import os import streamlit as st from openai import OpenAI from dotenv import load_dotenv from langchain_core.prompts import PromptTemplate # Load environment variables load_dotenv() ##openai_api_key = os.getenv("OPENAI_API_KEY") # Initialize the client client = OpenAI( base_url="https://api-inference.huggingface.co/v1", api_key=os.environ.get('TOKEN2') # Add your Huggingface token here ) # Initialize the OpenAI client ##client = OpenAI( ##base_url="https://api-inference.huggingface.co/v1", ##api_key=openai_api_key ##) # Define reset function for the conversation def reset_conversation(): st.session_state.conversation = [] st.session_state.messages = [] # Streamlit interface setup st.title("🤖 Text Data Generation & Labeling App") st.sidebar.title("Settings") # Sidebar settings selected_model = st.sidebar.selectbox("Select Model", ["meta-llama/Meta-Llama-3-8B-Instruct"]) temperature = st.sidebar.slider("Temperature", 0.0, 1.0, 0.5) st.sidebar.button("Reset Conversation", on_click=reset_conversation) st.sidebar.write(f"You're now chatting with **{selected_model}**") st.sidebar.markdown("*Note: Generated content may be inaccurate or false.*") # Initialize conversation state if "messages" not in st.session_state: st.session_state.messages = [] # Display conversation for message in st.session_state.messages: with st.chat_message(message["role"]): st.markdown(message["content"]) # Main logic: choose between Data Generation and Data Labeling task_choice = st.selectbox("Choose Task", ["Data Generation", "Data Labeling"]) if task_choice == "Data Generation": classification_type = st.selectbox( "Choose Classification Type", ["Sentiment Analysis", "Binary Classification", "Multi-Class Classification"] ) if classification_type == "Sentiment Analysis": labels = ["Positive", "Negative", "Neutral"] elif classification_type == "Binary Classification": label_1 = st.text_input("Enter first class") label_2 = st.text_input("Enter second class") labels = [label_1, label_2] else: # Multi-Class Classification num_classes = st.slider("How many classes?", 3, 10, 3) labels = [st.text_input(f"Class {i+1}") for i in range(num_classes)] domain = st.selectbox("Choose Domain", ["Restaurant reviews", "E-commerce reviews", "Custom"]) if domain == "Custom": domain = st.text_input("Specify custom domain") min_words = st.number_input("Minimum words per example", min_value=10, max_value=90, value=10) max_words = st.number_input("Maximum words per example", min_value=10, max_value=90, value=90) use_few_shot = st.radio("Use few-shot examples?", ["Yes", "No"]) few_shot_examples = [] if use_few_shot == "Yes": num_examples = st.slider("Number of few-shot examples", 1, 5, 1) for i in range(num_examples): content = st.text_area(f"Example {i+1} Content") label = st.selectbox(f"Example {i+1} Label", labels) few_shot_examples.append({"content": content, "label": label}) num_to_generate = st.number_input("Number of examples to generate", 1, 100, 10) user_prompt = st.text_area("Enter additional instructions", "") # Construct the LangChain prompt prompt_template = PromptTemplate( input_variables=["classification_type", "domain", "num_examples", "min_words", "max_words", "labels", "user_prompt"], template=( "You are a professional {classification_type} expert tasked with generating examples for {domain}.\n" "Use the following parameters:\n" "- Number of examples: {num_examples}\n" "- Word range: {min_words}-{max_words}\n" "- Labels: {labels}\n" "{user_prompt}" ) ) system_prompt = prompt_template.format( classification_type=classification_type, domain=domain, num_examples=num_to_generate, min_words=min_words, max_words=max_words, labels=", ".join(labels), user_prompt=user_prompt ) st.write("System Prompt:") st.code(system_prompt) if st.button("Generate Examples"): with st.spinner("Generating..."): st.session_state.messages.append({"role": "system", "content": system_prompt}) try: stream = client.chat.completions.create( model=selected_model, messages=[{"role": "system", "content": system_prompt}], temperature=temperature, stream=True, max_tokens=3000, ) response = st.write_stream(stream) st.session_state.messages.append({"role": "assistant", "content": response}) except Exception as e: st.error("An error occurred during generation.") st.error(f"Details: {e}") elif task_choice == "Data Labeling": # Labeling logic labeling_type = st.selectbox( "Classification Type for Labeling", ["Sentiment Analysis", "Binary Classification", "Multi-Class Classification"] ) if labeling_type == "Sentiment Analysis": labels = ["Positive", "Negative", "Neutral"] elif labeling_type == "Binary Classification": label_1 = st.text_input("First label for classification") label_2 = st.text_input("Second label for classification") labels = [label_1, label_2] else: # Multi-Class Classification num_classes = st.slider("Number of labels", 3, 10, 3) labels = [st.text_input(f"Label {i+1}") for i in range(num_classes)] use_few_shot_labeling = st.radio("Add few-shot examples for labeling?", ["Yes", "No"]) few_shot_labeling_examples = [] if use_few_shot_labeling == "Yes": num_labeling_examples = st.slider("Number of few-shot labeling examples", 1, 5, 1) for i in range(num_labeling_examples): content = st.text_area(f"Labeling Example {i+1} Content") label = st.selectbox(f"Label for Example {i+1}", labels) few_shot_labeling_examples.append({"content": content, "label": label}) text_to_classify = st.text_area("Enter text to classify") if st.button("Classify Text"): if text_to_classify: labeling_prompt = ( f"You are an expert in {labeling_type.lower()} classification. Classify this text using: {', '.join(labels)}.\n\n" ) if few_shot_labeling_examples: labeling_prompt += "Example classifications:\n" for ex in few_shot_labeling_examples: labeling_prompt += f"Text: {ex['content']} - Label: {ex['label']}\n" labeling_prompt += f"\nClassify this: {text_to_classify}" with st.spinner("Classifying..."): st.session_state.messages.append({"role": "system", "content": labeling_prompt}) try: stream = client.chat.completions.create( model=selected_model, messages=[{"role": "system", "content": labeling_prompt}], temperature=temperature, stream=True, max_tokens=3000, ) labeling_response = st.write_stream(stream) st.write("Label:", labeling_response) except Exception as e: st.error("An error occurred during classification.") st.error(f"Details: {e}") else: st.warning("Please enter text to classify.")