Upload edit\Qwen3-TTS-test\reference\README.md with huggingface_hub

edit//Qwen3-TTS-test//reference//README.md

@@ -0,0 +1,32 @@
+# Reference audio
+
+Drop two files here before running `python ../clone.py`:
+
+## 1. `reference.wav`
+
+A **10–15 second** clip of your voice — your cleanest, most natural sample.
+
+Quality rules:
+- Quiet room, no AC/fan hum, no background music
+- Consistent mic distance
+- Normal speaking pace and intonation (not reading flat)
+- 24 kHz or 44.1 kHz mono WAV is ideal (F5-TTS handles resampling but cleaner input = cleaner output)
+
+From your 30-min recording, pick the *best* 10–15 seconds. Don't just take the first clip.
+
+To trim with ffmpeg:
+```
+ffmpeg -i your-30min.wav -ss 00:05:12 -t 12 -ac 1 -ar 24000 reference.wav
+```
+(Adjust `-ss` start time and `-t` length.)
+
+## 2. `reference.txt`
+
+The **exact transcript** of `reference.wav`, on a single line. Punctuation matters. No extra commentary.
+
+Example:
+```
+Hey, my name is Palash and I make videos about AI and coding tools.
+```
+
+That's it. Once both files are here, run `python ../clone.py`.