diff --git "a/trainer_state.json" "b/trainer_state.json"
new file mode 100644--- /dev/null
+++ "b/trainer_state.json"
@@ -0,0 +1,10185 @@
+{
+  "best_metric": null,
+  "best_model_checkpoint": null,
+  "epoch": 2.9979310344827588,
+  "eval_steps": 500,
+  "global_step": 1449,
+  "is_hyper_param_search": false,
+  "is_local_process_zero": true,
+  "is_world_process_zero": true,
+  "log_history": [
+    {
+      "epoch": 0.0020689655172413794,
+      "grad_norm": 8.16248893737793,
+      "learning_rate": 6.896551724137931e-08,
+      "loss": 1.3047,
+      "step": 1
+    },
+    {
+      "epoch": 0.004137931034482759,
+      "grad_norm": 8.501221656799316,
+      "learning_rate": 1.3793103448275863e-07,
+      "loss": 1.3364,
+      "step": 2
+    },
+    {
+      "epoch": 0.006206896551724138,
+      "grad_norm": 8.676358222961426,
+      "learning_rate": 2.0689655172413796e-07,
+      "loss": 1.3572,
+      "step": 3
+    },
+    {
+      "epoch": 0.008275862068965517,
+      "grad_norm": 8.630147933959961,
+      "learning_rate": 2.7586206896551726e-07,
+      "loss": 1.3387,
+      "step": 4
+    },
+    {
+      "epoch": 0.010344827586206896,
+      "grad_norm": 8.223169326782227,
+      "learning_rate": 3.4482758620689656e-07,
+      "loss": 1.3156,
+      "step": 5
+    },
+    {
+      "epoch": 0.012413793103448275,
+      "grad_norm": 8.330610275268555,
+      "learning_rate": 4.137931034482759e-07,
+      "loss": 1.2883,
+      "step": 6
+    },
+    {
+      "epoch": 0.014482758620689656,
+      "grad_norm": 9.017132759094238,
+      "learning_rate": 4.827586206896552e-07,
+      "loss": 1.3781,
+      "step": 7
+    },
+    {
+      "epoch": 0.016551724137931035,
+      "grad_norm": 8.214015007019043,
+      "learning_rate": 5.517241379310345e-07,
+      "loss": 1.2708,
+      "step": 8
+    },
+    {
+      "epoch": 0.018620689655172412,
+      "grad_norm": 8.30260181427002,
+      "learning_rate": 6.206896551724139e-07,
+      "loss": 1.327,
+      "step": 9
+    },
+    {
+      "epoch": 0.020689655172413793,
+      "grad_norm": 7.864175319671631,
+      "learning_rate": 6.896551724137931e-07,
+      "loss": 1.2873,
+      "step": 10
+    },
+    {
+      "epoch": 0.022758620689655173,
+      "grad_norm": 7.549825191497803,
+      "learning_rate": 7.586206896551725e-07,
+      "loss": 1.2672,
+      "step": 11
+    },
+    {
+      "epoch": 0.02482758620689655,
+      "grad_norm": 7.3428802490234375,
+      "learning_rate": 8.275862068965518e-07,
+      "loss": 1.2312,
+      "step": 12
+    },
+    {
+      "epoch": 0.02689655172413793,
+      "grad_norm": 6.168758392333984,
+      "learning_rate": 8.965517241379311e-07,
+      "loss": 1.2331,
+      "step": 13
+    },
+    {
+      "epoch": 0.028965517241379312,
+      "grad_norm": 6.219021320343018,
+      "learning_rate": 9.655172413793103e-07,
+      "loss": 1.2173,
+      "step": 14
+    },
+    {
+      "epoch": 0.03103448275862069,
+      "grad_norm": 5.889226913452148,
+      "learning_rate": 1.0344827586206898e-06,
+      "loss": 1.2164,
+      "step": 15
+    },
+    {
+      "epoch": 0.03310344827586207,
+      "grad_norm": 5.6153411865234375,
+      "learning_rate": 1.103448275862069e-06,
+      "loss": 1.2284,
+      "step": 16
+    },
+    {
+      "epoch": 0.03517241379310345,
+      "grad_norm": 3.499351978302002,
+      "learning_rate": 1.1724137931034483e-06,
+      "loss": 1.1728,
+      "step": 17
+    },
+    {
+      "epoch": 0.037241379310344824,
+      "grad_norm": 3.140091896057129,
+      "learning_rate": 1.2413793103448277e-06,
+      "loss": 1.1509,
+      "step": 18
+    },
+    {
+      "epoch": 0.039310344827586205,
+      "grad_norm": 3.1030735969543457,
+      "learning_rate": 1.3103448275862072e-06,
+      "loss": 1.1557,
+      "step": 19
+    },
+    {
+      "epoch": 0.041379310344827586,
+      "grad_norm": 2.808349370956421,
+      "learning_rate": 1.3793103448275862e-06,
+      "loss": 1.099,
+      "step": 20
+    },
+    {
+      "epoch": 0.043448275862068966,
+      "grad_norm": 2.8193469047546387,
+      "learning_rate": 1.4482758620689657e-06,
+      "loss": 1.1728,
+      "step": 21
+    },
+    {
+      "epoch": 0.04551724137931035,
+      "grad_norm": 2.7212672233581543,
+      "learning_rate": 1.517241379310345e-06,
+      "loss": 1.1511,
+      "step": 22
+    },
+    {
+      "epoch": 0.04758620689655172,
+      "grad_norm": 2.5926826000213623,
+      "learning_rate": 1.5862068965517244e-06,
+      "loss": 1.1317,
+      "step": 23
+    },
+    {
+      "epoch": 0.0496551724137931,
+      "grad_norm": 3.0901269912719727,
+      "learning_rate": 1.6551724137931037e-06,
+      "loss": 1.0843,
+      "step": 24
+    },
+    {
+      "epoch": 0.05172413793103448,
+      "grad_norm": 3.495481252670288,
+      "learning_rate": 1.724137931034483e-06,
+      "loss": 1.1062,
+      "step": 25
+    },
+    {
+      "epoch": 0.05379310344827586,
+      "grad_norm": 3.160388231277466,
+      "learning_rate": 1.7931034482758622e-06,
+      "loss": 1.0713,
+      "step": 26
+    },
+    {
+      "epoch": 0.05586206896551724,
+      "grad_norm": 3.0356714725494385,
+      "learning_rate": 1.8620689655172416e-06,
+      "loss": 1.0513,
+      "step": 27
+    },
+    {
+      "epoch": 0.057931034482758624,
+      "grad_norm": 2.9434399604797363,
+      "learning_rate": 1.9310344827586207e-06,
+      "loss": 1.0907,
+      "step": 28
+    },
+    {
+      "epoch": 0.06,
+      "grad_norm": 2.5999321937561035,
+      "learning_rate": 2.0000000000000003e-06,
+      "loss": 1.0615,
+      "step": 29
+    },
+    {
+      "epoch": 0.06206896551724138,
+      "grad_norm": 2.188593864440918,
+      "learning_rate": 2.0689655172413796e-06,
+      "loss": 1.0549,
+      "step": 30
+    },
+    {
+      "epoch": 0.06413793103448276,
+      "grad_norm": 1.513140320777893,
+      "learning_rate": 2.137931034482759e-06,
+      "loss": 1.0179,
+      "step": 31
+    },
+    {
+      "epoch": 0.06620689655172414,
+      "grad_norm": 1.3605451583862305,
+      "learning_rate": 2.206896551724138e-06,
+      "loss": 1.0188,
+      "step": 32
+    },
+    {
+      "epoch": 0.06827586206896552,
+      "grad_norm": 1.7352983951568604,
+      "learning_rate": 2.2758620689655173e-06,
+      "loss": 1.0239,
+      "step": 33
+    },
+    {
+      "epoch": 0.0703448275862069,
+      "grad_norm": 1.7656395435333252,
+      "learning_rate": 2.3448275862068966e-06,
+      "loss": 0.9941,
+      "step": 34
+    },
+    {
+      "epoch": 0.07241379310344828,
+      "grad_norm": 1.684338092803955,
+      "learning_rate": 2.4137931034482762e-06,
+      "loss": 0.9742,
+      "step": 35
+    },
+    {
+      "epoch": 0.07448275862068965,
+      "grad_norm": 1.4101715087890625,
+      "learning_rate": 2.4827586206896555e-06,
+      "loss": 0.9809,
+      "step": 36
+    },
+    {
+      "epoch": 0.07655172413793103,
+      "grad_norm": 1.2714829444885254,
+      "learning_rate": 2.5517241379310347e-06,
+      "loss": 0.9693,
+      "step": 37
+    },
+    {
+      "epoch": 0.07862068965517241,
+      "grad_norm": 1.1786231994628906,
+      "learning_rate": 2.6206896551724144e-06,
+      "loss": 0.9956,
+      "step": 38
+    },
+    {
+      "epoch": 0.08068965517241379,
+      "grad_norm": 1.0718473196029663,
+      "learning_rate": 2.6896551724137932e-06,
+      "loss": 0.9836,
+      "step": 39
+    },
+    {
+      "epoch": 0.08275862068965517,
+      "grad_norm": 0.9965543746948242,
+      "learning_rate": 2.7586206896551725e-06,
+      "loss": 0.9186,
+      "step": 40
+    },
+    {
+      "epoch": 0.08482758620689655,
+      "grad_norm": 0.9652248620986938,
+      "learning_rate": 2.827586206896552e-06,
+      "loss": 0.9265,
+      "step": 41
+    },
+    {
+      "epoch": 0.08689655172413793,
+      "grad_norm": 1.0196917057037354,
+      "learning_rate": 2.8965517241379314e-06,
+      "loss": 0.9483,
+      "step": 42
+    },
+    {
+      "epoch": 0.08896551724137931,
+      "grad_norm": 1.2638320922851562,
+      "learning_rate": 2.9655172413793102e-06,
+      "loss": 0.9579,
+      "step": 43
+    },
+    {
+      "epoch": 0.0910344827586207,
+      "grad_norm": 1.0733128786087036,
+      "learning_rate": 3.03448275862069e-06,
+      "loss": 0.9739,
+      "step": 44
+    },
+    {
+      "epoch": 0.09310344827586207,
+      "grad_norm": 0.9054429531097412,
+      "learning_rate": 3.103448275862069e-06,
+      "loss": 0.9671,
+      "step": 45
+    },
+    {
+      "epoch": 0.09517241379310344,
+      "grad_norm": 0.9127644896507263,
+      "learning_rate": 3.172413793103449e-06,
+      "loss": 0.9145,
+      "step": 46
+    },
+    {
+      "epoch": 0.09724137931034482,
+      "grad_norm": 0.8835800886154175,
+      "learning_rate": 3.2413793103448277e-06,
+      "loss": 0.9169,
+      "step": 47
+    },
+    {
+      "epoch": 0.0993103448275862,
+      "grad_norm": 1.0521382093429565,
+      "learning_rate": 3.3103448275862073e-06,
+      "loss": 0.9341,
+      "step": 48
+    },
+    {
+      "epoch": 0.10137931034482758,
+      "grad_norm": 0.8984640836715698,
+      "learning_rate": 3.3793103448275866e-06,
+      "loss": 0.9007,
+      "step": 49
+    },
+    {
+      "epoch": 0.10344827586206896,
+      "grad_norm": 0.8801516890525818,
+      "learning_rate": 3.448275862068966e-06,
+      "loss": 0.9447,
+      "step": 50
+    },
+    {
+      "epoch": 0.10551724137931034,
+      "grad_norm": 0.873184859752655,
+      "learning_rate": 3.517241379310345e-06,
+      "loss": 0.9134,
+      "step": 51
+    },
+    {
+      "epoch": 0.10758620689655173,
+      "grad_norm": 0.8689625263214111,
+      "learning_rate": 3.5862068965517243e-06,
+      "loss": 0.9404,
+      "step": 52
+    },
+    {
+      "epoch": 0.1096551724137931,
+      "grad_norm": 0.8642359972000122,
+      "learning_rate": 3.655172413793104e-06,
+      "loss": 0.8982,
+      "step": 53
+    },
+    {
+      "epoch": 0.11172413793103449,
+      "grad_norm": 0.9032179117202759,
+      "learning_rate": 3.7241379310344832e-06,
+      "loss": 0.8848,
+      "step": 54
+    },
+    {
+      "epoch": 0.11379310344827587,
+      "grad_norm": 0.8974177837371826,
+      "learning_rate": 3.793103448275862e-06,
+      "loss": 0.9485,
+      "step": 55
+    },
+    {
+      "epoch": 0.11586206896551725,
+      "grad_norm": 0.9621221423149109,
+      "learning_rate": 3.862068965517241e-06,
+      "loss": 0.8727,
+      "step": 56
+    },
+    {
+      "epoch": 0.11793103448275861,
+      "grad_norm": 0.9366543889045715,
+      "learning_rate": 3.931034482758621e-06,
+      "loss": 0.926,
+      "step": 57
+    },
+    {
+      "epoch": 0.12,
+      "grad_norm": 0.8639587163925171,
+      "learning_rate": 4.000000000000001e-06,
+      "loss": 0.8623,
+      "step": 58
+    },
+    {
+      "epoch": 0.12206896551724138,
+      "grad_norm": 0.9308229684829712,
+      "learning_rate": 4.0689655172413795e-06,
+      "loss": 0.8973,
+      "step": 59
+    },
+    {
+      "epoch": 0.12413793103448276,
+      "grad_norm": 0.8033000230789185,
+      "learning_rate": 4.137931034482759e-06,
+      "loss": 0.8861,
+      "step": 60
+    },
+    {
+      "epoch": 0.12620689655172415,
+      "grad_norm": 0.9005990028381348,
+      "learning_rate": 4.206896551724138e-06,
+      "loss": 0.8823,
+      "step": 61
+    },
+    {
+      "epoch": 0.12827586206896552,
+      "grad_norm": 0.9020888805389404,
+      "learning_rate": 4.275862068965518e-06,
+      "loss": 0.8914,
+      "step": 62
+    },
+    {
+      "epoch": 0.13034482758620689,
+      "grad_norm": 0.8798072338104248,
+      "learning_rate": 4.3448275862068965e-06,
+      "loss": 0.9076,
+      "step": 63
+    },
+    {
+      "epoch": 0.13241379310344828,
+      "grad_norm": 0.8254591226577759,
+      "learning_rate": 4.413793103448276e-06,
+      "loss": 0.9246,
+      "step": 64
+    },
+    {
+      "epoch": 0.13448275862068965,
+      "grad_norm": 0.7802205085754395,
+      "learning_rate": 4.482758620689656e-06,
+      "loss": 0.8635,
+      "step": 65
+    },
+    {
+      "epoch": 0.13655172413793104,
+      "grad_norm": 0.8493502736091614,
+      "learning_rate": 4.551724137931035e-06,
+      "loss": 0.8793,
+      "step": 66
+    },
+    {
+      "epoch": 0.1386206896551724,
+      "grad_norm": 0.8410840034484863,
+      "learning_rate": 4.620689655172414e-06,
+      "loss": 0.8593,
+      "step": 67
+    },
+    {
+      "epoch": 0.1406896551724138,
+      "grad_norm": 0.796180009841919,
+      "learning_rate": 4.689655172413793e-06,
+      "loss": 0.8797,
+      "step": 68
+    },
+    {
+      "epoch": 0.14275862068965517,
+      "grad_norm": 0.7212627530097961,
+      "learning_rate": 4.758620689655173e-06,
+      "loss": 0.8661,
+      "step": 69
+    },
+    {
+      "epoch": 0.14482758620689656,
+      "grad_norm": 0.8363891243934631,
+      "learning_rate": 4.8275862068965525e-06,
+      "loss": 0.8679,
+      "step": 70
+    },
+    {
+      "epoch": 0.14689655172413793,
+      "grad_norm": 0.921846866607666,
+      "learning_rate": 4.896551724137931e-06,
+      "loss": 0.8643,
+      "step": 71
+    },
+    {
+      "epoch": 0.1489655172413793,
+      "grad_norm": 0.7880343794822693,
+      "learning_rate": 4.965517241379311e-06,
+      "loss": 0.878,
+      "step": 72
+    },
+    {
+      "epoch": 0.1510344827586207,
+      "grad_norm": 0.8128662705421448,
+      "learning_rate": 5.03448275862069e-06,
+      "loss": 0.8772,
+      "step": 73
+    },
+    {
+      "epoch": 0.15310344827586206,
+      "grad_norm": 0.8667871952056885,
+      "learning_rate": 5.1034482758620695e-06,
+      "loss": 0.8731,
+      "step": 74
+    },
+    {
+      "epoch": 0.15517241379310345,
+      "grad_norm": 0.8730148673057556,
+      "learning_rate": 5.172413793103449e-06,
+      "loss": 0.876,
+      "step": 75
+    },
+    {
+      "epoch": 0.15724137931034482,
+      "grad_norm": 0.8205820918083191,
+      "learning_rate": 5.241379310344829e-06,
+      "loss": 0.8415,
+      "step": 76
+    },
+    {
+      "epoch": 0.15931034482758621,
+      "grad_norm": 0.8799615502357483,
+      "learning_rate": 5.310344827586207e-06,
+      "loss": 0.8482,
+      "step": 77
+    },
+    {
+      "epoch": 0.16137931034482758,
+      "grad_norm": 0.936702311038971,
+      "learning_rate": 5.3793103448275865e-06,
+      "loss": 0.8677,
+      "step": 78
+    },
+    {
+      "epoch": 0.16344827586206898,
+      "grad_norm": 0.8674628734588623,
+      "learning_rate": 5.448275862068966e-06,
+      "loss": 0.8973,
+      "step": 79
+    },
+    {
+      "epoch": 0.16551724137931034,
+      "grad_norm": 0.8501123785972595,
+      "learning_rate": 5.517241379310345e-06,
+      "loss": 0.8603,
+      "step": 80
+    },
+    {
+      "epoch": 0.16758620689655174,
+      "grad_norm": 0.9434123039245605,
+      "learning_rate": 5.586206896551725e-06,
+      "loss": 0.8983,
+      "step": 81
+    },
+    {
+      "epoch": 0.1696551724137931,
+      "grad_norm": 0.8477321267127991,
+      "learning_rate": 5.655172413793104e-06,
+      "loss": 0.8621,
+      "step": 82
+    },
+    {
+      "epoch": 0.17172413793103447,
+      "grad_norm": 0.8743482232093811,
+      "learning_rate": 5.724137931034483e-06,
+      "loss": 0.8854,
+      "step": 83
+    },
+    {
+      "epoch": 0.17379310344827587,
+      "grad_norm": 0.7993690967559814,
+      "learning_rate": 5.793103448275863e-06,
+      "loss": 0.8605,
+      "step": 84
+    },
+    {
+      "epoch": 0.17586206896551723,
+      "grad_norm": 0.8073793649673462,
+      "learning_rate": 5.862068965517242e-06,
+      "loss": 0.8867,
+      "step": 85
+    },
+    {
+      "epoch": 0.17793103448275863,
+      "grad_norm": 0.8699036240577698,
+      "learning_rate": 5.9310344827586205e-06,
+      "loss": 0.8752,
+      "step": 86
+    },
+    {
+      "epoch": 0.18,
+      "grad_norm": 0.9213085770606995,
+      "learning_rate": 6e-06,
+      "loss": 0.8601,
+      "step": 87
+    },
+    {
+      "epoch": 0.1820689655172414,
+      "grad_norm": 0.8272461295127869,
+      "learning_rate": 6.06896551724138e-06,
+      "loss": 0.8613,
+      "step": 88
+    },
+    {
+      "epoch": 0.18413793103448275,
+      "grad_norm": 0.8762076497077942,
+      "learning_rate": 6.1379310344827595e-06,
+      "loss": 0.8755,
+      "step": 89
+    },
+    {
+      "epoch": 0.18620689655172415,
+      "grad_norm": 0.8086516261100769,
+      "learning_rate": 6.206896551724138e-06,
+      "loss": 0.8779,
+      "step": 90
+    },
+    {
+      "epoch": 0.18827586206896552,
+      "grad_norm": 0.8407052755355835,
+      "learning_rate": 6.275862068965518e-06,
+      "loss": 0.869,
+      "step": 91
+    },
+    {
+      "epoch": 0.19034482758620688,
+      "grad_norm": 0.8412452340126038,
+      "learning_rate": 6.344827586206898e-06,
+      "loss": 0.8937,
+      "step": 92
+    },
+    {
+      "epoch": 0.19241379310344828,
+      "grad_norm": 0.8644373416900635,
+      "learning_rate": 6.413793103448276e-06,
+      "loss": 0.8627,
+      "step": 93
+    },
+    {
+      "epoch": 0.19448275862068964,
+      "grad_norm": 0.867426335811615,
+      "learning_rate": 6.482758620689655e-06,
+      "loss": 0.8714,
+      "step": 94
+    },
+    {
+      "epoch": 0.19655172413793104,
+      "grad_norm": 0.8877378702163696,
+      "learning_rate": 6.551724137931035e-06,
+      "loss": 0.8603,
+      "step": 95
+    },
+    {
+      "epoch": 0.1986206896551724,
+      "grad_norm": 0.7677062749862671,
+      "learning_rate": 6.620689655172415e-06,
+      "loss": 0.8273,
+      "step": 96
+    },
+    {
+      "epoch": 0.2006896551724138,
+      "grad_norm": 0.8652222752571106,
+      "learning_rate": 6.6896551724137935e-06,
+      "loss": 0.8786,
+      "step": 97
+    },
+    {
+      "epoch": 0.20275862068965517,
+      "grad_norm": 0.8587886691093445,
+      "learning_rate": 6.758620689655173e-06,
+      "loss": 0.8541,
+      "step": 98
+    },
+    {
+      "epoch": 0.20482758620689656,
+      "grad_norm": 0.8325220942497253,
+      "learning_rate": 6.827586206896553e-06,
+      "loss": 0.8389,
+      "step": 99
+    },
+    {
+      "epoch": 0.20689655172413793,
+      "grad_norm": 0.9126191735267639,
+      "learning_rate": 6.896551724137932e-06,
+      "loss": 0.8588,
+      "step": 100
+    },
+    {
+      "epoch": 0.20896551724137932,
+      "grad_norm": 0.8659350872039795,
+      "learning_rate": 6.9655172413793105e-06,
+      "loss": 0.8649,
+      "step": 101
+    },
+    {
+      "epoch": 0.2110344827586207,
+      "grad_norm": 0.80833899974823,
+      "learning_rate": 7.03448275862069e-06,
+      "loss": 0.8668,
+      "step": 102
+    },
+    {
+      "epoch": 0.21310344827586206,
+      "grad_norm": 0.8443086743354797,
+      "learning_rate": 7.103448275862069e-06,
+      "loss": 0.852,
+      "step": 103
+    },
+    {
+      "epoch": 0.21517241379310345,
+      "grad_norm": 0.9743650555610657,
+      "learning_rate": 7.172413793103449e-06,
+      "loss": 0.8727,
+      "step": 104
+    },
+    {
+      "epoch": 0.21724137931034482,
+      "grad_norm": 0.8710004687309265,
+      "learning_rate": 7.241379310344828e-06,
+      "loss": 0.8605,
+      "step": 105
+    },
+    {
+      "epoch": 0.2193103448275862,
+      "grad_norm": 0.9044391512870789,
+      "learning_rate": 7.310344827586208e-06,
+      "loss": 0.8395,
+      "step": 106
+    },
+    {
+      "epoch": 0.22137931034482758,
+      "grad_norm": 0.7920845746994019,
+      "learning_rate": 7.379310344827587e-06,
+      "loss": 0.8361,
+      "step": 107
+    },
+    {
+      "epoch": 0.22344827586206897,
+      "grad_norm": 0.9043503403663635,
+      "learning_rate": 7.4482758620689665e-06,
+      "loss": 0.8326,
+      "step": 108
+    },
+    {
+      "epoch": 0.22551724137931034,
+      "grad_norm": 0.914540708065033,
+      "learning_rate": 7.517241379310345e-06,
+      "loss": 0.8592,
+      "step": 109
+    },
+    {
+      "epoch": 0.22758620689655173,
+      "grad_norm": 0.889606773853302,
+      "learning_rate": 7.586206896551724e-06,
+      "loss": 0.871,
+      "step": 110
+    },
+    {
+      "epoch": 0.2296551724137931,
+      "grad_norm": 0.8037165403366089,
+      "learning_rate": 7.655172413793104e-06,
+      "loss": 0.8059,
+      "step": 111
+    },
+    {
+      "epoch": 0.2317241379310345,
+      "grad_norm": 0.9440395832061768,
+      "learning_rate": 7.724137931034483e-06,
+      "loss": 0.8511,
+      "step": 112
+    },
+    {
+      "epoch": 0.23379310344827586,
+      "grad_norm": 0.8397315740585327,
+      "learning_rate": 7.793103448275863e-06,
+      "loss": 0.8363,
+      "step": 113
+    },
+    {
+      "epoch": 0.23586206896551723,
+      "grad_norm": 0.9024421572685242,
+      "learning_rate": 7.862068965517242e-06,
+      "loss": 0.8276,
+      "step": 114
+    },
+    {
+      "epoch": 0.23793103448275862,
+      "grad_norm": 0.9368509650230408,
+      "learning_rate": 7.93103448275862e-06,
+      "loss": 0.8896,
+      "step": 115
+    },
+    {
+      "epoch": 0.24,
+      "grad_norm": 0.9322966933250427,
+      "learning_rate": 8.000000000000001e-06,
+      "loss": 0.8054,
+      "step": 116
+    },
+    {
+      "epoch": 0.24206896551724139,
+      "grad_norm": 0.9048160910606384,
+      "learning_rate": 8.06896551724138e-06,
+      "loss": 0.8448,
+      "step": 117
+    },
+    {
+      "epoch": 0.24413793103448275,
+      "grad_norm": 0.8558708429336548,
+      "learning_rate": 8.137931034482759e-06,
+      "loss": 0.8275,
+      "step": 118
+    },
+    {
+      "epoch": 0.24620689655172415,
+      "grad_norm": 0.8424068093299866,
+      "learning_rate": 8.206896551724138e-06,
+      "loss": 0.8694,
+      "step": 119
+    },
+    {
+      "epoch": 0.2482758620689655,
+      "grad_norm": 1.001757025718689,
+      "learning_rate": 8.275862068965518e-06,
+      "loss": 0.8453,
+      "step": 120
+    },
+    {
+      "epoch": 0.2503448275862069,
+      "grad_norm": 0.9482073783874512,
+      "learning_rate": 8.344827586206897e-06,
+      "loss": 0.8696,
+      "step": 121
+    },
+    {
+      "epoch": 0.2524137931034483,
+      "grad_norm": 0.9771032929420471,
+      "learning_rate": 8.413793103448276e-06,
+      "loss": 0.8373,
+      "step": 122
+    },
+    {
+      "epoch": 0.25448275862068964,
+      "grad_norm": 1.0352799892425537,
+      "learning_rate": 8.482758620689656e-06,
+      "loss": 0.837,
+      "step": 123
+    },
+    {
+      "epoch": 0.25655172413793104,
+      "grad_norm": 0.9529556035995483,
+      "learning_rate": 8.551724137931035e-06,
+      "loss": 0.8286,
+      "step": 124
+    },
+    {
+      "epoch": 0.25862068965517243,
+      "grad_norm": 0.8518896698951721,
+      "learning_rate": 8.620689655172414e-06,
+      "loss": 0.847,
+      "step": 125
+    },
+    {
+      "epoch": 0.26068965517241377,
+      "grad_norm": 0.8487024903297424,
+      "learning_rate": 8.689655172413793e-06,
+      "loss": 0.8257,
+      "step": 126
+    },
+    {
+      "epoch": 0.26275862068965516,
+      "grad_norm": 1.0520756244659424,
+      "learning_rate": 8.758620689655173e-06,
+      "loss": 0.823,
+      "step": 127
+    },
+    {
+      "epoch": 0.26482758620689656,
+      "grad_norm": 1.002064824104309,
+      "learning_rate": 8.827586206896552e-06,
+      "loss": 0.8618,
+      "step": 128
+    },
+    {
+      "epoch": 0.26689655172413795,
+      "grad_norm": 0.9487592577934265,
+      "learning_rate": 8.896551724137931e-06,
+      "loss": 0.8344,
+      "step": 129
+    },
+    {
+      "epoch": 0.2689655172413793,
+      "grad_norm": 0.9966208338737488,
+      "learning_rate": 8.965517241379312e-06,
+      "loss": 0.8372,
+      "step": 130
+    },
+    {
+      "epoch": 0.2710344827586207,
+      "grad_norm": 0.8904616832733154,
+      "learning_rate": 9.03448275862069e-06,
+      "loss": 0.8547,
+      "step": 131
+    },
+    {
+      "epoch": 0.2731034482758621,
+      "grad_norm": 0.8750537037849426,
+      "learning_rate": 9.10344827586207e-06,
+      "loss": 0.8423,
+      "step": 132
+    },
+    {
+      "epoch": 0.2751724137931034,
+      "grad_norm": 1.0919158458709717,
+      "learning_rate": 9.172413793103448e-06,
+      "loss": 0.8295,
+      "step": 133
+    },
+    {
+      "epoch": 0.2772413793103448,
+      "grad_norm": 0.8703680038452148,
+      "learning_rate": 9.241379310344829e-06,
+      "loss": 0.7951,
+      "step": 134
+    },
+    {
+      "epoch": 0.2793103448275862,
+      "grad_norm": 0.8651219606399536,
+      "learning_rate": 9.310344827586207e-06,
+      "loss": 0.8656,
+      "step": 135
+    },
+    {
+      "epoch": 0.2813793103448276,
+      "grad_norm": 1.1044590473175049,
+      "learning_rate": 9.379310344827586e-06,
+      "loss": 0.8359,
+      "step": 136
+    },
+    {
+      "epoch": 0.28344827586206894,
+      "grad_norm": 0.8297178149223328,
+      "learning_rate": 9.448275862068967e-06,
+      "loss": 0.8055,
+      "step": 137
+    },
+    {
+      "epoch": 0.28551724137931034,
+      "grad_norm": 0.8873213529586792,
+      "learning_rate": 9.517241379310346e-06,
+      "loss": 0.8346,
+      "step": 138
+    },
+    {
+      "epoch": 0.28758620689655173,
+      "grad_norm": 0.9231070280075073,
+      "learning_rate": 9.586206896551724e-06,
+      "loss": 0.8647,
+      "step": 139
+    },
+    {
+      "epoch": 0.2896551724137931,
+      "grad_norm": 0.9084100127220154,
+      "learning_rate": 9.655172413793105e-06,
+      "loss": 0.8342,
+      "step": 140
+    },
+    {
+      "epoch": 0.29172413793103447,
+      "grad_norm": 0.9144700765609741,
+      "learning_rate": 9.724137931034484e-06,
+      "loss": 0.8468,
+      "step": 141
+    },
+    {
+      "epoch": 0.29379310344827586,
+      "grad_norm": 0.9257712364196777,
+      "learning_rate": 9.793103448275863e-06,
+      "loss": 0.8192,
+      "step": 142
+    },
+    {
+      "epoch": 0.29586206896551726,
+      "grad_norm": 0.8467687368392944,
+      "learning_rate": 9.862068965517241e-06,
+      "loss": 0.827,
+      "step": 143
+    },
+    {
+      "epoch": 0.2979310344827586,
+      "grad_norm": 0.925663411617279,
+      "learning_rate": 9.931034482758622e-06,
+      "loss": 0.8436,
+      "step": 144
+    },
+    {
+      "epoch": 0.3,
+      "grad_norm": 0.9307589530944824,
+      "learning_rate": 1e-05,
+      "loss": 0.7953,
+      "step": 145
+    },
+    {
+      "epoch": 0.3020689655172414,
+      "grad_norm": 0.9580163955688477,
+      "learning_rate": 9.999985489433723e-06,
+      "loss": 0.829,
+      "step": 146
+    },
+    {
+      "epoch": 0.3041379310344828,
+      "grad_norm": 0.8726505637168884,
+      "learning_rate": 9.999941957819112e-06,
+      "loss": 0.8571,
+      "step": 147
+    },
+    {
+      "epoch": 0.3062068965517241,
+      "grad_norm": 0.934336245059967,
+      "learning_rate": 9.999869405408833e-06,
+      "loss": 0.8127,
+      "step": 148
+    },
+    {
+      "epoch": 0.3082758620689655,
+      "grad_norm": 1.0289698839187622,
+      "learning_rate": 9.999767832624e-06,
+      "loss": 0.8592,
+      "step": 149
+    },
+    {
+      "epoch": 0.3103448275862069,
+      "grad_norm": 0.8795868754386902,
+      "learning_rate": 9.999637240054163e-06,
+      "loss": 0.8207,
+      "step": 150
+    },
+    {
+      "epoch": 0.3124137931034483,
+      "grad_norm": 0.8612985610961914,
+      "learning_rate": 9.99947762845731e-06,
+      "loss": 0.8548,
+      "step": 151
+    },
+    {
+      "epoch": 0.31448275862068964,
+      "grad_norm": 0.90495365858078,
+      "learning_rate": 9.999288998759865e-06,
+      "loss": 0.8346,
+      "step": 152
+    },
+    {
+      "epoch": 0.31655172413793103,
+      "grad_norm": 1.010977029800415,
+      "learning_rate": 9.999071352056676e-06,
+      "loss": 0.8504,
+      "step": 153
+    },
+    {
+      "epoch": 0.31862068965517243,
+      "grad_norm": 0.9661422967910767,
+      "learning_rate": 9.998824689611012e-06,
+      "loss": 0.8264,
+      "step": 154
+    },
+    {
+      "epoch": 0.32068965517241377,
+      "grad_norm": 1.0099589824676514,
+      "learning_rate": 9.998549012854562e-06,
+      "loss": 0.8054,
+      "step": 155
+    },
+    {
+      "epoch": 0.32275862068965516,
+      "grad_norm": 0.9441735744476318,
+      "learning_rate": 9.998244323387412e-06,
+      "loss": 0.803,
+      "step": 156
+    },
+    {
+      "epoch": 0.32482758620689656,
+      "grad_norm": 0.9308339953422546,
+      "learning_rate": 9.99791062297805e-06,
+      "loss": 0.8267,
+      "step": 157
+    },
+    {
+      "epoch": 0.32689655172413795,
+      "grad_norm": 0.9326705932617188,
+      "learning_rate": 9.997547913563352e-06,
+      "loss": 0.8291,
+      "step": 158
+    },
+    {
+      "epoch": 0.3289655172413793,
+      "grad_norm": 0.8982317447662354,
+      "learning_rate": 9.997156197248564e-06,
+      "loss": 0.837,
+      "step": 159
+    },
+    {
+      "epoch": 0.3310344827586207,
+      "grad_norm": 0.93187016248703,
+      "learning_rate": 9.996735476307292e-06,
+      "loss": 0.812,
+      "step": 160
+    },
+    {
+      "epoch": 0.3331034482758621,
+      "grad_norm": 0.9523342251777649,
+      "learning_rate": 9.996285753181499e-06,
+      "loss": 0.8545,
+      "step": 161
+    },
+    {
+      "epoch": 0.3351724137931035,
+      "grad_norm": 0.8485278487205505,
+      "learning_rate": 9.995807030481481e-06,
+      "loss": 0.8405,
+      "step": 162
+    },
+    {
+      "epoch": 0.3372413793103448,
+      "grad_norm": 0.9064662456512451,
+      "learning_rate": 9.99529931098585e-06,
+      "loss": 0.8164,
+      "step": 163
+    },
+    {
+      "epoch": 0.3393103448275862,
+      "grad_norm": 0.8660606145858765,
+      "learning_rate": 9.99476259764153e-06,
+      "loss": 0.816,
+      "step": 164
+    },
+    {
+      "epoch": 0.3413793103448276,
+      "grad_norm": 0.9558955430984497,
+      "learning_rate": 9.994196893563722e-06,
+      "loss": 0.8365,
+      "step": 165
+    },
+    {
+      "epoch": 0.34344827586206894,
+      "grad_norm": 0.838666558265686,
+      "learning_rate": 9.993602202035902e-06,
+      "loss": 0.7915,
+      "step": 166
+    },
+    {
+      "epoch": 0.34551724137931034,
+      "grad_norm": 0.8363468050956726,
+      "learning_rate": 9.992978526509794e-06,
+      "loss": 0.7789,
+      "step": 167
+    },
+    {
+      "epoch": 0.34758620689655173,
+      "grad_norm": 0.8468877077102661,
+      "learning_rate": 9.992325870605353e-06,
+      "loss": 0.8306,
+      "step": 168
+    },
+    {
+      "epoch": 0.3496551724137931,
+      "grad_norm": 0.8892214298248291,
+      "learning_rate": 9.991644238110741e-06,
+      "loss": 0.8433,
+      "step": 169
+    },
+    {
+      "epoch": 0.35172413793103446,
+      "grad_norm": 0.9307257533073425,
+      "learning_rate": 9.990933632982307e-06,
+      "loss": 0.8402,
+      "step": 170
+    },
+    {
+      "epoch": 0.35379310344827586,
+      "grad_norm": 0.8565515279769897,
+      "learning_rate": 9.990194059344565e-06,
+      "loss": 0.8403,
+      "step": 171
+    },
+    {
+      "epoch": 0.35586206896551725,
+      "grad_norm": 0.89169842004776,
+      "learning_rate": 9.989425521490169e-06,
+      "loss": 0.8387,
+      "step": 172
+    },
+    {
+      "epoch": 0.35793103448275865,
+      "grad_norm": 0.9352335333824158,
+      "learning_rate": 9.988628023879883e-06,
+      "loss": 0.8292,
+      "step": 173
+    },
+    {
+      "epoch": 0.36,
+      "grad_norm": 0.8871222734451294,
+      "learning_rate": 9.987801571142569e-06,
+      "loss": 0.8245,
+      "step": 174
+    },
+    {
+      "epoch": 0.3620689655172414,
+      "grad_norm": 0.7852333188056946,
+      "learning_rate": 9.98694616807514e-06,
+      "loss": 0.8523,
+      "step": 175
+    },
+    {
+      "epoch": 0.3641379310344828,
+      "grad_norm": 0.763839840888977,
+      "learning_rate": 9.986061819642554e-06,
+      "loss": 0.8027,
+      "step": 176
+    },
+    {
+      "epoch": 0.3662068965517241,
+      "grad_norm": 0.9269424080848694,
+      "learning_rate": 9.985148530977767e-06,
+      "loss": 0.8507,
+      "step": 177
+    },
+    {
+      "epoch": 0.3682758620689655,
+      "grad_norm": 0.9032247066497803,
+      "learning_rate": 9.984206307381714e-06,
+      "loss": 0.82,
+      "step": 178
+    },
+    {
+      "epoch": 0.3703448275862069,
+      "grad_norm": 0.8867712020874023,
+      "learning_rate": 9.983235154323276e-06,
+      "loss": 0.8395,
+      "step": 179
+    },
+    {
+      "epoch": 0.3724137931034483,
+      "grad_norm": 0.8791627883911133,
+      "learning_rate": 9.982235077439242e-06,
+      "loss": 0.8677,
+      "step": 180
