Efficient Knowledge Distillation for LLMs: Offline Top-K Logits and a Fused Chunked KL Loss
Paper • 2608.03796 • Published • 4
Jose Luis Cantarero
jlcanta
·
AI & ML interests
None yet
Recent Activity
liked a model 7 months ago
MultiverseComputingCAI/HyperNova-60B liked a model 8 months ago
MultiverseComputingCAI/Qwen3-Next-80B-A3B-Thinking-UncensoredOrganizations
None yet
upvoted a paper 1 day ago