This repository was archived by the owner on Feb 24, 2026. It is now read-only.
[Bugfix] Enhance LowerAsyncCopy Pass to handle INT8 dma copy with predicate #219
Merged
LeiWang1999 merged 7 commits intoOct 11, 2024
Conversation
Increase MAX_ERROR_MESSAGE_LENGTH from 200 to 500 Improve thread binding in MatmulTensorizationMMAWithDequantizeInfo
This file contains hidden or bidirectional Unicode text that may be interpreted or compiled differently than what appears below. To review, open the file in an editor that reveals hidden Unicode characters.
Learn more about bidirectional Unicode characters
Sign up for free
to subscribe to this conversation on GitHub.
Already have an account?
Sign in.
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Outdated suggestions cannot be applied.This suggestion has been applied or marked resolved.Suggestions cannot be applied from pending reviews.Suggestions cannot be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 across multiple files to enhance error handling, improve GPU matrix multiplication logic, and update integration benchmarks. The most important changes include increasing the maximum error message length, refining the logic for GPU matrix operations, and updating integration benchmarks.
Error Handling Improvements:
MAX_ERROR_MESSAGE_LENGTHfrom 200 to 500 inbitblas/common.py.GPU Matrix Multiplication Logic Enhancements:
block_reduction_depthand added a default value of 1 ifblock_reduction_depthisNoneinbitblas/gpu/matmul_mma_dequantize.py. [1] [2]reduce_kvalue inbitblas/gpu/matmul_mma_dequantize.py. [1] [2] [3] [4] [5] [6] [7] [8]Integration Benchmark Updates:
model.quantize()andtorch.compile(model)inintegration/BitNet/benchmark_inference_latency.py.Import Optimization:
integration/pytorch/bitblas_linear.pyby updating the import statement forMatmulConfigandMatmul.Submodule Update:
3rdparty/tvm.Ref to Issue #218