diff --git "a/trainer_state.json" "b/trainer_state.json"
new file mode 100644--- /dev/null
+++ "b/trainer_state.json"
@@ -0,0 +1,22729 @@
+{
+  "best_metric": null,
+  "best_model_checkpoint": null,
+  "epoch": 0.9998457504241863,
+  "eval_steps": 500,
+  "global_step": 3241,
+  "is_hyper_param_search": false,
+  "is_local_process_zero": true,
+  "is_world_process_zero": true,
+  "log_history": [
+    {
+      "epoch": 0.00030849915162733303,
+      "grad_norm": 0.0,
+      "learning_rate": 2.0408163265306121e-07,
+      "loss": 2.1748,
+      "step": 1
+    },
+    {
+      "epoch": 0.0006169983032546661,
+      "grad_norm": 0.0,
+      "learning_rate": 4.0816326530612243e-07,
+      "loss": 2.3846,
+      "step": 2
+    },
+    {
+      "epoch": 0.0009254974548819991,
+      "grad_norm": 0.0,
+      "learning_rate": 6.122448979591837e-07,
+      "loss": 2.3896,
+      "step": 3
+    },
+    {
+      "epoch": 0.0012339966065093321,
+      "grad_norm": 0.0,
+      "learning_rate": 8.163265306122449e-07,
+      "loss": 2.2741,
+      "step": 4
+    },
+    {
+      "epoch": 0.001542495758136665,
+      "grad_norm": 0.0,
+      "learning_rate": 1.0204081632653063e-06,
+      "loss": 2.4158,
+      "step": 5
+    },
+    {
+      "epoch": 0.0018509949097639982,
+      "grad_norm": 0.0,
+      "learning_rate": 1.2244897959183673e-06,
+      "loss": 1.4049,
+      "step": 6
+    },
+    {
+      "epoch": 0.0021594940613913313,
+      "grad_norm": 0.0,
+      "learning_rate": 1.4285714285714286e-06,
+      "loss": 2.1787,
+      "step": 7
+    },
+    {
+      "epoch": 0.0024679932130186643,
+      "grad_norm": 0.0,
+      "learning_rate": 1.6326530612244897e-06,
+      "loss": 2.5618,
+      "step": 8
+    },
+    {
+      "epoch": 0.002776492364645997,
+      "grad_norm": 0.0,
+      "learning_rate": 1.8367346938775512e-06,
+      "loss": 2.3799,
+      "step": 9
+    },
+    {
+      "epoch": 0.00308499151627333,
+      "grad_norm": 0.0,
+      "learning_rate": 2.0408163265306125e-06,
+      "loss": 2.1604,
+      "step": 10
+    },
+    {
+      "epoch": 0.0033934906679006635,
+      "grad_norm": 0.0,
+      "learning_rate": 2.244897959183674e-06,
+      "loss": 1.9946,
+      "step": 11
+    },
+    {
+      "epoch": 0.0037019898195279964,
+      "grad_norm": 0.0,
+      "learning_rate": 2.4489795918367347e-06,
+      "loss": 1.996,
+      "step": 12
+    },
+    {
+      "epoch": 0.004010488971155329,
+      "grad_norm": 0.0,
+      "learning_rate": 2.6530612244897964e-06,
+      "loss": 2.084,
+      "step": 13
+    },
+    {
+      "epoch": 0.004318988122782663,
+      "grad_norm": 0.0,
+      "learning_rate": 2.8571428571428573e-06,
+      "loss": 1.7759,
+      "step": 14
+    },
+    {
+      "epoch": 0.004627487274409995,
+      "grad_norm": 0.0,
+      "learning_rate": 3.0612244897959185e-06,
+      "loss": 1.6765,
+      "step": 15
+    },
+    {
+      "epoch": 0.0049359864260373285,
+      "grad_norm": 0.0,
+      "learning_rate": 3.2653061224489794e-06,
+      "loss": 1.6134,
+      "step": 16
+    },
+    {
+      "epoch": 0.005244485577664661,
+      "grad_norm": 0.0,
+      "learning_rate": 3.469387755102041e-06,
+      "loss": 1.6945,
+      "step": 17
+    },
+    {
+      "epoch": 0.005552984729291994,
+      "grad_norm": 0.0,
+      "learning_rate": 3.6734693877551024e-06,
+      "loss": 1.367,
+      "step": 18
+    },
+    {
+      "epoch": 0.005861483880919328,
+      "grad_norm": 0.0,
+      "learning_rate": 3.877551020408164e-06,
+      "loss": 1.5206,
+      "step": 19
+    },
+    {
+      "epoch": 0.00616998303254666,
+      "grad_norm": 0.0,
+      "learning_rate": 4.081632653061225e-06,
+      "loss": 1.4788,
+      "step": 20
+    },
+    {
+      "epoch": 0.006478482184173994,
+      "grad_norm": 0.0,
+      "learning_rate": 4.2857142857142855e-06,
+      "loss": 1.4541,
+      "step": 21
+    },
+    {
+      "epoch": 0.006786981335801327,
+      "grad_norm": 0.0,
+      "learning_rate": 4.489795918367348e-06,
+      "loss": 1.3626,
+      "step": 22
+    },
+    {
+      "epoch": 0.007095480487428659,
+      "grad_norm": 0.0,
+      "learning_rate": 4.693877551020409e-06,
+      "loss": 1.3584,
+      "step": 23
+    },
+    {
+      "epoch": 0.007403979639055993,
+      "grad_norm": 0.0,
+      "learning_rate": 4.897959183673469e-06,
+      "loss": 1.3817,
+      "step": 24
+    },
+    {
+      "epoch": 0.007712478790683325,
+      "grad_norm": 0.0,
+      "learning_rate": 5.1020408163265315e-06,
+      "loss": 1.4,
+      "step": 25
+    },
+    {
+      "epoch": 0.008020977942310659,
+      "grad_norm": 0.0,
+      "learning_rate": 5.306122448979593e-06,
+      "loss": 1.2967,
+      "step": 26
+    },
+    {
+      "epoch": 0.008329477093937991,
+      "grad_norm": 0.0,
+      "learning_rate": 5.510204081632653e-06,
+      "loss": 1.3896,
+      "step": 27
+    },
+    {
+      "epoch": 0.008637976245565325,
+      "grad_norm": 0.0,
+      "learning_rate": 5.7142857142857145e-06,
+      "loss": 1.3068,
+      "step": 28
+    },
+    {
+      "epoch": 0.008946475397192658,
+      "grad_norm": 0.0,
+      "learning_rate": 5.918367346938776e-06,
+      "loss": 1.2201,
+      "step": 29
+    },
+    {
+      "epoch": 0.00925497454881999,
+      "grad_norm": 0.0,
+      "learning_rate": 6.122448979591837e-06,
+      "loss": 1.262,
+      "step": 30
+    },
+    {
+      "epoch": 0.009563473700447325,
+      "grad_norm": 0.0,
+      "learning_rate": 6.326530612244899e-06,
+      "loss": 1.3774,
+      "step": 31
+    },
+    {
+      "epoch": 0.009871972852074657,
+      "grad_norm": 0.0,
+      "learning_rate": 6.530612244897959e-06,
+      "loss": 1.2003,
+      "step": 32
+    },
+    {
+      "epoch": 0.01018047200370199,
+      "grad_norm": 0.0,
+      "learning_rate": 6.734693877551021e-06,
+      "loss": 1.2146,
+      "step": 33
+    },
+    {
+      "epoch": 0.010488971155329322,
+      "grad_norm": 0.0,
+      "learning_rate": 6.938775510204082e-06,
+      "loss": 1.1563,
+      "step": 34
+    },
+    {
+      "epoch": 0.010797470306956656,
+      "grad_norm": 0.0,
+      "learning_rate": 7.1428571428571436e-06,
+      "loss": 1.2303,
+      "step": 35
+    },
+    {
+      "epoch": 0.011105969458583989,
+      "grad_norm": 0.0,
+      "learning_rate": 7.346938775510205e-06,
+      "loss": 1.0958,
+      "step": 36
+    },
+    {
+      "epoch": 0.011414468610211321,
+      "grad_norm": 0.0,
+      "learning_rate": 7.551020408163265e-06,
+      "loss": 1.1453,
+      "step": 37
+    },
+    {
+      "epoch": 0.011722967761838655,
+      "grad_norm": 0.0,
+      "learning_rate": 7.755102040816327e-06,
+      "loss": 1.1907,
+      "step": 38
+    },
+    {
+      "epoch": 0.012031466913465988,
+      "grad_norm": 0.0,
+      "learning_rate": 7.959183673469388e-06,
+      "loss": 1.181,
+      "step": 39
+    },
+    {
+      "epoch": 0.01233996606509332,
+      "grad_norm": 0.0,
+      "learning_rate": 8.16326530612245e-06,
+      "loss": 1.147,
+      "step": 40
+    },
+    {
+      "epoch": 0.012648465216720655,
+      "grad_norm": 0.0,
+      "learning_rate": 8.36734693877551e-06,
+      "loss": 1.1413,
+      "step": 41
+    },
+    {
+      "epoch": 0.012956964368347987,
+      "grad_norm": 0.0,
+      "learning_rate": 8.571428571428571e-06,
+      "loss": 1.1279,
+      "step": 42
+    },
+    {
+      "epoch": 0.01326546351997532,
+      "grad_norm": 0.0,
+      "learning_rate": 8.775510204081633e-06,
+      "loss": 1.1662,
+      "step": 43
+    },
+    {
+      "epoch": 0.013573962671602654,
+      "grad_norm": 0.0,
+      "learning_rate": 8.979591836734695e-06,
+      "loss": 1.2089,
+      "step": 44
+    },
+    {
+      "epoch": 0.013882461823229986,
+      "grad_norm": 0.0,
+      "learning_rate": 9.183673469387756e-06,
+      "loss": 1.1917,
+      "step": 45
+    },
+    {
+      "epoch": 0.014190960974857319,
+      "grad_norm": 0.0,
+      "learning_rate": 9.387755102040818e-06,
+      "loss": 1.1331,
+      "step": 46
+    },
+    {
+      "epoch": 0.014499460126484651,
+      "grad_norm": 0.0,
+      "learning_rate": 9.591836734693878e-06,
+      "loss": 1.0763,
+      "step": 47
+    },
+    {
+      "epoch": 0.014807959278111986,
+      "grad_norm": 0.0,
+      "learning_rate": 9.795918367346939e-06,
+      "loss": 1.1443,
+      "step": 48
+    },
+    {
+      "epoch": 0.015116458429739318,
+      "grad_norm": 0.0,
+      "learning_rate": 1e-05,
+      "loss": 1.1878,
+      "step": 49
+    },
+    {
+      "epoch": 0.01542495758136665,
+      "grad_norm": 0.0,
+      "learning_rate": 1.0204081632653063e-05,
+      "loss": 0.9932,
+      "step": 50
+    },
+    {
+      "epoch": 0.015733456732993985,
+      "grad_norm": 0.0,
+      "learning_rate": 1.0408163265306123e-05,
+      "loss": 1.1024,
+      "step": 51
+    },
+    {
+      "epoch": 0.016041955884621317,
+      "grad_norm": 0.0,
+      "learning_rate": 1.0612244897959186e-05,
+      "loss": 1.132,
+      "step": 52
+    },
+    {
+      "epoch": 0.01635045503624865,
+      "grad_norm": 0.0,
+      "learning_rate": 1.0816326530612246e-05,
+      "loss": 1.0862,
+      "step": 53
+    },
+    {
+      "epoch": 0.016658954187875982,
+      "grad_norm": 0.0,
+      "learning_rate": 1.1020408163265306e-05,
+      "loss": 1.0846,
+      "step": 54
+    },
+    {
+      "epoch": 0.016967453339503315,
+      "grad_norm": 0.0,
+      "learning_rate": 1.1224489795918367e-05,
+      "loss": 1.1018,
+      "step": 55
+    },
+    {
+      "epoch": 0.01727595249113065,
+      "grad_norm": 0.0,
+      "learning_rate": 1.1428571428571429e-05,
+      "loss": 1.06,
+      "step": 56
+    },
+    {
+      "epoch": 0.017584451642757983,
+      "grad_norm": 0.0,
+      "learning_rate": 1.1632653061224491e-05,
+      "loss": 1.0426,
+      "step": 57
+    },
+    {
+      "epoch": 0.017892950794385316,
+      "grad_norm": 0.0,
+      "learning_rate": 1.1836734693877552e-05,
+      "loss": 1.1386,
+      "step": 58
+    },
+    {
+      "epoch": 0.018201449946012648,
+      "grad_norm": 0.0,
+      "learning_rate": 1.2040816326530614e-05,
+      "loss": 1.0103,
+      "step": 59
+    },
+    {
+      "epoch": 0.01850994909763998,
+      "grad_norm": 0.0,
+      "learning_rate": 1.2244897959183674e-05,
+      "loss": 1.0551,
+      "step": 60
+    },
+    {
+      "epoch": 0.018818448249267313,
+      "grad_norm": 0.0,
+      "learning_rate": 1.2448979591836736e-05,
+      "loss": 0.9686,
+      "step": 61
+    },
+    {
+      "epoch": 0.01912694740089465,
+      "grad_norm": 0.0,
+      "learning_rate": 1.2653061224489798e-05,
+      "loss": 1.0008,
+      "step": 62
+    },
+    {
+      "epoch": 0.01943544655252198,
+      "grad_norm": 0.0,
+      "learning_rate": 1.2857142857142859e-05,
+      "loss": 1.0213,
+      "step": 63
+    },
+    {
+      "epoch": 0.019743945704149314,
+      "grad_norm": 0.0,
+      "learning_rate": 1.3061224489795918e-05,
+      "loss": 1.0082,
+      "step": 64
+    },
+    {
+      "epoch": 0.020052444855776647,
+      "grad_norm": 0.0,
+      "learning_rate": 1.326530612244898e-05,
+      "loss": 1.0559,
+      "step": 65
+    },
+    {
+      "epoch": 0.02036094400740398,
+      "grad_norm": 0.0,
+      "learning_rate": 1.3469387755102042e-05,
+      "loss": 1.0025,
+      "step": 66
+    },
+    {
+      "epoch": 0.02066944315903131,
+      "grad_norm": 0.0,
+      "learning_rate": 1.3673469387755102e-05,
+      "loss": 0.8757,
+      "step": 67
+    },
+    {
+      "epoch": 0.020977942310658644,
+      "grad_norm": 0.0,
+      "learning_rate": 1.3877551020408165e-05,
+      "loss": 1.0191,
+      "step": 68
+    },
+    {
+      "epoch": 0.02128644146228598,
+      "grad_norm": 0.0,
+      "learning_rate": 1.4081632653061225e-05,
+      "loss": 1.0413,
+      "step": 69
+    },
+    {
+      "epoch": 0.021594940613913313,
+      "grad_norm": 0.0,
+      "learning_rate": 1.4285714285714287e-05,
+      "loss": 1.002,
+      "step": 70
+    },
+    {
+      "epoch": 0.021903439765540645,
+      "grad_norm": 0.0,
+      "learning_rate": 1.448979591836735e-05,
+      "loss": 1.0391,
+      "step": 71
+    },
+    {
+      "epoch": 0.022211938917167977,
+      "grad_norm": 0.0,
+      "learning_rate": 1.469387755102041e-05,
+      "loss": 1.0641,
+      "step": 72
+    },
+    {
+      "epoch": 0.02252043806879531,
+      "grad_norm": 0.0,
+      "learning_rate": 1.4897959183673472e-05,
+      "loss": 1.0567,
+      "step": 73
+    },
+    {
+      "epoch": 0.022828937220422642,
+      "grad_norm": 0.0,
+      "learning_rate": 1.510204081632653e-05,
+      "loss": 1.0336,
+      "step": 74
+    },
+    {
+      "epoch": 0.02313743637204998,
+      "grad_norm": 0.0,
+      "learning_rate": 1.530612244897959e-05,
+      "loss": 1.0058,
+      "step": 75
+    },
+    {
+      "epoch": 0.02344593552367731,
+      "grad_norm": 0.0,
+      "learning_rate": 1.5510204081632655e-05,
+      "loss": 1.119,
+      "step": 76
+    },
+    {
+      "epoch": 0.023754434675304643,
+      "grad_norm": 0.0,
+      "learning_rate": 1.5714285714285715e-05,
+      "loss": 1.0142,
+      "step": 77
+    },
+    {
+      "epoch": 0.024062933826931976,
+      "grad_norm": 0.0,
+      "learning_rate": 1.5918367346938776e-05,
+      "loss": 1.0398,
+      "step": 78
+    },
+    {
+      "epoch": 0.02437143297855931,
+      "grad_norm": 0.0,
+      "learning_rate": 1.612244897959184e-05,
+      "loss": 0.9653,
+      "step": 79
+    },
+    {
+      "epoch": 0.02467993213018664,
+      "grad_norm": 0.0,
+      "learning_rate": 1.63265306122449e-05,
+      "loss": 1.2841,
+      "step": 80
+    },
+    {
+      "epoch": 0.024988431281813973,
+      "grad_norm": 0.0,
+      "learning_rate": 1.653061224489796e-05,
+      "loss": 1.02,
+      "step": 81
+    },
+    {
+      "epoch": 0.02529693043344131,
+      "grad_norm": 0.0,
+      "learning_rate": 1.673469387755102e-05,
+      "loss": 0.9886,
+      "step": 82
+    },
+    {
+      "epoch": 0.025605429585068642,
+      "grad_norm": 0.0,
+      "learning_rate": 1.6938775510204085e-05,
+      "loss": 0.9234,
+      "step": 83
+    },
+    {
+      "epoch": 0.025913928736695974,
+      "grad_norm": 0.0,
+      "learning_rate": 1.7142857142857142e-05,
+      "loss": 0.9129,
+      "step": 84
+    },
+    {
+      "epoch": 0.026222427888323307,
+      "grad_norm": 0.0,
+      "learning_rate": 1.7346938775510206e-05,
+      "loss": 0.9701,
+      "step": 85
+    },
+    {
+      "epoch": 0.02653092703995064,
+      "grad_norm": 0.0,
+      "learning_rate": 1.7551020408163266e-05,
+      "loss": 1.0317,
+      "step": 86
+    },
+    {
+      "epoch": 0.026839426191577972,
+      "grad_norm": 0.0,
+      "learning_rate": 1.7755102040816327e-05,
+      "loss": 0.9648,
+      "step": 87
+    },
+    {
+      "epoch": 0.027147925343205308,
+      "grad_norm": 0.0,
+      "learning_rate": 1.795918367346939e-05,
+      "loss": 1.2338,
+      "step": 88
+    },
+    {
+      "epoch": 0.02745642449483264,
+      "grad_norm": 0.0,
+      "learning_rate": 1.816326530612245e-05,
+      "loss": 0.9986,
+      "step": 89
+    },
+    {
+      "epoch": 0.027764923646459973,
+      "grad_norm": 0.0,
+      "learning_rate": 1.836734693877551e-05,
+      "loss": 1.0245,
+      "step": 90
+    },
+    {
+      "epoch": 0.028073422798087305,
+      "grad_norm": 0.0,
+      "learning_rate": 1.8571428571428575e-05,
+      "loss": 0.9307,
+      "step": 91
+    },
+    {
+      "epoch": 0.028381921949714638,
+      "grad_norm": 0.0,
+      "learning_rate": 1.8775510204081636e-05,
+      "loss": 1.0043,
+      "step": 92
+    },
+    {
+      "epoch": 0.02869042110134197,
+      "grad_norm": 0.0,
+      "learning_rate": 1.8979591836734696e-05,
+      "loss": 1.0452,
+      "step": 93
+    },
+    {
+      "epoch": 0.028998920252969303,
+      "grad_norm": 0.0,
+      "learning_rate": 1.9183673469387756e-05,
+      "loss": 0.9908,
+      "step": 94
+    },
+    {
+      "epoch": 0.02930741940459664,
+      "grad_norm": 0.0,
+      "learning_rate": 1.9387755102040817e-05,
+      "loss": 0.933,
+      "step": 95
+    },
+    {
+      "epoch": 0.02961591855622397,
+      "grad_norm": 0.0,
+      "learning_rate": 1.9591836734693877e-05,
+      "loss": 0.9802,
+      "step": 96
+    },
+    {
+      "epoch": 0.029924417707851304,
+      "grad_norm": 0.0,
+      "learning_rate": 1.979591836734694e-05,
+      "loss": 1.0628,
+      "step": 97
+    },
+    {
+      "epoch": 0.030232916859478636,
+      "grad_norm": 0.0,
+      "learning_rate": 2e-05,
+      "loss": 1.0759,
+      "step": 98
+    },
+    {
+      "epoch": 0.03054141601110597,
+      "grad_norm": 0.0,
+      "learning_rate": 1.999999500447713e-05,
+      "loss": 1.026,
+      "step": 99
+    },
+    {
+      "epoch": 0.0308499151627333,
+      "grad_norm": 0.0,
+      "learning_rate": 1.9999980017913514e-05,
+      "loss": 0.9354,
+      "step": 100
+    },
+    {
+      "epoch": 0.031158414314360637,
+      "grad_norm": 0.0,
+      "learning_rate": 1.999995504032412e-05,
+      "loss": 1.0366,
+      "step": 101
+    },
+    {
+      "epoch": 0.03146691346598797,
+      "grad_norm": 0.0,
+      "learning_rate": 1.999992007173391e-05,
+      "loss": 0.9773,
+      "step": 102
+    },
+    {
+      "epoch": 0.0317754126176153,
+      "grad_norm": 0.0,
+      "learning_rate": 1.999987511217781e-05,
+      "loss": 0.9804,
+      "step": 103
+    },
+    {
+      "epoch": 0.032083911769242635,
+      "grad_norm": 0.0,
+      "learning_rate": 1.999982016170075e-05,
+      "loss": 0.9758,
+      "step": 104
+    },
+    {
+      "epoch": 0.03239241092086997,
+      "grad_norm": 0.0,
+      "learning_rate": 1.999975522035763e-05,
+      "loss": 0.9966,
+      "step": 105
+    },
+    {
+      "epoch": 0.0327009100724973,
+      "grad_norm": 0.0,
+      "learning_rate": 1.999968028821333e-05,
+      "loss": 0.9426,
+      "step": 106
+    },
+    {
+      "epoch": 0.03300940922412463,
+      "grad_norm": 0.0,
+      "learning_rate": 1.9999595365342715e-05,
+      "loss": 1.0034,
+      "step": 107
+    },
+    {
+      "epoch": 0.033317908375751965,
+      "grad_norm": 0.0,
+      "learning_rate": 1.9999500451830634e-05,
+      "loss": 0.9326,
+      "step": 108
+    },
+    {
+      "epoch": 0.0336264075273793,
+      "grad_norm": 0.0,
+      "learning_rate": 1.9999395547771914e-05,
+      "loss": 1.0403,
+      "step": 109
+    },
+    {
+      "epoch": 0.03393490667900663,
+      "grad_norm": 0.0,
+      "learning_rate": 1.9999280653271364e-05,
+      "loss": 1.0244,
+      "step": 110
+    },
+    {
+      "epoch": 0.03424340583063397,
+      "grad_norm": 0.0,
+      "learning_rate": 1.9999155768443777e-05,
+      "loss": 0.9906,
+      "step": 111
+    },
+    {
+      "epoch": 0.0345519049822613,
+      "grad_norm": 0.0,
+      "learning_rate": 1.999902089341393e-05,
+      "loss": 0.9533,
+      "step": 112
+    },
+    {
+      "epoch": 0.034860404133888634,
+      "grad_norm": 0.0,
+      "learning_rate": 1.999887602831657e-05,
+      "loss": 1.0805,
+      "step": 113
+    },
+    {
+      "epoch": 0.035168903285515966,
+      "grad_norm": 0.0,
+      "learning_rate": 1.999872117329644e-05,
+      "loss": 1.3153,
+      "step": 114
+    },
+    {
+      "epoch": 0.0354774024371433,
+      "grad_norm": 0.0,
+      "learning_rate": 1.999855632850825e-05,
+      "loss": 0.9861,
+      "step": 115
+    },
+    {
+      "epoch": 0.03578590158877063,
+      "grad_norm": 0.0,
+      "learning_rate": 1.9998381494116693e-05,
+      "loss": 1.0677,
+      "step": 116
+    },
+    {
+      "epoch": 0.036094400740397964,
+      "grad_norm": 0.0,
+      "learning_rate": 1.999819667029646e-05,
+      "loss": 0.9798,
+      "step": 117
+    },
+    {
+      "epoch": 0.036402899892025296,
+      "grad_norm": 0.0,
+      "learning_rate": 1.99980018572322e-05,
+      "loss": 0.9872,
+      "step": 118
+    },
+    {
+      "epoch": 0.03671139904365263,
+      "grad_norm": 0.0,
+      "learning_rate": 1.999779705511856e-05,
+      "loss": 0.9897,
+      "step": 119
+    },
+    {
+      "epoch": 0.03701989819527996,
+      "grad_norm": 0.0,
+      "learning_rate": 1.9997582264160147e-05,
+      "loss": 1.0007,
+      "step": 120
+    },
+    {
+      "epoch": 0.037328397346907294,
+      "grad_norm": 0.0,
+      "learning_rate": 1.9997357484571566e-05,
+      "loss": 0.9305,
+      "step": 121
+    },
+    {
+      "epoch": 0.037636896498534626,
+      "grad_norm": 0.0,
+      "learning_rate": 1.9997122716577397e-05,
+      "loss": 0.9443,
+      "step": 122
+    },
+    {
+      "epoch": 0.03794539565016196,
+      "grad_norm": 0.0,
+      "learning_rate": 1.9996877960412192e-05,
+      "loss": 0.9924,
+      "step": 123
+    },
+    {
+      "epoch": 0.0382538948017893,
+      "grad_norm": 0.0,
+      "learning_rate": 1.9996623216320496e-05,
+      "loss": 1.0336,
+      "step": 124
+    },
+    {
+      "epoch": 0.03856239395341663,
+      "grad_norm": 0.0,
+      "learning_rate": 1.9996358484556818e-05,
+      "loss": 0.9855,
+      "step": 125
+    },
+    {
+      "epoch": 0.03887089310504396,
+      "grad_norm": 0.0,
+      "learning_rate": 1.9996083765385656e-05,
+      "loss": 0.9731,
+      "step": 126
+    },
+    {
+      "epoch": 0.039179392256671296,
+      "grad_norm": 0.0,
+      "learning_rate": 1.999579905908148e-05,
+      "loss": 1.0336,
+      "step": 127
+    },
+    {
+      "epoch": 0.03948789140829863,
+      "grad_norm": 0.0,
+      "learning_rate": 1.9995504365928746e-05,
+      "loss": 0.9698,
+      "step": 128
+    },
+    {
+      "epoch": 0.03979639055992596,
+      "grad_norm": 0.0,
+      "learning_rate": 1.999519968622188e-05,
+      "loss": 1.0549,
+      "step": 129
+    },
+    {
+      "epoch": 0.04010488971155329,
+      "grad_norm": 0.0,
+      "learning_rate": 1.9994885020265293e-05,
+      "loss": 0.906,
+      "step": 130
+    },
+    {
+      "epoch": 0.040413388863180626,
+      "grad_norm": 0.0,
+      "learning_rate": 1.9994560368373366e-05,
+      "loss": 1.0536,
+      "step": 131
+    },
+    {
+      "epoch": 0.04072188801480796,
+      "grad_norm": 0.0,
+      "learning_rate": 1.999422573087046e-05,
+      "loss": 0.9394,
+      "step": 132
+    },
+    {
+      "epoch": 0.04103038716643529,
+      "grad_norm": 0.0,
+      "learning_rate": 1.999388110809091e-05,
+      "loss": 1.0359,
+      "step": 133
+    },
+    {
+      "epoch": 0.04133888631806262,
+      "grad_norm": 0.0,
+      "learning_rate": 1.9993526500379037e-05,
+      "loss": 1.0034,
+      "step": 134
+    },
+    {
+      "epoch": 0.041647385469689956,
+      "grad_norm": 0.0,
+      "learning_rate": 1.9993161908089125e-05,
+      "loss": 0.9976,
+      "step": 135
+    },
+    {
+      "epoch": 0.04195588462131729,
+      "grad_norm": 0.0,
+      "learning_rate": 1.9992787331585447e-05,
+      "loss": 0.9835,
+      "step": 136
+    },
+    {
+      "epoch": 0.04226438377294463,
+      "grad_norm": 0.0,
+      "learning_rate": 1.9992402771242236e-05,
+      "loss": 0.9441,
+      "step": 137
+    },
+    {
+      "epoch": 0.04257288292457196,
+      "grad_norm": 0.0,
+      "learning_rate": 1.999200822744371e-05,
+      "loss": 0.873,
+      "step": 138
+    },
+    {
+      "epoch": 0.04288138207619929,
+      "grad_norm": 0.0,
+      "learning_rate": 1.9991603700584063e-05,
+      "loss": 1.0707,
+      "step": 139
+    },
+    {
+      "epoch": 0.043189881227826625,
+      "grad_norm": 0.0,
+      "learning_rate": 1.9991189191067457e-05,
+      "loss": 0.9523,
+      "step": 140
+    },
+    {
+      "epoch": 0.04349838037945396,
+      "grad_norm": 0.0,
+      "learning_rate": 1.9990764699308033e-05,
+      "loss": 0.9397,
+      "step": 141
+    },
+    {
+      "epoch": 0.04380687953108129,
+      "grad_norm": 0.0,
+      "learning_rate": 1.99903302257299e-05,
+      "loss": 0.9902,
+      "step": 142
+    },
+    {
+      "epoch": 0.04411537868270862,
+      "grad_norm": 0.0,
+      "learning_rate": 1.9989885770767143e-05,
+      "loss": 0.9969,
+      "step": 143
+    },
+    {
+      "epoch": 0.044423877834335955,
+      "grad_norm": 0.0,
+      "learning_rate": 1.9989431334863817e-05,
+      "loss": 1.0595,
+      "step": 144
+    },
+    {
+      "epoch": 0.04473237698596329,
+      "grad_norm": 0.0,
+      "learning_rate": 1.9988966918473957e-05,
+      "loss": 0.9438,
+      "step": 145
+    },
+    {
+      "epoch": 0.04504087613759062,
+      "grad_norm": 0.0,
+      "learning_rate": 1.998849252206156e-05,
+      "loss": 0.9605,
+      "step": 146
+    },
+    {
+      "epoch": 0.04534937528921795,
+      "grad_norm": 0.0,
+      "learning_rate": 1.9988008146100594e-05,
+      "loss": 1.1187,
+      "step": 147
+    },
+    {
+      "epoch": 0.045657874440845285,
+      "grad_norm": 0.0,
+      "learning_rate": 1.9987513791075007e-05,
+      "loss": 0.9843,
+      "step": 148
+    },
+    {
+      "epoch": 0.04596637359247262,
+      "grad_norm": 0.0,
+      "learning_rate": 1.998700945747871e-05,
+      "loss": 0.9245,
+      "step": 149
+    },
+    {
+      "epoch": 0.04627487274409996,
+      "grad_norm": 0.0,
+      "learning_rate": 1.998649514581558e-05,
+      "loss": 0.9899,
+      "step": 150
+    },
+    {
+      "epoch": 0.04658337189572729,
+      "grad_norm": 0.0,
+      "learning_rate": 1.9985970856599476e-05,
+      "loss": 1.0891,
+      "step": 151
+    },
+    {
+      "epoch": 0.04689187104735462,
+      "grad_norm": 0.0,
+      "learning_rate": 1.9985436590354207e-05,
+      "loss": 1.0649,
+      "step": 152
+    },
+    {
+      "epoch": 0.047200370198981954,
+      "grad_norm": 0.0,
+      "learning_rate": 1.9984892347613577e-05,
+      "loss": 1.0248,
+      "step": 153
+    },
+    {
+      "epoch": 0.04750886935060929,
+      "grad_norm": 0.0,
+      "learning_rate": 1.9984338128921327e-05,
+      "loss": 0.9197,
+      "step": 154
+    },
+    {
+      "epoch": 0.04781736850223662,
+      "grad_norm": 0.0,
+      "learning_rate": 1.9983773934831185e-05,
+      "loss": 1.0391,
+      "step": 155
+    },
+    {
+      "epoch": 0.04812586765386395,
+      "grad_norm": 0.0,
+      "learning_rate": 1.998319976590684e-05,
+      "loss": 1.0306,
+      "step": 156
+    },
+    {
+      "epoch": 0.048434366805491284,
+      "grad_norm": 0.0,
+      "learning_rate": 1.9982615622721948e-05,
+      "loss": 0.8968,
+      "step": 157
+    },
+    {
+      "epoch": 0.04874286595711862,
+      "grad_norm": 0.0,
+      "learning_rate": 1.9982021505860128e-05,
+      "loss": 0.9711,
+      "step": 158
+    },
+    {
+      "epoch": 0.04905136510874595,
+      "grad_norm": 0.0,
+      "learning_rate": 1.9981417415914957e-05,
+      "loss": 0.963,
+      "step": 159
+    },
+    {
+      "epoch": 0.04935986426037328,
+      "grad_norm": 0.0,
+      "learning_rate": 1.9980803353490002e-05,
+      "loss": 1.0086,
+      "step": 160
+    },
+    {
+      "epoch": 0.049668363412000614,
+      "grad_norm": 0.0,
+      "learning_rate": 1.9980179319198757e-05,
+      "loss": 0.9816,
+      "step": 161
+    },
+    {
+      "epoch": 0.04997686256362795,
+      "grad_norm": 0.0,
+      "learning_rate": 1.997954531366471e-05,
+      "loss": 1.0639,
+      "step": 162
+    },
+    {
+      "epoch": 0.050285361715255286,
+      "grad_norm": 0.0,
+      "learning_rate": 1.997890133752129e-05,
+      "loss": 0.9421,
+      "step": 163
+    },
+    {
+      "epoch": 0.05059386086688262,
+      "grad_norm": 0.0,
+      "learning_rate": 1.9978247391411905e-05,
+      "loss": 0.9511,
+      "step": 164
+    },
+    {
+      "epoch": 0.05090236001850995,
+      "grad_norm": 0.0,
+      "learning_rate": 1.9977583475989907e-05,
+      "loss": 0.9865,
+      "step": 165
+    },
+    {
+      "epoch": 0.051210859170137284,
+      "grad_norm": 0.0,
+      "learning_rate": 1.9976909591918624e-05,
+      "loss": 0.9379,
+      "step": 166
+    },
+    {
+      "epoch": 0.051519358321764616,
+      "grad_norm": 0.0,
+      "learning_rate": 1.9976225739871335e-05,
+      "loss": 1.0527,
+      "step": 167
+    },
+    {
+      "epoch": 0.05182785747339195,
+      "grad_norm": 0.0,
+      "learning_rate": 1.9975531920531277e-05,
+      "loss": 0.9074,
+      "step": 168
+    },
+    {
+      "epoch": 0.05213635662501928,
+      "grad_norm": 0.0,
+      "learning_rate": 1.997482813459165e-05,
+      "loss": 0.949,
+      "step": 169
+    },
+    {
+      "epoch": 0.052444855776646614,
+      "grad_norm": 0.0,
+      "learning_rate": 1.997411438275561e-05,
+      "loss": 1.0138,
+      "step": 170
+    },
+    {
+      "epoch": 0.052753354928273946,
+      "grad_norm": 0.0,
+      "learning_rate": 1.997339066573627e-05,
+      "loss": 0.9684,
+      "step": 171
+    },
+    {
+      "epoch": 0.05306185407990128,
+      "grad_norm": 0.0,
+      "learning_rate": 1.9972656984256695e-05,
+      "loss": 0.8981,
+      "step": 172
+    },
+    {
+      "epoch": 0.05337035323152861,
+      "grad_norm": 0.0,
+      "learning_rate": 1.9971913339049914e-05,
+      "loss": 0.9672,
+      "step": 173
+    },
+    {
+      "epoch": 0.053678852383155944,
+      "grad_norm": 0.0,
+      "learning_rate": 1.9971159730858903e-05,
+      "loss": 1.0605,
+      "step": 174
+    },
+    {
+      "epoch": 0.053987351534783276,
+      "grad_norm": 0.0,
+      "learning_rate": 1.9970396160436602e-05,
+      "loss": 0.8753,
+      "step": 175
+    },
+    {
+      "epoch": 0.054295850686410616,
+      "grad_norm": 0.0,
+      "learning_rate": 1.996962262854589e-05,
+      "loss": 0.9577,
+      "step": 176
+    },
+    {
+      "epoch": 0.05460434983803795,
+      "grad_norm": 0.0,
+      "learning_rate": 1.996883913595961e-05,
+      "loss": 1.0418,
+      "step": 177
+    },
+    {
+      "epoch": 0.05491284898966528,
+      "grad_norm": 0.0,
+      "learning_rate": 1.9968045683460554e-05,
+      "loss": 0.8894,
+      "step": 178
+    },
+    {
+      "epoch": 0.05522134814129261,
+      "grad_norm": 0.0,
+      "learning_rate": 1.9967242271841458e-05,
+      "loss": 0.9815,
+      "step": 179
+    },
+    {
+      "epoch": 0.055529847292919945,
+      "grad_norm": 0.0,
+      "learning_rate": 1.9966428901905025e-05,
+      "loss": 0.9228,
+      "step": 180
+    },
+    {
+      "epoch": 0.05583834644454728,
+      "grad_norm": 0.0,
+      "learning_rate": 1.9965605574463885e-05,
+      "loss": 0.9451,
+      "step": 181
+    },
+    {
+      "epoch": 0.05614684559617461,
+      "grad_norm": 0.0,
+      "learning_rate": 1.9964772290340633e-05,
+      "loss": 1.0085,
+      "step": 182
+    },
+    {
+      "epoch": 0.05645534474780194,
+      "grad_norm": 0.0,
+      "learning_rate": 1.996392905036781e-05,
+      "loss": 0.9334,
+      "step": 183
+    },
+    {
+      "epoch": 0.056763843899429275,
+      "grad_norm": 0.0,
+      "learning_rate": 1.9963075855387898e-05,
+      "loss": 1.0261,
+      "step": 184
+    },
+    {
+      "epoch": 0.05707234305105661,
+      "grad_norm": 0.0,
+      "learning_rate": 1.9962212706253327e-05,
+      "loss": 1.0683,
+      "step": 185
+    },
+    {
+      "epoch": 0.05738084220268394,
+      "grad_norm": 0.0,
+      "learning_rate": 1.9961339603826476e-05,
+      "loss": 1.0197,
+      "step": 186
+    },
+    {
+      "epoch": 0.05768934135431127,
+      "grad_norm": 0.0,
+      "learning_rate": 1.9960456548979664e-05,
+      "loss": 0.9587,
+      "step": 187
+    },
+    {
+      "epoch": 0.057997840505938605,
+      "grad_norm": 0.0,
+      "learning_rate": 1.9959563542595157e-05,
+      "loss": 0.9745,
+      "step": 188
+    },
+    {
+      "epoch": 0.058306339657565945,
+      "grad_norm": 0.0,
+      "learning_rate": 1.995866058556516e-05,
+      "loss": 0.9802,
+      "step": 189
+    },
+    {
+      "epoch": 0.05861483880919328,
+      "grad_norm": 0.0,
+      "learning_rate": 1.995774767879182e-05,
+      "loss": 0.9877,
+      "step": 190
+    },
+    {
+      "epoch": 0.05892333796082061,
+      "grad_norm": 0.0,
+      "learning_rate": 1.995682482318723e-05,
+      "loss": 0.967,
+      "step": 191
+    },
+    {
+      "epoch": 0.05923183711244794,
+      "grad_norm": 0.0,
+      "learning_rate": 1.9955892019673412e-05,
+      "loss": 1.065,
+      "step": 192
+    },
+    {
+      "epoch": 0.059540336264075275,
+      "grad_norm": 0.0,
+      "learning_rate": 1.9954949269182343e-05,
+      "loss": 0.9254,
+      "step": 193
+    },
+    {
+      "epoch": 0.05984883541570261,
+      "grad_norm": 0.0,
+      "learning_rate": 1.9953996572655928e-05,
+      "loss": 0.947,
+      "step": 194
+    },
+    {
+      "epoch": 0.06015733456732994,
+      "grad_norm": 0.0,
+      "learning_rate": 1.9953033931046005e-05,
+      "loss": 1.0769,
+      "step": 195
+    },
+    {
+      "epoch": 0.06046583371895727,
+      "grad_norm": 0.0,
+      "learning_rate": 1.9952061345314355e-05,
+      "loss": 1.0375,
+      "step": 196
+    },
+    {
+      "epoch": 0.060774332870584605,
+      "grad_norm": 0.0,
+      "learning_rate": 1.99510788164327e-05,
+      "loss": 0.9665,
+      "step": 197
+    },
+    {
+      "epoch": 0.06108283202221194,
+      "grad_norm": 0.0,
+      "learning_rate": 1.995008634538268e-05,
+      "loss": 0.9598,
+      "step": 198
+    },
+    {
+      "epoch": 0.06139133117383927,
+      "grad_norm": 0.0,
+      "learning_rate": 1.994908393315588e-05,
+      "loss": 0.9486,
+      "step": 199
+    },
+    {
+      "epoch": 0.0616998303254666,
+      "grad_norm": 0.0,
+      "learning_rate": 1.994807158075382e-05,
+      "loss": 0.9422,
+      "step": 200
+    },
+    {
+      "epoch": 0.062008329477093935,
+      "grad_norm": 0.0,
+      "learning_rate": 1.9947049289187942e-05,
+      "loss": 0.8275,
+      "step": 201
+    },
+    {
+      "epoch": 0.062316828628721274,
+      "grad_norm": 0.0,
+      "learning_rate": 1.9946017059479617e-05,
+      "loss": 0.9641,
+      "step": 202
+    },
+    {
+      "epoch": 0.0626253277803486,
+      "grad_norm": 0.0,
+      "learning_rate": 1.9944974892660158e-05,
+      "loss": 1.0779,
+      "step": 203
+    },
+    {
+      "epoch": 0.06293382693197594,
+      "grad_norm": 0.0,
+      "learning_rate": 1.9943922789770797e-05,
+      "loss": 1.0373,
+      "step": 204
+    },
+    {
+      "epoch": 0.06324232608360326,
+      "grad_norm": 0.0,
+      "learning_rate": 1.9942860751862696e-05,
+      "loss": 1.0035,
+      "step": 205
+    },
+    {
+      "epoch": 0.0635508252352306,
+      "grad_norm": 0.0,
+      "learning_rate": 1.9941788779996937e-05,
+      "loss": 1.0022,
+      "step": 206
+    },
+    {
+      "epoch": 0.06385932438685793,
+      "grad_norm": 0.0,
+      "learning_rate": 1.994070687524453e-05,
+      "loss": 0.9576,
+      "step": 207
+    },
+    {
+      "epoch": 0.06416782353848527,
+      "grad_norm": 0.0,
+      "learning_rate": 1.9939615038686423e-05,
+      "loss": 0.9669,
+      "step": 208
+    },
+    {
+      "epoch": 0.06447632269011261,
+      "grad_norm": 0.0,
+      "learning_rate": 1.9938513271413464e-05,
+      "loss": 0.9042,
+      "step": 209
+    },
+    {
+      "epoch": 0.06478482184173993,
+      "grad_norm": 0.0,
+      "learning_rate": 1.993740157452644e-05,
+      "loss": 0.8647,
+      "step": 210
+    },
+    {
+      "epoch": 0.06509332099336727,
+      "grad_norm": 0.0,
+      "learning_rate": 1.9936279949136047e-05,
+      "loss": 0.9643,
+      "step": 211
+    },
+    {
+      "epoch": 0.0654018201449946,
+      "grad_norm": 0.0,
+      "learning_rate": 1.993514839636291e-05,
+      "loss": 0.8882,
+      "step": 212
+    },
+    {
+      "epoch": 0.06571031929662194,
+      "grad_norm": 0.0,
+      "learning_rate": 1.993400691733757e-05,
+      "loss": 1.0237,
+      "step": 213
+    },
+    {
+      "epoch": 0.06601881844824926,
+      "grad_norm": 0.0,
+      "learning_rate": 1.993285551320048e-05,
+      "loss": 0.96,
+      "step": 214
+    },
+    {
+      "epoch": 0.0663273175998766,
+      "grad_norm": 0.0,
+      "learning_rate": 1.993169418510201e-05,
+      "loss": 0.9605,
+      "step": 215
+    },
+    {
+      "epoch": 0.06663581675150393,
+      "grad_norm": 0.0,
+      "learning_rate": 1.9930522934202455e-05,
+      "loss": 0.837,
+      "step": 216
+    },
+    {
+      "epoch": 0.06694431590313127,
+      "grad_norm": 0.0,
+      "learning_rate": 1.9929341761672017e-05,
+      "loss": 0.9478,
+      "step": 217
+    },
+    {
+      "epoch": 0.0672528150547586,
+      "grad_norm": 0.0,
+      "learning_rate": 1.9928150668690807e-05,
+      "loss": 1.013,
+      "step": 218
+    },
+    {
+      "epoch": 0.06756131420638593,
+      "grad_norm": 0.0,
+      "learning_rate": 1.992694965644885e-05,
+      "loss": 0.8715,
+      "step": 219
+    },
+    {
+      "epoch": 0.06786981335801326,
+      "grad_norm": 0.0,
+      "learning_rate": 1.9925738726146094e-05,
+      "loss": 0.9242,
+      "step": 220
+    },
+    {
+      "epoch": 0.0681783125096406,
+      "grad_norm": 0.0,
+      "learning_rate": 1.992451787899237e-05,
+      "loss": 0.9094,
+      "step": 221
+    },
+    {
+      "epoch": 0.06848681166126794,
+      "grad_norm": 0.0,
+      "learning_rate": 1.9923287116207442e-05,
+      "loss": 0.9811,
+      "step": 222
+    },
+    {
+      "epoch": 0.06879531081289526,
+      "grad_norm": 0.0,
+      "learning_rate": 1.9922046439020967e-05,
+      "loss": 0.927,
+      "step": 223
+    },
+    {
+      "epoch": 0.0691038099645226,
+      "grad_norm": 0.0,
+      "learning_rate": 1.9920795848672512e-05,
+      "loss": 0.9411,
+      "step": 224
+    },
+    {
+      "epoch": 0.06941230911614993,
+      "grad_norm": 0.0,
+      "learning_rate": 1.9919535346411546e-05,
+      "loss": 0.9485,
+      "step": 225
+    },
+    {
+      "epoch": 0.06972080826777727,
+      "grad_norm": 0.0,
+      "learning_rate": 1.9918264933497444e-05,
+      "loss": 0.97,
+      "step": 226
+    },
+    {
+      "epoch": 0.0700293074194046,
+      "grad_norm": 0.0,
+      "learning_rate": 1.9916984611199485e-05,
+      "loss": 0.9,
+      "step": 227
+    },
+    {
+      "epoch": 0.07033780657103193,
+      "grad_norm": 0.0,
+      "learning_rate": 1.9915694380796838e-05,
+      "loss": 1.0275,
+      "step": 228
+    },
+    {
+      "epoch": 0.07064630572265926,
+      "grad_norm": 0.0,
+      "learning_rate": 1.9914394243578582e-05,
+      "loss": 0.8988,
+      "step": 229
+    },
+    {
+      "epoch": 0.0709548048742866,
+      "grad_norm": 0.0,
+      "learning_rate": 1.991308420084369e-05,
+      "loss": 1.0184,
+      "step": 230
+    },
+    {
+      "epoch": 0.07126330402591392,
+      "grad_norm": 0.0,
+      "learning_rate": 1.9911764253901032e-05,
+      "loss": 0.9032,
+      "step": 231
+    },
+    {
+      "epoch": 0.07157180317754126,
+      "grad_norm": 0.0,
+      "learning_rate": 1.9910434404069368e-05,
+      "loss": 0.9098,
+      "step": 232
+    },
+    {
+      "epoch": 0.07188030232916859,
+      "grad_norm": 0.0,
+      "learning_rate": 1.9909094652677364e-05,
+      "loss": 0.9501,
+      "step": 233
+    },
+    {
+      "epoch": 0.07218880148079593,
+      "grad_norm": 0.0,
+      "learning_rate": 1.990774500106357e-05,
+      "loss": 0.9101,
+      "step": 234
+    },
+    {
+      "epoch": 0.07249730063242327,
+      "grad_norm": 0.0,
+      "learning_rate": 1.9906385450576424e-05,
+      "loss": 0.9228,
+      "step": 235
+    },
+    {
+      "epoch": 0.07280579978405059,
+      "grad_norm": 0.0,
+      "learning_rate": 1.9905016002574266e-05,
+      "loss": 1.0251,
+      "step": 236
+    },
+    {
+      "epoch": 0.07311429893567793,
+      "grad_norm": 0.0,
+      "learning_rate": 1.9903636658425316e-05,
+      "loss": 0.9231,
+      "step": 237
+    },
+    {
+      "epoch": 0.07342279808730526,
+      "grad_norm": 0.0,
+      "learning_rate": 1.990224741950768e-05,
+      "loss": 0.9371,
+      "step": 238
+    },
+    {
+      "epoch": 0.0737312972389326,
+      "grad_norm": 0.0,
+      "learning_rate": 1.9900848287209356e-05,
+      "loss": 1.0318,
+      "step": 239
+    },
+    {
+      "epoch": 0.07403979639055992,
+      "grad_norm": 0.0,
+      "learning_rate": 1.989943926292822e-05,
+      "loss": 0.9119,
+      "step": 240
+    },
+    {
+      "epoch": 0.07434829554218726,
+      "grad_norm": 0.0,
+      "learning_rate": 1.989802034807204e-05,
+      "loss": 1.0122,
+      "step": 241
+    },
+    {
+      "epoch": 0.07465679469381459,
+      "grad_norm": 0.0,
+      "learning_rate": 1.9896591544058458e-05,
+      "loss": 1.006,
+      "step": 242
+    },
+    {
+      "epoch": 0.07496529384544193,
+      "grad_norm": 0.0,
+      "learning_rate": 1.9895152852314995e-05,
+      "loss": 0.9525,
+      "step": 243
+    },
+    {
+      "epoch": 0.07527379299706925,
+      "grad_norm": 0.0,
+      "learning_rate": 1.9893704274279057e-05,
+      "loss": 0.8847,
+      "step": 244
+    },
+    {
+      "epoch": 0.07558229214869659,
+      "grad_norm": 0.0,
+      "learning_rate": 1.9892245811397924e-05,
+      "loss": 1.0279,
+      "step": 245
+    },
+    {
+      "epoch": 0.07589079130032392,
+      "grad_norm": 0.0,
+      "learning_rate": 1.989077746512876e-05,
+      "loss": 0.9846,
+      "step": 246
+    },
+    {
+      "epoch": 0.07619929045195126,
+      "grad_norm": 0.0,
+      "learning_rate": 1.9889299236938585e-05,
+      "loss": 1.023,
+      "step": 247
+    },
+    {
+      "epoch": 0.0765077896035786,
+      "grad_norm": 0.0,
+      "learning_rate": 1.9887811128304312e-05,
+      "loss": 0.9629,
+      "step": 248
+    },
+    {
+      "epoch": 0.07681628875520592,
+      "grad_norm": 0.0,
+      "learning_rate": 1.9886313140712717e-05,
+      "loss": 1.0206,
+      "step": 249
+    },
+    {
+      "epoch": 0.07712478790683326,
+      "grad_norm": 0.0,
+      "learning_rate": 1.988480527566044e-05,
+      "loss": 0.9065,
+      "step": 250
+    },
+    {
+      "epoch": 0.07743328705846059,
+      "grad_norm": 0.0,
+      "learning_rate": 1.9883287534654e-05,
+      "loss": 0.951,
+      "step": 251
+    },
+    {
+      "epoch": 0.07774178621008793,
+      "grad_norm": 0.0,
+      "learning_rate": 1.988175991920978e-05,
+      "loss": 1.0458,
+      "step": 252
+    },
+    {
+      "epoch": 0.07805028536171525,
+      "grad_norm": 0.0,
+      "learning_rate": 1.9880222430854025e-05,
+      "loss": 0.9257,
+      "step": 253
+    },
+    {
+      "epoch": 0.07835878451334259,
+      "grad_norm": 0.0,
+      "learning_rate": 1.9878675071122848e-05,
+      "loss": 0.9411,
+      "step": 254
+    },
+    {
+      "epoch": 0.07866728366496992,
+      "grad_norm": 0.0,
+      "learning_rate": 1.9877117841562222e-05,
+      "loss": 1.002,
+      "step": 255
+    },
+    {
+      "epoch": 0.07897578281659726,
+      "grad_norm": 0.0,
+      "learning_rate": 1.9875550743727982e-05,
+      "loss": 0.8932,
+      "step": 256
+    },
+    {
+      "epoch": 0.07928428196822458,
+      "grad_norm": 0.0,
+      "learning_rate": 1.9873973779185828e-05,
+      "loss": 0.9698,
+      "step": 257
+    },
+    {
+      "epoch": 0.07959278111985192,
+      "grad_norm": 0.0,
+      "learning_rate": 1.9872386949511308e-05,
+      "loss": 0.8699,
+      "step": 258
+    },
+    {
+      "epoch": 0.07990128027147925,
+      "grad_norm": 0.0,
+      "learning_rate": 1.9870790256289827e-05,
+      "loss": 0.9997,
+      "step": 259
+    },
+    {
+      "epoch": 0.08020977942310659,
+      "grad_norm": 0.0,
+      "learning_rate": 1.9869183701116655e-05,
+      "loss": 0.9943,
+      "step": 260
+    },
+    {
+      "epoch": 0.08051827857473393,
+      "grad_norm": 0.0,
+      "learning_rate": 1.9867567285596903e-05,
+      "loss": 1.0419,
+      "step": 261
+    },
+    {
+      "epoch": 0.08082677772636125,
+      "grad_norm": 0.0,
+      "learning_rate": 1.9865941011345547e-05,
+      "loss": 1.3233,
+      "step": 262
+    },
+    {
+      "epoch": 0.08113527687798859,
+      "grad_norm": 0.0,
+      "learning_rate": 1.9864304879987397e-05,
+      "loss": 0.8918,
+      "step": 263
+    },
+    {
+      "epoch": 0.08144377602961592,
+      "grad_norm": 0.0,
+      "learning_rate": 1.9862658893157124e-05,
+      "loss": 0.9424,
+      "step": 264
+    },
+    {
+      "epoch": 0.08175227518124326,
+      "grad_norm": 0.0,
+      "learning_rate": 1.986100305249924e-05,
+      "loss": 0.9601,
+      "step": 265
+    },
+    {
+      "epoch": 0.08206077433287058,
+      "grad_norm": 0.0,
+      "learning_rate": 1.9859337359668102e-05,
+      "loss": 0.9477,
+      "step": 266
+    },
+    {
+      "epoch": 0.08236927348449792,
+      "grad_norm": 0.0,
+      "learning_rate": 1.9857661816327913e-05,
+      "loss": 0.9606,
+      "step": 267
+    },
+    {
+      "epoch": 0.08267777263612525,
+      "grad_norm": 0.0,
+      "learning_rate": 1.9855976424152713e-05,
+      "loss": 1.0228,
+      "step": 268
+    },
+    {
+      "epoch": 0.08298627178775259,
+      "grad_norm": 0.0,
+      "learning_rate": 1.9854281184826386e-05,
+      "loss": 0.9454,
+      "step": 269
+    },
+    {
+      "epoch": 0.08329477093937991,
+      "grad_norm": 0.0,
+      "learning_rate": 1.9852576100042656e-05,
+      "loss": 0.9233,
+      "step": 270
+    },
+    {
+      "epoch": 0.08360327009100725,
+      "grad_norm": 0.0,
+      "learning_rate": 1.985086117150508e-05,
+      "loss": 0.9667,
+      "step": 271
+    },
+    {
+      "epoch": 0.08391176924263458,
+      "grad_norm": 0.0,
+      "learning_rate": 1.9849136400927048e-05,
+      "loss": 0.9803,
+      "step": 272
+    },
+    {
+      "epoch": 0.08422026839426192,
+      "grad_norm": 0.0,
+      "learning_rate": 1.9847401790031792e-05,
+      "loss": 0.861,
+      "step": 273
+    },
+    {
+      "epoch": 0.08452876754588926,
+      "grad_norm": 0.0,
+      "learning_rate": 1.9845657340552366e-05,
+      "loss": 0.9085,
+      "step": 274
+    },
+    {
+      "epoch": 0.08483726669751658,
+      "grad_norm": 0.0,
+      "learning_rate": 1.9843903054231653e-05,
+      "loss": 0.996,
+      "step": 275
+    },
+    {
+      "epoch": 0.08514576584914392,
+      "grad_norm": 0.0,
+      "learning_rate": 1.9842138932822377e-05,
+      "loss": 0.9158,
+      "step": 276
+    },
+    {
+      "epoch": 0.08545426500077125,
+      "grad_norm": 0.0,
+      "learning_rate": 1.9840364978087073e-05,
+      "loss": 0.842,
+      "step": 277
+    },
+    {
+      "epoch": 0.08576276415239859,
+      "grad_norm": 0.0,
+      "learning_rate": 1.9838581191798117e-05,
+      "loss": 0.9036,
+      "step": 278
+    },
+    {
+      "epoch": 0.08607126330402591,
+      "grad_norm": 0.0,
+      "learning_rate": 1.9836787575737683e-05,
+      "loss": 0.8667,
+      "step": 279
+    },
+    {
+      "epoch": 0.08637976245565325,
+      "grad_norm": 0.0,
+      "learning_rate": 1.9834984131697796e-05,
+      "loss": 0.9167,
+      "step": 280
+    },
+    {
+      "epoch": 0.08668826160728058,
+      "grad_norm": 0.0,
+      "learning_rate": 1.983317086148027e-05,
+      "loss": 0.9307,
+      "step": 281
+    },
+    {
+      "epoch": 0.08699676075890792,
+      "grad_norm": 0.0,
+      "learning_rate": 1.9831347766896766e-05,
+      "loss": 1.3152,
+      "step": 282
+    },
+    {
+      "epoch": 0.08730525991053524,
+      "grad_norm": 0.0,
+      "learning_rate": 1.9829514849768737e-05,
+      "loss": 0.9292,
+      "step": 283
+    },
+    {
+      "epoch": 0.08761375906216258,
+      "grad_norm": 0.0,
+      "learning_rate": 1.9827672111927466e-05,
+      "loss": 0.9636,
+      "step": 284
+    },
+    {
+      "epoch": 0.0879222582137899,
+      "grad_norm": 0.0,
+      "learning_rate": 1.9825819555214035e-05,
+      "loss": 0.9255,
+      "step": 285
+    },
+    {
+      "epoch": 0.08823075736541725,
+      "grad_norm": 0.0,
+      "learning_rate": 1.982395718147934e-05,
+      "loss": 0.8859,
+      "step": 286
+    },
+    {
+      "epoch": 0.08853925651704458,
+      "grad_norm": 0.0,
+      "learning_rate": 1.9822084992584098e-05,
+      "loss": 0.8519,
+      "step": 287
+    },
+    {
+      "epoch": 0.08884775566867191,
+      "grad_norm": 0.0,
+      "learning_rate": 1.982020299039881e-05,
+      "loss": 0.9121,
+      "step": 288
+    },
+    {
+      "epoch": 0.08915625482029925,
+      "grad_norm": 0.0,
+      "learning_rate": 1.9818311176803796e-05,
+      "loss": 0.9241,
+      "step": 289
+    },
+    {
+      "epoch": 0.08946475397192657,
+      "grad_norm": 0.0,
+      "learning_rate": 1.9816409553689178e-05,
+      "loss": 0.9331,
+      "step": 290
+    },
+    {
+      "epoch": 0.08977325312355391,
+      "grad_norm": 0.0,
+      "learning_rate": 1.9814498122954875e-05,
+      "loss": 0.9826,
+      "step": 291
+    },
+    {
+      "epoch": 0.09008175227518124,
+      "grad_norm": 0.0,
+      "learning_rate": 1.9812576886510607e-05,
+      "loss": 0.9549,
+      "step": 292
+    },
+    {
+      "epoch": 0.09039025142680858,
+      "grad_norm": 0.0,
+      "learning_rate": 1.981064584627589e-05,
+      "loss": 0.9691,
+      "step": 293
+    },
+    {
+      "epoch": 0.0906987505784359,
+      "grad_norm": 0.0,
+      "learning_rate": 1.9808705004180032e-05,
+      "loss": 0.907,
+      "step": 294
+    },
+    {
+      "epoch": 0.09100724973006324,
+      "grad_norm": 0.0,
+      "learning_rate": 1.9806754362162143e-05,
+      "loss": 0.8657,
+      "step": 295
+    },
+    {
+      "epoch": 0.09131574888169057,
+      "grad_norm": 0.0,
+      "learning_rate": 1.9804793922171114e-05,
+      "loss": 1.0198,
+      "step": 296
+    },
+    {
+      "epoch": 0.09162424803331791,
+      "grad_norm": 0.0,
+      "learning_rate": 1.980282368616563e-05,
+      "loss": 0.8932,
+      "step": 297
+    },
+    {
+      "epoch": 0.09193274718494523,
+      "grad_norm": 0.0,
+      "learning_rate": 1.9800843656114167e-05,
+      "loss": 0.8682,
+      "step": 298
+    },
+    {
+      "epoch": 0.09224124633657257,
+      "grad_norm": 0.0,
+      "learning_rate": 1.9798853833994975e-05,
+      "loss": 0.9221,
+      "step": 299
+    },
+    {
+      "epoch": 0.09254974548819991,
+      "grad_norm": 0.0,
+      "learning_rate": 1.9796854221796097e-05,
+      "loss": 1.0219,
+      "step": 300
+    },
+    {
+      "epoch": 0.09285824463982724,
+      "grad_norm": 0.0,
+      "learning_rate": 1.979484482151536e-05,
+      "loss": 1.0476,
+      "step": 301
+    },
+    {
+      "epoch": 0.09316674379145458,
+      "grad_norm": 0.0,
+      "learning_rate": 1.9792825635160357e-05,
+      "loss": 0.8485,
+      "step": 302
+    },
+    {
+      "epoch": 0.0934752429430819,
+      "grad_norm": 0.0,
+      "learning_rate": 1.979079666474847e-05,
+      "loss": 0.9505,
+      "step": 303
+    },
+    {
+      "epoch": 0.09378374209470924,
+      "grad_norm": 0.0,
+      "learning_rate": 1.9788757912306856e-05,
+      "loss": 0.926,
+      "step": 304
+    },
+    {
+      "epoch": 0.09409224124633657,
+      "grad_norm": 0.0,
+      "learning_rate": 1.9786709379872436e-05,
+      "loss": 0.9169,
+      "step": 305
+    },
+    {
+      "epoch": 0.09440074039796391,
+      "grad_norm": 0.0,
+      "learning_rate": 1.9784651069491914e-05,
+      "loss": 0.8907,
+      "step": 306
+    },
+    {
+      "epoch": 0.09470923954959123,
+      "grad_norm": 0.0,
+      "learning_rate": 1.978258298322175e-05,
+      "loss": 0.9022,
+      "step": 307
+    },
+    {
+      "epoch": 0.09501773870121857,
+      "grad_norm": 0.0,
+      "learning_rate": 1.9780505123128187e-05,
+      "loss": 0.8888,
+      "step": 308
+    },
+    {
+      "epoch": 0.0953262378528459,
+      "grad_norm": 0.0,
+      "learning_rate": 1.9778417491287217e-05,
+      "loss": 0.967,
+      "step": 309
+    },
+    {
+      "epoch": 0.09563473700447324,
+      "grad_norm": 0.0,
+      "learning_rate": 1.9776320089784605e-05,
+      "loss": 0.9804,
+      "step": 310
+    },
+    {
+      "epoch": 0.09594323615610056,
+      "grad_norm": 0.0,
+      "learning_rate": 1.9774212920715876e-05,
+      "loss": 0.9284,
+      "step": 311
+    },
+    {
+      "epoch": 0.0962517353077279,
+      "grad_norm": 0.0,
+      "learning_rate": 1.977209598618631e-05,
+      "loss": 0.9764,
+      "step": 312
+    },
+    {
+      "epoch": 0.09656023445935524,
+      "grad_norm": 0.0,
+      "learning_rate": 1.976996928831095e-05,
+      "loss": 0.9649,
+      "step": 313
+    },
+    {
+      "epoch": 0.09686873361098257,
+      "grad_norm": 0.0,
+      "learning_rate": 1.9767832829214588e-05,
+      "loss": 0.9338,
+      "step": 314
+    },
+    {
+      "epoch": 0.09717723276260991,
+      "grad_norm": 0.0,
+      "learning_rate": 1.976568661103176e-05,
+      "loss": 0.9494,
+      "step": 315
+    },
+    {
+      "epoch": 0.09748573191423723,
+      "grad_norm": 0.0,
+      "learning_rate": 1.976353063590678e-05,
+      "loss": 0.9908,
+      "step": 316
+    },
+    {
+      "epoch": 0.09779423106586457,
+      "grad_norm": 0.0,
+      "learning_rate": 1.976136490599368e-05,
+      "loss": 0.9387,
+      "step": 317
+    },
+    {
+      "epoch": 0.0981027302174919,
+      "grad_norm": 0.0,
+      "learning_rate": 1.975918942345626e-05,
+      "loss": 0.8793,
+      "step": 318
+    },
+    {
+      "epoch": 0.09841122936911924,
+      "grad_norm": 0.0,
+      "learning_rate": 1.975700419046804e-05,
+      "loss": 0.8569,
+      "step": 319
+    },
+    {
+      "epoch": 0.09871972852074656,
+      "grad_norm": 0.0,
+      "learning_rate": 1.975480920921231e-05,
+      "loss": 1.0324,
+      "step": 320
+    },
+    {
+      "epoch": 0.0990282276723739,
+      "grad_norm": 0.0,
+      "learning_rate": 1.975260448188208e-05,
+      "loss": 0.9234,
+      "step": 321
+    },
+    {
+      "epoch": 0.09933672682400123,
+      "grad_norm": 0.0,
+      "learning_rate": 1.975039001068011e-05,
+      "loss": 0.9475,
+      "step": 322
+    },
+    {
+      "epoch": 0.09964522597562857,
+      "grad_norm": 0.0,
+      "learning_rate": 1.9748165797818875e-05,
+      "loss": 0.9887,
+      "step": 323
+    },
+    {
+      "epoch": 0.0999537251272559,
+      "grad_norm": 0.0,
+      "learning_rate": 1.974593184552061e-05,
+      "loss": 0.9338,
+      "step": 324
+    },
+    {
+      "epoch": 0.10026222427888323,
+      "grad_norm": 0.0,
+      "learning_rate": 1.9743688156017254e-05,
+      "loss": 0.9126,
+      "step": 325
+    },
+    {
+      "epoch": 0.10057072343051057,
+      "grad_norm": 0.0,
+      "learning_rate": 1.9741434731550498e-05,
+      "loss": 0.9219,
+      "step": 326
+    },
+    {
+      "epoch": 0.1008792225821379,
+      "grad_norm": 0.0,
+      "learning_rate": 1.9739171574371744e-05,
+      "loss": 0.8622,
+      "step": 327
+    },
+    {
+      "epoch": 0.10118772173376524,
+      "grad_norm": 0.0,
+      "learning_rate": 1.9736898686742125e-05,
+      "loss": 0.9795,
+      "step": 328
+    },
+    {
+      "epoch": 0.10149622088539256,
+      "grad_norm": 0.0,
+      "learning_rate": 1.973461607093249e-05,
+      "loss": 0.8908,
+      "step": 329
+    },
+    {
+      "epoch": 0.1018047200370199,
+      "grad_norm": 0.0,
+      "learning_rate": 1.9732323729223414e-05,
+      "loss": 0.9311,
+      "step": 330
+    },
+    {
+      "epoch": 0.10211321918864723,
+      "grad_norm": 0.0,
+      "learning_rate": 1.9730021663905186e-05,
+      "loss": 0.9054,
+      "step": 331
+    },
+    {
+      "epoch": 0.10242171834027457,
+      "grad_norm": 0.0,
+      "learning_rate": 1.972770987727781e-05,
+      "loss": 0.9287,
+      "step": 332
+    },
+    {
+      "epoch": 0.10273021749190189,
+      "grad_norm": 0.0,
+      "learning_rate": 1.9725388371651e-05,
+      "loss": 0.9404,
+      "step": 333
+    },
+    {
+      "epoch": 0.10303871664352923,
+      "grad_norm": 0.0,
+      "learning_rate": 1.9723057149344184e-05,
+      "loss": 0.9188,
+      "step": 334
+    },
+    {
+      "epoch": 0.10334721579515656,
+      "grad_norm": 0.0,
+      "learning_rate": 1.97207162126865e-05,
+      "loss": 1.0049,
+      "step": 335
+    },
+    {
+      "epoch": 0.1036557149467839,
+      "grad_norm": 0.0,
+      "learning_rate": 1.9718365564016785e-05,
+      "loss": 0.9919,
+      "step": 336
+    },
+    {
+      "epoch": 0.10396421409841122,
+      "grad_norm": 0.0,
+      "learning_rate": 1.971600520568358e-05,
+      "loss": 0.8696,
+      "step": 337
+    },
+    {
+      "epoch": 0.10427271325003856,
+      "grad_norm": 0.0,
+      "learning_rate": 1.9713635140045134e-05,
+      "loss": 0.9094,
+      "step": 338
+    },
+    {
+      "epoch": 0.1045812124016659,
+      "grad_norm": 0.0,
+      "learning_rate": 1.9711255369469394e-05,
+      "loss": 0.9393,
+      "step": 339
+    },
+    {
+      "epoch": 0.10488971155329323,
+      "grad_norm": 0.0,
+      "learning_rate": 1.9708865896333993e-05,
+      "loss": 0.9828,
+      "step": 340
+    },
+    {
+      "epoch": 0.10519821070492057,
+      "grad_norm": 0.0,
+      "learning_rate": 1.970646672302627e-05,
+      "loss": 0.8656,
+      "step": 341
+    },
+    {
+      "epoch": 0.10550670985654789,
+      "grad_norm": 0.0,
+      "learning_rate": 1.9704057851943244e-05,
+      "loss": 0.9252,
+      "step": 342
+    },
+    {
+      "epoch": 0.10581520900817523,
+      "grad_norm": 0.0,
+      "learning_rate": 1.9701639285491633e-05,
+      "loss": 0.8904,
+      "step": 343
+    },
+    {
+      "epoch": 0.10612370815980256,
+      "grad_norm": 0.0,
+      "learning_rate": 1.969921102608784e-05,
+      "loss": 0.906,
+      "step": 344
+    },
+    {
+      "epoch": 0.1064322073114299,
+      "grad_norm": 0.0,
+      "learning_rate": 1.9696773076157942e-05,
+      "loss": 1.0674,
+      "step": 345
+    },
+    {
+      "epoch": 0.10674070646305722,
+      "grad_norm": 0.0,
+      "learning_rate": 1.9694325438137716e-05,
+      "loss": 0.9022,
+      "step": 346
+    },
+    {
+      "epoch": 0.10704920561468456,
+      "grad_norm": 0.0,
+      "learning_rate": 1.9691868114472602e-05,
+      "loss": 0.9366,
+      "step": 347
+    },
+    {
+      "epoch": 0.10735770476631189,
+      "grad_norm": 0.0,
+      "learning_rate": 1.9689401107617722e-05,
+      "loss": 0.8698,
+      "step": 348
+    },
+    {
+      "epoch": 0.10766620391793923,
+      "grad_norm": 0.0,
+      "learning_rate": 1.9686924420037877e-05,
+      "loss": 0.9393,
+      "step": 349
+    },
+    {
+      "epoch": 0.10797470306956655,
+      "grad_norm": 0.0,
+      "learning_rate": 1.968443805420754e-05,
+      "loss": 0.9386,
+      "step": 350
+    },
+    {
+      "epoch": 0.10828320222119389,
+      "grad_norm": 0.0,
+      "learning_rate": 1.9681942012610847e-05,
+      "loss": 0.8884,
+      "step": 351
+    },
+    {
+      "epoch": 0.10859170137282123,
+      "grad_norm": 0.0,
+      "learning_rate": 1.96794362977416e-05,
+      "loss": 0.9948,
+      "step": 352
+    },
+    {
+      "epoch": 0.10890020052444856,
+      "grad_norm": 0.0,
+      "learning_rate": 1.9676920912103278e-05,
+      "loss": 0.9183,
+      "step": 353
+    },
+    {
+      "epoch": 0.1092086996760759,
+      "grad_norm": 0.0,
+      "learning_rate": 1.9674395858209014e-05,
+      "loss": 0.9598,
+      "step": 354
+    },
+    {
+      "epoch": 0.10951719882770322,
+      "grad_norm": 0.0,
+      "learning_rate": 1.9671861138581594e-05,
+      "loss": 0.9646,
+      "step": 355
+    },
+    {
+      "epoch": 0.10982569797933056,
+      "grad_norm": 0.0,
+      "learning_rate": 1.966931675575347e-05,
+      "loss": 0.9969,
+      "step": 356
+    },
+    {
+      "epoch": 0.11013419713095789,
+      "grad_norm": 0.0,
+      "learning_rate": 1.966676271226675e-05,
+      "loss": 0.9328,
+      "step": 357
+    },
+    {
+      "epoch": 0.11044269628258523,
+      "grad_norm": 0.0,
+      "learning_rate": 1.9664199010673192e-05,
+      "loss": 0.8916,
+      "step": 358
+    },
+    {
+      "epoch": 0.11075119543421255,
+      "grad_norm": 0.0,
+      "learning_rate": 1.9661625653534196e-05,
+      "loss": 0.9142,
+      "step": 359
+    },
+    {
+      "epoch": 0.11105969458583989,
+      "grad_norm": 0.0,
+      "learning_rate": 1.9659042643420817e-05,
+      "loss": 0.9221,
+      "step": 360
+    },
+    {
+      "epoch": 0.11136819373746722,
+      "grad_norm": 0.0,
+      "learning_rate": 1.9656449982913757e-05,
+      "loss": 0.9171,
+      "step": 361
+    },
+    {
+      "epoch": 0.11167669288909456,
+      "grad_norm": 0.0,
+      "learning_rate": 1.9653847674603348e-05,
+      "loss": 0.8394,
+      "step": 362
+    },
+    {
+      "epoch": 0.11198519204072188,
+      "grad_norm": 0.0,
+      "learning_rate": 1.9651235721089575e-05,
+      "loss": 0.8754,
+      "step": 363
+    },
+    {
+      "epoch": 0.11229369119234922,
+      "grad_norm": 0.0,
+      "learning_rate": 1.9648614124982044e-05,
+      "loss": 0.9579,
+      "step": 364
+    },
+    {
+      "epoch": 0.11260219034397656,
+      "grad_norm": 0.0,
+      "learning_rate": 1.964598288890001e-05,
+      "loss": 0.9248,
+      "step": 365
+    },
+    {
+      "epoch": 0.11291068949560389,
+      "grad_norm": 0.0,
+      "learning_rate": 1.964334201547235e-05,
+      "loss": 1.0278,
+      "step": 366
+    },
+    {
+      "epoch": 0.11321918864723123,
+      "grad_norm": 0.0,
+      "learning_rate": 1.964069150733758e-05,
+      "loss": 0.8592,
+      "step": 367
+    },
+    {
+      "epoch": 0.11352768779885855,
+      "grad_norm": 0.0,
+      "learning_rate": 1.963803136714382e-05,
+      "loss": 0.9319,
+      "step": 368
+    },
+    {
+      "epoch": 0.11383618695048589,
+      "grad_norm": 0.0,
+      "learning_rate": 1.9635361597548844e-05,
+      "loss": 1.0014,
+      "step": 369
+    },
+    {
+      "epoch": 0.11414468610211322,
+      "grad_norm": 0.0,
+      "learning_rate": 1.9632682201220022e-05,
+      "loss": 0.9013,
+      "step": 370
+    },
+    {
+      "epoch": 0.11445318525374056,
+      "grad_norm": 0.0,
+      "learning_rate": 1.9629993180834356e-05,
+      "loss": 0.8994,
+      "step": 371
+    },
+    {
+      "epoch": 0.11476168440536788,
+      "grad_norm": 0.0,
+      "learning_rate": 1.9627294539078454e-05,
+      "loss": 0.901,
+      "step": 372
+    },
+    {
+      "epoch": 0.11507018355699522,
+      "grad_norm": 0.0,
+      "learning_rate": 1.9624586278648544e-05,
+      "loss": 1.0434,
+      "step": 373
+    },
+    {
+      "epoch": 0.11537868270862255,
+      "grad_norm": 0.0,
+      "learning_rate": 1.962186840225046e-05,
+      "loss": 0.9425,
+      "step": 374
+    },
+    {
+      "epoch": 0.11568718186024989,
+      "grad_norm": 0.0,
+      "learning_rate": 1.9619140912599643e-05,
+      "loss": 0.89,
+      "step": 375
+    },
+    {
+      "epoch": 0.11599568101187721,
+      "grad_norm": 0.0,
+      "learning_rate": 1.9616403812421147e-05,
+      "loss": 0.997,
+      "step": 376
+    },
+    {
+      "epoch": 0.11630418016350455,
+      "grad_norm": 0.0,
+      "learning_rate": 1.9613657104449615e-05,
+      "loss": 0.866,
+      "step": 377
+    },
+    {
+      "epoch": 0.11661267931513189,
+      "grad_norm": 0.0,
+      "learning_rate": 1.96109007914293e-05,
+      "loss": 0.9479,
+      "step": 378
+    },
+    {
+      "epoch": 0.11692117846675922,
+      "grad_norm": 0.0,
+      "learning_rate": 1.960813487611404e-05,
+      "loss": 0.9128,
+      "step": 379
+    },
+    {
+      "epoch": 0.11722967761838655,
+      "grad_norm": 0.0,
+      "learning_rate": 1.9605359361267282e-05,
+      "loss": 0.9501,
+      "step": 380
+    },
+    {
+      "epoch": 0.11753817677001388,
+      "grad_norm": 0.0,
+      "learning_rate": 1.960257424966205e-05,
+      "loss": 0.9546,
+      "step": 381
+    },
+    {
+      "epoch": 0.11784667592164122,
+      "grad_norm": 0.0,
+      "learning_rate": 1.9599779544080966e-05,
+      "loss": 0.9422,
+      "step": 382
+    },
+    {
+      "epoch": 0.11815517507326855,
+      "grad_norm": 0.0,
+      "learning_rate": 1.9596975247316226e-05,
+      "loss": 0.949,
+      "step": 383
+    },
+    {
+      "epoch": 0.11846367422489588,
+      "grad_norm": 0.0,
+      "learning_rate": 1.9594161362169627e-05,
+      "loss": 0.9639,
+      "step": 384
+    },
+    {
+      "epoch": 0.11877217337652321,
+      "grad_norm": 0.0,
+      "learning_rate": 1.9591337891452525e-05,
+      "loss": 0.8488,
+      "step": 385
+    },
+    {
+      "epoch": 0.11908067252815055,
+      "grad_norm": 0.0,
+      "learning_rate": 1.958850483798586e-05,
+      "loss": 0.9729,
+      "step": 386
+    },
+    {
+      "epoch": 0.11938917167977788,
+      "grad_norm": 0.0,
+      "learning_rate": 1.958566220460016e-05,
+      "loss": 0.944,
+      "step": 387
+    },
+    {
+      "epoch": 0.11969767083140521,
+      "grad_norm": 0.0,
+      "learning_rate": 1.9582809994135505e-05,
+      "loss": 0.996,
+      "step": 388
+    },
+    {
+      "epoch": 0.12000616998303254,
+      "grad_norm": 0.0,
+      "learning_rate": 1.9579948209441558e-05,
+      "loss": 0.891,
+      "step": 389
+    },
+    {
+      "epoch": 0.12031466913465988,
+      "grad_norm": 0.0,
+      "learning_rate": 1.957707685337753e-05,
+      "loss": 0.933,
+      "step": 390
+    },
+    {
+      "epoch": 0.12062316828628722,
+      "grad_norm": 0.0,
+      "learning_rate": 1.957419592881222e-05,
+      "loss": 0.9058,
+      "step": 391
+    },
+    {
+      "epoch": 0.12093166743791454,
+      "grad_norm": 0.0,
+      "learning_rate": 1.957130543862396e-05,
+      "loss": 0.9237,
+      "step": 392
+    },
+    {
+      "epoch": 0.12124016658954188,
+      "grad_norm": 0.0,
+      "learning_rate": 1.9568405385700658e-05,
+      "loss": 0.9637,
+      "step": 393
+    },
+    {
+      "epoch": 0.12154866574116921,
+      "grad_norm": 0.0,
+      "learning_rate": 1.956549577293977e-05,
+      "loss": 1.0328,
+      "step": 394
+    },
+    {
+      "epoch": 0.12185716489279655,
+      "grad_norm": 0.0,
+      "learning_rate": 1.9562576603248306e-05,
+      "loss": 0.9848,
+      "step": 395
+    },
+    {
+      "epoch": 0.12216566404442387,
+      "grad_norm": 0.0,
+      "learning_rate": 1.9559647879542817e-05,
+      "loss": 0.973,
+      "step": 396
+    },
+    {
+      "epoch": 0.12247416319605121,
+      "grad_norm": 0.0,
+      "learning_rate": 1.9556709604749408e-05,
+      "loss": 0.9057,
+      "step": 397
+    },
+    {
+      "epoch": 0.12278266234767854,
+      "grad_norm": 0.0,
+      "learning_rate": 1.9553761781803718e-05,
+      "loss": 0.8314,
+      "step": 398
+    },
+    {
+      "epoch": 0.12309116149930588,
+      "grad_norm": 0.0,
+      "learning_rate": 1.9550804413650935e-05,
+      "loss": 0.9103,
+      "step": 399
+    },
+    {
+      "epoch": 0.1233996606509332,
+      "grad_norm": 0.0,
+      "learning_rate": 1.954783750324578e-05,
+      "loss": 0.9178,
+      "step": 400
+    },
+    {
+      "epoch": 0.12370815980256054,
+      "grad_norm": 0.0,
+      "learning_rate": 1.9544861053552503e-05,
+      "loss": 0.9996,
+      "step": 401
+    },
+    {
+      "epoch": 0.12401665895418787,
+      "grad_norm": 0.0,
+      "learning_rate": 1.954187506754489e-05,
+      "loss": 0.9381,
+      "step": 402
+    },
+    {
+      "epoch": 0.12432515810581521,
+      "grad_norm": 0.0,
+      "learning_rate": 1.953887954820625e-05,
+      "loss": 0.9707,
+      "step": 403
+    },
+    {
+      "epoch": 0.12463365725744255,
+      "grad_norm": 0.0,
