Forecast SAM 3D slat velocities to skip backbone evaluations

SAM 3D Objects reconstructs full 3D shape, texture, and layout from a single masked image. This fork adds training-free HiCache++ caching to the slat-stage flow-matching sampler, forecasting velocity from cached anchors and calling the backbone only every interval steps.

# create sam3d-objects environment
mamba env create -f environments/default.yml
mamba activate sam3d-objects

# for pytorch/cuda dependencies
export PIP_EXTRA_INDEX_URL="https://pypi.ngc.nvidia.com https://download.pytorch.org/whl/cu121"

# install sam3d-objects and core dependencies
pip install -e '.[dev]'
pip install -e '.[p3d]' # pytorch3d dependency on pytorch is broken, this 2-step approach solves it

# for inference
export PIP_FIND_LINKS="https://nvidia-kaolin.s3.us-east-2.amazonaws.com/torch-2.5.1_cu121.html"
pip install -e '.[inference]'

# patch things that aren't yet in official pip packages
./patching/hydra # https://github.com/facebookresearch/hydra/pull/2863

# CPU contract tests — expect 4 passed
pytest tests/ -q
# accel self-test, no GPU or weights — expect ALL TESTS PASSED
python -m sam3d_objects.model.backbone.generator.flow_matching.accel