diff --git "a/checkpoint-924/trainer_state.json" "b/checkpoint-924/trainer_state.json"
new file mode 100644--- /dev/null
+++ "b/checkpoint-924/trainer_state.json"
@@ -0,0 +1,6489 @@
+{
+  "best_metric": null,
+  "best_model_checkpoint": null,
+  "epoch": 0.6007802340702211,
+  "eval_steps": 500,
+  "global_step": 924,
+  "is_hyper_param_search": false,
+  "is_local_process_zero": true,
+  "is_world_process_zero": true,
+  "log_history": [
+    {
+      "epoch": 0.0,
+      "grad_norm": 0.0303955078125,
+      "learning_rate": 4.2553191489361704e-07,
+      "loss": 1.11,
+      "step": 1
+    },
+    {
+      "epoch": 0.0,
+      "grad_norm": 0.0294189453125,
+      "learning_rate": 8.510638297872341e-07,
+      "loss": 0.9825,
+      "step": 2
+    },
+    {
+      "epoch": 0.0,
+      "grad_norm": 0.028076171875,
+      "learning_rate": 1.276595744680851e-06,
+      "loss": 1.0375,
+      "step": 3
+    },
+    {
+      "epoch": 0.0,
+      "grad_norm": 0.036865234375,
+      "learning_rate": 1.7021276595744682e-06,
+      "loss": 1.042,
+      "step": 4
+    },
+    {
+      "epoch": 0.0,
+      "grad_norm": 0.0311279296875,
+      "learning_rate": 2.1276595744680853e-06,
+      "loss": 0.9769,
+      "step": 5
+    },
+    {
+      "epoch": 0.0,
+      "grad_norm": 0.036865234375,
+      "learning_rate": 2.553191489361702e-06,
+      "loss": 0.9316,
+      "step": 6
+    },
+    {
+      "epoch": 0.0,
+      "grad_norm": 0.0299072265625,
+      "learning_rate": 2.978723404255319e-06,
+      "loss": 1.0077,
+      "step": 7
+    },
+    {
+      "epoch": 0.01,
+      "grad_norm": 0.03369140625,
+      "learning_rate": 3.4042553191489363e-06,
+      "loss": 1.0346,
+      "step": 8
+    },
+    {
+      "epoch": 0.01,
+      "grad_norm": 0.0301513671875,
+      "learning_rate": 3.8297872340425535e-06,
+      "loss": 1.0193,
+      "step": 9
+    },
+    {
+      "epoch": 0.01,
+      "grad_norm": 0.0267333984375,
+      "learning_rate": 4.255319148936171e-06,
+      "loss": 1.0297,
+      "step": 10
+    },
+    {
+      "epoch": 0.01,
+      "grad_norm": 0.0291748046875,
+      "learning_rate": 4.680851063829788e-06,
+      "loss": 1.0868,
+      "step": 11
+    },
+    {
+      "epoch": 0.01,
+      "grad_norm": 0.0296630859375,
+      "learning_rate": 5.106382978723404e-06,
+      "loss": 1.0641,
+      "step": 12
+    },
+    {
+      "epoch": 0.01,
+      "grad_norm": 0.028564453125,
+      "learning_rate": 5.531914893617022e-06,
+      "loss": 0.9389,
+      "step": 13
+    },
+    {
+      "epoch": 0.01,
+      "grad_norm": 0.03515625,
+      "learning_rate": 5.957446808510638e-06,
+      "loss": 1.033,
+      "step": 14
+    },
+    {
+      "epoch": 0.01,
+      "grad_norm": 0.0283203125,
+      "learning_rate": 6.382978723404256e-06,
+      "loss": 1.0216,
+      "step": 15
+    },
+    {
+      "epoch": 0.01,
+      "grad_norm": 0.0322265625,
+      "learning_rate": 6.808510638297873e-06,
+      "loss": 1.1086,
+      "step": 16
+    },
+    {
+      "epoch": 0.01,
+      "grad_norm": 0.0322265625,
+      "learning_rate": 7.234042553191491e-06,
+      "loss": 1.0577,
+      "step": 17
+    },
+    {
+      "epoch": 0.01,
+      "grad_norm": 0.030517578125,
+      "learning_rate": 7.659574468085107e-06,
+      "loss": 1.0733,
+      "step": 18
+    },
+    {
+      "epoch": 0.01,
+      "grad_norm": 0.0303955078125,
+      "learning_rate": 8.085106382978723e-06,
+      "loss": 0.9865,
+      "step": 19
+    },
+    {
+      "epoch": 0.01,
+      "grad_norm": 0.0291748046875,
+      "learning_rate": 8.510638297872341e-06,
+      "loss": 1.0125,
+      "step": 20
+    },
+    {
+      "epoch": 0.01,
+      "grad_norm": 0.034423828125,
+      "learning_rate": 8.936170212765958e-06,
+      "loss": 1.1245,
+      "step": 21
+    },
+    {
+      "epoch": 0.01,
+      "grad_norm": 0.0260009765625,
+      "learning_rate": 9.361702127659576e-06,
+      "loss": 1.0024,
+      "step": 22
+    },
+    {
+      "epoch": 0.01,
+      "grad_norm": 0.0306396484375,
+      "learning_rate": 9.787234042553192e-06,
+      "loss": 1.0131,
+      "step": 23
+    },
+    {
+      "epoch": 0.02,
+      "grad_norm": 0.033447265625,
+      "learning_rate": 1.0212765957446808e-05,
+      "loss": 1.0171,
+      "step": 24
+    },
+    {
+      "epoch": 0.02,
+      "grad_norm": 0.033203125,
+      "learning_rate": 1.0638297872340426e-05,
+      "loss": 0.9613,
+      "step": 25
+    },
+    {
+      "epoch": 0.02,
+      "grad_norm": 0.03857421875,
+      "learning_rate": 1.1063829787234044e-05,
+      "loss": 1.1312,
+      "step": 26
+    },
+    {
+      "epoch": 0.02,
+      "grad_norm": 0.036865234375,
+      "learning_rate": 1.1489361702127662e-05,
+      "loss": 1.0187,
+      "step": 27
+    },
+    {
+      "epoch": 0.02,
+      "grad_norm": 0.03515625,
+      "learning_rate": 1.1914893617021277e-05,
+      "loss": 0.9934,
+      "step": 28
+    },
+    {
+      "epoch": 0.02,
+      "grad_norm": 0.036376953125,
+      "learning_rate": 1.2340425531914895e-05,
+      "loss": 1.0872,
+      "step": 29
+    },
+    {
+      "epoch": 0.02,
+      "grad_norm": 0.035888671875,
+      "learning_rate": 1.2765957446808513e-05,
+      "loss": 0.9591,
+      "step": 30
+    },
+    {
+      "epoch": 0.02,
+      "grad_norm": 0.033203125,
+      "learning_rate": 1.3191489361702127e-05,
+      "loss": 0.9589,
+      "step": 31
+    },
+    {
+      "epoch": 0.02,
+      "grad_norm": 0.040771484375,
+      "learning_rate": 1.3617021276595745e-05,
+      "loss": 1.0093,
+      "step": 32
+    },
+    {
+      "epoch": 0.02,
+      "grad_norm": 0.0400390625,
+      "learning_rate": 1.4042553191489363e-05,
+      "loss": 1.0195,
+      "step": 33
+    },
+    {
+      "epoch": 0.02,
+      "grad_norm": 0.038330078125,
+      "learning_rate": 1.4468085106382981e-05,
+      "loss": 0.8936,
+      "step": 34
+    },
+    {
+      "epoch": 0.02,
+      "grad_norm": 0.04443359375,
+      "learning_rate": 1.4893617021276596e-05,
+      "loss": 0.9958,
+      "step": 35
+    },
+    {
+      "epoch": 0.02,
+      "grad_norm": 0.045654296875,
+      "learning_rate": 1.5319148936170214e-05,
+      "loss": 0.9279,
+      "step": 36
+    },
+    {
+      "epoch": 0.02,
+      "grad_norm": 0.03662109375,
+      "learning_rate": 1.5744680851063832e-05,
+      "loss": 1.0153,
+      "step": 37
+    },
+    {
+      "epoch": 0.02,
+      "grad_norm": 0.04638671875,
+      "learning_rate": 1.6170212765957446e-05,
+      "loss": 0.9862,
+      "step": 38
+    },
+    {
+      "epoch": 0.03,
+      "grad_norm": 0.042236328125,
+      "learning_rate": 1.6595744680851064e-05,
+      "loss": 1.0962,
+      "step": 39
+    },
+    {
+      "epoch": 0.03,
+      "grad_norm": 0.040771484375,
+      "learning_rate": 1.7021276595744682e-05,
+      "loss": 0.956,
+      "step": 40
+    },
+    {
+      "epoch": 0.03,
+      "grad_norm": 0.035888671875,
+      "learning_rate": 1.74468085106383e-05,
+      "loss": 1.0559,
+      "step": 41
+    },
+    {
+      "epoch": 0.03,
+      "grad_norm": 0.04736328125,
+      "learning_rate": 1.7872340425531915e-05,
+      "loss": 1.0014,
+      "step": 42
+    },
+    {
+      "epoch": 0.03,
+      "grad_norm": 0.051025390625,
+      "learning_rate": 1.8297872340425533e-05,
+      "loss": 1.0252,
+      "step": 43
+    },
+    {
+      "epoch": 0.03,
+      "grad_norm": 0.04931640625,
+      "learning_rate": 1.872340425531915e-05,
+      "loss": 0.9541,
+      "step": 44
+    },
+    {
+      "epoch": 0.03,
+      "grad_norm": 0.0556640625,
+      "learning_rate": 1.914893617021277e-05,
+      "loss": 0.9603,
+      "step": 45
+    },
+    {
+      "epoch": 0.03,
+      "grad_norm": 0.0419921875,
+      "learning_rate": 1.9574468085106384e-05,
+      "loss": 1.0601,
+      "step": 46
+    },
+    {
+      "epoch": 0.03,
+      "grad_norm": 0.0478515625,
+      "learning_rate": 2e-05,
+      "loss": 0.9919,
+      "step": 47
+    },
+    {
+      "epoch": 0.03,
+      "grad_norm": 0.047119140625,
+      "learning_rate": 1.9999977801976743e-05,
+      "loss": 1.0247,
+      "step": 48
+    },
+    {
+      "epoch": 0.03,
+      "grad_norm": 0.048095703125,
+      "learning_rate": 1.999991120800551e-05,
+      "loss": 0.9936,
+      "step": 49
+    },
+    {
+      "epoch": 0.03,
+      "grad_norm": 0.05419921875,
+      "learning_rate": 1.9999800218381958e-05,
+      "loss": 1.0315,
+      "step": 50
+    },
+    {
+      "epoch": 0.03,
+      "grad_norm": 0.0478515625,
+      "learning_rate": 1.9999644833598836e-05,
+      "loss": 0.9392,
+      "step": 51
+    },
+    {
+      "epoch": 0.03,
+      "grad_norm": 0.0546875,
+      "learning_rate": 1.9999445054345993e-05,
+      "loss": 1.0716,
+      "step": 52
+    },
+    {
+      "epoch": 0.03,
+      "grad_norm": 0.05224609375,
+      "learning_rate": 1.9999200881510366e-05,
+      "loss": 0.9724,
+      "step": 53
+    },
+    {
+      "epoch": 0.04,
+      "grad_norm": 0.04736328125,
+      "learning_rate": 1.999891231617599e-05,
+      "loss": 0.9966,
+      "step": 54
+    },
+    {
+      "epoch": 0.04,
+      "grad_norm": 0.049072265625,
+      "learning_rate": 1.9998579359623977e-05,
+      "loss": 0.969,
+      "step": 55
+    },
+    {
+      "epoch": 0.04,
+      "grad_norm": 0.051513671875,
+      "learning_rate": 1.9998202013332525e-05,
+      "loss": 0.972,
+      "step": 56
+    },
+    {
+      "epoch": 0.04,
+      "grad_norm": 0.043701171875,
+      "learning_rate": 1.99977802789769e-05,
+      "loss": 0.9705,
+      "step": 57
+    },
+    {
+      "epoch": 0.04,
+      "grad_norm": 0.044189453125,
+      "learning_rate": 1.999731415842944e-05,
+      "loss": 1.002,
+      "step": 58
+    },
+    {
+      "epoch": 0.04,
+      "grad_norm": 0.039794921875,
+      "learning_rate": 1.9996803653759534e-05,
+      "loss": 0.9508,
+      "step": 59
+    },
+    {
+      "epoch": 0.04,
+      "grad_norm": 0.03759765625,
+      "learning_rate": 1.9996248767233616e-05,
+      "loss": 0.9232,
+      "step": 60
+    },
+    {
+      "epoch": 0.04,
+      "grad_norm": 0.0390625,
+      "learning_rate": 1.9995649501315172e-05,
+      "loss": 1.0054,
+      "step": 61
+    },
+    {
+      "epoch": 0.04,
+      "grad_norm": 0.034423828125,
+      "learning_rate": 1.9995005858664696e-05,
+      "loss": 0.9685,
+      "step": 62
+    },
+    {
+      "epoch": 0.04,
+      "grad_norm": 0.03369140625,
+      "learning_rate": 1.9994317842139715e-05,
+      "loss": 0.9313,
+      "step": 63
+    },
+    {
+      "epoch": 0.04,
+      "grad_norm": 0.0311279296875,
+      "learning_rate": 1.9993585454794748e-05,
+      "loss": 0.9463,
+      "step": 64
+    },
+    {
+      "epoch": 0.04,
+      "grad_norm": 0.0311279296875,
+      "learning_rate": 1.9992808699881303e-05,
+      "loss": 0.9049,
+      "step": 65
+    },
+    {
+      "epoch": 0.04,
+      "grad_norm": 0.0322265625,
+      "learning_rate": 1.999198758084787e-05,
+      "loss": 0.9088,
+      "step": 66
+    },
+    {
+      "epoch": 0.04,
+      "grad_norm": 0.033203125,
+      "learning_rate": 1.9991122101339885e-05,
+      "loss": 0.9369,
+      "step": 67
+    },
+    {
+      "epoch": 0.04,
+      "grad_norm": 0.0478515625,
+      "learning_rate": 1.9990212265199738e-05,
+      "loss": 0.9902,
+      "step": 68
+    },
+    {
+      "epoch": 0.04,
+      "grad_norm": 0.03466796875,
+      "learning_rate": 1.9989258076466743e-05,
+      "loss": 0.9569,
+      "step": 69
+    },
+    {
+      "epoch": 0.05,
+      "grad_norm": 0.042724609375,
+      "learning_rate": 1.998825953937712e-05,
+      "loss": 0.9779,
+      "step": 70
+    },
+    {
+      "epoch": 0.05,
+      "grad_norm": 0.0380859375,
+      "learning_rate": 1.9987216658363983e-05,
+      "loss": 0.9505,
+      "step": 71
+    },
+    {
+      "epoch": 0.05,
+      "grad_norm": 0.036376953125,
+      "learning_rate": 1.9986129438057306e-05,
+      "loss": 0.9374,
+      "step": 72
+    },
+    {
+      "epoch": 0.05,
+      "grad_norm": 0.0361328125,
+      "learning_rate": 1.998499788328392e-05,
+      "loss": 1.0086,
+      "step": 73
+    },
+    {
+      "epoch": 0.05,
+      "grad_norm": 0.034912109375,
+      "learning_rate": 1.9983821999067478e-05,
+      "loss": 1.046,
+      "step": 74
+    },
+    {
+      "epoch": 0.05,
+      "grad_norm": 0.031005859375,
+      "learning_rate": 1.998260179062844e-05,
+      "loss": 0.9375,
+      "step": 75
+    },
+    {
+      "epoch": 0.05,
+      "grad_norm": 0.032958984375,
+      "learning_rate": 1.9981337263384057e-05,
+      "loss": 0.9514,
+      "step": 76
+    },
+    {
+      "epoch": 0.05,
+      "grad_norm": 0.031982421875,
+      "learning_rate": 1.9980028422948323e-05,
+      "loss": 0.8629,
+      "step": 77
+    },
+    {
+      "epoch": 0.05,
+      "grad_norm": 0.03125,
+      "learning_rate": 1.9978675275131975e-05,
+      "loss": 0.933,
+      "step": 78
+    },
+    {
+      "epoch": 0.05,
+      "grad_norm": 0.0299072265625,
+      "learning_rate": 1.9977277825942453e-05,
+      "loss": 0.9408,
+      "step": 79
+    },
+    {
+      "epoch": 0.05,
+      "grad_norm": 0.031494140625,
+      "learning_rate": 1.997583608158388e-05,
+      "loss": 1.0041,
+      "step": 80
+    },
+    {
+      "epoch": 0.05,
+      "grad_norm": 0.031982421875,
+      "learning_rate": 1.997435004845703e-05,
+      "loss": 0.9605,
+      "step": 81
+    },
+    {
+      "epoch": 0.05,
+      "grad_norm": 0.03271484375,
+      "learning_rate": 1.99728197331593e-05,
+      "loss": 0.9256,
+      "step": 82
+    },
+    {
+      "epoch": 0.05,
+      "grad_norm": 0.034912109375,
+      "learning_rate": 1.9971245142484693e-05,
+      "loss": 1.0026,
+      "step": 83
+    },
+    {
+      "epoch": 0.05,
+      "grad_norm": 0.031494140625,
+      "learning_rate": 1.996962628342376e-05,
+      "loss": 0.9789,
+      "step": 84
+    },
+    {
+      "epoch": 0.06,
+      "grad_norm": 0.0341796875,
+      "learning_rate": 1.99679631631636e-05,
+      "loss": 0.9437,
+      "step": 85
+    },
+    {
+      "epoch": 0.06,
+      "grad_norm": 0.030517578125,
+      "learning_rate": 1.996625578908781e-05,
+      "loss": 0.9487,
+      "step": 86
+    },
+    {
+      "epoch": 0.06,
+      "grad_norm": 0.033447265625,
+      "learning_rate": 1.9964504168776454e-05,
+      "loss": 0.9645,
+      "step": 87
+    },
+    {
+      "epoch": 0.06,
+      "grad_norm": 0.03271484375,
+      "learning_rate": 1.9962708310006032e-05,
+      "loss": 0.9967,
+      "step": 88
+    },
+    {
+      "epoch": 0.06,
+      "grad_norm": 0.0296630859375,
+      "learning_rate": 1.996086822074945e-05,
+      "loss": 1.0195,
+      "step": 89
+    },
+    {
+      "epoch": 0.06,
+      "grad_norm": 0.030517578125,
+      "learning_rate": 1.9958983909175977e-05,
+      "loss": 0.8769,
+      "step": 90
+    },
+    {
+      "epoch": 0.06,
+      "grad_norm": 0.031494140625,
+      "learning_rate": 1.995705538365121e-05,
+      "loss": 0.8407,
+      "step": 91
+    },
+    {
+      "epoch": 0.06,
+      "grad_norm": 0.033203125,
+      "learning_rate": 1.995508265273704e-05,
+      "loss": 0.9368,
+      "step": 92
+    },
+    {
+      "epoch": 0.06,
+      "grad_norm": 0.031982421875,
+      "learning_rate": 1.9953065725191613e-05,
+      "loss": 0.9308,
+      "step": 93
+    },
+    {
+      "epoch": 0.06,
+      "grad_norm": 0.03076171875,
+      "learning_rate": 1.9951004609969286e-05,
+      "loss": 0.9235,
+      "step": 94
+    },
+    {
+      "epoch": 0.06,
+      "grad_norm": 0.032958984375,
+      "learning_rate": 1.9948899316220603e-05,
+      "loss": 0.9008,
+      "step": 95
+    },
+    {
+      "epoch": 0.06,
+      "grad_norm": 0.03173828125,
+      "learning_rate": 1.9946749853292233e-05,
+      "loss": 0.9735,
+      "step": 96
+    },
+    {
+      "epoch": 0.06,
+      "grad_norm": 0.033447265625,
+      "learning_rate": 1.994455623072694e-05,
+      "loss": 0.9328,
+      "step": 97
+    },
+    {
+      "epoch": 0.06,
+      "grad_norm": 0.033203125,
+      "learning_rate": 1.994231845826354e-05,
+      "loss": 0.8967,
+      "step": 98
+    },
+    {
+      "epoch": 0.06,
+      "grad_norm": 0.03173828125,
+      "learning_rate": 1.994003654583686e-05,
+      "loss": 0.8363,
+      "step": 99
+    },
+    {
+      "epoch": 0.07,
+      "grad_norm": 0.033447265625,
+      "learning_rate": 1.993771050357769e-05,
+      "loss": 0.9072,
+      "step": 100
+    },
+    {
+      "epoch": 0.07,
+      "grad_norm": 0.03369140625,
+      "learning_rate": 1.9935340341812737e-05,
+      "loss": 0.9502,
+      "step": 101
+    },
+    {
+      "epoch": 0.07,
+      "grad_norm": 0.03271484375,
+      "learning_rate": 1.993292607106458e-05,
+      "loss": 0.8794,
+      "step": 102
+    },
+    {
+      "epoch": 0.07,
+      "grad_norm": 0.03466796875,
+      "learning_rate": 1.9930467702051632e-05,
+      "loss": 0.9601,
+      "step": 103
+    },
+    {
+      "epoch": 0.07,
+      "grad_norm": 0.0341796875,
+      "learning_rate": 1.9927965245688073e-05,
+      "loss": 0.9099,
+      "step": 104
+    },
+    {
+      "epoch": 0.07,
+      "grad_norm": 0.033935546875,
+      "learning_rate": 1.9925418713083824e-05,
+      "loss": 0.929,
+      "step": 105
+    },
+    {
+      "epoch": 0.07,
+      "grad_norm": 0.033447265625,
+      "learning_rate": 1.992282811554448e-05,
+      "loss": 0.9046,
+      "step": 106
+    },
+    {
+      "epoch": 0.07,
+      "grad_norm": 0.031005859375,
+      "learning_rate": 1.9920193464571277e-05,
+      "loss": 0.9393,
+      "step": 107
+    },
+    {
+      "epoch": 0.07,
+      "grad_norm": 0.03515625,
+      "learning_rate": 1.9917514771861015e-05,
+      "loss": 0.9933,
+      "step": 108
+    },
+    {
+      "epoch": 0.07,
+      "grad_norm": 0.035400390625,
+      "learning_rate": 1.9914792049306034e-05,
+      "loss": 0.8865,
+      "step": 109
+    },
+    {
+      "epoch": 0.07,
+      "grad_norm": 0.032958984375,
+      "learning_rate": 1.9912025308994146e-05,
+      "loss": 0.9158,
+      "step": 110
+    },
+    {
+      "epoch": 0.07,
+      "grad_norm": 0.035888671875,
+      "learning_rate": 1.990921456320859e-05,
+      "loss": 0.9143,
+      "step": 111
+    },
+    {
+      "epoch": 0.07,
+      "grad_norm": 0.03369140625,
+      "learning_rate": 1.9906359824427953e-05,
+      "loss": 0.9707,
+      "step": 112
+    },
+    {
+      "epoch": 0.07,
+      "grad_norm": 0.0341796875,
+      "learning_rate": 1.9903461105326155e-05,
+      "loss": 0.8894,
+      "step": 113
+    },
+    {
+      "epoch": 0.07,
+      "grad_norm": 0.0341796875,
+      "learning_rate": 1.9900518418772364e-05,
