Spaces:
Sleeping
Sleeping
import openai | |
import streamlit as st | |
from youtube_transcript_api import YouTubeTranscriptApi | |
import re | |
import tempfile | |
import os | |
from pydub import AudioSegment | |
import logging | |
import warnings | |
def convert_to_supported_format(file_path): | |
audio = AudioSegment.from_file(file_path) | |
with tempfile.NamedTemporaryFile(suffix=".wav", delete=False) as temp: | |
audio.export(temp.name, format="wav") | |
return temp.name | |
def transcribe_audio(file_path): | |
logging.info(f"Transcribing audio file: {file_path}") | |
file_path = convert_to_supported_format(file_path) | |
logging.info(f"Converted file path: {file_path}") | |
with warnings.catch_warnings(): | |
warnings.simplefilter("ignore") | |
with open(file_path, "rb") as audio_file: | |
transcript = openai.Audio.transcribe("whisper-1", audio_file) | |
os.remove(file_path) # Clean up temporary file | |
return transcript["text"] | |
def get_transcript(url): | |
try: | |
video_id_match = re.search(r"(?:v=|\/)([0-9A-Za-z_-]{11}).*", url) | |
if video_id_match: | |
video_id = video_id_match.group(1) | |
else: | |
return "Error: Invalid YouTube URL" | |
transcript = YouTubeTranscriptApi.get_transcript(video_id) | |
transcript_text = ' '.join([entry['text'] for entry in transcript]) | |
return transcript_text | |
except Exception as e: | |
return str(e) | |
def summarize_text(text): | |
response = openai.ChatCompletion.create( | |
model="gpt-3.5-turbo", | |
messages=[ | |
{"role": "system", "content": "You are a helpful assistant."}, | |
{"role": "user", "content": f"Summarize the following text:\n\n{text}"} | |
], | |
max_tokens=150 | |
) | |
summary = response['choices'][0]['message']['content'].strip() | |
return summary | |
def generate_quiz_questions(text): | |
response = openai.ChatCompletion.create( | |
model="gpt-3.5-turbo", | |
messages=[ | |
{"role": "system", "content": "You are a helpful assistant that generates quiz questions. Your task is to generate ten quiz questions and four multiple choice answers for each question from the given text. Make sure to mark the correct answer with an asterisk (*) at the beginning of the answer line. Use the following format for each question:\n\n1. Question\n a) Answer 1\n b) Answer 2\n c) Answer 3\n d) Answer 4\n\n2. Question\n a) Answer 1\n b) Answer 2\n c) Answer 3\n d) Answer 4\n\n..."}, | |
{"role": "user", "content": f"Generate quiz questions from the following text:\n\n{text}"} | |
], | |
max_tokens=300 | |
) | |
quiz_questions = response['choices'][0]['message']['content'].strip() | |
return quiz_questions | |
def parse_quiz_questions(quiz_text): | |
questions = [] | |
question_blocks = quiz_text.split("\n\n") | |
for block in question_blocks: | |
lines = block.strip().split("\n") | |
if len(lines) >= 5: | |
question = lines[0].split(". ")[1] | |
choices = [line.split(") ")[1].strip() for line in lines[1:5]] | |
correct_answer_lines = [line for line in lines[1:5] if "*" in line] | |
if correct_answer_lines: | |
correct_answer = correct_answer_lines[0].split(") ")[1].replace("*", "").strip() | |
else: | |
correct_answer = "No correct answer provided" | |
questions.append({"question": question, "choices": choices, "correct_answer": correct_answer}) | |
return questions | |
def generate_explanation(question, correct_answer, user_answer): | |
prompt = f"Explain why the correct answer to the following question is '{correct_answer}' and not '{user_answer}':\n\n{question}" | |
response = openai.ChatCompletion.create( | |
model="gpt-3.5-turbo", | |
messages=[ | |
{"role": "system", "content": "You are a helpful assistant."}, | |
{"role": "user", "content": prompt} | |
], | |
max_tokens=150 | |
) | |
explanation = response['choices'][0]['message']['content'].strip() | |
