heartwise/EfficientNetV2_AFIB_5y

DeeECG EfficientNet V2 Fine Tuned to detect AFIB at 5y

• —modelname: `efficientnetv2afib_5y`
• —Utilizes the EfficientNetV2 architecture for processing ECG signals.
• —Pre-trained model is loaded from a specified directory, and the model is used to process ECG signal tensors.
• —The model takes an input signal tensor and outputs the logits for the binary classification of incident AFIB at 5 years.
• —Model compiled with JIT,
• —Frozen weights no tunable

To load use

import torch
model = torch.jit.load(model_path)