EigenReporter and VINC algorithm #124
Merged
Conversation
This file contains bidirectional Unicode text that may be interpreted or compiled differently than what appears below. To review, open the file in an editor that reveals hidden Unicode characters.
Learn more about bidirectional Unicode characters
…ill equal CCS loss." string to be accurate
AlexTMallen
reviewed
Mar 13, 2023
AlexTMallen
reviewed
Mar 14, 2023
| num_heads: The number of reporter heads to fit. In other words, the number | ||
| of eigenvectors to compute from the VINC matrix. | ||
| """ | ||
|
|
||
| inv_weight: float = 5.0 | ||
| neg_cov_weight: float = 5.0 | ||
|
|
There was a problem hiding this comment.
I think it makes sense to add a var_weight as well, because we might want to set var_weight=0 (the confidence term).
There was a problem hiding this comment.
That's now resolved, right?
There was a problem hiding this comment.
actually it looks like it's not
There was a problem hiding this comment.
I didn't see it earlier for some reason
for more information, see https://pre-commit.ci
…to eigen-reporter
lauritowal
approved these changes
Mar 15, 2023
| LOSSES[name](logit0, logit1, coef) for name, coef in self.loss_dict.items() | ||
| ) | ||
| return assert_type(torch.Tensor, loss) | ||
|
|
||
| def reset_parameters(self): |
There was a problem hiding this comment.
@abstractmethod missing?
There was a problem hiding this comment.
It's not missing because EigenReporter doesn't need it
| torch.zeros(in_features, in_features, device=device, dtype=dtype), | ||
| ) | ||
| self.register_buffer( | ||
| "intracluster_cov", |
There was a problem hiding this comment.
@norabelrose @AlexTMallen I changed the name of this, since it was "intercluster_cov_M2" before... which was already registered in line 77. The fix prevents the error:
File "/mnt/ssd-1/nora/miniconda3/lib/python3.10/multiprocessing/pool.py", line 125, in worker
result = (True, func(*args, **kwds))
File "/mnt/ssd-1/notodai/walter/elk/elk/training/train.py", line 103, in train_reporter
reporter = EigenReporter(x0.shape[-1], cfg.net, device=device)
File "/mnt/ssd-1/notodai/walter/elk/elk/training/eigen_reporter.py", line 74, in __init__
self.intracluster_cov
File "/mnt/ssd-1/notodai/walter/elk/.venv/lib/python3.10/site-packages/torch/nn/modules/module.py", line 1269, in __getattr__
raise AttributeError("'{}' object has no attribute '{}'".format(
AttributeError: 'EigenReporter' object has no attribute 'intracluster_cov'
"""
The above exception was the direct cause of the following exception:
Traceback (most recent call last):
File "/mnt/ssd-1/notodai/walter/elk/.venv/bin/elk", line 8, in <module>
sys.exit(run())
File "/mnt/ssd-1/notodai/walter/elk/elk/__main__.py", line 57, in run
train(args.run, args.output)
File "/mnt/ssd-1/notodai/walter/elk/elk/training/train.py", line 195, in train
for i, *stats in tqdm(mapper(fn, layers), total=len(layers)):
File "/mnt/ssd-1/notodai/walter/elk/.venv/lib/python3.10/site-packages/tqdm/std.py", line 1195, in __iter__
for obj in iterable:
File "/mnt/ssd-1/nora/miniconda3/lib/python3.10/multiprocessing/pool.py", line 873, in next
raise value
AttributeError: 'EigenReporter' object has no attribute 'intracluster_cov'
AlexTMallen
approved these changes
Mar 15, 2023
Sign up for free
to join this conversation on GitHub.
Already have an account?
Sign in to comment
Add this suggestion to a batch that can be applied as a single commit.
This suggestion is invalid because no changes were made to the code.
Suggestions cannot be applied while the pull request is closed.
Suggestions cannot be applied while viewing a subset of changes.
Only one suggestion per line can be applied in a batch.
Add this suggestion to a batch that can be applied as a single commit.
Applying suggestions on deleted lines is not supported.
You must change the existing code in this line in order to create a valid suggestion.
Outdated suggestions cannot be applied.
This suggestion has been applied or marked resolved.
Suggestions cannot be applied from pending reviews.
Suggestions cannot be applied on multi-line comments.
Suggestions cannot be applied while the pull request is queued to merge.
Suggestion cannot be applied right now. Please check back later.
No description provided.