diff --git "a/checkpoint-594/trainer_state.json" "b/checkpoint-594/trainer_state.json"
new file mode 100644--- /dev/null
+++ "b/checkpoint-594/trainer_state.json"
@@ -0,0 +1,4191 @@
+{
+  "best_metric": null,
+  "best_model_checkpoint": null,
+  "epoch": 3.0,
+  "eval_steps": 500,
+  "global_step": 594,
+  "is_hyper_param_search": false,
+  "is_local_process_zero": true,
+  "is_world_process_zero": true,
+  "log_history": [
+    {
+      "epoch": 0.005050505050505051,
+      "grad_norm": 35.239383697509766,
+      "learning_rate": 5.0000000000000004e-08,
+      "loss": 2.6493,
+      "step": 1
+    },
+    {
+      "epoch": 0.010101010101010102,
+      "grad_norm": 35.41305923461914,
+      "learning_rate": 1.0000000000000001e-07,
+      "loss": 2.6401,
+      "step": 2
+    },
+    {
+      "epoch": 0.015151515151515152,
+      "grad_norm": 35.47914505004883,
+      "learning_rate": 1.5000000000000002e-07,
+      "loss": 2.6401,
+      "step": 3
+    },
+    {
+      "epoch": 0.020202020202020204,
+      "grad_norm": 34.923484802246094,
+      "learning_rate": 2.0000000000000002e-07,
+      "loss": 2.6202,
+      "step": 4
+    },
+    {
+      "epoch": 0.025252525252525252,
+      "grad_norm": 34.7836799621582,
+      "learning_rate": 2.5000000000000004e-07,
+      "loss": 2.6071,
+      "step": 5
+    },
+    {
+      "epoch": 0.030303030303030304,
+      "grad_norm": 36.033050537109375,
+      "learning_rate": 3.0000000000000004e-07,
+      "loss": 2.6612,
+      "step": 6
+    },
+    {
+      "epoch": 0.03535353535353535,
+      "grad_norm": 34.26911926269531,
+      "learning_rate": 3.5000000000000004e-07,
+      "loss": 2.5487,
+      "step": 7
+    },
+    {
+      "epoch": 0.04040404040404041,
+      "grad_norm": 33.940216064453125,
+      "learning_rate": 4.0000000000000003e-07,
+      "loss": 2.5171,
+      "step": 8
+    },
+    {
+      "epoch": 0.045454545454545456,
+      "grad_norm": 33.55999755859375,
+      "learning_rate": 4.5000000000000003e-07,
+      "loss": 2.5295,
+      "step": 9
+    },
+    {
+      "epoch": 0.050505050505050504,
+      "grad_norm": 35.171165466308594,
+      "learning_rate": 5.000000000000001e-07,
+      "loss": 2.6192,
+      "step": 10
+    },
+    {
+      "epoch": 0.05555555555555555,
+      "grad_norm": 32.917484283447266,
+      "learning_rate": 5.5e-07,
+      "loss": 2.4022,
+      "step": 11
+    },
+    {
+      "epoch": 0.06060606060606061,
+      "grad_norm": 34.53117752075195,
+      "learning_rate": 6.000000000000001e-07,
+      "loss": 2.5785,
+      "step": 12
+    },
+    {
+      "epoch": 0.06565656565656566,
+      "grad_norm": 34.89581298828125,
+      "learning_rate": 6.5e-07,
+      "loss": 2.52,
+      "step": 13
+    },
+    {
+      "epoch": 0.0707070707070707,
+      "grad_norm": 32.23255920410156,
+      "learning_rate": 7.000000000000001e-07,
+      "loss": 2.2731,
+      "step": 14
+    },
+    {
+      "epoch": 0.07575757575757576,
+      "grad_norm": 33.4830207824707,
+      "learning_rate": 7.5e-07,
+      "loss": 2.3016,
+      "step": 15
+    },
+    {
+      "epoch": 0.08080808080808081,
+      "grad_norm": 33.69716262817383,
+      "learning_rate": 8.000000000000001e-07,
+      "loss": 2.2107,
+      "step": 16
+    },
+    {
+      "epoch": 0.08585858585858586,
+      "grad_norm": 33.970333099365234,
+      "learning_rate": 8.500000000000001e-07,
+      "loss": 2.1916,
+      "step": 17
+    },
+    {
+      "epoch": 0.09090909090909091,
+      "grad_norm": 32.690433502197266,
+      "learning_rate": 9.000000000000001e-07,
+      "loss": 1.9767,
+      "step": 18
+    },
+    {
+      "epoch": 0.09595959595959595,
+      "grad_norm": 34.36090850830078,
+      "learning_rate": 9.500000000000001e-07,
+      "loss": 1.9856,
+      "step": 19
+    },
+    {
+      "epoch": 0.10101010101010101,
+      "grad_norm": 34.156280517578125,
+      "learning_rate": 1.0000000000000002e-06,
+      "loss": 1.7905,
+      "step": 20
+    },
+    {
+      "epoch": 0.10606060606060606,
+      "grad_norm": 34.23146438598633,
+      "learning_rate": 1.0500000000000001e-06,
+      "loss": 1.6931,
+      "step": 21
+    },
+    {
+      "epoch": 0.1111111111111111,
+      "grad_norm": 33.700294494628906,
+      "learning_rate": 1.1e-06,
+      "loss": 1.5142,
+      "step": 22
+    },
+    {
+      "epoch": 0.11616161616161616,
+      "grad_norm": 32.60850143432617,
+      "learning_rate": 1.1500000000000002e-06,
+      "loss": 1.4211,
+      "step": 23
+    },
+    {
+      "epoch": 0.12121212121212122,
+      "grad_norm": 30.005735397338867,
+      "learning_rate": 1.2000000000000002e-06,
+      "loss": 1.2172,
+      "step": 24
+    },
+    {
+      "epoch": 0.12626262626262627,
+      "grad_norm": 29.669198989868164,
+      "learning_rate": 1.25e-06,
+      "loss": 1.1584,
+      "step": 25
+    },
+    {
+      "epoch": 0.13131313131313133,
+      "grad_norm": 30.327133178710938,
+      "learning_rate": 1.3e-06,
+      "loss": 0.8788,
+      "step": 26
+    },
+    {
+      "epoch": 0.13636363636363635,
+      "grad_norm": 29.821216583251953,
+      "learning_rate": 1.3500000000000002e-06,
+      "loss": 0.7566,
+      "step": 27
+    },
+    {
+      "epoch": 0.1414141414141414,
+      "grad_norm": 27.04122543334961,
+      "learning_rate": 1.4000000000000001e-06,
+      "loss": 0.5775,
+      "step": 28
+    },
+    {
+      "epoch": 0.14646464646464646,
+      "grad_norm": 24.890750885009766,
+      "learning_rate": 1.45e-06,
+      "loss": 0.5193,
+      "step": 29
+    },
+    {
+      "epoch": 0.15151515151515152,
+      "grad_norm": 21.845609664916992,
+      "learning_rate": 1.5e-06,
+      "loss": 0.3785,
+      "step": 30
+    },
+    {
+      "epoch": 0.15656565656565657,
+      "grad_norm": 17.295909881591797,
+      "learning_rate": 1.5500000000000002e-06,
+      "loss": 0.3091,
+      "step": 31
+    },
+    {
+      "epoch": 0.16161616161616163,
+      "grad_norm": 11.265312194824219,
+      "learning_rate": 1.6000000000000001e-06,
+      "loss": 0.1745,
+      "step": 32
+    },
+    {
+      "epoch": 0.16666666666666666,
+      "grad_norm": 5.728430271148682,
+      "learning_rate": 1.6500000000000003e-06,
+      "loss": 0.1413,
+      "step": 33
+    },
+    {
+      "epoch": 0.1717171717171717,
+      "grad_norm": 7.306532859802246,
+      "learning_rate": 1.7000000000000002e-06,
+      "loss": 0.1613,
+      "step": 34
+    },
+    {
+      "epoch": 0.17676767676767677,
+      "grad_norm": 5.182886600494385,
+      "learning_rate": 1.75e-06,
+      "loss": 0.1155,
+      "step": 35
+    },
+    {
+      "epoch": 0.18181818181818182,
+      "grad_norm": 3.8576388359069824,
+      "learning_rate": 1.8000000000000001e-06,
+      "loss": 0.0764,
+      "step": 36
+    },
+    {
+      "epoch": 0.18686868686868688,
+      "grad_norm": 2.638275623321533,
+      "learning_rate": 1.85e-06,
+      "loss": 0.0745,
+      "step": 37
+    },
+    {
+      "epoch": 0.1919191919191919,
+      "grad_norm": 2.225215435028076,
+      "learning_rate": 1.9000000000000002e-06,
+      "loss": 0.0719,
+      "step": 38
+    },
+    {
+      "epoch": 0.19696969696969696,
+      "grad_norm": 1.7595651149749756,
+      "learning_rate": 1.9500000000000004e-06,
+      "loss": 0.0633,
+      "step": 39
+    },
+    {
+      "epoch": 0.20202020202020202,
+      "grad_norm": 1.4759680032730103,
+      "learning_rate": 2.0000000000000003e-06,
+      "loss": 0.0529,
+      "step": 40
+    },
+    {
+      "epoch": 0.20707070707070707,
+      "grad_norm": 1.3384065628051758,
+      "learning_rate": 2.05e-06,
+      "loss": 0.0433,
+      "step": 41
+    },
+    {
+      "epoch": 0.21212121212121213,
+      "grad_norm": 1.564224362373352,
+      "learning_rate": 2.1000000000000002e-06,
+      "loss": 0.061,
+      "step": 42
+    },
+    {
+      "epoch": 0.21717171717171718,
+      "grad_norm": 1.419657826423645,
+      "learning_rate": 2.15e-06,
+      "loss": 0.0615,
+      "step": 43
+    },
+    {
+      "epoch": 0.2222222222222222,
+      "grad_norm": 1.1758321523666382,
+      "learning_rate": 2.2e-06,
+      "loss": 0.059,
+      "step": 44
+    },
+    {
+      "epoch": 0.22727272727272727,
+      "grad_norm": 1.3546487092971802,
+      "learning_rate": 2.25e-06,
+      "loss": 0.0574,
+      "step": 45
+    },
+    {
+      "epoch": 0.23232323232323232,
+      "grad_norm": 1.0917863845825195,
+      "learning_rate": 2.3000000000000004e-06,
+      "loss": 0.0601,
+      "step": 46
+    },
+    {
+      "epoch": 0.23737373737373738,
+      "grad_norm": 0.990371823310852,
+      "learning_rate": 2.35e-06,
+      "loss": 0.0515,
+      "step": 47
+    },
+    {
+      "epoch": 0.24242424242424243,
+      "grad_norm": 1.3220105171203613,
+      "learning_rate": 2.4000000000000003e-06,
+      "loss": 0.0584,
+      "step": 48
+    },
+    {
+      "epoch": 0.2474747474747475,
+      "grad_norm": 1.1640170812606812,
+      "learning_rate": 2.4500000000000003e-06,
+      "loss": 0.0487,
+      "step": 49
+    },
+    {
+      "epoch": 0.25252525252525254,
+      "grad_norm": 1.0264198780059814,
+      "learning_rate": 2.5e-06,
+      "loss": 0.0456,
+      "step": 50
+    },
+    {
+      "epoch": 0.25757575757575757,
+      "grad_norm": 1.1187573671340942,
+      "learning_rate": 2.55e-06,
+      "loss": 0.0588,
+      "step": 51
+    },
+    {
+      "epoch": 0.26262626262626265,
+      "grad_norm": 0.8740547299385071,
+      "learning_rate": 2.6e-06,
+      "loss": 0.0415,
+      "step": 52
+    },
+    {
+      "epoch": 0.2676767676767677,
+      "grad_norm": 1.128773808479309,
+      "learning_rate": 2.6500000000000005e-06,
+      "loss": 0.0535,
+      "step": 53
+    },
+    {
+      "epoch": 0.2727272727272727,
+      "grad_norm": 1.1845512390136719,
+      "learning_rate": 2.7000000000000004e-06,
+      "loss": 0.0543,
+      "step": 54
+    },
+    {
+      "epoch": 0.2777777777777778,
+      "grad_norm": 1.0151381492614746,
+      "learning_rate": 2.7500000000000004e-06,
+      "loss": 0.0467,
+      "step": 55
+    },
+    {
+      "epoch": 0.2828282828282828,
+      "grad_norm": 1.0951204299926758,
+      "learning_rate": 2.8000000000000003e-06,
+      "loss": 0.0432,
+      "step": 56
+    },
+    {
+      "epoch": 0.2878787878787879,
+      "grad_norm": 0.8145187497138977,
+      "learning_rate": 2.85e-06,
+      "loss": 0.0433,
+      "step": 57
+    },
+    {
+      "epoch": 0.29292929292929293,
+      "grad_norm": 0.8338307738304138,
+      "learning_rate": 2.9e-06,
+      "loss": 0.0442,
+      "step": 58
+    },
+    {
+      "epoch": 0.29797979797979796,
+      "grad_norm": 1.2009202241897583,
+      "learning_rate": 2.95e-06,
+      "loss": 0.0471,
+      "step": 59
+    },
+    {
+      "epoch": 0.30303030303030304,
+      "grad_norm": 1.0247856378555298,
+      "learning_rate": 3e-06,
+      "loss": 0.0418,
+      "step": 60
+    },
+    {
+      "epoch": 0.30808080808080807,
+      "grad_norm": 0.8393287062644958,
+      "learning_rate": 3.05e-06,
+      "loss": 0.0411,
+      "step": 61
+    },
+    {
+      "epoch": 0.31313131313131315,
+      "grad_norm": 0.807152271270752,
+      "learning_rate": 3.1000000000000004e-06,
+      "loss": 0.0428,
+      "step": 62
+    },
+    {
+      "epoch": 0.3181818181818182,
+      "grad_norm": 1.0362982749938965,
+      "learning_rate": 3.1500000000000003e-06,
+      "loss": 0.0412,
+      "step": 63
+    },
+    {
+      "epoch": 0.32323232323232326,
+      "grad_norm": 0.8612103462219238,
+      "learning_rate": 3.2000000000000003e-06,
+      "loss": 0.0374,
+      "step": 64
+    },
+    {
+      "epoch": 0.3282828282828283,
+      "grad_norm": 0.763271152973175,
+      "learning_rate": 3.2500000000000002e-06,
+      "loss": 0.038,
+      "step": 65
+    },
+    {
+      "epoch": 0.3333333333333333,
+      "grad_norm": 0.993196427822113,
+      "learning_rate": 3.3000000000000006e-06,
+      "loss": 0.0476,
+      "step": 66
+    },
+    {
+      "epoch": 0.3383838383838384,
+      "grad_norm": 1.1754580736160278,
+      "learning_rate": 3.3500000000000005e-06,
+      "loss": 0.0496,
+      "step": 67
+    },
+    {
+      "epoch": 0.3434343434343434,
+      "grad_norm": 0.7608982920646667,
+      "learning_rate": 3.4000000000000005e-06,
+      "loss": 0.0371,
+      "step": 68
+    },
+    {
+      "epoch": 0.3484848484848485,
+      "grad_norm": 0.8498206734657288,
+      "learning_rate": 3.45e-06,
+      "loss": 0.0422,
+      "step": 69
+    },
+    {
+      "epoch": 0.35353535353535354,
+      "grad_norm": 0.8418146371841431,
+      "learning_rate": 3.5e-06,
+      "loss": 0.0361,
+      "step": 70
+    },
+    {
+      "epoch": 0.35858585858585856,
+      "grad_norm": 0.8102655410766602,
+      "learning_rate": 3.5500000000000003e-06,
+      "loss": 0.0356,
+      "step": 71
+    },
+    {
+      "epoch": 0.36363636363636365,
+      "grad_norm": 0.7071998119354248,
+      "learning_rate": 3.6000000000000003e-06,
+      "loss": 0.0352,
+      "step": 72
+    },
+    {
+      "epoch": 0.3686868686868687,
+      "grad_norm": 0.8242170810699463,
+      "learning_rate": 3.65e-06,
+      "loss": 0.0393,
+      "step": 73
+    },
+    {
+      "epoch": 0.37373737373737376,
+      "grad_norm": 0.93561190366745,
+      "learning_rate": 3.7e-06,
+      "loss": 0.047,
+      "step": 74
+    },
+    {
