Code

feat: parallel scan extension for CPU#17

Code

Open

yoyolicoris wants to merge 13 commits intomainDiffAPF/torchlpc:mainfrom feat-cpp-scanDiffAPF/torchlpc:feat-cpp-scanCopy head branch name to clipboard

+222-35

Commits

Commits on Jan 22, 2025

• draft: scan extension on cpu

  yoyolicoris

  committed
  1f392ba

  Copy full SHA for 1f392ba

• include cpp extension in setup.py

  yoyolicoris

  committed
  d77846a

  Copy full SHA for d77846a

• fix: remove extra arg

  yoyolicoris

  committed
  37b3340

  Copy full SHA for 37b3340

Commits on Jan 23, 2025

• fix: compile errors

  yoyolicoris

  committed
  e110d14

  Copy full SHA for e110d14

• use dev versioning

  yoyolicoris

  committed
  4611ce2

  Copy full SHA for 4611ce2

• load library file when being imported

  yoyolicoris

  committed
  5e28b30

  Copy full SHA for 5e28b30

• test equivalence to numba version

  yoyolicoris

  committed
  5b69d2a

  Copy full SHA for 5b69d2a

• refactor: return tensor instead of void function

  yoyolicoris

  committed
  7b54fb1

  Copy full SHA for 7b54fb1

• refactor: rename RecurrenceCUDA to Recurrence to cover CPU device

  yoyolicoris

  committed
  1e7ea16

  Copy full SHA for 1e7ea16

• refactor: update functions to use Recurrence for CPU and CUDA devices

  yoyolicoris

  committed
  314c5a2

  Copy full SHA for 314c5a2

• refactor: remove contiguous check besides output tensor

  yoyolicoris

  committed
  1472acf

  Copy full SHA for 1472acf

• refactor: add warning for missing _C*.so file and check extension loading in Recurrence

  yoyolicoris

  committed
  5e7a914

  Copy full SHA for 5e7a914

• ci: add workflow step to build CPP extension and copy shared objects

  yoyolicoris

  committed
  6cde45b

  Copy full SHA for 6cde45b