": 8.535263228597086e-06,
+      "loss": 0.834,
+      "step": 9612
+    },
+    {
+      "epoch": 1.6363032342413786,
+      "grad_norm": 1.8046875,
+      "learning_rate": 8.53347680203208e-06,
+      "loss": 0.9463,
+      "step": 9613
+    },
+    {
+      "epoch": 1.6364746983303684,
+      "grad_norm": 1.6328125,
+      "learning_rate": 8.531690423295917e-06,
+      "loss": 0.8724,
+      "step": 9614
+    },
+    {
+      "epoch": 1.6366461624193582,
+      "grad_norm": 1.7265625,
+      "learning_rate": 8.529904092446862e-06,
+      "loss": 0.8486,
+      "step": 9615
+    },
+    {
+      "epoch": 1.636817626508348,
+      "grad_norm": 1.71875,
+      "learning_rate": 8.528117809543168e-06,
+      "loss": 0.8814,
+      "step": 9616
+    },
+    {
+      "epoch": 1.6369890905973379,
+      "grad_norm": 1.5859375,
+      "learning_rate": 8.526331574643096e-06,
+      "loss": 0.9166,
+      "step": 9617
+    },
+    {
+      "epoch": 1.637160554686328,
+      "grad_norm": 1.71875,
+      "learning_rate": 8.5245453878049e-06,
+      "loss": 0.8858,
+      "step": 9618
+    },
+    {
+      "epoch": 1.6373320187753178,
+      "grad_norm": 1.5859375,
+      "learning_rate": 8.522759249086835e-06,
+      "loss": 0.7996,
+      "step": 9619
+    },
+    {
+      "epoch": 1.6375034828643076,
+      "grad_norm": 1.625,
+      "learning_rate": 8.520973158547154e-06,
+      "loss": 0.8665,
+      "step": 9620
+    },
+    {
+      "epoch": 1.6376749469532976,
+      "grad_norm": 1.8671875,
+      "learning_rate": 8.519187116244107e-06,
+      "loss": 0.91,
+      "step": 9621
+    },
+    {
+      "epoch": 1.6378464110422875,
+      "grad_norm": 1.703125,
+      "learning_rate": 8.517401122235945e-06,
+      "loss": 0.9372,
+      "step": 9622
+    },
+    {
+      "epoch": 1.6380178751312773,
+      "grad_norm": 1.7421875,
+      "learning_rate": 8.515615176580917e-06,
+      "loss": 0.9159,
+      "step": 9623
+    },
+    {
+      "epoch": 1.638189339220267,
+      "grad_norm": 1.734375,
+      "learning_rate": 8.513829279337267e-06,
+      "loss": 0.8983,
+      "step": 9624
+    },
+    {
+      "epoch": 1.638360803309257,
+      "grad_norm": 1.671875,
+      "learning_rate": 8.512043430563239e-06,
+      "loss": 0.8373,
+      "step": 9625
+    },
+    {
+      "epoch": 1.6385322673982468,
+      "grad_norm": 1.65625,
+      "learning_rate": 8.510257630317079e-06,
+      "loss": 0.8467,
+      "step": 9626
+    },
+    {
+      "epoch": 1.6387037314872366,
+      "grad_norm": 1.734375,
+      "learning_rate": 8.508471878657028e-06,
+      "loss": 0.9898,
+      "step": 9627
+    },
+    {
+      "epoch": 1.6388751955762264,
+      "grad_norm": 1.625,
+      "learning_rate": 8.506686175641324e-06,
+      "loss": 0.7781,
+      "step": 9628
+    },
+    {
+      "epoch": 1.6390466596652162,
+      "grad_norm": 1.6640625,
+      "learning_rate": 8.504900521328207e-06,
+      "loss": 0.7417,
+      "step": 9629
+    },
+    {
+      "epoch": 1.6392181237542063,
+      "grad_norm": 1.6171875,
+      "learning_rate": 8.503114915775915e-06,
+      "loss": 0.7881,
+      "step": 9630
+    },
+    {
+      "epoch": 1.639389587843196,
+      "grad_norm": 1.6953125,
+      "learning_rate": 8.501329359042683e-06,
+      "loss": 0.8748,
+      "step": 9631
+    },
+    {
+      "epoch": 1.639561051932186,
+      "grad_norm": 1.765625,
+      "learning_rate": 8.499543851186742e-06,
+      "loss": 0.8101,
+      "step": 9632
+    },
+    {
+      "epoch": 1.639732516021176,
+      "grad_norm": 1.828125,
+      "learning_rate": 8.497758392266328e-06,
+      "loss": 0.8246,
+      "step": 9633
+    },
+    {
+      "epoch": 1.6399039801101658,
+      "grad_norm": 1.7109375,
+      "learning_rate": 8.49597298233967e-06,
+      "loss": 0.7272,
+      "step": 9634
+    },
+    {
+      "epoch": 1.6400754441991556,
+      "grad_norm": 1.7109375,
+      "learning_rate": 8.494187621464997e-06,
+      "loss": 0.907,
+      "step": 9635
+    },
+    {
+      "epoch": 1.6402469082881455,
+      "grad_norm": 1.78125,
+      "learning_rate": 8.492402309700535e-06,
+      "loss": 0.8957,
+      "step": 9636
+    },
+    {
+      "epoch": 1.6404183723771353,
+      "grad_norm": 1.7578125,
+      "learning_rate": 8.490617047104511e-06,
+      "loss": 0.9228,
+      "step": 9637
+    },
+    {
+      "epoch": 1.640589836466125,
+      "grad_norm": 1.7734375,
+      "learning_rate": 8.488831833735149e-06,
+      "loss": 0.9267,
+      "step": 9638
+    },
+    {
+      "epoch": 1.640761300555115,
+      "grad_norm": 1.59375,
+      "learning_rate": 8.48704666965067e-06,
+      "loss": 0.8438,
+      "step": 9639
+    },
+    {
+      "epoch": 1.6409327646441048,
+      "grad_norm": 1.6796875,
+      "learning_rate": 8.485261554909298e-06,
+      "loss": 0.8037,
+      "step": 9640
+    },
+    {
+      "epoch": 1.6411042287330946,
+      "grad_norm": 1.75,
+      "learning_rate": 8.48347648956925e-06,
+      "loss": 0.8805,
+      "step": 9641
+    },
+    {
+      "epoch": 1.6412756928220844,
+      "grad_norm": 1.7578125,
+      "learning_rate": 8.481691473688745e-06,
+      "loss": 0.9045,
+      "step": 9642
+    },
+    {
+      "epoch": 1.6414471569110745,
+      "grad_norm": 1.7421875,
+      "learning_rate": 8.479906507325997e-06,
+      "loss": 0.8374,
+      "step": 9643
+    },
+    {
+      "epoch": 1.6416186210000643,
+      "grad_norm": 1.6484375,
+      "learning_rate": 8.478121590539221e-06,
+      "loss": 0.7615,
+      "step": 9644
+    },
+    {
+      "epoch": 1.641790085089054,
+      "grad_norm": 1.6875,
+      "learning_rate": 8.476336723386632e-06,
+      "loss": 0.8641,
+      "step": 9645
+    },
+    {
+      "epoch": 1.6419615491780442,
+      "grad_norm": 1.7265625,
+      "learning_rate": 8.474551905926446e-06,
+      "loss": 0.89,
+      "step": 9646
+    },
+    {
+      "epoch": 1.642133013267034,
+      "grad_norm": 1.78125,
+      "learning_rate": 8.47276713821686e-06,
+      "loss": 0.7849,
+      "step": 9647
+    },
+    {
+      "epoch": 1.6423044773560238,
+      "grad_norm": 1.53125,
+      "learning_rate": 8.47098242031609e-06,
+      "loss": 0.8519,
+      "step": 9648
+    },
+    {
+      "epoch": 1.6424759414450136,
+      "grad_norm": 1.6953125,
+      "learning_rate": 8.469197752282343e-06,
+      "loss": 0.8685,
+      "step": 9649
+    },
+    {
+      "epoch": 1.6426474055340035,
+      "grad_norm": 1.6875,
+      "learning_rate": 8.467413134173819e-06,
+      "loss": 0.813,
+      "step": 9650
+    },
+    {
+      "epoch": 1.6428188696229933,
+      "grad_norm": 1.6953125,
+      "learning_rate": 8.465628566048724e-06,
+      "loss": 0.8472,
+      "step": 9651
+    },
+    {
+      "epoch": 1.642990333711983,
+      "grad_norm": 1.6796875,
+      "learning_rate": 8.46384404796526e-06,
+      "loss": 0.8151,
+      "step": 9652
+    },
+    {
+      "epoch": 1.643161797800973,
+      "grad_norm": 1.7265625,
+      "learning_rate": 8.462059579981624e-06,
+      "loss": 0.8978,
+      "step": 9653
+    },
+    {
+      "epoch": 1.6433332618899628,
+      "grad_norm": 1.7890625,
+      "learning_rate": 8.460275162156019e-06,
+      "loss": 0.8837,
+      "step": 9654
+    },
+    {
+      "epoch": 1.6435047259789528,
+      "grad_norm": 1.7421875,
+      "learning_rate": 8.458490794546638e-06,
+      "loss": 0.852,
+      "step": 9655
+    },
+    {
+      "epoch": 1.6436761900679426,
+      "grad_norm": 1.671875,
+      "learning_rate": 8.456706477211677e-06,
+      "loss": 0.819,
+      "step": 9656
+    },
+    {
+      "epoch": 1.6438476541569325,
+      "grad_norm": 1.71875,
+      "learning_rate": 8.454922210209332e-06,
+      "loss": 0.9105,
+      "step": 9657
+    },
+    {
+      "epoch": 1.6440191182459225,
+      "grad_norm": 1.6875,
+      "learning_rate": 8.453137993597792e-06,
+      "loss": 0.8514,
+      "step": 9658
+    },
+    {
+      "epoch": 1.6441905823349123,
+      "grad_norm": 1.765625,
+      "learning_rate": 8.451353827435247e-06,
+      "loss": 0.8103,
+      "step": 9659
+    },
+    {
+      "epoch": 1.6443620464239022,
+      "grad_norm": 1.6640625,
+      "learning_rate": 8.449569711779883e-06,
+      "loss": 0.8241,
+      "step": 9660
+    },
+    {
+      "epoch": 1.644533510512892,
+      "grad_norm": 1.71875,
+      "learning_rate": 8.447785646689887e-06,
+      "loss": 0.8696,
+      "step": 9661
+    },
+    {
+      "epoch": 1.6447049746018818,
+      "grad_norm": 1.6640625,
+      "learning_rate": 8.446001632223448e-06,
+      "loss": 0.7978,
+      "step": 9662
+    },
+    {
+      "epoch": 1.6448764386908716,
+      "grad_norm": 1.7578125,
+      "learning_rate": 8.444217668438748e-06,
+      "loss": 0.7798,
+      "step": 9663
+    },
+    {
+      "epoch": 1.6450479027798615,
+      "grad_norm": 1.796875,
+      "learning_rate": 8.442433755393968e-06,
+      "loss": 0.9269,
+      "step": 9664
+    },
+    {
+      "epoch": 1.6452193668688513,
+      "grad_norm": 1.671875,
+      "learning_rate": 8.440649893147289e-06,
+      "loss": 0.842,
+      "step": 9665
+    },
+    {
+      "epoch": 1.645390830957841,
+      "grad_norm": 1.625,
+      "learning_rate": 8.438866081756889e-06,
+      "loss": 0.8264,
+      "step": 9666
+    },
+    {
+      "epoch": 1.6455622950468312,
+      "grad_norm": 1.6796875,
+      "learning_rate": 8.437082321280945e-06,
+      "loss": 0.7968,
+      "step": 9667
+    },
+    {
+      "epoch": 1.645733759135821,
+      "grad_norm": 1.703125,
+      "learning_rate": 8.435298611777632e-06,
+      "loss": 0.8673,
+      "step": 9668
+    },
+    {
+      "epoch": 1.6459052232248108,
+      "grad_norm": 1.7890625,
+      "learning_rate": 8.433514953305124e-06,
+      "loss": 0.9696,
+      "step": 9669
+    },
+    {
+      "epoch": 1.6460766873138009,
+      "grad_norm": 1.7265625,
+      "learning_rate": 8.431731345921592e-06,
+      "loss": 0.8855,
+      "step": 9670
+    },
+    {
+      "epoch": 1.6462481514027907,
+      "grad_norm": 1.640625,
+      "learning_rate": 8.429947789685206e-06,
+      "loss": 0.822,
+      "step": 9671
+    },
+    {
+      "epoch": 1.6464196154917805,
+      "grad_norm": 1.7890625,
+      "learning_rate": 8.428164284654133e-06,
+      "loss": 0.8785,
+      "step": 9672
+    },
+    {
+      "epoch": 1.6465910795807703,
+      "grad_norm": 1.6484375,
+      "learning_rate": 8.426380830886544e-06,
+      "loss": 0.8603,
+      "step": 9673
+    },
+    {
+      "epoch": 1.6467625436697602,
+      "grad_norm": 1.6875,
+      "learning_rate": 8.4245974284406e-06,
+      "loss": 0.8427,
+      "step": 9674
+    },
+    {
+      "epoch": 1.64693400775875,
+      "grad_norm": 1.8203125,
+      "learning_rate": 8.422814077374468e-06,
+      "loss": 0.8966,
+      "step": 9675
+    },
+    {
+      "epoch": 1.6471054718477398,
+      "grad_norm": 1.6484375,
+      "learning_rate": 8.421030777746306e-06,
+      "loss": 0.8551,
+      "step": 9676
+    },
+    {
+      "epoch": 1.6472769359367296,
+      "grad_norm": 1.7421875,
+      "learning_rate": 8.419247529614278e-06,
+      "loss": 0.8953,
+      "step": 9677
+    },
+    {
+      "epoch": 1.6474484000257195,
+      "grad_norm": 1.703125,
+      "learning_rate": 8.41746433303654e-06,
+      "loss": 0.8488,
+      "step": 9678
+    },
+    {
+      "epoch": 1.6476198641147095,
+      "grad_norm": 1.640625,
+      "learning_rate": 8.41568118807125e-06,
+      "loss": 0.8468,
+      "step": 9679
+    },
+    {
+      "epoch": 1.6477913282036993,
+      "grad_norm": 1.640625,
+      "learning_rate": 8.41389809477656e-06,
+      "loss": 0.7708,
+      "step": 9680
+    },
+    {
+      "epoch": 1.6479627922926892,
+      "grad_norm": 1.75,
+      "learning_rate": 8.412115053210631e-06,
+      "loss": 0.9317,
+      "step": 9681
+    },
+    {
+      "epoch": 1.6481342563816792,
+      "grad_norm": 1.703125,
+      "learning_rate": 8.410332063431606e-06,
+      "loss": 0.8848,
+      "step": 9682
+    },
+    {
+      "epoch": 1.648305720470669,
+      "grad_norm": 1.734375,
+      "learning_rate": 8.408549125497638e-06,
+      "loss": 0.8291,
+      "step": 9683
+    },
+    {
+      "epoch": 1.6484771845596589,
+      "grad_norm": 1.703125,
+      "learning_rate": 8.406766239466878e-06,
+      "loss": 0.8033,
+      "step": 9684
+    },
+    {
+      "epoch": 1.6486486486486487,
+      "grad_norm": 1.6015625,
+      "learning_rate": 8.404983405397468e-06,
+      "loss": 0.8135,
+      "step": 9685
+    },
+    {
+      "epoch": 1.6488201127376385,
+      "grad_norm": 1.6875,
+      "learning_rate": 8.403200623347556e-06,
+      "loss": 0.8502,
+      "step": 9686
+    },
+    {
+      "epoch": 1.6489915768266283,
+      "grad_norm": 1.6796875,
+      "learning_rate": 8.401417893375286e-06,
+      "loss": 0.7815,
+      "step": 9687
+    },
+    {
+      "epoch": 1.6491630409156182,
+      "grad_norm": 1.6875,
+      "learning_rate": 8.399635215538798e-06,
+      "loss": 0.8002,
+      "step": 9688
+    },
+    {
+      "epoch": 1.649334505004608,
+      "grad_norm": 1.734375,
+      "learning_rate": 8.39785258989623e-06,
+      "loss": 0.8572,
+      "step": 9689
+    },
+    {
+      "epoch": 1.6495059690935978,
+      "grad_norm": 1.7578125,
+      "learning_rate": 8.396070016505725e-06,
+      "loss": 0.9321,
+      "step": 9690
+    },
+    {
+      "epoch": 1.6496774331825879,
+      "grad_norm": 1.75,
+      "learning_rate": 8.394287495425412e-06,
+      "loss": 0.7589,
+      "step": 9691
+    },
+    {
+      "epoch": 1.6498488972715777,
+      "grad_norm": 1.7265625,
+      "learning_rate": 8.392505026713434e-06,
+      "loss": 0.8176,
+      "step": 9692
+    },
+    {
+      "epoch": 1.6500203613605675,
+      "grad_norm": 1.65625,
+      "learning_rate": 8.390722610427923e-06,
+      "loss": 0.7663,
+      "step": 9693
+    },
+    {
+      "epoch": 1.6501918254495576,
+      "grad_norm": 1.71875,
+      "learning_rate": 8.388940246627005e-06,
+      "loss": 0.8577,
+      "step": 9694
+    },
+    {
+      "epoch": 1.6503632895385474,
+      "grad_norm": 1.765625,
+      "learning_rate": 8.387157935368811e-06,
+      "loss": 0.8385,
+      "step": 9695
+    },
+    {
+      "epoch": 1.6505347536275372,
+      "grad_norm": 1.84375,
+      "learning_rate": 8.385375676711472e-06,
+      "loss": 0.9336,
+      "step": 9696
+    },
+    {
+      "epoch": 1.650706217716527,
+      "grad_norm": 1.6171875,
+      "learning_rate": 8.38359347071311e-06,
+      "loss": 0.7651,
+      "step": 9697
+    },
+    {
+      "epoch": 1.6508776818055169,
+      "grad_norm": 1.828125,
+      "learning_rate": 8.381811317431853e-06,
+      "loss": 0.9214,
+      "step": 9698
+    },
+    {
+      "epoch": 1.6510491458945067,
+      "grad_norm": 1.640625,
+      "learning_rate": 8.380029216925824e-06,
+      "loss": 0.8356,
+      "step": 9699
+    },
+    {
+      "epoch": 1.6512206099834965,
+      "grad_norm": 1.6484375,
+      "learning_rate": 8.378247169253138e-06,
+      "loss": 0.8845,
+      "step": 9700
+    },
+    {
+      "epoch": 1.6513920740724863,
+      "grad_norm": 1.7421875,
+      "learning_rate": 8.376465174471922e-06,
+      "loss": 0.8047,
+      "step": 9701
+    },
+    {
+      "epoch": 1.6515635381614762,
+      "grad_norm": 1.6328125,
+      "learning_rate": 8.374683232640289e-06,
+      "loss": 0.8342,
+      "step": 9702
+    },
+    {
+      "epoch": 1.6517350022504662,
+      "grad_norm": 1.75,
+      "learning_rate": 8.372901343816357e-06,
+      "loss": 0.9709,
+      "step": 9703
+    },
+    {
+      "epoch": 1.651906466339456,
+      "grad_norm": 1.6171875,
+      "learning_rate": 8.371119508058245e-06,
+      "loss": 0.9,
+      "step": 9704
+    },
+    {
+      "epoch": 1.6520779304284459,
+      "grad_norm": 1.6328125,
+      "learning_rate": 8.369337725424054e-06,
+      "loss": 0.8676,
+      "step": 9705
+    },
+    {
+      "epoch": 1.652249394517436,
+      "grad_norm": 1.6796875,
+      "learning_rate": 8.3675559959719e-06,
+      "loss": 0.8422,
+      "step": 9706
+    },
+    {
+      "epoch": 1.6524208586064257,
+      "grad_norm": 1.6796875,
+      "learning_rate": 8.365774319759892e-06,
+      "loss": 0.8387,
+      "step": 9707
+    },
+    {
+      "epoch": 1.6525923226954156,
+      "grad_norm": 1.7109375,
+      "learning_rate": 8.363992696846136e-06,
+      "loss": 0.8843,
+      "step": 9708
+    },
+    {
+      "epoch": 1.6527637867844054,
+      "grad_norm": 1.6015625,
+      "learning_rate": 8.36221112728874e-06,
+      "loss": 0.839,
+      "step": 9709
+    },
+    {
+      "epoch": 1.6529352508733952,
+      "grad_norm": 1.703125,
+      "learning_rate": 8.360429611145808e-06,
+      "loss": 0.7779,
+      "step": 9710
+    },
+    {
+      "epoch": 1.653106714962385,
+      "grad_norm": 1.7109375,
+      "learning_rate": 8.35864814847544e-06,
+      "loss": 0.8224,
+      "step": 9711
+    },
+    {
+      "epoch": 1.6532781790513749,
+      "grad_norm": 1.78125,
+      "learning_rate": 8.356866739335735e-06,
+      "loss": 0.8112,
+      "step": 9712
+    },
+    {
+      "epoch": 1.6534496431403647,
+      "grad_norm": 1.703125,
+      "learning_rate": 8.355085383784794e-06,
+      "loss": 0.8244,
+      "step": 9713
+    },
+    {
+      "epoch": 1.6536211072293545,
+      "grad_norm": 1.6953125,
+      "learning_rate": 8.353304081880713e-06,
+      "loss": 0.8558,
+      "step": 9714
+    },
+    {
+      "epoch": 1.6537925713183446,
+      "grad_norm": 1.6875,
+      "learning_rate": 8.351522833681587e-06,
+      "loss": 0.8791,
+      "step": 9715
+    },
+    {
+      "epoch": 1.6539640354073344,
+      "grad_norm": 1.7890625,
+      "learning_rate": 8.34974163924551e-06,
+      "loss": 0.8825,
+      "step": 9716
+    },
+    {
+      "epoch": 1.6541354994963242,
+      "grad_norm": 1.6953125,
+      "learning_rate": 8.347960498630574e-06,
+      "loss": 0.8269,
+      "step": 9717
+    },
+    {
+      "epoch": 1.6543069635853143,
+      "grad_norm": 1.6953125,
+      "learning_rate": 8.346179411894864e-06,
+      "loss": 0.8492,
+      "step": 9718
+    },
+    {
+      "epoch": 1.654478427674304,
+      "grad_norm": 1.734375,
+      "learning_rate": 8.34439837909647e-06,
+      "loss": 0.8248,
+      "step": 9719
+    },
+    {
+      "epoch": 1.654649891763294,
+      "grad_norm": 1.7265625,
+      "learning_rate": 8.342617400293482e-06,
+      "loss": 0.8771,
+      "step": 9720
+    },
+    {
+      "epoch": 1.6548213558522837,
+      "grad_norm": 1.6875,
+      "learning_rate": 8.34083647554398e-06,
+      "loss": 0.8594,
+      "step": 9721
+    },
+    {
+      "epoch": 1.6549928199412736,
+      "grad_norm": 1.671875,
+      "learning_rate": 8.33905560490605e-06,
+      "loss": 0.8927,
+      "step": 9722
+    },
+    {
+      "epoch": 1.6551642840302634,
+      "grad_norm": 1.640625,
+      "learning_rate": 8.337274788437769e-06,
+      "loss": 0.789,
+      "step": 9723
+    },
+    {
+      "epoch": 1.6553357481192532,
+      "grad_norm": 1.7109375,
+      "learning_rate": 8.33549402619722e-06,
+      "loss": 0.7861,
+      "step": 9724
+    },
+    {
+      "epoch": 1.655507212208243,
+      "grad_norm": 1.75,
+      "learning_rate": 8.333713318242477e-06,
+      "loss": 0.7888,
+      "step": 9725
+    },
+    {
+      "epoch": 1.6556786762972329,
+      "grad_norm": 1.6796875,
+      "learning_rate": 8.331932664631619e-06,
+      "loss": 0.8536,
+      "step": 9726
+    },
+    {
+      "epoch": 1.655850140386223,
+      "grad_norm": 1.6953125,
+      "learning_rate": 8.33015206542272e-06,
+      "loss": 0.7891,
+      "step": 9727
+    },
+    {
+      "epoch": 1.6560216044752127,
+      "grad_norm": 1.734375,
+      "learning_rate": 8.328371520673848e-06,
+      "loss": 0.887,
+      "step": 9728
+    },
+    {
+      "epoch": 1.6561930685642026,
+      "grad_norm": 1.7109375,
+      "learning_rate": 8.326591030443075e-06,
+      "loss": 0.8448,
+      "step": 9729
+    },
+    {
+      "epoch": 1.6563645326531926,
+      "grad_norm": 1.96875,
+      "learning_rate": 8.324810594788471e-06,
+      "loss": 0.848,
+      "step": 9730
+    },
+    {
+      "epoch": 1.6565359967421824,
+      "grad_norm": 1.7421875,
+      "learning_rate": 8.3230302137681e-06,
+      "loss": 0.9769,
+      "step": 9731
+    },
+    {
+      "epoch": 1.6567074608311723,
+      "grad_norm": 1.6484375,
+      "learning_rate": 8.32124988744003e-06,
+      "loss": 0.826,
+      "step": 9732
+    },
+    {
+      "epoch": 1.656878924920162,
+      "grad_norm": 1.84375,
+      "learning_rate": 8.319469615862324e-06,
+      "loss": 0.9164,
+      "step": 9733
+    },
+    {
+      "epoch": 1.657050389009152,
+      "grad_norm": 1.703125,
+      "learning_rate": 8.317689399093039e-06,
+      "loss": 0.8394,
+      "step": 9734
+    },
+    {
+      "epoch": 1.6572218530981417,
+      "grad_norm": 1.6796875,
+      "learning_rate": 8.315909237190241e-06,
+      "loss": 0.858,
+      "step": 9735
+    },
+    {
+      "epoch": 1.6573933171871316,
+      "grad_norm": 1.703125,
+      "learning_rate": 8.314129130211981e-06,
+      "loss": 0.8047,
+      "step": 9736
+    },
+    {
+      "epoch": 1.6575647812761214,
+      "grad_norm": 1.625,
+      "learning_rate": 8.31234907821632e-06,
+      "loss": 0.8094,
+      "step": 9737
+    },
+    {
+      "epoch": 1.6577362453651112,
+      "grad_norm": 1.78125,
+      "learning_rate": 8.31056908126131e-06,
+      "loss": 0.9054,
+      "step": 9738
+    },
+    {
+      "epoch": 1.657907709454101,
+      "grad_norm": 1.6953125,
+      "learning_rate": 8.308789139405008e-06,
+      "loss": 0.8816,
+      "step": 9739
+    },
+    {
+      "epoch": 1.658079173543091,
+      "grad_norm": 1.7265625,
+      "learning_rate": 8.30700925270546e-06,
+      "loss": 0.744,
+      "step": 9740
+    },
+    {
+      "epoch": 1.658250637632081,
+      "grad_norm": 1.7734375,
+      "learning_rate": 8.305229421220712e-06,
+      "loss": 0.9439,
+      "step": 9741
+    },
+    {
+      "epoch": 1.6584221017210707,
+      "grad_norm": 1.7421875,
+      "learning_rate": 8.303449645008815e-06,
+      "loss": 0.8557,
+      "step": 9742
+    },
+    {
+      "epoch": 1.6585935658100608,
+      "grad_norm": 1.7265625,
+      "learning_rate": 8.301669924127815e-06,
+      "loss": 0.8596,
+      "step": 9743
+    },
+    {
+      "epoch": 1.6587650298990506,
+      "grad_norm": 1.7734375,
+      "learning_rate": 8.299890258635753e-06,
+      "loss": 0.9133,
+      "step": 9744
+    },
+    {
+      "epoch": 1.6589364939880404,
+      "grad_norm": 1.703125,
+      "learning_rate": 8.298110648590673e-06,
+      "loss": 0.8114,
+      "step": 9745
+    },
+    {
+      "epoch": 1.6591079580770303,
+      "grad_norm": 1.65625,
+      "learning_rate": 8.296331094050609e-06,
+      "loss": 0.8435,
+      "step": 9746
+    },
+    {
+      "epoch": 1.65927942216602,
+      "grad_norm": 1.734375,
+      "learning_rate": 8.294551595073608e-06,
+      "loss": 0.8475,
+      "step": 9747
+    },
+    {
+      "epoch": 1.65945088625501,
+      "grad_norm": 1.6640625,
+      "learning_rate": 8.2927721517177e-06,
+      "loss": 0.8274,
+      "step": 9748
+    },
+    {
+      "epoch": 1.6596223503439997,
+      "grad_norm": 1.7265625,
+      "learning_rate": 8.290992764040922e-06,
+      "loss": 0.8764,
+      "step": 9749
+    },
+    {
+      "epoch": 1.6597938144329896,
+      "grad_norm": 1.6875,
+      "learning_rate": 8.28921343210131e-06,
+      "loss": 0.8514,
+      "step": 9750
+    },
+    {
+      "epoch": 1.6599652785219794,
+      "grad_norm": 1.703125,
+      "learning_rate": 8.287434155956885e-06,
+      "loss": 0.8684,
+      "step": 9751
+    },
+    {
+      "epoch": 1.6601367426109694,
+      "grad_norm": 1.7421875,
+      "learning_rate": 8.28565493566568e-06,
+      "loss": 0.9562,
+      "step": 9752
+    },
+    {
+      "epoch": 1.6603082066999593,
+      "grad_norm": 1.6796875,
+      "learning_rate": 8.283875771285725e-06,
+      "loss": 0.8543,
+      "step": 9753
+    },
+    {
+      "epoch": 1.660479670788949,
+      "grad_norm": 1.7734375,
+      "learning_rate": 8.282096662875042e-06,
+      "loss": 0.9567,
+      "step": 9754
+    },
+    {
+      "epoch": 1.6606511348779391,
+      "grad_norm": 1.640625,
+      "learning_rate": 8.280317610491655e-06,
+      "loss": 0.8085,
+      "step": 9755
+    },
+    {
+      "epoch": 1.660822598966929,
+      "grad_norm": 1.8203125,
+      "learning_rate": 8.27853861419359e-06,
+      "loss": 0.9327,
+      "step": 9756
+    },
+    {
+      "epoch": 1.6609940630559188,
+      "grad_norm": 1.7265625,
+      "learning_rate": 8.276759674038861e-06,
+      "loss": 0.8291,
+      "step": 9757
+    },
+    {
+      "epoch": 1.6611655271449086,
+      "grad_norm": 1.6640625,
+      "learning_rate": 8.274980790085489e-06,
+      "loss": 0.8373,
+      "step": 9758
+    },
+    {
+      "epoch": 1.6613369912338984,
+      "grad_norm": 1.71875,
+      "learning_rate": 8.273201962391488e-06,
+      "loss": 0.8789,
+      "step": 9759
+    },
+    {
+      "epoch": 1.6615084553228883,
+      "grad_norm": 1.734375,
+      "learning_rate": 8.271423191014874e-06,
+      "loss": 0.868,
+      "step": 9760
+    },
+    {
+      "epoch": 1.661679919411878,
+      "grad_norm": 1.7578125,
+      "learning_rate": 8.269644476013661e-06,
+      "loss": 0.8498,
+      "step": 9761
+    },
+    {
+      "epoch": 1.661851383500868,
+      "grad_norm": 1.6484375,
+      "learning_rate": 8.26786581744586e-06,
+      "loss": 0.8065,
+      "step": 9762
+    },
+    {
+      "epoch": 1.6620228475898577,
+      "grad_norm": 1.75,
+      "learning_rate": 8.26608721536947e-06,
+      "loss": 0.8592,
+      "step": 9763
+    },
+    {
+      "epoch": 1.6621943116788478,
+      "grad_norm": 1.78125,
+      "learning_rate": 8.26430866984251e-06,
+      "loss": 0.8404,
+      "step": 9764
+    },
+    {
+      "epoch": 1.6623657757678376,
+      "grad_norm": 1.6796875,
+      "learning_rate": 8.262530180922978e-06,
+      "loss": 0.824,
+      "step": 9765
+    },
+    {
+      "epoch": 1.6625372398568274,
+      "grad_norm": 1.796875,
+      "learning_rate": 8.260751748668881e-06,
+      "loss": 0.8971,
+      "step": 9766
+    },
+    {
+      "epoch": 1.6627087039458175,
+      "grad_norm": 1.7578125,
+      "learning_rate": 8.258973373138218e-06,
+      "loss": 0.924,
+      "step": 9767
+    },
+    {
+      "epoch": 1.6628801680348073,
+      "grad_norm": 1.765625,
+      "learning_rate": 8.25719505438899e-06,
+      "loss": 0.8984,
+      "step": 9768
+    },
+    {
+      "epoch": 1.6630516321237971,
+      "grad_norm": 1.765625,
+      "learning_rate": 8.255416792479192e-06,
+      "loss": 0.747,
+      "step": 9769
+    },
+    {
+      "epoch": 1.663223096212787,
+      "grad_norm": 1.7421875,
+      "learning_rate": 8.253638587466823e-06,
+      "loss": 0.9168,
+      "step": 9770
+    },
+    {
+      "epoch": 1.6633945603017768,
+      "grad_norm": 1.71875,
+      "learning_rate": 8.251860439409877e-06,
+      "loss": 0.8687,
+      "step": 9771
+    },
+    {
+      "epoch": 1.6635660243907666,
+      "grad_norm": 1.84375,
+      "learning_rate": 8.250082348366343e-06,
+      "loss": 0.9425,
+      "step": 9772
+    },
+    {
+      "epoch": 1.6637374884797564,
+      "grad_norm": 1.6953125,
+      "learning_rate": 8.248304314394218e-06,
+      "loss": 0.8669,
+      "step": 9773
+    },
+    {
+      "epoch": 1.6639089525687463,
+      "grad_norm": 1.7109375,
+      "learning_rate": 8.246526337551482e-06,
+      "loss": 0.8023,
+      "step": 9774
+    },
+    {
+      "epoch": 1.664080416657736,
+      "grad_norm": 1.7109375,
+      "learning_rate": 8.244748417896126e-06,
+      "loss": 0.9566,
+      "step": 9775
+    },
+    {
+      "epoch": 1.6642518807467261,
+      "grad_norm": 1.703125,
+      "learning_rate": 8.242970555486131e-06,
+      "loss": 0.8965,
+      "step": 9776
+    },
+    {
+      "epoch": 1.664423344835716,
+      "grad_norm": 1.671875,
+      "learning_rate": 8.241192750379484e-06,
+      "loss": 0.8908,
+      "step": 9777
+    },
+    {
+      "epoch": 1.6645948089247058,
+      "grad_norm": 1.6953125,
+      "learning_rate": 8.239415002634166e-06,
+      "loss": 0.8447,
+      "step": 9778
+    },
+    {
+      "epoch": 1.6647662730136958,
+      "grad_norm": 1.640625,
+      "learning_rate": 8.237637312308154e-06,
+      "loss": 0.8,
+      "step": 9779
+    },
+    {
+      "epoch": 1.6649377371026857,
+      "grad_norm": 1.7109375,
+      "learning_rate": 8.235859679459427e-06,
+      "loss": 0.8898,
+      "step": 9780
+    },
+    {
+      "epoch": 1.6651092011916755,
+      "grad_norm": 1.671875,
+      "learning_rate": 8.234082104145956e-06,
+      "loss": 0.9017,
+      "step": 9781
+    },
+    {
+      "epoch": 1.6652806652806653,
+      "grad_norm": 1.78125,
+      "learning_rate": 8.232304586425717e-06,
+      "loss": 0.8863,
+      "step": 9782
+    },
+    {
+      "epoch": 1.6654521293696551,
+      "grad_norm": 1.625,
+      "learning_rate": 8.230527126356684e-06,
+      "loss": 0.8047,
+      "step": 9783
+    },
+    {
+      "epoch": 1.665623593458645,
+      "grad_norm": 1.7265625,
+      "learning_rate": 8.228749723996825e-06,
+      "loss": 0.9204,
+      "step": 9784
+    },
+    {
+      "epoch": 1.6657950575476348,
+      "grad_norm": 1.7734375,
+      "learning_rate": 8.226972379404108e-06,
+      "loss": 0.9076,
+      "step": 9785
+    },
+    {
+      "epoch": 1.6659665216366246,
+      "grad_norm": 1.6875,
+      "learning_rate": 8.225195092636497e-06,
+      "loss": 0.8587,
+      "step": 9786
+    },
+    {
+      "epoch": 1.6661379857256144,
+      "grad_norm": 1.671875,
+      "learning_rate": 8.223417863751956e-06,
+      "loss": 0.8356,
+      "step": 9787
+    },
+    {
+      "epoch": 1.6663094498146045,
+      "grad_norm": 1.6796875,
+      "learning_rate": 8.221640692808451e-06,
+      "loss": 0.8909,
+      "step": 9788
+    },
+    {
+      "epoch": 1.6664809139035943,
+      "grad_norm": 1.625,
+      "learning_rate": 8.219863579863939e-06,
+      "loss": 0.7785,
+      "step": 9789
+    },
+    {
+      "epoch": 1.6666523779925841,
+      "grad_norm": 1.6484375,
+      "learning_rate": 8.218086524976377e-06,
+      "loss": 0.8028,
+      "step": 9790
+    },
+    {
+      "epoch": 1.6668238420815742,
+      "grad_norm": 1.6328125,
+      "learning_rate": 8.216309528203726e-06,
+      "loss": 0.8187,
+      "step": 9791
+    },
+    {
+      "epoch": 1.666995306170564,
+      "grad_norm": 1.75,
+      "learning_rate": 8.214532589603936e-06,
+      "loss": 0.9681,
+      "step": 9792
+    },
+    {
+      "epoch": 1.6671667702595538,
+      "grad_norm": 1.734375,
+      "learning_rate": 8.212755709234962e-06,
+      "loss": 0.8375,
+      "step": 9793
+    },
+    {
+      "epoch": 1.6673382343485437,
+      "grad_norm": 1.78125,
+      "learning_rate": 8.210978887154753e-06,
+      "loss": 0.9124,
+      "step": 9794
+    },
+    {
+      "epoch": 1.6675096984375335,
+      "grad_norm": 1.6796875,
+      "learning_rate": 8.209202123421262e-06,
+      "loss": 0.8024,
+      "step": 9795
+    },
+    {
+      "epoch": 1.6676811625265233,
+      "grad_norm": 1.625,
+      "learning_rate": 8.207425418092438e-06,
+      "loss": 0.8746,
+      "step": 9796
+    },
+    {
+      "epoch": 1.6678526266155131,
+      "grad_norm": 1.6328125,
+      "learning_rate": 8.205648771226215e-06,
+      "loss": 0.8642,
+      "step": 9797
+    },
+    {
+      "epoch": 1.668024090704503,
+      "grad_norm": 1.8125,
+      "learning_rate": 8.203872182880544e-06,
+      "loss": 0.8817,
+      "step": 9798
+    },
+    {
+      "epoch": 1.6681955547934928,
+      "grad_norm": 1.6796875,
+      "learning_rate": 8.202095653113365e-06,
+      "loss": 0.8011,
+      "step": 9799
+    },
+    {
+      "epoch": 1.6683670188824828,
+      "grad_norm": 1.7109375,
+      "learning_rate": 8.200319181982615e-06,
+      "loss": 0.8085,
+      "step": 9800
+    },
+    {
+      "epoch": 1.6683670188824828,
+      "eval_loss": 0.8372142910957336,
+      "eval_runtime": 835.8053,
+      "eval_samples_per_second": 2.99,
+      "eval_steps_per_second": 2.99,
+      "step": 9800
+    },
+    {
+      "epoch": 1.6685384829714727,
+      "grad_norm": 1.7109375,
+      "learning_rate": 8.198542769546235e-06,
+      "loss": 0.8591,
+      "step": 9801
+    },
+    {
+      "epoch": 1.6687099470604625,
+      "grad_norm": 1.7578125,
+      "learning_rate": 8.196766415862157e-06,
+      "loss": 0.9123,
+      "step": 9802
+    },
+    {
+      "epoch": 1.6688814111494525,
+      "grad_norm": 1.6953125,
+      "learning_rate": 8.194990120988317e-06,
+      "loss": 0.7732,
+      "step": 9803
+    },
+    {
+      "epoch": 1.6690528752384424,
+      "grad_norm": 1.6796875,
+      "learning_rate": 8.193213884982648e-06,
+      "loss": 0.8743,
+      "step": 9804
+    },
+    {
+      "epoch": 1.6692243393274322,
+      "grad_norm": 1.8203125,
+      "learning_rate": 8.191437707903077e-06,
+      "loss": 0.9104,
+      "step": 9805
+    },
+    {
+      "epoch": 1.669395803416422,
+      "grad_norm": 1.7578125,
+      "learning_rate": 8.189661589807532e-06,
+      "loss": 0.9285,
+      "step": 9806
+    },
+    {
+      "epoch": 1.6695672675054118,
+      "grad_norm": 1.703125,
+      "learning_rate": 8.187885530753942e-06,
+      "loss": 0.7954,
+      "step": 9807
+    },
+    {
+      "epoch": 1.6697387315944017,
+      "grad_norm": 1.7109375,
+      "learning_rate": 8.186109530800231e-06,
+      "loss": 0.9223,
+      "step": 9808
+    },
+    {
+      "epoch": 1.6699101956833915,
+      "grad_norm": 1.7734375,
+      "learning_rate": 8.184333590004315e-06,
+      "loss": 0.8979,
+      "step": 9809
+    },
+    {
+      "epoch": 1.6700816597723813,
+      "grad_norm": 1.625,
+      "learning_rate": 8.182557708424116e-06,
+      "loss": 0.8499,
+      "step": 9810
+    },
+    {
+      "epoch": 1.6702531238613711,
+      "grad_norm": 1.6171875,
+      "learning_rate": 8.180781886117554e-06,
+      "loss": 0.8487,
+      "step": 9811
+    },
+    {
+      "epoch": 1.6704245879503612,
+      "grad_norm": 1.7421875,
+      "learning_rate": 8.179006123142548e-06,
+      "loss": 0.8923,
+      "step": 9812
+    },
+    {
+      "epoch": 1.670596052039351,
+      "grad_norm": 1.6953125,
+      "learning_rate": 8.177230419557009e-06,
+      "loss": 0.8345,
+      "step": 9813
+    },
+    {
+      "epoch": 1.6707675161283408,
+      "grad_norm": 1.765625,
+      "learning_rate": 8.17545477541885e-06,
+      "loss": 0.9097,
+      "step": 9814
+    },
+    {
+      "epoch": 1.670938980217331,
+      "grad_norm": 1.7421875,
+      "learning_rate": 8.17367919078598e-06,
+      "loss": 0.8291,
+      "step": 9815
+    },
+    {
+      "epoch": 1.6711104443063207,
+      "grad_norm": 1.5703125,
+      "learning_rate": 8.17190366571631e-06,
+      "loss": 0.7673,
+      "step": 9816
+    },
+    {
+      "epoch": 1.6712819083953105,
+      "grad_norm": 1.6875,
+      "learning_rate": 8.170128200267745e-06,
+      "loss": 0.8326,
+      "step": 9817
+    },
+    {
+      "epoch": 1.6714533724843004,
+      "grad_norm": 1.6875,
+      "learning_rate": 8.16835279449819e-06,
+      "loss": 0.7833,
+      "step": 9818
+    },
+    {
+      "epoch": 1.6716248365732902,
+      "grad_norm": 1.828125,
+      "learning_rate": 8.16657744846555e-06,
+      "loss": 0.8009,
+      "step": 9819
+    },
+    {
+      "epoch": 1.67179630066228,
+      "grad_norm": 1.7265625,
+      "learning_rate": 8.164802162227722e-06,
+      "loss": 0.8816,
+      "step": 9820
+    },
+    {
+      "epoch": 1.6719677647512698,
+      "grad_norm": 1.6640625,
+      "learning_rate": 8.163026935842604e-06,
+      "loss": 0.8289,
+      "step": 9821
+    },
+    {
+      "epoch": 1.6721392288402597,
+      "grad_norm": 1.7109375,
+      "learning_rate": 8.161251769368098e-06,
+      "loss": 0.9218,
+      "step": 9822
+    },
+    {
+      "epoch": 1.6723106929292495,
+      "grad_norm": 1.703125,
+      "learning_rate": 8.159476662862094e-06,
+      "loss": 0.8668,
+      "step": 9823
+    },
+    {
+      "epoch": 1.6724821570182395,
+      "grad_norm": 1.6796875,
+      "learning_rate": 8.157701616382487e-06,
+      "loss": 0.8952,
+      "step": 9824
+    },
+    {
+      "epoch": 1.6726536211072294,
+      "grad_norm": 1.71875,
+      "learning_rate": 8.155926629987169e-06,
+      "loss": 0.9246,
+      "step": 9825
+    },
+    {
+      "epoch": 1.6728250851962192,
+      "grad_norm": 1.7265625,
+      "learning_rate": 8.154151703734026e-06,
+      "loss": 0.8608,
+      "step": 9826
+    },
+    {
+      "epoch": 1.6729965492852092,
+      "grad_norm": 1.7890625,
+      "learning_rate": 8.152376837680947e-06,
+      "loss": 0.8636,
+      "step": 9827
+    },
+    {
+      "epoch": 1.673168013374199,
+      "grad_norm": 1.75,
+      "learning_rate": 8.150602031885818e-06,
+      "loss": 0.9174,
+      "step": 9828
+    },
+    {
+      "epoch": 1.673339477463189,
+      "grad_norm": 1.65625,
+      "learning_rate": 8.14882728640652e-06,
+      "loss": 0.8828,
+      "step": 9829
+    },
+    {
+      "epoch": 1.6735109415521787,
+      "grad_norm": 1.65625,
+      "learning_rate": 8.147052601300936e-06,
+      "loss": 0.8424,
+      "step": 9830
+    },
+    {
+      "epoch": 1.6736824056411685,
+      "grad_norm": 1.7421875,
+      "learning_rate": 8.145277976626945e-06,
+      "loss": 0.8268,
+      "step": 9831
+    },
+    {
+      "epoch": 1.6738538697301584,
+      "grad_norm": 1.671875,
+      "learning_rate": 8.143503412442423e-06,
+      "loss": 0.9055,
+      "step": 9832
+    },
+    {
+      "epoch": 1.6740253338191482,
+      "grad_norm": 1.7109375,
+      "learning_rate": 8.141728908805244e-06,
+      "loss": 0.854,
+      "step": 9833
+    },
+    {
+      "epoch": 1.674196797908138,
+      "grad_norm": 1.71875,
+      "learning_rate": 8.139954465773283e-06,
+      "loss": 0.8855,
+      "step": 9834
+    },
+    {
+      "epoch": 1.6743682619971278,
+      "grad_norm": 1.734375,
+      "learning_rate": 8.138180083404412e-06,
+      "loss": 0.8924,
+      "step": 9835
+    },
+    {
+      "epoch": 1.6745397260861177,
+      "grad_norm": 1.6640625,
+      "learning_rate": 8.1364057617565e-06,
+      "loss": 0.8561,
+      "step": 9836
+    },
+    {
+      "epoch": 1.6747111901751077,
+      "grad_norm": 1.7265625,
+      "learning_rate": 8.134631500887412e-06,
+      "loss": 0.9067,
+      "step": 9837
+    },
+    {
+      "epoch": 1.6748826542640975,
+      "grad_norm": 1.8359375,
+      "learning_rate": 8.132857300855016e-06,
+      "loss": 0.8568,
+      "step": 9838
+    },
+    {
+      "epoch": 1.6750541183530874,
+      "grad_norm": 1.7578125,
+      "learning_rate": 8.131083161717175e-06,
+      "loss": 0.8184,
+      "step": 9839
+    },
+    {
+      "epoch": 1.6752255824420774,
+      "grad_norm": 1.6328125,
+      "learning_rate": 8.129309083531746e-06,
+      "loss": 0.8091,
+      "step": 9840
+    },
+    {
+      "epoch": 1.6753970465310672,
+      "grad_norm": 1.7109375,
+      "learning_rate": 8.127535066356595e-06,
+      "loss": 0.8772,
+      "step": 9841
+    },
+    {
+      "epoch": 1.675568510620057,
+      "grad_norm": 1.5859375,
+      "learning_rate": 8.12576111024958e-06,
+      "loss": 0.7298,
+      "step": 9842
+    },
+    {
+      "epoch": 1.675739974709047,
+      "grad_norm": 1.5625,
+      "learning_rate": 8.123987215268551e-06,
+      "loss": 0.806,
+      "step": 9843
+    },
+    {
+      "epoch": 1.6759114387980367,
+      "grad_norm": 1.640625,
+      "learning_rate": 8.122213381471363e-06,
+      "loss": 0.8501,
+      "step": 9844
+    },
+    {
+      "epoch": 1.6760829028870265,
+      "grad_norm": 1.7265625,
+      "learning_rate": 8.120439608915866e-06,
+      "loss": 0.8018,
+      "step": 9845
+    },
+    {
+      "epoch": 1.6762543669760164,
+      "grad_norm": 1.78125,
+      "learning_rate": 8.118665897659912e-06,
+      "loss": 0.8389,
+      "step": 9846
+    },
+    {
+      "epoch": 1.6764258310650062,
+      "grad_norm": 1.7578125,
+      "learning_rate": 8.116892247761348e-06,
+      "loss": 0.8466,
+      "step": 9847
+    },
+    {
+      "epoch": 1.676597295153996,
+      "grad_norm": 1.703125,
+      "learning_rate": 8.115118659278019e-06,
+      "loss": 0.8985,
+      "step": 9848
+    },
+    {
+      "epoch": 1.676768759242986,
+      "grad_norm": 1.7578125,
+      "learning_rate": 8.113345132267765e-06,
+      "loss": 0.8728,
+      "step": 9849
+    },
+    {
+      "epoch": 1.676940223331976,
+      "grad_norm": 1.7421875,
+      "learning_rate": 8.111571666788433e-06,
+      "loss": 0.7813,
+      "step": 9850
+    },
+    {
+      "epoch": 1.6771116874209657,
+      "grad_norm": 1.671875,
+      "learning_rate": 8.10979826289786e-06,
+      "loss": 0.779,
+      "step": 9851
+    },
+    {
+      "epoch": 1.6772831515099558,
+      "grad_norm": 1.765625,
+      "learning_rate": 8.108024920653885e-06,
+      "loss": 0.8168,
+      "step": 9852
+    },
+    {
+      "epoch": 1.6774546155989456,
+      "grad_norm": 1.6484375,
+      "learning_rate": 8.106251640114342e-06,
+      "loss": 0.8528,
+      "step": 9853
+    },
+    {
+      "epoch": 1.6776260796879354,
+      "grad_norm": 1.71875,
+      "learning_rate": 8.104478421337065e-06,
+      "loss": 0.8275,
+      "step": 9854
+    },
+    {
+      "epoch": 1.6777975437769252,
+      "grad_norm": 1.578125,
+      "learning_rate": 8.102705264379884e-06,
+      "loss": 0.7311,
+      "step": 9855
+    },
+    {
+      "epoch": 1.677969007865915,
+      "grad_norm": 1.78125,
+      "learning_rate": 8.100932169300627e-06,
+      "loss": 0.9246,
+      "step": 9856
+    },
+    {
+      "epoch": 1.678140471954905,
+      "grad_norm": 1.640625,
+      "learning_rate": 8.099159136157122e-06,
+      "loss": 0.8039,
+      "step": 9857
+    },
+    {
+      "epoch": 1.6783119360438947,
+      "grad_norm": 1.6328125,
+      "learning_rate": 8.097386165007197e-06,
+      "loss": 0.8484,
+      "step": 9858
+    },
+    {
+      "epoch": 1.6784834001328846,
+      "grad_norm": 1.7578125,
+      "learning_rate": 8.095613255908674e-06,
+      "loss": 0.878,
+      "step": 9859
+    },
+    {
+      "epoch": 1.6786548642218744,
+      "grad_norm": 1.6953125,
+      "learning_rate": 8.093840408919373e-06,
+      "loss": 0.9227,
+      "step": 9860
+    },
+    {
+      "epoch": 1.6788263283108644,
+      "grad_norm": 1.828125,
+      "learning_rate": 8.092067624097116e-06,
+      "loss": 0.9157,
+      "step": 9861
+    },
+    {
+      "epoch": 1.6789977923998542,
+      "grad_norm": 1.6484375,
+      "learning_rate": 8.090294901499718e-06,
+      "loss": 0.8615,
+      "step": 9862
+    },
+    {
+      "epoch": 1.679169256488844,
+      "grad_norm": 1.7109375,
+      "learning_rate": 8.088522241184992e-06,
+      "loss": 0.9062,
+      "step": 9863
+    },
+    {
+      "epoch": 1.6793407205778341,
+      "grad_norm": 1.75,
+      "learning_rate": 8.086749643210758e-06,
+      "loss": 0.8219,
+      "step": 9864
+    },
+    {
+      "epoch": 1.679512184666824,
+      "grad_norm": 1.7109375,
+      "learning_rate": 8.08497710763482e-06,
+      "loss": 0.8179,
+      "step": 9865
+    },
+    {
+      "epoch": 1.6796836487558138,
+      "grad_norm": 1.671875,
+      "learning_rate": 8.083204634514991e-06,
+      "loss": 0.7425,
+      "step": 9866
+    },
+    {
+      "epoch": 1.6798551128448036,
+      "grad_norm": 1.71875,
+      "learning_rate": 8.081432223909076e-06,
+      "loss": 0.8082,
+      "step": 9867
+    },
+    {
+      "epoch": 1.6800265769337934,
+      "grad_norm": 1.6640625,
+      "learning_rate": 8.079659875874883e-06,
+      "loss": 0.8319,
+      "step": 9868
+    },
+    {
+      "epoch": 1.6801980410227833,
+      "grad_norm": 1.8203125,
+      "learning_rate": 8.07788759047021e-06,
+      "loss": 0.8149,
+      "step": 9869
+    },
+    {
+      "epoch": 1.680369505111773,
