Add missing algorithms, add stylistic improvements, update with spec conventions #446
Conversation
Signed-off-by: Zoltan Kis <[email protected]>
Squashed from the following commits:
Replace MLOperand.[[descriptor]] with type and dimensions
Clarify the algorithm for only setting up the op
Improve the clamp() algorithm, use the prose assuming the create steps for MLOperand and MLActivation
Rework clamp with polymorphic behavior. Update for changes in MLOperand.
Rework clamp() like constant(), polymorphic forms in separate sections, argument and return descriptions as notes.
Fix platform related steps and reference to internal slots
Address review, remove note
Remove back quotes from title
Signed-off-by: Zoltan Kis <[email protected]>
Signed-off-by: Zoltan Kis <[email protected]>
Squashed from the following commits:
Add reference to the platform operand object
Address review comment, change note.
Signed-off-by: Zoltan Kis <[email protected]>
Signed-off-by: Zoltan Kis <[email protected]>
Signed-off-by: Zoltan Kis <[email protected]>
Signed-off-by: Zoltan Kis <[email protected]>
Signed-off-by: Zoltan Kis <[email protected]>
…ithm Add the batch normalization algorithm
Add algorithm to concat()
Clamp algorithm
Signed-off-by: Zoltan Kis <[email protected]>
…ing algorithms Signed-off-by: Zoltan Kis <[email protected]>
Signed-off-by: Zoltan Kis <[email protected]>
…integration Add stylistic changes
…to zk-conventions-integration
Signed-off-by: Zoltan Kis <[email protected]>
Squashed from the following commits:
Add reference to the platform operand object
Address review comment, change note.
Signed-off-by: Zoltan Kis <[email protected]>
Signed-off-by: Zoltan Kis <[email protected]>
Signed-off-by: Zoltan Kis <[email protected]>
Signed-off-by: Zoltan Kis <[email protected]>
Squashed from the following commits:
Replace MLOperand.[[descriptor]] with type and dimensions
Clarify the algorithm for only setting up the op
Improve the clamp() algorithm, use the prose assuming the create steps for MLOperand and MLActivation
Rework clamp with polymorphic behavior. Update for changes in MLOperand.
Rework clamp() like constant(), polymorphic forms in separate sections, argument and return descriptions as notes.
Fix platform related steps and reference to internal slots
Address review, remove note
Remove back quotes from title
Signed-off-by: Zoltan Kis <[email protected]>
Signed-off-by: Zoltan Kis <[email protected]>
Signed-off-by: Zoltan Kis <[email protected]>
Signed-off-by: Zoltan Kis <[email protected]>
…ing algorithms Signed-off-by: Zoltan Kis <[email protected]>
Signed-off-by: Zoltan Kis <[email protected]>
…s/webnn into zk-conventions-integration
Signed-off-by: Zoltan Kis <[email protected]>
…, clamp, concat, batch norm Signed-off-by: Zoltan Kis <[email protected]>
Signed-off-by: Zoltan Kis <[email protected]>
Signed-off-by: Zoltan Kis <[email protected]>
Signed-off-by: Zoltan Kis <[email protected]>
Signed-off-by: Zoltan Kis <[email protected]>
zolkis
force-pushed
the
zk-conventions-integration
branch
from
d04f364 to
6818a1c
Compare
Signed-off-by: Zoltan Kis <[email protected]>
Signed-off-by: Zoltan Kis <[email protected]>
Signed-off-by: Zoltan Kis <[email protected]>
| <div class=algorithm-steps> | ||
| 1. [=Assert=]: |op| is one of "reduceL1", "reduceL2", "reduceLogSum", "reduceLogSumExp", "reduceMax", "reduceMean", "reduceMin", "reduceProduct", "reduceSum", "reduceSumSquare". | ||
| 1. [=Assert=]: the type of |input| is {{MLOperand}}. | ||
| 1. If |options|.{{MLReduceOptions/axes}} [=map/exists=], if any of its elements is not in [=the range=] 0 to the [=rank=] of |input|, exclusive, then [=exception/throw=] a "{{DataError}}" {{DOMException}}. |
There was a problem hiding this comment.
in [=the range=] 0 to N-1 where N is the [=rank=] of the input tensor.
There was a problem hiding this comment.
The "exclusive" word does that. :)
The full expression is
in [=the range=] 0 to the [=rank=] of |input|, exclusive
There was a problem hiding this comment.
After the final check with other reviewers, all the comments are either addressed or tracked by issues. With that, this PR is ready for merging.
Thanks all for the productive review! Special thanks to @zolkis 's great work!
@wchao1115 , you may want to take another look before merging. Thanks!
|
Thank you everyone for taking this major effort to completion with special thanks to @zolkis! I asked @zolkis to acknowledge the contributors in 18442ad -- if any names are missing (or misspelled), please let us know and we'll fix. @wchao1115, feel free to proceed at your earliest convenience. |
|
@wchao1115 @huningxin thank you for your approval. As the editors, I will let you discuss among yourself whether you would like to squash this big PR or retain the full commit history. You're free to merge when you are decided on the best path forward. I have a slight preference for retaining the commits to help future spec archeologists. There's this idiom "how the sausage is made". :) |
|
I suppose even if we squash this PR and merge it as one commit, the full commit history in this PR would still be preserved, correct? |
|
@huningxin, if that works it would be best of both worlds: keep the main commit history clean while retain the detailed commit history in the PR. I don't remember whether we need to keep the integration branch around (not delete it) for this to work. Perhaps you want to test this in a test repo to find out. Sometimes we do a clean up of stale branches and this branch might accidentally get deleted in the future. |
I'd keep the separate commits because 1. there are different authors, and 2. one such big commit would be impractical for its sheer size. Up to the editors, though. |
I do not have a strong preference one way or the other. Either works for me. |
|
Thanks for the feedbacks. Let's merge this pull request with all commits history. |
SHA: 34a4d80 Reason: push, by huningxin Co-authored-by: github-actions[bot] <41898282+github-actions[bot]@users.noreply.github.com>
A series of changes including:
Preview | Diff