diff --git "a/trainer_state.json" "b/trainer_state.json"
new file mode 100644--- /dev/null
+++ "b/trainer_state.json"
@@ -0,0 +1,10533 @@
+{
+  "best_metric": null,
+  "best_model_checkpoint": null,
+  "epoch": 0.12800136534789705,
+  "eval_steps": 500,
+  "global_step": 1500,
+  "is_hyper_param_search": false,
+  "is_local_process_zero": true,
+  "is_world_process_zero": true,
+  "log_history": [
+    {
+      "epoch": 8.53342435652647e-05,
+      "grad_norm": 0.82421875,
+      "learning_rate": 1.0000000000000002e-06,
+      "loss": 1.8,
+      "step": 1
+    },
+    {
+      "epoch": 0.0001706684871305294,
+      "grad_norm": 0.9140625,
+      "learning_rate": 2.0000000000000003e-06,
+      "loss": 1.8859,
+      "step": 2
+    },
+    {
+      "epoch": 0.0002560027306957941,
+      "grad_norm": 0.90625,
+      "learning_rate": 3e-06,
+      "loss": 1.7003,
+      "step": 3
+    },
+    {
+      "epoch": 0.0003413369742610588,
+      "grad_norm": 0.76171875,
+      "learning_rate": 4.000000000000001e-06,
+      "loss": 1.6892,
+      "step": 4
+    },
+    {
+      "epoch": 0.0004266712178263235,
+      "grad_norm": 0.77734375,
+      "learning_rate": 5e-06,
+      "loss": 1.6493,
+      "step": 5
+    },
+    {
+      "epoch": 0.0005120054613915882,
+      "grad_norm": 0.80859375,
+      "learning_rate": 6e-06,
+      "loss": 1.7291,
+      "step": 6
+    },
+    {
+      "epoch": 0.0005973397049568529,
+      "grad_norm": 0.8203125,
+      "learning_rate": 7.000000000000001e-06,
+      "loss": 1.7985,
+      "step": 7
+    },
+    {
+      "epoch": 0.0006826739485221176,
+      "grad_norm": 0.875,
+      "learning_rate": 8.000000000000001e-06,
+      "loss": 1.6985,
+      "step": 8
+    },
+    {
+      "epoch": 0.0007680081920873823,
+      "grad_norm": 0.8671875,
+      "learning_rate": 9e-06,
+      "loss": 1.9519,
+      "step": 9
+    },
+    {
+      "epoch": 0.000853342435652647,
+      "grad_norm": 0.83984375,
+      "learning_rate": 1e-05,
+      "loss": 1.8502,
+      "step": 10
+    },
+    {
+      "epoch": 0.0009386766792179117,
+      "grad_norm": 0.921875,
+      "learning_rate": 1.1000000000000001e-05,
+      "loss": 1.6864,
+      "step": 11
+    },
+    {
+      "epoch": 0.0010240109227831764,
+      "grad_norm": 0.90625,
+      "learning_rate": 1.2e-05,
+      "loss": 2.0213,
+      "step": 12
+    },
+    {
+      "epoch": 0.001109345166348441,
+      "grad_norm": 0.8046875,
+      "learning_rate": 1.3000000000000001e-05,
+      "loss": 1.7538,
+      "step": 13
+    },
+    {
+      "epoch": 0.0011946794099137058,
+      "grad_norm": 0.921875,
+      "learning_rate": 1.4000000000000001e-05,
+      "loss": 1.8914,
+      "step": 14
+    },
+    {
+      "epoch": 0.0012800136534789704,
+      "grad_norm": 0.890625,
+      "learning_rate": 1.5e-05,
+      "loss": 1.7576,
+      "step": 15
+    },
+    {
+      "epoch": 0.0013653478970442352,
+      "grad_norm": 0.84375,
+      "learning_rate": 1.6000000000000003e-05,
+      "loss": 1.6817,
+      "step": 16
+    },
+    {
+      "epoch": 0.0014506821406094998,
+      "grad_norm": 0.91015625,
+      "learning_rate": 1.7000000000000003e-05,
+      "loss": 1.9613,
+      "step": 17
+    },
+    {
+      "epoch": 0.0015360163841747646,
+      "grad_norm": 0.86328125,
+      "learning_rate": 1.8e-05,
+      "loss": 1.692,
+      "step": 18
+    },
+    {
+      "epoch": 0.0016213506277400291,
+      "grad_norm": 0.7578125,
+      "learning_rate": 1.9e-05,
+      "loss": 1.5385,
+      "step": 19
+    },
+    {
+      "epoch": 0.001706684871305294,
+      "grad_norm": 0.8515625,
+      "learning_rate": 2e-05,
+      "loss": 1.8615,
+      "step": 20
+    },
+    {
+      "epoch": 0.0017920191148705585,
+      "grad_norm": 0.83984375,
+      "learning_rate": 2.1e-05,
+      "loss": 1.7127,
+      "step": 21
+    },
+    {
+      "epoch": 0.0018773533584358233,
+      "grad_norm": 0.80859375,
+      "learning_rate": 2.2000000000000003e-05,
+      "loss": 1.8537,
+      "step": 22
+    },
+    {
+      "epoch": 0.001962687602001088,
+      "grad_norm": 0.80078125,
+      "learning_rate": 2.3000000000000003e-05,
+      "loss": 2.0063,
+      "step": 23
+    },
+    {
+      "epoch": 0.0020480218455663527,
+      "grad_norm": 0.79296875,
+      "learning_rate": 2.4e-05,
+      "loss": 1.636,
+      "step": 24
+    },
+    {
+      "epoch": 0.0021333560891316173,
+      "grad_norm": 0.98046875,
+      "learning_rate": 2.5e-05,
+      "loss": 1.7672,
+      "step": 25
+    },
+    {
+      "epoch": 0.002218690332696882,
+      "grad_norm": 0.81640625,
+      "learning_rate": 2.6000000000000002e-05,
+      "loss": 2.0052,
+      "step": 26
+    },
+    {
+      "epoch": 0.002304024576262147,
+      "grad_norm": 0.9453125,
+      "learning_rate": 2.7000000000000002e-05,
+      "loss": 1.9785,
+      "step": 27
+    },
+    {
+      "epoch": 0.0023893588198274115,
+      "grad_norm": 0.85546875,
+      "learning_rate": 2.8000000000000003e-05,
+      "loss": 1.7909,
+      "step": 28
+    },
+    {
+      "epoch": 0.002474693063392676,
+      "grad_norm": 0.765625,
+      "learning_rate": 2.9e-05,
+      "loss": 1.8871,
+      "step": 29
+    },
+    {
+      "epoch": 0.0025600273069579407,
+      "grad_norm": 0.8203125,
+      "learning_rate": 3e-05,
+      "loss": 1.707,
+      "step": 30
+    },
+    {
+      "epoch": 0.0026453615505232057,
+      "grad_norm": 0.796875,
+      "learning_rate": 3.1e-05,
+      "loss": 1.9111,
+      "step": 31
+    },
+    {
+      "epoch": 0.0027306957940884703,
+      "grad_norm": 0.73828125,
+      "learning_rate": 3.2000000000000005e-05,
+      "loss": 1.6338,
+      "step": 32
+    },
+    {
+      "epoch": 0.002816030037653735,
+      "grad_norm": 0.80078125,
+      "learning_rate": 3.3e-05,
+      "loss": 1.8687,
+      "step": 33
+    },
+    {
+      "epoch": 0.0029013642812189995,
+      "grad_norm": 0.72265625,
+      "learning_rate": 3.4000000000000007e-05,
+      "loss": 1.7552,
+      "step": 34
+    },
+    {
+      "epoch": 0.0029866985247842645,
+      "grad_norm": 0.71875,
+      "learning_rate": 3.5e-05,
+      "loss": 1.6459,
+      "step": 35
+    },
+    {
+      "epoch": 0.003072032768349529,
+      "grad_norm": 0.89453125,
+      "learning_rate": 3.6e-05,
+      "loss": 1.8945,
+      "step": 36
+    },
+    {
+      "epoch": 0.0031573670119147937,
+      "grad_norm": 0.86328125,
+      "learning_rate": 3.7e-05,
+      "loss": 1.8111,
+      "step": 37
+    },
+    {
+      "epoch": 0.0032427012554800583,
+      "grad_norm": 0.84765625,
+      "learning_rate": 3.8e-05,
+      "loss": 1.8449,
+      "step": 38
+    },
+    {
+      "epoch": 0.0033280354990453233,
+      "grad_norm": 0.73828125,
+      "learning_rate": 3.9000000000000006e-05,
+      "loss": 1.7649,
+      "step": 39
+    },
+    {
+      "epoch": 0.003413369742610588,
+      "grad_norm": 0.82421875,
+      "learning_rate": 4e-05,
+      "loss": 1.8078,
+      "step": 40
+    },
+    {
+      "epoch": 0.0034987039861758525,
+      "grad_norm": 0.74609375,
+      "learning_rate": 4.1e-05,
+      "loss": 1.6198,
+      "step": 41
+    },
+    {
+      "epoch": 0.003584038229741117,
+      "grad_norm": 0.90625,
+      "learning_rate": 4.2e-05,
+      "loss": 1.7853,
+      "step": 42
+    },
+    {
+      "epoch": 0.003669372473306382,
+      "grad_norm": 0.828125,
+      "learning_rate": 4.3e-05,
+      "loss": 1.8162,
+      "step": 43
+    },
+    {
+      "epoch": 0.0037547067168716467,
+      "grad_norm": 0.7265625,
+      "learning_rate": 4.4000000000000006e-05,
+      "loss": 1.8106,
+      "step": 44
+    },
+    {
+      "epoch": 0.0038400409604369113,
+      "grad_norm": 0.69921875,
+      "learning_rate": 4.5e-05,
+      "loss": 1.6102,
+      "step": 45
+    },
+    {
+      "epoch": 0.003925375204002176,
+      "grad_norm": 0.6875,
+      "learning_rate": 4.600000000000001e-05,
+      "loss": 1.6463,
+      "step": 46
+    },
+    {
+      "epoch": 0.004010709447567441,
+      "grad_norm": 0.66015625,
+      "learning_rate": 4.7e-05,
+      "loss": 1.546,
+      "step": 47
+    },
+    {
+      "epoch": 0.0040960436911327055,
+      "grad_norm": 0.703125,
+      "learning_rate": 4.8e-05,
+      "loss": 1.7966,
+      "step": 48
+    },
+    {
+      "epoch": 0.00418137793469797,
+      "grad_norm": 0.796875,
+      "learning_rate": 4.9e-05,
+      "loss": 1.9185,
+      "step": 49
+    },
+    {
+      "epoch": 0.004266712178263235,
+      "grad_norm": 0.70703125,
+      "learning_rate": 5e-05,
+      "loss": 1.6677,
+      "step": 50
+    },
+    {
+      "epoch": 0.004352046421828499,
+      "grad_norm": 0.765625,
+      "learning_rate": 5.1000000000000006e-05,
+      "loss": 1.7566,
+      "step": 51
+    },
+    {
+      "epoch": 0.004437380665393764,
+      "grad_norm": 0.7578125,
+      "learning_rate": 5.2000000000000004e-05,
+      "loss": 1.6925,
+      "step": 52
+    },
+    {
+      "epoch": 0.004522714908959029,
+      "grad_norm": 0.8515625,
+      "learning_rate": 5.300000000000001e-05,
+      "loss": 1.8328,
+      "step": 53
+    },
+    {
+      "epoch": 0.004608049152524294,
+      "grad_norm": 0.73046875,
+      "learning_rate": 5.4000000000000005e-05,
+      "loss": 1.906,
+      "step": 54
+    },
+    {
+      "epoch": 0.0046933833960895585,
+      "grad_norm": 0.68359375,
+      "learning_rate": 5.500000000000001e-05,
+      "loss": 1.6741,
+      "step": 55
+    },
+    {
+      "epoch": 0.004778717639654823,
+      "grad_norm": 0.67578125,
+      "learning_rate": 5.6000000000000006e-05,
+      "loss": 1.8366,
+      "step": 56
+    },
+    {
+      "epoch": 0.004864051883220088,
+      "grad_norm": 0.69921875,
+      "learning_rate": 5.6999999999999996e-05,
+      "loss": 1.733,
+      "step": 57
+    },
+    {
+      "epoch": 0.004949386126785352,
+      "grad_norm": 0.6171875,
+      "learning_rate": 5.8e-05,
+      "loss": 1.8005,
+      "step": 58
+    },
+    {
+      "epoch": 0.005034720370350617,
+      "grad_norm": 0.734375,
+      "learning_rate": 5.9e-05,
+      "loss": 1.7329,
+      "step": 59
+    },
+    {
+      "epoch": 0.005120054613915881,
+      "grad_norm": 0.69140625,
+      "learning_rate": 6e-05,
+      "loss": 1.7772,
+      "step": 60
+    },
+    {
+      "epoch": 0.005205388857481147,
+      "grad_norm": 0.671875,
+      "learning_rate": 6.1e-05,
+      "loss": 1.6311,
+      "step": 61
+    },
+    {
+      "epoch": 0.0052907231010464115,
+      "grad_norm": 0.59765625,
+      "learning_rate": 6.2e-05,
+      "loss": 1.506,
+      "step": 62
+    },
+    {
+      "epoch": 0.005376057344611676,
+      "grad_norm": 0.6796875,
+      "learning_rate": 6.3e-05,
+      "loss": 1.6361,
+      "step": 63
+    },
+    {
+      "epoch": 0.005461391588176941,
+      "grad_norm": 0.65234375,
+      "learning_rate": 6.400000000000001e-05,
+      "loss": 1.6185,
+      "step": 64
+    },
+    {
+      "epoch": 0.005546725831742205,
+      "grad_norm": 0.6328125,
+      "learning_rate": 6.500000000000001e-05,
+      "loss": 1.7215,
+      "step": 65
+    },
+    {
+      "epoch": 0.00563206007530747,
+      "grad_norm": 0.671875,
+      "learning_rate": 6.6e-05,
+      "loss": 1.8679,
+      "step": 66
+    },
+    {
+      "epoch": 0.005717394318872734,
+      "grad_norm": 0.5859375,
+      "learning_rate": 6.7e-05,
+      "loss": 1.7063,
+      "step": 67
+    },
+    {
+      "epoch": 0.005802728562437999,
+      "grad_norm": 0.58984375,
+      "learning_rate": 6.800000000000001e-05,
+      "loss": 1.6245,
+      "step": 68
+    },
+    {
+      "epoch": 0.0058880628060032645,
+      "grad_norm": 0.578125,
+      "learning_rate": 6.9e-05,
+      "loss": 1.8109,
+      "step": 69
+    },
+    {
+      "epoch": 0.005973397049568529,
+      "grad_norm": 0.67578125,
+      "learning_rate": 7e-05,
+      "loss": 1.9045,
+      "step": 70
+    },
+    {
+      "epoch": 0.006058731293133794,
+      "grad_norm": 0.70703125,
+      "learning_rate": 7.1e-05,
+      "loss": 1.6802,
+      "step": 71
+    },
+    {
+      "epoch": 0.006144065536699058,
+      "grad_norm": 0.66015625,
+      "learning_rate": 7.2e-05,
+      "loss": 1.757,
+      "step": 72
+    },
+    {
+      "epoch": 0.006229399780264323,
+      "grad_norm": 0.6171875,
+      "learning_rate": 7.3e-05,
+      "loss": 1.7513,
+      "step": 73
+    },
+    {
+      "epoch": 0.006314734023829587,
+      "grad_norm": 0.60546875,
+      "learning_rate": 7.4e-05,
+      "loss": 1.8557,
+      "step": 74
+    },
+    {
+      "epoch": 0.006400068267394852,
+      "grad_norm": 0.5625,
+      "learning_rate": 7.500000000000001e-05,
+      "loss": 1.6747,
+      "step": 75
+    },
+    {
+      "epoch": 0.006485402510960117,
+      "grad_norm": 0.6015625,
+      "learning_rate": 7.6e-05,
+      "loss": 1.6629,
+      "step": 76
+    },
+    {
+      "epoch": 0.006570736754525381,
+      "grad_norm": 0.78515625,
+      "learning_rate": 7.7e-05,
+      "loss": 1.8731,
+      "step": 77
+    },
+    {
+      "epoch": 0.006656070998090647,
+      "grad_norm": 0.59765625,
+      "learning_rate": 7.800000000000001e-05,
+      "loss": 1.6409,
+      "step": 78
+    },
+    {
+      "epoch": 0.006741405241655911,
+      "grad_norm": 0.63671875,
+      "learning_rate": 7.900000000000001e-05,
+      "loss": 1.6955,
+      "step": 79
+    },
+    {
+      "epoch": 0.006826739485221176,
+      "grad_norm": 0.58203125,
+      "learning_rate": 8e-05,
+      "loss": 1.6168,
+      "step": 80
+    },
+    {
+      "epoch": 0.00691207372878644,
+      "grad_norm": 0.60546875,
+      "learning_rate": 8.1e-05,
+      "loss": 1.8731,
+      "step": 81
+    },
+    {
+      "epoch": 0.006997407972351705,
+      "grad_norm": 0.5625,
+      "learning_rate": 8.2e-05,
+      "loss": 1.8115,
+      "step": 82
+    },
+    {
+      "epoch": 0.00708274221591697,
+      "grad_norm": 0.6328125,
+      "learning_rate": 8.3e-05,
+      "loss": 1.8266,
+      "step": 83
+    },
+    {
+      "epoch": 0.007168076459482234,
+      "grad_norm": 0.5390625,
+      "learning_rate": 8.4e-05,
+      "loss": 1.6428,
+      "step": 84
+    },
+    {
+      "epoch": 0.007253410703047499,
+      "grad_norm": 0.53515625,
+      "learning_rate": 8.5e-05,
+      "loss": 1.6117,
+      "step": 85
+    },
+    {
+      "epoch": 0.007338744946612764,
+      "grad_norm": 0.5,
+      "learning_rate": 8.6e-05,
+      "loss": 1.5626,
+      "step": 86
+    },
+    {
+      "epoch": 0.007424079190178029,
+      "grad_norm": 0.51953125,
+      "learning_rate": 8.7e-05,
+      "loss": 1.686,
+      "step": 87
+    },
+    {
+      "epoch": 0.007509413433743293,
+      "grad_norm": 0.50390625,
+      "learning_rate": 8.800000000000001e-05,
+      "loss": 1.6377,
+      "step": 88
+    },
+    {
+      "epoch": 0.007594747677308558,
+      "grad_norm": 0.48046875,
+      "learning_rate": 8.900000000000001e-05,
+      "loss": 1.5294,
+      "step": 89
+    },
+    {
+      "epoch": 0.007680081920873823,
+      "grad_norm": 0.50390625,
+      "learning_rate": 9e-05,
+      "loss": 1.6331,
+      "step": 90
+    },
+    {
+      "epoch": 0.007765416164439087,
+      "grad_norm": 0.5078125,
+      "learning_rate": 9.1e-05,
+      "loss": 1.725,
+      "step": 91
+    },
+    {
+      "epoch": 0.007850750408004353,
+      "grad_norm": 0.5,
+      "learning_rate": 9.200000000000001e-05,
+      "loss": 1.6189,
+      "step": 92
+    },
+    {
+      "epoch": 0.007936084651569617,
+      "grad_norm": 0.546875,
+      "learning_rate": 9.300000000000001e-05,
+      "loss": 1.5317,
+      "step": 93
+    },
+    {
+      "epoch": 0.008021418895134882,
+      "grad_norm": 0.48828125,
+      "learning_rate": 9.4e-05,
+      "loss": 1.4674,
+      "step": 94
+    },
+    {
+      "epoch": 0.008106753138700146,
+      "grad_norm": 0.55078125,
+      "learning_rate": 9.5e-05,
+      "loss": 1.8128,
+      "step": 95
+    },
+    {
+      "epoch": 0.008192087382265411,
+      "grad_norm": 0.54296875,
+      "learning_rate": 9.6e-05,
+      "loss": 1.5884,
+      "step": 96
+    },
+    {
+      "epoch": 0.008277421625830676,
+      "grad_norm": 0.46875,
+      "learning_rate": 9.7e-05,
+      "loss": 1.6662,
+      "step": 97
+    },
+    {
+      "epoch": 0.00836275586939594,
+      "grad_norm": 0.5078125,
+      "learning_rate": 9.8e-05,
+      "loss": 1.5444,
+      "step": 98
+    },
+    {
+      "epoch": 0.008448090112961205,
+      "grad_norm": 0.57421875,
+      "learning_rate": 9.900000000000001e-05,
+      "loss": 1.8447,
+      "step": 99
+    },
+    {
+      "epoch": 0.00853342435652647,
+      "grad_norm": 0.515625,
+      "learning_rate": 0.0001,
+      "loss": 1.5976,
+      "step": 100
+    },
+    {
+      "epoch": 0.008618758600091734,
+      "grad_norm": 0.6640625,
+      "learning_rate": 9.999497487437187e-05,
+      "loss": 1.4566,
+      "step": 101
+    },
+    {
+      "epoch": 0.008704092843656999,
+      "grad_norm": 0.5078125,
+      "learning_rate": 9.998994974874373e-05,
+      "loss": 1.9021,
+      "step": 102
+    },
+    {
+      "epoch": 0.008789427087222263,
+      "grad_norm": 0.435546875,
+      "learning_rate": 9.998492462311558e-05,
+      "loss": 1.4717,
+      "step": 103
+    },
+    {
+      "epoch": 0.008874761330787528,
+      "grad_norm": 0.482421875,
+      "learning_rate": 9.997989949748744e-05,
+      "loss": 1.5066,
+      "step": 104
+    },
+    {
+      "epoch": 0.008960095574352792,
+      "grad_norm": 0.578125,
+      "learning_rate": 9.99748743718593e-05,
+      "loss": 1.553,
+      "step": 105
+    },
+    {
+      "epoch": 0.009045429817918059,
+      "grad_norm": 0.4609375,
+      "learning_rate": 9.996984924623116e-05,
+      "loss": 1.4151,
+      "step": 106
+    },
+    {
+      "epoch": 0.009130764061483323,
+      "grad_norm": 0.4921875,
+      "learning_rate": 9.996482412060301e-05,
+      "loss": 1.7529,
+      "step": 107
+    },
+    {
+      "epoch": 0.009216098305048588,
+      "grad_norm": 0.486328125,
+      "learning_rate": 9.995979899497487e-05,
+      "loss": 1.6218,
+      "step": 108
+    },
+    {
+      "epoch": 0.009301432548613852,
+      "grad_norm": 0.4765625,
+      "learning_rate": 9.995477386934674e-05,
+      "loss": 1.6131,
+      "step": 109
+    },
+    {
+      "epoch": 0.009386766792179117,
+      "grad_norm": 0.55078125,
+      "learning_rate": 9.99497487437186e-05,
+      "loss": 1.7024,
+      "step": 110
+    },
+    {
+      "epoch": 0.009472101035744382,
+      "grad_norm": 0.4453125,
+      "learning_rate": 9.994472361809045e-05,
+      "loss": 1.5234,
+      "step": 111
+    },
+    {
+      "epoch": 0.009557435279309646,
+      "grad_norm": 0.482421875,
+      "learning_rate": 9.993969849246232e-05,
+      "loss": 1.4346,
+      "step": 112
+    },
+    {
+      "epoch": 0.00964276952287491,
+      "grad_norm": 0.427734375,
+      "learning_rate": 9.993467336683417e-05,
+      "loss": 1.5256,
+      "step": 113
+    },
+    {
+      "epoch": 0.009728103766440175,
+      "grad_norm": 0.486328125,
+      "learning_rate": 9.992964824120603e-05,
+      "loss": 1.7505,
+      "step": 114
+    },
+    {
+      "epoch": 0.00981343801000544,
+      "grad_norm": 0.462890625,
+      "learning_rate": 9.99246231155779e-05,
+      "loss": 1.534,
+      "step": 115
+    },
+    {
+      "epoch": 0.009898772253570704,
+      "grad_norm": 0.5546875,
+      "learning_rate": 9.991959798994976e-05,
+      "loss": 1.7212,
+      "step": 116
+    },
+    {
+      "epoch": 0.009984106497135969,
+      "grad_norm": 0.4609375,
+      "learning_rate": 9.99145728643216e-05,
+      "loss": 1.4882,
+      "step": 117
+    },
+    {
+      "epoch": 0.010069440740701234,
+      "grad_norm": 0.443359375,
+      "learning_rate": 9.990954773869348e-05,
+      "loss": 1.6496,
+      "step": 118
+    },
+    {
+      "epoch": 0.010154774984266498,
+      "grad_norm": 0.47265625,
+      "learning_rate": 9.990452261306533e-05,
+      "loss": 1.547,
+      "step": 119
+    },
+    {
+      "epoch": 0.010240109227831763,
+      "grad_norm": 0.47265625,
+      "learning_rate": 9.989949748743719e-05,
+      "loss": 1.548,
+      "step": 120
+    },
+    {
+      "epoch": 0.010325443471397027,
+      "grad_norm": 0.41015625,
+      "learning_rate": 9.989447236180905e-05,
+      "loss": 1.5738,
+      "step": 121
+    },
+    {
+      "epoch": 0.010410777714962294,
+      "grad_norm": 0.4296875,
+      "learning_rate": 9.988944723618092e-05,
+      "loss": 1.4623,
+      "step": 122
+    },
+    {
+      "epoch": 0.010496111958527558,
+      "grad_norm": 0.482421875,
+      "learning_rate": 9.988442211055276e-05,
+      "loss": 1.5758,
+      "step": 123
+    },
+    {
+      "epoch": 0.010581446202092823,
+      "grad_norm": 0.396484375,
+      "learning_rate": 9.987939698492463e-05,
+      "loss": 1.4766,
+      "step": 124
+    },
+    {
+      "epoch": 0.010666780445658088,
+      "grad_norm": 0.45703125,
+      "learning_rate": 9.987437185929649e-05,
+      "loss": 1.5325,
+      "step": 125
+    },
+    {
+      "epoch": 0.010752114689223352,
+      "grad_norm": 0.52734375,
+      "learning_rate": 9.986934673366835e-05,
+      "loss": 1.5734,
+      "step": 126
+    },
+    {
+      "epoch": 0.010837448932788617,
+      "grad_norm": 0.43359375,
+      "learning_rate": 9.98643216080402e-05,
+      "loss": 1.5174,
+      "step": 127
+    },
+    {
+      "epoch": 0.010922783176353881,
+      "grad_norm": 0.384765625,
+      "learning_rate": 9.985929648241207e-05,
+      "loss": 1.4156,
+      "step": 128
+    },
+    {
+      "epoch": 0.011008117419919146,
+      "grad_norm": 0.392578125,
+      "learning_rate": 9.985427135678392e-05,
+      "loss": 1.5875,
+      "step": 129
+    },
+    {
+      "epoch": 0.01109345166348441,
+      "grad_norm": 0.37109375,
+      "learning_rate": 9.984924623115578e-05,
+      "loss": 1.4329,
+      "step": 130
+    },
+    {
+      "epoch": 0.011178785907049675,
+      "grad_norm": 0.4453125,
+      "learning_rate": 9.984422110552765e-05,
+      "loss": 1.6382,
+      "step": 131
+    },
+    {
+      "epoch": 0.01126412015061494,
+      "grad_norm": 0.435546875,
+      "learning_rate": 9.983919597989951e-05,
+      "loss": 1.5618,
+      "step": 132
+    },
+    {
+      "epoch": 0.011349454394180204,
+      "grad_norm": 0.41796875,
+      "learning_rate": 9.983417085427136e-05,
+      "loss": 1.4175,
+      "step": 133
+    },
+    {
+      "epoch": 0.011434788637745469,
+      "grad_norm": 0.44921875,
+      "learning_rate": 9.982914572864322e-05,
+      "loss": 1.5475,
+      "step": 134
+    },
+    {
+      "epoch": 0.011520122881310733,
+      "grad_norm": 0.4765625,
+      "learning_rate": 9.982412060301508e-05,
+      "loss": 1.5458,
+      "step": 135
+    },
+    {
+      "epoch": 0.011605457124875998,
+      "grad_norm": 0.46875,
+      "learning_rate": 9.981909547738694e-05,
+      "loss": 1.5643,
+      "step": 136
+    },
+    {
+      "epoch": 0.011690791368441263,
+      "grad_norm": 0.44140625,
+      "learning_rate": 9.98140703517588e-05,
+      "loss": 1.7032,
+      "step": 137
+    },
+    {
+      "epoch": 0.011776125612006529,
+      "grad_norm": 0.384765625,
+      "learning_rate": 9.980904522613065e-05,
+      "loss": 1.4243,
+      "step": 138
+    },
+    {
+      "epoch": 0.011861459855571794,
+      "grad_norm": 0.392578125,
+      "learning_rate": 9.980402010050252e-05,
+      "loss": 1.4475,
+      "step": 139
+    },
+    {
+      "epoch": 0.011946794099137058,
+      "grad_norm": 0.427734375,
+      "learning_rate": 9.979899497487438e-05,
+      "loss": 1.5821,
+      "step": 140
+    },
+    {
+      "epoch": 0.012032128342702323,
+      "grad_norm": 0.435546875,
+      "learning_rate": 9.979396984924624e-05,
+      "loss": 1.5276,
+      "step": 141
+    },
+    {
+      "epoch": 0.012117462586267587,
+      "grad_norm": 0.4296875,
+      "learning_rate": 9.978894472361809e-05,
+      "loss": 1.5591,
+      "step": 142
+    },
+    {
+      "epoch": 0.012202796829832852,
+      "grad_norm": 0.392578125,
+      "learning_rate": 9.978391959798995e-05,
+      "loss": 1.3586,
+      "step": 143
+    },
+    {
+      "epoch": 0.012288131073398116,
+      "grad_norm": 0.45703125,
+      "learning_rate": 9.977889447236181e-05,
+      "loss": 1.506,
+      "step": 144
+    },
+    {
+      "epoch": 0.012373465316963381,
+      "grad_norm": 0.451171875,
+      "learning_rate": 9.977386934673367e-05,
+      "loss": 1.5435,
+      "step": 145
+    },
+    {
+      "epoch": 0.012458799560528646,
+      "grad_norm": 0.431640625,
+      "learning_rate": 9.976884422110552e-05,
+      "loss": 1.5242,
+      "step": 146
+    },
+    {
+      "epoch": 0.01254413380409391,
+      "grad_norm": 0.439453125,
+      "learning_rate": 9.97638190954774e-05,
+      "loss": 1.5617,
+      "step": 147
+    },
+    {
+      "epoch": 0.012629468047659175,
+      "grad_norm": 0.392578125,
+      "learning_rate": 9.975879396984925e-05,
+      "loss": 1.6439,
+      "step": 148
+    },
+    {
+      "epoch": 0.01271480229122444,
+      "grad_norm": 0.40625,
+      "learning_rate": 9.975376884422111e-05,
+      "loss": 1.4132,
+      "step": 149
+    },
+    {
+      "epoch": 0.012800136534789704,
+      "grad_norm": 0.40234375,
+      "learning_rate": 9.974874371859297e-05,
+      "loss": 1.5351,
+      "step": 150
+    },
+    {
+      "epoch": 0.012885470778354969,
+      "grad_norm": 0.396484375,
+      "learning_rate": 9.974371859296483e-05,
+      "loss": 1.5083,
+      "step": 151
+    },
+    {
+      "epoch": 0.012970805021920233,
+      "grad_norm": 0.41796875,
+      "learning_rate": 9.973869346733668e-05,
+      "loss": 1.3979,
+      "step": 152
+    },
+    {
+      "epoch": 0.013056139265485498,
+      "grad_norm": 0.43359375,
+      "learning_rate": 9.973366834170856e-05,
+      "loss": 1.4763,
+      "step": 153
+    },
+    {
+      "epoch": 0.013141473509050762,
+      "grad_norm": 0.427734375,
+      "learning_rate": 9.97286432160804e-05,
+      "loss": 1.3848,
+      "step": 154
+    },
+    {
+      "epoch": 0.013226807752616029,
+      "grad_norm": 0.37109375,
+      "learning_rate": 9.972361809045227e-05,
+      "loss": 1.4794,
+      "step": 155
+    },
+    {
+      "epoch": 0.013312141996181293,
+      "grad_norm": 0.392578125,
+      "learning_rate": 9.971859296482412e-05,
+      "loss": 1.6615,
+      "step": 156
+    },
+    {
+      "epoch": 0.013397476239746558,
+      "grad_norm": 0.404296875,
+      "learning_rate": 9.971356783919599e-05,
+      "loss": 1.459,
+      "step": 157
+    },
+    {
+      "epoch": 0.013482810483311822,
+      "grad_norm": 0.458984375,
+      "learning_rate": 9.970854271356784e-05,
+      "loss": 1.4856,
+      "step": 158
+    },
+    {
+      "epoch": 0.013568144726877087,
+      "grad_norm": 0.390625,
+      "learning_rate": 9.97035175879397e-05,
+      "loss": 1.5139,
+      "step": 159
+    },
+    {
+      "epoch": 0.013653478970442352,
+      "grad_norm": 0.455078125,
+      "learning_rate": 9.969849246231156e-05,
+      "loss": 1.4959,
+      "step": 160
+    },
+    {
+      "epoch": 0.013738813214007616,
+      "grad_norm": 0.400390625,
+      "learning_rate": 9.969346733668343e-05,
+      "loss": 1.5225,
+      "step": 161
+    },
+    {
+      "epoch": 0.01382414745757288,
+      "grad_norm": 0.5078125,
+      "learning_rate": 9.968844221105527e-05,
+      "loss": 1.4846,
+      "step": 162
+    },
+    {
+      "epoch": 0.013909481701138145,
+      "grad_norm": 0.369140625,
+      "learning_rate": 9.968341708542715e-05,
+      "loss": 1.461,
+      "step": 163
+    },
+    {
+      "epoch": 0.01399481594470341,
+      "grad_norm": 0.388671875,
+      "learning_rate": 9.9678391959799e-05,
+      "loss": 1.6441,
+      "step": 164
+    },
+    {
+      "epoch": 0.014080150188268675,
+      "grad_norm": 0.421875,
+      "learning_rate": 9.967336683417086e-05,
+      "loss": 1.5616,
+      "step": 165
+    },
+    {
+      "epoch": 0.01416548443183394,
+      "grad_norm": 0.37890625,
+      "learning_rate": 9.966834170854272e-05,
+      "loss": 1.4335,
+      "step": 166
+    },
+    {
+      "epoch": 0.014250818675399204,
+      "grad_norm": 0.3828125,
+      "learning_rate": 9.966331658291458e-05,
+      "loss": 1.3676,
+      "step": 167
+    },
+    {
+      "epoch": 0.014336152918964468,
+      "grad_norm": 0.4140625,
+      "learning_rate": 9.965829145728643e-05,
+      "loss": 1.6303,
+      "step": 168
+    },
+    {
+      "epoch": 0.014421487162529733,
+      "grad_norm": 0.380859375,
+      "learning_rate": 9.96532663316583e-05,
+      "loss": 1.5304,
+      "step": 169
+    },
+    {
+      "epoch": 0.014506821406094998,
+      "grad_norm": 0.400390625,
+      "learning_rate": 9.964824120603016e-05,
+      "loss": 1.5157,
+      "step": 170
+    },
+    {
+      "epoch": 0.014592155649660264,
+      "grad_norm": 0.455078125,
+      "learning_rate": 9.964321608040202e-05,
+      "loss": 1.5339,
+      "step": 171
+    },
+    {
+      "epoch": 0.014677489893225528,
+      "grad_norm": 0.375,
+      "learning_rate": 9.963819095477387e-05,
+      "loss": 1.4739,
+      "step": 172
+    },
+    {
+      "epoch": 0.014762824136790793,
+      "grad_norm": 0.4375,
+      "learning_rate": 9.963316582914573e-05,
+      "loss": 1.3935,
+      "step": 173
+    },
+    {
+      "epoch": 0.014848158380356058,
+      "grad_norm": 0.37890625,
+      "learning_rate": 9.962814070351759e-05,
+      "loss": 1.4655,
+      "step": 174
+    },
+    {
+      "epoch": 0.014933492623921322,
+      "grad_norm": 0.41796875,
+      "learning_rate": 9.962311557788945e-05,
+      "loss": 1.4473,
+      "step": 175
+    },
+    {
+      "epoch": 0.015018826867486587,
+      "grad_norm": 0.435546875,
+      "learning_rate": 9.961809045226132e-05,
+      "loss": 1.4698,
+      "step": 176
+    },
+    {
+      "epoch": 0.015104161111051851,
+      "grad_norm": 0.37109375,
+      "learning_rate": 9.961306532663316e-05,
+      "loss": 1.5137,
+      "step": 177
+    },
+    {
+      "epoch": 0.015189495354617116,
+      "grad_norm": 0.3671875,
+      "learning_rate": 9.960804020100503e-05,
+      "loss": 1.5024,
+      "step": 178
+    },
+    {
+      "epoch": 0.01527482959818238,
+      "grad_norm": 0.416015625,
+      "learning_rate": 9.960301507537689e-05,
+      "loss": 1.4906,
+      "step": 179
+    },
+    {
+      "epoch": 0.015360163841747645,
+      "grad_norm": 0.396484375,
+      "learning_rate": 9.959798994974875e-05,
+      "loss": 1.4285,
+      "step": 180
+    },
+    {
+      "epoch": 0.01544549808531291,
+      "grad_norm": 0.373046875,
+      "learning_rate": 9.95929648241206e-05,
+      "loss": 1.3331,
+      "step": 181
+    },
+    {
+      "epoch": 0.015530832328878174,
+      "grad_norm": 0.421875,
+      "learning_rate": 9.958793969849247e-05,
+      "loss": 1.5696,
+      "step": 182
+    },
+    {
+      "epoch": 0.015616166572443439,
+      "grad_norm": 0.4453125,
+      "learning_rate": 9.958291457286432e-05,
+      "loss": 1.7032,
+      "step": 183
+    },
+    {
+      "epoch": 0.015701500816008705,
+      "grad_norm": 0.396484375,
+      "learning_rate": 9.957788944723619e-05,
+      "loss": 1.4965,
+      "step": 184
+    },
+    {
+      "epoch": 0.01578683505957397,
+      "grad_norm": 0.3984375,
+      "learning_rate": 9.957286432160805e-05,
+      "loss": 1.4278,
+      "step": 185
+    },
+    {
+      "epoch": 0.015872169303139234,
+      "grad_norm": 0.353515625,
+      "learning_rate": 9.956783919597991e-05,
+      "loss": 1.3238,
+      "step": 186
+    },
+    {
+      "epoch": 0.0159575035467045,
+      "grad_norm": 0.4296875,
+      "learning_rate": 9.956281407035176e-05,
+      "loss": 1.5803,
+      "step": 187
+    },
+    {
+      "epoch": 0.016042837790269764,
+      "grad_norm": 0.396484375,
+      "learning_rate": 9.955778894472362e-05,
+      "loss": 1.5097,
+      "step": 188
+    },
+    {
+      "epoch": 0.016128172033835028,
+      "grad_norm": 0.376953125,
+      "learning_rate": 9.955276381909548e-05,
+      "loss": 1.5302,
+      "step": 189
+    },
+    {
+      "epoch": 0.016213506277400293,
+      "grad_norm": 0.421875,
+      "learning_rate": 9.954773869346734e-05,
+      "loss": 1.4605,
+      "step": 190
+    },
+    {
+      "epoch": 0.016298840520965557,
+      "grad_norm": 0.431640625,
+      "learning_rate": 9.954271356783919e-05,
