-
Notifications
You must be signed in to change notification settings - Fork 35
New issue
Have a question about this project? Sign up for a free GitHub account to open an issue and contact its maintainers and the community.
By clicking “Sign up for GitHub”, you agree to our terms of service and privacy statement. We’ll occasionally send you account related emails.
Already on GitHub? Sign in to your account
Automated GitHub Actions Test for gRPC Training #148
Conversation
There was a problem hiding this comment.
Choose a reason for hiding this comment
The reason will be displayed to describe this comment to others. Learn more.
Also, can you please fix the merge conflicts
.vscode/settings.json
Outdated
"python.analysis.typeCheckingMode": "strict", | ||
"sshfs.configs": [ |
There was a problem hiding this comment.
Choose a reason for hiding this comment
The reason will be displayed to describe this comment to others. Learn more.
can you remove this?
src/configs/sys_config.py
Outdated
@@ -158,16 +158,20 @@ def get_digit_five_support(num_users: int, domains: List[str] = DIGIT_FIVE): | |||
CIAR10_DPATH = "./datasets/imgs/cifar10/" | |||
|
|||
NUM_COLLABORATORS = 1 | |||
DUMP_DIR = "/mas/camera/Experiments/SONAR/abhi/" | |||
# DUMP_DIR = "../../../../../../../home/" | |||
DUMP_DIR = "./" |
There was a problem hiding this comment.
Choose a reason for hiding this comment
The reason will be displayed to describe this comment to others. Learn more.
Let's make the default dump directory as /tmp
The risk with keeping "./" as dump directory is that the dump_dir
might try to copy itself in dump_dir
kicking off a recursive cascade
src/configs/sys_config.py
Outdated
"num_collaborators": NUM_COLLABORATORS, | ||
"dset": CIFAR10_DSET, | ||
"dump_dir": DUMP_DIR, | ||
"dpath": CIAR10_DPATH, | ||
"seed": 32, | ||
# "seed": 32, |
There was a problem hiding this comment.
Choose a reason for hiding this comment
The reason will be displayed to describe this comment to others. Learn more.
make sure to not include any unnecessary comments
This enables the ability to automate testing of gRPC training whenever new changes are being made to the repo. You can change the configuration of the test using algo_config_test.py and sys_config_test.py (do not alter the DUMP_DIR). You can add/change the algorithms as well as number of rounds and samples for testing in GitHub actions. This test ensures that new code being pushed does not break existing working functionality. Changes in send_quorum have been made to ensure the tests work in a Linux environment.
NOTE: This PR includes code from the MPI PR.