support timbre confusion

inference.py

@@ -66,10 +66,22 @@ def build_parser():
         default=None,
     )
     parser.add_argument(
-        "--speaker_name",
+        "--speaker_name_1",
         type=str,
         default=None,
-        help="speaker name for multi-speaker synthesis, for single-sentence mode only",
+        help="first speaker name for multi-speaker synthesis, for single-sentence mode only",
+    )
+    parser.add_argument(
+        "--speaker_name_2",
+        type=str,
+        default=None,
+        help="second speaker name for multi-speaker synthesis, for single-sentence mode only",
+    )
+    parser.add_argument(
+        "--alpha",
+        type=float,
+        default=0.5,
+        help="confusion degree, takes values within [0,1]",
     )
     parser.add_argument(
         "--text",