Matthijs Van Keirsbilck

Latest

• Fast-SLM: Towards Latency-Optimal Hybrid Small Language Models
• Hymba: A Hybrid-head Architecture for Small Language Models