return explanation | |
def check_answers(questions, user_answers): | |
feedback = [] | |
correct_count = 0 | |
for i, question in enumerate(questions): | |
correct_answer = question['correct_answer'] | |
user_answer = user_answers.get(f"question_{i+1}", "") | |
if user_answer == correct_answer: | |
feedback.append({ | |
"question": question['question'], | |
"user_answer": user_answer, | |
"correct_answer": correct_answer, | |
"status": "Correct" | |
}) | |
correct_count += 1 | |
else: | |
explanation = generate_explanation(question['question'], correct_answer, user_answer) | |
feedback.append({ | |
"question": question['question'], | |
"user_answer": user_answer, | |
"correct_answer": correct_answer, | |
"status": "Incorrect", | |
"explanation": explanation | |
}) | |
return feedback | |
def handle_uploaded_file(uploaded_file): | |
with tempfile.NamedTemporaryFile(delete=False) as tmp_file: | |
tmp_file.write(uploaded_file.read()) | |
tmp_file_path = tmp_file.name | |
return tmp_file_path | |
st.title("YouTube Transcript Quiz Generator") | |
st.markdown("**Instructions:** Paste a YouTube link or upload a media file to generate a quiz.") | |
option = st.selectbox("Choose input type", ("YouTube URL", "Upload audio/video file")) | |
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 url: | |
if st.button("Generate Quiz"): | |
transcript_text = get_transcript(url) | |
if "Error" not in transcript_text: | |
summary = summarize_text(transcript_text) | |
quiz_text = generate_quiz_questions(transcript_text) | |
questions = parse_quiz_questions(quiz_text) | |
st.write("## Summary") | |
st.write(summary) | |
st.write("## Quiz Questions") | |
st.session_state.questions = questions | |
st.session_state.user_answers = {} | |
st.session_state.generated_quiz = True | |
for i, question in enumerate(questions): | |
st.write(f"### Question {i+1}") | |
st.write(question['question']) | |
st.session_state.user_answers[f"question_{i+1}"] = st.radio( | |
label="", | |
options=question['choices'], | |
key=f"question_{i+1}" | |
) | |
elif option == "Upload audio/video file": | |
uploaded_file = st.file_uploader("Choose an audio or video file", type=["mp3", "wav", "mp4", "mov"]) | |
if uploaded_file: | |
tmp_file_path = handle_uploaded_file(uploaded_file) | |
transcript_text = transcribe_audio(tmp_file_path) | |
os.remove(tmp_file_path) | |
if "Error" not in transcript_text: | |
summary = summarize_text(transcript_text) | |
quiz_text = generate_quiz_questions(transcript_text) | |
questions = parse_quiz_questions(quiz_text) | |
st.write("## Summary") | |
st.write(summary) | |
st.write("## Quiz Questions") | |
st.session_state.questions = questions | |
st.session_state.user_answers = {} | |
st.session_state.generated_quiz = True | |
for i, question in enumerate(questions): | |
st.write(f"### Question {i+1}") | |
st.write(question['question']) | |
st.session_state.user_answers[f"question_{i+1}"] = st.radio( | |
label="", | |
options=question['choices'], | |
key=f"question_{i+1}" | |
) | |
if st.session_state.generated_quiz: | |
if st.button("Submit Answers"): | |
if "questions" in st.session_state and st.session_state.questions: | |
with st.spinner('Processing your answers...'): | |
feedback = check_answers(st.session_state.questions, st.session_state.user_answers) | |
st.write("## Feedback") | |
for i, item in enumerate(feedback): | |
with st.expander(f"Question {i+1} Feedback"): | |
st.write(f"### {item['question']}") | |
st.write(f"**Your answer:** {item['user_answer']}") | |
st.write(f"**Correct answer:** {item['correct_answer']}") | |
if item['status'] == "Incorrect": | |
st.write(f"**Explanation:** {item['explanation']}") | |
else: | |
st.write("Please generate the quiz first.") | |