This project aims to segment single-cells within diseased NF1 organoids by developing a deep-learning segmentation model.
Training results are logged in MLflow and are not currently available in the github. Each model's code is tracked within git's commit comment history. A trained model's code can be identified by a singular commit. Model commits can be identified by the commit message, which indicate the model's Experiment ID, Run ID, and possibly the model's codename in MLflow. These model commits may be located in main or other branches.
Each model is trained by executing train.py from the MLproject file so the results can be logged to MLflow. The paths inside train.py are needed to create the dataset for training the model.
All models will be trained with Experiment ID: 310992065458859481
These are the models developed in reverse chronological order:
Commit: 25f5b66e2b56dbf9c489645f6eb56610120c4ad6
Codename: luxuriant-shad-666
- Architecture: UNet Generator
- Task: One-to-One slice segmentation mask prediction
- QC / Filtering: Does not perform any QC or filtering of images or slices
- Data: Trained and evaluated on all slices of patient NF0016
- Preprocessing:
- Each input slice is normalized
- Each input is padded to preserve dimensionality (height and width divisible by 16)
Commit: b9c5ab4af0ef5db82f3575988d3e9eaeb891bd98
Codename: casual-stork-830
Important!!! The computer crashed during this run due to a device-side CUDA error
Note: The first two sub-runs completed without error
- Architecture: UNet Generator
- Task: One-to-One slice segmentation mask prediction
- QC / Filtering: Does not perform any QC or filtering of images or slices
- Data: Trained and evaluated on all slices of patient NF0016
- Preprocessing:
- Each input slice is normalized
- Each input is padded to preserve dimensionality (height and width divisible by 16)
Commit: d993cd928bb181fd723d9be4bd8f28ff87ec14a0
Codename: bright-calf-792
Important!!! The pyproject.toml file and the uv.lock file were not included in this commit
Important!!! The computer crashed during this run
- Architecture: UNet Generator
- Task: One-to-One slice segmentation mask prediction
- QC / Filtering: Does not perform any QC or filtering of images or slices
- Data: Trained and evaluated on all slices of patient NF0016
- Preprocessing:
- Each input slice is normalized
- Each input is padded to preserve dimensionality (height and width divisible by 16)
Commit: 1ccccd1595607ddbebac95f93aa6ecad6184c414
Codename: stylish-flea-695
Note: Manually ended early because images weren't being saved
Important!!! Did not commit the collator, so this model will also need the collator committed later
Important!!! The pyproject.toml file and the uv.lock file were not included in this commit
- Architecture: UNet Generator
- Task: One-to-One slice segmentation mask prediction
- QC / Filtering: Does not perform any QC or filtering of images or slices
- Data: Trained and evaluated on all slices of patient NF0016
- Preprocessing:
- Each input slice is normalized
- Each input is padded to preserve dimensionality (height and width divisible by 16)