diff --git "a/trainer_state.json" "b/trainer_state.json"
new file mode 100644--- /dev/null
+++ "b/trainer_state.json"
@@ -0,0 +1,9408 @@
+{
+  "best_metric": null,
+  "best_model_checkpoint": null,
+  "epoch": 1.0,
+  "eval_steps": 500,
+  "global_step": 1563,
+  "is_hyper_param_search": false,
+  "is_local_process_zero": true,
+  "is_world_process_zero": true,
+  "log_history": [
+    {
+      "epoch": 0.0,
+      "learning_rate": 4.2553191489361704e-07,
+      "loss": 0.317,
+      "step": 1
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 8.510638297872341e-07,
+      "loss": 0.3577,
+      "step": 2
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 1.276595744680851e-06,
+      "loss": 0.2838,
+      "step": 3
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 1.7021276595744682e-06,
+      "loss": 0.3348,
+      "step": 4
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 2.1276595744680853e-06,
+      "loss": 0.3636,
+      "step": 5
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 2.553191489361702e-06,
+      "loss": 0.3244,
+      "step": 6
+    },
+    {
+      "epoch": 0.0,
+      "learning_rate": 2.978723404255319e-06,
+      "loss": 0.3568,
+      "step": 7
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 3.4042553191489363e-06,
+      "loss": 0.3503,
+      "step": 8
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 3.8297872340425535e-06,
+      "loss": 0.3317,
+      "step": 9
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.255319148936171e-06,
+      "loss": 0.3605,
+      "step": 10
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 4.680851063829788e-06,
+      "loss": 0.3293,
+      "step": 11
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 5.106382978723404e-06,
+      "loss": 0.3454,
+      "step": 12
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 5.531914893617022e-06,
+      "loss": 0.3324,
+      "step": 13
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 5.957446808510638e-06,
+      "loss": 0.3036,
+      "step": 14
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 6.382978723404256e-06,
+      "loss": 0.323,
+      "step": 15
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 6.808510638297873e-06,
+      "loss": 0.3073,
+      "step": 16
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 7.234042553191491e-06,
+      "loss": 0.3251,
+      "step": 17
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 7.659574468085107e-06,
+      "loss": 0.296,
+      "step": 18
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 8.085106382978723e-06,
+      "loss": 0.3195,
+      "step": 19
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 8.510638297872341e-06,
+      "loss": 0.3427,
+      "step": 20
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 8.936170212765958e-06,
+      "loss": 0.3443,
+      "step": 21
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 9.361702127659576e-06,
+      "loss": 0.3195,
+      "step": 22
+    },
+    {
+      "epoch": 0.01,
+      "learning_rate": 9.787234042553192e-06,
+      "loss": 0.3327,
+      "step": 23
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 1.0212765957446808e-05,
+      "loss": 0.3121,
+      "step": 24
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 1.0638297872340426e-05,
+      "loss": 0.3161,
+      "step": 25
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 1.1063829787234044e-05,
+      "loss": 0.2794,
+      "step": 26
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 1.1489361702127662e-05,
+      "loss": 0.2609,
+      "step": 27
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 1.1914893617021277e-05,
+      "loss": 0.3522,
+      "step": 28
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 1.2340425531914895e-05,
+      "loss": 0.2945,
+      "step": 29
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 1.2765957446808513e-05,
+      "loss": 0.3045,
+      "step": 30
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 1.3191489361702127e-05,
+      "loss": 0.2544,
+      "step": 31
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 1.3617021276595745e-05,
+      "loss": 0.3052,
+      "step": 32
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 1.4042553191489363e-05,
+      "loss": 0.2538,
+      "step": 33
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 1.4468085106382981e-05,
+      "loss": 0.2716,
+      "step": 34
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 1.4893617021276596e-05,
+      "loss": 0.2709,
+      "step": 35
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 1.5319148936170214e-05,
+      "loss": 0.2853,
+      "step": 36
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 1.5744680851063832e-05,
+      "loss": 0.2911,
+      "step": 37
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 1.6170212765957446e-05,
+      "loss": 0.276,
+      "step": 38
+    },
+    {
+      "epoch": 0.02,
+      "learning_rate": 1.6595744680851064e-05,
+      "loss": 0.2942,
+      "step": 39
+    },
+    {
+      "epoch": 0.03,
+      "learning_rate": 1.7021276595744682e-05,
+      "loss": 0.2862,
+      "step": 40
+    },
+    {
+      "epoch": 0.03,
+      "learning_rate": 1.74468085106383e-05,
+      "loss": 0.2796,
+      "step": 41
+    },
+    {
+      "epoch": 0.03,
+      "learning_rate": 1.7872340425531915e-05,
+      "loss": 0.256,
+      "step": 42
+    },
+    {
+      "epoch": 0.03,
+      "learning_rate": 1.8297872340425533e-05,
+      "loss": 0.3059,
+      "step": 43
+    },
+    {
+      "epoch": 0.03,
+      "learning_rate": 1.872340425531915e-05,
+      "loss": 0.2971,
+      "step": 44
+    },
+    {
+      "epoch": 0.03,
+      "learning_rate": 1.914893617021277e-05,
+      "loss": 0.2945,
+      "step": 45
+    },
+    {
+      "epoch": 0.03,
+      "learning_rate": 1.9574468085106384e-05,
+      "loss": 0.2483,
+      "step": 46
+    },
+    {
+      "epoch": 0.03,
+      "learning_rate": 2e-05,
+      "loss": 0.2811,
+      "step": 47
+    },
+    {
+      "epoch": 0.03,
+      "learning_rate": 1.999997852806461e-05,
+      "loss": 0.2765,
+      "step": 48
+    },
+    {
+      "epoch": 0.03,
+      "learning_rate": 1.9999914112350643e-05,
+      "loss": 0.3074,
+      "step": 49
+    },
+    {
+      "epoch": 0.03,
+      "learning_rate": 1.9999806753134728e-05,
+      "loss": 0.2711,
+      "step": 50
+    },
+    {
+      "epoch": 0.03,
+      "learning_rate": 1.999965645087791e-05,
+      "loss": 0.2875,
+      "step": 51
+    },
+    {
+      "epoch": 0.03,
+      "learning_rate": 1.9999463206225637e-05,
+      "loss": 0.2765,
+      "step": 52
+    },
+    {
+      "epoch": 0.03,
+      "learning_rate": 1.9999227020007783e-05,
+      "loss": 0.2769,
+      "step": 53
+    },
+    {
+      "epoch": 0.03,
+      "learning_rate": 1.999894789323862e-05,
+      "loss": 0.2761,
+      "step": 54
+    },
+    {
+      "epoch": 0.04,
+      "learning_rate": 1.9998625827116828e-05,
+      "loss": 0.2997,
+      "step": 55
+    },
+    {
+      "epoch": 0.04,
+      "learning_rate": 1.999826082302548e-05,
+      "loss": 0.2871,
+      "step": 56
+    },
+    {
+      "epoch": 0.04,
+      "learning_rate": 1.9997852882532053e-05,
+      "loss": 0.2803,
+      "step": 57
+    },
+    {
+      "epoch": 0.04,
+      "learning_rate": 1.9997402007388392e-05,
+      "loss": 0.283,
+      "step": 58
+    },
+    {
+      "epoch": 0.04,
+      "learning_rate": 1.9996908199530737e-05,
+      "loss": 0.2686,
+      "step": 59
+    },
+    {
+      "epoch": 0.04,
+      "learning_rate": 1.9996371461079686e-05,
+      "loss": 0.2681,
+      "step": 60
+    },
+    {
+      "epoch": 0.04,
+      "learning_rate": 1.99957917943402e-05,
+      "loss": 0.2857,
+      "step": 61
+    },
+    {
+      "epoch": 0.04,
+      "learning_rate": 1.9995169201801595e-05,
+      "loss": 0.2643,
+      "step": 62
+    },
+    {
+      "epoch": 0.04,
+      "learning_rate": 1.9994503686137526e-05,
+      "loss": 0.2558,
+      "step": 63
+    },
+    {
+      "epoch": 0.04,
+      "learning_rate": 1.999379525020597e-05,
+      "loss": 0.2507,
+      "step": 64
+    },
+    {
+      "epoch": 0.04,
+      "learning_rate": 1.999304389704923e-05,
+      "loss": 0.2478,
+      "step": 65
+    },
+    {
+      "epoch": 0.04,
+      "learning_rate": 1.999224962989391e-05,
+      "loss": 0.2141,
+      "step": 66
+    },
+    {
+      "epoch": 0.04,
+      "learning_rate": 1.999141245215089e-05,
+      "loss": 0.2704,
+      "step": 67
+    },
+    {
+      "epoch": 0.04,
+      "learning_rate": 1.9990532367415342e-05,
+      "loss": 0.2848,
+      "step": 68
+    },
+    {
+      "epoch": 0.04,
+      "learning_rate": 1.9989609379466688e-05,
+      "loss": 0.3175,
+      "step": 69
+    },
+    {
+      "epoch": 0.04,
+      "learning_rate": 1.99886434922686e-05,
+      "loss": 0.257,
+      "step": 70
+    },
+    {
+      "epoch": 0.05,
+      "learning_rate": 1.998763470996897e-05,
+      "loss": 0.2825,
+      "step": 71
+    },
+    {
+      "epoch": 0.05,
+      "learning_rate": 1.9986583036899896e-05,
+      "loss": 0.2394,
+      "step": 72
+    },
+    {
+      "epoch": 0.05,
+      "learning_rate": 1.998548847757767e-05,
+      "loss": 0.2995,
+      "step": 73
+    },
+    {
+      "epoch": 0.05,
+      "learning_rate": 1.9984351036702758e-05,
+      "loss": 0.2799,
+      "step": 74
+    },
+    {
+      "epoch": 0.05,
+      "learning_rate": 1.998317071915977e-05,
+      "loss": 0.2518,
+      "step": 75
+    },
+    {
+      "epoch": 0.05,
+      "learning_rate": 1.998194753001744e-05,
+      "loss": 0.2672,
+      "step": 76
+    },
+    {
+      "epoch": 0.05,
+      "learning_rate": 1.9980681474528624e-05,
+      "loss": 0.2544,
+      "step": 77
+    },
+    {
+      "epoch": 0.05,
+      "learning_rate": 1.997937255813025e-05,
+      "loss": 0.2846,
+      "step": 78
+    },
+    {
+      "epoch": 0.05,
+      "learning_rate": 1.9978020786443314e-05,
+      "loss": 0.2551,
+      "step": 79
+    },
+    {
+      "epoch": 0.05,
+      "learning_rate": 1.9976626165272844e-05,
+      "loss": 0.2362,
+      "step": 80
+    },
+    {
+      "epoch": 0.05,
+      "learning_rate": 1.9975188700607885e-05,
+      "loss": 0.2657,
+      "step": 81
+    },
+    {
+      "epoch": 0.05,
+      "learning_rate": 1.9973708398621464e-05,
+      "loss": 0.3069,
+      "step": 82
+    },
+    {
+      "epoch": 0.05,
+      "learning_rate": 1.9972185265670573e-05,
+      "loss": 0.2927,
+      "step": 83
+    },
+    {
+      "epoch": 0.05,
+      "learning_rate": 1.9970619308296137e-05,
+      "loss": 0.2828,
+      "step": 84
+    },
+    {
+      "epoch": 0.05,
+      "learning_rate": 1.996901053322298e-05,
+      "loss": 0.2599,
+      "step": 85
+    },
+    {
+      "epoch": 0.06,
+      "learning_rate": 1.9967358947359805e-05,
+      "loss": 0.2778,
+      "step": 86
+    },
+    {
+      "epoch": 0.06,
+      "learning_rate": 1.9965664557799164e-05,
+      "loss": 0.265,
+      "step": 87
+    },
+    {
+      "epoch": 0.06,
+      "learning_rate": 1.9963927371817412e-05,
+      "loss": 0.3038,
+      "step": 88
+    },
+    {
+      "epoch": 0.06,
+      "learning_rate": 1.996214739687471e-05,
+      "loss": 0.244,
+      "step": 89
+    },
+    {
+      "epoch": 0.06,
+      "learning_rate": 1.9960324640614953e-05,
+      "loss": 0.2746,
+      "step": 90
+    },
+    {
+      "epoch": 0.06,
+      "learning_rate": 1.9958459110865767e-05,
+      "loss": 0.2865,
+      "step": 91
+    },
+    {
+      "epoch": 0.06,
+      "learning_rate": 1.9956550815638448e-05,
+      "loss": 0.2393,
+      "step": 92
+    },
+    {
+      "epoch": 0.06,
+      "learning_rate": 1.995459976312797e-05,
+      "loss": 0.2989,
+      "step": 93
+    },
+    {
+      "epoch": 0.06,
+      "learning_rate": 1.9952605961712894e-05,
+      "loss": 0.2476,
+      "step": 94
+    },
+    {
+      "epoch": 0.06,
+      "learning_rate": 1.995056941995538e-05,
+      "loss": 0.2742,
+      "step": 95
+    },
+    {
+      "epoch": 0.06,
+      "learning_rate": 1.994849014660113e-05,
+      "loss": 0.2543,
+      "step": 96
+    },
+    {
+      "epoch": 0.06,
+      "learning_rate": 1.9946368150579343e-05,
+      "loss": 0.2402,
+      "step": 97
+    },
+    {
+      "epoch": 0.06,
+      "learning_rate": 1.9944203441002696e-05,
+      "loss": 0.2591,
+      "step": 98
+    },
+    {
+      "epoch": 0.06,
+      "learning_rate": 1.9941996027167286e-05,
+      "loss": 0.2233,
+      "step": 99
+    },
+    {
+      "epoch": 0.06,
+      "learning_rate": 1.9939745918552602e-05,
+      "loss": 0.2582,
+      "step": 100
+    },
+    {
+      "epoch": 0.06,
+      "learning_rate": 1.9937453124821487e-05,
+      "loss": 0.2826,
+      "step": 101
+    },
+    {
+      "epoch": 0.07,
+      "learning_rate": 1.993511765582008e-05,
+      "loss": 0.2645,
+      "step": 102
+    },
+    {
+      "epoch": 0.07,
+      "learning_rate": 1.993273952157779e-05,
+      "loss": 0.2636,
+      "step": 103
+    },
+    {
+      "epoch": 0.07,
+      "learning_rate": 1.9930318732307244e-05,
+      "loss": 0.2577,
+      "step": 104
+    },
+    {
+      "epoch": 0.07,
+      "learning_rate": 1.9927855298404255e-05,
+      "loss": 0.2676,
+      "step": 105
+    },
+    {
+      "epoch": 0.07,
+      "learning_rate": 1.9925349230447756e-05,
+      "loss": 0.2693,
+      "step": 106
+    },
+    {
+      "epoch": 0.07,
+      "learning_rate": 1.992280053919977e-05,
+      "loss": 0.2309,
+      "step": 107
+    },
+    {
+      "epoch": 0.07,
+      "learning_rate": 1.9920209235605373e-05,
+      "loss": 0.2303,
+      "step": 108
+    },
+    {
+      "epoch": 0.07,
+      "learning_rate": 1.9917575330792616e-05,
+      "loss": 0.2509,
+      "step": 109
+    },
+    {
+      "epoch": 0.07,
+      "learning_rate": 1.9914898836072513e-05,
+      "loss": 0.2351,
+      "step": 110
+    },
+    {
+      "epoch": 0.07,
+      "learning_rate": 1.9912179762938962e-05,
+      "loss": 0.2896,
+      "step": 111
+    },
+    {
+      "epoch": 0.07,
+      "learning_rate": 1.990941812306872e-05,
+      "loss": 0.3193,
+      "step": 112
+    },
+    {
+      "epoch": 0.07,
+      "learning_rate": 1.990661392832134e-05,
+      "loss": 0.2692,
+      "step": 113
+    },
+    {
+      "epoch": 0.07,
+      "learning_rate": 1.990376719073911e-05,
+      "loss": 0.2526,
+      "step": 114
+    },
+    {
+      "epoch": 0.07,
+      "learning_rate": 1.9900877922547034e-05,
+      "loss": 0.2349,
+      "step": 115
+    },
+    {
+      "epoch": 0.07,
+      "learning_rate": 1.989794613615274e-05,
+      "loss": 0.2378,
+      "step": 116
+    },
+    {
+      "epoch": 0.07,
+      "learning_rate": 1.989497184414646e-05,
+      "loss": 0.2622,
+      "step": 117
+    },
+    {
+      "epoch": 0.08,
+      "learning_rate": 1.9891955059300955e-05,
+      "loss": 0.2843,
+      "step": 118
+    },
+    {
+      "epoch": 0.08,
+      "learning_rate": 1.988889579457146e-05,
+      "loss": 0.2444,
+      "step": 119
+    },
+    {
+      "epoch": 0.08,
+      "learning_rate": 1.9885794063095646e-05,
+      "loss": 0.2693,
+      "step": 120
+    },
+    {
+      "epoch": 0.08,
+      "learning_rate": 1.9882649878193544e-05,
+      "loss": 0.2427,
+      "step": 121
+    },
+    {
+      "epoch": 0.08,
+      "learning_rate": 1.987946325336751e-05,
+      "loss": 0.2393,
+      "step": 122
+    },
+    {
+      "epoch": 0.08,
+      "learning_rate": 1.987623420230214e-05,
+      "loss": 0.2405,
+      "step": 123
+    },
+    {
+      "epoch": 0.08,
+      "learning_rate": 1.987296273886423e-05,
+      "loss": 0.2423,
+      "step": 124
+    },
+    {
+      "epoch": 0.08,
+      "learning_rate": 1.9869648877102707e-05,
+      "loss": 0.2426,
+      "step": 125
+    },
+    {
+      "epoch": 0.08,
+      "learning_rate": 1.9866292631248578e-05,
+      "loss": 0.2165,
+      "step": 126
+    },
+    {
+      "epoch": 0.08,
+      "learning_rate": 1.9862894015714866e-05,
+      "loss": 0.2439,
+      "step": 127
+    },
+    {
+      "epoch": 0.08,
+      "learning_rate": 1.9859453045096536e-05,
+      "loss": 0.2634,
+      "step": 128
+    },
+    {
+      "epoch": 0.08,
+      "learning_rate": 1.9855969734170448e-05,
+      "loss": 0.2478,
+      "step": 129
+    },
+    {
+      "epoch": 0.08,
+      "learning_rate": 1.985244409789529e-05,
+      "loss": 0.2713,
+      "step": 130
+    },
+    {
+      "epoch": 0.08,
+      "learning_rate": 1.984887615141151e-05,
+      "loss": 0.2166,
+      "step": 131
+    },
+    {
+      "epoch": 0.08,
+      "learning_rate": 1.9845265910041247e-05,
+      "loss": 0.2674,
+      "step": 132
+    },
+    {
+      "epoch": 0.09,
+      "learning_rate": 1.9841613389288278e-05,
+      "loss": 0.2595,
+      "step": 133
+    },
+    {
+      "epoch": 0.09,
+      "learning_rate": 1.983791860483794e-05,
+      "loss": 0.2638,
+      "step": 134
+    },
+    {
+      "epoch": 0.09,
+      "learning_rate": 1.9834181572557066e-05,
+      "loss": 0.2763,
+      "step": 135
+    },
+    {
+      "epoch": 0.09,
+      "learning_rate": 1.9830402308493927e-05,
+      "loss": 0.2742,
+      "step": 136
+    },
+    {
+      "epoch": 0.09,
+      "learning_rate": 1.9826580828878134e-05,
+      "loss": 0.2097,
+      "step": 137
+    },
+    {
+      "epoch": 0.09,
+      "learning_rate": 1.9822717150120607e-05,
+      "loss": 0.2682,
+      "step": 138
+    },
+    {
+      "epoch": 0.09,
+      "learning_rate": 1.9818811288813475e-05,
+      "loss": 0.2479,
+      "step": 139
+    },
+    {
+      "epoch": 0.09,
+      "learning_rate": 1.981486326173002e-05,
+      "loss": 0.2511,
+      "step": 140
+    },
+    {
+      "epoch": 0.09,
+      "learning_rate": 1.9810873085824604e-05,
+      "loss": 0.2488,
+      "step": 141
+    },
+    {
+      "epoch": 0.09,
+      "learning_rate": 1.9806840778232575e-05,
+      "loss": 0.2538,
+      "step": 142
+    },
+    {
+      "epoch": 0.09,
+      "learning_rate": 1.9802766356270226e-05,
+      "loss": 0.2702,
+      "step": 143
+    },
+    {
+      "epoch": 0.09,
+      "learning_rate": 1.979864983743471e-05,
+      "loss": 0.2321,
+      "step": 144
+    },
+    {
+      "epoch": 0.09,
+      "learning_rate": 1.9794491239403946e-05,
+      "loss": 0.2402,
+      "step": 145
+    },
+    {
+      "epoch": 0.09,
+      "learning_rate": 1.9790290580036563e-05,
+      "loss": 0.2024,
+      "step": 146
+    },
+    {
+      "epoch": 0.09,
+      "learning_rate": 1.9786047877371823e-05,
+      "loss": 0.2763,
+      "step": 147
+    },
+    {
+      "epoch": 0.09,
+      "learning_rate": 1.978176314962953e-05,
+      "loss": 0.2671,
+      "step": 148
+    },
+    {
+      "epoch": 0.1,
+      "learning_rate": 1.977743641520996e-05,
+      "loss": 0.2449,
+      "step": 149
+    },
+    {
+      "epoch": 0.1,
+      "learning_rate": 1.9773067692693795e-05,
+      "loss": 0.2821,
+      "step": 150
+    },
+    {
+      "epoch": 0.1,
+      "learning_rate": 1.976865700084201e-05,
+      "loss": 0.2495,
+      "step": 151
+    },
+    {
+      "epoch": 0.1,
+      "learning_rate": 1.976420435859583e-05,
+      "loss": 0.2605,
+      "step": 152
+    },
+    {
+      "epoch": 0.1,
+      "learning_rate": 1.9759709785076622e-05,
+      "loss": 0.2422,
+      "step": 153
+    },
+    {
+      "epoch": 0.1,
+      "learning_rate": 1.9755173299585822e-05,
+      "loss": 0.2473,
+      "step": 154
+    },
+    {
+      "epoch": 0.1,
+      "learning_rate": 1.975059492160486e-05,
+      "loss": 0.2604,
+      "step": 155
+    },
+    {
+      "epoch": 0.1,
+      "learning_rate": 1.9745974670795063e-05,
+      "loss": 0.2356,
+      "step": 156
+    },
+    {
+      "epoch": 0.1,
+      "learning_rate": 1.9741312566997573e-05,
+      "loss": 0.2638,
+      "step": 157
+    },
+    {
+      "epoch": 0.1,
+      "learning_rate": 1.9736608630233267e-05,
+      "loss": 0.263,
+      "step": 158
+    },
+    {
+      "epoch": 0.1,
+      "learning_rate": 1.9731862880702676e-05,
+      "loss": 0.2174,
+      "step": 159
+    },
+    {
+      "epoch": 0.1,
+      "learning_rate": 1.972707533878588e-05,
+      "loss": 0.2401,
+      "step": 160
+    },
+    {
+      "epoch": 0.1,
+      "learning_rate": 1.972224602504244e-05,
+      "loss": 0.2391,
+      "step": 161
+    },
+    {
+      "epoch": 0.1,
+      "learning_rate": 1.971737496021129e-05,
+      "loss": 0.2321,
+      "step": 162
+    },
+    {
+      "epoch": 0.1,
+      "learning_rate": 1.9712462165210687e-05,
+      "loss": 0.2572,
+      "step": 163
+    },
+    {
+      "epoch": 0.1,
+      "learning_rate": 1.9707507661138058e-05,
+      "loss": 0.2798,
+      "step": 164
+    },
+    {
+      "epoch": 0.11,
+      "learning_rate": 1.9702511469269964e-05,
+      "loss": 0.2399,
+      "step": 165
+    },
+    {
+      "epoch": 0.11,
+      "learning_rate": 1.969747361106199e-05,
+      "loss": 0.2364,
+      "step": 166
+    },
+    {
+      "epoch": 0.11,
+      "learning_rate": 1.969239410814865e-05,
+      "loss": 0.2483,
+      "step": 167
+    },
+    {
+      "epoch": 0.11,
+      "learning_rate": 1.9687272982343297e-05,
+      "loss": 0.271,
+      "step": 168
+    },
+    {
+      "epoch": 0.11,
+      "learning_rate": 1.968211025563802e-05,
+      "loss": 0.2399,
+      "step": 169
+    },
+    {
+      "epoch": 0.11,
+      "learning_rate": 1.9676905950203568e-05,
+      "loss": 0.2516,
+      "step": 170
+    },
+    {
+      "epoch": 0.11,
+      "learning_rate": 1.967166008838925e-05,
+      "loss": 0.2899,
+      "step": 171
+    },
+    {
+      "epoch": 0.11,
+      "learning_rate": 1.9666372692722818e-05,
+      "loss": 0.2467,
+      "step": 172
+    },
+    {
+      "epoch": 0.11,
+      "learning_rate": 1.9661043785910404e-05,
+      "loss": 0.2388,
+      "step": 173
+    },
+    {
+      "epoch": 0.11,
+      "learning_rate": 1.9655673390836392e-05,
+      "loss": 0.2287,
+      "step": 174
+    },
+    {
+      "epoch": 0.11,
+      "learning_rate": 1.9650261530563336e-05,
+      "loss": 0.2801,
+      "step": 175
+    },
+    {
+      "epoch": 0.11,
+      "learning_rate": 1.9644808228331863e-05,
+      "loss": 0.252,
+      "step": 176
+    },
+    {
+      "epoch": 0.11,
+      "learning_rate": 1.963931350756056e-05,
+      "loss": 0.2713,
+      "step": 177
+    },
+    {
+      "epoch": 0.11,
+      "learning_rate": 1.9633777391845887e-05,
+      "loss": 0.2666,
+      "step": 178
+    },
+    {
+      "epoch": 0.11,
+      "learning_rate": 1.9628199904962063e-05,
+      "loss": 0.2509,
+      "step": 179
+    },
+    {
+      "epoch": 0.12,
+      "learning_rate": 1.9622581070860983e-05,
+      "loss": 0.2684,
+      "step": 180
+    },
+    {
+      "epoch": 0.12,
+      "learning_rate": 1.9616920913672092e-05,
+      "loss": 0.2466,
+      "step": 181
+    },
+    {
+      "epoch": 0.12,
+      "learning_rate": 1.9611219457702298e-05,
+      "loss": 0.2753,
+      "step": 182
+    },
+    {
+      "epoch": 0.12,
+      "learning_rate": 1.9605476727435854e-05,
+      "loss": 0.2793,
+      "step": 183
+    },
+    {
+      "epoch": 0.12,
+      "learning_rate": 1.959969274753427e-05,
+      "loss": 0.2549,
+      "step": 184
+    },
+    {
+      "epoch": 0.12,
+      "learning_rate": 1.9593867542836198e-05,
+      "loss": 0.2464,
+      "step": 185
+    },
+    {
+      "epoch": 0.12,
+      "learning_rate": 1.958800113835732e-05,
+      "loss": 0.265,
+      "step": 186
+    },
+    {
+      "epoch": 0.12,
+      "learning_rate": 1.9582093559290242e-05,
+      "loss": 0.261,
+      "step": 187
+    },
+    {
+      "epoch": 0.12,
+      "learning_rate": 1.95761448310044e-05,
+      "loss": 0.2777,
+      "step": 188
+    },
+    {
+      "epoch": 0.12,
+      "learning_rate": 1.9570154979045938e-05,
+      "loss": 0.2412,
+      "step": 189
+    },
+    {
+      "epoch": 0.12,
+      "learning_rate": 1.9564124029137598e-05,
+      "loss": 0.28,
+      "step": 190
+    },
+    {
+      "epoch": 0.12,
+      "learning_rate": 1.9558052007178608e-05,
+      "loss": 0.1896,
+      "step": 191
+    },
+    {
+      "epoch": 0.12,
+      "learning_rate": 1.9551938939244585e-05,
+      "loss": 0.2366,
+      "step": 192
+    },
+    {
+      "epoch": 0.12,
+      "learning_rate": 1.954578485158741e-05,
+      "loss": 0.2215,
+      "step": 193
+    },
+    {
+      "epoch": 0.12,
+      "learning_rate": 1.9539589770635117e-05,
+      "loss": 0.2395,
+      "step": 194
+    },
+    {
+      "epoch": 0.12,
+      "learning_rate": 1.9533353722991775e-05,
+      "loss": 0.2924,
+      "step": 195
+    },
+    {
+      "epoch": 0.13,
+      "learning_rate": 1.95270767354374e-05,
+      "loss": 0.2533,
+      "step": 196
+    },
+    {
