search

EricGuo5513 / momask-codes

Official implementation of "MoMask: Generative Masked Modeling of 3D Human Motions (CVPR2024)"
https://ericguo5513.github.io/momask/
MIT License
690 stars 56 forks source link

Will you help to verify the RVQAVE result on testset of humanml3D? #21

Open hdjsjyl opened 5 months ago

hdjsjyl commented 5 months ago

Hi Author, Would you help to verify whether the result I obtained is correct or not?

net_best_fid.tar final result, epoch -1 FID: 0.019, conf. 0.000 Diversity: 9.597, conf. 0.070 TOP1: 0.509, conf. 0.002, TOP2. 0.700, conf. 0.003, TOP3. 0.793, conf. 0.002 Matching: 3.009, conf. 0.007 MAE:0.029, conf.0.000

This is the result I obtained. Thanks

Murrol commented 1 month ago

Hi, sorry for the late response. The results look good. The minor deviation is normal.

HitBadTrap commented 1 week ago

Hi Author, Would you help to verify whether the result I obtained is correct or not?

net_best_fid.tar final result, epoch -1 FID: 0.019, conf. 0.000 Diversity: 9.597, conf. 0.070 TOP1: 0.509, conf. 0.002, TOP2. 0.700, conf. 0.003, TOP3. 0.793, conf. 0.002 Matching: 3.009, conf. 0.007 MAE:0.029, conf.0.000

This is the result I obtained. Thanks

Could you share your configuration files (including opt.txt, base_option.py, train_option.py, and vq_option.py) and commands? I cannot reproduce the results in the paper. @hdjsjyl

Thanks.