Spaces:
Sleeping
Sleeping
File size: 2,088 Bytes
bdcf6b5 | 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 | import logging
from langchain.prompts import ChatPromptTemplate
from langchain.schema.output_parser import StrOutputParser
from langchain_community.chat_models import ChatOpenAI
from config import template
import gradio as gr
log = logging.getLogger(__name__)
log.setLevel(logging.DEBUG)
# Dictionary to store conversation history for each user
user_conversations = {}
def dispatch(username: str, input_text: str):
global user_conversations
# Initialize conversation history for new users
if username not in user_conversations:
user_conversations[username] = []
model_name = 'gpt-3.5-turbo-16k'
model = ChatOpenAI(model_name=model_name, verbose=True)
# Append user's input to the conversation history
user_conversations[username].append(f"{username}: {input_text}")
# Generate the full prompt including conversation history
full_prompt = "\n".join(user_conversations[username])
prompt = ChatPromptTemplate.from_template(template)
log.info("\n\n================ UNFILLED PROMPT:\n" + template)
chain = (
{
"question": lambda query: full_prompt,
}
| prompt
| model
| StrOutputParser()
)
result = chain.invoke(input_text)
# Append model's response to the conversation history without adding "Model:"
user_conversations[username].append(result)
# Safely format the conversation history for output as Markdown
chat_history_md = "\n\n".join([
f"**{msg.split(': ')[0]}**: {msg.split(': ', 1)[1]}" if ':' in msg else f"**{username}**: {msg}"
for msg in user_conversations[username]
])
return chat_history_md
# Create Gradio Interface with an additional input for username
iface = gr.Interface(
fn=dispatch,
inputs=[
gr.Textbox(placeholder="Enter your username", label="Username"), # Username input with label
gr.Textbox(placeholder="Enter your message", label="Message") # User input with label
],
outputs=gr.Markdown() # Use Markdown to display chat history
)
iface.launch(share=True)
|