diff --git "a/trainer_state.json" "b/trainer_state.json"
new file mode 100644--- /dev/null
+++ "b/trainer_state.json"
@@ -0,0 +1,6816 @@
+{
+  "best_metric": null,
+  "best_model_checkpoint": null,
+  "epoch": 3.0,
+  "eval_steps": 500,
+  "global_step": 1131,
+  "is_hyper_param_search": false,
+  "is_local_process_zero": true,
+  "is_world_process_zero": true,
+  "log_history": [
+    {
+      "epoch": 0.0,
+      "learning_rate": 5.882352941176471e-07,
+      "loss": 0.9127,
+      "step": 1
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 1.1764705882352942e-06,
+      "loss": 0.9942,
+      "step": 2
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 1.7647058823529414e-06,
+      "loss": 0.9241,
+      "step": 3
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 2.3529411764705885e-06,
+      "loss": 0.9443,
+      "step": 4
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 2.9411764705882355e-06,
+      "loss": 0.8932,
+      "step": 5
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 3.529411764705883e-06,
+      "loss": 0.9872,
+      "step": 6
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.11764705882353e-06,
+      "loss": 0.9327,
+      "step": 7
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 4.705882352941177e-06,
+      "loss": 0.9247,
+      "step": 8
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 5.294117647058824e-06,
+      "loss": 0.9039,
+      "step": 9
+    },
+    {
+      "epoch": 0.03,
+      "learning_rate": 5.882352941176471e-06,
+      "loss": 0.8923,
+      "step": 10
+    },
+    {
+      "epoch": 0.03,
+      "learning_rate": 6.470588235294119e-06,
+      "loss": 0.9064,
+      "step": 11
+    },
+    {
+      "epoch": 0.03,
+      "learning_rate": 7.058823529411766e-06,
+      "loss": 0.9073,
+      "step": 12
+    },
+    {
+      "epoch": 0.03,
+      "learning_rate": 7.647058823529411e-06,
+      "loss": 0.8631,
+      "step": 13
+    },
+    {
+      "epoch": 0.04,
+      "learning_rate": 8.23529411764706e-06,
+      "loss": 0.8912,
+      "step": 14
+    },
+    {
+      "epoch": 0.04,
+      "learning_rate": 8.823529411764707e-06,
+      "loss": 0.8646,
+      "step": 15
+    },
+    {
+      "epoch": 0.04,
+      "learning_rate": 9.411764705882354e-06,
+      "loss": 0.8017,
+      "step": 16
+    },
+    {
+      "epoch": 0.05,
+      "learning_rate": 1e-05,
+      "loss": 0.8463,
+      "step": 17
+    },
+    {
+      "epoch": 0.05,
+      "learning_rate": 1.0588235294117648e-05,
+      "loss": 0.7901,
+      "step": 18
+    },
+    {
+      "epoch": 0.05,
+      "learning_rate": 1.1176470588235295e-05,
+      "loss": 0.7786,
+      "step": 19
+    },
+    {
+      "epoch": 0.05,
+      "learning_rate": 1.1764705882352942e-05,
+      "loss": 0.7813,
+      "step": 20
+    },
+    {
+      "epoch": 0.06,
+      "learning_rate": 1.235294117647059e-05,
+      "loss": 0.7729,
+      "step": 21
+    },
+    {
+      "epoch": 0.06,
+      "learning_rate": 1.2941176470588238e-05,
+      "loss": 0.7978,
+      "step": 22
+    },
+    {
+      "epoch": 0.06,
+      "learning_rate": 1.3529411764705885e-05,
+      "loss": 0.7806,
+      "step": 23
+    },
+    {
+      "epoch": 0.06,
+      "learning_rate": 1.4117647058823532e-05,
+      "loss": 0.7714,
+      "step": 24
+    },
+    {
+      "epoch": 0.07,
+      "learning_rate": 1.4705882352941179e-05,
+      "loss": 0.7831,
+      "step": 25
+    },
+    {
+      "epoch": 0.07,
+      "learning_rate": 1.5294117647058822e-05,
+      "loss": 0.767,
+      "step": 26
+    },
+    {
+      "epoch": 0.07,
+      "learning_rate": 1.5882352941176473e-05,
+      "loss": 0.7876,
+      "step": 27
+    },
+    {
+      "epoch": 0.07,
+      "learning_rate": 1.647058823529412e-05,
+      "loss": 0.7497,
+      "step": 28
+    },
+    {
+      "epoch": 0.08,
+      "learning_rate": 1.7058823529411767e-05,
+      "loss": 0.7382,
+      "step": 29
+    },
+    {
+      "epoch": 0.08,
+      "learning_rate": 1.7647058823529414e-05,
+      "loss": 0.7975,
+      "step": 30
+    },
+    {
+      "epoch": 0.08,
+      "learning_rate": 1.823529411764706e-05,
+      "loss": 0.7451,
+      "step": 31
+    },
+    {
+      "epoch": 0.08,
+      "learning_rate": 1.8823529411764708e-05,
+      "loss": 0.7531,
+      "step": 32
+    },
+    {
+      "epoch": 0.09,
+      "learning_rate": 1.9411764705882355e-05,
+      "loss": 0.7473,
+      "step": 33
+    },
+    {
+      "epoch": 0.09,
+      "learning_rate": 2e-05,
+      "loss": 0.7735,
+      "step": 34
+    },
+    {
+      "epoch": 0.09,
+      "learning_rate": 1.9999958993170007e-05,
+      "loss": 0.7137,
+      "step": 35
+    },
+    {
+      "epoch": 0.1,
+      "learning_rate": 1.9999835973016332e-05,
+      "loss": 0.7505,
+      "step": 36
+    },
+    {
+      "epoch": 0.1,
+      "learning_rate": 1.9999630940547908e-05,
+      "loss": 0.7466,
+      "step": 37
+    },
+    {
+      "epoch": 0.1,
+      "learning_rate": 1.9999343897446283e-05,
+      "loss": 0.7257,
+      "step": 38
+    },
+    {
+      "epoch": 0.1,
+      "learning_rate": 1.9998974846065607e-05,
+      "loss": 0.7447,
+      "step": 39
+    },
+    {
+      "epoch": 0.11,
+      "learning_rate": 1.9998523789432597e-05,
+      "loss": 0.7943,
+      "step": 40
+    },
+    {
+      "epoch": 0.11,
+      "learning_rate": 1.9997990731246545e-05,
+      "loss": 0.7445,
+      "step": 41
+    },
+    {
+      "epoch": 0.11,
+      "learning_rate": 1.999737567587924e-05,
+      "loss": 0.7174,
+      "step": 42
+    },
+    {
+      "epoch": 0.11,
+      "learning_rate": 1.9996678628374994e-05,
+      "loss": 0.7652,
+      "step": 43
+    },
+    {
+      "epoch": 0.12,
+      "learning_rate": 1.9995899594450538e-05,
+      "loss": 0.7495,
+      "step": 44
+    },
+    {
+      "epoch": 0.12,
+      "learning_rate": 1.9995038580495014e-05,
+      "loss": 0.721,
+      "step": 45
+    },
+    {
+      "epoch": 0.12,
+      "learning_rate": 1.9994095593569917e-05,
+      "loss": 0.7738,
+      "step": 46
+    },
+    {
+      "epoch": 0.12,
+      "learning_rate": 1.9993070641409026e-05,
+      "loss": 0.7717,
+      "step": 47
+    },
+    {
+      "epoch": 0.13,
+      "learning_rate": 1.999196373241835e-05,
+      "loss": 0.7431,
+      "step": 48
+    },
+    {
+      "epoch": 0.13,
+      "learning_rate": 1.9990774875676055e-05,
+      "loss": 0.742,
+      "step": 49
+    },
+    {
+      "epoch": 0.13,
+      "learning_rate": 1.9989504080932382e-05,
+      "loss": 0.7281,
+      "step": 50
+    },
+    {
+      "epoch": 0.14,
+      "learning_rate": 1.9988151358609598e-05,
+      "loss": 0.7273,
+      "step": 51
+    },
+    {
+      "epoch": 0.14,
+      "learning_rate": 1.9986716719801863e-05,
+      "loss": 0.7508,
+      "step": 52
+    },
+    {
+      "epoch": 0.14,
+      "learning_rate": 1.998520017627518e-05,
+      "loss": 0.6883,
+      "step": 53
+    },
+    {
+      "epoch": 0.14,
+      "learning_rate": 1.9983601740467277e-05,
+      "loss": 0.7484,
+      "step": 54
+    },
+    {
+      "epoch": 0.15,
+      "learning_rate": 1.9981921425487506e-05,
+      "loss": 0.7047,
+      "step": 55
+    },
+    {
+      "epoch": 0.15,
+      "learning_rate": 1.9980159245116756e-05,
+      "loss": 0.7326,
+      "step": 56
+    },
+    {
+      "epoch": 0.15,
+      "learning_rate": 1.99783152138073e-05,
+      "loss": 0.6988,
+      "step": 57
+    },
+    {
+      "epoch": 0.15,
+      "learning_rate": 1.9976389346682722e-05,
+      "loss": 0.7231,
+      "step": 58
+    },
+    {
+      "epoch": 0.16,
+      "learning_rate": 1.9974381659537764e-05,
+      "loss": 0.7159,
+      "step": 59
+    },
+    {
+      "epoch": 0.16,
+      "learning_rate": 1.99722921688382e-05,
+      "loss": 0.7071,
+      "step": 60
+    },
+    {
+      "epoch": 0.16,
+      "learning_rate": 1.9970120891720707e-05,
+      "loss": 0.7108,
+      "step": 61
+    },
+    {
+      "epoch": 0.16,
+      "learning_rate": 1.9967867845992722e-05,
+      "loss": 0.6794,
+      "step": 62
+    },
+    {
+      "epoch": 0.17,
+      "learning_rate": 1.9965533050132303e-05,
+      "loss": 0.7418,
+      "step": 63
+    },
+    {
+      "epoch": 0.17,
+      "learning_rate": 1.9963116523287965e-05,
+      "loss": 0.7714,
+      "step": 64
+    },
+    {
+      "epoch": 0.17,
+      "learning_rate": 1.9960618285278524e-05,
+      "loss": 0.7378,
+      "step": 65
+    },
+    {
+      "epoch": 0.18,
+      "learning_rate": 1.9958038356592944e-05,
+      "loss": 0.7545,
+      "step": 66
+    },
+    {
+      "epoch": 0.18,
+      "learning_rate": 1.9955376758390173e-05,
+      "loss": 0.7821,
+      "step": 67
+    },
+    {
+      "epoch": 0.18,
+      "learning_rate": 1.9952633512498945e-05,
+      "loss": 0.7502,
+      "step": 68
+    },
+    {
+      "epoch": 0.18,
+      "learning_rate": 1.9949808641417622e-05,
+      "loss": 0.7033,
+      "step": 69
+    },
+    {
+      "epoch": 0.19,
+      "learning_rate": 1.9946902168314014e-05,
+      "loss": 0.7443,
+      "step": 70
+    },
+    {
+      "epoch": 0.19,
+      "learning_rate": 1.9943914117025163e-05,
+      "loss": 0.7467,
+      "step": 71
+    },
+    {
+      "epoch": 0.19,
+      "learning_rate": 1.9940844512057174e-05,
+      "loss": 0.7393,
+      "step": 72
+    },
+    {
+      "epoch": 0.19,
+      "learning_rate": 1.9937693378585e-05,
+      "loss": 0.7566,
+      "step": 73
+    },
+    {
+      "epoch": 0.2,
+      "learning_rate": 1.993446074245224e-05,
+      "loss": 0.7237,
+      "step": 74
+    },
+    {
+      "epoch": 0.2,
+      "learning_rate": 1.9931146630170926e-05,
+      "loss": 0.6914,
+      "step": 75
+    },
+    {
+      "epoch": 0.2,
+      "learning_rate": 1.9927751068921308e-05,
+      "loss": 0.6906,
+      "step": 76
+    },
+    {
+      "epoch": 0.2,
+      "learning_rate": 1.9924274086551623e-05,
+      "loss": 0.7284,
+      "step": 77
+    },
+    {
+      "epoch": 0.21,
+      "learning_rate": 1.992071571157788e-05,
+      "loss": 0.7526,
+      "step": 78
+    },
+    {
+      "epoch": 0.21,
+      "learning_rate": 1.9917075973183608e-05,
+      "loss": 0.7322,
+      "step": 79
+    },
+    {
+      "epoch": 0.21,
+      "learning_rate": 1.9913354901219643e-05,
+      "loss": 0.7484,
+      "step": 80
+    },
+    {
+      "epoch": 0.21,
+      "learning_rate": 1.990955252620385e-05,
+      "loss": 0.78,
+      "step": 81
+    },
+    {
+      "epoch": 0.22,
+      "learning_rate": 1.9905668879320902e-05,
+      "loss": 0.7542,
+      "step": 82
+    },
+    {
+      "epoch": 0.22,
+      "learning_rate": 1.9901703992422007e-05,
+      "loss": 0.6869,
+      "step": 83
+    },
+    {
+      "epoch": 0.22,
+      "learning_rate": 1.9897657898024656e-05,
+      "loss": 0.6993,
+      "step": 84
+    },
+    {
+      "epoch": 0.23,
+      "learning_rate": 1.989353062931235e-05,
+      "loss": 0.7562,
+      "step": 85
+    },
+    {
+      "epoch": 0.23,
+      "learning_rate": 1.9889322220134325e-05,
+      "loss": 0.7031,
+      "step": 86
+    },
+    {
+      "epoch": 0.23,
+      "learning_rate": 1.988503270500529e-05,
+      "loss": 0.7221,
+      "step": 87
+    },
+    {
+      "epoch": 0.23,
+      "learning_rate": 1.9880662119105128e-05,
+      "loss": 0.7337,
+      "step": 88
+    },
+    {
+      "epoch": 0.24,
+      "learning_rate": 1.9876210498278612e-05,
+      "loss": 0.7506,
+      "step": 89
+    },
+    {
+      "epoch": 0.24,
+      "learning_rate": 1.9871677879035114e-05,
+      "loss": 0.7115,
+      "step": 90
+    },
+    {
+      "epoch": 0.24,
+      "learning_rate": 1.9867064298548305e-05,
+      "loss": 0.7215,
+      "step": 91
+    },
+    {
+      "epoch": 0.24,
+      "learning_rate": 1.9862369794655845e-05,
+      "loss": 0.769,
+      "step": 92
+    },
+    {
+      "epoch": 0.25,
+      "learning_rate": 1.985759440585908e-05,
+      "loss": 0.7495,
+      "step": 93
+    },
+    {
+      "epoch": 0.25,
+      "learning_rate": 1.985273817132272e-05,
+      "loss": 0.7467,
+      "step": 94
+    },
+    {
+      "epoch": 0.25,
+      "learning_rate": 1.9847801130874525e-05,
+      "loss": 0.7317,
+      "step": 95
+    },
+    {
+      "epoch": 0.25,
+      "learning_rate": 1.9842783325004966e-05,
+      "loss": 0.7458,
+      "step": 96
+    },
+    {
+      "epoch": 0.26,
+      "learning_rate": 1.983768479486691e-05,
+      "loss": 0.7239,
+      "step": 97
+    },
+    {
+      "epoch": 0.26,
+      "learning_rate": 1.9832505582275263e-05,
+      "loss": 0.7116,
+      "step": 98
+    },
+    {
+      "epoch": 0.26,
+      "learning_rate": 1.982724572970665e-05,
+      "loss": 0.7062,
+      "step": 99
+    },
+    {
+      "epoch": 0.27,
+      "learning_rate": 1.982190528029904e-05,
+      "loss": 0.6864,
+      "step": 100
+    },
+    {
+      "epoch": 0.27,
+      "learning_rate": 1.981648427785142e-05,
+      "loss": 0.7703,
+      "step": 101
+    },
+    {
+      "epoch": 0.27,
+      "learning_rate": 1.981098276682341e-05,
+      "loss": 0.6996,
+      "step": 102
+    },
+    {
+      "epoch": 0.27,
+      "learning_rate": 1.980540079233492e-05,
+      "loss": 0.7316,
+      "step": 103
+    },
+    {
+      "epoch": 0.28,
+      "learning_rate": 1.979973840016576e-05,
+      "loss": 0.7342,
+      "step": 104
+    },
+    {
+      "epoch": 0.28,
+      "learning_rate": 1.9793995636755287e-05,
+      "loss": 0.7499,
+      "step": 105
+    },
+    {
+      "epoch": 0.28,
+      "learning_rate": 1.9788172549202002e-05,
+      "loss": 0.6878,
+      "step": 106
+    },
+    {
+      "epoch": 0.28,
+      "learning_rate": 1.9782269185263175e-05,
+      "loss": 0.732,
+      "step": 107
+    },
+    {
+      "epoch": 0.29,
+      "learning_rate": 1.977628559335446e-05,
+      "loss": 0.696,
+      "step": 108
+    },
+    {
+      "epoch": 0.29,
+      "learning_rate": 1.9770221822549477e-05,
+      "loss": 0.7371,
+      "step": 109
+    },
+    {
+      "epoch": 0.29,
+      "learning_rate": 1.976407792257944e-05,
+      "loss": 0.6985,
+      "step": 110
+    },
+    {
+      "epoch": 0.29,
+      "learning_rate": 1.9757853943832707e-05,
+      "loss": 0.7627,
+      "step": 111
+    },
+    {
+      "epoch": 0.3,
+      "learning_rate": 1.9751549937354418e-05,
+      "loss": 0.722,
+      "step": 112
+    },
+    {
+      "epoch": 0.3,
+      "learning_rate": 1.9745165954846035e-05,
+      "loss": 0.7404,
+      "step": 113
+    },
+    {
+      "epoch": 0.3,
+      "learning_rate": 1.9738702048664936e-05,
+      "loss": 0.7318,
+      "step": 114
+    },
+    {
+      "epoch": 0.31,
+      "learning_rate": 1.973215827182397e-05,
+      "loss": 0.7172,
+      "step": 115
+    },
+    {
+      "epoch": 0.31,
+      "learning_rate": 1.9725534677991062e-05,
+      "loss": 0.7415,
+      "step": 116
+    },
+    {
+      "epoch": 0.31,
+      "learning_rate": 1.971883132148872e-05,
+      "loss": 0.7313,
+      "step": 117
+    },
+    {
+      "epoch": 0.31,
+      "learning_rate": 1.9712048257293625e-05,
+      "loss": 0.7244,
+      "step": 118
+    },
+    {
+      "epoch": 0.32,
+      "learning_rate": 1.970518554103617e-05,
+      "loss": 0.7372,
+      "step": 119
+    },
+    {
+      "epoch": 0.32,
+      "learning_rate": 1.9698243229000004e-05,
+      "loss": 0.7178,
+      "step": 120
+    },
+    {
+      "epoch": 0.32,
+      "learning_rate": 1.9691221378121565e-05,
+      "loss": 0.6709,
+      "step": 121
+    },
+    {
+      "epoch": 0.32,
+      "learning_rate": 1.9684120045989628e-05,
+      "loss": 0.7374,
+      "step": 122
+    },
+    {
+      "epoch": 0.33,
+      "learning_rate": 1.9676939290844813e-05,
+      "loss": 0.7067,
+      "step": 123
+    },
+    {
+      "epoch": 0.33,
+      "learning_rate": 1.966967917157912e-05,
+      "loss": 0.7454,
+      "step": 124
+    },
+    {
+      "epoch": 0.33,
+      "learning_rate": 1.9662339747735447e-05,
+      "loss": 0.7275,
+      "step": 125
+    },
+    {
+      "epoch": 0.33,
+      "learning_rate": 1.965492107950709e-05,
+      "loss": 0.7296,
+      "step": 126
+    },
+    {
+      "epoch": 0.34,
+      "learning_rate": 1.9647423227737273e-05,
+      "loss": 0.7348,
+      "step": 127
+    },
+    {
+      "epoch": 0.34,
+      "learning_rate": 1.9639846253918614e-05,
+      "loss": 0.7549,
+      "step": 128
+    },
+    {
+      "epoch": 0.34,
+      "learning_rate": 1.963219022019265e-05,
+      "loss": 0.6658,
+      "step": 129
+    },
+    {
+      "epoch": 0.34,
+      "learning_rate": 1.9624455189349318e-05,
+      "loss": 0.7123,
+      "step": 130
+    },
+    {
+      "epoch": 0.35,
+      "learning_rate": 1.9616641224826434e-05,
+      "loss": 0.6769,
+      "step": 131
+    },
+    {
+      "epoch": 0.35,
+      "learning_rate": 1.9608748390709182e-05,
+      "loss": 0.7563,
+      "step": 132
+    },
+    {
+      "epoch": 0.35,
+      "learning_rate": 1.9600776751729582e-05,
+      "loss": 0.7463,
+      "step": 133
+    },
+    {
+      "epoch": 0.36,
+      "learning_rate": 1.959272637326597e-05,
+      "loss": 0.6794,
+      "step": 134
+    },
+    {
+      "epoch": 0.36,
+      "learning_rate": 1.9584597321342434e-05,
+      "loss": 0.6992,
+      "step": 135
+    },
+    {
+      "epoch": 0.36,
+      "learning_rate": 1.9576389662628316e-05,
+      "loss": 0.7092,
+      "step": 136
+    },
+    {
+      "epoch": 0.36,
+      "learning_rate": 1.956810346443762e-05,
+      "loss": 0.7031,
+      "step": 137
+    },
+    {
+      "epoch": 0.37,
+      "learning_rate": 1.9559738794728492e-05,
+      "loss": 0.7276,
+      "step": 138
+    },
+    {
+      "epoch": 0.37,
+      "learning_rate": 1.9551295722102655e-05,
+      "loss": 0.7264,
+      "step": 139
+    },
+    {
+      "epoch": 0.37,
+      "learning_rate": 1.9542774315804834e-05,
+      "loss": 0.7306,
+      "step": 140
+    },
+    {
+      "epoch": 0.37,
+      "learning_rate": 1.9534174645722204e-05,
