diff --git "a/trainer_state.json" "b/trainer_state.json"
new file mode 100644--- /dev/null
+++ "b/trainer_state.json"
@@ -0,0 +1,7026 @@
+{
+  "best_metric": null,
+  "best_model_checkpoint": null,
+  "epoch": 3.0,
+  "eval_steps": 0,
+  "global_step": 999,
+  "is_hyper_param_search": false,
+  "is_local_process_zero": true,
+  "is_world_process_zero": true,
+  "log_history": [
+    {
+      "epoch": 0.003003003003003003,
+      "grad_norm": 0.423828125,
+      "learning_rate": 9.989989989989992e-06,
+      "loss": 1.8145,
+      "step": 1
+    },
+    {
+      "epoch": 0.006006006006006006,
+      "grad_norm": 0.400390625,
+      "learning_rate": 9.979979979979981e-06,
+      "loss": 1.8558,
+      "step": 2
+    },
+    {
+      "epoch": 0.009009009009009009,
+      "grad_norm": 0.375,
+      "learning_rate": 9.96996996996997e-06,
+      "loss": 1.8834,
+      "step": 3
+    },
+    {
+      "epoch": 0.012012012012012012,
+      "grad_norm": 0.388671875,
+      "learning_rate": 9.95995995995996e-06,
+      "loss": 1.8832,
+      "step": 4
+    },
+    {
+      "epoch": 0.015015015015015015,
+      "grad_norm": 0.388671875,
+      "learning_rate": 9.949949949949951e-06,
+      "loss": 1.8962,
+      "step": 5
+    },
+    {
+      "epoch": 0.018018018018018018,
+      "grad_norm": 0.357421875,
+      "learning_rate": 9.93993993993994e-06,
+      "loss": 1.7917,
+      "step": 6
+    },
+    {
+      "epoch": 0.021021021021021023,
+      "grad_norm": 0.34765625,
+      "learning_rate": 9.929929929929931e-06,
+      "loss": 1.7484,
+      "step": 7
+    },
+    {
+      "epoch": 0.024024024024024024,
+      "grad_norm": 0.33984375,
+      "learning_rate": 9.91991991991992e-06,
+      "loss": 1.7862,
+      "step": 8
+    },
+    {
+      "epoch": 0.02702702702702703,
+      "grad_norm": 0.365234375,
+      "learning_rate": 9.90990990990991e-06,
+      "loss": 1.8331,
+      "step": 9
+    },
+    {
+      "epoch": 0.03003003003003003,
+      "grad_norm": 0.326171875,
+      "learning_rate": 9.899899899899901e-06,
+      "loss": 1.7138,
+      "step": 10
+    },
+    {
+      "epoch": 0.03303303303303303,
+      "grad_norm": 0.322265625,
+      "learning_rate": 9.88988988988989e-06,
+      "loss": 1.8368,
+      "step": 11
+    },
+    {
+      "epoch": 0.036036036036036036,
+      "grad_norm": 0.294921875,
+      "learning_rate": 9.879879879879881e-06,
+      "loss": 1.6865,
+      "step": 12
+    },
+    {
+      "epoch": 0.03903903903903904,
+      "grad_norm": 0.3203125,
+      "learning_rate": 9.86986986986987e-06,
+      "loss": 1.6875,
+      "step": 13
+    },
+    {
+      "epoch": 0.042042042042042045,
+      "grad_norm": 0.2890625,
+      "learning_rate": 9.85985985985986e-06,
+      "loss": 1.7557,
+      "step": 14
+    },
+    {
+      "epoch": 0.04504504504504504,
+      "grad_norm": 0.283203125,
+      "learning_rate": 9.849849849849851e-06,
+      "loss": 1.6358,
+      "step": 15
+    },
+    {
+      "epoch": 0.04804804804804805,
+      "grad_norm": 1.1640625,
+      "learning_rate": 9.83983983983984e-06,
+      "loss": 1.6761,
+      "step": 16
+    },
+    {
+      "epoch": 0.05105105105105105,
+      "grad_norm": 0.28125,
+      "learning_rate": 9.829829829829831e-06,
+      "loss": 1.545,
+      "step": 17
+    },
+    {
+      "epoch": 0.05405405405405406,
+      "grad_norm": 0.296875,
+      "learning_rate": 9.81981981981982e-06,
+      "loss": 1.7298,
+      "step": 18
+    },
+    {
+      "epoch": 0.057057057057057055,
+      "grad_norm": 0.28125,
+      "learning_rate": 9.80980980980981e-06,
+      "loss": 1.6444,
+      "step": 19
+    },
+    {
+      "epoch": 0.06006006006006006,
+      "grad_norm": 0.263671875,
+      "learning_rate": 9.799799799799801e-06,
+      "loss": 1.5753,
+      "step": 20
+    },
+    {
+      "epoch": 0.06306306306306306,
+      "grad_norm": 0.25,
+      "learning_rate": 9.78978978978979e-06,
+      "loss": 1.5862,
+      "step": 21
+    },
+    {
+      "epoch": 0.06606606606606606,
+      "grad_norm": 0.2431640625,
+      "learning_rate": 9.779779779779781e-06,
+      "loss": 1.5533,
+      "step": 22
+    },
+    {
+      "epoch": 0.06906906906906907,
+      "grad_norm": 0.25390625,
+      "learning_rate": 9.76976976976977e-06,
+      "loss": 1.629,
+      "step": 23
+    },
+    {
+      "epoch": 0.07207207207207207,
+      "grad_norm": 0.23828125,
+      "learning_rate": 9.75975975975976e-06,
+      "loss": 1.6044,
+      "step": 24
+    },
+    {
+      "epoch": 0.07507507507507508,
+      "grad_norm": 0.2373046875,
+      "learning_rate": 9.749749749749751e-06,
+      "loss": 1.5288,
+      "step": 25
+    },
+    {
+      "epoch": 0.07807807807807808,
+      "grad_norm": 0.2177734375,
+      "learning_rate": 9.73973973973974e-06,
+      "loss": 1.4683,
+      "step": 26
+    },
+    {
+      "epoch": 0.08108108108108109,
+      "grad_norm": 0.232421875,
+      "learning_rate": 9.729729729729732e-06,
+      "loss": 1.5755,
+      "step": 27
+    },
+    {
+      "epoch": 0.08408408408408409,
+      "grad_norm": 0.2177734375,
+      "learning_rate": 9.719719719719721e-06,
+      "loss": 1.4974,
+      "step": 28
+    },
+    {
+      "epoch": 0.08708708708708708,
+      "grad_norm": 0.232421875,
+      "learning_rate": 9.70970970970971e-06,
+      "loss": 1.5748,
+      "step": 29
+    },
+    {
+      "epoch": 0.09009009009009009,
+      "grad_norm": 0.216796875,
+      "learning_rate": 9.699699699699701e-06,
+      "loss": 1.5117,
+      "step": 30
+    },
+    {
+      "epoch": 0.09309309309309309,
+      "grad_norm": 0.2373046875,
+      "learning_rate": 9.68968968968969e-06,
+      "loss": 1.444,
+      "step": 31
+    },
+    {
+      "epoch": 0.0960960960960961,
+      "grad_norm": 0.2158203125,
+      "learning_rate": 9.67967967967968e-06,
+      "loss": 1.5435,
+      "step": 32
+    },
+    {
+      "epoch": 0.0990990990990991,
+      "grad_norm": 0.19921875,
+      "learning_rate": 9.669669669669671e-06,
+      "loss": 1.4584,
+      "step": 33
+    },
+    {
+      "epoch": 0.1021021021021021,
+      "grad_norm": 0.1953125,
+      "learning_rate": 9.65965965965966e-06,
+      "loss": 1.4598,
+      "step": 34
+    },
+    {
+      "epoch": 0.10510510510510511,
+      "grad_norm": 0.1982421875,
+      "learning_rate": 9.649649649649651e-06,
+      "loss": 1.4071,
+      "step": 35
+    },
+    {
+      "epoch": 0.10810810810810811,
+      "grad_norm": 0.1904296875,
+      "learning_rate": 9.63963963963964e-06,
+      "loss": 1.3843,
+      "step": 36
+    },
+    {
+      "epoch": 0.1111111111111111,
+      "grad_norm": 0.2275390625,
+      "learning_rate": 9.62962962962963e-06,
+      "loss": 1.5031,
+      "step": 37
+    },
+    {
+      "epoch": 0.11411411411411411,
+      "grad_norm": 0.2158203125,
+      "learning_rate": 9.61961961961962e-06,
+      "loss": 1.4281,
+      "step": 38
+    },
+    {
+      "epoch": 0.11711711711711711,
+      "grad_norm": 0.4453125,
+      "learning_rate": 9.60960960960961e-06,
+      "loss": 1.4293,
+      "step": 39
+    },
+    {
+      "epoch": 0.12012012012012012,
+      "grad_norm": 0.17578125,
+      "learning_rate": 9.5995995995996e-06,
+      "loss": 1.3327,
+      "step": 40
+    },
+    {
+      "epoch": 0.12312312312312312,
+      "grad_norm": 0.1845703125,
+      "learning_rate": 9.58958958958959e-06,
+      "loss": 1.3545,
+      "step": 41
+    },
+    {
+      "epoch": 0.12612612612612611,
+      "grad_norm": 0.2138671875,
+      "learning_rate": 9.57957957957958e-06,
+      "loss": 1.3482,
+      "step": 42
+    },
+    {
+      "epoch": 0.12912912912912913,
+      "grad_norm": 0.1787109375,
+      "learning_rate": 9.56956956956957e-06,
+      "loss": 1.3785,
+      "step": 43
+    },
+    {
+      "epoch": 0.13213213213213212,
+      "grad_norm": 0.18359375,
+      "learning_rate": 9.55955955955956e-06,
+      "loss": 1.4396,
+      "step": 44
+    },
+    {
+      "epoch": 0.13513513513513514,
+      "grad_norm": 0.189453125,
+      "learning_rate": 9.54954954954955e-06,
+      "loss": 1.4395,
+      "step": 45
+    },
+    {
+      "epoch": 0.13813813813813813,
+      "grad_norm": 0.1650390625,
+      "learning_rate": 9.53953953953954e-06,
+      "loss": 1.2935,
+      "step": 46
+    },
+    {
+      "epoch": 0.14114114114114115,
+      "grad_norm": 0.1669921875,
+      "learning_rate": 9.52952952952953e-06,
+      "loss": 1.3763,
+      "step": 47
+    },
+    {
+      "epoch": 0.14414414414414414,
+      "grad_norm": 0.1787109375,
+      "learning_rate": 9.51951951951952e-06,
+      "loss": 1.3694,
+      "step": 48
+    },
+    {
+      "epoch": 0.14714714714714713,
+      "grad_norm": 0.19921875,
+      "learning_rate": 9.50950950950951e-06,
+      "loss": 1.3317,
+      "step": 49
+    },
+    {
+      "epoch": 0.15015015015015015,
+      "grad_norm": 0.171875,
+      "learning_rate": 9.4994994994995e-06,
+      "loss": 1.3111,
+      "step": 50
+    },
+    {
+      "epoch": 0.15315315315315314,
+      "grad_norm": 0.1669921875,
+      "learning_rate": 9.489489489489491e-06,
+      "loss": 1.3323,
+      "step": 51
+    },
+    {
+      "epoch": 0.15615615615615616,
+      "grad_norm": 0.166015625,
+      "learning_rate": 9.47947947947948e-06,
+      "loss": 1.3333,
+      "step": 52
+    },
+    {
+      "epoch": 0.15915915915915915,
+      "grad_norm": 0.166015625,
+      "learning_rate": 9.46946946946947e-06,
+      "loss": 1.3519,
+      "step": 53
+    },
+    {
+      "epoch": 0.16216216216216217,
+      "grad_norm": 0.22265625,
+      "learning_rate": 9.45945945945946e-06,
+      "loss": 1.3028,
+      "step": 54
+    },
+    {
+      "epoch": 0.16516516516516516,
+      "grad_norm": 0.169921875,
+      "learning_rate": 9.44944944944945e-06,
+      "loss": 1.3397,
+      "step": 55
+    },
+    {
+      "epoch": 0.16816816816816818,
+      "grad_norm": 0.1708984375,
+      "learning_rate": 9.439439439439441e-06,
+      "loss": 1.3241,
+      "step": 56
+    },
+    {
+      "epoch": 0.17117117117117117,
+      "grad_norm": 0.1669921875,
+      "learning_rate": 9.42942942942943e-06,
+      "loss": 1.2565,
+      "step": 57
+    },
+    {
+      "epoch": 0.17417417417417416,
+      "grad_norm": 0.1962890625,
+      "learning_rate": 9.41941941941942e-06,
+      "loss": 1.3746,
+      "step": 58
+    },
+    {
+      "epoch": 0.17717717717717718,
+      "grad_norm": 0.17578125,
+      "learning_rate": 9.40940940940941e-06,
+      "loss": 1.2678,
+      "step": 59
+    },
+    {
+      "epoch": 0.18018018018018017,
+      "grad_norm": 1.625,
+      "learning_rate": 9.3993993993994e-06,
+      "loss": 1.2557,
+      "step": 60
+    },
+    {
+      "epoch": 0.1831831831831832,
+      "grad_norm": 0.1787109375,
+      "learning_rate": 9.389389389389391e-06,
+      "loss": 1.2964,
+      "step": 61
+    },
+    {
+      "epoch": 0.18618618618618618,
+      "grad_norm": 0.16015625,
+      "learning_rate": 9.37937937937938e-06,
+      "loss": 1.2543,
+      "step": 62
+    },
+    {
+      "epoch": 0.1891891891891892,
+      "grad_norm": 0.158203125,
+      "learning_rate": 9.36936936936937e-06,
+      "loss": 1.2757,
+      "step": 63
+    },
+    {
+      "epoch": 0.1921921921921922,
+      "grad_norm": 0.2451171875,
+      "learning_rate": 9.35935935935936e-06,
+      "loss": 1.2978,
+      "step": 64
+    },
+    {
+      "epoch": 0.19519519519519518,
+      "grad_norm": 0.1689453125,
+      "learning_rate": 9.34934934934935e-06,
+      "loss": 1.2093,
+      "step": 65
+    },
+    {
+      "epoch": 0.1981981981981982,
+      "grad_norm": 0.193359375,
+      "learning_rate": 9.339339339339341e-06,
+      "loss": 1.2606,
+      "step": 66
+    },
+    {
+      "epoch": 0.2012012012012012,
+      "grad_norm": 0.181640625,
+      "learning_rate": 9.32932932932933e-06,
+      "loss": 1.2633,
+      "step": 67
+    },
+    {
+      "epoch": 0.2042042042042042,
+      "grad_norm": 0.1650390625,
+      "learning_rate": 9.31931931931932e-06,
+      "loss": 1.2949,
+      "step": 68
+    },
+    {
+      "epoch": 0.2072072072072072,
+      "grad_norm": 0.1474609375,
+      "learning_rate": 9.30930930930931e-06,
+      "loss": 1.2058,
+      "step": 69
+    },
+    {
+      "epoch": 0.21021021021021022,
+      "grad_norm": 0.1767578125,
+      "learning_rate": 9.2992992992993e-06,
+      "loss": 1.2525,
+      "step": 70
+    },
+    {
+      "epoch": 0.2132132132132132,
+      "grad_norm": 0.177734375,
+      "learning_rate": 9.289289289289291e-06,
+      "loss": 1.1823,
+      "step": 71
+    },
+    {
+      "epoch": 0.21621621621621623,
+      "grad_norm": 0.1806640625,
+      "learning_rate": 9.27927927927928e-06,
+      "loss": 1.3055,
+      "step": 72
+    },
+    {
+      "epoch": 0.21921921921921922,
+      "grad_norm": 0.15625,
+      "learning_rate": 9.26926926926927e-06,
+      "loss": 1.2541,
+      "step": 73
+    },
+    {
+      "epoch": 0.2222222222222222,
+      "grad_norm": 0.1572265625,
+      "learning_rate": 9.25925925925926e-06,
+      "loss": 1.1968,
+      "step": 74
+    },
+    {
+      "epoch": 0.22522522522522523,
+      "grad_norm": 0.173828125,
+      "learning_rate": 9.24924924924925e-06,
+      "loss": 1.2326,
+      "step": 75
+    },
+    {
+      "epoch": 0.22822822822822822,
+      "grad_norm": 0.173828125,
+      "learning_rate": 9.239239239239241e-06,
+      "loss": 1.2638,
+      "step": 76
+    },
+    {
+      "epoch": 0.23123123123123124,
+      "grad_norm": 0.216796875,
+      "learning_rate": 9.229229229229229e-06,
+      "loss": 1.249,
+      "step": 77
+    },
+    {
+      "epoch": 0.23423423423423423,
+      "grad_norm": 0.16796875,
+      "learning_rate": 9.21921921921922e-06,
+      "loss": 1.2606,
+      "step": 78
+    },
+    {
+      "epoch": 0.23723723723723725,
+      "grad_norm": 0.1591796875,
+      "learning_rate": 9.20920920920921e-06,
+      "loss": 1.2457,
+      "step": 79
+    },
+    {
+      "epoch": 0.24024024024024024,
+      "grad_norm": 0.1572265625,
+      "learning_rate": 9.1991991991992e-06,
+      "loss": 1.1868,
+      "step": 80
+    },
+    {
+      "epoch": 0.24324324324324326,
+      "grad_norm": 0.33203125,
+      "learning_rate": 9.189189189189191e-06,
+      "loss": 1.206,
+      "step": 81
+    },
+    {
+      "epoch": 0.24624624624624625,
+      "grad_norm": 0.1669921875,
+      "learning_rate": 9.179179179179179e-06,
+      "loss": 1.2223,
+      "step": 82
+    },
+    {
+      "epoch": 0.24924924924924924,
+      "grad_norm": 0.1796875,
+      "learning_rate": 9.16916916916917e-06,
+      "loss": 1.2091,
+      "step": 83
+    },
+    {
+      "epoch": 0.25225225225225223,
+      "grad_norm": 0.1533203125,
+      "learning_rate": 9.15915915915916e-06,
+      "loss": 1.1928,
+      "step": 84
+    },
+    {
+      "epoch": 0.2552552552552553,
+      "grad_norm": 0.376953125,
+      "learning_rate": 9.14914914914915e-06,
+      "loss": 1.1768,
+      "step": 85
+    },
+    {
+      "epoch": 0.25825825825825827,
+      "grad_norm": 0.1533203125,
+      "learning_rate": 9.13913913913914e-06,
+      "loss": 1.1964,
+      "step": 86
+    },
+    {
+      "epoch": 0.26126126126126126,
+      "grad_norm": 0.1611328125,
+      "learning_rate": 9.129129129129129e-06,
+      "loss": 1.1781,
+      "step": 87
+    },
+    {
+      "epoch": 0.26426426426426425,
+      "grad_norm": 0.15234375,
+      "learning_rate": 9.11911911911912e-06,
+      "loss": 1.226,
+      "step": 88
+    },
+    {
+      "epoch": 0.2672672672672673,
+      "grad_norm": 0.1748046875,
+      "learning_rate": 9.10910910910911e-06,
+      "loss": 1.1588,
+      "step": 89
+    },
+    {
+      "epoch": 0.2702702702702703,
+      "grad_norm": 0.17578125,
+      "learning_rate": 9.0990990990991e-06,
+      "loss": 1.1871,
+      "step": 90
+    },
+    {
+      "epoch": 0.2732732732732733,
+      "grad_norm": 0.15234375,
+      "learning_rate": 9.08908908908909e-06,
+      "loss": 1.1918,
+      "step": 91
+    },
+    {
+      "epoch": 0.27627627627627627,
+      "grad_norm": 0.189453125,
+      "learning_rate": 9.079079079079079e-06,
+      "loss": 1.1998,
+      "step": 92
+    },
+    {
+      "epoch": 0.27927927927927926,
+      "grad_norm": 0.15234375,
+      "learning_rate": 9.06906906906907e-06,
+      "loss": 1.1627,
+      "step": 93
+    },
+    {
+      "epoch": 0.2822822822822823,
+      "grad_norm": 0.1494140625,
+      "learning_rate": 9.05905905905906e-06,
+      "loss": 1.1679,
+      "step": 94
+    },
+    {
+      "epoch": 0.2852852852852853,
+      "grad_norm": 0.1806640625,
+      "learning_rate": 9.04904904904905e-06,
+      "loss": 1.1452,
+      "step": 95
+    },
+    {
+      "epoch": 0.2882882882882883,
+      "grad_norm": 0.16015625,
+      "learning_rate": 9.03903903903904e-06,
+      "loss": 1.1469,
+      "step": 96
+    },
+    {
+      "epoch": 0.2912912912912913,
+      "grad_norm": 0.193359375,
+      "learning_rate": 9.029029029029029e-06,
+      "loss": 1.1455,
+      "step": 97
+    },
+    {
+      "epoch": 0.29429429429429427,
+      "grad_norm": 0.1591796875,
+      "learning_rate": 9.01901901901902e-06,
+      "loss": 1.223,
+      "step": 98
+    },
+    {
+      "epoch": 0.2972972972972973,
+      "grad_norm": 0.2138671875,
+      "learning_rate": 9.00900900900901e-06,
+      "loss": 1.2792,
+      "step": 99
+    },
+    {
+      "epoch": 0.3003003003003003,
+      "grad_norm": 0.20703125,
+      "learning_rate": 8.998998998999e-06,
+      "loss": 1.1862,
+      "step": 100
+    },
+    {
+      "epoch": 0.3033033033033033,
+      "grad_norm": 0.17578125,
+      "learning_rate": 8.98898898898899e-06,
+      "loss": 1.1485,
+      "step": 101
+    },
+    {
+      "epoch": 0.3063063063063063,
+      "grad_norm": 0.166015625,
+      "learning_rate": 8.97897897897898e-06,
+      "loss": 1.184,
+      "step": 102
+    },
+    {
+      "epoch": 0.30930930930930933,
+      "grad_norm": 0.158203125,
+      "learning_rate": 8.96896896896897e-06,
+      "loss": 1.2003,
+      "step": 103
+    },
+    {
+      "epoch": 0.3123123123123123,
+      "grad_norm": 0.1640625,
+      "learning_rate": 8.95895895895896e-06,
+      "loss": 1.1451,
+      "step": 104
+    },
+    {
+      "epoch": 0.3153153153153153,
+      "grad_norm": 0.2109375,
+      "learning_rate": 8.94894894894895e-06,
+      "loss": 1.1214,
+      "step": 105
+    },
+    {
+      "epoch": 0.3183183183183183,
+      "grad_norm": 0.1650390625,
+      "learning_rate": 8.93893893893894e-06,
+      "loss": 1.1792,
+      "step": 106
+    },
+    {
+      "epoch": 0.3213213213213213,
+      "grad_norm": 0.1767578125,
+      "learning_rate": 8.92892892892893e-06,
+      "loss": 1.1123,
+      "step": 107
+    },
+    {
+      "epoch": 0.32432432432432434,
+      "grad_norm": 0.16015625,
+      "learning_rate": 8.91891891891892e-06,
+      "loss": 1.1537,
+      "step": 108
+    },
+    {
+      "epoch": 0.32732732732732733,
+      "grad_norm": 0.1689453125,
+      "learning_rate": 8.90890890890891e-06,
+      "loss": 1.164,
+      "step": 109
+    },
+    {
+      "epoch": 0.3303303303303303,
+      "grad_norm": 0.16796875,
+      "learning_rate": 8.8988988988989e-06,
+      "loss": 1.1731,
+      "step": 110
+    },
+    {
+      "epoch": 0.3333333333333333,
+      "grad_norm": 0.18359375,
+      "learning_rate": 8.888888888888888e-06,
+      "loss": 1.1784,
+      "step": 111
+    },
+    {
+      "epoch": 0.33633633633633636,
+      "grad_norm": 0.171875,
+      "learning_rate": 8.87887887887888e-06,
+      "loss": 1.1335,
+      "step": 112
+    },
+    {
+      "epoch": 0.33933933933933935,
+      "grad_norm": 0.162109375,
+      "learning_rate": 8.86886886886887e-06,
+      "loss": 1.1386,
+      "step": 113
+    },
+    {
+      "epoch": 0.34234234234234234,
+      "grad_norm": 0.16015625,
+      "learning_rate": 8.85885885885886e-06,
+      "loss": 1.137,
+      "step": 114
+    },
+    {
+      "epoch": 0.34534534534534533,
+      "grad_norm": 0.1611328125,
+      "learning_rate": 8.84884884884885e-06,
+      "loss": 1.1727,
+      "step": 115
+    },
+    {
+      "epoch": 0.3483483483483483,
+      "grad_norm": 0.15625,
+      "learning_rate": 8.838838838838838e-06,
+      "loss": 1.1203,
+      "step": 116
+    },
+    {
+      "epoch": 0.35135135135135137,
+      "grad_norm": 0.267578125,
+      "learning_rate": 8.82882882882883e-06,
+      "loss": 1.1205,
+      "step": 117
+    },
+    {
+      "epoch": 0.35435435435435436,
+      "grad_norm": 0.16015625,
+      "learning_rate": 8.818818818818819e-06,
+      "loss": 1.1002,
+      "step": 118
+    },
+    {
+      "epoch": 0.35735735735735735,
+      "grad_norm": 0.224609375,
+      "learning_rate": 8.80880880880881e-06,
+      "loss": 1.0944,
+      "step": 119
+    },
+    {
+      "epoch": 0.36036036036036034,
+      "grad_norm": 0.1630859375,
+      "learning_rate": 8.798798798798799e-06,
+      "loss": 1.1291,
+      "step": 120
+    },
+    {
+      "epoch": 0.3633633633633634,
+      "grad_norm": 0.18359375,
+      "learning_rate": 8.788788788788788e-06,
+      "loss": 1.1387,
+      "step": 121
+    },
+    {
+      "epoch": 0.3663663663663664,
+      "grad_norm": 0.193359375,
+      "learning_rate": 8.77877877877878e-06,
+      "loss": 1.191,
+      "step": 122
+    },
+    {
+      "epoch": 0.36936936936936937,
+      "grad_norm": 0.189453125,
+      "learning_rate": 8.768768768768769e-06,
+      "loss": 1.1816,
+      "step": 123
+    },
+    {
+      "epoch": 0.37237237237237236,
+      "grad_norm": 0.1728515625,
+      "learning_rate": 8.75875875875876e-06,
+      "loss": 1.1521,
