Closed dxoigmn closed 1 year ago
It looks like there're merged changes in
mart/models/modular.py
. I will review again after you merge main into this.
Should be good to review.
Leave it for now because some modules still use it. Making everything tidier should be a different PR.
What does this PR do?
Right now
SequentialDict
only returns theoutput
module. This requires hardcoding a module with that name and often makes configuration more verbose than necessary. This PR removes that requirement and simplifies configuration by returning all outputs in the form of aDotDict
. This is a breaking change and also removesReturnKwargs
since it is unnecessary.This PR depends upon the following:
Type of change
Please check all relevant options.
Testing
Please describe the tests that you ran to verify your changes. Consider listing any relevant details of your test configuration.
pytest
CUDA_VISIBLE_DEVICES=0 python -m mart experiment=CIFAR10_CNN_Adv trainer=gpu trainer.precision=16
reports 70% (21 sec/epoch).CUDA_VISIBLE_DEVICES=0,1 python -m mart experiment=CIFAR10_CNN_Adv trainer=ddp trainer.precision=16 trainer.devices=2 model.optimizer.lr=0.2 trainer.max_steps=2925 datamodule.ims_per_batch=256 datamodule.world_size=2
reports 70% (14 sec/epoch).Before submitting
pre-commit run -a
command without errorsDid you have fun?
Make sure you had fun coding 🙃