diff --git "a/trainer_state.json" "b/trainer_state.json"
new file mode 100644--- /dev/null
+++ "b/trainer_state.json"
@@ -0,0 +1,8629 @@
+{
+  "best_metric": null,
+  "best_model_checkpoint": null,
+  "epoch": 3.0,
+  "global_step": 1434,
+  "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": 1.1181,
+      "step": 1
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 9.090909090909091e-07,
+      "loss": 1.1064,
+      "step": 2
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 1.3636363636363636e-06,
+      "loss": 1.0985,
+      "step": 3
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 1.8181818181818183e-06,
+      "loss": 1.1074,
+      "step": 4
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 2.2727272727272728e-06,
+      "loss": 1.111,
+      "step": 5
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 2.7272727272727272e-06,
+      "loss": 1.1003,
+      "step": 6
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 3.181818181818182e-06,
+      "loss": 1.0564,
+      "step": 7
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 3.6363636363636366e-06,
+      "loss": 1.0259,
+      "step": 8
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.0909090909090915e-06,
+      "loss": 1.0098,
+      "step": 9
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.5454545454545455e-06,
+      "loss": 0.9849,
+      "step": 10
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 5e-06,
+      "loss": 1.0319,
+      "step": 11
+    },
+    {
+      "epoch": 0.03,
+      "learning_rate": 5.4545454545454545e-06,
+      "loss": 0.9849,
+      "step": 12
+    },
+    {
+      "epoch": 0.03,
+      "learning_rate": 5.90909090909091e-06,
+      "loss": 0.9733,
+      "step": 13
+    },
+    {
+      "epoch": 0.03,
+      "learning_rate": 6.363636363636364e-06,
+      "loss": 0.9598,
+      "step": 14
+    },
+    {
+      "epoch": 0.03,
+      "learning_rate": 6.818181818181818e-06,
+      "loss": 0.9581,
+      "step": 15
+    },
+    {
+      "epoch": 0.03,
+      "learning_rate": 7.272727272727273e-06,
+      "loss": 0.9664,
+      "step": 16
+    },
+    {
+      "epoch": 0.04,
+      "learning_rate": 7.727272727272727e-06,
+      "loss": 0.9423,
+      "step": 17
+    },
+    {
+      "epoch": 0.04,
+      "learning_rate": 8.181818181818183e-06,
+      "loss": 0.9156,
+      "step": 18
+    },
+    {
+      "epoch": 0.04,
+      "learning_rate": 8.636363636363637e-06,
+      "loss": 0.9261,
+      "step": 19
+    },
+    {
+      "epoch": 0.04,
+      "learning_rate": 9.090909090909091e-06,
+      "loss": 0.9207,
+      "step": 20
+    },
+    {
+      "epoch": 0.04,
+      "learning_rate": 9.545454545454547e-06,
+      "loss": 0.9141,
+      "step": 21
+    },
+    {
+      "epoch": 0.05,
+      "learning_rate": 1e-05,
+      "loss": 0.8864,
+      "step": 22
+    },
+    {
+      "epoch": 0.05,
+      "learning_rate": 1.0454545454545455e-05,
+      "loss": 0.878,
+      "step": 23
+    },
+    {
+      "epoch": 0.05,
+      "learning_rate": 1.0909090909090909e-05,
+      "loss": 0.8756,
+      "step": 24
+    },
+    {
+      "epoch": 0.05,
+      "learning_rate": 1.1363636363636366e-05,
+      "loss": 0.8778,
+      "step": 25
+    },
+    {
+      "epoch": 0.05,
+      "learning_rate": 1.181818181818182e-05,
+      "loss": 0.8622,
+      "step": 26
+    },
+    {
+      "epoch": 0.06,
+      "learning_rate": 1.2272727272727274e-05,
+      "loss": 0.861,
+      "step": 27
+    },
+    {
+      "epoch": 0.06,
+      "learning_rate": 1.2727272727272728e-05,
+      "loss": 0.8553,
+      "step": 28
+    },
+    {
+      "epoch": 0.06,
+      "learning_rate": 1.3181818181818183e-05,
+      "loss": 0.8693,
+      "step": 29
+    },
+    {
+      "epoch": 0.06,
+      "learning_rate": 1.3636363636363637e-05,
+      "loss": 0.8427,
+      "step": 30
+    },
+    {
+      "epoch": 0.06,
+      "learning_rate": 1.4090909090909092e-05,
+      "loss": 0.845,
+      "step": 31
+    },
+    {
+      "epoch": 0.07,
+      "learning_rate": 1.4545454545454546e-05,
+      "loss": 0.8239,
+      "step": 32
+    },
+    {
+      "epoch": 0.07,
+      "learning_rate": 1.5000000000000002e-05,
+      "loss": 0.8182,
+      "step": 33
+    },
+    {
+      "epoch": 0.07,
+      "learning_rate": 1.5454545454545454e-05,
+      "loss": 0.8311,
+      "step": 34
+    },
+    {
+      "epoch": 0.07,
+      "learning_rate": 1.590909090909091e-05,
+      "loss": 0.8175,
+      "step": 35
+    },
+    {
+      "epoch": 0.08,
+      "learning_rate": 1.6363636363636366e-05,
+      "loss": 0.814,
+      "step": 36
+    },
+    {
+      "epoch": 0.08,
+      "learning_rate": 1.681818181818182e-05,
+      "loss": 0.828,
+      "step": 37
+    },
+    {
+      "epoch": 0.08,
+      "learning_rate": 1.7272727272727274e-05,
+      "loss": 0.7977,
+      "step": 38
+    },
+    {
+      "epoch": 0.08,
+      "learning_rate": 1.772727272727273e-05,
+      "loss": 0.8009,
+      "step": 39
+    },
+    {
+      "epoch": 0.08,
+      "learning_rate": 1.8181818181818182e-05,
+      "loss": 0.8167,
+      "step": 40
+    },
+    {
+      "epoch": 0.09,
+      "learning_rate": 1.8636363636363638e-05,
+      "loss": 0.818,
+      "step": 41
+    },
+    {
+      "epoch": 0.09,
+      "learning_rate": 1.9090909090909094e-05,
+      "loss": 0.8077,
+      "step": 42
+    },
+    {
+      "epoch": 0.09,
+      "learning_rate": 1.9545454545454546e-05,
+      "loss": 0.8119,
+      "step": 43
+    },
+    {
+      "epoch": 0.09,
+      "learning_rate": 2e-05,
+      "loss": 0.7857,
+      "step": 44
+    },
+    {
+      "epoch": 0.09,
+      "learning_rate": 1.9999974458878423e-05,
+      "loss": 0.7913,
+      "step": 45
+    },
+    {
+      "epoch": 0.1,
+      "learning_rate": 1.9999897835644166e-05,
+      "loss": 0.7919,
+      "step": 46
+    },
+    {
+      "epoch": 0.1,
+      "learning_rate": 1.9999770130688628e-05,
+      "loss": 0.7924,
+      "step": 47
+    },
+    {
+      "epoch": 0.1,
+      "learning_rate": 1.9999591344664163e-05,
+      "loss": 0.7735,
+      "step": 48
+    },
+    {
+      "epoch": 0.1,
+      "learning_rate": 1.9999361478484043e-05,
+      "loss": 0.7979,
+      "step": 49
+    },
+    {
+      "epoch": 0.1,
+      "learning_rate": 1.9999080533322486e-05,
+      "loss": 0.7764,
+      "step": 50
+    },
+    {
+      "epoch": 0.11,
+      "learning_rate": 1.9998748510614615e-05,
+      "loss": 0.778,
+      "step": 51
+    },
+    {
+      "epoch": 0.11,
+      "learning_rate": 1.9998365412056476e-05,
+      "loss": 0.7824,
+      "step": 52
+    },
+    {
+      "epoch": 0.11,
+      "learning_rate": 1.999793123960503e-05,
+      "loss": 0.7659,
+      "step": 53
+    },
+    {
+      "epoch": 0.11,
+      "learning_rate": 1.999744599547812e-05,
+      "loss": 0.761,
+      "step": 54
+    },
+    {
+      "epoch": 0.12,
+      "learning_rate": 1.9996909682154487e-05,
+      "loss": 0.7694,
+      "step": 55
+    },
+    {
+      "epoch": 0.12,
+      "learning_rate": 1.999632230237373e-05,
+      "loss": 0.7806,
+      "step": 56
+    },
+    {
+      "epoch": 0.12,
+      "learning_rate": 1.999568385913633e-05,
+      "loss": 0.7704,
+      "step": 57
+    },
+    {
+      "epoch": 0.12,
+      "learning_rate": 1.999499435570359e-05,
+      "loss": 0.7873,
+      "step": 58
+    },
+    {
+      "epoch": 0.12,
+      "learning_rate": 1.999425379559765e-05,
+      "loss": 0.7592,
+      "step": 59
+    },
+    {
+      "epoch": 0.13,
+      "learning_rate": 1.999346218260146e-05,
+      "loss": 0.7697,
+      "step": 60
+    },
+    {
+      "epoch": 0.13,
+      "learning_rate": 1.999261952075875e-05,
+      "loss": 0.7655,
+      "step": 61
+    },
+    {
+      "epoch": 0.13,
+      "learning_rate": 1.999172581437403e-05,
+      "loss": 0.7805,
+      "step": 62
+    },
+    {
+      "epoch": 0.13,
+      "learning_rate": 1.9990781068012554e-05,
+      "loss": 0.7829,
+      "step": 63
+    },
+    {
+      "epoch": 0.13,
+      "learning_rate": 1.9989785286500294e-05,
+      "loss": 0.7738,
+      "step": 64
+    },
+    {
+      "epoch": 0.14,
+      "learning_rate": 1.998873847492393e-05,
+      "loss": 0.7671,
+      "step": 65
+    },
+    {
+      "epoch": 0.14,
+      "learning_rate": 1.9987640638630812e-05,
+      "loss": 0.7636,
+      "step": 66
+    },
+    {
+      "epoch": 0.14,
+      "learning_rate": 1.9986491783228925e-05,
+      "loss": 0.7738,
+      "step": 67
+    },
+    {
+      "epoch": 0.14,
+      "learning_rate": 1.998529191458689e-05,
+      "loss": 0.7546,
+      "step": 68
+    },
+    {
+      "epoch": 0.14,
+      "learning_rate": 1.99840410388339e-05,
+      "loss": 0.7745,
+      "step": 69
+    },
+    {
+      "epoch": 0.15,
+      "learning_rate": 1.9982739162359707e-05,
+      "loss": 0.7799,
+      "step": 70
+    },
+    {
+      "epoch": 0.15,
+      "learning_rate": 1.998138629181459e-05,
+      "loss": 0.757,
+      "step": 71
+    },
+    {
+      "epoch": 0.15,
+      "learning_rate": 1.997998243410932e-05,
+      "loss": 0.764,
+      "step": 72
+    },
+    {
+      "epoch": 0.15,
+      "learning_rate": 1.9978527596415108e-05,
+      "loss": 0.7727,
+      "step": 73
+    },
+    {
+      "epoch": 0.15,
+      "learning_rate": 1.99770217861636e-05,
+      "loss": 0.7546,
+      "step": 74
+    },
+    {
+      "epoch": 0.16,
+      "learning_rate": 1.9975465011046807e-05,
+      "loss": 0.7719,
+      "step": 75
+    },
+    {
+      "epoch": 0.16,
+      "learning_rate": 1.9973857279017092e-05,
+      "loss": 0.7743,
+      "step": 76
+    },
+    {
+      "epoch": 0.16,
+      "learning_rate": 1.99721985982871e-05,
+      "loss": 0.7633,
+      "step": 77
+    },
+    {
+      "epoch": 0.16,
+      "learning_rate": 1.9970488977329757e-05,
+      "loss": 0.7516,
+      "step": 78
+    },
+    {
+      "epoch": 0.17,
+      "learning_rate": 1.9968728424878178e-05,
+      "loss": 0.7624,
+      "step": 79
+    },
+    {
+      "epoch": 0.17,
+      "learning_rate": 1.996691694992567e-05,
+      "loss": 0.7474,
+      "step": 80
+    },
+    {
+      "epoch": 0.17,
+      "learning_rate": 1.9965054561725644e-05,
+      "loss": 0.7462,
+      "step": 81
+    },
+    {
+      "epoch": 0.17,
+      "learning_rate": 1.9963141269791606e-05,
+      "loss": 0.7602,
+      "step": 82
+    },
+    {
+      "epoch": 0.17,
+      "learning_rate": 1.9961177083897074e-05,
+      "loss": 0.7634,
+      "step": 83
+    },
+    {
+      "epoch": 0.18,
+      "learning_rate": 1.9959162014075553e-05,
+      "loss": 0.7625,
+      "step": 84
+    },
+    {
+      "epoch": 0.18,
+      "learning_rate": 1.995709607062047e-05,
+      "loss": 0.7415,
+      "step": 85
+    },
+    {
+      "epoch": 0.18,
+      "learning_rate": 1.995497926408513e-05,
+      "loss": 0.7613,
+      "step": 86
+    },
+    {
+      "epoch": 0.18,
+      "learning_rate": 1.9952811605282653e-05,
+      "loss": 0.764,
+      "step": 87
+    },
+    {
+      "epoch": 0.18,
+      "learning_rate": 1.9950593105285927e-05,
+      "loss": 0.7484,
+      "step": 88
+    },
+    {
+      "epoch": 0.19,
+      "learning_rate": 1.994832377542755e-05,
+      "loss": 0.759,
+      "step": 89
+    },
+    {
+      "epoch": 0.19,
+      "learning_rate": 1.9946003627299766e-05,
+      "loss": 0.7428,
+      "step": 90
+    },
+    {
+      "epoch": 0.19,
+      "learning_rate": 1.9943632672754408e-05,
+      "loss": 0.7441,
+      "step": 91
+    },
+    {
+      "epoch": 0.19,
+      "learning_rate": 1.9941210923902853e-05,
+      "loss": 0.7255,
+      "step": 92
+    },
+    {
+      "epoch": 0.19,
+      "learning_rate": 1.993873839311593e-05,
+      "loss": 0.7654,
+      "step": 93
+    },
+    {
+      "epoch": 0.2,
+      "learning_rate": 1.9936215093023884e-05,
+      "loss": 0.7543,
+      "step": 94
+    },
+    {
+      "epoch": 0.2,
+      "learning_rate": 1.99336410365163e-05,
+      "loss": 0.7573,
+      "step": 95
+    },
+    {
+      "epoch": 0.2,
+      "learning_rate": 1.9931016236742026e-05,
+      "loss": 0.741,
+      "step": 96
+    },
+    {
+      "epoch": 0.2,
+      "learning_rate": 1.992834070710914e-05,
+      "loss": 0.7459,
+      "step": 97
+    },
+    {
+      "epoch": 0.21,
+      "learning_rate": 1.992561446128484e-05,
+      "loss": 0.7668,
+      "step": 98
+    },
+    {
+      "epoch": 0.21,
+      "learning_rate": 1.9922837513195406e-05,
+      "loss": 0.7649,
+      "step": 99
+    },
+    {
+      "epoch": 0.21,
+      "learning_rate": 1.9920009877026106e-05,
+      "loss": 0.747,
+      "step": 100
+    },
+    {
+      "epoch": 0.21,
+      "learning_rate": 1.9917131567221147e-05,
+      "loss": 0.744,
+      "step": 101
+    },
+    {
+      "epoch": 0.21,
+      "learning_rate": 1.9914202598483576e-05,
+      "loss": 0.7382,
+      "step": 102
+    },
+    {
+      "epoch": 0.22,
+      "learning_rate": 1.9911222985775226e-05,
+      "loss": 0.7225,
+      "step": 103
+    },
+    {
+      "epoch": 0.22,
+      "learning_rate": 1.990819274431662e-05,
+      "loss": 0.7246,
+      "step": 104
+    },
+    {
+      "epoch": 0.22,
+      "learning_rate": 1.990511188958692e-05,
+      "loss": 0.7497,
+      "step": 105
+    },
+    {
+      "epoch": 0.22,
+      "learning_rate": 1.9901980437323818e-05,
+      "loss": 0.747,
+      "step": 106
+    },
+    {
+      "epoch": 0.22,
+      "learning_rate": 1.9898798403523475e-05,
+      "loss": 0.7321,
+      "step": 107
+    },
+    {
+      "epoch": 0.23,
+      "learning_rate": 1.9895565804440435e-05,
+      "loss": 0.7482,
+      "step": 108
+    },
+    {
+      "epoch": 0.23,
+      "learning_rate": 1.989228265658754e-05,
+      "loss": 0.7495,
+      "step": 109
+    },
+    {
+      "epoch": 0.23,
+      "learning_rate": 1.9888948976735843e-05,
+      "loss": 0.748,
+      "step": 110
+    },
+    {
+      "epoch": 0.23,
+      "learning_rate": 1.9885564781914528e-05,
+      "loss": 0.734,
+      "step": 111
+    },
+    {
+      "epoch": 0.23,
+      "learning_rate": 1.9882130089410822e-05,
+      "loss": 0.7519,
+      "step": 112
+    },
+    {
+      "epoch": 0.24,
+      "learning_rate": 1.987864491676991e-05,
+      "loss": 0.7432,
+      "step": 113
+    },
+    {
+      "epoch": 0.24,
+      "learning_rate": 1.9875109281794828e-05,
+      "loss": 0.7334,
+      "step": 114
+    },
+    {
+      "epoch": 0.24,
+      "learning_rate": 1.9871523202546395e-05,
+      "loss": 0.7445,
+      "step": 115
+    },
+    {
+      "epoch": 0.24,
+      "learning_rate": 1.986788669734311e-05,
+      "loss": 0.7301,
+      "step": 116
+    },
+    {
+      "epoch": 0.24,
+      "learning_rate": 1.986419978476106e-05,
+      "loss": 0.7491,
+      "step": 117
+    },
+    {
+      "epoch": 0.25,
+      "learning_rate": 1.986046248363381e-05,
+      "loss": 0.7429,
+      "step": 118
+    },
+    {
+      "epoch": 0.25,
+      "learning_rate": 1.9856674813052345e-05,
+      "loss": 0.7413,
+      "step": 119
+    },
+    {
+      "epoch": 0.25,
+      "learning_rate": 1.985283679236493e-05,
+      "loss": 0.7593,
+      "step": 120
+    },
+    {
+      "epoch": 0.25,
+      "learning_rate": 1.984894844117704e-05,
+      "loss": 0.7468,
+      "step": 121
+    },
+    {
+      "epoch": 0.26,
+      "learning_rate": 1.9845009779351235e-05,
+      "loss": 0.7476,
+      "step": 122
+    },
+    {
+      "epoch": 0.26,
+      "learning_rate": 1.9841020827007096e-05,
+      "loss": 0.752,
+      "step": 123
+    },
+    {
+      "epoch": 0.26,
+      "learning_rate": 1.9836981604521077e-05,
+      "loss": 0.7437,
+      "step": 124
+    },
+    {
+      "epoch": 0.26,
+      "learning_rate": 1.983289213252644e-05,
+      "loss": 0.7343,
+      "step": 125
+    },
+    {
+      "epoch": 0.26,
+      "learning_rate": 1.9828752431913116e-05,
+      "loss": 0.7393,
+      "step": 126
+    },
+    {
+      "epoch": 0.27,
+      "learning_rate": 1.9824562523827634e-05,
+      "loss": 0.7467,
+      "step": 127
+    },
+    {
+      "epoch": 0.27,
+      "learning_rate": 1.9820322429672978e-05,
+      "loss": 0.7313,
+      "step": 128
+    },
+    {
+      "epoch": 0.27,
+      "learning_rate": 1.98160321711085e-05,
+      "loss": 0.7424,
+      "step": 129
+    },
+    {
+      "epoch": 0.27,
+      "learning_rate": 1.9811691770049806e-05,
+      "loss": 0.7484,
+      "step": 130
+    },
+    {
+      "epoch": 0.27,
+      "learning_rate": 1.9807301248668638e-05,
+      "loss": 0.7341,
+      "step": 131
+    },
+    {
+      "epoch": 0.28,
+      "learning_rate": 1.9802860629392765e-05,
+      "loss": 0.718,
+      "step": 132
+    },
+    {
+      "epoch": 0.28,
+      "learning_rate": 1.979836993490586e-05,
+      "loss": 0.7383,
+      "step": 133
+    },
+    {
+      "epoch": 0.28,
+      "learning_rate": 1.9793829188147406e-05,
+      "loss": 0.7424,
+      "step": 134
+    },
+    {
+      "epoch": 0.28,
+      "learning_rate": 1.9789238412312553e-05,
+      "loss": 0.7311,
+      "step": 135
+    },
+    {
+      "epoch": 0.28,
+      "learning_rate": 1.9784597630852008e-05,
+      "loss": 0.7389,
+      "step": 136
+    },
+    {
+      "epoch": 0.29,
+      "learning_rate": 1.9779906867471937e-05,
+      "loss": 0.7473,
+      "step": 137
+    },
+    {
+      "epoch": 0.29,
+      "learning_rate": 1.97751661461338e-05,
+      "loss": 0.7188,
+      "step": 138
+    },
+    {
+      "epoch": 0.29,
+      "learning_rate": 1.9770375491054264e-05,
+      "loss": 0.7223,
+      "step": 139
+    },
+    {
+      "epoch": 0.29,
+      "learning_rate": 1.9765534926705082e-05,
+      "loss": 0.7408,
+      "step": 140
+    },
+    {
+      "epoch": 0.29,
+      "learning_rate": 1.9760644477812928e-05,
+      "loss": 0.7439,
+      "step": 141
+    },
+    {
+      "epoch": 0.3,
+      "learning_rate": 1.975570416935932e-05,
+      "loss": 0.7505,
+      "step": 142
+    },
+    {
+      "epoch": 0.3,
+      "learning_rate": 1.9750714026580465e-05,
+      "loss": 0.74,
+      "step": 143
+    },
+    {
+      "epoch": 0.3,
+      "learning_rate": 1.974567407496712e-05,
+      "loss": 0.7376,
+      "step": 144
+    },
+    {
+      "epoch": 0.3,
+      "learning_rate": 1.97405843402645e-05,
+      "loss": 0.7253,
+      "step": 145
+    },
+    {
+      "epoch": 0.31,
+      "learning_rate": 1.9735444848472108e-05,
+      "loss": 0.7321,
+      "step": 146
+    },
+    {
+      "epoch": 0.31,
+      "learning_rate": 1.973025562584362e-05,
+      "loss": 0.729,
+      "step": 147
+    },
+    {
+      "epoch": 0.31,
+      "learning_rate": 1.9725016698886748e-05,
+      "loss": 0.7354,
+      "step": 148
+    },
+    {
+      "epoch": 0.31,
+      "learning_rate": 1.9719728094363103e-05,
+      "loss": 0.7612,
+      "step": 149
+    },
+    {
+      "epoch": 0.31,
+      "learning_rate": 1.9714389839288073e-05,
+      "loss": 0.7162,
+      "step": 150
+    },
+    {
+      "epoch": 0.32,
+      "learning_rate": 1.9709001960930655e-05,
+      "loss": 0.7172,
+      "step": 151
+    },
+    {
+      "epoch": 0.32,
+      "learning_rate": 1.9703564486813342e-05,
+      "loss": 0.7298,
+      "step": 152
+    },
+    {
+      "epoch": 0.32,
+      "learning_rate": 1.9698077444711973e-05,
+      "loss": 0.7428,
+      "step": 153
+    },
+    {
+      "epoch": 0.32,
+      "learning_rate": 1.9692540862655587e-05,
+      "loss": 0.7398,
+      "step": 154
+    },
+    {
+      "epoch": 0.32,
+      "learning_rate": 1.9686954768926288e-05,
+      "loss": 0.7346,
+      "step": 155
+    },
+    {
+      "epoch": 0.33,
+      "learning_rate": 1.96813191920591e-05,
+      "loss": 0.7289,
+      "step": 156
+    },
+    {
+      "epoch": 0.33,
+      "learning_rate": 1.9675634160841807e-05,
+      "loss": 0.7331,
+      "step": 157
+    },
+    {
+      "epoch": 0.33,
+      "learning_rate": 1.9669899704314828e-05,
+      "loss": 0.7309,
+      "step": 158
+    },
+    {
+      "epoch": 0.33,
+      "learning_rate": 1.966411585177105e-05,
+      "loss": 0.7285,
+      "step": 159
+    },
+    {
+      "epoch": 0.33,
+      "learning_rate": 1.9658282632755694e-05,
+      "loss": 0.7305,
+      "step": 160
+    },
+    {
+      "epoch": 0.34,
+      "learning_rate": 1.9652400077066146e-05,
+      "loss": 0.7214,
+      "step": 161
+    },
+    {
+      "epoch": 0.34,
+      "learning_rate": 1.964646821475183e-05,
+      "loss": 0.7282,
+      "step": 162
+    },
+    {
+      "epoch": 0.34,
+      "learning_rate": 1.9640487076114012e-05,
+      "loss": 0.7328,
+      "step": 163
+    },
+    {
+      "epoch": 0.34,
+      "learning_rate": 1.9634456691705705e-05,
+      "loss": 0.7354,
+      "step": 164
+    },
+    {
+      "epoch": 0.35,
+      "learning_rate": 1.962837709233146e-05,
+      "loss": 0.7267,
+      "step": 165
+    },
+    {
+      "epoch": 0.35,
+      "learning_rate": 1.9622248309047233e-05,
+      "loss": 0.7172,
+      "step": 166
+    },
+    {
+      "epoch": 0.35,
+      "learning_rate": 1.9616070373160224e-05,
+      "loss": 0.7389,
+      "step": 167
+    },
+    {
+      "epoch": 0.35,
+      "learning_rate": 1.960984331622872e-05,
+      "loss": 0.7414,
+      "step": 168
+    },
+    {
+      "epoch": 0.35,
+      "learning_rate": 1.9603567170061918e-05,
+      "loss": 0.7029,
+      "step": 169
+    },
+    {
+      "epoch": 0.36,
+      "learning_rate": 1.959724196671978e-05,
+      "loss": 0.7262,
+      "step": 170
+    },
+    {
+      "epoch": 0.36,
+      "learning_rate": 1.9590867738512874e-05,
+      "loss": 0.7221,
+      "step": 171
+    },
+    {
+      "epoch": 0.36,
+      "learning_rate": 1.9584444518002178e-05,
+      "loss": 0.7217,
+      "step": 172
+    },
+    {
+      "epoch": 0.36,
+      "learning_rate": 1.9577972337998944e-05,
+      "loss": 0.7225,
+      "step": 173
+    },
+    {
+      "epoch": 0.36,
+      "learning_rate": 1.9571451231564523e-05,
+      "loss": 0.7284,
+      "step": 174
+    },
+    {
+      "epoch": 0.37,
+      "learning_rate": 1.956488123201019e-05,
+      "loss": 0.7369,
+      "step": 175
+    },
+    {
+      "epoch": 0.37,
+      "learning_rate": 1.955826237289697e-05,
+      "loss": 0.7242,
+      "step": 176
+    },
+    {
+      "epoch": 0.37,
+      "learning_rate": 1.9551594688035486e-05,
+      "loss": 0.7164,
+      "step": 177
+    },
+    {
+      "epoch": 0.37,
+      "learning_rate": 1.9544878211485763e-05,
+      "loss": 0.735,
+      "step": 178
+    },
+    {
+      "epoch": 0.37,
+      "learning_rate": 1.9538112977557077e-05,
+      "loss": 0.7408,
+      "step": 179
+    },
+    {
+      "epoch": 0.38,
+      "learning_rate": 1.9531299020807752e-05,
+      "loss": 0.7292,
+      "step": 180