+      "step": 124
+    },
+    {
+      "epoch": 0.37537537537537535,
+      "grad_norm": 0.1689453125,
+      "learning_rate": 8.74874874874875e-06,
+      "loss": 1.1581,
+      "step": 125
+    },
+    {
+      "epoch": 0.3783783783783784,
+      "grad_norm": 0.1630859375,
+      "learning_rate": 8.738738738738739e-06,
+      "loss": 1.1394,
+      "step": 126
+    },
+    {
+      "epoch": 0.3813813813813814,
+      "grad_norm": 0.181640625,
+      "learning_rate": 8.72872872872873e-06,
+      "loss": 1.1064,
+      "step": 127
+    },
+    {
+      "epoch": 0.3843843843843844,
+      "grad_norm": 0.166015625,
+      "learning_rate": 8.718718718718719e-06,
+      "loss": 1.1951,
+      "step": 128
+    },
+    {
+      "epoch": 0.38738738738738737,
+      "grad_norm": 0.15625,
+      "learning_rate": 8.70870870870871e-06,
+      "loss": 1.1213,
+      "step": 129
+    },
+    {
+      "epoch": 0.39039039039039036,
+      "grad_norm": 0.2421875,
+      "learning_rate": 8.6986986986987e-06,
+      "loss": 1.2004,
+      "step": 130
+    },
+    {
+      "epoch": 0.3933933933933934,
+      "grad_norm": 0.2080078125,
+      "learning_rate": 8.688688688688689e-06,
+      "loss": 1.0872,
+      "step": 131
+    },
+    {
+      "epoch": 0.3963963963963964,
+      "grad_norm": 0.1845703125,
+      "learning_rate": 8.67867867867868e-06,
+      "loss": 1.1877,
+      "step": 132
+    },
+    {
+      "epoch": 0.3993993993993994,
+      "grad_norm": 0.16796875,
+      "learning_rate": 8.668668668668669e-06,
+      "loss": 1.1522,
+      "step": 133
+    },
+    {
+      "epoch": 0.4024024024024024,
+      "grad_norm": 0.169921875,
+      "learning_rate": 8.65865865865866e-06,
+      "loss": 1.1393,
+      "step": 134
+    },
+    {
+      "epoch": 0.40540540540540543,
+      "grad_norm": 0.2021484375,
+      "learning_rate": 8.64864864864865e-06,
+      "loss": 1.0876,
+      "step": 135
+    },
+    {
+      "epoch": 0.4084084084084084,
+      "grad_norm": 0.1630859375,
+      "learning_rate": 8.638638638638639e-06,
+      "loss": 1.1139,
+      "step": 136
+    },
+    {
+      "epoch": 0.4114114114114114,
+      "grad_norm": 0.1962890625,
+      "learning_rate": 8.62862862862863e-06,
+      "loss": 1.1235,
+      "step": 137
+    },
+    {
+      "epoch": 0.4144144144144144,
+      "grad_norm": 0.169921875,
+      "learning_rate": 8.618618618618619e-06,
+      "loss": 1.1276,
+      "step": 138
+    },
+    {
+      "epoch": 0.4174174174174174,
+      "grad_norm": 0.1728515625,
+      "learning_rate": 8.60860860860861e-06,
+      "loss": 1.1457,
+      "step": 139
+    },
+    {
+      "epoch": 0.42042042042042044,
+      "grad_norm": 0.1630859375,
+      "learning_rate": 8.5985985985986e-06,
+      "loss": 1.0484,
+      "step": 140
+    },
+    {
+      "epoch": 0.42342342342342343,
+      "grad_norm": 0.169921875,
+      "learning_rate": 8.588588588588589e-06,
+      "loss": 1.1576,
+      "step": 141
+    },
+    {
+      "epoch": 0.4264264264264264,
+      "grad_norm": 0.2099609375,
+      "learning_rate": 8.57857857857858e-06,
+      "loss": 1.1088,
+      "step": 142
+    },
+    {
+      "epoch": 0.4294294294294294,
+      "grad_norm": 0.193359375,
+      "learning_rate": 8.568568568568569e-06,
+      "loss": 1.1248,
+      "step": 143
+    },
+    {
+      "epoch": 0.43243243243243246,
+      "grad_norm": 0.1787109375,
+      "learning_rate": 8.55855855855856e-06,
+      "loss": 1.0912,
+      "step": 144
+    },
+    {
+      "epoch": 0.43543543543543545,
+      "grad_norm": 0.2236328125,
+      "learning_rate": 8.54854854854855e-06,
+      "loss": 1.1049,
+      "step": 145
+    },
+    {
+      "epoch": 0.43843843843843844,
+      "grad_norm": 0.1767578125,
+      "learning_rate": 8.538538538538539e-06,
+      "loss": 1.0877,
+      "step": 146
+    },
+    {
+      "epoch": 0.44144144144144143,
+      "grad_norm": 0.275390625,
+      "learning_rate": 8.52852852852853e-06,
+      "loss": 1.1453,
+      "step": 147
+    },
+    {
+      "epoch": 0.4444444444444444,
+      "grad_norm": 0.1875,
+      "learning_rate": 8.518518518518519e-06,
+      "loss": 1.1215,
+      "step": 148
+    },
+    {
+      "epoch": 0.44744744744744747,
+      "grad_norm": 0.17578125,
+      "learning_rate": 8.50850850850851e-06,
+      "loss": 1.105,
+      "step": 149
+    },
+    {
+      "epoch": 0.45045045045045046,
+      "grad_norm": 0.17578125,
+      "learning_rate": 8.4984984984985e-06,
+      "loss": 1.1493,
+      "step": 150
+    },
+    {
+      "epoch": 0.45345345345345345,
+      "grad_norm": 0.16015625,
+      "learning_rate": 8.488488488488489e-06,
+      "loss": 1.0728,
+      "step": 151
+    },
+    {
+      "epoch": 0.45645645645645644,
+      "grad_norm": 0.169921875,
+      "learning_rate": 8.47847847847848e-06,
+      "loss": 1.0975,
+      "step": 152
+    },
+    {
+      "epoch": 0.4594594594594595,
+      "grad_norm": 0.1650390625,
+      "learning_rate": 8.46846846846847e-06,
+      "loss": 1.0759,
+      "step": 153
+    },
+    {
+      "epoch": 0.4624624624624625,
+      "grad_norm": 0.236328125,
+      "learning_rate": 8.45845845845846e-06,
+      "loss": 1.0969,
+      "step": 154
+    },
+    {
+      "epoch": 0.46546546546546547,
+      "grad_norm": 0.1767578125,
+      "learning_rate": 8.44844844844845e-06,
+      "loss": 1.1365,
+      "step": 155
+    },
+    {
+      "epoch": 0.46846846846846846,
+      "grad_norm": 0.1728515625,
+      "learning_rate": 8.438438438438439e-06,
+      "loss": 1.021,
+      "step": 156
+    },
+    {
+      "epoch": 0.47147147147147145,
+      "grad_norm": 0.228515625,
+      "learning_rate": 8.428428428428428e-06,
+      "loss": 1.1151,
+      "step": 157
+    },
+    {
+      "epoch": 0.4744744744744745,
+      "grad_norm": 0.1728515625,
+      "learning_rate": 8.41841841841842e-06,
+      "loss": 1.081,
+      "step": 158
+    },
+    {
+      "epoch": 0.4774774774774775,
+      "grad_norm": 0.189453125,
+      "learning_rate": 8.408408408408409e-06,
+      "loss": 1.121,
+      "step": 159
+    },
+    {
+      "epoch": 0.4804804804804805,
+      "grad_norm": 0.1865234375,
+      "learning_rate": 8.398398398398398e-06,
+      "loss": 1.0411,
+      "step": 160
+    },
+    {
+      "epoch": 0.48348348348348347,
+      "grad_norm": 0.1650390625,
+      "learning_rate": 8.388388388388389e-06,
+      "loss": 1.054,
+      "step": 161
+    },
+    {
+      "epoch": 0.4864864864864865,
+      "grad_norm": 0.1865234375,
+      "learning_rate": 8.378378378378378e-06,
+      "loss": 1.1347,
+      "step": 162
+    },
+    {
+      "epoch": 0.4894894894894895,
+      "grad_norm": 0.1884765625,
+      "learning_rate": 8.36836836836837e-06,
+      "loss": 1.1171,
+      "step": 163
+    },
+    {
+      "epoch": 0.4924924924924925,
+      "grad_norm": 0.1796875,
+      "learning_rate": 8.358358358358359e-06,
+      "loss": 1.0354,
+      "step": 164
+    },
+    {
+      "epoch": 0.4954954954954955,
+      "grad_norm": 0.166015625,
+      "learning_rate": 8.348348348348348e-06,
+      "loss": 1.0594,
+      "step": 165
+    },
+    {
+      "epoch": 0.4984984984984985,
+      "grad_norm": 0.169921875,
+      "learning_rate": 8.338338338338339e-06,
+      "loss": 1.0887,
+      "step": 166
+    },
+    {
+      "epoch": 0.5015015015015015,
+      "grad_norm": 0.185546875,
+      "learning_rate": 8.328328328328328e-06,
+      "loss": 1.0816,
+      "step": 167
+    },
+    {
+      "epoch": 0.5045045045045045,
+      "grad_norm": 0.30078125,
+      "learning_rate": 8.31831831831832e-06,
+      "loss": 1.05,
+      "step": 168
+    },
+    {
+      "epoch": 0.5075075075075075,
+      "grad_norm": 0.2275390625,
+      "learning_rate": 8.308308308308309e-06,
+      "loss": 1.0499,
+      "step": 169
+    },
+    {
+      "epoch": 0.5105105105105106,
+      "grad_norm": 0.1826171875,
+      "learning_rate": 8.298298298298298e-06,
+      "loss": 1.0935,
+      "step": 170
+    },
+    {
+      "epoch": 0.5135135135135135,
+      "grad_norm": 0.2060546875,
+      "learning_rate": 8.288288288288289e-06,
+      "loss": 1.0311,
+      "step": 171
+    },
+    {
+      "epoch": 0.5165165165165165,
+      "grad_norm": 0.181640625,
+      "learning_rate": 8.278278278278278e-06,
+      "loss": 1.0845,
+      "step": 172
+    },
+    {
+      "epoch": 0.5195195195195195,
+      "grad_norm": 0.1865234375,
+      "learning_rate": 8.26826826826827e-06,
+      "loss": 1.0856,
+      "step": 173
+    },
+    {
+      "epoch": 0.5225225225225225,
+      "grad_norm": 0.2138671875,
+      "learning_rate": 8.258258258258259e-06,
+      "loss": 1.1044,
+      "step": 174
+    },
+    {
+      "epoch": 0.5255255255255256,
+      "grad_norm": 0.1826171875,
+      "learning_rate": 8.248248248248248e-06,
+      "loss": 0.9977,
+      "step": 175
+    },
+    {
+      "epoch": 0.5285285285285285,
+      "grad_norm": 0.169921875,
+      "learning_rate": 8.23823823823824e-06,
+      "loss": 1.069,
+      "step": 176
+    },
+    {
+      "epoch": 0.5315315315315315,
+      "grad_norm": 0.1875,
+      "learning_rate": 8.228228228228229e-06,
+      "loss": 1.0809,
+      "step": 177
+    },
+    {
+      "epoch": 0.5345345345345346,
+      "grad_norm": 0.25390625,
+      "learning_rate": 8.21821821821822e-06,
+      "loss": 1.0526,
+      "step": 178
+    },
+    {
+      "epoch": 0.5375375375375375,
+      "grad_norm": 0.1845703125,
+      "learning_rate": 8.208208208208209e-06,
+      "loss": 1.0714,
+      "step": 179
+    },
+    {
+      "epoch": 0.5405405405405406,
+      "grad_norm": 0.173828125,
+      "learning_rate": 8.198198198198198e-06,
+      "loss": 1.0111,
+      "step": 180
+    },
+    {
+      "epoch": 0.5435435435435435,
+      "grad_norm": 0.1982421875,
+      "learning_rate": 8.18818818818819e-06,
+      "loss": 1.1091,
+      "step": 181
+    },
+    {
+      "epoch": 0.5465465465465466,
+      "grad_norm": 0.1669921875,
+      "learning_rate": 8.178178178178179e-06,
+      "loss": 1.0574,
+      "step": 182
+    },
+    {
+      "epoch": 0.5495495495495496,
+      "grad_norm": 0.271484375,
+      "learning_rate": 8.16816816816817e-06,
+      "loss": 1.0202,
+      "step": 183
+    },
+    {
+      "epoch": 0.5525525525525525,
+      "grad_norm": 0.1748046875,
+      "learning_rate": 8.158158158158159e-06,
+      "loss": 1.0849,
+      "step": 184
+    },
+    {
+      "epoch": 0.5555555555555556,
+      "grad_norm": 0.17578125,
+      "learning_rate": 8.148148148148148e-06,
+      "loss": 1.0444,
+      "step": 185
+    },
+    {
+      "epoch": 0.5585585585585585,
+      "grad_norm": 0.17578125,
+      "learning_rate": 8.13813813813814e-06,
+      "loss": 1.0647,
+      "step": 186
+    },
+    {
+      "epoch": 0.5615615615615616,
+      "grad_norm": 0.1787109375,
+      "learning_rate": 8.128128128128129e-06,
+      "loss": 1.0746,
+      "step": 187
+    },
+    {
+      "epoch": 0.5645645645645646,
+      "grad_norm": 0.251953125,
+      "learning_rate": 8.11811811811812e-06,
+      "loss": 1.1021,
+      "step": 188
+    },
+    {
+      "epoch": 0.5675675675675675,
+      "grad_norm": 0.212890625,
+      "learning_rate": 8.108108108108109e-06,
+      "loss": 1.1208,
+      "step": 189
+    },
+    {
+      "epoch": 0.5705705705705706,
+      "grad_norm": 0.181640625,
+      "learning_rate": 8.098098098098098e-06,
+      "loss": 1.0423,
+      "step": 190
+    },
+    {
+      "epoch": 0.5735735735735735,
+      "grad_norm": 0.1748046875,
+      "learning_rate": 8.088088088088088e-06,
+      "loss": 1.1179,
+      "step": 191
+    },
+    {
+      "epoch": 0.5765765765765766,
+      "grad_norm": 0.193359375,
+      "learning_rate": 8.078078078078079e-06,
+      "loss": 1.0608,
+      "step": 192
+    },
+    {
+      "epoch": 0.5795795795795796,
+      "grad_norm": 0.2080078125,
+      "learning_rate": 8.06806806806807e-06,
+      "loss": 1.0238,
+      "step": 193
+    },
+    {
+      "epoch": 0.5825825825825826,
+      "grad_norm": 0.1875,
+      "learning_rate": 8.058058058058059e-06,
+      "loss": 1.0827,
+      "step": 194
+    },
+    {
+      "epoch": 0.5855855855855856,
+      "grad_norm": 0.181640625,
+      "learning_rate": 8.048048048048048e-06,
+      "loss": 1.061,
+      "step": 195
+    },
+    {
+      "epoch": 0.5885885885885885,
+      "grad_norm": 0.197265625,
+      "learning_rate": 8.038038038038038e-06,
+      "loss": 1.0505,
+      "step": 196
+    },
+    {
+      "epoch": 0.5915915915915916,
+      "grad_norm": 0.2412109375,
+      "learning_rate": 8.028028028028029e-06,
+      "loss": 0.9871,
+      "step": 197
+    },
+    {
+      "epoch": 0.5945945945945946,
+      "grad_norm": 0.2392578125,
+      "learning_rate": 8.018018018018018e-06,
+      "loss": 1.103,
+      "step": 198
+    },
+    {
+      "epoch": 0.5975975975975976,
+      "grad_norm": 0.1904296875,
+      "learning_rate": 8.00800800800801e-06,
+      "loss": 1.0735,
+      "step": 199
+    },
+    {
+      "epoch": 0.6006006006006006,
+      "grad_norm": 0.267578125,
+      "learning_rate": 7.997997997997999e-06,
+      "loss": 1.1141,
+      "step": 200
+    },
+    {
+      "epoch": 0.6036036036036037,
+      "grad_norm": 0.173828125,
+      "learning_rate": 7.987987987987988e-06,
+      "loss": 1.0269,
+      "step": 201
+    },
+    {
+      "epoch": 0.6066066066066066,
+      "grad_norm": 0.259765625,
+      "learning_rate": 7.977977977977979e-06,
+      "loss": 1.1204,
+      "step": 202
+    },
+    {
+      "epoch": 0.6096096096096096,
+      "grad_norm": 0.171875,
+      "learning_rate": 7.967967967967968e-06,
+      "loss": 1.0658,
+      "step": 203
+    },
+    {
+      "epoch": 0.6126126126126126,
+      "grad_norm": 0.255859375,
+      "learning_rate": 7.95795795795796e-06,
+      "loss": 1.0282,
+      "step": 204
+    },
+    {
+      "epoch": 0.6156156156156156,
+      "grad_norm": 0.2158203125,
+      "learning_rate": 7.947947947947949e-06,
+      "loss": 1.0413,
+      "step": 205
+    },
+    {
+      "epoch": 0.6186186186186187,
+      "grad_norm": 0.1806640625,
+      "learning_rate": 7.937937937937938e-06,
+      "loss": 1.037,
+      "step": 206
+    },
+    {
+      "epoch": 0.6216216216216216,
+      "grad_norm": 0.19140625,
+      "learning_rate": 7.927927927927929e-06,
+      "loss": 1.0571,
+      "step": 207
+    },
+    {
+      "epoch": 0.6246246246246246,
+      "grad_norm": 0.1787109375,
+      "learning_rate": 7.917917917917918e-06,
+      "loss": 1.0461,
+      "step": 208
+    },
+    {
+      "epoch": 0.6276276276276276,
+      "grad_norm": 0.19140625,
+      "learning_rate": 7.90790790790791e-06,
+      "loss": 1.0284,
+      "step": 209
+    },
+    {
+      "epoch": 0.6306306306306306,
+      "grad_norm": 0.189453125,
+      "learning_rate": 7.897897897897899e-06,
+      "loss": 1.0301,
+      "step": 210
+    },
+    {
+      "epoch": 0.6336336336336337,
+      "grad_norm": 0.314453125,
+      "learning_rate": 7.887887887887888e-06,
+      "loss": 1.0039,
+      "step": 211
+    },
+    {
+      "epoch": 0.6366366366366366,
+      "grad_norm": 0.18359375,
+      "learning_rate": 7.877877877877879e-06,
+      "loss": 1.0684,
+      "step": 212
+    },
+    {
+      "epoch": 0.6396396396396397,
+      "grad_norm": 0.44921875,
+      "learning_rate": 7.867867867867868e-06,
+      "loss": 1.0595,
+      "step": 213
+    },
+    {
+      "epoch": 0.6426426426426426,
+      "grad_norm": 0.267578125,
+      "learning_rate": 7.85785785785786e-06,
+      "loss": 1.0739,
+      "step": 214
+    },
+    {
+      "epoch": 0.6456456456456456,
+      "grad_norm": 0.2099609375,
+      "learning_rate": 7.847847847847849e-06,
+      "loss": 1.0644,
+      "step": 215
+    },
+    {
+      "epoch": 0.6486486486486487,
+      "grad_norm": 0.2021484375,
+      "learning_rate": 7.837837837837838e-06,
+      "loss": 1.0784,
+      "step": 216
+    },
+    {
+      "epoch": 0.6516516516516516,
+      "grad_norm": 0.2353515625,
+      "learning_rate": 7.827827827827829e-06,
+      "loss": 1.1088,
+      "step": 217
+    },
+    {
+      "epoch": 0.6546546546546547,
+      "grad_norm": 0.2138671875,
+      "learning_rate": 7.817817817817818e-06,
+      "loss": 1.1102,
+      "step": 218
+    },
+    {
+      "epoch": 0.6576576576576577,
+      "grad_norm": 0.2060546875,
+      "learning_rate": 7.807807807807808e-06,
+      "loss": 1.0586,
+      "step": 219
+    },
+    {
+      "epoch": 0.6606606606606606,
+      "grad_norm": 0.1923828125,
+      "learning_rate": 7.797797797797799e-06,
+      "loss": 1.0627,
+      "step": 220
+    },
+    {
+      "epoch": 0.6636636636636637,
+      "grad_norm": 0.189453125,
+      "learning_rate": 7.787787787787788e-06,
+      "loss": 1.0713,
+      "step": 221
+    },
+    {
+      "epoch": 0.6666666666666666,
+      "grad_norm": 0.25390625,
+      "learning_rate": 7.77777777777778e-06,
+      "loss": 1.0472,
+      "step": 222
+    },
+    {
+      "epoch": 0.6696696696696697,
+      "grad_norm": 0.208984375,
+      "learning_rate": 7.767767767767769e-06,
+      "loss": 1.0131,
+      "step": 223
+    },
+    {
+      "epoch": 0.6726726726726727,
+      "grad_norm": 0.1796875,
+      "learning_rate": 7.757757757757758e-06,
+      "loss": 1.0199,
+      "step": 224
+    },
+    {
+      "epoch": 0.6756756756756757,
+      "grad_norm": 0.19140625,
+      "learning_rate": 7.747747747747749e-06,
+      "loss": 1.04,
+      "step": 225
+    },
+    {
+      "epoch": 0.6786786786786787,
+      "grad_norm": 0.22265625,
+      "learning_rate": 7.737737737737738e-06,
+      "loss": 1.0174,
+      "step": 226
+    },
+    {
+      "epoch": 0.6816816816816816,
+      "grad_norm": 0.1904296875,
+      "learning_rate": 7.72772772772773e-06,
+      "loss": 1.0253,
+      "step": 227
+    },
+    {
+      "epoch": 0.6846846846846847,
+      "grad_norm": 0.1953125,
+      "learning_rate": 7.717717717717719e-06,
+      "loss": 1.0346,
+      "step": 228
+    },
+    {
+      "epoch": 0.6876876876876877,
+      "grad_norm": 0.185546875,
+      "learning_rate": 7.707707707707708e-06,
+      "loss": 1.0772,
+      "step": 229
+    },
+    {
+      "epoch": 0.6906906906906907,
+      "grad_norm": 0.189453125,
+      "learning_rate": 7.697697697697697e-06,
+      "loss": 1.0482,
+      "step": 230
+    },
+    {
+      "epoch": 0.6936936936936937,
+      "grad_norm": 0.2041015625,
+      "learning_rate": 7.687687687687688e-06,
+      "loss": 1.117,
+      "step": 231
+    },
+    {
+      "epoch": 0.6966966966966966,
+      "grad_norm": 0.17578125,
+      "learning_rate": 7.67767767767768e-06,
+      "loss": 1.0211,
+      "step": 232
+    },
+    {
+      "epoch": 0.6996996996996997,
+      "grad_norm": 0.19140625,
+      "learning_rate": 7.667667667667669e-06,
+      "loss": 1.0425,
+      "step": 233
+    },
+    {
+      "epoch": 0.7027027027027027,
+      "grad_norm": 0.1923828125,
+      "learning_rate": 7.657657657657658e-06,
+      "loss": 1.0539,
+      "step": 234
+    },
+    {
+      "epoch": 0.7057057057057057,
+      "grad_norm": 0.2119140625,
+      "learning_rate": 7.647647647647647e-06,
+      "loss": 1.0409,
+      "step": 235
+    },
+    {
+      "epoch": 0.7087087087087087,
+      "grad_norm": 0.23046875,
+      "learning_rate": 7.637637637637638e-06,
+      "loss": 1.043,
+      "step": 236
+    },
+    {
+      "epoch": 0.7117117117117117,
+      "grad_norm": 0.2314453125,
+      "learning_rate": 7.6276276276276285e-06,
+      "loss": 1.0784,
+      "step": 237
+    },
+    {
+      "epoch": 0.7147147147147147,
+      "grad_norm": 0.1884765625,
+      "learning_rate": 7.617617617617619e-06,
+      "loss": 1.0248,
+      "step": 238
+    },
+    {
+      "epoch": 0.7177177177177178,
+      "grad_norm": 0.1943359375,
+      "learning_rate": 7.607607607607608e-06,
+      "loss": 1.03,
+      "step": 239
+    },
+    {
+      "epoch": 0.7207207207207207,
+      "grad_norm": 0.19140625,
+      "learning_rate": 7.597597597597598e-06,
+      "loss": 0.9948,
+      "step": 240
+    },
+    {
+      "epoch": 0.7237237237237237,
+      "grad_norm": 0.212890625,
+      "learning_rate": 7.587587587587588e-06,
+      "loss": 1.018,
+      "step": 241
+    },
+    {
+      "epoch": 0.7267267267267268,
+      "grad_norm": 0.2421875,
+      "learning_rate": 7.577577577577579e-06,
+      "loss": 1.0235,
+      "step": 242
+    },
+    {
+      "epoch": 0.7297297297297297,
+      "grad_norm": 0.2333984375,
+      "learning_rate": 7.567567567567569e-06,
+      "loss": 1.0374,
+      "step": 243
+    },
+    {
+      "epoch": 0.7327327327327328,
+      "grad_norm": 0.48046875,
+      "learning_rate": 7.557557557557558e-06,
+      "loss": 1.04,
+      "step": 244
+    },
+    {
+      "epoch": 0.7357357357357357,
+      "grad_norm": 0.2578125,
+      "learning_rate": 7.547547547547548e-06,
+      "loss": 0.9977,
+      "step": 245
+    },
+    {
+      "epoch": 0.7387387387387387,
+      "grad_norm": 0.189453125,
+      "learning_rate": 7.5375375375375385e-06,
+      "loss": 1.027,
+      "step": 246
+    },
+    {
+      "epoch": 0.7417417417417418,
+      "grad_norm": 0.25390625,
+      "learning_rate": 7.527527527527529e-06,
+      "loss": 1.0186,
+      "step": 247
+    },
+    {
+      "epoch": 0.7447447447447447,
+      "grad_norm": 0.212890625,
+      "learning_rate": 7.517517517517519e-06,
+      "loss": 1.0176,
+      "step": 248
+    },
+    {
+      "epoch": 0.7477477477477478,
+      "grad_norm": 0.193359375,
+      "learning_rate": 7.507507507507507e-06,
+      "loss": 1.0427,
+      "step": 249
+    },
+    {
+      "epoch": 0.7507507507507507,
