diff --git "a/trainer_state.json" "b/trainer_state.json"
new file mode 100644--- /dev/null
+++ "b/trainer_state.json"
@@ -0,0 +1,6016 @@
+{
+  "best_metric": null,
+  "best_model_checkpoint": null,
+  "epoch": 2.0465592223074953,
+  "global_step": 1000,
+  "is_hyper_param_search": false,
+  "is_local_process_zero": true,
+  "is_world_process_zero": true,
+  "log_history": [
+    {
+      "epoch": 0.0,
+      "learning_rate": 4.5454545454545457e-07,
+      "loss": 1.7447,
+      "step": 1
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 9.090909090909091e-07,
+      "loss": 1.7545,
+      "step": 2
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 1.3636363636363636e-06,
+      "loss": 1.729,
+      "step": 3
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 1.8181818181818183e-06,
+      "loss": 1.6094,
+      "step": 4
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 2.2727272727272728e-06,
+      "loss": 1.6613,
+      "step": 5
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 2.7272727272727272e-06,
+      "loss": 1.5168,
+      "step": 6
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 3.181818181818182e-06,
+      "loss": 1.4354,
+      "step": 7
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 3.6363636363636366e-06,
+      "loss": 1.218,
+      "step": 8
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.0909090909090915e-06,
+      "loss": 1.2204,
+      "step": 9
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.5454545454545455e-06,
+      "loss": 1.0702,
+      "step": 10
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 5e-06,
+      "loss": 0.9516,
+      "step": 11
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 5.4545454545454545e-06,
+      "loss": 1.0377,
+      "step": 12
+    },
+    {
+      "epoch": 0.03,
+      "learning_rate": 5.90909090909091e-06,
+      "loss": 0.9798,
+      "step": 13
+    },
+    {
+      "epoch": 0.03,
+      "learning_rate": 6.363636363636364e-06,
+      "loss": 0.9245,
+      "step": 14
+    },
+    {
+      "epoch": 0.03,
+      "learning_rate": 6.818181818181818e-06,
+      "loss": 0.9279,
+      "step": 15
+    },
+    {
+      "epoch": 0.03,
+      "learning_rate": 7.272727272727273e-06,
+      "loss": 0.9215,
+      "step": 16
+    },
+    {
+      "epoch": 0.03,
+      "learning_rate": 7.727272727272727e-06,
+      "loss": 0.9659,
+      "step": 17
+    },
+    {
+      "epoch": 0.04,
+      "learning_rate": 8.181818181818183e-06,
+      "loss": 0.9475,
+      "step": 18
+    },
+    {
+      "epoch": 0.04,
+      "learning_rate": 8.636363636363637e-06,
+      "loss": 0.9297,
+      "step": 19
+    },
+    {
+      "epoch": 0.04,
+      "learning_rate": 9.090909090909091e-06,
+      "loss": 0.93,
+      "step": 20
+    },
+    {
+      "epoch": 0.04,
+      "learning_rate": 9.545454545454547e-06,
+      "loss": 0.8521,
+      "step": 21
+    },
+    {
+      "epoch": 0.05,
+      "learning_rate": 1e-05,
+      "loss": 0.8539,
+      "step": 22
+    },
+    {
+      "epoch": 0.05,
+      "learning_rate": 1.0454545454545455e-05,
+      "loss": 0.8036,
+      "step": 23
+    },
+    {
+      "epoch": 0.05,
+      "learning_rate": 1.0909090909090909e-05,
+      "loss": 0.8456,
+      "step": 24
+    },
+    {
+      "epoch": 0.05,
+      "learning_rate": 1.1363636363636366e-05,
+      "loss": 0.819,
+      "step": 25
+    },
+    {
+      "epoch": 0.05,
+      "learning_rate": 1.181818181818182e-05,
+      "loss": 0.8771,
+      "step": 26
+    },
+    {
+      "epoch": 0.06,
+      "learning_rate": 1.2272727272727274e-05,
+      "loss": 0.8038,
+      "step": 27
+    },
+    {
+      "epoch": 0.06,
+      "learning_rate": 1.2727272727272728e-05,
+      "loss": 0.8301,
+      "step": 28
+    },
+    {
+      "epoch": 0.06,
+      "learning_rate": 1.3181818181818183e-05,
+      "loss": 0.8483,
+      "step": 29
+    },
+    {
+      "epoch": 0.06,
+      "learning_rate": 1.3636363636363637e-05,
+      "loss": 0.8009,
+      "step": 30
+    },
+    {
+      "epoch": 0.06,
+      "learning_rate": 1.4090909090909092e-05,
+      "loss": 0.7494,
+      "step": 31
+    },
+    {
+      "epoch": 0.07,
+      "learning_rate": 1.4545454545454546e-05,
+      "loss": 0.8033,
+      "step": 32
+    },
+    {
+      "epoch": 0.07,
+      "learning_rate": 1.5000000000000002e-05,
+      "loss": 0.7652,
+      "step": 33
+    },
+    {
+      "epoch": 0.07,
+      "learning_rate": 1.5454545454545454e-05,
+      "loss": 0.8586,
+      "step": 34
+    },
+    {
+      "epoch": 0.07,
+      "learning_rate": 1.590909090909091e-05,
+      "loss": 0.7789,
+      "step": 35
+    },
+    {
+      "epoch": 0.07,
+      "learning_rate": 1.6363636363636366e-05,
+      "loss": 0.8003,
+      "step": 36
+    },
+    {
+      "epoch": 0.08,
+      "learning_rate": 1.681818181818182e-05,
+      "loss": 0.7443,
+      "step": 37
+    },
+    {
+      "epoch": 0.08,
+      "learning_rate": 1.7272727272727274e-05,
+      "loss": 0.8285,
+      "step": 38
+    },
+    {
+      "epoch": 0.08,
+      "learning_rate": 1.772727272727273e-05,
+      "loss": 0.7628,
+      "step": 39
+    },
+    {
+      "epoch": 0.08,
+      "learning_rate": 1.8181818181818182e-05,
+      "loss": 0.7868,
+      "step": 40
+    },
+    {
+      "epoch": 0.08,
+      "learning_rate": 1.8636363636363638e-05,
+      "loss": 0.7481,
+      "step": 41
+    },
+    {
+      "epoch": 0.09,
+      "learning_rate": 1.9090909090909094e-05,
+      "loss": 0.7855,
+      "step": 42
+    },
+    {
+      "epoch": 0.09,
+      "learning_rate": 1.9545454545454546e-05,
+      "loss": 0.7506,
+      "step": 43
+    },
+    {
+      "epoch": 0.09,
+      "learning_rate": 2e-05,
+      "loss": 0.7504,
+      "step": 44
+    },
+    {
+      "epoch": 0.09,
+      "learning_rate": 1.9999975526680286e-05,
+      "loss": 0.7607,
+      "step": 45
+    },
+    {
+      "epoch": 0.09,
+      "learning_rate": 1.999990210684092e-05,
+      "loss": 0.8022,
+      "step": 46
+    },
+    {
+      "epoch": 0.1,
+      "learning_rate": 1.999977974084128e-05,
+      "loss": 0.8513,
+      "step": 47
+    },
+    {
+      "epoch": 0.1,
+      "learning_rate": 1.9999608429280302e-05,
+      "loss": 0.7948,
+      "step": 48
+    },
+    {
+      "epoch": 0.1,
+      "learning_rate": 1.9999388172996495e-05,
+      "loss": 0.7338,
+      "step": 49
+    },
+    {
+      "epoch": 0.1,
+      "learning_rate": 1.9999118973067942e-05,
+      "loss": 0.7139,
+      "step": 50
+    },
+    {
+      "epoch": 0.1,
+      "learning_rate": 1.999880083081229e-05,
+      "loss": 0.7114,
+      "step": 51
+    },
+    {
+      "epoch": 0.11,
+      "learning_rate": 1.999843374778673e-05,
+      "loss": 0.7345,
+      "step": 52
+    },
+    {
+      "epoch": 0.11,
+      "learning_rate": 1.999801772578801e-05,
+      "loss": 0.6277,
+      "step": 53
+    },
+    {
+      "epoch": 0.11,
+      "learning_rate": 1.9997552766852434e-05,
+      "loss": 0.7228,
+      "step": 54
+    },
+    {
+      "epoch": 0.11,
+      "learning_rate": 1.9997038873255802e-05,
+      "loss": 0.7264,
+      "step": 55
+    },
+    {
+      "epoch": 0.11,
+      "learning_rate": 1.9996476047513454e-05,
+      "loss": 0.7726,
+      "step": 56
+    },
+    {
+      "epoch": 0.12,
+      "learning_rate": 1.999586429238024e-05,
+      "loss": 0.7114,
+      "step": 57
+    },
+    {
+      "epoch": 0.12,
+      "learning_rate": 1.999520361085049e-05,
+      "loss": 0.7448,
+      "step": 58
+    },
+    {
+      "epoch": 0.12,
+      "learning_rate": 1.9994494006158018e-05,
+      "loss": 0.7882,
+      "step": 59
+    },
+    {
+      "epoch": 0.12,
+      "learning_rate": 1.9993735481776105e-05,
+      "loss": 0.7219,
+      "step": 60
+    },
+    {
+      "epoch": 0.12,
+      "learning_rate": 1.9992928041417468e-05,
+      "loss": 0.7117,
+      "step": 61
+    },
+    {
+      "epoch": 0.13,
+      "learning_rate": 1.9992071689034257e-05,
+      "loss": 0.6483,
+      "step": 62
+    },
+    {
+      "epoch": 0.13,
+      "learning_rate": 1.9991166428818036e-05,
+      "loss": 0.7375,
+      "step": 63
+    },
+    {
+      "epoch": 0.13,
+      "learning_rate": 1.9990212265199738e-05,
+      "loss": 0.6829,
+      "step": 64
+    },
+    {
+      "epoch": 0.13,
+      "learning_rate": 1.998920920284968e-05,
+      "loss": 0.7334,
+      "step": 65
+    },
+    {
+      "epoch": 0.14,
+      "learning_rate": 1.9988157246677514e-05,
+      "loss": 0.7891,
+      "step": 66
+    },
+    {
+      "epoch": 0.14,
+      "learning_rate": 1.9987056401832213e-05,
+      "loss": 0.7708,
+      "step": 67
+    },
+    {
+      "epoch": 0.14,
+      "learning_rate": 1.998590667370204e-05,
+      "loss": 0.787,
+      "step": 68
+    },
+    {
+      "epoch": 0.14,
+      "learning_rate": 1.9984708067914533e-05,
+      "loss": 0.7632,
+      "step": 69
+    },
+    {
+      "epoch": 0.14,
+      "learning_rate": 1.9983460590336458e-05,
+      "loss": 0.6609,
+      "step": 70
+    },
+    {
+      "epoch": 0.15,
+      "learning_rate": 1.9982164247073803e-05,
+      "loss": 0.7458,
+      "step": 71
+    },
+    {
+      "epoch": 0.15,
+      "learning_rate": 1.9980819044471732e-05,
+      "loss": 0.6935,
+      "step": 72
+    },
+    {
+      "epoch": 0.15,
+      "learning_rate": 1.9979424989114555e-05,
+      "loss": 0.6969,
+      "step": 73
+    },
+    {
+      "epoch": 0.15,
+      "learning_rate": 1.9977982087825714e-05,
+      "loss": 0.697,
+      "step": 74
+    },
+    {
+      "epoch": 0.15,
+      "learning_rate": 1.9976490347667713e-05,
+      "loss": 0.6652,
+      "step": 75
+    },
+    {
+      "epoch": 0.16,
+      "learning_rate": 1.9974949775942134e-05,
+      "loss": 0.7321,
+      "step": 76
+    },
+    {
+      "epoch": 0.16,
+      "learning_rate": 1.9973360380189547e-05,
+      "loss": 0.7344,
+      "step": 77
+    },
+    {
+      "epoch": 0.16,
+      "learning_rate": 1.997172216818951e-05,
+      "loss": 0.6714,
+      "step": 78
+    },
+    {
+      "epoch": 0.16,
+      "learning_rate": 1.9970035147960524e-05,
+      "loss": 0.7586,
+      "step": 79
+    },
+    {
+      "epoch": 0.16,
+      "learning_rate": 1.996829932775999e-05,
+      "loss": 0.6622,
+      "step": 80
+    },
+    {
+      "epoch": 0.17,
+      "learning_rate": 1.996651471608415e-05,
+      "loss": 0.7297,
+      "step": 81
+    },
+    {
+      "epoch": 0.17,
+      "learning_rate": 1.9964681321668097e-05,
+      "loss": 0.6464,
+      "step": 82
+    },
+    {
+      "epoch": 0.17,
+      "learning_rate": 1.996279915348567e-05,
+      "loss": 0.7213,
+      "step": 83
+    },
+    {
+      "epoch": 0.17,
+      "learning_rate": 1.996086822074945e-05,
+      "loss": 0.6702,
+      "step": 84
+    },
+    {
+      "epoch": 0.17,
+      "learning_rate": 1.9958888532910708e-05,
+      "loss": 0.7002,
+      "step": 85
+    },
+    {
+      "epoch": 0.18,
+      "learning_rate": 1.9956860099659345e-05,
+      "loss": 0.6698,
+      "step": 86
+    },
+    {
+      "epoch": 0.18,
+      "learning_rate": 1.9954782930923864e-05,
+      "loss": 0.7386,
+      "step": 87
+    },
+    {
+      "epoch": 0.18,
+      "learning_rate": 1.995265703687131e-05,
+      "loss": 0.7486,
+      "step": 88
+    },
+    {
+      "epoch": 0.18,
+      "learning_rate": 1.995048242790721e-05,
+      "loss": 0.6778,
+      "step": 89
+    },
+    {
+      "epoch": 0.18,
+      "learning_rate": 1.9948259114675553e-05,
+      "loss": 0.7191,
+      "step": 90
+    },
+    {
+      "epoch": 0.19,
+      "learning_rate": 1.9945987108058713e-05,
+      "loss": 0.6939,
+      "step": 91
+    },
+    {
+      "epoch": 0.19,
+      "learning_rate": 1.9943666419177392e-05,
+      "loss": 0.6387,
+      "step": 92
+    },
+    {
+      "epoch": 0.19,
+      "learning_rate": 1.9941297059390586e-05,
+      "loss": 0.6718,
+      "step": 93
+    },
+    {
+      "epoch": 0.19,
+      "learning_rate": 1.9938879040295508e-05,
+      "loss": 0.683,
+      "step": 94
+    },
+    {
+      "epoch": 0.19,
+      "learning_rate": 1.993641237372756e-05,
+      "loss": 0.7169,
+      "step": 95
+    },
+    {
+      "epoch": 0.2,
+      "learning_rate": 1.9933897071760236e-05,
+      "loss": 0.678,
+      "step": 96
+    },
+    {
+      "epoch": 0.2,
+      "learning_rate": 1.9931333146705097e-05,
+      "loss": 0.7516,
+      "step": 97
+    },
+    {
+      "epoch": 0.2,
+      "learning_rate": 1.99287206111117e-05,
+      "loss": 0.7245,
+      "step": 98
+    },
+    {
+      "epoch": 0.2,
+      "learning_rate": 1.992605947776752e-05,
+      "loss": 0.6419,
+      "step": 99
+    },
+    {
+      "epoch": 0.2,
+      "learning_rate": 1.9923349759697915e-05,
+      "loss": 0.6566,
+      "step": 100
+    },
+    {
+      "epoch": 0.21,
+      "learning_rate": 1.9920591470166047e-05,
+      "loss": 0.7252,
+      "step": 101
+    },
+    {
+      "epoch": 0.21,
+      "learning_rate": 1.991778462267281e-05,
+      "loss": 0.6467,
+      "step": 102
+    },
+    {
+      "epoch": 0.21,
+      "learning_rate": 1.9914929230956785e-05,
+      "loss": 0.6693,
+      "step": 103
+    },
+    {
+      "epoch": 0.21,
+      "learning_rate": 1.9912025308994146e-05,
+      "loss": 0.671,
+      "step": 104
+    },
+    {
+      "epoch": 0.21,
+      "learning_rate": 1.9909072870998625e-05,
+      "loss": 0.6422,
+      "step": 105
+    },
+    {
+      "epoch": 0.22,
+      "learning_rate": 1.9906071931421413e-05,
+      "loss": 0.7121,
+      "step": 106
+    },
+    {
+      "epoch": 0.22,
+      "learning_rate": 1.9903022504951098e-05,
+      "loss": 0.7052,
+      "step": 107
+    },
+    {
+      "epoch": 0.22,
+      "learning_rate": 1.9899924606513593e-05,
+      "loss": 0.6915,
+      "step": 108
+    },
+    {
+      "epoch": 0.22,
+      "learning_rate": 1.989677825127208e-05,
+      "loss": 0.7078,
+      "step": 109
+    },
+    {
+      "epoch": 0.23,
+      "learning_rate": 1.9893583454626905e-05,
+      "loss": 0.6683,
+      "step": 110
+    },
+    {
+      "epoch": 0.23,
+      "learning_rate": 1.9890340232215517e-05,
+      "loss": 0.6911,
+      "step": 111
+    },
+    {
+      "epoch": 0.23,
+      "learning_rate": 1.9887048599912412e-05,
+      "loss": 0.6564,
+      "step": 112
+    },
+    {
+      "epoch": 0.23,
+      "learning_rate": 1.988370857382902e-05,
+      "loss": 0.6423,
+      "step": 113
+    },
+    {
+      "epoch": 0.23,
+      "learning_rate": 1.988032017031364e-05,
+      "loss": 0.6684,
+      "step": 114
+    },
+    {
+      "epoch": 0.24,
+      "learning_rate": 1.9876883405951378e-05,
+      "loss": 0.725,
+      "step": 115
+    },
+    {
+      "epoch": 0.24,
+      "learning_rate": 1.9873398297564036e-05,
+      "loss": 0.6663,
+      "step": 116
+    },
+    {
+      "epoch": 0.24,
+      "learning_rate": 1.9869864862210045e-05,
+      "loss": 0.6671,
+      "step": 117
+    },
+    {
+      "epoch": 0.24,
+      "learning_rate": 1.9866283117184393e-05,
+      "loss": 0.6233,
+      "step": 118
+    },
+    {
+      "epoch": 0.24,
+      "learning_rate": 1.9862653080018508e-05,
+      "loss": 0.6376,
+      "step": 119
+    },
+    {
+      "epoch": 0.25,
+      "learning_rate": 1.9858974768480206e-05,
+      "loss": 0.6803,
+      "step": 120
+    },
+    {
+      "epoch": 0.25,
+      "learning_rate": 1.985524820057359e-05,
+      "loss": 0.6626,
+      "step": 121
+    },
+    {
+      "epoch": 0.25,
+      "learning_rate": 1.985147339453895e-05,
+      "loss": 0.6307,
+      "step": 122
+    },
+    {
+      "epoch": 0.25,
+      "learning_rate": 1.9847650368852697e-05,
+      "loss": 0.6416,
+      "step": 123
+    },
+    {
+      "epoch": 0.25,
+      "learning_rate": 1.9843779142227258e-05,
+      "loss": 0.6992,
+      "step": 124