+      "grad_norm": 1.7890625,
+      "learning_rate": 8.07611536775286e-06,
+      "loss": 0.8184,
+      "step": 9870
+    },
+    {
+      "epoch": 1.680540969200763,
+      "grad_norm": 1.7578125,
+      "learning_rate": 8.074343207780635e-06,
+      "loss": 0.8821,
+      "step": 9871
+    },
+    {
+      "epoch": 1.6807124332897527,
+      "grad_norm": 1.640625,
+      "learning_rate": 8.072571110611329e-06,
+      "loss": 0.8301,
+      "step": 9872
+    },
+    {
+      "epoch": 1.6808838973787428,
+      "grad_norm": 1.7421875,
+      "learning_rate": 8.070799076302735e-06,
+      "loss": 0.9121,
+      "step": 9873
+    },
+    {
+      "epoch": 1.6810553614677326,
+      "grad_norm": 1.65625,
+      "learning_rate": 8.06902710491265e-06,
+      "loss": 0.8864,
+      "step": 9874
+    },
+    {
+      "epoch": 1.6812268255567224,
+      "grad_norm": 1.671875,
+      "learning_rate": 8.067255196498862e-06,
+      "loss": 0.8234,
+      "step": 9875
+    },
+    {
+      "epoch": 1.6813982896457125,
+      "grad_norm": 1.796875,
+      "learning_rate": 8.06548335111916e-06,
+      "loss": 0.9018,
+      "step": 9876
+    },
+    {
+      "epoch": 1.6815697537347023,
+      "grad_norm": 1.7421875,
+      "learning_rate": 8.063711568831332e-06,
+      "loss": 0.9811,
+      "step": 9877
+    },
+    {
+      "epoch": 1.6817412178236921,
+      "grad_norm": 1.640625,
+      "learning_rate": 8.06193984969316e-06,
+      "loss": 0.8669,
+      "step": 9878
+    },
+    {
+      "epoch": 1.681912681912682,
+      "grad_norm": 1.7265625,
+      "learning_rate": 8.060168193762428e-06,
+      "loss": 0.89,
+      "step": 9879
+    },
+    {
+      "epoch": 1.6820841460016718,
+      "grad_norm": 1.6875,
+      "learning_rate": 8.058396601096914e-06,
+      "loss": 0.8921,
+      "step": 9880
+    },
+    {
+      "epoch": 1.6822556100906616,
+      "grad_norm": 1.6328125,
+      "learning_rate": 8.0566250717544e-06,
+      "loss": 0.8404,
+      "step": 9881
+    },
+    {
+      "epoch": 1.6824270741796514,
+      "grad_norm": 1.7578125,
+      "learning_rate": 8.05485360579266e-06,
+      "loss": 0.9198,
+      "step": 9882
+    },
+    {
+      "epoch": 1.6825985382686413,
+      "grad_norm": 1.6015625,
+      "learning_rate": 8.053082203269467e-06,
+      "loss": 0.8214,
+      "step": 9883
+    },
+    {
+      "epoch": 1.682770002357631,
+      "grad_norm": 1.671875,
+      "learning_rate": 8.051310864242598e-06,
+      "loss": 0.853,
+      "step": 9884
+    },
+    {
+      "epoch": 1.6829414664466211,
+      "grad_norm": 1.6875,
+      "learning_rate": 8.049539588769816e-06,
+      "loss": 0.8835,
+      "step": 9885
+    },
+    {
+      "epoch": 1.683112930535611,
+      "grad_norm": 1.7265625,
+      "learning_rate": 8.047768376908896e-06,
+      "loss": 0.7677,
+      "step": 9886
+    },
+    {
+      "epoch": 1.6832843946246008,
+      "grad_norm": 1.6640625,
+      "learning_rate": 8.045997228717597e-06,
+      "loss": 0.8603,
+      "step": 9887
+    },
+    {
+      "epoch": 1.6834558587135908,
+      "grad_norm": 1.6953125,
+      "learning_rate": 8.04422614425369e-06,
+      "loss": 0.9081,
+      "step": 9888
+    },
+    {
+      "epoch": 1.6836273228025807,
+      "grad_norm": 1.7578125,
+      "learning_rate": 8.042455123574936e-06,
+      "loss": 0.8705,
+      "step": 9889
+    },
+    {
+      "epoch": 1.6837987868915705,
+      "grad_norm": 1.640625,
+      "learning_rate": 8.040684166739088e-06,
+      "loss": 0.8418,
+      "step": 9890
+    },
+    {
+      "epoch": 1.6839702509805603,
+      "grad_norm": 1.7265625,
+      "learning_rate": 8.038913273803906e-06,
+      "loss": 0.8933,
+      "step": 9891
+    },
+    {
+      "epoch": 1.6841417150695501,
+      "grad_norm": 1.6640625,
+      "learning_rate": 8.03714244482715e-06,
+      "loss": 0.9062,
+      "step": 9892
+    },
+    {
+      "epoch": 1.68431317915854,
+      "grad_norm": 1.5546875,
+      "learning_rate": 8.035371679866569e-06,
+      "loss": 0.7517,
+      "step": 9893
+    },
+    {
+      "epoch": 1.6844846432475298,
+      "grad_norm": 1.671875,
+      "learning_rate": 8.033600978979913e-06,
+      "loss": 0.826,
+      "step": 9894
+    },
+    {
+      "epoch": 1.6846561073365196,
+      "grad_norm": 1.7265625,
+      "learning_rate": 8.031830342224935e-06,
+      "loss": 0.8717,
+      "step": 9895
+    },
+    {
+      "epoch": 1.6848275714255094,
+      "grad_norm": 1.6875,
+      "learning_rate": 8.030059769659382e-06,
+      "loss": 0.8776,
+      "step": 9896
+    },
+    {
+      "epoch": 1.6849990355144995,
+      "grad_norm": 1.6484375,
+      "learning_rate": 8.028289261340998e-06,
+      "loss": 0.8231,
+      "step": 9897
+    },
+    {
+      "epoch": 1.6851704996034893,
+      "grad_norm": 1.640625,
+      "learning_rate": 8.026518817327527e-06,
+      "loss": 0.8734,
+      "step": 9898
+    },
+    {
+      "epoch": 1.6853419636924791,
+      "grad_norm": 1.6875,
+      "learning_rate": 8.024748437676707e-06,
+      "loss": 0.8245,
+      "step": 9899
+    },
+    {
+      "epoch": 1.6855134277814692,
+      "grad_norm": 1.71875,
+      "learning_rate": 8.022978122446284e-06,
+      "loss": 0.8654,
+      "step": 9900
+    },
+    {
+      "epoch": 1.685684891870459,
+      "grad_norm": 1.703125,
+      "learning_rate": 8.021207871693984e-06,
+      "loss": 0.8636,
+      "step": 9901
+    },
+    {
+      "epoch": 1.6858563559594488,
+      "grad_norm": 1.6015625,
+      "learning_rate": 8.019437685477548e-06,
+      "loss": 0.9033,
+      "step": 9902
+    },
+    {
+      "epoch": 1.6860278200484387,
+      "grad_norm": 1.75,
+      "learning_rate": 8.017667563854706e-06,
+      "loss": 0.9517,
+      "step": 9903
+    },
+    {
+      "epoch": 1.6861992841374285,
+      "grad_norm": 1.6484375,
+      "learning_rate": 8.015897506883188e-06,
+      "loss": 0.8501,
+      "step": 9904
+    },
+    {
+      "epoch": 1.6863707482264183,
+      "grad_norm": 1.625,
+      "learning_rate": 8.014127514620726e-06,
+      "loss": 0.828,
+      "step": 9905
+    },
+    {
+      "epoch": 1.6865422123154081,
+      "grad_norm": 1.765625,
+      "learning_rate": 8.012357587125043e-06,
+      "loss": 0.8238,
+      "step": 9906
+    },
+    {
+      "epoch": 1.686713676404398,
+      "grad_norm": 1.734375,
+      "learning_rate": 8.010587724453865e-06,
+      "loss": 0.8278,
+      "step": 9907
+    },
+    {
+      "epoch": 1.6868851404933878,
+      "grad_norm": 1.78125,
+      "learning_rate": 8.008817926664912e-06,
+      "loss": 0.8806,
+      "step": 9908
+    },
+    {
+      "epoch": 1.6870566045823778,
+      "grad_norm": 1.6796875,
+      "learning_rate": 8.007048193815905e-06,
+      "loss": 0.8198,
+      "step": 9909
+    },
+    {
+      "epoch": 1.6872280686713677,
+      "grad_norm": 1.578125,
+      "learning_rate": 8.005278525964562e-06,
+      "loss": 0.7713,
+      "step": 9910
+    },
+    {
+      "epoch": 1.6873995327603575,
+      "grad_norm": 1.7109375,
+      "learning_rate": 8.003508923168596e-06,
+      "loss": 0.8293,
+      "step": 9911
+    },
+    {
+      "epoch": 1.6875709968493475,
+      "grad_norm": 1.7109375,
+      "learning_rate": 8.001739385485724e-06,
+      "loss": 0.8553,
+      "step": 9912
+    },
+    {
+      "epoch": 1.6877424609383374,
+      "grad_norm": 1.7109375,
+      "learning_rate": 7.999969912973656e-06,
+      "loss": 0.7869,
+      "step": 9913
+    },
+    {
+      "epoch": 1.6879139250273272,
+      "grad_norm": 1.65625,
+      "learning_rate": 7.998200505690097e-06,
+      "loss": 0.7464,
+      "step": 9914
+    },
+    {
+      "epoch": 1.688085389116317,
+      "grad_norm": 1.625,
+      "learning_rate": 7.99643116369276e-06,
+      "loss": 0.8032,
+      "step": 9915
+    },
+    {
+      "epoch": 1.6882568532053068,
+      "grad_norm": 1.6796875,
+      "learning_rate": 7.994661887039347e-06,
+      "loss": 0.8418,
+      "step": 9916
+    },
+    {
+      "epoch": 1.6884283172942967,
+      "grad_norm": 1.6875,
+      "learning_rate": 7.99289267578756e-06,
+      "loss": 0.7836,
+      "step": 9917
+    },
+    {
+      "epoch": 1.6885997813832865,
+      "grad_norm": 1.7734375,
+      "learning_rate": 7.991123529995102e-06,
+      "loss": 0.872,
+      "step": 9918
+    },
+    {
+      "epoch": 1.6887712454722763,
+      "grad_norm": 1.7265625,
+      "learning_rate": 7.989354449719671e-06,
+      "loss": 0.8624,
+      "step": 9919
+    },
+    {
+      "epoch": 1.6889427095612661,
+      "grad_norm": 1.7734375,
+      "learning_rate": 7.987585435018963e-06,
+      "loss": 0.8893,
+      "step": 9920
+    },
+    {
+      "epoch": 1.6891141736502562,
+      "grad_norm": 1.6953125,
+      "learning_rate": 7.98581648595067e-06,
+      "loss": 0.8371,
+      "step": 9921
+    },
+    {
+      "epoch": 1.689285637739246,
+      "grad_norm": 1.7578125,
+      "learning_rate": 7.984047602572486e-06,
+      "loss": 0.9117,
+      "step": 9922
+    },
+    {
+      "epoch": 1.6894571018282358,
+      "grad_norm": 1.734375,
+      "learning_rate": 7.982278784942106e-06,
+      "loss": 0.8182,
+      "step": 9923
+    },
+    {
+      "epoch": 1.6896285659172259,
+      "grad_norm": 1.78125,
+      "learning_rate": 7.980510033117208e-06,
+      "loss": 0.857,
+      "step": 9924
+    },
+    {
+      "epoch": 1.6898000300062157,
+      "grad_norm": 1.578125,
+      "learning_rate": 7.978741347155484e-06,
+      "loss": 0.8085,
+      "step": 9925
+    },
+    {
+      "epoch": 1.6899714940952055,
+      "grad_norm": 1.6484375,
+      "learning_rate": 7.976972727114615e-06,
+      "loss": 0.8537,
+      "step": 9926
+    },
+    {
+      "epoch": 1.6901429581841954,
+      "grad_norm": 1.78125,
+      "learning_rate": 7.975204173052284e-06,
+      "loss": 0.8811,
+      "step": 9927
+    },
+    {
+      "epoch": 1.6903144222731852,
+      "grad_norm": 1.7109375,
+      "learning_rate": 7.973435685026171e-06,
+      "loss": 0.8205,
+      "step": 9928
+    },
+    {
+      "epoch": 1.690485886362175,
+      "grad_norm": 1.671875,
+      "learning_rate": 7.97166726309395e-06,
+      "loss": 0.8305,
+      "step": 9929
+    },
+    {
+      "epoch": 1.6906573504511648,
+      "grad_norm": 1.6953125,
+      "learning_rate": 7.969898907313298e-06,
+      "loss": 0.8199,
+      "step": 9930
+    },
+    {
+      "epoch": 1.6908288145401547,
+      "grad_norm": 1.6953125,
+      "learning_rate": 7.968130617741887e-06,
+      "loss": 0.7937,
+      "step": 9931
+    },
+    {
+      "epoch": 1.6910002786291445,
+      "grad_norm": 1.71875,
+      "learning_rate": 7.966362394437389e-06,
+      "loss": 0.7968,
+      "step": 9932
+    },
+    {
+      "epoch": 1.6911717427181343,
+      "grad_norm": 1.6875,
+      "learning_rate": 7.964594237457469e-06,
+      "loss": 0.7697,
+      "step": 9933
+    },
+    {
+      "epoch": 1.6913432068071244,
+      "grad_norm": 1.6328125,
+      "learning_rate": 7.962826146859794e-06,
+      "loss": 0.8717,
+      "step": 9934
+    },
+    {
+      "epoch": 1.6915146708961142,
+      "grad_norm": 1.8671875,
+      "learning_rate": 7.961058122702037e-06,
+      "loss": 0.8174,
+      "step": 9935
+    },
+    {
+      "epoch": 1.691686134985104,
+      "grad_norm": 1.7890625,
+      "learning_rate": 7.959290165041848e-06,
+      "loss": 0.8491,
+      "step": 9936
+    },
+    {
+      "epoch": 1.691857599074094,
+      "grad_norm": 1.65625,
+      "learning_rate": 7.957522273936892e-06,
+      "loss": 0.8569,
+      "step": 9937
+    },
+    {
+      "epoch": 1.6920290631630839,
+      "grad_norm": 1.65625,
+      "learning_rate": 7.955754449444827e-06,
+      "loss": 0.8288,
+      "step": 9938
+    },
+    {
+      "epoch": 1.6922005272520737,
+      "grad_norm": 1.6640625,
+      "learning_rate": 7.953986691623305e-06,
+      "loss": 0.7834,
+      "step": 9939
+    },
+    {
+      "epoch": 1.6923719913410635,
+      "grad_norm": 1.734375,
+      "learning_rate": 7.952219000529982e-06,
+      "loss": 0.8607,
+      "step": 9940
+    },
+    {
+      "epoch": 1.6925434554300534,
+      "grad_norm": 1.703125,
+      "learning_rate": 7.950451376222508e-06,
+      "loss": 0.8376,
+      "step": 9941
+    },
+    {
+      "epoch": 1.6927149195190432,
+      "grad_norm": 1.7109375,
+      "learning_rate": 7.948683818758531e-06,
+      "loss": 0.7824,
+      "step": 9942
+    },
+    {
+      "epoch": 1.692886383608033,
+      "grad_norm": 1.6875,
+      "learning_rate": 7.946916328195701e-06,
+      "loss": 0.8443,
+      "step": 9943
+    },
+    {
+      "epoch": 1.6930578476970228,
+      "grad_norm": 1.6328125,
+      "learning_rate": 7.945148904591663e-06,
+      "loss": 0.8699,
+      "step": 9944
+    },
+    {
+      "epoch": 1.6932293117860127,
+      "grad_norm": 1.703125,
+      "learning_rate": 7.943381548004054e-06,
+      "loss": 0.8301,
+      "step": 9945
+    },
+    {
+      "epoch": 1.6934007758750027,
+      "grad_norm": 1.796875,
+      "learning_rate": 7.941614258490524e-06,
+      "loss": 0.9083,
+      "step": 9946
+    },
+    {
+      "epoch": 1.6935722399639925,
+      "grad_norm": 1.65625,
+      "learning_rate": 7.939847036108698e-06,
+      "loss": 0.7606,
+      "step": 9947
+    },
+    {
+      "epoch": 1.6937437040529824,
+      "grad_norm": 1.6796875,
+      "learning_rate": 7.938079880916219e-06,
+      "loss": 0.8062,
+      "step": 9948
+    },
+    {
+      "epoch": 1.6939151681419724,
+      "grad_norm": 1.640625,
+      "learning_rate": 7.936312792970719e-06,
+      "loss": 0.8749,
+      "step": 9949
+    },
+    {
+      "epoch": 1.6940866322309622,
+      "grad_norm": 1.625,
+      "learning_rate": 7.934545772329828e-06,
+      "loss": 0.8071,
+      "step": 9950
+    },
+    {
+      "epoch": 1.694258096319952,
+      "grad_norm": 1.703125,
+      "learning_rate": 7.93277881905118e-06,
+      "loss": 0.8501,
+      "step": 9951
+    },
+    {
+      "epoch": 1.6944295604089419,
+      "grad_norm": 1.8671875,
+      "learning_rate": 7.931011933192398e-06,
+      "loss": 0.8819,
+      "step": 9952
+    },
+    {
+      "epoch": 1.6946010244979317,
+      "grad_norm": 1.7109375,
+      "learning_rate": 7.929245114811108e-06,
+      "loss": 0.7806,
+      "step": 9953
+    },
+    {
+      "epoch": 1.6947724885869215,
+      "grad_norm": 1.609375,
+      "learning_rate": 7.927478363964933e-06,
+      "loss": 0.8344,
+      "step": 9954
+    },
+    {
+      "epoch": 1.6949439526759114,
+      "grad_norm": 1.8203125,
+      "learning_rate": 7.925711680711493e-06,
+      "loss": 0.9316,
+      "step": 9955
+    },
+    {
+      "epoch": 1.6951154167649012,
+      "grad_norm": 1.671875,
+      "learning_rate": 7.923945065108406e-06,
+      "loss": 0.9082,
+      "step": 9956
+    },
+    {
+      "epoch": 1.695286880853891,
+      "grad_norm": 1.7265625,
+      "learning_rate": 7.922178517213288e-06,
+      "loss": 0.8015,
+      "step": 9957
+    },
+    {
+      "epoch": 1.695458344942881,
+      "grad_norm": 1.734375,
+      "learning_rate": 7.920412037083757e-06,
+      "loss": 0.9053,
+      "step": 9958
+    },
+    {
+      "epoch": 1.6956298090318709,
+      "grad_norm": 1.6875,
+      "learning_rate": 7.918645624777415e-06,
+      "loss": 0.9377,
+      "step": 9959
+    },
+    {
+      "epoch": 1.6958012731208607,
+      "grad_norm": 1.7734375,
+      "learning_rate": 7.916879280351878e-06,
+      "loss": 0.8941,
+      "step": 9960
+    },
+    {
+      "epoch": 1.6959727372098508,
+      "grad_norm": 1.625,
+      "learning_rate": 7.915113003864753e-06,
+      "loss": 0.8981,
+      "step": 9961
+    },
+    {
+      "epoch": 1.6961442012988406,
+      "grad_norm": 1.734375,
+      "learning_rate": 7.913346795373643e-06,
+      "loss": 0.8599,
+      "step": 9962
+    },
+    {
+      "epoch": 1.6963156653878304,
+      "grad_norm": 1.6328125,
+      "learning_rate": 7.911580654936152e-06,
+      "loss": 0.7965,
+      "step": 9963
+    },
+    {
+      "epoch": 1.6964871294768202,
+      "grad_norm": 1.6640625,
+      "learning_rate": 7.90981458260988e-06,
+      "loss": 0.8015,
+      "step": 9964
+    },
+    {
+      "epoch": 1.69665859356581,
+      "grad_norm": 1.734375,
+      "learning_rate": 7.908048578452426e-06,
+      "loss": 0.9537,
+      "step": 9965
+    },
+    {
+      "epoch": 1.6968300576547999,
+      "grad_norm": 1.6953125,
+      "learning_rate": 7.906282642521384e-06,
+      "loss": 0.9249,
+      "step": 9966
+    },
+    {
+      "epoch": 1.6970015217437897,
+      "grad_norm": 1.6484375,
+      "learning_rate": 7.90451677487435e-06,
+      "loss": 0.8379,
+      "step": 9967
+    },
+    {
+      "epoch": 1.6971729858327795,
+      "grad_norm": 1.8203125,
+      "learning_rate": 7.902750975568914e-06,
+      "loss": 0.9602,
+      "step": 9968
+    },
+    {
+      "epoch": 1.6973444499217694,
+      "grad_norm": 1.6875,
+      "learning_rate": 7.90098524466267e-06,
+      "loss": 0.8189,
+      "step": 9969
+    },
+    {
+      "epoch": 1.6975159140107594,
+      "grad_norm": 1.7421875,
+      "learning_rate": 7.899219582213198e-06,
+      "loss": 0.8329,
+      "step": 9970
+    },
+    {
+      "epoch": 1.6976873780997492,
+      "grad_norm": 1.75,
+      "learning_rate": 7.897453988278087e-06,
+      "loss": 0.8815,
+      "step": 9971
+    },
+    {
+      "epoch": 1.697858842188739,
+      "grad_norm": 1.625,
+      "learning_rate": 7.895688462914919e-06,
+      "loss": 0.7879,
+      "step": 9972
+    },
+    {
+      "epoch": 1.698030306277729,
+      "grad_norm": 1.6328125,
+      "learning_rate": 7.893923006181274e-06,
+      "loss": 0.8663,
+      "step": 9973
+    },
+    {
+      "epoch": 1.698201770366719,
+      "grad_norm": 1.6484375,
+      "learning_rate": 7.892157618134729e-06,
+      "loss": 0.8129,
+      "step": 9974
+    },
+    {
+      "epoch": 1.6983732344557088,
+      "grad_norm": 1.6015625,
+      "learning_rate": 7.890392298832863e-06,
+      "loss": 0.8256,
+      "step": 9975
+    },
+    {
+      "epoch": 1.6985446985446986,
+      "grad_norm": 1.6796875,
+      "learning_rate": 7.888627048333248e-06,
+      "loss": 0.7792,
+      "step": 9976
+    },
+    {
+      "epoch": 1.6987161626336884,
+      "grad_norm": 1.7421875,
+      "learning_rate": 7.886861866693457e-06,
+      "loss": 0.8493,
+      "step": 9977
+    },
+    {
+      "epoch": 1.6988876267226782,
+      "grad_norm": 1.6796875,
+      "learning_rate": 7.885096753971056e-06,
+      "loss": 0.87,
+      "step": 9978
+    },
+    {
+      "epoch": 1.699059090811668,
+      "grad_norm": 1.75,
+      "learning_rate": 7.883331710223614e-06,
+      "loss": 0.9304,
+      "step": 9979
+    },
+    {
+      "epoch": 1.6992305549006579,
+      "grad_norm": 1.9453125,
+      "learning_rate": 7.881566735508696e-06,
+      "loss": 0.8853,
+      "step": 9980
+    },
+    {
+      "epoch": 1.6994020189896477,
+      "grad_norm": 1.7578125,
+      "learning_rate": 7.879801829883867e-06,
+      "loss": 0.8327,
+      "step": 9981
+    },
+    {
+      "epoch": 1.6995734830786378,
+      "grad_norm": 1.640625,
+      "learning_rate": 7.87803699340668e-06,
+      "loss": 0.7743,
+      "step": 9982
+    },
+    {
+      "epoch": 1.6997449471676276,
+      "grad_norm": 1.6875,
+      "learning_rate": 7.876272226134698e-06,
+      "loss": 0.8026,
+      "step": 9983
+    },
+    {
+      "epoch": 1.6999164112566174,
+      "grad_norm": 1.6953125,
+      "learning_rate": 7.874507528125476e-06,
+      "loss": 0.8312,
+      "step": 9984
+    },
+    {
+      "epoch": 1.7000878753456075,
+      "grad_norm": 1.65625,
+      "learning_rate": 7.872742899436568e-06,
+      "loss": 0.809,
+      "step": 9985
+    },
+    {
+      "epoch": 1.7002593394345973,
+      "grad_norm": 1.7421875,
+      "learning_rate": 7.870978340125524e-06,
+      "loss": 0.8874,
+      "step": 9986
+    },
+    {
+      "epoch": 1.700430803523587,
+      "grad_norm": 1.6171875,
+      "learning_rate": 7.869213850249895e-06,
+      "loss": 0.861,
+      "step": 9987
+    },
+    {
+      "epoch": 1.700602267612577,
+      "grad_norm": 1.75,
+      "learning_rate": 7.867449429867224e-06,
+      "loss": 0.9106,
+      "step": 9988
+    },
+    {
+      "epoch": 1.7007737317015668,
+      "grad_norm": 1.6328125,
+      "learning_rate": 7.865685079035058e-06,
+      "loss": 0.8218,
+      "step": 9989
+    },
+    {
+      "epoch": 1.7009451957905566,
+      "grad_norm": 1.65625,
+      "learning_rate": 7.863920797810938e-06,
+      "loss": 0.8162,
+      "step": 9990
+    },
+    {
+      "epoch": 1.7011166598795464,
+      "grad_norm": 1.6484375,
+      "learning_rate": 7.862156586252405e-06,
+      "loss": 0.8495,
+      "step": 9991
+    },
+    {
+      "epoch": 1.7012881239685362,
+      "grad_norm": 1.6796875,
+      "learning_rate": 7.860392444417001e-06,
+      "loss": 0.7733,
+      "step": 9992
+    },
+    {
+      "epoch": 1.701459588057526,
+      "grad_norm": 1.6484375,
+      "learning_rate": 7.858628372362253e-06,
+      "loss": 0.8968,
+      "step": 9993
+    },
+    {
+      "epoch": 1.701631052146516,
+      "grad_norm": 1.71875,
+      "learning_rate": 7.856864370145696e-06,
+      "loss": 0.7907,
+      "step": 9994
+    },
+    {
+      "epoch": 1.701802516235506,
+      "grad_norm": 1.75,
+      "learning_rate": 7.855100437824863e-06,
+      "loss": 0.8403,
+      "step": 9995
+    },
+    {
+      "epoch": 1.7019739803244958,
+      "grad_norm": 1.6953125,
+      "learning_rate": 7.853336575457281e-06,
+      "loss": 0.8308,
+      "step": 9996
+    },
+    {
+      "epoch": 1.7021454444134858,
+      "grad_norm": 1.78125,
+      "learning_rate": 7.851572783100475e-06,
+      "loss": 0.943,
+      "step": 9997
+    },
+    {
+      "epoch": 1.7023169085024756,
+      "grad_norm": 1.75,
+      "learning_rate": 7.849809060811973e-06,
+      "loss": 0.8719,
+      "step": 9998
+    },
+    {
+      "epoch": 1.7024883725914655,
+      "grad_norm": 1.71875,
+      "learning_rate": 7.848045408649295e-06,
+      "loss": 0.8588,
+      "step": 9999
+    },
+    {
+      "epoch": 1.7026598366804553,
+      "grad_norm": 1.6328125,
+      "learning_rate": 7.84628182666996e-06,
+      "loss": 0.82,
+      "step": 10000
+    },
+    {
+      "epoch": 1.702831300769445,
+      "grad_norm": 1.6875,
+      "learning_rate": 7.844518314931483e-06,
+      "loss": 0.901,
+      "step": 10001
+    },
+    {
+      "epoch": 1.703002764858435,
+      "grad_norm": 1.8125,
+      "learning_rate": 7.842754873491381e-06,
+      "loss": 0.7771,
+      "step": 10002
+    },
+    {
+      "epoch": 1.7031742289474248,
+      "grad_norm": 1.640625,
+      "learning_rate": 7.840991502407168e-06,
+      "loss": 0.8143,
+      "step": 10003
+    },
+    {
+      "epoch": 1.7033456930364146,
+      "grad_norm": 1.765625,
+      "learning_rate": 7.839228201736354e-06,
+      "loss": 0.8319,
+      "step": 10004
+    },
+    {
+      "epoch": 1.7035171571254044,
+      "grad_norm": 1.7421875,
+      "learning_rate": 7.837464971536442e-06,
+      "loss": 0.8533,
+      "step": 10005
+    },
+    {
+      "epoch": 1.7036886212143945,
+      "grad_norm": 1.6640625,
+      "learning_rate": 7.835701811864937e-06,
+      "loss": 0.8322,
+      "step": 10006
+    },
+    {
+      "epoch": 1.7038600853033843,
+      "grad_norm": 1.6875,
+      "learning_rate": 7.83393872277935e-06,
+      "loss": 0.9125,
+      "step": 10007
+    },
+    {
+      "epoch": 1.7040315493923741,
+      "grad_norm": 1.6875,
+      "learning_rate": 7.832175704337177e-06,
+      "loss": 0.8254,
+      "step": 10008
+    },
+    {
+      "epoch": 1.7042030134813642,
+      "grad_norm": 1.6484375,
+      "learning_rate": 7.830412756595917e-06,
+      "loss": 0.8034,
+      "step": 10009
+    },
+    {
+      "epoch": 1.704374477570354,
+      "grad_norm": 1.59375,
+      "learning_rate": 7.828649879613069e-06,
+      "loss": 0.7845,
+      "step": 10010
+    },
+    {
+      "epoch": 1.7045459416593438,
+      "grad_norm": 1.703125,
+      "learning_rate": 7.826887073446122e-06,
+      "loss": 0.8672,
+      "step": 10011
+    },
+    {
+      "epoch": 1.7047174057483336,
+      "grad_norm": 1.7265625,
+      "learning_rate": 7.825124338152574e-06,
+      "loss": 0.8432,
+      "step": 10012
+    },
+    {
+      "epoch": 1.7048888698373235,
+      "grad_norm": 1.609375,
+      "learning_rate": 7.823361673789909e-06,
+      "loss": 0.7993,
+      "step": 10013
+    },
+    {
+      "epoch": 1.7050603339263133,
+      "grad_norm": 1.65625,
+      "learning_rate": 7.821599080415618e-06,
+      "loss": 0.8738,
+      "step": 10014
+    },
+    {
+      "epoch": 1.7052317980153031,
+      "grad_norm": 1.71875,
+      "learning_rate": 7.81983655808718e-06,
+      "loss": 0.8296,
+      "step": 10015
+    },
+    {
+      "epoch": 1.705403262104293,
+      "grad_norm": 1.640625,
+      "learning_rate": 7.818074106862088e-06,
+      "loss": 0.8741,
+      "step": 10016
+    },
+    {
+      "epoch": 1.7055747261932828,
+      "grad_norm": 1.734375,
+      "learning_rate": 7.816311726797813e-06,
+      "loss": 0.8226,
+      "step": 10017
+    },
+    {
+      "epoch": 1.7057461902822728,
+      "grad_norm": 1.734375,
+      "learning_rate": 7.814549417951835e-06,
+      "loss": 0.8786,
+      "step": 10018
+    },
+    {
+      "epoch": 1.7059176543712626,
+      "grad_norm": 1.640625,
+      "learning_rate": 7.81278718038163e-06,
+      "loss": 0.8289,
+      "step": 10019
+    },
+    {
+      "epoch": 1.7060891184602525,
+      "grad_norm": 1.609375,
+      "learning_rate": 7.811025014144671e-06,
+      "loss": 0.7693,
+      "step": 10020
+    },
+    {
+      "epoch": 1.7062605825492423,
+      "grad_norm": 1.765625,
+      "learning_rate": 7.809262919298428e-06,
+      "loss": 0.8542,
+      "step": 10021
+    },
+    {
+      "epoch": 1.7064320466382323,
+      "grad_norm": 1.7890625,
+      "learning_rate": 7.807500895900373e-06,
+      "loss": 0.9365,
+      "step": 10022
+    },
+    {
+      "epoch": 1.7066035107272222,
+      "grad_norm": 1.671875,
+      "learning_rate": 7.805738944007968e-06,
+      "loss": 0.9101,
+      "step": 10023
+    },
+    {
+      "epoch": 1.706774974816212,
+      "grad_norm": 1.6796875,
+      "learning_rate": 7.803977063678682e-06,
+      "loss": 0.8732,
+      "step": 10024
+    },
+    {
+      "epoch": 1.7069464389052018,
+      "grad_norm": 1.7578125,
+      "learning_rate": 7.80221525496997e-06,
+      "loss": 0.8287,
+      "step": 10025
+    },
+    {
+      "epoch": 1.7071179029941916,
+      "grad_norm": 1.7578125,
+      "learning_rate": 7.800453517939298e-06,
+      "loss": 0.8571,
+      "step": 10026
+    },
+    {
+      "epoch": 1.7072893670831815,
+      "grad_norm": 1.7109375,
+      "learning_rate": 7.798691852644118e-06,
+      "loss": 0.8613,
+      "step": 10027
+    },
+    {
+      "epoch": 1.7074608311721713,
+      "grad_norm": 1.6953125,
+      "learning_rate": 7.796930259141885e-06,
+      "loss": 0.8497,
+      "step": 10028
+    },
+    {
+      "epoch": 1.7076322952611611,
+      "grad_norm": 1.625,
+      "learning_rate": 7.795168737490054e-06,
+      "loss": 0.7839,
+      "step": 10029
+    },
+    {
+      "epoch": 1.707803759350151,
+      "grad_norm": 1.6484375,
+      "learning_rate": 7.793407287746072e-06,
+      "loss": 0.9061,
+      "step": 10030
+    },
+    {
+      "epoch": 1.707975223439141,
+      "grad_norm": 1.6796875,
+      "learning_rate": 7.791645909967387e-06,
+      "loss": 0.8451,
+      "step": 10031
+    },
+    {
+      "epoch": 1.7081466875281308,
+      "grad_norm": 1.7109375,
+      "learning_rate": 7.789884604211447e-06,
+      "loss": 0.8251,
+      "step": 10032
+    },
+    {
+      "epoch": 1.7083181516171206,
+      "grad_norm": 1.6953125,
+      "learning_rate": 7.78812337053569e-06,
+      "loss": 0.883,
+      "step": 10033
+    },
+    {
+      "epoch": 1.7084896157061107,
+      "grad_norm": 1.6015625,
+      "learning_rate": 7.78636220899756e-06,
+      "loss": 0.7842,
+      "step": 10034
+    },
+    {
+      "epoch": 1.7086610797951005,
+      "grad_norm": 1.703125,
+      "learning_rate": 7.784601119654494e-06,
+      "loss": 0.8951,
+      "step": 10035
+    },
+    {
+      "epoch": 1.7088325438840903,
+      "grad_norm": 1.703125,
+      "learning_rate": 7.782840102563925e-06,
+      "loss": 0.9254,
+      "step": 10036
+    },
+    {
+      "epoch": 1.7090040079730802,
+      "grad_norm": 1.6953125,
+      "learning_rate": 7.781079157783292e-06,
+      "loss": 0.8714,
+      "step": 10037
+    },
+    {
+      "epoch": 1.70917547206207,
+      "grad_norm": 1.6640625,
+      "learning_rate": 7.779318285370024e-06,
+      "loss": 0.7217,
+      "step": 10038
+    },
+    {
+      "epoch": 1.7093469361510598,
+      "grad_norm": 1.734375,
+      "learning_rate": 7.777557485381553e-06,
+      "loss": 0.8525,
+      "step": 10039
+    },
+    {
+      "epoch": 1.7095184002400496,
+      "grad_norm": 1.71875,
+      "learning_rate": 7.775796757875298e-06,
+      "loss": 0.8348,
+      "step": 10040
+    },
+    {
+      "epoch": 1.7096898643290395,
+      "grad_norm": 1.6875,
+      "learning_rate": 7.774036102908685e-06,
+      "loss": 0.8365,
+      "step": 10041
+    },
+    {
+      "epoch": 1.7098613284180293,
+      "grad_norm": 1.6953125,
+      "learning_rate": 7.772275520539136e-06,
+      "loss": 0.8373,
+      "step": 10042
+    },
+    {
+      "epoch": 1.7100327925070193,
+      "grad_norm": 1.640625,
+      "learning_rate": 7.770515010824074e-06,
+      "loss": 0.8351,
+      "step": 10043
+    },
+    {
+      "epoch": 1.7102042565960092,
+      "grad_norm": 1.6640625,
+      "learning_rate": 7.768754573820908e-06,
+      "loss": 0.8265,
+      "step": 10044
+    },
+    {
+      "epoch": 1.710375720684999,
+      "grad_norm": 1.796875,
+      "learning_rate": 7.766994209587062e-06,
+      "loss": 0.9264,
+      "step": 10045
+    },
+    {
+      "epoch": 1.710547184773989,
+      "grad_norm": 1.6875,
+      "learning_rate": 7.765233918179942e-06,
+      "loss": 0.8793,
+      "step": 10046
+    },
+    {
+      "epoch": 1.7107186488629789,
+      "grad_norm": 1.7109375,
+      "learning_rate": 7.763473699656959e-06,
+      "loss": 0.9455,
+      "step": 10047
+    },
+    {
+      "epoch": 1.7108901129519687,
+      "grad_norm": 1.75,
+      "learning_rate": 7.761713554075521e-06,
+      "loss": 0.8932,
+      "step": 10048
+    },
+    {
+      "epoch": 1.7110615770409585,
+      "grad_norm": 1.7421875,
+      "learning_rate": 7.759953481493033e-06,
+      "loss": 0.8766,
+      "step": 10049
+    },
+    {
+      "epoch": 1.7112330411299483,
+      "grad_norm": 1.6875,
+      "learning_rate": 7.758193481966901e-06,
+      "loss": 0.8285,
+      "step": 10050
+    },
+    {
+      "epoch": 1.7114045052189382,
+      "grad_norm": 1.6953125,
+      "learning_rate": 7.756433555554518e-06,
+      "loss": 0.8557,
+      "step": 10051
+    },
+    {
+      "epoch": 1.711575969307928,
+      "grad_norm": 1.625,
+      "learning_rate": 7.754673702313284e-06,
+      "loss": 0.8181,
+      "step": 10052
+    },
+    {
+      "epoch": 1.7117474333969178,
+      "grad_norm": 1.703125,
+      "learning_rate": 7.752913922300595e-06,
+      "loss": 0.8187,
+      "step": 10053
+    },
+    {
+      "epoch": 1.7119188974859076,
+      "grad_norm": 1.65625,
+      "learning_rate": 7.751154215573845e-06,
+      "loss": 0.8839,
+      "step": 10054
+    },
+    {
+      "epoch": 1.7120903615748977,
+      "grad_norm": 1.671875,
+      "learning_rate": 7.749394582190426e-06,
+      "loss": 0.8244,
+      "step": 10055
+    },
+    {
+      "epoch": 1.7122618256638875,
+      "grad_norm": 1.671875,
+      "learning_rate": 7.747635022207724e-06,
+      "loss": 0.8911,
+      "step": 10056
+    },
+    {
+      "epoch": 1.7124332897528773,
+      "grad_norm": 1.7265625,
+      "learning_rate": 7.745875535683126e-06,
+      "loss": 0.8736,
+      "step": 10057
+    },
+    {
+      "epoch": 1.7126047538418674,
+      "grad_norm": 1.75,
+      "learning_rate": 7.744116122674015e-06,
+      "loss": 0.8851,
+      "step": 10058
+    },
+    {
+      "epoch": 1.7127762179308572,
+      "grad_norm": 1.6953125,
+      "learning_rate": 7.742356783237772e-06,
+      "loss": 0.8383,
+      "step": 10059
+    },
+    {
+      "epoch": 1.712947682019847,
+      "grad_norm": 1.6640625,
+      "learning_rate": 7.740597517431775e-06,
+      "loss": 0.8672,
+      "step": 10060
+    },
+    {
+      "epoch": 1.7131191461088369,
+      "grad_norm": 1.734375,
+      "learning_rate": 7.738838325313402e-06,
+      "loss": 0.8921,
+      "step": 10061
+    },
+    {
+      "epoch": 1.7132906101978267,
+      "grad_norm": 1.7421875,
+      "learning_rate": 7.737079206940027e-06,
+      "loss": 0.8335,
+      "step": 10062
+    },
+    {
+      "epoch": 1.7134620742868165,
+      "grad_norm": 1.671875,
+      "learning_rate": 7.73532016236902e-06,
+      "loss": 0.8776,
+      "step": 10063
+    },
+    {
+      "epoch": 1.7136335383758063,
+      "grad_norm": 1.703125,
+      "learning_rate": 7.733561191657748e-06,
+      "loss": 0.8359,
+      "step": 10064
+    },
+    {
+      "epoch": 1.7138050024647962,
+      "grad_norm": 1.78125,
+      "learning_rate": 7.731802294863583e-06,
+      "loss": 0.883,
+      "step": 10065
+    },
+    {
+      "epoch": 1.713976466553786,
+      "grad_norm": 1.765625,
+      "learning_rate": 7.730043472043884e-06,
+      "loss": 0.8665,
+      "step": 10066
+    },
+    {
+      "epoch": 1.714147930642776,
+      "grad_norm": 1.6953125,
+      "learning_rate": 7.728284723256017e-06,
+      "loss": 0.8383,
+      "step": 10067
+    },
+    {
+      "epoch": 1.7143193947317659,
+      "grad_norm": 1.6953125,
+      "learning_rate": 7.726526048557338e-06,
+      "loss": 0.8075,
+      "step": 10068
+    },
+    {
+      "epoch": 1.7144908588207557,
+      "grad_norm": 1.78125,
+      "learning_rate": 7.724767448005208e-06,
+      "loss": 0.8819,
+      "step": 10069
+    },
+    {
+      "epoch": 1.7146623229097457,
+      "grad_norm": 1.7421875,
+      "learning_rate": 7.723008921656977e-06,
+      "loss": 0.7742,
+      "step": 10070
+    },
+    {
+      "epoch": 1.7148337869987356,
+      "grad_norm": 1.640625,
+      "learning_rate": 7.72125046957e-06,
+      "loss": 0.8344,
+      "step": 10071
+    },
+    {
+      "epoch": 1.7150052510877254,
+      "grad_norm": 1.6875,
+      "learning_rate": 7.719492091801628e-06,
+      "loss": 0.8342,
+      "step": 10072
+    },
+    {
+      "epoch": 1.7151767151767152,
+      "grad_norm": 1.734375,
+      "learning_rate": 7.717733788409207e-06,
+      "loss": 0.8667,
+      "step": 10073
+    },
+    {
+      "epoch": 1.715348179265705,
+      "grad_norm": 1.71875,
+      "learning_rate": 7.71597555945008e-06,
+      "loss": 0.8732,
+      "step": 10074
+    },
+    {
+      "epoch": 1.7155196433546949,
+      "grad_norm": 1.671875,
+      "learning_rate": 7.71421740498159e-06,
+      "loss": 0.915,
+      "step": 10075
+    },
+    {
+      "epoch": 1.7156911074436847,
+      "grad_norm": 1.75,
+      "learning_rate": 7.712459325061078e-06,
+      "loss": 0.8567,
+      "step": 10076
+    },
+    {
+      "epoch": 1.7158625715326745,
+      "grad_norm": 1.8984375,
+      "learning_rate": 7.710701319745881e-06,
+      "loss": 0.8153,
+      "step": 10077
+    },
+    {
+      "epoch": 1.7160340356216643,
+      "grad_norm": 1.6875,
+      "learning_rate": 7.708943389093337e-06,
+      "loss": 0.8616,
+      "step": 10078
+    },
+    {
+      "epoch": 1.7162054997106544,
+      "grad_norm": 1.734375,
+      "learning_rate": 7.707185533160774e-06,
+      "loss": 0.8309,
+      "step": 10079
+    },
+    {
+      "epoch": 1.7163769637996442,
+      "grad_norm": 1.671875,
+      "learning_rate": 7.705427752005525e-06,
+      "loss": 0.8271,
+      "step": 10080
+    },
+    {
+      "epoch": 1.716548427888634,
+      "grad_norm": 1.703125,
+      "learning_rate": 7.703670045684919e-06,
+      "loss": 0.9102,
+      "step": 10081
+    },
+    {
+      "epoch": 1.716719891977624,
+      "grad_norm": 1.734375,
+      "learning_rate": 7.701912414256279e-06,
+      "loss": 0.9635,
+      "step": 10082
+    },
+    {
+      "epoch": 1.716891356066614,
+      "grad_norm": 1.6484375,
+      "learning_rate": 7.700154857776925e-06,
+      "loss": 0.8498,
+      "step": 10083
+    },
+    {
+      "epoch": 1.7170628201556037,
+      "grad_norm": 1.671875,
+      "learning_rate": 7.698397376304186e-06,
+      "loss": 0.8869,
+      "step": 10084
+    },
+    {
+      "epoch": 1.7172342842445936,
+      "grad_norm": 1.828125,
+      "learning_rate": 7.69663996989538e-06,
+      "loss": 0.8821,
+      "step": 10085
+    },
+    {
+      "epoch": 1.7174057483335834,
+      "grad_norm": 1.765625,
+      "learning_rate": 7.694882638607813e-06,
+      "loss": 0.9002,
+      "step": 10086
+    },
+    {
+      "epoch": 1.7175772124225732,
+      "grad_norm": 1.6953125,
+      "learning_rate": 7.693125382498804e-06,
+      "loss": 0.8881,
+      "step": 10087
+    },
+    {
+      "epoch": 1.717748676511563,
+      "grad_norm": 1.8359375,
+      "learning_rate": 7.691368201625662e-06,
+      "loss": 0.8636,
+      "step": 10088
+    },
+    {
+      "epoch": 1.7179201406005529,
+      "grad_norm": 1.703125,
+      "learning_rate": 7.689611096045698e-06,
+      "loss": 0.8352,
+      "step": 10089
+    },
+    {
+      "epoch": 1.7180916046895427,
+      "grad_norm": 1.71875,
+      "learning_rate": 7.687854065816216e-06,
+      "loss": 0.9081,
+      "step": 10090
+    },
+    {
+      "epoch": 1.7182630687785327,
+      "grad_norm": 1.65625,
+      "learning_rate": 7.686097110994516e-06,
+      "loss": 0.8152,
+      "step": 10091
+    },
+    {
+      "epoch": 1.7184345328675226,
+      "grad_norm": 1.609375,
+      "learning_rate": 7.684340231637905e-06,
+      "loss": 0.8166,
+      "step": 10092
+    },
+    {
+      "epoch": 1.7186059969565124,
+      "grad_norm": 1.59375,
+      "learning_rate": 7.68258342780368e-06,
+      "loss": 0.8214,
+      "step": 10093
+    },
+    {
+      "epoch": 1.7187774610455024,
+      "grad_norm": 1.71875,
+      "learning_rate": 7.680826699549136e-06,
+      "loss": 0.9223,
+      "step": 10094
+    },
+    {
+      "epoch": 1.7189489251344923,
+      "grad_norm": 1.6875,
+      "learning_rate": 7.679070046931567e-06,
+      "loss": 0.869,
+      "step": 10095
+    },
+    {
+      "epoch": 1.719120389223482,
+      "grad_norm": 1.6328125,
+      "learning_rate": 7.677313470008268e-06,
+      "loss": 0.8313,
+      "step": 10096
+    },
+    {
+      "epoch": 1.719291853312472,
+      "grad_norm": 1.6640625,
+      "learning_rate": 7.675556968836517e-06,
+      "loss": 0.8957,
+      "step": 10097
+    },
+    {
+      "epoch": 1.7194633174014617,
+      "grad_norm": 1.7421875,
+      "learning_rate": 7.673800543473608e-06,
+      "loss": 0.7927,
+      "step": 10098
+    },
+    {
+      "epoch": 1.7196347814904516,
+      "grad_norm": 1.625,
+      "learning_rate": 7.672044193976822e-06,
+      "loss": 0.8355,
+      "step": 10099
+    },
+    {
+      "epoch": 1.7198062455794414,
+      "grad_norm": 1.703125,
+      "learning_rate": 7.670287920403439e-06,
+      "loss": 0.819,
+      "step": 10100
+    },
+    {
+      "epoch": 1.7199777096684312,
+      "grad_norm": 1.8125,
+      "learning_rate": 7.668531722810742e-06,
+      "loss": 0.9703,
+      "step": 10101
+    },
+    {
+      "epoch": 1.720149173757421,
+      "grad_norm": 1.875,
+      "learning_rate": 7.666775601256006e-06,
+      "loss": 0.882,
+      "step": 10102
+    },
+    {
+      "epoch": 1.720320637846411,
+      "grad_norm": 1.671875,
+      "learning_rate": 7.665019555796502e-06,
+      "loss": 0.8179,
+      "step": 10103
+    },
+    {
+      "epoch": 1.720492101935401,
+      "grad_norm": 1.78125,
+      "learning_rate": 7.663263586489504e-06,
+      "loss": 0.9177,
+      "step": 10104
+    },
+    {
+      "epoch": 1.7206635660243907,
+      "grad_norm": 1.640625,
+      "learning_rate": 7.66150769339228e-06,
+      "loss": 0.7969,
+      "step": 10105
+    },
+    {
+      "epoch": 1.7208350301133808,
+      "grad_norm": 1.71875,
+      "learning_rate": 7.659751876562096e-06,
+      "loss": 0.8631,
+      "step": 10106
+    },
+    {
+      "epoch": 1.7210064942023706,
+      "grad_norm": 1.6171875,
+      "learning_rate": 7.657996136056216e-06,
+      "loss": 0.862,
+      "step": 10107
+    },
+    {
+      "epoch": 1.7211779582913604,
+      "grad_norm": 1.8046875,
+      "learning_rate": 7.656240471931904e-06,
+      "loss": 0.9143,
+      "step": 10108
+    },
+    {
+      "epoch": 1.7213494223803503,
+      "grad_norm": 1.640625,
+      "learning_rate": 7.654484884246412e-06,
+      "loss": 0.8633,
+      "step": 10109
+    },
+    {
+      "epoch": 1.72152088646934,
+      "grad_norm": 1.6328125,
+      "learning_rate": 7.652729373057001e-06,
+      "loss": 0.8728,
+      "step": 10110
+    },
+    {
+      "epoch": 1.72169235055833,
+      "grad_norm": 1.671875,
+      "learning_rate": 7.650973938420924e-06,
+      "loss": 0.7864,
+      "step": 10111
+    },
+    {
+      "epoch": 1.7218638146473197,
+      "grad_norm": 1.7265625,
+      "learning_rate": 7.649218580395433e-06,
+      "loss": 0.8282,
+      "step": 10112
+    },
+    {
+      "epoch": 1.7220352787363096,
+      "grad_norm": 1.7578125,
+      "learning_rate": 7.647463299037777e-06,
+      "loss": 0.8676,
+      "step": 10113
+    },
+    {
+      "epoch": 1.7222067428252994,
+      "grad_norm": 1.6953125,
+      "learning_rate": 7.6457080944052e-06,
+      "loss": 0.8708,
+      "step": 10114
+    },
+    {
+      "epoch": 1.7223782069142892,
+      "grad_norm": 1.703125,
+      "learning_rate": 7.643952966554948e-06,
+      "loss": 0.8541,
+      "step": 10115
+    },
+    {
+      "epoch": 1.7225496710032793,
+      "grad_norm": 1.703125,
+      "learning_rate": 7.642197915544263e-06,
+      "loss": 0.7929,
+      "step": 10116
+    },
+    {
+      "epoch": 1.722721135092269,
+      "grad_norm": 1.625,
+      "learning_rate": 7.640442941430382e-06,
+      "loss": 0.8458,
+      "step": 10117
+    },
+    {
+      "epoch": 1.722892599181259,
+      "grad_norm": 1.71875,
+      "learning_rate": 7.638688044270542e-06,
+      "loss": 0.8399,
+      "step": 10118
+    },
+    {
+      "epoch": 1.723064063270249,
+      "grad_norm": 1.6640625,
+      "learning_rate": 7.636933224121977e-06,
+      "loss": 0.8385,
+      "step": 10119
+    },
+    {
+      "epoch": 1.7232355273592388,
+      "grad_norm": 1.671875,
+      "learning_rate": 7.635178481041917e-06,
+      "loss": 0.8743,
+      "step": 10120
+    },
+    {
+      "epoch": 1.7234069914482286,
+      "grad_norm": 1.625,
+      "learning_rate": 7.633423815087593e-06,
+      "loss": 0.8821,
+      "step": 10121
+    },
+    {
+      "epoch": 1.7235784555372184,
+      "grad_norm": 1.6640625,
+      "learning_rate": 7.63166922631623e-06,
+      "loss": 0.7966,
+      "step": 10122
+    },
+    {
+      "epoch": 1.7237499196262083,
+      "grad_norm": 1.71875,
+      "learning_rate": 7.62991471478505e-06,
+      "loss": 0.8014,
+      "step": 10123
+    },
+    {
+      "epoch": 1.723921383715198,
+      "grad_norm": 1.6171875,
+      "learning_rate": 7.628160280551278e-06,
+      "loss": 0.8135,
+      "step": 10124
+    },
+    {
+      "epoch": 1.724092847804188,
+      "grad_norm": 1.65625,
+      "learning_rate": 7.62640592367213e-06,
+      "loss": 0.8701,
+      "step": 10125
+    },
+    {
+      "epoch": 1.7242643118931777,
+      "grad_norm": 1.671875,
+      "learning_rate": 7.624651644204823e-06,
+      "loss": 0.8518,
+      "step": 10126
+    },
+    {
+      "epoch": 1.7244357759821676,
+      "grad_norm": 1.6328125,
+      "learning_rate": 7.62289744220657e-06,
+      "loss": 0.8435,
+      "step": 10127
+    },
+    {
+      "epoch": 1.7246072400711576,
+      "grad_norm": 1.6640625,
+      "learning_rate": 7.621143317734584e-06,