+      "epoch": 0.3787878787878788,
+      "grad_norm": 0.8524022698402405,
+      "learning_rate": 3.7500000000000005e-06,
+      "loss": 0.0325,
+      "step": 75
+    },
+    {
+      "epoch": 0.3838383838383838,
+      "grad_norm": 0.8019789457321167,
+      "learning_rate": 3.8000000000000005e-06,
+      "loss": 0.0358,
+      "step": 76
+    },
+    {
+      "epoch": 0.3888888888888889,
+      "grad_norm": 0.9955350160598755,
+      "learning_rate": 3.85e-06,
+      "loss": 0.0373,
+      "step": 77
+    },
+    {
+      "epoch": 0.3939393939393939,
+      "grad_norm": 1.2481234073638916,
+      "learning_rate": 3.900000000000001e-06,
+      "loss": 0.0353,
+      "step": 78
+    },
+    {
+      "epoch": 0.398989898989899,
+      "grad_norm": 0.7706320285797119,
+      "learning_rate": 3.95e-06,
+      "loss": 0.0352,
+      "step": 79
+    },
+    {
+      "epoch": 0.40404040404040403,
+      "grad_norm": 0.8946953415870667,
+      "learning_rate": 4.000000000000001e-06,
+      "loss": 0.0389,
+      "step": 80
+    },
+    {
+      "epoch": 0.4090909090909091,
+      "grad_norm": 0.7142760753631592,
+      "learning_rate": 4.05e-06,
+      "loss": 0.0347,
+      "step": 81
+    },
+    {
+      "epoch": 0.41414141414141414,
+      "grad_norm": 0.6389093399047852,
+      "learning_rate": 4.1e-06,
+      "loss": 0.0352,
+      "step": 82
+    },
+    {
+      "epoch": 0.41919191919191917,
+      "grad_norm": 0.9447190165519714,
+      "learning_rate": 4.15e-06,
+      "loss": 0.0392,
+      "step": 83
+    },
+    {
+      "epoch": 0.42424242424242425,
+      "grad_norm": 0.832013726234436,
+      "learning_rate": 4.2000000000000004e-06,
+      "loss": 0.0361,
+      "step": 84
+    },
+    {
+      "epoch": 0.4292929292929293,
+      "grad_norm": 0.8298827409744263,
+      "learning_rate": 4.25e-06,
+      "loss": 0.0342,
+      "step": 85
+    },
+    {
+      "epoch": 0.43434343434343436,
+      "grad_norm": 0.8309102654457092,
+      "learning_rate": 4.3e-06,
+      "loss": 0.0253,
+      "step": 86
+    },
+    {
+      "epoch": 0.4393939393939394,
+      "grad_norm": 0.8921272158622742,
+      "learning_rate": 4.350000000000001e-06,
+      "loss": 0.0356,
+      "step": 87
+    },
+    {
+      "epoch": 0.4444444444444444,
+      "grad_norm": 0.9582182765007019,
+      "learning_rate": 4.4e-06,
+      "loss": 0.0307,
+      "step": 88
+    },
+    {
+      "epoch": 0.4494949494949495,
+      "grad_norm": 0.9414967894554138,
+      "learning_rate": 4.450000000000001e-06,
+      "loss": 0.0347,
+      "step": 89
+    },
+    {
+      "epoch": 0.45454545454545453,
+      "grad_norm": 0.8414708375930786,
+      "learning_rate": 4.5e-06,
+      "loss": 0.0324,
+      "step": 90
+    },
+    {
+      "epoch": 0.4595959595959596,
+      "grad_norm": 0.97618168592453,
+      "learning_rate": 4.5500000000000005e-06,
+      "loss": 0.035,
+      "step": 91
+    },
+    {
+      "epoch": 0.46464646464646464,
+      "grad_norm": 0.7519140839576721,
+      "learning_rate": 4.600000000000001e-06,
+      "loss": 0.0291,
+      "step": 92
+    },
+    {
+      "epoch": 0.4696969696969697,
+      "grad_norm": 0.7026158571243286,
+      "learning_rate": 4.65e-06,
+      "loss": 0.0343,
+      "step": 93
+    },
+    {
+      "epoch": 0.47474747474747475,
+      "grad_norm": 0.8826281428337097,
+      "learning_rate": 4.7e-06,
+      "loss": 0.0283,
+      "step": 94
+    },
+    {
+      "epoch": 0.4797979797979798,
+      "grad_norm": 0.8762742877006531,
+      "learning_rate": 4.75e-06,
+      "loss": 0.027,
+      "step": 95
+    },
+    {
+      "epoch": 0.48484848484848486,
+      "grad_norm": 0.7252739667892456,
+      "learning_rate": 4.800000000000001e-06,
+      "loss": 0.0286,
+      "step": 96
+    },
+    {
+      "epoch": 0.4898989898989899,
+      "grad_norm": 0.7852551341056824,
+      "learning_rate": 4.85e-06,
+      "loss": 0.0287,
+      "step": 97
+    },
+    {
+      "epoch": 0.494949494949495,
+      "grad_norm": 0.6870710849761963,
+      "learning_rate": 4.9000000000000005e-06,
+      "loss": 0.0324,
+      "step": 98
+    },
+    {
+      "epoch": 0.5,
+      "grad_norm": 1.0492205619812012,
+      "learning_rate": 4.95e-06,
+      "loss": 0.0384,
+      "step": 99
+    },
+    {
+      "epoch": 0.5050505050505051,
+      "grad_norm": 0.9691818952560425,
+      "learning_rate": 5e-06,
+      "loss": 0.0324,
+      "step": 100
+    },
+    {
+      "epoch": 0.51010101010101,
+      "grad_norm": 0.7680443525314331,
+      "learning_rate": 4.999989577985671e-06,
+      "loss": 0.0239,
+      "step": 101
+    },
+    {
+      "epoch": 0.5151515151515151,
+      "grad_norm": 0.7711728811264038,
+      "learning_rate": 4.99995831202958e-06,
+      "loss": 0.0339,
+      "step": 102
+    },
+    {
+      "epoch": 0.5202020202020202,
+      "grad_norm": 0.664016842842102,
+      "learning_rate": 4.999906202392409e-06,
+      "loss": 0.0242,
+      "step": 103
+    },
+    {
+      "epoch": 0.5252525252525253,
+      "grad_norm": 0.6629706621170044,
+      "learning_rate": 4.999833249508629e-06,
+      "loss": 0.0272,
+      "step": 104
+    },
+    {
+      "epoch": 0.5303030303030303,
+      "grad_norm": 0.8449888229370117,
+      "learning_rate": 4.999739453986491e-06,
+      "loss": 0.0313,
+      "step": 105
+    },
+    {
+      "epoch": 0.5353535353535354,
+      "grad_norm": 0.9404881000518799,
+      "learning_rate": 4.999624816608027e-06,
+      "loss": 0.0333,
+      "step": 106
+    },
+    {
+      "epoch": 0.5404040404040404,
+      "grad_norm": 0.8782823085784912,
+      "learning_rate": 4.99948933832904e-06,
+      "loss": 0.0298,
+      "step": 107
+    },
+    {
+      "epoch": 0.5454545454545454,
+      "grad_norm": 0.8479216694831848,
+      "learning_rate": 4.999333020279094e-06,
+      "loss": 0.0305,
+      "step": 108
+    },
+    {
+      "epoch": 0.5505050505050505,
+      "grad_norm": 0.7747307419776917,
+      "learning_rate": 4.999155863761507e-06,
+      "loss": 0.0275,
+      "step": 109
+    },
+    {
+      "epoch": 0.5555555555555556,
+      "grad_norm": 0.8345880508422852,
+      "learning_rate": 4.998957870253344e-06,
+      "loss": 0.0326,
+      "step": 110
+    },
+    {
+      "epoch": 0.5606060606060606,
+      "grad_norm": 0.84605473279953,
+      "learning_rate": 4.998739041405395e-06,
+      "loss": 0.0302,
+      "step": 111
+    },
+    {
+      "epoch": 0.5656565656565656,
+      "grad_norm": 0.7808102369308472,
+      "learning_rate": 4.998499379042172e-06,
+      "loss": 0.0262,
+      "step": 112
+    },
+    {
+      "epoch": 0.5707070707070707,
+      "grad_norm": 0.8007850050926208,
+      "learning_rate": 4.998238885161886e-06,
+      "loss": 0.0272,
+      "step": 113
+    },
+    {
+      "epoch": 0.5757575757575758,
+      "grad_norm": 0.8897403478622437,
+      "learning_rate": 4.997957561936433e-06,
+      "loss": 0.0261,
+      "step": 114
+    },
+    {
+      "epoch": 0.5808080808080808,
+      "grad_norm": 0.8403050899505615,
+      "learning_rate": 4.997655411711378e-06,
+      "loss": 0.0292,
+      "step": 115
+    },
+    {
+      "epoch": 0.5858585858585859,
+      "grad_norm": 0.7349573969841003,
+      "learning_rate": 4.997332437005932e-06,
+      "loss": 0.0243,
+      "step": 116
+    },
+    {
+      "epoch": 0.5909090909090909,
+      "grad_norm": 0.8440062999725342,
+      "learning_rate": 4.996988640512931e-06,
+      "loss": 0.0253,
+      "step": 117
+    },
+    {
+      "epoch": 0.5959595959595959,
+      "grad_norm": 0.8826310038566589,
+      "learning_rate": 4.996624025098819e-06,
+      "loss": 0.0209,
+      "step": 118
+    },
+    {
+      "epoch": 0.601010101010101,
+      "grad_norm": 1.1104847192764282,
+      "learning_rate": 4.996238593803616e-06,
+      "loss": 0.0242,
+      "step": 119
+    },
+    {
+      "epoch": 0.6060606060606061,
+      "grad_norm": 1.3400338888168335,
+      "learning_rate": 4.9958323498409e-06,
+      "loss": 0.0347,
+      "step": 120
+    },
+    {
+      "epoch": 0.6111111111111112,
+      "grad_norm": 0.9243490099906921,
+      "learning_rate": 4.9954052965977725e-06,
+      "loss": 0.0202,
+      "step": 121
+    },
+    {
+      "epoch": 0.6161616161616161,
+      "grad_norm": 1.075374960899353,
+      "learning_rate": 4.99495743763484e-06,
+      "loss": 0.0309,
+      "step": 122
+    },
+    {
+      "epoch": 0.6212121212121212,
+      "grad_norm": 1.0591344833374023,
+      "learning_rate": 4.9944887766861765e-06,
+      "loss": 0.0241,
+      "step": 123
+    },
+    {
+      "epoch": 0.6262626262626263,
+      "grad_norm": 0.7506961226463318,
+      "learning_rate": 4.993999317659293e-06,
+      "loss": 0.025,
+      "step": 124
+    },
+    {
+      "epoch": 0.6313131313131313,
+      "grad_norm": 1.016758680343628,
+      "learning_rate": 4.993489064635109e-06,
+      "loss": 0.0219,
+      "step": 125
+    },
+    {
+      "epoch": 0.6363636363636364,
+      "grad_norm": 0.7476636171340942,
+      "learning_rate": 4.9929580218679195e-06,
+      "loss": 0.021,
+      "step": 126
+    },
+    {
+      "epoch": 0.6414141414141414,
+      "grad_norm": 0.7349154949188232,
+      "learning_rate": 4.992406193785348e-06,
+      "loss": 0.0228,
+      "step": 127
+    },
+    {
+      "epoch": 0.6464646464646465,
+      "grad_norm": 0.672728419303894,
+      "learning_rate": 4.991833584988326e-06,
+      "loss": 0.0209,
+      "step": 128
+    },
+    {
+      "epoch": 0.6515151515151515,
+      "grad_norm": 0.5250627398490906,
+      "learning_rate": 4.991240200251041e-06,
+      "loss": 0.0124,
+      "step": 129
+    },
+    {
+      "epoch": 0.6565656565656566,
+      "grad_norm": 1.0610628128051758,
+      "learning_rate": 4.990626044520905e-06,
+      "loss": 0.0176,
+      "step": 130
+    },
+    {
+      "epoch": 0.6616161616161617,
+      "grad_norm": 1.0880004167556763,
+      "learning_rate": 4.98999112291851e-06,
+      "loss": 0.024,
+      "step": 131
+    },
+    {
+      "epoch": 0.6666666666666666,
+      "grad_norm": 0.9670020937919617,
+      "learning_rate": 4.989335440737587e-06,
+      "loss": 0.0224,
+      "step": 132
+    },
+    {
+      "epoch": 0.6717171717171717,
+      "grad_norm": 0.869041383266449,
+      "learning_rate": 4.988659003444956e-06,
+      "loss": 0.0184,
+      "step": 133
+    },
+    {
+      "epoch": 0.6767676767676768,
+      "grad_norm": 0.9090948700904846,
+      "learning_rate": 4.987961816680493e-06,
+      "loss": 0.0192,
+      "step": 134
+    },
+    {
+      "epoch": 0.6818181818181818,
+      "grad_norm": 1.0600134134292603,
+      "learning_rate": 4.987243886257066e-06,
+      "loss": 0.029,
+      "step": 135
+    },
+    {
+      "epoch": 0.6868686868686869,
+      "grad_norm": 0.5932410359382629,
+      "learning_rate": 4.986505218160502e-06,
+      "loss": 0.0172,
+      "step": 136
+    },
+    {
+      "epoch": 0.6919191919191919,
+      "grad_norm": 0.8293594121932983,
+      "learning_rate": 4.985745818549527e-06,
+      "loss": 0.0207,
+      "step": 137
+    },
+    {
+      "epoch": 0.696969696969697,
+      "grad_norm": 0.863837718963623,
+      "learning_rate": 4.984965693755723e-06,
+      "loss": 0.0262,
+      "step": 138
+    },
+    {
+      "epoch": 0.702020202020202,
+      "grad_norm": 0.8079712390899658,
+      "learning_rate": 4.984164850283465e-06,
+      "loss": 0.0226,
+      "step": 139
+    },
+    {
+      "epoch": 0.7070707070707071,
+      "grad_norm": 0.7353380918502808,
+      "learning_rate": 4.983343294809875e-06,
+      "loss": 0.031,
+      "step": 140
+    },
+    {
+      "epoch": 0.7121212121212122,
+      "grad_norm": 0.6296212673187256,
+      "learning_rate": 4.9825010341847644e-06,
+      "loss": 0.016,
+      "step": 141
+    },
+    {
+      "epoch": 0.7171717171717171,
+      "grad_norm": 0.9319673180580139,
+      "learning_rate": 4.981638075430572e-06,
+      "loss": 0.0213,
+      "step": 142
+    },
+    {
+      "epoch": 0.7222222222222222,
+      "grad_norm": 0.6460096836090088,
+      "learning_rate": 4.980754425742318e-06,
+      "loss": 0.0203,
+      "step": 143
+    },
+    {
+      "epoch": 0.7272727272727273,
+      "grad_norm": 0.6757596731185913,
+      "learning_rate": 4.979850092487525e-06,
+      "loss": 0.0175,
+      "step": 144
+    },
+    {
+      "epoch": 0.7323232323232324,
+      "grad_norm": 0.7001456022262573,
+      "learning_rate": 4.978925083206174e-06,
+      "loss": 0.021,
+      "step": 145
+    },
+    {
+      "epoch": 0.7373737373737373,
+      "grad_norm": 0.602177619934082,
+      "learning_rate": 4.977979405610635e-06,
+      "loss": 0.0168,
+      "step": 146
+    },
+    {
+      "epoch": 0.7424242424242424,
+      "grad_norm": 0.7525290250778198,
+      "learning_rate": 4.977013067585597e-06,
+      "loss": 0.022,
+      "step": 147
+    },
+    {
+      "epoch": 0.7474747474747475,
+      "grad_norm": 0.6703810095787048,
+      "learning_rate": 4.976026077188013e-06,
+      "loss": 0.016,
+      "step": 148
+    },
+    {
+      "epoch": 0.7525252525252525,
+      "grad_norm": 0.7743527293205261,
+      "learning_rate": 4.975018442647026e-06,
+      "loss": 0.0226,
+      "step": 149
+    },
+    {
+      "epoch": 0.7575757575757576,
+      "grad_norm": 0.6670228838920593,
+      "learning_rate": 4.973990172363899e-06,
+      "loss": 0.0181,
+      "step": 150
