This repository was archived by the owner on Mar 30, 2022. It is now read-only.
v0.3 release. #175
Merged
v0.3 release. #175
Conversation
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
Member
dan-zheng
commented
Apr 23, 2019
dan-zheng
commented
- Add RELEASES.md: release notes (new features, acknowledgements, link to known issues).
- Add KNOWN_ISSUES.md: curated list of issues.
- Update README.md.
- Minor edits, add links to new docs.
- Add RELEASES.md: release notes (new features, acknowledgements, link to known issues). - Add KNOWN_ISSUES.md: curated list of issues. - Update README.md. - Minor edits, add links to new docs.
rxwei
reviewed
Apr 23, 2019
rxwei
reviewed
Apr 23, 2019
|
|
||
| ### TensorFlow Library | ||
|
|
||
| * [ ] `Tensor` advanced indexing and striding are not supported on GPU. |
There was a problem hiding this comment.
They are actually supported. What's not supported on GPU are the subscript setters only (i.e., assignment).
|
|
||
| ### Swift Standard Library Enhancements | ||
|
|
||
| * `AnyDerivative` has been added, representing a type-erased derivative. |
|
Hi,
I have repeatedly tried to mute this thread, but keep receiving emails
anyway. What do I need to do to be removed from this list? Thanks!
Joseph
…On Tue, Apr 23, 2019, 1:18 PM Richard Wei ***@***.***> wrote:
***@***.**** commented on this pull request.
------------------------------
In KNOWN_ISSUES.md
<#175 (comment)>:
> @@ -0,0 +1,95 @@
+# Swift for TensorFlow Known Issues
+
+This is a curated list of known Swift for TensorFlow issues. With every release,
+new issues are added and resolved issues are updated.
+
+Please see the [JIRA issue tracker](https://bugs.swift.org/projects/TF/issues)
+for a full list of known issues.
+
+## Version 0.3
+
+### TensorFlow Library
+
+* [ ] `Tensor` advanced indexing and striding are not supported on GPU.
Is there an API link?
------------------------------
In RELEASES.md
<#175 (comment)>:
> + [swift-apis#65](tensorflow/swift-apis#65),
+ [swift-apis#72](tensorflow/swift-apis#72))
+* Optimizer stored properties (e.g. `learningRate`) are now mutable.
+ ([swift-apis#81](tensorflow/swift-apis#81))
+
+### Automatic Differentiation
+
+* `Array` now conforms to `Differentiable`.
+ ([swiftlang/swift#23183](swiftlang/swift#23183))
+* The ***@***.***` attribute now works when the derivative function has a
+ generic context that is more constrained than the original function's generic
+ context.
+* The error `function is differentiable only with respect to a smaller subset of
+ arguments` is now obsolete.
+ ([swiftlang/swift#23887](swiftlang/swift#23887))
+* Differentiation-related memory leaks have been fixed.
⬇️ Suggested change
-* Differentiation-related memory leaks have been fixed.
+* Some differentiation-related memory leaks have been fixed.
------------------------------
In RELEASES.md
<#175 (comment)>:
> + ([swift-apis#61](tensorflow/swift-apis#61))
+ * `TransposedConv2D`.
+ ([swift-apis#64](tensorflow/swift-apis#64))
+ * `GlobalAveragePooling1D`, `GlobalAveragePooling2D`,
+ `GlobalAveragePooling3D`.
+ ([swift-apis#66](tensorflow/swift-apis#66),
+ [swift-apis#65](tensorflow/swift-apis#65),
+ [swift-apis#72](tensorflow/swift-apis#72))
+* Optimizer stored properties (e.g. `learningRate`) are now mutable.
+ ([swift-apis#81](tensorflow/swift-apis#81))
+
+### Automatic Differentiation
+
+* `Array` now conforms to `Differentiable`.
+ ([swiftlang/swift#23183](swiftlang/swift#23183))
+* The ***@***.***` attribute now works when the derivative function has a
And wrt: is now supported.
------------------------------
In RELEASES.md
<#175 (comment)>:
> +and refining APIs.