+      "loss": 0.7021,
+      "step": 141
+    },
+    {
+      "epoch": 0.38,
+      "learning_rate": 1.95254967823838e-05,
+      "loss": 0.7173,
+      "step": 142
+    },
+    {
+      "epoch": 0.38,
+      "learning_rate": 1.9516740796959963e-05,
+      "loss": 0.7302,
+      "step": 143
+    },
+    {
+      "epoch": 0.38,
+      "learning_rate": 1.9507906761261727e-05,
+      "loss": 0.6993,
+      "step": 144
+    },
+    {
+      "epoch": 0.38,
+      "learning_rate": 1.949899474774026e-05,
+      "loss": 0.7269,
+      "step": 145
+    },
+    {
+      "epoch": 0.39,
+      "learning_rate": 1.949000482948624e-05,
+      "loss": 0.682,
+      "step": 146
+    },
+    {
+      "epoch": 0.39,
+      "learning_rate": 1.9480937080229277e-05,
+      "loss": 0.6963,
+      "step": 147
+    },
+    {
+      "epoch": 0.39,
+      "learning_rate": 1.9471791574337304e-05,
+      "loss": 0.7218,
+      "step": 148
+    },
+    {
+      "epoch": 0.4,
+      "learning_rate": 1.9462568386815964e-05,
+      "loss": 0.6921,
+      "step": 149
+    },
+    {
+      "epoch": 0.4,
+      "learning_rate": 1.945326759330799e-05,
+      "loss": 0.6943,
+      "step": 150
+    },
+    {
+      "epoch": 0.4,
+      "learning_rate": 1.9443889270092595e-05,
+      "loss": 0.7063,
+      "step": 151
+    },
+    {
+      "epoch": 0.4,
+      "learning_rate": 1.943443349408484e-05,
+      "loss": 0.6983,
+      "step": 152
+    },
+    {
+      "epoch": 0.41,
+      "learning_rate": 1.9424900342835004e-05,
+      "loss": 0.7664,
+      "step": 153
+    },
+    {
+      "epoch": 0.41,
+      "learning_rate": 1.941528989452795e-05,
+      "loss": 0.7031,
+      "step": 154
+    },
+    {
+      "epoch": 0.41,
+      "learning_rate": 1.9405602227982487e-05,
+      "loss": 0.7473,
+      "step": 155
+    },
+    {
+      "epoch": 0.41,
+      "learning_rate": 1.9395837422650707e-05,
+      "loss": 0.7074,
+      "step": 156
+    },
+    {
+      "epoch": 0.42,
+      "learning_rate": 1.9385995558617355e-05,
+      "loss": 0.728,
+      "step": 157
+    },
+    {
+      "epoch": 0.42,
+      "learning_rate": 1.9376076716599157e-05,
+      "loss": 0.7425,
+      "step": 158
+    },
+    {
+      "epoch": 0.42,
+      "learning_rate": 1.9366080977944172e-05,
+      "loss": 0.6808,
+      "step": 159
+    },
+    {
+      "epoch": 0.42,
+      "learning_rate": 1.935600842463111e-05,
+      "loss": 0.6985,
+      "step": 160
+    },
+    {
+      "epoch": 0.43,
+      "learning_rate": 1.9345859139268664e-05,
+      "loss": 0.6866,
+      "step": 161
+    },
+    {
+      "epoch": 0.43,
+      "learning_rate": 1.933563320509484e-05,
+      "loss": 0.7243,
+      "step": 162
+    },
+    {
+      "epoch": 0.43,
+      "learning_rate": 1.9325330705976267e-05,
+      "loss": 0.7236,
+      "step": 163
+    },
+    {
+      "epoch": 0.44,
+      "learning_rate": 1.931495172640751e-05,
+      "loss": 0.696,
+      "step": 164
+    },
+    {
+      "epoch": 0.44,
+      "learning_rate": 1.930449635151038e-05,
+      "loss": 0.7068,
+      "step": 165
+    },
+    {
+      "epoch": 0.44,
+      "learning_rate": 1.9293964667033235e-05,
+      "loss": 0.7202,
+      "step": 166
+    },
+    {
+      "epoch": 0.44,
+      "learning_rate": 1.928335675935027e-05,
+      "loss": 0.6921,
+      "step": 167
+    },
+    {
+      "epoch": 0.45,
+      "learning_rate": 1.9272672715460822e-05,
+      "loss": 0.7461,
+      "step": 168
+    },
+    {
+      "epoch": 0.45,
+      "learning_rate": 1.9261912622988642e-05,
+      "loss": 0.7397,
+      "step": 169
+    },
+    {
+      "epoch": 0.45,
+      "learning_rate": 1.9251076570181188e-05,
+      "loss": 0.6802,
+      "step": 170
+    },
+    {
+      "epoch": 0.45,
+      "learning_rate": 1.9240164645908898e-05,
+      "loss": 0.7231,
+      "step": 171
+    },
+    {
+      "epoch": 0.46,
+      "learning_rate": 1.922917693966445e-05,
+      "loss": 0.7013,
+      "step": 172
+    },
+    {
+      "epoch": 0.46,
+      "learning_rate": 1.9218113541562052e-05,
+      "loss": 0.7054,
+      "step": 173
+    },
+    {
+      "epoch": 0.46,
+      "learning_rate": 1.9206974542336673e-05,
+      "loss": 0.7029,
+      "step": 174
+    },
+    {
+      "epoch": 0.46,
+      "learning_rate": 1.919576003334333e-05,
+      "loss": 0.6876,
+      "step": 175
+    },
+    {
+      "epoch": 0.47,
+      "learning_rate": 1.9184470106556315e-05,
+      "loss": 0.7253,
+      "step": 176
+    },
+    {
+      "epoch": 0.47,
+      "learning_rate": 1.9173104854568446e-05,
+      "loss": 0.7339,
+      "step": 177
+    },
+    {
+      "epoch": 0.47,
+      "learning_rate": 1.9161664370590312e-05,
+      "loss": 0.6693,
+      "step": 178
+    },
+    {
+      "epoch": 0.47,
+      "learning_rate": 1.9150148748449516e-05,
+      "loss": 0.7066,
+      "step": 179
+    },
+    {
+      "epoch": 0.48,
+      "learning_rate": 1.913855808258989e-05,
+      "loss": 0.7135,
+      "step": 180
+    },
+    {
+      "epoch": 0.48,
+      "learning_rate": 1.9126892468070718e-05,
+      "loss": 0.7125,
+      "step": 181
+    },
+    {
+      "epoch": 0.48,
+      "learning_rate": 1.9115152000565982e-05,
+      "loss": 0.7445,
+      "step": 182
+    },
+    {
+      "epoch": 0.49,
+      "learning_rate": 1.910333677636355e-05,
+      "loss": 0.7319,
+      "step": 183
+    },
+    {
+      "epoch": 0.49,
+      "learning_rate": 1.9091446892364406e-05,
+      "loss": 0.727,
+      "step": 184
+    },
+    {
+      "epoch": 0.49,
+      "learning_rate": 1.907948244608183e-05,
+      "loss": 0.6924,
+      "step": 185
+    },
+    {
+      "epoch": 0.49,
+      "learning_rate": 1.9067443535640635e-05,
+      "loss": 0.6977,
+      "step": 186
+    },
+    {
+      "epoch": 0.5,
+      "learning_rate": 1.905533025977632e-05,
+      "loss": 0.7166,
+      "step": 187
+    },
+    {
+      "epoch": 0.5,
+      "learning_rate": 1.9043142717834306e-05,
+      "loss": 0.7165,
+      "step": 188
+    },
+    {
+      "epoch": 0.5,
+      "learning_rate": 1.903088100976908e-05,
+      "loss": 0.6808,
+      "step": 189
+    },
+    {
+      "epoch": 0.5,
+      "learning_rate": 1.9018545236143398e-05,
+      "loss": 0.7437,
+      "step": 190
+    },
+    {
+      "epoch": 0.51,
+      "learning_rate": 1.9006135498127452e-05,
+      "loss": 0.7316,
+      "step": 191
+    },
+    {
+      "epoch": 0.51,
+      "learning_rate": 1.899365189749805e-05,
+      "loss": 0.6841,
+      "step": 192
+    },
+    {
+      "epoch": 0.51,
+      "learning_rate": 1.898109453663776e-05,
+      "loss": 0.7055,
+      "step": 193
+    },
+    {
+      "epoch": 0.51,
+      "learning_rate": 1.896846351853411e-05,
+      "loss": 0.7141,
+      "step": 194
+    },
+    {
+      "epoch": 0.52,
+      "learning_rate": 1.895575894677869e-05,
+      "loss": 0.7071,
+      "step": 195
+    },
+    {
+      "epoch": 0.52,
+      "learning_rate": 1.894298092556635e-05,
+      "loss": 0.6967,
+      "step": 196
+    },
+    {
+      "epoch": 0.52,
+      "learning_rate": 1.8930129559694312e-05,
+      "loss": 0.7167,
+      "step": 197
+    },
+    {
+      "epoch": 0.53,
+      "learning_rate": 1.8917204954561337e-05,
+      "loss": 0.6736,
+      "step": 198
+    },
+    {
+      "epoch": 0.53,
+      "learning_rate": 1.8904207216166837e-05,
+      "loss": 0.6796,
+      "step": 199
+    },
+    {
+      "epoch": 0.53,
+      "learning_rate": 1.8891136451110027e-05,
+      "loss": 0.6615,
+      "step": 200
+    },
+    {
+      "epoch": 0.53,
+      "learning_rate": 1.887799276658903e-05,
+      "loss": 0.7325,
+      "step": 201
+    },
+    {
+      "epoch": 0.54,
+      "learning_rate": 1.886477627040002e-05,
+      "loss": 0.6827,
+      "step": 202
+    },
+    {
+      "epoch": 0.54,
+      "learning_rate": 1.8851487070936314e-05,
+      "loss": 0.7264,
+      "step": 203
+    },
+    {
+      "epoch": 0.54,
+      "learning_rate": 1.88381252771875e-05,
+      "loss": 0.7269,
+      "step": 204
+    },
+    {
+      "epoch": 0.54,
+      "learning_rate": 1.8824690998738543e-05,
+      "loss": 0.6851,
+      "step": 205
+    },
+    {
+      "epoch": 0.55,
+      "learning_rate": 1.8811184345768875e-05,
+      "loss": 0.7106,
+      "step": 206
+    },
+    {
+      "epoch": 0.55,
+      "learning_rate": 1.87976054290515e-05,
+      "loss": 0.6996,
+      "step": 207
+    },
+    {
+      "epoch": 0.55,
+      "learning_rate": 1.8783954359952088e-05,
+      "loss": 0.7013,
+      "step": 208
+    },
+    {
+      "epoch": 0.55,
+      "learning_rate": 1.8770231250428046e-05,
+      "loss": 0.7316,
+      "step": 209
+    },
+    {
+      "epoch": 0.56,
+      "learning_rate": 1.8756436213027625e-05,
+      "loss": 0.6962,
+      "step": 210
+    },
+    {
+      "epoch": 0.56,
+      "learning_rate": 1.8742569360888966e-05,
+      "loss": 0.701,
+      "step": 211
+    },
+    {
+      "epoch": 0.56,
+      "learning_rate": 1.8728630807739212e-05,
+      "loss": 0.7039,
+      "step": 212
+    },
+    {
+      "epoch": 0.56,
+      "learning_rate": 1.871462066789353e-05,
+      "loss": 0.7335,
+      "step": 213
+    },
+    {
+      "epoch": 0.57,
+      "learning_rate": 1.8700539056254204e-05,
+      "loss": 0.7368,
+      "step": 214
+    },
+    {
+      "epoch": 0.57,
+      "learning_rate": 1.8686386088309686e-05,
+      "loss": 0.6989,
+      "step": 215
+    },
+    {
+      "epoch": 0.57,
+      "learning_rate": 1.8672161880133654e-05,
+      "loss": 0.6648,
+      "step": 216
+    },
+    {
+      "epoch": 0.58,
+      "learning_rate": 1.865786654838403e-05,
+      "loss": 0.7051,
+      "step": 217
+    },
+    {
+      "epoch": 0.58,
+      "learning_rate": 1.8643500210302077e-05,
+      "loss": 0.7133,
+      "step": 218
+    },
+    {
+      "epoch": 0.58,
+      "learning_rate": 1.8629062983711383e-05,
+      "loss": 0.7101,
+      "step": 219
+    },
+    {
+      "epoch": 0.58,
+      "learning_rate": 1.8614554987016933e-05,
+      "loss": 0.6877,
+      "step": 220
+    },
+    {
+      "epoch": 0.59,
+      "learning_rate": 1.8599976339204107e-05,
+      "loss": 0.693,
+      "step": 221
+    },
+    {
+      "epoch": 0.59,
+      "learning_rate": 1.8585327159837747e-05,
+      "loss": 0.7275,
+      "step": 222
+    },
+    {
+      "epoch": 0.59,
+      "learning_rate": 1.857060756906112e-05,
+      "loss": 0.6677,
+      "step": 223
+    },
+    {
+      "epoch": 0.59,
+      "learning_rate": 1.8555817687594986e-05,
+      "loss": 0.6888,
+      "step": 224
+    },
+    {
+      "epoch": 0.6,
+      "learning_rate": 1.8540957636736574e-05,
+      "loss": 0.7139,
+      "step": 225
+    },
+    {
+      "epoch": 0.6,
+      "learning_rate": 1.85260275383586e-05,
+      "loss": 0.7075,
+      "step": 226
+    },
+    {
+      "epoch": 0.6,
+      "learning_rate": 1.8511027514908262e-05,
+      "loss": 0.692,
+      "step": 227
+    },
+    {
+      "epoch": 0.6,
+      "learning_rate": 1.8495957689406244e-05,
+      "loss": 0.6889,
+      "step": 228
+    },
+    {
+      "epoch": 0.61,
+      "learning_rate": 1.8480818185445707e-05,
+      "loss": 0.6722,
+      "step": 229
+    },
+    {
+      "epoch": 0.61,
+      "learning_rate": 1.8465609127191253e-05,
+      "loss": 0.6917,
+      "step": 230
+    },
+    {
+      "epoch": 0.61,
+      "learning_rate": 1.8450330639377942e-05,
+      "loss": 0.7066,
+      "step": 231
+    },
+    {
+      "epoch": 0.62,
+      "learning_rate": 1.8434982847310245e-05,
+      "loss": 0.7257,
+      "step": 232
+    },
+    {
+      "epoch": 0.62,
+      "learning_rate": 1.841956587686102e-05,
+      "loss": 0.7625,
+      "step": 233
+    },
+    {
+      "epoch": 0.62,
+      "learning_rate": 1.840407985447048e-05,
+      "loss": 0.6815,
+      "step": 234
+    },
+    {
+      "epoch": 0.62,
+      "learning_rate": 1.8388524907145172e-05,
+      "loss": 0.7389,
+      "step": 235
+    },
+    {
+      "epoch": 0.63,
+      "learning_rate": 1.837290116245691e-05,
+      "loss": 0.6997,
+      "step": 236
+    },
+    {
+      "epoch": 0.63,
+      "learning_rate": 1.835720874854173e-05,
+      "loss": 0.7163,
+      "step": 237
+    },
+    {
+      "epoch": 0.63,
+      "learning_rate": 1.834144779409888e-05,
+      "loss": 0.6948,
+      "step": 238
+    },
+    {
+      "epoch": 0.63,
+      "learning_rate": 1.8325618428389704e-05,
+      "loss": 0.6918,
+      "step": 239
+    },
+    {
+      "epoch": 0.64,
+      "learning_rate": 1.8309720781236627e-05,
+      "loss": 0.696,
+      "step": 240
+    },
+    {
+      "epoch": 0.64,
+      "learning_rate": 1.8293754983022073e-05,
+      "loss": 0.7244,
+      "step": 241
+    },
+    {
+      "epoch": 0.64,
+      "learning_rate": 1.8277721164687393e-05,
+      "loss": 0.703,
+      "step": 242
+    },
+    {
+      "epoch": 0.64,
+      "learning_rate": 1.8261619457731804e-05,
+      "loss": 0.6727,
+      "step": 243
+    },
+    {
+      "epoch": 0.65,
+      "learning_rate": 1.8245449994211295e-05,
+      "loss": 0.7425,
+      "step": 244
+    },
+    {
+      "epoch": 0.65,
+      "learning_rate": 1.8229212906737556e-05,
+      "loss": 0.7196,
+      "step": 245
+    },
+    {
+      "epoch": 0.65,
+      "learning_rate": 1.8212908328476885e-05,
+      "loss": 0.6612,
+      "step": 246
+    },
+    {
+      "epoch": 0.66,
+      "learning_rate": 1.8196536393149095e-05,
+      "loss": 0.7293,
+      "step": 247
+    },
+    {
+      "epoch": 0.66,
+      "learning_rate": 1.8180097235026416e-05,
+      "loss": 0.7252,
+      "step": 248
+    },
+    {
+      "epoch": 0.66,
+      "learning_rate": 1.8163590988932402e-05,
+      "loss": 0.7388,
+      "step": 249
+    },
+    {
+      "epoch": 0.66,
+      "learning_rate": 1.8147017790240825e-05,
+      "loss": 0.6647,
+      "step": 250
+    },
+    {
+      "epoch": 0.67,
+      "learning_rate": 1.8130377774874545e-05,
+      "loss": 0.6971,
+      "step": 251
+    },
+    {
+      "epoch": 0.67,
+      "learning_rate": 1.811367107930442e-05,
+      "loss": 0.6929,
+      "step": 252
+    },
+    {
+      "epoch": 0.67,
+      "learning_rate": 1.8096897840548176e-05,
+      "loss": 0.6738,
+      "step": 253
+    },
+    {
+      "epoch": 0.67,
+      "learning_rate": 1.808005819616929e-05,
+      "loss": 0.6767,
+      "step": 254
+    },
+    {
+      "epoch": 0.68,
+      "learning_rate": 1.8063152284275837e-05,
+      "loss": 0.6992,
+      "step": 255
+    },
+    {
+      "epoch": 0.68,
+      "learning_rate": 1.80461802435194e-05,
+      "loss": 0.6611,
+      "step": 256
+    },
+    {
+      "epoch": 0.68,
+      "learning_rate": 1.802914221309388e-05,
+      "loss": 0.6698,
+      "step": 257
+    },
+    {
+      "epoch": 0.68,
+      "learning_rate": 1.8012038332734417e-05,
+      "loss": 0.6605,
+      "step": 258
+    },
+    {
+      "epoch": 0.69,
+      "learning_rate": 1.7994868742716186e-05,
+      "loss": 0.6865,
+      "step": 259
+    },
+    {
+      "epoch": 0.69,
+      "learning_rate": 1.797763358385328e-05,
+      "loss": 0.7283,
+      "step": 260
+    },
+    {
+      "epoch": 0.69,
+      "learning_rate": 1.7960332997497548e-05,
+      "loss": 0.768,
+      "step": 261
+    },
+    {
+      "epoch": 0.69,
+      "learning_rate": 1.7942967125537426e-05,
+      "loss": 0.6642,
+      "step": 262
+    },
+    {
+      "epoch": 0.7,
+      "learning_rate": 1.7925536110396787e-05,
+      "loss": 0.7428,
+      "step": 263
+    },
+    {
+      "epoch": 0.7,
+      "learning_rate": 1.790804009503377e-05,
+      "loss": 0.6567,
+      "step": 264
+    },
+    {
+      "epoch": 0.7,
+      "learning_rate": 1.7890479222939594e-05,
+      "loss": 0.7206,
+      "step": 265
+    },
+    {
+      "epoch": 0.71,
+      "learning_rate": 1.78728536381374e-05,
+      "loss": 0.7375,
+      "step": 266
+    },
+    {
+      "epoch": 0.71,
+      "learning_rate": 1.7855163485181065e-05,
+      "loss": 0.7377,
+      "step": 267
+    },
+    {
+      "epoch": 0.71,
+      "learning_rate": 1.7837408909154006e-05,
+      "loss": 0.6782,
+      "step": 268
+    },
+    {
+      "epoch": 0.71,
+      "learning_rate": 1.7819590055667995e-05,
+      "loss": 0.705,
+      "step": 269
+    },
+    {
+      "epoch": 0.72,
+      "learning_rate": 1.7801707070861972e-05,
+      "loss": 0.6659,
+      "step": 270
+    },
+    {
+      "epoch": 0.72,
+      "learning_rate": 1.7783760101400846e-05,
+      "loss": 0.7313,
+      "step": 271
+    },
+    {
+      "epoch": 0.72,
+      "learning_rate": 1.7765749294474276e-05,
+      "loss": 0.6779,
+      "step": 272
+    },
+    {
+      "epoch": 0.72,
+      "learning_rate": 1.7747674797795483e-05,
+      "loss": 0.721,
+      "step": 273
+    },
+    {
+      "epoch": 0.73,
+      "learning_rate": 1.7729536759600033e-05,
+      "loss": 0.6872,
+      "step": 274
+    },
+    {
+      "epoch": 0.73,
+      "learning_rate": 1.7711335328644613e-05,
+      "loss": 0.7057,
+      "step": 275
+    },
+    {
+      "epoch": 0.73,
+      "learning_rate": 1.769307065420582e-05,
+      "loss": 0.6899,
+      "step": 276
+    },
+    {
+      "epoch": 0.73,
+      "learning_rate": 1.7674742886078934e-05,
+      "loss": 0.6848,
+      "step": 277
+    },
+    {
+      "epoch": 0.74,
+      "learning_rate": 1.765635217457669e-05,
+      "loss": 0.7153,
+      "step": 278
+    },
+    {
+      "epoch": 0.74,
+      "learning_rate": 1.7637898670528043e-05,
+      "loss": 0.7433,
+      "step": 279
+    },
+    {
+      "epoch": 0.74,
+      "learning_rate": 1.7619382525276935e-05,
+      "loss": 0.6842,
+      "step": 280
+    },
+    {
+      "epoch": 0.75,
+      "learning_rate": 1.7600803890681047e-05,
+      "loss": 0.6968,
+      "step": 281
+    },
+    {
+      "epoch": 0.75,
+      "learning_rate": 1.7582162919110564e-05,
+      "loss": 0.7256,
+      "step": 282
+    },
+    {
+      "epoch": 0.75,
