Spaces:
Running
Running
Create app.py
Browse files
app.py
ADDED
@@ -0,0 +1,43 @@
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
1 |
+
from fastapi import FastAPI, HTTPException, UploadFile, File
|
2 |
+
from fastapi.responses import FileResponse
|
3 |
+
from kokoro import KPipeline
|
4 |
+
import soundfile as sf
|
5 |
+
import os
|
6 |
+
import tempfile
|
7 |
+
|
8 |
+
app = FastAPI(title="Text-to-Speech Converter")
|
9 |
+
|
10 |
+
# Initialize pipeline once at startup
|
11 |
+
pipeline = KPipeline(lang_code='a')
|
12 |
+
|
13 |
+
@app.post("/generate_audio/")
|
14 |
+
async def generate_audio(text: str = None):
|
15 |
+
if not text:
|
16 |
+
raise HTTPException(status_code=400, detail="No text provided")
|
17 |
+
|
18 |
+
try:
|
19 |
+
# Create temporary directory
|
20 |
+
with tempfile.TemporaryDirectory() as tmpdir:
|
21 |
+
# Generate audio
|
22 |
+
generator = pipeline(
|
23 |
+
text,
|
24 |
+
voice='af_heart',
|
25 |
+
speed=1,
|
26 |
+
split_pattern=r'\n+'
|
27 |
+
)
|
28 |
+
|
29 |
+
# Process first audio segment only (modify as needed)
|
30 |
+
i, (gs, ps, audio) = next(enumerate(generator))
|
31 |
+
|
32 |
+
# Save to temporary file
|
33 |
+
output_path = f"{tmpdir}/output.wav"
|
34 |
+
sf.write(output_path, audio, 24000)
|
35 |
+
|
36 |
+
return FileResponse(
|
37 |
+
output_path,
|
38 |
+
media_type='audio/wav',
|
39 |
+
filename="generated_audio.wav"
|
40 |
+
)
|
41 |
+
|
42 |
+
except Exception as e:
|
43 |
+
raise HTTPException(status_code=500, detail=str(e))
|