Recent Releases of onnx2torch

onnx2torch -

feat: squeeze-21 feat: ArgMax and ArgMin

Thanks to @rosemanmeilof and @ajstarna for their contributions.

:cake:

- Python
Published by senysenyseny16 over 1 year ago

onnx2torch -

feat: isinf, isnan and nonzero operations fix: reduce keepdim type (int -> bool)

Thanks to @karalinkw and @francis2tm for their contributions.

:cake:

- Python
Published by senysenyseny16 almost 2 years ago

onnx2torch -

fix: div return type for integer arguments docs: missing [import onnx] in main example

Thanks @seriousran for his contribution.

:cake:

- Python
Published by senysenyseny16 over 2 years ago

onnx2torch -

fix: export to onnx ConstantOfShape in case of bool value attribute

:cake:

- Python
Published by senysenyseny16 over 2 years ago

onnx2torch -

feat: DepthToSpace

Thanks @sushanthr for his contribution.

:cake:

- Python
Published by senysenyseny16 over 2 years ago

onnx2torch - v1.5.10

chore: fix broken build (missing submodules)

:cake:

- Python
Published by senysenyseny16 over 2 years ago

onnx2torch -

feat: LayerNormalization

:cake:

Thanks @JohnMasoner for his contribution.

- Python
Published by senysenyseny16 over 2 years ago

onnx2torch -

fix: op Range export to onnx (keeping dynamic inputs)

:cake:

- Python
Published by senysenyseny16 over 2 years ago

onnx2torch -

``` feat: Mod feat: EyeLike feat: human-readable onnx node naming

fix: gemm weight swapped dimension ``` :cake:

Thanks @JohnMasoner and @jan-haug for their contributions.

- Python
Published by senysenyseny16 over 2 years ago

onnx2torch -

``` feat: InstanceNormalization feat: LeakyRelu opset 16 feat: RoiAlign for opset 16 feat: ScatterND for opset 16 (reduction none only) feat: GatherND (only for parameter batch_dims=0)

fix: lambda not found error when using torch.save fix: upcast indices to int64 fix: GEMM dim typo (no parameters for nn.Linear)

docs: workaround for old opset versions ```

:cake:

Thanks @SuperSecureHuman and @JohnMasoner for their contributions.

- Python
Published by senysenyseny16 about 3 years ago

onnx2torch - v1.5.4

Minor release to support torch 1.13 export.

- Python
Published by irkjero over 3 years ago

onnx2torch - v1.5.3

FIXED: - Gemm operation updated according onnx specs

- Python
Published by irkjero over 3 years ago

onnx2torch - v1.5.2

Support for new OPS :rocket: :

Dropout
LRN

Thanks @yan12125 and @monicathieu for their contributions.

- Python
Published by senysenyseny16 over 3 years ago

onnx2torch - v1.5.1

Include requirements.txt and exclude tests for sdist

- Python
Published by irkjero over 3 years ago

onnx2torch - v1.5.0

You've been waiting, you've asked... And here it is a new release(1.5.0)!

In this release: - Fixed shape inference of big models (now you can load your lovely big language model, like gptj6B); - Fixed BarchNorm converter according onnx specs; - Added asymmetric padding for both Conv and MaxPool.

Thank you for your discussions and issues! We always ready to help you!

- Python
Published by irkjero over 3 years ago

onnx2torch - v1.4.1

Minor updates: * Fix prelu according onnx specs, * Fix Global average pooling.

- Python
Published by irkjero over 3 years ago

onnx2torch - v1.4.0

:rocket: New day - new release! :rocket: New features in release 1.4.0: - Added operations: - Einsum, - Reciprocal, - Neg, - Prelu, - Mean, - Min, - Max, - CumSum, - Sum.

Fixes:
- GlobalAveragePool compatible with PyTorch FX QAT,
- Compatibility with torch 1.12,
- Allow to ignore bs+ch dimensions for size input in Resize.
- Check roi only for tfcropand_resize mode.
Code style and workflows:
- Pre-commit hook,
- Black code-style now is used in project,
- Linter and formatter with precommit hook,
- Code style checks for PRs in Github actions.

We thank all contributors for their work! Together we will make onnx2torch great! :star: 10^6

- Python
Published by irkjero over 3 years ago

onnx2torch - v1.3.0

New features in release 1.3.0:

Add Pad operation
Add Hardswish, Celu, Elu, Selu, Softplus, Softsign functions
Add GatherElements operation
Add new model tests

Fixes:

Fix ConvTranspose operation

- Python
Published by niobeus almost 4 years ago

onnx2torch - v1.2.5

Hi! New release 1.2.5. In this release we add: - ConvTranspose

And fix: - RoiAlign export

- Python
Published by irkjero almost 4 years ago

onnx2torch - v1.2.0

New features in release 1.2.0: - Support VIT, SWIN, FasterRcnn models; - add RoiAlign operation; - Add Floor, Ceil, Round and trigonometric functions; - New model tests for classification, detection and segmentation;

Fixes: - Fix ScatterND and its speed; - Fix names of plaseholders; - Fix optional arguments if last arguments are not passed; - Fix checking of empty inputs; - Fix avg pool operation;

- Python
Published by irkjero almost 4 years ago

onnx2torch - v1.1.0

New features in release 1.1.0:

Support of optional input arguments;

Fixed dynamic axes for squeeze and unsqueeze.

Also added new operations listed below:

Resize Reduces operations (ReduceL1, ReduceL2, ReduceLogSum, ReduceLogSumExp, ReduceMax, ReduceMean, ReduceMin, ReduceProd, ReduceSum, ReduceSumSquare) Logical Operaations (Or, Xor, And, Not) HardSigmoid, LeakyRelu Pow, Sqrt

- Python
Published by irkjero about 4 years ago

onnx2torch - v1.0.0

Supported models: * Resnet50 * ssdlite (backbone MobilenetV2)

- Python
Published by ivkalgin about 4 years ago