Google Introduces Compact Language Model Gemma 2 2B

Google has unveiled Gemma 2 2B, a compact yet powerful AI model that can compete with industry leaders despite its small size. The new language model contains only 2.6 billion parameters, but demonstrates performance on par or better than much larger models, including GPT-3.5 and Mistral AI Mixtral 8×7B.

In the LMSYS Chatbot Arena test, a popular online platform for benchmarking and evaluating the quality of AI models, Gemma 2 2B scored 1130 points. This result is slightly ahead of the results of GPT-3.5-Turbo-0613 (1117 points) and Mixtral-8x7B (1114 points) – models with ten times more parameters.

Google reports that Gemma 2 2B also scored 56.1 on the MMLU (Massive Multitask Language Understanding) test and 36.6 on the MBPP (Mostly Basic Python Programming) test, which is a significant improvement over the previous version.

Google trained Gemma 2 2B on a massive dataset of 2 trillion tokens using systems powered by its proprietary TPU v5e AI accelerators. Support for multiple languages ​​expands its potential for use in global applications.

The success of the Gemma 2 2B could change the direction of research, shifting the focus from the race to larger models to improving smaller, more energy-efficient solutions that can run on consumer devices.