MoisesHer opened a new pull request #17652: Use multi-tensor sumSQ in clip_global_norm URL: https://github.com/apache/incubator-mxnet/pull/17652 ## Description ## Using multi-tensor sum of squares in gluon: clip_global_norm. Instead of computing the sum of squares of each input array sequentially, compute them in parallel (multi-tensor). ## Checklist ## ### Essentials ### Please feel free to remove inapplicable items for your PR. - [x] Changes are complete (i.e. I finished coding on this PR) - [x] All changes have test coverage - [x] Code is well-documented - [x] 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 ### - [x] gluon: clip_global_norm was modified to use mxnet.nd.multi_sum_sq op and compute the sumSq of several arrays in parallel - [x] Test in tests/python/gpu/test_gluon_gpu:test_global_norm_clip_multi_device was extended to have 2 arrays on gpu(0) and 2 arrays on cpu(0), so that multi-tensor sumSq is tested for each of these contexts
---------------------------------------------------------------- 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
