File size: 76 Bytes
5fa1a76
 
1
2
to the model, which it will
use to perform cross-attention with the latents.