+      "loss": 0.966,
+      "step": 114
+    },
+    {
+      "epoch": 0.07,
+      "grad_norm": 0.035888671875,
+      "learning_rate": 1.989753177783094e-05,
+      "loss": 0.9201,
+      "step": 115
+    },
+    {
+      "epoch": 0.08,
+      "grad_norm": 0.0341796875,
+      "learning_rate": 1.9894501195761393e-05,
+      "loss": 0.9299,
+      "step": 116
+    },
+    {
+      "epoch": 0.08,
+      "grad_norm": 0.03369140625,
+      "learning_rate": 1.9891426686018308e-05,
+      "loss": 0.8812,
+      "step": 117
+    },
+    {
+      "epoch": 0.08,
+      "grad_norm": 0.035888671875,
+      "learning_rate": 1.9888308262251286e-05,
+      "loss": 0.9995,
+      "step": 118
+    },
+    {
+      "epoch": 0.08,
+      "grad_norm": 0.031982421875,
+      "learning_rate": 1.9885145938304905e-05,
+      "loss": 0.8804,
+      "step": 119
+    },
+    {
+      "epoch": 0.08,
+      "grad_norm": 0.03955078125,
+      "learning_rate": 1.988193972821863e-05,
+      "loss": 0.9021,
+      "step": 120
+    },
+    {
+      "epoch": 0.08,
+      "grad_norm": 0.039306640625,
+      "learning_rate": 1.987868964622676e-05,
+      "loss": 0.8066,
+      "step": 121
+    },
+    {
+      "epoch": 0.08,
+      "grad_norm": 0.0361328125,
+      "learning_rate": 1.9875395706758388e-05,
+      "loss": 0.909,
+      "step": 122
+    },
+    {
+      "epoch": 0.08,
+      "grad_norm": 0.037109375,
+      "learning_rate": 1.987205792443729e-05,
+      "loss": 0.8611,
+      "step": 123
+    },
+    {
+      "epoch": 0.08,
+      "grad_norm": 0.037841796875,
+      "learning_rate": 1.9868676314081907e-05,
+      "loss": 0.9249,
+      "step": 124
+    },
+    {
+      "epoch": 0.08,
+      "grad_norm": 0.0322265625,
+      "learning_rate": 1.986525089070525e-05,
+      "loss": 0.837,
+      "step": 125
+    },
+    {
+      "epoch": 0.08,
+      "grad_norm": 0.03515625,
+      "learning_rate": 1.986178166951484e-05,
+      "loss": 0.8653,
+      "step": 126
+    },
+    {
+      "epoch": 0.08,
+      "grad_norm": 0.03662109375,
+      "learning_rate": 1.9858268665912653e-05,
+      "loss": 0.9011,
+      "step": 127
+    },
+    {
+      "epoch": 0.08,
+      "grad_norm": 0.035888671875,
+      "learning_rate": 1.9854711895495034e-05,
+      "loss": 0.9942,
+      "step": 128
+    },
+    {
+      "epoch": 0.08,
+      "grad_norm": 0.032958984375,
+      "learning_rate": 1.985111137405264e-05,
+      "loss": 0.9303,
+      "step": 129
+    },
+    {
+      "epoch": 0.08,
+      "grad_norm": 0.03369140625,
+      "learning_rate": 1.9847467117570364e-05,
+      "loss": 0.9206,
+      "step": 130
+    },
+    {
+      "epoch": 0.09,
+      "grad_norm": 0.033935546875,
+      "learning_rate": 1.9843779142227258e-05,
+      "loss": 0.8366,
+      "step": 131
+    },
+    {
+      "epoch": 0.09,
+      "grad_norm": 0.0380859375,
+      "learning_rate": 1.9840047464396477e-05,
+      "loss": 0.8988,
+      "step": 132
+    },
+    {
+      "epoch": 0.09,
+      "grad_norm": 0.06005859375,
+      "learning_rate": 1.98362721006452e-05,
+      "loss": 0.9719,
+      "step": 133
+    },
+    {
+      "epoch": 0.09,
+      "grad_norm": 0.03466796875,
+      "learning_rate": 1.983245306773454e-05,
+      "loss": 0.9629,
+      "step": 134
+    },
+    {
+      "epoch": 0.09,
+      "grad_norm": 0.0380859375,
+      "learning_rate": 1.98285903826195e-05,
+      "loss": 0.8384,
+      "step": 135
+    },
+    {
+      "epoch": 0.09,
+      "grad_norm": 0.034423828125,
+      "learning_rate": 1.9824684062448876e-05,
+      "loss": 0.8031,
+      "step": 136
+    },
+    {
+      "epoch": 0.09,
+      "grad_norm": 0.036376953125,
+      "learning_rate": 1.982073412456518e-05,
+      "loss": 0.8623,
+      "step": 137
+    },
+    {
+      "epoch": 0.09,
+      "grad_norm": 0.033935546875,
+      "learning_rate": 1.981674058650458e-05,
+      "loss": 0.8357,
+      "step": 138
+    },
+    {
+      "epoch": 0.09,
+      "grad_norm": 0.036376953125,
+      "learning_rate": 1.98127034659968e-05,
+      "loss": 0.9306,
+      "step": 139
+    },
+    {
+      "epoch": 0.09,
+      "grad_norm": 0.03564453125,
+      "learning_rate": 1.9808622780965064e-05,
+      "loss": 0.9464,
+      "step": 140
+    },
+    {
+      "epoch": 0.09,
+      "grad_norm": 0.033935546875,
+      "learning_rate": 1.9804498549526e-05,
+      "loss": 0.9146,
+      "step": 141
+    },
+    {
+      "epoch": 0.09,
+      "grad_norm": 0.034912109375,
+      "learning_rate": 1.980033078998956e-05,
+      "loss": 0.8999,
+      "step": 142
+    },
+    {
+      "epoch": 0.09,
+      "grad_norm": 0.03564453125,
+      "learning_rate": 1.9796119520858957e-05,
+      "loss": 0.9932,
+      "step": 143
+    },
+    {
+      "epoch": 0.09,
+      "grad_norm": 0.035888671875,
+      "learning_rate": 1.9791864760830554e-05,
+      "loss": 0.8976,
+      "step": 144
+    },
+    {
+      "epoch": 0.09,
+      "grad_norm": 0.03369140625,
+      "learning_rate": 1.9787566528793806e-05,
+      "loss": 0.9024,
+      "step": 145
+    },
+    {
+      "epoch": 0.09,
+      "grad_norm": 0.033447265625,
+      "learning_rate": 1.9783224843831162e-05,
+      "loss": 0.8262,
+      "step": 146
+    },
+    {
+      "epoch": 0.1,
+      "grad_norm": 0.036376953125,
+      "learning_rate": 1.977883972521799e-05,
+      "loss": 0.9491,
+      "step": 147
+    },
+    {
+      "epoch": 0.1,
+      "grad_norm": 0.0361328125,
+      "learning_rate": 1.9774411192422486e-05,
+      "loss": 0.9347,
+      "step": 148
+    },
+    {
+      "epoch": 0.1,
+      "grad_norm": 0.0390625,
+      "learning_rate": 1.9769939265105573e-05,
+      "loss": 0.8401,
+      "step": 149
+    },
+    {
+      "epoch": 0.1,
+      "grad_norm": 0.03466796875,
+      "learning_rate": 1.976542396312085e-05,
+      "loss": 0.8949,
+      "step": 150
+    },
+    {
+      "epoch": 0.1,
+      "grad_norm": 0.03369140625,
+      "learning_rate": 1.976086530651447e-05,
+      "loss": 0.8675,
+      "step": 151
+    },
+    {
+      "epoch": 0.1,
+      "grad_norm": 0.0322265625,
+      "learning_rate": 1.975626331552507e-05,
+      "loss": 0.8617,
+      "step": 152
+    },
+    {
+      "epoch": 0.1,
+      "grad_norm": 0.034423828125,
+      "learning_rate": 1.9751618010583665e-05,
+      "loss": 0.8374,
+      "step": 153
+    },
+    {
+      "epoch": 0.1,
+      "grad_norm": 0.036865234375,
+      "learning_rate": 1.974692941231357e-05,
+      "loss": 0.8396,
+      "step": 154
+    },
+    {
+      "epoch": 0.1,
+      "grad_norm": 0.034912109375,
+      "learning_rate": 1.974219754153032e-05,
+      "loss": 0.9553,
+      "step": 155
+    },
+    {
+      "epoch": 0.1,
+      "grad_norm": 0.0341796875,
+      "learning_rate": 1.9737422419241538e-05,
+      "loss": 0.8821,
+      "step": 156
+    },
+    {
+      "epoch": 0.1,
+      "grad_norm": 0.03466796875,
+      "learning_rate": 1.9732604066646882e-05,
+      "loss": 0.8778,
+      "step": 157
+    },
+    {
+      "epoch": 0.1,
+      "grad_norm": 0.03955078125,
+      "learning_rate": 1.9727742505137936e-05,
+      "loss": 0.8552,
+      "step": 158
+    },
+    {
+      "epoch": 0.1,
+      "grad_norm": 0.03515625,
+      "learning_rate": 1.9722837756298112e-05,
+      "loss": 0.9358,
+      "step": 159
+    },
+    {
+      "epoch": 0.1,
+      "grad_norm": 0.03466796875,
+      "learning_rate": 1.9717889841902553e-05,
+      "loss": 0.9171,
+      "step": 160
+    },
+    {
+      "epoch": 0.1,
+      "grad_norm": 0.033447265625,
+      "learning_rate": 1.971289878391804e-05,
+      "loss": 0.8395,
+      "step": 161
+    },
+    {
+      "epoch": 0.11,
+      "grad_norm": 0.035400390625,
+      "learning_rate": 1.97078646045029e-05,
+      "loss": 0.8955,
+      "step": 162
+    },
+    {
+      "epoch": 0.11,
+      "grad_norm": 0.0419921875,
+      "learning_rate": 1.9702787326006906e-05,
+      "loss": 0.8192,
+      "step": 163
+    },
+    {
+      "epoch": 0.11,
+      "grad_norm": 0.037353515625,
+      "learning_rate": 1.9697666970971153e-05,
+      "loss": 0.8264,
+      "step": 164
+    },
+    {
+      "epoch": 0.11,
+      "grad_norm": 0.03857421875,
+      "learning_rate": 1.9692503562128004e-05,
+      "loss": 0.9093,
+      "step": 165
+    },
+    {
+      "epoch": 0.11,
+      "grad_norm": 0.044189453125,
+      "learning_rate": 1.9687297122400952e-05,
+      "loss": 0.9446,
+      "step": 166
+    },
+    {
+      "epoch": 0.11,
+      "grad_norm": 0.03466796875,
+      "learning_rate": 1.9682047674904527e-05,
+      "loss": 0.8802,
+      "step": 167
+    },
+    {
+      "epoch": 0.11,
+      "grad_norm": 0.03564453125,
+      "learning_rate": 1.9676755242944202e-05,
+      "loss": 0.9152,
+      "step": 168
+    },
+    {
+      "epoch": 0.11,
+      "grad_norm": 0.032958984375,
+      "learning_rate": 1.9671419850016283e-05,
+      "loss": 0.8396,
+      "step": 169
+    },
+    {
+      "epoch": 0.11,
+      "grad_norm": 0.0341796875,
+      "learning_rate": 1.9666041519807802e-05,
+      "loss": 0.7976,
+      "step": 170
+    },
+    {
+      "epoch": 0.11,
+      "grad_norm": 0.036376953125,
+      "learning_rate": 1.966062027619643e-05,
+      "loss": 0.8979,
+      "step": 171
+    },
+    {
+      "epoch": 0.11,
+      "grad_norm": 0.032470703125,
+      "learning_rate": 1.9655156143250328e-05,
+      "loss": 0.8632,
+      "step": 172
+    },
+    {
+      "epoch": 0.11,
+      "grad_norm": 0.0390625,
+      "learning_rate": 1.96496491452281e-05,
+      "loss": 0.9456,
+      "step": 173
+    },
+    {
+      "epoch": 0.11,
+      "grad_norm": 0.0361328125,
+      "learning_rate": 1.9644099306578636e-05,
+      "loss": 0.837,
+      "step": 174
+    },
+    {
+      "epoch": 0.11,
+      "grad_norm": 0.031982421875,
+      "learning_rate": 1.9638506651941024e-05,
+      "loss": 0.7911,
+      "step": 175
+    },
+    {
+      "epoch": 0.11,
+      "grad_norm": 0.03564453125,
+      "learning_rate": 1.963287120614444e-05,
+      "loss": 0.8926,
+      "step": 176
+    },
+    {
+      "epoch": 0.12,
+      "grad_norm": 0.0341796875,
+      "learning_rate": 1.9627192994208038e-05,
+      "loss": 0.8054,
+      "step": 177
+    },
+    {
+      "epoch": 0.12,
+      "grad_norm": 0.034423828125,
+      "learning_rate": 1.962147204134083e-05,
+      "loss": 0.9226,
+      "step": 178
+    },
+    {
+      "epoch": 0.12,
+      "grad_norm": 0.035400390625,
+      "learning_rate": 1.9615708372941588e-05,
+      "loss": 0.8987,
+      "step": 179
+    },
+    {
+      "epoch": 0.12,
+      "grad_norm": 0.040283203125,
+      "learning_rate": 1.960990201459872e-05,
+      "loss": 0.8729,
+      "step": 180
+    },
+    {
+      "epoch": 0.12,
+      "grad_norm": 0.039306640625,
+      "learning_rate": 1.960405299209016e-05,
+      "loss": 0.9454,
+      "step": 181
+    },
+    {
+      "epoch": 0.12,
+      "grad_norm": 0.035888671875,
+      "learning_rate": 1.9598161331383258e-05,
+      "loss": 0.9157,
+      "step": 182
+    },
+    {
+      "epoch": 0.12,
+      "grad_norm": 0.03857421875,
+      "learning_rate": 1.9592227058634655e-05,
+      "loss": 0.8724,
+      "step": 183
+    },
+    {
+      "epoch": 0.12,
+      "grad_norm": 0.0361328125,
+      "learning_rate": 1.958625020019018e-05,
+      "loss": 0.8446,
+      "step": 184
+    },
+    {
+      "epoch": 0.12,
+      "grad_norm": 0.0341796875,
+      "learning_rate": 1.9580230782584722e-05,
+      "loss": 0.8441,
+      "step": 185
+    },
+    {
+      "epoch": 0.12,
+      "grad_norm": 0.037841796875,
+      "learning_rate": 1.957416883254211e-05,
+      "loss": 0.9078,
+      "step": 186
+    },
+    {
+      "epoch": 0.12,
+      "grad_norm": 0.037109375,
+      "learning_rate": 1.9568064376975013e-05,
+      "loss": 0.9075,
+      "step": 187
+    },
+    {
+      "epoch": 0.12,
+      "grad_norm": 0.036376953125,
+      "learning_rate": 1.956191744298479e-05,
+      "loss": 0.8932,
+      "step": 188
+    },
+    {
+      "epoch": 0.12,
+      "grad_norm": 0.03466796875,
+      "learning_rate": 1.955572805786141e-05,
+      "loss": 0.8577,
+      "step": 189
+    },
+    {
+      "epoch": 0.12,
+      "grad_norm": 0.03564453125,
+      "learning_rate": 1.9549496249083288e-05,
+      "loss": 0.8257,
+      "step": 190
+    },
+    {
+      "epoch": 0.12,
+      "grad_norm": 0.03369140625,
+      "learning_rate": 1.954322204431719e-05,
+      "loss": 0.7848,
+      "step": 191
+    },
+    {
+      "epoch": 0.12,
+      "grad_norm": 0.037353515625,
+      "learning_rate": 1.953690547141811e-05,
+      "loss": 0.8617,
+      "step": 192
+    },
+    {
+      "epoch": 0.13,
+      "grad_norm": 0.034912109375,
+      "learning_rate": 1.953054655842913e-05,
+      "loss": 0.7992,
+      "step": 193
+    },
+    {
+      "epoch": 0.13,
+      "grad_norm": 0.03515625,
+      "learning_rate": 1.9524145333581315e-05,
+      "loss": 0.8101,
+      "step": 194
+    },
+    {
+      "epoch": 0.13,
+      "grad_norm": 0.038330078125,
+      "learning_rate": 1.951770182529357e-05,
+      "loss": 0.8669,
+      "step": 195
+    },
+    {
+      "epoch": 0.13,
+      "grad_norm": 0.03564453125,
+      "learning_rate": 1.951121606217252e-05,
+      "loss": 0.8589,
+      "step": 196
+    },
+    {
+      "epoch": 0.13,
+      "grad_norm": 0.036376953125,
+      "learning_rate": 1.9504688073012397e-05,
+      "loss": 0.9205,
+      "step": 197
+    },
+    {
+      "epoch": 0.13,
+      "grad_norm": 0.039794921875,
+      "learning_rate": 1.9498117886794885e-05,
+      "loss": 0.9052,
+      "step": 198
+    },
+    {
+      "epoch": 0.13,
+      "grad_norm": 0.037109375,
+      "learning_rate": 1.9491505532689017e-05,
+      "loss": 0.8167,
+      "step": 199
+    },
+    {
+      "epoch": 0.13,
+      "grad_norm": 0.03662109375,
+      "learning_rate": 1.948485104005103e-05,
+      "loss": 0.9358,
+      "step": 200
+    },
+    {
+      "epoch": 0.13,
+      "grad_norm": 0.03759765625,
+      "learning_rate": 1.947815443842424e-05,
+      "loss": 0.8639,
+      "step": 201
+    },
+    {
+      "epoch": 0.13,
+      "grad_norm": 0.034423828125,
+      "learning_rate": 1.9471415757538918e-05,
+      "loss": 0.8684,
+      "step": 202
+    },
+    {
+      "epoch": 0.13,
+      "grad_norm": 0.032470703125,
+      "learning_rate": 1.946463502731213e-05,
+      "loss": 0.7762,
+      "step": 203
+    },
+    {
+      "epoch": 0.13,
+      "grad_norm": 0.034912109375,
+      "learning_rate": 1.9457812277847645e-05,
+      "loss": 0.8664,
+      "step": 204
+    },
+    {
+      "epoch": 0.13,
+      "grad_norm": 0.038330078125,
+      "learning_rate": 1.945094753943577e-05,
+      "loss": 0.9964,
+      "step": 205
+    },
+    {
+      "epoch": 0.13,
+      "grad_norm": 0.037353515625,
+      "learning_rate": 1.944404084255324e-05,
+      "loss": 0.8768,
+      "step": 206
+    },
+    {
+      "epoch": 0.13,
+      "grad_norm": 0.0380859375,
+      "learning_rate": 1.9437092217863043e-05,
+      "loss": 0.8999,
+      "step": 207
+    },
+    {
+      "epoch": 0.14,
+      "grad_norm": 0.036376953125,
+      "learning_rate": 1.9430101696214335e-05,
+      "loss": 0.8437,
+      "step": 208
+    },
+    {
+      "epoch": 0.14,
+      "grad_norm": 0.037841796875,
+      "learning_rate": 1.9423069308642267e-05,
+      "loss": 0.8273,
+      "step": 209
+    },
+    {
+      "epoch": 0.14,
+      "grad_norm": 0.04052734375,
+      "learning_rate": 1.9415995086367858e-05,
+      "loss": 0.9275,
+      "step": 210
+    },
+    {
+      "epoch": 0.14,
+      "grad_norm": 0.036865234375,
+      "learning_rate": 1.940887906079786e-05,
+      "loss": 0.8938,
+      "step": 211
+    },
+    {
+      "epoch": 0.14,
+      "grad_norm": 0.033203125,
+      "learning_rate": 1.9401721263524616e-05,
+      "loss": 0.8414,
+      "step": 212
+    },
+    {
+      "epoch": 0.14,
+      "grad_norm": 0.037109375,
+      "learning_rate": 1.9394521726325907e-05,
+      "loss": 0.9055,
+      "step": 213
+    },
+    {
+      "epoch": 0.14,
+      "grad_norm": 0.042236328125,
+      "learning_rate": 1.938728048116484e-05,
+      "loss": 0.9002,
+      "step": 214
+    },
+    {
+      "epoch": 0.14,
+      "grad_norm": 0.034912109375,
+      "learning_rate": 1.9379997560189677e-05,
+      "loss": 0.8598,
+      "step": 215
+    },
+    {
+      "epoch": 0.14,
+      "grad_norm": 0.038330078125,
+      "learning_rate": 1.9372672995733706e-05,
+      "loss": 0.8557,
+      "step": 216
+    },
+    {
+      "epoch": 0.14,
+      "grad_norm": 0.036376953125,
+      "learning_rate": 1.9365306820315104e-05,
+      "loss": 0.9001,
+      "step": 217
+    },
+    {
+      "epoch": 0.14,
+      "grad_norm": 0.037109375,
+      "learning_rate": 1.9357899066636774e-05,
+      "loss": 0.842,
+      "step": 218
+    },
+    {
+      "epoch": 0.14,
+      "grad_norm": 0.036865234375,
+      "learning_rate": 1.935044976758621e-05,
+      "loss": 0.8759,
+      "step": 219
+    },
+    {
+      "epoch": 0.14,
+      "grad_norm": 0.03759765625,
+      "learning_rate": 1.9342958956235365e-05,
+      "loss": 0.8306,
+      "step": 220
+    },
+    {
+      "epoch": 0.14,
+      "grad_norm": 0.03759765625,
+      "learning_rate": 1.933542666584047e-05,
+      "loss": 0.8322,
+      "step": 221
+    },
+    {
+      "epoch": 0.14,
+      "grad_norm": 0.03515625,
+      "learning_rate": 1.9327852929841918e-05,
+      "loss": 0.8149,
+      "step": 222
+    },
+    {
+      "epoch": 0.14,
+      "grad_norm": 0.03955078125,
+      "learning_rate": 1.9320237781864106e-05,
+      "loss": 0.8458,
+      "step": 223
+    },
+    {
+      "epoch": 0.15,
+      "grad_norm": 0.03759765625,
+      "learning_rate": 1.9312581255715276e-05,
+      "loss": 0.84,
+      "step": 224
+    },
+    {
+      "epoch": 0.15,
+      "grad_norm": 0.038818359375,
+      "learning_rate": 1.9304883385387383e-05,
+      "loss": 0.8254,
+      "step": 225
+    },
+    {
+      "epoch": 0.15,
+      "grad_norm": 0.03466796875,
+      "learning_rate": 1.9297144205055925e-05,
+      "loss": 0.8898,
+      "step": 226
+    },
+    {
+      "epoch": 0.15,
+      "grad_norm": 0.037353515625,
+      "learning_rate": 1.9289363749079798e-05,
+      "loss": 0.8231,
+      "step": 227
+    },
+    {
+      "epoch": 0.15,
+      "grad_norm": 0.041015625,
+      "learning_rate": 1.928154205200116e-05,
+      "loss": 0.8764,
+      "step": 228
+    },
+    {
+      "epoch": 0.15,
+      "grad_norm": 0.037353515625,
+      "learning_rate": 1.9273679148545246e-05,
+      "loss": 0.8436,
+      "step": 229
+    },
+    {
+      "epoch": 0.15,
+      "grad_norm": 0.037841796875,
+      "learning_rate": 1.9265775073620244e-05,
