MeZO: Fine-Tuning Language Models with Just Forward Passes
A memory-efficient method to fine-tune LMs.
Princeton NLP Blog
2023
FireAct: Toward Language Agent Fine-tuning
Blog Post by Baian Chen and Shunyu Yao.
Sheared LLaMA: Accelerating Language Model Pre-training via Structured Pruning
Blog Post by Mengzhou Xia and Tianyu Gao from Princeton University
Flash-Decoding for Long-Context Inference
Blog Post by Tri Dao (Princeton), Daniel Haziza (Meta), Francisco Massa (Meta), and Grigory Sizov (Meta).
Language Agents in the Digital World: Opportunities and Risks
Blog Post by Shunyu Yao and Karthik Narasimhan.
FlashAttention-2: Faster Attention with Better Parallelism and Work Partitioning
Blog Post by Professor Tri Dao.
The Socratic Method for Self-Discovery in Large Language Models
Is there a Theory of Anamnesis of Large Language Models?
2021
Phrase Retrieval and Beyond
Blog Post by Jinhyuk Lee.
Prompting: Better Ways of Using Language Models for NLP Tasks
Blog Post by Tianyu Gao.