+    },
+    {
+      "epoch": 0.38,
+      "learning_rate": 1.9524436376045014e-05,
+      "loss": 0.7254,
+      "step": 181
+    },
+    {
+      "epoch": 0.38,
+      "learning_rate": 1.9517525078324787e-05,
+      "loss": 0.7063,
+      "step": 182
+    },
+    {
+      "epoch": 0.38,
+      "learning_rate": 1.9510565162951538e-05,
+      "loss": 0.7231,
+      "step": 183
+    },
+    {
+      "epoch": 0.38,
+      "learning_rate": 1.9503556665478066e-05,
+      "loss": 0.7298,
+      "step": 184
+    },
+    {
+      "epoch": 0.39,
+      "learning_rate": 1.9496499621705357e-05,
+      "loss": 0.7211,
+      "step": 185
+    },
+    {
+      "epoch": 0.39,
+      "learning_rate": 1.9489394067682365e-05,
+      "loss": 0.7225,
+      "step": 186
+    },
+    {
+      "epoch": 0.39,
+      "learning_rate": 1.948224003970586e-05,
+      "loss": 0.7186,
+      "step": 187
+    },
+    {
+      "epoch": 0.39,
+      "learning_rate": 1.9475037574320217e-05,
+      "loss": 0.7036,
+      "step": 188
+    },
+    {
+      "epoch": 0.4,
+      "learning_rate": 1.9467786708317257e-05,
+      "loss": 0.7264,
+      "step": 189
+    },
+    {
+      "epoch": 0.4,
+      "learning_rate": 1.946048747873601e-05,
+      "loss": 0.7261,
+      "step": 190
+    },
+    {
+      "epoch": 0.4,
+      "learning_rate": 1.9453139922862595e-05,
+      "loss": 0.7326,
+      "step": 191
+    },
+    {
+      "epoch": 0.4,
+      "learning_rate": 1.9445744078229967e-05,
+      "loss": 0.7296,
+      "step": 192
+    },
+    {
+      "epoch": 0.4,
+      "learning_rate": 1.9438299982617766e-05,
+      "loss": 0.7007,
+      "step": 193
+    },
+    {
+      "epoch": 0.41,
+      "learning_rate": 1.9430807674052092e-05,
+      "loss": 0.7248,
+      "step": 194
+    },
+    {
+      "epoch": 0.41,
+      "learning_rate": 1.9423267190805346e-05,
+      "loss": 0.7255,
+      "step": 195
+    },
+    {
+      "epoch": 0.41,
+      "learning_rate": 1.9415678571396006e-05,
+      "loss": 0.7142,
+      "step": 196
+    },
+    {
+      "epoch": 0.41,
+      "learning_rate": 1.940804185458844e-05,
+      "loss": 0.7283,
+      "step": 197
+    },
+    {
+      "epoch": 0.41,
+      "learning_rate": 1.9400357079392714e-05,
+      "loss": 0.7171,
+      "step": 198
+    },
+    {
+      "epoch": 0.42,
+      "learning_rate": 1.939262428506438e-05,
+      "loss": 0.7255,
+      "step": 199
+    },
+    {
+      "epoch": 0.42,
+      "learning_rate": 1.9384843511104294e-05,
+      "loss": 0.719,
+      "step": 200
+    },
+    {
+      "epoch": 0.42,
+      "learning_rate": 1.9377014797258384e-05,
+      "loss": 0.739,
+      "step": 201
+    },
+    {
+      "epoch": 0.42,
+      "learning_rate": 1.936913818351748e-05,
+      "loss": 0.747,
+      "step": 202
+    },
+    {
+      "epoch": 0.42,
+      "learning_rate": 1.9361213710117094e-05,
+      "loss": 0.7137,
+      "step": 203
+    },
+    {
+      "epoch": 0.43,
+      "learning_rate": 1.9353241417537216e-05,
+      "loss": 0.7109,
+      "step": 204
+    },
+    {
+      "epoch": 0.43,
+      "learning_rate": 1.9345221346502093e-05,
+      "loss": 0.7254,
+      "step": 205
+    },
+    {
+      "epoch": 0.43,
+      "learning_rate": 1.933715353798006e-05,
+      "loss": 0.7196,
+      "step": 206
+    },
+    {
+      "epoch": 0.43,
+      "learning_rate": 1.932903803318329e-05,
+      "loss": 0.7315,
+      "step": 207
+    },
+    {
+      "epoch": 0.44,
+      "learning_rate": 1.9320874873567598e-05,
+      "loss": 0.7175,
+      "step": 208
+    },
+    {
+      "epoch": 0.44,
+      "learning_rate": 1.9312664100832236e-05,
+      "loss": 0.7311,
+      "step": 209
+    },
+    {
+      "epoch": 0.44,
+      "learning_rate": 1.930440575691967e-05,
+      "loss": 0.7278,
+      "step": 210
+    },
+    {
+      "epoch": 0.44,
+      "learning_rate": 1.9296099884015382e-05,
+      "loss": 0.7381,
+      "step": 211
+    },
+    {
+      "epoch": 0.44,
+      "learning_rate": 1.9287746524547627e-05,
+      "loss": 0.7282,
+      "step": 212
+    },
+    {
+      "epoch": 0.45,
+      "learning_rate": 1.9279345721187236e-05,
+      "loss": 0.7259,
+      "step": 213
+    },
+    {
+      "epoch": 0.45,
+      "learning_rate": 1.9270897516847406e-05,
+      "loss": 0.7164,
+      "step": 214
+    },
+    {
+      "epoch": 0.45,
+      "learning_rate": 1.9262401954683452e-05,
+      "loss": 0.7314,
+      "step": 215
+    },
+    {
+      "epoch": 0.45,
+      "learning_rate": 1.9253859078092616e-05,
+      "loss": 0.7221,
+      "step": 216
+    },
+    {
+      "epoch": 0.45,
+      "learning_rate": 1.9245268930713825e-05,
+      "loss": 0.7225,
+      "step": 217
+    },
+    {
+      "epoch": 0.46,
+      "learning_rate": 1.923663155642748e-05,
+      "loss": 0.706,
+      "step": 218
+    },
+    {
+      "epoch": 0.46,
+      "learning_rate": 1.9227946999355226e-05,
+      "loss": 0.725,
+      "step": 219
+    },
+    {
+      "epoch": 0.46,
+      "learning_rate": 1.9219215303859732e-05,
+      "loss": 0.7158,
+      "step": 220
+    },
+    {
+      "epoch": 0.46,
+      "learning_rate": 1.921043651454445e-05,
+      "loss": 0.7271,
+      "step": 221
+    },
+    {
+      "epoch": 0.46,
+      "learning_rate": 1.9201610676253412e-05,
+      "loss": 0.7144,
+      "step": 222
+    },
+    {
+      "epoch": 0.47,
+      "learning_rate": 1.919273783407098e-05,
+      "loss": 0.7284,
+      "step": 223
+    },
+    {
+      "epoch": 0.47,
+      "learning_rate": 1.9183818033321612e-05,
+      "loss": 0.7191,
+      "step": 224
+    },
+    {
+      "epoch": 0.47,
+      "learning_rate": 1.9174851319569662e-05,
+      "loss": 0.718,
+      "step": 225
+    },
+    {
+      "epoch": 0.47,
+      "learning_rate": 1.916583773861911e-05,
+      "loss": 0.7289,
+      "step": 226
+    },
+    {
+      "epoch": 0.47,
+      "learning_rate": 1.9156777336513352e-05,
+      "loss": 0.7219,
+      "step": 227
+    },
+    {
+      "epoch": 0.48,
+      "learning_rate": 1.9147670159534953e-05,
+      "loss": 0.719,
+      "step": 228
+    },
+    {
+      "epoch": 0.48,
+      "learning_rate": 1.9138516254205416e-05,
+      "loss": 0.7049,
+      "step": 229
+    },
+    {
+      "epoch": 0.48,
+      "learning_rate": 1.912931566728494e-05,
+      "loss": 0.7068,
+      "step": 230
+    },
+    {
+      "epoch": 0.48,
+      "learning_rate": 1.9120068445772192e-05,
+      "loss": 0.7203,
+      "step": 231
+    },
+    {
+      "epoch": 0.49,
+      "learning_rate": 1.9110774636904052e-05,
+      "loss": 0.7264,
+      "step": 232
+    },
+    {
+      "epoch": 0.49,
+      "learning_rate": 1.910143428815538e-05,
+      "loss": 0.7209,
+      "step": 233
+    },
+    {
+      "epoch": 0.49,
+      "learning_rate": 1.9092047447238775e-05,
+      "loss": 0.713,
+      "step": 234
+    },
+    {
+      "epoch": 0.49,
+      "learning_rate": 1.9082614162104316e-05,
+      "loss": 0.7095,
+      "step": 235
+    },
+    {
+      "epoch": 0.49,
+      "learning_rate": 1.9073134480939353e-05,
+      "loss": 0.7091,
+      "step": 236
+    },
+    {
+      "epoch": 0.5,
+      "learning_rate": 1.9063608452168214e-05,
+      "loss": 0.7291,
+      "step": 237
+    },
+    {
+      "epoch": 0.5,
+      "learning_rate": 1.9054036124452e-05,
+      "loss": 0.7027,
+      "step": 238
+    },
+    {
+      "epoch": 0.5,
+      "learning_rate": 1.9044417546688295e-05,
+      "loss": 0.6992,
+      "step": 239
+    },
+    {
+      "epoch": 0.5,
+      "learning_rate": 1.9034752768010965e-05,
+      "loss": 0.7173,
+      "step": 240
+    },
+    {
+      "epoch": 0.5,
+      "learning_rate": 1.902504183778986e-05,
+      "loss": 0.7219,
+      "step": 241
+    },
+    {
+      "epoch": 0.51,
+      "learning_rate": 1.901528480563059e-05,
+      "loss": 0.7362,
+      "step": 242
+    },
+    {
+      "epoch": 0.51,
+      "learning_rate": 1.9005481721374272e-05,
+      "loss": 0.7022,
+      "step": 243
+    },
+    {
+      "epoch": 0.51,
+      "learning_rate": 1.899563263509725e-05,
+      "loss": 0.7205,
+      "step": 244
+    },
+    {
+      "epoch": 0.51,
+      "learning_rate": 1.898573759711087e-05,
+      "loss": 0.7173,
+      "step": 245
+    },
+    {
+      "epoch": 0.51,
+      "learning_rate": 1.89757966579612e-05,
+      "loss": 0.7027,
+      "step": 246
+    },
+    {
+      "epoch": 0.52,
+      "learning_rate": 1.8965809868428798e-05,
+      "loss": 0.7157,
+      "step": 247
+    },
+    {
+      "epoch": 0.52,
+      "learning_rate": 1.8955777279528414e-05,
+      "loss": 0.7295,
+      "step": 248
+    },
+    {
+      "epoch": 0.52,
+      "learning_rate": 1.894569894250877e-05,
+      "loss": 0.7103,
+      "step": 249
+    },
+    {
+      "epoch": 0.52,
+      "learning_rate": 1.8935574908852272e-05,
+      "loss": 0.7241,
+      "step": 250
+    },
+    {
+      "epoch": 0.53,
+      "learning_rate": 1.892540523027475e-05,
+      "loss": 0.7181,
+      "step": 251
+    },
+    {
+      "epoch": 0.53,
+      "learning_rate": 1.8915189958725207e-05,
+      "loss": 0.7303,
+      "step": 252
+    },
+    {
+      "epoch": 0.53,
+      "learning_rate": 1.890492914638554e-05,
+      "loss": 0.711,
+      "step": 253
+    },
+    {
+      "epoch": 0.53,
+      "learning_rate": 1.8894622845670282e-05,
+      "loss": 0.7165,
+      "step": 254
+    },
+    {
+      "epoch": 0.53,
+      "learning_rate": 1.8884271109226327e-05,
+      "loss": 0.7254,
+      "step": 255
+    },
+    {
+      "epoch": 0.54,
+      "learning_rate": 1.8873873989932666e-05,
+      "loss": 0.7035,
+      "step": 256
+    },
+    {
+      "epoch": 0.54,
+      "learning_rate": 1.886343154090012e-05,
+      "loss": 0.7162,
+      "step": 257
+    },
+    {
+      "epoch": 0.54,
+      "learning_rate": 1.8852943815471058e-05,
+      "loss": 0.7178,
+      "step": 258
+    },
+    {
+      "epoch": 0.54,
+      "learning_rate": 1.8842410867219137e-05,
+      "loss": 0.7356,
+      "step": 259
+    },
+    {
+      "epoch": 0.54,
+      "learning_rate": 1.8831832749949015e-05,
+      "loss": 0.6981,
+      "step": 260
+    },
+    {
+      "epoch": 0.55,
+      "learning_rate": 1.882120951769609e-05,
+      "loss": 0.7178,
+      "step": 261
+    },
+    {
+      "epoch": 0.55,
+      "learning_rate": 1.8810541224726217e-05,
+      "loss": 0.7156,
+      "step": 262
+    },
+    {
+      "epoch": 0.55,
+      "learning_rate": 1.879982792553543e-05,
+      "loss": 0.7276,
+      "step": 263
+    },
+    {
+      "epoch": 0.55,
+      "learning_rate": 1.878906967484966e-05,
+      "loss": 0.7329,
+      "step": 264
+    },
+    {
+      "epoch": 0.55,
+      "learning_rate": 1.8778266527624466e-05,
+      "loss": 0.7136,
+      "step": 265
+    },
+    {
+      "epoch": 0.56,
+      "learning_rate": 1.8767418539044753e-05,
+      "loss": 0.6963,
+      "step": 266
+    },
+    {
+      "epoch": 0.56,
+      "learning_rate": 1.8756525764524475e-05,
+      "loss": 0.7085,
+      "step": 267
+    },
+    {
+      "epoch": 0.56,
+      "learning_rate": 1.8745588259706366e-05,
+      "loss": 0.7139,
+      "step": 268
+    },
+    {
+      "epoch": 0.56,
+      "learning_rate": 1.8734606080461657e-05,
+      "loss": 0.6993,
+      "step": 269
+    },
+    {
+      "epoch": 0.56,
+      "learning_rate": 1.8723579282889784e-05,
+      "loss": 0.6875,
+      "step": 270
+    },
+    {
+      "epoch": 0.57,
+      "learning_rate": 1.87125079233181e-05,
+      "loss": 0.7082,
+      "step": 271
+    },
+    {
+      "epoch": 0.57,
+      "learning_rate": 1.8701392058301595e-05,
+      "loss": 0.7173,
+      "step": 272
+    },
+    {
+      "epoch": 0.57,
+      "learning_rate": 1.86902317446226e-05,
+      "loss": 0.7096,
+      "step": 273
+    },
+    {
+      "epoch": 0.57,
+      "learning_rate": 1.86790270392905e-05,
+      "loss": 0.7285,
+      "step": 274
+    },
+    {
+      "epoch": 0.58,
+      "learning_rate": 1.8667777999541444e-05,
+      "loss": 0.7212,
+      "step": 275
+    },
+    {
+      "epoch": 0.58,
+      "learning_rate": 1.865648468283805e-05,
+      "loss": 0.7229,
+      "step": 276
+    },
+    {
+      "epoch": 0.58,
+      "learning_rate": 1.8645147146869114e-05,
+      "loss": 0.7159,
+      "step": 277
+    },
+    {
+      "epoch": 0.58,
+      "learning_rate": 1.863376544954931e-05,
+      "loss": 0.6971,
+      "step": 278
+    },
+    {
+      "epoch": 0.58,
+      "learning_rate": 1.8622339649018907e-05,
+      "loss": 0.7014,
+      "step": 279
+    },
+    {
+      "epoch": 0.59,
+      "learning_rate": 1.8610869803643454e-05,
+      "loss": 0.7148,
+      "step": 280
+    },
+    {
+      "epoch": 0.59,
+      "learning_rate": 1.8599355972013493e-05,
+      "loss": 0.7196,
+      "step": 281
+    },
+    {
+      "epoch": 0.59,
+      "learning_rate": 1.8587798212944255e-05,
+      "loss": 0.7228,
+      "step": 282
+    },
+    {
+      "epoch": 0.59,
+      "learning_rate": 1.8576196585475376e-05,
+      "loss": 0.6995,
+      "step": 283
+    },
+    {
+      "epoch": 0.59,
+      "learning_rate": 1.856455114887056e-05,
+      "loss": 0.7173,
+      "step": 284
+    },
+    {
+      "epoch": 0.6,
+      "learning_rate": 1.8552861962617318e-05,
+      "loss": 0.7003,
+      "step": 285
+    },
+    {
+      "epoch": 0.6,
+      "learning_rate": 1.854112908642663e-05,
+      "loss": 0.7225,
+      "step": 286
+    },
+    {
+      "epoch": 0.6,
+      "learning_rate": 1.8529352580232668e-05,
+      "loss": 0.7159,
+      "step": 287
+    },
+    {
+      "epoch": 0.6,
+      "learning_rate": 1.8517532504192456e-05,
+      "loss": 0.705,
+      "step": 288
+    },
+    {
+      "epoch": 0.6,
+      "learning_rate": 1.8505668918685603e-05,
+      "loss": 0.7108,
+      "step": 289
+    },
+    {
+      "epoch": 0.61,
+      "learning_rate": 1.849376188431396e-05,
+      "loss": 0.716,
+      "step": 290
+    },
+    {
+      "epoch": 0.61,
+      "learning_rate": 1.8481811461901323e-05,
+      "loss": 0.7201,
+      "step": 291
+    },
+    {
+      "epoch": 0.61,
+      "learning_rate": 1.8469817712493148e-05,
+      "loss": 0.7089,
+      "step": 292
+    },
+    {
+      "epoch": 0.61,
+      "learning_rate": 1.845778069735618e-05,
+      "loss": 0.7112,
+      "step": 293
+    },
+    {
+      "epoch": 0.62,
+      "learning_rate": 1.8445700477978207e-05,
+      "loss": 0.6926,
+      "step": 294
+    },
+    {
+      "epoch": 0.62,
+      "learning_rate": 1.8433577116067684e-05,
+      "loss": 0.6866,
+      "step": 295
+    },
+    {
+      "epoch": 0.62,
+      "learning_rate": 1.8421410673553475e-05,
+      "loss": 0.6995,
+      "step": 296
+    },
+    {
+      "epoch": 0.62,
+      "learning_rate": 1.8409201212584493e-05,
+      "loss": 0.7087,
+      "step": 297
+    },
+    {
+      "epoch": 0.62,
+      "learning_rate": 1.8396948795529405e-05,
+      "loss": 0.7152,
+      "step": 298
+    },
+    {
+      "epoch": 0.63,
+      "learning_rate": 1.8384653484976305e-05,
+      "loss": 0.6933,
+      "step": 299
+    },
+    {
+      "epoch": 0.63,
+      "learning_rate": 1.8372315343732395e-05,
+      "loss": 0.7159,
+      "step": 300
+    },
+    {
+      "epoch": 0.63,
+      "learning_rate": 1.8359934434823672e-05,
+      "loss": 0.6924,
+      "step": 301
+    },
+    {
+      "epoch": 0.63,
+      "learning_rate": 1.8347510821494593e-05,
+      "loss": 0.7104,
+      "step": 302
+    },
+    {
+      "epoch": 0.63,
+      "learning_rate": 1.8335044567207763e-05,
+      "loss": 0.7067,
+      "step": 303
+    },
+    {
+      "epoch": 0.64,
+      "learning_rate": 1.8322535735643604e-05,
+      "loss": 0.6926,
+      "step": 304
+    },
+    {
+      "epoch": 0.64,
+      "learning_rate": 1.8309984390700038e-05,
+      "loss": 0.7108,
+      "step": 305
+    },
+    {
+      "epoch": 0.64,
+      "learning_rate": 1.8297390596492143e-05,
+      "loss": 0.7126,
+      "step": 306
+    },
+    {
+      "epoch": 0.64,
+      "learning_rate": 1.828475441735185e-05,
+      "loss": 0.6864,
+      "step": 307
+    },
+    {
+      "epoch": 0.64,
+      "learning_rate": 1.8272075917827597e-05,
+      "loss": 0.7076,
+      "step": 308
+    },
+    {
+      "epoch": 0.65,
+      "learning_rate": 1.8259355162684e-05,
+      "loss": 0.7303,
+      "step": 309
+    },
+    {
+      "epoch": 0.65,
+      "learning_rate": 1.824659221690153e-05,
+      "loss": 0.7014,
+      "step": 310
+    },
+    {
+      "epoch": 0.65,
+      "learning_rate": 1.823378714567618e-05,
+      "loss": 0.7113,
+      "step": 311
+    },
+    {
+      "epoch": 0.65,
+      "learning_rate": 1.822094001441913e-05,
+      "loss": 0.7065,
+      "step": 312
+    },
+    {
+      "epoch": 0.65,
+      "learning_rate": 1.82080508887564e-05,
+      "loss": 0.7064,
+      "step": 313
+    },
+    {
+      "epoch": 0.66,
+      "learning_rate": 1.8195119834528535e-05,
+      "loss": 0.7077,
+      "step": 314
+    },
+    {
+      "epoch": 0.66,
+      "learning_rate": 1.8182146917790264e-05,
+      "loss": 0.7021,
+      "step": 315
+    },
+    {
+      "epoch": 0.66,
+      "learning_rate": 1.8169132204810157e-05,
+      "loss": 0.6888,
+      "step": 316
+    },
+    {
+      "epoch": 0.66,
+      "learning_rate": 1.815607576207029e-05,
+      "loss": 0.6907,
+      "step": 317
+    },
+    {
+      "epoch": 0.67,
+      "learning_rate": 1.814297765626589e-05,
+      "loss": 0.7016,
+      "step": 318
+    },
+    {
+      "epoch": 0.67,
+      "learning_rate": 1.8129837954305033e-05,
+      "loss": 0.7091,
+      "step": 319
+    },
+    {
+      "epoch": 0.67,
+      "learning_rate": 1.8116656723308253e-05,
+      "loss": 0.683,
+      "step": 320
+    },
+    {
+      "epoch": 0.67,
+      "learning_rate": 1.810343403060824e-05,
+      "loss": 0.6885,
+      "step": 321
+    },
+    {
+      "epoch": 0.67,
+      "learning_rate": 1.8090169943749477e-05,
+      "loss": 0.708,
+      "step": 322
+    },
+    {
+      "epoch": 0.68,
+      "learning_rate": 1.8076864530487886e-05,
+      "loss": 0.7212,
+      "step": 323
+    },
+    {
+      "epoch": 0.68,
+      "learning_rate": 1.8063517858790517e-05,
+      "loss": 0.6991,
+      "step": 324
+    },
+    {
+      "epoch": 0.68,
+      "learning_rate": 1.8050129996835147e-05,
+      "loss": 0.7052,
+      "step": 325
+    },
+    {
+      "epoch": 0.68,
+      "learning_rate": 1.8036701013009988e-05,
+      "loss": 0.6902,
+      "step": 326
+    },
+    {
+      "epoch": 0.68,
+      "learning_rate": 1.80232309759133e-05,
+      "loss": 0.6952,
+      "step": 327
+    },
+    {
+      "epoch": 0.69,
+      "learning_rate": 1.800971995435305e-05,
+      "loss": 0.7168,
+      "step": 328
+    },
+    {
+      "epoch": 0.69,
+      "learning_rate": 1.799616801734657e-05,
+      "loss": 0.7164,
+      "step": 329
+    },
+    {
+      "epoch": 0.69,
+      "learning_rate": 1.7982575234120196e-05,
+      "loss": 0.6915,
+      "step": 330
+    },
+    {
+      "epoch": 0.69,
+      "learning_rate": 1.796894167410891e-05,
+      "loss": 0.7078,
+      "step": 331
+    },
+    {
+      "epoch": 0.69,
+      "learning_rate": 1.7955267406955997e-05,
+      "loss": 0.7046,
+      "step": 332
+    },
+    {
+      "epoch": 0.7,
+      "learning_rate": 1.7941552502512684e-05,
+      "loss": 0.711,
+      "step": 333
+    },
+    {
+      "epoch": 0.7,
+      "learning_rate": 1.792779703083777e-05,
+      "loss": 0.697,
+      "step": 334
+    },
+    {
+      "epoch": 0.7,
+      "learning_rate": 1.7914001062197298e-05,
+      "loss": 0.6866,
+      "step": 335
+    },
+    {
+      "epoch": 0.7,
+      "learning_rate": 1.790016466706417e-05,
+      "loss": 0.6973,
+      "step": 336
+    },
+    {
+      "epoch": 0.71,
+      "learning_rate": 1.788628791611779e-05,
+      "loss": 0.6973,
+      "step": 337
+    },
+    {
+      "epoch": 0.71,
+      "learning_rate": 1.787237088024372e-05,
+      "loss": 0.7045,
+      "step": 338
+    },
+    {
+      "epoch": 0.71,
+      "learning_rate": 1.7858413630533305e-05,
+      "loss": 0.7087,
+      "step": 339
+    },
+    {
+      "epoch": 0.71,
+      "learning_rate": 1.78444162382833e-05,
+      "loss": 0.7013,
+      "step": 340
+    },
+    {
+      "epoch": 0.71,
+      "learning_rate": 1.7830378774995524e-05,
+      "loss": 0.6973,
+      "step": 341
+    },
+    {
+      "epoch": 0.72,
+      "learning_rate": 1.781630131237649e-05,
+      "loss": 0.6941,
+      "step": 342
+    },
+    {
+      "epoch": 0.72,
+      "learning_rate": 1.780218392233704e-05,
+      "loss": 0.7032,
+      "step": 343
+    },
+    {
+      "epoch": 0.72,
+      "learning_rate": 1.778802667699196e-05,
+      "loss": 0.712,
+      "step": 344
+    },
+    {
+      "epoch": 0.72,
+      "learning_rate": 1.7773829648659645e-05,
+      "loss": 0.6927,
+      "step": 345
+    },
+    {
+      "epoch": 0.72,
+      "learning_rate": 1.7759592909861694e-05,
+      "loss": 0.6865,
+      "step": 346
+    },
+    {
+      "epoch": 0.73,
+      "learning_rate": 1.774531653332256e-05,
+      "loss": 0.6721,
+      "step": 347
+    },
+    {
+      "epoch": 0.73,
+      "learning_rate": 1.7731000591969182e-05,
+      "loss": 0.7076,
+      "step": 348
+    },
+    {
+      "epoch": 0.73,
+      "learning_rate": 1.77166451589306e-05,
+      "loss": 0.7061,
+      "step": 349
+    },
+    {
+      "epoch": 0.73,
+      "learning_rate": 1.7702250307537583e-05,
+      "loss": 0.7087,
+      "step": 350
+    },
+    {
+      "epoch": 0.73,
+      "learning_rate": 1.768781611132226e-05,
+      "loss": 0.7046,
+      "step": 351
+    },
+    {
+      "epoch": 0.74,
+      "learning_rate": 1.7673342644017744e-05,
+      "loss": 0.7029,
+      "step": 352
+    },
+    {
+      "epoch": 0.74,
+      "learning_rate": 1.7658829979557757e-05,
+      "loss": 0.7102,
+      "step": 353
+    },
+    {
+      "epoch": 0.74,
+      "learning_rate": 1.764427819207624e-05,
+      "loss": 0.7063,
+      "step": 354
+    },
+    {
+      "epoch": 0.74,
+      "learning_rate": 1.7629687355906988e-05,
+      "loss": 0.6969,
+      "step": 355
+    },
+    {
+      "epoch": 0.74,
+      "learning_rate": 1.761505754558327e-05,
+      "loss": 0.6944,
+      "step": 356
+    },
+    {
+      "epoch": 0.75,
+      "learning_rate": 1.7600388835837427e-05,
+      "loss": 0.6825,
+      "step": 357
+    },
+    {
+      "epoch": 0.75,
+      "learning_rate": 1.758568130160053e-05,
+      "loss": 0.6947,
+      "step": 358
+    },
+    {
+      "epoch": 0.75,
+      "learning_rate": 1.757093501800196e-05,
+      "loss": 0.7019,
+      "step": 359
+    },
+    {
+      "epoch": 0.75,
+      "learning_rate": 1.755615006036904e-05,
+      "loss": 0.6927,