+      "loss": 0.8622,
+      "step": 230
+    },
+    {
+      "epoch": 0.15,
+      "grad_norm": 0.036865234375,
+      "learning_rate": 1.9257829862317118e-05,
+      "loss": 0.8484,
+      "step": 231
+    },
+    {
+      "epoch": 0.15,
+      "grad_norm": 0.037841796875,
+      "learning_rate": 1.9249843549909467e-05,
+      "loss": 0.8765,
+      "step": 232
+    },
+    {
+      "epoch": 0.15,
+      "grad_norm": 0.03759765625,
+      "learning_rate": 1.9241816171853362e-05,
+      "loss": 0.8762,
+      "step": 233
+    },
+    {
+      "epoch": 0.15,
+      "grad_norm": 0.03955078125,
+      "learning_rate": 1.9233747763787187e-05,
+      "loss": 0.8716,
+      "step": 234
+    },
+    {
+      "epoch": 0.15,
+      "grad_norm": 0.04443359375,
+      "learning_rate": 1.9225638361531482e-05,
+      "loss": 0.8453,
+      "step": 235
+    },
+    {
+      "epoch": 0.15,
+      "grad_norm": 0.037109375,
+      "learning_rate": 1.9217488001088784e-05,
+      "loss": 0.7992,
+      "step": 236
+    },
+    {
+      "epoch": 0.15,
+      "grad_norm": 0.036865234375,
+      "learning_rate": 1.920929671864348e-05,
+      "loss": 0.9607,
+      "step": 237
+    },
+    {
+      "epoch": 0.15,
+      "grad_norm": 0.037841796875,
+      "learning_rate": 1.920106455056162e-05,
+      "loss": 0.8416,
+      "step": 238
+    },
+    {
+      "epoch": 0.16,
+      "grad_norm": 0.0380859375,
+      "learning_rate": 1.9192791533390778e-05,
+      "loss": 0.7983,
+      "step": 239
+    },
+    {
+      "epoch": 0.16,
+      "grad_norm": 0.040283203125,
+      "learning_rate": 1.9184477703859876e-05,
+      "loss": 0.8942,
+      "step": 240
+    },
+    {
+      "epoch": 0.16,
+      "grad_norm": 0.037841796875,
+      "learning_rate": 1.9176123098879035e-05,
+      "loss": 0.8849,
+      "step": 241
+    },
+    {
+      "epoch": 0.16,
+      "grad_norm": 0.037109375,
+      "learning_rate": 1.9167727755539393e-05,
+      "loss": 0.83,
+      "step": 242
+    },
+    {
+      "epoch": 0.16,
+      "grad_norm": 0.036865234375,
+      "learning_rate": 1.9159291711112962e-05,
+      "loss": 0.7999,
+      "step": 243
+    },
+    {
+      "epoch": 0.16,
+      "grad_norm": 0.035400390625,
+      "learning_rate": 1.9150815003052436e-05,
+      "loss": 0.8281,
+      "step": 244
+    },
+    {
+      "epoch": 0.16,
+      "grad_norm": 0.038818359375,
+      "learning_rate": 1.9142297668991053e-05,
+      "loss": 0.884,
+      "step": 245
+    },
+    {
+      "epoch": 0.16,
+      "grad_norm": 0.044189453125,
+      "learning_rate": 1.913373974674241e-05,
+      "loss": 0.8701,
+      "step": 246
+    },
+    {
+      "epoch": 0.16,
+      "grad_norm": 0.039306640625,
+      "learning_rate": 1.9125141274300293e-05,
+      "loss": 0.8734,
+      "step": 247
+    },
+    {
+      "epoch": 0.16,
+      "grad_norm": 0.03857421875,
+      "learning_rate": 1.9116502289838524e-05,
+      "loss": 0.8851,
+      "step": 248
+    },
+    {
+      "epoch": 0.16,
+      "grad_norm": 0.044189453125,
+      "learning_rate": 1.910782283171078e-05,
+      "loss": 0.9402,
+      "step": 249
+    },
+    {
+      "epoch": 0.16,
+      "grad_norm": 0.038818359375,
+      "learning_rate": 1.909910293845042e-05,
+      "loss": 0.831,
+      "step": 250
+    },
+    {
+      "epoch": 0.16,
+      "grad_norm": 0.038330078125,
+      "learning_rate": 1.909034264877032e-05,
+      "loss": 0.8093,
+      "step": 251
+    },
+    {
+      "epoch": 0.16,
+      "grad_norm": 0.039794921875,
+      "learning_rate": 1.9081542001562713e-05,
+      "loss": 0.9085,
+      "step": 252
+    },
+    {
+      "epoch": 0.16,
+      "grad_norm": 0.038330078125,
+      "learning_rate": 1.9072701035898985e-05,
+      "loss": 0.8466,
+      "step": 253
+    },
+    {
+      "epoch": 0.17,
+      "grad_norm": 0.0419921875,
+      "learning_rate": 1.906381979102953e-05,
+      "loss": 0.8938,
+      "step": 254
+    },
+    {
+      "epoch": 0.17,
+      "grad_norm": 0.0400390625,
+      "learning_rate": 1.9054898306383568e-05,
+      "loss": 0.8787,
+      "step": 255
+    },
+    {
+      "epoch": 0.17,
+      "grad_norm": 0.040283203125,
+      "learning_rate": 1.904593662156896e-05,
+      "loss": 0.882,
+      "step": 256
+    },
+    {
+      "epoch": 0.17,
+      "grad_norm": 0.03759765625,
+      "learning_rate": 1.903693477637204e-05,
+      "loss": 0.7803,
+      "step": 257
+    },
+    {
+      "epoch": 0.17,
+      "grad_norm": 0.037353515625,
+      "learning_rate": 1.902789281075745e-05,
+      "loss": 0.8078,
+      "step": 258
+    },
+    {
+      "epoch": 0.17,
+      "grad_norm": 0.03857421875,
+      "learning_rate": 1.9018810764867935e-05,
+      "loss": 0.8318,
+      "step": 259
+    },
+    {
+      "epoch": 0.17,
+      "grad_norm": 0.04150390625,
+      "learning_rate": 1.900968867902419e-05,
+      "loss": 0.8728,
+      "step": 260
+    },
+    {
+      "epoch": 0.17,
+      "grad_norm": 0.0390625,
+      "learning_rate": 1.9000526593724678e-05,
+      "loss": 0.836,
+      "step": 261
+    },
+    {
+      "epoch": 0.17,
+      "grad_norm": 0.0380859375,
+      "learning_rate": 1.8991324549645424e-05,
+      "loss": 0.9197,
+      "step": 262
+    },
+    {
+      "epoch": 0.17,
+      "grad_norm": 0.04150390625,
+      "learning_rate": 1.898208258763987e-05,
+      "loss": 0.7965,
+      "step": 263
+    },
+    {
+      "epoch": 0.17,
+      "grad_norm": 0.041015625,
+      "learning_rate": 1.897280074873868e-05,
+      "loss": 0.8078,
+      "step": 264
+    },
+    {
+      "epoch": 0.17,
+      "grad_norm": 0.042236328125,
+      "learning_rate": 1.8963479074149537e-05,
+      "loss": 0.9035,
+      "step": 265
+    },
+    {
+      "epoch": 0.17,
+      "grad_norm": 0.040771484375,
+      "learning_rate": 1.8954117605257e-05,
+      "loss": 0.8515,
+      "step": 266
+    },
+    {
+      "epoch": 0.17,
+      "grad_norm": 0.041015625,
+      "learning_rate": 1.8944716383622288e-05,
+      "loss": 0.8147,
+      "step": 267
+    },
+    {
+      "epoch": 0.17,
+      "grad_norm": 0.040771484375,
+      "learning_rate": 1.8935275450983102e-05,
+      "loss": 0.8121,
+      "step": 268
+    },
+    {
+      "epoch": 0.17,
+      "grad_norm": 0.04150390625,
+      "learning_rate": 1.8925794849253462e-05,
+      "loss": 0.858,
+      "step": 269
+    },
+    {
+      "epoch": 0.18,
+      "grad_norm": 0.044921875,
+      "learning_rate": 1.8916274620523482e-05,
+      "loss": 0.8502,
+      "step": 270
+    },
+    {
+      "epoch": 0.18,
+      "grad_norm": 0.04150390625,
+      "learning_rate": 1.8906714807059218e-05,
+      "loss": 0.8438,
+      "step": 271
+    },
+    {
+      "epoch": 0.18,
+      "grad_norm": 0.0390625,
+      "learning_rate": 1.889711545130246e-05,
+      "loss": 0.8464,
+      "step": 272
+    },
+    {
+      "epoch": 0.18,
+      "grad_norm": 0.05712890625,
+      "learning_rate": 1.8887476595870558e-05,
+      "loss": 0.8227,
+      "step": 273
+    },
+    {
+      "epoch": 0.18,
+      "grad_norm": 0.036865234375,
+      "learning_rate": 1.887779828355621e-05,
+      "loss": 0.8546,
+      "step": 274
+    },
+    {
+      "epoch": 0.18,
+      "grad_norm": 0.037109375,
+      "learning_rate": 1.8868080557327305e-05,
+      "loss": 0.8932,
+      "step": 275
+    },
+    {
+      "epoch": 0.18,
+      "grad_norm": 0.041748046875,
+      "learning_rate": 1.8858323460326704e-05,
+      "loss": 0.889,
+      "step": 276
+    },
+    {
+      "epoch": 0.18,
+      "grad_norm": 0.039306640625,
+      "learning_rate": 1.8848527035872057e-05,
+      "loss": 0.8174,
+      "step": 277
+    },
+    {
+      "epoch": 0.18,
+      "grad_norm": 0.038818359375,
+      "learning_rate": 1.883869132745561e-05,
+      "loss": 0.8183,
+      "step": 278
+    },
+    {
+      "epoch": 0.18,
+      "grad_norm": 0.0390625,
+      "learning_rate": 1.8828816378744035e-05,
+      "loss": 0.8924,
+      "step": 279
+    },
+    {
+      "epoch": 0.18,
+      "grad_norm": 0.038330078125,
+      "learning_rate": 1.8818902233578188e-05,
+      "loss": 0.7906,
+      "step": 280
+    },
+    {
+      "epoch": 0.18,
+      "grad_norm": 0.04248046875,
+      "learning_rate": 1.8808948935972965e-05,
+      "loss": 0.8118,
+      "step": 281
+    },
+    {
+      "epoch": 0.18,
+      "grad_norm": 0.03857421875,
+      "learning_rate": 1.8798956530117058e-05,
+      "loss": 0.8512,
+      "step": 282
+    },
+    {
+      "epoch": 0.18,
+      "grad_norm": 0.044677734375,
+      "learning_rate": 1.8788925060372806e-05,
+      "loss": 0.8224,
+      "step": 283
+    },
+    {
+      "epoch": 0.18,
+      "grad_norm": 0.047607421875,
+      "learning_rate": 1.8778854571275972e-05,
+      "loss": 0.8207,
+      "step": 284
+    },
+    {
+      "epoch": 0.19,
+      "grad_norm": 0.038330078125,
+      "learning_rate": 1.876874510753554e-05,
+      "loss": 0.8011,
+      "step": 285
+    },
+    {
+      "epoch": 0.19,
+      "grad_norm": 0.04296875,
+      "learning_rate": 1.875859671403354e-05,
+      "loss": 0.8132,
+      "step": 286
+    },
+    {
+      "epoch": 0.19,
+      "grad_norm": 0.042236328125,
+      "learning_rate": 1.874840943582482e-05,
+      "loss": 0.9056,
+      "step": 287
+    },
+    {
+      "epoch": 0.19,
+      "grad_norm": 0.0400390625,
+      "learning_rate": 1.8738183318136867e-05,
+      "loss": 0.8353,
+      "step": 288
+    },
+    {
+      "epoch": 0.19,
+      "grad_norm": 0.040771484375,
+      "learning_rate": 1.872791840636961e-05,
+      "loss": 0.7943,
+      "step": 289
+    },
+    {
+      "epoch": 0.19,
+      "grad_norm": 0.045654296875,
+      "learning_rate": 1.871761474609519e-05,
+      "loss": 0.8207,
+      "step": 290
+    },
+    {
+      "epoch": 0.19,
+      "grad_norm": 0.04345703125,
+      "learning_rate": 1.8707272383057785e-05,
+      "loss": 0.8415,
+      "step": 291
+    },
+    {
+      "epoch": 0.19,
+      "grad_norm": 0.04052734375,
+      "learning_rate": 1.8696891363173405e-05,
+      "loss": 0.797,
+      "step": 292
+    },
+    {
+      "epoch": 0.19,
+      "grad_norm": 0.046142578125,
+      "learning_rate": 1.8686471732529667e-05,
+      "loss": 0.8248,
+      "step": 293
+    },
+    {
+      "epoch": 0.19,
+      "grad_norm": 0.041259765625,
+      "learning_rate": 1.8676013537385614e-05,
+      "loss": 0.76,
+      "step": 294
+    },
+    {
+      "epoch": 0.19,
+      "grad_norm": 0.04150390625,
+      "learning_rate": 1.8665516824171497e-05,
+      "loss": 0.8362,
+      "step": 295
+    },
+    {
+      "epoch": 0.19,
+      "grad_norm": 0.040771484375,
+      "learning_rate": 1.865498163948858e-05,
+      "loss": 0.8093,
+      "step": 296
+    },
+    {
+      "epoch": 0.19,
+      "grad_norm": 0.0380859375,
+      "learning_rate": 1.864440803010891e-05,
+      "loss": 0.7735,
+      "step": 297
+    },
+    {
+      "epoch": 0.19,
+      "grad_norm": 0.041015625,
+      "learning_rate": 1.863379604297513e-05,
+      "loss": 0.8824,
+      "step": 298
+    },
+    {
+      "epoch": 0.19,
+      "grad_norm": 0.039794921875,
+      "learning_rate": 1.862314572520028e-05,
+      "loss": 0.8157,
+      "step": 299
+    },
+    {
+      "epoch": 0.2,
+      "grad_norm": 0.03857421875,
+      "learning_rate": 1.861245712406755e-05,
+      "loss": 0.8084,
+      "step": 300
+    },
+    {
+      "epoch": 0.2,
+      "grad_norm": 0.049072265625,
+      "learning_rate": 1.86017302870301e-05,
+      "loss": 0.7976,
+      "step": 301
+    },
+    {
+      "epoch": 0.2,
+      "grad_norm": 0.041259765625,
+      "learning_rate": 1.8590965261710856e-05,
+      "loss": 0.8406,
+      "step": 302
+    },
+    {
+      "epoch": 0.2,
+      "grad_norm": 0.041015625,
+      "learning_rate": 1.858016209590227e-05,
+      "loss": 0.8145,
+      "step": 303
+    },
+    {
+      "epoch": 0.2,
+      "grad_norm": 0.038818359375,
+      "learning_rate": 1.8569320837566128e-05,
+      "loss": 0.8142,
+      "step": 304
+    },
+    {
+      "epoch": 0.2,
+      "grad_norm": 0.04052734375,
+      "learning_rate": 1.8558441534833327e-05,
+      "loss": 0.8894,
+      "step": 305
+    },
+    {
+      "epoch": 0.2,
+      "grad_norm": 0.04296875,
+      "learning_rate": 1.8547524236003675e-05,
+      "loss": 0.8793,
+      "step": 306
+    },
+    {
+      "epoch": 0.2,
+      "grad_norm": 0.0380859375,
+      "learning_rate": 1.8536568989545662e-05,
+      "loss": 0.868,
+      "step": 307
+    },
+    {
+      "epoch": 0.2,
+      "grad_norm": 0.0419921875,
+      "learning_rate": 1.8525575844096243e-05,
+      "loss": 0.8572,
+      "step": 308
+    },
+    {
+      "epoch": 0.2,
+      "grad_norm": 0.04931640625,
+      "learning_rate": 1.8514544848460653e-05,
+      "loss": 0.7933,
+      "step": 309
+    },
+    {
+      "epoch": 0.2,
+      "grad_norm": 0.0390625,
+      "learning_rate": 1.8503476051612138e-05,
+      "loss": 0.8017,
+      "step": 310
+    },
+    {
+      "epoch": 0.2,
+      "grad_norm": 0.042724609375,
+      "learning_rate": 1.8492369502691785e-05,
+      "loss": 0.8317,
+      "step": 311
+    },
+    {
+      "epoch": 0.2,
+      "grad_norm": 0.04052734375,
+      "learning_rate": 1.8481225251008284e-05,
+      "loss": 0.8201,
+      "step": 312
+    },
+    {
+      "epoch": 0.2,
+      "grad_norm": 0.041748046875,
+      "learning_rate": 1.8470043346037698e-05,
+      "loss": 0.8258,
+      "step": 313
+    },
+    {
+      "epoch": 0.2,
+      "grad_norm": 0.0419921875,
+      "learning_rate": 1.8458823837423274e-05,
+      "loss": 0.8402,
+      "step": 314
+    },
+    {
+      "epoch": 0.2,
+      "grad_norm": 0.044921875,
+      "learning_rate": 1.8447566774975187e-05,
+      "loss": 0.9293,
+      "step": 315
+    },
+    {
+      "epoch": 0.21,
+      "grad_norm": 0.048583984375,
+      "learning_rate": 1.8436272208670346e-05,
+      "loss": 0.8716,
+      "step": 316
+    },
+    {
+      "epoch": 0.21,
+      "grad_norm": 0.0458984375,
+      "learning_rate": 1.842494018865216e-05,
+      "loss": 0.8868,
+      "step": 317
+    },
+    {
+      "epoch": 0.21,
+      "grad_norm": 0.047607421875,
+      "learning_rate": 1.841357076523032e-05,
+      "loss": 0.9027,
+      "step": 318
+    },
+    {
+      "epoch": 0.21,
+      "grad_norm": 0.044189453125,
+      "learning_rate": 1.840216398888057e-05,
+      "loss": 0.7936,
+      "step": 319
+    },
+    {
+      "epoch": 0.21,
+      "grad_norm": 0.0458984375,
+      "learning_rate": 1.8390719910244487e-05,
+      "loss": 0.8498,
+      "step": 320
+    },
+    {
+      "epoch": 0.21,
+      "grad_norm": 0.044677734375,
+      "learning_rate": 1.8379238580129256e-05,
+      "loss": 0.798,
+      "step": 321
+    },
+    {
+      "epoch": 0.21,
+      "grad_norm": 0.044677734375,
+      "learning_rate": 1.836772004950744e-05,
+      "loss": 0.8746,
+      "step": 322
+    },
+    {
+      "epoch": 0.21,
+      "grad_norm": 0.04541015625,
+      "learning_rate": 1.8356164369516772e-05,
+      "loss": 0.8658,
+      "step": 323
+    },
+    {
+      "epoch": 0.21,
+      "grad_norm": 0.0400390625,
+      "learning_rate": 1.834457159145989e-05,
+      "loss": 0.8299,
+      "step": 324
+    },
+    {
+      "epoch": 0.21,
+      "grad_norm": 0.0458984375,
+      "learning_rate": 1.8332941766804152e-05,
+      "loss": 0.8723,
+      "step": 325
+    },
+    {
+      "epoch": 0.21,
+      "grad_norm": 0.04150390625,
+      "learning_rate": 1.832127494718138e-05,
+      "loss": 0.8311,
+      "step": 326
+    },
+    {
+      "epoch": 0.21,
+      "grad_norm": 0.0439453125,
+      "learning_rate": 1.830957118438764e-05,
+      "loss": 0.8159,
+      "step": 327
+    },
+    {
+      "epoch": 0.21,
+      "grad_norm": 0.044921875,
+      "learning_rate": 1.829783053038301e-05,
+      "loss": 0.8351,
+      "step": 328
+    },
+    {
+      "epoch": 0.21,
+      "grad_norm": 0.0419921875,
+      "learning_rate": 1.8286053037291356e-05,
+      "loss": 0.7679,
+      "step": 329
+    },
+    {
+      "epoch": 0.21,
+      "grad_norm": 0.04345703125,
+      "learning_rate": 1.8274238757400096e-05,
+      "loss": 0.7848,
+      "step": 330
+    },
+    {
+      "epoch": 0.22,
+      "grad_norm": 0.04150390625,
+      "learning_rate": 1.826238774315995e-05,
+      "loss": 0.8741,
+      "step": 331
+    },
+    {
+      "epoch": 0.22,
+      "grad_norm": 0.042236328125,
+      "learning_rate": 1.8250500047184744e-05,
+      "loss": 0.8517,
+      "step": 332
+    },
+    {
+      "epoch": 0.22,
+      "grad_norm": 0.046875,
+      "learning_rate": 1.8238575722251144e-05,
+      "loss": 0.8602,
+      "step": 333
+    },
+    {
+      "epoch": 0.22,
+      "grad_norm": 0.041259765625,
+      "learning_rate": 1.8226614821298444e-05,
+      "loss": 0.8087,
+      "step": 334
+    },
+    {
+      "epoch": 0.22,
+      "grad_norm": 0.042724609375,
+      "learning_rate": 1.821461739742831e-05,
+      "loss": 0.8301,
+      "step": 335
+    },
+    {
+      "epoch": 0.22,
+      "grad_norm": 0.046875,
+      "learning_rate": 1.820258350390456e-05,
+      "loss": 0.8342,
+      "step": 336
+    },
+    {
+      "epoch": 0.22,
+      "grad_norm": 0.043701171875,
+      "learning_rate": 1.819051319415293e-05,
+      "loss": 0.8249,
+      "step": 337
+    },
+    {
+      "epoch": 0.22,
+      "grad_norm": 0.041748046875,
+      "learning_rate": 1.817840652176082e-05,
+      "loss": 0.7909,
+      "step": 338
+    },
+    {
+      "epoch": 0.22,
+      "grad_norm": 0.04248046875,
+      "learning_rate": 1.8166263540477068e-05,
+      "loss": 0.8071,
+      "step": 339
+    },
+    {
+      "epoch": 0.22,
+      "grad_norm": 0.043212890625,
+      "learning_rate": 1.815408430421171e-05,
+      "loss": 0.7983,
+      "step": 340
+    },
+    {
+      "epoch": 0.22,
+      "grad_norm": 0.041748046875,
+      "learning_rate": 1.8141868867035745e-05,
+      "loss": 0.7877,
+      "step": 341
+    },
+    {
+      "epoch": 0.22,
+      "grad_norm": 0.04443359375,
+      "learning_rate": 1.8129617283180878e-05,
+      "loss": 0.9056,
+      "step": 342
+    },
+    {
+      "epoch": 0.22,
+      "grad_norm": 0.043212890625,
+      "learning_rate": 1.81173296070393e-05,
+      "loss": 0.8708,
+      "step": 343
+    },
+    {
+      "epoch": 0.22,
+      "grad_norm": 0.04541015625,
+      "learning_rate": 1.8105005893163436e-05,
+      "loss": 0.8387,
+      "step": 344
+    },
+    {
+      "epoch": 0.22,
+      "grad_norm": 0.042724609375,
+      "learning_rate": 1.8092646196265705e-05,
+      "loss": 0.8578,
+      "step": 345
+    },
+    {
+      "epoch": 0.22,
+      "grad_norm": 0.042236328125,
+      "learning_rate": 1.808025057121827e-05,
