How important is the grouped_topk?

by dzhulgakov - opened Jul 4

Jul 4

Hi,

I ran the model with just regular top_k softmax and got pretty sensible results. How important is it to include the hierchical grouped top_k?

dzhulgakov changed discussion status to closed Jul 4

Upload images, audio, and videos by dragging in the text input, pasting, or clicking here.

Tap or paste here to upload images

· Sign up or log in to comment