File size: 1,974 Bytes
c33d1d0 018fb30 e5633a7 c33d1d0 018fb30 f7493dd 037c950 66fc16c e5633a7 c33d1d0 1f5e9cb 037c950 c33d1d0 037c950 018fb30 38e2fac cbed288 18cb8f3 142d17f ea07eae c33d1d0 142d17f 68b31c9 018fb30 037c950 68b31c9 018fb30 037c950 018fb30 c33d1d0 |
1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 16 17 18 19 20 21 22 23 24 25 26 27 28 29 30 31 32 33 34 35 36 37 38 39 40 41 42 43 44 45 46 47 48 49 50 51 52 53 54 55 56 57 58 |
import os
import gradio as gr
from dotenv import load_dotenv
from langchain.vectorstores.faiss import FAISS # Import FAISS
from langchain.vectorstores.chroma import Chroma # Import Chroma
from langchain.document_loaders import PyPDFLoader
from langchain.text_splitter import CharacterTextSplitter
from langchain.embeddings import HuggingFaceInferenceAPIEmbeddings
from langchain.embeddings import HuggingFaceBgeEmbeddings
# Load environment variables
load_dotenv()
# Use Hugging Face Inference API embeddings
inference_api_key = os.getenv('HF') # Use getenv to retrieve environment variable
api_hf_embeddings = HuggingFaceInferenceAPIEmbeddings(
api_key=inference_api_key,
model_name="sentence-transformers/all-MiniLM-l6-v2"
)
# Load and process the PDF files
loader = PyPDFLoader("./new_papers/ALiBi.pdf")
documents = loader.load()
# Split the documents into chunks and embed them using the HfApiEmbeddingTool
text_splitter = CharacterTextSplitter(chunk_size=100, chunk_overlap=0)
vdocuments = text_splitter.split_documents(documents)
model = "BAAI/bge-base-en-v1.5"
encode_kwargs = {
"normalize_embeddings": True
} # set True to compute cosine similarity
embeddings = HuggingFaceBgeEmbeddings(
model_name=model, encode_kwargs=encode_kwargs, model_kwargs={"device": "cpu"}
)
# Create FAISS vector store for API embeddings
api_db = FAISS.from_texts(texts=vdocuments, embedding=embeddings)
# Define the PDF retrieval function
def pdf_retrieval(query):
# Run the query through the retriever
response = api_db.similarity_search(query)
return response
# Create Gradio interface for the API retriever
api_tool = gr.Interface(
fn=pdf_retrieval,
inputs=[gr.Textbox()],
outputs=gr.Textbox(),
live=True,
title="API PDF Retrieval Tool",
description="This tool indexes PDF documents and retrieves relevant answers based on a given query (HF Inference API Embeddings).",
)
# Launch the Gradio interface
api_tool.launch()
|