GitLab

Co-authored-by: Woosuk Kwon <woosuk.kwon@berkeley.edu>

[Bugfix] Allow ScalarType to be compiled with pytorch 2.3 and add checks for registering FakeScalarType and dynamo support. (#7886)

Co-authored-by: Alexander Matveev <alexm@neuralmagic.com>

Signed-off-by: omrishiv <327609+omrishiv@users.noreply.github.com>

Co-authored-by: Cody Yu <hao.yu.cody@gmail.com>

Co-authored-by: afeldman-nm <156691304+afeldman-nm@users.noreply.github.com>

[Core] Fix tracking of model forward time to the span traces in case of PP>1 (#7440)

Signed-off-by: omrishiv <327609+omrishiv@users.noreply.github.com>
Co-authored-by: Cyrus Leung <cyrus.tl.leung@gmail.com>

Co-authored-by: Mor Zusman <morz@ai21.com>

[Core] Subclass ModelRunner to support cross-attention & encoder sequences (towards eventual encoder/decoder model support) (#4942)
Co-authored-by: Andrew Feldman <afeld2012@gmail.com>
Co-authored-by: Nick Hill <nickhill@us.ibm.com>