+      "learning_rate": 1.7563459763446916e-05,
+      "loss": 0.7315,
+      "step": 283
+    },
+    {
+      "epoch": 0.75,
+      "learning_rate": 1.7544694577081526e-05,
+      "loss": 0.6996,
+      "step": 284
+    },
+    {
+      "epoch": 0.76,
+      "learning_rate": 1.752586751391456e-05,
+      "loss": 0.7242,
+      "step": 285
+    },
+    {
+      "epoch": 0.76,
+      "learning_rate": 1.7506978728353653e-05,
+      "loss": 0.6893,
+      "step": 286
+    },
+    {
+      "epoch": 0.76,
+      "learning_rate": 1.748802837531264e-05,
+      "loss": 0.7117,
+      "step": 287
+    },
+    {
+      "epoch": 0.76,
+      "learning_rate": 1.7469016610210316e-05,
+      "loss": 0.7131,
+      "step": 288
+    },
+    {
+      "epoch": 0.77,
+      "learning_rate": 1.744994358896911e-05,
+      "loss": 0.7092,
+      "step": 289
+    },
+    {
+      "epoch": 0.77,
+      "learning_rate": 1.7430809468013863e-05,
+      "loss": 0.6709,
+      "step": 290
+    },
+    {
+      "epoch": 0.77,
+      "learning_rate": 1.74116144042705e-05,
+      "loss": 0.7048,
+      "step": 291
+    },
+    {
+      "epoch": 0.77,
+      "learning_rate": 1.739235855516476e-05,
+      "loss": 0.6825,
+      "step": 292
+    },
+    {
+      "epoch": 0.78,
+      "learning_rate": 1.7373042078620906e-05,
+      "loss": 0.6992,
+      "step": 293
+    },
+    {
+      "epoch": 0.78,
+      "learning_rate": 1.7353665133060443e-05,
+      "loss": 0.7204,
+      "step": 294
+    },
+    {
+      "epoch": 0.78,
+      "learning_rate": 1.733422787740079e-05,
+      "loss": 0.6945,
+      "step": 295
+    },
+    {
+      "epoch": 0.79,
+      "learning_rate": 1.731473047105399e-05,
+      "loss": 0.7133,
+      "step": 296
+    },
+    {
+      "epoch": 0.79,
+      "learning_rate": 1.7295173073925417e-05,
+      "loss": 0.696,
+      "step": 297
+    },
+    {
+      "epoch": 0.79,
+      "learning_rate": 1.727555584641243e-05,
+      "loss": 0.7015,
+      "step": 298
+    },
+    {
+      "epoch": 0.79,
+      "learning_rate": 1.72558789494031e-05,
+      "loss": 0.6816,
+      "step": 299
+    },
+    {
+      "epoch": 0.8,
+      "learning_rate": 1.723614254427486e-05,
+      "loss": 0.6822,
+      "step": 300
+    },
+    {
+      "epoch": 0.8,
+      "learning_rate": 1.7216346792893196e-05,
+      "loss": 0.6796,
+      "step": 301
+    },
+    {
+      "epoch": 0.8,
+      "learning_rate": 1.7196491857610304e-05,
+      "loss": 0.7258,
+      "step": 302
+    },
+    {
+      "epoch": 0.8,
+      "learning_rate": 1.717657790126378e-05,
+      "loss": 0.7766,
+      "step": 303
+    },
+    {
+      "epoch": 0.81,
+      "learning_rate": 1.7156605087175264e-05,
+      "loss": 0.6758,
+      "step": 304
+    },
+    {
+      "epoch": 0.81,
+      "learning_rate": 1.7136573579149115e-05,
+      "loss": 0.694,
+      "step": 305
+    },
+    {
+      "epoch": 0.81,
+      "learning_rate": 1.7116483541471065e-05,
+      "loss": 0.6936,
+      "step": 306
+    },
+    {
+      "epoch": 0.81,
+      "learning_rate": 1.7096335138906857e-05,
+      "loss": 0.6857,
+      "step": 307
+    },
+    {
+      "epoch": 0.82,
+      "learning_rate": 1.707612853670093e-05,
+      "loss": 0.6917,
+      "step": 308
+    },
+    {
+      "epoch": 0.82,
+      "learning_rate": 1.7055863900575014e-05,
+      "loss": 0.7102,
+      "step": 309
+    },
+    {
+      "epoch": 0.82,
+      "learning_rate": 1.703554139672681e-05,
+      "loss": 0.7113,
+      "step": 310
+    },
+    {
+      "epoch": 0.82,
+      "learning_rate": 1.7015161191828608e-05,
+      "loss": 0.6865,
+      "step": 311
+    },
+    {
+      "epoch": 0.83,
+      "learning_rate": 1.699472345302593e-05,
+      "loss": 0.7487,
+      "step": 312
+    },
+    {
+      "epoch": 0.83,
+      "learning_rate": 1.6974228347936148e-05,
+      "loss": 0.6962,
+      "step": 313
+    },
+    {
+      "epoch": 0.83,
+      "learning_rate": 1.695367604464712e-05,
+      "loss": 0.6851,
+      "step": 314
+    },
+    {
+      "epoch": 0.84,
+      "learning_rate": 1.6933066711715815e-05,
+      "loss": 0.7054,
+      "step": 315
+    },
+    {
+      "epoch": 0.84,
+      "learning_rate": 1.6912400518166907e-05,
+      "loss": 0.7169,
+      "step": 316
+    },
+    {
+      "epoch": 0.84,
+      "learning_rate": 1.6891677633491418e-05,
+      "loss": 0.6679,
+      "step": 317
+    },
+    {
+      "epoch": 0.84,
+      "learning_rate": 1.6870898227645308e-05,
+      "loss": 0.6934,
+      "step": 318
+    },
+    {
+      "epoch": 0.85,
+      "learning_rate": 1.6850062471048093e-05,
+      "loss": 0.7144,
+      "step": 319
+    },
+    {
+      "epoch": 0.85,
+      "learning_rate": 1.6829170534581432e-05,
+      "loss": 0.6778,
+      "step": 320
+    },
+    {
+      "epoch": 0.85,
+      "learning_rate": 1.6808222589587745e-05,
+      "loss": 0.7337,
+      "step": 321
+    },
+    {
+      "epoch": 0.85,
+      "learning_rate": 1.6787218807868797e-05,
+      "loss": 0.7197,
+      "step": 322
+    },
+    {
+      "epoch": 0.86,
+      "learning_rate": 1.676615936168429e-05,
+      "loss": 0.6604,
+      "step": 323
+    },
+    {
+      "epoch": 0.86,
+      "learning_rate": 1.6745044423750448e-05,
+      "loss": 0.6784,
+      "step": 324
+    },
+    {
+      "epoch": 0.86,
+      "learning_rate": 1.672387416723861e-05,
+      "loss": 0.7171,
+      "step": 325
+    },
+    {
+      "epoch": 0.86,
+      "learning_rate": 1.670264876577379e-05,
+      "loss": 0.685,
+      "step": 326
+    },
+    {
+      "epoch": 0.87,
+      "learning_rate": 1.668136839343328e-05,
+      "loss": 0.6835,
+      "step": 327
+    },
+    {
+      "epoch": 0.87,
+      "learning_rate": 1.6660033224745197e-05,
+      "loss": 0.7291,
+      "step": 328
+    },
+    {
+      "epoch": 0.87,
+      "learning_rate": 1.663864343468707e-05,
+      "loss": 0.7151,
+      "step": 329
+    },
+    {
+      "epoch": 0.88,
+      "learning_rate": 1.6617199198684403e-05,
+      "loss": 0.6924,
+      "step": 330
+    },
+    {
+      "epoch": 0.88,
+      "learning_rate": 1.6595700692609216e-05,
+      "loss": 0.7131,
+      "step": 331
+    },
+    {
+      "epoch": 0.88,
+      "learning_rate": 1.6574148092778625e-05,
+      "loss": 0.684,
+      "step": 332
+    },
+    {
+      "epoch": 0.88,
+      "learning_rate": 1.6552541575953395e-05,
+      "loss": 0.7235,
+      "step": 333
+    },
+    {
+      "epoch": 0.89,
+      "learning_rate": 1.6530881319336472e-05,
+      "loss": 0.6951,
+      "step": 334
+    },
+    {
+      "epoch": 0.89,
+      "learning_rate": 1.6509167500571553e-05,
+      "loss": 0.6609,
+      "step": 335
+    },
+    {
+      "epoch": 0.89,
+      "learning_rate": 1.6487400297741613e-05,
+      "loss": 0.6704,
+      "step": 336
+    },
+    {
+      "epoch": 0.89,
+      "learning_rate": 1.6465579889367447e-05,
+      "loss": 0.6696,
+      "step": 337
+    },
+    {
+      "epoch": 0.9,
+      "learning_rate": 1.6443706454406213e-05,
+      "loss": 0.7034,
+      "step": 338
+    },
+    {
+      "epoch": 0.9,
+      "learning_rate": 1.6421780172249955e-05,
+      "loss": 0.7081,
+      "step": 339
+    },
+    {
+      "epoch": 0.9,
+      "learning_rate": 1.6399801222724137e-05,
+      "loss": 0.6883,
+      "step": 340
+    },
+    {
+      "epoch": 0.9,
+      "learning_rate": 1.6377769786086165e-05,
+      "loss": 0.7224,
+      "step": 341
+    },
+    {
+      "epoch": 0.91,
+      "learning_rate": 1.6355686043023926e-05,
+      "loss": 0.7377,
+      "step": 342
+    },
+    {
+      "epoch": 0.91,
+      "learning_rate": 1.6333550174654266e-05,
+      "loss": 0.6724,
+      "step": 343
+    },
+    {
+      "epoch": 0.91,
+      "learning_rate": 1.6311362362521554e-05,
+      "loss": 0.7034,
+      "step": 344
+    },
+    {
+      "epoch": 0.92,
+      "learning_rate": 1.628912278859615e-05,
+      "loss": 0.7172,
+      "step": 345
+    },
+    {
+      "epoch": 0.92,
+      "learning_rate": 1.6266831635272945e-05,
+      "loss": 0.6961,
+      "step": 346
+    },
+    {
+      "epoch": 0.92,
+      "learning_rate": 1.6244489085369842e-05,
+      "loss": 0.6823,
+      "step": 347
+    },
+    {
+      "epoch": 0.92,
+      "learning_rate": 1.6222095322126272e-05,
+      "loss": 0.6961,
+      "step": 348
+    },
+    {
+      "epoch": 0.93,
+      "learning_rate": 1.6199650529201686e-05,
+      "loss": 0.6841,
+      "step": 349
+    },
+    {
+      "epoch": 0.93,
+      "learning_rate": 1.6177154890674042e-05,
+      "loss": 0.7198,
+      "step": 350
+    },
+    {
+      "epoch": 0.93,
+      "learning_rate": 1.6154608591038304e-05,
+      "loss": 0.6891,
+      "step": 351
+    },
+    {
+      "epoch": 0.93,
+      "learning_rate": 1.613201181520493e-05,
+      "loss": 0.6825,
+      "step": 352
+    },
+    {
+      "epoch": 0.94,
+      "learning_rate": 1.610936474849835e-05,
+      "loss": 0.7094,
+      "step": 353
+    },
+    {
+      "epoch": 0.94,
+      "learning_rate": 1.6086667576655442e-05,
+      "loss": 0.6611,
+      "step": 354
+    },
+    {
+      "epoch": 0.94,
+      "learning_rate": 1.6063920485824023e-05,
+      "loss": 0.6996,
+      "step": 355
+    },
+    {
+      "epoch": 0.94,
+      "learning_rate": 1.604112366256131e-05,
+      "loss": 0.6903,
+      "step": 356
+    },
+    {
+      "epoch": 0.95,
+      "learning_rate": 1.6018277293832395e-05,
+      "loss": 0.653,
+      "step": 357
+    },
+    {
+      "epoch": 0.95,
+      "learning_rate": 1.5995381567008704e-05,
+      "loss": 0.7033,
+      "step": 358
+    },
+    {
+      "epoch": 0.95,
+      "learning_rate": 1.5972436669866484e-05,
+      "loss": 0.6447,
+      "step": 359
+    },
+    {
+      "epoch": 0.95,
+      "learning_rate": 1.594944279058522e-05,
+      "loss": 0.7068,
+      "step": 360
+    },
+    {
+      "epoch": 0.96,
+      "learning_rate": 1.5926400117746145e-05,
+      "loss": 0.6839,
+      "step": 361
+    },
+    {
+      "epoch": 0.96,
+      "learning_rate": 1.5903308840330646e-05,
+      "loss": 0.6575,
+      "step": 362
+    },
+    {
+      "epoch": 0.96,
+      "learning_rate": 1.588016914771874e-05,
+      "loss": 0.7181,
+      "step": 363
+    },
+    {
+      "epoch": 0.97,
+      "learning_rate": 1.5856981229687518e-05,
+      "loss": 0.7439,
+      "step": 364
+    },
+    {
+      "epoch": 0.97,
+      "learning_rate": 1.5833745276409575e-05,
+      "loss": 0.6368,
+      "step": 365
+    },
+    {
+      "epoch": 0.97,
+      "learning_rate": 1.581046147845148e-05,
+      "loss": 0.7047,
+      "step": 366
+    },
+    {
+      "epoch": 0.97,
+      "learning_rate": 1.5787130026772172e-05,
+      "loss": 0.6834,
+      "step": 367
+    },
+    {
+      "epoch": 0.98,
+      "learning_rate": 1.5763751112721433e-05,
+      "loss": 0.6911,
+      "step": 368
+    },
+    {
+      "epoch": 0.98,
+      "learning_rate": 1.5740324928038287e-05,
+      "loss": 0.6901,
+      "step": 369
+    },
+    {
+      "epoch": 0.98,
+      "learning_rate": 1.5716851664849457e-05,
+      "loss": 0.7076,
+      "step": 370
+    },
+    {
+      "epoch": 0.98,
+      "learning_rate": 1.5693331515667755e-05,
+      "loss": 0.7038,
+      "step": 371
+    },
+    {
+      "epoch": 0.99,
+      "learning_rate": 1.566976467339054e-05,
+      "loss": 0.7068,
+      "step": 372
+    },
+    {
+      "epoch": 0.99,
+      "learning_rate": 1.564615133129811e-05,
+      "loss": 0.7015,
+      "step": 373
+    },
+    {
+      "epoch": 0.99,
+      "learning_rate": 1.5622491683052126e-05,
+      "loss": 0.6992,
+      "step": 374
+    },
+    {
+      "epoch": 0.99,
+      "learning_rate": 1.559878592269402e-05,
+      "loss": 0.7193,
+      "step": 375
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 1.5575034244643412e-05,
+      "loss": 0.7418,
+      "step": 376
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 1.5551236843696504e-05,
+      "loss": 0.69,
+      "step": 377
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 1.5527393915024493e-05,
+      "loss": 0.6714,
+      "step": 378
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 1.5503505654171963e-05,
+      "loss": 0.6553,
+      "step": 379
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 1.5479572257055285e-05,
+      "loss": 0.6921,
+      "step": 380
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 1.545559391996101e-05,
+      "loss": 0.6479,
+      "step": 381
+    },
+    {
+      "epoch": 1.01,
+      "learning_rate": 1.5431570839544252e-05,
+      "loss": 0.6552,
+      "step": 382
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 1.5407503212827087e-05,
+      "loss": 0.6502,
+      "step": 383
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 1.5383391237196935e-05,
+      "loss": 0.6593,
+      "step": 384
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 1.5359235110404927e-05,
+      "loss": 0.6538,
+      "step": 385
+    },
+    {
+      "epoch": 1.02,
+      "learning_rate": 1.5335035030564303e-05,
+      "loss": 0.6542,
+      "step": 386
+    },
+    {
+      "epoch": 1.03,
+      "learning_rate": 1.5310791196148776e-05,
+      "loss": 0.6495,
+      "step": 387
+    },
+    {
+      "epoch": 1.03,
+      "learning_rate": 1.5286503805990906e-05,
+      "loss": 0.6988,
+      "step": 388
+    },
+    {
+      "epoch": 1.03,
+      "learning_rate": 1.5262173059280462e-05,
+      "loss": 0.6205,
+      "step": 389
+    },
+    {
+      "epoch": 1.03,
+      "learning_rate": 1.5237799155562809e-05,
+      "loss": 0.6545,
+      "step": 390
+    },
+    {
+      "epoch": 1.04,
+      "learning_rate": 1.5213382294737253e-05,
+      "loss": 0.6377,
+      "step": 391
+    },
+    {
+      "epoch": 1.04,
+      "learning_rate": 1.5188922677055401e-05,
+      "loss": 0.6612,
+      "step": 392
+    },
+    {
+      "epoch": 1.04,
+      "learning_rate": 1.5164420503119539e-05,
+      "loss": 0.649,
+      "step": 393
+    },
+    {
+      "epoch": 1.05,
+      "learning_rate": 1.5139875973880953e-05,
+      "loss": 0.6704,
+      "step": 394
+    },
+    {
+      "epoch": 1.05,
+      "learning_rate": 1.5115289290638315e-05,
+      "loss": 0.6602,
+      "step": 395
+    },
+    {
+      "epoch": 1.05,
+      "learning_rate": 1.5090660655036016e-05,
+      "loss": 0.6813,
+      "step": 396
+    },
+    {
+      "epoch": 1.05,
+      "learning_rate": 1.5065990269062503e-05,
+      "loss": 0.6454,
+      "step": 397
+    },
+    {
+      "epoch": 1.06,
+      "learning_rate": 1.5041278335048644e-05,
+      "loss": 0.6745,
+      "step": 398
+    },
+    {
+      "epoch": 1.06,
+      "learning_rate": 1.5016525055666058e-05,
+      "loss": 0.6826,
+      "step": 399
+    },
+    {
+      "epoch": 1.06,
+      "learning_rate": 1.4991730633925446e-05,
+      "loss": 0.6866,
+      "step": 400
+    },
+    {
+      "epoch": 1.06,
+      "learning_rate": 1.4966895273174934e-05,
+      "loss": 0.6837,
+      "step": 401
+    },
+    {
+      "epoch": 1.07,
+      "learning_rate": 1.494201917709841e-05,
+      "loss": 0.6688,
+      "step": 402
+    },
+    {
+      "epoch": 1.07,
+      "learning_rate": 1.4917102549713838e-05,
+      "loss": 0.6618,
+      "step": 403
+    },
+    {
+      "epoch": 1.07,
+      "learning_rate": 1.4892145595371599e-05,
+      "loss": 0.6649,
+      "step": 404
+    },
+    {
+      "epoch": 1.07,
+      "learning_rate": 1.4867148518752814e-05,
+      "loss": 0.6478,
+      "step": 405
+    },
+    {
+      "epoch": 1.08,
+      "learning_rate": 1.4842111524867654e-05,
+      "loss": 0.6396,
+      "step": 406
+    },
+    {
+      "epoch": 1.08,
+      "learning_rate": 1.4817034819053669e-05,
+      "loss": 0.6338,
+      "step": 407
+    },
+    {
+      "epoch": 1.08,
+      "learning_rate": 1.47919186069741e-05,
+      "loss": 0.6601,
+      "step": 408
+    },
+    {
+      "epoch": 1.08,
+      "learning_rate": 1.4766763094616199e-05,
+      "loss": 0.5969,
+      "step": 409
+    },
+    {
+      "epoch": 1.09,
+      "learning_rate": 1.4741568488289524e-05,
+      "loss": 0.6485,
+      "step": 410
+    },
+    {
+      "epoch": 1.09,
+      "learning_rate": 1.4716334994624269e-05,
+      "loss": 0.6812,
+      "step": 411
+    },
+    {
+      "epoch": 1.09,
+      "learning_rate": 1.4691062820569548e-05,
+      "loss": 0.6519,
+      "step": 412
+    },
+    {
+      "epoch": 1.1,
+      "learning_rate": 1.466575217339171e-05,
+      "loss": 0.6459,
+      "step": 413
+    },
+    {
+      "epoch": 1.1,
+      "learning_rate": 1.4640403260672637e-05,
+      "loss": 0.6461,
+      "step": 414
+    },
+    {
+      "epoch": 1.1,
+      "learning_rate": 1.4615016290308037e-05,
+      "loss": 0.6757,
+      "step": 415
+    },
+    {
+      "epoch": 1.1,
+      "learning_rate": 1.458959147050575e-05,
+      "loss": 0.6517,
+      "step": 416
+    },
+    {
+      "epoch": 1.11,
+      "learning_rate": 1.456412900978403e-05,
+      "loss": 0.6587,
+      "step": 417
+    },
+    {
+      "epoch": 1.11,
+      "learning_rate": 1.4538629116969828e-05,
+      "loss": 0.6417,
+      "step": 418
+    },
+    {
+      "epoch": 1.11,
+      "learning_rate": 1.4513092001197105e-05,
+      "loss": 0.658,
+      "step": 419
+    },
+    {
+      "epoch": 1.11,
+      "learning_rate": 1.4487517871905095e-05,
+      "loss": 0.6509,
+      "step": 420
+    },
+    {
+      "epoch": 1.12,
+      "learning_rate": 1.4461906938836585e-05,
+      "loss": 0.6364,
+      "step": 421
+    },
+    {
+      "epoch": 1.12,
+      "learning_rate": 1.4436259412036218e-05,
+      "loss": 0.6031,
+      "step": 422
+    },
+    {
+      "epoch": 1.12,
+      "learning_rate": 1.4410575501848743e-05,
+      "loss": 0.659,
+      "step": 423
+    },
+    {
+      "epoch": 1.12,
+      "learning_rate": 1.4384855418917312e-05,
+      "loss": 0.6377,
+      "step": 424
+    },
+    {
+      "epoch": 1.13,
+      "learning_rate": 1.4359099374181736e-05,