+      "epoch": 0.13,
+      "learning_rate": 1.952075883492779e-05,
+      "loss": 0.2928,
+      "step": 197
+    },
+    {
+      "epoch": 0.13,
+      "learning_rate": 1.9514400048594467e-05,
+      "loss": 0.225,
+      "step": 198
+    },
+    {
+      "epoch": 0.13,
+      "learning_rate": 1.9508000403744516e-05,
+      "loss": 0.2317,
+      "step": 199
+    },
+    {
+      "epoch": 0.13,
+      "learning_rate": 1.950155992786049e-05,
+      "loss": 0.2669,
+      "step": 200
+    },
+    {
+      "epoch": 0.13,
+      "learning_rate": 1.9495078648600286e-05,
+      "loss": 0.2497,
+      "step": 201
+    },
+    {
+      "epoch": 0.13,
+      "learning_rate": 1.948855659379703e-05,
+      "loss": 0.3071,
+      "step": 202
+    },
+    {
+      "epoch": 0.13,
+      "learning_rate": 1.948199379145894e-05,
+      "loss": 0.2567,
+      "step": 203
+    },
+    {
+      "epoch": 0.13,
+      "learning_rate": 1.947539026976924e-05,
+      "loss": 0.2686,
+      "step": 204
+    },
+    {
+      "epoch": 0.13,
+      "learning_rate": 1.9468746057086e-05,
+      "loss": 0.2797,
+      "step": 205
+    },
+    {
+      "epoch": 0.13,
+      "learning_rate": 1.9462061181942045e-05,
+      "loss": 0.2571,
+      "step": 206
+    },
+    {
+      "epoch": 0.13,
+      "learning_rate": 1.9455335673044817e-05,
+      "loss": 0.2707,
+      "step": 207
+    },
+    {
+      "epoch": 0.13,
+      "learning_rate": 1.9448569559276252e-05,
+      "loss": 0.2652,
+      "step": 208
+    },
+    {
+      "epoch": 0.13,
+      "learning_rate": 1.9441762869692664e-05,
+      "loss": 0.2485,
+      "step": 209
+    },
+    {
+      "epoch": 0.13,
+      "learning_rate": 1.9434915633524612e-05,
+      "loss": 0.2528,
+      "step": 210
+    },
+    {
+      "epoch": 0.13,
+      "learning_rate": 1.9428027880176775e-05,
+      "loss": 0.2437,
+      "step": 211
+    },
+    {
+      "epoch": 0.14,
+      "learning_rate": 1.9421099639227837e-05,
+      "loss": 0.2301,
+      "step": 212
+    },
+    {
+      "epoch": 0.14,
+      "learning_rate": 1.9414130940430347e-05,
+      "loss": 0.2319,
+      "step": 213
+    },
+    {
+      "epoch": 0.14,
+      "learning_rate": 1.9407121813710593e-05,
+      "loss": 0.2705,
+      "step": 214
+    },
+    {
+      "epoch": 0.14,
+      "learning_rate": 1.9400072289168474e-05,
+      "loss": 0.2713,
+      "step": 215
+    },
+    {
+      "epoch": 0.14,
+      "learning_rate": 1.9392982397077383e-05,
+      "loss": 0.2702,
+      "step": 216
+    },
+    {
+      "epoch": 0.14,
+      "learning_rate": 1.9385852167884057e-05,
+      "loss": 0.2878,
+      "step": 217
+    },
+    {
+      "epoch": 0.14,
+      "learning_rate": 1.9378681632208467e-05,
+      "loss": 0.2606,
+      "step": 218
+    },
+    {
+      "epoch": 0.14,
+      "learning_rate": 1.937147082084366e-05,
+      "loss": 0.2393,
+      "step": 219
+    },
+    {
+      "epoch": 0.14,
+      "learning_rate": 1.9364219764755655e-05,
+      "loss": 0.2547,
+      "step": 220
+    },
+    {
+      "epoch": 0.14,
+      "learning_rate": 1.9356928495083295e-05,
+      "loss": 0.2607,
+      "step": 221
+    },
+    {
+      "epoch": 0.14,
+      "learning_rate": 1.9349597043138114e-05,
+      "loss": 0.2707,
+      "step": 222
+    },
+    {
+      "epoch": 0.14,
+      "learning_rate": 1.93422254404042e-05,
+      "loss": 0.2461,
+      "step": 223
+    },
+    {
+      "epoch": 0.14,
+      "learning_rate": 1.9334813718538075e-05,
+      "loss": 0.218,
+      "step": 224
+    },
+    {
+      "epoch": 0.14,
+      "learning_rate": 1.9327361909368535e-05,
+      "loss": 0.2486,
+      "step": 225
+    },
+    {
+      "epoch": 0.14,
+      "learning_rate": 1.931987004489654e-05,
+      "loss": 0.2298,
+      "step": 226
+    },
+    {
+      "epoch": 0.15,
+      "learning_rate": 1.931233815729505e-05,
+      "loss": 0.2571,
+      "step": 227
+    },
+    {
+      "epoch": 0.15,
+      "learning_rate": 1.930476627890891e-05,
+      "loss": 0.2634,
+      "step": 228
+    },
+    {
+      "epoch": 0.15,
+      "learning_rate": 1.9297154442254692e-05,
+      "loss": 0.2523,
+      "step": 229
+    },
+    {
+      "epoch": 0.15,
+      "learning_rate": 1.9289502680020575e-05,
+      "loss": 0.2841,
+      "step": 230
+    },
+    {
+      "epoch": 0.15,
+      "learning_rate": 1.9281811025066184e-05,
+      "loss": 0.2435,
+      "step": 231
+    },
+    {
+      "epoch": 0.15,
+      "learning_rate": 1.9274079510422466e-05,
+      "loss": 0.2841,
+      "step": 232
+    },
+    {
+      "epoch": 0.15,
+      "learning_rate": 1.9266308169291533e-05,
+      "loss": 0.2403,
+      "step": 233
+    },
+    {
+      "epoch": 0.15,
+      "learning_rate": 1.9258497035046534e-05,
+      "loss": 0.2572,
+      "step": 234
+    },
+    {
+      "epoch": 0.15,
+      "learning_rate": 1.9250646141231503e-05,
+      "loss": 0.2547,
+      "step": 235
+    },
+    {
+      "epoch": 0.15,
+      "learning_rate": 1.9242755521561213e-05,
+      "loss": 0.2249,
+      "step": 236
+    },
+    {
+      "epoch": 0.15,
+      "learning_rate": 1.9234825209921045e-05,
+      "loss": 0.2607,
+      "step": 237
+    },
+    {
+      "epoch": 0.15,
+      "learning_rate": 1.9226855240366826e-05,
+      "loss": 0.2581,
+      "step": 238
+    },
+    {
+      "epoch": 0.15,
+      "learning_rate": 1.921884564712469e-05,
+      "loss": 0.2475,
+      "step": 239
+    },
+    {
+      "epoch": 0.15,
+      "learning_rate": 1.921079646459093e-05,
+      "loss": 0.2687,
+      "step": 240
+    },
+    {
+      "epoch": 0.15,
+      "learning_rate": 1.920270772733185e-05,
+      "loss": 0.2575,
+      "step": 241
+    },
+    {
+      "epoch": 0.15,
+      "learning_rate": 1.919457947008362e-05,
+      "loss": 0.2681,
+      "step": 242
+    },
+    {
+      "epoch": 0.16,
+      "learning_rate": 1.9186411727752124e-05,
+      "loss": 0.2305,
+      "step": 243
+    },
+    {
+      "epoch": 0.16,
+      "learning_rate": 1.917820453541281e-05,
+      "loss": 0.2634,
+      "step": 244
+    },
+    {
+      "epoch": 0.16,
+      "learning_rate": 1.9169957928310533e-05,
+      "loss": 0.2263,
+      "step": 245
+    },
+    {
+      "epoch": 0.16,
+      "learning_rate": 1.916167194185942e-05,
+      "loss": 0.2595,
+      "step": 246
+    },
+    {
+      "epoch": 0.16,
+      "learning_rate": 1.9153346611642706e-05,
+      "loss": 0.2615,
+      "step": 247
+    },
+    {
+      "epoch": 0.16,
+      "learning_rate": 1.914498197341258e-05,
+      "loss": 0.2743,
+      "step": 248
+    },
+    {
+      "epoch": 0.16,
+      "learning_rate": 1.9136578063090035e-05,
+      "loss": 0.2499,
+      "step": 249
+    },
+    {
+      "epoch": 0.16,
+      "learning_rate": 1.9128134916764715e-05,
+      "loss": 0.2591,
+      "step": 250
+    },
+    {
+      "epoch": 0.16,
+      "learning_rate": 1.911965257069476e-05,
+      "loss": 0.262,
+      "step": 251
+    },
+    {
+      "epoch": 0.16,
+      "learning_rate": 1.9111131061306644e-05,
+      "loss": 0.2711,
+      "step": 252
+    },
+    {
+      "epoch": 0.16,
+      "learning_rate": 1.9102570425195032e-05,
+      "loss": 0.2391,
+      "step": 253
+    },
+    {
+      "epoch": 0.16,
+      "learning_rate": 1.9093970699122606e-05,
+      "loss": 0.2858,
+      "step": 254
+    },
+    {
+      "epoch": 0.16,
+      "learning_rate": 1.908533192001992e-05,
+      "loss": 0.2455,
+      "step": 255
+    },
+    {
+      "epoch": 0.16,
+      "learning_rate": 1.907665412498523e-05,
+      "loss": 0.3186,
+      "step": 256
+    },
+    {
+      "epoch": 0.16,
+      "learning_rate": 1.9067937351284355e-05,
+      "loss": 0.1974,
+      "step": 257
+    },
+    {
+      "epoch": 0.17,
+      "learning_rate": 1.9059181636350493e-05,
+      "loss": 0.2504,
+      "step": 258
+    },
+    {
+      "epoch": 0.17,
+      "learning_rate": 1.9050387017784067e-05,
+      "loss": 0.2488,
+      "step": 259
+    },
+    {
+      "epoch": 0.17,
+      "learning_rate": 1.904155353335258e-05,
+      "loss": 0.229,
+      "step": 260
+    },
+    {
+      "epoch": 0.17,
+      "learning_rate": 1.903268122099043e-05,
+      "loss": 0.2179,
+      "step": 261
+    },
+    {
+      "epoch": 0.17,
+      "learning_rate": 1.9023770118798767e-05,
+      "loss": 0.2281,
+      "step": 262
+    },
+    {
+      "epoch": 0.17,
+      "learning_rate": 1.9014820265045304e-05,
+      "loss": 0.22,
+      "step": 263
+    },
+    {
+      "epoch": 0.17,
+      "learning_rate": 1.900583169816418e-05,
+      "loss": 0.2224,
+      "step": 264
+    },
+    {
+      "epoch": 0.17,
+      "learning_rate": 1.8996804456755783e-05,
+      "loss": 0.2158,
+      "step": 265
+    },
+    {
+      "epoch": 0.17,
+      "learning_rate": 1.8987738579586582e-05,
+      "loss": 0.2499,
+      "step": 266
+    },
+    {
+      "epoch": 0.17,
+      "learning_rate": 1.8978634105588963e-05,
+      "loss": 0.2398,
+      "step": 267
+    },
+    {
+      "epoch": 0.17,
+      "learning_rate": 1.8969491073861057e-05,
+      "loss": 0.2392,
+      "step": 268
+    },
+    {
+      "epoch": 0.17,
+      "learning_rate": 1.8960309523666583e-05,
+      "loss": 0.2504,
+      "step": 269
+    },
+    {
+      "epoch": 0.17,
+      "learning_rate": 1.8951089494434675e-05,
+      "loss": 0.2711,
+      "step": 270
+    },
+    {
+      "epoch": 0.17,
+      "learning_rate": 1.8941831025759705e-05,
+      "loss": 0.2602,
+      "step": 271
+    },
+    {
+      "epoch": 0.17,
+      "learning_rate": 1.8932534157401122e-05,
+      "loss": 0.2743,
+      "step": 272
+    },
+    {
+      "epoch": 0.17,
+      "learning_rate": 1.8923198929283275e-05,
+      "loss": 0.2628,
+      "step": 273
+    },
+    {
+      "epoch": 0.18,
+      "learning_rate": 1.891382538149525e-05,
+      "loss": 0.2044,
+      "step": 274
+    },
+    {
+      "epoch": 0.18,
+      "learning_rate": 1.8904413554290687e-05,
+      "loss": 0.2605,
+      "step": 275
+    },
+    {
+      "epoch": 0.18,
+      "learning_rate": 1.8894963488087618e-05,
+      "loss": 0.2448,
+      "step": 276
+    },
+    {
+      "epoch": 0.18,
+      "learning_rate": 1.888547522346828e-05,
+      "loss": 0.229,
+      "step": 277
+    },
+    {
+      "epoch": 0.18,
+      "learning_rate": 1.887594880117896e-05,
+      "loss": 0.259,
+      "step": 278
+    },
+    {
+      "epoch": 0.18,
+      "learning_rate": 1.88663842621298e-05,
+      "loss": 0.208,
+      "step": 279
+    },
+    {
+      "epoch": 0.18,
+      "learning_rate": 1.8856781647394634e-05,
+      "loss": 0.2415,
+      "step": 280
+    },
+    {
+      "epoch": 0.18,
+      "learning_rate": 1.8847140998210806e-05,
+      "loss": 0.2401,
+      "step": 281
+    },
+    {
+      "epoch": 0.18,
+      "learning_rate": 1.8837462355978996e-05,
+      "loss": 0.2542,
+      "step": 282
+    },
+    {
+      "epoch": 0.18,
+      "learning_rate": 1.8827745762263038e-05,
+      "loss": 0.2642,
+      "step": 283
+    },
+    {
+      "epoch": 0.18,
+      "learning_rate": 1.8817991258789745e-05,
+      "loss": 0.2363,
+      "step": 284
+    },
+    {
+      "epoch": 0.18,
+      "learning_rate": 1.8808198887448737e-05,
+      "loss": 0.2366,
+      "step": 285
+    },
+    {
+      "epoch": 0.18,
+      "learning_rate": 1.8798368690292243e-05,
+      "loss": 0.2245,
+      "step": 286
+    },
+    {
+      "epoch": 0.18,
+      "learning_rate": 1.8788500709534934e-05,
+      "loss": 0.2367,
+      "step": 287
+    },
+    {
+      "epoch": 0.18,
+      "learning_rate": 1.8778594987553738e-05,
+      "loss": 0.2259,
+      "step": 288
+    },
+    {
+      "epoch": 0.18,
+      "learning_rate": 1.8768651566887667e-05,
+      "loss": 0.2533,
+      "step": 289
+    },
+    {
+      "epoch": 0.19,
+      "learning_rate": 1.8758670490237606e-05,
+      "loss": 0.2644,
+      "step": 290
+    },
+    {
+      "epoch": 0.19,
+      "learning_rate": 1.8748651800466174e-05,
+      "loss": 0.2624,
+      "step": 291
+    },
+    {
+      "epoch": 0.19,
+      "learning_rate": 1.8738595540597494e-05,
+      "loss": 0.2167,
+      "step": 292
+    },
+    {
+      "epoch": 0.19,
+      "learning_rate": 1.8728501753817044e-05,
+      "loss": 0.2475,
+      "step": 293
+    },
+    {
+      "epoch": 0.19,
+      "learning_rate": 1.871837048347145e-05,
+      "loss": 0.256,
+      "step": 294
+    },
+    {
+      "epoch": 0.19,
+      "learning_rate": 1.8708201773068303e-05,
+      "loss": 0.2315,
+      "step": 295
+    },
+    {
+      "epoch": 0.19,
+      "learning_rate": 1.869799566627599e-05,
+      "loss": 0.2385,
+      "step": 296
+    },
+    {
+      "epoch": 0.19,
+      "learning_rate": 1.868775220692348e-05,
+      "loss": 0.2335,
+      "step": 297
+    },
+    {
+      "epoch": 0.19,
+      "learning_rate": 1.867747143900015e-05,
+      "loss": 0.2635,
+      "step": 298
+    },
+    {
+      "epoch": 0.19,
+      "learning_rate": 1.8667153406655606e-05,
+      "loss": 0.2691,
+      "step": 299
+    },
+    {
+      "epoch": 0.19,
+      "learning_rate": 1.865679815419946e-05,
+      "loss": 0.2322,
+      "step": 300
+    },
+    {
+      "epoch": 0.19,
+      "learning_rate": 1.864640572610118e-05,
+      "loss": 0.2417,
+      "step": 301
+    },
+    {
+      "epoch": 0.19,
+      "learning_rate": 1.8635976166989883e-05,
+      "loss": 0.2446,
+      "step": 302
+    },
+    {
+      "epoch": 0.19,
+      "learning_rate": 1.8625509521654122e-05,
+      "loss": 0.2533,
+      "step": 303
+    },
+    {
+      "epoch": 0.19,
+      "learning_rate": 1.8615005835041726e-05,
+      "loss": 0.2686,
+      "step": 304
+    },
+    {
+      "epoch": 0.2,
+      "learning_rate": 1.8604465152259595e-05,
+      "loss": 0.2723,
+      "step": 305
+    },
+    {
+      "epoch": 0.2,
+      "learning_rate": 1.8593887518573495e-05,
+      "loss": 0.2648,
+      "step": 306
+    },
+    {
+      "epoch": 0.2,
+      "learning_rate": 1.8583272979407885e-05,
+      "loss": 0.2354,
+      "step": 307
+    },
+    {
+      "epoch": 0.2,
+      "learning_rate": 1.85726215803457e-05,
+      "loss": 0.2491,
+      "step": 308
+    },
+    {
+      "epoch": 0.2,
+      "learning_rate": 1.8561933367128175e-05,
+      "loss": 0.1985,
+      "step": 309
+    },
+    {
+      "epoch": 0.2,
+      "learning_rate": 1.8551208385654634e-05,
+      "loss": 0.2527,
+      "step": 310
+    },
+    {
+      "epoch": 0.2,
+      "learning_rate": 1.8540446681982295e-05,
+      "loss": 0.2302,
+      "step": 311
+    },
+    {
+      "epoch": 0.2,
+      "learning_rate": 1.8529648302326084e-05,
+      "loss": 0.2198,
+      "step": 312
+    },
+    {
+      "epoch": 0.2,
+      "learning_rate": 1.851881329305842e-05,
+      "loss": 0.2186,
+      "step": 313
+    },
+    {
+      "epoch": 0.2,
+      "learning_rate": 1.8507941700709024e-05,
+      "loss": 0.2359,
+      "step": 314
+    },
+    {
+      "epoch": 0.2,
+      "learning_rate": 1.849703357196473e-05,
+      "loss": 0.2639,
+      "step": 315
+    },
+    {
+      "epoch": 0.2,
+      "learning_rate": 1.8486088953669255e-05,
+      "loss": 0.2281,
+      "step": 316
+    },
+    {
+      "epoch": 0.2,
+      "learning_rate": 1.8475107892823038e-05,
+      "loss": 0.2139,
+      "step": 317
+    },
+    {
+      "epoch": 0.2,
+      "learning_rate": 1.8464090436582998e-05,
+      "loss": 0.2318,
+      "step": 318
+    },
+    {
+      "epoch": 0.2,
+      "learning_rate": 1.8453036632262352e-05,
+      "loss": 0.2338,
+      "step": 319
+    },
+    {
+      "epoch": 0.2,
+      "learning_rate": 1.8441946527330424e-05,
+      "loss": 0.2498,
+      "step": 320
+    },
+    {
+      "epoch": 0.21,
+      "learning_rate": 1.8430820169412415e-05,
+      "loss": 0.2593,
+      "step": 321
+    },
+    {
+      "epoch": 0.21,
+      "learning_rate": 1.8419657606289204e-05,
+      "loss": 0.2376,
+      "step": 322
+    },
+    {
+      "epoch": 0.21,
+      "learning_rate": 1.840845888589717e-05,
+      "loss": 0.2662,
+      "step": 323
+    },
+    {
+      "epoch": 0.21,
+      "learning_rate": 1.8397224056327948e-05,
+      "loss": 0.1955,
+      "step": 324
+    },
+    {
+      "epoch": 0.21,
+      "learning_rate": 1.8385953165828242e-05,
+      "loss": 0.2079,
+      "step": 325
+    },
+    {
+      "epoch": 0.21,
+      "learning_rate": 1.837464626279962e-05,
+      "loss": 0.2327,
+      "step": 326
+    },
+    {
+      "epoch": 0.21,
+      "learning_rate": 1.8363303395798305e-05,
+      "loss": 0.2396,
+      "step": 327
+    },
+    {
+      "epoch": 0.21,
+      "learning_rate": 1.835192461353495e-05,
+      "loss": 0.2457,
+      "step": 328
+    },
+    {
+      "epoch": 0.21,
+      "learning_rate": 1.834050996487446e-05,
+      "loss": 0.2785,
+      "step": 329
+    },
+    {
+      "epoch": 0.21,
+      "learning_rate": 1.832905949883575e-05,
+      "loss": 0.242,
+      "step": 330
+    },
+    {
+      "epoch": 0.21,
+      "learning_rate": 1.8317573264591553e-05,
+      "loss": 0.2364,
+      "step": 331
+    },
+    {
+      "epoch": 0.21,
+      "learning_rate": 1.8306051311468204e-05,
+      "loss": 0.2462,
+      "step": 332
+    },
+    {
+      "epoch": 0.21,
+      "learning_rate": 1.8294493688945435e-05,
+      "loss": 0.2402,
+      "step": 333
+    },
+    {
+      "epoch": 0.21,
+      "learning_rate": 1.828290044665614e-05,
+      "loss": 0.2702,
+      "step": 334
+    },
+    {
+      "epoch": 0.21,
+      "learning_rate": 1.82712716343862e-05,
+      "loss": 0.2152,
+      "step": 335
+    },
+    {
+      "epoch": 0.21,
+      "learning_rate": 1.8259607302074232e-05,
+      "loss": 0.2179,
+      "step": 336
+    },
+    {
+      "epoch": 0.22,
+      "learning_rate": 1.8247907499811395e-05,
+      "loss": 0.2295,
+      "step": 337
+    },
+    {
+      "epoch": 0.22,
+      "learning_rate": 1.8236172277841166e-05,
+      "loss": 0.249,
+      "step": 338
+    },
+    {
+      "epoch": 0.22,
+      "learning_rate": 1.822440168655913e-05,
+      "loss": 0.2539,
+      "step": 339
+    },
+    {
+      "epoch": 0.22,
+      "learning_rate": 1.8212595776512767e-05,
+      "loss": 0.2497,
+      "step": 340
+    },
+    {
+      "epoch": 0.22,
+      "learning_rate": 1.8200754598401223e-05,
+      "loss": 0.2522,
+      "step": 341
+    },
+    {
+      "epoch": 0.22,
+      "learning_rate": 1.8188878203075097e-05,
+      "loss": 0.2584,
+      "step": 342
+    },
+    {
+      "epoch": 0.22,
+      "learning_rate": 1.8176966641536228e-05,
+      "loss": 0.2594,
+      "step": 343
+    },
+    {
+      "epoch": 0.22,
+      "learning_rate": 1.816501996493748e-05,
+      "loss": 0.2553,
+      "step": 344
+    },
+    {
+      "epoch": 0.22,
+      "learning_rate": 1.8153038224582493e-05,
+      "loss": 0.2443,
+      "step": 345
+    },
+    {
+      "epoch": 0.22,
+      "learning_rate": 1.814102147192551e-05,
+      "loss": 0.25,
+      "step": 346
+    },
+    {
+      "epoch": 0.22,
+      "learning_rate": 1.812896975857111e-05,
+      "loss": 0.2635,
+      "step": 347
+    },
+    {
+      "epoch": 0.22,
+      "learning_rate": 1.811688313627402e-05,
+      "loss": 0.2518,
+      "step": 348
+    },
+    {
+      "epoch": 0.22,
+      "learning_rate": 1.8104761656938872e-05,
+      "loss": 0.2859,
+      "step": 349
+    },
+    {
+      "epoch": 0.22,
+      "learning_rate": 1.8092605372619996e-05,
+      "loss": 0.2242,
+      "step": 350
+    },
+    {
+      "epoch": 0.22,
+      "learning_rate": 1.8080414335521174e-05,
+      "loss": 0.2297,
+      "step": 351
+    },
+    {
+      "epoch": 0.23,
+      "learning_rate": 1.8068188597995443e-05,
+      "loss": 0.2478,
+      "step": 352
+    },
+    {
+      "epoch": 0.23,
+      "learning_rate": 1.805592821254485e-05,
+      "loss": 0.2207,
+      "step": 353
+    },
+    {
+      "epoch": 0.23,
+      "learning_rate": 1.8043633231820237e-05,
+      "loss": 0.2502,
+      "step": 354
+    },
+    {
+      "epoch": 0.23,
+      "learning_rate": 1.803130370862101e-05,
+      "loss": 0.2357,
+      "step": 355
+    },
+    {
+      "epoch": 0.23,
+      "learning_rate": 1.801893969589492e-05,
+      "loss": 0.2094,
+      "step": 356
+    },
+    {
+      "epoch": 0.23,
+      "learning_rate": 1.800654124673781e-05,
+      "loss": 0.222,
+      "step": 357
+    },
+    {
+      "epoch": 0.23,
+      "learning_rate": 1.7994108414393434e-05,
+      "loss": 0.2668,
+      "step": 358
+    },
+    {
+      "epoch": 0.23,
+      "learning_rate": 1.798164125225318e-05,
+      "loss": 0.24,
+      "step": 359
+    },
+    {
+      "epoch": 0.23,
+      "learning_rate": 1.7969139813855867e-05,
+      "loss": 0.2398,
+      "step": 360
+    },
+    {
+      "epoch": 0.23,
+      "learning_rate": 1.795660415288751e-05,
+      "loss": 0.2647,
+      "step": 361
+    },
+    {
+      "epoch": 0.23,
+      "learning_rate": 1.7944034323181094e-05,
+      "loss": 0.268,
+      "step": 362
+    },
+    {
+      "epoch": 0.23,
+      "learning_rate": 1.793143037871633e-05,
+      "loss": 0.248,
+      "step": 363
+    },
+    {
+      "epoch": 0.23,
+      "learning_rate": 1.7918792373619433e-05,
+      "loss": 0.2303,
+      "step": 364
+    },
+    {
+      "epoch": 0.23,
+      "learning_rate": 1.7906120362162895e-05,
+      "loss": 0.2276,
+      "step": 365
+    },
+    {
+      "epoch": 0.23,
+      "learning_rate": 1.789341439876523e-05,
+      "loss": 0.26,
+      "step": 366
+    },
+    {
+      "epoch": 0.23,
+      "learning_rate": 1.788067453799077e-05,
+      "loss": 0.2601,
+      "step": 367
+    },
+    {
+      "epoch": 0.24,
+      "learning_rate": 1.78679008345494e-05,
+      "loss": 0.2431,
+      "step": 368
+    },
+    {
+      "epoch": 0.24,
+      "learning_rate": 1.785509334329636e-05,
+      "loss": 0.241,
+      "step": 369
+    },
+    {
+      "epoch": 0.24,
+      "learning_rate": 1.7842252119231965e-05,
+      "loss": 0.2344,
+      "step": 370
+    },
+    {
+      "epoch": 0.24,
+      "learning_rate": 1.78293772175014e-05,
+      "loss": 0.2241,
+      "step": 371
+    },
+    {
+      "epoch": 0.24,
+      "learning_rate": 1.781646869339449e-05,
+      "loss": 0.311,
+      "step": 372
+    },
+    {
+      "epoch": 0.24,
+      "learning_rate": 1.7803526602345415e-05,
+      "loss": 0.2659,
+      "step": 373
+    },
+    {
+      "epoch": 0.24,
+      "learning_rate": 1.779055099993254e-05,
+      "loss": 0.218,
+      "step": 374
+    },
+    {
+      "epoch": 0.24,
+      "learning_rate": 1.7777541941878116e-05,
+      "loss": 0.2445,
+      "step": 375
+    },
+    {
+      "epoch": 0.24,
+      "learning_rate": 1.7764499484048074e-05,
+      "loss": 0.2523,
+      "step": 376
+    },
+    {
+      "epoch": 0.24,
+      "learning_rate": 1.775142368245178e-05,
+      "loss": 0.2756,
+      "step": 377
+    },
+    {
+      "epoch": 0.24,
+      "learning_rate": 1.7738314593241782e-05,
+      "loss": 0.2236,
+      "step": 378
+    },
+    {
+      "epoch": 0.24,
+      "learning_rate": 1.772517227271359e-05,
+      "loss": 0.2294,
+      "step": 379
+    },
+    {
+      "epoch": 0.24,
+      "learning_rate": 1.771199677730541e-05,
+      "loss": 0.2555,
+      "step": 380
+    },
+    {
+      "epoch": 0.24,
+      "learning_rate": 1.7698788163597923e-05,
+      "loss": 0.2545,
+      "step": 381
+    },
+    {
+      "epoch": 0.24,
+      "learning_rate": 1.7685546488314028e-05,
+      "loss": 0.2406,
+      "step": 382
+    },
+    {
+      "epoch": 0.25,
+      "learning_rate": 1.7672271808318603e-05,
+      "loss": 0.2439,
+      "step": 383
+    },
+    {
+      "epoch": 0.25,
+      "learning_rate": 1.7658964180618263e-05,
+      "loss": 0.2511,
+      "step": 384
+    },
+    {
+      "epoch": 0.25,
+      "learning_rate": 1.7645623662361112e-05,
+      "loss": 0.218,
+      "step": 385
+    },
+    {
+      "epoch": 0.25,
+      "learning_rate": 1.7632250310836498e-05,
+      "loss": 0.2495,
+      "step": 386
+    },
+    {
+      "epoch": 0.25,
+      "learning_rate": 1.7618844183474775e-05,
+      "loss": 0.2575,
+      "step": 387
+    },
+    {
+      "epoch": 0.25,
+      "learning_rate": 1.7605405337847036e-05,
+      "loss": 0.2303,
+      "step": 388
+    },
+    {
+      "epoch": 0.25,
+      "learning_rate": 1.7591933831664887e-05,
+      "loss": 0.1903,
+      "step": 389
+    },
+    {
+      "epoch": 0.25,
+      "learning_rate": 1.7578429722780194e-05,
+      "loss": 0.2617,
+      "step": 390
+    },
+    {
+      "epoch": 0.25,
+      "learning_rate": 1.7564893069184824e-05,
+      "loss": 0.207,
+      "step": 391
+    },
+    {
+      "epoch": 0.25,
+      "learning_rate": 1.7551323929010407e-05,
+      "loss": 0.2628,
+      "step": 392
+    },
+    {
+      "epoch": 0.25,
