5 papers
SearchSkill: Teaching LLMs to Use Search Tools with Evolving Skill Banks
Jinchao Hu, Meizhi Zhong, Kehai Chen +1
Teaching language models to use search tools is not only a question of whether they search, but also of whether they issue good queries. This is especially important in open-domain…
Adaptive Inner Speech-Text Alignment for LLM-based Speech Translation
Henglyu Liu, Andong Chen, Kehai Chen +4
Recent advancement of large language models (LLMs) has led to significant breakthroughs across various tasks, laying the foundation for the development of LLM-based speech translat…
ZigZagkv: Dynamic KV Cache Compression for Long-context Modeling based on Layer Uncertainty
Meizhi Zhong, Xikai Liu, Chen Zhang +5
Large Language models (LLMs) have become a research hotspot. To accelerate the inference of LLMs, storing computed caches in memory has become the standard technique. However, as t…
Understanding the RoPE Extensions of Long-Context LLMs: An Attention Perspective
Meizhi Zhong, Chen Zhang, Yikun Lei +5
Enabling LLMs to handle lengthy context is currently a research hotspot. Most LLMs are built upon rotary position embedding (RoPE), a popular position encoding method. Therefore, a…
MoDification: Mixture of Depths Made Easy
Chen Zhang, Meizhi Zhong, Qimeng Wang +8
Long-context efficiency has recently become a trending topic in serving large language models (LLMs). And mixture of depths (MoD) is proposed as a perfect fit to bring down both la…