Update app.py

app.py

@@ -1,27 +1,18 @@
-# app.py
 from fastapi import FastAPI, UploadFile, File, Form
-from fastapi.responses import FileResponse, HTMLResponse
+from fastapi.responses import FileResponse
 import torch
 import torchaudio
 import os
-from pathlib import Path
 from TTS.tts.models.xtts import Xtts
 from TTS.tts.configs.xtts_config import XttsConfig
-import gradio as gr
-import uvicorn
 
-# ------------------------
-# Setup paths
-# ------------------------
-MODEL_DIR = "my_model"          # folder with config.json, vocab.json, model.pth
+MODEL_DIR = "my_model"
 OUTPUT_DIR = "outputs"
 os.makedirs(OUTPUT_DIR, exist_ok=True)
 
 device = "cuda" if torch.cuda.is_available() else "cpu"
 
-# ------------------------
-# Load TTS model
-# ------------------------
+# Load model
 config = XttsConfig()
 config.load_json(os.path.join(MODEL_DIR, "config.json"))
 
@@ -34,9 +25,6 @@ model.load_checkpoint(
 )
 model.to(device)
 
-# ------------------------
-# TTS function
-# ------------------------
 def tts_arabic(text: str, audio_file: str) -> str:
     gpt_cond_latent, speaker_embedding = model.get_conditioning_latents(audio_path=[audio_file])
     out = model.inference(
@@ -54,58 +42,20 @@ def tts_arabic(text: str, audio_file: str) -> str:
     torchaudio.save(output_wav, torch.tensor(out["wav"]).unsqueeze(0), 24000)
     return output_wav
 
-# ------------------------
-# FastAPI setup
-# ------------------------
-app = FastAPI(title="EGTTS TTS API")
+app = FastAPI(title="EGTTS Arabic TTS API")
 
-@app.get("/", response_class=HTMLResponse)
-def index():
-    """Return simple HTML that links to Gradio UI"""
-    return """
-    <h2>Welcome to EGTTS TTS API</h2>
-    <p>Swagger docs available at <a href="/docs">/docs</a></p>
-    <p>Try the Gradio interface at <a href="/gradio">/gradio</a></p>
-    """
+@app.get("/")
+def root():
+    return {"message": "Welcome! Visit /docs for Swagger UI."}
 
 @app.post("/tts/")
 async def tts_endpoint(
     text: str = Form(...),
     audio_file: UploadFile = File(...)
 ):
-    # Save uploaded file
     file_path = os.path.join(OUTPUT_DIR, audio_file.filename)
     with open(file_path, "wb") as f:
         f.write(await audio_file.read())
 
     output_wav = tts_arabic(text, file_path)
-    return FileResponse(output_wav, media_type="audio/wav", filename="output.wav")
-
-# ------------------------
-# Gradio interface
-# ------------------------
-def gradio_fn(text, audio_file):
-    return tts_arabic(text, audio_file.name)
-
-gradio_interface = gr.Interface(
-    fn=gradio_fn,
-    inputs=[
-        gr.Textbox(label="Arabic Text", placeholder="اكتب النص هنا..."),
-        gr.File(label="Speaker Audio (.wav)")
-    ],
-    outputs=gr.Audio(label="Generated Speech"),
-    live=True,
-    title="EGTTS Arabic TTS",
-    description="Generate Arabic speech from text using your fine-tuned EGTTS model."
-)
-
-# Mount Gradio inside FastAPI
-@app.get("/gradio", response_class=HTMLResponse)
-def gradio_ui():
-    return gradio_interface.launch(inline=True, share=False, prevent_thread_lock=True).read()
-
-# ------------------------
-# Run server
-# ------------------------
-if __name__ == "__main__":
-    uvicorn.run(app, host="0.0.0.0", port=7860)
+    return FileResponse(output_wav, media_type="audio/wav", filename="output.wav")