+      "loss": 1.4818,
+      "step": 191
+    },
+    {
+      "epoch": 0.016384174764530822,
+      "grad_norm": 0.44921875,
+      "learning_rate": 9.953768844221107e-05,
+      "loss": 1.5929,
+      "step": 192
+    },
+    {
+      "epoch": 0.016469509008096087,
+      "grad_norm": 0.373046875,
+      "learning_rate": 9.953266331658292e-05,
+      "loss": 1.6972,
+      "step": 193
+    },
+    {
+      "epoch": 0.01655484325166135,
+      "grad_norm": 0.423828125,
+      "learning_rate": 9.952763819095478e-05,
+      "loss": 1.7035,
+      "step": 194
+    },
+    {
+      "epoch": 0.016640177495226616,
+      "grad_norm": 0.4140625,
+      "learning_rate": 9.952261306532664e-05,
+      "loss": 1.522,
+      "step": 195
+    },
+    {
+      "epoch": 0.01672551173879188,
+      "grad_norm": 0.40234375,
+      "learning_rate": 9.95175879396985e-05,
+      "loss": 1.5499,
+      "step": 196
+    },
+    {
+      "epoch": 0.016810845982357145,
+      "grad_norm": 0.396484375,
+      "learning_rate": 9.951256281407035e-05,
+      "loss": 1.3498,
+      "step": 197
+    },
+    {
+      "epoch": 0.01689618022592241,
+      "grad_norm": 0.423828125,
+      "learning_rate": 9.950753768844223e-05,
+      "loss": 1.4946,
+      "step": 198
+    },
+    {
+      "epoch": 0.016981514469487674,
+      "grad_norm": 0.412109375,
+      "learning_rate": 9.950251256281408e-05,
+      "loss": 1.4106,
+      "step": 199
+    },
+    {
+      "epoch": 0.01706684871305294,
+      "grad_norm": 0.34375,
+      "learning_rate": 9.949748743718594e-05,
+      "loss": 1.3061,
+      "step": 200
+    },
+    {
+      "epoch": 0.017152182956618203,
+      "grad_norm": 0.3671875,
+      "learning_rate": 9.94924623115578e-05,
+      "loss": 1.6192,
+      "step": 201
+    },
+    {
+      "epoch": 0.017237517200183468,
+      "grad_norm": 0.376953125,
+      "learning_rate": 9.948743718592966e-05,
+      "loss": 1.6002,
+      "step": 202
+    },
+    {
+      "epoch": 0.017322851443748732,
+      "grad_norm": 0.44921875,
+      "learning_rate": 9.948241206030151e-05,
+      "loss": 1.4208,
+      "step": 203
+    },
+    {
+      "epoch": 0.017408185687313997,
+      "grad_norm": 0.359375,
+      "learning_rate": 9.947738693467337e-05,
+      "loss": 1.4471,
+      "step": 204
+    },
+    {
+      "epoch": 0.01749351993087926,
+      "grad_norm": 0.3828125,
+      "learning_rate": 9.947236180904523e-05,
+      "loss": 1.5423,
+      "step": 205
+    },
+    {
+      "epoch": 0.017578854174444526,
+      "grad_norm": 0.390625,
+      "learning_rate": 9.94673366834171e-05,
+      "loss": 1.4858,
+      "step": 206
+    },
+    {
+      "epoch": 0.01766418841800979,
+      "grad_norm": 0.369140625,
+      "learning_rate": 9.946231155778894e-05,
+      "loss": 1.492,
+      "step": 207
+    },
+    {
+      "epoch": 0.017749522661575055,
+      "grad_norm": 0.35546875,
+      "learning_rate": 9.94572864321608e-05,
+      "loss": 1.4228,
+      "step": 208
+    },
+    {
+      "epoch": 0.01783485690514032,
+      "grad_norm": 0.359375,
+      "learning_rate": 9.945226130653267e-05,
+      "loss": 1.3717,
+      "step": 209
+    },
+    {
+      "epoch": 0.017920191148705585,
+      "grad_norm": 0.36328125,
+      "learning_rate": 9.944723618090453e-05,
+      "loss": 1.4904,
+      "step": 210
+    },
+    {
+      "epoch": 0.018005525392270853,
+      "grad_norm": 0.34375,
+      "learning_rate": 9.944221105527639e-05,
+      "loss": 1.3978,
+      "step": 211
+    },
+    {
+      "epoch": 0.018090859635836117,
+      "grad_norm": 0.41015625,
+      "learning_rate": 9.943718592964824e-05,
+      "loss": 1.4583,
+      "step": 212
+    },
+    {
+      "epoch": 0.018176193879401382,
+      "grad_norm": 0.36328125,
+      "learning_rate": 9.94321608040201e-05,
+      "loss": 1.5892,
+      "step": 213
+    },
+    {
+      "epoch": 0.018261528122966646,
+      "grad_norm": 0.37109375,
+      "learning_rate": 9.942713567839197e-05,
+      "loss": 1.5942,
+      "step": 214
+    },
+    {
+      "epoch": 0.01834686236653191,
+      "grad_norm": 0.35546875,
+      "learning_rate": 9.942211055276383e-05,
+      "loss": 1.5318,
+      "step": 215
+    },
+    {
+      "epoch": 0.018432196610097176,
+      "grad_norm": 0.359375,
+      "learning_rate": 9.941708542713568e-05,
+      "loss": 1.4243,
+      "step": 216
+    },
+    {
+      "epoch": 0.01851753085366244,
+      "grad_norm": 0.392578125,
+      "learning_rate": 9.941206030150754e-05,
+      "loss": 1.3628,
+      "step": 217
+    },
+    {
+      "epoch": 0.018602865097227705,
+      "grad_norm": 0.3359375,
+      "learning_rate": 9.94070351758794e-05,
+      "loss": 1.4799,
+      "step": 218
+    },
+    {
+      "epoch": 0.01868819934079297,
+      "grad_norm": 0.34765625,
+      "learning_rate": 9.940201005025126e-05,
+      "loss": 1.5227,
+      "step": 219
+    },
+    {
+      "epoch": 0.018773533584358234,
+      "grad_norm": 0.341796875,
+      "learning_rate": 9.939698492462311e-05,
+      "loss": 1.436,
+      "step": 220
+    },
+    {
+      "epoch": 0.0188588678279235,
+      "grad_norm": 0.3125,
+      "learning_rate": 9.939195979899499e-05,
+      "loss": 1.4322,
+      "step": 221
+    },
+    {
+      "epoch": 0.018944202071488763,
+      "grad_norm": 0.373046875,
+      "learning_rate": 9.938693467336683e-05,
+      "loss": 1.5913,
+      "step": 222
+    },
+    {
+      "epoch": 0.019029536315054028,
+      "grad_norm": 0.373046875,
+      "learning_rate": 9.93819095477387e-05,
+      "loss": 1.5052,
+      "step": 223
+    },
+    {
+      "epoch": 0.019114870558619292,
+      "grad_norm": 0.390625,
+      "learning_rate": 9.937688442211056e-05,
+      "loss": 1.6388,
+      "step": 224
+    },
+    {
+      "epoch": 0.019200204802184557,
+      "grad_norm": 0.416015625,
+      "learning_rate": 9.937185929648242e-05,
+      "loss": 1.5977,
+      "step": 225
+    },
+    {
+      "epoch": 0.01928553904574982,
+      "grad_norm": 0.400390625,
+      "learning_rate": 9.936683417085427e-05,
+      "loss": 1.4548,
+      "step": 226
+    },
+    {
+      "epoch": 0.019370873289315086,
+      "grad_norm": 0.369140625,
+      "learning_rate": 9.936180904522614e-05,
+      "loss": 1.4918,
+      "step": 227
+    },
+    {
+      "epoch": 0.01945620753288035,
+      "grad_norm": 0.4296875,
+      "learning_rate": 9.935678391959799e-05,
+      "loss": 1.5212,
+      "step": 228
+    },
+    {
+      "epoch": 0.019541541776445615,
+      "grad_norm": 0.5234375,
+      "learning_rate": 9.935175879396985e-05,
+      "loss": 1.5645,
+      "step": 229
+    },
+    {
+      "epoch": 0.01962687602001088,
+      "grad_norm": 0.37109375,
+      "learning_rate": 9.934673366834172e-05,
+      "loss": 1.4794,
+      "step": 230
+    },
+    {
+      "epoch": 0.019712210263576144,
+      "grad_norm": 0.4296875,
+      "learning_rate": 9.934170854271358e-05,
+      "loss": 1.4586,
+      "step": 231
+    },
+    {
+      "epoch": 0.01979754450714141,
+      "grad_norm": 0.369140625,
+      "learning_rate": 9.933668341708543e-05,
+      "loss": 1.5006,
+      "step": 232
+    },
+    {
+      "epoch": 0.019882878750706674,
+      "grad_norm": 0.388671875,
+      "learning_rate": 9.933165829145729e-05,
+      "loss": 1.5719,
+      "step": 233
+    },
+    {
+      "epoch": 0.019968212994271938,
+      "grad_norm": 0.37109375,
+      "learning_rate": 9.932663316582915e-05,
+      "loss": 1.3318,
+      "step": 234
+    },
+    {
+      "epoch": 0.020053547237837203,
+      "grad_norm": 0.388671875,
+      "learning_rate": 9.932160804020101e-05,
+      "loss": 1.5506,
+      "step": 235
+    },
+    {
+      "epoch": 0.020138881481402467,
+      "grad_norm": 0.357421875,
+      "learning_rate": 9.931658291457286e-05,
+      "loss": 1.4031,
+      "step": 236
+    },
+    {
+      "epoch": 0.020224215724967732,
+      "grad_norm": 0.3828125,
+      "learning_rate": 9.931155778894474e-05,
+      "loss": 1.5384,
+      "step": 237
+    },
+    {
+      "epoch": 0.020309549968532997,
+      "grad_norm": 0.3515625,
+      "learning_rate": 9.930653266331659e-05,
+      "loss": 1.3319,
+      "step": 238
+    },
+    {
+      "epoch": 0.02039488421209826,
+      "grad_norm": 0.39453125,
+      "learning_rate": 9.930150753768845e-05,
+      "loss": 1.441,
+      "step": 239
+    },
+    {
+      "epoch": 0.020480218455663526,
+      "grad_norm": 0.353515625,
+      "learning_rate": 9.929648241206031e-05,
+      "loss": 1.4311,
+      "step": 240
+    },
+    {
+      "epoch": 0.02056555269922879,
+      "grad_norm": 0.380859375,
+      "learning_rate": 9.929145728643217e-05,
+      "loss": 1.4405,
+      "step": 241
+    },
+    {
+      "epoch": 0.020650886942794055,
+      "grad_norm": 0.322265625,
+      "learning_rate": 9.928643216080402e-05,
+      "loss": 1.3873,
+      "step": 242
+    },
+    {
+      "epoch": 0.02073622118635932,
+      "grad_norm": 0.380859375,
+      "learning_rate": 9.928140703517588e-05,
+      "loss": 1.6589,
+      "step": 243
+    },
+    {
+      "epoch": 0.020821555429924588,
+      "grad_norm": 0.396484375,
+      "learning_rate": 9.927638190954774e-05,
+      "loss": 1.4386,
+      "step": 244
+    },
+    {
+      "epoch": 0.020906889673489852,
+      "grad_norm": 0.384765625,
+      "learning_rate": 9.927135678391961e-05,
+      "loss": 1.4542,
+      "step": 245
+    },
+    {
+      "epoch": 0.020992223917055117,
+      "grad_norm": 0.36328125,
+      "learning_rate": 9.926633165829147e-05,
+      "loss": 1.4885,
+      "step": 246
+    },
+    {
+      "epoch": 0.02107755816062038,
+      "grad_norm": 0.39453125,
+      "learning_rate": 9.926130653266332e-05,
+      "loss": 1.5343,
+      "step": 247
+    },
+    {
+      "epoch": 0.021162892404185646,
+      "grad_norm": 0.38671875,
+      "learning_rate": 9.925628140703518e-05,
+      "loss": 1.4289,
+      "step": 248
+    },
+    {
+      "epoch": 0.02124822664775091,
+      "grad_norm": 0.3515625,
+      "learning_rate": 9.925125628140703e-05,
+      "loss": 1.5846,
+      "step": 249
+    },
+    {
+      "epoch": 0.021333560891316175,
+      "grad_norm": 0.40625,
+      "learning_rate": 9.92462311557789e-05,
+      "loss": 1.308,
+      "step": 250
+    },
+    {
+      "epoch": 0.02141889513488144,
+      "grad_norm": 0.34375,
+      "learning_rate": 9.924120603015075e-05,
+      "loss": 1.6103,
+      "step": 251
+    },
+    {
+      "epoch": 0.021504229378446704,
+      "grad_norm": 0.365234375,
+      "learning_rate": 9.923618090452261e-05,
+      "loss": 1.5454,
+      "step": 252
+    },
+    {
+      "epoch": 0.02158956362201197,
+      "grad_norm": 0.380859375,
+      "learning_rate": 9.923115577889448e-05,
+      "loss": 1.5327,
+      "step": 253
+    },
+    {
+      "epoch": 0.021674897865577233,
+      "grad_norm": 0.3828125,
+      "learning_rate": 9.922613065326634e-05,
+      "loss": 1.4348,
+      "step": 254
+    },
+    {
+      "epoch": 0.021760232109142498,
+      "grad_norm": 0.341796875,
+      "learning_rate": 9.922110552763819e-05,
+      "loss": 1.5543,
+      "step": 255
+    },
+    {
+      "epoch": 0.021845566352707763,
+      "grad_norm": 0.39453125,
+      "learning_rate": 9.921608040201006e-05,
+      "loss": 1.4348,
+      "step": 256
+    },
+    {
+      "epoch": 0.021930900596273027,
+      "grad_norm": 0.34375,
+      "learning_rate": 9.921105527638191e-05,
+      "loss": 1.3951,
+      "step": 257
+    },
+    {
+      "epoch": 0.022016234839838292,
+      "grad_norm": 0.3671875,
+      "learning_rate": 9.920603015075377e-05,
+      "loss": 1.3103,
+      "step": 258
+    },
+    {
+      "epoch": 0.022101569083403556,
+      "grad_norm": 0.353515625,
+      "learning_rate": 9.920100502512563e-05,
+      "loss": 1.3765,
+      "step": 259
+    },
+    {
+      "epoch": 0.02218690332696882,
+      "grad_norm": 0.341796875,
+      "learning_rate": 9.91959798994975e-05,
+      "loss": 1.3495,
+      "step": 260
+    },
+    {
+      "epoch": 0.022272237570534086,
+      "grad_norm": 0.3515625,
+      "learning_rate": 9.919095477386935e-05,
+      "loss": 1.3266,
+      "step": 261
+    },
+    {
+      "epoch": 0.02235757181409935,
+      "grad_norm": 0.30078125,
+      "learning_rate": 9.918592964824122e-05,
+      "loss": 1.2198,
+      "step": 262
+    },
+    {
+      "epoch": 0.022442906057664615,
+      "grad_norm": 0.32421875,
+      "learning_rate": 9.918090452261307e-05,
+      "loss": 1.419,
+      "step": 263
+    },
+    {
+      "epoch": 0.02252824030122988,
+      "grad_norm": 0.34765625,
+      "learning_rate": 9.917587939698493e-05,
+      "loss": 1.341,
+      "step": 264
+    },
+    {
+      "epoch": 0.022613574544795144,
+      "grad_norm": 0.30859375,
+      "learning_rate": 9.917085427135678e-05,
+      "loss": 1.323,
+      "step": 265
+    },
+    {
+      "epoch": 0.02269890878836041,
+      "grad_norm": 0.37890625,
+      "learning_rate": 9.916582914572866e-05,
+      "loss": 1.3533,
+      "step": 266
+    },
+    {
+      "epoch": 0.022784243031925673,
+      "grad_norm": 0.3515625,
+      "learning_rate": 9.91608040201005e-05,
+      "loss": 1.3956,
+      "step": 267
+    },
+    {
+      "epoch": 0.022869577275490938,
+      "grad_norm": 0.41015625,
+      "learning_rate": 9.915577889447237e-05,
+      "loss": 1.5848,
+      "step": 268
+    },
+    {
+      "epoch": 0.022954911519056202,
+      "grad_norm": 0.40625,
+      "learning_rate": 9.915075376884423e-05,
+      "loss": 1.499,
+      "step": 269
+    },
+    {
+      "epoch": 0.023040245762621467,
+      "grad_norm": 0.3671875,
+      "learning_rate": 9.914572864321609e-05,
+      "loss": 1.4106,
+      "step": 270
+    },
+    {
+      "epoch": 0.02312558000618673,
+      "grad_norm": 0.37109375,
+      "learning_rate": 9.914070351758794e-05,
+      "loss": 1.4622,
+      "step": 271
+    },
+    {
+      "epoch": 0.023210914249751996,
+      "grad_norm": 0.359375,
+      "learning_rate": 9.913567839195981e-05,
+      "loss": 1.3194,
+      "step": 272
+    },
+    {
+      "epoch": 0.02329624849331726,
+      "grad_norm": 0.37890625,
+      "learning_rate": 9.913065326633166e-05,
+      "loss": 1.432,
+      "step": 273
+    },
+    {
+      "epoch": 0.023381582736882525,
+      "grad_norm": 0.376953125,
+      "learning_rate": 9.912562814070352e-05,
+      "loss": 1.4197,
+      "step": 274
+    },
+    {
+      "epoch": 0.02346691698044779,
+      "grad_norm": 0.345703125,
+      "learning_rate": 9.912060301507539e-05,
+      "loss": 1.3807,
+      "step": 275
+    },
+    {
+      "epoch": 0.023552251224013058,
+      "grad_norm": 0.380859375,
+      "learning_rate": 9.911557788944725e-05,
+      "loss": 1.4024,
+      "step": 276
+    },
+    {
+      "epoch": 0.023637585467578322,
+      "grad_norm": 0.380859375,
+      "learning_rate": 9.91105527638191e-05,
+      "loss": 1.4863,
+      "step": 277
+    },
+    {
+      "epoch": 0.023722919711143587,
+      "grad_norm": 0.375,
+      "learning_rate": 9.910552763819096e-05,
+      "loss": 1.314,
+      "step": 278
+    },
+    {
+      "epoch": 0.02380825395470885,
+      "grad_norm": 0.375,
+      "learning_rate": 9.910050251256282e-05,
+      "loss": 1.4456,
+      "step": 279
+    },
+    {
+      "epoch": 0.023893588198274116,
+      "grad_norm": 0.333984375,
+      "learning_rate": 9.909547738693468e-05,
+      "loss": 1.5877,
+      "step": 280
+    },
+    {
+      "epoch": 0.02397892244183938,
+      "grad_norm": 0.400390625,
+      "learning_rate": 9.909045226130653e-05,
+      "loss": 1.5869,
+      "step": 281
+    },
+    {
+      "epoch": 0.024064256685404645,
+      "grad_norm": 0.357421875,
+      "learning_rate": 9.90854271356784e-05,
+      "loss": 1.5391,
+      "step": 282
+    },
+    {
+      "epoch": 0.02414959092896991,
+      "grad_norm": 0.322265625,
+      "learning_rate": 9.908040201005026e-05,
+      "loss": 1.3901,
+      "step": 283
+    },
+    {
+      "epoch": 0.024234925172535175,
+      "grad_norm": 0.345703125,
+      "learning_rate": 9.90753768844221e-05,
+      "loss": 1.4891,
+      "step": 284
+    },
+    {
+      "epoch": 0.02432025941610044,
+      "grad_norm": 0.306640625,
+      "learning_rate": 9.907035175879398e-05,
+      "loss": 1.3531,
+      "step": 285
+    },
+    {
+      "epoch": 0.024405593659665704,
+      "grad_norm": 0.421875,
+      "learning_rate": 9.906532663316583e-05,
+      "loss": 1.3986,
+      "step": 286
+    },
+    {
+      "epoch": 0.02449092790323097,
+      "grad_norm": 0.34765625,
+      "learning_rate": 9.906030150753769e-05,
+      "loss": 1.5053,
+      "step": 287
+    },
+    {
+      "epoch": 0.024576262146796233,
+      "grad_norm": 0.361328125,
+      "learning_rate": 9.905527638190955e-05,
+      "loss": 1.4861,
+      "step": 288
+    },
+    {
+      "epoch": 0.024661596390361498,
+      "grad_norm": 0.361328125,
+      "learning_rate": 9.905025125628141e-05,
+      "loss": 1.3849,
+      "step": 289
+    },
+    {
+      "epoch": 0.024746930633926762,
+      "grad_norm": 0.357421875,
+      "learning_rate": 9.904522613065326e-05,
+      "loss": 1.4285,
+      "step": 290
+    },
+    {
+      "epoch": 0.024832264877492027,
+      "grad_norm": 0.341796875,
+      "learning_rate": 9.904020100502514e-05,
+      "loss": 1.3654,
+      "step": 291
+    },
+    {
+      "epoch": 0.02491759912105729,
+      "grad_norm": 0.357421875,
+      "learning_rate": 9.903517587939699e-05,
+      "loss": 1.453,
+      "step": 292
+    },
+    {
+      "epoch": 0.025002933364622556,
+      "grad_norm": 0.3515625,
+      "learning_rate": 9.903015075376885e-05,
+      "loss": 1.4705,
+      "step": 293
+    },
+    {
+      "epoch": 0.02508826760818782,
+      "grad_norm": 0.353515625,
+      "learning_rate": 9.902512562814071e-05,
+      "loss": 1.4251,
+      "step": 294
+    },
+    {
+      "epoch": 0.025173601851753085,
+      "grad_norm": 0.310546875,
+      "learning_rate": 9.902010050251257e-05,
+      "loss": 1.2653,
+      "step": 295
+    },
+    {
+      "epoch": 0.02525893609531835,
+      "grad_norm": 0.34765625,
+      "learning_rate": 9.901507537688442e-05,
+      "loss": 1.4037,
+      "step": 296
+    },
+    {
+      "epoch": 0.025344270338883614,
+      "grad_norm": 0.345703125,
+      "learning_rate": 9.901005025125628e-05,
+      "loss": 1.522,
+      "step": 297
+    },
+    {
+      "epoch": 0.02542960458244888,
+      "grad_norm": 0.369140625,
+      "learning_rate": 9.900502512562815e-05,
+      "loss": 1.2837,
+      "step": 298
+    },
+    {
+      "epoch": 0.025514938826014143,
+      "grad_norm": 0.33984375,
+      "learning_rate": 9.900000000000001e-05,
+      "loss": 1.3896,
+      "step": 299
+    },
+    {
+      "epoch": 0.025600273069579408,
+      "grad_norm": 0.3984375,
+      "learning_rate": 9.899497487437186e-05,
+      "loss": 1.5719,
+      "step": 300
+    },
+    {
+      "epoch": 0.025685607313144673,
+      "grad_norm": 0.4296875,
+      "learning_rate": 9.898994974874373e-05,
+      "loss": 1.6433,
+      "step": 301
+    },
+    {
+      "epoch": 0.025770941556709937,
+      "grad_norm": 0.3984375,
+      "learning_rate": 9.898492462311558e-05,
+      "loss": 1.4654,
+      "step": 302
+    },
+    {
+      "epoch": 0.025856275800275202,
+      "grad_norm": 0.34375,
+      "learning_rate": 9.897989949748744e-05,
+      "loss": 1.5617,
+      "step": 303
+    },
+    {
+      "epoch": 0.025941610043840466,
+      "grad_norm": 0.3671875,
+      "learning_rate": 9.89748743718593e-05,
+      "loss": 1.3666,
+      "step": 304
+    },
+    {
+      "epoch": 0.02602694428740573,
+      "grad_norm": 0.423828125,
+      "learning_rate": 9.896984924623117e-05,
+      "loss": 1.3791,
+      "step": 305
+    },
+    {
+      "epoch": 0.026112278530970996,
+      "grad_norm": 0.326171875,
+      "learning_rate": 9.896482412060301e-05,
+      "loss": 1.9473,
+      "step": 306
+    },
+    {
+      "epoch": 0.02619761277453626,
+      "grad_norm": 0.400390625,
+      "learning_rate": 9.895979899497489e-05,
+      "loss": 1.3326,
+      "step": 307
+    },
+    {
+      "epoch": 0.026282947018101525,
+      "grad_norm": 0.341796875,
+      "learning_rate": 9.895477386934674e-05,
+      "loss": 1.4549,
+      "step": 308
+    },
+    {
+      "epoch": 0.026368281261666793,
+      "grad_norm": 0.357421875,
+      "learning_rate": 9.89497487437186e-05,
+      "loss": 1.3408,
+      "step": 309
+    },
+    {
+      "epoch": 0.026453615505232057,
+      "grad_norm": 0.375,
+      "learning_rate": 9.894472361809046e-05,
+      "loss": 1.4272,
+      "step": 310
+    },
+    {
+      "epoch": 0.026538949748797322,
+      "grad_norm": 0.322265625,
+      "learning_rate": 9.893969849246232e-05,
+      "loss": 1.3771,
+      "step": 311
+    },
+    {
+      "epoch": 0.026624283992362587,
+      "grad_norm": 0.384765625,
+      "learning_rate": 9.893467336683417e-05,
+      "loss": 1.406,
+      "step": 312
+    },
+    {
+      "epoch": 0.02670961823592785,
+      "grad_norm": 0.32421875,
+      "learning_rate": 9.892964824120604e-05,
+      "loss": 1.2111,
+      "step": 313
+    },
+    {
+      "epoch": 0.026794952479493116,
+      "grad_norm": 0.50390625,
+      "learning_rate": 9.89246231155779e-05,
+      "loss": 1.44,
+      "step": 314
+    },
+    {
+      "epoch": 0.02688028672305838,
+      "grad_norm": 0.34765625,
+      "learning_rate": 9.891959798994975e-05,
+      "loss": 1.3375,
+      "step": 315
+    },
+    {
+      "epoch": 0.026965620966623645,
+      "grad_norm": 0.4453125,
+      "learning_rate": 9.891457286432161e-05,
+      "loss": 1.3964,
+      "step": 316
+    },
+    {
+      "epoch": 0.02705095521018891,
+      "grad_norm": 0.40234375,
+      "learning_rate": 9.890954773869347e-05,
+      "loss": 1.2975,
+      "step": 317
+    },
+    {
+      "epoch": 0.027136289453754174,
+      "grad_norm": 0.43359375,
+      "learning_rate": 9.890452261306533e-05,
+      "loss": 1.5118,
+      "step": 318
+    },
+    {
+      "epoch": 0.02722162369731944,
+      "grad_norm": 0.322265625,
+      "learning_rate": 9.889949748743718e-05,
+      "loss": 1.4032,
+      "step": 319
+    },
+    {
+      "epoch": 0.027306957940884703,
+      "grad_norm": 0.330078125,
+      "learning_rate": 9.889447236180906e-05,
+      "loss": 1.355,
+      "step": 320
+    },
+    {
+      "epoch": 0.027392292184449968,
+      "grad_norm": 0.328125,
+      "learning_rate": 9.88894472361809e-05,
+      "loss": 1.4382,
+      "step": 321
+    },
+    {
+      "epoch": 0.027477626428015232,
+      "grad_norm": 0.322265625,
+      "learning_rate": 9.888442211055277e-05,
+      "loss": 1.3795,
+      "step": 322
+    },
+    {
+      "epoch": 0.027562960671580497,
+      "grad_norm": 0.37109375,
+      "learning_rate": 9.887939698492463e-05,
+      "loss": 1.4173,
+      "step": 323
+    },
+    {
+      "epoch": 0.02764829491514576,
+      "grad_norm": 0.37109375,
+      "learning_rate": 9.887437185929649e-05,
+      "loss": 1.6702,
+      "step": 324
+    },
+    {
+      "epoch": 0.027733629158711026,
+      "grad_norm": 0.341796875,
+      "learning_rate": 9.886934673366834e-05,
+      "loss": 1.5437,
+      "step": 325
+    },
+    {
+      "epoch": 0.02781896340227629,
+      "grad_norm": 0.357421875,
+      "learning_rate": 9.886432160804021e-05,
+      "loss": 1.5578,
+      "step": 326
+    },
+    {
+      "epoch": 0.027904297645841555,
+      "grad_norm": 0.45703125,
+      "learning_rate": 9.885929648241206e-05,
+      "loss": 1.5042,
+      "step": 327
+    },
+    {
+      "epoch": 0.02798963188940682,
+      "grad_norm": 0.349609375,
+      "learning_rate": 9.885427135678393e-05,
+      "loss": 1.5055,
+      "step": 328
+    },
+    {
+      "epoch": 0.028074966132972085,
+      "grad_norm": 0.34765625,
+      "learning_rate": 9.884924623115577e-05,
+      "loss": 1.4351,
+      "step": 329
+    },
+    {
+      "epoch": 0.02816030037653735,
+      "grad_norm": 0.326171875,
+      "learning_rate": 9.884422110552765e-05,
+      "loss": 1.2851,
+      "step": 330
+    },
+    {
+      "epoch": 0.028245634620102614,
+      "grad_norm": 0.39453125,
+      "learning_rate": 9.88391959798995e-05,
+      "loss": 1.5372,
+      "step": 331
+    },
+    {
+      "epoch": 0.02833096886366788,
+      "grad_norm": 0.35546875,
+      "learning_rate": 9.883417085427136e-05,
+      "loss": 1.4504,
+      "step": 332
+    },
+    {
+      "epoch": 0.028416303107233143,
+      "grad_norm": 0.361328125,
+      "learning_rate": 9.882914572864322e-05,
+      "loss": 1.4873,
+      "step": 333
+    },
+    {
+      "epoch": 0.028501637350798407,
+      "grad_norm": 0.34765625,
+      "learning_rate": 9.882412060301508e-05,
+      "loss": 1.5475,
+      "step": 334
+    },
+    {
+      "epoch": 0.028586971594363672,
+      "grad_norm": 0.349609375,
+      "learning_rate": 9.881909547738693e-05,
+      "loss": 1.3558,
+      "step": 335
+    },
+    {
+      "epoch": 0.028672305837928937,
+      "grad_norm": 0.33203125,
+      "learning_rate": 9.881407035175881e-05,
+      "loss": 1.441,
+      "step": 336
+    },
+    {
+      "epoch": 0.0287576400814942,
+      "grad_norm": 0.337890625,
+      "learning_rate": 9.880904522613066e-05,
+      "loss": 1.4192,
+      "step": 337
+    },
+    {
+      "epoch": 0.028842974325059466,
+      "grad_norm": 0.37109375,
+      "learning_rate": 9.880402010050252e-05,
+      "loss": 1.4601,
+      "step": 338
+    },
+    {
+      "epoch": 0.02892830856862473,
+      "grad_norm": 0.37109375,
+      "learning_rate": 9.879899497487438e-05,
+      "loss": 1.4178,
+      "step": 339
+    },
+    {
+      "epoch": 0.029013642812189995,
+      "grad_norm": 0.318359375,
+      "learning_rate": 9.879396984924624e-05,
+      "loss": 1.416,
+      "step": 340
+    },
+    {
+      "epoch": 0.029098977055755263,
+      "grad_norm": 0.50390625,
+      "learning_rate": 9.878894472361809e-05,
+      "loss": 1.3266,
+      "step": 341
+    },
+    {
+      "epoch": 0.029184311299320528,
+      "grad_norm": 0.373046875,
+      "learning_rate": 9.878391959798995e-05,
+      "loss": 1.4582,
+      "step": 342
+    },
+    {
+      "epoch": 0.029269645542885792,
+      "grad_norm": 0.34375,
+      "learning_rate": 9.877889447236182e-05,
+      "loss": 1.461,
+      "step": 343
+    },
+    {
+      "epoch": 0.029354979786451057,
+      "grad_norm": 0.3359375,
+      "learning_rate": 9.877386934673368e-05,
+      "loss": 1.4722,
+      "step": 344
+    },
+    {
+      "epoch": 0.02944031403001632,
+      "grad_norm": 0.33203125,
+      "learning_rate": 9.876884422110553e-05,
+      "loss": 1.2707,
+      "step": 345
+    },
+    {
+      "epoch": 0.029525648273581586,
+      "grad_norm": 0.32421875,
+      "learning_rate": 9.87638190954774e-05,
+      "loss": 1.2089,
+      "step": 346
+    },
+    {
+      "epoch": 0.02961098251714685,
+      "grad_norm": 0.384765625,
+      "learning_rate": 9.875879396984925e-05,
+      "loss": 1.5013,
+      "step": 347
+    },
+    {
+      "epoch": 0.029696316760712115,
+      "grad_norm": 0.34375,
+      "learning_rate": 9.875376884422111e-05,
+      "loss": 1.46,
+      "step": 348
+    },
+    {
+      "epoch": 0.02978165100427738,
+      "grad_norm": 0.361328125,
+      "learning_rate": 9.874874371859297e-05,
+      "loss": 1.527,
+      "step": 349
+    },
+    {
+      "epoch": 0.029866985247842644,
+      "grad_norm": 0.35546875,
+      "learning_rate": 9.874371859296482e-05,
+      "loss": 1.4721,
+      "step": 350
+    },
+    {
+      "epoch": 0.02995231949140791,
+      "grad_norm": 0.33984375,
+      "learning_rate": 9.873869346733668e-05,
+      "loss": 1.4308,
+      "step": 351
+    },
+    {
+      "epoch": 0.030037653734973174,
+      "grad_norm": 0.34375,
+      "learning_rate": 9.873366834170855e-05,
+      "loss": 1.423,
+      "step": 352
+    },
+    {
+      "epoch": 0.030122987978538438,
+      "grad_norm": 0.361328125,
+      "learning_rate": 9.872864321608041e-05,
+      "loss": 1.3206,
+      "step": 353
+    },
+    {
+      "epoch": 0.030208322222103703,
+      "grad_norm": 0.3671875,
+      "learning_rate": 9.872361809045226e-05,
+      "loss": 1.5606,
+      "step": 354
+    },
+    {
+      "epoch": 0.030293656465668967,
+      "grad_norm": 0.373046875,
+      "learning_rate": 9.871859296482413e-05,
+      "loss": 1.4185,
+      "step": 355
+    },
+    {
+      "epoch": 0.030378990709234232,
+      "grad_norm": 0.357421875,
+      "learning_rate": 9.871356783919598e-05,
+      "loss": 1.3768,
+      "step": 356
+    },
+    {
+      "epoch": 0.030464324952799497,
+      "grad_norm": 0.341796875,
+      "learning_rate": 9.870854271356784e-05,
+      "loss": 1.4831,
+      "step": 357
+    },
+    {
+      "epoch": 0.03054965919636476,
+      "grad_norm": 0.318359375,
+      "learning_rate": 9.870351758793969e-05,
+      "loss": 1.4651,
+      "step": 358
+    },
+    {
+      "epoch": 0.030634993439930026,
+      "grad_norm": 0.328125,
+      "learning_rate": 9.869849246231157e-05,
+      "loss": 1.3874,
+      "step": 359
+    },
+    {
+      "epoch": 0.03072032768349529,
+      "grad_norm": 0.322265625,
+      "learning_rate": 9.869346733668342e-05,
+      "loss": 1.5274,
+      "step": 360
+    },
+    {
+      "epoch": 0.030805661927060555,
+      "grad_norm": 0.337890625,
+      "learning_rate": 9.868844221105528e-05,
+      "loss": 1.462,
+      "step": 361
+    },
+    {
+      "epoch": 0.03089099617062582,
+      "grad_norm": 0.37890625,
+      "learning_rate": 9.868341708542714e-05,
+      "loss": 1.5312,
+      "step": 362
+    },
+    {
+      "epoch": 0.030976330414191084,
+      "grad_norm": 0.375,
+      "learning_rate": 9.8678391959799e-05,
+      "loss": 1.44,
+      "step": 363
+    },
+    {
+      "epoch": 0.03106166465775635,
+      "grad_norm": 0.33203125,
+      "learning_rate": 9.867336683417085e-05,
+      "loss": 1.4294,
+      "step": 364
+    },
+    {
+      "epoch": 0.031146998901321613,
+      "grad_norm": 0.390625,
+      "learning_rate": 9.866834170854273e-05,
+      "loss": 1.4829,
+      "step": 365
+    },
+    {
+      "epoch": 0.031232333144886878,
+      "grad_norm": 0.37109375,
+      "learning_rate": 9.866331658291457e-05,
+      "loss": 1.5956,
+      "step": 366
+    },
+    {
+      "epoch": 0.031317667388452146,
+      "grad_norm": 0.333984375,
+      "learning_rate": 9.865829145728644e-05,
+      "loss": 1.4825,
+      "step": 367
+    },
+    {
+      "epoch": 0.03140300163201741,
+      "grad_norm": 0.369140625,
+      "learning_rate": 9.86532663316583e-05,
+      "loss": 1.4988,
+      "step": 368
+    },
+    {
+      "epoch": 0.031488335875582675,
+      "grad_norm": 0.3046875,
+      "learning_rate": 9.864824120603016e-05,
+      "loss": 1.3063,
+      "step": 369
+    },
+    {
+      "epoch": 0.03157367011914794,
+      "grad_norm": 0.3125,
+      "learning_rate": 9.864321608040201e-05,
+      "loss": 1.267,
+      "step": 370
+    },
+    {
+      "epoch": 0.031659004362713204,
+      "grad_norm": 0.333984375,
+      "learning_rate": 9.863819095477388e-05,
+      "loss": 1.2518,
+      "step": 371
+    },
+    {
+      "epoch": 0.03174433860627847,
+      "grad_norm": 0.337890625,
+      "learning_rate": 9.863316582914573e-05,
+      "loss": 1.3486,
+      "step": 372
+    },
+    {
+      "epoch": 0.03182967284984373,
+      "grad_norm": 0.35546875,
+      "learning_rate": 9.86281407035176e-05,
+      "loss": 1.4999,
+      "step": 373
+    },
+    {
+      "epoch": 0.031915007093409,
+      "grad_norm": 0.3671875,
+      "learning_rate": 9.862311557788944e-05,
+      "loss": 1.4815,
+      "step": 374
+    },
+    {
+      "epoch": 0.03200034133697426,
+      "grad_norm": 0.35546875,
+      "learning_rate": 9.861809045226132e-05,
+      "loss": 1.5226,
+      "step": 375
+    },
+    {
+      "epoch": 0.03208567558053953,
+      "grad_norm": 0.359375,
+      "learning_rate": 9.861306532663317e-05,
+      "loss": 1.4169,
+      "step": 376
+    },
+    {
+      "epoch": 0.03217100982410479,
+      "grad_norm": 0.3359375,
+      "learning_rate": 9.860804020100503e-05,
+      "loss": 1.2659,
+      "step": 377
+    },
+    {
+      "epoch": 0.032256344067670056,
+      "grad_norm": 0.328125,
+      "learning_rate": 9.860301507537689e-05,
+      "loss": 1.3325,
+      "step": 378
+    },
+    {
