You signed in with another tab or window. Reload to refresh your session.You signed out in another tab or window. Reload to refresh your session.You switched accounts on another tab or window. Reload to refresh your session.Dismiss alert
The number of parameters used on the leaderboard is intended (in general) to correlate with compute cost. Removing the embedding layer would make high-throughput models like the minishlab embedding stand out more clearly.
The text was updated successfully, but these errors were encountered:
The number of parameters used on the leaderboard is intended (in general) to correlate with compute cost. Removing the embedding layer would make high-throughput models like the minishlab embedding stand out more clearly.
The text was updated successfully, but these errors were encountered: