We’re excited to announce the second batch of a16z Open Source AI Grant recipients.
This program is designed to support a thriving open source ecosystem around modern AI. We provide grant funding (not an investment) to developers and small teams who are building critical pieces of the open source AI stack. More information about the program and a list of prior recipients are available here.
This cohort focuses mainly on two areas: tools for training, hosting, and evaluating language models; and models and communities built around visual AI. This complements many of the language model fine-tuning projects included in the first batch.
The recipients are:
- Common Crawl: open repository of web crawl data used for training many LLMs
- Axolotl (Wing Lian): framework for fine-tuning LLMs
- SkyPilot: standard framework for multicloud AI training and serving
- LMSys (Lianmin Zheng, Wei-Lin Chiang, and Ying Sheng): open models, systems, and evaluation platforms including Vicuna, FastChat, and Chatbot Arena
- LLaVA (Haotian Liu): open source multimodal model (language and vision)
- Deforum (Huemin): platform and open source community for AI animation
- Lucidrains: open implementations of high-impact AI models
Thank you to all of the above individuals and organizations for their contributions, as well as everyone working toward strengthening the open source AI ecosystem.