Igor Safonov
Igor Safonov
Hi @SNikhilVashista , Thanks for the RFC. I have a few questions: - What are the requirements for AMD support to work (minimal dpc++ version, hipDNN version, etc.) - I...
> As far as I can tell, the pipeline failures are the same as on master Yes, I created an issue for our team, thanks for the information.
@jondea @nSircombe The changes landed into master and were backported into rls-v2.7. Thanks for the contribution!
The changes landed into master and were backported into rls-v2.7. Thanks for the contribution! There were some issues with Gihub updating the PR changes once the #1330 was promoted (it...
The build fails for oneDNN v2.2 as well
tagging some folks from CPU team: @kwiersch @msotoflo @akharito Please take a look
@kawakami-k @jondea @nSircombe @cfRod Please review the RFC because it brings a lot of changes to quantization API in oneDNN in future major release (3.0). #### 1. As part of...
> Hi @igorsafo, > > Thanks for the heads-up. It's LGTM. Are there any other RFCs or documents regarding API changes in version 3.0? Hi @kawakami-k , This is a...
> We have a general propose for scales and zero points from PyTorch GPU team. > > With current design of oneDNN, it always assumes the scale and zero points...
> @igorsafo The RFC proposed both short-term and long-term changes. May I know what changes target v3.0? Some comments from PyTorch side: Long-term changes are the changes for v3.0. Short-term...