Google has introduced three new merchandise which can be a part of the Gemma 2 household, a sequence of open AI fashions that had been launched in June. The brand new choices embrace Gemma 2 2B, ShieldGemma, and Gemma Scope.
Gemma 2 2B is a 2 billion parameter possibility, becoming a member of the prevailing 27 billion and 9 billion parameter sizes. In keeping with Google, this new measurement balances efficiency with effectivity, and may outperform different fashions in its class, together with all GPT-3.5 fashions.
It’s optimized with the NVIDIA TensorRT-LLM library and is offered as an NVIDIA NIM, making it preferrred for quite a lot of deployment varieties, comparable to information facilities, cloud, native workstations, PCs, and edge units. Gemma 2 2B additionally integrates with Keras, JAX, Hugging Face, NVIDIA NeMo, Ollama, and Gemma.cpp, and can quickly combine with MediaPipe as effectively.
And due to its small measurement, it will possibly run on the free tier of T4 GPUs in Google Colab, which Google believes will make “experimentation and improvement simpler than ever.”
It’s accessible now through Kaggle, Hugging Face, or Vertex AI Mannequin Backyard, and can be utilized inside Google AI Studio.
Subsequent, ShieldGemma is a sequence of security classifiers for detecting dangerous content material in mannequin inputs and outputs. It particularly targets hate speech, harassment, sexually specific content material, and harmful content material. The ShieldGemma fashions are open and designed to allow collaboration and transparency within the AI improvement neighborhood, and add to the prevailing suite of security classifiers within the firm’s Accountable AI Toolkit.
It’s accessible in several mannequin sizes to satisfy totally different wants. For instance, the 2B mannequin is good for on-line classification, whereas the 9B and 27B can present higher efficiency for offline eventualities the place latency isn’t a priority. In keeping with Google, all mannequin sizes use NVIDIA velocity optimizations to enhance efficiency.
And at last, Gemma Scope supplies higher transparency into how Gemma 2 fashions come to their selections, it will possibly allow researchers to grasp how Gemma 2 identifies patterns, processes info, and makes predictions. It makes use of sparse autoencoders (SAEs) to have a look at particular factors within the mannequin and “unpack the dense, complicated info processed by Gemma 2, increasing it right into a type that’s simpler to investigate and perceive,” Google defined in a weblog submit.
“These releases characterize our ongoing dedication to offering the AI neighborhood with the instruments and assets wanted to construct a future the place AI advantages everybody. We consider that open entry, transparency, and collaboration are important for creating secure and useful AI,” Google wrote.
You may additionally like…
Google releases Gemma, a brand new AI mannequin designed with AI researchers in thoughts
RAG is the following thrilling development for LLMs