+      "loss": 0.8237,
+      "step": 10128
+    },
+    {
+      "epoch": 1.7247787041601474,
+      "grad_norm": 1.734375,
+      "learning_rate": 7.619389270846071e-06,
+      "loss": 0.91,
+      "step": 10129
+    },
+    {
+      "epoch": 1.7249501682491373,
+      "grad_norm": 1.6875,
+      "learning_rate": 7.617635301598237e-06,
+      "loss": 0.8793,
+      "step": 10130
+    },
+    {
+      "epoch": 1.7251216323381273,
+      "grad_norm": 1.7734375,
+      "learning_rate": 7.615881410048294e-06,
+      "loss": 0.9091,
+      "step": 10131
+    },
+    {
+      "epoch": 1.7252930964271171,
+      "grad_norm": 1.734375,
+      "learning_rate": 7.614127596253431e-06,
+      "loss": 0.8372,
+      "step": 10132
+    },
+    {
+      "epoch": 1.725464560516107,
+      "grad_norm": 1.71875,
+      "learning_rate": 7.612373860270852e-06,
+      "loss": 0.8641,
+      "step": 10133
+    },
+    {
+      "epoch": 1.7256360246050968,
+      "grad_norm": 1.65625,
+      "learning_rate": 7.610620202157751e-06,
+      "loss": 0.8591,
+      "step": 10134
+    },
+    {
+      "epoch": 1.7258074886940866,
+      "grad_norm": 1.7109375,
+      "learning_rate": 7.608866621971325e-06,
+      "loss": 0.8311,
+      "step": 10135
+    },
+    {
+      "epoch": 1.7259789527830764,
+      "grad_norm": 1.625,
+      "learning_rate": 7.6071131197687606e-06,
+      "loss": 0.8582,
+      "step": 10136
+    },
+    {
+      "epoch": 1.7261504168720663,
+      "grad_norm": 1.5546875,
+      "learning_rate": 7.605359695607248e-06,
+      "loss": 0.8795,
+      "step": 10137
+    },
+    {
+      "epoch": 1.726321880961056,
+      "grad_norm": 1.7421875,
+      "learning_rate": 7.603606349543973e-06,
+      "loss": 0.9352,
+      "step": 10138
+    },
+    {
+      "epoch": 1.726493345050046,
+      "grad_norm": 1.7734375,
+      "learning_rate": 7.601853081636119e-06,
+      "loss": 0.9355,
+      "step": 10139
+    },
+    {
+      "epoch": 1.726664809139036,
+      "grad_norm": 1.71875,
+      "learning_rate": 7.600099891940869e-06,
+      "loss": 0.8794,
+      "step": 10140
+    },
+    {
+      "epoch": 1.7268362732280258,
+      "grad_norm": 1.625,
+      "learning_rate": 7.598346780515396e-06,
+      "loss": 0.8277,
+      "step": 10141
+    },
+    {
+      "epoch": 1.7270077373170156,
+      "grad_norm": 1.71875,
+      "learning_rate": 7.596593747416885e-06,
+      "loss": 0.8309,
+      "step": 10142
+    },
+    {
+      "epoch": 1.7271792014060057,
+      "grad_norm": 1.703125,
+      "learning_rate": 7.5948407927024955e-06,
+      "loss": 0.8031,
+      "step": 10143
+    },
+    {
+      "epoch": 1.7273506654949955,
+      "grad_norm": 1.75,
+      "learning_rate": 7.593087916429407e-06,
+      "loss": 0.851,
+      "step": 10144
+    },
+    {
+      "epoch": 1.7275221295839853,
+      "grad_norm": 1.6484375,
+      "learning_rate": 7.591335118654784e-06,
+      "loss": 0.8321,
+      "step": 10145
+    },
+    {
+      "epoch": 1.7276935936729751,
+      "grad_norm": 1.765625,
+      "learning_rate": 7.589582399435791e-06,
+      "loss": 0.8972,
+      "step": 10146
+    },
+    {
+      "epoch": 1.727865057761965,
+      "grad_norm": 1.7109375,
+      "learning_rate": 7.587829758829594e-06,
+      "loss": 0.805,
+      "step": 10147
+    },
+    {
+      "epoch": 1.7280365218509548,
+      "grad_norm": 1.6328125,
+      "learning_rate": 7.586077196893353e-06,
+      "loss": 0.8274,
+      "step": 10148
+    },
+    {
+      "epoch": 1.7282079859399446,
+      "grad_norm": 1.734375,
+      "learning_rate": 7.5843247136842245e-06,
+      "loss": 0.926,
+      "step": 10149
+    },
+    {
+      "epoch": 1.7283794500289344,
+      "grad_norm": 1.6640625,
+      "learning_rate": 7.582572309259364e-06,
+      "loss": 0.8003,
+      "step": 10150
+    },
+    {
+      "epoch": 1.7285509141179243,
+      "grad_norm": 1.6796875,
+      "learning_rate": 7.5808199836759235e-06,
+      "loss": 0.8239,
+      "step": 10151
+    },
+    {
+      "epoch": 1.7287223782069143,
+      "grad_norm": 1.8203125,
+      "learning_rate": 7.579067736991053e-06,
+      "loss": 0.8298,
+      "step": 10152
+    },
+    {
+      "epoch": 1.7288938422959041,
+      "grad_norm": 1.6796875,
+      "learning_rate": 7.5773155692618995e-06,
+      "loss": 0.8657,
+      "step": 10153
+    },
+    {
+      "epoch": 1.729065306384894,
+      "grad_norm": 1.6015625,
+      "learning_rate": 7.575563480545613e-06,
+      "loss": 0.8066,
+      "step": 10154
+    },
+    {
+      "epoch": 1.729236770473884,
+      "grad_norm": 1.7109375,
+      "learning_rate": 7.573811470899325e-06,
+      "loss": 0.8659,
+      "step": 10155
+    },
+    {
+      "epoch": 1.7294082345628738,
+      "grad_norm": 1.703125,
+      "learning_rate": 7.572059540380182e-06,
+      "loss": 0.9177,
+      "step": 10156
+    },
+    {
+      "epoch": 1.7295796986518637,
+      "grad_norm": 1.6953125,
+      "learning_rate": 7.570307689045322e-06,
+      "loss": 0.9513,
+      "step": 10157
+    },
+    {
+      "epoch": 1.7297511627408535,
+      "grad_norm": 1.6015625,
+      "learning_rate": 7.568555916951877e-06,
+      "loss": 0.787,
+      "step": 10158
+    },
+    {
+      "epoch": 1.7299226268298433,
+      "grad_norm": 1.703125,
+      "learning_rate": 7.566804224156978e-06,
+      "loss": 0.8231,
+      "step": 10159
+    },
+    {
+      "epoch": 1.7300940909188331,
+      "grad_norm": 1.6484375,
+      "learning_rate": 7.565052610717757e-06,
+      "loss": 0.8836,
+      "step": 10160
+    },
+    {
+      "epoch": 1.730265555007823,
+      "grad_norm": 1.6171875,
+      "learning_rate": 7.563301076691339e-06,
+      "loss": 0.8266,
+      "step": 10161
+    },
+    {
+      "epoch": 1.7304370190968128,
+      "grad_norm": 1.75,
+      "learning_rate": 7.5615496221348495e-06,
+      "loss": 0.8571,
+      "step": 10162
+    },
+    {
+      "epoch": 1.7306084831858026,
+      "grad_norm": 1.75,
+      "learning_rate": 7.559798247105409e-06,
+      "loss": 0.8509,
+      "step": 10163
+    },
+    {
+      "epoch": 1.7307799472747927,
+      "grad_norm": 1.78125,
+      "learning_rate": 7.558046951660136e-06,
+      "loss": 0.8736,
+      "step": 10164
+    },
+    {
+      "epoch": 1.7309514113637825,
+      "grad_norm": 1.765625,
+      "learning_rate": 7.55629573585615e-06,
+      "loss": 0.8781,
+      "step": 10165
+    },
+    {
+      "epoch": 1.7311228754527723,
+      "grad_norm": 1.6640625,
+      "learning_rate": 7.554544599750559e-06,
+      "loss": 0.8972,
+      "step": 10166
+    },
+    {
+      "epoch": 1.7312943395417624,
+      "grad_norm": 1.734375,
+      "learning_rate": 7.55279354340048e-06,
+      "loss": 0.8911,
+      "step": 10167
+    },
+    {
+      "epoch": 1.7314658036307522,
+      "grad_norm": 1.734375,
+      "learning_rate": 7.551042566863017e-06,
+      "loss": 0.8434,
+      "step": 10168
+    },
+    {
+      "epoch": 1.731637267719742,
+      "grad_norm": 1.703125,
+      "learning_rate": 7.549291670195278e-06,
+      "loss": 0.8142,
+      "step": 10169
+    },
+    {
+      "epoch": 1.7318087318087318,
+      "grad_norm": 1.7265625,
+      "learning_rate": 7.547540853454366e-06,
+      "loss": 0.8171,
+      "step": 10170
+    },
+    {
+      "epoch": 1.7319801958977217,
+      "grad_norm": 1.6484375,
+      "learning_rate": 7.5457901166973825e-06,
+      "loss": 0.8392,
+      "step": 10171
+    },
+    {
+      "epoch": 1.7321516599867115,
+      "grad_norm": 1.75,
+      "learning_rate": 7.544039459981425e-06,
+      "loss": 0.8475,
+      "step": 10172
+    },
+    {
+      "epoch": 1.7323231240757013,
+      "grad_norm": 1.7109375,
+      "learning_rate": 7.542288883363587e-06,
+      "loss": 0.8659,
+      "step": 10173
+    },
+    {
+      "epoch": 1.7324945881646912,
+      "grad_norm": 1.7734375,
+      "learning_rate": 7.540538386900966e-06,
+      "loss": 0.9016,
+      "step": 10174
+    },
+    {
+      "epoch": 1.732666052253681,
+      "grad_norm": 1.703125,
+      "learning_rate": 7.538787970650648e-06,
+      "loss": 0.9046,
+      "step": 10175
+    },
+    {
+      "epoch": 1.732837516342671,
+      "grad_norm": 1.6328125,
+      "learning_rate": 7.5370376346697235e-06,
+      "loss": 0.8584,
+      "step": 10176
+    },
+    {
+      "epoch": 1.7330089804316609,
+      "grad_norm": 1.71875,
+      "learning_rate": 7.535287379015278e-06,
+      "loss": 0.8298,
+      "step": 10177
+    },
+    {
+      "epoch": 1.7331804445206507,
+      "grad_norm": 1.703125,
+      "learning_rate": 7.53353720374439e-06,
+      "loss": 0.845,
+      "step": 10178
+    },
+    {
+      "epoch": 1.7333519086096407,
+      "grad_norm": 1.6796875,
+      "learning_rate": 7.5317871089141415e-06,
+      "loss": 0.7766,
+      "step": 10179
+    },
+    {
+      "epoch": 1.7335233726986305,
+      "grad_norm": 1.640625,
+      "learning_rate": 7.53003709458161e-06,
+      "loss": 0.7899,
+      "step": 10180
+    },
+    {
+      "epoch": 1.7336948367876204,
+      "grad_norm": 1.625,
+      "learning_rate": 7.528287160803871e-06,
+      "loss": 0.8287,
+      "step": 10181
+    },
+    {
+      "epoch": 1.7338663008766102,
+      "grad_norm": 1.6796875,
+      "learning_rate": 7.5265373076379934e-06,
+      "loss": 0.8382,
+      "step": 10182
+    },
+    {
+      "epoch": 1.7340377649656,
+      "grad_norm": 1.7109375,
+      "learning_rate": 7.524787535141049e-06,
+      "loss": 0.8667,
+      "step": 10183
+    },
+    {
+      "epoch": 1.7342092290545899,
+      "grad_norm": 1.703125,
+      "learning_rate": 7.523037843370104e-06,
+      "loss": 0.7994,
+      "step": 10184
+    },
+    {
+      "epoch": 1.7343806931435797,
+      "grad_norm": 1.671875,
+      "learning_rate": 7.521288232382221e-06,
+      "loss": 0.8789,
+      "step": 10185
+    },
+    {
+      "epoch": 1.7345521572325695,
+      "grad_norm": 1.6953125,
+      "learning_rate": 7.519538702234464e-06,
+      "loss": 0.8457,
+      "step": 10186
+    },
+    {
+      "epoch": 1.7347236213215593,
+      "grad_norm": 1.65625,
+      "learning_rate": 7.517789252983891e-06,
+      "loss": 0.8667,
+      "step": 10187
+    },
+    {
+      "epoch": 1.7348950854105494,
+      "grad_norm": 1.71875,
+      "learning_rate": 7.516039884687557e-06,
+      "loss": 0.8285,
+      "step": 10188
+    },
+    {
+      "epoch": 1.7350665494995392,
+      "grad_norm": 1.6328125,
+      "learning_rate": 7.51429059740252e-06,
+      "loss": 0.8443,
+      "step": 10189
+    },
+    {
+      "epoch": 1.735238013588529,
+      "grad_norm": 1.78125,
+      "learning_rate": 7.512541391185823e-06,
+      "loss": 0.8528,
+      "step": 10190
+    },
+    {
+      "epoch": 1.735409477677519,
+      "grad_norm": 1.703125,
+      "learning_rate": 7.51079226609452e-06,
+      "loss": 0.7553,
+      "step": 10191
+    },
+    {
+      "epoch": 1.735580941766509,
+      "grad_norm": 1.75,
+      "learning_rate": 7.5090432221856526e-06,
+      "loss": 0.8326,
+      "step": 10192
+    },
+    {
+      "epoch": 1.7357524058554987,
+      "grad_norm": 1.671875,
+      "learning_rate": 7.507294259516265e-06,
+      "loss": 0.8594,
+      "step": 10193
+    },
+    {
+      "epoch": 1.7359238699444886,
+      "grad_norm": 1.65625,
+      "learning_rate": 7.505545378143398e-06,
+      "loss": 0.8385,
+      "step": 10194
+    },
+    {
+      "epoch": 1.7360953340334784,
+      "grad_norm": 1.796875,
+      "learning_rate": 7.503796578124092e-06,
+      "loss": 0.8266,
+      "step": 10195
+    },
+    {
+      "epoch": 1.7362667981224682,
+      "grad_norm": 1.71875,
+      "learning_rate": 7.502047859515378e-06,
+      "loss": 0.8895,
+      "step": 10196
+    },
+    {
+      "epoch": 1.736438262211458,
+      "grad_norm": 1.6328125,
+      "learning_rate": 7.50029922237429e-06,
+      "loss": 0.7925,
+      "step": 10197
+    },
+    {
+      "epoch": 1.7366097263004479,
+      "grad_norm": 1.78125,
+      "learning_rate": 7.498550666757858e-06,
+      "loss": 0.8161,
+      "step": 10198
+    },
+    {
+      "epoch": 1.7367811903894377,
+      "grad_norm": 1.6875,
+      "learning_rate": 7.496802192723107e-06,
+      "loss": 0.8297,
+      "step": 10199
+    },
+    {
+      "epoch": 1.7369526544784277,
+      "grad_norm": 1.703125,
+      "learning_rate": 7.495053800327068e-06,
+      "loss": 0.8645,
+      "step": 10200
+    },
+    {
+      "epoch": 1.7371241185674176,
+      "grad_norm": 1.6484375,
+      "learning_rate": 7.493305489626753e-06,
+      "loss": 0.8653,
+      "step": 10201
+    },
+    {
+      "epoch": 1.7372955826564074,
+      "grad_norm": 1.6875,
+      "learning_rate": 7.491557260679183e-06,
+      "loss": 0.7951,
+      "step": 10202
+    },
+    {
+      "epoch": 1.7374670467453974,
+      "grad_norm": 1.6640625,
+      "learning_rate": 7.489809113541379e-06,
+      "loss": 0.7885,
+      "step": 10203
+    },
+    {
+      "epoch": 1.7376385108343873,
+      "grad_norm": 1.75,
+      "learning_rate": 7.488061048270352e-06,
+      "loss": 0.9382,
+      "step": 10204
+    },
+    {
+      "epoch": 1.737809974923377,
+      "grad_norm": 1.734375,
+      "learning_rate": 7.486313064923114e-06,
+      "loss": 0.8568,
+      "step": 10205
+    },
+    {
+      "epoch": 1.737981439012367,
+      "grad_norm": 1.6796875,
+      "learning_rate": 7.484565163556672e-06,
+      "loss": 0.8941,
+      "step": 10206
+    },
+    {
+      "epoch": 1.7381529031013567,
+      "grad_norm": 1.6328125,
+      "learning_rate": 7.482817344228031e-06,
+      "loss": 0.804,
+      "step": 10207
+    },
+    {
+      "epoch": 1.7383243671903466,
+      "grad_norm": 1.671875,
+      "learning_rate": 7.481069606994198e-06,
+      "loss": 0.8119,
+      "step": 10208
+    },
+    {
+      "epoch": 1.7384958312793364,
+      "grad_norm": 1.734375,
+      "learning_rate": 7.479321951912168e-06,
+      "loss": 0.839,
+      "step": 10209
+    },
+    {
+      "epoch": 1.7386672953683262,
+      "grad_norm": 1.640625,
+      "learning_rate": 7.477574379038943e-06,
+      "loss": 0.8308,
+      "step": 10210
+    },
+    {
+      "epoch": 1.738838759457316,
+      "grad_norm": 1.78125,
+      "learning_rate": 7.475826888431515e-06,
+      "loss": 0.8357,
+      "step": 10211
+    },
+    {
+      "epoch": 1.7390102235463059,
+      "grad_norm": 1.6875,
+      "learning_rate": 7.47407948014688e-06,
+      "loss": 0.8502,
+      "step": 10212
+    },
+    {
+      "epoch": 1.739181687635296,
+      "grad_norm": 1.6796875,
+      "learning_rate": 7.472332154242023e-06,
+      "loss": 0.8749,
+      "step": 10213
+    },
+    {
+      "epoch": 1.7393531517242857,
+      "grad_norm": 1.7265625,
+      "learning_rate": 7.470584910773931e-06,
+      "loss": 0.8231,
+      "step": 10214
+    },
+    {
+      "epoch": 1.7395246158132756,
+      "grad_norm": 1.734375,
+      "learning_rate": 7.4688377497995915e-06,
+      "loss": 0.8878,
+      "step": 10215
+    },
+    {
+      "epoch": 1.7396960799022656,
+      "grad_norm": 1.703125,
+      "learning_rate": 7.467090671375985e-06,
+      "loss": 0.9022,
+      "step": 10216
+    },
+    {
+      "epoch": 1.7398675439912554,
+      "grad_norm": 1.71875,
+      "learning_rate": 7.465343675560088e-06,
+      "loss": 0.8323,
+      "step": 10217
+    },
+    {
+      "epoch": 1.7400390080802453,
+      "grad_norm": 1.7578125,
+      "learning_rate": 7.4635967624088805e-06,
+      "loss": 0.8797,
+      "step": 10218
+    },
+    {
+      "epoch": 1.740210472169235,
+      "grad_norm": 1.6015625,
+      "learning_rate": 7.461849931979332e-06,
+      "loss": 0.8438,
+      "step": 10219
+    },
+    {
+      "epoch": 1.740381936258225,
+      "grad_norm": 1.6796875,
+      "learning_rate": 7.4601031843284155e-06,
+      "loss": 0.903,
+      "step": 10220
+    },
+    {
+      "epoch": 1.7405534003472147,
+      "grad_norm": 1.6796875,
+      "learning_rate": 7.4583565195130995e-06,
+      "loss": 0.7802,
+      "step": 10221
+    },
+    {
+      "epoch": 1.7407248644362046,
+      "grad_norm": 1.734375,
+      "learning_rate": 7.456609937590347e-06,
+      "loss": 0.8862,
+      "step": 10222
+    },
+    {
+      "epoch": 1.7408963285251944,
+      "grad_norm": 1.7734375,
+      "learning_rate": 7.454863438617124e-06,
+      "loss": 0.849,
+      "step": 10223
+    },
+    {
+      "epoch": 1.7410677926141842,
+      "grad_norm": 1.796875,
+      "learning_rate": 7.453117022650387e-06,
+      "loss": 0.8742,
+      "step": 10224
+    },
+    {
+      "epoch": 1.7412392567031743,
+      "grad_norm": 1.6484375,
+      "learning_rate": 7.451370689747095e-06,
+      "loss": 0.8438,
+      "step": 10225
+    },
+    {
+      "epoch": 1.741410720792164,
+      "grad_norm": 1.6640625,
+      "learning_rate": 7.4496244399642005e-06,
+      "loss": 0.9366,
+      "step": 10226
+    },
+    {
+      "epoch": 1.741582184881154,
+      "grad_norm": 1.6875,
+      "learning_rate": 7.447878273358657e-06,
+      "loss": 0.8167,
+      "step": 10227
+    },
+    {
+      "epoch": 1.741753648970144,
+      "grad_norm": 1.78125,
+      "learning_rate": 7.4461321899874155e-06,
+      "loss": 0.8649,
+      "step": 10228
+    },
+    {
+      "epoch": 1.7419251130591338,
+      "grad_norm": 1.6484375,
+      "learning_rate": 7.444386189907418e-06,
+      "loss": 0.8296,
+      "step": 10229
+    },
+    {
+      "epoch": 1.7420965771481236,
+      "grad_norm": 1.65625,
+      "learning_rate": 7.442640273175612e-06,
+      "loss": 0.8709,
+      "step": 10230
+    },
+    {
+      "epoch": 1.7422680412371134,
+      "grad_norm": 1.6484375,
+      "learning_rate": 7.440894439848935e-06,
+      "loss": 0.8295,
+      "step": 10231
+    },
+    {
+      "epoch": 1.7424395053261033,
+      "grad_norm": 1.7109375,
+      "learning_rate": 7.439148689984327e-06,
+      "loss": 0.8687,
+      "step": 10232
+    },
+    {
+      "epoch": 1.742610969415093,
+      "grad_norm": 1.6640625,
+      "learning_rate": 7.437403023638725e-06,
+      "loss": 0.8493,
+      "step": 10233
+    },
+    {
+      "epoch": 1.742782433504083,
+      "grad_norm": 1.7421875,
+      "learning_rate": 7.43565744086906e-06,
+      "loss": 0.8864,
+      "step": 10234
+    },
+    {
+      "epoch": 1.7429538975930727,
+      "grad_norm": 1.78125,
+      "learning_rate": 7.433911941732266e-06,
+      "loss": 0.8556,
+      "step": 10235
+    },
+    {
+      "epoch": 1.7431253616820626,
+      "grad_norm": 1.609375,
+      "learning_rate": 7.4321665262852626e-06,
+      "loss": 0.7473,
+      "step": 10236
+    },
+    {
+      "epoch": 1.7432968257710526,
+      "grad_norm": 1.5859375,
+      "learning_rate": 7.430421194584978e-06,
+      "loss": 0.9047,
+      "step": 10237
+    },
+    {
+      "epoch": 1.7434682898600424,
+      "grad_norm": 1.65625,
+      "learning_rate": 7.428675946688335e-06,
+      "loss": 0.8799,
+      "step": 10238
+    },
+    {
+      "epoch": 1.7436397539490323,
+      "grad_norm": 1.7890625,
+      "learning_rate": 7.426930782652253e-06,
+      "loss": 0.8822,
+      "step": 10239
+    },
+    {
+      "epoch": 1.7438112180380223,
+      "grad_norm": 1.625,
+      "learning_rate": 7.425185702533644e-06,
+      "loss": 0.8088,
+      "step": 10240
+    },
+    {
+      "epoch": 1.7439826821270121,
+      "grad_norm": 1.71875,
+      "learning_rate": 7.423440706389427e-06,
+      "loss": 0.8863,
+      "step": 10241
+    },
+    {
+      "epoch": 1.744154146216002,
+      "grad_norm": 1.6640625,
+      "learning_rate": 7.421695794276511e-06,
+      "loss": 0.788,
+      "step": 10242
+    },
+    {
+      "epoch": 1.7443256103049918,
+      "grad_norm": 1.671875,
+      "learning_rate": 7.4199509662518054e-06,
+      "loss": 0.859,
+      "step": 10243
+    },
+    {
+      "epoch": 1.7444970743939816,
+      "grad_norm": 1.7578125,
+      "learning_rate": 7.418206222372213e-06,
+      "loss": 0.8653,
+      "step": 10244
+    },
+    {
+      "epoch": 1.7446685384829714,
+      "grad_norm": 1.7578125,
+      "learning_rate": 7.416461562694639e-06,
+      "loss": 0.8734,
+      "step": 10245
+    },
+    {
+      "epoch": 1.7448400025719613,
+      "grad_norm": 1.6953125,
+      "learning_rate": 7.414716987275985e-06,
+      "loss": 0.8341,
+      "step": 10246
+    },
+    {
+      "epoch": 1.745011466660951,
+      "grad_norm": 1.8203125,
+      "learning_rate": 7.412972496173143e-06,
+      "loss": 0.8791,
+      "step": 10247
+    },
+    {
+      "epoch": 1.745182930749941,
+      "grad_norm": 1.8203125,
+      "learning_rate": 7.4112280894430076e-06,
+      "loss": 0.8111,
+      "step": 10248
+    },
+    {
+      "epoch": 1.745354394838931,
+      "grad_norm": 1.6640625,
+      "learning_rate": 7.409483767142473e-06,
+      "loss": 0.8054,
+      "step": 10249
+    },
+    {
+      "epoch": 1.7455258589279208,
+      "grad_norm": 1.8359375,
+      "learning_rate": 7.4077395293284285e-06,
+      "loss": 0.8333,
+      "step": 10250
+    },
+    {
+      "epoch": 1.7456973230169106,
+      "grad_norm": 1.6640625,
+      "learning_rate": 7.405995376057758e-06,
+      "loss": 0.7959,
+      "step": 10251
+    },
+    {
+      "epoch": 1.7458687871059007,
+      "grad_norm": 1.6328125,
+      "learning_rate": 7.404251307387349e-06,
+      "loss": 0.8575,
+      "step": 10252
+    },
+    {
+      "epoch": 1.7460402511948905,
+      "grad_norm": 1.6015625,
+      "learning_rate": 7.402507323374077e-06,
+      "loss": 0.8788,
+      "step": 10253
+    },
+    {
+      "epoch": 1.7462117152838803,
+      "grad_norm": 1.765625,
+      "learning_rate": 7.400763424074824e-06,
+      "loss": 0.8249,
+      "step": 10254
+    },
+    {
+      "epoch": 1.7463831793728701,
+      "grad_norm": 1.6640625,
+      "learning_rate": 7.399019609546464e-06,
+      "loss": 0.9007,
+      "step": 10255
+    },
+    {
+      "epoch": 1.74655464346186,
+      "grad_norm": 1.6640625,
+      "learning_rate": 7.397275879845868e-06,
+      "loss": 0.8715,
+      "step": 10256
+    },
+    {
+      "epoch": 1.7467261075508498,
+      "grad_norm": 1.7109375,
+      "learning_rate": 7.395532235029908e-06,
+      "loss": 0.8172,
+      "step": 10257
+    },
+    {
+      "epoch": 1.7468975716398396,
+      "grad_norm": 1.7265625,
+      "learning_rate": 7.393788675155449e-06,
+      "loss": 0.922,
+      "step": 10258
+    },
+    {
+      "epoch": 1.7470690357288294,
+      "grad_norm": 1.609375,
+      "learning_rate": 7.392045200279354e-06,
+      "loss": 0.7991,
+      "step": 10259
+    },
+    {
+      "epoch": 1.7472404998178193,
+      "grad_norm": 1.7265625,
+      "learning_rate": 7.390301810458487e-06,
+      "loss": 0.8249,
+      "step": 10260
+    },
+    {
+      "epoch": 1.7474119639068093,
+      "grad_norm": 1.734375,
+      "learning_rate": 7.388558505749703e-06,
+      "loss": 0.8807,
+      "step": 10261
+    },
+    {
+      "epoch": 1.7475834279957991,
+      "grad_norm": 1.625,
+      "learning_rate": 7.386815286209862e-06,
+      "loss": 0.8831,
+      "step": 10262
+    },
+    {
+      "epoch": 1.747754892084789,
+      "grad_norm": 1.640625,
+      "learning_rate": 7.385072151895814e-06,
+      "loss": 0.854,
+      "step": 10263
+    },
+    {
+      "epoch": 1.747926356173779,
+      "grad_norm": 1.5859375,
+      "learning_rate": 7.383329102864411e-06,
+      "loss": 0.8495,
+      "step": 10264
+    },
+    {
+      "epoch": 1.7480978202627688,
+      "grad_norm": 1.6875,
+      "learning_rate": 7.381586139172499e-06,
+      "loss": 0.8507,
+      "step": 10265
+    },
+    {
+      "epoch": 1.7482692843517587,
+      "grad_norm": 1.625,
+      "learning_rate": 7.379843260876922e-06,
+      "loss": 0.8643,
+      "step": 10266
+    },
+    {
+      "epoch": 1.7484407484407485,
+      "grad_norm": 1.734375,
+      "learning_rate": 7.3781004680345235e-06,
+      "loss": 0.7792,
+      "step": 10267
+    },
+    {
+      "epoch": 1.7486122125297383,
+      "grad_norm": 1.5546875,
+      "learning_rate": 7.376357760702142e-06,
+      "loss": 0.7385,
+      "step": 10268
+    },
+    {
+      "epoch": 1.7487836766187281,
+      "grad_norm": 1.6796875,
+      "learning_rate": 7.374615138936615e-06,
+      "loss": 0.8638,
+      "step": 10269
+    },
+    {
+      "epoch": 1.748955140707718,
+      "grad_norm": 1.7421875,
+      "learning_rate": 7.372872602794772e-06,
+      "loss": 0.8629,
+      "step": 10270
+    },
+    {
+      "epoch": 1.7491266047967078,
+      "grad_norm": 1.6953125,
+      "learning_rate": 7.371130152333448e-06,
+      "loss": 0.8324,
+      "step": 10271
+    },
+    {
+      "epoch": 1.7492980688856976,
+      "grad_norm": 1.65625,
+      "learning_rate": 7.369387787609469e-06,
+      "loss": 0.8626,
+      "step": 10272
+    },
+    {
+      "epoch": 1.7494695329746877,
+      "grad_norm": 1.671875,
+      "learning_rate": 7.367645508679659e-06,
+      "loss": 0.8506,
+      "step": 10273
+    },
+    {
+      "epoch": 1.7496409970636775,
+      "grad_norm": 1.703125,
+      "learning_rate": 7.365903315600842e-06,
+      "loss": 0.7514,
+      "step": 10274
+    },
+    {
+      "epoch": 1.7498124611526673,
+      "grad_norm": 1.7109375,
+      "learning_rate": 7.364161208429838e-06,
+      "loss": 0.8505,
+      "step": 10275
+    },
+    {
+      "epoch": 1.7499839252416574,
+      "grad_norm": 1.65625,
+      "learning_rate": 7.362419187223462e-06,
+      "loss": 0.8002,
+      "step": 10276
+    },
+    {
+      "epoch": 1.7501553893306472,
+      "grad_norm": 1.6015625,
+      "learning_rate": 7.360677252038529e-06,
+      "loss": 0.7803,
+      "step": 10277
+    },
+    {
+      "epoch": 1.750326853419637,
+      "grad_norm": 1.71875,
+      "learning_rate": 7.358935402931848e-06,
+      "loss": 0.8861,
+      "step": 10278
+    },
+    {
+      "epoch": 1.7504983175086268,
+      "grad_norm": 1.7578125,
+      "learning_rate": 7.357193639960227e-06,
+      "loss": 0.8101,
+      "step": 10279
+    },
+    {
+      "epoch": 1.7506697815976167,
+      "grad_norm": 1.7109375,
+      "learning_rate": 7.355451963180477e-06,
+      "loss": 0.829,
+      "step": 10280
+    },
+    {
+      "epoch": 1.7508412456866065,
+      "grad_norm": 1.6875,
+      "learning_rate": 7.353710372649399e-06,
+      "loss": 0.8157,
+      "step": 10281
+    },
+    {
+      "epoch": 1.7510127097755963,
+      "grad_norm": 1.6953125,
+      "learning_rate": 7.351968868423789e-06,
+      "loss": 0.9102,
+      "step": 10282
+    },
+    {
+      "epoch": 1.7511841738645861,
+      "grad_norm": 1.6640625,
+      "learning_rate": 7.350227450560443e-06,
+      "loss": 0.8951,
+      "step": 10283
+    },
+    {
+      "epoch": 1.751355637953576,
+      "grad_norm": 1.6171875,
+      "learning_rate": 7.348486119116161e-06,
+      "loss": 0.8913,
+      "step": 10284
+    },
+    {
+      "epoch": 1.751527102042566,
+      "grad_norm": 1.671875,
+      "learning_rate": 7.346744874147729e-06,
+      "loss": 0.8598,
+      "step": 10285
+    },
+    {
+      "epoch": 1.7516985661315558,
+      "grad_norm": 1.703125,
+      "learning_rate": 7.345003715711938e-06,
+      "loss": 0.8271,
+      "step": 10286
+    },
+    {
+      "epoch": 1.7518700302205457,
+      "grad_norm": 1.6796875,
+      "learning_rate": 7.3432626438655726e-06,
+      "loss": 0.8592,
+      "step": 10287
+    },
+    {
+      "epoch": 1.7520414943095357,
+      "grad_norm": 1.6796875,
+      "learning_rate": 7.3415216586654184e-06,
+      "loss": 0.8075,
+      "step": 10288
+    },
+    {
+      "epoch": 1.7522129583985255,
+      "grad_norm": 1.6640625,
+      "learning_rate": 7.339780760168254e-06,
+      "loss": 0.8372,
+      "step": 10289
+    },
+    {
+      "epoch": 1.7523844224875154,
+      "grad_norm": 1.65625,
+      "learning_rate": 7.338039948430857e-06,
+      "loss": 0.8529,
+      "step": 10290
+    },
+    {
+      "epoch": 1.7525558865765052,
+      "grad_norm": 1.671875,
+      "learning_rate": 7.336299223509999e-06,
+      "loss": 0.7569,
+      "step": 10291
+    },
+    {
+      "epoch": 1.752727350665495,
+      "grad_norm": 1.6796875,
+      "learning_rate": 7.334558585462461e-06,
+      "loss": 0.8094,
+      "step": 10292
+    },
+    {
+      "epoch": 1.7528988147544848,
+      "grad_norm": 1.828125,
+      "learning_rate": 7.332818034344999e-06,
+      "loss": 0.897,
+      "step": 10293
+    },
+    {
+      "epoch": 1.7530702788434747,
+      "grad_norm": 1.7578125,
+      "learning_rate": 7.331077570214385e-06,
+      "loss": 0.8805,
+      "step": 10294
+    },
+    {
+      "epoch": 1.7532417429324645,
+      "grad_norm": 1.6796875,
+      "learning_rate": 7.329337193127379e-06,
+      "loss": 0.8222,
+      "step": 10295
+    },
+    {
+      "epoch": 1.7534132070214543,
+      "grad_norm": 1.75,
+      "learning_rate": 7.327596903140746e-06,
+      "loss": 0.8482,
+      "step": 10296
+    },
+    {
+      "epoch": 1.7535846711104444,
+      "grad_norm": 1.7265625,
+      "learning_rate": 7.325856700311243e-06,
+      "loss": 0.8909,
+      "step": 10297
+    },
+    {
+      "epoch": 1.7537561351994342,
+      "grad_norm": 1.7109375,
+      "learning_rate": 7.32411658469562e-06,
+      "loss": 0.8621,
+      "step": 10298
+    },
+    {
+      "epoch": 1.753927599288424,
+      "grad_norm": 1.703125,
+      "learning_rate": 7.322376556350633e-06,
+      "loss": 0.8418,
+      "step": 10299
+    },
+    {
+      "epoch": 1.754099063377414,
+      "grad_norm": 1.671875,
+      "learning_rate": 7.32063661533303e-06,
+      "loss": 0.8549,
+      "step": 10300
+    },
+    {
+      "epoch": 1.7542705274664039,
+      "grad_norm": 1.78125,
+      "learning_rate": 7.318896761699557e-06,
+      "loss": 0.9614,
+      "step": 10301
+    },
+    {
+      "epoch": 1.7544419915553937,
+      "grad_norm": 1.6171875,
+      "learning_rate": 7.317156995506955e-06,
+      "loss": 0.8664,
+      "step": 10302
+    },
+    {
+      "epoch": 1.7546134556443835,
+      "grad_norm": 1.6953125,
+      "learning_rate": 7.315417316811967e-06,
+      "loss": 0.8552,
+      "step": 10303
+    },
+    {
+      "epoch": 1.7547849197333734,
+      "grad_norm": 1.6640625,
+      "learning_rate": 7.313677725671331e-06,
+      "loss": 0.8504,
+      "step": 10304
+    },
+    {
+      "epoch": 1.7549563838223632,
+      "grad_norm": 1.6484375,
+      "learning_rate": 7.311938222141779e-06,
+      "loss": 0.7789,
+      "step": 10305
+    },
+    {
+      "epoch": 1.755127847911353,
+      "grad_norm": 1.671875,
+      "learning_rate": 7.3101988062800435e-06,
+      "loss": 0.8515,
+      "step": 10306
+    },
+    {
+      "epoch": 1.7552993120003428,
+      "grad_norm": 1.7265625,
+      "learning_rate": 7.308459478142853e-06,
+      "loss": 0.8759,
+      "step": 10307
+    },
+    {
+      "epoch": 1.7554707760893327,
+      "grad_norm": 1.6640625,
+      "learning_rate": 7.3067202377869354e-06,
+      "loss": 0.8369,
+      "step": 10308
+    },
+    {
+      "epoch": 1.7556422401783225,
+      "grad_norm": 1.6640625,
+      "learning_rate": 7.304981085269012e-06,
+      "loss": 0.8632,
+      "step": 10309
+    },
+    {
+      "epoch": 1.7558137042673125,
+      "grad_norm": 1.765625,
+      "learning_rate": 7.303242020645804e-06,
+      "loss": 0.8468,
+      "step": 10310
+    },
+    {
+      "epoch": 1.7559851683563024,
+      "grad_norm": 1.71875,
+      "learning_rate": 7.30150304397403e-06,
+      "loss": 0.7487,
+      "step": 10311
+    },
+    {
+      "epoch": 1.7561566324452922,
+      "grad_norm": 1.71875,
+      "learning_rate": 7.2997641553104025e-06,
+      "loss": 0.8561,
+      "step": 10312
+    },
+    {
+      "epoch": 1.7563280965342822,
+      "grad_norm": 1.6328125,
+      "learning_rate": 7.298025354711633e-06,
+      "loss": 0.8081,
+      "step": 10313
+    },
+    {
+      "epoch": 1.756499560623272,
+      "grad_norm": 1.6953125,
+      "learning_rate": 7.296286642234434e-06,
+      "loss": 0.9342,
+      "step": 10314
+    },
+    {
+      "epoch": 1.7566710247122619,
+      "grad_norm": 1.7265625,
+      "learning_rate": 7.2945480179355075e-06,
+      "loss": 0.8533,
+      "step": 10315
+    },
+    {
+      "epoch": 1.7568424888012517,
+      "grad_norm": 1.71875,
+      "learning_rate": 7.292809481871559e-06,
+      "loss": 0.8413,
+      "step": 10316
+    },
+    {
+      "epoch": 1.7570139528902415,
+      "grad_norm": 1.6875,
+      "learning_rate": 7.291071034099285e-06,
+      "loss": 0.8442,
+      "step": 10317
+    },
+    {
+      "epoch": 1.7571854169792314,
+      "grad_norm": 1.8203125,
+      "learning_rate": 7.289332674675386e-06,
+      "loss": 0.919,
+      "step": 10318
+    },
+    {
+      "epoch": 1.7573568810682212,
+      "grad_norm": 1.7109375,
+      "learning_rate": 7.287594403656557e-06,
+      "loss": 0.829,
+      "step": 10319
+    },
+    {
+      "epoch": 1.757528345157211,
+      "grad_norm": 1.625,
+      "learning_rate": 7.2858562210994886e-06,
+      "loss": 0.8363,
+      "step": 10320
+    },
+    {
+      "epoch": 1.7576998092462008,
+      "grad_norm": 1.6875,
+      "learning_rate": 7.284118127060868e-06,
+      "loss": 0.8166,
+      "step": 10321
+    },
+    {
+      "epoch": 1.7578712733351909,
+      "grad_norm": 1.5859375,
+      "learning_rate": 7.282380121597384e-06,
+      "loss": 0.7882,
+      "step": 10322
+    },
+    {
+      "epoch": 1.7580427374241807,
+      "grad_norm": 1.6953125,
+      "learning_rate": 7.2806422047657165e-06,
+      "loss": 0.8613,
+      "step": 10323
+    },
+    {
+      "epoch": 1.7582142015131705,
+      "grad_norm": 1.609375,
+      "learning_rate": 7.278904376622548e-06,
+      "loss": 0.8299,
+      "step": 10324
+    },
+    {
+      "epoch": 1.7583856656021606,
+      "grad_norm": 1.671875,
+      "learning_rate": 7.277166637224553e-06,
+      "loss": 0.7943,
+      "step": 10325
+    },
+    {
+      "epoch": 1.7585571296911504,
+      "grad_norm": 1.8203125,
+      "learning_rate": 7.275428986628406e-06,
+      "loss": 0.9177,
+      "step": 10326
+    },
+    {
+      "epoch": 1.7587285937801402,
+      "grad_norm": 1.734375,
+      "learning_rate": 7.273691424890786e-06,
+      "loss": 0.8536,
+      "step": 10327
+    },
+    {
+      "epoch": 1.75890005786913,
+      "grad_norm": 1.8046875,
+      "learning_rate": 7.271953952068351e-06,
+      "loss": 0.8851,
+      "step": 10328
+    },
+    {
+      "epoch": 1.7590715219581199,
+      "grad_norm": 1.7421875,
+      "learning_rate": 7.270216568217769e-06,
+      "loss": 0.8791,
+      "step": 10329
+    },
+    {
+      "epoch": 1.7592429860471097,
+      "grad_norm": 1.734375,
+      "learning_rate": 7.2684792733957055e-06,
+      "loss": 0.8501,
+      "step": 10330
+    },
+    {
+      "epoch": 1.7594144501360995,
+      "grad_norm": 1.625,
+      "learning_rate": 7.2667420676588185e-06,
+      "loss": 0.8799,
+      "step": 10331
+    },
+    {
+      "epoch": 1.7595859142250894,
+      "grad_norm": 1.703125,
+      "learning_rate": 7.265004951063765e-06,
+      "loss": 0.8702,
+      "step": 10332
+    },
+    {
+      "epoch": 1.7597573783140792,
+      "grad_norm": 1.765625,
+      "learning_rate": 7.263267923667199e-06,
+      "loss": 0.9555,
+      "step": 10333
+    },
+    {
+      "epoch": 1.7599288424030692,
+      "grad_norm": 1.65625,
+      "learning_rate": 7.261530985525771e-06,
+      "loss": 0.9094,
+      "step": 10334
+    },
+    {
+      "epoch": 1.760100306492059,
+      "grad_norm": 1.71875,
+      "learning_rate": 7.25979413669613e-06,
+      "loss": 0.8269,
+      "step": 10335
+    },
+    {
+      "epoch": 1.760271770581049,
+      "grad_norm": 1.6796875,
+      "learning_rate": 7.258057377234922e-06,
+      "loss": 0.8336,
+      "step": 10336
+    },
+    {
+      "epoch": 1.760443234670039,
+      "grad_norm": 1.828125,
+      "learning_rate": 7.256320707198786e-06,
+      "loss": 0.8184,
+      "step": 10337
+    },
+    {
+      "epoch": 1.7606146987590288,
+      "grad_norm": 1.65625,
+      "learning_rate": 7.254584126644366e-06,
+      "loss": 0.8485,
+      "step": 10338
+    },
+    {
+      "epoch": 1.7607861628480186,
+      "grad_norm": 1.7109375,
+      "learning_rate": 7.252847635628298e-06,
+      "loss": 0.849,
+      "step": 10339
+    },
+    {
+      "epoch": 1.7609576269370084,
+      "grad_norm": 1.7421875,
+      "learning_rate": 7.251111234207211e-06,
+      "loss": 0.8736,
+      "step": 10340
+    },
+    {
+      "epoch": 1.7611290910259982,
+      "grad_norm": 1.6015625,
+      "learning_rate": 7.249374922437737e-06,
+      "loss": 0.877,
+      "step": 10341
+    },
+    {
+      "epoch": 1.761300555114988,
+      "grad_norm": 1.640625,
+      "learning_rate": 7.247638700376503e-06,
+      "loss": 0.8993,
+      "step": 10342
+    },
+    {
+      "epoch": 1.761472019203978,
+      "grad_norm": 1.75,
+      "learning_rate": 7.2459025680801365e-06,
+      "loss": 0.8747,
+      "step": 10343
+    },
+    {
+      "epoch": 1.7616434832929677,
+      "grad_norm": 1.765625,
+      "learning_rate": 7.2441665256052606e-06,
+      "loss": 0.8566,
+      "step": 10344
+    },
+    {
+      "epoch": 1.7618149473819575,
+      "grad_norm": 1.703125,
+      "learning_rate": 7.24243057300849e-06,
+      "loss": 0.9118,
+      "step": 10345
+    },
+    {
+      "epoch": 1.7619864114709476,
+      "grad_norm": 1.6796875,
+      "learning_rate": 7.240694710346443e-06,
+      "loss": 0.8452,
+      "step": 10346
+    },
+    {
+      "epoch": 1.7621578755599374,
+      "grad_norm": 1.71875,
+      "learning_rate": 7.238958937675731e-06,
+      "loss": 0.8191,
+      "step": 10347
+    },
+    {
+      "epoch": 1.7623293396489272,
+      "grad_norm": 1.6328125,
+      "learning_rate": 7.237223255052967e-06,
+      "loss": 0.8403,
+      "step": 10348
+    },
+    {
+      "epoch": 1.7625008037379173,
+      "grad_norm": 1.6015625,
+      "learning_rate": 7.235487662534755e-06,
+      "loss": 0.7737,
+      "step": 10349
+    },
+    {
+      "epoch": 1.7626722678269071,
+      "grad_norm": 1.6796875,
+      "learning_rate": 7.233752160177705e-06,
+      "loss": 0.8608,
+      "step": 10350
+    },
+    {
+      "epoch": 1.762843731915897,
+      "grad_norm": 1.7265625,
+      "learning_rate": 7.232016748038408e-06,
+      "loss": 0.8762,
+      "step": 10351
+    },
+    {
+      "epoch": 1.7630151960048868,
+      "grad_norm": 1.625,
+      "learning_rate": 7.230281426173469e-06,
+      "loss": 0.8423,
+      "step": 10352
+    },
+    {
+      "epoch": 1.7631866600938766,
+      "grad_norm": 1.7109375,
+      "learning_rate": 7.228546194639483e-06,
+      "loss": 0.8653,
+      "step": 10353
+    },
+    {
+      "epoch": 1.7633581241828664,
+      "grad_norm": 1.703125,
+      "learning_rate": 7.226811053493041e-06,
+      "loss": 0.8833,
+      "step": 10354
+    },
+    {
+      "epoch": 1.7635295882718562,
+      "grad_norm": 1.765625,
+      "learning_rate": 7.225076002790736e-06,
+      "loss": 0.8177,
+      "step": 10355
+    },
+    {
+      "epoch": 1.763701052360846,
+      "grad_norm": 1.640625,
+      "learning_rate": 7.223341042589151e-06,
+      "loss": 0.751,
+      "step": 10356
+    },
+    {
+      "epoch": 1.763872516449836,
+      "grad_norm": 1.796875,
+      "learning_rate": 7.221606172944869e-06,
+      "loss": 0.8932,
+      "step": 10357
+    },
+    {
+      "epoch": 1.764043980538826,
+      "grad_norm": 1.6796875,
+      "learning_rate": 7.219871393914473e-06,
+      "loss": 0.9326,
+      "step": 10358
+    },
+    {
+      "epoch": 1.7642154446278158,
+      "grad_norm": 1.765625,
+      "learning_rate": 7.218136705554541e-06,
+      "loss": 0.9091,
+      "step": 10359
+    },
+    {
+      "epoch": 1.7643869087168056,
+      "grad_norm": 1.7265625,
+      "learning_rate": 7.216402107921645e-06,
+      "loss": 0.7611,
+      "step": 10360
+    },
+    {
+      "epoch": 1.7645583728057956,
+      "grad_norm": 1.6171875,
+      "learning_rate": 7.214667601072358e-06,
+      "loss": 0.8232,
+      "step": 10361
+    },
+    {
+      "epoch": 1.7647298368947855,
+      "grad_norm": 1.71875,
+      "learning_rate": 7.2129331850632535e-06,
+      "loss": 0.8851,
+      "step": 10362
+    },
+    {
+      "epoch": 1.7649013009837753,
+      "grad_norm": 1.7265625,
+      "learning_rate": 7.21119885995089e-06,
+      "loss": 0.8625,
+      "step": 10363
+    },
+    {
+      "epoch": 1.7650727650727651,
+      "grad_norm": 1.6796875,
+      "learning_rate": 7.209464625791831e-06,
+      "loss": 0.8253,
+      "step": 10364
+    },
+    {
+      "epoch": 1.765244229161755,
+      "grad_norm": 1.609375,
+      "learning_rate": 7.207730482642641e-06,
+      "loss": 0.8143,
+      "step": 10365
+    },
+    {
+      "epoch": 1.7654156932507448,
+      "grad_norm": 1.6953125,
+      "learning_rate": 7.205996430559874e-06,
+      "loss": 0.8331,
+      "step": 10366
+    },
+    {
+      "epoch": 1.7655871573397346,
+      "grad_norm": 1.6953125,
+      "learning_rate": 7.2042624696000855e-06,
+      "loss": 0.9238,
+      "step": 10367
+    },
+    {
+      "epoch": 1.7657586214287244,
+      "grad_norm": 1.78125,
+      "learning_rate": 7.202528599819825e-06,
+      "loss": 0.8804,
+      "step": 10368
+    },
+    {
+      "epoch": 1.7659300855177142,
+      "grad_norm": 1.734375,
+      "learning_rate": 7.200794821275641e-06,
+      "loss": 0.8217,
+      "step": 10369
+    },
+    {
+      "epoch": 1.7661015496067043,
+      "grad_norm": 1.7265625,
+      "learning_rate": 7.199061134024079e-06,
+      "loss": 0.8658,
+      "step": 10370
+    },
+    {
+      "epoch": 1.7662730136956941,
+      "grad_norm": 1.796875,
+      "learning_rate": 7.197327538121681e-06,
+      "loss": 0.9574,
+      "step": 10371
+    },
+    {
+      "epoch": 1.766444477784684,
+      "grad_norm": 1.640625,
+      "learning_rate": 7.195594033624985e-06,
+      "loss": 0.8551,
+      "step": 10372
+    },
+    {
+      "epoch": 1.766615941873674,
+      "grad_norm": 1.6953125,
+      "learning_rate": 7.193860620590532e-06,
+      "loss": 0.8563,
+      "step": 10373
+    },
+    {
+      "epoch": 1.7667874059626638,
+      "grad_norm": 1.59375,
+      "learning_rate": 7.192127299074847e-06,
+      "loss": 0.7884,
+      "step": 10374
+    },
+    {
+      "epoch": 1.7669588700516536,
+      "grad_norm": 1.7578125,
+      "learning_rate": 7.190394069134464e-06,
+      "loss": 0.9111,
+      "step": 10375
+    },
+    {
+      "epoch": 1.7671303341406435,
+      "grad_norm": 1.8203125,
+      "learning_rate": 7.188660930825911e-06,
+      "loss": 0.8985,
+      "step": 10376
+    },
+    {
+      "epoch": 1.7673017982296333,
+      "grad_norm": 1.6640625,
+      "learning_rate": 7.186927884205712e-06,
+      "loss": 0.8794,
+      "step": 10377
+    },
+    {
+      "epoch": 1.7674732623186231,
+      "grad_norm": 1.75,
+      "learning_rate": 7.185194929330388e-06,
+      "loss": 0.8902,
+      "step": 10378
+    },
+    {
+      "epoch": 1.767644726407613,
+      "grad_norm": 1.6171875,
+      "learning_rate": 7.183462066256457e-06,
+      "loss": 0.8045,
+      "step": 10379
+    },
+    {
+      "epoch": 1.7678161904966028,
+      "grad_norm": 1.78125,
+      "learning_rate": 7.1817292950404325e-06,
+      "loss": 0.8288,
+      "step": 10380
+    },
+    {
+      "epoch": 1.7679876545855926,
+      "grad_norm": 1.8515625,
+      "learning_rate": 7.179996615738828e-06,
+      "loss": 0.8861,
+      "step": 10381
+    },
+    {
+      "epoch": 1.7681591186745826,
+      "grad_norm": 1.671875,
+      "learning_rate": 7.178264028408154e-06,
+      "loss": 0.8331,
+      "step": 10382
+    },
+    {
+      "epoch": 1.7683305827635725,
+      "grad_norm": 1.625,
+      "learning_rate": 7.176531533104916e-06,
+      "loss": 0.7918,
+      "step": 10383
+    },
+    {
+      "epoch": 1.7685020468525623,
+      "grad_norm": 1.6953125,
+      "learning_rate": 7.174799129885617e-06,
+      "loss": 0.8849,
+      "step": 10384
+    },
+    {
+      "epoch": 1.7686735109415523,
+      "grad_norm": 1.7734375,
+      "learning_rate": 7.173066818806762e-06,
