onnxruntime
ec0e4d3b
- Parallel Transpose_BSNH_to_BNSH (#19406)
Go
Login via GitHub
Home
Pricing
FAQ
Install
Login
via GitHub
Commit
View On
GitHub
Commit
2 years ago
Parallel Transpose_BSNH_to_BNSH (#19406) Achieved a speedup of 1.098 in MultiHeadAttention and an end-to-end speedup of 1.021 in the OCR model through parallelization of the Transpose_BSNH_to_BNSH operation.
References
#19406 - Parallel Transpose_BSNH_to_BNSH
Author
yihonglyu
Parents
937cdd65
Loading