+      "grad_norm": 0.24609375,
+      "learning_rate": 7.4974974974974975e-06,
+      "loss": 1.0405,
+      "step": 250
+    },
+    {
+      "epoch": 0.7537537537537538,
+      "grad_norm": 0.203125,
+      "learning_rate": 7.487487487487488e-06,
+      "loss": 0.9866,
+      "step": 251
+    },
+    {
+      "epoch": 0.7567567567567568,
+      "grad_norm": 0.1962890625,
+      "learning_rate": 7.477477477477479e-06,
+      "loss": 1.0189,
+      "step": 252
+    },
+    {
+      "epoch": 0.7597597597597597,
+      "grad_norm": 0.205078125,
+      "learning_rate": 7.467467467467469e-06,
+      "loss": 1.0394,
+      "step": 253
+    },
+    {
+      "epoch": 0.7627627627627628,
+      "grad_norm": 0.1806640625,
+      "learning_rate": 7.457457457457457e-06,
+      "loss": 0.9809,
+      "step": 254
+    },
+    {
+      "epoch": 0.7657657657657657,
+      "grad_norm": 0.1953125,
+      "learning_rate": 7.447447447447448e-06,
+      "loss": 1.0057,
+      "step": 255
+    },
+    {
+      "epoch": 0.7687687687687688,
+      "grad_norm": 0.1943359375,
+      "learning_rate": 7.437437437437438e-06,
+      "loss": 1.0275,
+      "step": 256
+    },
+    {
+      "epoch": 0.7717717717717718,
+      "grad_norm": 0.21875,
+      "learning_rate": 7.427427427427428e-06,
+      "loss": 1.038,
+      "step": 257
+    },
+    {
+      "epoch": 0.7747747747747747,
+      "grad_norm": 0.263671875,
+      "learning_rate": 7.417417417417418e-06,
+      "loss": 1.0308,
+      "step": 258
+    },
+    {
+      "epoch": 0.7777777777777778,
+      "grad_norm": 0.1865234375,
+      "learning_rate": 7.4074074074074075e-06,
+      "loss": 0.9862,
+      "step": 259
+    },
+    {
+      "epoch": 0.7807807807807807,
+      "grad_norm": 0.1875,
+      "learning_rate": 7.397397397397398e-06,
+      "loss": 1.0437,
+      "step": 260
+    },
+    {
+      "epoch": 0.7837837837837838,
+      "grad_norm": 0.2265625,
+      "learning_rate": 7.387387387387388e-06,
+      "loss": 1.0832,
+      "step": 261
+    },
+    {
+      "epoch": 0.7867867867867868,
+      "grad_norm": 0.201171875,
+      "learning_rate": 7.377377377377378e-06,
+      "loss": 1.0836,
+      "step": 262
+    },
+    {
+      "epoch": 0.7897897897897898,
+      "grad_norm": 0.23828125,
+      "learning_rate": 7.367367367367368e-06,
+      "loss": 1.0496,
+      "step": 263
+    },
+    {
+      "epoch": 0.7927927927927928,
+      "grad_norm": 0.2138671875,
+      "learning_rate": 7.3573573573573575e-06,
+      "loss": 1.0285,
+      "step": 264
+    },
+    {
+      "epoch": 0.7957957957957958,
+      "grad_norm": 0.2294921875,
+      "learning_rate": 7.347347347347348e-06,
+      "loss": 1.035,
+      "step": 265
+    },
+    {
+      "epoch": 0.7987987987987988,
+      "grad_norm": 0.21875,
+      "learning_rate": 7.337337337337338e-06,
+      "loss": 1.0155,
+      "step": 266
+    },
+    {
+      "epoch": 0.8018018018018018,
+      "grad_norm": 0.2021484375,
+      "learning_rate": 7.327327327327328e-06,
+      "loss": 1.0005,
+      "step": 267
+    },
+    {
+      "epoch": 0.8048048048048048,
+      "grad_norm": 0.27734375,
+      "learning_rate": 7.317317317317318e-06,
+      "loss": 1.0519,
+      "step": 268
+    },
+    {
+      "epoch": 0.8078078078078078,
+      "grad_norm": 0.2138671875,
+      "learning_rate": 7.307307307307308e-06,
+      "loss": 1.0075,
+      "step": 269
+    },
+    {
+      "epoch": 0.8108108108108109,
+      "grad_norm": 0.220703125,
+      "learning_rate": 7.297297297297298e-06,
+      "loss": 1.0231,
+      "step": 270
+    },
+    {
+      "epoch": 0.8138138138138138,
+      "grad_norm": 0.2216796875,
+      "learning_rate": 7.287287287287288e-06,
+      "loss": 1.0124,
+      "step": 271
+    },
+    {
+      "epoch": 0.8168168168168168,
+      "grad_norm": 0.310546875,
+      "learning_rate": 7.277277277277278e-06,
+      "loss": 1.0538,
+      "step": 272
+    },
+    {
+      "epoch": 0.8198198198198198,
+      "grad_norm": 0.1953125,
+      "learning_rate": 7.267267267267268e-06,
+      "loss": 1.0014,
+      "step": 273
+    },
+    {
+      "epoch": 0.8228228228228228,
+      "grad_norm": 0.2080078125,
+      "learning_rate": 7.257257257257258e-06,
+      "loss": 0.9939,
+      "step": 274
+    },
+    {
+      "epoch": 0.8258258258258259,
+      "grad_norm": 0.1982421875,
+      "learning_rate": 7.247247247247248e-06,
+      "loss": 1.0402,
+      "step": 275
+    },
+    {
+      "epoch": 0.8288288288288288,
+      "grad_norm": 0.2119140625,
+      "learning_rate": 7.237237237237238e-06,
+      "loss": 1.0227,
+      "step": 276
+    },
+    {
+      "epoch": 0.8318318318318318,
+      "grad_norm": 0.197265625,
+      "learning_rate": 7.227227227227228e-06,
+      "loss": 1.0187,
+      "step": 277
+    },
+    {
+      "epoch": 0.8348348348348348,
+      "grad_norm": 0.234375,
+      "learning_rate": 7.217217217217218e-06,
+      "loss": 1.0286,
+      "step": 278
+    },
+    {
+      "epoch": 0.8378378378378378,
+      "grad_norm": 0.20703125,
+      "learning_rate": 7.207207207207208e-06,
+      "loss": 1.0365,
+      "step": 279
+    },
+    {
+      "epoch": 0.8408408408408409,
+      "grad_norm": 0.1962890625,
+      "learning_rate": 7.197197197197198e-06,
+      "loss": 1.0134,
+      "step": 280
+    },
+    {
+      "epoch": 0.8438438438438438,
+      "grad_norm": 0.240234375,
+      "learning_rate": 7.187187187187188e-06,
+      "loss": 1.0246,
+      "step": 281
+    },
+    {
+      "epoch": 0.8468468468468469,
+      "grad_norm": 0.197265625,
+      "learning_rate": 7.177177177177178e-06,
+      "loss": 1.0268,
+      "step": 282
+    },
+    {
+      "epoch": 0.8498498498498499,
+      "grad_norm": 0.29296875,
+      "learning_rate": 7.167167167167167e-06,
+      "loss": 1.0321,
+      "step": 283
+    },
+    {
+      "epoch": 0.8528528528528528,
+      "grad_norm": 0.2099609375,
+      "learning_rate": 7.157157157157158e-06,
+      "loss": 0.9869,
+      "step": 284
+    },
+    {
+      "epoch": 0.8558558558558559,
+      "grad_norm": 0.1865234375,
+      "learning_rate": 7.147147147147148e-06,
+      "loss": 1.0049,
+      "step": 285
+    },
+    {
+      "epoch": 0.8588588588588588,
+      "grad_norm": 0.267578125,
+      "learning_rate": 7.137137137137138e-06,
+      "loss": 1.0048,
+      "step": 286
+    },
+    {
+      "epoch": 0.8618618618618619,
+      "grad_norm": 0.2109375,
+      "learning_rate": 7.127127127127128e-06,
+      "loss": 1.0181,
+      "step": 287
+    },
+    {
+      "epoch": 0.8648648648648649,
+      "grad_norm": 0.2109375,
+      "learning_rate": 7.117117117117117e-06,
+      "loss": 1.0484,
+      "step": 288
+    },
+    {
+      "epoch": 0.8678678678678678,
+      "grad_norm": 0.2099609375,
+      "learning_rate": 7.107107107107107e-06,
+      "loss": 1.0126,
+      "step": 289
+    },
+    {
+      "epoch": 0.8708708708708709,
+      "grad_norm": 0.2099609375,
+      "learning_rate": 7.097097097097097e-06,
+      "loss": 0.9785,
+      "step": 290
+    },
+    {
+      "epoch": 0.8738738738738738,
+      "grad_norm": 0.212890625,
+      "learning_rate": 7.087087087087087e-06,
+      "loss": 1.0099,
+      "step": 291
+    },
+    {
+      "epoch": 0.8768768768768769,
+      "grad_norm": 0.19921875,
+      "learning_rate": 7.0770770770770784e-06,
+      "loss": 1.0611,
+      "step": 292
+    },
+    {
+      "epoch": 0.8798798798798799,
+      "grad_norm": 0.2001953125,
+      "learning_rate": 7.067067067067067e-06,
+      "loss": 0.995,
+      "step": 293
+    },
+    {
+      "epoch": 0.8828828828828829,
+      "grad_norm": 0.1806640625,
+      "learning_rate": 7.057057057057057e-06,
+      "loss": 0.9593,
+      "step": 294
+    },
+    {
+      "epoch": 0.8858858858858859,
+      "grad_norm": 0.232421875,
+      "learning_rate": 7.047047047047047e-06,
+      "loss": 0.9441,
+      "step": 295
+    },
+    {
+      "epoch": 0.8888888888888888,
+      "grad_norm": 0.26171875,
+      "learning_rate": 7.0370370370370375e-06,
+      "loss": 0.9935,
+      "step": 296
+    },
+    {
+      "epoch": 0.8918918918918919,
+      "grad_norm": 0.185546875,
+      "learning_rate": 7.027027027027028e-06,
+      "loss": 0.9788,
+      "step": 297
+    },
+    {
+      "epoch": 0.8948948948948949,
+      "grad_norm": 0.2138671875,
+      "learning_rate": 7.017017017017017e-06,
+      "loss": 0.9942,
+      "step": 298
+    },
+    {
+      "epoch": 0.8978978978978979,
+      "grad_norm": 0.1962890625,
+      "learning_rate": 7.007007007007007e-06,
+      "loss": 1.0299,
+      "step": 299
+    },
+    {
+      "epoch": 0.9009009009009009,
+      "grad_norm": 0.2265625,
+      "learning_rate": 6.996996996996997e-06,
+      "loss": 0.9859,
+      "step": 300
+    },
+    {
+      "epoch": 0.9039039039039038,
+      "grad_norm": 0.2216796875,
+      "learning_rate": 6.9869869869869876e-06,
+      "loss": 1.07,
+      "step": 301
+    },
+    {
+      "epoch": 0.9069069069069069,
+      "grad_norm": 0.25,
+      "learning_rate": 6.976976976976978e-06,
+      "loss": 1.0524,
+      "step": 302
+    },
+    {
+      "epoch": 0.9099099099099099,
+      "grad_norm": 0.1982421875,
+      "learning_rate": 6.966966966966967e-06,
+      "loss": 1.0087,
+      "step": 303
+    },
+    {
+      "epoch": 0.9129129129129129,
+      "grad_norm": 0.2041015625,
+      "learning_rate": 6.956956956956957e-06,
+      "loss": 1.002,
+      "step": 304
+    },
+    {
+      "epoch": 0.9159159159159159,
+      "grad_norm": 0.248046875,
+      "learning_rate": 6.9469469469469474e-06,
+      "loss": 1.0169,
+      "step": 305
+    },
+    {
+      "epoch": 0.918918918918919,
+      "grad_norm": 0.2265625,
+      "learning_rate": 6.936936936936938e-06,
+      "loss": 0.9833,
+      "step": 306
+    },
+    {
+      "epoch": 0.9219219219219219,
+      "grad_norm": 0.224609375,
+      "learning_rate": 6.926926926926928e-06,
+      "loss": 0.9473,
+      "step": 307
+    },
+    {
+      "epoch": 0.924924924924925,
+      "grad_norm": 0.189453125,
+      "learning_rate": 6.916916916916917e-06,
+      "loss": 1.0037,
+      "step": 308
+    },
+    {
+      "epoch": 0.9279279279279279,
+      "grad_norm": 0.193359375,
+      "learning_rate": 6.906906906906907e-06,
+      "loss": 1.0052,
+      "step": 309
+    },
+    {
+      "epoch": 0.9309309309309309,
+      "grad_norm": 0.1962890625,
+      "learning_rate": 6.8968968968968975e-06,
+      "loss": 1.0066,
+      "step": 310
+    },
+    {
+      "epoch": 0.933933933933934,
+      "grad_norm": 0.2021484375,
+      "learning_rate": 6.886886886886888e-06,
+      "loss": 1.0151,
+      "step": 311
+    },
+    {
+      "epoch": 0.9369369369369369,
+      "grad_norm": 0.201171875,
+      "learning_rate": 6.876876876876878e-06,
+      "loss": 1.0345,
+      "step": 312
+    },
+    {
+      "epoch": 0.93993993993994,
+      "grad_norm": 0.224609375,
+      "learning_rate": 6.866866866866867e-06,
+      "loss": 0.9866,
+      "step": 313
+    },
+    {
+      "epoch": 0.9429429429429429,
+      "grad_norm": 0.228515625,
+      "learning_rate": 6.856856856856857e-06,
+      "loss": 1.0282,
+      "step": 314
+    },
+    {
+      "epoch": 0.9459459459459459,
+      "grad_norm": 0.1953125,
+      "learning_rate": 6.846846846846848e-06,
+      "loss": 0.9906,
+      "step": 315
+    },
+    {
+      "epoch": 0.948948948948949,
+      "grad_norm": 0.189453125,
+      "learning_rate": 6.836836836836838e-06,
+      "loss": 1.0186,
+      "step": 316
+    },
+    {
+      "epoch": 0.9519519519519519,
+      "grad_norm": 0.265625,
+      "learning_rate": 6.826826826826828e-06,
+      "loss": 0.9693,
+      "step": 317
+    },
+    {
+      "epoch": 0.954954954954955,
+      "grad_norm": 0.2041015625,
+      "learning_rate": 6.816816816816817e-06,
+      "loss": 1.009,
+      "step": 318
+    },
+    {
+      "epoch": 0.9579579579579579,
+      "grad_norm": 0.267578125,
+      "learning_rate": 6.8068068068068075e-06,
+      "loss": 1.005,
+      "step": 319
+    },
+    {
+      "epoch": 0.960960960960961,
+      "grad_norm": 0.205078125,
+      "learning_rate": 6.796796796796798e-06,
+      "loss": 1.086,
+      "step": 320
+    },
+    {
+      "epoch": 0.963963963963964,
+      "grad_norm": 0.2080078125,
+      "learning_rate": 6.786786786786788e-06,
+      "loss": 1.0025,
+      "step": 321
+    },
+    {
+      "epoch": 0.9669669669669669,
+      "grad_norm": 0.2109375,
+      "learning_rate": 6.776776776776778e-06,
+      "loss": 1.0748,
+      "step": 322
+    },
+    {
+      "epoch": 0.96996996996997,
+      "grad_norm": 0.2275390625,
+      "learning_rate": 6.7667667667667665e-06,
+      "loss": 1.02,
+      "step": 323
+    },
+    {
+      "epoch": 0.972972972972973,
+      "grad_norm": 0.2119140625,
+      "learning_rate": 6.7567567567567575e-06,
+      "loss": 1.0112,
+      "step": 324
+    },
+    {
+      "epoch": 0.975975975975976,
+      "grad_norm": 0.1962890625,
+      "learning_rate": 6.746746746746748e-06,
+      "loss": 1.0305,
+      "step": 325
+    },
+    {
+      "epoch": 0.978978978978979,
+      "grad_norm": 0.2431640625,
+      "learning_rate": 6.736736736736738e-06,
+      "loss": 1.0332,
+      "step": 326
+    },
+    {
+      "epoch": 0.9819819819819819,
+      "grad_norm": 0.2109375,
+      "learning_rate": 6.726726726726728e-06,
+      "loss": 1.0257,
+      "step": 327
+    },
+    {
+      "epoch": 0.984984984984985,
+      "grad_norm": 0.310546875,
+      "learning_rate": 6.716716716716717e-06,
+      "loss": 0.9938,
+      "step": 328
+    },
+    {
+      "epoch": 0.987987987987988,
+      "grad_norm": 0.2041015625,
+      "learning_rate": 6.706706706706707e-06,
+      "loss": 1.0773,
+      "step": 329
+    },
+    {
+      "epoch": 0.990990990990991,
+      "grad_norm": 0.236328125,
+      "learning_rate": 6.696696696696697e-06,
+      "loss": 1.0023,
+      "step": 330
+    },
+    {
+      "epoch": 0.993993993993994,
+      "grad_norm": 0.2099609375,
+      "learning_rate": 6.686686686686687e-06,
+      "loss": 1.0111,
+      "step": 331
+    },
+    {
+      "epoch": 0.996996996996997,
+      "grad_norm": 0.20703125,
+      "learning_rate": 6.676676676676678e-06,
+      "loss": 0.9683,
+      "step": 332
+    },
+    {
+      "epoch": 1.0,
+      "grad_norm": 0.234375,
+      "learning_rate": 6.666666666666667e-06,
+      "loss": 1.0111,
+      "step": 333
+    },
+    {
+      "epoch": 1.003003003003003,
+      "grad_norm": 0.2021484375,
+      "learning_rate": 6.656656656656657e-06,
+      "loss": 0.9873,
+      "step": 334
+    },
+    {
+      "epoch": 1.006006006006006,
+      "grad_norm": 0.1904296875,
+      "learning_rate": 6.646646646646647e-06,
+      "loss": 0.9928,
+      "step": 335
+    },
+    {
+      "epoch": 1.009009009009009,
+      "grad_norm": 0.2333984375,
+      "learning_rate": 6.636636636636637e-06,
+      "loss": 0.9473,
+      "step": 336
+    },
+    {
+      "epoch": 1.012012012012012,
+      "grad_norm": 0.298828125,
+      "learning_rate": 6.626626626626627e-06,
+      "loss": 1.0064,
+      "step": 337
+    },
+    {
+      "epoch": 1.015015015015015,
+      "grad_norm": 0.205078125,
+      "learning_rate": 6.616616616616617e-06,
+      "loss": 0.955,
+      "step": 338
+    },
+    {
+      "epoch": 1.018018018018018,
+      "grad_norm": 0.2119140625,
+      "learning_rate": 6.606606606606607e-06,
+      "loss": 1.0592,
+      "step": 339
+    },
+    {
+      "epoch": 1.021021021021021,
+      "grad_norm": 0.255859375,
+      "learning_rate": 6.596596596596597e-06,
+      "loss": 1.0287,
+      "step": 340
+    },
+    {
+      "epoch": 1.024024024024024,
+      "grad_norm": 0.232421875,
+      "learning_rate": 6.586586586586587e-06,
+      "loss": 0.9764,
+      "step": 341
+    },
+    {
+      "epoch": 1.027027027027027,
+      "grad_norm": 0.212890625,
+      "learning_rate": 6.5765765765765775e-06,
+      "loss": 1.006,
+      "step": 342
+    },
+    {
+      "epoch": 1.03003003003003,
+      "grad_norm": 0.2314453125,
+      "learning_rate": 6.566566566566567e-06,
+      "loss": 1.0015,
+      "step": 343
+    },
+    {
+      "epoch": 1.033033033033033,
+      "grad_norm": 0.26171875,
+      "learning_rate": 6.556556556556557e-06,
+      "loss": 1.0186,
+      "step": 344
+    },
+    {
+      "epoch": 1.0360360360360361,
+      "grad_norm": 0.220703125,
+      "learning_rate": 6.546546546546547e-06,
+      "loss": 0.99,
+      "step": 345
+    },
+    {
+      "epoch": 1.039039039039039,
+      "grad_norm": 0.2138671875,
+      "learning_rate": 6.536536536536537e-06,
+      "loss": 1.0623,
+      "step": 346
+    },
+    {
+      "epoch": 1.042042042042042,
+      "grad_norm": 0.2138671875,
+      "learning_rate": 6.526526526526527e-06,
+      "loss": 0.9854,
+      "step": 347
+    },
+    {
+      "epoch": 1.045045045045045,
+      "grad_norm": 0.22265625,
+      "learning_rate": 6.516516516516517e-06,
+      "loss": 0.9782,
+      "step": 348
+    },
+    {
+      "epoch": 1.048048048048048,
+      "grad_norm": 0.2216796875,
+      "learning_rate": 6.506506506506507e-06,
+      "loss": 1.018,
+      "step": 349
+    },
+    {
+      "epoch": 1.0510510510510511,
+      "grad_norm": 0.208984375,
+      "learning_rate": 6.496496496496497e-06,
+      "loss": 0.9709,
+      "step": 350
+    },
+    {
+      "epoch": 1.054054054054054,
+      "grad_norm": 0.29296875,
+      "learning_rate": 6.486486486486487e-06,
+      "loss": 1.0351,
+      "step": 351
+    },
+    {
+      "epoch": 1.057057057057057,
+      "grad_norm": 0.2294921875,
+      "learning_rate": 6.476476476476477e-06,
+      "loss": 0.9985,
+      "step": 352
+    },
+    {
+      "epoch": 1.06006006006006,
+      "grad_norm": 0.2138671875,
+      "learning_rate": 6.466466466466467e-06,
+      "loss": 0.938,
+      "step": 353
+    },
+    {
+      "epoch": 1.063063063063063,
+      "grad_norm": 0.359375,
+      "learning_rate": 6.456456456456457e-06,
+      "loss": 0.937,
+      "step": 354
+    },
+    {
+      "epoch": 1.0660660660660661,
+      "grad_norm": 0.2119140625,
+      "learning_rate": 6.446446446446447e-06,
+      "loss": 0.9929,
+      "step": 355
+    },
+    {
+      "epoch": 1.069069069069069,
+      "grad_norm": 0.2109375,
+      "learning_rate": 6.4364364364364375e-06,
+      "loss": 0.9783,
+      "step": 356
+    },
+    {
+      "epoch": 1.072072072072072,
+      "grad_norm": 0.26171875,
+      "learning_rate": 6.426426426426427e-06,
+      "loss": 0.9743,
+      "step": 357
+    },
+    {
+      "epoch": 1.075075075075075,
+      "grad_norm": 0.2392578125,
+      "learning_rate": 6.416416416416417e-06,
+      "loss": 0.9639,
+      "step": 358
+    },
+    {
+      "epoch": 1.078078078078078,
+      "grad_norm": 0.2236328125,
+      "learning_rate": 6.406406406406407e-06,
+      "loss": 0.999,
+      "step": 359
+    },
+    {
+      "epoch": 1.0810810810810811,
+      "grad_norm": 0.224609375,
+      "learning_rate": 6.396396396396397e-06,
+      "loss": 0.9957,
+      "step": 360
+    },
+    {
+      "epoch": 1.0840840840840842,
+      "grad_norm": 0.2021484375,
+      "learning_rate": 6.3863863863863875e-06,
+      "loss": 1.0111,
+      "step": 361
+    },
+    {
+      "epoch": 1.087087087087087,
+      "grad_norm": 0.2109375,
+      "learning_rate": 6.376376376376376e-06,
+      "loss": 1.0141,
+      "step": 362
+    },
+    {
+      "epoch": 1.09009009009009,
+      "grad_norm": 0.28125,
+      "learning_rate": 6.366366366366366e-06,
+      "loss": 1.0369,
+      "step": 363
+    },
+    {
+      "epoch": 1.093093093093093,
+      "grad_norm": 0.21875,
+      "learning_rate": 6.356356356356357e-06,
+      "loss": 0.9964,
+      "step": 364
+    },
+    {
+      "epoch": 1.0960960960960962,
+      "grad_norm": 0.220703125,
+      "learning_rate": 6.3463463463463474e-06,
+      "loss": 1.004,
+      "step": 365
+    },
+    {
+      "epoch": 1.0990990990990992,
+      "grad_norm": 0.216796875,
+      "learning_rate": 6.336336336336338e-06,
+      "loss": 0.9596,
+      "step": 366
+    },
+    {
+      "epoch": 1.102102102102102,
+      "grad_norm": 0.2119140625,
+      "learning_rate": 6.326326326326326e-06,
+      "loss": 1.0075,
+      "step": 367
+    },
+    {
+      "epoch": 1.105105105105105,
+      "grad_norm": 0.28515625,
+      "learning_rate": 6.316316316316316e-06,
+      "loss": 1.0396,
+      "step": 368
+    },
+    {
+      "epoch": 1.1081081081081081,
+      "grad_norm": 0.205078125,
+      "learning_rate": 6.3063063063063065e-06,
+      "loss": 0.9559,
+      "step": 369
+    },
+    {
+      "epoch": 1.1111111111111112,
+      "grad_norm": 0.2060546875,
+      "learning_rate": 6.296296296296297e-06,
+      "loss": 0.951,
+      "step": 370
+    },
+    {
+      "epoch": 1.1141141141141142,
+      "grad_norm": 0.2265625,
+      "learning_rate": 6.286286286286287e-06,
+      "loss": 0.9737,
+      "step": 371
+    },
+    {
+      "epoch": 1.117117117117117,
+      "grad_norm": 0.2236328125,
+      "learning_rate": 6.276276276276276e-06,
+      "loss": 0.9998,
+      "step": 372
+    },
+    {
+      "epoch": 1.12012012012012,
+      "grad_norm": 0.279296875,
+      "learning_rate": 6.266266266266266e-06,
+      "loss": 0.9445,
+      "step": 373
+    },
+    {
+      "epoch": 1.1231231231231231,
+      "grad_norm": 0.236328125,
+      "learning_rate": 6.2562562562562565e-06,
+      "loss": 0.9871,
+      "step": 374
+    },
+    {
+      "epoch": 1.1261261261261262,
+      "grad_norm": 0.265625,
+      "learning_rate": 6.246246246246247e-06,
+      "loss": 1.0417,
+      "step": 375
