[Dev] Improve Dequant performance on CUDA Simt #189
Merged
Add this suggestion to a batch that can be applied as a single commit.
This suggestion is invalid because no changes were made to the code.
Suggestions cannot be applied while the pull request is closed.
Suggestions cannot be applied while viewing a subset of changes.
Only one suggestion per line can be applied in a batch.
Add this suggestion to a batch that can be applied as a single commit.
Applying suggestions on deleted lines is not supported.
You must change the existing code in this line in order to create a valid suggestion.
Outdated suggestions cannot be applied.
This suggestion has been applied or marked resolved.
Suggestions cannot be applied from pending reviews.
Suggestions cannot be applied on multi-line comments.
Suggestions cannot be applied while the pull request is queued to merge.
Suggestion cannot be applied right now. Please check back later.
This pull request includes several changes to improve error handling, code clarity, and functionality in the
bitblas
module. The most important changes include adding error handling when loading runtime modules, fixing a typo in comments, introducing alignment factors for vectorization, and adding new methods for fetching data to local memory.Error Handling:
bitblas/cache/operator.py
: Added error handling for loading runtime modules, logging an error message if the module fails to load.Code Clarity:
bitblas/gpu/matmul.py
: Fixed a typo in comments, changing "vtrhead" to "vthread".Functionality Enhancements:
bitblas/gpu/matmul.py
: Introduced analign_factor
to handle unalignedK
andBK
values, and expanded the vectorization factor. [1] [2]bitblas/gpu/matmul.py
: Added a new methodfetch_to_local
to fetch data to local memory with alignment considerations.bitblas/ops/operator.py
: Modifiedupdate_runtime_module
to handle cases wherert_mod
isNone
and added conditional evaluation oftime_evaluator
.