+    },
+    {
+      "epoch": 0.7626262626262627,
+      "grad_norm": 0.7247374653816223,
+      "learning_rate": 4.972941274911953e-06,
+      "loss": 0.0146,
+      "step": 151
+    },
+    {
+      "epoch": 0.7676767676767676,
+      "grad_norm": 0.6329397559165955,
+      "learning_rate": 4.9718717590364855e-06,
+      "loss": 0.0165,
+      "step": 152
+    },
+    {
+      "epoch": 0.7727272727272727,
+      "grad_norm": 0.859639048576355,
+      "learning_rate": 4.9707816336547045e-06,
+      "loss": 0.0216,
+      "step": 153
+    },
+    {
+      "epoch": 0.7777777777777778,
+      "grad_norm": 0.7974834442138672,
+      "learning_rate": 4.969670907855651e-06,
+      "loss": 0.0212,
+      "step": 154
+    },
+    {
+      "epoch": 0.7828282828282829,
+      "grad_norm": 1.1814897060394287,
+      "learning_rate": 4.968539590900128e-06,
+      "loss": 0.0231,
+      "step": 155
+    },
+    {
+      "epoch": 0.7878787878787878,
+      "grad_norm": 0.7747610211372375,
+      "learning_rate": 4.967387692220615e-06,
+      "loss": 0.0137,
+      "step": 156
+    },
+    {
+      "epoch": 0.7929292929292929,
+      "grad_norm": 0.5962601900100708,
+      "learning_rate": 4.966215221421195e-06,
+      "loss": 0.0139,
+      "step": 157
+    },
+    {
+      "epoch": 0.797979797979798,
+      "grad_norm": 1.1249876022338867,
+      "learning_rate": 4.965022188277474e-06,
+      "loss": 0.0192,
+      "step": 158
+    },
+    {
+      "epoch": 0.803030303030303,
+      "grad_norm": 0.559429407119751,
+      "learning_rate": 4.9638086027365005e-06,
+      "loss": 0.0128,
+      "step": 159
+    },
+    {
+      "epoch": 0.8080808080808081,
+      "grad_norm": 1.0353529453277588,
+      "learning_rate": 4.962574474916678e-06,
+      "loss": 0.0179,
+      "step": 160
+    },
+    {
+      "epoch": 0.8131313131313131,
+      "grad_norm": 0.5366432666778564,
+      "learning_rate": 4.961319815107685e-06,
+      "loss": 0.0098,
+      "step": 161
+    },
+    {
+      "epoch": 0.8181818181818182,
+      "grad_norm": 0.673726499080658,
+      "learning_rate": 4.960044633770387e-06,
+      "loss": 0.0144,
+      "step": 162
+    },
+    {
+      "epoch": 0.8232323232323232,
+      "grad_norm": 0.517419695854187,
+      "learning_rate": 4.95874894153675e-06,
+      "loss": 0.016,
+      "step": 163
+    },
+    {
+      "epoch": 0.8282828282828283,
+      "grad_norm": 0.5697750449180603,
+      "learning_rate": 4.957432749209755e-06,
+      "loss": 0.0117,
+      "step": 164
+    },
+    {
+      "epoch": 0.8333333333333334,
+      "grad_norm": 1.0551936626434326,
+      "learning_rate": 4.9560960677633e-06,
+      "loss": 0.0179,
+      "step": 165
+    },
+    {
+      "epoch": 0.8383838383838383,
+      "grad_norm": 0.564612865447998,
+      "learning_rate": 4.954738908342116e-06,
+      "loss": 0.0112,
+      "step": 166
+    },
+    {
+      "epoch": 0.8434343434343434,
+      "grad_norm": 0.7717934846878052,
+      "learning_rate": 4.953361282261671e-06,
+      "loss": 0.0121,
+      "step": 167
+    },
+    {
+      "epoch": 0.8484848484848485,
+      "grad_norm": 0.6095231175422668,
+      "learning_rate": 4.9519632010080765e-06,
+      "loss": 0.0071,
+      "step": 168
+    },
+    {
+      "epoch": 0.8535353535353535,
+      "grad_norm": 0.7816880941390991,
+      "learning_rate": 4.95054467623799e-06,
+      "loss": 0.0107,
+      "step": 169
+    },
+    {
+      "epoch": 0.8585858585858586,
+      "grad_norm": 0.5223932266235352,
+      "learning_rate": 4.9491057197785205e-06,
+      "loss": 0.0106,
+      "step": 170
+    },
+    {
+      "epoch": 0.8636363636363636,
+      "grad_norm": 0.9016621708869934,
+      "learning_rate": 4.947646343627128e-06,
+      "loss": 0.0118,
+      "step": 171
+    },
+    {
+      "epoch": 0.8686868686868687,
+      "grad_norm": 0.4834199547767639,
+      "learning_rate": 4.946166559951523e-06,
+      "loss": 0.0063,
+      "step": 172
+    },
+    {
+      "epoch": 0.8737373737373737,
+      "grad_norm": 0.8252310752868652,
+      "learning_rate": 4.944666381089567e-06,
+      "loss": 0.0175,
+      "step": 173
+    },
+    {
+      "epoch": 0.8787878787878788,
+      "grad_norm": 0.6991069912910461,
+      "learning_rate": 4.943145819549169e-06,
+      "loss": 0.0104,
+      "step": 174
+    },
+    {
+      "epoch": 0.8838383838383839,
+      "grad_norm": 0.6017345786094666,
+      "learning_rate": 4.941604888008181e-06,
+      "loss": 0.0117,
+      "step": 175
+    },
+    {
+      "epoch": 0.8888888888888888,
+      "grad_norm": 0.69279944896698,
+      "learning_rate": 4.9400435993142895e-06,
+      "loss": 0.0137,
+      "step": 176
+    },
+    {
+      "epoch": 0.8939393939393939,
+      "grad_norm": 1.0663758516311646,
+      "learning_rate": 4.938461966484914e-06,
+      "loss": 0.01,
+      "step": 177
+    },
+    {
+      "epoch": 0.898989898989899,
+      "grad_norm": 0.9510229229927063,
+      "learning_rate": 4.936860002707096e-06,
+      "loss": 0.0085,
+      "step": 178
+    },
+    {
+      "epoch": 0.9040404040404041,
+      "grad_norm": 0.8557943105697632,
+      "learning_rate": 4.935237721337384e-06,
+      "loss": 0.0141,
+      "step": 179
+    },
+    {
+      "epoch": 0.9090909090909091,
+      "grad_norm": 0.7711794972419739,
+      "learning_rate": 4.933595135901733e-06,
+      "loss": 0.0117,
+      "step": 180
+    },
+    {
+      "epoch": 0.9141414141414141,
+      "grad_norm": 0.8024590015411377,
+      "learning_rate": 4.931932260095379e-06,
+      "loss": 0.01,
+      "step": 181
+    },
+    {
+      "epoch": 0.9191919191919192,
+      "grad_norm": 0.8047268390655518,
+      "learning_rate": 4.9302491077827366e-06,
+      "loss": 0.0192,
+      "step": 182
+    },
+    {
+      "epoch": 0.9242424242424242,
+      "grad_norm": 0.5250588655471802,
+      "learning_rate": 4.928545692997275e-06,
+      "loss": 0.0096,
+      "step": 183
+    },
+    {
+      "epoch": 0.9292929292929293,
+      "grad_norm": 0.45513755083084106,
+      "learning_rate": 4.926822029941406e-06,
+      "loss": 0.0087,
+      "step": 184
+    },
+    {
+      "epoch": 0.9343434343434344,
+      "grad_norm": 0.4509701728820801,
+      "learning_rate": 4.925078132986361e-06,
+      "loss": 0.0118,
+      "step": 185
+    },
+    {
+      "epoch": 0.9393939393939394,
+      "grad_norm": 0.5309934616088867,
+      "learning_rate": 4.923314016672075e-06,
+      "loss": 0.0084,
+      "step": 186
+    },
+    {
+      "epoch": 0.9444444444444444,
+      "grad_norm": 0.4668470621109009,
+      "learning_rate": 4.921529695707065e-06,
+      "loss": 0.0076,
+      "step": 187
+    },
+    {
+      "epoch": 0.9494949494949495,
+      "grad_norm": 0.5394138097763062,
+      "learning_rate": 4.919725184968307e-06,
+      "loss": 0.0129,
+      "step": 188
+    },
+    {
+      "epoch": 0.9545454545454546,
+      "grad_norm": 0.5527902841567993,
+      "learning_rate": 4.917900499501109e-06,
+      "loss": 0.0099,
+      "step": 189
+    },
+    {
+      "epoch": 0.9595959595959596,
+      "grad_norm": 0.4318401515483856,
+      "learning_rate": 4.9160556545189895e-06,
+      "loss": 0.0116,
+      "step": 190
+    },
+    {
+      "epoch": 0.9646464646464646,
+      "grad_norm": 0.5038670301437378,
+      "learning_rate": 4.91419066540355e-06,
+      "loss": 0.0089,
+      "step": 191
+    },
+    {
+      "epoch": 0.9696969696969697,
+      "grad_norm": 0.5830440521240234,
+      "learning_rate": 4.9123055477043454e-06,
+      "loss": 0.0072,
+      "step": 192
+    },
+    {
+      "epoch": 0.9747474747474747,
+      "grad_norm": 0.3858352303504944,
+      "learning_rate": 4.910400317138752e-06,
+      "loss": 0.0078,
+      "step": 193
+    },
+    {
+      "epoch": 0.9797979797979798,
+      "grad_norm": 0.7046027183532715,
+      "learning_rate": 4.908474989591846e-06,
+      "loss": 0.0111,
+      "step": 194
+    },
+    {
+      "epoch": 0.9848484848484849,
+      "grad_norm": 0.8276668787002563,
+      "learning_rate": 4.906529581116259e-06,
+      "loss": 0.011,
+      "step": 195
+    },
+    {
+      "epoch": 0.98989898989899,
+      "grad_norm": 0.8134333491325378,
+      "learning_rate": 4.904564107932048e-06,
+      "loss": 0.0073,
+      "step": 196
+    },
+    {
+      "epoch": 0.9949494949494949,
+      "grad_norm": 0.41067278385162354,
+      "learning_rate": 4.902578586426569e-06,
+      "loss": 0.0073,
+      "step": 197
+    },
+    {
+      "epoch": 1.0,
+      "grad_norm": 0.6972447633743286,
+      "learning_rate": 4.900573033154325e-06,
+      "loss": 0.0113,
+      "step": 198
+    },
+    {
+      "epoch": 1.005050505050505,
+      "grad_norm": 0.7097163200378418,
+      "learning_rate": 4.898547464836844e-06,
+      "loss": 0.0105,
+      "step": 199
+    },
+    {
+      "epoch": 1.0101010101010102,
+      "grad_norm": 0.3972426652908325,
+      "learning_rate": 4.8965018983625245e-06,
+      "loss": 0.0059,
+      "step": 200
+    },
+    {
+      "epoch": 1.0151515151515151,
+      "grad_norm": 0.32025033235549927,
+      "learning_rate": 4.8944363507865065e-06,
+      "loss": 0.0025,
+      "step": 201
+    },
+    {
+      "epoch": 1.02020202020202,
+      "grad_norm": 0.5081179141998291,
+      "learning_rate": 4.8923508393305224e-06,
+      "loss": 0.0037,
+      "step": 202
+    },
+    {
+      "epoch": 1.0252525252525253,
+      "grad_norm": 0.24843256175518036,
+      "learning_rate": 4.890245381382757e-06,
+      "loss": 0.0021,
+      "step": 203
+    },
+    {
+      "epoch": 1.0303030303030303,
+      "grad_norm": 0.2488572746515274,
+      "learning_rate": 4.888119994497701e-06,
+      "loss": 0.0032,
+      "step": 204
+    },
+    {
+      "epoch": 1.0353535353535352,
+      "grad_norm": 0.8642904758453369,
+      "learning_rate": 4.885974696396002e-06,
+      "loss": 0.0079,
+      "step": 205
+    },
+    {
+      "epoch": 1.0404040404040404,
+      "grad_norm": 0.7817063331604004,
+      "learning_rate": 4.883809504964325e-06,
+      "loss": 0.0052,
+      "step": 206
+    },
+    {
+      "epoch": 1.0454545454545454,
+      "grad_norm": 0.5490451455116272,
+      "learning_rate": 4.881624438255194e-06,
+      "loss": 0.007,
+      "step": 207
+    },
+    {
+      "epoch": 1.0505050505050506,
+      "grad_norm": 0.9993918538093567,
+      "learning_rate": 4.879419514486846e-06,
+      "loss": 0.0103,
+      "step": 208
+    },
+    {
+      "epoch": 1.0555555555555556,
+      "grad_norm": 0.6243229508399963,
+      "learning_rate": 4.8771947520430785e-06,
+      "loss": 0.002,
+      "step": 209
+    },
+    {
+      "epoch": 1.0606060606060606,
+      "grad_norm": 0.5708845853805542,
+      "learning_rate": 4.874950169473097e-06,
+      "loss": 0.0026,
+      "step": 210
+    },
+    {
+      "epoch": 1.0656565656565657,
+      "grad_norm": 0.29404523968696594,
+      "learning_rate": 4.872685785491357e-06,
+      "loss": 0.0035,
+      "step": 211
+    },
+    {
+      "epoch": 1.0707070707070707,
+      "grad_norm": 0.44978681206703186,
+      "learning_rate": 4.870401618977415e-06,
+      "loss": 0.0033,
+      "step": 212
+    },
+    {
+      "epoch": 1.0757575757575757,
+      "grad_norm": 0.34965094923973083,
+      "learning_rate": 4.868097688975763e-06,
+      "loss": 0.0053,
+      "step": 213
+    },
+    {
+      "epoch": 1.0808080808080809,
+      "grad_norm": 0.4668480455875397,
+      "learning_rate": 4.8657740146956724e-06,
+      "loss": 0.0046,
+      "step": 214
+    },
+    {
+      "epoch": 1.0858585858585859,
+      "grad_norm": 0.6791878342628479,
+      "learning_rate": 4.863430615511039e-06,
+      "loss": 0.004,
+      "step": 215
+    },
+    {
+      "epoch": 1.0909090909090908,
+      "grad_norm": 0.4435560703277588,
+      "learning_rate": 4.8610675109602135e-06,
+      "loss": 0.0045,
+      "step": 216
+    },
+    {
+      "epoch": 1.095959595959596,
+      "grad_norm": 0.8708865642547607,
+      "learning_rate": 4.858684720745843e-06,
+      "loss": 0.0065,
+      "step": 217
+    },
+    {
+      "epoch": 1.101010101010101,
+      "grad_norm": 0.6024020314216614,
+      "learning_rate": 4.856282264734708e-06,
+      "loss": 0.0079,
+      "step": 218
+    },
+    {
+      "epoch": 1.106060606060606,
+      "grad_norm": 0.9265580773353577,
+      "learning_rate": 4.8538601629575525e-06,
+      "loss": 0.0047,
+      "step": 219
+    },
+    {
+      "epoch": 1.1111111111111112,
+      "grad_norm": 0.4640383720397949,
+      "learning_rate": 4.851418435608919e-06,
+      "loss": 0.0061,
+      "step": 220
+    },
+    {
+      "epoch": 1.1161616161616161,
+      "grad_norm": 0.36199602484703064,
+      "learning_rate": 4.848957103046984e-06,
+      "loss": 0.0051,
+      "step": 221
+    },
+    {
+      "epoch": 1.121212121212121,
+      "grad_norm": 0.4664932191371918,
+      "learning_rate": 4.84647618579338e-06,
+      "loss": 0.0033,
+      "step": 222
+    },
+    {
+      "epoch": 1.1262626262626263,
+      "grad_norm": 0.23365701735019684,
+      "learning_rate": 4.843975704533031e-06,
+      "loss": 0.0031,
+      "step": 223
+    },
+    {
+      "epoch": 1.1313131313131313,
+      "grad_norm": 0.40371668338775635,
+      "learning_rate": 4.841455680113979e-06,
+      "loss": 0.0072,
+      "step": 224
+    },
+    {
+      "epoch": 1.1363636363636362,
+      "grad_norm": 0.8710446357727051,
