Repository navigation

Cache load_inline Torch headers in a Modal Volume - #527

Merged

msaroufim merged 8 commits into

mainfrom

codex/modal-pch-volume

Sep 7, 2026

Merged

Cache load_inline Torch headers in a Modal Volume#527
msaroufim merged 8 commits into
mainfrom
codex/modal-pch-volume

Commits on Sep 7, 2026
- Cache PyTorch inline headers on Modal volumes
  msaroufim
  committed
  b16acc3
- Warm header profiles before the existing Modal rollout
  msaroufim
  committed
  38b9892
- Record fresh T4 PCH benchmarks on current main
  msaroufim
  committed
  6c502bd
- Test PCH hits, misses, changed headers and non-inline builds on Modal
  msaroufim
  committed
  3f9cd13
- Explain automatic PCH usage for submission authors
  msaroufim
  committed
  3b8fcb8
- Benchmark instantiated CUTLASS GEMMs and trim PCH experiment artifacts
  msaroufim
  committed
  8232215
- Measure CUTLASS with default and minimal CUDA headers
  msaroufim
  committed
  e226b86
- Focus PCH results on load_inline submissions
  msaroufim
  committed
  034fda3