+      "epoch": 0.03234167831123532,
+      "grad_norm": 0.3359375,
+      "learning_rate": 9.859798994974875e-05,
+      "loss": 1.2241,
+      "step": 379
+    },
+    {
+      "epoch": 0.032427012554800586,
+      "grad_norm": 0.390625,
+      "learning_rate": 9.85929648241206e-05,
+      "loss": 1.6017,
+      "step": 380
+    },
+    {
+      "epoch": 0.03251234679836585,
+      "grad_norm": 0.359375,
+      "learning_rate": 9.858793969849246e-05,
+      "loss": 1.436,
+      "step": 381
+    },
+    {
+      "epoch": 0.032597681041931115,
+      "grad_norm": 0.337890625,
+      "learning_rate": 9.858291457286433e-05,
+      "loss": 1.4302,
+      "step": 382
+    },
+    {
+      "epoch": 0.03268301528549638,
+      "grad_norm": 0.353515625,
+      "learning_rate": 9.857788944723619e-05,
+      "loss": 1.3834,
+      "step": 383
+    },
+    {
+      "epoch": 0.032768349529061644,
+      "grad_norm": 0.3359375,
+      "learning_rate": 9.857286432160805e-05,
+      "loss": 1.3019,
+      "step": 384
+    },
+    {
+      "epoch": 0.03285368377262691,
+      "grad_norm": 0.380859375,
+      "learning_rate": 9.85678391959799e-05,
+      "loss": 1.487,
+      "step": 385
+    },
+    {
+      "epoch": 0.03293901801619217,
+      "grad_norm": 0.318359375,
+      "learning_rate": 9.856281407035176e-05,
+      "loss": 1.3938,
+      "step": 386
+    },
+    {
+      "epoch": 0.03302435225975744,
+      "grad_norm": 0.43359375,
+      "learning_rate": 9.855778894472362e-05,
+      "loss": 1.488,
+      "step": 387
+    },
+    {
+      "epoch": 0.0331096865033227,
+      "grad_norm": 0.337890625,
+      "learning_rate": 9.855276381909548e-05,
+      "loss": 1.3579,
+      "step": 388
+    },
+    {
+      "epoch": 0.03319502074688797,
+      "grad_norm": 0.326171875,
+      "learning_rate": 9.854773869346733e-05,
+      "loss": 1.3319,
+      "step": 389
+    },
+    {
+      "epoch": 0.03328035499045323,
+      "grad_norm": 0.345703125,
+      "learning_rate": 9.85427135678392e-05,
+      "loss": 1.3677,
+      "step": 390
+    },
+    {
+      "epoch": 0.033365689234018496,
+      "grad_norm": 0.380859375,
+      "learning_rate": 9.853768844221106e-05,
+      "loss": 1.4472,
+      "step": 391
+    },
+    {
+      "epoch": 0.03345102347758376,
+      "grad_norm": 0.326171875,
+      "learning_rate": 9.853266331658292e-05,
+      "loss": 1.388,
+      "step": 392
+    },
+    {
+      "epoch": 0.033536357721149025,
+      "grad_norm": 0.333984375,
+      "learning_rate": 9.852763819095477e-05,
+      "loss": 1.2981,
+      "step": 393
+    },
+    {
+      "epoch": 0.03362169196471429,
+      "grad_norm": 0.3203125,
+      "learning_rate": 9.852261306532664e-05,
+      "loss": 1.5465,
+      "step": 394
+    },
+    {
+      "epoch": 0.033707026208279554,
+      "grad_norm": 0.306640625,
+      "learning_rate": 9.851758793969849e-05,
+      "loss": 1.3837,
+      "step": 395
+    },
+    {
+      "epoch": 0.03379236045184482,
+      "grad_norm": 0.37109375,
+      "learning_rate": 9.851256281407035e-05,
+      "loss": 1.5756,
+      "step": 396
+    },
+    {
+      "epoch": 0.033877694695410084,
+      "grad_norm": 0.33203125,
+      "learning_rate": 9.850753768844222e-05,
+      "loss": 1.3774,
+      "step": 397
+    },
+    {
+      "epoch": 0.03396302893897535,
+      "grad_norm": 0.34375,
+      "learning_rate": 9.850251256281408e-05,
+      "loss": 1.3807,
+      "step": 398
+    },
+    {
+      "epoch": 0.03404836318254061,
+      "grad_norm": 0.35546875,
+      "learning_rate": 9.849748743718593e-05,
+      "loss": 1.4118,
+      "step": 399
+    },
+    {
+      "epoch": 0.03413369742610588,
+      "grad_norm": 0.384765625,
+      "learning_rate": 9.84924623115578e-05,
+      "loss": 1.5277,
+      "step": 400
+    },
+    {
+      "epoch": 0.03421903166967114,
+      "grad_norm": 0.380859375,
+      "learning_rate": 9.848743718592965e-05,
+      "loss": 1.5215,
+      "step": 401
+    },
+    {
+      "epoch": 0.034304365913236406,
+      "grad_norm": 0.326171875,
+      "learning_rate": 9.848241206030151e-05,
+      "loss": 1.5074,
+      "step": 402
+    },
+    {
+      "epoch": 0.03438970015680167,
+      "grad_norm": 0.3671875,
+      "learning_rate": 9.847738693467337e-05,
+      "loss": 1.3304,
+      "step": 403
+    },
+    {
+      "epoch": 0.034475034400366936,
+      "grad_norm": 0.333984375,
+      "learning_rate": 9.847236180904524e-05,
+      "loss": 1.3198,
+      "step": 404
+    },
+    {
+      "epoch": 0.0345603686439322,
+      "grad_norm": 0.34765625,
+      "learning_rate": 9.846733668341709e-05,
+      "loss": 1.3872,
+      "step": 405
+    },
+    {
+      "epoch": 0.034645702887497465,
+      "grad_norm": 0.353515625,
+      "learning_rate": 9.846231155778895e-05,
+      "loss": 1.4763,
+      "step": 406
+    },
+    {
+      "epoch": 0.03473103713106273,
+      "grad_norm": 0.3203125,
+      "learning_rate": 9.845728643216081e-05,
+      "loss": 1.366,
+      "step": 407
+    },
+    {
+      "epoch": 0.034816371374627994,
+      "grad_norm": 0.400390625,
+      "learning_rate": 9.845226130653267e-05,
+      "loss": 1.3394,
+      "step": 408
+    },
+    {
+      "epoch": 0.03490170561819326,
+      "grad_norm": 0.33984375,
+      "learning_rate": 9.844723618090452e-05,
+      "loss": 1.4859,
+      "step": 409
+    },
+    {
+      "epoch": 0.03498703986175852,
+      "grad_norm": 0.37109375,
+      "learning_rate": 9.84422110552764e-05,
+      "loss": 1.4893,
+      "step": 410
+    },
+    {
+      "epoch": 0.03507237410532379,
+      "grad_norm": 0.349609375,
+      "learning_rate": 9.843718592964824e-05,
+      "loss": 1.4897,
+      "step": 411
+    },
+    {
+      "epoch": 0.03515770834888905,
+      "grad_norm": 0.328125,
+      "learning_rate": 9.84321608040201e-05,
+      "loss": 1.4898,
+      "step": 412
+    },
+    {
+      "epoch": 0.03524304259245432,
+      "grad_norm": 0.3203125,
+      "learning_rate": 9.842713567839197e-05,
+      "loss": 1.2922,
+      "step": 413
+    },
+    {
+      "epoch": 0.03532837683601958,
+      "grad_norm": 0.3359375,
+      "learning_rate": 9.842211055276383e-05,
+      "loss": 1.4774,
+      "step": 414
+    },
+    {
+      "epoch": 0.035413711079584846,
+      "grad_norm": 0.298828125,
+      "learning_rate": 9.841708542713568e-05,
+      "loss": 1.3156,
+      "step": 415
+    },
+    {
+      "epoch": 0.03549904532315011,
+      "grad_norm": 0.294921875,
+      "learning_rate": 9.841206030150754e-05,
+      "loss": 1.219,
+      "step": 416
+    },
+    {
+      "epoch": 0.035584379566715375,
+      "grad_norm": 0.31640625,
+      "learning_rate": 9.84070351758794e-05,
+      "loss": 1.346,
+      "step": 417
+    },
+    {
+      "epoch": 0.03566971381028064,
+      "grad_norm": 0.369140625,
+      "learning_rate": 9.840201005025126e-05,
+      "loss": 1.3559,
+      "step": 418
+    },
+    {
+      "epoch": 0.035755048053845905,
+      "grad_norm": 0.345703125,
+      "learning_rate": 9.839698492462313e-05,
+      "loss": 1.3243,
+      "step": 419
+    },
+    {
+      "epoch": 0.03584038229741117,
+      "grad_norm": 0.33984375,
+      "learning_rate": 9.839195979899497e-05,
+      "loss": 1.492,
+      "step": 420
+    },
+    {
+      "epoch": 0.035925716540976434,
+      "grad_norm": 0.326171875,
+      "learning_rate": 9.838693467336684e-05,
+      "loss": 1.324,
+      "step": 421
+    },
+    {
+      "epoch": 0.036011050784541705,
+      "grad_norm": 0.353515625,
+      "learning_rate": 9.83819095477387e-05,
+      "loss": 1.4559,
+      "step": 422
+    },
+    {
+      "epoch": 0.03609638502810697,
+      "grad_norm": 0.3359375,
+      "learning_rate": 9.837688442211056e-05,
+      "loss": 1.2916,
+      "step": 423
+    },
+    {
+      "epoch": 0.036181719271672234,
+      "grad_norm": 0.4609375,
+      "learning_rate": 9.837185929648241e-05,
+      "loss": 1.5289,
+      "step": 424
+    },
+    {
+      "epoch": 0.0362670535152375,
+      "grad_norm": 0.33984375,
+      "learning_rate": 9.836683417085427e-05,
+      "loss": 1.4793,
+      "step": 425
+    },
+    {
+      "epoch": 0.036352387758802764,
+      "grad_norm": 0.3203125,
+      "learning_rate": 9.836180904522613e-05,
+      "loss": 1.3065,
+      "step": 426
+    },
+    {
+      "epoch": 0.03643772200236803,
+      "grad_norm": 0.3984375,
+      "learning_rate": 9.8356783919598e-05,
+      "loss": 1.5644,
+      "step": 427
+    },
+    {
+      "epoch": 0.03652305624593329,
+      "grad_norm": 0.328125,
+      "learning_rate": 9.835175879396984e-05,
+      "loss": 1.4515,
+      "step": 428
+    },
+    {
+      "epoch": 0.03660839048949856,
+      "grad_norm": 0.341796875,
+      "learning_rate": 9.834673366834172e-05,
+      "loss": 1.3688,
+      "step": 429
+    },
+    {
+      "epoch": 0.03669372473306382,
+      "grad_norm": 0.3125,
+      "learning_rate": 9.834170854271357e-05,
+      "loss": 1.4453,
+      "step": 430
+    },
+    {
+      "epoch": 0.036779058976629087,
+      "grad_norm": 0.302734375,
+      "learning_rate": 9.833668341708543e-05,
+      "loss": 1.4015,
+      "step": 431
+    },
+    {
+      "epoch": 0.03686439322019435,
+      "grad_norm": 0.306640625,
+      "learning_rate": 9.833165829145729e-05,
+      "loss": 1.3695,
+      "step": 432
+    },
+    {
+      "epoch": 0.036949727463759616,
+      "grad_norm": 0.33203125,
+      "learning_rate": 9.832663316582915e-05,
+      "loss": 1.2991,
+      "step": 433
+    },
+    {
+      "epoch": 0.03703506170732488,
+      "grad_norm": 0.365234375,
+      "learning_rate": 9.8321608040201e-05,
+      "loss": 1.4742,
+      "step": 434
+    },
+    {
+      "epoch": 0.037120395950890145,
+      "grad_norm": 0.318359375,
+      "learning_rate": 9.831658291457288e-05,
+      "loss": 1.3426,
+      "step": 435
+    },
+    {
+      "epoch": 0.03720573019445541,
+      "grad_norm": 0.3515625,
+      "learning_rate": 9.831155778894473e-05,
+      "loss": 1.3282,
+      "step": 436
+    },
+    {
+      "epoch": 0.037291064438020674,
+      "grad_norm": 0.322265625,
+      "learning_rate": 9.830653266331659e-05,
+      "loss": 1.4286,
+      "step": 437
+    },
+    {
+      "epoch": 0.03737639868158594,
+      "grad_norm": 0.333984375,
+      "learning_rate": 9.830150753768844e-05,
+      "loss": 1.5547,
+      "step": 438
+    },
+    {
+      "epoch": 0.0374617329251512,
+      "grad_norm": 0.384765625,
+      "learning_rate": 9.829648241206031e-05,
+      "loss": 1.3978,
+      "step": 439
+    },
+    {
+      "epoch": 0.03754706716871647,
+      "grad_norm": 0.34375,
+      "learning_rate": 9.829145728643216e-05,
+      "loss": 1.4236,
+      "step": 440
+    },
+    {
+      "epoch": 0.03763240141228173,
+      "grad_norm": 0.306640625,
+      "learning_rate": 9.828643216080402e-05,
+      "loss": 1.3891,
+      "step": 441
+    },
+    {
+      "epoch": 0.037717735655847,
+      "grad_norm": 0.369140625,
+      "learning_rate": 9.828140703517589e-05,
+      "loss": 1.5679,
+      "step": 442
+    },
+    {
+      "epoch": 0.03780306989941226,
+      "grad_norm": 0.3515625,
+      "learning_rate": 9.827638190954775e-05,
+      "loss": 1.377,
+      "step": 443
+    },
+    {
+      "epoch": 0.037888404142977526,
+      "grad_norm": 0.34375,
+      "learning_rate": 9.82713567839196e-05,
+      "loss": 1.4544,
+      "step": 444
+    },
+    {
+      "epoch": 0.03797373838654279,
+      "grad_norm": 0.333984375,
+      "learning_rate": 9.826633165829147e-05,
+      "loss": 1.4464,
+      "step": 445
+    },
+    {
+      "epoch": 0.038059072630108055,
+      "grad_norm": 0.314453125,
+      "learning_rate": 9.826130653266332e-05,
+      "loss": 1.3857,
+      "step": 446
+    },
+    {
+      "epoch": 0.03814440687367332,
+      "grad_norm": 0.33984375,
+      "learning_rate": 9.825628140703518e-05,
+      "loss": 1.4554,
+      "step": 447
+    },
+    {
+      "epoch": 0.038229741117238585,
+      "grad_norm": 0.41796875,
+      "learning_rate": 9.825125628140704e-05,
+      "loss": 1.4405,
+      "step": 448
+    },
+    {
+      "epoch": 0.03831507536080385,
+      "grad_norm": 0.326171875,
+      "learning_rate": 9.82462311557789e-05,
+      "loss": 1.4376,
+      "step": 449
+    },
+    {
+      "epoch": 0.038400409604369114,
+      "grad_norm": 0.484375,
+      "learning_rate": 9.824120603015075e-05,
+      "loss": 1.5418,
+      "step": 450
+    },
+    {
+      "epoch": 0.03848574384793438,
+      "grad_norm": 0.318359375,
+      "learning_rate": 9.823618090452262e-05,
+      "loss": 1.3599,
+      "step": 451
+    },
+    {
+      "epoch": 0.03857107809149964,
+      "grad_norm": 0.341796875,
+      "learning_rate": 9.823115577889448e-05,
+      "loss": 1.2787,
+      "step": 452
+    },
+    {
+      "epoch": 0.03865641233506491,
+      "grad_norm": 0.34375,
+      "learning_rate": 9.822613065326634e-05,
+      "loss": 1.3217,
+      "step": 453
+    },
+    {
+      "epoch": 0.03874174657863017,
+      "grad_norm": 0.330078125,
+      "learning_rate": 9.822110552763819e-05,
+      "loss": 1.2746,
+      "step": 454
+    },
+    {
+      "epoch": 0.03882708082219544,
+      "grad_norm": 0.404296875,
+      "learning_rate": 9.821608040201005e-05,
+      "loss": 1.3369,
+      "step": 455
+    },
+    {
+      "epoch": 0.0389124150657607,
+      "grad_norm": 0.333984375,
+      "learning_rate": 9.821105527638191e-05,
+      "loss": 1.4238,
+      "step": 456
+    },
+    {
+      "epoch": 0.038997749309325966,
+      "grad_norm": 0.34765625,
+      "learning_rate": 9.820603015075378e-05,
+      "loss": 1.4062,
+      "step": 457
+    },
+    {
+      "epoch": 0.03908308355289123,
+      "grad_norm": 0.42578125,
+      "learning_rate": 9.820100502512564e-05,
+      "loss": 1.5235,
+      "step": 458
+    },
+    {
+      "epoch": 0.039168417796456495,
+      "grad_norm": 0.326171875,
+      "learning_rate": 9.819597989949749e-05,
+      "loss": 1.411,
+      "step": 459
+    },
+    {
+      "epoch": 0.03925375204002176,
+      "grad_norm": 0.390625,
+      "learning_rate": 9.819095477386935e-05,
+      "loss": 1.3977,
+      "step": 460
+    },
+    {
+      "epoch": 0.039339086283587024,
+      "grad_norm": 0.322265625,
+      "learning_rate": 9.818592964824121e-05,
+      "loss": 1.5666,
+      "step": 461
+    },
+    {
+      "epoch": 0.03942442052715229,
+      "grad_norm": 0.33984375,
+      "learning_rate": 9.818090452261307e-05,
+      "loss": 1.4605,
+      "step": 462
+    },
+    {
+      "epoch": 0.03950975477071755,
+      "grad_norm": 0.373046875,
+      "learning_rate": 9.817587939698492e-05,
+      "loss": 1.4669,
+      "step": 463
+    },
+    {
+      "epoch": 0.03959508901428282,
+      "grad_norm": 0.2578125,
+      "learning_rate": 9.81708542713568e-05,
+      "loss": 2.2431,
+      "step": 464
+    },
+    {
+      "epoch": 0.03968042325784808,
+      "grad_norm": 0.34765625,
+      "learning_rate": 9.816582914572864e-05,
+      "loss": 1.3993,
+      "step": 465
+    },
+    {
+      "epoch": 0.03976575750141335,
+      "grad_norm": 0.3359375,
+      "learning_rate": 9.81608040201005e-05,
+      "loss": 1.3912,
+      "step": 466
+    },
+    {
+      "epoch": 0.03985109174497861,
+      "grad_norm": 0.38671875,
+      "learning_rate": 9.815577889447236e-05,
+      "loss": 1.3504,
+      "step": 467
+    },
+    {
+      "epoch": 0.039936425988543876,
+      "grad_norm": 0.318359375,
+      "learning_rate": 9.815075376884423e-05,
+      "loss": 1.2783,
+      "step": 468
+    },
+    {
+      "epoch": 0.04002176023210914,
+      "grad_norm": 0.29296875,
+      "learning_rate": 9.814572864321608e-05,
+      "loss": 1.4368,
+      "step": 469
+    },
+    {
+      "epoch": 0.040107094475674405,
+      "grad_norm": 0.33203125,
+      "learning_rate": 9.814070351758794e-05,
+      "loss": 1.2483,
+      "step": 470
+    },
+    {
+      "epoch": 0.04019242871923967,
+      "grad_norm": 0.3984375,
+      "learning_rate": 9.81356783919598e-05,
+      "loss": 1.2745,
+      "step": 471
+    },
+    {
+      "epoch": 0.040277762962804935,
+      "grad_norm": 0.35546875,
+      "learning_rate": 9.813065326633167e-05,
+      "loss": 1.1939,
+      "step": 472
+    },
+    {
+      "epoch": 0.0403630972063702,
+      "grad_norm": 0.34375,
+      "learning_rate": 9.812562814070351e-05,
+      "loss": 1.3728,
+      "step": 473
+    },
+    {
+      "epoch": 0.040448431449935464,
+      "grad_norm": 0.3359375,
+      "learning_rate": 9.812060301507539e-05,
+      "loss": 1.4527,
+      "step": 474
+    },
+    {
+      "epoch": 0.04053376569350073,
+      "grad_norm": 0.326171875,
+      "learning_rate": 9.811557788944724e-05,
+      "loss": 1.5409,
+      "step": 475
+    },
+    {
+      "epoch": 0.04061909993706599,
+      "grad_norm": 0.33984375,
+      "learning_rate": 9.81105527638191e-05,
+      "loss": 1.4426,
+      "step": 476
+    },
+    {
+      "epoch": 0.04070443418063126,
+      "grad_norm": 0.333984375,
+      "learning_rate": 9.810552763819096e-05,
+      "loss": 1.3423,
+      "step": 477
+    },
+    {
+      "epoch": 0.04078976842419652,
+      "grad_norm": 0.34765625,
+      "learning_rate": 9.810050251256282e-05,
+      "loss": 1.4133,
+      "step": 478
+    },
+    {
+      "epoch": 0.04087510266776179,
+      "grad_norm": 0.34765625,
+      "learning_rate": 9.809547738693467e-05,
+      "loss": 1.3434,
+      "step": 479
+    },
+    {
+      "epoch": 0.04096043691132705,
+      "grad_norm": 0.302734375,
+      "learning_rate": 9.809045226130655e-05,
+      "loss": 1.5089,
+      "step": 480
+    },
+    {
+      "epoch": 0.041045771154892316,
+      "grad_norm": 0.33203125,
+      "learning_rate": 9.80854271356784e-05,
+      "loss": 1.185,
+      "step": 481
+    },
+    {
+      "epoch": 0.04113110539845758,
+      "grad_norm": 0.419921875,
+      "learning_rate": 9.808040201005026e-05,
+      "loss": 1.4114,
+      "step": 482
+    },
+    {
+      "epoch": 0.041216439642022845,
+      "grad_norm": 0.35546875,
+      "learning_rate": 9.807537688442211e-05,
+      "loss": 1.2592,
+      "step": 483
+    },
+    {
+      "epoch": 0.04130177388558811,
+      "grad_norm": 0.35546875,
+      "learning_rate": 9.807035175879398e-05,
+      "loss": 1.3393,
+      "step": 484
+    },
+    {
+      "epoch": 0.041387108129153374,
+      "grad_norm": 0.322265625,
+      "learning_rate": 9.806532663316583e-05,
+      "loss": 1.5252,
+      "step": 485
+    },
+    {
+      "epoch": 0.04147244237271864,
+      "grad_norm": 0.3671875,
+      "learning_rate": 9.806030150753769e-05,
+      "loss": 1.5749,
+      "step": 486
+    },
+    {
+      "epoch": 0.04155777661628391,
+      "grad_norm": 0.3046875,
+      "learning_rate": 9.805527638190956e-05,
+      "loss": 1.3505,
+      "step": 487
+    },
+    {
+      "epoch": 0.041643110859849175,
+      "grad_norm": 0.431640625,
+      "learning_rate": 9.805025125628142e-05,
+      "loss": 1.3747,
+      "step": 488
+    },
+    {
+      "epoch": 0.04172844510341444,
+      "grad_norm": 0.30859375,
+      "learning_rate": 9.804522613065327e-05,
+      "loss": 1.393,
+      "step": 489
+    },
+    {
+      "epoch": 0.041813779346979704,
+      "grad_norm": 0.34375,
+      "learning_rate": 9.804020100502513e-05,
+      "loss": 1.3142,
+      "step": 490
+    },
+    {
+      "epoch": 0.04189911359054497,
+      "grad_norm": 0.33203125,
+      "learning_rate": 9.803517587939699e-05,
+      "loss": 1.2547,
+      "step": 491
+    },
+    {
+      "epoch": 0.04198444783411023,
+      "grad_norm": 0.31640625,
+      "learning_rate": 9.803015075376885e-05,
+      "loss": 1.3805,
+      "step": 492
+    },
+    {
+      "epoch": 0.0420697820776755,
+      "grad_norm": 0.412109375,
+      "learning_rate": 9.802512562814071e-05,
+      "loss": 1.4887,
+      "step": 493
+    },
+    {
+      "epoch": 0.04215511632124076,
+      "grad_norm": 0.322265625,
+      "learning_rate": 9.802010050251256e-05,
+      "loss": 1.5492,
+      "step": 494
+    },
+    {
+      "epoch": 0.04224045056480603,
+      "grad_norm": 0.310546875,
+      "learning_rate": 9.801507537688442e-05,
+      "loss": 1.4367,
+      "step": 495
+    },
+    {
+      "epoch": 0.04232578480837129,
+      "grad_norm": 0.31640625,
+      "learning_rate": 9.801005025125629e-05,
+      "loss": 1.3598,
+      "step": 496
+    },
+    {
+      "epoch": 0.042411119051936556,
+      "grad_norm": 0.341796875,
+      "learning_rate": 9.800502512562815e-05,
+      "loss": 1.5329,
+      "step": 497
+    },
+    {
+      "epoch": 0.04249645329550182,
+      "grad_norm": 0.33203125,
+      "learning_rate": 9.8e-05,
+      "loss": 1.3892,
+      "step": 498
+    },
+    {
+      "epoch": 0.042581787539067086,
+      "grad_norm": 0.345703125,
+      "learning_rate": 9.799497487437186e-05,
+      "loss": 1.4771,
+      "step": 499
+    },
+    {
+      "epoch": 0.04266712178263235,
+      "grad_norm": 0.318359375,
+      "learning_rate": 9.798994974874372e-05,
+      "loss": 1.4531,
+      "step": 500
+    },
+    {
+      "epoch": 0.042752456026197615,
+      "grad_norm": 0.361328125,
+      "learning_rate": 9.798492462311558e-05,
+      "loss": 1.4213,
+      "step": 501
+    },
+    {
+      "epoch": 0.04283779026976288,
+      "grad_norm": 0.333984375,
+      "learning_rate": 9.797989949748743e-05,
+      "loss": 1.445,
+      "step": 502
+    },
+    {
+      "epoch": 0.042923124513328144,
+      "grad_norm": 0.36328125,
+      "learning_rate": 9.797487437185931e-05,
+      "loss": 1.4574,
+      "step": 503
+    },
+    {
+      "epoch": 0.04300845875689341,
+      "grad_norm": 0.466796875,
+      "learning_rate": 9.796984924623116e-05,
+      "loss": 1.6703,
+      "step": 504
+    },
+    {
+      "epoch": 0.04309379300045867,
+      "grad_norm": 0.39453125,
+      "learning_rate": 9.796482412060302e-05,
+      "loss": 1.3828,
+      "step": 505
+    },
+    {
+      "epoch": 0.04317912724402394,
+      "grad_norm": 0.337890625,
+      "learning_rate": 9.795979899497488e-05,
+      "loss": 1.4373,
+      "step": 506
+    },
+    {
+      "epoch": 0.0432644614875892,
+      "grad_norm": 0.376953125,
+      "learning_rate": 9.795477386934674e-05,
+      "loss": 1.5437,
+      "step": 507
+    },
+    {
+      "epoch": 0.04334979573115447,
+      "grad_norm": 0.330078125,
+      "learning_rate": 9.794974874371859e-05,
+      "loss": 1.4256,
+      "step": 508
+    },
+    {
+      "epoch": 0.04343512997471973,
+      "grad_norm": 0.365234375,
+      "learning_rate": 9.794472361809047e-05,
+      "loss": 1.3799,
+      "step": 509
+    },
+    {
+      "epoch": 0.043520464218284996,
+      "grad_norm": 0.33984375,
+      "learning_rate": 9.793969849246231e-05,
+      "loss": 1.3046,
+      "step": 510
+    },
+    {
+      "epoch": 0.04360579846185026,
+      "grad_norm": 0.35546875,
+      "learning_rate": 9.793467336683418e-05,
+      "loss": 1.4718,
+      "step": 511
+    },
+    {
+      "epoch": 0.043691132705415525,
+      "grad_norm": 0.34765625,
+      "learning_rate": 9.792964824120604e-05,
+      "loss": 1.3132,
+      "step": 512
+    },
+    {
+      "epoch": 0.04377646694898079,
+      "grad_norm": 0.3359375,
+      "learning_rate": 9.79246231155779e-05,
+      "loss": 1.4986,
+      "step": 513
+    },
+    {
+      "epoch": 0.043861801192546054,
+      "grad_norm": 0.451171875,
+      "learning_rate": 9.791959798994975e-05,
+      "loss": 1.3882,
+      "step": 514
+    },
+    {
+      "epoch": 0.04394713543611132,
+      "grad_norm": 0.3359375,
+      "learning_rate": 9.791457286432161e-05,
+      "loss": 1.4305,
+      "step": 515
+    },
+    {
+      "epoch": 0.044032469679676584,
+      "grad_norm": 0.359375,
+      "learning_rate": 9.790954773869347e-05,
+      "loss": 1.2547,
+      "step": 516
+    },
+    {
+      "epoch": 0.04411780392324185,
+      "grad_norm": 0.291015625,
+      "learning_rate": 9.790452261306533e-05,
+      "loss": 1.4184,
+      "step": 517
+    },
+    {
+      "epoch": 0.04420313816680711,
+      "grad_norm": 0.318359375,
+      "learning_rate": 9.789949748743718e-05,
+      "loss": 1.3546,
+      "step": 518
+    },
+    {
+      "epoch": 0.04428847241037238,
+      "grad_norm": 0.3359375,
+      "learning_rate": 9.789447236180906e-05,
+      "loss": 1.5497,
+      "step": 519
+    },
+    {
+      "epoch": 0.04437380665393764,
+      "grad_norm": 0.326171875,
+      "learning_rate": 9.788944723618091e-05,
+      "loss": 1.3478,
+      "step": 520
+    },
+    {
+      "epoch": 0.044459140897502906,
+      "grad_norm": 0.345703125,
+      "learning_rate": 9.788442211055277e-05,
+      "loss": 1.3972,
+      "step": 521
+    },
+    {
+      "epoch": 0.04454447514106817,
+      "grad_norm": 0.314453125,
+      "learning_rate": 9.787939698492463e-05,
+      "loss": 1.2856,
+      "step": 522
+    },
+    {
+      "epoch": 0.044629809384633436,
+      "grad_norm": 0.31640625,
+      "learning_rate": 9.78743718592965e-05,
+      "loss": 1.2654,
+      "step": 523
+    },
+    {
+      "epoch": 0.0447151436281987,
+      "grad_norm": 0.34375,
+      "learning_rate": 9.786934673366834e-05,
+      "loss": 1.38,
+      "step": 524
+    },
+    {
+      "epoch": 0.044800477871763965,
+      "grad_norm": 0.287109375,
+      "learning_rate": 9.78643216080402e-05,
+      "loss": 1.1772,
+      "step": 525
+    },
+    {
+      "epoch": 0.04488581211532923,
+      "grad_norm": 0.330078125,
+      "learning_rate": 9.785929648241207e-05,
+      "loss": 1.368,
+      "step": 526
+    },
+    {
+      "epoch": 0.044971146358894494,
+      "grad_norm": 0.306640625,
+      "learning_rate": 9.785427135678393e-05,
+      "loss": 1.3993,
+      "step": 527
+    },
+    {
+      "epoch": 0.04505648060245976,
+      "grad_norm": 0.349609375,
+      "learning_rate": 9.784924623115579e-05,
+      "loss": 1.3716,
+      "step": 528
+    },
+    {
+      "epoch": 0.04514181484602502,
+      "grad_norm": 0.345703125,
+      "learning_rate": 9.784422110552764e-05,
+      "loss": 1.4416,
+      "step": 529
+    },
+    {
+      "epoch": 0.04522714908959029,
+      "grad_norm": 0.3515625,
+      "learning_rate": 9.78391959798995e-05,
+      "loss": 1.3523,
+      "step": 530
+    },
+    {
+      "epoch": 0.04531248333315555,
+      "grad_norm": 0.33984375,
+      "learning_rate": 9.783417085427136e-05,
+      "loss": 1.373,
+      "step": 531
+    },
+    {
+      "epoch": 0.04539781757672082,
+      "grad_norm": 0.30859375,
+      "learning_rate": 9.782914572864322e-05,
+      "loss": 1.3263,
+      "step": 532
+    },
+    {
+      "epoch": 0.04548315182028608,
+      "grad_norm": 0.3515625,
+      "learning_rate": 9.782412060301507e-05,
+      "loss": 1.3304,
+      "step": 533
+    },
+    {
+      "epoch": 0.045568486063851346,
+      "grad_norm": 0.33203125,
+      "learning_rate": 9.781909547738694e-05,
+      "loss": 1.4301,
+      "step": 534
+    },
+    {
+      "epoch": 0.04565382030741661,
+      "grad_norm": 0.396484375,
+      "learning_rate": 9.78140703517588e-05,
+      "loss": 1.3432,
+      "step": 535
+    },
+    {
+      "epoch": 0.045739154550981875,
+      "grad_norm": 0.318359375,
+      "learning_rate": 9.780904522613066e-05,
+      "loss": 1.3278,
+      "step": 536
+    },
+    {
+      "epoch": 0.04582448879454714,
+      "grad_norm": 0.296875,
+      "learning_rate": 9.780402010050251e-05,
+      "loss": 1.3933,
+      "step": 537
+    },
+    {
+      "epoch": 0.045909823038112405,
+      "grad_norm": 0.30859375,
+      "learning_rate": 9.779899497487438e-05,
+      "loss": 1.3747,
+      "step": 538
+    },
+    {
+      "epoch": 0.04599515728167767,
+      "grad_norm": 0.349609375,
+      "learning_rate": 9.779396984924623e-05,
+      "loss": 1.4705,
+      "step": 539
+    },
+    {
+      "epoch": 0.046080491525242934,
+      "grad_norm": 0.294921875,
+      "learning_rate": 9.77889447236181e-05,
+      "loss": 1.3608,
+      "step": 540
+    },
+    {
+      "epoch": 0.0461658257688082,
+      "grad_norm": 0.2890625,
+      "learning_rate": 9.778391959798996e-05,
+      "loss": 1.3604,
+      "step": 541
+    },
+    {
+      "epoch": 0.04625116001237346,
+      "grad_norm": 0.357421875,
+      "learning_rate": 9.777889447236182e-05,
+      "loss": 1.366,
+      "step": 542
+    },
+    {
+      "epoch": 0.04633649425593873,
+      "grad_norm": 0.330078125,
+      "learning_rate": 9.777386934673367e-05,
+      "loss": 1.3359,
+      "step": 543
+    },
+    {
+      "epoch": 0.04642182849950399,
+      "grad_norm": 0.330078125,
+      "learning_rate": 9.776884422110554e-05,
+      "loss": 1.3241,
+      "step": 544
+    },
+    {
+      "epoch": 0.04650716274306926,
+      "grad_norm": 0.36328125,
+      "learning_rate": 9.776381909547739e-05,
+      "loss": 1.3774,
+      "step": 545
+    },
+    {
+      "epoch": 0.04659249698663452,
+      "grad_norm": 0.330078125,
+      "learning_rate": 9.775879396984925e-05,
+      "loss": 1.4116,
+      "step": 546
+    },
+    {
+      "epoch": 0.046677831230199786,
+      "grad_norm": 0.30859375,
+      "learning_rate": 9.77537688442211e-05,
+      "loss": 1.2552,
+      "step": 547
+    },
+    {
+      "epoch": 0.04676316547376505,
+      "grad_norm": 0.3203125,
+      "learning_rate": 9.774874371859298e-05,
+      "loss": 1.3401,
+      "step": 548
+    },
+    {
+      "epoch": 0.046848499717330315,
+      "grad_norm": 0.34765625,
+      "learning_rate": 9.774371859296483e-05,
+      "loss": 1.31,
+      "step": 549
+    },
+    {
+      "epoch": 0.04693383396089558,
+      "grad_norm": 0.361328125,
+      "learning_rate": 9.773869346733669e-05,
+      "loss": 1.4162,
+      "step": 550
+    },
+    {
+      "epoch": 0.047019168204460844,
+      "grad_norm": 0.279296875,
+      "learning_rate": 9.773366834170855e-05,
+      "loss": 1.3854,
+      "step": 551
+    },
+    {
+      "epoch": 0.047104502448026116,
+      "grad_norm": 0.333984375,
+      "learning_rate": 9.772864321608041e-05,
+      "loss": 1.3167,
+      "step": 552
+    },
+    {
+      "epoch": 0.04718983669159138,
+      "grad_norm": 0.310546875,
+      "learning_rate": 9.772361809045226e-05,
+      "loss": 1.4153,
+      "step": 553
+    },
+    {
+      "epoch": 0.047275170935156645,
+      "grad_norm": 0.306640625,
+      "learning_rate": 9.771859296482414e-05,
+      "loss": 1.3265,
+      "step": 554
+    },
+    {
+      "epoch": 0.04736050517872191,
+      "grad_norm": 0.333984375,
+      "learning_rate": 9.771356783919598e-05,
+      "loss": 1.4612,
+      "step": 555
+    },
+    {
+      "epoch": 0.047445839422287174,
+      "grad_norm": 0.333984375,
+      "learning_rate": 9.770854271356785e-05,
+      "loss": 1.3782,
+      "step": 556
+    },
+    {
+      "epoch": 0.04753117366585244,
+      "grad_norm": 0.326171875,
+      "learning_rate": 9.770351758793971e-05,
+      "loss": 1.3727,
+      "step": 557
+    },
+    {
+      "epoch": 0.0476165079094177,
+      "grad_norm": 0.375,
+      "learning_rate": 9.769849246231157e-05,
+      "loss": 1.4461,
+      "step": 558
+    },
+    {
+      "epoch": 0.04770184215298297,
+      "grad_norm": 0.32421875,
+      "learning_rate": 9.769346733668342e-05,
+      "loss": 1.1829,
+      "step": 559
+    },
+    {
+      "epoch": 0.04778717639654823,
+      "grad_norm": 0.302734375,
+      "learning_rate": 9.768844221105528e-05,
+      "loss": 1.3462,
+      "step": 560
+    },
+    {
+      "epoch": 0.0478725106401135,
+      "grad_norm": 0.330078125,
+      "learning_rate": 9.768341708542714e-05,
+      "loss": 1.4409,
+      "step": 561
+    },
+    {
+      "epoch": 0.04795784488367876,
+      "grad_norm": 0.328125,
+      "learning_rate": 9.7678391959799e-05,
+      "loss": 1.4291,
+      "step": 562
+    },
+    {
+      "epoch": 0.048043179127244026,
+      "grad_norm": 0.345703125,
+      "learning_rate": 9.767336683417085e-05,
+      "loss": 1.2478,
+      "step": 563
+    },
+    {
+      "epoch": 0.04812851337080929,
+      "grad_norm": 0.328125,
+      "learning_rate": 9.766834170854271e-05,
+      "loss": 1.5019,
+      "step": 564
+    },
+    {
+      "epoch": 0.048213847614374555,
+      "grad_norm": 0.314453125,
+      "learning_rate": 9.766331658291458e-05,
+      "loss": 1.3371,
+      "step": 565
+    },
+    {
+      "epoch": 0.04829918185793982,
+      "grad_norm": 0.42578125,
+      "learning_rate": 9.765829145728644e-05,
