app.py

import gradio as gr
from huggingface_hub import InferenceClient
from optimum.intel import OVModelForCausalLM
from transformers import AutoTokenizer, pipeline

# 載入模型和標記器
model_id = "hsuwill000/DeepSeek-R1-Distill-Qwen-1.5B-openvino"
model = OVModelForCausalLM.from_pretrained(model_id, device_map="auto")
tokenizer = AutoTokenizer.from_pretrained(model_id)

def respond(prompt, history):
    # 建立初始的系統訊息
    messages = [
        {"role": "system", "content": "You are Qwen, created by Alibaba Cloud. You are a helpful assistant."}
    ]

    # 將歷史對話內容整合進 messages（注意這裡轉換成字典）
    for msg in history:
        # 假設 history 中每個元素原本是字典格式 (如果不是，請自行轉換)
        messages.append({"role": "user", "content": msg["user"]})
        messages.append({"role": "assistant", "content": msg["assistant"]})

    # 加入當前用戶輸入
    messages.append({"role": "user", "content": prompt})

    # 構造模型輸入並生成回覆
    text = tokenizer.apply_chat_template(
        messages,
        tokenize=False,
        add_generation_prompt=True
    )
    model_inputs = tokenizer([text], return_tensors="pt").to(model.device)
    generated_ids = model.generate(
        **model_inputs,
        max_new_tokens=512
    )
    generated_ids = [
        output_ids[len(input_ids):] for input_ids, output_ids in zip(model_inputs.input_ids, generated_ids)
    ]
    response = tokenizer.batch_decode(generated_ids, skip_special_tokens=True)[0]

    print(f"Messages: {messages}")
    print(f"Reply: {response}")
    return response
    '''''
    # 更新並返回完整的聊天歷史（改為字典列表）
    new_history = history.copy()
    new_history.append({"user": prompt, "assistant": response})
    # 最後返回的是一個消息列表，每一條消息為字典格式（可進一步轉換為 ChatMessage 格式）
    final_messages = []
    # 如果需要顯示完整對話，可將歷史中每一條對話分拆成兩條消息
    for item in new_history:
        final_messages.append({"role": "user", "content": item["user"]})
        final_messages.append({"role": "assistant", "content": item["assistant"]})
    return final_messages
    '''''
   
# 設定 Gradio 的聊天界面
demo = gr.ChatInterface(
    fn=respond,
    title="DeepSeek-R1-Distill-Qwen-1.5B-openvino Chat",
    description="Chat with DeepSeek-R1-Distill-Qwen-1.5B-openvino model.",
    type="messages"
)

if __name__ == "__main__":
    demo.launch()