Skip to content
New issue

Have a question about this project? Sign up for a free GitHub account to open an issue and contact its maintainers and the community.

By clicking “Sign up for GitHub”, you agree to our terms of service and privacy statement. We’ll occasionally send you account related emails.

Already on GitHub? Sign in to your account

Object Detection using TorchSharp #6605

Merged
merged 14 commits into from Apr 26, 2023

Conversation

michaelgsharp
Copy link
Member

Adds in Object Detection using TorchSharp with an AutoFormer model.

Main files to look at are the Object Detection Trainer, Utility methods, and the refactors to TextClassification and Sentence Similarity.

We also pulled in local/internal copies of System.Range/Index to make the future model conversion process much easier.

Copy link
Member

@ericstj ericstj left a comment

Choose a reason for hiding this comment

The reason will be displayed to describe this comment to others. Learn more.

Small comments about index/range

src/Microsoft.ML.TorchSharp/Utils/Range.cs Outdated Show resolved Hide resolved
src/Microsoft.ML.TorchSharp/Utils/Index.cs Outdated Show resolved Hide resolved
src/Microsoft.ML.TorchSharp/Utils/Range.cs Outdated Show resolved Hide resolved
@codecov
Copy link

codecov bot commented Mar 24, 2023

Codecov Report

Merging #6605 (792407d) into main (d239fda) will increase coverage by 0.13%.
The diff coverage is 86.51%.

Additional details and impacted files
@@            Coverage Diff             @@
##             main    #6605      +/-   ##
==========================================
+ Coverage   68.39%   68.52%   +0.13%     
==========================================
  Files        1174     1200      +26     
  Lines      248047   250288    +2241     
  Branches    25909    26093     +184     
==========================================
+ Hits       169647   171521    +1874     
- Misses      71626    71932     +306     
- Partials     6774     6835      +61     
Flag Coverage Δ
Debug 68.52% <86.51%> (+0.13%) ⬆️
production 63.00% <87.36%> (+0.22%) ⬆️
test 88.83% <63.95%> (-0.06%) ⬇️

Flags with carried forward coverage won't be shown. Click here to find out more.

Impacted Files Coverage Δ
...icrosoft.ML.TorchSharp/NasBert/Models/BaseModel.cs 66.66% <ø> (ø)
...osoft.ML.TorchSharp/NasBert/Models/NasBertModel.cs 87.75% <ø> (ø)
...Microsoft.ML.TorchSharp/NasBert/Optimizers/Adam.cs 100.00% <ø> (ø)
src/Microsoft.ML.TorchSharp/Utils/HashHelpers.cs 0.00% <0.00%> (ø)
test/Microsoft.ML.Tests/TextClassificationTests.cs 86.01% <0.00%> (-8.75%) ⬇️
src/Microsoft.ML.TorchSharp/Utils/Index.cs 29.41% <29.41%> (ø)
src/Microsoft.ML.TorchSharp/Utils/Range.cs 31.57% <31.57%> (ø)
src/Microsoft.ML.TorchSharp/TorchSharpCatalog.cs 77.77% <71.42%> (-22.23%) ⬇️
src/Microsoft.ML.TorchSharp/Utils/ImageUtils.cs 83.56% <83.56%> (ø)
...c/Microsoft.ML.TorchSharp/TorchSharpBaseTrainer.cs 83.75% <83.75%> (ø)
... and 24 more

... and 34 files with indirect coverage changes

@ericstj ericstj changed the title Obj Detection using TorchSharp Object Detection using TorchSharp Apr 3, 2023
Copy link
Member

@ericstj ericstj left a comment

Choose a reason for hiding this comment

The reason will be displayed to describe this comment to others. Learn more.

None of my feedback is blocking, it's all just cleanup. I'd like a review from other area experts before merging though.

…s.cs

Co-authored-by: Jake <31937616+JakeRadMSFT@users.noreply.github.com>
Copy link
Contributor

@luisquintanilla luisquintanilla left a comment

Choose a reason for hiding this comment

The reason will be displayed to describe this comment to others. Learn more.

LGTM. Thanks everyone for the work involved in making this happen.

:shipit:

@michaelgsharp michaelgsharp merged commit cb2e495 into dotnet:main Apr 26, 2023
22 checks passed
@dotnet dotnet locked as resolved and limited conversation to collaborators May 26, 2023
Sign up for free to subscribe to this conversation on GitHub. Already have an account? Sign in.
Labels
None yet
Projects
None yet
Development

Successfully merging this pull request may close these issues.

None yet

6 participants