+    },
+    {
+      "epoch": 0.37448275862068964,
+      "grad_norm": 0.9103829264640808,
+      "learning_rate": 9.981206082534287e-06,
+      "loss": 0.8136,
+      "step": 181
+    },
+    {
+      "epoch": 0.37655172413793103,
+      "grad_norm": 0.8131744265556335,
+      "learning_rate": 9.98014817558093e-06,
+      "loss": 0.835,
+      "step": 182
+    },
+    {
+      "epoch": 0.3786206896551724,
+      "grad_norm": 0.8920416235923767,
+      "learning_rate": 9.979061362719502e-06,
+      "loss": 0.7958,
+      "step": 183
+    },
+    {
+      "epoch": 0.38068965517241377,
+      "grad_norm": 0.8333418369293213,
+      "learning_rate": 9.977945650258111e-06,
+      "loss": 0.8238,
+      "step": 184
+    },
+    {
+      "epoch": 0.38275862068965516,
+      "grad_norm": 0.8242124915122986,
+      "learning_rate": 9.976801044672608e-06,
+      "loss": 0.8412,
+      "step": 185
+    },
+    {
+      "epoch": 0.38482758620689655,
+      "grad_norm": 0.9200892448425293,
+      "learning_rate": 9.975627552606538e-06,
+      "loss": 0.8429,
+      "step": 186
+    },
+    {
+      "epoch": 0.38689655172413795,
+      "grad_norm": 0.8656523823738098,
+      "learning_rate": 9.974425180871118e-06,
+      "loss": 0.8268,
+      "step": 187
+    },
+    {
+      "epoch": 0.3889655172413793,
+      "grad_norm": 0.9208945631980896,
+      "learning_rate": 9.973193936445185e-06,
+      "loss": 0.8338,
+      "step": 188
+    },
+    {
+      "epoch": 0.3910344827586207,
+      "grad_norm": 0.873414158821106,
+      "learning_rate": 9.971933826475162e-06,
+      "loss": 0.8236,
+      "step": 189
+    },
+    {
+      "epoch": 0.3931034482758621,
+      "grad_norm": 0.9169439673423767,
+      "learning_rate": 9.970644858275008e-06,
+      "loss": 0.8242,
+      "step": 190
+    },
+    {
+      "epoch": 0.39517241379310347,
+      "grad_norm": 0.8280132412910461,
+      "learning_rate": 9.969327039326193e-06,
+      "loss": 0.8269,
+      "step": 191
+    },
+    {
+      "epoch": 0.3972413793103448,
+      "grad_norm": 0.8578395247459412,
+      "learning_rate": 9.967980377277632e-06,
+      "loss": 0.7664,
+      "step": 192
+    },
+    {
+      "epoch": 0.3993103448275862,
+      "grad_norm": 0.8905231952667236,
+      "learning_rate": 9.966604879945659e-06,
+      "loss": 0.8239,
+      "step": 193
+    },
+    {
+      "epoch": 0.4013793103448276,
+      "grad_norm": 0.8388824462890625,
+      "learning_rate": 9.96520055531397e-06,
+      "loss": 0.8128,
+      "step": 194
+    },
+    {
+      "epoch": 0.40344827586206894,
+      "grad_norm": 0.7973628640174866,
+      "learning_rate": 9.963767411533584e-06,
+      "loss": 0.7949,
+      "step": 195
+    },
+    {
+      "epoch": 0.40551724137931033,
+      "grad_norm": 0.8202701807022095,
+      "learning_rate": 9.962305456922792e-06,
+      "loss": 0.8213,
+      "step": 196
+    },
+    {
+      "epoch": 0.40758620689655173,
+      "grad_norm": 0.7976661324501038,
+      "learning_rate": 9.960814699967112e-06,
+      "loss": 0.7915,
+      "step": 197
+    },
+    {
+      "epoch": 0.4096551724137931,
+      "grad_norm": 0.7930517196655273,
+      "learning_rate": 9.959295149319233e-06,
+      "loss": 0.805,
+      "step": 198
+    },
+    {
+      "epoch": 0.41172413793103446,
+      "grad_norm": 0.809988260269165,
+      "learning_rate": 9.95774681379897e-06,
+      "loss": 0.7896,
+      "step": 199
+    },
+    {
+      "epoch": 0.41379310344827586,
+      "grad_norm": 0.8766646981239319,
+      "learning_rate": 9.956169702393215e-06,
+      "loss": 0.8189,
+      "step": 200
+    },
+    {
+      "epoch": 0.41586206896551725,
+      "grad_norm": 0.9457526803016663,
+      "learning_rate": 9.954563824255879e-06,
+      "loss": 0.8189,
+      "step": 201
+    },
+    {
+      "epoch": 0.41793103448275865,
+      "grad_norm": 0.9090128540992737,
+      "learning_rate": 9.952929188707843e-06,
+      "loss": 0.8614,
+      "step": 202
+    },
+    {
+      "epoch": 0.42,
+      "grad_norm": 0.8124152421951294,
+      "learning_rate": 9.951265805236903e-06,
+      "loss": 0.8186,
+      "step": 203
+    },
+    {
+      "epoch": 0.4220689655172414,
+      "grad_norm": 0.8770976066589355,
+      "learning_rate": 9.94957368349771e-06,
+      "loss": 0.7963,
+      "step": 204
+    },
+    {
+      "epoch": 0.4241379310344828,
+      "grad_norm": 0.8936490416526794,
+      "learning_rate": 9.947852833311725e-06,
+      "loss": 0.8278,
+      "step": 205
+    },
+    {
+      "epoch": 0.4262068965517241,
+      "grad_norm": 0.9065835475921631,
+      "learning_rate": 9.946103264667151e-06,
+      "loss": 0.8026,
+      "step": 206
+    },
+    {
+      "epoch": 0.4282758620689655,
+      "grad_norm": 0.755096435546875,
+      "learning_rate": 9.94432498771888e-06,
+      "loss": 0.7987,
+      "step": 207
+    },
+    {
+      "epoch": 0.4303448275862069,
+      "grad_norm": 0.8373960256576538,
+      "learning_rate": 9.942518012788438e-06,
+      "loss": 0.7967,
+      "step": 208
+    },
+    {
+      "epoch": 0.4324137931034483,
+      "grad_norm": 0.8678641319274902,
+      "learning_rate": 9.940682350363913e-06,
+      "loss": 0.8212,
+      "step": 209
+    },
+    {
+      "epoch": 0.43448275862068964,
+      "grad_norm": 0.8203259706497192,
+      "learning_rate": 9.938818011099906e-06,
+      "loss": 0.7922,
+      "step": 210
+    },
+    {
+      "epoch": 0.43655172413793103,
+      "grad_norm": 0.8179094195365906,
+      "learning_rate": 9.936925005817464e-06,
+      "loss": 0.8271,
+      "step": 211
+    },
+    {
+      "epoch": 0.4386206896551724,
+      "grad_norm": 0.8908324241638184,
+      "learning_rate": 9.93500334550402e-06,
+      "loss": 0.7867,
+      "step": 212
+    },
+    {
+      "epoch": 0.4406896551724138,
+      "grad_norm": 0.8492280840873718,
+      "learning_rate": 9.933053041313325e-06,
+      "loss": 0.8225,
+      "step": 213
+    },
+    {
+      "epoch": 0.44275862068965516,
+      "grad_norm": 0.8480848670005798,
+      "learning_rate": 9.931074104565387e-06,
+      "loss": 0.8418,
+      "step": 214
+    },
+    {
+      "epoch": 0.44482758620689655,
+      "grad_norm": 0.857255756855011,
+      "learning_rate": 9.929066546746403e-06,
+      "loss": 0.8297,
+      "step": 215
+    },
+    {
+      "epoch": 0.44689655172413795,
+      "grad_norm": 0.9532951712608337,
+      "learning_rate": 9.927030379508691e-06,
+      "loss": 0.8094,
+      "step": 216
+    },
+    {
+      "epoch": 0.4489655172413793,
+      "grad_norm": 0.8437204360961914,
+      "learning_rate": 9.924965614670629e-06,
+      "loss": 0.7996,
+      "step": 217
+    },
+    {
+      "epoch": 0.4510344827586207,
+      "grad_norm": 0.7982957363128662,
+      "learning_rate": 9.922872264216581e-06,
+      "loss": 0.7962,
+      "step": 218
+    },
+    {
+      "epoch": 0.4531034482758621,
+      "grad_norm": 0.8589974045753479,
+      "learning_rate": 9.920750340296823e-06,
+      "loss": 0.8311,
+      "step": 219
+    },
+    {
+      "epoch": 0.45517241379310347,
+      "grad_norm": 0.8524751663208008,
+      "learning_rate": 9.918599855227487e-06,
+      "loss": 0.8113,
+      "step": 220
+    },
+    {
+      "epoch": 0.4572413793103448,
+      "grad_norm": 0.8594533205032349,
+      "learning_rate": 9.916420821490474e-06,
+      "loss": 0.8201,
+      "step": 221
+    },
+    {
+      "epoch": 0.4593103448275862,
+      "grad_norm": 0.872061550617218,
+      "learning_rate": 9.914213251733387e-06,
+      "loss": 0.8014,
+      "step": 222
+    },
+    {
+      "epoch": 0.4613793103448276,
+      "grad_norm": 0.8578385710716248,
+      "learning_rate": 9.911977158769461e-06,
+      "loss": 0.787,
+      "step": 223
+    },
+    {
+      "epoch": 0.463448275862069,
+      "grad_norm": 0.9662230610847473,
+      "learning_rate": 9.909712555577491e-06,
+      "loss": 0.7924,
+      "step": 224
+    },
+    {
+      "epoch": 0.46551724137931033,
+      "grad_norm": 0.909537672996521,
+      "learning_rate": 9.90741945530174e-06,
+      "loss": 0.8267,
+      "step": 225
+    },
+    {
+      "epoch": 0.4675862068965517,
+      "grad_norm": 0.8204202651977539,
+      "learning_rate": 9.905097871251887e-06,
+      "loss": 0.8142,
+      "step": 226
+    },
+    {
+      "epoch": 0.4696551724137931,
+      "grad_norm": 0.8849340081214905,
+      "learning_rate": 9.90274781690293e-06,
+      "loss": 0.8364,
+      "step": 227
+    },
+    {
+      "epoch": 0.47172413793103446,
+      "grad_norm": 0.825469970703125,
+      "learning_rate": 9.900369305895115e-06,
+      "loss": 0.7874,
+      "step": 228
+    },
+    {
+      "epoch": 0.47379310344827585,
+      "grad_norm": 0.799996018409729,
+      "learning_rate": 9.89796235203386e-06,
+      "loss": 0.8044,
+      "step": 229
+    },
+    {
+      "epoch": 0.47586206896551725,
+      "grad_norm": 0.8811237812042236,
+      "learning_rate": 9.895526969289671e-06,
+      "loss": 0.8027,
+      "step": 230
+    },
+    {
+      "epoch": 0.47793103448275864,
+      "grad_norm": 0.9074158072471619,
+      "learning_rate": 9.89306317179806e-06,
+      "loss": 0.8172,
+      "step": 231
+    },
+    {
+      "epoch": 0.48,
+      "grad_norm": 0.7685100436210632,
+      "learning_rate": 9.890570973859468e-06,
+      "loss": 0.8144,
+      "step": 232
+    },
+    {
+      "epoch": 0.4820689655172414,
+      "grad_norm": 0.8316488265991211,
+      "learning_rate": 9.888050389939172e-06,
+      "loss": 0.8188,
+      "step": 233
+    },
+    {
+      "epoch": 0.48413793103448277,
+      "grad_norm": 0.8519396185874939,
+      "learning_rate": 9.885501434667215e-06,
+      "loss": 0.8242,
+      "step": 234
+    },
+    {
+      "epoch": 0.4862068965517241,
+      "grad_norm": 0.8774333000183105,
+      "learning_rate": 9.882924122838311e-06,
+      "loss": 0.8083,
+      "step": 235
+    },
+    {
+      "epoch": 0.4882758620689655,
+      "grad_norm": 0.8599486351013184,
+      "learning_rate": 9.88031846941176e-06,
+      "loss": 0.8024,
+      "step": 236
+    },
+    {
+      "epoch": 0.4903448275862069,
+      "grad_norm": 0.8138340711593628,
+      "learning_rate": 9.877684489511367e-06,
+      "loss": 0.8242,
+      "step": 237
+    },
+    {
+      "epoch": 0.4924137931034483,
+      "grad_norm": 0.9175668954849243,
+      "learning_rate": 9.875022198425347e-06,
+      "loss": 0.816,
+      "step": 238
+    },
+    {
+      "epoch": 0.49448275862068963,
+      "grad_norm": 0.8907325267791748,
+      "learning_rate": 9.872331611606238e-06,
+      "loss": 0.7923,
+      "step": 239
+    },
+    {
+      "epoch": 0.496551724137931,
+      "grad_norm": 0.8826651573181152,
+      "learning_rate": 9.869612744670817e-06,
+      "loss": 0.8359,
+      "step": 240
+    },
+    {
+      "epoch": 0.4986206896551724,
+      "grad_norm": 0.866460919380188,
+      "learning_rate": 9.866865613400008e-06,
+      "loss": 0.7803,
+      "step": 241
+    },
+    {
+      "epoch": 0.5006896551724138,
+      "grad_norm": 0.9463049173355103,
+      "learning_rate": 9.864090233738776e-06,
+      "loss": 0.8221,
+      "step": 242
+    },
+    {
+      "epoch": 0.5027586206896552,
+      "grad_norm": 0.8751413226127625,
+      "learning_rate": 9.861286621796056e-06,
+      "loss": 0.8299,
+      "step": 243
+    },
+    {
+      "epoch": 0.5048275862068966,
+      "grad_norm": 0.9759247303009033,
+      "learning_rate": 9.858454793844648e-06,
+      "loss": 0.8147,
+      "step": 244
+    },
+    {
+      "epoch": 0.506896551724138,
+      "grad_norm": 0.899433434009552,
+      "learning_rate": 9.855594766321122e-06,
+      "loss": 0.8172,
+      "step": 245
+    },
+    {
+      "epoch": 0.5089655172413793,
+      "grad_norm": 0.8528696298599243,
+      "learning_rate": 9.852706555825727e-06,
+      "loss": 0.7903,
+      "step": 246
+    },
+    {
+      "epoch": 0.5110344827586207,
+      "grad_norm": 0.9641590714454651,
+      "learning_rate": 9.849790179122287e-06,
+      "loss": 0.7741,
+      "step": 247
+    },
+    {
+      "epoch": 0.5131034482758621,
+      "grad_norm": 0.8884571194648743,
+      "learning_rate": 9.846845653138117e-06,
+      "loss": 0.8316,
+      "step": 248
+    },
+    {
+      "epoch": 0.5151724137931034,
+      "grad_norm": 0.8643737435340881,
+      "learning_rate": 9.843872994963912e-06,
+      "loss": 0.7956,
+      "step": 249
+    },
+    {
+      "epoch": 0.5172413793103449,
+      "grad_norm": 0.9500787854194641,
+      "learning_rate": 9.84087222185365e-06,
+      "loss": 0.8379,
+      "step": 250
+    },
+    {
+      "epoch": 0.5193103448275862,
+      "grad_norm": 0.8723439574241638,
+      "learning_rate": 9.837843351224504e-06,
+      "loss": 0.8169,
+      "step": 251
+    },
+    {
+      "epoch": 0.5213793103448275,
+      "grad_norm": 0.9550707936286926,
+      "learning_rate": 9.834786400656719e-06,
+      "loss": 0.8016,
+      "step": 252
+    },
+    {
+      "epoch": 0.523448275862069,
+      "grad_norm": 0.9410514831542969,
+      "learning_rate": 9.831701387893533e-06,
+      "loss": 0.7718,
+      "step": 253
+    },
+    {
+      "epoch": 0.5255172413793103,
+      "grad_norm": 0.8164445757865906,
+      "learning_rate": 9.828588330841055e-06,
+      "loss": 0.8035,
+      "step": 254
+    },
+    {
+      "epoch": 0.5275862068965518,
+      "grad_norm": 0.8466964364051819,
+      "learning_rate": 9.825447247568177e-06,
+      "loss": 0.8199,
+      "step": 255
+    },
+    {
+      "epoch": 0.5296551724137931,
+      "grad_norm": 1.039279818534851,
+      "learning_rate": 9.822278156306455e-06,
+      "loss": 0.8012,
+      "step": 256
+    },
+    {
+      "epoch": 0.5317241379310345,
+      "grad_norm": 0.9037715196609497,
+      "learning_rate": 9.819081075450014e-06,
+      "loss": 0.8094,
+      "step": 257
+    },
+    {
+      "epoch": 0.5337931034482759,
+      "grad_norm": 0.8636988401412964,
+      "learning_rate": 9.815856023555437e-06,
+      "loss": 0.8395,
+      "step": 258
+    },
+    {
+      "epoch": 0.5358620689655172,
+      "grad_norm": 0.883048415184021,
+      "learning_rate": 9.812603019341652e-06,
+      "loss": 0.7844,
+      "step": 259
+    },
+    {
+      "epoch": 0.5379310344827586,
+      "grad_norm": 0.861579954624176,
+      "learning_rate": 9.809322081689834e-06,
+      "loss": 0.8218,
+      "step": 260
+    },
+    {
+      "epoch": 0.54,
+      "grad_norm": 0.9384899139404297,
+      "learning_rate": 9.80601322964329e-06,
+      "loss": 0.802,
+      "step": 261
+    },
+    {
+      "epoch": 0.5420689655172414,
+      "grad_norm": 0.8394620418548584,
+      "learning_rate": 9.802676482407343e-06,
+      "loss": 0.7974,
+      "step": 262
+    },
+    {
+      "epoch": 0.5441379310344827,
+      "grad_norm": 0.8985764384269714,
+      "learning_rate": 9.799311859349235e-06,
+      "loss": 0.8001,
+      "step": 263
+    },
+    {
+      "epoch": 0.5462068965517242,
+      "grad_norm": 0.8577210307121277,
+      "learning_rate": 9.795919379997993e-06,
+      "loss": 0.8312,
+      "step": 264
+    },
+    {
+      "epoch": 0.5482758620689655,
+      "grad_norm": 0.8467931151390076,
+      "learning_rate": 9.792499064044343e-06,
+      "loss": 0.8007,
+      "step": 265
+    },
+    {
+      "epoch": 0.5503448275862068,
+      "grad_norm": 0.9455611705780029,
+      "learning_rate": 9.789050931340572e-06,
+      "loss": 0.8143,
+      "step": 266
+    },
+    {
+      "epoch": 0.5524137931034483,
+      "grad_norm": 0.8457427620887756,
+      "learning_rate": 9.785575001900418e-06,
+      "loss": 0.7889,
+      "step": 267
+    },
+    {
+      "epoch": 0.5544827586206896,
+      "grad_norm": 0.8870904445648193,
+      "learning_rate": 9.78207129589897e-06,
+      "loss": 0.823,
+      "step": 268
+    },
+    {
+      "epoch": 0.5565517241379311,
+      "grad_norm": 0.8103102445602417,
+      "learning_rate": 9.778539833672525e-06,
+      "loss": 0.7962,
+      "step": 269
+    },
+    {
+      "epoch": 0.5586206896551724,
+      "grad_norm": 0.8109989166259766,
+      "learning_rate": 9.774980635718494e-06,
+      "loss": 0.8107,
+      "step": 270
+    },
+    {
+      "epoch": 0.5606896551724138,
+      "grad_norm": 0.8663826584815979,
+      "learning_rate": 9.771393722695267e-06,
+      "loss": 0.8158,
+      "step": 271
+    },
+    {
+      "epoch": 0.5627586206896552,
+      "grad_norm": 1.5541844367980957,
+      "learning_rate": 9.767779115422097e-06,
+      "loss": 0.7898,
+      "step": 272
+    },
+    {
+      "epoch": 0.5648275862068965,
+      "grad_norm": 0.8373553156852722,
+      "learning_rate": 9.764136834878987e-06,
+      "loss": 0.7963,
+      "step": 273
+    },
+    {
+      "epoch": 0.5668965517241379,
+      "grad_norm": 0.7791042923927307,
+      "learning_rate": 9.760466902206557e-06,
+      "loss": 0.8085,
+      "step": 274
+    },
+    {
+      "epoch": 0.5689655172413793,
+      "grad_norm": 0.8393211960792542,
+      "learning_rate": 9.756769338705927e-06,
+      "loss": 0.8136,
+      "step": 275
+    },
+    {
+      "epoch": 0.5710344827586207,
+      "grad_norm": 0.811062216758728,
+      "learning_rate": 9.753044165838594e-06,
+      "loss": 0.8075,
+      "step": 276
+    },
+    {
+      "epoch": 0.573103448275862,
+      "grad_norm": 0.8894556760787964,
+      "learning_rate": 9.749291405226304e-06,
+      "loss": 0.8371,
+      "step": 277
+    },
+    {
+      "epoch": 0.5751724137931035,
+      "grad_norm": 0.8580628037452698,
+      "learning_rate": 9.745511078650931e-06,
+      "loss": 0.7695,
+      "step": 278
+    },
+    {
+      "epoch": 0.5772413793103448,
+      "grad_norm": 0.7854070067405701,
+      "learning_rate": 9.741703208054347e-06,
+      "loss": 0.8076,
+      "step": 279
+    },
+    {
+      "epoch": 0.5793103448275863,
+      "grad_norm": 0.8000266551971436,
+      "learning_rate": 9.737867815538292e-06,
+      "loss": 0.7894,
+      "step": 280
+    },
+    {
+      "epoch": 0.5813793103448276,
+      "grad_norm": 0.8094033002853394,
+      "learning_rate": 9.734004923364258e-06,
+      "loss": 0.8035,
+      "step": 281
+    },
+    {
+      "epoch": 0.5834482758620689,
+      "grad_norm": 0.8682005405426025,
+      "learning_rate": 9.73011455395334e-06,
+      "loss": 0.8219,
+      "step": 282
+    },
+    {
+      "epoch": 0.5855172413793104,
+      "grad_norm": 0.8397322297096252,
+      "learning_rate": 9.72619672988613e-06,
+      "loss": 0.8241,
+      "step": 283
+    },
+    {
+      "epoch": 0.5875862068965517,
+      "grad_norm": 0.8110212087631226,
+      "learning_rate": 9.722251473902562e-06,
+      "loss": 0.8099,
+      "step": 284
+    },
+    {
+      "epoch": 0.5896551724137931,
+      "grad_norm": 0.8143454790115356,
+      "learning_rate": 9.718278808901797e-06,
+      "loss": 0.7744,
+      "step": 285
+    },
+    {
+      "epoch": 0.5917241379310345,
+      "grad_norm": 0.8721611499786377,
+      "learning_rate": 9.714278757942082e-06,
+      "loss": 0.785,
+      "step": 286
+    },
+    {
+      "epoch": 0.5937931034482758,
+      "grad_norm": 0.7908191084861755,
+      "learning_rate": 9.710251344240619e-06,
+      "loss": 0.7983,
+      "step": 287
+    },
+    {
+      "epoch": 0.5958620689655172,
+      "grad_norm": 0.8809441924095154,
+      "learning_rate": 9.70619659117343e-06,
+      "loss": 0.8428,
+      "step": 288
+    },
+    {
+      "epoch": 0.5979310344827586,
+      "grad_norm": 0.8129044771194458,
+      "learning_rate": 9.702114522275216e-06,
+      "loss": 0.8149,
+      "step": 289
+    },
+    {
+      "epoch": 0.6,
+      "grad_norm": 0.826924204826355,
+      "learning_rate": 9.698005161239237e-06,
+      "loss": 0.8091,
+      "step": 290
+    },
+    {
+      "epoch": 0.6020689655172414,
+      "grad_norm": 0.8344675302505493,
+      "learning_rate": 9.69386853191715e-06,
+      "loss": 0.7913,
+      "step": 291
+    },
+    {
+      "epoch": 0.6041379310344828,
+      "grad_norm": 0.8673633933067322,
+      "learning_rate": 9.689704658318888e-06,
+      "loss": 0.8,
+      "step": 292
+    },
+    {
+      "epoch": 0.6062068965517241,
+      "grad_norm": 0.7886486053466797,
+      "learning_rate": 9.685513564612521e-06,
+      "loss": 0.7967,
+      "step": 293
+    },
+    {
+      "epoch": 0.6082758620689656,
+      "grad_norm": 0.8335657715797424,
+      "learning_rate": 9.681295275124104e-06,
+      "loss": 0.7843,
+      "step": 294
+    },
+    {
+      "epoch": 0.6103448275862069,
+      "grad_norm": 0.934149980545044,
+      "learning_rate": 9.67704981433754e-06,
+      "loss": 0.7967,
+      "step": 295
+    },
+    {
+      "epoch": 0.6124137931034482,
+      "grad_norm": 0.8753807544708252,
+      "learning_rate": 9.672777206894452e-06,
+      "loss": 0.8086,
+      "step": 296
+    },
+    {
+      "epoch": 0.6144827586206897,
+      "grad_norm": 0.8328104615211487,
+      "learning_rate": 9.668477477594021e-06,
+      "loss": 0.8006,
+      "step": 297
+    },
+    {
+      "epoch": 0.616551724137931,
+      "grad_norm": 0.8407275080680847,
+      "learning_rate": 9.664150651392846e-06,
+      "loss": 0.7968,
+      "step": 298
+    },
+    {
+      "epoch": 0.6186206896551724,
+      "grad_norm": 0.8628553748130798,
+      "learning_rate": 9.659796753404806e-06,
+      "loss": 0.8256,
+      "step": 299
+    },
+    {
+      "epoch": 0.6206896551724138,
+      "grad_norm": 0.7844734191894531,
+      "learning_rate": 9.655415808900915e-06,
+      "loss": 0.784,
+      "step": 300
+    },
+    {
+      "epoch": 0.6227586206896552,
+      "grad_norm": 0.8278802633285522,
+      "learning_rate": 9.651007843309164e-06,
+      "loss": 0.7548,
+      "step": 301
+    },
+    {
+      "epoch": 0.6248275862068966,
+      "grad_norm": 0.8672357201576233,
+      "learning_rate": 9.646572882214385e-06,
+      "loss": 0.8006,
+      "step": 302
+    },
+    {
+      "epoch": 0.6268965517241379,
+      "grad_norm": 0.8474316000938416,
+      "learning_rate": 9.642110951358097e-06,
+      "loss": 0.8056,
+      "step": 303
+    },
+    {
+      "epoch": 0.6289655172413793,
+      "grad_norm": 0.8918145895004272,
+      "learning_rate": 9.637622076638358e-06,
+      "loss": 0.8086,
+      "step": 304
+    },
+    {
+      "epoch": 0.6310344827586207,
+      "grad_norm": 0.7701090574264526,
+      "learning_rate": 9.633106284109612e-06,
+      "loss": 0.7971,
+      "step": 305
+    },
+    {
+      "epoch": 0.6331034482758621,
+      "grad_norm": 0.7959749102592468,
+      "learning_rate": 9.62856359998254e-06,
+      "loss": 0.7914,
+      "step": 306
+    },
+    {
+      "epoch": 0.6351724137931034,
+      "grad_norm": 0.8288137912750244,
+      "learning_rate": 9.623994050623917e-06,
+      "loss": 0.8074,
+      "step": 307
+    },
+    {
+      "epoch": 0.6372413793103449,
+      "grad_norm": 0.7820798754692078,
+      "learning_rate": 9.619397662556434e-06,
+      "loss": 0.764,
+      "step": 308
+    },
+    {
+      "epoch": 0.6393103448275862,
+      "grad_norm": 0.8969067335128784,
+      "learning_rate": 9.614774462458573e-06,
+      "loss": 0.8464,
+      "step": 309
+    },
+    {
+      "epoch": 0.6413793103448275,
+      "grad_norm": 0.8294548988342285,
+      "learning_rate": 9.610124477164436e-06,
+      "loss": 0.8576,
+      "step": 310
+    },
+    {
+      "epoch": 0.643448275862069,
+      "grad_norm": 0.8388926982879639,
+      "learning_rate": 9.605447733663586e-06,
+      "loss": 0.7603,
+      "step": 311
+    },
+    {
+      "epoch": 0.6455172413793103,
+      "grad_norm": 0.8787766098976135,
+      "learning_rate": 9.600744259100905e-06,
+      "loss": 0.7991,
+      "step": 312
+    },
+    {
+      "epoch": 0.6475862068965518,
+      "grad_norm": 0.8679440021514893,
+      "learning_rate": 9.596014080776424e-06,
+      "loss": 0.7977,
+      "step": 313
+    },
+    {
+      "epoch": 0.6496551724137931,
+      "grad_norm": 0.8179478049278259,
+      "learning_rate": 9.59125722614517e-06,
+      "loss": 0.8009,
+      "step": 314
+    },
+    {
+      "epoch": 0.6517241379310345,
+      "grad_norm": 0.8390472531318665,
+      "learning_rate": 9.586473722817002e-06,
+      "loss": 0.8023,
+      "step": 315
+    },
+    {
+      "epoch": 0.6537931034482759,
+      "grad_norm": 0.9145618677139282,
+      "learning_rate": 9.581663598556461e-06,
+      "loss": 0.7888,
+      "step": 316
+    },
+    {
+      "epoch": 0.6558620689655172,
+      "grad_norm": 0.7960366010665894,
+      "learning_rate": 9.576826881282595e-06,
+      "loss": 0.8047,
+      "step": 317
+    },
+    {
+      "epoch": 0.6579310344827586,
+      "grad_norm": 0.8886500000953674,
+      "learning_rate": 9.571963599068807e-06,
+      "loss": 0.7772,
+      "step": 318
+    },
+    {
+      "epoch": 0.66,
+      "grad_norm": 0.8854326009750366,
+      "learning_rate": 9.56707378014269e-06,
+      "loss": 0.783,
+      "step": 319
+    },
+    {
+      "epoch": 0.6620689655172414,
+      "grad_norm": 0.7884778380393982,
+      "learning_rate": 9.56215745288586e-06,
+      "loss": 0.7783,
+      "step": 320
+    },
+    {
+      "epoch": 0.6641379310344827,
+      "grad_norm": 0.9254597425460815,
+      "learning_rate": 9.557214645833792e-06,
+      "loss": 0.7903,
+      "step": 321
+    },
+    {
+      "epoch": 0.6662068965517242,
+      "grad_norm": 1.0196994543075562,
+      "learning_rate": 9.55224538767566e-06,
+      "loss": 0.7955,
+      "step": 322
+    },
+    {
+      "epoch": 0.6682758620689655,
+      "grad_norm": 0.8419172167778015,
+      "learning_rate": 9.547249707254166e-06,
+      "loss": 0.8135,
+      "step": 323
+    },
+    {
+      "epoch": 0.670344827586207,
+      "grad_norm": 0.85199373960495,
+      "learning_rate": 9.542227633565364e-06,
+      "loss": 0.7882,
+      "step": 324
+    },
+    {
+      "epoch": 0.6724137931034483,
+      "grad_norm": 0.8458907008171082,
+      "learning_rate": 9.537179195758513e-06,
+      "loss": 0.7803,
+      "step": 325
+    },
+    {
+      "epoch": 0.6744827586206896,
+      "grad_norm": 0.8774712681770325,
+      "learning_rate": 9.532104423135885e-06,
+      "loss": 0.8011,
+      "step": 326
+    },
+    {
+      "epoch": 0.6765517241379311,
+      "grad_norm": 0.8695656657218933,
+      "learning_rate": 9.527003345152616e-06,
+      "loss": 0.792,
+      "step": 327
+    },
+    {
+      "epoch": 0.6786206896551724,
+      "grad_norm": 0.799929141998291,
+      "learning_rate": 9.521875991416514e-06,
+      "loss": 0.7651,
+      "step": 328
+    },
+    {
+      "epoch": 0.6806896551724138,
+      "grad_norm": 0.8962751626968384,
+      "learning_rate": 9.516722391687903e-06,
+      "loss": 0.7951,
+      "step": 329
+    },
+    {
+      "epoch": 0.6827586206896552,
+      "grad_norm": 0.8643319010734558,
+      "learning_rate": 9.51154257587944e-06,
+      "loss": 0.7967,
+      "step": 330
+    },
+    {
+      "epoch": 0.6848275862068965,
+      "grad_norm": 0.9759703874588013,
+      "learning_rate": 9.506336574055954e-06,
+      "loss": 0.7929,
+      "step": 331
+    },
+    {
+      "epoch": 0.6868965517241379,
+      "grad_norm": 0.8432797193527222,
+      "learning_rate": 9.501104416434255e-06,
+      "loss": 0.7779,
+      "step": 332
+    },
+    {
+      "epoch": 0.6889655172413793,
+      "grad_norm": 0.9403534531593323,
+      "learning_rate": 9.495846133382973e-06,
+      "loss": 0.8123,
+      "step": 333
+    },
+    {
+      "epoch": 0.6910344827586207,
+      "grad_norm": 0.8679063320159912,
+      "learning_rate": 9.490561755422374e-06,
+      "loss": 0.7462,
+      "step": 334
+    },
+    {
+      "epoch": 0.6931034482758621,
+      "grad_norm": 0.791249692440033,
+      "learning_rate": 9.485251313224181e-06,
+      "loss": 0.7982,
+      "step": 335
+    },
+    {
+      "epoch": 0.6951724137931035,
+      "grad_norm": 0.830558180809021,
+      "learning_rate": 9.479914837611409e-06,
+      "loss": 0.7568,
+      "step": 336
+    },
+    {
+      "epoch": 0.6972413793103448,
+      "grad_norm": 0.8038316369056702,
+      "learning_rate": 9.474552359558167e-06,
+      "loss": 0.7626,
+      "step": 337
+    },
+    {
+      "epoch": 0.6993103448275862,
+      "grad_norm": 0.7749502658843994,
+      "learning_rate": 9.469163910189494e-06,
+      "loss": 0.79,
+      "step": 338
+    },
+    {
+      "epoch": 0.7013793103448276,
+      "grad_norm": 0.9494272470474243,
+      "learning_rate": 9.463749520781172e-06,
+      "loss": 0.7803,
+      "step": 339
+    },
+    {
+      "epoch": 0.7034482758620689,
+      "grad_norm": 0.8260956406593323,
+      "learning_rate": 9.458309222759542e-06,
+      "loss": 0.7729,
+      "step": 340
+    },
+    {
+      "epoch": 0.7055172413793104,
+      "grad_norm": 0.8298391699790955,
+      "learning_rate": 9.452843047701324e-06,
+      "loss": 0.8141,
+      "step": 341
+    },
+    {
+      "epoch": 0.7075862068965517,
+      "grad_norm": 0.9029043912887573,
+      "learning_rate": 9.44735102733344e-06,
+      "loss": 0.797,
+      "step": 342
+    },
+    {
+      "epoch": 0.7096551724137931,
+      "grad_norm": 0.8290296196937561,
+      "learning_rate": 9.441833193532817e-06,
+      "loss": 0.7708,
+      "step": 343
+    },
+    {
+      "epoch": 0.7117241379310345,
+      "grad_norm": 0.7795094847679138,
+      "learning_rate": 9.436289578326215e-06,
+      "loss": 0.8196,
+      "step": 344
+    },
+    {
+      "epoch": 0.7137931034482758,
+      "grad_norm": 0.7970453500747681,
+      "learning_rate": 9.43072021389003e-06,
+      "loss": 0.822,
+      "step": 345
+    },
+    {
+      "epoch": 0.7158620689655173,
+      "grad_norm": 0.8111122250556946,
+      "learning_rate": 9.425125132550115e-06,
+      "loss": 0.771,
+      "step": 346
+    },
+    {
+      "epoch": 0.7179310344827586,
+      "grad_norm": 0.782451868057251,
+      "learning_rate": 9.419504366781591e-06,
+      "loss": 0.7865,
+      "step": 347
+    },
+    {
+      "epoch": 0.72,
+      "grad_norm": 0.7692107558250427,
+      "learning_rate": 9.413857949208655e-06,
+      "loss": 0.7748,
+      "step": 348
+    },
+    {
+      "epoch": 0.7220689655172414,
+      "grad_norm": 0.9149669408798218,
+      "learning_rate": 9.408185912604395e-06,
+      "loss": 0.7847,
+      "step": 349
+    },
+    {
+      "epoch": 0.7241379310344828,
+      "grad_norm": 0.812156081199646,
+      "learning_rate": 9.402488289890593e-06,
+      "loss": 0.7921,
+      "step": 350
+    },
+    {
+      "epoch": 0.7262068965517241,
+      "grad_norm": 0.8430172801017761,
+      "learning_rate": 9.396765114137544e-06,
+      "loss": 0.7752,
+      "step": 351
+    },
+    {
+      "epoch": 0.7282758620689656,
+      "grad_norm": 0.8315995931625366,
+      "learning_rate": 9.391016418563856e-06,
+      "loss": 0.7867,
+      "step": 352
+    },
+    {
+      "epoch": 0.7303448275862069,
+      "grad_norm": 0.8469542860984802,
+      "learning_rate": 9.385242236536259e-06,
+      "loss": 0.7573,
+      "step": 353
+    },
+    {
+      "epoch": 0.7324137931034482,
+      "grad_norm": 0.8568265438079834,
+      "learning_rate": 9.379442601569415e-06,
+      "loss": 0.7873,
+      "step": 354
+    },
+    {
+      "epoch": 0.7344827586206897,
+      "grad_norm": 0.8971392512321472,
+      "learning_rate": 9.37361754732572e-06,
+      "loss": 0.8196,
+      "step": 355
+    },
+    {
+      "epoch": 0.736551724137931,
+      "grad_norm": 0.8264282941818237,
+      "learning_rate": 9.367767107615106e-06,
+      "loss": 0.7785,
+      "step": 356
+    },
+    {
+      "epoch": 0.7386206896551725,
+      "grad_norm": 0.852152407169342,
+      "learning_rate": 9.36189131639485e-06,
+      "loss": 0.7655,
+      "step": 357
+    },
+    {
+      "epoch": 0.7406896551724138,
+      "grad_norm": 0.7565860152244568,
+      "learning_rate": 9.355990207769378e-06,
+      "loss": 0.7672,
+      "step": 358
+    },
+    {
+      "epoch": 0.7427586206896551,
+      "grad_norm": 0.8384962677955627,
+      "learning_rate": 9.35006381599006e-06,
+      "loss": 0.7768,
+      "step": 359
+    },
+    {
+      "epoch": 0.7448275862068966,
+      "grad_norm": 0.806664764881134,
+      "learning_rate": 9.344112175455013e-06,
+      "loss": 0.7716,
+      "step": 360
+    },
+    {
+      "epoch": 0.7468965517241379,
+      "grad_norm": 0.848978579044342,
+      "learning_rate": 9.338135320708912e-06,
+      "loss": 0.817,
+      "step": 361
+    },
+    {
+      "epoch": 0.7489655172413793,
+      "grad_norm": 0.8164881467819214,
+      "learning_rate": 9.332133286442772e-06,
+      "loss": 0.7967,
+      "step": 362
+    },
+    {
+      "epoch": 0.7510344827586207,