+      "loss": 0.638,
+      "step": 425
+    },
+    {
+      "epoch": 1.13,
+      "learning_rate": 1.4333307578876764e-05,
+      "loss": 0.668,
+      "step": 426
+    },
+    {
+      "epoch": 1.13,
+      "learning_rate": 1.4307480244530352e-05,
+      "loss": 0.6596,
+      "step": 427
+    },
+    {
+      "epoch": 1.14,
+      "learning_rate": 1.428161758296192e-05,
+      "loss": 0.6492,
+      "step": 428
+    },
+    {
+      "epoch": 1.14,
+      "learning_rate": 1.4255719806280624e-05,
+      "loss": 0.6881,
+      "step": 429
+    },
+    {
+      "epoch": 1.14,
+      "learning_rate": 1.4229787126883606e-05,
+      "loss": 0.6174,
+      "step": 430
+    },
+    {
+      "epoch": 1.14,
+      "learning_rate": 1.4203819757454258e-05,
+      "loss": 0.6543,
+      "step": 431
+    },
+    {
+      "epoch": 1.15,
+      "learning_rate": 1.4177817910960487e-05,
+      "loss": 0.6757,
+      "step": 432
+    },
+    {
+      "epoch": 1.15,
+      "learning_rate": 1.415178180065295e-05,
+      "loss": 0.6086,
+      "step": 433
+    },
+    {
+      "epoch": 1.15,
+      "learning_rate": 1.4125711640063319e-05,
+      "loss": 0.6928,
+      "step": 434
+    },
+    {
+      "epoch": 1.15,
+      "learning_rate": 1.4099607643002518e-05,
+      "loss": 0.6479,
+      "step": 435
+    },
+    {
+      "epoch": 1.16,
+      "learning_rate": 1.4073470023558984e-05,
+      "loss": 0.6625,
+      "step": 436
+    },
+    {
+      "epoch": 1.16,
+      "learning_rate": 1.40472989960969e-05,
+      "loss": 0.6469,
+      "step": 437
+    },
+    {
+      "epoch": 1.16,
+      "learning_rate": 1.402109477525444e-05,
+      "loss": 0.6373,
+      "step": 438
+    },
+    {
+      "epoch": 1.16,
+      "learning_rate": 1.3994857575942012e-05,
+      "loss": 0.7088,
+      "step": 439
+    },
+    {
+      "epoch": 1.17,
+      "learning_rate": 1.3968587613340489e-05,
+      "loss": 0.6787,
+      "step": 440
+    },
+    {
+      "epoch": 1.17,
+      "learning_rate": 1.394228510289945e-05,
+      "loss": 0.6482,
+      "step": 441
+    },
+    {
+      "epoch": 1.17,
+      "learning_rate": 1.3915950260335409e-05,
+      "loss": 0.6753,
+      "step": 442
+    },
+    {
+      "epoch": 1.18,
+      "learning_rate": 1.3889583301630047e-05,
+      "loss": 0.6356,
+      "step": 443
+    },
+    {
+      "epoch": 1.18,
+      "learning_rate": 1.3863184443028443e-05,
+      "loss": 0.6844,
+      "step": 444
+    },
+    {
+      "epoch": 1.18,
+      "learning_rate": 1.38367539010373e-05,
+      "loss": 0.6432,
+      "step": 445
+    },
+    {
+      "epoch": 1.18,
+      "learning_rate": 1.3810291892423167e-05,
+      "loss": 0.6151,
+      "step": 446
+    },
+    {
+      "epoch": 1.19,
+      "learning_rate": 1.3783798634210659e-05,
+      "loss": 0.7102,
+      "step": 447
+    },
+    {
+      "epoch": 1.19,
+      "learning_rate": 1.3757274343680686e-05,
+      "loss": 0.6978,
+      "step": 448
+    },
+    {
+      "epoch": 1.19,
+      "learning_rate": 1.3730719238368662e-05,
+      "loss": 0.6761,
+      "step": 449
+    },
+    {
+      "epoch": 1.19,
+      "learning_rate": 1.370413353606272e-05,
+      "loss": 0.6355,
+      "step": 450
+    },
+    {
+      "epoch": 1.2,
+      "learning_rate": 1.3677517454801941e-05,
+      "loss": 0.6479,
+      "step": 451
+    },
+    {
+      "epoch": 1.2,
+      "learning_rate": 1.3650871212874546e-05,
+      "loss": 0.6922,
+      "step": 452
+    },
+    {
+      "epoch": 1.2,
+      "learning_rate": 1.362419502881612e-05,
+      "loss": 0.6266,
+      "step": 453
+    },
+    {
+      "epoch": 1.2,
+      "learning_rate": 1.359748912140781e-05,
+      "loss": 0.6847,
+      "step": 454
+    },
+    {
+      "epoch": 1.21,
+      "learning_rate": 1.3570753709674536e-05,
+      "loss": 0.6572,
+      "step": 455
+    },
+    {
+      "epoch": 1.21,
+      "learning_rate": 1.3543989012883196e-05,
+      "loss": 0.6789,
+      "step": 456
+    },
+    {
+      "epoch": 1.21,
+      "learning_rate": 1.3517195250540865e-05,
+      "loss": 0.6352,
+      "step": 457
+    },
+    {
+      "epoch": 1.21,
+      "learning_rate": 1.3490372642392991e-05,
+      "loss": 0.6376,
+      "step": 458
+    },
+    {
+      "epoch": 1.22,
+      "learning_rate": 1.3463521408421604e-05,
+      "loss": 0.6104,
+      "step": 459
+    },
+    {
+      "epoch": 1.22,
+      "learning_rate": 1.3436641768843503e-05,
+      "loss": 0.6447,
+      "step": 460
+    },
+    {
+      "epoch": 1.22,
+      "learning_rate": 1.3409733944108447e-05,
+      "loss": 0.6805,
+      "step": 461
+    },
+    {
+      "epoch": 1.23,
+      "learning_rate": 1.3382798154897353e-05,
+      "loss": 0.6506,
+      "step": 462
+    },
+    {
+      "epoch": 1.23,
+      "learning_rate": 1.3355834622120494e-05,
+      "loss": 0.6824,
+      "step": 463
+    },
+    {
+      "epoch": 1.23,
+      "learning_rate": 1.3328843566915661e-05,
+      "loss": 0.6162,
+      "step": 464
+    },
+    {
+      "epoch": 1.23,
+      "learning_rate": 1.3301825210646386e-05,
+      "loss": 0.6689,
+      "step": 465
+    },
+    {
+      "epoch": 1.24,
+      "learning_rate": 1.327477977490009e-05,
+      "loss": 0.6327,
+      "step": 466
+    },
+    {
+      "epoch": 1.24,
+      "learning_rate": 1.3247707481486293e-05,
+      "loss": 0.6603,
+      "step": 467
+    },
+    {
+      "epoch": 1.24,
+      "learning_rate": 1.3220608552434783e-05,
+      "loss": 0.687,
+      "step": 468
+    },
+    {
+      "epoch": 1.24,
+      "learning_rate": 1.3193483209993796e-05,
+      "loss": 0.6887,
+      "step": 469
+    },
+    {
+      "epoch": 1.25,
+      "learning_rate": 1.316633167662819e-05,
+      "loss": 0.6534,
+      "step": 470
+    },
+    {
+      "epoch": 1.25,
+      "learning_rate": 1.3139154175017627e-05,
+      "loss": 0.6849,
+      "step": 471
+    },
+    {
+      "epoch": 1.25,
+      "learning_rate": 1.3111950928054751e-05,
+      "loss": 0.6634,
+      "step": 472
+    },
+    {
+      "epoch": 1.25,
+      "learning_rate": 1.3084722158843339e-05,
+      "loss": 0.6648,
+      "step": 473
+    },
+    {
+      "epoch": 1.26,
+      "learning_rate": 1.3057468090696497e-05,
+      "loss": 0.687,
+      "step": 474
+    },
+    {
+      "epoch": 1.26,
+      "learning_rate": 1.3030188947134816e-05,
+      "loss": 0.6756,
+      "step": 475
+    },
+    {
+      "epoch": 1.26,
+      "learning_rate": 1.3002884951884528e-05,
+      "loss": 0.6698,
+      "step": 476
+    },
+    {
+      "epoch": 1.27,
+      "learning_rate": 1.29755563288757e-05,
+      "loss": 0.6711,
+      "step": 477
+    },
+    {
+      "epoch": 1.27,
+      "learning_rate": 1.2948203302240367e-05,
+      "loss": 0.6858,
+      "step": 478
+    },
+    {
+      "epoch": 1.27,
+      "learning_rate": 1.292082609631071e-05,
+      "loss": 0.6616,
+      "step": 479
+    },
+    {
+      "epoch": 1.27,
+      "learning_rate": 1.2893424935617218e-05,
+      "loss": 0.6448,
+      "step": 480
+    },
+    {
+      "epoch": 1.28,
+      "learning_rate": 1.2866000044886837e-05,
+      "loss": 0.7047,
+      "step": 481
+    },
+    {
+      "epoch": 1.28,
+      "learning_rate": 1.2838551649041136e-05,
+      "loss": 0.665,
+      "step": 482
+    },
+    {
+      "epoch": 1.28,
+      "learning_rate": 1.2811079973194452e-05,
+      "loss": 0.6713,
+      "step": 483
+    },
+    {
+      "epoch": 1.28,
+      "learning_rate": 1.2783585242652054e-05,
+      "loss": 0.6626,
+      "step": 484
+    },
+    {
+      "epoch": 1.29,
+      "learning_rate": 1.2756067682908294e-05,
+      "loss": 0.6769,
+      "step": 485
+    },
+    {
+      "epoch": 1.29,
+      "learning_rate": 1.2728527519644744e-05,
+      "loss": 0.6276,
+      "step": 486
+    },
+    {
+      "epoch": 1.29,
+      "learning_rate": 1.2700964978728372e-05,
+      "loss": 0.6227,
+      "step": 487
+    },
+    {
+      "epoch": 1.29,
+      "learning_rate": 1.2673380286209656e-05,
+      "loss": 0.6644,
+      "step": 488
+    },
+    {
+      "epoch": 1.3,
+      "learning_rate": 1.2645773668320755e-05,
+      "loss": 0.6615,
+      "step": 489
+    },
+    {
+      "epoch": 1.3,
+      "learning_rate": 1.2618145351473653e-05,
+      "loss": 0.6886,
+      "step": 490
+    },
+    {
+      "epoch": 1.3,
+      "learning_rate": 1.2590495562258282e-05,
+      "loss": 0.6614,
+      "step": 491
+    },
+    {
+      "epoch": 1.31,
+      "learning_rate": 1.2562824527440687e-05,
+      "loss": 0.6511,
+      "step": 492
+    },
+    {
+      "epoch": 1.31,
+      "learning_rate": 1.2535132473961149e-05,
+      "loss": 0.6307,
+      "step": 493
+    },
+    {
+      "epoch": 1.31,
+      "learning_rate": 1.2507419628932336e-05,
+      "loss": 0.6482,
+      "step": 494
+    },
+    {
+      "epoch": 1.31,
+      "learning_rate": 1.2479686219637432e-05,
+      "loss": 0.6554,
+      "step": 495
+    },
+    {
+      "epoch": 1.32,
+      "learning_rate": 1.2451932473528278e-05,
+      "loss": 0.6598,
+      "step": 496
+    },
+    {
+      "epoch": 1.32,
+      "learning_rate": 1.2424158618223506e-05,
+      "loss": 0.6428,
+      "step": 497
+    },
+    {
+      "epoch": 1.32,
+      "learning_rate": 1.2396364881506662e-05,
+      "loss": 0.6541,
+      "step": 498
+    },
+    {
+      "epoch": 1.32,
+      "learning_rate": 1.236855149132436e-05,
+      "loss": 0.6663,
+      "step": 499
+    },
+    {
+      "epoch": 1.33,
+      "learning_rate": 1.234071867578439e-05,
+      "loss": 0.6118,
+      "step": 500
+    },
+    {
+      "epoch": 1.33,
+      "learning_rate": 1.2312866663153855e-05,
+      "loss": 0.6371,
+      "step": 501
+    },
+    {
+      "epoch": 1.33,
+      "learning_rate": 1.2284995681857313e-05,
+      "loss": 0.6445,
+      "step": 502
+    },
+    {
+      "epoch": 1.33,
+      "learning_rate": 1.2257105960474876e-05,
+      "loss": 0.6908,
+      "step": 503
+    },
+    {
+      "epoch": 1.34,
+      "learning_rate": 1.2229197727740356e-05,
+      "loss": 0.6942,
+      "step": 504
+    },
+    {
+      "epoch": 1.34,
+      "learning_rate": 1.2201271212539388e-05,
+      "loss": 0.6046,
+      "step": 505
+    },
+    {
+      "epoch": 1.34,
+      "learning_rate": 1.217332664390754e-05,
+      "loss": 0.6585,
+      "step": 506
+    },
+    {
+      "epoch": 1.34,
+      "learning_rate": 1.2145364251028452e-05,
+      "loss": 0.6169,
+      "step": 507
+    },
+    {
+      "epoch": 1.35,
+      "learning_rate": 1.2117384263231938e-05,
+      "loss": 0.6851,
+      "step": 508
+    },
+    {
+      "epoch": 1.35,
+      "learning_rate": 1.208938690999212e-05,
+      "loss": 0.6398,
+      "step": 509
+    },
+    {
+      "epoch": 1.35,
+      "learning_rate": 1.2061372420925542e-05,
+      "loss": 0.6403,
+      "step": 510
+    },
+    {
+      "epoch": 1.36,
+      "learning_rate": 1.2033341025789276e-05,
+      "loss": 0.6459,
+      "step": 511
+    },
+    {
+      "epoch": 1.36,
+      "learning_rate": 1.200529295447906e-05,
+      "loss": 0.6304,
+      "step": 512
+    },
+    {
+      "epoch": 1.36,
+      "learning_rate": 1.1977228437027385e-05,
+      "loss": 0.6736,
+      "step": 513
+    },
+    {
+      "epoch": 1.36,
+      "learning_rate": 1.1949147703601635e-05,
+      "loss": 0.6583,
+      "step": 514
+    },
+    {
+      "epoch": 1.37,
+      "learning_rate": 1.1921050984502185e-05,
+      "loss": 0.6608,
+      "step": 515
+    },
+    {
+      "epoch": 1.37,
+      "learning_rate": 1.1892938510160504e-05,
+      "loss": 0.6435,
+      "step": 516
+    },
+    {
+      "epoch": 1.37,
+      "learning_rate": 1.1864810511137288e-05,
+      "loss": 0.6265,
+      "step": 517
+    },
+    {
+      "epoch": 1.37,
+      "learning_rate": 1.1836667218120551e-05,
+      "loss": 0.6408,
+      "step": 518
+    },
+    {
+      "epoch": 1.38,
+      "learning_rate": 1.180850886192374e-05,
+      "loss": 0.67,
+      "step": 519
+    },
+    {
+      "epoch": 1.38,
+      "learning_rate": 1.1780335673483839e-05,
+      "loss": 0.6598,
+      "step": 520
+    },
+    {
+      "epoch": 1.38,
+      "learning_rate": 1.1752147883859477e-05,
+      "loss": 0.6738,
+      "step": 521
+    },
+    {
+      "epoch": 1.38,
+      "learning_rate": 1.1723945724229036e-05,
+      "loss": 0.6281,
+      "step": 522
+    },
+    {
+      "epoch": 1.39,
+      "learning_rate": 1.1695729425888747e-05,
+      "loss": 0.6585,
+      "step": 523
+    },
+    {
+      "epoch": 1.39,
+      "learning_rate": 1.1667499220250805e-05,
+      "loss": 0.6549,
+      "step": 524
+    },
+    {
+      "epoch": 1.39,
+      "learning_rate": 1.1639255338841449e-05,
+      "loss": 0.6305,
+      "step": 525
+    },
+    {
+      "epoch": 1.4,
+      "learning_rate": 1.1610998013299096e-05,
+      "loss": 0.6486,
+      "step": 526
+    },
+    {
+      "epoch": 1.4,
+      "learning_rate": 1.158272747537241e-05,
+      "loss": 0.6337,
+      "step": 527
+    },
+    {
+      "epoch": 1.4,
+      "learning_rate": 1.1554443956918421e-05,
+      "loss": 0.6767,
+      "step": 528
+    },
+    {
+      "epoch": 1.4,
+      "learning_rate": 1.1526147689900619e-05,
+      "loss": 0.6414,
+      "step": 529
+    },
+    {
+      "epoch": 1.41,
+      "learning_rate": 1.1497838906387042e-05,
+      "loss": 0.6719,
+      "step": 530
+    },
+    {
+      "epoch": 1.41,
+      "learning_rate": 1.1469517838548385e-05,
+      "loss": 0.6585,
+      "step": 531
+    },
+    {
+      "epoch": 1.41,
+      "learning_rate": 1.1441184718656093e-05,
+      "loss": 0.6627,
+      "step": 532
+    },
+    {
+      "epoch": 1.41,
+      "learning_rate": 1.1412839779080451e-05,
+      "loss": 0.6243,
+      "step": 533
+    },
+    {
+      "epoch": 1.42,
+      "learning_rate": 1.1384483252288681e-05,
+      "loss": 0.6989,
+      "step": 534
+    },
+    {
+      "epoch": 1.42,
+      "learning_rate": 1.135611537084304e-05,
+      "loss": 0.6839,
+      "step": 535
+    },
+    {
+      "epoch": 1.42,
+      "learning_rate": 1.1327736367398908e-05,
+      "loss": 0.6299,
+      "step": 536
+    },
+    {
+      "epoch": 1.42,
+      "learning_rate": 1.1299346474702872e-05,
+      "loss": 0.6789,
+      "step": 537
+    },
+    {
+      "epoch": 1.43,
+      "learning_rate": 1.1270945925590839e-05,
+      "loss": 0.6323,
+      "step": 538
+    },
+    {
+      "epoch": 1.43,
+      "learning_rate": 1.1242534952986106e-05,
+      "loss": 0.6194,
+      "step": 539
+    },
+    {
+      "epoch": 1.43,
+      "learning_rate": 1.1214113789897457e-05,
+      "loss": 0.6102,
+      "step": 540
+    },
+    {
+      "epoch": 1.44,
+      "learning_rate": 1.1185682669417248e-05,
+      "loss": 0.6378,
+      "step": 541
+    },
+    {
+      "epoch": 1.44,
+      "learning_rate": 1.1157241824719515e-05,
+      "loss": 0.6426,
+      "step": 542
+    },
+    {
+      "epoch": 1.44,
+      "learning_rate": 1.1128791489058024e-05,
+      "loss": 0.6515,
+      "step": 543
+    },
+    {
+      "epoch": 1.44,
+      "learning_rate": 1.1100331895764393e-05,
+      "loss": 0.6312,
+      "step": 544
+    },
+    {
+      "epoch": 1.45,
+      "learning_rate": 1.1071863278246166e-05,
+      "loss": 0.6821,
+      "step": 545
+    },
+    {
+      "epoch": 1.45,
+      "learning_rate": 1.1043385869984895e-05,
+      "loss": 0.66,
+      "step": 546
+    },
+    {
+      "epoch": 1.45,
+      "learning_rate": 1.1014899904534222e-05,
+      "loss": 0.6806,
+      "step": 547
+    },
+    {
+      "epoch": 1.45,
+      "learning_rate": 1.0986405615517983e-05,
+      "loss": 0.6402,
+      "step": 548
+    },
+    {
+      "epoch": 1.46,
+      "learning_rate": 1.0957903236628266e-05,
+      "loss": 0.6493,
+      "step": 549
+    },
+    {
+      "epoch": 1.46,
+      "learning_rate": 1.0929393001623513e-05,
+      "loss": 0.6807,
+      "step": 550
+    },
+    {
+      "epoch": 1.46,
+      "learning_rate": 1.0900875144326597e-05,
+      "loss": 0.6623,
+      "step": 551
+    },
+    {
+      "epoch": 1.46,
+      "learning_rate": 1.0872349898622903e-05,
+      "loss": 0.687,
+      "step": 552
+    },
+    {
+      "epoch": 1.47,
+      "learning_rate": 1.084381749845841e-05,
+      "loss": 0.6267,
+      "step": 553
+    },
+    {
+      "epoch": 1.47,
+      "learning_rate": 1.0815278177837775e-05,
+      "loss": 0.6981,
+      "step": 554
+    },
+    {
+      "epoch": 1.47,
+      "learning_rate": 1.0786732170822416e-05,
+      "loss": 0.6243,
+      "step": 555
+    },
+    {
+      "epoch": 1.47,
+      "learning_rate": 1.075817971152858e-05,
+      "loss": 0.67,
+      "step": 556
+    },
+    {
+      "epoch": 1.48,
+      "learning_rate": 1.0729621034125435e-05,
+      "loss": 0.6392,
+      "step": 557
+    },
+    {
+      "epoch": 1.48,
+      "learning_rate": 1.0701056372833147e-05,
+      "loss": 0.6487,
+      "step": 558
+    },
+    {
+      "epoch": 1.48,
+      "learning_rate": 1.0672485961920961e-05,
+      "loss": 0.6942,
+      "step": 559
+    },
+    {
+      "epoch": 1.49,
+      "learning_rate": 1.0643910035705273e-05,
+      "loss": 0.6314,
+      "step": 560
+    },
+    {
+      "epoch": 1.49,
+      "learning_rate": 1.061532882854771e-05,
+      "loss": 0.6025,
+      "step": 561
+    },
+    {
+      "epoch": 1.49,
+      "learning_rate": 1.0586742574853217e-05,
+      "loss": 0.6467,
+      "step": 562
+    },
+    {
+      "epoch": 1.49,
+      "learning_rate": 1.0558151509068117e-05,
+      "loss": 0.6816,
+      "step": 563
+    },
+    {
+      "epoch": 1.5,
+      "learning_rate": 1.052955586567821e-05,
+      "loss": 0.6724,
+      "step": 564
+    },
+    {
+      "epoch": 1.5,
+      "learning_rate": 1.0500955879206828e-05,
+      "loss": 0.6399,
+      "step": 565
+    },
+    {
+      "epoch": 1.5,
+      "learning_rate": 1.0472351784212936e-05,
+      "loss": 0.6272,
+      "step": 566
+    },
+    {
+      "epoch": 1.5,
+      "learning_rate": 1.0443743815289179e-05,
