-
Download the model and the configuration file at the link: https://drive.google.com/drive/folders/1AkhAJPA3ZA5AoI9QQvo6m7RC9Zv0lhH3?usp=sharing
-
The
model_test.ipynbnotebook has some code to load the model and its config file. In essence, we can use the Whisper API to load the fine-tuned model, provided we have thedecode_options. More information can be found in themodel_loader.pyfile. -
To run fine-tuning, download the dataset from the torgo database https://www.cs.toronto.edu/~complingweb/data/TORGO/torgo.html, and then simply do
python train.py.
Folders and files
| Name | Name | Last commit date | ||
|---|---|---|---|---|