+      "grad_norm": 0.855451226234436,
+      "learning_rate": 9.326106107493762e-06,
+      "loss": 0.7519,
+      "step": 363
+    },
+    {
+      "epoch": 0.7531034482758621,
+      "grad_norm": 0.8127675652503967,
+      "learning_rate": 9.32005381884499e-06,
+      "loss": 0.758,
+      "step": 364
+    },
+    {
+      "epoch": 0.7551724137931034,
+      "grad_norm": 0.8386479020118713,
+      "learning_rate": 9.313976455625316e-06,
+      "loss": 0.7947,
+      "step": 365
+    },
+    {
+      "epoch": 0.7572413793103449,
+      "grad_norm": 0.9732878804206848,
+      "learning_rate": 9.307874053109127e-06,
+      "loss": 0.7968,
+      "step": 366
+    },
+    {
+      "epoch": 0.7593103448275862,
+      "grad_norm": 0.9192622900009155,
+      "learning_rate": 9.301746646716152e-06,
+      "loss": 0.7927,
+      "step": 367
+    },
+    {
+      "epoch": 0.7613793103448275,
+      "grad_norm": 0.8056217432022095,
+      "learning_rate": 9.295594272011246e-06,
+      "loss": 0.7892,
+      "step": 368
+    },
+    {
+      "epoch": 0.763448275862069,
+      "grad_norm": 0.8506579399108887,
+      "learning_rate": 9.289416964704186e-06,
+      "loss": 0.7746,
+      "step": 369
+    },
+    {
+      "epoch": 0.7655172413793103,
+      "grad_norm": 0.8824191689491272,
+      "learning_rate": 9.283214760649462e-06,
+      "loss": 0.7749,
+      "step": 370
+    },
+    {
+      "epoch": 0.7675862068965518,
+      "grad_norm": 0.894450306892395,
+      "learning_rate": 9.27698769584607e-06,
+      "loss": 0.7792,
+      "step": 371
+    },
+    {
+      "epoch": 0.7696551724137931,
+      "grad_norm": 0.8521785140037537,
+      "learning_rate": 9.270735806437306e-06,
+      "loss": 0.8031,
+      "step": 372
+    },
+    {
+      "epoch": 0.7717241379310344,
+      "grad_norm": 0.9518225789070129,
+      "learning_rate": 9.26445912871055e-06,
+      "loss": 0.8018,
+      "step": 373
+    },
+    {
+      "epoch": 0.7737931034482759,
+      "grad_norm": 0.82342129945755,
+      "learning_rate": 9.258157699097065e-06,
+      "loss": 0.7745,
+      "step": 374
+    },
+    {
+      "epoch": 0.7758620689655172,
+      "grad_norm": 0.9217233657836914,
+      "learning_rate": 9.251831554171773e-06,
+      "loss": 0.8179,
+      "step": 375
+    },
+    {
+      "epoch": 0.7779310344827586,
+      "grad_norm": 0.9028921127319336,
+      "learning_rate": 9.245480730653055e-06,
+      "loss": 0.7961,
+      "step": 376
+    },
+    {
+      "epoch": 0.78,
+      "grad_norm": 0.7794080376625061,
+      "learning_rate": 9.239105265402525e-06,
+      "loss": 0.7874,
+      "step": 377
+    },
+    {
+      "epoch": 0.7820689655172414,
+      "grad_norm": 0.8010353446006775,
+      "learning_rate": 9.232705195424832e-06,
+      "loss": 0.7673,
+      "step": 378
+    },
+    {
+      "epoch": 0.7841379310344827,
+      "grad_norm": 0.8443918824195862,
+      "learning_rate": 9.226280557867428e-06,
+      "loss": 0.7922,
+      "step": 379
+    },
+    {
+      "epoch": 0.7862068965517242,
+      "grad_norm": 0.8193161487579346,
+      "learning_rate": 9.219831390020367e-06,
+      "loss": 0.7668,
+      "step": 380
+    },
+    {
+      "epoch": 0.7882758620689655,
+      "grad_norm": 0.8224409818649292,
+      "learning_rate": 9.213357729316077e-06,
+      "loss": 0.7643,
+      "step": 381
+    },
+    {
+      "epoch": 0.7903448275862069,
+      "grad_norm": 0.8491160273551941,
+      "learning_rate": 9.206859613329155e-06,
+      "loss": 0.7801,
+      "step": 382
+    },
+    {
+      "epoch": 0.7924137931034483,
+      "grad_norm": 0.8366249203681946,
+      "learning_rate": 9.200337079776136e-06,
+      "loss": 0.7496,
+      "step": 383
+    },
+    {
+      "epoch": 0.7944827586206896,
+      "grad_norm": 0.7923499345779419,
+      "learning_rate": 9.193790166515282e-06,
+      "loss": 0.7988,
+      "step": 384
+    },
+    {
+      "epoch": 0.7965517241379311,
+      "grad_norm": 0.8000715970993042,
+      "learning_rate": 9.187218911546363e-06,
+      "loss": 0.7583,
+      "step": 385
+    },
+    {
+      "epoch": 0.7986206896551724,
+      "grad_norm": 0.7873345017433167,
+      "learning_rate": 9.180623353010428e-06,
+      "loss": 0.7631,
+      "step": 386
+    },
+    {
+      "epoch": 0.8006896551724138,
+      "grad_norm": 0.8444647789001465,
+      "learning_rate": 9.174003529189592e-06,
+      "loss": 0.7837,
+      "step": 387
+    },
+    {
+      "epoch": 0.8027586206896552,
+      "grad_norm": 0.7419801354408264,
+      "learning_rate": 9.167359478506816e-06,
+      "loss": 0.7626,
+      "step": 388
+    },
+    {
+      "epoch": 0.8048275862068965,
+      "grad_norm": 0.774566113948822,
+      "learning_rate": 9.160691239525675e-06,
+      "loss": 0.7728,
+      "step": 389
+    },
+    {
+      "epoch": 0.8068965517241379,
+      "grad_norm": 0.730536699295044,
+      "learning_rate": 9.153998850950133e-06,
+      "loss": 0.7907,
+      "step": 390
+    },
+    {
+      "epoch": 0.8089655172413793,
+      "grad_norm": 0.726324737071991,
+      "learning_rate": 9.147282351624334e-06,
+      "loss": 0.7827,
+      "step": 391
+    },
+    {
+      "epoch": 0.8110344827586207,
+      "grad_norm": 0.7642649412155151,
+      "learning_rate": 9.140541780532362e-06,
+      "loss": 0.7417,
+      "step": 392
+    },
+    {
+      "epoch": 0.8131034482758621,
+      "grad_norm": 0.8225309252738953,
+      "learning_rate": 9.133777176798013e-06,
+      "loss": 0.8105,
+      "step": 393
+    },
+    {
+      "epoch": 0.8151724137931035,
+      "grad_norm": 0.7588528394699097,
+      "learning_rate": 9.126988579684587e-06,
+      "loss": 0.781,
+      "step": 394
+    },
+    {
+      "epoch": 0.8172413793103448,
+      "grad_norm": 0.8489255309104919,
+      "learning_rate": 9.120176028594632e-06,
+      "loss": 0.7515,
+      "step": 395
+    },
+    {
+      "epoch": 0.8193103448275862,
+      "grad_norm": 0.8431716561317444,
+      "learning_rate": 9.113339563069742e-06,
+      "loss": 0.7931,
+      "step": 396
+    },
+    {
+      "epoch": 0.8213793103448276,
+      "grad_norm": 0.8133612871170044,
+      "learning_rate": 9.106479222790312e-06,
+      "loss": 0.7877,
+      "step": 397
+    },
+    {
+      "epoch": 0.8234482758620689,
+      "grad_norm": 0.8307138085365295,
+      "learning_rate": 9.099595047575308e-06,
+      "loss": 0.7774,
+      "step": 398
+    },
+    {
+      "epoch": 0.8255172413793104,
+      "grad_norm": 0.7466663122177124,
+      "learning_rate": 9.092687077382043e-06,
+      "loss": 0.7667,
+      "step": 399
+    },
+    {
+      "epoch": 0.8275862068965517,
+      "grad_norm": 0.8137718439102173,
+      "learning_rate": 9.08575535230594e-06,
+      "loss": 0.8056,
+      "step": 400
+    },
+    {
+      "epoch": 0.829655172413793,
+      "grad_norm": 0.8068400621414185,
+      "learning_rate": 9.078799912580305e-06,
+      "loss": 0.7588,
+      "step": 401
+    },
+    {
+      "epoch": 0.8317241379310345,
+      "grad_norm": 0.8182150721549988,
+      "learning_rate": 9.071820798576083e-06,
+      "loss": 0.7748,
+      "step": 402
+    },
+    {
+      "epoch": 0.8337931034482758,
+      "grad_norm": 0.8240687251091003,
+      "learning_rate": 9.064818050801634e-06,
+      "loss": 0.7854,
+      "step": 403
+    },
+    {
+      "epoch": 0.8358620689655173,
+      "grad_norm": 0.8210726976394653,
+      "learning_rate": 9.057791709902491e-06,
+      "loss": 0.7809,
+      "step": 404
+    },
+    {
+      "epoch": 0.8379310344827586,
+      "grad_norm": 0.8129264712333679,
+      "learning_rate": 9.050741816661128e-06,
+      "loss": 0.7747,
+      "step": 405
+    },
+    {
+      "epoch": 0.84,
+      "grad_norm": 0.7455905079841614,
+      "learning_rate": 9.04366841199672e-06,
+      "loss": 0.777,
+      "step": 406
+    },
+    {
+      "epoch": 0.8420689655172414,
+      "grad_norm": 0.8041906356811523,
+      "learning_rate": 9.036571536964915e-06,
+      "loss": 0.7634,
+      "step": 407
+    },
+    {
+      "epoch": 0.8441379310344828,
+      "grad_norm": 0.7791020274162292,
+      "learning_rate": 9.029451232757581e-06,
+      "loss": 0.803,
+      "step": 408
+    },
+    {
+      "epoch": 0.8462068965517241,
+      "grad_norm": 0.8423375487327576,
+      "learning_rate": 9.022307540702576e-06,
+      "loss": 0.7263,
+      "step": 409
+    },
+    {
+      "epoch": 0.8482758620689655,
+      "grad_norm": 0.8119176030158997,
+      "learning_rate": 9.015140502263506e-06,
+      "loss": 0.7459,
+      "step": 410
+    },
+    {
+      "epoch": 0.8503448275862069,
+      "grad_norm": 0.8209352493286133,
+      "learning_rate": 9.007950159039488e-06,
+      "loss": 0.7858,
+      "step": 411
+    },
+    {
+      "epoch": 0.8524137931034482,
+      "grad_norm": 0.8128482699394226,
+      "learning_rate": 9.0007365527649e-06,
+      "loss": 0.7707,
+      "step": 412
+    },
+    {
+      "epoch": 0.8544827586206897,
+      "grad_norm": 0.8830553293228149,
+      "learning_rate": 8.993499725309148e-06,
+      "loss": 0.7579,
+      "step": 413
+    },
+    {
+      "epoch": 0.856551724137931,
+      "grad_norm": 0.8541165590286255,
+      "learning_rate": 8.98623971867642e-06,
+      "loss": 0.7638,
+      "step": 414
+    },
+    {
+      "epoch": 0.8586206896551725,
+      "grad_norm": 0.7872908115386963,
+      "learning_rate": 8.978956575005432e-06,
+      "loss": 0.7967,
+      "step": 415
+    },
+    {
+      "epoch": 0.8606896551724138,
+      "grad_norm": 0.7946335673332214,
+      "learning_rate": 8.971650336569208e-06,
+      "loss": 0.7769,
+      "step": 416
+    },
+    {
+      "epoch": 0.8627586206896551,
+      "grad_norm": 0.8130462169647217,
+      "learning_rate": 8.964321045774808e-06,
+      "loss": 0.7889,
+      "step": 417
+    },
+    {
+      "epoch": 0.8648275862068966,
+      "grad_norm": 0.7823083400726318,
+      "learning_rate": 8.956968745163093e-06,
+      "loss": 0.7814,
+      "step": 418
+    },
+    {
+      "epoch": 0.8668965517241379,
+      "grad_norm": 0.8023502230644226,
+      "learning_rate": 8.949593477408485e-06,
+      "loss": 0.7647,
+      "step": 419
+    },
+    {
+      "epoch": 0.8689655172413793,
+      "grad_norm": 0.7985740900039673,
+      "learning_rate": 8.942195285318705e-06,
+      "loss": 0.7794,
+      "step": 420
+    },
+    {
+      "epoch": 0.8710344827586207,
+      "grad_norm": 0.8288792371749878,
+      "learning_rate": 8.934774211834538e-06,
+      "loss": 0.7891,
+      "step": 421
+    },
+    {
+      "epoch": 0.8731034482758621,
+      "grad_norm": 0.8776261806488037,
+      "learning_rate": 8.927330300029575e-06,
+      "loss": 0.7723,
+      "step": 422
+    },
+    {
+      "epoch": 0.8751724137931034,
+      "grad_norm": 0.8444452285766602,
+      "learning_rate": 8.919863593109967e-06,
+      "loss": 0.7812,
+      "step": 423
+    },
+    {
+      "epoch": 0.8772413793103448,
+      "grad_norm": 0.8049442768096924,
+      "learning_rate": 8.912374134414171e-06,
+      "loss": 0.7765,
+      "step": 424
+    },
+    {
+      "epoch": 0.8793103448275862,
+      "grad_norm": 0.8133780360221863,
+      "learning_rate": 8.904861967412702e-06,
+      "loss": 0.7289,
+      "step": 425
+    },
+    {
+      "epoch": 0.8813793103448276,
+      "grad_norm": 0.7836434245109558,
+      "learning_rate": 8.897327135707881e-06,
+      "loss": 0.7887,
+      "step": 426
+    },
+    {
+      "epoch": 0.883448275862069,
+      "grad_norm": 0.8233901858329773,
+      "learning_rate": 8.889769683033573e-06,
+      "loss": 0.7623,
+      "step": 427
+    },
+    {
+      "epoch": 0.8855172413793103,
+      "grad_norm": 0.8158826231956482,
+      "learning_rate": 8.882189653254951e-06,
+      "loss": 0.7821,
+      "step": 428
+    },
+    {
+      "epoch": 0.8875862068965518,
+      "grad_norm": 0.8711895942687988,
+      "learning_rate": 8.874587090368221e-06,
+      "loss": 0.7668,
+      "step": 429
+    },
+    {
+      "epoch": 0.8896551724137931,
+      "grad_norm": 0.7956521511077881,
+      "learning_rate": 8.86696203850038e-06,
+      "loss": 0.762,
+      "step": 430
+    },
+    {
+      "epoch": 0.8917241379310344,
+      "grad_norm": 0.7590909600257874,
+      "learning_rate": 8.859314541908957e-06,
+      "loss": 0.7269,
+      "step": 431
+    },
+    {
+      "epoch": 0.8937931034482759,
+      "grad_norm": 0.74117511510849,
+      "learning_rate": 8.851644644981756e-06,
+      "loss": 0.7703,
+      "step": 432
+    },
+    {
+      "epoch": 0.8958620689655172,
+      "grad_norm": 0.7806484699249268,
+      "learning_rate": 8.843952392236595e-06,
+      "loss": 0.7608,
+      "step": 433
+    },
+    {
+      "epoch": 0.8979310344827586,
+      "grad_norm": 0.7991644740104675,
+      "learning_rate": 8.83623782832105e-06,
+      "loss": 0.7932,
+      "step": 434
+    },
+    {
+      "epoch": 0.9,
+      "grad_norm": 0.7456468939781189,
+      "learning_rate": 8.8285009980122e-06,
+      "loss": 0.7638,
+      "step": 435
+    },
+    {
+      "epoch": 0.9020689655172414,
+      "grad_norm": 0.814285933971405,
+      "learning_rate": 8.820741946216357e-06,
+      "loss": 0.7711,
+      "step": 436
+    },
+    {
+      "epoch": 0.9041379310344828,
+      "grad_norm": 0.767855703830719,
+      "learning_rate": 8.81296071796882e-06,
+      "loss": 0.792,
+      "step": 437
+    },
+    {
+      "epoch": 0.9062068965517242,
+      "grad_norm": 0.8761329054832458,
+      "learning_rate": 8.805157358433595e-06,
+      "loss": 0.7736,
+      "step": 438
+    },
+    {
+      "epoch": 0.9082758620689655,
+      "grad_norm": 0.9907826781272888,
+      "learning_rate": 8.797331912903152e-06,
+      "loss": 0.7725,
+      "step": 439
+    },
+    {
+      "epoch": 0.9103448275862069,
+      "grad_norm": 0.8171151280403137,
+      "learning_rate": 8.789484426798148e-06,
+      "loss": 0.7593,
+      "step": 440
+    },
+    {
+      "epoch": 0.9124137931034483,
+      "grad_norm": 0.8003857731819153,
+      "learning_rate": 8.78161494566717e-06,
+      "loss": 0.7596,
+      "step": 441
+    },
+    {
+      "epoch": 0.9144827586206896,
+      "grad_norm": 0.9617862105369568,
+      "learning_rate": 8.77372351518647e-06,
+      "loss": 0.7872,
+      "step": 442
+    },
+    {
+      "epoch": 0.9165517241379311,
+      "grad_norm": 0.8046567440032959,
+      "learning_rate": 8.765810181159696e-06,
+      "loss": 0.7482,
+      "step": 443
+    },
+    {
+      "epoch": 0.9186206896551724,
+      "grad_norm": 0.7760241031646729,
+      "learning_rate": 8.757874989517631e-06,
+      "loss": 0.7843,
+      "step": 444
+    },
+    {
+      "epoch": 0.9206896551724137,
+      "grad_norm": 0.7372346520423889,
+      "learning_rate": 8.74991798631793e-06,
+      "loss": 0.7882,
+      "step": 445
+    },
+    {
+      "epoch": 0.9227586206896552,
+      "grad_norm": 0.8651936650276184,
+      "learning_rate": 8.741939217744835e-06,
+      "loss": 0.7872,
+      "step": 446
+    },
+    {
+      "epoch": 0.9248275862068965,
+      "grad_norm": 0.7748751640319824,
+      "learning_rate": 8.733938730108928e-06,
+      "loss": 0.7801,
+      "step": 447
+    },
+    {
+      "epoch": 0.926896551724138,
+      "grad_norm": 0.8208494782447815,
+      "learning_rate": 8.725916569846854e-06,
+      "loss": 0.7878,
+      "step": 448
+    },
+    {
+      "epoch": 0.9289655172413793,
+      "grad_norm": 0.7929314374923706,
+      "learning_rate": 8.717872783521048e-06,
+      "loss": 0.7709,
+      "step": 449
+    },
+    {
+      "epoch": 0.9310344827586207,
+      "grad_norm": 0.7704197764396667,
+      "learning_rate": 8.709807417819464e-06,
+      "loss": 0.7916,
+      "step": 450
+    },
+    {
+      "epoch": 0.9331034482758621,
+      "grad_norm": 0.8212770223617554,
+      "learning_rate": 8.701720519555315e-06,
+      "loss": 0.7711,
+      "step": 451
+    },
+    {
+      "epoch": 0.9351724137931035,
+      "grad_norm": 0.8029863238334656,
+      "learning_rate": 8.69361213566679e-06,
+      "loss": 0.7774,
+      "step": 452
+    },
+    {
+      "epoch": 0.9372413793103448,
+      "grad_norm": 0.7679287791252136,
+      "learning_rate": 8.685482313216784e-06,
+      "loss": 0.7461,
+      "step": 453
+    },
+    {
+      "epoch": 0.9393103448275862,
+      "grad_norm": 0.771782398223877,
+      "learning_rate": 8.67733109939263e-06,
+      "loss": 0.7808,
+      "step": 454
+    },
+    {
+      "epoch": 0.9413793103448276,
+      "grad_norm": 0.7877687811851501,
+      "learning_rate": 8.669158541505817e-06,
+      "loss": 0.7683,
+      "step": 455
+    },
+    {
+      "epoch": 0.9434482758620689,
+      "grad_norm": 0.7960418462753296,
+      "learning_rate": 8.660964686991725e-06,
+      "loss": 0.7719,
+      "step": 456
+    },
+    {
+      "epoch": 0.9455172413793104,
+      "grad_norm": 0.761481761932373,
+      "learning_rate": 8.65274958340934e-06,
+      "loss": 0.7635,
+      "step": 457
+    },
+    {
+      "epoch": 0.9475862068965517,
+      "grad_norm": 0.846750795841217,
+      "learning_rate": 8.644513278440984e-06,
+      "loss": 0.7917,
+      "step": 458
+    },
+    {
+      "epoch": 0.9496551724137932,
+      "grad_norm": 0.8491324782371521,
+      "learning_rate": 8.636255819892037e-06,
+      "loss": 0.7826,
+      "step": 459
+    },
+    {
+      "epoch": 0.9517241379310345,
+      "grad_norm": 0.8026092648506165,
+      "learning_rate": 8.627977255690658e-06,
+      "loss": 0.7715,
+      "step": 460
+    },
+    {
+      "epoch": 0.9537931034482758,
+      "grad_norm": 0.8143620491027832,
+      "learning_rate": 8.61967763388751e-06,
+      "loss": 0.7724,
+      "step": 461
+    },
+    {
+      "epoch": 0.9558620689655173,
+      "grad_norm": 0.7983601689338684,
+      "learning_rate": 8.611357002655475e-06,
+      "loss": 0.8044,
+      "step": 462
+    },
+    {
+      "epoch": 0.9579310344827586,
+      "grad_norm": 0.9372730851173401,
+      "learning_rate": 8.603015410289387e-06,
+      "loss": 0.7635,
+      "step": 463
+    },
+    {
+      "epoch": 0.96,
+      "grad_norm": 0.809413731098175,
+      "learning_rate": 8.594652905205731e-06,
+      "loss": 0.7524,
+      "step": 464
+    },
+    {
+      "epoch": 0.9620689655172414,
+      "grad_norm": 0.78413987159729,
+      "learning_rate": 8.586269535942386e-06,
+      "loss": 0.7623,
+      "step": 465
+    },
+    {
+      "epoch": 0.9641379310344828,
+      "grad_norm": 0.8303465247154236,
+      "learning_rate": 8.577865351158324e-06,
+      "loss": 0.7832,
+      "step": 466
+    },
+    {
+      "epoch": 0.9662068965517241,
+      "grad_norm": 0.9170937538146973,
+      "learning_rate": 8.569440399633337e-06,
+      "loss": 0.7817,
+      "step": 467
+    },
+    {
+      "epoch": 0.9682758620689655,
+      "grad_norm": 0.835443377494812,
+      "learning_rate": 8.560994730267751e-06,
+      "loss": 0.7658,
+      "step": 468
+    },
+    {
+      "epoch": 0.9703448275862069,
+      "grad_norm": 0.8764263987541199,
+      "learning_rate": 8.552528392082147e-06,
+      "loss": 0.7479,
+      "step": 469
+    },
+    {
+      "epoch": 0.9724137931034482,
+      "grad_norm": 0.8065820336341858,
+      "learning_rate": 8.544041434217067e-06,
+      "loss": 0.7469,
+      "step": 470
+    },
+    {
+      "epoch": 0.9744827586206897,
+      "grad_norm": 0.895577073097229,
+      "learning_rate": 8.535533905932739e-06,
+      "loss": 0.7716,
+      "step": 471
+    },
+    {
+      "epoch": 0.976551724137931,
+      "grad_norm": 0.844038188457489,
+      "learning_rate": 8.52700585660878e-06,
+      "loss": 0.7586,
+      "step": 472
+    },
+    {
+      "epoch": 0.9786206896551725,
+      "grad_norm": 0.8391221761703491,
+      "learning_rate": 8.518457335743927e-06,
+      "loss": 0.785,
+      "step": 473
+    },
+    {
+      "epoch": 0.9806896551724138,
+      "grad_norm": 0.802529513835907,
+      "learning_rate": 8.509888392955725e-06,
+      "loss": 0.7786,
+      "step": 474
+    },
+    {
+      "epoch": 0.9827586206896551,
+      "grad_norm": 0.793171763420105,
+      "learning_rate": 8.50129907798026e-06,
+      "loss": 0.7542,
+      "step": 475
+    },
+    {
+      "epoch": 0.9848275862068966,
+      "grad_norm": 0.7531145811080933,
+      "learning_rate": 8.492689440671865e-06,
+      "loss": 0.7837,
+      "step": 476
+    },
+    {
+      "epoch": 0.9868965517241379,
+      "grad_norm": 0.7774952054023743,
+      "learning_rate": 8.484059531002822e-06,
+      "loss": 0.7702,
+      "step": 477
+    },
+    {
+      "epoch": 0.9889655172413793,
+      "grad_norm": 0.84196537733078,
+      "learning_rate": 8.475409399063081e-06,
+      "loss": 0.7628,
+      "step": 478
+    },
+    {
+      "epoch": 0.9910344827586207,
+      "grad_norm": 0.760192334651947,
+      "learning_rate": 8.46673909505997e-06,
+      "loss": 0.7549,
+      "step": 479
+    },
+    {
+      "epoch": 0.993103448275862,
+      "grad_norm": 0.7565990686416626,
+      "learning_rate": 8.458048669317897e-06,
+      "loss": 0.7473,
+      "step": 480
+    },
+    {
+      "epoch": 0.9951724137931034,
+      "grad_norm": 0.792762041091919,
+      "learning_rate": 8.44933817227806e-06,
+      "loss": 0.7631,
+      "step": 481
+    },
+    {
+      "epoch": 0.9972413793103448,
+      "grad_norm": 0.7413747906684875,
+      "learning_rate": 8.440607654498156e-06,
+      "loss": 0.778,
+      "step": 482
+    },
+    {
+      "epoch": 0.9993103448275862,
+      "grad_norm": 0.7627434730529785,
+      "learning_rate": 8.43185716665209e-06,
+      "loss": 0.7737,
+      "step": 483
+    },
+    {
+      "epoch": 1.0013793103448276,
+      "grad_norm": 0.8035737872123718,
+      "learning_rate": 8.423086759529676e-06,
+      "loss": 0.7007,
+      "step": 484
+    },
+    {
+      "epoch": 1.0034482758620689,
+      "grad_norm": 0.7981570363044739,
+      "learning_rate": 8.41429648403634e-06,
+      "loss": 0.6635,
+      "step": 485
+    },
+    {
+      "epoch": 1.0055172413793103,
+      "grad_norm": 0.7854894995689392,
+      "learning_rate": 8.405486391192834e-06,
+      "loss": 0.6601,
+      "step": 486
+    },
+    {
+      "epoch": 1.0075862068965518,
+      "grad_norm": 0.7979158759117126,
+      "learning_rate": 8.396656532134934e-06,
+      "loss": 0.6369,
+      "step": 487
+    },
+    {
+      "epoch": 1.0096551724137932,
+      "grad_norm": 0.8349688053131104,
+      "learning_rate": 8.387806958113139e-06,
+      "loss": 0.6569,
+      "step": 488
+    },
+    {
+      "epoch": 1.0117241379310344,
+      "grad_norm": 0.8813797235488892,
+      "learning_rate": 8.378937720492384e-06,
+      "loss": 0.6697,
+      "step": 489
+    },
+    {
+      "epoch": 1.013793103448276,
+      "grad_norm": 0.8171966671943665,
+      "learning_rate": 8.370048870751732e-06,
+      "loss": 0.6545,
+      "step": 490
+    },
+    {
+      "epoch": 1.0158620689655173,
+      "grad_norm": 0.8515467047691345,
+      "learning_rate": 8.361140460484081e-06,
+      "loss": 0.6476,
+      "step": 491
+    },
+    {
+      "epoch": 1.0179310344827586,
+      "grad_norm": 0.8771365284919739,
+      "learning_rate": 8.352212541395859e-06,
+      "loss": 0.6641,
+      "step": 492
+    },
+    {
+      "epoch": 1.02,
+      "grad_norm": 0.8186706304550171,
+      "learning_rate": 8.343265165306736e-06,
+      "loss": 0.6558,
+      "step": 493
+    },
+    {
+      "epoch": 1.0220689655172415,
+      "grad_norm": 0.821365237236023,
+      "learning_rate": 8.334298384149305e-06,
+      "loss": 0.6526,
+      "step": 494
+    },
+    {
+      "epoch": 1.0241379310344827,
+      "grad_norm": 0.8323445320129395,
+      "learning_rate": 8.325312249968795e-06,
+      "loss": 0.6788,
+      "step": 495
+    },
+    {
+      "epoch": 1.0262068965517241,
+      "grad_norm": 0.7858123183250427,
+      "learning_rate": 8.316306814922769e-06,
+      "loss": 0.6476,
+      "step": 496
+    },
+    {
+      "epoch": 1.0282758620689656,
+      "grad_norm": 0.8636062741279602,
+      "learning_rate": 8.307282131280805e-06,
+      "loss": 0.6494,
+      "step": 497
+    },
+    {
+      "epoch": 1.0303448275862068,
+      "grad_norm": 0.8002017736434937,
+      "learning_rate": 8.298238251424216e-06,
+      "loss": 0.6432,
+      "step": 498
+    },
+    {
+      "epoch": 1.0324137931034483,
+      "grad_norm": 0.852461040019989,
+      "learning_rate": 8.289175227845727e-06,
+      "loss": 0.6602,
+      "step": 499
+    },
+    {
+      "epoch": 1.0344827586206897,
+      "grad_norm": 0.782200813293457,
+      "learning_rate": 8.280093113149182e-06,
+      "loss": 0.6347,
+      "step": 500
+    },
+    {
+      "epoch": 1.036551724137931,
+      "grad_norm": 0.7889745831489563,
+      "learning_rate": 8.270991960049231e-06,
+      "loss": 0.6626,
+      "step": 501
+    },
+    {
+      "epoch": 1.0386206896551724,
+      "grad_norm": 0.7759472131729126,
+      "learning_rate": 8.261871821371026e-06,
+      "loss": 0.6374,
+      "step": 502
+    },
+    {
+      "epoch": 1.0406896551724139,
+      "grad_norm": 0.7325034737586975,
+      "learning_rate": 8.252732750049918e-06,
+      "loss": 0.6531,
+      "step": 503
+    },
+    {
+      "epoch": 1.042758620689655,
+      "grad_norm": 0.7706113457679749,
+      "learning_rate": 8.24357479913115e-06,
+      "loss": 0.6526,
+      "step": 504
+    },
+    {
+      "epoch": 1.0448275862068965,
+      "grad_norm": 0.7855322360992432,
+      "learning_rate": 8.234398021769541e-06,
+      "loss": 0.6443,
+      "step": 505
+    },
+    {
+      "epoch": 1.046896551724138,
+      "grad_norm": 0.7973263263702393,
+      "learning_rate": 8.225202471229187e-06,
+      "loss": 0.6422,
+      "step": 506
+    },
+    {
+      "epoch": 1.0489655172413792,
+      "grad_norm": 0.7677010297775269,
+      "learning_rate": 8.215988200883145e-06,
+      "loss": 0.6378,
+      "step": 507
+    },
+    {
+      "epoch": 1.0510344827586207,
+      "grad_norm": 0.8065165877342224,
+      "learning_rate": 8.206755264213128e-06,
+      "loss": 0.6291,
+      "step": 508
+    },
+    {
+      "epoch": 1.053103448275862,
+      "grad_norm": 0.7429850101470947,
+      "learning_rate": 8.19750371480919e-06,
+      "loss": 0.6671,
+      "step": 509
+    },
+    {
+      "epoch": 1.0551724137931036,
+      "grad_norm": 0.8575379252433777,
+      "learning_rate": 8.188233606369422e-06,
+      "loss": 0.6462,
+      "step": 510
+    },
+    {
+      "epoch": 1.0572413793103448,
+      "grad_norm": 0.7762051224708557,
+      "learning_rate": 8.178944992699632e-06,
+      "loss": 0.6743,
+      "step": 511
+    },
+    {
+      "epoch": 1.0593103448275862,
+      "grad_norm": 0.7541788220405579,
+      "learning_rate": 8.169637927713037e-06,
+      "loss": 0.6351,
+      "step": 512
+    },
+    {
+      "epoch": 1.0613793103448277,
+      "grad_norm": 0.7777592539787292,
+      "learning_rate": 8.160312465429952e-06,
+      "loss": 0.6506,
+      "step": 513
+    },
+    {
+      "epoch": 1.063448275862069,
+      "grad_norm": 0.7478598952293396,
+      "learning_rate": 8.150968659977472e-06,
+      "loss": 0.6781,
+      "step": 514
+    },
+    {
+      "epoch": 1.0655172413793104,
+      "grad_norm": 0.7685233950614929,
+      "learning_rate": 8.141606565589157e-06,
+      "loss": 0.6802,
+      "step": 515
+    },
+    {
+      "epoch": 1.0675862068965518,
+      "grad_norm": 0.7681663632392883,
+      "learning_rate": 8.132226236604728e-06,
+      "loss": 0.6421,
+      "step": 516
+    },
+    {
+      "epoch": 1.069655172413793,
+      "grad_norm": 0.7492855787277222,
+      "learning_rate": 8.122827727469737e-06,
+      "loss": 0.6445,
+      "step": 517
+    },
+    {
+      "epoch": 1.0717241379310345,
+      "grad_norm": 0.8009100556373596,
+      "learning_rate": 8.113411092735262e-06,
+      "loss": 0.6424,
+      "step": 518
+    },
+    {
+      "epoch": 1.073793103448276,
+      "grad_norm": 0.7607418894767761,
+      "learning_rate": 8.103976387057579e-06,
+      "loss": 0.6456,
+      "step": 519
+    },
+    {
+      "epoch": 1.0758620689655172,
+      "grad_norm": 0.805308997631073,
+      "learning_rate": 8.094523665197862e-06,
+      "loss": 0.6639,
+      "step": 520
+    },
+    {
+      "epoch": 1.0779310344827586,
+      "grad_norm": 0.7429332137107849,
+      "learning_rate": 8.085052982021849e-06,
+      "loss": 0.644,
+      "step": 521
+    },
+    {
+      "epoch": 1.08,
+      "grad_norm": 0.7760939598083496,
+      "learning_rate": 8.075564392499527e-06,
+      "loss": 0.6472,
+      "step": 522
+    },
+    {
+      "epoch": 1.0820689655172413,
+      "grad_norm": 0.8237507343292236,
+      "learning_rate": 8.066057951704821e-06,
+      "loss": 0.643,
+      "step": 523
+    },
+    {
+      "epoch": 1.0841379310344827,
+      "grad_norm": 0.7856370806694031,
+      "learning_rate": 8.056533714815268e-06,
+      "loss": 0.658,
+      "step": 524
+    },
+    {
+      "epoch": 1.0862068965517242,
+      "grad_norm": 0.7537881731987,
+      "learning_rate": 8.046991737111696e-06,
+      "loss": 0.6441,
+      "step": 525
+    },
+    {
+      "epoch": 1.0882758620689654,
+      "grad_norm": 0.7884009480476379,
+      "learning_rate": 8.037432073977902e-06,
+      "loss": 0.6253,
+      "step": 526
+    },
+    {
+      "epoch": 1.0903448275862069,
+      "grad_norm": 0.9002606868743896,
+      "learning_rate": 8.027854780900339e-06,
+      "loss": 0.6652,
+      "step": 527
+    },
+    {
+      "epoch": 1.0924137931034483,
+      "grad_norm": 0.7838542461395264,
+      "learning_rate": 8.018259913467785e-06,
+      "loss": 0.6783,
+      "step": 528
+    },
+    {
+      "epoch": 1.0944827586206896,
+      "grad_norm": 0.74989253282547,
+      "learning_rate": 8.008647527371022e-06,
+      "loss": 0.655,
+      "step": 529
+    },
+    {
+      "epoch": 1.096551724137931,
+      "grad_norm": 0.9094736576080322,
+      "learning_rate": 7.999017678402518e-06,
+      "loss": 0.6587,
+      "step": 530
+    },
+    {
+      "epoch": 1.0986206896551725,
+      "grad_norm": 0.7564888000488281,
+      "learning_rate": 7.989370422456098e-06,
+      "loss": 0.6577,
+      "step": 531
+    },
+    {
+      "epoch": 1.1006896551724137,
+      "grad_norm": 0.806409478187561,
+      "learning_rate": 7.979705815526621e-06,
+      "loss": 0.6666,
+      "step": 532
+    },
+    {
+      "epoch": 1.1027586206896551,
+      "grad_norm": 0.7861296534538269,
+      "learning_rate": 7.970023913709652e-06,
+      "loss": 0.6849,
+      "step": 533
+    },
+    {
+      "epoch": 1.1048275862068966,
+      "grad_norm": 0.7604372501373291,
+      "learning_rate": 7.960324773201147e-06,
+      "loss": 0.6363,
+      "step": 534
+    },
+    {
+      "epoch": 1.106896551724138,
+      "grad_norm": 0.778635561466217,
+      "learning_rate": 7.95060845029711e-06,
+      "loss": 0.6551,
+      "step": 535
+    },
+    {
+      "epoch": 1.1089655172413793,
+      "grad_norm": 0.8555791974067688,
+      "learning_rate": 7.94087500139328e-06,
+      "loss": 0.6587,
+      "step": 536
+    },
+    {
+      "epoch": 1.1110344827586207,
+      "grad_norm": 0.7580198645591736,
+      "learning_rate": 7.931124482984802e-06,
+      "loss": 0.678,
+      "step": 537
+    },
+    {
+      "epoch": 1.1131034482758622,
+      "grad_norm": 0.749212384223938,
+      "learning_rate": 7.921356951665894e-06,
+      "loss": 0.6235,
+      "step": 538
+    },
+    {
+      "epoch": 1.1151724137931034,
+      "grad_norm": 0.7692828178405762,
+      "learning_rate": 7.911572464129515e-06,
+      "loss": 0.663,
+      "step": 539
+    },
+    {
+      "epoch": 1.1172413793103448,
+      "grad_norm": 0.7994730472564697,
+      "learning_rate": 7.901771077167051e-06,
+      "loss": 0.6318,
+      "step": 540
+    },
+    {
+      "epoch": 1.1193103448275863,
+      "grad_norm": 0.7523401975631714,
+      "learning_rate": 7.891952847667973e-06,
+      "loss": 0.6528,
+      "step": 541
+    },
+    {
+      "epoch": 1.1213793103448275,
+      "grad_norm": 0.7888882160186768,
+      "learning_rate": 7.882117832619506e-06,
+      "loss": 0.6238,
+      "step": 542
+    },
+    {
+      "epoch": 1.123448275862069,
+      "grad_norm": 0.7809697389602661,
+      "learning_rate": 7.872266089106309e-06,
+      "loss": 0.634,
+      "step": 543
+    },
+    {
+      "epoch": 1.1255172413793104,
+      "grad_norm": 0.7669185400009155,
+      "learning_rate": 7.862397674310127e-06,
+      "loss": 0.6938,
+      "step": 544
+    },
+    {
+      "epoch": 1.1275862068965516,
+      "grad_norm": 0.7955731749534607,
+      "learning_rate": 7.85251264550948e-06,