+      "step": 360
+    },
+    {
+      "epoch": 0.76,
+      "learning_rate": 1.7541326504226654e-05,
+      "loss": 0.7018,
+      "step": 361
+    },
+    {
+      "epoch": 0.76,
+      "learning_rate": 1.7526464425296846e-05,
+      "loss": 0.6907,
+      "step": 362
+    },
+    {
+      "epoch": 0.76,
+      "learning_rate": 1.7511563899498455e-05,
+      "loss": 0.7156,
+      "step": 363
+    },
+    {
+      "epoch": 0.76,
+      "learning_rate": 1.7496625002946702e-05,
+      "loss": 0.6977,
+      "step": 364
+    },
+    {
+      "epoch": 0.76,
+      "learning_rate": 1.7481647811952828e-05,
+      "loss": 0.6857,
+      "step": 365
+    },
+    {
+      "epoch": 0.77,
+      "learning_rate": 1.746663240302368e-05,
+      "loss": 0.6984,
+      "step": 366
+    },
+    {
+      "epoch": 0.77,
+      "learning_rate": 1.7451578852861337e-05,
+      "loss": 0.7048,
+      "step": 367
+    },
+    {
+      "epoch": 0.77,
+      "learning_rate": 1.743648723836271e-05,
+      "loss": 0.6897,
+      "step": 368
+    },
+    {
+      "epoch": 0.77,
+      "learning_rate": 1.7421357636619153e-05,
+      "loss": 0.6938,
+      "step": 369
+    },
+    {
+      "epoch": 0.77,
+      "learning_rate": 1.7406190124916064e-05,
+      "loss": 0.69,
+      "step": 370
+    },
+    {
+      "epoch": 0.78,
+      "learning_rate": 1.7390984780732493e-05,
+      "loss": 0.697,
+      "step": 371
+    },
+    {
+      "epoch": 0.78,
+      "learning_rate": 1.737574168174075e-05,
+      "loss": 0.7051,
+      "step": 372
+    },
+    {
+      "epoch": 0.78,
+      "learning_rate": 1.7360460905806004e-05,
+      "loss": 0.6941,
+      "step": 373
+    },
+    {
+      "epoch": 0.78,
+      "learning_rate": 1.734514253098589e-05,
+      "loss": 0.703,
+      "step": 374
+    },
+    {
+      "epoch": 0.78,
+      "learning_rate": 1.7329786635530094e-05,
+      "loss": 0.6968,
+      "step": 375
+    },
+    {
+      "epoch": 0.79,
+      "learning_rate": 1.7314393297879982e-05,
+      "loss": 0.7027,
+      "step": 376
+    },
+    {
+      "epoch": 0.79,
+      "learning_rate": 1.7298962596668176e-05,
+      "loss": 0.7058,
+      "step": 377
+    },
+    {
+      "epoch": 0.79,
+      "learning_rate": 1.7283494610718153e-05,
+      "loss": 0.708,
+      "step": 378
+    },
+    {
+      "epoch": 0.79,
+      "learning_rate": 1.726798941904386e-05,
+      "loss": 0.7115,
+      "step": 379
+    },
+    {
+      "epoch": 0.79,
+      "learning_rate": 1.7252447100849294e-05,
+      "loss": 0.6876,
+      "step": 380
+    },
+    {
+      "epoch": 0.8,
+      "learning_rate": 1.72368677355281e-05,
+      "loss": 0.6998,
+      "step": 381
+    },
+    {
+      "epoch": 0.8,
+      "learning_rate": 1.7221251402663176e-05,
+      "loss": 0.702,
+      "step": 382
+    },
+    {
+      "epoch": 0.8,
+      "learning_rate": 1.7205598182026243e-05,
+      "loss": 0.6964,
+      "step": 383
+    },
+    {
+      "epoch": 0.8,
+      "learning_rate": 1.7189908153577473e-05,
+      "loss": 0.7002,
+      "step": 384
+    },
+    {
+      "epoch": 0.81,
+      "learning_rate": 1.7174181397465047e-05,
+      "loss": 0.6979,
+      "step": 385
+    },
+    {
+      "epoch": 0.81,
+      "learning_rate": 1.7158417994024766e-05,
+      "loss": 0.6765,
+      "step": 386
+    },
+    {
+      "epoch": 0.81,
+      "learning_rate": 1.7142618023779623e-05,
+      "loss": 0.7065,
+      "step": 387
+    },
+    {
+      "epoch": 0.81,
+      "learning_rate": 1.7126781567439418e-05,
+      "loss": 0.7059,
+      "step": 388
+    },
+    {
+      "epoch": 0.81,
+      "learning_rate": 1.7110908705900322e-05,
+      "loss": 0.6866,
+      "step": 389
+    },
+    {
+      "epoch": 0.82,
+      "learning_rate": 1.709499952024447e-05,
+      "loss": 0.6954,
+      "step": 390
+    },
+    {
+      "epoch": 0.82,
+      "learning_rate": 1.707905409173955e-05,
+      "loss": 0.6981,
+      "step": 391
+    },
+    {
+      "epoch": 0.82,
+      "learning_rate": 1.7063072501838388e-05,
+      "loss": 0.709,
+      "step": 392
+    },
+    {
+      "epoch": 0.82,
+      "learning_rate": 1.7047054832178533e-05,
+      "loss": 0.6945,
+      "step": 393
+    },
+    {
+      "epoch": 0.82,
+      "learning_rate": 1.7031001164581828e-05,
+      "loss": 0.701,
+      "step": 394
+    },
+    {
+      "epoch": 0.83,
+      "learning_rate": 1.7014911581054016e-05,
+      "loss": 0.6843,
+      "step": 395
+    },
+    {
+      "epoch": 0.83,
+      "learning_rate": 1.6998786163784295e-05,
+      "loss": 0.6801,
+      "step": 396
+    },
+    {
+      "epoch": 0.83,
+      "learning_rate": 1.6982624995144913e-05,
+      "loss": 0.6974,
+      "step": 397
+    },
+    {
+      "epoch": 0.83,
+      "learning_rate": 1.696642815769075e-05,
+      "loss": 0.709,
+      "step": 398
+    },
+    {
+      "epoch": 0.83,
+      "learning_rate": 1.6950195734158874e-05,
+      "loss": 0.6962,
+      "step": 399
+    },
+    {
+      "epoch": 0.84,
+      "learning_rate": 1.6933927807468155e-05,
+      "loss": 0.6842,
+      "step": 400
+    },
+    {
+      "epoch": 0.84,
+      "learning_rate": 1.6917624460718812e-05,
+      "loss": 0.7046,
+      "step": 401
+    },
+    {
+      "epoch": 0.84,
+      "learning_rate": 1.690128577719199e-05,
+      "loss": 0.6955,
+      "step": 402
+    },
+    {
+      "epoch": 0.84,
+      "learning_rate": 1.6884911840349354e-05,
+      "loss": 0.6904,
+      "step": 403
+    },
+    {
+      "epoch": 0.85,
+      "learning_rate": 1.6868502733832647e-05,
+      "loss": 0.7004,
+      "step": 404
+    },
+    {
+      "epoch": 0.85,
+      "learning_rate": 1.6852058541463263e-05,
+      "loss": 0.6765,
+      "step": 405
+    },
+    {
+      "epoch": 0.85,
+      "learning_rate": 1.683557934724183e-05,
+      "loss": 0.6846,
+      "step": 406
+    },
+    {
+      "epoch": 0.85,
+      "learning_rate": 1.6819065235347763e-05,
+      "loss": 0.6922,
+      "step": 407
+    },
+    {
+      "epoch": 0.85,
+      "learning_rate": 1.680251629013885e-05,
+      "loss": 0.6964,
+      "step": 408
+    },
+    {
+      "epoch": 0.86,
+      "learning_rate": 1.6785932596150827e-05,
+      "loss": 0.6768,
+      "step": 409
+    },
+    {
+      "epoch": 0.86,
+      "learning_rate": 1.6769314238096906e-05,
+      "loss": 0.6939,
+      "step": 410
+    },
+    {
+      "epoch": 0.86,
+      "learning_rate": 1.67526613008674e-05,
+      "loss": 0.6773,
+      "step": 411
+    },
+    {
+      "epoch": 0.86,
+      "learning_rate": 1.673597386952924e-05,
+      "loss": 0.6981,
+      "step": 412
+    },
+    {
+      "epoch": 0.86,
+      "learning_rate": 1.6719252029325575e-05,
+      "loss": 0.6839,
+      "step": 413
+    },
+    {
+      "epoch": 0.87,
+      "learning_rate": 1.670249586567531e-05,
+      "loss": 0.7189,
+      "step": 414
+    },
+    {
+      "epoch": 0.87,
+      "learning_rate": 1.668570546417269e-05,
+      "loss": 0.6852,
+      "step": 415
+    },
+    {
+      "epoch": 0.87,
+      "learning_rate": 1.6668880910586853e-05,
+      "loss": 0.694,
+      "step": 416
+    },
+    {
+      "epoch": 0.87,
+      "learning_rate": 1.6652022290861393e-05,
+      "loss": 0.6889,
+      "step": 417
+    },
+    {
+      "epoch": 0.87,
+      "learning_rate": 1.663512969111392e-05,
+      "loss": 0.6873,
+      "step": 418
+    },
+    {
+      "epoch": 0.88,
+      "learning_rate": 1.6618203197635624e-05,
+      "loss": 0.7,
+      "step": 419
+    },
+    {
+      "epoch": 0.88,
+      "learning_rate": 1.6601242896890832e-05,
+      "loss": 0.6775,
+      "step": 420
+    },
+    {
+      "epoch": 0.88,
+      "learning_rate": 1.658424887551656e-05,
+      "loss": 0.6782,
+      "step": 421
+    },
+    {
+      "epoch": 0.88,
+      "learning_rate": 1.6567221220322082e-05,
+      "loss": 0.6575,
+      "step": 422
+    },
+    {
+      "epoch": 0.88,
+      "learning_rate": 1.655016001828848e-05,
+      "loss": 0.6967,
+      "step": 423
+    },
+    {
+      "epoch": 0.89,
+      "learning_rate": 1.6533065356568206e-05,
+      "loss": 0.6996,
+      "step": 424
+    },
+    {
+      "epoch": 0.89,
+      "learning_rate": 1.6515937322484627e-05,
+      "loss": 0.7136,
+      "step": 425
+    },
+    {
+      "epoch": 0.89,
+      "learning_rate": 1.6498776003531575e-05,
+      "loss": 0.6882,
+      "step": 426
+    },
+    {
+      "epoch": 0.89,
+      "learning_rate": 1.6481581487372925e-05,
+      "loss": 0.6822,
+      "step": 427
+    },
+    {
+      "epoch": 0.9,
+      "learning_rate": 1.6464353861842115e-05,
+      "loss": 0.6967,
+      "step": 428
+    },
+    {
+      "epoch": 0.9,
+      "learning_rate": 1.6447093214941727e-05,
+      "loss": 0.6928,
+      "step": 429
+    },
+    {
+      "epoch": 0.9,
+      "learning_rate": 1.6429799634843012e-05,
+      "loss": 0.6836,
+      "step": 430
+    },
+    {
+      "epoch": 0.9,
+      "learning_rate": 1.6412473209885466e-05,
+      "loss": 0.6847,
+      "step": 431
+    },
+    {
+      "epoch": 0.9,
+      "learning_rate": 1.6395114028576344e-05,
+      "loss": 0.7033,
+      "step": 432
+    },
+    {
+      "epoch": 0.91,
+      "learning_rate": 1.6377722179590237e-05,
+      "loss": 0.6873,
+      "step": 433
+    },
+    {
+      "epoch": 0.91,
+      "learning_rate": 1.636029775176862e-05,
+      "loss": 0.6816,
+      "step": 434
+    },
+    {
+      "epoch": 0.91,
+      "learning_rate": 1.6342840834119376e-05,
+      "loss": 0.6945,
+      "step": 435
+    },
+    {
+      "epoch": 0.91,
+      "learning_rate": 1.6325351515816353e-05,
+      "loss": 0.688,
+      "step": 436
+    },
+    {
+      "epoch": 0.91,
+      "learning_rate": 1.6307829886198913e-05,
+      "loss": 0.6869,
+      "step": 437
+    },
+    {
+      "epoch": 0.92,
+      "learning_rate": 1.629027603477147e-05,
+      "loss": 0.6821,
+      "step": 438
+    },
+    {
+      "epoch": 0.92,
+      "learning_rate": 1.627269005120304e-05,
+      "loss": 0.6828,
+      "step": 439
+    },
+    {
+      "epoch": 0.92,
+      "learning_rate": 1.6255072025326763e-05,
+      "loss": 0.7066,
+      "step": 440
+    },
+    {
+      "epoch": 0.92,
+      "learning_rate": 1.6237422047139472e-05,
+      "loss": 0.6772,
+      "step": 441
+    },
+    {
+      "epoch": 0.92,
+      "learning_rate": 1.621974020680122e-05,
+      "loss": 0.7041,
+      "step": 442
+    },
+    {
+      "epoch": 0.93,
+      "learning_rate": 1.6202026594634804e-05,
+      "loss": 0.6861,
+      "step": 443
+    },
+    {
+      "epoch": 0.93,
+      "learning_rate": 1.618428130112533e-05,
+      "loss": 0.6878,
+      "step": 444
+    },
+    {
+      "epoch": 0.93,
+      "learning_rate": 1.6166504416919747e-05,
+      "loss": 0.7052,
+      "step": 445
+    },
+    {
+      "epoch": 0.93,
+      "learning_rate": 1.6148696032826354e-05,
+      "loss": 0.6772,
+      "step": 446
+    },
+    {
+      "epoch": 0.94,
+      "learning_rate": 1.6130856239814382e-05,
+      "loss": 0.6928,
+      "step": 447
+    },
+    {
+      "epoch": 0.94,
+      "learning_rate": 1.611298512901349e-05,
+      "loss": 0.6924,
+      "step": 448
+    },
+    {
+      "epoch": 0.94,
+      "learning_rate": 1.6095082791713322e-05,
+      "loss": 0.6906,
+      "step": 449
+    },
+    {
+      "epoch": 0.94,
+      "learning_rate": 1.6077149319363035e-05,
+      "loss": 0.6986,
+      "step": 450
+    },
+    {
+      "epoch": 0.94,
+      "learning_rate": 1.6059184803570826e-05,
+      "loss": 0.6844,
+      "step": 451
+    },
+    {
+      "epoch": 0.95,
+      "learning_rate": 1.6041189336103475e-05,
+      "loss": 0.674,
+      "step": 452
+    },
+    {
+      "epoch": 0.95,
+      "learning_rate": 1.6023163008885858e-05,
+      "loss": 0.6804,
+      "step": 453
+    },
+    {
+      "epoch": 0.95,
+      "learning_rate": 1.6005105914000508e-05,
+      "loss": 0.6954,
+      "step": 454
+    },
+    {
+      "epoch": 0.95,
+      "learning_rate": 1.598701814368711e-05,
+      "loss": 0.6808,
+      "step": 455
+    },
+    {
+      "epoch": 0.95,
+      "learning_rate": 1.596889979034205e-05,
+      "loss": 0.6834,
+      "step": 456
+    },
+    {
+      "epoch": 0.96,
+      "learning_rate": 1.595075094651795e-05,
+      "loss": 0.6769,
+      "step": 457
+    },
+    {
+      "epoch": 0.96,
+      "learning_rate": 1.5932571704923168e-05,
+      "loss": 0.688,
+      "step": 458
+    },
+    {
+      "epoch": 0.96,
+      "learning_rate": 1.5914362158421352e-05,
+      "loss": 0.6731,
+      "step": 459
+    },
+    {
+      "epoch": 0.96,
+      "learning_rate": 1.589612240003095e-05,
+      "loss": 0.6967,
+      "step": 460
+    },
+    {
+      "epoch": 0.96,
+      "learning_rate": 1.5877852522924733e-05,
+      "loss": 0.6872,
+      "step": 461
+    },
+    {
+      "epoch": 0.97,
+      "learning_rate": 1.585955262042934e-05,
+      "loss": 0.6883,
+      "step": 462
+    },
+    {
+      "epoch": 0.97,
+      "learning_rate": 1.584122278602477e-05,
+      "loss": 0.6768,
+      "step": 463
+    },
+    {
+      "epoch": 0.97,
+      "learning_rate": 1.5822863113343934e-05,
+      "loss": 0.6772,
+      "step": 464
+    },
+    {
+      "epoch": 0.97,
+      "learning_rate": 1.580447369617216e-05,
+      "loss": 0.6882,
+      "step": 465
+    },
+    {
+      "epoch": 0.97,
+      "learning_rate": 1.5786054628446712e-05,
+      "loss": 0.6652,
+      "step": 466
+    },
+    {
+      "epoch": 0.98,
+      "learning_rate": 1.576760600425632e-05,
+      "loss": 0.6749,
+      "step": 467
+    },
+    {
+      "epoch": 0.98,
+      "learning_rate": 1.57491279178407e-05,
+      "loss": 0.6701,
+      "step": 468
+    },
+    {
+      "epoch": 0.98,
+      "learning_rate": 1.5730620463590052e-05,
+      "loss": 0.6911,
+      "step": 469
+    },
+    {
+      "epoch": 0.98,
+      "learning_rate": 1.5712083736044613e-05,
+      "loss": 0.6741,
+      "step": 470
+    },
+    {
+      "epoch": 0.99,
+      "learning_rate": 1.5693517829894138e-05,
+      "loss": 0.6942,
+      "step": 471
+    },
+    {
+      "epoch": 0.99,
+      "learning_rate": 1.5674922839977446e-05,
+      "loss": 0.7019,
+      "step": 472
+    },
+    {
+      "epoch": 0.99,
+      "learning_rate": 1.5656298861281917e-05,
+      "loss": 0.6996,
+      "step": 473
+    },
+    {
+      "epoch": 0.99,
+      "learning_rate": 1.5637645988943008e-05,
+      "loss": 0.6794,
+      "step": 474
+    },
+    {
+      "epoch": 0.99,
+      "learning_rate": 1.5618964318243775e-05,
+      "loss": 0.6802,
+      "step": 475
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 1.560025394461439e-05,
+      "loss": 0.6794,
+      "step": 476
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 1.558151496363163e-05,
+      "loss": 0.7043,
+      "step": 477
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 1.5562747471018415e-05,
+      "loss": 0.6799,
+      "step": 478
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 1.554395156264331e-05,
+      "loss": 0.5765,
+      "step": 479
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 1.552512733452003e-05,
+      "loss": 0.592,
+      "step": 480
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 1.550627488280695e-05,
+      "loss": 0.5588,
+      "step": 481
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 1.5487394303806632e-05,
+      "loss": 0.5646,
+      "step": 482
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 1.54684856939653e-05,
+      "loss": 0.5682,
+      "step": 483
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 1.544954914987238e-05,
+      "loss": 0.5536,
+      "step": 484
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 1.543058476825998e-05,
+      "loss": 0.5545,
+      "step": 485
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 1.541159264600242e-05,
+      "loss": 0.5699,
+      "step": 486
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 1.5392572880115726e-05,
+      "loss": 0.558,
+      "step": 487
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 1.5373525567757124e-05,
+      "loss": 0.5509,
+      "step": 488
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 1.5354450806224553e-05,
+      "loss": 0.5698,
+      "step": 489
+    },
+    {
+      "epoch": 1.03,
+      "learning_rate": 1.5335348692956177e-05,
+      "loss": 0.556,
+      "step": 490
+    },
+    {
+      "epoch": 1.03,
+      "learning_rate": 1.5316219325529882e-05,
+      "loss": 0.5686,
+      "step": 491
+    },
+    {
+      "epoch": 1.03,
+      "learning_rate": 1.529706280166276e-05,
+      "loss": 0.561,
+      "step": 492
+    },
+    {
+      "epoch": 1.03,
+      "learning_rate": 1.5277879219210637e-05,
+      "loss": 0.5768,
+      "step": 493
+    },
+    {
+      "epoch": 1.03,
+      "learning_rate": 1.5258668676167548e-05,
+      "loss": 0.5578,
+      "step": 494
+    },
+    {
+      "epoch": 1.04,
+      "learning_rate": 1.5239431270665263e-05,
+      "loss": 0.5582,
+      "step": 495
+    },
+    {
+      "epoch": 1.04,
+      "learning_rate": 1.5220167100972763e-05,
+      "loss": 0.5679,
+      "step": 496
+    },
+    {
+      "epoch": 1.04,
+      "learning_rate": 1.5200876265495745e-05,
+      "loss": 0.5593,
+      "step": 497
+    },
+    {
+      "epoch": 1.04,
+      "learning_rate": 1.518155886277613e-05,
+      "loss": 0.5677,
+      "step": 498
+    },
+    {
+      "epoch": 1.04,
+      "learning_rate": 1.516221499149154e-05,
+      "loss": 0.5428,
+      "step": 499
+    },
+    {
+      "epoch": 1.05,
+      "learning_rate": 1.5142844750454807e-05,
+      "loss": 0.5555,
+      "step": 500
+    },
+    {
+      "epoch": 1.05,
+      "learning_rate": 1.512344823861347e-05,
+      "loss": 0.5559,
+      "step": 501
+    },
+    {
+      "epoch": 1.05,
+      "learning_rate": 1.5104025555049262e-05,
+      "loss": 0.5627,
+      "step": 502
+    },
+    {
+      "epoch": 1.05,
+      "learning_rate": 1.5084576798977608e-05,
+      "loss": 0.5679,
+      "step": 503
+    },
+    {
+      "epoch": 1.05,
+      "learning_rate": 1.5065102069747117e-05,
+      "loss": 0.5572,
+      "step": 504
+    },
+    {
+      "epoch": 1.06,
+      "learning_rate": 1.504560146683907e-05,
+      "loss": 0.5635,
+      "step": 505
+    },
+    {
+      "epoch": 1.06,
+      "learning_rate": 1.502607508986693e-05,
+      "loss": 0.5625,
+      "step": 506
+    },
+    {
+      "epoch": 1.06,
+      "learning_rate": 1.50065230385758e-05,
+      "loss": 0.5557,
+      "step": 507
+    },
+    {
+      "epoch": 1.06,
+      "learning_rate": 1.498694541284195e-05,
+      "loss": 0.567,
+      "step": 508
+    },
+    {
+      "epoch": 1.06,
+      "learning_rate": 1.4967342312672283e-05,
+      "loss": 0.54,
+      "step": 509
+    },
+    {
+      "epoch": 1.07,
+      "learning_rate": 1.4947713838203835e-05,
+      "loss": 0.5484,
+      "step": 510
+    },
+    {
+      "epoch": 1.07,
+      "learning_rate": 1.492806008970325e-05,
+      "loss": 0.5825,
+      "step": 511
+    },
+    {
+      "epoch": 1.07,
+      "learning_rate": 1.4908381167566286e-05,
+      "loss": 0.5479,
+      "step": 512
+    },
+    {
+      "epoch": 1.07,
+      "learning_rate": 1.4888677172317294e-05,
+      "loss": 0.5599,
+      "step": 513
+    },
+    {
+      "epoch": 1.08,
+      "learning_rate": 1.48689482046087e-05,
+      "loss": 0.5585,
+      "step": 514
+    },
+    {
+      "epoch": 1.08,
+      "learning_rate": 1.4849194365220497e-05,
+      "loss": 0.5612,
+      "step": 515
+    },
+    {
+      "epoch": 1.08,
+      "learning_rate": 1.4829415755059726e-05,
+      "loss": 0.5633,
+      "step": 516
+    },
+    {
+      "epoch": 1.08,
+      "learning_rate": 1.4809612475159968e-05,
+      "loss": 0.5448,
+      "step": 517
+    },
+    {
+      "epoch": 1.08,
+      "learning_rate": 1.4789784626680819e-05,
+      "loss": 0.5633,
+      "step": 518
+    },
+    {
+      "epoch": 1.09,
+      "learning_rate": 1.4769932310907372e-05,
+      "loss": 0.5536,
+      "step": 519
+    },
+    {
+      "epoch": 1.09,
+      "learning_rate": 1.475005562924971e-05,
+      "loss": 0.5806,
+      "step": 520
+    },
+    {
+      "epoch": 1.09,
+      "learning_rate": 1.4730154683242385e-05,
+      "loss": 0.5553,
+      "step": 521
+    },
+    {
+      "epoch": 1.09,
+      "learning_rate": 1.4710229574543893e-05,
+      "loss": 0.5574,
+      "step": 522
+    },
+    {
+      "epoch": 1.09,
+      "learning_rate": 1.4690280404936155e-05,
+      "loss": 0.5673,
+      "step": 523
+    },
+    {
+      "epoch": 1.1,
+      "learning_rate": 1.467030727632401e-05,
+      "loss": 0.5646,
+      "step": 524
+    },
+    {
+      "epoch": 1.1,
+      "learning_rate": 1.465031029073467e-05,
+      "loss": 0.5609,
+      "step": 525
+    },
+    {
+      "epoch": 1.1,
+      "learning_rate": 1.4630289550317234e-05,
+      "loss": 0.5495,
+      "step": 526
+    },
+    {
+      "epoch": 1.1,
+      "learning_rate": 1.4610245157342127e-05,
+      "loss": 0.5645,
+      "step": 527
+    },
+    {
+      "epoch": 1.1,
+      "learning_rate": 1.4590177214200609e-05,
+      "loss": 0.5541,
+      "step": 528
+    },
+    {
+      "epoch": 1.11,
+      "learning_rate": 1.4570085823404232e-05,
+      "loss": 0.544,
+      "step": 529
+    },
+    {
+      "epoch": 1.11,
+      "learning_rate": 1.4549971087584329e-05,
+      "loss": 0.5432,
+      "step": 530
+    },
+    {
+      "epoch": 1.11,
+      "learning_rate": 1.4529833109491485e-05,
+      "loss": 0.5594,
+      "step": 531
+    },
+    {
+      "epoch": 1.11,
+      "learning_rate": 1.4509671991995003e-05,
+      "loss": 0.553,
+      "step": 532
+    },
+    {
+      "epoch": 1.12,
+      "learning_rate": 1.4489487838082396e-05,
+      "loss": 0.5567,
+      "step": 533
+    },
+    {
+      "epoch": 1.12,
+      "learning_rate": 1.4469280750858854e-05,
+      "loss": 0.5529,
+      "step": 534
+    },
+    {
+      "epoch": 1.12,
+      "learning_rate": 1.4449050833546709e-05,
+      "loss": 0.5696,
+      "step": 535
+    },
+    {
+      "epoch": 1.12,
+      "learning_rate": 1.4428798189484914e-05,
+      "loss": 0.57,
+      "step": 536
+    },
+    {
+      "epoch": 1.12,
+      "learning_rate": 1.4408522922128518e-05,
+      "loss": 0.5617,
+      "step": 537
+    },
+    {
+      "epoch": 1.13,
+      "learning_rate": 1.4388225135048137e-05,
+      "loss": 0.5603,
+      "step": 538
+    },
+    {
+      "epoch": 1.13,
+      "learning_rate": 1.4367904931929422e-05,
+      "loss": 0.5607,
+      "step": 539
+    },
+    {
+      "epoch": 1.13,
+      "learning_rate": 1.4347562416572525e-05,
+      "loss": 0.55,
+      "step": 540
+    },
+    {
+      "epoch": 1.13,