+      "learning_rate": 1.9535874498529423e-05,
+      "loss": 0.8752,
+      "step": 404
+    },
+    {
+      "epoch": 0.12494215640906987,
+      "grad_norm": 0.0,
+      "learning_rate": 1.953285992151677e-05,
+      "loss": 0.9752,
+      "step": 405
+    },
+    {
+      "epoch": 0.1252506555606972,
+      "grad_norm": 0.0,
+      "learning_rate": 1.9529835820180166e-05,
+      "loss": 0.9104,
+      "step": 406
+    },
+    {
+      "epoch": 0.12555915471232454,
+      "grad_norm": 0.0,
+      "learning_rate": 1.952680219754101e-05,
+      "loss": 0.9053,
+      "step": 407
+    },
+    {
+      "epoch": 0.12586765386395188,
+      "grad_norm": 0.0,
+      "learning_rate": 1.9523759056630196e-05,
+      "loss": 0.8727,
+      "step": 408
+    },
+    {
+      "epoch": 0.12617615301557922,
+      "grad_norm": 0.0,
+      "learning_rate": 1.952070640048815e-05,
+      "loss": 0.871,
+      "step": 409
+    },
+    {
+      "epoch": 0.12648465216720653,
+      "grad_norm": 0.0,
+      "learning_rate": 1.9517644232164793e-05,
+      "loss": 0.8591,
+      "step": 410
+    },
+    {
+      "epoch": 0.12679315131883387,
+      "grad_norm": 0.0,
+      "learning_rate": 1.951457255471955e-05,
+      "loss": 0.9535,
+      "step": 411
+    },
+    {
+      "epoch": 0.1271016504704612,
+      "grad_norm": 0.0,
+      "learning_rate": 1.9511491371221347e-05,
+      "loss": 0.9831,
+      "step": 412
+    },
+    {
+      "epoch": 0.12741014962208855,
+      "grad_norm": 0.0,
+      "learning_rate": 1.9508400684748615e-05,
+      "loss": 0.9473,
+      "step": 413
+    },
+    {
+      "epoch": 0.12771864877371586,
+      "grad_norm": 0.0,
+      "learning_rate": 1.950530049838926e-05,
+      "loss": 1.0712,
+      "step": 414
+    },
+    {
+      "epoch": 0.1280271479253432,
+      "grad_norm": 0.0,
+      "learning_rate": 1.9502190815240708e-05,
+      "loss": 0.9336,
+      "step": 415
+    },
+    {
+      "epoch": 0.12833564707697054,
+      "grad_norm": 0.0,
+      "learning_rate": 1.9499071638409845e-05,
+      "loss": 0.8945,
+      "step": 416
+    },
+    {
+      "epoch": 0.12864414622859788,
+      "grad_norm": 0.0,
+      "learning_rate": 1.949594297101306e-05,
+      "loss": 0.8821,
+      "step": 417
+    },
+    {
+      "epoch": 0.12895264538022522,
+      "grad_norm": 0.0,
+      "learning_rate": 1.9492804816176223e-05,
+      "loss": 0.8441,
+      "step": 418
+    },
+    {
+      "epoch": 0.12926114453185253,
+      "grad_norm": 0.0,
+      "learning_rate": 1.9489657177034673e-05,
+      "loss": 0.8545,
+      "step": 419
+    },
+    {
+      "epoch": 0.12956964368347987,
+      "grad_norm": 0.0,
+      "learning_rate": 1.948650005673323e-05,
+      "loss": 0.9616,
+      "step": 420
+    },
+    {
+      "epoch": 0.1298781428351072,
+      "grad_norm": 0.0,
+      "learning_rate": 1.9483333458426192e-05,
+      "loss": 1.0049,
+      "step": 421
+    },
+    {
+      "epoch": 0.13018664198673455,
+      "grad_norm": 0.0,
+      "learning_rate": 1.948015738527732e-05,
+      "loss": 0.9311,
+      "step": 422
+    },
+    {
+      "epoch": 0.13049514113836186,
+      "grad_norm": 0.0,
+      "learning_rate": 1.947697184045984e-05,
+      "loss": 0.9357,
+      "step": 423
+    },
+    {
+      "epoch": 0.1308036402899892,
+      "grad_norm": 0.0,
+      "learning_rate": 1.947377682715645e-05,
+      "loss": 0.8408,
+      "step": 424
+    },
+    {
+      "epoch": 0.13111213944161654,
+      "grad_norm": 0.0,
+      "learning_rate": 1.9470572348559295e-05,
+      "loss": 0.885,
+      "step": 425
+    },
+    {
+      "epoch": 0.13142063859324388,
+      "grad_norm": 0.0,
+      "learning_rate": 1.946735840786999e-05,
+      "loss": 0.9128,
+      "step": 426
+    },
+    {
+      "epoch": 0.1317291377448712,
+      "grad_norm": 0.0,
+      "learning_rate": 1.94641350082996e-05,
+      "loss": 0.9241,
+      "step": 427
+    },
+    {
+      "epoch": 0.13203763689649853,
+      "grad_norm": 0.0,
+      "learning_rate": 1.9460902153068633e-05,
+      "loss": 0.975,
+      "step": 428
+    },
+    {
+      "epoch": 0.13234613604812587,
+      "grad_norm": 0.0,
+      "learning_rate": 1.945765984540705e-05,
+      "loss": 0.8414,
+      "step": 429
+    },
+    {
+      "epoch": 0.1326546351997532,
+      "grad_norm": 0.0,
+      "learning_rate": 1.945440808855426e-05,
+      "loss": 0.9713,
+      "step": 430
+    },
+    {
+      "epoch": 0.13296313435138055,
+      "grad_norm": 0.0,
+      "learning_rate": 1.94511468857591e-05,
+      "loss": 0.9718,
+      "step": 431
+    },
+    {
+      "epoch": 0.13327163350300786,
+      "grad_norm": 0.0,
+      "learning_rate": 1.944787624027986e-05,
+      "loss": 0.8874,
+      "step": 432
+    },
+    {
+      "epoch": 0.1335801326546352,
+      "grad_norm": 0.0,
+      "learning_rate": 1.9444596155384253e-05,
+      "loss": 0.9717,
+      "step": 433
+    },
+    {
+      "epoch": 0.13388863180626254,
+      "grad_norm": 0.0,
+      "learning_rate": 1.944130663434943e-05,
+      "loss": 0.9274,
+      "step": 434
+    },
+    {
+      "epoch": 0.13419713095788988,
+      "grad_norm": 0.0,
+      "learning_rate": 1.9438007680461965e-05,
+      "loss": 0.879,
+      "step": 435
+    },
+    {
+      "epoch": 0.1345056301095172,
+      "grad_norm": 0.0,
+      "learning_rate": 1.9434699297017855e-05,
+      "loss": 0.8611,
+      "step": 436
+    },
+    {
+      "epoch": 0.13481412926114453,
+      "grad_norm": 0.0,
+      "learning_rate": 1.9431381487322527e-05,
+      "loss": 0.9236,
+      "step": 437
+    },
+    {
+      "epoch": 0.13512262841277187,
+      "grad_norm": 0.0,
+      "learning_rate": 1.9428054254690812e-05,
+      "loss": 0.9343,
+      "step": 438
+    },
+    {
+      "epoch": 0.1354311275643992,
+      "grad_norm": 0.0,
+      "learning_rate": 1.9424717602446973e-05,
+      "loss": 0.831,
+      "step": 439
+    },
+    {
+      "epoch": 0.13573962671602652,
+      "grad_norm": 0.0,
+      "learning_rate": 1.9421371533924664e-05,
+      "loss": 0.8447,
+      "step": 440
+    },
+    {
+      "epoch": 0.13604812586765386,
+      "grad_norm": 0.0,
+      "learning_rate": 1.9418016052466965e-05,
+      "loss": 0.9319,
+      "step": 441
+    },
+    {
+      "epoch": 0.1363566250192812,
+      "grad_norm": 0.0,
+      "learning_rate": 1.941465116142635e-05,
+      "loss": 1.0648,
+      "step": 442
+    },
+    {
+      "epoch": 0.13666512417090854,
+      "grad_norm": 0.0,
+      "learning_rate": 1.94112768641647e-05,
+      "loss": 0.8477,
+      "step": 443
+    },
+    {
+      "epoch": 0.13697362332253588,
+      "grad_norm": 0.0,
+      "learning_rate": 1.9407893164053285e-05,
+      "loss": 0.8829,
+      "step": 444
+    },
+    {
+      "epoch": 0.1372821224741632,
+      "grad_norm": 0.0,
+      "learning_rate": 1.940450006447278e-05,
+      "loss": 0.9168,
+      "step": 445
+    },
+    {
+      "epoch": 0.13759062162579053,
+      "grad_norm": 0.0,
+      "learning_rate": 1.9401097568813244e-05,
+      "loss": 0.936,
+      "step": 446
+    },
+    {
+      "epoch": 0.13789912077741787,
+      "grad_norm": 0.0,
+      "learning_rate": 1.939768568047413e-05,
+      "loss": 0.8714,
+      "step": 447
+    },
+    {
+      "epoch": 0.1382076199290452,
+      "grad_norm": 0.0,
+      "learning_rate": 1.9394264402864265e-05,
+      "loss": 0.9071,
+      "step": 448
+    },
+    {
+      "epoch": 0.13851611908067252,
+      "grad_norm": 0.0,
+      "learning_rate": 1.939083373940187e-05,
+      "loss": 0.8958,
+      "step": 449
+    },
+    {
+      "epoch": 0.13882461823229986,
+      "grad_norm": 0.0,
+      "learning_rate": 1.9387393693514532e-05,
+      "loss": 0.9762,
+      "step": 450
+    },
+    {
+      "epoch": 0.1391331173839272,
+      "grad_norm": 0.0,
+      "learning_rate": 1.9383944268639213e-05,
+      "loss": 0.942,
+      "step": 451
+    },
+    {
+      "epoch": 0.13944161653555454,
+      "grad_norm": 0.0,
+      "learning_rate": 1.9380485468222257e-05,
+      "loss": 0.9979,
+      "step": 452
+    },
+    {
+      "epoch": 0.13975011568718185,
+      "grad_norm": 0.0,
+      "learning_rate": 1.9377017295719362e-05,
+      "loss": 0.9849,
+      "step": 453
+    },
+    {
+      "epoch": 0.1400586148388092,
+      "grad_norm": 0.0,
+      "learning_rate": 1.9373539754595598e-05,
+      "loss": 0.9884,
+      "step": 454
+    },
+    {
+      "epoch": 0.14036711399043653,
+      "grad_norm": 0.0,
+      "learning_rate": 1.9370052848325392e-05,
+      "loss": 0.9648,
+      "step": 455
+    },
+    {
+      "epoch": 0.14067561314206387,
+      "grad_norm": 0.0,
+      "learning_rate": 1.9366556580392527e-05,
+      "loss": 0.893,
+      "step": 456
+    },
+    {
+      "epoch": 0.1409841122936912,
+      "grad_norm": 0.0,
+      "learning_rate": 1.936305095429014e-05,
+      "loss": 0.8304,
+      "step": 457
+    },
+    {
+      "epoch": 0.14129261144531852,
+      "grad_norm": 0.0,
+      "learning_rate": 1.935953597352072e-05,
+      "loss": 0.936,
+      "step": 458
+    },
+    {
+      "epoch": 0.14160111059694586,
+      "grad_norm": 0.0,
+      "learning_rate": 1.9356011641596096e-05,
+      "loss": 0.8423,
+      "step": 459
+    },
+    {
+      "epoch": 0.1419096097485732,
+      "grad_norm": 0.0,
+      "learning_rate": 1.9352477962037448e-05,
+      "loss": 0.9692,
+      "step": 460
+    },
+    {
+      "epoch": 0.14221810890020053,
+      "grad_norm": 0.0,
+      "learning_rate": 1.934893493837529e-05,
+      "loss": 0.9955,
+      "step": 461
+    },
+    {
+      "epoch": 0.14252660805182785,
+      "grad_norm": 0.0,
+      "learning_rate": 1.9345382574149473e-05,
+      "loss": 1.0374,
+      "step": 462
+    },
+    {
+      "epoch": 0.14283510720345519,
+      "grad_norm": 0.0,
+      "learning_rate": 1.9341820872909184e-05,
+      "loss": 0.9431,
+      "step": 463
+    },
+    {
+      "epoch": 0.14314360635508253,
+      "grad_norm": 0.0,
+      "learning_rate": 1.933824983821293e-05,
+      "loss": 0.951,
+      "step": 464
+    },
+    {
+      "epoch": 0.14345210550670986,
+      "grad_norm": 0.0,
+      "learning_rate": 1.933466947362855e-05,
+      "loss": 0.8325,
+      "step": 465
+    },
+    {
+      "epoch": 0.14376060465833718,
+      "grad_norm": 0.0,
+      "learning_rate": 1.9331079782733204e-05,
+      "loss": 0.859,
+      "step": 466
+    },
+    {
+      "epoch": 0.14406910380996452,
+      "grad_norm": 0.0,
+      "learning_rate": 1.9327480769113366e-05,
+      "loss": 0.8799,
+      "step": 467
+    },
+    {
+      "epoch": 0.14437760296159186,
+      "grad_norm": 0.0,
+      "learning_rate": 1.9323872436364825e-05,
+      "loss": 0.8794,
+      "step": 468
+    },
+    {
+      "epoch": 0.1446861021132192,
+      "grad_norm": 0.0,
+      "learning_rate": 1.932025478809269e-05,
+      "loss": 0.928,
+      "step": 469
+    },
+    {
+      "epoch": 0.14499460126484653,
+      "grad_norm": 0.0,
+      "learning_rate": 1.9316627827911366e-05,
+      "loss": 0.895,
+      "step": 470
+    },
+    {
+      "epoch": 0.14530310041647385,
+      "grad_norm": 0.0,
+      "learning_rate": 1.9312991559444565e-05,
+      "loss": 0.9044,
+      "step": 471
+    },
+    {
+      "epoch": 0.14561159956810119,
+      "grad_norm": 0.0,
+      "learning_rate": 1.9309345986325298e-05,
+      "loss": 0.975,
+      "step": 472
+    },
+    {
+      "epoch": 0.14592009871972852,
+      "grad_norm": 0.0,
+      "learning_rate": 1.9305691112195876e-05,
+      "loss": 0.9109,
+      "step": 473
+    },
+    {
+      "epoch": 0.14622859787135586,
+      "grad_norm": 0.0,
+      "learning_rate": 1.9302026940707903e-05,
+      "loss": 0.8694,
+      "step": 474
+    },
+    {
+      "epoch": 0.14653709702298318,
+      "grad_norm": 0.0,
+      "learning_rate": 1.9298353475522263e-05,
+      "loss": 0.8612,
+      "step": 475
+    },
+    {
+      "epoch": 0.14684559617461052,
+      "grad_norm": 0.0,
+      "learning_rate": 1.929467072030914e-05,
+      "loss": 0.9448,
+      "step": 476
+    },
+    {
+      "epoch": 0.14715409532623785,
+      "grad_norm": 0.0,
+      "learning_rate": 1.9290978678747984e-05,
+      "loss": 0.9417,
+      "step": 477
+    },
+    {
+      "epoch": 0.1474625944778652,
+      "grad_norm": 0.0,
+      "learning_rate": 1.9287277354527535e-05,
+      "loss": 0.8778,
+      "step": 478
+    },
+    {
+      "epoch": 0.1477710936294925,
+      "grad_norm": 0.0,
+      "learning_rate": 1.92835667513458e-05,
+      "loss": 0.9273,
+      "step": 479
+    },
+    {
+      "epoch": 0.14807959278111985,
+      "grad_norm": 0.0,
+      "learning_rate": 1.927984687291006e-05,
+      "loss": 0.9406,
+      "step": 480
+    },
+    {
+      "epoch": 0.14838809193274718,
+      "grad_norm": 0.0,
+      "learning_rate": 1.9276117722936867e-05,
+      "loss": 0.9262,
+      "step": 481
+    },
+    {
+      "epoch": 0.14869659108437452,
+      "grad_norm": 0.0,
+      "learning_rate": 1.9272379305152026e-05,
+      "loss": 1.0137,
+      "step": 482
+    },
+    {
+      "epoch": 0.14900509023600186,
+      "grad_norm": 0.0,
+      "learning_rate": 1.926863162329061e-05,
+      "loss": 0.8827,
+      "step": 483
+    },
+    {
+      "epoch": 0.14931358938762918,
+      "grad_norm": 0.0,
+      "learning_rate": 1.9264874681096948e-05,
+      "loss": 0.8382,
+      "step": 484
+    },
+    {
+      "epoch": 0.14962208853925651,
+      "grad_norm": 0.0,
+      "learning_rate": 1.9261108482324612e-05,
+      "loss": 0.8825,
+      "step": 485
+    },
+    {
+      "epoch": 0.14993058769088385,
+      "grad_norm": 0.0,
+      "learning_rate": 1.9257333030736433e-05,
+      "loss": 0.8738,
+      "step": 486
+    },
+    {
+      "epoch": 0.1502390868425112,
+      "grad_norm": 0.0,
+      "learning_rate": 1.925354833010448e-05,
+      "loss": 0.9005,
+      "step": 487
+    },
+    {
+      "epoch": 0.1505475859941385,
+      "grad_norm": 0.0,
+      "learning_rate": 1.9249754384210066e-05,
+      "loss": 0.9739,
+      "step": 488
+    },
+    {
+      "epoch": 0.15085608514576584,
+      "grad_norm": 0.0,
+      "learning_rate": 1.9245951196843736e-05,
+      "loss": 0.9109,
+      "step": 489
+    },
+    {
+      "epoch": 0.15116458429739318,
+      "grad_norm": 0.0,
+      "learning_rate": 1.924213877180528e-05,
+      "loss": 0.9328,
+      "step": 490
+    },
+    {
+      "epoch": 0.15147308344902052,
+      "grad_norm": 0.0,
+      "learning_rate": 1.92383171129037e-05,
+      "loss": 0.9959,
+      "step": 491
+    },
+    {
+      "epoch": 0.15178158260064784,
+      "grad_norm": 0.0,
+      "learning_rate": 1.9234486223957238e-05,
+      "loss": 0.8328,
+      "step": 492
+    },
+    {
+      "epoch": 0.15209008175227517,
+      "grad_norm": 0.0,
+      "learning_rate": 1.9230646108793353e-05,
+      "loss": 1.26,
+      "step": 493
+    },
+    {
+      "epoch": 0.15239858090390251,
+      "grad_norm": 0.0,
+      "learning_rate": 1.922679677124872e-05,
+      "loss": 0.9531,
+      "step": 494
+    },
+    {
+      "epoch": 0.15270708005552985,
+      "grad_norm": 0.0,
+      "learning_rate": 1.9222938215169227e-05,
+      "loss": 0.8436,
+      "step": 495
+    },
+    {
+      "epoch": 0.1530155792071572,
+      "grad_norm": 0.0,
+      "learning_rate": 1.921907044440998e-05,
+      "loss": 0.909,
+      "step": 496
+    },
+    {
+      "epoch": 0.1533240783587845,
+      "grad_norm": 0.0,
+      "learning_rate": 1.9215193462835285e-05,
+      "loss": 1.0067,
+      "step": 497
+    },
+    {
+      "epoch": 0.15363257751041184,
+      "grad_norm": 0.0,
+      "learning_rate": 1.921130727431865e-05,
+      "loss": 0.8809,
+      "step": 498
+    },
+    {
+      "epoch": 0.15394107666203918,
+      "grad_norm": 0.0,
+      "learning_rate": 1.9207411882742784e-05,
+      "loss": 1.0474,
+      "step": 499
+    },
+    {
+      "epoch": 0.15424957581366652,
+      "grad_norm": 0.0,
+      "learning_rate": 1.9203507291999598e-05,
+      "loss": 0.9858,
+      "step": 500
+    },
+    {
+      "epoch": 0.15455807496529383,
+      "grad_norm": 0.0,
+      "learning_rate": 1.9199593505990173e-05,
+      "loss": 0.9519,
+      "step": 501
+    },
+    {
+      "epoch": 0.15486657411692117,
+      "grad_norm": 0.0,
+      "learning_rate": 1.91956705286248e-05,
+      "loss": 0.9109,
+      "step": 502
+    },
+    {
+      "epoch": 0.1551750732685485,
+      "grad_norm": 0.0,
+      "learning_rate": 1.9191738363822943e-05,
+      "loss": 0.9116,
+      "step": 503
+    },
+    {
+      "epoch": 0.15548357242017585,
+      "grad_norm": 0.0,
+      "learning_rate": 1.9187797015513244e-05,
+      "loss": 0.9334,
+      "step": 504
+    },
+    {
+      "epoch": 0.15579207157180316,
+      "grad_norm": 0.0,
+      "learning_rate": 1.9183846487633524e-05,
+      "loss": 0.9153,
+      "step": 505
+    },
+    {
+      "epoch": 0.1561005707234305,
+      "grad_norm": 0.0,
+      "learning_rate": 1.9179886784130767e-05,
+      "loss": 0.9827,
+      "step": 506
+    },
+    {
+      "epoch": 0.15640906987505784,
+      "grad_norm": 0.0,
+      "learning_rate": 1.917591790896114e-05,
+      "loss": 0.9264,
+      "step": 507
+    },
+    {
+      "epoch": 0.15671756902668518,
+      "grad_norm": 0.0,
+      "learning_rate": 1.917193986608996e-05,
+      "loss": 0.8661,
+      "step": 508
+    },
+    {
+      "epoch": 0.15702606817831252,
+      "grad_norm": 0.0,
+      "learning_rate": 1.9167952659491703e-05,
+      "loss": 1.0055,
+      "step": 509
+    },
+    {
+      "epoch": 0.15733456732993983,
+      "grad_norm": 0.0,
+      "learning_rate": 1.9163956293150017e-05,
+      "loss": 0.984,
+      "step": 510
+    },
+    {
+      "epoch": 0.15764306648156717,
+      "grad_norm": 0.0,
+      "learning_rate": 1.915995077105768e-05,
+      "loss": 1.028,
+      "step": 511
+    },
+    {
+      "epoch": 0.1579515656331945,
+      "grad_norm": 0.0,
+      "learning_rate": 1.915593609721663e-05,
+      "loss": 0.9713,
+      "step": 512
+    },
+    {
+      "epoch": 0.15826006478482185,
+      "grad_norm": 0.0,
+      "learning_rate": 1.9151912275637946e-05,
+      "loss": 0.941,
+      "step": 513
+    },
+    {
+      "epoch": 0.15856856393644916,
+      "grad_norm": 0.0,
+      "learning_rate": 1.914787931034185e-05,
+      "loss": 0.9633,
+      "step": 514
+    },
+    {
+      "epoch": 0.1588770630880765,
+      "grad_norm": 0.0,
+      "learning_rate": 1.914383720535769e-05,
+      "loss": 0.9754,
+      "step": 515
+    },
+    {
+      "epoch": 0.15918556223970384,
+      "grad_norm": 0.0,
+      "learning_rate": 1.9139785964723955e-05,
+      "loss": 0.9942,
+      "step": 516
+    },
+    {
+      "epoch": 0.15949406139133118,
+      "grad_norm": 0.0,
+      "learning_rate": 1.913572559248826e-05,
+      "loss": 1.3288,
+      "step": 517
+    },
+    {
+      "epoch": 0.1598025605429585,
+      "grad_norm": 0.0,
+      "learning_rate": 1.9131656092707337e-05,
+      "loss": 0.8349,
+      "step": 518
+    },
+    {
+      "epoch": 0.16011105969458583,
+      "grad_norm": 0.0,
+      "learning_rate": 1.9127577469447045e-05,
+      "loss": 1.0034,
+      "step": 519
+    },
+    {
+      "epoch": 0.16041955884621317,
+      "grad_norm": 0.0,
+      "learning_rate": 1.912348972678235e-05,
+      "loss": 1.005,
+      "step": 520
+    },
+    {
+      "epoch": 0.1607280579978405,
+      "grad_norm": 0.0,
+      "learning_rate": 1.9119392868797347e-05,
+      "loss": 0.8971,
+      "step": 521
+    },
+    {
+      "epoch": 0.16103655714946785,
+      "grad_norm": 0.0,
+      "learning_rate": 1.9115286899585214e-05,
+      "loss": 0.9835,
+      "step": 522
+    },
+    {
+      "epoch": 0.16134505630109516,
+      "grad_norm": 0.0,
+      "learning_rate": 1.9111171823248243e-05,
+      "loss": 0.9274,
+      "step": 523
+    },
+    {
+      "epoch": 0.1616535554527225,
+      "grad_norm": 0.0,
+      "learning_rate": 1.9107047643897835e-05,
+      "loss": 0.8566,
+      "step": 524
+    },
+    {
+      "epoch": 0.16196205460434984,
+      "grad_norm": 0.0,
+      "learning_rate": 1.9102914365654465e-05,
+      "loss": 0.933,
+      "step": 525
+    },
+    {
+      "epoch": 0.16227055375597718,
+      "grad_norm": 0.0,
+      "learning_rate": 1.909877199264772e-05,
+      "loss": 0.9228,
+      "step": 526
+    },
+    {
+      "epoch": 0.1625790529076045,
+      "grad_norm": 0.0,
+      "learning_rate": 1.909462052901626e-05,
+      "loss": 0.9014,
+      "step": 527
+    },
+    {
+      "epoch": 0.16288755205923183,
+      "grad_norm": 0.0,
+      "learning_rate": 1.909045997890783e-05,
+      "loss": 1.0034,
+      "step": 528
+    },
+    {
+      "epoch": 0.16319605121085917,
+      "grad_norm": 0.0,
+      "learning_rate": 1.9086290346479254e-05,
+      "loss": 0.906,
+      "step": 529
+    },
+    {
+      "epoch": 0.1635045503624865,
+      "grad_norm": 0.0,
+      "learning_rate": 1.9082111635896438e-05,
+      "loss": 0.9278,
+      "step": 530
+    },
+    {
+      "epoch": 0.16381304951411382,
+      "grad_norm": 0.0,
+      "learning_rate": 1.9077923851334342e-05,
+      "loss": 0.9236,
+      "step": 531
+    },
+    {
+      "epoch": 0.16412154866574116,
+      "grad_norm": 0.0,
+      "learning_rate": 1.9073726996977004e-05,
+      "loss": 0.9511,
+      "step": 532
+    },
+    {
+      "epoch": 0.1644300478173685,
+      "grad_norm": 0.0,
+      "learning_rate": 1.906952107701752e-05,
+      "loss": 0.7983,
+      "step": 533
+    },
+    {
+      "epoch": 0.16473854696899584,
+      "grad_norm": 0.0,
+      "learning_rate": 1.9065306095658048e-05,
+      "loss": 0.9158,
+      "step": 534
+    },
+    {
+      "epoch": 0.16504704612062318,
+      "grad_norm": 0.0,
+      "learning_rate": 1.9061082057109787e-05,
+      "loss": 0.9543,
+      "step": 535
+    },
+    {
+      "epoch": 0.1653555452722505,
+      "grad_norm": 0.0,
+      "learning_rate": 1.9056848965593e-05,
+      "loss": 0.9502,
+      "step": 536
+    },
+    {
+      "epoch": 0.16566404442387783,
+      "grad_norm": 0.0,
+      "learning_rate": 1.905260682533699e-05,
+      "loss": 0.9167,
+      "step": 537
+    },
+    {
+      "epoch": 0.16597254357550517,
+      "grad_norm": 0.0,
+      "learning_rate": 1.9048355640580087e-05,
+      "loss": 1.0018,
+      "step": 538
+    },
+    {
+      "epoch": 0.1662810427271325,
+      "grad_norm": 0.0,
+      "learning_rate": 1.904409541556968e-05,
+      "loss": 0.8541,
+      "step": 539
+    },
+    {
+      "epoch": 0.16658954187875982,
+      "grad_norm": 0.0,
+      "learning_rate": 1.903982615456218e-05,
+      "loss": 0.8704,
+      "step": 540
+    },
+    {
+      "epoch": 0.16689804103038716,
+      "grad_norm": 0.0,
+      "learning_rate": 1.9035547861823016e-05,
+      "loss": 0.9562,
+      "step": 541
+    },
+    {
+      "epoch": 0.1672065401820145,
+      "grad_norm": 0.0,
+      "learning_rate": 1.903126054162666e-05,
+      "loss": 0.9997,
+      "step": 542
+    },
+    {
+      "epoch": 0.16751503933364184,
+      "grad_norm": 0.0,
+      "learning_rate": 1.9026964198256583e-05,
+      "loss": 0.9693,
+      "step": 543
+    },
+    {
+      "epoch": 0.16782353848526915,
+      "grad_norm": 0.0,
+      "learning_rate": 1.902265883600529e-05,
+      "loss": 0.8984,
+      "step": 544
+    },
+    {
+      "epoch": 0.1681320376368965,
+      "grad_norm": 0.0,
+      "learning_rate": 1.9018344459174285e-05,
+      "loss": 0.8469,
+      "step": 545
+    },
+    {
+      "epoch": 0.16844053678852383,
+      "grad_norm": 0.0,
+      "learning_rate": 1.901402107207408e-05,
+      "loss": 0.8924,
+      "step": 546
+    },
+    {
+      "epoch": 0.16874903594015117,
+      "grad_norm": 0.0,
+      "learning_rate": 1.900968867902419e-05,
+      "loss": 0.9235,
+      "step": 547
+    },
+    {
+      "epoch": 0.1690575350917785,
+      "grad_norm": 0.0,
+      "learning_rate": 1.9005347284353136e-05,
+      "loss": 1.2873,
+      "step": 548
+    },
+    {
+      "epoch": 0.16936603424340582,
+      "grad_norm": 0.0,
+      "learning_rate": 1.9000996892398418e-05,
+      "loss": 0.9419,
+      "step": 549
+    },
+    {
+      "epoch": 0.16967453339503316,
+      "grad_norm": 0.0,
+      "learning_rate": 1.8996637507506538e-05,
+      "loss": 0.8735,
+      "step": 550
+    },
+    {
+      "epoch": 0.1699830325466605,
+      "grad_norm": 0.0,
+      "learning_rate": 1.899226913403297e-05,
+      "loss": 0.8714,
+      "step": 551
+    },
+    {
+      "epoch": 0.17029153169828784,
+      "grad_norm": 0.0,
+      "learning_rate": 1.898789177634218e-05,
+      "loss": 0.9299,
+      "step": 552
+    },
+    {
+      "epoch": 0.17060003084991515,
+      "grad_norm": 0.0,
+      "learning_rate": 1.898350543880761e-05,
+      "loss": 0.8985,
+      "step": 553
+    },
+    {
+      "epoch": 0.1709085300015425,
+      "grad_norm": 0.0,
+      "learning_rate": 1.897911012581166e-05,
+      "loss": 0.936,
+      "step": 554
+    },
+    {
+      "epoch": 0.17121702915316983,
+      "grad_norm": 0.0,
+      "learning_rate": 1.8974705841745717e-05,
+      "loss": 0.9537,
+      "step": 555
+    },
+    {
+      "epoch": 0.17152552830479717,
+      "grad_norm": 0.0,
+      "learning_rate": 1.897029259101012e-05,
+      "loss": 1.2198,
+      "step": 556
+    },
+    {
+      "epoch": 0.17183402745642448,
+      "grad_norm": 0.0,
+      "learning_rate": 1.896587037801416e-05,
+      "loss": 0.9152,
+      "step": 557
+    },
+    {
+      "epoch": 0.17214252660805182,
+      "grad_norm": 0.0,
+      "learning_rate": 1.89614392071761e-05,
+      "loss": 0.9322,
+      "step": 558
+    },
+    {
+      "epoch": 0.17245102575967916,
+      "grad_norm": 0.0,
+      "learning_rate": 1.895699908292314e-05,
+      "loss": 0.9327,
+      "step": 559
+    },
+    {
+      "epoch": 0.1727595249113065,
+      "grad_norm": 0.0,
+      "learning_rate": 1.8952550009691424e-05,
+      "loss": 0.9372,
+      "step": 560
+    },
+    {
+      "epoch": 0.17306802406293384,
+      "grad_norm": 0.0,
+      "learning_rate": 1.894809199192605e-05,
+      "loss": 0.9337,
+      "step": 561
+    },
+    {
+      "epoch": 0.17337652321456115,
+      "grad_norm": 0.0,
+      "learning_rate": 1.8943625034081032e-05,
+      "loss": 0.9088,
+      "step": 562
+    },
+    {
+      "epoch": 0.1736850223661885,
+      "grad_norm": 0.0,
+      "learning_rate": 1.8939149140619338e-05,
+      "loss": 0.9155,
+      "step": 563
+    },
+    {
+      "epoch": 0.17399352151781583,
+      "grad_norm": 0.0,
+      "learning_rate": 1.8934664316012853e-05,
+      "loss": 0.9342,
+      "step": 564
+    },
+    {
+      "epoch": 0.17430202066944317,
+      "grad_norm": 0.0,
+      "learning_rate": 1.8930170564742377e-05,
+      "loss": 0.9431,
+      "step": 565
+    },
+    {
+      "epoch": 0.17461051982107048,
+      "grad_norm": 0.0,
+      "learning_rate": 1.8925667891297646e-05,
+      "loss": 0.843,
+      "step": 566
+    },
+    {
+      "epoch": 0.17491901897269782,
+      "grad_norm": 0.0,
+      "learning_rate": 1.8921156300177303e-05,
+      "loss": 0.881,
+      "step": 567
+    },
+    {
+      "epoch": 0.17522751812432516,
+      "grad_norm": 0.0,
+      "learning_rate": 1.8916635795888895e-05,
+      "loss": 0.9073,
+      "step": 568
+    },
+    {
+      "epoch": 0.1755360172759525,
+      "grad_norm": 0.0,
+      "learning_rate": 1.8912106382948875e-05,
+      "loss": 0.9315,
+      "step": 569
+    },
+    {
+      "epoch": 0.1758445164275798,
+      "grad_norm": 0.0,
+      "learning_rate": 1.890756806588261e-05,
+      "loss": 0.9232,
+      "step": 570
+    },
+    {
+      "epoch": 0.17615301557920715,
+      "grad_norm": 0.0,
+      "learning_rate": 1.8903020849224343e-05,
+      "loss": 0.8926,
+      "step": 571
+    },
+    {
+      "epoch": 0.1764615147308345,
+      "grad_norm": 0.0,
+      "learning_rate": 1.8898464737517225e-05,
+      "loss": 0.8986,
+      "step": 572
+    },
+    {
+      "epoch": 0.17677001388246183,
+      "grad_norm": 0.0,
+      "learning_rate": 1.8893899735313285e-05,
+      "loss": 0.9223,
+      "step": 573
+    },
+    {
+      "epoch": 0.17707851303408917,
+      "grad_norm": 0.0,
+      "learning_rate": 1.8889325847173444e-05,
+      "loss": 0.8557,
+      "step": 574
+    },
+    {
+      "epoch": 0.17738701218571648,
+      "grad_norm": 0.0,
+      "learning_rate": 1.8884743077667487e-05,
+      "loss": 0.924,
+      "step": 575
+    },
+    {
+      "epoch": 0.17769551133734382,
+      "grad_norm": 0.0,
+      "learning_rate": 1.8880151431374082e-05,
+      "loss": 0.9458,
+      "step": 576
+    },
+    {
+      "epoch": 0.17800401048897116,
+      "grad_norm": 0.0,
+      "learning_rate": 1.8875550912880766e-05,
+      "loss": 0.9072,
+      "step": 577
+    },
+    {
+      "epoch": 0.1783125096405985,
+      "grad_norm": 0.0,
+      "learning_rate": 1.8870941526783933e-05,
+      "loss": 0.895,
+      "step": 578
+    },
+    {
+      "epoch": 0.1786210087922258,
+      "grad_norm": 0.0,
+      "learning_rate": 1.886632327768885e-05,
+      "loss": 0.9086,
+      "step": 579
+    },
+    {
+      "epoch": 0.17892950794385315,
+      "grad_norm": 0.0,
+      "learning_rate": 1.8861696170209626e-05,
+      "loss": 0.8276,
+      "step": 580
+    },
+    {
+      "epoch": 0.1792380070954805,
+      "grad_norm": 0.0,
+      "learning_rate": 1.8857060208969225e-05,
+      "loss": 0.9947,
+      "step": 581
+    },
+    {
+      "epoch": 0.17954650624710783,
+      "grad_norm": 0.0,
+      "learning_rate": 1.8852415398599455e-05,
+      "loss": 0.9668,
+      "step": 582
+    },
+    {
+      "epoch": 0.17985500539873514,
+      "grad_norm": 0.0,
+      "learning_rate": 1.8847761743740972e-05,
+      "loss": 0.9431,
+      "step": 583
+    },
+    {
+      "epoch": 0.18016350455036248,
+      "grad_norm": 0.0,
+      "learning_rate": 1.8843099249043258e-05,
+      "loss": 0.9453,
+      "step": 584
+    },
+    {
+      "epoch": 0.18047200370198982,
+      "grad_norm": 0.0,
+      "learning_rate": 1.8838427919164643e-05,
+      "loss": 0.9127,
+      "step": 585
+    },
+    {
+      "epoch": 0.18078050285361716,
+      "grad_norm": 0.0,
+      "learning_rate": 1.8833747758772264e-05,
+      "loss": 0.9754,
+      "step": 586
+    },
+    {
+      "epoch": 0.1810890020052445,
+      "grad_norm": 0.0,
+      "learning_rate": 1.8829058772542092e-05,
+      "loss": 0.9359,
+      "step": 587
+    },
+    {
+      "epoch": 0.1813975011568718,
+      "grad_norm": 0.0,
+      "learning_rate": 1.882436096515892e-05,
+      "loss": 0.9143,
+      "step": 588
+    },
+    {
+      "epoch": 0.18170600030849915,
+      "grad_norm": 0.0,
+      "learning_rate": 1.8819654341316344e-05,
+      "loss": 0.9026,
+      "step": 589
+    },
+    {
+      "epoch": 0.1820144994601265,
+      "grad_norm": 0.0,
+      "learning_rate": 1.8814938905716778e-05,
+      "loss": 0.9464,
+      "step": 590
+    },
+    {
+      "epoch": 0.18232299861175383,
+      "grad_norm": 0.0,
+      "learning_rate": 1.8810214663071428e-05,
+      "loss": 0.9162,
+      "step": 591
+    },
+    {
+      "epoch": 0.18263149776338114,
+      "grad_norm": 0.0,
+      "learning_rate": 1.8805481618100314e-05,
+      "loss": 1.0065,
+      "step": 592
+    },
+    {
+      "epoch": 0.18293999691500848,
+      "grad_norm": 0.0,
+      "learning_rate": 1.8800739775532238e-05,
+      "loss": 0.92,
+      "step": 593
+    },
+    {
+      "epoch": 0.18324849606663582,
+      "grad_norm": 0.0,
+      "learning_rate": 1.8795989140104797e-05,
+      "loss": 0.936,
+      "step": 594
+    },
+    {
+      "epoch": 0.18355699521826316,
+      "grad_norm": 0.0,
+      "learning_rate": 1.8791229716564374e-05,
+      "loss": 0.9698,
+      "step": 595
+    },
+    {
+      "epoch": 0.18386549436989047,
+      "grad_norm": 0.0,
+      "learning_rate": 1.8786461509666135e-05,
+      "loss": 0.9489,
+      "step": 596
+    },
+    {
+      "epoch": 0.1841739935215178,
+      "grad_norm": 0.0,
+      "learning_rate": 1.8781684524174006e-05,
+      "loss": 0.9536,
+      "step": 597
+    },
+    {
+      "epoch": 0.18448249267314515,
+      "grad_norm": 0.0,
+      "learning_rate": 1.8776898764860707e-05,
+      "loss": 0.9249,
+      "step": 598
+    },
+    {
+      "epoch": 0.1847909918247725,
+      "grad_norm": 0.0,
+      "learning_rate": 1.8772104236507703e-05,
+      "loss": 0.9112,
+      "step": 599
+    },
+    {
+      "epoch": 0.18509949097639983,
+      "grad_norm": 0.0,
+      "learning_rate": 1.8767300943905238e-05,
+      "loss": 0.9114,
+      "step": 600
+    },
+    {
+      "epoch": 0.18540799012802714,
+      "grad_norm": 0.0,
+      "learning_rate": 1.8762488891852296e-05,
+      "loss": 0.9409,
+      "step": 601
+    },
+    {
+      "epoch": 0.18571648927965448,
+      "grad_norm": 0.0,
+      "learning_rate": 1.8757668085156628e-05,
+      "loss": 0.9692,
+      "step": 602
+    },
+    {
+      "epoch": 0.18602498843128182,
+      "grad_norm": 0.0,
+      "learning_rate": 1.875283852863471e-05,
+      "loss": 0.8082,
+      "step": 603
+    },
+    {
+      "epoch": 0.18633348758290916,
+      "grad_norm": 0.0,
+      "learning_rate": 1.874800022711179e-05,
+      "loss": 1.0027,
+      "step": 604
+    },
+    {
+      "epoch": 0.18664198673453647,
+      "grad_norm": 0.0,
+      "learning_rate": 1.8743153185421826e-05,
+      "loss": 0.9285,
+      "step": 605
+    },
+    {
+      "epoch": 0.1869504858861638,
+      "grad_norm": 0.0,
+      "learning_rate": 1.8738297408407525e-05,
+      "loss": 0.867,
+      "step": 606
+    },
+    {
+      "epoch": 0.18725898503779115,
+      "grad_norm": 0.0,
+      "learning_rate": 1.8733432900920316e-05,
+      "loss": 0.9368,
+      "step": 607
+    },
+    {
+      "epoch": 0.1875674841894185,
+      "grad_norm": 0.0,
+      "learning_rate": 1.872855966782035e-05,
+      "loss": 1.0111,
+      "step": 608
+    },
+    {
+      "epoch": 0.1878759833410458,
+      "grad_norm": 0.0,
+      "learning_rate": 1.8723677713976494e-05,
+      "loss": 0.9438,
+      "step": 609
+    },
+    {
+      "epoch": 0.18818448249267314,
+      "grad_norm": 0.0,
+      "learning_rate": 1.8718787044266334e-05,
+      "loss": 0.9616,
+      "step": 610
+    },
+    {
+      "epoch": 0.18849298164430048,
+      "grad_norm": 0.0,
+      "learning_rate": 1.8713887663576156e-05,
+      "loss": 1.1784,
+      "step": 611
+    },
+    {
+      "epoch": 0.18880148079592782,
+      "grad_norm": 0.0,
+      "learning_rate": 1.8708979576800962e-05,
+      "loss": 0.9202,
+      "step": 612
+    },
+    {
+      "epoch": 0.18910997994755516,
+      "grad_norm": 0.0,
+      "learning_rate": 1.8704062788844435e-05,
+      "loss": 0.9447,
+      "step": 613
+    },
+    {
+      "epoch": 0.18941847909918247,
+      "grad_norm": 0.0,
+      "learning_rate": 1.8699137304618963e-05,
+      "loss": 0.8955,
+      "step": 614
+    },
+    {
+      "epoch": 0.1897269782508098,
+      "grad_norm": 0.0,
+      "learning_rate": 1.869420312904562e-05,
+      "loss": 0.9599,
+      "step": 615
+    },
+    {
+      "epoch": 0.19003547740243715,
+      "grad_norm": 0.0,
+      "learning_rate": 1.868926026705417e-05,
+      "loss": 0.8844,
+      "step": 616
+    },
+    {
+      "epoch": 0.1903439765540645,
+      "grad_norm": 0.0,
+      "learning_rate": 1.8684308723583037e-05,
+      "loss": 0.8832,
+      "step": 617
+    },
+    {
+      "epoch": 0.1906524757056918,
+      "grad_norm": 0.0,
+      "learning_rate": 1.8679348503579336e-05,
+      "loss": 0.9653,
+      "step": 618
+    },
+    {
+      "epoch": 0.19096097485731914,
+      "grad_norm": 0.0,
+      "learning_rate": 1.867437961199885e-05,
+      "loss": 0.8365,
+      "step": 619
+    },
+    {
+      "epoch": 0.19126947400894648,
+      "grad_norm": 0.0,
+      "learning_rate": 1.8669402053806016e-05,
+      "loss": 0.8715,
+      "step": 620
+    },
+    {
+      "epoch": 0.19157797316057382,
+      "grad_norm": 0.0,
+      "learning_rate": 1.8664415833973937e-05,
+      "loss": 0.9752,
+      "step": 621
+    },
+    {
+      "epoch": 0.19188647231220113,
+      "grad_norm": 0.0,
+      "learning_rate": 1.8659420957484367e-05,
+      "loss": 0.9336,
+      "step": 622
+    },
+    {
+      "epoch": 0.19219497146382847,
+      "grad_norm": 0.0,
+      "learning_rate": 1.865441742932771e-05,
+      "loss": 0.9283,
+      "step": 623
+    },
+    {
+      "epoch": 0.1925034706154558,
+      "grad_norm": 0.0,
+      "learning_rate": 1.8649405254503018e-05,
+      "loss": 0.879,
+      "step": 624
+    },
+    {
+      "epoch": 0.19281196976708315,
+      "grad_norm": 0.0,
+      "learning_rate": 1.864438443801797e-05,
+      "loss": 1.036,
+      "step": 625
+    },
+    {
+      "epoch": 0.1931204689187105,
+      "grad_norm": 0.0,
+      "learning_rate": 1.863935498488889e-05,
+      "loss": 1.0019,
+      "step": 626
+    },
+    {
+      "epoch": 0.1934289680703378,
+      "grad_norm": 0.0,
+      "learning_rate": 1.8634316900140733e-05,
+      "loss": 0.9301,
+      "step": 627
+    },
+    {
+      "epoch": 0.19373746722196514,
+      "grad_norm": 0.0,
+      "learning_rate": 1.862927018880707e-05,
+      "loss": 1.1868,
+      "step": 628
+    },
+    {
+      "epoch": 0.19404596637359248,
+      "grad_norm": 0.0,
+      "learning_rate": 1.862421485593009e-05,
+      "loss": 1.0311,
+      "step": 629
+    },
+    {
+      "epoch": 0.19435446552521982,
+      "grad_norm": 0.0,
+      "learning_rate": 1.86191509065606e-05,
+      "loss": 0.959,
+      "step": 630
+    },
+    {
+      "epoch": 0.19466296467684713,
+      "grad_norm": 0.0,
+      "learning_rate": 1.8614078345758014e-05,
+      "loss": 0.8862,
+      "step": 631
+    },
+    {
+      "epoch": 0.19497146382847447,
+      "grad_norm": 0.0,
+      "learning_rate": 1.8608997178590354e-05,
+      "loss": 0.8945,
+      "step": 632
+    },
+    {
+      "epoch": 0.1952799629801018,
+      "grad_norm": 0.0,
+      "learning_rate": 1.8603907410134233e-05,
+      "loss": 0.889,
+      "step": 633
+    },
+    {
+      "epoch": 0.19558846213172915,
+      "grad_norm": 0.0,
+      "learning_rate": 1.8598809045474868e-05,
+      "loss": 0.8235,
+      "step": 634
+    },
+    {
+      "epoch": 0.19589696128335646,
+      "grad_norm": 0.0,
+      "learning_rate": 1.859370208970605e-05,
+      "loss": 0.8782,
+      "step": 635
+    },
+    {
+      "epoch": 0.1962054604349838,
+      "grad_norm": 0.0,
+      "learning_rate": 1.8588586547930176e-05,
+      "loss": 0.9183,
+      "step": 636
+    },
+    {
+      "epoch": 0.19651395958661114,
+      "grad_norm": 0.0,
+      "learning_rate": 1.8583462425258192e-05,
+      "loss": 0.866,
+      "step": 637
+    },
+    {
+      "epoch": 0.19682245873823848,
+      "grad_norm": 0.0,
+      "learning_rate": 1.857832972680964e-05,
+      "loss": 0.9157,
+      "step": 638
+    },
+    {
+      "epoch": 0.19713095788986582,
+      "grad_norm": 0.0,
+      "learning_rate": 1.8573188457712624e-05,
+      "loss": 0.8942,
+      "step": 639
+    },
+    {
+      "epoch": 0.19743945704149313,
+      "grad_norm": 0.0,
+      "learning_rate": 1.8568038623103802e-05,
+      "loss": 0.95,
+      "step": 640
+    },
+    {
+      "epoch": 0.19774795619312047,
+      "grad_norm": 0.0,
+      "learning_rate": 1.8562880228128402e-05,
+      "loss": 0.9938,
+      "step": 641
+    },
+    {
+      "epoch": 0.1980564553447478,
+      "grad_norm": 0.0,
+      "learning_rate": 1.8557713277940203e-05,
+      "loss": 0.8951,
+      "step": 642
+    },
+    {
+      "epoch": 0.19836495449637515,
+      "grad_norm": 0.0,
+      "learning_rate": 1.8552537777701525e-05,
+      "loss": 0.8841,
+      "step": 643
+    },
+    {
+      "epoch": 0.19867345364800246,
+      "grad_norm": 0.0,
+      "learning_rate": 1.8547353732583234e-05,
+      "loss": 0.923,
+      "step": 644
+    },
+    {
+      "epoch": 0.1989819527996298,
+      "grad_norm": 0.0,
+      "learning_rate": 1.8542161147764735e-05,
+      "loss": 0.8871,
+      "step": 645
+    },
+    {
+      "epoch": 0.19929045195125714,
+      "grad_norm": 0.0,
+      "learning_rate": 1.853696002843396e-05,
+      "loss": 0.806,
+      "step": 646
+    },
+    {
+      "epoch": 0.19959895110288448,
+      "grad_norm": 0.0,
+      "learning_rate": 1.8531750379787375e-05,
+      "loss": 0.8171,
+      "step": 647
+    },
+    {
+      "epoch": 0.1999074502545118,
+      "grad_norm": 0.0,
+      "learning_rate": 1.852653220702996e-05,
+      "loss": 0.8989,
+      "step": 648
+    },
+    {
+      "epoch": 0.20021594940613913,
+      "grad_norm": 0.0,
+      "learning_rate": 1.8521305515375217e-05,
+      "loss": 0.8877,
+      "step": 649
+    },
+    {
+      "epoch": 0.20052444855776647,
+      "grad_norm": 0.0,
+      "learning_rate": 1.8516070310045157e-05,
+      "loss": 0.9185,
+      "step": 650
+    },
+    {
+      "epoch": 0.2008329477093938,
+      "grad_norm": 0.0,
+      "learning_rate": 1.85108265962703e-05,
+      "loss": 0.883,
+      "step": 651
+    },
+    {
+      "epoch": 0.20114144686102114,
+      "grad_norm": 0.0,
+      "learning_rate": 1.850557437928966e-05,
+      "loss": 0.8563,
+      "step": 652
+    },
+    {
+      "epoch": 0.20144994601264846,
+      "grad_norm": 0.0,
+      "learning_rate": 1.8500313664350758e-05,
+      "loss": 0.9672,
+      "step": 653
+    },
+    {
+      "epoch": 0.2017584451642758,
+      "grad_norm": 0.0,
+      "learning_rate": 1.849504445670959e-05,
+      "loss": 0.9432,
+      "step": 654
+    },
+    {
+      "epoch": 0.20206694431590314,
+      "grad_norm": 0.0,
+      "learning_rate": 1.848976676163065e-05,
+      "loss": 0.9075,
+      "step": 655
+    },
+    {
+      "epoch": 0.20237544346753047,
+      "grad_norm": 0.0,
+      "learning_rate": 1.8484480584386907e-05,
+      "loss": 0.9218,
+      "step": 656
+    },
+    {
+      "epoch": 0.2026839426191578,
+      "grad_norm": 0.0,
+      "learning_rate": 1.847918593025981e-05,
+      "loss": 0.849,
+      "step": 657
+    },
+    {
+      "epoch": 0.20299244177078513,
+      "grad_norm": 0.0,
+      "learning_rate": 1.847388280453926e-05,
+      "loss": 0.8514,
+      "step": 658
+    },
+    {
+      "epoch": 0.20330094092241247,
+      "grad_norm": 0.0,
+      "learning_rate": 1.846857121252364e-05,
+      "loss": 0.8227,
+      "step": 659
+    },
+    {
+      "epoch": 0.2036094400740398,
+      "grad_norm": 0.0,
+      "learning_rate": 1.8463251159519793e-05,
+      "loss": 0.8503,
+      "step": 660
+    },
+    {
+      "epoch": 0.20391793922566712,
+      "grad_norm": 0.0,
+      "learning_rate": 1.8457922650842998e-05,
+      "loss": 0.9469,
+      "step": 661
+    },
+    {
+      "epoch": 0.20422643837729446,
+      "grad_norm": 0.0,
+      "learning_rate": 1.8452585691817e-05,
+      "loss": 0.8618,
+      "step": 662
+    },
+    {
+      "epoch": 0.2045349375289218,
+      "grad_norm": 0.0,
+      "learning_rate": 1.8447240287773973e-05,
+      "loss": 0.9251,
+      "step": 663
+    },
+    {
+      "epoch": 0.20484343668054913,
+      "grad_norm": 0.0,
+      "learning_rate": 1.8441886444054534e-05,
+      "loss": 1.0748,
+      "step": 664
+    },
+    {
+      "epoch": 0.20515193583217647,
+      "grad_norm": 0.0,
+      "learning_rate": 1.8436524166007742e-05,
+      "loss": 0.9532,
+      "step": 665
+    },
+    {
+      "epoch": 0.20546043498380379,
+      "grad_norm": 0.0,
+      "learning_rate": 1.8431153458991066e-05,
+      "loss": 0.9112,
+      "step": 666
+    },
+    {
+      "epoch": 0.20576893413543113,
+      "grad_norm": 0.0,
+      "learning_rate": 1.8425774328370402e-05,
+      "loss": 0.9206,
+      "step": 667
+    },
+    {
+      "epoch": 0.20607743328705846,
+      "grad_norm": 0.0,
+      "learning_rate": 1.8420386779520073e-05,
+      "loss": 0.8469,
+      "step": 668
+    },
+    {
+      "epoch": 0.2063859324386858,
+      "grad_norm": 0.0,
+      "learning_rate": 1.8414990817822794e-05,
+      "loss": 0.8662,
+      "step": 669
+    },
+    {
+      "epoch": 0.20669443159031312,
+      "grad_norm": 0.0,
+      "learning_rate": 1.84095864486697e-05,
+      "loss": 0.9517,
+      "step": 670
+    },
+    {
+      "epoch": 0.20700293074194046,
+      "grad_norm": 0.0,
+      "learning_rate": 1.840417367746032e-05,
+      "loss": 1.2155,
+      "step": 671
+    },
+    {
+      "epoch": 0.2073114298935678,
+      "grad_norm": 0.0,
+      "learning_rate": 1.839875250960258e-05,
+      "loss": 0.882,
+      "step": 672
+    },
+    {
+      "epoch": 0.20761992904519513,
+      "grad_norm": 0.0,
+      "learning_rate": 1.8393322950512793e-05,
+      "loss": 0.9063,
+      "step": 673
+    },
+    {
+      "epoch": 0.20792842819682245,
+      "grad_norm": 0.0,
+      "learning_rate": 1.8387885005615652e-05,
+      "loss": 0.9307,
+      "step": 674
+    },
+    {
+      "epoch": 0.20823692734844979,
+      "grad_norm": 0.0,
+      "learning_rate": 1.838243868034424e-05,
+      "loss": 0.9967,
+      "step": 675
+    },
+    {
+      "epoch": 0.20854542650007712,
+      "grad_norm": 0.0,
+      "learning_rate": 1.837698398014e-05,
+      "loss": 0.885,
+      "step": 676
+    },
+    {
+      "epoch": 0.20885392565170446,
+      "grad_norm": 0.0,
+      "learning_rate": 1.837152091045275e-05,
+      "loss": 0.923,
+      "step": 677
+    },
+    {
+      "epoch": 0.2091624248033318,
+      "grad_norm": 0.0,
+      "learning_rate": 1.8366049476740667e-05,
+      "loss": 0.8932,
+      "step": 678
+    },
+    {
+      "epoch": 0.20947092395495912,
+      "grad_norm": 0.0,
+      "learning_rate": 1.8360569684470286e-05,
+      "loss": 0.8664,
+      "step": 679
+    },
+    {
+      "epoch": 0.20977942310658645,
+      "grad_norm": 0.0,
+      "learning_rate": 1.8355081539116492e-05,
+      "loss": 0.9405,
+      "step": 680
+    },
+    {
+      "epoch": 0.2100879222582138,
+      "grad_norm": 0.0,
+      "learning_rate": 1.8349585046162517e-05,
+      "loss": 0.9038,
+      "step": 681
+    },
+    {
+      "epoch": 0.21039642140984113,
+      "grad_norm": 0.0,
+      "learning_rate": 1.8344080211099934e-05,
+      "loss": 0.986,
+      "step": 682
+    },
+    {
+      "epoch": 0.21070492056146845,
+      "grad_norm": 0.0,
+      "learning_rate": 1.8338567039428642e-05,
+      "loss": 0.871,
+      "step": 683
+    },
+    {
+      "epoch": 0.21101341971309578,
+      "grad_norm": 0.0,
+      "learning_rate": 1.8333045536656882e-05,
+      "loss": 0.9415,
+      "step": 684
+    },
+    {
+      "epoch": 0.21132191886472312,
+      "grad_norm": 0.0,
+      "learning_rate": 1.8327515708301217e-05,
+      "loss": 0.8756,
+      "step": 685
+    },
+    {
+      "epoch": 0.21163041801635046,
+      "grad_norm": 0.0,
+      "learning_rate": 1.8321977559886514e-05,
+      "loss": 1.0333,
+      "step": 686
+    },
+    {
+      "epoch": 0.21193891716797778,
+      "grad_norm": 0.0,
+      "learning_rate": 1.8316431096945965e-05,
+      "loss": 0.892,
+      "step": 687
+    },
+    {
+      "epoch": 0.21224741631960511,
+      "grad_norm": 0.0,
+      "learning_rate": 1.831087632502107e-05,
+      "loss": 0.9391,
+      "step": 688
+    },
+    {
+      "epoch": 0.21255591547123245,
+      "grad_norm": 0.0,
+      "learning_rate": 1.8305313249661628e-05,
+      "loss": 1.0055,
+      "step": 689
+    },
+    {
+      "epoch": 0.2128644146228598,
+      "grad_norm": 0.0,
+      "learning_rate": 1.8299741876425725e-05,
+      "loss": 0.9493,
+      "step": 690
+    },
+    {
+      "epoch": 0.21317291377448713,
+      "grad_norm": 0.0,
+      "learning_rate": 1.8294162210879753e-05,
+      "loss": 0.9286,
+      "step": 691
+    },
+    {
+      "epoch": 0.21348141292611444,
+      "grad_norm": 0.0,
+      "learning_rate": 1.8288574258598378e-05,
+      "loss": 0.8777,
+      "step": 692
+    },
+    {
+      "epoch": 0.21378991207774178,
+      "grad_norm": 0.0,
+      "learning_rate": 1.8282978025164553e-05,
+      "loss": 0.9229,
+      "step": 693
+    },
+    {
+      "epoch": 0.21409841122936912,
+      "grad_norm": 0.0,
+      "learning_rate": 1.8277373516169493e-05,
+      "loss": 0.907,
+      "step": 694
+    },
+    {
+      "epoch": 0.21440691038099646,
+      "grad_norm": 0.0,
+      "learning_rate": 1.8271760737212697e-05,
+      "loss": 0.9317,
+      "step": 695
+    },
+    {
+      "epoch": 0.21471540953262377,
+      "grad_norm": 0.0,
+      "learning_rate": 1.8266139693901914e-05,
+      "loss": 0.865,
+      "step": 696
+    },
+    {
+      "epoch": 0.2150239086842511,
+      "grad_norm": 0.0,
+      "learning_rate": 1.826051039185315e-05,
+      "loss": 0.9932,
+      "step": 697
+    },
+    {
+      "epoch": 0.21533240783587845,
+      "grad_norm": 0.0,
+      "learning_rate": 1.8254872836690672e-05,
+      "loss": 0.8622,
+      "step": 698
+    },
+    {
+      "epoch": 0.2156409069875058,
+      "grad_norm": 0.0,
+      "learning_rate": 1.8249227034046986e-05,
+      "loss": 0.9023,
+      "step": 699
+    },
+    {
+      "epoch": 0.2159494061391331,
+      "grad_norm": 0.0,
+      "learning_rate": 1.824357298956284e-05,
+      "loss": 0.8664,
+      "step": 700
+    },
+    {
+      "epoch": 0.21625790529076044,
+      "grad_norm": 0.0,
+      "learning_rate": 1.823791070888721e-05,
+      "loss": 0.9108,
+      "step": 701
+    },
+    {
+      "epoch": 0.21656640444238778,
+      "grad_norm": 0.0,
+      "learning_rate": 1.8232240197677318e-05,
+      "loss": 0.9412,
+      "step": 702
+    },
+    {
+      "epoch": 0.21687490359401512,
+      "grad_norm": 0.0,
+      "learning_rate": 1.8226561461598583e-05,
+      "loss": 0.9271,
+      "step": 703
+    },
+    {
+      "epoch": 0.21718340274564246,
+      "grad_norm": 0.0,
+      "learning_rate": 1.8220874506324667e-05,
+      "loss": 0.9417,
+      "step": 704
+    },
+    {
+      "epoch": 0.21749190189726977,
+      "grad_norm": 0.0,
+      "learning_rate": 1.8215179337537433e-05,
+      "loss": 0.9904,
+      "step": 705
+    },
+    {
+      "epoch": 0.2178004010488971,
+      "grad_norm": 0.0,
+      "learning_rate": 1.8209475960926946e-05,
+      "loss": 0.8971,
+      "step": 706
+    },
+    {
+      "epoch": 0.21810890020052445,
+      "grad_norm": 0.0,
+      "learning_rate": 1.8203764382191476e-05,
+      "loss": 0.9891,
+      "step": 707
+    },
+    {
+      "epoch": 0.2184173993521518,
+      "grad_norm": 0.0,
+      "learning_rate": 1.8198044607037486e-05,
+      "loss": 0.8612,
+      "step": 708
+    },
+    {
+      "epoch": 0.2187258985037791,
+      "grad_norm": 0.0,
+      "learning_rate": 1.8192316641179634e-05,
+      "loss": 0.865,
+      "step": 709
+    },
+    {
+      "epoch": 0.21903439765540644,
+      "grad_norm": 0.0,
+      "learning_rate": 1.8186580490340754e-05,
+      "loss": 0.9252,
+      "step": 710
+    },
+    {
+      "epoch": 0.21934289680703378,
+      "grad_norm": 0.0,
+      "learning_rate": 1.8180836160251863e-05,
+      "loss": 0.9649,
+      "step": 711
+    },
+    {
+      "epoch": 0.21965139595866112,
+      "grad_norm": 0.0,
+      "learning_rate": 1.8175083656652143e-05,
+      "loss": 0.7586,
+      "step": 712
+    },
+    {
+      "epoch": 0.21995989511028843,
+      "grad_norm": 0.0,
+      "learning_rate": 1.816932298528895e-05,
+      "loss": 0.8505,
+      "step": 713
+    },
+    {
+      "epoch": 0.22026839426191577,
+      "grad_norm": 0.0,
+      "learning_rate": 1.8163554151917796e-05,
+      "loss": 0.8742,
+      "step": 714
+    },
+    {
+      "epoch": 0.2205768934135431,
+      "grad_norm": 0.0,
+      "learning_rate": 1.815777716230235e-05,
+      "loss": 0.8806,
+      "step": 715
+    },
+    {
+      "epoch": 0.22088539256517045,
+      "grad_norm": 0.0,
+      "learning_rate": 1.8151992022214425e-05,
+      "loss": 0.8527,
+      "step": 716
+    },
+    {
+      "epoch": 0.2211938917167978,
+      "grad_norm": 0.0,
+      "learning_rate": 1.8146198737433993e-05,
+      "loss": 0.9671,
+      "step": 717
+    },
+    {
+      "epoch": 0.2215023908684251,
+      "grad_norm": 0.0,
+      "learning_rate": 1.8140397313749134e-05,
+      "loss": 0.9928,
+      "step": 718
+    },
+    {
+      "epoch": 0.22181089002005244,
+      "grad_norm": 0.0,
+      "learning_rate": 1.8134587756956084e-05,
+      "loss": 1.0005,
+      "step": 719
+    },
+    {
+      "epoch": 0.22211938917167978,
+      "grad_norm": 0.0,
+      "learning_rate": 1.8128770072859202e-05,
+      "loss": 0.8197,
+      "step": 720
+    },
+    {
+      "epoch": 0.22242788832330712,
+      "grad_norm": 0.0,
+      "learning_rate": 1.812294426727096e-05,
+      "loss": 0.8596,
+      "step": 721
+    },
+    {
+      "epoch": 0.22273638747493443,
+      "grad_norm": 0.0,
+      "learning_rate": 1.8117110346011946e-05,
+      "loss": 0.947,
+      "step": 722
+    },
+    {
+      "epoch": 0.22304488662656177,
+      "grad_norm": 0.0,
+      "learning_rate": 1.8111268314910857e-05,
+      "loss": 0.9231,
+      "step": 723
+    },
+    {
+      "epoch": 0.2233533857781891,
+      "grad_norm": 0.0,
+      "learning_rate": 1.81054181798045e-05,
+      "loss": 0.9239,
+      "step": 724
+    },
+    {
+      "epoch": 0.22366188492981645,
+      "grad_norm": 0.0,
+      "learning_rate": 1.809955994653776e-05,
+      "loss": 0.9616,
+      "step": 725
+    },
+    {
+      "epoch": 0.22397038408144376,
+      "grad_norm": 0.0,
+      "learning_rate": 1.809369362096363e-05,
+      "loss": 0.8898,
+      "step": 726
+    },
+    {
+      "epoch": 0.2242788832330711,
+      "grad_norm": 0.0,
+      "learning_rate": 1.8087819208943186e-05,
+      "loss": 0.8848,
+      "step": 727
+    },
+    {
+      "epoch": 0.22458738238469844,
+      "grad_norm": 0.0,
+      "learning_rate": 1.8081936716345574e-05,
+      "loss": 0.8048,
+      "step": 728
+    },
+    {
+      "epoch": 0.22489588153632578,
+      "grad_norm": 0.0,
+      "learning_rate": 1.8076046149048024e-05,
+      "loss": 0.8379,
+      "step": 729
+    },
+    {
+      "epoch": 0.22520438068795312,
+      "grad_norm": 0.0,
+      "learning_rate": 1.8070147512935828e-05,
+      "loss": 0.8821,
+      "step": 730
+    },
+    {
+      "epoch": 0.22551287983958043,
+      "grad_norm": 0.0,
+      "learning_rate": 1.806424081390234e-05,
+      "loss": 0.8625,
+      "step": 731
+    },
+    {
+      "epoch": 0.22582137899120777,
+      "grad_norm": 0.0,
+      "learning_rate": 1.8058326057848966e-05,
+      "loss": 0.9501,
+      "step": 732
+    },
+    {
+      "epoch": 0.2261298781428351,
+      "grad_norm": 0.0,
+      "learning_rate": 1.8052403250685172e-05,
+      "loss": 0.8423,
+      "step": 733
+    },
+    {
+      "epoch": 0.22643837729446245,
+      "grad_norm": 0.0,
+      "learning_rate": 1.804647239832846e-05,
+      "loss": 0.884,
+      "step": 734
+    },
+    {
+      "epoch": 0.22674687644608976,
+      "grad_norm": 0.0,
+      "learning_rate": 1.804053350670437e-05,
+      "loss": 0.9397,
+      "step": 735
+    },
+    {
+      "epoch": 0.2270553755977171,
+      "grad_norm": 0.0,
+      "learning_rate": 1.8034586581746474e-05,
+      "loss": 0.9681,
+      "step": 736
+    },
+    {
+      "epoch": 0.22736387474934444,
+      "grad_norm": 0.0,
+      "learning_rate": 1.8028631629396377e-05,
+      "loss": 0.9377,
+      "step": 737
+    },
+    {
+      "epoch": 0.22767237390097178,
+      "grad_norm": 0.0,
+      "learning_rate": 1.8022668655603696e-05,
+      "loss": 0.9687,
+      "step": 738
+    },
+    {
+      "epoch": 0.2279808730525991,
+      "grad_norm": 0.0,
+      "learning_rate": 1.8016697666326066e-05,
+      "loss": 0.9338,
+      "step": 739
+    },
+    {
+      "epoch": 0.22828937220422643,
+      "grad_norm": 0.0,
+      "learning_rate": 1.801071866752913e-05,
+      "loss": 0.8938,
+      "step": 740
+    },
+    {
+      "epoch": 0.22859787135585377,
+      "grad_norm": 0.0,
+      "learning_rate": 1.8004731665186532e-05,
+      "loss": 0.8316,
+      "step": 741
+    },
+    {
+      "epoch": 0.2289063705074811,
+      "grad_norm": 0.0,
+      "learning_rate": 1.7998736665279914e-05,
+      "loss": 0.9167,
+      "step": 742
+    },
+    {
+      "epoch": 0.22921486965910845,
+      "grad_norm": 0.0,
+      "learning_rate": 1.799273367379891e-05,
+      "loss": 0.9561,
+      "step": 743
+    },
+    {
+      "epoch": 0.22952336881073576,
+      "grad_norm": 0.0,
+      "learning_rate": 1.7986722696741132e-05,
+      "loss": 0.9086,
+      "step": 744
+    },
+    {
+      "epoch": 0.2298318679623631,
+      "grad_norm": 0.0,
+      "learning_rate": 1.798070374011218e-05,
+      "loss": 0.8398,
+      "step": 745
+    },
+    {
+      "epoch": 0.23014036711399044,
+      "grad_norm": 0.0,
+      "learning_rate": 1.7974676809925613e-05,
+      "loss": 0.8583,
+      "step": 746
+    },
+    {
+      "epoch": 0.23044886626561778,
+      "grad_norm": 0.0,
+      "learning_rate": 1.7968641912202973e-05,
+      "loss": 0.9205,
+      "step": 747
+    },
+    {
+      "epoch": 0.2307573654172451,
+      "grad_norm": 0.0,
+      "learning_rate": 1.7962599052973746e-05,
+      "loss": 0.9376,
+      "step": 748
+    },
+    {
+      "epoch": 0.23106586456887243,
+      "grad_norm": 0.0,
+      "learning_rate": 1.7956548238275387e-05,
+      "loss": 0.9171,
+      "step": 749
+    },
+    {
+      "epoch": 0.23137436372049977,
+      "grad_norm": 0.0,
+      "learning_rate": 1.7950489474153293e-05,
+      "loss": 0.8935,
+      "step": 750
+    },
+    {
+      "epoch": 0.2316828628721271,
+      "grad_norm": 0.0,
+      "learning_rate": 1.7944422766660797e-05,
+      "loss": 1.0081,
+      "step": 751
+    },
+    {
+      "epoch": 0.23199136202375442,
+      "grad_norm": 0.0,
+      "learning_rate": 1.7938348121859183e-05,
+      "loss": 0.8754,
+      "step": 752
+    },
+    {
+      "epoch": 0.23229986117538176,
+      "grad_norm": 0.0,
+      "learning_rate": 1.7932265545817645e-05,
+      "loss": 0.8547,
+      "step": 753
+    },
+    {
+      "epoch": 0.2326083603270091,
+      "grad_norm": 0.0,
+      "learning_rate": 1.792617504461332e-05,
+      "loss": 0.8784,
+      "step": 754
+    },
+    {
+      "epoch": 0.23291685947863644,
+      "grad_norm": 0.0,
+      "learning_rate": 1.7920076624331254e-05,
+      "loss": 0.9178,
+      "step": 755
+    },
+    {
+      "epoch": 0.23322535863026378,
+      "grad_norm": 0.0,
+      "learning_rate": 1.791397029106441e-05,
+      "loss": 0.8957,
+      "step": 756
+    },
+    {
+      "epoch": 0.2335338577818911,
+      "grad_norm": 0.0,
+      "learning_rate": 1.7907856050913644e-05,
+      "loss": 1.1664,
+      "step": 757
+    },
+    {
+      "epoch": 0.23384235693351843,
+      "grad_norm": 0.0,
+      "learning_rate": 1.7901733909987735e-05,
+      "loss": 0.8931,
+      "step": 758
+    },
+    {
+      "epoch": 0.23415085608514577,
+      "grad_norm": 0.0,
+      "learning_rate": 1.7895603874403326e-05,
+      "loss": 0.9065,
+      "step": 759
+    },
+    {
+      "epoch": 0.2344593552367731,
+      "grad_norm": 0.0,
+      "learning_rate": 1.788946595028498e-05,
+      "loss": 0.8369,
+      "step": 760
+    },
+    {
+      "epoch": 0.23476785438840042,
+      "grad_norm": 0.0,
+      "learning_rate": 1.7883320143765113e-05,
+      "loss": 0.9097,
+      "step": 761
+    },
+    {
+      "epoch": 0.23507635354002776,
+      "grad_norm": 0.0,
+      "learning_rate": 1.787716646098403e-05,
+      "loss": 0.9211,
+      "step": 762
+    },
+    {
+      "epoch": 0.2353848526916551,
+      "grad_norm": 0.0,
+      "learning_rate": 1.787100490808991e-05,
+      "loss": 0.9252,
+      "step": 763
+    },
+    {
+      "epoch": 0.23569335184328244,
+      "grad_norm": 0.0,
+      "learning_rate": 1.7864835491238785e-05,
+      "loss": 0.8593,
+      "step": 764
+    },
+    {
+      "epoch": 0.23600185099490975,
+      "grad_norm": 0.0,
+      "learning_rate": 1.7858658216594545e-05,
+      "loss": 0.9216,
+      "step": 765
+    },
+    {
+      "epoch": 0.2363103501465371,
+      "grad_norm": 0.0,
+      "learning_rate": 1.7852473090328937e-05,
+      "loss": 0.8814,
+      "step": 766
+    },
+    {
+      "epoch": 0.23661884929816443,
+      "grad_norm": 0.0,
+      "learning_rate": 1.7846280118621547e-05,
+      "loss": 0.9103,
+      "step": 767
+    },
+    {
+      "epoch": 0.23692734844979177,
+      "grad_norm": 0.0,
+      "learning_rate": 1.7840079307659803e-05,
+      "loss": 0.8907,
+      "step": 768
+    },
+    {
+      "epoch": 0.2372358476014191,
+      "grad_norm": 0.0,
+      "learning_rate": 1.783387066363896e-05,
+      "loss": 0.8361,
+      "step": 769
+    },
+    {
+      "epoch": 0.23754434675304642,
+      "grad_norm": 0.0,
+      "learning_rate": 1.782765419276211e-05,
+      "loss": 0.8302,
+      "step": 770
+    },
+    {
+      "epoch": 0.23785284590467376,
+      "grad_norm": 0.0,
+      "learning_rate": 1.782142990124015e-05,
+      "loss": 0.9117,
+      "step": 771
+    },
+    {
+      "epoch": 0.2381613450563011,
+      "grad_norm": 0.0,
+      "learning_rate": 1.78151977952918e-05,
+      "loss": 0.952,
+      "step": 772
+    },
+    {
+      "epoch": 0.23846984420792844,
+      "grad_norm": 0.0,
+      "learning_rate": 1.7808957881143588e-05,
+      "loss": 0.8969,
+      "step": 773
+    },
+    {
+      "epoch": 0.23877834335955575,
+      "grad_norm": 0.0,
+      "learning_rate": 1.780271016502984e-05,
+      "loss": 0.8482,
+      "step": 774
+    },
+    {
+      "epoch": 0.2390868425111831,
+      "grad_norm": 0.0,
+      "learning_rate": 1.7796454653192675e-05,
+      "loss": 0.7801,
+      "step": 775
+    },
+    {
+      "epoch": 0.23939534166281043,
+      "grad_norm": 0.0,
+      "learning_rate": 1.7790191351882006e-05,
+      "loss": 0.9085,
+      "step": 776
+    },
+    {
+      "epoch": 0.23970384081443777,
+      "grad_norm": 0.0,
+      "learning_rate": 1.7783920267355527e-05,
+      "loss": 0.8567,
+      "step": 777
+    },
+    {
+      "epoch": 0.24001233996606508,
+      "grad_norm": 0.0,
+      "learning_rate": 1.7777641405878706e-05,
+      "loss": 0.8105,
+      "step": 778
+    },
+    {
+      "epoch": 0.24032083911769242,
+      "grad_norm": 0.0,
+      "learning_rate": 1.777135477372478e-05,
+      "loss": 0.8481,
+      "step": 779
+    },
+    {
+      "epoch": 0.24062933826931976,
+      "grad_norm": 0.0,
+      "learning_rate": 1.7765060377174753e-05,
+      "loss": 0.8734,
+      "step": 780
+    },
+    {
+      "epoch": 0.2409378374209471,
+      "grad_norm": 0.0,
+      "learning_rate": 1.7758758222517387e-05,
+      "loss": 0.8233,
+      "step": 781
+    },
+    {
+      "epoch": 0.24124633657257444,
+      "grad_norm": 0.0,
+      "learning_rate": 1.7752448316049197e-05,
+      "loss": 0.8362,
+      "step": 782
+    },
+    {
+      "epoch": 0.24155483572420175,
+      "grad_norm": 0.0,
+      "learning_rate": 1.774613066407443e-05,
+      "loss": 0.9261,
+      "step": 783
+    },
+    {
+      "epoch": 0.2418633348758291,
+      "grad_norm": 0.0,
+      "learning_rate": 1.7739805272905087e-05,
+      "loss": 0.9551,
+      "step": 784
+    },
+    {
+      "epoch": 0.24217183402745643,
+      "grad_norm": 0.0,
+      "learning_rate": 1.7733472148860893e-05,
+      "loss": 0.8842,
+      "step": 785
+    },
+    {
+      "epoch": 0.24248033317908377,
+      "grad_norm": 0.0,
+      "learning_rate": 1.7727131298269306e-05,
+      "loss": 0.9109,
+      "step": 786
+    },
+    {
+      "epoch": 0.24278883233071108,
+      "grad_norm": 0.0,
+      "learning_rate": 1.7720782727465494e-05,
+      "loss": 0.845,
+      "step": 787
+    },
+    {
+      "epoch": 0.24309733148233842,
+      "grad_norm": 0.0,
+      "learning_rate": 1.7714426442792344e-05,
+      "loss": 0.9728,
+      "step": 788
+    },
+    {
+      "epoch": 0.24340583063396576,
+      "grad_norm": 0.0,
+      "learning_rate": 1.770806245060045e-05,
+      "loss": 0.9432,
+      "step": 789
+    },
+    {
+      "epoch": 0.2437143297855931,
+      "grad_norm": 0.0,
+      "learning_rate": 1.770169075724811e-05,
+      "loss": 0.7779,
+      "step": 790
+    },
+    {
+      "epoch": 0.2440228289372204,
+      "grad_norm": 0.0,
+      "learning_rate": 1.76953113691013e-05,
+      "loss": 0.7957,
+      "step": 791
+    },
+    {
+      "epoch": 0.24433132808884775,
+      "grad_norm": 0.0,
+      "learning_rate": 1.7688924292533706e-05,
+      "loss": 0.8393,
+      "step": 792
+    },
+    {
+      "epoch": 0.2446398272404751,
+      "grad_norm": 0.0,
+      "learning_rate": 1.768252953392668e-05,
+      "loss": 0.8052,
+      "step": 793
+    },
+    {
+      "epoch": 0.24494832639210243,
+      "grad_norm": 0.0,
+      "learning_rate": 1.7676127099669265e-05,
+      "loss": 0.7727,
+      "step": 794
+    },
+    {
+      "epoch": 0.24525682554372977,
+      "grad_norm": 0.0,
+      "learning_rate": 1.7669716996158148e-05,
+      "loss": 0.8535,
+      "step": 795
+    },
+    {
+      "epoch": 0.24556532469535708,
+      "grad_norm": 0.0,
+      "learning_rate": 1.76632992297977e-05,
+      "loss": 0.8624,
+      "step": 796
+    },
+    {
+      "epoch": 0.24587382384698442,
+      "grad_norm": 0.0,
+      "learning_rate": 1.765687380699994e-05,
+      "loss": 0.8997,
+      "step": 797
+    },
+    {
+      "epoch": 0.24618232299861176,
+      "grad_norm": 0.0,
+      "learning_rate": 1.765044073418454e-05,
+      "loss": 0.8142,
+      "step": 798
+    },
+    {
+      "epoch": 0.2464908221502391,
+      "grad_norm": 0.0,
+      "learning_rate": 1.7644000017778807e-05,
+      "loss": 0.9157,
+      "step": 799
+    },
+    {
+      "epoch": 0.2467993213018664,
+      "grad_norm": 0.0,
+      "learning_rate": 1.7637551664217695e-05,
+      "loss": 0.8581,
+      "step": 800
+    },
+    {
+      "epoch": 0.24710782045349375,
+      "grad_norm": 0.0,
+      "learning_rate": 1.763109567994378e-05,
+      "loss": 0.9645,
+      "step": 801
+    },
+    {
+      "epoch": 0.2474163196051211,
+      "grad_norm": 0.0,
+      "learning_rate": 1.7624632071407268e-05,
+      "loss": 0.8924,
+      "step": 802
+    },
+    {
+      "epoch": 0.24772481875674843,
+      "grad_norm": 0.0,
+      "learning_rate": 1.7618160845065978e-05,
+      "loss": 0.919,
+      "step": 803
+    },
+    {
+      "epoch": 0.24803331790837574,
+      "grad_norm": 0.0,
+      "learning_rate": 1.7611682007385345e-05,
+      "loss": 0.889,
+      "step": 804
+    },
+    {
+      "epoch": 0.24834181706000308,
+      "grad_norm": 0.0,
+      "learning_rate": 1.76051955648384e-05,
+      "loss": 0.9173,
+      "step": 805
+    },
+    {
+      "epoch": 0.24865031621163042,
+      "grad_norm": 0.0,
+      "learning_rate": 1.7598701523905783e-05,
+      "loss": 0.9212,
+      "step": 806
+    },
+    {
+      "epoch": 0.24895881536325776,
+      "grad_norm": 0.0,
+      "learning_rate": 1.7592199891075714e-05,
+      "loss": 0.7934,
+      "step": 807
+    },
+    {
+      "epoch": 0.2492673145148851,
+      "grad_norm": 0.0,
+      "learning_rate": 1.758569067284401e-05,
+      "loss": 0.8689,
+      "step": 808
+    },
+    {
+      "epoch": 0.2495758136665124,
+      "grad_norm": 0.0,
+      "learning_rate": 1.7579173875714058e-05,
+      "loss": 0.8654,
+      "step": 809
+    },
+    {
+      "epoch": 0.24988431281813975,
+      "grad_norm": 0.0,
+      "learning_rate": 1.757264950619682e-05,