+    },
+    {
+      "epoch": 1.1291291291291292,
+      "grad_norm": 0.2353515625,
+      "learning_rate": 6.236236236236237e-06,
+      "loss": 1.0213,
+      "step": 376
+    },
+    {
+      "epoch": 1.132132132132132,
+      "grad_norm": 0.2255859375,
+      "learning_rate": 6.226226226226226e-06,
+      "loss": 1.0245,
+      "step": 377
+    },
+    {
+      "epoch": 1.135135135135135,
+      "grad_norm": 0.2099609375,
+      "learning_rate": 6.2162162162162164e-06,
+      "loss": 0.9639,
+      "step": 378
+    },
+    {
+      "epoch": 1.1381381381381381,
+      "grad_norm": 0.2314453125,
+      "learning_rate": 6.206206206206207e-06,
+      "loss": 0.9978,
+      "step": 379
+    },
+    {
+      "epoch": 1.1411411411411412,
+      "grad_norm": 0.216796875,
+      "learning_rate": 6.196196196196197e-06,
+      "loss": 0.9651,
+      "step": 380
+    },
+    {
+      "epoch": 1.1441441441441442,
+      "grad_norm": 0.1982421875,
+      "learning_rate": 6.186186186186187e-06,
+      "loss": 0.9537,
+      "step": 381
+    },
+    {
+      "epoch": 1.147147147147147,
+      "grad_norm": 0.2138671875,
+      "learning_rate": 6.176176176176176e-06,
+      "loss": 0.9689,
+      "step": 382
+    },
+    {
+      "epoch": 1.15015015015015,
+      "grad_norm": 0.224609375,
+      "learning_rate": 6.1661661661661665e-06,
+      "loss": 1.0359,
+      "step": 383
+    },
+    {
+      "epoch": 1.1531531531531531,
+      "grad_norm": 0.2099609375,
+      "learning_rate": 6.156156156156157e-06,
+      "loss": 0.9635,
+      "step": 384
+    },
+    {
+      "epoch": 1.1561561561561562,
+      "grad_norm": 0.2392578125,
+      "learning_rate": 6.146146146146147e-06,
+      "loss": 0.9649,
+      "step": 385
+    },
+    {
+      "epoch": 1.1591591591591592,
+      "grad_norm": 0.212890625,
+      "learning_rate": 6.136136136136137e-06,
+      "loss": 0.9773,
+      "step": 386
+    },
+    {
+      "epoch": 1.1621621621621623,
+      "grad_norm": 0.263671875,
+      "learning_rate": 6.126126126126126e-06,
+      "loss": 0.9508,
+      "step": 387
+    },
+    {
+      "epoch": 1.165165165165165,
+      "grad_norm": 0.291015625,
+      "learning_rate": 6.1161161161161166e-06,
+      "loss": 0.9806,
+      "step": 388
+    },
+    {
+      "epoch": 1.1681681681681682,
+      "grad_norm": 0.271484375,
+      "learning_rate": 6.106106106106107e-06,
+      "loss": 0.9758,
+      "step": 389
+    },
+    {
+      "epoch": 1.1711711711711712,
+      "grad_norm": 0.2158203125,
+      "learning_rate": 6.096096096096097e-06,
+      "loss": 0.9403,
+      "step": 390
+    },
+    {
+      "epoch": 1.1741741741741742,
+      "grad_norm": 0.2060546875,
+      "learning_rate": 6.086086086086087e-06,
+      "loss": 1.0073,
+      "step": 391
+    },
+    {
+      "epoch": 1.1771771771771773,
+      "grad_norm": 0.2255859375,
+      "learning_rate": 6.0760760760760765e-06,
+      "loss": 0.9605,
+      "step": 392
+    },
+    {
+      "epoch": 1.1801801801801801,
+      "grad_norm": 0.232421875,
+      "learning_rate": 6.066066066066067e-06,
+      "loss": 0.956,
+      "step": 393
+    },
+    {
+      "epoch": 1.1831831831831832,
+      "grad_norm": 0.24609375,
+      "learning_rate": 6.056056056056057e-06,
+      "loss": 0.992,
+      "step": 394
+    },
+    {
+      "epoch": 1.1861861861861862,
+      "grad_norm": 0.2265625,
+      "learning_rate": 6.046046046046047e-06,
+      "loss": 1.002,
+      "step": 395
+    },
+    {
+      "epoch": 1.1891891891891893,
+      "grad_norm": 0.2099609375,
+      "learning_rate": 6.036036036036037e-06,
+      "loss": 0.9834,
+      "step": 396
+    },
+    {
+      "epoch": 1.1921921921921923,
+      "grad_norm": 0.212890625,
+      "learning_rate": 6.0260260260260265e-06,
+      "loss": 0.9428,
+      "step": 397
+    },
+    {
+      "epoch": 1.1951951951951951,
+      "grad_norm": 0.2080078125,
+      "learning_rate": 6.016016016016017e-06,
+      "loss": 0.9774,
+      "step": 398
+    },
+    {
+      "epoch": 1.1981981981981982,
+      "grad_norm": 0.357421875,
+      "learning_rate": 6.006006006006007e-06,
+      "loss": 0.9665,
+      "step": 399
+    },
+    {
+      "epoch": 1.2012012012012012,
+      "grad_norm": 0.21875,
+      "learning_rate": 5.995995995995997e-06,
+      "loss": 0.9895,
+      "step": 400
+    },
+    {
+      "epoch": 1.2042042042042043,
+      "grad_norm": 0.216796875,
+      "learning_rate": 5.985985985985987e-06,
+      "loss": 0.9463,
+      "step": 401
+    },
+    {
+      "epoch": 1.2072072072072073,
+      "grad_norm": 0.24609375,
+      "learning_rate": 5.975975975975976e-06,
+      "loss": 1.0247,
+      "step": 402
+    },
+    {
+      "epoch": 1.2102102102102101,
+      "grad_norm": 0.216796875,
+      "learning_rate": 5.965965965965966e-06,
+      "loss": 0.9705,
+      "step": 403
+    },
+    {
+      "epoch": 1.2132132132132132,
+      "grad_norm": 0.2265625,
+      "learning_rate": 5.955955955955957e-06,
+      "loss": 0.9489,
+      "step": 404
+    },
+    {
+      "epoch": 1.2162162162162162,
+      "grad_norm": 0.2158203125,
+      "learning_rate": 5.945945945945947e-06,
+      "loss": 0.992,
+      "step": 405
+    },
+    {
+      "epoch": 1.2192192192192193,
+      "grad_norm": 0.216796875,
+      "learning_rate": 5.935935935935936e-06,
+      "loss": 0.9664,
+      "step": 406
+    },
+    {
+      "epoch": 1.2222222222222223,
+      "grad_norm": 0.263671875,
+      "learning_rate": 5.925925925925926e-06,
+      "loss": 0.9687,
+      "step": 407
+    },
+    {
+      "epoch": 1.2252252252252251,
+      "grad_norm": 0.23046875,
+      "learning_rate": 5.915915915915916e-06,
+      "loss": 0.9521,
+      "step": 408
+    },
+    {
+      "epoch": 1.2282282282282282,
+      "grad_norm": 0.314453125,
+      "learning_rate": 5.905905905905906e-06,
+      "loss": 1.0079,
+      "step": 409
+    },
+    {
+      "epoch": 1.2312312312312312,
+      "grad_norm": 0.20703125,
+      "learning_rate": 5.895895895895896e-06,
+      "loss": 0.9922,
+      "step": 410
+    },
+    {
+      "epoch": 1.2342342342342343,
+      "grad_norm": 0.2060546875,
+      "learning_rate": 5.885885885885886e-06,
+      "loss": 0.9572,
+      "step": 411
+    },
+    {
+      "epoch": 1.2372372372372373,
+      "grad_norm": 0.2158203125,
+      "learning_rate": 5.875875875875876e-06,
+      "loss": 0.9603,
+      "step": 412
+    },
+    {
+      "epoch": 1.2402402402402402,
+      "grad_norm": 0.33984375,
+      "learning_rate": 5.865865865865866e-06,
+      "loss": 0.9716,
+      "step": 413
+    },
+    {
+      "epoch": 1.2432432432432432,
+      "grad_norm": 0.224609375,
+      "learning_rate": 5.855855855855856e-06,
+      "loss": 0.9517,
+      "step": 414
+    },
+    {
+      "epoch": 1.2462462462462462,
+      "grad_norm": 0.2236328125,
+      "learning_rate": 5.8458458458458464e-06,
+      "loss": 1.0133,
+      "step": 415
+    },
+    {
+      "epoch": 1.2492492492492493,
+      "grad_norm": 0.2412109375,
+      "learning_rate": 5.835835835835836e-06,
+      "loss": 1.0868,
+      "step": 416
+    },
+    {
+      "epoch": 1.2522522522522523,
+      "grad_norm": 0.244140625,
+      "learning_rate": 5.825825825825826e-06,
+      "loss": 1.0069,
+      "step": 417
+    },
+    {
+      "epoch": 1.2552552552552552,
+      "grad_norm": 0.21484375,
+      "learning_rate": 5.815815815815816e-06,
+      "loss": 0.9714,
+      "step": 418
+    },
+    {
+      "epoch": 1.2582582582582582,
+      "grad_norm": 0.21875,
+      "learning_rate": 5.805805805805806e-06,
+      "loss": 1.0243,
+      "step": 419
+    },
+    {
+      "epoch": 1.2612612612612613,
+      "grad_norm": 0.291015625,
+      "learning_rate": 5.7957957957957965e-06,
+      "loss": 0.9103,
+      "step": 420
+    },
+    {
+      "epoch": 1.2642642642642643,
+      "grad_norm": 0.2041015625,
+      "learning_rate": 5.785785785785786e-06,
+      "loss": 0.9576,
+      "step": 421
+    },
+    {
+      "epoch": 1.2672672672672673,
+      "grad_norm": 0.220703125,
+      "learning_rate": 5.775775775775776e-06,
+      "loss": 0.9476,
+      "step": 422
+    },
+    {
+      "epoch": 1.2702702702702702,
+      "grad_norm": 0.1943359375,
+      "learning_rate": 5.765765765765766e-06,
+      "loss": 0.9265,
+      "step": 423
+    },
+    {
+      "epoch": 1.2732732732732732,
+      "grad_norm": 0.2294921875,
+      "learning_rate": 5.755755755755756e-06,
+      "loss": 1.0366,
+      "step": 424
+    },
+    {
+      "epoch": 1.2762762762762763,
+      "grad_norm": 0.255859375,
+      "learning_rate": 5.7457457457457466e-06,
+      "loss": 0.9873,
+      "step": 425
+    },
+    {
+      "epoch": 1.2792792792792793,
+      "grad_norm": 0.2109375,
+      "learning_rate": 5.735735735735736e-06,
+      "loss": 0.9633,
+      "step": 426
+    },
+    {
+      "epoch": 1.2822822822822824,
+      "grad_norm": 0.236328125,
+      "learning_rate": 5.725725725725726e-06,
+      "loss": 0.9714,
+      "step": 427
+    },
+    {
+      "epoch": 1.2852852852852852,
+      "grad_norm": 0.2373046875,
+      "learning_rate": 5.715715715715716e-06,
+      "loss": 0.9301,
+      "step": 428
+    },
+    {
+      "epoch": 1.2882882882882882,
+      "grad_norm": 0.216796875,
+      "learning_rate": 5.7057057057057065e-06,
+      "loss": 0.9624,
+      "step": 429
+    },
+    {
+      "epoch": 1.2912912912912913,
+      "grad_norm": 0.27734375,
+      "learning_rate": 5.695695695695697e-06,
+      "loss": 0.9304,
+      "step": 430
+    },
+    {
+      "epoch": 1.2942942942942943,
+      "grad_norm": 0.203125,
+      "learning_rate": 5.685685685685686e-06,
+      "loss": 0.9892,
+      "step": 431
+    },
+    {
+      "epoch": 1.2972972972972974,
+      "grad_norm": 0.2255859375,
+      "learning_rate": 5.675675675675676e-06,
+      "loss": 0.9703,
+      "step": 432
+    },
+    {
+      "epoch": 1.3003003003003002,
+      "grad_norm": 0.2158203125,
+      "learning_rate": 5.665665665665666e-06,
+      "loss": 1.0354,
+      "step": 433
+    },
+    {
+      "epoch": 1.3033033033033032,
+      "grad_norm": 0.2177734375,
+      "learning_rate": 5.6556556556556565e-06,
+      "loss": 0.9887,
+      "step": 434
+    },
+    {
+      "epoch": 1.3063063063063063,
+      "grad_norm": 0.244140625,
+      "learning_rate": 5.645645645645647e-06,
+      "loss": 0.9452,
+      "step": 435
+    },
+    {
+      "epoch": 1.3093093093093093,
+      "grad_norm": 0.2138671875,
+      "learning_rate": 5.635635635635636e-06,
+      "loss": 0.9697,
+      "step": 436
+    },
+    {
+      "epoch": 1.3123123123123124,
+      "grad_norm": 0.251953125,
+      "learning_rate": 5.625625625625626e-06,
+      "loss": 0.9527,
+      "step": 437
+    },
+    {
+      "epoch": 1.3153153153153152,
+      "grad_norm": 0.2265625,
+      "learning_rate": 5.615615615615616e-06,
+      "loss": 0.9735,
+      "step": 438
+    },
+    {
+      "epoch": 1.3183183183183182,
+      "grad_norm": 0.2470703125,
+      "learning_rate": 5.605605605605607e-06,
+      "loss": 0.9782,
+      "step": 439
+    },
+    {
+      "epoch": 1.3213213213213213,
+      "grad_norm": 0.2099609375,
+      "learning_rate": 5.595595595595597e-06,
+      "loss": 0.9532,
+      "step": 440
+    },
+    {
+      "epoch": 1.3243243243243243,
+      "grad_norm": 0.2109375,
+      "learning_rate": 5.585585585585585e-06,
+      "loss": 0.9498,
+      "step": 441
+    },
+    {
+      "epoch": 1.3273273273273274,
+      "grad_norm": 0.2392578125,
+      "learning_rate": 5.5755755755755755e-06,
+      "loss": 0.9785,
+      "step": 442
+    },
+    {
+      "epoch": 1.3303303303303302,
+      "grad_norm": 0.2109375,
+      "learning_rate": 5.565565565565566e-06,
+      "loss": 0.9595,
+      "step": 443
+    },
+    {
+      "epoch": 1.3333333333333333,
+      "grad_norm": 0.205078125,
+      "learning_rate": 5.555555555555557e-06,
+      "loss": 0.9411,
+      "step": 444
+    },
+    {
+      "epoch": 1.3363363363363363,
+      "grad_norm": 0.212890625,
+      "learning_rate": 5.545545545545547e-06,
+      "loss": 0.9908,
+      "step": 445
+    },
+    {
+      "epoch": 1.3393393393393394,
+      "grad_norm": 0.2236328125,
+      "learning_rate": 5.535535535535535e-06,
+      "loss": 0.9618,
+      "step": 446
+    },
+    {
+      "epoch": 1.3423423423423424,
+      "grad_norm": 0.216796875,
+      "learning_rate": 5.5255255255255255e-06,
+      "loss": 0.9095,
+      "step": 447
+    },
+    {
+      "epoch": 1.3453453453453452,
+      "grad_norm": 0.26953125,
+      "learning_rate": 5.515515515515516e-06,
+      "loss": 0.9738,
+      "step": 448
+    },
+    {
+      "epoch": 1.3483483483483483,
+      "grad_norm": 0.248046875,
+      "learning_rate": 5.505505505505506e-06,
+      "loss": 0.9569,
+      "step": 449
+    },
+    {
+      "epoch": 1.3513513513513513,
+      "grad_norm": 0.2119140625,
+      "learning_rate": 5.495495495495496e-06,
+      "loss": 0.9864,
+      "step": 450
+    },
+    {
+      "epoch": 1.3543543543543544,
+      "grad_norm": 0.251953125,
+      "learning_rate": 5.485485485485485e-06,
+      "loss": 0.9969,
+      "step": 451
+    },
+    {
+      "epoch": 1.3573573573573574,
+      "grad_norm": 0.267578125,
+      "learning_rate": 5.475475475475476e-06,
+      "loss": 0.9218,
+      "step": 452
+    },
+    {
+      "epoch": 1.3603603603603602,
+      "grad_norm": 0.333984375,
+      "learning_rate": 5.465465465465466e-06,
+      "loss": 0.9345,
+      "step": 453
+    },
+    {
+      "epoch": 1.3633633633633635,
+      "grad_norm": 0.29296875,
+      "learning_rate": 5.455455455455456e-06,
+      "loss": 1.0285,
+      "step": 454
+    },
+    {
+      "epoch": 1.3663663663663663,
+      "grad_norm": 0.39453125,
+      "learning_rate": 5.445445445445446e-06,
+      "loss": 1.0049,
+      "step": 455
+    },
+    {
+      "epoch": 1.3693693693693694,
+      "grad_norm": 0.2197265625,
+      "learning_rate": 5.4354354354354355e-06,
+      "loss": 0.9251,
+      "step": 456
+    },
+    {
+      "epoch": 1.3723723723723724,
+      "grad_norm": 0.208984375,
+      "learning_rate": 5.425425425425426e-06,
+      "loss": 0.9518,
+      "step": 457
+    },
+    {
+      "epoch": 1.3753753753753752,
+      "grad_norm": 0.22265625,
+      "learning_rate": 5.415415415415416e-06,
+      "loss": 0.9486,
+      "step": 458
+    },
+    {
+      "epoch": 1.3783783783783785,
+      "grad_norm": 0.2236328125,
+      "learning_rate": 5.405405405405406e-06,
+      "loss": 0.9681,
+      "step": 459
+    },
+    {
+      "epoch": 1.3813813813813813,
+      "grad_norm": 0.2275390625,
+      "learning_rate": 5.395395395395396e-06,
+      "loss": 0.9777,
+      "step": 460
+    },
+    {
+      "epoch": 1.3843843843843844,
+      "grad_norm": 0.22265625,
+      "learning_rate": 5.3853853853853856e-06,
+      "loss": 0.9428,
+      "step": 461
+    },
+    {
+      "epoch": 1.3873873873873874,
+      "grad_norm": 0.2216796875,
+      "learning_rate": 5.375375375375376e-06,
+      "loss": 0.9608,
+      "step": 462
+    },
+    {
+      "epoch": 1.3903903903903903,
+      "grad_norm": 0.21484375,
+      "learning_rate": 5.365365365365366e-06,
+      "loss": 0.9433,
+      "step": 463
+    },
+    {
+      "epoch": 1.3933933933933935,
+      "grad_norm": 0.271484375,
+      "learning_rate": 5.355355355355356e-06,
+      "loss": 0.97,
+      "step": 464
+    },
+    {
+      "epoch": 1.3963963963963963,
+      "grad_norm": 0.2265625,
+      "learning_rate": 5.345345345345346e-06,
+      "loss": 0.9998,
+      "step": 465
+    },
+    {
+      "epoch": 1.3993993993993994,
+      "grad_norm": 0.2177734375,
+      "learning_rate": 5.335335335335336e-06,
+      "loss": 0.9415,
+      "step": 466
+    },
+    {
+      "epoch": 1.4024024024024024,
+      "grad_norm": 0.212890625,
+      "learning_rate": 5.325325325325326e-06,
+      "loss": 1.0013,
+      "step": 467
+    },
+    {
+      "epoch": 1.4054054054054055,
+      "grad_norm": 0.2138671875,
+      "learning_rate": 5.315315315315316e-06,
+      "loss": 0.9571,
+      "step": 468
+    },
+    {
+      "epoch": 1.4084084084084085,
+      "grad_norm": 0.2119140625,
+      "learning_rate": 5.305305305305306e-06,
+      "loss": 0.932,
+      "step": 469
+    },
+    {
+      "epoch": 1.4114114114114114,
+      "grad_norm": 0.2255859375,
+      "learning_rate": 5.2952952952952955e-06,
+      "loss": 0.9422,
+      "step": 470
+    },
+    {
+      "epoch": 1.4144144144144144,
+      "grad_norm": 0.2216796875,
+      "learning_rate": 5.285285285285286e-06,
+      "loss": 0.9769,
+      "step": 471
+    },
+    {
+      "epoch": 1.4174174174174174,
+      "grad_norm": 0.21875,
+      "learning_rate": 5.275275275275276e-06,
+      "loss": 0.9564,
+      "step": 472
+    },
+    {
+      "epoch": 1.4204204204204205,
+      "grad_norm": 2.953125,
+      "learning_rate": 5.265265265265266e-06,
+      "loss": 0.9256,
+      "step": 473
+    },
+    {
+      "epoch": 1.4234234234234235,
+      "grad_norm": 0.21875,
+      "learning_rate": 5.255255255255256e-06,
+      "loss": 0.9841,
+      "step": 474
+    },
+    {
+      "epoch": 1.4264264264264264,
+      "grad_norm": 0.263671875,
+      "learning_rate": 5.245245245245245e-06,
+      "loss": 1.0006,
+      "step": 475
+    },
+    {
+      "epoch": 1.4294294294294294,
+      "grad_norm": 2.140625,
+      "learning_rate": 5.235235235235236e-06,
+      "loss": 0.9967,
+      "step": 476
+    },
+    {
+      "epoch": 1.4324324324324325,
+      "grad_norm": 0.2216796875,
+      "learning_rate": 5.225225225225226e-06,
+      "loss": 0.9575,
+      "step": 477
+    },
+    {
+      "epoch": 1.4354354354354355,
+      "grad_norm": 0.30859375,
+      "learning_rate": 5.215215215215216e-06,
+      "loss": 0.9568,
+      "step": 478
+    },
+    {
+      "epoch": 1.4384384384384385,
+      "grad_norm": 0.21484375,
+      "learning_rate": 5.205205205205206e-06,
+      "loss": 0.9879,
+      "step": 479
+    },
+    {
+      "epoch": 1.4414414414414414,
+      "grad_norm": 0.2158203125,
+      "learning_rate": 5.195195195195195e-06,
+      "loss": 0.9682,
+      "step": 480
+    },
+    {
+      "epoch": 1.4444444444444444,
+      "grad_norm": 0.216796875,
+      "learning_rate": 5.185185185185185e-06,
+      "loss": 0.9758,
+      "step": 481
+    },
+    {
+      "epoch": 1.4474474474474475,
+      "grad_norm": 0.232421875,
+      "learning_rate": 5.175175175175175e-06,
+      "loss": 0.9181,
+      "step": 482
+    },
+    {
+      "epoch": 1.4504504504504505,
+      "grad_norm": 0.26953125,
+      "learning_rate": 5.165165165165165e-06,
+      "loss": 0.939,
+      "step": 483
+    },
+    {
+      "epoch": 1.4534534534534536,
+      "grad_norm": 0.31640625,
+      "learning_rate": 5.155155155155156e-06,
+      "loss": 0.9442,
+      "step": 484
+    },
+    {
+      "epoch": 1.4564564564564564,
+      "grad_norm": 0.2119140625,
+      "learning_rate": 5.145145145145145e-06,
+      "loss": 0.9697,
+      "step": 485
+    },
+    {
+      "epoch": 1.4594594594594594,
+      "grad_norm": 0.23828125,
+      "learning_rate": 5.135135135135135e-06,
+      "loss": 0.9684,
+      "step": 486
+    },
+    {
+      "epoch": 1.4624624624624625,
+      "grad_norm": 0.2392578125,
+      "learning_rate": 5.125125125125125e-06,
+      "loss": 0.9156,
+      "step": 487
+    },
+    {
+      "epoch": 1.4654654654654655,
+      "grad_norm": 0.232421875,
+      "learning_rate": 5.115115115115115e-06,
+      "loss": 0.9542,
+      "step": 488
+    },
+    {
+      "epoch": 1.4684684684684686,
+      "grad_norm": 0.25,
+      "learning_rate": 5.105105105105106e-06,
+      "loss": 0.9389,
+      "step": 489
+    },
+    {
+      "epoch": 1.4714714714714714,
+      "grad_norm": 0.2197265625,
+      "learning_rate": 5.095095095095095e-06,
+      "loss": 0.9677,
+      "step": 490
+    },
+    {
+      "epoch": 1.4744744744744744,
+      "grad_norm": 0.21875,
+      "learning_rate": 5.085085085085085e-06,
+      "loss": 0.97,
+      "step": 491
+    },
+    {
+      "epoch": 1.4774774774774775,
+      "grad_norm": 0.298828125,
+      "learning_rate": 5.075075075075075e-06,
+      "loss": 0.9723,
+      "step": 492
+    },
+    {
+      "epoch": 1.4804804804804805,
+      "grad_norm": 0.33984375,
+      "learning_rate": 5.0650650650650655e-06,
+      "loss": 0.9777,
+      "step": 493
+    },
+    {
+      "epoch": 1.4834834834834836,
+      "grad_norm": 0.2197265625,
+      "learning_rate": 5.055055055055056e-06,
+      "loss": 0.9669,
+      "step": 494
+    },
+    {
+      "epoch": 1.4864864864864864,
+      "grad_norm": 0.236328125,
+      "learning_rate": 5.045045045045045e-06,
+      "loss": 0.9796,
+      "step": 495
+    },
+    {
+      "epoch": 1.4894894894894894,
+      "grad_norm": 0.21875,
+      "learning_rate": 5.035035035035035e-06,
+      "loss": 0.9367,
+      "step": 496
+    },
+    {
+      "epoch": 1.4924924924924925,
+      "grad_norm": 0.2236328125,
+      "learning_rate": 5.025025025025025e-06,
+      "loss": 0.9917,
+      "step": 497
+    },
+    {
+      "epoch": 1.4954954954954955,
+      "grad_norm": 0.2265625,
+      "learning_rate": 5.0150150150150156e-06,
+      "loss": 0.9724,
+      "step": 498
+    },
+    {
+      "epoch": 1.4984984984984986,
+      "grad_norm": 0.287109375,
+      "learning_rate": 5.005005005005006e-06,
+      "loss": 0.9433,
+      "step": 499
+    },
+    {
+      "epoch": 1.5015015015015014,
+      "grad_norm": 0.212890625,
+      "learning_rate": 4.994994994994996e-06,
+      "loss": 0.9664,
+      "step": 500
+    },
+    {
+      "epoch": 1.5045045045045045,