+      "loss": 0.6357,
+      "step": 567
+    },
+    {
+      "epoch": 1.51,
+      "learning_rate": 1.0415132207059981e-05,
+      "loss": 0.6322,
+      "step": 568
+    },
+    {
+      "epoch": 1.51,
+      "learning_rate": 1.038651719417962e-05,
+      "loss": 0.6553,
+      "step": 569
+    },
+    {
+      "epoch": 1.51,
+      "learning_rate": 1.035789901133028e-05,
+      "loss": 0.6599,
+      "step": 570
+    },
+    {
+      "epoch": 1.51,
+      "learning_rate": 1.032927789322016e-05,
+      "loss": 0.6379,
+      "step": 571
+    },
+    {
+      "epoch": 1.52,
+      "learning_rate": 1.0300654074581522e-05,
+      "loss": 0.6086,
+      "step": 572
+    },
+    {
+      "epoch": 1.52,
+      "learning_rate": 1.0272027790168778e-05,
+      "loss": 0.6625,
+      "step": 573
+    },
+    {
+      "epoch": 1.52,
+      "learning_rate": 1.0243399274756564e-05,
+      "loss": 0.631,
+      "step": 574
+    },
+    {
+      "epoch": 1.53,
+      "learning_rate": 1.0214768763137815e-05,
+      "loss": 0.6641,
+      "step": 575
+    },
+    {
+      "epoch": 1.53,
+      "learning_rate": 1.0186136490121834e-05,
+      "loss": 0.6348,
+      "step": 576
+    },
+    {
+      "epoch": 1.53,
+      "learning_rate": 1.0157502690532373e-05,
+      "loss": 0.5912,
+      "step": 577
+    },
+    {
+      "epoch": 1.53,
+      "learning_rate": 1.0128867599205698e-05,
+      "loss": 0.6712,
+      "step": 578
+    },
+    {
+      "epoch": 1.54,
+      "learning_rate": 1.010023145098868e-05,
+      "loss": 0.6819,
+      "step": 579
+    },
+    {
+      "epoch": 1.54,
+      "learning_rate": 1.0071594480736846e-05,
+      "loss": 0.644,
+      "step": 580
+    },
+    {
+      "epoch": 1.54,
+      "learning_rate": 1.004295692331247e-05,
+      "loss": 0.6353,
+      "step": 581
+    },
+    {
+      "epoch": 1.54,
+      "learning_rate": 1.001431901358265e-05,
+      "loss": 0.6387,
+      "step": 582
+    },
+    {
+      "epoch": 1.55,
+      "learning_rate": 9.985680986417354e-06,
+      "loss": 0.6804,
+      "step": 583
+    },
+    {
+      "epoch": 1.55,
+      "learning_rate": 9.95704307668753e-06,
+      "loss": 0.6212,
+      "step": 584
+    },
+    {
+      "epoch": 1.55,
+      "learning_rate": 9.928405519263159e-06,
+      "loss": 0.6933,
+      "step": 585
+    },
+    {
+      "epoch": 1.55,
+      "learning_rate": 9.899768549011322e-06,
+      "loss": 0.6669,
+      "step": 586
+    },
+    {
+      "epoch": 1.56,
+      "learning_rate": 9.871132400794305e-06,
+      "loss": 0.6201,
+      "step": 587
+    },
+    {
+      "epoch": 1.56,
+      "learning_rate": 9.842497309467629e-06,
+      "loss": 0.6769,
+      "step": 588
+    },
+    {
+      "epoch": 1.56,
+      "learning_rate": 9.813863509878169e-06,
+      "loss": 0.6457,
+      "step": 589
+    },
+    {
+      "epoch": 1.56,
+      "learning_rate": 9.78523123686219e-06,
+      "loss": 0.6428,
+      "step": 590
+    },
+    {
+      "epoch": 1.57,
+      "learning_rate": 9.756600725243439e-06,
+      "loss": 0.6278,
+      "step": 591
+    },
+    {
+      "epoch": 1.57,
+      "learning_rate": 9.727972209831227e-06,
+      "loss": 0.6851,
+      "step": 592
+    },
+    {
+      "epoch": 1.57,
+      "learning_rate": 9.699345925418481e-06,
+      "loss": 0.6483,
+      "step": 593
+    },
+    {
+      "epoch": 1.58,
+      "learning_rate": 9.670722106779844e-06,
+      "loss": 0.6277,
+      "step": 594
+    },
+    {
+      "epoch": 1.58,
+      "learning_rate": 9.642100988669723e-06,
+      "loss": 0.6239,
+      "step": 595
+    },
+    {
+      "epoch": 1.58,
+      "learning_rate": 9.613482805820386e-06,
+      "loss": 0.66,
+      "step": 596
+    },
+    {
+      "epoch": 1.58,
+      "learning_rate": 9.584867792940019e-06,
+      "loss": 0.6582,
+      "step": 597
+    },
+    {
+      "epoch": 1.59,
+      "learning_rate": 9.556256184710825e-06,
+      "loss": 0.6166,
+      "step": 598
+    },
+    {
+      "epoch": 1.59,
+      "learning_rate": 9.527648215787066e-06,
+      "loss": 0.6297,
+      "step": 599
+    },
+    {
+      "epoch": 1.59,
+      "learning_rate": 9.499044120793173e-06,
+      "loss": 0.6966,
+      "step": 600
+    },
+    {
+      "epoch": 1.59,
+      "learning_rate": 9.470444134321792e-06,
+      "loss": 0.6453,
+      "step": 601
+    },
+    {
+      "epoch": 1.6,
+      "learning_rate": 9.441848490931886e-06,
+      "loss": 0.65,
+      "step": 602
+    },
+    {
+      "epoch": 1.6,
+      "learning_rate": 9.41325742514679e-06,
+      "loss": 0.6187,
+      "step": 603
+    },
+    {
+      "epoch": 1.6,
+      "learning_rate": 9.384671171452292e-06,
+      "loss": 0.6999,
+      "step": 604
+    },
+    {
+      "epoch": 1.6,
+      "learning_rate": 9.35608996429473e-06,
+      "loss": 0.6098,
+      "step": 605
+    },
+    {
+      "epoch": 1.61,
+      "learning_rate": 9.32751403807904e-06,
+      "loss": 0.6554,
+      "step": 606
+    },
+    {
+      "epoch": 1.61,
+      "learning_rate": 9.298943627166854e-06,
+      "loss": 0.6599,
+      "step": 607
+    },
+    {
+      "epoch": 1.61,
+      "learning_rate": 9.270378965874568e-06,
+      "loss": 0.644,
+      "step": 608
+    },
+    {
+      "epoch": 1.62,
+      "learning_rate": 9.241820288471426e-06,
+      "loss": 0.6323,
+      "step": 609
+    },
+    {
+      "epoch": 1.62,
+      "learning_rate": 9.213267829177586e-06,
+      "loss": 0.6113,
+      "step": 610
+    },
+    {
+      "epoch": 1.62,
+      "learning_rate": 9.184721822162226e-06,
+      "loss": 0.6189,
+      "step": 611
+    },
+    {
+      "epoch": 1.62,
+      "learning_rate": 9.156182501541589e-06,
+      "loss": 0.6783,
+      "step": 612
+    },
+    {
+      "epoch": 1.63,
+      "learning_rate": 9.127650101377098e-06,
+      "loss": 0.6556,
+      "step": 613
+    },
+    {
+      "epoch": 1.63,
+      "learning_rate": 9.099124855673403e-06,
+      "loss": 0.6789,
+      "step": 614
+    },
+    {
+      "epoch": 1.63,
+      "learning_rate": 9.070606998376489e-06,
+      "loss": 0.6343,
+      "step": 615
+    },
+    {
+      "epoch": 1.63,
+      "learning_rate": 9.042096763371735e-06,
+      "loss": 0.6103,
+      "step": 616
+    },
+    {
+      "epoch": 1.64,
+      "learning_rate": 9.013594384482019e-06,
+      "loss": 0.6258,
+      "step": 617
+    },
+    {
+      "epoch": 1.64,
+      "learning_rate": 8.98510009546578e-06,
+      "loss": 0.6547,
+      "step": 618
+    },
+    {
+      "epoch": 1.64,
+      "learning_rate": 8.956614130015109e-06,
+      "loss": 0.6157,
+      "step": 619
+    },
+    {
+      "epoch": 1.64,
+      "learning_rate": 8.928136721753837e-06,
+      "loss": 0.6346,
+      "step": 620
+    },
+    {
+      "epoch": 1.65,
+      "learning_rate": 8.899668104235608e-06,
+      "loss": 0.6707,
+      "step": 621
+    },
+    {
+      "epoch": 1.65,
+      "learning_rate": 8.87120851094198e-06,
+      "loss": 0.6335,
+      "step": 622
+    },
+    {
+      "epoch": 1.65,
+      "learning_rate": 8.842758175280488e-06,
+      "loss": 0.6305,
+      "step": 623
+    },
+    {
+      "epoch": 1.66,
+      "learning_rate": 8.814317330582754e-06,
+      "loss": 0.6711,
+      "step": 624
+    },
+    {
+      "epoch": 1.66,
+      "learning_rate": 8.785886210102545e-06,
+      "loss": 0.6349,
+      "step": 625
+    },
+    {
+      "epoch": 1.66,
+      "learning_rate": 8.757465047013898e-06,
+      "loss": 0.6327,
+      "step": 626
+    },
+    {
+      "epoch": 1.66,
+      "learning_rate": 8.729054074409161e-06,
+      "loss": 0.6308,
+      "step": 627
+    },
+    {
+      "epoch": 1.67,
+      "learning_rate": 8.700653525297131e-06,
+      "loss": 0.6642,
+      "step": 628
+    },
+    {
+      "epoch": 1.67,
+      "learning_rate": 8.672263632601097e-06,
+      "loss": 0.6555,
+      "step": 629
+    },
+    {
+      "epoch": 1.67,
+      "learning_rate": 8.643884629156964e-06,
+      "loss": 0.6266,
+      "step": 630
+    },
+    {
+      "epoch": 1.67,
+      "learning_rate": 8.615516747711322e-06,
+      "loss": 0.6248,
+      "step": 631
+    },
+    {
+      "epoch": 1.68,
+      "learning_rate": 8.587160220919552e-06,
+      "loss": 0.6873,
+      "step": 632
+    },
+    {
+      "epoch": 1.68,
+      "learning_rate": 8.558815281343912e-06,
+      "loss": 0.6248,
+      "step": 633
+    },
+    {
+      "epoch": 1.68,
+      "learning_rate": 8.530482161451617e-06,
+      "loss": 0.7246,
+      "step": 634
+    },
+    {
+      "epoch": 1.68,
+      "learning_rate": 8.502161093612961e-06,
+      "loss": 0.6256,
+      "step": 635
+    },
+    {
+      "epoch": 1.69,
+      "learning_rate": 8.473852310099383e-06,
+      "loss": 0.6465,
+      "step": 636
+    },
+    {
+      "epoch": 1.69,
+      "learning_rate": 8.44555604308158e-06,
+      "loss": 0.627,
+      "step": 637
+    },
+    {
+      "epoch": 1.69,
+      "learning_rate": 8.417272524627591e-06,
+      "loss": 0.6535,
+      "step": 638
+    },
+    {
+      "epoch": 1.69,
+      "learning_rate": 8.389001986700908e-06,
+      "loss": 0.6926,
+      "step": 639
+    },
+    {
+      "epoch": 1.7,
+      "learning_rate": 8.360744661158553e-06,
+      "loss": 0.6355,
+      "step": 640
+    },
+    {
+      "epoch": 1.7,
+      "learning_rate": 8.332500779749199e-06,
+      "loss": 0.5965,
+      "step": 641
+    },
+    {
+      "epoch": 1.7,
+      "learning_rate": 8.304270574111253e-06,
+      "loss": 0.6441,
+      "step": 642
+    },
+    {
+      "epoch": 1.71,
+      "learning_rate": 8.276054275770967e-06,
+      "loss": 0.6334,
+      "step": 643
+    },
+    {
+      "epoch": 1.71,
+      "learning_rate": 8.247852116140528e-06,
+      "loss": 0.6298,
+      "step": 644
+    },
+    {
+      "epoch": 1.71,
+      "learning_rate": 8.219664326516165e-06,
+      "loss": 0.6447,
+      "step": 645
+    },
+    {
+      "epoch": 1.71,
+      "learning_rate": 8.191491138076267e-06,
+      "loss": 0.65,
+      "step": 646
+    },
+    {
+      "epoch": 1.72,
+      "learning_rate": 8.163332781879452e-06,
+      "loss": 0.6327,
+      "step": 647
+    },
+    {
+      "epoch": 1.72,
+      "learning_rate": 8.135189488862717e-06,
+      "loss": 0.6533,
+      "step": 648
+    },
+    {
+      "epoch": 1.72,
+      "learning_rate": 8.107061489839498e-06,
+      "loss": 0.61,
+      "step": 649
+    },
+    {
+      "epoch": 1.72,
+      "learning_rate": 8.07894901549782e-06,
+      "loss": 0.6716,
+      "step": 650
+    },
+    {
+      "epoch": 1.73,
+      "learning_rate": 8.050852296398365e-06,
+      "loss": 0.6501,
+      "step": 651
+    },
+    {
+      "epoch": 1.73,
+      "learning_rate": 8.022771562972616e-06,
+      "loss": 0.6773,
+      "step": 652
+    },
+    {
+      "epoch": 1.73,
+      "learning_rate": 7.994707045520944e-06,
+      "loss": 0.6228,
+      "step": 653
+    },
+    {
+      "epoch": 1.73,
+      "learning_rate": 7.966658974210727e-06,
+      "loss": 0.6381,
+      "step": 654
+    },
+    {
+      "epoch": 1.74,
+      "learning_rate": 7.93862757907446e-06,
+      "loss": 0.6644,
+      "step": 655
+    },
+    {
+      "epoch": 1.74,
+      "learning_rate": 7.910613090007883e-06,
+      "loss": 0.6403,
+      "step": 656
+    },
+    {
+      "epoch": 1.74,
+      "learning_rate": 7.882615736768067e-06,
+      "loss": 0.602,
+      "step": 657
+    },
+    {
+      "epoch": 1.75,
+      "learning_rate": 7.854635748971552e-06,
+      "loss": 0.6359,
+      "step": 658
+    },
+    {
+      "epoch": 1.75,
+      "learning_rate": 7.826673356092465e-06,
+      "loss": 0.6523,
+      "step": 659
+    },
+    {
+      "epoch": 1.75,
+      "learning_rate": 7.798728787460617e-06,
+      "loss": 0.6636,
+      "step": 660
+    },
+    {
+      "epoch": 1.75,
+      "learning_rate": 7.77080227225965e-06,
+      "loss": 0.6542,
+      "step": 661
+    },
+    {
+      "epoch": 1.76,
+      "learning_rate": 7.74289403952513e-06,
+      "loss": 0.667,
+      "step": 662
+    },
+    {
+      "epoch": 1.76,
+      "learning_rate": 7.71500431814269e-06,
+      "loss": 0.6263,
+      "step": 663
+    },
+    {
+      "epoch": 1.76,
+      "learning_rate": 7.687133336846146e-06,
+      "loss": 0.6228,
+      "step": 664
+    },
+    {
+      "epoch": 1.76,
+      "learning_rate": 7.659281324215614e-06,
+      "loss": 0.6408,
+      "step": 665
+    },
+    {
+      "epoch": 1.77,
+      "learning_rate": 7.631448508675642e-06,
+      "loss": 0.6744,
+      "step": 666
+    },
+    {
+      "epoch": 1.77,
+      "learning_rate": 7.60363511849334e-06,
+      "loss": 0.669,
+      "step": 667
+    },
+    {
+      "epoch": 1.77,
+      "learning_rate": 7.575841381776497e-06,
+      "loss": 0.6293,
+      "step": 668
+    },
+    {
+      "epoch": 1.77,
+      "learning_rate": 7.548067526471723e-06,
+      "loss": 0.6406,
+      "step": 669
+    },
+    {
+      "epoch": 1.78,
+      "learning_rate": 7.520313780362567e-06,
+      "loss": 0.6119,
+      "step": 670
+    },
+    {
+      "epoch": 1.78,
+      "learning_rate": 7.492580371067666e-06,
+      "loss": 0.6729,
+      "step": 671
+    },
+    {
+      "epoch": 1.78,
+      "learning_rate": 7.4648675260388556e-06,
+      "loss": 0.6689,
+      "step": 672
+    },
+    {
+      "epoch": 1.79,
+      "learning_rate": 7.437175472559316e-06,
+      "loss": 0.6498,
+      "step": 673
+    },
+    {
+      "epoch": 1.79,
+      "learning_rate": 7.409504437741722e-06,
+      "loss": 0.6471,
+      "step": 674
+    },
+    {
+      "epoch": 1.79,
+      "learning_rate": 7.38185464852635e-06,
+      "loss": 0.6771,
+      "step": 675
+    },
+    {
+      "epoch": 1.79,
+      "learning_rate": 7.354226331679247e-06,
+      "loss": 0.6467,
+      "step": 676
+    },
+    {
+      "epoch": 1.8,
+      "learning_rate": 7.326619713790348e-06,
+      "loss": 0.6699,
+      "step": 677
+    },
+    {
+      "epoch": 1.8,
+      "learning_rate": 7.299035021271631e-06,
+      "loss": 0.6521,
+      "step": 678
+    },
+    {
+      "epoch": 1.8,
+      "learning_rate": 7.271472480355255e-06,
+      "loss": 0.6536,
+      "step": 679
+    },
+    {
+      "epoch": 1.8,
+      "learning_rate": 7.2439323170917096e-06,
+      "loss": 0.6417,
+      "step": 680
+    },
+    {
+      "epoch": 1.81,
+      "learning_rate": 7.216414757347946e-06,
+      "loss": 0.6795,
+      "step": 681
+    },
+    {
+      "epoch": 1.81,
+      "learning_rate": 7.188920026805551e-06,
+      "loss": 0.6573,
+      "step": 682
+    },
+    {
+      "epoch": 1.81,
+      "learning_rate": 7.161448350958865e-06,
+      "loss": 0.7009,
+      "step": 683
+    },
+    {
+      "epoch": 1.81,
+      "learning_rate": 7.133999955113164e-06,
+      "loss": 0.6306,
+      "step": 684
+    },
+    {
+      "epoch": 1.82,
+      "learning_rate": 7.106575064382787e-06,
+      "loss": 0.6343,
+      "step": 685
+    },
+    {
+      "epoch": 1.82,
+      "learning_rate": 7.079173903689292e-06,
+      "loss": 0.6411,
+      "step": 686
+    },
+    {
+      "epoch": 1.82,
+      "learning_rate": 7.051796697759639e-06,
+      "loss": 0.6552,
+      "step": 687
+    },
+    {
+      "epoch": 1.82,
+      "learning_rate": 7.0244436711243035e-06,
+      "loss": 0.6417,
+      "step": 688
+    },
+    {
+      "epoch": 1.83,
+      "learning_rate": 6.997115048115473e-06,
+      "loss": 0.6287,
+      "step": 689
+    },
+    {
+      "epoch": 1.83,
+      "learning_rate": 6.969811052865188e-06,
+      "loss": 0.6667,
+      "step": 690
+    },
+    {
+      "epoch": 1.83,
+      "learning_rate": 6.942531909303505e-06,
+      "loss": 0.653,
+      "step": 691
+    },
+    {
+      "epoch": 1.84,
+      "learning_rate": 6.915277841156661e-06,
+      "loss": 0.6387,
+      "step": 692
+    },
+    {
+      "epoch": 1.84,
+      "learning_rate": 6.888049071945253e-06,
+      "loss": 0.6732,
+      "step": 693
+    },
+    {
+      "epoch": 1.84,
+      "learning_rate": 6.860845824982372e-06,
+      "loss": 0.6307,
+      "step": 694
+    },
+    {
+      "epoch": 1.84,
+      "learning_rate": 6.8336683233718134e-06,
+      "loss": 0.6654,
+      "step": 695
+    },
+    {
+      "epoch": 1.85,
+      "learning_rate": 6.806516790006205e-06,
+      "loss": 0.6537,
+      "step": 696
+    },
+    {
+      "epoch": 1.85,
+      "learning_rate": 6.779391447565218e-06,
+      "loss": 0.6582,
+      "step": 697
+    },
+    {
+      "epoch": 1.85,
+      "learning_rate": 6.75229251851371e-06,
+      "loss": 0.6546,
+      "step": 698
+    },
+    {
+      "epoch": 1.85,
+      "learning_rate": 6.725220225099911e-06,
+      "loss": 0.6062,
+      "step": 699
+    },
+    {
+      "epoch": 1.86,
+      "learning_rate": 6.698174789353619e-06,
+      "loss": 0.6556,
+      "step": 700
+    },
+    {
+      "epoch": 1.86,
+      "learning_rate": 6.67115643308434e-06,
+      "loss": 0.6273,
+      "step": 701
+    },
+    {
+      "epoch": 1.86,
+      "learning_rate": 6.64416537787951e-06,
+      "loss": 0.655,
+      "step": 702
+    },
+    {
+      "epoch": 1.86,
+      "learning_rate": 6.6172018451026474e-06,
+      "loss": 0.6526,
+      "step": 703
+    },
+    {
+      "epoch": 1.87,
+      "learning_rate": 6.590266055891557e-06,
+      "loss": 0.6514,
+      "step": 704
+    },
+    {
+      "epoch": 1.87,
+      "learning_rate": 6.563358231156498e-06,
+      "loss": 0.6301,
+      "step": 705
+    },
+    {
+      "epoch": 1.87,
+      "learning_rate": 6.536478591578396e-06,
+      "loss": 0.6398,
+      "step": 706
+    },
+    {
+      "epoch": 1.88,
+      "learning_rate": 6.509627357607009e-06,
+      "loss": 0.6658,
+      "step": 707
+    },
+    {
+      "epoch": 1.88,
+      "learning_rate": 6.4828047494591395e-06,
+      "loss": 0.6563,
+      "step": 708
+    },
+    {
+      "epoch": 1.88,
+      "learning_rate": 6.456010987116806e-06,
+      "loss": 0.6445,
+      "step": 709
+    },
+    {