+      "loss": 0.6711,
+      "step": 545
+    },
+    {
+      "epoch": 1.129655172413793,
+      "grad_norm": 0.7307600378990173,
+      "learning_rate": 7.842611060079312e-06,
+      "loss": 0.6459,
+      "step": 546
+    },
+    {
+      "epoch": 1.1317241379310345,
+      "grad_norm": 0.8730869889259338,
+      "learning_rate": 7.832692975490668e-06,
+      "loss": 0.6805,
+      "step": 547
+    },
+    {
+      "epoch": 1.1337931034482758,
+      "grad_norm": 0.7648000121116638,
+      "learning_rate": 7.822758449310357e-06,
+      "loss": 0.6502,
+      "step": 548
+    },
+    {
+      "epoch": 1.1358620689655172,
+      "grad_norm": 0.7544475197792053,
+      "learning_rate": 7.812807539200622e-06,
+      "loss": 0.6807,
+      "step": 549
+    },
+    {
+      "epoch": 1.1379310344827587,
+      "grad_norm": 0.7903164029121399,
+      "learning_rate": 7.802840302918794e-06,
+      "loss": 0.6667,
+      "step": 550
+    },
+    {
+      "epoch": 1.1400000000000001,
+      "grad_norm": 0.7563683390617371,
+      "learning_rate": 7.792856798316975e-06,
+      "loss": 0.633,
+      "step": 551
+    },
+    {
+      "epoch": 1.1420689655172414,
+      "grad_norm": 0.8001116514205933,
+      "learning_rate": 7.782857083341685e-06,
+      "loss": 0.6297,
+      "step": 552
+    },
+    {
+      "epoch": 1.1441379310344828,
+      "grad_norm": 0.8235406279563904,
+      "learning_rate": 7.772841216033534e-06,
+      "loss": 0.6372,
+      "step": 553
+    },
+    {
+      "epoch": 1.1462068965517243,
+      "grad_norm": 0.7986905574798584,
+      "learning_rate": 7.762809254526884e-06,
+      "loss": 0.6727,
+      "step": 554
+    },
+    {
+      "epoch": 1.1482758620689655,
+      "grad_norm": 0.7706049680709839,
+      "learning_rate": 7.752761257049517e-06,
+      "loss": 0.6775,
+      "step": 555
+    },
+    {
+      "epoch": 1.150344827586207,
+      "grad_norm": 0.8530354499816895,
+      "learning_rate": 7.74269728192228e-06,
+      "loss": 0.6551,
+      "step": 556
+    },
+    {
+      "epoch": 1.1524137931034484,
+      "grad_norm": 0.8466417789459229,
+      "learning_rate": 7.732617387558769e-06,
+      "loss": 0.6418,
+      "step": 557
+    },
+    {
+      "epoch": 1.1544827586206896,
+      "grad_norm": 0.7542173862457275,
+      "learning_rate": 7.722521632464972e-06,
+      "loss": 0.6017,
+      "step": 558
+    },
+    {
+      "epoch": 1.156551724137931,
+      "grad_norm": 0.7655349373817444,
+      "learning_rate": 7.712410075238936e-06,
+      "loss": 0.6454,
+      "step": 559
+    },
+    {
+      "epoch": 1.1586206896551725,
+      "grad_norm": 0.7444957494735718,
+      "learning_rate": 7.702282774570434e-06,
+      "loss": 0.653,
+      "step": 560
+    },
+    {
+      "epoch": 1.1606896551724137,
+      "grad_norm": 0.7831370830535889,
+      "learning_rate": 7.692139789240611e-06,
+      "loss": 0.6443,
+      "step": 561
+    },
+    {
+      "epoch": 1.1627586206896552,
+      "grad_norm": 0.7594817876815796,
+      "learning_rate": 7.681981178121652e-06,
+      "loss": 0.6422,
+      "step": 562
+    },
+    {
+      "epoch": 1.1648275862068966,
+      "grad_norm": 0.7494958639144897,
+      "learning_rate": 7.671807000176434e-06,
+      "loss": 0.6217,
+      "step": 563
+    },
+    {
+      "epoch": 1.1668965517241379,
+      "grad_norm": 0.862666130065918,
+      "learning_rate": 7.661617314458194e-06,
+      "loss": 0.6465,
+      "step": 564
+    },
+    {
+      "epoch": 1.1689655172413793,
+      "grad_norm": 0.7856999635696411,
+      "learning_rate": 7.651412180110176e-06,
+      "loss": 0.6482,
+      "step": 565
+    },
+    {
+      "epoch": 1.1710344827586208,
+      "grad_norm": 0.7493500709533691,
+      "learning_rate": 7.641191656365288e-06,
+      "loss": 0.6378,
+      "step": 566
+    },
+    {
+      "epoch": 1.173103448275862,
+      "grad_norm": 0.7933719754219055,
+      "learning_rate": 7.630955802545768e-06,
+      "loss": 0.6481,
+      "step": 567
+    },
+    {
+      "epoch": 1.1751724137931034,
+      "grad_norm": 0.7814990282058716,
+      "learning_rate": 7.6207046780628295e-06,
+      "loss": 0.6434,
+      "step": 568
+    },
+    {
+      "epoch": 1.177241379310345,
+      "grad_norm": 0.7389763593673706,
+      "learning_rate": 7.61043834241632e-06,
+      "loss": 0.6241,
+      "step": 569
+    },
+    {
+      "epoch": 1.1793103448275861,
+      "grad_norm": 0.7658818364143372,
+      "learning_rate": 7.600156855194376e-06,
+      "loss": 0.6446,
+      "step": 570
+    },
+    {
+      "epoch": 1.1813793103448276,
+      "grad_norm": 0.8013659119606018,
+      "learning_rate": 7.589860276073082e-06,
+      "loss": 0.6253,
+      "step": 571
+    },
+    {
+      "epoch": 1.183448275862069,
+      "grad_norm": 0.7773886919021606,
+      "learning_rate": 7.579548664816112e-06,
+      "loss": 0.6574,
+      "step": 572
+    },
+    {
+      "epoch": 1.1855172413793102,
+      "grad_norm": 0.8565280437469482,
+      "learning_rate": 7.569222081274396e-06,
+      "loss": 0.6781,
+      "step": 573
+    },
+    {
+      "epoch": 1.1875862068965517,
+      "grad_norm": 0.75011146068573,
+      "learning_rate": 7.558880585385762e-06,
+      "loss": 0.6628,
+      "step": 574
+    },
+    {
+      "epoch": 1.1896551724137931,
+      "grad_norm": 0.854688823223114,
+      "learning_rate": 7.548524237174595e-06,
+      "loss": 0.6442,
+      "step": 575
+    },
+    {
+      "epoch": 1.1917241379310344,
+      "grad_norm": 0.8342920541763306,
+      "learning_rate": 7.538153096751486e-06,
+      "loss": 0.6493,
+      "step": 576
+    },
+    {
+      "epoch": 1.1937931034482758,
+      "grad_norm": 0.7875855565071106,
+      "learning_rate": 7.527767224312883e-06,
+      "loss": 0.6428,
+      "step": 577
+    },
+    {
+      "epoch": 1.1958620689655173,
+      "grad_norm": 0.8200221061706543,
+      "learning_rate": 7.517366680140743e-06,
+      "loss": 0.6735,
+      "step": 578
+    },
+    {
+      "epoch": 1.1979310344827585,
+      "grad_norm": 0.8583515882492065,
+      "learning_rate": 7.50695152460218e-06,
+      "loss": 0.6432,
+      "step": 579
+    },
+    {
+      "epoch": 1.2,
+      "grad_norm": 0.7881141304969788,
+      "learning_rate": 7.496521818149114e-06,
+      "loss": 0.6455,
+      "step": 580
+    },
+    {
+      "epoch": 1.2020689655172414,
+      "grad_norm": 0.8400844931602478,
+      "learning_rate": 7.4860776213179264e-06,
+      "loss": 0.6569,
+      "step": 581
+    },
+    {
+      "epoch": 1.2041379310344829,
+      "grad_norm": 0.8359171748161316,
+      "learning_rate": 7.4756189947291e-06,
+      "loss": 0.637,
+      "step": 582
+    },
+    {
+      "epoch": 1.206206896551724,
+      "grad_norm": 0.8451258540153503,
+      "learning_rate": 7.465145999086874e-06,
+      "loss": 0.6504,
+      "step": 583
+    },
+    {
+      "epoch": 1.2082758620689655,
+      "grad_norm": 0.7891321778297424,
+      "learning_rate": 7.454658695178885e-06,
+      "loss": 0.6712,
+      "step": 584
+    },
+    {
+      "epoch": 1.210344827586207,
+      "grad_norm": 0.8851990103721619,
+      "learning_rate": 7.44415714387582e-06,
+      "loss": 0.6334,
+      "step": 585
+    },
+    {
+      "epoch": 1.2124137931034482,
+      "grad_norm": 0.9024987816810608,
+      "learning_rate": 7.433641406131065e-06,
+      "loss": 0.6604,
+      "step": 586
+    },
+    {
+      "epoch": 1.2144827586206897,
+      "grad_norm": 0.7987050414085388,
+      "learning_rate": 7.4231115429803405e-06,
+      "loss": 0.6633,
+      "step": 587
+    },
+    {
+      "epoch": 1.2165517241379311,
+      "grad_norm": 0.7416420578956604,
+      "learning_rate": 7.41256761554136e-06,
+      "loss": 0.6401,
+      "step": 588
+    },
+    {
+      "epoch": 1.2186206896551723,
+      "grad_norm": 0.8680900931358337,
+      "learning_rate": 7.4020096850134635e-06,
+      "loss": 0.6553,
+      "step": 589
+    },
+    {
+      "epoch": 1.2206896551724138,
+      "grad_norm": 0.8396124243736267,
+      "learning_rate": 7.391437812677274e-06,
+      "loss": 0.6374,
+      "step": 590
+    },
+    {
+      "epoch": 1.2227586206896552,
+      "grad_norm": 0.7313915491104126,
+      "learning_rate": 7.380852059894331e-06,
+      "loss": 0.6733,
+      "step": 591
+    },
+    {
+      "epoch": 1.2248275862068965,
+      "grad_norm": 0.8062239289283752,
+      "learning_rate": 7.370252488106744e-06,
+      "loss": 0.6697,
+      "step": 592
+    },
+    {
+      "epoch": 1.226896551724138,
+      "grad_norm": 0.8707490563392639,
+      "learning_rate": 7.359639158836828e-06,
+      "loss": 0.6517,
+      "step": 593
+    },
+    {
+      "epoch": 1.2289655172413794,
+      "grad_norm": 0.8169229030609131,
+      "learning_rate": 7.349012133686748e-06,
+      "loss": 0.6507,
+      "step": 594
+    },
+    {
+      "epoch": 1.2310344827586206,
+      "grad_norm": 0.8145736455917358,
+      "learning_rate": 7.3383714743381664e-06,
+      "loss": 0.6594,
+      "step": 595
+    },
+    {
+      "epoch": 1.233103448275862,
+      "grad_norm": 0.7559329271316528,
+      "learning_rate": 7.327717242551882e-06,
+      "loss": 0.6685,
+      "step": 596
+    },
+    {
+      "epoch": 1.2351724137931035,
+      "grad_norm": 0.840366542339325,
+      "learning_rate": 7.317049500167466e-06,
+      "loss": 0.6152,
+      "step": 597
+    },
+    {
+      "epoch": 1.237241379310345,
+      "grad_norm": 0.8220683932304382,
+      "learning_rate": 7.306368309102914e-06,
+      "loss": 0.6615,
+      "step": 598
+    },
+    {
+      "epoch": 1.2393103448275862,
+      "grad_norm": 0.75614994764328,
+      "learning_rate": 7.295673731354278e-06,
+      "loss": 0.6451,
+      "step": 599
+    },
+    {
+      "epoch": 1.2413793103448276,
+      "grad_norm": 0.8044620156288147,
+      "learning_rate": 7.2849658289953084e-06,
+      "loss": 0.6482,
+      "step": 600
+    },
+    {
+      "epoch": 1.243448275862069,
+      "grad_norm": 0.7796791791915894,
+      "learning_rate": 7.2742446641770985e-06,
+      "loss": 0.6371,
+      "step": 601
+    },
+    {
+      "epoch": 1.2455172413793103,
+      "grad_norm": 0.739811897277832,
+      "learning_rate": 7.263510299127714e-06,
+      "loss": 0.6489,
+      "step": 602
+    },
+    {
+      "epoch": 1.2475862068965518,
+      "grad_norm": 0.7257277965545654,
+      "learning_rate": 7.252762796151843e-06,
+      "loss": 0.648,
+      "step": 603
+    },
+    {
+      "epoch": 1.2496551724137932,
+      "grad_norm": 0.7809808850288391,
+      "learning_rate": 7.242002217630428e-06,
+      "loss": 0.6584,
+      "step": 604
+    },
+    {
+      "epoch": 1.2517241379310344,
+      "grad_norm": 0.8291626572608948,
+      "learning_rate": 7.231228626020303e-06,
+      "loss": 0.652,
+      "step": 605
+    },
+    {
+      "epoch": 1.2537931034482759,
+      "grad_norm": 0.7942583560943604,
+      "learning_rate": 7.220442083853834e-06,
+      "loss": 0.6538,
+      "step": 606
+    },
+    {
+      "epoch": 1.2558620689655173,
+      "grad_norm": 0.7212340235710144,
+      "learning_rate": 7.209642653738555e-06,
+      "loss": 0.6351,
+      "step": 607
+    },
+    {
+      "epoch": 1.2579310344827586,
+      "grad_norm": 0.8261105418205261,
+      "learning_rate": 7.198830398356807e-06,
+      "loss": 0.6725,
+      "step": 608
+    },
+    {
+      "epoch": 1.26,
+      "grad_norm": 0.8113548755645752,
+      "learning_rate": 7.188005380465365e-06,
+      "loss": 0.6654,
+      "step": 609
+    },
+    {
+      "epoch": 1.2620689655172415,
+      "grad_norm": 0.76350337266922,
+      "learning_rate": 7.177167662895087e-06,
+      "loss": 0.6775,
+      "step": 610
+    },
+    {
+      "epoch": 1.2641379310344827,
+      "grad_norm": 0.6983400583267212,
+      "learning_rate": 7.166317308550541e-06,
+      "loss": 0.6494,
+      "step": 611
+    },
+    {
+      "epoch": 1.2662068965517241,
+      "grad_norm": 0.7737719416618347,
+      "learning_rate": 7.1554543804096425e-06,
+      "loss": 0.6344,
+      "step": 612
+    },
+    {
+      "epoch": 1.2682758620689656,
+      "grad_norm": 0.9658696055412292,
+      "learning_rate": 7.144578941523283e-06,
+      "loss": 0.6472,
+      "step": 613
+    },
+    {
+      "epoch": 1.2703448275862068,
+      "grad_norm": 0.7466915845870972,
+      "learning_rate": 7.1336910550149775e-06,
+      "loss": 0.6659,
+      "step": 614
+    },
+    {
+      "epoch": 1.2724137931034483,
+      "grad_norm": 0.7958390116691589,
+      "learning_rate": 7.122790784080482e-06,
+      "loss": 0.6792,
+      "step": 615
+    },
+    {
+      "epoch": 1.2744827586206897,
+      "grad_norm": 0.8669654726982117,
+      "learning_rate": 7.111878191987441e-06,
+      "loss": 0.6508,
+      "step": 616
+    },
+    {
+      "epoch": 1.276551724137931,
+      "grad_norm": 0.8021039366722107,
+      "learning_rate": 7.10095334207501e-06,
+      "loss": 0.6379,
+      "step": 617
+    },
+    {
+      "epoch": 1.2786206896551724,
+      "grad_norm": 0.8026660680770874,
+      "learning_rate": 7.090016297753492e-06,
+      "loss": 0.6554,
+      "step": 618
+    },
+    {
+      "epoch": 1.2806896551724138,
+      "grad_norm": 0.7808569669723511,
+      "learning_rate": 7.079067122503969e-06,
+      "loss": 0.6693,
+      "step": 619
+    },
+    {
+      "epoch": 1.282758620689655,
+      "grad_norm": 0.8397783637046814,
+      "learning_rate": 7.0681058798779335e-06,
+      "loss": 0.6484,
+      "step": 620
+    },
+    {
+      "epoch": 1.2848275862068965,
+      "grad_norm": 0.9213140606880188,
+      "learning_rate": 7.057132633496924e-06,
+      "loss": 0.6603,
+      "step": 621
+    },
+    {
+      "epoch": 1.286896551724138,
+      "grad_norm": 0.7552454471588135,
+      "learning_rate": 7.046147447052146e-06,
+      "loss": 0.6536,
+      "step": 622
+    },
+    {
+      "epoch": 1.2889655172413792,
+      "grad_norm": 0.7723549604415894,
+      "learning_rate": 7.03515038430411e-06,
+      "loss": 0.6549,
+      "step": 623
+    },
+    {
+      "epoch": 1.2910344827586206,
+      "grad_norm": 0.965324878692627,
+      "learning_rate": 7.024141509082259e-06,
+      "loss": 0.6552,
+      "step": 624
+    },
+    {
+      "epoch": 1.293103448275862,
+      "grad_norm": 0.8971583843231201,
+      "learning_rate": 7.013120885284599e-06,
+      "loss": 0.6667,
+      "step": 625
+    },
+    {
+      "epoch": 1.2951724137931033,
+      "grad_norm": 0.8151547908782959,
+      "learning_rate": 7.002088576877325e-06,
+      "loss": 0.6451,
+      "step": 626
+    },
+    {
+      "epoch": 1.2972413793103448,
+      "grad_norm": 0.7296470999717712,
+      "learning_rate": 6.991044647894456e-06,
+      "loss": 0.6517,
+      "step": 627
+    },
+    {
+      "epoch": 1.2993103448275862,
+      "grad_norm": 0.9281187057495117,
+      "learning_rate": 6.9799891624374565e-06,
+      "loss": 0.6777,
+      "step": 628
+    },
+    {
+      "epoch": 1.3013793103448275,
+      "grad_norm": 0.9756149053573608,
+      "learning_rate": 6.968922184674868e-06,
+      "loss": 0.6432,
+      "step": 629
+    },
+    {
+      "epoch": 1.303448275862069,
+      "grad_norm": 0.7917487025260925,
+      "learning_rate": 6.957843778841937e-06,
+      "loss": 0.6461,
+      "step": 630
+    },
+    {
+      "epoch": 1.3055172413793104,
+      "grad_norm": 0.7548885941505432,
+      "learning_rate": 6.946754009240241e-06,
+      "loss": 0.6323,
+      "step": 631
+    },
+    {
+      "epoch": 1.3075862068965518,
+      "grad_norm": 0.8275778889656067,
+      "learning_rate": 6.935652940237313e-06,
+      "loss": 0.6651,
+      "step": 632
+    },
+    {
+      "epoch": 1.309655172413793,
+      "grad_norm": 0.8038797974586487,
+      "learning_rate": 6.924540636266272e-06,
+      "loss": 0.6631,
+      "step": 633
+    },
+    {
+      "epoch": 1.3117241379310345,
+      "grad_norm": 0.7313756346702576,
+      "learning_rate": 6.913417161825449e-06,
+      "loss": 0.6292,
+      "step": 634
+    },
+    {
+      "epoch": 1.313793103448276,
+      "grad_norm": 0.7938227653503418,
+      "learning_rate": 6.902282581478009e-06,
+      "loss": 0.643,
+      "step": 635
+    },
+    {
+      "epoch": 1.3158620689655172,
+      "grad_norm": 0.7737557291984558,
+      "learning_rate": 6.891136959851576e-06,
+      "loss": 0.6506,
+      "step": 636
+    },
+    {
+      "epoch": 1.3179310344827586,
+      "grad_norm": 0.7778476476669312,
+      "learning_rate": 6.879980361637865e-06,
+      "loss": 0.6683,
+      "step": 637
+    },
+    {
+      "epoch": 1.32,
+      "grad_norm": 0.8388012051582336,
+      "learning_rate": 6.868812851592299e-06,
+      "loss": 0.6508,
+      "step": 638
+    },
+    {
+      "epoch": 1.3220689655172415,
+      "grad_norm": 0.8127637505531311,
+      "learning_rate": 6.857634494533636e-06,
+      "loss": 0.6521,
+      "step": 639
+    },
+    {
+      "epoch": 1.3241379310344827,
+      "grad_norm": 0.718436062335968,
+      "learning_rate": 6.846445355343591e-06,
+      "loss": 0.6625,
+      "step": 640
+    },
+    {
+      "epoch": 1.3262068965517242,
+      "grad_norm": 0.7375659346580505,
+      "learning_rate": 6.835245498966461e-06,
+      "loss": 0.6293,
+      "step": 641
+    },
+    {
+      "epoch": 1.3282758620689656,
+      "grad_norm": 0.7901964783668518,
+      "learning_rate": 6.824034990408753e-06,
+      "loss": 0.6281,
+      "step": 642
+    },
+    {
+      "epoch": 1.3303448275862069,
+      "grad_norm": 0.7931618094444275,
+      "learning_rate": 6.8128138947387966e-06,
+      "loss": 0.6634,
+      "step": 643
+    },
+    {
+      "epoch": 1.3324137931034483,
+      "grad_norm": 0.7482447624206543,
+      "learning_rate": 6.801582277086371e-06,
+      "loss": 0.6446,
+      "step": 644
+    },
+    {
+      "epoch": 1.3344827586206898,
+      "grad_norm": 0.8103066086769104,
+      "learning_rate": 6.790340202642333e-06,
+      "loss": 0.6436,
+      "step": 645
+    },
+    {
+      "epoch": 1.336551724137931,
+      "grad_norm": 0.7419809699058533,
+      "learning_rate": 6.7790877366582255e-06,
+      "loss": 0.6632,
+      "step": 646
+    },
+    {
+      "epoch": 1.3386206896551724,
+      "grad_norm": 0.7484923005104065,
+      "learning_rate": 6.76782494444591e-06,
+      "loss": 0.6609,
+      "step": 647
+    },
+    {
+      "epoch": 1.340689655172414,
+      "grad_norm": 0.7377297282218933,
+      "learning_rate": 6.756551891377185e-06,
+      "loss": 0.6557,
+      "step": 648
+    },
+    {
+      "epoch": 1.3427586206896551,
+      "grad_norm": 0.8004779815673828,
+      "learning_rate": 6.7452686428834045e-06,
+      "loss": 0.6413,
+      "step": 649
+    },
+    {
+      "epoch": 1.3448275862068966,
+      "grad_norm": 0.7940952181816101,
+      "learning_rate": 6.733975264455097e-06,
+      "loss": 0.6604,
+      "step": 650
+    },
+    {
+      "epoch": 1.346896551724138,
+      "grad_norm": 0.734728991985321,
+      "learning_rate": 6.722671821641591e-06,
+      "loss": 0.6445,
+      "step": 651
+    },
+    {
+      "epoch": 1.3489655172413793,
+      "grad_norm": 0.811326801776886,
+      "learning_rate": 6.711358380050628e-06,
+      "loss": 0.6637,
+      "step": 652
+    },
+    {
+      "epoch": 1.3510344827586207,
+      "grad_norm": 0.8057741522789001,
+      "learning_rate": 6.700035005347983e-06,
+      "loss": 0.6385,
+      "step": 653
+    },
+    {
+      "epoch": 1.3531034482758622,
+      "grad_norm": 0.7723029851913452,
+      "learning_rate": 6.68870176325709e-06,
+      "loss": 0.6514,
+      "step": 654
+    },
+    {
+      "epoch": 1.3551724137931034,
+      "grad_norm": 0.7846975326538086,
+      "learning_rate": 6.677358719558655e-06,
+      "loss": 0.6312,
+      "step": 655
+    },
+    {
+      "epoch": 1.3572413793103448,
+      "grad_norm": 0.8255680203437805,
+      "learning_rate": 6.666005940090271e-06,
+      "loss": 0.6651,
+      "step": 656
+    },
+    {
+      "epoch": 1.3593103448275863,
+      "grad_norm": 0.7784463763237,
+      "learning_rate": 6.654643490746042e-06,
+      "loss": 0.6654,
+      "step": 657
+    },
+    {
+      "epoch": 1.3613793103448275,
+      "grad_norm": 0.8567957878112793,
+      "learning_rate": 6.643271437476196e-06,
+      "loss": 0.6493,
+      "step": 658
+    },
+    {
+      "epoch": 1.363448275862069,
+      "grad_norm": 0.8170894980430603,
+      "learning_rate": 6.6318898462867086e-06,
+      "loss": 0.674,
+      "step": 659
+    },
+    {
+      "epoch": 1.3655172413793104,
+      "grad_norm": 0.777654230594635,
+      "learning_rate": 6.6204987832389115e-06,
+      "loss": 0.7003,
+      "step": 660
+    },
+    {
+      "epoch": 1.3675862068965516,
+      "grad_norm": 0.7541632056236267,
+      "learning_rate": 6.609098314449116e-06,
+      "loss": 0.6488,
+      "step": 661
+    },
+    {
+      "epoch": 1.369655172413793,
+      "grad_norm": 0.806873083114624,
+      "learning_rate": 6.597688506088224e-06,
+      "loss": 0.6484,
+      "step": 662
+    },
+    {
+      "epoch": 1.3717241379310345,
+      "grad_norm": 0.7729332447052002,
+      "learning_rate": 6.586269424381349e-06,
+      "loss": 0.63,
+      "step": 663
+    },
+    {
+      "epoch": 1.3737931034482758,
+      "grad_norm": 0.7325103282928467,
+      "learning_rate": 6.574841135607426e-06,
+      "loss": 0.6377,
+      "step": 664
+    },
+    {
+      "epoch": 1.3758620689655172,
+      "grad_norm": 0.7824835777282715,
+      "learning_rate": 6.563403706098833e-06,
+      "loss": 0.6493,
+      "step": 665
+    },
+    {
+      "epoch": 1.3779310344827587,
+      "grad_norm": 0.8269271850585938,
+      "learning_rate": 6.551957202241001e-06,
+      "loss": 0.6763,
+      "step": 666
+    },
+    {
+      "epoch": 1.38,
+      "grad_norm": 0.7548254728317261,
+      "learning_rate": 6.540501690472032e-06,
+      "loss": 0.6463,
+      "step": 667
+    },
+    {
+      "epoch": 1.3820689655172413,
+      "grad_norm": 0.728628396987915,
+      "learning_rate": 6.529037237282309e-06,
+      "loss": 0.6571,
+      "step": 668
+    },
+    {
+      "epoch": 1.3841379310344828,
+      "grad_norm": 0.7487698793411255,
+      "learning_rate": 6.517563909214119e-06,
+      "loss": 0.6819,
+      "step": 669
+    },
+    {
+      "epoch": 1.386206896551724,
+      "grad_norm": 0.7777491807937622,
+      "learning_rate": 6.5060817728612544e-06,
+      "loss": 0.6384,
+      "step": 670
+    },
+    {
+      "epoch": 1.3882758620689655,
+      "grad_norm": 0.7405235171318054,
+      "learning_rate": 6.4945908948686355e-06,
+      "loss": 0.6637,
+      "step": 671
+    },
+    {
+      "epoch": 1.390344827586207,
+      "grad_norm": 0.730686366558075,
+      "learning_rate": 6.4830913419319205e-06,
+      "loss": 0.6264,
+      "step": 672
+    },
+    {
+      "epoch": 1.3924137931034481,
+      "grad_norm": 0.7911974191665649,
+      "learning_rate": 6.471583180797121e-06,
+      "loss": 0.6251,
+      "step": 673
+    },
+    {
+      "epoch": 1.3944827586206896,
+      "grad_norm": 0.7909592390060425,
+      "learning_rate": 6.460066478260209e-06,
+      "loss": 0.6639,
+      "step": 674
+    },
+    {
+      "epoch": 1.396551724137931,
+      "grad_norm": 0.7824447154998779,
+      "learning_rate": 6.448541301166737e-06,
+      "loss": 0.6771,
+      "step": 675
+    },
+    {
+      "epoch": 1.3986206896551723,
+      "grad_norm": 0.7654966711997986,
+      "learning_rate": 6.437007716411441e-06,
+      "loss": 0.6661,
+      "step": 676
+    },
+    {
+      "epoch": 1.4006896551724137,
+      "grad_norm": 0.7122900485992432,
+      "learning_rate": 6.4254657909378615e-06,
+      "loss": 0.6724,
+      "step": 677
+    },
+    {
+      "epoch": 1.4027586206896552,
+      "grad_norm": 0.7968724370002747,
+      "learning_rate": 6.4139155917379445e-06,
+      "loss": 0.6348,
+      "step": 678
+    },
+    {
+      "epoch": 1.4048275862068966,
+      "grad_norm": 0.7387704849243164,
+      "learning_rate": 6.402357185851669e-06,
+      "loss": 0.6537,
+      "step": 679
+    },
+    {
+      "epoch": 1.4068965517241379,
+      "grad_norm": 0.7359346747398376,
+      "learning_rate": 6.3907906403666355e-06,
+      "loss": 0.6649,
+      "step": 680
+    },
+    {
+      "epoch": 1.4089655172413793,
+      "grad_norm": 0.7153935432434082,
+      "learning_rate": 6.379216022417695e-06,
+      "loss": 0.6574,
+      "step": 681
+    },
+    {
+      "epoch": 1.4110344827586208,
+      "grad_norm": 0.7676519751548767,
+      "learning_rate": 6.367633399186554e-06,
+      "loss": 0.6811,
+      "step": 682
+    },
+    {
+      "epoch": 1.4131034482758622,
+      "grad_norm": 0.8044560551643372,
+      "learning_rate": 6.3560428379013795e-06,
+      "loss": 0.6496,
+      "step": 683
+    },
+    {
+      "epoch": 1.4151724137931034,
+      "grad_norm": 0.7096127867698669,
+      "learning_rate": 6.344444405836414e-06,
+      "loss": 0.6493,
+      "step": 684
+    },
+    {
+      "epoch": 1.4172413793103449,
+      "grad_norm": 0.7820800542831421,
+      "learning_rate": 6.332838170311586e-06,
+      "loss": 0.6561,
+      "step": 685
+    },
+    {
+      "epoch": 1.4193103448275863,
+      "grad_norm": 0.7546415328979492,
+      "learning_rate": 6.321224198692115e-06,
+      "loss": 0.672,
+      "step": 686
+    },
+    {
+      "epoch": 1.4213793103448276,
+      "grad_norm": 0.7297517657279968,
+      "learning_rate": 6.309602558388122e-06,
+      "loss": 0.6697,
+      "step": 687
+    },
+    {
+      "epoch": 1.423448275862069,
+      "grad_norm": 0.8077932000160217,
+      "learning_rate": 6.297973316854241e-06,
+      "loss": 0.6603,
+      "step": 688
+    },
+    {
+      "epoch": 1.4255172413793105,
+      "grad_norm": 0.7859988808631897,
+      "learning_rate": 6.286336541589224e-06,
+      "loss": 0.6724,
+      "step": 689
+    },
+    {
+      "epoch": 1.4275862068965517,
+      "grad_norm": 0.7754270434379578,
+      "learning_rate": 6.27469230013555e-06,
+      "loss": 0.6529,
+      "step": 690
+    },
+    {
+      "epoch": 1.4296551724137931,
+      "grad_norm": 0.7747129797935486,
+      "learning_rate": 6.263040660079031e-06,
+      "loss": 0.635,
+      "step": 691
+    },
+    {
+      "epoch": 1.4317241379310346,
+      "grad_norm": 0.7329117655754089,
+      "learning_rate": 6.25138168904843e-06,
+      "loss": 0.6489,
+      "step": 692
+    },
+    {
+      "epoch": 1.4337931034482758,
+      "grad_norm": 0.7887305021286011,
+      "learning_rate": 6.239715454715054e-06,
+      "loss": 0.6317,
+      "step": 693
+    },
+    {
+      "epoch": 1.4358620689655173,
+      "grad_norm": 0.7606728672981262,
+      "learning_rate": 6.22804202479237e-06,
+      "loss": 0.6604,
+      "step": 694
+    },
+    {
+      "epoch": 1.4379310344827587,
+      "grad_norm": 0.7513334155082703,
+      "learning_rate": 6.216361467035608e-06,
+      "loss": 0.6437,
+      "step": 695
+    },
+    {
+      "epoch": 1.44,
+      "grad_norm": 0.8091455698013306,
+      "learning_rate": 6.204673849241372e-06,
+      "loss": 0.6305,
+      "step": 696
+    },
+    {
+      "epoch": 1.4420689655172414,
+      "grad_norm": 0.7892200946807861,
+      "learning_rate": 6.192979239247243e-06,
+      "loss": 0.6549,
+      "step": 697
+    },
+    {
+      "epoch": 1.4441379310344828,
+      "grad_norm": 0.7235889434814453,
+      "learning_rate": 6.181277704931386e-06,
+      "loss": 0.651,
+      "step": 698
+    },
+    {
+      "epoch": 1.446206896551724,
+      "grad_norm": 0.7446338534355164,
+      "learning_rate": 6.169569314212157e-06,
+      "loss": 0.636,
+      "step": 699
+    },
+    {
+      "epoch": 1.4482758620689655,
+      "grad_norm": 0.7943323254585266,
+      "learning_rate": 6.1578541350477076e-06,
+      "loss": 0.6331,
+      "step": 700
+    },
+    {
+      "epoch": 1.450344827586207,
+      "grad_norm": 0.6900318264961243,
+      "learning_rate": 6.146132235435591e-06,
+      "loss": 0.6569,
+      "step": 701
+    },
+    {
+      "epoch": 1.4524137931034482,
+      "grad_norm": 0.7237994074821472,
+      "learning_rate": 6.1344036834123695e-06,
+      "loss": 0.6224,
+      "step": 702
+    },
+    {
+      "epoch": 1.4544827586206897,
+      "grad_norm": 0.7555676102638245,
+      "learning_rate": 6.1226685470532125e-06,
+      "loss": 0.6633,
+      "step": 703
+    },
+    {
+      "epoch": 1.456551724137931,
+      "grad_norm": 0.72216796875,
+      "learning_rate": 6.1109268944715125e-06,
+      "loss": 0.6551,
+      "step": 704
+    },
+    {
+      "epoch": 1.4586206896551723,
+      "grad_norm": 0.7122531533241272,
+      "learning_rate": 6.099178793818479e-06,
+      "loss": 0.6363,
+      "step": 705
+    },
+    {
+      "epoch": 1.4606896551724138,
+      "grad_norm": 0.7397927045822144,
+      "learning_rate": 6.0874243132827505e-06,
+      "loss": 0.6569,
+      "step": 706
+    },
+    {
+      "epoch": 1.4627586206896552,
+      "grad_norm": 0.729145348072052,
+      "learning_rate": 6.075663521089994e-06,
+      "loss": 0.628,
+      "step": 707
+    },
+    {
+      "epoch": 1.4648275862068965,
+      "grad_norm": 0.7200533151626587,
+      "learning_rate": 6.06389648550251e-06,
+      "loss": 0.6554,
+      "step": 708
+    },
+    {
+      "epoch": 1.466896551724138,
+      "grad_norm": 0.7103410959243774,
+      "learning_rate": 6.0521232748188416e-06,
+      "loss": 0.6481,
+      "step": 709
+    },
+    {
+      "epoch": 1.4689655172413794,
+      "grad_norm": 0.7074390053749084,
+      "learning_rate": 6.040343957373367e-06,
+      "loss": 0.6274,
+      "step": 710
+    },
+    {
+      "epoch": 1.4710344827586206,
+      "grad_norm": 0.734455406665802,
+      "learning_rate": 6.028558601535915e-06,
+      "loss": 0.661,
+      "step": 711
+    },
+    {
+      "epoch": 1.473103448275862,
+      "grad_norm": 0.7593621015548706,
+      "learning_rate": 6.016767275711359e-06,
+      "loss": 0.6339,
+      "step": 712
+    },
+    {
+      "epoch": 1.4751724137931035,
+      "grad_norm": 0.7968884706497192,
+      "learning_rate": 6.0049700483392256e-06,
+      "loss": 0.6564,
+      "step": 713
+    },
+    {
+      "epoch": 1.4772413793103447,
+      "grad_norm": 0.7559491991996765,
+      "learning_rate": 5.993166987893294e-06,
+      "loss": 0.6525,
+      "step": 714
+    },
+    {
+      "epoch": 1.4793103448275862,
+      "grad_norm": 0.7332967519760132,
+      "learning_rate": 5.981358162881202e-06,
+      "loss": 0.6591,
+      "step": 715
+    },
+    {
+      "epoch": 1.4813793103448276,
+      "grad_norm": 0.7570938467979431,
+      "learning_rate": 5.969543641844044e-06,
+      "loss": 0.6558,
+      "step": 716
+    },
+    {
+      "epoch": 1.4834482758620688,
+      "grad_norm": 0.7297980189323425,
+      "learning_rate": 5.957723493355977e-06,
+      "loss": 0.6396,
+      "step": 717
+    },
+    {
+      "epoch": 1.4855172413793103,
+      "grad_norm": 0.7381563782691956,
+      "learning_rate": 5.945897786023817e-06,
+      "loss": 0.6539,
+      "step": 718
+    },
+    {
+      "epoch": 1.4875862068965517,
+      "grad_norm": 0.7127878665924072,
+      "learning_rate": 5.9340665884866535e-06,
+      "loss": 0.6587,
+      "step": 719
+    },
+    {
+      "epoch": 1.489655172413793,
+      "grad_norm": 0.7142752408981323,
+      "learning_rate": 5.922229969415432e-06,
+      "loss": 0.6867,
+      "step": 720
+    },
+    {
+      "epoch": 1.4917241379310344,
+      "grad_norm": 0.7629202604293823,
+      "learning_rate": 5.910387997512573e-06,
+      "loss": 0.6381,
+      "step": 721
+    },
+    {
+      "epoch": 1.4937931034482759,
+      "grad_norm": 0.716674268245697,
+      "learning_rate": 5.898540741511564e-06,
+      "loss": 0.6406,
+      "step": 722
+    },
+    {
+      "epoch": 1.4958620689655173,
+      "grad_norm": 0.732208251953125,
+      "learning_rate": 5.8866882701765605e-06,
+      "loss": 0.6366,
+      "step": 723
+    },
+    {
+      "epoch": 1.4979310344827585,
+      "grad_norm": 0.7054344415664673,
+      "learning_rate": 5.8748306523019925e-06,
+      "loss": 0.6491,
+      "step": 724
+    },
+    {
+      "epoch": 1.5,
+      "grad_norm": 0.7091671228408813,
+      "learning_rate": 5.86296795671216e-06,
+      "loss": 0.6448,
+      "step": 725
+    },
+    {
+      "epoch": 1.5020689655172412,
+      "grad_norm": 0.7083839178085327,
+      "learning_rate": 5.851100252260835e-06,
+      "loss": 0.6302,
+      "step": 726
+    },
+    {
+      "epoch": 1.504137931034483,
+      "grad_norm": 0.727642834186554,
+      "learning_rate": 5.839227607830862e-06,
+      "loss": 0.6375,
+      "step": 727
+    },
+    {