+      "grad_norm": 0.228515625,
+      "learning_rate": 4.984984984984985e-06,
+      "loss": 0.9829,
+      "step": 501
+    },
+    {
+      "epoch": 1.5075075075075075,
+      "grad_norm": 0.2255859375,
+      "learning_rate": 4.9749749749749754e-06,
+      "loss": 0.9586,
+      "step": 502
+    },
+    {
+      "epoch": 1.5105105105105106,
+      "grad_norm": 0.21875,
+      "learning_rate": 4.964964964964966e-06,
+      "loss": 0.9388,
+      "step": 503
+    },
+    {
+      "epoch": 1.5135135135135136,
+      "grad_norm": 0.2470703125,
+      "learning_rate": 4.954954954954955e-06,
+      "loss": 0.9922,
+      "step": 504
+    },
+    {
+      "epoch": 1.5165165165165164,
+      "grad_norm": 0.2109375,
+      "learning_rate": 4.944944944944945e-06,
+      "loss": 0.9538,
+      "step": 505
+    },
+    {
+      "epoch": 1.5195195195195195,
+      "grad_norm": 0.259765625,
+      "learning_rate": 4.934934934934935e-06,
+      "loss": 0.9507,
+      "step": 506
+    },
+    {
+      "epoch": 1.5225225225225225,
+      "grad_norm": 0.2041015625,
+      "learning_rate": 4.9249249249249255e-06,
+      "loss": 0.9844,
+      "step": 507
+    },
+    {
+      "epoch": 1.5255255255255256,
+      "grad_norm": 0.21875,
+      "learning_rate": 4.914914914914916e-06,
+      "loss": 0.9786,
+      "step": 508
+    },
+    {
+      "epoch": 1.5285285285285286,
+      "grad_norm": 0.236328125,
+      "learning_rate": 4.904904904904905e-06,
+      "loss": 0.9647,
+      "step": 509
+    },
+    {
+      "epoch": 1.5315315315315314,
+      "grad_norm": 0.2099609375,
+      "learning_rate": 4.894894894894895e-06,
+      "loss": 0.9486,
+      "step": 510
+    },
+    {
+      "epoch": 1.5345345345345347,
+      "grad_norm": 0.2275390625,
+      "learning_rate": 4.884884884884885e-06,
+      "loss": 0.9999,
+      "step": 511
+    },
+    {
+      "epoch": 1.5375375375375375,
+      "grad_norm": 0.2255859375,
+      "learning_rate": 4.874874874874876e-06,
+      "loss": 0.9741,
+      "step": 512
+    },
+    {
+      "epoch": 1.5405405405405406,
+      "grad_norm": 0.212890625,
+      "learning_rate": 4.864864864864866e-06,
+      "loss": 0.9684,
+      "step": 513
+    },
+    {
+      "epoch": 1.5435435435435436,
+      "grad_norm": 0.345703125,
+      "learning_rate": 4.854854854854855e-06,
+      "loss": 0.9236,
+      "step": 514
+    },
+    {
+      "epoch": 1.5465465465465464,
+      "grad_norm": 0.216796875,
+      "learning_rate": 4.844844844844845e-06,
+      "loss": 0.962,
+      "step": 515
+    },
+    {
+      "epoch": 1.5495495495495497,
+      "grad_norm": 0.236328125,
+      "learning_rate": 4.8348348348348355e-06,
+      "loss": 0.913,
+      "step": 516
+    },
+    {
+      "epoch": 1.5525525525525525,
+      "grad_norm": 0.255859375,
+      "learning_rate": 4.824824824824826e-06,
+      "loss": 1.0614,
+      "step": 517
+    },
+    {
+      "epoch": 1.5555555555555556,
+      "grad_norm": 0.22265625,
+      "learning_rate": 4.814814814814815e-06,
+      "loss": 0.9623,
+      "step": 518
+    },
+    {
+      "epoch": 1.5585585585585586,
+      "grad_norm": 0.314453125,
+      "learning_rate": 4.804804804804805e-06,
+      "loss": 0.9721,
+      "step": 519
+    },
+    {
+      "epoch": 1.5615615615615615,
+      "grad_norm": 0.216796875,
+      "learning_rate": 4.794794794794795e-06,
+      "loss": 0.9289,
+      "step": 520
+    },
+    {
+      "epoch": 1.5645645645645647,
+      "grad_norm": 0.263671875,
+      "learning_rate": 4.784784784784785e-06,
+      "loss": 0.961,
+      "step": 521
+    },
+    {
+      "epoch": 1.5675675675675675,
+      "grad_norm": 0.2177734375,
+      "learning_rate": 4.774774774774775e-06,
+      "loss": 1.0066,
+      "step": 522
+    },
+    {
+      "epoch": 1.5705705705705706,
+      "grad_norm": 0.236328125,
+      "learning_rate": 4.764764764764765e-06,
+      "loss": 0.9969,
+      "step": 523
+    },
+    {
+      "epoch": 1.5735735735735736,
+      "grad_norm": 0.248046875,
+      "learning_rate": 4.754754754754755e-06,
+      "loss": 0.9688,
+      "step": 524
+    },
+    {
+      "epoch": 1.5765765765765765,
+      "grad_norm": 0.2177734375,
+      "learning_rate": 4.7447447447447454e-06,
+      "loss": 0.9719,
+      "step": 525
+    },
+    {
+      "epoch": 1.5795795795795797,
+      "grad_norm": 0.265625,
+      "learning_rate": 4.734734734734735e-06,
+      "loss": 0.9253,
+      "step": 526
+    },
+    {
+      "epoch": 1.5825825825825826,
+      "grad_norm": 0.2177734375,
+      "learning_rate": 4.724724724724725e-06,
+      "loss": 0.9707,
+      "step": 527
+    },
+    {
+      "epoch": 1.5855855855855856,
+      "grad_norm": 0.236328125,
+      "learning_rate": 4.714714714714715e-06,
+      "loss": 0.9427,
+      "step": 528
+    },
+    {
+      "epoch": 1.5885885885885886,
+      "grad_norm": 0.2236328125,
+      "learning_rate": 4.704704704704705e-06,
+      "loss": 0.9845,
+      "step": 529
+    },
+    {
+      "epoch": 1.5915915915915915,
+      "grad_norm": 0.23046875,
+      "learning_rate": 4.6946946946946955e-06,
+      "loss": 0.931,
+      "step": 530
+    },
+    {
+      "epoch": 1.5945945945945947,
+      "grad_norm": 0.212890625,
+      "learning_rate": 4.684684684684685e-06,
+      "loss": 0.9458,
+      "step": 531
+    },
+    {
+      "epoch": 1.5975975975975976,
+      "grad_norm": 0.2294921875,
+      "learning_rate": 4.674674674674675e-06,
+      "loss": 1.0014,
+      "step": 532
+    },
+    {
+      "epoch": 1.6006006006006006,
+      "grad_norm": 0.23828125,
+      "learning_rate": 4.664664664664665e-06,
+      "loss": 0.9649,
+      "step": 533
+    },
+    {
+      "epoch": 1.6036036036036037,
+      "grad_norm": 0.2216796875,
+      "learning_rate": 4.654654654654655e-06,
+      "loss": 0.9237,
+      "step": 534
+    },
+    {
+      "epoch": 1.6066066066066065,
+      "grad_norm": 0.2236328125,
+      "learning_rate": 4.6446446446446456e-06,
+      "loss": 0.9965,
+      "step": 535
+    },
+    {
+      "epoch": 1.6096096096096097,
+      "grad_norm": 0.23046875,
+      "learning_rate": 4.634634634634635e-06,
+      "loss": 0.9584,
+      "step": 536
+    },
+    {
+      "epoch": 1.6126126126126126,
+      "grad_norm": 0.263671875,
+      "learning_rate": 4.624624624624625e-06,
+      "loss": 0.9746,
+      "step": 537
+    },
+    {
+      "epoch": 1.6156156156156156,
+      "grad_norm": 0.2177734375,
+      "learning_rate": 4.614614614614614e-06,
+      "loss": 0.9986,
+      "step": 538
+    },
+    {
+      "epoch": 1.6186186186186187,
+      "grad_norm": 0.216796875,
+      "learning_rate": 4.604604604604605e-06,
+      "loss": 0.9686,
+      "step": 539
+    },
+    {
+      "epoch": 1.6216216216216215,
+      "grad_norm": 0.2275390625,
+      "learning_rate": 4.594594594594596e-06,
+      "loss": 1.003,
+      "step": 540
+    },
+    {
+      "epoch": 1.6246246246246248,
+      "grad_norm": 0.2099609375,
+      "learning_rate": 4.584584584584585e-06,
+      "loss": 0.9655,
+      "step": 541
+    },
+    {
+      "epoch": 1.6276276276276276,
+      "grad_norm": 0.255859375,
+      "learning_rate": 4.574574574574575e-06,
+      "loss": 0.9058,
+      "step": 542
+    },
+    {
+      "epoch": 1.6306306306306306,
+      "grad_norm": 0.2236328125,
+      "learning_rate": 4.5645645645645645e-06,
+      "loss": 0.9205,
+      "step": 543
+    },
+    {
+      "epoch": 1.6336336336336337,
+      "grad_norm": 0.2451171875,
+      "learning_rate": 4.554554554554555e-06,
+      "loss": 0.9501,
+      "step": 544
+    },
+    {
+      "epoch": 1.6366366366366365,
+      "grad_norm": 0.23046875,
+      "learning_rate": 4.544544544544545e-06,
+      "loss": 0.9908,
+      "step": 545
+    },
+    {
+      "epoch": 1.6396396396396398,
+      "grad_norm": 0.224609375,
+      "learning_rate": 4.534534534534535e-06,
+      "loss": 0.9617,
+      "step": 546
+    },
+    {
+      "epoch": 1.6426426426426426,
+      "grad_norm": 0.2197265625,
+      "learning_rate": 4.524524524524525e-06,
+      "loss": 0.9957,
+      "step": 547
+    },
+    {
+      "epoch": 1.6456456456456456,
+      "grad_norm": 0.212890625,
+      "learning_rate": 4.5145145145145146e-06,
+      "loss": 0.9649,
+      "step": 548
+    },
+    {
+      "epoch": 1.6486486486486487,
+      "grad_norm": 0.29296875,
+      "learning_rate": 4.504504504504505e-06,
+      "loss": 0.9867,
+      "step": 549
+    },
+    {
+      "epoch": 1.6516516516516515,
+      "grad_norm": 0.2138671875,
+      "learning_rate": 4.494494494494495e-06,
+      "loss": 0.9191,
+      "step": 550
+    },
+    {
+      "epoch": 1.6546546546546548,
+      "grad_norm": 0.2490234375,
+      "learning_rate": 4.484484484484485e-06,
+      "loss": 0.9917,
+      "step": 551
+    },
+    {
+      "epoch": 1.6576576576576576,
+      "grad_norm": 0.302734375,
+      "learning_rate": 4.474474474474475e-06,
+      "loss": 0.9647,
+      "step": 552
+    },
+    {
+      "epoch": 1.6606606606606606,
+      "grad_norm": 0.2265625,
+      "learning_rate": 4.464464464464465e-06,
+      "loss": 0.985,
+      "step": 553
+    },
+    {
+      "epoch": 1.6636636636636637,
+      "grad_norm": 0.23046875,
+      "learning_rate": 4.454454454454455e-06,
+      "loss": 0.9358,
+      "step": 554
+    },
+    {
+      "epoch": 1.6666666666666665,
+      "grad_norm": 0.2314453125,
+      "learning_rate": 4.444444444444444e-06,
+      "loss": 0.938,
+      "step": 555
+    },
+    {
+      "epoch": 1.6696696696696698,
+      "grad_norm": 0.21875,
+      "learning_rate": 4.434434434434435e-06,
+      "loss": 0.9703,
+      "step": 556
+    },
+    {
+      "epoch": 1.6726726726726726,
+      "grad_norm": 0.2197265625,
+      "learning_rate": 4.424424424424425e-06,
+      "loss": 0.9716,
+      "step": 557
+    },
+    {
+      "epoch": 1.6756756756756757,
+      "grad_norm": 0.263671875,
+      "learning_rate": 4.414414414414415e-06,
+      "loss": 0.9434,
+      "step": 558
+    },
+    {
+      "epoch": 1.6786786786786787,
+      "grad_norm": 0.2197265625,
+      "learning_rate": 4.404404404404405e-06,
+      "loss": 0.9733,
+      "step": 559
+    },
+    {
+      "epoch": 1.6816816816816815,
+      "grad_norm": 0.2490234375,
+      "learning_rate": 4.394394394394394e-06,
+      "loss": 0.8826,
+      "step": 560
+    },
+    {
+      "epoch": 1.6846846846846848,
+      "grad_norm": 0.267578125,
+      "learning_rate": 4.384384384384384e-06,
+      "loss": 0.9659,
+      "step": 561
+    },
+    {
+      "epoch": 1.6876876876876876,
+      "grad_norm": 0.267578125,
+      "learning_rate": 4.374374374374375e-06,
+      "loss": 0.939,
+      "step": 562
+    },
+    {
+      "epoch": 1.6906906906906907,
+      "grad_norm": 0.2412109375,
+      "learning_rate": 4.364364364364365e-06,
+      "loss": 0.9483,
+      "step": 563
+    },
+    {
+      "epoch": 1.6936936936936937,
+      "grad_norm": 0.2353515625,
+      "learning_rate": 4.354354354354355e-06,
+      "loss": 0.9953,
+      "step": 564
+    },
+    {
+      "epoch": 1.6966966966966965,
+      "grad_norm": 0.236328125,
+      "learning_rate": 4.344344344344344e-06,
+      "loss": 0.9582,
+      "step": 565
+    },
+    {
+      "epoch": 1.6996996996996998,
+      "grad_norm": 0.296875,
+      "learning_rate": 4.3343343343343345e-06,
+      "loss": 0.9178,
+      "step": 566
+    },
+    {
+      "epoch": 1.7027027027027026,
+      "grad_norm": 0.2158203125,
+      "learning_rate": 4.324324324324325e-06,
+      "loss": 1.0095,
+      "step": 567
+    },
+    {
+      "epoch": 1.7057057057057057,
+      "grad_norm": 0.2373046875,
+      "learning_rate": 4.314314314314315e-06,
+      "loss": 0.946,
+      "step": 568
+    },
+    {
+      "epoch": 1.7087087087087087,
+      "grad_norm": 0.359375,
+      "learning_rate": 4.304304304304305e-06,
+      "loss": 0.9361,
+      "step": 569
+    },
+    {
+      "epoch": 1.7117117117117115,
+      "grad_norm": 0.2236328125,
+      "learning_rate": 4.294294294294294e-06,
+      "loss": 0.9628,
+      "step": 570
+    },
+    {
+      "epoch": 1.7147147147147148,
+      "grad_norm": 0.2265625,
+      "learning_rate": 4.2842842842842845e-06,
+      "loss": 0.9502,
+      "step": 571
+    },
+    {
+      "epoch": 1.7177177177177176,
+      "grad_norm": 0.2236328125,
+      "learning_rate": 4.274274274274275e-06,
+      "loss": 0.8967,
+      "step": 572
+    },
+    {
+      "epoch": 1.7207207207207207,
+      "grad_norm": 0.2421875,
+      "learning_rate": 4.264264264264265e-06,
+      "loss": 0.9369,
+      "step": 573
+    },
+    {
+      "epoch": 1.7237237237237237,
+      "grad_norm": 0.26171875,
+      "learning_rate": 4.254254254254255e-06,
+      "loss": 0.9129,
+      "step": 574
+    },
+    {
+      "epoch": 1.7267267267267268,
+      "grad_norm": 0.2353515625,
+      "learning_rate": 4.2442442442442444e-06,
+      "loss": 0.9549,
+      "step": 575
+    },
+    {
+      "epoch": 1.7297297297297298,
+      "grad_norm": 0.259765625,
+      "learning_rate": 4.234234234234235e-06,
+      "loss": 0.9616,
+      "step": 576
+    },
+    {
+      "epoch": 1.7327327327327327,
+      "grad_norm": 0.2119140625,
+      "learning_rate": 4.224224224224225e-06,
+      "loss": 0.971,
+      "step": 577
+    },
+    {
+      "epoch": 1.7357357357357357,
+      "grad_norm": 0.2177734375,
+      "learning_rate": 4.214214214214214e-06,
+      "loss": 0.9133,
+      "step": 578
+    },
+    {
+      "epoch": 1.7387387387387387,
+      "grad_norm": 0.28125,
+      "learning_rate": 4.204204204204204e-06,
+      "loss": 0.9309,
+      "step": 579
+    },
+    {
+      "epoch": 1.7417417417417418,
+      "grad_norm": 0.2119140625,
+      "learning_rate": 4.1941941941941945e-06,
+      "loss": 0.9594,
+      "step": 580
+    },
+    {
+      "epoch": 1.7447447447447448,
+      "grad_norm": 0.2275390625,
+      "learning_rate": 4.184184184184185e-06,
+      "loss": 0.9586,
+      "step": 581
+    },
+    {
+      "epoch": 1.7477477477477477,
+      "grad_norm": 0.24609375,
+      "learning_rate": 4.174174174174174e-06,
+      "loss": 1.0037,
+      "step": 582
+    },
+    {
+      "epoch": 1.7507507507507507,
+      "grad_norm": 0.216796875,
+      "learning_rate": 4.164164164164164e-06,
+      "loss": 0.9639,
+      "step": 583
+    },
+    {
+      "epoch": 1.7537537537537538,
+      "grad_norm": 0.2265625,
+      "learning_rate": 4.154154154154154e-06,
+      "loss": 0.9415,
+      "step": 584
+    },
+    {
+      "epoch": 1.7567567567567568,
+      "grad_norm": 0.29296875,
+      "learning_rate": 4.1441441441441446e-06,
+      "loss": 0.9258,
+      "step": 585
+    },
+    {
+      "epoch": 1.7597597597597598,
+      "grad_norm": 0.298828125,
+      "learning_rate": 4.134134134134135e-06,
+      "loss": 0.976,
+      "step": 586
+    },
+    {
+      "epoch": 1.7627627627627627,
+      "grad_norm": 0.2431640625,
+      "learning_rate": 4.124124124124124e-06,
+      "loss": 0.9293,
+      "step": 587
+    },
+    {
+      "epoch": 1.7657657657657657,
+      "grad_norm": 0.2333984375,
+      "learning_rate": 4.114114114114114e-06,
+      "loss": 0.9705,
+      "step": 588
+    },
+    {
+      "epoch": 1.7687687687687688,
+      "grad_norm": 0.2236328125,
+      "learning_rate": 4.1041041041041045e-06,
+      "loss": 0.9798,
+      "step": 589
+    },
+    {
+      "epoch": 1.7717717717717718,
+      "grad_norm": 0.2353515625,
+      "learning_rate": 4.094094094094095e-06,
+      "loss": 0.913,
+      "step": 590
+    },
+    {
+      "epoch": 1.7747747747747749,
+      "grad_norm": 0.220703125,
+      "learning_rate": 4.084084084084085e-06,
+      "loss": 0.9123,
+      "step": 591
+    },
+    {
+      "epoch": 1.7777777777777777,
+      "grad_norm": 0.220703125,
+      "learning_rate": 4.074074074074074e-06,
+      "loss": 0.9468,
+      "step": 592
+    },
+    {
+      "epoch": 1.7807807807807807,
+      "grad_norm": 0.2275390625,
+      "learning_rate": 4.064064064064064e-06,
+      "loss": 0.9565,
+      "step": 593
+    },
+    {
+      "epoch": 1.7837837837837838,
+      "grad_norm": 0.265625,
+      "learning_rate": 4.0540540540540545e-06,
+      "loss": 1.0266,
+      "step": 594
+    },
+    {
+      "epoch": 1.7867867867867868,
+      "grad_norm": 0.283203125,
+      "learning_rate": 4.044044044044044e-06,
+      "loss": 0.9144,
+      "step": 595
+    },
+    {
+      "epoch": 1.7897897897897899,
+      "grad_norm": 0.2373046875,
+      "learning_rate": 4.034034034034035e-06,
+      "loss": 0.9498,
+      "step": 596
+    },
+    {
+      "epoch": 1.7927927927927927,
+      "grad_norm": 0.2177734375,
+      "learning_rate": 4.024024024024024e-06,
+      "loss": 0.9722,
+      "step": 597
+    },
+    {
+      "epoch": 1.795795795795796,
+      "grad_norm": 0.2119140625,
+      "learning_rate": 4.014014014014014e-06,
+      "loss": 0.9294,
+      "step": 598
+    },
+    {
+      "epoch": 1.7987987987987988,
+      "grad_norm": 0.220703125,
+      "learning_rate": 4.004004004004005e-06,
+      "loss": 0.9385,
+      "step": 599
+    },
+    {
+      "epoch": 1.8018018018018018,
+      "grad_norm": 0.271484375,
+      "learning_rate": 3.993993993993994e-06,
+      "loss": 0.9427,
+      "step": 600
+    },
+    {
+      "epoch": 1.8048048048048049,
+      "grad_norm": 0.259765625,
+      "learning_rate": 3.983983983983984e-06,
+      "loss": 0.9611,
+      "step": 601
+    },
+    {
+      "epoch": 1.8078078078078077,
+      "grad_norm": 0.2734375,
+      "learning_rate": 3.973973973973974e-06,
+      "loss": 0.9794,
+      "step": 602
+    },
+    {
+      "epoch": 1.810810810810811,
+      "grad_norm": 0.234375,
+      "learning_rate": 3.9639639639639645e-06,
+      "loss": 0.9201,
+      "step": 603
+    },
+    {
+      "epoch": 1.8138138138138138,
+      "grad_norm": 0.2158203125,
+      "learning_rate": 3.953953953953955e-06,
+      "loss": 0.9977,
+      "step": 604
+    },
+    {
+      "epoch": 1.8168168168168168,
+      "grad_norm": 0.2197265625,
+      "learning_rate": 3.943943943943944e-06,
+      "loss": 0.9671,
+      "step": 605
+    },
+    {
+      "epoch": 1.8198198198198199,
+      "grad_norm": 0.228515625,
+      "learning_rate": 3.933933933933934e-06,
+      "loss": 0.9715,
+      "step": 606
+    },
+    {
+      "epoch": 1.8228228228228227,
+      "grad_norm": 0.2451171875,
+      "learning_rate": 3.923923923923924e-06,
+      "loss": 0.924,
+      "step": 607
+    },
+    {
+      "epoch": 1.825825825825826,
+      "grad_norm": 0.2275390625,
+      "learning_rate": 3.9139139139139145e-06,
+      "loss": 0.9593,
+      "step": 608
+    },
+    {
+      "epoch": 1.8288288288288288,
+      "grad_norm": 0.3828125,
+      "learning_rate": 3.903903903903904e-06,
+      "loss": 0.9516,
+      "step": 609
+    },
+    {
+      "epoch": 1.8318318318318318,
+      "grad_norm": 0.2353515625,
+      "learning_rate": 3.893893893893894e-06,
+      "loss": 0.9193,
+      "step": 610
+    },
+    {
+      "epoch": 1.834834834834835,
+      "grad_norm": 0.2255859375,
+      "learning_rate": 3.883883883883884e-06,
+      "loss": 0.9455,
+      "step": 611
+    },
+    {
+      "epoch": 1.8378378378378377,
+      "grad_norm": 0.271484375,
+      "learning_rate": 3.8738738738738744e-06,
+      "loss": 0.9487,
+      "step": 612
+    },
+    {
+      "epoch": 1.840840840840841,
+      "grad_norm": 0.255859375,
+      "learning_rate": 3.863863863863865e-06,
+      "loss": 0.9176,
+      "step": 613
+    },
+    {
+      "epoch": 1.8438438438438438,
+      "grad_norm": 0.2255859375,
+      "learning_rate": 3.853853853853854e-06,
+      "loss": 0.9437,
+      "step": 614
+    },
+    {
+      "epoch": 1.8468468468468469,
+      "grad_norm": 0.232421875,
+      "learning_rate": 3.843843843843844e-06,
+      "loss": 0.9615,
+      "step": 615
+    },
+    {
+      "epoch": 1.84984984984985,
+      "grad_norm": 0.2431640625,
+      "learning_rate": 3.833833833833834e-06,
+      "loss": 0.9702,
+      "step": 616
+    },
+    {
+      "epoch": 1.8528528528528527,
+      "grad_norm": 0.259765625,
+      "learning_rate": 3.823823823823824e-06,
+      "loss": 1.0011,
+      "step": 617
+    },
+    {
+      "epoch": 1.855855855855856,
+      "grad_norm": 0.23046875,
+      "learning_rate": 3.8138138138138143e-06,
+      "loss": 0.9659,
+      "step": 618
+    },
+    {
+      "epoch": 1.8588588588588588,
+      "grad_norm": 0.23046875,
+      "learning_rate": 3.803803803803804e-06,
+      "loss": 0.982,
+      "step": 619
+    },
+    {
+      "epoch": 1.8618618618618619,
+      "grad_norm": 0.22265625,
+      "learning_rate": 3.793793793793794e-06,
+      "loss": 0.9389,
+      "step": 620
+    },
+    {
+      "epoch": 1.864864864864865,
+      "grad_norm": 0.2255859375,
+      "learning_rate": 3.7837837837837844e-06,
+      "loss": 0.978,
+      "step": 621
+    },
+    {
+      "epoch": 1.8678678678678677,
+      "grad_norm": 0.2314453125,
+      "learning_rate": 3.773773773773774e-06,
+      "loss": 0.9151,
+      "step": 622
+    },
+    {
+      "epoch": 1.870870870870871,
+      "grad_norm": 0.287109375,
+      "learning_rate": 3.7637637637637643e-06,
+      "loss": 1.0688,
+      "step": 623
+    },
+    {
+      "epoch": 1.8738738738738738,
+      "grad_norm": 0.2255859375,
+      "learning_rate": 3.7537537537537537e-06,
+      "loss": 1.0145,
+      "step": 624
+    },
+    {
+      "epoch": 1.8768768768768769,
+      "grad_norm": 0.2109375,
+      "learning_rate": 3.743743743743744e-06,
+      "loss": 0.9682,
+      "step": 625
+    },
+    {
+      "epoch": 1.87987987987988,
+      "grad_norm": 0.2255859375,
