出版物
- Macaron-V1: Towards Open Continual Learning with Self-Improvement and Mixture-of-LoRA阅读
- LongStraw: Long-Context RL Beyond 2M Tokens under a Fixed GPU Budget阅读
- On the Scaling of PEFT: Towards Million Personal Models of Trillion Parameters阅读
- Macaron-A2UI: A Model for Generative UI in Personal Agents阅读
- MinT: Managed Infrastructure for Training and Serving Millions of LLMs阅读
- δ-mem: Efficient Online Memory for Large Language Models阅读