app.py

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

model_name = "fla-hub/rwkv7-2.9B-world"

print("Loading tokenizer...")
tokenizer = AutoTokenizer.from_pretrained(model_name, trust_remote_code=True)
print("Loading model...")
model = AutoModelForCausalLM.from_pretrained(
    model_name,
    trust_remote_code=True,
    torch_dtype=torch.float32,
    low_cpu_mem_usage=True,
    device_map="cpu"
)
print("Model loaded!")

def respond(message, history, system_message, max_tokens, temperature, top_p):
    messages = [{"role": "system", "content": system_message}]
    for human, assistant in history:
        messages.append({"role": "user", "content": human})
        messages.append({"role": "assistant", "content": assistant})
    messages.append({"role": "user", "content": message})
    
    text = tokenizer.apply_chat_template(messages, tokenize=False, add_generation_prompt=True)
    inputs = tokenizer(text, return_tensors="pt")
    
    with torch.no_grad():
        outputs = model.generate(
            **inputs,
            max_new_tokens=max_tokens,
            temperature=temperature,
            top_p=top_p,
            do_sample=True,
            pad_token_id=tokenizer.eos_token_id
        )
    
    response = tokenizer.decode(outputs[0][inputs.input_ids.shape[1]:], skip_special_tokens=True)
    return response

chatbot = gr.ChatInterface(
    respond,
    type="messages",
    additional_inputs=[
        gr.Textbox(value="You are a helpful assistant.", label="System message"),
        gr.Slider(1, 512, 256, step=1, label="Max tokens"),
        gr.Slider(0.1, 2.0, 0.7, step=0.1, label="Temperature"),
        gr.Slider(0.1, 1.0, 0.9, step=0.05, label="Top-p"),
    ],
)

chatbot.launch()