+      "loss": 1.7739,
+      "step": 566
+    },
+    {
+      "epoch": 0.048384516101505085,
+      "grad_norm": 0.3359375,
+      "learning_rate": 9.76532663316583e-05,
+      "loss": 1.431,
+      "step": 567
+    },
+    {
+      "epoch": 0.04846985034507035,
+      "grad_norm": 0.298828125,
+      "learning_rate": 9.764824120603015e-05,
+      "loss": 1.3027,
+      "step": 568
+    },
+    {
+      "epoch": 0.048555184588635614,
+      "grad_norm": 0.345703125,
+      "learning_rate": 9.764321608040201e-05,
+      "loss": 1.4325,
+      "step": 569
+    },
+    {
+      "epoch": 0.04864051883220088,
+      "grad_norm": 0.3515625,
+      "learning_rate": 9.763819095477387e-05,
+      "loss": 1.4314,
+      "step": 570
+    },
+    {
+      "epoch": 0.04872585307576614,
+      "grad_norm": 0.34375,
+      "learning_rate": 9.763316582914574e-05,
+      "loss": 1.438,
+      "step": 571
+    },
+    {
+      "epoch": 0.04881118731933141,
+      "grad_norm": 0.3125,
+      "learning_rate": 9.762814070351758e-05,
+      "loss": 1.2522,
+      "step": 572
+    },
+    {
+      "epoch": 0.04889652156289667,
+      "grad_norm": 0.36328125,
+      "learning_rate": 9.762311557788946e-05,
+      "loss": 1.4434,
+      "step": 573
+    },
+    {
+      "epoch": 0.04898185580646194,
+      "grad_norm": 0.373046875,
+      "learning_rate": 9.761809045226131e-05,
+      "loss": 1.4041,
+      "step": 574
+    },
+    {
+      "epoch": 0.0490671900500272,
+      "grad_norm": 0.388671875,
+      "learning_rate": 9.761306532663317e-05,
+      "loss": 1.3432,
+      "step": 575
+    },
+    {
+      "epoch": 0.049152524293592466,
+      "grad_norm": 0.302734375,
+      "learning_rate": 9.760804020100502e-05,
+      "loss": 1.2897,
+      "step": 576
+    },
+    {
+      "epoch": 0.04923785853715773,
+      "grad_norm": 0.34765625,
+      "learning_rate": 9.76030150753769e-05,
+      "loss": 1.4813,
+      "step": 577
+    },
+    {
+      "epoch": 0.049323192780722995,
+      "grad_norm": 0.369140625,
+      "learning_rate": 9.759798994974874e-05,
+      "loss": 1.6552,
+      "step": 578
+    },
+    {
+      "epoch": 0.04940852702428826,
+      "grad_norm": 0.30859375,
+      "learning_rate": 9.75929648241206e-05,
+      "loss": 1.3383,
+      "step": 579
+    },
+    {
+      "epoch": 0.049493861267853524,
+      "grad_norm": 0.328125,
+      "learning_rate": 9.758793969849247e-05,
+      "loss": 1.4323,
+      "step": 580
+    },
+    {
+      "epoch": 0.04957919551141879,
+      "grad_norm": 0.328125,
+      "learning_rate": 9.758291457286433e-05,
+      "loss": 1.3754,
+      "step": 581
+    },
+    {
+      "epoch": 0.04966452975498405,
+      "grad_norm": 0.30859375,
+      "learning_rate": 9.757788944723618e-05,
+      "loss": 1.2542,
+      "step": 582
+    },
+    {
+      "epoch": 0.04974986399854932,
+      "grad_norm": 0.333984375,
+      "learning_rate": 9.757286432160805e-05,
+      "loss": 1.3337,
+      "step": 583
+    },
+    {
+      "epoch": 0.04983519824211458,
+      "grad_norm": 0.34375,
+      "learning_rate": 9.75678391959799e-05,
+      "loss": 1.5184,
+      "step": 584
+    },
+    {
+      "epoch": 0.04992053248567985,
+      "grad_norm": 0.357421875,
+      "learning_rate": 9.756281407035176e-05,
+      "loss": 1.3454,
+      "step": 585
+    },
+    {
+      "epoch": 0.05000586672924511,
+      "grad_norm": 0.271484375,
+      "learning_rate": 9.755778894472363e-05,
+      "loss": 1.5811,
+      "step": 586
+    },
+    {
+      "epoch": 0.050091200972810376,
+      "grad_norm": 0.310546875,
+      "learning_rate": 9.755276381909549e-05,
+      "loss": 1.3913,
+      "step": 587
+    },
+    {
+      "epoch": 0.05017653521637564,
+      "grad_norm": 0.30078125,
+      "learning_rate": 9.754773869346734e-05,
+      "loss": 1.2978,
+      "step": 588
+    },
+    {
+      "epoch": 0.050261869459940905,
+      "grad_norm": 0.3359375,
+      "learning_rate": 9.754271356783921e-05,
+      "loss": 1.4808,
+      "step": 589
+    },
+    {
+      "epoch": 0.05034720370350617,
+      "grad_norm": 0.388671875,
+      "learning_rate": 9.753768844221106e-05,
+      "loss": 1.3697,
+      "step": 590
+    },
+    {
+      "epoch": 0.050432537947071435,
+      "grad_norm": 0.310546875,
+      "learning_rate": 9.753266331658292e-05,
+      "loss": 1.428,
+      "step": 591
+    },
+    {
+      "epoch": 0.0505178721906367,
+      "grad_norm": 0.341796875,
+      "learning_rate": 9.752763819095477e-05,
+      "loss": 1.3562,
+      "step": 592
+    },
+    {
+      "epoch": 0.050603206434201964,
+      "grad_norm": 0.314453125,
+      "learning_rate": 9.752261306532665e-05,
+      "loss": 1.1277,
+      "step": 593
+    },
+    {
+      "epoch": 0.05068854067776723,
+      "grad_norm": 0.32421875,
+      "learning_rate": 9.75175879396985e-05,
+      "loss": 1.2116,
+      "step": 594
+    },
+    {
+      "epoch": 0.05077387492133249,
+      "grad_norm": 0.35546875,
+      "learning_rate": 9.751256281407036e-05,
+      "loss": 1.4622,
+      "step": 595
+    },
+    {
+      "epoch": 0.05085920916489776,
+      "grad_norm": 0.28515625,
+      "learning_rate": 9.750753768844222e-05,
+      "loss": 1.3907,
+      "step": 596
+    },
+    {
+      "epoch": 0.05094454340846302,
+      "grad_norm": 0.35546875,
+      "learning_rate": 9.750251256281408e-05,
+      "loss": 1.5221,
+      "step": 597
+    },
+    {
+      "epoch": 0.05102987765202829,
+      "grad_norm": 0.306640625,
+      "learning_rate": 9.749748743718593e-05,
+      "loss": 1.318,
+      "step": 598
+    },
+    {
+      "epoch": 0.05111521189559355,
+      "grad_norm": 0.31640625,
+      "learning_rate": 9.749246231155779e-05,
+      "loss": 1.4771,
+      "step": 599
+    },
+    {
+      "epoch": 0.051200546139158816,
+      "grad_norm": 0.322265625,
+      "learning_rate": 9.748743718592965e-05,
+      "loss": 1.2633,
+      "step": 600
+    },
+    {
+      "epoch": 0.05128588038272408,
+      "grad_norm": 0.337890625,
+      "learning_rate": 9.748241206030152e-05,
+      "loss": 1.4679,
+      "step": 601
+    },
+    {
+      "epoch": 0.051371214626289345,
+      "grad_norm": 0.328125,
+      "learning_rate": 9.747738693467338e-05,
+      "loss": 1.4968,
+      "step": 602
+    },
+    {
+      "epoch": 0.05145654886985461,
+      "grad_norm": 0.32421875,
+      "learning_rate": 9.747236180904523e-05,
+      "loss": 1.3159,
+      "step": 603
+    },
+    {
+      "epoch": 0.051541883113419874,
+      "grad_norm": 0.294921875,
+      "learning_rate": 9.746733668341709e-05,
+      "loss": 1.3781,
+      "step": 604
+    },
+    {
+      "epoch": 0.05162721735698514,
+      "grad_norm": 0.314453125,
+      "learning_rate": 9.746231155778895e-05,
+      "loss": 1.4195,
+      "step": 605
+    },
+    {
+      "epoch": 0.051712551600550404,
+      "grad_norm": 0.314453125,
+      "learning_rate": 9.745728643216081e-05,
+      "loss": 1.4356,
+      "step": 606
+    },
+    {
+      "epoch": 0.05179788584411567,
+      "grad_norm": 0.357421875,
+      "learning_rate": 9.745226130653266e-05,
+      "loss": 1.3743,
+      "step": 607
+    },
+    {
+      "epoch": 0.05188322008768093,
+      "grad_norm": 0.310546875,
+      "learning_rate": 9.744723618090452e-05,
+      "loss": 1.3155,
+      "step": 608
+    },
+    {
+      "epoch": 0.0519685543312462,
+      "grad_norm": 0.337890625,
+      "learning_rate": 9.744221105527638e-05,
+      "loss": 1.4083,
+      "step": 609
+    },
+    {
+      "epoch": 0.05205388857481146,
+      "grad_norm": 0.30859375,
+      "learning_rate": 9.743718592964825e-05,
+      "loss": 1.3564,
+      "step": 610
+    },
+    {
+      "epoch": 0.052139222818376726,
+      "grad_norm": 0.333984375,
+      "learning_rate": 9.74321608040201e-05,
+      "loss": 1.5243,
+      "step": 611
+    },
+    {
+      "epoch": 0.05222455706194199,
+      "grad_norm": 0.298828125,
+      "learning_rate": 9.742713567839197e-05,
+      "loss": 1.3934,
+      "step": 612
+    },
+    {
+      "epoch": 0.052309891305507256,
+      "grad_norm": 0.330078125,
+      "learning_rate": 9.742211055276382e-05,
+      "loss": 1.3353,
+      "step": 613
+    },
+    {
+      "epoch": 0.05239522554907252,
+      "grad_norm": 0.32421875,
+      "learning_rate": 9.741708542713568e-05,
+      "loss": 1.353,
+      "step": 614
+    },
+    {
+      "epoch": 0.052480559792637785,
+      "grad_norm": 0.29296875,
+      "learning_rate": 9.741206030150754e-05,
+      "loss": 1.3131,
+      "step": 615
+    },
+    {
+      "epoch": 0.05256589403620305,
+      "grad_norm": 0.333984375,
+      "learning_rate": 9.74070351758794e-05,
+      "loss": 1.3904,
+      "step": 616
+    },
+    {
+      "epoch": 0.05265122827976832,
+      "grad_norm": 0.3359375,
+      "learning_rate": 9.740201005025125e-05,
+      "loss": 1.3883,
+      "step": 617
+    },
+    {
+      "epoch": 0.052736562523333586,
+      "grad_norm": 0.3046875,
+      "learning_rate": 9.739698492462313e-05,
+      "loss": 1.3321,
+      "step": 618
+    },
+    {
+      "epoch": 0.05282189676689885,
+      "grad_norm": 0.30859375,
+      "learning_rate": 9.739195979899498e-05,
+      "loss": 1.4628,
+      "step": 619
+    },
+    {
+      "epoch": 0.052907231010464115,
+      "grad_norm": 0.30078125,
+      "learning_rate": 9.738693467336684e-05,
+      "loss": 1.381,
+      "step": 620
+    },
+    {
+      "epoch": 0.05299256525402938,
+      "grad_norm": 0.38671875,
+      "learning_rate": 9.73819095477387e-05,
+      "loss": 1.3673,
+      "step": 621
+    },
+    {
+      "epoch": 0.053077899497594644,
+      "grad_norm": 0.341796875,
+      "learning_rate": 9.737688442211056e-05,
+      "loss": 1.3133,
+      "step": 622
+    },
+    {
+      "epoch": 0.05316323374115991,
+      "grad_norm": 0.3046875,
+      "learning_rate": 9.737185929648241e-05,
+      "loss": 1.5619,
+      "step": 623
+    },
+    {
+      "epoch": 0.05324856798472517,
+      "grad_norm": 0.3046875,
+      "learning_rate": 9.736683417085427e-05,
+      "loss": 1.3775,
+      "step": 624
+    },
+    {
+      "epoch": 0.05333390222829044,
+      "grad_norm": 0.330078125,
+      "learning_rate": 9.736180904522614e-05,
+      "loss": 1.2801,
+      "step": 625
+    },
+    {
+      "epoch": 0.0534192364718557,
+      "grad_norm": 0.3359375,
+      "learning_rate": 9.7356783919598e-05,
+      "loss": 1.2728,
+      "step": 626
+    },
+    {
+      "epoch": 0.05350457071542097,
+      "grad_norm": 0.32421875,
+      "learning_rate": 9.735175879396985e-05,
+      "loss": 1.3673,
+      "step": 627
+    },
+    {
+      "epoch": 0.05358990495898623,
+      "grad_norm": 0.353515625,
+      "learning_rate": 9.734673366834172e-05,
+      "loss": 1.5111,
+      "step": 628
+    },
+    {
+      "epoch": 0.053675239202551496,
+      "grad_norm": 0.34375,
+      "learning_rate": 9.734170854271357e-05,
+      "loss": 1.3328,
+      "step": 629
+    },
+    {
+      "epoch": 0.05376057344611676,
+      "grad_norm": 0.28125,
+      "learning_rate": 9.733668341708543e-05,
+      "loss": 1.3164,
+      "step": 630
+    },
+    {
+      "epoch": 0.053845907689682025,
+      "grad_norm": 0.314453125,
+      "learning_rate": 9.73316582914573e-05,
+      "loss": 1.4929,
+      "step": 631
+    },
+    {
+      "epoch": 0.05393124193324729,
+      "grad_norm": 0.359375,
+      "learning_rate": 9.732663316582916e-05,
+      "loss": 1.1437,
+      "step": 632
+    },
+    {
+      "epoch": 0.054016576176812554,
+      "grad_norm": 0.3515625,
+      "learning_rate": 9.7321608040201e-05,
+      "loss": 1.4573,
+      "step": 633
+    },
+    {
+      "epoch": 0.05410191042037782,
+      "grad_norm": 0.345703125,
+      "learning_rate": 9.731658291457287e-05,
+      "loss": 1.5108,
+      "step": 634
+    },
+    {
+      "epoch": 0.054187244663943084,
+      "grad_norm": 0.349609375,
+      "learning_rate": 9.731155778894473e-05,
+      "loss": 1.3038,
+      "step": 635
+    },
+    {
+      "epoch": 0.05427257890750835,
+      "grad_norm": 0.318359375,
+      "learning_rate": 9.730653266331659e-05,
+      "loss": 1.2705,
+      "step": 636
+    },
+    {
+      "epoch": 0.05435791315107361,
+      "grad_norm": 0.361328125,
+      "learning_rate": 9.730150753768845e-05,
+      "loss": 1.3002,
+      "step": 637
+    },
+    {
+      "epoch": 0.05444324739463888,
+      "grad_norm": 0.34765625,
+      "learning_rate": 9.72964824120603e-05,
+      "loss": 1.2631,
+      "step": 638
+    },
+    {
+      "epoch": 0.05452858163820414,
+      "grad_norm": 0.3359375,
+      "learning_rate": 9.729145728643216e-05,
+      "loss": 1.4492,
+      "step": 639
+    },
+    {
+      "epoch": 0.054613915881769406,
+      "grad_norm": 0.3515625,
+      "learning_rate": 9.728643216080403e-05,
+      "loss": 1.2772,
+      "step": 640
+    },
+    {
+      "epoch": 0.05469925012533467,
+      "grad_norm": 0.31640625,
+      "learning_rate": 9.728140703517589e-05,
+      "loss": 1.2737,
+      "step": 641
+    },
+    {
+      "epoch": 0.054784584368899936,
+      "grad_norm": 0.3125,
+      "learning_rate": 9.727638190954774e-05,
+      "loss": 1.3252,
+      "step": 642
+    },
+    {
+      "epoch": 0.0548699186124652,
+      "grad_norm": 0.326171875,
+      "learning_rate": 9.72713567839196e-05,
+      "loss": 1.5284,
+      "step": 643
+    },
+    {
+      "epoch": 0.054955252856030465,
+      "grad_norm": 0.345703125,
+      "learning_rate": 9.726633165829146e-05,
+      "loss": 1.2824,
+      "step": 644
+    },
+    {
+      "epoch": 0.05504058709959573,
+      "grad_norm": 0.41015625,
+      "learning_rate": 9.726130653266332e-05,
+      "loss": 1.519,
+      "step": 645
+    },
+    {
+      "epoch": 0.055125921343160994,
+      "grad_norm": 0.333984375,
+      "learning_rate": 9.725628140703517e-05,
+      "loss": 1.4557,
+      "step": 646
+    },
+    {
+      "epoch": 0.05521125558672626,
+      "grad_norm": 0.376953125,
+      "learning_rate": 9.725125628140705e-05,
+      "loss": 1.2523,
+      "step": 647
+    },
+    {
+      "epoch": 0.05529658983029152,
+      "grad_norm": 0.32421875,
+      "learning_rate": 9.72462311557789e-05,
+      "loss": 1.2909,
+      "step": 648
+    },
+    {
+      "epoch": 0.05538192407385679,
+      "grad_norm": 0.314453125,
+      "learning_rate": 9.724120603015076e-05,
+      "loss": 1.4421,
+      "step": 649
+    },
+    {
+      "epoch": 0.05546725831742205,
+      "grad_norm": 0.322265625,
+      "learning_rate": 9.723618090452262e-05,
+      "loss": 1.4081,
+      "step": 650
+    },
+    {
+      "epoch": 0.05555259256098732,
+      "grad_norm": 0.353515625,
+      "learning_rate": 9.723115577889448e-05,
+      "loss": 1.3829,
+      "step": 651
+    },
+    {
+      "epoch": 0.05563792680455258,
+      "grad_norm": 0.345703125,
+      "learning_rate": 9.722613065326633e-05,
+      "loss": 1.4267,
+      "step": 652
+    },
+    {
+      "epoch": 0.055723261048117846,
+      "grad_norm": 0.337890625,
+      "learning_rate": 9.72211055276382e-05,
+      "loss": 1.4874,
+      "step": 653
+    },
+    {
+      "epoch": 0.05580859529168311,
+      "grad_norm": 0.326171875,
+      "learning_rate": 9.721608040201005e-05,
+      "loss": 1.31,
+      "step": 654
+    },
+    {
+      "epoch": 0.055893929535248375,
+      "grad_norm": 0.322265625,
+      "learning_rate": 9.721105527638192e-05,
+      "loss": 1.3142,
+      "step": 655
+    },
+    {
+      "epoch": 0.05597926377881364,
+      "grad_norm": 0.3515625,
+      "learning_rate": 9.720603015075376e-05,
+      "loss": 1.2545,
+      "step": 656
+    },
+    {
+      "epoch": 0.056064598022378904,
+      "grad_norm": 0.376953125,
+      "learning_rate": 9.720100502512564e-05,
+      "loss": 1.3988,
+      "step": 657
+    },
+    {
+      "epoch": 0.05614993226594417,
+      "grad_norm": 0.38671875,
+      "learning_rate": 9.719597989949749e-05,
+      "loss": 1.4185,
+      "step": 658
+    },
+    {
+      "epoch": 0.056235266509509434,
+      "grad_norm": 0.31640625,
+      "learning_rate": 9.719095477386935e-05,
+      "loss": 2.0381,
+      "step": 659
+    },
+    {
+      "epoch": 0.0563206007530747,
+      "grad_norm": 0.373046875,
+      "learning_rate": 9.718592964824121e-05,
+      "loss": 1.5739,
+      "step": 660
+    },
+    {
+      "epoch": 0.05640593499663996,
+      "grad_norm": 0.29296875,
+      "learning_rate": 9.718090452261307e-05,
+      "loss": 1.3885,
+      "step": 661
+    },
+    {
+      "epoch": 0.05649126924020523,
+      "grad_norm": 0.3046875,
+      "learning_rate": 9.717587939698492e-05,
+      "loss": 1.3189,
+      "step": 662
+    },
+    {
+      "epoch": 0.05657660348377049,
+      "grad_norm": 0.326171875,
+      "learning_rate": 9.71708542713568e-05,
+      "loss": 1.42,
+      "step": 663
+    },
+    {
+      "epoch": 0.05666193772733576,
+      "grad_norm": 0.306640625,
+      "learning_rate": 9.716582914572865e-05,
+      "loss": 1.2109,
+      "step": 664
+    },
+    {
+      "epoch": 0.05674727197090102,
+      "grad_norm": 0.30859375,
+      "learning_rate": 9.716080402010051e-05,
+      "loss": 1.3912,
+      "step": 665
+    },
+    {
+      "epoch": 0.056832606214466286,
+      "grad_norm": 0.29296875,
+      "learning_rate": 9.715577889447237e-05,
+      "loss": 1.3246,
+      "step": 666
+    },
+    {
+      "epoch": 0.05691794045803155,
+      "grad_norm": 0.318359375,
+      "learning_rate": 9.715075376884423e-05,
+      "loss": 1.4655,
+      "step": 667
+    },
+    {
+      "epoch": 0.057003274701596815,
+      "grad_norm": 0.345703125,
+      "learning_rate": 9.714572864321608e-05,
+      "loss": 1.3685,
+      "step": 668
+    },
+    {
+      "epoch": 0.05708860894516208,
+      "grad_norm": 0.330078125,
+      "learning_rate": 9.714070351758794e-05,
+      "loss": 1.4323,
+      "step": 669
+    },
+    {
+      "epoch": 0.057173943188727344,
+      "grad_norm": 0.412109375,
+      "learning_rate": 9.71356783919598e-05,
+      "loss": 1.3601,
+      "step": 670
+    },
+    {
+      "epoch": 0.05725927743229261,
+      "grad_norm": 0.3359375,
+      "learning_rate": 9.713065326633167e-05,
+      "loss": 1.4362,
+      "step": 671
+    },
+    {
+      "epoch": 0.05734461167585787,
+      "grad_norm": 0.330078125,
+      "learning_rate": 9.712562814070352e-05,
+      "loss": 2.746,
+      "step": 672
+    },
+    {
+      "epoch": 0.05742994591942314,
+      "grad_norm": 0.314453125,
+      "learning_rate": 9.712060301507538e-05,
+      "loss": 1.3721,
+      "step": 673
+    },
+    {
+      "epoch": 0.0575152801629884,
+      "grad_norm": 0.2890625,
+      "learning_rate": 9.711557788944724e-05,
+      "loss": 1.2796,
+      "step": 674
+    },
+    {
+      "epoch": 0.05760061440655367,
+      "grad_norm": 0.322265625,
+      "learning_rate": 9.71105527638191e-05,
+      "loss": 1.5044,
+      "step": 675
+    },
+    {
+      "epoch": 0.05768594865011893,
+      "grad_norm": 0.328125,
+      "learning_rate": 9.710552763819096e-05,
+      "loss": 1.2025,
+      "step": 676
+    },
+    {
+      "epoch": 0.057771282893684196,
+      "grad_norm": 0.46484375,
+      "learning_rate": 9.710050251256281e-05,
+      "loss": 1.5289,
+      "step": 677
+    },
+    {
+      "epoch": 0.05785661713724946,
+      "grad_norm": 0.3203125,
+      "learning_rate": 9.709547738693468e-05,
+      "loss": 1.413,
+      "step": 678
+    },
+    {
+      "epoch": 0.057941951380814725,
+      "grad_norm": 0.33203125,
+      "learning_rate": 9.709045226130654e-05,
+      "loss": 1.4112,
+      "step": 679
+    },
+    {
+      "epoch": 0.05802728562437999,
+      "grad_norm": 0.28515625,
+      "learning_rate": 9.70854271356784e-05,
+      "loss": 1.3151,
+      "step": 680
+    },
+    {
+      "epoch": 0.058112619867945255,
+      "grad_norm": 0.37109375,
+      "learning_rate": 9.708040201005025e-05,
+      "loss": 1.3537,
+      "step": 681
+    },
+    {
+      "epoch": 0.058197954111510526,
+      "grad_norm": 0.34765625,
+      "learning_rate": 9.707537688442212e-05,
+      "loss": 1.3003,
+      "step": 682
+    },
+    {
+      "epoch": 0.05828328835507579,
+      "grad_norm": 0.333984375,
+      "learning_rate": 9.707035175879397e-05,
+      "loss": 1.3926,
+      "step": 683
+    },
+    {
+      "epoch": 0.058368622598641055,
+      "grad_norm": 0.310546875,
+      "learning_rate": 9.706532663316583e-05,
+      "loss": 1.3556,
+      "step": 684
+    },
+    {
+      "epoch": 0.05845395684220632,
+      "grad_norm": 0.306640625,
+      "learning_rate": 9.706030150753768e-05,
+      "loss": 1.2722,
+      "step": 685
+    },
+    {
+      "epoch": 0.058539291085771585,
+      "grad_norm": 0.34765625,
+      "learning_rate": 9.705527638190956e-05,
+      "loss": 1.3896,
+      "step": 686
+    },
+    {
+      "epoch": 0.05862462532933685,
+      "grad_norm": 0.333984375,
+      "learning_rate": 9.70502512562814e-05,
+      "loss": 1.2862,
+      "step": 687
+    },
+    {
+      "epoch": 0.058709959572902114,
+      "grad_norm": 0.322265625,
+      "learning_rate": 9.704522613065327e-05,
+      "loss": 1.46,
+      "step": 688
+    },
+    {
+      "epoch": 0.05879529381646738,
+      "grad_norm": 0.31640625,
+      "learning_rate": 9.704020100502513e-05,
+      "loss": 1.4346,
+      "step": 689
+    },
+    {
+      "epoch": 0.05888062806003264,
+      "grad_norm": 0.3671875,
+      "learning_rate": 9.703517587939699e-05,
+      "loss": 1.3685,
+      "step": 690
+    },
+    {
+      "epoch": 0.05896596230359791,
+      "grad_norm": 0.33984375,
+      "learning_rate": 9.703015075376884e-05,
+      "loss": 1.3749,
+      "step": 691
+    },
+    {
+      "epoch": 0.05905129654716317,
+      "grad_norm": 0.3046875,
+      "learning_rate": 9.702512562814072e-05,
+      "loss": 1.1969,
+      "step": 692
+    },
+    {
+      "epoch": 0.05913663079072844,
+      "grad_norm": 0.35546875,
+      "learning_rate": 9.702010050251257e-05,
+      "loss": 1.3639,
+      "step": 693
+    },
+    {
+      "epoch": 0.0592219650342937,
+      "grad_norm": 0.34765625,
+      "learning_rate": 9.701507537688443e-05,
+      "loss": 1.5146,
+      "step": 694
+    },
+    {
+      "epoch": 0.059307299277858966,
+      "grad_norm": 0.314453125,
+      "learning_rate": 9.701005025125629e-05,
+      "loss": 1.3232,
+      "step": 695
+    },
+    {
+      "epoch": 0.05939263352142423,
+      "grad_norm": 0.322265625,
+      "learning_rate": 9.700502512562815e-05,
+      "loss": 1.2522,
+      "step": 696
+    },
+    {
+      "epoch": 0.059477967764989495,
+      "grad_norm": 0.28515625,
+      "learning_rate": 9.7e-05,
+      "loss": 1.4553,
+      "step": 697
+    },
+    {
+      "epoch": 0.05956330200855476,
+      "grad_norm": 0.310546875,
+      "learning_rate": 9.699497487437188e-05,
+      "loss": 1.0855,
+      "step": 698
+    },
+    {
+      "epoch": 0.059648636252120024,
+      "grad_norm": 0.28125,
+      "learning_rate": 9.698994974874372e-05,
+      "loss": 1.3087,
+      "step": 699
+    },
+    {
+      "epoch": 0.05973397049568529,
+      "grad_norm": 0.369140625,
+      "learning_rate": 9.698492462311559e-05,
+      "loss": 1.4624,
+      "step": 700
+    },
+    {
+      "epoch": 0.05981930473925055,
+      "grad_norm": 0.35546875,
+      "learning_rate": 9.697989949748743e-05,
+      "loss": 1.4732,
+      "step": 701
+    },
+    {
+      "epoch": 0.05990463898281582,
+      "grad_norm": 0.3125,
+      "learning_rate": 9.697487437185931e-05,
+      "loss": 1.3887,
+      "step": 702
+    },
+    {
+      "epoch": 0.05998997322638108,
+      "grad_norm": 0.310546875,
+      "learning_rate": 9.696984924623116e-05,
+      "loss": 1.4674,
+      "step": 703
+    },
+    {
+      "epoch": 0.06007530746994635,
+      "grad_norm": 0.322265625,
+      "learning_rate": 9.696482412060302e-05,
+      "loss": 1.3927,
+      "step": 704
+    },
+    {
+      "epoch": 0.06016064171351161,
+      "grad_norm": 0.298828125,
+      "learning_rate": 9.695979899497488e-05,
+      "loss": 1.2565,
+      "step": 705
+    },
+    {
+      "epoch": 0.060245975957076876,
+      "grad_norm": 0.302734375,
+      "learning_rate": 9.695477386934674e-05,
+      "loss": 1.2928,
+      "step": 706
+    },
+    {
+      "epoch": 0.06033131020064214,
+      "grad_norm": 0.322265625,
+      "learning_rate": 9.694974874371859e-05,
+      "loss": 1.3397,
+      "step": 707
+    },
+    {
+      "epoch": 0.060416644444207405,
+      "grad_norm": 0.322265625,
+      "learning_rate": 9.694472361809045e-05,
+      "loss": 1.3858,
+      "step": 708
+    },
+    {
+      "epoch": 0.06050197868777267,
+      "grad_norm": 0.318359375,
+      "learning_rate": 9.693969849246232e-05,
+      "loss": 1.4811,
+      "step": 709
+    },
+    {
+      "epoch": 0.060587312931337935,
+      "grad_norm": 0.322265625,
+      "learning_rate": 9.693467336683418e-05,
+      "loss": 1.4033,
+      "step": 710
+    },
+    {
+      "epoch": 0.0606726471749032,
+      "grad_norm": 0.330078125,
+      "learning_rate": 9.692964824120604e-05,
+      "loss": 1.46,
+      "step": 711
+    },
+    {
+      "epoch": 0.060757981418468464,
+      "grad_norm": 0.34765625,
+      "learning_rate": 9.692462311557789e-05,
+      "loss": 1.3448,
+      "step": 712
+    },
+    {
+      "epoch": 0.06084331566203373,
+      "grad_norm": 0.326171875,
+      "learning_rate": 9.691959798994975e-05,
+      "loss": 1.4918,
+      "step": 713
+    },
+    {
+      "epoch": 0.06092864990559899,
+      "grad_norm": 0.30078125,
+      "learning_rate": 9.691457286432161e-05,
+      "loss": 1.2975,
+      "step": 714
+    },
+    {
+      "epoch": 0.06101398414916426,
+      "grad_norm": 0.34765625,
+      "learning_rate": 9.690954773869348e-05,
+      "loss": 1.3874,
+      "step": 715
+    },
+    {
+      "epoch": 0.06109931839272952,
+      "grad_norm": 0.365234375,
+      "learning_rate": 9.690452261306532e-05,
+      "loss": 1.4599,
+      "step": 716
+    },
+    {
+      "epoch": 0.06118465263629479,
+      "grad_norm": 0.302734375,
+      "learning_rate": 9.689949748743719e-05,
+      "loss": 1.3242,
+      "step": 717
+    },
+    {
+      "epoch": 0.06126998687986005,
+      "grad_norm": 0.32421875,
+      "learning_rate": 9.689447236180905e-05,
+      "loss": 1.4184,
+      "step": 718
+    },
+    {
+      "epoch": 0.061355321123425316,
+      "grad_norm": 0.337890625,
+      "learning_rate": 9.688944723618091e-05,
+      "loss": 1.5936,
+      "step": 719
+    },
+    {
+      "epoch": 0.06144065536699058,
+      "grad_norm": 0.2734375,
+      "learning_rate": 9.688442211055276e-05,
+      "loss": 1.1906,
+      "step": 720
+    },
+    {
+      "epoch": 0.061525989610555845,
+      "grad_norm": 0.294921875,
+      "learning_rate": 9.687939698492463e-05,
+      "loss": 1.2857,
+      "step": 721
+    },
+    {
+      "epoch": 0.06161132385412111,
+      "grad_norm": 0.298828125,
+      "learning_rate": 9.687437185929648e-05,
+      "loss": 1.2152,
+      "step": 722
+    },
+    {
+      "epoch": 0.061696658097686374,
+      "grad_norm": 0.3046875,
+      "learning_rate": 9.686934673366834e-05,
+      "loss": 1.3036,
+      "step": 723
+    },
+    {
+      "epoch": 0.06178199234125164,
+      "grad_norm": 0.294921875,
+      "learning_rate": 9.686432160804021e-05,
+      "loss": 1.3545,
+      "step": 724
+    },
+    {
+      "epoch": 0.061867326584816904,
+      "grad_norm": 0.353515625,
+      "learning_rate": 9.685929648241207e-05,
+      "loss": 1.6225,
+      "step": 725
+    },
+    {
+      "epoch": 0.06195266082838217,
+      "grad_norm": 0.30859375,
+      "learning_rate": 9.685427135678392e-05,
+      "loss": 1.1705,
+      "step": 726
+    },
+    {
+      "epoch": 0.06203799507194743,
+      "grad_norm": 0.298828125,
+      "learning_rate": 9.684924623115579e-05,
+      "loss": 1.2082,
+      "step": 727
+    },
+    {
+      "epoch": 0.0621233293155127,
+      "grad_norm": 0.318359375,
+      "learning_rate": 9.684422110552764e-05,
+      "loss": 1.2985,
+      "step": 728
+    },
+    {
+      "epoch": 0.06220866355907796,
+      "grad_norm": 0.306640625,
+      "learning_rate": 9.68391959798995e-05,
+      "loss": 1.2335,
+      "step": 729
+    },
+    {
+      "epoch": 0.062293997802643226,
+      "grad_norm": 0.3125,
+      "learning_rate": 9.683417085427137e-05,
+      "loss": 1.2785,
+      "step": 730
+    },
+    {
+      "epoch": 0.06237933204620849,
+      "grad_norm": 0.380859375,
+      "learning_rate": 9.682914572864323e-05,
+      "loss": 1.4843,
+      "step": 731
+    },
+    {
+      "epoch": 0.062464666289773756,
+      "grad_norm": 0.333984375,
+      "learning_rate": 9.682412060301508e-05,
+      "loss": 1.2533,
+      "step": 732
+    },
+    {
+      "epoch": 0.06255000053333902,
+      "grad_norm": 0.337890625,
+      "learning_rate": 9.681909547738694e-05,
+      "loss": 1.4767,
+      "step": 733
+    },
+    {
+      "epoch": 0.06263533477690429,
+      "grad_norm": 0.373046875,
+      "learning_rate": 9.68140703517588e-05,
+      "loss": 1.2843,
+      "step": 734
+    },
+    {
+      "epoch": 0.06272066902046955,
+      "grad_norm": 0.32421875,
+      "learning_rate": 9.680904522613066e-05,
+      "loss": 1.514,
+      "step": 735
+    },
+    {
+      "epoch": 0.06280600326403482,
+      "grad_norm": 0.32421875,
+      "learning_rate": 9.680402010050251e-05,
+      "loss": 1.4528,
+      "step": 736
+    },
+    {
+      "epoch": 0.06289133750760008,
+      "grad_norm": 0.3046875,
+      "learning_rate": 9.679899497487439e-05,
+      "loss": 1.3571,
+      "step": 737
+    },
+    {
+      "epoch": 0.06297667175116535,
+      "grad_norm": 0.35546875,
+      "learning_rate": 9.679396984924623e-05,
+      "loss": 1.2832,
+      "step": 738
+    },
+    {
+      "epoch": 0.06306200599473061,
+      "grad_norm": 0.3203125,
+      "learning_rate": 9.67889447236181e-05,
+      "loss": 1.4314,
+      "step": 739
+    },
+    {
+      "epoch": 0.06314734023829588,
+      "grad_norm": 0.38671875,
+      "learning_rate": 9.678391959798996e-05,
+      "loss": 1.5038,
+      "step": 740
+    },
+    {
+      "epoch": 0.06323267448186114,
+      "grad_norm": 0.314453125,
+      "learning_rate": 9.677889447236182e-05,
+      "loss": 1.2716,
+      "step": 741
+    },
+    {
+      "epoch": 0.06331800872542641,
+      "grad_norm": 0.328125,
+      "learning_rate": 9.677386934673367e-05,
+      "loss": 1.2287,
+      "step": 742
+    },
+    {
+      "epoch": 0.06340334296899167,
+      "grad_norm": 0.326171875,
+      "learning_rate": 9.676884422110553e-05,
+      "loss": 1.3997,
+      "step": 743
+    },
+    {
+      "epoch": 0.06348867721255694,
+      "grad_norm": 0.34765625,
+      "learning_rate": 9.67638190954774e-05,
+      "loss": 1.5631,
+      "step": 744
+    },
+    {
+      "epoch": 0.0635740114561222,
+      "grad_norm": 0.33984375,
+      "learning_rate": 9.675879396984924e-05,
+      "loss": 1.4241,
+      "step": 745
+    },
+    {
+      "epoch": 0.06365934569968747,
+      "grad_norm": 0.345703125,
+      "learning_rate": 9.675376884422112e-05,
+      "loss": 1.2748,
+      "step": 746
+    },
+    {
+      "epoch": 0.06374467994325272,
+      "grad_norm": 0.361328125,
+      "learning_rate": 9.674874371859297e-05,
+      "loss": 1.3819,
+      "step": 747
+    },
+    {
+      "epoch": 0.063830014186818,
+      "grad_norm": 0.35546875,
+      "learning_rate": 9.674371859296483e-05,
+      "loss": 1.3393,
+      "step": 748
+    },
+    {
+      "epoch": 0.06391534843038325,
+      "grad_norm": 0.333984375,
+      "learning_rate": 9.673869346733668e-05,
+      "loss": 1.3685,
+      "step": 749
+    },
+    {
+      "epoch": 0.06400068267394853,
+      "grad_norm": 0.34375,
+      "learning_rate": 9.673366834170855e-05,
+      "loss": 1.3799,
+      "step": 750
+    },
+    {
+      "epoch": 0.06408601691751378,
+      "grad_norm": 0.314453125,
+      "learning_rate": 9.67286432160804e-05,
+      "loss": 1.265,
+      "step": 751
+    },
+    {
+      "epoch": 0.06417135116107905,
+      "grad_norm": 0.37109375,
+      "learning_rate": 9.672361809045226e-05,
+      "loss": 1.5527,
+      "step": 752
+    },
+    {
+      "epoch": 0.06425668540464431,
+      "grad_norm": 0.32421875,
+      "learning_rate": 9.671859296482412e-05,
+      "loss": 1.3506,
+      "step": 753
+    },
+    {
+      "epoch": 0.06434201964820958,