+      "loss": 0.8837,
+      "step": 810
+    },
+    {
+      "epoch": 0.25019281196976706,
+      "grad_norm": 0.0,
+      "learning_rate": 1.7566117570810822e-05,
+      "loss": 0.8707,
+      "step": 811
+    },
+    {
+      "epoch": 0.2505013111213944,
+      "grad_norm": 0.0,
+      "learning_rate": 1.7559578076082156e-05,
+      "loss": 0.9307,
+      "step": 812
+    },
+    {
+      "epoch": 0.25080981027302174,
+      "grad_norm": 0.0,
+      "learning_rate": 1.7553031028544452e-05,
+      "loss": 0.9386,
+      "step": 813
+    },
+    {
+      "epoch": 0.2511183094246491,
+      "grad_norm": 0.0,
+      "learning_rate": 1.7546476434738904e-05,
+      "loss": 0.8065,
+      "step": 814
+    },
+    {
+      "epoch": 0.2514268085762764,
+      "grad_norm": 0.0,
+      "learning_rate": 1.7539914301214233e-05,
+      "loss": 0.8347,
+      "step": 815
+    },
+    {
+      "epoch": 0.25173530772790376,
+      "grad_norm": 0.0,
+      "learning_rate": 1.7533344634526693e-05,
+      "loss": 0.8738,
+      "step": 816
+    },
+    {
+      "epoch": 0.2520438068795311,
+      "grad_norm": 0.0,
+      "learning_rate": 1.7526767441240075e-05,
+      "loss": 0.8786,
+      "step": 817
+    },
+    {
+      "epoch": 0.25235230603115844,
+      "grad_norm": 0.0,
+      "learning_rate": 1.7520182727925678e-05,
+      "loss": 0.8297,
+      "step": 818
+    },
+    {
+      "epoch": 0.2526608051827858,
+      "grad_norm": 0.0,
+      "learning_rate": 1.751359050116232e-05,
+      "loss": 0.9052,
+      "step": 819
+    },
+    {
+      "epoch": 0.25296930433441306,
+      "grad_norm": 0.0,
+      "learning_rate": 1.7506990767536326e-05,
+      "loss": 0.8696,
+      "step": 820
+    },
+    {
+      "epoch": 0.2532778034860404,
+      "grad_norm": 0.0,
+      "learning_rate": 1.750038353364152e-05,
+      "loss": 0.8608,
+      "step": 821
+    },
+    {
+      "epoch": 0.25358630263766774,
+      "grad_norm": 0.0,
+      "learning_rate": 1.749376880607922e-05,
+      "loss": 0.8843,
+      "step": 822
+    },
+    {
+      "epoch": 0.2538948017892951,
+      "grad_norm": 0.0,
+      "learning_rate": 1.748714659145823e-05,
+      "loss": 0.9389,
+      "step": 823
+    },
+    {
+      "epoch": 0.2542033009409224,
+      "grad_norm": 0.0,
+      "learning_rate": 1.7480516896394833e-05,
+      "loss": 0.9542,
+      "step": 824
+    },
+    {
+      "epoch": 0.25451180009254976,
+      "grad_norm": 0.0,
+      "learning_rate": 1.747387972751279e-05,
+      "loss": 0.9504,
+      "step": 825
+    },
+    {
+      "epoch": 0.2548202992441771,
+      "grad_norm": 0.0,
+      "learning_rate": 1.7467235091443326e-05,
+      "loss": 0.934,
+      "step": 826
+    },
+    {
+      "epoch": 0.25512879839580443,
+      "grad_norm": 0.0,
+      "learning_rate": 1.7460582994825127e-05,
+      "loss": 0.8076,
+      "step": 827
+    },
+    {
+      "epoch": 0.2554372975474317,
+      "grad_norm": 0.0,
+      "learning_rate": 1.7453923444304334e-05,
+      "loss": 0.879,
+      "step": 828
+    },
+    {
+      "epoch": 0.25574579669905906,
+      "grad_norm": 0.0,
+      "learning_rate": 1.7447256446534534e-05,
+      "loss": 0.7934,
+      "step": 829
+    },
+    {
+      "epoch": 0.2560542958506864,
+      "grad_norm": 0.0,
+      "learning_rate": 1.7440582008176756e-05,
+      "loss": 0.8131,
+      "step": 830
+    },
+    {
+      "epoch": 0.25636279500231374,
+      "grad_norm": 0.0,
+      "learning_rate": 1.743390013589946e-05,
+      "loss": 0.8817,
+      "step": 831
+    },
+    {
+      "epoch": 0.2566712941539411,
+      "grad_norm": 0.0,
+      "learning_rate": 1.7427210836378535e-05,
+      "loss": 0.8438,
+      "step": 832
+    },
+    {
+      "epoch": 0.2569797933055684,
+      "grad_norm": 0.0,
+      "learning_rate": 1.7420514116297294e-05,
+      "loss": 0.8695,
+      "step": 833
+    },
+    {
+      "epoch": 0.25728829245719576,
+      "grad_norm": 0.0,
+      "learning_rate": 1.7413809982346458e-05,
+      "loss": 0.8446,
+      "step": 834
+    },
+    {
+      "epoch": 0.2575967916088231,
+      "grad_norm": 0.0,
+      "learning_rate": 1.7407098441224154e-05,
+      "loss": 0.8197,
+      "step": 835
+    },
+    {
+      "epoch": 0.25790529076045043,
+      "grad_norm": 0.0,
+      "learning_rate": 1.7400379499635926e-05,
+      "loss": 0.8765,
+      "step": 836
+    },
+    {
+      "epoch": 0.2582137899120777,
+      "grad_norm": 0.0,
+      "learning_rate": 1.7393653164294685e-05,
+      "loss": 0.8708,
+      "step": 837
+    },
+    {
+      "epoch": 0.25852228906370506,
+      "grad_norm": 0.0,
+      "learning_rate": 1.7386919441920748e-05,
+      "loss": 0.7986,
+      "step": 838
+    },
+    {
+      "epoch": 0.2588307882153324,
+      "grad_norm": 0.0,
+      "learning_rate": 1.738017833924181e-05,
+      "loss": 0.8454,
+      "step": 839
+    },
+    {
+      "epoch": 0.25913928736695974,
+      "grad_norm": 0.0,
+      "learning_rate": 1.737342986299294e-05,
+      "loss": 0.8292,
+      "step": 840
+    },
+    {
+      "epoch": 0.2594477865185871,
+      "grad_norm": 0.0,
+      "learning_rate": 1.7366674019916567e-05,
+      "loss": 0.8716,
+      "step": 841
+    },
+    {
+      "epoch": 0.2597562856702144,
+      "grad_norm": 0.0,
+      "learning_rate": 1.7359910816762487e-05,
+      "loss": 0.7871,
+      "step": 842
+    },
+    {
+      "epoch": 0.26006478482184175,
+      "grad_norm": 0.0,
+      "learning_rate": 1.7353140260287845e-05,
+      "loss": 0.9423,
+      "step": 843
+    },
+    {
+      "epoch": 0.2603732839734691,
+      "grad_norm": 0.0,
+      "learning_rate": 1.7346362357257135e-05,
+      "loss": 0.8139,
+      "step": 844
+    },
+    {
+      "epoch": 0.26068178312509643,
+      "grad_norm": 0.0,
+      "learning_rate": 1.7339577114442194e-05,
+      "loss": 0.8117,
+      "step": 845
+    },
+    {
+      "epoch": 0.2609902822767237,
+      "grad_norm": 0.0,
+      "learning_rate": 1.7332784538622184e-05,
+      "loss": 0.9124,
+      "step": 846
+    },
+    {
+      "epoch": 0.26129878142835106,
+      "grad_norm": 0.0,
+      "learning_rate": 1.7325984636583606e-05,
+      "loss": 0.8608,
+      "step": 847
+    },
+    {
+      "epoch": 0.2616072805799784,
+      "grad_norm": 0.0,
+      "learning_rate": 1.731917741512027e-05,
+      "loss": 0.8616,
+      "step": 848
+    },
+    {
+      "epoch": 0.26191577973160574,
+      "grad_norm": 0.0,
+      "learning_rate": 1.7312362881033293e-05,
+      "loss": 0.8625,
+      "step": 849
+    },
+    {
+      "epoch": 0.2622242788832331,
+      "grad_norm": 0.0,
+      "learning_rate": 1.730554104113112e-05,
+      "loss": 0.8,
+      "step": 850
+    },
+    {
+      "epoch": 0.2625327780348604,
+      "grad_norm": 0.0,
+      "learning_rate": 1.7298711902229478e-05,
+      "loss": 0.7432,
+      "step": 851
+    },
+    {
+      "epoch": 0.26284127718648775,
+      "grad_norm": 0.0,
+      "learning_rate": 1.7291875471151392e-05,
+      "loss": 0.93,
+      "step": 852
+    },
+    {
+      "epoch": 0.2631497763381151,
+      "grad_norm": 0.0,
+      "learning_rate": 1.728503175472717e-05,
+      "loss": 0.8965,
+      "step": 853
+    },
+    {
+      "epoch": 0.2634582754897424,
+      "grad_norm": 0.0,
+      "learning_rate": 1.7278180759794397e-05,
+      "loss": 0.8166,
+      "step": 854
+    },
+    {
+      "epoch": 0.2637667746413697,
+      "grad_norm": 0.0,
+      "learning_rate": 1.727132249319794e-05,
+      "loss": 0.8628,
+      "step": 855
+    },
+    {
+      "epoch": 0.26407527379299706,
+      "grad_norm": 0.0,
+      "learning_rate": 1.726445696178992e-05,
+      "loss": 0.8244,
+      "step": 856
+    },
+    {
+      "epoch": 0.2643837729446244,
+      "grad_norm": 0.0,
+      "learning_rate": 1.7257584172429723e-05,
+      "loss": 0.7755,
+      "step": 857
+    },
+    {
+      "epoch": 0.26469227209625174,
+      "grad_norm": 0.0,
+      "learning_rate": 1.7250704131983984e-05,
+      "loss": 0.7426,
+      "step": 858
+    },
+    {
+      "epoch": 0.2650007712478791,
+      "grad_norm": 0.0,
+      "learning_rate": 1.724381684732658e-05,
+      "loss": 0.889,
+      "step": 859
+    },
+    {
+      "epoch": 0.2653092703995064,
+      "grad_norm": 0.0,
+      "learning_rate": 1.723692232533863e-05,
+      "loss": 0.9441,
+      "step": 860
+    },
+    {
+      "epoch": 0.26561776955113375,
+      "grad_norm": 0.0,
+      "learning_rate": 1.723002057290849e-05,
+      "loss": 0.8384,
+      "step": 861
+    },
+    {
+      "epoch": 0.2659262687027611,
+      "grad_norm": 0.0,
+      "learning_rate": 1.7223111596931722e-05,
+      "loss": 0.8027,
+      "step": 862
+    },
+    {
+      "epoch": 0.2662347678543884,
+      "grad_norm": 0.0,
+      "learning_rate": 1.721619540431112e-05,
+      "loss": 0.8804,
+      "step": 863
+    },
+    {
+      "epoch": 0.2665432670060157,
+      "grad_norm": 0.0,
+      "learning_rate": 1.720927200195668e-05,
+      "loss": 0.8112,
+      "step": 864
+    },
+    {
+      "epoch": 0.26685176615764306,
+      "grad_norm": 0.0,
+      "learning_rate": 1.7202341396785613e-05,
+      "loss": 0.9347,
+      "step": 865
+    },
+    {
+      "epoch": 0.2671602653092704,
+      "grad_norm": 0.0,
+      "learning_rate": 1.719540359572231e-05,
+      "loss": 0.8728,
+      "step": 866
+    },
+    {
+      "epoch": 0.26746876446089773,
+      "grad_norm": 0.0,
+      "learning_rate": 1.7188458605698365e-05,
+      "loss": 0.8349,
+      "step": 867
+    },
+    {
+      "epoch": 0.2677772636125251,
+      "grad_norm": 0.0,
+      "learning_rate": 1.7181506433652545e-05,
+      "loss": 0.7284,
+      "step": 868
+    },
+    {
+      "epoch": 0.2680857627641524,
+      "grad_norm": 0.0,
+      "learning_rate": 1.71745470865308e-05,
+      "loss": 0.8875,
+      "step": 869
+    },
+    {
+      "epoch": 0.26839426191577975,
+      "grad_norm": 0.0,
+      "learning_rate": 1.7167580571286247e-05,
+      "loss": 0.8623,
+      "step": 870
+    },
+    {
+      "epoch": 0.2687027610674071,
+      "grad_norm": 0.0,
+      "learning_rate": 1.716060689487916e-05,
+      "loss": 0.817,
+      "step": 871
+    },
+    {
+      "epoch": 0.2690112602190344,
+      "grad_norm": 0.0,
+      "learning_rate": 1.7153626064276972e-05,
+      "loss": 0.8666,
+      "step": 872
+    },
+    {
+      "epoch": 0.2693197593706617,
+      "grad_norm": 0.0,
+      "learning_rate": 1.7146638086454264e-05,
+      "loss": 0.8331,
+      "step": 873
+    },
+    {
+      "epoch": 0.26962825852228905,
+      "grad_norm": 0.0,
+      "learning_rate": 1.7139642968392754e-05,
+      "loss": 0.8649,
+      "step": 874
+    },
+    {
+      "epoch": 0.2699367576739164,
+      "grad_norm": 0.0,
+      "learning_rate": 1.71326407170813e-05,
+      "loss": 0.9272,
+      "step": 875
+    },
+    {
+      "epoch": 0.27024525682554373,
+      "grad_norm": 0.0,
+      "learning_rate": 1.712563133951588e-05,
+      "loss": 0.8894,
+      "step": 876
+    },
+    {
+      "epoch": 0.2705537559771711,
+      "grad_norm": 0.0,
+      "learning_rate": 1.7118614842699595e-05,
+      "loss": 0.8173,
+      "step": 877
+    },
+    {
+      "epoch": 0.2708622551287984,
+      "grad_norm": 0.0,
+      "learning_rate": 1.711159123364266e-05,
+      "loss": 0.8226,
+      "step": 878
+    },
+    {
+      "epoch": 0.27117075428042575,
+      "grad_norm": 0.0,
+      "learning_rate": 1.7104560519362398e-05,
+      "loss": 0.8828,
+      "step": 879
+    },
+    {
+      "epoch": 0.27147925343205304,
+      "grad_norm": 0.0,
+      "learning_rate": 1.7097522706883225e-05,
+      "loss": 0.7943,
+      "step": 880
+    },
+    {
+      "epoch": 0.2717877525836804,
+      "grad_norm": 0.0,
+      "learning_rate": 1.709047780323665e-05,
+      "loss": 0.8355,
+      "step": 881
+    },
+    {
+      "epoch": 0.2720962517353077,
+      "grad_norm": 0.0,
+      "learning_rate": 1.7083425815461273e-05,
+      "loss": 0.8576,
+      "step": 882
+    },
+    {
+      "epoch": 0.27240475088693505,
+      "grad_norm": 0.0,
+      "learning_rate": 1.707636675060276e-05,
+      "loss": 0.8617,
+      "step": 883
+    },
+    {
+      "epoch": 0.2727132500385624,
+      "grad_norm": 0.0,
+      "learning_rate": 1.7069300615713866e-05,
+      "loss": 0.7715,
+      "step": 884
+    },
+    {
+      "epoch": 0.27302174919018973,
+      "grad_norm": 0.0,
+      "learning_rate": 1.7062227417854388e-05,
+      "loss": 0.8574,
+      "step": 885
+    },
+    {
+      "epoch": 0.2733302483418171,
+      "grad_norm": 0.0,
+      "learning_rate": 1.7055147164091197e-05,
+      "loss": 0.8169,
+      "step": 886
+    },
+    {
+      "epoch": 0.2736387474934444,
+      "grad_norm": 0.0,
+      "learning_rate": 1.7048059861498205e-05,
+      "loss": 0.8913,
+      "step": 887
+    },
+    {
+      "epoch": 0.27394724664507175,
+      "grad_norm": 0.0,
+      "learning_rate": 1.7040965517156365e-05,
+      "loss": 0.898,
+      "step": 888
+    },
+    {
+      "epoch": 0.27425574579669904,
+      "grad_norm": 0.0,
+      "learning_rate": 1.703386413815367e-05,
+      "loss": 0.8039,
+      "step": 889
+    },
+    {
+      "epoch": 0.2745642449483264,
+      "grad_norm": 0.0,
+      "learning_rate": 1.7026755731585146e-05,
+      "loss": 0.8275,
+      "step": 890
+    },
+    {
+      "epoch": 0.2748727440999537,
+      "grad_norm": 0.0,
+      "learning_rate": 1.7019640304552832e-05,
+      "loss": 0.8026,
+      "step": 891
+    },
+    {
+      "epoch": 0.27518124325158105,
+      "grad_norm": 0.0,
+      "learning_rate": 1.7012517864165778e-05,
+      "loss": 0.9627,
+      "step": 892
+    },
+    {
+      "epoch": 0.2754897424032084,
+      "grad_norm": 0.0,
+      "learning_rate": 1.7005388417540055e-05,
+      "loss": 0.7905,
+      "step": 893
+    },
+    {
+      "epoch": 0.27579824155483573,
+      "grad_norm": 0.0,
+      "learning_rate": 1.6998251971798717e-05,
+      "loss": 0.8794,
+      "step": 894
+    },
+    {
+      "epoch": 0.27610674070646307,
+      "grad_norm": 0.0,
+      "learning_rate": 1.699110853407183e-05,
+      "loss": 0.8351,
+      "step": 895
+    },
+    {
+      "epoch": 0.2764152398580904,
+      "grad_norm": 0.0,
+      "learning_rate": 1.6983958111496428e-05,
+      "loss": 0.8739,
+      "step": 896
+    },
+    {
+      "epoch": 0.27672373900971775,
+      "grad_norm": 0.0,
+      "learning_rate": 1.6976800711216527e-05,
+      "loss": 0.9094,
+      "step": 897
+    },
+    {
+      "epoch": 0.27703223816134503,
+      "grad_norm": 0.0,
+      "learning_rate": 1.6969636340383134e-05,
+      "loss": 0.8181,
+      "step": 898
+    },
+    {
+      "epoch": 0.2773407373129724,
+      "grad_norm": 0.0,
+      "learning_rate": 1.6962465006154186e-05,
+      "loss": 0.8547,
+      "step": 899
+    },
+    {
+      "epoch": 0.2776492364645997,
+      "grad_norm": 0.0,
+      "learning_rate": 1.695528671569461e-05,
+      "loss": 0.8662,
+      "step": 900
+    },
+    {
+      "epoch": 0.27795773561622705,
+      "grad_norm": 0.0,
+      "learning_rate": 1.694810147617626e-05,
+      "loss": 0.8293,
+      "step": 901
+    },
+    {
+      "epoch": 0.2782662347678544,
+      "grad_norm": 0.0,
+      "learning_rate": 1.6940909294777945e-05,
+      "loss": 0.8625,
+      "step": 902
+    },
+    {
+      "epoch": 0.27857473391948173,
+      "grad_norm": 0.0,
+      "learning_rate": 1.6933710178685406e-05,
+      "loss": 0.8169,
+      "step": 903
+    },
+    {
+      "epoch": 0.27888323307110907,
+      "grad_norm": 0.0,
+      "learning_rate": 1.6926504135091315e-05,
+      "loss": 0.7649,
+      "step": 904
+    },
+    {
+      "epoch": 0.2791917322227364,
+      "grad_norm": 0.0,
+      "learning_rate": 1.691929117119526e-05,
+      "loss": 0.8727,
+      "step": 905
+    },
+    {
+      "epoch": 0.2795002313743637,
+      "grad_norm": 0.0,
+      "learning_rate": 1.6912071294203746e-05,
+      "loss": 0.8505,
+      "step": 906
+    },
+    {
+      "epoch": 0.27980873052599103,
+      "grad_norm": 0.0,
+      "learning_rate": 1.690484451133019e-05,
+      "loss": 0.817,
+      "step": 907
+    },
+    {
+      "epoch": 0.2801172296776184,
+      "grad_norm": 0.0,
+      "learning_rate": 1.6897610829794898e-05,
+      "loss": 0.8791,
+      "step": 908
+    },
+    {
+      "epoch": 0.2804257288292457,
+      "grad_norm": 0.0,
+      "learning_rate": 1.6890370256825077e-05,
+      "loss": 0.7989,
+      "step": 909
+    },
+    {
+      "epoch": 0.28073422798087305,
+      "grad_norm": 0.0,
+      "learning_rate": 1.6883122799654814e-05,
+      "loss": 0.8845,
+      "step": 910
+    },
+    {
+      "epoch": 0.2810427271325004,
+      "grad_norm": 0.0,
+      "learning_rate": 1.6875868465525084e-05,
+      "loss": 0.8552,
+      "step": 911
+    },
+    {
+      "epoch": 0.28135122628412773,
+      "grad_norm": 0.0,
+      "learning_rate": 1.6868607261683716e-05,
+      "loss": 0.9005,
+      "step": 912
+    },
+    {
+      "epoch": 0.28165972543575507,
+      "grad_norm": 0.0,
+      "learning_rate": 1.686133919538542e-05,
+      "loss": 0.8005,
+      "step": 913
+    },
+    {
+      "epoch": 0.2819682245873824,
+      "grad_norm": 0.0,
+      "learning_rate": 1.685406427389175e-05,
+      "loss": 0.8123,
+      "step": 914
+    },
+    {
+      "epoch": 0.2822767237390097,
+      "grad_norm": 0.0,
+      "learning_rate": 1.6846782504471112e-05,
+      "loss": 0.8071,
+      "step": 915
+    },
+    {
+      "epoch": 0.28258522289063703,
+      "grad_norm": 0.0,
+      "learning_rate": 1.6839493894398753e-05,
+      "loss": 0.8606,
+      "step": 916
+    },
+    {
+      "epoch": 0.2828937220422644,
+      "grad_norm": 0.0,
+      "learning_rate": 1.6832198450956766e-05,
+      "loss": 0.8498,
+      "step": 917
+    },
+    {
+      "epoch": 0.2832022211938917,
+      "grad_norm": 0.0,
+      "learning_rate": 1.6824896181434055e-05,
+      "loss": 0.8226,
+      "step": 918
+    },
+    {
+      "epoch": 0.28351072034551905,
+      "grad_norm": 0.0,
+      "learning_rate": 1.6817587093126354e-05,
+      "loss": 0.9171,
+      "step": 919
+    },
+    {
+      "epoch": 0.2838192194971464,
+      "grad_norm": 0.0,
+      "learning_rate": 1.6810271193336203e-05,
+      "loss": 0.7303,
+      "step": 920
+    },
+    {
+      "epoch": 0.28412771864877373,
+      "grad_norm": 0.0,
+      "learning_rate": 1.6802948489372956e-05,
+      "loss": 0.8212,
+      "step": 921
+    },
+    {
+      "epoch": 0.28443621780040107,
+      "grad_norm": 0.0,
+      "learning_rate": 1.6795618988552754e-05,
+      "loss": 0.7725,
+      "step": 922
+    },
+    {
+      "epoch": 0.2847447169520284,
+      "grad_norm": 0.0,
+      "learning_rate": 1.6788282698198536e-05,
+      "loss": 0.8705,
+      "step": 923
+    },
+    {
+      "epoch": 0.2850532161036557,
+      "grad_norm": 0.0,
+      "learning_rate": 1.678093962564003e-05,
+      "loss": 0.7047,
+      "step": 924
+    },
+    {
+      "epoch": 0.28536171525528303,
+      "grad_norm": 0.0,
+      "learning_rate": 1.6773589778213724e-05,
+      "loss": 0.7995,
+      "step": 925
+    },
+    {
+      "epoch": 0.28567021440691037,
+      "grad_norm": 0.0,
+      "learning_rate": 1.6766233163262893e-05,
+      "loss": 0.8124,
+      "step": 926
+    },
+    {
+      "epoch": 0.2859787135585377,
+      "grad_norm": 0.0,
+      "learning_rate": 1.675886978813756e-05,
+      "loss": 0.9003,
+      "step": 927
+    },
+    {
+      "epoch": 0.28628721271016505,
+      "grad_norm": 0.0,
+      "learning_rate": 1.6751499660194502e-05,
+      "loss": 0.8927,
+      "step": 928
+    },
+    {
+      "epoch": 0.2865957118617924,
+      "grad_norm": 0.0,
+      "learning_rate": 1.6744122786797254e-05,
+      "loss": 0.8025,
+      "step": 929
+    },
+    {
+      "epoch": 0.28690421101341973,
+      "grad_norm": 0.0,
+      "learning_rate": 1.6736739175316086e-05,
+      "loss": 0.9125,
+      "step": 930
+    },
+    {
+      "epoch": 0.28721271016504707,
+      "grad_norm": 0.0,
+      "learning_rate": 1.672934883312799e-05,
+      "loss": 0.8202,
+      "step": 931
+    },
+    {
+      "epoch": 0.28752120931667435,
+      "grad_norm": 0.0,
+      "learning_rate": 1.6721951767616696e-05,
+      "loss": 0.8578,
+      "step": 932
+    },
+    {
+      "epoch": 0.2878297084683017,
+      "grad_norm": 0.0,
+      "learning_rate": 1.671454798617265e-05,
+      "loss": 0.8437,
+      "step": 933
+    },
+    {
+      "epoch": 0.28813820761992903,
+      "grad_norm": 0.0,
+      "learning_rate": 1.6707137496192994e-05,
+      "loss": 0.8265,
+      "step": 934
+    },
+    {
+      "epoch": 0.28844670677155637,
+      "grad_norm": 0.0,
+      "learning_rate": 1.669972030508159e-05,
+      "loss": 0.8087,
+      "step": 935
+    },
+    {
+      "epoch": 0.2887552059231837,
+      "grad_norm": 0.0,
+      "learning_rate": 1.6692296420248985e-05,
+      "loss": 0.7741,
+      "step": 936
+    },
+    {
+      "epoch": 0.28906370507481105,
+      "grad_norm": 0.0,
+      "learning_rate": 1.6684865849112414e-05,
+      "loss": 0.8904,
+      "step": 937
+    },
+    {
+      "epoch": 0.2893722042264384,
+      "grad_norm": 0.0,
+      "learning_rate": 1.6677428599095796e-05,
+      "loss": 0.867,
+      "step": 938
+    },
+    {
+      "epoch": 0.28968070337806573,
+      "grad_norm": 0.0,
+      "learning_rate": 1.666998467762973e-05,
+      "loss": 0.765,
+      "step": 939
+    },
+    {
+      "epoch": 0.28998920252969307,
+      "grad_norm": 0.0,
+      "learning_rate": 1.6662534092151457e-05,
+      "loss": 0.7802,
+      "step": 940
+    },
+    {
+      "epoch": 0.29029770168132035,
+      "grad_norm": 0.0,
+      "learning_rate": 1.6655076850104902e-05,
+      "loss": 0.8506,
+      "step": 941
+    },
+    {
+      "epoch": 0.2906062008329477,
+      "grad_norm": 0.0,
+      "learning_rate": 1.6647612958940622e-05,
+      "loss": 0.8015,
+      "step": 942
+    },
+    {
+      "epoch": 0.29091469998457503,
+      "grad_norm": 0.0,
+      "learning_rate": 1.6640142426115833e-05,
+      "loss": 0.8601,
+      "step": 943
+    },
+    {
+      "epoch": 0.29122319913620237,
+      "grad_norm": 0.0,
+      "learning_rate": 1.663266525909437e-05,
+      "loss": 0.8014,
+      "step": 944
+    },
+    {
+      "epoch": 0.2915316982878297,
+      "grad_norm": 0.0,
+      "learning_rate": 1.6625181465346717e-05,
+      "loss": 0.7378,
+      "step": 945
+    },
+    {
+      "epoch": 0.29184019743945705,
+      "grad_norm": 0.0,
+      "learning_rate": 1.6617691052349954e-05,
+      "loss": 0.9138,
+      "step": 946
+    },
+    {
+      "epoch": 0.2921486965910844,
+      "grad_norm": 0.0,
+      "learning_rate": 1.661019402758779e-05,
+      "loss": 0.8472,
+      "step": 947
+    },
+    {
+      "epoch": 0.29245719574271173,
+      "grad_norm": 0.0,
+      "learning_rate": 1.6602690398550542e-05,
+      "loss": 0.8478,
+      "step": 948
+    },
+    {
+      "epoch": 0.29276569489433907,
+      "grad_norm": 0.0,
+      "learning_rate": 1.6595180172735116e-05,
+      "loss": 0.8009,
+      "step": 949
+    },
+    {
+      "epoch": 0.29307419404596635,
+      "grad_norm": 0.0,
+      "learning_rate": 1.658766335764501e-05,
+      "loss": 0.8826,
+      "step": 950
+    },
+    {
+      "epoch": 0.2933826931975937,
+      "grad_norm": 0.0,
+      "learning_rate": 1.6580139960790316e-05,
+      "loss": 0.89,
+      "step": 951
+    },
+    {
+      "epoch": 0.29369119234922103,
+      "grad_norm": 0.0,
+      "learning_rate": 1.6572609989687687e-05,
+      "loss": 0.8615,
+      "step": 952
+    },
+    {
+      "epoch": 0.29399969150084837,
+      "grad_norm": 0.0,
+      "learning_rate": 1.6565073451860355e-05,
+      "loss": 0.8855,
+      "step": 953
+    },
+    {
+      "epoch": 0.2943081906524757,
+      "grad_norm": 0.0,
+      "learning_rate": 1.6557530354838108e-05,
+      "loss": 0.824,
+      "step": 954
+    },
+    {
+      "epoch": 0.29461668980410305,
+      "grad_norm": 0.0,
+      "learning_rate": 1.6549980706157295e-05,
+      "loss": 0.8551,
+      "step": 955
+    },
+    {
+      "epoch": 0.2949251889557304,
+      "grad_norm": 0.0,
+      "learning_rate": 1.6542424513360793e-05,
+      "loss": 0.8421,
+      "step": 956
+    },
+    {
+      "epoch": 0.29523368810735773,
+      "grad_norm": 0.0,
+      "learning_rate": 1.653486178399804e-05,
+      "loss": 0.8077,
+      "step": 957
+    },
+    {
+      "epoch": 0.295542187258985,
+      "grad_norm": 0.0,
+      "learning_rate": 1.6527292525624986e-05,
+      "loss": 0.8392,
+      "step": 958
+    },
+    {
+      "epoch": 0.29585068641061235,
+      "grad_norm": 0.0,
+      "learning_rate": 1.6519716745804112e-05,
+      "loss": 0.7634,
+      "step": 959
+    },
+    {
+      "epoch": 0.2961591855622397,
+      "grad_norm": 0.0,
+      "learning_rate": 1.651213445210442e-05,
+      "loss": 0.7943,
+      "step": 960
+    },
+    {
+      "epoch": 0.29646768471386703,
+      "grad_norm": 0.0,
+      "learning_rate": 1.650454565210141e-05,
+      "loss": 0.848,
+      "step": 961
+    },
+    {
+      "epoch": 0.29677618386549437,
+      "grad_norm": 0.0,
+      "learning_rate": 1.649695035337709e-05,
+      "loss": 0.8384,
+      "step": 962
+    },
+    {
+      "epoch": 0.2970846830171217,
+      "grad_norm": 0.0,
+      "learning_rate": 1.648934856351995e-05,
+      "loss": 0.8023,
+      "step": 963
+    },
+    {
+      "epoch": 0.29739318216874905,
+      "grad_norm": 0.0,
+      "learning_rate": 1.648174029012498e-05,
+      "loss": 0.872,
+      "step": 964
+    },
+    {
+      "epoch": 0.2977016813203764,
+      "grad_norm": 0.0,
+      "learning_rate": 1.647412554079364e-05,
+      "loss": 0.7895,
+      "step": 965
+    },
+    {
+      "epoch": 0.2980101804720037,
+      "grad_norm": 0.0,
+      "learning_rate": 1.6466504323133857e-05,
+      "loss": 0.8942,
+      "step": 966
+    },
+    {
+      "epoch": 0.298318679623631,
+      "grad_norm": 0.0,
+      "learning_rate": 1.6458876644760033e-05,
+      "loss": 0.877,
+      "step": 967
+    },
+    {
+      "epoch": 0.29862717877525835,
+      "grad_norm": 0.0,
+      "learning_rate": 1.6451242513293005e-05,
+      "loss": 0.7891,
+      "step": 968
+    },
+    {
+      "epoch": 0.2989356779268857,
+      "grad_norm": 0.0,
+      "learning_rate": 1.644360193636008e-05,
+      "loss": 0.863,
+      "step": 969
+    },
+    {
+      "epoch": 0.29924417707851303,
+      "grad_norm": 0.0,
+      "learning_rate": 1.6435954921594985e-05,
+      "loss": 0.8448,
+      "step": 970
+    },
+    {
+      "epoch": 0.29955267623014037,
+      "grad_norm": 0.0,
+      "learning_rate": 1.642830147663789e-05,
+      "loss": 0.8145,
+      "step": 971
+    },
+    {
+      "epoch": 0.2998611753817677,
+      "grad_norm": 0.0,
+      "learning_rate": 1.6420641609135388e-05,
+      "loss": 0.8186,
+      "step": 972
+    },
+    {
+      "epoch": 0.30016967453339505,
+      "grad_norm": 0.0,
+      "learning_rate": 1.6412975326740485e-05,
+      "loss": 0.7535,
+      "step": 973
+    },
+    {
+      "epoch": 0.3004781736850224,
+      "grad_norm": 0.0,
+      "learning_rate": 1.6405302637112598e-05,
+      "loss": 0.8412,
+      "step": 974
+    },
+    {
+      "epoch": 0.3007866728366497,
+      "grad_norm": 0.0,
+      "learning_rate": 1.6397623547917553e-05,
+      "loss": 0.7729,
+      "step": 975
+    },
+    {
+      "epoch": 0.301095171988277,
+      "grad_norm": 0.0,
+      "learning_rate": 1.6389938066827556e-05,
+      "loss": 0.7915,
+      "step": 976
+    },
+    {
+      "epoch": 0.30140367113990435,
+      "grad_norm": 0.0,
+      "learning_rate": 1.6382246201521213e-05,
+      "loss": 0.7591,
+      "step": 977
+    },
+    {
+      "epoch": 0.3017121702915317,
+      "grad_norm": 0.0,
+      "learning_rate": 1.6374547959683497e-05,
+      "loss": 0.8292,
+      "step": 978
+    },
+    {
+      "epoch": 0.30202066944315903,
+      "grad_norm": 0.0,
+      "learning_rate": 1.6366843349005755e-05,
+      "loss": 0.7889,
+      "step": 979
+    },
+    {
+      "epoch": 0.30232916859478637,
+      "grad_norm": 0.0,
+      "learning_rate": 1.63591323771857e-05,
+      "loss": 0.7864,
+      "step": 980
+    },
+    {
+      "epoch": 0.3026376677464137,
+      "grad_norm": 0.0,
+      "learning_rate": 1.6351415051927407e-05,
+      "loss": 0.8059,
+      "step": 981
+    },
+    {
+      "epoch": 0.30294616689804105,
+      "grad_norm": 0.0,
+      "learning_rate": 1.634369138094128e-05,
+      "loss": 0.8261,
+      "step": 982
+    },
+    {
+      "epoch": 0.3032546660496684,
+      "grad_norm": 0.0,
+      "learning_rate": 1.6335961371944084e-05,
+      "loss": 0.8465,
+      "step": 983
+    },
+    {
+      "epoch": 0.30356316520129567,
+      "grad_norm": 0.0,
+      "learning_rate": 1.6328225032658892e-05,
+      "loss": 0.8349,
+      "step": 984
+    },
+    {
+      "epoch": 0.303871664352923,
+      "grad_norm": 0.0,
+      "learning_rate": 1.6320482370815132e-05,
+      "loss": 0.9298,
+      "step": 985
+    },
+    {
+      "epoch": 0.30418016350455035,
+      "grad_norm": 0.0,
+      "learning_rate": 1.6312733394148524e-05,
+      "loss": 0.7644,
+      "step": 986
+    },
+    {
+      "epoch": 0.3044886626561777,
+      "grad_norm": 0.0,
+      "learning_rate": 1.6304978110401106e-05,
+      "loss": 0.8733,
+      "step": 987
+    },
+    {
+      "epoch": 0.30479716180780503,
+      "grad_norm": 0.0,
+      "learning_rate": 1.6297216527321223e-05,
+      "loss": 0.7945,
+      "step": 988
+    },
+    {
+      "epoch": 0.30510566095943237,
+      "grad_norm": 0.0,
+      "learning_rate": 1.62894486526635e-05,
+      "loss": 0.8094,
+      "step": 989
+    },
+    {
+      "epoch": 0.3054141601110597,
+      "grad_norm": 0.0,
+      "learning_rate": 1.6281674494188863e-05,
+      "loss": 0.8753,
+      "step": 990
+    },
+    {
+      "epoch": 0.30572265926268705,
+      "grad_norm": 0.0,
+      "learning_rate": 1.6273894059664507e-05,
+      "loss": 0.7734,
+      "step": 991
+    },
+    {
+      "epoch": 0.3060311584143144,
+      "grad_norm": 0.0,
+      "learning_rate": 1.62661073568639e-05,
+      "loss": 0.8767,
+      "step": 992
+    },
+    {
+      "epoch": 0.30633965756594167,
+      "grad_norm": 0.0,
+      "learning_rate": 1.625831439356677e-05,
+      "loss": 0.843,
+      "step": 993
+    },
+    {
+      "epoch": 0.306648156717569,
+      "grad_norm": 0.0,
+      "learning_rate": 1.6250515177559106e-05,
+      "loss": 0.8475,
+      "step": 994
+    },
+    {
+      "epoch": 0.30695665586919635,
+      "grad_norm": 0.0,
+      "learning_rate": 1.6242709716633137e-05,
+      "loss": 0.806,
+      "step": 995
+    },
+    {
+      "epoch": 0.3072651550208237,
+      "grad_norm": 0.0,
+      "learning_rate": 1.6234898018587336e-05,
+      "loss": 0.8225,
+      "step": 996
+    },
+    {
+      "epoch": 0.307573654172451,
+      "grad_norm": 0.0,
+      "learning_rate": 1.622708009122641e-05,
+      "loss": 0.8084,
+      "step": 997
+    },
+    {
+      "epoch": 0.30788215332407837,
+      "grad_norm": 0.0,
+      "learning_rate": 1.621925594236128e-05,
+      "loss": 0.8576,
+      "step": 998
+    },
+    {
+      "epoch": 0.3081906524757057,
+      "grad_norm": 0.0,
+      "learning_rate": 1.621142557980909e-05,
+      "loss": 0.917,
+      "step": 999
+    },
+    {
+      "epoch": 0.30849915162733305,
+      "grad_norm": 0.0,
+      "learning_rate": 1.6203589011393198e-05,
+      "loss": 0.8328,
+      "step": 1000
+    },
+    {
+      "epoch": 0.3088076507789604,
+      "grad_norm": 0.0,
+      "learning_rate": 1.6195746244943142e-05,
+      "loss": 1.1333,
+      "step": 1001
+    },
+    {
+      "epoch": 0.30911614993058767,
+      "grad_norm": 0.0,
+      "learning_rate": 1.618789728829468e-05,
+      "loss": 0.8886,
+      "step": 1002
+    },
+    {
+      "epoch": 0.309424649082215,
+      "grad_norm": 0.0,
+      "learning_rate": 1.618004214928973e-05,
+      "loss": 0.7299,
+      "step": 1003
+    },
+    {
+      "epoch": 0.30973314823384235,
+      "grad_norm": 0.0,
+      "learning_rate": 1.6172180835776404e-05,
+      "loss": 0.8005,
+      "step": 1004
+    },
+    {
+      "epoch": 0.3100416473854697,
+      "grad_norm": 0.0,
+      "learning_rate": 1.6164313355608974e-05,
+      "loss": 0.7836,
+      "step": 1005
+    },
+    {
+      "epoch": 0.310350146537097,
+      "grad_norm": 0.0,
+      "learning_rate": 1.6156439716647875e-05,
+      "loss": 0.8636,
+      "step": 1006
+    },
+    {
+      "epoch": 0.31065864568872437,
+      "grad_norm": 0.0,
+      "learning_rate": 1.6148559926759694e-05,
+      "loss": 0.8338,
+      "step": 1007
+    },
+    {
+      "epoch": 0.3109671448403517,
+      "grad_norm": 0.0,
+      "learning_rate": 1.614067399381717e-05,
+      "loss": 0.8371,
+      "step": 1008
+    },
+    {
+      "epoch": 0.31127564399197905,
+      "grad_norm": 0.0,
+      "learning_rate": 1.6132781925699168e-05,
+      "loss": 1.1121,
+      "step": 1009
+    },
+    {
+      "epoch": 0.31158414314360633,
+      "grad_norm": 0.0,
+      "learning_rate": 1.6124883730290695e-05,
+      "loss": 0.8098,
+      "step": 1010
+    },
+    {
+      "epoch": 0.31189264229523367,
+      "grad_norm": 0.0,
+      "learning_rate": 1.6116979415482875e-05,
+      "loss": 0.9123,
+      "step": 1011
+    },
+    {
+      "epoch": 0.312201141446861,
+      "grad_norm": 0.0,
+      "learning_rate": 1.6109068989172937e-05,
+      "loss": 0.8123,
+      "step": 1012
+    },
+    {
+      "epoch": 0.31250964059848835,
+      "grad_norm": 0.0,
+      "learning_rate": 1.610115245926423e-05,
+      "loss": 0.8503,
+      "step": 1013
+    },
+    {
+      "epoch": 0.3128181397501157,
+      "grad_norm": 0.0,
+      "learning_rate": 1.60932298336662e-05,
+      "loss": 0.7881,
+      "step": 1014
+    },
+    {
+      "epoch": 0.313126638901743,
+      "grad_norm": 0.0,
+      "learning_rate": 1.608530112029437e-05,
+      "loss": 0.786,
+      "step": 1015
+    },
+    {
+      "epoch": 0.31343513805337037,
+      "grad_norm": 0.0,
+      "learning_rate": 1.6077366327070354e-05,
+      "loss": 0.8752,
+      "step": 1016
+    },
+    {
+      "epoch": 0.3137436372049977,
+      "grad_norm": 0.0,
+      "learning_rate": 1.606942546192185e-05,
+      "loss": 0.8288,
+      "step": 1017
+    },
+    {
+      "epoch": 0.31405213635662504,
+      "grad_norm": 0.0,
+      "learning_rate": 1.60614785327826e-05,
+      "loss": 0.8726,
+      "step": 1018
+    },
+    {
+      "epoch": 0.31436063550825233,
+      "grad_norm": 0.0,
+      "learning_rate": 1.6053525547592424e-05,
+      "loss": 0.7704,
+      "step": 1019
+    },
+    {
+      "epoch": 0.31466913465987967,
+      "grad_norm": 0.0,
+      "learning_rate": 1.6045566514297184e-05,
+      "loss": 0.8248,
+      "step": 1020
+    },
+    {
+      "epoch": 0.314977633811507,
+      "grad_norm": 0.0,
+      "learning_rate": 1.603760144084879e-05,
+      "loss": 0.8371,
+      "step": 1021
+    },
+    {
+      "epoch": 0.31528613296313435,
+      "grad_norm": 0.0,
+      "learning_rate": 1.602963033520518e-05,
+      "loss": 0.829,
+      "step": 1022
+    },
+    {
+      "epoch": 0.3155946321147617,
+      "grad_norm": 0.0,
+      "learning_rate": 1.602165320533032e-05,
+      "loss": 0.7444,
+      "step": 1023
+    },
+    {
+      "epoch": 0.315903131266389,
+      "grad_norm": 0.0,
+      "learning_rate": 1.6013670059194203e-05,
+      "loss": 0.7885,
+      "step": 1024
+    },
+    {
+      "epoch": 0.31621163041801637,
+      "grad_norm": 0.0,
+      "learning_rate": 1.6005680904772822e-05,
+      "loss": 0.7537,
+      "step": 1025
+    },
+    {
+      "epoch": 0.3165201295696437,
+      "grad_norm": 0.0,
+      "learning_rate": 1.5997685750048183e-05,
+      "loss": 0.8151,
+      "step": 1026
+    },
+    {
+      "epoch": 0.31682862872127104,
+      "grad_norm": 0.0,
+      "learning_rate": 1.5989684603008274e-05,
+      "loss": 0.7634,
+      "step": 1027
+    },
+    {
+      "epoch": 0.31713712787289833,
+      "grad_norm": 0.0,
+      "learning_rate": 1.5981677471647085e-05,
+      "loss": 0.802,
+      "step": 1028
+    },
+    {
+      "epoch": 0.31744562702452567,
+      "grad_norm": 0.0,
+      "learning_rate": 1.5973664363964573e-05,
+      "loss": 0.811,
+      "step": 1029
+    },
+    {
+      "epoch": 0.317754126176153,
+      "grad_norm": 0.0,
+      "learning_rate": 1.5965645287966674e-05,
+      "loss": 0.8126,
+      "step": 1030
+    },
+    {
+      "epoch": 0.31806262532778035,
+      "grad_norm": 0.0,
+      "learning_rate": 1.5957620251665272e-05,
+      "loss": 0.8259,
+      "step": 1031
+    },
+    {
+      "epoch": 0.3183711244794077,
+      "grad_norm": 0.0,
+      "learning_rate": 1.594958926307824e-05,
+      "loss": 0.7898,
+      "step": 1032
+    },
+    {
+      "epoch": 0.318679623631035,
+      "grad_norm": 0.0,
+      "learning_rate": 1.5941552330229352e-05,
+      "loss": 0.8517,
+      "step": 1033
+    },
+    {
+      "epoch": 0.31898812278266236,
+      "grad_norm": 0.0,
+      "learning_rate": 1.593350946114836e-05,
+      "loss": 0.7696,
+      "step": 1034
+    },
+    {
+      "epoch": 0.3192966219342897,
+      "grad_norm": 0.0,
+      "learning_rate": 1.592546066387092e-05,
+      "loss": 0.7539,
+      "step": 1035
+    },
+    {
+      "epoch": 0.319605121085917,
+      "grad_norm": 0.0,
+      "learning_rate": 1.5917405946438635e-05,
+      "loss": 0.9824,
+      "step": 1036
+    },
+    {
+      "epoch": 0.3199136202375443,
+      "grad_norm": 0.0,
+      "learning_rate": 1.5909345316899e-05,
+      "loss": 0.862,
+      "step": 1037
+    },
+    {
+      "epoch": 0.32022211938917167,
+      "grad_norm": 0.0,
+      "learning_rate": 1.590127878330543e-05,
+      "loss": 0.8534,
+      "step": 1038
+    },
+    {
+      "epoch": 0.320530618540799,
+      "grad_norm": 0.0,
+      "learning_rate": 1.5893206353717234e-05,
+      "loss": 0.8594,
+      "step": 1039
+    },
+    {
+      "epoch": 0.32083911769242635,
+      "grad_norm": 0.0,
+      "learning_rate": 1.5885128036199615e-05,
+      "loss": 0.8474,
+      "step": 1040
+    },
+    {
+      "epoch": 0.3211476168440537,
+      "grad_norm": 0.0,
+      "learning_rate": 1.587704383882366e-05,
+      "loss": 0.7864,
+      "step": 1041
+    },
+    {
+      "epoch": 0.321456115995681,
+      "grad_norm": 0.0,
+      "learning_rate": 1.586895376966632e-05,
+      "loss": 0.9179,
+      "step": 1042
+    },
+    {
+      "epoch": 0.32176461514730836,
+      "grad_norm": 0.0,
+      "learning_rate": 1.5860857836810427e-05,
+      "loss": 0.7946,
+      "step": 1043
+    },
+    {
+      "epoch": 0.3220731142989357,
+      "grad_norm": 0.0,
+      "learning_rate": 1.585275604834466e-05,
+      "loss": 0.8053,
+      "step": 1044
+    },
+    {
+      "epoch": 0.322381613450563,
+      "grad_norm": 0.0,
+      "learning_rate": 1.584464841236356e-05,
+      "loss": 0.8583,
+      "step": 1045
+    },
+    {
+      "epoch": 0.3226901126021903,
+      "grad_norm": 0.0,
+      "learning_rate": 1.5836534936967493e-05,
+      "loss": 0.796,
+      "step": 1046
+    },
+    {
+      "epoch": 0.32299861175381767,
+      "grad_norm": 0.0,
+      "learning_rate": 1.5828415630262678e-05,
+      "loss": 0.864,
+      "step": 1047
+    },
+    {
+      "epoch": 0.323307110905445,
+      "grad_norm": 0.0,
+      "learning_rate": 1.5820290500361147e-05,
+      "loss": 0.791,
+      "step": 1048
+    },
+    {
+      "epoch": 0.32361561005707234,
+      "grad_norm": 0.0,
+      "learning_rate": 1.5812159555380752e-05,
+      "loss": 0.8613,
+      "step": 1049
+    },
+    {
+      "epoch": 0.3239241092086997,
+      "grad_norm": 0.0,
+      "learning_rate": 1.5804022803445164e-05,
+      "loss": 0.9215,
+      "step": 1050
+    },
+    {
+      "epoch": 0.324232608360327,
+      "grad_norm": 0.0,
+      "learning_rate": 1.5795880252683848e-05,
+      "loss": 0.8645,
+      "step": 1051
+    },
+    {
+      "epoch": 0.32454110751195436,
+      "grad_norm": 0.0,
+      "learning_rate": 1.5787731911232057e-05,
+      "loss": 0.8505,
+      "step": 1052
+    },
+    {
+      "epoch": 0.3248496066635817,
+      "grad_norm": 0.0,
+      "learning_rate": 1.5779577787230843e-05,
+      "loss": 0.8863,
+      "step": 1053
+    },
+    {
+      "epoch": 0.325158105815209,
+      "grad_norm": 0.0,
+      "learning_rate": 1.5771417888827026e-05,
+      "loss": 0.7677,
+      "step": 1054
+    },
+    {
+      "epoch": 0.3254666049668363,
+      "grad_norm": 0.0,
+      "learning_rate": 1.5763252224173196e-05,
+      "loss": 0.7598,
+      "step": 1055
+    },
+    {
+      "epoch": 0.32577510411846367,
+      "grad_norm": 0.0,
+      "learning_rate": 1.575508080142771e-05,
+      "loss": 0.8464,
+      "step": 1056
+    },
+    {
+      "epoch": 0.326083603270091,
+      "grad_norm": 0.0,
+      "learning_rate": 1.5746903628754672e-05,
+      "loss": 0.7798,
+      "step": 1057
+    },
+    {
+      "epoch": 0.32639210242171834,
+      "grad_norm": 0.0,
+      "learning_rate": 1.5738720714323935e-05,
+      "loss": 0.8526,
+      "step": 1058
+    },
+    {
+      "epoch": 0.3267006015733457,
+      "grad_norm": 0.0,
+      "learning_rate": 1.573053206631108e-05,
+      "loss": 0.8373,
+      "step": 1059
+    },
+    {
+      "epoch": 0.327009100724973,
+      "grad_norm": 0.0,
+      "learning_rate": 1.5722337692897428e-05,
+      "loss": 0.7411,
+      "step": 1060
+    },
+    {
+      "epoch": 0.32731759987660036,
+      "grad_norm": 0.0,
+      "learning_rate": 1.571413760227001e-05,
+      "loss": 0.7517,
+      "step": 1061
+    },
+    {
+      "epoch": 0.32762609902822765,
+      "grad_norm": 0.0,
+      "learning_rate": 1.5705931802621583e-05,
+      "loss": 0.7368,
+      "step": 1062
+    },
+    {
+      "epoch": 0.327934598179855,
+      "grad_norm": 0.0,
+      "learning_rate": 1.569772030215059e-05,
+      "loss": 0.8144,
+      "step": 1063
+    },
+    {
+      "epoch": 0.3282430973314823,
+      "grad_norm": 0.0,
+      "learning_rate": 1.5689503109061185e-05,
+      "loss": 0.8972,
+      "step": 1064
+    },
+    {
+      "epoch": 0.32855159648310966,
+      "grad_norm": 0.0,
+      "learning_rate": 1.5681280231563196e-05,
+      "loss": 0.8487,
+      "step": 1065
+    },
+    {
+      "epoch": 0.328860095634737,
+      "grad_norm": 0.0,
+      "learning_rate": 1.5673051677872143e-05,
+      "loss": 0.8097,
+      "step": 1066
+    },
+    {
+      "epoch": 0.32916859478636434,
+      "grad_norm": 0.0,
+      "learning_rate": 1.566481745620921e-05,
+      "loss": 0.8631,
+      "step": 1067
+    },
+    {
+      "epoch": 0.3294770939379917,
+      "grad_norm": 0.0,
+      "learning_rate": 1.565657757480125e-05,
+      "loss": 0.7832,
+      "step": 1068
+    },
+    {
+      "epoch": 0.329785593089619,
+      "grad_norm": 0.0,
+      "learning_rate": 1.564833204188076e-05,
+      "loss": 0.8752,
+      "step": 1069
+    },
+    {
+      "epoch": 0.33009409224124636,
+      "grad_norm": 0.0,
+      "learning_rate": 1.5640080865685888e-05,
+      "loss": 0.8046,
+      "step": 1070
+    },
+    {
+      "epoch": 0.33040259139287365,
+      "grad_norm": 0.0,
+      "learning_rate": 1.563182405446043e-05,
+      "loss": 0.8132,
+      "step": 1071
+    },
+    {
+      "epoch": 0.330711090544501,
+      "grad_norm": 0.0,
+      "learning_rate": 1.5623561616453798e-05,
+      "loss": 0.8263,
+      "step": 1072
+    },
+    {
+      "epoch": 0.3310195896961283,
+      "grad_norm": 0.0,
+      "learning_rate": 1.5615293559921037e-05,
+      "loss": 0.8533,
+      "step": 1073
+    },
+    {
+      "epoch": 0.33132808884775566,
+      "grad_norm": 0.0,
+      "learning_rate": 1.5607019893122792e-05,
+      "loss": 0.7371,
+      "step": 1074
+    },
+    {
+      "epoch": 0.331636587999383,
+      "grad_norm": 0.0,
+      "learning_rate": 1.5598740624325325e-05,
+      "loss": 0.9294,
+      "step": 1075
+    },
+    {
+      "epoch": 0.33194508715101034,
+      "grad_norm": 0.0,
+      "learning_rate": 1.5590455761800494e-05,
+      "loss": 0.7979,
+      "step": 1076
+    },
+    {
+      "epoch": 0.3322535863026377,
+      "grad_norm": 0.0,
+      "learning_rate": 1.558216531382574e-05,
+      "loss": 0.8072,
+      "step": 1077
+    },
+    {
+      "epoch": 0.332562085454265,
+      "grad_norm": 0.0,
+      "learning_rate": 1.5573869288684087e-05,
+      "loss": 0.8374,
+      "step": 1078
+    },
+    {
+      "epoch": 0.33287058460589236,
+      "grad_norm": 0.0,
+      "learning_rate": 1.556556769466414e-05,
+      "loss": 0.7954,
+      "step": 1079
+    },
+    {
+      "epoch": 0.33317908375751965,
+      "grad_norm": 0.0,
+      "learning_rate": 1.5557260540060047e-05,
+      "loss": 0.7761,
+      "step": 1080
+    },
+    {
+      "epoch": 0.333487582909147,
+      "grad_norm": 0.0,
+      "learning_rate": 1.554894783317153e-05,
+      "loss": 0.8671,
+      "step": 1081
+    },
+    {
+      "epoch": 0.3337960820607743,
+      "grad_norm": 0.0,
+      "learning_rate": 1.554062958230385e-05,
+      "loss": 0.8552,
+      "step": 1082
+    },
+    {
+      "epoch": 0.33410458121240166,
+      "grad_norm": 0.0,
+      "learning_rate": 1.5532305795767817e-05,
+      "loss": 0.844,
+      "step": 1083
+    },
+    {
+      "epoch": 0.334413080364029,
+      "grad_norm": 0.0,
+      "learning_rate": 1.5523976481879754e-05,
+      "loss": 0.7294,
+      "step": 1084
+    },
+    {
+      "epoch": 0.33472157951565634,
+      "grad_norm": 0.0,
+      "learning_rate": 1.5515641648961526e-05,
+      "loss": 0.8419,
+      "step": 1085
+    },
+    {
+      "epoch": 0.3350300786672837,
+      "grad_norm": 0.0,
+      "learning_rate": 1.5507301305340496e-05,
+      "loss": 0.8752,
+      "step": 1086
+    },
+    {
+      "epoch": 0.335338577818911,
+      "grad_norm": 0.0,
+      "learning_rate": 1.549895545934954e-05,
+      "loss": 0.7934,
+      "step": 1087
+    },
+    {
+      "epoch": 0.3356470769705383,
+      "grad_norm": 0.0,
+      "learning_rate": 1.549060411932704e-05,
+      "loss": 0.928,
+      "step": 1088
+    },
+    {
+      "epoch": 0.33595557612216564,
+      "grad_norm": 0.0,
+      "learning_rate": 1.5482247293616843e-05,
+      "loss": 0.859,
+      "step": 1089
+    },
+    {
+      "epoch": 0.336264075273793,
+      "grad_norm": 0.0,
+      "learning_rate": 1.5473884990568298e-05,
+      "loss": 0.9184,
+      "step": 1090
+    },
+    {
+      "epoch": 0.3365725744254203,
+      "grad_norm": 0.0,
+      "learning_rate": 1.5465517218536228e-05,
+      "loss": 0.7502,
+      "step": 1091
+    },
+    {
+      "epoch": 0.33688107357704766,
+      "grad_norm": 0.0,
+      "learning_rate": 1.5457143985880905e-05,
+      "loss": 0.7997,
+      "step": 1092
+    },
+    {
+      "epoch": 0.337189572728675,
+      "grad_norm": 0.0,
+      "learning_rate": 1.5448765300968066e-05,
+      "loss": 0.7976,
+      "step": 1093
+    },
+    {
+      "epoch": 0.33749807188030234,
+      "grad_norm": 0.0,
+      "learning_rate": 1.544038117216889e-05,
+      "loss": 0.7606,
+      "step": 1094
+    },
+    {
+      "epoch": 0.3378065710319297,
+      "grad_norm": 0.0,
+      "learning_rate": 1.5431991607859997e-05,
+      "loss": 0.782,
+      "step": 1095
+    },
+    {
+      "epoch": 0.338115070183557,
+      "grad_norm": 0.0,
+      "learning_rate": 1.542359661642345e-05,
+      "loss": 0.7834,
+      "step": 1096
+    },
+    {
+      "epoch": 0.3384235693351843,
+      "grad_norm": 0.0,
+      "learning_rate": 1.5415196206246712e-05,
+      "loss": 0.8193,
+      "step": 1097
+    },
+    {
+      "epoch": 0.33873206848681164,
+      "grad_norm": 0.0,
+      "learning_rate": 1.5406790385722676e-05,
+      "loss": 0.8953,
+      "step": 1098
+    },
+    {
+      "epoch": 0.339040567638439,
+      "grad_norm": 0.0,
+      "learning_rate": 1.5398379163249636e-05,
+      "loss": 0.7241,
+      "step": 1099
+    },
+    {
+      "epoch": 0.3393490667900663,
+      "grad_norm": 0.0,
+      "learning_rate": 1.5389962547231286e-05,
+      "loss": 0.8359,
+      "step": 1100
+    },
+    {
+      "epoch": 0.33965756594169366,
+      "grad_norm": 0.0,
+      "learning_rate": 1.5381540546076694e-05,
+      "loss": 0.7672,
+      "step": 1101
+    },
+    {
+      "epoch": 0.339966065093321,
+      "grad_norm": 0.0,
+      "learning_rate": 1.5373113168200332e-05,
+      "loss": 0.7012,
+      "step": 1102
+    },
+    {
+      "epoch": 0.34027456424494834,
+      "grad_norm": 0.0,
+      "learning_rate": 1.536468042202203e-05,
+      "loss": 0.7484,
+      "step": 1103
+    },
+    {
+      "epoch": 0.3405830633965757,
+      "grad_norm": 0.0,
+      "learning_rate": 1.5356242315966974e-05,
+      "loss": 0.8494,
+      "step": 1104
+    },
+    {
+      "epoch": 0.340891562548203,
+      "grad_norm": 0.0,
+      "learning_rate": 1.5347798858465727e-05,
+      "loss": 0.8927,
+      "step": 1105
+    },
+    {
+      "epoch": 0.3412000616998303,
+      "grad_norm": 0.0,
+      "learning_rate": 1.5339350057954178e-05,
+      "loss": 0.8585,
+      "step": 1106
+    },
+    {
+      "epoch": 0.34150856085145764,
+      "grad_norm": 0.0,
+      "learning_rate": 1.5330895922873562e-05,
+      "loss": 0.8511,
+      "step": 1107
+    },
+    {
+      "epoch": 0.341817060003085,
+      "grad_norm": 0.0,
+      "learning_rate": 1.5322436461670445e-05,
+      "loss": 0.8373,
+      "step": 1108
+    },
+    {
+      "epoch": 0.3421255591547123,
+      "grad_norm": 0.0,
+      "learning_rate": 1.531397168279672e-05,
+      "loss": 0.7488,
+      "step": 1109
+    },
+    {
+      "epoch": 0.34243405830633966,
+      "grad_norm": 0.0,
+      "learning_rate": 1.5305501594709578e-05,
+      "loss": 0.78,
+      "step": 1110
+    },
+    {
+      "epoch": 0.342742557457967,
+      "grad_norm": 0.0,
+      "learning_rate": 1.5297026205871528e-05,
+      "loss": 0.8525,
+      "step": 1111
+    },
+    {
+      "epoch": 0.34305105660959434,
+      "grad_norm": 0.0,
+      "learning_rate": 1.5288545524750366e-05,
+      "loss": 0.7766,
+      "step": 1112
+    },
+    {
+      "epoch": 0.3433595557612217,
+      "grad_norm": 0.0,
+      "learning_rate": 1.5280059559819177e-05,
+      "loss": 0.7915,
+      "step": 1113
+    },
+    {
+      "epoch": 0.34366805491284896,
+      "grad_norm": 0.0,
+      "learning_rate": 1.5271568319556336e-05,
+      "loss": 0.8904,
+      "step": 1114
+    },
+    {
+      "epoch": 0.3439765540644763,
+      "grad_norm": 0.0,
+      "learning_rate": 1.5263071812445475e-05,
+      "loss": 0.7938,
+      "step": 1115
+    },
+    {
+      "epoch": 0.34428505321610364,
+      "grad_norm": 0.0,
+      "learning_rate": 1.525457004697549e-05,
+      "loss": 0.7644,
+      "step": 1116
+    },
+    {
+      "epoch": 0.344593552367731,
+      "grad_norm": 0.0,
+      "learning_rate": 1.524606303164054e-05,
+      "loss": 0.829,
+      "step": 1117
+    },
+    {
+      "epoch": 0.3449020515193583,
+      "grad_norm": 0.0,
+      "learning_rate": 1.5237550774940018e-05,
+      "loss": 0.8274,
+      "step": 1118
+    },
+    {
+      "epoch": 0.34521055067098566,
+      "grad_norm": 0.0,
+      "learning_rate": 1.522903328537856e-05,
+      "loss": 0.7864,
+      "step": 1119
+    },
+    {
+      "epoch": 0.345519049822613,
+      "grad_norm": 0.0,
+      "learning_rate": 1.522051057146603e-05,
+      "loss": 0.7428,
+      "step": 1120
+    },
+    {
+      "epoch": 0.34582754897424034,
+      "grad_norm": 0.0,
+      "learning_rate": 1.5211982641717509e-05,
+      "loss": 0.7813,
+      "step": 1121
+    },
+    {
+      "epoch": 0.3461360481258677,
+      "grad_norm": 0.0,
+      "learning_rate": 1.5203449504653294e-05,
+      "loss": 0.8042,
+      "step": 1122
+    },
+    {
+      "epoch": 0.34644454727749496,
+      "grad_norm": 0.0,
+      "learning_rate": 1.5194911168798876e-05,
+      "loss": 0.8232,
+      "step": 1123
+    },
+    {
+      "epoch": 0.3467530464291223,
+      "grad_norm": 0.0,
+      "learning_rate": 1.5186367642684952e-05,
+      "loss": 0.8271,
+      "step": 1124
+    },
+    {
+      "epoch": 0.34706154558074964,
+      "grad_norm": 0.0,
+      "learning_rate": 1.517781893484739e-05,
+      "loss": 0.8243,
+      "step": 1125
+    },
+    {
+      "epoch": 0.347370044732377,
+      "grad_norm": 0.0,
+      "learning_rate": 1.5169265053827246e-05,
+      "loss": 0.7355,
+      "step": 1126
+    },
+    {
+      "epoch": 0.3476785438840043,
+      "grad_norm": 0.0,
+      "learning_rate": 1.5160706008170744e-05,
+      "loss": 0.7534,
+      "step": 1127
+    },
+    {
+      "epoch": 0.34798704303563166,
+      "grad_norm": 0.0,
+      "learning_rate": 1.5152141806429268e-05,
+      "loss": 0.7316,
+      "step": 1128
+    },
+    {
+      "epoch": 0.348295542187259,
+      "grad_norm": 0.0,
+      "learning_rate": 1.5143572457159344e-05,
+      "loss": 0.8143,
+      "step": 1129
+    },
+    {
+      "epoch": 0.34860404133888634,
+      "grad_norm": 0.0,
+      "learning_rate": 1.5134997968922655e-05,
+      "loss": 0.7644,
+      "step": 1130
+    },
+    {
+      "epoch": 0.3489125404905137,
+      "grad_norm": 0.0,
+      "learning_rate": 1.5126418350286005e-05,
+      "loss": 0.8303,
+      "step": 1131
+    },
+    {
+      "epoch": 0.34922103964214096,
+      "grad_norm": 0.0,
+      "learning_rate": 1.5117833609821333e-05,
+      "loss": 0.7803,
+      "step": 1132
+    },
+    {
+      "epoch": 0.3495295387937683,
+      "grad_norm": 0.0,
+      "learning_rate": 1.5109243756105692e-05,
+      "loss": 0.8298,
+      "step": 1133
+    },
+    {
+      "epoch": 0.34983803794539564,
+      "grad_norm": 0.0,
+      "learning_rate": 1.510064879772125e-05,
+      "loss": 0.8462,
+      "step": 1134
+    },
+    {
+      "epoch": 0.350146537097023,
+      "grad_norm": 0.0,
+      "learning_rate": 1.5092048743255258e-05,
+      "loss": 0.8431,
+      "step": 1135
+    },
+    {
+      "epoch": 0.3504550362486503,
+      "grad_norm": 0.0,
+      "learning_rate": 1.5083443601300078e-05,
+      "loss": 0.8221,
+      "step": 1136
+    },
+    {
+      "epoch": 0.35076353540027766,
+      "grad_norm": 0.0,
+      "learning_rate": 1.5074833380453146e-05,
+      "loss": 0.8576,
+      "step": 1137
+    },
+    {
+      "epoch": 0.351072034551905,
+      "grad_norm": 0.0,
+      "learning_rate": 1.5066218089316972e-05,
+      "loss": 0.8344,
+      "step": 1138
+    },
+    {
+      "epoch": 0.35138053370353234,
+      "grad_norm": 0.0,
+      "learning_rate": 1.505759773649913e-05,
+      "loss": 0.8229,
+      "step": 1139
+    },
+    {
+      "epoch": 0.3516890328551596,
+      "grad_norm": 0.0,
+      "learning_rate": 1.5048972330612256e-05,
+      "loss": 0.7867,
+      "step": 1140
+    },
+    {
+      "epoch": 0.35199753200678696,
+      "grad_norm": 0.0,
+      "learning_rate": 1.5040341880274038e-05,
+      "loss": 0.8067,
+      "step": 1141
+    },
+    {
+      "epoch": 0.3523060311584143,
+      "grad_norm": 0.0,
+      "learning_rate": 1.5031706394107188e-05,
+      "loss": 0.8575,
+      "step": 1142
+    },
+    {
+      "epoch": 0.35261453031004164,
+      "grad_norm": 0.0,
+      "learning_rate": 1.502306588073947e-05,
+      "loss": 0.8686,
+      "step": 1143
+    },
+    {
+      "epoch": 0.352923029461669,
+      "grad_norm": 0.0,
+      "learning_rate": 1.5014420348803649e-05,
+      "loss": 0.7929,
+      "step": 1144
+    },
+    {
+      "epoch": 0.3532315286132963,
+      "grad_norm": 0.0,
+      "learning_rate": 1.5005769806937523e-05,
+      "loss": 0.6688,
+      "step": 1145
+    },
+    {
+      "epoch": 0.35354002776492366,
+      "grad_norm": 0.0,
+      "learning_rate": 1.4997114263783887e-05,
+      "loss": 0.8364,
+      "step": 1146
+    },
+    {
+      "epoch": 0.353848526916551,
+      "grad_norm": 0.0,
+      "learning_rate": 1.4988453727990537e-05,
+      "loss": 0.7549,
+      "step": 1147
+    },
+    {
+      "epoch": 0.35415702606817834,
+      "grad_norm": 0.0,
+      "learning_rate": 1.4979788208210249e-05,
+      "loss": 0.8211,
+      "step": 1148
+    },
+    {
+      "epoch": 0.3544655252198056,
+      "grad_norm": 0.0,
+      "learning_rate": 1.4971117713100785e-05,
+      "loss": 0.8714,
+      "step": 1149
+    },
+    {
+      "epoch": 0.35477402437143296,
+      "grad_norm": 0.0,
+      "learning_rate": 1.4962442251324876e-05,
+      "loss": 0.8186,
+      "step": 1150
+    },
+    {
+      "epoch": 0.3550825235230603,
+      "grad_norm": 0.0,
+      "learning_rate": 1.4953761831550212e-05,
+      "loss": 0.7695,
+      "step": 1151
+    },
+    {
+      "epoch": 0.35539102267468764,
+      "grad_norm": 0.0,
+      "learning_rate": 1.4945076462449448e-05,
+      "loss": 0.8127,
+      "step": 1152
+    },
+    {
+      "epoch": 0.355699521826315,
+      "grad_norm": 0.0,
+      "learning_rate": 1.493638615270017e-05,
+      "loss": 0.7623,
+      "step": 1153
+    },
+    {
+      "epoch": 0.3560080209779423,
+      "grad_norm": 0.0,
+      "learning_rate": 1.4927690910984911e-05,
+      "loss": 0.7749,
+      "step": 1154
+    },
+    {
+      "epoch": 0.35631652012956966,
+      "grad_norm": 0.0,
+      "learning_rate": 1.4918990745991122e-05,
+      "loss": 0.8062,
+      "step": 1155
+    },
+    {
+      "epoch": 0.356625019281197,
+      "grad_norm": 0.0,
+      "learning_rate": 1.491028566641118e-05,
+      "loss": 0.8323,
+      "step": 1156
+    },
+    {
+      "epoch": 0.35693351843282434,
+      "grad_norm": 0.0,
+      "learning_rate": 1.4901575680942368e-05,
+      "loss": 0.791,
+      "step": 1157
+    },
+    {
+      "epoch": 0.3572420175844516,
+      "grad_norm": 0.0,
+      "learning_rate": 1.4892860798286875e-05,
+      "loss": 0.8861,
+      "step": 1158
+    },
+    {
+      "epoch": 0.35755051673607896,
+      "grad_norm": 0.0,
+      "learning_rate": 1.4884141027151778e-05,
+      "loss": 0.8189,
+      "step": 1159
+    },
+    {
+      "epoch": 0.3578590158877063,
+      "grad_norm": 0.0,
+      "learning_rate": 1.487541637624904e-05,
+      "loss": 1.1021,
+      "step": 1160
+    },
+    {
+      "epoch": 0.35816751503933364,
+      "grad_norm": 0.0,
+      "learning_rate": 1.4866686854295502e-05,
+      "loss": 0.8152,
+      "step": 1161
+    },
+    {
+      "epoch": 0.358476014190961,
+      "grad_norm": 0.0,
+      "learning_rate": 1.4857952470012871e-05,
+      "loss": 0.7575,
+      "step": 1162
+    },
+    {
+      "epoch": 0.3587845133425883,
+      "grad_norm": 0.0,
+      "learning_rate": 1.4849213232127701e-05,
+      "loss": 0.681,
+      "step": 1163
+    },
+    {
+      "epoch": 0.35909301249421566,
+      "grad_norm": 0.0,
+      "learning_rate": 1.4840469149371414e-05,
+      "loss": 0.7734,
+      "step": 1164
+    },
+    {
+      "epoch": 0.359401511645843,
+      "grad_norm": 0.0,
+      "learning_rate": 1.483172023048026e-05,
+      "loss": 0.7589,
+      "step": 1165
+    },
+    {
+      "epoch": 0.3597100107974703,
+      "grad_norm": 0.0,
+      "learning_rate": 1.4822966484195323e-05,
+      "loss": 0.8426,
+      "step": 1166
+    },
+    {
+      "epoch": 0.3600185099490976,
+      "grad_norm": 0.0,
+      "learning_rate": 1.4814207919262513e-05,
+      "loss": 0.7434,
+      "step": 1167
+    },
+    {
+      "epoch": 0.36032700910072496,
+      "grad_norm": 0.0,
+      "learning_rate": 1.4805444544432547e-05,
+      "loss": 0.7557,
+      "step": 1168
+    },
+    {
+      "epoch": 0.3606355082523523,
+      "grad_norm": 0.0,
+      "learning_rate": 1.4796676368460963e-05,
+      "loss": 0.7985,
+      "step": 1169
+    },
+    {
+      "epoch": 0.36094400740397964,
+      "grad_norm": 0.0,
+      "learning_rate": 1.4787903400108074e-05,
+      "loss": 0.7233,
+      "step": 1170
+    },
+    {
+      "epoch": 0.361252506555607,
+      "grad_norm": 0.0,
+      "learning_rate": 1.4779125648139002e-05,
+      "loss": 0.774,
+      "step": 1171
+    },
+    {
+      "epoch": 0.3615610057072343,
+      "grad_norm": 0.0,
+      "learning_rate": 1.4770343121323633e-05,
+      "loss": 0.783,
+      "step": 1172
+    },
+    {
+      "epoch": 0.36186950485886166,
+      "grad_norm": 0.0,
+      "learning_rate": 1.4761555828436635e-05,
+      "loss": 0.7645,
+      "step": 1173
+    },
+    {
+      "epoch": 0.362178004010489,
+      "grad_norm": 0.0,
+      "learning_rate": 1.4752763778257427e-05,
+      "loss": 0.8663,
+      "step": 1174
+    },
+    {
+      "epoch": 0.3624865031621163,
+      "grad_norm": 0.0,
+      "learning_rate": 1.474396697957019e-05,
+      "loss": 0.7965,
+      "step": 1175
+    },
+    {
+      "epoch": 0.3627950023137436,
+      "grad_norm": 0.0,
+      "learning_rate": 1.4735165441163846e-05,
+      "loss": 0.8533,
+      "step": 1176
+    },
+    {
+      "epoch": 0.36310350146537096,
+      "grad_norm": 0.0,
+      "learning_rate": 1.472635917183205e-05,
+      "loss": 0.8312,
+      "step": 1177
+    },
+    {
+      "epoch": 0.3634120006169983,
+      "grad_norm": 0.0,
+      "learning_rate": 1.4717548180373187e-05,
+      "loss": 0.8175,
+      "step": 1178
+    },
+    {
+      "epoch": 0.36372049976862564,
+      "grad_norm": 0.0,
+      "learning_rate": 1.4708732475590361e-05,
+      "loss": 0.9286,
+      "step": 1179
+    },
+    {
+      "epoch": 0.364028998920253,
+      "grad_norm": 0.0,
+      "learning_rate": 1.4699912066291383e-05,
+      "loss": 0.6973,
+      "step": 1180
+    },
+    {
+      "epoch": 0.3643374980718803,
+      "grad_norm": 0.0,
+      "learning_rate": 1.4691086961288758e-05,
+      "loss": 0.7972,
+      "step": 1181
+    },
+    {
+      "epoch": 0.36464599722350766,
+      "grad_norm": 0.0,
+      "learning_rate": 1.4682257169399697e-05,
+      "loss": 0.7783,
+      "step": 1182
+    },
+    {
+      "epoch": 0.364954496375135,
+      "grad_norm": 0.0,
+      "learning_rate": 1.4673422699446078e-05,
+      "loss": 0.9437,
+      "step": 1183
+    },
+    {
+      "epoch": 0.3652629955267623,
+      "grad_norm": 0.0,
+      "learning_rate": 1.4664583560254465e-05,
+      "loss": 0.8215,
+      "step": 1184
+    },
+    {
+      "epoch": 0.3655714946783896,
+      "grad_norm": 0.0,
+      "learning_rate": 1.4655739760656082e-05,
+      "loss": 0.7997,
+      "step": 1185
+    },
+    {
+      "epoch": 0.36587999383001696,
+      "grad_norm": 0.0,
+      "learning_rate": 1.464689130948681e-05,
+      "loss": 0.7802,
+      "step": 1186
+    },
+    {
+      "epoch": 0.3661884929816443,
+      "grad_norm": 0.0,
+      "learning_rate": 1.4638038215587176e-05,
+      "loss": 0.8049,
+      "step": 1187
+    },
+    {
+      "epoch": 0.36649699213327164,
+      "grad_norm": 0.0,
+      "learning_rate": 1.4629180487802348e-05,
+      "loss": 0.8285,
+      "step": 1188
+    },
+    {
+      "epoch": 0.366805491284899,
+      "grad_norm": 0.0,
+      "learning_rate": 1.4620318134982114e-05,
+      "loss": 0.743,
+      "step": 1189
+    },
+    {
+      "epoch": 0.3671139904365263,
+      "grad_norm": 0.0,
+      "learning_rate": 1.4611451165980905e-05,
+      "loss": 0.7363,
+      "step": 1190
+    },
+    {
+      "epoch": 0.36742248958815366,
+      "grad_norm": 0.0,
+      "learning_rate": 1.4602579589657742e-05,
+      "loss": 0.8358,
+      "step": 1191
+    },
+    {
+      "epoch": 0.36773098873978094,
+      "grad_norm": 0.0,
+      "learning_rate": 1.4593703414876262e-05,
+      "loss": 0.7217,
+      "step": 1192
+    },
+    {
+      "epoch": 0.3680394878914083,
+      "grad_norm": 0.0,
+      "learning_rate": 1.4584822650504685e-05,
+      "loss": 0.7513,
+      "step": 1193
+    },
+    {
+      "epoch": 0.3683479870430356,
+      "grad_norm": 0.0,
+      "learning_rate": 1.4575937305415829e-05,
+      "loss": 0.8975,
+      "step": 1194
+    },
+    {
+      "epoch": 0.36865648619466296,
+      "grad_norm": 0.0,
+      "learning_rate": 1.4567047388487077e-05,
+      "loss": 0.8446,
+      "step": 1195
+    },
+    {
+      "epoch": 0.3689649853462903,
+      "grad_norm": 0.0,
+      "learning_rate": 1.4558152908600394e-05,
+      "loss": 0.7981,
+      "step": 1196
+    },
+    {
+      "epoch": 0.36927348449791764,
+      "grad_norm": 0.0,
+      "learning_rate": 1.4549253874642289e-05,
+      "loss": 0.749,
+      "step": 1197
+    },
+    {
+      "epoch": 0.369581983649545,
+      "grad_norm": 0.0,
+      "learning_rate": 1.4540350295503834e-05,
+      "loss": 0.8389,
+      "step": 1198
+    },
+    {
+      "epoch": 0.3698904828011723,
+      "grad_norm": 0.0,
+      "learning_rate": 1.4531442180080625e-05,
+      "loss": 0.8098,
+      "step": 1199
+    },
+    {
+      "epoch": 0.37019898195279966,
+      "grad_norm": 0.0,
+      "learning_rate": 1.4522529537272813e-05,
+      "loss": 0.84,
+      "step": 1200
+    },
+    {
+      "epoch": 0.37050748110442694,
+      "grad_norm": 0.0,
+      "learning_rate": 1.451361237598505e-05,
+      "loss": 0.7923,
+      "step": 1201
+    },
+    {
+      "epoch": 0.3708159802560543,
+      "grad_norm": 0.0,
+      "learning_rate": 1.4504690705126519e-05,
+      "loss": 0.7732,
+      "step": 1202
+    },
+    {
+      "epoch": 0.3711244794076816,
+      "grad_norm": 0.0,
+      "learning_rate": 1.4495764533610902e-05,
+      "loss": 0.8072,
+      "step": 1203
+    },
+    {
+      "epoch": 0.37143297855930896,
+      "grad_norm": 0.0,
+      "learning_rate": 1.4486833870356374e-05,
+      "loss": 0.7858,
+      "step": 1204
+    },
+    {
+      "epoch": 0.3717414777109363,
+      "grad_norm": 0.0,
+      "learning_rate": 1.4477898724285603e-05,
+      "loss": 0.7929,
+      "step": 1205
+    },
+    {
+      "epoch": 0.37204997686256364,
+      "grad_norm": 0.0,
+      "learning_rate": 1.4468959104325737e-05,
+      "loss": 0.7335,
+      "step": 1206
+    },
+    {
+      "epoch": 0.372358476014191,
+      "grad_norm": 0.0,
+      "learning_rate": 1.446001501940839e-05,
+      "loss": 0.9012,
+      "step": 1207
+    },
+    {
+      "epoch": 0.3726669751658183,
+      "grad_norm": 0.0,
+      "learning_rate": 1.4451066478469633e-05,
+      "loss": 0.8734,
+      "step": 1208
+    },
+    {
+      "epoch": 0.37297547431744565,
+      "grad_norm": 0.0,
+      "learning_rate": 1.4442113490450002e-05,
+      "loss": 0.8447,
+      "step": 1209
+    },
+    {
+      "epoch": 0.37328397346907294,
+      "grad_norm": 0.0,
+      "learning_rate": 1.4433156064294465e-05,
+      "loss": 0.7566,
+      "step": 1210
+    },
+    {
+      "epoch": 0.3735924726207003,
+      "grad_norm": 0.0,
+      "learning_rate": 1.4424194208952427e-05,
+      "loss": 0.8284,
+      "step": 1211
+    },
+    {
+      "epoch": 0.3739009717723276,
+      "grad_norm": 0.0,
+      "learning_rate": 1.4415227933377715e-05,
+      "loss": 0.8023,
+      "step": 1212
+    },
+    {
+      "epoch": 0.37420947092395496,
+      "grad_norm": 0.0,
+      "learning_rate": 1.4406257246528584e-05,
+      "loss": 0.8064,
+      "step": 1213
+    },
+    {
+      "epoch": 0.3745179700755823,
+      "grad_norm": 0.0,
+      "learning_rate": 1.4397282157367682e-05,
+      "loss": 0.7575,
+      "step": 1214
+    },
+    {
+      "epoch": 0.37482646922720964,
+      "grad_norm": 0.0,
+      "learning_rate": 1.4388302674862065e-05,
+      "loss": 0.8027,
+      "step": 1215
+    },
+    {
+      "epoch": 0.375134968378837,
+      "grad_norm": 0.0,
