Spaces:
Sleeping
Sleeping
| import streamlit as st | |
| from transformers import AutoTokenizer, AutoModelForSeq2SeqLM | |
| # Load the tokenizer and model | |
| tokenizer = AutoTokenizer.from_pretrained("suriya7/bart-finetuned-text-summarization") | |
| model = AutoModelForSeq2SeqLM.from_pretrained("suriya7/bart-finetuned-text-summarization") | |
| def generate_user_stories(text, prompt): | |
| try: | |
| # Combine prompt with the text to guide the summarization | |
| combined_input = f"Prompt: {prompt}\n\nText: {text}" | |
| # Tokenize input with truncation to fit model requirements | |
| inputs = tokenizer([combined_input], max_length=1024, return_tensors='pt', truncation=True) | |
| # Generate summary | |
| summary_ids = model.generate(inputs['input_ids'], max_new_tokens=150, do_sample=False) | |
| summary = tokenizer.decode(summary_ids[0], skip_special_tokens=True) | |
| # Post-process to format as user stories | |
| user_stories = format_as_user_stories(summary) | |
| return user_stories | |
| except Exception as e: | |
| st.error(f"An error occurred: {e}") | |
| return "" | |
| def format_as_user_stories(summary): | |
| user_stories = [] | |
| lines = summary.split('. ') | |
| for line in lines: | |
| line = line.strip() | |
| if "as a" in line.lower() and "i want" in line.lower(): | |
| # Extract the parts of the user story | |
| parts = line.split("so that") | |
| if len(parts) == 2: | |
| story = { | |
| "As a": parts[0].strip().capitalize(), | |
| "I want to": parts[1].strip().capitalize(), | |
| "So that": parts[1].strip().capitalize() | |
| } | |
| user_stories.append(story) | |
| # Format user stories | |
| formatted_stories = "" | |
| for story in user_stories: | |
| formatted_stories += f"**User Story:**\n\n" \ | |
| f"**As a:** {story['As a']}\n" \ | |
| f"**I want to:** {story['I want to']}\n" \ | |
| f"**So that:** {story['So that']}\n\n" | |
| return formatted_stories | |
| # Initialize session state for input history if it doesn't exist | |
| if 'input_history' not in st.session_state: | |
| st.session_state['input_history'] = [] | |
| # Streamlit interface | |
| st.title("User Story Generator") | |
| # User text inputs | |
| bulk_text = st.text_area("Enter the bulk text (e.g., client calls, meeting transcripts)", height=300) | |
| prompt = st.text_input("Enter the prompt for the user stories", "Extract user stories from the following text.") | |
| if st.button("Generate User Stories"): | |
| if bulk_text and prompt: | |
| with st.spinner("Generating user stories..."): | |
| user_stories = generate_user_stories(bulk_text, prompt) | |
| if user_stories: | |
| # Save the input and user stories to the session state history | |
| st.session_state['input_history'].append({"text": bulk_text, "prompt": prompt, "user_stories": user_stories}) | |
| st.subheader("Generated User Stories:") | |
| st.write(user_stories) | |
| else: | |
| st.warning("No user stories were generated. Please check the input and try again.") | |
| else: | |
| st.warning("Please enter both the bulk text and the prompt.") | |
| # Display the history of inputs and user stories | |
| if st.session_state['input_history']: | |
| st.subheader("History") | |
| for i, entry in enumerate(st.session_state['input_history']): | |
| st.write(f"**Input {i+1} (Text):** {entry['text']}") | |
| st.write(f"**Prompt {i+1}:** {entry['prompt']}") | |
| st.write(f"**User Stories {i+1}:** {entry['user_stories']}") | |
| st.write("---") | |
| # Instructions for using the app | |
| st.write("Enter your bulk text and a prompt for user story extraction, then click 'Generate User Stories' to get user stories from the text.") | |