+      "learning_rate": 1.753772236052809e-05,
+      "loss": 0.2236,
+      "step": 393
+    },
+    {
+      "epoch": 0.25,
+      "learning_rate": 1.7524088422148266e-05,
+      "loss": 0.235,
+      "step": 394
+    },
+    {
+      "epoch": 0.25,
+      "learning_rate": 1.7510422172420344e-05,
+      "loss": 0.247,
+      "step": 395
+    },
+    {
+      "epoch": 0.25,
+      "learning_rate": 1.7496723670032495e-05,
+      "loss": 0.2415,
+      "step": 396
+    },
+    {
+      "epoch": 0.25,
+      "learning_rate": 1.7482992973811384e-05,
+      "loss": 0.2101,
+      "step": 397
+    },
+    {
+      "epoch": 0.25,
+      "learning_rate": 1.746923014272194e-05,
+      "loss": 0.2305,
+      "step": 398
+    },
+    {
+      "epoch": 0.26,
+      "learning_rate": 1.745543523586709e-05,
+      "loss": 0.2231,
+      "step": 399
+    },
+    {
+      "epoch": 0.26,
+      "learning_rate": 1.7441608312487494e-05,
+      "loss": 0.2256,
+      "step": 400
+    },
+    {
+      "epoch": 0.26,
+      "learning_rate": 1.7427749431961324e-05,
+      "loss": 0.2608,
+      "step": 401
+    },
+    {
+      "epoch": 0.26,
+      "learning_rate": 1.7413858653803964e-05,
+      "loss": 0.2543,
+      "step": 402
+    },
+    {
+      "epoch": 0.26,
+      "learning_rate": 1.7399936037667807e-05,
+      "loss": 0.2283,
+      "step": 403
+    },
+    {
+      "epoch": 0.26,
+      "learning_rate": 1.7385981643341948e-05,
+      "loss": 0.2525,
+      "step": 404
+    },
+    {
+      "epoch": 0.26,
+      "learning_rate": 1.737199553075196e-05,
+      "loss": 0.2392,
+      "step": 405
+    },
+    {
+      "epoch": 0.26,
+      "learning_rate": 1.7357977759959626e-05,
+      "loss": 0.2519,
+      "step": 406
+    },
+    {
+      "epoch": 0.26,
+      "learning_rate": 1.7343928391162673e-05,
+      "loss": 0.2419,
+      "step": 407
+    },
+    {
+      "epoch": 0.26,
+      "learning_rate": 1.7329847484694533e-05,
+      "loss": 0.2276,
+      "step": 408
+    },
+    {
+      "epoch": 0.26,
+      "learning_rate": 1.7315735101024068e-05,
+      "loss": 0.2582,
+      "step": 409
+    },
+    {
+      "epoch": 0.26,
+      "learning_rate": 1.7301591300755317e-05,
+      "loss": 0.2585,
+      "step": 410
+    },
+    {
+      "epoch": 0.26,
+      "learning_rate": 1.7287416144627237e-05,
+      "loss": 0.2482,
+      "step": 411
+    },
+    {
+      "epoch": 0.26,
+      "learning_rate": 1.7273209693513427e-05,
+      "loss": 0.1936,
+      "step": 412
+    },
+    {
+      "epoch": 0.26,
+      "learning_rate": 1.725897200842189e-05,
+      "loss": 0.2266,
+      "step": 413
+    },
+    {
+      "epoch": 0.26,
+      "learning_rate": 1.7244703150494762e-05,
+      "loss": 0.2267,
+      "step": 414
+    },
+    {
+      "epoch": 0.27,
+      "learning_rate": 1.7230403181008032e-05,
+      "loss": 0.2376,
+      "step": 415
+    },
+    {
+      "epoch": 0.27,
+      "learning_rate": 1.721607216137132e-05,
+      "loss": 0.2547,
+      "step": 416
+    },
+    {
+      "epoch": 0.27,
+      "learning_rate": 1.7201710153127554e-05,
+      "loss": 0.2292,
+      "step": 417
+    },
+    {
+      "epoch": 0.27,
+      "learning_rate": 1.7187317217952765e-05,
+      "loss": 0.2251,
+      "step": 418
+    },
+    {
+      "epoch": 0.27,
+      "learning_rate": 1.7172893417655792e-05,
+      "loss": 0.2209,
+      "step": 419
+    },
+    {
+      "epoch": 0.27,
+      "learning_rate": 1.7158438814178005e-05,
+      "loss": 0.2237,
+      "step": 420
+    },
+    {
+      "epoch": 0.27,
+      "learning_rate": 1.714395346959308e-05,
+      "loss": 0.2141,
+      "step": 421
+    },
+    {
+      "epoch": 0.27,
+      "learning_rate": 1.7129437446106683e-05,
+      "loss": 0.2393,
+      "step": 422
+    },
+    {
+      "epoch": 0.27,
+      "learning_rate": 1.7114890806056245e-05,
+      "loss": 0.2021,
+      "step": 423
+    },
+    {
+      "epoch": 0.27,
+      "learning_rate": 1.7100313611910664e-05,
+      "loss": 0.2183,
+      "step": 424
+    },
+    {
+      "epoch": 0.27,
+      "learning_rate": 1.708570592627006e-05,
+      "loss": 0.2284,
+      "step": 425
+    },
+    {
+      "epoch": 0.27,
+      "learning_rate": 1.7071067811865477e-05,
+      "loss": 0.2401,
+      "step": 426
+    },
+    {
+      "epoch": 0.27,
+      "learning_rate": 1.705639933155866e-05,
+      "loss": 0.2771,
+      "step": 427
+    },
+    {
+      "epoch": 0.27,
+      "learning_rate": 1.704170054834173e-05,
+      "loss": 0.2381,
+      "step": 428
+    },
+    {
+      "epoch": 0.27,
+      "learning_rate": 1.702697152533695e-05,
+      "loss": 0.2691,
+      "step": 429
+    },
+    {
+      "epoch": 0.28,
+      "learning_rate": 1.7012212325796463e-05,
+      "loss": 0.2444,
+      "step": 430
+    },
+    {
+      "epoch": 0.28,
+      "learning_rate": 1.6997423013101967e-05,
+      "loss": 0.2373,
+      "step": 431
+    },
+    {
+      "epoch": 0.28,
+      "learning_rate": 1.6982603650764508e-05,
+      "loss": 0.2295,
+      "step": 432
+    },
+    {
+      "epoch": 0.28,
+      "learning_rate": 1.6967754302424153e-05,
+      "loss": 0.2136,
+      "step": 433
+    },
+    {
+      "epoch": 0.28,
+      "learning_rate": 1.6952875031849757e-05,
+      "loss": 0.2635,
+      "step": 434
+    },
+    {
+      "epoch": 0.28,
+      "learning_rate": 1.6937965902938667e-05,
+      "loss": 0.2449,
+      "step": 435
+    },
+    {
+      "epoch": 0.28,
+      "learning_rate": 1.6923026979716455e-05,
+      "loss": 0.237,
+      "step": 436
+    },
+    {
+      "epoch": 0.28,
+      "learning_rate": 1.6908058326336634e-05,
+      "loss": 0.2688,
+      "step": 437
+    },
+    {
+      "epoch": 0.28,
+      "learning_rate": 1.6893060007080406e-05,
+      "loss": 0.2362,
+      "step": 438
+    },
+    {
+      "epoch": 0.28,
+      "learning_rate": 1.687803208635635e-05,
+      "loss": 0.2523,
+      "step": 439
+    },
+    {
+      "epoch": 0.28,
+      "learning_rate": 1.6862974628700183e-05,
+      "loss": 0.2483,
+      "step": 440
+    },
+    {
+      "epoch": 0.28,
+      "learning_rate": 1.6847887698774446e-05,
+      "loss": 0.2382,
+      "step": 441
+    },
+    {
+      "epoch": 0.28,
+      "learning_rate": 1.6832771361368263e-05,
+      "loss": 0.246,
+      "step": 442
+    },
+    {
+      "epoch": 0.28,
+      "learning_rate": 1.6817625681397035e-05,
+      "loss": 0.2195,
+      "step": 443
+    },
+    {
+      "epoch": 0.28,
+      "learning_rate": 1.680245072390218e-05,
+      "loss": 0.2038,
+      "step": 444
+    },
+    {
+      "epoch": 0.28,
+      "learning_rate": 1.678724655405083e-05,
+      "loss": 0.2073,
+      "step": 445
+    },
+    {
+      "epoch": 0.29,
+      "learning_rate": 1.6772013237135585e-05,
+      "loss": 0.2551,
+      "step": 446
+    },
+    {
+      "epoch": 0.29,
+      "learning_rate": 1.6756750838574197e-05,
+      "loss": 0.2466,
+      "step": 447
+    },
+    {
+      "epoch": 0.29,
+      "learning_rate": 1.674145942390932e-05,
+      "loss": 0.2268,
+      "step": 448
+    },
+    {
+      "epoch": 0.29,
+      "learning_rate": 1.6726139058808206e-05,
+      "loss": 0.2355,
+      "step": 449
+    },
+    {
+      "epoch": 0.29,
+      "learning_rate": 1.671078980906242e-05,
+      "loss": 0.2134,
+      "step": 450
+    },
+    {
+      "epoch": 0.29,
+      "learning_rate": 1.66954117405876e-05,
+      "loss": 0.2047,
+      "step": 451
+    },
+    {
+      "epoch": 0.29,
+      "learning_rate": 1.668000491942312e-05,
+      "loss": 0.2211,
+      "step": 452
+    },
+    {
+      "epoch": 0.29,
+      "learning_rate": 1.6664569411731828e-05,
+      "loss": 0.2305,
+      "step": 453
+    },
+    {
+      "epoch": 0.29,
+      "learning_rate": 1.664910528379977e-05,
+      "loss": 0.2274,
+      "step": 454
+    },
+    {
+      "epoch": 0.29,
+      "learning_rate": 1.66336126020359e-05,
+      "loss": 0.2109,
+      "step": 455
+    },
+    {
+      "epoch": 0.29,
+      "learning_rate": 1.661809143297179e-05,
+      "loss": 0.2421,
+      "step": 456
+    },
+    {
+      "epoch": 0.29,
+      "learning_rate": 1.6602541843261345e-05,
+      "loss": 0.2554,
+      "step": 457
+    },
+    {
+      "epoch": 0.29,
+      "learning_rate": 1.6586963899680525e-05,
+      "loss": 0.2698,
+      "step": 458
+    },
+    {
+      "epoch": 0.29,
+      "learning_rate": 1.6571357669127047e-05,
+      "loss": 0.2775,
+      "step": 459
+    },
+    {
+      "epoch": 0.29,
+      "learning_rate": 1.6555723218620113e-05,
+      "loss": 0.2345,
+      "step": 460
+    },
+    {
+      "epoch": 0.29,
+      "learning_rate": 1.6540060615300096e-05,
+      "loss": 0.2279,
+      "step": 461
+    },
+    {
+      "epoch": 0.3,
+      "learning_rate": 1.6524369926428282e-05,
+      "loss": 0.2553,
+      "step": 462
+    },
+    {
+      "epoch": 0.3,
+      "learning_rate": 1.650865121938656e-05,
+      "loss": 0.266,
+      "step": 463
+    },
+    {
+      "epoch": 0.3,
+      "learning_rate": 1.6492904561677146e-05,
+      "loss": 0.2451,
+      "step": 464
+    },
+    {
+      "epoch": 0.3,
+      "learning_rate": 1.6477130020922277e-05,
+      "loss": 0.2501,
+      "step": 465
+    },
+    {
+      "epoch": 0.3,
+      "learning_rate": 1.6461327664863948e-05,
+      "loss": 0.2129,
+      "step": 466
+    },
+    {
+      "epoch": 0.3,
+      "learning_rate": 1.644549756136358e-05,
+      "loss": 0.2137,
+      "step": 467
+    },
+    {
+      "epoch": 0.3,
+      "learning_rate": 1.6429639778401777e-05,
+      "loss": 0.2337,
+      "step": 468
+    },
+    {
+      "epoch": 0.3,
+      "learning_rate": 1.6413754384077992e-05,
+      "loss": 0.1939,
+      "step": 469
+    },
+    {
+      "epoch": 0.3,
+      "learning_rate": 1.639784144661025e-05,
+      "loss": 0.2282,
+      "step": 470
+    },
+    {
+      "epoch": 0.3,
+      "learning_rate": 1.6381901034334876e-05,
+      "loss": 0.2388,
+      "step": 471
+    },
+    {
+      "epoch": 0.3,
+      "learning_rate": 1.636593321570616e-05,
+      "loss": 0.2156,
+      "step": 472
+    },
+    {
+      "epoch": 0.3,
+      "learning_rate": 1.6349938059296106e-05,
+      "loss": 0.2639,
+      "step": 473
+    },
+    {
+      "epoch": 0.3,
+      "learning_rate": 1.63339156337941e-05,
+      "loss": 0.2228,
+      "step": 474
+    },
+    {
+      "epoch": 0.3,
+      "learning_rate": 1.631786600800664e-05,
+      "loss": 0.2268,
+      "step": 475
+    },
+    {
+      "epoch": 0.3,
+      "learning_rate": 1.630178925085703e-05,
+      "loss": 0.2439,
+      "step": 476
+    },
+    {
+      "epoch": 0.31,
+      "learning_rate": 1.6285685431385096e-05,
+      "loss": 0.2887,
+      "step": 477
+    },
+    {
+      "epoch": 0.31,
+      "learning_rate": 1.6269554618746867e-05,
+      "loss": 0.2329,
+      "step": 478
+    },
+    {
+      "epoch": 0.31,
+      "learning_rate": 1.6253396882214293e-05,
+      "loss": 0.2368,
+      "step": 479
+    },
+    {
+      "epoch": 0.31,
+      "learning_rate": 1.6237212291174955e-05,
+      "loss": 0.257,
+      "step": 480
+    },
+    {
+      "epoch": 0.31,
+      "learning_rate": 1.6221000915131746e-05,
+      "loss": 0.2293,
+      "step": 481
+    },
+    {
+      "epoch": 0.31,
+      "learning_rate": 1.6204762823702594e-05,
+      "loss": 0.194,
+      "step": 482
+    },
+    {
+      "epoch": 0.31,
+      "learning_rate": 1.6188498086620148e-05,
+      "loss": 0.2542,
+      "step": 483
+    },
+    {
+      "epoch": 0.31,
+      "learning_rate": 1.6172206773731486e-05,
+      "loss": 0.2649,
+      "step": 484
+    },
+    {
+      "epoch": 0.31,
+      "learning_rate": 1.615588895499781e-05,
+      "loss": 0.2132,
+      "step": 485
+    },
+    {
+      "epoch": 0.31,
+      "learning_rate": 1.6139544700494147e-05,
+      "loss": 0.2238,
+      "step": 486
+    },
+    {
+      "epoch": 0.31,
+      "learning_rate": 1.6123174080409055e-05,
+      "loss": 0.1704,
+      "step": 487
+    },
+    {
+      "epoch": 0.31,
+      "learning_rate": 1.610677716504432e-05,
+      "loss": 0.2287,
+      "step": 488
+    },
+    {
+      "epoch": 0.31,
+      "learning_rate": 1.609035402481463e-05,
+      "loss": 0.2062,
+      "step": 489
+    },
+    {
+      "epoch": 0.31,
+      "learning_rate": 1.6073904730247317e-05,
+      "loss": 0.2217,
+      "step": 490
+    },
+    {
+      "epoch": 0.31,
+      "learning_rate": 1.6057429351982016e-05,
+      "loss": 0.2132,
+      "step": 491
+    },
+    {
+      "epoch": 0.31,
+      "learning_rate": 1.6040927960770378e-05,
+      "loss": 0.2355,
+      "step": 492
+    },
+    {
+      "epoch": 0.32,
+      "learning_rate": 1.602440062747576e-05,
+      "loss": 0.2316,
+      "step": 493
+    },
+    {
+      "epoch": 0.32,
+      "learning_rate": 1.6007847423072938e-05,
+      "loss": 0.2277,
+      "step": 494
+    },
+    {
+      "epoch": 0.32,
+      "learning_rate": 1.5991268418647772e-05,
+      "loss": 0.2483,
+      "step": 495
+    },
+    {
+      "epoch": 0.32,
+      "learning_rate": 1.5974663685396926e-05,
+      "loss": 0.2538,
+      "step": 496
+    },
+    {
+      "epoch": 0.32,
+      "learning_rate": 1.595803329462755e-05,
+      "loss": 0.2134,
+      "step": 497
+    },
+    {
+      "epoch": 0.32,
+      "learning_rate": 1.5941377317756983e-05,
+      "loss": 0.255,
+      "step": 498
+    },
+    {
+      "epoch": 0.32,
+      "learning_rate": 1.5924695826312434e-05,
+      "loss": 0.2602,
+      "step": 499
+    },
+    {
+      "epoch": 0.32,
+      "learning_rate": 1.5907988891930686e-05,
+      "loss": 0.2518,
+      "step": 500
+    },
+    {
+      "epoch": 0.32,
+      "learning_rate": 1.5891256586357783e-05,
+      "loss": 0.2868,
+      "step": 501
+    },
+    {
+      "epoch": 0.32,
+      "learning_rate": 1.587449898144872e-05,
+      "loss": 0.2131,
+      "step": 502
+    },
+    {
+      "epoch": 0.32,
+      "learning_rate": 1.585771614916714e-05,
+      "loss": 0.2458,
+      "step": 503
+    },
+    {
+      "epoch": 0.32,
+      "learning_rate": 1.584090816158502e-05,
+      "loss": 0.2525,
+      "step": 504
+    },
+    {
+      "epoch": 0.32,
+      "learning_rate": 1.5824075090882364e-05,
+      "loss": 0.2321,
+      "step": 505
+    },
+    {
+      "epoch": 0.32,
+      "learning_rate": 1.5807217009346893e-05,
+      "loss": 0.2179,
+      "step": 506
+    },
+    {
+      "epoch": 0.32,
+      "learning_rate": 1.5790333989373738e-05,
+      "loss": 0.2326,
+      "step": 507
+    },
+    {
+      "epoch": 0.33,
+      "learning_rate": 1.577342610346512e-05,
+      "loss": 0.2215,
+      "step": 508
+    },
+    {
+      "epoch": 0.33,
+      "learning_rate": 1.5756493424230046e-05,
+      "loss": 0.2386,
+      "step": 509
+    },
+    {
+      "epoch": 0.33,
+      "learning_rate": 1.5739536024383994e-05,
+      "loss": 0.2419,
+      "step": 510
+    },
+    {
+      "epoch": 0.33,
+      "learning_rate": 1.5722553976748605e-05,
+      "loss": 0.2408,
+      "step": 511
+    },
+    {
+      "epoch": 0.33,
+      "learning_rate": 1.570554735425136e-05,
+      "loss": 0.2413,
+      "step": 512
+    },
+    {
+      "epoch": 0.33,
+      "learning_rate": 1.5688516229925282e-05,
+      "loss": 0.2191,
+      "step": 513
+    },
+    {
+      "epoch": 0.33,
+      "learning_rate": 1.5671460676908617e-05,
+      "loss": 0.2226,
+      "step": 514
+    },
+    {
+      "epoch": 0.33,
+      "learning_rate": 1.5654380768444498e-05,
+      "loss": 0.1892,
+      "step": 515
+    },
+    {
+      "epoch": 0.33,
+      "learning_rate": 1.5637276577880677e-05,
+      "loss": 0.2603,
+      "step": 516
+    },
+    {
+      "epoch": 0.33,
+      "learning_rate": 1.562014817866916e-05,
+      "loss": 0.2275,
+      "step": 517
+    },
+    {
+      "epoch": 0.33,
+      "learning_rate": 1.5602995644365924e-05,
+      "loss": 0.2707,
+      "step": 518
+    },
+    {
+      "epoch": 0.33,
+      "learning_rate": 1.55858190486306e-05,
+      "loss": 0.2495,
+      "step": 519
+    },
+    {
+      "epoch": 0.33,
+      "learning_rate": 1.5568618465226125e-05,
+      "loss": 0.247,
+      "step": 520
+    },
+    {
+      "epoch": 0.33,
+      "learning_rate": 1.555139396801847e-05,
+      "loss": 0.2542,
+      "step": 521
+    },
+    {
+      "epoch": 0.33,
+      "learning_rate": 1.5534145630976293e-05,
+      "loss": 0.2213,
+      "step": 522
+    },
+    {
+      "epoch": 0.33,
+      "learning_rate": 1.551687352817063e-05,
+      "loss": 0.1998,
+      "step": 523
+    },
+    {
+      "epoch": 0.34,
+      "learning_rate": 1.5499577733774573e-05,
+      "loss": 0.2195,
+      "step": 524
+    },
+    {
+      "epoch": 0.34,
+      "learning_rate": 1.548225832206296e-05,
+      "loss": 0.2062,
+      "step": 525
+    },
+    {
+      "epoch": 0.34,
+      "learning_rate": 1.546491536741205e-05,
+      "loss": 0.2168,
+      "step": 526
+    },
+    {
+      "epoch": 0.34,
+      "learning_rate": 1.5447548944299203e-05,
+      "loss": 0.276,
+      "step": 527
+    },
+    {
+      "epoch": 0.34,
+      "learning_rate": 1.5430159127302558e-05,
+      "loss": 0.2524,
+      "step": 528
+    },
+    {
+      "epoch": 0.34,
+      "learning_rate": 1.5412745991100724e-05,
+      "loss": 0.2595,
+      "step": 529
+    },
+    {
+      "epoch": 0.34,
+      "learning_rate": 1.539530961047245e-05,
+      "loss": 0.2394,
+      "step": 530
+    },
+    {
+      "epoch": 0.34,
+      "learning_rate": 1.53778500602963e-05,
+      "loss": 0.2277,
+      "step": 531
+    },
+    {
+      "epoch": 0.34,
+      "learning_rate": 1.536036741555034e-05,
+      "loss": 0.2232,
+      "step": 532
+    },
+    {
+      "epoch": 0.34,
+      "learning_rate": 1.5342861751311816e-05,
+      "loss": 0.2168,
+      "step": 533
+    },
+    {
+      "epoch": 0.34,
+      "learning_rate": 1.532533314275683e-05,
+      "loss": 0.1988,
+      "step": 534
+    },
+    {
+      "epoch": 0.34,
+      "learning_rate": 1.5307781665160005e-05,
+      "loss": 0.2384,
+      "step": 535
+    },
+    {
+      "epoch": 0.34,
+      "learning_rate": 1.529020739389418e-05,
+      "loss": 0.2167,
+      "step": 536
+    },
+    {
+      "epoch": 0.34,
+      "learning_rate": 1.527261040443008e-05,
+      "loss": 0.2601,
+      "step": 537
+    },
+    {
+      "epoch": 0.34,
+      "learning_rate": 1.5254990772335996e-05,
+      "loss": 0.2237,
+      "step": 538
+    },
+    {
+      "epoch": 0.34,
+      "learning_rate": 1.523734857327744e-05,
+      "loss": 0.2338,
+      "step": 539
+    },
+    {
+      "epoch": 0.35,
+      "learning_rate": 1.5219683883016844e-05,
+      "loss": 0.1802,
+      "step": 540
+    },
+    {
+      "epoch": 0.35,
+      "learning_rate": 1.5201996777413231e-05,
+      "loss": 0.2518,
+      "step": 541
+    },
+    {
+      "epoch": 0.35,
+      "learning_rate": 1.5184287332421872e-05,
+      "loss": 0.2175,
+      "step": 542
+    },
+    {
+      "epoch": 0.35,
+      "learning_rate": 1.5166555624093986e-05,
+      "loss": 0.2648,
+      "step": 543
+    },
+    {
+      "epoch": 0.35,
+      "learning_rate": 1.5148801728576388e-05,
+      "loss": 0.197,
+      "step": 544
+    },
+    {
+      "epoch": 0.35,
+      "learning_rate": 1.5131025722111176e-05,
+      "loss": 0.2243,
+      "step": 545
+    },
+    {
+      "epoch": 0.35,
+      "learning_rate": 1.5113227681035408e-05,
+      "loss": 0.2516,
+      "step": 546
+    },
+    {
+      "epoch": 0.35,
+      "learning_rate": 1.5095407681780753e-05,
+      "loss": 0.2171,
+      "step": 547
+    },
+    {
+      "epoch": 0.35,
+      "learning_rate": 1.5077565800873193e-05,
+      "loss": 0.2269,
+      "step": 548
+    },
+    {
+      "epoch": 0.35,
+      "learning_rate": 1.505970211493267e-05,
+      "loss": 0.1996,
+      "step": 549
+    },
+    {
+      "epoch": 0.35,
+      "learning_rate": 1.5041816700672762e-05,
+      "loss": 0.2273,
+      "step": 550
+    },
+    {
+      "epoch": 0.35,
+      "learning_rate": 1.5023909634900362e-05,
+      "loss": 0.2173,
+      "step": 551
+    },
+    {
+      "epoch": 0.35,
+      "learning_rate": 1.5005980994515346e-05,
+      "loss": 0.2094,
+      "step": 552
+    },
+    {
+      "epoch": 0.35,
+      "learning_rate": 1.4988030856510231e-05,
+      "loss": 0.2285,
+      "step": 553
+    },
+    {
+      "epoch": 0.35,
+      "learning_rate": 1.497005929796986e-05,
+      "loss": 0.2295,
+      "step": 554
+    },
+    {
+      "epoch": 0.36,
+      "learning_rate": 1.4952066396071064e-05,
+      "loss": 0.2757,
+      "step": 555
+    },
+    {
+      "epoch": 0.36,
+      "learning_rate": 1.4934052228082324e-05,
+      "loss": 0.2609,
+      "step": 556
+    },
+    {
+      "epoch": 0.36,
+      "learning_rate": 1.491601687136345e-05,
+      "loss": 0.2253,
+      "step": 557
+    },
+    {
+      "epoch": 0.36,
+      "learning_rate": 1.489796040336525e-05,
+      "loss": 0.2147,
+      "step": 558
+    },
+    {
+      "epoch": 0.36,
+      "learning_rate": 1.4879882901629181e-05,
+      "loss": 0.2234,
+      "step": 559
+    },
+    {
+      "epoch": 0.36,
+      "learning_rate": 1.4861784443787034e-05,
+      "loss": 0.2183,
+      "step": 560
+    },
+    {
+      "epoch": 0.36,
+      "learning_rate": 1.4843665107560596e-05,
+      "loss": 0.2091,
+      "step": 561
+    },
+    {
+      "epoch": 0.36,
+      "learning_rate": 1.482552497076131e-05,
+      "loss": 0.2258,
+      "step": 562
+    },
+    {
+      "epoch": 0.36,
+      "learning_rate": 1.4807364111289941e-05,
+      "loss": 0.2308,
+      "step": 563
+    },
+    {
+      "epoch": 0.36,
+      "learning_rate": 1.4789182607136252e-05,
+      "loss": 0.2189,
+      "step": 564
+    },
+    {
+      "epoch": 0.36,
+      "learning_rate": 1.477098053637866e-05,
+      "loss": 0.2005,
+      "step": 565
+    },
+    {
+      "epoch": 0.36,
+      "learning_rate": 1.4752757977183901e-05,
+      "loss": 0.2182,
+      "step": 566
+    },
+    {
+      "epoch": 0.36,
+      "learning_rate": 1.4734515007806698e-05,
+      "loss": 0.2496,
+      "step": 567
+    },
+    {
+      "epoch": 0.36,
+      "learning_rate": 1.4716251706589426e-05,
+      "loss": 0.2298,
+      "step": 568
+    },
+    {
+      "epoch": 0.36,
+      "learning_rate": 1.4697968151961763e-05,
+      "loss": 0.2479,
+      "step": 569
+    },
+    {
+      "epoch": 0.36,
+      "learning_rate": 1.4679664422440377e-05,
+      "loss": 0.2277,
+      "step": 570
+    },
+    {
+      "epoch": 0.37,
+      "learning_rate": 1.4661340596628564e-05,
+      "loss": 0.1948,
+      "step": 571
+    },
+    {
+      "epoch": 0.37,
+      "learning_rate": 1.4642996753215923e-05,
+      "loss": 0.2466,
+      "step": 572
+    },
+    {
+      "epoch": 0.37,
+      "learning_rate": 1.4624632970978021e-05,
+      "loss": 0.2049,
+      "step": 573
+    },
+    {
+      "epoch": 0.37,
+      "learning_rate": 1.4606249328776048e-05,
+      "loss": 0.2266,
+      "step": 574
+    },
+    {
+      "epoch": 0.37,
+      "learning_rate": 1.4587845905556477e-05,
+      "loss": 0.2467,
+      "step": 575
+    },
+    {
+      "epoch": 0.37,
+      "learning_rate": 1.456942278035073e-05,
+      "loss": 0.2374,
+      "step": 576
+    },
+    {
+      "epoch": 0.37,
+      "learning_rate": 1.4550980032274842e-05,
+      "loss": 0.1895,
+      "step": 577
+    },
+    {
+      "epoch": 0.37,
+      "learning_rate": 1.4532517740529108e-05,
+      "loss": 0.2576,
+      "step": 578
+    },
+    {
+      "epoch": 0.37,
+      "learning_rate": 1.4514035984397759e-05,
+      "loss": 0.2218,
+      "step": 579
+    },
+    {
+      "epoch": 0.37,
+      "learning_rate": 1.4495534843248606e-05,
+      "loss": 0.2028,
+      "step": 580
+    },
+    {
+      "epoch": 0.37,
+      "learning_rate": 1.4477014396532711e-05,
+      "loss": 0.2227,
+      "step": 581
+    },
+    {
+      "epoch": 0.37,
+      "learning_rate": 1.4458474723784045e-05,
+      "loss": 0.1985,
+      "step": 582
+    },
+    {
+      "epoch": 0.37,
+      "learning_rate": 1.4439915904619136e-05,
+      "loss": 0.2437,
+      "step": 583
+    },
+    {
+      "epoch": 0.37,
+      "learning_rate": 1.4421338018736732e-05,
+      "loss": 0.2238,
+      "step": 584
+    },
+    {
+      "epoch": 0.37,
+      "learning_rate": 1.4402741145917475e-05,
+      "loss": 0.2397,
+      "step": 585
+    },
+    {
+      "epoch": 0.37,
+      "learning_rate": 1.4384125366023533e-05,
+      "loss": 0.2452,
+      "step": 586
+    },
+    {
+      "epoch": 0.38,
+      "learning_rate": 1.4365490758998268e-05,
+      "loss": 0.252,
+      "step": 587
+    },
+    {
+      "epoch": 0.38,
+      "learning_rate": 1.4346837404865897e-05,
+      "loss": 0.2328,
+      "step": 588
+    },
+    {
+      "epoch": 0.38,
