This repository was archived by the owner on Feb 24, 2026. It is now read-only.

[Dev] Add support and test case for Ladder Weight only Transformation Matmul Operator#212

Merged

LeiWang1999 merged 39 commits into

microsoft:mainfrom

LeiWang1999:tl_ops_dynamic

Oct 3, 2024

Commits on Sep 30, 2024

remove debug related items.
LeiWang1999
committed

Commits on Oct 1, 2024

Commits on Oct 2, 2024

Update submodule reference in 3rdparty/tvm
LeiWang1999
committed
Refactor matmul_finetune function to use topk=20 for hardware-aware finetuning
LeiWang1999
committed
Refactor submodule reference in 3rdparty/tvm
LeiWang1999
committed
lint fix
LeiWang1999
committed
Refactor test_general_matmul_tilelang_impl.py and test_tilelang_gemm.py
LeiWang1999
committed
Refactor MatmulConfig to enable weight propagation on supported devices
LeiWang1999
committed
Merge branch 'main' of https://github.com/microsoft/BitBLAS into tl_ops_dynamic
LeiWang1999
committed
Refactor test_general_matmul_tilelang_impl.py and test_general_matmul_tilelang_kernel.py to use centered random values for input tensors
LeiWang1999
committed
test fix
LeiWang1999
committed
Merge branch 'main' of https://github.com/microsoft/BitBLAS into tl_ops_dynamic
LeiWang1999
committed
test fix
LeiWang1999
committed
Refactor flash attention tests to use centered random values for input tensors
LeiWang1999
committed
Refactor flash attention tests to use centered random values for input tensors
LeiWang1999
committed
Refactor flash attention tests to skip test if flash_attn is not installed
LeiWang1999
committed

Commits on Oct 3, 2024