This project mainly focuses on the baseline model for classification/segmentation/object detection, including resnet, resnext, densenet; deeplab, unet; faster-rcnn, yolo...
This is an in-progress project, if you are interested in it, thanks for starring it:)
-
In-implement metrics
-
friendly metrics with auto-grad and auto-log
-
ResNet Family & ResNext Family
-
DenseNet Family
-
Learning Rate Schdule and automatical finder
-
DeepLab Family (Need test)
-
Image loggers (Some issueses)