|
import gradio as gr |
|
from huggingface_hub import InferenceClient |
|
from unsloth.chat_templates import get_chat_template |
|
|
|
|
|
client = InferenceClient("wop/kosmox") |
|
|
|
|
|
tokenizer = get_chat_template( |
|
tokenizer=None, |
|
chat_template="phi-3", |
|
mapping={"role": "from", "content": "value", "user": "human", "assistant": "gpt"}, |
|
) |
|
|
|
def format_messages(system_message, history, user_message): |
|
|
|
formatted_message = "<s>\n" |
|
if system_message: |
|
formatted_message += f"{system_message}\n" |
|
|
|
for user_msg, assistant_msg in history: |
|
if user_msg: |
|
formatted_message += f"{user_msg}\n" |
|
if assistant_msg: |
|
formatted_message += f"{assistant_msg}\n" |
|
|
|
formatted_message += f"{user_message}\n" |
|
return formatted_message |
|
|
|
def respond( |
|
message: str, |
|
history: list[tuple[str, str]], |
|
system_message: str, |
|
max_tokens: int, |
|
temperature: float, |
|
top_p: float, |
|
): |
|
|
|
formatted_message = format_messages(system_message, history, message) |
|
|
|
response = "" |
|
|
|
|
|
for message in client.chat_completion( |
|
formatted_message, |
|
max_tokens=max_tokens, |
|
stream=True, |
|
temperature=temperature, |
|
top_p=top_p, |
|
): |
|
token = message.choices[0].delta.content |
|
response += token |
|
yield response |
|
|
|
|
|
demo = gr.ChatInterface( |
|
fn=respond, |
|
additional_inputs=[ |
|
gr.Textbox(value="You are AI.", label="System message"), |
|
gr.Slider(minimum=1, maximum=2048, value=512, step=1, label="Max new tokens"), |
|
gr.Slider(minimum=0.1, maximum=4.0, value=0.7, step=0.1, label="Temperature"), |
|
gr.Slider(minimum=0.1, maximum=1.0, value=0.95, step=0.05, label="Top-p (nucleus sampling)"), |
|
], |
|
) |
|
|
|
if __name__ == "__main__": |
|
demo.launch() |
|
|