Spaces:
Runtime error
Runtime error
File size: 3,760 Bytes
e7127be 5dec041 e7127be 5dec041 e7127be 5dec041 e7127be 5dec041 |
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 59 60 61 62 63 64 65 66 67 68 69 70 71 72 73 74 75 76 77 78 79 80 81 82 83 84 85 86 87 88 89 90 91 92 93 94 95 96 97 98 99 100 101 102 103 104 105 106 107 108 109 110 111 112 113 114 115 116 117 118 119 120 121 122 |
import os, dotenv, anthropic, panel, platform
from langchain_community.vectorstores import Chroma
from langchain.embeddings import HuggingFaceEmbeddings
from langchain.text_splitter import RecursiveCharacterTextSplitter
from langchain_community.document_loaders import DirectoryLoader, PyPDFLoader
panel.extension()
# Set API key
dotenv.load_dotenv()
ANTHROPIC_API_KEY = os.getenv('ANTHROPIC_API_KEY')
@panel.cache
def load_vectorstore():
if "macOS" in platform.platform():
device="mps"
else:
device="cpu"
# Create the HF embeddings
model_name = "sentence-transformers/all-mpnet-base-v2"
model_kwargs = {'device': device}
encode_kwargs = {'normalize_embeddings': False}
hf_embeddings = HuggingFaceEmbeddings(
model_name=model_name,
model_kwargs=model_kwargs,
encode_kwargs=encode_kwargs
)
# If the vector embeddings of the documents have not been created
if not os.path.isfile('chroma_db/chroma.sqlite3'):
# Load the documents
loader = DirectoryLoader('Docs/', glob="./*.pdf", loader_cls=PyPDFLoader)
data = loader.load()
# Split the docs into chunks
splitter = RecursiveCharacterTextSplitter(
chunk_size=1000,
chunk_overlap=50
)
docs = splitter.split_documents(data)
# Embed the documents and store them in a Chroma DB
vectorstore = Chroma.from_documents(documents=docs,embedding=hf_embeddings, persist_directory="./chroma_db")
else:
# load ChromaDB from disk
vectorstore = Chroma(persist_directory="./chroma_db", embedding_function=hf_embeddings)
return vectorstore
# Initialize the chat history
chat_history = []
async def get_response(contents, user, instance):
# Load the vectorstore
vectorstore = load_vectorstore()
question = contents
# Get the relevant information to form the context on which to query with the LLM
docs = vectorstore.similarity_search(question)
context = "\n"
for doc in docs:
context += "\n" + doc.page_content + "\n"
# Update the global chat_history with the user's question
global chat_history
chat_history.append({"role": "user", "content": question})
# Define prompt template
prompt = f"""
Here are the Task Context and History
- Context: {context}
- Chat History: {chat_history}
- User Question: {question}
"""
# Create the Anthropic client
client = anthropic.Anthropic(api_key=ANTHROPIC_API_KEY)
response = ''
# Generate the completion with the updated chat_history
with client.messages.stream(
max_tokens=1024,
messages=[
{"role": "user", "content": prompt}
],
model="claude-3-haiku-20240307",
) as stream:
for text in stream.text_stream:
response += text
yield response
# Append the assistant's response to the chat_history
chat_history.append({"role": "assistant", "content": response})
chat_interface = panel.chat.ChatInterface(
callback=get_response,
callback_user="Sarathi",
sizing_mode="stretch_width",
callback_exception='verbose',
message_params=dict(
default_avatars={"Sarathi": "S", "User": "U"},
reaction_icons={"like": "thumb-up"},
),
)
chat_interface.send(
{"user": "Sarathi", "value": '''Welcome to Sarathi, your personal assistant for Assam Tourism.'''},
respond=False,
)
template = panel.template.BootstrapTemplate(title="Sarathi", favicon="favicon.png", header_background = "#000000", main=[panel.Tabs( ('Chat', chat_interface), dynamic=True )])
template.servable()
|