+      "loss": 0.9,
+      "step": 10385
+    },
+    {
+      "epoch": 1.7688449750305422,
+      "grad_norm": 1.6640625,
+      "learning_rate": 7.171334599924837e-06,
+      "loss": 0.8639,
+      "step": 10386
+    },
+    {
+      "epoch": 1.769016439119532,
+      "grad_norm": 1.71875,
+      "learning_rate": 7.1696024732963445e-06,
+      "loss": 0.8004,
+      "step": 10387
+    },
+    {
+      "epoch": 1.7691879032085218,
+      "grad_norm": 1.6796875,
+      "learning_rate": 7.1678704389777735e-06,
+      "loss": 0.8962,
+      "step": 10388
+    },
+    {
+      "epoch": 1.7693593672975116,
+      "grad_norm": 1.6484375,
+      "learning_rate": 7.16613849702561e-06,
+      "loss": 0.8478,
+      "step": 10389
+    },
+    {
+      "epoch": 1.7695308313865015,
+      "grad_norm": 1.71875,
+      "learning_rate": 7.164406647496342e-06,
+      "loss": 0.8258,
+      "step": 10390
+    },
+    {
+      "epoch": 1.7697022954754913,
+      "grad_norm": 1.671875,
+      "learning_rate": 7.162674890446453e-06,
+      "loss": 0.8445,
+      "step": 10391
+    },
+    {
+      "epoch": 1.7698737595644811,
+      "grad_norm": 1.6484375,
+      "learning_rate": 7.16094322593242e-06,
+      "loss": 0.7893,
+      "step": 10392
+    },
+    {
+      "epoch": 1.770045223653471,
+      "grad_norm": 1.6484375,
+      "learning_rate": 7.1592116540107185e-06,
+      "loss": 0.8214,
+      "step": 10393
+    },
+    {
+      "epoch": 1.770216687742461,
+      "grad_norm": 1.7265625,
+      "learning_rate": 7.157480174737823e-06,
+      "loss": 0.8983,
+      "step": 10394
+    },
+    {
+      "epoch": 1.7703881518314508,
+      "grad_norm": 1.609375,
+      "learning_rate": 7.155748788170202e-06,
+      "loss": 0.8466,
+      "step": 10395
+    },
+    {
+      "epoch": 1.7705596159204406,
+      "grad_norm": 1.6640625,
+      "learning_rate": 7.154017494364329e-06,
+      "loss": 0.8173,
+      "step": 10396
+    },
+    {
+      "epoch": 1.7707310800094307,
+      "grad_norm": 1.609375,
+      "learning_rate": 7.1522862933766555e-06,
+      "loss": 0.817,
+      "step": 10397
+    },
+    {
+      "epoch": 1.7709025440984205,
+      "grad_norm": 1.6640625,
+      "learning_rate": 7.150555185263653e-06,
+      "loss": 0.8219,
+      "step": 10398
+    },
+    {
+      "epoch": 1.7710740081874103,
+      "grad_norm": 1.6328125,
+      "learning_rate": 7.148824170081774e-06,
+      "loss": 0.8267,
+      "step": 10399
+    },
+    {
+      "epoch": 1.7712454722764002,
+      "grad_norm": 1.734375,
+      "learning_rate": 7.147093247887476e-06,
+      "loss": 0.8318,
+      "step": 10400
+    },
+    {
+      "epoch": 1.77141693636539,
+      "grad_norm": 1.6171875,
+      "learning_rate": 7.145362418737209e-06,
+      "loss": 0.8338,
+      "step": 10401
+    },
+    {
+      "epoch": 1.7715884004543798,
+      "grad_norm": 1.7421875,
+      "learning_rate": 7.143631682687424e-06,
+      "loss": 0.9461,
+      "step": 10402
+    },
+    {
+      "epoch": 1.7717598645433696,
+      "grad_norm": 1.6796875,
+      "learning_rate": 7.141901039794566e-06,
+      "loss": 0.8248,
+      "step": 10403
+    },
+    {
+      "epoch": 1.7719313286323595,
+      "grad_norm": 1.7421875,
+      "learning_rate": 7.140170490115078e-06,
+      "loss": 0.8534,
+      "step": 10404
+    },
+    {
+      "epoch": 1.7721027927213493,
+      "grad_norm": 1.78125,
+      "learning_rate": 7.1384400337054e-06,
+      "loss": 0.8852,
+      "step": 10405
+    },
+    {
+      "epoch": 1.7722742568103391,
+      "grad_norm": 1.640625,
+      "learning_rate": 7.1367096706219665e-06,
+      "loss": 0.8271,
+      "step": 10406
+    },
+    {
+      "epoch": 1.7724457208993292,
+      "grad_norm": 1.7421875,
+      "learning_rate": 7.134979400921214e-06,
+      "loss": 0.7973,
+      "step": 10407
+    },
+    {
+      "epoch": 1.772617184988319,
+      "grad_norm": 1.8046875,
+      "learning_rate": 7.133249224659574e-06,
+      "loss": 0.7925,
+      "step": 10408
+    },
+    {
+      "epoch": 1.7727886490773088,
+      "grad_norm": 1.8046875,
+      "learning_rate": 7.131519141893469e-06,
+      "loss": 0.8688,
+      "step": 10409
+    },
+    {
+      "epoch": 1.7729601131662989,
+      "grad_norm": 1.6328125,
+      "learning_rate": 7.1297891526793276e-06,
+      "loss": 0.7748,
+      "step": 10410
+    },
+    {
+      "epoch": 1.7731315772552887,
+      "grad_norm": 1.671875,
+      "learning_rate": 7.128059257073569e-06,
+      "loss": 0.8827,
+      "step": 10411
+    },
+    {
+      "epoch": 1.7733030413442785,
+      "grad_norm": 1.6328125,
+      "learning_rate": 7.1263294551326145e-06,
+      "loss": 0.8463,
+      "step": 10412
+    },
+    {
+      "epoch": 1.7734745054332683,
+      "grad_norm": 1.75,
+      "learning_rate": 7.1245997469128754e-06,
+      "loss": 0.8997,
+      "step": 10413
+    },
+    {
+      "epoch": 1.7736459695222582,
+      "grad_norm": 1.6875,
+      "learning_rate": 7.122870132470769e-06,
+      "loss": 0.8661,
+      "step": 10414
+    },
+    {
+      "epoch": 1.773817433611248,
+      "grad_norm": 1.59375,
+      "learning_rate": 7.121140611862699e-06,
+      "loss": 0.7815,
+      "step": 10415
+    },
+    {
+      "epoch": 1.7739888977002378,
+      "grad_norm": 1.7109375,
+      "learning_rate": 7.119411185145075e-06,
+      "loss": 0.8995,
+      "step": 10416
+    },
+    {
+      "epoch": 1.7741603617892276,
+      "grad_norm": 1.6484375,
+      "learning_rate": 7.117681852374301e-06,
+      "loss": 0.9213,
+      "step": 10417
+    },
+    {
+      "epoch": 1.7743318258782175,
+      "grad_norm": 1.671875,
+      "learning_rate": 7.1159526136067755e-06,
+      "loss": 0.876,
+      "step": 10418
+    },
+    {
+      "epoch": 1.7745032899672075,
+      "grad_norm": 1.7421875,
+      "learning_rate": 7.114223468898897e-06,
+      "loss": 0.8722,
+      "step": 10419
+    },
+    {
+      "epoch": 1.7746747540561973,
+      "grad_norm": 1.7578125,
+      "learning_rate": 7.112494418307056e-06,
+      "loss": 0.8296,
+      "step": 10420
+    },
+    {
+      "epoch": 1.7748462181451872,
+      "grad_norm": 1.6875,
+      "learning_rate": 7.110765461887645e-06,
+      "loss": 0.807,
+      "step": 10421
+    },
+    {
+      "epoch": 1.7750176822341772,
+      "grad_norm": 1.6796875,
+      "learning_rate": 7.1090365996970526e-06,
+      "loss": 0.8302,
+      "step": 10422
+    },
+    {
+      "epoch": 1.775189146323167,
+      "grad_norm": 1.7421875,
+      "learning_rate": 7.107307831791663e-06,
+      "loss": 0.8044,
+      "step": 10423
+    },
+    {
+      "epoch": 1.7753606104121569,
+      "grad_norm": 1.6796875,
+      "learning_rate": 7.105579158227858e-06,
+      "loss": 0.8336,
+      "step": 10424
+    },
+    {
+      "epoch": 1.7755320745011467,
+      "grad_norm": 1.640625,
+      "learning_rate": 7.103850579062015e-06,
+      "loss": 0.8537,
+      "step": 10425
+    },
+    {
+      "epoch": 1.7757035385901365,
+      "grad_norm": 1.765625,
+      "learning_rate": 7.1021220943505124e-06,
+      "loss": 0.8888,
+      "step": 10426
+    },
+    {
+      "epoch": 1.7758750026791263,
+      "grad_norm": 1.7578125,
+      "learning_rate": 7.10039370414972e-06,
+      "loss": 0.877,
+      "step": 10427
+    },
+    {
+      "epoch": 1.7760464667681162,
+      "grad_norm": 1.6875,
+      "learning_rate": 7.098665408516004e-06,
+      "loss": 0.8352,
+      "step": 10428
+    },
+    {
+      "epoch": 1.776217930857106,
+      "grad_norm": 1.703125,
+      "learning_rate": 7.0969372075057385e-06,
+      "loss": 0.8651,
+      "step": 10429
+    },
+    {
+      "epoch": 1.7763893949460958,
+      "grad_norm": 1.6640625,
+      "learning_rate": 7.095209101175282e-06,
+      "loss": 0.8052,
+      "step": 10430
+    },
+    {
+      "epoch": 1.7765608590350859,
+      "grad_norm": 1.75,
+      "learning_rate": 7.093481089580999e-06,
+      "loss": 0.8881,
+      "step": 10431
+    },
+    {
+      "epoch": 1.7767323231240757,
+      "grad_norm": 1.640625,
+      "learning_rate": 7.0917531727792385e-06,
+      "loss": 0.9111,
+      "step": 10432
+    },
+    {
+      "epoch": 1.7769037872130655,
+      "grad_norm": 1.640625,
+      "learning_rate": 7.090025350826359e-06,
+      "loss": 0.8268,
+      "step": 10433
+    },
+    {
+      "epoch": 1.7770752513020556,
+      "grad_norm": 1.703125,
+      "learning_rate": 7.08829762377871e-06,
+      "loss": 0.7837,
+      "step": 10434
+    },
+    {
+      "epoch": 1.7772467153910454,
+      "grad_norm": 1.6796875,
+      "learning_rate": 7.086569991692641e-06,
+      "loss": 0.8403,
+      "step": 10435
+    },
+    {
+      "epoch": 1.7774181794800352,
+      "grad_norm": 1.703125,
+      "learning_rate": 7.084842454624493e-06,
+      "loss": 0.8457,
+      "step": 10436
+    },
+    {
+      "epoch": 1.777589643569025,
+      "grad_norm": 1.65625,
+      "learning_rate": 7.0831150126306125e-06,
+      "loss": 0.8311,
+      "step": 10437
+    },
+    {
+      "epoch": 1.7777611076580149,
+      "grad_norm": 1.71875,
+      "learning_rate": 7.081387665767334e-06,
+      "loss": 0.8601,
+      "step": 10438
+    },
+    {
+      "epoch": 1.7779325717470047,
+      "grad_norm": 1.6796875,
+      "learning_rate": 7.079660414090995e-06,
+      "loss": 0.8963,
+      "step": 10439
+    },
+    {
+      "epoch": 1.7781040358359945,
+      "grad_norm": 1.6953125,
+      "learning_rate": 7.077933257657927e-06,
+      "loss": 0.886,
+      "step": 10440
+    },
+    {
+      "epoch": 1.7782754999249843,
+      "grad_norm": 1.65625,
+      "learning_rate": 7.07620619652446e-06,
+      "loss": 0.8143,
+      "step": 10441
+    },
+    {
+      "epoch": 1.7784469640139742,
+      "grad_norm": 1.75,
+      "learning_rate": 7.074479230746921e-06,
+      "loss": 0.8514,
+      "step": 10442
+    },
+    {
+      "epoch": 1.7786184281029642,
+      "grad_norm": 1.7265625,
+      "learning_rate": 7.0727523603816276e-06,
+      "loss": 0.844,
+      "step": 10443
+    },
+    {
+      "epoch": 1.778789892191954,
+      "grad_norm": 1.671875,
+      "learning_rate": 7.071025585484901e-06,
+      "loss": 0.822,
+      "step": 10444
+    },
+    {
+      "epoch": 1.7789613562809439,
+      "grad_norm": 1.6328125,
+      "learning_rate": 7.069298906113061e-06,
+      "loss": 0.8756,
+      "step": 10445
+    },
+    {
+      "epoch": 1.779132820369934,
+      "grad_norm": 1.71875,
+      "learning_rate": 7.06757232232242e-06,
+      "loss": 0.8601,
+      "step": 10446
+    },
+    {
+      "epoch": 1.7793042844589237,
+      "grad_norm": 1.59375,
+      "learning_rate": 7.065845834169288e-06,
+      "loss": 0.8132,
+      "step": 10447
+    },
+    {
+      "epoch": 1.7794757485479136,
+      "grad_norm": 1.734375,
+      "learning_rate": 7.064119441709972e-06,
+      "loss": 0.885,
+      "step": 10448
+    },
+    {
+      "epoch": 1.7796472126369034,
+      "grad_norm": 1.6875,
+      "learning_rate": 7.062393145000776e-06,
+      "loss": 0.8066,
+      "step": 10449
+    },
+    {
+      "epoch": 1.7798186767258932,
+      "grad_norm": 1.6640625,
+      "learning_rate": 7.060666944098004e-06,
+      "loss": 0.8237,
+      "step": 10450
+    },
+    {
+      "epoch": 1.779990140814883,
+      "grad_norm": 1.65625,
+      "learning_rate": 7.05894083905795e-06,
+      "loss": 0.8584,
+      "step": 10451
+    },
+    {
+      "epoch": 1.7801616049038729,
+      "grad_norm": 1.6640625,
+      "learning_rate": 7.057214829936909e-06,
+      "loss": 0.8027,
+      "step": 10452
+    },
+    {
+      "epoch": 1.7803330689928627,
+      "grad_norm": 1.7734375,
+      "learning_rate": 7.055488916791176e-06,
+      "loss": 0.8564,
+      "step": 10453
+    },
+    {
+      "epoch": 1.7805045330818525,
+      "grad_norm": 1.71875,
+      "learning_rate": 7.053763099677038e-06,
+      "loss": 0.9454,
+      "step": 10454
+    },
+    {
+      "epoch": 1.7806759971708426,
+      "grad_norm": 1.703125,
+      "learning_rate": 7.052037378650778e-06,
+      "loss": 0.869,
+      "step": 10455
+    },
+    {
+      "epoch": 1.7808474612598324,
+      "grad_norm": 1.6640625,
+      "learning_rate": 7.050311753768681e-06,
+      "loss": 0.9125,
+      "step": 10456
+    },
+    {
+      "epoch": 1.7810189253488222,
+      "grad_norm": 1.671875,
+      "learning_rate": 7.048586225087024e-06,
+      "loss": 0.8282,
+      "step": 10457
+    },
+    {
+      "epoch": 1.7811903894378123,
+      "grad_norm": 1.640625,
+      "learning_rate": 7.0468607926620845e-06,
+      "loss": 0.7893,
+      "step": 10458
+    },
+    {
+      "epoch": 1.781361853526802,
+      "grad_norm": 1.6875,
+      "learning_rate": 7.0451354565501364e-06,
+      "loss": 0.9097,
+      "step": 10459
+    },
+    {
+      "epoch": 1.781533317615792,
+      "grad_norm": 1.65625,
+      "learning_rate": 7.043410216807447e-06,
+      "loss": 0.8195,
+      "step": 10460
+    },
+    {
+      "epoch": 1.7817047817047817,
+      "grad_norm": 1.6171875,
+      "learning_rate": 7.041685073490283e-06,
+      "loss": 0.8134,
+      "step": 10461
+    },
+    {
+      "epoch": 1.7818762457937716,
+      "grad_norm": 1.59375,
+      "learning_rate": 7.039960026654911e-06,
+      "loss": 0.7667,
+      "step": 10462
+    },
+    {
+      "epoch": 1.7820477098827614,
+      "grad_norm": 1.703125,
+      "learning_rate": 7.038235076357587e-06,
+      "loss": 0.8867,
+      "step": 10463
+    },
+    {
+      "epoch": 1.7822191739717512,
+      "grad_norm": 1.765625,
+      "learning_rate": 7.036510222654571e-06,
+      "loss": 0.8679,
+      "step": 10464
+    },
+    {
+      "epoch": 1.782390638060741,
+      "grad_norm": 1.703125,
+      "learning_rate": 7.034785465602118e-06,
+      "loss": 0.9346,
+      "step": 10465
+    },
+    {
+      "epoch": 1.7825621021497309,
+      "grad_norm": 1.625,
+      "learning_rate": 7.0330608052564744e-06,
+      "loss": 0.7896,
+      "step": 10466
+    },
+    {
+      "epoch": 1.782733566238721,
+      "grad_norm": 1.65625,
+      "learning_rate": 7.0313362416738905e-06,
+      "loss": 0.8179,
+      "step": 10467
+    },
+    {
+      "epoch": 1.7829050303277107,
+      "grad_norm": 1.640625,
+      "learning_rate": 7.0296117749106105e-06,
+      "loss": 0.8323,
+      "step": 10468
+    },
+    {
+      "epoch": 1.7830764944167006,
+      "grad_norm": 1.7265625,
+      "learning_rate": 7.0278874050228775e-06,
+      "loss": 0.8471,
+      "step": 10469
+    },
+    {
+      "epoch": 1.7832479585056906,
+      "grad_norm": 1.7890625,
+      "learning_rate": 7.026163132066927e-06,
+      "loss": 0.857,
+      "step": 10470
+    },
+    {
+      "epoch": 1.7834194225946804,
+      "grad_norm": 1.625,
+      "learning_rate": 7.024438956098996e-06,
+      "loss": 0.7765,
+      "step": 10471
+    },
+    {
+      "epoch": 1.7835908866836703,
+      "grad_norm": 1.7109375,
+      "learning_rate": 7.022714877175314e-06,
+      "loss": 0.8874,
+      "step": 10472
+    },
+    {
+      "epoch": 1.78376235077266,
+      "grad_norm": 1.6953125,
+      "learning_rate": 7.020990895352112e-06,
+      "loss": 0.8134,
+      "step": 10473
+    },
+    {
+      "epoch": 1.78393381486165,
+      "grad_norm": 1.5625,
+      "learning_rate": 7.019267010685615e-06,
+      "loss": 0.781,
+      "step": 10474
+    },
+    {
+      "epoch": 1.7841052789506398,
+      "grad_norm": 1.734375,
+      "learning_rate": 7.017543223232043e-06,
+      "loss": 0.8688,
+      "step": 10475
+    },
+    {
+      "epoch": 1.7842767430396296,
+      "grad_norm": 1.765625,
+      "learning_rate": 7.015819533047619e-06,
+      "loss": 0.886,
+      "step": 10476
+    },
+    {
+      "epoch": 1.7844482071286194,
+      "grad_norm": 1.7421875,
+      "learning_rate": 7.0140959401885635e-06,
+      "loss": 0.8504,
+      "step": 10477
+    },
+    {
+      "epoch": 1.7846196712176092,
+      "grad_norm": 1.703125,
+      "learning_rate": 7.012372444711078e-06,
+      "loss": 0.833,
+      "step": 10478
+    },
+    {
+      "epoch": 1.7847911353065993,
+      "grad_norm": 1.7734375,
+      "learning_rate": 7.010649046671376e-06,
+      "loss": 0.9199,
+      "step": 10479
+    },
+    {
+      "epoch": 1.784962599395589,
+      "grad_norm": 1.6875,
+      "learning_rate": 7.008925746125667e-06,
+      "loss": 0.8332,
+      "step": 10480
+    },
+    {
+      "epoch": 1.785134063484579,
+      "grad_norm": 1.6171875,
+      "learning_rate": 7.007202543130152e-06,
+      "loss": 0.8157,
+      "step": 10481
+    },
+    {
+      "epoch": 1.785305527573569,
+      "grad_norm": 1.671875,
+      "learning_rate": 7.005479437741032e-06,
+      "loss": 0.8156,
+      "step": 10482
+    },
+    {
+      "epoch": 1.7854769916625588,
+      "grad_norm": 1.71875,
+      "learning_rate": 7.003756430014502e-06,
+      "loss": 0.8972,
+      "step": 10483
+    },
+    {
+      "epoch": 1.7856484557515486,
+      "grad_norm": 1.6875,
+      "learning_rate": 7.00203352000676e-06,
+      "loss": 0.8747,
+      "step": 10484
+    },
+    {
+      "epoch": 1.7858199198405385,
+      "grad_norm": 1.7265625,
+      "learning_rate": 7.000310707773994e-06,
+      "loss": 0.9243,
+      "step": 10485
+    },
+    {
+      "epoch": 1.7859913839295283,
+      "grad_norm": 1.609375,
+      "learning_rate": 6.998587993372392e-06,
+      "loss": 0.7961,
+      "step": 10486
+    },
+    {
+      "epoch": 1.786162848018518,
+      "grad_norm": 1.6484375,
+      "learning_rate": 6.996865376858137e-06,
+      "loss": 0.8212,
+      "step": 10487
+    },
+    {
+      "epoch": 1.786334312107508,
+      "grad_norm": 1.765625,
+      "learning_rate": 6.995142858287416e-06,
+      "loss": 0.7968,
+      "step": 10488
+    },
+    {
+      "epoch": 1.7865057761964978,
+      "grad_norm": 1.8125,
+      "learning_rate": 6.993420437716396e-06,
+      "loss": 0.8075,
+      "step": 10489
+    },
+    {
+      "epoch": 1.7866772402854876,
+      "grad_norm": 1.6796875,
+      "learning_rate": 6.991698115201257e-06,
+      "loss": 0.8797,
+      "step": 10490
+    },
+    {
+      "epoch": 1.7868487043744776,
+      "grad_norm": 1.6328125,
+      "learning_rate": 6.9899758907981706e-06,
+      "loss": 0.818,
+      "step": 10491
+    },
+    {
+      "epoch": 1.7870201684634675,
+      "grad_norm": 1.6875,
+      "learning_rate": 6.988253764563307e-06,
+      "loss": 0.7987,
+      "step": 10492
+    },
+    {
+      "epoch": 1.7871916325524573,
+      "grad_norm": 1.703125,
+      "learning_rate": 6.986531736552829e-06,
+      "loss": 0.8601,
+      "step": 10493
+    },
+    {
+      "epoch": 1.787363096641447,
+      "grad_norm": 1.7265625,
+      "learning_rate": 6.984809806822897e-06,
+      "loss": 0.8766,
+      "step": 10494
+    },
+    {
+      "epoch": 1.7875345607304371,
+      "grad_norm": 1.65625,
+      "learning_rate": 6.983087975429673e-06,
+      "loss": 0.914,
+      "step": 10495
+    },
+    {
+      "epoch": 1.787706024819427,
+      "grad_norm": 1.8203125,
+      "learning_rate": 6.98136624242931e-06,
+      "loss": 0.9354,
+      "step": 10496
+    },
+    {
+      "epoch": 1.7878774889084168,
+      "grad_norm": 1.734375,
+      "learning_rate": 6.979644607877962e-06,
+      "loss": 0.9315,
+      "step": 10497
+    },
+    {
+      "epoch": 1.7880489529974066,
+      "grad_norm": 1.6953125,
+      "learning_rate": 6.977923071831776e-06,
+      "loss": 0.7592,
+      "step": 10498
+    },
+    {
+      "epoch": 1.7882204170863965,
+      "grad_norm": 1.84375,
+      "learning_rate": 6.9762016343469e-06,
+      "loss": 0.9585,
+      "step": 10499
+    },
+    {
+      "epoch": 1.7883918811753863,
+      "grad_norm": 1.734375,
+      "learning_rate": 6.974480295479476e-06,
+      "loss": 0.8544,
+      "step": 10500
+    },
+    {
+      "epoch": 1.7883918811753863,
+      "eval_loss": 0.8336145877838135,
+      "eval_runtime": 835.9359,
+      "eval_samples_per_second": 2.989,
+      "eval_steps_per_second": 2.989,
+      "step": 10500
+    },
+    {
+      "epoch": 1.788563345264376,
+      "grad_norm": 1.765625,
+      "learning_rate": 6.9727590552856405e-06,
+      "loss": 0.8694,
+      "step": 10501
+    },
+    {
+      "epoch": 1.788734809353366,
+      "grad_norm": 1.671875,
+      "learning_rate": 6.971037913821533e-06,
+      "loss": 0.7916,
+      "step": 10502
+    },
+    {
+      "epoch": 1.7889062734423558,
+      "grad_norm": 1.6875,
+      "learning_rate": 6.9693168711432835e-06,
+      "loss": 0.8967,
+      "step": 10503
+    },
+    {
+      "epoch": 1.7890777375313458,
+      "grad_norm": 1.671875,
+      "learning_rate": 6.967595927307025e-06,
+      "loss": 0.7497,
+      "step": 10504
+    },
+    {
+      "epoch": 1.7892492016203356,
+      "grad_norm": 1.65625,
+      "learning_rate": 6.96587508236888e-06,
+      "loss": 0.8398,
+      "step": 10505
+    },
+    {
+      "epoch": 1.7894206657093255,
+      "grad_norm": 1.8125,
+      "learning_rate": 6.964154336384976e-06,
+      "loss": 0.8664,
+      "step": 10506
+    },
+    {
+      "epoch": 1.7895921297983155,
+      "grad_norm": 1.6875,
+      "learning_rate": 6.962433689411431e-06,
+      "loss": 0.8373,
+      "step": 10507
+    },
+    {
+      "epoch": 1.7897635938873053,
+      "grad_norm": 1.6796875,
+      "learning_rate": 6.96071314150436e-06,
+      "loss": 0.8511,
+      "step": 10508
+    },
+    {
+      "epoch": 1.7899350579762952,
+      "grad_norm": 1.7109375,
+      "learning_rate": 6.9589926927198805e-06,
+      "loss": 0.8263,
+      "step": 10509
+    },
+    {
+      "epoch": 1.790106522065285,
+      "grad_norm": 1.7265625,
+      "learning_rate": 6.9572723431141e-06,
+      "loss": 0.934,
+      "step": 10510
+    },
+    {
+      "epoch": 1.7902779861542748,
+      "grad_norm": 1.71875,
+      "learning_rate": 6.955552092743125e-06,
+      "loss": 0.8711,
+      "step": 10511
+    },
+    {
+      "epoch": 1.7904494502432646,
+      "grad_norm": 1.65625,
+      "learning_rate": 6.953831941663065e-06,
+      "loss": 0.836,
+      "step": 10512
+    },
+    {
+      "epoch": 1.7906209143322545,
+      "grad_norm": 1.578125,
+      "learning_rate": 6.9521118899300135e-06,
+      "loss": 0.8665,
+      "step": 10513
+    },
+    {
+      "epoch": 1.7907923784212443,
+      "grad_norm": 1.671875,
+      "learning_rate": 6.95039193760007e-06,
+      "loss": 0.816,
+      "step": 10514
+    },
+    {
+      "epoch": 1.790963842510234,
+      "grad_norm": 1.7890625,
+      "learning_rate": 6.948672084729328e-06,
+      "loss": 0.8766,
+      "step": 10515
+    },
+    {
+      "epoch": 1.7911353065992242,
+      "grad_norm": 1.640625,
+      "learning_rate": 6.946952331373881e-06,
+      "loss": 0.802,
+      "step": 10516
+    },
+    {
+      "epoch": 1.791306770688214,
+      "grad_norm": 1.6953125,
+      "learning_rate": 6.945232677589815e-06,
+      "loss": 0.8039,
+      "step": 10517
+    },
+    {
+      "epoch": 1.7914782347772038,
+      "grad_norm": 1.734375,
+      "learning_rate": 6.943513123433214e-06,
+      "loss": 0.8979,
+      "step": 10518
+    },
+    {
+      "epoch": 1.7916496988661939,
+      "grad_norm": 1.734375,
+      "learning_rate": 6.94179366896016e-06,
+      "loss": 0.8817,
+      "step": 10519
+    },
+    {
+      "epoch": 1.7918211629551837,
+      "grad_norm": 1.6640625,
+      "learning_rate": 6.94007431422673e-06,
+      "loss": 0.8062,
+      "step": 10520
+    },
+    {
+      "epoch": 1.7919926270441735,
+      "grad_norm": 1.703125,
+      "learning_rate": 6.938355059289e-06,
+      "loss": 0.8331,
+      "step": 10521
+    },
+    {
+      "epoch": 1.7921640911331633,
+      "grad_norm": 1.6640625,
+      "learning_rate": 6.936635904203039e-06,
+      "loss": 0.812,
+      "step": 10522
+    },
+    {
+      "epoch": 1.7923355552221532,
+      "grad_norm": 1.671875,
+      "learning_rate": 6.934916849024922e-06,
+      "loss": 0.9065,
+      "step": 10523
+    },
+    {
+      "epoch": 1.792507019311143,
+      "grad_norm": 1.734375,
+      "learning_rate": 6.933197893810706e-06,
+      "loss": 0.8648,
+      "step": 10524
+    },
+    {
+      "epoch": 1.7926784834001328,
+      "grad_norm": 1.7578125,
+      "learning_rate": 6.931479038616453e-06,
+      "loss": 0.8835,
+      "step": 10525
+    },
+    {
+      "epoch": 1.7928499474891226,
+      "grad_norm": 1.6796875,
+      "learning_rate": 6.929760283498225e-06,
+      "loss": 0.893,
+      "step": 10526
+    },
+    {
+      "epoch": 1.7930214115781125,
+      "grad_norm": 1.65625,
+      "learning_rate": 6.928041628512074e-06,
+      "loss": 0.8337,
+      "step": 10527
+    },
+    {
+      "epoch": 1.7931928756671025,
+      "grad_norm": 1.7109375,
+      "learning_rate": 6.926323073714055e-06,
+      "loss": 0.8407,
+      "step": 10528
+    },
+    {
+      "epoch": 1.7933643397560923,
+      "grad_norm": 1.7265625,
+      "learning_rate": 6.924604619160214e-06,
+      "loss": 0.8472,
+      "step": 10529
+    },
+    {
+      "epoch": 1.7935358038450822,
+      "grad_norm": 1.5859375,
+      "learning_rate": 6.922886264906597e-06,
+      "loss": 0.8395,
+      "step": 10530
+    },
+    {
+      "epoch": 1.7937072679340722,
+      "grad_norm": 1.6640625,
+      "learning_rate": 6.921168011009247e-06,
+      "loss": 0.8467,
+      "step": 10531
+    },
+    {
+      "epoch": 1.793878732023062,
+      "grad_norm": 1.625,
+      "learning_rate": 6.919449857524201e-06,
+      "loss": 0.8281,
+      "step": 10532
+    },
+    {
+      "epoch": 1.7940501961120519,
+      "grad_norm": 1.71875,
+      "learning_rate": 6.917731804507497e-06,
+      "loss": 0.85,
+      "step": 10533
+    },
+    {
+      "epoch": 1.7942216602010417,
+      "grad_norm": 1.7421875,
+      "learning_rate": 6.916013852015165e-06,
+      "loss": 0.8706,
+      "step": 10534
+    },
+    {
+      "epoch": 1.7943931242900315,
+      "grad_norm": 1.71875,
+      "learning_rate": 6.914296000103238e-06,
+      "loss": 0.8551,
+      "step": 10535
+    },
+    {
+      "epoch": 1.7945645883790213,
+      "grad_norm": 1.7421875,
+      "learning_rate": 6.9125782488277345e-06,
+      "loss": 0.7838,
+      "step": 10536
+    },
+    {
+      "epoch": 1.7947360524680112,
+      "grad_norm": 1.6015625,
+      "learning_rate": 6.910860598244682e-06,
+      "loss": 0.8028,
+      "step": 10537
+    },
+    {
+      "epoch": 1.794907516557001,
+      "grad_norm": 1.6640625,
+      "learning_rate": 6.909143048410094e-06,
+      "loss": 0.8455,
+      "step": 10538
+    },
+    {
+      "epoch": 1.7950789806459908,
+      "grad_norm": 1.6640625,
+      "learning_rate": 6.907425599379993e-06,
+      "loss": 0.8057,
+      "step": 10539
+    },
+    {
+      "epoch": 1.7952504447349809,
+      "grad_norm": 1.7734375,
+      "learning_rate": 6.905708251210388e-06,
+      "loss": 0.788,
+      "step": 10540
+    },
+    {
+      "epoch": 1.7954219088239707,
+      "grad_norm": 1.8515625,
+      "learning_rate": 6.90399100395729e-06,
+      "loss": 0.8614,
+      "step": 10541
+    },
+    {
+      "epoch": 1.7955933729129605,
+      "grad_norm": 1.59375,
+      "learning_rate": 6.902273857676703e-06,
+      "loss": 0.8715,
+      "step": 10542
+    },
+    {
+      "epoch": 1.7957648370019506,
+      "grad_norm": 1.71875,
+      "learning_rate": 6.900556812424631e-06,
+      "loss": 0.9178,
+      "step": 10543
+    },
+    {
+      "epoch": 1.7959363010909404,
+      "grad_norm": 1.6796875,
+      "learning_rate": 6.898839868257072e-06,
+      "loss": 0.7955,
+      "step": 10544
+    },
+    {
+      "epoch": 1.7961077651799302,
+      "grad_norm": 1.703125,
+      "learning_rate": 6.897123025230023e-06,
+      "loss": 0.8021,
+      "step": 10545
+    },
+    {
+      "epoch": 1.79627922926892,
+      "grad_norm": 1.609375,
+      "learning_rate": 6.89540628339948e-06,
+      "loss": 0.7967,
+      "step": 10546
+    },
+    {
+      "epoch": 1.7964506933579099,
+      "grad_norm": 1.7578125,
+      "learning_rate": 6.893689642821426e-06,
+      "loss": 0.9277,
+      "step": 10547
+    },
+    {
+      "epoch": 1.7966221574468997,
+      "grad_norm": 1.703125,
+      "learning_rate": 6.89197310355185e-06,
+      "loss": 0.8216,
+      "step": 10548
+    },
+    {
+      "epoch": 1.7967936215358895,
+      "grad_norm": 1.625,
+      "learning_rate": 6.890256665646735e-06,
+      "loss": 0.7643,
+      "step": 10549
+    },
+    {
+      "epoch": 1.7969650856248793,
+      "grad_norm": 1.7890625,
+      "learning_rate": 6.88854032916206e-06,
+      "loss": 0.8166,
+      "step": 10550
+    },
+    {
+      "epoch": 1.7971365497138692,
+      "grad_norm": 1.609375,
+      "learning_rate": 6.886824094153801e-06,
+      "loss": 0.8987,
+      "step": 10551
+    },
+    {
+      "epoch": 1.7973080138028592,
+      "grad_norm": 1.71875,
+      "learning_rate": 6.885107960677933e-06,
+      "loss": 0.8656,
+      "step": 10552
+    },
+    {
+      "epoch": 1.797479477891849,
+      "grad_norm": 1.828125,
+      "learning_rate": 6.883391928790423e-06,
+      "loss": 0.8586,
+      "step": 10553
+    },
+    {
+      "epoch": 1.7976509419808389,
+      "grad_norm": 1.7421875,
+      "learning_rate": 6.881675998547238e-06,
+      "loss": 0.9042,
+      "step": 10554
+    },
+    {
+      "epoch": 1.797822406069829,
+      "grad_norm": 1.65625,
+      "learning_rate": 6.879960170004341e-06,
+      "loss": 0.852,
+      "step": 10555
+    },
+    {
+      "epoch": 1.7979938701588187,
+      "grad_norm": 1.75,
+      "learning_rate": 6.878244443217693e-06,
+      "loss": 0.8558,
+      "step": 10556
+    },
+    {
+      "epoch": 1.7981653342478086,
+      "grad_norm": 1.7265625,
+      "learning_rate": 6.876528818243249e-06,
+      "loss": 0.9334,
+      "step": 10557
+    },
+    {
+      "epoch": 1.7983367983367984,
+      "grad_norm": 1.71875,
+      "learning_rate": 6.874813295136964e-06,
+      "loss": 0.8312,
+      "step": 10558
+    },
+    {
+      "epoch": 1.7985082624257882,
+      "grad_norm": 1.7109375,
+      "learning_rate": 6.873097873954783e-06,
+      "loss": 0.7822,
+      "step": 10559
+    },
+    {
+      "epoch": 1.798679726514778,
+      "grad_norm": 1.59375,
+      "learning_rate": 6.871382554752655e-06,
+      "loss": 0.8809,
+      "step": 10560
+    },
+    {
+      "epoch": 1.7988511906037679,
+      "grad_norm": 1.640625,
+      "learning_rate": 6.869667337586524e-06,
+      "loss": 0.8643,
+      "step": 10561
+    },
+    {
+      "epoch": 1.7990226546927577,
+      "grad_norm": 1.75,
+      "learning_rate": 6.8679522225123286e-06,
+      "loss": 0.8073,
+      "step": 10562
+    },
+    {
+      "epoch": 1.7991941187817475,
+      "grad_norm": 1.625,
+      "learning_rate": 6.8662372095860055e-06,
+      "loss": 0.8247,
+      "step": 10563
+    },
+    {
+      "epoch": 1.7993655828707376,
+      "grad_norm": 1.6171875,
+      "learning_rate": 6.864522298863488e-06,
+      "loss": 0.8408,
+      "step": 10564
+    },
+    {
+      "epoch": 1.7995370469597274,
+      "grad_norm": 1.75,
+      "learning_rate": 6.862807490400705e-06,
+      "loss": 0.8944,
+      "step": 10565
+    },
+    {
+      "epoch": 1.7997085110487172,
+      "grad_norm": 1.7734375,
+      "learning_rate": 6.8610927842535825e-06,
+      "loss": 0.8626,
+      "step": 10566
+    },
+    {
+      "epoch": 1.7998799751377073,
+      "grad_norm": 1.671875,
+      "learning_rate": 6.859378180478046e-06,
+      "loss": 0.8183,
+      "step": 10567
+    },
+    {
+      "epoch": 1.800051439226697,
+      "grad_norm": 1.7421875,
+      "learning_rate": 6.857663679130013e-06,
+      "loss": 0.8266,
+      "step": 10568
+    },
+    {
+      "epoch": 1.800222903315687,
+      "grad_norm": 1.7109375,
+      "learning_rate": 6.855949280265402e-06,
+      "loss": 0.8345,
+      "step": 10569
+    },
+    {
+      "epoch": 1.8003943674046767,
+      "grad_norm": 1.7109375,
+      "learning_rate": 6.854234983940123e-06,
+      "loss": 0.8341,
+      "step": 10570
+    },
+    {
+      "epoch": 1.8005658314936666,
+      "grad_norm": 1.7734375,
+      "learning_rate": 6.8525207902100865e-06,
+      "loss": 0.8453,
+      "step": 10571
+    },
+    {
+      "epoch": 1.8007372955826564,
+      "grad_norm": 1.7421875,
+      "learning_rate": 6.850806699131198e-06,
+      "loss": 0.9124,
+      "step": 10572
+    },
+    {
+      "epoch": 1.8009087596716462,
+      "grad_norm": 1.625,
+      "learning_rate": 6.849092710759364e-06,
+      "loss": 0.826,
+      "step": 10573
+    },
+    {
+      "epoch": 1.801080223760636,
+      "grad_norm": 1.71875,
+      "learning_rate": 6.847378825150481e-06,
+      "loss": 0.9096,
+      "step": 10574
+    },
+    {
+      "epoch": 1.8012516878496259,
+      "grad_norm": 1.671875,
+      "learning_rate": 6.845665042360445e-06,
+      "loss": 0.8523,
+      "step": 10575
+    },
+    {
+      "epoch": 1.801423151938616,
+      "grad_norm": 1.6328125,
+      "learning_rate": 6.843951362445153e-06,
+      "loss": 0.916,
+      "step": 10576
+    },
+    {
+      "epoch": 1.8015946160276057,
+      "grad_norm": 1.7109375,
+      "learning_rate": 6.842237785460487e-06,
+      "loss": 0.8603,
+      "step": 10577
+    },
+    {
+      "epoch": 1.8017660801165956,
+      "grad_norm": 1.7265625,
+      "learning_rate": 6.840524311462341e-06,
+      "loss": 0.8717,
+      "step": 10578
+    },
+    {
+      "epoch": 1.8019375442055856,
+      "grad_norm": 1.703125,
+      "learning_rate": 6.838810940506595e-06,
+      "loss": 0.82,
+      "step": 10579
+    },
+    {
+      "epoch": 1.8021090082945754,
+      "grad_norm": 1.625,
+      "learning_rate": 6.837097672649126e-06,
+      "loss": 0.7833,
+      "step": 10580
+    },
+    {
+      "epoch": 1.8022804723835653,
+      "grad_norm": 1.8359375,
+      "learning_rate": 6.8353845079458195e-06,
+      "loss": 0.8223,
+      "step": 10581
+    },
+    {
+      "epoch": 1.802451936472555,
+      "grad_norm": 1.8203125,
+      "learning_rate": 6.833671446452535e-06,
+      "loss": 0.8815,
+      "step": 10582
+    },
+    {
+      "epoch": 1.802623400561545,
+      "grad_norm": 1.65625,
+      "learning_rate": 6.831958488225149e-06,
+      "loss": 0.8606,
+      "step": 10583
+    },
+    {
+      "epoch": 1.8027948646505347,
+      "grad_norm": 1.625,
+      "learning_rate": 6.8302456333195255e-06,
+      "loss": 0.8732,
+      "step": 10584
+    },
+    {
+      "epoch": 1.8029663287395246,
+      "grad_norm": 1.6640625,
+      "learning_rate": 6.828532881791528e-06,
+      "loss": 0.8402,
+      "step": 10585
+    },
+    {
+      "epoch": 1.8031377928285144,
+      "grad_norm": 1.6796875,
+      "learning_rate": 6.826820233697015e-06,
+      "loss": 0.8685,
+      "step": 10586
+    },
+    {
+      "epoch": 1.8033092569175042,
+      "grad_norm": 1.65625,
+      "learning_rate": 6.825107689091846e-06,
+      "loss": 0.8481,
+      "step": 10587
+    },
+    {
+      "epoch": 1.803480721006494,
+      "grad_norm": 1.65625,
+      "learning_rate": 6.823395248031867e-06,
+      "loss": 0.8237,
+      "step": 10588
+    },
+    {
+      "epoch": 1.803652185095484,
+      "grad_norm": 1.6875,
+      "learning_rate": 6.821682910572934e-06,
+      "loss": 0.9003,
+      "step": 10589
+    },
+    {
+      "epoch": 1.803823649184474,
+      "grad_norm": 1.6328125,
+      "learning_rate": 6.81997067677089e-06,
+      "loss": 0.8977,
+      "step": 10590
+    },
+    {
+      "epoch": 1.8039951132734637,
+      "grad_norm": 1.609375,
+      "learning_rate": 6.818258546681575e-06,
+      "loss": 0.7993,
+      "step": 10591
+    },
+    {
+      "epoch": 1.8041665773624538,
+      "grad_norm": 1.703125,
+      "learning_rate": 6.816546520360836e-06,
+      "loss": 0.9014,
+      "step": 10592
+    },
+    {
+      "epoch": 1.8043380414514436,
+      "grad_norm": 1.71875,
+      "learning_rate": 6.814834597864497e-06,
+      "loss": 0.8515,
+      "step": 10593
+    },
+    {
+      "epoch": 1.8045095055404334,
+      "grad_norm": 1.6953125,
+      "learning_rate": 6.8131227792483956e-06,
+      "loss": 0.9445,
+      "step": 10594
+    },
+    {
+      "epoch": 1.8046809696294233,
+      "grad_norm": 1.71875,
+      "learning_rate": 6.8114110645683625e-06,
+      "loss": 0.8309,
+      "step": 10595
+    },
+    {
+      "epoch": 1.804852433718413,
+      "grad_norm": 1.6015625,
+      "learning_rate": 6.809699453880223e-06,
+      "loss": 0.8086,
+      "step": 10596
+    },
+    {
+      "epoch": 1.805023897807403,
+      "grad_norm": 1.6875,
+      "learning_rate": 6.807987947239796e-06,
+      "loss": 0.9167,
+      "step": 10597
+    },
+    {
+      "epoch": 1.8051953618963927,
+      "grad_norm": 1.6484375,
+      "learning_rate": 6.806276544702902e-06,
+      "loss": 0.8983,
+      "step": 10598
+    },
+    {
+      "epoch": 1.8053668259853826,
+      "grad_norm": 1.59375,
+      "learning_rate": 6.804565246325358e-06,
+      "loss": 0.8876,
+      "step": 10599
+    },
+    {
+      "epoch": 1.8055382900743724,
+      "grad_norm": 1.625,
+      "learning_rate": 6.802854052162973e-06,
+      "loss": 0.846,
+      "step": 10600
+    },
+    {
+      "epoch": 1.8057097541633624,
+      "grad_norm": 1.6640625,
+      "learning_rate": 6.801142962271556e-06,
+      "loss": 0.8577,
+      "step": 10601
+    },
+    {
+      "epoch": 1.8058812182523523,
+      "grad_norm": 1.609375,
+      "learning_rate": 6.799431976706914e-06,
+      "loss": 0.8461,
+      "step": 10602
+    },
+    {
+      "epoch": 1.806052682341342,
+      "grad_norm": 1.5859375,
+      "learning_rate": 6.797721095524847e-06,
+      "loss": 0.8367,
+      "step": 10603
+    },
+    {
+      "epoch": 1.8062241464303321,
+      "grad_norm": 1.734375,
+      "learning_rate": 6.796010318781157e-06,
+      "loss": 0.8702,
+      "step": 10604
+    },
+    {
+      "epoch": 1.806395610519322,
+      "grad_norm": 1.7421875,
+      "learning_rate": 6.794299646531631e-06,
+      "loss": 0.8586,
+      "step": 10605
+    },
+    {
+      "epoch": 1.8065670746083118,
+      "grad_norm": 1.6015625,
+      "learning_rate": 6.792589078832066e-06,
+      "loss": 0.8048,
+      "step": 10606
+    },
+    {
+      "epoch": 1.8067385386973016,
+      "grad_norm": 1.75,
+      "learning_rate": 6.79087861573825e-06,
+      "loss": 0.8556,
+      "step": 10607
+    },
+    {
+      "epoch": 1.8069100027862914,
+      "grad_norm": 1.59375,
+      "learning_rate": 6.789168257305967e-06,
+      "loss": 0.8026,
+      "step": 10608
+    },
+    {
+      "epoch": 1.8070814668752813,
+      "grad_norm": 1.5859375,
+      "learning_rate": 6.787458003590996e-06,
+      "loss": 0.7729,
+      "step": 10609
+    },
+    {
+      "epoch": 1.807252930964271,
+      "grad_norm": 1.734375,
+      "learning_rate": 6.78574785464912e-06,
+      "loss": 0.8806,
+      "step": 10610
+    },
+    {
+      "epoch": 1.807424395053261,
+      "grad_norm": 1.6328125,
+      "learning_rate": 6.784037810536108e-06,
+      "loss": 0.86,
+      "step": 10611
+    },
+    {
+      "epoch": 1.8075958591422507,
+      "grad_norm": 1.8046875,
+      "learning_rate": 6.782327871307733e-06,
+      "loss": 0.9071,
+      "step": 10612
+    },
+    {
+      "epoch": 1.8077673232312408,
+      "grad_norm": 1.6640625,
+      "learning_rate": 6.780618037019764e-06,
+      "loss": 0.8872,
+      "step": 10613
+    },
+    {
+      "epoch": 1.8079387873202306,
+      "grad_norm": 1.6796875,
+      "learning_rate": 6.778908307727962e-06,
+      "loss": 0.8096,
+      "step": 10614
+    },
+    {
+      "epoch": 1.8081102514092204,
+      "grad_norm": 1.71875,
+      "learning_rate": 6.777198683488093e-06,
+      "loss": 0.8722,
+      "step": 10615
+    },
+    {
+      "epoch": 1.8082817154982105,
+      "grad_norm": 1.6328125,
+      "learning_rate": 6.7754891643559095e-06,
+      "loss": 0.8332,
+      "step": 10616
+    },
+    {
+      "epoch": 1.8084531795872003,
+      "grad_norm": 1.640625,
+      "learning_rate": 6.773779750387166e-06,
+      "loss": 0.7589,
+      "step": 10617
+    },
+    {
+      "epoch": 1.8086246436761901,
+      "grad_norm": 1.7421875,
+      "learning_rate": 6.772070441637614e-06,
+      "loss": 0.8632,
+      "step": 10618
+    },
+    {
+      "epoch": 1.80879610776518,
+      "grad_norm": 1.75,
+      "learning_rate": 6.7703612381629994e-06,
+      "loss": 0.8868,
+      "step": 10619
+    },
+    {
+      "epoch": 1.8089675718541698,
+      "grad_norm": 1.734375,
+      "learning_rate": 6.768652140019067e-06,
+      "loss": 0.8385,
+      "step": 10620
+    },
+    {
+      "epoch": 1.8091390359431596,
+      "grad_norm": 1.625,
+      "learning_rate": 6.7669431472615565e-06,
+      "loss": 0.8399,
+      "step": 10621
+    },
+    {
+      "epoch": 1.8093105000321494,
+      "grad_norm": 1.6796875,
+      "learning_rate": 6.765234259946204e-06,
+      "loss": 0.9131,
+      "step": 10622
+    },
+    {
+      "epoch": 1.8094819641211393,
+      "grad_norm": 1.78125,
+      "learning_rate": 6.763525478128744e-06,
+      "loss": 0.8054,
+      "step": 10623
+    },
+    {
+      "epoch": 1.809653428210129,
+      "grad_norm": 1.6640625,
+      "learning_rate": 6.7618168018649024e-06,
+      "loss": 0.8744,
+      "step": 10624
+    },
+    {
+      "epoch": 1.8098248922991191,
+      "grad_norm": 1.6640625,
+      "learning_rate": 6.760108231210411e-06,
+      "loss": 0.836,
+      "step": 10625
+    },
+    {
+      "epoch": 1.809996356388109,
+      "grad_norm": 1.703125,
+      "learning_rate": 6.758399766220992e-06,
+      "loss": 0.8886,
+      "step": 10626
+    },
+    {
+      "epoch": 1.8101678204770988,
+      "grad_norm": 1.78125,
+      "learning_rate": 6.756691406952367e-06,
+      "loss": 0.7741,
+      "step": 10627
+    },
+    {
+      "epoch": 1.8103392845660888,
+      "grad_norm": 1.703125,
+      "learning_rate": 6.754983153460244e-06,
+      "loss": 0.8985,
+      "step": 10628
+    },
+    {
+      "epoch": 1.8105107486550787,
+      "grad_norm": 1.7421875,
+      "learning_rate": 6.7532750058003395e-06,
+      "loss": 0.7809,
+      "step": 10629
+    },
+    {
+      "epoch": 1.8106822127440685,
+      "grad_norm": 1.7734375,
+      "learning_rate": 6.751566964028363e-06,
+      "loss": 0.9026,
+      "step": 10630
+    },
+    {
+      "epoch": 1.8108536768330583,
+      "grad_norm": 1.703125,
+      "learning_rate": 6.749859028200021e-06,
+      "loss": 0.8978,
+      "step": 10631
+    },
+    {
+      "epoch": 1.8110251409220481,
+      "grad_norm": 1.7421875,
+      "learning_rate": 6.7481511983710125e-06,
+      "loss": 0.8433,
+      "step": 10632
+    },
+    {
+      "epoch": 1.811196605011038,
+      "grad_norm": 1.640625,
+      "learning_rate": 6.746443474597041e-06,
+      "loss": 0.8048,
+      "step": 10633
+    },
+    {
+      "epoch": 1.8113680691000278,
+      "grad_norm": 1.7578125,
+      "learning_rate": 6.744735856933799e-06,
+      "loss": 0.8588,
+      "step": 10634
+    },
+    {
+      "epoch": 1.8115395331890176,
+      "grad_norm": 1.65625,
+      "learning_rate": 6.7430283454369795e-06,
+      "loss": 0.8494,
+      "step": 10635
+    },
+    {
+      "epoch": 1.8117109972780074,
+      "grad_norm": 1.6328125,
+      "learning_rate": 6.741320940162271e-06,
+      "loss": 0.8099,
+      "step": 10636
+    },
+    {
+      "epoch": 1.8118824613669975,
+      "grad_norm": 1.6640625,
+      "learning_rate": 6.7396136411653566e-06,
+      "loss": 0.827,
+      "step": 10637
+    },
+    {
+      "epoch": 1.8120539254559873,
+      "grad_norm": 1.671875,
+      "learning_rate": 6.737906448501923e-06,
+      "loss": 0.7802,
+      "step": 10638
+    },
+    {
+      "epoch": 1.8122253895449771,
+      "grad_norm": 1.65625,
+      "learning_rate": 6.73619936222764e-06,
+      "loss": 0.8083,
+      "step": 10639
+    },
+    {
+      "epoch": 1.8123968536339672,
+      "grad_norm": 1.75,
+      "learning_rate": 6.734492382398184e-06,
+      "loss": 0.8436,
+      "step": 10640
+    },
+    {
+      "epoch": 1.812568317722957,
+      "grad_norm": 1.8125,
+      "learning_rate": 6.7327855090692305e-06,
