Use PyTorch cu124 to match system CUDA 12.4

Install CPU-only PyTorch to fix CUDA symbol conflict

Fix cuDNN version: use cuda:12.4.1-cudnn for cuDNN 9

CUDA base image, pre-cache models, fix double loading

Add deploy documentation

Fix OMP_NUM_THREADS at Python level before rembg import

Enable CUDA GPU providers for T4

Fix OMP_NUM_THREADS and increase health check timeout

Deploy Django background removal API

initial commit