Spaces:
Sleeping
Sleeping
File size: 2,655 Bytes
1b41e6d f1e71b2 abd911f 1b41e6d f1e71b2 1b41e6d f1e71b2 100a302 f1e71b2 1b41e6d f1e71b2 1b41e6d f1e71b2 1b41e6d f1e71b2 100a302 f1e71b2 100a302 f1e71b2 100a302 f1e71b2 1b41e6d f1e71b2 1b41e6d f1e71b2 |
1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 16 17 18 19 20 21 22 23 24 25 26 27 28 29 30 31 32 33 34 35 36 37 38 39 40 41 42 43 44 45 46 47 48 49 50 51 52 53 54 55 56 57 58 59 60 61 62 63 64 65 66 67 68 69 70 71 72 73 74 75 76 77 78 79 80 |
import gradio as gr
import edge_tts
import asyncio
import tempfile
# 1) Fetch voices once, synchronously at startup
def load_voices():
loop = asyncio.get_event_loop()
voices = loop.run_until_complete(edge_tts.list_voices())
return {
f"{v['ShortName']} - {v['Locale']} ({v['Gender']})": v['ShortName']
for v in voices
}
VOICES = load_voices()
# 2) Async function to drive Edge TTS
async def _text_to_speech(text, short_name, rate_str, pitch_str):
communicate = edge_tts.Communicate(text, short_name, rate=rate_str, pitch=pitch_str)
with tempfile.NamedTemporaryFile(delete=False, suffix=".mp3") as tmp:
await communicate.save(tmp.name)
return tmp.name
# 3) Sync wrapper for the click callback
def tts_interface(text, voice_choice, rate, pitch):
if not text.strip():
return None, "🚨 Please enter some text."
if not voice_choice:
return None, "🚨 Please select a voice."
short_name = voice_choice.split(" - ")[0]
rate_str = f"{rate:+d}%"
pitch_str = f"{pitch:+d}Hz"
try:
# Run the async TTS call synchronously
audio_path = asyncio.get_event_loop().run_until_complete(
_text_to_speech(text, short_name, rate_str, pitch_str)
)
return audio_path, ""
except Exception as e:
return None, f"❌ TTS failed: {e}"
# 4) Build the Gradio Blocks UI
def create_demo():
with gr.Blocks(analytics_enabled=False) as demo:
gr.Markdown("# 🎙️ Edge TTS on Hugging Face Spaces")
with gr.Row():
gr.Markdown("""
**Convert your text to speech** using Microsoft Edge's neural voices.
Adjust rate and pitch to fine-tune the output.
""")
with gr.Row():
txt = gr.Textbox(label="Input Text", lines=5, placeholder="Type something…")
vox = gr.Dropdown(choices=list(VOICES.keys()), label="Voice")
rate = gr.Slider(-50, 50, value=0, label="Rate (%)")
pitch = gr.Slider(-20, 20, value=0, label="Pitch (Hz)")
btn = gr.Button("Generate Speech")
audio_out = gr.Audio(type="filepath", label="Audio Output")
warn_md = gr.Markdown("", label="Warnings / Errors")
# IMPORTANT: wire queue here so the frontend finds /api/predict
btn.click(
fn=tts_interface,
inputs=[txt, vox, rate, pitch],
outputs=[audio_out, warn_md]
)
# Enable the queue on the entire app
demo.queue()
return demo
# 5) Launch
if __name__ == "__main__":
demo = create_demo()
# On Spaces, this will serve on 0.0.0.0:7860 automatically
demo.launch()
|