Skip to content
This repository has been archived by the owner on Feb 7, 2023. It is now read-only.

Issues: onnx/onnx-coreml

Clear current search query, filters, and sorts
51 Open 119 Closed
51 Open 119 Closed
Author
Filter by author
Loading
Label
Filter by label
Loading
Use alt + click/return to exclude labels
or ⇧ + click/return for logical OR
Projects
Filter by project
Loading
Milestones
Filter by milestone
Loading
Assignee
Filter by who’s assigned
Sort
Sort by
Newest Oldest Most commented Least commented Recently updated Least recently updated Best match
Most reactions
👍 👎 😄 🎉 😕 ❤️ 🚀 👀

Issues list

Log accurary with CoreML3 needs discussion Collaboration needed to clarify prior to completing (status)
#430 opened Jul 29, 2019 by iiSeymour updated Sep 5, 2019
1
dilated convolution does not support weight as input tensor after convert to CoreML needs discussion Collaboration needed to clarify prior to completing (status) question Response providing clarification needed. Will not be assigned to a release. (type)
#462 opened Sep 3, 2019 by edis219 updated Sep 6, 2019
2
ValueError: Unsupported ONNX attribute: name: "body" clarification Issue resolution required comment of why and/or issue arose and what has been done to resolve it
#443 opened Aug 10, 2019 by zlinwei updated Oct 8, 2019
1
NotImplementedError: Unsupported ONNX ops of type: FC clarification Issue resolution required comment of why and/or issue arose and what has been done to resolve it
#398 opened Feb 23, 2019 by jjpd777 updated Oct 8, 2019
2
Question about is_bgr flag clarification Issue resolution required comment of why and/or issue arose and what has been done to resolve it repro needed More information or ability to reproduce the issue is needed (status)
#406 opened May 9, 2019 by lymanblue updated Oct 8, 2019
5
Check the ONNX version number and error out if not supported by the converter 1.1 A part of release 1.1 (release) enhancement Improvement to existing functionality (type) good first issue Issue for a contributor to self-assign and complete help wanted Available to be contributed to by any member. triaged Reviewed and examined, release as been assigned if applicable (status)
#322 opened Jul 9, 2018 by aseemw updated Oct 9, 2019
Reproduce pytorch output on iPhone 6s question Response providing clarification needed. Will not be assigned to a release. (type)
#486 opened Oct 6, 2019 by mirth updated Oct 14, 2019
1
Segfaulting tests with Catalina blocked Dependent on other factors (type) bug Unexpected behaviour that should be corrected (type) investigation Being looked into in order to complete (status)
#426 opened Jul 25, 2019 by iiSeymour updated Oct 16, 2019
2
Padding ops is not working as expected for 3D tensors. bug Unexpected behaviour that should be corrected (type)
#498 opened Oct 19, 2019 by marcoleewow updated Oct 19, 2019
gather + matmul ops duplicates data to 2nd axis. bug Unexpected behaviour that should be corrected (type)
#500 opened Oct 21, 2019 by marcoleewow updated Oct 21, 2019
_add_conv_like_op with convolution on 3d tensors produces ambiguous names for expanded inputs and outputs bug Unexpected behaviour that should be corrected (type)
#502 opened Oct 24, 2019 by richard-vogl updated Oct 24, 2019
1
Model works with target_ios='12' but not target_ios='13' bug Unexpected behaviour that should be corrected (type)
#499 opened Oct 21, 2019 by SumatPurewal updated Nov 15, 2019
1
Ops required for PyTorch Faster-RCNN: Range, NonMaxSuppression, and ScatterElements feature request Functionality does not currently exist, would need to be created as a new feature (type)
#508 opened Nov 18, 2019 by ccbrown updated Nov 18, 2019
Travis CI broken bug Unexpected behaviour that should be corrected (type)
#513 opened Dec 7, 2019 by bhushan23 updated Dec 9, 2019
1
Support Path from pathlib feature request Functionality does not currently exist, would need to be created as a new feature (type) good first issue Issue for a contributor to self-assign and complete help wanted Available to be contributed to by any member.
#516 opened Dec 10, 2019 by gunjunlee updated Dec 11, 2019
Issue converting Slice op when comes after Expand op bug Unexpected behaviour that should be corrected (type)
#520 opened Dec 16, 2019 by dkossnick-figma updated Jan 3, 2020
1
output_feature has no shape when output has len(shape)==2 question Response providing clarification needed. Will not be assigned to a release. (type)
#501 opened Oct 23, 2019 by yuvdar updated Jan 14, 2020
4
convert to CoreML and then get wrong predictions bug Unexpected behaviour that should be corrected (type)
#526 opened Dec 26, 2019 by MarcSteven updated Feb 15, 2020
5
convert to onnx and then convert to CoreML , but the prediction is wrong bug Unexpected behaviour that should be corrected (type)
#556 opened Mar 17, 2020 by douli9862 updated Mar 17, 2020
1
Add support for BatchNorm2d (and more options for Slice) feature request Functionality does not currently exist, would need to be created as a new feature (type)
#559 opened Mar 29, 2020 by jbmaxwell updated Mar 29, 2020
1
Using PixelShuffle operation causes error of computing minimum sequence length bug Unexpected behaviour that should be corrected (type)
#560 opened Apr 13, 2020 by oavrahami updated Apr 13, 2020
Error in neural network compiler computing minimum sequence length for the model question Response providing clarification needed. Will not be assigned to a release. (type)
#533 opened Jan 18, 2020 by Penelope1003 updated Apr 13, 2020
1
ONNX dynamic_axes cause compile time errors in final CoreML model bug Unexpected behaviour that should be corrected (type)
#565 opened Apr 17, 2020 by BorisKourt updated Apr 17, 2020
2
IndexError: list assignment index out of range bug Unexpected behaviour that should be corrected (type)
#550 opened Feb 24, 2020 by dzyjjpy updated Apr 18, 2020
2
KeyError: '498' W_shape = graph.shape_dict[W_name] awaiting response Please respond to this issue to provide further clarification (status) bug Unexpected behaviour that should be corrected (type)
#561 opened Apr 14, 2020 by fcqfcq updated Apr 20, 2020
1
ProTip! no:milestone will show everything without a milestone.