llama.cpp/examples/gradio_chat/server.py
2024-02-25 21:09:13 -05:00

56 lines
1.2 KiB
Python

import gradio as gr
from openai import OpenAI
client = OpenAI(
base_url="http://localhost:8000/v1",
api_key="llama.cpp"
)
model = "gpt-3.5-turbo"
def predict(message, history):
messages = []
for user_message, assistant_message in history:
messages.append({"role": "user", "content": user_message})
messages.append({"role": "assistant", "content": assistant_message})
messages.append({"role": "user", "content": message})
response = client.chat.completions.create(
model=model,
messages=messages,
stream=True
)
text = ""
for chunk in response:
content = chunk.choices[0].delta.content
if content:
text += content
yield text
js = """function () {
gradioURL = window.location.href
if (!gradioURL.endsWith('?__theme=dark')) {
window.location.replace(gradioURL + '?__theme=dark');
}
}"""
css = """
footer {
visibility: hidden;
}
full-height {
height: 100%;
}
"""
with gr.Blocks(theme=gr.themes.Soft(), js=js, css=css, fill_height=True) as demo:
gr.ChatInterface(predict, fill_height=True, examples=["What is the capital of France?", "Who was the first person on the moon?"])
if __name__ == "__main__":
demo.launch()