+      "learning_rate": 1.4328165383731145e-05,
+      "loss": 0.2365,
+      "step": 589
+    },
+    {
+      "epoch": 0.38,
+      "learning_rate": 1.430947477577889e-05,
+      "loss": 0.2117,
+      "step": 590
+    },
+    {
+      "epoch": 0.38,
+      "learning_rate": 1.429076566127385e-05,
+      "loss": 0.2335,
+      "step": 591
+    },
+    {
+      "epoch": 0.38,
+      "learning_rate": 1.4272038120560199e-05,
+      "loss": 0.2048,
+      "step": 592
+    },
+    {
+      "epoch": 0.38,
+      "learning_rate": 1.4253292234061237e-05,
+      "loss": 0.2172,
+      "step": 593
+    },
+    {
+      "epoch": 0.38,
+      "learning_rate": 1.4234528082279072e-05,
+      "loss": 0.2261,
+      "step": 594
+    },
+    {
+      "epoch": 0.38,
+      "learning_rate": 1.4215745745794224e-05,
+      "loss": 0.2205,
+      "step": 595
+    },
+    {
+      "epoch": 0.38,
+      "learning_rate": 1.4196945305265322e-05,
+      "loss": 0.2198,
+      "step": 596
+    },
+    {
+      "epoch": 0.38,
+      "learning_rate": 1.4178126841428732e-05,
+      "loss": 0.2238,
+      "step": 597
+    },
+    {
+      "epoch": 0.38,
+      "learning_rate": 1.415929043509822e-05,
+      "loss": 0.2063,
+      "step": 598
+    },
+    {
+      "epoch": 0.38,
+      "learning_rate": 1.4140436167164611e-05,
+      "loss": 0.196,
+      "step": 599
+    },
+    {
+      "epoch": 0.38,
+      "learning_rate": 1.4121564118595426e-05,
+      "loss": 0.2163,
+      "step": 600
+    },
+    {
+      "epoch": 0.38,
+      "learning_rate": 1.4102674370434549e-05,
+      "loss": 0.2161,
+      "step": 601
+    },
+    {
+      "epoch": 0.39,
+      "learning_rate": 1.4083767003801867e-05,
+      "loss": 0.2296,
+      "step": 602
+    },
+    {
+      "epoch": 0.39,
+      "learning_rate": 1.4064842099892934e-05,
+      "loss": 0.2194,
+      "step": 603
+    },
+    {
+      "epoch": 0.39,
+      "learning_rate": 1.4045899739978612e-05,
+      "loss": 0.2365,
+      "step": 604
+    },
+    {
+      "epoch": 0.39,
+      "learning_rate": 1.4026940005404726e-05,
+      "loss": 0.2224,
+      "step": 605
+    },
+    {
+      "epoch": 0.39,
+      "learning_rate": 1.4007962977591719e-05,
+      "loss": 0.2113,
+      "step": 606
+    },
+    {
+      "epoch": 0.39,
+      "learning_rate": 1.3988968738034285e-05,
+      "loss": 0.2814,
+      "step": 607
+    },
+    {
+      "epoch": 0.39,
+      "learning_rate": 1.3969957368301048e-05,
+      "loss": 0.2484,
+      "step": 608
+    },
+    {
+      "epoch": 0.39,
+      "learning_rate": 1.3950928950034189e-05,
+      "loss": 0.2132,
+      "step": 609
+    },
+    {
+      "epoch": 0.39,
+      "learning_rate": 1.3931883564949098e-05,
+      "loss": 0.2293,
+      "step": 610
+    },
+    {
+      "epoch": 0.39,
+      "learning_rate": 1.3912821294834033e-05,
+      "loss": 0.2233,
+      "step": 611
+    },
+    {
+      "epoch": 0.39,
+      "learning_rate": 1.389374222154976e-05,
+      "loss": 0.2398,
+      "step": 612
+    },
+    {
+      "epoch": 0.39,
+      "learning_rate": 1.3874646427029204e-05,
+      "loss": 0.2273,
+      "step": 613
+    },
+    {
+      "epoch": 0.39,
+      "learning_rate": 1.3855533993277097e-05,
+      "loss": 0.2212,
+      "step": 614
+    },
+    {
+      "epoch": 0.39,
+      "learning_rate": 1.3836405002369631e-05,
+      "loss": 0.2562,
+      "step": 615
+    },
+    {
+      "epoch": 0.39,
+      "learning_rate": 1.3817259536454097e-05,
+      "loss": 0.2157,
+      "step": 616
+    },
+    {
+      "epoch": 0.39,
+      "learning_rate": 1.3798097677748535e-05,
+      "loss": 0.2266,
+      "step": 617
+    },
+    {
+      "epoch": 0.4,
+      "learning_rate": 1.377891950854138e-05,
+      "loss": 0.2002,
+      "step": 618
+    },
+    {
+      "epoch": 0.4,
+      "learning_rate": 1.3759725111191118e-05,
+      "loss": 0.2204,
+      "step": 619
+    },
+    {
+      "epoch": 0.4,
+      "learning_rate": 1.3740514568125923e-05,
+      "loss": 0.2347,
+      "step": 620
+    },
+    {
+      "epoch": 0.4,
+      "learning_rate": 1.3721287961843298e-05,
+      "loss": 0.2257,
+      "step": 621
+    },
+    {
+      "epoch": 0.4,
+      "learning_rate": 1.3702045374909735e-05,
+      "loss": 0.2362,
+      "step": 622
+    },
+    {
+      "epoch": 0.4,
+      "learning_rate": 1.3682786889960354e-05,
+      "loss": 0.22,
+      "step": 623
+    },
+    {
+      "epoch": 0.4,
+      "learning_rate": 1.3663512589698536e-05,
+      "loss": 0.2376,
+      "step": 624
+    },
+    {
+      "epoch": 0.4,
+      "learning_rate": 1.3644222556895591e-05,
+      "loss": 0.2162,
+      "step": 625
+    },
+    {
+      "epoch": 0.4,
+      "learning_rate": 1.3624916874390395e-05,
+      "loss": 0.2397,
+      "step": 626
+    },
+    {
+      "epoch": 0.4,
+      "learning_rate": 1.3605595625089006e-05,
+      "loss": 0.2476,
+      "step": 627
+    },
+    {
+      "epoch": 0.4,
+      "learning_rate": 1.3586258891964356e-05,
+      "loss": 0.2221,
+      "step": 628
+    },
+    {
+      "epoch": 0.4,
+      "learning_rate": 1.3566906758055862e-05,
+      "loss": 0.2278,
+      "step": 629
+    },
+    {
+      "epoch": 0.4,
+      "learning_rate": 1.3547539306469077e-05,
+      "loss": 0.2329,
+      "step": 630
+    },
+    {
+      "epoch": 0.4,
+      "learning_rate": 1.3528156620375335e-05,
+      "loss": 0.1962,
+      "step": 631
+    },
+    {
+      "epoch": 0.4,
+      "learning_rate": 1.350875878301139e-05,
+      "loss": 0.211,
+      "step": 632
+    },
+    {
+      "epoch": 0.4,
+      "learning_rate": 1.3489345877679068e-05,
+      "loss": 0.1956,
+      "step": 633
+    },
+    {
+      "epoch": 0.41,
+      "learning_rate": 1.3469917987744894e-05,
+      "loss": 0.1839,
+      "step": 634
+    },
+    {
+      "epoch": 0.41,
+      "learning_rate": 1.3450475196639753e-05,
+      "loss": 0.1693,
+      "step": 635
+    },
+    {
+      "epoch": 0.41,
+      "learning_rate": 1.3431017587858515e-05,
+      "loss": 0.2624,
+      "step": 636
+    },
+    {
+      "epoch": 0.41,
+      "learning_rate": 1.341154524495968e-05,
+      "loss": 0.2186,
+      "step": 637
+    },
+    {
+      "epoch": 0.41,
+      "learning_rate": 1.339205825156503e-05,
+      "loss": 0.2107,
+      "step": 638
+    },
+    {
+      "epoch": 0.41,
+      "learning_rate": 1.3372556691359251e-05,
+      "loss": 0.2329,
+      "step": 639
+    },
+    {
+      "epoch": 0.41,
+      "learning_rate": 1.33530406480896e-05,
+      "loss": 0.2185,
+      "step": 640
+    },
+    {
+      "epoch": 0.41,
+      "learning_rate": 1.3333510205565516e-05,
+      "loss": 0.2148,
+      "step": 641
+    },
+    {
+      "epoch": 0.41,
+      "learning_rate": 1.3313965447658279e-05,
+      "loss": 0.2341,
+      "step": 642
+    },
+    {
+      "epoch": 0.41,
+      "learning_rate": 1.3294406458300646e-05,
+      "loss": 0.2566,
+      "step": 643
+    },
+    {
+      "epoch": 0.41,
+      "learning_rate": 1.3274833321486487e-05,
+      "loss": 0.2329,
+      "step": 644
+    },
+    {
+      "epoch": 0.41,
+      "learning_rate": 1.325524612127043e-05,
+      "loss": 0.2315,
+      "step": 645
+    },
+    {
+      "epoch": 0.41,
+      "learning_rate": 1.3235644941767492e-05,
+      "loss": 0.2146,
+      "step": 646
+    },
+    {
+      "epoch": 0.41,
+      "learning_rate": 1.3216029867152724e-05,
+      "loss": 0.2407,
+      "step": 647
+    },
+    {
+      "epoch": 0.41,
+      "learning_rate": 1.3196400981660853e-05,
+      "loss": 0.2238,
+      "step": 648
+    },
+    {
+      "epoch": 0.42,
+      "learning_rate": 1.3176758369585909e-05,
+      "loss": 0.2512,
+      "step": 649
+    },
+    {
+      "epoch": 0.42,
+      "learning_rate": 1.3157102115280873e-05,
+      "loss": 0.2073,
+      "step": 650
+    },
+    {
+      "epoch": 0.42,
+      "learning_rate": 1.3137432303157306e-05,
+      "loss": 0.2134,
+      "step": 651
+    },
+    {
+      "epoch": 0.42,
+      "learning_rate": 1.3117749017684998e-05,
+      "loss": 0.2378,
+      "step": 652
+    },
+    {
+      "epoch": 0.42,
+      "learning_rate": 1.3098052343391596e-05,
+      "loss": 0.2401,
+      "step": 653
+    },
+    {
+      "epoch": 0.42,
+      "learning_rate": 1.3078342364862242e-05,
+      "loss": 0.2123,
+      "step": 654
+    },
+    {
+      "epoch": 0.42,
+      "learning_rate": 1.3058619166739212e-05,
+      "loss": 0.188,
+      "step": 655
+    },
+    {
+      "epoch": 0.42,
+      "learning_rate": 1.3038882833721557e-05,
+      "loss": 0.2142,
+      "step": 656
+    },
+    {
+      "epoch": 0.42,
+      "learning_rate": 1.3019133450564725e-05,
+      "loss": 0.2075,
+      "step": 657
+    },
+    {
+      "epoch": 0.42,
+      "learning_rate": 1.2999371102080218e-05,
+      "loss": 0.2214,
+      "step": 658
+    },
+    {
+      "epoch": 0.42,
+      "learning_rate": 1.2979595873135208e-05,
+      "loss": 0.2149,
+      "step": 659
+    },
+    {
+      "epoch": 0.42,
+      "learning_rate": 1.295980784865218e-05,
+      "loss": 0.2593,
+      "step": 660
+    },
+    {
+      "epoch": 0.42,
+      "learning_rate": 1.2940007113608572e-05,
+      "loss": 0.1885,
+      "step": 661
+    },
+    {
+      "epoch": 0.42,
+      "learning_rate": 1.2920193753036408e-05,
+      "loss": 0.2286,
+      "step": 662
+    },
+    {
+      "epoch": 0.42,
+      "learning_rate": 1.290036785202192e-05,
+      "loss": 0.2748,
+      "step": 663
+    },
+    {
+      "epoch": 0.42,
+      "learning_rate": 1.288052949570521e-05,
+      "loss": 0.2218,
+      "step": 664
+    },
+    {
+      "epoch": 0.43,
+      "learning_rate": 1.2860678769279855e-05,
+      "loss": 0.2124,
+      "step": 665
+    },
+    {
+      "epoch": 0.43,
+      "learning_rate": 1.2840815757992556e-05,
+      "loss": 0.1945,
+      "step": 666
+    },
+    {
+      "epoch": 0.43,
+      "learning_rate": 1.2820940547142773e-05,
+      "loss": 0.2369,
+      "step": 667
+    },
+    {
+      "epoch": 0.43,
+      "learning_rate": 1.2801053222082355e-05,
+      "loss": 0.2186,
+      "step": 668
+    },
+    {
+      "epoch": 0.43,
+      "learning_rate": 1.278115386821518e-05,
+      "loss": 0.2208,
+      "step": 669
+    },
+    {
+      "epoch": 0.43,
+      "learning_rate": 1.2761242570996765e-05,
+      "loss": 0.2114,
+      "step": 670
+    },
+    {
+      "epoch": 0.43,
+      "learning_rate": 1.2741319415933934e-05,
+      "loss": 0.2086,
+      "step": 671
+    },
+    {
+      "epoch": 0.43,
+      "learning_rate": 1.2721384488584424e-05,
+      "loss": 0.2163,
+      "step": 672
+    },
+    {
+      "epoch": 0.43,
+      "learning_rate": 1.2701437874556537e-05,
+      "loss": 0.2168,
+      "step": 673
+    },
+    {
+      "epoch": 0.43,
+      "learning_rate": 1.2681479659508746e-05,
+      "loss": 0.2196,
+      "step": 674
+    },
+    {
+      "epoch": 0.43,
+      "learning_rate": 1.2661509929149351e-05,
+      "loss": 0.2227,
+      "step": 675
+    },
+    {
+      "epoch": 0.43,
+      "learning_rate": 1.2641528769236111e-05,
+      "loss": 0.195,
+      "step": 676
+    },
+    {
+      "epoch": 0.43,
+      "learning_rate": 1.2621536265575855e-05,
+      "loss": 0.2063,
+      "step": 677
+    },
+    {
+      "epoch": 0.43,
+      "learning_rate": 1.2601532504024138e-05,
+      "loss": 0.2012,
+      "step": 678
+    },
+    {
+      "epoch": 0.43,
+      "learning_rate": 1.258151757048485e-05,
+      "loss": 0.2483,
+      "step": 679
+    },
+    {
+      "epoch": 0.44,
+      "learning_rate": 1.2561491550909865e-05,
+      "loss": 0.2422,
+      "step": 680
+    },
+    {
+      "epoch": 0.44,
+      "learning_rate": 1.2541454531298661e-05,
+      "loss": 0.2188,
+      "step": 681
+    },
+    {
+      "epoch": 0.44,
+      "learning_rate": 1.2521406597697959e-05,
+      "loss": 0.2103,
+      "step": 682
+    },
+    {
+      "epoch": 0.44,
+      "learning_rate": 1.2501347836201343e-05,
+      "loss": 0.2203,
+      "step": 683
+    },
+    {
+      "epoch": 0.44,
+      "learning_rate": 1.2481278332948902e-05,
+      "loss": 0.2042,
+      "step": 684
+    },
+    {
+      "epoch": 0.44,
+      "learning_rate": 1.2461198174126851e-05,
+      "loss": 0.2365,
+      "step": 685
+    },
+    {
+      "epoch": 0.44,
+      "learning_rate": 1.2441107445967164e-05,
+      "loss": 0.2442,
+      "step": 686
+    },
+    {
+      "epoch": 0.44,
+      "learning_rate": 1.2421006234747202e-05,
+      "loss": 0.218,
+      "step": 687
+    },
+    {
+      "epoch": 0.44,
+      "learning_rate": 1.2400894626789353e-05,
+      "loss": 0.2188,
+      "step": 688
+    },
+    {
+      "epoch": 0.44,
+      "learning_rate": 1.238077270846064e-05,
+      "loss": 0.2286,
+      "step": 689
+    },
+    {
+      "epoch": 0.44,
+      "learning_rate": 1.2360640566172368e-05,
+      "loss": 0.2171,
+      "step": 690
+    },
+    {
+      "epoch": 0.44,
+      "learning_rate": 1.2340498286379757e-05,
+      "loss": 0.2229,
+      "step": 691
+    },
+    {
+      "epoch": 0.44,
+      "learning_rate": 1.2320345955581545e-05,
+      "loss": 0.2134,
+      "step": 692
+    },
+    {
+      "epoch": 0.44,
+      "learning_rate": 1.2300183660319647e-05,
+      "loss": 0.208,
+      "step": 693
+    },
+    {
+      "epoch": 0.44,
+      "learning_rate": 1.2280011487178758e-05,
+      "loss": 0.2363,
+      "step": 694
+    },
+    {
+      "epoch": 0.44,
+      "learning_rate": 1.2259829522786003e-05,
+      "loss": 0.259,
+      "step": 695
+    },
+    {
+      "epoch": 0.45,
+      "learning_rate": 1.2239637853810539e-05,
+      "loss": 0.2261,
+      "step": 696
+    },
+    {
+      "epoch": 0.45,
+      "learning_rate": 1.2219436566963221e-05,
+      "loss": 0.2183,
+      "step": 697
+    },
+    {
+      "epoch": 0.45,
+      "learning_rate": 1.219922574899619e-05,
+      "loss": 0.2222,
+      "step": 698
+    },
+    {
+      "epoch": 0.45,
+      "learning_rate": 1.2179005486702518e-05,
+      "loss": 0.2295,
+      "step": 699
+    },
+    {
+      "epoch": 0.45,
+      "learning_rate": 1.2158775866915839e-05,
+      "loss": 0.2068,
+      "step": 700
+    },
+    {
+      "epoch": 0.45,
+      "learning_rate": 1.2138536976509974e-05,
+      "loss": 0.2268,
+      "step": 701
+    },
+    {
+      "epoch": 0.45,
+      "learning_rate": 1.2118288902398549e-05,
+      "loss": 0.2406,
+      "step": 702
+    },
+    {
+      "epoch": 0.45,
+      "learning_rate": 1.2098031731534636e-05,
+      "loss": 0.244,
+      "step": 703
+    },
+    {
+      "epoch": 0.45,
+      "learning_rate": 1.2077765550910363e-05,
+      "loss": 0.2186,
+      "step": 704
+    },
+    {
+      "epoch": 0.45,
+      "learning_rate": 1.2057490447556556e-05,
+      "loss": 0.1868,
+      "step": 705
+    },
+    {
+      "epoch": 0.45,
+      "learning_rate": 1.2037206508542358e-05,
+      "loss": 0.2333,
+      "step": 706
+    },
+    {
+      "epoch": 0.45,
+      "learning_rate": 1.2016913820974855e-05,
+      "loss": 0.1802,
+      "step": 707
+    },
+    {
+      "epoch": 0.45,
+      "learning_rate": 1.19966124719987e-05,
+      "loss": 0.2037,
+      "step": 708
+    },
+    {
+      "epoch": 0.45,
+      "learning_rate": 1.1976302548795746e-05,
+      "loss": 0.2676,
+      "step": 709
+    },
+    {
+      "epoch": 0.45,
+      "learning_rate": 1.1955984138584661e-05,
+      "loss": 0.1908,
+      "step": 710
+    },
+    {
+      "epoch": 0.45,
+      "learning_rate": 1.1935657328620566e-05,
+      "loss": 0.2509,
+      "step": 711
+    },
+    {
+      "epoch": 0.46,
+      "learning_rate": 1.1915322206194654e-05,
+      "loss": 0.2188,
+      "step": 712
+    },
+    {
+      "epoch": 0.46,
+      "learning_rate": 1.1894978858633807e-05,
+      "loss": 0.2103,
+      "step": 713
+    },
+    {
+      "epoch": 0.46,
+      "learning_rate": 1.1874627373300236e-05,
+      "loss": 0.2399,
+      "step": 714
+    },
+    {
+      "epoch": 0.46,
+      "learning_rate": 1.1854267837591095e-05,
+      "loss": 0.2505,
+      "step": 715
+    },
+    {
+      "epoch": 0.46,
+      "learning_rate": 1.1833900338938117e-05,
+      "loss": 0.2239,
+      "step": 716
+    },
+    {
+      "epoch": 0.46,
+      "learning_rate": 1.1813524964807216e-05,
+      "loss": 0.1988,
+      "step": 717
+    },
+    {
+      "epoch": 0.46,
+      "learning_rate": 1.1793141802698141e-05,
+      "loss": 0.2499,
+      "step": 718
+    },
+    {
+      "epoch": 0.46,
+      "learning_rate": 1.1772750940144079e-05,
+      "loss": 0.2408,
+      "step": 719
+    },
+    {
+      "epoch": 0.46,
+      "learning_rate": 1.1752352464711286e-05,
+      "loss": 0.2,
+      "step": 720
+    },
+    {
+      "epoch": 0.46,
+      "learning_rate": 1.1731946463998712e-05,
+      "loss": 0.2467,
+      "step": 721
+    },
+    {
+      "epoch": 0.46,
+      "learning_rate": 1.1711533025637626e-05,
+      "loss": 0.2349,
+      "step": 722
+    },
+    {
+      "epoch": 0.46,
+      "learning_rate": 1.1691112237291225e-05,
+      "loss": 0.2367,
+      "step": 723
+    },
+    {
+      "epoch": 0.46,
+      "learning_rate": 1.1670684186654287e-05,
+      "loss": 0.2431,
+      "step": 724
+    },
+    {
+      "epoch": 0.46,
+      "learning_rate": 1.1650248961452765e-05,
+      "loss": 0.2019,
+      "step": 725
+    },
+    {
+      "epoch": 0.46,
+      "learning_rate": 1.1629806649443428e-05,
+      "loss": 0.2254,
+      "step": 726
+    },
+    {
+      "epoch": 0.47,
+      "learning_rate": 1.1609357338413476e-05,
+      "loss": 0.2083,
+      "step": 727
+    },
+    {
+      "epoch": 0.47,
+      "learning_rate": 1.1588901116180165e-05,
+      "loss": 0.2307,
+      "step": 728
+    },
+    {
+      "epoch": 0.47,
+      "learning_rate": 1.1568438070590428e-05,
+      "loss": 0.2408,
+      "step": 729
+    },
+    {
+      "epoch": 0.47,
+      "learning_rate": 1.1547968289520513e-05,
+      "loss": 0.1983,
+      "step": 730
+    },
+    {
+      "epoch": 0.47,
+      "learning_rate": 1.1527491860875574e-05,
+      "loss": 0.1988,
+      "step": 731
+    },
+    {
+      "epoch": 0.47,
+      "learning_rate": 1.150700887258933e-05,
+      "loss": 0.2198,
+      "step": 732
+    },
+    {
+      "epoch": 0.47,
+      "learning_rate": 1.1486519412623653e-05,
+      "loss": 0.2511,
+      "step": 733
+    },
+    {
+      "epoch": 0.47,
+      "learning_rate": 1.146602356896822e-05,
+      "loss": 0.2198,
+      "step": 734
+    },
+    {
+      "epoch": 0.47,
+      "learning_rate": 1.1445521429640113e-05,
+      "loss": 0.2139,
+      "step": 735
+    },
+    {
+      "epoch": 0.47,
+      "learning_rate": 1.1425013082683462e-05,
+      "loss": 0.2207,
+      "step": 736
+    },
+    {
+      "epoch": 0.47,
+      "learning_rate": 1.1404498616169039e-05,
+      "loss": 0.2451,
+      "step": 737
+    },
+    {
+      "epoch": 0.47,
+      "learning_rate": 1.1383978118193908e-05,
+      "loss": 0.2095,
+      "step": 738
+    },
+    {
+      "epoch": 0.47,
+      "learning_rate": 1.136345167688103e-05,
+      "loss": 0.1927,
+      "step": 739
+    },
+    {
+      "epoch": 0.47,
+      "learning_rate": 1.1342919380378888e-05,
+      "loss": 0.2305,
+      "step": 740
+    },
+    {
+      "epoch": 0.47,
+      "learning_rate": 1.1322381316861114e-05,
+      "loss": 0.1902,
+      "step": 741
+    },
+    {
+      "epoch": 0.47,
+      "learning_rate": 1.1301837574526097e-05,
+      "loss": 0.1851,
+      "step": 742
+    },
+    {
+      "epoch": 0.48,
+      "learning_rate": 1.1281288241596624e-05,
+      "loss": 0.2015,
+      "step": 743
+    },
+    {
+      "epoch": 0.48,
+      "learning_rate": 1.126073340631948e-05,
+      "loss": 0.2052,
+      "step": 744
+    },
+    {
+      "epoch": 0.48,
+      "learning_rate": 1.1240173156965089e-05,
+      "loss": 0.2049,
+      "step": 745
+    },
+    {
+      "epoch": 0.48,
+      "learning_rate": 1.1219607581827119e-05,
+      "loss": 0.2354,
+      "step": 746
+    },
+    {
+      "epoch": 0.48,
+      "learning_rate": 1.1199036769222104e-05,
+      "loss": 0.2238,
+      "step": 747
+    },
+    {
+      "epoch": 0.48,
+      "learning_rate": 1.1178460807489086e-05,
+      "loss": 0.198,
+      "step": 748
+    },
+    {
+      "epoch": 0.48,
+      "learning_rate": 1.1157879784989203e-05,
+      "loss": 0.1999,
+      "step": 749
+    },
+    {
+      "epoch": 0.48,
+      "learning_rate": 1.1137293790105332e-05,
+      "loss": 0.2248,
+      "step": 750
+    },
+    {
+      "epoch": 0.48,
+      "learning_rate": 1.1116702911241704e-05,
+      "loss": 0.2034,
+      "step": 751
+    },
+    {
+      "epoch": 0.48,
+      "learning_rate": 1.1096107236823525e-05,
+      "loss": 0.2192,
+      "step": 752
+    },
+    {
+      "epoch": 0.48,
+      "learning_rate": 1.107550685529659e-05,
+      "loss": 0.1934,
+      "step": 753
+    },
+    {
+      "epoch": 0.48,
+      "learning_rate": 1.1054901855126914e-05,
+      "loss": 0.1908,
+      "step": 754
+    },
+    {
+      "epoch": 0.48,
+      "learning_rate": 1.1034292324800342e-05,
+      "loss": 0.1883,
+      "step": 755
+    },
+    {
+      "epoch": 0.48,
+      "learning_rate": 1.1013678352822176e-05,
+      "loss": 0.2147,
+      "step": 756
+    },
+    {
+      "epoch": 0.48,
+      "learning_rate": 1.099306002771679e-05,
+      "loss": 0.2325,
+      "step": 757
+    },
+    {
+      "epoch": 0.48,
+      "learning_rate": 1.0972437438027252e-05,
+      "loss": 0.2515,
+      "step": 758
+    },
+    {
+      "epoch": 0.49,
+      "learning_rate": 1.0951810672314946e-05,
+      "loss": 0.236,
+      "step": 759
+    },
+    {
+      "epoch": 0.49,
+      "learning_rate": 1.093117981915919e-05,
+      "loss": 0.2078,
+      "step": 760
+    },
+    {
+      "epoch": 0.49,
+      "learning_rate": 1.091054496715685e-05,
+      "loss": 0.2216,
+      "step": 761
+    },
+    {
+      "epoch": 0.49,
+      "learning_rate": 1.0889906204921965e-05,
+      "loss": 0.21,
+      "step": 762
+    },
+    {
+      "epoch": 0.49,
+      "learning_rate": 1.0869263621085375e-05,
+      "loss": 0.2326,
+      "step": 763
+    },
+    {
+      "epoch": 0.49,
+      "learning_rate": 1.0848617304294324e-05,
+      "loss": 0.1953,
+      "step": 764
+    },
+    {
+      "epoch": 0.49,
+      "learning_rate": 1.0827967343212087e-05,
+      "loss": 0.2061,
+      "step": 765
+    },
+    {
+      "epoch": 0.49,
+      "learning_rate": 1.0807313826517592e-05,
+      "loss": 0.2392,
+      "step": 766
+    },
+    {
+      "epoch": 0.49,
+      "learning_rate": 1.0786656842905028e-05,
+      "loss": 0.2041,
+      "step": 767
+    },
+    {
+      "epoch": 0.49,
+      "learning_rate": 1.0765996481083483e-05,
+      "loss": 0.1861,
+      "step": 768
+    },
+    {
+      "epoch": 0.49,
+      "learning_rate": 1.074533282977655e-05,
+      "loss": 0.2084,
+      "step": 769
+    },
+    {
+      "epoch": 0.49,
+      "learning_rate": 1.0724665977721944e-05,
+      "loss": 0.1985,
+      "step": 770
+    },
+    {
+      "epoch": 0.49,
+      "learning_rate": 1.0703996013671126e-05,
+      "loss": 0.245,
+      "step": 771
+    },
+    {
+      "epoch": 0.49,
+      "learning_rate": 1.0683323026388923e-05,
+      "loss": 0.1873,
+      "step": 772
+    },
+    {
+      "epoch": 0.49,
+      "learning_rate": 1.0662647104653145e-05,
+      "loss": 0.2217,
+      "step": 773
+    },
+    {
+      "epoch": 0.5,
+      "learning_rate": 1.0641968337254204e-05,
+      "loss": 0.2155,
+      "step": 774
+    },
+    {
+      "epoch": 0.5,
+      "learning_rate": 1.0621286812994733e-05,
+      "loss": 0.2134,
+      "step": 775
+    },
+    {
+      "epoch": 0.5,
+      "learning_rate": 1.0600602620689198e-05,
+      "loss": 0.2092,
+      "step": 776
+    },
+    {
+      "epoch": 0.5,
+      "learning_rate": 1.057991584916353e-05,
+      "loss": 0.2513,
+      "step": 777
+    },
+    {
+      "epoch": 0.5,
+      "learning_rate": 1.0559226587254732e-05,
+      "loss": 0.1817,
+      "step": 778
+    },
+    {
+      "epoch": 0.5,
+      "learning_rate": 1.0538534923810506e-05,
+      "loss": 0.2246,
+      "step": 779
+    },
+    {
+      "epoch": 0.5,
+      "learning_rate": 1.0517840947688858e-05,
+      "loss": 0.2419,
+      "step": 780
+    },
+    {
+      "epoch": 0.5,
+      "learning_rate": 1.049714474775774e-05,
+      "loss": 0.2216,
+      "step": 781
+    },
+    {
+      "epoch": 0.5,
+      "learning_rate": 1.0476446412894639e-05,
+      "loss": 0.2076,
+      "step": 782
+    },
+    {
+      "epoch": 0.5,
+      "learning_rate": 1.0455746031986215e-05,
+      "loss": 0.2397,
+      "step": 783
+    },
+    {
+      "epoch": 0.5,
+      "learning_rate": 1.0435043693927929e-05,
+      "loss": 0.2027,
+      "step": 784
+    },
