FX: update README

README.md

@@ -50,7 +50,7 @@ By default, when you first run the script, the model will be downloaded automati
 - Download VoxCPM-0.5B
     ```
     from huggingface_hub import snapshot_download
-    snapshot_download("openbmb/VoxCPM-0.5B",local_files_only=local_files_only)
+    snapshot_download("openbmb/VoxCPM-0.5B")
     ```
 - Download ZipEnhancer and SenseVoice-Small. We use ZipEnhancer to enhance speech prompts and SenseVoice-Small for speech prompt ASR in the web demo.
     ```
@@ -98,6 +98,13 @@ voxcpm --text "VoxCPM is an innovative end-to-end TTS model from ModelBest, desi
   --output out.wav \
   --denoise
 
+# (Optinal) Voice cloning (reference audio + transcript file)
+voxcpm --text "VoxCPM is an innovative end-to-end TTS model from ModelBest, designed to generate highly expressive speech." \
+  --prompt-audio path/to/voice.wav \
+  --prompt-file "/path/to/text-file" \
+  --output out.wav \
+  --denoise
+
 # 3) Batch processing (one text per line)
 voxcpm --input examples/input.txt --output-dir outs
 # (optional) Batch + cloning