diff options
| author | Gustaf Rydholm <gustaf.rydholm@gmail.com> | 2022-09-13 21:56:59 +0200 |
|---|---|---|
| committer | Gustaf Rydholm <gustaf.rydholm@gmail.com> | 2022-09-13 21:56:59 +0200 |
| commit | 55a29c6c707c289449445c7f15964e9dbbdcf818 (patch) | |
| tree | ec3b766e39815161345928965dd300c1025a731d | |
| parent | 7db6e95f5ae404e276f22ef2c07e72285a33c490 (diff) | |
Update README
| -rw-r--r-- | README.md | 1 |
1 files changed, 1 insertions, 0 deletions
@@ -53,6 +53,7 @@ Ideas of mine that did not work unfortunately: * Efficientnet was apparently a terrible choice of an encoder - A ConvNext module heavily copied from lucidrains [x-unet](https://github.com/lucidrains/x-unet) + was incredibly much better at encoding the images to a better representation. * Use VQVAE to create pre-train a good latent representation - Tests with various compressions did not show any performance increase compared to training directly e2e, more like decrease to be honest |