+      "learning_rate": 1.4379318807983172e-05,
+      "loss": 0.8194,
+      "step": 1216
+    },
+    {
+      "epoch": 0.3754434675304643,
+      "grad_norm": 0.0,
+      "learning_rate": 1.4370330565706826e-05,
+      "loss": 0.8225,
+      "step": 1217
+    },
+    {
+      "epoch": 0.3757519666820916,
+      "grad_norm": 0.0,
+      "learning_rate": 1.4361337957013227e-05,
+      "loss": 0.7896,
+      "step": 1218
+    },
+    {
+      "epoch": 0.37606046583371894,
+      "grad_norm": 0.0,
+      "learning_rate": 1.4352340990886924e-05,
+      "loss": 0.7724,
+      "step": 1219
+    },
+    {
+      "epoch": 0.3763689649853463,
+      "grad_norm": 0.0,
+      "learning_rate": 1.434333967631683e-05,
+      "loss": 0.7446,
+      "step": 1220
+    },
+    {
+      "epoch": 0.3766774641369736,
+      "grad_norm": 0.0,
+      "learning_rate": 1.4334334022296196e-05,
+      "loss": 0.8344,
+      "step": 1221
+    },
+    {
+      "epoch": 0.37698596328860096,
+      "grad_norm": 0.0,
+      "learning_rate": 1.432532403782262e-05,
+      "loss": 0.7863,
+      "step": 1222
+    },
+    {
+      "epoch": 0.3772944624402283,
+      "grad_norm": 0.0,
+      "learning_rate": 1.431630973189801e-05,
+      "loss": 0.8061,
+      "step": 1223
+    },
+    {
+      "epoch": 0.37760296159185563,
+      "grad_norm": 0.0,
+      "learning_rate": 1.430729111352861e-05,
+      "loss": 0.7785,
+      "step": 1224
+    },
+    {
+      "epoch": 0.377911460743483,
+      "grad_norm": 0.0,
+      "learning_rate": 1.4298268191724951e-05,
+      "loss": 0.8627,
+      "step": 1225
+    },
+    {
+      "epoch": 0.3782199598951103,
+      "grad_norm": 0.0,
+      "learning_rate": 1.4289240975501885e-05,
+      "loss": 0.714,
+      "step": 1226
+    },
+    {
+      "epoch": 0.3785284590467376,
+      "grad_norm": 0.0,
+      "learning_rate": 1.4280209473878541e-05,
+      "loss": 0.8796,
+      "step": 1227
+    },
+    {
+      "epoch": 0.37883695819836494,
+      "grad_norm": 0.0,
+      "learning_rate": 1.4271173695878335e-05,
+      "loss": 0.8085,
+      "step": 1228
+    },
+    {
+      "epoch": 0.3791454573499923,
+      "grad_norm": 0.0,
+      "learning_rate": 1.4262133650528951e-05,
+      "loss": 0.7804,
+      "step": 1229
+    },
+    {
+      "epoch": 0.3794539565016196,
+      "grad_norm": 0.0,
+      "learning_rate": 1.4253089346862346e-05,
+      "loss": 0.8191,
+      "step": 1230
+    },
+    {
+      "epoch": 0.37976245565324696,
+      "grad_norm": 0.0,
+      "learning_rate": 1.4244040793914717e-05,
+      "loss": 0.7658,
+      "step": 1231
+    },
+    {
+      "epoch": 0.3800709548048743,
+      "grad_norm": 0.0,
+      "learning_rate": 1.423498800072652e-05,
+      "loss": 0.7949,
+      "step": 1232
+    },
+    {
+      "epoch": 0.38037945395650163,
+      "grad_norm": 0.0,
+      "learning_rate": 1.422593097634244e-05,
+      "loss": 0.8312,
+      "step": 1233
+    },
+    {
+      "epoch": 0.380687953108129,
+      "grad_norm": 0.0,
+      "learning_rate": 1.4216869729811393e-05,
+      "loss": 0.7824,
+      "step": 1234
+    },
+    {
+      "epoch": 0.3809964522597563,
+      "grad_norm": 0.0,
+      "learning_rate": 1.4207804270186514e-05,
+      "loss": 0.8182,
+      "step": 1235
+    },
+    {
+      "epoch": 0.3813049514113836,
+      "grad_norm": 0.0,
+      "learning_rate": 1.419873460652514e-05,
+      "loss": 0.8145,
+      "step": 1236
+    },
+    {
+      "epoch": 0.38161345056301094,
+      "grad_norm": 0.0,
+      "learning_rate": 1.4189660747888816e-05,
+      "loss": 0.7986,
+      "step": 1237
+    },
+    {
+      "epoch": 0.3819219497146383,
+      "grad_norm": 0.0,
+      "learning_rate": 1.4180582703343276e-05,
+      "loss": 0.7957,
+      "step": 1238
+    },
+    {
+      "epoch": 0.3822304488662656,
+      "grad_norm": 0.0,
+      "learning_rate": 1.4171500481958433e-05,
+      "loss": 0.8269,
+      "step": 1239
+    },
+    {
+      "epoch": 0.38253894801789295,
+      "grad_norm": 0.0,
+      "learning_rate": 1.416241409280838e-05,
+      "loss": 0.7489,
+      "step": 1240
+    },
+    {
+      "epoch": 0.3828474471695203,
+      "grad_norm": 0.0,
+      "learning_rate": 1.4153323544971371e-05,
+      "loss": 0.7288,
+      "step": 1241
+    },
+    {
+      "epoch": 0.38315594632114763,
+      "grad_norm": 0.0,
+      "learning_rate": 1.4144228847529809e-05,
+      "loss": 0.9038,
+      "step": 1242
+    },
+    {
+      "epoch": 0.383464445472775,
+      "grad_norm": 0.0,
+      "learning_rate": 1.4135130009570251e-05,
+      "loss": 0.7882,
+      "step": 1243
+    },
+    {
+      "epoch": 0.38377294462440226,
+      "grad_norm": 0.0,
+      "learning_rate": 1.4126027040183382e-05,
+      "loss": 0.7397,
+      "step": 1244
+    },
+    {
+      "epoch": 0.3840814437760296,
+      "grad_norm": 0.0,
+      "learning_rate": 1.411691994846403e-05,
+      "loss": 0.7556,
+      "step": 1245
+    },
+    {
+      "epoch": 0.38438994292765694,
+      "grad_norm": 0.0,
+      "learning_rate": 1.4107808743511124e-05,
+      "loss": 0.7622,
+      "step": 1246
+    },
+    {
+      "epoch": 0.3846984420792843,
+      "grad_norm": 0.0,
+      "learning_rate": 1.4098693434427715e-05,
+      "loss": 0.9192,
+      "step": 1247
+    },
+    {
+      "epoch": 0.3850069412309116,
+      "grad_norm": 0.0,
+      "learning_rate": 1.4089574030320947e-05,
+      "loss": 0.7547,
+      "step": 1248
+    },
+    {
+      "epoch": 0.38531544038253895,
+      "grad_norm": 0.0,
+      "learning_rate": 1.4080450540302061e-05,
+      "loss": 0.8098,
+      "step": 1249
+    },
+    {
+      "epoch": 0.3856239395341663,
+      "grad_norm": 0.0,
+      "learning_rate": 1.4071322973486376e-05,
+      "loss": 0.7672,
+      "step": 1250
+    },
+    {
+      "epoch": 0.38593243868579363,
+      "grad_norm": 0.0,
+      "learning_rate": 1.4062191338993288e-05,
+      "loss": 0.7344,
+      "step": 1251
+    },
+    {
+      "epoch": 0.386240937837421,
+      "grad_norm": 0.0,
+      "learning_rate": 1.405305564594625e-05,
+      "loss": 0.7607,
+      "step": 1252
+    },
+    {
+      "epoch": 0.38654943698904826,
+      "grad_norm": 0.0,
+      "learning_rate": 1.4043915903472777e-05,
+      "loss": 0.7472,
+      "step": 1253
+    },
+    {
+      "epoch": 0.3868579361406756,
+      "grad_norm": 0.0,
+      "learning_rate": 1.4034772120704431e-05,
+      "loss": 0.9083,
+      "step": 1254
+    },
+    {
+      "epoch": 0.38716643529230294,
+      "grad_norm": 0.0,
+      "learning_rate": 1.4025624306776806e-05,
+      "loss": 0.8289,
+      "step": 1255
+    },
+    {
+      "epoch": 0.3874749344439303,
+      "grad_norm": 0.0,
+      "learning_rate": 1.401647247082952e-05,
+      "loss": 0.7318,
+      "step": 1256
+    },
+    {
+      "epoch": 0.3877834335955576,
+      "grad_norm": 0.0,
+      "learning_rate": 1.400731662200622e-05,
+      "loss": 0.8786,
+      "step": 1257
+    },
+    {
+      "epoch": 0.38809193274718495,
+      "grad_norm": 0.0,
+      "learning_rate": 1.3998156769454549e-05,
+      "loss": 0.8683,
+      "step": 1258
+    },
+    {
+      "epoch": 0.3884004318988123,
+      "grad_norm": 0.0,
+      "learning_rate": 1.3988992922326166e-05,
+      "loss": 1.088,
+      "step": 1259
+    },
+    {
+      "epoch": 0.38870893105043963,
+      "grad_norm": 0.0,
+      "learning_rate": 1.3979825089776708e-05,
+      "loss": 0.8816,
+      "step": 1260
+    },
+    {
+      "epoch": 0.38901743020206697,
+      "grad_norm": 0.0,
+      "learning_rate": 1.39706532809658e-05,
+      "loss": 0.8478,
+      "step": 1261
+    },
+    {
+      "epoch": 0.38932592935369426,
+      "grad_norm": 0.0,
+      "learning_rate": 1.396147750505704e-05,
+      "loss": 0.7809,
+      "step": 1262
+    },
+    {
+      "epoch": 0.3896344285053216,
+      "grad_norm": 0.0,
+      "learning_rate": 1.395229777121798e-05,
+      "loss": 0.7805,
+      "step": 1263
+    },
+    {
+      "epoch": 0.38994292765694893,
+      "grad_norm": 0.0,
+      "learning_rate": 1.394311408862014e-05,
+      "loss": 0.8007,
+      "step": 1264
+    },
+    {
+      "epoch": 0.3902514268085763,
+      "grad_norm": 0.0,
+      "learning_rate": 1.3933926466438979e-05,
+      "loss": 0.7991,
+      "step": 1265
+    },
+    {
+      "epoch": 0.3905599259602036,
+      "grad_norm": 0.0,
+      "learning_rate": 1.3924734913853893e-05,
+      "loss": 0.7925,
+      "step": 1266
+    },
+    {
+      "epoch": 0.39086842511183095,
+      "grad_norm": 0.0,
+      "learning_rate": 1.3915539440048203e-05,
+      "loss": 0.8606,
+      "step": 1267
+    },
+    {
+      "epoch": 0.3911769242634583,
+      "grad_norm": 0.0,
+      "learning_rate": 1.3906340054209149e-05,
+      "loss": 0.693,
+      "step": 1268
+    },
+    {
+      "epoch": 0.39148542341508563,
+      "grad_norm": 0.0,
+      "learning_rate": 1.3897136765527879e-05,
+      "loss": 0.8273,
+      "step": 1269
+    },
+    {
+      "epoch": 0.3917939225667129,
+      "grad_norm": 0.0,
+      "learning_rate": 1.388792958319944e-05,
+      "loss": 0.7817,
+      "step": 1270
+    },
+    {
+      "epoch": 0.39210242171834025,
+      "grad_norm": 0.0,
+      "learning_rate": 1.3878718516422773e-05,
+      "loss": 0.7842,
+      "step": 1271
+    },
+    {
+      "epoch": 0.3924109208699676,
+      "grad_norm": 0.0,
+      "learning_rate": 1.3869503574400694e-05,
+      "loss": 0.7567,
+      "step": 1272
+    },
+    {
+      "epoch": 0.39271942002159493,
+      "grad_norm": 0.0,
+      "learning_rate": 1.3860284766339896e-05,
+      "loss": 0.7805,
+      "step": 1273
+    },
+    {
+      "epoch": 0.3930279191732223,
+      "grad_norm": 0.0,
+      "learning_rate": 1.385106210145093e-05,
+      "loss": 0.7966,
+      "step": 1274
+    },
+    {
+      "epoch": 0.3933364183248496,
+      "grad_norm": 0.0,
+      "learning_rate": 1.3841835588948208e-05,
+      "loss": 0.8702,
+      "step": 1275
+    },
+    {
+      "epoch": 0.39364491747647695,
+      "grad_norm": 0.0,
+      "learning_rate": 1.3832605238049972e-05,
+      "loss": 0.8001,
+      "step": 1276
+    },
+    {
+      "epoch": 0.3939534166281043,
+      "grad_norm": 0.0,
+      "learning_rate": 1.3823371057978312e-05,
+      "loss": 0.8095,
+      "step": 1277
+    },
+    {
+      "epoch": 0.39426191577973163,
+      "grad_norm": 0.0,
+      "learning_rate": 1.3814133057959143e-05,
+      "loss": 0.8103,
+      "step": 1278
+    },
+    {
+      "epoch": 0.3945704149313589,
+      "grad_norm": 0.0,
+      "learning_rate": 1.380489124722219e-05,
+      "loss": 0.7678,
+      "step": 1279
+    },
+    {
+      "epoch": 0.39487891408298625,
+      "grad_norm": 0.0,
+      "learning_rate": 1.3795645635000987e-05,
+      "loss": 0.8294,
+      "step": 1280
+    },
+    {
+      "epoch": 0.3951874132346136,
+      "grad_norm": 0.0,
+      "learning_rate": 1.3786396230532871e-05,
+      "loss": 0.7952,
+      "step": 1281
+    },
+    {
+      "epoch": 0.39549591238624093,
+      "grad_norm": 0.0,
+      "learning_rate": 1.377714304305896e-05,
+      "loss": 0.7974,
+      "step": 1282
+    },
+    {
+      "epoch": 0.3958044115378683,
+      "grad_norm": 0.0,
+      "learning_rate": 1.3767886081824158e-05,
+      "loss": 0.7424,
+      "step": 1283
+    },
+    {
+      "epoch": 0.3961129106894956,
+      "grad_norm": 0.0,
+      "learning_rate": 1.375862535607714e-05,
+      "loss": 0.7725,
+      "step": 1284
+    },
+    {
+      "epoch": 0.39642140984112295,
+      "grad_norm": 0.0,
+      "learning_rate": 1.3749360875070337e-05,
+      "loss": 0.7731,
+      "step": 1285
+    },
+    {
+      "epoch": 0.3967299089927503,
+      "grad_norm": 0.0,
+      "learning_rate": 1.3740092648059933e-05,
+      "loss": 0.7925,
+      "step": 1286
+    },
+    {
+      "epoch": 0.39703840814437763,
+      "grad_norm": 0.0,
+      "learning_rate": 1.3730820684305857e-05,
+      "loss": 0.7795,
+      "step": 1287
+    },
+    {
+      "epoch": 0.3973469072960049,
+      "grad_norm": 0.0,
+      "learning_rate": 1.3721544993071774e-05,
+      "loss": 0.7334,
+      "step": 1288
+    },
+    {
+      "epoch": 0.39765540644763225,
+      "grad_norm": 0.0,
+      "learning_rate": 1.3712265583625059e-05,
+      "loss": 0.7917,
+      "step": 1289
+    },
+    {
+      "epoch": 0.3979639055992596,
+      "grad_norm": 0.0,
+      "learning_rate": 1.3702982465236827e-05,
+      "loss": 0.7484,
+      "step": 1290
+    },
+    {
+      "epoch": 0.39827240475088693,
+      "grad_norm": 0.0,
+      "learning_rate": 1.3693695647181873e-05,
+      "loss": 0.8103,
+      "step": 1291
+    },
+    {
+      "epoch": 0.39858090390251427,
+      "grad_norm": 0.0,
+      "learning_rate": 1.3684405138738705e-05,
+      "loss": 0.759,
+      "step": 1292
+    },
+    {
+      "epoch": 0.3988894030541416,
+      "grad_norm": 0.0,
+      "learning_rate": 1.367511094918951e-05,
+      "loss": 0.8789,
+      "step": 1293
+    },
+    {
+      "epoch": 0.39919790220576895,
+      "grad_norm": 0.0,
+      "learning_rate": 1.3665813087820157e-05,
+      "loss": 0.8449,
+      "step": 1294
+    },
+    {
+      "epoch": 0.3995064013573963,
+      "grad_norm": 0.0,
+      "learning_rate": 1.3656511563920178e-05,
+      "loss": 0.7016,
+      "step": 1295
+    },
+    {
+      "epoch": 0.3998149005090236,
+      "grad_norm": 0.0,
+      "learning_rate": 1.3647206386782774e-05,
+      "loss": 0.7889,
+      "step": 1296
+    },
+    {
+      "epoch": 0.4001233996606509,
+      "grad_norm": 0.0,
+      "learning_rate": 1.3637897565704785e-05,
+      "loss": 0.7689,
+      "step": 1297
+    },
+    {
+      "epoch": 0.40043189881227825,
+      "grad_norm": 0.0,
+      "learning_rate": 1.3628585109986699e-05,
+      "loss": 0.7998,
+      "step": 1298
+    },
+    {
+      "epoch": 0.4007403979639056,
+      "grad_norm": 0.0,
+      "learning_rate": 1.3619269028932633e-05,
+      "loss": 0.763,
+      "step": 1299
+    },
+    {
+      "epoch": 0.40104889711553293,
+      "grad_norm": 0.0,
+      "learning_rate": 1.3609949331850323e-05,
+      "loss": 0.8125,
+      "step": 1300
+    },
+    {
+      "epoch": 0.40135739626716027,
+      "grad_norm": 0.0,
+      "learning_rate": 1.3600626028051127e-05,
+      "loss": 0.7986,
+      "step": 1301
+    },
+    {
+      "epoch": 0.4016658954187876,
+      "grad_norm": 0.0,
+      "learning_rate": 1.3591299126849994e-05,
+      "loss": 1.0926,
+      "step": 1302
+    },
+    {
+      "epoch": 0.40197439457041495,
+      "grad_norm": 0.0,
+      "learning_rate": 1.3581968637565481e-05,
+      "loss": 0.8111,
+      "step": 1303
+    },
+    {
+      "epoch": 0.4022828937220423,
+      "grad_norm": 0.0,
+      "learning_rate": 1.3572634569519718e-05,
+      "loss": 0.8507,
+      "step": 1304
+    },
+    {
+      "epoch": 0.4025913928736696,
+      "grad_norm": 0.0,
+      "learning_rate": 1.3563296932038416e-05,
+      "loss": 0.7634,
+      "step": 1305
+    },
+    {
+      "epoch": 0.4028998920252969,
+      "grad_norm": 0.0,
+      "learning_rate": 1.355395573445085e-05,
+      "loss": 0.8496,
+      "step": 1306
+    },
+    {
+      "epoch": 0.40320839117692425,
+      "grad_norm": 0.0,
+      "learning_rate": 1.3544610986089853e-05,
+      "loss": 0.808,
+      "step": 1307
+    },
+    {
+      "epoch": 0.4035168903285516,
+      "grad_norm": 0.0,
+      "learning_rate": 1.3535262696291805e-05,
+      "loss": 0.7791,
+      "step": 1308
+    },
+    {
+      "epoch": 0.40382538948017893,
+      "grad_norm": 0.0,
+      "learning_rate": 1.352591087439663e-05,
+      "loss": 0.8124,
+      "step": 1309
+    },
+    {
+      "epoch": 0.40413388863180627,
+      "grad_norm": 0.0,
+      "learning_rate": 1.3516555529747772e-05,
+      "loss": 0.7973,
+      "step": 1310
+    },
+    {
+      "epoch": 0.4044423877834336,
+      "grad_norm": 0.0,
+      "learning_rate": 1.3507196671692202e-05,
+      "loss": 0.8126,
+      "step": 1311
+    },
+    {
+      "epoch": 0.40475088693506095,
+      "grad_norm": 0.0,
+      "learning_rate": 1.3497834309580396e-05,
+      "loss": 0.78,
+      "step": 1312
+    },
+    {
+      "epoch": 0.4050593860866883,
+      "grad_norm": 0.0,
+      "learning_rate": 1.3488468452766328e-05,
+      "loss": 0.9037,
+      "step": 1313
+    },
+    {
+      "epoch": 0.4053678852383156,
+      "grad_norm": 0.0,
+      "learning_rate": 1.3479099110607475e-05,
+      "loss": 0.7797,
+      "step": 1314
+    },
+    {
+      "epoch": 0.4056763843899429,
+      "grad_norm": 0.0,
+      "learning_rate": 1.3469726292464788e-05,
+      "loss": 0.7638,
+      "step": 1315
+    },
+    {
+      "epoch": 0.40598488354157025,
+      "grad_norm": 0.0,
+      "learning_rate": 1.3460350007702691e-05,
+      "loss": 0.7942,
+      "step": 1316
+    },
+    {
+      "epoch": 0.4062933826931976,
+      "grad_norm": 0.0,
+      "learning_rate": 1.3450970265689074e-05,
+      "loss": 0.8901,
+      "step": 1317
+    },
+    {
+      "epoch": 0.40660188184482493,
+      "grad_norm": 0.0,
+      "learning_rate": 1.3441587075795281e-05,
+      "loss": 0.7406,
+      "step": 1318
+    },
+    {
+      "epoch": 0.40691038099645227,
+      "grad_norm": 0.0,
+      "learning_rate": 1.34322004473961e-05,
+      "loss": 0.7988,
+      "step": 1319
+    },
+    {
+      "epoch": 0.4072188801480796,
+      "grad_norm": 0.0,
+      "learning_rate": 1.3422810389869749e-05,
+      "loss": 0.7412,
+      "step": 1320
+    },
+    {
+      "epoch": 0.40752737929970695,
+      "grad_norm": 0.0,
+      "learning_rate": 1.3413416912597885e-05,
+      "loss": 0.8367,
+      "step": 1321
+    },
+    {
+      "epoch": 0.40783587845133423,
+      "grad_norm": 0.0,
+      "learning_rate": 1.340402002496557e-05,
+      "loss": 0.7749,
+      "step": 1322
+    },
+    {
+      "epoch": 0.40814437760296157,
+      "grad_norm": 0.0,
+      "learning_rate": 1.3394619736361278e-05,
+      "loss": 0.7756,
+      "step": 1323
+    },
+    {
+      "epoch": 0.4084528767545889,
+      "grad_norm": 0.0,
+      "learning_rate": 1.3385216056176882e-05,
+      "loss": 0.7485,
+      "step": 1324
+    },
+    {
+      "epoch": 0.40876137590621625,
+      "grad_norm": 0.0,
+      "learning_rate": 1.337580899380764e-05,
+      "loss": 0.732,
+      "step": 1325
+    },
+    {
+      "epoch": 0.4090698750578436,
+      "grad_norm": 0.0,
+      "learning_rate": 1.3366398558652192e-05,
+      "loss": 0.77,
+      "step": 1326
+    },
+    {
+      "epoch": 0.40937837420947093,
+      "grad_norm": 0.0,
+      "learning_rate": 1.3356984760112543e-05,
+      "loss": 0.7426,
+      "step": 1327
+    },
+    {
+      "epoch": 0.40968687336109827,
+      "grad_norm": 0.0,
+      "learning_rate": 1.3347567607594069e-05,
+      "loss": 0.7481,
+      "step": 1328
+    },
+    {
+      "epoch": 0.4099953725127256,
+      "grad_norm": 0.0,
+      "learning_rate": 1.3338147110505486e-05,
+      "loss": 0.8337,
+      "step": 1329
+    },
+    {
+      "epoch": 0.41030387166435295,
+      "grad_norm": 0.0,
+      "learning_rate": 1.3328723278258857e-05,
+      "loss": 0.8113,
+      "step": 1330
+    },
+    {
+      "epoch": 0.41061237081598023,
+      "grad_norm": 0.0,
+      "learning_rate": 1.3319296120269573e-05,
+      "loss": 0.8169,
+      "step": 1331
+    },
+    {
+      "epoch": 0.41092086996760757,
+      "grad_norm": 0.0,
+      "learning_rate": 1.3309865645956355e-05,
+      "loss": 0.8443,
+      "step": 1332
+    },
+    {
+      "epoch": 0.4112293691192349,
+      "grad_norm": 0.0,
+      "learning_rate": 1.3300431864741229e-05,
+      "loss": 0.7986,
+      "step": 1333
+    },
+    {
+      "epoch": 0.41153786827086225,
+      "grad_norm": 0.0,
+      "learning_rate": 1.3290994786049532e-05,
+      "loss": 0.6888,
+      "step": 1334
+    },
+    {
+      "epoch": 0.4118463674224896,
+      "grad_norm": 0.0,
+      "learning_rate": 1.3281554419309892e-05,
+      "loss": 0.7991,
+      "step": 1335
+    },
+    {
+      "epoch": 0.41215486657411693,
+      "grad_norm": 0.0,
+      "learning_rate": 1.3272110773954221e-05,
+      "loss": 0.7479,
+      "step": 1336
+    },
+    {
+      "epoch": 0.41246336572574427,
+      "grad_norm": 0.0,
+      "learning_rate": 1.326266385941771e-05,
+      "loss": 0.6592,
+      "step": 1337
+    },
+    {
+      "epoch": 0.4127718648773716,
+      "grad_norm": 0.0,
+      "learning_rate": 1.3253213685138814e-05,
+      "loss": 0.7235,
+      "step": 1338
+    },
+    {
+      "epoch": 0.41308036402899895,
+      "grad_norm": 0.0,
+      "learning_rate": 1.3243760260559245e-05,
+      "loss": 0.8486,
+      "step": 1339
+    },
+    {
+      "epoch": 0.41338886318062623,
+      "grad_norm": 0.0,
+      "learning_rate": 1.323430359512396e-05,
+      "loss": 0.7809,
+      "step": 1340
+    },
+    {
+      "epoch": 0.41369736233225357,
+      "grad_norm": 0.0,
+      "learning_rate": 1.3224843698281165e-05,
+      "loss": 0.7706,
+      "step": 1341
+    },
+    {
+      "epoch": 0.4140058614838809,
+      "grad_norm": 0.0,
+      "learning_rate": 1.3215380579482277e-05,
+      "loss": 0.7855,
+      "step": 1342
+    },
+    {
+      "epoch": 0.41431436063550825,
+      "grad_norm": 0.0,
+      "learning_rate": 1.3205914248181948e-05,
+      "loss": 0.8558,
+      "step": 1343
+    },
+    {
+      "epoch": 0.4146228597871356,
+      "grad_norm": 0.0,
+      "learning_rate": 1.3196444713838028e-05,
+      "loss": 0.7852,
+      "step": 1344
+    },
+    {
+      "epoch": 0.41493135893876293,
+      "grad_norm": 0.0,
+      "learning_rate": 1.3186971985911575e-05,
+      "loss": 0.7275,
+      "step": 1345
+    },
+    {
+      "epoch": 0.41523985809039027,
+      "grad_norm": 0.0,
+      "learning_rate": 1.317749607386683e-05,
+      "loss": 0.8248,
+      "step": 1346
+    },
+    {
+      "epoch": 0.4155483572420176,
+      "grad_norm": 0.0,
+      "learning_rate": 1.316801698717123e-05,
+      "loss": 0.8147,
+      "step": 1347
+    },
+    {
+      "epoch": 0.4158568563936449,
+      "grad_norm": 0.0,
+      "learning_rate": 1.3158534735295365e-05,
+      "loss": 0.8669,
+      "step": 1348
+    },
+    {
+      "epoch": 0.41616535554527223,
+      "grad_norm": 0.0,
+      "learning_rate": 1.3149049327712996e-05,
+      "loss": 0.7874,
+      "step": 1349
+    },
+    {
+      "epoch": 0.41647385469689957,
+      "grad_norm": 0.0,
+      "learning_rate": 1.3139560773901045e-05,
+      "loss": 0.7368,
+      "step": 1350
+    },
+    {
+      "epoch": 0.4167823538485269,
+      "grad_norm": 0.0,
+      "learning_rate": 1.3130069083339563e-05,
+      "loss": 0.7996,
+      "step": 1351
+    },
+    {
+      "epoch": 0.41709085300015425,
+      "grad_norm": 0.0,
+      "learning_rate": 1.3120574265511744e-05,
+      "loss": 0.7961,
+      "step": 1352
+    },
+    {
+      "epoch": 0.4173993521517816,
+      "grad_norm": 0.0,
+      "learning_rate": 1.3111076329903898e-05,
+      "loss": 0.7844,
+      "step": 1353
+    },
+    {
+      "epoch": 0.41770785130340893,
+      "grad_norm": 0.0,
+      "learning_rate": 1.3101575286005467e-05,
+      "loss": 0.7793,
+      "step": 1354
+    },
+    {
+      "epoch": 0.41801635045503627,
+      "grad_norm": 0.0,
+      "learning_rate": 1.309207114330898e-05,
+      "loss": 0.7862,
+      "step": 1355
+    },
+    {
+      "epoch": 0.4183248496066636,
+      "grad_norm": 0.0,
+      "learning_rate": 1.3082563911310069e-05,
+      "loss": 0.8153,
+      "step": 1356
+    },
+    {
+      "epoch": 0.4186333487582909,
+      "grad_norm": 0.0,
+      "learning_rate": 1.3073053599507455e-05,
+      "loss": 0.752,
+      "step": 1357
+    },
+    {
+      "epoch": 0.41894184790991823,
+      "grad_norm": 0.0,
+      "learning_rate": 1.3063540217402934e-05,
+      "loss": 0.8408,
+      "step": 1358
+    },
+    {
+      "epoch": 0.41925034706154557,
+      "grad_norm": 0.0,
+      "learning_rate": 1.305402377450137e-05,
+      "loss": 0.8508,
+      "step": 1359
+    },
+    {
+      "epoch": 0.4195588462131729,
+      "grad_norm": 0.0,
+      "learning_rate": 1.3044504280310683e-05,
+      "loss": 0.6705,
+      "step": 1360
+    },
+    {
+      "epoch": 0.41986734536480025,
+      "grad_norm": 0.0,
+      "learning_rate": 1.3034981744341843e-05,
+      "loss": 0.8153,
+      "step": 1361
+    },
+    {
+      "epoch": 0.4201758445164276,
+      "grad_norm": 0.0,
+      "learning_rate": 1.3025456176108864e-05,
+      "loss": 0.8187,
+      "step": 1362
+    },
+    {
+      "epoch": 0.4204843436680549,
+      "grad_norm": 0.0,
+      "learning_rate": 1.301592758512878e-05,
+      "loss": 0.8127,
+      "step": 1363
+    },
+    {
+      "epoch": 0.42079284281968227,
+      "grad_norm": 0.0,
+      "learning_rate": 1.3006395980921649e-05,
+      "loss": 0.6688,
+      "step": 1364
+    },
+    {
+      "epoch": 0.4211013419713096,
+      "grad_norm": 0.0,
+      "learning_rate": 1.2996861373010543e-05,
+      "loss": 0.8386,
+      "step": 1365
+    },
+    {
+      "epoch": 0.4214098411229369,
+      "grad_norm": 0.0,
+      "learning_rate": 1.2987323770921531e-05,
+      "loss": 0.8533,
+      "step": 1366
+    },
+    {
+      "epoch": 0.42171834027456423,
+      "grad_norm": 0.0,
+      "learning_rate": 1.2977783184183679e-05,
+      "loss": 0.8437,
+      "step": 1367
+    },
+    {
+      "epoch": 0.42202683942619157,
+      "grad_norm": 0.0,
+      "learning_rate": 1.2968239622329027e-05,
+      "loss": 0.8493,
+      "step": 1368
+    },
+    {
+      "epoch": 0.4223353385778189,
+      "grad_norm": 0.0,
+      "learning_rate": 1.2958693094892589e-05,
+      "loss": 0.7388,
+      "step": 1369
+    },
+    {
+      "epoch": 0.42264383772944625,
+      "grad_norm": 0.0,
+      "learning_rate": 1.294914361141235e-05,
+      "loss": 0.7706,
+      "step": 1370
+    },
+    {
+      "epoch": 0.4229523368810736,
+      "grad_norm": 0.0,
+      "learning_rate": 1.2939591181429235e-05,
+      "loss": 0.7611,
+      "step": 1371
+    },
+    {
+      "epoch": 0.4232608360327009,
+      "grad_norm": 0.0,
+      "learning_rate": 1.2930035814487128e-05,
+      "loss": 0.784,
+      "step": 1372
+    },
+    {
+      "epoch": 0.42356933518432827,
+      "grad_norm": 0.0,
+      "learning_rate": 1.292047752013284e-05,
+      "loss": 0.7931,
+      "step": 1373
+    },
+    {
+      "epoch": 0.42387783433595555,
+      "grad_norm": 0.0,
+      "learning_rate": 1.29109163079161e-05,
+      "loss": 0.7797,
+      "step": 1374
+    },
+    {
+      "epoch": 0.4241863334875829,
+      "grad_norm": 0.0,
+      "learning_rate": 1.2901352187389562e-05,
+      "loss": 0.7803,
+      "step": 1375
+    },
+    {
+      "epoch": 0.42449483263921023,
+      "grad_norm": 0.0,
+      "learning_rate": 1.2891785168108784e-05,
+      "loss": 0.7076,
+      "step": 1376
+    },
+    {
+      "epoch": 0.42480333179083757,
+      "grad_norm": 0.0,
+      "learning_rate": 1.2882215259632217e-05,
+      "loss": 0.7885,
+      "step": 1377
+    },
+    {
+      "epoch": 0.4251118309424649,
+      "grad_norm": 0.0,
+      "learning_rate": 1.28726424715212e-05,
+      "loss": 0.7596,
+      "step": 1378
+    },
+    {
+      "epoch": 0.42542033009409225,
+      "grad_norm": 0.0,
+      "learning_rate": 1.2863066813339953e-05,
+      "loss": 0.773,
+      "step": 1379
+    },
+    {
+      "epoch": 0.4257288292457196,
+      "grad_norm": 0.0,
+      "learning_rate": 1.2853488294655556e-05,
+      "loss": 0.8012,
+      "step": 1380
+    },
+    {
+      "epoch": 0.4260373283973469,
+      "grad_norm": 0.0,
+      "learning_rate": 1.2843906925037955e-05,
+      "loss": 0.766,
+      "step": 1381
+    },
+    {
+      "epoch": 0.42634582754897427,
+      "grad_norm": 0.0,
+      "learning_rate": 1.2834322714059932e-05,
+      "loss": 0.7433,
+      "step": 1382
+    },
+    {
+      "epoch": 0.42665432670060155,
+      "grad_norm": 0.0,
+      "learning_rate": 1.2824735671297121e-05,
+      "loss": 0.7938,
+      "step": 1383
+    },
+    {
+      "epoch": 0.4269628258522289,
+      "grad_norm": 0.0,
+      "learning_rate": 1.2815145806327986e-05,
+      "loss": 0.7867,
+      "step": 1384
+    },
+    {
+      "epoch": 0.42727132500385623,
+      "grad_norm": 0.0,
+      "learning_rate": 1.2805553128733797e-05,
+      "loss": 0.697,
+      "step": 1385
+    },
+    {
+      "epoch": 0.42757982415548357,
+      "grad_norm": 0.0,
+      "learning_rate": 1.2795957648098645e-05,
+      "loss": 0.7269,
+      "step": 1386
+    },
+    {
+      "epoch": 0.4278883233071109,
+      "grad_norm": 0.0,
+      "learning_rate": 1.2786359374009418e-05,
+      "loss": 0.7957,
+      "step": 1387
+    },
+    {
+      "epoch": 0.42819682245873825,
+      "grad_norm": 0.0,
+      "learning_rate": 1.2776758316055797e-05,
+      "loss": 0.7505,
+      "step": 1388
+    },
+    {
+      "epoch": 0.4285053216103656,
+      "grad_norm": 0.0,
+      "learning_rate": 1.276715448383024e-05,
+      "loss": 0.7421,
+      "step": 1389
+    },
+    {
+      "epoch": 0.4288138207619929,
+      "grad_norm": 0.0,
+      "learning_rate": 1.2757547886927982e-05,
+      "loss": 0.766,
+      "step": 1390
+    },
+    {
+      "epoch": 0.42912231991362026,
+      "grad_norm": 0.0,
+      "learning_rate": 1.2747938534947015e-05,
+      "loss": 0.7551,
+      "step": 1391
+    },
+    {
+      "epoch": 0.42943081906524755,
+      "grad_norm": 0.0,
+      "learning_rate": 1.2738326437488093e-05,
+      "loss": 0.7754,
+      "step": 1392
+    },
+    {
+      "epoch": 0.4297393182168749,
+      "grad_norm": 0.0,
+      "learning_rate": 1.2728711604154702e-05,
+      "loss": 0.7794,
+      "step": 1393
+    },
+    {
+      "epoch": 0.4300478173685022,
+      "grad_norm": 0.0,
+      "learning_rate": 1.2719094044553066e-05,
+      "loss": 0.7675,
+      "step": 1394
+    },
+    {
+      "epoch": 0.43035631652012957,
+      "grad_norm": 0.0,
+      "learning_rate": 1.270947376829213e-05,
+      "loss": 0.7542,
+      "step": 1395
+    },
+    {
+      "epoch": 0.4306648156717569,
+      "grad_norm": 0.0,
+      "learning_rate": 1.269985078498356e-05,
+      "loss": 0.7695,
+      "step": 1396
+    },
+    {
+      "epoch": 0.43097331482338425,
+      "grad_norm": 0.0,
+      "learning_rate": 1.2690225104241722e-05,
+      "loss": 0.7702,
+      "step": 1397
+    },
+    {
+      "epoch": 0.4312818139750116,
+      "grad_norm": 0.0,
+      "learning_rate": 1.2680596735683679e-05,
+      "loss": 0.7786,
+      "step": 1398
+    },
+    {
+      "epoch": 0.4315903131266389,
+      "grad_norm": 0.0,
+      "learning_rate": 1.2670965688929176e-05,
+      "loss": 0.7759,
+      "step": 1399
+    },
+    {
+      "epoch": 0.4318988122782662,
+      "grad_norm": 0.0,
+      "learning_rate": 1.2661331973600635e-05,
+      "loss": 0.7005,
+      "step": 1400
+    },
+    {
+      "epoch": 0.43220731142989355,
+      "grad_norm": 0.0,
+      "learning_rate": 1.2651695599323147e-05,
+      "loss": 0.7682,
+      "step": 1401
+    },
+    {
+      "epoch": 0.4325158105815209,
+      "grad_norm": 0.0,
+      "learning_rate": 1.2642056575724457e-05,
+      "loss": 0.7238,
+      "step": 1402
+    },
+    {
+      "epoch": 0.4328243097331482,
+      "grad_norm": 0.0,
+      "learning_rate": 1.2632414912434961e-05,
+      "loss": 0.7869,
+      "step": 1403
+    },
+    {
+      "epoch": 0.43313280888477557,
+      "grad_norm": 0.0,
+      "learning_rate": 1.262277061908768e-05,
+      "loss": 0.852,
+      "step": 1404
+    },
+    {
+      "epoch": 0.4334413080364029,
+      "grad_norm": 0.0,
+      "learning_rate": 1.2613123705318281e-05,
+      "loss": 0.7744,
+      "step": 1405
+    },
+    {
+      "epoch": 0.43374980718803025,
+      "grad_norm": 0.0,
+      "learning_rate": 1.2603474180765035e-05,
+      "loss": 0.7881,
+      "step": 1406
+    },
+    {
+      "epoch": 0.4340583063396576,
+      "grad_norm": 0.0,
+      "learning_rate": 1.2593822055068825e-05,
+      "loss": 0.7329,
+      "step": 1407
+    },
+    {
+      "epoch": 0.4343668054912849,
+      "grad_norm": 0.0,
+      "learning_rate": 1.2584167337873137e-05,
+      "loss": 0.8911,
+      "step": 1408
+    },
+    {
+      "epoch": 0.4346753046429122,
+      "grad_norm": 0.0,
+      "learning_rate": 1.2574510038824042e-05,
+      "loss": 0.742,
+      "step": 1409
+    },
+    {
+      "epoch": 0.43498380379453955,
+      "grad_norm": 0.0,
+      "learning_rate": 1.256485016757019e-05,
+      "loss": 0.791,
+      "step": 1410
+    },
+    {
+      "epoch": 0.4352923029461669,
+      "grad_norm": 0.0,
+      "learning_rate": 1.2555187733762808e-05,
+      "loss": 0.7505,
+      "step": 1411
+    },
+    {
+      "epoch": 0.4356008020977942,
+      "grad_norm": 0.0,
+      "learning_rate": 1.2545522747055669e-05,
+      "loss": 0.7848,
+      "step": 1412
+    },
+    {
+      "epoch": 0.43590930124942157,
+      "grad_norm": 0.0,
+      "learning_rate": 1.2535855217105114e-05,
+      "loss": 0.6734,
+      "step": 1413
+    },
+    {
+      "epoch": 0.4362178004010489,
+      "grad_norm": 0.0,
+      "learning_rate": 1.252618515357001e-05,
+      "loss": 0.7401,
+      "step": 1414
+    },
+    {
+      "epoch": 0.43652629955267624,
+      "grad_norm": 0.0,
+      "learning_rate": 1.2516512566111766e-05,
+      "loss": 0.7617,
+      "step": 1415
+    },
+    {
+      "epoch": 0.4368347987043036,
+      "grad_norm": 0.0,
+      "learning_rate": 1.2506837464394307e-05,
+      "loss": 0.7688,
+      "step": 1416
+    },
+    {
+      "epoch": 0.4371432978559309,
+      "grad_norm": 0.0,
+      "learning_rate": 1.2497159858084072e-05,
+      "loss": 0.7622,
+      "step": 1417
+    },
+    {
+      "epoch": 0.4374517970075582,
+      "grad_norm": 0.0,
+      "learning_rate": 1.2487479756850001e-05,
+      "loss": 0.6837,
+      "step": 1418
+    },
+    {
+      "epoch": 0.43776029615918555,
+      "grad_norm": 0.0,
+      "learning_rate": 1.2477797170363524e-05,
+      "loss": 0.8013,
+      "step": 1419
+    },
+    {
+      "epoch": 0.4380687953108129,
+      "grad_norm": 0.0,
+      "learning_rate": 1.2468112108298563e-05,
+      "loss": 0.8371,
+      "step": 1420
+    },
+    {
+      "epoch": 0.4383772944624402,
+      "grad_norm": 0.0,
+      "learning_rate": 1.2458424580331505e-05,
+      "loss": 0.6998,
+      "step": 1421
+    },
+    {
+      "epoch": 0.43868579361406757,
+      "grad_norm": 0.0,
+      "learning_rate": 1.2448734596141205e-05,
+      "loss": 0.7428,
+      "step": 1422
+    },
+    {
+      "epoch": 0.4389942927656949,
+      "grad_norm": 0.0,
+      "learning_rate": 1.243904216540897e-05,
+      "loss": 0.749,
+      "step": 1423
+    },
+    {
+      "epoch": 0.43930279191732224,
+      "grad_norm": 0.0,
+      "learning_rate": 1.2429347297818551e-05,
+      "loss": 0.8241,
+      "step": 1424
+    },
+    {
+      "epoch": 0.4396112910689496,
+      "grad_norm": 0.0,
+      "learning_rate": 1.2419650003056134e-05,
+      "loss": 1.087,
+      "step": 1425
+    },
+    {
+      "epoch": 0.43991979022057687,
+      "grad_norm": 0.0,
+      "learning_rate": 1.2409950290810332e-05,
+      "loss": 0.786,
+      "step": 1426
+    },
+    {
+      "epoch": 0.4402282893722042,
+      "grad_norm": 0.0,
+      "learning_rate": 1.240024817077217e-05,
+      "loss": 0.8391,
+      "step": 1427
+    },
+    {
+      "epoch": 0.44053678852383155,
+      "grad_norm": 0.0,
+      "learning_rate": 1.2390543652635088e-05,
+      "loss": 0.8163,
+      "step": 1428
+    },
+    {
+      "epoch": 0.4408452876754589,
+      "grad_norm": 0.0,
+      "learning_rate": 1.2380836746094903e-05,
+      "loss": 0.7677,
+      "step": 1429
+    },
+    {
+      "epoch": 0.4411537868270862,
+      "grad_norm": 0.0,
+      "learning_rate": 1.2371127460849837e-05,
+      "loss": 0.7921,
+      "step": 1430
+    },
+    {
+      "epoch": 0.44146228597871356,
+      "grad_norm": 0.0,
+      "learning_rate": 1.236141580660048e-05,
+      "loss": 0.8214,
+      "step": 1431
+    },
+    {
+      "epoch": 0.4417707851303409,
+      "grad_norm": 0.0,
+      "learning_rate": 1.2351701793049786e-05,
+      "loss": 0.7893,
+      "step": 1432
+    },
+    {
+      "epoch": 0.44207928428196824,
+      "grad_norm": 0.0,
+      "learning_rate": 1.2341985429903075e-05,
+      "loss": 0.7722,
+      "step": 1433
+    },
+    {
+      "epoch": 0.4423877834335956,
+      "grad_norm": 0.0,
+      "learning_rate": 1.2332266726868013e-05,
+      "loss": 0.7685,
+      "step": 1434
+    },
+    {
+      "epoch": 0.44269628258522287,
+      "grad_norm": 0.0,
+      "learning_rate": 1.2322545693654595e-05,
+      "loss": 0.8296,
+      "step": 1435
+    },
+    {
+      "epoch": 0.4430047817368502,
+      "grad_norm": 0.0,
+      "learning_rate": 1.2312822339975147e-05,
+      "loss": 0.7838,
+      "step": 1436
+    },
+    {
+      "epoch": 0.44331328088847755,
+      "grad_norm": 0.0,
+      "learning_rate": 1.2303096675544325e-05,
+      "loss": 0.8509,
+      "step": 1437
+    },
+    {
+      "epoch": 0.4436217800401049,
+      "grad_norm": 0.0,
+      "learning_rate": 1.229336871007908e-05,
+      "loss": 0.7628,
+      "step": 1438
+    },
+    {
+      "epoch": 0.4439302791917322,
+      "grad_norm": 0.0,
+      "learning_rate": 1.2283638453298665e-05,
+      "loss": 0.7456,
+      "step": 1439
+    },
+    {
+      "epoch": 0.44423877834335956,
+      "grad_norm": 0.0,
+      "learning_rate": 1.2273905914924627e-05,
+      "loss": 0.8279,
+      "step": 1440
+    },
+    {
+      "epoch": 0.4445472774949869,
+      "grad_norm": 0.0,
+      "learning_rate": 1.2264171104680791e-05,
+      "loss": 0.742,
+      "step": 1441
+    },
+    {
+      "epoch": 0.44485577664661424,
+      "grad_norm": 0.0,
+      "learning_rate": 1.2254434032293245e-05,
+      "loss": 0.7001,
+      "step": 1442
+    },
+    {
+      "epoch": 0.4451642757982416,
+      "grad_norm": 0.0,
+      "learning_rate": 1.2244694707490348e-05,
+      "loss": 0.7747,
+      "step": 1443
+    },
+    {
+      "epoch": 0.44547277494986887,
+      "grad_norm": 0.0,
+      "learning_rate": 1.2234953140002698e-05,
+      "loss": 0.8176,
+      "step": 1444
+    },
+    {
+      "epoch": 0.4457812741014962,
+      "grad_norm": 0.0,
+      "learning_rate": 1.2225209339563144e-05,
+      "loss": 0.8175,
+      "step": 1445
+    },
+    {
+      "epoch": 0.44608977325312354,
+      "grad_norm": 0.0,
+      "learning_rate": 1.2215463315906764e-05,
+      "loss": 0.8301,
+      "step": 1446
+    },
+    {
+      "epoch": 0.4463982724047509,
+      "grad_norm": 0.0,
+      "learning_rate": 1.220571507877085e-05,
+      "loss": 0.7825,
+      "step": 1447
+    },
+    {
+      "epoch": 0.4467067715563782,
+      "grad_norm": 0.0,
+      "learning_rate": 1.2195964637894913e-05,
+      "loss": 0.8353,
+      "step": 1448
+    },
+    {
+      "epoch": 0.44701527070800556,
+      "grad_norm": 0.0,
+      "learning_rate": 1.2186212003020658e-05,
+      "loss": 0.8001,
+      "step": 1449
+    },
+    {
+      "epoch": 0.4473237698596329,
+      "grad_norm": 0.0,
+      "learning_rate": 1.2176457183891996e-05,
+      "loss": 0.7869,
+      "step": 1450
+    },
+    {
+      "epoch": 0.44763226901126024,
+      "grad_norm": 0.0,
+      "learning_rate": 1.2166700190255006e-05,
+      "loss": 0.768,
+      "step": 1451
+    },
+    {
+      "epoch": 0.4479407681628875,
+      "grad_norm": 0.0,
+      "learning_rate": 1.2156941031857943e-05,
+      "loss": 0.774,
+      "step": 1452
+    },
+    {
+      "epoch": 0.44824926731451487,
+      "grad_norm": 0.0,
+      "learning_rate": 1.214717971845123e-05,
+      "loss": 0.7134,
+      "step": 1453
+    },
+    {
+      "epoch": 0.4485577664661422,
+      "grad_norm": 0.0,
+      "learning_rate": 1.2137416259787441e-05,
+      "loss": 0.8372,
+      "step": 1454
+    },
+    {
+      "epoch": 0.44886626561776954,
+      "grad_norm": 0.0,
+      "learning_rate": 1.212765066562129e-05,
+      "loss": 0.7524,
+      "step": 1455
+    },
+    {
+      "epoch": 0.4491747647693969,
+      "grad_norm": 0.0,
+      "learning_rate": 1.2117882945709626e-05,
+      "loss": 0.6967,
+      "step": 1456
+    },
+    {
+      "epoch": 0.4494832639210242,
+      "grad_norm": 0.0,
+      "learning_rate": 1.2108113109811426e-05,
+      "loss": 0.7692,
+      "step": 1457
+    },
+    {
+      "epoch": 0.44979176307265156,
+      "grad_norm": 0.0,
+      "learning_rate": 1.2098341167687773e-05,
+      "loss": 1.0787,
+      "step": 1458
+    },
+    {
+      "epoch": 0.4501002622242789,
+      "grad_norm": 0.0,
+      "learning_rate": 1.2088567129101861e-05,
+      "loss": 0.7832,
+      "step": 1459
+    },
+    {
+      "epoch": 0.45040876137590624,
+      "grad_norm": 0.0,
+      "learning_rate": 1.2078791003818981e-05,
+      "loss": 0.7352,
+      "step": 1460
+    },
+    {
+      "epoch": 0.4507172605275335,
+      "grad_norm": 0.0,
+      "learning_rate": 1.20690128016065e-05,
+      "loss": 0.7652,
+      "step": 1461
+    },
+    {
+      "epoch": 0.45102575967916086,
+      "grad_norm": 0.0,
+      "learning_rate": 1.2059232532233864e-05,
+      "loss": 0.7968,
+      "step": 1462
+    },
+    {
+      "epoch": 0.4513342588307882,
+      "grad_norm": 0.0,
+      "learning_rate": 1.2049450205472585e-05,
+      "loss": 0.7473,
+      "step": 1463
+    },
+    {
+      "epoch": 0.45164275798241554,
+      "grad_norm": 0.0,
+      "learning_rate": 1.203966583109623e-05,
+      "loss": 0.7823,
+      "step": 1464
+    },
+    {
+      "epoch": 0.4519512571340429,
+      "grad_norm": 0.0,
+      "learning_rate": 1.2029879418880419e-05,
+      "loss": 0.7657,
+      "step": 1465
+    },
+    {
+      "epoch": 0.4522597562856702,
+      "grad_norm": 0.0,
+      "learning_rate": 1.2020090978602795e-05,
+      "loss": 0.7575,
+      "step": 1466
+    },
+    {
+      "epoch": 0.45256825543729756,
+      "grad_norm": 0.0,
+      "learning_rate": 1.2010300520043035e-05,
+      "loss": 0.8245,
+      "step": 1467
+    },
+    {
+      "epoch": 0.4528767545889249,
+      "grad_norm": 0.0,
+      "learning_rate": 1.2000508052982827e-05,
+      "loss": 0.7677,
+      "step": 1468
+    },
+    {
+      "epoch": 0.45318525374055224,
+      "grad_norm": 0.0,
+      "learning_rate": 1.1990713587205876e-05,
+      "loss": 0.8229,
+      "step": 1469
+    },
+    {
+      "epoch": 0.4534937528921795,
+      "grad_norm": 0.0,
+      "learning_rate": 1.1980917132497871e-05,
+      "loss": 0.7153,
+      "step": 1470
+    },
+    {
+      "epoch": 0.45380225204380686,
+      "grad_norm": 0.0,
+      "learning_rate": 1.1971118698646502e-05,
+      "loss": 0.7822,
+      "step": 1471
+    },
+    {
+      "epoch": 0.4541107511954342,
+      "grad_norm": 0.0,
+      "learning_rate": 1.1961318295441424e-05,
+      "loss": 0.7923,
+      "step": 1472
+    },
+    {
+      "epoch": 0.45441925034706154,
+      "grad_norm": 0.0,
+      "learning_rate": 1.1951515932674267e-05,
+      "loss": 0.8233,
+      "step": 1473
+    },
+    {
+      "epoch": 0.4547277494986889,
+      "grad_norm": 0.0,
+      "learning_rate": 1.1941711620138614e-05,
+      "loss": 0.7519,
+      "step": 1474
+    },
+    {
+      "epoch": 0.4550362486503162,
+      "grad_norm": 0.0,
+      "learning_rate": 1.1931905367630004e-05,
+      "loss": 0.6626,
+      "step": 1475
+    },
+    {
+      "epoch": 0.45534474780194356,
+      "grad_norm": 0.0,
+      "learning_rate": 1.1922097184945897e-05,
+      "loss": 0.8072,
+      "step": 1476
+    },
+    {
+      "epoch": 0.4556532469535709,
+      "grad_norm": 0.0,
+      "learning_rate": 1.1912287081885708e-05,
+      "loss": 0.6687,
+      "step": 1477
+    },
+    {
+      "epoch": 0.4559617461051982,
+      "grad_norm": 0.0,
+      "learning_rate": 1.1902475068250747e-05,
+      "loss": 0.8199,
+      "step": 1478
+    },
+    {
+      "epoch": 0.4562702452568255,
+      "grad_norm": 0.0,
+      "learning_rate": 1.1892661153844243e-05,
+      "loss": 0.7257,
+      "step": 1479
+    },
+    {
+      "epoch": 0.45657874440845286,
+      "grad_norm": 0.0,
+      "learning_rate": 1.1882845348471324e-05,
+      "loss": 0.6936,
+      "step": 1480
+    },
+    {
+      "epoch": 0.4568872435600802,
+      "grad_norm": 0.0,
+      "learning_rate": 1.1873027661939003e-05,
+      "loss": 0.6945,
+      "step": 1481
+    },
+    {
+      "epoch": 0.45719574271170754,
+      "grad_norm": 0.0,
+      "learning_rate": 1.1863208104056179e-05,
+      "loss": 0.7709,
+      "step": 1482
+    },
+    {
+      "epoch": 0.4575042418633349,
+      "grad_norm": 0.0,
+      "learning_rate": 1.1853386684633616e-05,
+      "loss": 0.748,
+      "step": 1483
+    },
+    {
+      "epoch": 0.4578127410149622,
+      "grad_norm": 0.0,
+      "learning_rate": 1.1843563413483941e-05,
+      "loss": 0.7041,
+      "step": 1484
+    },
+    {
+      "epoch": 0.45812124016658956,
+      "grad_norm": 0.0,
+      "learning_rate": 1.1833738300421625e-05,
+      "loss": 1.0664,
+      "step": 1485
+    },
+    {
+      "epoch": 0.4584297393182169,
+      "grad_norm": 0.0,
+      "learning_rate": 1.1823911355262986e-05,
+      "loss": 0.8065,
+      "step": 1486
+    },
+    {
+      "epoch": 0.4587382384698442,
+      "grad_norm": 0.0,
+      "learning_rate": 1.181408258782617e-05,
+      "loss": 0.8089,
+      "step": 1487
+    },
+    {
+      "epoch": 0.4590467376214715,
+      "grad_norm": 0.0,
+      "learning_rate": 1.1804252007931142e-05,
+      "loss": 0.7305,
+      "step": 1488
+    },
+    {
+      "epoch": 0.45935523677309886,
+      "grad_norm": 0.0,
+      "learning_rate": 1.1794419625399682e-05,
+      "loss": 0.6922,
+      "step": 1489
+    },
+    {
+      "epoch": 0.4596637359247262,
+      "grad_norm": 0.0,
+      "learning_rate": 1.1784585450055367e-05,
+      "loss": 0.6915,
+      "step": 1490
+    },
+    {
+      "epoch": 0.45997223507635354,
+      "grad_norm": 0.0,
+      "learning_rate": 1.1774749491723565e-05,
+      "loss": 0.7147,
+      "step": 1491
+    },
+    {
+      "epoch": 0.4602807342279809,
+      "grad_norm": 0.0,
+      "learning_rate": 1.176491176023143e-05,
+      "loss": 0.7419,
+      "step": 1492
+    },
+    {
+      "epoch": 0.4605892333796082,
+      "grad_norm": 0.0,
+      "learning_rate": 1.1755072265407881e-05,
+      "loss": 0.8442,
+      "step": 1493
+    },
+    {
+      "epoch": 0.46089773253123556,
+      "grad_norm": 0.0,
+      "learning_rate": 1.1745231017083603e-05,
+      "loss": 0.748,
+      "step": 1494
+    },
+    {
+      "epoch": 0.4612062316828629,
+      "grad_norm": 0.0,
+      "learning_rate": 1.1735388025091032e-05,
+      "loss": 0.8019,
+      "step": 1495
+    },
+    {
+      "epoch": 0.4615147308344902,
+      "grad_norm": 0.0,
+      "learning_rate": 1.1725543299264352e-05,
+      "loss": 0.7985,
+      "step": 1496
+    },
+    {
+      "epoch": 0.4618232299861175,
+      "grad_norm": 0.0,
+      "learning_rate": 1.1715696849439465e-05,
+      "loss": 0.8069,
+      "step": 1497
+    },
+    {
+      "epoch": 0.46213172913774486,
+      "grad_norm": 0.0,
+      "learning_rate": 1.1705848685454011e-05,
+      "loss": 0.8735,
+      "step": 1498
+    },
+    {
+      "epoch": 0.4624402282893722,
+      "grad_norm": 0.0,
+      "learning_rate": 1.1695998817147334e-05,
+      "loss": 0.8315,
+      "step": 1499
+    },
+    {
+      "epoch": 0.46274872744099954,
+      "grad_norm": 0.0,
+      "learning_rate": 1.168614725436048e-05,
+      "loss": 0.7874,
+      "step": 1500
+    },
+    {
+      "epoch": 0.4630572265926269,
+      "grad_norm": 0.0,
+      "learning_rate": 1.167629400693619e-05,
+      "loss": 0.7339,
+      "step": 1501
+    },
+    {
+      "epoch": 0.4633657257442542,
+      "grad_norm": 0.0,
+      "learning_rate": 1.1666439084718893e-05,
+      "loss": 0.8196,
+      "step": 1502
+    },
+    {
+      "epoch": 0.46367422489588156,
+      "grad_norm": 0.0,
+      "learning_rate": 1.1656582497554682e-05,
+      "loss": 0.8023,
+      "step": 1503
+    },
+    {
+      "epoch": 0.46398272404750884,
+      "grad_norm": 0.0,
+      "learning_rate": 1.1646724255291321e-05,
+      "loss": 0.7816,
+      "step": 1504
+    },
+    {
+      "epoch": 0.4642912231991362,
+      "grad_norm": 0.0,
+      "learning_rate": 1.1636864367778226e-05,
+      "loss": 0.7904,
+      "step": 1505
+    },
+    {
+      "epoch": 0.4645997223507635,
+      "grad_norm": 0.0,
+      "learning_rate": 1.1627002844866455e-05,
+      "loss": 0.8138,
+      "step": 1506
+    },
+    {
+      "epoch": 0.46490822150239086,
+      "grad_norm": 0.0,
+      "learning_rate": 1.1617139696408696e-05,
+      "loss": 0.7566,
+      "step": 1507
+    },
+    {
+      "epoch": 0.4652167206540182,
+      "grad_norm": 0.0,
+      "learning_rate": 1.1607274932259273e-05,
+      "loss": 0.8639,
+      "step": 1508
+    },
+    {
+      "epoch": 0.46552521980564554,
+      "grad_norm": 0.0,
+      "learning_rate": 1.1597408562274113e-05,
+      "loss": 0.7517,
+      "step": 1509
+    },
+    {
+      "epoch": 0.4658337189572729,
+      "grad_norm": 0.0,
+      "learning_rate": 1.1587540596310751e-05,
+      "loss": 0.8349,
+      "step": 1510
+    },
+    {
+      "epoch": 0.4661422181089002,
+      "grad_norm": 0.0,
+      "learning_rate": 1.157767104422832e-05,
+      "loss": 0.7613,
+      "step": 1511
+    },
+    {
+      "epoch": 0.46645071726052756,
+      "grad_norm": 0.0,
+      "learning_rate": 1.1567799915887532e-05,
+      "loss": 0.7602,
+      "step": 1512
+    },
+    {
+      "epoch": 0.46675921641215484,
+      "grad_norm": 0.0,
+      "learning_rate": 1.1557927221150674e-05,
+      "loss": 0.7969,
+      "step": 1513
+    },
+    {
+      "epoch": 0.4670677155637822,
+      "grad_norm": 0.0,
+      "learning_rate": 1.1548052969881604e-05,
+      "loss": 0.6685,
+      "step": 1514
+    },
+    {
+      "epoch": 0.4673762147154095,
+      "grad_norm": 0.0,
+      "learning_rate": 1.1538177171945733e-05,
+      "loss": 0.7429,
+      "step": 1515
+    },
+    {
+      "epoch": 0.46768471386703686,
+      "grad_norm": 0.0,
+      "learning_rate": 1.1528299837210015e-05,
+      "loss": 0.7313,
+      "step": 1516
+    },
+    {
+      "epoch": 0.4679932130186642,
+      "grad_norm": 0.0,
+      "learning_rate": 1.1518420975542937e-05,
+      "loss": 0.7563,
+      "step": 1517
+    },
+    {
+      "epoch": 0.46830171217029154,
+      "grad_norm": 0.0,
+      "learning_rate": 1.150854059681452e-05,
+      "loss": 0.7186,
+      "step": 1518
+    },
+    {
+      "epoch": 0.4686102113219189,
+      "grad_norm": 0.0,
+      "learning_rate": 1.149865871089629e-05,
+      "loss": 0.8129,
+      "step": 1519
+    },
+    {
+      "epoch": 0.4689187104735462,
+      "grad_norm": 0.0,
+      "learning_rate": 1.1488775327661288e-05,
+      "loss": 0.7723,
+      "step": 1520
+    },
+    {
+      "epoch": 0.46922720962517356,
+      "grad_norm": 0.0,
+      "learning_rate": 1.1478890456984044e-05,
+      "loss": 0.7515,
+      "step": 1521
+    },
+    {
+      "epoch": 0.46953570877680084,
+      "grad_norm": 0.0,
+      "learning_rate": 1.1469004108740584e-05,
+      "loss": 0.7922,
+      "step": 1522
+    },
+    {
+      "epoch": 0.4698442079284282,
+      "grad_norm": 0.0,
+      "learning_rate": 1.14591162928084e-05,
+      "loss": 0.7006,
+      "step": 1523
+    },
+    {
+      "epoch": 0.4701527070800555,
+      "grad_norm": 0.0,
+      "learning_rate": 1.1449227019066452e-05,
+      "loss": 0.7337,
+      "step": 1524
+    },
+    {
+      "epoch": 0.47046120623168286,
+      "grad_norm": 0.0,
+      "learning_rate": 1.1439336297395161e-05,
+      "loss": 0.756,
+      "step": 1525
+    },
+    {
+      "epoch": 0.4707697053833102,
+      "grad_norm": 0.0,
+      "learning_rate": 1.1429444137676391e-05,
+      "loss": 0.7558,
+      "step": 1526
+    },
+    {
+      "epoch": 0.47107820453493754,
+      "grad_norm": 0.0,
+      "learning_rate": 1.1419550549793443e-05,
+      "loss": 0.7017,
+      "step": 1527
+    },
+    {
+      "epoch": 0.4713867036865649,
+      "grad_norm": 0.0,
+      "learning_rate": 1.1409655543631054e-05,
+      "loss": 0.6988,
+      "step": 1528
+    },
+    {
+      "epoch": 0.4716952028381922,
+      "grad_norm": 0.0,
+      "learning_rate": 1.1399759129075358e-05,
+      "loss": 0.8331,
+      "step": 1529
+    },
+    {
+      "epoch": 0.4720037019898195,
+      "grad_norm": 0.0,
+      "learning_rate": 1.1389861316013914e-05,
+      "loss": 0.7593,
+      "step": 1530
+    },
+    {
+      "epoch": 0.47231220114144684,
+      "grad_norm": 0.0,
+      "learning_rate": 1.1379962114335676e-05,
+      "loss": 0.8505,
+      "step": 1531
+    },
+    {
+      "epoch": 0.4726207002930742,
+      "grad_norm": 0.0,
+      "learning_rate": 1.1370061533930974e-05,
+      "loss": 0.7795,
+      "step": 1532
+    },
+    {
+      "epoch": 0.4729291994447015,
+      "grad_norm": 0.0,
+      "learning_rate": 1.1360159584691529e-05,
+      "loss": 0.7112,
+      "step": 1533
+    },
+    {
+      "epoch": 0.47323769859632886,
+      "grad_norm": 0.0,
+      "learning_rate": 1.1350256276510421e-05,
+      "loss": 0.7215,
+      "step": 1534
+    },
+    {
+      "epoch": 0.4735461977479562,
+      "grad_norm": 0.0,
+      "learning_rate": 1.1340351619282091e-05,
+      "loss": 0.7342,
+      "step": 1535
+    },
+    {
+      "epoch": 0.47385469689958354,
+      "grad_norm": 0.0,
+      "learning_rate": 1.1330445622902326e-05,
+      "loss": 0.7108,
+      "step": 1536
+    },
+    {
+      "epoch": 0.4741631960512109,
+      "grad_norm": 0.0,
+      "learning_rate": 1.1320538297268258e-05,
+      "loss": 0.8285,
+      "step": 1537
+    },
+    {
+      "epoch": 0.4744716952028382,
+      "grad_norm": 0.0,
+      "learning_rate": 1.1310629652278334e-05,
+      "loss": 0.7805,
+      "step": 1538
+    },
+    {
+      "epoch": 0.4747801943544655,
+      "grad_norm": 0.0,
+      "learning_rate": 1.1300719697832329e-05,
+      "loss": 0.724,
+      "step": 1539
+    },
+    {
+      "epoch": 0.47508869350609284,
+      "grad_norm": 0.0,
+      "learning_rate": 1.1290808443831324e-05,
+      "loss": 0.7394,
+      "step": 1540
+    },
+    {
+      "epoch": 0.4753971926577202,
+      "grad_norm": 0.0,
+      "learning_rate": 1.1280895900177704e-05,
+      "loss": 0.7535,
+      "step": 1541
+    },
+    {
+      "epoch": 0.4757056918093475,
+      "grad_norm": 0.0,
+      "learning_rate": 1.1270982076775126e-05,
+      "loss": 0.7777,
+      "step": 1542
+    },
+    {
+      "epoch": 0.47601419096097486,
+      "grad_norm": 0.0,
+      "learning_rate": 1.1261066983528544e-05,
+      "loss": 0.7488,
+      "step": 1543
+    },
+    {
+      "epoch": 0.4763226901126022,
+      "grad_norm": 0.0,
+      "learning_rate": 1.125115063034417e-05,
+      "loss": 0.7654,
+      "step": 1544
+    },
+    {
+      "epoch": 0.47663118926422954,
+      "grad_norm": 0.0,
+      "learning_rate": 1.124123302712948e-05,
+      "loss": 0.7223,
+      "step": 1545
+    },
+    {
+      "epoch": 0.4769396884158569,
+      "grad_norm": 0.0,
+      "learning_rate": 1.1231314183793192e-05,
+      "loss": 0.7698,
+      "step": 1546
+    },
+    {
+      "epoch": 0.4772481875674842,
+      "grad_norm": 0.0,
+      "learning_rate": 1.1221394110245271e-05,
+      "loss": 0.7299,
+      "step": 1547
+    },
+    {
+      "epoch": 0.4775566867191115,
+      "grad_norm": 0.0,
+      "learning_rate": 1.1211472816396912e-05,
+      "loss": 0.7969,
+      "step": 1548
+    },
+    {
+      "epoch": 0.47786518587073884,
+      "grad_norm": 0.0,
+      "learning_rate": 1.120155031216052e-05,
+      "loss": 0.7025,
+      "step": 1549
+    },
+    {
+      "epoch": 0.4781736850223662,
+      "grad_norm": 0.0,
+      "learning_rate": 1.1191626607449713e-05,
+      "loss": 0.7603,
+      "step": 1550
+    },
+    {
+      "epoch": 0.4784821841739935,
+      "grad_norm": 0.0,
+      "learning_rate": 1.1181701712179311e-05,
+      "loss": 0.7384,
+      "step": 1551
+    },
+    {
+      "epoch": 0.47879068332562086,
+      "grad_norm": 0.0,
+      "learning_rate": 1.1171775636265326e-05,
+      "loss": 0.7992,
+      "step": 1552
+    },
+    {
+      "epoch": 0.4790991824772482,
+      "grad_norm": 0.0,
+      "learning_rate": 1.1161848389624942e-05,
+      "loss": 0.7818,
+      "step": 1553
+    },
+    {
+      "epoch": 0.47940768162887554,
+      "grad_norm": 0.0,
+      "learning_rate": 1.115191998217652e-05,
+      "loss": 0.7237,
+      "step": 1554
+    },
+    {
+      "epoch": 0.4797161807805029,
+      "grad_norm": 0.0,
+      "learning_rate": 1.1141990423839573e-05,
+      "loss": 0.7212,
+      "step": 1555
+    },
+    {
+      "epoch": 0.48002467993213016,
+      "grad_norm": 0.0,
+      "learning_rate": 1.1132059724534772e-05,
+      "loss": 0.7296,
+      "step": 1556
+    },
+    {
+      "epoch": 0.4803331790837575,
+      "grad_norm": 0.0,
+      "learning_rate": 1.112212789418392e-05,
+      "loss": 0.7456,
+      "step": 1557
+    },
+    {
+      "epoch": 0.48064167823538484,
+      "grad_norm": 0.0,
+      "learning_rate": 1.1112194942709956e-05,
+      "loss": 0.7829,
+      "step": 1558
+    },
+    {
+      "epoch": 0.4809501773870122,
+      "grad_norm": 0.0,
+      "learning_rate": 1.1102260880036937e-05,
+      "loss": 0.7913,
+      "step": 1559
+    },
+    {
+      "epoch": 0.4812586765386395,
+      "grad_norm": 0.0,
+      "learning_rate": 1.1092325716090035e-05,
+      "loss": 0.7316,
+      "step": 1560
+    },
+    {
+      "epoch": 0.48156717569026686,
+      "grad_norm": 0.0,
+      "learning_rate": 1.108238946079551e-05,
+      "loss": 0.8268,
+      "step": 1561
+    },
+    {
+      "epoch": 0.4818756748418942,
+      "grad_norm": 0.0,
+      "learning_rate": 1.1072452124080728e-05,
+      "loss": 0.7784,
+      "step": 1562
+    },
+    {
+      "epoch": 0.48218417399352154,
+      "grad_norm": 0.0,
+      "learning_rate": 1.1062513715874119e-05,
+      "loss": 0.8224,
+      "step": 1563
+    },
+    {
+      "epoch": 0.4824926731451489,
+      "grad_norm": 0.0,
+      "learning_rate": 1.1052574246105196e-05,
+      "loss": 0.7781,
+      "step": 1564
+    },
+    {
+      "epoch": 0.48280117229677616,
+      "grad_norm": 0.0,
+      "learning_rate": 1.1042633724704534e-05,
+      "loss": 0.6717,
+      "step": 1565
+    },
+    {
+      "epoch": 0.4831096714484035,
+      "grad_norm": 0.0,
+      "learning_rate": 1.1032692161603746e-05,
+      "loss": 0.8659,
+      "step": 1566
+    },
+    {
+      "epoch": 0.48341817060003084,
+      "grad_norm": 0.0,
+      "learning_rate": 1.1022749566735497e-05,
+      "loss": 0.7775,
+      "step": 1567
+    },
+    {
+      "epoch": 0.4837266697516582,
+      "grad_norm": 0.0,
+      "learning_rate": 1.1012805950033476e-05,
+      "loss": 0.7473,
+      "step": 1568
+    },
+    {
+      "epoch": 0.4840351689032855,
+      "grad_norm": 0.0,
+      "learning_rate": 1.1002861321432401e-05,
+      "loss": 0.7479,
+      "step": 1569
+    },
+    {
+      "epoch": 0.48434366805491286,
+      "grad_norm": 0.0,
+      "learning_rate": 1.0992915690867989e-05,
+      "loss": 0.721,
+      "step": 1570
+    },
+    {
+      "epoch": 0.4846521672065402,
+      "grad_norm": 0.0,
+      "learning_rate": 1.0982969068276971e-05,
+      "loss": 0.7837,
+      "step": 1571
+    },
+    {
+      "epoch": 0.48496066635816754,
+      "grad_norm": 0.0,
+      "learning_rate": 1.097302146359706e-05,
+      "loss": 0.7825,
+      "step": 1572
+    },
+    {
+      "epoch": 0.4852691655097949,
+      "grad_norm": 0.0,
+      "learning_rate": 1.0963072886766956e-05,
+      "loss": 0.7759,
+      "step": 1573
+    },
+    {
+      "epoch": 0.48557766466142216,
+      "grad_norm": 0.0,
+      "learning_rate": 1.0953123347726325e-05,
+      "loss": 0.7175,
+      "step": 1574
+    },
+    {
+      "epoch": 0.4858861638130495,
+      "grad_norm": 0.0,
+      "learning_rate": 1.0943172856415798e-05,
+      "loss": 0.7636,
+      "step": 1575
+    },
+    {
+      "epoch": 0.48619466296467684,
+      "grad_norm": 0.0,
+      "learning_rate": 1.0933221422776953e-05,
+      "loss": 0.7212,
+      "step": 1576
+    },
+    {
+      "epoch": 0.4865031621163042,
+      "grad_norm": 0.0,
+      "learning_rate": 1.092326905675232e-05,
+      "loss": 0.7146,
+      "step": 1577
+    },
+    {
+      "epoch": 0.4868116612679315,
+      "grad_norm": 0.0,
+      "learning_rate": 1.0913315768285347e-05,
+      "loss": 0.7574,
+      "step": 1578
+    },
+    {
+      "epoch": 0.48712016041955886,
+      "grad_norm": 0.0,
+      "learning_rate": 1.0903361567320411e-05,
+      "loss": 0.7246,
+      "step": 1579
+    },
+    {
+      "epoch": 0.4874286595711862,
+      "grad_norm": 0.0,
+      "learning_rate": 1.0893406463802805e-05,
+      "loss": 0.785,
+      "step": 1580
+    },
+    {
+      "epoch": 0.48773715872281354,
+      "grad_norm": 0.0,
+      "learning_rate": 1.0883450467678711e-05,
+      "loss": 0.7738,
+      "step": 1581
+    },
+    {
+      "epoch": 0.4880456578744408,
+      "grad_norm": 0.0,
+      "learning_rate": 1.0873493588895216e-05,
+      "loss": 0.732,
+      "step": 1582
+    },
+    {
+      "epoch": 0.48835415702606816,
+      "grad_norm": 0.0,
+      "learning_rate": 1.0863535837400281e-05,
+      "loss": 0.7488,
+      "step": 1583
+    },
+    {
+      "epoch": 0.4886626561776955,
+      "grad_norm": 0.0,
+      "learning_rate": 1.0853577223142742e-05,
+      "loss": 0.8508,
+      "step": 1584
+    },
+    {
+      "epoch": 0.48897115532932284,
+      "grad_norm": 0.0,
+      "learning_rate": 1.0843617756072294e-05,
+      "loss": 0.8112,
+      "step": 1585
+    },
+    {
+      "epoch": 0.4892796544809502,
+      "grad_norm": 0.0,
+      "learning_rate": 1.0833657446139487e-05,
+      "loss": 0.7818,
+      "step": 1586
+    },
+    {
+      "epoch": 0.4895881536325775,
+      "grad_norm": 0.0,
+      "learning_rate": 1.0823696303295712e-05,
+      "loss": 0.7754,
+      "step": 1587
+    },
+    {
+      "epoch": 0.48989665278420486,
+      "grad_norm": 0.0,
+      "learning_rate": 1.0813734337493194e-05,
+      "loss": 0.7389,
+      "step": 1588
+    },
+    {
+      "epoch": 0.4902051519358322,
+      "grad_norm": 0.0,
+      "learning_rate": 1.0803771558684977e-05,
+      "loss": 0.6946,
+      "step": 1589
+    },
+    {
+      "epoch": 0.49051365108745953,
+      "grad_norm": 0.0,
+      "learning_rate": 1.079380797682492e-05,
+      "loss": 0.7306,
+      "step": 1590
+    },
+    {
+      "epoch": 0.4908221502390868,
+      "grad_norm": 0.0,
+      "learning_rate": 1.0783843601867681e-05,
+      "loss": 0.681,
+      "step": 1591
+    },
+    {
+      "epoch": 0.49113064939071416,
+      "grad_norm": 0.0,
+      "learning_rate": 1.0773878443768716e-05,
+      "loss": 0.7666,
+      "step": 1592
+    },
+    {
+      "epoch": 0.4914391485423415,
+      "grad_norm": 0.0,
+      "learning_rate": 1.0763912512484257e-05,
+      "loss": 0.807,
+      "step": 1593
+    },
+    {
+      "epoch": 0.49174764769396884,
+      "grad_norm": 0.0,
+      "learning_rate": 1.0753945817971311e-05,
+      "loss": 0.8633,
+      "step": 1594
+    },
+    {
+      "epoch": 0.4920561468455962,
+      "grad_norm": 0.0,
+      "learning_rate": 1.0743978370187651e-05,
+      "loss": 0.8063,
+      "step": 1595
+    },
+    {
+      "epoch": 0.4923646459972235,
+      "grad_norm": 0.0,
+      "learning_rate": 1.0734010179091798e-05,
+      "loss": 0.7307,
+      "step": 1596
+    },
+    {
+      "epoch": 0.49267314514885086,
+      "grad_norm": 0.0,
+      "learning_rate": 1.072404125464302e-05,
+      "loss": 0.747,
+      "step": 1597
+    },
+    {
+      "epoch": 0.4929816443004782,
+      "grad_norm": 0.0,
+      "learning_rate": 1.0714071606801314e-05,
+      "loss": 0.7696,
+      "step": 1598
+    },
+    {
+      "epoch": 0.49329014345210553,
+      "grad_norm": 0.0,
+      "learning_rate": 1.0704101245527396e-05,
+      "loss": 0.774,
+      "step": 1599
+    },
+    {
+      "epoch": 0.4935986426037328,
+      "grad_norm": 0.0,
+      "learning_rate": 1.0694130180782705e-05,
+      "loss": 0.7023,
+      "step": 1600
+    },
+    {
+      "epoch": 0.49390714175536016,
+      "grad_norm": 0.0,
+      "learning_rate": 1.0684158422529374e-05,
+      "loss": 0.7351,
+      "step": 1601
+    },
+    {
+      "epoch": 0.4942156409069875,
+      "grad_norm": 0.0,
+      "learning_rate": 1.0674185980730234e-05,
+      "loss": 0.7549,
+      "step": 1602
+    },
+    {
+      "epoch": 0.49452414005861484,
+      "grad_norm": 0.0,
+      "learning_rate": 1.0664212865348798e-05,
+      "loss": 0.7447,
+      "step": 1603
+    },
+    {
+      "epoch": 0.4948326392102422,
+      "grad_norm": 0.0,
+      "learning_rate": 1.065423908634925e-05,
+      "loss": 0.7676,
+      "step": 1604
+    },
+    {
+      "epoch": 0.4951411383618695,
+      "grad_norm": 0.0,
+      "learning_rate": 1.064426465369644e-05,
+      "loss": 0.7771,
+      "step": 1605
+    },
+    {
+      "epoch": 0.49544963751349685,
+      "grad_norm": 0.0,
+      "learning_rate": 1.0634289577355864e-05,
+      "loss": 0.7675,
+      "step": 1606
+    },
+    {
+      "epoch": 0.4957581366651242,
+      "grad_norm": 0.0,
+      "learning_rate": 1.062431386729367e-05,
+      "loss": 0.7554,
+      "step": 1607
+    },
+    {
+      "epoch": 0.4960666358167515,
+      "grad_norm": 0.0,
+      "learning_rate": 1.0614337533476635e-05,
+      "loss": 0.8138,
+      "step": 1608
+    },
+    {
+      "epoch": 0.4963751349683788,
+      "grad_norm": 0.0,
+      "learning_rate": 1.0604360585872165e-05,
+      "loss": 0.7168,
+      "step": 1609
+    },
+    {
+      "epoch": 0.49668363412000616,
+      "grad_norm": 0.0,
+      "learning_rate": 1.0594383034448266e-05,
+      "loss": 0.7353,
+      "step": 1610
+    },
+    {
+      "epoch": 0.4969921332716335,
+      "grad_norm": 0.0,
+      "learning_rate": 1.058440488917356e-05,
+      "loss": 0.7012,
+      "step": 1611
+    },
+    {
+      "epoch": 0.49730063242326084,
+      "grad_norm": 0.0,
+      "learning_rate": 1.0574426160017257e-05,
+      "loss": 0.8099,
+      "step": 1612
+    },
+    {
+      "epoch": 0.4976091315748882,
+      "grad_norm": 0.0,
+      "learning_rate": 1.0564446856949146e-05,
+      "loss": 0.827,
+      "step": 1613
+    },
+    {
+      "epoch": 0.4979176307265155,
+      "grad_norm": 0.0,
+      "learning_rate": 1.0554466989939602e-05,
+      "loss": 0.855,
+      "step": 1614
+    },
+    {
+      "epoch": 0.49822612987814285,
+      "grad_norm": 0.0,
+      "learning_rate": 1.0544486568959551e-05,
+      "loss": 0.7281,
+      "step": 1615
+    },
+    {
+      "epoch": 0.4985346290297702,
+      "grad_norm": 0.0,
+      "learning_rate": 1.0534505603980481e-05,
+      "loss": 0.7121,
+      "step": 1616
+    },
+    {
+      "epoch": 0.4988431281813975,
+      "grad_norm": 0.0,
+      "learning_rate": 1.0524524104974414e-05,
+      "loss": 0.7575,
+      "step": 1617
+    },
+    {
+      "epoch": 0.4991516273330248,
+      "grad_norm": 0.0,
+      "learning_rate": 1.0514542081913916e-05,
+      "loss": 0.8202,
+      "step": 1618
+    },
+    {
+      "epoch": 0.49946012648465216,
+      "grad_norm": 0.0,
+      "learning_rate": 1.0504559544772071e-05,
+      "loss": 0.858,
+      "step": 1619
+    },
+    {
+      "epoch": 0.4997686256362795,