+    {
+      "epoch": 0.5,
+      "learning_rate": 1.041433948762362e-05,
+      "loss": 0.2239,
+      "step": 785
+    },
+    {
+      "epoch": 0.5,
+      "learning_rate": 1.0393633501985167e-05,
+      "loss": 0.1996,
+      "step": 786
+    },
+    {
+      "epoch": 0.5,
+      "learning_rate": 1.0372925825932095e-05,
+      "loss": 0.1947,
+      "step": 787
+    },
+    {
+      "epoch": 0.5,
+      "learning_rate": 1.0352216548391173e-05,
+      "loss": 0.1646,
+      "step": 788
+    },
+    {
+      "epoch": 0.5,
+      "learning_rate": 1.0331505758296054e-05,
+      "loss": 0.2107,
+      "step": 789
+    },
+    {
+      "epoch": 0.51,
+      "learning_rate": 1.0310793544586895e-05,
+      "loss": 0.2089,
+      "step": 790
+    },
+    {
+      "epoch": 0.51,
+      "learning_rate": 1.029007999620995e-05,
+      "loss": 0.1913,
+      "step": 791
+    },
+    {
+      "epoch": 0.51,
+      "learning_rate": 1.026936520211722e-05,
+      "loss": 0.2098,
+      "step": 792
+    },
+    {
+      "epoch": 0.51,
+      "learning_rate": 1.0248649251266043e-05,
+      "loss": 0.2134,
+      "step": 793
+    },
+    {
+      "epoch": 0.51,
+      "learning_rate": 1.022793223261874e-05,
+      "loss": 0.2235,
+      "step": 794
+    },
+    {
+      "epoch": 0.51,
+      "learning_rate": 1.0207214235142197e-05,
+      "loss": 0.2288,
+      "step": 795
+    },
+    {
+      "epoch": 0.51,
+      "learning_rate": 1.0186495347807523e-05,
+      "loss": 0.2257,
+      "step": 796
+    },
+    {
+      "epoch": 0.51,
+      "learning_rate": 1.0165775659589638e-05,
+      "loss": 0.2282,
+      "step": 797
+    },
+    {
+      "epoch": 0.51,
+      "learning_rate": 1.0145055259466902e-05,
+      "loss": 0.184,
+      "step": 798
+    },
+    {
+      "epoch": 0.51,
+      "learning_rate": 1.0124334236420735e-05,
+      "loss": 0.1496,
+      "step": 799
+    },
+    {
+      "epoch": 0.51,
+      "learning_rate": 1.0103612679435226e-05,
+      "loss": 0.214,
+      "step": 800
+    },
+    {
+      "epoch": 0.51,
+      "learning_rate": 1.0082890677496766e-05,
+      "loss": 0.2316,
+      "step": 801
+    },
+    {
+      "epoch": 0.51,
+      "learning_rate": 1.0062168319593655e-05,
+      "loss": 0.1972,
+      "step": 802
+    },
+    {
+      "epoch": 0.51,
+      "learning_rate": 1.0041445694715717e-05,
+      "loss": 0.2079,
+      "step": 803
+    },
+    {
+      "epoch": 0.51,
+      "learning_rate": 1.0020722891853918e-05,
+      "loss": 0.2555,
+      "step": 804
+    },
+    {
+      "epoch": 0.52,
+      "learning_rate": 1e-05,
+      "loss": 0.2107,
+      "step": 805
+    },
+    {
+      "epoch": 0.52,
+      "learning_rate": 9.979277108146087e-06,
+      "loss": 0.1843,
+      "step": 806
+    },
+    {
+      "epoch": 0.52,
+      "learning_rate": 9.95855430528429e-06,
+      "loss": 0.2259,
+      "step": 807
+    },
+    {
+      "epoch": 0.52,
+      "learning_rate": 9.937831680406348e-06,
+      "loss": 0.2687,
+      "step": 808
+    },
+    {
+      "epoch": 0.52,
+      "learning_rate": 9.917109322503234e-06,
+      "loss": 0.2012,
+      "step": 809
+    },
+    {
+      "epoch": 0.52,
+      "learning_rate": 9.896387320564776e-06,
+      "loss": 0.2196,
+      "step": 810
+    },
+    {
+      "epoch": 0.52,
+      "learning_rate": 9.87566576357927e-06,
+      "loss": 0.1938,
+      "step": 811
+    },
+    {
+      "epoch": 0.52,
+      "learning_rate": 9.854944740533101e-06,
+      "loss": 0.2288,
+      "step": 812
+    },
+    {
+      "epoch": 0.52,
+      "learning_rate": 9.834224340410367e-06,
+      "loss": 0.2103,
+      "step": 813
+    },
+    {
+      "epoch": 0.52,
+      "learning_rate": 9.813504652192482e-06,
+      "loss": 0.2382,
+      "step": 814
+    },
+    {
+      "epoch": 0.52,
+      "learning_rate": 9.792785764857803e-06,
+      "loss": 0.2012,
+      "step": 815
+    },
+    {
+      "epoch": 0.52,
+      "learning_rate": 9.772067767381263e-06,
+      "loss": 0.2309,
+      "step": 816
+    },
+    {
+      "epoch": 0.52,
+      "learning_rate": 9.751350748733958e-06,
+      "loss": 0.2142,
+      "step": 817
+    },
+    {
+      "epoch": 0.52,
+      "learning_rate": 9.730634797882786e-06,
+      "loss": 0.1948,
+      "step": 818
+    },
+    {
+      "epoch": 0.52,
+      "learning_rate": 9.709920003790054e-06,
+      "loss": 0.2318,
+      "step": 819
+    },
+    {
+      "epoch": 0.52,
+      "learning_rate": 9.689206455413112e-06,
+      "loss": 0.212,
+      "step": 820
+    },
+    {
+      "epoch": 0.53,
+      "learning_rate": 9.668494241703946e-06,
+      "loss": 0.1916,
+      "step": 821
+    },
+    {
+      "epoch": 0.53,
+      "learning_rate": 9.64778345160883e-06,
+      "loss": 0.2301,
+      "step": 822
+    },
+    {
+      "epoch": 0.53,
+      "learning_rate": 9.627074174067909e-06,
+      "loss": 0.1976,
+      "step": 823
+    },
+    {
+      "epoch": 0.53,
+      "learning_rate": 9.606366498014834e-06,
+      "loss": 0.2048,
+      "step": 824
+    },
+    {
+      "epoch": 0.53,
+      "learning_rate": 9.585660512376385e-06,
+      "loss": 0.1812,
+      "step": 825
+    },
+    {
+      "epoch": 0.53,
+      "learning_rate": 9.564956306072078e-06,
+      "loss": 0.2087,
+      "step": 826
+    },
+    {
+      "epoch": 0.53,
+      "learning_rate": 9.544253968013785e-06,
+      "loss": 0.2072,
+      "step": 827
+    },
+    {
+      "epoch": 0.53,
+      "learning_rate": 9.523553587105366e-06,
+      "loss": 0.2201,
+      "step": 828
+    },
+    {
+      "epoch": 0.53,
+      "learning_rate": 9.502855252242265e-06,
+      "loss": 0.2116,
+      "step": 829
+    },
+    {
+      "epoch": 0.53,
+      "learning_rate": 9.482159052311144e-06,
+      "loss": 0.2386,
+      "step": 830
+    },
+    {
+      "epoch": 0.53,
+      "learning_rate": 9.461465076189499e-06,
+      "loss": 0.2024,
+      "step": 831
+    },
+    {
+      "epoch": 0.53,
+      "learning_rate": 9.440773412745273e-06,
+      "loss": 0.195,
+      "step": 832
+    },
+    {
+      "epoch": 0.53,
+      "learning_rate": 9.420084150836472e-06,
+      "loss": 0.225,
+      "step": 833
+    },
+    {
+      "epoch": 0.53,
+      "learning_rate": 9.399397379310805e-06,
+      "loss": 0.165,
+      "step": 834
+    },
+    {
+      "epoch": 0.53,
+      "learning_rate": 9.37871318700527e-06,
+      "loss": 0.2306,
+      "step": 835
+    },
+    {
+      "epoch": 0.53,
+      "learning_rate": 9.358031662745797e-06,
+      "loss": 0.205,
+      "step": 836
+    },
+    {
+      "epoch": 0.54,
+      "learning_rate": 9.337352895346858e-06,
+      "loss": 0.2146,
+      "step": 837
+    },
+    {
+      "epoch": 0.54,
+      "learning_rate": 9.316676973611082e-06,
+      "loss": 0.2242,
+      "step": 838
+    },
+    {
+      "epoch": 0.54,
+      "learning_rate": 9.296003986328876e-06,
+      "loss": 0.224,
+      "step": 839
+    },
+    {
+      "epoch": 0.54,
+      "learning_rate": 9.275334022278058e-06,
+      "loss": 0.2504,
+      "step": 840
+    },
+    {
+      "epoch": 0.54,
+      "learning_rate": 9.254667170223454e-06,
+      "loss": 0.1996,
+      "step": 841
+    },
+    {
+      "epoch": 0.54,
+      "learning_rate": 9.23400351891652e-06,
+      "loss": 0.191,
+      "step": 842
+    },
+    {
+      "epoch": 0.54,
+      "learning_rate": 9.213343157094975e-06,
+      "loss": 0.203,
+      "step": 843
+    },
+    {
+      "epoch": 0.54,
+      "learning_rate": 9.192686173482415e-06,
+      "loss": 0.2228,
+      "step": 844
+    },
+    {
+      "epoch": 0.54,
+      "learning_rate": 9.172032656787913e-06,
+      "loss": 0.2307,
+      "step": 845
+    },
+    {
+      "epoch": 0.54,
+      "learning_rate": 9.151382695705679e-06,
+      "loss": 0.2301,
+      "step": 846
+    },
+    {
+      "epoch": 0.54,
+      "learning_rate": 9.130736378914626e-06,
+      "loss": 0.2246,
+      "step": 847
+    },
+    {
+      "epoch": 0.54,
+      "learning_rate": 9.110093795078037e-06,
+      "loss": 0.2177,
+      "step": 848
+    },
+    {
+      "epoch": 0.54,
+      "learning_rate": 9.089455032843156e-06,
+      "loss": 0.2173,
+      "step": 849
+    },
+    {
+      "epoch": 0.54,
+      "learning_rate": 9.068820180840814e-06,
+      "loss": 0.2091,
+      "step": 850
+    },
+    {
+      "epoch": 0.54,
+      "learning_rate": 9.048189327685056e-06,
+      "loss": 0.2124,
+      "step": 851
+    },
+    {
+      "epoch": 0.55,
+      "learning_rate": 9.027562561972751e-06,
+      "loss": 0.1923,
+      "step": 852
+    },
+    {
+      "epoch": 0.55,
+      "learning_rate": 9.006939972283213e-06,
+      "loss": 0.2212,
+      "step": 853
+    },
+    {
+      "epoch": 0.55,
+      "learning_rate": 8.986321647177827e-06,
+      "loss": 0.2112,
+      "step": 854
+    },
+    {
+      "epoch": 0.55,
+      "learning_rate": 8.965707675199661e-06,
+      "loss": 0.2202,
+      "step": 855
+    },
+    {
+      "epoch": 0.55,
+      "learning_rate": 8.94509814487309e-06,
+      "loss": 0.2352,
+      "step": 856
+    },
+    {
+      "epoch": 0.55,
+      "learning_rate": 8.92449314470341e-06,
+      "loss": 0.2354,
+      "step": 857
+    },
+    {
+      "epoch": 0.55,
+      "learning_rate": 8.903892763176478e-06,
+      "loss": 0.2079,
+      "step": 858
+    },
+    {
+      "epoch": 0.55,
+      "learning_rate": 8.883297088758298e-06,
+      "loss": 0.212,
+      "step": 859
+    },
+    {
+      "epoch": 0.55,
+      "learning_rate": 8.862706209894672e-06,
+      "loss": 0.207,
+      "step": 860
+    },
+    {
+      "epoch": 0.55,
+      "learning_rate": 8.842120215010804e-06,
+      "loss": 0.2178,
+      "step": 861
+    },
+    {
+      "epoch": 0.55,
+      "learning_rate": 8.82153919251092e-06,
+      "loss": 0.238,
+      "step": 862
+    },
+    {
+      "epoch": 0.55,
+      "learning_rate": 8.800963230777895e-06,
+      "loss": 0.2373,
+      "step": 863
+    },
+    {
+      "epoch": 0.55,
+      "learning_rate": 8.780392418172885e-06,
+      "loss": 0.2281,
+      "step": 864
+    },
+    {
+      "epoch": 0.55,
+      "learning_rate": 8.759826843034916e-06,
+      "loss": 0.2166,
+      "step": 865
+    },
+    {
+      "epoch": 0.55,
+      "learning_rate": 8.739266593680524e-06,
+      "loss": 0.2215,
+      "step": 866
+    },
+    {
+      "epoch": 0.55,
+      "learning_rate": 8.718711758403382e-06,
+      "loss": 0.21,
+      "step": 867
+    },
+    {
+      "epoch": 0.56,
+      "learning_rate": 8.698162425473908e-06,
+      "loss": 0.1957,
+      "step": 868
+    },
+    {
+      "epoch": 0.56,
+      "learning_rate": 8.677618683138888e-06,
+      "loss": 0.213,
+      "step": 869
+    },
+    {
+      "epoch": 0.56,
+      "learning_rate": 8.657080619621115e-06,
+      "loss": 0.2343,
+      "step": 870
+    },
+    {
+      "epoch": 0.56,
+      "learning_rate": 8.636548323118974e-06,
+      "loss": 0.2107,
+      "step": 871
+    },
+    {
+      "epoch": 0.56,
+      "learning_rate": 8.616021881806095e-06,
+      "loss": 0.208,
+      "step": 872
+    },
+    {
+      "epoch": 0.56,
+      "learning_rate": 8.595501383830964e-06,
+      "loss": 0.2278,
+      "step": 873
+    },
+    {
+      "epoch": 0.56,
+      "learning_rate": 8.574986917316543e-06,
+      "loss": 0.2158,
+      "step": 874
+    },
+    {
+      "epoch": 0.56,
+      "learning_rate": 8.554478570359888e-06,
+      "loss": 0.247,
+      "step": 875
+    },
+    {
+      "epoch": 0.56,
+      "learning_rate": 8.533976431031783e-06,
+      "loss": 0.2057,
+      "step": 876
+    },
+    {
+      "epoch": 0.56,
+      "learning_rate": 8.51348058737635e-06,
+      "loss": 0.2149,
+      "step": 877
+    },
+    {
+      "epoch": 0.56,
+      "learning_rate": 8.492991127410674e-06,
+      "loss": 0.2143,
+      "step": 878
+    },
+    {
+      "epoch": 0.56,
+      "learning_rate": 8.472508139124427e-06,
+      "loss": 0.1739,
+      "step": 879
+    },
+    {
+      "epoch": 0.56,
+      "learning_rate": 8.45203171047949e-06,
+      "loss": 0.1869,
+      "step": 880
+    },
+    {
+      "epoch": 0.56,
+      "learning_rate": 8.431561929409572e-06,
+      "loss": 0.2167,
+      "step": 881
+    },
+    {
+      "epoch": 0.56,
+      "learning_rate": 8.411098883819838e-06,
+      "loss": 0.2085,
+      "step": 882
+    },
+    {
+      "epoch": 0.56,
+      "learning_rate": 8.390642661586527e-06,
+      "loss": 0.2199,
+      "step": 883
+    },
+    {
+      "epoch": 0.57,
+      "learning_rate": 8.370193350556575e-06,
+      "loss": 0.2045,
+      "step": 884
+    },
+    {
+      "epoch": 0.57,
+      "learning_rate": 8.34975103854724e-06,
+      "loss": 0.2229,
+      "step": 885
+    },
+    {
+      "epoch": 0.57,
+      "learning_rate": 8.329315813345718e-06,
+      "loss": 0.1863,
+      "step": 886
+    },
+    {
+      "epoch": 0.57,
+      "learning_rate": 8.308887762708776e-06,
+      "loss": 0.2011,
+      "step": 887
+    },
+    {
+      "epoch": 0.57,
+      "learning_rate": 8.288466974362378e-06,
+      "loss": 0.188,
+      "step": 888
+    },
+    {
+      "epoch": 0.57,
+      "learning_rate": 8.268053536001291e-06,
+      "loss": 0.1992,
+      "step": 889
+    },
+    {
+      "epoch": 0.57,
+      "learning_rate": 8.24764753528872e-06,
+      "loss": 0.2387,
+      "step": 890
+    },
+    {
+      "epoch": 0.57,
+      "learning_rate": 8.227249059855926e-06,
+      "loss": 0.2175,
+      "step": 891
+    },
+    {
+      "epoch": 0.57,
+      "learning_rate": 8.206858197301864e-06,
+      "loss": 0.2155,
+      "step": 892
+    },
+    {
+      "epoch": 0.57,
+      "learning_rate": 8.186475035192788e-06,
+      "loss": 0.2403,
+      "step": 893
+    },
+    {
+      "epoch": 0.57,
+      "learning_rate": 8.166099661061888e-06,
+      "loss": 0.2306,
+      "step": 894
+    },
+    {
+      "epoch": 0.57,
+      "learning_rate": 8.145732162408906e-06,
+      "loss": 0.1834,
+      "step": 895
+    },
+    {
+      "epoch": 0.57,
+      "learning_rate": 8.125372626699768e-06,
+      "loss": 0.2053,
+      "step": 896
+    },
+    {
+      "epoch": 0.57,
+      "learning_rate": 8.105021141366196e-06,
+      "loss": 0.226,
+      "step": 897
+    },
+    {
+      "epoch": 0.57,
+      "learning_rate": 8.08467779380535e-06,
+      "loss": 0.2041,
+      "step": 898
+    },
+    {
+      "epoch": 0.58,
+      "learning_rate": 8.064342671379436e-06,
+      "loss": 0.2172,
+      "step": 899
+    },
+    {
+      "epoch": 0.58,
+      "learning_rate": 8.04401586141534e-06,
+      "loss": 0.2279,
+      "step": 900
+    },
+    {
+      "epoch": 0.58,
+      "learning_rate": 8.023697451204257e-06,
+      "loss": 0.2096,
+      "step": 901
+    },
+    {
+      "epoch": 0.58,
+      "learning_rate": 8.003387528001303e-06,
+      "loss": 0.1796,
+      "step": 902
+    },
+    {
+      "epoch": 0.58,
+      "learning_rate": 7.983086179025148e-06,
+      "loss": 0.2585,
+      "step": 903
+    },
+    {
+      "epoch": 0.58,
+      "learning_rate": 7.962793491457644e-06,
+      "loss": 0.1851,
+      "step": 904
+    },
+    {
+      "epoch": 0.58,
+      "learning_rate": 7.942509552443445e-06,
+      "loss": 0.1994,
+      "step": 905
+    },
+    {
+      "epoch": 0.58,
+      "learning_rate": 7.922234449089638e-06,
+      "loss": 0.226,
+      "step": 906
+    },
+    {
+      "epoch": 0.58,
+      "learning_rate": 7.901968268465367e-06,
+      "loss": 0.241,
+      "step": 907
+    },
+    {
+      "epoch": 0.58,
+      "learning_rate": 7.881711097601453e-06,
+      "loss": 0.1748,
+      "step": 908
+    },
+    {
+      "epoch": 0.58,
+      "learning_rate": 7.86146302349003e-06,
+      "loss": 0.2262,
+      "step": 909
+    },
+    {
+      "epoch": 0.58,
+      "learning_rate": 7.841224133084166e-06,
+      "loss": 0.2176,
+      "step": 910
+    },
+    {
+      "epoch": 0.58,
+      "learning_rate": 7.820994513297484e-06,
+      "loss": 0.2328,
+      "step": 911
+    },
+    {
+      "epoch": 0.58,
+      "learning_rate": 7.800774251003811e-06,
+      "loss": 0.1894,
+      "step": 912
+    },
+    {
+      "epoch": 0.58,
+      "learning_rate": 7.78056343303678e-06,
+      "loss": 0.1924,
+      "step": 913
+    },
+    {
+      "epoch": 0.58,
+      "learning_rate": 7.760362146189463e-06,
+      "loss": 0.2255,
+      "step": 914
+    },
+    {
+      "epoch": 0.59,
+      "learning_rate": 7.740170477214004e-06,
+      "loss": 0.2165,
+      "step": 915
+    },
+    {
+      "epoch": 0.59,
+      "learning_rate": 7.719988512821247e-06,
+      "loss": 0.19,
+      "step": 916
+    },
+    {
+      "epoch": 0.59,
+      "learning_rate": 7.699816339680357e-06,
+      "loss": 0.2195,
+      "step": 917
+    },
+    {
+      "epoch": 0.59,
+      "learning_rate": 7.679654044418456e-06,
+      "loss": 0.2023,
+      "step": 918
+    },
+    {
+      "epoch": 0.59,
+      "learning_rate": 7.659501713620247e-06,
+      "loss": 0.2099,
+      "step": 919
+    },
+    {
+      "epoch": 0.59,
+      "learning_rate": 7.639359433827634e-06,
+      "loss": 0.1745,
+      "step": 920
+    },
+    {
+      "epoch": 0.59,
+      "learning_rate": 7.619227291539365e-06,
+      "loss": 0.1974,
+      "step": 921
+    },
+    {
+      "epoch": 0.59,
+      "learning_rate": 7.5991053732106526e-06,
+      "loss": 0.2028,
+      "step": 922
+    },
+    {
+      "epoch": 0.59,
+      "learning_rate": 7.578993765252799e-06,
+      "loss": 0.1714,
+      "step": 923
+    },
+    {
+      "epoch": 0.59,
+      "learning_rate": 7.558892554032838e-06,
+      "loss": 0.1764,
+      "step": 924
+    },
+    {
+      "epoch": 0.59,
+      "learning_rate": 7.538801825873151e-06,
+      "loss": 0.2114,
+      "step": 925
+    },
+    {
+      "epoch": 0.59,
+      "learning_rate": 7.5187216670511e-06,
+      "loss": 0.1754,
+      "step": 926
+    },
+    {
+      "epoch": 0.59,
+      "learning_rate": 7.498652163798659e-06,
+      "loss": 0.197,
+      "step": 927
+    },
+    {
+      "epoch": 0.59,
+      "learning_rate": 7.4785934023020454e-06,
+      "loss": 0.1991,
+      "step": 928
+    },
+    {
+      "epoch": 0.59,
+      "learning_rate": 7.458545468701342e-06,
+      "loss": 0.2345,
+      "step": 929
+    },
+    {
+      "epoch": 0.6,
+      "learning_rate": 7.438508449090138e-06,
+      "loss": 0.2088,
+      "step": 930
+    },
+    {
+      "epoch": 0.6,
+      "learning_rate": 7.418482429515153e-06,
+      "loss": 0.1949,
+      "step": 931
+    },
+    {
+      "epoch": 0.6,
+      "learning_rate": 7.398467495975865e-06,
+      "loss": 0.2169,
+      "step": 932
+    },
+    {
+      "epoch": 0.6,
+      "learning_rate": 7.378463734424148e-06,
+      "loss": 0.2153,
+      "step": 933
+    },
+    {
+      "epoch": 0.6,
+      "learning_rate": 7.358471230763894e-06,
+      "loss": 0.2124,
+      "step": 934
+    },
+    {
+      "epoch": 0.6,
+      "learning_rate": 7.338490070850649e-06,
+      "loss": 0.2029,
+      "step": 935
+    },
+    {
+      "epoch": 0.6,
+      "learning_rate": 7.318520340491256e-06,
+      "loss": 0.2351,
+      "step": 936
+    },
+    {
+      "epoch": 0.6,
+      "learning_rate": 7.298562125443467e-06,
+      "loss": 0.2023,
+      "step": 937
+    },
+    {
+      "epoch": 0.6,
+      "learning_rate": 7.2786155114155765e-06,
+      "loss": 0.2098,
+      "step": 938
+    },
+    {
+      "epoch": 0.6,
+      "learning_rate": 7.258680584066069e-06,
+      "loss": 0.1866,
+      "step": 939
+    },
+    {
+      "epoch": 0.6,
+      "learning_rate": 7.23875742900324e-06,
+      "loss": 0.2195,
+      "step": 940
+    },
+    {
+      "epoch": 0.6,
+      "learning_rate": 7.218846131784824e-06,
+      "loss": 0.2184,
+      "step": 941
+    },
+    {
+      "epoch": 0.6,
+      "learning_rate": 7.198946777917646e-06,
+      "loss": 0.1992,
+      "step": 942
+    },
+    {
+      "epoch": 0.6,
+      "learning_rate": 7.17905945285723e-06,
+      "loss": 0.1672,
+      "step": 943
+    },
+    {
+      "epoch": 0.6,
+      "learning_rate": 7.159184242007449e-06,
+      "loss": 0.1796,
+      "step": 944
+    },
+    {
+      "epoch": 0.6,
+      "learning_rate": 7.139321230720151e-06,
+      "loss": 0.1996,
+      "step": 945
+    },
+    {
+      "epoch": 0.61,
+      "learning_rate": 7.1194705042947945e-06,
+      "loss": 0.2178,
+      "step": 946
+    },
+    {
+      "epoch": 0.61,
+      "learning_rate": 7.09963214797808e-06,
+      "loss": 0.2082,
+      "step": 947
+    },
+    {
+      "epoch": 0.61,
+      "learning_rate": 7.079806246963596e-06,
+      "loss": 0.1975,
+      "step": 948
+    },
+    {
+      "epoch": 0.61,
+      "learning_rate": 7.059992886391429e-06,
+      "loss": 0.1721,
+      "step": 949
+    },
+    {
+      "epoch": 0.61,
+      "learning_rate": 7.040192151347822e-06,
+      "loss": 0.1827,
+      "step": 950
+    },
+    {
+      "epoch": 0.61,
+      "learning_rate": 7.020404126864795e-06,
+      "loss": 0.2102,
+      "step": 951
+    },
+    {
+      "epoch": 0.61,
+      "learning_rate": 7.000628897919784e-06,
+      "loss": 0.2007,
+      "step": 952
+    },
+    {
+      "epoch": 0.61,
+      "learning_rate": 6.980866549435274e-06,
+      "loss": 0.2254,
+      "step": 953
+    },
+    {
+      "epoch": 0.61,
+      "learning_rate": 6.961117166278446e-06,
+      "loss": 0.188,
+      "step": 954
+    },
+    {
+      "epoch": 0.61,
+      "learning_rate": 6.94138083326079e-06,
+      "loss": 0.1973,
+      "step": 955
+    },
+    {
+      "epoch": 0.61,
+      "learning_rate": 6.921657635137762e-06,
+      "loss": 0.2365,
+      "step": 956
+    },
+    {
+      "epoch": 0.61,
+      "learning_rate": 6.90194765660841e-06,
+      "loss": 0.199,
+      "step": 957
+    },
+    {
+      "epoch": 0.61,
+      "learning_rate": 6.882250982315006e-06,
+      "loss": 0.2193,
+      "step": 958
+    },
+    {
+      "epoch": 0.61,
+      "learning_rate": 6.862567696842694e-06,
+      "loss": 0.1749,
+      "step": 959
+    },
+    {
+      "epoch": 0.61,
+      "learning_rate": 6.842897884719128e-06,
+      "loss": 0.227,
+      "step": 960
+    },
+    {
+      "epoch": 0.61,
+      "learning_rate": 6.823241630414095e-06,
+      "loss": 0.2001,
+      "step": 961
+    },
+    {
+      "epoch": 0.62,
+      "learning_rate": 6.803599018339151e-06,
+      "loss": 0.2254,
+      "step": 962
+    },
+    {
+      "epoch": 0.62,
+      "learning_rate": 6.78397013284728e-06,
+      "loss": 0.1886,
+      "step": 963
+    },
+    {
+      "epoch": 0.62,
+      "learning_rate": 6.764355058232515e-06,
+      "loss": 0.2081,
+      "step": 964
+    },
+    {
+      "epoch": 0.62,
+      "learning_rate": 6.744753878729574e-06,
+      "loss": 0.2229,
+      "step": 965
+    },
+    {
+      "epoch": 0.62,
+      "learning_rate": 6.725166678513516e-06,
+      "loss": 0.2054,
+      "step": 966
+    },
+    {
+      "epoch": 0.62,
+      "learning_rate": 6.705593541699358e-06,
+      "loss": 0.1915,
+      "step": 967
+    },
+    {
+      "epoch": 0.62,
+      "learning_rate": 6.6860345523417244e-06,
+      "loss": 0.2239,
+      "step": 968
+    },
+    {
+      "epoch": 0.62,
+      "learning_rate": 6.666489794434488e-06,
+      "loss": 0.2165,
+      "step": 969
+    },
+    {
+      "epoch": 0.62,
+      "learning_rate": 6.646959351910403e-06,
+      "loss": 0.2146,
+      "step": 970
+    },
+    {
+      "epoch": 0.62,
+      "learning_rate": 6.627443308640749e-06,
+      "loss": 0.2063,
+      "step": 971
+    },
+    {
+      "epoch": 0.62,
+      "learning_rate": 6.6079417484349736e-06,
+      "loss": 0.2239,
+      "step": 972
+    },
+    {
+      "epoch": 0.62,
+      "learning_rate": 6.588454755040322e-06,
+      "loss": 0.1693,
+      "step": 973
+    },
+    {
+      "epoch": 0.62,
+      "learning_rate": 6.568982412141488e-06,
+      "loss": 0.2201,
+      "step": 974
+    },
+    {
+      "epoch": 0.62,
+      "learning_rate": 6.549524803360249e-06,
+      "loss": 0.195,
+      "step": 975
+    },
+    {
+      "epoch": 0.62,
+      "learning_rate": 6.530082012255108e-06,
+      "loss": 0.2119,
+      "step": 976
+    },
+    {
+      "epoch": 0.63,
+      "learning_rate": 6.5106541223209344e-06,
+      "loss": 0.2035,
+      "step": 977
+    },
+    {
+      "epoch": 0.63,
+      "learning_rate": 6.491241216988612e-06,
+      "loss": 0.2095,
+      "step": 978
+    },
+    {
+      "epoch": 0.63,
+      "learning_rate": 6.4718433796246685e-06,
+      "loss": 0.2201,
+      "step": 979
+    },
+    {
+      "epoch": 0.63,
+      "learning_rate": 6.452460693530925e-06,
+      "loss": 0.1818,
+      "step": 980
+    },
+    {
+      "epoch": 0.63,
+      "learning_rate": 6.433093241944142e-06,
+      "loss": 0.164,
+      "step": 981
+    },
+    {
+      "epoch": 0.63,
