Fig. 4: Schematic of the overall training process.
From: Using large language models as a scalable mental status evaluation technique
The training dataset is augmented and then used to fine-tune a transformer model. The model is optimized through hyperparameter tuning before the final model is attained.