+      "grad_norm": 0.33984375,
+      "learning_rate": 9.671356783919599e-05,
+      "loss": 1.4867,
+      "step": 754
+    },
+    {
+      "epoch": 0.06442735389177484,
+      "grad_norm": 0.3671875,
+      "learning_rate": 9.670854271356784e-05,
+      "loss": 1.3884,
+      "step": 755
+    },
+    {
+      "epoch": 0.06451268813534011,
+      "grad_norm": 0.353515625,
+      "learning_rate": 9.670351758793971e-05,
+      "loss": 1.3691,
+      "step": 756
+    },
+    {
+      "epoch": 0.06459802237890537,
+      "grad_norm": 0.34375,
+      "learning_rate": 9.669849246231156e-05,
+      "loss": 1.3279,
+      "step": 757
+    },
+    {
+      "epoch": 0.06468335662247064,
+      "grad_norm": 0.34375,
+      "learning_rate": 9.669346733668342e-05,
+      "loss": 1.3145,
+      "step": 758
+    },
+    {
+      "epoch": 0.0647686908660359,
+      "grad_norm": 0.494140625,
+      "learning_rate": 9.668844221105528e-05,
+      "loss": 1.4105,
+      "step": 759
+    },
+    {
+      "epoch": 0.06485402510960117,
+      "grad_norm": 0.412109375,
+      "learning_rate": 9.668341708542715e-05,
+      "loss": 1.3765,
+      "step": 760
+    },
+    {
+      "epoch": 0.06493935935316643,
+      "grad_norm": 0.349609375,
+      "learning_rate": 9.6678391959799e-05,
+      "loss": 1.3581,
+      "step": 761
+    },
+    {
+      "epoch": 0.0650246935967317,
+      "grad_norm": 0.330078125,
+      "learning_rate": 9.667336683417087e-05,
+      "loss": 1.3577,
+      "step": 762
+    },
+    {
+      "epoch": 0.06511002784029696,
+      "grad_norm": 0.31640625,
+      "learning_rate": 9.666834170854272e-05,
+      "loss": 1.3019,
+      "step": 763
+    },
+    {
+      "epoch": 0.06519536208386223,
+      "grad_norm": 0.32421875,
+      "learning_rate": 9.666331658291458e-05,
+      "loss": 1.4107,
+      "step": 764
+    },
+    {
+      "epoch": 0.06528069632742749,
+      "grad_norm": 0.30078125,
+      "learning_rate": 9.665829145728643e-05,
+      "loss": 1.2924,
+      "step": 765
+    },
+    {
+      "epoch": 0.06536603057099276,
+      "grad_norm": 0.33203125,
+      "learning_rate": 9.66532663316583e-05,
+      "loss": 1.3826,
+      "step": 766
+    },
+    {
+      "epoch": 0.06545136481455802,
+      "grad_norm": 0.318359375,
+      "learning_rate": 9.664824120603015e-05,
+      "loss": 1.2879,
+      "step": 767
+    },
+    {
+      "epoch": 0.06553669905812329,
+      "grad_norm": 0.33984375,
+      "learning_rate": 9.664321608040201e-05,
+      "loss": 1.4199,
+      "step": 768
+    },
+    {
+      "epoch": 0.06562203330168855,
+      "grad_norm": 0.298828125,
+      "learning_rate": 9.663819095477388e-05,
+      "loss": 1.342,
+      "step": 769
+    },
+    {
+      "epoch": 0.06570736754525382,
+      "grad_norm": 0.314453125,
+      "learning_rate": 9.663316582914574e-05,
+      "loss": 1.3338,
+      "step": 770
+    },
+    {
+      "epoch": 0.06579270178881907,
+      "grad_norm": 0.306640625,
+      "learning_rate": 9.662814070351759e-05,
+      "loss": 1.3405,
+      "step": 771
+    },
+    {
+      "epoch": 0.06587803603238435,
+      "grad_norm": 0.337890625,
+      "learning_rate": 9.662311557788946e-05,
+      "loss": 1.4719,
+      "step": 772
+    },
+    {
+      "epoch": 0.0659633702759496,
+      "grad_norm": 0.31640625,
+      "learning_rate": 9.661809045226131e-05,
+      "loss": 1.3346,
+      "step": 773
+    },
+    {
+      "epoch": 0.06604870451951488,
+      "grad_norm": 0.3046875,
+      "learning_rate": 9.661306532663317e-05,
+      "loss": 1.3901,
+      "step": 774
+    },
+    {
+      "epoch": 0.06613403876308013,
+      "grad_norm": 0.3125,
+      "learning_rate": 9.660804020100504e-05,
+      "loss": 1.3265,
+      "step": 775
+    },
+    {
+      "epoch": 0.0662193730066454,
+      "grad_norm": 0.3046875,
+      "learning_rate": 9.66030150753769e-05,
+      "loss": 1.2461,
+      "step": 776
+    },
+    {
+      "epoch": 0.06630470725021066,
+      "grad_norm": 0.31640625,
+      "learning_rate": 9.659798994974875e-05,
+      "loss": 1.3661,
+      "step": 777
+    },
+    {
+      "epoch": 0.06639004149377593,
+      "grad_norm": 0.35546875,
+      "learning_rate": 9.659296482412061e-05,
+      "loss": 1.4615,
+      "step": 778
+    },
+    {
+      "epoch": 0.0664753757373412,
+      "grad_norm": 0.296875,
+      "learning_rate": 9.658793969849247e-05,
+      "loss": 1.3211,
+      "step": 779
+    },
+    {
+      "epoch": 0.06656070998090646,
+      "grad_norm": 0.310546875,
+      "learning_rate": 9.658291457286432e-05,
+      "loss": 1.3174,
+      "step": 780
+    },
+    {
+      "epoch": 0.06664604422447173,
+      "grad_norm": 0.333984375,
+      "learning_rate": 9.657788944723618e-05,
+      "loss": 1.3867,
+      "step": 781
+    },
+    {
+      "epoch": 0.06673137846803699,
+      "grad_norm": 0.310546875,
+      "learning_rate": 9.657286432160804e-05,
+      "loss": 1.318,
+      "step": 782
+    },
+    {
+      "epoch": 0.06681671271160226,
+      "grad_norm": 0.349609375,
+      "learning_rate": 9.65678391959799e-05,
+      "loss": 1.449,
+      "step": 783
+    },
+    {
+      "epoch": 0.06690204695516752,
+      "grad_norm": 0.3828125,
+      "learning_rate": 9.656281407035175e-05,
+      "loss": 1.5028,
+      "step": 784
+    },
+    {
+      "epoch": 0.06698738119873279,
+      "grad_norm": 0.28125,
+      "learning_rate": 9.655778894472363e-05,
+      "loss": 1.2669,
+      "step": 785
+    },
+    {
+      "epoch": 0.06707271544229805,
+      "grad_norm": 0.32421875,
+      "learning_rate": 9.655276381909548e-05,
+      "loss": 1.3624,
+      "step": 786
+    },
+    {
+      "epoch": 0.06715804968586332,
+      "grad_norm": 0.271484375,
+      "learning_rate": 9.654773869346734e-05,
+      "loss": 1.1874,
+      "step": 787
+    },
+    {
+      "epoch": 0.06724338392942858,
+      "grad_norm": 0.306640625,
+      "learning_rate": 9.65427135678392e-05,
+      "loss": 1.2312,
+      "step": 788
+    },
+    {
+      "epoch": 0.06732871817299385,
+      "grad_norm": 0.353515625,
+      "learning_rate": 9.653768844221106e-05,
+      "loss": 1.3271,
+      "step": 789
+    },
+    {
+      "epoch": 0.06741405241655911,
+      "grad_norm": 0.333984375,
+      "learning_rate": 9.653266331658291e-05,
+      "loss": 1.3787,
+      "step": 790
+    },
+    {
+      "epoch": 0.06749938666012438,
+      "grad_norm": 0.291015625,
+      "learning_rate": 9.652763819095479e-05,
+      "loss": 1.2081,
+      "step": 791
+    },
+    {
+      "epoch": 0.06758472090368964,
+      "grad_norm": 0.32421875,
+      "learning_rate": 9.652261306532664e-05,
+      "loss": 1.2338,
+      "step": 792
+    },
+    {
+      "epoch": 0.06767005514725491,
+      "grad_norm": 0.302734375,
+      "learning_rate": 9.65175879396985e-05,
+      "loss": 1.3317,
+      "step": 793
+    },
+    {
+      "epoch": 0.06775538939082017,
+      "grad_norm": 0.330078125,
+      "learning_rate": 9.651256281407036e-05,
+      "loss": 1.0677,
+      "step": 794
+    },
+    {
+      "epoch": 0.06784072363438544,
+      "grad_norm": 0.3125,
+      "learning_rate": 9.650753768844222e-05,
+      "loss": 1.3388,
+      "step": 795
+    },
+    {
+      "epoch": 0.0679260578779507,
+      "grad_norm": 0.328125,
+      "learning_rate": 9.650251256281407e-05,
+      "loss": 1.2781,
+      "step": 796
+    },
+    {
+      "epoch": 0.06801139212151597,
+      "grad_norm": 0.31640625,
+      "learning_rate": 9.649748743718593e-05,
+      "loss": 1.3382,
+      "step": 797
+    },
+    {
+      "epoch": 0.06809672636508123,
+      "grad_norm": 0.30859375,
+      "learning_rate": 9.64924623115578e-05,
+      "loss": 1.3792,
+      "step": 798
+    },
+    {
+      "epoch": 0.0681820606086465,
+      "grad_norm": 0.330078125,
+      "learning_rate": 9.648743718592966e-05,
+      "loss": 1.2831,
+      "step": 799
+    },
+    {
+      "epoch": 0.06826739485221175,
+      "grad_norm": 0.31640625,
+      "learning_rate": 9.64824120603015e-05,
+      "loss": 1.3086,
+      "step": 800
+    },
+    {
+      "epoch": 0.06835272909577703,
+      "grad_norm": 0.333984375,
+      "learning_rate": 9.647738693467338e-05,
+      "loss": 1.5027,
+      "step": 801
+    },
+    {
+      "epoch": 0.06843806333934228,
+      "grad_norm": 0.330078125,
+      "learning_rate": 9.647236180904523e-05,
+      "loss": 1.2856,
+      "step": 802
+    },
+    {
+      "epoch": 0.06852339758290756,
+      "grad_norm": 0.3359375,
+      "learning_rate": 9.646733668341709e-05,
+      "loss": 1.4071,
+      "step": 803
+    },
+    {
+      "epoch": 0.06860873182647281,
+      "grad_norm": 0.318359375,
+      "learning_rate": 9.646231155778895e-05,
+      "loss": 1.3371,
+      "step": 804
+    },
+    {
+      "epoch": 0.06869406607003808,
+      "grad_norm": 0.310546875,
+      "learning_rate": 9.645728643216081e-05,
+      "loss": 1.3227,
+      "step": 805
+    },
+    {
+      "epoch": 0.06877940031360334,
+      "grad_norm": 0.306640625,
+      "learning_rate": 9.645226130653266e-05,
+      "loss": 1.5008,
+      "step": 806
+    },
+    {
+      "epoch": 0.06886473455716861,
+      "grad_norm": 0.310546875,
+      "learning_rate": 9.644723618090454e-05,
+      "loss": 1.2905,
+      "step": 807
+    },
+    {
+      "epoch": 0.06895006880073387,
+      "grad_norm": 0.3125,
+      "learning_rate": 9.644221105527639e-05,
+      "loss": 1.3807,
+      "step": 808
+    },
+    {
+      "epoch": 0.06903540304429914,
+      "grad_norm": 0.3125,
+      "learning_rate": 9.643718592964825e-05,
+      "loss": 1.4242,
+      "step": 809
+    },
+    {
+      "epoch": 0.0691207372878644,
+      "grad_norm": 0.3046875,
+      "learning_rate": 9.64321608040201e-05,
+      "loss": 1.4082,
+      "step": 810
+    },
+    {
+      "epoch": 0.06920607153142967,
+      "grad_norm": 0.35546875,
+      "learning_rate": 9.642713567839196e-05,
+      "loss": 1.4116,
+      "step": 811
+    },
+    {
+      "epoch": 0.06929140577499493,
+      "grad_norm": 0.32421875,
+      "learning_rate": 9.642211055276382e-05,
+      "loss": 1.4158,
+      "step": 812
+    },
+    {
+      "epoch": 0.0693767400185602,
+      "grad_norm": 0.291015625,
+      "learning_rate": 9.641708542713568e-05,
+      "loss": 1.2911,
+      "step": 813
+    },
+    {
+      "epoch": 0.06946207426212546,
+      "grad_norm": 0.34765625,
+      "learning_rate": 9.641206030150755e-05,
+      "loss": 1.3035,
+      "step": 814
+    },
+    {
+      "epoch": 0.06954740850569073,
+      "grad_norm": 0.375,
+      "learning_rate": 9.64070351758794e-05,
+      "loss": 1.5684,
+      "step": 815
+    },
+    {
+      "epoch": 0.06963274274925599,
+      "grad_norm": 0.294921875,
+      "learning_rate": 9.640201005025126e-05,
+      "loss": 1.387,
+      "step": 816
+    },
+    {
+      "epoch": 0.06971807699282126,
+      "grad_norm": 0.314453125,
+      "learning_rate": 9.639698492462312e-05,
+      "loss": 1.3563,
+      "step": 817
+    },
+    {
+      "epoch": 0.06980341123638652,
+      "grad_norm": 0.3203125,
+      "learning_rate": 9.639195979899498e-05,
+      "loss": 1.2745,
+      "step": 818
+    },
+    {
+      "epoch": 0.06988874547995179,
+      "grad_norm": 0.3203125,
+      "learning_rate": 9.638693467336683e-05,
+      "loss": 1.4439,
+      "step": 819
+    },
+    {
+      "epoch": 0.06997407972351705,
+      "grad_norm": 0.345703125,
+      "learning_rate": 9.63819095477387e-05,
+      "loss": 1.4016,
+      "step": 820
+    },
+    {
+      "epoch": 0.07005941396708232,
+      "grad_norm": 0.314453125,
+      "learning_rate": 9.637688442211055e-05,
+      "loss": 1.2745,
+      "step": 821
+    },
+    {
+      "epoch": 0.07014474821064758,
+      "grad_norm": 0.294921875,
+      "learning_rate": 9.637185929648242e-05,
+      "loss": 1.2946,
+      "step": 822
+    },
+    {
+      "epoch": 0.07023008245421285,
+      "grad_norm": 0.302734375,
+      "learning_rate": 9.636683417085428e-05,
+      "loss": 1.2733,
+      "step": 823
+    },
+    {
+      "epoch": 0.0703154166977781,
+      "grad_norm": 0.310546875,
+      "learning_rate": 9.636180904522614e-05,
+      "loss": 1.3107,
+      "step": 824
+    },
+    {
+      "epoch": 0.07040075094134338,
+      "grad_norm": 0.3359375,
+      "learning_rate": 9.635678391959799e-05,
+      "loss": 1.3516,
+      "step": 825
+    },
+    {
+      "epoch": 0.07048608518490863,
+      "grad_norm": 0.3359375,
+      "learning_rate": 9.635175879396985e-05,
+      "loss": 1.3862,
+      "step": 826
+    },
+    {
+      "epoch": 0.0705714194284739,
+      "grad_norm": 0.28125,
+      "learning_rate": 9.634673366834171e-05,
+      "loss": 1.3199,
+      "step": 827
+    },
+    {
+      "epoch": 0.07065675367203916,
+      "grad_norm": 0.28515625,
+      "learning_rate": 9.634170854271357e-05,
+      "loss": 1.3849,
+      "step": 828
+    },
+    {
+      "epoch": 0.07074208791560443,
+      "grad_norm": 0.3515625,
+      "learning_rate": 9.633668341708542e-05,
+      "loss": 1.4096,
+      "step": 829
+    },
+    {
+      "epoch": 0.07082742215916969,
+      "grad_norm": 0.333984375,
+      "learning_rate": 9.63316582914573e-05,
+      "loss": 1.3158,
+      "step": 830
+    },
+    {
+      "epoch": 0.07091275640273496,
+      "grad_norm": 0.302734375,
+      "learning_rate": 9.632663316582915e-05,
+      "loss": 1.1482,
+      "step": 831
+    },
+    {
+      "epoch": 0.07099809064630022,
+      "grad_norm": 0.32421875,
+      "learning_rate": 9.632160804020101e-05,
+      "loss": 1.2551,
+      "step": 832
+    },
+    {
+      "epoch": 0.07108342488986549,
+      "grad_norm": 0.296875,
+      "learning_rate": 9.631658291457287e-05,
+      "loss": 1.3848,
+      "step": 833
+    },
+    {
+      "epoch": 0.07116875913343075,
+      "grad_norm": 0.345703125,
+      "learning_rate": 9.631155778894473e-05,
+      "loss": 1.4597,
+      "step": 834
+    },
+    {
+      "epoch": 0.07125409337699602,
+      "grad_norm": 0.302734375,
+      "learning_rate": 9.630653266331658e-05,
+      "loss": 1.2314,
+      "step": 835
+    },
+    {
+      "epoch": 0.07133942762056128,
+      "grad_norm": 0.3125,
+      "learning_rate": 9.630150753768846e-05,
+      "loss": 1.3928,
+      "step": 836
+    },
+    {
+      "epoch": 0.07142476186412655,
+      "grad_norm": 0.318359375,
+      "learning_rate": 9.62964824120603e-05,
+      "loss": 1.4368,
+      "step": 837
+    },
+    {
+      "epoch": 0.07151009610769181,
+      "grad_norm": 0.34375,
+      "learning_rate": 9.629145728643217e-05,
+      "loss": 1.5076,
+      "step": 838
+    },
+    {
+      "epoch": 0.07159543035125708,
+      "grad_norm": 0.337890625,
+      "learning_rate": 9.628643216080403e-05,
+      "loss": 1.3296,
+      "step": 839
+    },
+    {
+      "epoch": 0.07168076459482234,
+      "grad_norm": 0.34765625,
+      "learning_rate": 9.628140703517589e-05,
+      "loss": 1.302,
+      "step": 840
+    },
+    {
+      "epoch": 0.07176609883838761,
+      "grad_norm": 0.314453125,
+      "learning_rate": 9.627638190954774e-05,
+      "loss": 1.5028,
+      "step": 841
+    },
+    {
+      "epoch": 0.07185143308195287,
+      "grad_norm": 0.30859375,
+      "learning_rate": 9.62713567839196e-05,
+      "loss": 1.3537,
+      "step": 842
+    },
+    {
+      "epoch": 0.07193676732551814,
+      "grad_norm": 0.3359375,
+      "learning_rate": 9.626633165829146e-05,
+      "loss": 1.415,
+      "step": 843
+    },
+    {
+      "epoch": 0.07202210156908341,
+      "grad_norm": 0.29296875,
+      "learning_rate": 9.626130653266333e-05,
+      "loss": 1.3223,
+      "step": 844
+    },
+    {
+      "epoch": 0.07210743581264867,
+      "grad_norm": 0.32421875,
+      "learning_rate": 9.625628140703517e-05,
+      "loss": 1.396,
+      "step": 845
+    },
+    {
+      "epoch": 0.07219277005621394,
+      "grad_norm": 0.3203125,
+      "learning_rate": 9.625125628140704e-05,
+      "loss": 1.2808,
+      "step": 846
+    },
+    {
+      "epoch": 0.0722781042997792,
+      "grad_norm": 0.30859375,
+      "learning_rate": 9.62462311557789e-05,
+      "loss": 1.2221,
+      "step": 847
+    },
+    {
+      "epoch": 0.07236343854334447,
+      "grad_norm": 0.287109375,
+      "learning_rate": 9.624120603015076e-05,
+      "loss": 1.4332,
+      "step": 848
+    },
+    {
+      "epoch": 0.07244877278690973,
+      "grad_norm": 0.318359375,
+      "learning_rate": 9.623618090452262e-05,
+      "loss": 1.3363,
+      "step": 849
+    },
+    {
+      "epoch": 0.072534107030475,
+      "grad_norm": 0.30078125,
+      "learning_rate": 9.623115577889447e-05,
+      "loss": 1.3362,
+      "step": 850
+    },
+    {
+      "epoch": 0.07261944127404026,
+      "grad_norm": 0.353515625,
+      "learning_rate": 9.622613065326633e-05,
+      "loss": 1.4657,
+      "step": 851
+    },
+    {
+      "epoch": 0.07270477551760553,
+      "grad_norm": 0.333984375,
+      "learning_rate": 9.62211055276382e-05,
+      "loss": 1.5472,
+      "step": 852
+    },
+    {
+      "epoch": 0.07279010976117078,
+      "grad_norm": 0.31640625,
+      "learning_rate": 9.621608040201006e-05,
+      "loss": 1.3259,
+      "step": 853
+    },
+    {
+      "epoch": 0.07287544400473606,
+      "grad_norm": 0.296875,
+      "learning_rate": 9.62110552763819e-05,
+      "loss": 1.2589,
+      "step": 854
+    },
+    {
+      "epoch": 0.07296077824830131,
+      "grad_norm": 0.296875,
+      "learning_rate": 9.620603015075378e-05,
+      "loss": 1.2032,
+      "step": 855
+    },
+    {
+      "epoch": 0.07304611249186659,
+      "grad_norm": 0.3359375,
+      "learning_rate": 9.620100502512563e-05,
+      "loss": 1.4066,
+      "step": 856
+    },
+    {
+      "epoch": 0.07313144673543184,
+      "grad_norm": 0.30859375,
+      "learning_rate": 9.619597989949749e-05,
+      "loss": 1.3098,
+      "step": 857
+    },
+    {
+      "epoch": 0.07321678097899711,
+      "grad_norm": 0.400390625,
+      "learning_rate": 9.619095477386934e-05,
+      "loss": 1.4806,
+      "step": 858
+    },
+    {
+      "epoch": 0.07330211522256237,
+      "grad_norm": 0.302734375,
+      "learning_rate": 9.618592964824122e-05,
+      "loss": 1.3792,
+      "step": 859
+    },
+    {
+      "epoch": 0.07338744946612764,
+      "grad_norm": 0.32421875,
+      "learning_rate": 9.618090452261306e-05,
+      "loss": 1.4728,
+      "step": 860
+    },
+    {
+      "epoch": 0.0734727837096929,
+      "grad_norm": 0.3046875,
+      "learning_rate": 9.617587939698493e-05,
+      "loss": 1.4013,
+      "step": 861
+    },
+    {
+      "epoch": 0.07355811795325817,
+      "grad_norm": 0.3046875,
+      "learning_rate": 9.617085427135679e-05,
+      "loss": 1.2793,
+      "step": 862
+    },
+    {
+      "epoch": 0.07364345219682343,
+      "grad_norm": 0.337890625,
+      "learning_rate": 9.616582914572865e-05,
+      "loss": 1.3481,
+      "step": 863
+    },
+    {
+      "epoch": 0.0737287864403887,
+      "grad_norm": 0.314453125,
+      "learning_rate": 9.61608040201005e-05,
+      "loss": 1.4183,
+      "step": 864
+    },
+    {
+      "epoch": 0.07381412068395396,
+      "grad_norm": 0.33203125,
+      "learning_rate": 9.615577889447237e-05,
+      "loss": 1.3814,
+      "step": 865
+    },
+    {
+      "epoch": 0.07389945492751923,
+      "grad_norm": 0.32421875,
+      "learning_rate": 9.615075376884422e-05,
+      "loss": 1.354,
+      "step": 866
+    },
+    {
+      "epoch": 0.07398478917108449,
+      "grad_norm": 0.359375,
+      "learning_rate": 9.614572864321608e-05,
+      "loss": 1.3322,
+      "step": 867
+    },
+    {
+      "epoch": 0.07407012341464976,
+      "grad_norm": 0.306640625,
+      "learning_rate": 9.614070351758795e-05,
+      "loss": 1.301,
+      "step": 868
+    },
+    {
+      "epoch": 0.07415545765821502,
+      "grad_norm": 0.29296875,
+      "learning_rate": 9.613567839195981e-05,
+      "loss": 1.4269,
+      "step": 869
+    },
+    {
+      "epoch": 0.07424079190178029,
+      "grad_norm": 0.3125,
+      "learning_rate": 9.613065326633166e-05,
+      "loss": 1.4436,
+      "step": 870
+    },
+    {
+      "epoch": 0.07432612614534555,
+      "grad_norm": 0.32421875,
+      "learning_rate": 9.612562814070353e-05,
+      "loss": 1.376,
+      "step": 871
+    },
+    {
+      "epoch": 0.07441146038891082,
+      "grad_norm": 0.306640625,
+      "learning_rate": 9.612060301507538e-05,
+      "loss": 1.2913,
+      "step": 872
+    },
+    {
+      "epoch": 0.07449679463247608,
+      "grad_norm": 0.294921875,
+      "learning_rate": 9.611557788944724e-05,
+      "loss": 1.1825,
+      "step": 873
+    },
+    {
+      "epoch": 0.07458212887604135,
+      "grad_norm": 0.35546875,
+      "learning_rate": 9.611055276381909e-05,
+      "loss": 1.4007,
+      "step": 874
+    },
+    {
+      "epoch": 0.0746674631196066,
+      "grad_norm": 0.359375,
+      "learning_rate": 9.610552763819097e-05,
+      "loss": 1.3464,
+      "step": 875
+    },
+    {
+      "epoch": 0.07475279736317188,
+      "grad_norm": 0.32421875,
+      "learning_rate": 9.610050251256282e-05,
+      "loss": 1.4501,
+      "step": 876
+    },
+    {
+      "epoch": 0.07483813160673713,
+      "grad_norm": 0.345703125,
+      "learning_rate": 9.609547738693468e-05,
+      "loss": 1.3211,
+      "step": 877
+    },
+    {
+      "epoch": 0.0749234658503024,
+      "grad_norm": 0.2890625,
+      "learning_rate": 9.609045226130654e-05,
+      "loss": 1.2537,
+      "step": 878
+    },
+    {
+      "epoch": 0.07500880009386766,
+      "grad_norm": 0.298828125,
+      "learning_rate": 9.60854271356784e-05,
+      "loss": 1.3352,
+      "step": 879
+    },
+    {
+      "epoch": 0.07509413433743294,
+      "grad_norm": 0.3046875,
+      "learning_rate": 9.608040201005025e-05,
+      "loss": 1.2167,
+      "step": 880
+    },
+    {
+      "epoch": 0.0751794685809982,
+      "grad_norm": 0.30859375,
+      "learning_rate": 9.607537688442211e-05,
+      "loss": 1.2767,
+      "step": 881
+    },
+    {
+      "epoch": 0.07526480282456346,
+      "grad_norm": 0.330078125,
+      "learning_rate": 9.607035175879397e-05,
+      "loss": 1.4463,
+      "step": 882
+    },
+    {
+      "epoch": 0.07535013706812872,
+      "grad_norm": 0.306640625,
+      "learning_rate": 9.606532663316584e-05,
+      "loss": 1.2891,
+      "step": 883
+    },
+    {
+      "epoch": 0.075435471311694,
+      "grad_norm": 0.27734375,
+      "learning_rate": 9.60603015075377e-05,
+      "loss": 1.3248,
+      "step": 884
+    },
+    {
+      "epoch": 0.07552080555525925,
+      "grad_norm": 0.341796875,
+      "learning_rate": 9.605527638190955e-05,
+      "loss": 1.4177,
+      "step": 885
+    },
+    {
+      "epoch": 0.07560613979882452,
+      "grad_norm": 0.3203125,
+      "learning_rate": 9.605025125628141e-05,
+      "loss": 1.2561,
+      "step": 886
+    },
+    {
+      "epoch": 0.07569147404238978,
+      "grad_norm": 0.310546875,
+      "learning_rate": 9.604522613065327e-05,
+      "loss": 1.7487,
+      "step": 887
+    },
+    {
+      "epoch": 0.07577680828595505,
+      "grad_norm": 0.291015625,
+      "learning_rate": 9.604020100502513e-05,
+      "loss": 1.3968,
+      "step": 888
+    },
+    {
+      "epoch": 0.07586214252952031,
+      "grad_norm": 0.5703125,
+      "learning_rate": 9.603517587939698e-05,
+      "loss": 1.5181,
+      "step": 889
+    },
+    {
+      "epoch": 0.07594747677308558,
+      "grad_norm": 0.310546875,
+      "learning_rate": 9.603015075376884e-05,
+      "loss": 1.3616,
+      "step": 890
+    },
+    {
+      "epoch": 0.07603281101665084,
+      "grad_norm": 0.349609375,
+      "learning_rate": 9.60251256281407e-05,
+      "loss": 1.451,
+      "step": 891
+    },
+    {
+      "epoch": 0.07611814526021611,
+      "grad_norm": 0.35546875,
+      "learning_rate": 9.602010050251257e-05,
+      "loss": 1.3771,
+      "step": 892
+    },
+    {
+      "epoch": 0.07620347950378137,
+      "grad_norm": 0.3515625,
+      "learning_rate": 9.601507537688442e-05,
+      "loss": 1.3326,
+      "step": 893
+    },
+    {
+      "epoch": 0.07628881374734664,
+      "grad_norm": 0.3203125,
+      "learning_rate": 9.601005025125629e-05,
+      "loss": 1.359,
+      "step": 894
+    },
+    {
+      "epoch": 0.0763741479909119,
+      "grad_norm": 0.3125,
+      "learning_rate": 9.600502512562814e-05,
+      "loss": 1.4046,
+      "step": 895
+    },
+    {
+      "epoch": 0.07645948223447717,
+      "grad_norm": 0.31640625,
+      "learning_rate": 9.6e-05,
+      "loss": 1.4966,
+      "step": 896
+    },
+    {
+      "epoch": 0.07654481647804243,
+      "grad_norm": 0.3046875,
+      "learning_rate": 9.599497487437186e-05,
+      "loss": 1.4438,
+      "step": 897
+    },
+    {
+      "epoch": 0.0766301507216077,
+      "grad_norm": 0.341796875,
+      "learning_rate": 9.598994974874373e-05,
+      "loss": 1.3693,
+      "step": 898
+    },
+    {
+      "epoch": 0.07671548496517296,
+      "grad_norm": 0.3046875,
+      "learning_rate": 9.598492462311557e-05,
+      "loss": 1.3065,
+      "step": 899
+    },
+    {
+      "epoch": 0.07680081920873823,
+      "grad_norm": 0.322265625,
+      "learning_rate": 9.597989949748745e-05,
+      "loss": 1.3884,
+      "step": 900
+    },
+    {
+      "epoch": 0.07688615345230349,
+      "grad_norm": 0.318359375,
+      "learning_rate": 9.59748743718593e-05,
+      "loss": 1.297,
+      "step": 901
+    },
+    {
+      "epoch": 0.07697148769586876,
+      "grad_norm": 0.279296875,
+      "learning_rate": 9.596984924623116e-05,
+      "loss": 1.3187,
+      "step": 902
+    },
+    {
+      "epoch": 0.07705682193943401,
+      "grad_norm": 0.349609375,
+      "learning_rate": 9.596482412060302e-05,
+      "loss": 1.6206,
+      "step": 903
+    },
+    {
+      "epoch": 0.07714215618299929,
+      "grad_norm": 0.30078125,
+      "learning_rate": 9.595979899497489e-05,
+      "loss": 1.3043,
+      "step": 904
+    },
+    {
+      "epoch": 0.07722749042656454,
+      "grad_norm": 0.359375,
+      "learning_rate": 9.595477386934673e-05,
+      "loss": 1.5438,
+      "step": 905
+    },
+    {
+      "epoch": 0.07731282467012981,
+      "grad_norm": 0.3125,
+      "learning_rate": 9.59497487437186e-05,
+      "loss": 1.2039,
+      "step": 906
+    },
+    {
+      "epoch": 0.07739815891369507,
+      "grad_norm": 0.31640625,
+      "learning_rate": 9.594472361809046e-05,
+      "loss": 1.5157,
+      "step": 907
+    },
+    {
+      "epoch": 0.07748349315726034,
+      "grad_norm": 0.314453125,
+      "learning_rate": 9.593969849246232e-05,
+      "loss": 1.1498,
+      "step": 908
+    },
+    {
+      "epoch": 0.07756882740082562,
+      "grad_norm": 0.34765625,
+      "learning_rate": 9.593467336683417e-05,
+      "loss": 1.5209,
+      "step": 909
+    },
+    {
+      "epoch": 0.07765416164439087,
+      "grad_norm": 0.322265625,
+      "learning_rate": 9.592964824120604e-05,
+      "loss": 1.4398,
+      "step": 910
+    },
+    {
+      "epoch": 0.07773949588795614,
+      "grad_norm": 0.341796875,
+      "learning_rate": 9.592462311557789e-05,
+      "loss": 1.3306,
+      "step": 911
+    },
+    {
+      "epoch": 0.0778248301315214,
+      "grad_norm": 0.353515625,
+      "learning_rate": 9.591959798994975e-05,
+      "loss": 1.3401,
+      "step": 912
+    },
+    {
+      "epoch": 0.07791016437508667,
+      "grad_norm": 0.3125,
+      "learning_rate": 9.591457286432162e-05,
+      "loss": 1.3406,
+      "step": 913
+    },
+    {
+      "epoch": 0.07799549861865193,
+      "grad_norm": 0.33984375,
+      "learning_rate": 9.590954773869348e-05,
+      "loss": 1.3676,
+      "step": 914
+    },
+    {
+      "epoch": 0.0780808328622172,
+      "grad_norm": 0.31640625,
+      "learning_rate": 9.590452261306533e-05,
+      "loss": 1.2864,
+      "step": 915
+    },
+    {
+      "epoch": 0.07816616710578246,
+      "grad_norm": 0.359375,
+      "learning_rate": 9.589949748743719e-05,
+      "loss": 1.4584,
+      "step": 916
+    },
+    {
+      "epoch": 0.07825150134934773,
+      "grad_norm": 0.3203125,
+      "learning_rate": 9.589447236180905e-05,
+      "loss": 1.3652,
+      "step": 917
+    },
+    {
+      "epoch": 0.07833683559291299,
+      "grad_norm": 0.3046875,
+      "learning_rate": 9.588944723618091e-05,
+      "loss": 1.3157,
+      "step": 918
+    },
+    {
+      "epoch": 0.07842216983647826,
+      "grad_norm": 0.296875,
+      "learning_rate": 9.588442211055276e-05,
+      "loss": 1.3047,
+      "step": 919
+    },
+    {
+      "epoch": 0.07850750408004352,
+      "grad_norm": 0.31640625,
+      "learning_rate": 9.587939698492462e-05,
+      "loss": 1.3499,
+      "step": 920
+    },
+    {
+      "epoch": 0.07859283832360879,
+      "grad_norm": 0.314453125,
+      "learning_rate": 9.587437185929649e-05,
+      "loss": 1.3093,
+      "step": 921
+    },
+    {
+      "epoch": 0.07867817256717405,
+      "grad_norm": 0.318359375,
+      "learning_rate": 9.586934673366835e-05,
+      "loss": 1.5205,
+      "step": 922
+    },
+    {
+      "epoch": 0.07876350681073932,
+      "grad_norm": 0.33203125,
+      "learning_rate": 9.586432160804021e-05,
+      "loss": 1.3924,
+      "step": 923
+    },
+    {
+      "epoch": 0.07884884105430458,
+      "grad_norm": 0.341796875,
+      "learning_rate": 9.585929648241206e-05,
+      "loss": 1.1967,
+      "step": 924
+    },
+    {
+      "epoch": 0.07893417529786985,
+      "grad_norm": 0.29296875,
+      "learning_rate": 9.585427135678392e-05,
+      "loss": 1.3746,
+      "step": 925
+    },
+    {
+      "epoch": 0.0790195095414351,
+      "grad_norm": 0.298828125,
+      "learning_rate": 9.584924623115578e-05,
+      "loss": 1.274,
+      "step": 926
+    },
+    {
+      "epoch": 0.07910484378500038,
+      "grad_norm": 0.330078125,
+      "learning_rate": 9.584422110552764e-05,
+      "loss": 1.3245,
+      "step": 927
+    },
+    {
+      "epoch": 0.07919017802856564,
+      "grad_norm": 0.32421875,
+      "learning_rate": 9.583919597989949e-05,
+      "loss": 1.4353,
+      "step": 928
+    },
+    {
+      "epoch": 0.07927551227213091,
+      "grad_norm": 0.287109375,
+      "learning_rate": 9.583417085427137e-05,
+      "loss": 1.3904,
+      "step": 929
+    },
+    {
+      "epoch": 0.07936084651569617,
+      "grad_norm": 0.3359375,
+      "learning_rate": 9.582914572864322e-05,
+      "loss": 1.3832,
+      "step": 930
+    },
+    {
+      "epoch": 0.07944618075926144,
+      "grad_norm": 0.2890625,
+      "learning_rate": 9.582412060301508e-05,
+      "loss": 1.2578,
+      "step": 931
+    },
+    {
+      "epoch": 0.0795315150028267,
+      "grad_norm": 0.341796875,
+      "learning_rate": 9.581909547738694e-05,
+      "loss": 1.248,
+      "step": 932
+    },
+    {
+      "epoch": 0.07961684924639197,
+      "grad_norm": 0.333984375,
+      "learning_rate": 9.58140703517588e-05,
+      "loss": 1.3457,
+      "step": 933
+    },
+    {
+      "epoch": 0.07970218348995722,
+      "grad_norm": 0.302734375,
+      "learning_rate": 9.580904522613065e-05,
+      "loss": 1.3833,
+      "step": 934
+    },
+    {
+      "epoch": 0.0797875177335225,
+      "grad_norm": 0.318359375,
+      "learning_rate": 9.580402010050251e-05,
+      "loss": 1.3946,
+      "step": 935
+    },
+    {
+      "epoch": 0.07987285197708775,
+      "grad_norm": 0.30078125,
+      "learning_rate": 9.579899497487438e-05,
+      "loss": 1.237,
+      "step": 936
+    },
+    {
+      "epoch": 0.07995818622065302,
+      "grad_norm": 0.3359375,
+      "learning_rate": 9.579396984924624e-05,
+      "loss": 1.4489,
+      "step": 937
+    },
+    {
+      "epoch": 0.08004352046421828,
+      "grad_norm": 0.37109375,
+      "learning_rate": 9.578894472361809e-05,
+      "loss": 1.6768,
+      "step": 938
+    },
+    {
+      "epoch": 0.08012885470778355,
+      "grad_norm": 0.298828125,
+      "learning_rate": 9.578391959798996e-05,
+      "loss": 1.3023,
+      "step": 939
+    },
+    {
+      "epoch": 0.08021418895134881,
+      "grad_norm": 0.322265625,
+      "learning_rate": 9.577889447236181e-05,
+      "loss": 1.4101,
+      "step": 940
+    },
+    {
+      "epoch": 0.08029952319491408,
+      "grad_norm": 0.384765625,
+      "learning_rate": 9.577386934673367e-05,
+      "loss": 1.5069,
+      "step": 941
+    },
+    {
+      "epoch": 0.08038485743847934,
+      "grad_norm": 0.306640625,