+    },
+    {
+      "epoch": 0.26,
+      "learning_rate": 1.9839859733610984e-05,
+      "loss": 0.6593,
+      "step": 125
+    },
+    {
+      "epoch": 0.26,
+      "learning_rate": 1.9835892162188065e-05,
+      "loss": 0.6937,
+      "step": 126
+    },
+    {
+      "epoch": 0.26,
+      "learning_rate": 1.983187644737843e-05,
+      "loss": 0.6826,
+      "step": 127
+    },
+    {
+      "epoch": 0.26,
+      "learning_rate": 1.982781260883765e-05,
+      "loss": 0.6342,
+      "step": 128
+    },
+    {
+      "epoch": 0.26,
+      "learning_rate": 1.9823700666456854e-05,
+      "loss": 0.6551,
+      "step": 129
+    },
+    {
+      "epoch": 0.27,
+      "learning_rate": 1.9819540640362612e-05,
+      "loss": 0.6724,
+      "step": 130
+    },
+    {
+      "epoch": 0.27,
+      "learning_rate": 1.9815332550916864e-05,
+      "loss": 0.687,
+      "step": 131
+    },
+    {
+      "epoch": 0.27,
+      "learning_rate": 1.981107641871678e-05,
+      "loss": 0.6661,
+      "step": 132
+    },
+    {
+      "epoch": 0.27,
+      "learning_rate": 1.980677226459471e-05,
+      "loss": 0.6977,
+      "step": 133
+    },
+    {
+      "epoch": 0.27,
+      "learning_rate": 1.980242010961803e-05,
+      "loss": 0.7067,
+      "step": 134
+    },
+    {
+      "epoch": 0.28,
+      "learning_rate": 1.9798019975089083e-05,
+      "loss": 0.685,
+      "step": 135
+    },
+    {
+      "epoch": 0.28,
+      "learning_rate": 1.9793571882545048e-05,
+      "loss": 0.6531,
+      "step": 136
+    },
+    {
+      "epoch": 0.28,
+      "learning_rate": 1.9789075853757845e-05,
+      "loss": 0.6709,
+      "step": 137
+    },
+    {
+      "epoch": 0.28,
+      "learning_rate": 1.978453191073402e-05,
+      "loss": 0.6399,
+      "step": 138
+    },
+    {
+      "epoch": 0.28,
+      "learning_rate": 1.977994007571465e-05,
+      "loss": 0.6996,
+      "step": 139
+    },
+    {
+      "epoch": 0.29,
+      "learning_rate": 1.9775300371175225e-05,
+      "loss": 0.6513,
+      "step": 140
+    },
+    {
+      "epoch": 0.29,
+      "learning_rate": 1.9770612819825534e-05,
+      "loss": 0.692,
+      "step": 141
+    },
+    {
+      "epoch": 0.29,
+      "learning_rate": 1.976587744460957e-05,
+      "loss": 0.5954,
+      "step": 142
+    },
+    {
+      "epoch": 0.29,
+      "learning_rate": 1.97610942687054e-05,
+      "loss": 0.5954,
+      "step": 143
+    },
+    {
+      "epoch": 0.29,
+      "learning_rate": 1.975626331552507e-05,
+      "loss": 0.7209,
+      "step": 144
+    },
+    {
+      "epoch": 0.3,
+      "learning_rate": 1.9751384608714466e-05,
+      "loss": 0.6777,
+      "step": 145
+    },
+    {
+      "epoch": 0.3,
+      "learning_rate": 1.974645817215322e-05,
+      "loss": 0.6925,
+      "step": 146
+    },
+    {
+      "epoch": 0.3,
+      "learning_rate": 1.9741484029954582e-05,
+      "loss": 0.6353,
+      "step": 147
+    },
+    {
+      "epoch": 0.3,
+      "learning_rate": 1.973646220646531e-05,
+      "loss": 0.6056,
+      "step": 148
+    },
+    {
+      "epoch": 0.3,
+      "learning_rate": 1.9731392726265538e-05,
+      "loss": 0.7167,
+      "step": 149
+    },
+    {
+      "epoch": 0.31,
+      "learning_rate": 1.972627561416867e-05,
+      "loss": 0.7169,
+      "step": 150
+    },
+    {
+      "epoch": 0.31,
+      "learning_rate": 1.9721110895221252e-05,
+      "loss": 0.6759,
+      "step": 151
+    },
+    {
+      "epoch": 0.31,
+      "learning_rate": 1.9715898594702845e-05,
+      "loss": 0.663,
+      "step": 152
+    },
+    {
+      "epoch": 0.31,
+      "learning_rate": 1.9710638738125908e-05,
+      "loss": 0.6235,
+      "step": 153
+    },
+    {
+      "epoch": 0.32,
+      "learning_rate": 1.9705331351235673e-05,
+      "loss": 0.5991,
+      "step": 154
+    },
+    {
+      "epoch": 0.32,
+      "learning_rate": 1.9699976460010016e-05,
+      "loss": 0.6922,
+      "step": 155
+    },
+    {
+      "epoch": 0.32,
+      "learning_rate": 1.969457409065933e-05,
+      "loss": 0.6976,
+      "step": 156
+    },
+    {
+      "epoch": 0.32,
+      "learning_rate": 1.9689124269626393e-05,
+      "loss": 0.6361,
+      "step": 157
+    },
+    {
+      "epoch": 0.32,
+      "learning_rate": 1.9683627023586252e-05,
+      "loss": 0.6306,
+      "step": 158
+    },
+    {
+      "epoch": 0.33,
+      "learning_rate": 1.967808237944608e-05,
+      "loss": 0.6463,
+      "step": 159
+    },
+    {
+      "epoch": 0.33,
+      "learning_rate": 1.967249036434504e-05,
+      "loss": 0.6487,
+      "step": 160
+    },
+    {
+      "epoch": 0.33,
+      "learning_rate": 1.9666851005654174e-05,
+      "loss": 0.6386,
+      "step": 161
+    },
+    {
+      "epoch": 0.33,
+      "learning_rate": 1.9661164330976246e-05,
+      "loss": 0.6127,
+      "step": 162
+    },
+    {
+      "epoch": 0.33,
+      "learning_rate": 1.965543036814561e-05,
+      "loss": 0.6314,
+      "step": 163
+    },
+    {
+      "epoch": 0.34,
+      "learning_rate": 1.96496491452281e-05,
+      "loss": 0.6329,
+      "step": 164
+    },
+    {
+      "epoch": 0.34,
+      "learning_rate": 1.964382069052085e-05,
+      "loss": 0.6284,
+      "step": 165
+    },
+    {
+      "epoch": 0.34,
+      "learning_rate": 1.963794503255219e-05,
+      "loss": 0.639,
+      "step": 166
+    },
+    {
+      "epoch": 0.34,
+      "learning_rate": 1.9632022200081496e-05,
+      "loss": 0.6172,
+      "step": 167
+    },
+    {
+      "epoch": 0.34,
+      "learning_rate": 1.962605222209903e-05,
+      "loss": 0.6282,
+      "step": 168
+    },
+    {
+      "epoch": 0.35,
+      "learning_rate": 1.962003512782584e-05,
+      "loss": 0.6409,
+      "step": 169
+    },
+    {
+      "epoch": 0.35,
+      "learning_rate": 1.9613970946713574e-05,
+      "loss": 0.5883,
+      "step": 170
+    },
+    {
+      "epoch": 0.35,
+      "learning_rate": 1.9607859708444364e-05,
+      "loss": 0.6287,
+      "step": 171
+    },
+    {
+      "epoch": 0.35,
+      "learning_rate": 1.9601701442930667e-05,
+      "loss": 0.6535,
+      "step": 172
+    },
+    {
+      "epoch": 0.35,
+      "learning_rate": 1.959549618031512e-05,
+      "loss": 0.6392,
+      "step": 173
+    },
+    {
+      "epoch": 0.36,
+      "learning_rate": 1.95892439509704e-05,
+      "loss": 0.6127,
+      "step": 174
+    },
+    {
+      "epoch": 0.36,
+      "learning_rate": 1.9582944785499076e-05,
+      "loss": 0.6815,
+      "step": 175
+    },
+    {
+      "epoch": 0.36,
+      "learning_rate": 1.9576598714733434e-05,
+      "loss": 0.5827,
+      "step": 176
+    },
+    {
+      "epoch": 0.36,
+      "learning_rate": 1.957020576973536e-05,
+      "loss": 0.6628,
+      "step": 177
+    },
+    {
+      "epoch": 0.36,
+      "learning_rate": 1.956376598179618e-05,
+      "loss": 0.6723,
+      "step": 178
+    },
+    {
+      "epoch": 0.37,
+      "learning_rate": 1.9557279382436483e-05,
+      "loss": 0.6952,
+      "step": 179
+    },
+    {
+      "epoch": 0.37,
+      "learning_rate": 1.9550746003405996e-05,
+      "loss": 0.7105,
+      "step": 180
+    },
+    {
+      "epoch": 0.37,
+      "learning_rate": 1.9544165876683413e-05,
+      "loss": 0.6247,
+      "step": 181
+    },
+    {
+      "epoch": 0.37,
+      "learning_rate": 1.9537539034476247e-05,
+      "loss": 0.6238,
+      "step": 182
+    },
+    {
+      "epoch": 0.37,
+      "learning_rate": 1.953086550922066e-05,
+      "loss": 0.6857,
+      "step": 183
+    },
+    {
+      "epoch": 0.38,
+      "learning_rate": 1.9524145333581315e-05,
+      "loss": 0.6661,
+      "step": 184
+    },
+    {
+      "epoch": 0.38,
+      "learning_rate": 1.9517378540451215e-05,
+      "loss": 0.6574,
+      "step": 185
+    },
+    {
+      "epoch": 0.38,
+      "learning_rate": 1.9510565162951538e-05,
+      "loss": 0.6571,
+      "step": 186
+    },
+    {
+      "epoch": 0.38,
+      "learning_rate": 1.9503705234431477e-05,
+      "loss": 0.6454,
+      "step": 187
+    },
+    {
+      "epoch": 0.38,
+      "learning_rate": 1.9496798788468077e-05,
+      "loss": 0.6569,
+      "step": 188
+    },
+    {
+      "epoch": 0.39,
+      "learning_rate": 1.9489845858866066e-05,
+      "loss": 0.572,
+      "step": 189
+    },
+    {
+      "epoch": 0.39,
+      "learning_rate": 1.9482846479657707e-05,
+      "loss": 0.6289,
+      "step": 190
+    },
+    {
+      "epoch": 0.39,
+      "learning_rate": 1.94758006851026e-05,
+      "loss": 0.6365,
+      "step": 191
+    },
+    {
+      "epoch": 0.39,
+      "learning_rate": 1.9468708509687545e-05,
+      "loss": 0.6705,
+      "step": 192
+    },
+    {
+      "epoch": 0.39,
+      "learning_rate": 1.946156998812636e-05,
+      "loss": 0.6304,
+      "step": 193
+    },
+    {
+      "epoch": 0.4,
+      "learning_rate": 1.9454385155359704e-05,
+      "loss": 0.6822,
+      "step": 194
+    },
+    {
+      "epoch": 0.4,
+      "learning_rate": 1.9447154046554923e-05,
+      "loss": 0.6451,
+      "step": 195
+    },
+    {
+      "epoch": 0.4,
+      "learning_rate": 1.943987669710586e-05,
+      "loss": 0.6208,
+      "step": 196
+    },
+    {
+      "epoch": 0.4,
+      "learning_rate": 1.9432553142632704e-05,
+      "loss": 0.6142,
+      "step": 197
+    },
+    {
+      "epoch": 0.41,
+      "learning_rate": 1.9425183418981783e-05,
+      "loss": 0.5998,
+      "step": 198
+    },
+    {
+      "epoch": 0.41,
+      "learning_rate": 1.9417767562225422e-05,
+      "loss": 0.6414,
+      "step": 199
+    },
+    {
+      "epoch": 0.41,
+      "learning_rate": 1.9410305608661745e-05,
+      "loss": 0.6145,
+      "step": 200
+    },
+    {
+      "epoch": 0.41,
+      "learning_rate": 1.9402797594814517e-05,
+      "loss": 0.5703,
+      "step": 201
+    },
+    {
+      "epoch": 0.41,
+      "learning_rate": 1.939524355743293e-05,
+      "loss": 0.5993,
+      "step": 202
+    },
+    {
+      "epoch": 0.42,
+      "learning_rate": 1.9387643533491462e-05,
+      "loss": 0.5606,
+      "step": 203
+    },
+    {
+      "epoch": 0.42,
+      "learning_rate": 1.9379997560189677e-05,
+      "loss": 0.5842,
+      "step": 204
+    },
+    {
+      "epoch": 0.42,
+      "learning_rate": 1.937230567495205e-05,
+      "loss": 0.6151,
+      "step": 205
+    },
+    {
+      "epoch": 0.42,
+      "learning_rate": 1.936456791542776e-05,
+      "loss": 0.6354,
+      "step": 206
+    },
+    {
+      "epoch": 0.42,
+      "learning_rate": 1.935678431949056e-05,
+      "loss": 0.6016,
+      "step": 207
+    },
+    {
+      "epoch": 0.43,
+      "learning_rate": 1.934895492523852e-05,
+      "loss": 0.6464,
+      "step": 208
+    },
+    {
+      "epoch": 0.43,
+      "learning_rate": 1.93410797709939e-05,
+      "loss": 0.6306,
+      "step": 209
+    },
+    {
+      "epoch": 0.43,
+      "learning_rate": 1.9333158895302934e-05,
+      "loss": 0.712,
+      "step": 210
+    },
+    {
+      "epoch": 0.43,
+      "learning_rate": 1.9325192336935646e-05,
+      "loss": 0.6244,
+      "step": 211
+    },
+    {
+      "epoch": 0.43,
+      "learning_rate": 1.931718013488566e-05,
+      "loss": 0.6237,
+      "step": 212
+    },
+    {
+      "epoch": 0.44,
+      "learning_rate": 1.9309122328370013e-05,
+      "loss": 0.5864,
+      "step": 213
+    },
+    {
+      "epoch": 0.44,
+      "learning_rate": 1.9301018956828966e-05,
+      "loss": 0.6459,
+      "step": 214
+    },
+    {
+      "epoch": 0.44,
+      "learning_rate": 1.929287005992579e-05,
+      "loss": 0.6109,
+      "step": 215
+    },
+    {
+      "epoch": 0.44,
+      "learning_rate": 1.92846756775466e-05,
+      "loss": 0.606,
+      "step": 216
+    },
+    {
+      "epoch": 0.44,
+      "learning_rate": 1.927643584980015e-05,
+      "loss": 0.6363,
+      "step": 217
+    },
+    {
+      "epoch": 0.45,
+      "learning_rate": 1.926815061701762e-05,
+      "loss": 0.6476,
+      "step": 218
+    },
+    {
+      "epoch": 0.45,
+      "learning_rate": 1.9259820019752445e-05,
+      "loss": 0.5801,
+      "step": 219
+    },
+    {
+      "epoch": 0.45,
+      "learning_rate": 1.9251444098780095e-05,
+      "loss": 0.6469,
+      "step": 220
+    },
+    {
+      "epoch": 0.45,
+      "learning_rate": 1.924302289509789e-05,
+      "loss": 0.5569,
+      "step": 221
+    },
+    {
+      "epoch": 0.45,
+      "learning_rate": 1.9234556449924796e-05,
+      "loss": 0.5868,
+      "step": 222
+    },
+    {
+      "epoch": 0.46,
+      "learning_rate": 1.9226044804701214e-05,
+      "loss": 0.6507,
+      "step": 223
+    },
+    {
+      "epoch": 0.46,
+      "learning_rate": 1.9217488001088784e-05,
+      "loss": 0.6277,
+      "step": 224
+    },
+    {
+      "epoch": 0.46,
+      "learning_rate": 1.9208886080970187e-05,
+      "loss": 0.6239,
+      "step": 225
+    },
+    {
+      "epoch": 0.46,
+      "learning_rate": 1.9200239086448934e-05,
+      "loss": 0.624,
+      "step": 226
+    },
+    {
+      "epoch": 0.46,
+      "learning_rate": 1.9191547059849155e-05,
+      "loss": 0.624,
+      "step": 227
+    },
+    {
+      "epoch": 0.47,
+      "learning_rate": 1.9182810043715394e-05,
+      "loss": 0.6085,
+      "step": 228
+    },
+    {
+      "epoch": 0.47,
+      "learning_rate": 1.9174028080812415e-05,
+      "loss": 0.6327,
+      "step": 229
+    },
+    {
+      "epoch": 0.47,
+      "learning_rate": 1.9165201214124972e-05,
+      "loss": 0.5803,
+      "step": 230
+    },
+    {
+      "epoch": 0.47,
+      "learning_rate": 1.9156329486857615e-05,
+      "loss": 0.6179,
+      "step": 231
+    },
+    {
+      "epoch": 0.47,
+      "learning_rate": 1.9147412942434463e-05,
+      "loss": 0.5817,
+      "step": 232
+    },
+    {
+      "epoch": 0.48,
+      "learning_rate": 1.9138451624499007e-05,
+      "loss": 0.6135,
+      "step": 233
+    },
+    {
+      "epoch": 0.48,
+      "learning_rate": 1.9129445576913886e-05,
+      "loss": 0.5763,
+      "step": 234
+    },
+    {
+      "epoch": 0.48,
+      "learning_rate": 1.912039484376068e-05,
+      "loss": 0.5978,
+      "step": 235
+    },
+    {
+      "epoch": 0.48,
+      "learning_rate": 1.911129946933968e-05,
+      "loss": 0.6477,
+      "step": 236
+    },
+    {
+      "epoch": 0.49,
+      "learning_rate": 1.910215949816969e-05,
+      "loss": 0.6291,
+      "step": 237
+    },
+    {
+      "epoch": 0.49,
+      "learning_rate": 1.90929749749878e-05,
+      "loss": 0.6731,
+      "step": 238
+    },
+    {
+      "epoch": 0.49,
+      "learning_rate": 1.9083745944749163e-05,
+      "loss": 0.7298,
+      "step": 239
+    },
+    {
+      "epoch": 0.49,
+      "learning_rate": 1.907447245262678e-05,
+      "loss": 0.6642,
+      "step": 240
+    },
+    {
+      "epoch": 0.49,
+      "learning_rate": 1.9065154544011273e-05,
+      "loss": 0.6143,
+      "step": 241
+    },
+    {
+      "epoch": 0.5,
+      "learning_rate": 1.9055792264510683e-05,
+      "loss": 0.6343,
+      "step": 242
+    },
+    {
+      "epoch": 0.5,
+      "learning_rate": 1.9046385659950213e-05,
+      "loss": 0.6532,
+      "step": 243
+    },
+    {
+      "epoch": 0.5,
+      "learning_rate": 1.903693477637204e-05,
+      "loss": 0.6067,
+      "step": 244
+    },
+    {
+      "epoch": 0.5,
+      "learning_rate": 1.902743966003506e-05,
+      "loss": 0.5724,
+      "step": 245
+    },
+    {
+      "epoch": 0.5,
+      "learning_rate": 1.901790035741467e-05,
+      "loss": 0.5935,
+      "step": 246
+    },
+    {
+      "epoch": 0.51,
+      "learning_rate": 1.9008316915202555e-05,
+      "loss": 0.646,
+      "step": 247
+    },
+    {
+      "epoch": 0.51,
+      "learning_rate": 1.899868938030645e-05,
+      "loss": 0.6535,
+      "step": 248
+    },
+    {
+      "epoch": 0.51,
+      "learning_rate": 1.8989017799849896e-05,
+      "loss": 0.6427,
+      "step": 249
+    },