+      "loss": 0.8642,
+      "step": 346
+    },
+    {
+      "epoch": 0.23,
+      "grad_norm": 0.048095703125,
+      "learning_rate": 1.8067819073052813e-05,
+      "loss": 0.8058,
+      "step": 347
+    },
+    {
+      "epoch": 0.23,
+      "grad_norm": 0.041259765625,
+      "learning_rate": 1.8055351756960262e-05,
+      "loss": 0.8128,
+      "step": 348
+    },
+    {
+      "epoch": 0.23,
+      "grad_norm": 0.04296875,
+      "learning_rate": 1.804284867829058e-05,
+      "loss": 0.8387,
+      "step": 349
+    },
+    {
+      "epoch": 0.23,
+      "grad_norm": 0.0458984375,
+      "learning_rate": 1.8030309892552488e-05,
+      "loss": 0.9106,
+      "step": 350
+    },
+    {
+      "epoch": 0.23,
+      "grad_norm": 0.041259765625,
+      "learning_rate": 1.801773545541324e-05,
+      "loss": 0.752,
+      "step": 351
+    },
+    {
+      "epoch": 0.23,
+      "grad_norm": 0.044189453125,
+      "learning_rate": 1.800512542269836e-05,
+      "loss": 0.881,
+      "step": 352
+    },
+    {
+      "epoch": 0.23,
+      "grad_norm": 0.045166015625,
+      "learning_rate": 1.7992479850391416e-05,
+      "loss": 0.8004,
+      "step": 353
+    },
+    {
+      "epoch": 0.23,
+      "grad_norm": 0.046875,
+      "learning_rate": 1.797979879463375e-05,
+      "loss": 0.8075,
+      "step": 354
+    },
+    {
+      "epoch": 0.23,
+      "grad_norm": 0.044677734375,
+      "learning_rate": 1.796708231172423e-05,
+      "loss": 0.8315,
+      "step": 355
+    },
+    {
+      "epoch": 0.23,
+      "grad_norm": 0.041259765625,
+      "learning_rate": 1.795433045811901e-05,
+      "loss": 0.8506,
+      "step": 356
+    },
+    {
+      "epoch": 0.23,
+      "grad_norm": 0.051025390625,
+      "learning_rate": 1.7941543290431286e-05,
+      "loss": 0.8314,
+      "step": 357
+    },
+    {
+      "epoch": 0.23,
+      "grad_norm": 0.045166015625,
+      "learning_rate": 1.792872086543103e-05,
+      "loss": 0.7697,
+      "step": 358
+    },
+    {
+      "epoch": 0.23,
+      "grad_norm": 0.04541015625,
+      "learning_rate": 1.7915863240044727e-05,
+      "loss": 0.9001,
+      "step": 359
+    },
+    {
+      "epoch": 0.23,
+      "grad_norm": 0.04443359375,
+      "learning_rate": 1.7902970471355162e-05,
+      "loss": 0.7685,
+      "step": 360
+    },
+    {
+      "epoch": 0.23,
+      "grad_norm": 0.0458984375,
+      "learning_rate": 1.7890042616601125e-05,
+      "loss": 0.8105,
+      "step": 361
+    },
+    {
+      "epoch": 0.24,
+      "grad_norm": 0.044677734375,
+      "learning_rate": 1.7877079733177185e-05,
+      "loss": 0.9061,
+      "step": 362
+    },
+    {
+      "epoch": 0.24,
+      "grad_norm": 0.043212890625,
+      "learning_rate": 1.7864081878633414e-05,
+      "loss": 0.813,
+      "step": 363
+    },
+    {
+      "epoch": 0.24,
+      "grad_norm": 0.04345703125,
+      "learning_rate": 1.785104911067515e-05,
+      "loss": 0.8197,
+      "step": 364
+    },
+    {
+      "epoch": 0.24,
+      "grad_norm": 0.044189453125,
+      "learning_rate": 1.783798148716273e-05,
+      "loss": 0.894,
+      "step": 365
+    },
+    {
+      "epoch": 0.24,
+      "grad_norm": 0.04296875,
+      "learning_rate": 1.782487906611124e-05,
+      "loss": 0.7809,
+      "step": 366
+    },
+    {
+      "epoch": 0.24,
+      "grad_norm": 0.04296875,
+      "learning_rate": 1.781174190569024e-05,
+      "loss": 0.8428,
+      "step": 367
+    },
+    {
+      "epoch": 0.24,
+      "grad_norm": 0.049072265625,
+      "learning_rate": 1.7798570064223536e-05,
+      "loss": 0.8276,
+      "step": 368
+    },
+    {
+      "epoch": 0.24,
+      "grad_norm": 0.04541015625,
+      "learning_rate": 1.7785363600188894e-05,
+      "loss": 0.7937,
+      "step": 369
+    },
+    {
+      "epoch": 0.24,
+      "grad_norm": 0.03955078125,
+      "learning_rate": 1.7772122572217796e-05,
+      "loss": 0.7835,
+      "step": 370
+    },
+    {
+      "epoch": 0.24,
+      "grad_norm": 0.04833984375,
+      "learning_rate": 1.7758847039095167e-05,
+      "loss": 0.8456,
+      "step": 371
+    },
+    {
+      "epoch": 0.24,
+      "grad_norm": 0.0439453125,
+      "learning_rate": 1.774553705975913e-05,
+      "loss": 0.8483,
+      "step": 372
+    },
+    {
+      "epoch": 0.24,
+      "grad_norm": 0.041259765625,
+      "learning_rate": 1.773219269330073e-05,
+      "loss": 0.7902,
+      "step": 373
+    },
+    {
+      "epoch": 0.24,
+      "grad_norm": 0.04296875,
+      "learning_rate": 1.7718813998963678e-05,
+      "loss": 0.8734,
+      "step": 374
+    },
+    {
+      "epoch": 0.24,
+      "grad_norm": 0.044921875,
+      "learning_rate": 1.7705401036144086e-05,
+      "loss": 0.8646,
+      "step": 375
+    },
+    {
+      "epoch": 0.24,
+      "grad_norm": 0.04345703125,
+      "learning_rate": 1.7691953864390208e-05,
+      "loss": 0.8005,
+      "step": 376
+    },
+    {
+      "epoch": 0.25,
+      "grad_norm": 0.0419921875,
+      "learning_rate": 1.7678472543402166e-05,
+      "loss": 0.8701,
+      "step": 377
+    },
+    {
+      "epoch": 0.25,
+      "grad_norm": 0.045166015625,
+      "learning_rate": 1.7664957133031705e-05,
+      "loss": 0.8099,
+      "step": 378
+    },
+    {
+      "epoch": 0.25,
+      "grad_norm": 0.057373046875,
+      "learning_rate": 1.7651407693281896e-05,
+      "loss": 0.8524,
+      "step": 379
+    },
+    {
+      "epoch": 0.25,
+      "grad_norm": 0.05224609375,
+      "learning_rate": 1.7637824284306898e-05,
+      "loss": 0.8456,
+      "step": 380
+    },
+    {
+      "epoch": 0.25,
+      "grad_norm": 0.05078125,
+      "learning_rate": 1.762420696641167e-05,
+      "loss": 0.7977,
+      "step": 381
+    },
+    {
+      "epoch": 0.25,
+      "grad_norm": 0.044189453125,
+      "learning_rate": 1.7610555800051727e-05,
+      "loss": 0.7834,
+      "step": 382
+    },
+    {
+      "epoch": 0.25,
+      "grad_norm": 0.045166015625,
+      "learning_rate": 1.759687084583285e-05,
+      "loss": 0.7946,
+      "step": 383
+    },
+    {
+      "epoch": 0.25,
+      "grad_norm": 0.04443359375,
+      "learning_rate": 1.7583152164510827e-05,
+      "loss": 0.7456,
+      "step": 384
+    },
+    {
+      "epoch": 0.25,
+      "grad_norm": 0.044677734375,
+      "learning_rate": 1.7569399816991174e-05,
+      "loss": 0.8358,
+      "step": 385
+    },
+    {
+      "epoch": 0.25,
+      "grad_norm": 0.047119140625,
+      "learning_rate": 1.7555613864328876e-05,
+      "loss": 0.7976,
+      "step": 386
+    },
+    {
+      "epoch": 0.25,
+      "grad_norm": 0.09423828125,
+      "learning_rate": 1.754179436772812e-05,
+      "loss": 0.9486,
+      "step": 387
+    },
+    {
+      "epoch": 0.25,
+      "grad_norm": 0.046142578125,
+      "learning_rate": 1.7527941388542006e-05,
+      "loss": 0.7898,
+      "step": 388
+    },
+    {
+      "epoch": 0.25,
+      "grad_norm": 0.04150390625,
+      "learning_rate": 1.751405498827228e-05,
+      "loss": 0.7644,
+      "step": 389
+    },
+    {
+      "epoch": 0.25,
+      "grad_norm": 0.04931640625,
+      "learning_rate": 1.7500135228569067e-05,
+      "loss": 0.8363,
+      "step": 390
+    },
+    {
+      "epoch": 0.25,
+      "grad_norm": 0.047119140625,
+      "learning_rate": 1.748618217123061e-05,
+      "loss": 0.801,
+      "step": 391
+    },
+    {
+      "epoch": 0.25,
+      "grad_norm": 0.044677734375,
+      "learning_rate": 1.7472195878202955e-05,
+      "loss": 0.8487,
+      "step": 392
+    },
+    {
+      "epoch": 0.26,
+      "grad_norm": 0.046142578125,
+      "learning_rate": 1.7458176411579715e-05,
+      "loss": 0.8884,
+      "step": 393
+    },
+    {
+      "epoch": 0.26,
+      "grad_norm": 0.0439453125,
+      "learning_rate": 1.7444123833601784e-05,
+      "loss": 0.8484,
+      "step": 394
+    },
+    {
+      "epoch": 0.26,
+      "grad_norm": 0.043701171875,
+      "learning_rate": 1.743003820665705e-05,
+      "loss": 0.8325,
+      "step": 395
+    },
+    {
+      "epoch": 0.26,
+      "grad_norm": 0.048095703125,
+      "learning_rate": 1.741591959328013e-05,
+      "loss": 0.8061,
+      "step": 396
+    },
+    {
+      "epoch": 0.26,
+      "grad_norm": 0.047607421875,
+      "learning_rate": 1.7401768056152083e-05,
+      "loss": 0.7888,
+      "step": 397
+    },
+    {
+      "epoch": 0.26,
+      "grad_norm": 0.047119140625,
+      "learning_rate": 1.7387583658100144e-05,
+      "loss": 0.8564,
+      "step": 398
+    },
+    {
+      "epoch": 0.26,
+      "grad_norm": 0.044677734375,
+      "learning_rate": 1.737336646209742e-05,
+      "loss": 0.8412,
+      "step": 399
+    },
+    {
+      "epoch": 0.26,
+      "grad_norm": 0.0439453125,
+      "learning_rate": 1.7359116531262654e-05,
+      "loss": 0.9182,
+      "step": 400
+    },
+    {
+      "epoch": 0.26,
+      "grad_norm": 0.047119140625,
+      "learning_rate": 1.73448339288599e-05,
+      "loss": 0.8653,
+      "step": 401
+    },
+    {
+      "epoch": 0.26,
+      "grad_norm": 0.0478515625,
+      "learning_rate": 1.7330518718298263e-05,
+      "loss": 0.8174,
+      "step": 402
+    },
+    {
+      "epoch": 0.26,
+      "grad_norm": 0.05859375,
+      "learning_rate": 1.7316170963131627e-05,
+      "loss": 0.8621,
+      "step": 403
+    },
+    {
+      "epoch": 0.26,
+      "grad_norm": 0.047119140625,
+      "learning_rate": 1.7301790727058344e-05,
+      "loss": 0.7991,
+      "step": 404
+    },
+    {
+      "epoch": 0.26,
+      "grad_norm": 0.046142578125,
+      "learning_rate": 1.728737807392098e-05,
+      "loss": 0.8706,
+      "step": 405
+    },
+    {
+      "epoch": 0.26,
+      "grad_norm": 0.046142578125,
+      "learning_rate": 1.727293306770602e-05,
+      "loss": 0.824,
+      "step": 406
+    },
+    {
+      "epoch": 0.26,
+      "grad_norm": 0.0498046875,
+      "learning_rate": 1.7258455772543573e-05,
+      "loss": 0.9865,
+      "step": 407
+    },
+    {
+      "epoch": 0.27,
+      "grad_norm": 0.0537109375,
+      "learning_rate": 1.7243946252707115e-05,
+      "loss": 0.844,
+      "step": 408
+    },
+    {
+      "epoch": 0.27,
+      "grad_norm": 0.04052734375,
+      "learning_rate": 1.7229404572613174e-05,
+      "loss": 0.7566,
+      "step": 409
+    },
+    {
+      "epoch": 0.27,
+      "grad_norm": 0.0439453125,
+      "learning_rate": 1.721483079682106e-05,
+      "loss": 0.8393,
+      "step": 410
+    },
+    {
+      "epoch": 0.27,
+      "grad_norm": 0.0458984375,
+      "learning_rate": 1.7200224990032577e-05,
+      "loss": 0.7992,
+      "step": 411
+    },
+    {
+      "epoch": 0.27,
+      "grad_norm": 0.04443359375,
+      "learning_rate": 1.7185587217091727e-05,
+      "loss": 0.8862,
+      "step": 412
+    },
+    {
+      "epoch": 0.27,
+      "grad_norm": 0.046630859375,
+      "learning_rate": 1.7170917542984445e-05,
+      "loss": 0.8859,
+      "step": 413
+    },
+    {
+      "epoch": 0.27,
+      "grad_norm": 0.050537109375,
+      "learning_rate": 1.7156216032838275e-05,
+      "loss": 0.8738,
+      "step": 414
+    },
+    {
+      "epoch": 0.27,
+      "grad_norm": 0.047119140625,
+      "learning_rate": 1.7141482751922117e-05,
+      "loss": 0.8702,
+      "step": 415
+    },
+    {
+      "epoch": 0.27,
+      "grad_norm": 0.044677734375,
+      "learning_rate": 1.7126717765645908e-05,
+      "loss": 0.8496,
+      "step": 416
+    },
+    {
+      "epoch": 0.27,
+      "grad_norm": 0.047607421875,
+      "learning_rate": 1.7111921139560356e-05,
+      "loss": 0.8402,
+      "step": 417
+    },
+    {
+      "epoch": 0.27,
+      "grad_norm": 0.044677734375,
+      "learning_rate": 1.7097092939356622e-05,
+      "loss": 0.8719,
+      "step": 418
+    },
+    {
+      "epoch": 0.27,
+      "grad_norm": 0.05126953125,
+      "learning_rate": 1.7082233230866064e-05,
+      "loss": 0.865,
+      "step": 419
+    },
+    {
+      "epoch": 0.27,
+      "grad_norm": 0.0517578125,
+      "learning_rate": 1.7067342080059904e-05,
+      "loss": 0.8876,
+      "step": 420
+    },
+    {
+      "epoch": 0.27,
+      "grad_norm": 0.0478515625,
+      "learning_rate": 1.7052419553048965e-05,
+      "loss": 0.8594,
+      "step": 421
+    },
+    {
+      "epoch": 0.27,
+      "grad_norm": 0.041015625,
+      "learning_rate": 1.703746571608337e-05,
+      "loss": 0.7774,
+      "step": 422
+    },
+    {
+      "epoch": 0.28,
+      "grad_norm": 0.04638671875,
+      "learning_rate": 1.7022480635552243e-05,
+      "loss": 0.8357,
+      "step": 423
+    },
+    {
+      "epoch": 0.28,
+      "grad_norm": 0.044189453125,
+      "learning_rate": 1.700746437798342e-05,
+      "loss": 0.8365,
+      "step": 424
+    },
+    {
+      "epoch": 0.28,
+      "grad_norm": 0.0517578125,
+      "learning_rate": 1.6992417010043144e-05,
+      "loss": 0.7916,
+      "step": 425
+    },
+    {
+      "epoch": 0.28,
+      "grad_norm": 0.0439453125,
+      "learning_rate": 1.6977338598535776e-05,
+      "loss": 0.886,
+      "step": 426
+    },
+    {
+      "epoch": 0.28,
+      "grad_norm": 0.04638671875,
+      "learning_rate": 1.696222921040351e-05,
+      "loss": 0.8391,
+      "step": 427
+    },
+    {
+      "epoch": 0.28,
+      "grad_norm": 0.045166015625,
+      "learning_rate": 1.6947088912726054e-05,
+      "loss": 0.8403,
+      "step": 428
+    },
+    {
+      "epoch": 0.28,
+      "grad_norm": 0.046875,
+      "learning_rate": 1.693191777272034e-05,
+      "loss": 0.8048,
+      "step": 429
+    },
+    {
+      "epoch": 0.28,
+      "grad_norm": 0.049072265625,
+      "learning_rate": 1.6916715857740234e-05,
+      "loss": 0.7742,
+      "step": 430
+    },
+    {
+      "epoch": 0.28,
+      "grad_norm": 0.045166015625,
+      "learning_rate": 1.690148323527623e-05,
+      "loss": 0.7859,
+      "step": 431
+    },
+    {
+      "epoch": 0.28,
+      "grad_norm": 0.045166015625,
+      "learning_rate": 1.688621997295515e-05,
+      "loss": 0.7956,
+      "step": 432
+    },
+    {
+      "epoch": 0.28,
+      "grad_norm": 0.04443359375,
+      "learning_rate": 1.6870926138539837e-05,
+      "loss": 0.8672,
+      "step": 433
+    },
+    {
+      "epoch": 0.28,
+      "grad_norm": 0.060546875,
+      "learning_rate": 1.6855601799928877e-05,
+      "loss": 0.848,
+      "step": 434
+    },
+    {
+      "epoch": 0.28,
+      "grad_norm": 0.046875,
+      "learning_rate": 1.6840247025156272e-05,
+      "loss": 0.8125,
+      "step": 435
+    },
+    {
+      "epoch": 0.28,
+      "grad_norm": 0.052001953125,
+      "learning_rate": 1.6824861882391154e-05,
+      "loss": 0.8359,
+      "step": 436
+    },
+    {
+      "epoch": 0.28,
+      "grad_norm": 0.048583984375,
+      "learning_rate": 1.6809446439937472e-05,
+      "loss": 0.877,
+      "step": 437
+    },
+    {
+      "epoch": 0.28,
+      "grad_norm": 0.04833984375,
+      "learning_rate": 1.6794000766233697e-05,
+      "loss": 0.8408,
+      "step": 438
+    },
+    {
+      "epoch": 0.29,
+      "grad_norm": 0.04833984375,
+      "learning_rate": 1.6778524929852513e-05,
+      "loss": 0.8381,
+      "step": 439
+    },
+    {
+      "epoch": 0.29,
+      "grad_norm": 0.052490234375,
+      "learning_rate": 1.676301899950052e-05,
+      "loss": 0.782,
+      "step": 440
+    },
+    {
+      "epoch": 0.29,
+      "grad_norm": 0.04345703125,
+      "learning_rate": 1.674748304401791e-05,
+      "loss": 0.8621,
+      "step": 441
+    },
+    {
+      "epoch": 0.29,
+      "grad_norm": 0.04345703125,
+      "learning_rate": 1.673191713237819e-05,
+      "loss": 0.8012,
+      "step": 442
+    },
+    {
+      "epoch": 0.29,
+      "grad_norm": 0.05322265625,
+      "learning_rate": 1.671632133368785e-05,
+      "loss": 0.8245,
+      "step": 443
+    },
+    {
+      "epoch": 0.29,
+      "grad_norm": 0.0478515625,
+      "learning_rate": 1.670069571718607e-05,
+      "loss": 0.7882,
+      "step": 444
+    },
+    {
+      "epoch": 0.29,
+      "grad_norm": 0.045166015625,
+      "learning_rate": 1.6685040352244414e-05,
+      "loss": 0.8387,
+      "step": 445
+    },
+    {
+      "epoch": 0.29,
+      "grad_norm": 0.055419921875,
+      "learning_rate": 1.666935530836651e-05,
+      "loss": 0.7766,
+      "step": 446
+    },
+    {
+      "epoch": 0.29,
+      "grad_norm": 0.04931640625,
+      "learning_rate": 1.665364065518775e-05,
+      "loss": 0.8204,
+      "step": 447
+    },
+    {
+      "epoch": 0.29,
+      "grad_norm": 0.046630859375,
+      "learning_rate": 1.6637896462474986e-05,
+      "loss": 0.8133,
+      "step": 448
+    },
+    {
+      "epoch": 0.29,
+      "grad_norm": 0.048828125,
+      "learning_rate": 1.662212280012621e-05,
+      "loss": 0.85,
+      "step": 449
+    },
+    {
+      "epoch": 0.29,
+      "grad_norm": 0.045654296875,
+      "learning_rate": 1.660631973817024e-05,
+      "loss": 0.8247,
+      "step": 450
+    },
+    {
+      "epoch": 0.29,
+      "grad_norm": 0.048583984375,
+      "learning_rate": 1.6590487346766426e-05,
+      "loss": 0.8977,
+      "step": 451
+    },
+    {
+      "epoch": 0.29,
+      "grad_norm": 0.05712890625,
+      "learning_rate": 1.657462569620433e-05,
+      "loss": 0.8456,
+      "step": 452
+    },
+    {
+      "epoch": 0.29,
+      "grad_norm": 0.048583984375,
+      "learning_rate": 1.6558734856903406e-05,
+      "loss": 0.8369,
+      "step": 453
+    },
+    {
+      "epoch": 0.3,
+      "grad_norm": 0.04638671875,
+      "learning_rate": 1.6542814899412694e-05,
+      "loss": 0.8055,
+      "step": 454
+    },
+    {
+      "epoch": 0.3,
+      "grad_norm": 0.04443359375,
+      "learning_rate": 1.6526865894410526e-05,
+      "loss": 0.8358,
+      "step": 455
+    },
+    {
+      "epoch": 0.3,
+      "grad_norm": 0.046630859375,
+      "learning_rate": 1.651088791270416e-05,
+      "loss": 0.8094,
+      "step": 456
+    },
+    {
+      "epoch": 0.3,
+      "grad_norm": 0.04833984375,
+      "learning_rate": 1.6494881025229535e-05,
+      "loss": 0.8518,
+      "step": 457
+    },
+    {
+      "epoch": 0.3,
+      "grad_norm": 0.048583984375,
+      "learning_rate": 1.647884530305089e-05,
+      "loss": 0.9644,
+      "step": 458
+    },
+    {
+      "epoch": 0.3,
+      "grad_norm": 0.046630859375,
+      "learning_rate": 1.6462780817360502e-05,
+      "loss": 0.8415,
+      "step": 459
+    },
+    {
+      "epoch": 0.3,
+      "grad_norm": 0.050537109375,
+      "learning_rate": 1.644668763947833e-05,
+      "loss": 0.8764,
+      "step": 460
+    },
+    {
+      "epoch": 0.3,
+      "grad_norm": 0.044921875,
+      "learning_rate": 1.6430565840851723e-05,
+      "loss": 0.7737,
+      "step": 461
+    },
+    {
+      "epoch": 0.3,
+      "grad_norm": 0.047607421875,
+      "learning_rate": 1.641441549305509e-05,
