Test / app.py
JPraj's picture
Update app.py
9eae408 verified
Raw History Blame Contribute Delete
1.84 kB
import gradio as gr
from huggingface_hub import InferenceClient
def respond(
user_message,
history: list[dict],
system_message,
max_tokens,
temperature,
top_p,
hf_token: gr.OAuthToken,
):
# create the HF client (make sure hf_token is valid)
client = InferenceClient(token=hf_token.token, model="openai/gpt-oss-20b")
messages = [{"role": "system", "content": system_message}]
# history (if using type="messages", history should be list of dicts)
messages.extend(history or [])
messages.append({"role": "user", "content": user_message})
response = ""
# iterate streamed chunks from HF
for event in client.chat_completion(
messages,
max_tokens=max_tokens,
stream=True,
temperature=temperature,
top_p=top_p,
):
# event contains choices; don't shadow `message`
choices = getattr(event, "choices", None) or event.get("choices", [])
if choices and getattr(choices[0], "delta", None):
token = choices[0].delta.content or ""
else:
token = ""
response += token
# yield the full response string (ChatInterface expects that)
yield response
# build ChatInterface
chatbot = gr.ChatInterface(
respond,
type="messages",
additional_inputs=[
gr.Textbox(value="You are a friendly Chatbot.", 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)"),
],
)
with gr.Blocks() as demo:
with gr.Sidebar():
gr.LoginButton()
chatbot.render()
if __name__ == "__main__":
demo.launch()