Instructions to use Edoardo-Coppola/HuBERT-ECG-SFT-CardioLearning-base with libraries, inference providers, notebooks, and local apps. Follow these links to get started.
- Libraries
- Transformers
How to use Edoardo-Coppola/HuBERT-ECG-SFT-CardioLearning-base with Transformers:
# Load model directly from transformers import HuBERTECGForClassification model = HuBERTECGForClassification.from_pretrained("Edoardo-Coppola/HuBERT-ECG-SFT-CardioLearning-base", device_map="auto") - Notebooks
- Google Colab
- Kaggle
Mapping of Labels
#1
by UMCU - opened
Hi Edoardo,
thanks for publishing this state-of-the-art model. I would like to use it for inference but I cannot find any mapping
for the labels. Is the mapping still available and open?
Greetings,
Bram
Hi Bram,
I think you're referring to cardiolearning_labels.pkl at this path: https://github.com/Edoar-do/HuBERT-ECG/tree/master/reproducibility
That contains an ordered list of 164 label strings corresponding to disease/condition names matching the 164 outputs of HuBERT-ECG-SFT-CardioLearning-{small,base,large}.
Best and thanks for your interest in our work