AlibabaResearch / AlibabaResearch/AdvancedLiterateMachinery
VGT training memory requirement
Ouverte
- Langage dominant
- C++
- Étoiles
- 1.8k
- Forks
- 195
- Métriques de merge des PR
- Aucune PR mergée en 30 j
Description
I am able to train it with one 2080Ti with 12 GB of VRAM but I would have to resort to having a batch size of 1.
Training with a batch size of 4 and 4 2080Tis throws out of memory after some time even with mixed precision enabled.
Please let me know about the:
1) Normal memory requirements for training
2) Any ways to reduce the memory required.
Thanks very much in advance.
Guide de contribution
Aucun guide de contribution indexé pour ce dépôt
Évaluation
Cette issue n'a pas encore été évaluée.