Benchmark Gemma models

[vmakarov-pistoia]

Gemma model family https://huggingface.co/google/gemma-7b https://huggingface.co/google/gemma-2b https://huggingface.co/google/codegemma-7b https://huggingface.co/google/codegemma-2b

see also https://www.kaggle.com/models/google/gemma

[slobentanzer]

Thanks, @vmakarov-pistoia. Gemma chat is available in Xinference standard models as gemma-it I believe (https://inference.readthedocs.io/en/latest/models/builtin/llm/gemma-it.html), so this should be quick. I didn't see codegemma though, that may take a little longer.

[slobentanzer]

I have to note that #134 from @matthijsz asks for overlapping models, we should probably only have one issue for Gemma.

[slobentanzer]

Closing this; Gemma models are accounted for by #134