+      "learning_rate": 1.432719769289158e-05,
+      "loss": 0.5711,
+      "step": 541
+    },
+    {
+      "epoch": 1.13,
+      "learning_rate": 1.430681086491416e-05,
+      "loss": 0.5636,
+      "step": 542
+    },
+    {
+      "epoch": 1.14,
+      "learning_rate": 1.4286402036780762e-05,
+      "loss": 0.5859,
+      "step": 543
+    },
+    {
+      "epoch": 1.14,
+      "learning_rate": 1.4265971312744252e-05,
+      "loss": 0.5608,
+      "step": 544
+    },
+    {
+      "epoch": 1.14,
+      "learning_rate": 1.4245518797169353e-05,
+      "loss": 0.5574,
+      "step": 545
+    },
+    {
+      "epoch": 1.14,
+      "learning_rate": 1.4225044594532104e-05,
+      "loss": 0.547,
+      "step": 546
+    },
+    {
+      "epoch": 1.14,
+      "learning_rate": 1.4204548809419323e-05,
+      "loss": 0.5554,
+      "step": 547
+    },
+    {
+      "epoch": 1.15,
+      "learning_rate": 1.4184031546528077e-05,
+      "loss": 0.5751,
+      "step": 548
+    },
+    {
+      "epoch": 1.15,
+      "learning_rate": 1.4163492910665153e-05,
+      "loss": 0.5721,
+      "step": 549
+    },
+    {
+      "epoch": 1.15,
+      "learning_rate": 1.4142933006746502e-05,
+      "loss": 0.559,
+      "step": 550
+    },
+    {
+      "epoch": 1.15,
+      "learning_rate": 1.4122351939796733e-05,
+      "loss": 0.5644,
+      "step": 551
+    },
+    {
+      "epoch": 1.15,
+      "learning_rate": 1.4101749814948544e-05,
+      "loss": 0.5602,
+      "step": 552
+    },
+    {
+      "epoch": 1.16,
+      "learning_rate": 1.4081126737442216e-05,
+      "loss": 0.5584,
+      "step": 553
+    },
+    {
+      "epoch": 1.16,
+      "learning_rate": 1.4060482812625055e-05,
+      "loss": 0.5637,
+      "step": 554
+    },
+    {
+      "epoch": 1.16,
+      "learning_rate": 1.4039818145950856e-05,
+      "loss": 0.5534,
+      "step": 555
+    },
+    {
+      "epoch": 1.16,
+      "learning_rate": 1.4019132842979375e-05,
+      "loss": 0.5546,
+      "step": 556
+    },
+    {
+      "epoch": 1.17,
+      "learning_rate": 1.399842700937578e-05,
+      "loss": 0.5667,
+      "step": 557
+    },
+    {
+      "epoch": 1.17,
+      "learning_rate": 1.3977700750910112e-05,
+      "loss": 0.5567,
+      "step": 558
+    },
+    {
+      "epoch": 1.17,
+      "learning_rate": 1.395695417345675e-05,
+      "loss": 0.5499,
+      "step": 559
+    },
+    {
+      "epoch": 1.17,
+      "learning_rate": 1.3936187382993862e-05,
+      "loss": 0.5571,
+      "step": 560
+    },
+    {
+      "epoch": 1.17,
+      "learning_rate": 1.3915400485602876e-05,
+      "loss": 0.5678,
+      "step": 561
+    },
+    {
+      "epoch": 1.18,
+      "learning_rate": 1.3894593587467924e-05,
+      "loss": 0.5765,
+      "step": 562
+    },
+    {
+      "epoch": 1.18,
+      "learning_rate": 1.3873766794875309e-05,
+      "loss": 0.5722,
+      "step": 563
+    },
+    {
+      "epoch": 1.18,
+      "learning_rate": 1.3852920214212966e-05,
+      "loss": 0.5698,
+      "step": 564
+    },
+    {
+      "epoch": 1.18,
+      "learning_rate": 1.3832053951969893e-05,
+      "loss": 0.5617,
+      "step": 565
+    },
+    {
+      "epoch": 1.18,
+      "learning_rate": 1.3811168114735647e-05,
+      "loss": 0.5531,
+      "step": 566
+    },
+    {
+      "epoch": 1.19,
+      "learning_rate": 1.3790262809199768e-05,
+      "loss": 0.5606,
+      "step": 567
+    },
+    {
+      "epoch": 1.19,
+      "learning_rate": 1.3769338142151245e-05,
+      "loss": 0.5681,
+      "step": 568
+    },
+    {
+      "epoch": 1.19,
+      "learning_rate": 1.3748394220477972e-05,
+      "loss": 0.5546,
+      "step": 569
+    },
+    {
+      "epoch": 1.19,
+      "learning_rate": 1.3727431151166196e-05,
+      "loss": 0.5686,
+      "step": 570
+    },
+    {
+      "epoch": 1.19,
+      "learning_rate": 1.3706449041299984e-05,
+      "loss": 0.5805,
+      "step": 571
+    },
+    {
+      "epoch": 1.2,
+      "learning_rate": 1.3685447998060651e-05,
+      "loss": 0.57,
+      "step": 572
+    },
+    {
+      "epoch": 1.2,
+      "learning_rate": 1.3664428128726242e-05,
+      "loss": 0.5436,
+      "step": 573
+    },
+    {
+      "epoch": 1.2,
+      "learning_rate": 1.3643389540670963e-05,
+      "loss": 0.557,
+      "step": 574
+    },
+    {
+      "epoch": 1.2,
+      "learning_rate": 1.3622332341364646e-05,
+      "loss": 0.5589,
+      "step": 575
+    },
+    {
+      "epoch": 1.21,
+      "learning_rate": 1.3601256638372182e-05,
+      "loss": 0.5618,
+      "step": 576
+    },
+    {
+      "epoch": 1.21,
+      "learning_rate": 1.3580162539352993e-05,
+      "loss": 0.559,
+      "step": 577
+    },
+    {
+      "epoch": 1.21,
+      "learning_rate": 1.3559050152060465e-05,
+      "loss": 0.5564,
+      "step": 578
+    },
+    {
+      "epoch": 1.21,
+      "learning_rate": 1.3537919584341413e-05,
+      "loss": 0.5672,
+      "step": 579
+    },
+    {
+      "epoch": 1.21,
+      "learning_rate": 1.3516770944135514e-05,
+      "loss": 0.5695,
+      "step": 580
+    },
+    {
+      "epoch": 1.22,
+      "learning_rate": 1.3495604339474763e-05,
+      "loss": 0.5595,
+      "step": 581
+    },
+    {
+      "epoch": 1.22,
+      "learning_rate": 1.3474419878482935e-05,
+      "loss": 0.5612,
+      "step": 582
+    },
+    {
+      "epoch": 1.22,
+      "learning_rate": 1.3453217669374997e-05,
+      "loss": 0.5507,
+      "step": 583
+    },
+    {
+      "epoch": 1.22,
+      "learning_rate": 1.3431997820456592e-05,
+      "loss": 0.551,
+      "step": 584
+    },
+    {
+      "epoch": 1.22,
+      "learning_rate": 1.3410760440123475e-05,
+      "loss": 0.5621,
+      "step": 585
+    },
+    {
+      "epoch": 1.23,
+      "learning_rate": 1.3389505636860944e-05,
+      "loss": 0.5586,
+      "step": 586
+    },
+    {
+      "epoch": 1.23,
+      "learning_rate": 1.3368233519243301e-05,
+      "loss": 0.5711,
+      "step": 587
+    },
+    {
+      "epoch": 1.23,
+      "learning_rate": 1.3346944195933294e-05,
+      "loss": 0.5714,
+      "step": 588
+    },
+    {
+      "epoch": 1.23,
+      "learning_rate": 1.3325637775681561e-05,
+      "loss": 0.571,
+      "step": 589
+    },
+    {
+      "epoch": 1.23,
+      "learning_rate": 1.330431436732608e-05,
+      "loss": 0.5475,
+      "step": 590
+    },
+    {
+      "epoch": 1.24,
+      "learning_rate": 1.3282974079791603e-05,
+      "loss": 0.5628,
+      "step": 591
+    },
+    {
+      "epoch": 1.24,
+      "learning_rate": 1.3261617022089103e-05,
+      "loss": 0.5696,
+      "step": 592
+    },
+    {
+      "epoch": 1.24,
+      "learning_rate": 1.3240243303315227e-05,
+      "loss": 0.5617,
+      "step": 593
+    },
+    {
+      "epoch": 1.24,
+      "learning_rate": 1.3218853032651719e-05,
+      "loss": 0.5584,
+      "step": 594
+    },
+    {
+      "epoch": 1.24,
+      "learning_rate": 1.3197446319364882e-05,
+      "loss": 0.5752,
+      "step": 595
+    },
+    {
+      "epoch": 1.25,
+      "learning_rate": 1.3176023272805008e-05,
+      "loss": 0.5622,
+      "step": 596
+    },
+    {
+      "epoch": 1.25,
+      "learning_rate": 1.3154584002405828e-05,
+      "loss": 0.5609,
+      "step": 597
+    },
+    {
+      "epoch": 1.25,
+      "learning_rate": 1.313312861768394e-05,
+      "loss": 0.5522,
+      "step": 598
+    },
+    {
+      "epoch": 1.25,
+      "learning_rate": 1.3111657228238263e-05,
+      "loss": 0.5721,
+      "step": 599
+    },
+    {
+      "epoch": 1.26,
+      "learning_rate": 1.3090169943749475e-05,
+      "loss": 0.5539,
+      "step": 600
+    },
+    {
+      "epoch": 1.26,
+      "learning_rate": 1.3068666873979443e-05,
+      "loss": 0.5619,
+      "step": 601
+    },
+    {
+      "epoch": 1.26,
+      "learning_rate": 1.3047148128770664e-05,
+      "loss": 0.5527,
+      "step": 602
+    },
+    {
+      "epoch": 1.26,
+      "learning_rate": 1.3025613818045723e-05,
+      "loss": 0.5681,
+      "step": 603
+    },
+    {
+      "epoch": 1.26,
+      "learning_rate": 1.3004064051806712e-05,
+      "loss": 0.5591,
+      "step": 604
+    },
+    {
+      "epoch": 1.27,
+      "learning_rate": 1.298249894013466e-05,
+      "loss": 0.5512,
+      "step": 605
+    },
+    {
+      "epoch": 1.27,
+      "learning_rate": 1.2960918593189005e-05,
+      "loss": 0.561,
+      "step": 606
+    },
+    {
+      "epoch": 1.27,
+      "learning_rate": 1.2939323121206997e-05,
+      "loss": 0.5514,
+      "step": 607
+    },
+    {
+      "epoch": 1.27,
+      "learning_rate": 1.2917712634503148e-05,
+      "loss": 0.5679,
+      "step": 608
+    },
+    {
+      "epoch": 1.27,
+      "learning_rate": 1.2896087243468673e-05,
+      "loss": 0.5756,
+      "step": 609
+    },
+    {
+      "epoch": 1.28,
+      "learning_rate": 1.2874447058570927e-05,
+      "loss": 0.5607,
+      "step": 610
+    },
+    {
+      "epoch": 1.28,
+      "learning_rate": 1.2852792190352821e-05,
+      "loss": 0.57,
+      "step": 611
+    },
+    {
+      "epoch": 1.28,
+      "learning_rate": 1.2831122749432278e-05,
+      "loss": 0.5732,
+      "step": 612
+    },
+    {
+      "epoch": 1.28,
+      "learning_rate": 1.280943884650167e-05,
+      "loss": 0.5648,
+      "step": 613
+    },
+    {
+      "epoch": 1.28,
+      "learning_rate": 1.2787740592327232e-05,
+      "loss": 0.5746,
+      "step": 614
+    },
+    {
+      "epoch": 1.29,
+      "learning_rate": 1.2766028097748518e-05,
+      "loss": 0.5735,
+      "step": 615
+    },
+    {
+      "epoch": 1.29,
+      "learning_rate": 1.2744301473677814e-05,
+      "loss": 0.5823,
+      "step": 616
+    },
+    {
+      "epoch": 1.29,
+      "learning_rate": 1.2722560831099595e-05,
+      "loss": 0.5482,
+      "step": 617
+    },
+    {
+      "epoch": 1.29,
+      "learning_rate": 1.2700806281069942e-05,
+      "loss": 0.5633,
+      "step": 618
+    },
+    {
+      "epoch": 1.29,
+      "learning_rate": 1.267903793471597e-05,
+      "loss": 0.5591,
+      "step": 619
+    },
+    {
+      "epoch": 1.3,
+      "learning_rate": 1.2657255903235278e-05,
+      "loss": 0.5582,
+      "step": 620
+    },
+    {
+      "epoch": 1.3,
+      "learning_rate": 1.2635460297895367e-05,
+      "loss": 0.553,
+      "step": 621
+    },
+    {
+      "epoch": 1.3,
+      "learning_rate": 1.2613651230033085e-05,
+      "loss": 0.5552,
+      "step": 622
+    },
+    {
+      "epoch": 1.3,
+      "learning_rate": 1.2591828811054034e-05,
+      "loss": 0.5647,
+      "step": 623
+    },
+    {
+      "epoch": 1.31,
+      "learning_rate": 1.2569993152432028e-05,
+      "loss": 0.5701,
+      "step": 624
+    },
+    {
+      "epoch": 1.31,
+      "learning_rate": 1.2548144365708514e-05,
+      "loss": 0.5518,
+      "step": 625
+    },
+    {
+      "epoch": 1.31,
+      "learning_rate": 1.2526282562491991e-05,
+      "loss": 0.569,
+      "step": 626
+    },
+    {
+      "epoch": 1.31,
+      "learning_rate": 1.2504407854457455e-05,
+      "loss": 0.5688,
+      "step": 627
+    },
+    {
+      "epoch": 1.31,
+      "learning_rate": 1.2482520353345819e-05,
+      "loss": 0.57,
+      "step": 628
+    },
+    {
+      "epoch": 1.32,
+      "learning_rate": 1.2460620170963353e-05,
+      "loss": 0.5712,
+      "step": 629
+    },
+    {
+      "epoch": 1.32,
+      "learning_rate": 1.2438707419181097e-05,
+      "loss": 0.5536,
+      "step": 630
+    },
+    {
+      "epoch": 1.32,
+      "learning_rate": 1.2416782209934304e-05,
+      "loss": 0.5574,
+      "step": 631
+    },
+    {
+      "epoch": 1.32,
+      "learning_rate": 1.2394844655221863e-05,
+      "loss": 0.5639,
+      "step": 632
+    },
+    {
+      "epoch": 1.32,
+      "learning_rate": 1.2372894867105722e-05,
+      "loss": 0.5731,
+      "step": 633
+    },
+    {
+      "epoch": 1.33,
+      "learning_rate": 1.2350932957710322e-05,
+      "loss": 0.563,
+      "step": 634
+    },
+    {
+      "epoch": 1.33,
+      "learning_rate": 1.2328959039222026e-05,
+      "loss": 0.5604,
+      "step": 635
+    },
+    {
+      "epoch": 1.33,
+      "learning_rate": 1.2306973223888535e-05,
+      "loss": 0.5515,
+      "step": 636
+    },
+    {
+      "epoch": 1.33,
+      "learning_rate": 1.2284975624018327e-05,
+      "loss": 0.5615,
+      "step": 637
+    },
+    {
+      "epoch": 1.33,
+      "learning_rate": 1.2262966351980075e-05,
+      "loss": 0.5626,
+      "step": 638
+    },
+    {
+      "epoch": 1.34,
+      "learning_rate": 1.2240945520202079e-05,
+      "loss": 0.5548,
+      "step": 639
+    },
+    {
+      "epoch": 1.34,
+      "learning_rate": 1.2218913241171691e-05,
+      "loss": 0.5504,
+      "step": 640
+    },
+    {
+      "epoch": 1.34,
+      "learning_rate": 1.2196869627434725e-05,
+      "loss": 0.5579,
+      "step": 641
+    },
+    {
+      "epoch": 1.34,
+      "learning_rate": 1.2174814791594913e-05,
+      "loss": 0.5532,
+      "step": 642
+    },
+    {
+      "epoch": 1.35,
+      "learning_rate": 1.2152748846313302e-05,
+      "loss": 0.5739,
+      "step": 643
+    },
+    {
+      "epoch": 1.35,
+      "learning_rate": 1.2130671904307692e-05,
+      "loss": 0.5707,
+      "step": 644
+    },
+    {
+      "epoch": 1.35,
+      "learning_rate": 1.2108584078352048e-05,
+      "loss": 0.5626,
+      "step": 645
+    },
+    {
+      "epoch": 1.35,
+      "learning_rate": 1.2086485481275943e-05,
+      "loss": 0.5734,
+      "step": 646
+    },
+    {
+      "epoch": 1.35,
+      "learning_rate": 1.2064376225963972e-05,
+      "loss": 0.5608,
+      "step": 647
+    },
+    {
+      "epoch": 1.36,
+      "learning_rate": 1.2042256425355165e-05,
+      "loss": 0.5595,
+      "step": 648
+    },
+    {
+      "epoch": 1.36,
+      "learning_rate": 1.202012619244243e-05,
+      "loss": 0.5614,
+      "step": 649
+    },
+    {
+      "epoch": 1.36,
+      "learning_rate": 1.1997985640271956e-05,
+      "loss": 0.5652,
+      "step": 650
+    },
+    {
+      "epoch": 1.36,
+      "learning_rate": 1.1975834881942656e-05,
+      "loss": 0.5578,
+      "step": 651
+    },
+    {
+      "epoch": 1.36,
+      "learning_rate": 1.1953674030605568e-05,
+      "loss": 0.5608,
+      "step": 652
+    },
+    {
+      "epoch": 1.37,
+      "learning_rate": 1.1931503199463292e-05,
+      "loss": 0.5554,
+      "step": 653
+    },
+    {
+      "epoch": 1.37,
+      "learning_rate": 1.1909322501769407e-05,
+      "loss": 0.5679,
+      "step": 654
+    },
+    {
+      "epoch": 1.37,
+      "learning_rate": 1.1887132050827893e-05,
+      "loss": 0.5611,
+      "step": 655
+    },
+    {
+      "epoch": 1.37,
+      "learning_rate": 1.186493195999255e-05,
+      "loss": 0.5747,
+      "step": 656
+    },
+    {
+      "epoch": 1.37,
+      "learning_rate": 1.1842722342666421e-05,
+      "loss": 0.5693,
+      "step": 657
+    },
+    {
+      "epoch": 1.38,
+      "learning_rate": 1.1820503312301218e-05,
+      "loss": 0.5708,
+      "step": 658
+    },
+    {
+      "epoch": 1.38,
+      "learning_rate": 1.1798274982396728e-05,
+      "loss": 0.5664,
+      "step": 659
+    },
+    {
+      "epoch": 1.38,
+      "learning_rate": 1.1776037466500245e-05,
+      "loss": 0.5686,
+      "step": 660
+    },
+    {
+      "epoch": 1.38,
+      "learning_rate": 1.1753790878205993e-05,
+      "loss": 0.5622,
+      "step": 661
+    },
+    {
+      "epoch": 1.38,
+      "learning_rate": 1.1731535331154532e-05,
+      "loss": 0.5498,
+      "step": 662
+    },
+    {
+      "epoch": 1.39,
+      "learning_rate": 1.1709270939032189e-05,
+      "loss": 0.565,
+      "step": 663
+    },
+    {
+      "epoch": 1.39,
+      "learning_rate": 1.1686997815570473e-05,
+      "loss": 0.5698,
+      "step": 664
+    },
+    {
+      "epoch": 1.39,
+      "learning_rate": 1.1664716074545498e-05,
+      "loss": 0.5586,
+      "step": 665
+    },
+    {
+      "epoch": 1.39,
+      "learning_rate": 1.1642425829777391e-05,
+      "loss": 0.5605,
+      "step": 666
+    },
+    {
+      "epoch": 1.4,
+      "learning_rate": 1.1620127195129724e-05,
+      "loss": 0.5622,
+      "step": 667
+    },
+    {
+      "epoch": 1.4,
+      "learning_rate": 1.1597820284508927e-05,
+      "loss": 0.5746,
+      "step": 668
+    },
+    {
+      "epoch": 1.4,
+      "learning_rate": 1.15755052118637e-05,
+      "loss": 0.5623,
+      "step": 669
+    },
+    {
+      "epoch": 1.4,
+      "learning_rate": 1.1553182091184439e-05,
+      "loss": 0.5673,
+      "step": 670
+    },
+    {
+      "epoch": 1.4,
+      "learning_rate": 1.1530851036502655e-05,
+      "loss": 0.5567,
+      "step": 671
+    },
+    {
+      "epoch": 1.41,
+      "learning_rate": 1.1508512161890381e-05,
+      "loss": 0.5537,
+      "step": 672
+    },
+    {
+      "epoch": 1.41,
+      "learning_rate": 1.1486165581459605e-05,
+      "loss": 0.555,
+      "step": 673
+    },
+    {
+      "epoch": 1.41,
+      "learning_rate": 1.1463811409361667e-05,
+      "loss": 0.5615,
+      "step": 674
+    },
+    {
+      "epoch": 1.41,
+      "learning_rate": 1.1441449759786693e-05,
+      "loss": 0.574,
+      "step": 675
+    },
+    {
+      "epoch": 1.41,
+      "learning_rate": 1.1419080746963012e-05,
+      "loss": 0.5762,
+      "step": 676
+    },
+    {
+      "epoch": 1.42,
+      "learning_rate": 1.1396704485156552e-05,
+      "loss": 0.5617,
+      "step": 677
+    },
+    {
+      "epoch": 1.42,
+      "learning_rate": 1.1374321088670277e-05,
+      "loss": 0.5499,
+      "step": 678
+    },
+    {
+      "epoch": 1.42,
+      "learning_rate": 1.13519306718436e-05,
+      "loss": 0.5557,
+      "step": 679
+    },
+    {
+      "epoch": 1.42,
+      "learning_rate": 1.1329533349051794e-05,
+      "loss": 0.5621,
+      "step": 680
+    },
+    {
+      "epoch": 1.42,
+      "learning_rate": 1.1307129234705404e-05,
+      "loss": 0.5566,
+      "step": 681
+    },
+    {
+      "epoch": 1.43,
+      "learning_rate": 1.1284718443249676e-05,
+      "loss": 0.5651,
+      "step": 682
+    },
+    {
+      "epoch": 1.43,
+      "learning_rate": 1.1262301089163958e-05,
+      "loss": 0.5516,
+      "step": 683
+    },
+    {
+      "epoch": 1.43,
+      "learning_rate": 1.1239877286961123e-05,
+      "loss": 0.5525,
+      "step": 684
+    },
+    {
+      "epoch": 1.43,
+      "learning_rate": 1.1217447151186983e-05,
+      "loss": 0.5618,
+      "step": 685
+    },
+    {
+      "epoch": 1.44,
+      "learning_rate": 1.11950107964197e-05,
+      "loss": 0.5471,
+      "step": 686
+    },
+    {
+      "epoch": 1.44,
+      "learning_rate": 1.1172568337269212e-05,
+      "loss": 0.5543,
+      "step": 687
+    },
+    {
+      "epoch": 1.44,
+      "learning_rate": 1.1150119888376631e-05,
+      "loss": 0.5752,
+      "step": 688
+    },
+    {
+      "epoch": 1.44,
+      "learning_rate": 1.112766556441367e-05,
+      "loss": 0.5644,
+      "step": 689
+    },
+    {
+      "epoch": 1.44,
+      "learning_rate": 1.1105205480082052e-05,
+      "loss": 0.5726,
+      "step": 690
+    },
+    {
+      "epoch": 1.45,
+      "learning_rate": 1.1082739750112932e-05,
+      "loss": 0.5555,
+      "step": 691
+    },
+    {
+      "epoch": 1.45,
+      "learning_rate": 1.106026848926629e-05,
+      "loss": 0.5567,
+      "step": 692
+    },
+    {
+      "epoch": 1.45,
+      "learning_rate": 1.1037791812330372e-05,
+      "loss": 0.5636,
+      "step": 693
+    },
+    {
+      "epoch": 1.45,
+      "learning_rate": 1.1015309834121083e-05,
+      "loss": 0.5613,
+      "step": 694
+    },
+    {
+      "epoch": 1.45,
+      "learning_rate": 1.099282266948141e-05,
+      "loss": 0.5573,
+      "step": 695
+    },
+    {
+      "epoch": 1.46,
+      "learning_rate": 1.0970330433280838e-05,
+      "loss": 0.5554,
+      "step": 696
+    },
+    {
+      "epoch": 1.46,
+      "learning_rate": 1.0947833240414751e-05,
+      "loss": 0.5677,
+      "step": 697
+    },
+    {
+      "epoch": 1.46,
+      "learning_rate": 1.0925331205803861e-05,
+      "loss": 0.5648,
+      "step": 698
+    },
+    {
+      "epoch": 1.46,
+      "learning_rate": 1.0902824444393602e-05,
+      "loss": 0.5476,
+      "step": 699
+    },
+    {
+      "epoch": 1.46,
+      "learning_rate": 1.0880313071153568e-05,
+      "loss": 0.5745,
+      "step": 700
+    },
+    {
+      "epoch": 1.47,
+      "learning_rate": 1.0857797201076897e-05,
+      "loss": 0.5613,
+      "step": 701
+    },
+    {
+      "epoch": 1.47,
+      "learning_rate": 1.0835276949179707e-05,
+      "loss": 0.5563,
+      "step": 702
+    },
+    {
+      "epoch": 1.47,
+      "learning_rate": 1.0812752430500492e-05,
+      "loss": 0.5492,
+      "step": 703
+    },
+    {
+      "epoch": 1.47,
+      "learning_rate": 1.079022376009955e-05,
+      "loss": 0.5606,
+      "step": 704
+    },
+    {
+      "epoch": 1.47,
+      "learning_rate": 1.0767691053058385e-05,
+      "loss": 0.5648,
+      "step": 705
+    },
+    {
+      "epoch": 1.48,
+      "learning_rate": 1.0745154424479112e-05,
+      "loss": 0.5626,
+      "step": 706
+    },
+    {
+      "epoch": 1.48,
+      "learning_rate": 1.0722613989483889e-05,
+      "loss": 0.5533,
+      "step": 707
+    },
+    {
+      "epoch": 1.48,
+      "learning_rate": 1.0700069863214317e-05,
+      "loss": 0.5373,
+      "step": 708
+    },
+    {
+      "epoch": 1.48,
+      "learning_rate": 1.067752216083085e-05,
+      "loss": 0.5522,
+      "step": 709
+    },
+    {
+      "epoch": 1.49,
+      "learning_rate": 1.0654970997512201e-05,
+      "loss": 0.5571,
+      "step": 710
+    },
+    {
+      "epoch": 1.49,
+      "learning_rate": 1.063241648845478e-05,
+      "loss": 0.5669,
+      "step": 711
+    },
+    {
+      "epoch": 1.49,
+      "learning_rate": 1.0609858748872073e-05,
+      "loss": 0.5633,
+      "step": 712
+    },
+    {
+      "epoch": 1.49,
+      "learning_rate": 1.058729789399408e-05,
+      "loss": 0.5628,
+      "step": 713
+    },
+    {
+      "epoch": 1.49,
+      "learning_rate": 1.05647340390667e-05,
+      "loss": 0.5652,
+      "step": 714
+    },
+    {
+      "epoch": 1.5,
+      "learning_rate": 1.0542167299351171e-05,
+      "loss": 0.5621,
+      "step": 715
+    },
+    {
+      "epoch": 1.5,
+      "learning_rate": 1.0519597790123463e-05,
+      "loss": 0.5522,
+      "step": 716
+    },
+    {
+      "epoch": 1.5,
+      "learning_rate": 1.0497025626673687e-05,
+      "loss": 0.569,
+      "step": 717
+    },
+    {
+      "epoch": 1.5,
+      "learning_rate": 1.047445092430552e-05,
+      "loss": 0.5357,
+      "step": 718
+    },
+    {
+      "epoch": 1.5,
+      "learning_rate": 1.0451873798335605e-05,
+      "loss": 0.5501,
+      "step": 719
+    },
+    {
+      "epoch": 1.51,
+      "learning_rate": 1.0429294364092968e-05,