+      "learning_rate": 9.576884422110553e-05,
+      "loss": 1.414,
+      "step": 942
+    },
+    {
+      "epoch": 0.08047019168204461,
+      "grad_norm": 0.2734375,
+      "learning_rate": 9.57638190954774e-05,
+      "loss": 1.4306,
+      "step": 943
+    },
+    {
+      "epoch": 0.08055552592560987,
+      "grad_norm": 0.318359375,
+      "learning_rate": 9.575879396984924e-05,
+      "loss": 1.2921,
+      "step": 944
+    },
+    {
+      "epoch": 0.08064086016917514,
+      "grad_norm": 0.28515625,
+      "learning_rate": 9.575376884422112e-05,
+      "loss": 1.2511,
+      "step": 945
+    },
+    {
+      "epoch": 0.0807261944127404,
+      "grad_norm": 0.2734375,
+      "learning_rate": 9.574874371859297e-05,
+      "loss": 1.2058,
+      "step": 946
+    },
+    {
+      "epoch": 0.08081152865630567,
+      "grad_norm": 0.30859375,
+      "learning_rate": 9.574371859296483e-05,
+      "loss": 1.2351,
+      "step": 947
+    },
+    {
+      "epoch": 0.08089686289987093,
+      "grad_norm": 0.345703125,
+      "learning_rate": 9.573869346733669e-05,
+      "loss": 1.4025,
+      "step": 948
+    },
+    {
+      "epoch": 0.0809821971434362,
+      "grad_norm": 0.32421875,
+      "learning_rate": 9.573366834170855e-05,
+      "loss": 1.3862,
+      "step": 949
+    },
+    {
+      "epoch": 0.08106753138700146,
+      "grad_norm": 0.318359375,
+      "learning_rate": 9.57286432160804e-05,
+      "loss": 1.5174,
+      "step": 950
+    },
+    {
+      "epoch": 0.08115286563056673,
+      "grad_norm": 0.294921875,
+      "learning_rate": 9.572361809045227e-05,
+      "loss": 1.3707,
+      "step": 951
+    },
+    {
+      "epoch": 0.08123819987413199,
+      "grad_norm": 0.3203125,
+      "learning_rate": 9.571859296482413e-05,
+      "loss": 1.2754,
+      "step": 952
+    },
+    {
+      "epoch": 0.08132353411769726,
+      "grad_norm": 0.37109375,
+      "learning_rate": 9.571356783919599e-05,
+      "loss": 1.2164,
+      "step": 953
+    },
+    {
+      "epoch": 0.08140886836126252,
+      "grad_norm": 0.298828125,
+      "learning_rate": 9.570854271356784e-05,
+      "loss": 1.3342,
+      "step": 954
+    },
+    {
+      "epoch": 0.08149420260482779,
+      "grad_norm": 0.296875,
+      "learning_rate": 9.57035175879397e-05,
+      "loss": 1.2289,
+      "step": 955
+    },
+    {
+      "epoch": 0.08157953684839304,
+      "grad_norm": 0.34375,
+      "learning_rate": 9.569849246231156e-05,
+      "loss": 1.4531,
+      "step": 956
+    },
+    {
+      "epoch": 0.08166487109195832,
+      "grad_norm": 0.306640625,
+      "learning_rate": 9.569346733668342e-05,
+      "loss": 1.2437,
+      "step": 957
+    },
+    {
+      "epoch": 0.08175020533552357,
+      "grad_norm": 0.3125,
+      "learning_rate": 9.568844221105529e-05,
+      "loss": 1.2783,
+      "step": 958
+    },
+    {
+      "epoch": 0.08183553957908885,
+      "grad_norm": 0.32421875,
+      "learning_rate": 9.568341708542713e-05,
+      "loss": 1.3499,
+      "step": 959
+    },
+    {
+      "epoch": 0.0819208738226541,
+      "grad_norm": 0.31640625,
+      "learning_rate": 9.5678391959799e-05,
+      "loss": 1.4048,
+      "step": 960
+    },
+    {
+      "epoch": 0.08200620806621937,
+      "grad_norm": 0.341796875,
+      "learning_rate": 9.567336683417086e-05,
+      "loss": 1.397,
+      "step": 961
+    },
+    {
+      "epoch": 0.08209154230978463,
+      "grad_norm": 0.3359375,
+      "learning_rate": 9.566834170854272e-05,
+      "loss": 1.3072,
+      "step": 962
+    },
+    {
+      "epoch": 0.0821768765533499,
+      "grad_norm": 0.29296875,
+      "learning_rate": 9.566331658291457e-05,
+      "loss": 1.2147,
+      "step": 963
+    },
+    {
+      "epoch": 0.08226221079691516,
+      "grad_norm": 0.296875,
+      "learning_rate": 9.565829145728644e-05,
+      "loss": 1.3292,
+      "step": 964
+    },
+    {
+      "epoch": 0.08234754504048043,
+      "grad_norm": 0.365234375,
+      "learning_rate": 9.565326633165829e-05,
+      "loss": 1.348,
+      "step": 965
+    },
+    {
+      "epoch": 0.08243287928404569,
+      "grad_norm": 0.32421875,
+      "learning_rate": 9.564824120603016e-05,
+      "loss": 1.4118,
+      "step": 966
+    },
+    {
+      "epoch": 0.08251821352761096,
+      "grad_norm": 0.31640625,
+      "learning_rate": 9.5643216080402e-05,
+      "loss": 1.372,
+      "step": 967
+    },
+    {
+      "epoch": 0.08260354777117622,
+      "grad_norm": 0.357421875,
+      "learning_rate": 9.563819095477388e-05,
+      "loss": 1.3445,
+      "step": 968
+    },
+    {
+      "epoch": 0.08268888201474149,
+      "grad_norm": 0.330078125,
+      "learning_rate": 9.563316582914573e-05,
+      "loss": 1.4447,
+      "step": 969
+    },
+    {
+      "epoch": 0.08277421625830675,
+      "grad_norm": 0.322265625,
+      "learning_rate": 9.562814070351759e-05,
+      "loss": 1.3272,
+      "step": 970
+    },
+    {
+      "epoch": 0.08285955050187202,
+      "grad_norm": 0.31640625,
+      "learning_rate": 9.562311557788945e-05,
+      "loss": 1.2451,
+      "step": 971
+    },
+    {
+      "epoch": 0.08294488474543728,
+      "grad_norm": 0.322265625,
+      "learning_rate": 9.561809045226131e-05,
+      "loss": 1.3698,
+      "step": 972
+    },
+    {
+      "epoch": 0.08303021898900255,
+      "grad_norm": 0.34375,
+      "learning_rate": 9.561306532663316e-05,
+      "loss": 1.323,
+      "step": 973
+    },
+    {
+      "epoch": 0.08311555323256782,
+      "grad_norm": 0.28125,
+      "learning_rate": 9.560804020100504e-05,
+      "loss": 1.2683,
+      "step": 974
+    },
+    {
+      "epoch": 0.08320088747613308,
+      "grad_norm": 0.31640625,
+      "learning_rate": 9.560301507537689e-05,
+      "loss": 1.2856,
+      "step": 975
+    },
+    {
+      "epoch": 0.08328622171969835,
+      "grad_norm": 0.306640625,
+      "learning_rate": 9.559798994974875e-05,
+      "loss": 1.3442,
+      "step": 976
+    },
+    {
+      "epoch": 0.08337155596326361,
+      "grad_norm": 0.314453125,
+      "learning_rate": 9.559296482412061e-05,
+      "loss": 1.3305,
+      "step": 977
+    },
+    {
+      "epoch": 0.08345689020682888,
+      "grad_norm": 0.333984375,
+      "learning_rate": 9.558793969849247e-05,
+      "loss": 1.3871,
+      "step": 978
+    },
+    {
+      "epoch": 0.08354222445039414,
+      "grad_norm": 0.3046875,
+      "learning_rate": 9.558291457286432e-05,
+      "loss": 1.1914,
+      "step": 979
+    },
+    {
+      "epoch": 0.08362755869395941,
+      "grad_norm": 0.359375,
+      "learning_rate": 9.55778894472362e-05,
+      "loss": 1.5542,
+      "step": 980
+    },
+    {
+      "epoch": 0.08371289293752467,
+      "grad_norm": 0.35546875,
+      "learning_rate": 9.557286432160804e-05,
+      "loss": 1.3395,
+      "step": 981
+    },
+    {
+      "epoch": 0.08379822718108994,
+      "grad_norm": 0.31640625,
+      "learning_rate": 9.556783919597991e-05,
+      "loss": 1.2492,
+      "step": 982
+    },
+    {
+      "epoch": 0.0838835614246552,
+      "grad_norm": 0.318359375,
+      "learning_rate": 9.556281407035176e-05,
+      "loss": 1.3243,
+      "step": 983
+    },
+    {
+      "epoch": 0.08396889566822047,
+      "grad_norm": 0.330078125,
+      "learning_rate": 9.555778894472363e-05,
+      "loss": 1.4897,
+      "step": 984
+    },
+    {
+      "epoch": 0.08405422991178572,
+      "grad_norm": 0.296875,
+      "learning_rate": 9.555276381909548e-05,
+      "loss": 1.2071,
+      "step": 985
+    },
+    {
+      "epoch": 0.084139564155351,
+      "grad_norm": 0.306640625,
+      "learning_rate": 9.554773869346734e-05,
+      "loss": 1.4489,
+      "step": 986
+    },
+    {
+      "epoch": 0.08422489839891625,
+      "grad_norm": 0.3515625,
+      "learning_rate": 9.55427135678392e-05,
+      "loss": 1.2155,
+      "step": 987
+    },
+    {
+      "epoch": 0.08431023264248153,
+      "grad_norm": 0.294921875,
+      "learning_rate": 9.553768844221107e-05,
+      "loss": 1.3922,
+      "step": 988
+    },
+    {
+      "epoch": 0.08439556688604678,
+      "grad_norm": 0.359375,
+      "learning_rate": 9.553266331658291e-05,
+      "loss": 1.3206,
+      "step": 989
+    },
+    {
+      "epoch": 0.08448090112961205,
+      "grad_norm": 0.357421875,
+      "learning_rate": 9.552763819095478e-05,
+      "loss": 1.7447,
+      "step": 990
+    },
+    {
+      "epoch": 0.08456623537317731,
+      "grad_norm": 0.298828125,
+      "learning_rate": 9.552261306532664e-05,
+      "loss": 1.349,
+      "step": 991
+    },
+    {
+      "epoch": 0.08465156961674258,
+      "grad_norm": 0.314453125,
+      "learning_rate": 9.55175879396985e-05,
+      "loss": 1.4593,
+      "step": 992
+    },
+    {
+      "epoch": 0.08473690386030784,
+      "grad_norm": 0.390625,
+      "learning_rate": 9.551256281407036e-05,
+      "loss": 1.2754,
+      "step": 993
+    },
+    {
+      "epoch": 0.08482223810387311,
+      "grad_norm": 0.34375,
+      "learning_rate": 9.550753768844221e-05,
+      "loss": 1.3154,
+      "step": 994
+    },
+    {
+      "epoch": 0.08490757234743837,
+      "grad_norm": 0.291015625,
+      "learning_rate": 9.550251256281407e-05,
+      "loss": 1.3023,
+      "step": 995
+    },
+    {
+      "epoch": 0.08499290659100364,
+      "grad_norm": 0.3671875,
+      "learning_rate": 9.549748743718593e-05,
+      "loss": 1.3258,
+      "step": 996
+    },
+    {
+      "epoch": 0.0850782408345689,
+      "grad_norm": 0.318359375,
+      "learning_rate": 9.54924623115578e-05,
+      "loss": 1.4054,
+      "step": 997
+    },
+    {
+      "epoch": 0.08516357507813417,
+      "grad_norm": 0.314453125,
+      "learning_rate": 9.548743718592965e-05,
+      "loss": 1.4067,
+      "step": 998
+    },
+    {
+      "epoch": 0.08524890932169943,
+      "grad_norm": 0.302734375,
+      "learning_rate": 9.548241206030151e-05,
+      "loss": 1.2185,
+      "step": 999
+    },
+    {
+      "epoch": 0.0853342435652647,
+      "grad_norm": 0.34765625,
+      "learning_rate": 9.547738693467337e-05,
+      "loss": 1.4602,
+      "step": 1000
+    },
+    {
+      "epoch": 0.08541957780882996,
+      "grad_norm": 0.265625,
+      "learning_rate": 9.547236180904523e-05,
+      "loss": 1.1682,
+      "step": 1001
+    },
+    {
+      "epoch": 0.08550491205239523,
+      "grad_norm": 0.3125,
+      "learning_rate": 9.546733668341708e-05,
+      "loss": 1.5237,
+      "step": 1002
+    },
+    {
+      "epoch": 0.08559024629596049,
+      "grad_norm": 0.330078125,
+      "learning_rate": 9.546231155778896e-05,
+      "loss": 1.4155,
+      "step": 1003
+    },
+    {
+      "epoch": 0.08567558053952576,
+      "grad_norm": 0.32421875,
+      "learning_rate": 9.54572864321608e-05,
+      "loss": 1.3246,
+      "step": 1004
+    },
+    {
+      "epoch": 0.08576091478309102,
+      "grad_norm": 0.29296875,
+      "learning_rate": 9.545226130653267e-05,
+      "loss": 1.4683,
+      "step": 1005
+    },
+    {
+      "epoch": 0.08584624902665629,
+      "grad_norm": 0.326171875,
+      "learning_rate": 9.544723618090453e-05,
+      "loss": 1.3286,
+      "step": 1006
+    },
+    {
+      "epoch": 0.08593158327022155,
+      "grad_norm": 0.306640625,
+      "learning_rate": 9.544221105527639e-05,
+      "loss": 1.3636,
+      "step": 1007
+    },
+    {
+      "epoch": 0.08601691751378682,
+      "grad_norm": 0.33984375,
+      "learning_rate": 9.543718592964824e-05,
+      "loss": 1.4854,
+      "step": 1008
+    },
+    {
+      "epoch": 0.08610225175735207,
+      "grad_norm": 0.328125,
+      "learning_rate": 9.543216080402011e-05,
+      "loss": 1.3744,
+      "step": 1009
+    },
+    {
+      "epoch": 0.08618758600091735,
+      "grad_norm": 0.310546875,
+      "learning_rate": 9.542713567839196e-05,
+      "loss": 1.2739,
+      "step": 1010
+    },
+    {
+      "epoch": 0.0862729202444826,
+      "grad_norm": 0.373046875,
+      "learning_rate": 9.542211055276382e-05,
+      "loss": 1.2112,
+      "step": 1011
+    },
+    {
+      "epoch": 0.08635825448804788,
+      "grad_norm": 0.3203125,
+      "learning_rate": 9.541708542713569e-05,
+      "loss": 1.3066,
+      "step": 1012
+    },
+    {
+      "epoch": 0.08644358873161313,
+      "grad_norm": 0.32421875,
+      "learning_rate": 9.541206030150755e-05,
+      "loss": 1.6257,
+      "step": 1013
+    },
+    {
+      "epoch": 0.0865289229751784,
+      "grad_norm": 0.369140625,
+      "learning_rate": 9.54070351758794e-05,
+      "loss": 1.315,
+      "step": 1014
+    },
+    {
+      "epoch": 0.08661425721874366,
+      "grad_norm": 0.318359375,
+      "learning_rate": 9.540201005025126e-05,
+      "loss": 1.4225,
+      "step": 1015
+    },
+    {
+      "epoch": 0.08669959146230893,
+      "grad_norm": 0.279296875,
+      "learning_rate": 9.539698492462312e-05,
+      "loss": 1.3597,
+      "step": 1016
+    },
+    {
+      "epoch": 0.08678492570587419,
+      "grad_norm": 0.30859375,
+      "learning_rate": 9.539195979899498e-05,
+      "loss": 1.4484,
+      "step": 1017
+    },
+    {
+      "epoch": 0.08687025994943946,
+      "grad_norm": 0.380859375,
+      "learning_rate": 9.538693467336683e-05,
+      "loss": 1.4764,
+      "step": 1018
+    },
+    {
+      "epoch": 0.08695559419300472,
+      "grad_norm": 0.333984375,
+      "learning_rate": 9.538190954773871e-05,
+      "loss": 1.4303,
+      "step": 1019
+    },
+    {
+      "epoch": 0.08704092843656999,
+      "grad_norm": 0.38671875,
+      "learning_rate": 9.537688442211056e-05,
+      "loss": 1.2781,
+      "step": 1020
+    },
+    {
+      "epoch": 0.08712626268013525,
+      "grad_norm": 0.294921875,
+      "learning_rate": 9.537185929648242e-05,
+      "loss": 1.1953,
+      "step": 1021
+    },
+    {
+      "epoch": 0.08721159692370052,
+      "grad_norm": 0.306640625,
+      "learning_rate": 9.536683417085428e-05,
+      "loss": 1.1877,
+      "step": 1022
+    },
+    {
+      "epoch": 0.08729693116726578,
+      "grad_norm": 0.328125,
+      "learning_rate": 9.536180904522614e-05,
+      "loss": 1.2809,
+      "step": 1023
+    },
+    {
+      "epoch": 0.08738226541083105,
+      "grad_norm": 0.3125,
+      "learning_rate": 9.535678391959799e-05,
+      "loss": 1.3766,
+      "step": 1024
+    },
+    {
+      "epoch": 0.08746759965439631,
+      "grad_norm": 0.314453125,
+      "learning_rate": 9.535175879396985e-05,
+      "loss": 1.3612,
+      "step": 1025
+    },
+    {
+      "epoch": 0.08755293389796158,
+      "grad_norm": 0.359375,
+      "learning_rate": 9.534673366834171e-05,
+      "loss": 1.4097,
+      "step": 1026
+    },
+    {
+      "epoch": 0.08763826814152684,
+      "grad_norm": 0.359375,
+      "learning_rate": 9.534170854271358e-05,
+      "loss": 1.4027,
+      "step": 1027
+    },
+    {
+      "epoch": 0.08772360238509211,
+      "grad_norm": 0.283203125,
+      "learning_rate": 9.533668341708543e-05,
+      "loss": 1.2262,
+      "step": 1028
+    },
+    {
+      "epoch": 0.08780893662865737,
+      "grad_norm": 0.31640625,
+      "learning_rate": 9.533165829145729e-05,
+      "loss": 1.4883,
+      "step": 1029
+    },
+    {
+      "epoch": 0.08789427087222264,
+      "grad_norm": 0.28515625,
+      "learning_rate": 9.532663316582915e-05,
+      "loss": 1.2614,
+      "step": 1030
+    },
+    {
+      "epoch": 0.0879796051157879,
+      "grad_norm": 0.359375,
+      "learning_rate": 9.532160804020101e-05,
+      "loss": 1.2919,
+      "step": 1031
+    },
+    {
+      "epoch": 0.08806493935935317,
+      "grad_norm": 0.345703125,
+      "learning_rate": 9.531658291457287e-05,
+      "loss": 1.5493,
+      "step": 1032
+    },
+    {
+      "epoch": 0.08815027360291842,
+      "grad_norm": 0.31640625,
+      "learning_rate": 9.531155778894472e-05,
+      "loss": 1.2417,
+      "step": 1033
+    },
+    {
+      "epoch": 0.0882356078464837,
+      "grad_norm": 0.3203125,
+      "learning_rate": 9.530653266331658e-05,
+      "loss": 1.2649,
+      "step": 1034
+    },
+    {
+      "epoch": 0.08832094209004895,
+      "grad_norm": 0.30859375,
+      "learning_rate": 9.530150753768845e-05,
+      "loss": 1.4258,
+      "step": 1035
+    },
+    {
+      "epoch": 0.08840627633361423,
+      "grad_norm": 0.310546875,
+      "learning_rate": 9.529648241206031e-05,
+      "loss": 1.3556,
+      "step": 1036
+    },
+    {
+      "epoch": 0.08849161057717948,
+      "grad_norm": 0.310546875,
+      "learning_rate": 9.529145728643216e-05,
+      "loss": 1.448,
+      "step": 1037
+    },
+    {
+      "epoch": 0.08857694482074475,
+      "grad_norm": 0.291015625,
+      "learning_rate": 9.528643216080403e-05,
+      "loss": 1.2203,
+      "step": 1038
+    },
+    {
+      "epoch": 0.08866227906431003,
+      "grad_norm": 0.330078125,
+      "learning_rate": 9.528140703517588e-05,
+      "loss": 1.3738,
+      "step": 1039
+    },
+    {
+      "epoch": 0.08874761330787528,
+      "grad_norm": 0.3671875,
+      "learning_rate": 9.527638190954774e-05,
+      "loss": 1.2509,
+      "step": 1040
+    },
+    {
+      "epoch": 0.08883294755144056,
+      "grad_norm": 0.353515625,
+      "learning_rate": 9.52713567839196e-05,
+      "loss": 1.4614,
+      "step": 1041
+    },
+    {
+      "epoch": 0.08891828179500581,
+      "grad_norm": 0.287109375,
+      "learning_rate": 9.526633165829147e-05,
+      "loss": 1.1727,
+      "step": 1042
+    },
+    {
+      "epoch": 0.08900361603857108,
+      "grad_norm": 0.298828125,
+      "learning_rate": 9.526130653266331e-05,
+      "loss": 1.4499,
+      "step": 1043
+    },
+    {
+      "epoch": 0.08908895028213634,
+      "grad_norm": 0.3046875,
+      "learning_rate": 9.525628140703518e-05,
+      "loss": 1.2172,
+      "step": 1044
+    },
+    {
+      "epoch": 0.08917428452570161,
+      "grad_norm": 0.296875,
+      "learning_rate": 9.525125628140704e-05,
+      "loss": 1.2474,
+      "step": 1045
+    },
+    {
+      "epoch": 0.08925961876926687,
+      "grad_norm": 0.27734375,
+      "learning_rate": 9.52462311557789e-05,
+      "loss": 1.3159,
+      "step": 1046
+    },
+    {
+      "epoch": 0.08934495301283214,
+      "grad_norm": 0.3203125,
+      "learning_rate": 9.524120603015075e-05,
+      "loss": 1.3855,
+      "step": 1047
+    },
+    {
+      "epoch": 0.0894302872563974,
+      "grad_norm": 0.33203125,
+      "learning_rate": 9.523618090452263e-05,
+      "loss": 1.1638,
+      "step": 1048
+    },
+    {
+      "epoch": 0.08951562149996267,
+      "grad_norm": 0.337890625,
+      "learning_rate": 9.523115577889447e-05,
+      "loss": 1.442,
+      "step": 1049
+    },
+    {
+      "epoch": 0.08960095574352793,
+      "grad_norm": 0.306640625,
+      "learning_rate": 9.522613065326634e-05,
+      "loss": 1.1956,
+      "step": 1050
+    },
+    {
+      "epoch": 0.0896862899870932,
+      "grad_norm": 0.310546875,
+      "learning_rate": 9.52211055276382e-05,
+      "loss": 1.2994,
+      "step": 1051
+    },
+    {
+      "epoch": 0.08977162423065846,
+      "grad_norm": 0.365234375,
+      "learning_rate": 9.521608040201006e-05,
+      "loss": 1.3061,
+      "step": 1052
+    },
+    {
+      "epoch": 0.08985695847422373,
+      "grad_norm": 0.3203125,
+      "learning_rate": 9.521105527638191e-05,
+      "loss": 1.435,
+      "step": 1053
+    },
+    {
+      "epoch": 0.08994229271778899,
+      "grad_norm": 0.3203125,
+      "learning_rate": 9.520603015075378e-05,
+      "loss": 1.4175,
+      "step": 1054
+    },
+    {
+      "epoch": 0.09002762696135426,
+      "grad_norm": 0.326171875,
+      "learning_rate": 9.520100502512563e-05,
+      "loss": 1.2454,
+      "step": 1055
+    },
+    {
+      "epoch": 0.09011296120491952,
+      "grad_norm": 0.306640625,
+      "learning_rate": 9.51959798994975e-05,
+      "loss": 1.2274,
+      "step": 1056
+    },
+    {
+      "epoch": 0.09019829544848479,
+      "grad_norm": 0.3046875,
+      "learning_rate": 9.519095477386936e-05,
+      "loss": 1.3122,
+      "step": 1057
+    },
+    {
+      "epoch": 0.09028362969205005,
+      "grad_norm": 0.275390625,
+      "learning_rate": 9.518592964824122e-05,
+      "loss": 1.2216,
+      "step": 1058
+    },
+    {
+      "epoch": 0.09036896393561532,
+      "grad_norm": 0.296875,
+      "learning_rate": 9.518090452261307e-05,
+      "loss": 1.3435,
+      "step": 1059
+    },
+    {
+      "epoch": 0.09045429817918058,
+      "grad_norm": 0.30859375,
+      "learning_rate": 9.517587939698493e-05,
+      "loss": 1.1502,
+      "step": 1060
+    },
+    {
+      "epoch": 0.09053963242274585,
+      "grad_norm": 0.31640625,
+      "learning_rate": 9.517085427135679e-05,
+      "loss": 1.1575,
+      "step": 1061
+    },
+    {
+      "epoch": 0.0906249666663111,
+      "grad_norm": 0.298828125,
+      "learning_rate": 9.516582914572865e-05,
+      "loss": 1.209,
+      "step": 1062
+    },
+    {
+      "epoch": 0.09071030090987638,
+      "grad_norm": 0.28515625,
+      "learning_rate": 9.51608040201005e-05,
+      "loss": 1.2504,
+      "step": 1063
+    },
+    {
+      "epoch": 0.09079563515344163,
+      "grad_norm": 0.30859375,
+      "learning_rate": 9.515577889447236e-05,
+      "loss": 1.3055,
+      "step": 1064
+    },
+    {
+      "epoch": 0.0908809693970069,
+      "grad_norm": 0.306640625,
+      "learning_rate": 9.515075376884423e-05,
+      "loss": 1.2386,
+      "step": 1065
+    },
+    {
+      "epoch": 0.09096630364057216,
+      "grad_norm": 0.3359375,
+      "learning_rate": 9.514572864321609e-05,
+      "loss": 1.2577,
+      "step": 1066
+    },
+    {
+      "epoch": 0.09105163788413743,
+      "grad_norm": 0.349609375,
+      "learning_rate": 9.514070351758795e-05,
+      "loss": 1.3365,
+      "step": 1067
+    },
+    {
+      "epoch": 0.09113697212770269,
+      "grad_norm": 0.306640625,
+      "learning_rate": 9.51356783919598e-05,
+      "loss": 1.2264,
+      "step": 1068
+    },
+    {
+      "epoch": 0.09122230637126796,
+      "grad_norm": 0.30078125,
+      "learning_rate": 9.513065326633166e-05,
+      "loss": 1.4203,
+      "step": 1069
+    },
+    {
+      "epoch": 0.09130764061483322,
+      "grad_norm": 0.306640625,
+      "learning_rate": 9.512562814070352e-05,
+      "loss": 1.4671,
+      "step": 1070
+    },
+    {
+      "epoch": 0.09139297485839849,
+      "grad_norm": 0.34765625,
+      "learning_rate": 9.512060301507538e-05,
+      "loss": 1.2146,
+      "step": 1071
+    },
+    {
+      "epoch": 0.09147830910196375,
+      "grad_norm": 0.314453125,
+      "learning_rate": 9.511557788944723e-05,
+      "loss": 1.4928,
+      "step": 1072
+    },
+    {
+      "epoch": 0.09156364334552902,
+      "grad_norm": 0.33203125,
+      "learning_rate": 9.511055276381911e-05,
+      "loss": 1.2909,
+      "step": 1073
+    },
+    {
+      "epoch": 0.09164897758909428,
+      "grad_norm": 0.298828125,
+      "learning_rate": 9.510552763819096e-05,
+      "loss": 1.4021,
+      "step": 1074
+    },
+    {
+      "epoch": 0.09173431183265955,
+      "grad_norm": 0.283203125,
+      "learning_rate": 9.510050251256282e-05,
+      "loss": 1.2152,
+      "step": 1075
+    },
+    {
+      "epoch": 0.09181964607622481,
+      "grad_norm": 0.46875,
+      "learning_rate": 9.509547738693467e-05,
+      "loss": 1.499,
+      "step": 1076
+    },
+    {
+      "epoch": 0.09190498031979008,
+      "grad_norm": 0.298828125,
+      "learning_rate": 9.509045226130654e-05,
+      "loss": 1.3289,
+      "step": 1077
+    },
+    {
+      "epoch": 0.09199031456335534,
+      "grad_norm": 0.34375,
+      "learning_rate": 9.508542713567839e-05,
+      "loss": 1.3843,
+      "step": 1078
+    },
+    {
+      "epoch": 0.09207564880692061,
+      "grad_norm": 0.2890625,
+      "learning_rate": 9.508040201005025e-05,
+      "loss": 1.3072,
+      "step": 1079
+    },
+    {
+      "epoch": 0.09216098305048587,
+      "grad_norm": 0.30859375,
+      "learning_rate": 9.507537688442212e-05,
+      "loss": 1.3309,
+      "step": 1080
+    },
+    {
+      "epoch": 0.09224631729405114,
+      "grad_norm": 0.3203125,
+      "learning_rate": 9.507035175879398e-05,
+      "loss": 1.3491,
+      "step": 1081
+    },
+    {
+      "epoch": 0.0923316515376164,
+      "grad_norm": 0.310546875,
+      "learning_rate": 9.506532663316583e-05,
+      "loss": 1.4421,
+      "step": 1082
+    },
+    {
+      "epoch": 0.09241698578118167,
+      "grad_norm": 0.34375,
+      "learning_rate": 9.50603015075377e-05,
+      "loss": 1.279,
+      "step": 1083
+    },
+    {
+      "epoch": 0.09250232002474693,
+      "grad_norm": 0.27734375,
+      "learning_rate": 9.505527638190955e-05,
+      "loss": 1.1988,
+      "step": 1084
+    },
+    {
+      "epoch": 0.0925876542683122,
+      "grad_norm": 0.265625,
+      "learning_rate": 9.505025125628141e-05,
+      "loss": 1.1984,
+      "step": 1085
+    },
+    {
+      "epoch": 0.09267298851187745,
+      "grad_norm": 0.341796875,
+      "learning_rate": 9.504522613065327e-05,
+      "loss": 1.2976,
+      "step": 1086
+    },
+    {
+      "epoch": 0.09275832275544273,
+      "grad_norm": 0.306640625,
+      "learning_rate": 9.504020100502514e-05,
+      "loss": 1.3143,
+      "step": 1087
+    },
+    {
+      "epoch": 0.09284365699900798,
+      "grad_norm": 0.291015625,
+      "learning_rate": 9.503517587939698e-05,
+      "loss": 1.222,
+      "step": 1088
+    },
+    {
+      "epoch": 0.09292899124257326,
+      "grad_norm": 0.384765625,
+      "learning_rate": 9.503015075376886e-05,
+      "loss": 1.4583,
+      "step": 1089
+    },
+    {
+      "epoch": 0.09301432548613851,
+      "grad_norm": 0.3046875,
+      "learning_rate": 9.502512562814071e-05,
+      "loss": 1.3027,
+      "step": 1090
+    },
+    {
+      "epoch": 0.09309965972970378,
+      "grad_norm": 0.326171875,
+      "learning_rate": 9.502010050251257e-05,
+      "loss": 1.2766,
+      "step": 1091
+    },
+    {
+      "epoch": 0.09318499397326904,
+      "grad_norm": 0.333984375,
+      "learning_rate": 9.501507537688442e-05,
+      "loss": 1.367,
+      "step": 1092
+    },
+    {
+      "epoch": 0.09327032821683431,
+      "grad_norm": 0.330078125,
+      "learning_rate": 9.50100502512563e-05,
+      "loss": 1.318,
+      "step": 1093
+    },
+    {
+      "epoch": 0.09335566246039957,
+      "grad_norm": 0.322265625,
+      "learning_rate": 9.500502512562814e-05,
+      "loss": 1.3332,
+      "step": 1094
+    },
+    {
+      "epoch": 0.09344099670396484,
+      "grad_norm": 0.287109375,
+      "learning_rate": 9.5e-05,
+      "loss": 1.156,
+      "step": 1095
+    },
+    {
+      "epoch": 0.0935263309475301,
+      "grad_norm": 0.291015625,
+      "learning_rate": 9.499497487437187e-05,
+      "loss": 1.3936,
+      "step": 1096
+    },
+    {
+      "epoch": 0.09361166519109537,
+      "grad_norm": 0.28125,
+      "learning_rate": 9.498994974874373e-05,
+      "loss": 1.4093,
+      "step": 1097
+    },
+    {
+      "epoch": 0.09369699943466063,
+      "grad_norm": 0.298828125,
+      "learning_rate": 9.498492462311558e-05,
+      "loss": 1.1657,
+      "step": 1098
+    },
+    {
+      "epoch": 0.0937823336782259,
+      "grad_norm": 0.310546875,
+      "learning_rate": 9.497989949748744e-05,
+      "loss": 1.1372,
+      "step": 1099
+    },
+    {
+      "epoch": 0.09386766792179116,
+      "grad_norm": 0.353515625,
+      "learning_rate": 9.49748743718593e-05,
+      "loss": 1.4971,
+      "step": 1100
+    },
+    {
+      "epoch": 0.09395300216535643,
+      "grad_norm": 0.306640625,
+      "learning_rate": 9.496984924623116e-05,
+      "loss": 1.4222,
+      "step": 1101
+    },
+    {
+      "epoch": 0.09403833640892169,
+      "grad_norm": 0.408203125,
+      "learning_rate": 9.496482412060303e-05,
+      "loss": 1.3926,
+      "step": 1102
+    },
+    {
+      "epoch": 0.09412367065248696,
+      "grad_norm": 0.32421875,
+      "learning_rate": 9.495979899497487e-05,
+      "loss": 1.3004,
+      "step": 1103
+    },
+    {
+      "epoch": 0.09420900489605223,
+      "grad_norm": 0.298828125,
+      "learning_rate": 9.495477386934674e-05,
+      "loss": 1.2803,
+      "step": 1104
+    },
+    {
+      "epoch": 0.09429433913961749,
+      "grad_norm": 0.32421875,
+      "learning_rate": 9.49497487437186e-05,
+      "loss": 1.3128,
+      "step": 1105
+    },
+    {
+      "epoch": 0.09437967338318276,
+      "grad_norm": 0.357421875,
+      "learning_rate": 9.494472361809046e-05,
+      "loss": 1.3343,
+      "step": 1106
+    },
+    {
+      "epoch": 0.09446500762674802,
+      "grad_norm": 0.306640625,
+      "learning_rate": 9.493969849246231e-05,
+      "loss": 1.2732,
+      "step": 1107
+    },
+    {
+      "epoch": 0.09455034187031329,
+      "grad_norm": 0.287109375,
+      "learning_rate": 9.493467336683417e-05,
+      "loss": 1.3553,
+      "step": 1108
+    },
+    {
+      "epoch": 0.09463567611387855,
+      "grad_norm": 0.328125,
+      "learning_rate": 9.492964824120603e-05,
+      "loss": 1.2705,
+      "step": 1109
+    },
+    {
+      "epoch": 0.09472101035744382,
+      "grad_norm": 0.34375,
+      "learning_rate": 9.49246231155779e-05,
+      "loss": 1.3513,
+      "step": 1110
+    },
+    {
+      "epoch": 0.09480634460100908,
+      "grad_norm": 0.310546875,
+      "learning_rate": 9.491959798994974e-05,
+      "loss": 1.3373,
+      "step": 1111
+    },
+    {
+      "epoch": 0.09489167884457435,
+      "grad_norm": 0.29296875,
+      "learning_rate": 9.491457286432162e-05,
+      "loss": 1.2673,
+      "step": 1112
+    },
+    {
+      "epoch": 0.0949770130881396,
+      "grad_norm": 0.306640625,
+      "learning_rate": 9.490954773869347e-05,
+      "loss": 1.3147,
+      "step": 1113
+    },
+    {
+      "epoch": 0.09506234733170488,
+      "grad_norm": 0.3203125,
+      "learning_rate": 9.490452261306533e-05,
+      "loss": 1.2911,
+      "step": 1114
+    },
+    {
+      "epoch": 0.09514768157527013,
+      "grad_norm": 0.302734375,
+      "learning_rate": 9.489949748743719e-05,
+      "loss": 1.2077,
+      "step": 1115
+    },
+    {
+      "epoch": 0.0952330158188354,
+      "grad_norm": 0.326171875,
+      "learning_rate": 9.489447236180905e-05,
+      "loss": 1.3118,
+      "step": 1116
+    },
+    {
+      "epoch": 0.09531835006240066,
+      "grad_norm": 0.326171875,
+      "learning_rate": 9.48894472361809e-05,
+      "loss": 1.2258,
+      "step": 1117
+    },
+    {
+      "epoch": 0.09540368430596594,
+      "grad_norm": 0.412109375,
+      "learning_rate": 9.488442211055278e-05,
+      "loss": 1.5091,
+      "step": 1118
+    },
+    {
+      "epoch": 0.0954890185495312,
+      "grad_norm": 0.337890625,
+      "learning_rate": 9.487939698492463e-05,
+      "loss": 1.3728,
+      "step": 1119
+    },
+    {
+      "epoch": 0.09557435279309646,
+      "grad_norm": 0.30859375,
+      "learning_rate": 9.487437185929649e-05,
+      "loss": 1.2788,
+      "step": 1120
+    },
+    {
+      "epoch": 0.09565968703666172,
+      "grad_norm": 0.287109375,
+      "learning_rate": 9.486934673366835e-05,
+      "loss": 1.1832,
+      "step": 1121
+    },
+    {
+      "epoch": 0.095745021280227,
+      "grad_norm": 0.310546875,
+      "learning_rate": 9.486432160804021e-05,
+      "loss": 1.3591,
+      "step": 1122
+    },
+    {
+      "epoch": 0.09583035552379225,
+      "grad_norm": 0.283203125,
+      "learning_rate": 9.485929648241206e-05,
+      "loss": 1.2637,
+      "step": 1123
+    },
+    {
+      "epoch": 0.09591568976735752,
+      "grad_norm": 0.314453125,
+      "learning_rate": 9.485427135678392e-05,
+      "loss": 1.2691,
+      "step": 1124
+    },
+    {
+      "epoch": 0.09600102401092278,
+      "grad_norm": 0.294921875,
+      "learning_rate": 9.484924623115578e-05,
+      "loss": 1.3089,
+      "step": 1125
+    },
+    {
+      "epoch": 0.09608635825448805,
+      "grad_norm": 0.33984375,
+      "learning_rate": 9.484422110552765e-05,
+      "loss": 1.2502,
+      "step": 1126
+    },
+    {
+      "epoch": 0.09617169249805331,
+      "grad_norm": 0.42578125,
+      "learning_rate": 9.48391959798995e-05,
+      "loss": 1.2362,
+      "step": 1127
+    },
+    {
+      "epoch": 0.09625702674161858,
+      "grad_norm": 0.326171875,
+      "learning_rate": 9.483417085427137e-05,
+      "loss": 1.4225,
+      "step": 1128
+    },
+    {
+      "epoch": 0.09634236098518384,
+      "grad_norm": 0.30078125,
+      "learning_rate": 9.482914572864322e-05,
+      "loss": 1.2085,