+      "epoch": 1.5062068965517241,
+      "grad_norm": 0.7836138606071472,
+      "learning_rate": 5.827350092333758e-06,
+      "loss": 0.6429,
+      "step": 728
+    },
+    {
+      "epoch": 1.5082758620689654,
+      "grad_norm": 0.7098393440246582,
+      "learning_rate": 5.815467774709314e-06,
+      "loss": 0.6579,
+      "step": 729
+    },
+    {
+      "epoch": 1.510344827586207,
+      "grad_norm": 0.749321460723877,
+      "learning_rate": 5.803580723925193e-06,
+      "loss": 0.6361,
+      "step": 730
+    },
+    {
+      "epoch": 1.5124137931034483,
+      "grad_norm": 0.7477707862854004,
+      "learning_rate": 5.791689008976531e-06,
+      "loss": 0.6525,
+      "step": 731
+    },
+    {
+      "epoch": 1.5144827586206897,
+      "grad_norm": 0.7622563242912292,
+      "learning_rate": 5.779792698885534e-06,
+      "loss": 0.667,
+      "step": 732
+    },
+    {
+      "epoch": 1.5165517241379312,
+      "grad_norm": 0.7400147318840027,
+      "learning_rate": 5.767891862701081e-06,
+      "loss": 0.618,
+      "step": 733
+    },
+    {
+      "epoch": 1.5186206896551724,
+      "grad_norm": 0.7715649008750916,
+      "learning_rate": 5.755986569498321e-06,
+      "loss": 0.6501,
+      "step": 734
+    },
+    {
+      "epoch": 1.5206896551724138,
+      "grad_norm": 0.7207207679748535,
+      "learning_rate": 5.744076888378272e-06,
+      "loss": 0.6616,
+      "step": 735
+    },
+    {
+      "epoch": 1.5227586206896553,
+      "grad_norm": 0.7672625780105591,
+      "learning_rate": 5.732162888467421e-06,
+      "loss": 0.6117,
+      "step": 736
+    },
+    {
+      "epoch": 1.5248275862068965,
+      "grad_norm": 0.8607432246208191,
+      "learning_rate": 5.7202446389173225e-06,
+      "loss": 0.6272,
+      "step": 737
+    },
+    {
+      "epoch": 1.526896551724138,
+      "grad_norm": 0.8004416227340698,
+      "learning_rate": 5.708322208904196e-06,
+      "loss": 0.6538,
+      "step": 738
+    },
+    {
+      "epoch": 1.5289655172413794,
+      "grad_norm": 0.7796547412872314,
+      "learning_rate": 5.696395667628526e-06,
+      "loss": 0.6301,
+      "step": 739
+    },
+    {
+      "epoch": 1.5310344827586206,
+      "grad_norm": 0.7893610596656799,
+      "learning_rate": 5.6844650843146595e-06,
+      "loss": 0.6299,
+      "step": 740
+    },
+    {
+      "epoch": 1.533103448275862,
+      "grad_norm": 0.7845308184623718,
+      "learning_rate": 5.672530528210405e-06,
+      "loss": 0.6635,
+      "step": 741
+    },
+    {
+      "epoch": 1.5351724137931035,
+      "grad_norm": 0.7558560371398926,
+      "learning_rate": 5.660592068586629e-06,
+      "loss": 0.6637,
+      "step": 742
+    },
+    {
+      "epoch": 1.5372413793103448,
+      "grad_norm": 0.7075406908988953,
+      "learning_rate": 5.648649774736855e-06,
+      "loss": 0.6433,
+      "step": 743
+    },
+    {
+      "epoch": 1.5393103448275862,
+      "grad_norm": 0.7514592409133911,
+      "learning_rate": 5.6367037159768625e-06,
+      "loss": 0.6547,
+      "step": 744
+    },
+    {
+      "epoch": 1.5413793103448277,
+      "grad_norm": 0.7782887816429138,
+      "learning_rate": 5.624753961644281e-06,
+      "loss": 0.6541,
+      "step": 745
+    },
+    {
+      "epoch": 1.543448275862069,
+      "grad_norm": 0.7897704243659973,
+      "learning_rate": 5.612800581098193e-06,
+      "loss": 0.6582,
+      "step": 746
+    },
+    {
+      "epoch": 1.5455172413793103,
+      "grad_norm": 0.7592941522598267,
+      "learning_rate": 5.600843643718728e-06,
+      "loss": 0.6309,
+      "step": 747
+    },
+    {
+      "epoch": 1.5475862068965518,
+      "grad_norm": 0.7635563015937805,
+      "learning_rate": 5.588883218906653e-06,
+      "loss": 0.6407,
+      "step": 748
+    },
+    {
+      "epoch": 1.549655172413793,
+      "grad_norm": 0.7806128859519958,
+      "learning_rate": 5.57691937608299e-06,
+      "loss": 0.659,
+      "step": 749
+    },
+    {
+      "epoch": 1.5517241379310345,
+      "grad_norm": 0.7472797632217407,
+      "learning_rate": 5.564952184688588e-06,
+      "loss": 0.661,
+      "step": 750
+    },
+    {
+      "epoch": 1.553793103448276,
+      "grad_norm": 0.7442753314971924,
+      "learning_rate": 5.552981714183738e-06,
+      "loss": 0.6582,
+      "step": 751
+    },
+    {
+      "epoch": 1.5558620689655172,
+      "grad_norm": 0.7324591279029846,
+      "learning_rate": 5.5410080340477634e-06,
+      "loss": 0.6454,
+      "step": 752
+    },
+    {
+      "epoch": 1.5579310344827586,
+      "grad_norm": 0.7217804789543152,
+      "learning_rate": 5.529031213778615e-06,
+      "loss": 0.6441,
+      "step": 753
+    },
+    {
+      "epoch": 1.56,
+      "grad_norm": 0.7129271626472473,
+      "learning_rate": 5.517051322892468e-06,
+      "loss": 0.6642,
+      "step": 754
+    },
+    {
+      "epoch": 1.5620689655172413,
+      "grad_norm": 0.6992490291595459,
+      "learning_rate": 5.505068430923327e-06,
+      "loss": 0.6382,
+      "step": 755
+    },
+    {
+      "epoch": 1.5641379310344827,
+      "grad_norm": 0.7824325561523438,
+      "learning_rate": 5.4930826074226085e-06,
+      "loss": 0.6657,
+      "step": 756
+    },
+    {
+      "epoch": 1.5662068965517242,
+      "grad_norm": 0.7302783727645874,
+      "learning_rate": 5.481093921958749e-06,
+      "loss": 0.6643,
+      "step": 757
+    },
+    {
+      "epoch": 1.5682758620689654,
+      "grad_norm": 0.7289795875549316,
+      "learning_rate": 5.469102444116791e-06,
+      "loss": 0.6429,
+      "step": 758
+    },
+    {
+      "epoch": 1.5703448275862069,
+      "grad_norm": 0.7278138995170593,
+      "learning_rate": 5.457108243497993e-06,
+      "loss": 0.6518,
+      "step": 759
+    },
+    {
+      "epoch": 1.5724137931034483,
+      "grad_norm": 0.7361022233963013,
+      "learning_rate": 5.445111389719408e-06,
+      "loss": 0.6445,
+      "step": 760
+    },
+    {
+      "epoch": 1.5744827586206895,
+      "grad_norm": 0.7398816347122192,
+      "learning_rate": 5.433111952413496e-06,
+      "loss": 0.6601,
+      "step": 761
+    },
+    {
+      "epoch": 1.576551724137931,
+      "grad_norm": 0.7518286108970642,
+      "learning_rate": 5.421110001227705e-06,
+      "loss": 0.6119,
+      "step": 762
+    },
+    {
+      "epoch": 1.5786206896551724,
+      "grad_norm": 0.766607403755188,
+      "learning_rate": 5.409105605824082e-06,
+      "loss": 0.6555,
+      "step": 763
+    },
+    {
+      "epoch": 1.5806896551724137,
+      "grad_norm": 0.7912081480026245,
+      "learning_rate": 5.3970988358788565e-06,
+      "loss": 0.6652,
+      "step": 764
+    },
+    {
+      "epoch": 1.5827586206896553,
+      "grad_norm": 0.7829515337944031,
+      "learning_rate": 5.385089761082039e-06,
+      "loss": 0.6272,
+      "step": 765
+    },
+    {
+      "epoch": 1.5848275862068966,
+      "grad_norm": 0.7310935854911804,
+      "learning_rate": 5.3730784511370204e-06,
+      "loss": 0.6714,
+      "step": 766
+    },
+    {
+      "epoch": 1.5868965517241378,
+      "grad_norm": 0.7284607291221619,
+      "learning_rate": 5.361064975760166e-06,
+      "loss": 0.6563,
+      "step": 767
+    },
+    {
+      "epoch": 1.5889655172413795,
+      "grad_norm": 0.7712138891220093,
+      "learning_rate": 5.349049404680407e-06,
+      "loss": 0.6611,
+      "step": 768
+    },
+    {
+      "epoch": 1.5910344827586207,
+      "grad_norm": 0.7651541233062744,
+      "learning_rate": 5.3370318076388405e-06,
+      "loss": 0.6278,
+      "step": 769
+    },
+    {
+      "epoch": 1.593103448275862,
+      "grad_norm": 0.7919555902481079,
+      "learning_rate": 5.32501225438832e-06,
+      "loss": 0.6497,
+      "step": 770
+    },
+    {
+      "epoch": 1.5951724137931036,
+      "grad_norm": 0.6936086416244507,
+      "learning_rate": 5.3129908146930565e-06,
+      "loss": 0.6327,
+      "step": 771
+    },
+    {
+      "epoch": 1.5972413793103448,
+      "grad_norm": 0.7541401386260986,
+      "learning_rate": 5.300967558328208e-06,
+      "loss": 0.6141,
+      "step": 772
+    },
+    {
+      "epoch": 1.599310344827586,
+      "grad_norm": 0.785559892654419,
+      "learning_rate": 5.288942555079479e-06,
+      "loss": 0.6763,
+      "step": 773
+    },
+    {
+      "epoch": 1.6013793103448277,
+      "grad_norm": 0.7877768874168396,
+      "learning_rate": 5.2769158747427115e-06,
+      "loss": 0.6647,
+      "step": 774
+    },
+    {
+      "epoch": 1.603448275862069,
+      "grad_norm": 0.7559152245521545,
+      "learning_rate": 5.264887587123483e-06,
+      "loss": 0.6589,
+      "step": 775
+    },
+    {
+      "epoch": 1.6055172413793104,
+      "grad_norm": 0.7801457643508911,
+      "learning_rate": 5.2528577620366986e-06,
+      "loss": 0.6643,
+      "step": 776
+    },
+    {
+      "epoch": 1.6075862068965519,
+      "grad_norm": 0.7516634464263916,
+      "learning_rate": 5.240826469306187e-06,
+      "loss": 0.6471,
+      "step": 777
+    },
+    {
+      "epoch": 1.609655172413793,
+      "grad_norm": 0.7641648054122925,
+      "learning_rate": 5.228793778764297e-06,
+      "loss": 0.6796,
+      "step": 778
+    },
+    {
+      "epoch": 1.6117241379310345,
+      "grad_norm": 0.7582610845565796,
+      "learning_rate": 5.216759760251493e-06,
+      "loss": 0.6602,
+      "step": 779
+    },
+    {
+      "epoch": 1.613793103448276,
+      "grad_norm": 0.7532828450202942,
+      "learning_rate": 5.204724483615941e-06,
+      "loss": 0.6594,
+      "step": 780
+    },
+    {
+      "epoch": 1.6158620689655172,
+      "grad_norm": 0.774940550327301,
+      "learning_rate": 5.1926880187131134e-06,
+      "loss": 0.656,
+      "step": 781
+    },
+    {
+      "epoch": 1.6179310344827587,
+      "grad_norm": 0.7041785717010498,
+      "learning_rate": 5.18065043540538e-06,
+      "loss": 0.645,
+      "step": 782
+    },
+    {
+      "epoch": 1.62,
+      "grad_norm": 0.7526170611381531,
+      "learning_rate": 5.168611803561599e-06,
+      "loss": 0.658,
+      "step": 783
+    },
+    {
+      "epoch": 1.6220689655172413,
+      "grad_norm": 0.8174700736999512,
+      "learning_rate": 5.156572193056718e-06,
+      "loss": 0.6502,
+      "step": 784
+    },
+    {
+      "epoch": 1.6241379310344828,
+      "grad_norm": 0.7578688263893127,
+      "learning_rate": 5.144531673771364e-06,
+      "loss": 0.6664,
+      "step": 785
+    },
+    {
+      "epoch": 1.6262068965517242,
+      "grad_norm": 0.7004210948944092,
+      "learning_rate": 5.132490315591437e-06,
+      "loss": 0.6351,
+      "step": 786
+    },
+    {
+      "epoch": 1.6282758620689655,
+      "grad_norm": 0.7185120582580566,
+      "learning_rate": 5.1204481884077075e-06,
+      "loss": 0.6148,
+      "step": 787
+    },
+    {
+      "epoch": 1.630344827586207,
+      "grad_norm": 0.8190487623214722,
+      "learning_rate": 5.10840536211541e-06,
+      "loss": 0.6241,
+      "step": 788
+    },
+    {
+      "epoch": 1.6324137931034484,
+      "grad_norm": 0.8276094198226929,
+      "learning_rate": 5.096361906613836e-06,
+      "loss": 0.6241,
+      "step": 789
+    },
+    {
+      "epoch": 1.6344827586206896,
+      "grad_norm": 0.7432799339294434,
+      "learning_rate": 5.084317891805928e-06,
+      "loss": 0.6408,
+      "step": 790
+    },
+    {
+      "epoch": 1.636551724137931,
+      "grad_norm": 0.7277430891990662,
+      "learning_rate": 5.072273387597877e-06,
+      "loss": 0.6406,
+      "step": 791
+    },
+    {
+      "epoch": 1.6386206896551725,
+      "grad_norm": 0.7360126972198486,
+      "learning_rate": 5.0602284638987145e-06,
+      "loss": 0.6422,
+      "step": 792
+    },
+    {
+      "epoch": 1.6406896551724137,
+      "grad_norm": 0.8584643006324768,
+      "learning_rate": 5.048183190619904e-06,
+      "loss": 0.6329,
+      "step": 793
+    },
+    {
+      "epoch": 1.6427586206896552,
+      "grad_norm": 0.7649242281913757,
+      "learning_rate": 5.036137637674943e-06,
+      "loss": 0.6356,
+      "step": 794
+    },
+    {
+      "epoch": 1.6448275862068966,
+      "grad_norm": 0.7129114270210266,
+      "learning_rate": 5.0240918749789455e-06,
+      "loss": 0.6334,
+      "step": 795
+    },
+    {
+      "epoch": 1.6468965517241378,
+      "grad_norm": 0.7862198352813721,
+      "learning_rate": 5.01204597244825e-06,
+      "loss": 0.675,
+      "step": 796
+    },
+    {
+      "epoch": 1.6489655172413793,
+      "grad_norm": 0.7846953272819519,
+      "learning_rate": 5e-06,
+      "loss": 0.6573,
+      "step": 797
+    },
+    {
+      "epoch": 1.6510344827586207,
+      "grad_norm": 0.7969672083854675,
+      "learning_rate": 4.987954027551751e-06,
+      "loss": 0.652,
+      "step": 798
+    },
+    {
+      "epoch": 1.653103448275862,
+      "grad_norm": 0.6976673007011414,
+      "learning_rate": 4.975908125021055e-06,
+      "loss": 0.6251,
+      "step": 799
+    },
+    {
+      "epoch": 1.6551724137931034,
+      "grad_norm": 0.7470307946205139,
+      "learning_rate": 4.963862362325058e-06,
+      "loss": 0.6489,
+      "step": 800
+    },
+    {
+      "epoch": 1.6572413793103449,
+      "grad_norm": 0.7507964372634888,
+      "learning_rate": 4.951816809380098e-06,
+      "loss": 0.6355,
+      "step": 801
+    },
+    {
+      "epoch": 1.659310344827586,
+      "grad_norm": 0.7489534616470337,
+      "learning_rate": 4.939771536101286e-06,
+      "loss": 0.6555,
+      "step": 802
+    },
+    {
+      "epoch": 1.6613793103448276,
+      "grad_norm": 0.7056592106819153,
+      "learning_rate": 4.9277266124021245e-06,
+      "loss": 0.6497,
+      "step": 803
+    },
+    {
+      "epoch": 1.663448275862069,
+      "grad_norm": 0.7045999765396118,
+      "learning_rate": 4.915682108194073e-06,
+      "loss": 0.6547,
+      "step": 804
+    },
+    {
+      "epoch": 1.6655172413793102,
+      "grad_norm": 0.7271073460578918,
+      "learning_rate": 4.903638093386167e-06,
+      "loss": 0.6267,
+      "step": 805
+    },
+    {
+      "epoch": 1.6675862068965517,
+      "grad_norm": 0.7735942602157593,
+      "learning_rate": 4.891594637884591e-06,
+      "loss": 0.6369,
+      "step": 806
+    },
+    {
+      "epoch": 1.6696551724137931,
+      "grad_norm": 0.7303586006164551,
+      "learning_rate": 4.879551811592295e-06,
+      "loss": 0.6438,
+      "step": 807
+    },
+    {
+      "epoch": 1.6717241379310344,
+      "grad_norm": 0.734630286693573,
+      "learning_rate": 4.867509684408564e-06,
+      "loss": 0.6273,
+      "step": 808
+    },
+    {
+      "epoch": 1.6737931034482758,
+      "grad_norm": 0.7351275682449341,
+      "learning_rate": 4.855468326228638e-06,
+      "loss": 0.6628,
+      "step": 809
+    },
+    {
+      "epoch": 1.6758620689655173,
+      "grad_norm": 0.7179361581802368,
+      "learning_rate": 4.843427806943283e-06,
+      "loss": 0.6422,
+      "step": 810
+    },
+    {
+      "epoch": 1.6779310344827585,
+      "grad_norm": 0.6784632802009583,
+      "learning_rate": 4.831388196438402e-06,
+      "loss": 0.6407,
+      "step": 811
+    },
+    {
+      "epoch": 1.6800000000000002,
+      "grad_norm": 0.73552006483078,
+      "learning_rate": 4.819349564594622e-06,
+      "loss": 0.6448,
+      "step": 812
+    },
+    {
+      "epoch": 1.6820689655172414,
+      "grad_norm": 0.7421280145645142,
+      "learning_rate": 4.807311981286888e-06,
+      "loss": 0.6621,
+      "step": 813
+    },
+    {
+      "epoch": 1.6841379310344826,
+      "grad_norm": 0.6813828349113464,
+      "learning_rate": 4.79527551638406e-06,
+      "loss": 0.6442,
+      "step": 814
+    },
+    {
+      "epoch": 1.6862068965517243,
+      "grad_norm": 0.7029781937599182,
+      "learning_rate": 4.783240239748509e-06,
+      "loss": 0.6351,
+      "step": 815
+    },
+    {
+      "epoch": 1.6882758620689655,
+      "grad_norm": 0.6841586828231812,
+      "learning_rate": 4.7712062212357045e-06,
+      "loss": 0.6373,
+      "step": 816
+    },
+    {
+      "epoch": 1.6903448275862067,
+      "grad_norm": 0.7366838455200195,
+      "learning_rate": 4.7591735306938144e-06,
+      "loss": 0.6733,
+      "step": 817
+    },
+    {
+      "epoch": 1.6924137931034484,
+      "grad_norm": 0.7882569432258606,
+      "learning_rate": 4.747142237963305e-06,
+      "loss": 0.6266,
+      "step": 818
+    },
+    {
+      "epoch": 1.6944827586206896,
+      "grad_norm": 0.6938716173171997,
+      "learning_rate": 4.7351124128765185e-06,
+      "loss": 0.6318,
+      "step": 819
+    },
+    {
+      "epoch": 1.6965517241379309,
+      "grad_norm": 0.7524465918540955,
+      "learning_rate": 4.723084125257291e-06,
+      "loss": 0.6437,
+      "step": 820
+    },
+    {
+      "epoch": 1.6986206896551725,
+      "grad_norm": 0.7383887767791748,
+      "learning_rate": 4.711057444920522e-06,
+      "loss": 0.6525,
+      "step": 821
+    },
+    {
+      "epoch": 1.7006896551724138,
+      "grad_norm": 0.7824906706809998,
+      "learning_rate": 4.699032441671794e-06,
+      "loss": 0.6458,
+      "step": 822
+    },
+    {
+      "epoch": 1.7027586206896552,
+      "grad_norm": 0.7252950072288513,
+      "learning_rate": 4.687009185306945e-06,
+      "loss": 0.6536,
+      "step": 823
+    },
+    {
+      "epoch": 1.7048275862068967,
+      "grad_norm": 0.7240720391273499,
+      "learning_rate": 4.67498774561168e-06,
+      "loss": 0.6583,
+      "step": 824
+    },
+    {
+      "epoch": 1.706896551724138,
+      "grad_norm": 0.739566445350647,
+      "learning_rate": 4.662968192361161e-06,
+      "loss": 0.6758,
+      "step": 825
+    },
+    {
+      "epoch": 1.7089655172413794,
+      "grad_norm": 0.7385093569755554,
+      "learning_rate": 4.650950595319593e-06,
+      "loss": 0.6273,
+      "step": 826
+    },
+    {
+      "epoch": 1.7110344827586208,
+      "grad_norm": 0.7286804914474487,
+      "learning_rate": 4.6389350242398354e-06,
+      "loss": 0.6436,
+      "step": 827
+    },
+    {
+      "epoch": 1.713103448275862,
+      "grad_norm": 0.7011599540710449,
+      "learning_rate": 4.6269215488629795e-06,
+      "loss": 0.6532,
+      "step": 828
+    },
+    {
+      "epoch": 1.7151724137931035,
+      "grad_norm": 0.7191659212112427,
+      "learning_rate": 4.614910238917963e-06,
+      "loss": 0.6544,
+      "step": 829
+    },
+    {
+      "epoch": 1.717241379310345,
+      "grad_norm": 0.7047889232635498,
+      "learning_rate": 4.602901164121145e-06,
+      "loss": 0.6421,
+      "step": 830
+    },
+    {
+      "epoch": 1.7193103448275862,
+      "grad_norm": 0.7209073901176453,
+      "learning_rate": 4.59089439417592e-06,
+      "loss": 0.6289,
+      "step": 831
+    },
+    {
+      "epoch": 1.7213793103448276,
+      "grad_norm": 0.7176074385643005,
+      "learning_rate": 4.578889998772296e-06,
+      "loss": 0.642,
+      "step": 832
+    },
+    {
+      "epoch": 1.723448275862069,
+      "grad_norm": 0.7749189138412476,
+      "learning_rate": 4.5668880475865074e-06,
+      "loss": 0.6755,
+      "step": 833
+    },
+    {
+      "epoch": 1.7255172413793103,
+      "grad_norm": 0.689206600189209,
+      "learning_rate": 4.554888610280593e-06,
+      "loss": 0.6552,
+      "step": 834
+    },
+    {
+      "epoch": 1.7275862068965517,
+      "grad_norm": 0.7584705352783203,
+      "learning_rate": 4.542891756502008e-06,
+      "loss": 0.6109,
+      "step": 835
+    },
+    {
+      "epoch": 1.7296551724137932,
+      "grad_norm": 0.7560999393463135,
+      "learning_rate": 4.5308975558832095e-06,
+      "loss": 0.6619,
+      "step": 836
+    },
+    {
+      "epoch": 1.7317241379310344,
+      "grad_norm": 0.8325402140617371,
+      "learning_rate": 4.518906078041252e-06,
+      "loss": 0.6344,
+      "step": 837
+    },
+    {
+      "epoch": 1.7337931034482759,
+      "grad_norm": 0.7222752571105957,
+      "learning_rate": 4.506917392577393e-06,
+      "loss": 0.6255,
+      "step": 838
+    },
+    {
+      "epoch": 1.7358620689655173,
+      "grad_norm": 0.7850416898727417,
+      "learning_rate": 4.494931569076674e-06,
+      "loss": 0.6213,
+      "step": 839
+    },
+    {
+      "epoch": 1.7379310344827585,
+      "grad_norm": 0.7065498232841492,
+      "learning_rate": 4.482948677107533e-06,
+      "loss": 0.6146,
+      "step": 840
+    },
+    {
+      "epoch": 1.74,
+      "grad_norm": 0.6791179776191711,
+      "learning_rate": 4.4709687862213866e-06,
+      "loss": 0.6461,
+      "step": 841
+    },
+    {
+      "epoch": 1.7420689655172414,
+      "grad_norm": 0.750182032585144,
+      "learning_rate": 4.458991965952238e-06,
+      "loss": 0.652,
+      "step": 842
+    },
+    {
+      "epoch": 1.7441379310344827,
+      "grad_norm": 0.700097382068634,
+      "learning_rate": 4.447018285816263e-06,
+      "loss": 0.6437,
+      "step": 843
+    },
+    {
+      "epoch": 1.7462068965517241,
+      "grad_norm": 0.7411333322525024,
+      "learning_rate": 4.435047815311414e-06,
+      "loss": 0.6502,
+      "step": 844
+    },
+    {
+      "epoch": 1.7482758620689656,
+      "grad_norm": 0.7353593707084656,
+      "learning_rate": 4.423080623917012e-06,
+      "loss": 0.6502,
+      "step": 845
+    },
+    {
+      "epoch": 1.7503448275862068,
+      "grad_norm": 0.8049408793449402,
+      "learning_rate": 4.411116781093348e-06,
+      "loss": 0.6482,
+      "step": 846
+    },
+    {
+      "epoch": 1.7524137931034482,
+      "grad_norm": 0.7221590876579285,
+      "learning_rate": 4.399156356281274e-06,
+      "loss": 0.6549,
+      "step": 847
+    },
+    {
+      "epoch": 1.7544827586206897,
+      "grad_norm": 0.7125415802001953,
+      "learning_rate": 4.3871994189018075e-06,
+      "loss": 0.6409,
+      "step": 848
+    },
+    {
+      "epoch": 1.756551724137931,
+      "grad_norm": 0.7120684385299683,
+      "learning_rate": 4.3752460383557195e-06,
+      "loss": 0.6476,
+      "step": 849
+    },
+    {
+      "epoch": 1.7586206896551724,
+      "grad_norm": 0.7340549230575562,
+      "learning_rate": 4.363296284023139e-06,
+      "loss": 0.6606,
+      "step": 850
+    },
+    {
+      "epoch": 1.7606896551724138,
+      "grad_norm": 0.8672581911087036,
+      "learning_rate": 4.351350225263147e-06,
+      "loss": 0.6447,
+      "step": 851
+    },
+    {
+      "epoch": 1.762758620689655,
+      "grad_norm": 0.7489596009254456,
+      "learning_rate": 4.339407931413372e-06,
+      "loss": 0.634,
+      "step": 852
+    },
+    {
+      "epoch": 1.7648275862068965,
+      "grad_norm": 0.7057939171791077,
+      "learning_rate": 4.327469471789597e-06,
+      "loss": 0.6683,
+      "step": 853
+    },
+    {
+      "epoch": 1.766896551724138,
+      "grad_norm": 0.7291073799133301,
+      "learning_rate": 4.315534915685341e-06,
+      "loss": 0.6766,
+      "step": 854
+    },
+    {
+      "epoch": 1.7689655172413792,
+      "grad_norm": 0.7649169564247131,
+      "learning_rate": 4.303604332371476e-06,
+      "loss": 0.649,
+      "step": 855
+    },
+    {
+      "epoch": 1.7710344827586209,
+      "grad_norm": 0.7476722002029419,
+      "learning_rate": 4.291677791095805e-06,
+      "loss": 0.6111,
+      "step": 856
+    },
+    {
+      "epoch": 1.773103448275862,
+      "grad_norm": 0.7485122680664062,
+      "learning_rate": 4.27975536108268e-06,
+      "loss": 0.6375,
+      "step": 857
+    },
+    {
+      "epoch": 1.7751724137931033,
+      "grad_norm": 0.7042410373687744,
+      "learning_rate": 4.26783711153258e-06,
+      "loss": 0.6406,
+      "step": 858
+    },
+    {
+      "epoch": 1.777241379310345,
+      "grad_norm": 0.7430243492126465,
+      "learning_rate": 4.2559231116217305e-06,
+      "loss": 0.6511,
+      "step": 859
+    },
+    {
+      "epoch": 1.7793103448275862,
+      "grad_norm": 0.6917579174041748,
+      "learning_rate": 4.244013430501681e-06,
+      "loss": 0.6316,
+      "step": 860
+    },
+    {
+      "epoch": 1.7813793103448274,
+      "grad_norm": 0.6834431290626526,
+      "learning_rate": 4.2321081372989195e-06,
+      "loss": 0.6413,
+      "step": 861
+    },
+    {
+      "epoch": 1.783448275862069,
+      "grad_norm": 0.6975676417350769,
+      "learning_rate": 4.220207301114467e-06,
+      "loss": 0.6331,
+      "step": 862
+    },
+    {
+      "epoch": 1.7855172413793103,
+      "grad_norm": 0.7284821271896362,
+      "learning_rate": 4.208310991023469e-06,
+      "loss": 0.6431,
+      "step": 863
+    },
+    {
+      "epoch": 1.7875862068965516,
+      "grad_norm": 0.742372453212738,
+      "learning_rate": 4.1964192760748085e-06,
+      "loss": 0.6307,
+      "step": 864
+    },
+    {
+      "epoch": 1.7896551724137932,
+      "grad_norm": 0.7629566788673401,
+      "learning_rate": 4.184532225290687e-06,
+      "loss": 0.6186,
+      "step": 865
+    },
+    {
+      "epoch": 1.7917241379310345,
+      "grad_norm": 0.7036967277526855,
+      "learning_rate": 4.172649907666244e-06,
+      "loss": 0.6212,
+      "step": 866
+    },
+    {
+      "epoch": 1.793793103448276,
+      "grad_norm": 0.6951998472213745,
+      "learning_rate": 4.16077239216914e-06,
+      "loss": 0.6023,
+      "step": 867
+    },
+    {
+      "epoch": 1.7958620689655174,
+      "grad_norm": 0.728022575378418,
+      "learning_rate": 4.148899747739168e-06,
+      "loss": 0.6444,
+      "step": 868
+    },
+    {
+      "epoch": 1.7979310344827586,
+      "grad_norm": 0.6999719142913818,
+      "learning_rate": 4.137032043287841e-06,
+      "loss": 0.6368,
+      "step": 869
+    },
+    {
+      "epoch": 1.8,
+      "grad_norm": 0.7263539433479309,
+      "learning_rate": 4.125169347698009e-06,
+      "loss": 0.6584,
+      "step": 870
+    },
+    {
+      "epoch": 1.8020689655172415,
+      "grad_norm": 0.7751326560974121,
+      "learning_rate": 4.11331172982344e-06,
+      "loss": 0.6245,
+      "step": 871
+    },
+    {
+      "epoch": 1.8041379310344827,
+      "grad_norm": 0.7183523178100586,
+      "learning_rate": 4.101459258488438e-06,
+      "loss": 0.5938,
+      "step": 872
+    },
+    {
+      "epoch": 1.8062068965517242,
+      "grad_norm": 0.7222279906272888,
+      "learning_rate": 4.089612002487428e-06,
+      "loss": 0.6162,
+      "step": 873
+    },
+    {
+      "epoch": 1.8082758620689656,
+      "grad_norm": 0.7456408143043518,
+      "learning_rate": 4.077770030584569e-06,
+      "loss": 0.6602,
+      "step": 874
+    },
+    {
+      "epoch": 1.8103448275862069,
+      "grad_norm": 0.7470499873161316,
+      "learning_rate": 4.065933411513349e-06,
+      "loss": 0.653,
+      "step": 875
+    },
+    {
+      "epoch": 1.8124137931034483,
+      "grad_norm": 0.7037544846534729,
+      "learning_rate": 4.054102213976184e-06,
+      "loss": 0.6216,
+      "step": 876
+    },
+    {
+      "epoch": 1.8144827586206898,
+      "grad_norm": 0.7381490468978882,
+      "learning_rate": 4.042276506644024e-06,
+      "loss": 0.6422,
+      "step": 877
+    },
+    {
+      "epoch": 1.816551724137931,
+      "grad_norm": 0.7278012037277222,
+      "learning_rate": 4.030456358155957e-06,
+      "loss": 0.6475,
+      "step": 878
+    },
+    {
+      "epoch": 1.8186206896551724,
+      "grad_norm": 0.7335582971572876,
+      "learning_rate": 4.018641837118799e-06,
+      "loss": 0.643,
+      "step": 879
+    },
+    {
+      "epoch": 1.8206896551724139,
+      "grad_norm": 0.6921178698539734,
+      "learning_rate": 4.006833012106707e-06,
+      "loss": 0.6241,
+      "step": 880
+    },
+    {
+      "epoch": 1.822758620689655,
+      "grad_norm": 0.7078073024749756,
+      "learning_rate": 3.995029951660777e-06,
+      "loss": 0.6295,
+      "step": 881
+    },
+    {
+      "epoch": 1.8248275862068966,
+      "grad_norm": 0.6960477232933044,
+      "learning_rate": 3.983232724288642e-06,
+      "loss": 0.6596,
+      "step": 882
+    },
+    {
+      "epoch": 1.826896551724138,
+      "grad_norm": 0.7017378211021423,
+      "learning_rate": 3.971441398464088e-06,
+      "loss": 0.611,
+      "step": 883
+    },
+    {
+      "epoch": 1.8289655172413792,
+      "grad_norm": 0.729068398475647,
+      "learning_rate": 3.959656042626634e-06,
+      "loss": 0.6184,
+      "step": 884
+    },
+    {
+      "epoch": 1.8310344827586207,
+      "grad_norm": 0.7174490690231323,
+      "learning_rate": 3.94787672518116e-06,
+      "loss": 0.6383,
+      "step": 885
+    },
+    {
+      "epoch": 1.8331034482758621,
+      "grad_norm": 0.777277946472168,
+      "learning_rate": 3.9361035144974905e-06,
+      "loss": 0.6508,
+      "step": 886
+    },
+    {
+      "epoch": 1.8351724137931034,
+      "grad_norm": 0.7446042895317078,
+      "learning_rate": 3.924336478910007e-06,
+      "loss": 0.6588,
+      "step": 887
+    },
+    {
+      "epoch": 1.8372413793103448,
+      "grad_norm": 0.7238491773605347,
+      "learning_rate": 3.912575686717251e-06,
+      "loss": 0.6394,
+      "step": 888
+    },
+    {
+      "epoch": 1.8393103448275863,
+      "grad_norm": 0.7243683338165283,
+      "learning_rate": 3.900821206181521e-06,
+      "loss": 0.6529,
+      "step": 889
+    },
+    {
+      "epoch": 1.8413793103448275,
+      "grad_norm": 0.719829261302948,
+      "learning_rate": 3.889073105528489e-06,
+      "loss": 0.6535,
+      "step": 890
+    },
+    {
+      "epoch": 1.843448275862069,
+      "grad_norm": 0.7592584490776062,
+      "learning_rate": 3.8773314529467875e-06,
+      "loss": 0.6391,
+      "step": 891
+    },
+    {
+      "epoch": 1.8455172413793104,
+      "grad_norm": 0.7423802018165588,
+      "learning_rate": 3.865596316587633e-06,
+      "loss": 0.643,
+      "step": 892
+    },
+    {
+      "epoch": 1.8475862068965516,
+      "grad_norm": 0.7308037281036377,
+      "learning_rate": 3.853867764564409e-06,
+      "loss": 0.6329,
+      "step": 893
+    },
+    {
+      "epoch": 1.849655172413793,
+      "grad_norm": 0.7450445294380188,
+      "learning_rate": 3.842145864952295e-06,
+      "loss": 0.6516,
+      "step": 894
+    },
+    {
+      "epoch": 1.8517241379310345,
+      "grad_norm": 0.8020148277282715,
+      "learning_rate": 3.830430685787844e-06,
+      "loss": 0.6378,
+      "step": 895
+    },
+    {
+      "epoch": 1.8537931034482757,
+      "grad_norm": 0.6937201023101807,
+      "learning_rate": 3.818722295068616e-06,
+      "loss": 0.646,
+      "step": 896
+    },
+    {
+      "epoch": 1.8558620689655172,
+      "grad_norm": 0.7302286028862,
+      "learning_rate": 3.8070207607527587e-06,
+      "loss": 0.6489,
+      "step": 897
+    },
+    {
+      "epoch": 1.8579310344827586,
+      "grad_norm": 0.7561166286468506,
+      "learning_rate": 3.7953261507586282e-06,
+      "loss": 0.6249,
+      "step": 898
+    },
+    {
+      "epoch": 1.8599999999999999,
+      "grad_norm": 0.6970481276512146,
+      "learning_rate": 3.7836385329643933e-06,
+      "loss": 0.6355,
+      "step": 899
+    },
+    {
+      "epoch": 1.8620689655172413,
+      "grad_norm": 0.7228543162345886,
+      "learning_rate": 3.77195797520763e-06,
+      "loss": 0.6435,
+      "step": 900
+    },
+    {
+      "epoch": 1.8641379310344828,
+      "grad_norm": 0.7981168627738953,
+      "learning_rate": 3.760284545284947e-06,
+      "loss": 0.6294,
+      "step": 901
+    },
+    {
+      "epoch": 1.866206896551724,
+      "grad_norm": 0.758095383644104,
+      "learning_rate": 3.7486183109515696e-06,
+      "loss": 0.6652,
+      "step": 902
+    },
+    {
+      "epoch": 1.8682758620689657,
+      "grad_norm": 0.7205570936203003,
+      "learning_rate": 3.7369593399209704e-06,
+      "loss": 0.6236,
+      "step": 903
+    },
+    {
+      "epoch": 1.870344827586207,
+      "grad_norm": 0.7436397075653076,
+      "learning_rate": 3.725307699864452e-06,
+      "loss": 0.639,
+      "step": 904
+    },
+    {
+      "epoch": 1.8724137931034481,
+      "grad_norm": 0.7406452298164368,
+      "learning_rate": 3.7136634584107787e-06,
+      "loss": 0.6274,
+      "step": 905
+    },
+    {
+      "epoch": 1.8744827586206898,
+      "grad_norm": 0.7269127368927002,
+      "learning_rate": 3.7020266831457598e-06,
+      "loss": 0.6544,
+      "step": 906
+    },
+    {
+      "epoch": 1.876551724137931,
+      "grad_norm": 0.732607901096344,
+      "learning_rate": 3.6903974416118786e-06,
+      "loss": 0.6784,
+      "step": 907
+    },
+    {
+      "epoch": 1.8786206896551723,
+      "grad_norm": 0.7445371150970459,
+      "learning_rate": 3.6787758013078863e-06,
+      "loss": 0.6298,
+      "step": 908
+    },
+    {
+      "epoch": 1.880689655172414,
+      "grad_norm": 0.7168945670127869,
+      "learning_rate": 3.6671618296884147e-06,
+      "loss": 0.6331,
+      "step": 909
+    },
+    {
+      "epoch": 1.8827586206896552,