+      "loss": 0.845,
+      "step": 10641
+    },
+    {
+      "epoch": 1.8127397818119468,
+      "grad_norm": 1.6796875,
+      "learning_rate": 6.731078742296444e-06,
+      "loss": 0.8076,
+      "step": 10642
+    },
+    {
+      "epoch": 1.8129112459009367,
+      "grad_norm": 1.671875,
+      "learning_rate": 6.72937208213549e-06,
+      "loss": 0.8896,
+      "step": 10643
+    },
+    {
+      "epoch": 1.8130827099899265,
+      "grad_norm": 1.7109375,
+      "learning_rate": 6.727665528642029e-06,
+      "loss": 0.8489,
+      "step": 10644
+    },
+    {
+      "epoch": 1.8132541740789163,
+      "grad_norm": 1.6484375,
+      "learning_rate": 6.725959081871717e-06,
+      "loss": 0.8434,
+      "step": 10645
+    },
+    {
+      "epoch": 1.8134256381679061,
+      "grad_norm": 1.7109375,
+      "learning_rate": 6.724252741880208e-06,
+      "loss": 0.884,
+      "step": 10646
+    },
+    {
+      "epoch": 1.813597102256896,
+      "grad_norm": 1.71875,
+      "learning_rate": 6.722546508723154e-06,
+      "loss": 0.8877,
+      "step": 10647
+    },
+    {
+      "epoch": 1.8137685663458858,
+      "grad_norm": 1.75,
+      "learning_rate": 6.720840382456198e-06,
+      "loss": 0.8864,
+      "step": 10648
+    },
+    {
+      "epoch": 1.8139400304348758,
+      "grad_norm": 1.703125,
+      "learning_rate": 6.719134363134986e-06,
+      "loss": 0.8983,
+      "step": 10649
+    },
+    {
+      "epoch": 1.8141114945238657,
+      "grad_norm": 1.8125,
+      "learning_rate": 6.717428450815159e-06,
+      "loss": 0.8891,
+      "step": 10650
+    },
+    {
+      "epoch": 1.8142829586128555,
+      "grad_norm": 1.6953125,
+      "learning_rate": 6.715722645552349e-06,
+      "loss": 0.8517,
+      "step": 10651
+    },
+    {
+      "epoch": 1.8144544227018455,
+      "grad_norm": 1.546875,
+      "learning_rate": 6.714016947402191e-06,
+      "loss": 0.8189,
+      "step": 10652
+    },
+    {
+      "epoch": 1.8146258867908354,
+      "grad_norm": 1.7109375,
+      "learning_rate": 6.712311356420315e-06,
+      "loss": 0.7795,
+      "step": 10653
+    },
+    {
+      "epoch": 1.8147973508798252,
+      "grad_norm": 1.640625,
+      "learning_rate": 6.710605872662346e-06,
+      "loss": 0.8564,
+      "step": 10654
+    },
+    {
+      "epoch": 1.814968814968815,
+      "grad_norm": 1.6328125,
+      "learning_rate": 6.708900496183906e-06,
+      "loss": 0.8641,
+      "step": 10655
+    },
+    {
+      "epoch": 1.8151402790578048,
+      "grad_norm": 1.7578125,
+      "learning_rate": 6.707195227040612e-06,
+      "loss": 0.8355,
+      "step": 10656
+    },
+    {
+      "epoch": 1.8153117431467947,
+      "grad_norm": 1.71875,
+      "learning_rate": 6.705490065288083e-06,
+      "loss": 0.7409,
+      "step": 10657
+    },
+    {
+      "epoch": 1.8154832072357845,
+      "grad_norm": 1.75,
+      "learning_rate": 6.7037850109819294e-06,
+      "loss": 0.8776,
+      "step": 10658
+    },
+    {
+      "epoch": 1.8156546713247743,
+      "grad_norm": 1.78125,
+      "learning_rate": 6.702080064177759e-06,
+      "loss": 0.7984,
+      "step": 10659
+    },
+    {
+      "epoch": 1.8158261354137641,
+      "grad_norm": 1.7421875,
+      "learning_rate": 6.700375224931175e-06,
+      "loss": 0.8178,
+      "step": 10660
+    },
+    {
+      "epoch": 1.8159975995027542,
+      "grad_norm": 1.71875,
+      "learning_rate": 6.698670493297781e-06,
+      "loss": 0.7873,
+      "step": 10661
+    },
+    {
+      "epoch": 1.816169063591744,
+      "grad_norm": 1.703125,
+      "learning_rate": 6.696965869333176e-06,
+      "loss": 0.8951,
+      "step": 10662
+    },
+    {
+      "epoch": 1.8163405276807338,
+      "grad_norm": 1.71875,
+      "learning_rate": 6.69526135309295e-06,
+      "loss": 0.8174,
+      "step": 10663
+    },
+    {
+      "epoch": 1.8165119917697239,
+      "grad_norm": 1.6484375,
+      "learning_rate": 6.693556944632696e-06,
+      "loss": 0.7665,
+      "step": 10664
+    },
+    {
+      "epoch": 1.8166834558587137,
+      "grad_norm": 1.7265625,
+      "learning_rate": 6.691852644008e-06,
+      "loss": 0.8757,
+      "step": 10665
+    },
+    {
+      "epoch": 1.8168549199477035,
+      "grad_norm": 1.6796875,
+      "learning_rate": 6.690148451274447e-06,
+      "loss": 0.8875,
+      "step": 10666
+    },
+    {
+      "epoch": 1.8170263840366934,
+      "grad_norm": 1.625,
+      "learning_rate": 6.688444366487618e-06,
+      "loss": 0.7827,
+      "step": 10667
+    },
+    {
+      "epoch": 1.8171978481256832,
+      "grad_norm": 1.671875,
+      "learning_rate": 6.686740389703087e-06,
+      "loss": 0.8972,
+      "step": 10668
+    },
+    {
+      "epoch": 1.817369312214673,
+      "grad_norm": 1.6875,
+      "learning_rate": 6.685036520976429e-06,
+      "loss": 0.8712,
+      "step": 10669
+    },
+    {
+      "epoch": 1.8175407763036628,
+      "grad_norm": 1.6796875,
+      "learning_rate": 6.683332760363212e-06,
+      "loss": 0.8629,
+      "step": 10670
+    },
+    {
+      "epoch": 1.8177122403926527,
+      "grad_norm": 1.75,
+      "learning_rate": 6.681629107919002e-06,
+      "loss": 0.8683,
+      "step": 10671
+    },
+    {
+      "epoch": 1.8178837044816425,
+      "grad_norm": 1.7109375,
+      "learning_rate": 6.679925563699365e-06,
+      "loss": 0.8437,
+      "step": 10672
+    },
+    {
+      "epoch": 1.8180551685706325,
+      "grad_norm": 1.7734375,
+      "learning_rate": 6.678222127759862e-06,
+      "loss": 0.8247,
+      "step": 10673
+    },
+    {
+      "epoch": 1.8182266326596224,
+      "grad_norm": 1.7265625,
+      "learning_rate": 6.676518800156039e-06,
+      "loss": 0.9106,
+      "step": 10674
+    },
+    {
+      "epoch": 1.8183980967486122,
+      "grad_norm": 1.578125,
+      "learning_rate": 6.674815580943453e-06,
+      "loss": 0.8461,
+      "step": 10675
+    },
+    {
+      "epoch": 1.8185695608376022,
+      "grad_norm": 1.7421875,
+      "learning_rate": 6.673112470177653e-06,
+      "loss": 0.8636,
+      "step": 10676
+    },
+    {
+      "epoch": 1.818741024926592,
+      "grad_norm": 1.8203125,
+      "learning_rate": 6.671409467914183e-06,
+      "loss": 0.8077,
+      "step": 10677
+    },
+    {
+      "epoch": 1.818912489015582,
+      "grad_norm": 1.640625,
+      "learning_rate": 6.669706574208585e-06,
+      "loss": 0.8084,
+      "step": 10678
+    },
+    {
+      "epoch": 1.8190839531045717,
+      "grad_norm": 1.640625,
+      "learning_rate": 6.668003789116393e-06,
+      "loss": 0.789,
+      "step": 10679
+    },
+    {
+      "epoch": 1.8192554171935615,
+      "grad_norm": 1.59375,
+      "learning_rate": 6.666301112693147e-06,
+      "loss": 0.8396,
+      "step": 10680
+    },
+    {
+      "epoch": 1.8194268812825514,
+      "grad_norm": 1.7265625,
+      "learning_rate": 6.664598544994376e-06,
+      "loss": 0.9069,
+      "step": 10681
+    },
+    {
+      "epoch": 1.8195983453715412,
+      "grad_norm": 1.65625,
+      "learning_rate": 6.662896086075607e-06,
+      "loss": 0.8007,
+      "step": 10682
+    },
+    {
+      "epoch": 1.819769809460531,
+      "grad_norm": 1.640625,
+      "learning_rate": 6.6611937359923615e-06,
+      "loss": 0.8241,
+      "step": 10683
+    },
+    {
+      "epoch": 1.8199412735495208,
+      "grad_norm": 1.7109375,
+      "learning_rate": 6.659491494800161e-06,
+      "loss": 0.8853,
+      "step": 10684
+    },
+    {
+      "epoch": 1.8201127376385107,
+      "grad_norm": 1.71875,
+      "learning_rate": 6.657789362554527e-06,
+      "loss": 0.8538,
+      "step": 10685
+    },
+    {
+      "epoch": 1.8202842017275007,
+      "grad_norm": 1.703125,
+      "learning_rate": 6.656087339310964e-06,
+      "loss": 0.797,
+      "step": 10686
+    },
+    {
+      "epoch": 1.8204556658164905,
+      "grad_norm": 1.71875,
+      "learning_rate": 6.654385425124981e-06,
+      "loss": 0.8743,
+      "step": 10687
+    },
+    {
+      "epoch": 1.8206271299054804,
+      "grad_norm": 1.8203125,
+      "learning_rate": 6.652683620052092e-06,
+      "loss": 0.8449,
+      "step": 10688
+    },
+    {
+      "epoch": 1.8207985939944704,
+      "grad_norm": 1.6640625,
+      "learning_rate": 6.6509819241477924e-06,
+      "loss": 0.7926,
+      "step": 10689
+    },
+    {
+      "epoch": 1.8209700580834602,
+      "grad_norm": 1.765625,
+      "learning_rate": 6.649280337467585e-06,
+      "loss": 0.8931,
+      "step": 10690
+    },
+    {
+      "epoch": 1.82114152217245,
+      "grad_norm": 1.6875,
+      "learning_rate": 6.6475788600669635e-06,
+      "loss": 0.8234,
+      "step": 10691
+    },
+    {
+      "epoch": 1.82131298626144,
+      "grad_norm": 1.625,
+      "learning_rate": 6.645877492001419e-06,
+      "loss": 0.8517,
+      "step": 10692
+    },
+    {
+      "epoch": 1.8214844503504297,
+      "grad_norm": 1.640625,
+      "learning_rate": 6.64417623332644e-06,
+      "loss": 0.8323,
+      "step": 10693
+    },
+    {
+      "epoch": 1.8216559144394195,
+      "grad_norm": 1.75,
+      "learning_rate": 6.6424750840975115e-06,
+      "loss": 0.8254,
+      "step": 10694
+    },
+    {
+      "epoch": 1.8218273785284094,
+      "grad_norm": 1.5546875,
+      "learning_rate": 6.640774044370113e-06,
+      "loss": 0.7968,
+      "step": 10695
+    },
+    {
+      "epoch": 1.8219988426173992,
+      "grad_norm": 1.6796875,
+      "learning_rate": 6.6390731141997255e-06,
+      "loss": 0.7962,
+      "step": 10696
+    },
+    {
+      "epoch": 1.822170306706389,
+      "grad_norm": 1.7109375,
+      "learning_rate": 6.6373722936418176e-06,
+      "loss": 0.8424,
+      "step": 10697
+    },
+    {
+      "epoch": 1.822341770795379,
+      "grad_norm": 1.6875,
+      "learning_rate": 6.6356715827518615e-06,
+      "loss": 0.9017,
+      "step": 10698
+    },
+    {
+      "epoch": 1.822513234884369,
+      "grad_norm": 1.703125,
+      "learning_rate": 6.633970981585323e-06,
+      "loss": 0.8439,
+      "step": 10699
+    },
+    {
+      "epoch": 1.8226846989733587,
+      "grad_norm": 1.8203125,
+      "learning_rate": 6.632270490197667e-06,
+      "loss": 0.8899,
+      "step": 10700
+    },
+    {
+      "epoch": 1.8228561630623488,
+      "grad_norm": 1.671875,
+      "learning_rate": 6.630570108644352e-06,
+      "loss": 0.8732,
+      "step": 10701
+    },
+    {
+      "epoch": 1.8230276271513386,
+      "grad_norm": 1.578125,
+      "learning_rate": 6.628869836980833e-06,
+      "loss": 0.7281,
+      "step": 10702
+    },
+    {
+      "epoch": 1.8231990912403284,
+      "grad_norm": 1.65625,
+      "learning_rate": 6.627169675262563e-06,
+      "loss": 0.8485,
+      "step": 10703
+    },
+    {
+      "epoch": 1.8233705553293182,
+      "grad_norm": 1.6875,
+      "learning_rate": 6.625469623544994e-06,
+      "loss": 0.8677,
+      "step": 10704
+    },
+    {
+      "epoch": 1.823542019418308,
+      "grad_norm": 1.828125,
+      "learning_rate": 6.623769681883565e-06,
+      "loss": 0.8255,
+      "step": 10705
+    },
+    {
+      "epoch": 1.823713483507298,
+      "grad_norm": 1.796875,
+      "learning_rate": 6.62206985033372e-06,
+      "loss": 0.897,
+      "step": 10706
+    },
+    {
+      "epoch": 1.8238849475962877,
+      "grad_norm": 1.6875,
+      "learning_rate": 6.620370128950898e-06,
+      "loss": 0.8786,
+      "step": 10707
+    },
+    {
+      "epoch": 1.8240564116852775,
+      "grad_norm": 1.7265625,
+      "learning_rate": 6.618670517790535e-06,
+      "loss": 0.8853,
+      "step": 10708
+    },
+    {
+      "epoch": 1.8242278757742674,
+      "grad_norm": 1.796875,
+      "learning_rate": 6.6169710169080585e-06,
+      "loss": 0.8861,
+      "step": 10709
+    },
+    {
+      "epoch": 1.8243993398632574,
+      "grad_norm": 1.625,
+      "learning_rate": 6.615271626358895e-06,
+      "loss": 0.8048,
+      "step": 10710
+    },
+    {
+      "epoch": 1.8245708039522472,
+      "grad_norm": 1.71875,
+      "learning_rate": 6.613572346198469e-06,
+      "loss": 0.8505,
+      "step": 10711
+    },
+    {
+      "epoch": 1.824742268041237,
+      "grad_norm": 1.7109375,
+      "learning_rate": 6.611873176482201e-06,
+      "loss": 0.8112,
+      "step": 10712
+    },
+    {
+      "epoch": 1.8249137321302271,
+      "grad_norm": 1.671875,
+      "learning_rate": 6.610174117265507e-06,
+      "loss": 0.8226,
+      "step": 10713
+    },
+    {
+      "epoch": 1.825085196219217,
+      "grad_norm": 1.6015625,
+      "learning_rate": 6.608475168603801e-06,
+      "loss": 0.863,
+      "step": 10714
+    },
+    {
+      "epoch": 1.8252566603082068,
+      "grad_norm": 1.765625,
+      "learning_rate": 6.606776330552491e-06,
+      "loss": 0.8547,
+      "step": 10715
+    },
+    {
+      "epoch": 1.8254281243971966,
+      "grad_norm": 1.6328125,
+      "learning_rate": 6.605077603166981e-06,
+      "loss": 0.7697,
+      "step": 10716
+    },
+    {
+      "epoch": 1.8255995884861864,
+      "grad_norm": 1.65625,
+      "learning_rate": 6.603378986502677e-06,
+      "loss": 0.8732,
+      "step": 10717
+    },
+    {
+      "epoch": 1.8257710525751762,
+      "grad_norm": 1.59375,
+      "learning_rate": 6.6016804806149715e-06,
+      "loss": 0.8357,
+      "step": 10718
+    },
+    {
+      "epoch": 1.825942516664166,
+      "grad_norm": 1.7265625,
+      "learning_rate": 6.5999820855592675e-06,
+      "loss": 0.8082,
+      "step": 10719
+    },
+    {
+      "epoch": 1.826113980753156,
+      "grad_norm": 1.6875,
+      "learning_rate": 6.598283801390948e-06,
+      "loss": 0.9118,
+      "step": 10720
+    },
+    {
+      "epoch": 1.8262854448421457,
+      "grad_norm": 1.6796875,
+      "learning_rate": 6.596585628165404e-06,
+      "loss": 0.8519,
+      "step": 10721
+    },
+    {
+      "epoch": 1.8264569089311358,
+      "grad_norm": 1.5859375,
+      "learning_rate": 6.594887565938018e-06,
+      "loss": 0.7778,
+      "step": 10722
+    },
+    {
+      "epoch": 1.8266283730201256,
+      "grad_norm": 1.6796875,
+      "learning_rate": 6.593189614764171e-06,
+      "loss": 0.893,
+      "step": 10723
+    },
+    {
+      "epoch": 1.8267998371091154,
+      "grad_norm": 1.6640625,
+      "learning_rate": 6.591491774699239e-06,
+      "loss": 0.9023,
+      "step": 10724
+    },
+    {
+      "epoch": 1.8269713011981055,
+      "grad_norm": 1.7421875,
+      "learning_rate": 6.589794045798596e-06,
+      "loss": 0.8617,
+      "step": 10725
+    },
+    {
+      "epoch": 1.8271427652870953,
+      "grad_norm": 1.65625,
+      "learning_rate": 6.588096428117608e-06,
+      "loss": 0.8279,
+      "step": 10726
+    },
+    {
+      "epoch": 1.8273142293760851,
+      "grad_norm": 1.671875,
+      "learning_rate": 6.586398921711646e-06,
+      "loss": 0.8542,
+      "step": 10727
+    },
+    {
+      "epoch": 1.827485693465075,
+      "grad_norm": 1.734375,
+      "learning_rate": 6.5847015266360706e-06,
+      "loss": 0.8184,
+      "step": 10728
+    },
+    {
+      "epoch": 1.8276571575540648,
+      "grad_norm": 1.6796875,
+      "learning_rate": 6.5830042429462386e-06,
+      "loss": 0.81,
+      "step": 10729
+    },
+    {
+      "epoch": 1.8278286216430546,
+      "grad_norm": 1.7421875,
+      "learning_rate": 6.581307070697505e-06,
+      "loss": 0.8955,
+      "step": 10730
+    },
+    {
+      "epoch": 1.8280000857320444,
+      "grad_norm": 1.7265625,
+      "learning_rate": 6.579610009945225e-06,
+      "loss": 0.9194,
+      "step": 10731
+    },
+    {
+      "epoch": 1.8281715498210342,
+      "grad_norm": 1.59375,
+      "learning_rate": 6.57791306074474e-06,
+      "loss": 0.8302,
+      "step": 10732
+    },
+    {
+      "epoch": 1.828343013910024,
+      "grad_norm": 1.625,
+      "learning_rate": 6.576216223151395e-06,
+      "loss": 0.8272,
+      "step": 10733
+    },
+    {
+      "epoch": 1.8285144779990141,
+      "grad_norm": 1.796875,
+      "learning_rate": 6.57451949722053e-06,
+      "loss": 0.8894,
+      "step": 10734
+    },
+    {
+      "epoch": 1.828685942088004,
+      "grad_norm": 1.734375,
+      "learning_rate": 6.572822883007486e-06,
+      "loss": 0.932,
+      "step": 10735
+    },
+    {
+      "epoch": 1.8288574061769938,
+      "grad_norm": 1.640625,
+      "learning_rate": 6.571126380567594e-06,
+      "loss": 0.8385,
+      "step": 10736
+    },
+    {
+      "epoch": 1.8290288702659838,
+      "grad_norm": 1.7265625,
+      "learning_rate": 6.569429989956182e-06,
+      "loss": 0.8257,
+      "step": 10737
+    },
+    {
+      "epoch": 1.8292003343549736,
+      "grad_norm": 1.796875,
+      "learning_rate": 6.567733711228577e-06,
+      "loss": 0.9127,
+      "step": 10738
+    },
+    {
+      "epoch": 1.8293717984439635,
+      "grad_norm": 1.7265625,
+      "learning_rate": 6.566037544440098e-06,
+      "loss": 0.8548,
+      "step": 10739
+    },
+    {
+      "epoch": 1.8295432625329533,
+      "grad_norm": 1.8125,
+      "learning_rate": 6.564341489646068e-06,
+      "loss": 0.9063,
+      "step": 10740
+    },
+    {
+      "epoch": 1.8297147266219431,
+      "grad_norm": 1.75,
+      "learning_rate": 6.562645546901798e-06,
+      "loss": 0.8873,
+      "step": 10741
+    },
+    {
+      "epoch": 1.829886190710933,
+      "grad_norm": 1.6171875,
+      "learning_rate": 6.5609497162626044e-06,
+      "loss": 0.8054,
+      "step": 10742
+    },
+    {
+      "epoch": 1.8300576547999228,
+      "grad_norm": 1.71875,
+      "learning_rate": 6.5592539977837875e-06,
+      "loss": 0.8719,
+      "step": 10743
+    },
+    {
+      "epoch": 1.8302291188889126,
+      "grad_norm": 1.609375,
+      "learning_rate": 6.557558391520655e-06,
+      "loss": 0.793,
+      "step": 10744
+    },
+    {
+      "epoch": 1.8304005829779024,
+      "grad_norm": 1.6953125,
+      "learning_rate": 6.555862897528507e-06,
+      "loss": 0.8555,
+      "step": 10745
+    },
+    {
+      "epoch": 1.8305720470668925,
+      "grad_norm": 1.671875,
+      "learning_rate": 6.554167515862637e-06,
+      "loss": 0.8055,
+      "step": 10746
+    },
+    {
+      "epoch": 1.8307435111558823,
+      "grad_norm": 1.7890625,
+      "learning_rate": 6.552472246578343e-06,
+      "loss": 0.8477,
+      "step": 10747
+    },
+    {
+      "epoch": 1.8309149752448721,
+      "grad_norm": 1.734375,
+      "learning_rate": 6.55077708973091e-06,
+      "loss": 0.8425,
+      "step": 10748
+    },
+    {
+      "epoch": 1.8310864393338622,
+      "grad_norm": 1.78125,
+      "learning_rate": 6.549082045375624e-06,
+      "loss": 0.8409,
+      "step": 10749
+    },
+    {
+      "epoch": 1.831257903422852,
+      "grad_norm": 1.6484375,
+      "learning_rate": 6.547387113567768e-06,
+      "loss": 0.8289,
+      "step": 10750
+    },
+    {
+      "epoch": 1.8314293675118418,
+      "grad_norm": 1.75,
+      "learning_rate": 6.54569229436262e-06,
+      "loss": 0.9509,
+      "step": 10751
+    },
+    {
+      "epoch": 1.8316008316008316,
+      "grad_norm": 1.6796875,
+      "learning_rate": 6.543997587815454e-06,
+      "loss": 0.8315,
+      "step": 10752
+    },
+    {
+      "epoch": 1.8317722956898215,
+      "grad_norm": 1.6875,
+      "learning_rate": 6.5423029939815394e-06,
+      "loss": 0.8989,
+      "step": 10753
+    },
+    {
+      "epoch": 1.8319437597788113,
+      "grad_norm": 1.6796875,
+      "learning_rate": 6.5406085129161475e-06,
+      "loss": 0.8576,
+      "step": 10754
+    },
+    {
+      "epoch": 1.8321152238678011,
+      "grad_norm": 1.7734375,
+      "learning_rate": 6.538914144674537e-06,
+      "loss": 0.8601,
+      "step": 10755
+    },
+    {
+      "epoch": 1.832286687956791,
+      "grad_norm": 1.671875,
+      "learning_rate": 6.53721988931197e-06,
+      "loss": 0.8074,
+      "step": 10756
+    },
+    {
+      "epoch": 1.8324581520457808,
+      "grad_norm": 1.7578125,
+      "learning_rate": 6.535525746883702e-06,
+      "loss": 0.8987,
+      "step": 10757
+    },
+    {
+      "epoch": 1.8326296161347708,
+      "grad_norm": 1.6328125,
+      "learning_rate": 6.533831717444984e-06,
+      "loss": 0.7944,
+      "step": 10758
+    },
+    {
+      "epoch": 1.8328010802237606,
+      "grad_norm": 1.7578125,
+      "learning_rate": 6.5321378010510675e-06,
+      "loss": 0.8961,
+      "step": 10759
+    },
+    {
+      "epoch": 1.8329725443127505,
+      "grad_norm": 1.7421875,
+      "learning_rate": 6.5304439977571945e-06,
+      "loss": 0.8753,
+      "step": 10760
+    },
+    {
+      "epoch": 1.8331440084017405,
+      "grad_norm": 1.609375,
+      "learning_rate": 6.52875030761861e-06,
+      "loss": 0.825,
+      "step": 10761
+    },
+    {
+      "epoch": 1.8333154724907303,
+      "grad_norm": 1.5859375,
+      "learning_rate": 6.527056730690547e-06,
+      "loss": 0.8377,
+      "step": 10762
+    },
+    {
+      "epoch": 1.8334869365797202,
+      "grad_norm": 1.59375,
+      "learning_rate": 6.525363267028244e-06,
+      "loss": 0.7996,
+      "step": 10763
+    },
+    {
+      "epoch": 1.83365840066871,
+      "grad_norm": 1.671875,
+      "learning_rate": 6.523669916686928e-06,
+      "loss": 0.8381,
+      "step": 10764
+    },
+    {
+      "epoch": 1.8338298647576998,
+      "grad_norm": 1.8125,
+      "learning_rate": 6.521976679721829e-06,
+      "loss": 0.8257,
+      "step": 10765
+    },
+    {
+      "epoch": 1.8340013288466896,
+      "grad_norm": 1.6484375,
+      "learning_rate": 6.520283556188166e-06,
+      "loss": 0.8644,
+      "step": 10766
+    },
+    {
+      "epoch": 1.8341727929356795,
+      "grad_norm": 1.7578125,
+      "learning_rate": 6.518590546141159e-06,
+      "loss": 0.8382,
+      "step": 10767
+    },
+    {
+      "epoch": 1.8343442570246693,
+      "grad_norm": 1.7578125,
+      "learning_rate": 6.516897649636025e-06,
+      "loss": 0.8607,
+      "step": 10768
+    },
+    {
+      "epoch": 1.8345157211136591,
+      "grad_norm": 1.765625,
+      "learning_rate": 6.5152048667279735e-06,
+      "loss": 0.8957,
+      "step": 10769
+    },
+    {
+      "epoch": 1.8346871852026492,
+      "grad_norm": 1.8125,
+      "learning_rate": 6.513512197472214e-06,
+      "loss": 0.9375,
+      "step": 10770
+    },
+    {
+      "epoch": 1.834858649291639,
+      "grad_norm": 1.703125,
+      "learning_rate": 6.511819641923951e-06,
+      "loss": 0.8961,
+      "step": 10771
+    },
+    {
+      "epoch": 1.8350301133806288,
+      "grad_norm": 1.6171875,
+      "learning_rate": 6.510127200138385e-06,
+      "loss": 0.8588,
+      "step": 10772
+    },
+    {
+      "epoch": 1.8352015774696189,
+      "grad_norm": 1.7890625,
+      "learning_rate": 6.50843487217071e-06,
+      "loss": 0.8759,
+      "step": 10773
+    },
+    {
+      "epoch": 1.8353730415586087,
+      "grad_norm": 1.6796875,
+      "learning_rate": 6.506742658076124e-06,
+      "loss": 0.8354,
+      "step": 10774
+    },
+    {
+      "epoch": 1.8355445056475985,
+      "grad_norm": 1.7265625,
+      "learning_rate": 6.505050557909816e-06,
+      "loss": 0.7801,
+      "step": 10775
+    },
+    {
+      "epoch": 1.8357159697365883,
+      "grad_norm": 1.859375,
+      "learning_rate": 6.503358571726968e-06,
+      "loss": 0.8814,
+      "step": 10776
+    },
+    {
+      "epoch": 1.8358874338255782,
+      "grad_norm": 1.6953125,
+      "learning_rate": 6.501666699582769e-06,
+      "loss": 0.8261,
+      "step": 10777
+    },
+    {
+      "epoch": 1.836058897914568,
+      "grad_norm": 1.828125,
+      "learning_rate": 6.499974941532387e-06,
+      "loss": 0.8936,
+      "step": 10778
+    },
+    {
+      "epoch": 1.8362303620035578,
+      "grad_norm": 1.75,
+      "learning_rate": 6.498283297631004e-06,
+      "loss": 0.8725,
+      "step": 10779
+    },
+    {
+      "epoch": 1.8364018260925477,
+      "grad_norm": 1.7421875,
+      "learning_rate": 6.496591767933789e-06,
+      "loss": 0.7837,
+      "step": 10780
+    },
+    {
+      "epoch": 1.8365732901815375,
+      "grad_norm": 1.71875,
+      "learning_rate": 6.4949003524959055e-06,
+      "loss": 0.8273,
+      "step": 10781
+    },
+    {
+      "epoch": 1.8367447542705273,
+      "grad_norm": 1.8046875,
+      "learning_rate": 6.493209051372522e-06,
+      "loss": 0.8467,
+      "step": 10782
+    },
+    {
+      "epoch": 1.8369162183595173,
+      "grad_norm": 1.71875,
+      "learning_rate": 6.491517864618799e-06,
+      "loss": 0.8545,
+      "step": 10783
+    },
+    {
+      "epoch": 1.8370876824485072,
+      "grad_norm": 1.6484375,
+      "learning_rate": 6.4898267922898884e-06,
+      "loss": 0.8044,
+      "step": 10784
+    },
+    {
+      "epoch": 1.837259146537497,
+      "grad_norm": 1.6953125,
+      "learning_rate": 6.488135834440945e-06,
+      "loss": 0.8334,
+      "step": 10785
+    },
+    {
+      "epoch": 1.837430610626487,
+      "grad_norm": 1.6875,
+      "learning_rate": 6.4864449911271165e-06,
+      "loss": 0.884,
+      "step": 10786
+    },
+    {
+      "epoch": 1.8376020747154769,
+      "grad_norm": 1.8046875,
+      "learning_rate": 6.484754262403547e-06,
+      "loss": 0.8961,
+      "step": 10787
+    },
+    {
+      "epoch": 1.8377735388044667,
+      "grad_norm": 1.5859375,
+      "learning_rate": 6.483063648325383e-06,
+      "loss": 0.7778,
+      "step": 10788
+    },
+    {
+      "epoch": 1.8379450028934565,
+      "grad_norm": 1.7421875,
+      "learning_rate": 6.48137314894775e-06,
+      "loss": 0.8881,
+      "step": 10789
+    },
+    {
+      "epoch": 1.8381164669824464,
+      "grad_norm": 1.7109375,
+      "learning_rate": 6.479682764325792e-06,
+      "loss": 0.8727,
+      "step": 10790
+    },
+    {
+      "epoch": 1.8382879310714362,
+      "grad_norm": 1.6953125,
+      "learning_rate": 6.477992494514633e-06,
+      "loss": 0.8266,
+      "step": 10791
+    },
+    {
+      "epoch": 1.838459395160426,
+      "grad_norm": 1.703125,
+      "learning_rate": 6.476302339569405e-06,
+      "loss": 0.8603,
+      "step": 10792
+    },
+    {
+      "epoch": 1.8386308592494158,
+      "grad_norm": 1.6796875,
+      "learning_rate": 6.474612299545225e-06,
+      "loss": 0.7457,
+      "step": 10793
+    },
+    {
+      "epoch": 1.8388023233384057,
+      "grad_norm": 1.6796875,
+      "learning_rate": 6.472922374497211e-06,
+      "loss": 0.8563,
+      "step": 10794
+    },
+    {
+      "epoch": 1.8389737874273957,
+      "grad_norm": 1.59375,
+      "learning_rate": 6.471232564480483e-06,
+      "loss": 0.7924,
+      "step": 10795
+    },
+    {
+      "epoch": 1.8391452515163855,
+      "grad_norm": 1.6875,
+      "learning_rate": 6.469542869550147e-06,
+      "loss": 0.883,
+      "step": 10796
+    },
+    {
+      "epoch": 1.8393167156053754,
+      "grad_norm": 1.6640625,
+      "learning_rate": 6.467853289761315e-06,
+      "loss": 0.9064,
+      "step": 10797
+    },
+    {
+      "epoch": 1.8394881796943654,
+      "grad_norm": 1.7265625,
+      "learning_rate": 6.466163825169086e-06,
+      "loss": 0.9077,
+      "step": 10798
+    },
+    {
+      "epoch": 1.8396596437833552,
+      "grad_norm": 1.6875,
+      "learning_rate": 6.464474475828563e-06,
+      "loss": 0.8597,
+      "step": 10799
+    },
+    {
+      "epoch": 1.839831107872345,
+      "grad_norm": 1.71875,
+      "learning_rate": 6.4627852417948425e-06,
+      "loss": 0.8379,
+      "step": 10800
+    },
+    {
+      "epoch": 1.8400025719613349,
+      "grad_norm": 1.7421875,
+      "learning_rate": 6.461096123123013e-06,
+      "loss": 0.8876,
+      "step": 10801
+    },
+    {
+      "epoch": 1.8401740360503247,
+      "grad_norm": 1.7109375,
+      "learning_rate": 6.459407119868165e-06,
+      "loss": 0.8891,
+      "step": 10802
+    },
+    {
+      "epoch": 1.8403455001393145,
+      "grad_norm": 1.6640625,
+      "learning_rate": 6.4577182320853836e-06,
+      "loss": 0.8582,
+      "step": 10803
+    },
+    {
+      "epoch": 1.8405169642283044,
+      "grad_norm": 1.6796875,
+      "learning_rate": 6.456029459829751e-06,
+      "loss": 0.8022,
+      "step": 10804
+    },
+    {
+      "epoch": 1.8406884283172942,
+      "grad_norm": 1.71875,
+      "learning_rate": 6.454340803156341e-06,
+      "loss": 0.8657,
+      "step": 10805
+    },
+    {
+      "epoch": 1.840859892406284,
+      "grad_norm": 1.6171875,
+      "learning_rate": 6.452652262120231e-06,
+      "loss": 0.8819,
+      "step": 10806
+    },
+    {
+      "epoch": 1.841031356495274,
+      "grad_norm": 1.6484375,
+      "learning_rate": 6.450963836776487e-06,
+      "loss": 0.816,
+      "step": 10807
+    },
+    {
+      "epoch": 1.8412028205842639,
+      "grad_norm": 1.7578125,
+      "learning_rate": 6.449275527180178e-06,
+      "loss": 0.9231,
+      "step": 10808
+    },
+    {
+      "epoch": 1.8413742846732537,
+      "grad_norm": 1.7578125,
+      "learning_rate": 6.447587333386365e-06,
+      "loss": 0.9352,
+      "step": 10809
+    },
+    {
+      "epoch": 1.8415457487622438,
+      "grad_norm": 1.65625,
+      "learning_rate": 6.445899255450106e-06,
+      "loss": 0.8779,
+      "step": 10810
+    },
+    {
+      "epoch": 1.8417172128512336,
+      "grad_norm": 1.671875,
+      "learning_rate": 6.444211293426457e-06,
+      "loss": 0.8487,
+      "step": 10811
+    },
+    {
+      "epoch": 1.8418886769402234,
+      "grad_norm": 1.7421875,
+      "learning_rate": 6.442523447370466e-06,
+      "loss": 0.8554,
+      "step": 10812
+    },
+    {
+      "epoch": 1.8420601410292132,
+      "grad_norm": 1.7109375,
+      "learning_rate": 6.440835717337182e-06,
+      "loss": 0.9101,
+      "step": 10813
+    },
+    {
+      "epoch": 1.842231605118203,
+      "grad_norm": 1.796875,
+      "learning_rate": 6.439148103381647e-06,
+      "loss": 0.8077,
+      "step": 10814
+    },
+    {
+      "epoch": 1.8424030692071929,
+      "grad_norm": 1.78125,
+      "learning_rate": 6.437460605558902e-06,
+      "loss": 0.8836,
+      "step": 10815
+    },
+    {
+      "epoch": 1.8425745332961827,
+      "grad_norm": 1.640625,
+      "learning_rate": 6.435773223923982e-06,
+      "loss": 0.8436,
+      "step": 10816
+    },
+    {
+      "epoch": 1.8427459973851725,
+      "grad_norm": 1.8125,
+      "learning_rate": 6.434085958531918e-06,
+      "loss": 0.8848,
+      "step": 10817
+    },
+    {
+      "epoch": 1.8429174614741624,
+      "grad_norm": 1.6953125,
+      "learning_rate": 6.432398809437739e-06,
+      "loss": 0.8224,
+      "step": 10818
+    },
+    {
+      "epoch": 1.8430889255631524,
+      "grad_norm": 1.6796875,
+      "learning_rate": 6.43071177669647e-06,
+      "loss": 0.8838,
+      "step": 10819
+    },
+    {
+      "epoch": 1.8432603896521422,
+      "grad_norm": 1.6953125,
+      "learning_rate": 6.429024860363128e-06,
+      "loss": 0.8179,
+      "step": 10820
+    },
+    {
+      "epoch": 1.843431853741132,
+      "grad_norm": 1.6796875,
+      "learning_rate": 6.427338060492734e-06,
+      "loss": 0.8074,
+      "step": 10821
+    },
+    {
+      "epoch": 1.843603317830122,
+      "grad_norm": 1.71875,
+      "learning_rate": 6.4256513771403e-06,
+      "loss": 0.9034,
+      "step": 10822
+    },
+    {
+      "epoch": 1.843774781919112,
+      "grad_norm": 1.7578125,
+      "learning_rate": 6.4239648103608384e-06,
+      "loss": 0.8656,
+      "step": 10823
+    },
+    {
+      "epoch": 1.8439462460081018,
+      "grad_norm": 1.7265625,
+      "learning_rate": 6.4222783602093465e-06,
+      "loss": 0.8631,
+      "step": 10824
+    },
+    {
+      "epoch": 1.8441177100970916,
+      "grad_norm": 1.734375,
+      "learning_rate": 6.420592026740829e-06,
+      "loss": 0.7848,
+      "step": 10825
+    },
+    {
+      "epoch": 1.8442891741860814,
+      "grad_norm": 1.6484375,
+      "learning_rate": 6.418905810010285e-06,
+      "loss": 0.7932,
+      "step": 10826
+    },
+    {
+      "epoch": 1.8444606382750712,
+      "grad_norm": 1.71875,
+      "learning_rate": 6.4172197100727075e-06,
+      "loss": 0.7858,
+      "step": 10827
+    },
+    {
+      "epoch": 1.844632102364061,
+      "grad_norm": 1.671875,
+      "learning_rate": 6.4155337269830855e-06,
+      "loss": 0.8253,
+      "step": 10828
+    },
+    {
+      "epoch": 1.8448035664530509,
+      "grad_norm": 1.7265625,
+      "learning_rate": 6.413847860796407e-06,
+      "loss": 0.8208,
+      "step": 10829
+    },
+    {
+      "epoch": 1.8449750305420407,
+      "grad_norm": 1.6484375,
+      "learning_rate": 6.412162111567655e-06,
+      "loss": 0.8985,
+      "step": 10830
+    },
+    {
+      "epoch": 1.8451464946310308,
+      "grad_norm": 1.6640625,
+      "learning_rate": 6.410476479351806e-06,
+      "loss": 0.8778,
+      "step": 10831
+    },
+    {
+      "epoch": 1.8453179587200206,
+      "grad_norm": 1.6171875,
+      "learning_rate": 6.408790964203837e-06,
+      "loss": 0.8216,
+      "step": 10832
+    },
+    {
+      "epoch": 1.8454894228090104,
+      "grad_norm": 1.7265625,
+      "learning_rate": 6.407105566178717e-06,
+      "loss": 0.7939,
+      "step": 10833
+    },
+    {
+      "epoch": 1.8456608868980005,
+      "grad_norm": 1.71875,
+      "learning_rate": 6.405420285331414e-06,
+      "loss": 0.8711,
+      "step": 10834
+    },
+    {
+      "epoch": 1.8458323509869903,
+      "grad_norm": 1.65625,
+      "learning_rate": 6.4037351217168965e-06,
+      "loss": 0.8093,
+      "step": 10835
+    },
+    {
+      "epoch": 1.84600381507598,
+      "grad_norm": 1.7109375,
+      "learning_rate": 6.402050075390112e-06,
+      "loss": 0.8094,
+      "step": 10836
+    },
+    {
+      "epoch": 1.84617527916497,
+      "grad_norm": 1.7109375,
+      "learning_rate": 6.400365146406027e-06,
+      "loss": 0.8507,
+      "step": 10837
+    },
+    {
+      "epoch": 1.8463467432539598,
+      "grad_norm": 1.6015625,
+      "learning_rate": 6.398680334819587e-06,
+      "loss": 0.867,
+      "step": 10838
+    },
+    {
+      "epoch": 1.8465182073429496,
+      "grad_norm": 1.6796875,
+      "learning_rate": 6.396995640685744e-06,
+      "loss": 0.8413,
+      "step": 10839
+    },
+    {
+      "epoch": 1.8466896714319394,
+      "grad_norm": 1.6640625,
+      "learning_rate": 6.395311064059442e-06,
+      "loss": 0.8135,
+      "step": 10840
+    },
+    {
+      "epoch": 1.8468611355209292,
+      "grad_norm": 1.65625,
+      "learning_rate": 6.393626604995617e-06,
+      "loss": 0.8094,
+      "step": 10841
+    },
+    {
+      "epoch": 1.847032599609919,
+      "grad_norm": 1.796875,
+      "learning_rate": 6.391942263549211e-06,
+      "loss": 0.8335,
+      "step": 10842
+    },
+    {
+      "epoch": 1.847204063698909,
+      "grad_norm": 1.7890625,
+      "learning_rate": 6.390258039775155e-06,
+      "loss": 0.8797,
+      "step": 10843
+    },
+    {
+      "epoch": 1.847375527787899,
+      "grad_norm": 1.6796875,
+      "learning_rate": 6.388573933728376e-06,
+      "loss": 0.8325,
+      "step": 10844
+    },
+    {
+      "epoch": 1.8475469918768888,
+      "grad_norm": 1.828125,
+      "learning_rate": 6.386889945463801e-06,
+      "loss": 1.0021,
+      "step": 10845
+    },
+    {
+      "epoch": 1.8477184559658788,
+      "grad_norm": 1.6796875,
+      "learning_rate": 6.385206075036352e-06,
+      "loss": 0.853,
+      "step": 10846
+    },
+    {
+      "epoch": 1.8478899200548686,
+      "grad_norm": 1.609375,
+      "learning_rate": 6.383522322500943e-06,
+      "loss": 0.7751,
+      "step": 10847
+    },
+    {
+      "epoch": 1.8480613841438585,
+      "grad_norm": 1.6953125,
+      "learning_rate": 6.381838687912489e-06,
+      "loss": 0.913,
+      "step": 10848
+    },
+    {
+      "epoch": 1.8482328482328483,
+      "grad_norm": 1.703125,
+      "learning_rate": 6.3801551713259015e-06,
+      "loss": 0.8545,
+      "step": 10849
+    },
+    {
+      "epoch": 1.848404312321838,
+      "grad_norm": 1.7578125,
+      "learning_rate": 6.3784717727960844e-06,
+      "loss": 0.8677,
+      "step": 10850
+    },
+    {
+      "epoch": 1.848575776410828,
+      "grad_norm": 1.7421875,
+      "learning_rate": 6.376788492377941e-06,
+      "loss": 0.7775,
+      "step": 10851
+    },
+    {
+      "epoch": 1.8487472404998178,
+      "grad_norm": 1.625,
+      "learning_rate": 6.375105330126368e-06,
+      "loss": 0.7807,
+      "step": 10852
+    },
+    {
+      "epoch": 1.8489187045888076,
+      "grad_norm": 1.71875,
+      "learning_rate": 6.373422286096259e-06,
+      "loss": 0.8227,
+      "step": 10853
+    },
+    {
+      "epoch": 1.8490901686777974,
+      "grad_norm": 1.7109375,
+      "learning_rate": 6.371739360342507e-06,
+      "loss": 0.9536,
+      "step": 10854
+    },
+    {
+      "epoch": 1.8492616327667875,
+      "grad_norm": 1.7265625,
+      "learning_rate": 6.3700565529199965e-06,
+      "loss": 0.7772,
+      "step": 10855
+    },
+    {
+      "epoch": 1.8494330968557773,
+      "grad_norm": 1.640625,
+      "learning_rate": 6.3683738638836125e-06,
+      "loss": 0.8825,
+      "step": 10856
+    },
+    {
+      "epoch": 1.849604560944767,
+      "grad_norm": 1.6171875,
+      "learning_rate": 6.366691293288229e-06,
+      "loss": 0.8269,
+      "step": 10857
+    },
+    {
+      "epoch": 1.8497760250337572,
+      "grad_norm": 1.7578125,
+      "learning_rate": 6.365008841188729e-06,
+      "loss": 0.9076,
+      "step": 10858
+    },
+    {
+      "epoch": 1.849947489122747,
+      "grad_norm": 1.6875,
+      "learning_rate": 6.363326507639978e-06,
+      "loss": 0.9056,
+      "step": 10859
+    },
+    {
+      "epoch": 1.8501189532117368,
+      "grad_norm": 1.71875,
+      "learning_rate": 6.36164429269684e-06,
+      "loss": 0.9067,
+      "step": 10860
+    },
+    {
+      "epoch": 1.8502904173007266,
+      "grad_norm": 1.6640625,
+      "learning_rate": 6.359962196414185e-06,
+      "loss": 0.7929,
+      "step": 10861
+    },
+    {
+      "epoch": 1.8504618813897165,
+      "grad_norm": 1.5859375,
+      "learning_rate": 6.358280218846871e-06,
+      "loss": 0.766,
+      "step": 10862
+    },
+    {
+      "epoch": 1.8506333454787063,
+      "grad_norm": 1.59375,
+      "learning_rate": 6.356598360049751e-06,
+      "loss": 0.8052,
+      "step": 10863
+    },
+    {
+      "epoch": 1.850804809567696,
+      "grad_norm": 1.8671875,
+      "learning_rate": 6.354916620077677e-06,
+      "loss": 0.8696,
+      "step": 10864
+    },
+    {
+      "epoch": 1.850976273656686,
+      "grad_norm": 1.75,
+      "learning_rate": 6.353234998985499e-06,
+      "loss": 0.8394,
+      "step": 10865
+    },
+    {
+      "epoch": 1.8511477377456758,
+      "grad_norm": 1.7109375,
+      "learning_rate": 6.3515534968280604e-06,
+      "loss": 0.8423,
+      "step": 10866
+    },
+    {
+      "epoch": 1.8513192018346658,
+      "grad_norm": 1.6796875,
+      "learning_rate": 6.3498721136601994e-06,
+      "loss": 0.8579,
+      "step": 10867
+    },
+    {
+      "epoch": 1.8514906659236556,
+      "grad_norm": 1.703125,
+      "learning_rate": 6.348190849536755e-06,
+      "loss": 0.7728,
+      "step": 10868
+    },
+    {
+      "epoch": 1.8516621300126455,
+      "grad_norm": 1.5859375,
+      "learning_rate": 6.346509704512563e-06,
+      "loss": 0.779,
+      "step": 10869
+    },
+    {
+      "epoch": 1.8518335941016355,
+      "grad_norm": 1.65625,
+      "learning_rate": 6.344828678642444e-06,
+      "loss": 0.7862,
+      "step": 10870
+    },
+    {
+      "epoch": 1.8520050581906253,
+      "grad_norm": 1.6640625,
+      "learning_rate": 6.343147771981225e-06,
+      "loss": 0.825,
+      "step": 10871
+    },
+    {
+      "epoch": 1.8521765222796152,
+      "grad_norm": 1.625,
+      "learning_rate": 6.341466984583728e-06,
+      "loss": 0.8207,
+      "step": 10872
+    },
+    {
+      "epoch": 1.852347986368605,
+      "grad_norm": 1.6953125,
+      "learning_rate": 6.339786316504769e-06,
+      "loss": 0.8482,
+      "step": 10873
+    },
+    {
+      "epoch": 1.8525194504575948,
+      "grad_norm": 1.609375,
+      "learning_rate": 6.338105767799161e-06,
+      "loss": 0.854,
+      "step": 10874
+    },
+    {
+      "epoch": 1.8526909145465846,
+      "grad_norm": 1.7265625,
+      "learning_rate": 6.336425338521712e-06,
+      "loss": 0.8701,
+      "step": 10875
+    },
+    {
+      "epoch": 1.8528623786355745,
+      "grad_norm": 1.8046875,
+      "learning_rate": 6.33474502872723e-06,
+      "loss": 0.9802,
+      "step": 10876
+    },
+    {
+      "epoch": 1.8530338427245643,
+      "grad_norm": 1.7109375,
+      "learning_rate": 6.333064838470515e-06,
+      "loss": 0.8136,
+      "step": 10877
+    },
+    {
+      "epoch": 1.853205306813554,
+      "grad_norm": 1.6640625,
+      "learning_rate": 6.331384767806365e-06,
+      "loss": 0.808,
+      "step": 10878
+    },
+    {
+      "epoch": 1.853376770902544,
+      "grad_norm": 1.7421875,
+      "learning_rate": 6.3297048167895705e-06,
+      "loss": 0.9327,
+      "step": 10879
+    },
+    {
+      "epoch": 1.853548234991534,
+      "grad_norm": 1.6328125,
+      "learning_rate": 6.328024985474924e-06,
+      "loss": 0.8691,
+      "step": 10880
+    },
+    {
+      "epoch": 1.8537196990805238,
+      "grad_norm": 1.6796875,
+      "learning_rate": 6.326345273917214e-06,
+      "loss": 0.7949,
+      "step": 10881
+    },
+    {
+      "epoch": 1.8538911631695136,
+      "grad_norm": 1.65625,
+      "learning_rate": 6.324665682171214e-06,
+      "loss": 0.8325,
+      "step": 10882
+    },
+    {
+      "epoch": 1.8540626272585037,
+      "grad_norm": 1.796875,
+      "learning_rate": 6.322986210291705e-06,
+      "loss": 0.8751,
+      "step": 10883
+    },
+    {
+      "epoch": 1.8542340913474935,
+      "grad_norm": 1.7265625,
+      "learning_rate": 6.321306858333463e-06,
+      "loss": 0.8528,
+      "step": 10884
+    },
+    {
+      "epoch": 1.8544055554364833,
+      "grad_norm": 1.640625,
+      "learning_rate": 6.319627626351258e-06,
+      "loss": 0.8821,
+      "step": 10885
+    },
+    {
+      "epoch": 1.8545770195254732,
+      "grad_norm": 1.75,
+      "learning_rate": 6.317948514399854e-06,
+      "loss": 0.8082,
+      "step": 10886
+    },
+    {
+      "epoch": 1.854748483614463,
+      "grad_norm": 1.7890625,
+      "learning_rate": 6.3162695225340155e-06,
+      "loss": 0.8664,
+      "step": 10887
+    },
+    {
+      "epoch": 1.8549199477034528,
+      "grad_norm": 1.6875,
+      "learning_rate": 6.314590650808498e-06,
+      "loss": 0.8479,
+      "step": 10888
+    },
+    {
+      "epoch": 1.8550914117924426,
+      "grad_norm": 1.765625,
+      "learning_rate": 6.312911899278059e-06,
+      "loss": 0.8752,
+      "step": 10889
+    },
+    {
+      "epoch": 1.8552628758814325,
+      "grad_norm": 1.71875,
+      "learning_rate": 6.311233267997446e-06,
+      "loss": 0.8617,
+      "step": 10890
+    },
+    {
+      "epoch": 1.8554343399704223,
+      "grad_norm": 1.65625,
+      "learning_rate": 6.309554757021408e-06,
+      "loss": 0.8035,
+      "step": 10891
+    },
+    {
+      "epoch": 1.8556058040594123,
+      "grad_norm": 1.625,
+      "learning_rate": 6.307876366404687e-06,
+      "loss": 0.8573,
+      "step": 10892
+    },
+    {
+      "epoch": 1.8557772681484022,
+      "grad_norm": 1.6875,
+      "learning_rate": 6.30619809620202e-06,
+      "loss": 0.923,
+      "step": 10893
+    },
+    {
+      "epoch": 1.855948732237392,
+      "grad_norm": 1.7109375,
+      "learning_rate": 6.304519946468142e-06,
+      "loss": 0.7693,
+      "step": 10894
+    },
+    {
+      "epoch": 1.856120196326382,
+      "grad_norm": 1.640625,
+      "learning_rate": 6.302841917257785e-06,
+      "loss": 0.824,
+      "step": 10895
+    },
+    {
+      "epoch": 1.8562916604153719,
+      "grad_norm": 1.6171875,
+      "learning_rate": 6.301164008625674e-06,
+      "loss": 0.7486,
+      "step": 10896
+    },
+    {
+      "epoch": 1.8564631245043617,
+      "grad_norm": 1.734375,
+      "learning_rate": 6.299486220626534e-06,
+      "loss": 0.9049,
+      "step": 10897
+    },
+    {
+      "epoch": 1.8566345885933515,
+      "grad_norm": 1.765625,
+      "learning_rate": 6.297808553315084e-06,
