blchu opened a new pull request #17994: Tensor cores used only for fp16 in interleaved multihead attention URL: https://github.com/apache/incubator-mxnet/pull/17994 ## Description ## Fixed issue where fp32 inputs use tensor cores for the interleaved multihead attention operators, resulting in lower precision calculations and potential reduction in accuracy. ## Checklist ## ### Essentials ### Please feel free to remove inapplicable items for your PR. - [ ] Changes are complete (i.e. I finished coding on this PR) - [ ] All changes have test coverage: - Unit tests are added for small changes to verify correctness (e.g. adding a new operator) - Nightly tests are added for complicated/long-running ones (e.g. changing distributed kvstore) - Build tests will be added for build configuration changes (e.g. adding a new build option with NCCL) - [ ] Code is well-documented: - For user-facing API changes, API doc string has been updated. - For new C++ functions in header files, their functionalities and arguments are documented. - For new examples, README.md is added to explain the what the example does, the source of the dataset, expected performance on test set and reference to the original paper if applicable - Check the API doc at https://mxnet-ci-doc.s3-accelerate.dualstack.amazonaws.com/PR-$PR_ID/$BUILD_ID/index.html - [ ] To the best of my knowledge, examples are either not affected by this change, or have been fixed to be compatible with this change ### Changes ### - [ ] Set interleaved multihead attention GEMM default to not use tensor cores, and only use if input data type is fp16 - [ ] No longer checks for tensor input shape divisibility by 8 ## Comments ##
---------------------------------------------------------------- This is an automated message from the Apache Git Service. To respond to the message, please log on to GitHub and use the URL above to go to the specific comment. For queries about this service, please contact Infrastructure at: us...@infra.apache.org With regards, Apache Git Services
