Skip to content
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.

Sign up for GitHub

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

Jump to bottom

[WIP] Method refactor #8

Open
wants to merge 70 commits into
base: v2
Choose a base branch
from
Open

[WIP] Method refactor #8

wants to merge 70 commits into from

Conversation

Lakshita110
Copy link
Collaborator

No description provided.

PatrickFeeney and others added 30 commits November 15, 2024 15:17
@PatrickFeeney
Method Refactoring Sketch
2a58506
@Lakshita110
Update training script with TensorBoard logging and early stopping, a…
6c201d1 
…nd introduce utility functions for model training
@Lakshita110
Refactor training function to include class weights and improve early…
66e4832 
… stopping mechanism
@nikkifalicov
Preliminary work on refactoring
edbc4b1
@PatrickFeeney
MethodWrapper Class
46ceac4
Refactor main
b6f5ba9
Remove unwanted file
bfb85d0
@Lakshita110
Formatting
ab6bab5
@Lakshita110
Refactor hyper_search and utils
89a4a13
@Lakshita110
Implement SupervisedMethod class and refactor training loop
2518a4b
@nikkifalicov
Working on integrating CheXpert
a8da6d3
@nikkifalicov
Updated config file to include IDRID and CheXpert stats
a135826
@nikkifalicov
Finished and tested get_dataloaders
6089cae
@PatrickFeeney
Add TODO Comments
e35d2f5
@Lakshita110
Formatting and function contracts
9ac28fd
@Lakshita110
Refactor hyper_search.py for consistent formatting and indentation
8446ced
@Lakshita110
Missing imports and code formatting improvements
30efb7d
@Lakshita110
Enhance evaluation metrics by adding AUPRC and AUROC calculations;
763bc04
Refactor main function according to ray tune tutorial
0d61c09
Refactor main function according to ray tune tutorial
1e6a3eb
@PatrickFeeney
Add Ray to Pipenv
0c4cda2
@PatrickFeeney
Enable Dataset and Method Configurations
3065b4a
@nikkifalicov
Fixed TODO comments
70a56ff
@nikkifalicov
Merge branch 'method_refactor_dataloaders' of https://github.com/tuft…
cf330a5 
…s-ml/SSL-vs-SSL-benchmark into method_refactor_dataloaders

Conflicts:
	src/dataset_csv.py
@nikkifalicov
Applied flake8
3f8b989
@nikkifalicov
Applied flake8 formatting
6cd67a7
@PatrickFeeney
Fix Unlabeled and Remove Duped Code
1c861eb
@PatrickFeeney
Remove Duplicated Function Definitions from Unlabeled Dataset
38b4c0e
@nikkifalicov
Update CSV dataloader naming
3973802
@nikkifalicov
Updated dataloader calling in hyper_search.py
5914f11
@PatrickFeeney PatrickFeeney mentioned this pull request Feb 6, 2025
Merged
Lakshita110 and others added 26 commits February 6, 2025 11:04
@Lakshita110
Adjust import paths, and refine argument parsing in training utilities
c7b74ec
@Lakshita110
Refactor import paths and enhance logging in hyper_search
78354e6
@Lakshita110
Merge pull request #9 from tufts-ml/method_refactor_integrate
280fd80 
Refactor training and optimization logic in hyper_search.py
@Lakshita110
Refactor import paths, update training code, and enhance experiment s…
444fa7e 
…cripts
@PatrickFeeney
dataset_config -> dataset_configs
2450b07
@PatrickFeeney
Replacement for Ray Hyperparameter Sampling
b83a4b5
@Lakshita110
Add torch-tb-profiler to development dependencies
567a8e1
@Lakshita110
Adjust experiment resource allocation and update environment variables
d8d67a3
@Lakshita110
Remove debug print statements and improve evaluation function documen…
acb8f47 
…tation
@Lakshita110
Add method to generate directory name from hyperparameters and improv…
d7799c1 
…e DataLoader configurations
@Lakshita110
Change SLURM partition from compute to preempt for resource optimization
2307f7f
@nikkifalicov
Fixed a naming issue
3d20704
@nikkifalicov
Fixed folder naming issue
7b70cba
@Lakshita110
Refactor training setup and evaluation functions for clarity and effi…
bece16a 
…ciency
@Lakshita110
Refactor evaluation metrics handling and streamline eval_model function
d80c8e9
@Lakshita110
Fix --use_pretrained
f45cfee
@Lakshita110
Update experiment configuration and refactor argument handling for cl…
ba3f3f7 
…arity
@Lakshita110
Enhance model initialization by freezing layers for pretrained models…
c29c35c 
… and ensuring the last layer is trainable
@Lakshita110
Update SLURM configuration and optimize evaluation model function for…
408aa99 
… better performance
@Lakshita110
Refactor model import bug
bb8af5b
@Lakshita110
Update experiment configuration
52aae38
@Lakshita110
Refactor optimizer settings to use dynamic learning rate and weight d…
4cd27e5 
…ecay from arguments;
@Lakshita110
Add freeze_backbone argument and update data transformations for pret…
332daa1 
…rained models
@PatrickFeeney
Argparse strs to bools
6f59d88
@Lakshita110
Add help descriptions for pretrained weights and freeze backbone options
d86e707
@Lakshita110
Add SLURM scripts for hyperparameter search and experiment launching
c5bc0ee
Sign up for free to join this conversation on GitHub. Already have an account? Sign in to comment
Reviewers
No reviews
Assignees
No one assigned
Labels
None yet
Projects
None yet
Milestone
No milestone
Development

Successfully merging this pull request may close these issues.
4 participants
@Lakshita110 @PatrickFeeney @nikkifalicov @abarancikova