main.py

from fastapi import FastAPI, File, UploadFile
from fastapi.responses import StreamingResponse
import os
import io

from faster_whisper import WhisperModel
model_size = "large-v2"
transcripe_model = WhisperModel(model_size, device="cuda", compute_type="float16")


app = FastAPI()


@app.post("/speech_to_text/")
async def create_upload_file(file: UploadFile = File(...)):
    try:
        # Save the file with a specific name
        file_path = "inputvoice.mp3"

        with open(file_path, "wb") as f:
            f.write(file.file.read())

        # Read the content of the saved file
        #with open(file_path, "rb") as f:
            #file_content = f.read()
        segments, info = transcripe_model.transcribe(file_path, beam_size=5)
        
        print("Detected language '%s' with probability %f" % (info.language, info.language_probability))
        result = ""
        for segment in segments:
            result = result + segment.text
        # Return the content as a streaming response
        #return StreamingResponse(io.BytesIO(file_content), media_type="audio/mpeg", headers={"Content-Disposition": "inline; filename=inputvoice.mp3"})

        return "result"  
    except PermissionError as e:
        return {"error": f"PermissionError: {str(e)}"}