import os
import gradio as gr
from huggingface_hub import InferenceClient

# שליפת המפתח ממשתנה הסביבה
api_key = os.getenv("HF_API_KEY")
client = InferenceClient(api_key=api_key)

def chat_with_model(prompt):
    response_text = ""
    for message in client.chat_completion(
        model="google/gemma-2-2b-it",
        messages=[{"role": "user", "content": prompt}],
        max_tokens=250,
        stream=True,
    ):
        response_text += message.choices[0].delta.content
    return response_text

interface = gr.Interface(fn=chat_with_model, inputs="text", outputs="text", title="Chat with Hugging Face Model")
interface.launch()