Closed jonasd4 closed 2 years ago
Base: 79.67% // Head: 79.57% // Decreases project coverage by -0.09%
:warning:
Coverage data is based on head (
ce61c2e
) compared to base (bfb7439
). Patch coverage: 75.55% of modified lines in pull request are covered.
:umbrella: View full report at Codecov.
:loudspeaker: Do you have feedback about the report comment? Let us know in this issue.
This resolves issue #85
Addresses Issue #85
Creates the new VisslExtractor which downloads models from the VISSL library and converts them to torchvision models.
adds models from the VISSL library which are based on the ResNet50 architecture
Available models:
simclr-rn50
,mocov2-rn50
,jigsaw-rn50
,rotnet-rn50
,swav-rn50
,pirl-rn50
add these models to the tests
More models with also different architectures will be added in the future.