| __import__('pysqlite3') | |
| import sys | |
| sys.modules['sqlite3'] = sys.modules.pop('pysqlite3') | |
| import os | |
| import gradio as gr | |
| import chromadb | |
| from sentence_transformers import SentenceTransformer | |
| import pandas as pd | |
| import numpy as np | |
| from chromadb.utils import embedding_functions | |
| from huggingface_hub import InferenceClient | |
| from dotenv import load_dotenv, find_dotenv | |
| _ = load_dotenv(find_dotenv()) | |
| hf_api_key = os.environ['HF_API_KEY'] | |
| dfs = pd.read_csv('Patents.csv') | |
| ids= [str(x) for x in dfs.index.tolist()] | |
| docs = dfs['text'].tolist() | |
| client = chromadb.Client() | |
| collection = client.get_or_create_collection("patents") | |
| collection.add(documents=docs,ids=ids) | |
| def text_embedding(text)-> None: | |
| model = SentenceTransformer('sentence-transformers/all-MiniLM-L6-v2') | |
| return model.encode(text) | |
| def gen_context(query): | |
| vector = text_embedding(query).tolist() | |
| results = collection.query( | |
| query_embeddings=vector,n_results=15,include=["documents"]) | |
| res = "\n".join(str(item) for item in results['documents'][0]) | |
| return res | |
| def chat_completion(user_prompt): | |
| length = 1000 | |
| system_prompt = """\ | |
| You are a helpful AI assistant that can answer questions on the patents dataset. Answer based on the context provided.If you cannot find the correct answer, say I don't know. Be concise and just include the response""" | |
| final_prompt = f"""<s>[INST]<<SYS>> | |
| {system_prompt} | |
| <</SYS>> | |
| {user_prompt}[/INST]""" | |
| return client.text_generation(prompt=final_prompt,max_new_tokens = length).strip() | |
| client = InferenceClient(model = "mistralai/Mixtral-8x7B-Instruct-v0.1") | |
| def chat_completion(query): | |
| length = 1000 | |
| context = gen_context(query) | |
| user_prompt = f""" | |
| Based on the context: | |
| {context} | |
| Answer the below query: | |
| {query} | |
| """ | |
| system_prompt = """\ | |
| You are a helpful AI assistant that can answer questions on the patents dataset. Answer based on the context provided.If you cannot find the correct answer, say I don't know. Be concise and just include the response""" | |
| final_prompt = f"""<s>[INST]<<SYS>> | |
| {system_prompt} | |
| <</SYS>> | |
| {user_prompt}[/INST]""" | |
| return client.text_generation(prompt=final_prompt,max_new_tokens = length).strip() | |
| demo = gr.Interface(fn=chat_completion, | |
| inputs=[gr.Textbox(label="Query", lines=2)], | |
| outputs=[gr.Textbox(label="Result", lines=16)], | |
| title="Chat on Patents Data") | |
| demo.queue().launch(share=True) |