Tags: #large-language-models

An AI project extending Large Language Models with video understanding capabilities, enabling conversational AI to process and respond to queries about video content.

AirLLM optimizes large language model inference memory, enabling 70B LLMs on a single 4GB GPU without quantization, and 405B Llama3.1 on 8GB VRAM.

An efficient and user-friendly library for scaling Reinforcement Learning with Large Language Models on large-scale GPU resources.

A comprehensive collection of recipes and code for training various reward models crucial for Reinforcement Learning from Human Feedback (RLHF) in large language models.

A comprehensive GitHub repository serving as the official companion to the survey paper 'A Survey of Large Language Models,' curating papers and resources on LLMs.