Fix: Exact import test format, TF32 set directly (no try/except), verify allocator config includes max_split_size_mb=128

Fix: PyTorch 2.9.0 cu128, simplified setup.sh with fatal errors, TF32=tf32, cuDNN always on, TORCH_CUDA_ARCH_LIST=8.9

Fixed

Production build strategy: Docker support, pinned dependencies, OOM guardrails, optimized defaults

Speed optimizations: GPU aggregation, max-autotune compile, increased batch/ROI, GPU utilization monitoring, quality guardrails

Fix Gradio version and OMP_NUM_THREADS: Upgrade Gradio to >=4.44.0 in setup.sh, add runtime version check, and fix OMP_NUM_THREADS warning

Add setup script for mamba-ssm build and temporarily comment out mamba-ssm from requirements