import os import gradio as gr from llama_cpp import Llama from huggingface_hub import hf_hub_download model = Llama( model_path=hf_hub_download( repo_id=os.environ.get("REPO_ID", "bartowski/QwQ-32B-Preview-GGUF"), filename=os.environ.get("MODEL_FILE", "QwQ-32B-Preview-Q3_K_L.gguf"), ) ) DESCRIPTION = ''' # QwQ-32B-Preview | Duplicate the space and set it to private for faster & personal inference for free. Qwen/QwQ-32B-Preview: an experimental research model developed by the Qwen Team. Focused on advancing AI reasoning capabilities. **To start a new chat**, click "clear" and start a new dialog. ''' LICENSE = """ --- Apache 2.0 License --- """ def generate_text(message, history, max_tokens=512, temperature=0.9, top_p=0.95): """Generate a response using the Llama model.""" temp = "" response = model.create_chat_completion( messages=[{"role": "system", "content": "You are a helpful and harmless assistant. You are Qwen developed by Alibaba. You should think step-by-step."}, {"role": "user", "content": message}], temperature=temperature, max_tokens=max_tokens, top_p=top_p, stream=True, ) for streamed in response: delta = streamed["choices"][0].get("delta", {}) text_chunk = delta.get("content", "") temp += text_chunk yield temp with gr.Blocks() as demo: gr.Markdown(DESCRIPTION) chatbot = gr.ChatInterface( generate_text, title="Qwen/QwQ-32B-Preview | GGUF Demo", description=" settings below if needed.", examples=[ ["How many r's are in the word strawberry?"], ['What is the most optimal way to do Test-Time Scaling?'], ['Explain to me how gravity works like I am 5!'], ], cache_examples=False, fill_height=True ) with gr.Accordion("Adjust Parameters", open=False): gr.Slider(minimum=512, maximum=4096, value=1024, step=1, label="Max Tokens") gr.Slider(minimum=0.1, maximum=1.5, value=0.9, step=0.1, label="Temperature") gr.Slider(minimum=0.1, maximum=1.0, value=0.95, step=0.05, label="Top-p (nucleus sampling)") gr.Markdown(LICENSE) if __name__ == "__main__": demo.launch()