Google releases Gemma open LLM family with 2B and 7B models

AI Impact Summary

Google has released Gemma, a new family of open LLMs with two sizes (7B and 2B) designed for efficient deployment on consumer-grade hardware. The models offer instruction-tuned variants and a context length of 8K tokens, and support for both base and instruction-tuned models. The release includes integrations with Hugging Face and Google Cloud, offering options for deployment and training via Vertex AI, GKE, and Inference Endpoints, with quantization support for running on lower-spec hardware.