+      "loss": 0.7559,
+      "step": 462
+    },
+    {
+      "epoch": 0.3,
+      "grad_norm": 0.044677734375,
+      "learning_rate": 1.6398236667789595e-05,
+      "loss": 0.7893,
+      "step": 463
+    },
+    {
+      "epoch": 0.3,
+      "grad_norm": 0.045166015625,
+      "learning_rate": 1.6382029436882826e-05,
+      "loss": 0.8285,
+      "step": 464
+    },
+    {
+      "epoch": 0.3,
+      "grad_norm": 0.05419921875,
+      "learning_rate": 1.636579387228848e-05,
+      "loss": 0.9,
+      "step": 465
+    },
+    {
+      "epoch": 0.3,
+      "grad_norm": 0.052978515625,
+      "learning_rate": 1.634953004608604e-05,
+      "loss": 0.9457,
+      "step": 466
+    },
+    {
+      "epoch": 0.3,
+      "grad_norm": 0.044921875,
+      "learning_rate": 1.6333238030480473e-05,
+      "loss": 0.8015,
+      "step": 467
+    },
+    {
+      "epoch": 0.3,
+      "grad_norm": 0.056884765625,
+      "learning_rate": 1.631691789780188e-05,
+      "loss": 0.8978,
+      "step": 468
+    },
+    {
+      "epoch": 0.3,
+      "grad_norm": 0.04443359375,
+      "learning_rate": 1.6300569720505198e-05,
+      "loss": 0.8787,
+      "step": 469
+    },
+    {
+      "epoch": 0.31,
+      "grad_norm": 0.047119140625,
+      "learning_rate": 1.6284193571169878e-05,
+      "loss": 0.8,
+      "step": 470
+    },
+    {
+      "epoch": 0.31,
+      "grad_norm": 0.0498046875,
+      "learning_rate": 1.6267789522499545e-05,
+      "loss": 0.7745,
+      "step": 471
+    },
+    {
+      "epoch": 0.31,
+      "grad_norm": 0.051025390625,
+      "learning_rate": 1.6251357647321685e-05,
+      "loss": 0.8191,
+      "step": 472
+    },
+    {
+      "epoch": 0.31,
+      "grad_norm": 0.046142578125,
+      "learning_rate": 1.6234898018587336e-05,
+      "loss": 0.7503,
+      "step": 473
+    },
+    {
+      "epoch": 0.31,
+      "grad_norm": 0.048095703125,
+      "learning_rate": 1.6218410709370735e-05,
+      "loss": 0.8587,
+      "step": 474
+    },
+    {
+      "epoch": 0.31,
+      "grad_norm": 0.05029296875,
+      "learning_rate": 1.6201895792869023e-05,
+      "loss": 0.8287,
+      "step": 475
+    },
+    {
+      "epoch": 0.31,
+      "grad_norm": 0.0517578125,
+      "learning_rate": 1.6185353342401896e-05,
+      "loss": 0.8017,
+      "step": 476
+    },
+    {
+      "epoch": 0.31,
+      "grad_norm": 0.04833984375,
+      "learning_rate": 1.6168783431411295e-05,
+      "loss": 0.8638,
+      "step": 477
+    },
+    {
+      "epoch": 0.31,
+      "grad_norm": 0.0498046875,
+      "learning_rate": 1.6152186133461075e-05,
+      "loss": 0.8111,
+      "step": 478
+    },
+    {
+      "epoch": 0.31,
+      "grad_norm": 0.0458984375,
+      "learning_rate": 1.6135561522236675e-05,
+      "loss": 0.7919,
+      "step": 479
+    },
+    {
+      "epoch": 0.31,
+      "grad_norm": 0.05078125,
+      "learning_rate": 1.6118909671544797e-05,
+      "loss": 0.8731,
+      "step": 480
+    },
+    {
+      "epoch": 0.31,
+      "grad_norm": 0.049072265625,
+      "learning_rate": 1.6102230655313076e-05,
+      "loss": 0.9006,
+      "step": 481
+    },
+    {
+      "epoch": 0.31,
+      "grad_norm": 0.044921875,
+      "learning_rate": 1.6085524547589747e-05,
+      "loss": 0.7621,
+      "step": 482
+    },
+    {
+      "epoch": 0.31,
+      "grad_norm": 0.050537109375,
+      "learning_rate": 1.6068791422543327e-05,
+      "loss": 0.9065,
+      "step": 483
+    },
+    {
+      "epoch": 0.31,
+      "grad_norm": 0.051025390625,
+      "learning_rate": 1.6052031354462275e-05,
+      "loss": 0.8989,
+      "step": 484
+    },
+    {
+      "epoch": 0.32,
+      "grad_norm": 0.047119140625,
+      "learning_rate": 1.6035244417754666e-05,
+      "loss": 0.82,
+      "step": 485
+    },
+    {
+      "epoch": 0.32,
+      "grad_norm": 0.048095703125,
+      "learning_rate": 1.6018430686947865e-05,
+      "loss": 0.8926,
+      "step": 486
+    },
+    {
+      "epoch": 0.32,
+      "grad_norm": 0.047607421875,
+      "learning_rate": 1.6001590236688187e-05,
+      "loss": 0.8628,
+      "step": 487
+    },
+    {
+      "epoch": 0.32,
+      "grad_norm": 0.0517578125,
+      "learning_rate": 1.5984723141740578e-05,
+      "loss": 0.8583,
+      "step": 488
+    },
+    {
+      "epoch": 0.32,
+      "grad_norm": 0.048095703125,
+      "learning_rate": 1.596782947698826e-05,
+      "loss": 0.8314,
+      "step": 489
+    },
+    {
+      "epoch": 0.32,
+      "grad_norm": 0.058837890625,
+      "learning_rate": 1.5950909317432436e-05,
+      "loss": 0.8222,
+      "step": 490
+    },
+    {
+      "epoch": 0.32,
+      "grad_norm": 0.0458984375,
+      "learning_rate": 1.593396273819192e-05,
+      "loss": 0.8258,
+      "step": 491
+    },
+    {
+      "epoch": 0.32,
+      "grad_norm": 0.048583984375,
+      "learning_rate": 1.591698981450283e-05,
+      "loss": 0.8246,
+      "step": 492
+    },
+    {
+      "epoch": 0.32,
+      "grad_norm": 0.046142578125,
+      "learning_rate": 1.5899990621718232e-05,
+      "loss": 0.868,
+      "step": 493
+    },
+    {
+      "epoch": 0.32,
+      "grad_norm": 0.046630859375,
+      "learning_rate": 1.588296523530782e-05,
+      "loss": 0.835,
+      "step": 494
+    },
+    {
+      "epoch": 0.32,
+      "grad_norm": 0.04736328125,
+      "learning_rate": 1.5865913730857583e-05,
+      "loss": 0.8298,
+      "step": 495
+    },
+    {
+      "epoch": 0.32,
+      "grad_norm": 0.04833984375,
+      "learning_rate": 1.584883618406946e-05,
+      "loss": 0.8405,
+      "step": 496
+    },
+    {
+      "epoch": 0.32,
+      "grad_norm": 0.051025390625,
+      "learning_rate": 1.5831732670761e-05,
+      "loss": 0.8951,
+      "step": 497
+    },
+    {
+      "epoch": 0.32,
+      "grad_norm": 0.044921875,
+      "learning_rate": 1.5814603266865046e-05,
+      "loss": 0.8396,
+      "step": 498
+    },
+    {
+      "epoch": 0.32,
+      "grad_norm": 0.04736328125,
+      "learning_rate": 1.5797448048429377e-05,
+      "loss": 0.8433,
+      "step": 499
+    },
+    {
+      "epoch": 0.33,
+      "grad_norm": 0.046142578125,
+      "learning_rate": 1.5780267091616383e-05,
+      "loss": 0.869,
+      "step": 500
+    },
+    {
+      "epoch": 0.33,
+      "grad_norm": 0.048095703125,
+      "learning_rate": 1.576306047270272e-05,
+      "loss": 0.8454,
+      "step": 501
+    },
+    {
+      "epoch": 0.33,
+      "grad_norm": 0.0478515625,
+      "learning_rate": 1.574582826807897e-05,
+      "loss": 0.7873,
+      "step": 502
+    },
+    {
+      "epoch": 0.33,
+      "grad_norm": 0.056884765625,
+      "learning_rate": 1.5728570554249312e-05,
+      "loss": 0.9116,
+      "step": 503
+    },
+    {
+      "epoch": 0.33,
+      "grad_norm": 0.0546875,
+      "learning_rate": 1.571128740783117e-05,
+      "loss": 0.8564,
+      "step": 504
+    },
+    {
+      "epoch": 0.33,
+      "grad_norm": 0.04833984375,
+      "learning_rate": 1.5693978905554886e-05,
+      "loss": 0.8302,
+      "step": 505
+    },
+    {
+      "epoch": 0.33,
+      "grad_norm": 0.050537109375,
+      "learning_rate": 1.567664512426336e-05,
+      "loss": 0.8041,
+      "step": 506
+    },
+    {
+      "epoch": 0.33,
+      "grad_norm": 0.0537109375,
+      "learning_rate": 1.5659286140911733e-05,
+      "loss": 0.8421,
+      "step": 507
+    },
+    {
+      "epoch": 0.33,
+      "grad_norm": 0.05859375,
+      "learning_rate": 1.5641902032567023e-05,
+      "loss": 0.9291,
+      "step": 508
+    },
+    {
+      "epoch": 0.33,
+      "grad_norm": 0.05712890625,
+      "learning_rate": 1.562449287640781e-05,
+      "loss": 0.8886,
+      "step": 509
+    },
+    {
+      "epoch": 0.33,
+      "grad_norm": 0.052001953125,
+      "learning_rate": 1.560705874972385e-05,
+      "loss": 0.8472,
+      "step": 510
+    },
+    {
+      "epoch": 0.33,
+      "grad_norm": 0.0517578125,
+      "learning_rate": 1.5589599729915783e-05,
+      "loss": 0.8108,
+      "step": 511
+    },
+    {
+      "epoch": 0.33,
+      "grad_norm": 0.049072265625,
+      "learning_rate": 1.5572115894494752e-05,
+      "loss": 0.868,
+      "step": 512
+    },
+    {
+      "epoch": 0.33,
+      "grad_norm": 0.04833984375,
+      "learning_rate": 1.5554607321082077e-05,
+      "loss": 0.8719,
+      "step": 513
+    },
+    {
+      "epoch": 0.33,
+      "grad_norm": 0.0517578125,
+      "learning_rate": 1.5537074087408894e-05,
+      "loss": 0.8289,
+      "step": 514
+    },
+    {
+      "epoch": 0.33,
+      "grad_norm": 0.05615234375,
+      "learning_rate": 1.5519516271315834e-05,
+      "loss": 0.8315,
+      "step": 515
+    },
+    {
+      "epoch": 0.34,
+      "grad_norm": 0.049072265625,
+      "learning_rate": 1.5501933950752655e-05,
+      "loss": 0.7429,
+      "step": 516
+    },
+    {
+      "epoch": 0.34,
+      "grad_norm": 0.0556640625,
+      "learning_rate": 1.5484327203777917e-05,
+      "loss": 0.7782,
+      "step": 517
+    },
+    {
+      "epoch": 0.34,
+      "grad_norm": 0.044189453125,
+      "learning_rate": 1.5466696108558614e-05,
+      "loss": 0.8454,
+      "step": 518
+    },
+    {
+      "epoch": 0.34,
+      "grad_norm": 0.050537109375,
+      "learning_rate": 1.544904074336983e-05,
+      "loss": 0.7808,
+      "step": 519
+    },
+    {
+      "epoch": 0.34,
+      "grad_norm": 0.04736328125,
+      "learning_rate": 1.5431361186594415e-05,
+      "loss": 0.7698,
+      "step": 520
+    },
+    {
+      "epoch": 0.34,
+      "grad_norm": 0.044921875,
+      "learning_rate": 1.5413657516722607e-05,
+      "loss": 0.8511,
+      "step": 521
+    },
+    {
+      "epoch": 0.34,
+      "grad_norm": 0.0498046875,
+      "learning_rate": 1.53959298123517e-05,
+      "loss": 0.8284,
+      "step": 522
+    },
+    {
+      "epoch": 0.34,
+      "grad_norm": 0.054443359375,
+      "learning_rate": 1.5378178152185703e-05,
+      "loss": 0.852,
+      "step": 523
+    },
+    {
+      "epoch": 0.34,
+      "grad_norm": 0.05126953125,
+      "learning_rate": 1.5360402615034958e-05,
+      "loss": 0.8806,
+      "step": 524
+    },
+    {
+      "epoch": 0.34,
+      "grad_norm": 0.05224609375,
+      "learning_rate": 1.5342603279815826e-05,
+      "loss": 0.778,
+      "step": 525
+    },
+    {
+      "epoch": 0.34,
+      "grad_norm": 0.05126953125,
+      "learning_rate": 1.5324780225550316e-05,
+      "loss": 0.8209,
+      "step": 526
+    },
+    {
+      "epoch": 0.34,
+      "grad_norm": 0.049072265625,
+      "learning_rate": 1.5306933531365748e-05,
+      "loss": 0.8506,
+      "step": 527
+    },
+    {
+      "epoch": 0.34,
+      "grad_norm": 0.049072265625,
+      "learning_rate": 1.5289063276494384e-05,
+      "loss": 0.8609,
+      "step": 528
+    },
+    {
+      "epoch": 0.34,
+      "grad_norm": 0.051025390625,
+      "learning_rate": 1.5271169540273093e-05,
+      "loss": 0.848,
+      "step": 529
+    },
+    {
+      "epoch": 0.34,
+      "grad_norm": 0.049560546875,
+      "learning_rate": 1.5253252402142989e-05,
+      "loss": 0.7898,
+      "step": 530
+    },
+    {
+      "epoch": 0.35,
+      "grad_norm": 0.049560546875,
+      "learning_rate": 1.5235311941649085e-05,
+      "loss": 0.845,
+      "step": 531
+    },
+    {
+      "epoch": 0.35,
+      "grad_norm": 0.052490234375,
+      "learning_rate": 1.5217348238439922e-05,
+      "loss": 0.8216,
+      "step": 532
+    },
+    {
+      "epoch": 0.35,
+      "grad_norm": 0.050048828125,
+      "learning_rate": 1.5199361372267252e-05,
+      "loss": 0.8517,
+      "step": 533
+    },
+    {
+      "epoch": 0.35,
+      "grad_norm": 0.05029296875,
+      "learning_rate": 1.5181351422985646e-05,
+      "loss": 0.8148,
+      "step": 534
+    },
+    {
+      "epoch": 0.35,
+      "grad_norm": 0.050537109375,
+      "learning_rate": 1.516331847055216e-05,
+      "loss": 0.8649,
+      "step": 535
+    },
+    {
+      "epoch": 0.35,
+      "grad_norm": 0.046142578125,
+      "learning_rate": 1.514526259502597e-05,
+      "loss": 0.8321,
+      "step": 536
+    },
+    {
+      "epoch": 0.35,
+      "grad_norm": 0.048583984375,
+      "learning_rate": 1.5127183876568024e-05,
+      "loss": 0.8175,
+      "step": 537
+    },
+    {
+      "epoch": 0.35,
+      "grad_norm": 0.04931640625,
+      "learning_rate": 1.5109082395440689e-05,
+      "loss": 0.7704,
+      "step": 538
+    },
+    {
+      "epoch": 0.35,
+      "grad_norm": 0.0595703125,
+      "learning_rate": 1.5090958232007383e-05,
+      "loss": 0.9095,
+      "step": 539
+    },
+    {
+      "epoch": 0.35,
+      "grad_norm": 0.049560546875,
+      "learning_rate": 1.507281146673223e-05,
+      "loss": 0.8498,
+      "step": 540
+    },
+    {
+      "epoch": 0.35,
+      "grad_norm": 0.04833984375,
+      "learning_rate": 1.5054642180179684e-05,
+      "loss": 0.8709,
+      "step": 541
+    },
+    {
+      "epoch": 0.35,
+      "grad_norm": 0.050537109375,
+      "learning_rate": 1.5036450453014202e-05,
+      "loss": 0.8439,
+      "step": 542
+    },
+    {
+      "epoch": 0.35,
+      "grad_norm": 0.048583984375,
+      "learning_rate": 1.5018236365999862e-05,
+      "loss": 0.9223,
+      "step": 543
+    },
+    {
+      "epoch": 0.35,
+      "grad_norm": 0.045654296875,
+      "learning_rate": 1.5000000000000002e-05,
+      "loss": 0.8336,
+      "step": 544
+    },
+    {
+      "epoch": 0.35,
+      "grad_norm": 0.049072265625,
+      "learning_rate": 1.4981741435976882e-05,
+      "loss": 0.8523,
+      "step": 545
+    },
+    {
+      "epoch": 0.36,
+      "grad_norm": 0.048583984375,
+      "learning_rate": 1.4963460754991309e-05,
+      "loss": 0.8496,
+      "step": 546
+    },
+    {
+      "epoch": 0.36,
+      "grad_norm": 0.0478515625,
+      "learning_rate": 1.4945158038202274e-05,
+      "loss": 0.8392,
+      "step": 547
+    },
+    {
+      "epoch": 0.36,
+      "grad_norm": 0.050048828125,
+      "learning_rate": 1.4926833366866611e-05,
+      "loss": 0.7421,
+      "step": 548
+    },
+    {
+      "epoch": 0.36,
+      "grad_norm": 0.052734375,
+      "learning_rate": 1.4908486822338611e-05,
+      "loss": 0.8284,
+      "step": 549
+    },
+    {
+      "epoch": 0.36,
+      "grad_norm": 0.048583984375,
+      "learning_rate": 1.489011848606968e-05,
+      "loss": 0.8383,
+      "step": 550
+    },
+    {
+      "epoch": 0.36,
+      "grad_norm": 0.0517578125,
+      "learning_rate": 1.4871728439607967e-05,
+      "loss": 0.8099,
+      "step": 551
+    },
+    {
+      "epoch": 0.36,
+      "grad_norm": 0.047607421875,
+      "learning_rate": 1.4853316764598011e-05,
+      "loss": 0.9332,
+      "step": 552
+    },
+    {
+      "epoch": 0.36,
+      "grad_norm": 0.046875,
+      "learning_rate": 1.4834883542780367e-05,
+      "loss": 0.8748,
+      "step": 553
+    },
+    {
+      "epoch": 0.36,
+      "grad_norm": 0.050048828125,
+      "learning_rate": 1.4816428855991257e-05,
+      "loss": 0.7789,
+      "step": 554
+    },
+    {
+      "epoch": 0.36,
+      "grad_norm": 0.051025390625,
+      "learning_rate": 1.4797952786162188e-05,
+      "loss": 0.8209,
+      "step": 555
+    },
+    {
+      "epoch": 0.36,
+      "grad_norm": 0.051025390625,
+      "learning_rate": 1.4779455415319612e-05,
+      "loss": 0.8381,
+      "step": 556
+    },
+    {
+      "epoch": 0.36,
+      "grad_norm": 0.05029296875,
+      "learning_rate": 1.4760936825584535e-05,
+      "loss": 0.7684,
+      "step": 557
+    },
+    {
+      "epoch": 0.36,
+      "grad_norm": 0.048828125,
+      "learning_rate": 1.4742397099172183e-05,
+      "loss": 0.8275,
+      "step": 558
+    },
+    {
+      "epoch": 0.36,
+      "grad_norm": 0.05224609375,
+      "learning_rate": 1.4723836318391607e-05,
+      "loss": 0.9173,
+      "step": 559
+    },
+    {
+      "epoch": 0.36,
+      "grad_norm": 0.05322265625,
+      "learning_rate": 1.4705254565645335e-05,
+      "loss": 0.7534,
+      "step": 560
+    },
+    {
+      "epoch": 0.36,
+      "grad_norm": 0.052001953125,
+      "learning_rate": 1.4686651923429002e-05,
+      "loss": 0.7822,
+      "step": 561
+    },
+    {
+      "epoch": 0.37,
+      "grad_norm": 0.04931640625,
+      "learning_rate": 1.4668028474330989e-05,
+      "loss": 0.8575,
+      "step": 562
+    },
+    {
+      "epoch": 0.37,
+      "grad_norm": 0.052978515625,
+      "learning_rate": 1.4649384301032044e-05,
+      "loss": 0.8385,
+      "step": 563
+    },
+    {
+      "epoch": 0.37,
+      "grad_norm": 0.052490234375,
+      "learning_rate": 1.4630719486304928e-05,
+      "loss": 0.8523,
+      "step": 564
+    },
+    {
+      "epoch": 0.37,
+      "grad_norm": 0.050537109375,
+      "learning_rate": 1.4612034113014036e-05,
+      "loss": 0.8569,
+      "step": 565
+    },
+    {
+      "epoch": 0.37,
+      "grad_norm": 0.051025390625,
+      "learning_rate": 1.4593328264115044e-05,
+      "loss": 0.7559,
+      "step": 566
+    },
+    {
+      "epoch": 0.37,
+      "grad_norm": 0.052490234375,
+      "learning_rate": 1.4574602022654516e-05,
+      "loss": 0.822,
+      "step": 567
+    },
+    {
+      "epoch": 0.37,
+      "grad_norm": 0.04833984375,
+      "learning_rate": 1.4555855471769572e-05,
+      "loss": 0.7773,
+      "step": 568
+    },
+    {
+      "epoch": 0.37,
+      "grad_norm": 0.047607421875,
+      "learning_rate": 1.4537088694687476e-05,
+      "loss": 0.8264,
+      "step": 569
+    },
+    {
+      "epoch": 0.37,
+      "grad_norm": 0.04833984375,
+      "learning_rate": 1.4518301774725308e-05,
+      "loss": 0.8323,
+      "step": 570
+    },
+    {
+      "epoch": 0.37,
+      "grad_norm": 0.0576171875,
+      "learning_rate": 1.4499494795289562e-05,
+      "loss": 0.8173,
+      "step": 571
+    },
+    {
+      "epoch": 0.37,
+      "grad_norm": 0.050048828125,
+      "learning_rate": 1.4480667839875786e-05,
+      "loss": 0.8007,
+      "step": 572
+    },
+    {
+      "epoch": 0.37,
+      "grad_norm": 0.046630859375,
+      "learning_rate": 1.4461820992068224e-05,
+      "loss": 0.8152,
+      "step": 573
+    },
+    {
+      "epoch": 0.37,
+      "grad_norm": 0.049560546875,
+      "learning_rate": 1.4442954335539432e-05,
+      "loss": 0.8258,
+      "step": 574
+    },
+    {
+      "epoch": 0.37,
+      "grad_norm": 0.046875,
+      "learning_rate": 1.4424067954049903e-05,
+      "loss": 0.8003,
+      "step": 575
+    },
+    {
+      "epoch": 0.37,
+      "grad_norm": 0.046875,
+      "learning_rate": 1.4405161931447702e-05,
+      "loss": 0.7678,
+      "step": 576
+    },
+    {
+      "epoch": 0.38,
+      "grad_norm": 0.049072265625,
+      "learning_rate": 1.4386236351668095e-05,
+      "loss": 0.8295,
+      "step": 577
+    },
+    {
+      "epoch": 0.38,
+      "grad_norm": 0.06298828125,
