The Technology Innovation Institute (TII), the research pillar of Abu Dhabi’s Advanced Technology Research Council (ATRC), launched on Monday a second iteration of its large language model (LLM) – Falcon 2.
Within this series, TII has unveiled two versions of the LLM, the Falcon 2 11B, a more efficient and accessible LLM trained on 5.5 trillion tokens with 11 billion parameters, and Falcon 2 11B VLM, distinguished by its vision-to-language model (VLM) capabilities, which enable seamless conversion of visual inputs into textual outputs.
Multimodal model
While both models are multilingual, Falcon 2 11B VLM stands out as TII’s first multimodal model. Hence, it is currently the only LLM that has this image-to-text conversion capability, marking a significant advancement in AI innovation.
Moreover, in several tests, Falcon 2 11B surpasses the performance of Meta’s new Llama 3 with eight billion parameters and performs on par with Google’s Gemma 7B. Notably, both Falcon 2 11B and 11B VLM are open-source, empowering developers worldwide with unrestricted access.
“With other multimodal models soon coming to the market in various sizes, our aim is to ensure that developers and entities that value their privacy have access to one of the best AI models to enable their AI journey,” stated Faisal Al Bannai, Secretary-General of ATRC and Strategic Research and Advanced Technology Affairs Advisor to the President His Highness Sheikh Mohamed bin Zayed Al Nahyan.
Falcon 2 11B models handle tasks in English, French, Spanish, German, Portuguese, and various other languages. Meanwhile, Falcon 2 11B VLM has the capability to identify and interpret images and visuals, providing a wide range of applications across industries such as healthcare, finance, e-commerce, education, and legal sectors.
Read: UAE leads global fiber connectivity for 8 years in a row with 99.3 percent penetration
Future GenAI innovations
The success of Falcon 2 11B has led TII to look forward to working on more GenAI innovations. This includes adopting a new type of machine learning capability called “Mixture of Experts”. This method encompasses combining smaller networks with specific specializations to form the most knowledgeable domains. This innovation improves accuracy and accelerates decision-making, paving the way for more intelligent and efficient AI systems.
For more news on technology, click here.