Course Project of Machine Learning for Healthcare, Spring 2020
Blog post: https://sohamtamba.github.io/projects/deep-learning/2020/12/17/mlhc-tf.html
- Python: 3.8.5
- Pytorch: 1.7.0
- CUDA: 10.2
- Torchvision: 0.8.1
- Pytorch-Lightning: 1.0.8
- PIL: 8.0.1
To train a CBR-Large-Wide model on the entire dataset to solve the Consolidationt detection task, execute:
python train_scratch --model_type cbr_large_wide -- task Consolidation --num_train_samples -1
To train a CBR-Large-Wide model on 5,000 data points to solve the Edema detection task, execute:
python train_scratch --model_type cbr_large_wide -- task Edema --num_train_samples 5000
To fine-tune a CBR model to solve the Edema detection task using 5,000 datapoints, execute:
python finetune.py --task Edema --num_train_samples 5000 --backbone_path <Path to checkpoint created by train_scratch.py>