+      "loss": 0.8944,
+      "step": 10898
+    },
+    {
+      "epoch": 1.8568060526823413,
+      "grad_norm": 1.7109375,
+      "learning_rate": 6.296131006746037e-06,
+      "loss": 0.9116,
+      "step": 10899
+    },
+    {
+      "epoch": 1.8569775167713312,
+      "grad_norm": 1.7734375,
+      "learning_rate": 6.294453580974106e-06,
+      "loss": 0.9053,
+      "step": 10900
+    },
+    {
+      "epoch": 1.857148980860321,
+      "grad_norm": 1.7421875,
+      "learning_rate": 6.2927762760539975e-06,
+      "loss": 0.8544,
+      "step": 10901
+    },
+    {
+      "epoch": 1.8573204449493108,
+      "grad_norm": 2.25,
+      "learning_rate": 6.291099092040414e-06,
+      "loss": 0.8509,
+      "step": 10902
+    },
+    {
+      "epoch": 1.8574919090383006,
+      "grad_norm": 1.78125,
+      "learning_rate": 6.289422028988057e-06,
+      "loss": 0.815,
+      "step": 10903
+    },
+    {
+      "epoch": 1.8576633731272907,
+      "grad_norm": 1.6484375,
+      "learning_rate": 6.287745086951621e-06,
+      "loss": 0.7843,
+      "step": 10904
+    },
+    {
+      "epoch": 1.8578348372162805,
+      "grad_norm": 1.734375,
+      "learning_rate": 6.286068265985795e-06,
+      "loss": 0.8522,
+      "step": 10905
+    },
+    {
+      "epoch": 1.8580063013052703,
+      "grad_norm": 1.765625,
+      "learning_rate": 6.284391566145269e-06,
+      "loss": 0.8605,
+      "step": 10906
+    },
+    {
+      "epoch": 1.8581777653942604,
+      "grad_norm": 1.8046875,
+      "learning_rate": 6.282714987484725e-06,
+      "loss": 0.8064,
+      "step": 10907
+    },
+    {
+      "epoch": 1.8583492294832502,
+      "grad_norm": 1.7109375,
+      "learning_rate": 6.281038530058843e-06,
+      "loss": 0.832,
+      "step": 10908
+    },
+    {
+      "epoch": 1.85852069357224,
+      "grad_norm": 1.640625,
+      "learning_rate": 6.2793621939223e-06,
+      "loss": 0.7933,
+      "step": 10909
+    },
+    {
+      "epoch": 1.8586921576612299,
+      "grad_norm": 1.671875,
+      "learning_rate": 6.277685979129766e-06,
+      "loss": 0.8531,
+      "step": 10910
+    },
+    {
+      "epoch": 1.8588636217502197,
+      "grad_norm": 1.6953125,
+      "learning_rate": 6.276009885735909e-06,
+      "loss": 0.8221,
+      "step": 10911
+    },
+    {
+      "epoch": 1.8590350858392095,
+      "grad_norm": 1.7265625,
+      "learning_rate": 6.274333913795392e-06,
+      "loss": 0.8708,
+      "step": 10912
+    },
+    {
+      "epoch": 1.8592065499281993,
+      "grad_norm": 1.78125,
+      "learning_rate": 6.2726580633628775e-06,
+      "loss": 0.8046,
+      "step": 10913
+    },
+    {
+      "epoch": 1.8593780140171892,
+      "grad_norm": 1.671875,
+      "learning_rate": 6.270982334493017e-06,
+      "loss": 0.8399,
+      "step": 10914
+    },
+    {
+      "epoch": 1.859549478106179,
+      "grad_norm": 1.734375,
+      "learning_rate": 6.26930672724047e-06,
+      "loss": 0.9163,
+      "step": 10915
+    },
+    {
+      "epoch": 1.859720942195169,
+      "grad_norm": 1.6171875,
+      "learning_rate": 6.267631241659875e-06,
+      "loss": 0.9097,
+      "step": 10916
+    },
+    {
+      "epoch": 1.8598924062841589,
+      "grad_norm": 1.71875,
+      "learning_rate": 6.26595587780588e-06,
+      "loss": 0.8519,
+      "step": 10917
+    },
+    {
+      "epoch": 1.8600638703731487,
+      "grad_norm": 1.75,
+      "learning_rate": 6.2642806357331244e-06,
+      "loss": 0.9399,
+      "step": 10918
+    },
+    {
+      "epoch": 1.8602353344621387,
+      "grad_norm": 1.6875,
+      "learning_rate": 6.262605515496245e-06,
+      "loss": 0.8465,
+      "step": 10919
+    },
+    {
+      "epoch": 1.8604067985511286,
+      "grad_norm": 1.6953125,
+      "learning_rate": 6.260930517149875e-06,
+      "loss": 0.7655,
+      "step": 10920
+    },
+    {
+      "epoch": 1.8605782626401184,
+      "grad_norm": 1.6328125,
+      "learning_rate": 6.2592556407486394e-06,
+      "loss": 0.9003,
+      "step": 10921
+    },
+    {
+      "epoch": 1.8607497267291082,
+      "grad_norm": 1.703125,
+      "learning_rate": 6.257580886347162e-06,
+      "loss": 0.8566,
+      "step": 10922
+    },
+    {
+      "epoch": 1.860921190818098,
+      "grad_norm": 1.640625,
+      "learning_rate": 6.255906254000067e-06,
+      "loss": 0.839,
+      "step": 10923
+    },
+    {
+      "epoch": 1.8610926549070879,
+      "grad_norm": 1.5859375,
+      "learning_rate": 6.254231743761967e-06,
+      "loss": 0.7622,
+      "step": 10924
+    },
+    {
+      "epoch": 1.8612641189960777,
+      "grad_norm": 1.546875,
+      "learning_rate": 6.252557355687476e-06,
+      "loss": 0.7633,
+      "step": 10925
+    },
+    {
+      "epoch": 1.8614355830850675,
+      "grad_norm": 1.6484375,
+      "learning_rate": 6.250883089831202e-06,
+      "loss": 0.809,
+      "step": 10926
+    },
+    {
+      "epoch": 1.8616070471740573,
+      "grad_norm": 1.6953125,
+      "learning_rate": 6.2492089462477515e-06,
+      "loss": 0.862,
+      "step": 10927
+    },
+    {
+      "epoch": 1.8617785112630474,
+      "grad_norm": 1.7734375,
+      "learning_rate": 6.247534924991716e-06,
+      "loss": 0.8738,
+      "step": 10928
+    },
+    {
+      "epoch": 1.8619499753520372,
+      "grad_norm": 1.71875,
+      "learning_rate": 6.245861026117699e-06,
+      "loss": 0.8735,
+      "step": 10929
+    },
+    {
+      "epoch": 1.862121439441027,
+      "grad_norm": 1.6875,
+      "learning_rate": 6.244187249680287e-06,
+      "loss": 0.8735,
+      "step": 10930
+    },
+    {
+      "epoch": 1.862292903530017,
+      "grad_norm": 1.6953125,
+      "learning_rate": 6.242513595734075e-06,
+      "loss": 0.8638,
+      "step": 10931
+    },
+    {
+      "epoch": 1.862464367619007,
+      "grad_norm": 1.6796875,
+      "learning_rate": 6.240840064333644e-06,
+      "loss": 0.8454,
+      "step": 10932
+    },
+    {
+      "epoch": 1.8626358317079967,
+      "grad_norm": 1.7265625,
+      "learning_rate": 6.239166655533575e-06,
+      "loss": 0.8941,
+      "step": 10933
+    },
+    {
+      "epoch": 1.8628072957969866,
+      "grad_norm": 1.6015625,
+      "learning_rate": 6.237493369388441e-06,
+      "loss": 0.7839,
+      "step": 10934
+    },
+    {
+      "epoch": 1.8629787598859764,
+      "grad_norm": 1.796875,
+      "learning_rate": 6.235820205952818e-06,
+      "loss": 0.8964,
+      "step": 10935
+    },
+    {
+      "epoch": 1.8631502239749662,
+      "grad_norm": 1.75,
+      "learning_rate": 6.2341471652812734e-06,
+      "loss": 0.8891,
+      "step": 10936
+    },
+    {
+      "epoch": 1.863321688063956,
+      "grad_norm": 1.6640625,
+      "learning_rate": 6.2324742474283695e-06,
+      "loss": 0.8195,
+      "step": 10937
+    },
+    {
+      "epoch": 1.8634931521529459,
+      "grad_norm": 1.6875,
+      "learning_rate": 6.23080145244867e-06,
+      "loss": 0.7823,
+      "step": 10938
+    },
+    {
+      "epoch": 1.8636646162419357,
+      "grad_norm": 1.609375,
+      "learning_rate": 6.229128780396727e-06,
+      "loss": 0.8456,
+      "step": 10939
+    },
+    {
+      "epoch": 1.8638360803309257,
+      "grad_norm": 1.6171875,
+      "learning_rate": 6.227456231327094e-06,
+      "loss": 0.8666,
+      "step": 10940
+    },
+    {
+      "epoch": 1.8640075444199156,
+      "grad_norm": 1.8125,
+      "learning_rate": 6.225783805294319e-06,
+      "loss": 0.8293,
+      "step": 10941
+    },
+    {
+      "epoch": 1.8641790085089054,
+      "grad_norm": 1.6953125,
+      "learning_rate": 6.224111502352947e-06,
+      "loss": 0.9055,
+      "step": 10942
+    },
+    {
+      "epoch": 1.8643504725978954,
+      "grad_norm": 1.6875,
+      "learning_rate": 6.222439322557516e-06,
+      "loss": 0.7858,
+      "step": 10943
+    },
+    {
+      "epoch": 1.8645219366868853,
+      "grad_norm": 1.734375,
+      "learning_rate": 6.2207672659625665e-06,
+      "loss": 0.9471,
+      "step": 10944
+    },
+    {
+      "epoch": 1.864693400775875,
+      "grad_norm": 1.7265625,
+      "learning_rate": 6.219095332622626e-06,
+      "loss": 0.9263,
+      "step": 10945
+    },
+    {
+      "epoch": 1.864864864864865,
+      "grad_norm": 1.6875,
+      "learning_rate": 6.217423522592223e-06,
+      "loss": 0.8449,
+      "step": 10946
+    },
+    {
+      "epoch": 1.8650363289538547,
+      "grad_norm": 1.8359375,
+      "learning_rate": 6.215751835925885e-06,
+      "loss": 0.8745,
+      "step": 10947
+    },
+    {
+      "epoch": 1.8652077930428446,
+      "grad_norm": 1.578125,
+      "learning_rate": 6.2140802726781294e-06,
+      "loss": 0.8291,
+      "step": 10948
+    },
+    {
+      "epoch": 1.8653792571318344,
+      "grad_norm": 1.7109375,
+      "learning_rate": 6.2124088329034715e-06,
+      "loss": 0.7714,
+      "step": 10949
+    },
+    {
+      "epoch": 1.8655507212208242,
+      "grad_norm": 1.828125,
+      "learning_rate": 6.210737516656427e-06,
+      "loss": 0.849,
+      "step": 10950
+    },
+    {
+      "epoch": 1.865722185309814,
+      "grad_norm": 1.703125,
+      "learning_rate": 6.2090663239915e-06,
+      "loss": 0.8369,
+      "step": 10951
+    },
+    {
+      "epoch": 1.865893649398804,
+      "grad_norm": 1.71875,
+      "learning_rate": 6.207395254963193e-06,
+      "loss": 0.8538,
+      "step": 10952
+    },
+    {
+      "epoch": 1.866065113487794,
+      "grad_norm": 1.734375,
+      "learning_rate": 6.205724309626011e-06,
+      "loss": 0.8414,
+      "step": 10953
+    },
+    {
+      "epoch": 1.8662365775767837,
+      "grad_norm": 1.78125,
+      "learning_rate": 6.204053488034446e-06,
+      "loss": 0.8347,
+      "step": 10954
+    },
+    {
+      "epoch": 1.8664080416657738,
+      "grad_norm": 1.6015625,
+      "learning_rate": 6.2023827902429915e-06,
+      "loss": 0.7904,
+      "step": 10955
+    },
+    {
+      "epoch": 1.8665795057547636,
+      "grad_norm": 1.65625,
+      "learning_rate": 6.200712216306134e-06,
+      "loss": 0.8911,
+      "step": 10956
+    },
+    {
+      "epoch": 1.8667509698437534,
+      "grad_norm": 1.7265625,
+      "learning_rate": 6.1990417662783574e-06,
+      "loss": 0.8466,
+      "step": 10957
+    },
+    {
+      "epoch": 1.8669224339327433,
+      "grad_norm": 1.6484375,
+      "learning_rate": 6.197371440214144e-06,
+      "loss": 0.8515,
+      "step": 10958
+    },
+    {
+      "epoch": 1.867093898021733,
+      "grad_norm": 1.6875,
+      "learning_rate": 6.195701238167966e-06,
+      "loss": 0.8596,
+      "step": 10959
+    },
+    {
+      "epoch": 1.867265362110723,
+      "grad_norm": 1.6796875,
+      "learning_rate": 6.194031160194296e-06,
+      "loss": 0.8283,
+      "step": 10960
+    },
+    {
+      "epoch": 1.8674368261997127,
+      "grad_norm": 1.6484375,
+      "learning_rate": 6.192361206347603e-06,
+      "loss": 0.7804,
+      "step": 10961
+    },
+    {
+      "epoch": 1.8676082902887026,
+      "grad_norm": 1.6796875,
+      "learning_rate": 6.190691376682349e-06,
+      "loss": 0.8363,
+      "step": 10962
+    },
+    {
+      "epoch": 1.8677797543776924,
+      "grad_norm": 1.640625,
+      "learning_rate": 6.189021671252993e-06,
+      "loss": 0.8353,
+      "step": 10963
+    },
+    {
+      "epoch": 1.8679512184666824,
+      "grad_norm": 1.6484375,
+      "learning_rate": 6.187352090113992e-06,
+      "loss": 0.8868,
+      "step": 10964
+    },
+    {
+      "epoch": 1.8681226825556723,
+      "grad_norm": 1.765625,
+      "learning_rate": 6.185682633319796e-06,
+      "loss": 0.8438,
+      "step": 10965
+    },
+    {
+      "epoch": 1.868294146644662,
+      "grad_norm": 1.6796875,
+      "learning_rate": 6.184013300924852e-06,
+      "loss": 0.8665,
+      "step": 10966
+    },
+    {
+      "epoch": 1.868465610733652,
+      "grad_norm": 1.7578125,
+      "learning_rate": 6.1823440929836055e-06,
+      "loss": 0.8625,
+      "step": 10967
+    },
+    {
+      "epoch": 1.868637074822642,
+      "grad_norm": 1.6875,
+      "learning_rate": 6.180675009550492e-06,
+      "loss": 0.8351,
+      "step": 10968
+    },
+    {
+      "epoch": 1.8688085389116318,
+      "grad_norm": 1.7890625,
+      "learning_rate": 6.179006050679947e-06,
+      "loss": 0.8879,
+      "step": 10969
+    },
+    {
+      "epoch": 1.8689800030006216,
+      "grad_norm": 1.6484375,
+      "learning_rate": 6.177337216426407e-06,
+      "loss": 0.8025,
+      "step": 10970
+    },
+    {
+      "epoch": 1.8691514670896114,
+      "grad_norm": 1.7109375,
+      "learning_rate": 6.175668506844294e-06,
+      "loss": 0.8274,
+      "step": 10971
+    },
+    {
+      "epoch": 1.8693229311786013,
+      "grad_norm": 1.71875,
+      "learning_rate": 6.173999921988032e-06,
+      "loss": 0.8356,
+      "step": 10972
+    },
+    {
+      "epoch": 1.869494395267591,
+      "grad_norm": 1.6640625,
+      "learning_rate": 6.172331461912044e-06,
+      "loss": 0.7581,
+      "step": 10973
+    },
+    {
+      "epoch": 1.869665859356581,
+      "grad_norm": 1.6875,
+      "learning_rate": 6.170663126670737e-06,
+      "loss": 0.8047,
+      "step": 10974
+    },
+    {
+      "epoch": 1.8698373234455707,
+      "grad_norm": 1.703125,
+      "learning_rate": 6.1689949163185245e-06,
+      "loss": 0.809,
+      "step": 10975
+    },
+    {
+      "epoch": 1.8700087875345606,
+      "grad_norm": 1.6875,
+      "learning_rate": 6.167326830909815e-06,
+      "loss": 0.785,
+      "step": 10976
+    },
+    {
+      "epoch": 1.8701802516235506,
+      "grad_norm": 1.6015625,
+      "learning_rate": 6.1656588704990085e-06,
+      "loss": 0.7857,
+      "step": 10977
+    },
+    {
+      "epoch": 1.8703517157125404,
+      "grad_norm": 1.6875,
+      "learning_rate": 6.163991035140506e-06,
+      "loss": 0.8535,
+      "step": 10978
+    },
+    {
+      "epoch": 1.8705231798015303,
+      "grad_norm": 1.8046875,
+      "learning_rate": 6.162323324888702e-06,
+      "loss": 0.8995,
+      "step": 10979
+    },
+    {
+      "epoch": 1.8706946438905203,
+      "grad_norm": 1.6875,
+      "learning_rate": 6.160655739797985e-06,
+      "loss": 0.7442,
+      "step": 10980
+    },
+    {
+      "epoch": 1.8708661079795101,
+      "grad_norm": 1.6875,
+      "learning_rate": 6.158988279922741e-06,
+      "loss": 0.9137,
+      "step": 10981
+    },
+    {
+      "epoch": 1.8710375720685,
+      "grad_norm": 1.703125,
+      "learning_rate": 6.157320945317353e-06,
+      "loss": 0.8714,
+      "step": 10982
+    },
+    {
+      "epoch": 1.8712090361574898,
+      "grad_norm": 1.765625,
+      "learning_rate": 6.1556537360362014e-06,
+      "loss": 0.9099,
+      "step": 10983
+    },
+    {
+      "epoch": 1.8713805002464796,
+      "grad_norm": 1.7109375,
+      "learning_rate": 6.153986652133657e-06,
+      "loss": 0.8458,
+      "step": 10984
+    },
+    {
+      "epoch": 1.8715519643354694,
+      "grad_norm": 1.578125,
+      "learning_rate": 6.152319693664091e-06,
+      "loss": 0.8526,
+      "step": 10985
+    },
+    {
+      "epoch": 1.8717234284244593,
+      "grad_norm": 1.6640625,
+      "learning_rate": 6.150652860681869e-06,
+      "loss": 0.8533,
+      "step": 10986
+    },
+    {
+      "epoch": 1.871894892513449,
+      "grad_norm": 1.8125,
+      "learning_rate": 6.148986153241352e-06,
+      "loss": 0.9229,
+      "step": 10987
+    },
+    {
+      "epoch": 1.872066356602439,
+      "grad_norm": 1.796875,
+      "learning_rate": 6.147319571396897e-06,
+      "loss": 0.8749,
+      "step": 10988
+    },
+    {
+      "epoch": 1.872237820691429,
+      "grad_norm": 1.7890625,
+      "learning_rate": 6.14565311520286e-06,
+      "loss": 0.8774,
+      "step": 10989
+    },
+    {
+      "epoch": 1.8724092847804188,
+      "grad_norm": 1.734375,
+      "learning_rate": 6.143986784713588e-06,
+      "loss": 0.948,
+      "step": 10990
+    },
+    {
+      "epoch": 1.8725807488694086,
+      "grad_norm": 1.640625,
+      "learning_rate": 6.142320579983427e-06,
+      "loss": 0.8464,
+      "step": 10991
+    },
+    {
+      "epoch": 1.8727522129583987,
+      "grad_norm": 1.6796875,
+      "learning_rate": 6.14065450106672e-06,
+      "loss": 0.8901,
+      "step": 10992
+    },
+    {
+      "epoch": 1.8729236770473885,
+      "grad_norm": 1.7578125,
+      "learning_rate": 6.138988548017802e-06,
+      "loss": 0.805,
+      "step": 10993
+    },
+    {
+      "epoch": 1.8730951411363783,
+      "grad_norm": 1.7578125,
+      "learning_rate": 6.137322720891007e-06,
+      "loss": 0.823,
+      "step": 10994
+    },
+    {
+      "epoch": 1.8732666052253681,
+      "grad_norm": 1.671875,
+      "learning_rate": 6.135657019740663e-06,
+      "loss": 0.8371,
+      "step": 10995
+    },
+    {
+      "epoch": 1.873438069314358,
+      "grad_norm": 1.7421875,
+      "learning_rate": 6.133991444621097e-06,
+      "loss": 0.9085,
+      "step": 10996
+    },
+    {
+      "epoch": 1.8736095334033478,
+      "grad_norm": 1.6640625,
+      "learning_rate": 6.132325995586628e-06,
+      "loss": 0.8843,
+      "step": 10997
+    },
+    {
+      "epoch": 1.8737809974923376,
+      "grad_norm": 1.796875,
+      "learning_rate": 6.130660672691571e-06,
+      "loss": 0.9687,
+      "step": 10998
+    },
+    {
+      "epoch": 1.8739524615813274,
+      "grad_norm": 1.71875,
+      "learning_rate": 6.128995475990241e-06,
+      "loss": 0.8697,
+      "step": 10999
+    },
+    {
+      "epoch": 1.8741239256703173,
+      "grad_norm": 1.71875,
+      "learning_rate": 6.127330405536943e-06,
+      "loss": 0.8857,
+      "step": 11000
+    },
+    {
+      "epoch": 1.8742953897593073,
+      "grad_norm": 1.7421875,
+      "learning_rate": 6.125665461385986e-06,
+      "loss": 0.8626,
+      "step": 11001
+    },
+    {
+      "epoch": 1.8744668538482971,
+      "grad_norm": 1.6640625,
+      "learning_rate": 6.124000643591667e-06,
+      "loss": 0.7529,
+      "step": 11002
+    },
+    {
+      "epoch": 1.874638317937287,
+      "grad_norm": 1.703125,
+      "learning_rate": 6.122335952208283e-06,
+      "loss": 0.8251,
+      "step": 11003
+    },
+    {
+      "epoch": 1.874809782026277,
+      "grad_norm": 1.703125,
+      "learning_rate": 6.120671387290125e-06,
+      "loss": 0.8308,
+      "step": 11004
+    },
+    {
+      "epoch": 1.8749812461152668,
+      "grad_norm": 1.7265625,
+      "learning_rate": 6.1190069488914806e-06,
+      "loss": 0.8574,
+      "step": 11005
+    },
+    {
+      "epoch": 1.8751527102042567,
+      "grad_norm": 1.4921875,
+      "learning_rate": 6.117342637066635e-06,
+      "loss": 0.7836,
+      "step": 11006
+    },
+    {
+      "epoch": 1.8753241742932465,
+      "grad_norm": 1.71875,
+      "learning_rate": 6.115678451869866e-06,
+      "loss": 0.8602,
+      "step": 11007
+    },
+    {
+      "epoch": 1.8754956383822363,
+      "grad_norm": 1.6328125,
+      "learning_rate": 6.114014393355453e-06,
+      "loss": 0.8605,
+      "step": 11008
+    },
+    {
+      "epoch": 1.8756671024712261,
+      "grad_norm": 1.703125,
+      "learning_rate": 6.112350461577661e-06,
+      "loss": 0.8752,
+      "step": 11009
+    },
+    {
+      "epoch": 1.875838566560216,
+      "grad_norm": 1.671875,
+      "learning_rate": 6.110686656590761e-06,
+      "loss": 0.754,
+      "step": 11010
+    },
+    {
+      "epoch": 1.8760100306492058,
+      "grad_norm": 1.625,
+      "learning_rate": 6.109022978449013e-06,
+      "loss": 0.8233,
+      "step": 11011
+    },
+    {
+      "epoch": 1.8761814947381956,
+      "grad_norm": 1.6796875,
+      "learning_rate": 6.107359427206679e-06,
+      "loss": 0.8455,
+      "step": 11012
+    },
+    {
+      "epoch": 1.8763529588271857,
+      "grad_norm": 1.6953125,
+      "learning_rate": 6.105696002918012e-06,
+      "loss": 0.7755,
+      "step": 11013
+    },
+    {
+      "epoch": 1.8765244229161755,
+      "grad_norm": 1.71875,
+      "learning_rate": 6.104032705637264e-06,
+      "loss": 0.8127,
+      "step": 11014
+    },
+    {
+      "epoch": 1.8766958870051653,
+      "grad_norm": 1.6875,
+      "learning_rate": 6.102369535418679e-06,
+      "loss": 0.8543,
+      "step": 11015
+    },
+    {
+      "epoch": 1.8768673510941554,
+      "grad_norm": 1.8046875,
+      "learning_rate": 6.100706492316499e-06,
+      "loss": 0.9207,
+      "step": 11016
+    },
+    {
+      "epoch": 1.8770388151831452,
+      "grad_norm": 1.6953125,
+      "learning_rate": 6.099043576384966e-06,
+      "loss": 0.8285,
+      "step": 11017
+    },
+    {
+      "epoch": 1.877210279272135,
+      "grad_norm": 1.71875,
+      "learning_rate": 6.097380787678311e-06,
+      "loss": 0.7908,
+      "step": 11018
+    },
+    {
+      "epoch": 1.8773817433611248,
+      "grad_norm": 1.7734375,
+      "learning_rate": 6.095718126250769e-06,
+      "loss": 0.8402,
+      "step": 11019
+    },
+    {
+      "epoch": 1.8775532074501147,
+      "grad_norm": 1.921875,
+      "learning_rate": 6.094055592156557e-06,
+      "loss": 0.8721,
+      "step": 11020
+    },
+    {
+      "epoch": 1.8777246715391045,
+      "grad_norm": 1.71875,
+      "learning_rate": 6.092393185449901e-06,
+      "loss": 0.8221,
+      "step": 11021
+    },
+    {
+      "epoch": 1.8778961356280943,
+      "grad_norm": 1.6484375,
+      "learning_rate": 6.090730906185016e-06,
+      "loss": 0.9318,
+      "step": 11022
+    },
+    {
+      "epoch": 1.8780675997170841,
+      "grad_norm": 1.734375,
+      "learning_rate": 6.089068754416118e-06,
+      "loss": 0.8173,
+      "step": 11023
+    },
+    {
+      "epoch": 1.878239063806074,
+      "grad_norm": 1.796875,
+      "learning_rate": 6.087406730197414e-06,
+      "loss": 0.8182,
+      "step": 11024
+    },
+    {
+      "epoch": 1.878410527895064,
+      "grad_norm": 1.65625,
+      "learning_rate": 6.085744833583111e-06,
+      "loss": 0.9004,
+      "step": 11025
+    },
+    {
+      "epoch": 1.8785819919840538,
+      "grad_norm": 1.7109375,
+      "learning_rate": 6.08408306462741e-06,
+      "loss": 0.897,
+      "step": 11026
+    },
+    {
+      "epoch": 1.8787534560730437,
+      "grad_norm": 1.6796875,
+      "learning_rate": 6.082421423384505e-06,
+      "loss": 0.8675,
+      "step": 11027
+    },
+    {
+      "epoch": 1.8789249201620337,
+      "grad_norm": 1.6171875,
+      "learning_rate": 6.0807599099085915e-06,
+      "loss": 0.7529,
+      "step": 11028
+    },
+    {
+      "epoch": 1.8790963842510235,
+      "grad_norm": 1.671875,
+      "learning_rate": 6.079098524253853e-06,
+      "loss": 0.8412,
+      "step": 11029
+    },
+    {
+      "epoch": 1.8792678483400134,
+      "grad_norm": 1.6015625,
+      "learning_rate": 6.077437266474478e-06,
+      "loss": 0.85,
+      "step": 11030
+    },
+    {
+      "epoch": 1.8794393124290032,
+      "grad_norm": 1.796875,
+      "learning_rate": 6.075776136624649e-06,
+      "loss": 0.9143,
+      "step": 11031
+    },
+    {
+      "epoch": 1.879610776517993,
+      "grad_norm": 1.6953125,
+      "learning_rate": 6.074115134758532e-06,
+      "loss": 0.8095,
+      "step": 11032
+    },
+    {
+      "epoch": 1.8797822406069828,
+      "grad_norm": 1.59375,
+      "learning_rate": 6.0724542609303035e-06,
+      "loss": 0.7707,
+      "step": 11033
+    },
+    {
+      "epoch": 1.8799537046959727,
+      "grad_norm": 1.703125,
+      "learning_rate": 6.070793515194133e-06,
+      "loss": 0.8904,
+      "step": 11034
+    },
+    {
+      "epoch": 1.8801251687849625,
+      "grad_norm": 1.6640625,
+      "learning_rate": 6.069132897604182e-06,
+      "loss": 0.8905,
+      "step": 11035
+    },
+    {
+      "epoch": 1.8802966328739523,
+      "grad_norm": 1.7421875,
+      "learning_rate": 6.06747240821461e-06,
+      "loss": 0.8963,
+      "step": 11036
+    },
+    {
+      "epoch": 1.8804680969629424,
+      "grad_norm": 1.65625,
+      "learning_rate": 6.065812047079569e-06,
+      "loss": 0.9203,
+      "step": 11037
+    },
+    {
+      "epoch": 1.8806395610519322,
+      "grad_norm": 1.7421875,
+      "learning_rate": 6.064151814253214e-06,
+      "loss": 0.9206,
+      "step": 11038
+    },
+    {
+      "epoch": 1.880811025140922,
+      "grad_norm": 1.7578125,
+      "learning_rate": 6.062491709789688e-06,
+      "loss": 0.8701,
+      "step": 11039
+    },
+    {
+      "epoch": 1.880982489229912,
+      "grad_norm": 1.6796875,
+      "learning_rate": 6.060831733743136e-06,
+      "loss": 0.8292,
+      "step": 11040
+    },
+    {
+      "epoch": 1.881153953318902,
+      "grad_norm": 1.7421875,
+      "learning_rate": 6.059171886167694e-06,
+      "loss": 0.8637,
+      "step": 11041
+    },
+    {
+      "epoch": 1.8813254174078917,
+      "grad_norm": 1.7734375,
+      "learning_rate": 6.0575121671174985e-06,
+      "loss": 0.9126,
+      "step": 11042
+    },
+    {
+      "epoch": 1.8814968814968815,
+      "grad_norm": 1.6875,
+      "learning_rate": 6.055852576646677e-06,
+      "loss": 0.8704,
+      "step": 11043
+    },
+    {
+      "epoch": 1.8816683455858714,
+      "grad_norm": 1.703125,
+      "learning_rate": 6.0541931148093525e-06,
+      "loss": 0.9013,
+      "step": 11044
+    },
+    {
+      "epoch": 1.8818398096748612,
+      "grad_norm": 1.6796875,
+      "learning_rate": 6.052533781659651e-06,
+      "loss": 0.8459,
+      "step": 11045
+    },
+    {
+      "epoch": 1.882011273763851,
+      "grad_norm": 1.8046875,
+      "learning_rate": 6.050874577251686e-06,
+      "loss": 0.9123,
+      "step": 11046
+    },
+    {
+      "epoch": 1.8821827378528408,
+      "grad_norm": 1.65625,
+      "learning_rate": 6.049215501639574e-06,
+      "loss": 0.8889,
+      "step": 11047
+    },
+    {
+      "epoch": 1.8823542019418307,
+      "grad_norm": 1.6328125,
+      "learning_rate": 6.04755655487742e-06,
+      "loss": 0.7668,
+      "step": 11048
+    },
+    {
+      "epoch": 1.8825256660308207,
+      "grad_norm": 1.6640625,
+      "learning_rate": 6.0458977370193316e-06,
+      "loss": 0.809,
+      "step": 11049
+    },
+    {
+      "epoch": 1.8826971301198105,
+      "grad_norm": 1.6953125,
+      "learning_rate": 6.044239048119407e-06,
+      "loss": 0.8047,
+      "step": 11050
+    },
+    {
+      "epoch": 1.8828685942088004,
+      "grad_norm": 1.703125,
+      "learning_rate": 6.042580488231744e-06,
+      "loss": 0.7854,
+      "step": 11051
+    },
+    {
+      "epoch": 1.8830400582977904,
+      "grad_norm": 1.7578125,
+      "learning_rate": 6.040922057410432e-06,
+      "loss": 0.9012,
+      "step": 11052
+    },
+    {
+      "epoch": 1.8832115223867802,
+      "grad_norm": 1.625,
+      "learning_rate": 6.039263755709561e-06,
+      "loss": 0.8462,
+      "step": 11053
+    },
+    {
+      "epoch": 1.88338298647577,
+      "grad_norm": 1.6953125,
+      "learning_rate": 6.037605583183217e-06,
+      "loss": 0.8311,
+      "step": 11054
+    },
+    {
+      "epoch": 1.88355445056476,
+      "grad_norm": 1.734375,
+      "learning_rate": 6.035947539885472e-06,
+      "loss": 0.9026,
+      "step": 11055
+    },
+    {
+      "epoch": 1.8837259146537497,
+      "grad_norm": 1.890625,
+      "learning_rate": 6.034289625870405e-06,
+      "loss": 0.8557,
+      "step": 11056
+    },
+    {
+      "epoch": 1.8838973787427395,
+      "grad_norm": 1.7734375,
+      "learning_rate": 6.032631841192088e-06,
+      "loss": 0.8124,
+      "step": 11057
+    },
+    {
+      "epoch": 1.8840688428317294,
+      "grad_norm": 1.59375,
+      "learning_rate": 6.030974185904586e-06,
+      "loss": 0.8698,
+      "step": 11058
+    },
+    {
+      "epoch": 1.8842403069207192,
+      "grad_norm": 1.6953125,
+      "learning_rate": 6.029316660061961e-06,
+      "loss": 0.9407,
+      "step": 11059
+    },
+    {
+      "epoch": 1.884411771009709,
+      "grad_norm": 1.609375,
+      "learning_rate": 6.027659263718273e-06,
+      "loss": 0.7825,
+      "step": 11060
+    },
+    {
+      "epoch": 1.8845832350986989,
+      "grad_norm": 1.7109375,
+      "learning_rate": 6.026001996927574e-06,
+      "loss": 0.843,
+      "step": 11061
+    },
+    {
+      "epoch": 1.884754699187689,
+      "grad_norm": 1.703125,
+      "learning_rate": 6.0243448597439154e-06,
+      "loss": 0.9317,
+      "step": 11062
+    },
+    {
+      "epoch": 1.8849261632766787,
+      "grad_norm": 1.609375,
+      "learning_rate": 6.0226878522213385e-06,
+      "loss": 0.8107,
+      "step": 11063
+    },
+    {
+      "epoch": 1.8850976273656685,
+      "grad_norm": 1.6875,
+      "learning_rate": 6.02103097441389e-06,
+      "loss": 0.7993,
+      "step": 11064
+    },
+    {
+      "epoch": 1.8852690914546586,
+      "grad_norm": 1.71875,
+      "learning_rate": 6.0193742263756115e-06,
+      "loss": 0.8842,
+      "step": 11065
+    },
+    {
+      "epoch": 1.8854405555436484,
+      "grad_norm": 1.75,
+      "learning_rate": 6.017717608160522e-06,
+      "loss": 0.8449,
+      "step": 11066
+    },
+    {
+      "epoch": 1.8856120196326382,
+      "grad_norm": 1.78125,
+      "learning_rate": 6.0160611198226595e-06,
+      "loss": 0.8286,
+      "step": 11067
+    },
+    {
+      "epoch": 1.885783483721628,
+      "grad_norm": 1.671875,
+      "learning_rate": 6.014404761416044e-06,
+      "loss": 0.8558,
+      "step": 11068
+    },
+    {
+      "epoch": 1.885954947810618,
+      "grad_norm": 1.6484375,
+      "learning_rate": 6.012748532994699e-06,
+      "loss": 0.7994,
+      "step": 11069
+    },
+    {
+      "epoch": 1.8861264118996077,
+      "grad_norm": 1.6796875,
+      "learning_rate": 6.011092434612639e-06,
+      "loss": 0.8989,
+      "step": 11070
+    },
+    {
+      "epoch": 1.8862978759885975,
+      "grad_norm": 1.6796875,
+      "learning_rate": 6.009436466323873e-06,
+      "loss": 0.8535,
+      "step": 11071
+    },
+    {
+      "epoch": 1.8864693400775874,
+      "grad_norm": 1.6328125,
+      "learning_rate": 6.007780628182413e-06,
+      "loss": 0.846,
+      "step": 11072
+    },
+    {
+      "epoch": 1.8866408041665772,
+      "grad_norm": 1.703125,
+      "learning_rate": 6.00612492024226e-06,
+      "loss": 0.8123,
+      "step": 11073
+    },
+    {
+      "epoch": 1.8868122682555672,
+      "grad_norm": 1.6484375,
+      "learning_rate": 6.004469342557413e-06,
+      "loss": 0.7837,
+      "step": 11074
+    },
+    {
+      "epoch": 1.886983732344557,
+      "grad_norm": 1.6015625,
+      "learning_rate": 6.002813895181865e-06,
+      "loss": 0.83,
+      "step": 11075
+    },
+    {
+      "epoch": 1.887155196433547,
+      "grad_norm": 1.6953125,
+      "learning_rate": 6.00115857816961e-06,
+      "loss": 0.8554,
+      "step": 11076
+    },
+    {
+      "epoch": 1.887326660522537,
+      "grad_norm": 1.7421875,
+      "learning_rate": 5.999503391574635e-06,
+      "loss": 0.8573,
+      "step": 11077
+    },
+    {
+      "epoch": 1.8874981246115268,
+      "grad_norm": 1.6953125,
+      "learning_rate": 5.9978483354509155e-06,
+      "loss": 0.7935,
+      "step": 11078
+    },
+    {
+      "epoch": 1.8876695887005166,
+      "grad_norm": 1.765625,
+      "learning_rate": 5.99619340985243e-06,
+      "loss": 0.8641,
+      "step": 11079
+    },
+    {
+      "epoch": 1.8878410527895064,
+      "grad_norm": 1.6171875,
+      "learning_rate": 5.9945386148331565e-06,
+      "loss": 0.8151,
+      "step": 11080
+    },
+    {
+      "epoch": 1.8880125168784962,
+      "grad_norm": 1.625,
+      "learning_rate": 5.992883950447062e-06,
+      "loss": 0.8537,
+      "step": 11081
+    },
+    {
+      "epoch": 1.888183980967486,
+      "grad_norm": 1.703125,
+      "learning_rate": 5.99122941674811e-06,
+      "loss": 0.8656,
+      "step": 11082
+    },
+    {
+      "epoch": 1.888355445056476,
+      "grad_norm": 1.6875,
+      "learning_rate": 5.989575013790264e-06,
+      "loss": 0.8776,
+      "step": 11083
+    },
+    {
+      "epoch": 1.8885269091454657,
+      "grad_norm": 1.8125,
+      "learning_rate": 5.987920741627479e-06,
+      "loss": 0.8657,
+      "step": 11084
+    },
+    {
+      "epoch": 1.8886983732344556,
+      "grad_norm": 1.6484375,
+      "learning_rate": 5.986266600313706e-06,
+      "loss": 0.9305,
+      "step": 11085
+    },
+    {
+      "epoch": 1.8888698373234456,
+      "grad_norm": 1.7890625,
+      "learning_rate": 5.984612589902893e-06,
+      "loss": 0.9352,
+      "step": 11086
+    },
+    {
+      "epoch": 1.8890413014124354,
+      "grad_norm": 1.6328125,
+      "learning_rate": 5.982958710448984e-06,
+      "loss": 0.8567,
+      "step": 11087
+    },
+    {
+      "epoch": 1.8892127655014253,
+      "grad_norm": 1.71875,
+      "learning_rate": 5.9813049620059206e-06,
+      "loss": 0.8659,
+      "step": 11088
+    },
+    {
+      "epoch": 1.8893842295904153,
+      "grad_norm": 1.734375,
+      "learning_rate": 5.979651344627633e-06,
+      "loss": 0.8445,
+      "step": 11089
+    },
+    {
+      "epoch": 1.8895556936794051,
+      "grad_norm": 1.84375,
+      "learning_rate": 5.977997858368055e-06,
+      "loss": 0.7571,
+      "step": 11090
+    },
+    {
+      "epoch": 1.889727157768395,
+      "grad_norm": 1.6953125,
+      "learning_rate": 5.976344503281113e-06,
+      "loss": 0.7836,
+      "step": 11091
+    },
+    {
+      "epoch": 1.8898986218573848,
+      "grad_norm": 1.5703125,
+      "learning_rate": 5.974691279420727e-06,
+      "loss": 0.8175,
+      "step": 11092
+    },
+    {
+      "epoch": 1.8900700859463746,
+      "grad_norm": 1.671875,
+      "learning_rate": 5.973038186840816e-06,
+      "loss": 0.8845,
+      "step": 11093
+    },
+    {
+      "epoch": 1.8902415500353644,
+      "grad_norm": 1.7734375,
+      "learning_rate": 5.971385225595294e-06,
+      "loss": 0.8002,
+      "step": 11094
+    },
+    {
+      "epoch": 1.8904130141243543,
+      "grad_norm": 1.671875,
+      "learning_rate": 5.969732395738071e-06,
+      "loss": 0.8699,
+      "step": 11095
+    },
+    {
+      "epoch": 1.890584478213344,
+      "grad_norm": 1.7109375,
+      "learning_rate": 5.968079697323052e-06,
+      "loss": 0.7574,
+      "step": 11096
+    },
+    {
+      "epoch": 1.890755942302334,
+      "grad_norm": 1.65625,
+      "learning_rate": 5.966427130404136e-06,
+      "loss": 0.8268,
+      "step": 11097
+    },
+    {
+      "epoch": 1.890927406391324,
+      "grad_norm": 1.7109375,
+      "learning_rate": 5.964774695035219e-06,
+      "loss": 0.8457,
+      "step": 11098
+    },
+    {
+      "epoch": 1.8910988704803138,
+      "grad_norm": 1.96875,
+      "learning_rate": 5.963122391270195e-06,
+      "loss": 0.9298,
+      "step": 11099
+    },
+    {
+      "epoch": 1.8912703345693036,
+      "grad_norm": 1.7265625,
+      "learning_rate": 5.961470219162955e-06,
+      "loss": 0.8533,
+      "step": 11100
+    },
+    {
+      "epoch": 1.8914417986582936,
+      "grad_norm": 1.765625,
+      "learning_rate": 5.959818178767376e-06,
+      "loss": 0.9166,
+      "step": 11101
+    },
+    {
+      "epoch": 1.8916132627472835,
+      "grad_norm": 1.6875,
+      "learning_rate": 5.9581662701373386e-06,
+      "loss": 0.8582,
+      "step": 11102
+    },
+    {
+      "epoch": 1.8917847268362733,
+      "grad_norm": 1.78125,
+      "learning_rate": 5.9565144933267205e-06,
+      "loss": 0.7822,
+      "step": 11103
+    },
+    {
+      "epoch": 1.8919561909252631,
+      "grad_norm": 1.7421875,
+      "learning_rate": 5.9548628483893915e-06,
+      "loss": 0.9107,
+      "step": 11104
+    },
+    {
+      "epoch": 1.892127655014253,
+      "grad_norm": 1.796875,
+      "learning_rate": 5.953211335379217e-06,
+      "loss": 0.9657,
+      "step": 11105
+    },
+    {
+      "epoch": 1.8922991191032428,
+      "grad_norm": 1.6796875,
+      "learning_rate": 5.951559954350059e-06,
+      "loss": 0.9031,
+      "step": 11106
+    },
+    {
+      "epoch": 1.8924705831922326,
+      "grad_norm": 1.6171875,
+      "learning_rate": 5.949908705355778e-06,
+      "loss": 0.7995,
+      "step": 11107
+    },
+    {
+      "epoch": 1.8926420472812224,
+      "grad_norm": 1.7578125,
+      "learning_rate": 5.948257588450224e-06,
+      "loss": 0.8743,
+      "step": 11108
+    },
+    {
+      "epoch": 1.8928135113702123,
+      "grad_norm": 1.671875,
+      "learning_rate": 5.946606603687246e-06,
+      "loss": 0.8933,
+      "step": 11109
+    },
+    {
+      "epoch": 1.8929849754592023,
+      "grad_norm": 1.7109375,
+      "learning_rate": 5.944955751120691e-06,
+      "loss": 0.8185,
+      "step": 11110
+    },
+    {
+      "epoch": 1.8931564395481921,
+      "grad_norm": 1.6875,
+      "learning_rate": 5.943305030804403e-06,
+      "loss": 0.8492,
+      "step": 11111
+    },
+    {
+      "epoch": 1.893327903637182,
+      "grad_norm": 1.7109375,
+      "learning_rate": 5.94165444279221e-06,
+      "loss": 0.8794,
+      "step": 11112
+    },
+    {
+      "epoch": 1.893499367726172,
+      "grad_norm": 1.6796875,
+      "learning_rate": 5.94000398713795e-06,
+      "loss": 0.8339,
+      "step": 11113
+    },
+    {
+      "epoch": 1.8936708318151618,
+      "grad_norm": 1.625,
+      "learning_rate": 5.938353663895447e-06,
+      "loss": 0.8516,
+      "step": 11114
+    },
+    {
+      "epoch": 1.8938422959041517,
+      "grad_norm": 1.65625,
+      "learning_rate": 5.936703473118526e-06,
+      "loss": 0.8399,
+      "step": 11115
+    },
+    {
+      "epoch": 1.8940137599931415,
+      "grad_norm": 1.6484375,
+      "learning_rate": 5.935053414861005e-06,
+      "loss": 0.7824,
+      "step": 11116
+    },
+    {
+      "epoch": 1.8941852240821313,
+      "grad_norm": 1.6875,
+      "learning_rate": 5.933403489176701e-06,
+      "loss": 0.8342,
+      "step": 11117
+    },
+    {
+      "epoch": 1.8943566881711211,
+      "grad_norm": 1.640625,
+      "learning_rate": 5.931753696119419e-06,
+      "loss": 0.8584,
+      "step": 11118
+    },
+    {
+      "epoch": 1.894528152260111,
+      "grad_norm": 1.7109375,
+      "learning_rate": 5.930104035742972e-06,
+      "loss": 0.8126,
+      "step": 11119
+    },
+    {
+      "epoch": 1.8946996163491008,
+      "grad_norm": 1.6796875,
+      "learning_rate": 5.928454508101156e-06,
+      "loss": 0.8786,
+      "step": 11120
+    },
+    {
+      "epoch": 1.8948710804380906,
+      "grad_norm": 1.671875,
+      "learning_rate": 5.926805113247772e-06,
+      "loss": 0.8117,
+      "step": 11121
+    },
+    {
+      "epoch": 1.8950425445270807,
+      "grad_norm": 1.71875,
+      "learning_rate": 5.9251558512366115e-06,
+      "loss": 0.8673,
+      "step": 11122
+    },
+    {
+      "epoch": 1.8952140086160705,
+      "grad_norm": 1.6328125,
+      "learning_rate": 5.923506722121467e-06,
+      "loss": 0.898,
+      "step": 11123
+    },
+    {
+      "epoch": 1.8953854727050603,
+      "grad_norm": 1.609375,
+      "learning_rate": 5.9218577259561146e-06,
+      "loss": 0.7834,
+      "step": 11124
+    },
+    {
+      "epoch": 1.8955569367940504,
+      "grad_norm": 1.7265625,
+      "learning_rate": 5.920208862794339e-06,
+      "loss": 0.8138,
+      "step": 11125
+    },
+    {
+      "epoch": 1.8957284008830402,
+      "grad_norm": 1.703125,
+      "learning_rate": 5.9185601326899145e-06,
+      "loss": 0.8742,
+      "step": 11126
+    },
+    {
+      "epoch": 1.89589986497203,
+      "grad_norm": 1.7734375,
+      "learning_rate": 5.9169115356966135e-06,
+      "loss": 0.8483,
+      "step": 11127
+    },
+    {
+      "epoch": 1.8960713290610198,
+      "grad_norm": 1.7265625,
+      "learning_rate": 5.915263071868203e-06,
+      "loss": 0.8284,
+      "step": 11128
+    },
+    {
+      "epoch": 1.8962427931500097,
+      "grad_norm": 1.6640625,
+      "learning_rate": 5.913614741258446e-06,
+      "loss": 0.8855,
+      "step": 11129
+    },
+    {
+      "epoch": 1.8964142572389995,
+      "grad_norm": 1.7734375,
+      "learning_rate": 5.911966543921101e-06,
+      "loss": 0.8719,
+      "step": 11130
+    },
+    {
+      "epoch": 1.8965857213279893,
+      "grad_norm": 1.671875,
+      "learning_rate": 5.91031847990992e-06,
+      "loss": 0.8622,
+      "step": 11131
+    },
+    {
+      "epoch": 1.8967571854169791,
+      "grad_norm": 1.6875,
+      "learning_rate": 5.908670549278655e-06,
+      "loss": 0.832,
+      "step": 11132
+    },
+    {
+      "epoch": 1.896928649505969,
+      "grad_norm": 1.703125,
+      "learning_rate": 5.907022752081047e-06,
+      "loss": 0.8111,
+      "step": 11133
+    },
+    {
+      "epoch": 1.897100113594959,
+      "grad_norm": 1.625,
+      "learning_rate": 5.905375088370842e-06,
+      "loss": 0.8047,
+      "step": 11134
+    },
+    {
+      "epoch": 1.8972715776839488,
+      "grad_norm": 1.7421875,
+      "learning_rate": 5.903727558201776e-06,
+      "loss": 0.9245,
+      "step": 11135
+    },
+    {
+      "epoch": 1.8974430417729387,
+      "grad_norm": 1.703125,
+      "learning_rate": 5.902080161627577e-06,
+      "loss": 0.9039,
+      "step": 11136
+    },
+    {
+      "epoch": 1.8976145058619287,
+      "grad_norm": 1.6796875,
+      "learning_rate": 5.900432898701977e-06,
+      "loss": 0.7891,
+      "step": 11137
+    },
+    {
+      "epoch": 1.8977859699509185,
+      "grad_norm": 1.6796875,
+      "learning_rate": 5.898785769478695e-06,
+      "loss": 0.8917,
+      "step": 11138
+    },
+    {
+      "epoch": 1.8979574340399084,
+      "grad_norm": 1.7109375,
+      "learning_rate": 5.897138774011455e-06,
+      "loss": 0.8707,
+      "step": 11139
+    },
+    {
+      "epoch": 1.8981288981288982,
+      "grad_norm": 1.7109375,
+      "learning_rate": 5.8954919123539675e-06,
+      "loss": 0.8902,
+      "step": 11140
+    },
+    {
+      "epoch": 1.898300362217888,
+      "grad_norm": 1.6796875,
+      "learning_rate": 5.893845184559948e-06,
+      "loss": 0.8544,
+      "step": 11141
+    },
+    {
+      "epoch": 1.8984718263068778,
+      "grad_norm": 1.6328125,
+      "learning_rate": 5.892198590683096e-06,
+      "loss": 0.811,
+      "step": 11142
+    },
+    {
+      "epoch": 1.8986432903958677,
+      "grad_norm": 1.6328125,
+      "learning_rate": 5.890552130777119e-06,
+      "loss": 0.8141,
+      "step": 11143
+    },
+    {
+      "epoch": 1.8988147544848575,
+      "grad_norm": 1.65625,
+      "learning_rate": 5.88890580489571e-06,
+      "loss": 0.6992,
+      "step": 11144
+    },
+    {
+      "epoch": 1.8989862185738473,
+      "grad_norm": 1.6875,
+      "learning_rate": 5.887259613092564e-06,
+      "loss": 0.7467,
+      "step": 11145
+    },
+    {
+      "epoch": 1.8991576826628374,
+      "grad_norm": 1.671875,
+      "learning_rate": 5.885613555421372e-06,
+      "loss": 0.8146,
+      "step": 11146
+    },
+    {
+      "epoch": 1.8993291467518272,
+      "grad_norm": 1.734375,
+      "learning_rate": 5.883967631935813e-06,
+      "loss": 0.7893,
+      "step": 11147
+    },
+    {
+      "epoch": 1.899500610840817,
+      "grad_norm": 1.859375,
+      "learning_rate": 5.882321842689569e-06,
+      "loss": 0.8589,
+      "step": 11148
+    },
+    {
+      "epoch": 1.899672074929807,
+      "grad_norm": 1.6171875,
+      "learning_rate": 5.880676187736316e-06,
+      "loss": 0.7366,
+      "step": 11149
+    },
+    {
+      "epoch": 1.8998435390187969,
+      "grad_norm": 1.6875,
+      "learning_rate": 5.8790306671297234e-06,
+      "loss": 0.8711,
+      "step": 11150
+    },
+    {
+      "epoch": 1.9000150031077867,
+      "grad_norm": 1.6484375,
+      "learning_rate": 5.87738528092346e-06,
+      "loss": 0.8846,
+      "step": 11151
+    },
+    {
+      "epoch": 1.9001864671967765,
+      "grad_norm": 1.65625,
+      "learning_rate": 5.875740029171185e-06,
+      "loss": 0.8274,
+      "step": 11152
+    },
+    {
+      "epoch": 1.9003579312857664,
+      "grad_norm": 1.78125,
+      "learning_rate": 5.87409491192656e-06,
+      "loss": 0.7669,
+      "step": 11153
+    },
+    {
+      "epoch": 1.9005293953747562,
+      "grad_norm": 1.6328125,
+      "learning_rate": 5.872449929243236e-06,
+      "loss": 0.8362,
+      "step": 11154
+    },
+    {
+      "epoch": 1.900700859463746,
+      "grad_norm": 1.7265625,
+      "learning_rate": 5.870805081174862e-06,