+      "loss": 0.5687,
+      "step": 720
+    },
+    {
+      "epoch": 1.51,
+      "learning_rate": 1.0406712736918418e-05,
+      "loss": 0.5815,
+      "step": 721
+    },
+    {
+      "epoch": 1.51,
+      "learning_rate": 1.0384129032163976e-05,
+      "loss": 0.562,
+      "step": 722
+    },
+    {
+      "epoch": 1.51,
+      "learning_rate": 1.0361543365192275e-05,
+      "loss": 0.5444,
+      "step": 723
+    },
+    {
+      "epoch": 1.51,
+      "learning_rate": 1.0338955851375962e-05,
+      "loss": 0.5569,
+      "step": 724
+    },
+    {
+      "epoch": 1.52,
+      "learning_rate": 1.0316366606097128e-05,
+      "loss": 0.5582,
+      "step": 725
+    },
+    {
+      "epoch": 1.52,
+      "learning_rate": 1.0293775744746705e-05,
+      "loss": 0.5583,
+      "step": 726
+    },
+    {
+      "epoch": 1.52,
+      "learning_rate": 1.0271183382723881e-05,
+      "loss": 0.5532,
+      "step": 727
+    },
+    {
+      "epoch": 1.52,
+      "learning_rate": 1.0248589635435505e-05,
+      "loss": 0.5509,
+      "step": 728
+    },
+    {
+      "epoch": 1.53,
+      "learning_rate": 1.0225994618295507e-05,
+      "loss": 0.5467,
+      "step": 729
+    },
+    {
+      "epoch": 1.53,
+      "learning_rate": 1.0203398446724306e-05,
+      "loss": 0.5683,
+      "step": 730
+    },
+    {
+      "epoch": 1.53,
+      "learning_rate": 1.0180801236148217e-05,
+      "loss": 0.5612,
+      "step": 731
+    },
+    {
+      "epoch": 1.53,
+      "learning_rate": 1.0158203101998854e-05,
+      "loss": 0.5625,
+      "step": 732
+    },
+    {
+      "epoch": 1.53,
+      "learning_rate": 1.0135604159712558e-05,
+      "loss": 0.5673,
+      "step": 733
+    },
+    {
+      "epoch": 1.54,
+      "learning_rate": 1.01130045247298e-05,
+      "loss": 0.5642,
+      "step": 734
+    },
+    {
+      "epoch": 1.54,
+      "learning_rate": 1.009040431249458e-05,
+      "loss": 0.5676,
+      "step": 735
+    },
+    {
+      "epoch": 1.54,
+      "learning_rate": 1.0067803638453847e-05,
+      "loss": 0.5663,
+      "step": 736
+    },
+    {
+      "epoch": 1.54,
+      "learning_rate": 1.0045202618056925e-05,
+      "loss": 0.5549,
+      "step": 737
+    },
+    {
+      "epoch": 1.54,
+      "learning_rate": 1.0022601366754889e-05,
+      "loss": 0.5462,
+      "step": 738
+    },
+    {
+      "epoch": 1.55,
+      "learning_rate": 1e-05,
+      "loss": 0.5611,
+      "step": 739
+    },
+    {
+      "epoch": 1.55,
+      "learning_rate": 9.977398633245116e-06,
+      "loss": 0.5503,
+      "step": 740
+    },
+    {
+      "epoch": 1.55,
+      "learning_rate": 9.954797381943079e-06,
+      "loss": 0.551,
+      "step": 741
+    },
+    {
+      "epoch": 1.55,
+      "learning_rate": 9.932196361546156e-06,
+      "loss": 0.5502,
+      "step": 742
+    },
+    {
+      "epoch": 1.55,
+      "learning_rate": 9.909595687505424e-06,
+      "loss": 0.5645,
+      "step": 743
+    },
+    {
+      "epoch": 1.56,
+      "learning_rate": 9.886995475270205e-06,
+      "loss": 0.5552,
+      "step": 744
+    },
+    {
+      "epoch": 1.56,
+      "learning_rate": 9.864395840287442e-06,
+      "loss": 0.5519,
+      "step": 745
+    },
+    {
+      "epoch": 1.56,
+      "learning_rate": 9.84179689800115e-06,
+      "loss": 0.569,
+      "step": 746
+    },
+    {
+      "epoch": 1.56,
+      "learning_rate": 9.819198763851785e-06,
+      "loss": 0.5684,
+      "step": 747
+    },
+    {
+      "epoch": 1.56,
+      "learning_rate": 9.796601553275697e-06,
+      "loss": 0.5702,
+      "step": 748
+    },
+    {
+      "epoch": 1.57,
+      "learning_rate": 9.774005381704498e-06,
+      "loss": 0.5661,
+      "step": 749
+    },
+    {
+      "epoch": 1.57,
+      "learning_rate": 9.751410364564499e-06,
+      "loss": 0.5479,
+      "step": 750
+    },
+    {
+      "epoch": 1.57,
+      "learning_rate": 9.728816617276124e-06,
+      "loss": 0.5484,
+      "step": 751
+    },
+    {
+      "epoch": 1.57,
+      "learning_rate": 9.706224255253297e-06,
+      "loss": 0.5652,
+      "step": 752
+    },
+    {
+      "epoch": 1.58,
+      "learning_rate": 9.683633393902876e-06,
+      "loss": 0.5735,
+      "step": 753
+    },
+    {
+      "epoch": 1.58,
+      "learning_rate": 9.661044148624038e-06,
+      "loss": 0.5518,
+      "step": 754
+    },
+    {
+      "epoch": 1.58,
+      "learning_rate": 9.638456634807728e-06,
+      "loss": 0.5407,
+      "step": 755
+    },
+    {
+      "epoch": 1.58,
+      "learning_rate": 9.615870967836026e-06,
+      "loss": 0.5642,
+      "step": 756
+    },
+    {
+      "epoch": 1.58,
+      "learning_rate": 9.593287263081586e-06,
+      "loss": 0.5612,
+      "step": 757
+    },
+    {
+      "epoch": 1.59,
+      "learning_rate": 9.570705635907038e-06,
+      "loss": 0.5736,
+      "step": 758
+    },
+    {
+      "epoch": 1.59,
+      "learning_rate": 9.548126201664398e-06,
+      "loss": 0.5603,
+      "step": 759
+    },
+    {
+      "epoch": 1.59,
+      "learning_rate": 9.525549075694484e-06,
+      "loss": 0.5587,
+      "step": 760
+    },
+    {
+      "epoch": 1.59,
+      "learning_rate": 9.502974373326315e-06,
+      "loss": 0.5611,
+      "step": 761
+    },
+    {
+      "epoch": 1.59,
+      "learning_rate": 9.48040220987654e-06,
+      "loss": 0.5577,
+      "step": 762
+    },
+    {
+      "epoch": 1.6,
+      "learning_rate": 9.45783270064883e-06,
+      "loss": 0.5634,
+      "step": 763
+    },
+    {
+      "epoch": 1.6,
+      "learning_rate": 9.435265960933304e-06,
+      "loss": 0.5515,
+      "step": 764
+    },
+    {
+      "epoch": 1.6,
+      "learning_rate": 9.412702106005927e-06,
+      "loss": 0.5542,
+      "step": 765
+    },
+    {
+      "epoch": 1.6,
+      "learning_rate": 9.39014125112793e-06,
+      "loss": 0.5555,
+      "step": 766
+    },
+    {
+      "epoch": 1.6,
+      "learning_rate": 9.367583511545225e-06,
+      "loss": 0.5602,
+      "step": 767
+    },
+    {
+      "epoch": 1.61,
+      "learning_rate": 9.3450290024878e-06,
+      "loss": 0.5595,
+      "step": 768
+    },
+    {
+      "epoch": 1.61,
+      "learning_rate": 9.322477839169156e-06,
+      "loss": 0.561,
+      "step": 769
+    },
+    {
+      "epoch": 1.61,
+      "learning_rate": 9.299930136785685e-06,
+      "loss": 0.5517,
+      "step": 770
+    },
+    {
+      "epoch": 1.61,
+      "learning_rate": 9.277386010516113e-06,
+      "loss": 0.5496,
+      "step": 771
+    },
+    {
+      "epoch": 1.62,
+      "learning_rate": 9.25484557552089e-06,
+      "loss": 0.5579,
+      "step": 772
+    },
+    {
+      "epoch": 1.62,
+      "learning_rate": 9.232308946941618e-06,
+      "loss": 0.5648,
+      "step": 773
+    },
+    {
+      "epoch": 1.62,
+      "learning_rate": 9.209776239900453e-06,
+      "loss": 0.5471,
+      "step": 774
+    },
+    {
+      "epoch": 1.62,
+      "learning_rate": 9.187247569499511e-06,
+      "loss": 0.5511,
+      "step": 775
+    },
+    {
+      "epoch": 1.62,
+      "learning_rate": 9.164723050820298e-06,
+      "loss": 0.57,
+      "step": 776
+    },
+    {
+      "epoch": 1.63,
+      "learning_rate": 9.142202798923106e-06,
+      "loss": 0.564,
+      "step": 777
+    },
+    {
+      "epoch": 1.63,
+      "learning_rate": 9.119686928846437e-06,
+      "loss": 0.5496,
+      "step": 778
+    },
+    {
+      "epoch": 1.63,
+      "learning_rate": 9.097175555606396e-06,
+      "loss": 0.5619,
+      "step": 779
+    },
+    {
+      "epoch": 1.63,
+      "learning_rate": 9.074668794196142e-06,
+      "loss": 0.5417,
+      "step": 780
+    },
+    {
+      "epoch": 1.63,
+      "learning_rate": 9.052166759585252e-06,
+      "loss": 0.5444,
+      "step": 781
+    },
+    {
+      "epoch": 1.64,
+      "learning_rate": 9.029669566719165e-06,
+      "loss": 0.5628,
+      "step": 782
+    },
+    {
+      "epoch": 1.64,
+      "learning_rate": 9.007177330518595e-06,
+      "loss": 0.5571,
+      "step": 783
+    },
+    {
+      "epoch": 1.64,
+      "learning_rate": 8.98469016587892e-06,
+      "loss": 0.5668,
+      "step": 784
+    },
+    {
+      "epoch": 1.64,
+      "learning_rate": 8.962208187669633e-06,
+      "loss": 0.5576,
+      "step": 785
+    },
+    {
+      "epoch": 1.64,
+      "learning_rate": 8.939731510733711e-06,
+      "loss": 0.5549,
+      "step": 786
+    },
+    {
+      "epoch": 1.65,
+      "learning_rate": 8.917260249887072e-06,
+      "loss": 0.5657,
+      "step": 787
+    },
+    {
+      "epoch": 1.65,
+      "learning_rate": 8.894794519917947e-06,
+      "loss": 0.5605,
+      "step": 788
+    },
+    {
+      "epoch": 1.65,
+      "learning_rate": 8.872334435586333e-06,
+      "loss": 0.571,
+      "step": 789
+    },
+    {
+      "epoch": 1.65,
+      "learning_rate": 8.849880111623374e-06,
+      "loss": 0.5603,
+      "step": 790
+    },
+    {
+      "epoch": 1.65,
+      "learning_rate": 8.82743166273079e-06,
+      "loss": 0.5552,
+      "step": 791
+    },
+    {
+      "epoch": 1.66,
+      "learning_rate": 8.804989203580303e-06,
+      "loss": 0.5578,
+      "step": 792
+    },
+    {
+      "epoch": 1.66,
+      "learning_rate": 8.78255284881302e-06,
+      "loss": 0.563,
+      "step": 793
+    },
+    {
+      "epoch": 1.66,
+      "learning_rate": 8.76012271303888e-06,
+      "loss": 0.559,
+      "step": 794
+    },
+    {
+      "epoch": 1.66,
+      "learning_rate": 8.737698910836045e-06,
+      "loss": 0.5568,
+      "step": 795
+    },
+    {
+      "epoch": 1.67,
+      "learning_rate": 8.715281556750327e-06,
+      "loss": 0.5481,
+      "step": 796
+    },
+    {
+      "epoch": 1.67,
+      "learning_rate": 8.692870765294595e-06,
+      "loss": 0.5539,
+      "step": 797
+    },
+    {
+      "epoch": 1.67,
+      "learning_rate": 8.670466650948208e-06,
+      "loss": 0.5635,
+      "step": 798
+    },
+    {
+      "epoch": 1.67,
+      "learning_rate": 8.648069328156403e-06,
+      "loss": 0.5537,
+      "step": 799
+    },
+    {
+      "epoch": 1.67,
+      "learning_rate": 8.625678911329727e-06,
+      "loss": 0.5705,
+      "step": 800
+    },
+    {
+      "epoch": 1.68,
+      "learning_rate": 8.603295514843453e-06,
+      "loss": 0.5632,
+      "step": 801
+    },
+    {
+      "epoch": 1.68,
+      "learning_rate": 8.580919253036991e-06,
+      "loss": 0.5627,
+      "step": 802
+    },
+    {
+      "epoch": 1.68,
+      "learning_rate": 8.558550240213309e-06,
+      "loss": 0.5578,
+      "step": 803
+    },
+    {
+      "epoch": 1.68,
+      "learning_rate": 8.536188590638334e-06,
+      "loss": 0.5532,
+      "step": 804
+    },
+    {
+      "epoch": 1.68,
+      "learning_rate": 8.513834418540398e-06,
+      "loss": 0.5555,
+      "step": 805
+    },
+    {
+      "epoch": 1.69,
+      "learning_rate": 8.491487838109622e-06,
+      "loss": 0.5391,
+      "step": 806
+    },
+    {
+      "epoch": 1.69,
+      "learning_rate": 8.46914896349735e-06,
+      "loss": 0.581,
+      "step": 807
+    },
+    {
+      "epoch": 1.69,
+      "learning_rate": 8.446817908815566e-06,
+      "loss": 0.5506,
+      "step": 808
+    },
+    {
+      "epoch": 1.69,
+      "learning_rate": 8.424494788136303e-06,
+      "loss": 0.5536,
+      "step": 809
+    },
+    {
+      "epoch": 1.69,
+      "learning_rate": 8.402179715491078e-06,
+      "loss": 0.55,
+      "step": 810
+    },
+    {
+      "epoch": 1.7,
+      "learning_rate": 8.379872804870277e-06,
+      "loss": 0.5485,
+      "step": 811
+    },
+    {
+      "epoch": 1.7,
+      "learning_rate": 8.357574170222612e-06,
+      "loss": 0.5487,
+      "step": 812
+    },
+    {
+      "epoch": 1.7,
+      "learning_rate": 8.335283925454505e-06,
+      "loss": 0.5595,
+      "step": 813
+    },
+    {
+      "epoch": 1.7,
+      "learning_rate": 8.313002184429529e-06,
+      "loss": 0.555,
+      "step": 814
+    },
+    {
+      "epoch": 1.71,
+      "learning_rate": 8.290729060967817e-06,
+      "loss": 0.5596,
+      "step": 815
+    },
+    {
+      "epoch": 1.71,
+      "learning_rate": 8.268464668845471e-06,
+      "loss": 0.5579,
+      "step": 816
+    },
+    {
+      "epoch": 1.71,
+      "learning_rate": 8.24620912179401e-06,
+      "loss": 0.5627,
+      "step": 817
+    },
+    {
+      "epoch": 1.71,
+      "learning_rate": 8.223962533499757e-06,
+      "loss": 0.5463,
+      "step": 818
+    },
+    {
+      "epoch": 1.71,
+      "learning_rate": 8.201725017603277e-06,
+      "loss": 0.55,
+      "step": 819
+    },
+    {
+      "epoch": 1.72,
+      "learning_rate": 8.179496687698785e-06,
+      "loss": 0.5648,
+      "step": 820
+    },
+    {
+      "epoch": 1.72,
+      "learning_rate": 8.157277657333582e-06,
+      "loss": 0.5691,
+      "step": 821
+    },
+    {
+      "epoch": 1.72,
+      "learning_rate": 8.135068040007452e-06,
+      "loss": 0.5419,
+      "step": 822
+    },
+    {
+      "epoch": 1.72,
+      "learning_rate": 8.11286794917211e-06,
+      "loss": 0.5542,
+      "step": 823
+    },
+    {
+      "epoch": 1.72,
+      "learning_rate": 8.090677498230598e-06,
+      "loss": 0.5499,
+      "step": 824
+    },
+    {
+      "epoch": 1.73,
+      "learning_rate": 8.068496800536712e-06,
+      "loss": 0.5628,
+      "step": 825
+    },
+    {
+      "epoch": 1.73,
+      "learning_rate": 8.046325969394437e-06,
+      "loss": 0.5553,
+      "step": 826
+    },
+    {
+      "epoch": 1.73,
+      "learning_rate": 8.024165118057344e-06,
+      "loss": 0.5604,
+      "step": 827
+    },
+    {
+      "epoch": 1.73,
+      "learning_rate": 8.002014359728046e-06,
+      "loss": 0.5499,
+      "step": 828
+    },
+    {
+      "epoch": 1.73,
+      "learning_rate": 7.97987380755757e-06,
+      "loss": 0.543,
+      "step": 829
+    },
+    {
+      "epoch": 1.74,
+      "learning_rate": 7.957743574644837e-06,
+      "loss": 0.5604,
+      "step": 830
+    },
+    {
+      "epoch": 1.74,
+      "learning_rate": 7.93562377403603e-06,
+      "loss": 0.5536,
+      "step": 831
+    },
+    {
+      "epoch": 1.74,
+      "learning_rate": 7.913514518724059e-06,
+      "loss": 0.55,
+      "step": 832
+    },
+    {
+      "epoch": 1.74,
+      "learning_rate": 7.891415921647957e-06,
+      "loss": 0.5581,
+      "step": 833
+    },
+    {
+      "epoch": 1.74,
+      "learning_rate": 7.869328095692313e-06,
+      "loss": 0.5577,
+      "step": 834
+    },
+    {
+      "epoch": 1.75,
+      "learning_rate": 7.847251153686701e-06,
+      "loss": 0.5708,
+      "step": 835
+    },
+    {
+      "epoch": 1.75,
+      "learning_rate": 7.825185208405089e-06,
+      "loss": 0.5435,
+      "step": 836
+    },
+    {
+      "epoch": 1.75,
+      "learning_rate": 7.803130372565277e-06,
+      "loss": 0.5458,
+      "step": 837
+    },
+    {
+      "epoch": 1.75,
+      "learning_rate": 7.781086758828314e-06,
+      "loss": 0.5459,
+      "step": 838
+    },
+    {
+      "epoch": 1.76,
+      "learning_rate": 7.759054479797924e-06,
+      "loss": 0.5518,
+      "step": 839
+    },
+    {
+      "epoch": 1.76,
+      "learning_rate": 7.73703364801993e-06,
+      "loss": 0.5463,
+      "step": 840
+    },
+    {
+      "epoch": 1.76,
+      "learning_rate": 7.715024375981676e-06,
+      "loss": 0.5474,
+      "step": 841
+    },
+    {
+      "epoch": 1.76,
+      "learning_rate": 7.69302677611147e-06,
+      "loss": 0.5531,
+      "step": 842
+    },
+    {
+      "epoch": 1.76,
+      "learning_rate": 7.671040960777977e-06,
+      "loss": 0.5496,
+      "step": 843
+    },
+    {
+      "epoch": 1.77,
+      "learning_rate": 7.649067042289681e-06,
+      "loss": 0.5449,
+      "step": 844
+    },
+    {
+      "epoch": 1.77,
+      "learning_rate": 7.627105132894279e-06,
+      "loss": 0.5526,
+      "step": 845
+    },
+    {
+      "epoch": 1.77,
+      "learning_rate": 7.6051553447781415e-06,
+      "loss": 0.5527,
+      "step": 846
+    },
+    {
+      "epoch": 1.77,
+      "learning_rate": 7.583217790065697e-06,
+      "loss": 0.5473,
+      "step": 847
+    },
+    {
+      "epoch": 1.77,
+      "learning_rate": 7.561292580818906e-06,
+      "loss": 0.5564,
+      "step": 848
+    },
+    {
+      "epoch": 1.78,
+      "learning_rate": 7.539379829036652e-06,
+      "loss": 0.5372,
+      "step": 849
+    },
+    {
+      "epoch": 1.78,
+      "learning_rate": 7.517479646654184e-06,
+      "loss": 0.5572,
+      "step": 850
+    },
+    {
+      "epoch": 1.78,
+      "learning_rate": 7.4955921455425505e-06,
+      "loss": 0.5707,
+      "step": 851
+    },
+    {
+      "epoch": 1.78,
+      "learning_rate": 7.47371743750801e-06,
+      "loss": 0.553,
+      "step": 852
+    },
+    {
+      "epoch": 1.78,
+      "learning_rate": 7.451855634291488e-06,
+      "loss": 0.5574,
+      "step": 853
+    },
+    {
+      "epoch": 1.79,
+      "learning_rate": 7.430006847567972e-06,
+      "loss": 0.5297,
+      "step": 854
+    },
+    {
+      "epoch": 1.79,
+      "learning_rate": 7.4081711889459695e-06,
+      "loss": 0.5561,
+      "step": 855
+    },
+    {
+      "epoch": 1.79,
+      "learning_rate": 7.386348769966918e-06,
+      "loss": 0.547,
+      "step": 856
+    },
+    {
+      "epoch": 1.79,
+      "learning_rate": 7.364539702104635e-06,
+      "loss": 0.5486,
+      "step": 857
+    },
+    {
+      "epoch": 1.79,
+      "learning_rate": 7.342744096764727e-06,
+      "loss": 0.5511,
+      "step": 858
+    },
+    {
+      "epoch": 1.8,
+      "learning_rate": 7.320962065284032e-06,
+      "loss": 0.5406,
+      "step": 859
+    },
+    {
+      "epoch": 1.8,
+      "learning_rate": 7.299193718930062e-06,
+      "loss": 0.5668,
+      "step": 860
+    },
+    {
+      "epoch": 1.8,
+      "learning_rate": 7.277439168900404e-06,
+      "loss": 0.5645,
+      "step": 861
+    },
+    {
+      "epoch": 1.8,
+      "learning_rate": 7.255698526322188e-06,
+      "loss": 0.551,
+      "step": 862
+    },
+    {
+      "epoch": 1.81,
+      "learning_rate": 7.233971902251483e-06,
+      "loss": 0.5564,
+      "step": 863
+    },
+    {
+      "epoch": 1.81,
+      "learning_rate": 7.2122594076727705e-06,
+      "loss": 0.5549,
+      "step": 864
+    },
+    {
+      "epoch": 1.81,
+      "learning_rate": 7.190561153498334e-06,
+      "loss": 0.5553,
+      "step": 865
+    },
+    {
+      "epoch": 1.81,
+      "learning_rate": 7.1688772505677225e-06,
+      "loss": 0.5517,
+      "step": 866
+    },
+    {
+      "epoch": 1.81,
+      "learning_rate": 7.147207809647183e-06,
+      "loss": 0.5609,
+      "step": 867
+    },
+    {
+      "epoch": 1.82,
+      "learning_rate": 7.125552941429077e-06,
+      "loss": 0.5504,
+      "step": 868
+    },
+    {
+      "epoch": 1.82,
+      "learning_rate": 7.1039127565313285e-06,
+      "loss": 0.548,
+      "step": 869
+    },
+    {
+      "epoch": 1.82,
+      "learning_rate": 7.082287365496852e-06,
+      "loss": 0.5437,
+      "step": 870
+    },
+    {
+      "epoch": 1.82,
+      "learning_rate": 7.060676878793007e-06,
+      "loss": 0.5577,
+      "step": 871
+    },
+    {
+      "epoch": 1.82,
+      "learning_rate": 7.0390814068109965e-06,
+      "loss": 0.5808,
+      "step": 872
+    },
+    {
+      "epoch": 1.83,
+      "learning_rate": 7.0175010598653414e-06,
+      "loss": 0.5452,
+      "step": 873
+    },
+    {
+      "epoch": 1.83,
+      "learning_rate": 6.995935948193294e-06,
+      "loss": 0.5455,
+      "step": 874
+    },
+    {
+      "epoch": 1.83,
+      "learning_rate": 6.9743861819542794e-06,
+      "loss": 0.5358,
+      "step": 875
+    },
+    {
+      "epoch": 1.83,
+      "learning_rate": 6.9528518712293405e-06,
+      "loss": 0.5519,
+      "step": 876
+    },
+    {
+      "epoch": 1.83,
+      "learning_rate": 6.931333126020561e-06,
+      "loss": 0.554,
+      "step": 877
+    },
+    {
+      "epoch": 1.84,
+      "learning_rate": 6.909830056250527e-06,
+      "loss": 0.5583,
+      "step": 878
+    },
+    {
+      "epoch": 1.84,
+      "learning_rate": 6.888342771761737e-06,
+      "loss": 0.5485,
+      "step": 879
+    },
+    {
+      "epoch": 1.84,
+      "learning_rate": 6.866871382316063e-06,
+      "loss": 0.5513,
+      "step": 880
+    },
+    {
+      "epoch": 1.84,
+      "learning_rate": 6.8454159975941725e-06,
+      "loss": 0.5478,
+      "step": 881
+    },
+    {
+      "epoch": 1.85,
+      "learning_rate": 6.823976727194994e-06,
+      "loss": 0.546,
+      "step": 882
+    },
+    {
+      "epoch": 1.85,
+      "learning_rate": 6.8025536806351235e-06,
+      "loss": 0.5505,
+      "step": 883
+    },
+    {
+      "epoch": 1.85,
+      "learning_rate": 6.781146967348283e-06,
+      "loss": 0.5433,
+      "step": 884
+    },
+    {
+      "epoch": 1.85,
+      "learning_rate": 6.759756696684776e-06,
+      "loss": 0.5571,
+      "step": 885
+    },
+    {
+      "epoch": 1.85,
+      "learning_rate": 6.738382977910898e-06,
+      "loss": 0.5525,
+      "step": 886
+    },
+    {
+      "epoch": 1.86,
+      "learning_rate": 6.7170259202084e-06,
+      "loss": 0.5405,
+      "step": 887
+    },
+    {
+      "epoch": 1.86,
+      "learning_rate": 6.695685632673919e-06,
+      "loss": 0.5536,
+      "step": 888
+    },
+    {
+      "epoch": 1.86,
+      "learning_rate": 6.6743622243184405e-06,
+      "loss": 0.553,
+      "step": 889
+    },
+    {
+      "epoch": 1.86,
+      "learning_rate": 6.653055804066712e-06,
+      "loss": 0.5566,
+      "step": 890
+    },
+    {
+      "epoch": 1.86,
+      "learning_rate": 6.631766480756703e-06,
+      "loss": 0.5568,
+      "step": 891
+    },
+    {
+      "epoch": 1.87,
+      "learning_rate": 6.6104943631390596e-06,
+      "loss": 0.556,
+      "step": 892
+    },
+    {
+      "epoch": 1.87,
+      "learning_rate": 6.589239559876525e-06,
+      "loss": 0.5601,
+      "step": 893
+    },
+    {
+      "epoch": 1.87,
+      "learning_rate": 6.568002179543409e-06,
+      "loss": 0.5595,
+      "step": 894
+    },
+    {
+      "epoch": 1.87,
+      "learning_rate": 6.546782330625004e-06,
+      "loss": 0.5639,
+      "step": 895
+    },
+    {
+      "epoch": 1.87,
+      "learning_rate": 6.525580121517069e-06,
+      "loss": 0.5441,
+      "step": 896
+    },
+    {
+      "epoch": 1.88,
+      "learning_rate": 6.504395660525236e-06,
+      "loss": 0.5575,
+      "step": 897
+    },
+    {
+      "epoch": 1.88,
+      "learning_rate": 6.48322905586449e-06,
+      "loss": 0.5585,
+      "step": 898
+    },
+    {
+      "epoch": 1.88,
+      "learning_rate": 6.462080415658591e-06,
+      "loss": 0.5612,
+      "step": 899
+    },
+    {
+      "epoch": 1.88,
+      "learning_rate": 6.440949847939538e-06,
+      "loss": 0.5473,
+      "step": 900
+    },
+    {
+      "epoch": 1.88,
