Small Cellpose improvements#44
Merged
nilsmechtel merged 33 commits intoaicell-lab:mainfrom Feb 16, 2026
Merged
Conversation
hugokallander
changed the title
oeway
reviewed
Dec 16, 2025
oeway
reviewed
Dec 16, 2025
oeway
reviewed
Dec 18, 2025
oeway
reviewed
Dec 18, 2025
Contributor
There was a problem hiding this comment.
I would maybe swap the name, and keep manifest.yaml for the prod, and manifest-test.yaml for dev.
oeway
reviewed
Dec 18, 2025
oeway
reviewed
Dec 18, 2025
Contributor
oeway
left a comment
oeway
left a comment
There was a problem hiding this comment.
I did a quick review, the changes looks good to me! I assume you have already tested it?
I made a few comments, would be nice if you can take a stab on them. Thanks!
oeway
reviewed
Dec 18, 2025
Resolved conflict in update_status() keeping main's read-modify-write pattern while adding PR's test_metrics parameter. Co-Authored-By: Claude Opus 4.6 <noreply@anthropic.com>
- Fix _predict_and_encode() missing output=mask_np (broke inference) - Preserve update_status() read-modify-write pattern with test_metrics - Restore dataset_artifact_id in SessionStatus and start_training() - Restore training history inheritance including test_metrics - Fix SessionStatusWithId for Ray cloudpickle compatibility - Deduplicate metrics: extract _binary_threshold helper to module level - Remove all cast() calls (19 -> 0), reduce # type: ignore (90 -> 22) - Delete manifest-test.yaml (referenced non-existent class) - Restore deploy_cellpose.py with argparse and production defaults - Fix redeploy/test scripts to default to production target - Add wait-for-services logic in deploy/redeploy scripts - Add py.typed PEP 561 marker - Update README: validation metrics docs, CLI args, example scripts Deployed and tested: training (3 epochs), inference, dataset_artifact_id all verified on https://hypha.aicell.io. Co-Authored-By: Claude Opus 4.6 <noreply@anthropic.com>
Contributor
|
FYI: hugokallander#1 |
…cs, consistent sentinels - Add validation_interval parameter to start_training() API (default: every 10 epochs) - Add InstanceMetrics (AP@0.5/0.75/0.9) computed at end of training via cellpose.metrics - Change test_losses sentinel from 0.0 to None for non-evaluated epochs - Clarify ValidationMetrics docstring: pixel-level binary foreground, NOT instance segmentation - Update consumer scripts (test_service.py, test_callbacks.py) to filter with `is not None` - Add --validation-interval CLI arg to test_service.py - Update README with new parameters, metrics docs, and filter patterns - Deployed and tested end-to-end: 3 epochs, validation_interval=1, instance metrics computed Co-Authored-By: Claude Opus 4.6 <noreply@anthropic.com>
…x CLI args - Add stop_training() and list_training_sessions() API docs - Remove non-existent list_pretrained_models() from README - Add missing params: min_train_masks, return_flows, validation_interval in examples - Add missing status fields: current_batch, total_batches, exported_artifact_id, model_modified - Fix status_type values: add "stopped" and "unknown" - Fix export_model() model_name as optional - Fix save_application.py CLI: --server-url not --server_url - Fix list_training_sessions() to include dataset_artifact_id and instance_metrics - Add "Validation Metrics" to Features list Co-Authored-By: Claude Opus 4.6 <noreply@anthropic.com>
Contributor
|
@hugokallander could you resolve this and merge, then we deploy the merged version? |
Member
Author
|
Updated this PR to include changes from PR #46 and local fixes. This branch now contains the consolidated work for Cellpose finetuning improvements. |
nilsmechtel
approved these changes
Feb 16, 2026
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
Sign up for free
to join this conversation on GitHub.
Already have an account?
Sign in to comment
3 participants
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.
Uh oh!
There was an error while loading. Please reload this page.