app.py

import gradio as gr
from transformers import AutoModel, AutoTokenizer
import torch

# Load the model and tokenizer
model_name = "wop/kosmox-gguf"
tokenizer = AutoTokenizer.from_pretrained(model_name)
model = AutoModel.from_pretrained(model_name)

# Function to generate responses
def respond(message, history, system_message, max_tokens, temperature, top_p):
    # Prepare the chat history
    messages = [{"role": "system", "content": system_message}]
    
    for user_msg, bot_msg in history:
        if user_msg:
            messages.append({"role": "user", "content": user_msg})
        if bot_msg:
            messages.append({"role": "assistant", "content": bot_msg})
    
    messages.append({"role": "user", "content": message})

    # Create the chat input for the model
    chat_input = tokenizer.chat_template.format(
        bos_token=tokenizer.bos_token,
        messages=[{"from": "human", "value": m['content']} if m['role'] == 'user' else {"from": "gpt", "value": m['content']} for m in messages]
    )
    
    inputs = tokenizer(chat_input, return_tensors="pt")

    # Generate response
    with torch.no_grad():
        outputs = model.generate(
            input_ids=inputs['input_ids'],
            max_length=max_tokens,
            temperature=temperature,
            top_p=top_p,
            do_sample=True
        )
    
    response = tokenizer.decode(outputs[0], skip_special_tokens=True)
    yield response.strip()

# Define the Gradio interface
demo = gr.ChatInterface(
    respond,
    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)"),
    ],
)

# Launch the demo
if __name__ == "__main__":
    demo.launch()