+      "learning_rate": 4.838916133547208e-06,
+      "loss": 0.0062,
+      "step": 225
+    },
+    {
+      "epoch": 1.1414141414141414,
+      "grad_norm": 0.5429500341415405,
+      "learning_rate": 4.836357086006471e-06,
+      "loss": 0.0034,
+      "step": 226
+    },
+    {
+      "epoch": 1.1464646464646464,
+      "grad_norm": 0.7119227051734924,
+      "learning_rate": 4.8337785588281125e-06,
+      "loss": 0.0069,
+      "step": 227
+    },
+    {
+      "epoch": 1.1515151515151516,
+      "grad_norm": 0.35653722286224365,
+      "learning_rate": 4.83118057351089e-06,
+      "loss": 0.0044,
+      "step": 228
+    },
+    {
+      "epoch": 1.1565656565656566,
+      "grad_norm": 0.34784218668937683,
+      "learning_rate": 4.828563151715795e-06,
+      "loss": 0.0025,
+      "step": 229
+    },
+    {
+      "epoch": 1.1616161616161615,
+      "grad_norm": 0.6477975845336914,
+      "learning_rate": 4.825926315265874e-06,
+      "loss": 0.0059,
+      "step": 230
+    },
+    {
+      "epoch": 1.1666666666666667,
+      "grad_norm": 0.557756781578064,
+      "learning_rate": 4.823270086146045e-06,
+      "loss": 0.0048,
+      "step": 231
+    },
+    {
+      "epoch": 1.1717171717171717,
+      "grad_norm": 0.4651699960231781,
+      "learning_rate": 4.820594486502913e-06,
+      "loss": 0.0064,
+      "step": 232
+    },
+    {
+      "epoch": 1.1767676767676767,
+      "grad_norm": 0.32300034165382385,
+      "learning_rate": 4.81789953864459e-06,
+      "loss": 0.0017,
+      "step": 233
+    },
+    {
+      "epoch": 1.1818181818181819,
+      "grad_norm": 0.23726579546928406,
+      "learning_rate": 4.815185265040504e-06,
+      "loss": 0.0023,
+      "step": 234
+    },
+    {
+      "epoch": 1.1868686868686869,
+      "grad_norm": 0.6567448973655701,
+      "learning_rate": 4.812451688321213e-06,
+      "loss": 0.0028,
+      "step": 235
+    },
+    {
+      "epoch": 1.1919191919191918,
+      "grad_norm": 0.2410786747932434,
+      "learning_rate": 4.809698831278217e-06,
+      "loss": 0.0009,
+      "step": 236
+    },
+    {
+      "epoch": 1.196969696969697,
+      "grad_norm": 0.839699923992157,
+      "learning_rate": 4.80692671686377e-06,
+      "loss": 0.005,
+      "step": 237
+    },
+    {
+      "epoch": 1.202020202020202,
+      "grad_norm": 1.173795461654663,
+      "learning_rate": 4.804135368190684e-06,
+      "loss": 0.0021,
+      "step": 238
+    },
+    {
+      "epoch": 1.2070707070707072,
+      "grad_norm": 0.5123177170753479,
+      "learning_rate": 4.8013248085321394e-06,
+      "loss": 0.0043,
+      "step": 239
+    },
+    {
+      "epoch": 1.2121212121212122,
+      "grad_norm": 0.6547465920448303,
+      "learning_rate": 4.798495061321492e-06,
+      "loss": 0.0056,
+      "step": 240
+    },
+    {
+      "epoch": 1.2171717171717171,
+      "grad_norm": 1.0825695991516113,
+      "learning_rate": 4.795646150152072e-06,
+      "loss": 0.0046,
+      "step": 241
+    },
+    {
+      "epoch": 1.2222222222222223,
+      "grad_norm": 0.25643694400787354,
+      "learning_rate": 4.792778098776997e-06,
+      "loss": 0.0015,
+      "step": 242
+    },
+    {
+      "epoch": 1.2272727272727273,
+      "grad_norm": 0.35064786672592163,
+      "learning_rate": 4.789890931108963e-06,
+      "loss": 0.0018,
+      "step": 243
+    },
+    {
+      "epoch": 1.2323232323232323,
+      "grad_norm": 0.6860871315002441,
+      "learning_rate": 4.786984671220053e-06,
+      "loss": 0.0045,
+      "step": 244
+    },
+    {
+      "epoch": 1.2373737373737375,
+      "grad_norm": 0.5118893384933472,
+      "learning_rate": 4.784059343341531e-06,
+      "loss": 0.0026,
+      "step": 245
+    },
+    {
+      "epoch": 1.2424242424242424,
+      "grad_norm": 0.3446030616760254,
+      "learning_rate": 4.7811149718636475e-06,
+      "loss": 0.0026,
+      "step": 246
+    },
+    {
+      "epoch": 1.2474747474747474,
+      "grad_norm": 0.29767417907714844,
+      "learning_rate": 4.778151581335427e-06,
+      "loss": 0.0005,
+      "step": 247
+    },
+    {
+      "epoch": 1.2525252525252526,
+      "grad_norm": 0.41258764266967773,
+      "learning_rate": 4.7751691964644655e-06,
+      "loss": 0.0036,
+      "step": 248
+    },
+    {
+      "epoch": 1.2575757575757576,
+      "grad_norm": 0.1846575289964676,
+      "learning_rate": 4.772167842116732e-06,
+      "loss": 0.0014,
+      "step": 249
+    },
+    {
+      "epoch": 1.2626262626262625,
+      "grad_norm": 0.5577757358551025,
+      "learning_rate": 4.7691475433163515e-06,
+      "loss": 0.002,
+      "step": 250
+    },
+    {
+      "epoch": 1.2676767676767677,
+      "grad_norm": 2.5368878841400146,
+      "learning_rate": 4.766108325245403e-06,
+      "loss": 0.005,
+      "step": 251
+    },
+    {
+      "epoch": 1.2727272727272727,
+      "grad_norm": 0.24205638468265533,
+      "learning_rate": 4.763050213243705e-06,
+      "loss": 0.0022,
+      "step": 252
+    },
+    {
+      "epoch": 1.2777777777777777,
+      "grad_norm": 1.3182103633880615,
+      "learning_rate": 4.759973232808609e-06,
+      "loss": 0.005,
+      "step": 253
+    },
+    {
+      "epoch": 1.2828282828282829,
+      "grad_norm": 0.13516050577163696,
+      "learning_rate": 4.7568774095947804e-06,
+      "loss": 0.0007,
+      "step": 254
+    },
+    {
+      "epoch": 1.2878787878787878,
+      "grad_norm": 0.2980661392211914,
+      "learning_rate": 4.753762769413991e-06,
+      "loss": 0.0033,
+      "step": 255
+    },
+    {
+      "epoch": 1.2929292929292928,
+      "grad_norm": 0.25073671340942383,
+      "learning_rate": 4.7506293382349e-06,
+      "loss": 0.0008,
+      "step": 256
+    },
+    {
+      "epoch": 1.297979797979798,
+      "grad_norm": 0.21918445825576782,
+      "learning_rate": 4.747477142182839e-06,
+      "loss": 0.0009,
+      "step": 257
+    },
+    {
+      "epoch": 1.303030303030303,
+      "grad_norm": 0.18419255316257477,
+      "learning_rate": 4.744306207539595e-06,
+      "loss": 0.0014,
+      "step": 258
+    },
+    {
+      "epoch": 1.308080808080808,
+      "grad_norm": 0.5623617172241211,
+      "learning_rate": 4.741116560743189e-06,
+      "loss": 0.0036,
+      "step": 259
+    },
+    {
+      "epoch": 1.3131313131313131,
+      "grad_norm": 0.675177276134491,
+      "learning_rate": 4.737908228387656e-06,
+      "loss": 0.0058,
+      "step": 260
+    },
+    {
+      "epoch": 1.3181818181818181,
+      "grad_norm": 0.6051934361457825,
+      "learning_rate": 4.734681237222825e-06,
+      "loss": 0.0008,
+      "step": 261
+    },
+    {
+      "epoch": 1.3232323232323233,
+      "grad_norm": 0.4990449845790863,
+      "learning_rate": 4.731435614154094e-06,
+      "loss": 0.0021,
+      "step": 262
+    },
+    {
+      "epoch": 1.3282828282828283,
+      "grad_norm": 0.42459845542907715,
+      "learning_rate": 4.7281713862422074e-06,
+      "loss": 0.0025,
+      "step": 263
+    },
+    {
+      "epoch": 1.3333333333333333,
+      "grad_norm": 0.5847693085670471,
+      "learning_rate": 4.72488858070303e-06,
+      "loss": 0.0045,
+      "step": 264
+    },
+    {
+      "epoch": 1.3383838383838385,
+      "grad_norm": 0.512243390083313,
+      "learning_rate": 4.721587224907317e-06,
+      "loss": 0.0035,
+      "step": 265
+    },
+    {
+      "epoch": 1.3434343434343434,
+      "grad_norm": 0.4338761270046234,
+      "learning_rate": 4.718267346380492e-06,
+      "loss": 0.0022,
+      "step": 266
+    },
+    {
+      "epoch": 1.3484848484848486,
+      "grad_norm": 0.449686735868454,
+      "learning_rate": 4.714928972802412e-06,
+      "loss": 0.0035,
+      "step": 267
+    },
+    {
+      "epoch": 1.3535353535353536,
+      "grad_norm": 0.34395352005958557,
+      "learning_rate": 4.711572132007139e-06,
+      "loss": 0.0011,
+      "step": 268
+    },
+    {
+      "epoch": 1.3585858585858586,
+      "grad_norm": 0.5509817004203796,
+      "learning_rate": 4.708196851982706e-06,
+      "loss": 0.0055,
+      "step": 269
+    },
+    {
+      "epoch": 1.3636363636363638,
+      "grad_norm": 0.7635436058044434,
+      "learning_rate": 4.704803160870888e-06,
+      "loss": 0.0013,
+      "step": 270
+    },
+    {
+      "epoch": 1.3686868686868687,
+      "grad_norm": 0.6281613111495972,
+      "learning_rate": 4.701391086966962e-06,
+      "loss": 0.0024,
+      "step": 271
+    },
+    {
+      "epoch": 1.3737373737373737,
+      "grad_norm": 0.27791550755500793,
+      "learning_rate": 4.697960658719475e-06,
+      "loss": 0.0011,
+      "step": 272
+    },
+    {
+      "epoch": 1.378787878787879,
+      "grad_norm": 0.18364505469799042,
+      "learning_rate": 4.694511904730004e-06,
+      "loss": 0.0007,
+      "step": 273
+    },
+    {
+      "epoch": 1.3838383838383839,
+      "grad_norm": 0.3883489966392517,
+      "learning_rate": 4.69104485375292e-06,
+      "loss": 0.003,
+      "step": 274
+    },
+    {
+      "epoch": 1.3888888888888888,
+      "grad_norm": 0.36030399799346924,
+      "learning_rate": 4.687559534695149e-06,
+      "loss": 0.0026,
+      "step": 275
+    },
+    {
+      "epoch": 1.393939393939394,
+      "grad_norm": 0.6687548160552979,
+      "learning_rate": 4.684055976615924e-06,
+      "loss": 0.0036,
+      "step": 276
+    },
+    {
+      "epoch": 1.398989898989899,
+      "grad_norm": 0.3158160150051117,
+      "learning_rate": 4.6805342087265534e-06,
+      "loss": 0.0033,
+      "step": 277
+    },
+    {
+      "epoch": 1.404040404040404,
+      "grad_norm": 0.30688995122909546,
+      "learning_rate": 4.676994260390168e-06,
+      "loss": 0.0038,
+      "step": 278
+    },
+    {
+      "epoch": 1.4090909090909092,
+      "grad_norm": 3.210895538330078,
+      "learning_rate": 4.6734361611214825e-06,
+      "loss": 0.002,
+      "step": 279
+    },
+    {
+      "epoch": 1.4141414141414141,
+      "grad_norm": 0.2669866383075714,
+      "learning_rate": 4.6698599405865465e-06,
+      "loss": 0.0019,
+      "step": 280
+    },
+    {
+      "epoch": 1.4191919191919191,
+      "grad_norm": 0.15304385125637054,
+      "learning_rate": 4.666265628602496e-06,
+      "loss": 0.0008,
+      "step": 281
+    },
+    {
+      "epoch": 1.4242424242424243,
+      "grad_norm": 0.6389310359954834,
+      "learning_rate": 4.662653255137308e-06,
+      "loss": 0.0008,
+      "step": 282
+    },
+    {
+      "epoch": 1.4292929292929293,
+      "grad_norm": 0.5978128910064697,
+      "learning_rate": 4.65902285030955e-06,
+      "loss": 0.0018,
+      "step": 283
+    },
+    {
+      "epoch": 1.4343434343434343,
+      "grad_norm": 0.9742176532745361,
+      "learning_rate": 4.655374444388127e-06,
+      "loss": 0.0053,
+      "step": 284
+    },
+    {
+      "epoch": 1.4393939393939394,
+      "grad_norm": 0.6034579277038574,
+      "learning_rate": 4.651708067792029e-06,
+      "loss": 0.0024,
+      "step": 285
+    },
+    {
+      "epoch": 1.4444444444444444,
+      "grad_norm": 0.50716233253479,
+      "learning_rate": 4.648023751090079e-06,
+      "loss": 0.0023,
+      "step": 286
+    },
+    {
+      "epoch": 1.4494949494949494,
+      "grad_norm": 0.2467270791530609,
+      "learning_rate": 4.644321525000681e-06,
+      "loss": 0.001,
+      "step": 287
+    },
+    {
+      "epoch": 1.4545454545454546,
+      "grad_norm": 0.2956307530403137,
+      "learning_rate": 4.640601420391554e-06,
+      "loss": 0.0025,
+      "step": 288
+    },
+    {
+      "epoch": 1.4595959595959596,
+      "grad_norm": 0.25785166025161743,
+      "learning_rate": 4.6368634682794875e-06,
+      "loss": 0.0025,
+      "step": 289
+    },
+    {
+      "epoch": 1.4646464646464645,
+      "grad_norm": 0.5894668698310852,
+      "learning_rate": 4.633107699830073e-06,
+      "loss": 0.0025,
+      "step": 290
+    },
+    {
+      "epoch": 1.4696969696969697,
+      "grad_norm": 0.31446027755737305,
+      "learning_rate": 4.629334146357448e-06,
+      "loss": 0.0021,
+      "step": 291
+    },
+    {
+      "epoch": 1.4747474747474747,
+      "grad_norm": 1.2636518478393555,
+      "learning_rate": 4.625542839324036e-06,
+      "loss": 0.0014,
+      "step": 292
+    },
+    {
+      "epoch": 1.4797979797979797,
+      "grad_norm": 0.38668882846832275,
+      "learning_rate": 4.621733810340282e-06,
+      "loss": 0.0022,
+      "step": 293
+    },
+    {
+      "epoch": 1.4848484848484849,
+      "grad_norm": 0.32553184032440186,
+      "learning_rate": 4.617907091164389e-06,
+      "loss": 0.0037,
+      "step": 294
+    },
+    {
+      "epoch": 1.4898989898989898,
+      "grad_norm": 0.3489368259906769,
+      "learning_rate": 4.614062713702055e-06,
+      "loss": 0.0021,
+      "step": 295
+    },
+    {
+      "epoch": 1.494949494949495,
+      "grad_norm": 0.5913827419281006,
+      "learning_rate": 4.610200710006206e-06,
+      "loss": 0.0026,
+      "step": 296
+    },
+    {
+      "epoch": 1.5,
+      "grad_norm": 0.2569572925567627,
+      "learning_rate": 4.606321112276728e-06,
+      "loss": 0.0007,
+      "step": 297
+    },
+    {
+      "epoch": 1.5050505050505052,
+      "grad_norm": 0.2640990912914276,
+      "learning_rate": 4.602423952860199e-06,
+      "loss": 0.0011,
+      "step": 298
+    },
+    {
+      "epoch": 1.51010101010101,
+      "grad_norm": 0.13946537673473358,
+      "learning_rate": 4.598509264249621e-06,
+      "loss": 0.0008,
+      "step": 299