+      "grad_norm": 0.0,
+      "learning_rate": 1.0494576503522475e-05,
+      "loss": 0.7307,
+      "step": 1620
+    },
+    {
+      "epoch": 0.5000771247879068,
+      "grad_norm": 0.0,
+      "learning_rate": 1.0484592968139234e-05,
+      "loss": 0.7756,
+      "step": 1621
+    },
+    {
+      "epoch": 0.5003856239395341,
+      "grad_norm": 0.0,
+      "learning_rate": 1.0474608948596943e-05,
+      "loss": 0.7073,
+      "step": 1622
+    },
+    {
+      "epoch": 0.5006941230911615,
+      "grad_norm": 0.0,
+      "learning_rate": 1.0464624454870678e-05,
+      "loss": 0.7282,
+      "step": 1623
+    },
+    {
+      "epoch": 0.5010026222427888,
+      "grad_norm": 0.0,
+      "learning_rate": 1.0454639496935997e-05,
+      "loss": 0.8389,
+      "step": 1624
+    },
+    {
+      "epoch": 0.5013111213944161,
+      "grad_norm": 0.0,
+      "learning_rate": 1.0444654084768915e-05,
+      "loss": 0.7465,
+      "step": 1625
+    },
+    {
+      "epoch": 0.5016196205460435,
+      "grad_norm": 0.0,
+      "learning_rate": 1.0434668228345901e-05,
+      "loss": 0.8251,
+      "step": 1626
+    },
+    {
+      "epoch": 0.5019281196976708,
+      "grad_norm": 0.0,
+      "learning_rate": 1.0424681937643872e-05,
+      "loss": 0.8191,
+      "step": 1627
+    },
+    {
+      "epoch": 0.5022366188492982,
+      "grad_norm": 0.0,
+      "learning_rate": 1.041469522264018e-05,
+      "loss": 0.8236,
+      "step": 1628
+    },
+    {
+      "epoch": 0.5025451180009255,
+      "grad_norm": 0.0,
+      "learning_rate": 1.0404708093312593e-05,
+      "loss": 0.7842,
+      "step": 1629
+    },
+    {
+      "epoch": 0.5028536171525528,
+      "grad_norm": 0.0,
+      "learning_rate": 1.0394720559639295e-05,
+      "loss": 0.8182,
+      "step": 1630
+    },
+    {
+      "epoch": 0.5031621163041802,
+      "grad_norm": 0.0,
+      "learning_rate": 1.0384732631598886e-05,
+      "loss": 0.7417,
+      "step": 1631
+    },
+    {
+      "epoch": 0.5034706154558075,
+      "grad_norm": 0.0,
+      "learning_rate": 1.0374744319170342e-05,
+      "loss": 0.7373,
+      "step": 1632
+    },
+    {
+      "epoch": 0.5037791146074349,
+      "grad_norm": 0.0,
+      "learning_rate": 1.0364755632333034e-05,
+      "loss": 0.7286,
+      "step": 1633
+    },
+    {
+      "epoch": 0.5040876137590622,
+      "grad_norm": 0.0,
+      "learning_rate": 1.0354766581066706e-05,
+      "loss": 0.8905,
+      "step": 1634
+    },
+    {
+      "epoch": 0.5043961129106895,
+      "grad_norm": 0.0,
+      "learning_rate": 1.0344777175351467e-05,
+      "loss": 0.7573,
+      "step": 1635
+    },
+    {
+      "epoch": 0.5047046120623169,
+      "grad_norm": 0.0,
+      "learning_rate": 1.0334787425167772e-05,
+      "loss": 0.7438,
+      "step": 1636
+    },
+    {
+      "epoch": 0.5050131112139442,
+      "grad_norm": 0.0,
+      "learning_rate": 1.0324797340496432e-05,
+      "loss": 0.7304,
+      "step": 1637
+    },
+    {
+      "epoch": 0.5053216103655715,
+      "grad_norm": 0.0,
+      "learning_rate": 1.031480693131858e-05,
+      "loss": 0.6828,
+      "step": 1638
+    },
+    {
+      "epoch": 0.5056301095171988,
+      "grad_norm": 0.0,
+      "learning_rate": 1.0304816207615682e-05,
+      "loss": 0.762,
+      "step": 1639
+    },
+    {
+      "epoch": 0.5059386086688261,
+      "grad_norm": 0.0,
+      "learning_rate": 1.0294825179369518e-05,
+      "loss": 0.6984,
+      "step": 1640
+    },
+    {
+      "epoch": 0.5062471078204535,
+      "grad_norm": 0.0,
+      "learning_rate": 1.0284833856562173e-05,
+      "loss": 0.677,
+      "step": 1641
+    },
+    {
+      "epoch": 0.5065556069720808,
+      "grad_norm": 0.0,
+      "learning_rate": 1.0274842249176016e-05,
+      "loss": 0.7709,
+      "step": 1642
+    },
+    {
+      "epoch": 0.5068641061237081,
+      "grad_norm": 0.0,
+      "learning_rate": 1.0264850367193713e-05,
+      "loss": 0.748,
+      "step": 1643
+    },
+    {
+      "epoch": 0.5071726052753355,
+      "grad_norm": 0.0,
+      "learning_rate": 1.0254858220598194e-05,
+      "loss": 0.671,
+      "step": 1644
+    },
+    {
+      "epoch": 0.5074811044269628,
+      "grad_norm": 0.0,
+      "learning_rate": 1.024486581937266e-05,
+      "loss": 0.7853,
+      "step": 1645
+    },
+    {
+      "epoch": 0.5077896035785902,
+      "grad_norm": 0.0,
+      "learning_rate": 1.0234873173500566e-05,
+      "loss": 0.7042,
+      "step": 1646
+    },
+    {
+      "epoch": 0.5080981027302175,
+      "grad_norm": 0.0,
+      "learning_rate": 1.0224880292965611e-05,
+      "loss": 0.6942,
+      "step": 1647
+    },
+    {
+      "epoch": 0.5084066018818448,
+      "grad_norm": 0.0,
+      "learning_rate": 1.021488718775173e-05,
+      "loss": 0.7299,
+      "step": 1648
+    },
+    {
+      "epoch": 0.5087151010334722,
+      "grad_norm": 0.0,
+      "learning_rate": 1.0204893867843073e-05,
+      "loss": 0.7374,
+      "step": 1649
+    },
+    {
+      "epoch": 0.5090236001850995,
+      "grad_norm": 0.0,
+      "learning_rate": 1.019490034322402e-05,
+      "loss": 0.7705,
+      "step": 1650
+    },
+    {
+      "epoch": 0.5093320993367269,
+      "grad_norm": 0.0,
+      "learning_rate": 1.0184906623879137e-05,
+      "loss": 0.7399,
+      "step": 1651
+    },
+    {
+      "epoch": 0.5096405984883542,
+      "grad_norm": 0.0,
+      "learning_rate": 1.0174912719793202e-05,
+      "loss": 0.7825,
+      "step": 1652
+    },
+    {
+      "epoch": 0.5099490976399815,
+      "grad_norm": 0.0,
+      "learning_rate": 1.0164918640951169e-05,
+      "loss": 0.7848,
+      "step": 1653
+    },
+    {
+      "epoch": 0.5102575967916089,
+      "grad_norm": 0.0,
+      "learning_rate": 1.0154924397338169e-05,
+      "loss": 0.7622,
+      "step": 1654
+    },
+    {
+      "epoch": 0.5105660959432362,
+      "grad_norm": 0.0,
+      "learning_rate": 1.0144929998939491e-05,
+      "loss": 0.8247,
+      "step": 1655
+    },
+    {
+      "epoch": 0.5108745950948634,
+      "grad_norm": 0.0,
+      "learning_rate": 1.0134935455740593e-05,
+      "loss": 0.7686,
+      "step": 1656
+    },
+    {
+      "epoch": 0.5111830942464908,
+      "grad_norm": 0.0,
+      "learning_rate": 1.0124940777727061e-05,
+      "loss": 0.7389,
+      "step": 1657
+    },
+    {
+      "epoch": 0.5114915933981181,
+      "grad_norm": 0.0,
+      "learning_rate": 1.0114945974884627e-05,
+      "loss": 0.7494,
+      "step": 1658
+    },
+    {
+      "epoch": 0.5118000925497455,
+      "grad_norm": 0.0,
+      "learning_rate": 1.0104951057199144e-05,
+      "loss": 0.82,
+      "step": 1659
+    },
+    {
+      "epoch": 0.5121085917013728,
+      "grad_norm": 0.0,
+      "learning_rate": 1.0094956034656582e-05,
+      "loss": 0.749,
+      "step": 1660
+    },
+    {
+      "epoch": 0.5124170908530001,
+      "grad_norm": 0.0,
+      "learning_rate": 1.0084960917243008e-05,
+      "loss": 0.6975,
+      "step": 1661
+    },
+    {
+      "epoch": 0.5127255900046275,
+      "grad_norm": 0.0,
+      "learning_rate": 1.0074965714944593e-05,
+      "loss": 0.8068,
+      "step": 1662
+    },
+    {
+      "epoch": 0.5130340891562548,
+      "grad_norm": 0.0,
+      "learning_rate": 1.0064970437747592e-05,
+      "loss": 0.6657,
+      "step": 1663
+    },
+    {
+      "epoch": 0.5133425883078822,
+      "grad_norm": 0.0,
+      "learning_rate": 1.0054975095638327e-05,
+      "loss": 0.7297,
+      "step": 1664
+    },
+    {
+      "epoch": 0.5136510874595095,
+      "grad_norm": 0.0,
+      "learning_rate": 1.0044979698603195e-05,
+      "loss": 0.6419,
+      "step": 1665
+    },
+    {
+      "epoch": 0.5139595866111368,
+      "grad_norm": 0.0,
+      "learning_rate": 1.0034984256628637e-05,
+      "loss": 0.8095,
+      "step": 1666
+    },
+    {
+      "epoch": 0.5142680857627642,
+      "grad_norm": 0.0,
+      "learning_rate": 1.002498877970115e-05,
+      "loss": 0.7918,
+      "step": 1667
+    },
+    {
+      "epoch": 0.5145765849143915,
+      "grad_norm": 0.0,
+      "learning_rate": 1.001499327780726e-05,
+      "loss": 0.7891,
+      "step": 1668
+    },
+    {
+      "epoch": 0.5148850840660188,
+      "grad_norm": 0.0,
+      "learning_rate": 1.0004997760933517e-05,
+      "loss": 0.7117,
+      "step": 1669
+    },
+    {
+      "epoch": 0.5151935832176462,
+      "grad_norm": 0.0,
+      "learning_rate": 9.995002239066486e-06,
+      "loss": 0.7741,
+      "step": 1670
+    },
+    {
+      "epoch": 0.5155020823692735,
+      "grad_norm": 0.0,
+      "learning_rate": 9.985006722192742e-06,
+      "loss": 0.8163,
+      "step": 1671
+    },
+    {
+      "epoch": 0.5158105815209009,
+      "grad_norm": 0.0,
+      "learning_rate": 9.975011220298853e-06,
+      "loss": 0.7816,
+      "step": 1672
+    },
+    {
+      "epoch": 0.5161190806725281,
+      "grad_norm": 0.0,
+      "learning_rate": 9.965015743371368e-06,
+      "loss": 0.7656,
+      "step": 1673
+    },
+    {
+      "epoch": 0.5164275798241554,
+      "grad_norm": 0.0,
+      "learning_rate": 9.95502030139681e-06,
+      "loss": 0.8098,
+      "step": 1674
+    },
+    {
+      "epoch": 0.5167360789757828,
+      "grad_norm": 0.0,
+      "learning_rate": 9.945024904361675e-06,
+      "loss": 0.7805,
+      "step": 1675
+    },
+    {
+      "epoch": 0.5170445781274101,
+      "grad_norm": 0.0,
+      "learning_rate": 9.935029562252413e-06,
+      "loss": 0.7974,
+      "step": 1676
+    },
+    {
+      "epoch": 0.5173530772790375,
+      "grad_norm": 0.0,
+      "learning_rate": 9.92503428505541e-06,
+      "loss": 0.7451,
+      "step": 1677
+    },
+    {
+      "epoch": 0.5176615764306648,
+      "grad_norm": 0.0,
+      "learning_rate": 9.915039082756995e-06,
+      "loss": 0.7494,
+      "step": 1678
+    },
+    {
+      "epoch": 0.5179700755822921,
+      "grad_norm": 0.0,
+      "learning_rate": 9.905043965343421e-06,
+      "loss": 0.7199,
+      "step": 1679
+    },
+    {
+      "epoch": 0.5182785747339195,
+      "grad_norm": 0.0,
+      "learning_rate": 9.895048942800856e-06,
+      "loss": 0.7528,
+      "step": 1680
+    },
+    {
+      "epoch": 0.5185870738855468,
+      "grad_norm": 0.0,
+      "learning_rate": 9.885054025115378e-06,
+      "loss": 0.7486,
+      "step": 1681
+    },
+    {
+      "epoch": 0.5188955730371742,
+      "grad_norm": 0.0,
+      "learning_rate": 9.875059222272942e-06,
+      "loss": 0.6966,
+      "step": 1682
+    },
+    {
+      "epoch": 0.5192040721888015,
+      "grad_norm": 0.0,
+      "learning_rate": 9.865064544259409e-06,
+      "loss": 0.6967,
+      "step": 1683
+    },
+    {
+      "epoch": 0.5195125713404288,
+      "grad_norm": 0.0,
+      "learning_rate": 9.855070001060508e-06,
+      "loss": 0.8254,
+      "step": 1684
+    },
+    {
+      "epoch": 0.5198210704920562,
+      "grad_norm": 0.0,
+      "learning_rate": 9.845075602661836e-06,
+      "loss": 0.7459,
+      "step": 1685
+    },
+    {
+      "epoch": 0.5201295696436835,
+      "grad_norm": 0.0,
+      "learning_rate": 9.835081359048835e-06,
+      "loss": 0.7506,
+      "step": 1686
+    },
+    {
+      "epoch": 0.5204380687953108,
+      "grad_norm": 0.0,
+      "learning_rate": 9.825087280206801e-06,
+      "loss": 0.8133,
+      "step": 1687
+    },
+    {
+      "epoch": 0.5207465679469382,
+      "grad_norm": 0.0,
+      "learning_rate": 9.815093376120866e-06,
+      "loss": 0.7707,
+      "step": 1688
+    },
+    {
+      "epoch": 0.5210550670985655,
+      "grad_norm": 0.0,
+      "learning_rate": 9.805099656775987e-06,
+      "loss": 0.7421,
+      "step": 1689
+    },
+    {
+      "epoch": 0.5213635662501929,
+      "grad_norm": 0.0,
+      "learning_rate": 9.79510613215693e-06,
+      "loss": 0.6809,
+      "step": 1690
+    },
+    {
+      "epoch": 0.5216720654018201,
+      "grad_norm": 0.0,
+      "learning_rate": 9.785112812248274e-06,
+      "loss": 0.7263,
+      "step": 1691
+    },
+    {
+      "epoch": 0.5219805645534474,
+      "grad_norm": 0.0,
+      "learning_rate": 9.775119707034389e-06,
+      "loss": 0.7779,
+      "step": 1692
+    },
+    {
+      "epoch": 0.5222890637050748,
+      "grad_norm": 0.0,
+      "learning_rate": 9.765126826499437e-06,
+      "loss": 0.7037,
+      "step": 1693
+    },
+    {
+      "epoch": 0.5225975628567021,
+      "grad_norm": 0.0,
+      "learning_rate": 9.755134180627342e-06,
+      "loss": 0.78,
+      "step": 1694
+    },
+    {
+      "epoch": 0.5229060620083295,
+      "grad_norm": 0.0,
+      "learning_rate": 9.74514177940181e-06,
+      "loss": 0.7053,
+      "step": 1695
+    },
+    {
+      "epoch": 0.5232145611599568,
+      "grad_norm": 0.0,
+      "learning_rate": 9.73514963280629e-06,
+      "loss": 0.851,
+      "step": 1696
+    },
+    {
+      "epoch": 0.5235230603115841,
+      "grad_norm": 0.0,
+      "learning_rate": 9.725157750823984e-06,
+      "loss": 0.7866,
+      "step": 1697
+    },
+    {
+      "epoch": 0.5238315594632115,
+      "grad_norm": 0.0,
+      "learning_rate": 9.715166143437832e-06,
+      "loss": 0.7903,
+      "step": 1698
+    },
+    {
+      "epoch": 0.5241400586148388,
+      "grad_norm": 0.0,
+      "learning_rate": 9.705174820630483e-06,
+      "loss": 0.7754,
+      "step": 1699
+    },
+    {
+      "epoch": 0.5244485577664662,
+      "grad_norm": 0.0,
+      "learning_rate": 9.69518379238432e-06,
+      "loss": 0.7465,
+      "step": 1700
+    },
+    {
+      "epoch": 0.5247570569180935,
+      "grad_norm": 0.0,
+      "learning_rate": 9.685193068681423e-06,
+      "loss": 0.7097,
+      "step": 1701
+    },
+    {
+      "epoch": 0.5250655560697208,
+      "grad_norm": 0.0,
+      "learning_rate": 9.675202659503575e-06,
+      "loss": 0.737,
+      "step": 1702
+    },
+    {
+      "epoch": 0.5253740552213482,
+      "grad_norm": 0.0,
+      "learning_rate": 9.665212574832233e-06,
+      "loss": 0.7015,
+      "step": 1703
+    },
+    {
+      "epoch": 0.5256825543729755,
+      "grad_norm": 0.0,
+      "learning_rate": 9.655222824648537e-06,
+      "loss": 0.7699,
+      "step": 1704
+    },
+    {
+      "epoch": 0.5259910535246028,
+      "grad_norm": 0.0,
+      "learning_rate": 9.645233418933296e-06,
+      "loss": 0.6963,
+      "step": 1705
+    },
+    {
+      "epoch": 0.5262995526762302,
+      "grad_norm": 0.0,
+      "learning_rate": 9.63524436766697e-06,
+      "loss": 0.7432,
+      "step": 1706
+    },
+    {
+      "epoch": 0.5266080518278575,
+      "grad_norm": 0.0,
+      "learning_rate": 9.625255680829661e-06,
+      "loss": 0.7409,
+      "step": 1707
+    },
+    {
+      "epoch": 0.5269165509794848,
+      "grad_norm": 0.0,
+      "learning_rate": 9.615267368401118e-06,
+      "loss": 0.7743,
+      "step": 1708
+    },
+    {
+      "epoch": 0.5272250501311121,
+      "grad_norm": 0.0,
+      "learning_rate": 9.605279440360705e-06,
+      "loss": 0.8109,
+      "step": 1709
+    },
+    {
+      "epoch": 0.5275335492827394,
+      "grad_norm": 0.0,
+      "learning_rate": 9.595291906687414e-06,
+      "loss": 0.7718,
+      "step": 1710
+    },
+    {
+      "epoch": 0.5278420484343668,
+      "grad_norm": 0.0,
+      "learning_rate": 9.585304777359826e-06,
+      "loss": 0.7101,
+      "step": 1711
+    },
+    {
+      "epoch": 0.5281505475859941,
+      "grad_norm": 0.0,
+      "learning_rate": 9.57531806235613e-06,
+      "loss": 0.8157,
+      "step": 1712
+    },
+    {
+      "epoch": 0.5284590467376215,
+      "grad_norm": 0.0,
+      "learning_rate": 9.565331771654102e-06,
+      "loss": 0.7767,
+      "step": 1713
+    },
+    {
+      "epoch": 0.5287675458892488,
+      "grad_norm": 0.0,
+      "learning_rate": 9.55534591523109e-06,
+      "loss": 0.7739,
+      "step": 1714
+    },
+    {
+      "epoch": 0.5290760450408761,
+      "grad_norm": 0.0,
+      "learning_rate": 9.545360503064007e-06,
+      "loss": 0.7932,
+      "step": 1715
+    },
+    {
+      "epoch": 0.5293845441925035,
+      "grad_norm": 0.0,
+      "learning_rate": 9.535375545129326e-06,
+      "loss": 0.7008,
+      "step": 1716
+    },
+    {
+      "epoch": 0.5296930433441308,
+      "grad_norm": 0.0,
+      "learning_rate": 9.52539105140306e-06,
+      "loss": 0.7103,
+      "step": 1717
+    },
+    {
+      "epoch": 0.5300015424957581,
+      "grad_norm": 0.0,
+      "learning_rate": 9.515407031860768e-06,
+      "loss": 0.7694,
+      "step": 1718
+    },
+    {
+      "epoch": 0.5303100416473855,
+      "grad_norm": 0.0,
+      "learning_rate": 9.505423496477527e-06,
+      "loss": 0.6776,
+      "step": 1719
+    },
+    {
+      "epoch": 0.5306185407990128,
+      "grad_norm": 0.0,
+      "learning_rate": 9.495440455227932e-06,
+      "loss": 0.7044,
+      "step": 1720
+    },
+    {
+      "epoch": 0.5309270399506402,
+      "grad_norm": 0.0,
+      "learning_rate": 9.485457918086086e-06,
+      "loss": 0.7518,
+      "step": 1721
+    },
+    {
+      "epoch": 0.5312355391022675,
+      "grad_norm": 0.0,
+      "learning_rate": 9.475475895025586e-06,
+      "loss": 0.7754,
+      "step": 1722
+    },
+    {
+      "epoch": 0.5315440382538948,
+      "grad_norm": 0.0,
+      "learning_rate": 9.465494396019524e-06,
+      "loss": 0.6894,
+      "step": 1723
+    },
+    {
+      "epoch": 0.5318525374055222,
+      "grad_norm": 0.0,
+      "learning_rate": 9.455513431040452e-06,
+      "loss": 0.7407,
+      "step": 1724
+    },
+    {
+      "epoch": 0.5321610365571494,
+      "grad_norm": 0.0,
+      "learning_rate": 9.445533010060402e-06,
+      "loss": 0.7462,
+      "step": 1725
+    },
+    {
+      "epoch": 0.5324695357087768,
+      "grad_norm": 0.0,
+      "learning_rate": 9.435553143050856e-06,
+      "loss": 0.7176,
+      "step": 1726
+    },
+    {
+      "epoch": 0.5327780348604041,
+      "grad_norm": 0.0,
+      "learning_rate": 9.42557383998275e-06,
+      "loss": 0.7529,
+      "step": 1727
+    },
+    {
+      "epoch": 0.5330865340120314,
+      "grad_norm": 0.0,
+      "learning_rate": 9.415595110826444e-06,
+      "loss": 0.7301,
+      "step": 1728
+    },
+    {
+      "epoch": 0.5333950331636588,
+      "grad_norm": 0.0,
+      "learning_rate": 9.405616965551738e-06,
+      "loss": 0.7651,
+      "step": 1729
+    },
+    {
+      "epoch": 0.5337035323152861,
+      "grad_norm": 0.0,
+      "learning_rate": 9.395639414127837e-06,
+      "loss": 0.6992,
+      "step": 1730
+    },
+    {
+      "epoch": 0.5340120314669135,
+      "grad_norm": 0.0,
+      "learning_rate": 9.385662466523364e-06,
+      "loss": 0.7285,
+      "step": 1731
+    },
+    {
+      "epoch": 0.5343205306185408,
+      "grad_norm": 0.0,
+      "learning_rate": 9.375686132706333e-06,
+      "loss": 0.8753,
+      "step": 1732
+    },
+    {
+      "epoch": 0.5346290297701681,
+      "grad_norm": 0.0,
+      "learning_rate": 9.36571042264414e-06,
+      "loss": 0.767,
+      "step": 1733
+    },
+    {
+      "epoch": 0.5349375289217955,
+      "grad_norm": 0.0,
+      "learning_rate": 9.355735346303566e-06,
+      "loss": 0.7476,
+      "step": 1734
+    },
+    {
+      "epoch": 0.5352460280734228,
+      "grad_norm": 0.0,
+      "learning_rate": 9.345760913650752e-06,
+      "loss": 0.8124,
+      "step": 1735
+    },
+    {
+      "epoch": 0.5355545272250501,
+      "grad_norm": 0.0,
+      "learning_rate": 9.335787134651207e-06,
+      "loss": 0.8571,
+      "step": 1736
+    },
+    {
+      "epoch": 0.5358630263766775,
+      "grad_norm": 0.0,
+      "learning_rate": 9.32581401926977e-06,
+      "loss": 0.743,
+      "step": 1737
+    },
+    {
+      "epoch": 0.5361715255283048,
+      "grad_norm": 0.0,
+      "learning_rate": 9.31584157747063e-06,
+      "loss": 0.763,
+      "step": 1738
+    },
+    {
+      "epoch": 0.5364800246799322,
+      "grad_norm": 0.0,
+      "learning_rate": 9.305869819217299e-06,
+      "loss": 0.7776,
+      "step": 1739
+    },
+    {
+      "epoch": 0.5367885238315595,
+      "grad_norm": 0.0,
+      "learning_rate": 9.295898754472607e-06,
+      "loss": 0.7812,
+      "step": 1740
+    },
+    {
+      "epoch": 0.5370970229831868,
+      "grad_norm": 0.0,
+      "learning_rate": 9.285928393198691e-06,
+      "loss": 0.8133,
+      "step": 1741
+    },
+    {
+      "epoch": 0.5374055221348142,
+      "grad_norm": 0.0,
+      "learning_rate": 9.275958745356982e-06,
+      "loss": 0.7735,
+      "step": 1742
+    },
+    {
+      "epoch": 0.5377140212864414,
+      "grad_norm": 0.0,
+      "learning_rate": 9.2659898209082e-06,
+      "loss": 0.7431,
+      "step": 1743
+    },
+    {
+      "epoch": 0.5380225204380688,
+      "grad_norm": 0.0,
+      "learning_rate": 9.256021629812349e-06,
+      "loss": 0.7729,
+      "step": 1744
+    },
+    {
+      "epoch": 0.5383310195896961,
+      "grad_norm": 0.0,
+      "learning_rate": 9.24605418202869e-06,
+      "loss": 0.6907,
+      "step": 1745
+    },
+    {
+      "epoch": 0.5386395187413234,
+      "grad_norm": 0.0,
+      "learning_rate": 9.236087487515746e-06,
+      "loss": 0.7184,
+      "step": 1746
+    },
+    {
+      "epoch": 0.5389480178929508,
+      "grad_norm": 0.0,
+      "learning_rate": 9.226121556231287e-06,
+      "loss": 0.6686,
+      "step": 1747
+    },
+    {
+      "epoch": 0.5392565170445781,
+      "grad_norm": 0.0,
+      "learning_rate": 9.21615639813232e-06,
+      "loss": 0.7839,
+      "step": 1748
+    },
+    {
+      "epoch": 0.5395650161962054,
+      "grad_norm": 0.0,
+      "learning_rate": 9.206192023175086e-06,
+      "loss": 0.776,
+      "step": 1749
+    },
+    {
+      "epoch": 0.5398735153478328,
+      "grad_norm": 0.0,
+      "learning_rate": 9.196228441315028e-06,
+      "loss": 0.7371,
+      "step": 1750
+    },
+    {
+      "epoch": 0.5401820144994601,
+      "grad_norm": 0.0,
+      "learning_rate": 9.18626566250681e-06,
+      "loss": 0.6791,
+      "step": 1751
+    },
+    {
+      "epoch": 0.5404905136510875,
+      "grad_norm": 0.0,
+      "learning_rate": 9.176303696704288e-06,
+      "loss": 0.7811,
+      "step": 1752
+    },
+    {
+      "epoch": 0.5407990128027148,
+      "grad_norm": 0.0,
+      "learning_rate": 9.166342553860518e-06,
+      "loss": 0.7032,
+      "step": 1753
+    },
+    {
+      "epoch": 0.5411075119543421,
+      "grad_norm": 0.0,
+      "learning_rate": 9.15638224392771e-06,
+      "loss": 0.7617,
+      "step": 1754
+    },
+    {
+      "epoch": 0.5414160111059695,
+      "grad_norm": 0.0,
+      "learning_rate": 9.146422776857262e-06,
+      "loss": 0.6893,
+      "step": 1755
+    },
+    {
+      "epoch": 0.5417245102575968,
+      "grad_norm": 0.0,
+      "learning_rate": 9.13646416259972e-06,
+      "loss": 0.7536,
+      "step": 1756
+    },
+    {
+      "epoch": 0.5420330094092242,
+      "grad_norm": 0.0,
+      "learning_rate": 9.126506411104786e-06,
+      "loss": 0.756,
+      "step": 1757
+    },
+    {
+      "epoch": 0.5423415085608515,
+      "grad_norm": 0.0,
+      "learning_rate": 9.11654953232129e-06,
+      "loss": 0.81,
+      "step": 1758
+    },
+    {
+      "epoch": 0.5426500077124788,
+      "grad_norm": 0.0,
+      "learning_rate": 9.106593536197198e-06,
+      "loss": 0.7235,
+      "step": 1759
+    },
+    {
+      "epoch": 0.5429585068641061,
+      "grad_norm": 0.0,
+      "learning_rate": 9.09663843267959e-06,
+      "loss": 1.0373,
+      "step": 1760
+    },
+    {
+      "epoch": 0.5432670060157334,
+      "grad_norm": 0.0,
+      "learning_rate": 9.086684231714655e-06,
+      "loss": 0.751,
+      "step": 1761
+    },
+    {
+      "epoch": 0.5435755051673608,
+      "grad_norm": 0.0,
+      "learning_rate": 9.076730943247687e-06,
+      "loss": 0.7119,
+      "step": 1762
+    },
+    {
+      "epoch": 0.5438840043189881,
+      "grad_norm": 0.0,
+      "learning_rate": 9.06677857722305e-06,
+      "loss": 0.7603,
+      "step": 1763
+    },
+    {
+      "epoch": 0.5441925034706154,
+      "grad_norm": 0.0,
+      "learning_rate": 9.056827143584206e-06,
+      "loss": 0.6543,
+      "step": 1764
+    },
+    {
+      "epoch": 0.5445010026222428,
+      "grad_norm": 0.0,
+      "learning_rate": 9.046876652273677e-06,
+      "loss": 0.7651,
+      "step": 1765
+    },
+    {
+      "epoch": 0.5448095017738701,
+      "grad_norm": 0.0,
+      "learning_rate": 9.036927113233049e-06,
+      "loss": 0.8314,
+      "step": 1766
+    },
+    {
+      "epoch": 0.5451180009254974,
+      "grad_norm": 0.0,
+      "learning_rate": 9.026978536402943e-06,
+      "loss": 0.7353,
+      "step": 1767
+    },
+    {
+      "epoch": 0.5454265000771248,
+      "grad_norm": 0.0,
+      "learning_rate": 9.01703093172303e-06,
+      "loss": 0.7053,
+      "step": 1768
+    },
+    {
+      "epoch": 0.5457349992287521,
+      "grad_norm": 0.0,
+      "learning_rate": 9.007084309132013e-06,
+      "loss": 0.7845,
+      "step": 1769
+    },
+    {
+      "epoch": 0.5460434983803795,
+      "grad_norm": 0.0,
+      "learning_rate": 8.997138678567602e-06,
+      "loss": 0.778,
+      "step": 1770
+    },
+    {
+      "epoch": 0.5463519975320068,
+      "grad_norm": 0.0,
+      "learning_rate": 8.987194049966526e-06,
+      "loss": 0.7123,
+      "step": 1771
+    },
+    {
+      "epoch": 0.5466604966836341,
+      "grad_norm": 0.0,
+      "learning_rate": 8.977250433264504e-06,
+      "loss": 0.7844,
+      "step": 1772
+    },
+    {
+      "epoch": 0.5469689958352615,
+      "grad_norm": 0.0,
+      "learning_rate": 8.967307838396256e-06,
+      "loss": 0.7202,
+      "step": 1773
+    },
+    {
+      "epoch": 0.5472774949868888,
+      "grad_norm": 0.0,
+      "learning_rate": 8.957366275295471e-06,
+      "loss": 0.7484,
+      "step": 1774
+    },
+    {
+      "epoch": 0.5475859941385162,
+      "grad_norm": 0.0,
+      "learning_rate": 8.947425753894805e-06,
+      "loss": 0.77,
+      "step": 1775
+    },
+    {
+      "epoch": 0.5478944932901435,
+      "grad_norm": 0.0,
+      "learning_rate": 8.937486284125883e-06,
+      "loss": 0.8058,
+      "step": 1776
+    },
+    {
+      "epoch": 0.5482029924417707,
+      "grad_norm": 0.0,
+      "learning_rate": 8.927547875919275e-06,
+      "loss": 0.8029,
+      "step": 1777
+    },
+    {
+      "epoch": 0.5485114915933981,
+      "grad_norm": 0.0,
+      "learning_rate": 8.91761053920449e-06,
+      "loss": 0.7281,
+      "step": 1778
+    },
+    {
+      "epoch": 0.5488199907450254,
+      "grad_norm": 0.0,
+      "learning_rate": 8.90767428390997e-06,
+      "loss": 1.0318,
+      "step": 1779
+    },
+    {
+      "epoch": 0.5491284898966527,
+      "grad_norm": 0.0,
+      "learning_rate": 8.897739119963065e-06,
+      "loss": 0.8065,
+      "step": 1780
+    },
+    {
+      "epoch": 0.5494369890482801,
+      "grad_norm": 0.0,
+      "learning_rate": 8.887805057290048e-06,
+      "loss": 0.7575,
+      "step": 1781
+    },
+    {
+      "epoch": 0.5497454881999074,
+      "grad_norm": 0.0,
+      "learning_rate": 8.877872105816082e-06,
+      "loss": 0.7086,
+      "step": 1782
+    },
+    {
+      "epoch": 0.5500539873515348,
+      "grad_norm": 0.0,
+      "learning_rate": 8.867940275465233e-06,
+      "loss": 0.7722,
+      "step": 1783
+    },
+    {
+      "epoch": 0.5503624865031621,
+      "grad_norm": 0.0,
+      "learning_rate": 8.85800957616043e-06,
+      "loss": 0.7618,
+      "step": 1784
+    },
+    {
+      "epoch": 0.5506709856547894,
+      "grad_norm": 0.0,
+      "learning_rate": 8.848080017823482e-06,
+      "loss": 0.7556,
+      "step": 1785
+    },
+    {
+      "epoch": 0.5509794848064168,
+      "grad_norm": 0.0,
+      "learning_rate": 8.83815161037506e-06,
+      "loss": 0.8066,
+      "step": 1786
+    },
+    {
+      "epoch": 0.5512879839580441,
+      "grad_norm": 0.0,
+      "learning_rate": 8.828224363734679e-06,
+      "loss": 0.791,
+      "step": 1787
+    },
+    {
+      "epoch": 0.5515964831096715,
+      "grad_norm": 0.0,
+      "learning_rate": 8.81829828782069e-06,
+      "loss": 0.778,
+      "step": 1788
+    },
+    {
+      "epoch": 0.5519049822612988,
+      "grad_norm": 0.0,
+      "learning_rate": 8.80837339255029e-06,
+      "loss": 0.676,
+      "step": 1789
+    },
+    {
+      "epoch": 0.5522134814129261,
+      "grad_norm": 0.0,
+      "learning_rate": 8.798449687839482e-06,
+      "loss": 0.7408,
+      "step": 1790
+    },
+    {
+      "epoch": 0.5525219805645535,
+      "grad_norm": 0.0,
+      "learning_rate": 8.788527183603093e-06,
+      "loss": 0.7563,
+      "step": 1791
+    },
+    {
+      "epoch": 0.5528304797161808,
+      "grad_norm": 0.0,
+      "learning_rate": 8.77860588975473e-06,
+      "loss": 0.7402,
+      "step": 1792
+    },
+    {
+      "epoch": 0.5531389788678082,
+      "grad_norm": 0.0,
+      "learning_rate": 8.768685816206811e-06,
+      "loss": 0.7686,
+      "step": 1793
+    },
+    {
+      "epoch": 0.5534474780194355,
+      "grad_norm": 0.0,
+      "learning_rate": 8.758766972870525e-06,
+      "loss": 0.7298,
+      "step": 1794
+    },
+    {
+      "epoch": 0.5537559771710627,
+      "grad_norm": 0.0,
+      "learning_rate": 8.748849369655833e-06,
+      "loss": 0.7227,
+      "step": 1795
+    },
+    {
+      "epoch": 0.5540644763226901,
+      "grad_norm": 0.0,
+      "learning_rate": 8.738933016471461e-06,
+      "loss": 0.802,
+      "step": 1796
+    },
+    {
+      "epoch": 0.5543729754743174,
+      "grad_norm": 0.0,
+      "learning_rate": 8.729017923224878e-06,
+      "loss": 0.7057,
+      "step": 1797
+    },
+    {
+      "epoch": 0.5546814746259447,
+      "grad_norm": 0.0,
+      "learning_rate": 8.7191040998223e-06,
+      "loss": 0.708,
+      "step": 1798
+    },
+    {
+      "epoch": 0.5549899737775721,
+      "grad_norm": 0.0,
+      "learning_rate": 8.709191556168675e-06,
+      "loss": 0.6959,
+      "step": 1799
+    },
+    {
+      "epoch": 0.5552984729291994,
+      "grad_norm": 0.0,
+      "learning_rate": 8.699280302167674e-06,
+      "loss": 0.7177,
+      "step": 1800
+    },
+    {
+      "epoch": 0.5556069720808268,
+      "grad_norm": 0.0,
+      "learning_rate": 8.689370347721668e-06,
+      "loss": 0.731,
+      "step": 1801
+    },
+    {
+      "epoch": 0.5559154712324541,
+      "grad_norm": 0.0,
+      "learning_rate": 8.679461702731746e-06,
+      "loss": 0.8371,
+      "step": 1802
+    },
+    {
+      "epoch": 0.5562239703840814,
+      "grad_norm": 0.0,
+      "learning_rate": 8.669554377097674e-06,
+      "loss": 0.8826,
+      "step": 1803
+    },
+    {
+      "epoch": 0.5565324695357088,
+      "grad_norm": 0.0,
+      "learning_rate": 8.659648380717914e-06,
+      "loss": 0.6772,
+      "step": 1804
+    },
+    {
+      "epoch": 0.5568409686873361,
+      "grad_norm": 0.0,
+      "learning_rate": 8.649743723489582e-06,
+      "loss": 0.6543,
+      "step": 1805
+    },
+    {
+      "epoch": 0.5571494678389635,
+      "grad_norm": 0.0,
+      "learning_rate": 8.639840415308475e-06,
+      "loss": 0.7435,
+      "step": 1806
+    },
+    {
+      "epoch": 0.5574579669905908,
+      "grad_norm": 0.0,
+      "learning_rate": 8.629938466069028e-06,
+      "loss": 0.6827,
+      "step": 1807
+    },
+    {
+      "epoch": 0.5577664661422181,
+      "grad_norm": 0.0,
+      "learning_rate": 8.62003788566433e-06,
+      "loss": 0.7311,
+      "step": 1808
+    },
+    {
+      "epoch": 0.5580749652938455,
+      "grad_norm": 0.0,
+      "learning_rate": 8.610138683986088e-06,
+      "loss": 0.762,
+      "step": 1809
+    },
+    {
+      "epoch": 0.5583834644454728,
+      "grad_norm": 0.0,
+      "learning_rate": 8.600240870924645e-06,
+      "loss": 0.7338,
+      "step": 1810
+    },
+    {
+      "epoch": 0.5586919635971002,
+      "grad_norm": 0.0,
+      "learning_rate": 8.59034445636895e-06,
+      "loss": 0.7335,
+      "step": 1811
+    },
+    {
+      "epoch": 0.5590004627487274,
+      "grad_norm": 0.0,
+      "learning_rate": 8.580449450206555e-06,
+      "loss": 0.765,
+      "step": 1812
+    },
+    {
+      "epoch": 0.5593089619003547,
+      "grad_norm": 0.0,
+      "learning_rate": 8.570555862323612e-06,
+      "loss": 0.7169,
+      "step": 1813
+    },
+    {
+      "epoch": 0.5596174610519821,
+      "grad_norm": 0.0,
+      "learning_rate": 8.560663702604844e-06,
+      "loss": 0.7687,
+      "step": 1814
+    },
+    {
+      "epoch": 0.5599259602036094,
+      "grad_norm": 0.0,
+      "learning_rate": 8.55077298093355e-06,
+      "loss": 0.6824,
+      "step": 1815
+    },
+    {
+      "epoch": 0.5602344593552367,
+      "grad_norm": 0.0,
+      "learning_rate": 8.540883707191602e-06,
+      "loss": 0.7119,
+      "step": 1816
+    },
+    {
+      "epoch": 0.5605429585068641,
+      "grad_norm": 0.0,
+      "learning_rate": 8.53099589125942e-06,
+      "loss": 0.7964,
+      "step": 1817
+    },
+    {
+      "epoch": 0.5608514576584914,
+      "grad_norm": 0.0,
+      "learning_rate": 8.521109543015958e-06,
+      "loss": 0.7385,
+      "step": 1818
+    },
+    {
+      "epoch": 0.5611599568101188,
+      "grad_norm": 0.0,
+      "learning_rate": 8.511224672338715e-06,
+      "loss": 0.7068,
+      "step": 1819
+    },
+    {
+      "epoch": 0.5614684559617461,
+      "grad_norm": 0.0,
+      "learning_rate": 8.501341289103712e-06,
+      "loss": 1.0249,
+      "step": 1820
+    },
+    {
+      "epoch": 0.5617769551133734,
+      "grad_norm": 0.0,
+      "learning_rate": 8.491459403185485e-06,
+      "loss": 0.7633,
+      "step": 1821
+    },
+    {
+      "epoch": 0.5620854542650008,
+      "grad_norm": 0.0,
+      "learning_rate": 8.481579024457066e-06,
+      "loss": 0.6486,
+      "step": 1822
+    },
+    {
+      "epoch": 0.5623939534166281,
+      "grad_norm": 0.0,
+      "learning_rate": 8.471700162789989e-06,
+      "loss": 0.7773,
+      "step": 1823
+    },
+    {
+      "epoch": 0.5627024525682555,
+      "grad_norm": 0.0,
+      "learning_rate": 8.461822828054269e-06,
+      "loss": 0.7,
+      "step": 1824
+    },
+    {
+      "epoch": 0.5630109517198828,
+      "grad_norm": 0.0,
+      "learning_rate": 8.451947030118397e-06,
+      "loss": 0.7306,
+      "step": 1825
+    },
+    {
+      "epoch": 0.5633194508715101,
+      "grad_norm": 0.0,
+      "learning_rate": 8.442072778849328e-06,
+      "loss": 0.7438,
+      "step": 1826
+    },
+    {
+      "epoch": 0.5636279500231375,
+      "grad_norm": 0.0,
+      "learning_rate": 8.432200084112473e-06,
+      "loss": 0.6776,
+      "step": 1827
+    },
+    {
+      "epoch": 0.5639364491747648,
+      "grad_norm": 0.0,
+      "learning_rate": 8.422328955771683e-06,
+      "loss": 0.7962,
+      "step": 1828
+    },
+    {
+      "epoch": 0.564244948326392,
+      "grad_norm": 0.0,
+      "learning_rate": 8.412459403689249e-06,
+      "loss": 0.7924,
+      "step": 1829
+    },
+    {
+      "epoch": 0.5645534474780194,
+      "grad_norm": 0.0,
+      "learning_rate": 8.40259143772589e-06,
+      "loss": 0.7442,
+      "step": 1830
+    },
+    {
+      "epoch": 0.5648619466296467,
+      "grad_norm": 0.0,
+      "learning_rate": 8.39272506774073e-06,
+      "loss": 0.6769,
+      "step": 1831
+    },
+    {
+      "epoch": 0.5651704457812741,
+      "grad_norm": 0.0,
+      "learning_rate": 8.382860303591306e-06,
+      "loss": 0.7485,
+      "step": 1832
+    },
+    {
+      "epoch": 0.5654789449329014,
+      "grad_norm": 0.0,
+      "learning_rate": 8.372997155133548e-06,
+      "loss": 0.7878,
+      "step": 1833
+    },
+    {
+      "epoch": 0.5657874440845287,
+      "grad_norm": 0.0,
+      "learning_rate": 8.363135632221777e-06,
+      "loss": 0.8033,
+      "step": 1834
+    },
+    {
+      "epoch": 0.5660959432361561,
+      "grad_norm": 0.0,
+      "learning_rate": 8.35327574470868e-06,
+      "loss": 0.6677,
+      "step": 1835
+    },
+    {
+      "epoch": 0.5664044423877834,
+      "grad_norm": 0.0,
+      "learning_rate": 8.34341750244532e-06,
+      "loss": 0.7917,
+      "step": 1836
+    },
+    {
+      "epoch": 0.5667129415394108,
+      "grad_norm": 0.0,
+      "learning_rate": 8.333560915281109e-06,
+      "loss": 0.7731,
+      "step": 1837
+    },
+    {
+      "epoch": 0.5670214406910381,
+      "grad_norm": 0.0,
+      "learning_rate": 8.323705993063813e-06,
+      "loss": 0.7557,
+      "step": 1838
+    },
+    {
+      "epoch": 0.5673299398426654,
+      "grad_norm": 0.0,
+      "learning_rate": 8.313852745639523e-06,
+      "loss": 0.761,
+      "step": 1839
+    },
+    {
+      "epoch": 0.5676384389942928,
+      "grad_norm": 0.0,
+      "learning_rate": 8.304001182852668e-06,
+      "loss": 0.7892,
+      "step": 1840
+    },
+    {
+      "epoch": 0.5679469381459201,
+      "grad_norm": 0.0,
+      "learning_rate": 8.294151314545988e-06,
+      "loss": 0.7484,
+      "step": 1841
+    },
+    {
+      "epoch": 0.5682554372975475,
+      "grad_norm": 0.0,
+      "learning_rate": 8.284303150560538e-06,
+      "loss": 0.7573,
+      "step": 1842
+    },
+    {
+      "epoch": 0.5685639364491748,
+      "grad_norm": 0.0,
+      "learning_rate": 8.274456700735653e-06,
+      "loss": 0.6972,
+      "step": 1843
+    },
+    {
+      "epoch": 0.5688724356008021,
+      "grad_norm": 0.0,
+      "learning_rate": 8.26461197490897e-06,
+      "loss": 0.7,
+      "step": 1844
+    },
+    {
+      "epoch": 0.5691809347524295,
+      "grad_norm": 0.0,
+      "learning_rate": 8.2547689829164e-06,
+      "loss": 0.7058,
+      "step": 1845
+    },
+    {
+      "epoch": 0.5694894339040568,
+      "grad_norm": 0.0,
+      "learning_rate": 8.24492773459212e-06,
+      "loss": 0.7191,
+      "step": 1846
+    },
+    {
+      "epoch": 0.569797933055684,
+      "grad_norm": 0.0,
+      "learning_rate": 8.235088239768577e-06,
+      "loss": 0.7849,
+      "step": 1847
+    },
+    {
+      "epoch": 0.5701064322073114,
+      "grad_norm": 0.0,
+      "learning_rate": 8.225250508276439e-06,
+      "loss": 0.7633,
+      "step": 1848
+    },
+    {
+      "epoch": 0.5704149313589387,
+      "grad_norm": 0.0,
+      "learning_rate": 8.215414549944636e-06,
+      "loss": 0.6931,
+      "step": 1849
+    },
+    {
+      "epoch": 0.5707234305105661,
+      "grad_norm": 0.0,
+      "learning_rate": 8.20558037460032e-06,
+      "loss": 0.6946,
+      "step": 1850
+    },
+    {
+      "epoch": 0.5710319296621934,
+      "grad_norm": 0.0,
+      "learning_rate": 8.19574799206886e-06,
+      "loss": 0.6747,
+      "step": 1851
+    },
+    {
+      "epoch": 0.5713404288138207,
+      "grad_norm": 0.0,
+      "learning_rate": 8.185917412173832e-06,
+      "loss": 0.7037,
+      "step": 1852
+    },
+    {
+      "epoch": 0.5716489279654481,
+      "grad_norm": 0.0,
+      "learning_rate": 8.176088644737015e-06,
+      "loss": 0.7733,
+      "step": 1853
+    },
+    {
+      "epoch": 0.5719574271170754,
+      "grad_norm": 0.0,
+      "learning_rate": 8.166261699578375e-06,
+      "loss": 0.7072,
+      "step": 1854
+    },
+    {
+      "epoch": 0.5722659262687028,
+      "grad_norm": 0.0,
+      "learning_rate": 8.156436586516064e-06,
+      "loss": 0.7545,
+      "step": 1855
+    },
+    {
+      "epoch": 0.5725744254203301,
+      "grad_norm": 0.0,
+      "learning_rate": 8.146613315366387e-06,
+      "loss": 0.7408,
+      "step": 1856
+    },
+    {
+      "epoch": 0.5728829245719574,
+      "grad_norm": 0.0,
+      "learning_rate": 8.136791895943825e-06,
+      "loss": 0.7409,
+      "step": 1857
+    },
+    {
+      "epoch": 0.5731914237235848,
+      "grad_norm": 0.0,
+      "learning_rate": 8.126972338060997e-06,
+      "loss": 0.7324,
+      "step": 1858
+    },
+    {
+      "epoch": 0.5734999228752121,
+      "grad_norm": 0.0,
+      "learning_rate": 8.117154651528676e-06,
+      "loss": 0.8158,
+      "step": 1859
+    },
+    {
+      "epoch": 0.5738084220268395,
+      "grad_norm": 0.0,
+      "learning_rate": 8.107338846155762e-06,
+      "loss": 0.7496,
+      "step": 1860
+    },
+    {
+      "epoch": 0.5741169211784668,
+      "grad_norm": 0.0,
+      "learning_rate": 8.097524931749256e-06,
+      "loss": 0.7526,
+      "step": 1861
+    },
+    {
+      "epoch": 0.5744254203300941,
+      "grad_norm": 0.0,
+      "learning_rate": 8.087712918114294e-06,
+      "loss": 0.7048,
+      "step": 1862
+    },
+    {
+      "epoch": 0.5747339194817215,
+      "grad_norm": 0.0,
+      "learning_rate": 8.077902815054102e-06,
+      "loss": 0.7347,
+      "step": 1863
+    },
+    {
+      "epoch": 0.5750424186333487,
+      "grad_norm": 0.0,
+      "learning_rate": 8.06809463237e-06,
+      "loss": 1.0299,
+      "step": 1864
+    },
+    {
+      "epoch": 0.575350917784976,
+      "grad_norm": 0.0,
+      "learning_rate": 8.058288379861387e-06,
+      "loss": 0.7647,
+      "step": 1865
+    },
+    {
+      "epoch": 0.5756594169366034,
+      "grad_norm": 0.0,
+      "learning_rate": 8.048484067325735e-06,
+      "loss": 0.7311,
+      "step": 1866
+    },
+    {
+      "epoch": 0.5759679160882307,
+      "grad_norm": 0.0,
+      "learning_rate": 8.038681704558578e-06,
+      "loss": 0.6969,
+      "step": 1867
+    },
+    {
+      "epoch": 0.5762764152398581,
+      "grad_norm": 0.0,
+      "learning_rate": 8.028881301353503e-06,
+      "loss": 0.7247,
+      "step": 1868
+    },
+    {
+      "epoch": 0.5765849143914854,
+      "grad_norm": 0.0,
+      "learning_rate": 8.019082867502132e-06,
+      "loss": 0.713,
+      "step": 1869
+    },
+    {
+      "epoch": 0.5768934135431127,
+      "grad_norm": 0.0,
+      "learning_rate": 8.009286412794126e-06,
+      "loss": 0.7307,
+      "step": 1870
+    },
+    {
+      "epoch": 0.5772019126947401,
+      "grad_norm": 0.0,
+      "learning_rate": 7.999491947017174e-06,
+      "loss": 0.6996,
+      "step": 1871
+    },
+    {
+      "epoch": 0.5775104118463674,
+      "grad_norm": 0.0,
+      "learning_rate": 7.989699479956972e-06,
+      "loss": 0.7088,
+      "step": 1872
+    },
+    {
+      "epoch": 0.5778189109979948,
+      "grad_norm": 0.0,
+      "learning_rate": 7.97990902139721e-06,
+      "loss": 0.7401,
+      "step": 1873
+    },
+    {
+      "epoch": 0.5781274101496221,
+      "grad_norm": 0.0,
+      "learning_rate": 7.970120581119584e-06,
+      "loss": 0.7832,
+      "step": 1874
+    },
+    {
+      "epoch": 0.5784359093012494,
+      "grad_norm": 0.0,
+      "learning_rate": 7.960334168903769e-06,
+      "loss": 0.6926,
+      "step": 1875
+    },
+    {
+      "epoch": 0.5787444084528768,
+      "grad_norm": 0.0,
+      "learning_rate": 7.950549794527418e-06,
+      "loss": 0.7896,
+      "step": 1876
+    },
+    {
+      "epoch": 0.5790529076045041,
+      "grad_norm": 0.0,
+      "learning_rate": 7.940767467766142e-06,
+      "loss": 0.7093,
+      "step": 1877
+    },
+    {
+      "epoch": 0.5793614067561315,
+      "grad_norm": 0.0,
+      "learning_rate": 7.930987198393506e-06,
+      "loss": 0.7789,
+      "step": 1878
+    },
+    {
+      "epoch": 0.5796699059077588,
+      "grad_norm": 0.0,
+      "learning_rate": 7.921208996181022e-06,
+      "loss": 0.7411,
+      "step": 1879
+    },
+    {
+      "epoch": 0.5799784050593861,
+      "grad_norm": 0.0,
+      "learning_rate": 7.911432870898139e-06,
+      "loss": 0.7569,
+      "step": 1880
+    },
+    {
+      "epoch": 0.5802869042110134,
+      "grad_norm": 0.0,
+      "learning_rate": 7.901658832312234e-06,
+      "loss": 0.6853,
+      "step": 1881
+    },
+    {
+      "epoch": 0.5805954033626407,
+      "grad_norm": 0.0,
+      "learning_rate": 7.891886890188578e-06,
+      "loss": 0.8369,
+      "step": 1882
+    },
+    {
+      "epoch": 0.580903902514268,
+      "grad_norm": 0.0,
+      "learning_rate": 7.882117054290375e-06,
+      "loss": 0.804,
+      "step": 1883
+    },
+    {
+      "epoch": 0.5812124016658954,
+      "grad_norm": 0.0,
+      "learning_rate": 7.872349334378712e-06,
+      "loss": 0.7166,
+      "step": 1884
+    },
+    {
+      "epoch": 0.5815209008175227,
+      "grad_norm": 0.0,
+      "learning_rate": 7.862583740212564e-06,
+      "loss": 0.8346,
+      "step": 1885
+    },
+    {
+      "epoch": 0.5818293999691501,
+      "grad_norm": 0.0,
+      "learning_rate": 7.852820281548773e-06,
+      "loss": 0.6754,
+      "step": 1886
+    },
+    {
+      "epoch": 0.5821378991207774,
+      "grad_norm": 0.0,
+      "learning_rate": 7.84305896814206e-06,
+      "loss": 0.727,
+      "step": 1887
+    },
+    {
+      "epoch": 0.5824463982724047,
+      "grad_norm": 0.0,
+      "learning_rate": 7.833299809744997e-06,
+      "loss": 0.7691,
+      "step": 1888
+    },
+    {
+      "epoch": 0.5827548974240321,
+      "grad_norm": 0.0,
+      "learning_rate": 7.823542816108007e-06,
+      "loss": 0.7847,
+      "step": 1889
+    },
+    {
+      "epoch": 0.5830633965756594,
+      "grad_norm": 0.0,
+      "learning_rate": 7.813787996979343e-06,
+      "loss": 0.7657,
+      "step": 1890
+    },
+    {
+      "epoch": 0.5833718957272868,
+      "grad_norm": 0.0,
+      "learning_rate": 7.804035362105092e-06,
+      "loss": 0.7125,
+      "step": 1891
+    },
+    {
+      "epoch": 0.5836803948789141,
+      "grad_norm": 0.0,
+      "learning_rate": 7.794284921229151e-06,
+      "loss": 0.7207,
+      "step": 1892
+    },
+    {
+      "epoch": 0.5839888940305414,
+      "grad_norm": 0.0,
+      "learning_rate": 7.784536684093237e-06,
+      "loss": 0.7243,
+      "step": 1893
+    },
+    {
+      "epoch": 0.5842973931821688,
+      "grad_norm": 0.0,
+      "learning_rate": 7.774790660436857e-06,
+      "loss": 0.6716,
+      "step": 1894
+    },
+    {
+      "epoch": 0.5846058923337961,
+      "grad_norm": 0.0,
+      "learning_rate": 7.765046859997303e-06,
+      "loss": 0.7218,
+      "step": 1895
+    },
+    {
+      "epoch": 0.5849143914854235,
+      "grad_norm": 0.0,
+      "learning_rate": 7.755305292509656e-06,
+      "loss": 0.7652,
+      "step": 1896
+    },
+    {
+      "epoch": 0.5852228906370508,
+      "grad_norm": 0.0,
+      "learning_rate": 7.745565967706757e-06,
+      "loss": 0.7331,
+      "step": 1897
+    },
+    {
+      "epoch": 0.5855313897886781,
+      "grad_norm": 0.0,
+      "learning_rate": 7.735828895319215e-06,
+      "loss": 0.7257,
+      "step": 1898
+    },
+    {
+      "epoch": 0.5858398889403054,
+      "grad_norm": 0.0,
+      "learning_rate": 7.726094085075377e-06,
+      "loss": 0.7037,
+      "step": 1899
+    },
+    {
+      "epoch": 0.5861483880919327,
+      "grad_norm": 0.0,
+      "learning_rate": 7.716361546701337e-06,
+      "loss": 0.7186,
+      "step": 1900
+    },
+    {
+      "epoch": 0.58645688724356,
+      "grad_norm": 0.0,
+      "learning_rate": 7.706631289920923e-06,
+      "loss": 0.7586,
+      "step": 1901
+    },
+    {
+      "epoch": 0.5867653863951874,
+      "grad_norm": 0.0,
+      "learning_rate": 7.696903324455678e-06,
+      "loss": 0.7501,
+      "step": 1902
+    },
+    {
+      "epoch": 0.5870738855468147,
+      "grad_norm": 0.0,
+      "learning_rate": 7.687177660024854e-06,
+      "loss": 0.7695,
+      "step": 1903
+    },
+    {
+      "epoch": 0.5873823846984421,
+      "grad_norm": 0.0,
+      "learning_rate": 7.677454306345408e-06,
+      "loss": 0.7259,
+      "step": 1904
+    },
+    {
+      "epoch": 0.5876908838500694,
+      "grad_norm": 0.0,
+      "learning_rate": 7.667733273131989e-06,
+      "loss": 0.7603,
+      "step": 1905
+    },
+    {
+      "epoch": 0.5879993830016967,
+      "grad_norm": 0.0,
+      "learning_rate": 7.658014570096926e-06,
+      "loss": 0.6914,
+      "step": 1906
+    },
+    {
+      "epoch": 0.5883078821533241,
+      "grad_norm": 0.0,
+      "learning_rate": 7.648298206950216e-06,
+      "loss": 0.7278,
+      "step": 1907
+    },
+    {
+      "epoch": 0.5886163813049514,
+      "grad_norm": 0.0,
+      "learning_rate": 7.638584193399524e-06,
+      "loss": 0.7888,
+      "step": 1908
+    },
+    {
+      "epoch": 0.5889248804565788,
+      "grad_norm": 0.0,
+      "learning_rate": 7.628872539150165e-06,
+      "loss": 0.7208,
+      "step": 1909
+    },
+    {
+      "epoch": 0.5892333796082061,
+      "grad_norm": 0.0,
+      "learning_rate": 7.619163253905097e-06,
+      "loss": 0.6926,
+      "step": 1910
+    },
+    {
+      "epoch": 0.5895418787598334,
+      "grad_norm": 0.0,
+      "learning_rate": 7.609456347364919e-06,
+      "loss": 0.7166,
+      "step": 1911
+    },
+    {
+      "epoch": 0.5898503779114608,
+      "grad_norm": 0.0,
+      "learning_rate": 7.599751829227832e-06,
+      "loss": 0.7261,
+      "step": 1912
+    },
+    {
+      "epoch": 0.5901588770630881,
+      "grad_norm": 0.0,
+      "learning_rate": 7.590049709189671e-06,
+      "loss": 0.6931,
+      "step": 1913
+    },
+    {
+      "epoch": 0.5904673762147155,
+      "grad_norm": 0.0,
+      "learning_rate": 7.580349996943868e-06,
+      "loss": 0.6828,
+      "step": 1914
+    },
+    {
+      "epoch": 0.5907758753663428,
+      "grad_norm": 0.0,
+      "learning_rate": 7.570652702181454e-06,
+      "loss": 0.727,
+      "step": 1915
+    },
+    {
+      "epoch": 0.59108437451797,
+      "grad_norm": 0.0,
+      "learning_rate": 7.560957834591034e-06,
+      "loss": 0.7789,
+      "step": 1916
+    },
+    {
+      "epoch": 0.5913928736695974,
+      "grad_norm": 0.0,
+      "learning_rate": 7.551265403858797e-06,
+      "loss": 0.7021,
+      "step": 1917
+    },
+    {
+      "epoch": 0.5917013728212247,
+      "grad_norm": 0.0,
+      "learning_rate": 7.541575419668497e-06,
+      "loss": 0.6739,
+      "step": 1918
+    },
+    {
+      "epoch": 0.592009871972852,
+      "grad_norm": 0.0,
+      "learning_rate": 7.531887891701441e-06,
+      "loss": 0.8115,
+      "step": 1919
+    },
+    {
+      "epoch": 0.5923183711244794,
+      "grad_norm": 0.0,
+      "learning_rate": 7.522202829636478e-06,
+      "loss": 0.7618,
+      "step": 1920
+    },
+    {
+      "epoch": 0.5926268702761067,
+      "grad_norm": 0.0,
+      "learning_rate": 7.512520243150003e-06,
+      "loss": 0.7356,
+      "step": 1921
+    },
+    {
+      "epoch": 0.5929353694277341,
+      "grad_norm": 0.0,
+      "learning_rate": 7.50284014191593e-06,
+      "loss": 0.7187,
+      "step": 1922
+    },
+    {
+      "epoch": 0.5932438685793614,
+      "grad_norm": 0.0,
+      "learning_rate": 7.493162535605698e-06,
+      "loss": 0.7121,
+      "step": 1923
+    },
+    {
+      "epoch": 0.5935523677309887,
+      "grad_norm": 0.0,
+      "learning_rate": 7.483487433888238e-06,
+      "loss": 0.7162,
+      "step": 1924
+    },
+    {
+      "epoch": 0.5938608668826161,
+      "grad_norm": 0.0,
+      "learning_rate": 7.473814846429993e-06,
+      "loss": 0.6884,
+      "step": 1925
+    },
+    {
+      "epoch": 0.5941693660342434,
+      "grad_norm": 0.0,
+      "learning_rate": 7.46414478289489e-06,
+      "loss": 0.7447,
+      "step": 1926
+    },
+    {
+      "epoch": 0.5944778651858708,
+      "grad_norm": 0.0,
+      "learning_rate": 7.4544772529443295e-06,
+      "loss": 0.7128,
+      "step": 1927
+    },
+    {
+      "epoch": 0.5947863643374981,
+      "grad_norm": 0.0,
+      "learning_rate": 7.444812266237198e-06,
+      "loss": 0.665,
+      "step": 1928
+    },
+    {
+      "epoch": 0.5950948634891254,
+      "grad_norm": 0.0,
+      "learning_rate": 7.435149832429812e-06,
+      "loss": 0.7802,
+      "step": 1929
+    },
+    {
+      "epoch": 0.5954033626407528,
+      "grad_norm": 0.0,
+      "learning_rate": 7.4254899611759616e-06,
+      "loss": 1.0484,
+      "step": 1930
+    },
+    {
+      "epoch": 0.5957118617923801,
+      "grad_norm": 0.0,
+      "learning_rate": 7.415832662126865e-06,
+      "loss": 0.7477,
+      "step": 1931
+    },
+    {
+      "epoch": 0.5960203609440075,
+      "grad_norm": 0.0,
+      "learning_rate": 7.406177944931179e-06,
+      "loss": 0.6942,
+      "step": 1932
+    },
+    {
+      "epoch": 0.5963288600956347,
+      "grad_norm": 0.0,
+      "learning_rate": 7.396525819234969e-06,
+      "loss": 0.7013,
+      "step": 1933
+    },
+    {
+      "epoch": 0.596637359247262,
+      "grad_norm": 0.0,
+      "learning_rate": 7.386876294681722e-06,
+      "loss": 0.7359,
+      "step": 1934
+    },
+    {
+      "epoch": 0.5969458583988894,
+      "grad_norm": 0.0,
+      "learning_rate": 7.377229380912321e-06,
+      "loss": 0.7121,
+      "step": 1935
+    },
+    {
+      "epoch": 0.5972543575505167,
+      "grad_norm": 0.0,
+      "learning_rate": 7.367585087565046e-06,
+      "loss": 0.7173,
+      "step": 1936
+    },
+    {
+      "epoch": 0.597562856702144,
+      "grad_norm": 0.0,
+      "learning_rate": 7.357943424275547e-06,
+      "loss": 0.7273,
+      "step": 1937
+    },
+    {
+      "epoch": 0.5978713558537714,
+      "grad_norm": 0.0,
+      "learning_rate": 7.348304400676856e-06,
+      "loss": 0.7294,
+      "step": 1938
+    },
+    {
+      "epoch": 0.5981798550053987,
+      "grad_norm": 0.0,
+      "learning_rate": 7.338668026399365e-06,
+      "loss": 0.7416,
+      "step": 1939
+    },
+    {
+      "epoch": 0.5984883541570261,
+      "grad_norm": 0.0,
+      "learning_rate": 7.329034311070828e-06,
+      "loss": 0.8225,
+      "step": 1940
+    },
+    {
+      "epoch": 0.5987968533086534,
+      "grad_norm": 0.0,
+      "learning_rate": 7.319403264316325e-06,
+      "loss": 0.7122,
+      "step": 1941
+    },
+    {
+      "epoch": 0.5991053524602807,
+      "grad_norm": 0.0,
+      "learning_rate": 7.30977489575828e-06,
+      "loss": 0.6605,
+      "step": 1942
+    },
+    {
+      "epoch": 0.5994138516119081,
+      "grad_norm": 0.0,
+      "learning_rate": 7.300149215016442e-06,
+      "loss": 0.7746,
+      "step": 1943
+    },
+    {
+      "epoch": 0.5997223507635354,
+      "grad_norm": 0.0,
+      "learning_rate": 7.290526231707873e-06,
+      "loss": 0.6528,
+      "step": 1944
+    },
+    {
+      "epoch": 0.6000308499151628,
+      "grad_norm": 0.0,
+      "learning_rate": 7.28090595544694e-06,
+      "loss": 0.7022,
+      "step": 1945
+    },
+    {
+      "epoch": 0.6003393490667901,
+      "grad_norm": 0.0,
+      "learning_rate": 7.271288395845302e-06,
+      "loss": 0.6744,
+      "step": 1946
+    },
+    {
+      "epoch": 0.6006478482184174,
+      "grad_norm": 0.0,
+      "learning_rate": 7.2616735625119085e-06,
+      "loss": 0.6828,
+      "step": 1947
+    },
+    {
+      "epoch": 0.6009563473700448,
+      "grad_norm": 0.0,
+      "learning_rate": 7.252061465052984e-06,
+      "loss": 0.7555,
+      "step": 1948
+    },
+    {
+      "epoch": 0.6012648465216721,
+      "grad_norm": 0.0,
+      "learning_rate": 7.242452113072022e-06,
+      "loss": 0.7715,
+      "step": 1949
+    },
+    {
+      "epoch": 0.6015733456732995,
+      "grad_norm": 0.0,
+      "learning_rate": 7.232845516169764e-06,
+      "loss": 0.7449,
+      "step": 1950
+    },
+    {
+      "epoch": 0.6018818448249267,
+      "grad_norm": 0.0,
+      "learning_rate": 7.223241683944204e-06,
+      "loss": 0.7889,
+      "step": 1951
+    },
+    {
+      "epoch": 0.602190343976554,
+      "grad_norm": 0.0,
+      "learning_rate": 7.213640625990582e-06,
+      "loss": 0.6666,
+      "step": 1952
+    },
+    {
+      "epoch": 0.6024988431281814,
+      "grad_norm": 0.0,
+      "learning_rate": 7.204042351901359e-06,
+      "loss": 1.0587,
+      "step": 1953
+    },
+    {
+      "epoch": 0.6028073422798087,
+      "grad_norm": 0.0,
+      "learning_rate": 7.194446871266206e-06,
+      "loss": 0.6755,
+      "step": 1954
+    },
+    {
+      "epoch": 0.603115841431436,
+      "grad_norm": 0.0,
+      "learning_rate": 7.184854193672017e-06,
+      "loss": 0.7586,
+      "step": 1955
+    },
+    {
+      "epoch": 0.6034243405830634,
+      "grad_norm": 0.0,
+      "learning_rate": 7.175264328702878e-06,
+      "loss": 0.7476,
+      "step": 1956
+    },
+    {
+      "epoch": 0.6037328397346907,
+      "grad_norm": 0.0,
+      "learning_rate": 7.165677285940071e-06,
+      "loss": 0.6872,
+      "step": 1957
+    },
+    {
+      "epoch": 0.6040413388863181,
+      "grad_norm": 0.0,
+      "learning_rate": 7.156093074962052e-06,
+      "loss": 0.7498,
+      "step": 1958
+    },
+    {
+      "epoch": 0.6043498380379454,
+      "grad_norm": 0.0,
+      "learning_rate": 7.1465117053444465e-06,
+      "loss": 0.7997,
+      "step": 1959
+    },
+    {
+      "epoch": 0.6046583371895727,
+      "grad_norm": 0.0,
+      "learning_rate": 7.136933186660049e-06,
+      "loss": 0.6762,
+      "step": 1960
+    },
+    {
+      "epoch": 0.6049668363412001,
+      "grad_norm": 0.0,
+      "learning_rate": 7.1273575284788e-06,
+      "loss": 0.8435,
+      "step": 1961
+    },
+    {
+      "epoch": 0.6052753354928274,
+      "grad_norm": 0.0,
+      "learning_rate": 7.117784740367788e-06,
+      "loss": 0.8126,
+      "step": 1962
+    },
+    {
+      "epoch": 0.6055838346444548,
+      "grad_norm": 0.0,
+      "learning_rate": 7.108214831891219e-06,
+      "loss": 0.6898,
+      "step": 1963
+    },
+    {
+      "epoch": 0.6058923337960821,
+      "grad_norm": 0.0,
+      "learning_rate": 7.09864781261044e-06,
+      "loss": 0.7271,
+      "step": 1964
+    },
+    {
+      "epoch": 0.6062008329477094,
+      "grad_norm": 0.0,
+      "learning_rate": 7.089083692083902e-06,
+      "loss": 0.7512,
+      "step": 1965
+    },
+    {
+      "epoch": 0.6065093320993368,
+      "grad_norm": 0.0,
+      "learning_rate": 7.0795224798671666e-06,
+      "loss": 0.8055,
+      "step": 1966
+    },
+    {
+      "epoch": 0.6068178312509641,
+      "grad_norm": 0.0,
+      "learning_rate": 7.069964185512874e-06,
+      "loss": 0.6799,
+      "step": 1967
+    },
+    {
+      "epoch": 0.6071263304025913,
+      "grad_norm": 0.0,
+      "learning_rate": 7.060408818570768e-06,
+      "loss": 0.7722,
+      "step": 1968
+    },
+    {
+      "epoch": 0.6074348295542187,
+      "grad_norm": 0.0,
+      "learning_rate": 7.050856388587655e-06,
+      "loss": 0.7811,
+      "step": 1969
+    },
+    {
+      "epoch": 0.607743328705846,
+      "grad_norm": 0.0,
+      "learning_rate": 7.0413069051074146e-06,
+      "loss": 0.6801,
+      "step": 1970
+    },
+    {
+      "epoch": 0.6080518278574734,
+      "grad_norm": 0.0,
+      "learning_rate": 7.031760377670978e-06,
+      "loss": 0.7812,
+      "step": 1971
+    },
+    {
+      "epoch": 0.6083603270091007,
+      "grad_norm": 0.0,
+      "learning_rate": 7.022216815816323e-06,
+      "loss": 0.7256,
+      "step": 1972
+    },
+    {
+      "epoch": 0.608668826160728,
+      "grad_norm": 0.0,
+      "learning_rate": 7.012676229078469e-06,
+      "loss": 0.7172,
+      "step": 1973
+    },
+    {
+      "epoch": 0.6089773253123554,
+      "grad_norm": 0.0,
+      "learning_rate": 7.003138626989457e-06,
+      "loss": 1.0376,
+      "step": 1974
+    },
+    {
+      "epoch": 0.6092858244639827,
+      "grad_norm": 0.0,
+      "learning_rate": 6.993604019078354e-06,
+      "loss": 0.6692,
+      "step": 1975
+    },
+    {
+      "epoch": 0.6095943236156101,
+      "grad_norm": 0.0,
+      "learning_rate": 6.984072414871223e-06,
+      "loss": 0.7331,
+      "step": 1976
+    },
+    {
+      "epoch": 0.6099028227672374,
+      "grad_norm": 0.0,
+      "learning_rate": 6.974543823891138e-06,
+      "loss": 0.7152,
+      "step": 1977
+    },
+    {
+      "epoch": 0.6102113219188647,
+      "grad_norm": 0.0,
+      "learning_rate": 6.965018255658156e-06,
+      "loss": 0.7579,
+      "step": 1978
+    },
+    {
+      "epoch": 0.6105198210704921,
+      "grad_norm": 0.0,
+      "learning_rate": 6.955495719689321e-06,
+      "loss": 0.6,
+      "step": 1979
+    },
+    {
+      "epoch": 0.6108283202221194,
+      "grad_norm": 0.0,
+      "learning_rate": 6.9459762254986344e-06,
+      "loss": 0.8325,
+      "step": 1980
+    },
+    {
+      "epoch": 0.6111368193737468,
+      "grad_norm": 0.0,
+      "learning_rate": 6.936459782597069e-06,
+      "loss": 0.7452,
+      "step": 1981
+    },
+    {
+      "epoch": 0.6114453185253741,
+      "grad_norm": 0.0,
+      "learning_rate": 6.9269464004925466e-06,
+      "loss": 0.6906,
+      "step": 1982
+    },
+    {
+      "epoch": 0.6117538176770014,
+      "grad_norm": 0.0,
+      "learning_rate": 6.917436088689935e-06,
+      "loss": 0.7581,
+      "step": 1983
+    },
+    {
+      "epoch": 0.6120623168286288,
+      "grad_norm": 0.0,
+      "learning_rate": 6.907928856691024e-06,
+      "loss": 0.6741,
+      "step": 1984
+    },
+    {
+      "epoch": 0.612370815980256,
+      "grad_norm": 0.0,
+      "learning_rate": 6.898424713994536e-06,
+      "loss": 0.6596,
+      "step": 1985
+    },
+    {
+      "epoch": 0.6126793151318833,
+      "grad_norm": 0.0,
+      "learning_rate": 6.888923670096102e-06,
+      "loss": 0.748,
+      "step": 1986
+    },
+    {
+      "epoch": 0.6129878142835107,
+      "grad_norm": 0.0,
+      "learning_rate": 6.879425734488261e-06,
+      "loss": 0.7023,
+      "step": 1987
+    },
+    {
+      "epoch": 0.613296313435138,
+      "grad_norm": 0.0,
+      "learning_rate": 6.86993091666044e-06,
+      "loss": 0.7277,
+      "step": 1988
+    },
+    {
+      "epoch": 0.6136048125867654,
+      "grad_norm": 0.0,
+      "learning_rate": 6.860439226098956e-06,
+      "loss": 0.8299,
+      "step": 1989
+    },
+    {
+      "epoch": 0.6139133117383927,
+      "grad_norm": 0.0,
+      "learning_rate": 6.850950672287003e-06,
+      "loss": 0.8051,
+      "step": 1990
+    },
+    {
+      "epoch": 0.61422181089002,
+      "grad_norm": 0.0,
+      "learning_rate": 6.841465264704636e-06,
+      "loss": 0.8194,
+      "step": 1991
+    },
+    {
+      "epoch": 0.6145303100416474,
+      "grad_norm": 0.0,
+      "learning_rate": 6.831983012828775e-06,
+      "loss": 0.7889,
+      "step": 1992
+    },
+    {
+      "epoch": 0.6148388091932747,
+      "grad_norm": 0.0,
+      "learning_rate": 6.82250392613317e-06,
+      "loss": 0.7253,
+      "step": 1993
+    },
+    {
+      "epoch": 0.615147308344902,
+      "grad_norm": 0.0,
+      "learning_rate": 6.8130280140884286e-06,
+      "loss": 0.7615,
+      "step": 1994
+    },
+    {
+      "epoch": 0.6154558074965294,
+      "grad_norm": 0.0,
+      "learning_rate": 6.803555286161973e-06,
+      "loss": 0.6758,
+      "step": 1995
+    },
+    {
+      "epoch": 0.6157643066481567,
+      "grad_norm": 0.0,
+      "learning_rate": 6.7940857518180555e-06,
+      "loss": 0.8349,
+      "step": 1996
+    },
+    {
+      "epoch": 0.6160728057997841,
+      "grad_norm": 0.0,
+      "learning_rate": 6.784619420517724e-06,
+      "loss": 0.7385,
+      "step": 1997
+    },
+    {
+      "epoch": 0.6163813049514114,
+      "grad_norm": 0.0,
+      "learning_rate": 6.775156301718837e-06,
+      "loss": 0.6718,
+      "step": 1998
+    },
+    {
+      "epoch": 0.6166898041030388,
+      "grad_norm": 0.0,
+      "learning_rate": 6.765696404876039e-06,
+      "loss": 0.7217,
+      "step": 1999
+    },
+    {
+      "epoch": 0.6169983032546661,
+      "grad_norm": 0.0,
+      "learning_rate": 6.756239739440758e-06,
+      "loss": 0.8079,
+      "step": 2000
+    },
+    {
+      "epoch": 0.6173068024062934,
+      "grad_norm": 0.0,
+      "learning_rate": 6.746786314861189e-06,
+      "loss": 0.703,
+      "step": 2001
+    },
+    {
+      "epoch": 0.6176153015579208,
+      "grad_norm": 0.0,
+      "learning_rate": 6.737336140582291e-06,
+      "loss": 0.7275,
+      "step": 2002
+    },
+    {
+      "epoch": 0.617923800709548,
+      "grad_norm": 0.0,
+      "learning_rate": 6.72788922604578e-06,
+      "loss": 0.7693,
+      "step": 2003
+    },
+    {
+      "epoch": 0.6182322998611753,
+      "grad_norm": 0.0,
+      "learning_rate": 6.718445580690113e-06,
+      "loss": 0.7452,
+      "step": 2004
+    },
+    {
+      "epoch": 0.6185407990128027,
+      "grad_norm": 0.0,
+      "learning_rate": 6.709005213950472e-06,
+      "loss": 0.7144,
+      "step": 2005
+    },
+    {
+      "epoch": 0.61884929816443,
+      "grad_norm": 0.0,
+      "learning_rate": 6.699568135258774e-06,
+      "loss": 0.6836,
+      "step": 2006
+    },
+    {
+      "epoch": 0.6191577973160574,
+      "grad_norm": 0.0,
+      "learning_rate": 6.690134354043649e-06,
+      "loss": 0.7265,
+      "step": 2007
+    },
+    {
+      "epoch": 0.6194662964676847,
+      "grad_norm": 0.0,
+      "learning_rate": 6.68070387973043e-06,
+      "loss": 0.6829,
+      "step": 2008
+    },
+    {
+      "epoch": 0.619774795619312,
+      "grad_norm": 0.0,
+      "learning_rate": 6.671276721741149e-06,
+      "loss": 0.7093,
+      "step": 2009
+    },
+    {
+      "epoch": 0.6200832947709394,
+      "grad_norm": 0.0,
+      "learning_rate": 6.6618528894945175e-06,
+      "loss": 0.7282,
+      "step": 2010
+    },
+    {
+      "epoch": 0.6203917939225667,
+      "grad_norm": 0.0,
+      "learning_rate": 6.652432392405934e-06,
+      "loss": 0.7671,
+      "step": 2011
+    },
+    {
+      "epoch": 0.620700293074194,
+      "grad_norm": 0.0,
+      "learning_rate": 6.643015239887458e-06,
+      "loss": 0.7393,
+      "step": 2012
+    },
+    {
+      "epoch": 0.6210087922258214,
+      "grad_norm": 0.0,
+      "learning_rate": 6.633601441347812e-06,
+      "loss": 0.6363,
+      "step": 2013
+    },
+    {
+      "epoch": 0.6213172913774487,
+      "grad_norm": 0.0,
+      "learning_rate": 6.624191006192363e-06,
+      "loss": 0.7655,
+      "step": 2014
+    },
+    {
+      "epoch": 0.6216257905290761,
+      "grad_norm": 0.0,
+      "learning_rate": 6.61478394382312e-06,
+      "loss": 0.6773,
+      "step": 2015
+    },
+    {
+      "epoch": 0.6219342896807034,
+      "grad_norm": 0.0,
+      "learning_rate": 6.605380263638722e-06,
+      "loss": 0.6704,
+      "step": 2016
+    },
+    {
+      "epoch": 0.6222427888323308,
+      "grad_norm": 0.0,
+      "learning_rate": 6.595979975034434e-06,
+      "loss": 0.7343,
+      "step": 2017
+    },
+    {
+      "epoch": 0.6225512879839581,
+      "grad_norm": 0.0,
+      "learning_rate": 6.586583087402119e-06,
+      "loss": 0.7653,
+      "step": 2018
+    },
+    {
+      "epoch": 0.6228597871355854,
+      "grad_norm": 0.0,
+      "learning_rate": 6.577189610130254e-06,
+      "loss": 0.7775,
+      "step": 2019
+    },
+    {
+      "epoch": 0.6231682862872127,
+      "grad_norm": 0.0,
+      "learning_rate": 6.567799552603904e-06,
+      "loss": 0.7751,
+      "step": 2020
+    },
+    {
+      "epoch": 0.62347678543884,
+      "grad_norm": 0.0,
+      "learning_rate": 6.558412924204722e-06,
+      "loss": 0.676,
+      "step": 2021
+    },
+    {
+      "epoch": 0.6237852845904673,
+      "grad_norm": 0.0,
+      "learning_rate": 6.549029734310928e-06,
+      "loss": 0.6836,
+      "step": 2022
+    },
+    {
+      "epoch": 0.6240937837420947,
+      "grad_norm": 0.0,
+      "learning_rate": 6.539649992297311e-06,
+      "loss": 0.7058,
+      "step": 2023
+    },
+    {
+      "epoch": 0.624402282893722,
+      "grad_norm": 0.0,
+      "learning_rate": 6.530273707535214e-06,
+      "loss": 0.724,
+      "step": 2024
+    },
+    {
+      "epoch": 0.6247107820453494,
+      "grad_norm": 0.0,
+      "learning_rate": 6.520900889392525e-06,
+      "loss": 0.7247,
+      "step": 2025
+    },
+    {
+      "epoch": 0.6250192811969767,
+      "grad_norm": 0.0,
+      "learning_rate": 6.511531547233674e-06,
+      "loss": 0.761,
+      "step": 2026
+    },
+    {
+      "epoch": 0.625327780348604,
+      "grad_norm": 0.0,