+      "learning_rate": 1.436729129873318e-05,
+      "loss": 0.8286,
+      "step": 578
+    },
+    {
+      "epoch": 0.38,
+      "grad_norm": 0.05322265625,
+      "learning_rate": 1.4348326856751496e-05,
+      "loss": 0.827,
+      "step": 579
+    },
+    {
+      "epoch": 0.38,
+      "grad_norm": 0.046142578125,
+      "learning_rate": 1.4329343109917671e-05,
+      "loss": 0.7671,
+      "step": 580
+    },
+    {
+      "epoch": 0.38,
+      "grad_norm": 0.050537109375,
+      "learning_rate": 1.431034014251203e-05,
+      "loss": 0.8195,
+      "step": 581
+    },
+    {
+      "epoch": 0.38,
+      "grad_norm": 0.05029296875,
+      "learning_rate": 1.4291318038900243e-05,
+      "loss": 0.8331,
+      "step": 582
+    },
+    {
+      "epoch": 0.38,
+      "grad_norm": 0.050537109375,
+      "learning_rate": 1.4272276883532927e-05,
+      "loss": 0.8458,
+      "step": 583
+    },
+    {
+      "epoch": 0.38,
+      "grad_norm": 0.05712890625,
+      "learning_rate": 1.4253216760945284e-05,
+      "loss": 0.8311,
+      "step": 584
+    },
+    {
+      "epoch": 0.38,
+      "grad_norm": 0.054931640625,
+      "learning_rate": 1.423413775575672e-05,
+      "loss": 0.8339,
+      "step": 585
+    },
+    {
+      "epoch": 0.38,
+      "grad_norm": 0.05029296875,
+      "learning_rate": 1.4215039952670482e-05,
+      "loss": 0.811,
+      "step": 586
+    },
+    {
+      "epoch": 0.38,
+      "grad_norm": 0.060546875,
+      "learning_rate": 1.4195923436473257e-05,
+      "loss": 0.8431,
+      "step": 587
+    },
+    {
+      "epoch": 0.38,
+      "grad_norm": 0.04833984375,
+      "learning_rate": 1.4176788292034824e-05,
+      "loss": 0.7789,
+      "step": 588
+    },
+    {
+      "epoch": 0.38,
+      "grad_norm": 0.052734375,
+      "learning_rate": 1.4157634604307661e-05,
+      "loss": 0.8773,
+      "step": 589
+    },
+    {
+      "epoch": 0.38,
+      "grad_norm": 0.05126953125,
+      "learning_rate": 1.413846245832657e-05,
+      "loss": 0.8304,
+      "step": 590
+    },
+    {
+      "epoch": 0.38,
+      "grad_norm": 0.052001953125,
+      "learning_rate": 1.411927193920829e-05,
+      "loss": 0.8578,
+      "step": 591
+    },
+    {
+      "epoch": 0.38,
+      "grad_norm": 0.05029296875,
+      "learning_rate": 1.4100063132151148e-05,
+      "loss": 0.7993,
+      "step": 592
+    },
+    {
+      "epoch": 0.39,
+      "grad_norm": 0.052001953125,
+      "learning_rate": 1.408083612243465e-05,
+      "loss": 0.8845,
+      "step": 593
+    },
+    {
+      "epoch": 0.39,
+      "grad_norm": 0.0517578125,
+      "learning_rate": 1.4061590995419118e-05,
+      "loss": 0.8415,
+      "step": 594
+    },
+    {
+      "epoch": 0.39,
+      "grad_norm": 0.0517578125,
+      "learning_rate": 1.404232783654531e-05,
+      "loss": 0.7521,
+      "step": 595
+    },
+    {
+      "epoch": 0.39,
+      "grad_norm": 0.04931640625,
+      "learning_rate": 1.402304673133403e-05,
+      "loss": 0.8167,
+      "step": 596
+    },
+    {
+      "epoch": 0.39,
+      "grad_norm": 0.050048828125,
+      "learning_rate": 1.4003747765385767e-05,
+      "loss": 0.7807,
+      "step": 597
+    },
+    {
+      "epoch": 0.39,
+      "grad_norm": 0.0498046875,
+      "learning_rate": 1.3984431024380301e-05,
+      "loss": 0.8049,
+      "step": 598
+    },
+    {
+      "epoch": 0.39,
+      "grad_norm": 0.053466796875,
+      "learning_rate": 1.3965096594076322e-05,
+      "loss": 0.8303,
+      "step": 599
+    },
+    {
+      "epoch": 0.39,
+      "grad_norm": 0.050048828125,
+      "learning_rate": 1.3945744560311056e-05,
+      "loss": 0.8684,
+      "step": 600
+    },
+    {
+      "epoch": 0.39,
+      "grad_norm": 0.053466796875,
+      "learning_rate": 1.3926375008999887e-05,
+      "loss": 0.7436,
+      "step": 601
+    },
+    {
+      "epoch": 0.39,
+      "grad_norm": 0.05029296875,
+      "learning_rate": 1.3906988026135957e-05,
+      "loss": 0.8579,
+      "step": 602
+    },
+    {
+      "epoch": 0.39,
+      "grad_norm": 0.04736328125,
+      "learning_rate": 1.3887583697789815e-05,
+      "loss": 0.7671,
+      "step": 603
+    },
+    {
+      "epoch": 0.39,
+      "grad_norm": 0.049072265625,
+      "learning_rate": 1.3868162110109001e-05,
+      "loss": 0.8025,
+      "step": 604
+    },
+    {
+      "epoch": 0.39,
+      "grad_norm": 0.054443359375,
+      "learning_rate": 1.3848723349317688e-05,
+      "loss": 0.8687,
+      "step": 605
+    },
+    {
+      "epoch": 0.39,
+      "grad_norm": 0.053955078125,
+      "learning_rate": 1.382926750171629e-05,
+      "loss": 0.8175,
+      "step": 606
+    },
+    {
+      "epoch": 0.39,
+      "grad_norm": 0.05078125,
+      "learning_rate": 1.3809794653681075e-05,
+      "loss": 0.8536,
+      "step": 607
+    },
+    {
+      "epoch": 0.4,
+      "grad_norm": 0.0498046875,
+      "learning_rate": 1.3790304891663793e-05,
+      "loss": 0.8289,
+      "step": 608
+    },
+    {
+      "epoch": 0.4,
+      "grad_norm": 0.051513671875,
+      "learning_rate": 1.3770798302191279e-05,
+      "loss": 0.7827,
+      "step": 609
+    },
+    {
+      "epoch": 0.4,
+      "grad_norm": 0.050537109375,
+      "learning_rate": 1.3751274971865086e-05,
+      "loss": 0.8537,
+      "step": 610
+    },
+    {
+      "epoch": 0.4,
+      "grad_norm": 0.048095703125,
+      "learning_rate": 1.3731734987361069e-05,
+      "loss": 0.8194,
+      "step": 611
+    },
+    {
+      "epoch": 0.4,
+      "grad_norm": 0.048095703125,
+      "learning_rate": 1.3712178435429044e-05,
+      "loss": 0.7846,
+      "step": 612
+    },
+    {
+      "epoch": 0.4,
+      "grad_norm": 0.0478515625,
+      "learning_rate": 1.3692605402892369e-05,
+      "loss": 0.8813,
+      "step": 613
+    },
+    {
+      "epoch": 0.4,
+      "grad_norm": 0.0478515625,
+      "learning_rate": 1.367301597664757e-05,
+      "loss": 0.7698,
+      "step": 614
+    },
+    {
+      "epoch": 0.4,
+      "grad_norm": 0.051513671875,
+      "learning_rate": 1.3653410243663953e-05,
+      "loss": 0.7923,
+      "step": 615
+    },
+    {
+      "epoch": 0.4,
+      "grad_norm": 0.0478515625,
+      "learning_rate": 1.3633788290983221e-05,
+      "loss": 0.7843,
+      "step": 616
+    },
+    {
+      "epoch": 0.4,
+      "grad_norm": 0.051025390625,
+      "learning_rate": 1.3614150205719086e-05,
+      "loss": 0.8806,
+      "step": 617
+    },
+    {
+      "epoch": 0.4,
+      "grad_norm": 0.046142578125,
+      "learning_rate": 1.3594496075056886e-05,
+      "loss": 0.7934,
+      "step": 618
+    },
+    {
+      "epoch": 0.4,
+      "grad_norm": 0.049072265625,
+      "learning_rate": 1.3574825986253191e-05,
+      "loss": 0.8547,
+      "step": 619
+    },
+    {
+      "epoch": 0.4,
+      "grad_norm": 0.04833984375,
+      "learning_rate": 1.3555140026635415e-05,
+      "loss": 0.8556,
+      "step": 620
+    },
+    {
+      "epoch": 0.4,
+      "grad_norm": 0.04931640625,
+      "learning_rate": 1.3535438283601437e-05,
+      "loss": 0.8136,
+      "step": 621
+    },
+    {
+      "epoch": 0.4,
+      "grad_norm": 0.051025390625,
+      "learning_rate": 1.3515720844619206e-05,
+      "loss": 0.7409,
+      "step": 622
+    },
+    {
+      "epoch": 0.41,
+      "grad_norm": 0.046875,
+      "learning_rate": 1.3495987797226362e-05,
+      "loss": 0.7693,
+      "step": 623
+    },
+    {
+      "epoch": 0.41,
+      "grad_norm": 0.05322265625,
+      "learning_rate": 1.3476239229029826e-05,
+      "loss": 0.8478,
+      "step": 624
+    },
+    {
+      "epoch": 0.41,
+      "grad_norm": 0.05859375,
+      "learning_rate": 1.3456475227705442e-05,
+      "loss": 0.8639,
+      "step": 625
+    },
+    {
+      "epoch": 0.41,
+      "grad_norm": 0.059326171875,
+      "learning_rate": 1.3436695880997551e-05,
+      "loss": 0.8093,
+      "step": 626
+    },
+    {
+      "epoch": 0.41,
+      "grad_norm": 0.050048828125,
+      "learning_rate": 1.3416901276718643e-05,
+      "loss": 0.7741,
+      "step": 627
+    },
+    {
+      "epoch": 0.41,
+      "grad_norm": 0.0615234375,
+      "learning_rate": 1.339709150274893e-05,
+      "loss": 0.8319,
+      "step": 628
+    },
+    {
+      "epoch": 0.41,
+      "grad_norm": 0.052490234375,
+      "learning_rate": 1.3377266647035977e-05,
+      "loss": 0.8408,
+      "step": 629
+    },
+    {
+      "epoch": 0.41,
+      "grad_norm": 0.052734375,
+      "learning_rate": 1.3357426797594309e-05,
+      "loss": 0.7784,
+      "step": 630
+    },
+    {
+      "epoch": 0.41,
+      "grad_norm": 0.05712890625,
+      "learning_rate": 1.3337572042505007e-05,
+      "loss": 0.7569,
+      "step": 631
+    },
+    {
+      "epoch": 0.41,
+      "grad_norm": 0.052490234375,
+      "learning_rate": 1.331770246991534e-05,
+      "loss": 0.8598,
+      "step": 632
+    },
+    {
+      "epoch": 0.41,
+      "grad_norm": 0.053466796875,
+      "learning_rate": 1.3297818168038353e-05,
+      "loss": 0.8132,
+      "step": 633
+    },
+    {
+      "epoch": 0.41,
+      "grad_norm": 0.05029296875,
+      "learning_rate": 1.3277919225152486e-05,
+      "loss": 0.8481,
+      "step": 634
+    },
+    {
+      "epoch": 0.41,
+      "grad_norm": 0.049560546875,
+      "learning_rate": 1.3258005729601178e-05,
+      "loss": 0.8206,
+      "step": 635
+    },
+    {
+      "epoch": 0.41,
+      "grad_norm": 0.0546875,
+      "learning_rate": 1.3238077769792475e-05,
+      "loss": 0.8263,
+      "step": 636
+    },
+    {
+      "epoch": 0.41,
+      "grad_norm": 0.049560546875,
+      "learning_rate": 1.321813543419864e-05,
+      "loss": 0.8544,
+      "step": 637
+    },
+    {
+      "epoch": 0.41,
+      "grad_norm": 0.061279296875,
+      "learning_rate": 1.3198178811355762e-05,
+      "loss": 0.8752,
+      "step": 638
+    },
+    {
+      "epoch": 0.42,
+      "grad_norm": 0.054931640625,
+      "learning_rate": 1.3178207989863356e-05,
+      "loss": 0.8917,
+      "step": 639
+    },
+    {
+      "epoch": 0.42,
+      "grad_norm": 0.053466796875,
+      "learning_rate": 1.3158223058383972e-05,
+      "loss": 0.8589,
+      "step": 640
+    },
+    {
+      "epoch": 0.42,
+      "grad_norm": 0.053955078125,
+      "learning_rate": 1.3138224105642803e-05,
+      "loss": 0.7553,
+      "step": 641
+    },
+    {
+      "epoch": 0.42,
+      "grad_norm": 0.0625,
+      "learning_rate": 1.31182112204273e-05,
+      "loss": 0.8247,
+      "step": 642
+    },
+    {
+      "epoch": 0.42,
+      "grad_norm": 0.058349609375,
+      "learning_rate": 1.3098184491586752e-05,
+      "loss": 0.8437,
+      "step": 643
+    },
+    {
+      "epoch": 0.42,
+      "grad_norm": 0.0556640625,
+      "learning_rate": 1.3078144008031924e-05,
+      "loss": 0.8129,
+      "step": 644
+    },
+    {
+      "epoch": 0.42,
+      "grad_norm": 0.050537109375,
+      "learning_rate": 1.3058089858734637e-05,
+      "loss": 0.7883,
+      "step": 645
+    },
+    {
+      "epoch": 0.42,
+      "grad_norm": 0.051025390625,
+      "learning_rate": 1.3038022132727388e-05,
+      "loss": 0.889,
+      "step": 646
+    },
+    {
+      "epoch": 0.42,
+      "grad_norm": 0.052490234375,
+      "learning_rate": 1.3017940919102943e-05,
+      "loss": 0.8434,
+      "step": 647
+    },
+    {
+      "epoch": 0.42,
+      "grad_norm": 0.05859375,
+      "learning_rate": 1.2997846307013955e-05,
+      "loss": 0.8118,
+      "step": 648
+    },
+    {
+      "epoch": 0.42,
+      "grad_norm": 0.05322265625,
+      "learning_rate": 1.2977738385672558e-05,
+      "loss": 0.8348,
+      "step": 649
+    },
+    {
+      "epoch": 0.42,
+      "grad_norm": 0.05322265625,
+      "learning_rate": 1.295761724434997e-05,
+      "loss": 0.8419,
+      "step": 650
+    },
+    {
+      "epoch": 0.42,
+      "grad_norm": 0.057373046875,
+      "learning_rate": 1.2937482972376104e-05,
+      "loss": 0.8148,
+      "step": 651
+    },
+    {
+      "epoch": 0.42,
+      "grad_norm": 0.050048828125,
+      "learning_rate": 1.2917335659139166e-05,
+      "loss": 0.8083,
+      "step": 652
+    },
+    {
+      "epoch": 0.42,
+      "grad_norm": 0.05029296875,
+      "learning_rate": 1.2897175394085266e-05,
+      "loss": 0.7939,
+      "step": 653
+    },
+    {
+      "epoch": 0.43,
+      "grad_norm": 0.053955078125,
+      "learning_rate": 1.2877002266718011e-05,
+      "loss": 0.8044,
+      "step": 654
+    },
+    {
+      "epoch": 0.43,
+      "grad_norm": 0.0517578125,
+      "learning_rate": 1.2856816366598103e-05,
+      "loss": 0.809,
+      "step": 655
+    },
+    {
+      "epoch": 0.43,
+      "grad_norm": 0.048095703125,
+      "learning_rate": 1.2836617783342968e-05,
+      "loss": 0.88,
+      "step": 656
+    },
+    {
+      "epoch": 0.43,
+      "grad_norm": 0.050048828125,
+      "learning_rate": 1.2816406606626324e-05,
+      "loss": 0.7649,
+      "step": 657
+    },
+    {
+      "epoch": 0.43,
+      "grad_norm": 0.0517578125,
+      "learning_rate": 1.2796182926177809e-05,
+      "loss": 0.8551,
+      "step": 658
+    },
+    {
+      "epoch": 0.43,
+      "grad_norm": 0.05126953125,
+      "learning_rate": 1.2775946831782565e-05,
+      "loss": 0.8506,
+      "step": 659
+    },
+    {
+      "epoch": 0.43,
+      "grad_norm": 0.048095703125,
+      "learning_rate": 1.2755698413280853e-05,
+      "loss": 0.8209,
+      "step": 660
+    },
+    {
+      "epoch": 0.43,
+      "grad_norm": 0.052490234375,
+      "learning_rate": 1.2735437760567644e-05,
+      "loss": 0.847,
+      "step": 661
+    },
+    {
+      "epoch": 0.43,
+      "grad_norm": 0.048828125,
+      "learning_rate": 1.2715164963592228e-05,
+      "loss": 0.7587,
+      "step": 662
+    },
+    {
+      "epoch": 0.43,
+      "grad_norm": 0.05224609375,
+      "learning_rate": 1.2694880112357809e-05,
+      "loss": 0.7927,
+      "step": 663
+    },
+    {
+      "epoch": 0.43,
+      "grad_norm": 0.0478515625,
+      "learning_rate": 1.2674583296921109e-05,
+      "loss": 0.7415,
+      "step": 664
+    },
+    {
+      "epoch": 0.43,
+      "grad_norm": 0.048095703125,
+      "learning_rate": 1.2654274607391959e-05,
+      "loss": 0.7919,
+      "step": 665
+    },
+    {
+      "epoch": 0.43,
+      "grad_norm": 0.04931640625,
+      "learning_rate": 1.2633954133932913e-05,
+      "loss": 0.8203,
+      "step": 666
+    },
+    {
+      "epoch": 0.43,
+      "grad_norm": 0.0478515625,
+      "learning_rate": 1.2613621966758838e-05,
+      "loss": 0.8021,
+      "step": 667
+    },
+    {
+      "epoch": 0.43,
+      "grad_norm": 0.051513671875,
+      "learning_rate": 1.2593278196136525e-05,
+      "loss": 0.7899,
+      "step": 668
+    },
+    {
+      "epoch": 0.43,
+      "grad_norm": 0.056884765625,
+      "learning_rate": 1.257292291238427e-05,
+      "loss": 0.8331,
+      "step": 669
+    },
+    {
+      "epoch": 0.44,
+      "grad_norm": 0.05029296875,
+      "learning_rate": 1.255255620587148e-05,
+      "loss": 0.7661,
+      "step": 670
+    },
+    {
+      "epoch": 0.44,
+      "grad_norm": 0.056396484375,
+      "learning_rate": 1.2532178167018283e-05,
+      "loss": 0.8314,
+      "step": 671
+    },
+    {
+      "epoch": 0.44,
+      "grad_norm": 0.051513671875,
+      "learning_rate": 1.2511788886295115e-05,
+      "loss": 0.7618,
+      "step": 672
+    },
+    {
+      "epoch": 0.44,
+      "grad_norm": 0.0556640625,
+      "learning_rate": 1.2491388454222327e-05,
+      "loss": 0.8003,
+      "step": 673
+    },
+    {
+      "epoch": 0.44,
+      "grad_norm": 0.051025390625,
+      "learning_rate": 1.2470976961369765e-05,
+      "loss": 0.8458,
+      "step": 674
+    },
+    {
+      "epoch": 0.44,
+      "grad_norm": 0.0458984375,
+      "learning_rate": 1.2450554498356388e-05,
+      "loss": 0.7778,
+      "step": 675
+    },
+    {
+      "epoch": 0.44,
+      "grad_norm": 0.048095703125,
+      "learning_rate": 1.243012115584986e-05,
+      "loss": 0.8396,
+      "step": 676
+    },
+    {
+      "epoch": 0.44,
+      "grad_norm": 0.048828125,
+      "learning_rate": 1.2409677024566145e-05,
+      "loss": 0.8113,
+      "step": 677
+    },
+    {
+      "epoch": 0.44,
+      "grad_norm": 0.049072265625,
+      "learning_rate": 1.2389222195269102e-05,
+      "loss": 0.8426,
+      "step": 678
+    },
+    {
+      "epoch": 0.44,
+      "grad_norm": 0.052978515625,
+      "learning_rate": 1.2368756758770084e-05,
+      "loss": 0.8192,
+      "step": 679
+    },
+    {
+      "epoch": 0.44,
+      "grad_norm": 0.054443359375,
+      "learning_rate": 1.2348280805927541e-05,
+      "loss": 0.883,
+      "step": 680
+    },
+    {
+      "epoch": 0.44,
+      "grad_norm": 0.061279296875,
+      "learning_rate": 1.2327794427646607e-05,
+      "loss": 0.7733,
+      "step": 681
+    },
+    {
+      "epoch": 0.44,
+      "grad_norm": 0.06201171875,
+      "learning_rate": 1.2307297714878706e-05,
+      "loss": 0.8566,
+      "step": 682
+    },
+    {
+      "epoch": 0.44,
+      "grad_norm": 0.051513671875,
+      "learning_rate": 1.2286790758621132e-05,
+      "loss": 0.86,
+      "step": 683
+    },
+    {
+      "epoch": 0.44,
+      "grad_norm": 0.05615234375,
+      "learning_rate": 1.2266273649916669e-05,
+      "loss": 0.8344,
+      "step": 684
+    },
+    {
+      "epoch": 0.45,
+      "grad_norm": 0.07373046875,
+      "learning_rate": 1.2245746479853168e-05,
+      "loss": 0.9192,
+      "step": 685
+    },
+    {
+      "epoch": 0.45,
+      "grad_norm": 0.0537109375,
+      "learning_rate": 1.2225209339563144e-05,
+      "loss": 0.7797,
+      "step": 686
+    },
+    {
+      "epoch": 0.45,
+      "grad_norm": 0.051025390625,
+      "learning_rate": 1.2204662320223385e-05,
+      "loss": 0.7331,
+      "step": 687
+    },
+    {
+      "epoch": 0.45,
+      "grad_norm": 0.046875,
+      "learning_rate": 1.2184105513054539e-05,
+      "loss": 0.8126,
+      "step": 688
+    },
+    {
+      "epoch": 0.45,
+      "grad_norm": 0.05126953125,
+      "learning_rate": 1.2163539009320691e-05,
+      "loss": 0.8052,
+      "step": 689
+    },
+    {
+      "epoch": 0.45,
+      "grad_norm": 0.051025390625,
+      "learning_rate": 1.2142962900328994e-05,
+      "loss": 0.7866,
+      "step": 690
+    },
+    {
+      "epoch": 0.45,
+      "grad_norm": 0.052001953125,
+      "learning_rate": 1.2122377277429231e-05,
+      "loss": 0.8124,
+      "step": 691
+    },
+    {
+      "epoch": 0.45,
+      "grad_norm": 0.052734375,
+      "learning_rate": 1.2101782232013436e-05,
+      "loss": 0.8899,
+      "step": 692
+    },
+    {
+      "epoch": 0.45,
+      "grad_norm": 0.053955078125,
+      "learning_rate": 1.208117785551547e-05,
+      "loss": 0.7966,
+      "step": 693
+    },
+    {
+      "epoch": 0.45,