+      "epoch": 1.88,
+      "learning_rate": 6.429246290325467e-06,
+      "loss": 0.6284,
+      "step": 710
+    },
+    {
+      "epoch": 1.89,
+      "learning_rate": 6.402510878592193e-06,
+      "loss": 0.6355,
+      "step": 711
+    },
+    {
+      "epoch": 1.89,
+      "learning_rate": 6.375804971183882e-06,
+      "loss": 0.6797,
+      "step": 712
+    },
+    {
+      "epoch": 1.89,
+      "learning_rate": 6.349128787125456e-06,
+      "loss": 0.6357,
+      "step": 713
+    },
+    {
+      "epoch": 1.89,
+      "learning_rate": 6.322482545198062e-06,
+      "loss": 0.6148,
+      "step": 714
+    },
+    {
+      "epoch": 1.9,
+      "learning_rate": 6.2958664639372845e-06,
+      "loss": 0.6451,
+      "step": 715
+    },
+    {
+      "epoch": 1.9,
+      "learning_rate": 6.269280761631342e-06,
+      "loss": 0.6193,
+      "step": 716
+    },
+    {
+      "epoch": 1.9,
+      "learning_rate": 6.242725656319318e-06,
+      "loss": 0.6644,
+      "step": 717
+    },
+    {
+      "epoch": 1.9,
+      "learning_rate": 6.216201365789343e-06,
+      "loss": 0.6698,
+      "step": 718
+    },
+    {
+      "epoch": 1.91,
+      "learning_rate": 6.189708107576837e-06,
+      "loss": 0.6373,
+      "step": 719
+    },
+    {
+      "epoch": 1.91,
+      "learning_rate": 6.163246098962701e-06,
+      "loss": 0.6291,
+      "step": 720
+    },
+    {
+      "epoch": 1.91,
+      "learning_rate": 6.136815556971561e-06,
+      "loss": 0.6324,
+      "step": 721
+    },
+    {
+      "epoch": 1.92,
+      "learning_rate": 6.1104166983699545e-06,
+      "loss": 0.6124,
+      "step": 722
+    },
+    {
+      "epoch": 1.92,
+      "learning_rate": 6.0840497396645946e-06,
+      "loss": 0.6365,
+      "step": 723
+    },
+    {
+      "epoch": 1.92,
+      "learning_rate": 6.057714897100552e-06,
+      "loss": 0.6629,
+      "step": 724
+    },
+    {
+      "epoch": 1.92,
+      "learning_rate": 6.0314123866595145e-06,
+      "loss": 0.6504,
+      "step": 725
+    },
+    {
+      "epoch": 1.93,
+      "learning_rate": 6.0051424240579905e-06,
+      "loss": 0.6436,
+      "step": 726
+    },
+    {
+      "epoch": 1.93,
+      "learning_rate": 5.978905224745562e-06,
+      "loss": 0.6466,
+      "step": 727
+    },
+    {
+      "epoch": 1.93,
+      "learning_rate": 5.952701003903105e-06,
+      "loss": 0.6677,
+      "step": 728
+    },
+    {
+      "epoch": 1.93,
+      "learning_rate": 5.926529976441018e-06,
+      "loss": 0.6675,
+      "step": 729
+    },
+    {
+      "epoch": 1.94,
+      "learning_rate": 5.900392356997486e-06,
+      "loss": 0.6696,
+      "step": 730
+    },
+    {
+      "epoch": 1.94,
+      "learning_rate": 5.874288359936684e-06,
+      "loss": 0.6732,
+      "step": 731
+    },
+    {
+      "epoch": 1.94,
+      "learning_rate": 5.848218199347053e-06,
+      "loss": 0.6235,
+      "step": 732
+    },
+    {
+      "epoch": 1.94,
+      "learning_rate": 5.822182089039515e-06,
+      "loss": 0.6794,
+      "step": 733
+    },
+    {
+      "epoch": 1.95,
+      "learning_rate": 5.796180242545746e-06,
+      "loss": 0.6367,
+      "step": 734
+    },
+    {
+      "epoch": 1.95,
+      "learning_rate": 5.7702128731163966e-06,
+      "loss": 0.6516,
+      "step": 735
+    },
+    {
+      "epoch": 1.95,
+      "learning_rate": 5.744280193719381e-06,
+      "loss": 0.6402,
+      "step": 736
+    },
+    {
+      "epoch": 1.95,
+      "learning_rate": 5.718382417038079e-06,
+      "loss": 0.6823,
+      "step": 737
+    },
+    {
+      "epoch": 1.96,
+      "learning_rate": 5.69251975546965e-06,
+      "loss": 0.6722,
+      "step": 738
+    },
+    {
+      "epoch": 1.96,
+      "learning_rate": 5.666692421123239e-06,
+      "loss": 0.6632,
+      "step": 739
+    },
+    {
+      "epoch": 1.96,
+      "learning_rate": 5.640900625818268e-06,
+      "loss": 0.6072,
+      "step": 740
+    },
+    {
+      "epoch": 1.97,
+      "learning_rate": 5.615144581082692e-06,
+      "loss": 0.6414,
+      "step": 741
+    },
+    {
+      "epoch": 1.97,
+      "learning_rate": 5.5894244981512605e-06,
+      "loss": 0.6377,
+      "step": 742
+    },
+    {
+      "epoch": 1.97,
+      "learning_rate": 5.563740587963786e-06,
+      "loss": 0.6824,
+      "step": 743
+    },
+    {
+      "epoch": 1.97,
+      "learning_rate": 5.538093061163418e-06,
+      "loss": 0.6538,
+      "step": 744
+    },
+    {
+      "epoch": 1.98,
+      "learning_rate": 5.512482128094912e-06,
+      "loss": 0.6449,
+      "step": 745
+    },
+    {
+      "epoch": 1.98,
+      "learning_rate": 5.486907998802895e-06,
+      "loss": 0.6514,
+      "step": 746
+    },
+    {
+      "epoch": 1.98,
+      "learning_rate": 5.461370883030176e-06,
+      "loss": 0.6233,
+      "step": 747
+    },
+    {
+      "epoch": 1.98,
+      "learning_rate": 5.435870990215971e-06,
+      "loss": 0.639,
+      "step": 748
+    },
+    {
+      "epoch": 1.99,
+      "learning_rate": 5.410408529494251e-06,
+      "loss": 0.649,
+      "step": 749
+    },
+    {
+      "epoch": 1.99,
+      "learning_rate": 5.38498370969196e-06,
+      "loss": 0.6413,
+      "step": 750
+    },
+    {
+      "epoch": 1.99,
+      "learning_rate": 5.359596739327366e-06,
+      "loss": 0.6525,
+      "step": 751
+    },
+    {
+      "epoch": 1.99,
+      "learning_rate": 5.334247826608292e-06,
+      "loss": 0.647,
+      "step": 752
+    },
+    {
+      "epoch": 2.0,
+      "learning_rate": 5.308937179430454e-06,
+      "loss": 0.663,
+      "step": 753
+    },
+    {
+      "epoch": 2.0,
+      "learning_rate": 5.2836650053757336e-06,
+      "loss": 0.6579,
+      "step": 754
+    },
+    {
+      "epoch": 2.0,
+      "learning_rate": 5.258431511710478e-06,
+      "loss": 0.6314,
+      "step": 755
+    },
+    {
+      "epoch": 2.01,
+      "learning_rate": 5.23323690538381e-06,
+      "loss": 0.6359,
+      "step": 756
+    },
+    {
+      "epoch": 2.01,
+      "learning_rate": 5.208081393025903e-06,
+      "loss": 0.6419,
+      "step": 757
+    },
+    {
+      "epoch": 2.01,
+      "learning_rate": 5.182965180946338e-06,
+      "loss": 0.5908,
+      "step": 758
+    },
+    {
+      "epoch": 2.01,
+      "learning_rate": 5.157888475132349e-06,
+      "loss": 0.6471,
+      "step": 759
+    },
+    {
+      "epoch": 2.02,
+      "learning_rate": 5.1328514812471906e-06,
+      "loss": 0.6058,
+      "step": 760
+    },
+    {
+      "epoch": 2.02,
+      "learning_rate": 5.107854404628399e-06,
+      "loss": 0.6251,
+      "step": 761
+    },
+    {
+      "epoch": 2.02,
+      "learning_rate": 5.082897450286166e-06,
+      "loss": 0.6603,
+      "step": 762
+    },
+    {
+      "epoch": 2.02,
+      "learning_rate": 5.057980822901589e-06,
+      "loss": 0.6328,
+      "step": 763
+    },
+    {
+      "epoch": 2.03,
+      "learning_rate": 5.033104726825067e-06,
+      "loss": 0.6866,
+      "step": 764
+    },
+    {
+      "epoch": 2.03,
+      "learning_rate": 5.008269366074556e-06,
+      "loss": 0.6268,
+      "step": 765
+    },
+    {
+      "epoch": 2.03,
+      "learning_rate": 4.983474944333945e-06,
+      "loss": 0.6303,
+      "step": 766
+    },
+    {
+      "epoch": 2.03,
+      "learning_rate": 4.958721664951358e-06,
+      "loss": 0.6656,
+      "step": 767
+    },
+    {
+      "epoch": 2.04,
+      "learning_rate": 4.9340097309375e-06,
+      "loss": 0.6439,
+      "step": 768
+    },
+    {
+      "epoch": 2.04,
+      "learning_rate": 4.909339344963992e-06,
+      "loss": 0.6388,
+      "step": 769
+    },
+    {
+      "epoch": 2.04,
+      "learning_rate": 4.884710709361685e-06,
+      "loss": 0.6664,
+      "step": 770
+    },
+    {
+      "epoch": 2.05,
+      "learning_rate": 4.860124026119052e-06,
+      "loss": 0.5804,
+      "step": 771
+    },
+    {
+      "epoch": 2.05,
+      "learning_rate": 4.835579496880461e-06,
+      "loss": 0.6678,
+      "step": 772
+    },
+    {
+      "epoch": 2.05,
+      "learning_rate": 4.811077322944599e-06,
+      "loss": 0.6365,
+      "step": 773
+    },
+    {
+      "epoch": 2.05,
+      "learning_rate": 4.7866177052627464e-06,
+      "loss": 0.6438,
+      "step": 774
+    },
+    {
+      "epoch": 2.06,
+      "learning_rate": 4.762200844437193e-06,
+      "loss": 0.6478,
+      "step": 775
+    },
+    {
+      "epoch": 2.06,
+      "learning_rate": 4.737826940719541e-06,
+      "loss": 0.6041,
+      "step": 776
+    },
+    {
+      "epoch": 2.06,
+      "learning_rate": 4.713496194009098e-06,
+      "loss": 0.6133,
+      "step": 777
+    },
+    {
+      "epoch": 2.06,
+      "learning_rate": 4.689208803851226e-06,
+      "loss": 0.6221,
+      "step": 778
+    },
+    {
+      "epoch": 2.07,
+      "learning_rate": 4.664964969435699e-06,
+      "loss": 0.6047,
+      "step": 779
+    },
+    {
+      "epoch": 2.07,
+      "learning_rate": 4.6407648895950785e-06,
+      "loss": 0.6943,
+      "step": 780
+    },
+    {
+      "epoch": 2.07,
+      "learning_rate": 4.616608762803068e-06,
+      "loss": 0.6084,
+      "step": 781
+    },
+    {
+      "epoch": 2.07,
+      "learning_rate": 4.5924967871729175e-06,
+      "loss": 0.6189,
+      "step": 782
+    },
+    {
+      "epoch": 2.08,
+      "learning_rate": 4.56842916045575e-06,
+      "loss": 0.6191,
+      "step": 783
+    },
+    {
+      "epoch": 2.08,
+      "learning_rate": 4.544406080038996e-06,
+      "loss": 0.61,
+      "step": 784
+    },
+    {
+      "epoch": 2.08,
+      "learning_rate": 4.520427742944715e-06,
+      "loss": 0.5954,
+      "step": 785
+    },
+    {
+      "epoch": 2.08,
+      "learning_rate": 4.496494345828039e-06,
+      "loss": 0.6407,
+      "step": 786
+    },
+    {
+      "epoch": 2.09,
+      "learning_rate": 4.472606084975507e-06,
+      "loss": 0.6177,
+      "step": 787
+    },
+    {
+      "epoch": 2.09,
+      "learning_rate": 4.4487631563035e-06,
+      "loss": 0.6091,
+      "step": 788
+    },
+    {
+      "epoch": 2.09,
+      "learning_rate": 4.424965755356592e-06,
+      "loss": 0.6233,
+      "step": 789
+    },
+    {
+      "epoch": 2.1,
+      "learning_rate": 4.401214077305983e-06,
+      "loss": 0.6093,
+      "step": 790
+    },
+    {
+      "epoch": 2.1,
+      "learning_rate": 4.377508316947877e-06,
+      "loss": 0.6205,
+      "step": 791
+    },
+    {
+      "epoch": 2.1,
+      "learning_rate": 4.353848668701891e-06,
+      "loss": 0.6328,
+      "step": 792
+    },
+    {
+      "epoch": 2.1,
+      "learning_rate": 4.330235326609464e-06,
+      "loss": 0.6443,
+      "step": 793
+    },
+    {
+      "epoch": 2.11,
+      "learning_rate": 4.3066684843322455e-06,
+      "loss": 0.6383,
+      "step": 794
+    },
+    {
+      "epoch": 2.11,
+      "learning_rate": 4.2831483351505485e-06,
+      "loss": 0.6539,
+      "step": 795
+    },
+    {
+      "epoch": 2.11,
+      "learning_rate": 4.259675071961712e-06,
+      "loss": 0.6114,
+      "step": 796
+    },
+    {
+      "epoch": 2.11,
+      "learning_rate": 4.2362488872785715e-06,
+      "loss": 0.6435,
+      "step": 797
+    },
+    {
+      "epoch": 2.12,
+      "learning_rate": 4.212869973227827e-06,
+      "loss": 0.6073,
+      "step": 798
+    },
+    {
+      "epoch": 2.12,
+      "learning_rate": 4.189538521548524e-06,
+      "loss": 0.6324,
+      "step": 799
+    },
+    {
+      "epoch": 2.12,
+      "learning_rate": 4.166254723590427e-06,
+      "loss": 0.6471,
+      "step": 800
+    },
+    {
+      "epoch": 2.12,
+      "learning_rate": 4.143018770312488e-06,
+      "loss": 0.6283,
+      "step": 801
+    },
+    {
+      "epoch": 2.13,
+      "learning_rate": 4.119830852281264e-06,
+      "loss": 0.6105,
+      "step": 802
+    },
+    {
+      "epoch": 2.13,
+      "learning_rate": 4.096691159669357e-06,
+      "loss": 0.6532,
+      "step": 803
+    },
+    {
+      "epoch": 2.13,
+      "learning_rate": 4.073599882253857e-06,
+      "loss": 0.646,
+      "step": 804
+    },
+    {
+      "epoch": 2.14,
+      "learning_rate": 4.050557209414779e-06,
+      "loss": 0.644,
+      "step": 805
+    },
+    {
+      "epoch": 2.14,
+      "learning_rate": 4.027563330133523e-06,
+      "loss": 0.622,
+      "step": 806
+    },
+    {
+      "epoch": 2.14,
+      "learning_rate": 4.004618432991295e-06,
+      "loss": 0.6616,
+      "step": 807
+    },
+    {
+      "epoch": 2.14,
+      "learning_rate": 3.981722706167611e-06,
+      "loss": 0.6405,
+      "step": 808
+    },
+    {
+      "epoch": 2.15,
+      "learning_rate": 3.9588763374386914e-06,
+      "loss": 0.6162,
+      "step": 809
+    },
+    {
+      "epoch": 2.15,
+      "learning_rate": 3.9360795141759824e-06,
+      "loss": 0.6803,
+      "step": 810
+    },
+    {
+      "epoch": 2.15,
+      "learning_rate": 3.913332423344559e-06,
+      "loss": 0.6905,
+      "step": 811
+    },
+    {
+      "epoch": 2.15,
+      "learning_rate": 3.890635251501656e-06,
+      "loss": 0.6199,
+      "step": 812
+    },
+    {
+      "epoch": 2.16,
+      "learning_rate": 3.867988184795073e-06,
+      "loss": 0.6569,
+      "step": 813
+    },
+    {
+      "epoch": 2.16,
+      "learning_rate": 3.8453914089617e-06,
+      "loss": 0.6086,
+      "step": 814
+    },
+    {
+      "epoch": 2.16,
+      "learning_rate": 3.822845109325962e-06,
+      "loss": 0.5954,
+      "step": 815
+    },
+    {
+      "epoch": 2.16,
+      "learning_rate": 3.8003494707983166e-06,
+      "loss": 0.6494,
+      "step": 816
+    },
+    {
+      "epoch": 2.17,
+      "learning_rate": 3.7779046778737282e-06,
+      "loss": 0.6442,
+      "step": 817
+    },
+    {
+      "epoch": 2.17,
+      "learning_rate": 3.7555109146301593e-06,
+      "loss": 0.6316,
+      "step": 818
+    },
+    {
+      "epoch": 2.17,
+      "learning_rate": 3.7331683647270568e-06,
+      "loss": 0.665,
+      "step": 819
+    },
+    {
+      "epoch": 2.18,
+      "learning_rate": 3.7108772114038504e-06,
+      "loss": 0.6193,
+      "step": 820
+    },
+    {
+      "epoch": 2.18,
+      "learning_rate": 3.6886376374784514e-06,
+      "loss": 0.623,
+      "step": 821
+    },
+    {
+      "epoch": 2.18,
+      "learning_rate": 3.666449825345734e-06,
+      "loss": 0.6436,
+      "step": 822
+    },
+    {
+      "epoch": 2.18,
+      "learning_rate": 3.64431395697608e-06,
+      "loss": 0.6068,
+      "step": 823
+    },
+    {
+      "epoch": 2.19,
+      "learning_rate": 3.6222302139138364e-06,
+      "loss": 0.6384,
+      "step": 824
+    },
+    {
+      "epoch": 2.19,
+      "learning_rate": 3.6001987772758684e-06,
+      "loss": 0.5924,
+      "step": 825
+    },
+    {
+      "epoch": 2.19,
+      "learning_rate": 3.57821982775005e-06,
+      "loss": 0.6204,
+      "step": 826
+    },
+    {
+      "epoch": 2.19,
+      "learning_rate": 3.5562935455937908e-06,
+      "loss": 0.6103,
+      "step": 827
+    },
+    {
+      "epoch": 2.2,
+      "learning_rate": 3.5344201106325548e-06,
+      "loss": 0.6182,
+      "step": 828
+    },
+    {
+      "epoch": 2.2,
+      "learning_rate": 3.512599702258389e-06,
+      "loss": 0.6363,
+      "step": 829
+    },
+    {
+      "epoch": 2.2,
+      "learning_rate": 3.4908324994284483e-06,
+      "loss": 0.6123,
+      "step": 830
+    },
+    {
+      "epoch": 2.2,
+      "learning_rate": 3.46911868066353e-06,
+      "loss": 0.6288,
+      "step": 831
+    },
+    {
+      "epoch": 2.21,
+      "learning_rate": 3.4474584240466092e-06,
+      "loss": 0.6367,
+      "step": 832
+    },
+    {
+      "epoch": 2.21,
+      "learning_rate": 3.425851907221377e-06,
+      "loss": 0.6059,
+      "step": 833
+    },
+    {
+      "epoch": 2.21,
+      "learning_rate": 3.4042993073907904e-06,
+      "loss": 0.6327,
+      "step": 834
+    },
+    {
+      "epoch": 2.21,
+      "learning_rate": 3.3828008013155976e-06,
+      "loss": 0.6426,
+      "step": 835
+    },
+    {
+      "epoch": 2.22,
+      "learning_rate": 3.361356565312931e-06,
+      "loss": 0.62,
+      "step": 836
+    },
+    {
+      "epoch": 2.22,
+      "learning_rate": 3.339966775254806e-06,
+      "loss": 0.6661,
+      "step": 837
+    },
+    {
+      "epoch": 2.22,
+      "learning_rate": 3.3186316065667247e-06,
+      "loss": 0.6796,
+      "step": 838
+    },
+    {
+      "epoch": 2.23,
+      "learning_rate": 3.2973512342262127e-06,
+      "loss": 0.632,
+      "step": 839
+    },
+    {
+      "epoch": 2.23,
+      "learning_rate": 3.2761258327613945e-06,
+      "loss": 0.6025,
+      "step": 840
+    },
+    {
+      "epoch": 2.23,
+      "learning_rate": 3.254955576249552e-06,
+      "loss": 0.6207,
+      "step": 841
+    },
+    {
+      "epoch": 2.23,
+      "learning_rate": 3.2338406383157117e-06,
+      "loss": 0.6204,
+      "step": 842
+    },
+    {
+      "epoch": 2.24,
+      "learning_rate": 3.2127811921312045e-06,
+      "loss": 0.6627,
+      "step": 843
+    },
+    {
+      "epoch": 2.24,
+      "learning_rate": 3.191777410412257e-06,
+      "loss": 0.6081,
+      "step": 844
+    },
+    {
+      "epoch": 2.24,
+      "learning_rate": 3.1708294654185713e-06,
+      "loss": 0.6096,
+      "step": 845
+    },
+    {
+      "epoch": 2.24,
+      "learning_rate": 3.1499375289519107e-06,
+      "loss": 0.6143,
+      "step": 846
+    },
+    {
+      "epoch": 2.25,
+      "learning_rate": 3.129101772354696e-06,
+      "loss": 0.6,
+      "step": 847
+    },
+    {
+      "epoch": 2.25,
+      "learning_rate": 3.108322366508585e-06,
+      "loss": 0.6812,
+      "step": 848
+    },
+    {
+      "epoch": 2.25,
+      "learning_rate": 3.0875994818330955e-06,
+      "loss": 0.6438,
+      "step": 849
+    },
+    {
+      "epoch": 2.25,
+      "learning_rate": 3.0669332882841896e-06,
+      "loss": 0.6254,
+      "step": 850
+    },
+    {
+      "epoch": 2.26,
+      "learning_rate": 3.046323955352882e-06,
+      "loss": 0.6281,
+      "step": 851
+    },
+    {
+      "epoch": 2.26,
+      "learning_rate": 3.025771652063857e-06,
