AI models
LLaVA-CoT
Parameters
11B
Published
Nov 15, 2024
LLaVA-CoT is an AI model developed by Peking University, Tsinghua University, Peng Cheng Laboratory, Alibaba DAMO Academy and Lehigh University (China and United States), first published in November 2024. It works in the language, vision and multimodal domain, on tasks such as visual question answering, language modeling/generation and quantitative reasoning.
Epoch AI has no training-compute estimate for this model. The model has 11,000,000,000 parameters. It was trained on roughly 24M datapoints. Training ran on 8 NVIDIA H100 SXM5 80GB.
Access: Open weights (non-commercial). Its weights are openly available. It is built on top of Llama 3.2 11B. The reference paper has 459 citations. Epoch AI rates the confidence of this record as confident.
Full record
- Organization
- Peking University, Tsinghua University, Peng Cheng Laboratory, Alibaba DAMO Academy, Lehigh University
- Country of organization
- China, United States
- Domain
- Language, Vision, Multimodal
- Task
- Visual question answering, Language modeling/generation, Quantitative reasoning
- Parameters
- 11,000,000,000
- Dataset size
- 24M
- Training hardware
- NVIDIA H100 SXM5 80GB
- Chips used
- 8
- Training power draw
- 11.0 kW
- Model accessibility
- Open weights (non-commercial)
- Open weights
- Yes
- Base model
- Llama 3.2 11B
- Citations
- 459
- Epoch confidence
- Confident
SourceEpoch AI, 'AI Models'. Published online at epoch.ai. Retrieved 2026-07-29 from https://epoch.ai/data/ai-models. Licensed under CC BY 4.0.
← All ai models