+      "loss": 0.8204,
+      "step": 11155
+    },
+    {
+      "epoch": 1.9008723235527358,
+      "grad_norm": 1.7109375,
+      "learning_rate": 5.869160367775084e-06,
+      "loss": 0.7579,
+      "step": 11156
+    },
+    {
+      "epoch": 1.9010437876417257,
+      "grad_norm": 1.7578125,
+      "learning_rate": 5.86751578909754e-06,
+      "loss": 0.8474,
+      "step": 11157
+    },
+    {
+      "epoch": 1.9012152517307155,
+      "grad_norm": 1.6875,
+      "learning_rate": 5.865871345195875e-06,
+      "loss": 0.8747,
+      "step": 11158
+    },
+    {
+      "epoch": 1.9013867158197055,
+      "grad_norm": 1.640625,
+      "learning_rate": 5.8642270361237066e-06,
+      "loss": 0.8889,
+      "step": 11159
+    },
+    {
+      "epoch": 1.9015581799086954,
+      "grad_norm": 1.71875,
+      "learning_rate": 5.8625828619346695e-06,
+      "loss": 0.8505,
+      "step": 11160
+    },
+    {
+      "epoch": 1.9017296439976852,
+      "grad_norm": 1.7578125,
+      "learning_rate": 5.860938822682385e-06,
+      "loss": 0.9267,
+      "step": 11161
+    },
+    {
+      "epoch": 1.9019011080866752,
+      "grad_norm": 1.78125,
+      "learning_rate": 5.859294918420473e-06,
+      "loss": 0.8676,
+      "step": 11162
+    },
+    {
+      "epoch": 1.902072572175665,
+      "grad_norm": 1.6015625,
+      "learning_rate": 5.857651149202545e-06,
+      "loss": 0.8717,
+      "step": 11163
+    },
+    {
+      "epoch": 1.9022440362646549,
+      "grad_norm": 1.6796875,
+      "learning_rate": 5.856007515082212e-06,
+      "loss": 0.8625,
+      "step": 11164
+    },
+    {
+      "epoch": 1.9024155003536447,
+      "grad_norm": 1.640625,
+      "learning_rate": 5.8543640161130765e-06,
+      "loss": 0.7168,
+      "step": 11165
+    },
+    {
+      "epoch": 1.9025869644426345,
+      "grad_norm": 1.671875,
+      "learning_rate": 5.852720652348744e-06,
+      "loss": 0.826,
+      "step": 11166
+    },
+    {
+      "epoch": 1.9027584285316244,
+      "grad_norm": 1.6953125,
+      "learning_rate": 5.851077423842807e-06,
+      "loss": 0.81,
+      "step": 11167
+    },
+    {
+      "epoch": 1.9029298926206142,
+      "grad_norm": 1.671875,
+      "learning_rate": 5.8494343306488595e-06,
+      "loss": 0.9127,
+      "step": 11168
+    },
+    {
+      "epoch": 1.903101356709604,
+      "grad_norm": 1.7734375,
+      "learning_rate": 5.847791372820493e-06,
+      "loss": 0.9623,
+      "step": 11169
+    },
+    {
+      "epoch": 1.9032728207985938,
+      "grad_norm": 1.6171875,
+      "learning_rate": 5.846148550411279e-06,
+      "loss": 0.8097,
+      "step": 11170
+    },
+    {
+      "epoch": 1.9034442848875839,
+      "grad_norm": 1.8125,
+      "learning_rate": 5.8445058634748055e-06,
+      "loss": 0.8239,
+      "step": 11171
+    },
+    {
+      "epoch": 1.9036157489765737,
+      "grad_norm": 1.578125,
+      "learning_rate": 5.842863312064642e-06,
+      "loss": 0.7607,
+      "step": 11172
+    },
+    {
+      "epoch": 1.9037872130655635,
+      "grad_norm": 1.7734375,
+      "learning_rate": 5.841220896234358e-06,
+      "loss": 0.8831,
+      "step": 11173
+    },
+    {
+      "epoch": 1.9039586771545536,
+      "grad_norm": 1.6875,
+      "learning_rate": 5.839578616037525e-06,
+      "loss": 0.8309,
+      "step": 11174
+    },
+    {
+      "epoch": 1.9041301412435434,
+      "grad_norm": 1.734375,
+      "learning_rate": 5.837936471527701e-06,
+      "loss": 0.8537,
+      "step": 11175
+    },
+    {
+      "epoch": 1.9043016053325332,
+      "grad_norm": 1.6484375,
+      "learning_rate": 5.836294462758441e-06,
+      "loss": 0.7846,
+      "step": 11176
+    },
+    {
+      "epoch": 1.904473069421523,
+      "grad_norm": 1.71875,
+      "learning_rate": 5.834652589783295e-06,
+      "loss": 0.8714,
+      "step": 11177
+    },
+    {
+      "epoch": 1.9046445335105129,
+      "grad_norm": 1.828125,
+      "learning_rate": 5.833010852655815e-06,
+      "loss": 0.9001,
+      "step": 11178
+    },
+    {
+      "epoch": 1.9048159975995027,
+      "grad_norm": 1.65625,
+      "learning_rate": 5.8313692514295416e-06,
+      "loss": 0.8025,
+      "step": 11179
+    },
+    {
+      "epoch": 1.9049874616884925,
+      "grad_norm": 1.6875,
+      "learning_rate": 5.829727786158011e-06,
+      "loss": 0.88,
+      "step": 11180
+    },
+    {
+      "epoch": 1.9051589257774824,
+      "grad_norm": 1.6328125,
+      "learning_rate": 5.828086456894769e-06,
+      "loss": 0.8556,
+      "step": 11181
+    },
+    {
+      "epoch": 1.9053303898664722,
+      "grad_norm": 1.734375,
+      "learning_rate": 5.826445263693335e-06,
+      "loss": 0.8384,
+      "step": 11182
+    },
+    {
+      "epoch": 1.9055018539554622,
+      "grad_norm": 1.765625,
+      "learning_rate": 5.824804206607235e-06,
+      "loss": 0.8997,
+      "step": 11183
+    },
+    {
+      "epoch": 1.905673318044452,
+      "grad_norm": 1.7578125,
+      "learning_rate": 5.823163285689992e-06,
+      "loss": 0.8814,
+      "step": 11184
+    },
+    {
+      "epoch": 1.9058447821334419,
+      "grad_norm": 1.65625,
+      "learning_rate": 5.821522500995125e-06,
+      "loss": 0.8122,
+      "step": 11185
+    },
+    {
+      "epoch": 1.906016246222432,
+      "grad_norm": 1.6171875,
+      "learning_rate": 5.819881852576141e-06,
+      "loss": 0.7983,
+      "step": 11186
+    },
+    {
+      "epoch": 1.9061877103114218,
+      "grad_norm": 1.765625,
+      "learning_rate": 5.818241340486554e-06,
+      "loss": 0.8247,
+      "step": 11187
+    },
+    {
+      "epoch": 1.9063591744004116,
+      "grad_norm": 1.5625,
+      "learning_rate": 5.8166009647798616e-06,
+      "loss": 0.7229,
+      "step": 11188
+    },
+    {
+      "epoch": 1.9065306384894014,
+      "grad_norm": 1.671875,
+      "learning_rate": 5.8149607255095665e-06,
+      "loss": 0.7919,
+      "step": 11189
+    },
+    {
+      "epoch": 1.9067021025783912,
+      "grad_norm": 1.6796875,
+      "learning_rate": 5.813320622729159e-06,
+      "loss": 0.8322,
+      "step": 11190
+    },
+    {
+      "epoch": 1.906873566667381,
+      "grad_norm": 1.703125,
+      "learning_rate": 5.811680656492134e-06,
+      "loss": 0.8727,
+      "step": 11191
+    },
+    {
+      "epoch": 1.9070450307563709,
+      "grad_norm": 1.6796875,
+      "learning_rate": 5.810040826851978e-06,
+      "loss": 0.793,
+      "step": 11192
+    },
+    {
+      "epoch": 1.9072164948453607,
+      "grad_norm": 1.703125,
+      "learning_rate": 5.808401133862165e-06,
+      "loss": 0.7886,
+      "step": 11193
+    },
+    {
+      "epoch": 1.9073879589343505,
+      "grad_norm": 1.7265625,
+      "learning_rate": 5.8067615775761746e-06,
+      "loss": 0.8129,
+      "step": 11194
+    },
+    {
+      "epoch": 1.9075594230233406,
+      "grad_norm": 1.7890625,
+      "learning_rate": 5.8051221580474786e-06,
+      "loss": 0.9707,
+      "step": 11195
+    },
+    {
+      "epoch": 1.9077308871123304,
+      "grad_norm": 1.6328125,
+      "learning_rate": 5.803482875329543e-06,
+      "loss": 0.7983,
+      "step": 11196
+    },
+    {
+      "epoch": 1.9079023512013202,
+      "grad_norm": 1.6875,
+      "learning_rate": 5.801843729475836e-06,
+      "loss": 0.7935,
+      "step": 11197
+    },
+    {
+      "epoch": 1.9080738152903103,
+      "grad_norm": 1.6171875,
+      "learning_rate": 5.800204720539813e-06,
+      "loss": 0.8917,
+      "step": 11198
+    },
+    {
+      "epoch": 1.9082452793793,
+      "grad_norm": 1.6171875,
+      "learning_rate": 5.798565848574931e-06,
+      "loss": 0.799,
+      "step": 11199
+    },
+    {
+      "epoch": 1.90841674346829,
+      "grad_norm": 1.7734375,
+      "learning_rate": 5.796927113634637e-06,
+      "loss": 0.8679,
+      "step": 11200
+    },
+    {
+      "epoch": 1.90841674346829,
+      "eval_loss": 0.8319188952445984,
+      "eval_runtime": 835.9538,
+      "eval_samples_per_second": 2.989,
+      "eval_steps_per_second": 2.989,
+      "step": 11200
+    },
+    {
+      "epoch": 1.9085882075572798,
+      "grad_norm": 1.75,
+      "learning_rate": 5.795288515772377e-06,
+      "loss": 0.8503,
+      "step": 11201
+    },
+    {
+      "epoch": 1.9087596716462696,
+      "grad_norm": 1.71875,
+      "learning_rate": 5.7936500550415934e-06,
+      "loss": 0.8268,
+      "step": 11202
+    },
+    {
+      "epoch": 1.9089311357352594,
+      "grad_norm": 1.9453125,
+      "learning_rate": 5.792011731495719e-06,
+      "loss": 0.9486,
+      "step": 11203
+    },
+    {
+      "epoch": 1.9091025998242492,
+      "grad_norm": 1.6171875,
+      "learning_rate": 5.7903735451881935e-06,
+      "loss": 0.823,
+      "step": 11204
+    },
+    {
+      "epoch": 1.909274063913239,
+      "grad_norm": 1.6953125,
+      "learning_rate": 5.788735496172435e-06,
+      "loss": 0.8247,
+      "step": 11205
+    },
+    {
+      "epoch": 1.9094455280022289,
+      "grad_norm": 1.953125,
+      "learning_rate": 5.7870975845018685e-06,
+      "loss": 0.8613,
+      "step": 11206
+    },
+    {
+      "epoch": 1.909616992091219,
+      "grad_norm": 1.6640625,
+      "learning_rate": 5.785459810229914e-06,
+      "loss": 0.8429,
+      "step": 11207
+    },
+    {
+      "epoch": 1.9097884561802088,
+      "grad_norm": 1.7734375,
+      "learning_rate": 5.783822173409988e-06,
+      "loss": 0.8999,
+      "step": 11208
+    },
+    {
+      "epoch": 1.9099599202691986,
+      "grad_norm": 1.734375,
+      "learning_rate": 5.782184674095495e-06,
+      "loss": 0.8341,
+      "step": 11209
+    },
+    {
+      "epoch": 1.9101313843581886,
+      "grad_norm": 1.671875,
+      "learning_rate": 5.780547312339844e-06,
+      "loss": 0.8422,
+      "step": 11210
+    },
+    {
+      "epoch": 1.9103028484471785,
+      "grad_norm": 1.75,
+      "learning_rate": 5.778910088196435e-06,
+      "loss": 0.8455,
+      "step": 11211
+    },
+    {
+      "epoch": 1.9104743125361683,
+      "grad_norm": 1.703125,
+      "learning_rate": 5.77727300171866e-06,
+      "loss": 0.9009,
+      "step": 11212
+    },
+    {
+      "epoch": 1.910645776625158,
+      "grad_norm": 1.6953125,
+      "learning_rate": 5.775636052959916e-06,
+      "loss": 0.9527,
+      "step": 11213
+    },
+    {
+      "epoch": 1.910817240714148,
+      "grad_norm": 1.5859375,
+      "learning_rate": 5.773999241973587e-06,
+      "loss": 0.8251,
+      "step": 11214
+    },
+    {
+      "epoch": 1.9109887048031378,
+      "grad_norm": 1.6953125,
+      "learning_rate": 5.7723625688130565e-06,
+      "loss": 0.7843,
+      "step": 11215
+    },
+    {
+      "epoch": 1.9111601688921276,
+      "grad_norm": 1.6015625,
+      "learning_rate": 5.770726033531704e-06,
+      "loss": 0.8589,
+      "step": 11216
+    },
+    {
+      "epoch": 1.9113316329811174,
+      "grad_norm": 1.703125,
+      "learning_rate": 5.769089636182901e-06,
+      "loss": 0.8252,
+      "step": 11217
+    },
+    {
+      "epoch": 1.9115030970701072,
+      "grad_norm": 1.7265625,
+      "learning_rate": 5.767453376820016e-06,
+      "loss": 0.8801,
+      "step": 11218
+    },
+    {
+      "epoch": 1.9116745611590973,
+      "grad_norm": 1.6640625,
+      "learning_rate": 5.765817255496414e-06,
+      "loss": 0.8598,
+      "step": 11219
+    },
+    {
+      "epoch": 1.911846025248087,
+      "grad_norm": 1.609375,
+      "learning_rate": 5.764181272265456e-06,
+      "loss": 0.8349,
+      "step": 11220
+    },
+    {
+      "epoch": 1.912017489337077,
+      "grad_norm": 1.5859375,
+      "learning_rate": 5.762545427180499e-06,
+      "loss": 0.8272,
+      "step": 11221
+    },
+    {
+      "epoch": 1.912188953426067,
+      "grad_norm": 1.671875,
+      "learning_rate": 5.760909720294892e-06,
+      "loss": 0.8555,
+      "step": 11222
+    },
+    {
+      "epoch": 1.9123604175150568,
+      "grad_norm": 1.59375,
+      "learning_rate": 5.759274151661981e-06,
+      "loss": 0.8261,
+      "step": 11223
+    },
+    {
+      "epoch": 1.9125318816040466,
+      "grad_norm": 1.7265625,
+      "learning_rate": 5.757638721335111e-06,
+      "loss": 0.901,
+      "step": 11224
+    },
+    {
+      "epoch": 1.9127033456930365,
+      "grad_norm": 1.7109375,
+      "learning_rate": 5.756003429367615e-06,
+      "loss": 0.8994,
+      "step": 11225
+    },
+    {
+      "epoch": 1.9128748097820263,
+      "grad_norm": 1.71875,
+      "learning_rate": 5.7543682758128295e-06,
+      "loss": 0.8293,
+      "step": 11226
+    },
+    {
+      "epoch": 1.9130462738710161,
+      "grad_norm": 1.6640625,
+      "learning_rate": 5.752733260724086e-06,
+      "loss": 0.8326,
+      "step": 11227
+    },
+    {
+      "epoch": 1.913217737960006,
+      "grad_norm": 1.7109375,
+      "learning_rate": 5.751098384154701e-06,
+      "loss": 0.8523,
+      "step": 11228
+    },
+    {
+      "epoch": 1.9133892020489958,
+      "grad_norm": 1.6953125,
+      "learning_rate": 5.749463646157998e-06,
+      "loss": 0.8387,
+      "step": 11229
+    },
+    {
+      "epoch": 1.9135606661379856,
+      "grad_norm": 1.7109375,
+      "learning_rate": 5.74782904678729e-06,
+      "loss": 0.7988,
+      "step": 11230
+    },
+    {
+      "epoch": 1.9137321302269756,
+      "grad_norm": 1.75,
+      "learning_rate": 5.74619458609589e-06,
+      "loss": 0.7845,
+      "step": 11231
+    },
+    {
+      "epoch": 1.9139035943159655,
+      "grad_norm": 1.640625,
+      "learning_rate": 5.744560264137102e-06,
+      "loss": 0.9016,
+      "step": 11232
+    },
+    {
+      "epoch": 1.9140750584049553,
+      "grad_norm": 1.7265625,
+      "learning_rate": 5.7429260809642295e-06,
+      "loss": 0.8691,
+      "step": 11233
+    },
+    {
+      "epoch": 1.9142465224939453,
+      "grad_norm": 1.75,
+      "learning_rate": 5.741292036630568e-06,
+      "loss": 0.8374,
+      "step": 11234
+    },
+    {
+      "epoch": 1.9144179865829352,
+      "grad_norm": 1.6328125,
+      "learning_rate": 5.739658131189405e-06,
+      "loss": 0.8578,
+      "step": 11235
+    },
+    {
+      "epoch": 1.914589450671925,
+      "grad_norm": 1.7265625,
+      "learning_rate": 5.738024364694039e-06,
+      "loss": 0.9524,
+      "step": 11236
+    },
+    {
+      "epoch": 1.9147609147609148,
+      "grad_norm": 1.6796875,
+      "learning_rate": 5.736390737197745e-06,
+      "loss": 0.8665,
+      "step": 11237
+    },
+    {
+      "epoch": 1.9149323788499046,
+      "grad_norm": 1.7421875,
+      "learning_rate": 5.734757248753811e-06,
+      "loss": 0.8449,
+      "step": 11238
+    },
+    {
+      "epoch": 1.9151038429388945,
+      "grad_norm": 1.765625,
+      "learning_rate": 5.7331238994155e-06,
+      "loss": 0.9447,
+      "step": 11239
+    },
+    {
+      "epoch": 1.9152753070278843,
+      "grad_norm": 1.5859375,
+      "learning_rate": 5.7314906892360855e-06,
+      "loss": 0.8613,
+      "step": 11240
+    },
+    {
+      "epoch": 1.9154467711168741,
+      "grad_norm": 1.7109375,
+      "learning_rate": 5.729857618268832e-06,
+      "loss": 0.8167,
+      "step": 11241
+    },
+    {
+      "epoch": 1.915618235205864,
+      "grad_norm": 1.7265625,
+      "learning_rate": 5.728224686567004e-06,
+      "loss": 0.9106,
+      "step": 11242
+    },
+    {
+      "epoch": 1.915789699294854,
+      "grad_norm": 1.796875,
+      "learning_rate": 5.726591894183855e-06,
+      "loss": 0.7939,
+      "step": 11243
+    },
+    {
+      "epoch": 1.9159611633838438,
+      "grad_norm": 1.65625,
+      "learning_rate": 5.724959241172634e-06,
+      "loss": 0.7932,
+      "step": 11244
+    },
+    {
+      "epoch": 1.9161326274728336,
+      "grad_norm": 1.7421875,
+      "learning_rate": 5.723326727586593e-06,
+      "loss": 0.8271,
+      "step": 11245
+    },
+    {
+      "epoch": 1.9163040915618237,
+      "grad_norm": 1.828125,
+      "learning_rate": 5.721694353478971e-06,
+      "loss": 0.9315,
+      "step": 11246
+    },
+    {
+      "epoch": 1.9164755556508135,
+      "grad_norm": 1.6796875,
+      "learning_rate": 5.720062118903006e-06,
+      "loss": 0.8342,
+      "step": 11247
+    },
+    {
+      "epoch": 1.9166470197398033,
+      "grad_norm": 1.6484375,
+      "learning_rate": 5.718430023911932e-06,
+      "loss": 0.8443,
+      "step": 11248
+    },
+    {
+      "epoch": 1.9168184838287932,
+      "grad_norm": 1.734375,
+      "learning_rate": 5.7167980685589785e-06,
+      "loss": 0.8486,
+      "step": 11249
+    },
+    {
+      "epoch": 1.916989947917783,
+      "grad_norm": 1.6171875,
+      "learning_rate": 5.715166252897373e-06,
+      "loss": 0.8103,
+      "step": 11250
+    },
+    {
+      "epoch": 1.9171614120067728,
+      "grad_norm": 1.71875,
+      "learning_rate": 5.713534576980328e-06,
+      "loss": 0.8686,
+      "step": 11251
+    },
+    {
+      "epoch": 1.9173328760957626,
+      "grad_norm": 1.796875,
+      "learning_rate": 5.711903040861055e-06,
+      "loss": 0.8277,
+      "step": 11252
+    },
+    {
+      "epoch": 1.9175043401847525,
+      "grad_norm": 1.6953125,
+      "learning_rate": 5.7102716445927785e-06,
+      "loss": 0.8404,
+      "step": 11253
+    },
+    {
+      "epoch": 1.9176758042737423,
+      "grad_norm": 1.8671875,
+      "learning_rate": 5.708640388228697e-06,
+      "loss": 0.9247,
+      "step": 11254
+    },
+    {
+      "epoch": 1.9178472683627321,
+      "grad_norm": 1.671875,
+      "learning_rate": 5.707009271822011e-06,
+      "loss": 0.7726,
+      "step": 11255
+    },
+    {
+      "epoch": 1.9180187324517222,
+      "grad_norm": 1.703125,
+      "learning_rate": 5.705378295425919e-06,
+      "loss": 0.8043,
+      "step": 11256
+    },
+    {
+      "epoch": 1.918190196540712,
+      "grad_norm": 1.6953125,
+      "learning_rate": 5.703747459093611e-06,
+      "loss": 0.924,
+      "step": 11257
+    },
+    {
+      "epoch": 1.9183616606297018,
+      "grad_norm": 1.7109375,
+      "learning_rate": 5.702116762878278e-06,
+      "loss": 0.8636,
+      "step": 11258
+    },
+    {
+      "epoch": 1.9185331247186919,
+      "grad_norm": 1.640625,
+      "learning_rate": 5.700486206833103e-06,
+      "loss": 0.8355,
+      "step": 11259
+    },
+    {
+      "epoch": 1.9187045888076817,
+      "grad_norm": 1.6171875,
+      "learning_rate": 5.698855791011262e-06,
+      "loss": 0.817,
+      "step": 11260
+    },
+    {
+      "epoch": 1.9188760528966715,
+      "grad_norm": 1.5859375,
+      "learning_rate": 5.697225515465934e-06,
+      "loss": 0.8022,
+      "step": 11261
+    },
+    {
+      "epoch": 1.9190475169856613,
+      "grad_norm": 1.6953125,
+      "learning_rate": 5.69559538025028e-06,
+      "loss": 0.9162,
+      "step": 11262
+    },
+    {
+      "epoch": 1.9192189810746512,
+      "grad_norm": 1.6875,
+      "learning_rate": 5.693965385417471e-06,
+      "loss": 0.8811,
+      "step": 11263
+    },
+    {
+      "epoch": 1.919390445163641,
+      "grad_norm": 1.59375,
+      "learning_rate": 5.692335531020665e-06,
+      "loss": 0.7753,
+      "step": 11264
+    },
+    {
+      "epoch": 1.9195619092526308,
+      "grad_norm": 1.7578125,
+      "learning_rate": 5.69070581711302e-06,
+      "loss": 0.8442,
+      "step": 11265
+    },
+    {
+      "epoch": 1.9197333733416206,
+      "grad_norm": 1.7109375,
+      "learning_rate": 5.689076243747684e-06,
+      "loss": 0.8683,
+      "step": 11266
+    },
+    {
+      "epoch": 1.9199048374306105,
+      "grad_norm": 1.6953125,
+      "learning_rate": 5.687446810977806e-06,
+      "loss": 0.8505,
+      "step": 11267
+    },
+    {
+      "epoch": 1.9200763015196005,
+      "grad_norm": 1.6796875,
+      "learning_rate": 5.6858175188565266e-06,
+      "loss": 0.8469,
+      "step": 11268
+    },
+    {
+      "epoch": 1.9202477656085903,
+      "grad_norm": 1.6953125,
+      "learning_rate": 5.684188367436984e-06,
+      "loss": 0.8282,
+      "step": 11269
+    },
+    {
+      "epoch": 1.9204192296975802,
+      "grad_norm": 1.796875,
+      "learning_rate": 5.68255935677231e-06,
+      "loss": 0.8958,
+      "step": 11270
+    },
+    {
+      "epoch": 1.9205906937865702,
+      "grad_norm": 1.609375,
+      "learning_rate": 5.680930486915633e-06,
+      "loss": 0.7198,
+      "step": 11271
+    },
+    {
+      "epoch": 1.92076215787556,
+      "grad_norm": 1.6796875,
+      "learning_rate": 5.679301757920078e-06,
+      "loss": 0.8596,
+      "step": 11272
+    },
+    {
+      "epoch": 1.9209336219645499,
+      "grad_norm": 1.6953125,
+      "learning_rate": 5.677673169838762e-06,
+      "loss": 0.8341,
+      "step": 11273
+    },
+    {
+      "epoch": 1.9211050860535397,
+      "grad_norm": 1.71875,
+      "learning_rate": 5.676044722724801e-06,
+      "loss": 0.843,
+      "step": 11274
+    },
+    {
+      "epoch": 1.9212765501425295,
+      "grad_norm": 1.7109375,
+      "learning_rate": 5.674416416631304e-06,
+      "loss": 0.8879,
+      "step": 11275
+    },
+    {
+      "epoch": 1.9214480142315193,
+      "grad_norm": 1.6328125,
+      "learning_rate": 5.6727882516113765e-06,
+      "loss": 0.823,
+      "step": 11276
+    },
+    {
+      "epoch": 1.9216194783205092,
+      "grad_norm": 1.7734375,
+      "learning_rate": 5.671160227718118e-06,
+      "loss": 0.9125,
+      "step": 11277
+    },
+    {
+      "epoch": 1.921790942409499,
+      "grad_norm": 1.7109375,
+      "learning_rate": 5.669532345004627e-06,
+      "loss": 0.875,
+      "step": 11278
+    },
+    {
+      "epoch": 1.9219624064984888,
+      "grad_norm": 1.71875,
+      "learning_rate": 5.667904603523992e-06,
+      "loss": 0.8969,
+      "step": 11279
+    },
+    {
+      "epoch": 1.9221338705874789,
+      "grad_norm": 1.6953125,
+      "learning_rate": 5.666277003329301e-06,
+      "loss": 0.9196,
+      "step": 11280
+    },
+    {
+      "epoch": 1.9223053346764687,
+      "grad_norm": 1.6640625,
+      "learning_rate": 5.664649544473636e-06,
+      "loss": 0.9005,
+      "step": 11281
+    },
+    {
+      "epoch": 1.9224767987654585,
+      "grad_norm": 1.71875,
+      "learning_rate": 5.6630222270100755e-06,
+      "loss": 0.8406,
+      "step": 11282
+    },
+    {
+      "epoch": 1.9226482628544486,
+      "grad_norm": 1.6171875,
+      "learning_rate": 5.661395050991691e-06,
+      "loss": 0.8144,
+      "step": 11283
+    },
+    {
+      "epoch": 1.9228197269434384,
+      "grad_norm": 1.625,
+      "learning_rate": 5.659768016471556e-06,
+      "loss": 0.7749,
+      "step": 11284
+    },
+    {
+      "epoch": 1.9229911910324282,
+      "grad_norm": 1.671875,
+      "learning_rate": 5.658141123502726e-06,
+      "loss": 0.8384,
+      "step": 11285
+    },
+    {
+      "epoch": 1.923162655121418,
+      "grad_norm": 1.6328125,
+      "learning_rate": 5.656514372138265e-06,
+      "loss": 0.8448,
+      "step": 11286
+    },
+    {
+      "epoch": 1.9233341192104079,
+      "grad_norm": 1.6328125,
+      "learning_rate": 5.654887762431224e-06,
+      "loss": 0.8417,
+      "step": 11287
+    },
+    {
+      "epoch": 1.9235055832993977,
+      "grad_norm": 1.78125,
+      "learning_rate": 5.653261294434655e-06,
+      "loss": 0.8934,
+      "step": 11288
+    },
+    {
+      "epoch": 1.9236770473883875,
+      "grad_norm": 1.65625,
+      "learning_rate": 5.651634968201606e-06,
+      "loss": 0.9038,
+      "step": 11289
+    },
+    {
+      "epoch": 1.9238485114773773,
+      "grad_norm": 1.6328125,
+      "learning_rate": 5.65000878378511e-06,
+      "loss": 0.8409,
+      "step": 11290
+    },
+    {
+      "epoch": 1.9240199755663672,
+      "grad_norm": 1.65625,
+      "learning_rate": 5.648382741238212e-06,
+      "loss": 0.8006,
+      "step": 11291
+    },
+    {
+      "epoch": 1.9241914396553572,
+      "grad_norm": 1.7109375,
+      "learning_rate": 5.64675684061394e-06,
+      "loss": 0.8432,
+      "step": 11292
+    },
+    {
+      "epoch": 1.924362903744347,
+      "grad_norm": 1.7109375,
+      "learning_rate": 5.645131081965319e-06,
+      "loss": 0.8412,
+      "step": 11293
+    },
+    {
+      "epoch": 1.9245343678333369,
+      "grad_norm": 1.734375,
+      "learning_rate": 5.6435054653453735e-06,
+      "loss": 0.8535,
+      "step": 11294
+    },
+    {
+      "epoch": 1.924705831922327,
+      "grad_norm": 1.7265625,
+      "learning_rate": 5.64187999080712e-06,
+      "loss": 0.8394,
+      "step": 11295
+    },
+    {
+      "epoch": 1.9248772960113167,
+      "grad_norm": 1.6796875,
+      "learning_rate": 5.6402546584035744e-06,
+      "loss": 0.8817,
+      "step": 11296
+    },
+    {
+      "epoch": 1.9250487601003066,
+      "grad_norm": 1.6640625,
+      "learning_rate": 5.6386294681877375e-06,
+      "loss": 0.8096,
+      "step": 11297
+    },
+    {
+      "epoch": 1.9252202241892964,
+      "grad_norm": 1.7421875,
+      "learning_rate": 5.637004420212617e-06,
+      "loss": 0.8154,
+      "step": 11298
+    },
+    {
+      "epoch": 1.9253916882782862,
+      "grad_norm": 1.5859375,
+      "learning_rate": 5.635379514531212e-06,
+      "loss": 0.7501,
+      "step": 11299
+    },
+    {
+      "epoch": 1.925563152367276,
+      "grad_norm": 1.6875,
+      "learning_rate": 5.633754751196516e-06,
+      "loss": 0.8901,
+      "step": 11300
+    },
+    {
+      "epoch": 1.9257346164562659,
+      "grad_norm": 1.6640625,
+      "learning_rate": 5.632130130261522e-06,
+      "loss": 0.8765,
+      "step": 11301
+    },
+    {
+      "epoch": 1.9259060805452557,
+      "grad_norm": 1.59375,
+      "learning_rate": 5.63050565177921e-06,
+      "loss": 0.8111,
+      "step": 11302
+    },
+    {
+      "epoch": 1.9260775446342455,
+      "grad_norm": 1.6953125,
+      "learning_rate": 5.628881315802563e-06,
+      "loss": 0.8233,
+      "step": 11303
+    },
+    {
+      "epoch": 1.9262490087232356,
+      "grad_norm": 1.71875,
+      "learning_rate": 5.627257122384558e-06,
+      "loss": 0.8096,
+      "step": 11304
+    },
+    {
+      "epoch": 1.9264204728122254,
+      "grad_norm": 1.71875,
+      "learning_rate": 5.625633071578163e-06,
+      "loss": 0.8586,
+      "step": 11305
+    },
+    {
+      "epoch": 1.9265919369012152,
+      "grad_norm": 1.734375,
+      "learning_rate": 5.624009163436345e-06,
+      "loss": 0.8614,
+      "step": 11306
+    },
+    {
+      "epoch": 1.9267634009902053,
+      "grad_norm": 1.6796875,
+      "learning_rate": 5.622385398012067e-06,
+      "loss": 0.8562,
+      "step": 11307
+    },
+    {
+      "epoch": 1.926934865079195,
+      "grad_norm": 1.6484375,
+      "learning_rate": 5.620761775358287e-06,
+      "loss": 0.7839,
+      "step": 11308
+    },
+    {
+      "epoch": 1.927106329168185,
+      "grad_norm": 1.625,
+      "learning_rate": 5.619138295527955e-06,
+      "loss": 0.8728,
+      "step": 11309
+    },
+    {
+      "epoch": 1.9272777932571747,
+      "grad_norm": 1.6796875,
+      "learning_rate": 5.617514958574021e-06,
+      "loss": 0.84,
+      "step": 11310
+    },
+    {
+      "epoch": 1.9274492573461646,
+      "grad_norm": 1.7109375,
+      "learning_rate": 5.615891764549426e-06,
+      "loss": 0.9072,
+      "step": 11311
+    },
+    {
+      "epoch": 1.9276207214351544,
+      "grad_norm": 1.75,
+      "learning_rate": 5.61426871350711e-06,
+      "loss": 0.8774,
+      "step": 11312
+    },
+    {
+      "epoch": 1.9277921855241442,
+      "grad_norm": 1.703125,
+      "learning_rate": 5.612645805500008e-06,
+      "loss": 0.8319,
+      "step": 11313
+    },
+    {
+      "epoch": 1.927963649613134,
+      "grad_norm": 1.7109375,
+      "learning_rate": 5.611023040581045e-06,
+      "loss": 0.8352,
+      "step": 11314
+    },
+    {
+      "epoch": 1.9281351137021239,
+      "grad_norm": 1.7109375,
+      "learning_rate": 5.60940041880315e-06,
+      "loss": 0.7795,
+      "step": 11315
+    },
+    {
+      "epoch": 1.928306577791114,
+      "grad_norm": 1.6328125,
+      "learning_rate": 5.607777940219239e-06,
+      "loss": 0.8453,
+      "step": 11316
+    },
+    {
+      "epoch": 1.9284780418801037,
+      "grad_norm": 1.765625,
+      "learning_rate": 5.606155604882231e-06,
+      "loss": 0.9062,
+      "step": 11317
+    },
+    {
+      "epoch": 1.9286495059690936,
+      "grad_norm": 1.5859375,
+      "learning_rate": 5.604533412845032e-06,
+      "loss": 0.8051,
+      "step": 11318
+    },
+    {
+      "epoch": 1.9288209700580836,
+      "grad_norm": 1.71875,
+      "learning_rate": 5.602911364160557e-06,
+      "loss": 0.848,
+      "step": 11319
+    },
+    {
+      "epoch": 1.9289924341470734,
+      "grad_norm": 1.71875,
+      "learning_rate": 5.601289458881693e-06,
+      "loss": 0.888,
+      "step": 11320
+    },
+    {
+      "epoch": 1.9291638982360633,
+      "grad_norm": 1.71875,
+      "learning_rate": 5.599667697061346e-06,
+      "loss": 0.7835,
+      "step": 11321
+    },
+    {
+      "epoch": 1.929335362325053,
+      "grad_norm": 1.875,
+      "learning_rate": 5.598046078752406e-06,
+      "loss": 0.8079,
+      "step": 11322
+    },
+    {
+      "epoch": 1.929506826414043,
+      "grad_norm": 1.7421875,
+      "learning_rate": 5.596424604007758e-06,
+      "loss": 0.7706,
+      "step": 11323
+    },
+    {
+      "epoch": 1.9296782905030327,
+      "grad_norm": 1.6015625,
+      "learning_rate": 5.594803272880287e-06,
+      "loss": 0.8445,
+      "step": 11324
+    },
+    {
+      "epoch": 1.9298497545920226,
+      "grad_norm": 1.6640625,
+      "learning_rate": 5.59318208542287e-06,
+      "loss": 0.7663,
+      "step": 11325
+    },
+    {
+      "epoch": 1.9300212186810124,
+      "grad_norm": 1.734375,
+      "learning_rate": 5.591561041688378e-06,
+      "loss": 0.8155,
+      "step": 11326
+    },
+    {
+      "epoch": 1.9301926827700022,
+      "grad_norm": 1.625,
+      "learning_rate": 5.589940141729683e-06,
+      "loss": 0.7547,
+      "step": 11327
+    },
+    {
+      "epoch": 1.9303641468589923,
+      "grad_norm": 1.7578125,
+      "learning_rate": 5.588319385599645e-06,
+      "loss": 0.8048,
+      "step": 11328
+    },
+    {
+      "epoch": 1.930535610947982,
+      "grad_norm": 1.7890625,
+      "learning_rate": 5.586698773351122e-06,
+      "loss": 0.8576,
+      "step": 11329
+    },
+    {
+      "epoch": 1.930707075036972,
+      "grad_norm": 1.7109375,
+      "learning_rate": 5.585078305036975e-06,
+      "loss": 0.9041,
+      "step": 11330
+    },
+    {
+      "epoch": 1.930878539125962,
+      "grad_norm": 1.765625,
+      "learning_rate": 5.583457980710053e-06,
+      "loss": 0.8994,
+      "step": 11331
+    },
+    {
+      "epoch": 1.9310500032149518,
+      "grad_norm": 1.6953125,
+      "learning_rate": 5.581837800423193e-06,
+      "loss": 0.8983,
+      "step": 11332
+    },
+    {
+      "epoch": 1.9312214673039416,
+      "grad_norm": 1.71875,
+      "learning_rate": 5.580217764229241e-06,
+      "loss": 0.8648,
+      "step": 11333
+    },
+    {
+      "epoch": 1.9313929313929314,
+      "grad_norm": 1.6640625,
+      "learning_rate": 5.578597872181031e-06,
+      "loss": 0.7956,
+      "step": 11334
+    },
+    {
+      "epoch": 1.9315643954819213,
+      "grad_norm": 1.8515625,
+      "learning_rate": 5.5769781243313915e-06,
+      "loss": 0.9276,
+      "step": 11335
+    },
+    {
+      "epoch": 1.931735859570911,
+      "grad_norm": 1.6640625,
+      "learning_rate": 5.575358520733153e-06,
+      "loss": 0.7656,
+      "step": 11336
+    },
+    {
+      "epoch": 1.931907323659901,
+      "grad_norm": 1.6328125,
+      "learning_rate": 5.573739061439134e-06,
+      "loss": 0.826,
+      "step": 11337
+    },
+    {
+      "epoch": 1.9320787877488907,
+      "grad_norm": 1.640625,
+      "learning_rate": 5.572119746502152e-06,
+      "loss": 0.8464,
+      "step": 11338
+    },
+    {
+      "epoch": 1.9322502518378806,
+      "grad_norm": 1.6796875,
+      "learning_rate": 5.57050057597502e-06,
+      "loss": 0.8301,
+      "step": 11339
+    },
+    {
+      "epoch": 1.9324217159268706,
+      "grad_norm": 1.640625,
+      "learning_rate": 5.568881549910543e-06,
+      "loss": 0.9023,
+      "step": 11340
+    },
+    {
+      "epoch": 1.9325931800158604,
+      "grad_norm": 1.5390625,
+      "learning_rate": 5.567262668361525e-06,
+      "loss": 0.7431,
+      "step": 11341
+    },
+    {
+      "epoch": 1.9327646441048503,
+      "grad_norm": 1.609375,
+      "learning_rate": 5.5656439313807675e-06,
+      "loss": 0.884,
+      "step": 11342
+    },
+    {
+      "epoch": 1.9329361081938403,
+      "grad_norm": 1.765625,
+      "learning_rate": 5.564025339021055e-06,
+      "loss": 0.7854,
+      "step": 11343
+    },
+    {
+      "epoch": 1.9331075722828301,
+      "grad_norm": 1.734375,
+      "learning_rate": 5.562406891335181e-06,
+      "loss": 0.8516,
+      "step": 11344
+    },
+    {
+      "epoch": 1.93327903637182,
+      "grad_norm": 1.7734375,
+      "learning_rate": 5.560788588375925e-06,
+      "loss": 0.8559,
+      "step": 11345
+    },
+    {
+      "epoch": 1.9334505004608098,
+      "grad_norm": 1.59375,
+      "learning_rate": 5.559170430196074e-06,
+      "loss": 0.7799,
+      "step": 11346
+    },
+    {
+      "epoch": 1.9336219645497996,
+      "grad_norm": 1.6484375,
+      "learning_rate": 5.557552416848398e-06,
+      "loss": 0.7802,
+      "step": 11347
+    },
+    {
+      "epoch": 1.9337934286387894,
+      "grad_norm": 1.6953125,
+      "learning_rate": 5.555934548385665e-06,
+      "loss": 0.8102,
+      "step": 11348
+    },
+    {
+      "epoch": 1.9339648927277793,
+      "grad_norm": 1.7578125,
+      "learning_rate": 5.554316824860642e-06,
+      "loss": 0.8421,
+      "step": 11349
+    },
+    {
+      "epoch": 1.934136356816769,
+      "grad_norm": 1.6953125,
+      "learning_rate": 5.5526992463260884e-06,
+      "loss": 0.8531,
+      "step": 11350
+    },
+    {
+      "epoch": 1.934307820905759,
+      "grad_norm": 1.6171875,
+      "learning_rate": 5.55108181283476e-06,
+      "loss": 0.7784,
+      "step": 11351
+    },
+    {
+      "epoch": 1.9344792849947487,
+      "grad_norm": 1.515625,
+      "learning_rate": 5.549464524439406e-06,
+      "loss": 0.8066,
+      "step": 11352
+    },
+    {
+      "epoch": 1.9346507490837388,
+      "grad_norm": 1.6640625,
+      "learning_rate": 5.547847381192772e-06,
+      "loss": 0.8316,
+      "step": 11353
+    },
+    {
+      "epoch": 1.9348222131727286,
+      "grad_norm": 1.6484375,
+      "learning_rate": 5.546230383147606e-06,
+      "loss": 0.8106,
+      "step": 11354
+    },
+    {
+      "epoch": 1.9349936772617184,
+      "grad_norm": 1.6640625,
+      "learning_rate": 5.544613530356633e-06,
+      "loss": 0.8452,
+      "step": 11355
+    },
+    {
+      "epoch": 1.9351651413507085,
+      "grad_norm": 1.75,
+      "learning_rate": 5.54299682287259e-06,
+      "loss": 0.9008,
+      "step": 11356
+    },
+    {
+      "epoch": 1.9353366054396983,
+      "grad_norm": 1.6953125,
+      "learning_rate": 5.541380260748206e-06,
+      "loss": 0.8208,
+      "step": 11357
+    },
+    {
+      "epoch": 1.9355080695286881,
+      "grad_norm": 1.6875,
+      "learning_rate": 5.539763844036201e-06,
+      "loss": 0.8263,
+      "step": 11358
+    },
+    {
+      "epoch": 1.935679533617678,
+      "grad_norm": 1.71875,
+      "learning_rate": 5.538147572789292e-06,
+      "loss": 0.8089,
+      "step": 11359
+    },
+    {
+      "epoch": 1.9358509977066678,
+      "grad_norm": 1.6796875,
+      "learning_rate": 5.5365314470601915e-06,
+      "loss": 0.8441,
+      "step": 11360
+    },
+    {
+      "epoch": 1.9360224617956576,
+      "grad_norm": 1.765625,
+      "learning_rate": 5.534915466901609e-06,
+      "loss": 0.7927,
+      "step": 11361
+    },
+    {
+      "epoch": 1.9361939258846474,
+      "grad_norm": 1.625,
+      "learning_rate": 5.533299632366248e-06,
+      "loss": 0.8619,
+      "step": 11362
+    },
+    {
+      "epoch": 1.9363653899736373,
+      "grad_norm": 1.671875,
+      "learning_rate": 5.531683943506806e-06,
+      "loss": 0.7727,
+      "step": 11363
+    },
+    {
+      "epoch": 1.936536854062627,
+      "grad_norm": 1.734375,
+      "learning_rate": 5.530068400375976e-06,
+      "loss": 0.7626,
+      "step": 11364
+    },
+    {
+      "epoch": 1.9367083181516171,
+      "grad_norm": 1.6484375,
+      "learning_rate": 5.528453003026448e-06,
+      "loss": 0.8035,
+      "step": 11365
+    },
+    {
+      "epoch": 1.936879782240607,
+      "grad_norm": 1.71875,
+      "learning_rate": 5.526837751510905e-06,
+      "loss": 0.8574,
+      "step": 11366
+    },
+    {
+      "epoch": 1.9370512463295968,
+      "grad_norm": 1.7734375,
+      "learning_rate": 5.525222645882029e-06,
+      "loss": 0.9693,
+      "step": 11367
+    },
+    {
+      "epoch": 1.9372227104185868,
+      "grad_norm": 1.703125,
+      "learning_rate": 5.523607686192492e-06,
+      "loss": 0.8168,
+      "step": 11368
+    },
+    {
+      "epoch": 1.9373941745075767,
+      "grad_norm": 1.6953125,
+      "learning_rate": 5.521992872494965e-06,
+      "loss": 0.8063,
+      "step": 11369
+    },
+    {
+      "epoch": 1.9375656385965665,
+      "grad_norm": 1.828125,
+      "learning_rate": 5.5203782048421115e-06,
+      "loss": 0.8824,
+      "step": 11370
+    },
+    {
+      "epoch": 1.9377371026855563,
+      "grad_norm": 1.546875,
+      "learning_rate": 5.518763683286596e-06,
+      "loss": 0.7815,
+      "step": 11371
+    },
+    {
+      "epoch": 1.9379085667745461,
+      "grad_norm": 1.6796875,
+      "learning_rate": 5.5171493078810704e-06,
+      "loss": 0.8088,
+      "step": 11372
+    },
+    {
+      "epoch": 1.938080030863536,
+      "grad_norm": 1.5859375,
+      "learning_rate": 5.515535078678187e-06,
+      "loss": 0.8024,
+      "step": 11373
+    },
+    {
+      "epoch": 1.9382514949525258,
+      "grad_norm": 1.6484375,
+      "learning_rate": 5.51392099573059e-06,
+      "loss": 0.7431,
+      "step": 11374
+    },
+    {
+      "epoch": 1.9384229590415156,
+      "grad_norm": 1.8671875,
+      "learning_rate": 5.512307059090925e-06,
+      "loss": 0.7968,
+      "step": 11375
+    },
+    {
+      "epoch": 1.9385944231305055,
+      "grad_norm": 1.671875,
+      "learning_rate": 5.5106932688118245e-06,
+      "loss": 0.7958,
+      "step": 11376
+    },
+    {
+      "epoch": 1.9387658872194955,
+      "grad_norm": 2.640625,
+      "learning_rate": 5.509079624945926e-06,
+      "loss": 0.8416,
+      "step": 11377
+    },
+    {
+      "epoch": 1.9389373513084853,
+      "grad_norm": 1.6328125,
+      "learning_rate": 5.507466127545847e-06,
+      "loss": 0.8436,
+      "step": 11378
+    },
+    {
+      "epoch": 1.9391088153974751,
+      "grad_norm": 1.84375,
+      "learning_rate": 5.505852776664215e-06,
+      "loss": 0.8292,
+      "step": 11379
+    },
+    {
+      "epoch": 1.9392802794864652,
+      "grad_norm": 1.6875,
+      "learning_rate": 5.504239572353649e-06,
+      "loss": 0.8329,
+      "step": 11380
+    },
+    {
+      "epoch": 1.939451743575455,
+      "grad_norm": 1.640625,
+      "learning_rate": 5.5026265146667605e-06,
+      "loss": 0.7997,
+      "step": 11381
+    },
+    {
+      "epoch": 1.9396232076644448,
+      "grad_norm": 1.6328125,
+      "learning_rate": 5.501013603656155e-06,
+      "loss": 0.7976,
+      "step": 11382
+    },
+    {
+      "epoch": 1.9397946717534347,
+      "grad_norm": 1.6796875,
+      "learning_rate": 5.499400839374438e-06,
+      "loss": 0.8052,
+      "step": 11383
+    },
+    {
+      "epoch": 1.9399661358424245,
+      "grad_norm": 1.7578125,
+      "learning_rate": 5.497788221874203e-06,
+      "loss": 0.9537,
+      "step": 11384
+    },
+    {
+      "epoch": 1.9401375999314143,
+      "grad_norm": 1.6484375,
+      "learning_rate": 5.496175751208052e-06,
+      "loss": 0.8053,
+      "step": 11385
+    },
+    {
+      "epoch": 1.9403090640204042,
+      "grad_norm": 1.6171875,
+      "learning_rate": 5.494563427428569e-06,
+      "loss": 0.8405,
+      "step": 11386
+    },
+    {
+      "epoch": 1.940480528109394,
+      "grad_norm": 1.671875,
+      "learning_rate": 5.492951250588341e-06,
+      "loss": 0.8469,
+      "step": 11387
+    },
+    {
+      "epoch": 1.9406519921983838,
+      "grad_norm": 1.625,
+      "learning_rate": 5.491339220739947e-06,
+      "loss": 0.797,
+      "step": 11388
+    },
+    {
+      "epoch": 1.9408234562873738,
+      "grad_norm": 1.6953125,
+      "learning_rate": 5.489727337935955e-06,
+      "loss": 0.8585,
+      "step": 11389
+    },
+    {
+      "epoch": 1.9409949203763637,
+      "grad_norm": 1.7890625,
+      "learning_rate": 5.488115602228938e-06,
+      "loss": 0.9122,
+      "step": 11390
+    },
+    {
+      "epoch": 1.9411663844653535,
+      "grad_norm": 1.625,
+      "learning_rate": 5.486504013671463e-06,
+      "loss": 0.792,
+      "step": 11391
+    },
+    {
+      "epoch": 1.9413378485543435,
+      "grad_norm": 1.7265625,
+      "learning_rate": 5.484892572316088e-06,
+      "loss": 0.8458,
+      "step": 11392
+    },
+    {
+      "epoch": 1.9415093126433334,
+      "grad_norm": 1.8203125,
+      "learning_rate": 5.483281278215366e-06,
+      "loss": 0.9417,
+      "step": 11393
+    },
+    {
+      "epoch": 1.9416807767323232,
+      "grad_norm": 1.6484375,
+      "learning_rate": 5.481670131421853e-06,
+      "loss": 0.8389,
+      "step": 11394
+    },
+    {
+      "epoch": 1.941852240821313,
+      "grad_norm": 1.640625,
+      "learning_rate": 5.4800591319880905e-06,
+      "loss": 0.8018,
+      "step": 11395
+    },
+    {
+      "epoch": 1.9420237049103029,
+      "grad_norm": 1.671875,
+      "learning_rate": 5.47844827996662e-06,
+      "loss": 0.861,
+      "step": 11396
+    },
+    {
+      "epoch": 1.9421951689992927,
+      "grad_norm": 1.671875,
+      "learning_rate": 5.476837575409978e-06,
+      "loss": 0.8843,
+      "step": 11397
+    },
+    {
+      "epoch": 1.9423666330882825,
+      "grad_norm": 1.7265625,
+      "learning_rate": 5.475227018370695e-06,
+      "loss": 0.8738,
+      "step": 11398
+    },
+    {
+      "epoch": 1.9425380971772723,
+      "grad_norm": 1.59375,
+      "learning_rate": 5.473616608901298e-06,
+      "loss": 0.8311,
+      "step": 11399
+    },
+    {
+      "epoch": 1.9427095612662622,
+      "grad_norm": 1.71875,
+      "learning_rate": 5.4720063470543115e-06,
+      "loss": 0.8529,
+      "step": 11400
+    },
+    {
+      "epoch": 1.9428810253552522,
+      "grad_norm": 1.71875,
+      "learning_rate": 5.470396232882242e-06,
+      "loss": 0.8726,
+      "step": 11401
+    },
+    {
+      "epoch": 1.943052489444242,
+      "grad_norm": 1.6953125,
+      "learning_rate": 5.468786266437611e-06,
+      "loss": 0.8968,
+      "step": 11402
+    },
+    {
+      "epoch": 1.9432239535332319,
+      "grad_norm": 1.640625,
+      "learning_rate": 5.467176447772924e-06,
+      "loss": 0.864,
+      "step": 11403
+    },
+    {
+      "epoch": 1.943395417622222,
+      "grad_norm": 1.625,
+      "learning_rate": 5.4655667769406805e-06,
+      "loss": 0.9435,
+      "step": 11404
+    },
+    {
+      "epoch": 1.9435668817112117,
+      "grad_norm": 1.6015625,
+      "learning_rate": 5.46395725399338e-06,
+      "loss": 0.7874,
+      "step": 11405
+    },
+    {
+      "epoch": 1.9437383458002016,
+      "grad_norm": 1.6640625,
+      "learning_rate": 5.462347878983516e-06,
+      "loss": 0.8414,
+      "step": 11406
+    },
+    {
+      "epoch": 1.9439098098891914,
+      "grad_norm": 1.6328125,
+      "learning_rate": 5.460738651963573e-06,
+      "loss": 0.7489,
+      "step": 11407
+    },
+    {
+      "epoch": 1.9440812739781812,
+      "grad_norm": 1.6171875,
+      "learning_rate": 5.459129572986037e-06,
+      "loss": 0.7888,
+      "step": 11408
+    },
+    {
+      "epoch": 1.944252738067171,
+      "grad_norm": 1.765625,
+      "learning_rate": 5.457520642103385e-06,
+      "loss": 0.8051,
+      "step": 11409
+    },
+    {
+      "epoch": 1.9444242021561609,
+      "grad_norm": 1.8046875,
+      "learning_rate": 5.4559118593680905e-06,
+      "loss": 0.9306,
+      "step": 11410
+    },
+    {
+      "epoch": 1.9445956662451507,
