-
Notifications
You must be signed in to change notification settings - Fork 14
New issue
Have a question about this project? Sign up for a free GitHub account to open an issue and contact its maintainers and the community.
By clicking “Sign up for GitHub”, you agree to our terms of service and privacy statement. We’ll occasionally send you account related emails.
Already on GitHub? Sign in to your account
Add conversions for stablehlo ceil, sine and cosine Ops #939
Conversation
@uazizTT this look great! If these new ops are expected to work on silicon, can you add tests for them under Added Silicon tests. |
813f13f
to
3e5e976
Compare
e700b84
to
9525a8d
Compare
9525a8d
to
4caae05
Compare
There was a problem hiding this comment.
Choose a reason for hiding this comment
The reason will be displayed to describe this comment to others. Learn more.
Can you add a ceil, concat and cosine test under Silicon/TTNN/perf_unit?
4caae05
to
4fe5125
Compare
There was a problem hiding this comment.
Choose a reason for hiding this comment
The reason will be displayed to describe this comment to others. Learn more.
great! Looks good, thanks!
ed4c0b3
to
6434de2
Compare
6434de2
to
dea22dd
Compare
@uazizTT I don't think it's okay to merge PRs with unresolved comments without any discourse. I didn't want to block you so I didn't put "Requests changes" status, but I see no other way to prevent these situations in the future. I know I might be too pedantic on PRs and it is totally fine to disagree with my comments and explain why you won't fix them, or say "I will follow up with another PR", but I don't think it's okay to just ignore comments. |
Hey @mrakitaTT , I thought I covered all the comments and got a thumbs up from everyone before merging, but reading your last comment again, it seems you were asking to break down Anyways that is outside the scope of this PR, so let me create a new issue for it to decompose both |
Created an issue to split |
@uazizTT No problem, let's just please try in the future to have proper discourse in comments and to resolve all comments you think you've addressed. As I said it is totally fine to disagree with the comment or "won't fix" it, the important thing is to have a conversation about it. In this case there was no conversation and the comment was left unresolved, so I couldn't know what you've assumed. When you mark comment as resolved I can at least know that you think you've addressed it so I can ping you about it. The comment in question here wasn't about splitting the test file, you've marked that one as resolved and I didn't object. The comment in question here was about improving FILECHECK statements in conversion tests. |
OK will cover them in subsequent PR. |
Updated the tests in this PR #1122. |
Great, thank you Usman! |
Added tests and refactored unary ops.