Skip to content

DeepLens v2.5.4

Choose a tag to compare

View all tags
@singer-yang singer-yang released this 12 Sep 22:54
v2.5.4
278e4f7

DeepLens is a differentiable optical lens simulator for computational imaging and optical design.

This patch release:

  • Standardizes the environment on Python 3.12, PyTorch 2.10.0, and torchvision 0.25.0. Linux and Windows requirements select the exact CUDA 12.8 builds; Apple-silicon macOS uses native wheels.
  • Simplifies environment.yml to create Python 3.12 and install the shared requirements.txt, replacing a machine-specific environment export.
  • Aligns package metadata and installation documentation with the pinned runtime.
  • Fixes CPU/CUDA device placement in the far-ray regression tests so the full suite runs on an A100.

Install with Python 3.12:

python -m pip install --upgrade deeplens-core==2.5.4

For NVIDIA CUDA 12.8:

python -m pip install --upgrade deeplens-core==2.5.4 torch==2.10.0+cu128 torchvision==0.25.0+cu128 --extra-index-url https://download.pytorch.org/whl/cu128

Validation: all 605 tests passed with CUDA enabled on an NVIDIA A100 80GB using Python 3.12.13, PyTorch 2.10.0+cu128, and torchvision 0.25.0+cu128. Ruff lint/format and strict Twine metadata checks passed. The installed wheel also passed ray-tracing and normalized complex-field PSF smoke tests on macOS.

Full changelog

Assets 5
Loading