+    {
+      "epoch": 0.51,
+      "learning_rate": 1.8979302221172027e-05,
+      "loss": 0.6327,
+      "step": 250
+    },
+    {
+      "epoch": 0.51,
+      "learning_rate": 1.8969542691827346e-05,
+      "loss": 0.6216,
+      "step": 251
+    },
+    {
+      "epoch": 0.52,
+      "learning_rate": 1.8959739259585458e-05,
+      "loss": 0.6665,
+      "step": 252
+    },
+    {
+      "epoch": 0.52,
+      "learning_rate": 1.8949891972430878e-05,
+      "loss": 0.53,
+      "step": 253
+    },
+    {
+      "epoch": 0.52,
+      "learning_rate": 1.8940000878562758e-05,
+      "loss": 0.5817,
+      "step": 254
+    },
+    {
+      "epoch": 0.52,
+      "learning_rate": 1.893006602639469e-05,
+      "loss": 0.6036,
+      "step": 255
+    },
+    {
+      "epoch": 0.52,
+      "learning_rate": 1.8920087464554427e-05,
+      "loss": 0.6375,
+      "step": 256
+    },
+    {
+      "epoch": 0.53,
+      "learning_rate": 1.891006524188368e-05,
+      "loss": 0.5278,
+      "step": 257
+    },
+    {
+      "epoch": 0.53,
+      "learning_rate": 1.889999940743786e-05,
+      "loss": 0.5926,
+      "step": 258
+    },
+    {
+      "epoch": 0.53,
+      "learning_rate": 1.8889890010485847e-05,
+      "loss": 0.6125,
+      "step": 259
+    },
+    {
+      "epoch": 0.53,
+      "learning_rate": 1.887973710050974e-05,
+      "loss": 0.6059,
+      "step": 260
+    },
+    {
+      "epoch": 0.53,
+      "learning_rate": 1.8869540727204615e-05,
+      "loss": 0.6221,
+      "step": 261
+    },
+    {
+      "epoch": 0.54,
+      "learning_rate": 1.8859300940478304e-05,
+      "loss": 0.5928,
+      "step": 262
+    },
+    {
+      "epoch": 0.54,
+      "learning_rate": 1.8849017790451114e-05,
+      "loss": 0.5739,
+      "step": 263
+    },
+    {
+      "epoch": 0.54,
+      "learning_rate": 1.883869132745561e-05,
+      "loss": 0.5876,
+      "step": 264
+    },
+    {
+      "epoch": 0.54,
+      "learning_rate": 1.8828321602036363e-05,
+      "loss": 0.5671,
+      "step": 265
+    },
+    {
+      "epoch": 0.54,
+      "learning_rate": 1.881790866494969e-05,
+      "loss": 0.616,
+      "step": 266
+    },
+    {
+      "epoch": 0.55,
+      "learning_rate": 1.8807452567163413e-05,
+      "loss": 0.5381,
+      "step": 267
+    },
+    {
+      "epoch": 0.55,
+      "learning_rate": 1.8796953359856626e-05,
+      "loss": 0.5394,
+      "step": 268
+    },
+    {
+      "epoch": 0.55,
+      "learning_rate": 1.878641109441942e-05,
+      "loss": 0.5925,
+      "step": 269
+    },
+    {
+      "epoch": 0.55,
+      "learning_rate": 1.8775825822452636e-05,
+      "loss": 0.624,
+      "step": 270
+    },
+    {
+      "epoch": 0.55,
+      "learning_rate": 1.8765197595767623e-05,
+      "loss": 0.6122,
+      "step": 271
+    },
+    {
+      "epoch": 0.56,
+      "learning_rate": 1.8754526466385986e-05,
+      "loss": 0.6139,
+      "step": 272
+    },
+    {
+      "epoch": 0.56,
+      "learning_rate": 1.874381248653931e-05,
+      "loss": 0.6111,
+      "step": 273
+    },
+    {
+      "epoch": 0.56,
+      "learning_rate": 1.8733055708668928e-05,
+      "loss": 0.5524,
+      "step": 274
+    },
+    {
+      "epoch": 0.56,
+      "learning_rate": 1.8722256185425654e-05,
+      "loss": 0.5514,
+      "step": 275
+    },
+    {
+      "epoch": 0.56,
+      "learning_rate": 1.871141396966953e-05,
+      "loss": 0.6109,
+      "step": 276
+    },
+    {
+      "epoch": 0.57,
+      "learning_rate": 1.8700529114469545e-05,
+      "loss": 0.6452,
+      "step": 277
+    },
+    {
+      "epoch": 0.57,
+      "learning_rate": 1.868960167310342e-05,
+      "loss": 0.5266,
+      "step": 278
+    },
+    {
+      "epoch": 0.57,
+      "learning_rate": 1.86786316990573e-05,
+      "loss": 0.5495,
+      "step": 279
+    },
+    {
+      "epoch": 0.57,
+      "learning_rate": 1.8667619246025527e-05,
+      "loss": 0.588,
+      "step": 280
+    },
+    {
+      "epoch": 0.58,
+      "learning_rate": 1.8656564367910358e-05,
+      "loss": 0.5112,
+      "step": 281
+    },
+    {
+      "epoch": 0.58,
+      "learning_rate": 1.86454671188217e-05,
+      "loss": 0.6438,
+      "step": 282
+    },
+    {
+      "epoch": 0.58,
+      "learning_rate": 1.8634327553076864e-05,
+      "loss": 0.5494,
+      "step": 283
+    },
+    {
+      "epoch": 0.58,
+      "learning_rate": 1.862314572520028e-05,
+      "loss": 0.5849,
+      "step": 284
+    },
+    {
+      "epoch": 0.58,
+      "learning_rate": 1.8611921689923234e-05,
+      "loss": 0.5857,
+      "step": 285
+    },
+    {
+      "epoch": 0.59,
+      "learning_rate": 1.860065550218361e-05,
+      "loss": 0.6192,
+      "step": 286
+    },
+    {
+      "epoch": 0.59,
+      "learning_rate": 1.858934721712561e-05,
+      "loss": 0.4881,
+      "step": 287
+    },
+    {
+      "epoch": 0.59,
+      "learning_rate": 1.8577996890099494e-05,
+      "loss": 0.541,
+      "step": 288
+    },
+    {
+      "epoch": 0.59,
+      "learning_rate": 1.8566604576661288e-05,
+      "loss": 0.52,
+      "step": 289
+    },
+    {
+      "epoch": 0.59,
+      "learning_rate": 1.8555170332572544e-05,
+      "loss": 0.5997,
+      "step": 290
+    },
+    {
+      "epoch": 0.6,
+      "learning_rate": 1.8543694213800044e-05,
+      "loss": 0.5964,
+      "step": 291
+    },
+    {
+      "epoch": 0.6,
+      "learning_rate": 1.8532176276515538e-05,
+      "loss": 0.5221,
+      "step": 292
+    },
+    {
+      "epoch": 0.6,
+      "learning_rate": 1.8520616577095447e-05,
+      "loss": 0.5878,
+      "step": 293
+    },
+    {
+      "epoch": 0.6,
+      "learning_rate": 1.850901517212062e-05,
+      "loss": 0.5868,
+      "step": 294
+    },
+    {
+      "epoch": 0.6,
+      "learning_rate": 1.8497372118376043e-05,
+      "loss": 0.5946,
+      "step": 295
+    },
+    {
+      "epoch": 0.61,
+      "learning_rate": 1.8485687472850538e-05,
+      "loss": 0.6517,
+      "step": 296
+    },
+    {
+      "epoch": 0.61,
+      "learning_rate": 1.847396129273653e-05,
+      "loss": 0.5609,
+      "step": 297
+    },
+    {
+      "epoch": 0.61,
+      "learning_rate": 1.8462193635429723e-05,
+      "loss": 0.5922,
+      "step": 298
+    },
+    {
+      "epoch": 0.61,
+      "learning_rate": 1.8450384558528848e-05,
+      "loss": 0.5165,
+      "step": 299
+    },
+    {
+      "epoch": 0.61,
+      "learning_rate": 1.8438534119835365e-05,
+      "loss": 0.5359,
+      "step": 300
+    },
+    {
+      "epoch": 0.62,
+      "learning_rate": 1.842664237735319e-05,
+      "loss": 0.607,
+      "step": 301
+    },
+    {
+      "epoch": 0.62,
+      "learning_rate": 1.841470938928841e-05,
+      "loss": 0.5162,
+      "step": 302
+    },
+    {
+      "epoch": 0.62,
+      "learning_rate": 1.8402735214048988e-05,
+      "loss": 0.5648,
+      "step": 303
+    },
+    {
+      "epoch": 0.62,
+      "learning_rate": 1.8390719910244487e-05,
+      "loss": 0.5554,
+      "step": 304
+    },
+    {
+      "epoch": 0.62,
+      "learning_rate": 1.8378663536685784e-05,
+      "loss": 0.5435,
+      "step": 305
+    },
+    {
+      "epoch": 0.63,
+      "learning_rate": 1.8366566152384773e-05,
+      "loss": 0.5684,
+      "step": 306
+    },
+    {
+      "epoch": 0.63,
+      "learning_rate": 1.8354427816554085e-05,
+      "loss": 0.6131,
+      "step": 307
+    },
+    {
+      "epoch": 0.63,
+      "learning_rate": 1.83422485886068e-05,
+      "loss": 0.5604,
+      "step": 308
+    },
+    {
+      "epoch": 0.63,
+      "learning_rate": 1.8330028528156138e-05,
+      "loss": 0.5868,
+      "step": 309
+    },
+    {
+      "epoch": 0.63,
+      "learning_rate": 1.8317767695015195e-05,
+      "loss": 0.5601,
+      "step": 310
+    },
+    {
+      "epoch": 0.64,
+      "learning_rate": 1.8305466149196624e-05,
+      "loss": 0.646,
+      "step": 311
+    },
+    {
+      "epoch": 0.64,
+      "learning_rate": 1.8293123950912362e-05,
+      "loss": 0.5813,
+      "step": 312
+    },
+    {
+      "epoch": 0.64,
+      "learning_rate": 1.8280741160573318e-05,
+      "loss": 0.6303,
+      "step": 313
+    },
+    {
+      "epoch": 0.64,
+      "learning_rate": 1.826831783878909e-05,
+      "loss": 0.5233,
+      "step": 314
+    },
+    {
+      "epoch": 0.64,
+      "learning_rate": 1.825585404636766e-05,
+      "loss": 0.5722,
+      "step": 315
+    },
+    {
+      "epoch": 0.65,
+      "learning_rate": 1.8243349844315116e-05,
+      "loss": 0.6125,
+      "step": 316
+    },
+    {
+      "epoch": 0.65,
+      "learning_rate": 1.8230805293835315e-05,
+      "loss": 0.5913,
+      "step": 317
+    },
+    {
+      "epoch": 0.65,
+      "learning_rate": 1.8218220456329618e-05,
+      "loss": 0.5453,
+      "step": 318
+    },
+    {
+      "epoch": 0.65,
+      "learning_rate": 1.820559539339657e-05,
+      "loss": 0.6059,
+      "step": 319
+    },
+    {
+      "epoch": 0.65,
+      "learning_rate": 1.8192930166831618e-05,
+      "loss": 0.5906,
+      "step": 320
+    },
+    {
+      "epoch": 0.66,
+      "learning_rate": 1.818022483862679e-05,
+      "loss": 0.5697,
+      "step": 321
+    },
+    {
+      "epoch": 0.66,
+      "learning_rate": 1.8167479470970394e-05,
+      "loss": 0.5952,
+      "step": 322
+    },
+    {
+      "epoch": 0.66,
+      "learning_rate": 1.8154694126246725e-05,
+      "loss": 0.5524,
+      "step": 323
+    },
+    {
+      "epoch": 0.66,
+      "learning_rate": 1.8141868867035745e-05,
+      "loss": 0.5772,
+      "step": 324
+    },
+    {
+      "epoch": 0.67,
+      "learning_rate": 1.812900375611279e-05,
+      "loss": 0.5916,
+      "step": 325
+    },
+    {
+      "epoch": 0.67,
+      "learning_rate": 1.8116098856448256e-05,
+      "loss": 0.5707,
+      "step": 326
+    },
+    {
+      "epoch": 0.67,
+      "learning_rate": 1.8103154231207286e-05,
+      "loss": 0.5902,
+      "step": 327
+    },
+    {
+      "epoch": 0.67,
+      "learning_rate": 1.8090169943749477e-05,
+      "loss": 0.6478,
+      "step": 328
+    },
+    {
+      "epoch": 0.67,
+      "learning_rate": 1.8077146057628547e-05,
+      "loss": 0.5788,
+      "step": 329
+    },
+    {
+      "epoch": 0.68,
+      "learning_rate": 1.8064082636592043e-05,
+      "loss": 0.5773,
+      "step": 330
+    },
+    {
+      "epoch": 0.68,
+      "learning_rate": 1.8050979744581022e-05,
+      "loss": 0.5594,
+      "step": 331
+    },
+    {
+      "epoch": 0.68,
+      "learning_rate": 1.8037837445729733e-05,
+      "loss": 0.5812,
+      "step": 332
+    },
+    {
+      "epoch": 0.68,
+      "learning_rate": 1.802465580436532e-05,
+      "loss": 0.5748,
+      "step": 333
+    },
+    {
+      "epoch": 0.68,
+      "learning_rate": 1.8011434885007482e-05,
+      "loss": 0.5564,
+      "step": 334
+    },
+    {
+      "epoch": 0.69,
+      "learning_rate": 1.7998174752368178e-05,
+      "loss": 0.6678,
+      "step": 335
+    },
+    {
+      "epoch": 0.69,
+      "learning_rate": 1.79848754713513e-05,
+      "loss": 0.6761,
+      "step": 336
+    },
+    {
+      "epoch": 0.69,
+      "learning_rate": 1.7971537107052364e-05,
+      "loss": 0.5973,
+      "step": 337
+    },
+    {
+      "epoch": 0.69,
+      "learning_rate": 1.7958159724758173e-05,
+      "loss": 0.5899,
+      "step": 338
+    },
+    {
+      "epoch": 0.69,
+      "learning_rate": 1.7944743389946524e-05,
+      "loss": 0.5607,
+      "step": 339
+    },
+    {
+      "epoch": 0.7,
+      "learning_rate": 1.7931288168285863e-05,
+      "loss": 0.5449,
+      "step": 340
+    },
+    {
+      "epoch": 0.7,
+      "learning_rate": 1.7917794125634983e-05,
+      "loss": 0.6624,
+      "step": 341
+    },
+    {
+      "epoch": 0.7,
+      "learning_rate": 1.790426132804268e-05,
+      "loss": 0.6022,
+      "step": 342
+    },
+    {
+      "epoch": 0.7,
+      "learning_rate": 1.7890689841747463e-05,
+      "loss": 0.5466,
+      "step": 343
+    },
+    {
+      "epoch": 0.7,
+      "learning_rate": 1.7877079733177185e-05,
+      "loss": 0.5806,
+      "step": 344
+    },
+    {
+      "epoch": 0.71,
+      "learning_rate": 1.7863431068948757e-05,
+      "loss": 0.5916,
+      "step": 345
+    },
+    {
+      "epoch": 0.71,
+      "learning_rate": 1.784974391586781e-05,
+      "loss": 0.5027,
+      "step": 346
+    },
+    {
+      "epoch": 0.71,
+      "learning_rate": 1.7836018340928348e-05,
+      "loss": 0.5486,
+      "step": 347
+    },
+    {
+      "epoch": 0.71,
+      "learning_rate": 1.7822254411312455e-05,
+      "loss": 0.5552,
+      "step": 348
+    },
+    {
+      "epoch": 0.71,
+      "learning_rate": 1.780845219438994e-05,
+      "loss": 0.6191,
+      "step": 349
+    },
+    {
+      "epoch": 0.72,
+      "learning_rate": 1.7794611757718013e-05,
+      "loss": 0.5775,
+      "step": 350
+    },
+    {
+      "epoch": 0.72,
+      "learning_rate": 1.7780733169040964e-05,
+      "loss": 0.5991,
+      "step": 351
+    },
+    {
+      "epoch": 0.72,
+      "learning_rate": 1.7766816496289822e-05,
+      "loss": 0.6355,
+      "step": 352
+    },
+    {
+      "epoch": 0.72,
+      "learning_rate": 1.775286180758202e-05,
+      "loss": 0.5947,
+      "step": 353
+    },
+    {
+      "epoch": 0.72,
+      "learning_rate": 1.773886917122107e-05,
+      "loss": 0.6049,
+      "step": 354
+    },
+    {
+      "epoch": 0.73,
+      "learning_rate": 1.7724838655696225e-05,
+      "loss": 0.5597,
+      "step": 355
+    },
+    {
+      "epoch": 0.73,
+      "learning_rate": 1.7710770329682145e-05,
+      "loss": 0.6115,
+      "step": 356
+    },
+    {
+      "epoch": 0.73,
+      "learning_rate": 1.7696664262038554e-05,
+      "loss": 0.5728,
+      "step": 357
+    },
+    {
+      "epoch": 0.73,
+      "learning_rate": 1.7682520521809918e-05,
+      "loss": 0.5278,
+      "step": 358
+    },
+    {
+      "epoch": 0.73,
+      "learning_rate": 1.766833917822509e-05,
+      "loss": 0.5627,
+      "step": 359
+    },
+    {
+      "epoch": 0.74,
+      "learning_rate": 1.765412030069698e-05,
+      "loss": 0.6724,
+      "step": 360
+    },
+    {
+      "epoch": 0.74,
+      "learning_rate": 1.7639863958822213e-05,
+      "loss": 0.5225,
+      "step": 361
+    },
+    {
+      "epoch": 0.74,
+      "learning_rate": 1.76255702223808e-05,
+      "loss": 0.551,
+      "step": 362
+    },
+    {
+      "epoch": 0.74,
+      "learning_rate": 1.761123916133577e-05,
+      "loss": 0.5401,
+      "step": 363
+    },
+    {
+      "epoch": 0.74,
+      "learning_rate": 1.759687084583285e-05,
+      "loss": 0.6538,
+      "step": 364
+    },
+    {
+      "epoch": 0.75,
+      "learning_rate": 1.758246534620012e-05,
+      "loss": 0.5547,
+      "step": 365
+    },
+    {
+      "epoch": 0.75,
+      "learning_rate": 1.756802273294766e-05,
+      "loss": 0.6067,
+      "step": 366
+    },
+    {
+      "epoch": 0.75,
+      "learning_rate": 1.7553543076767207e-05,
+      "loss": 0.5354,
+      "step": 367
+    },
+    {
+      "epoch": 0.75,
+      "learning_rate": 1.753902644853181e-05,
+      "loss": 0.5518,
+      "step": 368
+    },
+    {
+      "epoch": 0.76,
+      "learning_rate": 1.7524472919295488e-05,
+      "loss": 0.5829,
+      "step": 369
+    },
+    {
+      "epoch": 0.76,
+      "learning_rate": 1.750988256029287e-05,
+      "loss": 0.5488,
+      "step": 370
+    },
+    {
+      "epoch": 0.76,
+      "learning_rate": 1.7495255442938873e-05,
+      "loss": 0.5554,
+      "step": 371
+    },
+    {
+      "epoch": 0.76,
+      "learning_rate": 1.748059163882831e-05,
+      "loss": 0.5965,
+      "step": 372
+    },
+    {
+      "epoch": 0.76,
+      "learning_rate": 1.7465891219735576e-05,
+      "loss": 0.5826,
+      "step": 373
+    },
+    {
+      "epoch": 0.77,
+      "learning_rate": 1.7451154257614287e-05,
+      "loss": 0.4955,
+      "step": 374
+    },
