from lmdeploy import pipeline, GenerationConfig, TurbomindEngineConfig from lmdeploy.vl import load_image import spaces import gradio as gr from PIL import Image import numpy as np @spaces.GPU def create_captions_llava_llama3_docci(image): pipe = pipeline('Lin-Chen/open-llava-next-llama3-8b') gen_config = GenerationConfig(repetition_penalty=1.10) image = Image.fromarray(np.uint8(image)).convert('RGB') response = pipe(('As an AI image annotation expert, please provide accurate annotations for images to enhance model understanding of the content.These prompts will be used for image reconstruction, so the closer the similarity to the original image, the better the prompts quality.Special prompts will receive a reward of $10 per image.', image), gen_config=gen_config) return response.text css = """ #mkd { height: 500px; overflow: auto; border: 1px solid #ccc; } """ with gr.Blocks(css=css) as demo: gr.HTML("