+      "learning_rate": 6.419837460647014e-06,
+      "loss": 0.5421,
+      "step": 901
+    },
+    {
+      "epoch": 1.89,
+      "learning_rate": 6.39874336162782e-06,
+      "loss": 0.542,
+      "step": 902
+    },
+    {
+      "epoch": 1.89,
+      "learning_rate": 6.377667658635358e-06,
+      "loss": 0.5466,
+      "step": 903
+    },
+    {
+      "epoch": 1.89,
+      "learning_rate": 6.356610459329038e-06,
+      "loss": 0.5661,
+      "step": 904
+    },
+    {
+      "epoch": 1.89,
+      "learning_rate": 6.335571871273762e-06,
+      "loss": 0.5504,
+      "step": 905
+    },
+    {
+      "epoch": 1.9,
+      "learning_rate": 6.314552001939351e-06,
+      "loss": 0.5533,
+      "step": 906
+    },
+    {
+      "epoch": 1.9,
+      "learning_rate": 6.293550958700021e-06,
+      "loss": 0.5505,
+      "step": 907
+    },
+    {
+      "epoch": 1.9,
+      "learning_rate": 6.272568848833809e-06,
+      "loss": 0.556,
+      "step": 908
+    },
+    {
+      "epoch": 1.9,
+      "learning_rate": 6.251605779522032e-06,
+      "loss": 0.5537,
+      "step": 909
+    },
+    {
+      "epoch": 1.9,
+      "learning_rate": 6.230661857848759e-06,
+      "loss": 0.5521,
+      "step": 910
+    },
+    {
+      "epoch": 1.91,
+      "learning_rate": 6.209737190800235e-06,
+      "loss": 0.544,
+      "step": 911
+    },
+    {
+      "epoch": 1.91,
+      "learning_rate": 6.188831885264357e-06,
+      "loss": 0.5516,
+      "step": 912
+    },
+    {
+      "epoch": 1.91,
+      "learning_rate": 6.167946048030107e-06,
+      "loss": 0.5547,
+      "step": 913
+    },
+    {
+      "epoch": 1.91,
+      "learning_rate": 6.147079785787038e-06,
+      "loss": 0.5685,
+      "step": 914
+    },
+    {
+      "epoch": 1.91,
+      "learning_rate": 6.126233205124692e-06,
+      "loss": 0.5523,
+      "step": 915
+    },
+    {
+      "epoch": 1.92,
+      "learning_rate": 6.105406412532078e-06,
+      "loss": 0.5459,
+      "step": 916
+    },
+    {
+      "epoch": 1.92,
+      "learning_rate": 6.084599514397127e-06,
+      "loss": 0.5669,
+      "step": 917
+    },
+    {
+      "epoch": 1.92,
+      "learning_rate": 6.06381261700614e-06,
+      "loss": 0.5329,
+      "step": 918
+    },
+    {
+      "epoch": 1.92,
+      "learning_rate": 6.043045826543254e-06,
+      "loss": 0.5512,
+      "step": 919
+    },
+    {
+      "epoch": 1.92,
+      "learning_rate": 6.022299249089889e-06,
+      "loss": 0.5564,
+      "step": 920
+    },
+    {
+      "epoch": 1.93,
+      "learning_rate": 6.001572990624222e-06,
+      "loss": 0.5562,
+      "step": 921
+    },
+    {
+      "epoch": 1.93,
+      "learning_rate": 5.980867157020624e-06,
+      "loss": 0.5575,
+      "step": 922
+    },
+    {
+      "epoch": 1.93,
+      "learning_rate": 5.960181854049146e-06,
+      "loss": 0.5634,
+      "step": 923
+    },
+    {
+      "epoch": 1.93,
+      "learning_rate": 5.93951718737495e-06,
+      "loss": 0.5555,
+      "step": 924
+    },
+    {
+      "epoch": 1.94,
+      "learning_rate": 5.918873262557785e-06,
+      "loss": 0.5478,
+      "step": 925
+    },
+    {
+      "epoch": 1.94,
+      "learning_rate": 5.8982501850514614e-06,
+      "loss": 0.5627,
+      "step": 926
+    },
+    {
+      "epoch": 1.94,
+      "learning_rate": 5.8776480602032715e-06,
+      "loss": 0.5545,
+      "step": 927
+    },
+    {
+      "epoch": 1.94,
+      "learning_rate": 5.857066993253501e-06,
+      "loss": 0.5373,
+      "step": 928
+    },
+    {
+      "epoch": 1.94,
+      "learning_rate": 5.836507089334849e-06,
+      "loss": 0.5583,
+      "step": 929
+    },
+    {
+      "epoch": 1.95,
+      "learning_rate": 5.815968453471923e-06,
+      "loss": 0.5566,
+      "step": 930
+    },
+    {
+      "epoch": 1.95,
+      "learning_rate": 5.79545119058068e-06,
+      "loss": 0.5516,
+      "step": 931
+    },
+    {
+      "epoch": 1.95,
+      "learning_rate": 5.7749554054679015e-06,
+      "loss": 0.5609,
+      "step": 932
+    },
+    {
+      "epoch": 1.95,
+      "learning_rate": 5.754481202830653e-06,
+      "loss": 0.5712,
+      "step": 933
+    },
+    {
+      "epoch": 1.95,
+      "learning_rate": 5.7340286872557515e-06,
+      "loss": 0.5548,
+      "step": 934
+    },
+    {
+      "epoch": 1.96,
+      "learning_rate": 5.713597963219243e-06,
+      "loss": 0.5436,
+      "step": 935
+    },
+    {
+      "epoch": 1.96,
+      "learning_rate": 5.693189135085839e-06,
+      "loss": 0.5478,
+      "step": 936
+    },
+    {
+      "epoch": 1.96,
+      "learning_rate": 5.672802307108423e-06,
+      "loss": 0.5521,
+      "step": 937
+    },
+    {
+      "epoch": 1.96,
+      "learning_rate": 5.652437583427478e-06,
+      "loss": 0.541,
+      "step": 938
+    },
+    {
+      "epoch": 1.96,
+      "learning_rate": 5.6320950680705826e-06,
+      "loss": 0.5485,
+      "step": 939
+    },
+    {
+      "epoch": 1.97,
+      "learning_rate": 5.6117748649518665e-06,
+      "loss": 0.5548,
+      "step": 940
+    },
+    {
+      "epoch": 1.97,
+      "learning_rate": 5.591477077871485e-06,
+      "loss": 0.5529,
+      "step": 941
+    },
+    {
+      "epoch": 1.97,
+      "learning_rate": 5.5712018105150914e-06,
+      "loss": 0.5494,
+      "step": 942
+    },
+    {
+      "epoch": 1.97,
+      "learning_rate": 5.550949166453293e-06,
+      "loss": 0.5433,
+      "step": 943
+    },
+    {
+      "epoch": 1.97,
+      "learning_rate": 5.530719249141148e-06,
+      "loss": 0.5588,
+      "step": 944
+    },
+    {
+      "epoch": 1.98,
+      "learning_rate": 5.5105121619176064e-06,
+      "loss": 0.5451,
+      "step": 945
+    },
+    {
+      "epoch": 1.98,
+      "learning_rate": 5.490328008005002e-06,
+      "loss": 0.5592,
+      "step": 946
+    },
+    {
+      "epoch": 1.98,
+      "learning_rate": 5.470166890508519e-06,
+      "loss": 0.5552,
+      "step": 947
+    },
+    {
+      "epoch": 1.98,
+      "learning_rate": 5.450028912415672e-06,
+      "loss": 0.5309,
+      "step": 948
+    },
+    {
+      "epoch": 1.99,
+      "learning_rate": 5.429914176595772e-06,
+      "loss": 0.5643,
+      "step": 949
+    },
+    {
+      "epoch": 1.99,
+      "learning_rate": 5.409822785799393e-06,
+      "loss": 0.5481,
+      "step": 950
+    },
+    {
+      "epoch": 1.99,
+      "learning_rate": 5.389754842657875e-06,
+      "loss": 0.5503,
+      "step": 951
+    },
+    {
+      "epoch": 1.99,
+      "learning_rate": 5.369710449682767e-06,
+      "loss": 0.5545,
+      "step": 952
+    },
+    {
+      "epoch": 1.99,
+      "learning_rate": 5.3496897092653335e-06,
+      "loss": 0.544,
+      "step": 953
+    },
+    {
+      "epoch": 2.0,
+      "learning_rate": 5.329692723675994e-06,
+      "loss": 0.5395,
+      "step": 954
+    },
+    {
+      "epoch": 2.0,
+      "learning_rate": 5.309719595063848e-06,
+      "loss": 0.5429,
+      "step": 955
+    },
+    {
+      "epoch": 2.0,
+      "learning_rate": 5.289770425456109e-06,
+      "loss": 0.4955,
+      "step": 956
+    },
+    {
+      "epoch": 2.0,
+      "learning_rate": 5.2698453167576155e-06,
+      "loss": 0.4129,
+      "step": 957
+    },
+    {
+      "epoch": 2.0,
+      "learning_rate": 5.249944370750293e-06,
+      "loss": 0.4203,
+      "step": 958
+    },
+    {
+      "epoch": 2.01,
+      "learning_rate": 5.230067689092629e-06,
+      "loss": 0.4053,
+      "step": 959
+    },
+    {
+      "epoch": 2.01,
+      "learning_rate": 5.210215373319183e-06,
+      "loss": 0.404,
+      "step": 960
+    },
+    {
+      "epoch": 2.01,
+      "learning_rate": 5.190387524840033e-06,
+      "loss": 0.3867,
+      "step": 961
+    },
+    {
+      "epoch": 2.01,
+      "learning_rate": 5.170584244940275e-06,
+      "loss": 0.4151,
+      "step": 962
+    },
+    {
+      "epoch": 2.01,
+      "learning_rate": 5.150805634779505e-06,
+      "loss": 0.4011,
+      "step": 963
+    },
+    {
+      "epoch": 2.02,
+      "learning_rate": 5.131051795391302e-06,
+      "loss": 0.3939,
+      "step": 964
+    },
+    {
+      "epoch": 2.02,
+      "learning_rate": 5.111322827682706e-06,
+      "loss": 0.3899,
+      "step": 965
+    },
+    {
+      "epoch": 2.02,
+      "learning_rate": 5.091618832433716e-06,
+      "loss": 0.3869,
+      "step": 966
+    },
+    {
+      "epoch": 2.02,
+      "learning_rate": 5.0719399102967536e-06,
+      "loss": 0.3885,
+      "step": 967
+    },
+    {
+      "epoch": 2.03,
+      "learning_rate": 5.0522861617961694e-06,
+      "loss": 0.3877,
+      "step": 968
+    },
+    {
+      "epoch": 2.03,
+      "learning_rate": 5.03265768732772e-06,
+      "loss": 0.404,
+      "step": 969
+    },
+    {
+      "epoch": 2.03,
+      "learning_rate": 5.0130545871580504e-06,
+      "loss": 0.3904,
+      "step": 970
+    },
+    {
+      "epoch": 2.03,
+      "learning_rate": 4.993476961424203e-06,
+      "loss": 0.4024,
+      "step": 971
+    },
+    {
+      "epoch": 2.03,
+      "learning_rate": 4.973924910133071e-06,
+      "loss": 0.3832,
+      "step": 972
+    },
+    {
+      "epoch": 2.04,
+      "learning_rate": 4.95439853316093e-06,
+      "loss": 0.4054,
+      "step": 973
+    },
+    {
+      "epoch": 2.04,
+      "learning_rate": 4.934897930252887e-06,
+      "loss": 0.3792,
+      "step": 974
+    },
+    {
+      "epoch": 2.04,
+      "learning_rate": 4.915423201022396e-06,
+      "loss": 0.3922,
+      "step": 975
+    },
+    {
+      "epoch": 2.04,
+      "learning_rate": 4.895974444950743e-06,
+      "loss": 0.3994,
+      "step": 976
+    },
+    {
+      "epoch": 2.04,
+      "learning_rate": 4.876551761386532e-06,
+      "loss": 0.3937,
+      "step": 977
+    },
+    {
+      "epoch": 2.05,
+      "learning_rate": 4.857155249545197e-06,
+      "loss": 0.3849,
+      "step": 978
+    },
+    {
+      "epoch": 2.05,
+      "learning_rate": 4.837785008508462e-06,
+      "loss": 0.3976,
+      "step": 979
+    },
+    {
+      "epoch": 2.05,
+      "learning_rate": 4.8184411372238724e-06,
+      "loss": 0.3929,
+      "step": 980
+    },
+    {
+      "epoch": 2.05,
+      "learning_rate": 4.799123734504258e-06,
+      "loss": 0.3868,
+      "step": 981
+    },
+    {
+      "epoch": 2.05,
+      "learning_rate": 4.779832899027243e-06,
+      "loss": 0.3957,
+      "step": 982
+    },
+    {
+      "epoch": 2.06,
+      "learning_rate": 4.7605687293347425e-06,
+      "loss": 0.3872,
+      "step": 983
+    },
+    {
+      "epoch": 2.06,
+      "learning_rate": 4.7413313238324556e-06,
+      "loss": 0.3951,
+      "step": 984
+    },
+    {
+      "epoch": 2.06,
+      "learning_rate": 4.722120780789369e-06,
+      "loss": 0.387,
+      "step": 985
+    },
+    {
+      "epoch": 2.06,
+      "learning_rate": 4.702937198337241e-06,
+      "loss": 0.3973,
+      "step": 986
+    },
+    {
+      "epoch": 2.06,
+      "learning_rate": 4.68378067447012e-06,
+      "loss": 0.3842,
+      "step": 987
+    },
+    {
+      "epoch": 2.07,
+      "learning_rate": 4.66465130704382e-06,
+      "loss": 0.4018,
+      "step": 988
+    },
+    {
+      "epoch": 2.07,
+      "learning_rate": 4.645549193775452e-06,
+      "loss": 0.4036,
+      "step": 989
+    },
+    {
+      "epoch": 2.07,
+      "learning_rate": 4.626474432242879e-06,
+      "loss": 0.3831,
+      "step": 990
+    },
+    {
+      "epoch": 2.07,
+      "learning_rate": 4.607427119884276e-06,
+      "loss": 0.3993,
+      "step": 991
+    },
+    {
+      "epoch": 2.08,
+      "learning_rate": 4.58840735399758e-06,
+      "loss": 0.3782,
+      "step": 992
+    },
+    {
+      "epoch": 2.08,
+      "learning_rate": 4.569415231740021e-06,
+      "loss": 0.3948,
+      "step": 993
+    },
+    {
+      "epoch": 2.08,
+      "learning_rate": 4.550450850127626e-06,
+      "loss": 0.3754,
+      "step": 994
+    },
+    {
+      "epoch": 2.08,
+      "learning_rate": 4.531514306034701e-06,
+      "loss": 0.3965,
+      "step": 995
+    },
+    {
+      "epoch": 2.08,
+      "learning_rate": 4.512605696193371e-06,
+      "loss": 0.3768,
+      "step": 996
+    },
+    {
+      "epoch": 2.09,
+      "learning_rate": 4.493725117193051e-06,
+      "loss": 0.3924,
+      "step": 997
+    },
+    {
+      "epoch": 2.09,
+      "learning_rate": 4.474872665479974e-06,
+      "loss": 0.408,
+      "step": 998
+    },
+    {
+      "epoch": 2.09,
+      "learning_rate": 4.4560484373566945e-06,
+      "loss": 0.397,
+      "step": 999
+    },
+    {
+      "epoch": 2.09,
+      "learning_rate": 4.437252528981586e-06,
+      "loss": 0.3982,
+      "step": 1000
+    },
+    {
+      "epoch": 2.09,
+      "learning_rate": 4.4184850363683726e-06,
+      "loss": 0.3833,
+      "step": 1001
+    },
+    {
+      "epoch": 2.1,
+      "learning_rate": 4.3997460553856095e-06,
+      "loss": 0.4035,
+      "step": 1002
+    },
+    {
+      "epoch": 2.1,
+      "learning_rate": 4.381035681756223e-06,
+      "loss": 0.3937,
+      "step": 1003
+    },
+    {
+      "epoch": 2.1,
+      "learning_rate": 4.3623540110569935e-06,
+      "loss": 0.3917,
+      "step": 1004
+    },
+    {
+      "epoch": 2.1,
+      "learning_rate": 4.343701138718086e-06,
+      "loss": 0.4103,
+      "step": 1005
+    },
+    {
+      "epoch": 2.1,
+      "learning_rate": 4.3250771600225536e-06,
+      "loss": 0.3882,
+      "step": 1006
+    },
+    {
+      "epoch": 2.11,
+      "learning_rate": 4.306482170105864e-06,
+      "loss": 0.3799,
+      "step": 1007
+    },
+    {
+      "epoch": 2.11,
+      "learning_rate": 4.2879162639553925e-06,
+      "loss": 0.3936,
+      "step": 1008
+    },
+    {
+      "epoch": 2.11,
+      "learning_rate": 4.26937953640995e-06,
+      "loss": 0.3822,
+      "step": 1009
+    },
+    {
+      "epoch": 2.11,
+      "learning_rate": 4.250872082159305e-06,
+      "loss": 0.3863,
+      "step": 1010
+    },
+    {
+      "epoch": 2.12,
+      "learning_rate": 4.232393995743682e-06,
+      "loss": 0.3842,
+      "step": 1011
+    },
+    {
+      "epoch": 2.12,
+      "learning_rate": 4.213945371553292e-06,
+      "loss": 0.3768,
+      "step": 1012
+    },
+    {
+      "epoch": 2.12,
+      "learning_rate": 4.195526303827842e-06,
+      "loss": 0.4035,
+      "step": 1013
+    },
+    {
+      "epoch": 2.12,
+      "learning_rate": 4.177136886656067e-06,
+      "loss": 0.3818,
+      "step": 1014
+    },
+    {
+      "epoch": 2.12,
+      "learning_rate": 4.158777213975231e-06,
+      "loss": 0.3972,
+      "step": 1015
+    },
+    {
+      "epoch": 2.13,
+      "learning_rate": 4.140447379570663e-06,
+      "loss": 0.3983,
+      "step": 1016
+    },
+    {
+      "epoch": 2.13,
+      "learning_rate": 4.12214747707527e-06,
+      "loss": 0.397,
+      "step": 1017
+    },
+    {
+      "epoch": 2.13,
+      "learning_rate": 4.103877599969056e-06,
+      "loss": 0.3859,
+      "step": 1018
+    },
+    {
+      "epoch": 2.13,
+      "learning_rate": 4.085637841578652e-06,
+      "loss": 0.4015,
+      "step": 1019
+    },
+    {
+      "epoch": 2.13,
+      "learning_rate": 4.067428295076833e-06,
+      "loss": 0.3806,
+      "step": 1020
+    },
+    {
+      "epoch": 2.14,
+      "learning_rate": 4.0492490534820525e-06,
+      "loss": 0.3847,
+      "step": 1021
+    },
+    {
+      "epoch": 2.14,
+      "learning_rate": 4.0311002096579486e-06,
+      "loss": 0.3835,
+      "step": 1022
+    },
+    {
+      "epoch": 2.14,
+      "learning_rate": 4.012981856312893e-06,
+      "loss": 0.4005,
+      "step": 1023
+    },
+    {
+      "epoch": 2.14,
+      "learning_rate": 3.9948940859994964e-06,
+      "loss": 0.3836,
+      "step": 1024
+    },
+    {
+      "epoch": 2.14,
+      "learning_rate": 3.976836991114146e-06,
+      "loss": 0.3956,
+      "step": 1025
+    },
+    {
+      "epoch": 2.15,
+      "learning_rate": 3.958810663896531e-06,
+      "loss": 0.3874,
+      "step": 1026
+    },
+    {
+      "epoch": 2.15,
+      "learning_rate": 3.9408151964291764e-06,
+      "loss": 0.3787,
+      "step": 1027
+    },
+    {
+      "epoch": 2.15,
+      "learning_rate": 3.922850680636968e-06,
+      "loss": 0.3895,
+      "step": 1028
+    },
+    {
+      "epoch": 2.15,
+      "learning_rate": 3.9049172082866786e-06,
+      "loss": 0.3959,
+      "step": 1029
+    },
+    {
+      "epoch": 2.15,
+      "learning_rate": 3.8870148709865115e-06,
+      "loss": 0.3917,
+      "step": 1030
+    },
+    {
+      "epoch": 2.16,
+      "learning_rate": 3.869143760185618e-06,
+      "loss": 0.3879,
+      "step": 1031
+    },
+    {
+      "epoch": 2.16,
+      "learning_rate": 3.851303967173647e-06,
+      "loss": 0.3776,
+      "step": 1032
+    },
+    {
+      "epoch": 2.16,
+      "learning_rate": 3.833495583080261e-06,
+      "loss": 0.3789,
+      "step": 1033
+    },
+    {
+      "epoch": 2.16,
+      "learning_rate": 3.815718698874672e-06,
+      "loss": 0.3818,
+      "step": 1034
+    },
+    {
+      "epoch": 2.17,
+      "learning_rate": 3.7979734053652028e-06,
+      "loss": 0.3818,
+      "step": 1035
+    },
+    {
+      "epoch": 2.17,
+      "learning_rate": 3.780259793198784e-06,
+      "loss": 0.387,
+      "step": 1036
+    },
+    {
+      "epoch": 2.17,
+      "learning_rate": 3.762577952860529e-06,
+      "loss": 0.4076,
+      "step": 1037
+    },
+    {
+      "epoch": 2.17,
+      "learning_rate": 3.744927974673237e-06,
+      "loss": 0.3906,
+      "step": 1038
+    },
+    {
+      "epoch": 2.17,
+      "learning_rate": 3.727309948796963e-06,
+      "loss": 0.3884,
+      "step": 1039
+    },
+    {
+      "epoch": 2.18,
+      "learning_rate": 3.709723965228531e-06,
+      "loss": 0.3964,
+      "step": 1040
+    },
+    {
+      "epoch": 2.18,
+      "learning_rate": 3.69217011380109e-06,
+      "loss": 0.3926,
+      "step": 1041
+    },
+    {
+      "epoch": 2.18,
+      "learning_rate": 3.6746484841836516e-06,
+      "loss": 0.3828,
+      "step": 1042
+    },
+    {
+      "epoch": 2.18,
+      "learning_rate": 3.657159165880626e-06,
+      "loss": 0.3923,
+      "step": 1043
+    },
+    {
+      "epoch": 2.18,
+      "learning_rate": 3.6397022482313804e-06,
+      "loss": 0.3869,
+      "step": 1044
+    },
+    {
+      "epoch": 2.19,
+      "learning_rate": 3.6222778204097607e-06,
+      "loss": 0.3958,
+      "step": 1045
+    },
+    {
+      "epoch": 2.19,
+      "learning_rate": 3.6048859714236597e-06,
+      "loss": 0.3935,
+      "step": 1046
+    },
+    {
+      "epoch": 2.19,
+      "learning_rate": 3.5875267901145383e-06,
+      "loss": 0.405,
+      "step": 1047
+    },
+    {
+      "epoch": 2.19,
+      "learning_rate": 3.5702003651569883e-06,
+      "loss": 0.4013,
+      "step": 1048
+    },
+    {
+      "epoch": 2.19,
+      "learning_rate": 3.552906785058278e-06,
+      "loss": 0.3931,
+      "step": 1049
+    },
+    {
+      "epoch": 2.2,
+      "learning_rate": 3.5356461381578865e-06,
+      "loss": 0.387,
+      "step": 1050
+    },
+    {
+      "epoch": 2.2,
+      "learning_rate": 3.51841851262708e-06,
+      "loss": 0.3862,
+      "step": 1051
+    },
+    {
+      "epoch": 2.2,
+      "learning_rate": 3.501223996468426e-06,
+      "loss": 0.3846,
+      "step": 1052
+    },
+    {
+      "epoch": 2.2,
+      "learning_rate": 3.484062677515376e-06,
+      "loss": 0.3927,
+      "step": 1053
+    },
+    {
+      "epoch": 2.21,
+      "learning_rate": 3.466934643431795e-06,
+      "loss": 0.38,
+      "step": 1054
+    },
+    {
+      "epoch": 2.21,
+      "learning_rate": 3.4498399817115213e-06,
+      "loss": 0.3874,
+      "step": 1055
+    },
+    {
+      "epoch": 2.21,
+      "learning_rate": 3.432778779677921e-06,
+      "loss": 0.3866,
+      "step": 1056
+    },
+    {
+      "epoch": 2.21,
+      "learning_rate": 3.415751124483445e-06,
+      "loss": 0.3962,
+      "step": 1057
+    },
+    {
+      "epoch": 2.21,
+      "learning_rate": 3.3987571031091735e-06,
+      "loss": 0.3934,
+      "step": 1058
+    },
+    {
+      "epoch": 2.22,
+      "learning_rate": 3.3817968023643766e-06,
+      "loss": 0.3887,
+      "step": 1059
+    },
+    {
+      "epoch": 2.22,
+      "learning_rate": 3.36487030888608e-06,
+      "loss": 0.3968,
+      "step": 1060
+    },
+    {
+      "epoch": 2.22,
+      "learning_rate": 3.3479777091386088e-06,
+      "loss": 0.3774,
+      "step": 1061
+    },
+    {
+      "epoch": 2.22,
+      "learning_rate": 3.3311190894131495e-06,
+      "loss": 0.3856,
+      "step": 1062
+    },
+    {
+      "epoch": 2.22,
+      "learning_rate": 3.3142945358273114e-06,
+      "loss": 0.3786,
+      "step": 1063
+    },
+    {
+      "epoch": 2.23,
+      "learning_rate": 3.2975041343246937e-06,
+      "loss": 0.3866,
+      "step": 1064
+    },
+    {
+      "epoch": 2.23,
+      "learning_rate": 3.280747970674427e-06,
+      "loss": 0.3814,
+      "step": 1065
+    },
+    {
+      "epoch": 2.23,
+      "learning_rate": 3.264026130470762e-06,
+      "loss": 0.3995,
+      "step": 1066
+    },
+    {
+      "epoch": 2.23,
+      "learning_rate": 3.247338699132604e-06,
+      "loss": 0.4061,
+      "step": 1067
+    },
+    {
+      "epoch": 2.23,
+      "learning_rate": 3.230685761903094e-06,
+      "loss": 0.387,
+      "step": 1068
+    },
+    {
+      "epoch": 2.24,
+      "learning_rate": 3.214067403849179e-06,
+      "loss": 0.3927,
+      "step": 1069
+    },
+    {
+      "epoch": 2.24,
+      "learning_rate": 3.1974837098611487e-06,
+      "loss": 0.394,
+      "step": 1070
+    },
+    {
+      "epoch": 2.24,
+      "learning_rate": 3.1809347646522414e-06,
+      "loss": 0.39,
+      "step": 1071
+    },
+    {
+      "epoch": 2.24,
+      "learning_rate": 3.1644206527581734e-06,
+      "loss": 0.3885,
+      "step": 1072
+    },
+    {
+      "epoch": 2.24,
+      "learning_rate": 3.1479414585367396e-06,
+      "loss": 0.3838,
+      "step": 1073
+    },
+    {
+      "epoch": 2.25,
+      "learning_rate": 3.1314972661673572e-06,
+      "loss": 0.3903,
+      "step": 1074
+    },
+    {
+      "epoch": 2.25,
+      "learning_rate": 3.115088159650648e-06,
+      "loss": 0.3879,
+      "step": 1075
+    },
+    {
+      "epoch": 2.25,
+      "learning_rate": 3.0987142228080137e-06,
+      "loss": 0.387,
+      "step": 1076
+    },
+    {
+      "epoch": 2.25,
+      "learning_rate": 3.082375539281194e-06,
+      "loss": 0.392,
+      "step": 1077
+    },
+    {
+      "epoch": 2.26,
+      "learning_rate": 3.0660721925318483e-06,
+      "loss": 0.3918,
+      "step": 1078
+    },
+    {
+      "epoch": 2.26,
+      "learning_rate": 3.0498042658411276e-06,
+      "loss": 0.4021,
+      "step": 1079
+    },
+    {
+      "epoch": 2.26,
+      "learning_rate": 3.0335718423092553e-06,
+      "loss": 0.3813,
+      "step": 1080
+    },