+    {
+      "epoch": 0.77,
+      "learning_rate": 1.7436380824596915e-05,
+      "loss": 0.5438,
+      "step": 375
+    },
+    {
+      "epoch": 0.77,
+      "learning_rate": 1.742157099299445e-05,
+      "loss": 0.5584,
+      "step": 376
+    },
+    {
+      "epoch": 0.77,
+      "learning_rate": 1.7406724835296045e-05,
+      "loss": 0.5441,
+      "step": 377
+    },
+    {
+      "epoch": 0.77,
+      "learning_rate": 1.7391842424168652e-05,
+      "loss": 0.611,
+      "step": 378
+    },
+    {
+      "epoch": 0.78,
+      "learning_rate": 1.7376923832456665e-05,
+      "loss": 0.5766,
+      "step": 379
+    },
+    {
+      "epoch": 0.78,
+      "learning_rate": 1.7361969133181585e-05,
+      "loss": 0.5686,
+      "step": 380
+    },
+    {
+      "epoch": 0.78,
+      "learning_rate": 1.7346978399541638e-05,
+      "loss": 0.4756,
+      "step": 381
+    },
+    {
+      "epoch": 0.78,
+      "learning_rate": 1.7331951704911426e-05,
+      "loss": 0.5605,
+      "step": 382
+    },
+    {
+      "epoch": 0.78,
+      "learning_rate": 1.731688912284157e-05,
+      "loss": 0.5667,
+      "step": 383
+    },
+    {
+      "epoch": 0.79,
+      "learning_rate": 1.7301790727058344e-05,
+      "loss": 0.545,
+      "step": 384
+    },
+    {
+      "epoch": 0.79,
+      "learning_rate": 1.7286656591463325e-05,
+      "loss": 0.5494,
+      "step": 385
+    },
+    {
+      "epoch": 0.79,
+      "learning_rate": 1.7271486790133022e-05,
+      "loss": 0.5661,
+      "step": 386
+    },
+    {
+      "epoch": 0.79,
+      "learning_rate": 1.725628139731851e-05,
+      "loss": 0.5165,
+      "step": 387
+    },
+    {
+      "epoch": 0.79,
+      "learning_rate": 1.7241040487445082e-05,
+      "loss": 0.5775,
+      "step": 388
+    },
+    {
+      "epoch": 0.8,
+      "learning_rate": 1.7225764135111867e-05,
+      "loss": 0.5586,
+      "step": 389
+    },
+    {
+      "epoch": 0.8,
+      "learning_rate": 1.7210452415091477e-05,
+      "loss": 0.515,
+      "step": 390
+    },
+    {
+      "epoch": 0.8,
+      "learning_rate": 1.7195105402329636e-05,
+      "loss": 0.5611,
+      "step": 391
+    },
+    {
+      "epoch": 0.8,
+      "learning_rate": 1.7179723171944812e-05,
+      "loss": 0.5666,
+      "step": 392
+    },
+    {
+      "epoch": 0.8,
+      "learning_rate": 1.716430579922786e-05,
+      "loss": 0.6011,
+      "step": 393
+    },
+    {
+      "epoch": 0.81,
+      "learning_rate": 1.7148853359641627e-05,
+      "loss": 0.588,
+      "step": 394
+    },
+    {
+      "epoch": 0.81,
+      "learning_rate": 1.713336592882062e-05,
+      "loss": 0.5629,
+      "step": 395
+    },
+    {
+      "epoch": 0.81,
+      "learning_rate": 1.7117843582570608e-05,
+      "loss": 0.5584,
+      "step": 396
+    },
+    {
+      "epoch": 0.81,
+      "learning_rate": 1.7102286396868254e-05,
+      "loss": 0.5626,
+      "step": 397
+    },
+    {
+      "epoch": 0.81,
+      "learning_rate": 1.708669444786076e-05,
+      "loss": 0.5522,
+      "step": 398
+    },
+    {
+      "epoch": 0.82,
+      "learning_rate": 1.7071067811865477e-05,
+      "loss": 0.5697,
+      "step": 399
+    },
+    {
+      "epoch": 0.82,
+      "learning_rate": 1.705540656536953e-05,
+      "loss": 0.5854,
+      "step": 400
+    },
+    {
+      "epoch": 0.82,
+      "learning_rate": 1.7039710785029468e-05,
+      "loss": 0.5569,
+      "step": 401
+    },
+    {
+      "epoch": 0.82,
+      "learning_rate": 1.702398054767085e-05,
+      "loss": 0.52,
+      "step": 402
+    },
+    {
+      "epoch": 0.82,
+      "learning_rate": 1.7008215930287906e-05,
+      "loss": 0.5795,
+      "step": 403
+    },
+    {
+      "epoch": 0.83,
+      "learning_rate": 1.6992417010043144e-05,
+      "loss": 0.5428,
+      "step": 404
+    },
+    {
+      "epoch": 0.83,
+      "learning_rate": 1.6976583864266964e-05,
+      "loss": 0.579,
+      "step": 405
+    },
+    {
+      "epoch": 0.83,
+      "learning_rate": 1.6960716570457295e-05,
+      "loss": 0.6103,
+      "step": 406
+    },
+    {
+      "epoch": 0.83,
+      "learning_rate": 1.694481520627921e-05,
+      "loss": 0.5511,
+      "step": 407
+    },
+    {
+      "epoch": 0.83,
+      "learning_rate": 1.692887984956454e-05,
+      "loss": 0.5227,
+      "step": 408
+    },
+    {
+      "epoch": 0.84,
+      "learning_rate": 1.6912910578311503e-05,
+      "loss": 0.5095,
+      "step": 409
+    },
+    {
+      "epoch": 0.84,
+      "learning_rate": 1.6896907470684316e-05,
+      "loss": 0.5619,
+      "step": 410
+    },
+    {
+      "epoch": 0.84,
+      "learning_rate": 1.688087060501281e-05,
+      "loss": 0.5409,
+      "step": 411
+    },
+    {
+      "epoch": 0.84,
+      "learning_rate": 1.6864800059792057e-05,
+      "loss": 0.5767,
+      "step": 412
+    },
+    {
+      "epoch": 0.85,
+      "learning_rate": 1.684869591368197e-05,
+      "loss": 0.5357,
+      "step": 413
+    },
+    {
+      "epoch": 0.85,
+      "learning_rate": 1.6832558245506937e-05,
+      "loss": 0.5583,
+      "step": 414
+    },
+    {
+      "epoch": 0.85,
+      "learning_rate": 1.6816387134255418e-05,
+      "loss": 0.5937,
+      "step": 415
+    },
+    {
+      "epoch": 0.85,
+      "learning_rate": 1.6800182659079568e-05,
+      "loss": 0.546,
+      "step": 416
+    },
+    {
+      "epoch": 0.85,
+      "learning_rate": 1.678394489929485e-05,
+      "loss": 0.5565,
+      "step": 417
+    },
+    {
+      "epoch": 0.86,
+      "learning_rate": 1.676767393437964e-05,
+      "loss": 0.569,
+      "step": 418
+    },
+    {
+      "epoch": 0.86,
+      "learning_rate": 1.6751369843974842e-05,
+      "loss": 0.5867,
+      "step": 419
+    },
+    {
+      "epoch": 0.86,
+      "learning_rate": 1.6735032707883502e-05,
+      "loss": 0.5091,
+      "step": 420
+    },
+    {
+      "epoch": 0.86,
+      "learning_rate": 1.671866260607041e-05,
+      "loss": 0.5186,
+      "step": 421
+    },
+    {
+      "epoch": 0.86,
+      "learning_rate": 1.670225961866171e-05,
+      "loss": 0.5205,
+      "step": 422
+    },
+    {
+      "epoch": 0.87,
+      "learning_rate": 1.6685823825944515e-05,
+      "loss": 0.5012,
+      "step": 423
+    },
+    {
+      "epoch": 0.87,
+      "learning_rate": 1.666935530836651e-05,
+      "loss": 0.5475,
+      "step": 424
+    },
+    {
+      "epoch": 0.87,
+      "learning_rate": 1.665285414653555e-05,
+      "loss": 0.5368,
+      "step": 425
+    },
+    {
+      "epoch": 0.87,
+      "learning_rate": 1.663632042121928e-05,
+      "loss": 0.5657,
+      "step": 426
+    },
+    {
+      "epoch": 0.87,
+      "learning_rate": 1.6619754213344727e-05,
+      "loss": 0.5456,
+      "step": 427
+    },
+    {
+      "epoch": 0.88,
+      "learning_rate": 1.6603155603997908e-05,
+      "loss": 0.5521,
+      "step": 428
+    },
+    {
+      "epoch": 0.88,
+      "learning_rate": 1.658652467442345e-05,
+      "loss": 0.535,
+      "step": 429
+    },
+    {
+      "epoch": 0.88,
+      "learning_rate": 1.656986150602415e-05,
+      "loss": 0.5959,
+      "step": 430
+    },
+    {
+      "epoch": 0.88,
+      "learning_rate": 1.6553166180360622e-05,
+      "loss": 0.5514,
+      "step": 431
+    },
+    {
+      "epoch": 0.88,
+      "learning_rate": 1.653643877915088e-05,
+      "loss": 0.5134,
+      "step": 432
+    },
+    {
+      "epoch": 0.89,
+      "learning_rate": 1.6519679384269927e-05,
+      "loss": 0.5502,
+      "step": 433
+    },
+    {
+      "epoch": 0.89,
+      "learning_rate": 1.650288807774937e-05,
+      "loss": 0.5709,
+      "step": 434
+    },
+    {
+      "epoch": 0.89,
+      "learning_rate": 1.648606494177701e-05,
+      "loss": 0.5654,
+      "step": 435
+    },
+    {
+      "epoch": 0.89,
+      "learning_rate": 1.6469210058696448e-05,
+      "loss": 0.5226,
+      "step": 436
+    },
+    {
+      "epoch": 0.89,
+      "learning_rate": 1.6452323511006664e-05,
+      "loss": 0.5576,
+      "step": 437
+    },
+    {
+      "epoch": 0.9,
+      "learning_rate": 1.6435405381361645e-05,
+      "loss": 0.638,
+      "step": 438
+    },
+    {
+      "epoch": 0.9,
+      "learning_rate": 1.6418455752569945e-05,
+      "loss": 0.6046,
+      "step": 439
+    },
+    {
+      "epoch": 0.9,
+      "learning_rate": 1.64014747075943e-05,
+      "loss": 0.5463,
+      "step": 440
+    },
+    {
+      "epoch": 0.9,
+      "learning_rate": 1.6384462329551217e-05,
+      "loss": 0.5615,
+      "step": 441
+    },
+    {
+      "epoch": 0.9,
+      "learning_rate": 1.6367418701710572e-05,
+      "loss": 0.5569,
+      "step": 442
+    },
+    {
+      "epoch": 0.91,
+      "learning_rate": 1.6350343907495195e-05,
+      "loss": 0.5712,
+      "step": 443
+    },
+    {
+      "epoch": 0.91,
+      "learning_rate": 1.6333238030480473e-05,
+      "loss": 0.5568,
+      "step": 444
+    },
+    {
+      "epoch": 0.91,
+      "learning_rate": 1.631610115439391e-05,
+      "loss": 0.532,
+      "step": 445
+    },
+    {
+      "epoch": 0.91,
+      "learning_rate": 1.629893336311477e-05,
+      "loss": 0.5412,
+      "step": 446
+    },
+    {
+      "epoch": 0.91,
+      "learning_rate": 1.6281734740673612e-05,
+      "loss": 0.5757,
+      "step": 447
+    },
+    {
+      "epoch": 0.92,
+      "learning_rate": 1.626450537125192e-05,
+      "loss": 0.5626,
+      "step": 448
+    },
+    {
+      "epoch": 0.92,
+      "learning_rate": 1.624724533918166e-05,
+      "loss": 0.4593,
+      "step": 449
+    },
+    {
+      "epoch": 0.92,
+      "learning_rate": 1.6229954728944895e-05,
+      "loss": 0.5205,
+      "step": 450
+    },
+    {
+      "epoch": 0.92,
+      "learning_rate": 1.6212633625173353e-05,
+      "loss": 0.5173,
+      "step": 451
+    },
+    {
+      "epoch": 0.93,
+      "learning_rate": 1.6195282112648007e-05,
+      "loss": 0.5298,
+      "step": 452
+    },
+    {
+      "epoch": 0.93,
+      "learning_rate": 1.617790027629869e-05,
+      "loss": 0.5513,
+      "step": 453
+    },
+    {
+      "epoch": 0.93,
+      "learning_rate": 1.6160488201203643e-05,
+      "loss": 0.5346,
+      "step": 454
+    },
+    {
+      "epoch": 0.93,
+      "learning_rate": 1.6143045972589123e-05,
+      "loss": 0.5459,
+      "step": 455
+    },
+    {
+      "epoch": 0.93,
+      "learning_rate": 1.6125573675828984e-05,
+      "loss": 0.5559,
+      "step": 456
+    },
+    {
+      "epoch": 0.94,
+      "learning_rate": 1.6108071396444237e-05,
+      "loss": 0.5823,
+      "step": 457
+    },
+    {
+      "epoch": 0.94,
+      "learning_rate": 1.6090539220102658e-05,
+      "loss": 0.575,
+      "step": 458
+    },
+    {
+      "epoch": 0.94,
+      "learning_rate": 1.607297723261837e-05,
+      "loss": 0.5139,
+      "step": 459
+    },
+    {
+      "epoch": 0.94,
+      "learning_rate": 1.605538551995139e-05,
+      "loss": 0.4767,
+      "step": 460
+    },
+    {
+      "epoch": 0.94,
+      "learning_rate": 1.603776416820724e-05,
+      "loss": 0.5164,
+      "step": 461
+    },
+    {
+      "epoch": 0.95,
+      "learning_rate": 1.6020113263636523e-05,
+      "loss": 0.5664,
+      "step": 462
+    },
+    {
+      "epoch": 0.95,
+      "learning_rate": 1.6002432892634476e-05,
+      "loss": 0.515,
+      "step": 463
+    },
+    {
+      "epoch": 0.95,
+      "learning_rate": 1.5984723141740578e-05,
+      "loss": 0.5396,
+      "step": 464
+    },
+    {
+      "epoch": 0.95,
+      "learning_rate": 1.5966984097638105e-05,
+      "loss": 0.5068,
+      "step": 465
+    },
+    {
+      "epoch": 0.95,
+      "learning_rate": 1.5949215847153718e-05,
+      "loss": 0.523,
+      "step": 466
+    },
+    {
+      "epoch": 0.96,
+      "learning_rate": 1.5931418477257035e-05,
+      "loss": 0.5126,
+      "step": 467
+    },
+    {
+      "epoch": 0.96,
+      "learning_rate": 1.59135920750602e-05,
+      "loss": 0.5549,
+      "step": 468
+    },
+    {
+      "epoch": 0.96,
+      "learning_rate": 1.5895736727817457e-05,
+      "loss": 0.515,
+      "step": 469
+    },
+    {
+      "epoch": 0.96,
+      "learning_rate": 1.5877852522924733e-05,
+      "loss": 0.5138,
+      "step": 470
+    },
+    {
+      "epoch": 0.96,
+      "learning_rate": 1.58599395479192e-05,
+      "loss": 0.5509,
+      "step": 471
+    },
+    {
+      "epoch": 0.97,
+      "learning_rate": 1.5841997890478853e-05,
+      "loss": 0.5435,
+      "step": 472
+    },
+    {
+      "epoch": 0.97,
+      "learning_rate": 1.5824027638422075e-05,
+      "loss": 0.5667,
+      "step": 473
+    },
+    {
+      "epoch": 0.97,
+      "learning_rate": 1.580602887970721e-05,
+      "loss": 0.5233,
+      "step": 474
+    },
+    {
+      "epoch": 0.97,
+      "learning_rate": 1.5788001702432132e-05,
+      "loss": 0.5261,
+      "step": 475
+    },
+    {
+      "epoch": 0.97,
+      "learning_rate": 1.5769946194833816e-05,
+      "loss": 0.5563,
+      "step": 476
+    },
+    {
+      "epoch": 0.98,
+      "learning_rate": 1.5751862445287907e-05,
+      "loss": 0.5314,
+      "step": 477
+    },
+    {
+      "epoch": 0.98,
+      "learning_rate": 1.573375054230828e-05,
+      "loss": 0.5157,
+      "step": 478
+    },
+    {
+      "epoch": 0.98,
+      "learning_rate": 1.5715610574546612e-05,
+      "loss": 0.5315,
+      "step": 479
+    },
+    {
+      "epoch": 0.98,
+      "learning_rate": 1.569744263079195e-05,
+      "loss": 0.5759,
+      "step": 480
+    },
+    {
+      "epoch": 0.98,
+      "learning_rate": 1.567924679997028e-05,
+      "loss": 0.564,
+      "step": 481
+    },
+    {
+      "epoch": 0.99,
+      "learning_rate": 1.566102317114406e-05,
+      "loss": 0.6112,
+      "step": 482
+    },
+    {
+      "epoch": 0.99,
+      "learning_rate": 1.5642771833511855e-05,
+      "loss": 0.5539,
+      "step": 483
+    },
+    {
+      "epoch": 0.99,
+      "learning_rate": 1.562449287640781e-05,
+      "loss": 0.5206,
+      "step": 484
+    },
+    {
+      "epoch": 0.99,
+      "learning_rate": 1.5606186389301284e-05,
+      "loss": 0.4999,
+      "step": 485
+    },
+    {
+      "epoch": 0.99,
+      "learning_rate": 1.5587852461796375e-05,
+      "loss": 0.5569,
+      "step": 486
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 1.5569491183631504e-05,
+      "loss": 0.5436,
+      "step": 487
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 1.555110264467895e-05,
+      "loss": 0.5182,
+      "step": 488
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 1.553268693494444e-05,
+      "loss": 0.4257,
+      "step": 489
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 1.5514244144566676e-05,
+      "loss": 0.3766,
+      "step": 490
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 1.5495774363816928e-05,
+      "loss": 0.3515,
+      "step": 491
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 1.5477277683098555e-05,
+      "loss": 0.3663,
+      "step": 492
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 1.54587541929466e-05,
+      "loss": 0.3483,
+      "step": 493
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 1.5440203984027323e-05,
+      "loss": 0.3346,
+      "step": 494
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 1.542162714713776e-05,
+      "loss": 0.3432,
+      "step": 495
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 1.5403023773205286e-05,
+      "loss": 0.3776,
+      "step": 496
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 1.5384393953287165e-05,
+      "loss": 0.3569,
+      "step": 497
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 1.5365737778570102e-05,
+      "loss": 0.3141,
+      "step": 498
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 1.5347055340369806e-05,
+      "loss": 0.3333,
+      "step": 499
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 1.532834673013053e-05,
+      "loss": 0.3472,