+      "grad_norm": 0.0546875,
+      "learning_rate": 1.2060564239410613e-05,
+      "loss": 0.8062,
+      "step": 694
+    },
+    {
+      "epoch": 0.45,
+      "grad_norm": 0.0634765625,
+      "learning_rate": 1.2039941475215169e-05,
+      "loss": 0.8438,
+      "step": 695
+    },
+    {
+      "epoch": 0.45,
+      "grad_norm": 0.05078125,
+      "learning_rate": 1.2019309654486065e-05,
+      "loss": 0.8648,
+      "step": 696
+    },
+    {
+      "epoch": 0.45,
+      "grad_norm": 0.0546875,
+      "learning_rate": 1.1998668868820422e-05,
+      "loss": 0.8274,
+      "step": 697
+    },
+    {
+      "epoch": 0.45,
+      "grad_norm": 0.05712890625,
+      "learning_rate": 1.1978019209855174e-05,
+      "loss": 0.8074,
+      "step": 698
+    },
+    {
+      "epoch": 0.45,
+      "grad_norm": 0.052490234375,
+      "learning_rate": 1.195736076926664e-05,
+      "loss": 0.8436,
+      "step": 699
+    },
+    {
+      "epoch": 0.46,
+      "grad_norm": 0.057861328125,
+      "learning_rate": 1.1936693638770127e-05,
+      "loss": 0.7833,
+      "step": 700
+    },
+    {
+      "epoch": 0.46,
+      "grad_norm": 0.054443359375,
+      "learning_rate": 1.1916017910119525e-05,
+      "loss": 0.8146,
+      "step": 701
+    },
+    {
+      "epoch": 0.46,
+      "grad_norm": 0.05419921875,
+      "learning_rate": 1.1895333675106897e-05,
+      "loss": 0.848,
+      "step": 702
+    },
+    {
+      "epoch": 0.46,
+      "grad_norm": 0.0537109375,
+      "learning_rate": 1.1874641025562065e-05,
+      "loss": 0.8167,
+      "step": 703
+    },
+    {
+      "epoch": 0.46,
+      "grad_norm": 0.05517578125,
+      "learning_rate": 1.185394005335222e-05,
+      "loss": 0.8499,
+      "step": 704
+    },
+    {
+      "epoch": 0.46,
+      "grad_norm": 0.0556640625,
+      "learning_rate": 1.1833230850381488e-05,
+      "loss": 0.7853,
+      "step": 705
+    },
+    {
+      "epoch": 0.46,
+      "grad_norm": 0.055908203125,
+      "learning_rate": 1.1812513508590541e-05,
+      "loss": 0.8321,
+      "step": 706
+    },
+    {
+      "epoch": 0.46,
+      "grad_norm": 0.05712890625,
+      "learning_rate": 1.1791788119956191e-05,
+      "loss": 0.7874,
+      "step": 707
+    },
+    {
+      "epoch": 0.46,
+      "grad_norm": 0.055419921875,
+      "learning_rate": 1.1771054776490968e-05,
+      "loss": 0.8356,
+      "step": 708
+    },
+    {
+      "epoch": 0.46,
+      "grad_norm": 0.053466796875,
+      "learning_rate": 1.1750313570242721e-05,
+      "loss": 0.8236,
+      "step": 709
+    },
+    {
+      "epoch": 0.46,
+      "grad_norm": 0.06103515625,
+      "learning_rate": 1.1729564593294203e-05,
+      "loss": 0.8301,
+      "step": 710
+    },
+    {
+      "epoch": 0.46,
+      "grad_norm": 0.048583984375,
+      "learning_rate": 1.170880793776267e-05,
+      "loss": 0.7926,
+      "step": 711
+    },
+    {
+      "epoch": 0.46,
+      "grad_norm": 0.055908203125,
+      "learning_rate": 1.168804369579947e-05,
+      "loss": 0.7654,
+      "step": 712
+    },
+    {
+      "epoch": 0.46,
+      "grad_norm": 0.050537109375,
+      "learning_rate": 1.1667271959589623e-05,
+      "loss": 0.7618,
+      "step": 713
+    },
+    {
+      "epoch": 0.46,
+      "grad_norm": 0.05126953125,
+      "learning_rate": 1.1646492821351428e-05,
+      "loss": 0.8027,
+      "step": 714
+    },
+    {
+      "epoch": 0.46,
+      "grad_norm": 0.0537109375,
+      "learning_rate": 1.1625706373336046e-05,
+      "loss": 0.8891,
+      "step": 715
+    },
+    {
+      "epoch": 0.47,
+      "grad_norm": 0.051513671875,
+      "learning_rate": 1.1604912707827083e-05,
+      "loss": 0.85,
+      "step": 716
+    },
+    {
+      "epoch": 0.47,
+      "grad_norm": 0.05419921875,
+      "learning_rate": 1.1584111917140197e-05,
+      "loss": 0.8244,
+      "step": 717
+    },
+    {
+      "epoch": 0.47,
+      "grad_norm": 0.0537109375,
+      "learning_rate": 1.1563304093622674e-05,
+      "loss": 0.7958,
+      "step": 718
+    },
+    {
+      "epoch": 0.47,
+      "grad_norm": 0.05126953125,
+      "learning_rate": 1.1542489329653024e-05,
+      "loss": 0.8329,
+      "step": 719
+    },
+    {
+      "epoch": 0.47,
+      "grad_norm": 0.055908203125,
+      "learning_rate": 1.1521667717640572e-05,
+      "loss": 0.7596,
+      "step": 720
+    },
+    {
+      "epoch": 0.47,
+      "grad_norm": 0.05712890625,
+      "learning_rate": 1.1500839350025039e-05,
+      "loss": 0.9115,
+      "step": 721
+    },
+    {
+      "epoch": 0.47,
+      "grad_norm": 0.0556640625,
+      "learning_rate": 1.1480004319276145e-05,
+      "loss": 0.8006,
+      "step": 722
+    },
+    {
+      "epoch": 0.47,
+      "grad_norm": 0.05615234375,
+      "learning_rate": 1.1459162717893193e-05,
+      "loss": 0.7918,
+      "step": 723
+    },
+    {
+      "epoch": 0.47,
+      "grad_norm": 0.054443359375,
+      "learning_rate": 1.143831463840465e-05,
+      "loss": 0.7406,
+      "step": 724
+    },
+    {
+      "epoch": 0.47,
+      "grad_norm": 0.05322265625,
+      "learning_rate": 1.1417460173367748e-05,
+      "loss": 0.7735,
+      "step": 725
+    },
+    {
+      "epoch": 0.47,
+      "grad_norm": 0.050048828125,
+      "learning_rate": 1.1396599415368062e-05,
+      "loss": 0.7926,
+      "step": 726
+    },
+    {
+      "epoch": 0.47,
+      "grad_norm": 0.05126953125,
+      "learning_rate": 1.1375732457019118e-05,
+      "loss": 0.8132,
+      "step": 727
+    },
+    {
+      "epoch": 0.47,
+      "grad_norm": 0.05224609375,
+      "learning_rate": 1.1354859390961958e-05,
+      "loss": 0.7791,
+      "step": 728
+    },
+    {
+      "epoch": 0.47,
+      "grad_norm": 0.052001953125,
+      "learning_rate": 1.1333980309864743e-05,
+      "loss": 0.8277,
+      "step": 729
+    },
+    {
+      "epoch": 0.47,
+      "grad_norm": 0.055908203125,
+      "learning_rate": 1.1313095306422336e-05,
+      "loss": 0.8324,
+      "step": 730
+    },
+    {
+      "epoch": 0.48,
+      "grad_norm": 0.050048828125,
+      "learning_rate": 1.1292204473355897e-05,
+      "loss": 0.8637,
+      "step": 731
+    },
+    {
+      "epoch": 0.48,
+      "grad_norm": 0.050048828125,
+      "learning_rate": 1.1271307903412469e-05,
+      "loss": 0.8237,
+      "step": 732
+    },
+    {
+      "epoch": 0.48,
+      "grad_norm": 0.054931640625,
+      "learning_rate": 1.1250405689364561e-05,
+      "loss": 0.8395,
+      "step": 733
+    },
+    {
+      "epoch": 0.48,
+      "grad_norm": 0.0517578125,
+      "learning_rate": 1.1229497924009731e-05,
+      "loss": 0.7743,
+      "step": 734
+    },
+    {
+      "epoch": 0.48,
+      "grad_norm": 0.1611328125,
+      "learning_rate": 1.1208584700170203e-05,
+      "loss": 0.7385,
+      "step": 735
+    },
+    {
+      "epoch": 0.48,
+      "grad_norm": 0.058837890625,
+      "learning_rate": 1.1187666110692417e-05,
+      "loss": 0.8611,
+      "step": 736
+    },
+    {
+      "epoch": 0.48,
+      "grad_norm": 0.06103515625,
+      "learning_rate": 1.116674224844664e-05,
+      "loss": 0.7648,
+      "step": 737
+    },
+    {
+      "epoch": 0.48,
+      "grad_norm": 0.057373046875,
+      "learning_rate": 1.1145813206326548e-05,
+      "loss": 0.8365,
+      "step": 738
+    },
+    {
+      "epoch": 0.48,
+      "grad_norm": 0.054931640625,
+      "learning_rate": 1.1124879077248815e-05,
+      "loss": 0.7706,
+      "step": 739
+    },
+    {
+      "epoch": 0.48,
+      "grad_norm": 0.0498046875,
+      "learning_rate": 1.11039399541527e-05,
+      "loss": 0.8028,
+      "step": 740
+    },
+    {
+      "epoch": 0.48,
+      "grad_norm": 0.0498046875,
+      "learning_rate": 1.1082995929999626e-05,
+      "loss": 0.7975,
+      "step": 741
+    },
+    {
+      "epoch": 0.48,
+      "grad_norm": 0.049072265625,
+      "learning_rate": 1.1062047097772783e-05,
+      "loss": 0.8077,
+      "step": 742
+    },
+    {
+      "epoch": 0.48,
+      "grad_norm": 0.053955078125,
+      "learning_rate": 1.1041093550476706e-05,
+      "loss": 0.794,
+      "step": 743
+    },
+    {
+      "epoch": 0.48,
+      "grad_norm": 0.056640625,
+      "learning_rate": 1.1020135381136858e-05,
+      "loss": 0.825,
+      "step": 744
+    },
+    {
+      "epoch": 0.48,
+      "grad_norm": 0.05224609375,
+      "learning_rate": 1.0999172682799227e-05,
+      "loss": 0.8103,
+      "step": 745
+    },
+    {
+      "epoch": 0.49,
+      "grad_norm": 0.05322265625,
+      "learning_rate": 1.0978205548529902e-05,
+      "loss": 0.838,
+      "step": 746
+    },
+    {
+      "epoch": 0.49,
+      "grad_norm": 0.05224609375,
+      "learning_rate": 1.0957234071414675e-05,
+      "loss": 0.8659,
+      "step": 747
+    },
+    {
+      "epoch": 0.49,
+      "grad_norm": 0.052978515625,
+      "learning_rate": 1.0936258344558613e-05,
+      "loss": 0.858,
+      "step": 748
+    },
+    {
+      "epoch": 0.49,
+      "grad_norm": 0.0576171875,
+      "learning_rate": 1.091527846108565e-05,
+      "loss": 0.7929,
+      "step": 749
+    },
+    {
+      "epoch": 0.49,
+      "grad_norm": 0.05029296875,
+      "learning_rate": 1.0894294514138169e-05,
+      "loss": 0.7513,
+      "step": 750
+    },
+    {
+      "epoch": 0.49,
+      "grad_norm": 0.051025390625,
+      "learning_rate": 1.0873306596876602e-05,
+      "loss": 0.8347,
+      "step": 751
+    },
+    {
+      "epoch": 0.49,
+      "grad_norm": 0.054443359375,
+      "learning_rate": 1.0852314802479009e-05,
+      "loss": 0.8225,
+      "step": 752
+    },
+    {
+      "epoch": 0.49,
+      "grad_norm": 0.049072265625,
+      "learning_rate": 1.0831319224140653e-05,
+      "loss": 0.8331,
+      "step": 753
+    },
+    {
+      "epoch": 0.49,
+      "grad_norm": 0.055908203125,
+      "learning_rate": 1.08103199550736e-05,
+      "loss": 0.8195,
+      "step": 754
+    },
+    {
+      "epoch": 0.49,
+      "grad_norm": 0.057373046875,
+      "learning_rate": 1.0789317088506307e-05,
+      "loss": 0.7831,
+      "step": 755
+    },
+    {
+      "epoch": 0.49,
+      "grad_norm": 0.052978515625,
+      "learning_rate": 1.0768310717683192e-05,
+      "loss": 0.8176,
+      "step": 756
+    },
+    {
+      "epoch": 0.49,
+      "grad_norm": 0.04638671875,
+      "learning_rate": 1.0747300935864245e-05,
+      "loss": 0.7595,
+      "step": 757
+    },
+    {
+      "epoch": 0.49,
+      "grad_norm": 0.04931640625,
+      "learning_rate": 1.0726287836324583e-05,
+      "loss": 0.8544,
+      "step": 758
+    },
+    {
+      "epoch": 0.49,
+      "grad_norm": 0.05029296875,
+      "learning_rate": 1.0705271512354068e-05,
+      "loss": 0.88,
+      "step": 759
+    },
+    {
+      "epoch": 0.49,
+      "grad_norm": 0.053955078125,
+      "learning_rate": 1.0684252057256861e-05,
+      "loss": 0.8787,
+      "step": 760
+    },
+    {
+      "epoch": 0.49,
+      "grad_norm": 0.06201171875,
+      "learning_rate": 1.066322956435104e-05,
+      "loss": 0.8907,
+      "step": 761
+    },
+    {
+      "epoch": 0.5,
+      "grad_norm": 0.0537109375,
+      "learning_rate": 1.0642204126968159e-05,
+      "loss": 0.7765,
+      "step": 762
+    },
+    {
+      "epoch": 0.5,
+      "grad_norm": 0.054443359375,
+      "learning_rate": 1.062117583845285e-05,
+      "loss": 0.7968,
+      "step": 763
+    },
+    {
+      "epoch": 0.5,
+      "grad_norm": 0.05029296875,
+      "learning_rate": 1.06001447921624e-05,
+      "loss": 0.8493,
+      "step": 764
+    },
+    {
+      "epoch": 0.5,
+      "grad_norm": 0.07421875,
+      "learning_rate": 1.0579111081466333e-05,
+      "loss": 0.789,
+      "step": 765
+    },
+    {
+      "epoch": 0.5,
+      "grad_norm": 0.051513671875,
+      "learning_rate": 1.0558074799746019e-05,
+      "loss": 0.8526,
+      "step": 766
+    },
+    {
+      "epoch": 0.5,
+      "grad_norm": 0.052490234375,
+      "learning_rate": 1.0537036040394226e-05,
+      "loss": 0.8511,
+      "step": 767
+    },
+    {
+      "epoch": 0.5,
+      "grad_norm": 0.051513671875,
+      "learning_rate": 1.0515994896814731e-05,
+      "loss": 0.8339,
+      "step": 768
+    },
+    {
+      "epoch": 0.5,
+      "grad_norm": 0.05419921875,
+      "learning_rate": 1.0494951462421893e-05,
+      "loss": 0.7972,
+      "step": 769
+    },
+    {
+      "epoch": 0.5,
+      "grad_norm": 0.05029296875,
+      "learning_rate": 1.0473905830640239e-05,
+      "loss": 0.8114,
+      "step": 770
+    },
+    {
+      "epoch": 0.5,
+      "grad_norm": 0.050048828125,
+      "learning_rate": 1.0452858094904053e-05,
+      "loss": 0.7742,
+      "step": 771
+    },
+    {
+      "epoch": 0.5,
+      "grad_norm": 0.05224609375,
+      "learning_rate": 1.0431808348656961e-05,
+      "loss": 0.8655,
+      "step": 772
+    },
+    {
+      "epoch": 0.5,
+      "grad_norm": 0.05224609375,
+      "learning_rate": 1.0410756685351517e-05,
+      "loss": 0.8543,
+      "step": 773
+    },
+    {
+      "epoch": 0.5,
+      "grad_norm": 0.052001953125,
+      "learning_rate": 1.0389703198448784e-05,
+      "loss": 0.8646,
+      "step": 774
+    },
+    {
+      "epoch": 0.5,
+      "grad_norm": 0.0546875,
+      "learning_rate": 1.0368647981417917e-05,
+      "loss": 0.8008,
+      "step": 775
+    },
+    {
+      "epoch": 0.5,
+      "grad_norm": 0.053466796875,
+      "learning_rate": 1.0347591127735754e-05,
+      "loss": 0.839,
+      "step": 776
+    },
+    {
+      "epoch": 0.51,
+      "grad_norm": 0.05517578125,
+      "learning_rate": 1.0326532730886405e-05,
+      "loss": 0.749,
+      "step": 777
+    },
+    {
+      "epoch": 0.51,
+      "grad_norm": 0.052978515625,
+      "learning_rate": 1.0305472884360825e-05,
+      "loss": 0.778,
+      "step": 778
+    },
+    {
+      "epoch": 0.51,
+      "grad_norm": 0.0546875,
+      "learning_rate": 1.0284411681656408e-05,
+      "loss": 0.7992,
+      "step": 779
+    },
+    {
+      "epoch": 0.51,
+      "grad_norm": 0.0517578125,
+      "learning_rate": 1.0263349216276564e-05,
+      "loss": 0.8138,
+      "step": 780
+    },
+    {
+      "epoch": 0.51,
+      "grad_norm": 0.056396484375,
+      "learning_rate": 1.0242285581730313e-05,
+      "loss": 0.8241,
+      "step": 781
+    },
+    {
+      "epoch": 0.51,
+      "grad_norm": 0.047119140625,
+      "learning_rate": 1.022122087153187e-05,
+      "loss": 0.7916,
+      "step": 782
+    },
+    {
+      "epoch": 0.51,
+      "grad_norm": 0.0517578125,
+      "learning_rate": 1.0200155179200214e-05,
+      "loss": 0.8253,
+      "step": 783
+    },
+    {
+      "epoch": 0.51,
+      "grad_norm": 0.05517578125,
+      "learning_rate": 1.0179088598258697e-05,
+      "loss": 0.8981,
+      "step": 784
+    },
+    {
+      "epoch": 0.51,
+      "grad_norm": 0.053466796875,
+      "learning_rate": 1.0158021222234602e-05,
+      "loss": 0.8234,
+      "step": 785
+    },
+    {
+      "epoch": 0.51,
+      "grad_norm": 0.053955078125,
+      "learning_rate": 1.0136953144658753e-05,
+      "loss": 0.8459,
+      "step": 786
+    },
+    {
+      "epoch": 0.51,
+      "grad_norm": 0.04931640625,
+      "learning_rate": 1.0115884459065088e-05,
+      "loss": 0.8396,
+      "step": 787
+    },
+    {
+      "epoch": 0.51,
+      "grad_norm": 0.055419921875,
+      "learning_rate": 1.009481525899024e-05,
+      "loss": 0.8258,
+      "step": 788
+    },
+    {
+      "epoch": 0.51,
+      "grad_norm": 0.050537109375,
+      "learning_rate": 1.0073745637973125e-05,
+      "loss": 0.7542,
+      "step": 789
+    },
+    {
+      "epoch": 0.51,
+      "grad_norm": 0.051513671875,
+      "learning_rate": 1.0052675689554534e-05,
+      "loss": 0.7882,
+      "step": 790
+    },
+    {
+      "epoch": 0.51,
+      "grad_norm": 0.052978515625,
+      "learning_rate": 1.0031605507276705e-05,
+      "loss": 0.8435,
+      "step": 791
+    },
+    {
+      "epoch": 0.51,
+      "grad_norm": 0.05029296875,
+      "learning_rate": 1.0010535184682921e-05,
+      "loss": 0.7813,
+      "step": 792
+    },
+    {
+      "epoch": 0.52,
+      "grad_norm": 0.0498046875,
+      "learning_rate": 9.98946481531708e-06,
+      "loss": 0.8568,
+      "step": 793
+    },
+    {
+      "epoch": 0.52,
+      "grad_norm": 0.053955078125,
+      "learning_rate": 9.968394492723298e-06,
+      "loss": 0.7625,
+      "step": 794
+    },
+    {
+      "epoch": 0.52,
+      "grad_norm": 0.0517578125,
+      "learning_rate": 9.947324310445467e-06,
+      "loss": 0.8384,
+      "step": 795
+    },
+    {
+      "epoch": 0.52,
+      "grad_norm": 0.052490234375,
+      "learning_rate": 9.926254362026875e-06,
+      "loss": 0.8727,
+      "step": 796
+    },
+    {
+      "epoch": 0.52,
+      "grad_norm": 0.057373046875,
+      "learning_rate": 9.905184741009765e-06,
+      "loss": 0.8494,
+      "step": 797
+    },
+    {
+      "epoch": 0.52,
+      "grad_norm": 0.06591796875,
+      "learning_rate": 9.884115540934915e-06,
+      "loss": 0.8167,
+      "step": 798
+    },
+    {
+      "epoch": 0.52,
+      "grad_norm": 0.0576171875,
+      "learning_rate": 9.863046855341247e-06,
+      "loss": 0.7945,
+      "step": 799
+    },
+    {
+      "epoch": 0.52,
+      "grad_norm": 0.05419921875,
+      "learning_rate": 9.841978777765401e-06,
+      "loss": 0.8663,
+      "step": 800
+    },
+    {
+      "epoch": 0.52,
+      "grad_norm": 0.0498046875,
+      "learning_rate": 9.820911401741306e-06,
+      "loss": 0.8504,
+      "step": 801
+    },
+    {
+      "epoch": 0.52,
+      "grad_norm": 0.055419921875,
+      "learning_rate": 9.79984482079979e-06,
+      "loss": 0.9436,
+      "step": 802
+    },
+    {
+      "epoch": 0.52,
+      "grad_norm": 0.05322265625,
+      "learning_rate": 9.778779128468133e-06,
+      "loss": 0.7959,
+      "step": 803
+    },
+    {
+      "epoch": 0.52,
+      "grad_norm": 0.055908203125,
+      "learning_rate": 9.757714418269687e-06,
+      "loss": 0.8337,
+      "step": 804
+    },
+    {
+      "epoch": 0.52,
+      "grad_norm": 0.053466796875,
+      "learning_rate": 9.73665078372344e-06,
+      "loss": 0.8399,
+      "step": 805
+    },
+    {
+      "epoch": 0.52,
+      "grad_norm": 0.053955078125,
+      "learning_rate": 9.715588318343594e-06,
+      "loss": 0.832,
+      "step": 806
+    },
+    {
+      "epoch": 0.52,
+      "grad_norm": 0.055908203125,
+      "learning_rate": 9.694527115639175e-06,
+      "loss": 0.8985,
+      "step": 807
+    },
+    {
+      "epoch": 0.53,
+      "grad_norm": 0.05029296875,
+      "learning_rate": 9.673467269113599e-06,
+      "loss": 0.7697,
+      "step": 808
+    },
+    {
+      "epoch": 0.53,
+      "grad_norm": 0.04931640625,
+      "learning_rate": 9.652408872264249e-06,
+      "loss": 0.8257,
+      "step": 809