+    },
+    {
+      "epoch": 1.5151515151515151,
+      "grad_norm": 0.3428565263748169,
+      "learning_rate": 4.594577079084146e-06,
+      "loss": 0.0027,
+      "step": 300
+    },
+    {
+      "epoch": 1.5202020202020203,
+      "grad_norm": 0.10775818675756454,
+      "learning_rate": 4.590627430148806e-06,
+      "loss": 0.0005,
+      "step": 301
+    },
+    {
+      "epoch": 1.5252525252525253,
+      "grad_norm": 0.10182857513427734,
+      "learning_rate": 4.58666035037424e-06,
+      "loss": 0.0005,
+      "step": 302
+    },
+    {
+      "epoch": 1.5303030303030303,
+      "grad_norm": 0.8627449870109558,
+      "learning_rate": 4.582675872836417e-06,
+      "loss": 0.0034,
+      "step": 303
+    },
+    {
+      "epoch": 1.5353535353535355,
+      "grad_norm": 0.1706748604774475,
+      "learning_rate": 4.578674030756364e-06,
+      "loss": 0.001,
+      "step": 304
+    },
+    {
+      "epoch": 1.5404040404040404,
+      "grad_norm": 0.22486886382102966,
+      "learning_rate": 4.574654857499883e-06,
+      "loss": 0.0006,
+      "step": 305
+    },
+    {
+      "epoch": 1.5454545454545454,
+      "grad_norm": 0.132318377494812,
+      "learning_rate": 4.57061838657728e-06,
+      "loss": 0.0002,
+      "step": 306
+    },
+    {
+      "epoch": 1.5505050505050506,
+      "grad_norm": 0.11632445454597473,
+      "learning_rate": 4.566564651643083e-06,
+      "loss": 0.0008,
+      "step": 307
+    },
+    {
+      "epoch": 1.5555555555555556,
+      "grad_norm": 0.5458851456642151,
+      "learning_rate": 4.562493686495756e-06,
+      "loss": 0.0017,
+      "step": 308
+    },
+    {
+      "epoch": 1.5606060606060606,
+      "grad_norm": 0.3893134593963623,
+      "learning_rate": 4.558405525077426e-06,
+      "loss": 0.0008,
+      "step": 309
+    },
+    {
+      "epoch": 1.5656565656565657,
+      "grad_norm": 0.14303813874721527,
+      "learning_rate": 4.5543002014735955e-06,
+      "loss": 0.0008,
+      "step": 310
+    },
+    {
+      "epoch": 1.5707070707070707,
+      "grad_norm": 0.06348168104887009,
+      "learning_rate": 4.550177749912857e-06,
+      "loss": 0.0001,
+      "step": 311
+    },
+    {
+      "epoch": 1.5757575757575757,
+      "grad_norm": 0.024890398606657982,
+      "learning_rate": 4.546038204766609e-06,
+      "loss": 0.0001,
+      "step": 312
+    },
+    {
+      "epoch": 1.5808080808080809,
+      "grad_norm": 0.2720322012901306,
+      "learning_rate": 4.541881600548771e-06,
+      "loss": 0.0007,
+      "step": 313
+    },
+    {
+      "epoch": 1.5858585858585859,
+      "grad_norm": 2.1944384574890137,
+      "learning_rate": 4.537707971915495e-06,
+      "loss": 0.0046,
+      "step": 314
+    },
+    {
+      "epoch": 1.5909090909090908,
+      "grad_norm": 0.7526996731758118,
+      "learning_rate": 4.533517353664874e-06,
+      "loss": 0.0035,
+      "step": 315
+    },
+    {
+      "epoch": 1.595959595959596,
+      "grad_norm": 0.5112221240997314,
+      "learning_rate": 4.529309780736654e-06,
+      "loss": 0.0028,
+      "step": 316
+    },
+    {
+      "epoch": 1.601010101010101,
+      "grad_norm": 0.514263391494751,
+      "learning_rate": 4.525085288211943e-06,
+      "loss": 0.0035,
+      "step": 317
+    },
+    {
+      "epoch": 1.606060606060606,
+      "grad_norm": 0.22527411580085754,
+      "learning_rate": 4.520843911312922e-06,
+      "loss": 0.0005,
+      "step": 318
+    },
+    {
+      "epoch": 1.6111111111111112,
+      "grad_norm": 0.3148433566093445,
+      "learning_rate": 4.5165856854025394e-06,
+      "loss": 0.0015,
+      "step": 319
+    },
+    {
+      "epoch": 1.6161616161616161,
+      "grad_norm": 0.186680406332016,
+      "learning_rate": 4.512310645984231e-06,
+      "loss": 0.0002,
+      "step": 320
+    },
+    {
+      "epoch": 1.621212121212121,
+      "grad_norm": 0.30500105023384094,
+      "learning_rate": 4.508018828701613e-06,
+      "loss": 0.003,
+      "step": 321
+    },
+    {
+      "epoch": 1.6262626262626263,
+      "grad_norm": 0.11456169933080673,
+      "learning_rate": 4.503710269338191e-06,
+      "loss": 0.0008,
+      "step": 322
+    },
+    {
+      "epoch": 1.6313131313131313,
+      "grad_norm": 0.1492011696100235,
+      "learning_rate": 4.4993850038170586e-06,
+      "loss": 0.0009,
+      "step": 323
+    },
+    {
+      "epoch": 1.6363636363636362,
+      "grad_norm": 0.059317488223314285,
+      "learning_rate": 4.4950430682005995e-06,
+      "loss": 0.0003,
+      "step": 324
+    },
+    {
+      "epoch": 1.6414141414141414,
+      "grad_norm": 0.12276207655668259,
+      "learning_rate": 4.490684498690186e-06,
+      "loss": 0.0005,
+      "step": 325
+    },
+    {
+      "epoch": 1.6464646464646466,
+      "grad_norm": 0.15645022690296173,
+      "learning_rate": 4.486309331625877e-06,
+      "loss": 0.0004,
+      "step": 326
+    },
+    {
+      "epoch": 1.6515151515151514,
+      "grad_norm": 0.2627776861190796,
+      "learning_rate": 4.4819176034861146e-06,
+      "loss": 0.0034,
+      "step": 327
+    },
+    {
+      "epoch": 1.6565656565656566,
+      "grad_norm": 0.23384395241737366,
+      "learning_rate": 4.477509350887424e-06,
+      "loss": 0.0007,
+      "step": 328
+    },
+    {
+      "epoch": 1.6616161616161618,
+      "grad_norm": 0.16817571222782135,
+      "learning_rate": 4.4730846105841e-06,
+      "loss": 0.0004,
+      "step": 329
+    },
+    {
+      "epoch": 1.6666666666666665,
+      "grad_norm": 0.291900634765625,
+      "learning_rate": 4.468643419467909e-06,
+      "loss": 0.0017,
+      "step": 330
+    },
+    {
+      "epoch": 1.6717171717171717,
+      "grad_norm": 0.34208860993385315,
+      "learning_rate": 4.464185814567779e-06,
+      "loss": 0.0013,
+      "step": 331
+    },
+    {
+      "epoch": 1.676767676767677,
+      "grad_norm": 2.485881805419922,
+      "learning_rate": 4.459711833049485e-06,
+      "loss": 0.002,
+      "step": 332
+    },
+    {
+      "epoch": 1.6818181818181817,
+      "grad_norm": 0.22386501729488373,
+      "learning_rate": 4.455221512215347e-06,
+      "loss": 0.0014,
+      "step": 333
+    },
+    {
+      "epoch": 1.6868686868686869,
+      "grad_norm": 0.6212871074676514,
+      "learning_rate": 4.4507148895039165e-06,
+      "loss": 0.0043,
+      "step": 334
+    },
+    {
+      "epoch": 1.691919191919192,
+      "grad_norm": 0.5749779343605042,
+      "learning_rate": 4.4461920024896624e-06,
+      "loss": 0.0026,
+      "step": 335
+    },
+    {
+      "epoch": 1.696969696969697,
+      "grad_norm": 0.05218998342752457,
+      "learning_rate": 4.4416528888826595e-06,
+      "loss": 0.0003,
+      "step": 336
+    },
+    {
+      "epoch": 1.702020202020202,
+      "grad_norm": 0.18819659948349,
+      "learning_rate": 4.437097586528271e-06,
+      "loss": 0.0003,
+      "step": 337
+    },
+    {
+      "epoch": 1.7070707070707072,
+      "grad_norm": 0.3252980709075928,
+      "learning_rate": 4.432526133406843e-06,
+      "loss": 0.0006,
+      "step": 338
+    },
+    {
+      "epoch": 1.7121212121212122,
+      "grad_norm": 0.05513511970639229,
+      "learning_rate": 4.427938567633371e-06,
+      "loss": 0.0002,
+      "step": 339
+    },
+    {
+      "epoch": 1.7171717171717171,
+      "grad_norm": 0.22534117102622986,
+      "learning_rate": 4.423334927457198e-06,
+      "loss": 0.0004,
+      "step": 340
+    },
+    {
+      "epoch": 1.7222222222222223,
+      "grad_norm": 0.5896280407905579,
+      "learning_rate": 4.418715251261686e-06,
+      "loss": 0.0044,
+      "step": 341
+    },
+    {
+      "epoch": 1.7272727272727273,
+      "grad_norm": 0.29156655073165894,
+      "learning_rate": 4.414079577563901e-06,
+      "loss": 0.0004,
+      "step": 342
+    },
+    {
+      "epoch": 1.7323232323232323,
+      "grad_norm": 0.40658339858055115,
+      "learning_rate": 4.409427945014289e-06,
+      "loss": 0.0013,
+      "step": 343
+    },
+    {
+      "epoch": 1.7373737373737375,
+      "grad_norm": 0.11740848422050476,
+      "learning_rate": 4.404760392396355e-06,
+      "loss": 0.0002,
+      "step": 344
+    },
+    {
+      "epoch": 1.7424242424242424,
+      "grad_norm": 0.32155877351760864,
+      "learning_rate": 4.4000769586263394e-06,
+      "loss": 0.0005,
+      "step": 345
+    },
+    {
+      "epoch": 1.7474747474747474,
+      "grad_norm": 0.32882601022720337,
+      "learning_rate": 4.3953776827528925e-06,
+      "loss": 0.0009,
+      "step": 346
+    },
+    {
+      "epoch": 1.7525252525252526,
+      "grad_norm": 0.02031019888818264,
+      "learning_rate": 4.390662603956751e-06,
+      "loss": 0.0001,
+      "step": 347
+    },
+    {
+      "epoch": 1.7575757575757576,
+      "grad_norm": 0.030053604394197464,
+      "learning_rate": 4.385931761550411e-06,
+      "loss": 0.0001,
+      "step": 348
+    },
+    {
+      "epoch": 1.7626262626262625,
+      "grad_norm": 0.04812119901180267,
+      "learning_rate": 4.381185194977796e-06,
+      "loss": 0.0001,
+      "step": 349
+    },
+    {
+      "epoch": 1.7676767676767677,
+      "grad_norm": 0.7133746147155762,
+      "learning_rate": 4.376422943813936e-06,
+      "loss": 0.0013,
+      "step": 350
+    },
+    {
+      "epoch": 1.7727272727272727,
+      "grad_norm": 0.13918650150299072,
+      "learning_rate": 4.37164504776463e-06,
+      "loss": 0.0004,
+      "step": 351
+    },
+    {
+      "epoch": 1.7777777777777777,
+      "grad_norm": 0.029156319797039032,
+      "learning_rate": 4.366851546666118e-06,
+      "loss": 0.0001,
+      "step": 352
+    },
+    {
+      "epoch": 1.7828282828282829,
+      "grad_norm": 0.23016220331192017,
+      "learning_rate": 4.362042480484753e-06,
+      "loss": 0.0014,
+      "step": 353
+    },
+    {
+      "epoch": 1.7878787878787878,
+      "grad_norm": 0.015141244977712631,
+      "learning_rate": 4.357217889316657e-06,
+      "loss": 0.0001,
+      "step": 354
+    },
+    {
+      "epoch": 1.7929292929292928,
+      "grad_norm": 0.04107781499624252,
+      "learning_rate": 4.352377813387398e-06,
+      "loss": 0.0002,
+      "step": 355
+    },
+    {
+      "epoch": 1.797979797979798,
+      "grad_norm": 0.12357518076896667,
+      "learning_rate": 4.3475222930516484e-06,
+      "loss": 0.0002,
+      "step": 356
+    },
+    {
+      "epoch": 1.803030303030303,
+      "grad_norm": 0.1132410392165184,
+      "learning_rate": 4.34265136879285e-06,
+      "loss": 0.0021,
+      "step": 357
+    },
+    {
+      "epoch": 1.808080808080808,
+      "grad_norm": 0.2640242278575897,
+      "learning_rate": 4.3377650812228765e-06,
+      "loss": 0.0028,
+      "step": 358
+    },
+    {
+      "epoch": 1.8131313131313131,
+      "grad_norm": 0.3503035008907318,
+      "learning_rate": 4.332863471081695e-06,
+      "loss": 0.0006,
+      "step": 359
+    },
+    {
+      "epoch": 1.8181818181818183,
+      "grad_norm": 0.20633821189403534,
+      "learning_rate": 4.327946579237028e-06,
+      "loss": 0.0005,
+      "step": 360
+    },
+    {
+      "epoch": 1.823232323232323,
+      "grad_norm": 0.7339881658554077,
+      "learning_rate": 4.323014446684007e-06,
+      "loss": 0.0013,
+      "step": 361
+    },
+    {
+      "epoch": 1.8282828282828283,
+      "grad_norm": 0.32832375168800354,
+      "learning_rate": 4.318067114544838e-06,
+      "loss": 0.0017,
+      "step": 362
+    },
+    {
+      "epoch": 1.8333333333333335,
+      "grad_norm": 0.2648046314716339,
+      "learning_rate": 4.313104624068456e-06,
+      "loss": 0.0011,
+      "step": 363
+    },
+    {
+      "epoch": 1.8383838383838382,
+      "grad_norm": 0.14635734260082245,
+      "learning_rate": 4.308127016630176e-06,
+      "loss": 0.0008,
+      "step": 364
+    },
+    {
+      "epoch": 1.8434343434343434,
+      "grad_norm": 0.7941786646842957,
+      "learning_rate": 4.303134333731354e-06,
+      "loss": 0.0027,
+      "step": 365
+    },
+    {
+      "epoch": 1.8484848484848486,
+      "grad_norm": 0.02228306420147419,
+      "learning_rate": 4.2981266169990436e-06,
+      "loss": 0.0001,
+      "step": 366
+    },
+    {
+      "epoch": 1.8535353535353534,
+      "grad_norm": 0.03166576102375984,
+      "learning_rate": 4.293103908185638e-06,
+      "loss": 0.0001,
+      "step": 367
+    },
+    {
+      "epoch": 1.8585858585858586,
+      "grad_norm": 0.055547453463077545,
+      "learning_rate": 4.2880662491685345e-06,
+      "loss": 0.0002,
+      "step": 368
+    },
+    {
+      "epoch": 1.8636363636363638,
+      "grad_norm": 0.2905639708042145,
+      "learning_rate": 4.283013681949774e-06,
+      "loss": 0.0002,
+      "step": 369
+    },
+    {
+      "epoch": 1.8686868686868687,
+      "grad_norm": 0.1792776733636856,
+      "learning_rate": 4.277946248655701e-06,
+      "loss": 0.0005,
+      "step": 370
+    },
+    {
+      "epoch": 1.8737373737373737,
+      "grad_norm": 0.1844152808189392,
+      "learning_rate": 4.272863991536604e-06,
+      "loss": 0.0004,
+      "step": 371
+    },
+    {
+      "epoch": 1.878787878787879,
+      "grad_norm": 0.5641722679138184,
+      "learning_rate": 4.267766952966369e-06,
+      "loss": 0.0012,
+      "step": 372
+    },
+    {
+      "epoch": 1.8838383838383839,
+      "grad_norm": 0.7322543859481812,
+      "learning_rate": 4.262655175442123e-06,
+      "loss": 0.0019,
+      "step": 373
+    },
+    {
+      "epoch": 1.8888888888888888,
