Skip to content
  • Heijmer, Thijs's avatar
    65e8f6da
    Refactor model training and evaluation pipeline · 65e8f6da
    Heijmer, Thijs authored
    - Updated `analysis.py` to include new model `DinoSCP8736_1` and adjusted data loading for test dataset.
    - Modified `main.py` to use `DinoSCP8736_1` for training and added loss calculation.
    - Enhanced `LitSCP8736` class in `model.py` to support per-class F1 score logging.
    - Removed obsolete `model_dino.py` file.
    - Added new Jupyter notebooks for charts and inference.
    - Updated data augmentation strategies in `analysis.py` and `main.py`.
    - Cleaned up utility functions in `utils.py` for better clarity and functionality.
    - Deleted `list.txt` and `test_loss_results.csv` as they are no longer needed.
    - Added new tensorboard output CSV files for tracking model performance.
    - Updated requirements to reflect changes in dependencies.
    65e8f6da
    Refactor model training and evaluation pipeline
    Heijmer, Thijs authored
    - Updated `analysis.py` to include new model `DinoSCP8736_1` and adjusted data loading for test dataset.
    - Modified `main.py` to use `DinoSCP8736_1` for training and added loss calculation.
    - Enhanced `LitSCP8736` class in `model.py` to support per-class F1 score logging.
    - Removed obsolete `model_dino.py` file.
    - Added new Jupyter notebooks for charts and inference.
    - Updated data augmentation strategies in `analysis.py` and `main.py`.
    - Cleaned up utility functions in `utils.py` for better clarity and functionality.
    - Deleted `list.txt` and `test_loss_results.csv` as they are no longer needed.
    - Added new tensorboard output CSV files for tracking model performance.
    - Updated requirements to reflect changes in dependencies.
Loading