Post

Replies

Boosts

Views

Activity

Reply to Large ML Models on the Neural Engine (Billion+ Parameters)
I also wonder the same thing. My guess is that ANE had a maximum tensor shape/size it can handle. If the size exceeds the maximum on any of the tensor dimensions, it would fail to use the ANE
Topic: Machine Learning & AI SubTopic: Core ML Tags:
Replies
Boosts
Views
Activity
Jul ’24