+      "grad_norm": 0.7478596568107605,
+      "learning_rate": 3.655555594163587e-06,
+      "loss": 0.6396,
+      "step": 910
+    },
+    {
+      "epoch": 1.8848275862068964,
+      "grad_norm": 0.7174217700958252,
+      "learning_rate": 3.6439571620986213e-06,
+      "loss": 0.6619,
+      "step": 911
+    },
+    {
+      "epoch": 1.886896551724138,
+      "grad_norm": 0.7665561437606812,
+      "learning_rate": 3.6323666008134465e-06,
+      "loss": 0.6513,
+      "step": 912
+    },
+    {
+      "epoch": 1.8889655172413793,
+      "grad_norm": 0.6955576539039612,
+      "learning_rate": 3.620783977582305e-06,
+      "loss": 0.6409,
+      "step": 913
+    },
+    {
+      "epoch": 1.8910344827586207,
+      "grad_norm": 0.6954272389411926,
+      "learning_rate": 3.6092093596333654e-06,
+      "loss": 0.6423,
+      "step": 914
+    },
+    {
+      "epoch": 1.8931034482758622,
+      "grad_norm": 0.744321882724762,
+      "learning_rate": 3.5976428141483326e-06,
+      "loss": 0.6142,
+      "step": 915
+    },
+    {
+      "epoch": 1.8951724137931034,
+      "grad_norm": 0.7438650131225586,
+      "learning_rate": 3.586084408262056e-06,
+      "loss": 0.6602,
+      "step": 916
+    },
+    {
+      "epoch": 1.8972413793103449,
+      "grad_norm": 0.6955951452255249,
+      "learning_rate": 3.5745342090621406e-06,
+      "loss": 0.6311,
+      "step": 917
+    },
+    {
+      "epoch": 1.8993103448275863,
+      "grad_norm": 0.6816698312759399,
+      "learning_rate": 3.5629922835885618e-06,
+      "loss": 0.6514,
+      "step": 918
+    },
+    {
+      "epoch": 1.9013793103448275,
+      "grad_norm": 0.6610486507415771,
+      "learning_rate": 3.551458698833265e-06,
+      "loss": 0.6401,
+      "step": 919
+    },
+    {
+      "epoch": 1.903448275862069,
+      "grad_norm": 0.7027138471603394,
+      "learning_rate": 3.5399335217397936e-06,
+      "loss": 0.6452,
+      "step": 920
+    },
+    {
+      "epoch": 1.9055172413793104,
+      "grad_norm": 0.715492844581604,
+      "learning_rate": 3.528416819202881e-06,
+      "loss": 0.6386,
+      "step": 921
+    },
+    {
+      "epoch": 1.9075862068965517,
+      "grad_norm": 0.7160024642944336,
+      "learning_rate": 3.5169086580680824e-06,
+      "loss": 0.6416,
+      "step": 922
+    },
+    {
+      "epoch": 1.9096551724137931,
+      "grad_norm": 0.7222967743873596,
+      "learning_rate": 3.5054091051313666e-06,
+      "loss": 0.6377,
+      "step": 923
+    },
+    {
+      "epoch": 1.9117241379310346,
+      "grad_norm": 0.6950160264968872,
+      "learning_rate": 3.493918227138746e-06,
+      "loss": 0.6458,
+      "step": 924
+    },
+    {
+      "epoch": 1.9137931034482758,
+      "grad_norm": 0.6922501921653748,
+      "learning_rate": 3.4824360907858824e-06,
+      "loss": 0.6528,
+      "step": 925
+    },
+    {
+      "epoch": 1.9158620689655173,
+      "grad_norm": 0.7411875128746033,
+      "learning_rate": 3.47096276271769e-06,
+      "loss": 0.6178,
+      "step": 926
+    },
+    {
+      "epoch": 1.9179310344827587,
+      "grad_norm": 0.6832643151283264,
+      "learning_rate": 3.45949830952797e-06,
+      "loss": 0.6365,
+      "step": 927
+    },
+    {
+      "epoch": 1.92,
+      "grad_norm": 0.7130672931671143,
+      "learning_rate": 3.4480427977590004e-06,
+      "loss": 0.6319,
+      "step": 928
+    },
+    {
+      "epoch": 1.9220689655172414,
+      "grad_norm": 0.6737676858901978,
+      "learning_rate": 3.43659629390117e-06,
+      "loss": 0.6644,
+      "step": 929
+    },
+    {
+      "epoch": 1.9241379310344828,
+      "grad_norm": 0.707264244556427,
+      "learning_rate": 3.4251588643925757e-06,
+      "loss": 0.659,
+      "step": 930
+    },
+    {
+      "epoch": 1.926206896551724,
+      "grad_norm": 0.687591016292572,
+      "learning_rate": 3.4137305756186543e-06,
+      "loss": 0.6175,
+      "step": 931
+    },
+    {
+      "epoch": 1.9282758620689655,
+      "grad_norm": 0.685309112071991,
+      "learning_rate": 3.4023114939117773e-06,
+      "loss": 0.6384,
+      "step": 932
+    },
+    {
+      "epoch": 1.930344827586207,
+      "grad_norm": 0.7224867343902588,
+      "learning_rate": 3.390901685550887e-06,
+      "loss": 0.6648,
+      "step": 933
+    },
+    {
+      "epoch": 1.9324137931034482,
+      "grad_norm": 0.7063190340995789,
+      "learning_rate": 3.3795012167610898e-06,
+      "loss": 0.6494,
+      "step": 934
+    },
+    {
+      "epoch": 1.9344827586206896,
+      "grad_norm": 0.7511458992958069,
+      "learning_rate": 3.3681101537132914e-06,
+      "loss": 0.6272,
+      "step": 935
+    },
+    {
+      "epoch": 1.936551724137931,
+      "grad_norm": 0.6883670091629028,
+      "learning_rate": 3.356728562523805e-06,
+      "loss": 0.6276,
+      "step": 936
+    },
+    {
+      "epoch": 1.9386206896551723,
+      "grad_norm": 0.7098174691200256,
+      "learning_rate": 3.3453565092539586e-06,
+      "loss": 0.6675,
+      "step": 937
+    },
+    {
+      "epoch": 1.9406896551724138,
+      "grad_norm": 0.7062398791313171,
+      "learning_rate": 3.33399405990973e-06,
+      "loss": 0.6408,
+      "step": 938
+    },
+    {
+      "epoch": 1.9427586206896552,
+      "grad_norm": 0.7228072881698608,
+      "learning_rate": 3.3226412804413444e-06,
+      "loss": 0.6422,
+      "step": 939
+    },
+    {
+      "epoch": 1.9448275862068964,
+      "grad_norm": 0.7137162685394287,
+      "learning_rate": 3.3112982367429105e-06,
+      "loss": 0.6474,
+      "step": 940
+    },
+    {
+      "epoch": 1.946896551724138,
+      "grad_norm": 0.6957323551177979,
+      "learning_rate": 3.299964994652017e-06,
+      "loss": 0.6492,
+      "step": 941
+    },
+    {
+      "epoch": 1.9489655172413793,
+      "grad_norm": 0.6673961281776428,
+      "learning_rate": 3.288641619949374e-06,
+      "loss": 0.6512,
+      "step": 942
+    },
+    {
+      "epoch": 1.9510344827586206,
+      "grad_norm": 0.7092306017875671,
+      "learning_rate": 3.2773281783584104e-06,
+      "loss": 0.6623,
+      "step": 943
+    },
+    {
+      "epoch": 1.953103448275862,
+      "grad_norm": 0.7391819953918457,
+      "learning_rate": 3.2660247355449037e-06,
+      "loss": 0.6698,
+      "step": 944
+    },
+    {
+      "epoch": 1.9551724137931035,
+      "grad_norm": 0.7239782214164734,
+      "learning_rate": 3.254731357116597e-06,
+      "loss": 0.6404,
+      "step": 945
+    },
+    {
+      "epoch": 1.9572413793103447,
+      "grad_norm": 0.735502302646637,
+      "learning_rate": 3.2434481086228163e-06,
+      "loss": 0.6278,
+      "step": 946
+    },
+    {
+      "epoch": 1.9593103448275864,
+      "grad_norm": 0.7158029079437256,
+      "learning_rate": 3.232175055554091e-06,
+      "loss": 0.6441,
+      "step": 947
+    },
+    {
+      "epoch": 1.9613793103448276,
+      "grad_norm": 0.7179998755455017,
+      "learning_rate": 3.2209122633417766e-06,
+      "loss": 0.6464,
+      "step": 948
+    },
+    {
+      "epoch": 1.9634482758620688,
+      "grad_norm": 0.6865692734718323,
+      "learning_rate": 3.209659797357669e-06,
+      "loss": 0.6373,
+      "step": 949
+    },
+    {
+      "epoch": 1.9655172413793105,
+      "grad_norm": 0.7228394150733948,
+      "learning_rate": 3.1984177229136287e-06,
+      "loss": 0.6583,
+      "step": 950
+    },
+    {
+      "epoch": 1.9675862068965517,
+      "grad_norm": 0.7167955636978149,
+      "learning_rate": 3.187186105261205e-06,
+      "loss": 0.6276,
+      "step": 951
+    },
+    {
+      "epoch": 1.969655172413793,
+      "grad_norm": 0.6937686204910278,
+      "learning_rate": 3.1759650095912474e-06,
+      "loss": 0.6203,
+      "step": 952
+    },
+    {
+      "epoch": 1.9717241379310346,
+      "grad_norm": 0.6669635772705078,
+      "learning_rate": 3.16475450103354e-06,
+      "loss": 0.6447,
+      "step": 953
+    },
+    {
+      "epoch": 1.9737931034482759,
+      "grad_norm": 0.7243427634239197,
+      "learning_rate": 3.1535546446564107e-06,
+      "loss": 0.6375,
+      "step": 954
+    },
+    {
+      "epoch": 1.975862068965517,
+      "grad_norm": 0.7239643931388855,
+      "learning_rate": 3.1423655054663665e-06,
+      "loss": 0.666,
+      "step": 955
+    },
+    {
+      "epoch": 1.9779310344827588,
+      "grad_norm": 0.7242691516876221,
+      "learning_rate": 3.1311871484077017e-06,
+      "loss": 0.6427,
+      "step": 956
+    },
+    {
+      "epoch": 1.98,
+      "grad_norm": 0.7019519209861755,
+      "learning_rate": 3.1200196383621363e-06,
+      "loss": 0.637,
+      "step": 957
+    },
+    {
+      "epoch": 1.9820689655172414,
+      "grad_norm": 0.7041625380516052,
+      "learning_rate": 3.1088630401484245e-06,
+      "loss": 0.6297,
+      "step": 958
+    },
+    {
+      "epoch": 1.9841379310344829,
+      "grad_norm": 0.7110829949378967,
+      "learning_rate": 3.0977174185219937e-06,
+      "loss": 0.6353,
+      "step": 959
+    },
+    {
+      "epoch": 1.986206896551724,
+      "grad_norm": 0.6561233401298523,
+      "learning_rate": 3.0865828381745515e-06,
+      "loss": 0.6184,
+      "step": 960
+    },
+    {
+      "epoch": 1.9882758620689656,
+      "grad_norm": 0.7260603904724121,
+      "learning_rate": 3.0754593637337276e-06,
+      "loss": 0.6435,
+      "step": 961
+    },
+    {
+      "epoch": 1.990344827586207,
+      "grad_norm": 0.7066130638122559,
+      "learning_rate": 3.0643470597626885e-06,
+      "loss": 0.6555,
+      "step": 962
+    },
+    {
+      "epoch": 1.9924137931034482,
+      "grad_norm": 0.6790770888328552,
+      "learning_rate": 3.05324599075976e-06,
+      "loss": 0.6216,
+      "step": 963
+    },
+    {
+      "epoch": 1.9944827586206897,
+      "grad_norm": 0.7200824022293091,
+      "learning_rate": 3.042156221158064e-06,
+      "loss": 0.6475,
+      "step": 964
+    },
+    {
+      "epoch": 1.9965517241379311,
+      "grad_norm": 0.6905136108398438,
+      "learning_rate": 3.0310778153251325e-06,
+      "loss": 0.6133,
+      "step": 965
+    },
+    {
+      "epoch": 1.9986206896551724,
+      "grad_norm": 0.7127056121826172,
+      "learning_rate": 3.020010837562546e-06,
+      "loss": 0.667,
+      "step": 966
+    },
+    {
+      "epoch": 2.0006896551724136,
+      "grad_norm": 0.6826462149620056,
+      "learning_rate": 3.0089553521055455e-06,
+      "loss": 0.6117,
+      "step": 967
+    },
+    {
+      "epoch": 2.0027586206896553,
+      "grad_norm": 0.803715169429779,
+      "learning_rate": 2.9979114231226776e-06,
+      "loss": 0.5496,
+      "step": 968
+    },
+    {
+      "epoch": 2.0048275862068965,
+      "grad_norm": 0.7920373678207397,
+      "learning_rate": 2.986879114715403e-06,
+      "loss": 0.5323,
+      "step": 969
+    },
+    {
+      "epoch": 2.0068965517241377,
+      "grad_norm": 0.7448180913925171,
+      "learning_rate": 2.9758584909177434e-06,
+      "loss": 0.5318,
+      "step": 970
+    },
+    {
+      "epoch": 2.0089655172413794,
+      "grad_norm": 0.7098000049591064,
+      "learning_rate": 2.964849615695892e-06,
+      "loss": 0.5341,
+      "step": 971
+    },
+    {
+      "epoch": 2.0110344827586206,
+      "grad_norm": 0.7263830304145813,
+      "learning_rate": 2.9538525529478557e-06,
+      "loss": 0.5605,
+      "step": 972
+    },
+    {
+      "epoch": 2.013103448275862,
+      "grad_norm": 0.7626384496688843,
+      "learning_rate": 2.9428673665030772e-06,
+      "loss": 0.5407,
+      "step": 973
+    },
+    {
+      "epoch": 2.0151724137931035,
+      "grad_norm": 0.8168633580207825,
+      "learning_rate": 2.9318941201220664e-06,
+      "loss": 0.5628,
+      "step": 974
+    },
+    {
+      "epoch": 2.0172413793103448,
+      "grad_norm": 0.8351486325263977,
+      "learning_rate": 2.9209328774960327e-06,
+      "loss": 0.5295,
+      "step": 975
+    },
+    {
+      "epoch": 2.0193103448275864,
+      "grad_norm": 0.7826777100563049,
+      "learning_rate": 2.90998370224651e-06,
+      "loss": 0.5141,
+      "step": 976
+    },
+    {
+      "epoch": 2.0213793103448277,
+      "grad_norm": 0.7777988314628601,
+      "learning_rate": 2.899046657924992e-06,
+      "loss": 0.5577,
+      "step": 977
+    },
+    {
+      "epoch": 2.023448275862069,
+      "grad_norm": 0.751868724822998,
+      "learning_rate": 2.8881218080125605e-06,
+      "loss": 0.5402,
+      "step": 978
+    },
+    {
+      "epoch": 2.0255172413793106,
+      "grad_norm": 0.6992483735084534,
+      "learning_rate": 2.8772092159195193e-06,
+      "loss": 0.5323,
+      "step": 979
+    },
+    {
+      "epoch": 2.027586206896552,
+      "grad_norm": 0.7283136248588562,
+      "learning_rate": 2.8663089449850246e-06,
+      "loss": 0.5364,
+      "step": 980
+    },
+    {
+      "epoch": 2.029655172413793,
+      "grad_norm": 0.7618082761764526,
+      "learning_rate": 2.8554210584767188e-06,
+      "loss": 0.5388,
+      "step": 981
+    },
+    {
+      "epoch": 2.0317241379310347,
+      "grad_norm": 0.7302858829498291,
+      "learning_rate": 2.844545619590361e-06,
+      "loss": 0.5247,
+      "step": 982
+    },
+    {
+      "epoch": 2.033793103448276,
+      "grad_norm": 0.706373393535614,
+      "learning_rate": 2.8336826914494607e-06,
+      "loss": 0.5138,
+      "step": 983
+    },
+    {
+      "epoch": 2.035862068965517,
+      "grad_norm": 0.6692239046096802,
+      "learning_rate": 2.822832337104915e-06,
+      "loss": 0.5363,
+      "step": 984
+    },
+    {
+      "epoch": 2.037931034482759,
+      "grad_norm": 0.7576499581336975,
+      "learning_rate": 2.8119946195346375e-06,
+      "loss": 0.5389,
+      "step": 985
+    },
+    {
+      "epoch": 2.04,
+      "grad_norm": 0.7035110592842102,
+      "learning_rate": 2.8011696016431967e-06,
+      "loss": 0.5333,
+      "step": 986
+    },
+    {
+      "epoch": 2.0420689655172413,
+      "grad_norm": 0.7431889176368713,
+      "learning_rate": 2.790357346261445e-06,
+      "loss": 0.5228,
+      "step": 987
+    },
+    {
+      "epoch": 2.044137931034483,
+      "grad_norm": 0.7341644167900085,
+      "learning_rate": 2.7795579161461685e-06,
+      "loss": 0.5153,
+      "step": 988
+    },
+    {
+      "epoch": 2.046206896551724,
+      "grad_norm": 0.7432145476341248,
+      "learning_rate": 2.7687713739796972e-06,
+      "loss": 0.5299,
+      "step": 989
+    },
+    {
+      "epoch": 2.0482758620689654,
+      "grad_norm": 0.6737946271896362,
+      "learning_rate": 2.7579977823695727e-06,
+      "loss": 0.5301,
+      "step": 990
+    },
+    {
+      "epoch": 2.050344827586207,
+      "grad_norm": 0.7434407472610474,
+      "learning_rate": 2.7472372038481575e-06,
+      "loss": 0.5442,
+      "step": 991
+    },
+    {
+      "epoch": 2.0524137931034483,
+      "grad_norm": 0.7228389382362366,
+      "learning_rate": 2.7364897008722867e-06,
+      "loss": 0.5392,
+      "step": 992
+    },
+    {
+      "epoch": 2.0544827586206895,
+      "grad_norm": 0.6981167793273926,
+      "learning_rate": 2.725755335822903e-06,
+      "loss": 0.5503,
+      "step": 993
+    },
+    {
+      "epoch": 2.056551724137931,
+      "grad_norm": 0.6900709867477417,
+      "learning_rate": 2.7150341710046924e-06,
+      "loss": 0.5448,
+      "step": 994
+    },
+    {
+      "epoch": 2.0586206896551724,
+      "grad_norm": 0.6905665993690491,
+      "learning_rate": 2.704326268645724e-06,
+      "loss": 0.5304,
+      "step": 995
+    },
+    {
+      "epoch": 2.0606896551724136,
+      "grad_norm": 0.720930814743042,
+      "learning_rate": 2.6936316908970874e-06,
+      "loss": 0.5304,
+      "step": 996
+    },
+    {
+      "epoch": 2.0627586206896553,
+      "grad_norm": 0.697242796421051,
+      "learning_rate": 2.6829504998325352e-06,
+      "loss": 0.5209,
+      "step": 997
+    },
+    {
+      "epoch": 2.0648275862068965,
+      "grad_norm": 0.6550352573394775,
+      "learning_rate": 2.67228275744812e-06,
+      "loss": 0.5319,
+      "step": 998
+    },
+    {
+      "epoch": 2.0668965517241378,
+      "grad_norm": 0.7028206586837769,
+      "learning_rate": 2.661628525661835e-06,
+      "loss": 0.5253,
+      "step": 999
+    },
+    {
+      "epoch": 2.0689655172413794,
+      "grad_norm": 0.7251700162887573,
+      "learning_rate": 2.6509878663132514e-06,
+      "loss": 0.5589,
+      "step": 1000
+    },
+    {
+      "epoch": 2.0710344827586207,
+      "grad_norm": 0.67062908411026,
+      "learning_rate": 2.6403608411631744e-06,
+      "loss": 0.5462,
+      "step": 1001
+    },
+    {
+      "epoch": 2.073103448275862,
+      "grad_norm": 0.7188374996185303,
+      "learning_rate": 2.629747511893255e-06,
+      "loss": 0.5201,
+      "step": 1002
+    },
+    {
+      "epoch": 2.0751724137931036,
+      "grad_norm": 0.6892343163490295,
+      "learning_rate": 2.61914794010567e-06,
+      "loss": 0.5357,
+      "step": 1003
+    },
+    {
+      "epoch": 2.077241379310345,
+      "grad_norm": 0.6836119294166565,
+      "learning_rate": 2.6085621873227264e-06,
+      "loss": 0.5123,
+      "step": 1004
+    },
+    {
+      "epoch": 2.079310344827586,
+      "grad_norm": 0.6911861896514893,
+      "learning_rate": 2.5979903149865386e-06,
+      "loss": 0.546,
+      "step": 1005
+    },
+    {
+      "epoch": 2.0813793103448277,
+      "grad_norm": 0.6947477459907532,
+      "learning_rate": 2.5874323844586414e-06,
+      "loss": 0.5408,
+      "step": 1006
+    },
+    {
+      "epoch": 2.083448275862069,
+      "grad_norm": 0.6671912670135498,
+      "learning_rate": 2.5768884570196615e-06,
+      "loss": 0.5256,
+      "step": 1007
+    },
+    {
+      "epoch": 2.08551724137931,
+      "grad_norm": 0.6696575284004211,
+      "learning_rate": 2.566358593868936e-06,
+      "loss": 0.5186,
+      "step": 1008
+    },
+    {
+      "epoch": 2.087586206896552,
+      "grad_norm": 0.6655759215354919,
+      "learning_rate": 2.555842856124182e-06,
+      "loss": 0.5686,
+      "step": 1009
+    },
+    {
+      "epoch": 2.089655172413793,
+      "grad_norm": 0.685197114944458,
+      "learning_rate": 2.5453413048211175e-06,
+      "loss": 0.4912,
+      "step": 1010
+    },
+    {
+      "epoch": 2.0917241379310343,
+      "grad_norm": 0.7344222664833069,
+      "learning_rate": 2.5348540009131283e-06,
+      "loss": 0.5489,
+      "step": 1011
+    },
+    {
+      "epoch": 2.093793103448276,
+      "grad_norm": 0.7047103047370911,
+      "learning_rate": 2.5243810052709006e-06,
+      "loss": 0.5366,
+      "step": 1012
+    },
+    {
+      "epoch": 2.095862068965517,
+      "grad_norm": 0.6715806722640991,
+      "learning_rate": 2.513922378682075e-06,
+      "loss": 0.5337,
+      "step": 1013
+    },
+    {
+      "epoch": 2.0979310344827584,
+      "grad_norm": 0.6645084619522095,
+      "learning_rate": 2.5034781818508867e-06,
+      "loss": 0.5451,
+      "step": 1014
+    },
+    {
+      "epoch": 2.1,
+      "grad_norm": 0.6680448651313782,
+      "learning_rate": 2.4930484753978224e-06,
+      "loss": 0.5141,
+      "step": 1015
+    },
+    {
+      "epoch": 2.1020689655172413,
+      "grad_norm": 0.6710419058799744,
+      "learning_rate": 2.482633319859259e-06,
+      "loss": 0.5233,
+      "step": 1016
+    },
+    {
+      "epoch": 2.1041379310344825,
+      "grad_norm": 0.6609877943992615,
+      "learning_rate": 2.472232775687119e-06,
+      "loss": 0.5343,
+      "step": 1017
+    },
+    {
+      "epoch": 2.106206896551724,
+      "grad_norm": 0.6360195279121399,
+      "learning_rate": 2.4618469032485164e-06,
+      "loss": 0.5549,
+      "step": 1018
+    },
+    {
+      "epoch": 2.1082758620689654,
+      "grad_norm": 0.6452751159667969,
+      "learning_rate": 2.4514757628254075e-06,
+      "loss": 0.5369,
+      "step": 1019
+    },
+    {
+      "epoch": 2.110344827586207,
+      "grad_norm": 0.6557055115699768,
+      "learning_rate": 2.441119414614241e-06,
+      "loss": 0.5277,
+      "step": 1020
+    },
+    {
+      "epoch": 2.1124137931034483,
+      "grad_norm": 0.7097801566123962,
+      "learning_rate": 2.4307779187256064e-06,
+      "loss": 0.5358,
+      "step": 1021
+    },
+    {
+      "epoch": 2.1144827586206896,
+      "grad_norm": 0.6572842597961426,
+      "learning_rate": 2.4204513351838897e-06,
+      "loss": 0.5394,
+      "step": 1022
+    },
+    {
+      "epoch": 2.1165517241379312,
+      "grad_norm": 0.6733007431030273,
+      "learning_rate": 2.4101397239269202e-06,
+      "loss": 0.5579,
+      "step": 1023
+    },
+    {
+      "epoch": 2.1186206896551725,
+      "grad_norm": 0.6480697393417358,
+      "learning_rate": 2.3998431448056237e-06,
+      "loss": 0.5635,
+      "step": 1024
+    },
+    {
+      "epoch": 2.1206896551724137,
+      "grad_norm": 0.6844390034675598,
+      "learning_rate": 2.389561657583681e-06,
+      "loss": 0.5477,
+      "step": 1025
+    },
+    {
+      "epoch": 2.1227586206896554,
+      "grad_norm": 0.6834397912025452,
+      "learning_rate": 2.3792953219371713e-06,
+      "loss": 0.5252,
+      "step": 1026
+    },
+    {
+      "epoch": 2.1248275862068966,
+      "grad_norm": 0.666690468788147,
+      "learning_rate": 2.3690441974542322e-06,
+      "loss": 0.574,
+      "step": 1027
+    },
+    {
+      "epoch": 2.126896551724138,
+      "grad_norm": 0.6561058759689331,
+      "learning_rate": 2.358808343634712e-06,
+      "loss": 0.5291,
+      "step": 1028
+    },
+    {
+      "epoch": 2.1289655172413795,
+      "grad_norm": 0.6525347828865051,
+      "learning_rate": 2.3485878198898253e-06,
+      "loss": 0.5626,
+      "step": 1029
+    },
+    {
+      "epoch": 2.1310344827586207,
+      "grad_norm": 0.6866719126701355,
+      "learning_rate": 2.3383826855418064e-06,
+      "loss": 0.524,
+      "step": 1030
+    },
+    {
+      "epoch": 2.133103448275862,
+      "grad_norm": 0.6798598170280457,
+      "learning_rate": 2.3281929998235664e-06,
+      "loss": 0.5184,
+      "step": 1031
+    },
+    {
+      "epoch": 2.1351724137931036,
+      "grad_norm": 0.6799730658531189,
+      "learning_rate": 2.31801882187835e-06,
+      "loss": 0.5086,
+      "step": 1032
+    },
+    {
+      "epoch": 2.137241379310345,
+      "grad_norm": 0.6736025214195251,
+      "learning_rate": 2.30786021075939e-06,
+      "loss": 0.5333,
+      "step": 1033
+    },
+    {
+      "epoch": 2.139310344827586,
+      "grad_norm": 0.6704842448234558,
+      "learning_rate": 2.297717225429567e-06,
+      "loss": 0.5065,
+      "step": 1034
+    },
+    {
+      "epoch": 2.1413793103448278,
+      "grad_norm": 0.6423225998878479,
+      "learning_rate": 2.287589924761065e-06,
+      "loss": 0.5292,
+      "step": 1035
+    },
+    {
+      "epoch": 2.143448275862069,
+      "grad_norm": 0.6814275979995728,
+      "learning_rate": 2.2774783675350303e-06,
+      "loss": 0.5418,
+      "step": 1036
+    },
+    {
+      "epoch": 2.14551724137931,
+      "grad_norm": 0.6510114073753357,
+      "learning_rate": 2.2673826124412314e-06,
+      "loss": 0.5175,
+      "step": 1037
+    },
+    {
+      "epoch": 2.147586206896552,
+      "grad_norm": 0.6611030697822571,
+      "learning_rate": 2.257302718077721e-06,
+      "loss": 0.5526,
+      "step": 1038
+    },
+    {
+      "epoch": 2.149655172413793,
+      "grad_norm": 0.6714975833892822,
+      "learning_rate": 2.247238742950483e-06,
+      "loss": 0.524,
+      "step": 1039
+    },
+    {
+      "epoch": 2.1517241379310343,
+      "grad_norm": 0.6750874519348145,
+      "learning_rate": 2.2371907454731168e-06,
+      "loss": 0.5529,
+      "step": 1040
+    },
+    {
+      "epoch": 2.153793103448276,
+      "grad_norm": 0.6794918179512024,
+      "learning_rate": 2.2271587839664673e-06,
+      "loss": 0.5034,
+      "step": 1041
+    },
+    {
+      "epoch": 2.1558620689655172,
+      "grad_norm": 0.6698225140571594,
+      "learning_rate": 2.217142916658318e-06,
+      "loss": 0.531,
+      "step": 1042
+    },
+    {
+      "epoch": 2.1579310344827585,
+      "grad_norm": 0.6470540761947632,
+      "learning_rate": 2.2071432016830257e-06,
+      "loss": 0.5631,
+      "step": 1043
+    },
+    {
+      "epoch": 2.16,
+      "grad_norm": 0.7159327268600464,
+      "learning_rate": 2.197159697081208e-06,
+      "loss": 0.5476,
+      "step": 1044
+    },
+    {
+      "epoch": 2.1620689655172414,
+      "grad_norm": 0.7387550473213196,
+      "learning_rate": 2.18719246079938e-06,
+      "loss": 0.5019,
+      "step": 1045
+    },
+    {
+      "epoch": 2.1641379310344826,
+      "grad_norm": 0.7351087331771851,
+      "learning_rate": 2.1772415506896447e-06,
+      "loss": 0.5607,
+      "step": 1046
+    },
+    {
+      "epoch": 2.1662068965517243,
+      "grad_norm": 0.6578730940818787,
+      "learning_rate": 2.1673070245093324e-06,
+      "loss": 0.5494,
+      "step": 1047
+    },
+    {
+      "epoch": 2.1682758620689655,
+      "grad_norm": 0.675528347492218,
+      "learning_rate": 2.157388939920689e-06,
+      "loss": 0.526,
+      "step": 1048
+    },
+    {
+      "epoch": 2.1703448275862067,
+      "grad_norm": 0.6719648838043213,
+      "learning_rate": 2.1474873544905204e-06,
+      "loss": 0.5334,
+      "step": 1049
+    },
+    {
+      "epoch": 2.1724137931034484,
+      "grad_norm": 0.6997095346450806,
+      "learning_rate": 2.137602325689873e-06,
+      "loss": 0.5274,
+      "step": 1050
+    },
+    {
+      "epoch": 2.1744827586206896,
+      "grad_norm": 0.7052441835403442,
+      "learning_rate": 2.127733910893693e-06,
+      "loss": 0.5378,
+      "step": 1051
+    },
+    {
+      "epoch": 2.176551724137931,
+      "grad_norm": 0.6952463388442993,
+      "learning_rate": 2.1178821673804943e-06,
+      "loss": 0.5244,
+      "step": 1052
+    },
+    {
+      "epoch": 2.1786206896551725,
+      "grad_norm": 0.664283037185669,
+      "learning_rate": 2.108047152332028e-06,
+      "loss": 0.5335,
+      "step": 1053
+    },
+    {
+      "epoch": 2.1806896551724138,
+      "grad_norm": 0.6841456294059753,
+      "learning_rate": 2.0982289228329495e-06,
+      "loss": 0.5367,
+      "step": 1054
+    },
+    {
+      "epoch": 2.182758620689655,
+      "grad_norm": 0.6873970031738281,
+      "learning_rate": 2.0884275358704863e-06,
+      "loss": 0.5242,
+      "step": 1055
+    },
+    {
+      "epoch": 2.1848275862068967,
+      "grad_norm": 0.704255998134613,
+      "learning_rate": 2.078643048334109e-06,
+      "loss": 0.5369,
+      "step": 1056
+    },
+    {
+      "epoch": 2.186896551724138,
+      "grad_norm": 0.6614420413970947,
+      "learning_rate": 2.0688755170152e-06,
+      "loss": 0.5397,
+      "step": 1057
+    },
+    {
+      "epoch": 2.188965517241379,
+      "grad_norm": 0.65423583984375,
+      "learning_rate": 2.0591249986067214e-06,
+      "loss": 0.5374,
+      "step": 1058
+    },
+    {
+      "epoch": 2.191034482758621,
+      "grad_norm": 0.6417624950408936,
+      "learning_rate": 2.0493915497028933e-06,
+      "loss": 0.5238,
+      "step": 1059
+    },
+    {
+      "epoch": 2.193103448275862,
+      "grad_norm": 0.6377779841423035,
+      "learning_rate": 2.039675226798854e-06,
+      "loss": 0.5377,
+      "step": 1060
+    },
+    {
+      "epoch": 2.1951724137931032,
+      "grad_norm": 0.6852798461914062,
+      "learning_rate": 2.029976086290347e-06,
+      "loss": 0.5386,
+      "step": 1061
+    },
+    {
+      "epoch": 2.197241379310345,
+      "grad_norm": 0.6332745552062988,
+      "learning_rate": 2.0202941844733792e-06,
+      "loss": 0.5541,
+      "step": 1062
+    },
+    {
+      "epoch": 2.199310344827586,
+      "grad_norm": 0.6756484508514404,
+      "learning_rate": 2.0106295775439018e-06,
+      "loss": 0.5092,
+      "step": 1063
+    },
+    {
+      "epoch": 2.2013793103448274,
+      "grad_norm": 0.6440528035163879,
+      "learning_rate": 2.0009823215974815e-06,
+      "loss": 0.5256,
+      "step": 1064
+    },
+    {
+      "epoch": 2.203448275862069,
+      "grad_norm": 0.6707040071487427,
+      "learning_rate": 1.9913524726289784e-06,
+      "loss": 0.5317,
+      "step": 1065
+    },
+    {
+      "epoch": 2.2055172413793103,
+      "grad_norm": 0.6765915155410767,
+      "learning_rate": 1.9817400865322166e-06,
+      "loss": 0.5319,
+      "step": 1066
+    },
+    {
+      "epoch": 2.2075862068965515,
+      "grad_norm": 0.6627142429351807,
+      "learning_rate": 1.972145219099662e-06,
+      "loss": 0.5246,
+      "step": 1067
+    },
+    {
+      "epoch": 2.209655172413793,
+      "grad_norm": 0.650766909122467,
+      "learning_rate": 1.962567926022099e-06,
+      "loss": 0.541,
+      "step": 1068
+    },
+    {
+      "epoch": 2.2117241379310344,
+      "grad_norm": 0.6753216981887817,
+      "learning_rate": 1.9530082628883058e-06,
+      "loss": 0.5472,
+      "step": 1069
+    },
+    {
+      "epoch": 2.213793103448276,
+      "grad_norm": 0.6764883399009705,
+      "learning_rate": 1.9434662851847325e-06,
+      "loss": 0.522,
+      "step": 1070
+    },
+    {
+      "epoch": 2.2158620689655173,
+      "grad_norm": 0.6282169818878174,
+      "learning_rate": 1.9339420482951794e-06,
+      "loss": 0.5127,
+      "step": 1071
+    },
+    {
+      "epoch": 2.2179310344827585,
+      "grad_norm": 0.6872683763504028,
+      "learning_rate": 1.9244356075004743e-06,
+      "loss": 0.5661,
+      "step": 1072
+    },
+    {
+      "epoch": 2.22,
+      "grad_norm": 0.6766654849052429,
+      "learning_rate": 1.914947017978153e-06,
+      "loss": 0.543,
+      "step": 1073
+    },
+    {
+      "epoch": 2.2220689655172414,
+      "grad_norm": 0.6761453747749329,
+      "learning_rate": 1.9054763348021372e-06,
+      "loss": 0.5139,
+      "step": 1074
+    },
+    {
+      "epoch": 2.2241379310344827,
+      "grad_norm": 0.6696473360061646,
+      "learning_rate": 1.896023612942422e-06,
+      "loss": 0.5224,
+      "step": 1075
+    },
+    {
+      "epoch": 2.2262068965517243,
+      "grad_norm": 0.7067750096321106,
+      "learning_rate": 1.8865889072647386e-06,
+      "loss": 0.5376,
+      "step": 1076
+    },
+    {
+      "epoch": 2.2282758620689656,
+      "grad_norm": 0.654654324054718,
+      "learning_rate": 1.8771722725302644e-06,
+      "loss": 0.5438,
+      "step": 1077
+    },
+    {
+      "epoch": 2.230344827586207,
+      "grad_norm": 0.6717188358306885,
+      "learning_rate": 1.8677737633952715e-06,
+      "loss": 0.5354,
+      "step": 1078
+    },
+    {
+      "epoch": 2.2324137931034485,
+      "grad_norm": 0.6855020523071289,
+      "learning_rate": 1.8583934344108446e-06,
+      "loss": 0.5294,
+      "step": 1079
+    },
+    {
+      "epoch": 2.2344827586206897,
+      "grad_norm": 0.6623315811157227,
+      "learning_rate": 1.8490313400225296e-06,
+      "loss": 0.5317,
+      "step": 1080
+    },
+    {
+      "epoch": 2.236551724137931,
+      "grad_norm": 0.687870979309082,
+      "learning_rate": 1.8396875345700498e-06,
+      "loss": 0.5539,
+      "step": 1081
+    },
+    {
+      "epoch": 2.2386206896551726,
+      "grad_norm": 0.6914966106414795,
+      "learning_rate": 1.830362072286963e-06,
+      "loss": 0.5537,
+      "step": 1082
+    },
+    {
+      "epoch": 2.240689655172414,
+      "grad_norm": 0.7108622193336487,
+      "learning_rate": 1.8210550073003701e-06,
+      "loss": 0.5252,
+      "step": 1083
+    },
+    {
+      "epoch": 2.242758620689655,
+      "grad_norm": 0.6858078241348267,
+      "learning_rate": 1.8117663936305785e-06,
+      "loss": 0.5542,
+      "step": 1084
+    },
+    {
+      "epoch": 2.2448275862068967,
+      "grad_norm": 0.6860451698303223,
+      "learning_rate": 1.8024962851908106e-06,
+      "loss": 0.5246,
+      "step": 1085
+    },
+    {
+      "epoch": 2.246896551724138,
+      "grad_norm": 0.6590988636016846,
+      "learning_rate": 1.7932447357868737e-06,
+      "loss": 0.5328,
+      "step": 1086
+    },
+    {
+      "epoch": 2.248965517241379,
+      "grad_norm": 0.7076303958892822,
+      "learning_rate": 1.7840117991168566e-06,
+      "loss": 0.5205,
+      "step": 1087
+    },
+    {
+      "epoch": 2.251034482758621,
+      "grad_norm": 0.6729241013526917,
+      "learning_rate": 1.7747975287708147e-06,
+      "loss": 0.5322,
+      "step": 1088
+    },
+    {
+      "epoch": 2.253103448275862,
+      "grad_norm": 0.6530643701553345,
+      "learning_rate": 1.7656019782304602e-06,
+      "loss": 0.552,
+      "step": 1089
+    },
+    {
+      "epoch": 2.2551724137931033,
+      "grad_norm": 0.6714941263198853,
+      "learning_rate": 1.7564252008688514e-06,
+      "loss": 0.5344,
+      "step": 1090
+    },
+    {
+      "epoch": 2.257241379310345,
+      "grad_norm": 0.6543527841567993,
+      "learning_rate": 1.7472672499500837e-06,