+      "learning_rate": 3.7337337337337345e-06,
+      "loss": 0.9673,
+      "step": 626
+    },
+    {
+      "epoch": 1.8828828828828827,
+      "grad_norm": 0.2314453125,
+      "learning_rate": 3.723723723723724e-06,
+      "loss": 0.9561,
+      "step": 627
+    },
+    {
+      "epoch": 1.885885885885886,
+      "grad_norm": 0.224609375,
+      "learning_rate": 3.713713713713714e-06,
+      "loss": 0.9702,
+      "step": 628
+    },
+    {
+      "epoch": 1.8888888888888888,
+      "grad_norm": 0.2412109375,
+      "learning_rate": 3.7037037037037037e-06,
+      "loss": 1.0111,
+      "step": 629
+    },
+    {
+      "epoch": 1.8918918918918919,
+      "grad_norm": 0.2353515625,
+      "learning_rate": 3.693693693693694e-06,
+      "loss": 0.9257,
+      "step": 630
+    },
+    {
+      "epoch": 1.894894894894895,
+      "grad_norm": 0.2294921875,
+      "learning_rate": 3.683683683683684e-06,
+      "loss": 0.9265,
+      "step": 631
+    },
+    {
+      "epoch": 1.8978978978978978,
+      "grad_norm": 0.2197265625,
+      "learning_rate": 3.673673673673674e-06,
+      "loss": 0.9402,
+      "step": 632
+    },
+    {
+      "epoch": 1.900900900900901,
+      "grad_norm": 0.25390625,
+      "learning_rate": 3.663663663663664e-06,
+      "loss": 0.9279,
+      "step": 633
+    },
+    {
+      "epoch": 1.9039039039039038,
+      "grad_norm": 0.2138671875,
+      "learning_rate": 3.653653653653654e-06,
+      "loss": 0.9453,
+      "step": 634
+    },
+    {
+      "epoch": 1.906906906906907,
+      "grad_norm": 0.33984375,
+      "learning_rate": 3.643643643643644e-06,
+      "loss": 0.9368,
+      "step": 635
+    },
+    {
+      "epoch": 1.90990990990991,
+      "grad_norm": 0.2275390625,
+      "learning_rate": 3.633633633633634e-06,
+      "loss": 0.9691,
+      "step": 636
+    },
+    {
+      "epoch": 1.9129129129129128,
+      "grad_norm": 0.2578125,
+      "learning_rate": 3.623623623623624e-06,
+      "loss": 0.9543,
+      "step": 637
+    },
+    {
+      "epoch": 1.915915915915916,
+      "grad_norm": 0.224609375,
+      "learning_rate": 3.613613613613614e-06,
+      "loss": 0.9794,
+      "step": 638
+    },
+    {
+      "epoch": 1.9189189189189189,
+      "grad_norm": 0.28125,
+      "learning_rate": 3.603603603603604e-06,
+      "loss": 0.9814,
+      "step": 639
+    },
+    {
+      "epoch": 1.921921921921922,
+      "grad_norm": 0.2314453125,
+      "learning_rate": 3.593593593593594e-06,
+      "loss": 0.9506,
+      "step": 640
+    },
+    {
+      "epoch": 1.924924924924925,
+      "grad_norm": 0.419921875,
+      "learning_rate": 3.5835835835835834e-06,
+      "loss": 0.9039,
+      "step": 641
+    },
+    {
+      "epoch": 1.9279279279279278,
+      "grad_norm": 0.228515625,
+      "learning_rate": 3.573573573573574e-06,
+      "loss": 0.9294,
+      "step": 642
+    },
+    {
+      "epoch": 1.930930930930931,
+      "grad_norm": 0.232421875,
+      "learning_rate": 3.563563563563564e-06,
+      "loss": 0.9973,
+      "step": 643
+    },
+    {
+      "epoch": 1.9339339339339339,
+      "grad_norm": 0.3046875,
+      "learning_rate": 3.5535535535535535e-06,
+      "loss": 0.9538,
+      "step": 644
+    },
+    {
+      "epoch": 1.936936936936937,
+      "grad_norm": 0.263671875,
+      "learning_rate": 3.5435435435435437e-06,
+      "loss": 0.9527,
+      "step": 645
+    },
+    {
+      "epoch": 1.93993993993994,
+      "grad_norm": 0.234375,
+      "learning_rate": 3.5335335335335335e-06,
+      "loss": 0.9536,
+      "step": 646
+    },
+    {
+      "epoch": 1.9429429429429428,
+      "grad_norm": 0.23828125,
+      "learning_rate": 3.5235235235235237e-06,
+      "loss": 0.9765,
+      "step": 647
+    },
+    {
+      "epoch": 1.945945945945946,
+      "grad_norm": 0.314453125,
+      "learning_rate": 3.513513513513514e-06,
+      "loss": 0.9174,
+      "step": 648
+    },
+    {
+      "epoch": 1.9489489489489489,
+      "grad_norm": 0.298828125,
+      "learning_rate": 3.5035035035035036e-06,
+      "loss": 0.9799,
+      "step": 649
+    },
+    {
+      "epoch": 1.951951951951952,
+      "grad_norm": 0.2412109375,
+      "learning_rate": 3.4934934934934938e-06,
+      "loss": 0.9356,
+      "step": 650
+    },
+    {
+      "epoch": 1.954954954954955,
+      "grad_norm": 0.228515625,
+      "learning_rate": 3.4834834834834835e-06,
+      "loss": 0.9389,
+      "step": 651
+    },
+    {
+      "epoch": 1.9579579579579578,
+      "grad_norm": 0.2294921875,
+      "learning_rate": 3.4734734734734737e-06,
+      "loss": 0.9917,
+      "step": 652
+    },
+    {
+      "epoch": 1.960960960960961,
+      "grad_norm": 0.2275390625,
+      "learning_rate": 3.463463463463464e-06,
+      "loss": 0.9692,
+      "step": 653
+    },
+    {
+      "epoch": 1.9639639639639639,
+      "grad_norm": 0.23046875,
+      "learning_rate": 3.4534534534534537e-06,
+      "loss": 0.9994,
+      "step": 654
+    },
+    {
+      "epoch": 1.966966966966967,
+      "grad_norm": 0.26953125,
+      "learning_rate": 3.443443443443444e-06,
+      "loss": 0.938,
+      "step": 655
+    },
+    {
+      "epoch": 1.96996996996997,
+      "grad_norm": 0.23828125,
+      "learning_rate": 3.4334334334334336e-06,
+      "loss": 0.9452,
+      "step": 656
+    },
+    {
+      "epoch": 1.972972972972973,
+      "grad_norm": 0.220703125,
+      "learning_rate": 3.423423423423424e-06,
+      "loss": 0.9398,
+      "step": 657
+    },
+    {
+      "epoch": 1.975975975975976,
+      "grad_norm": 0.2275390625,
+      "learning_rate": 3.413413413413414e-06,
+      "loss": 0.9553,
+      "step": 658
+    },
+    {
+      "epoch": 1.978978978978979,
+      "grad_norm": 0.2333984375,
+      "learning_rate": 3.4034034034034037e-06,
+      "loss": 0.9565,
+      "step": 659
+    },
+    {
+      "epoch": 1.981981981981982,
+      "grad_norm": 0.2294921875,
+      "learning_rate": 3.393393393393394e-06,
+      "loss": 1.0094,
+      "step": 660
+    },
+    {
+      "epoch": 1.984984984984985,
+      "grad_norm": 0.2333984375,
+      "learning_rate": 3.3833833833833833e-06,
+      "loss": 0.9229,
+      "step": 661
+    },
+    {
+      "epoch": 1.987987987987988,
+      "grad_norm": 0.2578125,
+      "learning_rate": 3.373373373373374e-06,
+      "loss": 0.9503,
+      "step": 662
+    },
+    {
+      "epoch": 1.990990990990991,
+      "grad_norm": 0.2734375,
+      "learning_rate": 3.363363363363364e-06,
+      "loss": 0.9876,
+      "step": 663
+    },
+    {
+      "epoch": 1.993993993993994,
+      "grad_norm": 0.373046875,
+      "learning_rate": 3.3533533533533534e-06,
+      "loss": 0.929,
+      "step": 664
+    },
+    {
+      "epoch": 1.996996996996997,
+      "grad_norm": 0.27734375,
+      "learning_rate": 3.3433433433433436e-06,
+      "loss": 0.9345,
+      "step": 665
+    },
+    {
+      "epoch": 2.0,
+      "grad_norm": 0.240234375,
+      "learning_rate": 3.3333333333333333e-06,
+      "loss": 0.9648,
+      "step": 666
+    },
+    {
+      "epoch": 2.003003003003003,
+      "grad_norm": 0.3984375,
+      "learning_rate": 3.3233233233233235e-06,
+      "loss": 0.942,
+      "step": 667
+    },
+    {
+      "epoch": 2.006006006006006,
+      "grad_norm": 0.240234375,
+      "learning_rate": 3.3133133133133137e-06,
+      "loss": 0.9186,
+      "step": 668
+    },
+    {
+      "epoch": 2.009009009009009,
+      "grad_norm": 0.2294921875,
+      "learning_rate": 3.3033033033033035e-06,
+      "loss": 0.9342,
+      "step": 669
+    },
+    {
+      "epoch": 2.012012012012012,
+      "grad_norm": 0.275390625,
+      "learning_rate": 3.2932932932932936e-06,
+      "loss": 0.9412,
+      "step": 670
+    },
+    {
+      "epoch": 2.015015015015015,
+      "grad_norm": 0.26171875,
+      "learning_rate": 3.2832832832832834e-06,
+      "loss": 0.9512,
+      "step": 671
+    },
+    {
+      "epoch": 2.018018018018018,
+      "grad_norm": 0.2294921875,
+      "learning_rate": 3.2732732732732736e-06,
+      "loss": 0.9225,
+      "step": 672
+    },
+    {
+      "epoch": 2.021021021021021,
+      "grad_norm": 0.2333984375,
+      "learning_rate": 3.2632632632632633e-06,
+      "loss": 0.9135,
+      "step": 673
+    },
+    {
+      "epoch": 2.024024024024024,
+      "grad_norm": 0.2197265625,
+      "learning_rate": 3.2532532532532535e-06,
+      "loss": 0.9622,
+      "step": 674
+    },
+    {
+      "epoch": 2.027027027027027,
+      "grad_norm": 0.255859375,
+      "learning_rate": 3.2432432432432437e-06,
+      "loss": 0.9385,
+      "step": 675
+    },
+    {
+      "epoch": 2.03003003003003,
+      "grad_norm": 0.216796875,
+      "learning_rate": 3.2332332332332335e-06,
+      "loss": 0.9188,
+      "step": 676
+    },
+    {
+      "epoch": 2.033033033033033,
+      "grad_norm": 0.275390625,
+      "learning_rate": 3.2232232232232236e-06,
+      "loss": 0.957,
+      "step": 677
+    },
+    {
+      "epoch": 2.036036036036036,
+      "grad_norm": 0.28125,
+      "learning_rate": 3.2132132132132134e-06,
+      "loss": 0.9151,
+      "step": 678
+    },
+    {
+      "epoch": 2.039039039039039,
+      "grad_norm": 0.2392578125,
+      "learning_rate": 3.2032032032032036e-06,
+      "loss": 0.9553,
+      "step": 679
+    },
+    {
+      "epoch": 2.042042042042042,
+      "grad_norm": 0.2333984375,
+      "learning_rate": 3.1931931931931938e-06,
+      "loss": 0.9592,
+      "step": 680
+    },
+    {
+      "epoch": 2.045045045045045,
+      "grad_norm": 0.2314453125,
+      "learning_rate": 3.183183183183183e-06,
+      "loss": 0.9462,
+      "step": 681
+    },
+    {
+      "epoch": 2.048048048048048,
+      "grad_norm": 0.318359375,
+      "learning_rate": 3.1731731731731737e-06,
+      "loss": 0.9197,
+      "step": 682
+    },
+    {
+      "epoch": 2.051051051051051,
+      "grad_norm": 0.24609375,
+      "learning_rate": 3.163163163163163e-06,
+      "loss": 0.9209,
+      "step": 683
+    },
+    {
+      "epoch": 2.054054054054054,
+      "grad_norm": 0.2314453125,
+      "learning_rate": 3.1531531531531532e-06,
+      "loss": 0.9634,
+      "step": 684
+    },
+    {
+      "epoch": 2.057057057057057,
+      "grad_norm": 0.25,
+      "learning_rate": 3.1431431431431434e-06,
+      "loss": 1.0004,
+      "step": 685
+    },
+    {
+      "epoch": 2.06006006006006,
+      "grad_norm": 0.2314453125,
+      "learning_rate": 3.133133133133133e-06,
+      "loss": 0.9304,
+      "step": 686
+    },
+    {
+      "epoch": 2.063063063063063,
+      "grad_norm": 0.240234375,
+      "learning_rate": 3.1231231231231234e-06,
+      "loss": 0.9261,
+      "step": 687
+    },
+    {
+      "epoch": 2.066066066066066,
+      "grad_norm": 0.236328125,
+      "learning_rate": 3.113113113113113e-06,
+      "loss": 0.9122,
+      "step": 688
+    },
+    {
+      "epoch": 2.069069069069069,
+      "grad_norm": 0.220703125,
+      "learning_rate": 3.1031031031031033e-06,
+      "loss": 0.9646,
+      "step": 689
+    },
+    {
+      "epoch": 2.0720720720720722,
+      "grad_norm": 0.21875,
+      "learning_rate": 3.0930930930930935e-06,
+      "loss": 0.9422,
+      "step": 690
+    },
+    {
+      "epoch": 2.075075075075075,
+      "grad_norm": 0.2353515625,
+      "learning_rate": 3.0830830830830832e-06,
+      "loss": 0.9317,
+      "step": 691
+    },
+    {
+      "epoch": 2.078078078078078,
+      "grad_norm": 0.236328125,
+      "learning_rate": 3.0730730730730734e-06,
+      "loss": 0.961,
+      "step": 692
+    },
+    {
+      "epoch": 2.081081081081081,
+      "grad_norm": 0.34375,
+      "learning_rate": 3.063063063063063e-06,
+      "loss": 0.9541,
+      "step": 693
+    },
+    {
+      "epoch": 2.084084084084084,
+      "grad_norm": 0.2333984375,
+      "learning_rate": 3.0530530530530534e-06,
+      "loss": 0.8907,
+      "step": 694
+    },
+    {
+      "epoch": 2.0870870870870872,
+      "grad_norm": 0.28515625,
+      "learning_rate": 3.0430430430430436e-06,
+      "loss": 0.9163,
+      "step": 695
+    },
+    {
+      "epoch": 2.09009009009009,
+      "grad_norm": 0.2734375,
+      "learning_rate": 3.0330330330330333e-06,
+      "loss": 1.051,
+      "step": 696
+    },
+    {
+      "epoch": 2.093093093093093,
+      "grad_norm": 0.2138671875,
+      "learning_rate": 3.0230230230230235e-06,
+      "loss": 0.9295,
+      "step": 697
+    },
+    {
+      "epoch": 2.096096096096096,
+      "grad_norm": 0.625,
+      "learning_rate": 3.0130130130130133e-06,
+      "loss": 0.9177,
+      "step": 698
+    },
+    {
+      "epoch": 2.099099099099099,
+      "grad_norm": 0.2333984375,
+      "learning_rate": 3.0030030030030034e-06,
+      "loss": 0.9697,
+      "step": 699
+    },
+    {
+      "epoch": 2.1021021021021022,
+      "grad_norm": 0.2490234375,
+      "learning_rate": 2.9929929929929936e-06,
+      "loss": 0.906,
+      "step": 700
+    },
+    {
+      "epoch": 2.105105105105105,
+      "grad_norm": 0.2890625,
+      "learning_rate": 2.982982982982983e-06,
+      "loss": 0.8809,
+      "step": 701
+    },
+    {
+      "epoch": 2.108108108108108,
+      "grad_norm": 0.2353515625,
+      "learning_rate": 2.9729729729729736e-06,
+      "loss": 0.9175,
+      "step": 702
+    },
+    {
+      "epoch": 2.111111111111111,
+      "grad_norm": 0.2451171875,
+      "learning_rate": 2.962962962962963e-06,
+      "loss": 0.926,
+      "step": 703
+    },
+    {
+      "epoch": 2.114114114114114,
+      "grad_norm": 0.27734375,
+      "learning_rate": 2.952952952952953e-06,
+      "loss": 0.9402,
+      "step": 704
+    },
+    {
+      "epoch": 2.1171171171171173,
+      "grad_norm": 0.2265625,
+      "learning_rate": 2.942942942942943e-06,
+      "loss": 0.8922,
+      "step": 705
+    },
+    {
+      "epoch": 2.12012012012012,
+      "grad_norm": 0.2392578125,
+      "learning_rate": 2.932932932932933e-06,
+      "loss": 0.9533,
+      "step": 706
+    },
+    {
+      "epoch": 2.123123123123123,
+      "grad_norm": 0.32421875,
+      "learning_rate": 2.9229229229229232e-06,
+      "loss": 1.0124,
+      "step": 707
+    },
+    {
+      "epoch": 2.126126126126126,
+      "grad_norm": 0.23046875,
+      "learning_rate": 2.912912912912913e-06,
+      "loss": 0.9754,
+      "step": 708
+    },
+    {
+      "epoch": 2.129129129129129,
+      "grad_norm": 0.2578125,
+      "learning_rate": 2.902902902902903e-06,
+      "loss": 0.9586,
+      "step": 709
+    },
+    {
+      "epoch": 2.1321321321321323,
+      "grad_norm": 0.2392578125,
+      "learning_rate": 2.892892892892893e-06,
+      "loss": 1.0196,
+      "step": 710
+    },
+    {
+      "epoch": 2.135135135135135,
+      "grad_norm": 0.22265625,
+      "learning_rate": 2.882882882882883e-06,
+      "loss": 0.955,
+      "step": 711
+    },
+    {
+      "epoch": 2.138138138138138,
+      "grad_norm": 0.2314453125,
+      "learning_rate": 2.8728728728728733e-06,
+      "loss": 0.9455,
+      "step": 712
+    },
+    {
+      "epoch": 2.141141141141141,
+      "grad_norm": 0.251953125,
+      "learning_rate": 2.862862862862863e-06,
+      "loss": 0.9567,
+      "step": 713
+    },
+    {
+      "epoch": 2.144144144144144,
+      "grad_norm": 0.2421875,
+      "learning_rate": 2.8528528528528532e-06,
+      "loss": 0.9519,
+      "step": 714
+    },
+    {
+      "epoch": 2.1471471471471473,
+      "grad_norm": 0.2265625,
+      "learning_rate": 2.842842842842843e-06,
+      "loss": 0.9927,
+      "step": 715
+    },
+    {
+      "epoch": 2.15015015015015,
+      "grad_norm": 0.2412109375,
+      "learning_rate": 2.832832832832833e-06,
+      "loss": 0.9362,
+      "step": 716
+    },
+    {
+      "epoch": 2.153153153153153,
+      "grad_norm": 0.2392578125,
+      "learning_rate": 2.8228228228228234e-06,
+      "loss": 0.9658,
+      "step": 717
+    },
+    {
+      "epoch": 2.156156156156156,
+      "grad_norm": 0.2333984375,
+      "learning_rate": 2.812812812812813e-06,
+      "loss": 0.9555,
+      "step": 718
+    },
+    {
+      "epoch": 2.159159159159159,
+      "grad_norm": 0.21875,
+      "learning_rate": 2.8028028028028033e-06,
+      "loss": 0.9543,
+      "step": 719
+    },
+    {
+      "epoch": 2.1621621621621623,
+      "grad_norm": 0.236328125,
+      "learning_rate": 2.7927927927927926e-06,
+      "loss": 0.922,
+      "step": 720
+    },
+    {
+      "epoch": 2.165165165165165,
+      "grad_norm": 0.224609375,
+      "learning_rate": 2.782782782782783e-06,
+      "loss": 0.9602,
+      "step": 721
+    },
+    {
+      "epoch": 2.1681681681681684,
+      "grad_norm": 0.2197265625,
+      "learning_rate": 2.7727727727727734e-06,
+      "loss": 0.924,
+      "step": 722
+    },
+    {
+      "epoch": 2.171171171171171,
+      "grad_norm": 0.298828125,
+      "learning_rate": 2.7627627627627628e-06,
+      "loss": 0.893,
+      "step": 723
+    },
+    {
+      "epoch": 2.174174174174174,
+      "grad_norm": 0.2265625,
+      "learning_rate": 2.752752752752753e-06,
+      "loss": 0.9968,
+      "step": 724
+    },
+    {
+      "epoch": 2.1771771771771773,
+      "grad_norm": 0.236328125,
+      "learning_rate": 2.7427427427427427e-06,
+      "loss": 0.933,
+      "step": 725
+    },
+    {
+      "epoch": 2.18018018018018,
+      "grad_norm": 0.283203125,
+      "learning_rate": 2.732732732732733e-06,
+      "loss": 0.941,
+      "step": 726
+    },
+    {
+      "epoch": 2.1831831831831834,
+      "grad_norm": 0.37109375,
+      "learning_rate": 2.722722722722723e-06,
+      "loss": 0.9271,
+      "step": 727
+    },
+    {
+      "epoch": 2.186186186186186,
+      "grad_norm": 0.2314453125,
+      "learning_rate": 2.712712712712713e-06,
+      "loss": 0.9313,
+      "step": 728
+    },
+    {
+      "epoch": 2.189189189189189,
+      "grad_norm": 0.251953125,
+      "learning_rate": 2.702702702702703e-06,
+      "loss": 0.9875,
+      "step": 729
+    },
+    {
+      "epoch": 2.1921921921921923,
+      "grad_norm": 0.25390625,
+      "learning_rate": 2.6926926926926928e-06,
+      "loss": 0.9949,
+      "step": 730
+    },
+    {
+      "epoch": 2.195195195195195,
+      "grad_norm": 0.232421875,
+      "learning_rate": 2.682682682682683e-06,
+      "loss": 0.9957,
+      "step": 731
+    },
+    {
+      "epoch": 2.1981981981981984,
+      "grad_norm": 0.302734375,
+      "learning_rate": 2.672672672672673e-06,
+      "loss": 0.936,
+      "step": 732
+    },
+    {
+      "epoch": 2.201201201201201,
+      "grad_norm": 0.2392578125,
+      "learning_rate": 2.662662662662663e-06,
+      "loss": 0.9227,
+      "step": 733
+    },
+    {
+      "epoch": 2.204204204204204,
+      "grad_norm": 0.236328125,
+      "learning_rate": 2.652652652652653e-06,
+      "loss": 0.9309,
+      "step": 734
+    },
+    {
+      "epoch": 2.2072072072072073,
+      "grad_norm": 0.232421875,
+      "learning_rate": 2.642642642642643e-06,
+      "loss": 0.9482,
+      "step": 735
+    },
+    {
+      "epoch": 2.21021021021021,
+      "grad_norm": 0.244140625,
+      "learning_rate": 2.632632632632633e-06,
+      "loss": 0.9681,
+      "step": 736
+    },
+    {
+      "epoch": 2.2132132132132134,
+      "grad_norm": 0.2216796875,
+      "learning_rate": 2.6226226226226224e-06,
+      "loss": 0.9192,
+      "step": 737
+    },
+    {
+      "epoch": 2.2162162162162162,
+      "grad_norm": 0.2578125,
+      "learning_rate": 2.612612612612613e-06,
+      "loss": 0.9129,
+      "step": 738
+    },
+    {
+      "epoch": 2.219219219219219,
+      "grad_norm": 0.232421875,
+      "learning_rate": 2.602602602602603e-06,
+      "loss": 1.0269,
+      "step": 739
+    },
+    {
+      "epoch": 2.2222222222222223,
+      "grad_norm": 0.25,
+      "learning_rate": 2.5925925925925925e-06,
+      "loss": 0.8884,
+      "step": 740
+    },
+    {
+      "epoch": 2.225225225225225,
+      "grad_norm": 0.294921875,
+      "learning_rate": 2.5825825825825827e-06,
+      "loss": 0.9719,
+      "step": 741
+    },
+    {
+      "epoch": 2.2282282282282284,
+      "grad_norm": 0.2216796875,
+      "learning_rate": 2.5725725725725724e-06,
+      "loss": 0.891,
+      "step": 742
+    },
+    {
+      "epoch": 2.2312312312312312,
+      "grad_norm": 0.2314453125,
+      "learning_rate": 2.5625625625625626e-06,
+      "loss": 0.9652,
+      "step": 743
+    },
+    {
+      "epoch": 2.234234234234234,
+      "grad_norm": 0.26953125,
+      "learning_rate": 2.552552552552553e-06,
+      "loss": 0.985,
+      "step": 744
+    },
+    {
+      "epoch": 2.2372372372372373,
+      "grad_norm": 0.2197265625,
+      "learning_rate": 2.5425425425425426e-06,
+      "loss": 0.9171,
+      "step": 745
+    },
+    {
+      "epoch": 2.24024024024024,
+      "grad_norm": 0.306640625,
+      "learning_rate": 2.5325325325325327e-06,
+      "loss": 0.8712,
+      "step": 746
+    },
+    {
+      "epoch": 2.2432432432432434,
+      "grad_norm": 0.2275390625,
+      "learning_rate": 2.5225225225225225e-06,
+      "loss": 0.9052,
+      "step": 747
+    },
+    {
+      "epoch": 2.2462462462462462,
+      "grad_norm": 0.232421875,
+      "learning_rate": 2.5125125125125127e-06,
+      "loss": 0.9449,
+      "step": 748
+    },
+    {
+      "epoch": 2.249249249249249,
+      "grad_norm": 0.353515625,
+      "learning_rate": 2.502502502502503e-06,
+      "loss": 0.909,
+      "step": 749
+    },
+    {
+      "epoch": 2.2522522522522523,
+      "grad_norm": 0.251953125,
+      "learning_rate": 2.4924924924924926e-06,
+      "loss": 0.9378,
+      "step": 750
+    },
+    {
+      "epoch": 2.255255255255255,
+      "grad_norm": 0.265625,
+      "learning_rate": 2.482482482482483e-06,
+      "loss": 0.936,
+      "step": 751
+    },
+    {
+      "epoch": 2.2582582582582584,