+      "step": 500
+    },
+    {
+      "epoch": 1.03,
+      "learning_rate": 1.530961203942464e-05,
+      "loss": 0.3899,
+      "step": 501
+    },
+    {
+      "epoch": 1.03,
+      "learning_rate": 1.5290851359952145e-05,
+      "loss": 0.3458,
+      "step": 502
+    },
+    {
+      "epoch": 1.03,
+      "learning_rate": 1.527206478354027e-05,
+      "loss": 0.3112,
+      "step": 503
+    },
+    {
+      "epoch": 1.03,
+      "learning_rate": 1.5253252402142989e-05,
+      "loss": 0.3338,
+      "step": 504
+    },
+    {
+      "epoch": 1.03,
+      "learning_rate": 1.5234414307840592e-05,
+      "loss": 0.3522,
+      "step": 505
+    },
+    {
+      "epoch": 1.04,
+      "learning_rate": 1.5215550592839219e-05,
+      "loss": 0.3615,
+      "step": 506
+    },
+    {
+      "epoch": 1.04,
+      "learning_rate": 1.5196661349470412e-05,
+      "loss": 0.338,
+      "step": 507
+    },
+    {
+      "epoch": 1.04,
+      "learning_rate": 1.5177746670190674e-05,
+      "loss": 0.3354,
+      "step": 508
+    },
+    {
+      "epoch": 1.04,
+      "learning_rate": 1.5158806647581e-05,
+      "loss": 0.3792,
+      "step": 509
+    },
+    {
+      "epoch": 1.04,
+      "learning_rate": 1.513984137434644e-05,
+      "loss": 0.37,
+      "step": 510
+    },
+    {
+      "epoch": 1.05,
+      "learning_rate": 1.5120850943315627e-05,
+      "loss": 0.3569,
+      "step": 511
+    },
+    {
+      "epoch": 1.05,
+      "learning_rate": 1.5101835447440346e-05,
+      "loss": 0.3368,
+      "step": 512
+    },
+    {
+      "epoch": 1.05,
+      "learning_rate": 1.5082794979795053e-05,
+      "loss": 0.3371,
+      "step": 513
+    },
+    {
+      "epoch": 1.05,
+      "learning_rate": 1.506372963357644e-05,
+      "loss": 0.3409,
+      "step": 514
+    },
+    {
+      "epoch": 1.05,
+      "learning_rate": 1.504463950210297e-05,
+      "loss": 0.3733,
+      "step": 515
+    },
+    {
+      "epoch": 1.06,
+      "learning_rate": 1.5025524678814428e-05,
+      "loss": 0.3653,
+      "step": 516
+    },
+    {
+      "epoch": 1.06,
+      "learning_rate": 1.5006385257271438e-05,
+      "loss": 0.2992,
+      "step": 517
+    },
+    {
+      "epoch": 1.06,
+      "learning_rate": 1.4987221331155044e-05,
+      "loss": 0.3438,
+      "step": 518
+    },
+    {
+      "epoch": 1.06,
+      "learning_rate": 1.4968032994266224e-05,
+      "loss": 0.3442,
+      "step": 519
+    },
+    {
+      "epoch": 1.06,
+      "learning_rate": 1.4948820340525439e-05,
+      "loss": 0.3223,
+      "step": 520
+    },
+    {
+      "epoch": 1.07,
+      "learning_rate": 1.4929583463972169e-05,
+      "loss": 0.3353,
+      "step": 521
+    },
+    {
+      "epoch": 1.07,
+      "learning_rate": 1.4910322458764461e-05,
+      "loss": 0.3557,
+      "step": 522
+    },
+    {
+      "epoch": 1.07,
+      "learning_rate": 1.4891037419178465e-05,
+      "loss": 0.3542,
+      "step": 523
+    },
+    {
+      "epoch": 1.07,
+      "learning_rate": 1.4871728439607967e-05,
+      "loss": 0.3496,
+      "step": 524
+    },
+    {
+      "epoch": 1.07,
+      "learning_rate": 1.4852395614563935e-05,
+      "loss": 0.4041,
+      "step": 525
+    },
+    {
+      "epoch": 1.08,
+      "learning_rate": 1.4833039038674049e-05,
+      "loss": 0.3431,
+      "step": 526
+    },
+    {
+      "epoch": 1.08,
+      "learning_rate": 1.4813658806682244e-05,
+      "loss": 0.3058,
+      "step": 527
+    },
+    {
+      "epoch": 1.08,
+      "learning_rate": 1.4794255013448242e-05,
+      "loss": 0.341,
+      "step": 528
+    },
+    {
+      "epoch": 1.08,
+      "learning_rate": 1.4774827753947088e-05,
+      "loss": 0.3834,
+      "step": 529
+    },
+    {
+      "epoch": 1.08,
+      "learning_rate": 1.4755377123268693e-05,
+      "loss": 0.3557,
+      "step": 530
+    },
+    {
+      "epoch": 1.09,
+      "learning_rate": 1.4735903216617357e-05,
+      "loss": 0.3165,
+      "step": 531
+    },
+    {
+      "epoch": 1.09,
+      "learning_rate": 1.4716406129311307e-05,
+      "loss": 0.3183,
+      "step": 532
+    },
+    {
+      "epoch": 1.09,
+      "learning_rate": 1.4696885956782233e-05,
+      "loss": 0.3391,
+      "step": 533
+    },
+    {
+      "epoch": 1.09,
+      "learning_rate": 1.4677342794574819e-05,
+      "loss": 0.3375,
+      "step": 534
+    },
+    {
+      "epoch": 1.09,
+      "learning_rate": 1.4657776738346279e-05,
+      "loss": 0.3321,
+      "step": 535
+    },
+    {
+      "epoch": 1.1,
+      "learning_rate": 1.4638187883865881e-05,
+      "loss": 0.3438,
+      "step": 536
+    },
+    {
+      "epoch": 1.1,
+      "learning_rate": 1.4618576327014486e-05,
+      "loss": 0.3431,
+      "step": 537
+    },
+    {
+      "epoch": 1.1,
+      "learning_rate": 1.4598942163784073e-05,
+      "loss": 0.3749,
+      "step": 538
+    },
+    {
+      "epoch": 1.1,
+      "learning_rate": 1.4579285490277275e-05,
+      "loss": 0.3647,
+      "step": 539
+    },
+    {
+      "epoch": 1.11,
+      "learning_rate": 1.45596064027069e-05,
+      "loss": 0.3772,
+      "step": 540
+    },
+    {
+      "epoch": 1.11,
+      "learning_rate": 1.4539904997395468e-05,
+      "loss": 0.3686,
+      "step": 541
+    },
+    {
+      "epoch": 1.11,
+      "learning_rate": 1.4520181370774742e-05,
+      "loss": 0.3605,
+      "step": 542
+    },
+    {
+      "epoch": 1.11,
+      "learning_rate": 1.4500435619385243e-05,
+      "loss": 0.3245,
+      "step": 543
+    },
+    {
+      "epoch": 1.11,
+      "learning_rate": 1.4480667839875786e-05,
+      "loss": 0.2829,
+      "step": 544
+    },
+    {
+      "epoch": 1.12,
+      "learning_rate": 1.4460878129003013e-05,
+      "loss": 0.3769,
+      "step": 545
+    },
+    {
+      "epoch": 1.12,
+      "learning_rate": 1.4441066583630905e-05,
+      "loss": 0.3177,
+      "step": 546
+    },
+    {
+      "epoch": 1.12,
+      "learning_rate": 1.4421233300730317e-05,
+      "loss": 0.3396,
+      "step": 547
+    },
+    {
+      "epoch": 1.12,
+      "learning_rate": 1.4401378377378512e-05,
+      "loss": 0.3668,
+      "step": 548
+    },
+    {
+      "epoch": 1.12,
+      "learning_rate": 1.4381501910758662e-05,
+      "loss": 0.2987,
+      "step": 549
+    },
+    {
+      "epoch": 1.13,
+      "learning_rate": 1.4361603998159388e-05,
+      "loss": 0.2946,
+      "step": 550
+    },
+    {
+      "epoch": 1.13,
+      "learning_rate": 1.434168473697429e-05,
+      "loss": 0.3625,
+      "step": 551
+    },
+    {
+      "epoch": 1.13,
+      "learning_rate": 1.432174422470146e-05,
+      "loss": 0.3463,
+      "step": 552
+    },
+    {
+      "epoch": 1.13,
+      "learning_rate": 1.4301782558942998e-05,
+      "loss": 0.3516,
+      "step": 553
+    },
+    {
+      "epoch": 1.13,
+      "learning_rate": 1.4281799837404553e-05,
+      "loss": 0.3743,
+      "step": 554
+    },
+    {
+      "epoch": 1.14,
+      "learning_rate": 1.4261796157894831e-05,
+      "loss": 0.3097,
+      "step": 555
+    },
+    {
+      "epoch": 1.14,
+      "learning_rate": 1.4241771618325123e-05,
+      "loss": 0.3439,
+      "step": 556
+    },
+    {
+      "epoch": 1.14,
+      "learning_rate": 1.4221726316708815e-05,
+      "loss": 0.3115,
+      "step": 557
+    },
+    {
+      "epoch": 1.14,
+      "learning_rate": 1.420166035116093e-05,
+      "loss": 0.405,
+      "step": 558
+    },
+    {
+      "epoch": 1.14,
+      "learning_rate": 1.4181573819897618e-05,
+      "loss": 0.306,
+      "step": 559
+    },
+    {
+      "epoch": 1.15,
+      "learning_rate": 1.4161466821235705e-05,
+      "loss": 0.3366,
+      "step": 560
+    },
+    {
+      "epoch": 1.15,
+      "learning_rate": 1.414133945359219e-05,
+      "loss": 0.3233,
+      "step": 561
+    },
+    {
+      "epoch": 1.15,
+      "learning_rate": 1.4121191815483777e-05,
+      "loss": 0.3669,
+      "step": 562
+    },
+    {
+      "epoch": 1.15,
+      "learning_rate": 1.4101024005526378e-05,
+      "loss": 0.3583,
+      "step": 563
+    },
+    {
+      "epoch": 1.15,
+      "learning_rate": 1.408083612243465e-05,
+      "loss": 0.3315,
+      "step": 564
+    },
+    {
+      "epoch": 1.16,
+      "learning_rate": 1.4060628265021494e-05,
+      "loss": 0.3605,
+      "step": 565
+    },
+    {
+      "epoch": 1.16,
+      "learning_rate": 1.4040400532197582e-05,
+      "loss": 0.2843,
+      "step": 566
+    },
+    {
+      "epoch": 1.16,
+      "learning_rate": 1.4020153022970868e-05,
+      "loss": 0.3142,
+      "step": 567
+    },
+    {
+      "epoch": 1.16,
+      "learning_rate": 1.3999885836446106e-05,
+      "loss": 0.348,
+      "step": 568
+    },
+    {
+      "epoch": 1.16,
+      "learning_rate": 1.3979599071824363e-05,
+      "loss": 0.3656,
+      "step": 569
+    },
+    {
+      "epoch": 1.17,
+      "learning_rate": 1.3959292828402534e-05,
+      "loss": 0.3027,
+      "step": 570
+    },
+    {
+      "epoch": 1.17,
+      "learning_rate": 1.3938967205572858e-05,
+      "loss": 0.2979,
+      "step": 571
+    },
+    {
+      "epoch": 1.17,
+      "learning_rate": 1.3918622302822425e-05,
+      "loss": 0.395,
+      "step": 572
+    },
+    {
+      "epoch": 1.17,
+      "learning_rate": 1.3898258219732698e-05,
+      "loss": 0.3468,
+      "step": 573
+    },
+    {
+      "epoch": 1.17,
+      "learning_rate": 1.3877875055979021e-05,
+      "loss": 0.3753,
+      "step": 574
+    },
+    {
+      "epoch": 1.18,
+      "learning_rate": 1.3857472911330131e-05,
+      "loss": 0.3558,
+      "step": 575
+    },
+    {
+      "epoch": 1.18,
+      "learning_rate": 1.3837051885647673e-05,
+      "loss": 0.3528,
+      "step": 576
+    },
+    {
+      "epoch": 1.18,
+      "learning_rate": 1.3816612078885698e-05,
+      "loss": 0.3321,
+      "step": 577
+    },
+    {
+      "epoch": 1.18,
+      "learning_rate": 1.3796153591090195e-05,
+      "loss": 0.3338,
+      "step": 578
+    },
+    {
+      "epoch": 1.18,
+      "learning_rate": 1.3775676522398587e-05,
+      "loss": 0.3063,
+      "step": 579
+    },
+    {
+      "epoch": 1.19,
+      "learning_rate": 1.3755180973039241e-05,
+      "loss": 0.2991,
+      "step": 580
+    },
+    {
+      "epoch": 1.19,
+      "learning_rate": 1.3734667043330989e-05,
+      "loss": 0.3341,
+      "step": 581
+    },
+    {
+      "epoch": 1.19,
+      "learning_rate": 1.3714134833682618e-05,
+      "loss": 0.3317,
+      "step": 582
+    },
+    {
+      "epoch": 1.19,
+      "learning_rate": 1.3693584444592397e-05,
+      "loss": 0.3128,
+      "step": 583
+    },
+    {
+      "epoch": 1.2,
+      "learning_rate": 1.367301597664757e-05,
+      "loss": 0.4056,
+      "step": 584
+    },
+    {
+      "epoch": 1.2,
+      "learning_rate": 1.3652429530523882e-05,
+      "loss": 0.3349,
+      "step": 585
+    },
+    {
+      "epoch": 1.2,
+      "learning_rate": 1.3631825206985063e-05,
+      "loss": 0.3484,
+      "step": 586
+    },
+    {
+      "epoch": 1.2,
+      "learning_rate": 1.3611203106882354e-05,
+      "loss": 0.3325,
+      "step": 587
+    },
+    {
+      "epoch": 1.2,
+      "learning_rate": 1.3590563331154008e-05,
+      "loss": 0.3837,
+      "step": 588
+    },
+    {
+      "epoch": 1.21,
+      "learning_rate": 1.3569905980824789e-05,
+      "loss": 0.3464,
+      "step": 589
+    },
+    {
+      "epoch": 1.21,
+      "learning_rate": 1.3549231157005482e-05,
+      "loss": 0.325,
+      "step": 590
+    },
+    {
+      "epoch": 1.21,
+      "learning_rate": 1.3528538960892405e-05,
+      "loss": 0.3354,
+      "step": 591
+    },
+    {
+      "epoch": 1.21,
+      "learning_rate": 1.3507829493766903e-05,
+      "loss": 0.3032,
+      "step": 592
+    },
+    {
+      "epoch": 1.21,
+      "learning_rate": 1.348710285699486e-05,
+      "loss": 0.2975,
+      "step": 593
+    },
+    {
+      "epoch": 1.22,
+      "learning_rate": 1.3466359152026197e-05,
+      "loss": 0.3204,
+      "step": 594
+    },
+    {
+      "epoch": 1.22,
+      "learning_rate": 1.3445598480394375e-05,
+      "loss": 0.3701,
+      "step": 595
+    },
+    {
+      "epoch": 1.22,
+      "learning_rate": 1.342482094371591e-05,
+      "loss": 0.3717,
+      "step": 596
+    },
+    {
+      "epoch": 1.22,
+      "learning_rate": 1.3404026643689858e-05,
+      "loss": 0.3556,
+      "step": 597
+    },
+    {
+      "epoch": 1.22,
+      "learning_rate": 1.338321568209733e-05,
+      "loss": 0.3537,
+      "step": 598
+    },
+    {
+      "epoch": 1.23,
+      "learning_rate": 1.3362388160800992e-05,
+      "loss": 0.2958,
+      "step": 599
+    },
+    {
+      "epoch": 1.23,
+      "learning_rate": 1.3341544181744559e-05,
+      "loss": 0.3392,
+      "step": 600
+    },
+    {
+      "epoch": 1.23,
+      "learning_rate": 1.3320683846952304e-05,
+      "loss": 0.3263,
+      "step": 601
+    },
+    {
+      "epoch": 1.23,
+      "learning_rate": 1.3299807258528557e-05,
+      "loss": 0.3154,
+      "step": 602
+    },
+    {
+      "epoch": 1.23,
+      "learning_rate": 1.3278914518657202e-05,
+      "loss": 0.349,
+      "step": 603
+    },
+    {
+      "epoch": 1.24,
+      "learning_rate": 1.3258005729601178e-05,
+      "loss": 0.272,
+      "step": 604
+    },
+    {
+      "epoch": 1.24,
+      "learning_rate": 1.3237080993701983e-05,
+      "loss": 0.3664,
+      "step": 605
+    },
+    {
+      "epoch": 1.24,
+      "learning_rate": 1.3216140413379166e-05,
+      "loss": 0.2923,
+      "step": 606
+    },
+    {
+      "epoch": 1.24,
+      "learning_rate": 1.3195184091129828e-05,
+      "loss": 0.2998,
+      "step": 607
+    },
+    {
+      "epoch": 1.24,
+      "learning_rate": 1.317421212952813e-05,
+      "loss": 0.3417,
+      "step": 608
+    },
+    {
+      "epoch": 1.25,
+      "learning_rate": 1.3153224631224772e-05,
+      "loss": 0.342,
+      "step": 609
+    },
+    {
+      "epoch": 1.25,
+      "learning_rate": 1.3132221698946508e-05,
+      "loss": 0.3199,
+      "step": 610
+    },
+    {
+      "epoch": 1.25,
+      "learning_rate": 1.311120343549563e-05,
+      "loss": 0.3353,
+      "step": 611
+    },
+    {
+      "epoch": 1.25,
+      "learning_rate": 1.3090169943749475e-05,
+      "loss": 0.3335,
+      "step": 612
+    },
+    {
+      "epoch": 1.25,
+      "learning_rate": 1.3069121326659919e-05,
+      "loss": 0.3024,
+      "step": 613
+    },
+    {
+      "epoch": 1.26,
+      "learning_rate": 1.3048057687252866e-05,
+      "loss": 0.3583,
+      "step": 614
+    },
+    {
+      "epoch": 1.26,
+      "learning_rate": 1.3026979128627757e-05,
+      "loss": 0.3051,
+      "step": 615
+    },
+    {
+      "epoch": 1.26,
+      "learning_rate": 1.3005885753957048e-05,
+      "loss": 0.3496,
+      "step": 616
+    },
+    {
+      "epoch": 1.26,
+      "learning_rate": 1.2984777666485722e-05,
+      "loss": 0.3295,
+      "step": 617
+    },
+    {
+      "epoch": 1.26,
+      "learning_rate": 1.2963654969530773e-05,
+      "loss": 0.2658,
+      "step": 618
+    },
+    {
+      "epoch": 1.27,
+      "learning_rate": 1.2942517766480702e-05,
+      "loss": 0.3044,
+      "step": 619
+    },
+    {
+      "epoch": 1.27,
+      "learning_rate": 1.2921366160795017e-05,
+      "loss": 0.3394,
+      "step": 620
+    },
+    {
+      "epoch": 1.27,
+      "learning_rate": 1.2900200256003722e-05,
+      "loss": 0.3473,
+      "step": 621
+    },
+    {
+      "epoch": 1.27,
+      "learning_rate": 1.2879020155706803e-05,
+      "loss": 0.2876,
+      "step": 622
+    },
+    {
+      "epoch": 1.28,
+      "learning_rate": 1.2857825963573736e-05,
+      "loss": 0.3297,
+      "step": 623
+    },
+    {
+      "epoch": 1.28,
+      "learning_rate": 1.2836617783342968e-05,
+      "loss": 0.2879,
+      "step": 624
+    },
+    {
+      "epoch": 1.28,
+      "learning_rate": 1.2815395718821415e-05,
