69 lines
No EOL
2.2 KiB
Python
69 lines
No EOL
2.2 KiB
Python
import argparse
|
|
|
|
import gradio as gr
|
|
from openai import OpenAI
|
|
|
|
# Argument parser setup
|
|
parser = argparse.ArgumentParser(
|
|
description='Chatbot Interface with Customizable Parameters')
|
|
parser.add_argument('--model-url',
|
|
type=str,
|
|
default='http://localhost:8000/v1',
|
|
help='Model URL')
|
|
parser.add_argument('-m',
|
|
'--model',
|
|
type=str,
|
|
required=True,
|
|
help='Model name for the chatbot')
|
|
parser.add_argument("--host", type=str, default=None)
|
|
parser.add_argument("--port", type=int, default=8001)
|
|
|
|
# Parse the arguments
|
|
args = parser.parse_args()
|
|
|
|
# Set OpenAI's API key and API base to use vLLM's API server.
|
|
openai_api_key = "EMPTY"
|
|
openai_api_base = args.model_url
|
|
|
|
# Create an OpenAI client to interact with the API server
|
|
client = OpenAI(
|
|
api_key=openai_api_key,
|
|
base_url=openai_api_base,
|
|
)
|
|
|
|
|
|
def predict(message, history):
|
|
# Convert chat history to OpenAI format
|
|
history_openai_format = [{
|
|
"role": "system",
|
|
"content": "You are a great ai assistant."
|
|
}]
|
|
for human, assistant in history:
|
|
history_openai_format.append({"role": "user", "content": human})
|
|
history_openai_format.append({
|
|
"role": "assistant",
|
|
"content": assistant
|
|
})
|
|
history_openai_format.append({"role": "user", "content": message})
|
|
|
|
# Create a chat completion request and send it to the API server
|
|
stream = client.chat.completions.create(
|
|
model=args.model, # Model name to use
|
|
messages=history_openai_format, # Chat history
|
|
stream=True, # Stream response
|
|
)
|
|
|
|
# Read and return generated text from response stream
|
|
partial_message = ""
|
|
for chunk in stream:
|
|
# import pdb
|
|
# pdb.set_trace()
|
|
# partial_message += (chunk.delta['content'] or "")
|
|
partial_message += (chunk.choices[0].delta.content or "")
|
|
yield partial_message
|
|
|
|
|
|
# Create and launch a chat interface with Gradio
|
|
gr.ChatInterface(predict).queue().launch(server_name=args.host,
|
|
server_port=args.port,
|
|
share=True) |