PipeRAG: Enhancing Efficiency in Retrieval-Augmented Generation
PipeRAG improves generation efficiency through pipeline parallelism, flexible retrieval intervals, and performance modeling.
Optimizing Transformers for Long-Text Classification on Limited GPU Service
Efficiently optimize Transformer models for long-text classification on limited GPU resources.
KIWI: A Dataset of Knowledge-Intensive Writing Instructions for Answering Research Questions
Current LLMs struggle to effectively follow expert-written instructions for revising long-form answers in the scientific domain.
Metaphor Understanding Challenge Dataset for Large Language Models (LLMs)
Metaphors in natural language are essential for large language models, and the Metaphor Understanding Challenge Dataset (MUNCH) provides a challenging task for LLMs to interpret metaphors accurately.
Decomposed Prompting: Unveiling Multilingual Linguistic Structure in English-Centric Large Language Models
English-centric Large Language Models demonstrate multilingual capabilities through decomposed prompting, surpassing iterative methods in efficacy and efficiency.
Analyzing Precision and Recall for LLMs Evaluation Framework
The author introduces a novel evaluation framework for Large Language Models (LLMs) by adapting Precision and Recall metrics from image generation to text generation. This approach provides insights into the quality and diversity of generated text without the need for aligned corpora.