+      "loss": 0.3296,
+      "step": 625
+    },
+    {
+      "epoch": 1.28,
+      "learning_rate": 1.279415987388395e-05,
+      "loss": 0.3378,
+      "step": 626
+    },
+    {
+      "epoch": 1.28,
+      "learning_rate": 1.27729103524729e-05,
+      "loss": 0.3366,
+      "step": 627
+    },
+    {
+      "epoch": 1.29,
+      "learning_rate": 1.2751647258597531e-05,
+      "loss": 0.3283,
+      "step": 628
+    },
+    {
+      "epoch": 1.29,
+      "learning_rate": 1.2730370696333538e-05,
+      "loss": 0.3257,
+      "step": 629
+    },
+    {
+      "epoch": 1.29,
+      "learning_rate": 1.2709080769822549e-05,
+      "loss": 0.3453,
+      "step": 630
+    },
+    {
+      "epoch": 1.29,
+      "learning_rate": 1.2687777583271595e-05,
+      "loss": 0.3763,
+      "step": 631
+    },
+    {
+      "epoch": 1.29,
+      "learning_rate": 1.2666461240952615e-05,
+      "loss": 0.3242,
+      "step": 632
+    },
+    {
+      "epoch": 1.3,
+      "learning_rate": 1.264513184720194e-05,
+      "loss": 0.3473,
+      "step": 633
+    },
+    {
+      "epoch": 1.3,
+      "learning_rate": 1.2623789506419792e-05,
+      "loss": 0.3616,
+      "step": 634
+    },
+    {
+      "epoch": 1.3,
+      "learning_rate": 1.260243432306975e-05,
+      "loss": 0.3552,
+      "step": 635
+    },
+    {
+      "epoch": 1.3,
+      "learning_rate": 1.2581066401678261e-05,
+      "loss": 0.3226,
+      "step": 636
+    },
+    {
+      "epoch": 1.3,
+      "learning_rate": 1.2559685846834122e-05,
+      "loss": 0.3472,
+      "step": 637
+    },
+    {
+      "epoch": 1.31,
+      "learning_rate": 1.253829276318796e-05,
+      "loss": 0.3617,
+      "step": 638
+    },
+    {
+      "epoch": 1.31,
+      "learning_rate": 1.2516887255451735e-05,
+      "loss": 0.3793,
+      "step": 639
+    },
+    {
+      "epoch": 1.31,
+      "learning_rate": 1.249546942839821e-05,
+      "loss": 0.3457,
+      "step": 640
+    },
+    {
+      "epoch": 1.31,
+      "learning_rate": 1.2474039386860453e-05,
+      "loss": 0.2998,
+      "step": 641
+    },
+    {
+      "epoch": 1.31,
+      "learning_rate": 1.2452597235731312e-05,
+      "loss": 0.3611,
+      "step": 642
+    },
+    {
+      "epoch": 1.32,
+      "learning_rate": 1.2431143079962917e-05,
+      "loss": 0.3333,
+      "step": 643
+    },
+    {
+      "epoch": 1.32,
+      "learning_rate": 1.2409677024566145e-05,
+      "loss": 0.3457,
+      "step": 644
+    },
+    {
+      "epoch": 1.32,
+      "learning_rate": 1.2388199174610126e-05,
+      "loss": 0.3519,
+      "step": 645
+    },
+    {
+      "epoch": 1.32,
+      "learning_rate": 1.236670963522172e-05,
+      "loss": 0.351,
+      "step": 646
+    },
+    {
+      "epoch": 1.32,
+      "learning_rate": 1.2345208511584993e-05,
+      "loss": 0.3347,
+      "step": 647
+    },
+    {
+      "epoch": 1.33,
+      "learning_rate": 1.2323695908940729e-05,
+      "loss": 0.2927,
+      "step": 648
+    },
+    {
+      "epoch": 1.33,
+      "learning_rate": 1.2302171932585885e-05,
+      "loss": 0.2979,
+      "step": 649
+    },
+    {
+      "epoch": 1.33,
+      "learning_rate": 1.2280636687873089e-05,
+      "loss": 0.3536,
+      "step": 650
+    },
+    {
+      "epoch": 1.33,
+      "learning_rate": 1.2259090280210129e-05,
+      "loss": 0.3331,
+      "step": 651
+    },
+    {
+      "epoch": 1.33,
+      "learning_rate": 1.2237532815059427e-05,
+      "loss": 0.3271,
+      "step": 652
+    },
+    {
+      "epoch": 1.34,
+      "learning_rate": 1.2215964397937537e-05,
+      "loss": 0.3449,
+      "step": 653
+    },
+    {
+      "epoch": 1.34,
+      "learning_rate": 1.2194385134414608e-05,
+      "loss": 0.3327,
+      "step": 654
+    },
+    {
+      "epoch": 1.34,
+      "learning_rate": 1.2172795130113884e-05,
+      "loss": 0.3044,
+      "step": 655
+    },
+    {
+      "epoch": 1.34,
+      "learning_rate": 1.2151194490711177e-05,
+      "loss": 0.3247,
+      "step": 656
+    },
+    {
+      "epoch": 1.34,
+      "learning_rate": 1.2129583321934364e-05,
+      "loss": 0.2612,
+      "step": 657
+    },
+    {
+      "epoch": 1.35,
+      "learning_rate": 1.2107961729562852e-05,
+      "loss": 0.3643,
+      "step": 658
+    },
+    {
+      "epoch": 1.35,
+      "learning_rate": 1.2086329819427065e-05,
+      "loss": 0.3353,
+      "step": 659
+    },
+    {
+      "epoch": 1.35,
+      "learning_rate": 1.2064687697407939e-05,
+      "loss": 0.3266,
+      "step": 660
+    },
+    {
+      "epoch": 1.35,
+      "learning_rate": 1.2043035469436387e-05,
+      "loss": 0.3187,
+      "step": 661
+    },
+    {
+      "epoch": 1.35,
+      "learning_rate": 1.2021373241492786e-05,
+      "loss": 0.3592,
+      "step": 662
+    },
+    {
+      "epoch": 1.36,
+      "learning_rate": 1.1999701119606464e-05,
+      "loss": 0.3128,
+      "step": 663
+    },
+    {
+      "epoch": 1.36,
+      "learning_rate": 1.1978019209855174e-05,
+      "loss": 0.3352,
+      "step": 664
+    },
+    {
+      "epoch": 1.36,
+      "learning_rate": 1.195632761836458e-05,
+      "loss": 0.3119,
+      "step": 665
+    },
+    {
+      "epoch": 1.36,
+      "learning_rate": 1.1934626451307727e-05,
+      "loss": 0.362,
+      "step": 666
+    },
+    {
+      "epoch": 1.37,
+      "learning_rate": 1.1912915814904542e-05,
+      "loss": 0.3756,
+      "step": 667
+    },
+    {
+      "epoch": 1.37,
+      "learning_rate": 1.1891195815421293e-05,
+      "loss": 0.3075,
+      "step": 668
+    },
+    {
+      "epoch": 1.37,
+      "learning_rate": 1.1869466559170073e-05,
+      "loss": 0.2956,
+      "step": 669
+    },
+    {
+      "epoch": 1.37,
+      "learning_rate": 1.1847728152508294e-05,
+      "loss": 0.3339,
+      "step": 670
+    },
+    {
+      "epoch": 1.37,
+      "learning_rate": 1.182598070183815e-05,
+      "loss": 0.3061,
+      "step": 671
+    },
+    {
+      "epoch": 1.38,
+      "learning_rate": 1.1804224313606104e-05,
+      "loss": 0.342,
+      "step": 672
+    },
+    {
+      "epoch": 1.38,
+      "learning_rate": 1.1782459094302362e-05,
+      "loss": 0.3183,
+      "step": 673
+    },
+    {
+      "epoch": 1.38,
+      "learning_rate": 1.1760685150460363e-05,
+      "loss": 0.2949,
+      "step": 674
+    },
+    {
+      "epoch": 1.38,
+      "learning_rate": 1.1738902588656242e-05,
+      "loss": 0.3344,
+      "step": 675
+    },
+    {
+      "epoch": 1.38,
+      "learning_rate": 1.171711151550832e-05,
+      "loss": 0.3551,
+      "step": 676
+    },
+    {
+      "epoch": 1.39,
+      "learning_rate": 1.1695312037676574e-05,
+      "loss": 0.3301,
+      "step": 677
+    },
+    {
+      "epoch": 1.39,
+      "learning_rate": 1.1673504261862125e-05,
+      "loss": 0.3514,
+      "step": 678
+    },
+    {
+      "epoch": 1.39,
+      "learning_rate": 1.1651688294806706e-05,
+      "loss": 0.2991,
+      "step": 679
+    },
+    {
+      "epoch": 1.39,
+      "learning_rate": 1.1629864243292149e-05,
+      "loss": 0.3646,
+      "step": 680
+    },
+    {
+      "epoch": 1.39,
+      "learning_rate": 1.1608032214139843e-05,
+      "loss": 0.2693,
+      "step": 681
+    },
+    {
+      "epoch": 1.4,
+      "learning_rate": 1.158619231421024e-05,
+      "loss": 0.3498,
+      "step": 682
+    },
+    {
+      "epoch": 1.4,
+      "learning_rate": 1.156434465040231e-05,
+      "loss": 0.3022,
+      "step": 683
+    },
+    {
+      "epoch": 1.4,
+      "learning_rate": 1.1542489329653024e-05,
+      "loss": 0.3682,
+      "step": 684
+    },
+    {
+      "epoch": 1.4,
+      "learning_rate": 1.1520626458936834e-05,
+      "loss": 0.2932,
+      "step": 685
+    },
+    {
+      "epoch": 1.4,
+      "learning_rate": 1.1498756145265145e-05,
+      "loss": 0.3045,
+      "step": 686
+    },
+    {
+      "epoch": 1.41,
+      "learning_rate": 1.147687849568579e-05,
+      "loss": 0.3469,
+      "step": 687
+    },
+    {
+      "epoch": 1.41,
+      "learning_rate": 1.1454993617282513e-05,
+      "loss": 0.3251,
+      "step": 688
+    },
+    {
+      "epoch": 1.41,
+      "learning_rate": 1.143310161717444e-05,
+      "loss": 0.3271,
+      "step": 689
+    },
+    {
+      "epoch": 1.41,
+      "learning_rate": 1.1411202602515556e-05,
+      "loss": 0.3398,
+      "step": 690
+    },
+    {
+      "epoch": 1.41,
+      "learning_rate": 1.1389296680494174e-05,
+      "loss": 0.2821,
+      "step": 691
+    },
+    {
+      "epoch": 1.42,
+      "learning_rate": 1.1367383958332427e-05,
+      "loss": 0.3378,
+      "step": 692
+    },
+    {
+      "epoch": 1.42,
+      "learning_rate": 1.1345464543285721e-05,
+      "loss": 0.3412,
+      "step": 693
+    },
+    {
+      "epoch": 1.42,
+      "learning_rate": 1.1323538542642227e-05,
+      "loss": 0.3316,
+      "step": 694
+    },
+    {
+      "epoch": 1.42,
+      "learning_rate": 1.130160606372235e-05,
+      "loss": 0.3495,
+      "step": 695
+    },
+    {
+      "epoch": 1.42,
+      "learning_rate": 1.1279667213878206e-05,
+      "loss": 0.3575,
+      "step": 696
+    },
+    {
+      "epoch": 1.43,
+      "learning_rate": 1.1257722100493089e-05,
+      "loss": 0.3264,
+      "step": 697
+    },
+    {
+      "epoch": 1.43,
+      "learning_rate": 1.123577083098096e-05,
+      "loss": 0.3527,
+      "step": 698
+    },
+    {
+      "epoch": 1.43,
+      "learning_rate": 1.1213813512785897e-05,
+      "loss": 0.3033,
+      "step": 699
+    },
+    {
+      "epoch": 1.43,
+      "learning_rate": 1.1191850253381602e-05,
+      "loss": 0.3448,
+      "step": 700
+    },
+    {
+      "epoch": 1.43,
+      "learning_rate": 1.1169881160270845e-05,
+      "loss": 0.3502,
+      "step": 701
+    },
+    {
+      "epoch": 1.44,
+      "learning_rate": 1.1147906340984955e-05,
+      "loss": 0.3656,
+      "step": 702
+    },
+    {
+      "epoch": 1.44,
+      "learning_rate": 1.1125925903083287e-05,
+      "loss": 0.3439,
+      "step": 703
+    },
+    {
+      "epoch": 1.44,
+      "learning_rate": 1.11039399541527e-05,
+      "loss": 0.2745,
+      "step": 704
+    },
+    {
+      "epoch": 1.44,
+      "learning_rate": 1.1081948601807022e-05,
+      "loss": 0.334,
+      "step": 705
+    },
+    {
+      "epoch": 1.44,
+      "learning_rate": 1.1059951953686537e-05,
+      "loss": 0.323,
+      "step": 706
+    },
+    {
+      "epoch": 1.45,
+      "learning_rate": 1.1037950117457437e-05,
+      "loss": 0.3513,
+      "step": 707
+    },
+    {
+      "epoch": 1.45,
+      "learning_rate": 1.1015943200811325e-05,
+      "loss": 0.3224,
+      "step": 708
+    },
+    {
+      "epoch": 1.45,
+      "learning_rate": 1.099393131146466e-05,
+      "loss": 0.2862,
+      "step": 709
+    },
+    {
+      "epoch": 1.45,
+      "learning_rate": 1.0971914557158243e-05,
+      "loss": 0.3239,
+      "step": 710
+    },
+    {
+      "epoch": 1.46,
+      "learning_rate": 1.0949893045656688e-05,
+      "loss": 0.3587,
+      "step": 711
+    },
+    {
+      "epoch": 1.46,
+      "learning_rate": 1.092786688474789e-05,
+      "loss": 0.306,
+      "step": 712
+    },
+    {
+      "epoch": 1.46,
+      "learning_rate": 1.0905836182242508e-05,
+      "loss": 0.3137,
+      "step": 713
+    },
+    {
+      "epoch": 1.46,
+      "learning_rate": 1.0883801045973425e-05,
+      "loss": 0.3412,
+      "step": 714
+    },
+    {
+      "epoch": 1.46,
+      "learning_rate": 1.0861761583795231e-05,
+      "loss": 0.2886,
+      "step": 715
+    },
+    {
+      "epoch": 1.47,
+      "learning_rate": 1.0839717903583684e-05,
+      "loss": 0.3354,
+      "step": 716
+    },
+    {
+      "epoch": 1.47,
+      "learning_rate": 1.0817670113235192e-05,
+      "loss": 0.3324,
+      "step": 717
+    },
+    {
+      "epoch": 1.47,
+      "learning_rate": 1.0795618320666282e-05,
+      "loss": 0.289,
+      "step": 718
+    },
+    {
+      "epoch": 1.47,
+      "learning_rate": 1.0773562633813061e-05,
+      "loss": 0.3321,
+      "step": 719
+    },
+    {
+      "epoch": 1.47,
+      "learning_rate": 1.075150316063071e-05,
+      "loss": 0.3079,
+      "step": 720
+    },
+    {
+      "epoch": 1.48,
+      "learning_rate": 1.0729440009092935e-05,
+      "loss": 0.3066,
+      "step": 721
+    },
+    {
+      "epoch": 1.48,
+      "learning_rate": 1.070737328719145e-05,
+      "loss": 0.3762,
+      "step": 722
+    },
+    {
+      "epoch": 1.48,
+      "learning_rate": 1.0685303102935437e-05,
+      "loss": 0.3437,
+      "step": 723
+    },
+    {
+      "epoch": 1.48,
+      "learning_rate": 1.066322956435104e-05,
+      "loss": 0.3024,
+      "step": 724
+    },
+    {
+      "epoch": 1.48,
+      "learning_rate": 1.0641152779480805e-05,
+      "loss": 0.3366,
+      "step": 725
+    },
+    {
+      "epoch": 1.49,
+      "learning_rate": 1.0619072856383181e-05,
+      "loss": 0.3112,
+      "step": 726
+    },
+    {
+      "epoch": 1.49,
+      "learning_rate": 1.0596989903131968e-05,
+      "loss": 0.3197,
+      "step": 727
+    },
+    {
+      "epoch": 1.49,
+      "learning_rate": 1.0574904027815802e-05,
+      "loss": 0.3269,
+      "step": 728
+    },
+    {
+      "epoch": 1.49,
+      "learning_rate": 1.0552815338537621e-05,
+      "loss": 0.3424,
+      "step": 729
+    },
+    {
+      "epoch": 1.49,
+      "learning_rate": 1.0530723943414135e-05,
+      "loss": 0.267,
+      "step": 730
+    },
+    {
+      "epoch": 1.5,
+      "learning_rate": 1.0508629950575297e-05,
+      "loss": 0.3163,
+      "step": 731
+    },
+    {
+      "epoch": 1.5,
+      "learning_rate": 1.0486533468163782e-05,
+      "loss": 0.3478,
+      "step": 732
+    },
+    {
+      "epoch": 1.5,
+      "learning_rate": 1.046443460433444e-05,
+      "loss": 0.3361,
+      "step": 733
+    },
+    {
+      "epoch": 1.5,
+      "learning_rate": 1.044233346725379e-05,
+      "loss": 0.3349,
+      "step": 734
+    },
+    {
+      "epoch": 1.5,
+      "learning_rate": 1.0420230165099464e-05,
+      "loss": 0.36,
+      "step": 735
+    },
+    {
+      "epoch": 1.51,
+      "learning_rate": 1.0398124806059702e-05,
+      "loss": 0.3069,
+      "step": 736
+    },
+    {
+      "epoch": 1.51,
+      "learning_rate": 1.0376017498332807e-05,
+      "loss": 0.3505,
+      "step": 737
+    },
+    {
+      "epoch": 1.51,
+      "learning_rate": 1.035390835012662e-05,
+      "loss": 0.3088,
+      "step": 738
+    },
+    {
+      "epoch": 1.51,
+      "learning_rate": 1.0331797469657992e-05,
+      "loss": 0.3462,
+      "step": 739
+    },
+    {
+      "epoch": 1.51,
+      "learning_rate": 1.0309684965152254e-05,
+      "loss": 0.3005,
+      "step": 740
+    },
+    {
+      "epoch": 1.52,
+      "learning_rate": 1.0287570944842682e-05,
+      "loss": 0.3206,
+      "step": 741
+    },
+    {
+      "epoch": 1.52,
+      "learning_rate": 1.0265455516969978e-05,
+      "loss": 0.3507,
+      "step": 742
+    },
+    {
+      "epoch": 1.52,
+      "learning_rate": 1.0243338789781722e-05,
+      "loss": 0.3474,
+      "step": 743
+    },
+    {
+      "epoch": 1.52,
+      "learning_rate": 1.022122087153187e-05,
+      "loss": 0.2976,
+      "step": 744
+    },
+    {
+      "epoch": 1.52,
+      "learning_rate": 1.0199101870480193e-05,
+      "loss": 0.3137,
+      "step": 745
+    },
+    {
+      "epoch": 1.53,
+      "learning_rate": 1.0176981894891768e-05,
+      "loss": 0.3088,
+      "step": 746
+    },
+    {
+      "epoch": 1.53,
+      "learning_rate": 1.0154861053036446e-05,
+      "loss": 0.3788,
+      "step": 747
+    },
+    {
+      "epoch": 1.53,
+      "learning_rate": 1.013273945318831e-05,
+      "loss": 0.336,
+      "step": 748
+    },
+    {
+      "epoch": 1.53,
+      "learning_rate": 1.011061720362516e-05,
+      "loss": 0.3105,
+      "step": 749
+    },
+    {
+      "epoch": 1.53,
+      "learning_rate": 1.0088494412627969e-05,