+      "learning_rate": 6.413741108035648e-06,
+      "loss": 0.2041,
+      "step": 982
+    },
+    {
+      "epoch": 0.63,
+      "learning_rate": 6.394404374910996e-06,
+      "loss": 0.2022,
+      "step": 983
+    },
+    {
+      "epoch": 0.63,
+      "learning_rate": 6.3750831256096094e-06,
+      "loss": 0.2063,
+      "step": 984
+    },
+    {
+      "epoch": 0.63,
+      "learning_rate": 6.355777443104409e-06,
+      "loss": 0.2461,
+      "step": 985
+    },
+    {
+      "epoch": 0.63,
+      "learning_rate": 6.336487410301468e-06,
+      "loss": 0.2137,
+      "step": 986
+    },
+    {
+      "epoch": 0.63,
+      "learning_rate": 6.317213110039652e-06,
+      "loss": 0.2263,
+      "step": 987
+    },
+    {
+      "epoch": 0.63,
+      "learning_rate": 6.297954625090268e-06,
+      "loss": 0.2295,
+      "step": 988
+    },
+    {
+      "epoch": 0.63,
+      "learning_rate": 6.278712038156705e-06,
+      "loss": 0.2162,
+      "step": 989
+    },
+    {
+      "epoch": 0.63,
+      "learning_rate": 6.2594854318740806e-06,
+      "loss": 0.1983,
+      "step": 990
+    },
+    {
+      "epoch": 0.63,
+      "learning_rate": 6.240274888808884e-06,
+      "loss": 0.2326,
+      "step": 991
+    },
+    {
+      "epoch": 0.63,
+      "learning_rate": 6.221080491458624e-06,
+      "loss": 0.1879,
+      "step": 992
+    },
+    {
+      "epoch": 0.64,
+      "learning_rate": 6.201902322251471e-06,
+      "loss": 0.2191,
+      "step": 993
+    },
+    {
+      "epoch": 0.64,
+      "learning_rate": 6.182740463545908e-06,
+      "loss": 0.1992,
+      "step": 994
+    },
+    {
+      "epoch": 0.64,
+      "learning_rate": 6.16359499763037e-06,
+      "loss": 0.1968,
+      "step": 995
+    },
+    {
+      "epoch": 0.64,
+      "learning_rate": 6.1444660067229044e-06,
+      "loss": 0.2155,
+      "step": 996
+    },
+    {
+      "epoch": 0.64,
+      "learning_rate": 6.125353572970798e-06,
+      "loss": 0.1901,
+      "step": 997
+    },
+    {
+      "epoch": 0.64,
+      "learning_rate": 6.106257778450242e-06,
+      "loss": 0.203,
+      "step": 998
+    },
+    {
+      "epoch": 0.64,
+      "learning_rate": 6.087178705165969e-06,
+      "loss": 0.2054,
+      "step": 999
+    },
+    {
+      "epoch": 0.64,
+      "learning_rate": 6.068116435050903e-06,
+      "loss": 0.1916,
+      "step": 1000
+    },
+    {
+      "epoch": 0.64,
+      "learning_rate": 6.049071049965811e-06,
+      "loss": 0.2317,
+      "step": 1001
+    },
+    {
+      "epoch": 0.64,
+      "learning_rate": 6.030042631698952e-06,
+      "loss": 0.2137,
+      "step": 1002
+    },
+    {
+      "epoch": 0.64,
+      "learning_rate": 6.011031261965716e-06,
+      "loss": 0.2333,
+      "step": 1003
+    },
+    {
+      "epoch": 0.64,
+      "learning_rate": 5.992037022408285e-06,
+      "loss": 0.1953,
+      "step": 1004
+    },
+    {
+      "epoch": 0.64,
+      "learning_rate": 5.973059994595277e-06,
+      "loss": 0.1874,
+      "step": 1005
+    },
+    {
+      "epoch": 0.64,
+      "learning_rate": 5.954100260021392e-06,
+      "loss": 0.1781,
+      "step": 1006
+    },
+    {
+      "epoch": 0.64,
+      "learning_rate": 5.9351579001070655e-06,
+      "loss": 0.2099,
+      "step": 1007
+    },
+    {
+      "epoch": 0.64,
+      "learning_rate": 5.916232996198132e-06,
+      "loss": 0.2212,
+      "step": 1008
+    },
+    {
+      "epoch": 0.65,
+      "learning_rate": 5.897325629565455e-06,
+      "loss": 0.2052,
+      "step": 1009
+    },
+    {
+      "epoch": 0.65,
+      "learning_rate": 5.878435881404576e-06,
+      "loss": 0.2043,
+      "step": 1010
+    },
+    {
+      "epoch": 0.65,
+      "learning_rate": 5.859563832835393e-06,
+      "loss": 0.2111,
+      "step": 1011
+    },
+    {
+      "epoch": 0.65,
+      "learning_rate": 5.840709564901783e-06,
+      "loss": 0.1983,
+      "step": 1012
+    },
+    {
+      "epoch": 0.65,
+      "learning_rate": 5.8218731585712716e-06,
+      "loss": 0.203,
+      "step": 1013
+    },
+    {
+      "epoch": 0.65,
+      "learning_rate": 5.80305469473468e-06,
+      "loss": 0.2029,
+      "step": 1014
+    },
+    {
+      "epoch": 0.65,
+      "learning_rate": 5.784254254205779e-06,
+      "loss": 0.2121,
+      "step": 1015
+    },
+    {
+      "epoch": 0.65,
+      "learning_rate": 5.76547191772093e-06,
+      "loss": 0.1975,
+      "step": 1016
+    },
+    {
+      "epoch": 0.65,
+      "learning_rate": 5.746707765938764e-06,
+      "loss": 0.2069,
+      "step": 1017
+    },
+    {
+      "epoch": 0.65,
+      "learning_rate": 5.7279618794398095e-06,
+      "loss": 0.1754,
+      "step": 1018
+    },
+    {
+      "epoch": 0.65,
+      "learning_rate": 5.70923433872615e-06,
+      "loss": 0.2156,
+      "step": 1019
+    },
+    {
+      "epoch": 0.65,
+      "learning_rate": 5.6905252242211095e-06,
+      "loss": 0.2064,
+      "step": 1020
+    },
+    {
+      "epoch": 0.65,
+      "learning_rate": 5.671834616268861e-06,
+      "loss": 0.1992,
+      "step": 1021
+    },
+    {
+      "epoch": 0.65,
+      "learning_rate": 5.653162595134105e-06,
+      "loss": 0.1741,
+      "step": 1022
+    },
+    {
+      "epoch": 0.65,
+      "learning_rate": 5.634509241001736e-06,
+      "loss": 0.1752,
+      "step": 1023
+    },
+    {
+      "epoch": 0.66,
+      "learning_rate": 5.61587463397647e-06,
+      "loss": 0.2099,
+      "step": 1024
+    },
+    {
+      "epoch": 0.66,
+      "learning_rate": 5.5972588540825245e-06,
+      "loss": 0.2264,
+      "step": 1025
+    },
+    {
+      "epoch": 0.66,
+      "learning_rate": 5.57866198126327e-06,
+      "loss": 0.185,
+      "step": 1026
+    },
+    {
+      "epoch": 0.66,
+      "learning_rate": 5.5600840953808675e-06,
+      "loss": 0.2077,
+      "step": 1027
+    },
+    {
+      "epoch": 0.66,
+      "learning_rate": 5.5415252762159595e-06,
+      "loss": 0.2165,
+      "step": 1028
+    },
+    {
+      "epoch": 0.66,
+      "learning_rate": 5.52298560346729e-06,
+      "loss": 0.2187,
+      "step": 1029
+    },
+    {
+      "epoch": 0.66,
+      "learning_rate": 5.5044651567513985e-06,
+      "loss": 0.1997,
+      "step": 1030
+    },
+    {
+      "epoch": 0.66,
+      "learning_rate": 5.485964015602243e-06,
+      "loss": 0.1857,
+      "step": 1031
+    },
+    {
+      "epoch": 0.66,
+      "learning_rate": 5.467482259470891e-06,
+      "loss": 0.1843,
+      "step": 1032
+    },
+    {
+      "epoch": 0.66,
+      "learning_rate": 5.449019967725161e-06,
+      "loss": 0.2058,
+      "step": 1033
+    },
+    {
+      "epoch": 0.66,
+      "learning_rate": 5.43057721964927e-06,
+      "loss": 0.2244,
+      "step": 1034
+    },
+    {
+      "epoch": 0.66,
+      "learning_rate": 5.412154094443527e-06,
+      "loss": 0.2049,
+      "step": 1035
+    },
+    {
+      "epoch": 0.66,
+      "learning_rate": 5.393750671223954e-06,
+      "loss": 0.2348,
+      "step": 1036
+    },
+    {
+      "epoch": 0.66,
+      "learning_rate": 5.375367029021979e-06,
+      "loss": 0.2008,
+      "step": 1037
+    },
+    {
+      "epoch": 0.66,
+      "learning_rate": 5.35700324678408e-06,
+      "loss": 0.1849,
+      "step": 1038
+    },
+    {
+      "epoch": 0.66,
+      "learning_rate": 5.338659403371438e-06,
+      "loss": 0.1906,
+      "step": 1039
+    },
+    {
+      "epoch": 0.67,
+      "learning_rate": 5.320335577559626e-06,
+      "loss": 0.1969,
+      "step": 1040
+    },
+    {
+      "epoch": 0.67,
+      "learning_rate": 5.302031848038241e-06,
+      "loss": 0.2276,
+      "step": 1041
+    },
+    {
+      "epoch": 0.67,
+      "learning_rate": 5.283748293410578e-06,
+      "loss": 0.1977,
+      "step": 1042
+    },
+    {
+      "epoch": 0.67,
+      "learning_rate": 5.265484992193302e-06,
+      "loss": 0.1982,
+      "step": 1043
+    },
+    {
+      "epoch": 0.67,
+      "learning_rate": 5.247242022816098e-06,
+      "loss": 0.1928,
+      "step": 1044
+    },
+    {
+      "epoch": 0.67,
+      "learning_rate": 5.229019463621341e-06,
+      "loss": 0.2242,
+      "step": 1045
+    },
+    {
+      "epoch": 0.67,
+      "learning_rate": 5.210817392863751e-06,
+      "loss": 0.2169,
+      "step": 1046
+    },
+    {
+      "epoch": 0.67,
+      "learning_rate": 5.19263588871006e-06,
+      "loss": 0.2085,
+      "step": 1047
+    },
+    {
+      "epoch": 0.67,
+      "learning_rate": 5.174475029238695e-06,
+      "loss": 0.1851,
+      "step": 1048
+    },
+    {
+      "epoch": 0.67,
+      "learning_rate": 5.156334892439405e-06,
+      "loss": 0.2068,
+      "step": 1049
+    },
+    {
+      "epoch": 0.67,
+      "learning_rate": 5.138215556212966e-06,
+      "loss": 0.2179,
+      "step": 1050
+    },
+    {
+      "epoch": 0.67,
+      "learning_rate": 5.120117098370824e-06,
+      "loss": 0.2341,
+      "step": 1051
+    },
+    {
+      "epoch": 0.67,
+      "learning_rate": 5.1020395966347534e-06,
+      "loss": 0.225,
+      "step": 1052
+    },
+    {
+      "epoch": 0.67,
+      "learning_rate": 5.0839831286365535e-06,
+      "loss": 0.2264,
+      "step": 1053
+    },
+    {
+      "epoch": 0.67,
+      "learning_rate": 5.065947771917679e-06,
+      "loss": 0.2116,
+      "step": 1054
+    },
+    {
+      "epoch": 0.67,
+      "learning_rate": 5.047933603928936e-06,
+      "loss": 0.2093,
+      "step": 1055
+    },
+    {
+      "epoch": 0.68,
+      "learning_rate": 5.02994070203014e-06,
+      "loss": 0.1985,
+      "step": 1056
+    },
+    {
+      "epoch": 0.68,
+      "learning_rate": 5.011969143489769e-06,
+      "loss": 0.2173,
+      "step": 1057
+    },
+    {
+      "epoch": 0.68,
+      "learning_rate": 4.994019005484657e-06,
+      "loss": 0.2366,
+      "step": 1058
+    },
+    {
+      "epoch": 0.68,
+      "learning_rate": 4.976090365099638e-06,
+      "loss": 0.2069,
+      "step": 1059
+    },
+    {
+      "epoch": 0.68,
+      "learning_rate": 4.958183299327243e-06,
+      "loss": 0.2124,
+      "step": 1060
+    },
+    {
+      "epoch": 0.68,
+      "learning_rate": 4.940297885067333e-06,
+      "loss": 0.1906,
+      "step": 1061
+    },
+    {
+      "epoch": 0.68,
+      "learning_rate": 4.9224341991268065e-06,
+      "loss": 0.1871,
+      "step": 1062
+    },
+    {
+      "epoch": 0.68,
+      "learning_rate": 4.904592318219249e-06,
+      "loss": 0.2158,
+      "step": 1063
+    },
+    {
+      "epoch": 0.68,
+      "learning_rate": 4.8867723189645945e-06,
+      "loss": 0.2334,
+      "step": 1064
+    },
+    {
+      "epoch": 0.68,
+      "learning_rate": 4.868974277888826e-06,
+      "loss": 0.1841,
+      "step": 1065
+    },
+    {
+      "epoch": 0.68,
+      "learning_rate": 4.851198271423617e-06,
+      "loss": 0.2119,
+      "step": 1066
+    },
+    {
+      "epoch": 0.68,
+      "learning_rate": 4.833444375906013e-06,
+      "loss": 0.1811,
+      "step": 1067
+    },
+    {
+      "epoch": 0.68,
+      "learning_rate": 4.815712667578128e-06,
+      "loss": 0.1728,
+      "step": 1068
+    },
+    {
+      "epoch": 0.68,
+      "learning_rate": 4.798003222586773e-06,
+      "loss": 0.1904,
+      "step": 1069
+    },
+    {
+      "epoch": 0.68,
+      "learning_rate": 4.780316116983157e-06,
+      "loss": 0.2109,
+      "step": 1070
+    },
+    {
+      "epoch": 0.69,
+      "learning_rate": 4.762651426722566e-06,
+      "loss": 0.215,
+      "step": 1071
+    },
+    {
+      "epoch": 0.69,
+      "learning_rate": 4.745009227664007e-06,
+      "loss": 0.1876,
+      "step": 1072
+    },
+    {
+      "epoch": 0.69,
+      "learning_rate": 4.727389595569919e-06,
+      "loss": 0.2372,
+      "step": 1073
+    },
+    {
+      "epoch": 0.69,
+      "learning_rate": 4.709792606105824e-06,
+      "loss": 0.2414,
+      "step": 1074
+    },
+    {
+      "epoch": 0.69,
+      "learning_rate": 4.6922183348399996e-06,
+      "loss": 0.1861,
+      "step": 1075
+    },
+    {
+      "epoch": 0.69,
+      "learning_rate": 4.674666857243176e-06,
+      "loss": 0.1745,
+      "step": 1076
+    },
+    {
+      "epoch": 0.69,
+      "learning_rate": 4.657138248688185e-06,
+      "loss": 0.2229,
+      "step": 1077
+    },
+    {
+      "epoch": 0.69,
+      "learning_rate": 4.6396325844496644e-06,
+      "loss": 0.1892,
+      "step": 1078
+    },
+    {
+      "epoch": 0.69,
+      "learning_rate": 4.622149939703704e-06,
+      "loss": 0.2044,
+      "step": 1079
+    },
+    {
+      "epoch": 0.69,
+      "learning_rate": 4.6046903895275515e-06,
+      "loss": 0.1856,
+      "step": 1080
+    },
+    {
+      "epoch": 0.69,
+      "learning_rate": 4.587254008899278e-06,
+      "loss": 0.1902,
+      "step": 1081
+    },
+    {
+      "epoch": 0.69,
+      "learning_rate": 4.5698408726974434e-06,
+      "loss": 0.19,
+      "step": 1082
+    },
+    {
+      "epoch": 0.69,
+      "learning_rate": 4.552451055700802e-06,
+      "loss": 0.195,
+      "step": 1083
+    },
+    {
+      "epoch": 0.69,
+      "learning_rate": 4.535084632587952e-06,
+      "loss": 0.1724,
+      "step": 1084
+    },
+    {
+      "epoch": 0.69,
+      "learning_rate": 4.517741677937039e-06,
+      "loss": 0.1865,
+      "step": 1085
+    },
+    {
+      "epoch": 0.69,
+      "learning_rate": 4.500422266225428e-06,
+      "loss": 0.1755,
+      "step": 1086
+    },
+    {
+      "epoch": 0.7,
+      "learning_rate": 4.483126471829371e-06,
+      "loss": 0.2106,
+      "step": 1087
+    },
+    {
+      "epoch": 0.7,
+      "learning_rate": 4.465854369023709e-06,
+      "loss": 0.1948,
+      "step": 1088
+    },
+    {
+      "epoch": 0.7,
+      "learning_rate": 4.448606031981534e-06,
+      "loss": 0.2005,
+      "step": 1089
+    },
+    {
+      "epoch": 0.7,
+      "learning_rate": 4.431381534773879e-06,
+      "loss": 0.2009,
+      "step": 1090
+    },
+    {
+      "epoch": 0.7,
+      "learning_rate": 4.414180951369405e-06,
+      "loss": 0.2067,
+      "step": 1091
+    },
+    {
+      "epoch": 0.7,
+      "learning_rate": 4.397004355634075e-06,
+      "loss": 0.1667,
+      "step": 1092
+    },
+    {
+      "epoch": 0.7,
+      "learning_rate": 4.379851821330842e-06,
+      "loss": 0.1952,
+      "step": 1093
+    },
+    {
+      "epoch": 0.7,
+      "learning_rate": 4.362723422119329e-06,
+      "loss": 0.21,
+      "step": 1094
+    },
+    {
+      "epoch": 0.7,
+      "learning_rate": 4.345619231555503e-06,
+      "loss": 0.2052,
+      "step": 1095
+    },
+    {
+      "epoch": 0.7,
+      "learning_rate": 4.328539323091391e-06,
+      "loss": 0.1809,
+      "step": 1096
+    },
+    {
+      "epoch": 0.7,
+      "learning_rate": 4.3114837700747205e-06,
+      "loss": 0.17,
+      "step": 1097
+    },
+    {
+      "epoch": 0.7,
+      "learning_rate": 4.294452645748641e-06,
+      "loss": 0.2061,
+      "step": 1098
+    },
+    {
+      "epoch": 0.7,
+      "learning_rate": 4.277446023251401e-06,
+      "loss": 0.2041,
+      "step": 1099
+    },
+    {
+      "epoch": 0.7,
+      "learning_rate": 4.260463975616007e-06,
+      "loss": 0.1923,
+      "step": 1100
+    },
+    {
+      "epoch": 0.7,
+      "learning_rate": 4.243506575769958e-06,
+      "loss": 0.2123,
+      "step": 1101
+    },
+    {
+      "epoch": 0.71,
+      "learning_rate": 4.2265738965348815e-06,
+      "loss": 0.2169,
+      "step": 1102
+    },
+    {
+      "epoch": 0.71,
+      "learning_rate": 4.209666010626262e-06,
+      "loss": 0.1803,
+      "step": 1103
+    },
+    {
+      "epoch": 0.71,
+      "learning_rate": 4.192782990653109e-06,
+      "loss": 0.1981,
+      "step": 1104
+    },
+    {
+      "epoch": 0.71,
+      "learning_rate": 4.175924909117638e-06,
+      "loss": 0.1958,
+      "step": 1105
+    },
+    {
+      "epoch": 0.71,
+      "learning_rate": 4.159091838414985e-06,
+      "loss": 0.2167,
+      "step": 1106
+    },
+    {
+      "epoch": 0.71,
+      "learning_rate": 4.142283850832862e-06,
+      "loss": 0.2132,
+      "step": 1107
+    },
+    {
+      "epoch": 0.71,
+      "learning_rate": 4.125501018551284e-06,
+      "loss": 0.1935,
+      "step": 1108
+    },
+    {
+      "epoch": 0.71,
+      "learning_rate": 4.108743413642219e-06,
+      "loss": 0.2129,
+      "step": 1109
+    },
+    {
+      "epoch": 0.71,
+      "learning_rate": 4.092011108069313e-06,
+      "loss": 0.207,
+      "step": 1110
+    },
+    {
+      "epoch": 0.71,
+      "learning_rate": 4.0753041736875675e-06,
+      "loss": 0.2149,
+      "step": 1111
+    },
+    {
+      "epoch": 0.71,
+      "learning_rate": 4.0586226822430175e-06,
+      "loss": 0.1846,
+      "step": 1112
+    },
+    {
+      "epoch": 0.71,
+      "learning_rate": 4.041966705372453e-06,
+      "loss": 0.1954,
+      "step": 1113
+    },
+    {
+      "epoch": 0.71,
+      "learning_rate": 4.02533631460308e-06,
+      "loss": 0.1982,
+      "step": 1114
+    },
+    {
+      "epoch": 0.71,
+      "learning_rate": 4.008731581352229e-06,
+      "loss": 0.189,
+      "step": 1115
+    },
+    {
+      "epoch": 0.71,
+      "learning_rate": 3.9921525769270644e-06,
+      "loss": 0.244,
+      "step": 1116
+    },
+    {
+      "epoch": 0.71,
+      "learning_rate": 3.975599372524242e-06,
+      "loss": 0.1967,
+      "step": 1117
+    },
+    {
+      "epoch": 0.72,
+      "learning_rate": 3.959072039229625e-06,
+      "loss": 0.1957,
+      "step": 1118
+    },
+    {
+      "epoch": 0.72,
+      "learning_rate": 3.942570648017988e-06,
+      "loss": 0.2084,
+      "step": 1119
+    },
+    {
+      "epoch": 0.72,
+      "learning_rate": 3.9260952697526846e-06,
+      "loss": 0.2001,
+      "step": 1120
+    },
+    {
+      "epoch": 0.72,
+      "learning_rate": 3.90964597518537e-06,
+      "loss": 0.1898,
+      "step": 1121
+    },
+    {
+      "epoch": 0.72,
+      "learning_rate": 3.893222834955684e-06,
+      "loss": 0.2277,
+      "step": 1122
+    },
+    {
+      "epoch": 0.72,
+      "learning_rate": 3.876825919590944e-06,
+      "loss": 0.2166,
+      "step": 1123
+    },
+    {
+      "epoch": 0.72,
+      "learning_rate": 3.860455299505856e-06,
+      "loss": 0.2075,
+      "step": 1124
+    },
+    {
+      "epoch": 0.72,
+      "learning_rate": 3.844111045002193e-06,
+      "loss": 0.1934,
+      "step": 1125
+    },
+    {
+      "epoch": 0.72,
+      "learning_rate": 3.827793226268518e-06,
+      "loss": 0.2147,
+      "step": 1126
+    },
+    {
+      "epoch": 0.72,
+      "learning_rate": 3.8115019133798535e-06,
+      "loss": 0.1768,
+      "step": 1127
+    },
+    {
+      "epoch": 0.72,
+      "learning_rate": 3.7952371762974063e-06,
+      "loss": 0.2011,
+      "step": 1128
+    },
+    {
+      "epoch": 0.72,
+      "learning_rate": 3.778999084868257e-06,
+      "loss": 0.2188,
+      "step": 1129
+    },
+    {
+      "epoch": 0.72,
+      "learning_rate": 3.7627877088250474e-06,
+      "loss": 0.2314,
+      "step": 1130
+    },
+    {
+      "epoch": 0.72,
+      "learning_rate": 3.74660311778571e-06,
+      "loss": 0.2162,
+      "step": 1131
+    },
+    {
+      "epoch": 0.72,
+      "learning_rate": 3.7304453812531393e-06,
+      "loss": 0.1876,
+      "step": 1132
+    },
+    {
+      "epoch": 0.72,
+      "learning_rate": 3.714314568614904e-06,
+      "loss": 0.1933,
+      "step": 1133
+    },
+    {
+      "epoch": 0.73,
+      "learning_rate": 3.6982107491429698e-06,
+      "loss": 0.1822,
+      "step": 1134
+    },
+    {
+      "epoch": 0.73,
+      "learning_rate": 3.682133991993361e-06,
+      "loss": 0.1886,
+      "step": 1135
+    },
+    {
+      "epoch": 0.73,
+      "learning_rate": 3.6660843662059033e-06,
+      "loss": 0.235,
+      "step": 1136
+    },
+    {
+      "epoch": 0.73,
+      "learning_rate": 3.6500619407038996e-06,
+      "loss": 0.2222,
+      "step": 1137
+    },
+    {
+      "epoch": 0.73,
+      "learning_rate": 3.6340667842938415e-06,
+      "loss": 0.1989,
+      "step": 1138
+    },
+    {
+      "epoch": 0.73,
+      "learning_rate": 3.6180989656651268e-06,
+      "loss": 0.1838,
+      "step": 1139
+    },
+    {
+      "epoch": 0.73,
+      "learning_rate": 3.6021585533897486e-06,
+      "loss": 0.1677,
+      "step": 1140
+    },
+    {
+      "epoch": 0.73,
+      "learning_rate": 3.5862456159220115e-06,
+      "loss": 0.2096,
+      "step": 1141
+    },
+    {
+      "epoch": 0.73,
+      "learning_rate": 3.5703602215982257e-06,
+      "loss": 0.2017,
+      "step": 1142
+    },
+    {
+      "epoch": 0.73,
+      "learning_rate": 3.5545024386364192e-06,
+      "loss": 0.1712,
+      "step": 1143
+    },
+    {
+      "epoch": 0.73,
+      "learning_rate": 3.538672335136056e-06,
+      "loss": 0.1908,
+      "step": 1144
+    },
+    {
+      "epoch": 0.73,
+      "learning_rate": 3.522869979077723e-06,
+      "loss": 0.2119,
+      "step": 1145
+    },
+    {
+      "epoch": 0.73,
+      "learning_rate": 3.507095438322856e-06,
+      "loss": 0.1873,
+      "step": 1146
+    },
+    {
+      "epoch": 0.73,
+      "learning_rate": 3.491348780613444e-06,
+      "loss": 0.1899,
+      "step": 1147
+    },
+    {
+      "epoch": 0.73,
+      "learning_rate": 3.4756300735717218e-06,
+      "loss": 0.1787,
+      "step": 1148
+    },
+    {
+      "epoch": 0.74,
+      "learning_rate": 3.4599393846999087e-06,
+      "loss": 0.2056,
+      "step": 1149
+    },
+    {
+      "epoch": 0.74,
+      "learning_rate": 3.4442767813798906e-06,
+      "loss": 0.1966,
+      "step": 1150
+    },
+    {
+      "epoch": 0.74,
+      "learning_rate": 3.4286423308729523e-06,
+      "loss": 0.2128,
+      "step": 1151
+    },
+    {
+      "epoch": 0.74,
+      "learning_rate": 3.413036100319478e-06,
+      "loss": 0.1947,
+      "step": 1152
+    },
+    {
+      "epoch": 0.74,
+      "learning_rate": 3.3974581567386567e-06,
+      "loss": 0.1903,
+      "step": 1153
+    },
+    {
+      "epoch": 0.74,
+      "learning_rate": 3.3819085670282147e-06,
+      "loss": 0.2181,
+      "step": 1154
+    },
+    {
+      "epoch": 0.74,
+      "learning_rate": 3.3663873979641014e-06,
+      "loss": 0.1909,
+      "step": 1155
+    },
+    {
+      "epoch": 0.74,
+      "learning_rate": 3.350894716200234e-06,
+      "loss": 0.1974,
+      "step": 1156
+    },
+    {
+      "epoch": 0.74,
+      "learning_rate": 3.3354305882681747e-06,
+      "loss": 0.1767,
+      "step": 1157
+    },
+    {
+      "epoch": 0.74,
+      "learning_rate": 3.319995080576881e-06,
+      "loss": 0.1913,
+      "step": 1158
+    },
+    {
+      "epoch": 0.74,
+      "learning_rate": 3.304588259412399e-06,
+      "loss": 0.1656,
+      "step": 1159
+    },
+    {
+      "epoch": 0.74,
+      "learning_rate": 3.2892101909375775e-06,
+      "loss": 0.1884,
+      "step": 1160
+    },
+    {
+      "epoch": 0.74,
+      "learning_rate": 3.2738609411918e-06,
+      "loss": 0.206,
+      "step": 1161
+    },
+    {
+      "epoch": 0.74,
+      "learning_rate": 3.2585405760906842e-06,
+      "loss": 0.2313,
+      "step": 1162
+    },
+    {
+      "epoch": 0.74,
+      "learning_rate": 3.243249161425801e-06,
+      "loss": 0.201,
+      "step": 1163
+    },
+    {
+      "epoch": 0.74,
+      "learning_rate": 3.227986762864417e-06,
+      "loss": 0.2028,
+      "step": 1164
+    },
+    {
+      "epoch": 0.75,
+      "learning_rate": 3.2127534459491715e-06,
+      "loss": 0.1952,
+      "step": 1165
+    },
+    {
+      "epoch": 0.75,
+      "learning_rate": 3.1975492760978233e-06,
+      "loss": 0.2166,
+      "step": 1166
+    },
+    {
+      "epoch": 0.75,
+      "learning_rate": 3.1823743186029676e-06,
+      "loss": 0.2012,
+      "step": 1167
+    },
+    {
+      "epoch": 0.75,
+      "learning_rate": 3.16722863863174e-06,
+      "loss": 0.2136,
+      "step": 1168
+    },
+    {
+      "epoch": 0.75,
+      "learning_rate": 3.1521123012255563e-06,
+      "loss": 0.1911,
+      "step": 1169
+    },
+    {
+      "epoch": 0.75,
+      "learning_rate": 3.137025371299822e-06,
+      "loss": 0.1659,
+      "step": 1170
+    },
+    {
+      "epoch": 0.75,
+      "learning_rate": 3.1219679136436496e-06,
+      "loss": 0.1906,
+      "step": 1171
+    },
+    {
+      "epoch": 0.75,
+      "learning_rate": 3.1069399929195966e-06,
+      "loss": 0.2049,
+      "step": 1172
+    },
+    {
+      "epoch": 0.75,
+      "learning_rate": 3.0919416736633643e-06,
+      "loss": 0.1677,
+      "step": 1173
+    },
+    {
+      "epoch": 0.75,
+      "learning_rate": 3.0769730202835493e-06,
+      "loss": 0.2085,
+      "step": 1174
+    },
+    {
+      "epoch": 0.75,
+      "learning_rate": 3.062034097061335e-06,
+      "loss": 0.1877,
+      "step": 1175
+    },
+    {
+      "epoch": 0.75,
+      "learning_rate": 3.0471249681502445e-06,
+      "loss": 0.1964,
+      "step": 1176
+    },
+    {
+      "epoch": 0.75,
+      "learning_rate": 3.0322456975758507e-06,
