Przemyslaw Wysocki

Results 43 issues of Przemyslaw Wysocki

Signed-off-by: p-wysocki **Description** - New optional Split attribute `num_outputs` determining the number of equally sized output tensors. **Motivation and Context** - Fixes an issue described at https://github.com/onnx/onnx/issues/4188

### Details: - Remove Python 3.6 from docs, use Python 3.7 on GitHub workflows - Not to be merged until explicit permission given (see ticket) ### Tickets: - 86687 Notes:...

do not merge

### Description - new `num_outputs` attribute - support for splitting tensors into uneven chunks added ### Motivation and Context - solves https://github.com/onnx/onnx/issues/4188 - solves https://github.com/onnx/onnx/issues/4438 Previously discussed in https://github.com/onnx/onnx/pull/4389

operator

### System information N/A ### What is the problem that this feature solves? Contributors are currently recommended to keep tests for previous opsets when adding a new operator. This comes...

documentation
enhancement

### Details: - Add Python 3.10 to the list of supported Python versions - Remove Python 3.6 from the list of supported Python versions ### Tickets: - 86685

category: Python API
category: docs
do not merge

### Details: - Upgrade ONNX to 1.12.0 in Python requirements - Upgrade ONNX to 1.12.0 in `openvino/thirdparty/onnx` - Mark tests of unsupported ops/features/bugs as xfails - Upgrading ONNX fixed `OnnxBackendNodeModelTest.test_mul_uint8_cpu`...

WIP
do_not_review

name: A White Paper on Neural Network Quantization about: Quantization - from absolute basics to advanced concepts labels: quantization, miscellaneous **Topic:** *quantization* **Paper title:** *A White Paper on Neural Network...

Signed-off-by: p-wysocki ### Description PR is a part of an ongoing 1.13.0 release. It's based on https://github.com/onnx/onnx/pull/4209.

Signed-off-by: p-wysocki ### Description Add `LpPool-18`, which now has `ceil_mode` and `dilations` attributes. Deprecate `auto_pad`, like it is done in the case of [AveragePool](https://github.com/onnx/onnx/blob/main/docs/Operators.md#averagepool) and [MaxPool](https://github.com/onnx/onnx/blob/main/docs/Operators.md#maxpool). ### Motivation and Context...

operator

### Description Introduce Split-20, which performs uneven tensor split in torch's manner: instead of reducing only the size of the last dimension of the output, it reduces the last n...

operator
auto update doc