+      "step": 1129
+    },
+    {
+      "epoch": 0.09642769522874911,
+      "grad_norm": 0.318359375,
+      "learning_rate": 9.482412060301508e-05,
+      "loss": 1.2116,
+      "step": 1130
+    },
+    {
+      "epoch": 0.09651302947231437,
+      "grad_norm": 0.302734375,
+      "learning_rate": 9.481909547738694e-05,
+      "loss": 1.3269,
+      "step": 1131
+    },
+    {
+      "epoch": 0.09659836371587964,
+      "grad_norm": 0.296875,
+      "learning_rate": 9.48140703517588e-05,
+      "loss": 1.3203,
+      "step": 1132
+    },
+    {
+      "epoch": 0.0966836979594449,
+      "grad_norm": 0.302734375,
+      "learning_rate": 9.480904522613065e-05,
+      "loss": 1.513,
+      "step": 1133
+    },
+    {
+      "epoch": 0.09676903220301017,
+      "grad_norm": 0.3046875,
+      "learning_rate": 9.480402010050252e-05,
+      "loss": 1.3283,
+      "step": 1134
+    },
+    {
+      "epoch": 0.09685436644657543,
+      "grad_norm": 0.322265625,
+      "learning_rate": 9.479899497487438e-05,
+      "loss": 1.4475,
+      "step": 1135
+    },
+    {
+      "epoch": 0.0969397006901407,
+      "grad_norm": 0.314453125,
+      "learning_rate": 9.479396984924624e-05,
+      "loss": 1.3202,
+      "step": 1136
+    },
+    {
+      "epoch": 0.09702503493370596,
+      "grad_norm": 0.302734375,
+      "learning_rate": 9.47889447236181e-05,
+      "loss": 1.2201,
+      "step": 1137
+    },
+    {
+      "epoch": 0.09711036917727123,
+      "grad_norm": 0.314453125,
+      "learning_rate": 9.478391959798995e-05,
+      "loss": 1.2863,
+      "step": 1138
+    },
+    {
+      "epoch": 0.09719570342083649,
+      "grad_norm": 0.314453125,
+      "learning_rate": 9.477889447236181e-05,
+      "loss": 1.3521,
+      "step": 1139
+    },
+    {
+      "epoch": 0.09728103766440176,
+      "grad_norm": 0.283203125,
+      "learning_rate": 9.477386934673366e-05,
+      "loss": 1.2675,
+      "step": 1140
+    },
+    {
+      "epoch": 0.09736637190796701,
+      "grad_norm": 0.33984375,
+      "learning_rate": 9.476884422110554e-05,
+      "loss": 1.3831,
+      "step": 1141
+    },
+    {
+      "epoch": 0.09745170615153229,
+      "grad_norm": 0.31640625,
+      "learning_rate": 9.476381909547739e-05,
+      "loss": 1.2899,
+      "step": 1142
+    },
+    {
+      "epoch": 0.09753704039509754,
+      "grad_norm": 0.28515625,
+      "learning_rate": 9.475879396984925e-05,
+      "loss": 1.3978,
+      "step": 1143
+    },
+    {
+      "epoch": 0.09762237463866281,
+      "grad_norm": 0.294921875,
+      "learning_rate": 9.475376884422111e-05,
+      "loss": 1.2152,
+      "step": 1144
+    },
+    {
+      "epoch": 0.09770770888222807,
+      "grad_norm": 0.3125,
+      "learning_rate": 9.474874371859297e-05,
+      "loss": 1.41,
+      "step": 1145
+    },
+    {
+      "epoch": 0.09779304312579334,
+      "grad_norm": 0.3671875,
+      "learning_rate": 9.474371859296482e-05,
+      "loss": 1.1995,
+      "step": 1146
+    },
+    {
+      "epoch": 0.0978783773693586,
+      "grad_norm": 0.333984375,
+      "learning_rate": 9.47386934673367e-05,
+      "loss": 1.3036,
+      "step": 1147
+    },
+    {
+      "epoch": 0.09796371161292387,
+      "grad_norm": 0.296875,
+      "learning_rate": 9.473366834170854e-05,
+      "loss": 1.3866,
+      "step": 1148
+    },
+    {
+      "epoch": 0.09804904585648913,
+      "grad_norm": 0.306640625,
+      "learning_rate": 9.47286432160804e-05,
+      "loss": 1.4042,
+      "step": 1149
+    },
+    {
+      "epoch": 0.0981343801000544,
+      "grad_norm": 0.3125,
+      "learning_rate": 9.472361809045227e-05,
+      "loss": 1.3863,
+      "step": 1150
+    },
+    {
+      "epoch": 0.09821971434361966,
+      "grad_norm": 0.31640625,
+      "learning_rate": 9.471859296482413e-05,
+      "loss": 1.5364,
+      "step": 1151
+    },
+    {
+      "epoch": 0.09830504858718493,
+      "grad_norm": 0.296875,
+      "learning_rate": 9.471356783919598e-05,
+      "loss": 1.1391,
+      "step": 1152
+    },
+    {
+      "epoch": 0.09839038283075019,
+      "grad_norm": 0.330078125,
+      "learning_rate": 9.470854271356784e-05,
+      "loss": 1.3903,
+      "step": 1153
+    },
+    {
+      "epoch": 0.09847571707431546,
+      "grad_norm": 0.306640625,
+      "learning_rate": 9.47035175879397e-05,
+      "loss": 1.3542,
+      "step": 1154
+    },
+    {
+      "epoch": 0.09856105131788072,
+      "grad_norm": 0.318359375,
+      "learning_rate": 9.469849246231156e-05,
+      "loss": 1.2436,
+      "step": 1155
+    },
+    {
+      "epoch": 0.09864638556144599,
+      "grad_norm": 0.302734375,
+      "learning_rate": 9.469346733668341e-05,
+      "loss": 1.313,
+      "step": 1156
+    },
+    {
+      "epoch": 0.09873171980501125,
+      "grad_norm": 0.291015625,
+      "learning_rate": 9.468844221105529e-05,
+      "loss": 1.3168,
+      "step": 1157
+    },
+    {
+      "epoch": 0.09881705404857652,
+      "grad_norm": 0.322265625,
+      "learning_rate": 9.468341708542714e-05,
+      "loss": 1.1924,
+      "step": 1158
+    },
+    {
+      "epoch": 0.09890238829214178,
+      "grad_norm": 0.326171875,
+      "learning_rate": 9.4678391959799e-05,
+      "loss": 1.4191,
+      "step": 1159
+    },
+    {
+      "epoch": 0.09898772253570705,
+      "grad_norm": 0.29296875,
+      "learning_rate": 9.467336683417086e-05,
+      "loss": 1.3188,
+      "step": 1160
+    },
+    {
+      "epoch": 0.0990730567792723,
+      "grad_norm": 0.28515625,
+      "learning_rate": 9.466834170854272e-05,
+      "loss": 1.2368,
+      "step": 1161
+    },
+    {
+      "epoch": 0.09915839102283758,
+      "grad_norm": 0.3359375,
+      "learning_rate": 9.466331658291457e-05,
+      "loss": 1.606,
+      "step": 1162
+    },
+    {
+      "epoch": 0.09924372526640284,
+      "grad_norm": 0.330078125,
+      "learning_rate": 9.465829145728645e-05,
+      "loss": 1.4095,
+      "step": 1163
+    },
+    {
+      "epoch": 0.0993290595099681,
+      "grad_norm": 0.302734375,
+      "learning_rate": 9.46532663316583e-05,
+      "loss": 1.2883,
+      "step": 1164
+    },
+    {
+      "epoch": 0.09941439375353336,
+      "grad_norm": 0.3125,
+      "learning_rate": 9.464824120603016e-05,
+      "loss": 1.3407,
+      "step": 1165
+    },
+    {
+      "epoch": 0.09949972799709864,
+      "grad_norm": 0.298828125,
+      "learning_rate": 9.464321608040202e-05,
+      "loss": 1.4751,
+      "step": 1166
+    },
+    {
+      "epoch": 0.0995850622406639,
+      "grad_norm": 0.3515625,
+      "learning_rate": 9.463819095477388e-05,
+      "loss": 1.2236,
+      "step": 1167
+    },
+    {
+      "epoch": 0.09967039648422917,
+      "grad_norm": 0.404296875,
+      "learning_rate": 9.463316582914573e-05,
+      "loss": 1.3881,
+      "step": 1168
+    },
+    {
+      "epoch": 0.09975573072779444,
+      "grad_norm": 0.31640625,
+      "learning_rate": 9.462814070351759e-05,
+      "loss": 1.3522,
+      "step": 1169
+    },
+    {
+      "epoch": 0.0998410649713597,
+      "grad_norm": 0.337890625,
+      "learning_rate": 9.462311557788945e-05,
+      "loss": 1.0582,
+      "step": 1170
+    },
+    {
+      "epoch": 0.09992639921492497,
+      "grad_norm": 0.328125,
+      "learning_rate": 9.461809045226132e-05,
+      "loss": 1.3409,
+      "step": 1171
+    },
+    {
+      "epoch": 0.10001173345849022,
+      "grad_norm": 0.357421875,
+      "learning_rate": 9.461306532663317e-05,
+      "loss": 1.2444,
+      "step": 1172
+    },
+    {
+      "epoch": 0.1000970677020555,
+      "grad_norm": 0.322265625,
+      "learning_rate": 9.460804020100503e-05,
+      "loss": 1.3634,
+      "step": 1173
+    },
+    {
+      "epoch": 0.10018240194562075,
+      "grad_norm": 0.34375,
+      "learning_rate": 9.460301507537689e-05,
+      "loss": 1.3582,
+      "step": 1174
+    },
+    {
+      "epoch": 0.10026773618918602,
+      "grad_norm": 0.275390625,
+      "learning_rate": 9.459798994974874e-05,
+      "loss": 1.303,
+      "step": 1175
+    },
+    {
+      "epoch": 0.10035307043275128,
+      "grad_norm": 0.2890625,
+      "learning_rate": 9.459296482412061e-05,
+      "loss": 1.3325,
+      "step": 1176
+    },
+    {
+      "epoch": 0.10043840467631655,
+      "grad_norm": 0.328125,
+      "learning_rate": 9.458793969849246e-05,
+      "loss": 1.3358,
+      "step": 1177
+    },
+    {
+      "epoch": 0.10052373891988181,
+      "grad_norm": 0.296875,
+      "learning_rate": 9.458291457286432e-05,
+      "loss": 1.2335,
+      "step": 1178
+    },
+    {
+      "epoch": 0.10060907316344708,
+      "grad_norm": 0.32421875,
+      "learning_rate": 9.457788944723619e-05,
+      "loss": 1.1906,
+      "step": 1179
+    },
+    {
+      "epoch": 0.10069440740701234,
+      "grad_norm": 0.2890625,
+      "learning_rate": 9.457286432160805e-05,
+      "loss": 1.4378,
+      "step": 1180
+    },
+    {
+      "epoch": 0.10077974165057761,
+      "grad_norm": 0.32421875,
+      "learning_rate": 9.45678391959799e-05,
+      "loss": 1.3206,
+      "step": 1181
+    },
+    {
+      "epoch": 0.10086507589414287,
+      "grad_norm": 0.349609375,
+      "learning_rate": 9.456281407035177e-05,
+      "loss": 1.3138,
+      "step": 1182
+    },
+    {
+      "epoch": 0.10095041013770814,
+      "grad_norm": 0.330078125,
+      "learning_rate": 9.455778894472362e-05,
+      "loss": 1.2635,
+      "step": 1183
+    },
+    {
+      "epoch": 0.1010357443812734,
+      "grad_norm": 0.30859375,
+      "learning_rate": 9.455276381909548e-05,
+      "loss": 1.3609,
+      "step": 1184
+    },
+    {
+      "epoch": 0.10112107862483867,
+      "grad_norm": 0.328125,
+      "learning_rate": 9.454773869346733e-05,
+      "loss": 1.494,
+      "step": 1185
+    },
+    {
+      "epoch": 0.10120641286840393,
+      "grad_norm": 0.3125,
+      "learning_rate": 9.45427135678392e-05,
+      "loss": 1.5018,
+      "step": 1186
+    },
+    {
+      "epoch": 0.1012917471119692,
+      "grad_norm": 0.310546875,
+      "learning_rate": 9.453768844221105e-05,
+      "loss": 1.2777,
+      "step": 1187
+    },
+    {
+      "epoch": 0.10137708135553446,
+      "grad_norm": 0.294921875,
+      "learning_rate": 9.453266331658292e-05,
+      "loss": 1.3416,
+      "step": 1188
+    },
+    {
+      "epoch": 0.10146241559909973,
+      "grad_norm": 0.333984375,
+      "learning_rate": 9.452763819095478e-05,
+      "loss": 1.5682,
+      "step": 1189
+    },
+    {
+      "epoch": 0.10154774984266499,
+      "grad_norm": 0.34765625,
+      "learning_rate": 9.452261306532664e-05,
+      "loss": 1.4004,
+      "step": 1190
+    },
+    {
+      "epoch": 0.10163308408623026,
+      "grad_norm": 0.361328125,
+      "learning_rate": 9.451758793969849e-05,
+      "loss": 1.3492,
+      "step": 1191
+    },
+    {
+      "epoch": 0.10171841832979552,
+      "grad_norm": 0.298828125,
+      "learning_rate": 9.451256281407037e-05,
+      "loss": 1.3,
+      "step": 1192
+    },
+    {
+      "epoch": 0.10180375257336079,
+      "grad_norm": 0.3203125,
+      "learning_rate": 9.450753768844221e-05,
+      "loss": 1.1861,
+      "step": 1193
+    },
+    {
+      "epoch": 0.10188908681692604,
+      "grad_norm": 0.3359375,
+      "learning_rate": 9.450251256281408e-05,
+      "loss": 1.342,
+      "step": 1194
+    },
+    {
+      "epoch": 0.10197442106049132,
+      "grad_norm": 0.318359375,
+      "learning_rate": 9.449748743718594e-05,
+      "loss": 1.3087,
+      "step": 1195
+    },
+    {
+      "epoch": 0.10205975530405657,
+      "grad_norm": 0.318359375,
+      "learning_rate": 9.44924623115578e-05,
+      "loss": 1.1589,
+      "step": 1196
+    },
+    {
+      "epoch": 0.10214508954762185,
+      "grad_norm": 0.306640625,
+      "learning_rate": 9.448743718592965e-05,
+      "loss": 1.2609,
+      "step": 1197
+    },
+    {
+      "epoch": 0.1022304237911871,
+      "grad_norm": 0.291015625,
+      "learning_rate": 9.448241206030152e-05,
+      "loss": 1.334,
+      "step": 1198
+    },
+    {
+      "epoch": 0.10231575803475237,
+      "grad_norm": 0.31640625,
+      "learning_rate": 9.447738693467337e-05,
+      "loss": 1.4618,
+      "step": 1199
+    },
+    {
+      "epoch": 0.10240109227831763,
+      "grad_norm": 0.326171875,
+      "learning_rate": 9.447236180904523e-05,
+      "loss": 1.2337,
+      "step": 1200
+    },
+    {
+      "epoch": 0.1024864265218829,
+      "grad_norm": 0.369140625,
+      "learning_rate": 9.446733668341708e-05,
+      "loss": 1.4728,
+      "step": 1201
+    },
+    {
+      "epoch": 0.10257176076544816,
+      "grad_norm": 0.30859375,
+      "learning_rate": 9.446231155778896e-05,
+      "loss": 1.3977,
+      "step": 1202
+    },
+    {
+      "epoch": 0.10265709500901343,
+      "grad_norm": 0.30078125,
+      "learning_rate": 9.445728643216081e-05,
+      "loss": 1.4819,
+      "step": 1203
+    },
+    {
+      "epoch": 0.10274242925257869,
+      "grad_norm": 0.3203125,
+      "learning_rate": 9.445226130653267e-05,
+      "loss": 1.24,
+      "step": 1204
+    },
+    {
+      "epoch": 0.10282776349614396,
+      "grad_norm": 0.33203125,
+      "learning_rate": 9.444723618090453e-05,
+      "loss": 1.4156,
+      "step": 1205
+    },
+    {
+      "epoch": 0.10291309773970922,
+      "grad_norm": 0.283203125,
+      "learning_rate": 9.444221105527638e-05,
+      "loss": 1.2449,
+      "step": 1206
+    },
+    {
+      "epoch": 0.10299843198327449,
+      "grad_norm": 0.318359375,
+      "learning_rate": 9.443718592964824e-05,
+      "loss": 1.4098,
+      "step": 1207
+    },
+    {
+      "epoch": 0.10308376622683975,
+      "grad_norm": 0.318359375,
+      "learning_rate": 9.44321608040201e-05,
+      "loss": 1.477,
+      "step": 1208
+    },
+    {
+      "epoch": 0.10316910047040502,
+      "grad_norm": 0.306640625,
+      "learning_rate": 9.442713567839197e-05,
+      "loss": 1.4231,
+      "step": 1209
+    },
+    {
+      "epoch": 0.10325443471397028,
+      "grad_norm": 0.294921875,
+      "learning_rate": 9.442211055276381e-05,
+      "loss": 1.1363,
+      "step": 1210
+    },
+    {
+      "epoch": 0.10333976895753555,
+      "grad_norm": 0.37109375,
+      "learning_rate": 9.441708542713569e-05,
+      "loss": 1.2383,
+      "step": 1211
+    },
+    {
+      "epoch": 0.10342510320110081,
+      "grad_norm": 0.353515625,
+      "learning_rate": 9.441206030150754e-05,
+      "loss": 1.4936,
+      "step": 1212
+    },
+    {
+      "epoch": 0.10351043744466608,
+      "grad_norm": 4.3125,
+      "learning_rate": 9.44070351758794e-05,
+      "loss": 1.5887,
+      "step": 1213
+    },
+    {
+      "epoch": 0.10359577168823134,
+      "grad_norm": 0.318359375,
+      "learning_rate": 9.440201005025126e-05,
+      "loss": 1.2671,
+      "step": 1214
+    },
+    {
+      "epoch": 0.10368110593179661,
+      "grad_norm": 0.306640625,
+      "learning_rate": 9.439698492462312e-05,
+      "loss": 1.3602,
+      "step": 1215
+    },
+    {
+      "epoch": 0.10376644017536187,
+      "grad_norm": 0.279296875,
+      "learning_rate": 9.439195979899497e-05,
+      "loss": 1.2588,
+      "step": 1216
+    },
+    {
+      "epoch": 0.10385177441892714,
+      "grad_norm": 0.310546875,
+      "learning_rate": 9.438693467336683e-05,
+      "loss": 1.3548,
+      "step": 1217
+    },
+    {
+      "epoch": 0.1039371086624924,
+      "grad_norm": 0.353515625,
+      "learning_rate": 9.43819095477387e-05,
+      "loss": 1.4364,
+      "step": 1218
+    },
+    {
+      "epoch": 0.10402244290605767,
+      "grad_norm": 0.283203125,
+      "learning_rate": 9.437688442211056e-05,
+      "loss": 1.2901,
+      "step": 1219
+    },
+    {
+      "epoch": 0.10410777714962292,
+      "grad_norm": 0.30859375,
+      "learning_rate": 9.437185929648241e-05,
+      "loss": 1.3073,
+      "step": 1220
+    },
+    {
+      "epoch": 0.1041931113931882,
+      "grad_norm": 0.3125,
+      "learning_rate": 9.436683417085428e-05,
+      "loss": 1.3083,
+      "step": 1221
+    },
+    {
+      "epoch": 0.10427844563675345,
+      "grad_norm": 0.3203125,
+      "learning_rate": 9.436180904522613e-05,
+      "loss": 1.3715,
+      "step": 1222
+    },
+    {
+      "epoch": 0.10436377988031872,
+      "grad_norm": 0.33984375,
+      "learning_rate": 9.4356783919598e-05,
+      "loss": 1.4082,
+      "step": 1223
+    },
+    {
+      "epoch": 0.10444911412388398,
+      "grad_norm": 0.3046875,
+      "learning_rate": 9.435175879396986e-05,
+      "loss": 1.1787,
+      "step": 1224
+    },
+    {
+      "epoch": 0.10453444836744925,
+      "grad_norm": 0.29296875,
+      "learning_rate": 9.434673366834172e-05,
+      "loss": 1.2793,
+      "step": 1225
+    },
+    {
+      "epoch": 0.10461978261101451,
+      "grad_norm": 0.3671875,
+      "learning_rate": 9.434170854271357e-05,
+      "loss": 1.3337,
+      "step": 1226
+    },
+    {
+      "epoch": 0.10470511685457978,
+      "grad_norm": 0.30859375,
+      "learning_rate": 9.433668341708544e-05,
+      "loss": 1.4945,
+      "step": 1227
+    },
+    {
+      "epoch": 0.10479045109814504,
+      "grad_norm": 0.3984375,
+      "learning_rate": 9.433165829145729e-05,
+      "loss": 1.462,
+      "step": 1228
+    },
+    {
+      "epoch": 0.10487578534171031,
+      "grad_norm": 0.30859375,
+      "learning_rate": 9.432663316582915e-05,
+      "loss": 1.2387,
+      "step": 1229
+    },
+    {
+      "epoch": 0.10496111958527557,
+      "grad_norm": 0.3125,
+      "learning_rate": 9.432160804020101e-05,
+      "loss": 1.3557,
+      "step": 1230
+    },
+    {
+      "epoch": 0.10504645382884084,
+      "grad_norm": 0.302734375,
+      "learning_rate": 9.431658291457288e-05,
+      "loss": 1.3068,
+      "step": 1231
+    },
+    {
+      "epoch": 0.1051317880724061,
+      "grad_norm": 0.32421875,
+      "learning_rate": 9.431155778894472e-05,
+      "loss": 1.3712,
+      "step": 1232
+    },
+    {
+      "epoch": 0.10521712231597137,
+      "grad_norm": 0.296875,
+      "learning_rate": 9.430653266331659e-05,
+      "loss": 1.2886,
+      "step": 1233
+    },
+    {
+      "epoch": 0.10530245655953664,
+      "grad_norm": 0.28515625,
+      "learning_rate": 9.430150753768845e-05,
+      "loss": 1.1449,
+      "step": 1234
+    },
+    {
+      "epoch": 0.1053877908031019,
+      "grad_norm": 0.314453125,
+      "learning_rate": 9.429648241206031e-05,
+      "loss": 1.3099,
+      "step": 1235
+    },
+    {
+      "epoch": 0.10547312504666717,
+      "grad_norm": 0.3046875,
+      "learning_rate": 9.429145728643216e-05,
+      "loss": 1.3578,
+      "step": 1236
+    },
+    {
+      "epoch": 0.10555845929023243,
+      "grad_norm": 0.3515625,
+      "learning_rate": 9.428643216080403e-05,
+      "loss": 1.2041,
+      "step": 1237
+    },
+    {
+      "epoch": 0.1056437935337977,
+      "grad_norm": 0.310546875,
+      "learning_rate": 9.428140703517588e-05,
+      "loss": 1.3481,
+      "step": 1238
+    },
+    {
+      "epoch": 0.10572912777736296,
+      "grad_norm": 0.259765625,
+      "learning_rate": 9.427638190954775e-05,
+      "loss": 1.1176,
+      "step": 1239
+    },
+    {
+      "epoch": 0.10581446202092823,
+      "grad_norm": 0.310546875,
+      "learning_rate": 9.427135678391961e-05,
+      "loss": 1.2424,
+      "step": 1240
+    },
+    {
+      "epoch": 0.10589979626449349,
+      "grad_norm": 0.29296875,
+      "learning_rate": 9.426633165829146e-05,
+      "loss": 1.1172,
+      "step": 1241
+    },
+    {
+      "epoch": 0.10598513050805876,
+      "grad_norm": 0.330078125,
+      "learning_rate": 9.426130653266332e-05,
+      "loss": 1.383,
+      "step": 1242
+    },
+    {
+      "epoch": 0.10607046475162402,
+      "grad_norm": 0.30859375,
+      "learning_rate": 9.425628140703518e-05,
+      "loss": 1.3154,
+      "step": 1243
+    },
+    {
+      "epoch": 0.10615579899518929,
+      "grad_norm": 0.29296875,
+      "learning_rate": 9.425125628140704e-05,
+      "loss": 1.1934,
+      "step": 1244
+    },
+    {
+      "epoch": 0.10624113323875455,
+      "grad_norm": 0.287109375,
+      "learning_rate": 9.424623115577889e-05,
+      "loss": 1.3165,
+      "step": 1245
+    },
+    {
+      "epoch": 0.10632646748231982,
+      "grad_norm": 0.302734375,
+      "learning_rate": 9.424120603015077e-05,
+      "loss": 1.1671,
+      "step": 1246
+    },
+    {
+      "epoch": 0.10641180172588507,
+      "grad_norm": 0.283203125,
+      "learning_rate": 9.423618090452261e-05,
+      "loss": 1.2845,
+      "step": 1247
+    },
+    {
+      "epoch": 0.10649713596945035,
+      "grad_norm": 0.333984375,
+      "learning_rate": 9.423115577889448e-05,
+      "loss": 1.3658,
+      "step": 1248
+    },
+    {
+      "epoch": 0.1065824702130156,
+      "grad_norm": 0.32421875,
+      "learning_rate": 9.422613065326632e-05,
+      "loss": 1.3176,
+      "step": 1249
+    },
+    {
+      "epoch": 0.10666780445658088,
+      "grad_norm": 0.3359375,
+      "learning_rate": 9.42211055276382e-05,
+      "loss": 1.3618,
+      "step": 1250
+    },
+    {
+      "epoch": 0.10675313870014613,
+      "grad_norm": 0.326171875,
+      "learning_rate": 9.421608040201005e-05,
+      "loss": 1.2907,
+      "step": 1251
+    },
+    {
+      "epoch": 0.1068384729437114,
+      "grad_norm": 0.32421875,
+      "learning_rate": 9.421105527638191e-05,
+      "loss": 1.4733,
+      "step": 1252
+    },
+    {
+      "epoch": 0.10692380718727666,
+      "grad_norm": 0.318359375,
+      "learning_rate": 9.420603015075377e-05,
+      "loss": 1.4793,
+      "step": 1253
+    },
+    {
+      "epoch": 0.10700914143084193,
+      "grad_norm": 0.3203125,
+      "learning_rate": 9.420100502512564e-05,
+      "loss": 1.2664,
+      "step": 1254
+    },
+    {
+      "epoch": 0.10709447567440719,
+      "grad_norm": 0.328125,
+      "learning_rate": 9.419597989949748e-05,
+      "loss": 1.4601,
+      "step": 1255
+    },
+    {
+      "epoch": 0.10717980991797246,
+      "grad_norm": 0.30078125,
+      "learning_rate": 9.419095477386936e-05,
+      "loss": 1.2501,
+      "step": 1256
+    },
+    {
+      "epoch": 0.10726514416153772,
+      "grad_norm": 0.30078125,
+      "learning_rate": 9.418592964824121e-05,
+      "loss": 1.3505,
+      "step": 1257
+    },
+    {
+      "epoch": 0.10735047840510299,
+      "grad_norm": 0.33203125,
+      "learning_rate": 9.418090452261307e-05,
+      "loss": 1.2536,
+      "step": 1258
+    },
+    {
+      "epoch": 0.10743581264866825,
+      "grad_norm": 0.353515625,
+      "learning_rate": 9.417587939698493e-05,
+      "loss": 1.4409,
+      "step": 1259
+    },
+    {
+      "epoch": 0.10752114689223352,
+      "grad_norm": 0.32421875,
+      "learning_rate": 9.41708542713568e-05,
+      "loss": 1.2701,
+      "step": 1260
+    },
+    {
+      "epoch": 0.10760648113579878,
+      "grad_norm": 0.2890625,
+      "learning_rate": 9.416582914572864e-05,
+      "loss": 1.3374,
+      "step": 1261
+    },
+    {
+      "epoch": 0.10769181537936405,
+      "grad_norm": 0.3359375,
+      "learning_rate": 9.41608040201005e-05,
+      "loss": 1.3722,
+      "step": 1262
+    },
+    {
+      "epoch": 0.10777714962292931,
+      "grad_norm": 0.30859375,
+      "learning_rate": 9.415577889447237e-05,
+      "loss": 1.3045,
+      "step": 1263
+    },
+    {
+      "epoch": 0.10786248386649458,
+      "grad_norm": 0.294921875,
+      "learning_rate": 9.415075376884423e-05,
+      "loss": 1.2823,
+      "step": 1264
+    },
+    {
+      "epoch": 0.10794781811005984,
+      "grad_norm": 0.318359375,
+      "learning_rate": 9.414572864321608e-05,
+      "loss": 1.4643,
+      "step": 1265
+    },
+    {
+      "epoch": 0.10803315235362511,
+      "grad_norm": 0.283203125,
+      "learning_rate": 9.414070351758795e-05,
+      "loss": 1.3461,
+      "step": 1266
+    },
+    {
+      "epoch": 0.10811848659719037,
+      "grad_norm": 0.3515625,
+      "learning_rate": 9.41356783919598e-05,
+      "loss": 1.4715,
+      "step": 1267
+    },
+    {
+      "epoch": 0.10820382084075564,
+      "grad_norm": 0.365234375,
+      "learning_rate": 9.413065326633166e-05,
+      "loss": 1.2455,
+      "step": 1268
+    },
+    {
+      "epoch": 0.1082891550843209,
+      "grad_norm": 0.310546875,
+      "learning_rate": 9.412562814070352e-05,
+      "loss": 1.2344,
+      "step": 1269
+    },
+    {
+      "epoch": 0.10837448932788617,
+      "grad_norm": 0.28125,
+      "learning_rate": 9.412060301507539e-05,
+      "loss": 1.2255,
+      "step": 1270
+    },
+    {
+      "epoch": 0.10845982357145142,
+      "grad_norm": 0.298828125,
+      "learning_rate": 9.411557788944724e-05,
+      "loss": 1.3467,
+      "step": 1271
+    },
+    {
+      "epoch": 0.1085451578150167,
+      "grad_norm": 0.328125,
+      "learning_rate": 9.41105527638191e-05,
+      "loss": 1.4181,
+      "step": 1272
+    },
+    {
+      "epoch": 0.10863049205858195,
+      "grad_norm": 0.28515625,
+      "learning_rate": 9.410552763819096e-05,
+      "loss": 1.1255,
+      "step": 1273
+    },
+    {
+      "epoch": 0.10871582630214723,
+      "grad_norm": 0.333984375,
+      "learning_rate": 9.410050251256282e-05,
+      "loss": 1.3963,
+      "step": 1274
+    },
+    {
+      "epoch": 0.10880116054571248,
+      "grad_norm": 0.3125,
+      "learning_rate": 9.409547738693468e-05,
+      "loss": 1.3,
+      "step": 1275
+    },
+    {
+      "epoch": 0.10888649478927775,
+      "grad_norm": 0.283203125,
+      "learning_rate": 9.409045226130653e-05,
+      "loss": 1.3346,
+      "step": 1276
+    },
+    {
+      "epoch": 0.10897182903284301,
+      "grad_norm": 0.345703125,
+      "learning_rate": 9.40854271356784e-05,
+      "loss": 1.4137,
+      "step": 1277
+    },
+    {
+      "epoch": 0.10905716327640828,
+      "grad_norm": 0.357421875,
+      "learning_rate": 9.408040201005026e-05,
+      "loss": 1.4214,
+      "step": 1278
+    },
+    {
+      "epoch": 0.10914249751997354,
+      "grad_norm": 0.28125,
+      "learning_rate": 9.407537688442212e-05,
+      "loss": 1.387,
+      "step": 1279
+    },
+    {
+      "epoch": 0.10922783176353881,
+      "grad_norm": 0.330078125,
+      "learning_rate": 9.407035175879397e-05,
+      "loss": 1.4503,
+      "step": 1280
+    },
+    {
+      "epoch": 0.10931316600710407,
+      "grad_norm": 0.3046875,
+      "learning_rate": 9.406532663316583e-05,
+      "loss": 1.3951,
+      "step": 1281
+    },
+    {
+      "epoch": 0.10939850025066934,
+      "grad_norm": 0.365234375,
+      "learning_rate": 9.406030150753769e-05,
+      "loss": 1.6485,
+      "step": 1282
+    },
+    {
+      "epoch": 0.1094838344942346,
+      "grad_norm": 0.3125,
+      "learning_rate": 9.405527638190955e-05,
+      "loss": 1.3528,
+      "step": 1283
+    },
+    {
+      "epoch": 0.10956916873779987,
+      "grad_norm": 0.31640625,
+      "learning_rate": 9.40502512562814e-05,
+      "loss": 1.3463,
+      "step": 1284
+    },
+    {
+      "epoch": 0.10965450298136513,
+      "grad_norm": 0.337890625,
+      "learning_rate": 9.404522613065328e-05,
+      "loss": 1.4435,
+      "step": 1285
+    },
+    {
+      "epoch": 0.1097398372249304,
+      "grad_norm": 0.3046875,
+      "learning_rate": 9.404020100502513e-05,
+      "loss": 1.2741,
+      "step": 1286
+    },
+    {
+      "epoch": 0.10982517146849566,
+      "grad_norm": 0.296875,
+      "learning_rate": 9.403517587939699e-05,
+      "loss": 1.2217,
+      "step": 1287
+    },
+    {
+      "epoch": 0.10991050571206093,
+      "grad_norm": 0.333984375,
+      "learning_rate": 9.403015075376885e-05,
+      "loss": 1.4001,
+      "step": 1288
+    },
+    {
+      "epoch": 0.10999583995562619,
+      "grad_norm": 0.333984375,
+      "learning_rate": 9.402512562814071e-05,
+      "loss": 1.5025,
+      "step": 1289
+    },
+    {
+      "epoch": 0.11008117419919146,
+      "grad_norm": 0.30859375,
+      "learning_rate": 9.402010050251256e-05,
+      "loss": 1.3302,
+      "step": 1290
+    },
+    {
+      "epoch": 0.11016650844275672,
+      "grad_norm": 0.29296875,
+      "learning_rate": 9.401507537688444e-05,
+      "loss": 1.3329,
+      "step": 1291
+    },
+    {
+      "epoch": 0.11025184268632199,
+      "grad_norm": 0.29296875,
+      "learning_rate": 9.401005025125628e-05,
+      "loss": 1.2649,
+      "step": 1292
+    },
+    {
+      "epoch": 0.11033717692988725,
+      "grad_norm": 0.322265625,
+      "learning_rate": 9.400502512562815e-05,
+      "loss": 1.4107,
+      "step": 1293
+    },
+    {
+      "epoch": 0.11042251117345252,
+      "grad_norm": 0.279296875,
+      "learning_rate": 9.4e-05,
+      "loss": 1.3736,
+      "step": 1294
+    },
+    {
+      "epoch": 0.11050784541701777,
+      "grad_norm": 0.322265625,
+      "learning_rate": 9.399497487437187e-05,
+      "loss": 1.2525,
+      "step": 1295
+    },
+    {
+      "epoch": 0.11059317966058305,
+      "grad_norm": 0.341796875,
+      "learning_rate": 9.398994974874372e-05,
+      "loss": 1.4613,
+      "step": 1296
+    },
+    {
+      "epoch": 0.1106785139041483,
+      "grad_norm": 0.326171875,
+      "learning_rate": 9.398492462311558e-05,
+      "loss": 1.4314,
+      "step": 1297
+    },
+    {
+      "epoch": 0.11076384814771358,
+      "grad_norm": 0.33203125,
+      "learning_rate": 9.397989949748744e-05,
+      "loss": 1.1376,
+      "step": 1298
+    },
+    {
+      "epoch": 0.11084918239127885,
+      "grad_norm": 0.3515625,
+      "learning_rate": 9.39748743718593e-05,
+      "loss": 1.2772,
+      "step": 1299
+    },
+    {
+      "epoch": 0.1109345166348441,
+      "grad_norm": 0.287109375,
+      "learning_rate": 9.396984924623115e-05,
+      "loss": 1.3063,
+      "step": 1300
+    },
+    {
+      "epoch": 0.11101985087840938,
+      "grad_norm": 0.30078125,
+      "learning_rate": 9.396482412060303e-05,
+      "loss": 1.3122,
+      "step": 1301
+    },
+    {
+      "epoch": 0.11110518512197463,
+      "grad_norm": 0.3203125,
+      "learning_rate": 9.395979899497488e-05,
+      "loss": 1.4451,
+      "step": 1302
+    },
+    {
+      "epoch": 0.1111905193655399,
+      "grad_norm": 0.326171875,
+      "learning_rate": 9.395477386934674e-05,
+      "loss": 1.327,
+      "step": 1303
+    },
+    {
+      "epoch": 0.11127585360910516,
+      "grad_norm": 0.306640625,
+      "learning_rate": 9.39497487437186e-05,
+      "loss": 1.3376,
+      "step": 1304
+    },
+    {
+      "epoch": 0.11136118785267043,
+      "grad_norm": 0.267578125,
+      "learning_rate": 9.394472361809046e-05,
+      "loss": 1.2783,
+      "step": 1305
+    },
+    {
+      "epoch": 0.11144652209623569,
+      "grad_norm": 0.34375,
+      "learning_rate": 9.393969849246231e-05,
+      "loss": 1.4416,
+      "step": 1306
+    },
+    {
+      "epoch": 0.11153185633980096,
+      "grad_norm": 0.333984375,
+      "learning_rate": 9.393467336683417e-05,
+      "loss": 1.434,
+      "step": 1307
+    },
+    {
+      "epoch": 0.11161719058336622,
+      "grad_norm": 0.287109375,
+      "learning_rate": 9.392964824120604e-05,
+      "loss": 1.3551,
+      "step": 1308
+    },
+    {
+      "epoch": 0.11170252482693149,
+      "grad_norm": 0.341796875,
+      "learning_rate": 9.39246231155779e-05,
+      "loss": 1.3086,
+      "step": 1309
+    },
+    {
+      "epoch": 0.11178785907049675,
+      "grad_norm": 0.298828125,
+      "learning_rate": 9.391959798994975e-05,
+      "loss": 1.2479,
+      "step": 1310
+    },
+    {
+      "epoch": 0.11187319331406202,
+      "grad_norm": 0.296875,
+      "learning_rate": 9.391457286432161e-05,
+      "loss": 1.2318,
+      "step": 1311
+    },
+    {
+      "epoch": 0.11195852755762728,
+      "grad_norm": 0.318359375,
+      "learning_rate": 9.390954773869347e-05,
+      "loss": 1.4304,
+      "step": 1312
+    },
+    {
+      "epoch": 0.11204386180119255,
+      "grad_norm": 0.31640625,
+      "learning_rate": 9.390452261306533e-05,
+      "loss": 1.3709,
+      "step": 1313
+    },
+    {
+      "epoch": 0.11212919604475781,
+      "grad_norm": 0.322265625,
+      "learning_rate": 9.38994974874372e-05,
+      "loss": 1.2007,
+      "step": 1314
+    },
+    {
+      "epoch": 0.11221453028832308,
+      "grad_norm": 0.29296875,
+      "learning_rate": 9.389447236180904e-05,
+      "loss": 1.3801,
+      "step": 1315
+    },
+    {
+      "epoch": 0.11229986453188834,
+      "grad_norm": 0.337890625,
+      "learning_rate": 9.38894472361809e-05,
+      "loss": 1.4037,
+      "step": 1316