+      "learning_rate": 6.502165690419608e-06,
+      "loss": 0.801,
+      "step": 2027
+    },
+    {
+      "epoch": 0.6256362795002314,
+      "grad_norm": 0.0,
+      "learning_rate": 6.492803328307799e-06,
+      "loss": 0.726,
+      "step": 2028
+    },
+    {
+      "epoch": 0.6259447786518587,
+      "grad_norm": 0.0,
+      "learning_rate": 6.483444470252227e-06,
+      "loss": 0.7648,
+      "step": 2029
+    },
+    {
+      "epoch": 0.626253277803486,
+      "grad_norm": 0.0,
+      "learning_rate": 6.4740891256033736e-06,
+      "loss": 0.7044,
+      "step": 2030
+    },
+    {
+      "epoch": 0.6265617769551134,
+      "grad_norm": 0.0,
+      "learning_rate": 6.464737303708197e-06,
+      "loss": 0.7878,
+      "step": 2031
+    },
+    {
+      "epoch": 0.6268702761067407,
+      "grad_norm": 0.0,
+      "learning_rate": 6.455389013910151e-06,
+      "loss": 0.6916,
+      "step": 2032
+    },
+    {
+      "epoch": 0.6271787752583681,
+      "grad_norm": 0.0,
+      "learning_rate": 6.4460442655491515e-06,
+      "loss": 0.6691,
+      "step": 2033
+    },
+    {
+      "epoch": 0.6274872744099954,
+      "grad_norm": 0.0,
+      "learning_rate": 6.436703067961589e-06,
+      "loss": 0.6556,
+      "step": 2034
+    },
+    {
+      "epoch": 0.6277957735616227,
+      "grad_norm": 0.0,
+      "learning_rate": 6.4273654304802844e-06,
+      "loss": 0.7585,
+      "step": 2035
+    },
+    {
+      "epoch": 0.6281042727132501,
+      "grad_norm": 0.0,
+      "learning_rate": 6.4180313624345205e-06,
+      "loss": 0.7429,
+      "step": 2036
+    },
+    {
+      "epoch": 0.6284127718648773,
+      "grad_norm": 0.0,
+      "learning_rate": 6.408700873150005e-06,
+      "loss": 0.6842,
+      "step": 2037
+    },
+    {
+      "epoch": 0.6287212710165047,
+      "grad_norm": 0.0,
+      "learning_rate": 6.399373971948877e-06,
+      "loss": 0.7534,
+      "step": 2038
+    },
+    {
+      "epoch": 0.629029770168132,
+      "grad_norm": 0.0,
+      "learning_rate": 6.3900506681496786e-06,
+      "loss": 0.7038,
+      "step": 2039
+    },
+    {
+      "epoch": 0.6293382693197593,
+      "grad_norm": 0.0,
+      "learning_rate": 6.38073097106737e-06,
+      "loss": 0.6684,
+      "step": 2040
+    },
+    {
+      "epoch": 0.6296467684713867,
+      "grad_norm": 0.0,
+      "learning_rate": 6.371414890013304e-06,
+      "loss": 0.6384,
+      "step": 2041
+    },
+    {
+      "epoch": 0.629955267623014,
+      "grad_norm": 0.0,
+      "learning_rate": 6.362102434295216e-06,
+      "loss": 0.816,
+      "step": 2042
+    },
+    {
+      "epoch": 0.6302637667746414,
+      "grad_norm": 0.0,
+      "learning_rate": 6.352793613217232e-06,
+      "loss": 0.6687,
+      "step": 2043
+    },
+    {
+      "epoch": 0.6305722659262687,
+      "grad_norm": 0.0,
+      "learning_rate": 6.3434884360798255e-06,
+      "loss": 0.7027,
+      "step": 2044
+    },
+    {
+      "epoch": 0.630880765077896,
+      "grad_norm": 0.0,
+      "learning_rate": 6.334186912179845e-06,
+      "loss": 0.7615,
+      "step": 2045
+    },
+    {
+      "epoch": 0.6311892642295234,
+      "grad_norm": 0.0,
+      "learning_rate": 6.3248890508104895e-06,
+      "loss": 0.6306,
+      "step": 2046
+    },
+    {
+      "epoch": 0.6314977633811507,
+      "grad_norm": 0.0,
+      "learning_rate": 6.315594861261299e-06,
+      "loss": 0.8275,
+      "step": 2047
+    },
+    {
+      "epoch": 0.631806262532778,
+      "grad_norm": 0.0,
+      "learning_rate": 6.3063043528181286e-06,
+      "loss": 0.7133,
+      "step": 2048
+    },
+    {
+      "epoch": 0.6321147616844054,
+      "grad_norm": 0.0,
+      "learning_rate": 6.297017534763175e-06,
+      "loss": 0.7637,
+      "step": 2049
+    },
+    {
+      "epoch": 0.6324232608360327,
+      "grad_norm": 0.0,
+      "learning_rate": 6.28773441637494e-06,
+      "loss": 1.0219,
+      "step": 2050
+    },
+    {
+      "epoch": 0.6327317599876601,
+      "grad_norm": 0.0,
+      "learning_rate": 6.278455006928233e-06,
+      "loss": 0.6952,
+      "step": 2051
+    },
+    {
+      "epoch": 0.6330402591392874,
+      "grad_norm": 0.0,
+      "learning_rate": 6.269179315694145e-06,
+      "loss": 0.6483,
+      "step": 2052
+    },
+    {
+      "epoch": 0.6333487582909147,
+      "grad_norm": 0.0,
+      "learning_rate": 6.259907351940069e-06,
+      "loss": 0.614,
+      "step": 2053
+    },
+    {
+      "epoch": 0.6336572574425421,
+      "grad_norm": 0.0,
+      "learning_rate": 6.250639124929665e-06,
+      "loss": 0.7949,
+      "step": 2054
+    },
+    {
+      "epoch": 0.6339657565941693,
+      "grad_norm": 0.0,
+      "learning_rate": 6.241374643922864e-06,
+      "loss": 0.6852,
+      "step": 2055
+    },
+    {
+      "epoch": 0.6342742557457967,
+      "grad_norm": 0.0,
+      "learning_rate": 6.232113918175845e-06,
+      "loss": 0.7307,
+      "step": 2056
+    },
+    {
+      "epoch": 0.634582754897424,
+      "grad_norm": 0.0,
+      "learning_rate": 6.222856956941041e-06,
+      "loss": 0.7534,
+      "step": 2057
+    },
+    {
+      "epoch": 0.6348912540490513,
+      "grad_norm": 0.0,
+      "learning_rate": 6.213603769467132e-06,
+      "loss": 0.6979,
+      "step": 2058
+    },
+    {
+      "epoch": 0.6351997532006787,
+      "grad_norm": 0.0,
+      "learning_rate": 6.204354364999014e-06,
+      "loss": 0.7602,
+      "step": 2059
+    },
+    {
+      "epoch": 0.635508252352306,
+      "grad_norm": 0.0,
+      "learning_rate": 6.195108752777814e-06,
+      "loss": 0.7576,
+      "step": 2060
+    },
+    {
+      "epoch": 0.6358167515039334,
+      "grad_norm": 0.0,
+      "learning_rate": 6.185866942040861e-06,
+      "loss": 0.7792,
+      "step": 2061
+    },
+    {
+      "epoch": 0.6361252506555607,
+      "grad_norm": 0.0,
+      "learning_rate": 6.17662894202169e-06,
+      "loss": 0.7114,
+      "step": 2062
+    },
+    {
+      "epoch": 0.636433749807188,
+      "grad_norm": 0.0,
+      "learning_rate": 6.167394761950032e-06,
+      "loss": 0.8525,
+      "step": 2063
+    },
+    {
+      "epoch": 0.6367422489588154,
+      "grad_norm": 0.0,
+      "learning_rate": 6.158164411051799e-06,
+      "loss": 0.6502,
+      "step": 2064
+    },
+    {
+      "epoch": 0.6370507481104427,
+      "grad_norm": 0.0,
+      "learning_rate": 6.148937898549072e-06,
+      "loss": 0.7319,
+      "step": 2065
+    },
+    {
+      "epoch": 0.63735924726207,
+      "grad_norm": 0.0,
+      "learning_rate": 6.139715233660106e-06,
+      "loss": 0.7326,
+      "step": 2066
+    },
+    {
+      "epoch": 0.6376677464136974,
+      "grad_norm": 0.0,
+      "learning_rate": 6.130496425599308e-06,
+      "loss": 0.7331,
+      "step": 2067
+    },
+    {
+      "epoch": 0.6379762455653247,
+      "grad_norm": 0.0,
+      "learning_rate": 6.121281483577233e-06,
+      "loss": 0.7268,
+      "step": 2068
+    },
+    {
+      "epoch": 0.6382847447169521,
+      "grad_norm": 0.0,
+      "learning_rate": 6.112070416800562e-06,
+      "loss": 0.7058,
+      "step": 2069
+    },
+    {
+      "epoch": 0.6385932438685794,
+      "grad_norm": 0.0,
+      "learning_rate": 6.102863234472124e-06,
+      "loss": 0.6867,
+      "step": 2070
+    },
+    {
+      "epoch": 0.6389017430202067,
+      "grad_norm": 0.0,
+      "learning_rate": 6.093659945790853e-06,
+      "loss": 0.7267,
+      "step": 2071
+    },
+    {
+      "epoch": 0.639210242171834,
+      "grad_norm": 0.0,
+      "learning_rate": 6.084460559951802e-06,
+      "loss": 0.6444,
+      "step": 2072
+    },
+    {
+      "epoch": 0.6395187413234613,
+      "grad_norm": 0.0,
+      "learning_rate": 6.075265086146111e-06,
+      "loss": 1.0321,
+      "step": 2073
+    },
+    {
+      "epoch": 0.6398272404750887,
+      "grad_norm": 0.0,
+      "learning_rate": 6.066073533561024e-06,
+      "loss": 0.6793,
+      "step": 2074
+    },
+    {
+      "epoch": 0.640135739626716,
+      "grad_norm": 0.0,
+      "learning_rate": 6.056885911379863e-06,
+      "loss": 0.6712,
+      "step": 2075
+    },
+    {
+      "epoch": 0.6404442387783433,
+      "grad_norm": 0.0,
+      "learning_rate": 6.047702228782023e-06,
+      "loss": 0.7343,
+      "step": 2076
+    },
+    {
+      "epoch": 0.6407527379299707,
+      "grad_norm": 0.0,
+      "learning_rate": 6.0385224949429666e-06,
+      "loss": 0.6725,
+      "step": 2077
+    },
+    {
+      "epoch": 0.641061237081598,
+      "grad_norm": 0.0,
+      "learning_rate": 6.029346719034203e-06,
+      "loss": 0.6364,
+      "step": 2078
+    },
+    {
+      "epoch": 0.6413697362332254,
+      "grad_norm": 0.0,
+      "learning_rate": 6.020174910223293e-06,
+      "loss": 0.7213,
+      "step": 2079
+    },
+    {
+      "epoch": 0.6416782353848527,
+      "grad_norm": 0.0,
+      "learning_rate": 6.011007077673835e-06,
+      "loss": 0.6837,
+      "step": 2080
+    },
+    {
+      "epoch": 0.64198673453648,
+      "grad_norm": 0.0,
+      "learning_rate": 6.001843230545452e-06,
+      "loss": 0.669,
+      "step": 2081
+    },
+    {
+      "epoch": 0.6422952336881074,
+      "grad_norm": 0.0,
+      "learning_rate": 5.992683377993784e-06,
+      "loss": 0.6777,
+      "step": 2082
+    },
+    {
+      "epoch": 0.6426037328397347,
+      "grad_norm": 0.0,
+      "learning_rate": 5.983527529170481e-06,
+      "loss": 0.7283,
+      "step": 2083
+    },
+    {
+      "epoch": 0.642912231991362,
+      "grad_norm": 0.0,
+      "learning_rate": 5.9743756932231955e-06,
+      "loss": 0.7848,
+      "step": 2084
+    },
+    {
+      "epoch": 0.6432207311429894,
+      "grad_norm": 0.0,
+      "learning_rate": 5.965227879295572e-06,
+      "loss": 0.7034,
+      "step": 2085
+    },
+    {
+      "epoch": 0.6435292302946167,
+      "grad_norm": 0.0,
+      "learning_rate": 5.956084096527224e-06,
+      "loss": 0.7339,
+      "step": 2086
+    },
+    {
+      "epoch": 0.6438377294462441,
+      "grad_norm": 0.0,
+      "learning_rate": 5.946944354053753e-06,
+      "loss": 0.6728,
+      "step": 2087
+    },
+    {
+      "epoch": 0.6441462285978714,
+      "grad_norm": 0.0,
+      "learning_rate": 5.937808661006715e-06,
+      "loss": 0.7308,
+      "step": 2088
+    },
+    {
+      "epoch": 0.6444547277494986,
+      "grad_norm": 0.0,
+      "learning_rate": 5.928677026513627e-06,
+      "loss": 0.6657,
+      "step": 2089
+    },
+    {
+      "epoch": 0.644763226901126,
+      "grad_norm": 0.0,
+      "learning_rate": 5.919549459697942e-06,
+      "loss": 0.7471,
+      "step": 2090
+    },
+    {
+      "epoch": 0.6450717260527533,
+      "grad_norm": 0.0,
+      "learning_rate": 5.910425969679056e-06,
+      "loss": 0.7049,
+      "step": 2091
+    },
+    {
+      "epoch": 0.6453802252043807,
+      "grad_norm": 0.0,
+      "learning_rate": 5.901306565572288e-06,
+      "loss": 0.6928,
+      "step": 2092
+    },
+    {
+      "epoch": 0.645688724356008,
+      "grad_norm": 0.0,
+      "learning_rate": 5.8921912564888775e-06,
+      "loss": 0.758,
+      "step": 2093
+    },
+    {
+      "epoch": 0.6459972235076353,
+      "grad_norm": 0.0,
+      "learning_rate": 5.883080051535974e-06,
+      "loss": 0.6874,
+      "step": 2094
+    },
+    {
+      "epoch": 0.6463057226592627,
+      "grad_norm": 0.0,
+      "learning_rate": 5.873972959816619e-06,
+      "loss": 0.7153,
+      "step": 2095
+    },
+    {
+      "epoch": 0.64661422181089,
+      "grad_norm": 0.0,
+      "learning_rate": 5.864869990429753e-06,
+      "loss": 0.6867,
+      "step": 2096
+    },
+    {
+      "epoch": 0.6469227209625174,
+      "grad_norm": 0.0,
+      "learning_rate": 5.855771152470193e-06,
+      "loss": 0.6989,
+      "step": 2097
+    },
+    {
+      "epoch": 0.6472312201141447,
+      "grad_norm": 0.0,
+      "learning_rate": 5.846676455028635e-06,
+      "loss": 0.7185,
+      "step": 2098
+    },
+    {
+      "epoch": 0.647539719265772,
+      "grad_norm": 0.0,
+      "learning_rate": 5.83758590719162e-06,
+      "loss": 0.7486,
+      "step": 2099
+    },
+    {
+      "epoch": 0.6478482184173994,
+      "grad_norm": 0.0,
+      "learning_rate": 5.8284995180415685e-06,
+      "loss": 0.6952,
+      "step": 2100
+    },
+    {
+      "epoch": 0.6481567175690267,
+      "grad_norm": 0.0,
+      "learning_rate": 5.819417296656724e-06,
+      "loss": 0.7227,
+      "step": 2101
+    },
+    {
+      "epoch": 0.648465216720654,
+      "grad_norm": 0.0,
+      "learning_rate": 5.81033925211119e-06,
+      "loss": 0.7285,
+      "step": 2102
+    },
+    {
+      "epoch": 0.6487737158722814,
+      "grad_norm": 0.0,
+      "learning_rate": 5.8012653934748644e-06,
+      "loss": 0.7128,
+      "step": 2103
+    },
+    {
+      "epoch": 0.6490822150239087,
+      "grad_norm": 0.0,
+      "learning_rate": 5.7921957298134865e-06,
+      "loss": 0.6555,
+      "step": 2104
+    },
+    {
+      "epoch": 0.6493907141755361,
+      "grad_norm": 0.0,
+      "learning_rate": 5.783130270188607e-06,
+      "loss": 0.7154,
+      "step": 2105
+    },
+    {
+      "epoch": 0.6496992133271634,
+      "grad_norm": 0.0,
+      "learning_rate": 5.774069023657558e-06,
+      "loss": 0.7147,
+      "step": 2106
+    },
+    {
+      "epoch": 0.6500077124787906,
+      "grad_norm": 0.0,
+      "learning_rate": 5.765011999273484e-06,
+      "loss": 0.7412,
+      "step": 2107
+    },
+    {
+      "epoch": 0.650316211630418,
+      "grad_norm": 0.0,
+      "learning_rate": 5.755959206085285e-06,
+      "loss": 0.7176,
+      "step": 2108
+    },
+    {
+      "epoch": 0.6506247107820453,
+      "grad_norm": 0.0,
+      "learning_rate": 5.746910653137659e-06,
+      "loss": 0.6684,
+      "step": 2109
+    },
+    {
+      "epoch": 0.6509332099336727,
+      "grad_norm": 0.0,
+      "learning_rate": 5.73786634947105e-06,
+      "loss": 0.717,
+      "step": 2110
+    },
+    {
+      "epoch": 0.6512417090853,
+      "grad_norm": 0.0,
+      "learning_rate": 5.7288263041216685e-06,
+      "loss": 0.7727,
+      "step": 2111
+    },
+    {
+      "epoch": 0.6515502082369273,
+      "grad_norm": 0.0,
+      "learning_rate": 5.719790526121462e-06,
+      "loss": 0.7078,
+      "step": 2112
+    },
+    {
+      "epoch": 0.6518587073885547,
+      "grad_norm": 0.0,
+      "learning_rate": 5.7107590244981156e-06,
+      "loss": 0.7385,
+      "step": 2113
+    },
+    {
+      "epoch": 0.652167206540182,
+      "grad_norm": 0.0,
+      "learning_rate": 5.70173180827505e-06,
+      "loss": 0.7426,
+      "step": 2114
+    },
+    {
+      "epoch": 0.6524757056918093,
+      "grad_norm": 0.0,
+      "learning_rate": 5.692708886471395e-06,
+      "loss": 0.6916,
+      "step": 2115
+    },
+    {
+      "epoch": 0.6527842048434367,
+      "grad_norm": 0.0,
+      "learning_rate": 5.683690268101989e-06,
+      "loss": 0.6724,
+      "step": 2116
+    },
+    {
+      "epoch": 0.653092703995064,
+      "grad_norm": 0.0,
+      "learning_rate": 5.674675962177383e-06,
+      "loss": 0.6995,
+      "step": 2117
+    },
+    {
+      "epoch": 0.6534012031466914,
+      "grad_norm": 0.0,
+      "learning_rate": 5.665665977703803e-06,
+      "loss": 0.6714,
+      "step": 2118
+    },
+    {
+      "epoch": 0.6537097022983187,
+      "grad_norm": 0.0,
+      "learning_rate": 5.656660323683177e-06,
+      "loss": 0.6629,
+      "step": 2119
+    },
+    {
+      "epoch": 0.654018201449946,
+      "grad_norm": 0.0,
+      "learning_rate": 5.647659009113079e-06,
+      "loss": 0.6691,
+      "step": 2120
+    },
+    {
+      "epoch": 0.6543267006015734,
+      "grad_norm": 0.0,
+      "learning_rate": 5.638662042986777e-06,
+      "loss": 0.6801,
+      "step": 2121
+    },
+    {
+      "epoch": 0.6546351997532007,
+      "grad_norm": 0.0,
+      "learning_rate": 5.629669434293172e-06,
+      "loss": 0.746,
+      "step": 2122
+    },
+    {
+      "epoch": 0.6549436989048281,
+      "grad_norm": 0.0,
+      "learning_rate": 5.62068119201683e-06,
+      "loss": 0.7055,
+      "step": 2123
+    },
+    {
+      "epoch": 0.6552521980564553,
+      "grad_norm": 0.0,
+      "learning_rate": 5.611697325137939e-06,
+      "loss": 0.7346,
+      "step": 2124
+    },
+    {
+      "epoch": 0.6555606972080826,
+      "grad_norm": 0.0,
+      "learning_rate": 5.602717842632319e-06,
+      "loss": 0.6536,
+      "step": 2125
+    },
+    {
+      "epoch": 0.65586919635971,
+      "grad_norm": 0.0,
+      "learning_rate": 5.5937427534714195e-06,
+      "loss": 0.6763,
+      "step": 2126
+    },
+    {
+      "epoch": 0.6561776955113373,
+      "grad_norm": 0.0,
+      "learning_rate": 5.584772066622284e-06,
+      "loss": 0.7353,
+      "step": 2127
+    },
+    {
+      "epoch": 0.6564861946629647,
+      "grad_norm": 0.0,
+      "learning_rate": 5.575805791047577e-06,
+      "loss": 0.7712,
+      "step": 2128
+    },
+    {
+      "epoch": 0.656794693814592,
+      "grad_norm": 0.0,
+      "learning_rate": 5.566843935705539e-06,
+      "loss": 0.754,
+      "step": 2129
+    },
+    {
+      "epoch": 0.6571031929662193,
+      "grad_norm": 0.0,
+      "learning_rate": 5.557886509549998e-06,
+      "loss": 0.7469,
+      "step": 2130
+    },
+    {
+      "epoch": 0.6574116921178467,
+      "grad_norm": 0.0,
+      "learning_rate": 5.5489335215303674e-06,
+      "loss": 0.6798,
+      "step": 2131
+    },
+    {
+      "epoch": 0.657720191269474,
+      "grad_norm": 0.0,
+      "learning_rate": 5.539984980591615e-06,
+      "loss": 0.6893,
+      "step": 2132
+    },
+    {
+      "epoch": 0.6580286904211013,
+      "grad_norm": 0.0,
+      "learning_rate": 5.531040895674267e-06,
+      "loss": 0.7365,
+      "step": 2133
+    },
+    {
+      "epoch": 0.6583371895727287,
+      "grad_norm": 0.0,
+      "learning_rate": 5.5221012757143974e-06,
+      "loss": 0.6681,
+      "step": 2134
+    },
+    {
+      "epoch": 0.658645688724356,
+      "grad_norm": 0.0,
+      "learning_rate": 5.51316612964363e-06,
+      "loss": 0.7298,
+      "step": 2135
+    },
+    {
+      "epoch": 0.6589541878759834,
+      "grad_norm": 0.0,
+      "learning_rate": 5.504235466389103e-06,
+      "loss": 0.6818,
+      "step": 2136
+    },
+    {
+      "epoch": 0.6592626870276107,
+      "grad_norm": 0.0,
+      "learning_rate": 5.495309294873483e-06,
+      "loss": 0.7411,
+      "step": 2137
+    },
+    {
+      "epoch": 0.659571186179238,
+      "grad_norm": 0.0,
+      "learning_rate": 5.486387624014952e-06,
+      "loss": 0.734,
+      "step": 2138
+    },
+    {
+      "epoch": 0.6598796853308654,
+      "grad_norm": 0.0,
+      "learning_rate": 5.47747046272719e-06,
+      "loss": 0.7049,
+      "step": 2139
+    },
+    {
+      "epoch": 0.6601881844824927,
+      "grad_norm": 0.0,
+      "learning_rate": 5.468557819919378e-06,
+      "loss": 0.7069,
+      "step": 2140
+    },
+    {
+      "epoch": 0.66049668363412,
+      "grad_norm": 0.0,
+      "learning_rate": 5.4596497044961725e-06,
+      "loss": 0.7255,
+      "step": 2141
+    },
+    {
+      "epoch": 0.6608051827857473,
+      "grad_norm": 0.0,
+      "learning_rate": 5.450746125357712e-06,
+      "loss": 0.7469,
+      "step": 2142
+    },
+    {
+      "epoch": 0.6611136819373746,
+      "grad_norm": 0.0,
+      "learning_rate": 5.44184709139961e-06,
+      "loss": 0.7603,
+      "step": 2143
+    },
+    {
+      "epoch": 0.661422181089002,
+      "grad_norm": 0.0,
+      "learning_rate": 5.432952611512923e-06,
+      "loss": 0.8311,
+      "step": 2144
+    },
+    {
+      "epoch": 0.6617306802406293,
+      "grad_norm": 0.0,
+      "learning_rate": 5.424062694584179e-06,
+      "loss": 0.6956,
+      "step": 2145
+    },
+    {
+      "epoch": 0.6620391793922566,
+      "grad_norm": 0.0,
+      "learning_rate": 5.41517734949532e-06,
+      "loss": 0.7034,
+      "step": 2146
+    },
+    {
+      "epoch": 0.662347678543884,
+      "grad_norm": 0.0,
+      "learning_rate": 5.406296585123745e-06,
+      "loss": 0.7404,
+      "step": 2147
+    },
+    {
+      "epoch": 0.6626561776955113,
+      "grad_norm": 0.0,
+      "learning_rate": 5.397420410342259e-06,
+      "loss": 0.7452,
+      "step": 2148
+    },
+    {
+      "epoch": 0.6629646768471387,
+      "grad_norm": 0.0,
+      "learning_rate": 5.388548834019097e-06,
+      "loss": 0.629,
+      "step": 2149
+    },
+    {
+      "epoch": 0.663273175998766,
+      "grad_norm": 0.0,
+      "learning_rate": 5.379681865017887e-06,
+      "loss": 0.7724,
+      "step": 2150
+    },
+    {
+      "epoch": 0.6635816751503933,
+      "grad_norm": 0.0,
+      "learning_rate": 5.370819512197656e-06,
+      "loss": 0.7365,
+      "step": 2151
+    },
+    {
+      "epoch": 0.6638901743020207,
+      "grad_norm": 0.0,
+      "learning_rate": 5.361961784412828e-06,
+      "loss": 0.6681,
+      "step": 2152
+    },
+    {
+      "epoch": 0.664198673453648,
+      "grad_norm": 0.0,
+      "learning_rate": 5.353108690513193e-06,
+      "loss": 0.7178,
+      "step": 2153
+    },
+    {
+      "epoch": 0.6645071726052754,
+      "grad_norm": 0.0,
+      "learning_rate": 5.344260239343919e-06,
+      "loss": 0.7258,
+      "step": 2154
+    },
+    {
+      "epoch": 0.6648156717569027,
+      "grad_norm": 0.0,
+      "learning_rate": 5.335416439745538e-06,
+      "loss": 0.7827,
+      "step": 2155
+    },
+    {
+      "epoch": 0.66512417090853,
+      "grad_norm": 0.0,
+      "learning_rate": 5.326577300553923e-06,
+      "loss": 0.757,
+      "step": 2156
+    },
+    {
+      "epoch": 0.6654326700601574,
+      "grad_norm": 0.0,
+      "learning_rate": 5.317742830600306e-06,
+      "loss": 0.824,
+      "step": 2157
+    },
+    {
+      "epoch": 0.6657411692117847,
+      "grad_norm": 0.0,
+      "learning_rate": 5.308913038711245e-06,
+      "loss": 0.7818,
+      "step": 2158
+    },
+    {
+      "epoch": 0.666049668363412,
+      "grad_norm": 0.0,
+      "learning_rate": 5.300087933708624e-06,
+      "loss": 0.7217,
+      "step": 2159
+    },
+    {
+      "epoch": 0.6663581675150393,
+      "grad_norm": 0.0,
+      "learning_rate": 5.29126752440964e-06,
+      "loss": 0.7242,
+      "step": 2160
+    },
+    {
+      "epoch": 0.6666666666666666,
+      "grad_norm": 0.0,
+      "learning_rate": 5.282451819626815e-06,
+      "loss": 0.7298,
+      "step": 2161
+    },
+    {
+      "epoch": 0.666975165818294,
+      "grad_norm": 0.0,
+      "learning_rate": 5.273640828167954e-06,
+      "loss": 0.7062,
+      "step": 2162
+    },
+    {
+      "epoch": 0.6672836649699213,
+      "grad_norm": 0.0,
+      "learning_rate": 5.264834558836156e-06,
+      "loss": 0.7027,
+      "step": 2163
+    },
+    {
+      "epoch": 0.6675921641215486,
+      "grad_norm": 0.0,
+      "learning_rate": 5.256033020429813e-06,
+      "loss": 0.7512,
+      "step": 2164
+    },
+    {
+      "epoch": 0.667900663273176,
+      "grad_norm": 0.0,
+      "learning_rate": 5.247236221742575e-06,
+      "loss": 0.7286,
+      "step": 2165
+    },
+    {
+      "epoch": 0.6682091624248033,
+      "grad_norm": 0.0,
+      "learning_rate": 5.238444171563368e-06,
+      "loss": 0.7256,
+      "step": 2166
+    },
+    {
+      "epoch": 0.6685176615764307,
+      "grad_norm": 0.0,
+      "learning_rate": 5.22965687867637e-06,
+      "loss": 0.744,
+      "step": 2167
+    },
+    {
+      "epoch": 0.668826160728058,
+      "grad_norm": 0.0,
+      "learning_rate": 5.220874351861001e-06,
+      "loss": 0.7358,
+      "step": 2168
+    },
+    {
+      "epoch": 0.6691346598796853,
+      "grad_norm": 0.0,
+      "learning_rate": 5.212096599891927e-06,
+      "loss": 0.7285,
+      "step": 2169
+    },
+    {
+      "epoch": 0.6694431590313127,
+      "grad_norm": 0.0,
+      "learning_rate": 5.203323631539042e-06,
+      "loss": 0.6504,
+      "step": 2170
+    },
+    {
+      "epoch": 0.66975165818294,
+      "grad_norm": 0.0,
+      "learning_rate": 5.194555455567456e-06,
+      "loss": 0.8199,
+      "step": 2171
+    },
+    {
+      "epoch": 0.6700601573345674,
+      "grad_norm": 0.0,
+      "learning_rate": 5.185792080737491e-06,
+      "loss": 0.7196,
+      "step": 2172
+    },
+    {
+      "epoch": 0.6703686564861947,
+      "grad_norm": 0.0,
+      "learning_rate": 5.177033515804682e-06,
+      "loss": 0.751,
+      "step": 2173
+    },
+    {
+      "epoch": 0.670677155637822,
+      "grad_norm": 0.0,
+      "learning_rate": 5.168279769519742e-06,
+      "loss": 0.771,
+      "step": 2174
+    },
+    {
+      "epoch": 0.6709856547894494,
+      "grad_norm": 0.0,
+      "learning_rate": 5.159530850628589e-06,
+      "loss": 0.7746,
+      "step": 2175
+    },
+    {
+      "epoch": 0.6712941539410766,
+      "grad_norm": 0.0,
+      "learning_rate": 5.150786767872302e-06,
+      "loss": 0.7745,
+      "step": 2176
+    },
+    {
+      "epoch": 0.671602653092704,
+      "grad_norm": 0.0,
+      "learning_rate": 5.142047529987133e-06,
+      "loss": 0.6888,
+      "step": 2177
+    },
+    {
+      "epoch": 0.6719111522443313,
+      "grad_norm": 0.0,
+      "learning_rate": 5.1333131457044995e-06,
+      "loss": 0.7485,
+      "step": 2178
+    },
+    {
+      "epoch": 0.6722196513959586,
+      "grad_norm": 0.0,
+      "learning_rate": 5.124583623750963e-06,
+      "loss": 0.7505,
+      "step": 2179
+    },
+    {
+      "epoch": 0.672528150547586,
+      "grad_norm": 0.0,
+      "learning_rate": 5.115858972848224e-06,
+      "loss": 0.6436,
+      "step": 2180
+    },
+    {
+      "epoch": 0.6728366496992133,
+      "grad_norm": 0.0,
+      "learning_rate": 5.107139201713128e-06,
+      "loss": 0.7271,
+      "step": 2181
+    },
+    {
+      "epoch": 0.6731451488508406,
+      "grad_norm": 0.0,
+      "learning_rate": 5.098424319057632e-06,
+      "loss": 0.7306,
+      "step": 2182
+    },
+    {
+      "epoch": 0.673453648002468,
+      "grad_norm": 0.0,
+      "learning_rate": 5.089714333588827e-06,
+      "loss": 1.0136,
+      "step": 2183
+    },
+    {
+      "epoch": 0.6737621471540953,
+      "grad_norm": 0.0,
+      "learning_rate": 5.081009254008882e-06,
+      "loss": 0.7328,
+      "step": 2184
+    },
+    {
+      "epoch": 0.6740706463057227,
+      "grad_norm": 0.0,
+      "learning_rate": 5.072309089015092e-06,
+      "loss": 0.7289,
+      "step": 2185
+    },
+    {
+      "epoch": 0.67437914545735,
+      "grad_norm": 0.0,
+      "learning_rate": 5.063613847299831e-06,
+      "loss": 0.6871,
+      "step": 2186
+    },
+    {
+      "epoch": 0.6746876446089773,
+      "grad_norm": 0.0,
+      "learning_rate": 5.054923537550554e-06,
+      "loss": 0.6143,
+      "step": 2187
+    },
+    {
+      "epoch": 0.6749961437606047,
+      "grad_norm": 0.0,
+      "learning_rate": 5.046238168449791e-06,
+      "loss": 0.6871,
+      "step": 2188
+    },
+    {
+      "epoch": 0.675304642912232,
+      "grad_norm": 0.0,
+      "learning_rate": 5.037557748675128e-06,
+      "loss": 0.6316,
+      "step": 2189
+    },
+    {
+      "epoch": 0.6756131420638594,
+      "grad_norm": 0.0,
+      "learning_rate": 5.028882286899219e-06,
+      "loss": 0.7132,
+      "step": 2190
+    },
+    {
+      "epoch": 0.6759216412154867,
+      "grad_norm": 0.0,
+      "learning_rate": 5.020211791789753e-06,
+      "loss": 0.6796,
+      "step": 2191
+    },
+    {
+      "epoch": 0.676230140367114,
+      "grad_norm": 0.0,
+      "learning_rate": 5.011546272009464e-06,
+      "loss": 0.7367,
+      "step": 2192
+    },
+    {
+      "epoch": 0.6765386395187413,
+      "grad_norm": 0.0,
+      "learning_rate": 5.0028857362161144e-06,
+      "loss": 0.7244,
+      "step": 2193
+    },
+    {
+      "epoch": 0.6768471386703686,
+      "grad_norm": 0.0,
+      "learning_rate": 4.994230193062477e-06,
+      "loss": 0.6804,
+      "step": 2194
+    },
+    {
+      "epoch": 0.677155637821996,
+      "grad_norm": 0.0,
+      "learning_rate": 4.985579651196354e-06,
+      "loss": 0.7504,
+      "step": 2195
+    },
+    {
+      "epoch": 0.6774641369736233,
+      "grad_norm": 0.0,
+      "learning_rate": 4.976934119260537e-06,
+      "loss": 0.6989,
+      "step": 2196
+    },
+    {
+      "epoch": 0.6777726361252506,
+      "grad_norm": 0.0,
+      "learning_rate": 4.968293605892817e-06,
+      "loss": 0.6624,
+      "step": 2197
+    },
+    {
+      "epoch": 0.678081135276878,
+      "grad_norm": 0.0,
+      "learning_rate": 4.959658119725965e-06,
+      "loss": 0.7303,
+      "step": 2198
+    },
+    {
+      "epoch": 0.6783896344285053,
+      "grad_norm": 0.0,
+      "learning_rate": 4.951027669387741e-06,
+      "loss": 0.7036,
+      "step": 2199
+    },
+    {
+      "epoch": 0.6786981335801326,
+      "grad_norm": 0.0,
+      "learning_rate": 4.942402263500874e-06,
+      "loss": 0.7572,
+      "step": 2200
+    },
+    {
+      "epoch": 0.67900663273176,
+      "grad_norm": 0.0,
+      "learning_rate": 4.933781910683031e-06,
+      "loss": 0.7067,
+      "step": 2201
+    },
+    {
+      "epoch": 0.6793151318833873,
+      "grad_norm": 0.0,
+      "learning_rate": 4.925166619546857e-06,
+      "loss": 0.8102,
+      "step": 2202
+    },
+    {
+      "epoch": 0.6796236310350147,
+      "grad_norm": 0.0,
+      "learning_rate": 4.916556398699922e-06,
+      "loss": 0.7812,
+      "step": 2203
+    },
+    {
+      "epoch": 0.679932130186642,
+      "grad_norm": 0.0,
+      "learning_rate": 4.907951256744744e-06,
+      "loss": 0.7577,
+      "step": 2204
+    },
+    {
+      "epoch": 0.6802406293382693,
+      "grad_norm": 0.0,
+      "learning_rate": 4.899351202278756e-06,
+      "loss": 0.7994,
+      "step": 2205
+    },
+    {
+      "epoch": 0.6805491284898967,
+      "grad_norm": 0.0,
+      "learning_rate": 4.890756243894308e-06,
+      "loss": 0.698,
+      "step": 2206
+    },
+    {
+      "epoch": 0.680857627641524,
+      "grad_norm": 0.0,
+      "learning_rate": 4.88216639017867e-06,
+      "loss": 0.7232,
+      "step": 2207
+    },
+    {
+      "epoch": 0.6811661267931514,
+      "grad_norm": 0.0,
+      "learning_rate": 4.873581649713996e-06,
+      "loss": 0.6626,
+      "step": 2208
+    },
+    {
+      "epoch": 0.6814746259447787,
+      "grad_norm": 0.0,
+      "learning_rate": 4.865002031077353e-06,
+      "loss": 0.774,
+      "step": 2209
+    },
+    {
+      "epoch": 0.681783125096406,
+      "grad_norm": 0.0,
+      "learning_rate": 4.856427542840658e-06,
+      "loss": 0.7072,
+      "step": 2210
+    },
+    {
+      "epoch": 0.6820916242480333,
+      "grad_norm": 0.0,
+      "learning_rate": 4.847858193570733e-06,
+      "loss": 0.7224,
+      "step": 2211
+    },
+    {
+      "epoch": 0.6824001233996606,
+      "grad_norm": 0.0,
+      "learning_rate": 4.839293991829256e-06,
+      "loss": 0.733,
+      "step": 2212
+    },
+    {
+      "epoch": 0.682708622551288,
+      "grad_norm": 0.0,
+      "learning_rate": 4.830734946172756e-06,
+      "loss": 0.6981,
+      "step": 2213
+    },
+    {
+      "epoch": 0.6830171217029153,
+      "grad_norm": 0.0,
+      "learning_rate": 4.8221810651526154e-06,
+      "loss": 0.7701,
+      "step": 2214
+    },
+    {
+      "epoch": 0.6833256208545426,
+      "grad_norm": 0.0,
+      "learning_rate": 4.8136323573150525e-06,
+      "loss": 0.7602,
+      "step": 2215
+    },
+    {
+      "epoch": 0.68363412000617,
+      "grad_norm": 0.0,
+      "learning_rate": 4.805088831201127e-06,
+      "loss": 0.7599,
+      "step": 2216
+    },
+    {
+      "epoch": 0.6839426191577973,
+      "grad_norm": 0.0,
+      "learning_rate": 4.796550495346711e-06,
+      "loss": 0.744,
+      "step": 2217
+    },
+    {
+      "epoch": 0.6842511183094246,
+      "grad_norm": 0.0,
+      "learning_rate": 4.788017358282492e-06,
+      "loss": 0.7434,
+      "step": 2218
+    },
+    {
+      "epoch": 0.684559617461052,
+      "grad_norm": 0.0,
+      "learning_rate": 4.779489428533973e-06,
+      "loss": 0.748,
+      "step": 2219
+    },
+    {
+      "epoch": 0.6848681166126793,
+      "grad_norm": 0.0,
+      "learning_rate": 4.770966714621441e-06,
+      "loss": 0.7359,
+      "step": 2220
+    },
+    {
+      "epoch": 0.6851766157643067,
+      "grad_norm": 0.0,
+      "learning_rate": 4.762449225059985e-06,
+      "loss": 0.7254,
+      "step": 2221
+    },
+    {
+      "epoch": 0.685485114915934,
+      "grad_norm": 0.0,
+      "learning_rate": 4.753936968359465e-06,
+      "loss": 0.7747,
+      "step": 2222
+    },
+    {
+      "epoch": 0.6857936140675613,
+      "grad_norm": 0.0,
+      "learning_rate": 4.745429953024511e-06,
+      "loss": 0.7263,
+      "step": 2223
+    },
+    {
+      "epoch": 0.6861021132191887,
+      "grad_norm": 0.0,
+      "learning_rate": 4.736928187554529e-06,
+      "loss": 0.7798,
+      "step": 2224
+    },
+    {
+      "epoch": 0.686410612370816,
+      "grad_norm": 0.0,
+      "learning_rate": 4.728431680443663e-06,
+      "loss": 0.7363,
+      "step": 2225
+    },
+    {
+      "epoch": 0.6867191115224434,
+      "grad_norm": 0.0,
+      "learning_rate": 4.719940440180827e-06,
+      "loss": 0.8307,
+      "step": 2226
+    },
+    {
+      "epoch": 0.6870276106740707,
+      "grad_norm": 0.0,
+      "learning_rate": 4.711454475249638e-06,
+      "loss": 0.7031,
+      "step": 2227
+    },
+    {
+      "epoch": 0.6873361098256979,
+      "grad_norm": 0.0,
+      "learning_rate": 4.702973794128477e-06,
+      "loss": 0.7277,
+      "step": 2228
+    },
+    {
+      "epoch": 0.6876446089773253,
+      "grad_norm": 0.0,
+      "learning_rate": 4.694498405290423e-06,
+      "loss": 0.6801,
+      "step": 2229
+    },
+    {
+      "epoch": 0.6879531081289526,
+      "grad_norm": 0.0,
+      "learning_rate": 4.686028317203283e-06,
+      "loss": 0.6709,
+      "step": 2230
+    },
+    {
+      "epoch": 0.68826160728058,
+      "grad_norm": 0.0,
+      "learning_rate": 4.6775635383295555e-06,
+      "loss": 0.8133,
+      "step": 2231
+    },
+    {
+      "epoch": 0.6885701064322073,
+      "grad_norm": 0.0,
+      "learning_rate": 4.669104077126439e-06,
+      "loss": 0.6551,
+      "step": 2232
+    },
+    {
+      "epoch": 0.6888786055838346,
+      "grad_norm": 0.0,
+      "learning_rate": 4.660649942045826e-06,
+      "loss": 0.7909,
+      "step": 2233
+    },
+    {
+      "epoch": 0.689187104735462,
+      "grad_norm": 0.0,
+      "learning_rate": 4.652201141534279e-06,
+      "loss": 0.6711,
+      "step": 2234
+    },
+    {
+      "epoch": 0.6894956038870893,
+      "grad_norm": 0.0,
+      "learning_rate": 4.643757684033026e-06,
+      "loss": 0.7137,
+      "step": 2235
+    },
+    {
+      "epoch": 0.6898041030387166,
+      "grad_norm": 0.0,
+      "learning_rate": 4.635319577977975e-06,
+      "loss": 0.615,
+      "step": 2236
+    },
+    {
+      "epoch": 0.690112602190344,
+      "grad_norm": 0.0,
+      "learning_rate": 4.626886831799668e-06,
+      "loss": 0.6953,
+      "step": 2237
+    },
+    {
+      "epoch": 0.6904211013419713,
+      "grad_norm": 0.0,
+      "learning_rate": 4.618459453923307e-06,
+      "loss": 0.7033,
+      "step": 2238
+    },
+    {
+      "epoch": 0.6907296004935987,
+      "grad_norm": 0.0,
+      "learning_rate": 4.6100374527687195e-06,
+      "loss": 0.7572,
+      "step": 2239
+    },
+    {
+      "epoch": 0.691038099645226,
+      "grad_norm": 0.0,
+      "learning_rate": 4.601620836750367e-06,
+      "loss": 0.7321,
+      "step": 2240
+    },
+    {
+      "epoch": 0.6913465987968533,
+      "grad_norm": 0.0,
+      "learning_rate": 4.593209614277325e-06,
+      "loss": 0.6992,
+      "step": 2241
+    },
+    {
+      "epoch": 0.6916550979484807,
+      "grad_norm": 0.0,
+      "learning_rate": 4.58480379375329e-06,
+      "loss": 1.0177,
+      "step": 2242
+    },
+    {
+      "epoch": 0.691963597100108,
+      "grad_norm": 0.0,
+      "learning_rate": 4.576403383576555e-06,
+      "loss": 0.6625,
+      "step": 2243
+    },
+    {
+      "epoch": 0.6922720962517354,
+      "grad_norm": 0.0,
+      "learning_rate": 4.568008392140003e-06,
+      "loss": 0.7334,
+      "step": 2244
+    },
+    {
+      "epoch": 0.6925805954033626,
+      "grad_norm": 0.0,
+      "learning_rate": 4.559618827831116e-06,
+      "loss": 0.6984,
+      "step": 2245
+    },
+    {
+      "epoch": 0.6928890945549899,
+      "grad_norm": 0.0,
+      "learning_rate": 4.551234699031938e-06,
+      "loss": 0.6868,
+      "step": 2246
+    },
+    {
+      "epoch": 0.6931975937066173,
+      "grad_norm": 0.0,
+      "learning_rate": 4.542856014119098e-06,
+      "loss": 0.7404,
+      "step": 2247
+    },
+    {
+      "epoch": 0.6935060928582446,
+      "grad_norm": 0.0,
+      "learning_rate": 4.534482781463775e-06,
+      "loss": 0.8335,
+      "step": 2248
+    },
+    {
+      "epoch": 0.6938145920098719,
+      "grad_norm": 0.0,
+      "learning_rate": 4.5261150094317e-06,
+      "loss": 0.7299,
+      "step": 2249
+    },
+    {
+      "epoch": 0.6941230911614993,
+      "grad_norm": 0.0,
+      "learning_rate": 4.517752706383159e-06,
+      "loss": 0.6716,
+      "step": 2250
+    },
+    {
+      "epoch": 0.6944315903131266,
+      "grad_norm": 0.0,
+      "learning_rate": 4.509395880672967e-06,
+      "loss": 0.6818,
+      "step": 2251
+    },
+    {
+      "epoch": 0.694740089464754,
+      "grad_norm": 0.0,
+      "learning_rate": 4.501044540650464e-06,
+      "loss": 0.7939,
+      "step": 2252
+    },
+    {
+      "epoch": 0.6950485886163813,
+      "grad_norm": 0.0,
+      "learning_rate": 4.4926986946595065e-06,
+      "loss": 0.7085,
+      "step": 2253
+    },
+    {
+      "epoch": 0.6953570877680086,
+      "grad_norm": 0.0,
+      "learning_rate": 4.484358351038478e-06,
+      "loss": 0.8178,
+      "step": 2254
+    },
+    {
+      "epoch": 0.695665586919636,
+      "grad_norm": 0.0,
+      "learning_rate": 4.4760235181202465e-06,
+      "loss": 0.7343,
+      "step": 2255
+    },
+    {
+      "epoch": 0.6959740860712633,
+      "grad_norm": 0.0,
+      "learning_rate": 4.467694204232187e-06,
+      "loss": 0.7586,
+      "step": 2256
+    },
+    {
+      "epoch": 0.6962825852228907,
+      "grad_norm": 0.0,
+      "learning_rate": 4.459370417696152e-06,
+      "loss": 0.8157,
+      "step": 2257
+    },
+    {
+      "epoch": 0.696591084374518,
+      "grad_norm": 0.0,
+      "learning_rate": 4.4510521668284736e-06,
+      "loss": 0.6318,
+      "step": 2258
+    },
+    {
+      "epoch": 0.6968995835261453,
+      "grad_norm": 0.0,
+      "learning_rate": 4.4427394599399575e-06,
+      "loss": 0.7224,
+      "step": 2259
+    },
+    {
+      "epoch": 0.6972080826777727,
+      "grad_norm": 0.0,
+      "learning_rate": 4.434432305335866e-06,
+      "loss": 0.6759,
+      "step": 2260
+    },
+    {
+      "epoch": 0.6975165818294,
+      "grad_norm": 0.0,
+      "learning_rate": 4.426130711315913e-06,
+      "loss": 0.7719,
+      "step": 2261
+    },
+    {
+      "epoch": 0.6978250809810274,
+      "grad_norm": 0.0,
+      "learning_rate": 4.417834686174263e-06,
+      "loss": 0.6957,
+      "step": 2262
+    },
+    {
+      "epoch": 0.6981335801326546,
+      "grad_norm": 0.0,
+      "learning_rate": 4.4095442381995055e-06,
+      "loss": 0.7719,
+      "step": 2263
+    },
+    {
+      "epoch": 0.6984420792842819,
+      "grad_norm": 0.0,
+      "learning_rate": 4.401259375674679e-06,
+      "loss": 0.6877,
+      "step": 2264
+    },
+    {
+      "epoch": 0.6987505784359093,
+      "grad_norm": 0.0,
+      "learning_rate": 4.392980106877212e-06,
+      "loss": 0.7441,
+      "step": 2265
+    },
+    {
+      "epoch": 0.6990590775875366,
+      "grad_norm": 0.0,
+      "learning_rate": 4.384706440078968e-06,
+      "loss": 0.7662,
+      "step": 2266
+    },
+    {
+      "epoch": 0.6993675767391639,
+      "grad_norm": 0.0,
+      "learning_rate": 4.376438383546202e-06,
+      "loss": 0.7548,
+      "step": 2267
+    },
+    {
+      "epoch": 0.6996760758907913,
+      "grad_norm": 0.0,
+      "learning_rate": 4.368175945539572e-06,
+      "loss": 0.6793,
+      "step": 2268
+    },
+    {
+      "epoch": 0.6999845750424186,
+      "grad_norm": 0.0,
+      "learning_rate": 4.359919134314113e-06,
+      "loss": 0.6686,
+      "step": 2269
+    },
+    {
+      "epoch": 0.700293074194046,
+      "grad_norm": 0.0,
+      "learning_rate": 4.351667958119242e-06,
+      "loss": 0.6941,
+      "step": 2270
+    },
+    {
+      "epoch": 0.7006015733456733,
+      "grad_norm": 0.0,
+      "learning_rate": 4.343422425198753e-06,
+      "loss": 0.7611,
+      "step": 2271
+    },
+    {
+      "epoch": 0.7009100724973006,
+      "grad_norm": 0.0,
+      "learning_rate": 4.335182543790788e-06,
+      "loss": 0.8059,
+      "step": 2272
+    },
+    {
+      "epoch": 0.701218571648928,
+      "grad_norm": 0.0,
+      "learning_rate": 4.326948322127858e-06,
+      "loss": 0.6992,
+      "step": 2273
+    },
+    {
+      "epoch": 0.7015270708005553,
+      "grad_norm": 0.0,
+      "learning_rate": 4.318719768436808e-06,
+      "loss": 0.7047,
+      "step": 2274
+    },
+    {
+      "epoch": 0.7018355699521827,
+      "grad_norm": 0.0,
+      "learning_rate": 4.3104968909388174e-06,
+      "loss": 0.7672,
+      "step": 2275
+    },
+    {
+      "epoch": 0.70214406910381,
+      "grad_norm": 0.0,
+      "learning_rate": 4.302279697849412e-06,
+      "loss": 0.7012,
+      "step": 2276
+    },
+    {
+      "epoch": 0.7024525682554373,
+      "grad_norm": 0.0,
+      "learning_rate": 4.29406819737842e-06,
+      "loss": 0.6696,
+      "step": 2277
+    },
+    {
+      "epoch": 0.7027610674070647,
+      "grad_norm": 0.0,
+      "learning_rate": 4.285862397729993e-06,
+      "loss": 0.7496,
+      "step": 2278
+    },
+    {
+      "epoch": 0.703069566558692,
+      "grad_norm": 0.0,
+      "learning_rate": 4.277662307102574e-06,
+      "loss": 0.7151,
+      "step": 2279
+    },
+    {
+      "epoch": 0.7033780657103192,
+      "grad_norm": 0.0,
+      "learning_rate": 4.26946793368892e-06,
+      "loss": 0.6707,
+      "step": 2280
+    },
+    {
+      "epoch": 0.7036865648619466,
+      "grad_norm": 0.0,
+      "learning_rate": 4.261279285676071e-06,
+      "loss": 0.6088,
+      "step": 2281
+    },
+    {
+      "epoch": 0.7039950640135739,
+      "grad_norm": 0.0,
+      "learning_rate": 4.253096371245329e-06,
+      "loss": 0.6599,
+      "step": 2282
+    },
+    {
+      "epoch": 0.7043035631652013,
+      "grad_norm": 0.0,
+      "learning_rate": 4.244919198572293e-06,
+      "loss": 0.7794,
+      "step": 2283
+    },
+    {
+      "epoch": 0.7046120623168286,
+      "grad_norm": 0.0,
+      "learning_rate": 4.236747775826804e-06,
+      "loss": 0.6954,
+      "step": 2284
+    },
+    {
+      "epoch": 0.7049205614684559,
+      "grad_norm": 0.0,
+      "learning_rate": 4.228582111172977e-06,
+      "loss": 0.6958,
+      "step": 2285
+    },
+    {
+      "epoch": 0.7052290606200833,
+      "grad_norm": 0.0,
+      "learning_rate": 4.220422212769161e-06,
+      "loss": 0.6756,
+      "step": 2286
+    },
+    {
+      "epoch": 0.7055375597717106,
+      "grad_norm": 0.0,
+      "learning_rate": 4.212268088767944e-06,
+      "loss": 0.9999,
+      "step": 2287
+    },
+    {
+      "epoch": 0.705846058923338,
+      "grad_norm": 0.0,
+      "learning_rate": 4.204119747316157e-06,
+      "loss": 0.759,
+      "step": 2288
+    },
+    {
+      "epoch": 0.7061545580749653,
+      "grad_norm": 0.0,
+      "learning_rate": 4.195977196554835e-06,
+      "loss": 0.7907,
+      "step": 2289
+    },
+    {
+      "epoch": 0.7064630572265926,
+      "grad_norm": 0.0,
+      "learning_rate": 4.187840444619251e-06,
+      "loss": 0.7364,
+      "step": 2290
+    },
+    {
+      "epoch": 0.70677155637822,
+      "grad_norm": 0.0,
+      "learning_rate": 4.179709499638857e-06,
+      "loss": 0.7315,
+      "step": 2291
+    },
+    {
+      "epoch": 0.7070800555298473,
+      "grad_norm": 0.0,
+      "learning_rate": 4.171584369737322e-06,
+      "loss": 0.7142,
+      "step": 2292
+    },
+    {
+      "epoch": 0.7073885546814747,
+      "grad_norm": 0.0,
+      "learning_rate": 4.163465063032507e-06,
+      "loss": 0.7166,
+      "step": 2293
+    },
+    {
+      "epoch": 0.707697053833102,
+      "grad_norm": 0.0,
+      "learning_rate": 4.1553515876364435e-06,
+      "loss": 0.6211,
+      "step": 2294
+    },
+    {
+      "epoch": 0.7080055529847293,
+      "grad_norm": 0.0,
+      "learning_rate": 4.147243951655341e-06,
+      "loss": 0.7314,
+      "step": 2295
+    },
+    {
+      "epoch": 0.7083140521363567,
+      "grad_norm": 0.0,
+      "learning_rate": 4.139142163189573e-06,
+      "loss": 0.7196,
+      "step": 2296
+    },
+    {
+      "epoch": 0.7086225512879839,
+      "grad_norm": 0.0,
+      "learning_rate": 4.131046230333682e-06,
+      "loss": 0.7261,
+      "step": 2297
+    },
+    {
+      "epoch": 0.7089310504396112,
+      "grad_norm": 0.0,
+      "learning_rate": 4.1229561611763445e-06,
+      "loss": 0.7069,
+      "step": 2298
+    },
+    {
+      "epoch": 0.7092395495912386,
+      "grad_norm": 0.0,
+      "learning_rate": 4.114871963800385e-06,
+      "loss": 0.6299,
+      "step": 2299
+    },
+    {
+      "epoch": 0.7095480487428659,
+      "grad_norm": 0.0,
+      "learning_rate": 4.106793646282769e-06,
+      "loss": 0.7613,
+      "step": 2300
+    },
+    {
+      "epoch": 0.7098565478944933,
+      "grad_norm": 0.0,
+      "learning_rate": 4.098721216694572e-06,
+      "loss": 0.7339,
+      "step": 2301
+    },
+    {
+      "epoch": 0.7101650470461206,
+      "grad_norm": 0.0,
+      "learning_rate": 4.090654683101007e-06,
+      "loss": 0.7247,
+      "step": 2302
+    },
+    {
+      "epoch": 0.7104735461977479,
+      "grad_norm": 0.0,
+      "learning_rate": 4.082594053561369e-06,
+      "loss": 0.664,
+      "step": 2303
+    },
+    {
+      "epoch": 0.7107820453493753,
+      "grad_norm": 0.0,
+      "learning_rate": 4.074539336129079e-06,
+      "loss": 0.7243,
+      "step": 2304
+    },
+    {
+      "epoch": 0.7110905445010026,
+      "grad_norm": 0.0,
+      "learning_rate": 4.066490538851644e-06,
+      "loss": 0.6925,
+      "step": 2305
+    },
+    {
+      "epoch": 0.71139904365263,
+      "grad_norm": 0.0,
+      "learning_rate": 4.0584476697706475e-06,
+      "loss": 0.7386,
+      "step": 2306
+    },
+    {
+      "epoch": 0.7117075428042573,
+      "grad_norm": 0.0,
+      "learning_rate": 4.0504107369217686e-06,
+      "loss": 0.6881,
+      "step": 2307
+    },
+    {
+      "epoch": 0.7120160419558846,
+      "grad_norm": 0.0,
+      "learning_rate": 4.042379748334727e-06,
+      "loss": 0.719,
+      "step": 2308
+    },
+    {
+      "epoch": 0.712324541107512,
+      "grad_norm": 0.0,
+      "learning_rate": 4.034354712033332e-06,
+      "loss": 0.7329,
+      "step": 2309
+    },
+    {
+      "epoch": 0.7126330402591393,
+      "grad_norm": 0.0,
+      "learning_rate": 4.026335636035429e-06,
+      "loss": 0.7455,
+      "step": 2310
+    },
+    {
+      "epoch": 0.7129415394107667,
+      "grad_norm": 0.0,
+      "learning_rate": 4.018322528352917e-06,
+      "loss": 0.7113,
+      "step": 2311
+    },
+    {
+      "epoch": 0.713250038562394,
+      "grad_norm": 0.0,
+      "learning_rate": 4.010315396991727e-06,
+      "loss": 0.6837,
+      "step": 2312
+    },
+    {
+      "epoch": 0.7135585377140213,
+      "grad_norm": 0.0,
+      "learning_rate": 4.002314249951819e-06,
+      "loss": 0.6868,
+      "step": 2313
+    },
+    {
+      "epoch": 0.7138670368656487,
+      "grad_norm": 0.0,
+      "learning_rate": 3.994319095227178e-06,
+      "loss": 0.7261,
+      "step": 2314
+    },
+    {
+      "epoch": 0.7141755360172759,
+      "grad_norm": 0.0,
+      "learning_rate": 3.986329940805799e-06,
+      "loss": 0.7157,
+      "step": 2315
+    },
+    {
+      "epoch": 0.7144840351689032,
+      "grad_norm": 0.0,
+      "learning_rate": 3.978346794669679e-06,
+      "loss": 0.741,
+      "step": 2316
+    },
+    {
+      "epoch": 0.7147925343205306,
+      "grad_norm": 0.0,
+      "learning_rate": 3.970369664794823e-06,
+      "loss": 0.7708,
+      "step": 2317
+    },
+    {
+      "epoch": 0.7151010334721579,
+      "grad_norm": 0.0,
+      "learning_rate": 3.9623985591512105e-06,
+      "loss": 0.7344,
+      "step": 2318
+    },
+    {
+      "epoch": 0.7154095326237853,
+      "grad_norm": 0.0,
+      "learning_rate": 3.95443348570282e-06,
+      "loss": 0.7342,
+      "step": 2319
+    },
+    {
+      "epoch": 0.7157180317754126,
+      "grad_norm": 0.0,
+      "learning_rate": 3.946474452407579e-06,
+      "loss": 0.6732,
+      "step": 2320
+    },
+    {
+      "epoch": 0.7160265309270399,
+      "grad_norm": 0.0,
+      "learning_rate": 3.938521467217405e-06,
+      "loss": 0.6827,
+      "step": 2321
+    },
+    {
+      "epoch": 0.7163350300786673,
+      "grad_norm": 0.0,
+      "learning_rate": 3.930574538078155e-06,
+      "loss": 0.7727,
+      "step": 2322
+    },
+    {
+      "epoch": 0.7166435292302946,
+      "grad_norm": 0.0,
+      "learning_rate": 3.922633672929648e-06,
+      "loss": 0.7735,
+      "step": 2323
+    },
+    {
+      "epoch": 0.716952028381922,
+      "grad_norm": 0.0,
+      "learning_rate": 3.914698879705635e-06,
+      "loss": 0.7188,
+      "step": 2324
+    },
+    {
+      "epoch": 0.7172605275335493,
+      "grad_norm": 0.0,
+      "learning_rate": 3.906770166333802e-06,
+      "loss": 0.644,
+      "step": 2325
+    },
+    {
+      "epoch": 0.7175690266851766,
+      "grad_norm": 0.0,
+      "learning_rate": 3.898847540735771e-06,
+      "loss": 0.6054,
+      "step": 2326
+    },
+    {
+      "epoch": 0.717877525836804,
+      "grad_norm": 0.0,
+      "learning_rate": 3.890931010827062e-06,
+      "loss": 0.6724,
+      "step": 2327
+    },
+    {
+      "epoch": 0.7181860249884313,
+      "grad_norm": 0.0,
+      "learning_rate": 3.883020584517129e-06,
+      "loss": 0.8042,
+      "step": 2328
+    },
+    {
+      "epoch": 0.7184945241400587,
+      "grad_norm": 0.0,
+      "learning_rate": 3.875116269709307e-06,
+      "loss": 0.7232,
+      "step": 2329
+    },
+    {
+      "epoch": 0.718803023291686,
+      "grad_norm": 0.0,
+      "learning_rate": 3.867218074300832e-06,
+      "loss": 0.7611,
+      "step": 2330
+    },
+    {
+      "epoch": 0.7191115224433133,
+      "grad_norm": 0.0,
+      "learning_rate": 3.859326006182833e-06,
+      "loss": 0.6313,
+      "step": 2331
+    },
+    {
+      "epoch": 0.7194200215949406,
+      "grad_norm": 0.0,
+      "learning_rate": 3.851440073240309e-06,
+      "loss": 0.7193,
+      "step": 2332
+    },
+    {
+      "epoch": 0.7197285207465679,
+      "grad_norm": 0.0,
+      "learning_rate": 3.84356028335213e-06,
+      "loss": 0.6549,
+      "step": 2333
+    },
+    {
+      "epoch": 0.7200370198981952,
+      "grad_norm": 0.0,
+      "learning_rate": 3.835686644391029e-06,
+      "loss": 0.6637,
+      "step": 2334
+    },
+    {
+      "epoch": 0.7203455190498226,
+      "grad_norm": 0.0,
+      "learning_rate": 3.827819164223599e-06,
+      "loss": 0.7928,
+      "step": 2335
+    },
+    {
+      "epoch": 0.7206540182014499,
+      "grad_norm": 0.0,
+      "learning_rate": 3.819957850710269e-06,
+      "loss": 0.6682,
+      "step": 2336
+    },
+    {
+      "epoch": 0.7209625173530773,
+      "grad_norm": 0.0,
+      "learning_rate": 3.812102711705323e-06,
+      "loss": 0.7203,
+      "step": 2337
+    },
+    {
+      "epoch": 0.7212710165047046,
+      "grad_norm": 0.0,
+      "learning_rate": 3.8042537550568603e-06,
+      "loss": 0.6589,
+      "step": 2338
+    },
+    {
+      "epoch": 0.7215795156563319,
+      "grad_norm": 0.0,
+      "learning_rate": 3.7964109886068066e-06,
+      "loss": 0.7143,
+      "step": 2339
+    },
+    {
+      "epoch": 0.7218880148079593,
+      "grad_norm": 0.0,
+      "learning_rate": 3.7885744201909115e-06,
+      "loss": 0.7169,
+      "step": 2340
+    },
+    {
+      "epoch": 0.7221965139595866,
+      "grad_norm": 0.0,
+      "learning_rate": 3.7807440576387244e-06,
+      "loss": 0.6901,
+      "step": 2341
+    },
+    {
+      "epoch": 0.722505013111214,
+      "grad_norm": 0.0,
+      "learning_rate": 3.7729199087735924e-06,
+      "loss": 0.7984,
+      "step": 2342
+    },
+    {
+      "epoch": 0.7228135122628413,
+      "grad_norm": 0.0,
+      "learning_rate": 3.7651019814126656e-06,
+      "loss": 0.693,
+      "step": 2343
+    },
+    {
+      "epoch": 0.7231220114144686,
+      "grad_norm": 0.0,
+      "learning_rate": 3.7572902833668635e-06,
+      "loss": 0.7642,
+      "step": 2344
+    },
+    {
+      "epoch": 0.723430510566096,
+      "grad_norm": 0.0,
+      "learning_rate": 3.7494848224408998e-06,
+      "loss": 0.6759,
+      "step": 2345
+    },
+    {
+      "epoch": 0.7237390097177233,
+      "grad_norm": 0.0,
+      "learning_rate": 3.741685606433233e-06,
+      "loss": 0.7235,
+      "step": 2346
+    },
+    {
+      "epoch": 0.7240475088693507,
+      "grad_norm": 0.0,
+      "learning_rate": 3.7338926431361055e-06,
+      "loss": 0.7229,
+      "step": 2347
+    },
+    {
+      "epoch": 0.724356008020978,
+      "grad_norm": 0.0,
+      "learning_rate": 3.726105940335495e-06,
+      "loss": 0.7559,
+      "step": 2348
+    },
+    {
+      "epoch": 0.7246645071726052,
+      "grad_norm": 0.0,
+      "learning_rate": 3.71832550581114e-06,
+      "loss": 0.7572,
+      "step": 2349
+    },
+    {
+      "epoch": 0.7249730063242326,
+      "grad_norm": 0.0,
+      "learning_rate": 3.710551347336504e-06,
+      "loss": 0.7353,
+      "step": 2350
+    },
+    {
+      "epoch": 0.7252815054758599,
+      "grad_norm": 0.0,
+      "learning_rate": 3.7027834726787806e-06,
+      "loss": 0.6815,
+      "step": 2351
+    },
+    {
+      "epoch": 0.7255900046274872,
+      "grad_norm": 0.0,
+      "learning_rate": 3.6950218895988966e-06,
+      "loss": 0.7229,
+      "step": 2352
+    },
+    {
+      "epoch": 0.7258985037791146,
+      "grad_norm": 0.0,
+      "learning_rate": 3.6872666058514783e-06,
+      "loss": 0.6987,
+      "step": 2353
+    },
+    {
+      "epoch": 0.7262070029307419,
+      "grad_norm": 0.0,
+      "learning_rate": 3.6795176291848713e-06,
+      "loss": 0.7232,
+      "step": 2354
+    },
+    {
+      "epoch": 0.7265155020823693,
+      "grad_norm": 0.0,
+      "learning_rate": 3.6717749673411096e-06,
+      "loss": 0.7647,
+      "step": 2355
+    },
+    {
+      "epoch": 0.7268240012339966,
+      "grad_norm": 0.0,
+      "learning_rate": 3.66403862805592e-06,
+      "loss": 0.7256,
+      "step": 2356
+    },
+    {
+      "epoch": 0.7271325003856239,
+      "grad_norm": 0.0,
+      "learning_rate": 3.6563086190587215e-06,
+      "loss": 0.6587,
+      "step": 2357
+    },
+    {
+      "epoch": 0.7274409995372513,
+      "grad_norm": 0.0,
+      "learning_rate": 3.6485849480725964e-06,
+      "loss": 0.6631,
+      "step": 2358
+    },
+    {
+      "epoch": 0.7277494986888786,
+      "grad_norm": 0.0,
+      "learning_rate": 3.6408676228143013e-06,
+      "loss": 0.7334,
+      "step": 2359
+    },
+    {
+      "epoch": 0.728057997840506,
+      "grad_norm": 0.0,
+      "learning_rate": 3.633156650994247e-06,
+      "loss": 0.9905,
+      "step": 2360
+    },
+    {
+      "epoch": 0.7283664969921333,
+      "grad_norm": 0.0,
+      "learning_rate": 3.625452040316505e-06,
+      "loss": 0.7015,
+      "step": 2361
+    },
+    {
+      "epoch": 0.7286749961437606,
+      "grad_norm": 0.0,
+      "learning_rate": 3.6177537984787924e-06,
+      "loss": 0.731,
+      "step": 2362
+    },
+    {
+      "epoch": 0.728983495295388,
+      "grad_norm": 0.0,
+      "learning_rate": 3.610061933172445e-06,
+      "loss": 0.7258,
+      "step": 2363
+    },
+    {
+      "epoch": 0.7292919944470153,
+      "grad_norm": 0.0,
+      "learning_rate": 3.602376452082451e-06,
+      "loss": 0.7024,
+      "step": 2364
+    },
+    {
+      "epoch": 0.7296004935986427,
+      "grad_norm": 0.0,
+      "learning_rate": 3.5946973628874026e-06,
+      "loss": 0.7763,
+      "step": 2365
+    },
+    {
+      "epoch": 0.72990899275027,
+      "grad_norm": 0.0,
+      "learning_rate": 3.587024673259519e-06,
+      "loss": 0.7543,
+      "step": 2366
+    },
+    {
+      "epoch": 0.7302174919018972,
+      "grad_norm": 0.0,
+      "learning_rate": 3.579358390864618e-06,
+      "loss": 0.9771,
+      "step": 2367
+    },
+    {
+      "epoch": 0.7305259910535246,
+      "grad_norm": 0.0,
+      "learning_rate": 3.571698523362113e-06,
+      "loss": 0.7753,
+      "step": 2368
+    },
+    {
+      "epoch": 0.7308344902051519,
+      "grad_norm": 0.0,
+      "learning_rate": 3.5640450784050185e-06,
+      "loss": 0.7656,
+      "step": 2369
+    },
+    {
+      "epoch": 0.7311429893567792,
+      "grad_norm": 0.0,
+      "learning_rate": 3.556398063639921e-06,
+      "loss": 0.7309,
+      "step": 2370
+    },
+    {
+      "epoch": 0.7314514885084066,
+      "grad_norm": 0.0,
+      "learning_rate": 3.548757486706997e-06,
+      "loss": 0.713,
+      "step": 2371
+    },
+    {
+      "epoch": 0.7317599876600339,
+      "grad_norm": 0.0,
+      "learning_rate": 3.5411233552399703e-06,
+      "loss": 0.6742,
+      "step": 2372
+    },
+    {
+      "epoch": 0.7320684868116613,
+      "grad_norm": 0.0,
+      "learning_rate": 3.533495676866141e-06,
+      "loss": 0.8157,
+      "step": 2373
+    },
+    {
+      "epoch": 0.7323769859632886,
+      "grad_norm": 0.0,
+      "learning_rate": 3.525874459206362e-06,
+      "loss": 0.7107,
+      "step": 2374
+    },
+    {
+      "epoch": 0.7326854851149159,
+      "grad_norm": 0.0,
+      "learning_rate": 3.518259709875023e-06,
+      "loss": 0.7892,
+      "step": 2375
+    },
+    {
+      "epoch": 0.7329939842665433,
+      "grad_norm": 0.0,
+      "learning_rate": 3.5106514364800548e-06,
+      "loss": 0.6195,
+      "step": 2376
+    },
+    {
+      "epoch": 0.7333024834181706,
+      "grad_norm": 0.0,
+      "learning_rate": 3.503049646622915e-06,
+      "loss": 0.7394,
+      "step": 2377
+    },
+    {
+      "epoch": 0.733610982569798,
+      "grad_norm": 0.0,
+      "learning_rate": 3.4954543478985926e-06,
+      "loss": 0.7432,
+      "step": 2378
+    },
+    {
+      "epoch": 0.7339194817214253,
+      "grad_norm": 0.0,
+      "learning_rate": 3.4878655478955838e-06,
+      "loss": 0.6289,
+      "step": 2379
+    },
+    {
+      "epoch": 0.7342279808730526,
+      "grad_norm": 0.0,
+      "learning_rate": 3.4802832541958886e-06,
+      "loss": 0.7075,
+      "step": 2380
+    },
+    {
+      "epoch": 0.73453648002468,
+      "grad_norm": 0.0,
+      "learning_rate": 3.472707474375018e-06,
+      "loss": 0.7032,
+      "step": 2381
+    },
+    {
+      "epoch": 0.7348449791763073,
+      "grad_norm": 0.0,
+      "learning_rate": 3.465138216001962e-06,
+      "loss": 0.7202,
+      "step": 2382
+    },
+    {
+      "epoch": 0.7351534783279347,
+      "grad_norm": 0.0,
+      "learning_rate": 3.457575486639211e-06,
+      "loss": 0.6907,
+      "step": 2383
+    },
+    {
+      "epoch": 0.7354619774795619,
+      "grad_norm": 0.0,
+      "learning_rate": 3.4500192938427093e-06,
+      "loss": 0.5944,
+      "step": 2384
+    },
+    {
+      "epoch": 0.7357704766311892,
+      "grad_norm": 0.0,
+      "learning_rate": 3.44246964516189e-06,
+      "loss": 0.7482,
+      "step": 2385
+    },
+    {
+      "epoch": 0.7360789757828166,
+      "grad_norm": 0.0,
+      "learning_rate": 3.4349265481396453e-06,
+      "loss": 0.7875,
+      "step": 2386
+    },
+    {
+      "epoch": 0.7363874749344439,
+      "grad_norm": 0.0,
+      "learning_rate": 3.4273900103123125e-06,
+      "loss": 0.7129,
+      "step": 2387
+    },
+    {
+      "epoch": 0.7366959740860712,
+      "grad_norm": 0.0,
+      "learning_rate": 3.4198600392096894e-06,
+      "loss": 0.7024,
+      "step": 2388
+    },
+    {
+      "epoch": 0.7370044732376986,
+      "grad_norm": 0.0,
+      "learning_rate": 3.4123366423549907e-06,
+      "loss": 0.7508,
+      "step": 2389
+    },
+    {
+      "epoch": 0.7373129723893259,
+      "grad_norm": 0.0,
+      "learning_rate": 3.404819827264888e-06,
+      "loss": 0.7088,
+      "step": 2390
+    },
+    {
+      "epoch": 0.7376214715409533,
+      "grad_norm": 0.0,
+      "learning_rate": 3.397309601449459e-06,
+      "loss": 0.766,
+      "step": 2391
+    },
+    {
+      "epoch": 0.7379299706925806,
+      "grad_norm": 0.0,
+      "learning_rate": 3.38980597241221e-06,
+      "loss": 0.7314,
+      "step": 2392
+    },
+    {
+      "epoch": 0.7382384698442079,
+      "grad_norm": 0.0,
+      "learning_rate": 3.3823089476500492e-06,
+      "loss": 0.7783,
+      "step": 2393
+    },
+    {
+      "epoch": 0.7385469689958353,
+      "grad_norm": 0.0,
+      "learning_rate": 3.3748185346532847e-06,
+      "loss": 0.6801,
+      "step": 2394
+    },
+    {
+      "epoch": 0.7388554681474626,
+      "grad_norm": 0.0,
+      "learning_rate": 3.367334740905629e-06,
+      "loss": 0.7453,
+      "step": 2395
+    },
+    {
+      "epoch": 0.73916396729909,
+      "grad_norm": 0.0,
+      "learning_rate": 3.3598575738841698e-06,
+      "loss": 0.7184,
+      "step": 2396
+    },
+    {
+      "epoch": 0.7394724664507173,
+      "grad_norm": 0.0,
+      "learning_rate": 3.352387041059377e-06,
+      "loss": 0.7691,
+      "step": 2397
+    },
+    {
+      "epoch": 0.7397809656023446,
+      "grad_norm": 0.0,
+      "learning_rate": 3.3449231498951016e-06,
+      "loss": 0.7078,
+      "step": 2398
+    },
+    {
+      "epoch": 0.740089464753972,
+      "grad_norm": 0.0,
+      "learning_rate": 3.337465907848544e-06,
+      "loss": 0.7818,
+      "step": 2399
+    },
+    {
+      "epoch": 0.7403979639055993,
+      "grad_norm": 0.0,
+      "learning_rate": 3.330015322370277e-06,
+      "loss": 0.7287,
+      "step": 2400
+    },
+    {
+      "epoch": 0.7407064630572265,
+      "grad_norm": 0.0,
+      "learning_rate": 3.3225714009042045e-06,
+      "loss": 0.6973,
+      "step": 2401
+    },
+    {
+      "epoch": 0.7410149622088539,
+      "grad_norm": 0.0,
+      "learning_rate": 3.31513415088759e-06,
+      "loss": 0.6757,
+      "step": 2402
+    },
+    {
+      "epoch": 0.7413234613604812,
+      "grad_norm": 0.0,
+      "learning_rate": 3.307703579751018e-06,
+      "loss": 0.661,
+      "step": 2403
+    },
+    {
+      "epoch": 0.7416319605121086,
+      "grad_norm": 0.0,
+      "learning_rate": 3.3002796949184135e-06,
+      "loss": 0.8085,
+      "step": 2404
+    },
+    {
+      "epoch": 0.7419404596637359,
+      "grad_norm": 0.0,
+      "learning_rate": 3.2928625038070093e-06,
+      "loss": 0.7602,
+      "step": 2405
+    },
+    {
+      "epoch": 0.7422489588153632,
+      "grad_norm": 0.0,
+      "learning_rate": 3.2854520138273526e-06,
+      "loss": 0.7372,
+      "step": 2406
+    },
+    {
+      "epoch": 0.7425574579669906,
+      "grad_norm": 0.0,
+      "learning_rate": 3.278048232383305e-06,
+      "loss": 0.6769,
+      "step": 2407
+    },
+    {
+      "epoch": 0.7428659571186179,
+      "grad_norm": 0.0,
+      "learning_rate": 3.27065116687201e-06,
+      "loss": 0.6679,
+      "step": 2408
+    },
+    {
+      "epoch": 0.7431744562702453,
+      "grad_norm": 0.0,
+      "learning_rate": 3.263260824683917e-06,
+      "loss": 0.7707,
+      "step": 2409
+    },
+    {
+      "epoch": 0.7434829554218726,
+      "grad_norm": 0.0,
+      "learning_rate": 3.255877213202748e-06,
+      "loss": 0.6441,
+      "step": 2410
+    },
+    {
+      "epoch": 0.7437914545734999,
+      "grad_norm": 0.0,
+      "learning_rate": 3.248500339805499e-06,
+      "loss": 0.7432,
+      "step": 2411
+    },
+    {
+      "epoch": 0.7440999537251273,
+      "grad_norm": 0.0,
+      "learning_rate": 3.2411302118624453e-06,
+      "loss": 0.7155,
+      "step": 2412
+    },
+    {
+      "epoch": 0.7444084528767546,
+      "grad_norm": 0.0,
+      "learning_rate": 3.233766836737111e-06,
+      "loss": 0.6777,
+      "step": 2413
+    },
+    {
+      "epoch": 0.744716952028382,
+      "grad_norm": 0.0,
+      "learning_rate": 3.2264102217862794e-06,
+      "loss": 0.6547,
+      "step": 2414
+    },
+    {
+      "epoch": 0.7450254511800093,
+      "grad_norm": 0.0,
+      "learning_rate": 3.2190603743599723e-06,
+      "loss": 0.724,
+      "step": 2415
+    },
+    {
+      "epoch": 0.7453339503316366,
+      "grad_norm": 0.0,
+      "learning_rate": 3.2117173018014656e-06,
+      "loss": 0.7582,
+      "step": 2416
+    },
+    {
+      "epoch": 0.745642449483264,
+      "grad_norm": 0.0,
+      "learning_rate": 3.2043810114472508e-06,
+      "loss": 0.7383,
+      "step": 2417
+    },
+    {
+      "epoch": 0.7459509486348913,
+      "grad_norm": 0.0,
+      "learning_rate": 3.1970515106270474e-06,
+      "loss": 0.712,
+      "step": 2418
+    },
+    {
+      "epoch": 0.7462594477865185,
+      "grad_norm": 0.0,
+      "learning_rate": 3.1897288066638e-06,
+      "loss": 0.7693,
+      "step": 2419
+    },
+    {
+      "epoch": 0.7465679469381459,
+      "grad_norm": 0.0,
+      "learning_rate": 3.1824129068736474e-06,
+      "loss": 0.6618,
+      "step": 2420
+    },
+    {
+      "epoch": 0.7468764460897732,
+      "grad_norm": 0.0,
+      "learning_rate": 3.1751038185659467e-06,
+      "loss": 0.7433,
+      "step": 2421
+    },
+    {
+      "epoch": 0.7471849452414006,
+      "grad_norm": 0.0,
+      "learning_rate": 3.1678015490432368e-06,
+      "loss": 0.6502,
+      "step": 2422
+    },
+    {
+      "epoch": 0.7474934443930279,
+      "grad_norm": 0.0,
+      "learning_rate": 3.1605061056012465e-06,
+      "loss": 0.7712,
+      "step": 2423
+    },
+    {
+      "epoch": 0.7478019435446552,
+      "grad_norm": 0.0,
+      "learning_rate": 3.153217495528893e-06,
+      "loss": 0.6721,
+      "step": 2424
+    },
+    {
+      "epoch": 0.7481104426962826,
+      "grad_norm": 0.0,
+      "learning_rate": 3.1459357261082525e-06,
+      "loss": 0.7167,
+      "step": 2425
+    },
+    {
+      "epoch": 0.7484189418479099,
+      "grad_norm": 0.0,
+      "learning_rate": 3.138660804614586e-06,
+      "loss": 0.6691,
+      "step": 2426
+    },
+    {
+      "epoch": 0.7487274409995373,
+      "grad_norm": 0.0,
+      "learning_rate": 3.131392738316286e-06,
+      "loss": 0.7755,
+      "step": 2427
+    },
+    {
+      "epoch": 0.7490359401511646,
+      "grad_norm": 0.0,
+      "learning_rate": 3.12413153447492e-06,
+      "loss": 0.6478,
+      "step": 2428
+    },
+    {
+      "epoch": 0.7493444393027919,
+      "grad_norm": 0.0,
+      "learning_rate": 3.1168772003451855e-06,
+      "loss": 0.7047,
+      "step": 2429
+    },
+    {
+      "epoch": 0.7496529384544193,
+      "grad_norm": 0.0,
+      "learning_rate": 3.1096297431749257e-06,
+      "loss": 0.6735,
+      "step": 2430
+    },
+    {
+      "epoch": 0.7499614376060466,
+      "grad_norm": 0.0,
+      "learning_rate": 3.1023891702051057e-06,
+      "loss": 1.0232,
+      "step": 2431
+    },
+    {
+      "epoch": 0.750269936757674,
+      "grad_norm": 0.0,
+      "learning_rate": 3.095155488669812e-06,
+      "loss": 0.6308,
+      "step": 2432
+    },
+    {
+      "epoch": 0.7505784359093013,
+      "grad_norm": 0.0,
+      "learning_rate": 3.087928705796256e-06,
+      "loss": 0.7367,
+      "step": 2433
+    },
+    {
+      "epoch": 0.7508869350609286,
+      "grad_norm": 0.0,
+      "learning_rate": 3.0807088288047447e-06,