+      "loss": 0.5282,
+      "step": 1091
+    },
+    {
+      "epoch": 2.259310344827586,
+      "grad_norm": 0.6821870803833008,
+      "learning_rate": 1.7381281786289767e-06,
+      "loss": 0.5009,
+      "step": 1092
+    },
+    {
+      "epoch": 2.2613793103448274,
+      "grad_norm": 0.6596633791923523,
+      "learning_rate": 1.729008039950772e-06,
+      "loss": 0.546,
+      "step": 1093
+    },
+    {
+      "epoch": 2.263448275862069,
+      "grad_norm": 0.6522555351257324,
+      "learning_rate": 1.71990688685082e-06,
+      "loss": 0.5464,
+      "step": 1094
+    },
+    {
+      "epoch": 2.2655172413793103,
+      "grad_norm": 0.668368399143219,
+      "learning_rate": 1.7108247721542725e-06,
+      "loss": 0.528,
+      "step": 1095
+    },
+    {
+      "epoch": 2.2675862068965515,
+      "grad_norm": 0.697405219078064,
+      "learning_rate": 1.701761748575786e-06,
+      "loss": 0.5657,
+      "step": 1096
+    },
+    {
+      "epoch": 2.269655172413793,
+      "grad_norm": 0.6927202939987183,
+      "learning_rate": 1.6927178687191953e-06,
+      "loss": 0.518,
+      "step": 1097
+    },
+    {
+      "epoch": 2.2717241379310344,
+      "grad_norm": 0.658198356628418,
+      "learning_rate": 1.6836931850772331e-06,
+      "loss": 0.5482,
+      "step": 1098
+    },
+    {
+      "epoch": 2.2737931034482757,
+      "grad_norm": 0.688456654548645,
+      "learning_rate": 1.6746877500312054e-06,
+      "loss": 0.5114,
+      "step": 1099
+    },
+    {
+      "epoch": 2.2758620689655173,
+      "grad_norm": 0.6654470562934875,
+      "learning_rate": 1.6657016158506966e-06,
+      "loss": 0.5417,
+      "step": 1100
+    },
+    {
+      "epoch": 2.2779310344827586,
+      "grad_norm": 0.6720593571662903,
+      "learning_rate": 1.656734834693266e-06,
+      "loss": 0.5243,
+      "step": 1101
+    },
+    {
+      "epoch": 2.2800000000000002,
+      "grad_norm": 0.6629576086997986,
+      "learning_rate": 1.6477874586041415e-06,
+      "loss": 0.5244,
+      "step": 1102
+    },
+    {
+      "epoch": 2.2820689655172415,
+      "grad_norm": 0.6911601424217224,
+      "learning_rate": 1.6388595395159207e-06,
+      "loss": 0.5541,
+      "step": 1103
+    },
+    {
+      "epoch": 2.2841379310344827,
+      "grad_norm": 0.6672130823135376,
+      "learning_rate": 1.6299511292482689e-06,
+      "loss": 0.5315,
+      "step": 1104
+    },
+    {
+      "epoch": 2.2862068965517244,
+      "grad_norm": 0.688179612159729,
+      "learning_rate": 1.621062279507617e-06,
+      "loss": 0.526,
+      "step": 1105
+    },
+    {
+      "epoch": 2.2882758620689656,
+      "grad_norm": 0.6778724193572998,
+      "learning_rate": 1.612193041886862e-06,
+      "loss": 0.5029,
+      "step": 1106
+    },
+    {
+      "epoch": 2.290344827586207,
+      "grad_norm": 0.6525766253471375,
+      "learning_rate": 1.6033434678650684e-06,
+      "loss": 0.5367,
+      "step": 1107
+    },
+    {
+      "epoch": 2.2924137931034485,
+      "grad_norm": 0.6853109002113342,
+      "learning_rate": 1.5945136088071677e-06,
+      "loss": 0.5411,
+      "step": 1108
+    },
+    {
+      "epoch": 2.2944827586206897,
+      "grad_norm": 0.6768192648887634,
+      "learning_rate": 1.5857035159636625e-06,
+      "loss": 0.5474,
+      "step": 1109
+    },
+    {
+      "epoch": 2.296551724137931,
+      "grad_norm": 0.6869329214096069,
+      "learning_rate": 1.5769132404703274e-06,
+      "loss": 0.5156,
+      "step": 1110
+    },
+    {
+      "epoch": 2.2986206896551726,
+      "grad_norm": 0.641927182674408,
+      "learning_rate": 1.5681428333479104e-06,
+      "loss": 0.5612,
+      "step": 1111
+    },
+    {
+      "epoch": 2.300689655172414,
+      "grad_norm": 0.6849982738494873,
+      "learning_rate": 1.5593923455018462e-06,
+      "loss": 0.524,
+      "step": 1112
+    },
+    {
+      "epoch": 2.302758620689655,
+      "grad_norm": 0.6639160513877869,
+      "learning_rate": 1.550661827721941e-06,
+      "loss": 0.5219,
+      "step": 1113
+    },
+    {
+      "epoch": 2.3048275862068968,
+      "grad_norm": 0.6571563482284546,
+      "learning_rate": 1.5419513306821055e-06,
+      "loss": 0.5213,
+      "step": 1114
+    },
+    {
+      "epoch": 2.306896551724138,
+      "grad_norm": 0.6656675338745117,
+      "learning_rate": 1.5332609049400298e-06,
+      "loss": 0.5142,
+      "step": 1115
+    },
+    {
+      "epoch": 2.308965517241379,
+      "grad_norm": 0.6509329080581665,
+      "learning_rate": 1.5245906009369211e-06,
+      "loss": 0.5379,
+      "step": 1116
+    },
+    {
+      "epoch": 2.311034482758621,
+      "grad_norm": 0.6804161667823792,
+      "learning_rate": 1.5159404689971797e-06,
+      "loss": 0.5348,
+      "step": 1117
+    },
+    {
+      "epoch": 2.313103448275862,
+      "grad_norm": 0.6578646898269653,
+      "learning_rate": 1.5073105593281385e-06,
+      "loss": 0.5218,
+      "step": 1118
+    },
+    {
+      "epoch": 2.3151724137931033,
+      "grad_norm": 0.6653922200202942,
+      "learning_rate": 1.4987009220197406e-06,
+      "loss": 0.5304,
+      "step": 1119
+    },
+    {
+      "epoch": 2.317241379310345,
+      "grad_norm": 0.6492714881896973,
+      "learning_rate": 1.4901116070442772e-06,
+      "loss": 0.5218,
+      "step": 1120
+    },
+    {
+      "epoch": 2.3193103448275862,
+      "grad_norm": 0.7024757862091064,
+      "learning_rate": 1.4815426642560753e-06,
+      "loss": 0.545,
+      "step": 1121
+    },
+    {
+      "epoch": 2.3213793103448275,
+      "grad_norm": 0.6717721223831177,
+      "learning_rate": 1.47299414339122e-06,
+      "loss": 0.5376,
+      "step": 1122
+    },
+    {
+      "epoch": 2.323448275862069,
+      "grad_norm": 0.6341143250465393,
+      "learning_rate": 1.4644660940672628e-06,
+      "loss": 0.5158,
+      "step": 1123
+    },
+    {
+      "epoch": 2.3255172413793104,
+      "grad_norm": 0.6935365200042725,
+      "learning_rate": 1.455958565782934e-06,
+      "loss": 0.537,
+      "step": 1124
+    },
+    {
+      "epoch": 2.3275862068965516,
+      "grad_norm": 0.6960786581039429,
+      "learning_rate": 1.4474716079178541e-06,
+      "loss": 0.5547,
+      "step": 1125
+    },
+    {
+      "epoch": 2.3296551724137933,
+      "grad_norm": 0.6549028754234314,
+      "learning_rate": 1.4390052697322499e-06,
+      "loss": 0.544,
+      "step": 1126
+    },
+    {
+      "epoch": 2.3317241379310345,
+      "grad_norm": 0.6705878973007202,
+      "learning_rate": 1.430559600366665e-06,
+      "loss": 0.5502,
+      "step": 1127
+    },
+    {
+      "epoch": 2.3337931034482757,
+      "grad_norm": 0.6721364855766296,
+      "learning_rate": 1.4221346488416765e-06,
+      "loss": 0.5532,
+      "step": 1128
+    },
+    {
+      "epoch": 2.3358620689655174,
+      "grad_norm": 0.6675882935523987,
+      "learning_rate": 1.4137304640576161e-06,
+      "loss": 0.5174,
+      "step": 1129
+    },
+    {
+      "epoch": 2.3379310344827586,
+      "grad_norm": 0.650718629360199,
+      "learning_rate": 1.4053470947942694e-06,
+      "loss": 0.5565,
+      "step": 1130
+    },
+    {
+      "epoch": 2.34,
+      "grad_norm": 0.6801978945732117,
+      "learning_rate": 1.3969845897106166e-06,
+      "loss": 0.5276,
+      "step": 1131
+    },
+    {
+      "epoch": 2.3420689655172415,
+      "grad_norm": 0.6504865288734436,
+      "learning_rate": 1.3886429973445253e-06,
+      "loss": 0.5126,
+      "step": 1132
+    },
+    {
+      "epoch": 2.3441379310344828,
+      "grad_norm": 0.6334866881370544,
+      "learning_rate": 1.3803223661124938e-06,
+      "loss": 0.5036,
+      "step": 1133
+    },
+    {
+      "epoch": 2.346206896551724,
+      "grad_norm": 0.6434235572814941,
+      "learning_rate": 1.3720227443093437e-06,
+      "loss": 0.5119,
+      "step": 1134
+    },
+    {
+      "epoch": 2.3482758620689657,
+      "grad_norm": 0.6684521436691284,
+      "learning_rate": 1.3637441801079666e-06,
+      "loss": 0.526,
+      "step": 1135
+    },
+    {
+      "epoch": 2.350344827586207,
+      "grad_norm": 0.6425202488899231,
+      "learning_rate": 1.3554867215590174e-06,
+      "loss": 0.5348,
+      "step": 1136
+    },
+    {
+      "epoch": 2.352413793103448,
+      "grad_norm": 0.6454555988311768,
+      "learning_rate": 1.3472504165906614e-06,
+      "loss": 0.5457,
+      "step": 1137
+    },
+    {
+      "epoch": 2.35448275862069,
+      "grad_norm": 0.6798628568649292,
+      "learning_rate": 1.3390353130082756e-06,
+      "loss": 0.5436,
+      "step": 1138
+    },
+    {
+      "epoch": 2.356551724137931,
+      "grad_norm": 0.6718437671661377,
+      "learning_rate": 1.330841458494183e-06,
+      "loss": 0.5232,
+      "step": 1139
+    },
+    {
+      "epoch": 2.3586206896551722,
+      "grad_norm": 0.6361545324325562,
+      "learning_rate": 1.3226689006073712e-06,
+      "loss": 0.5313,
+      "step": 1140
+    },
+    {
+      "epoch": 2.360689655172414,
+      "grad_norm": 0.6867586374282837,
+      "learning_rate": 1.3145176867832165e-06,
+      "loss": 0.5452,
+      "step": 1141
+    },
+    {
+      "epoch": 2.362758620689655,
+      "grad_norm": 0.665239691734314,
+      "learning_rate": 1.3063878643332112e-06,
+      "loss": 0.5136,
+      "step": 1142
+    },
+    {
+      "epoch": 2.3648275862068964,
+      "grad_norm": 0.6792328953742981,
+      "learning_rate": 1.2982794804446858e-06,
+      "loss": 0.5325,
+      "step": 1143
+    },
+    {
+      "epoch": 2.366896551724138,
+      "grad_norm": 0.6773557066917419,
+      "learning_rate": 1.2901925821805372e-06,
+      "loss": 0.5112,
+      "step": 1144
+    },
+    {
+      "epoch": 2.3689655172413793,
+      "grad_norm": 0.6913496851921082,
+      "learning_rate": 1.2821272164789544e-06,
+      "loss": 0.5547,
+      "step": 1145
+    },
+    {
+      "epoch": 2.3710344827586205,
+      "grad_norm": 0.6805727481842041,
+      "learning_rate": 1.2740834301531468e-06,
+      "loss": 0.5473,
+      "step": 1146
+    },
+    {
+      "epoch": 2.373103448275862,
+      "grad_norm": 0.7125365734100342,
+      "learning_rate": 1.266061269891073e-06,
+      "loss": 0.5248,
+      "step": 1147
+    },
+    {
+      "epoch": 2.3751724137931034,
+      "grad_norm": 0.6634067893028259,
+      "learning_rate": 1.2580607822551677e-06,
+      "loss": 0.5287,
+      "step": 1148
+    },
+    {
+      "epoch": 2.3772413793103446,
+      "grad_norm": 0.6796189546585083,
+      "learning_rate": 1.2500820136820735e-06,
+      "loss": 0.532,
+      "step": 1149
+    },
+    {
+      "epoch": 2.3793103448275863,
+      "grad_norm": 0.6692973971366882,
+      "learning_rate": 1.2421250104823683e-06,
+      "loss": 0.5265,
+      "step": 1150
+    },
+    {
+      "epoch": 2.3813793103448275,
+      "grad_norm": 0.6606705784797668,
+      "learning_rate": 1.2341898188403068e-06,
+      "loss": 0.5184,
+      "step": 1151
+    },
+    {
+      "epoch": 2.3834482758620688,
+      "grad_norm": 0.6459177136421204,
+      "learning_rate": 1.2262764848135316e-06,
+      "loss": 0.5578,
+      "step": 1152
+    },
+    {
+      "epoch": 2.3855172413793104,
+      "grad_norm": 0.67464679479599,
+      "learning_rate": 1.2183850543328313e-06,
+      "loss": 0.5367,
+      "step": 1153
+    },
+    {
+      "epoch": 2.3875862068965517,
+      "grad_norm": 0.6794815063476562,
+      "learning_rate": 1.2105155732018532e-06,
+      "loss": 0.5458,
+      "step": 1154
+    },
+    {
+      "epoch": 2.389655172413793,
+      "grad_norm": 0.6938785910606384,
+      "learning_rate": 1.2026680870968494e-06,
+      "loss": 0.5217,
+      "step": 1155
+    },
+    {
+      "epoch": 2.3917241379310346,
+      "grad_norm": 0.6906960010528564,
+      "learning_rate": 1.194842641566406e-06,
+      "loss": 0.5315,
+      "step": 1156
+    },
+    {
+      "epoch": 2.393793103448276,
+      "grad_norm": 0.6642972826957703,
+      "learning_rate": 1.187039282031182e-06,
+      "loss": 0.5315,
+      "step": 1157
+    },
+    {
+      "epoch": 2.395862068965517,
+      "grad_norm": 0.6678175926208496,
+      "learning_rate": 1.179258053783644e-06,
+      "loss": 0.5007,
+      "step": 1158
+    },
+    {
+      "epoch": 2.3979310344827587,
+      "grad_norm": 0.6204706430435181,
+      "learning_rate": 1.171499001987802e-06,
+      "loss": 0.5617,
+      "step": 1159
+    },
+    {
+      "epoch": 2.4,
+      "grad_norm": 0.6954528093338013,
+      "learning_rate": 1.163762171678951e-06,
+      "loss": 0.5211,
+      "step": 1160
+    },
+    {
+      "epoch": 2.402068965517241,
+      "grad_norm": 0.6939405798912048,
+      "learning_rate": 1.156047607763407e-06,
+      "loss": 0.5264,
+      "step": 1161
+    },
+    {
+      "epoch": 2.404137931034483,
+      "grad_norm": 0.7006513476371765,
+      "learning_rate": 1.1483553550182453e-06,
+      "loss": 0.5488,
+      "step": 1162
+    },
+    {
+      "epoch": 2.406206896551724,
+      "grad_norm": 0.67670738697052,
+      "learning_rate": 1.1406854580910426e-06,
+      "loss": 0.5304,
+      "step": 1163
+    },
+    {
+      "epoch": 2.4082758620689657,
+      "grad_norm": 0.681532084941864,
+      "learning_rate": 1.1330379614996223e-06,
+      "loss": 0.5307,
+      "step": 1164
+    },
+    {
+      "epoch": 2.410344827586207,
+      "grad_norm": 0.671879768371582,
+      "learning_rate": 1.1254129096317807e-06,
+      "loss": 0.5423,
+      "step": 1165
+    },
+    {
+      "epoch": 2.412413793103448,
+      "grad_norm": 0.6575496196746826,
+      "learning_rate": 1.1178103467450512e-06,
+      "loss": 0.5261,
+      "step": 1166
+    },
+    {
+      "epoch": 2.41448275862069,
+      "grad_norm": 0.692674458026886,
+      "learning_rate": 1.110230316966427e-06,
+      "loss": 0.522,
+      "step": 1167
+    },
+    {
+      "epoch": 2.416551724137931,
+      "grad_norm": 0.6568799018859863,
+      "learning_rate": 1.102672864292122e-06,
+      "loss": 0.5187,
+      "step": 1168
+    },
+    {
+      "epoch": 2.4186206896551723,
+      "grad_norm": 0.6846016049385071,
+      "learning_rate": 1.095138032587298e-06,
+      "loss": 0.5518,
+      "step": 1169
+    },
+    {
+      "epoch": 2.420689655172414,
+      "grad_norm": 0.6592535376548767,
+      "learning_rate": 1.0876258655858307e-06,
+      "loss": 0.5241,
+      "step": 1170
+    },
+    {
+      "epoch": 2.422758620689655,
+      "grad_norm": 0.6726316809654236,
+      "learning_rate": 1.0801364068900334e-06,
+      "loss": 0.5153,
+      "step": 1171
+    },
+    {
+      "epoch": 2.4248275862068964,
+      "grad_norm": 0.661979615688324,
+      "learning_rate": 1.0726696999704266e-06,
+      "loss": 0.5321,
+      "step": 1172
+    },
+    {
+      "epoch": 2.426896551724138,
+      "grad_norm": 0.6790592074394226,
+      "learning_rate": 1.0652257881654625e-06,
+      "loss": 0.5361,
+      "step": 1173
+    },
+    {
+      "epoch": 2.4289655172413793,
+      "grad_norm": 0.6957859396934509,
+      "learning_rate": 1.057804714681296e-06,
+      "loss": 0.5129,
+      "step": 1174
+    },
+    {
+      "epoch": 2.4310344827586206,
+      "grad_norm": 0.6732125282287598,
+      "learning_rate": 1.0504065225915166e-06,
+      "loss": 0.56,
+      "step": 1175
+    },
+    {
+      "epoch": 2.4331034482758622,
+      "grad_norm": 0.6757326126098633,
+      "learning_rate": 1.0430312548369076e-06,
+      "loss": 0.5469,
+      "step": 1176
+    },
+    {
+      "epoch": 2.4351724137931035,
+      "grad_norm": 0.673953115940094,
+      "learning_rate": 1.0356789542251939e-06,
+      "loss": 0.507,
+      "step": 1177
+    },
+    {
+      "epoch": 2.4372413793103447,
+      "grad_norm": 0.6575935482978821,
+      "learning_rate": 1.028349663430792e-06,
+      "loss": 0.5134,
+      "step": 1178
+    },
+    {
+      "epoch": 2.4393103448275864,
+      "grad_norm": 0.673143208026886,
+      "learning_rate": 1.0210434249945677e-06,
+      "loss": 0.5443,
+      "step": 1179
+    },
+    {
+      "epoch": 2.4413793103448276,
+      "grad_norm": 0.6751227974891663,
+      "learning_rate": 1.0137602813235824e-06,
+      "loss": 0.5535,
+      "step": 1180
+    },
+    {
+      "epoch": 2.443448275862069,
+      "grad_norm": 0.6402212977409363,
+      "learning_rate": 1.0065002746908532e-06,
+      "loss": 0.5517,
+      "step": 1181
+    },
+    {
+      "epoch": 2.4455172413793105,
+      "grad_norm": 0.6959084868431091,
+      "learning_rate": 9.992634472351016e-07,
+      "loss": 0.5104,
+      "step": 1182
+    },
+    {
+      "epoch": 2.4475862068965517,
+      "grad_norm": 0.673514187335968,
+      "learning_rate": 9.92049840960514e-07,
+      "loss": 0.5485,
+      "step": 1183
+    },
+    {
+      "epoch": 2.449655172413793,
+      "grad_norm": 0.64295893907547,
+      "learning_rate": 9.84859497736495e-07,
+      "loss": 0.5202,
+      "step": 1184
+    },
+    {
+      "epoch": 2.4517241379310346,
+      "grad_norm": 0.669701337814331,
+      "learning_rate": 9.776924592974257e-07,
+      "loss": 0.5201,
+      "step": 1185
+    },
+    {
+      "epoch": 2.453793103448276,
+      "grad_norm": 0.6700237393379211,
+      "learning_rate": 9.705487672424202e-07,
+      "loss": 0.5437,
+      "step": 1186
+    },
+    {
+      "epoch": 2.455862068965517,
+      "grad_norm": 0.6907758712768555,
+      "learning_rate": 9.63428463035085e-07,
+      "loss": 0.511,
+      "step": 1187
+    },
+    {
+      "epoch": 2.4579310344827587,
+      "grad_norm": 0.6732699275016785,
+      "learning_rate": 9.563315880032798e-07,
+      "loss": 0.5418,
+      "step": 1188
+    },
+    {
+      "epoch": 2.46,
+      "grad_norm": 0.7099267840385437,
+      "learning_rate": 9.492581833388736e-07,
+      "loss": 0.5403,
+      "step": 1189
+    },
+    {
+      "epoch": 2.462068965517241,
+      "grad_norm": 0.6354034543037415,
+      "learning_rate": 9.422082900975105e-07,
+      "loss": 0.5227,
+      "step": 1190
+    },
+    {
+      "epoch": 2.464137931034483,
+      "grad_norm": 0.6575520038604736,
+      "learning_rate": 9.351819491983671e-07,
+      "loss": 0.5327,
+      "step": 1191
+    },
+    {
+      "epoch": 2.466206896551724,
+      "grad_norm": 0.6583318114280701,
+      "learning_rate": 9.281792014239171e-07,
+      "loss": 0.5292,
+      "step": 1192
+    },
+    {
+      "epoch": 2.4682758620689658,
+      "grad_norm": 0.6778759360313416,
+      "learning_rate": 9.212000874196953e-07,
+      "loss": 0.5091,
+      "step": 1193
+    },
+    {
+      "epoch": 2.470344827586207,
+      "grad_norm": 0.6768158674240112,
+      "learning_rate": 9.142446476940598e-07,
+      "loss": 0.5043,
+      "step": 1194
+    },
+    {
+      "epoch": 2.472413793103448,
+      "grad_norm": 0.6922417879104614,
+      "learning_rate": 9.073129226179589e-07,
+      "loss": 0.5382,
+      "step": 1195
+    },
+    {
+      "epoch": 2.47448275862069,
+      "grad_norm": 0.6937352418899536,
+      "learning_rate": 9.004049524246944e-07,
+      "loss": 0.5334,
+      "step": 1196
+    },
+    {
+      "epoch": 2.476551724137931,
+      "grad_norm": 0.6940251588821411,
+      "learning_rate": 8.935207772096904e-07,
+      "loss": 0.5235,
+      "step": 1197
+    },
+    {
+      "epoch": 2.4786206896551723,
+      "grad_norm": 0.6610797643661499,
+      "learning_rate": 8.86660436930259e-07,
+      "loss": 0.5369,
+      "step": 1198
+    },
+    {
+      "epoch": 2.480689655172414,
+      "grad_norm": 0.6743729114532471,
+      "learning_rate": 8.798239714053692e-07,
+      "loss": 0.5494,
+      "step": 1199
+    },
+    {
+      "epoch": 2.4827586206896552,
+      "grad_norm": 0.6906112432479858,
+      "learning_rate": 8.730114203154144e-07,
+      "loss": 0.5221,
+      "step": 1200
+    },
+    {
+      "epoch": 2.4848275862068965,
+      "grad_norm": 0.6822516322135925,
+      "learning_rate": 8.662228232019876e-07,
+      "loss": 0.5364,
+      "step": 1201
+    },
+    {
+      "epoch": 2.486896551724138,
+      "grad_norm": 0.6591139435768127,
+      "learning_rate": 8.594582194676393e-07,
+      "loss": 0.5412,
+      "step": 1202
+    },
+    {
+      "epoch": 2.4889655172413794,
+      "grad_norm": 0.694092333316803,
+      "learning_rate": 8.527176483756671e-07,
+      "loss": 0.5233,
+      "step": 1203
+    },
+    {
+      "epoch": 2.4910344827586206,
+      "grad_norm": 0.6754273772239685,
+      "learning_rate": 8.460011490498671e-07,
+      "loss": 0.5501,
+      "step": 1204
+    },
+    {
+      "epoch": 2.4931034482758623,
+      "grad_norm": 0.670056164264679,
+      "learning_rate": 8.393087604743283e-07,
+      "loss": 0.528,
+      "step": 1205
+    },
+    {
+      "epoch": 2.4951724137931035,
+      "grad_norm": 0.6645668148994446,
+      "learning_rate": 8.326405214931838e-07,
+      "loss": 0.5586,
+      "step": 1206
+    },
+    {
+      "epoch": 2.4972413793103447,
+      "grad_norm": 0.6548720002174377,
+      "learning_rate": 8.259964708104091e-07,
+      "loss": 0.5316,
+      "step": 1207
+    },
+    {
+      "epoch": 2.4993103448275864,
+      "grad_norm": 0.6840012073516846,
+      "learning_rate": 8.193766469895737e-07,
+      "loss": 0.5274,
+      "step": 1208
+    },
+    {
+      "epoch": 2.5013793103448276,
+      "grad_norm": 0.655719518661499,
+      "learning_rate": 8.127810884536402e-07,
+      "loss": 0.521,
+      "step": 1209
+    },
+    {
+      "epoch": 2.503448275862069,
+      "grad_norm": 0.6610667109489441,
+      "learning_rate": 8.062098334847185e-07,
+      "loss": 0.5472,
+      "step": 1210
+    },
+    {
+      "epoch": 2.5055172413793105,
+      "grad_norm": 0.6627329587936401,
+      "learning_rate": 7.996629202238654e-07,
+      "loss": 0.5101,
+      "step": 1211
+    },
+    {
+      "epoch": 2.5075862068965518,
+      "grad_norm": 0.6512289047241211,
+      "learning_rate": 7.931403866708464e-07,
+      "loss": 0.5263,
+      "step": 1212
+    },
+    {
+      "epoch": 2.509655172413793,
+      "grad_norm": 0.6700037121772766,
+      "learning_rate": 7.866422706839239e-07,
+      "loss": 0.5265,
+      "step": 1213
+    },
+    {
+      "epoch": 2.5117241379310347,
+      "grad_norm": 0.649949312210083,
+      "learning_rate": 7.801686099796358e-07,
+      "loss": 0.5439,
+      "step": 1214
+    },
+    {
+      "epoch": 2.513793103448276,
+      "grad_norm": 0.6628941893577576,
+      "learning_rate": 7.737194421325739e-07,
+      "loss": 0.5112,
+      "step": 1215
+    },
+    {
+      "epoch": 2.515862068965517,
+      "grad_norm": 0.6837835311889648,
+      "learning_rate": 7.672948045751699e-07,
+      "loss": 0.5497,
+      "step": 1216
+    },
+    {
+      "epoch": 2.517931034482759,
+      "grad_norm": 0.6798170804977417,
+      "learning_rate": 7.60894734597476e-07,
+      "loss": 0.5439,
+      "step": 1217
+    },
+    {
+      "epoch": 2.52,
+      "grad_norm": 0.6628232002258301,
+      "learning_rate": 7.545192693469467e-07,
+      "loss": 0.5435,
+      "step": 1218
+    },
+    {
+      "epoch": 2.5220689655172412,
+      "grad_norm": 0.6693230867385864,
+      "learning_rate": 7.481684458282273e-07,
+      "loss": 0.5377,
+      "step": 1219
+    },
+    {
+      "epoch": 2.524137931034483,
+      "grad_norm": 0.6691813468933105,
+      "learning_rate": 7.418423009029363e-07,
+      "loss": 0.5463,
+      "step": 1220
+    },
+    {
+      "epoch": 2.526206896551724,
+      "grad_norm": 0.6763612031936646,
+      "learning_rate": 7.355408712894508e-07,
+      "loss": 0.5616,
+      "step": 1221
+    },
+    {
+      "epoch": 2.5282758620689654,
+      "grad_norm": 0.6607294082641602,
+      "learning_rate": 7.292641935626966e-07,
+      "loss": 0.5381,
+      "step": 1222
+    },
+    {
+      "epoch": 2.530344827586207,
+      "grad_norm": 0.6794742941856384,
+      "learning_rate": 7.23012304153931e-07,
+      "loss": 0.5427,
+      "step": 1223
+    },
+    {
+      "epoch": 2.5324137931034483,
+      "grad_norm": 0.6626706719398499,
+      "learning_rate": 7.167852393505393e-07,
+      "loss": 0.534,
+      "step": 1224
+    },
+    {
+      "epoch": 2.5344827586206895,
+      "grad_norm": 0.6796131730079651,
+      "learning_rate": 7.105830352958143e-07,
+      "loss": 0.5493,
+      "step": 1225
+    },
+    {
+      "epoch": 2.536551724137931,
+      "grad_norm": 0.6909650564193726,
+      "learning_rate": 7.04405727988754e-07,
+      "loss": 0.5359,
+      "step": 1226
+    },
+    {
+      "epoch": 2.5386206896551724,
+      "grad_norm": 0.6646261811256409,
+      "learning_rate": 6.982533532838481e-07,
+      "loss": 0.4803,
+      "step": 1227
+    },
+    {
+      "epoch": 2.5406896551724136,
+      "grad_norm": 0.6714354753494263,
+      "learning_rate": 6.921259468908742e-07,
+      "loss": 0.5642,
+      "step": 1228
+    },
+    {
+      "epoch": 2.5427586206896553,
+      "grad_norm": 0.6736488342285156,
+      "learning_rate": 6.86023544374686e-07,
+      "loss": 0.5513,
+      "step": 1229
+    },
+    {
+      "epoch": 2.5448275862068965,
+      "grad_norm": 0.7024829387664795,
+      "learning_rate": 6.799461811550101e-07,
+      "loss": 0.5382,
+      "step": 1230
+    },
+    {
+      "epoch": 2.5468965517241378,
+      "grad_norm": 0.670295000076294,
+      "learning_rate": 6.738938925062399e-07,
+      "loss": 0.5301,
+      "step": 1231
+    },
+    {
+      "epoch": 2.5489655172413794,
+      "grad_norm": 0.6387397050857544,
+      "learning_rate": 6.678667135572293e-07,
+      "loss": 0.5093,
+      "step": 1232
+    },
+    {
+      "epoch": 2.5510344827586207,
+      "grad_norm": 0.6766871809959412,
+      "learning_rate": 6.618646792910893e-07,
+      "loss": 0.5519,
+      "step": 1233
+    },
+    {
+      "epoch": 2.553103448275862,
+      "grad_norm": 0.7046573758125305,
+      "learning_rate": 6.558878245449879e-07,
+      "loss": 0.5218,
+      "step": 1234
+    },
+    {
+      "epoch": 2.5551724137931036,
+      "grad_norm": 0.6664929389953613,
+      "learning_rate": 6.499361840099421e-07,
+      "loss": 0.522,
+      "step": 1235
+    },
+    {
+      "epoch": 2.557241379310345,
+      "grad_norm": 0.6669551730155945,
+      "learning_rate": 6.440097922306232e-07,
+      "loss": 0.565,
+      "step": 1236
+    },
+    {
+      "epoch": 2.559310344827586,
+      "grad_norm": 0.6854833364486694,
+      "learning_rate": 6.381086836051498e-07,
+      "loss": 0.5561,
+      "step": 1237
+    },
+    {
+      "epoch": 2.5613793103448277,
+      "grad_norm": 0.6877058148384094,
+      "learning_rate": 6.322328923848958e-07,
+      "loss": 0.5313,
+      "step": 1238
+    },
+    {
+      "epoch": 2.563448275862069,
+      "grad_norm": 0.6529080867767334,
+      "learning_rate": 6.263824526742807e-07,
+      "loss": 0.5318,
+      "step": 1239
+    },
+    {
+      "epoch": 2.56551724137931,
+      "grad_norm": 0.6818994879722595,
+      "learning_rate": 6.205573984305857e-07,
+      "loss": 0.5387,
+      "step": 1240
+    },
+    {
+      "epoch": 2.567586206896552,
+      "grad_norm": 0.6354886889457703,
+      "learning_rate": 6.147577634637413e-07,
+      "loss": 0.5351,
+      "step": 1241
+    },
+    {
+      "epoch": 2.569655172413793,
+      "grad_norm": 0.6590231657028198,
+      "learning_rate": 6.089835814361461e-07,
+      "loss": 0.5288,
+      "step": 1242
+    },
+    {
+      "epoch": 2.5717241379310343,
+      "grad_norm": 0.6300837993621826,
+      "learning_rate": 6.03234885862457e-07,
+      "loss": 0.554,
+      "step": 1243
+    },
+    {
+      "epoch": 2.573793103448276,
+      "grad_norm": 0.64389568567276,
+      "learning_rate": 5.975117101094091e-07,
+      "loss": 0.5548,
+      "step": 1244
+    },
+    {
+      "epoch": 2.575862068965517,
+      "grad_norm": 0.65228271484375,
+      "learning_rate": 5.918140873956063e-07,
+      "loss": 0.5418,
+      "step": 1245
+    },
+    {
+      "epoch": 2.5779310344827584,
+      "grad_norm": 0.6716015934944153,
+      "learning_rate": 5.861420507913462e-07,
+      "loss": 0.5506,
+      "step": 1246
+    },
+    {
+      "epoch": 2.58,
+      "grad_norm": 0.6640960574150085,
+      "learning_rate": 5.804956332184092e-07,
+      "loss": 0.5215,
+      "step": 1247
+    },
+    {
+      "epoch": 2.5820689655172413,
+      "grad_norm": 0.6824872493743896,
+      "learning_rate": 5.748748674498855e-07,
+      "loss": 0.546,
+      "step": 1248
+    },
+    {
+      "epoch": 2.5841379310344825,
+      "grad_norm": 0.6688941121101379,
+      "learning_rate": 5.692797861099719e-07,
+      "loss": 0.5166,
+      "step": 1249
+    },
+    {
+      "epoch": 2.586206896551724,
+      "grad_norm": 0.6445797085762024,
+      "learning_rate": 5.637104216737871e-07,
+      "loss": 0.5286,
+      "step": 1250
+    },
+    {
+      "epoch": 2.5882758620689654,
+      "grad_norm": 0.6584656238555908,
+      "learning_rate": 5.581668064671847e-07,
+      "loss": 0.5364,
+      "step": 1251
+    },
+    {
+      "epoch": 2.5903448275862067,
+      "grad_norm": 0.6681441068649292,
+      "learning_rate": 5.526489726665624e-07,
+      "loss": 0.5203,
+      "step": 1252
+    },
+    {
+      "epoch": 2.5924137931034483,
+      "grad_norm": 0.6537905931472778,
+      "learning_rate": 5.471569522986775e-07,
+      "loss": 0.5203,
+      "step": 1253
+    },
+    {
+      "epoch": 2.5944827586206896,
+      "grad_norm": 0.613910973072052,
+      "learning_rate": 5.416907772404606e-07,
+      "loss": 0.5105,
+      "step": 1254
+    },
+    {
+      "epoch": 2.596551724137931,
+      "grad_norm": 0.6802027225494385,
+      "learning_rate": 5.3625047921883e-07,
+      "loss": 0.5231,
+      "step": 1255
+    },
+    {
+      "epoch": 2.5986206896551725,
+      "grad_norm": 0.6352335214614868,
+      "learning_rate": 5.30836089810507e-07,
+      "loss": 0.5125,
+      "step": 1256
+    },
+    {
+      "epoch": 2.6006896551724137,
+      "grad_norm": 0.6671028137207031,
+      "learning_rate": 5.254476404418341e-07,
+      "loss": 0.5221,
+      "step": 1257
+    },
+    {
+      "epoch": 2.602758620689655,
+      "grad_norm": 0.6607826948165894,
+      "learning_rate": 5.200851623885922e-07,
+      "loss": 0.5384,
+      "step": 1258
+    },
+    {
+      "epoch": 2.6048275862068966,
+      "grad_norm": 0.645964503288269,
+      "learning_rate": 5.147486867758201e-07,
+      "loss": 0.5306,
+      "step": 1259
+    },
+    {
+      "epoch": 2.606896551724138,
+      "grad_norm": 0.6596983671188354,
+      "learning_rate": 5.094382445776281e-07,
+      "loss": 0.5464,
+      "step": 1260
+    },
+    {
+      "epoch": 2.608965517241379,
+      "grad_norm": 0.678046464920044,
+      "learning_rate": 5.041538666170282e-07,
+      "loss": 0.524,
+      "step": 1261
+    },
+    {
+      "epoch": 2.6110344827586207,
+      "grad_norm": 0.6556622385978699,
+      "learning_rate": 4.988955835657456e-07,
+      "loss": 0.5395,
+      "step": 1262
+    },
+    {
+      "epoch": 2.613103448275862,
+      "grad_norm": 0.6770464181900024,
+      "learning_rate": 4.93663425944047e-07,
+      "loss": 0.5394,
+      "step": 1263
+    },
+    {
+      "epoch": 2.6151724137931036,
+      "grad_norm": 0.6753197312355042,
+      "learning_rate": 4.884574241205603e-07,
+      "loss": 0.5239,
+      "step": 1264
+    },
+    {
+      "epoch": 2.617241379310345,
+      "grad_norm": 0.655596911907196,
+      "learning_rate": 4.832776083120983e-07,
+      "loss": 0.5148,
+      "step": 1265
+    },
+    {
+      "epoch": 2.619310344827586,
+      "grad_norm": 0.6836827397346497,
+      "learning_rate": 4.781240085834865e-07,
+      "loss": 0.5453,
+      "step": 1266
+    },
+    {
+      "epoch": 2.6213793103448277,
+      "grad_norm": 0.680099368095398,
+      "learning_rate": 4.729966548473841e-07,
+      "loss": 0.5148,
+      "step": 1267
+    },
+    {
+      "epoch": 2.623448275862069,
+      "grad_norm": 0.6692262887954712,
+      "learning_rate": 4.678955768641147e-07,
+      "loss": 0.554,
+      "step": 1268
+    },
+    {
+      "epoch": 2.62551724137931,
+      "grad_norm": 0.6534875631332397,
+      "learning_rate": 4.628208042414889e-07,
+      "loss": 0.5548,
+      "step": 1269
+    },
+    {
+      "epoch": 2.627586206896552,
+      "grad_norm": 0.617273211479187,
+      "learning_rate": 4.5777236643463787e-07,
+      "loss": 0.5305,
+      "step": 1270
+    },
+    {
+      "epoch": 2.629655172413793,
+      "grad_norm": 0.6691027283668518,
+      "learning_rate": 4.5275029274583715e-07,
+      "loss": 0.5292,
+      "step": 1271
+    },
+    {
+      "epoch": 2.6317241379310343,
+      "grad_norm": 0.6685481071472168,
+      "learning_rate": 4.477546123243409e-07,
+      "loss": 0.5306,
+      "step": 1272
+    },
+    {