+    },
+    {
+      "epoch": 0.11238519877545361,
+      "grad_norm": 0.322265625,
+      "learning_rate": 9.388442211055277e-05,
+      "loss": 1.307,
+      "step": 1317
+    },
+    {
+      "epoch": 0.11247053301901887,
+      "grad_norm": 0.31640625,
+      "learning_rate": 9.387939698492463e-05,
+      "loss": 1.2305,
+      "step": 1318
+    },
+    {
+      "epoch": 0.11255586726258414,
+      "grad_norm": 0.416015625,
+      "learning_rate": 9.387437185929648e-05,
+      "loss": 1.5274,
+      "step": 1319
+    },
+    {
+      "epoch": 0.1126412015061494,
+      "grad_norm": 0.322265625,
+      "learning_rate": 9.386934673366835e-05,
+      "loss": 1.357,
+      "step": 1320
+    },
+    {
+      "epoch": 0.11272653574971467,
+      "grad_norm": 0.31640625,
+      "learning_rate": 9.38643216080402e-05,
+      "loss": 1.1687,
+      "step": 1321
+    },
+    {
+      "epoch": 0.11281186999327993,
+      "grad_norm": 0.3671875,
+      "learning_rate": 9.385929648241206e-05,
+      "loss": 1.4293,
+      "step": 1322
+    },
+    {
+      "epoch": 0.1128972042368452,
+      "grad_norm": 0.333984375,
+      "learning_rate": 9.385427135678393e-05,
+      "loss": 1.4857,
+      "step": 1323
+    },
+    {
+      "epoch": 0.11298253848041045,
+      "grad_norm": 0.37109375,
+      "learning_rate": 9.384924623115579e-05,
+      "loss": 1.4498,
+      "step": 1324
+    },
+    {
+      "epoch": 0.11306787272397573,
+      "grad_norm": 0.33203125,
+      "learning_rate": 9.384422110552764e-05,
+      "loss": 1.3203,
+      "step": 1325
+    },
+    {
+      "epoch": 0.11315320696754098,
+      "grad_norm": 0.31640625,
+      "learning_rate": 9.38391959798995e-05,
+      "loss": 1.3093,
+      "step": 1326
+    },
+    {
+      "epoch": 0.11323854121110626,
+      "grad_norm": 0.3046875,
+      "learning_rate": 9.383417085427136e-05,
+      "loss": 1.3309,
+      "step": 1327
+    },
+    {
+      "epoch": 0.11332387545467151,
+      "grad_norm": 0.359375,
+      "learning_rate": 9.382914572864322e-05,
+      "loss": 1.6233,
+      "step": 1328
+    },
+    {
+      "epoch": 0.11340920969823678,
+      "grad_norm": 0.298828125,
+      "learning_rate": 9.382412060301507e-05,
+      "loss": 1.4111,
+      "step": 1329
+    },
+    {
+      "epoch": 0.11349454394180204,
+      "grad_norm": 0.306640625,
+      "learning_rate": 9.381909547738695e-05,
+      "loss": 1.4237,
+      "step": 1330
+    },
+    {
+      "epoch": 0.11357987818536731,
+      "grad_norm": 0.302734375,
+      "learning_rate": 9.38140703517588e-05,
+      "loss": 1.4405,
+      "step": 1331
+    },
+    {
+      "epoch": 0.11366521242893257,
+      "grad_norm": 0.287109375,
+      "learning_rate": 9.380904522613066e-05,
+      "loss": 1.3108,
+      "step": 1332
+    },
+    {
+      "epoch": 0.11375054667249784,
+      "grad_norm": 0.314453125,
+      "learning_rate": 9.380402010050252e-05,
+      "loss": 1.2497,
+      "step": 1333
+    },
+    {
+      "epoch": 0.1138358809160631,
+      "grad_norm": 0.326171875,
+      "learning_rate": 9.379899497487438e-05,
+      "loss": 1.3772,
+      "step": 1334
+    },
+    {
+      "epoch": 0.11392121515962837,
+      "grad_norm": 0.29296875,
+      "learning_rate": 9.379396984924623e-05,
+      "loss": 1.3153,
+      "step": 1335
+    },
+    {
+      "epoch": 0.11400654940319363,
+      "grad_norm": 0.3203125,
+      "learning_rate": 9.37889447236181e-05,
+      "loss": 1.3436,
+      "step": 1336
+    },
+    {
+      "epoch": 0.1140918836467589,
+      "grad_norm": 0.3203125,
+      "learning_rate": 9.378391959798995e-05,
+      "loss": 1.4253,
+      "step": 1337
+    },
+    {
+      "epoch": 0.11417721789032416,
+      "grad_norm": 0.369140625,
+      "learning_rate": 9.377889447236182e-05,
+      "loss": 1.1631,
+      "step": 1338
+    },
+    {
+      "epoch": 0.11426255213388943,
+      "grad_norm": 0.2734375,
+      "learning_rate": 9.377386934673368e-05,
+      "loss": 1.1159,
+      "step": 1339
+    },
+    {
+      "epoch": 0.11434788637745469,
+      "grad_norm": 0.318359375,
+      "learning_rate": 9.376884422110554e-05,
+      "loss": 1.1696,
+      "step": 1340
+    },
+    {
+      "epoch": 0.11443322062101996,
+      "grad_norm": 0.328125,
+      "learning_rate": 9.376381909547739e-05,
+      "loss": 1.2474,
+      "step": 1341
+    },
+    {
+      "epoch": 0.11451855486458522,
+      "grad_norm": 0.283203125,
+      "learning_rate": 9.375879396984925e-05,
+      "loss": 1.3009,
+      "step": 1342
+    },
+    {
+      "epoch": 0.11460388910815049,
+      "grad_norm": 0.28515625,
+      "learning_rate": 9.375376884422111e-05,
+      "loss": 1.1385,
+      "step": 1343
+    },
+    {
+      "epoch": 0.11468922335171575,
+      "grad_norm": 0.3125,
+      "learning_rate": 9.374874371859297e-05,
+      "loss": 1.2862,
+      "step": 1344
+    },
+    {
+      "epoch": 0.11477455759528102,
+      "grad_norm": 0.306640625,
+      "learning_rate": 9.374371859296482e-05,
+      "loss": 1.2026,
+      "step": 1345
+    },
+    {
+      "epoch": 0.11485989183884628,
+      "grad_norm": 0.361328125,
+      "learning_rate": 9.373869346733668e-05,
+      "loss": 1.3274,
+      "step": 1346
+    },
+    {
+      "epoch": 0.11494522608241155,
+      "grad_norm": 0.27734375,
+      "learning_rate": 9.373366834170855e-05,
+      "loss": 1.3037,
+      "step": 1347
+    },
+    {
+      "epoch": 0.1150305603259768,
+      "grad_norm": 0.326171875,
+      "learning_rate": 9.372864321608041e-05,
+      "loss": 1.4523,
+      "step": 1348
+    },
+    {
+      "epoch": 0.11511589456954208,
+      "grad_norm": 0.306640625,
+      "learning_rate": 9.372361809045227e-05,
+      "loss": 1.2325,
+      "step": 1349
+    },
+    {
+      "epoch": 0.11520122881310733,
+      "grad_norm": 0.396484375,
+      "learning_rate": 9.371859296482412e-05,
+      "loss": 1.23,
+      "step": 1350
+    },
+    {
+      "epoch": 0.1152865630566726,
+      "grad_norm": 0.28515625,
+      "learning_rate": 9.371356783919598e-05,
+      "loss": 1.3123,
+      "step": 1351
+    },
+    {
+      "epoch": 0.11537189730023786,
+      "grad_norm": 0.27734375,
+      "learning_rate": 9.370854271356784e-05,
+      "loss": 1.108,
+      "step": 1352
+    },
+    {
+      "epoch": 0.11545723154380313,
+      "grad_norm": 0.30859375,
+      "learning_rate": 9.37035175879397e-05,
+      "loss": 1.3374,
+      "step": 1353
+    },
+    {
+      "epoch": 0.11554256578736839,
+      "grad_norm": 0.328125,
+      "learning_rate": 9.369849246231155e-05,
+      "loss": 1.4087,
+      "step": 1354
+    },
+    {
+      "epoch": 0.11562790003093366,
+      "grad_norm": 0.330078125,
+      "learning_rate": 9.369346733668343e-05,
+      "loss": 1.2411,
+      "step": 1355
+    },
+    {
+      "epoch": 0.11571323427449892,
+      "grad_norm": 0.30859375,
+      "learning_rate": 9.368844221105528e-05,
+      "loss": 1.2342,
+      "step": 1356
+    },
+    {
+      "epoch": 0.1157985685180642,
+      "grad_norm": 0.361328125,
+      "learning_rate": 9.368341708542714e-05,
+      "loss": 1.3929,
+      "step": 1357
+    },
+    {
+      "epoch": 0.11588390276162945,
+      "grad_norm": 0.322265625,
+      "learning_rate": 9.367839195979899e-05,
+      "loss": 1.3551,
+      "step": 1358
+    },
+    {
+      "epoch": 0.11596923700519472,
+      "grad_norm": 0.31640625,
+      "learning_rate": 9.367336683417086e-05,
+      "loss": 1.2237,
+      "step": 1359
+    },
+    {
+      "epoch": 0.11605457124875998,
+      "grad_norm": 0.302734375,
+      "learning_rate": 9.366834170854271e-05,
+      "loss": 1.2868,
+      "step": 1360
+    },
+    {
+      "epoch": 0.11613990549232525,
+      "grad_norm": 0.326171875,
+      "learning_rate": 9.366331658291457e-05,
+      "loss": 1.2755,
+      "step": 1361
+    },
+    {
+      "epoch": 0.11622523973589051,
+      "grad_norm": 0.322265625,
+      "learning_rate": 9.365829145728644e-05,
+      "loss": 1.4521,
+      "step": 1362
+    },
+    {
+      "epoch": 0.11631057397945578,
+      "grad_norm": 0.283203125,
+      "learning_rate": 9.36532663316583e-05,
+      "loss": 1.2697,
+      "step": 1363
+    },
+    {
+      "epoch": 0.11639590822302105,
+      "grad_norm": 0.296875,
+      "learning_rate": 9.364824120603015e-05,
+      "loss": 1.3259,
+      "step": 1364
+    },
+    {
+      "epoch": 0.11648124246658631,
+      "grad_norm": 0.3125,
+      "learning_rate": 9.364321608040202e-05,
+      "loss": 1.3184,
+      "step": 1365
+    },
+    {
+      "epoch": 0.11656657671015158,
+      "grad_norm": 0.322265625,
+      "learning_rate": 9.363819095477387e-05,
+      "loss": 1.2799,
+      "step": 1366
+    },
+    {
+      "epoch": 0.11665191095371684,
+      "grad_norm": 0.310546875,
+      "learning_rate": 9.363316582914573e-05,
+      "loss": 1.3662,
+      "step": 1367
+    },
+    {
+      "epoch": 0.11673724519728211,
+      "grad_norm": 0.279296875,
+      "learning_rate": 9.36281407035176e-05,
+      "loss": 1.2788,
+      "step": 1368
+    },
+    {
+      "epoch": 0.11682257944084737,
+      "grad_norm": 0.287109375,
+      "learning_rate": 9.362311557788946e-05,
+      "loss": 1.1637,
+      "step": 1369
+    },
+    {
+      "epoch": 0.11690791368441264,
+      "grad_norm": 0.306640625,
+      "learning_rate": 9.36180904522613e-05,
+      "loss": 1.2888,
+      "step": 1370
+    },
+    {
+      "epoch": 0.1169932479279779,
+      "grad_norm": 0.369140625,
+      "learning_rate": 9.361306532663318e-05,
+      "loss": 1.4161,
+      "step": 1371
+    },
+    {
+      "epoch": 0.11707858217154317,
+      "grad_norm": 0.3515625,
+      "learning_rate": 9.360804020100503e-05,
+      "loss": 1.4456,
+      "step": 1372
+    },
+    {
+      "epoch": 0.11716391641510843,
+      "grad_norm": 0.30859375,
+      "learning_rate": 9.360301507537689e-05,
+      "loss": 1.5161,
+      "step": 1373
+    },
+    {
+      "epoch": 0.1172492506586737,
+      "grad_norm": 0.310546875,
+      "learning_rate": 9.359798994974874e-05,
+      "loss": 1.3524,
+      "step": 1374
+    },
+    {
+      "epoch": 0.11733458490223896,
+      "grad_norm": 0.4453125,
+      "learning_rate": 9.359296482412062e-05,
+      "loss": 1.47,
+      "step": 1375
+    },
+    {
+      "epoch": 0.11741991914580423,
+      "grad_norm": 0.314453125,
+      "learning_rate": 9.358793969849246e-05,
+      "loss": 1.2959,
+      "step": 1376
+    },
+    {
+      "epoch": 0.11750525338936949,
+      "grad_norm": 0.369140625,
+      "learning_rate": 9.358291457286433e-05,
+      "loss": 1.5436,
+      "step": 1377
+    },
+    {
+      "epoch": 0.11759058763293476,
+      "grad_norm": 0.3046875,
+      "learning_rate": 9.357788944723619e-05,
+      "loss": 1.3796,
+      "step": 1378
+    },
+    {
+      "epoch": 0.11767592187650001,
+      "grad_norm": 0.322265625,
+      "learning_rate": 9.357286432160805e-05,
+      "loss": 1.082,
+      "step": 1379
+    },
+    {
+      "epoch": 0.11776125612006529,
+      "grad_norm": 0.298828125,
+      "learning_rate": 9.35678391959799e-05,
+      "loss": 1.3027,
+      "step": 1380
+    },
+    {
+      "epoch": 0.11784659036363054,
+      "grad_norm": 0.3125,
+      "learning_rate": 9.356281407035176e-05,
+      "loss": 1.3783,
+      "step": 1381
+    },
+    {
+      "epoch": 0.11793192460719581,
+      "grad_norm": 0.361328125,
+      "learning_rate": 9.355778894472362e-05,
+      "loss": 1.2626,
+      "step": 1382
+    },
+    {
+      "epoch": 0.11801725885076107,
+      "grad_norm": 0.291015625,
+      "learning_rate": 9.355276381909549e-05,
+      "loss": 1.4151,
+      "step": 1383
+    },
+    {
+      "epoch": 0.11810259309432634,
+      "grad_norm": 0.3046875,
+      "learning_rate": 9.354773869346735e-05,
+      "loss": 1.2773,
+      "step": 1384
+    },
+    {
+      "epoch": 0.1181879273378916,
+      "grad_norm": 0.310546875,
+      "learning_rate": 9.35427135678392e-05,
+      "loss": 1.3266,
+      "step": 1385
+    },
+    {
+      "epoch": 0.11827326158145687,
+      "grad_norm": 0.3203125,
+      "learning_rate": 9.353768844221106e-05,
+      "loss": 1.2559,
+      "step": 1386
+    },
+    {
+      "epoch": 0.11835859582502213,
+      "grad_norm": 0.3359375,
+      "learning_rate": 9.353266331658292e-05,
+      "loss": 1.3789,
+      "step": 1387
+    },
+    {
+      "epoch": 0.1184439300685874,
+      "grad_norm": 0.31640625,
+      "learning_rate": 9.352763819095478e-05,
+      "loss": 1.4742,
+      "step": 1388
+    },
+    {
+      "epoch": 0.11852926431215266,
+      "grad_norm": 0.2890625,
+      "learning_rate": 9.352261306532663e-05,
+      "loss": 1.4176,
+      "step": 1389
+    },
+    {
+      "epoch": 0.11861459855571793,
+      "grad_norm": 0.2890625,
+      "learning_rate": 9.351758793969849e-05,
+      "loss": 1.3704,
+      "step": 1390
+    },
+    {
+      "epoch": 0.11869993279928319,
+      "grad_norm": 0.26171875,
+      "learning_rate": 9.351256281407035e-05,
+      "loss": 1.22,
+      "step": 1391
+    },
+    {
+      "epoch": 0.11878526704284846,
+      "grad_norm": 0.294921875,
+      "learning_rate": 9.350753768844222e-05,
+      "loss": 1.2012,
+      "step": 1392
+    },
+    {
+      "epoch": 0.11887060128641372,
+      "grad_norm": 0.322265625,
+      "learning_rate": 9.350251256281406e-05,
+      "loss": 1.4565,
+      "step": 1393
+    },
+    {
+      "epoch": 0.11895593552997899,
+      "grad_norm": 0.3828125,
+      "learning_rate": 9.349748743718594e-05,
+      "loss": 1.5512,
+      "step": 1394
+    },
+    {
+      "epoch": 0.11904126977354425,
+      "grad_norm": 0.310546875,
+      "learning_rate": 9.349246231155779e-05,
+      "loss": 1.2607,
+      "step": 1395
+    },
+    {
+      "epoch": 0.11912660401710952,
+      "grad_norm": 0.322265625,
+      "learning_rate": 9.348743718592965e-05,
+      "loss": 1.3142,
+      "step": 1396
+    },
+    {
+      "epoch": 0.11921193826067478,
+      "grad_norm": 0.3203125,
+      "learning_rate": 9.348241206030151e-05,
+      "loss": 1.3598,
+      "step": 1397
+    },
+    {
+      "epoch": 0.11929727250424005,
+      "grad_norm": 0.294921875,
+      "learning_rate": 9.347738693467338e-05,
+      "loss": 1.3794,
+      "step": 1398
+    },
+    {
+      "epoch": 0.1193826067478053,
+      "grad_norm": 0.3359375,
+      "learning_rate": 9.347236180904522e-05,
+      "loss": 1.1818,
+      "step": 1399
+    },
+    {
+      "epoch": 0.11946794099137058,
+      "grad_norm": 0.36328125,
+      "learning_rate": 9.34673366834171e-05,
+      "loss": 1.4556,
+      "step": 1400
+    },
+    {
+      "epoch": 0.11955327523493584,
+      "grad_norm": 0.30078125,
+      "learning_rate": 9.346231155778895e-05,
+      "loss": 1.2601,
+      "step": 1401
+    },
+    {
+      "epoch": 0.1196386094785011,
+      "grad_norm": 0.275390625,
+      "learning_rate": 9.345728643216081e-05,
+      "loss": 1.2135,
+      "step": 1402
+    },
+    {
+      "epoch": 0.11972394372206636,
+      "grad_norm": 0.3359375,
+      "learning_rate": 9.345226130653266e-05,
+      "loss": 1.4675,
+      "step": 1403
+    },
+    {
+      "epoch": 0.11980927796563164,
+      "grad_norm": 0.298828125,
+      "learning_rate": 9.344723618090453e-05,
+      "loss": 1.4075,
+      "step": 1404
+    },
+    {
+      "epoch": 0.1198946122091969,
+      "grad_norm": 0.318359375,
+      "learning_rate": 9.344221105527638e-05,
+      "loss": 1.3385,
+      "step": 1405
+    },
+    {
+      "epoch": 0.11997994645276217,
+      "grad_norm": 0.287109375,
+      "learning_rate": 9.343718592964824e-05,
+      "loss": 1.3084,
+      "step": 1406
+    },
+    {
+      "epoch": 0.12006528069632742,
+      "grad_norm": 0.30859375,
+      "learning_rate": 9.34321608040201e-05,
+      "loss": 1.2108,
+      "step": 1407
+    },
+    {
+      "epoch": 0.1201506149398927,
+      "grad_norm": 0.3046875,
+      "learning_rate": 9.342713567839197e-05,
+      "loss": 1.4003,
+      "step": 1408
+    },
+    {
+      "epoch": 0.12023594918345795,
+      "grad_norm": 0.294921875,
+      "learning_rate": 9.342211055276382e-05,
+      "loss": 1.2786,
+      "step": 1409
+    },
+    {
+      "epoch": 0.12032128342702322,
+      "grad_norm": 0.326171875,
+      "learning_rate": 9.341708542713569e-05,
+      "loss": 1.3677,
+      "step": 1410
+    },
+    {
+      "epoch": 0.12040661767058848,
+      "grad_norm": 0.330078125,
+      "learning_rate": 9.341206030150754e-05,
+      "loss": 1.2692,
+      "step": 1411
+    },
+    {
+      "epoch": 0.12049195191415375,
+      "grad_norm": 0.310546875,
+      "learning_rate": 9.34070351758794e-05,
+      "loss": 1.2877,
+      "step": 1412
+    },
+    {
+      "epoch": 0.12057728615771901,
+      "grad_norm": 0.333984375,
+      "learning_rate": 9.340201005025126e-05,
+      "loss": 1.3741,
+      "step": 1413
+    },
+    {
+      "epoch": 0.12066262040128428,
+      "grad_norm": 0.328125,
+      "learning_rate": 9.339698492462313e-05,
+      "loss": 1.228,
+      "step": 1414
+    },
+    {
+      "epoch": 0.12074795464484954,
+      "grad_norm": 0.2734375,
+      "learning_rate": 9.339195979899498e-05,
+      "loss": 1.2013,
+      "step": 1415
+    },
+    {
+      "epoch": 0.12083328888841481,
+      "grad_norm": 0.5,
+      "learning_rate": 9.338693467336684e-05,
+      "loss": 1.3256,
+      "step": 1416
+    },
+    {
+      "epoch": 0.12091862313198007,
+      "grad_norm": 0.3203125,
+      "learning_rate": 9.33819095477387e-05,
+      "loss": 1.3276,
+      "step": 1417
+    },
+    {
+      "epoch": 0.12100395737554534,
+      "grad_norm": 0.33203125,
+      "learning_rate": 9.337688442211056e-05,
+      "loss": 1.3572,
+      "step": 1418
+    },
+    {
+      "epoch": 0.1210892916191106,
+      "grad_norm": 0.341796875,
+      "learning_rate": 9.337185929648241e-05,
+      "loss": 1.514,
+      "step": 1419
+    },
+    {
+      "epoch": 0.12117462586267587,
+      "grad_norm": 0.28125,
+      "learning_rate": 9.336683417085427e-05,
+      "loss": 1.2285,
+      "step": 1420
+    },
+    {
+      "epoch": 0.12125996010624113,
+      "grad_norm": 0.306640625,
+      "learning_rate": 9.336180904522613e-05,
+      "loss": 1.2514,
+      "step": 1421
+    },
+    {
+      "epoch": 0.1213452943498064,
+      "grad_norm": 0.326171875,
+      "learning_rate": 9.3356783919598e-05,
+      "loss": 1.3792,
+      "step": 1422
+    },
+    {
+      "epoch": 0.12143062859337166,
+      "grad_norm": 0.408203125,
+      "learning_rate": 9.335175879396986e-05,
+      "loss": 1.2483,
+      "step": 1423
+    },
+    {
+      "epoch": 0.12151596283693693,
+      "grad_norm": 0.345703125,
+      "learning_rate": 9.33467336683417e-05,
+      "loss": 1.4372,
+      "step": 1424
+    },
+    {
+      "epoch": 0.12160129708050219,
+      "grad_norm": 0.328125,
+      "learning_rate": 9.334170854271357e-05,
+      "loss": 1.1626,
+      "step": 1425
+    },
+    {
+      "epoch": 0.12168663132406746,
+      "grad_norm": 0.31640625,
+      "learning_rate": 9.333668341708543e-05,
+      "loss": 1.2953,
+      "step": 1426
+    },
+    {
+      "epoch": 0.12177196556763271,
+      "grad_norm": 0.322265625,
+      "learning_rate": 9.333165829145729e-05,
+      "loss": 1.3465,
+      "step": 1427
+    },
+    {
+      "epoch": 0.12185729981119799,
+      "grad_norm": 0.291015625,
+      "learning_rate": 9.332663316582914e-05,
+      "loss": 1.1558,
+      "step": 1428
+    },
+    {
+      "epoch": 0.12194263405476326,
+      "grad_norm": 0.29296875,
+      "learning_rate": 9.332160804020102e-05,
+      "loss": 1.1799,
+      "step": 1429
+    },
+    {
+      "epoch": 0.12202796829832852,
+      "grad_norm": 0.390625,
+      "learning_rate": 9.331658291457287e-05,
+      "loss": 1.3799,
+      "step": 1430
+    },
+    {
+      "epoch": 0.12211330254189379,
+      "grad_norm": 0.31640625,
+      "learning_rate": 9.331155778894473e-05,
+      "loss": 1.4263,
+      "step": 1431
+    },
+    {
+      "epoch": 0.12219863678545904,
+      "grad_norm": 0.302734375,
+      "learning_rate": 9.330653266331659e-05,
+      "loss": 1.3859,
+      "step": 1432
+    },
+    {
+      "epoch": 0.12228397102902432,
+      "grad_norm": 0.3671875,
+      "learning_rate": 9.330150753768845e-05,
+      "loss": 1.438,
+      "step": 1433
+    },
+    {
+      "epoch": 0.12236930527258957,
+      "grad_norm": 0.318359375,
+      "learning_rate": 9.32964824120603e-05,
+      "loss": 1.4246,
+      "step": 1434
+    },
+    {
+      "epoch": 0.12245463951615485,
+      "grad_norm": 0.3125,
+      "learning_rate": 9.329145728643216e-05,
+      "loss": 1.393,
+      "step": 1435
+    },
+    {
+      "epoch": 0.1225399737597201,
+      "grad_norm": 0.30859375,
+      "learning_rate": 9.328643216080402e-05,
+      "loss": 1.2203,
+      "step": 1436
+    },
+    {
+      "epoch": 0.12262530800328537,
+      "grad_norm": 0.310546875,
+      "learning_rate": 9.328140703517589e-05,
+      "loss": 1.2853,
+      "step": 1437
+    },
+    {
+      "epoch": 0.12271064224685063,
+      "grad_norm": 0.296875,
+      "learning_rate": 9.327638190954773e-05,
+      "loss": 1.2891,
+      "step": 1438
+    },
+    {
+      "epoch": 0.1227959764904159,
+      "grad_norm": 0.333984375,
+      "learning_rate": 9.327135678391961e-05,
+      "loss": 1.4089,
+      "step": 1439
+    },
+    {
+      "epoch": 0.12288131073398116,
+      "grad_norm": 0.35546875,
+      "learning_rate": 9.326633165829146e-05,
+      "loss": 1.3873,
+      "step": 1440
+    },
+    {
+      "epoch": 0.12296664497754643,
+      "grad_norm": 0.27734375,
+      "learning_rate": 9.326130653266332e-05,
+      "loss": 1.243,
+      "step": 1441
+    },
+    {
+      "epoch": 0.12305197922111169,
+      "grad_norm": 0.283203125,
+      "learning_rate": 9.325628140703518e-05,
+      "loss": 1.1859,
+      "step": 1442
+    },
+    {
+      "epoch": 0.12313731346467696,
+      "grad_norm": 0.28515625,
+      "learning_rate": 9.325125628140704e-05,
+      "loss": 1.309,
+      "step": 1443
+    },
+    {
+      "epoch": 0.12322264770824222,
+      "grad_norm": 0.291015625,
+      "learning_rate": 9.324623115577889e-05,
+      "loss": 1.1615,
+      "step": 1444
+    },
+    {
+      "epoch": 0.12330798195180749,
+      "grad_norm": 0.326171875,
+      "learning_rate": 9.324120603015077e-05,
+      "loss": 1.6309,
+      "step": 1445
+    },
+    {
+      "epoch": 0.12339331619537275,
+      "grad_norm": 0.30859375,
+      "learning_rate": 9.323618090452262e-05,
+      "loss": 1.3351,
+      "step": 1446
+    },
+    {
+      "epoch": 0.12347865043893802,
+      "grad_norm": 1.1171875,
+      "learning_rate": 9.323115577889448e-05,
+      "loss": 1.4337,
+      "step": 1447
+    },
+    {
+      "epoch": 0.12356398468250328,
+      "grad_norm": 0.3203125,
+      "learning_rate": 9.322613065326634e-05,
+      "loss": 1.4513,
+      "step": 1448
+    },
+    {
+      "epoch": 0.12364931892606855,
+      "grad_norm": 0.33203125,
+      "learning_rate": 9.32211055276382e-05,
+      "loss": 1.4264,
+      "step": 1449
+    },
+    {
+      "epoch": 0.12373465316963381,
+      "grad_norm": 0.298828125,
+      "learning_rate": 9.321608040201005e-05,
+      "loss": 1.1431,
+      "step": 1450
+    },
+    {
+      "epoch": 0.12381998741319908,
+      "grad_norm": 0.298828125,
+      "learning_rate": 9.321105527638191e-05,
+      "loss": 1.287,
+      "step": 1451
+    },
+    {
+      "epoch": 0.12390532165676434,
+      "grad_norm": 0.3046875,
+      "learning_rate": 9.320603015075378e-05,
+      "loss": 1.3962,
+      "step": 1452
+    },
+    {
+      "epoch": 0.12399065590032961,
+      "grad_norm": 0.306640625,
+      "learning_rate": 9.320100502512564e-05,
+      "loss": 1.2267,
+      "step": 1453
+    },
+    {
+      "epoch": 0.12407599014389487,
+      "grad_norm": 0.2890625,
+      "learning_rate": 9.319597989949749e-05,
+      "loss": 1.2399,
+      "step": 1454
+    },
+    {
+      "epoch": 0.12416132438746014,
+      "grad_norm": 0.35546875,
+      "learning_rate": 9.319095477386935e-05,
+      "loss": 1.3872,
+      "step": 1455
+    },
+    {
+      "epoch": 0.1242466586310254,
+      "grad_norm": 0.31640625,
+      "learning_rate": 9.318592964824121e-05,
+      "loss": 1.276,
+      "step": 1456
+    },
+    {
+      "epoch": 0.12433199287459067,
+      "grad_norm": 0.328125,
+      "learning_rate": 9.318090452261307e-05,
+      "loss": 1.227,
+      "step": 1457
+    },
+    {
+      "epoch": 0.12441732711815592,
+      "grad_norm": 0.37109375,
+      "learning_rate": 9.317587939698493e-05,
+      "loss": 1.3223,
+      "step": 1458
+    },
+    {
+      "epoch": 0.1245026613617212,
+      "grad_norm": 0.3125,
+      "learning_rate": 9.317085427135678e-05,
+      "loss": 1.2242,
+      "step": 1459
+    },
+    {
+      "epoch": 0.12458799560528645,
+      "grad_norm": 0.318359375,
+      "learning_rate": 9.316582914572864e-05,
+      "loss": 1.4762,
+      "step": 1460
+    },
+    {
+      "epoch": 0.12467332984885172,
+      "grad_norm": 0.34375,
+      "learning_rate": 9.316080402010051e-05,
+      "loss": 1.3564,
+      "step": 1461
+    },
+    {
+      "epoch": 0.12475866409241698,
+      "grad_norm": 0.3203125,
+      "learning_rate": 9.315577889447237e-05,
+      "loss": 1.4305,
+      "step": 1462
+    },
+    {
+      "epoch": 0.12484399833598225,
+      "grad_norm": 0.28515625,
+      "learning_rate": 9.315075376884422e-05,
+      "loss": 1.2938,
+      "step": 1463
+    },
+    {
+      "epoch": 0.12492933257954751,
+      "grad_norm": 0.294921875,
+      "learning_rate": 9.314572864321609e-05,
+      "loss": 1.2876,
+      "step": 1464
+    },
+    {
+      "epoch": 0.12501466682311277,
+      "grad_norm": 0.314453125,
+      "learning_rate": 9.314070351758794e-05,
+      "loss": 1.3643,
+      "step": 1465
+    },
+    {
+      "epoch": 0.12510000106667804,
+      "grad_norm": 0.28515625,
+      "learning_rate": 9.31356783919598e-05,
+      "loss": 1.2878,
+      "step": 1466
+    },
+    {
+      "epoch": 0.1251853353102433,
+      "grad_norm": 0.3125,
+      "learning_rate": 9.313065326633165e-05,
+      "loss": 1.3228,
+      "step": 1467
+    },
+    {
+      "epoch": 0.12527066955380858,
+      "grad_norm": 0.318359375,
+      "learning_rate": 9.312562814070353e-05,
+      "loss": 1.2678,
+      "step": 1468
+    },
+    {
+      "epoch": 0.12535600379737383,
+      "grad_norm": 0.3203125,
+      "learning_rate": 9.312060301507538e-05,
+      "loss": 1.3039,
+      "step": 1469
+    },
+    {
+      "epoch": 0.1254413380409391,
+      "grad_norm": 0.3125,
+      "learning_rate": 9.311557788944724e-05,
+      "loss": 1.3492,
+      "step": 1470
+    },
+    {
+      "epoch": 0.12552667228450437,
+      "grad_norm": 0.31640625,
+      "learning_rate": 9.31105527638191e-05,
+      "loss": 1.4216,
+      "step": 1471
+    },
+    {
+      "epoch": 0.12561200652806964,
+      "grad_norm": 0.337890625,
+      "learning_rate": 9.310552763819096e-05,
+      "loss": 1.2161,
+      "step": 1472
+    },
+    {
+      "epoch": 0.12569734077163489,
+      "grad_norm": 0.27734375,
+      "learning_rate": 9.310050251256281e-05,
+      "loss": 1.2469,
+      "step": 1473
+    },
+    {
+      "epoch": 0.12578267501520016,
+      "grad_norm": 0.314453125,
+      "learning_rate": 9.309547738693469e-05,
+      "loss": 1.4488,
+      "step": 1474
+    },
+    {
+      "epoch": 0.12586800925876543,
+      "grad_norm": 0.28515625,
+      "learning_rate": 9.309045226130653e-05,
+      "loss": 1.4823,
+      "step": 1475
+    },
+    {
+      "epoch": 0.1259533435023307,
+      "grad_norm": 0.3359375,
+      "learning_rate": 9.30854271356784e-05,
+      "loss": 1.3294,
+      "step": 1476
+    },
+    {
+      "epoch": 0.12603867774589594,
+      "grad_norm": 0.353515625,
+      "learning_rate": 9.308040201005026e-05,
+      "loss": 1.4159,
+      "step": 1477
+    },
+    {
+      "epoch": 0.12612401198946122,
+      "grad_norm": 0.306640625,
+      "learning_rate": 9.307537688442212e-05,
+      "loss": 1.2869,
+      "step": 1478
+    },
+    {
+      "epoch": 0.1262093462330265,
+      "grad_norm": 0.32421875,
+      "learning_rate": 9.307035175879397e-05,
+      "loss": 1.3068,
+      "step": 1479
+    },
+    {
+      "epoch": 0.12629468047659176,
+      "grad_norm": 0.2734375,
+      "learning_rate": 9.306532663316585e-05,
+      "loss": 1.2561,
+      "step": 1480
+    },
+    {
+      "epoch": 0.126380014720157,
+      "grad_norm": 0.345703125,
+      "learning_rate": 9.30603015075377e-05,
+      "loss": 1.3812,
+      "step": 1481
+    },
+    {
+      "epoch": 0.12646534896372227,
+      "grad_norm": 0.328125,
+      "learning_rate": 9.305527638190956e-05,
+      "loss": 1.4297,
+      "step": 1482
+    },
+    {
+      "epoch": 0.12655068320728755,
+      "grad_norm": 0.345703125,
+      "learning_rate": 9.30502512562814e-05,
+      "loss": 1.5409,
+      "step": 1483
+    },
+    {
+      "epoch": 0.12663601745085282,
+      "grad_norm": 0.314453125,
+      "learning_rate": 9.304522613065328e-05,
+      "loss": 1.2521,
+      "step": 1484
+    },
+    {
+      "epoch": 0.12672135169441806,
+      "grad_norm": 0.30078125,
+      "learning_rate": 9.304020100502513e-05,
+      "loss": 1.1897,
+      "step": 1485
+    },
+    {
+      "epoch": 0.12680668593798333,
+      "grad_norm": 0.310546875,
+      "learning_rate": 9.303517587939699e-05,
+      "loss": 1.2765,
+      "step": 1486
+    },
+    {
+      "epoch": 0.1268920201815486,
+      "grad_norm": 0.291015625,
+      "learning_rate": 9.303015075376885e-05,
+      "loss": 1.1291,
+      "step": 1487
+    },
+    {
+      "epoch": 0.12697735442511388,
+      "grad_norm": 0.341796875,
+      "learning_rate": 9.302512562814071e-05,
+      "loss": 1.373,
+      "step": 1488
+    },
+    {
+      "epoch": 0.12706268866867912,
+      "grad_norm": 0.357421875,
+      "learning_rate": 9.302010050251256e-05,
+      "loss": 1.1695,
+      "step": 1489
+    },
+    {
+      "epoch": 0.1271480229122444,
+      "grad_norm": 0.3671875,
+      "learning_rate": 9.301507537688442e-05,
+      "loss": 1.2835,
+      "step": 1490
+    },
+    {
+      "epoch": 0.12723335715580966,
+      "grad_norm": 0.3359375,
+      "learning_rate": 9.301005025125629e-05,
+      "loss": 1.351,
+      "step": 1491
+    },
+    {
+      "epoch": 0.12731869139937493,
+      "grad_norm": 0.32421875,
+      "learning_rate": 9.300502512562815e-05,
+      "loss": 1.3164,
+      "step": 1492
+    },
+    {
+      "epoch": 0.1274040256429402,
+      "grad_norm": 0.314453125,
+      "learning_rate": 9.300000000000001e-05,
+      "loss": 1.3673,
+      "step": 1493
+    },
+    {
+      "epoch": 0.12748935988650545,
+      "grad_norm": 0.337890625,
+      "learning_rate": 9.299497487437186e-05,
+      "loss": 1.449,
+      "step": 1494
+    },
+    {
+      "epoch": 0.12757469413007072,
+      "grad_norm": 0.35546875,
+      "learning_rate": 9.298994974874372e-05,
+      "loss": 1.2609,
+      "step": 1495
+    },
+    {
+      "epoch": 0.127660028373636,
+      "grad_norm": 0.302734375,
+      "learning_rate": 9.298492462311558e-05,
+      "loss": 1.3044,
+      "step": 1496
+    },
+    {
+      "epoch": 0.12774536261720126,
+      "grad_norm": 0.310546875,
+      "learning_rate": 9.297989949748745e-05,
+      "loss": 1.2145,
+      "step": 1497
+    },
+    {
+      "epoch": 0.1278306968607665,
+      "grad_norm": 0.28125,
+      "learning_rate": 9.29748743718593e-05,
+      "loss": 1.3209,
+      "step": 1498
+    },
+    {
+      "epoch": 0.12791603110433178,
+      "grad_norm": 0.3046875,
+      "learning_rate": 9.296984924623116e-05,
+      "loss": 1.2787,
+      "step": 1499
+    },
+    {
+      "epoch": 0.12800136534789705,
+      "grad_norm": 0.283203125,
+      "learning_rate": 9.296482412060302e-05,
+      "loss": 1.15,
+      "step": 1500
+    }
+  ],
+  "logging_steps": 1,
+  "max_steps": 20000,
+  "num_input_tokens_seen": 0,
+  "num_train_epochs": 2,
+  "save_steps": 500,
+  "stateful_callbacks": {
+    "TrainerControl": {
+      "args": {
+        "should_epoch_stop": false,
+        "should_evaluate": false,
+        "should_log": false,
+        "should_save": true,
+        "should_training_stop": false
+      },
+      "attributes": {}
+    }
+  },
+  "total_flos": 3.87471749677056e+17,
+  "train_batch_size": 2,
+  "trial_name": null,
+  "trial_params": null
+}