PyTorch ver 1.0: keras 3.+ with pytorch backend for nn_models.py and … #1515
Conversation
This file contains hidden or 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
added 2 commits
August 5, 2025 15:41
…associated files, pytorch replaces tensorflow for Volpy's mrcnn folder, parts of cnmf, and components evaluation. Additional testing files implemented and added, and demo folders updated. Lastly, the new pytorch models (weights-layers converted from tensorflow) have been added
…ents_evaluation.py
pgunn
reviewed
Aug 6, 2025
| ]) | ||
| display_table(table) | ||
| fname = fnames[ind] | ||
| return np.load(dir + fname)['img'], dir + fname |
pgunn
reviewed
Aug 13, 2025
caiman/utils/utils.py
Outdated
Comment on lines
16
to
17
| import inspect | ||
| import h5py |
There was a problem hiding this comment.
These two somehow ended up out of alphabetical order
pgunn
reviewed
Aug 13, 2025
caiman/components_evaluation.py
Outdated
| print(f"USING MODEL (tensorflow API): {model_file}") | ||
| loaded_model = caiman.utils.utils.load_graph(model_file) | ||
| raise FileNotFoundError(f"File for requested model {model_name} not found") | ||
| print(f"Using model: {model_file}") |
There was a problem hiding this comment.
This probably should be using the logger rather than print
| @@ -2009,8 +2020,7 @@ def get_candidate_components(sv, dims, Yres_buf, min_num_trial=3, gSig=(5, 5), | |||
| gHalf=(5, 5), sniper_mode=True, rval_thr=0.85, | |||
There was a problem hiding this comment.
I know it's not necessarily code you added, but as I assume you've had to understand this function to rework it, any docs you could add to code-dense parts of this function that might help with understanding would be really useful
There was a problem hiding this comment.
Sounds good. I have some notes, so I could add them in.
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.
…associated files, pytorch replaces tensorflow for Volpy's mrcnn folder, parts of cnmf, and components evaluation. Additional testing files implemented and added, and demo folders updated. Lastly, the new pytorch models (weights-layers converted from tensorflow) have been added
Description
PyTorch dependencies are added, Keras is updated to have PyTorch backend, and Tensorflow dependencies are removed. Models are updated into the
.ptformat and apytorch_model_arch.pyfile is added which contains the model architecture for these two files. The Ring CNN model (which usescaiman.utils.nn_models.pyandcaiman.source_extraction.cnmf.online_cmf.pyhas been updated for compatibility with torch and keras 3.+. Volpy's MRCNN module (and its respective files) have been updated with pytorch training, validation, and inferencing. Demo notebooks have been updated given these new changes.test_pytorch.pyis added to test the pytorch dependency withtest_tensorflow.pyremoved, whiletest_mrcnn.pyhas been updated to match the Volpy revisions.Fixes # (issue)
Dependency changes with revisions to Volpy, component evaluations, and Ring CNN.
Please delete options that are not relevant.
Branching
Has your PR been tested?
Yes, it has been tested.
If you're fixing a bug or introducing a new feature it is recommended you run the tests by typing
caimanmanager testand
caimanmanager demotestprior to submitting your pull request.
Please describe any additional tests that you ran to verify your changes. If they are fast you can also
include them in the folder 'caiman/tests/
and name themtest_***.py` so they can be included in our lists of tests.