Spaces:
Paused
Paused
| # app.py Made with https://github.co/TeamTonic/Vectonic | |
| import os | |
| import gradio as gr | |
| import requests | |
| from vectara_cli.core import VectaraClient | |
| # Load environment variables | |
| from dotenv import load_dotenv | |
| load_dotenv() | |
| # Configuration variables | |
| VECTARA_CUSTOMER_ID = os.getenv("VECTARA_CUSTOMER_ID") | |
| VECTARA_API_KEY = os.getenv("VECTARA_API_KEY") | |
| VECTARA_CORPUS_ID = os.getenv("VECTARA_CORPUS_ID") | |
| TOGETHER_API_TOKEN = os.getenv("TOGETHER_API_TOKEN") | |
| DEFAULT_SYSTEM_PROMPT = os.getenv("SYSTEM_PROMPT") | |
| MODEL_NAME = os.getenv("MODEL_NAME", "databricks/dbrx-instruct") | |
| TITLE = os.getenv("TITLE") | |
| DESCRIPTION = os.getenv("DESCRIPTION") | |
| class VectaraClientExtended(VectaraClient): | |
| """ Enhanced Vectara Client with methods tailored for chatbot application. """ | |
| def __init__(self, customer_id, api_key): | |
| super().__init__(customer_id, api_key) | |
| def retrieve_context(self, query, corpus_id): | |
| response = self.advanced_query(query, 1, corpus_id, {}, {}) | |
| if response.get('matches'): | |
| return response['matches'][0]['document']['text'] | |
| return None | |
| class TogetherAIInterface: | |
| """ Integration for querying TogetherAI LLM. """ | |
| def __init__(self, token): | |
| self.headers = {"Authorization": f"Bearer {token}"} | |
| def generate_response(self, prompt, model="curie", max_tokens=150, temperature=0.7): | |
| url = 'https://api.together.xyz/v1/completions' | |
| data = { | |
| "model": model, | |
| "prompt": prompt, | |
| "max_tokens": max_tokens, | |
| "temperature": temperature | |
| } | |
| response = requests.post(url, headers=self.headers, json=data) | |
| return response.json().get('choices')[0]['text'] | |
| # Application logic | |
| def chatbot_response(user_query): | |
| vectara_client = VectaraClientExtended(VECTARA_CUSTOMER_ID, VECTARA_API_KEY) | |
| together_ai = TogetherAIInterface(TOGETHER_API_TOKEN) | |
| # Retrieve context from Vectara based on the user's query | |
| context = vectara_client.retrieve_context(user_query, VECTARA_CORPUS_ID) | |
| if not context: | |
| context = "No relevant context was found for your query." | |
| # Prepare the prompt | |
| prompt = f"System Prompt: {DEFAULT_SYSTEM_PROMPT}\nContext: {context}\nQuery: {user_query}" | |
| # Get response from TogetherAI | |
| model_response = together_ai.generate_response(prompt) | |
| return model_response | |
| # Gradio interface setup | |
| def main(): | |
| iface = gr.Interface( | |
| chatbot_response, | |
| inputs=gr.Textbox(lines=2, placeholder="Enter your query here..."), | |
| outputs=gr.Textbox(), | |
| title="Vectonic ChatBot", | |
| description="Optimized Using Tonicai. Powered by Vectara and TogetherAI. Ask anything!", | |
| ) | |
| iface.launch() | |
| if __name__ == "__main__": | |
| main() |