heartwise/DeepECG-Tok_EfficientNetV2_77_Classes

DeepECG-Tok_EfficientNetV2_77_Classes

• DeepECG-Tok_EfficientNetV2_77_Classes is a repository for the EfficientNetV2 weights that classify embeddings from DeepECG-Tok into 77 classes.

• base_config.yaml file contains the parameters for the training and inference.

• Paths to be updated:

  • base_checkpoint_path: path to the checkpoints directory
  • pretrained_tokenizer_path: path to the pretrained tokenizer checkpoint
  • inference_dataset_path: path to the inference dataset
  • inference_checkpoint_path: path to the inference checkpoint

• To run inference, run the following command:

bash scripts/runner.sh --base_config config/linear_probing/base_config.yaml --selected_gpus 0,1 --use_wandb true --run_mode inference