Scott McKay

Checker should validate the Loop outputs have names.

1

# Bug Report ### Describe the bug An ONNX Runtime user was attempting to create a Loop node where the loop carried dependencies were made optional outputs of the Loop...

Support direct usage of ORT format model flatbuffer for initializers

6

**Description**: An ORT format model contains initializer data that we currently copy into a `TensorProto` during model load, and copy again into an `OrtValue` during session state finalization. We can...

**Description**: Use the onnx headers to find the latest opset for each operator. This allows the script to detect optimizers with `graph_utils::IsSupportedOptypeVersionAndDomain` calls that need updating when run during the...

The KNN conversions (SklearnKNNTransformerDistance and SklearnKNNTransformerConnectivity ) produce models with a Sqrt prior to a TopK. As the Sqrt won't change the values selected by TopK, it would be cheaper...

### Description Add check that bias and scale sizes match norm_size in LayerNormalization. ### Motivation and Context #12917

Build error when including XNNPACK using FetchContent

4

If I include XNNPACK as a checksummed archive for security reasons using FetchContent e.g. ```cmake FetchContent_Declare(googlexnnpack URL https://github.com/google/XNNPACK/archive/21df17dd436635621588ce9de1e79bae4c9d4a4b.zip URL_HASH SHA1=dbd0c6c161894d35f17df07e892f434b10146b07) FetchContent_MakeAvailable(googlexnnpack) ``` the recent change in #5031 results in the...

Documentation does not show iOS16 ops for conv and conv_transpose

9

## 🐞Describing the bug I have an ML Program with a conv node. If the pad_type is set to `same` or `same_lower` I get an error saying the 'pad' input...

In ONNX a named dimension must have a single value when the model is executed. Due to that the input/output shapes for the recurrent values are incorrect. https://github.com/PeterL1n/RobustVideoMatting/blob/ebead27cb683e157b2bea7ca869daa820a07ba8f/export_onnx.py#LL49C1-L73C15 I believe...

### Description Xcode UI tests seem to be flaky: https://github.com/orgs/community/discussions/68807 Add a couple of retries if we get a "Timed out while loading Accessibility." error which is transient. ### Motivation...

### Description A number of Qualcomm Snapdragon chipsets do not produce correct output if we skip the Reshape, which ironically was a performance optimization for Snapdragon chips. Perf testing showed...