+      "grad_norm": 1.5703125,
+      "learning_rate": 5.454303224832624e-06,
+      "loss": 0.7945,
+      "step": 11411
+    },
+    {
+      "epoch": 1.9447671303341405,
+      "grad_norm": 1.8125,
+      "learning_rate": 5.4526947385494445e-06,
+      "loss": 0.8615,
+      "step": 11412
+    },
+    {
+      "epoch": 1.9449385944231306,
+      "grad_norm": 1.75,
+      "learning_rate": 5.451086400571013e-06,
+      "loss": 0.8689,
+      "step": 11413
+    },
+    {
+      "epoch": 1.9451100585121204,
+      "grad_norm": 1.6953125,
+      "learning_rate": 5.449478210949784e-06,
+      "loss": 0.8464,
+      "step": 11414
+    },
+    {
+      "epoch": 1.9452815226011102,
+      "grad_norm": 1.734375,
+      "learning_rate": 5.447870169738205e-06,
+      "loss": 0.8437,
+      "step": 11415
+    },
+    {
+      "epoch": 1.9454529866901002,
+      "grad_norm": 1.6484375,
+      "learning_rate": 5.446262276988722e-06,
+      "loss": 0.8332,
+      "step": 11416
+    },
+    {
+      "epoch": 1.94562445077909,
+      "grad_norm": 1.6171875,
+      "learning_rate": 5.444654532753775e-06,
+      "loss": 0.8617,
+      "step": 11417
+    },
+    {
+      "epoch": 1.94579591486808,
+      "grad_norm": 1.625,
+      "learning_rate": 5.443046937085795e-06,
+      "loss": 0.8854,
+      "step": 11418
+    },
+    {
+      "epoch": 1.9459673789570697,
+      "grad_norm": 1.6328125,
+      "learning_rate": 5.441439490037217e-06,
+      "loss": 0.8435,
+      "step": 11419
+    },
+    {
+      "epoch": 1.9461388430460596,
+      "grad_norm": 1.7734375,
+      "learning_rate": 5.439832191660462e-06,
+      "loss": 0.8938,
+      "step": 11420
+    },
+    {
+      "epoch": 1.9463103071350494,
+      "grad_norm": 1.7890625,
+      "learning_rate": 5.438225042007949e-06,
+      "loss": 0.9738,
+      "step": 11421
+    },
+    {
+      "epoch": 1.9464817712240392,
+      "grad_norm": 1.734375,
+      "learning_rate": 5.436618041132099e-06,
+      "loss": 0.8825,
+      "step": 11422
+    },
+    {
+      "epoch": 1.946653235313029,
+      "grad_norm": 1.703125,
+      "learning_rate": 5.435011189085318e-06,
+      "loss": 0.9542,
+      "step": 11423
+    },
+    {
+      "epoch": 1.9468246994020189,
+      "grad_norm": 1.6953125,
+      "learning_rate": 5.433404485920011e-06,
+      "loss": 0.8039,
+      "step": 11424
+    },
+    {
+      "epoch": 1.946996163491009,
+      "grad_norm": 1.7421875,
+      "learning_rate": 5.431797931688581e-06,
+      "loss": 0.7727,
+      "step": 11425
+    },
+    {
+      "epoch": 1.9471676275799987,
+      "grad_norm": 1.7265625,
+      "learning_rate": 5.430191526443423e-06,
+      "loss": 0.8357,
+      "step": 11426
+    },
+    {
+      "epoch": 1.9473390916689886,
+      "grad_norm": 1.6796875,
+      "learning_rate": 5.428585270236925e-06,
+      "loss": 0.8357,
+      "step": 11427
+    },
+    {
+      "epoch": 1.9475105557579786,
+      "grad_norm": 1.7890625,
+      "learning_rate": 5.426979163121479e-06,
+      "loss": 0.8168,
+      "step": 11428
+    },
+    {
+      "epoch": 1.9476820198469684,
+      "grad_norm": 1.6953125,
+      "learning_rate": 5.425373205149461e-06,
+      "loss": 0.8457,
+      "step": 11429
+    },
+    {
+      "epoch": 1.9478534839359583,
+      "grad_norm": 1.734375,
+      "learning_rate": 5.42376739637325e-06,
+      "loss": 0.8166,
+      "step": 11430
+    },
+    {
+      "epoch": 1.948024948024948,
+      "grad_norm": 1.6640625,
+      "learning_rate": 5.4221617368452165e-06,
+      "loss": 0.8057,
+      "step": 11431
+    },
+    {
+      "epoch": 1.948196412113938,
+      "grad_norm": 1.7421875,
+      "learning_rate": 5.420556226617727e-06,
+      "loss": 0.7438,
+      "step": 11432
+    },
+    {
+      "epoch": 1.9483678762029277,
+      "grad_norm": 1.6015625,
+      "learning_rate": 5.4189508657431444e-06,
+      "loss": 0.7822,
+      "step": 11433
+    },
+    {
+      "epoch": 1.9485393402919176,
+      "grad_norm": 1.75,
+      "learning_rate": 5.417345654273827e-06,
+      "loss": 0.8917,
+      "step": 11434
+    },
+    {
+      "epoch": 1.9487108043809074,
+      "grad_norm": 1.671875,
+      "learning_rate": 5.415740592262122e-06,
+      "loss": 0.7634,
+      "step": 11435
+    },
+    {
+      "epoch": 1.9488822684698972,
+      "grad_norm": 1.65625,
+      "learning_rate": 5.41413567976038e-06,
+      "loss": 0.8146,
+      "step": 11436
+    },
+    {
+      "epoch": 1.9490537325588873,
+      "grad_norm": 1.7421875,
+      "learning_rate": 5.4125309168209394e-06,
+      "loss": 0.9364,
+      "step": 11437
+    },
+    {
+      "epoch": 1.949225196647877,
+      "grad_norm": 1.625,
+      "learning_rate": 5.410926303496142e-06,
+      "loss": 0.8589,
+      "step": 11438
+    },
+    {
+      "epoch": 1.949396660736867,
+      "grad_norm": 1.671875,
+      "learning_rate": 5.4093218398383154e-06,
+      "loss": 0.8342,
+      "step": 11439
+    },
+    {
+      "epoch": 1.9495681248258567,
+      "grad_norm": 1.8046875,
+      "learning_rate": 5.407717525899793e-06,
+      "loss": 0.8523,
+      "step": 11440
+    },
+    {
+      "epoch": 1.9497395889148468,
+      "grad_norm": 1.6171875,
+      "learning_rate": 5.4061133617328956e-06,
+      "loss": 0.8945,
+      "step": 11441
+    },
+    {
+      "epoch": 1.9499110530038366,
+      "grad_norm": 1.6953125,
+      "learning_rate": 5.40450934738994e-06,
+      "loss": 0.7959,
+      "step": 11442
+    },
+    {
+      "epoch": 1.9500825170928264,
+      "grad_norm": 1.65625,
+      "learning_rate": 5.402905482923239e-06,
+      "loss": 0.9032,
+      "step": 11443
+    },
+    {
+      "epoch": 1.9502539811818163,
+      "grad_norm": 1.6875,
+      "learning_rate": 5.4013017683851034e-06,
+      "loss": 0.8463,
+      "step": 11444
+    },
+    {
+      "epoch": 1.950425445270806,
+      "grad_norm": 1.7734375,
+      "learning_rate": 5.399698203827834e-06,
+      "loss": 0.8543,
+      "step": 11445
+    },
+    {
+      "epoch": 1.950596909359796,
+      "grad_norm": 1.7265625,
+      "learning_rate": 5.398094789303731e-06,
+      "loss": 0.9256,
+      "step": 11446
+    },
+    {
+      "epoch": 1.9507683734487857,
+      "grad_norm": 1.6640625,
+      "learning_rate": 5.396491524865084e-06,
+      "loss": 0.7521,
+      "step": 11447
+    },
+    {
+      "epoch": 1.9509398375377756,
+      "grad_norm": 1.75,
+      "learning_rate": 5.394888410564184e-06,
+      "loss": 0.8547,
+      "step": 11448
+    },
+    {
+      "epoch": 1.9511113016267654,
+      "grad_norm": 1.609375,
+      "learning_rate": 5.393285446453314e-06,
+      "loss": 0.801,
+      "step": 11449
+    },
+    {
+      "epoch": 1.9512827657157554,
+      "grad_norm": 1.7109375,
+      "learning_rate": 5.391682632584752e-06,
+      "loss": 0.8614,
+      "step": 11450
+    },
+    {
+      "epoch": 1.9514542298047453,
+      "grad_norm": 1.6875,
+      "learning_rate": 5.390079969010773e-06,
+      "loss": 0.8522,
+      "step": 11451
+    },
+    {
+      "epoch": 1.951625693893735,
+      "grad_norm": 1.671875,
+      "learning_rate": 5.388477455783645e-06,
+      "loss": 0.8962,
+      "step": 11452
+    },
+    {
+      "epoch": 1.9517971579827251,
+      "grad_norm": 1.75,
+      "learning_rate": 5.386875092955633e-06,
+      "loss": 0.8499,
+      "step": 11453
+    },
+    {
+      "epoch": 1.951968622071715,
+      "grad_norm": 1.734375,
+      "learning_rate": 5.385272880578996e-06,
+      "loss": 0.9148,
+      "step": 11454
+    },
+    {
+      "epoch": 1.9521400861607048,
+      "grad_norm": 1.8828125,
+      "learning_rate": 5.383670818705987e-06,
+      "loss": 0.792,
+      "step": 11455
+    },
+    {
+      "epoch": 1.9523115502496946,
+      "grad_norm": 1.6875,
+      "learning_rate": 5.382068907388856e-06,
+      "loss": 0.8232,
+      "step": 11456
+    },
+    {
+      "epoch": 1.9524830143386844,
+      "grad_norm": 1.6171875,
+      "learning_rate": 5.3804671466798486e-06,
+      "loss": 0.8675,
+      "step": 11457
+    },
+    {
+      "epoch": 1.9526544784276743,
+      "grad_norm": 1.71875,
+      "learning_rate": 5.378865536631203e-06,
+      "loss": 0.8182,
+      "step": 11458
+    },
+    {
+      "epoch": 1.952825942516664,
+      "grad_norm": 1.6796875,
+      "learning_rate": 5.377264077295152e-06,
+      "loss": 0.8782,
+      "step": 11459
+    },
+    {
+      "epoch": 1.952997406605654,
+      "grad_norm": 1.734375,
+      "learning_rate": 5.375662768723929e-06,
+      "loss": 0.7995,
+      "step": 11460
+    },
+    {
+      "epoch": 1.9531688706946437,
+      "grad_norm": 1.625,
+      "learning_rate": 5.374061610969756e-06,
+      "loss": 0.8436,
+      "step": 11461
+    },
+    {
+      "epoch": 1.9533403347836338,
+      "grad_norm": 1.71875,
+      "learning_rate": 5.372460604084854e-06,
+      "loss": 0.8517,
+      "step": 11462
+    },
+    {
+      "epoch": 1.9535117988726236,
+      "grad_norm": 1.6484375,
+      "learning_rate": 5.370859748121437e-06,
+      "loss": 0.8209,
+      "step": 11463
+    },
+    {
+      "epoch": 1.9536832629616134,
+      "grad_norm": 1.8046875,
+      "learning_rate": 5.369259043131716e-06,
+      "loss": 0.8313,
+      "step": 11464
+    },
+    {
+      "epoch": 1.9538547270506035,
+      "grad_norm": 1.6640625,
+      "learning_rate": 5.367658489167894e-06,
+      "loss": 0.8333,
+      "step": 11465
+    },
+    {
+      "epoch": 1.9540261911395933,
+      "grad_norm": 1.6640625,
+      "learning_rate": 5.366058086282173e-06,
+      "loss": 0.8558,
+      "step": 11466
+    },
+    {
+      "epoch": 1.9541976552285831,
+      "grad_norm": 1.8046875,
+      "learning_rate": 5.364457834526749e-06,
+      "loss": 0.8054,
+      "step": 11467
+    },
+    {
+      "epoch": 1.954369119317573,
+      "grad_norm": 1.6875,
+      "learning_rate": 5.362857733953809e-06,
+      "loss": 0.8077,
+      "step": 11468
+    },
+    {
+      "epoch": 1.9545405834065628,
+      "grad_norm": 1.6015625,
+      "learning_rate": 5.361257784615546e-06,
+      "loss": 0.8044,
+      "step": 11469
+    },
+    {
+      "epoch": 1.9547120474955526,
+      "grad_norm": 1.7578125,
+      "learning_rate": 5.3596579865641286e-06,
+      "loss": 0.9041,
+      "step": 11470
+    },
+    {
+      "epoch": 1.9548835115845424,
+      "grad_norm": 1.6484375,
+      "learning_rate": 5.358058339851739e-06,
+      "loss": 0.8412,
+      "step": 11471
+    },
+    {
+      "epoch": 1.9550549756735323,
+      "grad_norm": 1.6640625,
+      "learning_rate": 5.356458844530546e-06,
+      "loss": 0.8495,
+      "step": 11472
+    },
+    {
+      "epoch": 1.955226439762522,
+      "grad_norm": 1.6328125,
+      "learning_rate": 5.354859500652717e-06,
+      "loss": 0.8629,
+      "step": 11473
+    },
+    {
+      "epoch": 1.9553979038515121,
+      "grad_norm": 1.7421875,
+      "learning_rate": 5.3532603082704115e-06,
+      "loss": 0.8612,
+      "step": 11474
+    },
+    {
+      "epoch": 1.955569367940502,
+      "grad_norm": 1.8125,
+      "learning_rate": 5.351661267435786e-06,
+      "loss": 0.8293,
+      "step": 11475
+    },
+    {
+      "epoch": 1.9557408320294918,
+      "grad_norm": 1.7421875,
+      "learning_rate": 5.350062378200989e-06,
+      "loss": 0.8635,
+      "step": 11476
+    },
+    {
+      "epoch": 1.9559122961184818,
+      "grad_norm": 1.7109375,
+      "learning_rate": 5.348463640618169e-06,
+      "loss": 0.8071,
+      "step": 11477
+    },
+    {
+      "epoch": 1.9560837602074717,
+      "grad_norm": 1.6640625,
+      "learning_rate": 5.346865054739463e-06,
+      "loss": 0.8663,
+      "step": 11478
+    },
+    {
+      "epoch": 1.9562552242964615,
+      "grad_norm": 1.7890625,
+      "learning_rate": 5.345266620617011e-06,
+      "loss": 0.9402,
+      "step": 11479
+    },
+    {
+      "epoch": 1.9564266883854513,
+      "grad_norm": 1.7890625,
+      "learning_rate": 5.343668338302944e-06,
+      "loss": 0.8633,
+      "step": 11480
+    },
+    {
+      "epoch": 1.9565981524744411,
+      "grad_norm": 1.734375,
+      "learning_rate": 5.34207020784939e-06,
+      "loss": 0.8151,
+      "step": 11481
+    },
+    {
+      "epoch": 1.956769616563431,
+      "grad_norm": 1.703125,
+      "learning_rate": 5.340472229308465e-06,
+      "loss": 0.8865,
+      "step": 11482
+    },
+    {
+      "epoch": 1.9569410806524208,
+      "grad_norm": 1.671875,
+      "learning_rate": 5.338874402732284e-06,
+      "loss": 0.7956,
+      "step": 11483
+    },
+    {
+      "epoch": 1.9571125447414106,
+      "grad_norm": 1.59375,
+      "learning_rate": 5.337276728172962e-06,
+      "loss": 0.7595,
+      "step": 11484
+    },
+    {
+      "epoch": 1.9572840088304004,
+      "grad_norm": 1.671875,
+      "learning_rate": 5.335679205682604e-06,
+      "loss": 0.8831,
+      "step": 11485
+    },
+    {
+      "epoch": 1.9574554729193905,
+      "grad_norm": 1.71875,
+      "learning_rate": 5.3340818353133105e-06,
+      "loss": 0.8479,
+      "step": 11486
+    },
+    {
+      "epoch": 1.9576269370083803,
+      "grad_norm": 1.59375,
+      "learning_rate": 5.33248461711718e-06,
+      "loss": 0.8728,
+      "step": 11487
+    },
+    {
+      "epoch": 1.9577984010973701,
+      "grad_norm": 1.703125,
+      "learning_rate": 5.330887551146301e-06,
+      "loss": 0.952,
+      "step": 11488
+    },
+    {
+      "epoch": 1.9579698651863602,
+      "grad_norm": 1.6875,
+      "learning_rate": 5.329290637452762e-06,
+      "loss": 0.8699,
+      "step": 11489
+    },
+    {
+      "epoch": 1.95814132927535,
+      "grad_norm": 1.6484375,
+      "learning_rate": 5.327693876088643e-06,
+      "loss": 0.8186,
+      "step": 11490
+    },
+    {
+      "epoch": 1.9583127933643398,
+      "grad_norm": 1.8046875,
+      "learning_rate": 5.3260972671060215e-06,
+      "loss": 0.8201,
+      "step": 11491
+    },
+    {
+      "epoch": 1.9584842574533297,
+      "grad_norm": 1.734375,
+      "learning_rate": 5.324500810556971e-06,
+      "loss": 0.8688,
+      "step": 11492
+    },
+    {
+      "epoch": 1.9586557215423195,
+      "grad_norm": 1.65625,
+      "learning_rate": 5.322904506493554e-06,
+      "loss": 0.8751,
+      "step": 11493
+    },
+    {
+      "epoch": 1.9588271856313093,
+      "grad_norm": 1.7890625,
+      "learning_rate": 5.321308354967828e-06,
+      "loss": 0.8815,
+      "step": 11494
+    },
+    {
+      "epoch": 1.9589986497202991,
+      "grad_norm": 1.71875,
+      "learning_rate": 5.319712356031859e-06,
+      "loss": 0.8465,
+      "step": 11495
+    },
+    {
+      "epoch": 1.959170113809289,
+      "grad_norm": 1.703125,
+      "learning_rate": 5.318116509737694e-06,
+      "loss": 0.807,
+      "step": 11496
+    },
+    {
+      "epoch": 1.9593415778982788,
+      "grad_norm": 1.8046875,
+      "learning_rate": 5.31652081613738e-06,
+      "loss": 0.8358,
+      "step": 11497
+    },
+    {
+      "epoch": 1.9595130419872688,
+      "grad_norm": 1.6875,
+      "learning_rate": 5.314925275282959e-06,
+      "loss": 0.8653,
+      "step": 11498
+    },
+    {
+      "epoch": 1.9596845060762587,
+      "grad_norm": 1.6796875,
+      "learning_rate": 5.313329887226467e-06,
+      "loss": 0.8799,
+      "step": 11499
+    },
+    {
+      "epoch": 1.9598559701652485,
+      "grad_norm": 1.796875,
+      "learning_rate": 5.311734652019935e-06,
+      "loss": 0.8724,
+      "step": 11500
+    },
+    {
+      "epoch": 1.9600274342542385,
+      "grad_norm": 1.78125,
+      "learning_rate": 5.310139569715391e-06,
+      "loss": 0.8704,
+      "step": 11501
+    },
+    {
+      "epoch": 1.9601988983432284,
+      "grad_norm": 1.6171875,
+      "learning_rate": 5.308544640364856e-06,
+      "loss": 0.7946,
+      "step": 11502
+    },
+    {
+      "epoch": 1.9603703624322182,
+      "grad_norm": 1.6640625,
+      "learning_rate": 5.306949864020348e-06,
+      "loss": 0.7768,
+      "step": 11503
+    },
+    {
+      "epoch": 1.960541826521208,
+      "grad_norm": 1.828125,
+      "learning_rate": 5.305355240733881e-06,
+      "loss": 0.7978,
+      "step": 11504
+    },
+    {
+      "epoch": 1.9607132906101978,
+      "grad_norm": 1.609375,
+      "learning_rate": 5.303760770557452e-06,
+      "loss": 0.8094,
+      "step": 11505
+    },
+    {
+      "epoch": 1.9608847546991877,
+      "grad_norm": 1.6484375,
+      "learning_rate": 5.302166453543069e-06,
+      "loss": 0.8126,
+      "step": 11506
+    },
+    {
+      "epoch": 1.9610562187881775,
+      "grad_norm": 1.6484375,
+      "learning_rate": 5.30057228974273e-06,
+      "loss": 0.8152,
+      "step": 11507
+    },
+    {
+      "epoch": 1.9612276828771673,
+      "grad_norm": 1.6640625,
+      "learning_rate": 5.2989782792084236e-06,
+      "loss": 0.8209,
+      "step": 11508
+    },
+    {
+      "epoch": 1.9613991469661571,
+      "grad_norm": 1.6796875,
+      "learning_rate": 5.297384421992138e-06,
+      "loss": 0.8824,
+      "step": 11509
+    },
+    {
+      "epoch": 1.9615706110551472,
+      "grad_norm": 1.6875,
+      "learning_rate": 5.2957907181458554e-06,
+      "loss": 0.8342,
+      "step": 11510
+    },
+    {
+      "epoch": 1.961742075144137,
+      "grad_norm": 1.6640625,
+      "learning_rate": 5.294197167721552e-06,
+      "loss": 0.8246,
+      "step": 11511
+    },
+    {
+      "epoch": 1.9619135392331268,
+      "grad_norm": 1.6875,
+      "learning_rate": 5.292603770771199e-06,
+      "loss": 0.8105,
+      "step": 11512
+    },
+    {
+      "epoch": 1.9620850033221169,
+      "grad_norm": 1.546875,
+      "learning_rate": 5.291010527346763e-06,
+      "loss": 0.747,
+      "step": 11513
+    },
+    {
+      "epoch": 1.9622564674111067,
+      "grad_norm": 1.734375,
+      "learning_rate": 5.289417437500207e-06,
+      "loss": 0.8496,
+      "step": 11514
+    },
+    {
+      "epoch": 1.9624279315000965,
+      "grad_norm": 1.765625,
+      "learning_rate": 5.287824501283487e-06,
+      "loss": 0.8033,
+      "step": 11515
+    },
+    {
+      "epoch": 1.9625993955890864,
+      "grad_norm": 1.6640625,
+      "learning_rate": 5.286231718748553e-06,
+      "loss": 0.7926,
+      "step": 11516
+    },
+    {
+      "epoch": 1.9627708596780762,
+      "grad_norm": 1.6875,
+      "learning_rate": 5.284639089947355e-06,
+      "loss": 0.8827,
+      "step": 11517
+    },
+    {
+      "epoch": 1.962942323767066,
+      "grad_norm": 1.6796875,
+      "learning_rate": 5.28304661493183e-06,
+      "loss": 0.7797,
+      "step": 11518
+    },
+    {
+      "epoch": 1.9631137878560558,
+      "grad_norm": 1.6484375,
+      "learning_rate": 5.281454293753921e-06,
+      "loss": 0.827,
+      "step": 11519
+    },
+    {
+      "epoch": 1.9632852519450457,
+      "grad_norm": 1.640625,
+      "learning_rate": 5.279862126465552e-06,
+      "loss": 0.8803,
+      "step": 11520
+    },
+    {
+      "epoch": 1.9634567160340355,
+      "grad_norm": 1.6875,
+      "learning_rate": 5.278270113118655e-06,
+      "loss": 0.8489,
+      "step": 11521
+    },
+    {
+      "epoch": 1.9636281801230255,
+      "grad_norm": 1.7109375,
+      "learning_rate": 5.276678253765151e-06,
+      "loss": 0.8641,
+      "step": 11522
+    },
+    {
+      "epoch": 1.9637996442120154,
+      "grad_norm": 1.6796875,
+      "learning_rate": 5.275086548456954e-06,
+      "loss": 0.8714,
+      "step": 11523
+    },
+    {
+      "epoch": 1.9639711083010052,
+      "grad_norm": 1.8046875,
+      "learning_rate": 5.273494997245978e-06,
+      "loss": 0.9374,
+      "step": 11524
+    },
+    {
+      "epoch": 1.9641425723899952,
+      "grad_norm": 1.6875,
+      "learning_rate": 5.2719036001841295e-06,
+      "loss": 0.8625,
+      "step": 11525
+    },
+    {
+      "epoch": 1.964314036478985,
+      "grad_norm": 1.75,
+      "learning_rate": 5.270312357323307e-06,
+      "loss": 0.8165,
+      "step": 11526
+    },
+    {
+      "epoch": 1.9644855005679749,
+      "grad_norm": 1.6640625,
+      "learning_rate": 5.268721268715413e-06,
+      "loss": 0.8819,
+      "step": 11527
+    },
+    {
+      "epoch": 1.9646569646569647,
+      "grad_norm": 1.7265625,
+      "learning_rate": 5.2671303344123315e-06,
+      "loss": 0.8243,
+      "step": 11528
+    },
+    {
+      "epoch": 1.9648284287459545,
+      "grad_norm": 1.65625,
+      "learning_rate": 5.265539554465953e-06,
+      "loss": 0.8084,
+      "step": 11529
+    },
+    {
+      "epoch": 1.9649998928349444,
+      "grad_norm": 1.7265625,
+      "learning_rate": 5.263948928928157e-06,
+      "loss": 0.8177,
+      "step": 11530
+    },
+    {
+      "epoch": 1.9651713569239342,
+      "grad_norm": 1.7265625,
+      "learning_rate": 5.26235845785082e-06,
+      "loss": 0.8604,
+      "step": 11531
+    },
+    {
+      "epoch": 1.965342821012924,
+      "grad_norm": 1.78125,
+      "learning_rate": 5.260768141285814e-06,
+      "loss": 0.877,
+      "step": 11532
+    },
+    {
+      "epoch": 1.9655142851019138,
+      "grad_norm": 1.671875,
+      "learning_rate": 5.259177979285003e-06,
+      "loss": 0.8761,
+      "step": 11533
+    },
+    {
+      "epoch": 1.9656857491909037,
+      "grad_norm": 1.671875,
+      "learning_rate": 5.257587971900253e-06,
+      "loss": 0.8233,
+      "step": 11534
+    },
+    {
+      "epoch": 1.9658572132798937,
+      "grad_norm": 1.7734375,
+      "learning_rate": 5.255998119183419e-06,
+      "loss": 0.9104,
+      "step": 11535
+    },
+    {
+      "epoch": 1.9660286773688835,
+      "grad_norm": 1.6796875,
+      "learning_rate": 5.254408421186348e-06,
+      "loss": 0.8313,
+      "step": 11536
+    },
+    {
+      "epoch": 1.9662001414578734,
+      "grad_norm": 1.7734375,
+      "learning_rate": 5.25281887796089e-06,
+      "loss": 0.8515,
+      "step": 11537
+    },
+    {
+      "epoch": 1.9663716055468634,
+      "grad_norm": 1.7265625,
+      "learning_rate": 5.251229489558889e-06,
+      "loss": 0.8842,
+      "step": 11538
+    },
+    {
+      "epoch": 1.9665430696358532,
+      "grad_norm": 1.7421875,
+      "learning_rate": 5.249640256032172e-06,
+      "loss": 0.8579,
+      "step": 11539
+    },
+    {
+      "epoch": 1.966714533724843,
+      "grad_norm": 1.8046875,
+      "learning_rate": 5.248051177432575e-06,
+      "loss": 0.8923,
+      "step": 11540
+    },
+    {
+      "epoch": 1.9668859978138329,
+      "grad_norm": 1.7109375,
+      "learning_rate": 5.246462253811923e-06,
+      "loss": 0.8346,
+      "step": 11541
+    },
+    {
+      "epoch": 1.9670574619028227,
+      "grad_norm": 1.640625,
+      "learning_rate": 5.244873485222038e-06,
+      "loss": 0.8458,
+      "step": 11542
+    },
+    {
+      "epoch": 1.9672289259918125,
+      "grad_norm": 1.765625,
+      "learning_rate": 5.243284871714733e-06,
+      "loss": 0.8551,
+      "step": 11543
+    },
+    {
+      "epoch": 1.9674003900808024,
+      "grad_norm": 1.734375,
+      "learning_rate": 5.2416964133418216e-06,
+      "loss": 0.8916,
+      "step": 11544
+    },
+    {
+      "epoch": 1.9675718541697922,
+      "grad_norm": 1.7734375,
+      "learning_rate": 5.240108110155108e-06,
+      "loss": 0.8748,
+      "step": 11545
+    },
+    {
+      "epoch": 1.967743318258782,
+      "grad_norm": 1.6953125,
+      "learning_rate": 5.238519962206392e-06,
+      "loss": 0.8406,
+      "step": 11546
+    },
+    {
+      "epoch": 1.967914782347772,
+      "grad_norm": 1.75,
+      "learning_rate": 5.236931969547472e-06,
+      "loss": 0.8873,
+      "step": 11547
+    },
+    {
+      "epoch": 1.9680862464367619,
+      "grad_norm": 1.7578125,
+      "learning_rate": 5.2353441322301356e-06,
+      "loss": 0.8211,
+      "step": 11548
+    },
+    {
+      "epoch": 1.9682577105257517,
+      "grad_norm": 1.8671875,
+      "learning_rate": 5.2337564503061685e-06,
+      "loss": 0.8973,
+      "step": 11549
+    },
+    {
+      "epoch": 1.9684291746147418,
+      "grad_norm": 1.7890625,
+      "learning_rate": 5.232168923827352e-06,
+      "loss": 0.8692,
+      "step": 11550
+    },
+    {
+      "epoch": 1.9686006387037316,
+      "grad_norm": 1.734375,
+      "learning_rate": 5.230581552845462e-06,
+      "loss": 0.9202,
+      "step": 11551
+    },
+    {
+      "epoch": 1.9687721027927214,
+      "grad_norm": 1.78125,
+      "learning_rate": 5.228994337412266e-06,
+      "loss": 0.8969,
+      "step": 11552
+    },
+    {
+      "epoch": 1.9689435668817112,
+      "grad_norm": 1.609375,
+      "learning_rate": 5.227407277579531e-06,
+      "loss": 0.8916,
+      "step": 11553
+    },
+    {
+      "epoch": 1.969115030970701,
+      "grad_norm": 2.265625,
+      "learning_rate": 5.225820373399018e-06,
+      "loss": 0.8081,
+      "step": 11554
+    },
+    {
+      "epoch": 1.969286495059691,
+      "grad_norm": 1.671875,
+      "learning_rate": 5.224233624922479e-06,
+      "loss": 0.7951,
+      "step": 11555
+    },
+    {
+      "epoch": 1.9694579591486807,
+      "grad_norm": 1.703125,
+      "learning_rate": 5.222647032201664e-06,
+      "loss": 0.8401,
+      "step": 11556
+    },
+    {
+      "epoch": 1.9696294232376705,
+      "grad_norm": 1.7109375,
+      "learning_rate": 5.221060595288321e-06,
+      "loss": 0.8383,
+      "step": 11557
+    },
+    {
+      "epoch": 1.9698008873266604,
+      "grad_norm": 1.8046875,
+      "learning_rate": 5.219474314234187e-06,
+      "loss": 0.8254,
+      "step": 11558
+    },
+    {
+      "epoch": 1.9699723514156504,
+      "grad_norm": 1.6953125,
+      "learning_rate": 5.217888189090999e-06,
+      "loss": 0.8489,
+      "step": 11559
+    },
+    {
+      "epoch": 1.9701438155046402,
+      "grad_norm": 1.6484375,
+      "learning_rate": 5.216302219910484e-06,
+      "loss": 0.8711,
+      "step": 11560
+    },
+    {
+      "epoch": 1.97031527959363,
+      "grad_norm": 1.8125,
+      "learning_rate": 5.2147164067443715e-06,
+      "loss": 0.8555,
+      "step": 11561
+    },
+    {
+      "epoch": 1.9704867436826201,
+      "grad_norm": 1.7109375,
+      "learning_rate": 5.2131307496443724e-06,
+      "loss": 0.8815,
+      "step": 11562
+    },
+    {
+      "epoch": 1.97065820777161,
+      "grad_norm": 1.8515625,
+      "learning_rate": 5.211545248662205e-06,
+      "loss": 0.8769,
+      "step": 11563
+    },
+    {
+      "epoch": 1.9708296718605998,
+      "grad_norm": 1.703125,
+      "learning_rate": 5.20995990384958e-06,
+      "loss": 0.7547,
+      "step": 11564
+    },
+    {
+      "epoch": 1.9710011359495896,
+      "grad_norm": 1.7578125,
+      "learning_rate": 5.2083747152581975e-06,
+      "loss": 0.9023,
+      "step": 11565
+    },
+    {
+      "epoch": 1.9711726000385794,
+      "grad_norm": 1.6953125,
+      "learning_rate": 5.206789682939761e-06,
+      "loss": 0.8359,
+      "step": 11566
+    },
+    {
+      "epoch": 1.9713440641275692,
+      "grad_norm": 1.6484375,
+      "learning_rate": 5.205204806945963e-06,
+      "loss": 0.7655,
+      "step": 11567
+    },
+    {
+      "epoch": 1.971515528216559,
+      "grad_norm": 1.6328125,
+      "learning_rate": 5.203620087328491e-06,
+      "loss": 0.7909,
+      "step": 11568
+    },
+    {
+      "epoch": 1.971686992305549,
+      "grad_norm": 1.71875,
+      "learning_rate": 5.202035524139029e-06,
+      "loss": 0.8972,
+      "step": 11569
+    },
+    {
+      "epoch": 1.9718584563945387,
+      "grad_norm": 1.6171875,
+      "learning_rate": 5.200451117429257e-06,
+      "loss": 0.8527,
+      "step": 11570
+    },
+    {
+      "epoch": 1.9720299204835288,
+      "grad_norm": 1.703125,
+      "learning_rate": 5.198866867250847e-06,
+      "loss": 0.8325,
+      "step": 11571
+    },
+    {
+      "epoch": 1.9722013845725186,
+      "grad_norm": 1.6484375,
+      "learning_rate": 5.197282773655463e-06,
+      "loss": 0.8403,
+      "step": 11572
+    },
+    {
+      "epoch": 1.9723728486615084,
+      "grad_norm": 1.6796875,
+      "learning_rate": 5.195698836694783e-06,
+      "loss": 0.816,
+      "step": 11573
+    },
+    {
+      "epoch": 1.9725443127504985,
+      "grad_norm": 1.7421875,
+      "learning_rate": 5.19411505642045e-06,
+      "loss": 0.8329,
+      "step": 11574
+    },
+    {
+      "epoch": 1.9727157768394883,
+      "grad_norm": 1.703125,
+      "learning_rate": 5.192531432884122e-06,
+      "loss": 0.955,
+      "step": 11575
+    },
+    {
+      "epoch": 1.9728872409284781,
+      "grad_norm": 1.703125,
+      "learning_rate": 5.1909479661374475e-06,
+      "loss": 0.8504,
+      "step": 11576
+    },
+    {
+      "epoch": 1.973058705017468,
+      "grad_norm": 1.6796875,
+      "learning_rate": 5.18936465623207e-06,
+      "loss": 0.8312,
+      "step": 11577
+    },
+    {
+      "epoch": 1.9732301691064578,
+      "grad_norm": 1.84375,
+      "learning_rate": 5.1877815032196255e-06,
+      "loss": 0.8168,
+      "step": 11578
+    },
+    {
+      "epoch": 1.9734016331954476,
+      "grad_norm": 1.6875,
+      "learning_rate": 5.186198507151747e-06,
+      "loss": 0.8544,
+      "step": 11579
+    },
+    {
+      "epoch": 1.9735730972844374,
+      "grad_norm": 1.7109375,
+      "learning_rate": 5.184615668080063e-06,
+      "loss": 0.8645,
+      "step": 11580
+    },
+    {
+      "epoch": 1.9737445613734272,
+      "grad_norm": 1.6640625,
+      "learning_rate": 5.183032986056195e-06,
+      "loss": 0.7389,
+      "step": 11581
+    },
+    {
+      "epoch": 1.973916025462417,
+      "grad_norm": 1.671875,
+      "learning_rate": 5.181450461131759e-06,
+      "loss": 0.7831,
+      "step": 11582
+    },
+    {
+      "epoch": 1.9740874895514071,
+      "grad_norm": 1.671875,
+      "learning_rate": 5.17986809335837e-06,
+      "loss": 0.881,
+      "step": 11583
+    },
+    {
+      "epoch": 1.974258953640397,
+      "grad_norm": 1.6640625,
+      "learning_rate": 5.178285882787637e-06,
+      "loss": 0.8336,
+      "step": 11584
+    },
+    {
+      "epoch": 1.9744304177293868,
+      "grad_norm": 1.7578125,
+      "learning_rate": 5.176703829471155e-06,
+      "loss": 0.8578,
+      "step": 11585
+    },
+    {
+      "epoch": 1.9746018818183768,
+      "grad_norm": 1.6015625,
+      "learning_rate": 5.175121933460523e-06,
+      "loss": 0.8177,
+      "step": 11586
+    },
+    {
+      "epoch": 1.9747733459073666,
+      "grad_norm": 1.7265625,
+      "learning_rate": 5.173540194807334e-06,
+      "loss": 0.853,
+      "step": 11587
+    },
+    {
+      "epoch": 1.9749448099963565,
+      "grad_norm": 1.703125,
+      "learning_rate": 5.171958613563168e-06,
+      "loss": 0.8128,
+      "step": 11588
+    },
+    {
+      "epoch": 1.9751162740853463,
+      "grad_norm": 1.65625,
+      "learning_rate": 5.170377189779618e-06,
+      "loss": 0.8597,
+      "step": 11589
+    },
+    {
+      "epoch": 1.9752877381743361,
+      "grad_norm": 1.75,
+      "learning_rate": 5.1687959235082554e-06,
+      "loss": 0.7888,
+      "step": 11590
+    },
+    {
+      "epoch": 1.975459202263326,
+      "grad_norm": 1.734375,
+      "learning_rate": 5.1672148148006494e-06,
+      "loss": 0.8659,
+      "step": 11591
+    },
+    {
+      "epoch": 1.9756306663523158,
+      "grad_norm": 1.6796875,
+      "learning_rate": 5.165633863708368e-06,
+      "loss": 0.8922,
+      "step": 11592
+    },
+    {
+      "epoch": 1.9758021304413056,
+      "grad_norm": 1.6875,
+      "learning_rate": 5.164053070282968e-06,
+      "loss": 0.9071,
+      "step": 11593
+    },
+    {
+      "epoch": 1.9759735945302954,
+      "grad_norm": 1.703125,
+      "learning_rate": 5.162472434576011e-06,
+      "loss": 0.9287,
+      "step": 11594
+    },
+    {
+      "epoch": 1.9761450586192855,
+      "grad_norm": 1.765625,
+      "learning_rate": 5.160891956639042e-06,
+      "loss": 0.8846,
+      "step": 11595
+    },
+    {
+      "epoch": 1.9763165227082753,
+      "grad_norm": 1.6640625,
+      "learning_rate": 5.159311636523614e-06,
+      "loss": 0.8229,
+      "step": 11596
+    },
+    {
+      "epoch": 1.9764879867972651,
+      "grad_norm": 1.625,
+      "learning_rate": 5.1577314742812575e-06,
+      "loss": 0.8095,
+      "step": 11597
+    },
+    {
+      "epoch": 1.9766594508862552,
+      "grad_norm": 1.7421875,
+      "learning_rate": 5.1561514699635126e-06,
+      "loss": 0.8743,
+      "step": 11598
+    },
+    {
+      "epoch": 1.976830914975245,
+      "grad_norm": 1.65625,
+      "learning_rate": 5.1545716236219075e-06,
+      "loss": 0.7827,
+      "step": 11599
+    },
+    {
+      "epoch": 1.9770023790642348,
+      "grad_norm": 1.7109375,
+      "learning_rate": 5.152991935307969e-06,
+      "loss": 0.9119,
+      "step": 11600
+    },
+    {
+      "epoch": 1.9771738431532246,
+      "grad_norm": 1.6796875,
+      "learning_rate": 5.151412405073215e-06,
+      "loss": 0.8324,
+      "step": 11601
+    },
+    {
+      "epoch": 1.9773453072422145,
+      "grad_norm": 1.6484375,
+      "learning_rate": 5.1498330329691605e-06,
+      "loss": 0.8356,
+      "step": 11602
+    },
+    {
+      "epoch": 1.9775167713312043,
+      "grad_norm": 1.640625,
+      "learning_rate": 5.148253819047315e-06,
+      "loss": 0.833,
+      "step": 11603
+    },
+    {
+      "epoch": 1.9776882354201941,
+      "grad_norm": 1.65625,
+      "learning_rate": 5.146674763359183e-06,
+      "loss": 0.8032,
+      "step": 11604
+    },
+    {
+      "epoch": 1.977859699509184,
+      "grad_norm": 1.7890625,
+      "learning_rate": 5.145095865956262e-06,
+      "loss": 0.8332,
+      "step": 11605
+    },
+    {
+      "epoch": 1.9780311635981738,
+      "grad_norm": 1.7421875,
+      "learning_rate": 5.1435171268900474e-06,
+      "loss": 0.8906,
+      "step": 11606
+    },
+    {
+      "epoch": 1.9782026276871638,
+      "grad_norm": 1.640625,
+      "learning_rate": 5.1419385462120264e-06,
+      "loss": 0.9111,
+      "step": 11607
+    },
+    {
+      "epoch": 1.9783740917761536,
+      "grad_norm": 1.65625,
+      "learning_rate": 5.140360123973683e-06,
+      "loss": 0.8105,
+      "step": 11608
+    },
+    {
+      "epoch": 1.9785455558651435,
+      "grad_norm": 1.671875,
+      "learning_rate": 5.138781860226496e-06,
+      "loss": 0.8314,
+      "step": 11609
+    },
+    {
+      "epoch": 1.9787170199541335,
+      "grad_norm": 1.6328125,
+      "learning_rate": 5.137203755021938e-06,
+      "loss": 0.7603,
+      "step": 11610
+    },
+    {
+      "epoch": 1.9788884840431233,
+      "grad_norm": 1.671875,
+      "learning_rate": 5.135625808411478e-06,
+      "loss": 0.874,
+      "step": 11611
+    },
+    {
+      "epoch": 1.9790599481321132,
+      "grad_norm": 1.7265625,
+      "learning_rate": 5.134048020446578e-06,
+      "loss": 0.8629,
+      "step": 11612
+    },
+    {
+      "epoch": 1.979231412221103,
+      "grad_norm": 1.8125,
+      "learning_rate": 5.1324703911786946e-06,
+      "loss": 0.8188,
+      "step": 11613
+    },
+    {
+      "epoch": 1.9794028763100928,
+      "grad_norm": 1.7109375,
+      "learning_rate": 5.1308929206592805e-06,
+      "loss": 0.7701,
+      "step": 11614
+    },
+    {
+      "epoch": 1.9795743403990826,
+      "grad_norm": 1.703125,
+      "learning_rate": 5.129315608939786e-06,
+      "loss": 0.8955,
+      "step": 11615
+    },
+    {
+      "epoch": 1.9797458044880725,
+      "grad_norm": 1.7421875,
+      "learning_rate": 5.1277384560716495e-06,
+      "loss": 0.8836,
+      "step": 11616
+    },
+    {
+      "epoch": 1.9799172685770623,
+      "grad_norm": 1.703125,
+      "learning_rate": 5.126161462106308e-06,
+      "loss": 0.9262,
+      "step": 11617
+    },
+    {
+      "epoch": 1.9800887326660521,
+      "grad_norm": 1.7578125,
+      "learning_rate": 5.124584627095196e-06,
+      "loss": 0.8944,
+      "step": 11618
+    },
+    {
+      "epoch": 1.9802601967550422,
+      "grad_norm": 1.6484375,
+      "learning_rate": 5.123007951089743e-06,
+      "loss": 0.8255,
+      "step": 11619
+    },
+    {
+      "epoch": 1.980431660844032,
+      "grad_norm": 1.6796875,
+      "learning_rate": 5.12143143414136e-06,
+      "loss": 0.8353,
+      "step": 11620
+    },
+    {
+      "epoch": 1.9806031249330218,
+      "grad_norm": 1.7578125,
+      "learning_rate": 5.119855076301468e-06,
+      "loss": 0.8852,
+      "step": 11621
+    },
+    {
+      "epoch": 1.9807745890220119,
+      "grad_norm": 1.671875,
+      "learning_rate": 5.118278877621481e-06,
+      "loss": 0.8293,
+      "step": 11622
+    },
+    {
+      "epoch": 1.9809460531110017,
+      "grad_norm": 1.65625,
+      "learning_rate": 5.116702838152803e-06,
+      "loss": 0.8015,
+      "step": 11623
+    },
+    {
+      "epoch": 1.9811175171999915,
+      "grad_norm": 1.7265625,
+      "learning_rate": 5.115126957946832e-06,
+      "loss": 0.8779,
+      "step": 11624
+    },
+    {
+      "epoch": 1.9812889812889813,
+      "grad_norm": 1.8046875,
+      "learning_rate": 5.113551237054968e-06,
+      "loss": 0.869,
+      "step": 11625
+    },
+    {
+      "epoch": 1.9814604453779712,
+      "grad_norm": 1.796875,
+      "learning_rate": 5.111975675528598e-06,
+      "loss": 0.8304,
+      "step": 11626
+    },
+    {
+      "epoch": 1.981631909466961,
+      "grad_norm": 1.6171875,
+      "learning_rate": 5.110400273419104e-06,
+      "loss": 0.7922,
+      "step": 11627
+    },
+    {
+      "epoch": 1.9818033735559508,
+      "grad_norm": 1.625,
+      "learning_rate": 5.108825030777873e-06,
+      "loss": 0.8331,
+      "step": 11628
+    },
+    {
+      "epoch": 1.9819748376449406,
+      "grad_norm": 1.7421875,
+      "learning_rate": 5.107249947656276e-06,
+      "loss": 0.8969,
+      "step": 11629
+    },
+    {
+      "epoch": 1.9821463017339305,
+      "grad_norm": 1.7265625,
+      "learning_rate": 5.105675024105683e-06,
+      "loss": 0.8615,
+      "step": 11630
+    },
+    {
+      "epoch": 1.9823177658229203,
+      "grad_norm": 1.734375,
+      "learning_rate": 5.104100260177462e-06,
+      "loss": 0.8852,
+      "step": 11631
+    },
+    {
+      "epoch": 1.9824892299119103,
+      "grad_norm": 1.7890625,
+      "learning_rate": 5.102525655922963e-06,
+      "loss": 0.8795,
+      "step": 11632
+    },
+    {
+      "epoch": 1.9826606940009002,
+      "grad_norm": 1.6171875,
+      "learning_rate": 5.100951211393545e-06,
+      "loss": 0.8602,
+      "step": 11633
+    },
+    {
+      "epoch": 1.98283215808989,
+      "grad_norm": 1.71875,
+      "learning_rate": 5.099376926640558e-06,
+      "loss": 0.9049,
+      "step": 11634
+    },
+    {
+      "epoch": 1.98300362217888,
+      "grad_norm": 1.6484375,
+      "learning_rate": 5.097802801715341e-06,
+      "loss": 0.8136,
+      "step": 11635
+    },
+    {
+      "epoch": 1.9831750862678699,
+      "grad_norm": 1.6484375,
+      "learning_rate": 5.096228836669234e-06,
+      "loss": 0.8509,
+      "step": 11636
+    },
+    {
+      "epoch": 1.9833465503568597,
+      "grad_norm": 1.703125,
+      "learning_rate": 5.094655031553572e-06,
+      "loss": 0.8705,
+      "step": 11637
+    },
+    {
+      "epoch": 1.9835180144458495,
+      "grad_norm": 1.5703125,
+      "learning_rate": 5.093081386419682e-06,
+      "loss": 0.731,
+      "step": 11638
+    },
+    {
+      "epoch": 1.9836894785348393,
+      "grad_norm": 1.7109375,
+      "learning_rate": 5.091507901318885e-06,
+      "loss": 0.8829,
+      "step": 11639
+    },
+    {
+      "epoch": 1.9838609426238292,
+      "grad_norm": 1.734375,
+      "learning_rate": 5.0899345763024986e-06,
+      "loss": 0.8311,
+      "step": 11640
+    },
+    {
+      "epoch": 1.984032406712819,
+      "grad_norm": 1.71875,
+      "learning_rate": 5.088361411421836e-06,
+      "loss": 0.8326,
+      "step": 11641
+    },
+    {
+      "epoch": 1.9842038708018088,
+      "grad_norm": 1.6328125,
+      "learning_rate": 5.086788406728207e-06,
+      "loss": 0.8074,
+      "step": 11642
+    },
+    {
+      "epoch": 1.9843753348907986,
+      "grad_norm": 1.6640625,
+      "learning_rate": 5.085215562272901e-06,
+      "loss": 0.8817,
+      "step": 11643
+    },
+    {
+      "epoch": 1.9845467989797887,
+      "grad_norm": 1.7265625,
+      "learning_rate": 5.0836428781072266e-06,
+      "loss": 0.8249,
+      "step": 11644
+    },
+    {
+      "epoch": 1.9847182630687785,
+      "grad_norm": 1.734375,
+      "learning_rate": 5.082070354282473e-06,
+      "loss": 0.8485,
+      "step": 11645
+    },
+    {
+      "epoch": 1.9848897271577683,
+      "grad_norm": 1.75,
+      "learning_rate": 5.080497990849922e-06,
+      "loss": 0.9151,
+      "step": 11646
+    },
+    {
+      "epoch": 1.9850611912467584,
+      "grad_norm": 1.765625,
+      "learning_rate": 5.078925787860858e-06,
+      "loss": 0.7656,
+      "step": 11647
+    },
+    {
+      "epoch": 1.9852326553357482,
+      "grad_norm": 1.6484375,
+      "learning_rate": 5.077353745366555e-06,
+      "loss": 0.8011,
+      "step": 11648
+    },
+    {
+      "epoch": 1.985404119424738,
+      "grad_norm": 1.703125,
+      "learning_rate": 5.075781863418283e-06,
+      "loss": 0.8813,
+      "step": 11649
+    },
+    {
+      "epoch": 1.9855755835137279,
+      "grad_norm": 1.6171875,
+      "learning_rate": 5.074210142067309e-06,
+      "loss": 0.8026,
+      "step": 11650
+    },
+    {
+      "epoch": 1.9857470476027177,
+      "grad_norm": 1.7421875,
+      "learning_rate": 5.072638581364888e-06,
+      "loss": 0.8569,
+      "step": 11651
+    },
+    {
+      "epoch": 1.9859185116917075,
+      "grad_norm": 1.7578125,
+      "learning_rate": 5.071067181362279e-06,
+      "loss": 0.926,
+      "step": 11652
+    },
+    {
+      "epoch": 1.9860899757806973,
+      "grad_norm": 1.71875,
+      "learning_rate": 5.069495942110731e-06,
+      "loss": 0.8279,
+      "step": 11653
+    },
+    {
+      "epoch": 1.9862614398696872,
+      "grad_norm": 1.6953125,
+      "learning_rate": 5.06792486366149e-06,
+      "loss": 0.8042,
+      "step": 11654
+    },
+    {
+      "epoch": 1.986432903958677,
+      "grad_norm": 1.765625,
+      "learning_rate": 5.066353946065787e-06,
+      "loss": 0.8291,
+      "step": 11655
+    },
+    {
+      "epoch": 1.986604368047667,
+      "grad_norm": 1.6640625,
+      "learning_rate": 5.064783189374862e-06,
+      "loss": 0.8628,
+      "step": 11656
+    },
+    {
+      "epoch": 1.9867758321366569,
+      "grad_norm": 1.671875,
+      "learning_rate": 5.063212593639939e-06,
+      "loss": 0.8537,
+      "step": 11657
+    },
+    {
+      "epoch": 1.9869472962256467,
+      "grad_norm": 1.6875,
+      "learning_rate": 5.061642158912246e-06,
+      "loss": 0.8203,
+      "step": 11658
+    },
+    {
+      "epoch": 1.9871187603146367,
+      "grad_norm": 1.7578125,
+      "learning_rate": 5.060071885242996e-06,
+      "loss": 0.8631,
+      "step": 11659
+    },
+    {
+      "epoch": 1.9872902244036266,
+      "grad_norm": 1.765625,
+      "learning_rate": 5.0585017726834065e-06,
+      "loss": 0.801,
+      "step": 11660
+    },
+    {
+      "epoch": 1.9874616884926164,
+      "grad_norm": 1.6484375,
+      "learning_rate": 5.056931821284681e-06,
+      "loss": 0.87,
+      "step": 11661
+    },
+    {
+      "epoch": 1.9876331525816062,
+      "grad_norm": 1.8046875,
+      "learning_rate": 5.055362031098021e-06,
+      "loss": 0.8322,
+      "step": 11662
+    },
+    {
+      "epoch": 1.987804616670596,
+      "grad_norm": 1.671875,
+      "learning_rate": 5.053792402174627e-06,
+      "loss": 0.7921,
+      "step": 11663
+    },
+    {
+      "epoch": 1.9879760807595859,
+      "grad_norm": 1.7421875,
+      "learning_rate": 5.0522229345656885e-06,
+      "loss": 0.8372,
+      "step": 11664
+    }
+  ],
+  "logging_steps": 1,
+  "max_steps": 17496,
+  "num_input_tokens_seen": 0,
+  "num_train_epochs": 3,
+  "save_steps": 5832,
+  "stateful_callbacks": {
+    "TrainerControl": {
+      "args": {
+        "should_epoch_stop": false,
+        "should_evaluate": false,
+        "should_log": false,
+        "should_save": true,
+        "should_training_stop": false
+      },
+      "attributes": {}
+    }
+  },
+  "total_flos": 3.4421122547400573e+19,
+  "train_batch_size": 1,
+  "trial_name": null,
+  "trial_params": null
+}