+      "loss": 0.2016,
+      "step": 1177
+    },
+    {
+      "epoch": 0.75,
+      "learning_rate": 3.0173963492354954e-06,
+      "loss": 0.2094,
+      "step": 1178
+    },
+    {
+      "epoch": 0.75,
+      "learning_rate": 3.0025769868980337e-06,
+      "loss": 0.2255,
+      "step": 1179
+    },
+    {
+      "epoch": 0.75,
+      "learning_rate": 2.9877876742035416e-06,
+      "loss": 0.1795,
+      "step": 1180
+    },
+    {
+      "epoch": 0.76,
+      "learning_rate": 2.9730284746630454e-06,
+      "loss": 0.2321,
+      "step": 1181
+    },
+    {
+      "epoch": 0.76,
+      "learning_rate": 2.958299451658274e-06,
+      "loss": 0.2118,
+      "step": 1182
+    },
+    {
+      "epoch": 0.76,
+      "learning_rate": 2.9436006684413444e-06,
+      "loss": 0.1996,
+      "step": 1183
+    },
+    {
+      "epoch": 0.76,
+      "learning_rate": 2.9289321881345257e-06,
+      "loss": 0.1799,
+      "step": 1184
+    },
+    {
+      "epoch": 0.76,
+      "learning_rate": 2.9142940737299486e-06,
+      "loss": 0.1873,
+      "step": 1185
+    },
+    {
+      "epoch": 0.76,
+      "learning_rate": 2.899686388089339e-06,
+      "loss": 0.1949,
+      "step": 1186
+    },
+    {
+      "epoch": 0.76,
+      "learning_rate": 2.88510919394376e-06,
+      "loss": 0.1921,
+      "step": 1187
+    },
+    {
+      "epoch": 0.76,
+      "learning_rate": 2.8705625538933158e-06,
+      "loss": 0.1975,
+      "step": 1188
+    },
+    {
+      "epoch": 0.76,
+      "learning_rate": 2.8560465304069217e-06,
+      "loss": 0.1958,
+      "step": 1189
+    },
+    {
+      "epoch": 0.76,
+      "learning_rate": 2.841561185821996e-06,
+      "loss": 0.1792,
+      "step": 1190
+    },
+    {
+      "epoch": 0.76,
+      "learning_rate": 2.827106582344212e-06,
+      "loss": 0.1772,
+      "step": 1191
+    },
+    {
+      "epoch": 0.76,
+      "learning_rate": 2.812682782047238e-06,
+      "loss": 0.1969,
+      "step": 1192
+    },
+    {
+      "epoch": 0.76,
+      "learning_rate": 2.7982898468724496e-06,
+      "loss": 0.1898,
+      "step": 1193
+    },
+    {
+      "epoch": 0.76,
+      "learning_rate": 2.783927838628684e-06,
+      "loss": 0.1882,
+      "step": 1194
+    },
+    {
+      "epoch": 0.76,
+      "learning_rate": 2.7695968189919686e-06,
+      "loss": 0.2089,
+      "step": 1195
+    },
+    {
+      "epoch": 0.77,
+      "learning_rate": 2.755296849505241e-06,
+      "loss": 0.245,
+      "step": 1196
+    },
+    {
+      "epoch": 0.77,
+      "learning_rate": 2.7410279915781124e-06,
+      "loss": 0.1925,
+      "step": 1197
+    },
+    {
+      "epoch": 0.77,
+      "learning_rate": 2.7267903064865765e-06,
+      "loss": 0.2089,
+      "step": 1198
+    },
+    {
+      "epoch": 0.77,
+      "learning_rate": 2.712583855372769e-06,
+      "loss": 0.2154,
+      "step": 1199
+    },
+    {
+      "epoch": 0.77,
+      "learning_rate": 2.6984086992446834e-06,
+      "loss": 0.1862,
+      "step": 1200
+    },
+    {
+      "epoch": 0.77,
+      "learning_rate": 2.6842648989759325e-06,
+      "loss": 0.1882,
+      "step": 1201
+    },
+    {
+      "epoch": 0.77,
+      "learning_rate": 2.67015251530547e-06,
+      "loss": 0.1719,
+      "step": 1202
+    },
+    {
+      "epoch": 0.77,
+      "learning_rate": 2.6560716088373295e-06,
+      "loss": 0.1874,
+      "step": 1203
+    },
+    {
+      "epoch": 0.77,
+      "learning_rate": 2.642022240040378e-06,
+      "loss": 0.207,
+      "step": 1204
+    },
+    {
+      "epoch": 0.77,
+      "learning_rate": 2.628004469248043e-06,
+      "loss": 0.1905,
+      "step": 1205
+    },
+    {
+      "epoch": 0.77,
+      "learning_rate": 2.61401835665805e-06,
+      "loss": 0.1908,
+      "step": 1206
+    },
+    {
+      "epoch": 0.77,
+      "learning_rate": 2.6000639623321934e-06,
+      "loss": 0.1427,
+      "step": 1207
+    },
+    {
+      "epoch": 0.77,
+      "learning_rate": 2.5861413461960373e-06,
+      "loss": 0.1609,
+      "step": 1208
+    },
+    {
+      "epoch": 0.77,
+      "learning_rate": 2.572250568038681e-06,
+      "loss": 0.209,
+      "step": 1209
+    },
+    {
+      "epoch": 0.77,
+      "learning_rate": 2.5583916875125104e-06,
+      "loss": 0.1916,
+      "step": 1210
+    },
+    {
+      "epoch": 0.77,
+      "learning_rate": 2.5445647641329154e-06,
+      "loss": 0.2104,
+      "step": 1211
+    },
+    {
+      "epoch": 0.78,
+      "learning_rate": 2.5307698572780604e-06,
+      "loss": 0.1919,
+      "step": 1212
+    },
+    {
+      "epoch": 0.78,
+      "learning_rate": 2.517007026188619e-06,
+      "loss": 0.2098,
+      "step": 1213
+    },
+    {
+      "epoch": 0.78,
+      "learning_rate": 2.503276329967509e-06,
+      "loss": 0.2093,
+      "step": 1214
+    },
+    {
+      "epoch": 0.78,
+      "learning_rate": 2.489577827579659e-06,
+      "loss": 0.1987,
+      "step": 1215
+    },
+    {
+      "epoch": 0.78,
+      "learning_rate": 2.475911577851737e-06,
+      "loss": 0.1957,
+      "step": 1216
+    },
+    {
+      "epoch": 0.78,
+      "learning_rate": 2.462277639471914e-06,
+      "loss": 0.1982,
+      "step": 1217
+    },
+    {
+      "epoch": 0.78,
+      "learning_rate": 2.448676070989593e-06,
+      "loss": 0.1927,
+      "step": 1218
+    },
+    {
+      "epoch": 0.78,
+      "learning_rate": 2.4351069308151774e-06,
+      "loss": 0.1698,
+      "step": 1219
+    },
+    {
+      "epoch": 0.78,
+      "learning_rate": 2.421570277219809e-06,
+      "loss": 0.168,
+      "step": 1220
+    },
+    {
+      "epoch": 0.78,
+      "learning_rate": 2.4080661683351146e-06,
+      "loss": 0.1901,
+      "step": 1221
+    },
+    {
+      "epoch": 0.78,
+      "learning_rate": 2.3945946621529693e-06,
+      "loss": 0.1913,
+      "step": 1222
+    },
+    {
+      "epoch": 0.78,
+      "learning_rate": 2.381155816525228e-06,
+      "loss": 0.1815,
+      "step": 1223
+    },
+    {
+      "epoch": 0.78,
+      "learning_rate": 2.367749689163501e-06,
+      "loss": 0.1855,
+      "step": 1224
+    },
+    {
+      "epoch": 0.78,
+      "learning_rate": 2.3543763376388906e-06,
+      "loss": 0.1683,
+      "step": 1225
+    },
+    {
+      "epoch": 0.78,
+      "learning_rate": 2.3410358193817384e-06,
+      "loss": 0.227,
+      "step": 1226
+    },
+    {
+      "epoch": 0.79,
+      "learning_rate": 2.3277281916814e-06,
+      "loss": 0.1572,
+      "step": 1227
+    },
+    {
+      "epoch": 0.79,
+      "learning_rate": 2.3144535116859756e-06,
+      "loss": 0.1993,
+      "step": 1228
+    },
+    {
+      "epoch": 0.79,
+      "learning_rate": 2.3012118364020787e-06,
+      "loss": 0.188,
+      "step": 1229
+    },
+    {
+      "epoch": 0.79,
+      "learning_rate": 2.2880032226945903e-06,
+      "loss": 0.2077,
+      "step": 1230
+    },
+    {
+      "epoch": 0.79,
+      "learning_rate": 2.2748277272864108e-06,
+      "loss": 0.2132,
+      "step": 1231
+    },
+    {
+      "epoch": 0.79,
+      "learning_rate": 2.2616854067582195e-06,
+      "loss": 0.1932,
+      "step": 1232
+    },
+    {
+      "epoch": 0.79,
+      "learning_rate": 2.2485763175482255e-06,
+      "loss": 0.1655,
+      "step": 1233
+    },
+    {
+      "epoch": 0.79,
+      "learning_rate": 2.2355005159519294e-06,
+      "loss": 0.1956,
+      "step": 1234
+    },
+    {
+      "epoch": 0.79,
+      "learning_rate": 2.222458058121889e-06,
+      "loss": 0.1633,
+      "step": 1235
+    },
+    {
+      "epoch": 0.79,
+      "learning_rate": 2.2094490000674605e-06,
+      "loss": 0.2053,
+      "step": 1236
+    },
+    {
+      "epoch": 0.79,
+      "learning_rate": 2.196473397654585e-06,
+      "loss": 0.1924,
+      "step": 1237
+    },
+    {
+      "epoch": 0.79,
+      "learning_rate": 2.1835313066055153e-06,
+      "loss": 0.189,
+      "step": 1238
+    },
+    {
+      "epoch": 0.79,
+      "learning_rate": 2.170622782498598e-06,
+      "loss": 0.2059,
+      "step": 1239
+    },
+    {
+      "epoch": 0.79,
+      "learning_rate": 2.1577478807680384e-06,
+      "loss": 0.1872,
+      "step": 1240
+    },
+    {
+      "epoch": 0.79,
+      "learning_rate": 2.1449066567036413e-06,
+      "loss": 0.2283,
+      "step": 1241
+    },
+    {
+      "epoch": 0.79,
+      "learning_rate": 2.1320991654505975e-06,
+      "loss": 0.2033,
+      "step": 1242
+    },
+    {
+      "epoch": 0.8,
+      "learning_rate": 2.119325462009233e-06,
+      "loss": 0.2114,
+      "step": 1243
+    },
+    {
+      "epoch": 0.8,
+      "learning_rate": 2.106585601234771e-06,
+      "loss": 0.183,
+      "step": 1244
+    },
+    {
+      "epoch": 0.8,
+      "learning_rate": 2.0938796378371084e-06,
+      "loss": 0.2103,
+      "step": 1245
+    },
+    {
+      "epoch": 0.8,
+      "learning_rate": 2.081207626380567e-06,
+      "loss": 0.2018,
+      "step": 1246
+    },
+    {
+      "epoch": 0.8,
+      "learning_rate": 2.0685696212836737e-06,
+      "loss": 0.2021,
+      "step": 1247
+    },
+    {
+      "epoch": 0.8,
+      "learning_rate": 2.0559656768189084e-06,
+      "loss": 0.2091,
+      "step": 1248
+    },
+    {
+      "epoch": 0.8,
+      "learning_rate": 2.0433958471124903e-06,
+      "loss": 0.1561,
+      "step": 1249
+    },
+    {
+      "epoch": 0.8,
+      "learning_rate": 2.0308601861441367e-06,
+      "loss": 0.1937,
+      "step": 1250
+    },
+    {
+      "epoch": 0.8,
+      "learning_rate": 2.0183587477468226e-06,
+      "loss": 0.1966,
+      "step": 1251
+    },
+    {
+      "epoch": 0.8,
+      "learning_rate": 2.0058915856065688e-06,
+      "loss": 0.2052,
+      "step": 1252
+    },
+    {
+      "epoch": 0.8,
+      "learning_rate": 1.9934587532621917e-06,
+      "loss": 0.1747,
+      "step": 1253
+    },
+    {
+      "epoch": 0.8,
+      "learning_rate": 1.9810603041050825e-06,
+      "loss": 0.1888,
+      "step": 1254
+    },
+    {
+      "epoch": 0.8,
+      "learning_rate": 1.9686962913789897e-06,
+      "loss": 0.1854,
+      "step": 1255
+    },
+    {
+      "epoch": 0.8,
+      "learning_rate": 1.956366768179766e-06,
+      "loss": 0.1541,
+      "step": 1256
+    },
+    {
+      "epoch": 0.8,
+      "learning_rate": 1.9440717874551527e-06,
+      "loss": 0.1869,
+      "step": 1257
+    },
+    {
+      "epoch": 0.8,
+      "learning_rate": 1.931811402004562e-06,
+      "loss": 0.193,
+      "step": 1258
+    },
+    {
+      "epoch": 0.81,
+      "learning_rate": 1.91958566447883e-06,
+      "loss": 0.2031,
+      "step": 1259
+    },
+    {
+      "epoch": 0.81,
+      "learning_rate": 1.907394627380007e-06,
+      "loss": 0.1843,
+      "step": 1260
+    },
+    {
+      "epoch": 0.81,
+      "learning_rate": 1.8952383430611299e-06,
+      "loss": 0.1908,
+      "step": 1261
+    },
+    {
+      "epoch": 0.81,
+      "learning_rate": 1.8831168637259823e-06,
+      "loss": 0.1628,
+      "step": 1262
+    },
+    {
+      "epoch": 0.81,
+      "learning_rate": 1.871030241428894e-06,
+      "loss": 0.2136,
+      "step": 1263
+    },
+    {
+      "epoch": 0.81,
+      "learning_rate": 1.8589785280744944e-06,
+      "loss": 0.2158,
+      "step": 1264
+    },
+    {
+      "epoch": 0.81,
+      "learning_rate": 1.8469617754175106e-06,
+      "loss": 0.1779,
+      "step": 1265
+    },
+    {
+      "epoch": 0.81,
+      "learning_rate": 1.8349800350625258e-06,
+      "loss": 0.1749,
+      "step": 1266
+    },
+    {
+      "epoch": 0.81,
+      "learning_rate": 1.8230333584637715e-06,
+      "loss": 0.1983,
+      "step": 1267
+    },
+    {
+      "epoch": 0.81,
+      "learning_rate": 1.8111217969249062e-06,
+      "loss": 0.2122,
+      "step": 1268
+    },
+    {
+      "epoch": 0.81,
+      "learning_rate": 1.7992454015987791e-06,
+      "loss": 0.2273,
+      "step": 1269
+    },
+    {
+      "epoch": 0.81,
+      "learning_rate": 1.7874042234872358e-06,
+      "loss": 0.2053,
+      "step": 1270
+    },
+    {
+      "epoch": 0.81,
+      "learning_rate": 1.7755983134408706e-06,
+      "loss": 0.1875,
+      "step": 1271
+    },
+    {
+      "epoch": 0.81,
+      "learning_rate": 1.7638277221588351e-06,
+      "loss": 0.1855,
+      "step": 1272
+    },
+    {
+      "epoch": 0.81,
+      "learning_rate": 1.752092500188608e-06,
+      "loss": 0.1845,
+      "step": 1273
+    },
+    {
+      "epoch": 0.82,
+      "learning_rate": 1.740392697925768e-06,
+      "loss": 0.1647,
+      "step": 1274
+    },
+    {
+      "epoch": 0.82,
+      "learning_rate": 1.7287283656138009e-06,
+      "loss": 0.1971,
+      "step": 1275
+    },
+    {
+      "epoch": 0.82,
+      "learning_rate": 1.7170995533438628e-06,
+      "loss": 0.1945,
+      "step": 1276
+    },
+    {
+      "epoch": 0.82,
+      "learning_rate": 1.70550631105457e-06,
+      "loss": 0.2329,
+      "step": 1277
+    },
+    {
+      "epoch": 0.82,
+      "learning_rate": 1.693948688531797e-06,
+      "loss": 0.1999,
+      "step": 1278
+    },
+    {
+      "epoch": 0.82,
+      "learning_rate": 1.6824267354084478e-06,
+      "loss": 0.181,
+      "step": 1279
+    },
+    {
+      "epoch": 0.82,
+      "learning_rate": 1.6709405011642522e-06,
+      "loss": 0.1812,
+      "step": 1280
+    },
+    {
+      "epoch": 0.82,
+      "learning_rate": 1.6594900351255427e-06,
+      "loss": 0.2112,
+      "step": 1281
+    },
+    {
+      "epoch": 0.82,
+      "learning_rate": 1.6480753864650501e-06,
+      "loss": 0.1565,
+      "step": 1282
+    },
+    {
+      "epoch": 0.82,
+      "learning_rate": 1.6366966042016996e-06,
+      "loss": 0.2196,
+      "step": 1283
+    },
+    {
+      "epoch": 0.82,
+      "learning_rate": 1.6253537372003825e-06,
+      "loss": 0.1916,
+      "step": 1284
+    },
+    {
+      "epoch": 0.82,
+      "learning_rate": 1.6140468341717607e-06,
+      "loss": 0.1789,
+      "step": 1285
+    },
+    {
+      "epoch": 0.82,
+      "learning_rate": 1.602775943672057e-06,
+      "loss": 0.1889,
+      "step": 1286
+    },
+    {
+      "epoch": 0.82,
+      "learning_rate": 1.5915411141028325e-06,
+      "loss": 0.2073,
+      "step": 1287
+    },
+    {
+      "epoch": 0.82,
+      "learning_rate": 1.5803423937107975e-06,
+      "loss": 0.1959,
+      "step": 1288
+    },
+    {
+      "epoch": 0.82,
+      "learning_rate": 1.5691798305875893e-06,
+      "loss": 0.187,
+      "step": 1289
+    },
+    {
+      "epoch": 0.83,
+      "learning_rate": 1.558053472669576e-06,
+      "loss": 0.1921,
+      "step": 1290
+    },
+    {
+      "epoch": 0.83,
+      "learning_rate": 1.5469633677376494e-06,
+      "loss": 0.197,
+      "step": 1291
+    },
+    {
+      "epoch": 0.83,
+      "learning_rate": 1.5359095634170052e-06,
+      "loss": 0.2046,
+      "step": 1292
+    },
+    {
+      "epoch": 0.83,
+      "learning_rate": 1.524892107176964e-06,
+      "loss": 0.1895,
+      "step": 1293
+    },
+    {
+      "epoch": 0.83,
+      "learning_rate": 1.5139110463307438e-06,
+      "loss": 0.1844,
+      "step": 1294
+    },
+    {
+      "epoch": 0.83,
+      "learning_rate": 1.502966428035274e-06,
+      "loss": 0.2046,
+      "step": 1295
+    },
+    {
+      "epoch": 0.83,
+      "learning_rate": 1.492058299290977e-06,
+      "loss": 0.1743,
+      "step": 1296
+    },
+    {
+      "epoch": 0.83,
+      "learning_rate": 1.4811867069415832e-06,
+      "loss": 0.2051,
+      "step": 1297
+    },
+    {
+      "epoch": 0.83,
+      "learning_rate": 1.4703516976739197e-06,
+      "loss": 0.2313,
+      "step": 1298
+    },
+    {
+      "epoch": 0.83,
+      "learning_rate": 1.4595533180177058e-06,
+      "loss": 0.1696,
+      "step": 1299
+    },
+    {
+      "epoch": 0.83,
+      "learning_rate": 1.4487916143453684e-06,
+      "loss": 0.1607,
+      "step": 1300
+    },
+    {
+      "epoch": 0.83,
+      "learning_rate": 1.4380666328718274e-06,
+      "loss": 0.2208,
+      "step": 1301
+    },
+    {
+      "epoch": 0.83,
+      "learning_rate": 1.4273784196542983e-06,
+      "loss": 0.1967,
+      "step": 1302
+    },
+    {
+      "epoch": 0.83,
+      "learning_rate": 1.4167270205921168e-06,
+      "loss": 0.1729,
+      "step": 1303
+    },
+    {
+      "epoch": 0.83,
+      "learning_rate": 1.406112481426507e-06,
+      "loss": 0.209,
+      "step": 1304
+    },
+    {
+      "epoch": 0.83,
+      "learning_rate": 1.3955348477404073e-06,
+      "loss": 0.1956,
+      "step": 1305
+    },
+    {
+      "epoch": 0.84,
+      "learning_rate": 1.3849941649582766e-06,
+      "loss": 0.2028,
+      "step": 1306
+    },
+    {
+      "epoch": 0.84,
+      "learning_rate": 1.37449047834588e-06,
+      "loss": 0.1992,
+      "step": 1307
+    },
+    {
+      "epoch": 0.84,
+      "learning_rate": 1.3640238330101186e-06,
+      "loss": 0.183,
+      "step": 1308
+    },
+    {
+      "epoch": 0.84,
+      "learning_rate": 1.3535942738988194e-06,
+      "loss": 0.186,
+      "step": 1309
+    },
+    {
+      "epoch": 0.84,
+      "learning_rate": 1.3432018458005414e-06,
+      "loss": 0.2091,
+      "step": 1310
+    },
+    {
+      "epoch": 0.84,
+      "learning_rate": 1.332846593344399e-06,
+      "loss": 0.1895,
+      "step": 1311
+    },
+    {
+      "epoch": 0.84,
+      "learning_rate": 1.3225285609998507e-06,
+      "loss": 0.1659,
+      "step": 1312
+    },
+    {
+      "epoch": 0.84,
+      "learning_rate": 1.3122477930765243e-06,
+      "loss": 0.1934,
+      "step": 1313
+    },
+    {
+      "epoch": 0.84,
+      "learning_rate": 1.3020043337240117e-06,
+      "loss": 0.2145,
+      "step": 1314
+    },
+    {
+      "epoch": 0.84,
+      "learning_rate": 1.2917982269316975e-06,
+      "loss": 0.1796,
+      "step": 1315
+    },
+    {
+      "epoch": 0.84,
+      "learning_rate": 1.2816295165285541e-06,
+      "loss": 0.1888,
+      "step": 1316
+    },
+    {
+      "epoch": 0.84,
+      "learning_rate": 1.2714982461829572e-06,
+      "loss": 0.181,
+      "step": 1317
+    },
+    {
+      "epoch": 0.84,
+      "learning_rate": 1.2614044594025076e-06,
+      "loss": 0.186,
+      "step": 1318
+    },
+    {
+      "epoch": 0.84,
+      "learning_rate": 1.2513481995338284e-06,
+      "loss": 0.1773,
+      "step": 1319
+    },
+    {
+      "epoch": 0.84,
+      "learning_rate": 1.2413295097623922e-06,
+      "loss": 0.2173,
+      "step": 1320
+    },
+    {
+      "epoch": 0.85,
+      "learning_rate": 1.2313484331123371e-06,
+      "loss": 0.1751,
+      "step": 1321
+    },
+    {
+      "epoch": 0.85,
+      "learning_rate": 1.2214050124462617e-06,
+      "loss": 0.1684,
+      "step": 1322
+    },
+    {
+      "epoch": 0.85,
+      "learning_rate": 1.211499290465069e-06,
+      "loss": 0.1898,
+      "step": 1323
+    },
+    {
+      "epoch": 0.85,
+      "learning_rate": 1.2016313097077615e-06,
+      "loss": 0.1827,
+      "step": 1324
+    },
+    {
+      "epoch": 0.85,
+      "learning_rate": 1.1918011125512651e-06,
+      "loss": 0.1951,
+      "step": 1325
+    },
+    {
+      "epoch": 0.85,
+      "learning_rate": 1.1820087412102555e-06,
+      "loss": 0.1974,
+      "step": 1326
+    },
+    {
+      "epoch": 0.85,
+      "learning_rate": 1.1722542377369639e-06,
+      "loss": 0.1478,
+      "step": 1327
+    },
+    {
+      "epoch": 0.85,
+      "learning_rate": 1.1625376440210067e-06,
+      "loss": 0.1831,
+      "step": 1328
+    },
+    {
+      "epoch": 0.85,
+      "learning_rate": 1.152859001789196e-06,
+      "loss": 0.1652,
+      "step": 1329
+    },
+    {
+      "epoch": 0.85,
+      "learning_rate": 1.1432183526053664e-06,
+      "loss": 0.2023,
+      "step": 1330
+    },
+    {
+      "epoch": 0.85,
+      "learning_rate": 1.1336157378702017e-06,
+      "loss": 0.179,
+      "step": 1331
+    },
+    {
+      "epoch": 0.85,
+      "learning_rate": 1.12405119882104e-06,
+      "loss": 0.1858,
+      "step": 1332
+    },
+    {
+      "epoch": 0.85,
+      "learning_rate": 1.1145247765317192e-06,
+      "loss": 0.1951,
+      "step": 1333
+    },
+    {
+      "epoch": 0.85,
+      "learning_rate": 1.1050365119123851e-06,
+      "loss": 0.178,
+      "step": 1334
+    },
+    {
+      "epoch": 0.85,
+      "learning_rate": 1.0955864457093147e-06,
+      "loss": 0.1834,
+      "step": 1335
+    },
+    {
+      "epoch": 0.85,
+      "learning_rate": 1.0861746185047528e-06,
+      "loss": 0.1779,
+      "step": 1336
+    },
+    {
+      "epoch": 0.86,
+      "learning_rate": 1.0768010707167265e-06,
+      "loss": 0.182,
+      "step": 1337
+    },
+    {
+      "epoch": 0.86,
+      "learning_rate": 1.0674658425988793e-06,
+      "loss": 0.2102,
+      "step": 1338
+    },
+    {
+      "epoch": 0.86,
+      "learning_rate": 1.0581689742402967e-06,
+      "loss": 0.2034,
+      "step": 1339
+    },
+    {
+      "epoch": 0.86,
+      "learning_rate": 1.0489105055653247e-06,
+      "loss": 0.1788,
+      "step": 1340
+    },
+    {
+      "epoch": 0.86,
+      "learning_rate": 1.0396904763334182e-06,
+      "loss": 0.1959,
+      "step": 1341
+    },
+    {
+      "epoch": 0.86,
+      "learning_rate": 1.0305089261389444e-06,
+      "loss": 0.2121,
+      "step": 1342
+    },
+    {
+      "epoch": 0.86,
+      "learning_rate": 1.0213658944110406e-06,
+      "loss": 0.2051,
+      "step": 1343
+    },
+    {
+      "epoch": 0.86,
+      "learning_rate": 1.0122614204134184e-06,
+      "loss": 0.1772,
+      "step": 1344
+    },
+    {
+      "epoch": 0.86,
+      "learning_rate": 1.0031955432442153e-06,
+      "loss": 0.2121,
+      "step": 1345
+    },
+    {
+      "epoch": 0.86,
+      "learning_rate": 9.941683018358206e-07,
+      "loss": 0.2068,
+      "step": 1346
+    },
+    {
+      "epoch": 0.86,
+      "learning_rate": 9.851797349546977e-07,
+      "loss": 0.2192,
+      "step": 1347
+    },
+    {
+      "epoch": 0.86,
+      "learning_rate": 9.762298812012361e-07,
+      "loss": 0.1684,
+      "step": 1348
+    },
+    {
+      "epoch": 0.86,
+      "learning_rate": 9.673187790095707e-07,
+      "loss": 0.2194,
+      "step": 1349
+    },
+    {
+      "epoch": 0.86,
+      "learning_rate": 9.58446466647419e-07,
+      "loss": 0.196,
+      "step": 1350
+    },
+    {
+      "epoch": 0.86,
+      "learning_rate": 9.496129822159339e-07,
+      "loss": 0.1837,
+      "step": 1351
+    },
+    {
+      "epoch": 0.87,
+      "learning_rate": 9.408183636495116e-07,
+      "loss": 0.2189,
+      "step": 1352
+    },
+    {
+      "epoch": 0.87,
+      "learning_rate": 9.320626487156459e-07,
+      "loss": 0.2212,
+      "step": 1353
+    },
+    {
+      "epoch": 0.87,
+      "learning_rate": 9.233458750147717e-07,
+      "loss": 0.214,
+      "step": 1354
+    },
+    {
+      "epoch": 0.87,
+      "learning_rate": 9.146680799800833e-07,
+      "loss": 0.1779,
+      "step": 1355
+    },
+    {
+      "epoch": 0.87,
+      "learning_rate": 9.060293008773957e-07,
+      "loss": 0.1804,
+      "step": 1356
+    },
+    {
+      "epoch": 0.87,
+      "learning_rate": 8.974295748049711e-07,
+      "loss": 0.1932,
+      "step": 1357
+    },
+    {
+      "epoch": 0.87,
+      "learning_rate": 8.888689386933569e-07,
+      "loss": 0.1806,
+      "step": 1358
+    },
+    {
+      "epoch": 0.87,
+      "learning_rate": 8.803474293052438e-07,
+      "loss": 0.1577,
+      "step": 1359
+    },
+    {
+      "epoch": 0.87,
+      "learning_rate": 8.718650832352882e-07,
+      "loss": 0.2128,
+      "step": 1360
+    },
+    {
+      "epoch": 0.87,
+      "learning_rate": 8.634219369099694e-07,
+      "loss": 0.1948,
+      "step": 1361
+    },
+    {
+      "epoch": 0.87,
+      "learning_rate": 8.550180265874231e-07,
+      "loss": 0.194,
+      "step": 1362
+    },
+    {
+      "epoch": 0.87,
+      "learning_rate": 8.466533883572947e-07,
+      "loss": 0.1949,
+      "step": 1363
+    },
+    {
+      "epoch": 0.87,
+      "learning_rate": 8.383280581405806e-07,
+      "loss": 0.2057,
+      "step": 1364
+    },
+    {
+      "epoch": 0.87,
+      "learning_rate": 8.300420716894685e-07,
+      "loss": 0.1642,
+      "step": 1365
+    },
+    {
+      "epoch": 0.87,
+      "learning_rate": 8.217954645871939e-07,
+      "loss": 0.1865,
+      "step": 1366
+    },
+    {
+      "epoch": 0.87,
+      "learning_rate": 8.135882722478772e-07,
+      "loss": 0.1912,
+      "step": 1367
+    },
+    {
+      "epoch": 0.88,
+      "learning_rate": 8.054205299163797e-07,
+      "loss": 0.1879,
+      "step": 1368
+    },
+    {
+      "epoch": 0.88,
+      "learning_rate": 7.972922726681509e-07,
+      "loss": 0.1855,
+      "step": 1369
+    },
+    {
+      "epoch": 0.88,
+      "learning_rate": 7.89203535409071e-07,
+      "loss": 0.2127,
+      "step": 1370
+    },
+    {
+      "epoch": 0.88,
+      "learning_rate": 7.811543528753108e-07,
+      "loss": 0.1801,
+      "step": 1371
+    },
+    {
+      "epoch": 0.88,