+      "loss": 0.6943,
+      "step": 2434
+    },
+    {
+      "epoch": 0.751195434212556,
+      "grad_norm": 0.0,
+      "learning_rate": 3.0734958649086878e-06,
+      "loss": 0.6738,
+      "step": 2435
+    },
+    {
+      "epoch": 0.7515039333641832,
+      "grad_norm": 0.0,
+      "learning_rate": 3.0662898213145965e-06,
+      "loss": 0.7519,
+      "step": 2436
+    },
+    {
+      "epoch": 0.7518124325158105,
+      "grad_norm": 0.0,
+      "learning_rate": 3.0590907052220566e-06,
+      "loss": 0.6588,
+      "step": 2437
+    },
+    {
+      "epoch": 0.7521209316674379,
+      "grad_norm": 0.0,
+      "learning_rate": 3.051898523823743e-06,
+      "loss": 0.6724,
+      "step": 2438
+    },
+    {
+      "epoch": 0.7524294308190652,
+      "grad_norm": 0.0,
+      "learning_rate": 3.044713284305395e-06,
+      "loss": 0.7122,
+      "step": 2439
+    },
+    {
+      "epoch": 0.7527379299706926,
+      "grad_norm": 0.0,
+      "learning_rate": 3.037534993845818e-06,
+      "loss": 0.6708,
+      "step": 2440
+    },
+    {
+      "epoch": 0.7530464291223199,
+      "grad_norm": 0.0,
+      "learning_rate": 3.0303636596168706e-06,
+      "loss": 0.7263,
+      "step": 2441
+    },
+    {
+      "epoch": 0.7533549282739472,
+      "grad_norm": 0.0,
+      "learning_rate": 3.0231992887834725e-06,
+      "loss": 0.7195,
+      "step": 2442
+    },
+    {
+      "epoch": 0.7536634274255746,
+      "grad_norm": 0.0,
+      "learning_rate": 3.016041888503578e-06,
+      "loss": 0.7201,
+      "step": 2443
+    },
+    {
+      "epoch": 0.7539719265772019,
+      "grad_norm": 0.0,
+      "learning_rate": 3.008891465928172e-06,
+      "loss": 0.7869,
+      "step": 2444
+    },
+    {
+      "epoch": 0.7542804257288293,
+      "grad_norm": 0.0,
+      "learning_rate": 3.0017480282012836e-06,
+      "loss": 0.6756,
+      "step": 2445
+    },
+    {
+      "epoch": 0.7545889248804566,
+      "grad_norm": 0.0,
+      "learning_rate": 2.9946115824599465e-06,
+      "loss": 0.7122,
+      "step": 2446
+    },
+    {
+      "epoch": 0.7548974240320839,
+      "grad_norm": 0.0,
+      "learning_rate": 2.987482135834223e-06,
+      "loss": 0.6999,
+      "step": 2447
+    },
+    {
+      "epoch": 0.7552059231837113,
+      "grad_norm": 0.0,
+      "learning_rate": 2.9803596954471714e-06,
+      "loss": 0.6999,
+      "step": 2448
+    },
+    {
+      "epoch": 0.7555144223353386,
+      "grad_norm": 0.0,
+      "learning_rate": 2.9732442684148533e-06,
+      "loss": 0.7637,
+      "step": 2449
+    },
+    {
+      "epoch": 0.755822921486966,
+      "grad_norm": 0.0,
+      "learning_rate": 2.9661358618463297e-06,
+      "loss": 0.724,
+      "step": 2450
+    },
+    {
+      "epoch": 0.7561314206385933,
+      "grad_norm": 0.0,
+      "learning_rate": 2.959034482843639e-06,
+      "loss": 0.6431,
+      "step": 2451
+    },
+    {
+      "epoch": 0.7564399197902206,
+      "grad_norm": 0.0,
+      "learning_rate": 2.951940138501801e-06,
+      "loss": 0.9853,
+      "step": 2452
+    },
+    {
+      "epoch": 0.7567484189418479,
+      "grad_norm": 0.0,
+      "learning_rate": 2.944852835908806e-06,
+      "loss": 0.6493,
+      "step": 2453
+    },
+    {
+      "epoch": 0.7570569180934752,
+      "grad_norm": 0.0,
+      "learning_rate": 2.9377725821456138e-06,
+      "loss": 0.6325,
+      "step": 2454
+    },
+    {
+      "epoch": 0.7573654172451025,
+      "grad_norm": 0.0,
+      "learning_rate": 2.9306993842861354e-06,
+      "loss": 0.69,
+      "step": 2455
+    },
+    {
+      "epoch": 0.7576739163967299,
+      "grad_norm": 0.0,
+      "learning_rate": 2.9236332493972386e-06,
+      "loss": 0.6958,
+      "step": 2456
+    },
+    {
+      "epoch": 0.7579824155483572,
+      "grad_norm": 0.0,
+      "learning_rate": 2.916574184538731e-06,
+      "loss": 0.837,
+      "step": 2457
+    },
+    {
+      "epoch": 0.7582909146999846,
+      "grad_norm": 0.0,
+      "learning_rate": 2.9095221967633502e-06,
+      "loss": 0.6747,
+      "step": 2458
+    },
+    {
+      "epoch": 0.7585994138516119,
+      "grad_norm": 0.0,
+      "learning_rate": 2.9024772931167777e-06,
+      "loss": 0.6415,
+      "step": 2459
+    },
+    {
+      "epoch": 0.7589079130032392,
+      "grad_norm": 0.0,
+      "learning_rate": 2.8954394806376052e-06,
+      "loss": 0.6934,
+      "step": 2460
+    },
+    {
+      "epoch": 0.7592164121548666,
+      "grad_norm": 0.0,
+      "learning_rate": 2.8884087663573403e-06,
+      "loss": 0.6837,
+      "step": 2461
+    },
+    {
+      "epoch": 0.7595249113064939,
+      "grad_norm": 0.0,
+      "learning_rate": 2.881385157300408e-06,
+      "loss": 0.7331,
+      "step": 2462
+    },
+    {
+      "epoch": 0.7598334104581213,
+      "grad_norm": 0.0,
+      "learning_rate": 2.8743686604841227e-06,
+      "loss": 0.7135,
+      "step": 2463
+    },
+    {
+      "epoch": 0.7601419096097486,
+      "grad_norm": 0.0,
+      "learning_rate": 2.8673592829187057e-06,
+      "loss": 0.765,
+      "step": 2464
+    },
+    {
+      "epoch": 0.7604504087613759,
+      "grad_norm": 0.0,
+      "learning_rate": 2.8603570316072484e-06,
+      "loss": 0.7272,
+      "step": 2465
+    },
+    {
+      "epoch": 0.7607589079130033,
+      "grad_norm": 0.0,
+      "learning_rate": 2.8533619135457402e-06,
+      "loss": 0.637,
+      "step": 2466
+    },
+    {
+      "epoch": 0.7610674070646306,
+      "grad_norm": 0.0,
+      "learning_rate": 2.846373935723029e-06,
+      "loss": 0.7201,
+      "step": 2467
+    },
+    {
+      "epoch": 0.761375906216258,
+      "grad_norm": 0.0,
+      "learning_rate": 2.839393105120839e-06,
+      "loss": 0.747,
+      "step": 2468
+    },
+    {
+      "epoch": 0.7616844053678853,
+      "grad_norm": 0.0,
+      "learning_rate": 2.8324194287137565e-06,
+      "loss": 0.7197,
+      "step": 2469
+    },
+    {
+      "epoch": 0.7619929045195126,
+      "grad_norm": 0.0,
+      "learning_rate": 2.8254529134692e-06,
+      "loss": 1.0135,
+      "step": 2470
+    },
+    {
+      "epoch": 0.7623014036711399,
+      "grad_norm": 0.0,
+      "learning_rate": 2.8184935663474567e-06,
+      "loss": 0.7235,
+      "step": 2471
+    },
+    {
+      "epoch": 0.7626099028227672,
+      "grad_norm": 0.0,
+      "learning_rate": 2.8115413943016366e-06,
+      "loss": 0.6625,
+      "step": 2472
+    },
+    {
+      "epoch": 0.7629184019743945,
+      "grad_norm": 0.0,
+      "learning_rate": 2.804596404277692e-06,
+      "loss": 0.7001,
+      "step": 2473
+    },
+    {
+      "epoch": 0.7632269011260219,
+      "grad_norm": 0.0,
+      "learning_rate": 2.7976586032143914e-06,
+      "loss": 0.6663,
+      "step": 2474
+    },
+    {
+      "epoch": 0.7635354002776492,
+      "grad_norm": 0.0,
+      "learning_rate": 2.7907279980433197e-06,
+      "loss": 0.7304,
+      "step": 2475
+    },
+    {
+      "epoch": 0.7638438994292766,
+      "grad_norm": 0.0,
+      "learning_rate": 2.7838045956888836e-06,
+      "loss": 0.7103,
+      "step": 2476
+    },
+    {
+      "epoch": 0.7641523985809039,
+      "grad_norm": 0.0,
+      "learning_rate": 2.776888403068282e-06,
+      "loss": 0.6845,
+      "step": 2477
+    },
+    {
+      "epoch": 0.7644608977325312,
+      "grad_norm": 0.0,
+      "learning_rate": 2.769979427091515e-06,
+      "loss": 0.6667,
+      "step": 2478
+    },
+    {
+      "epoch": 0.7647693968841586,
+      "grad_norm": 0.0,
+      "learning_rate": 2.7630776746613696e-06,
+      "loss": 0.6478,
+      "step": 2479
+    },
+    {
+      "epoch": 0.7650778960357859,
+      "grad_norm": 0.0,
+      "learning_rate": 2.7561831526734207e-06,
+      "loss": 0.7323,
+      "step": 2480
+    },
+    {
+      "epoch": 0.7653863951874132,
+      "grad_norm": 0.0,
+      "learning_rate": 2.749295868016022e-06,
+      "loss": 0.6891,
+      "step": 2481
+    },
+    {
+      "epoch": 0.7656948943390406,
+      "grad_norm": 0.0,
+      "learning_rate": 2.7424158275702807e-06,
+      "loss": 0.7104,
+      "step": 2482
+    },
+    {
+      "epoch": 0.7660033934906679,
+      "grad_norm": 0.0,
+      "learning_rate": 2.735543038210083e-06,
+      "loss": 0.6838,
+      "step": 2483
+    },
+    {
+      "epoch": 0.7663118926422953,
+      "grad_norm": 0.0,
+      "learning_rate": 2.728677506802062e-06,
+      "loss": 0.7601,
+      "step": 2484
+    },
+    {
+      "epoch": 0.7666203917939226,
+      "grad_norm": 0.0,
+      "learning_rate": 2.7218192402056052e-06,
+      "loss": 0.6113,
+      "step": 2485
+    },
+    {
+      "epoch": 0.76692889094555,
+      "grad_norm": 0.0,
+      "learning_rate": 2.7149682452728354e-06,
+      "loss": 0.6649,
+      "step": 2486
+    },
+    {
+      "epoch": 0.7672373900971773,
+      "grad_norm": 0.0,
+      "learning_rate": 2.70812452884861e-06,
+      "loss": 0.7328,
+      "step": 2487
+    },
+    {
+      "epoch": 0.7675458892488045,
+      "grad_norm": 0.0,
+      "learning_rate": 2.7012880977705236e-06,
+      "loss": 0.6515,
+      "step": 2488
+    },
+    {
+      "epoch": 0.7678543884004319,
+      "grad_norm": 0.0,
+      "learning_rate": 2.69445895886888e-06,
+      "loss": 0.6507,
+      "step": 2489
+    },
+    {
+      "epoch": 0.7681628875520592,
+      "grad_norm": 0.0,
+      "learning_rate": 2.6876371189667107e-06,
+      "loss": 1.0037,
+      "step": 2490
+    },
+    {
+      "epoch": 0.7684713867036865,
+      "grad_norm": 0.0,
+      "learning_rate": 2.6808225848797365e-06,
+      "loss": 0.7197,
+      "step": 2491
+    },
+    {
+      "epoch": 0.7687798858553139,
+      "grad_norm": 0.0,
+      "learning_rate": 2.6740153634163955e-06,
+      "loss": 0.6876,
+      "step": 2492
+    },
+    {
+      "epoch": 0.7690883850069412,
+      "grad_norm": 0.0,
+      "learning_rate": 2.6672154613778166e-06,
+      "loss": 0.7228,
+      "step": 2493
+    },
+    {
+      "epoch": 0.7693968841585686,
+      "grad_norm": 0.0,
+      "learning_rate": 2.6604228855578096e-06,
+      "loss": 0.6671,
+      "step": 2494
+    },
+    {
+      "epoch": 0.7697053833101959,
+      "grad_norm": 0.0,
+      "learning_rate": 2.653637642742869e-06,
+      "loss": 0.7387,
+      "step": 2495
+    },
+    {
+      "epoch": 0.7700138824618232,
+      "grad_norm": 0.0,
+      "learning_rate": 2.6468597397121575e-06,
+      "loss": 0.6826,
+      "step": 2496
+    },
+    {
+      "epoch": 0.7703223816134506,
+      "grad_norm": 0.0,
+      "learning_rate": 2.6400891832375163e-06,
+      "loss": 0.7781,
+      "step": 2497
+    },
+    {
+      "epoch": 0.7706308807650779,
+      "grad_norm": 0.0,
+      "learning_rate": 2.6333259800834366e-06,
+      "loss": 0.6769,
+      "step": 2498
+    },
+    {
+      "epoch": 0.7709393799167052,
+      "grad_norm": 0.0,
+      "learning_rate": 2.626570137007061e-06,
+      "loss": 0.6654,
+      "step": 2499
+    },
+    {
+      "epoch": 0.7712478790683326,
+      "grad_norm": 0.0,
+      "learning_rate": 2.6198216607581907e-06,
+      "loss": 0.6779,
+      "step": 2500
+    },
+    {
+      "epoch": 0.7715563782199599,
+      "grad_norm": 0.0,
+      "learning_rate": 2.613080558079253e-06,
+      "loss": 0.6879,
+      "step": 2501
+    },
+    {
+      "epoch": 0.7718648773715873,
+      "grad_norm": 0.0,
+      "learning_rate": 2.606346835705319e-06,
+      "loss": 0.7331,
+      "step": 2502
+    },
+    {
+      "epoch": 0.7721733765232146,
+      "grad_norm": 0.0,
+      "learning_rate": 2.5996205003640796e-06,
+      "loss": 0.7353,
+      "step": 2503
+    },
+    {
+      "epoch": 0.772481875674842,
+      "grad_norm": 0.0,
+      "learning_rate": 2.5929015587758455e-06,
+      "loss": 0.6572,
+      "step": 2504
+    },
+    {
+      "epoch": 0.7727903748264692,
+      "grad_norm": 0.0,
+      "learning_rate": 2.586190017653546e-06,
+      "loss": 0.7388,
+      "step": 2505
+    },
+    {
+      "epoch": 0.7730988739780965,
+      "grad_norm": 0.0,
+      "learning_rate": 2.579485883702707e-06,
+      "loss": 0.695,
+      "step": 2506
+    },
+    {
+      "epoch": 0.7734073731297239,
+      "grad_norm": 0.0,
+      "learning_rate": 2.5727891636214684e-06,
+      "loss": 0.6227,
+      "step": 2507
+    },
+    {
+      "epoch": 0.7737158722813512,
+      "grad_norm": 0.0,
+      "learning_rate": 2.566099864100543e-06,
+      "loss": 0.7622,
+      "step": 2508
+    },
+    {
+      "epoch": 0.7740243714329785,
+      "grad_norm": 0.0,
+      "learning_rate": 2.5594179918232476e-06,
+      "loss": 0.7167,
+      "step": 2509
+    },
+    {
+      "epoch": 0.7743328705846059,
+      "grad_norm": 0.0,
+      "learning_rate": 2.5527435534654667e-06,
+      "loss": 0.6815,
+      "step": 2510
+    },
+    {
+      "epoch": 0.7746413697362332,
+      "grad_norm": 0.0,
+      "learning_rate": 2.546076555695668e-06,
+      "loss": 0.6534,
+      "step": 2511
+    },
+    {
+      "epoch": 0.7749498688878605,
+      "grad_norm": 0.0,
+      "learning_rate": 2.539417005174876e-06,
+      "loss": 0.661,
+      "step": 2512
+    },
+    {
+      "epoch": 0.7752583680394879,
+      "grad_norm": 0.0,
+      "learning_rate": 2.5327649085566763e-06,
+      "loss": 0.6576,
+      "step": 2513
+    },
+    {
+      "epoch": 0.7755668671911152,
+      "grad_norm": 0.0,
+      "learning_rate": 2.5261202724872126e-06,
+      "loss": 0.7524,
+      "step": 2514
+    },
+    {
+      "epoch": 0.7758753663427426,
+      "grad_norm": 0.0,
+      "learning_rate": 2.5194831036051716e-06,
+      "loss": 0.6885,
+      "step": 2515
+    },
+    {
+      "epoch": 0.7761838654943699,
+      "grad_norm": 0.0,
+      "learning_rate": 2.5128534085417734e-06,
+      "loss": 0.6119,
+      "step": 2516
+    },
+    {
+      "epoch": 0.7764923646459972,
+      "grad_norm": 0.0,
+      "learning_rate": 2.506231193920783e-06,
+      "loss": 0.6358,
+      "step": 2517
+    },
+    {
+      "epoch": 0.7768008637976246,
+      "grad_norm": 0.0,
+      "learning_rate": 2.49961646635848e-06,
+      "loss": 0.684,
+      "step": 2518
+    },
+    {
+      "epoch": 0.7771093629492519,
+      "grad_norm": 0.0,
+      "learning_rate": 2.4930092324636744e-06,
+      "loss": 0.7587,
+      "step": 2519
+    },
+    {
+      "epoch": 0.7774178621008793,
+      "grad_norm": 0.0,
+      "learning_rate": 2.486409498837683e-06,
+      "loss": 0.7202,
+      "step": 2520
+    },
+    {
+      "epoch": 0.7777263612525066,
+      "grad_norm": 0.0,
+      "learning_rate": 2.4798172720743262e-06,
+      "loss": 0.9812,
+      "step": 2521
+    },
+    {
+      "epoch": 0.7780348604041339,
+      "grad_norm": 0.0,
+      "learning_rate": 2.473232558759927e-06,
+      "loss": 0.6815,
+      "step": 2522
+    },
+    {
+      "epoch": 0.7783433595557612,
+      "grad_norm": 0.0,
+      "learning_rate": 2.4666553654733095e-06,
+      "loss": 0.7473,
+      "step": 2523
+    },
+    {
+      "epoch": 0.7786518587073885,
+      "grad_norm": 0.0,
+      "learning_rate": 2.4600856987857725e-06,
+      "loss": 0.692,
+      "step": 2524
+    },
+    {
+      "epoch": 0.7789603578590159,
+      "grad_norm": 0.0,
+      "learning_rate": 2.4535235652610976e-06,
+      "loss": 0.7427,
+      "step": 2525
+    },
+    {
+      "epoch": 0.7792688570106432,
+      "grad_norm": 0.0,
+      "learning_rate": 2.44696897145555e-06,
+      "loss": 0.6775,
+      "step": 2526
+    },
+    {
+      "epoch": 0.7795773561622705,
+      "grad_norm": 0.0,
+      "learning_rate": 2.440421923917846e-06,
+      "loss": 0.63,
+      "step": 2527
+    },
+    {
+      "epoch": 0.7798858553138979,
+      "grad_norm": 0.0,
+      "learning_rate": 2.4338824291891795e-06,
+      "loss": 0.6496,
+      "step": 2528
+    },
+    {
+      "epoch": 0.7801943544655252,
+      "grad_norm": 0.0,
+      "learning_rate": 2.4273504938031835e-06,
+      "loss": 0.7412,
+      "step": 2529
+    },
+    {
+      "epoch": 0.7805028536171525,
+      "grad_norm": 0.0,
+      "learning_rate": 2.420826124285943e-06,
+      "loss": 0.797,
+      "step": 2530
+    },
+    {
+      "epoch": 0.7808113527687799,
+      "grad_norm": 0.0,
+      "learning_rate": 2.4143093271559925e-06,
+      "loss": 0.7071,
+      "step": 2531
+    },
+    {
+      "epoch": 0.7811198519204072,
+      "grad_norm": 0.0,
+      "learning_rate": 2.4078001089242887e-06,
+      "loss": 0.6181,
+      "step": 2532
+    },
+    {
+      "epoch": 0.7814283510720346,
+      "grad_norm": 0.0,
+      "learning_rate": 2.4012984760942224e-06,
+      "loss": 0.6362,
+      "step": 2533
+    },
+    {
+      "epoch": 0.7817368502236619,
+      "grad_norm": 0.0,
+      "learning_rate": 2.394804435161603e-06,
+      "loss": 0.6895,
+      "step": 2534
+    },
+    {
+      "epoch": 0.7820453493752892,
+      "grad_norm": 0.0,
+      "learning_rate": 2.3883179926146593e-06,
+      "loss": 0.6591,
+      "step": 2535
+    },
+    {
+      "epoch": 0.7823538485269166,
+      "grad_norm": 0.0,
+      "learning_rate": 2.3818391549340225e-06,
+      "loss": 0.7316,
+      "step": 2536
+    },
+    {
+      "epoch": 0.7826623476785439,
+      "grad_norm": 0.0,
+      "learning_rate": 2.3753679285927345e-06,
+      "loss": 0.6605,
+      "step": 2537
+    },
+    {
+      "epoch": 0.7829708468301713,
+      "grad_norm": 0.0,
+      "learning_rate": 2.368904320056222e-06,
+      "loss": 0.7423,
+      "step": 2538
+    },
+    {
+      "epoch": 0.7832793459817986,
+      "grad_norm": 0.0,
+      "learning_rate": 2.362448335782307e-06,
+      "loss": 0.6575,
+      "step": 2539
+    },
+    {
+      "epoch": 0.7835878451334258,
+      "grad_norm": 0.0,
+      "learning_rate": 2.3559999822211943e-06,
+      "loss": 0.7509,
+      "step": 2540
+    },
+    {
+      "epoch": 0.7838963442850532,
+      "grad_norm": 0.0,
+      "learning_rate": 2.349559265815463e-06,
+      "loss": 0.6073,
+      "step": 2541
+    },
+    {
+      "epoch": 0.7842048434366805,
+      "grad_norm": 0.0,
+      "learning_rate": 2.3431261930000602e-06,
+      "loss": 0.7227,
+      "step": 2542
+    },
+    {
+      "epoch": 0.7845133425883078,
+      "grad_norm": 0.0,
+      "learning_rate": 2.3367007702023016e-06,
+      "loss": 0.6742,
+      "step": 2543
+    },
+    {
+      "epoch": 0.7848218417399352,
+      "grad_norm": 0.0,
+      "learning_rate": 2.330283003841853e-06,
+      "loss": 0.8234,
+      "step": 2544
+    },
+    {
+      "epoch": 0.7851303408915625,
+      "grad_norm": 0.0,
+      "learning_rate": 2.3238729003307412e-06,
+      "loss": 0.9959,
+      "step": 2545
+    },
+    {
+      "epoch": 0.7854388400431899,
+      "grad_norm": 0.0,
+      "learning_rate": 2.3174704660733192e-06,
+      "loss": 0.7392,
+      "step": 2546
+    },
+    {
+      "epoch": 0.7857473391948172,
+      "grad_norm": 0.0,
+      "learning_rate": 2.311075707466297e-06,
+      "loss": 0.7615,
+      "step": 2547
+    },
+    {
+      "epoch": 0.7860558383464445,
+      "grad_norm": 0.0,
+      "learning_rate": 2.3046886308987017e-06,
+      "loss": 0.6454,
+      "step": 2548
+    },
+    {
+      "epoch": 0.7863643374980719,
+      "grad_norm": 0.0,
+      "learning_rate": 2.298309242751896e-06,
+      "loss": 0.6645,
+      "step": 2549
+    },
+    {
+      "epoch": 0.7866728366496992,
+      "grad_norm": 0.0,
+      "learning_rate": 2.291937549399552e-06,
+      "loss": 0.7703,
+      "step": 2550
+    },
+    {
+      "epoch": 0.7869813358013266,
+      "grad_norm": 0.0,
+      "learning_rate": 2.2855735572076564e-06,
+      "loss": 0.6995,
+      "step": 2551
+    },
+    {
+      "epoch": 0.7872898349529539,
+      "grad_norm": 0.0,
+      "learning_rate": 2.2792172725345084e-06,
+      "loss": 0.6252,
+      "step": 2552
+    },
+    {
+      "epoch": 0.7875983341045812,
+      "grad_norm": 0.0,
+      "learning_rate": 2.272868701730695e-06,
+      "loss": 0.6929,
+      "step": 2553
+    },
+    {
+      "epoch": 0.7879068332562086,
+      "grad_norm": 0.0,
+      "learning_rate": 2.2665278511391075e-06,
+      "loss": 0.6335,
+      "step": 2554
+    },
+    {
+      "epoch": 0.7882153324078359,
+      "grad_norm": 0.0,
+      "learning_rate": 2.2601947270949164e-06,
+      "loss": 0.7216,
+      "step": 2555
+    },
+    {
+      "epoch": 0.7885238315594633,
+      "grad_norm": 0.0,
+      "learning_rate": 2.2538693359255724e-06,
+      "loss": 0.7121,
+      "step": 2556
+    },
+    {
+      "epoch": 0.7888323307110905,
+      "grad_norm": 0.0,
+      "learning_rate": 2.2475516839508072e-06,
+      "loss": 0.6249,
+      "step": 2557
+    },
+    {
+      "epoch": 0.7891408298627178,
+      "grad_norm": 0.0,
+      "learning_rate": 2.2412417774826144e-06,
+      "loss": 0.6763,
+      "step": 2558
+    },
+    {
+      "epoch": 0.7894493290143452,
+      "grad_norm": 0.0,
+      "learning_rate": 2.2349396228252506e-06,
+      "loss": 0.7082,
+      "step": 2559
+    },
+    {
+      "epoch": 0.7897578281659725,
+      "grad_norm": 0.0,
+      "learning_rate": 2.2286452262752223e-06,
+      "loss": 0.7742,
+      "step": 2560
+    },
+    {
+      "epoch": 0.7900663273175998,
+      "grad_norm": 0.0,
+      "learning_rate": 2.2223585941212956e-06,
+      "loss": 0.6681,
+      "step": 2561
+    },
+    {
+      "epoch": 0.7903748264692272,
+      "grad_norm": 0.0,
+      "learning_rate": 2.216079732644476e-06,
+      "loss": 0.7311,
+      "step": 2562
+    },
+    {
+      "epoch": 0.7906833256208545,
+      "grad_norm": 0.0,
+      "learning_rate": 2.2098086481179947e-06,
+      "loss": 0.6645,
+      "step": 2563
+    },
+    {
+      "epoch": 0.7909918247724819,
+      "grad_norm": 0.0,
+      "learning_rate": 2.2035453468073263e-06,
+      "loss": 0.557,
+      "step": 2564
+    },
+    {
+      "epoch": 0.7913003239241092,
+      "grad_norm": 0.0,
+      "learning_rate": 2.197289834970161e-06,
+      "loss": 0.6466,
+      "step": 2565
+    },
+    {
+      "epoch": 0.7916088230757365,
+      "grad_norm": 0.0,
+      "learning_rate": 2.1910421188564134e-06,
+      "loss": 0.6967,
+      "step": 2566
+    },
+    {
+      "epoch": 0.7919173222273639,
+      "grad_norm": 0.0,
+      "learning_rate": 2.184802204708202e-06,
+      "loss": 0.7423,
+      "step": 2567
+    },
+    {
+      "epoch": 0.7922258213789912,
+      "grad_norm": 0.0,
+      "learning_rate": 2.178570098759851e-06,
+      "loss": 0.7138,
+      "step": 2568
+    },
+    {
+      "epoch": 0.7925343205306186,
+      "grad_norm": 0.0,
+      "learning_rate": 2.1723458072378924e-06,
+      "loss": 0.6411,
+      "step": 2569
+    },
+    {
+      "epoch": 0.7928428196822459,
+      "grad_norm": 0.0,
+      "learning_rate": 2.166129336361038e-06,
+      "loss": 0.8109,
+      "step": 2570
+    },
+    {
+      "epoch": 0.7931513188338732,
+      "grad_norm": 0.0,
+      "learning_rate": 2.159920692340202e-06,
+      "loss": 0.7287,
+      "step": 2571
+    },
+    {
+      "epoch": 0.7934598179855006,
+      "grad_norm": 0.0,
+      "learning_rate": 2.1537198813784554e-06,
+      "loss": 0.7112,
+      "step": 2572
+    },
+    {
+      "epoch": 0.7937683171371279,
+      "grad_norm": 0.0,
+      "learning_rate": 2.147526909671064e-06,
+      "loss": 0.6677,
+      "step": 2573
+    },
+    {
+      "epoch": 0.7940768162887553,
+      "grad_norm": 0.0,
+      "learning_rate": 2.141341783405457e-06,
+      "loss": 0.7404,
+      "step": 2574
+    },
+    {
+      "epoch": 0.7943853154403825,
+      "grad_norm": 0.0,
+      "learning_rate": 2.1351645087612195e-06,
+      "loss": 0.7257,
+      "step": 2575
+    },
+    {
+      "epoch": 0.7946938145920098,
+      "grad_norm": 0.0,
+      "learning_rate": 2.1289950919100932e-06,
+      "loss": 0.7316,
+      "step": 2576
+    },
+    {
+      "epoch": 0.7950023137436372,
+      "grad_norm": 0.0,
+      "learning_rate": 2.1228335390159704e-06,
+      "loss": 0.7996,
+      "step": 2577
+    },
+    {
+      "epoch": 0.7953108128952645,
+      "grad_norm": 0.0,
+      "learning_rate": 2.116679856234892e-06,
+      "loss": 0.7102,
+      "step": 2578
+    },
+    {
+      "epoch": 0.7956193120468918,
+      "grad_norm": 0.0,
+      "learning_rate": 2.1105340497150264e-06,
+      "loss": 0.7584,
+      "step": 2579
+    },
+    {
+      "epoch": 0.7959278111985192,
+      "grad_norm": 0.0,
+      "learning_rate": 2.104396125596675e-06,
+      "loss": 0.7088,
+      "step": 2580
+    },
+    {
+      "epoch": 0.7962363103501465,
+      "grad_norm": 0.0,
+      "learning_rate": 2.098266090012271e-06,
+      "loss": 0.6948,
+      "step": 2581
+    },
+    {
+      "epoch": 0.7965448095017739,
+      "grad_norm": 0.0,
+      "learning_rate": 2.092143949086356e-06,
+      "loss": 0.6375,
+      "step": 2582
+    },
+    {
+      "epoch": 0.7968533086534012,
+      "grad_norm": 0.0,
+      "learning_rate": 2.0860297089355943e-06,
+      "loss": 0.7181,
+      "step": 2583
+    },
+    {
+      "epoch": 0.7971618078050285,
+      "grad_norm": 0.0,
+      "learning_rate": 2.0799233756687477e-06,
+      "loss": 0.7211,
+      "step": 2584
+    },
+    {
+      "epoch": 0.7974703069566559,
+      "grad_norm": 0.0,
+      "learning_rate": 2.0738249553866807e-06,
+      "loss": 0.6163,
+      "step": 2585
+    },
+    {
+      "epoch": 0.7977788061082832,
+      "grad_norm": 0.0,
+      "learning_rate": 2.0677344541823573e-06,
+      "loss": 0.7215,
+      "step": 2586
+    },
+    {
+      "epoch": 0.7980873052599106,
+      "grad_norm": 0.0,
+      "learning_rate": 2.061651878140819e-06,
+      "loss": 0.7583,
+      "step": 2587
+    },
+    {
+      "epoch": 0.7983958044115379,
+      "grad_norm": 0.0,
+      "learning_rate": 2.0555772333392055e-06,
+      "loss": 0.7149,
+      "step": 2588
+    },
+    {
+      "epoch": 0.7987043035631652,
+      "grad_norm": 0.0,
+      "learning_rate": 2.0495105258467085e-06,
+      "loss": 0.7232,
+      "step": 2589
+    },
+    {
+      "epoch": 0.7990128027147926,
+      "grad_norm": 0.0,
+      "learning_rate": 2.043451761724614e-06,
+      "loss": 0.7209,
+      "step": 2590
+    },
+    {
+      "epoch": 0.7993213018664199,
+      "grad_norm": 0.0,
+      "learning_rate": 2.0374009470262545e-06,
+      "loss": 0.7242,
+      "step": 2591
+    },
+    {
+      "epoch": 0.7996298010180471,
+      "grad_norm": 0.0,
+      "learning_rate": 2.031358087797032e-06,
+      "loss": 0.7512,
+      "step": 2592
+    },
+    {
+      "epoch": 0.7999383001696745,
+      "grad_norm": 0.0,
+      "learning_rate": 2.0253231900743906e-06,
+      "loss": 0.6766,
+      "step": 2593
+    },
+    {
+      "epoch": 0.8002467993213018,
+      "grad_norm": 0.0,
+      "learning_rate": 2.019296259887825e-06,
+      "loss": 0.7268,
+      "step": 2594
+    },
+    {
+      "epoch": 0.8005552984729292,
+      "grad_norm": 0.0,
+      "learning_rate": 2.0132773032588717e-06,
+      "loss": 0.7501,
+      "step": 2595
+    },
+    {
+      "epoch": 0.8008637976245565,
+      "grad_norm": 0.0,
+      "learning_rate": 2.007266326201095e-06,
+      "loss": 0.7455,
+      "step": 2596
+    },
+    {
+      "epoch": 0.8011722967761838,
+      "grad_norm": 0.0,
+      "learning_rate": 2.0012633347200873e-06,
+      "loss": 0.6826,
+      "step": 2597
+    },
+    {
+      "epoch": 0.8014807959278112,
+      "grad_norm": 0.0,
+      "learning_rate": 1.9952683348134704e-06,
+      "loss": 0.6854,
+      "step": 2598
+    },
+    {
+      "epoch": 0.8017892950794385,
+      "grad_norm": 0.0,
+      "learning_rate": 1.9892813324708705e-06,
+      "loss": 0.7561,
+      "step": 2599
+    },
+    {
+      "epoch": 0.8020977942310659,
+      "grad_norm": 0.0,
+      "learning_rate": 1.9833023336739354e-06,
+      "loss": 0.654,
+      "step": 2600
+    },
+    {
+      "epoch": 0.8024062933826932,
+      "grad_norm": 0.0,
+      "learning_rate": 1.977331344396306e-06,
+      "loss": 0.6279,
+      "step": 2601
+    },
+    {
+      "epoch": 0.8027147925343205,
+      "grad_norm": 0.0,
+      "learning_rate": 1.971368370603626e-06,
+      "loss": 0.7695,
+      "step": 2602
+    },
+    {
+      "epoch": 0.8030232916859479,
+      "grad_norm": 0.0,
+      "learning_rate": 1.9654134182535267e-06,
+      "loss": 0.7384,
+      "step": 2603
+    },
+    {
+      "epoch": 0.8033317908375752,
+      "grad_norm": 0.0,
+      "learning_rate": 1.959466493295633e-06,
+      "loss": 0.6961,
+      "step": 2604
+    },
+    {
+      "epoch": 0.8036402899892026,
+      "grad_norm": 0.0,
+      "learning_rate": 1.953527601671543e-06,
+      "loss": 0.7267,
+      "step": 2605
+    },
+    {
+      "epoch": 0.8039487891408299,
+      "grad_norm": 0.0,
+      "learning_rate": 1.9475967493148283e-06,
+      "loss": 0.6406,
+      "step": 2606
+    },
+    {
+      "epoch": 0.8042572882924572,
+      "grad_norm": 0.0,
+      "learning_rate": 1.9416739421510354e-06,
+      "loss": 0.674,
+      "step": 2607
+    },
+    {
+      "epoch": 0.8045657874440846,
+      "grad_norm": 0.0,
+      "learning_rate": 1.935759186097662e-06,
+      "loss": 0.63,
+      "step": 2608
+    },
+    {
+      "epoch": 0.8048742865957118,
+      "grad_norm": 0.0,
+      "learning_rate": 1.9298524870641734e-06,
+      "loss": 0.712,
+      "step": 2609
+    },
+    {
+      "epoch": 0.8051827857473391,
+      "grad_norm": 0.0,
+      "learning_rate": 1.9239538509519784e-06,
+      "loss": 0.7013,
+      "step": 2610
+    },
+    {
+      "epoch": 0.8054912848989665,
+      "grad_norm": 0.0,
+      "learning_rate": 1.9180632836544268e-06,
+      "loss": 0.7126,
+      "step": 2611
+    },
+    {
+      "epoch": 0.8057997840505938,
+      "grad_norm": 0.0,
+      "learning_rate": 1.912180791056817e-06,
+      "loss": 0.6296,
+      "step": 2612
+    },
+    {
+      "epoch": 0.8061082832022212,
+      "grad_norm": 0.0,
+      "learning_rate": 1.9063063790363733e-06,
+      "loss": 0.6604,
+      "step": 2613
+    },
+    {
+      "epoch": 0.8064167823538485,
+      "grad_norm": 0.0,
+      "learning_rate": 1.9004400534622458e-06,
+      "loss": 0.7148,
+      "step": 2614
+    },
+    {
+      "epoch": 0.8067252815054758,
+      "grad_norm": 0.0,
+      "learning_rate": 1.8945818201955047e-06,
+      "loss": 0.7608,
+      "step": 2615
+    },
+    {
+      "epoch": 0.8070337806571032,
+      "grad_norm": 0.0,
+      "learning_rate": 1.8887316850891446e-06,
+      "loss": 0.6483,
+      "step": 2616
+    },
+    {
+      "epoch": 0.8073422798087305,
+      "grad_norm": 0.0,
+      "learning_rate": 1.8828896539880549e-06,
+      "loss": 0.755,
+      "step": 2617
+    },
+    {
+      "epoch": 0.8076507789603579,
+      "grad_norm": 0.0,
+      "learning_rate": 1.8770557327290418e-06,
+      "loss": 0.7258,
+      "step": 2618
+    },
+    {
+      "epoch": 0.8079592781119852,
+      "grad_norm": 0.0,
+      "learning_rate": 1.8712299271407998e-06,
+      "loss": 0.597,
+      "step": 2619
+    },
+    {
+      "epoch": 0.8082677772636125,
+      "grad_norm": 0.0,
+      "learning_rate": 1.865412243043916e-06,
+      "loss": 0.7922,
+      "step": 2620
+    },
+    {
+      "epoch": 0.8085762764152399,
+      "grad_norm": 0.0,
+      "learning_rate": 1.8596026862508697e-06,
+      "loss": 0.7384,
+      "step": 2621
+    },
+    {
+      "epoch": 0.8088847755668672,
+      "grad_norm": 0.0,
+      "learning_rate": 1.853801262566014e-06,
+      "loss": 0.6699,
+      "step": 2622
+    },
+    {
+      "epoch": 0.8091932747184946,
+      "grad_norm": 0.0,
+      "learning_rate": 1.848007977785573e-06,
+      "loss": 0.6537,
+      "step": 2623
+    },
+    {
+      "epoch": 0.8095017738701219,
+      "grad_norm": 0.0,
+      "learning_rate": 1.8422228376976526e-06,
+      "loss": 0.6377,
+      "step": 2624
+    },
+    {
+      "epoch": 0.8098102730217492,
+      "grad_norm": 0.0,
+      "learning_rate": 1.8364458480822045e-06,
+      "loss": 0.7561,
+      "step": 2625
+    },
+    {
+      "epoch": 0.8101187721733766,
+      "grad_norm": 0.0,
+      "learning_rate": 1.8306770147110542e-06,
+      "loss": 0.6598,
+      "step": 2626
+    },
+    {
+      "epoch": 0.8104272713250038,
+      "grad_norm": 0.0,
+      "learning_rate": 1.8249163433478601e-06,
+      "loss": 0.665,
+      "step": 2627
+    },
+    {
+      "epoch": 0.8107357704766311,
+      "grad_norm": 0.0,
+      "learning_rate": 1.8191638397481415e-06,
+      "loss": 0.7308,
+      "step": 2628
+    },
+    {
+      "epoch": 0.8110442696282585,
+      "grad_norm": 0.0,
+      "learning_rate": 1.8134195096592466e-06,
+      "loss": 0.7206,
+      "step": 2629
+    },
+    {
+      "epoch": 0.8113527687798858,
+      "grad_norm": 0.0,
+      "learning_rate": 1.807683358820368e-06,
+      "loss": 0.722,
+      "step": 2630
+    },
+    {
+      "epoch": 0.8116612679315132,
+      "grad_norm": 0.0,
+      "learning_rate": 1.8019553929625166e-06,
+      "loss": 0.6939,
+      "step": 2631
+    },
+    {
+      "epoch": 0.8119697670831405,
+      "grad_norm": 0.0,
+      "learning_rate": 1.796235617808526e-06,
+      "loss": 0.6726,
+      "step": 2632
+    },
+    {
+      "epoch": 0.8122782662347678,
+      "grad_norm": 0.0,
+      "learning_rate": 1.7905240390730571e-06,
+      "loss": 0.6318,
+      "step": 2633
+    },
+    {
+      "epoch": 0.8125867653863952,
+      "grad_norm": 0.0,
+      "learning_rate": 1.7848206624625675e-06,
+      "loss": 0.7716,
+      "step": 2634
+    },
+    {
+      "epoch": 0.8128952645380225,
+      "grad_norm": 0.0,
+      "learning_rate": 1.7791254936753323e-06,
+      "loss": 0.6721,
+      "step": 2635
+    },
+    {
+      "epoch": 0.8132037636896499,
+      "grad_norm": 0.0,
+      "learning_rate": 1.7734385384014175e-06,
+      "loss": 0.6664,
+      "step": 2636
+    },
+    {
+      "epoch": 0.8135122628412772,
+      "grad_norm": 0.0,
+      "learning_rate": 1.767759802322685e-06,
+      "loss": 0.6266,
+      "step": 2637
+    },
+    {
+      "epoch": 0.8138207619929045,
+      "grad_norm": 0.0,
+      "learning_rate": 1.7620892911127897e-06,
+      "loss": 0.6798,
+      "step": 2638
+    },
+    {
+      "epoch": 0.8141292611445319,
+      "grad_norm": 0.0,
+      "learning_rate": 1.7564270104371638e-06,
+      "loss": 0.6637,
+      "step": 2639
+    },
+    {
+      "epoch": 0.8144377602961592,
+      "grad_norm": 0.0,
+      "learning_rate": 1.750772965953017e-06,
+      "loss": 0.7297,
+      "step": 2640
+    },
+    {
+      "epoch": 0.8147462594477866,
+      "grad_norm": 0.0,
+      "learning_rate": 1.7451271633093292e-06,
+      "loss": 0.6687,
+      "step": 2641
+    },
+    {
+      "epoch": 0.8150547585994139,
+      "grad_norm": 0.0,
+      "learning_rate": 1.7394896081468527e-06,
+      "loss": 0.7067,
+      "step": 2642
+    },
+    {
+      "epoch": 0.8153632577510412,
+      "grad_norm": 0.0,
+      "learning_rate": 1.733860306098092e-06,
+      "loss": 1.034,
+      "step": 2643
+    },
+    {
+      "epoch": 0.8156717569026685,
+      "grad_norm": 0.0,
+      "learning_rate": 1.7282392627873046e-06,
+      "loss": 0.6971,
+      "step": 2644
+    },
+    {
+      "epoch": 0.8159802560542958,
+      "grad_norm": 0.0,
+      "learning_rate": 1.7226264838305074e-06,
+      "loss": 0.7338,
+      "step": 2645
+    },
+    {
+      "epoch": 0.8162887552059231,
+      "grad_norm": 0.0,
+      "learning_rate": 1.7170219748354488e-06,
+      "loss": 0.7389,
+      "step": 2646
+    },
+    {
+      "epoch": 0.8165972543575505,
+      "grad_norm": 0.0,
+      "learning_rate": 1.7114257414016223e-06,
+      "loss": 0.7335,
+      "step": 2647
+    },
+    {
+      "epoch": 0.8169057535091778,
+      "grad_norm": 0.0,
+      "learning_rate": 1.7058377891202493e-06,
+      "loss": 0.6609,
+      "step": 2648
+    },
+    {
+      "epoch": 0.8172142526608052,
+      "grad_norm": 0.0,
+      "learning_rate": 1.7002581235742766e-06,
+      "loss": 0.6985,
+      "step": 2649
+    },
+    {
+      "epoch": 0.8175227518124325,
+      "grad_norm": 0.0,
+      "learning_rate": 1.6946867503383768e-06,
+      "loss": 0.7653,
+      "step": 2650
+    },
+    {
+      "epoch": 0.8178312509640598,
+      "grad_norm": 0.0,
+      "learning_rate": 1.6891236749789297e-06,
+      "loss": 0.7605,
+      "step": 2651
+    },
+    {
+      "epoch": 0.8181397501156872,
+      "grad_norm": 0.0,
+      "learning_rate": 1.6835689030540382e-06,
+      "loss": 0.6845,
+      "step": 2652
+    },
+    {
+      "epoch": 0.8184482492673145,
+      "grad_norm": 0.0,
+      "learning_rate": 1.6780224401134903e-06,
+      "loss": 0.687,
+      "step": 2653
+    },
+    {
+      "epoch": 0.8187567484189419,
+      "grad_norm": 0.0,
+      "learning_rate": 1.6724842916987882e-06,
+      "loss": 0.6749,
+      "step": 2654
+    },
+    {
+      "epoch": 0.8190652475705692,
+      "grad_norm": 0.0,
+      "learning_rate": 1.6669544633431178e-06,
+      "loss": 0.6974,
+      "step": 2655
+    },
+    {
+      "epoch": 0.8193737467221965,
+      "grad_norm": 0.0,
+      "learning_rate": 1.6614329605713597e-06,
+      "loss": 0.6402,
+      "step": 2656
+    },
+    {
+      "epoch": 0.8196822458738239,
+      "grad_norm": 0.0,
+      "learning_rate": 1.655919788900071e-06,
+      "loss": 0.6724,
+      "step": 2657
+    },
+    {
+      "epoch": 0.8199907450254512,
+      "grad_norm": 0.0,
+      "learning_rate": 1.6504149538374848e-06,
+      "loss": 0.7253,
+      "step": 2658
+    },
+    {
+      "epoch": 0.8202992441770786,
+      "grad_norm": 0.0,
+      "learning_rate": 1.6449184608835112e-06,
+      "loss": 0.7499,
+      "step": 2659
+    },
+    {
+      "epoch": 0.8206077433287059,
+      "grad_norm": 0.0,
+      "learning_rate": 1.6394303155297186e-06,
+      "loss": 0.6714,
+      "step": 2660
+    },
+    {
+      "epoch": 0.8209162424803331,
+      "grad_norm": 0.0,
+      "learning_rate": 1.6339505232593356e-06,
+      "loss": 0.6722,
+      "step": 2661
+    },
+    {
+      "epoch": 0.8212247416319605,
+      "grad_norm": 0.0,
+      "learning_rate": 1.6284790895472536e-06,
+      "loss": 0.7015,
+      "step": 2662
+    },
+    {
+      "epoch": 0.8215332407835878,
+      "grad_norm": 0.0,
+      "learning_rate": 1.623016019860002e-06,
+      "loss": 0.6303,
+      "step": 2663
+    },
+    {
+      "epoch": 0.8218417399352151,
+      "grad_norm": 0.0,
+      "learning_rate": 1.6175613196557648e-06,
+      "loss": 0.7468,
+      "step": 2664
+    },
+    {
+      "epoch": 0.8221502390868425,
+      "grad_norm": 0.0,
+      "learning_rate": 1.6121149943843494e-06,
+      "loss": 0.6882,
+      "step": 2665
+    },
+    {
+      "epoch": 0.8224587382384698,
+      "grad_norm": 0.0,
+      "learning_rate": 1.606677049487212e-06,
+      "loss": 0.6778,
+      "step": 2666
+    },
+    {
+      "epoch": 0.8227672373900972,
+      "grad_norm": 0.0,
+      "learning_rate": 1.6012474903974218e-06,
+      "loss": 0.6876,
+      "step": 2667
+    },
+    {
+      "epoch": 0.8230757365417245,
+      "grad_norm": 0.0,
+      "learning_rate": 1.5958263225396796e-06,
+      "loss": 0.6676,
+      "step": 2668
+    },
+    {
+      "epoch": 0.8233842356933518,
+      "grad_norm": 0.0,
+      "learning_rate": 1.5904135513303032e-06,
+      "loss": 0.6778,
+      "step": 2669
+    },
+    {
+      "epoch": 0.8236927348449792,
+      "grad_norm": 0.0,
+      "learning_rate": 1.5850091821772085e-06,
+      "loss": 0.7142,
+      "step": 2670
+    },
+    {
+      "epoch": 0.8240012339966065,
+      "grad_norm": 0.0,
+      "learning_rate": 1.579613220479932e-06,
+      "loss": 0.6722,
+      "step": 2671
+    },
+    {
+      "epoch": 0.8243097331482339,
+      "grad_norm": 0.0,
+      "learning_rate": 1.5742256716295978e-06,
+      "loss": 0.657,
+      "step": 2672
+    },
+    {
+      "epoch": 0.8246182322998612,
+      "grad_norm": 0.0,
+      "learning_rate": 1.5688465410089371e-06,
+      "loss": 0.7935,
+      "step": 2673
+    },
+    {
+      "epoch": 0.8249267314514885,
+      "grad_norm": 0.0,
+      "learning_rate": 1.5634758339922619e-06,
+      "loss": 0.7298,
+      "step": 2674
+    },
+    {
+      "epoch": 0.8252352306031159,
+      "grad_norm": 0.0,
+      "learning_rate": 1.5581135559454653e-06,
+      "loss": 0.6653,
+      "step": 2675
+    },
+    {
+      "epoch": 0.8255437297547432,
+      "grad_norm": 0.0,
+      "learning_rate": 1.5527597122260307e-06,
+      "loss": 0.6033,
+      "step": 2676
+    },
+    {
+      "epoch": 0.8258522289063706,
+      "grad_norm": 0.0,
+      "learning_rate": 1.547414308183005e-06,
+      "loss": 0.6975,
+      "step": 2677
+    },
+    {
+      "epoch": 0.8261607280579979,
+      "grad_norm": 0.0,
+      "learning_rate": 1.542077349157005e-06,
+      "loss": 0.6166,
+      "step": 2678
+    },
+    {
+      "epoch": 0.8264692272096251,
+      "grad_norm": 0.0,
+      "learning_rate": 1.536748840480209e-06,
+      "loss": 0.7719,
+      "step": 2679
+    },
+    {
+      "epoch": 0.8267777263612525,
+      "grad_norm": 0.0,
+      "learning_rate": 1.5314287874763578e-06,
+      "loss": 0.7122,
+      "step": 2680
+    },
+    {
+      "epoch": 0.8270862255128798,
+      "grad_norm": 0.0,
+      "learning_rate": 1.5261171954607435e-06,
+      "loss": 0.7992,
+      "step": 2681
+    },
+    {
+      "epoch": 0.8273947246645071,
+      "grad_norm": 0.0,
+      "learning_rate": 1.5208140697401942e-06,
+      "loss": 0.7702,
+      "step": 2682
+    },
+    {
+      "epoch": 0.8277032238161345,
+      "grad_norm": 0.0,
+      "learning_rate": 1.5155194156130936e-06,
+      "loss": 0.6515,
+      "step": 2683
+    },
+    {
+      "epoch": 0.8280117229677618,
+      "grad_norm": 0.0,
+      "learning_rate": 1.5102332383693496e-06,
+      "loss": 0.6377,
+      "step": 2684
+    },
+    {
+      "epoch": 0.8283202221193892,
+      "grad_norm": 0.0,
+      "learning_rate": 1.5049555432904118e-06,
+      "loss": 0.6163,
+      "step": 2685
+    },
+    {
+      "epoch": 0.8286287212710165,
+      "grad_norm": 0.0,
+      "learning_rate": 1.4996863356492464e-06,
+      "loss": 0.6824,
+      "step": 2686
+    },
+    {
+      "epoch": 0.8289372204226438,
+      "grad_norm": 0.0,
+      "learning_rate": 1.4944256207103403e-06,
+      "loss": 0.734,
+      "step": 2687
+    },
+    {
+      "epoch": 0.8292457195742712,
+      "grad_norm": 0.0,
+      "learning_rate": 1.4891734037297035e-06,
+      "loss": 0.7307,
+      "step": 2688
+    },
+    {
+      "epoch": 0.8295542187258985,
+      "grad_norm": 0.0,
+      "learning_rate": 1.4839296899548438e-06,
+      "loss": 0.6306,
+      "step": 2689
+    },
+    {
+      "epoch": 0.8298627178775259,
+      "grad_norm": 0.0,
+      "learning_rate": 1.4786944846247887e-06,
+      "loss": 0.7072,
+      "step": 2690
+    },
+    {
+      "epoch": 0.8301712170291532,
+      "grad_norm": 0.0,
+      "learning_rate": 1.4734677929700435e-06,
+      "loss": 0.7571,
+      "step": 2691
+    },
+    {
+      "epoch": 0.8304797161807805,
+      "grad_norm": 0.0,
+      "learning_rate": 1.4682496202126283e-06,
+      "loss": 0.723,
+      "step": 2692
+    },
+    {
+      "epoch": 0.8307882153324079,
+      "grad_norm": 0.0,
+      "learning_rate": 1.4630399715660425e-06,
+      "loss": 0.7182,
+      "step": 2693
+    },
+    {
+      "epoch": 0.8310967144840352,
+      "grad_norm": 0.0,
+      "learning_rate": 1.4578388522352682e-06,
+      "loss": 0.623,
+      "step": 2694
+    },
+    {
+      "epoch": 0.8314052136356626,
+      "grad_norm": 0.0,
+      "learning_rate": 1.4526462674167685e-06,
+      "loss": 0.6567,
+      "step": 2695
+    },
+    {
+      "epoch": 0.8317137127872898,
+      "grad_norm": 0.0,
+      "learning_rate": 1.447462222298477e-06,
+      "loss": 0.662,
+      "step": 2696
+    },
+    {
+      "epoch": 0.8320222119389171,
+      "grad_norm": 0.0,
+      "learning_rate": 1.442286722059799e-06,
+      "loss": 0.7251,
+      "step": 2697
+    },
+    {
+      "epoch": 0.8323307110905445,
+      "grad_norm": 0.0,
+      "learning_rate": 1.4371197718715967e-06,
+      "loss": 0.6979,
+      "step": 2698
+    },
+    {
+      "epoch": 0.8326392102421718,
+      "grad_norm": 0.0,
+      "learning_rate": 1.4319613768962003e-06,
+      "loss": 0.6503,
+      "step": 2699
+    },
+    {
+      "epoch": 0.8329477093937991,
+      "grad_norm": 0.0,
+      "learning_rate": 1.4268115422873807e-06,
+      "loss": 0.7083,
+      "step": 2700
+    },
+    {
+      "epoch": 0.8332562085454265,
+      "grad_norm": 0.0,
+      "learning_rate": 1.4216702731903608e-06,
+      "loss": 0.6817,
+      "step": 2701
+    },
+    {
+      "epoch": 0.8335647076970538,
+      "grad_norm": 0.0,
+      "learning_rate": 1.4165375747418097e-06,
+      "loss": 0.7009,
+      "step": 2702
+    },
+    {
+      "epoch": 0.8338732068486812,
+      "grad_norm": 0.0,
+      "learning_rate": 1.4114134520698286e-06,
+      "loss": 0.7503,
+      "step": 2703
+    },
+    {
+      "epoch": 0.8341817060003085,
+      "grad_norm": 0.0,
+      "learning_rate": 1.4062979102939478e-06,
+      "loss": 0.6682,
+      "step": 2704
+    },
+    {
+      "epoch": 0.8344902051519358,
+      "grad_norm": 0.0,
+      "learning_rate": 1.4011909545251357e-06,
+      "loss": 0.6859,
+      "step": 2705
+    },
+    {
+      "epoch": 0.8347987043035632,
+      "grad_norm": 0.0,
+      "learning_rate": 1.3960925898657674e-06,
+      "loss": 0.7062,
+      "step": 2706
+    },
+    {
+      "epoch": 0.8351072034551905,
+      "grad_norm": 0.0,
+      "learning_rate": 1.3910028214096517e-06,
+      "loss": 0.6987,
+      "step": 2707
+    },
+    {
+      "epoch": 0.8354157026068179,
+      "grad_norm": 0.0,
+      "learning_rate": 1.3859216542419907e-06,
+      "loss": 0.6976,
+      "step": 2708
+    },
+    {
+      "epoch": 0.8357242017584452,
+      "grad_norm": 0.0,
+      "learning_rate": 1.3808490934394058e-06,
+      "loss": 0.7817,
+      "step": 2709
+    },
+    {
+      "epoch": 0.8360327009100725,
+      "grad_norm": 0.0,
+      "learning_rate": 1.3757851440699132e-06,
+      "loss": 0.6481,
+      "step": 2710
+    },
+    {
+      "epoch": 0.8363412000616999,
+      "grad_norm": 0.0,
+      "learning_rate": 1.3707298111929335e-06,
+      "loss": 0.7614,
+      "step": 2711
+    },
+    {
+      "epoch": 0.8366496992133272,
+      "grad_norm": 0.0,
+      "learning_rate": 1.3656830998592674e-06,
+      "loss": 0.6758,
+      "step": 2712
+    },
+    {
+      "epoch": 0.8369581983649544,
+      "grad_norm": 0.0,
+      "learning_rate": 1.3606450151111083e-06,
+      "loss": 0.6121,
+      "step": 2713
+    },
+    {
+      "epoch": 0.8372666975165818,
+      "grad_norm": 0.0,
+      "learning_rate": 1.3556155619820322e-06,
+      "loss": 0.6688,
+      "step": 2714
+    },
+    {
+      "epoch": 0.8375751966682091,
+      "grad_norm": 0.0,
+      "learning_rate": 1.3505947454969849e-06,
+      "loss": 0.6738,
+      "step": 2715
+    },
+    {
+      "epoch": 0.8378836958198365,
+      "grad_norm": 0.0,
+      "learning_rate": 1.3455825706722913e-06,
+      "loss": 0.6668,
+      "step": 2716
+    },
+    {
+      "epoch": 0.8381921949714638,
+      "grad_norm": 0.0,
+      "learning_rate": 1.340579042515636e-06,
+      "loss": 0.7641,
+      "step": 2717
+    },
+    {
+      "epoch": 0.8385006941230911,
+      "grad_norm": 0.0,
+      "learning_rate": 1.3355841660260648e-06,
+      "loss": 0.677,
+      "step": 2718
+    },
+    {
+      "epoch": 0.8388091932747185,
+      "grad_norm": 0.0,
+      "learning_rate": 1.3305979461939866e-06,
+      "loss": 0.7054,
+      "step": 2719
+    },
+    {
+      "epoch": 0.8391176924263458,
+      "grad_norm": 0.0,
+      "learning_rate": 1.3256203880011531e-06,
+      "loss": 0.7803,
+      "step": 2720
+    },
+    {
+      "epoch": 0.8394261915779732,
+      "grad_norm": 0.0,
+      "learning_rate": 1.3206514964206664e-06,
+      "loss": 0.6283,
+      "step": 2721
+    },
+    {
+      "epoch": 0.8397346907296005,
+      "grad_norm": 0.0,
+      "learning_rate": 1.315691276416966e-06,
+      "loss": 0.6933,
+      "step": 2722
+    },
+    {
+      "epoch": 0.8400431898812278,
+      "grad_norm": 0.0,
+      "learning_rate": 1.3107397329458348e-06,
+      "loss": 0.6158,
+      "step": 2723
+    },
+    {
+      "epoch": 0.8403516890328552,
+      "grad_norm": 0.0,
+      "learning_rate": 1.3057968709543812e-06,
+      "loss": 0.6628,
+      "step": 2724
+    },
+    {
+      "epoch": 0.8406601881844825,
+      "grad_norm": 0.0,
+      "learning_rate": 1.3008626953810388e-06,
+      "loss": 0.7122,
+      "step": 2725
+    },
+    {
+      "epoch": 0.8409686873361099,
+      "grad_norm": 0.0,
+      "learning_rate": 1.295937211155568e-06,
+      "loss": 0.7095,
+      "step": 2726
+    },
+    {
+      "epoch": 0.8412771864877372,
+      "grad_norm": 0.0,
+      "learning_rate": 1.2910204231990397e-06,
+      "loss": 0.6882,
+      "step": 2727
+    },
+    {
+      "epoch": 0.8415856856393645,
+      "grad_norm": 0.0,
+      "learning_rate": 1.2861123364238438e-06,
+      "loss": 0.7383,
+      "step": 2728
+    },
+    {
+      "epoch": 0.8418941847909919,
+      "grad_norm": 0.0,
+      "learning_rate": 1.2812129557336684e-06,
+      "loss": 0.7447,
+      "step": 2729
+    },
+    {
+      "epoch": 0.8422026839426192,
+      "grad_norm": 0.0,
+      "learning_rate": 1.276322286023506e-06,
+      "loss": 0.6829,
+      "step": 2730
+    },
+    {
+      "epoch": 0.8425111830942464,
+      "grad_norm": 0.0,
+      "learning_rate": 1.271440332179652e-06,
+      "loss": 0.6144,
+      "step": 2731
+    },
+    {
+      "epoch": 0.8428196822458738,
+      "grad_norm": 0.0,
+      "learning_rate": 1.266567099079683e-06,
+      "loss": 0.7344,
+      "step": 2732
+    },
+    {
+      "epoch": 0.8431281813975011,
+      "grad_norm": 0.0,
+      "learning_rate": 1.2617025915924764e-06,
+      "loss": 0.6714,
+      "step": 2733
+    },
+    {
+      "epoch": 0.8434366805491285,
+      "grad_norm": 0.0,
+      "learning_rate": 1.2568468145781753e-06,
+      "loss": 0.6946,
+      "step": 2734
+    },
+    {
+      "epoch": 0.8437451797007558,
+      "grad_norm": 0.0,
+      "learning_rate": 1.2519997728882138e-06,
+      "loss": 0.7134,
+      "step": 2735
+    },
+    {
+      "epoch": 0.8440536788523831,
+      "grad_norm": 0.0,
+      "learning_rate": 1.2471614713652902e-06,
+      "loss": 0.7373,
+      "step": 2736
+    },
+    {
+      "epoch": 0.8443621780040105,
+      "grad_norm": 0.0,
+      "learning_rate": 1.2423319148433777e-06,
+      "loss": 0.7617,
+      "step": 2737
+    },
+    {
+      "epoch": 0.8446706771556378,
+      "grad_norm": 0.0,
+      "learning_rate": 1.2375111081477065e-06,
+      "loss": 0.6986,
+      "step": 2738
+    },
+    {
+      "epoch": 0.8449791763072652,
+      "grad_norm": 0.0,
+      "learning_rate": 1.2326990560947627e-06,
+      "loss": 0.6288,
+      "step": 2739
+    },
+    {
+      "epoch": 0.8452876754588925,
+      "grad_norm": 0.0,
+      "learning_rate": 1.2278957634922972e-06,
+      "loss": 0.6866,
+      "step": 2740
+    },
+    {
+      "epoch": 0.8455961746105198,
+      "grad_norm": 0.0,
+      "learning_rate": 1.2231012351392958e-06,
+      "loss": 0.664,
+      "step": 2741
+    },
+    {
+      "epoch": 0.8459046737621472,
+      "grad_norm": 0.0,
+      "learning_rate": 1.218315475825994e-06,
+      "loss": 0.6362,
+      "step": 2742
+    },
+    {
+      "epoch": 0.8462131729137745,
+      "grad_norm": 0.0,
+      "learning_rate": 1.2135384903338699e-06,
+      "loss": 0.7224,
+      "step": 2743
+    },
+    {
+      "epoch": 0.8465216720654019,
+      "grad_norm": 0.0,
+      "learning_rate": 1.2087702834356264e-06,
+      "loss": 0.6184,
+      "step": 2744
+    },
+    {
+      "epoch": 0.8468301712170292,
+      "grad_norm": 0.0,
+      "learning_rate": 1.2040108598952072e-06,
+      "loss": 0.7239,
+      "step": 2745
+    },
+    {
+      "epoch": 0.8471386703686565,
+      "grad_norm": 0.0,
+      "learning_rate": 1.1992602244677653e-06,
+      "loss": 0.7085,
+      "step": 2746
+    },
+    {
+      "epoch": 0.8474471695202839,
+      "grad_norm": 0.0,
+      "learning_rate": 1.194518381899691e-06,
+      "loss": 0.712,
+      "step": 2747
+    },
+    {
+      "epoch": 0.8477556686719111,
+      "grad_norm": 0.0,
+      "learning_rate": 1.1897853369285738e-06,
+      "loss": 0.6558,
+      "step": 2748
+    },
+    {
+      "epoch": 0.8480641678235384,
+      "grad_norm": 0.0,
+      "learning_rate": 1.185061094283224e-06,
+      "loss": 0.6414,
+      "step": 2749
+    },
+    {
+      "epoch": 0.8483726669751658,
+      "grad_norm": 0.0,
+      "learning_rate": 1.180345658683658e-06,
+      "loss": 0.6253,
+      "step": 2750
+    },
+    {
+      "epoch": 0.8486811661267931,
+      "grad_norm": 0.0,
+      "learning_rate": 1.1756390348410819e-06,
+      "loss": 0.699,
+      "step": 2751
+    },
+    {
+      "epoch": 0.8489896652784205,
+      "grad_norm": 0.0,
+      "learning_rate": 1.170941227457909e-06,
+      "loss": 0.6701,
+      "step": 2752
+    },
+    {
+      "epoch": 0.8492981644300478,
+      "grad_norm": 0.0,
+      "learning_rate": 1.1662522412277388e-06,
+      "loss": 0.7196,
+      "step": 2753
+    },
+    {
+      "epoch": 0.8496066635816751,
+      "grad_norm": 0.0,
+      "learning_rate": 1.1615720808353604e-06,
+      "loss": 0.7144,
+      "step": 2754
+    },
+    {
+      "epoch": 0.8499151627333025,
+      "grad_norm": 0.0,
+      "learning_rate": 1.156900750956742e-06,
+      "loss": 0.7172,
+      "step": 2755
+    },
+    {
+      "epoch": 0.8502236618849298,
+      "grad_norm": 0.0,
+      "learning_rate": 1.1522382562590305e-06,
+      "loss": 0.6855,
+      "step": 2756
+    },
+    {
+      "epoch": 0.8505321610365572,
+      "grad_norm": 0.0,
+      "learning_rate": 1.1475846014005477e-06,
+      "loss": 0.6519,
+      "step": 2757
+    },
+    {
+      "epoch": 0.8508406601881845,
+      "grad_norm": 0.0,
+      "learning_rate": 1.1429397910307794e-06,
+      "loss": 0.6767,
+      "step": 2758
+    },
+    {
+      "epoch": 0.8511491593398118,
+      "grad_norm": 0.0,
+      "learning_rate": 1.138303829790378e-06,
+      "loss": 0.7498,
+      "step": 2759
+    },
+    {
+      "epoch": 0.8514576584914392,
+      "grad_norm": 0.0,
+      "learning_rate": 1.1336767223111521e-06,
+      "loss": 0.7421,
+      "step": 2760
+    },
+    {
+      "epoch": 0.8517661576430665,
+      "grad_norm": 0.0,
+      "learning_rate": 1.129058473216066e-06,
+      "loss": 0.6301,
+      "step": 2761
+    },
+    {
+      "epoch": 0.8520746567946939,
+      "grad_norm": 0.0,
+      "learning_rate": 1.1244490871192382e-06,
+      "loss": 0.6477,
+      "step": 2762
+    },
+    {
+      "epoch": 0.8523831559463212,
+      "grad_norm": 0.0,
+      "learning_rate": 1.1198485686259208e-06,
+      "loss": 0.6272,
+      "step": 2763
+    },
+    {
+      "epoch": 0.8526916550979485,
+      "grad_norm": 0.0,
+      "learning_rate": 1.1152569223325171e-06,
+      "loss": 0.698,
+      "step": 2764
+    },
+    {
+      "epoch": 0.8530001542495758,
+      "grad_norm": 0.0,
+      "learning_rate": 1.1106741528265585e-06,
+      "loss": 0.6155,
+      "step": 2765
+    },
+    {
+      "epoch": 0.8533086534012031,
+      "grad_norm": 0.0,
+      "learning_rate": 1.1061002646867159e-06,
+      "loss": 0.6813,
+      "step": 2766
+    },
+    {
+      "epoch": 0.8536171525528304,
+      "grad_norm": 0.0,
+      "learning_rate": 1.101535262482778e-06,
+      "loss": 0.6624,
+      "step": 2767
+    },
+    {
+      "epoch": 0.8539256517044578,
+      "grad_norm": 0.0,
+      "learning_rate": 1.0969791507756588e-06,
+      "loss": 0.9815,
+      "step": 2768
+    },
+    {
+      "epoch": 0.8542341508560851,
+      "grad_norm": 0.0,
+      "learning_rate": 1.0924319341173938e-06,
+      "loss": 0.7458,
+      "step": 2769
+    },
+    {
+      "epoch": 0.8545426500077125,
+      "grad_norm": 0.0,
+      "learning_rate": 1.0878936170511246e-06,
+      "loss": 0.7608,
+      "step": 2770
+    },
+    {
+      "epoch": 0.8548511491593398,
+      "grad_norm": 0.0,
+      "learning_rate": 1.0833642041111102e-06,
+      "loss": 0.666,
+      "step": 2771
+    },
+    {
+      "epoch": 0.8551596483109671,
+      "grad_norm": 0.0,
+      "learning_rate": 1.0788436998226991e-06,
+      "loss": 0.7159,
+      "step": 2772
+    },
+    {
+      "epoch": 0.8554681474625945,
+      "grad_norm": 0.0,
+      "learning_rate": 1.0743321087023528e-06,
+      "loss": 0.6403,
+      "step": 2773
+    },
+    {
+      "epoch": 0.8557766466142218,
+      "grad_norm": 0.0,
+      "learning_rate": 1.0698294352576232e-06,
+      "loss": 0.706,
+      "step": 2774
+    },
+    {
+      "epoch": 0.8560851457658492,
+      "grad_norm": 0.0,
+      "learning_rate": 1.065335683987152e-06,
+      "loss": 0.634,
+      "step": 2775
+    },
+    {
+      "epoch": 0.8563936449174765,
+      "grad_norm": 0.0,
+      "learning_rate": 1.0608508593806655e-06,
+      "loss": 0.7068,
+      "step": 2776
+    },
+    {
+      "epoch": 0.8567021440691038,
+      "grad_norm": 0.0,
+      "learning_rate": 1.0563749659189703e-06,
+      "loss": 0.7045,
+      "step": 2777
+    },
+    {
+      "epoch": 0.8570106432207312,
+      "grad_norm": 0.0,
+      "learning_rate": 1.0519080080739562e-06,
+      "loss": 0.6228,
+      "step": 2778
+    },
+    {
+      "epoch": 0.8573191423723585,
+      "grad_norm": 0.0,
+      "learning_rate": 1.047449990308579e-06,
+      "loss": 0.767,
+      "step": 2779
+    },
+    {
+      "epoch": 0.8576276415239859,
+      "grad_norm": 0.0,
+      "learning_rate": 1.043000917076863e-06,
+      "loss": 0.714,
+      "step": 2780
+    },
+    {
+      "epoch": 0.8579361406756132,
+      "grad_norm": 0.0,
+      "learning_rate": 1.0385607928239017e-06,
+      "loss": 0.6613,
+      "step": 2781
+    },
+    {
+      "epoch": 0.8582446398272405,
+      "grad_norm": 0.0,
+      "learning_rate": 1.0341296219858398e-06,
+      "loss": 0.6747,
+      "step": 2782
+    },
+    {
+      "epoch": 0.8585531389788678,
+      "grad_norm": 0.0,
+      "learning_rate": 1.029707408989884e-06,
+      "loss": 0.7062,
+      "step": 2783
+    },
+    {
+      "epoch": 0.8588616381304951,
+      "grad_norm": 0.0,
+      "learning_rate": 1.0252941582542852e-06,
+      "loss": 0.5831,
+      "step": 2784
+    },
+    {
+      "epoch": 0.8591701372821224,
+      "grad_norm": 0.0,
+      "learning_rate": 1.02088987418834e-06,
+      "loss": 0.6918,
+      "step": 2785
+    },
+    {
+      "epoch": 0.8594786364337498,
+      "grad_norm": 0.0,
+      "learning_rate": 1.0164945611923948e-06,
+      "loss": 0.7209,
+      "step": 2786
+    },
+    {
+      "epoch": 0.8597871355853771,
+      "grad_norm": 0.0,
+      "learning_rate": 1.0121082236578205e-06,
+      "loss": 0.6809,
+      "step": 2787
+    },
+    {
+      "epoch": 0.8600956347370045,
+      "grad_norm": 0.0,
+      "learning_rate": 1.0077308659670348e-06,
+      "loss": 0.7467,
+      "step": 2788
+    },
+    {
+      "epoch": 0.8604041338886318,
+      "grad_norm": 0.0,
+      "learning_rate": 1.003362492493466e-06,
+      "loss": 0.6971,
+      "step": 2789
+    },
+    {
+      "epoch": 0.8607126330402591,
+      "grad_norm": 0.0,
+      "learning_rate": 9.990031076015838e-07,
+      "loss": 0.7136,
+      "step": 2790
+    },
+    {
+      "epoch": 0.8610211321918865,
+      "grad_norm": 0.0,
+      "learning_rate": 9.946527156468643e-07,
+      "loss": 0.7539,
+      "step": 2791
+    },
+    {
+      "epoch": 0.8613296313435138,
+      "grad_norm": 0.0,
+      "learning_rate": 9.903113209758098e-07,
+      "loss": 0.7467,
+      "step": 2792
+    },
+    {
+      "epoch": 0.8616381304951412,
+      "grad_norm": 0.0,
+      "learning_rate": 9.859789279259225e-07,
+      "loss": 0.6418,
+      "step": 2793
+    },
+    {
+      "epoch": 0.8619466296467685,
+      "grad_norm": 0.0,
+      "learning_rate": 9.816555408257177e-07,
+      "loss": 0.7238,
+      "step": 2794
+    },
+    {
+      "epoch": 0.8622551287983958,
+      "grad_norm": 0.0,
+      "learning_rate": 9.773411639947117e-07,
+      "loss": 0.6799,
+      "step": 2795
+    },
+    {
+      "epoch": 0.8625636279500232,
+      "grad_norm": 0.0,
+      "learning_rate": 9.730358017434194e-07,
+      "loss": 0.6993,
+      "step": 2796
+    },
+    {
+      "epoch": 0.8628721271016505,
+      "grad_norm": 0.0,
+      "learning_rate": 9.687394583733435e-07,
+      "loss": 0.659,
+      "step": 2797
+    },
+    {
+      "epoch": 0.8631806262532778,
+      "grad_norm": 0.0,
+      "learning_rate": 9.644521381769855e-07,
+      "loss": 0.7527,
+      "step": 2798
+    },
+    {
+      "epoch": 0.8634891254049052,
+      "grad_norm": 0.0,
+      "learning_rate": 9.601738454378229e-07,
+      "loss": 0.7055,
+      "step": 2799
+    },
+    {
+      "epoch": 0.8637976245565324,
+      "grad_norm": 0.0,
+      "learning_rate": 9.559045844303193e-07,
+      "loss": 0.9917,
+      "step": 2800
+    },
+    {
+      "epoch": 0.8641061237081598,
+      "grad_norm": 0.0,
+      "learning_rate": 9.516443594199143e-07,
+      "loss": 0.677,
+      "step": 2801
+    },
+    {
+      "epoch": 0.8644146228597871,
+      "grad_norm": 0.0,
+      "learning_rate": 9.47393174663016e-07,
+      "loss": 0.7474,
+      "step": 2802
+    },
+    {
+      "epoch": 0.8647231220114144,
+      "grad_norm": 0.0,
+      "learning_rate": 9.431510344070005e-07,
+      "loss": 0.7235,
+      "step": 2803
+    },
+    {
+      "epoch": 0.8650316211630418,
+      "grad_norm": 0.0,
+      "learning_rate": 9.389179428902139e-07,
+      "loss": 0.5918,
+      "step": 2804
+    },
+    {
+      "epoch": 0.8653401203146691,
+      "grad_norm": 0.0,
+      "learning_rate": 9.346939043419568e-07,
+      "loss": 0.6162,
+      "step": 2805
+    },
+    {
+      "epoch": 0.8656486194662965,
+      "grad_norm": 0.0,
+      "learning_rate": 9.304789229824806e-07,
+      "loss": 0.7186,
+      "step": 2806
+    },
+    {
+      "epoch": 0.8659571186179238,
+      "grad_norm": 0.0,
+      "learning_rate": 9.262730030229983e-07,
+      "loss": 0.9891,
+      "step": 2807
+    },
+    {
+      "epoch": 0.8662656177695511,
+      "grad_norm": 0.0,
+      "learning_rate": 9.220761486656593e-07,
+      "loss": 0.7034,
+      "step": 2808
+    },
+    {
+      "epoch": 0.8665741169211785,
+      "grad_norm": 0.0,
+      "learning_rate": 9.178883641035652e-07,
+      "loss": 0.7197,
+      "step": 2809
+    },
+    {
+      "epoch": 0.8668826160728058,
+      "grad_norm": 0.0,
+      "learning_rate": 9.137096535207457e-07,
+      "loss": 0.6205,
+      "step": 2810
+    },
+    {
+      "epoch": 0.8671911152244332,
+      "grad_norm": 0.0,
+      "learning_rate": 9.095400210921712e-07,
+      "loss": 0.6383,
+      "step": 2811
+    },
+    {
+      "epoch": 0.8674996143760605,
+      "grad_norm": 0.0,
+      "learning_rate": 9.053794709837415e-07,
+      "loss": 0.7297,
+      "step": 2812
+    },
+    {
+      "epoch": 0.8678081135276878,
+      "grad_norm": 0.0,
+      "learning_rate": 9.012280073522816e-07,
+      "loss": 0.6416,
+      "step": 2813
+    },
+    {
+      "epoch": 0.8681166126793152,
+      "grad_norm": 0.0,
+      "learning_rate": 8.97085634345537e-07,
+      "loss": 0.752,
+      "step": 2814
+    },
+    {
+      "epoch": 0.8684251118309425,
+      "grad_norm": 0.0,
+      "learning_rate": 8.929523561021691e-07,
+      "loss": 0.6442,
+      "step": 2815
+    },
+    {
+      "epoch": 0.8687336109825698,
+      "grad_norm": 0.0,
+      "learning_rate": 8.888281767517582e-07,
+      "loss": 0.6653,
+      "step": 2816
+    },
+    {
+      "epoch": 0.8690421101341971,
+      "grad_norm": 0.0,
+      "learning_rate": 8.847131004147891e-07,
+      "loss": 0.7447,
+      "step": 2817
+    },
+    {
+      "epoch": 0.8693506092858244,
+      "grad_norm": 0.0,
+      "learning_rate": 8.806071312026554e-07,
+      "loss": 0.7057,
+      "step": 2818
+    },
+    {
+      "epoch": 0.8696591084374518,
+      "grad_norm": 0.0,
+      "learning_rate": 8.765102732176489e-07,
+      "loss": 0.7113,
+      "step": 2819
+    },
+    {
+      "epoch": 0.8699676075890791,
+      "grad_norm": 0.0,
+      "learning_rate": 8.724225305529577e-07,
+      "loss": 0.7002,
+      "step": 2820
+    },
+    {
+      "epoch": 0.8702761067407064,
+      "grad_norm": 0.0,
+      "learning_rate": 8.683439072926658e-07,
+      "loss": 0.7051,
+      "step": 2821
+    },
+    {
+      "epoch": 0.8705846058923338,
+      "grad_norm": 0.0,
+      "learning_rate": 8.642744075117448e-07,
+      "loss": 0.6729,
+      "step": 2822
+    },
+    {
+      "epoch": 0.8708931050439611,
+      "grad_norm": 0.0,
+      "learning_rate": 8.602140352760469e-07,
+      "loss": 0.6511,
+      "step": 2823
+    },
+    {
+      "epoch": 0.8712016041955885,
+      "grad_norm": 0.0,
+      "learning_rate": 8.561627946423134e-07,
+      "loss": 0.6652,
+      "step": 2824
+    },
+    {
+      "epoch": 0.8715101033472158,
+      "grad_norm": 0.0,
+      "learning_rate": 8.521206896581535e-07,
+      "loss": 0.6922,
+      "step": 2825
+    },
+    {
+      "epoch": 0.8718186024988431,
+      "grad_norm": 0.0,
+      "learning_rate": 8.480877243620578e-07,
+      "loss": 0.7121,
+      "step": 2826
+    },
+    {
+      "epoch": 0.8721271016504705,
+      "grad_norm": 0.0,
+      "learning_rate": 8.44063902783373e-07,
+      "loss": 0.7083,
+      "step": 2827
+    },
+    {
+      "epoch": 0.8724356008020978,
+      "grad_norm": 0.0,
+      "learning_rate": 8.400492289423235e-07,
+      "loss": 0.8806,
+      "step": 2828
+    },
+    {
+      "epoch": 0.8727440999537251,
+      "grad_norm": 0.0,
+      "learning_rate": 8.360437068499849e-07,
+      "loss": 0.6684,
+      "step": 2829
+    },
+    {
+      "epoch": 0.8730525991053525,
+      "grad_norm": 0.0,
+      "learning_rate": 8.320473405082952e-07,
+      "loss": 0.7055,
+      "step": 2830
+    },
+    {
+      "epoch": 0.8733610982569798,
+      "grad_norm": 0.0,
+      "learning_rate": 8.280601339100436e-07,
+      "loss": 0.6907,
+      "step": 2831
+    },
+    {
+      "epoch": 0.8736695974086072,
+      "grad_norm": 0.0,
+      "learning_rate": 8.240820910388614e-07,
+      "loss": 0.6585,
+      "step": 2832
+    },
+    {
+      "epoch": 0.8739780965602345,
+      "grad_norm": 0.0,
+      "learning_rate": 8.201132158692327e-07,
+      "loss": 0.7069,
+      "step": 2833
+    },
+    {
+      "epoch": 0.8742865957118618,
+      "grad_norm": 0.0,
+      "learning_rate": 8.161535123664776e-07,
+      "loss": 0.6241,
+      "step": 2834
+    },
+    {
+      "epoch": 0.8745950948634891,
+      "grad_norm": 0.0,
+      "learning_rate": 8.122029844867562e-07,
+      "loss": 0.6609,
+      "step": 2835
+    },
+    {
+      "epoch": 0.8749035940151164,
+      "grad_norm": 0.0,
+      "learning_rate": 8.082616361770579e-07,
+      "loss": 0.7406,
+      "step": 2836
+    },
+    {
+      "epoch": 0.8752120931667438,
+      "grad_norm": 0.0,
+      "learning_rate": 8.043294713751992e-07,
+      "loss": 0.6594,
+      "step": 2837
+    },
+    {
+      "epoch": 0.8755205923183711,
+      "grad_norm": 0.0,
+      "learning_rate": 8.004064940098277e-07,
+      "loss": 0.6504,
+      "step": 2838
+    },
+    {
+      "epoch": 0.8758290914699984,
+      "grad_norm": 0.0,
+      "learning_rate": 7.964927080004059e-07,
+      "loss": 0.6821,
+      "step": 2839
+    },
+    {
+      "epoch": 0.8761375906216258,