+
+### Notebook Environments (Colab and Jupyter)
+
+* Install SwiftPM packages using `%install` directives. See [documentation in
+ README](https://github.com/google/swift-jupyter#install-directives).
+ ([swift-jupyter#45](google/swift-jupyter#45),
+ [swift-jupyter#48](google/swift-jupyter#48),
+ [swift-jupyter#52](google/swift-jupyter#52))
+* `swift-jupyter` can now be installed in a Conda environment. See
+ [documentation in
+ README](https://github.com/google/swift-jupyter#option-2-using-a-swift-for-tensorflow-toolchain-and-conda).
+
+### Swift Standard Library Enhancements
+
+* `AnyDerivative` has been added, representing a type-erased derivative.
Is there an API link?
------------------------------
In RELEASES.md
<#175 (comment)>:
> + ([swiftlang/swift#23868](swiftlang/swift#23868))
+
+### Swift for TensorFlow Deep Learning Library
+
+* The `Layer` protocol `applied(to:)` method has been renamed to `call(_:)`.
+ `Layer`s are now “callable” like functions, e.g. `layer(input)`.
+ * Note: this is experimental functionality that is currently [being proposed
+ through Swift
+ Evolution](https://forums.swift.org/t/se-0253-static-callables/22243).
+ Expect some changes.
+* The `context` argument has been removed from `Layer`’s `applied(to:)` method.
+ It is replaced by the `ContextManager` class, which manages a stack of
+ contexts for a single thread.
+ ([swift-apis#87](tensorflow/swift-apis#87))
+* A `RNNCell` protocol has been added, generalizing simple RNNs, LSTMs, GRUs,
+ and NTMs. ([swift-apis#80](tensorflow/swift-apis#80),
I'd suggest removing "and NTMs" to be more conservative until we have an
implementation.
—
You are receiving this because you are subscribed to this thread.
Reply to this email directly, view it on GitHub
<#175 (review)>,
or mute the thread
<https://github.com/notifications/unsubscribe-auth/ABOOLPQLCD73CKURPO6FB5DPR5VI7ANCNFSM4HH56QGQ>
.
|
rxwei
reviewed
Apr 23, 2019
|
@jcatanza You can click the "Unwatch" button and select "Ignoring". Sorry for the inconvenience! |
|
Hey @jcatanza ! It seems you are watching this repo. Clicking the "unwatch" button at the top and selecting "ignoring" or "not watching" will prevent you from getting any emails or github notifications. |
rxwei
approved these changes
Apr 23, 2019
rxwei
reviewed
Apr 23, 2019
rxwei
reviewed
Apr 23, 2019
|
Thanks Richard, no worries. I'll try that.
Muting via email didn't work for me.
Regards,
Joseph
…On Tue, Apr 23, 2019, 1:28 PM Richard Wei ***@***.***> wrote:
@jcatanza <https://github.com/jcatanza> You can click the "Unwatch"
button and select "Ignoring". Sorry for the inconvenience!
[image: image]
<https://user-images.githubusercontent.com/856507/56613473-8c697a00-65cb-11e9-8743-6b86ef484206.png>
—
You are receiving this because you were mentioned.
Reply to this email directly, view it on GitHub
<#175 (comment)>, or mute
the thread
<https://github.com/notifications/unsubscribe-auth/ABOOLPSU6EZJBFEZAGUB2XLPR5WNHANCNFSM4HH56QGQ>
.
|
|
I'm curious, with the API changing and new versions getting released often, how often will Google Colab be updated with a new Swift toolchain? I'm specifically wondering in relation to the tutorials we have from the TensorFlow Summit back in March as from what I recall, these updates will break some of those tutorials (like the model building one I think). Will those notebooks be deprecated, will they run using an old version of the toolchain on Colab, will someone be constantly updating them every time a breaking API change is made, etc. |
|
@bartchr808 we will be updating colab over the next couple days, and we will be updating the notebooks to match. :-) |
|
@saeta awesome sounds good! 😄 |
|
Thanks to reviewers and @marcrasi for updating |
Sign up for free
to subscribe to this conversation on GitHub.
Already have an account?
Sign in.
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.