VersaTune: An Efficient Data Composition Framework for Training Multi-Capability LLMs

Documenti analoghi

• DataSculpt: Crafting Data Landscapes for Long-Context LLMs through Multi-Objective Partitioning
  di: Lu, Keer, et al.
  Pubblicazione: (2024)
• Med-R$^2$: Crafting Trustworthy LLM Physicians via Retrieval and Reasoning of Evidence-Based Medicine
  di: Lu, Keer, et al.
  Pubblicazione: (2025)
• Med-R$^3$: Enhancing Medical Retrieval-Augmented Reasoning of LLMs via Progressive Reinforcement Learning
  di: Lu, Keer, et al.
  Pubblicazione: (2025)
• Data Proportion Detection for Optimized Data Management for Large Language Models
  di: Liang, Hao, et al.
  Pubblicazione: (2024)
• Facilitating Multi-turn Function Calling for LLMs via Compositional Instruction Tuning
  di: Chen, Mingyang, et al.
  Pubblicazione: (2024)