You don't seem to have optimized the code you publish, making it a little difficult to read. Can you tell me where the verb classifier for the model is in the code? And how does it work?
Hi,
The provided code makes interactiveness predictions.
In notebooks/eval.ipynb, the interactiveness predictions are used to apply non-interaction suppression (NIS) on verb classification results.
Hi,
You don't seem to have optimized the code you publish, making it a little difficult to read. Can you tell me where the verb classifier for the model is in the code? And how does it work?