+      "loss": 0.6122,
+      "step": 852
+    },
+    {
+      "epoch": 2.26,
+      "learning_rate": 3.0052765469740753e-06,
+      "loss": 0.6489,
+      "step": 853
+    },
+    {
+      "epoch": 2.27,
+      "learning_rate": 2.9848388081713954e-06,
+      "loss": 0.6354,
+      "step": 854
+    },
+    {
+      "epoch": 2.27,
+      "learning_rate": 2.9644586032731927e-06,
+      "loss": 0.6721,
+      "step": 855
+    },
+    {
+      "epoch": 2.27,
+      "learning_rate": 2.9441360994249866e-06,
+      "loss": 0.6176,
+      "step": 856
+    },
+    {
+      "epoch": 2.27,
+      "learning_rate": 2.9238714632990704e-06,
+      "loss": 0.65,
+      "step": 857
+    },
+    {
+      "epoch": 2.28,
+      "learning_rate": 2.9036648610931407e-06,
+      "loss": 0.6269,
+      "step": 858
+    },
+    {
+      "epoch": 2.28,
+      "learning_rate": 2.883516458528939e-06,
+      "loss": 0.6175,
+      "step": 859
+    },
+    {
+      "epoch": 2.28,
+      "learning_rate": 2.863426420850888e-06,
+      "loss": 0.6327,
+      "step": 860
+    },
+    {
+      "epoch": 2.28,
+      "learning_rate": 2.843394912824742e-06,
+      "loss": 0.6557,
+      "step": 861
+    },
+    {
+      "epoch": 2.29,
+      "learning_rate": 2.8234220987362248e-06,
+      "loss": 0.6727,
+      "step": 862
+    },
+    {
+      "epoch": 2.29,
+      "learning_rate": 2.8035081423896983e-06,
+      "loss": 0.661,
+      "step": 863
+    },
+    {
+      "epoch": 2.29,
+      "learning_rate": 2.7836532071068057e-06,
+      "loss": 0.627,
+      "step": 864
+    },
+    {
+      "epoch": 2.29,
+      "learning_rate": 2.7638574557251397e-06,
+      "loss": 0.6311,
+      "step": 865
+    },
+    {
+      "epoch": 2.3,
+      "learning_rate": 2.7441210505969008e-06,
+      "loss": 0.6369,
+      "step": 866
+    },
+    {
+      "epoch": 2.3,
+      "learning_rate": 2.724444153587572e-06,
+      "loss": 0.6404,
+      "step": 867
+    },
+    {
+      "epoch": 2.3,
+      "learning_rate": 2.7048269260745876e-06,
+      "loss": 0.6431,
+      "step": 868
+    },
+    {
+      "epoch": 2.31,
+      "learning_rate": 2.6852695289460095e-06,
+      "loss": 0.6613,
+      "step": 869
+    },
+    {
+      "epoch": 2.31,
+      "learning_rate": 2.6657721225992104e-06,
+      "loss": 0.6926,
+      "step": 870
+    },
+    {
+      "epoch": 2.31,
+      "learning_rate": 2.646334866939555e-06,
+      "loss": 0.6489,
+      "step": 871
+    },
+    {
+      "epoch": 2.31,
+      "learning_rate": 2.6269579213790952e-06,
+      "loss": 0.6021,
+      "step": 872
+    },
+    {
+      "epoch": 2.32,
+      "learning_rate": 2.607641444835243e-06,
+      "loss": 0.6212,
+      "step": 873
+    },
+    {
+      "epoch": 2.32,
+      "learning_rate": 2.5883855957295054e-06,
+      "loss": 0.6036,
+      "step": 874
+    },
+    {
+      "epoch": 2.32,
+      "learning_rate": 2.5691905319861386e-06,
+      "loss": 0.6812,
+      "step": 875
+    },
+    {
+      "epoch": 2.32,
+      "learning_rate": 2.55005641103089e-06,
+      "loss": 0.6218,
+      "step": 876
+    },
+    {
+      "epoch": 2.33,
+      "learning_rate": 2.5309833897896873e-06,
+      "loss": 0.62,
+      "step": 877
+    },
+    {
+      "epoch": 2.33,
+      "learning_rate": 2.5119716246873603e-06,
+      "loss": 0.6357,
+      "step": 878
+    },
+    {
+      "epoch": 2.33,
+      "learning_rate": 2.493021271646351e-06,
+      "loss": 0.6299,
+      "step": 879
+    },
+    {
+      "epoch": 2.33,
+      "learning_rate": 2.47413248608544e-06,
+      "loss": 0.6823,
+      "step": 880
+    },
+    {
+      "epoch": 2.34,
+      "learning_rate": 2.455305422918474e-06,
+      "loss": 0.6118,
+      "step": 881
+    },
+    {
+      "epoch": 2.34,
+      "learning_rate": 2.4365402365530856e-06,
+      "loss": 0.656,
+      "step": 882
+    },
+    {
+      "epoch": 2.34,
+      "learning_rate": 2.417837080889438e-06,
+      "loss": 0.5984,
+      "step": 883
+    },
+    {
+      "epoch": 2.34,
+      "learning_rate": 2.3991961093189553e-06,
+      "loss": 0.6398,
+      "step": 884
+    },
+    {
+      "epoch": 2.35,
+      "learning_rate": 2.3806174747230703e-06,
+      "loss": 0.6432,
+      "step": 885
+    },
+    {
+      "epoch": 2.35,
+      "learning_rate": 2.3621013294719586e-06,
+      "loss": 0.659,
+      "step": 886
+    },
+    {
+      "epoch": 2.35,
+      "learning_rate": 2.343647825423314e-06,
+      "loss": 0.6049,
+      "step": 887
+    },
+    {
+      "epoch": 2.36,
+      "learning_rate": 2.325257113921069e-06,
+      "loss": 0.6447,
+      "step": 888
+    },
+    {
+      "epoch": 2.36,
+      "learning_rate": 2.3069293457941823e-06,
+      "loss": 0.6505,
+      "step": 889
+    },
+    {
+      "epoch": 2.36,
+      "learning_rate": 2.2886646713553884e-06,
+      "loss": 0.6374,
+      "step": 890
+    },
+    {
+      "epoch": 2.36,
+      "learning_rate": 2.2704632403999683e-06,
+      "loss": 0.6297,
+      "step": 891
+    },
+    {
+      "epoch": 2.37,
+      "learning_rate": 2.252325202204517e-06,
+      "loss": 0.6413,
+      "step": 892
+    },
+    {
+      "epoch": 2.37,
+      "learning_rate": 2.2342507055257256e-06,
+      "loss": 0.6523,
+      "step": 893
+    },
+    {
+      "epoch": 2.37,
+      "learning_rate": 2.216239898599156e-06,
+      "loss": 0.662,
+      "step": 894
+    },
+    {
+      "epoch": 2.37,
+      "learning_rate": 2.1982929291380285e-06,
+      "loss": 0.6533,
+      "step": 895
+    },
+    {
+      "epoch": 2.38,
+      "learning_rate": 2.1804099443320094e-06,
+      "loss": 0.6684,
+      "step": 896
+    },
+    {
+      "epoch": 2.38,
+      "learning_rate": 2.1625910908459957e-06,
+      "loss": 0.6363,
+      "step": 897
+    },
+    {
+      "epoch": 2.38,
+      "learning_rate": 2.1448365148189363e-06,
+      "loss": 0.6659,
+      "step": 898
+    },
+    {
+      "epoch": 2.38,
+      "learning_rate": 2.1271463618625987e-06,
+      "loss": 0.6631,
+      "step": 899
+    },
+    {
+      "epoch": 2.39,
+      "learning_rate": 2.1095207770604097e-06,
+      "loss": 0.6504,
+      "step": 900
+    },
+    {
+      "epoch": 2.39,
+      "learning_rate": 2.091959904966234e-06,
+      "loss": 0.6113,
+      "step": 901
+    },
+    {
+      "epoch": 2.39,
+      "learning_rate": 2.0744638896032145e-06,
+      "loss": 0.6542,
+      "step": 902
+    },
+    {
+      "epoch": 2.4,
+      "learning_rate": 2.0570328744625765e-06,
+      "loss": 0.63,
+      "step": 903
+    },
+    {
+      "epoch": 2.4,
+      "learning_rate": 2.0396670025024545e-06,
+      "loss": 0.622,
+      "step": 904
+    },
+    {
+      "epoch": 2.4,
+      "learning_rate": 2.0223664161467206e-06,
+      "loss": 0.5942,
+      "step": 905
+    },
+    {
+      "epoch": 2.4,
+      "learning_rate": 2.0051312572838156e-06,
+      "loss": 0.6437,
+      "step": 906
+    },
+    {
+      "epoch": 2.41,
+      "learning_rate": 1.9879616672655845e-06,
+      "loss": 0.6571,
+      "step": 907
+    },
+    {
+      "epoch": 2.41,
+      "learning_rate": 1.9708577869061207e-06,
+      "loss": 0.6136,
+      "step": 908
+    },
+    {
+      "epoch": 2.41,
+      "learning_rate": 1.953819756480608e-06,
+      "loss": 0.6643,
+      "step": 909
+    },
+    {
+      "epoch": 2.41,
+      "learning_rate": 1.9368477157241626e-06,
+      "loss": 0.6627,
+      "step": 910
+    },
+    {
+      "epoch": 2.42,
+      "learning_rate": 1.9199418038307137e-06,
+      "loss": 0.6088,
+      "step": 911
+    },
+    {
+      "epoch": 2.42,
+      "learning_rate": 1.9031021594518206e-06,
+      "loss": 0.6456,
+      "step": 912
+    },
+    {
+      "epoch": 2.42,
+      "learning_rate": 1.8863289206955815e-06,
+      "loss": 0.6353,
+      "step": 913
+    },
+    {
+      "epoch": 2.42,
+      "learning_rate": 1.8696222251254559e-06,
+      "loss": 0.6524,
+      "step": 914
+    },
+    {
+      "epoch": 2.43,
+      "learning_rate": 1.8529822097591787e-06,
+      "loss": 0.6283,
+      "step": 915
+    },
+    {
+      "epoch": 2.43,
+      "learning_rate": 1.8364090110675992e-06,
+      "loss": 0.6337,
+      "step": 916
+    },
+    {
+      "epoch": 2.43,
+      "learning_rate": 1.819902764973588e-06,
+      "loss": 0.6333,
+      "step": 917
+    },
+    {
+      "epoch": 2.44,
+      "learning_rate": 1.8034636068509093e-06,
+      "loss": 0.6596,
+      "step": 918
+    },
+    {
+      "epoch": 2.44,
+      "learning_rate": 1.7870916715231167e-06,
+      "loss": 0.6217,
+      "step": 919
+    },
+    {
+      "epoch": 2.44,
+      "learning_rate": 1.7707870932624448e-06,
+      "loss": 0.639,
+      "step": 920
+    },
+    {
+      "epoch": 2.44,
+      "learning_rate": 1.7545500057887055e-06,
+      "loss": 0.6471,
+      "step": 921
+    },
+    {
+      "epoch": 2.45,
+      "learning_rate": 1.7383805422682e-06,
+      "loss": 0.6436,
+      "step": 922
+    },
+    {
+      "epoch": 2.45,
+      "learning_rate": 1.7222788353126075e-06,
+      "loss": 0.6313,
+      "step": 923
+    },
+    {
+      "epoch": 2.45,
+      "learning_rate": 1.706245016977931e-06,
+      "loss": 0.6472,
+      "step": 924
+    },
+    {
+      "epoch": 2.45,
+      "learning_rate": 1.6902792187633744e-06,
+      "loss": 0.6776,
+      "step": 925
+    },
+    {
+      "epoch": 2.46,
+      "learning_rate": 1.6743815716102995e-06,
+      "loss": 0.5891,
+      "step": 926
+    },
+    {
+      "epoch": 2.46,
+      "learning_rate": 1.6585522059011205e-06,
+      "loss": 0.6838,
+      "step": 927
+    },
+    {
+      "epoch": 2.46,
+      "learning_rate": 1.64279125145827e-06,
+      "loss": 0.6495,
+      "step": 928
+    },
+    {
+      "epoch": 2.46,
+      "learning_rate": 1.6270988375430953e-06,
+      "loss": 0.7168,
+      "step": 929
+    },
+    {
+      "epoch": 2.47,
+      "learning_rate": 1.6114750928548284e-06,
+      "loss": 0.6072,
+      "step": 930
+    },
+    {
+      "epoch": 2.47,
+      "learning_rate": 1.5959201455295192e-06,
+      "loss": 0.6306,
+      "step": 931
+    },
+    {
+      "epoch": 2.47,
+      "learning_rate": 1.580434123138983e-06,
+      "loss": 0.6326,
+      "step": 932
+    },
+    {
+      "epoch": 2.47,
+      "learning_rate": 1.5650171526897595e-06,
+      "loss": 0.6284,
+      "step": 933
+    },
+    {
+      "epoch": 2.48,
+      "learning_rate": 1.5496693606220593e-06,
+      "loss": 0.6189,
+      "step": 934
+    },
+    {
+      "epoch": 2.48,
+      "learning_rate": 1.5343908728087509e-06,
+      "loss": 0.6466,
+      "step": 935
+    },
+    {
+      "epoch": 2.48,
+      "learning_rate": 1.519181814554297e-06,
+      "loss": 0.6574,
+      "step": 936
+    },
+    {
+      "epoch": 2.49,
+      "learning_rate": 1.5040423105937585e-06,
+      "loss": 0.6423,
+      "step": 937
+    },
+    {
+      "epoch": 2.49,
+      "learning_rate": 1.4889724850917397e-06,
+      "loss": 0.6484,
+      "step": 938
+    },
+    {
+      "epoch": 2.49,
+      "learning_rate": 1.4739724616414053e-06,
+      "loss": 0.655,
+      "step": 939
+    },
+    {
+      "epoch": 2.49,
+      "learning_rate": 1.4590423632634298e-06,
+      "loss": 0.6639,
+      "step": 940
+    },
+    {
+      "epoch": 2.5,
+      "learning_rate": 1.444182312405017e-06,
+      "loss": 0.6374,
+      "step": 941
+    },
+    {
+      "epoch": 2.5,
+      "learning_rate": 1.429392430938882e-06,
+      "loss": 0.6198,
+      "step": 942
+    },
+    {
+      "epoch": 2.5,
+      "learning_rate": 1.4146728401622567e-06,
+      "loss": 0.665,
+      "step": 943
+    },
+    {
+      "epoch": 2.5,
+      "learning_rate": 1.4000236607958927e-06,
+      "loss": 0.6347,
+      "step": 944
+    },
+    {
+      "epoch": 2.51,
+      "learning_rate": 1.38544501298307e-06,
+      "loss": 0.6093,
+      "step": 945
+    },
+    {
+      "epoch": 2.51,
+      "learning_rate": 1.3709370162886194e-06,
+      "loss": 0.6172,
+      "step": 946
+    },
+    {
+      "epoch": 2.51,
+      "learning_rate": 1.3564997896979226e-06,
+      "loss": 0.6238,
+      "step": 947
+    },
+    {
+      "epoch": 2.51,
+      "learning_rate": 1.3421334516159701e-06,
+      "loss": 0.6314,
+      "step": 948
+    },
+    {
+      "epoch": 2.52,
+      "learning_rate": 1.3278381198663492e-06,
+      "loss": 0.6377,
+      "step": 949
+    },
+    {
+      "epoch": 2.52,
+      "learning_rate": 1.3136139116903147e-06,
+      "loss": 0.6319,
+      "step": 950
+    },
+    {
+      "epoch": 2.52,
+      "learning_rate": 1.2994609437457972e-06,
+      "loss": 0.6692,
+      "step": 951
+    },
+    {
+      "epoch": 2.53,
+      "learning_rate": 1.2853793321064746e-06,
+      "loss": 0.6452,
+      "step": 952
+    },
+    {
+      "epoch": 2.53,
+      "learning_rate": 1.2713691922607907e-06,
+      "loss": 0.6195,
+      "step": 953
+    },
+    {
+      "epoch": 2.53,
+      "learning_rate": 1.2574306391110347e-06,
+      "loss": 0.64,
+      "step": 954
+    },
+    {
+      "epoch": 2.53,
+      "learning_rate": 1.2435637869723794e-06,
+      "loss": 0.6359,
+      "step": 955
+    },
+    {
+      "epoch": 2.54,
+      "learning_rate": 1.2297687495719557e-06,
+      "loss": 0.5882,
+      "step": 956
+    },
+    {
+      "epoch": 2.54,
+      "learning_rate": 1.2160456400479171e-06,
+      "loss": 0.6237,
+      "step": 957
+    },
+    {
+      "epoch": 2.54,
+      "learning_rate": 1.2023945709485008e-06,
+      "loss": 0.6224,
+      "step": 958
+    },
+    {
+      "epoch": 2.54,
+      "learning_rate": 1.188815654231128e-06,
+      "loss": 0.673,
+      "step": 959
+    },
+    {
+      "epoch": 2.55,
+      "learning_rate": 1.1753090012614576e-06,
+      "loss": 0.6566,
+      "step": 960
+    },
+    {
+      "epoch": 2.55,
+      "learning_rate": 1.1618747228125016e-06,
+      "loss": 0.656,
+      "step": 961
+    },
+    {
+      "epoch": 2.55,
+      "learning_rate": 1.1485129290636866e-06,
+      "loss": 0.611,
+      "step": 962
+    },
+    {
+      "epoch": 2.55,
+      "learning_rate": 1.1352237295999824e-06,
+      "loss": 0.6599,
+      "step": 963
+    },
+    {
+      "epoch": 2.56,
+      "learning_rate": 1.1220072334109678e-06,
+      "loss": 0.6309,
+      "step": 964
+    },
+    {
+      "epoch": 2.56,
+      "learning_rate": 1.1088635488899746e-06,
+      "loss": 0.6327,
+      "step": 965
+    },
+    {
+      "epoch": 2.56,
+      "learning_rate": 1.095792783833164e-06,
+      "loss": 0.6205,
+      "step": 966
+    },
+    {
+      "epoch": 2.56,
+      "learning_rate": 1.082795045438666e-06,
+      "loss": 0.6593,
+      "step": 967
+    },
+    {
+      "epoch": 2.57,
+      "learning_rate": 1.0698704403056892e-06,
+      "loss": 0.6137,
+      "step": 968
+    },
+    {
+      "epoch": 2.57,
+      "learning_rate": 1.0570190744336529e-06,
+      "loss": 0.6297,
+      "step": 969
+    },
+    {
+      "epoch": 2.57,
+      "learning_rate": 1.0442410532213132e-06,
+      "loss": 0.6339,
+      "step": 970
+    },
+    {
+      "epoch": 2.58,
+      "learning_rate": 1.0315364814658913e-06,
+      "loss": 0.669,
+      "step": 971
+    },
+    {
+      "epoch": 2.58,
+      "learning_rate": 1.0189054633622407e-06,
+      "loss": 0.6184,
+      "step": 972
+    },
+    {
+      "epoch": 2.58,
+      "learning_rate": 1.0063481025019538e-06,
+      "loss": 0.6595,
+      "step": 973
+    },
+    {
+      "epoch": 2.58,
+      "learning_rate": 9.938645018725524e-07,
+      "loss": 0.6583,
+      "step": 974
+    },
+    {
+      "epoch": 2.59,
+      "learning_rate": 9.814547638566052e-07,
+      "loss": 0.6261,
+      "step": 975
+    },
+    {
+      "epoch": 2.59,
+      "learning_rate": 9.691189902309228e-07,
+      "loss": 0.6765,
+      "step": 976
+    },
+    {
+      "epoch": 2.59,
+      "learning_rate": 9.568572821656963e-07,
+      "loss": 0.6334,
+      "step": 977
+    },
+    {
+      "epoch": 2.59,
+      "learning_rate": 9.446697402236816e-07,
+      "loss": 0.6282,
+      "step": 978
+    },
+    {
+      "epoch": 2.6,
+      "learning_rate": 9.325564643593698e-07,
+      "loss": 0.6429,
+      "step": 979
+    },
+    {
+      "epoch": 2.6,
+      "learning_rate": 9.205175539181721e-07,
+      "loss": 0.6286,
+      "step": 980
+    },
+    {
+      "epoch": 2.6,
+      "learning_rate": 9.08553107635598e-07,
+      "loss": 0.6857,
+      "step": 981
+    },
+    {
+      "epoch": 2.6,
+      "learning_rate": 8.966632236364503e-07,
+      "loss": 0.6502,
+      "step": 982
+    },
+    {
+      "epoch": 2.61,
+      "learning_rate": 8.848479994340209e-07,
+      "loss": 0.627,
+      "step": 983
+    },
+    {
+      "epoch": 2.61,
+      "learning_rate": 8.731075319292836e-07,
+      "loss": 0.6063,
+      "step": 984
+    },
+    {
+      "epoch": 2.61,
+      "learning_rate": 8.614419174101152e-07,
+      "loss": 0.5699,
+      "step": 985
+    },
+    {
+      "epoch": 2.62,
+      "learning_rate": 8.498512515504842e-07,
+      "loss": 0.6457,
+      "step": 986
+    },
+    {
+      "epoch": 2.62,
+      "learning_rate": 8.383356294096889e-07,
+      "loss": 0.6528,
+      "step": 987
+    },
+    {
+      "epoch": 2.62,
+      "learning_rate": 8.268951454315555e-07,
+      "loss": 0.6519,
+      "step": 988
+    },
+    {
+      "epoch": 2.62,
+      "learning_rate": 8.155298934436873e-07,
+      "loss": 0.6201,
+      "step": 989
+    },
+    {
+      "epoch": 2.63,
+      "learning_rate": 8.042399666566702e-07,
+      "loss": 0.6164,
+      "step": 990
+    },
+    {
+      "epoch": 2.63,
+      "learning_rate": 7.930254576633267e-07,
+      "loss": 0.6109,
+      "step": 991
+    },
+    {
+      "epoch": 2.63,
+      "learning_rate": 7.818864584379515e-07,
+      "loss": 0.6657,
+      "step": 992
+    },
+    {
+      "epoch": 2.63,
+      "learning_rate": 7.708230603355516e-07,
+      "loss": 0.6315,
+      "step": 993
+    },
+    {
+      "epoch": 2.64,
+      "learning_rate": 7.598353540911052e-07,
