import subprocess subprocess.run('pip install flash-attn --no-build-isolation', env={'FLASH_ATTENTION_SKIP_CUDA_BUILD': "TRUE"}, shell=True) import gradio as gr from PIL import Image from transformers import AutoModelForCausalLM, AutoProcessor import torch device = torch.device("cuda" if torch.cuda.is_available() else "cpu") model_name = "Chan-Y/Florence-2-LaTex-v2" model = AutoModelForCausalLM.from_pretrained(model_name, trust_remote_code=True).to(device) processor = AutoProcessor.from_pretrained(model_name, trust_remote_code=True) def process_image(image): if image: inputs = processor(images=image, return_tensors="pt").to(device) outputs = model.generate(**inputs) text = processor.decode(outputs[0], skip_special_tokens=True) return text return "No image provided." with gr.Blocks() as demo: with gr.Row(): image_input = gr.Image(label="Upload an Image", type="pil") process_button = gr.Button("Process Image") output = gr.Textbox(label="Output Message") process_button.click(process_image, inputs=image_input, outputs=output) demo.launch()