Gemini and Gemma

Google is certainly aiming for the stars.  

Not only is their new AI model called Gemini which of course is the name of a northern constellation of stars but they are going to enhance this constellation with another star -Astra!

Google's owner Alphabet has revealed a new artificial intelligence agent that answers in real-time queries in multiple modes like video and audio as well as text.

This new AI multimodal assistant is called project Astra.

It was demonstrated by CEO Sundar Pichai during an annual developer conference on Tuesday, May 14th 2024.

Bard, Google's first drive into the generative ai space following OpenAi's ChatGpt success has now changed its name to Gemini.

Gemini is the next advancement of the models driving Bard and Google see the name change as significant and not being 'just a name change;' but a reflection of their more advanced technology. See more on Gemma and Gemini here.

Well its all G's: Google, Gemini, Gemma...getting confused?  Maybe this will help!

Google of course is the search engine company founded way back in 1998 part of the massive Alphabet Inc.

Gemini is a leading 'closed' AI model.

Gemma is an 'open source' model.

So how does this help:

Google

Google has an initiative 'Google DeepMind'.   This was the combination of two of the world's leading AI labs:  Google Brain and DeepMind.  This approach has resulted in new products such as Gemini and Gemma.  

It all started a little way back with LLaMA and then the outstanding LLaMA 2 the Large Language Model from Google that was released as a platform in open source for others to build on.

Now Google is well and truly back in the opensource LLM world with its new model Gemma!

Gemma has been trained on 6 million tokens compared to LLaMA 2's 4 million.

Gemma has been released with 4 new models.

The Base Models

Firstly a base model with a 7 billion or 2 billion parameters model. variants