+    {
+      "epoch": 2.26,
+      "learning_rate": 3.0173750048550866e-06,
+      "loss": 0.3825,
+      "step": 1081
+    },
+    {
+      "epoch": 2.26,
+      "learning_rate": 3.0012138362157062e-06,
+      "loss": 0.3931,
+      "step": 1082
+    },
+    {
+      "epoch": 2.27,
+      "learning_rate": 2.985088418945986e-06,
+      "loss": 0.3923,
+      "step": 1083
+    },
+    {
+      "epoch": 2.27,
+      "learning_rate": 2.9689988354181742e-06,
+      "loss": 0.3943,
+      "step": 1084
+    },
+    {
+      "epoch": 2.27,
+      "learning_rate": 2.9529451678214726e-06,
+      "loss": 0.3995,
+      "step": 1085
+    },
+    {
+      "epoch": 2.27,
+      "learning_rate": 2.9369274981616137e-06,
+      "loss": 0.3945,
+      "step": 1086
+    },
+    {
+      "epoch": 2.27,
+      "learning_rate": 2.920945908260453e-06,
+      "loss": 0.3835,
+      "step": 1087
+    },
+    {
+      "epoch": 2.28,
+      "learning_rate": 2.905000479755531e-06,
+      "loss": 0.3894,
+      "step": 1088
+    },
+    {
+      "epoch": 2.28,
+      "learning_rate": 2.8890912940996784e-06,
+      "loss": 0.3932,
+      "step": 1089
+    },
+    {
+      "epoch": 2.28,
+      "learning_rate": 2.8732184325605815e-06,
+      "loss": 0.3989,
+      "step": 1090
+    },
+    {
+      "epoch": 2.28,
+      "learning_rate": 2.857381976220379e-06,
+      "loss": 0.381,
+      "step": 1091
+    },
+    {
+      "epoch": 2.28,
+      "learning_rate": 2.8415820059752397e-06,
+      "loss": 0.4052,
+      "step": 1092
+    },
+    {
+      "epoch": 2.29,
+      "learning_rate": 2.8258186025349543e-06,
+      "loss": 0.3817,
+      "step": 1093
+    },
+    {
+      "epoch": 2.29,
+      "learning_rate": 2.8100918464225304e-06,
+      "loss": 0.3899,
+      "step": 1094
+    },
+    {
+      "epoch": 2.29,
+      "learning_rate": 2.7944018179737576e-06,
+      "loss": 0.3908,
+      "step": 1095
+    },
+    {
+      "epoch": 2.29,
+      "learning_rate": 2.7787485973368288e-06,
+      "loss": 0.3808,
+      "step": 1096
+    },
+    {
+      "epoch": 2.29,
+      "learning_rate": 2.763132264471903e-06,
+      "loss": 0.3884,
+      "step": 1097
+    },
+    {
+      "epoch": 2.3,
+      "learning_rate": 2.7475528991507106e-06,
+      "loss": 0.3968,
+      "step": 1098
+    },
+    {
+      "epoch": 2.3,
+      "learning_rate": 2.7320105809561415e-06,
+      "loss": 0.389,
+      "step": 1099
+    },
+    {
+      "epoch": 2.3,
+      "learning_rate": 2.7165053892818495e-06,
+      "loss": 0.396,
+      "step": 1100
+    },
+    {
+      "epoch": 2.3,
+      "learning_rate": 2.701037403331829e-06,
+      "loss": 0.3884,
+      "step": 1101
+    },
+    {
+      "epoch": 2.31,
+      "learning_rate": 2.685606702120019e-06,
+      "loss": 0.3859,
+      "step": 1102
+    },
+    {
+      "epoch": 2.31,
+      "learning_rate": 2.670213364469908e-06,
+      "loss": 0.3831,
+      "step": 1103
+    },
+    {
+      "epoch": 2.31,
+      "learning_rate": 2.654857469014113e-06,
+      "loss": 0.4027,
+      "step": 1104
+    },
+    {
+      "epoch": 2.31,
+      "learning_rate": 2.6395390941940002e-06,
+      "loss": 0.3974,
+      "step": 1105
+    },
+    {
+      "epoch": 2.31,
+      "learning_rate": 2.624258318259253e-06,
+      "loss": 0.3867,
+      "step": 1106
+    },
+    {
+      "epoch": 2.32,
+      "learning_rate": 2.609015219267511e-06,
+      "loss": 0.3941,
+      "step": 1107
+    },
+    {
+      "epoch": 2.32,
+      "learning_rate": 2.5938098750839414e-06,
+      "loss": 0.3841,
+      "step": 1108
+    },
+    {
+      "epoch": 2.32,
+      "learning_rate": 2.5786423633808487e-06,
+      "loss": 0.3905,
+      "step": 1109
+    },
+    {
+      "epoch": 2.32,
+      "learning_rate": 2.563512761637291e-06,
+      "loss": 0.3888,
+      "step": 1110
+    },
+    {
+      "epoch": 2.32,
+      "learning_rate": 2.5484211471386633e-06,
+      "loss": 0.3949,
+      "step": 1111
+    },
+    {
+      "epoch": 2.33,
+      "learning_rate": 2.5333675969763215e-06,
+      "loss": 0.385,
+      "step": 1112
+    },
+    {
+      "epoch": 2.33,
+      "learning_rate": 2.518352188047176e-06,
+      "loss": 0.3869,
+      "step": 1113
+    },
+    {
+      "epoch": 2.33,
+      "learning_rate": 2.5033749970533015e-06,
+      "loss": 0.3751,
+      "step": 1114
+    },
+    {
+      "epoch": 2.33,
+      "learning_rate": 2.488436100501549e-06,
+      "loss": 0.3846,
+      "step": 1115
+    },
+    {
+      "epoch": 2.33,
+      "learning_rate": 2.4735355747031566e-06,
+      "loss": 0.3997,
+      "step": 1116
+    },
+    {
+      "epoch": 2.34,
+      "learning_rate": 2.4586734957733495e-06,
+      "loss": 0.3842,
+      "step": 1117
+    },
+    {
+      "epoch": 2.34,
+      "learning_rate": 2.443849939630959e-06,
+      "loss": 0.3898,
+      "step": 1118
+    },
+    {
+      "epoch": 2.34,
+      "learning_rate": 2.4290649819980404e-06,
+      "loss": 0.411,
+      "step": 1119
+    },
+    {
+      "epoch": 2.34,
+      "learning_rate": 2.4143186983994715e-06,
+      "loss": 0.3942,
+      "step": 1120
+    },
+    {
+      "epoch": 2.35,
+      "learning_rate": 2.3996111641625762e-06,
+      "loss": 0.3958,
+      "step": 1121
+    },
+    {
+      "epoch": 2.35,
+      "learning_rate": 2.384942454416734e-06,
+      "loss": 0.381,
+      "step": 1122
+    },
+    {
+      "epoch": 2.35,
+      "learning_rate": 2.3703126440930137e-06,
+      "loss": 0.3989,
+      "step": 1123
+    },
+    {
+      "epoch": 2.35,
+      "learning_rate": 2.3557218079237608e-06,
+      "loss": 0.4029,
+      "step": 1124
+    },
+    {
+      "epoch": 2.35,
+      "learning_rate": 2.3411700204422437e-06,
+      "loss": 0.3879,
+      "step": 1125
+    },
+    {
+      "epoch": 2.36,
+      "learning_rate": 2.3266573559822568e-06,
+      "loss": 0.3772,
+      "step": 1126
+    },
+    {
+      "epoch": 2.36,
+      "learning_rate": 2.3121838886777437e-06,
+      "loss": 0.3859,
+      "step": 1127
+    },
+    {
+      "epoch": 2.36,
+      "learning_rate": 2.2977496924624223e-06,
+      "loss": 0.3922,
+      "step": 1128
+    },
+    {
+      "epoch": 2.36,
+      "learning_rate": 2.283354841069403e-06,
+      "loss": 0.3871,
+      "step": 1129
+    },
+    {
+      "epoch": 2.36,
+      "learning_rate": 2.26899940803082e-06,
+      "loss": 0.3807,
+      "step": 1130
+    },
+    {
+      "epoch": 2.37,
+      "learning_rate": 2.2546834666774397e-06,
+      "loss": 0.3798,
+      "step": 1131
+    },
+    {
+      "epoch": 2.37,
+      "learning_rate": 2.240407090138309e-06,
+      "loss": 0.3974,
+      "step": 1132
+    },
+    {
+      "epoch": 2.37,
+      "learning_rate": 2.226170351340358e-06,
+      "loss": 0.3925,
+      "step": 1133
+    },
+    {
+      "epoch": 2.37,
+      "learning_rate": 2.211973323008041e-06,
+      "loss": 0.389,
+      "step": 1134
+    },
+    {
+      "epoch": 2.37,
+      "learning_rate": 2.197816077662965e-06,
+      "loss": 0.3926,
+      "step": 1135
+    },
+    {
+      "epoch": 2.38,
+      "learning_rate": 2.183698687623511e-06,
+      "loss": 0.3863,
+      "step": 1136
+    },
+    {
+      "epoch": 2.38,
+      "learning_rate": 2.16962122500448e-06,
+      "loss": 0.3973,
+      "step": 1137
+    },
+    {
+      "epoch": 2.38,
+      "learning_rate": 2.155583761716703e-06,
+      "loss": 0.3789,
+      "step": 1138
+    },
+    {
+      "epoch": 2.38,
+      "learning_rate": 2.1415863694666973e-06,
+      "loss": 0.3854,
+      "step": 1139
+    },
+    {
+      "epoch": 2.38,
+      "learning_rate": 2.1276291197562772e-06,
+      "loss": 0.3803,
+      "step": 1140
+    },
+    {
+      "epoch": 2.39,
+      "learning_rate": 2.1137120838822124e-06,
+      "loss": 0.3896,
+      "step": 1141
+    },
+    {
+      "epoch": 2.39,
+      "learning_rate": 2.0998353329358355e-06,
+      "loss": 0.384,
+      "step": 1142
+    },
+    {
+      "epoch": 2.39,
+      "learning_rate": 2.0859989378027033e-06,
+      "loss": 0.3905,
+      "step": 1143
+    },
+    {
+      "epoch": 2.39,
+      "learning_rate": 2.072202969162234e-06,
+      "loss": 0.3945,
+      "step": 1144
+    },
+    {
+      "epoch": 2.4,
+      "learning_rate": 2.0584474974873204e-06,
+      "loss": 0.3864,
+      "step": 1145
+    },
+    {
+      "epoch": 2.4,
+      "learning_rate": 2.0447325930440043e-06,
+      "loss": 0.3813,
+      "step": 1146
+    },
+    {
+      "epoch": 2.4,
+      "learning_rate": 2.03105832589109e-06,
+      "loss": 0.389,
+      "step": 1147
+    },
+    {
+      "epoch": 2.4,
+      "learning_rate": 2.0174247658798054e-06,
+      "loss": 0.3832,
+      "step": 1148
+    },
+    {
+      "epoch": 2.4,
+      "learning_rate": 2.0038319826534312e-06,
+      "loss": 0.391,
+      "step": 1149
+    },
+    {
+      "epoch": 2.41,
+      "learning_rate": 1.990280045646954e-06,
+      "loss": 0.3766,
+      "step": 1150
+    },
+    {
+      "epoch": 2.41,
+      "learning_rate": 1.9767690240867054e-06,
+      "loss": 0.3752,
+      "step": 1151
+    },
+    {
+      "epoch": 2.41,
+      "learning_rate": 1.9632989869900145e-06,
+      "loss": 0.3939,
+      "step": 1152
+    },
+    {
+      "epoch": 2.41,
+      "learning_rate": 1.9498700031648554e-06,
+      "loss": 0.3786,
+      "step": 1153
+    },
+    {
+      "epoch": 2.41,
+      "learning_rate": 1.936482141209486e-06,
+      "loss": 0.4039,
+      "step": 1154
+    },
+    {
+      "epoch": 2.42,
+      "learning_rate": 1.923135469512113e-06,
+      "loss": 0.3944,
+      "step": 1155
+    },
+    {
+      "epoch": 2.42,
+      "learning_rate": 1.9098300562505266e-06,
+      "loss": 0.3835,
+      "step": 1156
+    },
+    {
+      "epoch": 2.42,
+      "learning_rate": 1.8965659693917626e-06,
+      "loss": 0.3903,
+      "step": 1157
+    },
+    {
+      "epoch": 2.42,
+      "learning_rate": 1.8833432766917514e-06,
+      "loss": 0.3786,
+      "step": 1158
+    },
+    {
+      "epoch": 2.42,
+      "learning_rate": 1.870162045694971e-06,
+      "loss": 0.3823,
+      "step": 1159
+    },
+    {
+      "epoch": 2.43,
+      "learning_rate": 1.8570223437341118e-06,
+      "loss": 0.3877,
+      "step": 1160
+    },
+    {
+      "epoch": 2.43,
+      "learning_rate": 1.8439242379297129e-06,
+      "loss": 0.3819,
+      "step": 1161
+    },
+    {
+      "epoch": 2.43,
+      "learning_rate": 1.8308677951898435e-06,
+      "loss": 0.3875,
+      "step": 1162
+    },
+    {
+      "epoch": 2.43,
+      "learning_rate": 1.8178530822097373e-06,
+      "loss": 0.391,
+      "step": 1163
+    },
+    {
+      "epoch": 2.44,
+      "learning_rate": 1.8048801654714687e-06,
+      "loss": 0.3783,
+      "step": 1164
+    },
+    {
+      "epoch": 2.44,
+      "learning_rate": 1.7919491112436038e-06,
+      "loss": 0.3867,
+      "step": 1165
+    },
+    {
+      "epoch": 2.44,
+      "learning_rate": 1.7790599855808732e-06,
+      "loss": 0.3859,
+      "step": 1166
+    },
+    {
+      "epoch": 2.44,
+      "learning_rate": 1.76621285432382e-06,
+      "loss": 0.381,
+      "step": 1167
+    },
+    {
+      "epoch": 2.44,
+      "learning_rate": 1.7534077830984697e-06,
+      "loss": 0.3796,
+      "step": 1168
+    },
+    {
+      "epoch": 2.45,
+      "learning_rate": 1.7406448373160024e-06,
+      "loss": 0.3771,
+      "step": 1169
+    },
+    {
+      "epoch": 2.45,
+      "learning_rate": 1.7279240821724063e-06,
+      "loss": 0.396,
+      "step": 1170
+    },
+    {
+      "epoch": 2.45,
+      "learning_rate": 1.7152455826481529e-06,
+      "loss": 0.3835,
+      "step": 1171
+    },
+    {
+      "epoch": 2.45,
+      "learning_rate": 1.7026094035078589e-06,
+      "loss": 0.3838,
+      "step": 1172
+    },
+    {
+      "epoch": 2.45,
+      "learning_rate": 1.6900156092999664e-06,
+      "loss": 0.3821,
+      "step": 1173
+    },
+    {
+      "epoch": 2.46,
+      "learning_rate": 1.6774642643563955e-06,
+      "loss": 0.3895,
+      "step": 1174
+    },
+    {
+      "epoch": 2.46,
+      "learning_rate": 1.664955432792239e-06,
+      "loss": 0.388,
+      "step": 1175
+    },
+    {
+      "epoch": 2.46,
+      "learning_rate": 1.6524891785054097e-06,
+      "loss": 0.3709,
+      "step": 1176
+    },
+    {
+      "epoch": 2.46,
+      "learning_rate": 1.6400655651763308e-06,
+      "loss": 0.3872,
+      "step": 1177
+    },
+    {
+      "epoch": 2.46,
+      "learning_rate": 1.6276846562676085e-06,
+      "loss": 0.3905,
+      "step": 1178
+    },
+    {
+      "epoch": 2.47,
+      "learning_rate": 1.615346515023698e-06,
+      "loss": 0.3909,
+      "step": 1179
+    },
+    {
+      "epoch": 2.47,
+      "learning_rate": 1.603051204470597e-06,
+      "loss": 0.3898,
+      "step": 1180
+    },
+    {
+      "epoch": 2.47,
+      "learning_rate": 1.5907987874155074e-06,
+      "loss": 0.3949,
+      "step": 1181
+    },
+    {
+      "epoch": 2.47,
+      "learning_rate": 1.5785893264465257e-06,
+      "loss": 0.3791,
+      "step": 1182
+    },
+    {
+      "epoch": 2.47,
+      "learning_rate": 1.566422883932317e-06,
+      "loss": 0.3978,
+      "step": 1183
+    },
+    {
+      "epoch": 2.48,
+      "learning_rate": 1.5542995220217961e-06,
+      "loss": 0.3908,
+      "step": 1184
+    },
+    {
+      "epoch": 2.48,
+      "learning_rate": 1.5422193026438225e-06,
+      "loss": 0.3837,
+      "step": 1185
+    },
+    {
+      "epoch": 2.48,
+      "learning_rate": 1.530182287506855e-06,
+      "loss": 0.3858,
+      "step": 1186
+    },
+    {
+      "epoch": 2.48,
+      "learning_rate": 1.518188538098677e-06,
+      "loss": 0.3923,
+      "step": 1187
+    },
+    {
+      "epoch": 2.49,
+      "learning_rate": 1.506238115686044e-06,
+      "loss": 0.379,
+      "step": 1188
+    },
+    {
+      "epoch": 2.49,
+      "learning_rate": 1.4943310813144006e-06,
+      "loss": 0.3946,
+      "step": 1189
+    },
+    {
+      "epoch": 2.49,
+      "learning_rate": 1.4824674958075436e-06,
+      "loss": 0.394,
+      "step": 1190
+    },
+    {
+      "epoch": 2.49,
+      "learning_rate": 1.4706474197673336e-06,
+      "loss": 0.3897,
+      "step": 1191
+    },
+    {
+      "epoch": 2.49,
+      "learning_rate": 1.458870913573368e-06,
+      "loss": 0.3851,
+      "step": 1192
+    },
+    {
+      "epoch": 2.5,
+      "learning_rate": 1.4471380373826826e-06,
+      "loss": 0.3807,
+      "step": 1193
+    },
+    {
+      "epoch": 2.5,
+      "learning_rate": 1.4354488511294418e-06,
+      "loss": 0.3864,
+      "step": 1194
+    },
+    {
+      "epoch": 2.5,
+      "learning_rate": 1.4238034145246271e-06,
+      "loss": 0.3789,
+      "step": 1195
+    },
+    {
+      "epoch": 2.5,
+      "learning_rate": 1.4122017870557458e-06,
+      "loss": 0.3962,
+      "step": 1196
+    },
+    {
+      "epoch": 2.5,
+      "learning_rate": 1.4006440279865085e-06,
+      "loss": 0.3751,
+      "step": 1197
+    },
+    {
+      "epoch": 2.51,
+      "learning_rate": 1.3891301963565473e-06,
+      "loss": 0.3829,
+      "step": 1198
+    },
+    {
+      "epoch": 2.51,
+      "learning_rate": 1.3776603509810938e-06,
+      "loss": 0.3854,
+      "step": 1199
+    },
+    {
+      "epoch": 2.51,
+      "learning_rate": 1.3662345504506903e-06,
+      "loss": 0.3855,
+      "step": 1200
+    },
+    {
+      "epoch": 2.51,
+      "learning_rate": 1.3548528531308892e-06,
+      "loss": 0.3804,
+      "step": 1201
+    },
+    {
+      "epoch": 2.51,
+      "learning_rate": 1.343515317161952e-06,
+      "loss": 0.3896,
+      "step": 1202
+    },
+    {
+      "epoch": 2.52,
+      "learning_rate": 1.3322220004585596e-06,
+      "loss": 0.3883,
+      "step": 1203
+    },
+    {
+      "epoch": 2.52,
+      "learning_rate": 1.3209729607095022e-06,
+      "loss": 0.3896,
+      "step": 1204
+    },
+    {
+      "epoch": 2.52,
+      "learning_rate": 1.3097682553774028e-06,
+      "loss": 0.3842,
+      "step": 1205
+    },
+    {
+      "epoch": 2.52,
+      "learning_rate": 1.2986079416984088e-06,
+      "loss": 0.3847,
+      "step": 1206
+    },
+    {
+      "epoch": 2.53,
+      "learning_rate": 1.2874920766819032e-06,
+      "loss": 0.3868,
+      "step": 1207
+    },
+    {
+      "epoch": 2.53,
+      "learning_rate": 1.2764207171102206e-06,
+      "loss": 0.4027,
+      "step": 1208
+    },
+    {
+      "epoch": 2.53,
+      "learning_rate": 1.2653939195383448e-06,
+      "loss": 0.3961,
+      "step": 1209
+    },
+    {
+      "epoch": 2.53,
+      "learning_rate": 1.2544117402936373e-06,
+      "loss": 0.3868,
+      "step": 1210
+    },
+    {
+      "epoch": 2.53,
+      "learning_rate": 1.243474235475528e-06,
+      "loss": 0.3853,
+      "step": 1211
+    },
+    {
+      "epoch": 2.54,
+      "learning_rate": 1.232581460955249e-06,
+      "loss": 0.3911,
+      "step": 1212
+    },
+    {
+      "epoch": 2.54,
+      "learning_rate": 1.2217334723755348e-06,
+      "loss": 0.3851,
+      "step": 1213
+    },
+    {
+      "epoch": 2.54,
+      "learning_rate": 1.2109303251503434e-06,
+      "loss": 0.3901,
+      "step": 1214
+    },
+    {
+      "epoch": 2.54,
+      "learning_rate": 1.2001720744645729e-06,
+      "loss": 0.3777,
+      "step": 1215
+    },
+    {
+      "epoch": 2.54,
+      "learning_rate": 1.189458775273784e-06,
+      "loss": 0.3872,
+      "step": 1216
+    },
+    {
+      "epoch": 2.55,
+      "learning_rate": 1.178790482303912e-06,
+      "loss": 0.3812,
+      "step": 1217
+    },
+    {
+      "epoch": 2.55,
+      "learning_rate": 1.1681672500509866e-06,
+      "loss": 0.3847,
+      "step": 1218
+    },
+    {
+      "epoch": 2.55,
+      "learning_rate": 1.1575891327808664e-06,
+      "loss": 0.3878,
+      "step": 1219
+    },
+    {
+      "epoch": 2.55,
+      "learning_rate": 1.147056184528943e-06,
+      "loss": 0.3816,
+      "step": 1220
+    },
+    {
+      "epoch": 2.55,
+      "learning_rate": 1.1365684590998837e-06,
+      "loss": 0.3927,
+      "step": 1221
+    },
+    {
+      "epoch": 2.56,
+      "learning_rate": 1.1261260100673355e-06,
+      "loss": 0.39,
+      "step": 1222
+    },
+    {
+      "epoch": 2.56,
+      "learning_rate": 1.1157288907736762e-06,
+      "loss": 0.381,
+      "step": 1223
+    },
+    {
+      "epoch": 2.56,
+      "learning_rate": 1.1053771543297198e-06,
+      "loss": 0.3771,
+      "step": 1224
+    },
+    {
+      "epoch": 2.56,
+      "learning_rate": 1.0950708536144616e-06,
+      "loss": 0.3794,
+      "step": 1225
+    },
+    {
+      "epoch": 2.56,
+      "learning_rate": 1.0848100412747954e-06,
+      "loss": 0.3893,
+      "step": 1226
+    },
+    {
+      "epoch": 2.57,
+      "learning_rate": 1.0745947697252512e-06,
+      "loss": 0.3721,
+      "step": 1227
+    },
+    {
+      "epoch": 2.57,
+      "learning_rate": 1.0644250911477306e-06,
+      "loss": 0.3848,
+      "step": 1228
+    },
+    {
+      "epoch": 2.57,
+      "learning_rate": 1.0543010574912305e-06,
+      "loss": 0.3889,
+      "step": 1229
+    },
+    {
+      "epoch": 2.57,
+      "learning_rate": 1.0442227204715872e-06,
+      "loss": 0.3755,
+      "step": 1230
+    },
+    {
+      "epoch": 2.58,
+      "learning_rate": 1.0341901315712044e-06,
+      "loss": 0.3751,
+      "step": 1231
+    },
+    {
+      "epoch": 2.58,
+      "learning_rate": 1.0242033420388008e-06,
+      "loss": 0.3756,
+      "step": 1232
+    },
+    {
+      "epoch": 2.58,
+      "learning_rate": 1.0142624028891334e-06,
+      "loss": 0.3848,
+      "step": 1233
+    },
+    {
+      "epoch": 2.58,
+      "learning_rate": 1.0043673649027519e-06,
+      "loss": 0.3736,
+      "step": 1234
+    },
+    {
+      "epoch": 2.58,
+      "learning_rate": 9.945182786257301e-07,
+      "loss": 0.3825,
+      "step": 1235
+    },
+    {
+      "epoch": 2.59,
+      "learning_rate": 9.8471519436941e-07,
+      "loss": 0.3878,
+      "step": 1236
+    },
+    {
+      "epoch": 2.59,
+      "learning_rate": 9.749581622101424e-07,
+      "loss": 0.3964,
+      "step": 1237
+    },
+    {
+      "epoch": 2.59,
+      "learning_rate": 9.652472319890372e-07,
+      "loss": 0.3759,
+      "step": 1238
+    },
+    {
+      "epoch": 2.59,
+      "learning_rate": 9.555824533117064e-07,
+      "loss": 0.3872,
+      "step": 1239
+    },
+    {
+      "epoch": 2.59,
+      "learning_rate": 9.459638755480038e-07,
+      "loss": 0.3946,
+      "step": 1240
+    },
+    {
+      "epoch": 2.6,
+      "learning_rate": 9.363915478317875e-07,
+      "loss": 0.3905,
+      "step": 1241
+    },
+    {
+      "epoch": 2.6,
+      "learning_rate": 9.268655190606501e-07,
+      "loss": 0.3941,
+      "step": 1242
+    },
+    {
+      "epoch": 2.6,
+      "learning_rate": 9.173858378956856e-07,
+      "loss": 0.392,
+      "step": 1243
+    },
+    {
+      "epoch": 2.6,
+      "learning_rate": 9.079525527612321e-07,
+      "loss": 0.371,
+      "step": 1244
+    },
+    {
+      "epoch": 2.6,
+      "learning_rate": 8.985657118446223e-07,
+      "loss": 0.3806,
+      "step": 1245
+    },
+    {
+      "epoch": 2.61,
+      "learning_rate": 8.892253630959502e-07,
+      "loss": 0.3856,
+      "step": 1246
+    },
+    {
+      "epoch": 2.61,
+      "learning_rate": 8.799315542278075e-07,
+      "loss": 0.3807,
+      "step": 1247
+    },
+    {
+      "epoch": 2.61,
+      "learning_rate": 8.706843327150605e-07,
+      "loss": 0.3832,
+      "step": 1248
+    },
+    {
+      "epoch": 2.61,
+      "learning_rate": 8.614837457945868e-07,
+      "loss": 0.3736,
+      "step": 1249
+    },
+    {
+      "epoch": 2.62,
+      "learning_rate": 8.523298404650504e-07,
+      "loss": 0.3914,
+      "step": 1250
+    },
+    {
+      "epoch": 2.62,
+      "learning_rate": 8.432226634866514e-07,
+      "loss": 0.4003,
+      "step": 1251
+    },
+    {
+      "epoch": 2.62,
+      "learning_rate": 8.34162261380892e-07,
+      "loss": 0.3841,
+      "step": 1252
+    },
+    {
+      "epoch": 2.62,
+      "learning_rate": 8.25148680430341e-07,
+      "loss": 0.3784,
+      "step": 1253
+    },
+    {
+      "epoch": 2.62,
+      "learning_rate": 8.161819666783888e-07,
+      "loss": 0.3942,
+      "step": 1254
+    },
+    {
+      "epoch": 2.63,
+      "learning_rate": 8.072621659290247e-07,
+      "loss": 0.3953,
+      "step": 1255
+    },
+    {
+      "epoch": 2.63,
+      "learning_rate": 7.983893237465878e-07,
+      "loss": 0.3904,
+      "step": 1256
+    },
+    {
+      "epoch": 2.63,
+      "learning_rate": 7.895634854555512e-07,
+      "loss": 0.3973,
+      "step": 1257
+    },
+    {
+      "epoch": 2.63,
+      "learning_rate": 7.807846961402699e-07,
+      "loss": 0.391,
+      "step": 1258
+    },
+    {
+      "epoch": 2.63,
