MM-LLMs
Paper Name: MM-LLMs: Recent Advances in MultiModal Large Language Models
Summary
3+ Most Important Things
1+ Deficiencies
3+ New Ideas
- Use my current work on ARL and develop a dataset for Supervised fine tuning that can be used on the stage 2 of RLHF
Important Figures
Components of usual MM-LLMs
Advancement in 2023
Mainstream MM-LLMs
Multimodal Pre-Training Dataset
Multimodal SFT or IT dataset
Multimodal Benchmark