* Add efficientnet backbones (B0-B3) * Remove `input_tensor` kwarg * Update README
Attach a file by drag & drop or click to upload
Pavel Yakubovskiy authored