Merge latest code #4
Open
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
Signed-off-by: Deyu Huang <[email protected]> Co-authored-by: Jay Zhang <[email protected]>
…ce_value_info (onnx#1930) Signed-off-by: Chun-Wei Chen <[email protected]>
Signed-off-by: Deyu Huang <[email protected]>
Signed-off-by: Francesco Di Natale <[email protected]> Co-authored-by: Jay Zhang <[email protected]>
* drop python36 unit tests Signed-off-by: Deyu Huang <[email protected]> * add doc change Signed-off-by: Deyu Huang <[email protected]> Co-authored-by: Jay Zhang <[email protected]>
* add opset 16 support and ci * python 3.7 requires tf>=1.13 Signed-off-by: Deyu Huang <[email protected]>
* Update the default opset version for tf2onnx. * Update README.md accordingly. Signed-off-by: Jay Zhang <[email protected]>
* Update the way to check input_signature in from_function(). Signed-off-by: Jay Zhang <[email protected]>
Signed-off-by: Jay Zhang <[email protected]>
* update protobuf version in ci Signed-off-by: Deyu Huang <[email protected]> * add change in unit_test Signed-off-by: Deyu Huang <[email protected]>
* add TensorScatterAdd op conversion in opset 16 Signed-off-by: Deyu Huang <[email protected]> * add unit test Signed-off-by: Deyu Huang <[email protected]> * fix pylint name Signed-off-by: Deyu Huang <[email protected]> * fix typo and test op check Signed-off-by: Deyu Huang <[email protected]> Co-authored-by: Jay Zhang <[email protected]> Co-authored-by: Jay Zhang <[email protected]>
* Add a new api from_tflite to improve user experience. Signed-off-by: Jay Zhang <[email protected]>
* upgrade opset 16 status doc Signed-off-by: Deyu Huang <[email protected]>
Signed-off-by: Jay Zhang <[email protected]>
increment main to 1.11 version Signed-off-by: Deyu Huang <[email protected]>
…nnx#1964) * Transpose optimization for Softmax and LogSoftmax (fixes onnx#1716) In opsets 13 and higher, the axis of the operation is arbitrary and can simply be changed according to the permutation of the Transpose. In lower opsets, Softmax always coerces its inputs to a 2D tensor, making Transpose operations necessary if the permutation moves axes between the coerced batch and feature dimensions. Signed-off-by: janbernloehr <[email protected]> Co-authored-by: fthielke <[email protected]>
…ut_names and output_names. (onnx#1967) * The from_tflite() function should not change the value of None to an empty list for input_names and output_names. * Change the way to validate a list is None or Emtpy. Signed-off-by: Jay Zhang <[email protected]>
…onnx#1971) * make Loop op with maximum iterations M equal to empty string to match onnx spec Signed-off-by: Deyu Huang <[email protected]>
Update README.md typos Signed-off-by: andife <[email protected]>
Update CLA -> DCO (as all onnx projects) Signed-off-by: andife <[email protected]>
* Update all Keras related tests to support latest TF version. Signed-off-by: Jay Zhang <[email protected]>
* Update tests in nightly keras-application-test. Signed-off-by: Jay Zhang <[email protected]>
L2_NORMALIZATION support for tflite Signed-off-by: Shesung <[email protected]> Co-authored-by: Deyu Huang <[email protected]>
Replace deprecated `np.object` with `object` Signed-off-by: Václav Volhejn <[email protected]> Co-authored-by: Deyu Huang <[email protected]>
…onnx#1979) * add output_as_nchw Signed-off-by: Deyu Huang <[email protected]> * fix node replace logic Signed-off-by: Deyu Huang <[email protected]> * add tests for outputs as nchw Signed-off-by: Deyu Huang <[email protected]> * add it into function and doc Signed-off-by: Deyu Huang <[email protected]> * fix output_names_with_port range Signed-off-by: Deyu Huang <[email protected]> * fix the input_as_nchw description Signed-off-by: Deyu Huang <[email protected]> * change tests name Signed-off-by: Deyu Huang <[email protected]>
* add tranposeOptimizer split for opset >= 13 Signed-off-by: Deyu Huang <[email protected]> * add test Signed-off-by: Deyu Huang <[email protected]> * fix comments Signed-off-by: Deyu Huang <[email protected]>
…nnx#2000) If the Graph object being constructed already contains a node of the name of a Const node in orginal graph, do not add as input. Signed-off-by: Yu Cong <[email protected]> Co-authored-by: Yu Cong <[email protected]> Co-authored-by: Deyu Huang <[email protected]>
…#2001) Add pattern matching and parsing where Keras LSTM uses `HardSigmoid` as the recurrent activation. Signed-off-by: Yu Cong <[email protected]>
* Add more tests to cover tf 2.9.x. * Set tensorflowjs to a fixed version 3.18.0. * Fix an issue of installing tensorflow-text component. Signed-off-by: Jay Zhang <[email protected]>
…nx#2007) Signed-off-by: southfreebird <[email protected]> Co-authored-by: iolkhovsky <[email protected]>
* Improve ZerosLike implementation for opset >= 9 Signed-off-by: Deyu Huang <[email protected]> Co-authored-by: Guenther Schmuelling <[email protected]> * add a blank line Signed-off-by: Deyu Huang <[email protected]> Co-authored-by: Guenther Schmuelling <[email protected]> Co-authored-by: Jay Zhang <[email protected]>
Signed-off-by: Jay Zhang <[email protected]> Co-authored-by: Deyu Huang <[email protected]>
* python3.10 support Signed-off-by: Deyu Huang <[email protected]> * add tests except keras Signed-off-by: Deyu Huang <[email protected]> * add protubuf limit for tf 2.9.1 and ort to 1.12.0 Signed-off-by: Deyu Huang <[email protected]> * onnxruntime-extensions/yaml is not support with python3.10 Signed-off-by: Deyu Huang <[email protected]> * remove onnxruntime-extensions support in py310 Signed-off-by: Deyu Huang <[email protected]> Co-authored-by: Jay Zhang <[email protected]>
Signed-off-by: Deyu Huang <[email protected]>
Signed-off-by: Deyu Huang <[email protected]>
* ONNX opset 17 with IR version 8 support Signed-off-by: Deyu Huang <[email protected]>
Signed-off-by: Deyu Huang <[email protected]>
Signed-off-by: Deyu Huang <[email protected]>
Signed-off-by: Deyu Huang <[email protected]>
* Turn on graph tf optimize grappler dependency Signed-off-by: Deyu Huang <[email protected]> * Aviod output name rewrite in random normal Signed-off-by: Deyu Huang <[email protected]> Co-authored-by: Jay Zhang <[email protected]>
Sign up for free
to join this conversation on GitHub.
Already have an account?
Sign in to comment
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.
No description provided.