+      "epoch": 2.633793103448276,
+      "grad_norm": 0.6642303466796875,
+      "learning_rate": 4.4278535416620914e-07,
+      "loss": 0.5276,
+      "step": 1273
+    },
+    {
+      "epoch": 2.6358620689655172,
+      "grad_norm": 0.6658126711845398,
+      "learning_rate": 4.3784254711414064e-07,
+      "loss": 0.5447,
+      "step": 1274
+    },
+    {
+      "epoch": 2.637931034482759,
+      "grad_norm": 0.6844912767410278,
+      "learning_rate": 4.329262198573109e-07,
+      "loss": 0.5419,
+      "step": 1275
+    },
+    {
+      "epoch": 2.64,
+      "grad_norm": 0.6531012654304504,
+      "learning_rate": 4.280364009311927e-07,
+      "loss": 0.5283,
+      "step": 1276
+    },
+    {
+      "epoch": 2.6420689655172414,
+      "grad_norm": 0.6809141635894775,
+      "learning_rate": 4.231731187174065e-07,
+      "loss": 0.5486,
+      "step": 1277
+    },
+    {
+      "epoch": 2.644137931034483,
+      "grad_norm": 0.6723445653915405,
+      "learning_rate": 4.183364014435398e-07,
+      "loss": 0.5441,
+      "step": 1278
+    },
+    {
+      "epoch": 2.6462068965517243,
+      "grad_norm": 0.6648197174072266,
+      "learning_rate": 4.1352627718299953e-07,
+      "loss": 0.5506,
+      "step": 1279
+    },
+    {
+      "epoch": 2.6482758620689655,
+      "grad_norm": 0.668096661567688,
+      "learning_rate": 4.087427738548322e-07,
+      "loss": 0.5363,
+      "step": 1280
+    },
+    {
+      "epoch": 2.650344827586207,
+      "grad_norm": 0.6981197595596313,
+      "learning_rate": 4.0398591922357787e-07,
+      "loss": 0.5673,
+      "step": 1281
+    },
+    {
+      "epoch": 2.6524137931034484,
+      "grad_norm": 0.6475323438644409,
+      "learning_rate": 3.9925574089909603e-07,
+      "loss": 0.5402,
+      "step": 1282
+    },
+    {
+      "epoch": 2.6544827586206896,
+      "grad_norm": 0.640293300151825,
+      "learning_rate": 3.945522663364154e-07,
+      "loss": 0.5302,
+      "step": 1283
+    },
+    {
+      "epoch": 2.6565517241379313,
+      "grad_norm": 0.6663315296173096,
+      "learning_rate": 3.898755228355661e-07,
+      "loss": 0.5379,
+      "step": 1284
+    },
+    {
+      "epoch": 2.6586206896551725,
+      "grad_norm": 0.6679863929748535,
+      "learning_rate": 3.852255375414271e-07,
+      "loss": 0.5605,
+      "step": 1285
+    },
+    {
+      "epoch": 2.6606896551724137,
+      "grad_norm": 0.6763351559638977,
+      "learning_rate": 3.8060233744356634e-07,
+      "loss": 0.5102,
+      "step": 1286
+    },
+    {
+      "epoch": 2.6627586206896554,
+      "grad_norm": 0.6907029151916504,
+      "learning_rate": 3.760059493760848e-07,
+      "loss": 0.522,
+      "step": 1287
+    },
+    {
+      "epoch": 2.6648275862068966,
+      "grad_norm": 0.6595637202262878,
+      "learning_rate": 3.7143640001745916e-07,
+      "loss": 0.549,
+      "step": 1288
+    },
+    {
+      "epoch": 2.666896551724138,
+      "grad_norm": 0.6693307161331177,
+      "learning_rate": 3.6689371589039013e-07,
+      "loss": 0.5257,
+      "step": 1289
+    },
+    {
+      "epoch": 2.6689655172413795,
+      "grad_norm": 0.6373291611671448,
+      "learning_rate": 3.6237792336164424e-07,
+      "loss": 0.5478,
+      "step": 1290
+    },
+    {
+      "epoch": 2.6710344827586208,
+      "grad_norm": 0.6235068440437317,
+      "learning_rate": 3.5788904864190346e-07,
+      "loss": 0.5484,
+      "step": 1291
+    },
+    {
+      "epoch": 2.673103448275862,
+      "grad_norm": 0.6607393026351929,
+      "learning_rate": 3.534271177856163e-07,
+      "loss": 0.5082,
+      "step": 1292
+    },
+    {
+      "epoch": 2.6751724137931037,
+      "grad_norm": 0.6282607316970825,
+      "learning_rate": 3.489921566908372e-07,
+      "loss": 0.5366,
+      "step": 1293
+    },
+    {
+      "epoch": 2.677241379310345,
+      "grad_norm": 0.6705973148345947,
+      "learning_rate": 3.445841910990877e-07,
+      "loss": 0.5382,
+      "step": 1294
+    },
+    {
+      "epoch": 2.679310344827586,
+      "grad_norm": 0.6531069278717041,
+      "learning_rate": 3.4020324659519487e-07,
+      "loss": 0.5556,
+      "step": 1295
+    },
+    {
+      "epoch": 2.681379310344828,
+      "grad_norm": 0.6508034467697144,
+      "learning_rate": 3.358493486071568e-07,
+      "loss": 0.5256,
+      "step": 1296
+    },
+    {
+      "epoch": 2.683448275862069,
+      "grad_norm": 0.6649227142333984,
+      "learning_rate": 3.315225224059809e-07,
+      "loss": 0.5104,
+      "step": 1297
+    },
+    {
+      "epoch": 2.6855172413793102,
+      "grad_norm": 0.6882157325744629,
+      "learning_rate": 3.2722279310554825e-07,
+      "loss": 0.542,
+      "step": 1298
+    },
+    {
+      "epoch": 2.687586206896552,
+      "grad_norm": 0.6619060039520264,
+      "learning_rate": 3.2295018566246007e-07,
+      "loss": 0.5379,
+      "step": 1299
+    },
+    {
+      "epoch": 2.689655172413793,
+      "grad_norm": 0.6553718447685242,
+      "learning_rate": 3.1870472487589845e-07,
+      "loss": 0.5122,
+      "step": 1300
+    },
+    {
+      "epoch": 2.6917241379310344,
+      "grad_norm": 0.691162645816803,
+      "learning_rate": 3.1448643538748045e-07,
+      "loss": 0.5102,
+      "step": 1301
+    },
+    {
+      "epoch": 2.693793103448276,
+      "grad_norm": 0.6377774477005005,
+      "learning_rate": 3.1029534168111217e-07,
+      "loss": 0.524,
+      "step": 1302
+    },
+    {
+      "epoch": 2.6958620689655173,
+      "grad_norm": 0.6641302704811096,
+      "learning_rate": 3.06131468082852e-07,
+      "loss": 0.5126,
+      "step": 1303
+    },
+    {
+      "epoch": 2.6979310344827585,
+      "grad_norm": 0.6569216251373291,
+      "learning_rate": 3.0199483876076465e-07,
+      "loss": 0.5347,
+      "step": 1304
+    },
+    {
+      "epoch": 2.7,
+      "grad_norm": 0.6541246175765991,
+      "learning_rate": 2.9788547772478416e-07,
+      "loss": 0.5384,
+      "step": 1305
+    },
+    {
+      "epoch": 2.7020689655172414,
+      "grad_norm": 0.6962401866912842,
+      "learning_rate": 2.938034088265723e-07,
+      "loss": 0.5321,
+      "step": 1306
+    },
+    {
+      "epoch": 2.7041379310344826,
+      "grad_norm": 0.6459115147590637,
+      "learning_rate": 2.897486557593826e-07,
+      "loss": 0.5477,
+      "step": 1307
+    },
+    {
+      "epoch": 2.7062068965517243,
+      "grad_norm": 0.6690937280654907,
+      "learning_rate": 2.8572124205791916e-07,
+      "loss": 0.5361,
+      "step": 1308
+    },
+    {
+      "epoch": 2.7082758620689655,
+      "grad_norm": 0.651241660118103,
+      "learning_rate": 2.817211910982037e-07,
+      "loss": 0.5231,
+      "step": 1309
+    },
+    {
+      "epoch": 2.7103448275862068,
+      "grad_norm": 0.6507266759872437,
+      "learning_rate": 2.7774852609743887e-07,
+      "loss": 0.5331,
+      "step": 1310
+    },
+    {
+      "epoch": 2.7124137931034484,
+      "grad_norm": 0.6396605968475342,
+      "learning_rate": 2.7380327011387096e-07,
+      "loss": 0.5418,
+      "step": 1311
+    },
+    {
+      "epoch": 2.7144827586206897,
+      "grad_norm": 0.635365903377533,
+      "learning_rate": 2.698854460466599e-07,
+      "loss": 0.5258,
+      "step": 1312
+    },
+    {
+      "epoch": 2.716551724137931,
+      "grad_norm": 0.6772063374519348,
+      "learning_rate": 2.6599507663574387e-07,
+      "loss": 0.5274,
+      "step": 1313
+    },
+    {
+      "epoch": 2.7186206896551726,
+      "grad_norm": 0.6762843132019043,
+      "learning_rate": 2.621321844617092e-07,
+      "loss": 0.5279,
+      "step": 1314
+    },
+    {
+      "epoch": 2.720689655172414,
+      "grad_norm": 0.6602565050125122,
+      "learning_rate": 2.582967919456547e-07,
+      "loss": 0.5343,
+      "step": 1315
+    },
+    {
+      "epoch": 2.722758620689655,
+      "grad_norm": 0.6536930799484253,
+      "learning_rate": 2.544889213490698e-07,
+      "loss": 0.5146,
+      "step": 1316
+    },
+    {
+      "epoch": 2.7248275862068967,
+      "grad_norm": 0.6757822632789612,
+      "learning_rate": 2.5070859477369645e-07,
+      "loss": 0.5247,
+      "step": 1317
+    },
+    {
+      "epoch": 2.726896551724138,
+      "grad_norm": 0.6500410437583923,
+      "learning_rate": 2.46955834161407e-07,
+      "loss": 0.5383,
+      "step": 1318
+    },
+    {
+      "epoch": 2.728965517241379,
+      "grad_norm": 0.6390625834465027,
+      "learning_rate": 2.432306612940738e-07,
+      "loss": 0.5254,
+      "step": 1319
+    },
+    {
+      "epoch": 2.731034482758621,
+      "grad_norm": 0.6645777821540833,
+      "learning_rate": 2.39533097793444e-07,
+      "loss": 0.4968,
+      "step": 1320
+    },
+    {
+      "epoch": 2.733103448275862,
+      "grad_norm": 0.6638719439506531,
+      "learning_rate": 2.3586316512101416e-07,
+      "loss": 0.5147,
+      "step": 1321
+    },
+    {
+      "epoch": 2.7351724137931033,
+      "grad_norm": 0.6468044519424438,
+      "learning_rate": 2.3222088457790448e-07,
+      "loss": 0.5321,
+      "step": 1322
+    },
+    {
+      "epoch": 2.737241379310345,
+      "grad_norm": 0.6644803285598755,
+      "learning_rate": 2.286062773047354e-07,
+      "loss": 0.5462,
+      "step": 1323
+    },
+    {
+      "epoch": 2.739310344827586,
+      "grad_norm": 0.6668086647987366,
+      "learning_rate": 2.2501936428150794e-07,
+      "loss": 0.5093,
+      "step": 1324
+    },
+    {
+      "epoch": 2.7413793103448274,
+      "grad_norm": 0.6860660910606384,
+      "learning_rate": 2.2146016632747624e-07,
+      "loss": 0.5485,
+      "step": 1325
+    },
+    {
+      "epoch": 2.743448275862069,
+      "grad_norm": 0.6771222949028015,
+      "learning_rate": 2.1792870410103195e-07,
+      "loss": 0.5049,
+      "step": 1326
+    },
+    {
+      "epoch": 2.7455172413793103,
+      "grad_norm": 0.6570666432380676,
+      "learning_rate": 2.1442499809958273e-07,
+      "loss": 0.5263,
+      "step": 1327
+    },
+    {
+      "epoch": 2.7475862068965515,
+      "grad_norm": 0.6926378011703491,
+      "learning_rate": 2.109490686594301e-07,
+      "loss": 0.5532,
+      "step": 1328
+    },
+    {
+      "epoch": 2.749655172413793,
+      "grad_norm": 0.6724755764007568,
+      "learning_rate": 2.0750093595565735e-07,
+      "loss": 0.5282,
+      "step": 1329
+    },
+    {
+      "epoch": 2.7517241379310344,
+      "grad_norm": 0.6550016403198242,
+      "learning_rate": 2.0408062000200623e-07,
+      "loss": 0.5475,
+      "step": 1330
+    },
+    {
+      "epoch": 2.7537931034482757,
+      "grad_norm": 0.6652045249938965,
+      "learning_rate": 2.006881406507677e-07,
+      "loss": 0.5517,
+      "step": 1331
+    },
+    {
+      "epoch": 2.7558620689655173,
+      "grad_norm": 0.6777829527854919,
+      "learning_rate": 1.9732351759265743e-07,
+      "loss": 0.5755,
+      "step": 1332
+    },
+    {
+      "epoch": 2.7579310344827586,
+      "grad_norm": 0.632814347743988,
+      "learning_rate": 1.9398677035671222e-07,
+      "loss": 0.5004,
+      "step": 1333
+    },
+    {
+      "epoch": 2.76,
+      "grad_norm": 0.6533367037773132,
+      "learning_rate": 1.9067791831016647e-07,
+      "loss": 0.5314,
+      "step": 1334
+    },
+    {
+      "epoch": 2.7620689655172415,
+      "grad_norm": 0.6759463548660278,
+      "learning_rate": 1.8739698065834977e-07,
+      "loss": 0.5277,
+      "step": 1335
+    },
+    {
+      "epoch": 2.7641379310344827,
+      "grad_norm": 0.6170943975448608,
+      "learning_rate": 1.841439764445646e-07,
+      "loss": 0.5207,
+      "step": 1336
+    },
+    {
+      "epoch": 2.766206896551724,
+      "grad_norm": 0.6548056602478027,
+      "learning_rate": 1.8091892454998595e-07,
+      "loss": 0.5539,
+      "step": 1337
+    },
+    {
+      "epoch": 2.7682758620689656,
+      "grad_norm": 0.6568596363067627,
+      "learning_rate": 1.7772184369354517e-07,
+      "loss": 0.5267,
+      "step": 1338
+    },
+    {
+      "epoch": 2.770344827586207,
+      "grad_norm": 0.6602894067764282,
+      "learning_rate": 1.7455275243182413e-07,
+      "loss": 0.5589,
+      "step": 1339
+    },
+    {
+      "epoch": 2.772413793103448,
+      "grad_norm": 0.6518042087554932,
+      "learning_rate": 1.714116691589457e-07,
+      "loss": 0.518,
+      "step": 1340
+    },
+    {
+      "epoch": 2.7744827586206897,
+      "grad_norm": 0.6620248556137085,
+      "learning_rate": 1.682986121064689e-07,
+      "loss": 0.5521,
+      "step": 1341
+    },
+    {
+      "epoch": 2.776551724137931,
+      "grad_norm": 0.6624969840049744,
+      "learning_rate": 1.6521359934328185e-07,
+      "loss": 0.5467,
+      "step": 1342
+    },
+    {
+      "epoch": 2.778620689655172,
+      "grad_norm": 0.6587342619895935,
+      "learning_rate": 1.6215664877549774e-07,
+      "loss": 0.5063,
+      "step": 1343
+    },
+    {
+      "epoch": 2.780689655172414,
+      "grad_norm": 0.6892203092575073,
+      "learning_rate": 1.5912777814635018e-07,
+      "loss": 0.5463,
+      "step": 1344
+    },
+    {
+      "epoch": 2.782758620689655,
+      "grad_norm": 0.6598497033119202,
+      "learning_rate": 1.561270050360897e-07,
+      "loss": 0.5265,
+      "step": 1345
+    },
+    {
+      "epoch": 2.7848275862068963,
+      "grad_norm": 0.6603181958198547,
+      "learning_rate": 1.5315434686188345e-07,
+      "loss": 0.5467,
+      "step": 1346
+    },
+    {
+      "epoch": 2.786896551724138,
+      "grad_norm": 0.6583660840988159,
+      "learning_rate": 1.5020982087771362e-07,
+      "loss": 0.5537,
+      "step": 1347
+    },
+    {
+      "epoch": 2.788965517241379,
+      "grad_norm": 0.6598078012466431,
+      "learning_rate": 1.472934441742746e-07,
+      "loss": 0.5563,
+      "step": 1348
+    },
+    {
+      "epoch": 2.7910344827586204,
+      "grad_norm": 0.6481250524520874,
+      "learning_rate": 1.444052336788787e-07,
+      "loss": 0.5573,
+      "step": 1349
+    },
+    {
+      "epoch": 2.793103448275862,
+      "grad_norm": 0.6612368226051331,
+      "learning_rate": 1.4154520615535185e-07,
+      "loss": 0.5604,
+      "step": 1350
+    },
+    {
+      "epoch": 2.7951724137931033,
+      "grad_norm": 0.6412969827651978,
+      "learning_rate": 1.3871337820394404e-07,
+      "loss": 0.5592,
+      "step": 1351
+    },
+    {
+      "epoch": 2.7972413793103446,
+      "grad_norm": 0.6546990275382996,
+      "learning_rate": 1.359097662612252e-07,
+      "loss": 0.5295,
+      "step": 1352
+    },
+    {
+      "epoch": 2.7993103448275862,
+      "grad_norm": 0.6505655646324158,
+      "learning_rate": 1.33134386599994e-07,
+      "loss": 0.5208,
+      "step": 1353
+    },
+    {
+      "epoch": 2.8013793103448275,
+      "grad_norm": 0.6737064123153687,
+      "learning_rate": 1.303872553291824e-07,
+      "loss": 0.529,
+      "step": 1354
+    },
+    {
+      "epoch": 2.803448275862069,
+      "grad_norm": 0.6469187140464783,
+      "learning_rate": 1.2766838839376294e-07,
+      "loss": 0.5435,
+      "step": 1355
+    },
+    {
+      "epoch": 2.8055172413793104,
+      "grad_norm": 0.6671730279922485,
+      "learning_rate": 1.24977801574655e-07,
+      "loss": 0.5453,
+      "step": 1356
+    },
+    {
+      "epoch": 2.8075862068965516,
+      "grad_norm": 0.656061053276062,
+      "learning_rate": 1.2231551048863421e-07,
+      "loss": 0.5404,
+      "step": 1357
+    },
+    {
+      "epoch": 2.8096551724137933,
+      "grad_norm": 0.6444647312164307,
+      "learning_rate": 1.196815305882404e-07,
+      "loss": 0.5321,
+      "step": 1358
+    },
+    {
+      "epoch": 2.8117241379310345,
+      "grad_norm": 0.6776153445243835,
+      "learning_rate": 1.1707587716169033e-07,
+      "loss": 0.5226,
+      "step": 1359
+    },
+    {
+      "epoch": 2.8137931034482757,
+      "grad_norm": 0.6640626192092896,
+      "learning_rate": 1.1449856533278624e-07,
+      "loss": 0.5381,
+      "step": 1360
+    },
+    {
+      "epoch": 2.8158620689655174,
+      "grad_norm": 0.6596142053604126,
+      "learning_rate": 1.1194961006082972e-07,
+      "loss": 0.52,
+      "step": 1361
+    },
+    {
+      "epoch": 2.8179310344827586,
+      "grad_norm": 0.6379115581512451,
+      "learning_rate": 1.0942902614053453e-07,
+      "loss": 0.5294,
+      "step": 1362
+    },
+    {
+      "epoch": 2.82,
+      "grad_norm": 0.6417571902275085,
+      "learning_rate": 1.0693682820194062e-07,
+      "loss": 0.56,
+      "step": 1363
+    },
+    {
+      "epoch": 2.8220689655172415,
+      "grad_norm": 0.672452986240387,
+      "learning_rate": 1.0447303071032977e-07,
+      "loss": 0.5091,
+      "step": 1364
+    },
+    {
+      "epoch": 2.8241379310344827,
+      "grad_norm": 0.6543784737586975,
+      "learning_rate": 1.0203764796614057e-07,
+      "loss": 0.5518,
+      "step": 1365
+    },
+    {
+      "epoch": 2.8262068965517244,
+      "grad_norm": 0.6682656407356262,
+      "learning_rate": 9.963069410488635e-08,
+      "loss": 0.5146,
+      "step": 1366
+    },
+    {
+      "epoch": 2.8282758620689656,
+      "grad_norm": 0.6736637949943542,
+      "learning_rate": 9.725218309707129e-08,
+      "loss": 0.5245,
+      "step": 1367
+    },
+    {
+      "epoch": 2.830344827586207,
+      "grad_norm": 0.6793366074562073,
+      "learning_rate": 9.490212874811388e-08,
+      "loss": 0.5301,
+      "step": 1368
+    },
+    {
+      "epoch": 2.8324137931034485,
+      "grad_norm": 0.6533759832382202,
+      "learning_rate": 9.258054469825972e-08,
+      "loss": 0.5245,
+      "step": 1369
+    },
+    {
+      "epoch": 2.8344827586206898,
+      "grad_norm": 0.6463536024093628,
+      "learning_rate": 9.028744442251103e-08,
+      "loss": 0.5174,
+      "step": 1370
+    },
+    {
+      "epoch": 2.836551724137931,
+      "grad_norm": 0.691880464553833,
+      "learning_rate": 8.802284123053894e-08,
+      "loss": 0.5345,
+      "step": 1371
+    },
+    {
+      "epoch": 2.8386206896551727,
+      "grad_norm": 0.6536058187484741,
+      "learning_rate": 8.578674826661527e-08,
+      "loss": 0.5371,
+      "step": 1372
+    },
+    {
+      "epoch": 2.840689655172414,
+      "grad_norm": 0.658385157585144,
+      "learning_rate": 8.357917850952802e-08,
+      "loss": 0.5064,
+      "step": 1373
+    },
+    {
+      "epoch": 2.842758620689655,
+      "grad_norm": 0.662955105304718,
+      "learning_rate": 8.140014477251379e-08,
+      "loss": 0.5247,
+      "step": 1374
+    },
+    {
+      "epoch": 2.844827586206897,
+      "grad_norm": 0.6408557295799255,
+      "learning_rate": 7.924965970317722e-08,
+      "loss": 0.5348,
+      "step": 1375
+    },
+    {
+      "epoch": 2.846896551724138,
+      "grad_norm": 0.6421958208084106,
+      "learning_rate": 7.712773578342047e-08,
+      "loss": 0.5583,
+      "step": 1376
+    },
+    {
+      "epoch": 2.8489655172413793,
+      "grad_norm": 0.6565936803817749,
+      "learning_rate": 7.503438532937169e-08,
+      "loss": 0.5502,
+      "step": 1377
+    },
+    {
+      "epoch": 2.851034482758621,
+      "grad_norm": 0.6656866073608398,
+      "learning_rate": 7.296962049131051e-08,
+      "loss": 0.5256,
+      "step": 1378
+    },
+    {
+      "epoch": 2.853103448275862,
+      "grad_norm": 0.662927508354187,
+      "learning_rate": 7.093345325359935e-08,
+      "loss": 0.5399,
+      "step": 1379
+    },
+    {
+      "epoch": 2.8551724137931034,
+      "grad_norm": 0.664005696773529,
+      "learning_rate": 6.892589543461392e-08,
+      "loss": 0.5233,
+      "step": 1380
+    },
+    {
+      "epoch": 2.857241379310345,
+      "grad_norm": 0.6524354219436646,
+      "learning_rate": 6.694695868667556e-08,
+      "loss": 0.5318,
+      "step": 1381
+    },
+    {
+      "epoch": 2.8593103448275863,
+      "grad_norm": 0.6665657758712769,
+      "learning_rate": 6.49966544959807e-08,
+      "loss": 0.5241,
+      "step": 1382
+    },
+    {
+      "epoch": 2.8613793103448275,
+      "grad_norm": 0.6510589122772217,
+      "learning_rate": 6.307499418253705e-08,
+      "loss": 0.5293,
+      "step": 1383
+    },
+    {
+      "epoch": 2.863448275862069,
+      "grad_norm": 0.6608887910842896,
+      "learning_rate": 6.118198890009586e-08,
+      "loss": 0.5257,
+      "step": 1384
+    },
+    {
+      "epoch": 2.8655172413793104,
+      "grad_norm": 0.6739917993545532,
+      "learning_rate": 5.9317649636088656e-08,
+      "loss": 0.5544,
+      "step": 1385
+    },
+    {
+      "epoch": 2.8675862068965516,
+      "grad_norm": 0.6483108401298523,
+      "learning_rate": 5.748198721156284e-08,
+      "loss": 0.5297,
+      "step": 1386
+    },
+    {
+      "epoch": 2.8696551724137933,
+      "grad_norm": 0.653653621673584,
+      "learning_rate": 5.5675012281119486e-08,
+      "loss": 0.5445,
+      "step": 1387
+    },
+    {
+      "epoch": 2.8717241379310345,
+      "grad_norm": 0.6576898097991943,
+      "learning_rate": 5.389673533284956e-08,
+      "loss": 0.4969,
+      "step": 1388
+    },
+    {
+      "epoch": 2.8737931034482758,
+      "grad_norm": 0.6513156294822693,
+      "learning_rate": 5.214716668827558e-08,
+      "loss": 0.547,
+      "step": 1389
+    },
+    {
+      "epoch": 2.8758620689655174,
+      "grad_norm": 0.6595351099967957,
+      "learning_rate": 5.042631650229057e-08,
+      "loss": 0.5567,
+      "step": 1390
+    },
+    {
+      "epoch": 2.8779310344827587,
+      "grad_norm": 0.644473671913147,
+      "learning_rate": 4.8734194763098706e-08,
+      "loss": 0.5005,
+      "step": 1391
+    },
+    {
+      "epoch": 2.88,
+      "grad_norm": 0.6628778576850891,
+      "learning_rate": 4.7070811292157513e-08,
+      "loss": 0.5162,
+      "step": 1392
+    },
+    {
+      "epoch": 2.8820689655172416,
+      "grad_norm": 0.6727180480957031,
+      "learning_rate": 4.543617574412185e-08,
+      "loss": 0.5441,
+      "step": 1393
+    },
+    {
+      "epoch": 2.884137931034483,
+      "grad_norm": 0.666273832321167,
+      "learning_rate": 4.383029760678614e-08,
+      "loss": 0.5409,
+      "step": 1394
+    },
+    {
+      "epoch": 2.886206896551724,
+      "grad_norm": 0.6508522629737854,
+      "learning_rate": 4.2253186201031135e-08,
+      "loss": 0.5193,
+      "step": 1395
+    },
+    {
+      "epoch": 2.8882758620689657,
+      "grad_norm": 0.6973995566368103,
+      "learning_rate": 4.07048506807689e-08,
+      "loss": 0.5415,
+      "step": 1396
+    },
+    {
+      "epoch": 2.890344827586207,
+      "grad_norm": 0.6525006890296936,
+      "learning_rate": 3.9185300032889005e-08,
+      "loss": 0.5379,
+      "step": 1397
+    },
+    {
+      "epoch": 2.892413793103448,
+      "grad_norm": 0.677098274230957,
+      "learning_rate": 3.7694543077208546e-08,
+      "loss": 0.5413,
+      "step": 1398
+    },
+    {
+      "epoch": 2.89448275862069,
+      "grad_norm": 0.6358975172042847,
+      "learning_rate": 3.6232588466417195e-08,
+      "loss": 0.5449,
+      "step": 1399
+    },
+    {
+      "epoch": 2.896551724137931,
+      "grad_norm": 0.6577171087265015,
+      "learning_rate": 3.479944468603169e-08,
+      "loss": 0.5363,
+      "step": 1400
+    },
+    {
+      "epoch": 2.8986206896551723,
+      "grad_norm": 0.6604806780815125,
+      "learning_rate": 3.339512005434309e-08,
+      "loss": 0.5307,
+      "step": 1401
+    },
+    {
+      "epoch": 2.900689655172414,
+      "grad_norm": 0.6712777614593506,
+      "learning_rate": 3.2019622722369024e-08,
+      "loss": 0.537,
+      "step": 1402
+    },
+    {
+      "epoch": 2.902758620689655,
+      "grad_norm": 0.6645734310150146,
+      "learning_rate": 3.0672960673808205e-08,
+      "loss": 0.5775,
+      "step": 1403
+    },
+    {
+      "epoch": 2.9048275862068964,
+      "grad_norm": 0.6652960777282715,
+      "learning_rate": 2.93551417249921e-08,
+      "loss": 0.5579,
+      "step": 1404
+    },
+    {
+      "epoch": 2.906896551724138,
+      "grad_norm": 0.65793377161026,
+      "learning_rate": 2.8066173524839978e-08,
+      "loss": 0.5572,
+      "step": 1405
+    },
+    {
+      "epoch": 2.9089655172413793,
+      "grad_norm": 0.6762325167655945,
+      "learning_rate": 2.6806063554815632e-08,
+      "loss": 0.5502,
+      "step": 1406
+    },
+    {
+      "epoch": 2.9110344827586205,
+      "grad_norm": 0.6445053219795227,
+      "learning_rate": 2.5574819128882933e-08,
+      "loss": 0.5287,
+      "step": 1407
+    },
+    {
+      "epoch": 2.913103448275862,
+      "grad_norm": 0.651478111743927,
+      "learning_rate": 2.4372447393462562e-08,
+      "loss": 0.5344,
+      "step": 1408
+    },
+    {
+      "epoch": 2.9151724137931034,
+      "grad_norm": 0.6526092290878296,
+      "learning_rate": 2.319895532739369e-08,
+      "loss": 0.5081,
+      "step": 1409
+    },
+    {
+      "epoch": 2.9172413793103447,
+      "grad_norm": 0.6735363602638245,
+      "learning_rate": 2.205434974188847e-08,
+      "loss": 0.5468,
+      "step": 1410
+    },
+    {
+      "epoch": 2.9193103448275863,
+      "grad_norm": 0.6453359127044678,
+      "learning_rate": 2.093863728049872e-08,
+      "loss": 0.5245,
+      "step": 1411
+    },
+    {
+      "epoch": 2.9213793103448276,
+      "grad_norm": 0.6538481116294861,
+      "learning_rate": 1.9851824419070965e-08,
+      "loss": 0.5291,
+      "step": 1412
+    },
+    {
+      "epoch": 2.923448275862069,
+      "grad_norm": 0.6372977495193481,
+      "learning_rate": 1.8793917465713686e-08,
+      "loss": 0.5489,
+      "step": 1413
+    },
+    {
+      "epoch": 2.9255172413793105,
+      "grad_norm": 0.6539329290390015,
+      "learning_rate": 1.7764922560759014e-08,
+      "loss": 0.5307,
+      "step": 1414
+    },
+    {
+      "epoch": 2.9275862068965517,
+      "grad_norm": 0.6570875644683838,
+      "learning_rate": 1.676484567672554e-08,
+      "loss": 0.5509,
+      "step": 1415
+    },
+    {
+      "epoch": 2.929655172413793,
+      "grad_norm": 0.6732608079910278,
+      "learning_rate": 1.5793692618286115e-08,
+      "loss": 0.5501,
+      "step": 1416
+    },
+    {
+      "epoch": 2.9317241379310346,
+      "grad_norm": 0.638278067111969,
+      "learning_rate": 1.4851469022234e-08,
+      "loss": 0.5377,
+      "step": 1417
+    },
+    {
+      "epoch": 2.933793103448276,
+      "grad_norm": 0.6890416741371155,
+      "learning_rate": 1.3938180357447318e-08,
+      "loss": 0.503,
+      "step": 1418
+    },
+    {
+      "epoch": 2.935862068965517,
+      "grad_norm": 0.6768400073051453,
+      "learning_rate": 1.3053831924861315e-08,
+      "loss": 0.531,
+      "step": 1419
+    },
+    {
+      "epoch": 2.9379310344827587,
+      "grad_norm": 0.6668310165405273,
+      "learning_rate": 1.2198428857433941e-08,
+      "loss": 0.5394,
+      "step": 1420
+    },
+    {
+      "epoch": 2.94,
+      "grad_norm": 0.6531161665916443,
+      "learning_rate": 1.137197612011809e-08,
+      "loss": 0.526,
+      "step": 1421
+    },
+    {
+      "epoch": 2.942068965517241,
+      "grad_norm": 0.6525026559829712,
+      "learning_rate": 1.0574478509833286e-08,
+      "loss": 0.5175,
+      "step": 1422
+    },
+    {
+      "epoch": 2.944137931034483,
+      "grad_norm": 0.6503228545188904,
+      "learning_rate": 9.805940655436274e-09,
+      "loss": 0.534,
+      "step": 1423
+    },
+    {
+      "epoch": 2.946206896551724,
+      "grad_norm": 0.6437545418739319,
+      "learning_rate": 9.066367017694366e-09,
+      "loss": 0.5318,
+      "step": 1424
+    },
+    {
+      "epoch": 2.9482758620689653,
+      "grad_norm": 0.6803710460662842,
+      "learning_rate": 8.35576188926046e-09,
+      "loss": 0.5196,
+      "step": 1425
+    },
+    {
+      "epoch": 2.950344827586207,
+      "grad_norm": 0.6403658390045166,
+      "learning_rate": 7.674129394648623e-09,
+      "loss": 0.5252,
+      "step": 1426
+    },
+    {
+      "epoch": 2.952413793103448,
+      "grad_norm": 0.6625465750694275,
+      "learning_rate": 7.0214734902074314e-09,
+      "loss": 0.5389,
+      "step": 1427
+    },
+    {
+      "epoch": 2.9544827586206894,
+      "grad_norm": 0.6770764589309692,
+      "learning_rate": 6.3977979640994504e-09,
+      "loss": 0.5674,
+      "step": 1428
+    },
+    {
+      "epoch": 2.956551724137931,
+      "grad_norm": 0.6394649147987366,
+      "learning_rate": 5.803106436279571e-09,
+      "loss": 0.5134,
+      "step": 1429
+    },
+    {
+      "epoch": 2.9586206896551723,
+      "grad_norm": 0.6759665608406067,
+      "learning_rate": 5.237402358471144e-09,
+      "loss": 0.5349,
+      "step": 1430
+    },
+    {
+      "epoch": 2.9606896551724136,
+      "grad_norm": 0.6301043629646301,
+      "learning_rate": 4.700689014149329e-09,
+      "loss": 0.5539,
+      "step": 1431
+    },
+    {
+      "epoch": 2.9627586206896552,
+      "grad_norm": 0.6656786799430847,
+      "learning_rate": 4.192969518519441e-09,
+      "loss": 0.534,
+      "step": 1432
+    },
+    {
+      "epoch": 2.9648275862068965,
+      "grad_norm": 0.681606113910675,
+      "learning_rate": 3.71424681850141e-09,
+      "loss": 0.5136,
+      "step": 1433
+    },
+    {
+      "epoch": 2.9668965517241377,
+      "grad_norm": 0.6622211933135986,
+      "learning_rate": 3.2645236927092426e-09,
+      "loss": 0.5406,
+      "step": 1434
+    },
+    {
+      "epoch": 2.9689655172413794,
+      "grad_norm": 0.6423860192298889,
+      "learning_rate": 2.8438027514382517e-09,
+      "loss": 0.5265,
+      "step": 1435
+    },
+    {
+      "epoch": 2.9710344827586206,
+      "grad_norm": 0.6584990620613098,
+      "learning_rate": 2.452086436648404e-09,
+      "loss": 0.5342,
+      "step": 1436
+    },
+    {
+      "epoch": 2.973103448275862,
+      "grad_norm": 0.6501452922821045,
+      "learning_rate": 2.0893770219493347e-09,
+      "loss": 0.544,
+      "step": 1437
+    },
+    {
+      "epoch": 2.9751724137931035,
+      "grad_norm": 0.6575037240982056,
+      "learning_rate": 1.755676612588686e-09,
+      "loss": 0.5459,
+      "step": 1438
+    },
+    {
+      "epoch": 2.9772413793103447,
+      "grad_norm": 0.6507884860038757,
+      "learning_rate": 1.450987145439342e-09,
+      "loss": 0.5337,
+      "step": 1439
+    },
+    {
+      "epoch": 2.979310344827586,
+      "grad_norm": 0.6674343347549438,
+      "learning_rate": 1.1753103889883267e-09,
+      "loss": 0.5298,
+      "step": 1440
+    },
+    {
+      "epoch": 2.9813793103448276,
+      "grad_norm": 0.6635849475860596,
+      "learning_rate": 9.286479433257e-10,
+      "loss": 0.5505,
+      "step": 1441
+    },
+    {
+      "epoch": 2.983448275862069,
+      "grad_norm": 0.6534397006034851,
+      "learning_rate": 7.110012401362332e-10,
+      "loss": 0.5287,
+      "step": 1442
+    },
+    {
+      "epoch": 2.98551724137931,
+      "grad_norm": 0.6629782319068909,
+      "learning_rate": 5.2237154268997e-10,
+      "loss": 0.5412,
+      "step": 1443
+    },
+    {
+      "epoch": 2.9875862068965517,
+      "grad_norm": 0.628498375415802,
+      "learning_rate": 3.627599458377873e-10,
+      "loss": 0.5274,
+      "step": 1444
+    },
+    {
+      "epoch": 2.989655172413793,
+      "grad_norm": 0.6589900851249695,
+      "learning_rate": 2.321673760002918e-10,
+      "loss": 0.5285,
+      "step": 1445
+    },
+    {
+      "epoch": 2.9917241379310346,
+      "grad_norm": 0.6807800531387329,
+      "learning_rate": 1.305945911672657e-10,
+      "loss": 0.5263,
+      "step": 1446
+    },
+    {
+      "epoch": 2.993793103448276,
+      "grad_norm": 0.6564441919326782,
+      "learning_rate": 5.804218088933944e-11,
+      "loss": 0.5142,
+      "step": 1447
+    },
+    {
+      "epoch": 2.995862068965517,
+      "grad_norm": 0.6432551145553589,
+      "learning_rate": 1.4510566277992077e-11,
+      "loss": 0.5098,
+      "step": 1448
+    },
+    {
+      "epoch": 2.9979310344827588,
+      "grad_norm": 0.6673936247825623,
+      "learning_rate": 0.0,
+      "loss": 0.5185,
+      "step": 1449
+    },
+    {
+      "epoch": 2.9979310344827588,
+      "step": 1449,
+      "total_flos": 7.092133536927252e+17,
+      "train_loss": 0.6742002179780608,
+      "train_runtime": 25544.3965,
+      "train_samples_per_second": 5.449,
+      "train_steps_per_second": 0.057
+    }
+  ],
+  "logging_steps": 1,
+  "max_steps": 1449,
+  "num_input_tokens_seen": 0,
+  "num_train_epochs": 3,
+  "save_steps": 500,
+  "stateful_callbacks": {
+    "TrainerControl": {
+      "args": {
+        "should_epoch_stop": false,
+        "should_evaluate": false,
+        "should_log": false,
+        "should_save": true,
+        "should_training_stop": true
+      },
+      "attributes": {}
+    }
+  },
+  "total_flos": 7.092133536927252e+17,
+  "train_batch_size": 1,
+  "trial_name": null,
+  "trial_params": null
+}