what is the performance gap of this model with the full precision model?
#7 opened 10 days ago
by 
yuimo
How do we use MTP for speculative decoding?
1
#6 opened 23 days ago
by 
pythiccoder
Raise per-image vision soft-token budget from 280 to 1120
#2 opened 30 days ago
by 
lucianommartins