Donglin Zhuang
Donglin Zhuang
May I expect any update on this? std::cout in device code would be a sweet feature to facilitate debugging.
Has the cherrypick PR merged in r2.10? This error still blocks building from the source with the r2.10 branch.
@rishikasinha-tf @mihaimaruseac Kindly ping... This issue is still present on the latest r2.10 branch. commit: 829f9ebd6bf7fa9ce4fb7e79c0142bda3cadc134
upvote this feature request.
Did some trials and looks like its related to TmaWarpSpecialized epilogue schedule. In comparison, NoSmemWarpSpecialized schedule works for small m.
Thank you so much for the explanation.
@terrychenism maybe not, https://github.com/facebookincubator/AITemplate/pull/126 is later than the commit I used.
@terrychenism Thank you!
@shibin2018 @JennieCheng Try this: python -m spacy download en