Updated code (contrastive learning) #130
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
ziw-liu
reviewed
Aug 8, 2024
ziw-liu
reviewed
Aug 8, 2024
ziw-liu
reviewed
Aug 8, 2024
ziw-liu
reviewed
Aug 8, 2024
| self.tracks = self._filter_tracks(tracks_tables) | ||
| self.tracks = ( | ||
| self._specific_cells(self.tracks) if self.predict_cells else self.tracks | ||
| ) | ||
|
|
||
| def _filter_tracks(self, tracks_tables: list[pd.DataFrame]) -> pd.DataFrame: |
There was a problem hiding this comment.
@ziw-liu please edit docstrings related to processing of tracking info for accuracy.
| channel_wise=False, | ||
| dtype=None, | ||
| allow_missing_keys=False | ||
| ), | ||
| ] | ||
|
|
||
| def main(hparams): |
There was a problem hiding this comment.
Let's clean up the predict.py in the next PR
Comment on lines
+588
to
+590
| features_output_path: str = "", | ||
| projections_output_path: str = "", | ||
| metadata_output_path: str = "", |
There was a problem hiding this comment.
Looks like these three paths are needed to save predictions. I suggest renaming these three to predict_output_path and assign these three variables a consistent names, predict_output_path + features.npy, projecttions.npy, meta.csv ...
| trainer.fit(model, datamodule=data_module) | ||
|
|
||
| # # Validate the model | ||
| trainer.validate(model, datamodule=data_module) | ||
|
|
||
| # # Test the model | ||
| # trainer.test(model, datamodule=data_module) | ||
|
|
||
| # Argument parser for command-line options |
There was a problem hiding this comment.
@alishbaimran In the next PR, this training script should be turned into a config for use with lightning CLI and then deleted.
There was a problem hiding this comment.
I am merging this PR to unblock the rest of the improvements in model construction. Let's revisit comments in this PR to:
- delete applications/contrastive_phenotyping/training_script.py
- clean up predict.py.
- document key methods in viscy.data.triplet.py
- clean up excess input parameters to viscy.light.engine.ContrastiveModule
mattersoflight
approved these changes
Aug 9, 2024
mattersoflight
approved these changes
Aug 9, 2024
edyoshikun
pushed a commit
that referenced
this pull request
Aug 15, 2024
* updated prediction code to test specific cells * updated kernel and stride * updated kernel and stride * new training script and predict takes in output path as parameter * passing ci requirements * fixed error for output path * fixed error for output path * fixed formatting * format resolved * docstrings for methods that parse tracks --------- Co-authored-by: Shalin Mehta <[email protected]>
edyoshikun
pushed a commit
that referenced
this pull request
Aug 15, 2024
* updated prediction code to test specific cells * updated kernel and stride * updated kernel and stride * new training script and predict takes in output path as parameter * passing ci requirements * fixed error for output path * fixed error for output path * fixed formatting * format resolved * docstrings for methods that parse tracks --------- Co-authored-by: Shalin Mehta <[email protected]>
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.
I made the following changes:
—> Updated triplet code such that during prediction evaluation can be done for specific pre-defined cells.
—> Changed stride to be the same as stem kernel size which was also changed to (5, 4, 4).
—> Output file name is configurable for the prediction.
—> Updated training_script.py and predict.py
To-do:
—> Should be merged before any other branches merged into main.