Efficient Knowledge Distillation for LLMs: Offline Top-K Logits and a Fused Chunked KL Loss
Paper • 2608.03796 • Published • 14
Jezabel Rodriguez Garcia
jezgarcia
AI & ML interests
None yet
Recent Activity
authored a paper 5 days ago
Efficient Knowledge Distillation for LLMs: Offline Top-K Logits and a Fused Chunked KL LossOrganizations
None yet
authored a
paper 5 days ago