app.py

import gradio as gr
import openai
import os
from io import BytesIO
import tempfile
import boto3

s3 = boto3.client(
    's3',
    aws_access_key_id=os.environ["AWS_ACCESS_KEY_ID"],
    aws_secret_access_key=os.environ["AWS_SECRET_ACCESS_KEY"],
    region_name='ap-northeast-1'
)

def create_meeting_summary(uploaded_audio):
    openai.api_key = os.environ["OPENAI_API_KEY"]
    transcript = openai.Audio.transcribe("whisper-1", open(uploaded_audio, "rb"), response_format="verbose_json")
    transcript_text = ""
    for segment in transcript.segments:
        transcript_text += f"{segment['text']}\n"

    bucket_name = "ling-transcribe"
    file_name = f"test.txt"
    s3.put_object(Bucket=bucket_name, Key=file_name, Body=transcript_text)

    download_url = os.environ["DOWNLOAD_URL"] + "test.txt"
    
    return download_url

inputs = [
    gr.Audio(type="filepath", label="音声ファイルをアップロード")
]

outputs = [
    gr.Textbox(label="ダウンロードURL")
]

app = gr.Interface(
    fn=create_meeting_summary,
    inputs=inputs,
    outputs=outputs,
    title="音声文字起こしアプリ",
    description="音声ファイルをアップロードして、文字起こしファイルを作成します。"
)

app.launch(debug=True, auth=(os.environ["USERNAME"], os.environ["PASSWORD"]))