search

HobbitLong / PyContrast

PyTorch implementation of Contrastive Learning methods
1.94k stars 186 forks source link

Can't reproduce results for detection weights rand init. 1x and 2x #34

Open JohannesTheo opened 3 years ago

JohannesTheo commented 3 years ago

Hey there, first of all a huge thank you for providing the pre-trained model, that's pretty cool :) One question regarding the detection weights:

I failed to reproduce the results for rand init. 1x and 2x (all other checkpoints work fine). From the log I saw that rand init. 1x and 2x are trained in "RGB" format with the respective mean and std. The configs will default to BGR and 1.0 std which results in a box/mask AP ~0. Strangely, even if I use RGB format, mean and std I only get box/mask AP of 0.257 / 0.235 (reported 32.8 / 29.9) for the 1x and 0.304 / 0.278 (reported 38.4 / 34.7) for the 2x model. The rand init. 6x models seems to be trained in BGR and works fine with the config.

Do I miss something here? Maybe these are not the final checkpoints? Thanks in advance

HobbitLong commented 3 years ago

Hey, sry for the late response.

For the rand init. experiments, I forgot all the details since it's long time ago. But giving that rand int. has nothing to do with pre-training, I would recommend you directly using Detectron2 here:

https://github.com/facebookresearch/detectron2