app.py

# import gradio as gr
# from huggingface_hub import InferenceClient

# """
# For more information on `huggingface_hub` Inference API support, please check the docs: https://huggingface.co/docs/huggingface_hub/v0.22.2/en/guides/inference
# """
# client = InferenceClient("HuggingFaceH4/zephyr-7b-beta")


# def respond(
#     message,
#     history: list[tuple[str, str]],
#     system_message,
#     max_tokens,
#     temperature,
#     top_p,
# ):
#     messages = [{"role": "system", "content": system_message}]

#     for val in history:
#         if val[0]:
#             messages.append({"role": "user", "content": val[0]})
#         if val[1]:
#             messages.append({"role": "assistant", "content": val[1]})

#     messages.append({"role": "user", "content": message})

#     response = ""

#     for message in client.chat_completion(
#         messages,
#         max_tokens=max_tokens,
#         stream=True,
#         temperature=temperature,
#         top_p=top_p,
#     ):
#         token = message.choices[0].delta.content

#         response += token
#         yield response


# """
# For information on how to customize the ChatInterface, peruse the gradio docs: https://www.gradio.app/docs/chatinterface
# """
# 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)",
#         ),
#     ],
# )


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

import gradio as gr
from PIL import Image

# —— 在这里根据你自己的 Agent 框架实现这个函数 ——  
def run_agent_on_image(original_img: Image.Image, prompt: str, agent_name: str) -> Image.Image:
    """
    调用指定 agent（模型/工具）处理图片并返回结果
    original_img: PIL Image
    prompt: 用户输入的描述
    agent_name: 在下拉框里选的模型名称
    """
    # 示例逻辑（请替换为真正的 agent 调用）
    # if agent_name == "Model A":
    #     return model_a.process(original_img, prompt)
    # elif agent_name == "Model B":
    #     return model_b.process(original_img, prompt)
    return original_img  # TODO: 删除这行

# 可选：把可用的 agent 列表写成一个文件或者直接在这里列出
MODEL_CHOICES = ["Model A", "Model B", "Model C"]


with gr.Blocks() as demo:
    gr.Markdown("## CV Agent Arena  🎨🤖\nUpload the image you want to process, provide your requirements, select two Agents, and click 'Run Agents' to compare the results!")

    with gr.Row():
        with gr.Column():
            original = gr.Image(type="pil", label="Upload Original Image")
            prompt = gr.Textbox(lines=2, placeholder="e.g. ‘Make it look like a sunny day’", label="Prompt")
        with gr.Column():
            agent1 = gr.Dropdown(choices=MODEL_CHOICES, label="Select Agent 1")
            agent2 = gr.Dropdown(choices=MODEL_CHOICES, label="Select Agent 2")

    # 处理按钮
    run_btn = gr.Button("Run Agents")    

    with gr.Row():
        # 左侧输出：Agent1 结果
        with gr.Column():
            out1 = gr.Image(type="pil", label="Agent 1 Output")
        # 右侧输出：Agent2 结果
        with gr.Column():
            out2 = gr.Image(type="pil", label="Agent 2 Output")

    # 按钮绑定
    run_btn.click(
        fn=lambda img, p, a1, a2: (run_agent_on_image(img, p, a1), run_agent_on_image(img, p, a2)),
        inputs=[original, prompt, agent1, agent2],
        outputs=[out1, out2],
    )

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