Update app.py

app.py

@@ -1,21 +1,19 @@
 import streamlit as st
+import requests
 from openai import OpenAI
 from youtube_transcript_api import YouTubeTranscriptApi
 import re
 import tempfile
 import os
-from transformers import pipeline
-import soundfile as sf
 
-# Initialize the pipeline with the model
-pipe = pipeline("automatic-speech-recognition", model="openai/whisper-small")
-
-# Function to transcribe audio using Hugging Face Whisper
-def transcribe_audio(file_path):
-    # Load audio file into NumPy array
-    audio_input, _ = sf.read(file_path)
-    transcription = pipe(audio_input)["text"]
-    return transcription
+# Function to transcribe audio using Hugging Face Inference API
+def transcribe_audio(api_key, file_path):
+    API_URL = "https://api-inference.huggingface.co/models/openai/whisper-large"
+    headers = {"Authorization": f"Bearer {api_key}"}
+    with open(file_path, "rb") as f:
+        data = f.read()
+    response = requests.post(API_URL, headers=headers, data=data)
+    return response.json().get("text", "Error: Could not transcribe audio")
 
 # Function to get YouTube transcript
 def get_transcript(url):
@@ -122,21 +120,21 @@ def handle_uploaded_file(uploaded_file):
 # Streamlit UI
 st.title("YouTube Transcript Quiz Generator")
 
-st.markdown("**Instructions:** Enter your OpenAI API key and paste a YouTube link or upload a media file to generate a quiz.")
-
-api_key = st.text_input("Enter your OpenAI API Key", type="password")
-
-if api_key:
-    client = OpenAI(api_key=api_key)
+st.markdown("**Instructions:** Enter your OpenAI and Hugging Face API keys, and paste a YouTube link or upload a media file to generate a quiz.")
 
+openai_api_key = st.text_input("Enter your OpenAI API Key", type="password")
+hf_api_key = st.text_input("Enter your Hugging Face API Key", type="password")
 option = st.selectbox("Choose input type", ("YouTube URL", "Upload audio/video file"))
 
+if openai_api_key:
+    client = OpenAI(api_key=openai_api_key)
+
 if "generated_quiz" not in st.session_state:
     st.session_state.generated_quiz = False
 
 if option == "YouTube URL":
     url = st.text_input("YouTube URL", value="")
-    if api_key and url:
+    if openai_api_key and hf_api_key and url:
         if st.button("Generate Quiz"):
             transcript_text = get_transcript(url)
             if "Error" not in transcript_text:
@@ -151,11 +149,11 @@ if option == "YouTube URL":
 
 if option == "Upload audio/video file":
     uploaded_file = st.file_uploader("Choose an audio or video file", type=["mp3", "wav", "mp4", "mov"])
-    if uploaded_file and api_key:
+    if uploaded_file and openai_api_key and hf_api_key:
         if st.button("Generate Quiz"):
             tmp_file_path = handle_uploaded_file(uploaded_file)
             with st.spinner('Transcribing audio...'):
-                transcript_text = transcribe_audio(tmp_file_path)
+                transcript_text = transcribe_audio(hf_api_key, tmp_file_path)
             os.remove(tmp_file_path)
             if "Error" not in transcript_text:
                 summary = summarize_text(client, transcript_text)