-
Notifications
You must be signed in to change notification settings - Fork 18
Issues: mmcdermott/EventStreamGPT
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
Author
Label
Projects
Milestones
Assignee
Sort
Issues list
Event type vocabulary should be filtered to UNKs too.
enhancement
New feature or request
#43
opened Jul 24, 2023 by
mmcdermott
ID_map should be saved in dataset construction
enhancement
New feature or request
#8
opened May 5, 2023 by
mmcdermott
Should enable retrieval metrics for classification as well
enhancement
New feature or request
#9
opened May 7, 2023 by
mmcdermott
Default model should be simplified and use the built-in PyTorch transformer layers (though still support structured attention)
refactor
Clean up code, testing interface, or API.
#10
opened May 7, 2023 by
mmcdermott
Torch.compile doesn't work for some reason.
bug
Something isn't working
#11
opened May 7, 2023 by
mmcdermott
CATEGORICAL_ONLY v. NUMERICAL_ONLY v. CATEGORICAL_AND_NUMERICAL is confusing.
refactor
Clean up code, testing interface, or API.
#13
opened May 9, 2023 by
mmcdermott
Dataset storage to disk should not rely on pickle.
enhancement
New feature or request
#14
opened May 10, 2023 by
mmcdermott
Test coverage needs to be expanded and interface simplified
refactor
Clean up code, testing interface, or API.
#15
opened May 10, 2023 by
mmcdermott
Upgrade to python 3.11 and eliminate the custom StrEnum class.
refactor
Clean up code, testing interface, or API.
#23
opened May 30, 2023 by
mmcdermott
Possibly replace Clean up code, testing interface, or API.
JSONableMixin
with OmegaConf
throughout.
refactor
#24
opened May 30, 2023 by
mmcdermott
Should allow use of NLL-Beta
enhancement
New feature or request
#26
opened Jun 13, 2023 by
mmcdermott
Benchmark results over standard EHR tasks like diagnosis prediction
enhancement
New feature or request
#41
opened Jul 21, 2023 by
avani17101
Warning during multi-gpu training re metrics and Something isn't working
sync_dist
bug
#5
opened May 1, 2023 by
mmcdermott
Vocabulary serialization into configs should not include observation frequencies.
#42
opened Jul 24, 2023 by
mmcdermott
Metrics in weights and biases should use New feature or request
good first issue
Good for newcomers
/
not _
to separate name parts.
enhancement
#60
opened Aug 24, 2023 by
mmcdermott
Modeling Configs Should be Re-arranged for Clarity, Better Defaults, and Ease-of-use
enhancement
New feature or request
#61
opened Aug 25, 2023 by
mmcdermott
DataLoader with num_workers > 0 increases memory consumption over time
bug
Something isn't working
resource constraints
For when the system is too resource hungry
#73
opened Oct 19, 2023 by
juancq
PytorchDataset doesn't have a way to extract subject_ids, start_time, and end_time all together
enhancement
New feature or request
#75
opened Oct 19, 2023 by
pargaw
Should be possible to resume pre-training and evaluate a pre-trained model post-hoc
enhancement
New feature or request
#91
opened Jan 2, 2024 by
mmcdermott
May be able to make temporal loss more stable by directly computing log-prob from pre-activation function outputs
enhancement
New feature or request
#93
opened Feb 19, 2024 by
mmcdermott
Getting started steps
documentation
Improvements or additions to documentation
enhancement
New feature or request
#96
opened Apr 17, 2024 by
rvandewater
Nested ragged tensors integration does not respect
seq_padding_side
#104
opened Apr 22, 2024 by
mmcdermott
Could not override 'config.task_specific_params.pooling_method' when lauching wandb agent
#108
opened May 24, 2024 by
Rhett-Ying
hydra.utils.instantiate(cfg,_convert_,"object") not working in finetune.py line 26
#123
opened Feb 13, 2025 by
filipols
Previous Next
ProTip!
no:milestone will show everything without a milestone.