+    },
+    {
+      "epoch": 0.53,
+      "grad_norm": 0.055908203125,
+      "learning_rate": 9.631352018582088e-06,
+      "loss": 0.8525,
+      "step": 810
+    },
+    {
+      "epoch": 0.53,
+      "grad_norm": 0.051513671875,
+      "learning_rate": 9.61029680155122e-06,
+      "loss": 0.758,
+      "step": 811
+    },
+    {
+      "epoch": 0.53,
+      "grad_norm": 0.051025390625,
+      "learning_rate": 9.589243314648483e-06,
+      "loss": 0.9023,
+      "step": 812
+    },
+    {
+      "epoch": 0.53,
+      "grad_norm": 0.050537109375,
+      "learning_rate": 9.568191651343042e-06,
+      "loss": 0.7768,
+      "step": 813
+    },
+    {
+      "epoch": 0.53,
+      "grad_norm": 0.0517578125,
+      "learning_rate": 9.54714190509595e-06,
+      "loss": 0.8373,
+      "step": 814
+    },
+    {
+      "epoch": 0.53,
+      "grad_norm": 0.059814453125,
+      "learning_rate": 9.526094169359766e-06,
+      "loss": 0.8108,
+      "step": 815
+    },
+    {
+      "epoch": 0.53,
+      "grad_norm": 0.052490234375,
+      "learning_rate": 9.50504853757811e-06,
+      "loss": 0.8059,
+      "step": 816
+    },
+    {
+      "epoch": 0.53,
+      "grad_norm": 0.05908203125,
+      "learning_rate": 9.48400510318527e-06,
+      "loss": 0.8751,
+      "step": 817
+    },
+    {
+      "epoch": 0.53,
+      "grad_norm": 0.05029296875,
+      "learning_rate": 9.462963959605777e-06,
+      "loss": 0.8535,
+      "step": 818
+    },
+    {
+      "epoch": 0.53,
+      "grad_norm": 0.0537109375,
+      "learning_rate": 9.441925200253985e-06,
+      "loss": 0.8148,
+      "step": 819
+    },
+    {
+      "epoch": 0.53,
+      "grad_norm": 0.05126953125,
+      "learning_rate": 9.420888918533669e-06,
+      "loss": 0.8368,
+      "step": 820
+    },
+    {
+      "epoch": 0.53,
+      "grad_norm": 0.048095703125,
+      "learning_rate": 9.399855207837606e-06,
+      "loss": 0.9126,
+      "step": 821
+    },
+    {
+      "epoch": 0.53,
+      "grad_norm": 0.0517578125,
+      "learning_rate": 9.378824161547152e-06,
+      "loss": 0.8548,
+      "step": 822
+    },
+    {
+      "epoch": 0.54,
+      "grad_norm": 0.055419921875,
+      "learning_rate": 9.357795873031841e-06,
+      "loss": 0.9106,
+      "step": 823
+    },
+    {
+      "epoch": 0.54,
+      "grad_norm": 0.058349609375,
+      "learning_rate": 9.336770435648963e-06,
+      "loss": 0.8111,
+      "step": 824
+    },
+    {
+      "epoch": 0.54,
+      "grad_norm": 0.056640625,
+      "learning_rate": 9.315747942743142e-06,
+      "loss": 0.8079,
+      "step": 825
+    },
+    {
+      "epoch": 0.54,
+      "grad_norm": 0.05126953125,
+      "learning_rate": 9.294728487645934e-06,
+      "loss": 0.7641,
+      "step": 826
+    },
+    {
+      "epoch": 0.54,
+      "grad_norm": 0.05908203125,
+      "learning_rate": 9.273712163675419e-06,
+      "loss": 0.7871,
+      "step": 827
+    },
+    {
+      "epoch": 0.54,
+      "grad_norm": 0.05615234375,
+      "learning_rate": 9.252699064135759e-06,
+      "loss": 0.8296,
+      "step": 828
+    },
+    {
+      "epoch": 0.54,
+      "grad_norm": 0.05078125,
+      "learning_rate": 9.23168928231681e-06,
+      "loss": 0.8139,
+      "step": 829
+    },
+    {
+      "epoch": 0.54,
+      "grad_norm": 0.053955078125,
+      "learning_rate": 9.210682911493697e-06,
+      "loss": 0.8108,
+      "step": 830
+    },
+    {
+      "epoch": 0.54,
+      "grad_norm": 0.0546875,
+      "learning_rate": 9.189680044926402e-06,
+      "loss": 0.843,
+      "step": 831
+    },
+    {
+      "epoch": 0.54,
+      "grad_norm": 0.061279296875,
+      "learning_rate": 9.168680775859352e-06,
+      "loss": 0.8102,
+      "step": 832
+    },
+    {
+      "epoch": 0.54,
+      "grad_norm": 0.0576171875,
+      "learning_rate": 9.147685197520995e-06,
+      "loss": 0.8662,
+      "step": 833
+    },
+    {
+      "epoch": 0.54,
+      "grad_norm": 0.05126953125,
+      "learning_rate": 9.126693403123398e-06,
+      "loss": 0.9419,
+      "step": 834
+    },
+    {
+      "epoch": 0.54,
+      "grad_norm": 0.051513671875,
+      "learning_rate": 9.105705485861834e-06,
+      "loss": 0.8292,
+      "step": 835
+    },
+    {
+      "epoch": 0.54,
+      "grad_norm": 0.0546875,
+      "learning_rate": 9.084721538914354e-06,
+      "loss": 0.8407,
+      "step": 836
+    },
+    {
+      "epoch": 0.54,
+      "grad_norm": 0.054443359375,
+      "learning_rate": 9.06374165544139e-06,
+      "loss": 0.818,
+      "step": 837
+    },
+    {
+      "epoch": 0.54,
+      "grad_norm": 0.052490234375,
+      "learning_rate": 9.042765928585327e-06,
+      "loss": 0.8263,
+      "step": 838
+    },
+    {
+      "epoch": 0.55,
+      "grad_norm": 0.05810546875,
+      "learning_rate": 9.0217944514701e-06,
+      "loss": 0.8929,
+      "step": 839
+    },
+    {
+      "epoch": 0.55,
+      "grad_norm": 0.048828125,
+      "learning_rate": 9.000827317200778e-06,
+      "loss": 0.8218,
+      "step": 840
+    },
+    {
+      "epoch": 0.55,
+      "grad_norm": 0.0595703125,
+      "learning_rate": 8.979864618863144e-06,
+      "loss": 0.9103,
+      "step": 841
+    },
+    {
+      "epoch": 0.55,
+      "grad_norm": 0.051513671875,
+      "learning_rate": 8.958906449523295e-06,
+      "loss": 0.7595,
+      "step": 842
+    },
+    {
+      "epoch": 0.55,
+      "grad_norm": 0.0498046875,
+      "learning_rate": 8.93795290222722e-06,
+      "loss": 0.8868,
+      "step": 843
+    },
+    {
+      "epoch": 0.55,
+      "grad_norm": 0.05712890625,
+      "learning_rate": 8.917004070000377e-06,
+      "loss": 0.8137,
+      "step": 844
+    },
+    {
+      "epoch": 0.55,
+      "grad_norm": 0.056640625,
+      "learning_rate": 8.896060045847305e-06,
+      "loss": 0.7945,
+      "step": 845
+    },
+    {
+      "epoch": 0.55,
+      "grad_norm": 0.05810546875,
+      "learning_rate": 8.875120922751186e-06,
+      "loss": 0.8517,
+      "step": 846
+    },
+    {
+      "epoch": 0.55,
+      "grad_norm": 0.05419921875,
+      "learning_rate": 8.854186793673454e-06,
+      "loss": 0.8248,
+      "step": 847
+    },
+    {
+      "epoch": 0.55,
+      "grad_norm": 0.05224609375,
+      "learning_rate": 8.833257751553365e-06,
+      "loss": 0.7861,
+      "step": 848
+    },
+    {
+      "epoch": 0.55,
+      "grad_norm": 0.049072265625,
+      "learning_rate": 8.812333889307586e-06,
+      "loss": 0.8321,
+      "step": 849
+    },
+    {
+      "epoch": 0.55,
+      "grad_norm": 0.053466796875,
+      "learning_rate": 8.791415299829798e-06,
+      "loss": 0.8041,
+      "step": 850
+    },
+    {
+      "epoch": 0.55,
+      "grad_norm": 0.0537109375,
+      "learning_rate": 8.77050207599027e-06,
+      "loss": 0.8564,
+      "step": 851
+    },
+    {
+      "epoch": 0.55,
+      "grad_norm": 0.059326171875,
+      "learning_rate": 8.749594310635442e-06,
+      "loss": 0.811,
+      "step": 852
+    },
+    {
+      "epoch": 0.55,
+      "grad_norm": 0.052001953125,
+      "learning_rate": 8.728692096587536e-06,
+      "loss": 0.8223,
+      "step": 853
+    },
+    {
+      "epoch": 0.56,
+      "grad_norm": 0.056884765625,
+      "learning_rate": 8.707795526644107e-06,
+      "loss": 0.8617,
+      "step": 854
+    },
+    {
+      "epoch": 0.56,
+      "grad_norm": 0.05078125,
+      "learning_rate": 8.686904693577668e-06,
+      "loss": 0.778,
+      "step": 855
+    },
+    {
+      "epoch": 0.56,
+      "grad_norm": 0.05322265625,
+      "learning_rate": 8.666019690135264e-06,
+      "loss": 0.7807,
+      "step": 856
+    },
+    {
+      "epoch": 0.56,
+      "grad_norm": 0.051513671875,
+      "learning_rate": 8.645140609038045e-06,
+      "loss": 0.8375,
+      "step": 857
+    },
+    {
+      "epoch": 0.56,
+      "grad_norm": 0.050048828125,
+      "learning_rate": 8.624267542980882e-06,
+      "loss": 0.8494,
+      "step": 858
+    },
+    {
+      "epoch": 0.56,
+      "grad_norm": 0.05078125,
+      "learning_rate": 8.60340058463194e-06,
+      "loss": 0.7645,
+      "step": 859
+    },
+    {
+      "epoch": 0.56,
+      "grad_norm": 0.055419921875,
+      "learning_rate": 8.582539826632253e-06,
+      "loss": 0.8531,
+      "step": 860
+    },
+    {
+      "epoch": 0.56,
+      "grad_norm": 0.053466796875,
+      "learning_rate": 8.561685361595353e-06,
+      "loss": 0.7901,
+      "step": 861
+    },
+    {
+      "epoch": 0.56,
+      "grad_norm": 0.04931640625,
+      "learning_rate": 8.540837282106809e-06,
+      "loss": 0.7988,
+      "step": 862
+    },
+    {
+      "epoch": 0.56,
+      "grad_norm": 0.05322265625,
+      "learning_rate": 8.519995680723853e-06,
+      "loss": 0.7846,
+      "step": 863
+    },
+    {
+      "epoch": 0.56,
+      "grad_norm": 0.0537109375,
+      "learning_rate": 8.499160649974964e-06,
+      "loss": 0.7196,
+      "step": 864
+    },
+    {
+      "epoch": 0.56,
+      "grad_norm": 0.055908203125,
+      "learning_rate": 8.47833228235943e-06,
+      "loss": 0.8331,
+      "step": 865
+    },
+    {
+      "epoch": 0.56,
+      "grad_norm": 0.055419921875,
+      "learning_rate": 8.457510670346976e-06,
+      "loss": 0.7706,
+      "step": 866
+    },
+    {
+      "epoch": 0.56,
+      "grad_norm": 0.053466796875,
+      "learning_rate": 8.43669590637733e-06,
+      "loss": 0.819,
+      "step": 867
+    },
+    {
+      "epoch": 0.56,
+      "grad_norm": 0.051025390625,
+      "learning_rate": 8.415888082859806e-06,
+      "loss": 0.8358,
+      "step": 868
+    },
+    {
+      "epoch": 0.57,
+      "grad_norm": 0.05615234375,
+      "learning_rate": 8.39508729217292e-06,
+      "loss": 0.9253,
+      "step": 869
+    },
+    {
+      "epoch": 0.57,
+      "grad_norm": 0.05859375,
+      "learning_rate": 8.374293626663958e-06,
+      "loss": 0.9424,
+      "step": 870
+    },
+    {
+      "epoch": 0.57,
+      "grad_norm": 0.04833984375,
+      "learning_rate": 8.353507178648572e-06,
+      "loss": 0.7783,
+      "step": 871
+    },
+    {
+      "epoch": 0.57,
+      "grad_norm": 0.054931640625,
+      "learning_rate": 8.33272804041038e-06,
+      "loss": 0.838,
+      "step": 872
+    },
+    {
+      "epoch": 0.57,
+      "grad_norm": 0.056640625,
+      "learning_rate": 8.311956304200532e-06,
+      "loss": 0.8241,
+      "step": 873
+    },
+    {
+      "epoch": 0.57,
+      "grad_norm": 0.07421875,
+      "learning_rate": 8.291192062237329e-06,
+      "loss": 0.8346,
+      "step": 874
+    },
+    {
+      "epoch": 0.57,
+      "grad_norm": 0.051513671875,
+      "learning_rate": 8.2704354067058e-06,
+      "loss": 0.752,
+      "step": 875
+    },
+    {
+      "epoch": 0.57,
+      "grad_norm": 0.052001953125,
+      "learning_rate": 8.249686429757282e-06,
+      "loss": 0.8214,
+      "step": 876
+    },
+    {
+      "epoch": 0.57,
+      "grad_norm": 0.05810546875,
+      "learning_rate": 8.228945223509037e-06,
+      "loss": 0.7654,
+      "step": 877
+    },
+    {
+      "epoch": 0.57,
+      "grad_norm": 0.058349609375,
+      "learning_rate": 8.208211880043812e-06,
+      "loss": 0.8251,
+      "step": 878
+    },
+    {
+      "epoch": 0.57,
+      "grad_norm": 0.052978515625,
+      "learning_rate": 8.187486491409462e-06,
+      "loss": 0.8461,
+      "step": 879
+    },
+    {
+      "epoch": 0.57,
+      "grad_norm": 0.052734375,
+      "learning_rate": 8.166769149618517e-06,
+      "loss": 0.7852,
+      "step": 880
+    },
+    {
+      "epoch": 0.57,
+      "grad_norm": 0.05419921875,
+      "learning_rate": 8.146059946647784e-06,
+      "loss": 0.8457,
+      "step": 881
+    },
+    {
+      "epoch": 0.57,
+      "grad_norm": 0.0537109375,
+      "learning_rate": 8.125358974437933e-06,
+      "loss": 0.868,
+      "step": 882
+    },
+    {
+      "epoch": 0.57,
+      "grad_norm": 0.053955078125,
+      "learning_rate": 8.104666324893106e-06,
+      "loss": 0.7821,
+      "step": 883
+    },
+    {
+      "epoch": 0.57,
+      "grad_norm": 0.052978515625,
+      "learning_rate": 8.083982089880477e-06,
+      "loss": 0.8308,
+      "step": 884
+    },
+    {
+      "epoch": 0.58,
+      "grad_norm": 0.055908203125,
+      "learning_rate": 8.063306361229876e-06,
+      "loss": 0.8241,
+      "step": 885
+    },
+    {
+      "epoch": 0.58,
+      "grad_norm": 0.0537109375,
+      "learning_rate": 8.042639230733364e-06,
+      "loss": 0.8983,
+      "step": 886
+    },
+    {
+      "epoch": 0.58,
+      "grad_norm": 0.0556640625,
+      "learning_rate": 8.021980790144828e-06,
+      "loss": 0.8106,
+      "step": 887
+    },
+    {
+      "epoch": 0.58,
+      "grad_norm": 0.05078125,
+      "learning_rate": 8.001331131179581e-06,
+      "loss": 0.7587,
+      "step": 888
+    },
+    {
+      "epoch": 0.58,
+      "grad_norm": 0.061767578125,
+      "learning_rate": 7.98069034551394e-06,
+      "loss": 0.8053,
+      "step": 889
+    },
+    {
+      "epoch": 0.58,
+      "grad_norm": 0.057373046875,
+      "learning_rate": 7.960058524784833e-06,
+      "loss": 0.7977,
+      "step": 890
+    },
+    {
+      "epoch": 0.58,
+      "grad_norm": 0.054443359375,
+      "learning_rate": 7.939435760589392e-06,
+      "loss": 0.8021,
+      "step": 891
+    },
+    {
+      "epoch": 0.58,
+      "grad_norm": 0.052001953125,
+      "learning_rate": 7.918822144484532e-06,
+      "loss": 0.832,
+      "step": 892
+    },
+    {
+      "epoch": 0.58,
+      "grad_norm": 0.061279296875,
+      "learning_rate": 7.898217767986562e-06,
+      "loss": 0.8482,
+      "step": 893
+    },
+    {
+      "epoch": 0.58,
+      "grad_norm": 0.054443359375,
+      "learning_rate": 7.877622722570772e-06,
+      "loss": 0.8368,
+      "step": 894
+    },
+    {
+      "epoch": 0.58,
+      "grad_norm": 0.056396484375,
+      "learning_rate": 7.857037099671008e-06,
+      "loss": 0.8214,
+      "step": 895
+    },
+    {
+      "epoch": 0.58,
+      "grad_norm": 0.056396484375,
+      "learning_rate": 7.836460990679312e-06,
+      "loss": 0.8223,
+      "step": 896
+    },
+    {
+      "epoch": 0.58,
+      "grad_norm": 0.051513671875,
+      "learning_rate": 7.815894486945466e-06,
+      "loss": 0.793,
+      "step": 897
+    },
+    {
+      "epoch": 0.58,
+      "grad_norm": 0.050537109375,
+      "learning_rate": 7.795337679776613e-06,
+      "loss": 0.8183,
+      "step": 898
+    },
+    {
+      "epoch": 0.58,
+      "grad_norm": 0.048828125,
+      "learning_rate": 7.774790660436857e-06,
+      "loss": 0.774,
+      "step": 899
+    },
+    {
+      "epoch": 0.59,
+      "grad_norm": 0.052734375,
+      "learning_rate": 7.754253520146835e-06,
+      "loss": 0.887,
+      "step": 900
+    },
+    {
+      "epoch": 0.59,
+      "grad_norm": 0.0498046875,
+      "learning_rate": 7.733726350083331e-06,
+      "loss": 0.7979,
+      "step": 901
+    },
+    {
+      "epoch": 0.59,
+      "grad_norm": 0.05615234375,
+      "learning_rate": 7.713209241378871e-06,
+      "loss": 0.8028,
+      "step": 902
+    },
+    {
+      "epoch": 0.59,
+      "grad_norm": 0.0546875,
+      "learning_rate": 7.692702285121299e-06,
+      "loss": 0.7976,
+      "step": 903
+    },
+    {
+      "epoch": 0.59,
+      "grad_norm": 0.0498046875,
+      "learning_rate": 7.672205572353394e-06,
+      "loss": 0.747,
+      "step": 904
+    },
+    {
+      "epoch": 0.59,
+      "grad_norm": 0.08251953125,
+      "learning_rate": 7.65171919407246e-06,
+      "loss": 0.7857,
+      "step": 905
+    },
+    {
+      "epoch": 0.59,
+      "grad_norm": 0.051513671875,
+      "learning_rate": 7.631243241229916e-06,
+      "loss": 0.7942,
+      "step": 906
+    },
+    {
+      "epoch": 0.59,
+      "grad_norm": 0.052001953125,
+      "learning_rate": 7.610777804730903e-06,
+      "loss": 0.8072,
+      "step": 907
+    },
+    {
+      "epoch": 0.59,
+      "grad_norm": 0.0537109375,
+      "learning_rate": 7.590322975433857e-06,
+      "loss": 0.7873,
+      "step": 908
+    },
+    {
+      "epoch": 0.59,
+      "grad_norm": 0.0537109375,
+      "learning_rate": 7.56987884415014e-06,
+      "loss": 0.8778,
+      "step": 909
+    },
+    {
+      "epoch": 0.59,
+      "grad_norm": 0.051513671875,
+      "learning_rate": 7.549445501643615e-06,
+      "loss": 0.7803,
+      "step": 910
+    },
+    {
+      "epoch": 0.59,
+      "grad_norm": 0.0517578125,
+      "learning_rate": 7.5290230386302384e-06,
+      "loss": 0.8263,
+      "step": 911
+    },
+    {
+      "epoch": 0.59,
+      "grad_norm": 0.05126953125,
+      "learning_rate": 7.508611545777679e-06,
+      "loss": 0.8951,
+      "step": 912
+    },
+    {
+      "epoch": 0.59,
+      "grad_norm": 0.0517578125,
+      "learning_rate": 7.488211113704886e-06,
+      "loss": 0.8128,
+      "step": 913
+    },
+    {
+      "epoch": 0.59,
+      "grad_norm": 0.054931640625,
+      "learning_rate": 7.46782183298172e-06,
+      "loss": 0.833,
+      "step": 914
+    },
+    {
+      "epoch": 0.59,
+      "grad_norm": 0.056640625,
+      "learning_rate": 7.447443794128525e-06,
+      "loss": 0.783,
+      "step": 915
+    },
+    {
+      "epoch": 0.6,
+      "grad_norm": 0.052734375,
+      "learning_rate": 7.427077087615735e-06,
+      "loss": 0.8378,
+      "step": 916
+    },
+    {
+      "epoch": 0.6,
+      "grad_norm": 0.056884765625,
+      "learning_rate": 7.406721803863475e-06,
+      "loss": 0.8127,
+      "step": 917
+    },
+    {
+      "epoch": 0.6,
+      "grad_norm": 0.06640625,
+      "learning_rate": 7.386378033241164e-06,
+      "loss": 0.7863,
+      "step": 918
+    },
+    {
+      "epoch": 0.6,
+      "grad_norm": 0.054931640625,
+      "learning_rate": 7.3660458660670905e-06,
+      "loss": 0.795,
+      "step": 919
+    },
+    {
+      "epoch": 0.6,
+      "grad_norm": 0.060791015625,
+      "learning_rate": 7.345725392608047e-06,
+      "loss": 0.9132,
+      "step": 920
+    },
+    {
+      "epoch": 0.6,
+      "grad_norm": 0.052001953125,
+      "learning_rate": 7.3254167030788955e-06,
+      "loss": 0.8416,
+      "step": 921
+    },
+    {
+      "epoch": 0.6,
+      "grad_norm": 0.055419921875,
+      "learning_rate": 7.305119887642191e-06,
+      "loss": 0.8038,
+      "step": 922
+    },
+    {
+      "epoch": 0.6,
+      "grad_norm": 0.053466796875,
+      "learning_rate": 7.284835036407776e-06,
+      "loss": 0.8545,
+      "step": 923
+    },
+    {
+      "epoch": 0.6,
+      "grad_norm": 0.049560546875,
+      "learning_rate": 7.26456223943236e-06,
+      "loss": 0.8192,
+      "step": 924
+    }
+  ],
+  "logging_steps": 1,
+  "max_steps": 1538,
+  "num_input_tokens_seen": 0,
+  "num_train_epochs": 1,
+  "save_steps": 462,
+  "total_flos": 2.5348887258259784e+18,
+  "train_batch_size": 16,
+  "trial_name": null,
+  "trial_params": null
+}