+      "loss": 0.328,
+      "step": 750
+    },
+    {
+      "epoch": 1.54,
+      "learning_rate": 1.0066371188480368e-05,
+      "loss": 0.2917,
+      "step": 751
+    },
+    {
+      "epoch": 1.54,
+      "learning_rate": 1.0044247639468106e-05,
+      "loss": 0.3444,
+      "step": 752
+    },
+    {
+      "epoch": 1.54,
+      "learning_rate": 1.0022123873878518e-05,
+      "loss": 0.331,
+      "step": 753
+    },
+    {
+      "epoch": 1.54,
+      "learning_rate": 1e-05,
+      "loss": 0.2771,
+      "step": 754
+    },
+    {
+      "epoch": 1.55,
+      "learning_rate": 9.977876126121485e-06,
+      "loss": 0.3524,
+      "step": 755
+    },
+    {
+      "epoch": 1.55,
+      "learning_rate": 9.955752360531896e-06,
+      "loss": 0.337,
+      "step": 756
+    },
+    {
+      "epoch": 1.55,
+      "learning_rate": 9.933628811519633e-06,
+      "loss": 0.3187,
+      "step": 757
+    },
+    {
+      "epoch": 1.55,
+      "learning_rate": 9.911505587372033e-06,
+      "loss": 0.3265,
+      "step": 758
+    },
+    {
+      "epoch": 1.55,
+      "learning_rate": 9.889382796374844e-06,
+      "loss": 0.3432,
+      "step": 759
+    },
+    {
+      "epoch": 1.56,
+      "learning_rate": 9.867260546811694e-06,
+      "loss": 0.3665,
+      "step": 760
+    },
+    {
+      "epoch": 1.56,
+      "learning_rate": 9.845138946963557e-06,
+      "loss": 0.3454,
+      "step": 761
+    },
+    {
+      "epoch": 1.56,
+      "learning_rate": 9.823018105108234e-06,
+      "loss": 0.3039,
+      "step": 762
+    },
+    {
+      "epoch": 1.56,
+      "learning_rate": 9.80089812951981e-06,
+      "loss": 0.3232,
+      "step": 763
+    },
+    {
+      "epoch": 1.56,
+      "learning_rate": 9.778779128468133e-06,
+      "loss": 0.3165,
+      "step": 764
+    },
+    {
+      "epoch": 1.57,
+      "learning_rate": 9.75666121021828e-06,
+      "loss": 0.3242,
+      "step": 765
+    },
+    {
+      "epoch": 1.57,
+      "learning_rate": 9.734544483030027e-06,
+      "loss": 0.3532,
+      "step": 766
+    },
+    {
+      "epoch": 1.57,
+      "learning_rate": 9.71242905515732e-06,
+      "loss": 0.3016,
+      "step": 767
+    },
+    {
+      "epoch": 1.57,
+      "learning_rate": 9.690315034847749e-06,
+      "loss": 0.3276,
+      "step": 768
+    },
+    {
+      "epoch": 1.57,
+      "learning_rate": 9.66820253034201e-06,
+      "loss": 0.319,
+      "step": 769
+    },
+    {
+      "epoch": 1.58,
+      "learning_rate": 9.646091649873384e-06,
+      "loss": 0.3451,
+      "step": 770
+    },
+    {
+      "epoch": 1.58,
+      "learning_rate": 9.623982501667196e-06,
+      "loss": 0.3364,
+      "step": 771
+    },
+    {
+      "epoch": 1.58,
+      "learning_rate": 9.601875193940301e-06,
+      "loss": 0.2921,
+      "step": 772
+    },
+    {
+      "epoch": 1.58,
+      "learning_rate": 9.57976983490054e-06,
+      "loss": 0.3406,
+      "step": 773
+    },
+    {
+      "epoch": 1.58,
+      "learning_rate": 9.557666532746214e-06,
+      "loss": 0.3051,
+      "step": 774
+    },
+    {
+      "epoch": 1.59,
+      "learning_rate": 9.535565395665563e-06,
+      "loss": 0.3153,
+      "step": 775
+    },
+    {
+      "epoch": 1.59,
+      "learning_rate": 9.513466531836223e-06,
+      "loss": 0.3328,
+      "step": 776
+    },
+    {
+      "epoch": 1.59,
+      "learning_rate": 9.491370049424704e-06,
+      "loss": 0.3573,
+      "step": 777
+    },
+    {
+      "epoch": 1.59,
+      "learning_rate": 9.469276056585869e-06,
+      "loss": 0.323,
+      "step": 778
+    },
+    {
+      "epoch": 1.59,
+      "learning_rate": 9.447184661462382e-06,
+      "loss": 0.3284,
+      "step": 779
+    },
+    {
+      "epoch": 1.6,
+      "learning_rate": 9.4250959721842e-06,
+      "loss": 0.2826,
+      "step": 780
+    },
+    {
+      "epoch": 1.6,
+      "learning_rate": 9.403010096868033e-06,
+      "loss": 0.2942,
+      "step": 781
+    },
+    {
+      "epoch": 1.6,
+      "learning_rate": 9.38092714361682e-06,
+      "loss": 0.2934,
+      "step": 782
+    },
+    {
+      "epoch": 1.6,
+      "learning_rate": 9.358847220519196e-06,
+      "loss": 0.303,
+      "step": 783
+    },
+    {
+      "epoch": 1.6,
+      "learning_rate": 9.336770435648963e-06,
+      "loss": 0.3518,
+      "step": 784
+    },
+    {
+      "epoch": 1.61,
+      "learning_rate": 9.314696897064566e-06,
+      "loss": 0.3143,
+      "step": 785
+    },
+    {
+      "epoch": 1.61,
+      "learning_rate": 9.292626712808557e-06,
+      "loss": 0.308,
+      "step": 786
+    },
+    {
+      "epoch": 1.61,
+      "learning_rate": 9.27055999090707e-06,
+      "loss": 0.2698,
+      "step": 787
+    },
+    {
+      "epoch": 1.61,
+      "learning_rate": 9.248496839369293e-06,
+      "loss": 0.3552,
+      "step": 788
+    },
+    {
+      "epoch": 1.61,
+      "learning_rate": 9.226437366186942e-06,
+      "loss": 0.3167,
+      "step": 789
+    },
+    {
+      "epoch": 1.62,
+      "learning_rate": 9.204381679333723e-06,
+      "loss": 0.363,
+      "step": 790
+    },
+    {
+      "epoch": 1.62,
+      "learning_rate": 9.182329886764811e-06,
+      "loss": 0.3381,
+      "step": 791
+    },
+    {
+      "epoch": 1.62,
+      "learning_rate": 9.160282096416317e-06,
+      "loss": 0.3084,
+      "step": 792
+    },
+    {
+      "epoch": 1.62,
+      "learning_rate": 9.13823841620477e-06,
+      "loss": 0.3399,
+      "step": 793
+    },
+    {
+      "epoch": 1.62,
+      "learning_rate": 9.116198954026577e-06,
+      "loss": 0.3185,
+      "step": 794
+    },
+    {
+      "epoch": 1.63,
+      "learning_rate": 9.094163817757495e-06,
+      "loss": 0.3098,
+      "step": 795
+    },
+    {
+      "epoch": 1.63,
+      "learning_rate": 9.072133115252113e-06,
+      "loss": 0.2861,
+      "step": 796
+    },
+    {
+      "epoch": 1.63,
+      "learning_rate": 9.050106954343316e-06,
+      "loss": 0.2954,
+      "step": 797
+    },
+    {
+      "epoch": 1.63,
+      "learning_rate": 9.02808544284176e-06,
+      "loss": 0.3474,
+      "step": 798
+    },
+    {
+      "epoch": 1.64,
+      "learning_rate": 9.006068688535342e-06,
+      "loss": 0.2994,
+      "step": 799
+    },
+    {
+      "epoch": 1.64,
+      "learning_rate": 8.984056799188678e-06,
+      "loss": 0.3349,
+      "step": 800
+    },
+    {
+      "epoch": 1.64,
+      "learning_rate": 8.962049882542566e-06,
+      "loss": 0.3506,
+      "step": 801
+    },
+    {
+      "epoch": 1.64,
+      "learning_rate": 8.94004804631347e-06,
+      "loss": 0.2862,
+      "step": 802
+    },
+    {
+      "epoch": 1.64,
+      "learning_rate": 8.918051398192982e-06,
+      "loss": 0.3139,
+      "step": 803
+    },
+    {
+      "epoch": 1.65,
+      "learning_rate": 8.896060045847305e-06,
+      "loss": 0.3711,
+      "step": 804
+    },
+    {
+      "epoch": 1.65,
+      "learning_rate": 8.874074096916716e-06,
+      "loss": 0.317,
+      "step": 805
+    },
+    {
+      "epoch": 1.65,
+      "learning_rate": 8.85209365901505e-06,
+      "loss": 0.2916,
+      "step": 806
+    },
+    {
+      "epoch": 1.65,
+      "learning_rate": 8.83011883972916e-06,
+      "loss": 0.3223,
+      "step": 807
+    },
+    {
+      "epoch": 1.65,
+      "learning_rate": 8.808149746618403e-06,
+      "loss": 0.3173,
+      "step": 808
+    },
+    {
+      "epoch": 1.66,
+      "learning_rate": 8.786186487214108e-06,
+      "loss": 0.3,
+      "step": 809
+    },
+    {
+      "epoch": 1.66,
+      "learning_rate": 8.764229169019047e-06,
+      "loss": 0.3337,
+      "step": 810
+    },
+    {
+      "epoch": 1.66,
+      "learning_rate": 8.742277899506914e-06,
+      "loss": 0.3205,
+      "step": 811
+    },
+    {
+      "epoch": 1.66,
+      "learning_rate": 8.7203327861218e-06,
+      "loss": 0.3316,
+      "step": 812
+    },
+    {
+      "epoch": 1.66,
+      "learning_rate": 8.698393936277655e-06,
+      "loss": 0.2895,
+      "step": 813
+    },
+    {
+      "epoch": 1.67,
+      "learning_rate": 8.676461457357777e-06,
+      "loss": 0.2884,
+      "step": 814
+    },
+    {
+      "epoch": 1.67,
+      "learning_rate": 8.654535456714282e-06,
+      "loss": 0.3226,
+      "step": 815
+    },
+    {
+      "epoch": 1.67,
+      "learning_rate": 8.632616041667577e-06,
+      "loss": 0.2972,
+      "step": 816
+    },
+    {
+      "epoch": 1.67,
+      "learning_rate": 8.610703319505824e-06,
+      "loss": 0.2968,
+      "step": 817
+    },
+    {
+      "epoch": 1.67,
+      "learning_rate": 8.588797397484446e-06,
+      "loss": 0.2679,
+      "step": 818
+    },
+    {
+      "epoch": 1.68,
+      "learning_rate": 8.566898382825558e-06,
+      "loss": 0.3664,
+      "step": 819
+    },
+    {
+      "epoch": 1.68,
+      "learning_rate": 8.545006382717487e-06,
+      "loss": 0.3131,
+      "step": 820
+    },
+    {
+      "epoch": 1.68,
+      "learning_rate": 8.523121504314212e-06,
+      "loss": 0.3254,
+      "step": 821
+    },
+    {
+      "epoch": 1.68,
+      "learning_rate": 8.501243854734857e-06,
+      "loss": 0.2907,
+      "step": 822
+    },
+    {
+      "epoch": 1.68,
+      "learning_rate": 8.479373541063166e-06,
+      "loss": 0.3237,
+      "step": 823
+    },
+    {
+      "epoch": 1.69,
+      "learning_rate": 8.457510670346976e-06,
+      "loss": 0.3237,
+      "step": 824
+    },
+    {
+      "epoch": 1.69,
+      "learning_rate": 8.43565534959769e-06,
+      "loss": 0.3112,
+      "step": 825
+    },
+    {
+      "epoch": 1.69,
+      "learning_rate": 8.41380768578976e-06,
+      "loss": 0.3462,
+      "step": 826
+    },
+    {
+      "epoch": 1.69,
+      "learning_rate": 8.391967785860158e-06,
+      "loss": 0.3212,
+      "step": 827
+    },
+    {
+      "epoch": 1.69,
+      "learning_rate": 8.370135756707853e-06,
+      "loss": 0.3074,
+      "step": 828
+    },
+    {
+      "epoch": 1.7,
+      "learning_rate": 8.348311705193294e-06,
+      "loss": 0.3369,
+      "step": 829
+    },
+    {
+      "epoch": 1.7,
+      "learning_rate": 8.326495738137875e-06,
+      "loss": 0.2984,
+      "step": 830
+    },
+    {
+      "epoch": 1.7,
+      "learning_rate": 8.304687962323426e-06,
+      "loss": 0.3003,
+      "step": 831
+    },
+    {
+      "epoch": 1.7,
+      "learning_rate": 8.282888484491682e-06,
+      "loss": 0.3242,
+      "step": 832
+    },
+    {
+      "epoch": 1.7,
+      "learning_rate": 8.26109741134376e-06,
+      "loss": 0.3402,
+      "step": 833
+    },
+    {
+      "epoch": 1.71,
+      "learning_rate": 8.239314849539639e-06,
+      "loss": 0.3357,
+      "step": 834
+    },
+    {
+      "epoch": 1.71,
+      "learning_rate": 8.21754090569764e-06,
+      "loss": 0.2947,
+      "step": 835
+    },
+    {
+      "epoch": 1.71,
+      "learning_rate": 8.195775686393898e-06,
+      "loss": 0.3186,
+      "step": 836
+    },
+    {
+      "epoch": 1.71,
+      "learning_rate": 8.17401929816185e-06,
+      "loss": 0.2997,
+      "step": 837
+    },
+    {
+      "epoch": 1.72,
+      "learning_rate": 8.152271847491706e-06,
+      "loss": 0.3464,
+      "step": 838
+    },
+    {
+      "epoch": 1.72,
+      "learning_rate": 8.130533440829927e-06,
+      "loss": 0.285,
+      "step": 839
+    },
+    {
+      "epoch": 1.72,
+      "learning_rate": 8.108804184578708e-06,
+      "loss": 0.3286,
+      "step": 840
+    },
+    {
+      "epoch": 1.72,
+      "learning_rate": 8.087084185095458e-06,
+      "loss": 0.2942,
+      "step": 841
+    },
+    {
+      "epoch": 1.72,
+      "learning_rate": 8.065373548692272e-06,
+      "loss": 0.2883,
+      "step": 842
+    },
+    {
+      "epoch": 1.73,
+      "learning_rate": 8.043672381635423e-06,
+      "loss": 0.3024,
+      "step": 843
+    },
+    {
+      "epoch": 1.73,
+      "learning_rate": 8.021980790144828e-06,
+      "loss": 0.3351,
+      "step": 844
+    },
+    {
+      "epoch": 1.73,
+      "learning_rate": 8.00029888039354e-06,
+      "loss": 0.3142,
+      "step": 845
+    },
+    {
+      "epoch": 1.73,
+      "learning_rate": 7.978626758507218e-06,
+      "loss": 0.3111,
+      "step": 846
+    },
+    {
+      "epoch": 1.73,
+      "learning_rate": 7.956964530563617e-06,
+      "loss": 0.3333,
+      "step": 847
+    },
+    {
+      "epoch": 1.74,
+      "learning_rate": 7.935312302592062e-06,
+      "loss": 0.3401,
+      "step": 848
+    },
+    {
+      "epoch": 1.74,
+      "learning_rate": 7.913670180572936e-06,
+      "loss": 0.2711,
+      "step": 849
+    },
+    {
+      "epoch": 1.74,
+      "learning_rate": 7.892038270437153e-06,
+      "loss": 0.3327,
+      "step": 850
+    },
+    {
+      "epoch": 1.74,
+      "learning_rate": 7.870416678065638e-06,
+      "loss": 0.343,
+      "step": 851
+    },
+    {
+      "epoch": 1.74,
+      "learning_rate": 7.848805509288824e-06,
+      "loss": 0.2779,
+      "step": 852
+    },
+    {
+      "epoch": 1.75,
+      "learning_rate": 7.827204869886121e-06,
+      "loss": 0.3348,
+      "step": 853
+    },
+    {
+      "epoch": 1.75,
+      "learning_rate": 7.805614865585397e-06,
+      "loss": 0.3044,
+      "step": 854
+    },
+    {
+      "epoch": 1.75,
+      "learning_rate": 7.784035602062465e-06,
+      "loss": 0.3639,
+      "step": 855
+    },
+    {
+      "epoch": 1.75,
+      "learning_rate": 7.762467184940574e-06,
+      "loss": 0.275,
+      "step": 856
+    },
+    {
+      "epoch": 1.75,
+      "learning_rate": 7.740909719789876e-06,
+      "loss": 0.3276,
+      "step": 857
+    },
+    {
+      "epoch": 1.76,
+      "learning_rate": 7.719363312126914e-06,
+      "loss": 0.3012,
+      "step": 858
+    },
+    {
+      "epoch": 1.76,
+      "learning_rate": 7.69782806741412e-06,
+      "loss": 0.2995,
+      "step": 859
+    },
+    {
+      "epoch": 1.76,
+      "learning_rate": 7.676304091059273e-06,
+      "loss": 0.3349,
+      "step": 860
+    },
+    {
+      "epoch": 1.76,
+      "learning_rate": 7.654791488415008e-06,
+      "loss": 0.3129,
+      "step": 861
+    },
+    {
+      "epoch": 1.76,
+      "learning_rate": 7.633290364778284e-06,
+      "loss": 0.3115,
+      "step": 862
+    },
+    {
+      "epoch": 1.77,
+      "learning_rate": 7.611800825389877e-06,
+      "loss": 0.3657,
+      "step": 863
+    },
+    {
+      "epoch": 1.77,
+      "learning_rate": 7.590322975433857e-06,
+      "loss": 0.3107,
+      "step": 864
+    },
+    {
+      "epoch": 1.77,
+      "learning_rate": 7.5688569200370866e-06,
+      "loss": 0.2923,
+      "step": 865
+    },
+    {
+      "epoch": 1.77,
+      "learning_rate": 7.54740276426869e-06,
+      "loss": 0.3494,
+      "step": 866
+    },
+    {
+      "epoch": 1.77,
+      "learning_rate": 7.52596061313955e-06,
+      "loss": 0.2754,
+      "step": 867
+    },
+    {
+      "epoch": 1.78,
+      "learning_rate": 7.504530571601792e-06,
+      "loss": 0.3288,
+      "step": 868
+    },
+    {
+      "epoch": 1.78,
+      "learning_rate": 7.4831127445482675e-06,
+      "loss": 0.3017,
+      "step": 869
+    },
+    {
+      "epoch": 1.78,
+      "learning_rate": 7.461707236812042e-06,
+      "loss": 0.3228,
+      "step": 870
+    },
+    {
+      "epoch": 1.78,
+      "learning_rate": 7.440314153165881e-06,
+      "loss": 0.2946,
+      "step": 871
+    },
+    {
+      "epoch": 1.78,
+      "learning_rate": 7.418933598321741e-06,
+      "loss": 0.3247,
+      "step": 872
+    },
+    {
+      "epoch": 1.79,
+      "learning_rate": 7.397565676930253e-06,
+      "loss": 0.2792,
+      "step": 873
+    },
+    {
+      "epoch": 1.79,
+      "learning_rate": 7.376210493580211e-06,
+      "loss": 0.3614,
+      "step": 874
+    },
+    {
+      "epoch": 1.79,
+      "learning_rate": 7.35486815279806e-06,
+      "loss": 0.3277,
+      "step": 875
+    },
+    {
+      "epoch": 1.79,
+      "learning_rate": 7.33353875904739e-06,