+      "loss": 0.6228,
+      "step": 994
+    },
+    {
+      "epoch": 2.64,
+      "learning_rate": 7.489234298188142e-07,
+      "loss": 0.6426,
+      "step": 995
+    },
+    {
+      "epoch": 2.64,
+      "learning_rate": 7.380873770113617e-07,
+      "loss": 0.5989,
+      "step": 996
+    },
+    {
+      "epoch": 2.64,
+      "learning_rate": 7.27327284539181e-07,
+      "loss": 0.6469,
+      "step": 997
+    },
+    {
+      "epoch": 2.65,
+      "learning_rate": 7.166432406497337e-07,
+      "loss": 0.6063,
+      "step": 998
+    },
+    {
+      "epoch": 2.65,
+      "learning_rate": 7.060353329667668e-07,
+      "loss": 0.6461,
+      "step": 999
+    },
+    {
+      "epoch": 2.65,
+      "learning_rate": 6.955036484896227e-07,
+      "loss": 0.6302,
+      "step": 1000
+    },
+    {
+      "epoch": 2.66,
+      "learning_rate": 6.850482735924923e-07,
+      "loss": 0.6567,
+      "step": 1001
+    },
+    {
+      "epoch": 2.66,
+      "learning_rate": 6.746692940237365e-07,
+      "loss": 0.621,
+      "step": 1002
+    },
+    {
+      "epoch": 2.66,
+      "learning_rate": 6.643667949051625e-07,
+      "loss": 0.6307,
+      "step": 1003
+    },
+    {
+      "epoch": 2.66,
+      "learning_rate": 6.541408607313393e-07,
+      "loss": 0.6815,
+      "step": 1004
+    },
+    {
+      "epoch": 2.67,
+      "learning_rate": 6.43991575368893e-07,
+      "loss": 0.6307,
+      "step": 1005
+    },
+    {
+      "epoch": 2.67,
+      "learning_rate": 6.3391902205583e-07,
+      "loss": 0.6425,
+      "step": 1006
+    },
+    {
+      "epoch": 2.67,
+      "learning_rate": 6.239232834008446e-07,
+      "loss": 0.5933,
+      "step": 1007
+    },
+    {
+      "epoch": 2.67,
+      "learning_rate": 6.140044413826485e-07,
+      "loss": 0.6741,
+      "step": 1008
+    },
+    {
+      "epoch": 2.68,
+      "learning_rate": 6.041625773492954e-07,
+      "loss": 0.6381,
+      "step": 1009
+    },
+    {
+      "epoch": 2.68,
+      "learning_rate": 5.943977720175143e-07,
+      "loss": 0.6525,
+      "step": 1010
+    },
+    {
+      "epoch": 2.68,
+      "learning_rate": 5.847101054720495e-07,
+      "loss": 0.623,
+      "step": 1011
+    },
+    {
+      "epoch": 2.68,
+      "learning_rate": 5.750996571649958e-07,
+      "loss": 0.6333,
+      "step": 1012
+    },
+    {
+      "epoch": 2.69,
+      "learning_rate": 5.655665059151627e-07,
+      "loss": 0.6664,
+      "step": 1013
+    },
+    {
+      "epoch": 2.69,
+      "learning_rate": 5.561107299074076e-07,
+      "loss": 0.622,
+      "step": 1014
+    },
+    {
+      "epoch": 2.69,
+      "learning_rate": 5.467324066920121e-07,
+      "loss": 0.631,
+      "step": 1015
+    },
+    {
+      "epoch": 2.69,
+      "learning_rate": 5.374316131840385e-07,
+      "loss": 0.6598,
+      "step": 1016
+    },
+    {
+      "epoch": 2.7,
+      "learning_rate": 5.282084256626973e-07,
+      "loss": 0.6132,
+      "step": 1017
+    },
+    {
+      "epoch": 2.7,
+      "learning_rate": 5.190629197707264e-07,
+      "loss": 0.6289,
+      "step": 1018
+    },
+    {
+      "epoch": 2.7,
+      "learning_rate": 5.099951705137641e-07,
+      "loss": 0.6403,
+      "step": 1019
+    },
+    {
+      "epoch": 2.71,
+      "learning_rate": 5.010052522597442e-07,
+      "loss": 0.6634,
+      "step": 1020
+    },
+    {
+      "epoch": 2.71,
+      "learning_rate": 4.920932387382738e-07,
+      "loss": 0.6477,
+      "step": 1021
+    },
+    {
+      "epoch": 2.71,
+      "learning_rate": 4.832592030400396e-07,
+      "loss": 0.6305,
+      "step": 1022
+    },
+    {
+      "epoch": 2.71,
+      "learning_rate": 4.745032176162012e-07,
+      "loss": 0.5913,
+      "step": 1023
+    },
+    {
+      "epoch": 2.72,
+      "learning_rate": 4.658253542777991e-07,
+      "loss": 0.6139,
+      "step": 1024
+    },
+    {
+      "epoch": 2.72,
+      "learning_rate": 4.5722568419516723e-07,
+      "loss": 0.6244,
+      "step": 1025
+    },
+    {
+      "epoch": 2.72,
+      "learning_rate": 4.4870427789734606e-07,
+      "loss": 0.6589,
+      "step": 1026
+    },
+    {
+      "epoch": 2.72,
+      "learning_rate": 4.402612052715083e-07,
+      "loss": 0.6659,
+      "step": 1027
+    },
+    {
+      "epoch": 2.73,
+      "learning_rate": 4.3189653556238385e-07,
+      "loss": 0.679,
+      "step": 1028
+    },
+    {
+      "epoch": 2.73,
+      "learning_rate": 4.236103373716893e-07,
+      "loss": 0.6441,
+      "step": 1029
+    },
+    {
+      "epoch": 2.73,
+      "learning_rate": 4.1540267865756825e-07,
+      "loss": 0.6259,
+      "step": 1030
+    },
+    {
+      "epoch": 2.73,
+      "learning_rate": 4.0727362673403493e-07,
+      "loss": 0.6314,
+      "step": 1031
+    },
+    {
+      "epoch": 2.74,
+      "learning_rate": 3.9922324827041837e-07,
+      "loss": 0.6616,
+      "step": 1032
+    },
+    {
+      "epoch": 2.74,
+      "learning_rate": 3.9125160929082027e-07,
+      "loss": 0.6101,
+      "step": 1033
+    },
+    {
+      "epoch": 2.74,
+      "learning_rate": 3.833587751735679e-07,
+      "loss": 0.6436,
+      "step": 1034
+    },
+    {
+      "epoch": 2.75,
+      "learning_rate": 3.7554481065068427e-07,
+      "loss": 0.6047,
+      "step": 1035
+    },
+    {
+      "epoch": 2.75,
+      "learning_rate": 3.678097798073499e-07,
+      "loss": 0.6413,
+      "step": 1036
+    },
+    {
+      "epoch": 2.75,
+      "learning_rate": 3.601537460813875e-07,
+      "loss": 0.6186,
+      "step": 1037
+    },
+    {
+      "epoch": 2.75,
+      "learning_rate": 3.525767722627282e-07,
+      "loss": 0.6734,
+      "step": 1038
+    },
+    {
+      "epoch": 2.76,
+      "learning_rate": 3.4507892049290945e-07,
+      "loss": 0.6741,
+      "step": 1039
+    },
+    {
+      "epoch": 2.76,
+      "learning_rate": 3.376602522645556e-07,
+      "loss": 0.6373,
+      "step": 1040
+    },
+    {
+      "epoch": 2.76,
+      "learning_rate": 3.303208284208825e-07,
+      "loss": 0.6504,
+      "step": 1041
+    },
+    {
+      "epoch": 2.76,
+      "learning_rate": 3.2306070915519074e-07,
+      "loss": 0.6245,
+      "step": 1042
+    },
+    {
+      "epoch": 2.77,
+      "learning_rate": 3.158799540103752e-07,
+      "loss": 0.6054,
+      "step": 1043
+    },
+    {
+      "epoch": 2.77,
+      "learning_rate": 3.0877862187843613e-07,
+      "loss": 0.6081,
+      "step": 1044
+    },
+    {
+      "epoch": 2.77,
+      "learning_rate": 3.017567709999991e-07,
+      "loss": 0.6208,
+      "step": 1045
+    },
+    {
+      "epoch": 2.77,
+      "learning_rate": 2.9481445896383244e-07,
+      "loss": 0.6646,
+      "step": 1046
+    },
+    {
+      "epoch": 2.78,
+      "learning_rate": 2.8795174270637714e-07,
+      "loss": 0.6592,
+      "step": 1047
+    },
+    {
+      "epoch": 2.78,
+      "learning_rate": 2.8116867851128215e-07,
+      "loss": 0.6746,
+      "step": 1048
+    },
+    {
+      "epoch": 2.78,
+      "learning_rate": 2.7446532200894103e-07,
+      "loss": 0.6784,
+      "step": 1049
+    },
+    {
+      "epoch": 2.79,
+      "learning_rate": 2.678417281760304e-07,
+      "loss": 0.6435,
+      "step": 1050
+    },
+    {
+      "epoch": 2.79,
+      "learning_rate": 2.6129795133506884e-07,
+      "loss": 0.6215,
+      "step": 1051
+    },
+    {
+      "epoch": 2.79,
+      "learning_rate": 2.5483404515396547e-07,
+      "loss": 0.6295,
+      "step": 1052
+    },
+    {
+      "epoch": 2.79,
+      "learning_rate": 2.4845006264558103e-07,
+      "loss": 0.6249,
+      "step": 1053
+    },
+    {
+      "epoch": 2.8,
+      "learning_rate": 2.421460561672928e-07,
+      "loss": 0.6334,
+      "step": 1054
+    },
+    {
+      "epoch": 2.8,
+      "learning_rate": 2.3592207742056396e-07,
+      "loss": 0.6476,
+      "step": 1055
+    },
+    {
+      "epoch": 2.8,
+      "learning_rate": 2.2977817745052366e-07,
+      "loss": 0.6537,
+      "step": 1056
+    },
+    {
+      "epoch": 2.8,
+      "learning_rate": 2.2371440664554322e-07,
+      "loss": 0.615,
+      "step": 1057
+    },
+    {
+      "epoch": 2.81,
+      "learning_rate": 2.1773081473682622e-07,
+      "loss": 0.6717,
+      "step": 1058
+    },
+    {
+      "epoch": 2.81,
+      "learning_rate": 2.1182745079800115e-07,
+      "loss": 0.6165,
+      "step": 1059
+    },
+    {
+      "epoch": 2.81,
+      "learning_rate": 2.0600436324471506e-07,
+      "loss": 0.697,
+      "step": 1060
+    },
+    {
+      "epoch": 2.81,
+      "learning_rate": 2.0026159983423944e-07,
+      "loss": 0.6163,
+      "step": 1061
+    },
+    {
+      "epoch": 2.82,
+      "learning_rate": 1.9459920766508266e-07,
+      "loss": 0.6131,
+      "step": 1062
+    },
+    {
+      "epoch": 2.82,
+      "learning_rate": 1.8901723317658938e-07,
+      "loss": 0.6064,
+      "step": 1063
+    },
+    {
+      "epoch": 2.82,
+      "learning_rate": 1.835157221485817e-07,
+      "loss": 0.6349,
+      "step": 1064
+    },
+    {
+      "epoch": 2.82,
+      "learning_rate": 1.780947197009608e-07,
+      "loss": 0.6461,
+      "step": 1065
+    },
+    {
+      "epoch": 2.83,
+      "learning_rate": 1.7275427029335267e-07,
+      "loss": 0.5979,
+      "step": 1066
+    },
+    {
+      "epoch": 2.83,
+      "learning_rate": 1.674944177247384e-07,
+      "loss": 0.629,
+      "step": 1067
+    },
+    {
+      "epoch": 2.83,
+      "learning_rate": 1.623152051330934e-07,
+      "loss": 0.6514,
+      "step": 1068
+    },
+    {
+      "epoch": 2.84,
+      "learning_rate": 1.5721667499503544e-07,
+      "loss": 0.6474,
+      "step": 1069
+    },
+    {
+      "epoch": 2.84,
+      "learning_rate": 1.521988691254772e-07,
+      "loss": 0.7004,
+      "step": 1070
+    },
+    {
+      "epoch": 2.84,
+      "learning_rate": 1.472618286772809e-07,
+      "loss": 0.6315,
+      "step": 1071
+    },
+    {
+      "epoch": 2.84,
+      "learning_rate": 1.4240559414092192e-07,
+      "loss": 0.6665,
+      "step": 1072
+    },
+    {
+      "epoch": 2.85,
+      "learning_rate": 1.3763020534415695e-07,
+      "loss": 0.6684,
+      "step": 1073
+    },
+    {
+      "epoch": 2.85,
+      "learning_rate": 1.3293570145169744e-07,
+      "loss": 0.6333,
+      "step": 1074
+    },
+    {
+      "epoch": 2.85,
+      "learning_rate": 1.2832212096488772e-07,
+      "loss": 0.6458,
+      "step": 1075
+    },
+    {
+      "epoch": 2.85,
+      "learning_rate": 1.237895017213897e-07,
+      "loss": 0.6204,
+      "step": 1076
+    },
+    {
+      "epoch": 2.86,
+      "learning_rate": 1.1933788089487418e-07,
+      "loss": 0.5991,
+      "step": 1077
+    },
+    {
+      "epoch": 2.86,
+      "learning_rate": 1.1496729499471226e-07,
+      "loss": 0.6066,
+      "step": 1078
+    },
+    {
+      "epoch": 2.86,
+      "learning_rate": 1.1067777986567663e-07,
+      "loss": 0.6574,
+      "step": 1079
+    },
+    {
+      "epoch": 2.86,
+      "learning_rate": 1.0646937068765407e-07,
+      "loss": 0.6512,
+      "step": 1080
+    },
+    {
+      "epoch": 2.87,
+      "learning_rate": 1.0234210197534567e-07,
+      "loss": 0.6617,
+      "step": 1081
+    },
+    {
+      "epoch": 2.87,
+      "learning_rate": 9.829600757799373e-08,
+      "loss": 0.6106,
+      "step": 1082
+    },
+    {
+      "epoch": 2.87,
+      "learning_rate": 9.433112067909977e-08,
+      "loss": 0.6568,
+      "step": 1083
+    },
+    {
+      "epoch": 2.88,
+      "learning_rate": 9.044747379615248e-08,
+      "loss": 0.631,
+      "step": 1084
+    },
+    {
+      "epoch": 2.88,
+      "learning_rate": 8.664509878036021e-08,
+      "loss": 0.5966,
+      "step": 1085
+    },
+    {
+      "epoch": 2.88,
+      "learning_rate": 8.292402681639333e-08,
+      "loss": 0.6299,
+      "step": 1086
+    },
+    {
+      "epoch": 2.88,
+      "learning_rate": 7.928428842212343e-08,
+      "loss": 0.6349,
+      "step": 1087
+    },
+    {
+      "epoch": 2.89,
+      "learning_rate": 7.57259134483801e-08,
+      "loss": 0.625,
+      "step": 1088
+    },
+    {
+      "epoch": 2.89,
+      "learning_rate": 7.224893107869446e-08,
+      "loss": 0.6291,
+      "step": 1089
+    },
+    {
+      "epoch": 2.89,
+      "learning_rate": 6.885336982907609e-08,
+      "loss": 0.647,
+      "step": 1090
+    },
+    {
+      "epoch": 2.89,
+      "learning_rate": 6.553925754776314e-08,
+      "loss": 0.647,
+      "step": 1091
+    },
+    {
+      "epoch": 2.9,
+      "learning_rate": 6.230662141500254e-08,
+      "loss": 0.631,
+      "step": 1092
+    },
+    {
+      "epoch": 2.9,
+      "learning_rate": 5.915548794282911e-08,
+      "loss": 0.6476,
+      "step": 1093
+    },
+    {
+      "epoch": 2.9,
+      "learning_rate": 5.608588297483897e-08,
+      "loss": 0.6398,
+      "step": 1094
+    },
+    {
+      "epoch": 2.9,
+      "learning_rate": 5.309783168598759e-08,
+      "loss": 0.6487,
+      "step": 1095
+    },
+    {
+      "epoch": 2.91,
+      "learning_rate": 5.019135858237767e-08,
+      "loss": 0.6044,
+      "step": 1096
+    },
+    {
+      "epoch": 2.91,
+      "learning_rate": 4.73664875010571e-08,
+      "loss": 0.6093,
+      "step": 1097
+    },
+    {
+      "epoch": 2.91,
+      "learning_rate": 4.4623241609828004e-08,
+      "loss": 0.6123,
+      "step": 1098
+    },
+    {
+      "epoch": 2.92,
+      "learning_rate": 4.1961643407055775e-08,
+      "loss": 0.6445,
+      "step": 1099
+    },
+    {
+      "epoch": 2.92,
+      "learning_rate": 3.938171472147922e-08,
+      "loss": 0.6379,
+      "step": 1100
+    },
+    {
+      "epoch": 2.92,
+      "learning_rate": 3.6883476712038514e-08,
+      "loss": 0.6287,
+      "step": 1101
+    },
+    {
+      "epoch": 2.92,
+      "learning_rate": 3.4466949867698606e-08,
+      "loss": 0.6091,
+      "step": 1102
+    },
+    {
+      "epoch": 2.93,
+      "learning_rate": 3.213215400727942e-08,
+      "loss": 0.6426,
+      "step": 1103
+    },
+    {
+      "epoch": 2.93,
+      "learning_rate": 2.987910827929708e-08,
+      "loss": 0.6582,
+      "step": 1104
+    },
+    {
+      "epoch": 2.93,
+      "learning_rate": 2.7707831161803977e-08,
+      "loss": 0.6309,
+      "step": 1105
+    },
+    {
+      "epoch": 2.93,
+      "learning_rate": 2.5618340462237877e-08,
+      "loss": 0.6401,
+      "step": 1106
+    },
+    {
+      "epoch": 2.94,
+      "learning_rate": 2.361065331727863e-08,
+      "loss": 0.6463,
+      "step": 1107
+    },
+    {
+      "epoch": 2.94,
+      "learning_rate": 2.1684786192701646e-08,
+      "loss": 0.6053,
+      "step": 1108
+    },
+    {
+      "epoch": 2.94,
+      "learning_rate": 1.9840754883247993e-08,
+      "loss": 0.639,
+      "step": 1109
+    },
+    {
+      "epoch": 2.94,
+      "learning_rate": 1.8078574512494508e-08,
+      "loss": 0.6274,
+      "step": 1110
+    },
+    {
+      "epoch": 2.95,
+      "learning_rate": 1.6398259532726115e-08,
+      "loss": 0.6257,
+      "step": 1111
+    },
+    {
+      "epoch": 2.95,
+      "learning_rate": 1.4799823724821472e-08,
+      "loss": 0.6344,
+      "step": 1112
+    },
+    {
+      "epoch": 2.95,
+      "learning_rate": 1.3283280198137516e-08,
+      "loss": 0.6533,
+      "step": 1113
+    },
+    {
+      "epoch": 2.95,
+      "learning_rate": 1.1848641390403981e-08,
+      "loss": 0.6709,
+      "step": 1114
+    },
+    {
+      "epoch": 2.96,
+      "learning_rate": 1.0495919067617933e-08,
+      "loss": 0.6407,
+      "step": 1115
+    },
+    {
+      "epoch": 2.96,
+      "learning_rate": 9.225124323949396e-09,
+      "loss": 0.6442,
+      "step": 1116
+    },
+    {
+      "epoch": 2.96,
+      "learning_rate": 8.036267581652546e-09,
+      "loss": 0.6102,
+      "step": 1117
+    },
+    {
+      "epoch": 2.97,
+      "learning_rate": 6.929358590974655e-09,
+      "loss": 0.6602,
+      "step": 1118
+    },
+    {
+      "epoch": 2.97,
+      "learning_rate": 5.904406430083942e-09,
+      "loss": 0.6439,
+      "step": 1119
+    },
+    {
+      "epoch": 2.97,
+      "learning_rate": 4.961419504987408e-09,
+      "loss": 0.6426,
+      "step": 1120
+    },
+    {
+      "epoch": 2.97,
+      "learning_rate": 4.100405549465336e-09,
+      "loss": 0.6005,
+      "step": 1121
+    },
+    {
+      "epoch": 2.98,
+      "learning_rate": 3.3213716250080073e-09,
+      "loss": 0.6417,
+      "step": 1122
+    },
+    {
+      "epoch": 2.98,
+      "learning_rate": 2.624324120759081e-09,
+      "loss": 0.6562,
+      "step": 1123
+    },
+    {
+      "epoch": 2.98,
+      "learning_rate": 2.0092687534589703e-09,
+      "loss": 0.6209,
+      "step": 1124
+    },
+    {
+      "epoch": 2.98,
+      "learning_rate": 1.4762105674026583e-09,
+      "loss": 0.6844,
+      "step": 1125
+    },
+    {
+      "epoch": 2.99,
+      "learning_rate": 1.0251539343952844e-09,
+      "loss": 0.5994,
+      "step": 1126
+    },
+    {
+      "epoch": 2.99,
+      "learning_rate": 6.561025537177301e-10,
+      "loss": 0.6534,
+      "step": 1127
+    },
+    {
+      "epoch": 2.99,
+      "learning_rate": 3.690594520944224e-10,
+      "loss": 0.6521,
+      "step": 1128
+    },
+    {
+      "epoch": 2.99,
+      "learning_rate": 1.6402698367223857e-10,
+      "loss": 0.6166,
+      "step": 1129
+    },
+    {
+      "epoch": 3.0,
+      "learning_rate": 4.1006829996081963e-11,
+      "loss": 0.6529,
+      "step": 1130
+    },
+    {
+      "epoch": 3.0,
+      "learning_rate": 0.0,
+      "loss": 0.621,
+      "step": 1131
+    },
+    {
+      "epoch": 3.0,
+      "step": 1131,
+      "total_flos": 1.4692330300432712e+18,
+      "train_loss": 0.6702391549510096,
+      "train_runtime": 12400.886,
+      "train_samples_per_second": 11.669,
+      "train_steps_per_second": 0.091
+    }
+  ],
+  "logging_steps": 1.0,
+  "max_steps": 1131,
+  "num_input_tokens_seen": 0,
+  "num_train_epochs": 3,
+  "save_steps": 1200,
+  "total_flos": 1.4692330300432712e+18,
+  "train_batch_size": 4,
+  "trial_name": null,
+  "trial_params": null
+}