+      "learning_rate": 7.720530006447735e-07,
+      "loss": 0.3753,
+      "step": 1259
+    },
+    {
+      "epoch": 2.64,
+      "learning_rate": 7.633684435725208e-07,
+      "loss": 0.3918,
+      "step": 1260
+    },
+    {
+      "epoch": 2.64,
+      "learning_rate": 7.547310692861754e-07,
+      "loss": 0.3836,
+      "step": 1261
+    },
+    {
+      "epoch": 2.64,
+      "learning_rate": 7.461409219073857e-07,
+      "loss": 0.3931,
+      "step": 1262
+    },
+    {
+      "epoch": 2.64,
+      "learning_rate": 7.375980453165487e-07,
+      "loss": 0.3756,
+      "step": 1263
+    },
+    {
+      "epoch": 2.64,
+      "learning_rate": 7.291024831525961e-07,
+      "loss": 0.3804,
+      "step": 1264
+    },
+    {
+      "epoch": 2.65,
+      "learning_rate": 7.206542788127646e-07,
+      "loss": 0.3757,
+      "step": 1265
+    },
+    {
+      "epoch": 2.65,
+      "learning_rate": 7.122534754523768e-07,
+      "loss": 0.3949,
+      "step": 1266
+    },
+    {
+      "epoch": 2.65,
+      "learning_rate": 7.039001159846215e-07,
+      "loss": 0.3915,
+      "step": 1267
+    },
+    {
+      "epoch": 2.65,
+      "learning_rate": 6.955942430803298e-07,
+      "loss": 0.3786,
+      "step": 1268
+    },
+    {
+      "epoch": 2.65,
+      "learning_rate": 6.87335899167767e-07,
+      "loss": 0.3835,
+      "step": 1269
+    },
+    {
+      "epoch": 2.66,
+      "learning_rate": 6.791251264324039e-07,
+      "loss": 0.3821,
+      "step": 1270
+    },
+    {
+      "epoch": 2.66,
+      "learning_rate": 6.709619668167122e-07,
+      "loss": 0.3888,
+      "step": 1271
+    },
+    {
+      "epoch": 2.66,
+      "learning_rate": 6.628464620199404e-07,
+      "loss": 0.3976,
+      "step": 1272
+    },
+    {
+      "epoch": 2.66,
+      "learning_rate": 6.547786534979083e-07,
+      "loss": 0.3857,
+      "step": 1273
+    },
+    {
+      "epoch": 2.67,
+      "learning_rate": 6.467585824627886e-07,
+      "loss": 0.3807,
+      "step": 1274
+    },
+    {
+      "epoch": 2.67,
+      "learning_rate": 6.387862898829078e-07,
+      "loss": 0.3993,
+      "step": 1275
+    },
+    {
+      "epoch": 2.67,
+      "learning_rate": 6.30861816482522e-07,
+      "loss": 0.3987,
+      "step": 1276
+    },
+    {
+      "epoch": 2.67,
+      "learning_rate": 6.229852027416183e-07,
+      "loss": 0.3762,
+      "step": 1277
+    },
+    {
+      "epoch": 2.67,
+      "learning_rate": 6.151564888957084e-07,
+      "loss": 0.3914,
+      "step": 1278
+    },
+    {
+      "epoch": 2.68,
+      "learning_rate": 6.073757149356185e-07,
+      "loss": 0.3817,
+      "step": 1279
+    },
+    {
+      "epoch": 2.68,
+      "learning_rate": 5.996429206072874e-07,
+      "loss": 0.3883,
+      "step": 1280
+    },
+    {
+      "epoch": 2.68,
+      "learning_rate": 5.919581454115598e-07,
+      "loss": 0.3928,
+      "step": 1281
+    },
+    {
+      "epoch": 2.68,
+      "learning_rate": 5.843214286039956e-07,
+      "loss": 0.3864,
+      "step": 1282
+    },
+    {
+      "epoch": 2.68,
+      "learning_rate": 5.767328091946544e-07,
+      "loss": 0.3862,
+      "step": 1283
+    },
+    {
+      "epoch": 2.69,
+      "learning_rate": 5.691923259479093e-07,
+      "loss": 0.3798,
+      "step": 1284
+    },
+    {
+      "epoch": 2.69,
+      "learning_rate": 5.617000173822385e-07,
+      "loss": 0.3819,
+      "step": 1285
+    },
+    {
+      "epoch": 2.69,
+      "learning_rate": 5.542559217700339e-07,
+      "loss": 0.3896,
+      "step": 1286
+    },
+    {
+      "epoch": 2.69,
+      "learning_rate": 5.468600771374077e-07,
+      "loss": 0.3801,
+      "step": 1287
+    },
+    {
+      "epoch": 2.69,
+      "learning_rate": 5.395125212639895e-07,
+      "loss": 0.3809,
+      "step": 1288
+    },
+    {
+      "epoch": 2.7,
+      "learning_rate": 5.322132916827483e-07,
+      "loss": 0.4017,
+      "step": 1289
+    },
+    {
+      "epoch": 2.7,
+      "learning_rate": 5.249624256797803e-07,
+      "loss": 0.3745,
+      "step": 1290
+    },
+    {
+      "epoch": 2.7,
+      "learning_rate": 5.177599602941407e-07,
+      "loss": 0.3815,
+      "step": 1291
+    },
+    {
+      "epoch": 2.7,
+      "learning_rate": 5.106059323176371e-07,
+      "loss": 0.3785,
+      "step": 1292
+    },
+    {
+      "epoch": 2.71,
+      "learning_rate": 5.035003782946468e-07,
+      "loss": 0.378,
+      "step": 1293
+    },
+    {
+      "epoch": 2.71,
+      "learning_rate": 4.964433345219354e-07,
+      "loss": 0.3809,
+      "step": 1294
+    },
+    {
+      "epoch": 2.71,
+      "learning_rate": 4.894348370484648e-07,
+      "loss": 0.3927,
+      "step": 1295
+    },
+    {
+      "epoch": 2.71,
+      "learning_rate": 4.824749216752134e-07,
+      "loss": 0.3773,
+      "step": 1296
+    },
+    {
+      "epoch": 2.71,
+      "learning_rate": 4.755636239549877e-07,
+      "loss": 0.3792,
+      "step": 1297
+    },
+    {
+      "epoch": 2.72,
+      "learning_rate": 4.6870097919224923e-07,
+      "loss": 0.3912,
+      "step": 1298
+    },
+    {
+      "epoch": 2.72,
+      "learning_rate": 4.6188702244292614e-07,
+      "loss": 0.3895,
+      "step": 1299
+    },
+    {
+      "epoch": 2.72,
+      "learning_rate": 4.551217885142378e-07,
+      "loss": 0.3889,
+      "step": 1300
+    },
+    {
+      "epoch": 2.72,
+      "learning_rate": 4.484053119645182e-07,
+      "loss": 0.3904,
+      "step": 1301
+    },
+    {
+      "epoch": 2.72,
+      "learning_rate": 4.417376271030327e-07,
+      "loss": 0.3805,
+      "step": 1302
+    },
+    {
+      "epoch": 2.73,
+      "learning_rate": 4.3511876798981387e-07,
+      "loss": 0.3748,
+      "step": 1303
+    },
+    {
+      "epoch": 2.73,
+      "learning_rate": 4.285487684354772e-07,
+      "loss": 0.3694,
+      "step": 1304
+    },
+    {
+      "epoch": 2.73,
+      "learning_rate": 4.220276620010566e-07,
+      "loss": 0.3993,
+      "step": 1305
+    },
+    {
+      "epoch": 2.73,
+      "learning_rate": 4.1555548199782357e-07,
+      "loss": 0.3964,
+      "step": 1306
+    },
+    {
+      "epoch": 2.73,
+      "learning_rate": 4.091322614871274e-07,
+      "loss": 0.3748,
+      "step": 1307
+    },
+    {
+      "epoch": 2.74,
+      "learning_rate": 4.0275803328021946e-07,
+      "loss": 0.3945,
+      "step": 1308
+    },
+    {
+      "epoch": 2.74,
+      "learning_rate": 3.96432829938086e-07,
+      "loss": 0.3868,
+      "step": 1309
+    },
+    {
+      "epoch": 2.74,
+      "learning_rate": 3.9015668377128446e-07,
+      "loss": 0.3864,
+      "step": 1310
+    },
+    {
+      "epoch": 2.74,
+      "learning_rate": 3.839296268397774e-07,
+      "loss": 0.3633,
+      "step": 1311
+    },
+    {
+      "epoch": 2.74,
+      "learning_rate": 3.777516909527701e-07,
+      "loss": 0.3856,
+      "step": 1312
+    },
+    {
+      "epoch": 2.75,
+      "learning_rate": 3.7162290766854293e-07,
+      "loss": 0.3771,
+      "step": 1313
+    },
+    {
+      "epoch": 2.75,
+      "learning_rate": 3.6554330829429716e-07,
+      "loss": 0.3725,
+      "step": 1314
+    },
+    {
+      "epoch": 2.75,
+      "learning_rate": 3.5951292388598956e-07,
+      "loss": 0.3817,
+      "step": 1315
+    },
+    {
+      "epoch": 2.75,
+      "learning_rate": 3.5353178524817566e-07,
+      "loss": 0.3983,
+      "step": 1316
+    },
+    {
+      "epoch": 2.76,
+      "learning_rate": 3.4759992293385446e-07,
+      "loss": 0.3792,
+      "step": 1317
+    },
+    {
+      "epoch": 2.76,
+      "learning_rate": 3.417173672443075e-07,
+      "loss": 0.393,
+      "step": 1318
+    },
+    {
+      "epoch": 2.76,
+      "learning_rate": 3.3588414822895097e-07,
+      "loss": 0.3771,
+      "step": 1319
+    },
+    {
+      "epoch": 2.76,
+      "learning_rate": 3.30100295685174e-07,
+      "loss": 0.3874,
+      "step": 1320
+    },
+    {
+      "epoch": 2.76,
+      "learning_rate": 3.24365839158195e-07,
+      "loss": 0.3851,
+      "step": 1321
+    },
+    {
+      "epoch": 2.77,
+      "learning_rate": 3.1868080794090316e-07,
+      "loss": 0.3781,
+      "step": 1322
+    },
+    {
+      "epoch": 2.77,
+      "learning_rate": 3.130452310737131e-07,
+      "loss": 0.395,
+      "step": 1323
+    },
+    {
+      "epoch": 2.77,
+      "learning_rate": 3.0745913734441357e-07,
+      "loss": 0.3821,
+      "step": 1324
+    },
+    {
+      "epoch": 2.77,
+      "learning_rate": 3.0192255528802894e-07,
+      "loss": 0.3876,
+      "step": 1325
+    },
+    {
+      "epoch": 2.77,
+      "learning_rate": 2.9643551318665917e-07,
+      "loss": 0.378,
+      "step": 1326
+    },
+    {
+      "epoch": 2.78,
+      "learning_rate": 2.9099803906934567e-07,
+      "loss": 0.3943,
+      "step": 1327
+    },
+    {
+      "epoch": 2.78,
+      "learning_rate": 2.8561016071192884e-07,
+      "loss": 0.3829,
+      "step": 1328
+    },
+    {
+      "epoch": 2.78,
+      "learning_rate": 2.8027190563689745e-07,
+      "loss": 0.3866,
+      "step": 1329
+    },
+    {
+      "epoch": 2.78,
+      "learning_rate": 2.7498330111325635e-07,
+      "loss": 0.3822,
+      "step": 1330
+    },
+    {
+      "epoch": 2.78,
+      "learning_rate": 2.697443741563832e-07,
+      "loss": 0.3809,
+      "step": 1331
+    },
+    {
+      "epoch": 2.79,
+      "learning_rate": 2.6455515152789435e-07,
+      "loss": 0.3868,
+      "step": 1332
+    },
+    {
+      "epoch": 2.79,
+      "learning_rate": 2.594156597355002e-07,
+      "loss": 0.3749,
+      "step": 1333
+    },
+    {
+      "epoch": 2.79,
+      "learning_rate": 2.5432592503288e-07,
+      "loss": 0.3734,
+      "step": 1334
+    },
+    {
+      "epoch": 2.79,
+      "learning_rate": 2.492859734195385e-07,
+      "loss": 0.3891,
+      "step": 1335
+    },
+    {
+      "epoch": 2.79,
+      "learning_rate": 2.442958306406795e-07,
+      "loss": 0.378,
+      "step": 1336
+    },
+    {
+      "epoch": 2.8,
+      "learning_rate": 2.393555221870747e-07,
+      "loss": 0.3792,
+      "step": 1337
+    },
+    {
+      "epoch": 2.8,
+      "learning_rate": 2.3446507329492274e-07,
+      "loss": 0.3718,
+      "step": 1338
+    },
+    {
+      "epoch": 2.8,
+      "learning_rate": 2.2962450894573606e-07,
+      "loss": 0.3825,
+      "step": 1339
+    },
+    {
+      "epoch": 2.8,
+      "learning_rate": 2.2483385386620317e-07,
+      "loss": 0.3774,
+      "step": 1340
+    },
+    {
+      "epoch": 2.81,
+      "learning_rate": 2.200931325280664e-07,
+      "loss": 0.381,
+      "step": 1341
+    },
+    {
+      "epoch": 2.81,
+      "learning_rate": 2.1540236914799116e-07,
+      "loss": 0.3821,
+      "step": 1342
+    },
+    {
+      "epoch": 2.81,
+      "learning_rate": 2.1076158768744914e-07,
+      "loss": 0.3842,
+      "step": 1343
+    },
+    {
+      "epoch": 2.81,
+      "learning_rate": 2.0617081185259512e-07,
+      "loss": 0.3848,
+      "step": 1344
+    },
+    {
+      "epoch": 2.81,
+      "learning_rate": 2.0163006509414052e-07,
+      "loss": 0.3862,
+      "step": 1345
+    },
+    {
+      "epoch": 2.82,
+      "learning_rate": 1.9713937060723887e-07,
+      "loss": 0.3804,
+      "step": 1346
+    },
+    {
+      "epoch": 2.82,
+      "learning_rate": 1.9269875133136384e-07,
+      "loss": 0.3843,
+      "step": 1347
+    },
+    {
+      "epoch": 2.82,
+      "learning_rate": 1.8830822995019482e-07,
+      "loss": 0.3857,
+      "step": 1348
+    },
+    {
+      "epoch": 2.82,
+      "learning_rate": 1.8396782889150144e-07,
+      "loss": 0.3823,
+      "step": 1349
+    },
+    {
+      "epoch": 2.82,
+      "learning_rate": 1.7967757032702481e-07,
+      "loss": 0.3758,
+      "step": 1350
+    },
+    {
+      "epoch": 2.83,
+      "learning_rate": 1.7543747617236873e-07,
+      "loss": 0.4084,
+      "step": 1351
+    },
+    {
+      "epoch": 2.83,
+      "learning_rate": 1.7124756808688525e-07,
+      "loss": 0.372,
+      "step": 1352
+    },
+    {
+      "epoch": 2.83,
+      "learning_rate": 1.6710786747356377e-07,
+      "loss": 0.3709,
+      "step": 1353
+    },
+    {
+      "epoch": 2.83,
+      "learning_rate": 1.630183954789233e-07,
+      "loss": 0.3896,
+      "step": 1354
+    },
+    {
+      "epoch": 2.83,
+      "learning_rate": 1.5897917299290578e-07,
+      "loss": 0.3794,
+      "step": 1355
+    },
+    {
+      "epoch": 2.84,
+      "learning_rate": 1.5499022064876412e-07,
+      "loss": 0.3814,
+      "step": 1356
+    },
+    {
+      "epoch": 2.84,
+      "learning_rate": 1.510515588229633e-07,
+      "loss": 0.3698,
+      "step": 1357
+    },
+    {
+      "epoch": 2.84,
+      "learning_rate": 1.4716320763507152e-07,
+      "loss": 0.3803,
+      "step": 1358
+    },
+    {
+      "epoch": 2.84,
+      "learning_rate": 1.4332518694765708e-07,
+      "loss": 0.3802,
+      "step": 1359
+    },
+    {
+      "epoch": 2.85,
+      "learning_rate": 1.3953751636619162e-07,
+      "loss": 0.3933,
+      "step": 1360
+    },
+    {
+      "epoch": 2.85,
+      "learning_rate": 1.3580021523894482e-07,
+      "loss": 0.3816,
+      "step": 1361
+    },
+    {
+      "epoch": 2.85,
+      "learning_rate": 1.3211330265689104e-07,
+      "loss": 0.3876,
+      "step": 1362
+    },
+    {
+      "epoch": 2.85,
+      "learning_rate": 1.2847679745360607e-07,
+      "loss": 0.3847,
+      "step": 1363
+    },
+    {
+      "epoch": 2.85,
+      "learning_rate": 1.2489071820517394e-07,
+      "loss": 0.3842,
+      "step": 1364
+    },
+    {
+      "epoch": 2.86,
+      "learning_rate": 1.2135508323009248e-07,
+      "loss": 0.3863,
+      "step": 1365
+    },
+    {
+      "epoch": 2.86,
+      "learning_rate": 1.1786991058917785e-07,
+      "loss": 0.3907,
+      "step": 1366
+    },
+    {
+      "epoch": 2.86,
+      "learning_rate": 1.1443521808547353e-07,
+      "loss": 0.3883,
+      "step": 1367
+    },
+    {
+      "epoch": 2.86,
+      "learning_rate": 1.1105102326415929e-07,
+      "loss": 0.3795,
+      "step": 1368
+    },
+    {
+      "epoch": 2.86,
+      "learning_rate": 1.0771734341246121e-07,
+      "loss": 0.3969,
+      "step": 1369
+    },
+    {
+      "epoch": 2.87,
+      "learning_rate": 1.0443419555956402e-07,
+      "loss": 0.3842,
+      "step": 1370
+    },
+    {
+      "epoch": 2.87,
+      "learning_rate": 1.012015964765245e-07,
+      "loss": 0.3854,
+      "step": 1371
+    },
+    {
+      "epoch": 2.87,
+      "learning_rate": 9.801956267618262e-08,
+      "loss": 0.3854,
+      "step": 1372
+    },
+    {
+      "epoch": 2.87,
+      "learning_rate": 9.488811041308166e-08,
+      "loss": 0.3956,
+      "step": 1373
+    },
+    {
+      "epoch": 2.87,
+      "learning_rate": 9.180725568338045e-08,
+      "loss": 0.3827,
+      "step": 1374
+    },
+    {
+      "epoch": 2.88,
+      "learning_rate": 8.87770142247768e-08,
+      "loss": 0.3835,
+      "step": 1375
+    },
+    {
+      "epoch": 2.88,
+      "learning_rate": 8.579740151642534e-08,
+      "loss": 0.3876,
+      "step": 1376
+    },
+    {
+      "epoch": 2.88,
+      "learning_rate": 8.286843277885426e-08,
+      "loss": 0.385,
+      "step": 1377
+    },
+    {
+      "epoch": 2.88,
+      "learning_rate": 7.999012297389419e-08,
+      "loss": 0.3735,
+      "step": 1378
+    },
+    {
+      "epoch": 2.88,
+      "learning_rate": 7.716248680459726e-08,
+      "loss": 0.3863,
+      "step": 1379
+    },
+    {
+      "epoch": 2.89,
+      "learning_rate": 7.438553871516152e-08,
+      "loss": 0.3819,
+      "step": 1380
+    },
+    {
+      "epoch": 2.89,
+      "learning_rate": 7.165929289086327e-08,
+      "loss": 0.3817,
+      "step": 1381
+    },
+    {
+      "epoch": 2.89,
+      "learning_rate": 6.898376325797596e-08,
+      "loss": 0.3771,
+      "step": 1382
+    },
+    {
+      "epoch": 2.89,
+      "learning_rate": 6.635896348370475e-08,
+      "loss": 0.369,
+      "step": 1383
+    },
+    {
+      "epoch": 2.9,
+      "learning_rate": 6.378490697611761e-08,
+      "loss": 0.3744,
+      "step": 1384
+    },
+    {
+      "epoch": 2.9,
+      "learning_rate": 6.126160688407101e-08,
+      "loss": 0.3884,
+      "step": 1385
+    },
+    {
+      "epoch": 2.9,
+      "learning_rate": 5.878907609714879e-08,
+      "loss": 0.382,
+      "step": 1386
+    },
+    {
+      "epoch": 2.9,
+      "learning_rate": 5.636732724559113e-08,
+      "loss": 0.3774,
+      "step": 1387
+    },
+    {
+      "epoch": 2.9,
+      "learning_rate": 5.399637270023683e-08,
+      "loss": 0.3801,
+      "step": 1388
+    },
+    {
+      "epoch": 2.91,
+      "learning_rate": 5.1676224572452246e-08,
+      "loss": 0.3739,
+      "step": 1389
+    },
+    {
+      "epoch": 2.91,
+      "learning_rate": 4.940689471407356e-08,
+      "loss": 0.3792,
+      "step": 1390
+    },
+    {
+      "epoch": 2.91,
+      "learning_rate": 4.718839471734904e-08,
+      "loss": 0.3887,
+      "step": 1391
+    },
+    {
+      "epoch": 2.91,
+      "learning_rate": 4.502073591487244e-08,
+      "loss": 0.3758,
+      "step": 1392
+    },
+    {
+      "epoch": 2.91,
+      "learning_rate": 4.2903929379530806e-08,
+      "loss": 0.4003,
+      "step": 1393
+    },
+    {
+      "epoch": 2.92,
+      "learning_rate": 4.083798592444899e-08,
+      "loss": 0.39,
+      "step": 1394
+    },
+    {
+      "epoch": 2.92,
+      "learning_rate": 3.882291610292854e-08,
+      "loss": 0.3847,
+      "step": 1395
+    },
+    {
+      "epoch": 2.92,
+      "learning_rate": 3.68587302083967e-08,
+      "loss": 0.3776,
+      "step": 1396
+    },
+    {
+      "epoch": 2.92,
+      "learning_rate": 3.494543827435748e-08,
+      "loss": 0.3872,
+      "step": 1397
+    },
+    {
+      "epoch": 2.92,
+      "learning_rate": 3.308305007433399e-08,
+      "loss": 0.3882,
+      "step": 1398
+    },
+    {
+      "epoch": 2.93,
+      "learning_rate": 3.127157512182288e-08,
+      "loss": 0.3938,
+      "step": 1399
+    },
+    {
+      "epoch": 2.93,
+      "learning_rate": 2.9511022670246635e-08,
+      "loss": 0.378,
+      "step": 1400
+    },
+    {
+      "epoch": 2.93,
+      "learning_rate": 2.7801401712900245e-08,
+      "loss": 0.3823,
+      "step": 1401
+    },
+    {
+      "epoch": 2.93,
+      "learning_rate": 2.6142720982911264e-08,
+      "loss": 0.3843,
+      "step": 1402
+    },
+    {
+      "epoch": 2.94,
+      "learning_rate": 2.4534988953192063e-08,
+      "loss": 0.3757,
+      "step": 1403
+    },
+    {
+      "epoch": 2.94,
+      "learning_rate": 2.2978213836400974e-08,
+      "loss": 0.3907,
+      "step": 1404
+    },
+    {
+      "epoch": 2.94,
+      "learning_rate": 2.1472403584892332e-08,
+      "loss": 0.377,
+      "step": 1405
+    },
+    {
+      "epoch": 2.94,
+      "learning_rate": 2.0017565890683154e-08,
+      "loss": 0.378,
+      "step": 1406
+    },
+    {
+      "epoch": 2.94,
+      "learning_rate": 1.8613708185410973e-08,
+      "loss": 0.3755,
+      "step": 1407
+    },
+    {
+      "epoch": 2.95,
+      "learning_rate": 1.726083764029607e-08,
+      "loss": 0.3972,
+      "step": 1408
+    },
+    {
+      "epoch": 2.95,
+      "learning_rate": 1.5958961166104847e-08,
+      "loss": 0.3744,
+      "step": 1409
+    },
+    {
+      "epoch": 2.95,
+      "learning_rate": 1.4708085413113194e-08,
+      "loss": 0.3741,
+      "step": 1410
+    },
+    {
+      "epoch": 2.95,
+      "learning_rate": 1.3508216771076499e-08,
+      "loss": 0.3799,
+      "step": 1411
+    },
+    {
+      "epoch": 2.95,
+      "learning_rate": 1.2359361369190804e-08,
+      "loss": 0.3713,
+      "step": 1412
+    },
+    {
+      "epoch": 2.96,
+      "learning_rate": 1.1261525076069479e-08,
+      "loss": 0.3854,
+      "step": 1413
+    },
+    {
+      "epoch": 2.96,
+      "learning_rate": 1.0214713499706596e-08,
+      "loss": 0.3796,
+      "step": 1414
+    },
+    {
+      "epoch": 2.96,
+      "learning_rate": 9.218931987448055e-09,
+      "loss": 0.3834,
+      "step": 1415
+    },
+    {
+      "epoch": 2.96,
+      "learning_rate": 8.274185625971598e-09,
+      "loss": 0.3774,
+      "step": 1416
+    },
+    {
+      "epoch": 2.96,
+      "learning_rate": 7.380479241253513e-09,
+      "loss": 0.3923,
+      "step": 1417
+    },
+    {
+      "epoch": 2.97,
+      "learning_rate": 6.5378173985441994e-09,
+      "loss": 0.3728,
+      "step": 1418
+    },
+    {
+      "epoch": 2.97,
+      "learning_rate": 5.7462044023515186e-09,
+      "loss": 0.3724,
+      "step": 1419
+    },
+    {
+      "epoch": 2.97,
+      "learning_rate": 5.0056442964119265e-09,
+      "loss": 0.3917,
+      "step": 1420
+    },
+    {
+      "epoch": 2.97,
+      "learning_rate": 4.316140863671603e-09,
+      "loss": 0.3836,
+      "step": 1421
+    },
+    {
+      "epoch": 2.97,
+      "learning_rate": 3.6776976262697937e-09,
+      "loss": 0.3877,
+      "step": 1422
+    },
+    {
+      "epoch": 2.98,
+      "learning_rate": 3.0903178455166106e-09,
+      "loss": 0.3872,
+      "step": 1423
+    },
+    {
+      "epoch": 2.98,
+      "learning_rate": 2.5540045218819256e-09,
+      "loss": 0.3776,
+      "step": 1424
+    },
+    {
+      "epoch": 2.98,
+      "learning_rate": 2.0687603949720583e-09,
+      "loss": 0.3921,
+      "step": 1425
+    },
+    {
+      "epoch": 2.98,
+      "learning_rate": 1.6345879435231138e-09,
+      "loss": 0.3845,
+      "step": 1426
+    },
+    {
+      "epoch": 2.99,
+      "learning_rate": 1.2514893853865506e-09,
+      "loss": 0.3829,
+      "step": 1427
+    },
+    {
+      "epoch": 2.99,
+      "learning_rate": 9.194666775158567e-10,
+      "loss": 0.3857,
+      "step": 1428
+    },
+    {
+      "epoch": 2.99,
+      "learning_rate": 6.385215159565583e-10,
+      "loss": 0.3928,
+      "step": 1429
+    },
+    {
+      "epoch": 2.99,
+      "learning_rate": 4.086553358395584e-10,
+      "loss": 0.3863,
+      "step": 1430
+    },
+    {
+      "epoch": 2.99,
+      "learning_rate": 2.2986931137447544e-10,
+      "loss": 0.3843,
+      "step": 1431
+    },
+    {
+      "epoch": 3.0,
+      "learning_rate": 1.0216435583743079e-10,
+      "loss": 0.3764,
+      "step": 1432
+    },
+    {
+      "epoch": 3.0,
+      "learning_rate": 2.5541121576599937e-11,
+      "loss": 0.3817,
+      "step": 1433
+    },
+    {
+      "epoch": 3.0,
+      "learning_rate": 0.0,
+      "loss": 0.3441,
+      "step": 1434
+    },
+    {
+      "epoch": 3.0,
+      "step": 1434,
+      "total_flos": 3191501654065152.0,
+      "train_loss": 0.5603317790450412,
+      "train_runtime": 20210.0223,
+      "train_samples_per_second": 9.073,
+      "train_steps_per_second": 0.071
+    }
+  ],
+  "max_steps": 1434,
+  "num_train_epochs": 3,
+  "total_flos": 3191501654065152.0,
+  "trial_name": null,
+  "trial_params": null
+}