Conversation
|
Thanks! Update: This error is now fixed 🟢 |
| :: | ||
| Whether {{MLTensor}}.{{MLTensor/destroy()}} has been called. Once destroyed, the {{MLTensor}} can no longer be used. | ||
|
|
||
| : <dfn>\[[data]]</dfn> of an [=implementation-defined=] type |
There was a problem hiding this comment.
I considered making this a Data Block but I haven't fully thought through the implications of that yet, so... I went for maximal hand-waviness for now
a-sully
force-pushed
the
mltensor-spec
branch
from
4e1dad9 to
e24f5c5
Compare
|
Apologies for the accidental force-push 🤦 Thanks for the feedback @inexorabletash! I believe I've addressed all your comments |
inexorabletash
left a comment
inexorabletash
left a comment
There was a problem hiding this comment.
Thanks! Most updates LGTM.
a-sully
force-pushed
the
mltensor-spec
branch
from
26748bf to
4043982
Compare
|
Thanks for the review! Feel free to merge at your convenience 🙏 |
|
@inexorabletash @reillyeon Any further comments? Also add @fdwr , PTAL. |
inexorabletash
left a comment
inexorabletash
left a comment
There was a problem hiding this comment.
LGTM! (I can't mark comment threads as resolved)
See discussion at: webmachinelearning/webnn#787 (comment) Change-Id: I9b00c041b2a12e72d4a7147d6b31dc4ad02d8b87 Reviewed-on: https://chromium-review.googlesource.com/c/chromium/src/+/6044441 Auto-Submit: Austin Sullivan <[email protected]> Reviewed-by: Joshua Bell <[email protected]> Commit-Queue: Joshua Bell <[email protected]> Cr-Commit-Position: refs/heads/main@{#1387182}
There was a problem hiding this comment.
Minor comments, else LGTM. Thanks for including in the spec.
@RafaelCintron, any thoughts? Rafael skimmed and messaged me.
SHA: dd5cbdc Reason: push, by fdwr Co-authored-by: github-actions[bot] <41898282+github-actions[bot]@users.noreply.github.com>
5 participants
This PR specifies key methods related to
MLTensor, while hand-waving over much of the juicy details, such as how to actually specify theMLContext's timeline (#529)My hope is that this is a mostly-straightforward improvement which is "good enough" to allow us to remove
MLContext.compute()as well as providing a more concrete starting point for more detailed discussions about e.g. specifying timelines in follow-up PRsPreview | Diff