biondizzle/nvfp4-megamoe-kernel
1
0
Fork 0
Code Issues Pull Requests Actions Packages Projects Releases Wiki Activity

Restructure: cutedsl/ -> dsv4/ with proper layering

Browse Source 
- Split bridge.py -> ops/quantize.py, ops/layouts.py, ops/gemm_runner.py
- Renamed classes: CuTeDSLNvfp4Linear -> Nvfp4Linear, etc.
- Moved kernel code to dsv4/kernels/ (gemm, attention, compressor, decode, cuda)
- Moved PyTorch bridges to dsv4/ops/
- Moved nn.Module layers to dsv4layers/
- Moved reference implementations to dsv4/reference/
- Moved vendored CUTLASS code to vendored/
- Archived ~190 debug tests to tests/archive/
- Kept ~15 canonical tests in tests/unit/
- Updated all import paths
- Added stubs for future components (model/, cache/, loader/)
- Updated pyproject.toml: dsv4-inference package name
This commit is contained in:
biondizzle
2026-05-21 17:30:44 +00:00
parent 99e143dd0e
commit 3fb3c925af
274 changed files with 715 additions and 556 deletions
Download Patch File Download Diff File Expand all files Collapse all files

2
dsv4/cache/paged_cache.py vendored Normal file
View File

@@ -0,0 +1,2 @@
"""Paged KV cache."""
# TODO: Phase 3