+      "learning_rate": 7.731447596331753e-07,
+      "loss": 0.1926,
+      "step": 1372
+    },
+    {
+      "epoch": 0.88,
+      "learning_rate": 7.651747900789552e-07,
+      "loss": 0.1797,
+      "step": 1373
+    },
+    {
+      "epoch": 0.88,
+      "learning_rate": 7.572444784387878e-07,
+      "loss": 0.181,
+      "step": 1374
+    },
+    {
+      "epoch": 0.88,
+      "learning_rate": 7.493538587685012e-07,
+      "loss": 0.1929,
+      "step": 1375
+    },
+    {
+      "epoch": 0.88,
+      "learning_rate": 7.415029649534699e-07,
+      "loss": 0.2045,
+      "step": 1376
+    },
+    {
+      "epoch": 0.88,
+      "learning_rate": 7.336918307084718e-07,
+      "loss": 0.175,
+      "step": 1377
+    },
+    {
+      "epoch": 0.88,
+      "learning_rate": 7.259204895775373e-07,
+      "loss": 0.1927,
+      "step": 1378
+    },
+    {
+      "epoch": 0.88,
+      "learning_rate": 7.181889749338178e-07,
+      "loss": 0.191,
+      "step": 1379
+    },
+    {
+      "epoch": 0.88,
+      "learning_rate": 7.104973199794252e-07,
+      "loss": 0.19,
+      "step": 1380
+    },
+    {
+      "epoch": 0.88,
+      "learning_rate": 7.028455577453075e-07,
+      "loss": 0.2013,
+      "step": 1381
+    },
+    {
+      "epoch": 0.88,
+      "learning_rate": 6.95233721091092e-07,
+      "loss": 0.2114,
+      "step": 1382
+    },
+    {
+      "epoch": 0.88,
+      "learning_rate": 6.876618427049509e-07,
+      "loss": 0.1942,
+      "step": 1383
+    },
+    {
+      "epoch": 0.89,
+      "learning_rate": 6.801299551034623e-07,
+      "loss": 0.175,
+      "step": 1384
+    },
+    {
+      "epoch": 0.89,
+      "learning_rate": 6.726380906314655e-07,
+      "loss": 0.1867,
+      "step": 1385
+    },
+    {
+      "epoch": 0.89,
+      "learning_rate": 6.65186281461927e-07,
+      "loss": 0.1425,
+      "step": 1386
+    },
+    {
+      "epoch": 0.89,
+      "learning_rate": 6.577745595958018e-07,
+      "loss": 0.186,
+      "step": 1387
+    },
+    {
+      "epoch": 0.89,
+      "learning_rate": 6.504029568618886e-07,
+      "loss": 0.1644,
+      "step": 1388
+    },
+    {
+      "epoch": 0.89,
+      "learning_rate": 6.430715049167069e-07,
+      "loss": 0.1692,
+      "step": 1389
+    },
+    {
+      "epoch": 0.89,
+      "learning_rate": 6.357802352443454e-07,
+      "loss": 0.1881,
+      "step": 1390
+    },
+    {
+      "epoch": 0.89,
+      "learning_rate": 6.285291791563431e-07,
+      "loss": 0.1851,
+      "step": 1391
+    },
+    {
+      "epoch": 0.89,
+      "learning_rate": 6.213183677915358e-07,
+      "loss": 0.179,
+      "step": 1392
+    },
+    {
+      "epoch": 0.89,
+      "learning_rate": 6.141478321159422e-07,
+      "loss": 0.1876,
+      "step": 1393
+    },
+    {
+      "epoch": 0.89,
+      "learning_rate": 6.070176029226194e-07,
+      "loss": 0.1763,
+      "step": 1394
+    },
+    {
+      "epoch": 0.89,
+      "learning_rate": 5.999277108315272e-07,
+      "loss": 0.214,
+      "step": 1395
+    },
+    {
+      "epoch": 0.89,
+      "learning_rate": 5.928781862894107e-07,
+      "loss": 0.1968,
+      "step": 1396
+    },
+    {
+      "epoch": 0.89,
+      "learning_rate": 5.858690595696559e-07,
+      "loss": 0.1808,
+      "step": 1397
+    },
+    {
+      "epoch": 0.89,
+      "learning_rate": 5.789003607721621e-07,
+      "loss": 0.214,
+      "step": 1398
+    },
+    {
+      "epoch": 0.9,
+      "learning_rate": 5.719721198232254e-07,
+      "loss": 0.2074,
+      "step": 1399
+    },
+    {
+      "epoch": 0.9,
+      "learning_rate": 5.650843664753924e-07,
+      "loss": 0.2042,
+      "step": 1400
+    },
+    {
+      "epoch": 0.9,
+      "learning_rate": 5.582371303073386e-07,
+      "loss": 0.2104,
+      "step": 1401
+    },
+    {
+      "epoch": 0.9,
+      "learning_rate": 5.514304407237503e-07,
+      "loss": 0.1717,
+      "step": 1402
+    },
+    {
+      "epoch": 0.9,
+      "learning_rate": 5.446643269551854e-07,
+      "loss": 0.1703,
+      "step": 1403
+    },
+    {
+      "epoch": 0.9,
+      "learning_rate": 5.379388180579559e-07,
+      "loss": 0.1898,
+      "step": 1404
+    },
+    {
+      "epoch": 0.9,
+      "learning_rate": 5.312539429140018e-07,
+      "loss": 0.197,
+      "step": 1405
+    },
+    {
+      "epoch": 0.9,
+      "learning_rate": 5.246097302307629e-07,
+      "loss": 0.2283,
+      "step": 1406
+    },
+    {
+      "epoch": 0.9,
+      "learning_rate": 5.180062085410609e-07,
+      "loss": 0.2104,
+      "step": 1407
+    },
+    {
+      "epoch": 0.9,
+      "learning_rate": 5.114434062029727e-07,
+      "loss": 0.1832,
+      "step": 1408
+    },
+    {
+      "epoch": 0.9,
+      "learning_rate": 5.049213513997142e-07,
+      "loss": 0.1925,
+      "step": 1409
+    },
+    {
+      "epoch": 0.9,
+      "learning_rate": 4.9844007213951e-07,
+      "loss": 0.1998,
+      "step": 1410
+    },
+    {
+      "epoch": 0.9,
+      "learning_rate": 4.919995962554846e-07,
+      "loss": 0.201,
+      "step": 1411
+    },
+    {
+      "epoch": 0.9,
+      "learning_rate": 4.855999514055354e-07,
+      "loss": 0.1961,
+      "step": 1412
+    },
+    {
+      "epoch": 0.9,
+      "learning_rate": 4.792411650722117e-07,
+      "loss": 0.1749,
+      "step": 1413
+    },
+    {
+      "epoch": 0.9,
+      "learning_rate": 4.7292326456260586e-07,
+      "loss": 0.1914,
+      "step": 1414
+    },
+    {
+      "epoch": 0.91,
+      "learning_rate": 4.6664627700822475e-07,
+      "loss": 0.1871,
+      "step": 1415
+    },
+    {
+      "epoch": 0.91,
+      "learning_rate": 4.60410229364886e-07,
+      "loss": 0.1818,
+      "step": 1416
+    },
+    {
+      "epoch": 0.91,
+      "learning_rate": 4.5421514841259115e-07,
+      "loss": 0.1984,
+      "step": 1417
+    },
+    {
+      "epoch": 0.91,
+      "learning_rate": 4.480610607554148e-07,
+      "loss": 0.1706,
+      "step": 1418
+    },
+    {
+      "epoch": 0.91,
+      "learning_rate": 4.4194799282139324e-07,
+      "loss": 0.2156,
+      "step": 1419
+    },
+    {
+      "epoch": 0.91,
+      "learning_rate": 4.3587597086240607e-07,
+      "loss": 0.2046,
+      "step": 1420
+    },
+    {
+      "epoch": 0.91,
+      "learning_rate": 4.298450209540628e-07,
+      "loss": 0.1691,
+      "step": 1421
+    },
+    {
+      "epoch": 0.91,
+      "learning_rate": 4.2385516899559944e-07,
+      "loss": 0.2141,
+      "step": 1422
+    },
+    {
+      "epoch": 0.91,
+      "learning_rate": 4.1790644070975884e-07,
+      "loss": 0.1955,
+      "step": 1423
+    },
+    {
+      "epoch": 0.91,
+      "learning_rate": 4.11998861642684e-07,
+      "loss": 0.1789,
+      "step": 1424
+    },
+    {
+      "epoch": 0.91,
+      "learning_rate": 4.061324571638048e-07,
+      "loss": 0.2374,
+      "step": 1425
+    },
+    {
+      "epoch": 0.91,
+      "learning_rate": 4.003072524657303e-07,
+      "loss": 0.2032,
+      "step": 1426
+    },
+    {
+      "epoch": 0.91,
+      "learning_rate": 3.94523272564149e-07,
+      "loss": 0.1962,
+      "step": 1427
+    },
+    {
+      "epoch": 0.91,
+      "learning_rate": 3.8878054229770643e-07,
+      "loss": 0.179,
+      "step": 1428
+    },
+    {
+      "epoch": 0.91,
+      "learning_rate": 3.830790863279088e-07,
+      "loss": 0.2065,
+      "step": 1429
+    },
+    {
+      "epoch": 0.91,
+      "learning_rate": 3.7741892913901847e-07,
+      "loss": 0.1774,
+      "step": 1430
+    },
+    {
+      "epoch": 0.92,
+      "learning_rate": 3.7180009503793745e-07,
+      "loss": 0.1875,
+      "step": 1431
+    },
+    {
+      "epoch": 0.92,
+      "learning_rate": 3.662226081541176e-07,
+      "loss": 0.1794,
+      "step": 1432
+    },
+    {
+      "epoch": 0.92,
+      "learning_rate": 3.6068649243944267e-07,
+      "loss": 0.1951,
+      "step": 1433
+    },
+    {
+      "epoch": 0.92,
+      "learning_rate": 3.5519177166813854e-07,
+      "loss": 0.2014,
+      "step": 1434
+    },
+    {
+      "epoch": 0.92,
+      "learning_rate": 3.4973846943666567e-07,
+      "loss": 0.1884,
+      "step": 1435
+    },
+    {
+      "epoch": 0.92,
+      "learning_rate": 3.443266091636099e-07,
+      "loss": 0.1604,
+      "step": 1436
+    },
+    {
+      "epoch": 0.92,
+      "learning_rate": 3.3895621408959744e-07,
+      "loss": 0.1852,
+      "step": 1437
+    },
+    {
+      "epoch": 0.92,
+      "learning_rate": 3.336273072771812e-07,
+      "loss": 0.1793,
+      "step": 1438
+    },
+    {
+      "epoch": 0.92,
+      "learning_rate": 3.283399116107533e-07,
+      "loss": 0.218,
+      "step": 1439
+    },
+    {
+      "epoch": 0.92,
+      "learning_rate": 3.23094049796433e-07,
+      "loss": 0.1709,
+      "step": 1440
+    },
+    {
+      "epoch": 0.92,
+      "learning_rate": 3.1788974436198326e-07,
+      "loss": 0.1971,
+      "step": 1441
+    },
+    {
+      "epoch": 0.92,
+      "learning_rate": 3.1272701765670766e-07,
+      "loss": 0.2222,
+      "step": 1442
+    },
+    {
+      "epoch": 0.92,
+      "learning_rate": 3.076058918513503e-07,
+      "loss": 0.167,
+      "step": 1443
+    },
+    {
+      "epoch": 0.92,
+      "learning_rate": 3.0252638893801034e-07,
+      "loss": 0.1761,
+      "step": 1444
+    },
+    {
+      "epoch": 0.92,
+      "learning_rate": 2.974885307300379e-07,
+      "loss": 0.1726,
+      "step": 1445
+    },
+    {
+      "epoch": 0.93,
+      "learning_rate": 2.9249233886194497e-07,
+      "loss": 0.1618,
+      "step": 1446
+    },
+    {
+      "epoch": 0.93,
+      "learning_rate": 2.875378347893165e-07,
+      "loss": 0.1822,
+      "step": 1447
+    },
+    {
+      "epoch": 0.93,
+      "learning_rate": 2.826250397887087e-07,
+      "loss": 0.1845,
+      "step": 1448
+    },
+    {
+      "epoch": 0.93,
+      "learning_rate": 2.7775397495756527e-07,
+      "loss": 0.1952,
+      "step": 1449
+    },
+    {
+      "epoch": 0.93,
+      "learning_rate": 2.7292466121412453e-07,
+      "loss": 0.1763,
+      "step": 1450
+    },
+    {
+      "epoch": 0.93,
+      "learning_rate": 2.681371192973281e-07,
+      "loss": 0.178,
+      "step": 1451
+    },
+    {
+      "epoch": 0.93,
+      "learning_rate": 2.633913697667334e-07,
+      "loss": 0.1603,
+      "step": 1452
+    },
+    {
+      "epoch": 0.93,
+      "learning_rate": 2.5868743300242913e-07,
+      "loss": 0.2244,
+      "step": 1453
+    },
+    {
+      "epoch": 0.93,
+      "learning_rate": 2.540253292049377e-07,
+      "loss": 0.2247,
+      "step": 1454
+    },
+    {
+      "epoch": 0.93,
+      "learning_rate": 2.494050783951396e-07,
+      "loss": 0.2148,
+      "step": 1455
+    },
+    {
+      "epoch": 0.93,
+      "learning_rate": 2.4482670041417687e-07,
+      "loss": 0.2011,
+      "step": 1456
+    },
+    {
+      "epoch": 0.93,
+      "learning_rate": 2.402902149233799e-07,
+      "loss": 0.2022,
+      "step": 1457
+    },
+    {
+      "epoch": 0.93,
+      "learning_rate": 2.3579564140417177e-07,
+      "loss": 0.1906,
+      "step": 1458
+    },
+    {
+      "epoch": 0.93,
+      "learning_rate": 2.3134299915799185e-07,
+      "loss": 0.1831,
+      "step": 1459
+    },
+    {
+      "epoch": 0.93,
+      "learning_rate": 2.2693230730620907e-07,
+      "loss": 0.1855,
+      "step": 1460
+    },
+    {
+      "epoch": 0.93,
+      "learning_rate": 2.2256358479004092e-07,
+      "loss": 0.1807,
+      "step": 1461
+    },
+    {
+      "epoch": 0.94,
+      "learning_rate": 2.1823685037047348e-07,
+      "loss": 0.1676,
+      "step": 1462
+    },
+    {
+      "epoch": 0.94,
+      "learning_rate": 2.1395212262817933e-07,
+      "loss": 0.2057,
+      "step": 1463
+    },
+    {
+      "epoch": 0.94,
+      "learning_rate": 2.0970941996343642e-07,
+      "loss": 0.1758,
+      "step": 1464
+    },
+    {
+      "epoch": 0.94,
+      "learning_rate": 2.0550876059605597e-07,
+      "loss": 0.185,
+      "step": 1465
+    },
+    {
+      "epoch": 0.94,
+      "learning_rate": 2.0135016256529138e-07,
+      "loss": 0.1825,
+      "step": 1466
+    },
+    {
+      "epoch": 0.94,
+      "learning_rate": 1.9723364372977395e-07,
+      "loss": 0.1809,
+      "step": 1467
+    },
+    {
+      "epoch": 0.94,
+      "learning_rate": 1.931592217674294e-07,
+      "loss": 0.1796,
+      "step": 1468
+    },
+    {
+      "epoch": 0.94,
+      "learning_rate": 1.8912691417540152e-07,
+      "loss": 0.1874,
+      "step": 1469
+    },
+    {
+      "epoch": 0.94,
+      "learning_rate": 1.851367382699798e-07,
+      "loss": 0.2031,
+      "step": 1470
+    },
+    {
+      "epoch": 0.94,
+      "learning_rate": 1.8118871118652516e-07,
+      "loss": 0.1907,
+      "step": 1471
+    },
+    {
+      "epoch": 0.94,
+      "learning_rate": 1.7728284987939437e-07,
+      "loss": 0.1814,
+      "step": 1472
+    },
+    {
+      "epoch": 0.94,
+      "learning_rate": 1.7341917112186801e-07,
+      "loss": 0.2039,
+      "step": 1473
+    },
+    {
+      "epoch": 0.94,
+      "learning_rate": 1.6959769150607598e-07,
+      "loss": 0.2232,
+      "step": 1474
+    },
+    {
+      "epoch": 0.94,
+      "learning_rate": 1.6581842744293309e-07,
+      "loss": 0.1995,
+      "step": 1475
+    },
+    {
+      "epoch": 0.94,
+      "learning_rate": 1.6208139516206034e-07,
+      "loss": 0.1707,
+      "step": 1476
+    },
+    {
+      "epoch": 0.94,
+      "learning_rate": 1.5838661071172268e-07,
+      "loss": 0.1929,
+      "step": 1477
+    },
+    {
+      "epoch": 0.95,
+      "learning_rate": 1.547340899587546e-07,
+      "loss": 0.17,
+      "step": 1478
+    },
+    {
+      "epoch": 0.95,
+      "learning_rate": 1.511238485884914e-07,
+      "loss": 0.1989,
+      "step": 1479
+    },
+    {
+      "epoch": 0.95,
+      "learning_rate": 1.4755590210471126e-07,
+      "loss": 0.1731,
+      "step": 1480
+    },
+    {
+      "epoch": 0.95,
+      "learning_rate": 1.4403026582955337e-07,
+      "loss": 0.2075,
+      "step": 1481
+    },
+    {
+      "epoch": 0.95,
+      "learning_rate": 1.4054695490346658e-07,
+      "loss": 0.1809,
+      "step": 1482
+    },
+    {
+      "epoch": 0.95,
+      "learning_rate": 1.3710598428513634e-07,
+      "loss": 0.1637,
+      "step": 1483
+    },
+    {
+      "epoch": 0.95,
+      "learning_rate": 1.337073687514212e-07,
+      "loss": 0.2204,
+      "step": 1484
+    },
+    {
+      "epoch": 0.95,
+      "learning_rate": 1.3035112289729423e-07,
+      "loss": 0.2295,
+      "step": 1485
+    },
+    {
+      "epoch": 0.95,
+      "learning_rate": 1.2703726113577175e-07,
+      "loss": 0.2111,
+      "step": 1486
+    },
+    {
+      "epoch": 0.95,
+      "learning_rate": 1.2376579769786124e-07,
+      "loss": 0.1824,
+      "step": 1487
+    },
+    {
+      "epoch": 0.95,
+      "learning_rate": 1.2053674663249027e-07,
+      "loss": 0.1787,
+      "step": 1488
+    },
+    {
+      "epoch": 0.95,
+      "learning_rate": 1.1735012180645544e-07,
+      "loss": 0.1826,
+      "step": 1489
+    },
+    {
+      "epoch": 0.95,
+      "learning_rate": 1.14205936904358e-07,
+      "loss": 0.1897,
+      "step": 1490
+    },
+    {
+      "epoch": 0.95,
+      "learning_rate": 1.1110420542854383e-07,
+      "loss": 0.1749,
+      "step": 1491
+    },
+    {
+      "epoch": 0.95,
+      "learning_rate": 1.0804494069904914e-07,
+      "loss": 0.2058,
+      "step": 1492
+    },
+    {
+      "epoch": 0.96,
+      "learning_rate": 1.0502815585354042e-07,
+      "loss": 0.1863,
+      "step": 1493
+    },
+    {
+      "epoch": 0.96,
+      "learning_rate": 1.020538638472579e-07,
+      "loss": 0.2048,
+      "step": 1494
+    },
+    {
+      "epoch": 0.96,
+      "learning_rate": 9.912207745296665e-08,
+      "loss": 0.2051,
+      "step": 1495
+    },
+    {
+      "epoch": 0.96,
+      "learning_rate": 9.623280926088995e-08,
+      "loss": 0.1947,
+      "step": 1496
+    },
+    {
+      "epoch": 0.96,
+      "learning_rate": 9.338607167866276e-08,
+      "loss": 0.2119,
+      "step": 1497
+    },
+    {
+      "epoch": 0.96,
+      "learning_rate": 9.058187693128051e-08,
+      "loss": 0.1953,
+      "step": 1498
+    },
+    {
+      "epoch": 0.96,
+      "learning_rate": 8.782023706103926e-08,
+      "loss": 0.1853,
+      "step": 1499
+    },
+    {
+      "epoch": 0.96,
+      "learning_rate": 8.510116392749013e-08,
+      "loss": 0.1848,
+      "step": 1500
+    },
+    {
+      "epoch": 0.96,
+      "learning_rate": 8.242466920738601e-08,
+      "loss": 0.1975,
+      "step": 1501
+    },
+    {
+      "epoch": 0.96,
+      "learning_rate": 7.97907643946294e-08,
+      "loss": 0.2083,
+      "step": 1502
+    },
+    {
+      "epoch": 0.96,
+      "learning_rate": 7.719946080023022e-08,
+      "loss": 0.1721,
+      "step": 1503
+    },
+    {
+      "epoch": 0.96,
+      "learning_rate": 7.465076955224693e-08,
+      "loss": 0.2095,
+      "step": 1504
+    },
+    {
+      "epoch": 0.96,
+      "learning_rate": 7.21447015957466e-08,
+      "loss": 0.1526,
+      "step": 1505
+    },
+    {
+      "epoch": 0.96,
+      "learning_rate": 6.968126769275496e-08,
+      "loss": 0.1904,
+      "step": 1506
+    },
+    {
+      "epoch": 0.96,
+      "learning_rate": 6.726047842221083e-08,
+      "loss": 0.1765,
+      "step": 1507
+    },
+    {
+      "epoch": 0.96,
+      "learning_rate": 6.488234417992179e-08,
+      "loss": 0.2153,
+      "step": 1508
+    },
+    {
+      "epoch": 0.97,
+      "learning_rate": 6.25468751785141e-08,
+      "loss": 0.1801,
+      "step": 1509
+    },
+    {
+      "epoch": 0.97,
+      "learning_rate": 6.025408144739842e-08,
+      "loss": 0.2081,
+      "step": 1510
+    },
+    {
+      "epoch": 0.97,
+      "learning_rate": 5.800397283271753e-08,
+      "loss": 0.1582,
+      "step": 1511
+    },
+    {
+      "epoch": 0.97,
+      "learning_rate": 5.5796558997306404e-08,
+      "loss": 0.1793,
+      "step": 1512
+    },
+    {
+      "epoch": 0.97,
+      "learning_rate": 5.363184942065891e-08,
+      "loss": 0.1986,
+      "step": 1513
+    },
+    {
+      "epoch": 0.97,
+      "learning_rate": 5.150985339887227e-08,
+      "loss": 0.1443,
+      "step": 1514
+    },
+    {
+      "epoch": 0.97,
+      "learning_rate": 4.9430580044620425e-08,
+      "loss": 0.216,
+      "step": 1515
+    },
+    {
+      "epoch": 0.97,
+      "learning_rate": 4.7394038287108534e-08,
+      "loss": 0.1905,
+      "step": 1516
+    },
+    {
+      "epoch": 0.97,
+      "learning_rate": 4.540023687203299e-08,
+      "loss": 0.1886,
+      "step": 1517
+    },
+    {
+      "epoch": 0.97,
+      "learning_rate": 4.3449184361551434e-08,
+      "loss": 0.1651,
+      "step": 1518
+    },
+    {
+      "epoch": 0.97,
+      "learning_rate": 4.154088913423615e-08,
+      "loss": 0.1709,
+      "step": 1519
+    },
+    {
+      "epoch": 0.97,
+      "learning_rate": 3.967535938504741e-08,
+      "loss": 0.1947,
+      "step": 1520
+    },
+    {
+      "epoch": 0.97,
+      "learning_rate": 3.7852603125291265e-08,
+      "loss": 0.2034,
+      "step": 1521
+    },
+    {
+      "epoch": 0.97,
+      "learning_rate": 3.6072628182588495e-08,
+      "loss": 0.185,
+      "step": 1522
+    },
+    {
+      "epoch": 0.97,
+      "learning_rate": 3.433544220084017e-08,
+      "loss": 0.1773,
+      "step": 1523
+    },
+    {
+      "epoch": 0.98,
+      "learning_rate": 3.264105264019657e-08,
+      "loss": 0.2042,
+      "step": 1524
+    },
+    {
+      "epoch": 0.98,
+      "learning_rate": 3.0989466777021634e-08,
+      "loss": 0.195,
+      "step": 1525
+    },
+    {
+      "epoch": 0.98,
+      "learning_rate": 2.938069170386415e-08,
+      "loss": 0.1875,
+      "step": 1526
+    },
+    {
+      "epoch": 0.98,
+      "learning_rate": 2.7814734329426607e-08,
+      "loss": 0.1796,
+      "step": 1527
+    },
+    {
+      "epoch": 0.98,
+      "learning_rate": 2.629160137853748e-08,
+      "loss": 0.2041,
+      "step": 1528
+    },
+    {
+      "epoch": 0.98,
+      "learning_rate": 2.4811299392117904e-08,
+      "loss": 0.2216,
+      "step": 1529
+    },
+    {
+      "epoch": 0.98,
+      "learning_rate": 2.3373834727158374e-08,
+      "loss": 0.2097,
+      "step": 1530
+    },
+    {
+      "epoch": 0.98,
+      "learning_rate": 2.197921355668875e-08,
+      "loss": 0.1704,
+      "step": 1531
+    },
+    {
+      "epoch": 0.98,
+      "learning_rate": 2.0627441869750518e-08,
+      "loss": 0.1751,
+      "step": 1532
+    },
+    {
+      "epoch": 0.98,
+      "learning_rate": 1.9318525471376803e-08,
+      "loss": 0.1716,
+      "step": 1533
+    },
+    {
+      "epoch": 0.98,
+      "learning_rate": 1.805246998256016e-08,
+      "loss": 0.2148,
+      "step": 1534
+    },
+    {
+      "epoch": 0.98,
+      "learning_rate": 1.682928084023261e-08,
+      "loss": 0.1766,
+      "step": 1535
+    },
+    {
+      "epoch": 0.98,
+      "learning_rate": 1.5648963297242302e-08,
+      "loss": 0.1939,
+      "step": 1536
+    },
+    {
+      "epoch": 0.98,
+      "learning_rate": 1.4511522422330226e-08,
+      "loss": 0.2156,
+      "step": 1537
+    },
+    {
+      "epoch": 0.98,
+      "learning_rate": 1.3416963100105762e-08,
+      "loss": 0.1689,
+      "step": 1538
+    },
+    {
+      "epoch": 0.98,
+      "learning_rate": 1.2365290031032261e-08,
+      "loss": 0.2005,
+      "step": 1539
+    },
+    {
+      "epoch": 0.99,
+      "learning_rate": 1.1356507731401512e-08,
+      "loss": 0.2108,
+      "step": 1540
+    },
+    {
+      "epoch": 0.99,
+      "learning_rate": 1.0390620533312635e-08,
+      "loss": 0.1962,
+      "step": 1541
+    },
+    {
+      "epoch": 0.99,
+      "learning_rate": 9.46763258466099e-09,
+      "loss": 0.1696,
+      "step": 1542
+    },
+    {
+      "epoch": 0.99,
+      "learning_rate": 8.587547849112643e-09,
+      "loss": 0.1842,
+      "step": 1543
+    },
+    {
+      "epoch": 0.99,
+      "learning_rate": 7.750370106094363e-09,
+      "loss": 0.2257,
+      "step": 1544
+    },
+    {
+      "epoch": 0.99,
+      "learning_rate": 6.956102950770316e-09,
+      "loss": 0.1884,
+      "step": 1545
+    },
+    {
+      "epoch": 0.99,
+      "learning_rate": 6.204749794029852e-09,
+      "loss": 0.2022,
+      "step": 1546
+    },
+    {
+      "epoch": 0.99,
+      "learning_rate": 5.496313862476399e-09,
+      "loss": 0.2041,
+      "step": 1547
+    },
+    {
+      "epoch": 0.99,
+      "learning_rate": 4.830798198406373e-09,
+      "loss": 0.2201,
+      "step": 1548
+    },
+    {
+      "epoch": 0.99,
+      "learning_rate": 4.208205659802511e-09,
+      "loss": 0.2041,
+      "step": 1549
+    },
+    {
+      "epoch": 0.99,
+      "learning_rate": 3.628538920317226e-09,
+      "loss": 0.1739,
+      "step": 1550
+    },
+    {
+      "epoch": 0.99,
+      "learning_rate": 3.0918004692648273e-09,
+      "loss": 0.166,
+      "step": 1551
+    },
+    {
+      "epoch": 0.99,
+      "learning_rate": 2.597992611608202e-09,
+      "loss": 0.2307,
+      "step": 1552
+    },
+    {
+      "epoch": 0.99,
+      "learning_rate": 2.147117467949933e-09,
+      "loss": 0.2121,
+      "step": 1553
+    },
+    {
+      "epoch": 0.99,
+      "learning_rate": 1.7391769745200848e-09,
+      "loss": 0.1956,
+      "step": 1554
+    },
+    {
+      "epoch": 0.99,
+      "learning_rate": 1.3741728831750956e-09,
+      "loss": 0.2184,
+      "step": 1555
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 1.0521067613822323e-09,
+      "loss": 0.189,
+      "step": 1556
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 7.729799922195913e-10,
+      "loss": 0.1858,
+      "step": 1557
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 5.367937743649964e-10,
+      "loss": 0.1947,
+      "step": 1558
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 3.435491220937781e-10,
+      "loss": 0.1797,
+      "step": 1559
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 1.93246865272112e-10,
+      "loss": 0.1643,
+      "step": 1560
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 8.588764935812955e-11,
+      "loss": 0.1633,
+      "step": 1561
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 2.1471935391925713e-11,
+      "loss": 0.1791,
+      "step": 1562
+    },
+    {
+      "epoch": 1.0,
+      "learning_rate": 0.0,
+      "loss": 0.2095,
+      "step": 1563
+    },
+    {
+      "epoch": 1.0,
+      "step": 1563,
+      "total_flos": 8.705675574102259e+22,
+      "train_loss": 0.2210330774939678,
+      "train_runtime": 11823.142,
+      "train_samples_per_second": 8.459,
+      "train_steps_per_second": 0.132
+    }
+  ],
+  "logging_steps": 1.0,
+  "max_steps": 1563,
+  "num_input_tokens_seen": 0,
+  "num_train_epochs": 1,
+  "save_steps": 2000,
+  "total_flos": 8.705675574102259e+22,
+  "train_batch_size": 1,
+  "trial_name": null,
+  "trial_params": null
+}