New blog post! Take a look at some best practices for efficient CNN training, and find out how you can apply them easily with our Composer library:
#EfficientML
@MosaicML
Great read! Have you looked into anything like AIStore to offload some cpu costs for augmentations?
Assuming you can deploy none gpu nodes at much cheaper rate than v100/a100 VMs you should get the augmentation win without the slowdown.