Qingyun

I'm working on multi-task learning too, I wonder if your experiment is supported by literature. Could you please share some ? @victoic

The work will be done by other contributor.

> Hi, thanks very much for your code. May I ask if you have met this problem? > > When did this happen ? at the beginning or after a...

@czczup I haven't fully trained yet, but I have tested 1000 iters training (2 GPU dist_train). I didn't met it. Could you please provide more information? Thanks for your attention.

@czczup Hi, the newest config file contains other incompatibility problems. ```python3 norm_cfg=dict(type='FrozenBatchNorm2d'), # should use BN instead conv_cfg=dict(type='Conv2d', bias=True), # should be deleted ``` I made modification in my local...

@czczup I'm working on experiments for aligning training results, there is still a few modification which i have not committed. You can comment the FrozenBatchNorm2d, use BN instead. I modified...

@czczup The label_embedding is used in DINOHead.forward_train() and DnQueryGenerator.forward(). I haven't found the bug yet, and maybe a few days later it'll be found. I'm pushing full speed ahead with...

@czczup Hi, the original repo has a sh file to launch training, [in which](https://github.com/IDEACVR/DINO/blob/24f3567e162c75f0323cf4a1ed2d5bf6e36bee52/scripts/DINO_train.sh#L6) the box_noise_scale is set 1. The description is at the last paragraph of the D.3 (Supplementary...

@czczup Thanks! Maybe it's the special case when the current sample has no target, I'll have a check.

@czczup Hi, I'm waiting for the queue of lab's slurm cluster, so can't experiment in time. Could you have a test of setting `filter_empty_gt=True`? ```Python data = dict( samples_per_gpu=2, workers_per_gpu=2,...