+      "grad_norm": 0.2265625,
+      "learning_rate": 2.4724724724724726e-06,
+      "loss": 0.9718,
+      "step": 752
+    },
+    {
+      "epoch": 2.2612612612612613,
+      "grad_norm": 0.376953125,
+      "learning_rate": 2.4624624624624628e-06,
+      "loss": 0.9267,
+      "step": 753
+    },
+    {
+      "epoch": 2.264264264264264,
+      "grad_norm": 0.2275390625,
+      "learning_rate": 2.4524524524524525e-06,
+      "loss": 0.9239,
+      "step": 754
+    },
+    {
+      "epoch": 2.2672672672672673,
+      "grad_norm": 0.33203125,
+      "learning_rate": 2.4424424424424427e-06,
+      "loss": 0.9231,
+      "step": 755
+    },
+    {
+      "epoch": 2.27027027027027,
+      "grad_norm": 0.26171875,
+      "learning_rate": 2.432432432432433e-06,
+      "loss": 0.9088,
+      "step": 756
+    },
+    {
+      "epoch": 2.2732732732732734,
+      "grad_norm": 0.23046875,
+      "learning_rate": 2.4224224224224226e-06,
+      "loss": 0.9087,
+      "step": 757
+    },
+    {
+      "epoch": 2.2762762762762763,
+      "grad_norm": 0.2236328125,
+      "learning_rate": 2.412412412412413e-06,
+      "loss": 0.9487,
+      "step": 758
+    },
+    {
+      "epoch": 2.279279279279279,
+      "grad_norm": 0.232421875,
+      "learning_rate": 2.4024024024024026e-06,
+      "loss": 1.0017,
+      "step": 759
+    },
+    {
+      "epoch": 2.2822822822822824,
+      "grad_norm": 0.25,
+      "learning_rate": 2.3923923923923923e-06,
+      "loss": 0.9996,
+      "step": 760
+    },
+    {
+      "epoch": 2.285285285285285,
+      "grad_norm": 0.294921875,
+      "learning_rate": 2.3823823823823825e-06,
+      "loss": 0.9802,
+      "step": 761
+    },
+    {
+      "epoch": 2.2882882882882885,
+      "grad_norm": 0.23828125,
+      "learning_rate": 2.3723723723723727e-06,
+      "loss": 0.9465,
+      "step": 762
+    },
+    {
+      "epoch": 2.2912912912912913,
+      "grad_norm": 0.244140625,
+      "learning_rate": 2.3623623623623625e-06,
+      "loss": 0.8918,
+      "step": 763
+    },
+    {
+      "epoch": 2.294294294294294,
+      "grad_norm": 0.2255859375,
+      "learning_rate": 2.3523523523523527e-06,
+      "loss": 0.931,
+      "step": 764
+    },
+    {
+      "epoch": 2.2972972972972974,
+      "grad_norm": 0.228515625,
+      "learning_rate": 2.3423423423423424e-06,
+      "loss": 0.9194,
+      "step": 765
+    },
+    {
+      "epoch": 2.3003003003003,
+      "grad_norm": 0.236328125,
+      "learning_rate": 2.3323323323323326e-06,
+      "loss": 0.9348,
+      "step": 766
+    },
+    {
+      "epoch": 2.3033033033033035,
+      "grad_norm": 0.2314453125,
+      "learning_rate": 2.3223223223223228e-06,
+      "loss": 0.9396,
+      "step": 767
+    },
+    {
+      "epoch": 2.3063063063063063,
+      "grad_norm": 0.236328125,
+      "learning_rate": 2.3123123123123125e-06,
+      "loss": 0.9705,
+      "step": 768
+    },
+    {
+      "epoch": 2.3093093093093096,
+      "grad_norm": 0.23046875,
+      "learning_rate": 2.3023023023023023e-06,
+      "loss": 0.9573,
+      "step": 769
+    },
+    {
+      "epoch": 2.3123123123123124,
+      "grad_norm": 0.337890625,
+      "learning_rate": 2.2922922922922925e-06,
+      "loss": 0.9243,
+      "step": 770
+    },
+    {
+      "epoch": 2.315315315315315,
+      "grad_norm": 0.2470703125,
+      "learning_rate": 2.2822822822822822e-06,
+      "loss": 0.9929,
+      "step": 771
+    },
+    {
+      "epoch": 2.3183183183183185,
+      "grad_norm": 0.2294921875,
+      "learning_rate": 2.2722722722722724e-06,
+      "loss": 0.9517,
+      "step": 772
+    },
+    {
+      "epoch": 2.3213213213213213,
+      "grad_norm": 0.2451171875,
+      "learning_rate": 2.2622622622622626e-06,
+      "loss": 0.9442,
+      "step": 773
+    },
+    {
+      "epoch": 2.3243243243243246,
+      "grad_norm": 0.220703125,
+      "learning_rate": 2.2522522522522524e-06,
+      "loss": 0.9385,
+      "step": 774
+    },
+    {
+      "epoch": 2.3273273273273274,
+      "grad_norm": 0.2236328125,
+      "learning_rate": 2.2422422422422426e-06,
+      "loss": 0.9035,
+      "step": 775
+    },
+    {
+      "epoch": 2.33033033033033,
+      "grad_norm": 0.3203125,
+      "learning_rate": 2.2322322322322323e-06,
+      "loss": 0.9366,
+      "step": 776
+    },
+    {
+      "epoch": 2.3333333333333335,
+      "grad_norm": 0.2255859375,
+      "learning_rate": 2.222222222222222e-06,
+      "loss": 0.9169,
+      "step": 777
+    },
+    {
+      "epoch": 2.3363363363363363,
+      "grad_norm": 0.2421875,
+      "learning_rate": 2.2122122122122127e-06,
+      "loss": 0.9368,
+      "step": 778
+    },
+    {
+      "epoch": 2.3393393393393396,
+      "grad_norm": 0.267578125,
+      "learning_rate": 2.2022022022022024e-06,
+      "loss": 0.9596,
+      "step": 779
+    },
+    {
+      "epoch": 2.3423423423423424,
+      "grad_norm": 0.2216796875,
+      "learning_rate": 2.192192192192192e-06,
+      "loss": 0.9477,
+      "step": 780
+    },
+    {
+      "epoch": 2.3453453453453452,
+      "grad_norm": 0.263671875,
+      "learning_rate": 2.1821821821821824e-06,
+      "loss": 0.9411,
+      "step": 781
+    },
+    {
+      "epoch": 2.3483483483483485,
+      "grad_norm": 0.2265625,
+      "learning_rate": 2.172172172172172e-06,
+      "loss": 0.9744,
+      "step": 782
+    },
+    {
+      "epoch": 2.3513513513513513,
+      "grad_norm": 0.26953125,
+      "learning_rate": 2.1621621621621623e-06,
+      "loss": 0.9418,
+      "step": 783
+    },
+    {
+      "epoch": 2.3543543543543546,
+      "grad_norm": 0.2236328125,
+      "learning_rate": 2.1521521521521525e-06,
+      "loss": 0.9063,
+      "step": 784
+    },
+    {
+      "epoch": 2.3573573573573574,
+      "grad_norm": 0.2421875,
+      "learning_rate": 2.1421421421421423e-06,
+      "loss": 0.9218,
+      "step": 785
+    },
+    {
+      "epoch": 2.3603603603603602,
+      "grad_norm": 0.2421875,
+      "learning_rate": 2.1321321321321325e-06,
+      "loss": 0.9419,
+      "step": 786
+    },
+    {
+      "epoch": 2.3633633633633635,
+      "grad_norm": 0.2431640625,
+      "learning_rate": 2.1221221221221222e-06,
+      "loss": 0.8884,
+      "step": 787
+    },
+    {
+      "epoch": 2.3663663663663663,
+      "grad_norm": 0.2314453125,
+      "learning_rate": 2.1121121121121124e-06,
+      "loss": 0.9619,
+      "step": 788
+    },
+    {
+      "epoch": 2.3693693693693696,
+      "grad_norm": 0.2216796875,
+      "learning_rate": 2.102102102102102e-06,
+      "loss": 0.9612,
+      "step": 789
+    },
+    {
+      "epoch": 2.3723723723723724,
+      "grad_norm": 0.2353515625,
+      "learning_rate": 2.0920920920920923e-06,
+      "loss": 0.9509,
+      "step": 790
+    },
+    {
+      "epoch": 2.3753753753753752,
+      "grad_norm": 0.2138671875,
+      "learning_rate": 2.082082082082082e-06,
+      "loss": 0.929,
+      "step": 791
+    },
+    {
+      "epoch": 2.3783783783783785,
+      "grad_norm": 0.2578125,
+      "learning_rate": 2.0720720720720723e-06,
+      "loss": 0.9511,
+      "step": 792
+    },
+    {
+      "epoch": 2.3813813813813813,
+      "grad_norm": 0.2412109375,
+      "learning_rate": 2.062062062062062e-06,
+      "loss": 0.9869,
+      "step": 793
+    },
+    {
+      "epoch": 2.3843843843843846,
+      "grad_norm": 0.2333984375,
+      "learning_rate": 2.0520520520520522e-06,
+      "loss": 0.9151,
+      "step": 794
+    },
+    {
+      "epoch": 2.3873873873873874,
+      "grad_norm": 0.2314453125,
+      "learning_rate": 2.0420420420420424e-06,
+      "loss": 0.9481,
+      "step": 795
+    },
+    {
+      "epoch": 2.3903903903903903,
+      "grad_norm": 0.2294921875,
+      "learning_rate": 2.032032032032032e-06,
+      "loss": 0.9263,
+      "step": 796
+    },
+    {
+      "epoch": 2.3933933933933935,
+      "grad_norm": 0.296875,
+      "learning_rate": 2.022022022022022e-06,
+      "loss": 0.9165,
+      "step": 797
+    },
+    {
+      "epoch": 2.3963963963963963,
+      "grad_norm": 0.2275390625,
+      "learning_rate": 2.012012012012012e-06,
+      "loss": 0.9355,
+      "step": 798
+    },
+    {
+      "epoch": 2.3993993993993996,
+      "grad_norm": 0.32421875,
+      "learning_rate": 2.0020020020020023e-06,
+      "loss": 0.932,
+      "step": 799
+    },
+    {
+      "epoch": 2.4024024024024024,
+      "grad_norm": 0.2275390625,
+      "learning_rate": 1.991991991991992e-06,
+      "loss": 0.8926,
+      "step": 800
+    },
+    {
+      "epoch": 2.4054054054054053,
+      "grad_norm": 0.22265625,
+      "learning_rate": 1.9819819819819822e-06,
+      "loss": 0.9439,
+      "step": 801
+    },
+    {
+      "epoch": 2.4084084084084085,
+      "grad_norm": 0.2255859375,
+      "learning_rate": 1.971971971971972e-06,
+      "loss": 0.9198,
+      "step": 802
+    },
+    {
+      "epoch": 2.4114114114114114,
+      "grad_norm": 0.24609375,
+      "learning_rate": 1.961961961961962e-06,
+      "loss": 0.9531,
+      "step": 803
+    },
+    {
+      "epoch": 2.4144144144144146,
+      "grad_norm": 0.23046875,
+      "learning_rate": 1.951951951951952e-06,
+      "loss": 1.0069,
+      "step": 804
+    },
+    {
+      "epoch": 2.4174174174174174,
+      "grad_norm": 0.23046875,
+      "learning_rate": 1.941941941941942e-06,
+      "loss": 0.9741,
+      "step": 805
+    },
+    {
+      "epoch": 2.4204204204204203,
+      "grad_norm": 0.28515625,
+      "learning_rate": 1.9319319319319323e-06,
+      "loss": 0.9224,
+      "step": 806
+    },
+    {
+      "epoch": 2.4234234234234235,
+      "grad_norm": 0.234375,
+      "learning_rate": 1.921921921921922e-06,
+      "loss": 0.9471,
+      "step": 807
+    },
+    {
+      "epoch": 2.4264264264264264,
+      "grad_norm": 0.248046875,
+      "learning_rate": 1.911911911911912e-06,
+      "loss": 0.971,
+      "step": 808
+    },
+    {
+      "epoch": 2.4294294294294296,
+      "grad_norm": 0.31640625,
+      "learning_rate": 1.901901901901902e-06,
+      "loss": 0.8754,
+      "step": 809
+    },
+    {
+      "epoch": 2.4324324324324325,
+      "grad_norm": 0.224609375,
+      "learning_rate": 1.8918918918918922e-06,
+      "loss": 0.9106,
+      "step": 810
+    },
+    {
+      "epoch": 2.4354354354354353,
+      "grad_norm": 0.2255859375,
+      "learning_rate": 1.8818818818818822e-06,
+      "loss": 0.9086,
+      "step": 811
+    },
+    {
+      "epoch": 2.4384384384384385,
+      "grad_norm": 0.2294921875,
+      "learning_rate": 1.871871871871872e-06,
+      "loss": 0.9803,
+      "step": 812
+    },
+    {
+      "epoch": 2.4414414414414414,
+      "grad_norm": 0.345703125,
+      "learning_rate": 1.861861861861862e-06,
+      "loss": 0.9193,
+      "step": 813
+    },
+    {
+      "epoch": 2.4444444444444446,
+      "grad_norm": 0.2333984375,
+      "learning_rate": 1.8518518518518519e-06,
+      "loss": 0.9394,
+      "step": 814
+    },
+    {
+      "epoch": 2.4474474474474475,
+      "grad_norm": 0.2275390625,
+      "learning_rate": 1.841841841841842e-06,
+      "loss": 0.9627,
+      "step": 815
+    },
+    {
+      "epoch": 2.4504504504504503,
+      "grad_norm": 0.390625,
+      "learning_rate": 1.831831831831832e-06,
+      "loss": 0.9574,
+      "step": 816
+    },
+    {
+      "epoch": 2.4534534534534536,
+      "grad_norm": 0.23828125,
+      "learning_rate": 1.821821821821822e-06,
+      "loss": 0.946,
+      "step": 817
+    },
+    {
+      "epoch": 2.4564564564564564,
+      "grad_norm": 0.232421875,
+      "learning_rate": 1.811811811811812e-06,
+      "loss": 0.8872,
+      "step": 818
+    },
+    {
+      "epoch": 2.4594594594594597,
+      "grad_norm": 0.232421875,
+      "learning_rate": 1.801801801801802e-06,
+      "loss": 0.9368,
+      "step": 819
+    },
+    {
+      "epoch": 2.4624624624624625,
+      "grad_norm": 0.232421875,
+      "learning_rate": 1.7917917917917917e-06,
+      "loss": 0.9225,
+      "step": 820
+    },
+    {
+      "epoch": 2.4654654654654653,
+      "grad_norm": 0.322265625,
+      "learning_rate": 1.781781781781782e-06,
+      "loss": 0.9093,
+      "step": 821
+    },
+    {
+      "epoch": 2.4684684684684686,
+      "grad_norm": 0.224609375,
+      "learning_rate": 1.7717717717717719e-06,
+      "loss": 0.9315,
+      "step": 822
+    },
+    {
+      "epoch": 2.4714714714714714,
+      "grad_norm": 0.228515625,
+      "learning_rate": 1.7617617617617618e-06,
+      "loss": 0.9838,
+      "step": 823
+    },
+    {
+      "epoch": 2.4744744744744747,
+      "grad_norm": 0.2490234375,
+      "learning_rate": 1.7517517517517518e-06,
+      "loss": 0.9513,
+      "step": 824
+    },
+    {
+      "epoch": 2.4774774774774775,
+      "grad_norm": 0.310546875,
+      "learning_rate": 1.7417417417417418e-06,
+      "loss": 0.8773,
+      "step": 825
+    },
+    {
+      "epoch": 2.4804804804804803,
+      "grad_norm": 0.2216796875,
+      "learning_rate": 1.731731731731732e-06,
+      "loss": 0.9534,
+      "step": 826
+    },
+    {
+      "epoch": 2.4834834834834836,
+      "grad_norm": 0.2265625,
+      "learning_rate": 1.721721721721722e-06,
+      "loss": 0.9282,
+      "step": 827
+    },
+    {
+      "epoch": 2.4864864864864864,
+      "grad_norm": 0.2451171875,
+      "learning_rate": 1.711711711711712e-06,
+      "loss": 0.8995,
+      "step": 828
+    },
+    {
+      "epoch": 2.4894894894894897,
+      "grad_norm": 0.2314453125,
+      "learning_rate": 1.7017017017017019e-06,
+      "loss": 0.9753,
+      "step": 829
+    },
+    {
+      "epoch": 2.4924924924924925,
+      "grad_norm": 0.310546875,
+      "learning_rate": 1.6916916916916916e-06,
+      "loss": 0.8878,
+      "step": 830
+    },
+    {
+      "epoch": 2.4954954954954953,
+      "grad_norm": 0.369140625,
+      "learning_rate": 1.681681681681682e-06,
+      "loss": 0.8518,
+      "step": 831
+    },
+    {
+      "epoch": 2.4984984984984986,
+      "grad_norm": 0.2265625,
+      "learning_rate": 1.6716716716716718e-06,
+      "loss": 0.926,
+      "step": 832
+    },
+    {
+      "epoch": 2.5015015015015014,
+      "grad_norm": 0.2373046875,
+      "learning_rate": 1.6616616616616618e-06,
+      "loss": 0.98,
+      "step": 833
+    },
+    {
+      "epoch": 2.5045045045045047,
+      "grad_norm": 0.23046875,
+      "learning_rate": 1.6516516516516517e-06,
+      "loss": 1.0084,
+      "step": 834
+    },
+    {
+      "epoch": 2.5075075075075075,
+      "grad_norm": 0.28515625,
+      "learning_rate": 1.6416416416416417e-06,
+      "loss": 0.9221,
+      "step": 835
+    },
+    {
+      "epoch": 2.5105105105105103,
+      "grad_norm": 0.236328125,
+      "learning_rate": 1.6316316316316317e-06,
+      "loss": 0.9311,
+      "step": 836
+    },
+    {
+      "epoch": 2.5135135135135136,
+      "grad_norm": 0.2255859375,
+      "learning_rate": 1.6216216216216219e-06,
+      "loss": 0.9595,
+      "step": 837
+    },
+    {
+      "epoch": 2.5165165165165164,
+      "grad_norm": 0.2314453125,
+      "learning_rate": 1.6116116116116118e-06,
+      "loss": 0.9407,
+      "step": 838
+    },
+    {
+      "epoch": 2.5195195195195197,
+      "grad_norm": 0.2255859375,
+      "learning_rate": 1.6016016016016018e-06,
+      "loss": 0.9109,
+      "step": 839
+    },
+    {
+      "epoch": 2.5225225225225225,
+      "grad_norm": 0.23046875,
+      "learning_rate": 1.5915915915915916e-06,
+      "loss": 0.9342,
+      "step": 840
+    },
+    {
+      "epoch": 2.5255255255255253,
+      "grad_norm": 0.232421875,
+      "learning_rate": 1.5815815815815815e-06,
+      "loss": 0.9442,
+      "step": 841
+    },
+    {
+      "epoch": 2.5285285285285286,
+      "grad_norm": 0.2294921875,
+      "learning_rate": 1.5715715715715717e-06,
+      "loss": 0.9587,
+      "step": 842
+    },
+    {
+      "epoch": 2.5315315315315314,
+      "grad_norm": 0.2314453125,
+      "learning_rate": 1.5615615615615617e-06,
+      "loss": 0.946,
+      "step": 843
+    },
+    {
+      "epoch": 2.5345345345345347,
+      "grad_norm": 0.21875,
+      "learning_rate": 1.5515515515515517e-06,
+      "loss": 0.9373,
+      "step": 844
+    },
+    {
+      "epoch": 2.5375375375375375,
+      "grad_norm": 0.2275390625,
+      "learning_rate": 1.5415415415415416e-06,
+      "loss": 0.9752,
+      "step": 845
+    },
+    {
+      "epoch": 2.5405405405405403,
+      "grad_norm": 0.2353515625,
+      "learning_rate": 1.5315315315315316e-06,
+      "loss": 0.9618,
+      "step": 846
+    },
+    {
+      "epoch": 2.5435435435435436,
+      "grad_norm": 0.2412109375,
+      "learning_rate": 1.5215215215215218e-06,
+      "loss": 0.9026,
+      "step": 847
+    },
+    {
+      "epoch": 2.5465465465465464,
+      "grad_norm": 0.2236328125,
+      "learning_rate": 1.5115115115115118e-06,
+      "loss": 0.9063,
+      "step": 848
+    },
+    {
+      "epoch": 2.5495495495495497,
+      "grad_norm": 0.474609375,
+      "learning_rate": 1.5015015015015017e-06,
+      "loss": 0.8977,
+      "step": 849
+    },
+    {
+      "epoch": 2.5525525525525525,
+      "grad_norm": 0.369140625,
+      "learning_rate": 1.4914914914914915e-06,
+      "loss": 0.9089,
+      "step": 850
+    },
+    {
+      "epoch": 2.5555555555555554,
+      "grad_norm": 0.263671875,
+      "learning_rate": 1.4814814814814815e-06,
+      "loss": 0.9708,
+      "step": 851
+    },
+    {
+      "epoch": 2.5585585585585586,
+      "grad_norm": 0.2275390625,
+      "learning_rate": 1.4714714714714714e-06,
+      "loss": 0.9526,
+      "step": 852
+    },
+    {
+      "epoch": 2.5615615615615615,
+      "grad_norm": 0.236328125,
+      "learning_rate": 1.4614614614614616e-06,
+      "loss": 0.9758,
+      "step": 853
+    },
+    {
+      "epoch": 2.5645645645645647,
+      "grad_norm": 0.2451171875,
+      "learning_rate": 1.4514514514514516e-06,
+      "loss": 0.9308,
+      "step": 854
+    },
+    {
+      "epoch": 2.5675675675675675,
+      "grad_norm": 0.2451171875,
+      "learning_rate": 1.4414414414414416e-06,
+      "loss": 0.9739,
+      "step": 855
+    },
+    {
+      "epoch": 2.5705705705705704,
+      "grad_norm": 0.24609375,
+      "learning_rate": 1.4314314314314315e-06,
+      "loss": 0.9982,
+      "step": 856
+    },
+    {
+      "epoch": 2.5735735735735736,
+      "grad_norm": 0.2333984375,
+      "learning_rate": 1.4214214214214215e-06,
+      "loss": 0.9415,
+      "step": 857
+    },
+    {
+      "epoch": 2.5765765765765765,
+      "grad_norm": 0.23046875,
+      "learning_rate": 1.4114114114114117e-06,
+      "loss": 0.9796,
+      "step": 858
+    },
+    {
+      "epoch": 2.5795795795795797,
+      "grad_norm": 0.240234375,
+      "learning_rate": 1.4014014014014016e-06,
+      "loss": 0.9521,
+      "step": 859
+    },
+    {
+      "epoch": 2.5825825825825826,
+      "grad_norm": 0.2373046875,
+      "learning_rate": 1.3913913913913914e-06,
+      "loss": 0.9432,
+      "step": 860
+    },
+    {
+      "epoch": 2.5855855855855854,
+      "grad_norm": 0.328125,
+      "learning_rate": 1.3813813813813814e-06,
+      "loss": 0.9018,
+      "step": 861
+    },
+    {
+      "epoch": 2.5885885885885886,
+      "grad_norm": 0.236328125,
+      "learning_rate": 1.3713713713713714e-06,
+      "loss": 0.959,
+      "step": 862
+    },
+    {
+      "epoch": 2.5915915915915915,
+      "grad_norm": 0.255859375,
+      "learning_rate": 1.3613613613613615e-06,
+      "loss": 0.9416,
+      "step": 863
+    },
+    {
+      "epoch": 2.5945945945945947,
+      "grad_norm": 0.2578125,
+      "learning_rate": 1.3513513513513515e-06,
+      "loss": 0.9138,
+      "step": 864
+    },
+    {
+      "epoch": 2.5975975975975976,
+      "grad_norm": 0.2197265625,
+      "learning_rate": 1.3413413413413415e-06,
+      "loss": 0.9317,
+      "step": 865
+    },
+    {
+      "epoch": 2.6006006006006004,
+      "grad_norm": 0.2265625,
+      "learning_rate": 1.3313313313313315e-06,
+      "loss": 0.9503,
+      "step": 866
+    },
+    {
+      "epoch": 2.6036036036036037,
+      "grad_norm": 0.279296875,
+      "learning_rate": 1.3213213213213214e-06,
+      "loss": 0.9666,
+      "step": 867
+    },
+    {
+      "epoch": 2.6066066066066065,
+      "grad_norm": 0.353515625,
+      "learning_rate": 1.3113113113113112e-06,
+      "loss": 0.9285,
+      "step": 868
+    },
+    {
+      "epoch": 2.6096096096096097,
+      "grad_norm": 0.2333984375,
+      "learning_rate": 1.3013013013013016e-06,
+      "loss": 0.911,
+      "step": 869
+    },
+    {
+      "epoch": 2.6126126126126126,
+      "grad_norm": 0.232421875,
+      "learning_rate": 1.2912912912912913e-06,
+      "loss": 0.9176,
+      "step": 870
+    },
+    {
+      "epoch": 2.6156156156156154,
+      "grad_norm": 0.2333984375,
+      "learning_rate": 1.2812812812812813e-06,
+      "loss": 0.9858,
+      "step": 871
+    },
+    {
+      "epoch": 2.6186186186186187,
+      "grad_norm": 0.263671875,
+      "learning_rate": 1.2712712712712713e-06,
+      "loss": 0.9085,
+      "step": 872
+    },
+    {
+      "epoch": 2.6216216216216215,
+      "grad_norm": 0.2392578125,
+      "learning_rate": 1.2612612612612613e-06,
+      "loss": 0.9729,
+      "step": 873
+    },
+    {
+      "epoch": 2.6246246246246248,
+      "grad_norm": 0.251953125,
+      "learning_rate": 1.2512512512512514e-06,
+      "loss": 0.9489,
+      "step": 874
+    },
+    {
+      "epoch": 2.6276276276276276,
+      "grad_norm": 0.271484375,
+      "learning_rate": 1.2412412412412414e-06,
+      "loss": 0.945,
+      "step": 875
+    },
+    {
+      "epoch": 2.6306306306306304,
+      "grad_norm": 0.24609375,
+      "learning_rate": 1.2312312312312314e-06,
+      "loss": 0.9313,
+      "step": 876
+    },
+    {
