deepinsight / deepinsight/insightface
The effect of VPL method
Open
- Dominant language
- Python
- Stars
- 29.7k
- Forks
- 6.1k
- PR merge metrics
- No merged PRs in 30d
Description
Hi ,
I had trained a FR model with arcFace_torch method in 8 GPUs, and the accuracy of it is LFW 99.85, CFP_FP 99.12, AGEDB_30 98.48;Simultaneously, I had train another model with VPL_arcface ( 'vpl': {'start_iters': 20000, 'allowed_delta': 80, 'lambda': 0.15, 'mode': 0, 'momentum': True}) in 4 GPUs, and the accuracy of it is LFW 99.85, CFP_FP 99.14, AGEDB_30 98.43. The VPL method don't seem to be working,I don't know the reason.
Contributor guide
No contributing guide indexed for this repository
Assessment
This issue has not been assessed yet.