+      "grad_norm": 0.1359657496213913,
+      "learning_rate": 4.257528701583882e-06,
+      "loss": 0.0004,
+      "step": 374
+    },
+    {
+      "epoch": 1.893939393939394,
+      "grad_norm": 0.06701353937387466,
+      "learning_rate": 4.2523875741341905e-06,
+      "loss": 0.0002,
+      "step": 375
+    },
+    {
+      "epoch": 1.898989898989899,
+      "grad_norm": 0.07855173945426941,
+      "learning_rate": 4.247231835957773e-06,
+      "loss": 0.0001,
+      "step": 376
+    },
+    {
+      "epoch": 1.904040404040404,
+      "grad_norm": 0.2692381739616394,
+      "learning_rate": 4.242061530041173e-06,
+      "loss": 0.0017,
+      "step": 377
+    },
+    {
+      "epoch": 1.9090909090909092,
+      "grad_norm": 0.059216227382421494,
+      "learning_rate": 4.236876699492391e-06,
+      "loss": 0.0001,
+      "step": 378
+    },
+    {
+      "epoch": 1.9141414141414141,
+      "grad_norm": 0.14110217988491058,
+      "learning_rate": 4.231677387540528e-06,
+      "loss": 0.0002,
+      "step": 379
+    },
+    {
+      "epoch": 1.9191919191919191,
+      "grad_norm": 0.02432100474834442,
+      "learning_rate": 4.226463637535429e-06,
+      "loss": 0.0001,
+      "step": 380
+    },
+    {
+      "epoch": 1.9242424242424243,
+      "grad_norm": 2.4970834255218506,
+      "learning_rate": 4.221235492947316e-06,
+      "loss": 0.0011,
+      "step": 381
+    },
+    {
+      "epoch": 1.9292929292929293,
+      "grad_norm": 0.029205068945884705,
+      "learning_rate": 4.215992997366425e-06,
+      "loss": 0.0001,
+      "step": 382
+    },
+    {
+      "epoch": 1.9343434343434343,
+      "grad_norm": 0.1291147619485855,
+      "learning_rate": 4.21073619450265e-06,
+      "loss": 0.0002,
+      "step": 383
+    },
+    {
+      "epoch": 1.9393939393939394,
+      "grad_norm": 0.24579289555549622,
+      "learning_rate": 4.2054651281851685e-06,
+      "loss": 0.0028,
+      "step": 384
+    },
+    {
+      "epoch": 1.9444444444444444,
+      "grad_norm": 0.10453365743160248,
+      "learning_rate": 4.200179842362084e-06,
+      "loss": 0.0003,
+      "step": 385
+    },
+    {
+      "epoch": 1.9494949494949494,
+      "grad_norm": 0.1425495594739914,
+      "learning_rate": 4.1948803811000585e-06,
+      "loss": 0.0007,
+      "step": 386
+    },
+    {
+      "epoch": 1.9545454545454546,
+      "grad_norm": 0.0607614740729332,
+      "learning_rate": 4.189566788583938e-06,
+      "loss": 0.0002,
+      "step": 387
+    },
+    {
+      "epoch": 1.9595959595959596,
+      "grad_norm": 0.1010780856013298,
+      "learning_rate": 4.184239109116393e-06,
+      "loss": 0.0002,
+      "step": 388
+    },
+    {
+      "epoch": 1.9646464646464645,
+      "grad_norm": 0.09742361307144165,
+      "learning_rate": 4.178897387117547e-06,
+      "loss": 0.0001,
+      "step": 389
+    },
+    {
+      "epoch": 1.9696969696969697,
+      "grad_norm": 0.13673080503940582,
+      "learning_rate": 4.173541667124599e-06,
+      "loss": 0.0002,
+      "step": 390
+    },
+    {
+      "epoch": 1.9747474747474747,
+      "grad_norm": 0.28466007113456726,
+      "learning_rate": 4.168171993791465e-06,
+      "loss": 0.0004,
+      "step": 391
+    },
+    {
+      "epoch": 1.9797979797979797,
+      "grad_norm": 0.07821296155452728,
+      "learning_rate": 4.1627884118883925e-06,
+      "loss": 0.0001,
+      "step": 392
+    },
+    {
+      "epoch": 1.9848484848484849,
+      "grad_norm": 0.42731398344039917,
+      "learning_rate": 4.157390966301597e-06,
+      "loss": 0.0014,
+      "step": 393
+    },
+    {
+      "epoch": 1.98989898989899,
+      "grad_norm": 0.19062280654907227,
+      "learning_rate": 4.1519797020328815e-06,
+      "loss": 0.002,
+      "step": 394
+    },
+    {
+      "epoch": 1.9949494949494948,
+      "grad_norm": 0.033713385462760925,
+      "learning_rate": 4.146554664199265e-06,
+      "loss": 0.0001,
+      "step": 395
+    },
+    {
+      "epoch": 2.0,
+      "grad_norm": 0.3922223746776581,
+      "learning_rate": 4.141115898032607e-06,
+      "loss": 0.002,
+      "step": 396
+    },
+    {
+      "epoch": 2.005050505050505,
+      "grad_norm": 0.09248466789722443,
+      "learning_rate": 4.135663448879225e-06,
+      "loss": 0.0002,
+      "step": 397
+    },
+    {
+      "epoch": 2.01010101010101,
+      "grad_norm": 0.013212003745138645,
+      "learning_rate": 4.130197362199521e-06,
+      "loss": 0.0001,
+      "step": 398
+    },
+    {
+      "epoch": 2.015151515151515,
+      "grad_norm": 0.02999984472990036,
+      "learning_rate": 4.124717683567605e-06,
+      "loss": 0.0001,
+      "step": 399
+    },
+    {
+      "epoch": 2.0202020202020203,
+      "grad_norm": 0.05465150251984596,
+      "learning_rate": 4.119224458670905e-06,
+      "loss": 0.0001,
+      "step": 400
+    },
+    {
+      "epoch": 2.025252525252525,
+      "grad_norm": 0.01147720031440258,
+      "learning_rate": 4.113717733309798e-06,
+      "loss": 0.0001,
+      "step": 401
+    },
+    {
+      "epoch": 2.0303030303030303,
+      "grad_norm": 0.3443642258644104,
+      "learning_rate": 4.1081975533972185e-06,
+      "loss": 0.0002,
+      "step": 402
+    },
+    {
+      "epoch": 2.0353535353535355,
+      "grad_norm": 0.012025265023112297,
+      "learning_rate": 4.102663964958285e-06,
+      "loss": 0.0,
+      "step": 403
+    },
+    {
+      "epoch": 2.04040404040404,
+      "grad_norm": 0.007129222620278597,
+      "learning_rate": 4.097117014129903e-06,
+      "loss": 0.0,
+      "step": 404
+    },
+    {
+      "epoch": 2.0454545454545454,
+      "grad_norm": 0.011770972982048988,
+      "learning_rate": 4.091556747160398e-06,
+      "loss": 0.0001,
+      "step": 405
+    },
+    {
+      "epoch": 2.0505050505050506,
+      "grad_norm": 0.01947842165827751,
+      "learning_rate": 4.085983210409114e-06,
+      "loss": 0.0001,
+      "step": 406
+    },
+    {
+      "epoch": 2.0555555555555554,
+      "grad_norm": 0.10189538449048996,
+      "learning_rate": 4.0803964503460345e-06,
+      "loss": 0.001,
+      "step": 407
+    },
+    {
+      "epoch": 2.0606060606060606,
+      "grad_norm": 0.009692806750535965,
+      "learning_rate": 4.074796513551395e-06,
+      "loss": 0.0,
+      "step": 408
+    },
+    {
+      "epoch": 2.0656565656565657,
+      "grad_norm": 0.3666267395019531,
+      "learning_rate": 4.069183446715292e-06,
+      "loss": 0.0022,
+      "step": 409
+    },
+    {
+      "epoch": 2.0707070707070705,
+      "grad_norm": 0.05267680808901787,
+      "learning_rate": 4.063557296637295e-06,
+      "loss": 0.0001,
+      "step": 410
+    },
+    {
+      "epoch": 2.0757575757575757,
+      "grad_norm": 0.01215702947229147,
+      "learning_rate": 4.057918110226059e-06,
+      "loss": 0.0,
+      "step": 411
+    },
+    {
+      "epoch": 2.080808080808081,
+      "grad_norm": 0.013252578675746918,
+      "learning_rate": 4.052265934498929e-06,
+      "loss": 0.0001,
+      "step": 412
+    },
+    {
+      "epoch": 2.0858585858585856,
+      "grad_norm": 0.007566288113594055,
+      "learning_rate": 4.04660081658155e-06,
+      "loss": 0.0,
+      "step": 413
+    },
+    {
+      "epoch": 2.090909090909091,
+      "grad_norm": 0.00914331991225481,
+      "learning_rate": 4.040922803707474e-06,
+      "loss": 0.0,
+      "step": 414
+    },
+    {
+      "epoch": 2.095959595959596,
+      "grad_norm": 0.007612581364810467,
+      "learning_rate": 4.035231943217765e-06,
+      "loss": 0.0,
+      "step": 415
+    },
+    {
+      "epoch": 2.101010101010101,
+      "grad_norm": 0.06358092278242111,
+      "learning_rate": 4.029528282560609e-06,
+      "loss": 0.0001,
+      "step": 416
+    },
+    {
+      "epoch": 2.106060606060606,
+      "grad_norm": 0.015951288864016533,
+      "learning_rate": 4.023811869290911e-06,
+      "loss": 0.0001,
+      "step": 417
+    },
+    {
+      "epoch": 2.111111111111111,
+      "grad_norm": 0.00994573812931776,
+      "learning_rate": 4.018082751069904e-06,
+      "loss": 0.0,
+      "step": 418
+    },
+    {
+      "epoch": 2.1161616161616164,
+      "grad_norm": 0.018615180626511574,
+      "learning_rate": 4.01234097566475e-06,
+      "loss": 0.0001,
+      "step": 419
+    },
+    {
+      "epoch": 2.121212121212121,
+      "grad_norm": 0.007956521585583687,
+      "learning_rate": 4.006586590948141e-06,
+      "loss": 0.0,
+      "step": 420
+    },
+    {
+      "epoch": 2.1262626262626263,
+      "grad_norm": 0.02259342558681965,
+      "learning_rate": 4.000819644897902e-06,
+      "loss": 0.0001,
+      "step": 421
+    },
+    {
+      "epoch": 2.1313131313131315,
+      "grad_norm": 0.011591397225856781,
+      "learning_rate": 3.995040185596588e-06,
+      "loss": 0.0,
+      "step": 422
+    },
+    {
+      "epoch": 2.1363636363636362,
+      "grad_norm": 0.007842698134481907,
+      "learning_rate": 3.989248261231084e-06,
+      "loss": 0.0,
+      "step": 423
+    },
+    {
+      "epoch": 2.1414141414141414,
+      "grad_norm": 0.07917871326208115,
+      "learning_rate": 3.983443920092206e-06,
+      "loss": 0.0,
+      "step": 424
+    },
+    {
+      "epoch": 2.1464646464646466,
+      "grad_norm": 0.0059897336177527905,
+      "learning_rate": 3.977627210574295e-06,
+      "loss": 0.0,
+      "step": 425
+    },
+    {
+      "epoch": 2.1515151515151514,
+      "grad_norm": 0.004839922301471233,
+      "learning_rate": 3.971798181174816e-06,
+      "loss": 0.0,
+      "step": 426
+    },
+    {
+      "epoch": 2.1565656565656566,
+      "grad_norm": 0.0059188189916312695,
+      "learning_rate": 3.96595688049395e-06,
+      "loss": 0.0,
+      "step": 427
+    },
+    {
+      "epoch": 2.1616161616161618,
+      "grad_norm": 0.010813100263476372,
+      "learning_rate": 3.960103357234192e-06,
+      "loss": 0.0,
+      "step": 428
+    },
+    {
+      "epoch": 2.1666666666666665,
+      "grad_norm": 0.008660750463604927,
+      "learning_rate": 3.954237660199946e-06,
+      "loss": 0.0,
+      "step": 429
+    },
+    {
+      "epoch": 2.1717171717171717,
+      "grad_norm": 0.009066971018910408,
+      "learning_rate": 3.948359838297115e-06,
+      "loss": 0.0,
+      "step": 430
+    },
+    {
+      "epoch": 2.176767676767677,
+      "grad_norm": 0.11104936897754669,
+      "learning_rate": 3.942469940532694e-06,
+      "loss": 0.0006,
+      "step": 431
+    },
+    {
+      "epoch": 2.1818181818181817,
+      "grad_norm": 0.12647657096385956,
+      "learning_rate": 3.9365680160143595e-06,
+      "loss": 0.0006,
+      "step": 432
+    },
+    {
+      "epoch": 2.186868686868687,
+      "grad_norm": 0.011482702568173409,
+      "learning_rate": 3.93065411395007e-06,
+      "loss": 0.0,
+      "step": 433
+    },
+    {
+      "epoch": 2.191919191919192,
+      "grad_norm": 0.006080338265746832,
+      "learning_rate": 3.924728283647638e-06,
+      "loss": 0.0,
+      "step": 434
+    },
+    {
+      "epoch": 2.196969696969697,
+      "grad_norm": 0.003556201932951808,
+      "learning_rate": 3.918790574514338e-06,
+      "loss": 0.0,
+      "step": 435
+    },
+    {
+      "epoch": 2.202020202020202,
+      "grad_norm": 0.037242766469717026,
+      "learning_rate": 3.91284103605648e-06,
+      "loss": 0.0,
+      "step": 436
+    },
+    {
+      "epoch": 2.207070707070707,
+      "grad_norm": 0.0035004725214093924,
+      "learning_rate": 3.906879717879005e-06,
+      "loss": 0.0,
+      "step": 437
+    },
+    {
+      "epoch": 2.212121212121212,
+      "grad_norm": 0.19019567966461182,
+      "learning_rate": 3.9009066696850664e-06,
+      "loss": 0.0009,
+      "step": 438
+    },
+    {
+      "epoch": 2.217171717171717,
+      "grad_norm": 0.3347993493080139,
+      "learning_rate": 3.89492194127562e-06,
+      "loss": 0.0036,
+      "step": 439
+    },
+    {
+      "epoch": 2.2222222222222223,
+      "grad_norm": 0.013379104435443878,
+      "learning_rate": 3.888925582549006e-06,
+      "loss": 0.0,
+      "step": 440
+    },
+    {
+      "epoch": 2.227272727272727,
+      "grad_norm": 0.22218060493469238,
+      "learning_rate": 3.882917643500534e-06,
+      "loss": 0.0022,
+      "step": 441
+    },
+    {
+      "epoch": 2.2323232323232323,
+      "grad_norm": 0.011584995314478874,
+      "learning_rate": 3.8768981742220646e-06,
+      "loss": 0.0,
+      "step": 442
+    },
+    {
+      "epoch": 2.2373737373737375,
+      "grad_norm": 0.012700509279966354,
+      "learning_rate": 3.870867224901595e-06,
+      "loss": 0.0,
+      "step": 443
+    },
+    {
+      "epoch": 2.242424242424242,
+      "grad_norm": 0.11896836757659912,
+      "learning_rate": 3.864824845822837e-06,
+      "loss": 0.0001,
+      "step": 444
+    },
+    {
+      "epoch": 2.2474747474747474,
+      "grad_norm": 0.010297846049070358,
+      "learning_rate": 3.8587710873647985e-06,
+      "loss": 0.0,
+      "step": 445
+    },
+    {
+      "epoch": 2.2525252525252526,
+      "grad_norm": 0.008279677480459213,
+      "learning_rate": 3.852706000001367e-06,
+      "loss": 0.0,
+      "step": 446
+    },
+    {
+      "epoch": 2.257575757575758,
+      "grad_norm": 0.25368809700012207,
+      "learning_rate": 3.846629634300883e-06,
+      "loss": 0.0005,
+      "step": 447
+    },
+    {
+      "epoch": 2.2626262626262625,
+      "grad_norm": 0.03778368607163429,
+      "learning_rate": 3.840542040925725e-06,