+      "epoch": 2.6336336336336337,
+      "grad_norm": 0.2421875,
+      "learning_rate": 1.2212212212212213e-06,
+      "loss": 0.9176,
+      "step": 877
+    },
+    {
+      "epoch": 2.6366366366366365,
+      "grad_norm": 0.236328125,
+      "learning_rate": 1.2112112112112113e-06,
+      "loss": 0.9444,
+      "step": 878
+    },
+    {
+      "epoch": 2.6396396396396398,
+      "grad_norm": 0.224609375,
+      "learning_rate": 1.2012012012012013e-06,
+      "loss": 0.8967,
+      "step": 879
+    },
+    {
+      "epoch": 2.6426426426426426,
+      "grad_norm": 0.26171875,
+      "learning_rate": 1.1911911911911913e-06,
+      "loss": 0.9457,
+      "step": 880
+    },
+    {
+      "epoch": 2.6456456456456454,
+      "grad_norm": 0.2275390625,
+      "learning_rate": 1.1811811811811812e-06,
+      "loss": 0.9372,
+      "step": 881
+    },
+    {
+      "epoch": 2.6486486486486487,
+      "grad_norm": 0.2451171875,
+      "learning_rate": 1.1711711711711712e-06,
+      "loss": 0.9334,
+      "step": 882
+    },
+    {
+      "epoch": 2.6516516516516515,
+      "grad_norm": 0.2265625,
+      "learning_rate": 1.1611611611611614e-06,
+      "loss": 0.9281,
+      "step": 883
+    },
+    {
+      "epoch": 2.6546546546546548,
+      "grad_norm": 0.28125,
+      "learning_rate": 1.1511511511511512e-06,
+      "loss": 0.9191,
+      "step": 884
+    },
+    {
+      "epoch": 2.6576576576576576,
+      "grad_norm": 0.30859375,
+      "learning_rate": 1.1411411411411411e-06,
+      "loss": 0.9695,
+      "step": 885
+    },
+    {
+      "epoch": 2.6606606606606604,
+      "grad_norm": 0.25390625,
+      "learning_rate": 1.1311311311311313e-06,
+      "loss": 1.0134,
+      "step": 886
+    },
+    {
+      "epoch": 2.6636636636636637,
+      "grad_norm": 0.23828125,
+      "learning_rate": 1.1211211211211213e-06,
+      "loss": 0.9441,
+      "step": 887
+    },
+    {
+      "epoch": 2.6666666666666665,
+      "grad_norm": 0.23046875,
+      "learning_rate": 1.111111111111111e-06,
+      "loss": 0.9236,
+      "step": 888
+    },
+    {
+      "epoch": 2.66966966966967,
+      "grad_norm": 0.263671875,
+      "learning_rate": 1.1011011011011012e-06,
+      "loss": 0.9699,
+      "step": 889
+    },
+    {
+      "epoch": 2.6726726726726726,
+      "grad_norm": 0.234375,
+      "learning_rate": 1.0910910910910912e-06,
+      "loss": 0.9747,
+      "step": 890
+    },
+    {
+      "epoch": 2.6756756756756754,
+      "grad_norm": 0.244140625,
+      "learning_rate": 1.0810810810810812e-06,
+      "loss": 0.9145,
+      "step": 891
+    },
+    {
+      "epoch": 2.6786786786786787,
+      "grad_norm": 0.23046875,
+      "learning_rate": 1.0710710710710711e-06,
+      "loss": 0.9617,
+      "step": 892
+    },
+    {
+      "epoch": 2.6816816816816815,
+      "grad_norm": 0.2333984375,
+      "learning_rate": 1.0610610610610611e-06,
+      "loss": 0.9833,
+      "step": 893
+    },
+    {
+      "epoch": 2.684684684684685,
+      "grad_norm": 0.259765625,
+      "learning_rate": 1.051051051051051e-06,
+      "loss": 0.9488,
+      "step": 894
+    },
+    {
+      "epoch": 2.6876876876876876,
+      "grad_norm": 0.37109375,
+      "learning_rate": 1.041041041041041e-06,
+      "loss": 0.903,
+      "step": 895
+    },
+    {
+      "epoch": 2.6906906906906904,
+      "grad_norm": 0.2373046875,
+      "learning_rate": 1.031031031031031e-06,
+      "loss": 0.9255,
+      "step": 896
+    },
+    {
+      "epoch": 2.6936936936936937,
+      "grad_norm": 0.35546875,
+      "learning_rate": 1.0210210210210212e-06,
+      "loss": 0.9133,
+      "step": 897
+    },
+    {
+      "epoch": 2.6966966966966965,
+      "grad_norm": 0.283203125,
+      "learning_rate": 1.011011011011011e-06,
+      "loss": 0.9845,
+      "step": 898
+    },
+    {
+      "epoch": 2.6996996996997,
+      "grad_norm": 0.2734375,
+      "learning_rate": 1.0010010010010011e-06,
+      "loss": 0.9086,
+      "step": 899
+    },
+    {
+      "epoch": 2.7027027027027026,
+      "grad_norm": 0.228515625,
+      "learning_rate": 9.909909909909911e-07,
+      "loss": 0.9376,
+      "step": 900
+    },
+    {
+      "epoch": 2.7057057057057055,
+      "grad_norm": 0.283203125,
+      "learning_rate": 9.80980980980981e-07,
+      "loss": 0.9886,
+      "step": 901
+    },
+    {
+      "epoch": 2.7087087087087087,
+      "grad_norm": 0.2265625,
+      "learning_rate": 9.70970970970971e-07,
+      "loss": 0.916,
+      "step": 902
+    },
+    {
+      "epoch": 2.7117117117117115,
+      "grad_norm": 0.275390625,
+      "learning_rate": 9.60960960960961e-07,
+      "loss": 1.0119,
+      "step": 903
+    },
+    {
+      "epoch": 2.714714714714715,
+      "grad_norm": 0.2890625,
+      "learning_rate": 9.50950950950951e-07,
+      "loss": 0.9048,
+      "step": 904
+    },
+    {
+      "epoch": 2.7177177177177176,
+      "grad_norm": 0.2265625,
+      "learning_rate": 9.409409409409411e-07,
+      "loss": 0.9336,
+      "step": 905
+    },
+    {
+      "epoch": 2.7207207207207205,
+      "grad_norm": 0.236328125,
+      "learning_rate": 9.30930930930931e-07,
+      "loss": 0.9378,
+      "step": 906
+    },
+    {
+      "epoch": 2.7237237237237237,
+      "grad_norm": 0.236328125,
+      "learning_rate": 9.20920920920921e-07,
+      "loss": 0.9558,
+      "step": 907
+    },
+    {
+      "epoch": 2.726726726726727,
+      "grad_norm": 0.2236328125,
+      "learning_rate": 9.10910910910911e-07,
+      "loss": 0.9445,
+      "step": 908
+    },
+    {
+      "epoch": 2.72972972972973,
+      "grad_norm": 0.2255859375,
+      "learning_rate": 9.00900900900901e-07,
+      "loss": 0.953,
+      "step": 909
+    },
+    {
+      "epoch": 2.7327327327327327,
+      "grad_norm": 0.23046875,
+      "learning_rate": 8.90890890890891e-07,
+      "loss": 0.9786,
+      "step": 910
+    },
+    {
+      "epoch": 2.7357357357357355,
+      "grad_norm": 0.2265625,
+      "learning_rate": 8.808808808808809e-07,
+      "loss": 0.9439,
+      "step": 911
+    },
+    {
+      "epoch": 2.7387387387387387,
+      "grad_norm": 0.234375,
+      "learning_rate": 8.708708708708709e-07,
+      "loss": 0.926,
+      "step": 912
+    },
+    {
+      "epoch": 2.741741741741742,
+      "grad_norm": 0.2421875,
+      "learning_rate": 8.60860860860861e-07,
+      "loss": 0.9654,
+      "step": 913
+    },
+    {
+      "epoch": 2.744744744744745,
+      "grad_norm": 0.24609375,
+      "learning_rate": 8.508508508508509e-07,
+      "loss": 0.9336,
+      "step": 914
+    },
+    {
+      "epoch": 2.7477477477477477,
+      "grad_norm": 0.2392578125,
+      "learning_rate": 8.40840840840841e-07,
+      "loss": 0.9318,
+      "step": 915
+    },
+    {
+      "epoch": 2.7507507507507505,
+      "grad_norm": 0.240234375,
+      "learning_rate": 8.308308308308309e-07,
+      "loss": 0.9396,
+      "step": 916
+    },
+    {
+      "epoch": 2.7537537537537538,
+      "grad_norm": 0.28125,
+      "learning_rate": 8.208208208208208e-07,
+      "loss": 1.0022,
+      "step": 917
+    },
+    {
+      "epoch": 2.756756756756757,
+      "grad_norm": 0.2294921875,
+      "learning_rate": 8.108108108108109e-07,
+      "loss": 0.9604,
+      "step": 918
+    },
+    {
+      "epoch": 2.75975975975976,
+      "grad_norm": 0.2470703125,
+      "learning_rate": 8.008008008008009e-07,
+      "loss": 0.9967,
+      "step": 919
+    },
+    {
+      "epoch": 2.7627627627627627,
+      "grad_norm": 0.287109375,
+      "learning_rate": 7.907907907907908e-07,
+      "loss": 1.003,
+      "step": 920
+    },
+    {
+      "epoch": 2.7657657657657655,
+      "grad_norm": 0.2353515625,
+      "learning_rate": 7.807807807807808e-07,
+      "loss": 0.9132,
+      "step": 921
+    },
+    {
+      "epoch": 2.7687687687687688,
+      "grad_norm": 0.2177734375,
+      "learning_rate": 7.707707707707708e-07,
+      "loss": 0.9363,
+      "step": 922
+    },
+    {
+      "epoch": 2.771771771771772,
+      "grad_norm": 0.2294921875,
+      "learning_rate": 7.607607607607609e-07,
+      "loss": 0.9135,
+      "step": 923
+    },
+    {
+      "epoch": 2.774774774774775,
+      "grad_norm": 0.23828125,
+      "learning_rate": 7.507507507507509e-07,
+      "loss": 0.9572,
+      "step": 924
+    },
+    {
+      "epoch": 2.7777777777777777,
+      "grad_norm": 0.2275390625,
+      "learning_rate": 7.407407407407407e-07,
+      "loss": 0.9178,
+      "step": 925
+    },
+    {
+      "epoch": 2.7807807807807805,
+      "grad_norm": 0.2490234375,
+      "learning_rate": 7.307307307307308e-07,
+      "loss": 0.9313,
+      "step": 926
+    },
+    {
+      "epoch": 2.7837837837837838,
+      "grad_norm": 0.263671875,
+      "learning_rate": 7.207207207207208e-07,
+      "loss": 0.8922,
+      "step": 927
+    },
+    {
+      "epoch": 2.786786786786787,
+      "grad_norm": 0.275390625,
+      "learning_rate": 7.107107107107107e-07,
+      "loss": 0.9348,
+      "step": 928
+    },
+    {
+      "epoch": 2.78978978978979,
+      "grad_norm": 0.2236328125,
+      "learning_rate": 7.007007007007008e-07,
+      "loss": 0.9164,
+      "step": 929
+    },
+    {
+      "epoch": 2.7927927927927927,
+      "grad_norm": 0.2294921875,
+      "learning_rate": 6.906906906906907e-07,
+      "loss": 0.9017,
+      "step": 930
+    },
+    {
+      "epoch": 2.795795795795796,
+      "grad_norm": 0.2578125,
+      "learning_rate": 6.806806806806808e-07,
+      "loss": 0.9833,
+      "step": 931
+    },
+    {
+      "epoch": 2.798798798798799,
+      "grad_norm": 0.2412109375,
+      "learning_rate": 6.706706706706707e-07,
+      "loss": 1.0166,
+      "step": 932
+    },
+    {
+      "epoch": 2.801801801801802,
+      "grad_norm": 0.2451171875,
+      "learning_rate": 6.606606606606607e-07,
+      "loss": 0.9839,
+      "step": 933
+    },
+    {
+      "epoch": 2.804804804804805,
+      "grad_norm": 0.22265625,
+      "learning_rate": 6.506506506506508e-07,
+      "loss": 0.9298,
+      "step": 934
+    },
+    {
+      "epoch": 2.8078078078078077,
+      "grad_norm": 0.2333984375,
+      "learning_rate": 6.406406406406407e-07,
+      "loss": 0.9642,
+      "step": 935
+    },
+    {
+      "epoch": 2.810810810810811,
+      "grad_norm": 0.25,
+      "learning_rate": 6.306306306306306e-07,
+      "loss": 0.9332,
+      "step": 936
+    },
+    {
+      "epoch": 2.813813813813814,
+      "grad_norm": 0.22265625,
+      "learning_rate": 6.206206206206207e-07,
+      "loss": 0.9122,
+      "step": 937
+    },
+    {
+      "epoch": 2.816816816816817,
+      "grad_norm": 0.2314453125,
+      "learning_rate": 6.106106106106107e-07,
+      "loss": 0.9278,
+      "step": 938
+    },
+    {
+      "epoch": 2.81981981981982,
+      "grad_norm": 1.390625,
+      "learning_rate": 6.006006006006006e-07,
+      "loss": 0.9441,
+      "step": 939
+    },
+    {
+      "epoch": 2.8228228228228227,
+      "grad_norm": 0.2392578125,
+      "learning_rate": 5.905905905905906e-07,
+      "loss": 0.9257,
+      "step": 940
+    },
+    {
+      "epoch": 2.825825825825826,
+      "grad_norm": 0.2421875,
+      "learning_rate": 5.805805805805807e-07,
+      "loss": 0.9897,
+      "step": 941
+    },
+    {
+      "epoch": 2.828828828828829,
+      "grad_norm": 0.2236328125,
+      "learning_rate": 5.705705705705706e-07,
+      "loss": 0.9019,
+      "step": 942
+    },
+    {
+      "epoch": 2.831831831831832,
+      "grad_norm": 0.259765625,
+      "learning_rate": 5.605605605605606e-07,
+      "loss": 0.9317,
+      "step": 943
+    },
+    {
+      "epoch": 2.834834834834835,
+      "grad_norm": 0.36328125,
+      "learning_rate": 5.505505505505506e-07,
+      "loss": 0.9096,
+      "step": 944
+    },
+    {
+      "epoch": 2.8378378378378377,
+      "grad_norm": 0.302734375,
+      "learning_rate": 5.405405405405406e-07,
+      "loss": 0.9026,
+      "step": 945
+    },
+    {
+      "epoch": 2.840840840840841,
+      "grad_norm": 0.2265625,
+      "learning_rate": 5.305305305305306e-07,
+      "loss": 0.9618,
+      "step": 946
+    },
+    {
+      "epoch": 2.843843843843844,
+      "grad_norm": 0.2421875,
+      "learning_rate": 5.205205205205205e-07,
+      "loss": 0.9015,
+      "step": 947
+    },
+    {
+      "epoch": 2.846846846846847,
+      "grad_norm": 0.2216796875,
+      "learning_rate": 5.105105105105106e-07,
+      "loss": 0.9351,
+      "step": 948
+    },
+    {
+      "epoch": 2.84984984984985,
+      "grad_norm": 0.322265625,
+      "learning_rate": 5.005005005005006e-07,
+      "loss": 0.9347,
+      "step": 949
+    },
+    {
+      "epoch": 2.8528528528528527,
+      "grad_norm": 0.240234375,
+      "learning_rate": 4.904904904904905e-07,
+      "loss": 0.9432,
+      "step": 950
+    },
+    {
+      "epoch": 2.855855855855856,
+      "grad_norm": 0.26171875,
+      "learning_rate": 4.804804804804805e-07,
+      "loss": 0.9455,
+      "step": 951
+    },
+    {
+      "epoch": 2.858858858858859,
+      "grad_norm": 0.271484375,
+      "learning_rate": 4.7047047047047054e-07,
+      "loss": 0.9354,
+      "step": 952
+    },
+    {
+      "epoch": 2.861861861861862,
+      "grad_norm": 0.232421875,
+      "learning_rate": 4.604604604604605e-07,
+      "loss": 0.9672,
+      "step": 953
+    },
+    {
+      "epoch": 2.864864864864865,
+      "grad_norm": 0.2333984375,
+      "learning_rate": 4.504504504504505e-07,
+      "loss": 0.9141,
+      "step": 954
+    },
+    {
+      "epoch": 2.8678678678678677,
+      "grad_norm": 0.259765625,
+      "learning_rate": 4.4044044044044046e-07,
+      "loss": 0.9208,
+      "step": 955
+    },
+    {
+      "epoch": 2.870870870870871,
+      "grad_norm": 0.234375,
+      "learning_rate": 4.304304304304305e-07,
+      "loss": 0.9691,
+      "step": 956
+    },
+    {
+      "epoch": 2.873873873873874,
+      "grad_norm": 0.2578125,
+      "learning_rate": 4.204204204204205e-07,
+      "loss": 0.9378,
+      "step": 957
+    },
+    {
+      "epoch": 2.876876876876877,
+      "grad_norm": 0.2265625,
+      "learning_rate": 4.104104104104104e-07,
+      "loss": 0.935,
+      "step": 958
+    },
+    {
+      "epoch": 2.87987987987988,
+      "grad_norm": 0.2578125,
+      "learning_rate": 4.0040040040040045e-07,
+      "loss": 0.8964,
+      "step": 959
+    },
+    {
+      "epoch": 2.8828828828828827,
+      "grad_norm": 0.310546875,
+      "learning_rate": 3.903903903903904e-07,
+      "loss": 0.9354,
+      "step": 960
+    },
+    {
+      "epoch": 2.885885885885886,
+      "grad_norm": 0.26171875,
+      "learning_rate": 3.8038038038038044e-07,
+      "loss": 0.9209,
+      "step": 961
+    },
+    {
+      "epoch": 2.888888888888889,
+      "grad_norm": 0.22265625,
+      "learning_rate": 3.7037037037037036e-07,
+      "loss": 0.9619,
+      "step": 962
+    },
+    {
+      "epoch": 2.891891891891892,
+      "grad_norm": 0.2421875,
+      "learning_rate": 3.603603603603604e-07,
+      "loss": 0.9273,
+      "step": 963
+    },
+    {
+      "epoch": 2.894894894894895,
+      "grad_norm": 0.2451171875,
+      "learning_rate": 3.503503503503504e-07,
+      "loss": 0.9459,
+      "step": 964
+    },
+    {
+      "epoch": 2.8978978978978978,
+      "grad_norm": 0.25,
+      "learning_rate": 3.403403403403404e-07,
+      "loss": 0.9549,
+      "step": 965
+    },
+    {
+      "epoch": 2.900900900900901,
+      "grad_norm": 0.30078125,
+      "learning_rate": 3.3033033033033036e-07,
+      "loss": 0.9524,
+      "step": 966
+    },
+    {
+      "epoch": 2.903903903903904,
+      "grad_norm": 0.2236328125,
+      "learning_rate": 3.2032032032032033e-07,
+      "loss": 0.9443,
+      "step": 967
+    },
+    {
+      "epoch": 2.906906906906907,
+      "grad_norm": 0.23828125,
+      "learning_rate": 3.1031031031031035e-07,
+      "loss": 0.9576,
+      "step": 968
+    },
+    {
+      "epoch": 2.90990990990991,
+      "grad_norm": 0.2275390625,
+      "learning_rate": 3.003003003003003e-07,
+      "loss": 0.9042,
+      "step": 969
+    },
+    {
+      "epoch": 2.9129129129129128,
+      "grad_norm": 0.23046875,
+      "learning_rate": 2.9029029029029035e-07,
+      "loss": 0.9236,
+      "step": 970
+    },
+    {
+      "epoch": 2.915915915915916,
+      "grad_norm": 0.2353515625,
+      "learning_rate": 2.802802802802803e-07,
+      "loss": 0.9501,
+      "step": 971
+    },
+    {
+      "epoch": 2.918918918918919,
+      "grad_norm": 0.26953125,
+      "learning_rate": 2.702702702702703e-07,
+      "loss": 0.9917,
+      "step": 972
+    },
+    {
+      "epoch": 2.921921921921922,
+      "grad_norm": 0.240234375,
+      "learning_rate": 2.6026026026026026e-07,
+      "loss": 0.929,
+      "step": 973
+    },
+    {
+      "epoch": 2.924924924924925,
+      "grad_norm": 0.30078125,
+      "learning_rate": 2.502502502502503e-07,
+      "loss": 0.9541,
+      "step": 974
+    },
+    {
+      "epoch": 2.9279279279279278,
+      "grad_norm": 0.2294921875,
+      "learning_rate": 2.4024024024024026e-07,
+      "loss": 0.9836,
+      "step": 975
+    },
+    {
+      "epoch": 2.930930930930931,
+      "grad_norm": 0.236328125,
+      "learning_rate": 2.3023023023023026e-07,
+      "loss": 1.0039,
+      "step": 976
+    },
+    {
+      "epoch": 2.933933933933934,
+      "grad_norm": 0.2314453125,
+      "learning_rate": 2.2022022022022023e-07,
+      "loss": 0.9617,
+      "step": 977
+    },
+    {
+      "epoch": 2.936936936936937,
+      "grad_norm": 0.248046875,
+      "learning_rate": 2.1021021021021025e-07,
+      "loss": 0.8823,
+      "step": 978
+    },
+    {
+      "epoch": 2.93993993993994,
+      "grad_norm": 0.3828125,
+      "learning_rate": 2.0020020020020022e-07,
+      "loss": 0.9444,
+      "step": 979
+    },
+    {
+      "epoch": 2.942942942942943,
+      "grad_norm": 0.228515625,
+      "learning_rate": 1.9019019019019022e-07,
+      "loss": 0.9653,
+      "step": 980
+    },
+    {
+      "epoch": 2.945945945945946,
+      "grad_norm": 0.2373046875,
+      "learning_rate": 1.801801801801802e-07,
+      "loss": 0.9174,
+      "step": 981
+    },
+    {
+      "epoch": 2.948948948948949,
+      "grad_norm": 0.2275390625,
+      "learning_rate": 1.701701701701702e-07,
+      "loss": 0.9446,
+      "step": 982
+    },
+    {
+      "epoch": 2.951951951951952,
+      "grad_norm": 0.2353515625,
+      "learning_rate": 1.6016016016016016e-07,
+      "loss": 0.8897,
+      "step": 983
+    },
+    {
+      "epoch": 2.954954954954955,
+      "grad_norm": 0.23828125,
+      "learning_rate": 1.5015015015015016e-07,
+      "loss": 0.9451,
+      "step": 984
+    },
+    {
+      "epoch": 2.957957957957958,
+      "grad_norm": 0.228515625,
+      "learning_rate": 1.4014014014014016e-07,
+      "loss": 0.9335,
+      "step": 985
+    },
+    {
+      "epoch": 2.960960960960961,
+      "grad_norm": 0.306640625,
+      "learning_rate": 1.3013013013013013e-07,
+      "loss": 0.9516,
+      "step": 986
+    },
+    {
+      "epoch": 2.963963963963964,
+      "grad_norm": 0.2314453125,
+      "learning_rate": 1.2012012012012013e-07,
+      "loss": 0.9919,
+      "step": 987
+    },
+    {
+      "epoch": 2.966966966966967,
+      "grad_norm": 0.26171875,
+      "learning_rate": 1.1011011011011011e-07,
+      "loss": 0.9401,
+      "step": 988
+    },
+    {
+      "epoch": 2.96996996996997,
+      "grad_norm": 0.234375,
+      "learning_rate": 1.0010010010010011e-07,
+      "loss": 0.9606,
+      "step": 989
+    },
+    {
+      "epoch": 2.972972972972973,
+      "grad_norm": 0.236328125,
+      "learning_rate": 9.00900900900901e-08,
+      "loss": 0.9594,
+      "step": 990
+    },
+    {
+      "epoch": 2.975975975975976,
+      "grad_norm": 0.291015625,
+      "learning_rate": 8.008008008008008e-08,
+      "loss": 0.9616,
+      "step": 991
+    },
+    {
+      "epoch": 2.978978978978979,
+      "grad_norm": 0.2216796875,
+      "learning_rate": 7.007007007007008e-08,
+      "loss": 0.9056,
+      "step": 992
+    },
+    {
+      "epoch": 2.981981981981982,
+      "grad_norm": 0.2392578125,
+      "learning_rate": 6.006006006006006e-08,
+      "loss": 0.9647,
+      "step": 993
+    },
+    {
+      "epoch": 2.984984984984985,
+      "grad_norm": 0.22265625,
+      "learning_rate": 5.0050050050050056e-08,
+      "loss": 0.9642,
+      "step": 994
+    },
+    {
+      "epoch": 2.987987987987988,
+      "grad_norm": 0.2333984375,
+      "learning_rate": 4.004004004004004e-08,
+      "loss": 0.9377,
+      "step": 995
+    },
+    {
+      "epoch": 2.990990990990991,
+      "grad_norm": 0.26953125,
+      "learning_rate": 3.003003003003003e-08,
+      "loss": 0.9236,
+      "step": 996
+    },
+    {
+      "epoch": 2.993993993993994,
+      "grad_norm": 0.32421875,
+      "learning_rate": 2.002002002002002e-08,
+      "loss": 0.9051,
+      "step": 997
+    },
+    {
+      "epoch": 2.996996996996997,
+      "grad_norm": 0.302734375,
+      "learning_rate": 1.001001001001001e-08,
+      "loss": 0.9192,
+      "step": 998
+    },
+    {
+      "epoch": 3.0,
+      "grad_norm": 0.232421875,
+      "learning_rate": 0.0,
+      "loss": 0.9868,
+      "step": 999
+    }
+  ],
+  "logging_steps": 1.0,
+  "max_steps": 999,
+  "num_input_tokens_seen": 0,
+  "num_train_epochs": 3,
+  "save_steps": 0,
+  "stateful_callbacks": {
+    "TrainerControl": {
+      "args": {
+        "should_epoch_stop": false,
+        "should_evaluate": false,
+        "should_log": false,
+        "should_save": true,
+        "should_training_stop": true
+      },
+      "attributes": {}
+    }
+  },
+  "total_flos": 3.157056495934243e+18,
+  "train_batch_size": 1,
+  "trial_name": null,
+  "trial_params": null
+}