Update app.py
Browse files
app.py
CHANGED
|
@@ -1,25 +1,45 @@
|
|
| 1 |
import gradio as gr
|
| 2 |
-
import
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
| 3 |
|
| 4 |
-
|
| 5 |
-
|
| 6 |
-
|
| 7 |
-
|
| 8 |
-
|
| 9 |
-
|
| 10 |
-
|
| 11 |
-
|
| 12 |
-
|
| 13 |
-
|
| 14 |
-
|
| 15 |
-
|
| 16 |
-
term.open(document.getElementById('terminal'));
|
| 17 |
-
term.write('Hello from \x1B[1;3;31mxterm.js\x1B[0m $ ')
|
| 18 |
-
</script>
|
| 19 |
-
</body>
|
| 20 |
-
</html>
|
| 21 |
-
"""
|
| 22 |
-
return gr.HTML(html_code)
|
| 23 |
|
| 24 |
-
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
| 25 |
iface.launch()
|
|
|
|
| 1 |
import gradio as gr
|
| 2 |
+
import bs4
|
| 3 |
+
from langchain.text_splitter import RecursiveCharacterTextSplitter
|
| 4 |
+
from langchain_community.document_loaders import WebBaseLoader
|
| 5 |
+
from langchain_community.vectorstores import Chroma
|
| 6 |
+
from langchain_community.embeddings import OllamaEmbeddings
|
| 7 |
+
import ollama
|
| 8 |
|
| 9 |
+
# Function to load, split, and retrieve documents
|
| 10 |
+
def load_and_retrieve_docs(url):
|
| 11 |
+
loader = WebBaseLoader(
|
| 12 |
+
web_paths=(url,),
|
| 13 |
+
bs_kwargs=dict()
|
| 14 |
+
)
|
| 15 |
+
docs = loader.load()
|
| 16 |
+
text_splitter = RecursiveCharacterTextSplitter(chunk_size=1000, chunk_overlap=200)
|
| 17 |
+
splits = text_splitter.split_documents(docs)
|
| 18 |
+
embeddings = OllamaEmbeddings(model="mistral")
|
| 19 |
+
vectorstore = Chroma.from_documents(documents=splits, embedding=embeddings)
|
| 20 |
+
return vectorstore.as_retriever()
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
| 21 |
|
| 22 |
+
# Function to format documents
|
| 23 |
+
def format_docs(docs):
|
| 24 |
+
return "\n\n".join(doc.page_content for doc in docs)
|
| 25 |
+
|
| 26 |
+
# Function that defines the RAG chain
|
| 27 |
+
def rag_chain(url, question):
|
| 28 |
+
retriever = load_and_retrieve_docs(url)
|
| 29 |
+
retrieved_docs = retriever.invoke(question)
|
| 30 |
+
formatted_context = format_docs(retrieved_docs)
|
| 31 |
+
formatted_prompt = f"Question: {question}\n\nContext: {formatted_context}"
|
| 32 |
+
response = ollama.chat(model='mistral', messages=[{'role': 'user', 'content': formatted_prompt}])
|
| 33 |
+
return response['message']['content']
|
| 34 |
+
|
| 35 |
+
# Gradio interface
|
| 36 |
+
iface = gr.Interface(
|
| 37 |
+
fn=rag_chain,
|
| 38 |
+
inputs=["text", "text"],
|
| 39 |
+
outputs="text",
|
| 40 |
+
title="RAG Chain Question Answering",
|
| 41 |
+
description="Enter a URL and a query to get answers from the RAG chain."
|
| 42 |
+
)
|
| 43 |
+
|
| 44 |
+
# Launch the app
|
| 45 |
iface.launch()
|