+      "loss": 0.0001,
+      "step": 448
+    },
+    {
+      "epoch": 2.2676767676767677,
+      "grad_norm": 0.017897954210639,
+      "learning_rate": 3.834443270631879e-06,
+      "loss": 0.0001,
+      "step": 449
+    },
+    {
+      "epoch": 2.2727272727272725,
+      "grad_norm": 0.011958397924900055,
+      "learning_rate": 3.828333374268523e-06,
+      "loss": 0.0001,
+      "step": 450
+    },
+    {
+      "epoch": 2.2777777777777777,
+      "grad_norm": 0.03435271605849266,
+      "learning_rate": 3.8222124027775994e-06,
+      "loss": 0.0001,
+      "step": 451
+    },
+    {
+      "epoch": 2.282828282828283,
+      "grad_norm": 0.012268775142729282,
+      "learning_rate": 3.81608040719339e-06,
+      "loss": 0.0001,
+      "step": 452
+    },
+    {
+      "epoch": 2.287878787878788,
+      "grad_norm": 0.03738567978143692,
+      "learning_rate": 3.8099374386420906e-06,
+      "loss": 0.0001,
+      "step": 453
+    },
+    {
+      "epoch": 2.292929292929293,
+      "grad_norm": 0.027487782761454582,
+      "learning_rate": 3.8037835483413877e-06,
+      "loss": 0.0001,
+      "step": 454
+    },
+    {
+      "epoch": 2.297979797979798,
+      "grad_norm": 0.02155883237719536,
+      "learning_rate": 3.7976187876000266e-06,
+      "loss": 0.0001,
+      "step": 455
+    },
+    {
+      "epoch": 2.303030303030303,
+      "grad_norm": 0.044139981269836426,
+      "learning_rate": 3.7914432078173867e-06,
+      "loss": 0.0001,
+      "step": 456
+    },
+    {
+      "epoch": 2.308080808080808,
+      "grad_norm": 0.10077373683452606,
+      "learning_rate": 3.785256860483054e-06,
+      "loss": 0.0001,
+      "step": 457
+    },
+    {
+      "epoch": 2.313131313131313,
+      "grad_norm": 0.01995331421494484,
+      "learning_rate": 3.7790597971763892e-06,
+      "loss": 0.0,
+      "step": 458
+    },
+    {
+      "epoch": 2.3181818181818183,
+      "grad_norm": 0.014978822320699692,
+      "learning_rate": 3.772852069566097e-06,
+      "loss": 0.0,
+      "step": 459
+    },
+    {
+      "epoch": 2.323232323232323,
+      "grad_norm": 0.0070231300778687,
+      "learning_rate": 3.7666337294097987e-06,
+      "loss": 0.0,
+      "step": 460
+    },
+    {
+      "epoch": 2.3282828282828283,
+      "grad_norm": 0.007256317883729935,
+      "learning_rate": 3.760404828553599e-06,
+      "loss": 0.0,
+      "step": 461
+    },
+    {
+      "epoch": 2.3333333333333335,
+      "grad_norm": 0.02269594930112362,
+      "learning_rate": 3.7541654189316525e-06,
+      "loss": 0.0001,
+      "step": 462
+    },
+    {
+      "epoch": 2.3383838383838382,
+      "grad_norm": 0.004164062440395355,
+      "learning_rate": 3.7479155525657323e-06,
+      "loss": 0.0,
+      "step": 463
+    },
+    {
+      "epoch": 2.3434343434343434,
+      "grad_norm": 0.00466076610609889,
+      "learning_rate": 3.741655281564796e-06,
+      "loss": 0.0,
+      "step": 464
+    },
+    {
+      "epoch": 2.3484848484848486,
+      "grad_norm": 0.009731332771480083,
+      "learning_rate": 3.735384658124551e-06,
+      "loss": 0.0,
+      "step": 465
+    },
+    {
+      "epoch": 2.3535353535353534,
+      "grad_norm": 0.46250560879707336,
+      "learning_rate": 3.72910373452702e-06,
+      "loss": 0.0003,
+      "step": 466
+    },
+    {
+      "epoch": 2.3585858585858586,
+      "grad_norm": 0.009878060780465603,
+      "learning_rate": 3.722812563140102e-06,
+      "loss": 0.0,
+      "step": 467
+    },
+    {
+      "epoch": 2.3636363636363638,
+      "grad_norm": 0.006603723857551813,
+      "learning_rate": 3.7165111964171407e-06,
+      "loss": 0.0,
+      "step": 468
+    },
+    {
+      "epoch": 2.3686868686868685,
+      "grad_norm": 0.17878268659114838,
+      "learning_rate": 3.7101996868964836e-06,
+      "loss": 0.001,
+      "step": 469
+    },
+    {
+      "epoch": 2.3737373737373737,
+      "grad_norm": 0.0032485495321452618,
+      "learning_rate": 3.703878087201044e-06,
+      "loss": 0.0,
+      "step": 470
+    },
+    {
+      "epoch": 2.378787878787879,
+      "grad_norm": 0.004560573026537895,
+      "learning_rate": 3.697546450037865e-06,
+      "loss": 0.0,
+      "step": 471
+    },
+    {
+      "epoch": 2.3838383838383836,
+      "grad_norm": 0.007358085829764605,
+      "learning_rate": 3.6912048281976764e-06,
+      "loss": 0.0,
+      "step": 472
+    },
+    {
+      "epoch": 2.388888888888889,
+      "grad_norm": 0.007086644880473614,
+      "learning_rate": 3.6848532745544574e-06,
+      "loss": 0.0,
+      "step": 473
+    },
+    {
+      "epoch": 2.393939393939394,
+      "grad_norm": 0.06341815739870071,
+      "learning_rate": 3.6784918420649952e-06,
+      "loss": 0.0003,
+      "step": 474
+    },
+    {
+      "epoch": 2.398989898989899,
+      "grad_norm": 0.005435152444988489,
+      "learning_rate": 3.67212058376844e-06,
+      "loss": 0.0,
+      "step": 475
+    },
+    {
+      "epoch": 2.404040404040404,
+      "grad_norm": 0.004161132499575615,
+      "learning_rate": 3.66573955278587e-06,
+      "loss": 0.0,
+      "step": 476
+    },
+    {
+      "epoch": 2.409090909090909,
+      "grad_norm": 0.006175799760967493,
+      "learning_rate": 3.6593488023198427e-06,
+      "loss": 0.0,
+      "step": 477
+    },
+    {
+      "epoch": 2.4141414141414144,
+      "grad_norm": 0.015126085840165615,
+      "learning_rate": 3.6529483856539512e-06,
+      "loss": 0.0,
+      "step": 478
+    },
+    {
+      "epoch": 2.419191919191919,
+      "grad_norm": 0.003908907063305378,
+      "learning_rate": 3.6465383561523825e-06,
+      "loss": 0.0,
+      "step": 479
+    },
+    {
+      "epoch": 2.4242424242424243,
+      "grad_norm": 0.0020491559989750385,
+      "learning_rate": 3.640118767259474e-06,
+      "loss": 0.0,
+      "step": 480
+    },
+    {
+      "epoch": 2.429292929292929,
+      "grad_norm": 0.00249357963912189,
+      "learning_rate": 3.6336896724992614e-06,
+      "loss": 0.0,
+      "step": 481
+    },
+    {
+      "epoch": 2.4343434343434343,
+      "grad_norm": 0.003509709145873785,
+      "learning_rate": 3.6272511254750403e-06,
+      "loss": 0.0,
+      "step": 482
+    },
+    {
+      "epoch": 2.4393939393939394,
+      "grad_norm": 0.04434852674603462,
+      "learning_rate": 3.620803179868913e-06,
+      "loss": 0.0001,
+      "step": 483
+    },
+    {
+      "epoch": 2.4444444444444446,
+      "grad_norm": 0.004402869381010532,
+      "learning_rate": 3.6143458894413463e-06,
+      "loss": 0.0,
+      "step": 484
+    },
+    {
+      "epoch": 2.4494949494949494,
+      "grad_norm": 0.19418330490589142,
+      "learning_rate": 3.6078793080307173e-06,
+      "loss": 0.0011,
+      "step": 485
+    },
+    {
+      "epoch": 2.4545454545454546,
+      "grad_norm": 0.006222233176231384,
+      "learning_rate": 3.6014034895528705e-06,
+      "loss": 0.0,
+      "step": 486
+    },
+    {
+      "epoch": 2.45959595959596,
+      "grad_norm": 0.2598673105239868,
+      "learning_rate": 3.594918488000664e-06,
+      "loss": 0.0013,
+      "step": 487
+    },
+    {
+      "epoch": 2.4646464646464645,
+      "grad_norm": 0.00211211945861578,
+      "learning_rate": 3.588424357443521e-06,
+      "loss": 0.0,
+      "step": 488
+    },
+    {
+      "epoch": 2.4696969696969697,
+      "grad_norm": 0.005605990532785654,
+      "learning_rate": 3.581921152026978e-06,
+      "loss": 0.0,
+      "step": 489
+    },
+    {
+      "epoch": 2.474747474747475,
+      "grad_norm": 0.007124903611838818,
+      "learning_rate": 3.5754089259722365e-06,
+      "loss": 0.0,
+      "step": 490
+    },
+    {
+      "epoch": 2.4797979797979797,
+      "grad_norm": 1.6332594156265259,
+      "learning_rate": 3.5688877335757055e-06,
+      "loss": 0.0024,
+      "step": 491
+    },
+    {
+      "epoch": 2.484848484848485,
+      "grad_norm": 0.002983370330184698,
+      "learning_rate": 3.5623576292085555e-06,
+      "loss": 0.0,
+      "step": 492
+    },
+    {
+      "epoch": 2.48989898989899,
+      "grad_norm": 0.04062391445040703,
+      "learning_rate": 3.5558186673162575e-06,
+      "loss": 0.0001,
+      "step": 493
+    },
+    {
+      "epoch": 2.494949494949495,
+      "grad_norm": 0.006106998771429062,
+      "learning_rate": 3.549270902418136e-06,
+      "loss": 0.0,
+      "step": 494
+    },
+    {
+      "epoch": 2.5,
+      "grad_norm": 0.04901747405529022,
+      "learning_rate": 3.542714389106911e-06,
+      "loss": 0.0001,
+      "step": 495
+    },
+    {
+      "epoch": 2.505050505050505,
+      "grad_norm": 0.006030434276908636,
+      "learning_rate": 3.536149182048243e-06,
+      "loss": 0.0,
+      "step": 496
+    },
+    {
+      "epoch": 2.51010101010101,
+      "grad_norm": 0.009680806659162045,
+      "learning_rate": 3.529575335980277e-06,
+      "loss": 0.0,
+      "step": 497
+    },
+    {
+      "epoch": 2.515151515151515,
+      "grad_norm": 0.003187371650710702,
+      "learning_rate": 3.5229929057131877e-06,
+      "loss": 0.0,
+      "step": 498
+    },
+    {
+      "epoch": 2.5202020202020203,
+      "grad_norm": 0.14173120260238647,
+      "learning_rate": 3.5164019461287215e-06,
+      "loss": 0.0004,
+      "step": 499
+    },
+    {
+      "epoch": 2.525252525252525,
+      "grad_norm": 0.006695139221847057,
+      "learning_rate": 3.5098025121797375e-06,
+      "loss": 0.0,
+      "step": 500
+    },
+    {
+      "epoch": 2.5303030303030303,
+      "grad_norm": 0.005796091165393591,
+      "learning_rate": 3.503194658889753e-06,
+      "loss": 0.0,
+      "step": 501
+    },
+    {
+      "epoch": 2.5353535353535355,
+      "grad_norm": 0.007152227684855461,
+      "learning_rate": 3.496578441352481e-06,
+      "loss": 0.0,
+      "step": 502
+    },
+    {
+      "epoch": 2.5404040404040407,
+      "grad_norm": 0.007902374491095543,
+      "learning_rate": 3.4899539147313727e-06,
+      "loss": 0.0,
+      "step": 503
+    },
+    {
+      "epoch": 2.5454545454545454,
+      "grad_norm": 0.04069540649652481,
+      "learning_rate": 3.4833211342591565e-06,
+      "loss": 0.0,
+      "step": 504
+    },
+    {
+      "epoch": 2.5505050505050506,
+      "grad_norm": 0.040364135056734085,
+      "learning_rate": 3.4766801552373804e-06,
+      "loss": 0.0001,
+      "step": 505
+    },
+    {
+      "epoch": 2.5555555555555554,
+      "grad_norm": 0.0079901572316885,
+      "learning_rate": 3.4700310330359456e-06,
+      "loss": 0.0,
+      "step": 506
+    },
+    {
+      "epoch": 2.5606060606060606,
+      "grad_norm": 0.004804165568202734,
+      "learning_rate": 3.4633738230926508e-06,
+      "loss": 0.0,
+      "step": 507
+    },
+    {
+      "epoch": 2.5656565656565657,
+      "grad_norm": 0.0025831114035099745,
+      "learning_rate": 3.4567085809127247e-06,
+      "loss": 0.0,
+      "step": 508
+    },
+    {
+      "epoch": 2.570707070707071,
+      "grad_norm": 0.012033859267830849,
+      "learning_rate": 3.450035362068368e-06,
+      "loss": 0.0,
+      "step": 509
+    },
+    {
+      "epoch": 2.5757575757575757,
+      "grad_norm": 0.002701899502426386,
+      "learning_rate": 3.4433542221982863e-06,
+      "loss": 0.0,
+      "step": 510
+    },
+    {
+      "epoch": 2.580808080808081,
+      "grad_norm": 0.008882194757461548,
+      "learning_rate": 3.4366652170072277e-06,
+      "loss": 0.0,
+      "step": 511
+    },
+    {
+      "epoch": 2.5858585858585856,
+      "grad_norm": 0.15618982911109924,
+      "learning_rate": 3.4299684022655196e-06,
+      "loss": 0.0006,
+      "step": 512
+    },
+    {
+      "epoch": 2.590909090909091,
+      "grad_norm": 0.0035035579930990934,
+      "learning_rate": 3.4232638338085996e-06,
+      "loss": 0.0,
+      "step": 513
+    },
+    {
+      "epoch": 2.595959595959596,
+      "grad_norm": 0.03973141312599182,
+      "learning_rate": 3.4165515675365558e-06,
+      "loss": 0.0001,
+      "step": 514
+    },
+    {
+      "epoch": 2.601010101010101,
+      "grad_norm": 0.0064284526742994785,
+      "learning_rate": 3.4098316594136554e-06,
+      "loss": 0.0,
+      "step": 515
+    },
+    {
+      "epoch": 2.606060606060606,
+      "grad_norm": 0.008207082748413086,
+      "learning_rate": 3.403104165467883e-06,
+      "loss": 0.0,
+      "step": 516
+    },
+    {
+      "epoch": 2.611111111111111,
+      "grad_norm": 0.02613808587193489,
+      "learning_rate": 3.3963691417904676e-06,
+      "loss": 0.0,
+      "step": 517
+    },
+    {
+      "epoch": 2.616161616161616,
+      "grad_norm": 0.0026936449576169252,
+      "learning_rate": 3.3896266445354208e-06,
+      "loss": 0.0,
+      "step": 518
+    },
+    {
+      "epoch": 2.621212121212121,
+      "grad_norm": 0.006232825573533773,
+      "learning_rate": 3.3828767299190646e-06,
+      "loss": 0.0,
+      "step": 519
+    },
+    {
+      "epoch": 2.6262626262626263,
+      "grad_norm": 0.004667866975069046,
+      "learning_rate": 3.376119454219565e-06,
+      "loss": 0.0,
+      "step": 520
+    },
+    {
+      "epoch": 2.6313131313131315,
+      "grad_norm": 0.00448667211458087,
+      "learning_rate": 3.3693548737764607e-06,
+      "loss": 0.0,
+      "step": 521
+    },
+    {
+      "epoch": 2.6363636363636362,
+      "grad_norm": 0.009849603287875652,
+      "learning_rate": 3.362583044990195e-06,
+      "loss": 0.0,
+      "step": 522
+    },
+    {
+      "epoch": 2.6414141414141414,
