Step-2 SFT reference policies (π_ref) used to initialize DDRO (MS MARCO / NQ; PQ and Title+URL DocIDs); use these for fair comparisons/ablations.
Kidist Amde Mekonnen
kiyam
AI & ML interests
AI (Generative models, Computer vision, NLP) ,XAI
Recent Activity
liked a dataset about 6 hours ago
dqmis/NQ320K-NCI updated a model 23 days ago
kiyam/ddro-nq-pq-sft updated a model 23 days ago
kiyam/ddro-msmarco-pq-sftOrganizations
None yet