fix: Add nl.zeros with buffer parameter support#609
Closed
mrkcath-aws wants to merge 1 commit intomainfrom
Closed
Conversation
nl.zeros was previously mapped to builtin.lang.zeros which was moved to builtin.tensor.zeros (a trace-time constant) in NKI-499, breaking the nl.zeros -> builtin.lang.zeros mapping. Customers using nl.zeros with buffer= (e.g. buffer=nl.sbuf) received 'unexpected keyword argument buffer' errors. Add builtin.lang.zeros to Lang.lean that allocates an on-device tensor (matching ndarray semantics) and emits a memset to zero-initialize it. Supports buffer=nl.sbuf (default), nl.psum, nl.hbm, nl.shared_hbm, and integer dtypes. Workaround for customers on affected releases: ZS = nl.ndarray(shape, dtype=dtype, buffer=buffer) nisa.memset(ZS, value=0.0) Fixes: V2117488695
mrkcath-aws
requested review from
govereau,
jiancheng-aws,
joonwonc,
jtristan,
kerrijoe-aws,
ppotapov-aws,
seanmcl,
waahm7 and
yongweiy
as code owners
Collaborator
Author
|
Implement in user library |
This file contains hidden or bidirectional Unicode text that may be interpreted or compiled differently than what appears below. To review, open the file in an editor that reveals hidden Unicode characters.
Learn more about bidirectional Unicode characters
Sign up for free
to join this conversation on GitHub.
Already have an account?
Sign in to comment
1 participant
Add this suggestion to a batch that can be applied as a single commit.This suggestion is invalid because no changes were made to the code.Suggestions cannot be applied while the pull request is closed.Suggestions cannot be applied while viewing a subset of changes.Only one suggestion per line can be applied in a batch.Add this suggestion to a batch that can be applied as a single commit.Applying suggestions on deleted lines is not supported.You must change the existing code in this line in order to create a valid suggestion.Outdated suggestions cannot be applied.This suggestion has been applied or marked resolved.Suggestions cannot be applied from pending reviews.Suggestions cannot be applied on multi-line comments.Suggestions cannot be applied while the pull request is queued to merge.Suggestion cannot be applied right now. Please check back later.
nl.zeros was previously mapped to builtin.lang.zeros which was moved to builtin.tensor.zeros (a trace-time constant) in NKI-499, breaking the nl.zeros -> builtin.lang.zeros mapping. Customers using nl.zeros with buffer= (e.g. buffer=nl.sbuf) received 'unexpected keyword argument buffer' errors.
Add builtin.lang.zeros to Lang.lean that allocates an on-device tensor (matching ndarray semantics) and emits a memset to zero-initialize it. Supports buffer=nl.sbuf (default), nl.psum, nl.hbm, nl.shared_hbm, and integer dtypes.
Workaround for customers on affected releases:
Fixes: V2117488695