- Item Prepare image encoder and text encoder. download the image encoder at: image encoder. Put it to backbones/image_encoder
backbones/image_encoder/
deit_base_patch16_224-b5f2ef4d.pth
ImageEncoder.py
vits.py
download the text encoder at: text encoder. unzip the Bio_ClinicalBERT.zip. Put it to backbones/bert_model
backbones/bert_model/
Bio_ClinicalBERT
config.py
flax_model.msgpack
...
bert_config.py
med.py
TextEncoder.py
- Item Run VLM pre-training, use the following command.
cd vlm_pretrain
python train_model.py
Run SSMIS, use the following command.
cd segmentation_train
python train_SSL.py