zimenglan comments

Results 111 comments of


                                            zimenglan

My results are lower than yours.

> Dear @zimenglan-sysu-512, > > Yes, setting data.sampler.train.sample_ratio to `[2, 4]` or `[3, 3]` should work well. The first number represents the number of labeled samples, while the second number...

My results are lower than yours.

> Dear @zimenglan-sysu-512 The results you provided are amazing. It would be very helpful if you could share the config and checkpoint. This experiment could be extremely useful for people...

My results are lower than yours.

hi @Adamdad why in the fulldata config, the backbone r50 freeze the BN layer, but the 10% config does not ?

微信群的二维码过期了

Please provide pre-trained models

+1. also can release the results like top1?

2D VQA and Image-Text Retrieval

hi @ziqipang any update?

How to disable flash_attention

+1, @OmkarThawakar did u solve it? thanks

how to train

想学习AnimateAnyone快加学习交流群

过期了，亲

why filter out samples that have less than 3 objects?

oh you're right. by the weight, why set `weight_loss_cap` to zero?