search

DanJun6737 / TransFace

[ICCV 2023] TransFace: Calibrating Transformer Training for Face Recognition from a Data-Centric Perspective
52 stars 7 forks source link

Vit-L config #7

Closed TranThanh96 closed 5 months ago

TranThanh96 commented 6 months ago

Can you provide VitL config to train glint360k?

DanJun6737 commented 5 months ago

Of course！ We will upload the corresponding configuration files in the next few days.

DanJun6737 commented 5 months ago

Hi~ @TranThanh96 We have uploaded the relevant configuration file.

TranThanh96 commented 4 months ago

@DanJun6737 can you provide your training log? my loss is a little bit high

DanJun6737 commented 3 months ago

Hi @TranThanh96 ~ Sorry, some training logs have been lost. Both the DPAP strategy and the EHSM strategy can cause the loss of the model to be higher in the early stages of training compared to a normal ViT model, but it will quickly decrease over time. Additionally, our configuration files are based on 8 V100 GPUs, so if a different number of GPUs is used, the corresponding configuration parameters need to be adjusted accordingly.