+      "grad_norm": 0.0,
+      "learning_rate": 7.925881172572169e-07,
+      "loss": 0.7181,
+      "step": 2840
+    },
+    {
+      "epoch": 0.8764460897732531,
+      "grad_norm": 0.0,
+      "learning_rate": 7.886927256813514e-07,
+      "loss": 0.6449,
+      "step": 2841
+    },
+    {
+      "epoch": 0.8767545889248805,
+      "grad_norm": 0.0,
+      "learning_rate": 7.848065371647162e-07,
+      "loss": 0.6459,
+      "step": 2842
+    },
+    {
+      "epoch": 0.8770630880765078,
+      "grad_norm": 0.0,
+      "learning_rate": 7.809295555900231e-07,
+      "loss": 0.7463,
+      "step": 2843
+    },
+    {
+      "epoch": 0.8773715872281351,
+      "grad_norm": 0.0,
+      "learning_rate": 7.770617848307749e-07,
+      "loss": 0.6198,
+      "step": 2844
+    },
+    {
+      "epoch": 0.8776800863797625,
+      "grad_norm": 0.0,
+      "learning_rate": 7.732032287512847e-07,
+      "loss": 0.7149,
+      "step": 2845
+    },
+    {
+      "epoch": 0.8779885855313898,
+      "grad_norm": 0.0,
+      "learning_rate": 7.6935389120665e-07,
+      "loss": 0.6899,
+      "step": 2846
+    },
+    {
+      "epoch": 0.8782970846830171,
+      "grad_norm": 0.0,
+      "learning_rate": 7.655137760427645e-07,
+      "loss": 0.6932,
+      "step": 2847
+    },
+    {
+      "epoch": 0.8786055838346445,
+      "grad_norm": 0.0,
+      "learning_rate": 7.616828870963023e-07,
+      "loss": 0.6516,
+      "step": 2848
+    },
+    {
+      "epoch": 0.8789140829862718,
+      "grad_norm": 0.0,
+      "learning_rate": 7.578612281947229e-07,
+      "loss": 0.6304,
+      "step": 2849
+    },
+    {
+      "epoch": 0.8792225821378992,
+      "grad_norm": 0.0,
+      "learning_rate": 7.540488031562642e-07,
+      "loss": 0.6516,
+      "step": 2850
+    },
+    {
+      "epoch": 0.8795310812895265,
+      "grad_norm": 0.0,
+      "learning_rate": 7.502456157899351e-07,
+      "loss": 0.6183,
+      "step": 2851
+    },
+    {
+      "epoch": 0.8798395804411537,
+      "grad_norm": 0.0,
+      "learning_rate": 7.464516698955226e-07,
+      "loss": 0.6367,
+      "step": 2852
+    },
+    {
+      "epoch": 0.8801480795927811,
+      "grad_norm": 0.0,
+      "learning_rate": 7.426669692635691e-07,
+      "loss": 0.673,
+      "step": 2853
+    },
+    {
+      "epoch": 0.8804565787444084,
+      "grad_norm": 0.0,
+      "learning_rate": 7.388915176753886e-07,
+      "loss": 0.7608,
+      "step": 2854
+    },
+    {
+      "epoch": 0.8807650778960358,
+      "grad_norm": 0.0,
+      "learning_rate": 7.351253189030538e-07,
+      "loss": 0.7474,
+      "step": 2855
+    },
+    {
+      "epoch": 0.8810735770476631,
+      "grad_norm": 0.0,
+      "learning_rate": 7.313683767093915e-07,
+      "loss": 0.669,
+      "step": 2856
+    },
+    {
+      "epoch": 0.8813820761992904,
+      "grad_norm": 0.0,
+      "learning_rate": 7.276206948479769e-07,
+      "loss": 0.633,
+      "step": 2857
+    },
+    {
+      "epoch": 0.8816905753509178,
+      "grad_norm": 0.0,
+      "learning_rate": 7.23882277063136e-07,
+      "loss": 0.6076,
+      "step": 2858
+    },
+    {
+      "epoch": 0.8819990745025451,
+      "grad_norm": 0.0,
+      "learning_rate": 7.20153127089942e-07,
+      "loss": 0.6447,
+      "step": 2859
+    },
+    {
+      "epoch": 0.8823075736541725,
+      "grad_norm": 0.0,
+      "learning_rate": 7.164332486542036e-07,
+      "loss": 0.7517,
+      "step": 2860
+    },
+    {
+      "epoch": 0.8826160728057998,
+      "grad_norm": 0.0,
+      "learning_rate": 7.127226454724689e-07,
+      "loss": 0.7321,
+      "step": 2861
+    },
+    {
+      "epoch": 0.8829245719574271,
+      "grad_norm": 0.0,
+      "learning_rate": 7.090213212520191e-07,
+      "loss": 0.6784,
+      "step": 2862
+    },
+    {
+      "epoch": 0.8832330711090545,
+      "grad_norm": 0.0,
+      "learning_rate": 7.053292796908629e-07,
+      "loss": 0.7292,
+      "step": 2863
+    },
+    {
+      "epoch": 0.8835415702606818,
+      "grad_norm": 0.0,
+      "learning_rate": 7.01646524477737e-07,
+      "loss": 0.6657,
+      "step": 2864
+    },
+    {
+      "epoch": 0.8838500694123091,
+      "grad_norm": 0.0,
+      "learning_rate": 6.979730592920997e-07,
+      "loss": 0.7044,
+      "step": 2865
+    },
+    {
+      "epoch": 0.8841585685639365,
+      "grad_norm": 0.0,
+      "learning_rate": 6.943088878041238e-07,
+      "loss": 0.7455,
+      "step": 2866
+    },
+    {
+      "epoch": 0.8844670677155638,
+      "grad_norm": 0.0,
+      "learning_rate": 6.906540136747031e-07,
+      "loss": 0.6747,
+      "step": 2867
+    },
+    {
+      "epoch": 0.8847755668671912,
+      "grad_norm": 0.0,
+      "learning_rate": 6.870084405554367e-07,
+      "loss": 0.6485,
+      "step": 2868
+    },
+    {
+      "epoch": 0.8850840660188184,
+      "grad_norm": 0.0,
+      "learning_rate": 6.833721720886377e-07,
+      "loss": 0.7415,
+      "step": 2869
+    },
+    {
+      "epoch": 0.8853925651704457,
+      "grad_norm": 0.0,
+      "learning_rate": 6.797452119073111e-07,
+      "loss": 0.6579,
+      "step": 2870
+    },
+    {
+      "epoch": 0.8857010643220731,
+      "grad_norm": 0.0,
+      "learning_rate": 6.76127563635176e-07,
+      "loss": 0.7267,
+      "step": 2871
+    },
+    {
+      "epoch": 0.8860095634737004,
+      "grad_norm": 0.0,
+      "learning_rate": 6.725192308866368e-07,
+      "loss": 0.7453,
+      "step": 2872
+    },
+    {
+      "epoch": 0.8863180626253278,
+      "grad_norm": 0.0,
+      "learning_rate": 6.689202172667996e-07,
+      "loss": 0.6308,
+      "step": 2873
+    },
+    {
+      "epoch": 0.8866265617769551,
+      "grad_norm": 0.0,
+      "learning_rate": 6.653305263714526e-07,
+      "loss": 0.6391,
+      "step": 2874
+    },
+    {
+      "epoch": 0.8869350609285824,
+      "grad_norm": 0.0,
+      "learning_rate": 6.617501617870726e-07,
+      "loss": 0.6722,
+      "step": 2875
+    },
+    {
+      "epoch": 0.8872435600802098,
+      "grad_norm": 0.0,
+      "learning_rate": 6.581791270908189e-07,
+      "loss": 0.657,
+      "step": 2876
+    },
+    {
+      "epoch": 0.8875520592318371,
+      "grad_norm": 0.0,
+      "learning_rate": 6.546174258505289e-07,
+      "loss": 0.6685,
+      "step": 2877
+    },
+    {
+      "epoch": 0.8878605583834644,
+      "grad_norm": 0.0,
+      "learning_rate": 6.510650616247116e-07,
+      "loss": 0.6154,
+      "step": 2878
+    },
+    {
+      "epoch": 0.8881690575350918,
+      "grad_norm": 0.0,
+      "learning_rate": 6.475220379625546e-07,
+      "loss": 0.7037,
+      "step": 2879
+    },
+    {
+      "epoch": 0.8884775566867191,
+      "grad_norm": 0.0,
+      "learning_rate": 6.439883584039064e-07,
+      "loss": 0.6844,
+      "step": 2880
+    },
+    {
+      "epoch": 0.8887860558383465,
+      "grad_norm": 0.0,
+      "learning_rate": 6.404640264792839e-07,
+      "loss": 0.7061,
+      "step": 2881
+    },
+    {
+      "epoch": 0.8890945549899738,
+      "grad_norm": 0.0,
+      "learning_rate": 6.369490457098626e-07,
+      "loss": 0.7333,
+      "step": 2882
+    },
+    {
+      "epoch": 0.8894030541416011,
+      "grad_norm": 0.0,
+      "learning_rate": 6.334434196074757e-07,
+      "loss": 0.7895,
+      "step": 2883
+    },
+    {
+      "epoch": 0.8897115532932285,
+      "grad_norm": 0.0,
+      "learning_rate": 6.299471516746092e-07,
+      "loss": 0.6718,
+      "step": 2884
+    },
+    {
+      "epoch": 0.8900200524448558,
+      "grad_norm": 0.0,
+      "learning_rate": 6.264602454044033e-07,
+      "loss": 0.7163,
+      "step": 2885
+    },
+    {
+      "epoch": 0.8903285515964832,
+      "grad_norm": 0.0,
+      "learning_rate": 6.229827042806403e-07,
+      "loss": 0.7093,
+      "step": 2886
+    },
+    {
+      "epoch": 0.8906370507481104,
+      "grad_norm": 0.0,
+      "learning_rate": 6.195145317777452e-07,
+      "loss": 0.6634,
+      "step": 2887
+    },
+    {
+      "epoch": 0.8909455498997377,
+      "grad_norm": 0.0,
+      "learning_rate": 6.160557313607884e-07,
+      "loss": 0.72,
+      "step": 2888
+    },
+    {
+      "epoch": 0.8912540490513651,
+      "grad_norm": 0.0,
+      "learning_rate": 6.126063064854715e-07,
+      "loss": 0.7211,
+      "step": 2889
+    },
+    {
+      "epoch": 0.8915625482029924,
+      "grad_norm": 0.0,
+      "learning_rate": 6.091662605981319e-07,
+      "loss": 0.7169,
+      "step": 2890
+    },
+    {
+      "epoch": 0.8918710473546198,
+      "grad_norm": 0.0,
+      "learning_rate": 6.057355971357359e-07,
+      "loss": 0.6511,
+      "step": 2891
+    },
+    {
+      "epoch": 0.8921795465062471,
+      "grad_norm": 0.0,
+      "learning_rate": 6.023143195258708e-07,
+      "loss": 0.7037,
+      "step": 2892
+    },
+    {
+      "epoch": 0.8924880456578744,
+      "grad_norm": 0.0,
+      "learning_rate": 5.989024311867553e-07,
+      "loss": 0.7011,
+      "step": 2893
+    },
+    {
+      "epoch": 0.8927965448095018,
+      "grad_norm": 0.0,
+      "learning_rate": 5.954999355272207e-07,
+      "loss": 0.7156,
+      "step": 2894
+    },
+    {
+      "epoch": 0.8931050439611291,
+      "grad_norm": 0.0,
+      "learning_rate": 5.921068359467164e-07,
+      "loss": 0.7205,
+      "step": 2895
+    },
+    {
+      "epoch": 0.8934135431127564,
+      "grad_norm": 0.0,
+      "learning_rate": 5.887231358353018e-07,
+      "loss": 0.7144,
+      "step": 2896
+    },
+    {
+      "epoch": 0.8937220422643838,
+      "grad_norm": 0.0,
+      "learning_rate": 5.853488385736506e-07,
+      "loss": 0.6505,
+      "step": 2897
+    },
+    {
+      "epoch": 0.8940305414160111,
+      "grad_norm": 0.0,
+      "learning_rate": 5.819839475330358e-07,
+      "loss": 0.69,
+      "step": 2898
+    },
+    {
+      "epoch": 0.8943390405676385,
+      "grad_norm": 0.0,
+      "learning_rate": 5.786284660753372e-07,
+      "loss": 0.7309,
+      "step": 2899
+    },
+    {
+      "epoch": 0.8946475397192658,
+      "grad_norm": 0.0,
+      "learning_rate": 5.752823975530308e-07,
+      "loss": 0.6849,
+      "step": 2900
+    },
+    {
+      "epoch": 0.8949560388708931,
+      "grad_norm": 0.0,
+      "learning_rate": 5.719457453091881e-07,
+      "loss": 0.6756,
+      "step": 2901
+    },
+    {
+      "epoch": 0.8952645380225205,
+      "grad_norm": 0.0,
+      "learning_rate": 5.686185126774757e-07,
+      "loss": 0.6996,
+      "step": 2902
+    },
+    {
+      "epoch": 0.8955730371741478,
+      "grad_norm": 0.0,
+      "learning_rate": 5.653007029821467e-07,
+      "loss": 0.719,
+      "step": 2903
+    },
+    {
+      "epoch": 0.895881536325775,
+      "grad_norm": 0.0,
+      "learning_rate": 5.619923195380372e-07,
+      "loss": 0.7229,
+      "step": 2904
+    },
+    {
+      "epoch": 0.8961900354774024,
+      "grad_norm": 0.0,
+      "learning_rate": 5.586933656505711e-07,
+      "loss": 0.7041,
+      "step": 2905
+    },
+    {
+      "epoch": 0.8964985346290297,
+      "grad_norm": 0.0,
+      "learning_rate": 5.554038446157473e-07,
+      "loss": 0.6751,
+      "step": 2906
+    },
+    {
+      "epoch": 0.8968070337806571,
+      "grad_norm": 0.0,
+      "learning_rate": 5.521237597201423e-07,
+      "loss": 0.7222,
+      "step": 2907
+    },
+    {
+      "epoch": 0.8971155329322844,
+      "grad_norm": 0.0,
+      "learning_rate": 5.488531142409015e-07,
+      "loss": 0.6142,
+      "step": 2908
+    },
+    {
+      "epoch": 0.8974240320839117,
+      "grad_norm": 0.0,
+      "learning_rate": 5.455919114457442e-07,
+      "loss": 0.6486,
+      "step": 2909
+    },
+    {
+      "epoch": 0.8977325312355391,
+      "grad_norm": 0.0,
+      "learning_rate": 5.423401545929518e-07,
+      "loss": 0.6912,
+      "step": 2910
+    },
+    {
+      "epoch": 0.8980410303871664,
+      "grad_norm": 0.0,
+      "learning_rate": 5.390978469313701e-07,
+      "loss": 0.7119,
+      "step": 2911
+    },
+    {
+      "epoch": 0.8983495295387938,
+      "grad_norm": 0.0,
+      "learning_rate": 5.358649917004033e-07,
+      "loss": 0.6782,
+      "step": 2912
+    },
+    {
+      "epoch": 0.8986580286904211,
+      "grad_norm": 0.0,
+      "learning_rate": 5.326415921300099e-07,
+      "loss": 0.7136,
+      "step": 2913
+    },
+    {
+      "epoch": 0.8989665278420484,
+      "grad_norm": 0.0,
+      "learning_rate": 5.294276514407071e-07,
+      "loss": 0.6287,
+      "step": 2914
+    },
+    {
+      "epoch": 0.8992750269936758,
+      "grad_norm": 0.0,
+      "learning_rate": 5.262231728435541e-07,
+      "loss": 0.732,
+      "step": 2915
+    },
+    {
+      "epoch": 0.8995835261453031,
+      "grad_norm": 0.0,
+      "learning_rate": 5.230281595401621e-07,
+      "loss": 0.6561,
+      "step": 2916
+    },
+    {
+      "epoch": 0.8998920252969305,
+      "grad_norm": 0.0,
+      "learning_rate": 5.198426147226843e-07,
+      "loss": 0.6318,
+      "step": 2917
+    },
+    {
+      "epoch": 0.9002005244485578,
+      "grad_norm": 0.0,
+      "learning_rate": 5.166665415738093e-07,
+      "loss": 0.6749,
+      "step": 2918
+    },
+    {
+      "epoch": 0.9005090236001851,
+      "grad_norm": 0.0,
+      "learning_rate": 5.134999432667708e-07,
+      "loss": 0.6787,
+      "step": 2919
+    },
+    {
+      "epoch": 0.9008175227518125,
+      "grad_norm": 0.0,
+      "learning_rate": 5.103428229653296e-07,
+      "loss": 0.7511,
+      "step": 2920
+    },
+    {
+      "epoch": 0.9011260219034397,
+      "grad_norm": 0.0,
+      "learning_rate": 5.071951838237787e-07,
+      "loss": 0.665,
+      "step": 2921
+    },
+    {
+      "epoch": 0.901434521055067,
+      "grad_norm": 0.0,
+      "learning_rate": 5.040570289869384e-07,
+      "loss": 0.6458,
+      "step": 2922
+    },
+    {
+      "epoch": 0.9017430202066944,
+      "grad_norm": 0.0,
+      "learning_rate": 5.009283615901561e-07,
+      "loss": 0.669,
+      "step": 2923
+    },
+    {
+      "epoch": 0.9020515193583217,
+      "grad_norm": 0.0,
+      "learning_rate": 4.978091847592959e-07,
+      "loss": 0.6912,
+      "step": 2924
+    },
+    {
+      "epoch": 0.9023600185099491,
+      "grad_norm": 0.0,
+      "learning_rate": 4.946995016107392e-07,
+      "loss": 0.6912,
+      "step": 2925
+    },
+    {
+      "epoch": 0.9026685176615764,
+      "grad_norm": 0.0,
+      "learning_rate": 4.915993152513887e-07,
+      "loss": 0.7632,
+      "step": 2926
+    },
+    {
+      "epoch": 0.9029770168132037,
+      "grad_norm": 0.0,
+      "learning_rate": 4.885086287786522e-07,
+      "loss": 0.6636,
+      "step": 2927
+    },
+    {
+      "epoch": 0.9032855159648311,
+      "grad_norm": 0.0,
+      "learning_rate": 4.854274452804519e-07,
+      "loss": 0.7389,
+      "step": 2928
+    },
+    {
+      "epoch": 0.9035940151164584,
+      "grad_norm": 0.0,
+      "learning_rate": 4.823557678352097e-07,
+      "loss": 0.6548,
+      "step": 2929
+    },
+    {
+      "epoch": 0.9039025142680858,
+      "grad_norm": 0.0,
+      "learning_rate": 4.792935995118509e-07,
+      "loss": 0.7511,
+      "step": 2930
+    },
+    {
+      "epoch": 0.9042110134197131,
+      "grad_norm": 0.0,
+      "learning_rate": 4.7624094336980565e-07,
+      "loss": 0.7295,
+      "step": 2931
+    },
+    {
+      "epoch": 0.9045195125713404,
+      "grad_norm": 0.0,
+      "learning_rate": 4.731978024589945e-07,
+      "loss": 0.65,
+      "step": 2932
+    },
+    {
+      "epoch": 0.9048280117229678,
+      "grad_norm": 0.0,
+      "learning_rate": 4.701641798198353e-07,
+      "loss": 0.6513,
+      "step": 2933
+    },
+    {
+      "epoch": 0.9051365108745951,
+      "grad_norm": 0.0,
+      "learning_rate": 4.671400784832314e-07,
+      "loss": 0.7507,
+      "step": 2934
+    },
+    {
+      "epoch": 0.9054450100262225,
+      "grad_norm": 0.0,
+      "learning_rate": 4.641255014705781e-07,
+      "loss": 0.6802,
+      "step": 2935
+    },
+    {
+      "epoch": 0.9057535091778498,
+      "grad_norm": 0.0,
+      "learning_rate": 4.6112045179375286e-07,
+      "loss": 0.6565,
+      "step": 2936
+    },
+    {
+      "epoch": 0.9060620083294771,
+      "grad_norm": 0.0,
+      "learning_rate": 4.58124932455114e-07,
+      "loss": 0.7055,
+      "step": 2937
+    },
+    {
+      "epoch": 0.9063705074811045,
+      "grad_norm": 0.0,
+      "learning_rate": 4.5513894644750087e-07,
+      "loss": 0.7324,
+      "step": 2938
+    },
+    {
+      "epoch": 0.9066790066327317,
+      "grad_norm": 0.0,
+      "learning_rate": 4.5216249675422153e-07,
+      "loss": 0.7231,
+      "step": 2939
+    },
+    {
+      "epoch": 0.906987505784359,
+      "grad_norm": 0.0,
+      "learning_rate": 4.491955863490649e-07,
+      "loss": 0.69,
+      "step": 2940
+    },
+    {
+      "epoch": 0.9072960049359864,
+      "grad_norm": 0.0,
+      "learning_rate": 4.462382181962832e-07,
+      "loss": 0.6883,
+      "step": 2941
+    },
+    {
+      "epoch": 0.9076045040876137,
+      "grad_norm": 0.0,
+      "learning_rate": 4.43290395250594e-07,
+      "loss": 0.6147,
+      "step": 2942
+    },
+    {
+      "epoch": 0.9079130032392411,
+      "grad_norm": 0.0,
+      "learning_rate": 4.4035212045718366e-07,
+      "loss": 0.7843,
+      "step": 2943
+    },
+    {
+      "epoch": 0.9082215023908684,
+      "grad_norm": 0.0,
+      "learning_rate": 4.37423396751695e-07,
+      "loss": 0.7698,
+      "step": 2944
+    },
+    {
+      "epoch": 0.9085300015424957,
+      "grad_norm": 0.0,
+      "learning_rate": 4.3450422706022956e-07,
+      "loss": 0.7347,
+      "step": 2945
+    },
+    {
+      "epoch": 0.9088385006941231,
+      "grad_norm": 0.0,
+      "learning_rate": 4.315946142993432e-07,
+      "loss": 0.6458,
+      "step": 2946
+    },
+    {
+      "epoch": 0.9091469998457504,
+      "grad_norm": 0.0,
+      "learning_rate": 4.286945613760429e-07,
+      "loss": 0.7321,
+      "step": 2947
+    },
+    {
+      "epoch": 0.9094554989973778,
+      "grad_norm": 0.0,
+      "learning_rate": 4.258040711877842e-07,
+      "loss": 0.7338,
+      "step": 2948
+    },
+    {
+      "epoch": 0.9097639981490051,
+      "grad_norm": 0.0,
+      "learning_rate": 4.229231466224693e-07,
+      "loss": 0.6162,
+      "step": 2949
+    },
+    {
+      "epoch": 0.9100724973006324,
+      "grad_norm": 0.0,
+      "learning_rate": 4.200517905584467e-07,
+      "loss": 0.632,
+      "step": 2950
+    },
+    {
+      "epoch": 0.9103809964522598,
+      "grad_norm": 0.0,
+      "learning_rate": 4.1719000586449507e-07,
+      "loss": 0.662,
+      "step": 2951
+    },
+    {
+      "epoch": 0.9106894956038871,
+      "grad_norm": 0.0,
+      "learning_rate": 4.143377953998418e-07,
+      "loss": 0.7428,
+      "step": 2952
+    },
+    {
+      "epoch": 0.9109979947555145,
+      "grad_norm": 0.0,
+      "learning_rate": 4.1149516201413944e-07,
+      "loss": 0.6866,
+      "step": 2953
+    },
+    {
+      "epoch": 0.9113064939071418,
+      "grad_norm": 0.0,
+      "learning_rate": 4.0866210854747956e-07,
+      "loss": 0.6504,
+      "step": 2954
+    },
+    {
+      "epoch": 0.9116149930587691,
+      "grad_norm": 0.0,
+      "learning_rate": 4.058386378303769e-07,
+      "loss": 0.6522,
+      "step": 2955
+    },
+    {
+      "epoch": 0.9119234922103964,
+      "grad_norm": 0.0,
+      "learning_rate": 4.0302475268377386e-07,
+      "loss": 0.6448,
+      "step": 2956
+    },
+    {
+      "epoch": 0.9122319913620237,
+      "grad_norm": 0.0,
+      "learning_rate": 4.00220455919037e-07,
+      "loss": 0.6592,
+      "step": 2957
+    },
+    {
+      "epoch": 0.912540490513651,
+      "grad_norm": 0.0,
+      "learning_rate": 3.974257503379508e-07,
+      "loss": 0.6448,
+      "step": 2958
+    },
+    {
+      "epoch": 0.9128489896652784,
+      "grad_norm": 0.0,
+      "learning_rate": 3.9464063873272064e-07,
+      "loss": 0.6848,
+      "step": 2959
+    },
+    {
+      "epoch": 0.9131574888169057,
+      "grad_norm": 0.0,
+      "learning_rate": 3.9186512388596053e-07,
+      "loss": 0.7141,
+      "step": 2960
+    },
+    {
+      "epoch": 0.9134659879685331,
+      "grad_norm": 0.0,
+      "learning_rate": 3.8909920857070237e-07,
+      "loss": 0.6235,
+      "step": 2961
+    },
+    {
+      "epoch": 0.9137744871201604,
+      "grad_norm": 0.0,
+      "learning_rate": 3.863428955503856e-07,
+      "loss": 0.6826,
+      "step": 2962
+    },
+    {
+      "epoch": 0.9140829862717877,
+      "grad_norm": 0.0,
+      "learning_rate": 3.835961875788552e-07,
+      "loss": 0.6908,
+      "step": 2963
+    },
+    {
+      "epoch": 0.9143914854234151,
+      "grad_norm": 0.0,
+      "learning_rate": 3.8085908740035706e-07,
+      "loss": 0.6579,
+      "step": 2964
+    },
+    {
+      "epoch": 0.9146999845750424,
+      "grad_norm": 0.0,
+      "learning_rate": 3.7813159774954256e-07,
+      "loss": 0.7451,
+      "step": 2965
+    },
+    {
+      "epoch": 0.9150084837266698,
+      "grad_norm": 0.0,
+      "learning_rate": 3.754137213514586e-07,
+      "loss": 0.6747,
+      "step": 2966
+    },
+    {
+      "epoch": 0.9153169828782971,
+      "grad_norm": 0.0,
+      "learning_rate": 3.7270546092154856e-07,
+      "loss": 0.7102,
+      "step": 2967
+    },
+    {
+      "epoch": 0.9156254820299244,
+      "grad_norm": 0.0,
+      "learning_rate": 3.7000681916564583e-07,
+      "loss": 0.7626,
+      "step": 2968
+    },
+    {
+      "epoch": 0.9159339811815518,
+      "grad_norm": 0.0,
+      "learning_rate": 3.673177987799781e-07,
+      "loss": 0.6208,
+      "step": 2969
+    },
+    {
+      "epoch": 0.9162424803331791,
+      "grad_norm": 0.0,
+      "learning_rate": 3.646384024511562e-07,
+      "loss": 0.7493,
+      "step": 2970
+    },
+    {
+      "epoch": 0.9165509794848065,
+      "grad_norm": 0.0,
+      "learning_rate": 3.6196863285618e-07,
+      "loss": 0.7094,
+      "step": 2971
+    },
+    {
+      "epoch": 0.9168594786364338,
+      "grad_norm": 0.0,
+      "learning_rate": 3.5930849266242465e-07,
+      "loss": 0.6915,
+      "step": 2972
+    },
+    {
+      "epoch": 0.917167977788061,
+      "grad_norm": 0.0,
+      "learning_rate": 3.5665798452764966e-07,
+      "loss": 0.6739,
+      "step": 2973
+    },
+    {
+      "epoch": 0.9174764769396884,
+      "grad_norm": 0.0,
+      "learning_rate": 3.540171110999913e-07,
+      "loss": 0.6801,
+      "step": 2974
+    },
+    {
+      "epoch": 0.9177849760913157,
+      "grad_norm": 0.0,
+      "learning_rate": 3.5138587501795884e-07,
+      "loss": 0.6038,
+      "step": 2975
+    },
+    {
+      "epoch": 0.918093475242943,
+      "grad_norm": 0.0,
+      "learning_rate": 3.487642789104295e-07,
+      "loss": 0.6319,
+      "step": 2976
+    },
+    {
+      "epoch": 0.9184019743945704,
+      "grad_norm": 0.0,
+      "learning_rate": 3.4615232539665254e-07,
+      "loss": 0.7123,
+      "step": 2977
+    },
+    {
+      "epoch": 0.9187104735461977,
+      "grad_norm": 0.0,
+      "learning_rate": 3.4355001708624493e-07,
+      "loss": 0.651,
+      "step": 2978
+    },
+    {
+      "epoch": 0.9190189726978251,
+      "grad_norm": 0.0,
+      "learning_rate": 3.4095735657918127e-07,
+      "loss": 0.6915,
+      "step": 2979
+    },
+    {
+      "epoch": 0.9193274718494524,
+      "grad_norm": 0.0,
+      "learning_rate": 3.3837434646580514e-07,
+      "loss": 0.8021,
+      "step": 2980
+    },
+    {
+      "epoch": 0.9196359710010797,
+      "grad_norm": 0.0,
+      "learning_rate": 3.3580098932680994e-07,
+      "loss": 0.6535,
+      "step": 2981
+    },
+    {
+      "epoch": 0.9199444701527071,
+      "grad_norm": 0.0,
+      "learning_rate": 3.3323728773324904e-07,
+      "loss": 0.6707,
+      "step": 2982
+    },
+    {
+      "epoch": 0.9202529693043344,
+      "grad_norm": 0.0,
+      "learning_rate": 3.306832442465302e-07,
+      "loss": 0.6688,
+      "step": 2983
+    },
+    {
+      "epoch": 0.9205614684559618,
+      "grad_norm": 0.0,
+      "learning_rate": 3.281388614184089e-07,
+      "loss": 0.6877,
+      "step": 2984
+    },
+    {
+      "epoch": 0.9208699676075891,
+      "grad_norm": 0.0,
+      "learning_rate": 3.256041417909894e-07,
+      "loss": 0.6977,
+      "step": 2985
+    },
+    {
+      "epoch": 0.9211784667592164,
+      "grad_norm": 0.0,
+      "learning_rate": 3.230790878967216e-07,
+      "loss": 0.6933,
+      "step": 2986
+    },
+    {
+      "epoch": 0.9214869659108438,
+      "grad_norm": 0.0,
+      "learning_rate": 3.205637022583985e-07,
+      "loss": 0.6552,
+      "step": 2987
+    },
+    {
+      "epoch": 0.9217954650624711,
+      "grad_norm": 0.0,
+      "learning_rate": 3.180579873891565e-07,
+      "loss": 0.6738,
+      "step": 2988
+    },
+    {
+      "epoch": 0.9221039642140985,
+      "grad_norm": 0.0,
+      "learning_rate": 3.155619457924608e-07,
+      "loss": 0.6119,
+      "step": 2989
+    },
+    {
+      "epoch": 0.9224124633657258,
+      "grad_norm": 0.0,
+      "learning_rate": 3.130755799621221e-07,
+      "loss": 0.643,
+      "step": 2990
+    },
+    {
+      "epoch": 0.922720962517353,
+      "grad_norm": 0.0,
+      "learning_rate": 3.1059889238227893e-07,
+      "loss": 0.6854,
+      "step": 2991
+    },
+    {
+      "epoch": 0.9230294616689804,
+      "grad_norm": 0.0,
+      "learning_rate": 3.0813188552740067e-07,
+      "loss": 0.6532,
+      "step": 2992
+    },
+    {
+      "epoch": 0.9233379608206077,
+      "grad_norm": 0.0,
+      "learning_rate": 3.0567456186228694e-07,
+      "loss": 0.6868,
+      "step": 2993
+    },
+    {
+      "epoch": 0.923646459972235,
+      "grad_norm": 0.0,
+      "learning_rate": 3.032269238420582e-07,
+      "loss": 0.7246,
+      "step": 2994
+    },
+    {
+      "epoch": 0.9239549591238624,
+      "grad_norm": 0.0,
+      "learning_rate": 3.0078897391216387e-07,
+      "loss": 0.6767,
+      "step": 2995
+    },
+    {
+      "epoch": 0.9242634582754897,
+      "grad_norm": 0.0,
+      "learning_rate": 2.9836071450836776e-07,
+      "loss": 0.6672,
+      "step": 2996
+    },
+    {
+      "epoch": 0.9245719574271171,
+      "grad_norm": 0.0,
+      "learning_rate": 2.9594214805675703e-07,
+      "loss": 0.6284,
+      "step": 2997
+    },
+    {
+      "epoch": 0.9248804565787444,
+      "grad_norm": 0.0,
+      "learning_rate": 2.935332769737331e-07,
+      "loss": 0.703,
+      "step": 2998
+    },
+    {
+      "epoch": 0.9251889557303717,
+      "grad_norm": 0.0,
+      "learning_rate": 2.911341036660065e-07,
+      "loss": 0.7389,
+      "step": 2999
+    },
+    {
+      "epoch": 0.9254974548819991,
+      "grad_norm": 0.0,
+      "learning_rate": 2.887446305306074e-07,
+      "loss": 0.6566,
+      "step": 3000
+    },
+    {
+      "epoch": 0.9258059540336264,
+      "grad_norm": 0.0,
+      "learning_rate": 2.863648599548652e-07,
+      "loss": 0.6818,
+      "step": 3001
+    },
+    {
+      "epoch": 0.9261144531852538,
+      "grad_norm": 0.0,
+      "learning_rate": 2.839947943164223e-07,
+      "loss": 0.6706,
+      "step": 3002
+    },
+    {
+      "epoch": 0.9264229523368811,
+      "grad_norm": 0.0,
+      "learning_rate": 2.81634435983219e-07,
+      "loss": 0.6665,
+      "step": 3003
+    },
+    {
+      "epoch": 0.9267314514885084,
+      "grad_norm": 0.0,
+      "learning_rate": 2.792837873135035e-07,
+      "loss": 0.6966,
+      "step": 3004
+    },
+    {
+      "epoch": 0.9270399506401358,
+      "grad_norm": 0.0,
+      "learning_rate": 2.7694285065581807e-07,
+      "loss": 0.7047,
+      "step": 3005
+    },
+    {
+      "epoch": 0.9273484497917631,
+      "grad_norm": 0.0,
+      "learning_rate": 2.746116283490019e-07,
+      "loss": 0.7193,
+      "step": 3006
+    },
+    {
+      "epoch": 0.9276569489433905,
+      "grad_norm": 0.0,
+      "learning_rate": 2.722901227221919e-07,
+      "loss": 0.6711,
+      "step": 3007
+    },
+    {
+      "epoch": 0.9279654480950177,
+      "grad_norm": 0.0,
+      "learning_rate": 2.6997833609481384e-07,
+      "loss": 0.6848,
+      "step": 3008
+    },
+    {
+      "epoch": 0.928273947246645,
+      "grad_norm": 0.0,
+      "learning_rate": 2.6767627077658563e-07,
+      "loss": 0.6251,
+      "step": 3009
+    },
+    {
+      "epoch": 0.9285824463982724,
+      "grad_norm": 0.0,
+      "learning_rate": 2.653839290675109e-07,
+      "loss": 0.6504,
+      "step": 3010
+    },
+    {
+      "epoch": 0.9288909455498997,
+      "grad_norm": 0.0,
+      "learning_rate": 2.6310131325787634e-07,
+      "loss": 0.5879,
+      "step": 3011
+    },
+    {
+      "epoch": 0.929199444701527,
+      "grad_norm": 0.0,
+      "learning_rate": 2.608284256282567e-07,
+      "loss": 0.7294,
+      "step": 3012
+    },
+    {
+      "epoch": 0.9295079438531544,
+      "grad_norm": 0.0,
+      "learning_rate": 2.5856526844950324e-07,
+      "loss": 0.7099,
+      "step": 3013
+    },
+    {
+      "epoch": 0.9298164430047817,
+      "grad_norm": 0.0,
+      "learning_rate": 2.5631184398274834e-07,
+      "loss": 0.6724,
+      "step": 3014
+    },
+    {
+      "epoch": 0.9301249421564091,
+      "grad_norm": 0.0,
+      "learning_rate": 2.540681544793955e-07,
+      "loss": 0.7407,
+      "step": 3015
+    },
+    {
+      "epoch": 0.9304334413080364,
+      "grad_norm": 0.0,
+      "learning_rate": 2.518342021811271e-07,
+      "loss": 0.6528,
+      "step": 3016
+    },
+    {
+      "epoch": 0.9307419404596637,
+      "grad_norm": 0.0,
+      "learning_rate": 2.496099893198944e-07,
+      "loss": 0.7048,
+      "step": 3017
+    },
+    {
+      "epoch": 0.9310504396112911,
+      "grad_norm": 0.0,
+      "learning_rate": 2.4739551811791996e-07,
+      "loss": 0.7183,
+      "step": 3018
+    },
+    {
+      "epoch": 0.9313589387629184,
+      "grad_norm": 0.0,
+      "learning_rate": 2.451907907876916e-07,
+      "loss": 0.6742,
+      "step": 3019
+    },
+    {
+      "epoch": 0.9316674379145458,
+      "grad_norm": 0.0,
+      "learning_rate": 2.4299580953196066e-07,
+      "loss": 0.6479,
+      "step": 3020
+    },
+    {
+      "epoch": 0.9319759370661731,
+      "grad_norm": 0.0,
+      "learning_rate": 2.408105765437452e-07,
+      "loss": 0.6814,
+      "step": 3021
+    },
+    {
+      "epoch": 0.9322844362178004,
+      "grad_norm": 0.0,
+      "learning_rate": 2.386350940063209e-07,
+      "loss": 0.6509,
+      "step": 3022
+    },
+    {
+      "epoch": 0.9325929353694278,
+      "grad_norm": 0.0,
+      "learning_rate": 2.364693640932214e-07,
+      "loss": 0.6871,
+      "step": 3023
+    },
+    {
+      "epoch": 0.9329014345210551,
+      "grad_norm": 0.0,
+      "learning_rate": 2.3431338896823917e-07,
+      "loss": 0.6811,
+      "step": 3024
+    },
+    {
+      "epoch": 0.9332099336726823,
+      "grad_norm": 0.0,
+      "learning_rate": 2.321671707854156e-07,
+      "loss": 0.6711,
+      "step": 3025
+    },
+    {
+      "epoch": 0.9335184328243097,
+      "grad_norm": 0.0,
+      "learning_rate": 2.300307116890521e-07,
+      "loss": 0.6324,
+      "step": 3026
+    },
+    {
+      "epoch": 0.933826931975937,
+      "grad_norm": 0.0,
+      "learning_rate": 2.279040138136901e-07,
+      "loss": 0.6814,
+      "step": 3027
+    },
+    {
+      "epoch": 0.9341354311275644,
+      "grad_norm": 0.0,
+      "learning_rate": 2.2578707928412545e-07,
+      "loss": 0.6596,
+      "step": 3028
+    },
+    {
+      "epoch": 0.9344439302791917,
+      "grad_norm": 0.0,
+      "learning_rate": 2.2367991021539637e-07,
+      "loss": 0.679,
+      "step": 3029
+    },
+    {
+      "epoch": 0.934752429430819,
+      "grad_norm": 0.0,
+      "learning_rate": 2.215825087127843e-07,
+      "loss": 0.601,
+      "step": 3030
+    },
+    {
+      "epoch": 0.9350609285824464,
+      "grad_norm": 0.0,
+      "learning_rate": 2.1949487687181525e-07,
+      "loss": 0.691,
+      "step": 3031
+    },
+    {
+      "epoch": 0.9353694277340737,
+      "grad_norm": 0.0,
+      "learning_rate": 2.1741701677824966e-07,
+      "loss": 0.7404,
+      "step": 3032
+    },
+    {
+      "epoch": 0.9356779268857011,
+      "grad_norm": 0.0,
+      "learning_rate": 2.1534893050808802e-07,
+      "loss": 0.654,
+      "step": 3033
+    },
+    {
+      "epoch": 0.9359864260373284,
+      "grad_norm": 0.0,
+      "learning_rate": 2.1329062012756308e-07,
+      "loss": 0.622,
+      "step": 3034
+    },
+    {
+      "epoch": 0.9362949251889557,
+      "grad_norm": 0.0,
+      "learning_rate": 2.1124208769314424e-07,
+      "loss": 0.7046,
+      "step": 3035
+    },
+    {
+      "epoch": 0.9366034243405831,
+      "grad_norm": 0.0,
+      "learning_rate": 2.092033352515288e-07,
+      "loss": 0.6661,
+      "step": 3036
+    },
+    {
+      "epoch": 0.9369119234922104,
+      "grad_norm": 0.0,
+      "learning_rate": 2.0717436483964293e-07,
+      "loss": 0.7741,
+      "step": 3037
+    },
+    {
+      "epoch": 0.9372204226438378,
+      "grad_norm": 0.0,
+      "learning_rate": 2.0515517848464284e-07,
+      "loss": 0.6348,
+      "step": 3038
+    },
+    {
+      "epoch": 0.9375289217954651,
+      "grad_norm": 0.0,
+      "learning_rate": 2.031457782039037e-07,
+      "loss": 0.6986,
+      "step": 3039
+    },
+    {
+      "epoch": 0.9378374209470924,
+      "grad_norm": 0.0,
+      "learning_rate": 2.0114616600502845e-07,
+      "loss": 0.6558,
+      "step": 3040
+    },
+    {
+      "epoch": 0.9381459200987198,
+      "grad_norm": 0.0,
+      "learning_rate": 1.991563438858368e-07,
+      "loss": 0.6811,
+      "step": 3041
+    },
+    {
+      "epoch": 0.9384544192503471,
+      "grad_norm": 0.0,
+      "learning_rate": 1.9717631383437075e-07,
+      "loss": 0.6629,
+      "step": 3042
+    },
+    {
+      "epoch": 0.9387629184019743,
+      "grad_norm": 0.0,
+      "learning_rate": 1.952060778288889e-07,
+      "loss": 0.6583,
+      "step": 3043
+    },
+    {
+      "epoch": 0.9390714175536017,
+      "grad_norm": 0.0,
+      "learning_rate": 1.9324563783785888e-07,
+      "loss": 0.6649,
+      "step": 3044
+    },
+    {
+      "epoch": 0.939379916705229,
+      "grad_norm": 0.0,
+      "learning_rate": 1.9129499581996945e-07,
+      "loss": 0.7522,
+      "step": 3045
+    },
+    {
+      "epoch": 0.9396884158568564,
+      "grad_norm": 0.0,
+      "learning_rate": 1.893541537241128e-07,
+      "loss": 0.6718,
+      "step": 3046
+    },
+    {
+      "epoch": 0.9399969150084837,
+      "grad_norm": 0.0,
+      "learning_rate": 1.8742311348939558e-07,
+      "loss": 0.6987,
+      "step": 3047
+    },
+    {
+      "epoch": 0.940305414160111,
+      "grad_norm": 0.0,
+      "learning_rate": 1.855018770451278e-07,
+      "loss": 0.6032,
+      "step": 3048
+    },
+    {
+      "epoch": 0.9406139133117384,
+      "grad_norm": 0.0,
+      "learning_rate": 1.83590446310824e-07,
+      "loss": 0.701,
+      "step": 3049
+    },
+    {
+      "epoch": 0.9409224124633657,
+      "grad_norm": 0.0,
+      "learning_rate": 1.8168882319620663e-07,
+      "loss": 0.7369,
+      "step": 3050
+    },
+    {
+      "epoch": 0.9412309116149931,
+      "grad_norm": 0.0,
+      "learning_rate": 1.797970096011936e-07,
+      "loss": 0.7331,
+      "step": 3051
+    },
+    {
+      "epoch": 0.9415394107666204,
+      "grad_norm": 0.0,
+      "learning_rate": 1.779150074159064e-07,
+      "loss": 0.6953,
+      "step": 3052
+    },
+    {
+      "epoch": 0.9418479099182477,
+      "grad_norm": 0.0,
+      "learning_rate": 1.7604281852065973e-07,
+      "loss": 0.7046,
+      "step": 3053
+    },
+    {
+      "epoch": 0.9421564090698751,
+      "grad_norm": 0.0,
+      "learning_rate": 1.7418044478596742e-07,
+      "loss": 0.7163,
+      "step": 3054
+    },
+    {
+      "epoch": 0.9424649082215024,
+      "grad_norm": 0.0,
+      "learning_rate": 1.7232788807253654e-07,
+      "loss": 0.6754,
+      "step": 3055
+    },
+    {
+      "epoch": 0.9427734073731298,
+      "grad_norm": 0.0,
+      "learning_rate": 1.7048515023126323e-07,
+      "loss": 0.6923,
+      "step": 3056
+    },
+    {
+      "epoch": 0.9430819065247571,
+      "grad_norm": 0.0,
+      "learning_rate": 1.6865223310323586e-07,
+      "loss": 0.6807,
+      "step": 3057
+    },
+    {
+      "epoch": 0.9433904056763844,
+      "grad_norm": 0.0,
+      "learning_rate": 1.6682913851972959e-07,
+      "loss": 0.6595,
+      "step": 3058
+    },
+    {
+      "epoch": 0.9436989048280118,
+      "grad_norm": 0.0,
+      "learning_rate": 1.6501586830220852e-07,
+      "loss": 0.7242,
+      "step": 3059
+    },
+    {
+      "epoch": 0.944007403979639,
+      "grad_norm": 0.0,
+      "learning_rate": 1.6321242426231787e-07,
+      "loss": 0.9792,
+      "step": 3060
+    },
+    {
+      "epoch": 0.9443159031312663,
+      "grad_norm": 0.0,
+      "learning_rate": 1.6141880820188749e-07,
+      "loss": 0.6886,
+      "step": 3061
+    },
+    {
+      "epoch": 0.9446244022828937,
+      "grad_norm": 0.0,
+      "learning_rate": 1.596350219129261e-07,
+      "loss": 0.6266,
+      "step": 3062
+    },
+    {
+      "epoch": 0.944932901434521,
+      "grad_norm": 0.0,
+      "learning_rate": 1.5786106717762373e-07,
+      "loss": 0.6921,
+      "step": 3063
+    },
+    {
+      "epoch": 0.9452414005861484,
+      "grad_norm": 0.0,
+      "learning_rate": 1.5609694576834698e-07,
+      "loss": 0.682,
+      "step": 3064
+    },
+    {
+      "epoch": 0.9455498997377757,
+      "grad_norm": 0.0,
+      "learning_rate": 1.5434265944763717e-07,
+      "loss": 0.609,
+      "step": 3065
+    },
+    {
+      "epoch": 0.945858398889403,
+      "grad_norm": 0.0,
+      "learning_rate": 1.5259820996820884e-07,
+      "loss": 0.6423,
+      "step": 3066
+    },
+    {
+      "epoch": 0.9461668980410304,
+      "grad_norm": 0.0,
+      "learning_rate": 1.5086359907295124e-07,
+      "loss": 0.6499,
+      "step": 3067
+    },
+    {
+      "epoch": 0.9464753971926577,
+      "grad_norm": 0.0,
+      "learning_rate": 1.4913882849492022e-07,
+      "loss": 0.6977,
+      "step": 3068
+    },
+    {
+      "epoch": 0.9467838963442851,
+      "grad_norm": 0.0,
+      "learning_rate": 1.4742389995734519e-07,
+      "loss": 0.7011,
+      "step": 3069
+    },
+    {
+      "epoch": 0.9470923954959124,
+      "grad_norm": 0.0,
+      "learning_rate": 1.457188151736144e-07,
+      "loss": 0.6401,
+      "step": 3070
+    },
+    {
+      "epoch": 0.9474008946475397,
+      "grad_norm": 0.0,
+      "learning_rate": 1.4402357584728966e-07,
+      "loss": 0.7007,
+      "step": 3071
+    },
+    {
+      "epoch": 0.9477093937991671,
+      "grad_norm": 0.0,
+      "learning_rate": 1.423381836720894e-07,
+      "loss": 0.6718,
+      "step": 3072
+    },
+    {
+      "epoch": 0.9480178929507944,
+      "grad_norm": 0.0,
+      "learning_rate": 1.4066264033190002e-07,
+      "loss": 0.6217,
+      "step": 3073
+    },
+    {
+      "epoch": 0.9483263921024218,
+      "grad_norm": 0.0,
+      "learning_rate": 1.3899694750076241e-07,
+      "loss": 0.6629,
+      "step": 3074
+    },
+    {
+      "epoch": 0.9486348912540491,
+      "grad_norm": 0.0,
+      "learning_rate": 1.3734110684287761e-07,
+      "loss": 0.6666,
+      "step": 3075
+    },
+    {
+      "epoch": 0.9489433904056764,
+      "grad_norm": 0.0,
+      "learning_rate": 1.3569512001260444e-07,
+      "loss": 0.7184,
+      "step": 3076
+    },
+    {
+      "epoch": 0.9492518895573037,
+      "grad_norm": 0.0,
+      "learning_rate": 1.3405898865445522e-07,
+      "loss": 0.6494,
+      "step": 3077
+    },
+    {
+      "epoch": 0.949560388708931,
+      "grad_norm": 0.0,
+      "learning_rate": 1.3243271440309679e-07,
+      "loss": 0.7152,
+      "step": 3078
+    },
+    {
+      "epoch": 0.9498688878605583,
+      "grad_norm": 0.0,
+      "learning_rate": 1.3081629888334834e-07,
+      "loss": 0.6792,
+      "step": 3079
+    },
+    {
+      "epoch": 0.9501773870121857,
+      "grad_norm": 0.0,
+      "learning_rate": 1.292097437101747e-07,
+      "loss": 0.6981,
+      "step": 3080
+    },
+    {
+      "epoch": 0.950485886163813,
+      "grad_norm": 0.0,
+      "learning_rate": 1.276130504886963e-07,
+      "loss": 0.6675,
+      "step": 3081
+    },
+    {
+      "epoch": 0.9507943853154404,
+      "grad_norm": 0.0,
+      "learning_rate": 1.260262208141738e-07,
+      "loss": 0.6856,
+      "step": 3082
+    },
+    {
+      "epoch": 0.9511028844670677,
+      "grad_norm": 0.0,
+      "learning_rate": 1.2444925627201786e-07,
+      "loss": 0.7035,
+      "step": 3083
+    },
+    {
+      "epoch": 0.951411383618695,
+      "grad_norm": 0.0,
+      "learning_rate": 1.2288215843777928e-07,
+      "loss": 0.652,
+      "step": 3084
+    },
+    {
+      "epoch": 0.9517198827703224,
+      "grad_norm": 0.0,
+      "learning_rate": 1.2132492887715454e-07,
+      "loss": 0.7783,
+      "step": 3085
+    },
+    {
+      "epoch": 0.9520283819219497,
+      "grad_norm": 0.0,
+      "learning_rate": 1.1977756914597792e-07,
+      "loss": 0.682,
+      "step": 3086
+    },
+    {
+      "epoch": 0.9523368810735771,
+      "grad_norm": 0.0,
+      "learning_rate": 1.1824008079022286e-07,
+      "loss": 0.6684,
+      "step": 3087
+    },
+    {
+      "epoch": 0.9526453802252044,
+      "grad_norm": 0.0,
+      "learning_rate": 1.1671246534600167e-07,
+      "loss": 0.6782,
+      "step": 3088
+    },
+    {
+      "epoch": 0.9529538793768317,
+      "grad_norm": 0.0,
+      "learning_rate": 1.1519472433956125e-07,
+      "loss": 0.6979,
+      "step": 3089
+    },
+    {
+      "epoch": 0.9532623785284591,
+      "grad_norm": 0.0,
+      "learning_rate": 1.136868592872864e-07,
+      "loss": 0.6788,
+      "step": 3090
+    },
+    {
+      "epoch": 0.9535708776800864,
+      "grad_norm": 0.0,
+      "learning_rate": 1.1218887169568871e-07,
+      "loss": 0.7168,
+      "step": 3091
+    },
+    {
+      "epoch": 0.9538793768317138,
+      "grad_norm": 0.0,
+      "learning_rate": 1.107007630614143e-07,
+      "loss": 0.6758,
+      "step": 3092
+    },
+    {
+      "epoch": 0.9541878759833411,
+      "grad_norm": 0.0,
+      "learning_rate": 1.0922253487124279e-07,
+      "loss": 0.7599,
+      "step": 3093
+    },
+    {
+      "epoch": 0.9544963751349684,
+      "grad_norm": 0.0,
+      "learning_rate": 1.0775418860207498e-07,
+      "loss": 0.6001,
+      "step": 3094
+    },
+    {
+      "epoch": 0.9548048742865957,
+      "grad_norm": 0.0,
+      "learning_rate": 1.0629572572094515e-07,
+      "loss": 0.6489,
+      "step": 3095
+    },
+    {
+      "epoch": 0.955113373438223,
+      "grad_norm": 0.0,
+      "learning_rate": 1.048471476850077e-07,
+      "loss": 0.6668,
+      "step": 3096
+    },
+    {
+      "epoch": 0.9554218725898503,
+      "grad_norm": 0.0,
+      "learning_rate": 1.0340845594154603e-07,
+      "loss": 0.6543,
+      "step": 3097
+    },
+    {
+      "epoch": 0.9557303717414777,
+      "grad_norm": 0.0,
+      "learning_rate": 1.0197965192796145e-07,
+      "loss": 0.7041,
+      "step": 3098
+    },
+    {
+      "epoch": 0.956038870893105,
+      "grad_norm": 0.0,
+      "learning_rate": 1.0056073707177982e-07,
+      "loss": 0.7049,
+      "step": 3099
+    },
+    {
+      "epoch": 0.9563473700447324,
+      "grad_norm": 0.0,
+      "learning_rate": 9.915171279064606e-08,
+      "loss": 0.7802,
+      "step": 3100
+    },
+    {
+      "epoch": 0.9566558691963597,
+      "grad_norm": 0.0,
+      "learning_rate": 9.775258049232072e-08,
+      "loss": 0.6623,
+      "step": 3101
+    },
+    {
+      "epoch": 0.956964368347987,
+      "grad_norm": 0.0,
+      "learning_rate": 9.636334157468563e-08,
+      "loss": 0.7155,
+      "step": 3102
+    },
+    {
+      "epoch": 0.9572728674996144,
+      "grad_norm": 0.0,
+      "learning_rate": 9.498399742573495e-08,
+      "loss": 0.7015,
+      "step": 3103
+    },
+    {
+      "epoch": 0.9575813666512417,
+      "grad_norm": 0.0,
+      "learning_rate": 9.361454942357628e-08,
+      "loss": 0.6937,
+      "step": 3104
+    },
+    {
+      "epoch": 0.9578898658028691,
+      "grad_norm": 0.0,
+      "learning_rate": 9.225499893643297e-08,
+      "loss": 0.6619,
+      "step": 3105
+    },
+    {
+      "epoch": 0.9581983649544964,
+      "grad_norm": 0.0,
+      "learning_rate": 9.090534732263734e-08,
+      "loss": 0.6644,
+      "step": 3106
+    },
+    {
+      "epoch": 0.9585068641061237,
+      "grad_norm": 0.0,
+      "learning_rate": 8.956559593063408e-08,
+      "loss": 0.6596,
+      "step": 3107
+    },
+    {
+      "epoch": 0.9588153632577511,
+      "grad_norm": 0.0,
+      "learning_rate": 8.823574609897134e-08,
+      "loss": 0.6561,
+      "step": 3108
+    },
+    {
+      "epoch": 0.9591238624093784,
+      "grad_norm": 0.0,
+      "learning_rate": 8.691579915631299e-08,
+      "loss": 0.6519,
+      "step": 3109
+    },
+    {
+      "epoch": 0.9594323615610058,
+      "grad_norm": 0.0,
+      "learning_rate": 8.560575642141966e-08,
+      "loss": 0.7131,
+      "step": 3110
+    },
+    {
+      "epoch": 0.9597408607126331,
+      "grad_norm": 0.0,
+      "learning_rate": 8.430561920316438e-08,
+      "loss": 0.755,
+      "step": 3111
+    },
+    {
+      "epoch": 0.9600493598642603,
+      "grad_norm": 0.0,
+      "learning_rate": 8.301538880051808e-08,
+      "loss": 0.6812,
+      "step": 3112
+    },
+    {
+      "epoch": 0.9603578590158877,
+      "grad_norm": 0.0,
+      "learning_rate": 8.173506650255625e-08,
+      "loss": 0.6656,
+      "step": 3113
+    },
+    {
+      "epoch": 0.960666358167515,
+      "grad_norm": 0.0,
+      "learning_rate": 8.046465358845568e-08,
+      "loss": 0.6969,
+      "step": 3114
+    },
+    {
+      "epoch": 0.9609748573191423,
+      "grad_norm": 0.0,
+      "learning_rate": 7.920415132748993e-08,
+      "loss": 0.6985,
+      "step": 3115
+    },
+    {
+      "epoch": 0.9612833564707697,
+      "grad_norm": 0.0,
+      "learning_rate": 7.795356097903495e-08,
+      "loss": 0.654,
+      "step": 3116
+    },
+    {
+      "epoch": 0.961591855622397,
+      "grad_norm": 0.0,
+      "learning_rate": 7.671288379256015e-08,
+      "loss": 0.783,
+      "step": 3117
+    },
+    {
+      "epoch": 0.9619003547740244,
+      "grad_norm": 0.0,
+      "learning_rate": 7.548212100763063e-08,
+      "loss": 0.7332,
+      "step": 3118
+    },
+    {
+      "epoch": 0.9622088539256517,
+      "grad_norm": 0.0,
+      "learning_rate": 7.426127385390947e-08,
+      "loss": 0.6841,
+      "step": 3119
+    },
+    {
+      "epoch": 0.962517353077279,
+      "grad_norm": 0.0,
+      "learning_rate": 7.30503435511487e-08,
+      "loss": 0.6868,
+      "step": 3120
+    },
+    {
+      "epoch": 0.9628258522289064,
+      "grad_norm": 0.0,
+      "learning_rate": 7.184933130919614e-08,
+      "loss": 0.6421,
+      "step": 3121
+    },
+    {
+      "epoch": 0.9631343513805337,
+      "grad_norm": 0.0,
+      "learning_rate": 7.065823832798524e-08,
+      "loss": 0.7666,
+      "step": 3122
+    },
+    {
+      "epoch": 0.963442850532161,
+      "grad_norm": 0.0,
+      "learning_rate": 6.947706579754632e-08,
+      "loss": 0.701,
+      "step": 3123
+    },
+    {
+      "epoch": 0.9637513496837884,
+      "grad_norm": 0.0,
+      "learning_rate": 6.830581489799204e-08,
+      "loss": 1.0016,
+      "step": 3124
+    },
+    {
+      "epoch": 0.9640598488354157,
+      "grad_norm": 0.0,
+      "learning_rate": 6.714448679952524e-08,
+      "loss": 0.6931,
+      "step": 3125
+    },
+    {
+      "epoch": 0.9643683479870431,
+      "grad_norm": 0.0,
+      "learning_rate": 6.599308266243443e-08,
+      "loss": 0.6796,
+      "step": 3126
+    },
+    {
+      "epoch": 0.9646768471386704,
+      "grad_norm": 0.0,
+      "learning_rate": 6.485160363709053e-08,
+      "loss": 0.6407,
+      "step": 3127
+    },
+    {
+      "epoch": 0.9649853462902978,
+      "grad_norm": 0.0,
+      "learning_rate": 6.37200508639535e-08,
+      "loss": 0.7245,
+      "step": 3128
+    },
+    {
+      "epoch": 0.965293845441925,
+      "grad_norm": 0.0,
+      "learning_rate": 6.259842547356231e-08,
+      "loss": 0.6961,
+      "step": 3129
+    },
+    {
+      "epoch": 0.9656023445935523,
+      "grad_norm": 0.0,
+      "learning_rate": 6.148672858653615e-08,
+      "loss": 0.7058,
+      "step": 3130
+    },
+    {
+      "epoch": 0.9659108437451797,
+      "grad_norm": 0.0,
+      "learning_rate": 6.038496131357874e-08,
+      "loss": 0.6755,
+      "step": 3131
+    },
+    {
+      "epoch": 0.966219342896807,
+      "grad_norm": 0.0,
+      "learning_rate": 5.929312475546845e-08,
+      "loss": 0.7858,
+      "step": 3132
+    },
+    {
+      "epoch": 0.9665278420484343,
+      "grad_norm": 0.0,
+      "learning_rate": 5.8211220003067114e-08,
+      "loss": 0.6399,
+      "step": 3133
+    },
+    {
+      "epoch": 0.9668363412000617,
+      "grad_norm": 0.0,
+      "learning_rate": 5.7139248137307865e-08,
+      "loss": 0.6506,
+      "step": 3134
+    },
+    {
+      "epoch": 0.967144840351689,
+      "grad_norm": 0.0,
+      "learning_rate": 5.607721022920398e-08,
+      "loss": 0.6636,
+      "step": 3135
+    },
+    {
+      "epoch": 0.9674533395033164,
+      "grad_norm": 0.0,
+      "learning_rate": 5.5025107339842234e-08,
+      "loss": 0.7279,
+      "step": 3136
+    },
+    {
+      "epoch": 0.9677618386549437,
+      "grad_norm": 0.0,
+      "learning_rate": 5.3982940520382885e-08,
+      "loss": 0.659,
+      "step": 3137
+    },
+    {
+      "epoch": 0.968070337806571,
+      "grad_norm": 0.0,
+      "learning_rate": 5.295071081206194e-08,
+      "loss": 0.6571,
+      "step": 3138
+    },
+    {
+      "epoch": 0.9683788369581984,
+      "grad_norm": 0.0,
+      "learning_rate": 5.1928419246181085e-08,
+      "loss": 0.6417,
+      "step": 3139
+    },
+    {
+      "epoch": 0.9686873361098257,
+      "grad_norm": 0.0,
+      "learning_rate": 5.091606684411998e-08,
+      "loss": 0.6781,
+      "step": 3140
+    },
+    {
+      "epoch": 0.968995835261453,
+      "grad_norm": 0.0,
+      "learning_rate": 4.9913654617322894e-08,
+      "loss": 0.7228,
+      "step": 3141
+    },
+    {
+      "epoch": 0.9693043344130804,
+      "grad_norm": 0.0,
+      "learning_rate": 4.892118356730313e-08,
+      "loss": 0.9788,
+      "step": 3142
+    },
+    {
+      "epoch": 0.9696128335647077,
+      "grad_norm": 0.0,
+      "learning_rate": 4.79386546856464e-08,
+      "loss": 0.6991,
+      "step": 3143
+    },
+    {
+      "epoch": 0.9699213327163351,
+      "grad_norm": 0.0,
+      "learning_rate": 4.696606895399858e-08,
+      "loss": 0.6679,
+      "step": 3144
+    },
+    {
+      "epoch": 0.9702298318679624,
+      "grad_norm": 0.0,
+      "learning_rate": 4.60034273440757e-08,
+      "loss": 0.7438,
+      "step": 3145
+    },
+    {
+      "epoch": 0.9705383310195898,
+      "grad_norm": 0.0,
+      "learning_rate": 4.505073081765843e-08,
+      "loss": 0.6752,
+      "step": 3146
+    },
+    {
+      "epoch": 0.970846830171217,
+      "grad_norm": 0.0,
+      "learning_rate": 4.4107980326589806e-08,
+      "loss": 0.6864,
+      "step": 3147
+    },
+    {
+      "epoch": 0.9711553293228443,
+      "grad_norm": 0.0,
+      "learning_rate": 4.317517681277528e-08,
+      "loss": 0.6253,
+      "step": 3148
+    },
+    {
+      "epoch": 0.9714638284744717,
+      "grad_norm": 0.0,
+      "learning_rate": 4.225232120818268e-08,
+      "loss": 0.6613,
+      "step": 3149
+    },
+    {
+      "epoch": 0.971772327626099,
+      "grad_norm": 0.0,
+      "learning_rate": 4.133941443484335e-08,
+      "loss": 0.6816,
+      "step": 3150
+    },
+    {
+      "epoch": 0.9720808267777263,
+      "grad_norm": 0.0,
+      "learning_rate": 4.043645740484436e-08,
+      "loss": 0.6524,
+      "step": 3151
+    },
+    {
+      "epoch": 0.9723893259293537,
+      "grad_norm": 0.0,
+      "learning_rate": 3.95434510203363e-08,
+      "loss": 0.6208,
+      "step": 3152
+    },
+    {
+      "epoch": 0.972697825080981,
+      "grad_norm": 0.0,
+      "learning_rate": 3.866039617352324e-08,
+      "loss": 0.7372,
+      "step": 3153
+    },
+    {
+      "epoch": 0.9730063242326084,
+      "grad_norm": 0.0,
+      "learning_rate": 3.778729374667278e-08,
+      "loss": 0.7214,
+      "step": 3154
+    },
+    {
+      "epoch": 0.9733148233842357,
+      "grad_norm": 0.0,
+      "learning_rate": 3.6924144612102695e-08,
+      "loss": 0.6511,
+      "step": 3155
+    },
+    {
+      "epoch": 0.973623322535863,
+      "grad_norm": 0.0,
+      "learning_rate": 3.6070949632190934e-08,
+      "loss": 0.6205,
+      "step": 3156
+    },
+    {
+      "epoch": 0.9739318216874904,
+      "grad_norm": 0.0,
+      "learning_rate": 3.5227709659367834e-08,
+      "loss": 0.7322,
+      "step": 3157
+    },
+    {
+      "epoch": 0.9742403208391177,
+      "grad_norm": 0.0,
+      "learning_rate": 3.439442553611727e-08,
+      "loss": 0.9699,
+      "step": 3158
+    },
+    {
+      "epoch": 0.974548819990745,
+      "grad_norm": 0.0,
+      "learning_rate": 3.3571098094978825e-08,
+      "loss": 0.6859,
+      "step": 3159
+    },
+    {
+      "epoch": 0.9748573191423724,
+      "grad_norm": 0.0,
+      "learning_rate": 3.275772815854228e-08,
+      "loss": 0.69,
+      "step": 3160
+    },
+    {
+      "epoch": 0.9751658182939997,
+      "grad_norm": 0.0,
+      "learning_rate": 3.195431653944869e-08,
+      "loss": 0.6999,
+      "step": 3161
+    },
+    {
+      "epoch": 0.9754743174456271,
+      "grad_norm": 0.0,
+      "learning_rate": 3.1160864040391534e-08,
+      "loss": 0.6989,
+      "step": 3162
+    },
+    {
+      "epoch": 0.9757828165972544,
+      "grad_norm": 0.0,
+      "learning_rate": 3.0377371454112234e-08,
+      "loss": 0.6986,
+      "step": 3163
+    },
+    {
+      "epoch": 0.9760913157488816,
+      "grad_norm": 0.0,
+      "learning_rate": 2.9603839563400183e-08,
+      "loss": 0.6149,
+      "step": 3164
+    },
+    {
+      "epoch": 0.976399814900509,
+      "grad_norm": 0.0,
+      "learning_rate": 2.8840269141097165e-08,
+      "loss": 0.729,
+      "step": 3165
+    },
+    {
+      "epoch": 0.9767083140521363,
+      "grad_norm": 0.0,
+      "learning_rate": 2.8086660950088497e-08,
+      "loss": 0.6686,
+      "step": 3166
+    },
+    {
+      "epoch": 0.9770168132037637,
+      "grad_norm": 0.0,
+      "learning_rate": 2.7343015743307443e-08,
+      "loss": 0.6762,
+      "step": 3167
+    },
+    {
+      "epoch": 0.977325312355391,
+      "grad_norm": 0.0,
+      "learning_rate": 2.660933426373413e-08,
+      "loss": 0.7482,
+      "step": 3168
+    },
+    {
+      "epoch": 0.9776338115070183,
+      "grad_norm": 0.0,
+      "learning_rate": 2.5885617244392204e-08,
+      "loss": 0.7199,
+      "step": 3169
+    },
+    {
+      "epoch": 0.9779423106586457,
+      "grad_norm": 0.0,
+      "learning_rate": 2.5171865408352147e-08,
+      "loss": 0.7257,
+      "step": 3170
+    },
+    {
+      "epoch": 0.978250809810273,
+      "grad_norm": 0.0,
+      "learning_rate": 2.4468079468724647e-08,
+      "loss": 0.6469,
+      "step": 3171
+    },
+    {
+      "epoch": 0.9785593089619004,
+      "grad_norm": 0.0,
+      "learning_rate": 2.3774260128667238e-08,
+      "loss": 0.8081,
+      "step": 3172
+    },
+    {
+      "epoch": 0.9788678081135277,
+      "grad_norm": 0.0,
+      "learning_rate": 2.3090408081376525e-08,
+      "loss": 0.6983,
+      "step": 3173
+    },
+    {
+      "epoch": 0.979176307265155,
+      "grad_norm": 0.0,
+      "learning_rate": 2.2416524010092644e-08,
+      "loss": 0.6102,
+      "step": 3174
+    },
+    {
+      "epoch": 0.9794848064167824,
+      "grad_norm": 0.0,
+      "learning_rate": 2.1752608588097024e-08,
+      "loss": 0.6399,
+      "step": 3175
+    },
+    {
+      "epoch": 0.9797933055684097,
+      "grad_norm": 0.0,
+      "learning_rate": 2.1098662478710173e-08,
+      "loss": 0.6928,
+      "step": 3176
+    },
+    {
+      "epoch": 0.980101804720037,
+      "grad_norm": 0.0,
+      "learning_rate": 2.0454686335292794e-08,
+      "loss": 0.8059,
+      "step": 3177
+    },
+    {
+      "epoch": 0.9804103038716644,
+      "grad_norm": 0.0,
+      "learning_rate": 1.9820680801243554e-08,
+      "loss": 0.6157,
+      "step": 3178
+    },
+    {
+      "epoch": 0.9807188030232917,
+      "grad_norm": 0.0,
+      "learning_rate": 1.9196646510001303e-08,
+      "loss": 0.6005,
+      "step": 3179
+    },
+    {
+      "epoch": 0.9810273021749191,
+      "grad_norm": 0.0,
+      "learning_rate": 1.8582584085041765e-08,
+      "loss": 0.7006,
+      "step": 3180
+    },
+    {
+      "epoch": 0.9813358013265463,
+      "grad_norm": 0.0,
+      "learning_rate": 1.797849413987529e-08,
+      "loss": 0.7217,
+      "step": 3181
+    },
+    {
+      "epoch": 0.9816443004781736,
+      "grad_norm": 0.0,
+      "learning_rate": 1.7384377278053533e-08,
+      "loss": 0.7453,
+      "step": 3182
+    },
+    {
+      "epoch": 0.981952799629801,
+      "grad_norm": 0.0,
+      "learning_rate": 1.680023409316056e-08,
+      "loss": 0.7559,
+      "step": 3183
+    },
+    {
+      "epoch": 0.9822612987814283,
+      "grad_norm": 0.0,
+      "learning_rate": 1.6226065168816198e-08,
+      "loss": 0.6353,
+      "step": 3184
+    },
+    {
+      "epoch": 0.9825697979330557,
+      "grad_norm": 0.0,
+      "learning_rate": 1.5661871078674895e-08,
+      "loss": 0.7237,
+      "step": 3185
+    },
+    {
+      "epoch": 0.982878297084683,
+      "grad_norm": 0.0,
+      "learning_rate": 1.510765238642575e-08,
+      "loss": 0.7412,
+      "step": 3186
+    },
+    {
+      "epoch": 0.9831867962363103,
+      "grad_norm": 0.0,
+      "learning_rate": 1.4563409645792503e-08,
+      "loss": 0.6923,
+      "step": 3187
+    },
+    {
+      "epoch": 0.9834952953879377,
+      "grad_norm": 0.0,
+      "learning_rate": 1.4029143400529077e-08,
+      "loss": 0.662,
+      "step": 3188
+    },
+    {
+      "epoch": 0.983803794539565,
+      "grad_norm": 0.0,
+      "learning_rate": 1.3504854184422932e-08,
+      "loss": 0.5827,
+      "step": 3189
+    },
+    {
+      "epoch": 0.9841122936911924,
+      "grad_norm": 0.0,
+      "learning_rate": 1.2990542521295057e-08,
+      "loss": 0.6758,
+      "step": 3190
+    },
+    {
+      "epoch": 0.9844207928428197,
+      "grad_norm": 0.0,
+      "learning_rate": 1.2486208924996634e-08,
+      "loss": 0.7292,
+      "step": 3191
+    },
+    {
+      "epoch": 0.984729291994447,
+      "grad_norm": 0.0,
+      "learning_rate": 1.1991853899409044e-08,
+      "loss": 0.6786,
+      "step": 3192
+    },
+    {
+      "epoch": 0.9850377911460744,
+      "grad_norm": 0.0,
+      "learning_rate": 1.1507477938443868e-08,
+      "loss": 0.6944,
+      "step": 3193
+    },
+    {
+      "epoch": 0.9853462902977017,
+      "grad_norm": 0.0,
+      "learning_rate": 1.1033081526045098e-08,
+      "loss": 0.6941,
+      "step": 3194
+    },
+    {
+      "epoch": 0.985654789449329,
+      "grad_norm": 0.0,
+      "learning_rate": 1.0568665136183597e-08,
+      "loss": 0.6545,
+      "step": 3195
+    },
+    {
+      "epoch": 0.9859632886009564,
+      "grad_norm": 0.0,
+      "learning_rate": 1.0114229232859319e-08,
+      "loss": 0.7082,
+      "step": 3196
+    },
+    {
+      "epoch": 0.9862717877525837,
+      "grad_norm": 0.0,
+      "learning_rate": 9.669774270102406e-09,
+      "loss": 0.6543,
+      "step": 3197
+    },
+    {
+      "epoch": 0.9865802869042111,
+      "grad_norm": 0.0,
+      "learning_rate": 9.235300691969872e-09,
+      "loss": 0.6542,
+      "step": 3198
+    },
+    {
+      "epoch": 0.9868887860558383,
+      "grad_norm": 0.0,
+      "learning_rate": 8.81080893254449e-09,
+      "loss": 0.6826,
+      "step": 3199
+    },
+    {
+      "epoch": 0.9871972852074656,
+      "grad_norm": 0.0,
+      "learning_rate": 8.396299415939223e-09,
+      "loss": 0.6522,
+      "step": 3200
+    },
+    {
+      "epoch": 0.987505784359093,
+      "grad_norm": 0.0,
+      "learning_rate": 7.991772556291689e-09,
+      "loss": 0.6831,
+      "step": 3201
+    },
+    {
+      "epoch": 0.9878142835107203,
+      "grad_norm": 0.0,
+      "learning_rate": 7.597228757767472e-09,
+      "loss": 0.6952,
+      "step": 3202
+    },
+    {
+      "epoch": 0.9881227826623477,
+      "grad_norm": 0.0,
+      "learning_rate": 7.212668414556811e-09,
+      "loss": 0.7034,
+      "step": 3203
+    },
+    {
+      "epoch": 0.988431281813975,
+      "grad_norm": 0.0,
+      "learning_rate": 6.8380919108757e-09,
+      "loss": 0.7407,
+      "step": 3204
+    },
+    {
+      "epoch": 0.9887397809656023,
+      "grad_norm": 0.0,
+      "learning_rate": 6.473499620965884e-09,
+      "loss": 0.7395,
+      "step": 3205
+    },
+    {
+      "epoch": 0.9890482801172297,
+      "grad_norm": 0.0,
+      "learning_rate": 6.11889190909043e-09,
+      "loss": 0.7463,
+      "step": 3206
+    },
+    {
+      "epoch": 0.989356779268857,
+      "grad_norm": 0.0,
+      "learning_rate": 5.774269129543708e-09,
+      "loss": 0.7032,
+      "step": 3207
+    },
+    {
+      "epoch": 0.9896652784204844,
+      "grad_norm": 0.0,
+      "learning_rate": 5.439631626635855e-09,
+      "loss": 0.6845,
+      "step": 3208
+    },
+    {
+      "epoch": 0.9899737775721117,
+      "grad_norm": 0.0,
+      "learning_rate": 5.114979734707204e-09,
+      "loss": 0.6526,
+      "step": 3209
+    },
+    {
+      "epoch": 0.990282276723739,
+      "grad_norm": 0.0,
+      "learning_rate": 4.8003137781194035e-09,
+      "loss": 0.6862,
+      "step": 3210
+    },
+    {
+      "epoch": 0.9905907758753664,
+      "grad_norm": 0.0,
+      "learning_rate": 4.495634071254307e-09,
+      "loss": 0.6675,
+      "step": 3211
+    },
+    {
+      "epoch": 0.9908992750269937,
+      "grad_norm": 0.0,
+      "learning_rate": 4.200940918520635e-09,
+      "loss": 0.6661,
+      "step": 3212
+    },
+    {
+      "epoch": 0.991207774178621,
+      "grad_norm": 0.0,
+      "learning_rate": 3.916234614346204e-09,
+      "loss": 0.685,
+      "step": 3213
+    },
+    {
+      "epoch": 0.9915162733302484,
+      "grad_norm": 0.0,
+      "learning_rate": 3.6415154431845846e-09,
+      "loss": 0.6182,
+      "step": 3214
+    },
+    {
+      "epoch": 0.9918247724818757,
+      "grad_norm": 0.0,
+      "learning_rate": 3.3767836795062235e-09,
+      "loss": 0.7722,
+      "step": 3215
+    },
+    {
+      "epoch": 0.992133271633503,
+      "grad_norm": 0.0,
+      "learning_rate": 3.1220395878084343e-09,
+      "loss": 0.6592,
+      "step": 3216
+    },
+    {
+      "epoch": 0.9924417707851303,
+      "grad_norm": 0.0,
+      "learning_rate": 2.8772834226054036e-09,
+      "loss": 0.6126,
+      "step": 3217
+    },
+    {
+      "epoch": 0.9927502699367576,
+      "grad_norm": 0.0,
+      "learning_rate": 2.6425154284359655e-09,
+      "loss": 0.7914,
+      "step": 3218
+    },
+    {
+      "epoch": 0.993058769088385,
+      "grad_norm": 0.0,
+      "learning_rate": 2.4177358398558283e-09,
+      "loss": 0.6343,
+      "step": 3219
+    },
+    {
+      "epoch": 0.9933672682400123,
+      "grad_norm": 0.0,
+      "learning_rate": 2.2029448814431254e-09,
+      "loss": 0.6759,
+      "step": 3220
+    },
+    {
+      "epoch": 0.9936757673916397,
+      "grad_norm": 0.0,
+      "learning_rate": 1.9981427677995267e-09,
+      "loss": 0.7549,
+      "step": 3221
+    },
+    {
+      "epoch": 0.993984266543267,
+      "grad_norm": 0.0,
+      "learning_rate": 1.8033297035402463e-09,
+      "loss": 0.7574,
+      "step": 3222
+    },
+    {
+      "epoch": 0.9942927656948943,
+      "grad_norm": 0.0,
+      "learning_rate": 1.6185058833062538e-09,
+      "loss": 0.6929,
+      "step": 3223
+    },
+    {
+      "epoch": 0.9946012648465217,
+      "grad_norm": 0.0,
+      "learning_rate": 1.4436714917542839e-09,
+      "loss": 0.6079,
+      "step": 3224
+    },
+    {
+      "epoch": 0.994909763998149,
+      "grad_norm": 0.0,
+      "learning_rate": 1.278826703563496e-09,
+      "loss": 0.6822,
+      "step": 3225
+    },
+    {
+      "epoch": 0.9952182631497763,
+      "grad_norm": 0.0,
+      "learning_rate": 1.1239716834310354e-09,
+      "loss": 0.9537,
+      "step": 3226
+    },
+    {
+      "epoch": 0.9955267623014037,
+      "grad_norm": 0.0,
+      "learning_rate": 9.791065860720316e-10,
+      "loss": 0.6772,
+      "step": 3227
+    },
+    {
+      "epoch": 0.995835261453031,
+      "grad_norm": 0.0,
+      "learning_rate": 8.442315562229298e-10,
+      "loss": 0.686,
+      "step": 3228
+    },
+    {
+      "epoch": 0.9961437606046584,
+      "grad_norm": 0.0,
+      "learning_rate": 7.193467286370492e-10,
+      "loss": 0.68,
+      "step": 3229
+    },
+    {
+      "epoch": 0.9964522597562857,
+      "grad_norm": 0.0,
+      "learning_rate": 6.044522280890252e-10,
+      "loss": 0.6778,
+      "step": 3230
+    },
+    {
+      "epoch": 0.996760758907913,
+      "grad_norm": 0.0,
+      "learning_rate": 4.995481693681469e-10,
+      "loss": 0.6982,
+      "step": 3231
+    },
+    {
+      "epoch": 0.9970692580595404,
+      "grad_norm": 0.0,
+      "learning_rate": 4.0463465728723947e-10,
+      "loss": 0.6542,
+      "step": 3232
+    },
+    {
+      "epoch": 0.9973777572111676,
+      "grad_norm": 0.0,
+      "learning_rate": 3.1971178667267176e-10,
+      "loss": 0.6921,
+      "step": 3233
+    },
+    {
+      "epoch": 0.997686256362795,
+      "grad_norm": 0.0,
+      "learning_rate": 2.4477964237212823e-10,
+      "loss": 0.71,
+      "step": 3234
+    },
+    {
+      "epoch": 0.9979947555144223,
+      "grad_norm": 0.0,
+      "learning_rate": 1.7983829925016793e-10,
+      "loss": 0.7023,
+      "step": 3235
+    },
+    {
+      "epoch": 0.9983032546660496,
+      "grad_norm": 0.0,
+      "learning_rate": 1.248878221904448e-10,
+      "loss": 0.6873,
+      "step": 3236
+    },
+    {
+      "epoch": 0.998611753817677,
+      "grad_norm": 0.0,
+      "learning_rate": 7.992826609459769e-11,
+      "loss": 0.6443,
+      "step": 3237
+    },
+    {
+      "epoch": 0.9989202529693043,
+      "grad_norm": 0.0,
+      "learning_rate": 4.4959675881139916e-11,
+      "loss": 0.6832,
+      "step": 3238
+    },
+    {
+      "epoch": 0.9992287521209317,
+      "grad_norm": 0.0,
+      "learning_rate": 1.9982086487679853e-11,
+      "loss": 0.6543,
+      "step": 3239
+    },
+    {
+      "epoch": 0.999537251272559,
+      "grad_norm": 0.0,
+      "learning_rate": 4.995522869810643e-12,
+      "loss": 0.7005,
+      "step": 3240
+    },
+    {
+      "epoch": 0.9998457504241863,
+      "grad_norm": 0.0,
+      "learning_rate": 0.0,
+      "loss": 0.8583,
+      "step": 3241
+    },
+    {
+      "epoch": 0.9998457504241863,
+      "step": 3241,
+      "total_flos": 1.4116677844331397e+19,
+      "train_loss": 0.5099010296678588,
+      "train_runtime": 75843.298,
+      "train_samples_per_second": 5.47,
+      "train_steps_per_second": 0.043
+    }
+  ],
+  "logging_steps": 1.0,
+  "max_steps": 3241,
+  "num_input_tokens_seen": 0,
+  "num_train_epochs": 1,
+  "save_steps": 1000,
+  "stateful_callbacks": {
+    "TrainerControl": {
+      "args": {
+        "should_epoch_stop": false,
+        "should_evaluate": false,
+        "should_log": false,
+        "should_save": true,
+        "should_training_stop": false
+      },
+      "attributes": {}
+    }
+  },
+  "total_flos": 1.4116677844331397e+19,
+  "train_batch_size": 1,
+  "trial_name": null,
+  "trial_params": null
+}