+      "grad_norm": 0.041966211050748825,
+      "learning_rate": 3.3558040243216463e-06,
+      "loss": 0.0,
+      "step": 523
+    },
+    {
+      "epoch": 2.6464646464646466,
+      "grad_norm": 0.021129852160811424,
+      "learning_rate": 3.3490178682916534e-06,
+      "loss": 0.0,
+      "step": 524
+    },
+    {
+      "epoch": 2.6515151515151514,
+      "grad_norm": 0.010130657814443111,
+      "learning_rate": 3.3422246334805504e-06,
+      "loss": 0.0,
+      "step": 525
+    },
+    {
+      "epoch": 2.6565656565656566,
+      "grad_norm": 0.004168142564594746,
+      "learning_rate": 3.335424376527688e-06,
+      "loss": 0.0,
+      "step": 526
+    },
+    {
+      "epoch": 2.6616161616161618,
+      "grad_norm": 0.002758739748969674,
+      "learning_rate": 3.3286171541309686e-06,
+      "loss": 0.0,
+      "step": 527
+    },
+    {
+      "epoch": 2.6666666666666665,
+      "grad_norm": 0.005857916083186865,
+      "learning_rate": 3.321803023046366e-06,
+      "loss": 0.0,
+      "step": 528
+    },
+    {
+      "epoch": 2.6717171717171717,
+      "grad_norm": 0.02048281952738762,
+      "learning_rate": 3.3149820400874575e-06,
+      "loss": 0.0,
+      "step": 529
+    },
+    {
+      "epoch": 2.676767676767677,
+      "grad_norm": 0.010321596637368202,
+      "learning_rate": 3.3081542621249503e-06,
+      "loss": 0.0,
+      "step": 530
+    },
+    {
+      "epoch": 2.6818181818181817,
+      "grad_norm": 0.008361139334738255,
+      "learning_rate": 3.301319746086203e-06,
+      "loss": 0.0,
+      "step": 531
+    },
+    {
+      "epoch": 2.686868686868687,
+      "grad_norm": 0.02190493792295456,
+      "learning_rate": 3.2944785489547544e-06,
+      "loss": 0.0,
+      "step": 532
+    },
+    {
+      "epoch": 2.691919191919192,
+      "grad_norm": 0.002024057786911726,
+      "learning_rate": 3.2876307277698495e-06,
+      "loss": 0.0,
+      "step": 533
+    },
+    {
+      "epoch": 2.6969696969696972,
+      "grad_norm": 0.008165915496647358,
+      "learning_rate": 3.2807763396259597e-06,
+      "loss": 0.0,
+      "step": 534
+    },
+    {
+      "epoch": 2.702020202020202,
+      "grad_norm": 0.0023303315974771976,
+      "learning_rate": 3.2739154416723113e-06,
+      "loss": 0.0,
+      "step": 535
+    },
+    {
+      "epoch": 2.707070707070707,
+      "grad_norm": 0.002119947923347354,
+      "learning_rate": 3.2670480911124045e-06,
+      "loss": 0.0,
+      "step": 536
+    },
+    {
+      "epoch": 2.712121212121212,
+      "grad_norm": 0.0013211388140916824,
+      "learning_rate": 3.2601743452035413e-06,
+      "loss": 0.0,
+      "step": 537
+    },
+    {
+      "epoch": 2.717171717171717,
+      "grad_norm": 0.003069607773795724,
+      "learning_rate": 3.2532942612563436e-06,
+      "loss": 0.0,
+      "step": 538
+    },
+    {
+      "epoch": 2.7222222222222223,
+      "grad_norm": 0.002737869042903185,
+      "learning_rate": 3.246407896634277e-06,
+      "loss": 0.0,
+      "step": 539
+    },
+    {
+      "epoch": 2.7272727272727275,
+      "grad_norm": 0.002814906882122159,
+      "learning_rate": 3.2395153087531767e-06,
+      "loss": 0.0,
+      "step": 540
+    },
+    {
+      "epoch": 2.7323232323232323,
+      "grad_norm": 0.00364164961501956,
+      "learning_rate": 3.23261655508076e-06,
+      "loss": 0.0,
+      "step": 541
+    },
+    {
+      "epoch": 2.7373737373737375,
+      "grad_norm": 0.002594191348180175,
+      "learning_rate": 3.225711693136156e-06,
+      "loss": 0.0,
+      "step": 542
+    },
+    {
+      "epoch": 2.742424242424242,
+      "grad_norm": 0.0014940836699679494,
+      "learning_rate": 3.2188007804894206e-06,
+      "loss": 0.0,
+      "step": 543
+    },
+    {
+      "epoch": 2.7474747474747474,
+      "grad_norm": 0.002774301450699568,
+      "learning_rate": 3.211883874761058e-06,
+      "loss": 0.0,
+      "step": 544
+    },
+    {
+      "epoch": 2.7525252525252526,
+      "grad_norm": 0.0019573576282709837,
+      "learning_rate": 3.2049610336215416e-06,
+      "loss": 0.0,
+      "step": 545
+    },
+    {
+      "epoch": 2.757575757575758,
+      "grad_norm": 0.0017838666681200266,
+      "learning_rate": 3.19803231479083e-06,
+      "loss": 0.0,
+      "step": 546
+    },
+    {
+      "epoch": 2.7626262626262625,
+      "grad_norm": 0.0023138851393014193,
+      "learning_rate": 3.1910977760378884e-06,
+      "loss": 0.0,
+      "step": 547
+    },
+    {
+      "epoch": 2.7676767676767677,
+      "grad_norm": 0.005525521468371153,
+      "learning_rate": 3.184157475180208e-06,
+      "loss": 0.0,
+      "step": 548
+    },
+    {
+      "epoch": 2.7727272727272725,
+      "grad_norm": 0.0011541108833625913,
+      "learning_rate": 3.1772114700833194e-06,
+      "loss": 0.0,
+      "step": 549
+    },
+    {
+      "epoch": 2.7777777777777777,
+      "grad_norm": 0.0012338707456365228,
+      "learning_rate": 3.1702598186603152e-06,
+      "loss": 0.0,
+      "step": 550
+    },
+    {
+      "epoch": 2.782828282828283,
+      "grad_norm": 0.0019040412735193968,
+      "learning_rate": 3.1633025788713632e-06,
+      "loss": 0.0,
+      "step": 551
+    },
+    {
+      "epoch": 2.787878787878788,
+      "grad_norm": 0.15179114043712616,
+      "learning_rate": 3.1563398087232265e-06,
+      "loss": 0.0003,
+      "step": 552
+    },
+    {
+      "epoch": 2.792929292929293,
+      "grad_norm": 0.0010482153156772256,
+      "learning_rate": 3.1493715662687774e-06,
+      "loss": 0.0,
+      "step": 553
+    },
+    {
+      "epoch": 2.797979797979798,
+      "grad_norm": 0.26141083240509033,
+      "learning_rate": 3.1423979096065134e-06,
+      "loss": 0.0009,
+      "step": 554
+    },
+    {
+      "epoch": 2.8030303030303028,
+      "grad_norm": 0.00429484061896801,
+      "learning_rate": 3.135418896880074e-06,
+      "loss": 0.0,
+      "step": 555
+    },
+    {
+      "epoch": 2.808080808080808,
+      "grad_norm": 0.010286355391144753,
+      "learning_rate": 3.1284345862777572e-06,
+      "loss": 0.0,
+      "step": 556
+    },
+    {
+      "epoch": 2.813131313131313,
+      "grad_norm": 0.0035182933788746595,
+      "learning_rate": 3.121445036032029e-06,
+      "loss": 0.0,
+      "step": 557
+    },
+    {
+      "epoch": 2.8181818181818183,
+      "grad_norm": 0.005269261077046394,
+      "learning_rate": 3.1144503044190456e-06,
+      "loss": 0.0,
+      "step": 558
+    },
+    {
+      "epoch": 2.823232323232323,
+      "grad_norm": 0.003896774258464575,
+      "learning_rate": 3.10745044975816e-06,
+      "loss": 0.0,
+      "step": 559
+    },
+    {
+      "epoch": 2.8282828282828283,
+      "grad_norm": 0.0011062441626563668,
+      "learning_rate": 3.100445530411442e-06,
+      "loss": 0.0,
+      "step": 560
+    },
+    {
+      "epoch": 2.8333333333333335,
+      "grad_norm": 0.003219842677935958,
+      "learning_rate": 3.093435604783186e-06,
+      "loss": 0.0,
+      "step": 561
+    },
+    {
+      "epoch": 2.8383838383838382,
+      "grad_norm": 0.0029111194889992476,
+      "learning_rate": 3.086420731319429e-06,
+      "loss": 0.0,
+      "step": 562
+    },
+    {
+      "epoch": 2.8434343434343434,
+      "grad_norm": 0.0015513853868469596,
+      "learning_rate": 3.0794009685074606e-06,
+      "loss": 0.0,
+      "step": 563
+    },
+    {
+      "epoch": 2.8484848484848486,
+      "grad_norm": 0.0007438853499479592,
+      "learning_rate": 3.0723763748753354e-06,
+      "loss": 0.0,
+      "step": 564
+    },
+    {
+      "epoch": 2.8535353535353534,
+      "grad_norm": 0.0032596453092992306,
+      "learning_rate": 3.065347008991386e-06,
+      "loss": 0.0,
+      "step": 565
+    },
+    {
+      "epoch": 2.8585858585858586,
+      "grad_norm": 0.0018467978807166219,
+      "learning_rate": 3.0583129294637342e-06,
+      "loss": 0.0,
+      "step": 566
+    },
+    {
+      "epoch": 2.8636363636363638,
+      "grad_norm": 0.004413960501551628,
+      "learning_rate": 3.051274194939802e-06,
+      "loss": 0.0,
+      "step": 567
+    },
+    {
+      "epoch": 2.8686868686868685,
+      "grad_norm": 0.020187318325042725,
+      "learning_rate": 3.044230864105821e-06,
+      "loss": 0.0,
+      "step": 568
+    },
+    {
+      "epoch": 2.8737373737373737,
+      "grad_norm": 0.0026113647036254406,
+      "learning_rate": 3.03718299568635e-06,
+      "loss": 0.0,
+      "step": 569
+    },
+    {
+      "epoch": 2.878787878787879,
+      "grad_norm": 0.006202157121151686,
+      "learning_rate": 3.030130648443777e-06,
+      "loss": 0.0,
+      "step": 570
+    },
+    {
+      "epoch": 2.883838383838384,
+      "grad_norm": 0.0044505237601697445,
+      "learning_rate": 3.023073881177833e-06,
+      "loss": 0.0,
+      "step": 571
+    },
+    {
+      "epoch": 2.888888888888889,
+      "grad_norm": 0.0023774108849465847,
+      "learning_rate": 3.0160127527250993e-06,
+      "loss": 0.0,
+      "step": 572
+    },
+    {
+      "epoch": 2.893939393939394,
+      "grad_norm": 0.005300168413668871,
+      "learning_rate": 3.0089473219585246e-06,
+      "loss": 0.0,
+      "step": 573
+    },
+    {
+      "epoch": 2.898989898989899,
+      "grad_norm": 0.0016464275540784001,
+      "learning_rate": 3.0018776477869244e-06,
+      "loss": 0.0,
+      "step": 574
+    },
+    {
+      "epoch": 2.904040404040404,
+      "grad_norm": 0.0017848126590251923,
+      "learning_rate": 2.994803789154495e-06,
+      "loss": 0.0,
+      "step": 575
+    },
+    {
+      "epoch": 2.909090909090909,
+      "grad_norm": 0.0014869156293570995,
+      "learning_rate": 2.9877258050403214e-06,
+      "loss": 0.0,
+      "step": 576
+    },
+    {
+      "epoch": 2.9141414141414144,
+      "grad_norm": 0.006105687469244003,
+      "learning_rate": 2.9806437544578844e-06,
+      "loss": 0.0,
+      "step": 577
+    },
+    {
+      "epoch": 2.919191919191919,
+      "grad_norm": 0.006673851516097784,
+      "learning_rate": 2.973557696454571e-06,
+      "loss": 0.0,
+      "step": 578
+    },
+    {
+      "epoch": 2.9242424242424243,
+      "grad_norm": 0.002130689565092325,
+      "learning_rate": 2.966467690111179e-06,
+      "loss": 0.0,
+      "step": 579
+    },
+    {
+      "epoch": 2.929292929292929,
+      "grad_norm": 0.0036351734306663275,
+      "learning_rate": 2.9593737945414264e-06,
+      "loss": 0.0,
+      "step": 580
+    },
+    {
+      "epoch": 2.9343434343434343,
+      "grad_norm": 0.08766695111989975,
+      "learning_rate": 2.9522760688914587e-06,
+      "loss": 0.0005,
+      "step": 581
+    },
+    {
+      "epoch": 2.9393939393939394,
+      "grad_norm": 0.0016932138241827488,
+      "learning_rate": 2.9451745723393547e-06,
+      "loss": 0.0,
+      "step": 582
+    },
+    {
+      "epoch": 2.9444444444444446,
+      "grad_norm": 0.04073040559887886,
+      "learning_rate": 2.938069364094634e-06,
+      "loss": 0.0,
+      "step": 583
+    },
+    {
+      "epoch": 2.9494949494949494,
+      "grad_norm": 0.0032754631247371435,
+      "learning_rate": 2.930960503397761e-06,
+      "loss": 0.0,
+      "step": 584
+    },
+    {
+      "epoch": 2.9545454545454546,
+      "grad_norm": 0.0012018376728519797,
+      "learning_rate": 2.9238480495196543e-06,
+      "loss": 0.0,
+      "step": 585
+    },
+    {
+      "epoch": 2.9595959595959593,
+      "grad_norm": 0.0016919056652113795,
+      "learning_rate": 2.916732061761192e-06,
+      "loss": 0.0,
+      "step": 586
+    },
+    {
+      "epoch": 2.9646464646464645,
+      "grad_norm": 0.0013172467006370425,
+      "learning_rate": 2.9096125994527147e-06,
+      "loss": 0.0,
+      "step": 587
+    },
+    {
+      "epoch": 2.9696969696969697,
+      "grad_norm": 0.007328435312956572,
+      "learning_rate": 2.9024897219535326e-06,
+      "loss": 0.0,
+      "step": 588
+    },
+    {
+      "epoch": 2.974747474747475,
+      "grad_norm": 0.0016528492560610175,
+      "learning_rate": 2.895363488651431e-06,
+      "loss": 0.0,
+      "step": 589
+    },
+    {
+      "epoch": 2.9797979797979797,
+      "grad_norm": 0.001276439055800438,
+      "learning_rate": 2.8882339589621742e-06,
+      "loss": 0.0,
+      "step": 590
+    },
+    {
+      "epoch": 2.984848484848485,
+      "grad_norm": 0.0013321200385689735,
+      "learning_rate": 2.8811011923290104e-06,
+      "loss": 0.0,
+      "step": 591
+    },
+    {
+      "epoch": 2.98989898989899,
+      "grad_norm": 0.002064246218651533,
+      "learning_rate": 2.873965248222178e-06,
+      "loss": 0.0,
+      "step": 592
+    },
+    {
+      "epoch": 2.994949494949495,
+      "grad_norm": 0.001075342413969338,
+      "learning_rate": 2.8668261861384045e-06,
+      "loss": 0.0,
+      "step": 593
+    },
+    {
+      "epoch": 3.0,
+      "grad_norm": 0.0037419938016682863,
+      "learning_rate": 2.859684065600417e-06,
+      "loss": 0.0,
+      "step": 594
+    }
+  ],
+  "logging_steps": 1,
+  "max_steps": 1188,
+  "num_input_tokens_seen": 0,
+  "num_train_epochs": 6,
+  "save_steps": 198,
+  "stateful_callbacks": {
+    "TrainerControl": {
+      "args": {
+        "should_epoch_stop": false,
+        "should_evaluate": false,
+        "should_log": false,
+        "should_save": true,
+        "should_training_stop": false
+      },
+      "attributes": {}
+    }
+  },
+  "total_flos": 1.50046579787563e+18,
+  "train_batch_size": 4,
+  "trial_name": null,
+  "trial_params": null
+}