+      "loss": 0.3608,
+      "step": 876
+    },
+    {
+      "epoch": 1.79,
+      "learning_rate": 7.3122224167284094e-06,
+      "loss": 0.345,
+      "step": 877
+    },
+    {
+      "epoch": 1.8,
+      "learning_rate": 7.290919230177455e-06,
+      "loss": 0.3057,
+      "step": 878
+    },
+    {
+      "epoch": 1.8,
+      "learning_rate": 7.269629303666463e-06,
+      "loss": 0.2788,
+      "step": 879
+    },
+    {
+      "epoch": 1.8,
+      "learning_rate": 7.2483527414024714e-06,
+      "loss": 0.329,
+      "step": 880
+    },
+    {
+      "epoch": 1.8,
+      "learning_rate": 7.227089647527101e-06,
+      "loss": 0.3128,
+      "step": 881
+    },
+    {
+      "epoch": 1.81,
+      "learning_rate": 7.20584012611605e-06,
+      "loss": 0.2957,
+      "step": 882
+    },
+    {
+      "epoch": 1.81,
+      "learning_rate": 7.184604281178588e-06,
+      "loss": 0.2876,
+      "step": 883
+    },
+    {
+      "epoch": 1.81,
+      "learning_rate": 7.163382216657033e-06,
+      "loss": 0.2758,
+      "step": 884
+    },
+    {
+      "epoch": 1.81,
+      "learning_rate": 7.142174036426267e-06,
+      "loss": 0.327,
+      "step": 885
+    },
+    {
+      "epoch": 1.81,
+      "learning_rate": 7.120979844293201e-06,
+      "loss": 0.3464,
+      "step": 886
+    },
+    {
+      "epoch": 1.82,
+      "learning_rate": 7.099799743996282e-06,
+      "loss": 0.3336,
+      "step": 887
+    },
+    {
+      "epoch": 1.82,
+      "learning_rate": 7.078633839204984e-06,
+      "loss": 0.309,
+      "step": 888
+    },
+    {
+      "epoch": 1.82,
+      "learning_rate": 7.0574822335193015e-06,
+      "loss": 0.2823,
+      "step": 889
+    },
+    {
+      "epoch": 1.82,
+      "learning_rate": 7.036345030469232e-06,
+      "loss": 0.2716,
+      "step": 890
+    },
+    {
+      "epoch": 1.82,
+      "learning_rate": 7.015222333514282e-06,
+      "loss": 0.2895,
+      "step": 891
+    },
+    {
+      "epoch": 1.83,
+      "learning_rate": 6.9941142460429555e-06,
+      "loss": 0.3543,
+      "step": 892
+    },
+    {
+      "epoch": 1.83,
+      "learning_rate": 6.9730208713722456e-06,
+      "loss": 0.2767,
+      "step": 893
+    },
+    {
+      "epoch": 1.83,
+      "learning_rate": 6.951942312747135e-06,
+      "loss": 0.3497,
+      "step": 894
+    },
+    {
+      "epoch": 1.83,
+      "learning_rate": 6.930878673340084e-06,
+      "loss": 0.313,
+      "step": 895
+    },
+    {
+      "epoch": 1.83,
+      "learning_rate": 6.909830056250527e-06,
+      "loss": 0.2855,
+      "step": 896
+    },
+    {
+      "epoch": 1.84,
+      "learning_rate": 6.888796564504373e-06,
+      "loss": 0.2798,
+      "step": 897
+    },
+    {
+      "epoch": 1.84,
+      "learning_rate": 6.867778301053496e-06,
+      "loss": 0.3395,
+      "step": 898
+    },
+    {
+      "epoch": 1.84,
+      "learning_rate": 6.846775368775231e-06,
+      "loss": 0.3373,
+      "step": 899
+    },
+    {
+      "epoch": 1.84,
+      "learning_rate": 6.825787870471873e-06,
+      "loss": 0.3302,
+      "step": 900
+    },
+    {
+      "epoch": 1.84,
+      "learning_rate": 6.804815908870175e-06,
+      "loss": 0.3389,
+      "step": 901
+    },
+    {
+      "epoch": 1.85,
+      "learning_rate": 6.783859586620839e-06,
+      "loss": 0.2978,
+      "step": 902
+    },
+    {
+      "epoch": 1.85,
+      "learning_rate": 6.762919006298022e-06,
+      "loss": 0.3461,
+      "step": 903
+    },
+    {
+      "epoch": 1.85,
+      "learning_rate": 6.741994270398826e-06,
+      "loss": 0.2705,
+      "step": 904
+    },
+    {
+      "epoch": 1.85,
+      "learning_rate": 6.721085481342804e-06,
+      "loss": 0.3264,
+      "step": 905
+    },
+    {
+      "epoch": 1.85,
+      "learning_rate": 6.700192741471447e-06,
+      "loss": 0.3322,
+      "step": 906
+    },
+    {
+      "epoch": 1.86,
+      "learning_rate": 6.6793161530477e-06,
+      "loss": 0.3171,
+      "step": 907
+    },
+    {
+      "epoch": 1.86,
+      "learning_rate": 6.658455818255445e-06,
+      "loss": 0.3121,
+      "step": 908
+    },
+    {
+      "epoch": 1.86,
+      "learning_rate": 6.637611839199013e-06,
+      "loss": 0.2939,
+      "step": 909
+    },
+    {
+      "epoch": 1.86,
+      "learning_rate": 6.616784317902673e-06,
+      "loss": 0.3272,
+      "step": 910
+    },
+    {
+      "epoch": 1.86,
+      "learning_rate": 6.595973356310148e-06,
+      "loss": 0.3363,
+      "step": 911
+    },
+    {
+      "epoch": 1.87,
+      "learning_rate": 6.575179056284095e-06,
+      "loss": 0.3172,
+      "step": 912
+    },
+    {
+      "epoch": 1.87,
+      "learning_rate": 6.554401519605629e-06,
+      "loss": 0.2938,
+      "step": 913
+    },
+    {
+      "epoch": 1.87,
+      "learning_rate": 6.533640847973809e-06,
+      "loss": 0.325,
+      "step": 914
+    },
+    {
+      "epoch": 1.87,
+      "learning_rate": 6.5128971430051434e-06,
+      "loss": 0.3093,
+      "step": 915
+    },
+    {
+      "epoch": 1.87,
+      "learning_rate": 6.4921705062331e-06,
+      "loss": 0.3371,
+      "step": 916
+    },
+    {
+      "epoch": 1.88,
+      "learning_rate": 6.471461039107599e-06,
+      "loss": 0.2857,
+      "step": 917
+    },
+    {
+      "epoch": 1.88,
+      "learning_rate": 6.450768842994523e-06,
+      "loss": 0.3078,
+      "step": 918
+    },
+    {
+      "epoch": 1.88,
+      "learning_rate": 6.430094019175217e-06,
+      "loss": 0.2963,
+      "step": 919
+    },
+    {
+      "epoch": 1.88,
+      "learning_rate": 6.409436668845997e-06,
+      "loss": 0.3197,
+      "step": 920
+    },
+    {
+      "epoch": 1.88,
+      "learning_rate": 6.3887968931176505e-06,
+      "loss": 0.2894,
+      "step": 921
+    },
+    {
+      "epoch": 1.89,
+      "learning_rate": 6.368174793014943e-06,
+      "loss": 0.3123,
+      "step": 922
+    },
+    {
+      "epoch": 1.89,
+      "learning_rate": 6.347570469476124e-06,
+      "loss": 0.3355,
+      "step": 923
+    },
+    {
+      "epoch": 1.89,
+      "learning_rate": 6.326984023352435e-06,
+      "loss": 0.3466,
+      "step": 924
+    },
+    {
+      "epoch": 1.89,
+      "learning_rate": 6.30641555540761e-06,
+      "loss": 0.3254,
+      "step": 925
+    },
+    {
+      "epoch": 1.9,
+      "learning_rate": 6.285865166317386e-06,
+      "loss": 0.3254,
+      "step": 926
+    },
+    {
+      "epoch": 1.9,
+      "learning_rate": 6.265332956669016e-06,
+      "loss": 0.3435,
+      "step": 927
+    },
+    {
+      "epoch": 1.9,
+      "learning_rate": 6.2448190269607625e-06,
+      "loss": 0.2964,
+      "step": 928
+    },
+    {
+      "epoch": 1.9,
+      "learning_rate": 6.224323477601417e-06,
+      "loss": 0.332,
+      "step": 929
+    },
+    {
+      "epoch": 1.9,
+      "learning_rate": 6.2038464089098095e-06,
+      "loss": 0.3451,
+      "step": 930
+    },
+    {
+      "epoch": 1.91,
+      "learning_rate": 6.183387921114307e-06,
+      "loss": 0.3159,
+      "step": 931
+    },
+    {
+      "epoch": 1.91,
+      "learning_rate": 6.162948114352328e-06,
+      "loss": 0.3014,
+      "step": 932
+    },
+    {
+      "epoch": 1.91,
+      "learning_rate": 6.142527088669868e-06,
+      "loss": 0.2643,
+      "step": 933
+    },
+    {
+      "epoch": 1.91,
+      "learning_rate": 6.122124944020978e-06,
+      "loss": 0.3337,
+      "step": 934
+    },
+    {
+      "epoch": 1.91,
+      "learning_rate": 6.101741780267303e-06,
+      "loss": 0.2863,
+      "step": 935
+    },
+    {
+      "epoch": 1.92,
+      "learning_rate": 6.081377697177576e-06,
+      "loss": 0.3193,
+      "step": 936
+    },
+    {
+      "epoch": 1.92,
+      "learning_rate": 6.061032794427144e-06,
+      "loss": 0.2779,
+      "step": 937
+    },
+    {
+      "epoch": 1.92,
+      "learning_rate": 6.040707171597466e-06,
+      "loss": 0.28,
+      "step": 938
+    },
+    {
+      "epoch": 1.92,
+      "learning_rate": 6.020400928175637e-06,
+      "loss": 0.2749,
+      "step": 939
+    },
+    {
+      "epoch": 1.92,
+      "learning_rate": 6.000114163553894e-06,
+      "loss": 0.3445,
+      "step": 940
+    },
+    {
+      "epoch": 1.93,
+      "learning_rate": 5.979846977029132e-06,
+      "loss": 0.2927,
+      "step": 941
+    },
+    {
+      "epoch": 1.93,
+      "learning_rate": 5.959599467802418e-06,
+      "loss": 0.2813,
+      "step": 942
+    },
+    {
+      "epoch": 1.93,
+      "learning_rate": 5.9393717349785064e-06,
+      "loss": 0.2619,
+      "step": 943
+    },
+    {
+      "epoch": 1.93,
+      "learning_rate": 5.919163877565351e-06,
+      "loss": 0.2857,
+      "step": 944
+    },
+    {
+      "epoch": 1.93,
+      "learning_rate": 5.898975994473622e-06,
+      "loss": 0.3253,
+      "step": 945
+    },
+    {
+      "epoch": 1.94,
+      "learning_rate": 5.878808184516225e-06,
+      "loss": 0.3377,
+      "step": 946
+    },
+    {
+      "epoch": 1.94,
+      "learning_rate": 5.858660546407809e-06,
+      "loss": 0.3062,
+      "step": 947
+    },
+    {
+      "epoch": 1.94,
+      "learning_rate": 5.8385331787642956e-06,
+      "loss": 0.3498,
+      "step": 948
+    },
+    {
+      "epoch": 1.94,
+      "learning_rate": 5.818426180102382e-06,
+      "loss": 0.3081,
+      "step": 949
+    },
+    {
+      "epoch": 1.94,
+      "learning_rate": 5.798339648839073e-06,
+      "loss": 0.2663,
+      "step": 950
+    },
+    {
+      "epoch": 1.95,
+      "learning_rate": 5.778273683291184e-06,
+      "loss": 0.2549,
+      "step": 951
+    },
+    {
+      "epoch": 1.95,
+      "learning_rate": 5.75822838167488e-06,
+      "loss": 0.298,
+      "step": 952
+    },
+    {
+      "epoch": 1.95,
+      "learning_rate": 5.73820384210517e-06,
+      "loss": 0.362,
+      "step": 953
+    },
+    {
+      "epoch": 1.95,
+      "learning_rate": 5.718200162595449e-06,
+      "loss": 0.2851,
+      "step": 954
+    },
+    {
+      "epoch": 1.95,
+      "learning_rate": 5.698217441057004e-06,
+      "loss": 0.286,
+      "step": 955
+    },
+    {
+      "epoch": 1.96,
+      "learning_rate": 5.678255775298542e-06,
+      "loss": 0.311,
+      "step": 956
+    },
+    {
+      "epoch": 1.96,
+      "learning_rate": 5.658315263025709e-06,
+      "loss": 0.3515,
+      "step": 957
+    },
+    {
+      "epoch": 1.96,
+      "learning_rate": 5.638396001840613e-06,
+      "loss": 0.2928,
+      "step": 958
+    },
+    {
+      "epoch": 1.96,
+      "learning_rate": 5.61849808924134e-06,
+      "loss": 0.3245,
+      "step": 959
+    },
+    {
+      "epoch": 1.96,
+      "learning_rate": 5.598621622621489e-06,
+      "loss": 0.3151,
+      "step": 960
+    },
+    {
+      "epoch": 1.97,
+      "learning_rate": 5.578766699269682e-06,
+      "loss": 0.3068,
+      "step": 961
+    },
+    {
+      "epoch": 1.97,
+      "learning_rate": 5.5589334163690975e-06,
+      "loss": 0.3193,
+      "step": 962
+    },
+    {
+      "epoch": 1.97,
+      "learning_rate": 5.539121870996991e-06,
+      "loss": 0.317,
+      "step": 963
+    },
+    {
+      "epoch": 1.97,
+      "learning_rate": 5.519332160124215e-06,
+      "loss": 0.3163,
+      "step": 964
+    },
+    {
+      "epoch": 1.97,
+      "learning_rate": 5.499564380614761e-06,
+      "loss": 0.2811,
+      "step": 965
+    },
+    {
+      "epoch": 1.98,
+      "learning_rate": 5.47981862922526e-06,
+      "loss": 0.3202,
+      "step": 966
+    },
+    {
+      "epoch": 1.98,
+      "learning_rate": 5.460095002604533e-06,
+      "loss": 0.3682,
+      "step": 967
+    },
+    {
+      "epoch": 1.98,
+      "learning_rate": 5.440393597293103e-06,
+      "loss": 0.3017,
+      "step": 968
+    },
+    {
+      "epoch": 1.98,
+      "learning_rate": 5.420714509722729e-06,
+      "loss": 0.2656,
+      "step": 969
+    },
+    {
+      "epoch": 1.99,
+      "learning_rate": 5.401057836215928e-06,
+      "loss": 0.3457,
+      "step": 970
+    },
+    {
+      "epoch": 1.99,
+      "learning_rate": 5.381423672985516e-06,
+      "loss": 0.3266,
+      "step": 971
+    },
+    {
+      "epoch": 1.99,
+      "learning_rate": 5.361812116134122e-06,
+      "loss": 0.3399,
+      "step": 972
+    },
+    {
+      "epoch": 1.99,
+      "learning_rate": 5.342223261653723e-06,
+      "loss": 0.2869,
+      "step": 973
+    },
+    {
+      "epoch": 1.99,
+      "learning_rate": 5.322657205425184e-06,
+      "loss": 0.3632,
+      "step": 974
+    },
+    {
+      "epoch": 2.0,
+      "learning_rate": 5.303114043217771e-06,
+      "loss": 0.3032,
+      "step": 975
+    },
+    {
+      "epoch": 2.0,
+      "learning_rate": 5.283593870688697e-06,
+      "loss": 0.2694,
+      "step": 976
+    },
+    {
+      "epoch": 2.0,
+      "learning_rate": 5.264096783382646e-06,
+      "loss": 0.294,
+      "step": 977
+    },
+    {
+      "epoch": 2.0,
+      "learning_rate": 5.244622876731309e-06,
+      "loss": 0.2232,
+      "step": 978
+    },
+    {
+      "epoch": 2.0,
+      "learning_rate": 5.225172246052914e-06,
+      "loss": 0.2009,
+      "step": 979
+    },
+    {
+      "epoch": 2.01,
+      "learning_rate": 5.205744986551763e-06,
+      "loss": 0.173,
+      "step": 980
+    },
+    {
+      "epoch": 2.01,
+      "learning_rate": 5.186341193317761e-06,
+      "loss": 0.1939,
+      "step": 981
+    },
+    {
+      "epoch": 2.01,
+      "learning_rate": 5.166960961325955e-06,
+      "loss": 0.159,
+      "step": 982
+    },
+    {
+      "epoch": 2.01,
+      "learning_rate": 5.147604385436069e-06,
+      "loss": 0.1942,
+      "step": 983
+    },
+    {
+      "epoch": 2.01,
+      "learning_rate": 5.128271560392037e-06,
+      "loss": 0.1608,
+      "step": 984
+    },
+    {
+      "epoch": 2.02,
+      "learning_rate": 5.10896258082154e-06,
+      "loss": 0.1567,
+      "step": 985
+    },
+    {
+      "epoch": 2.02,
+      "learning_rate": 5.089677541235544e-06,
+      "loss": 0.1514,
+      "step": 986
+    },
+    {
+      "epoch": 2.02,
+      "learning_rate": 5.0704165360278365e-06,
+      "loss": 0.1574,
+      "step": 987
+    },
+    {
+      "epoch": 2.02,
+      "learning_rate": 5.051179659474568e-06,
+      "loss": 0.1477,
+      "step": 988
+    },
+    {
+      "epoch": 2.02,
+      "learning_rate": 5.03196700573378e-06,
+      "loss": 0.1771,
+      "step": 989
+    },
+    {
+      "epoch": 2.03,
+      "learning_rate": 5.01277866884496e-06,
+      "loss": 0.143,
+      "step": 990
+    },
+    {
+      "epoch": 2.03,
+      "learning_rate": 4.993614742728566e-06,
+      "loss": 0.145,
+      "step": 991
+    },
+    {
+      "epoch": 2.03,
+      "learning_rate": 4.974475321185578e-06,
+      "loss": 0.1609,
+      "step": 992
+    },
+    {
+      "epoch": 2.03,
+      "learning_rate": 4.955360497897032e-06,
+      "loss": 0.1603,
+      "step": 993
+    },
+    {
+      "epoch": 2.03,
+      "learning_rate": 4.936270366423563e-06,
+      "loss": 0.1875,
+      "step": 994
+    },
+    {
+      "epoch": 2.04,
+      "learning_rate": 4.9172050202049524e-06,
+      "loss": 0.1439,
+      "step": 995
+    },
+    {
+      "epoch": 2.04,
+      "learning_rate": 4.89816455255966e-06,
+      "loss": 0.1747,
+      "step": 996
+    },
+    {
+      "epoch": 2.04,
+      "learning_rate": 4.879149056684377e-06,
+      "loss": 0.1581,
+      "step": 997
+    },
+    {
+      "epoch": 2.04,
+      "learning_rate": 4.8601586256535645e-06,
+      "loss": 0.1515,
+      "step": 998
+    },
+    {
+      "epoch": 2.04,
+      "learning_rate": 4.841193352419003e-06,
+      "loss": 0.1568,
+      "step": 999
+    },
+    {
+      "epoch": 2.05,
+      "learning_rate": 4.82225332980933e-06,
+      "loss": 0.1533,
+      "step": 1000
+    }
+  ],
+  "max_steps": 1464,
+  "num_train_epochs": 3,
+  "total_flos": 2.7459166715405926e+17,
+  "trial_name": null,
+  "trial_params": null
+}