All pull requests

Pull requests

• added MacOs support for pytorch (mlx)

  #119·fwild opened Oct 18, 2026·

  ·

• Skip completed per-frame predictions in batch mode

  #116·Aliciawque opened Oct 10, 2026·

  ·

• Preserve selected frame paths for sky-mask caching

  #115·Aliciawque opened Oct 10, 2026·

  ·

• feat(thor): add opt-in SM110 inference optimizations

  #109·buptxzz opened Sep 14, 2026·

  ·

• Fix GPU memory waste in demo.py and a viewer scene-corruption bug

  #104·Zawaer opened Aug 22, 2026·3 tasks done·

  ·

• Windows/ gmae test

  #103·3a1b2c3 opened Aug 21, 2026·

  ·

• docs: add Chinese and Japanese READMEs

  #95·tura-ai-agent opened Jul 22, 2026·

  ·

• Add Apple Silicon (MPS) support with bf16 autocast

  #88·andeplane opened Jul 17, 2026·

  ·
  1

• Add CPU-first inference with automatic SDPA fallback

  #85·Lakshya717 opened Jul 8, 2026·

  ·

• feat: defaulting CPU offload + GPU OOM fix + viewer PLY export

  #82·nvandamme opened Jul 1, 2026·

  ·

• Make FlashInfer cache dtype hardware-aware (fix bf16 on SM<80)

  #56·os-gabe opened Apr 29, 2026·

  ·

• Add Mac metal MLX inference

  #50·Ma-Dan opened Apr 26, 2026·

  ·
  2

• Low memory version demo

  #43·YoshiRi opened Apr 23, 2026·

  ·
  1

• add docker feature to the project

  #34·wild-deer opened Apr 21, 2026·

  ·
  1

• fix: prevent arbitrary code execution in torch.load()

  #25·rajpratham1 opened Apr 19, 2026·

  ·
  1

• Add viewer dependencies to pyproject.toml

  #19·HenkPoley opened Apr 18, 2026·

  ·

• Add Hugging Face checkpoint loading with -hf

  #18·HenkPoley opened Apr 18, 2026·

  ·
  1

• Add Mac MPS accelerated inference support

  #17·HenkPoley opened Apr 18, 2026·

  ·
  2

• Add saved prediction viewer workflow

  #16·HenkPoley opened Apr 18, 2026·

  ·

• Add (Mac) webcam capture input

  #15·HenkPoley opened Apr 18, 2026·

  ·
  1

• upgrade it.

  #14·saarors opened Apr 18, 2026·

  ·

• Add end-to-end I/O streaming demo with live browser udpates

  #13·L42ARO opened Apr 18, 2026·

  ·
  2