Direct Preference Heads (Preprint)
Collection
This collection contains the pre-trained, fine-tuned and aligned models for the Direct Preference Heads paper. • 4 items • Updated • 1
How to use Avelina/lovelace-medium-alpha1-dph with Transformers:
# Load model directly
from transformers import LSWTForDPH
model = LSWTForDPH.from_pretrained("Avelina/lovelace-medium-alpha1-dph", device_map="auto")