WISHLIST.md

A wish list of cookbooks showcasing:

• Inference

• Fine-tuning

  • Fine-tuning Gemma for function calling
  • Fine-tuning Gemma and/or CodeGemma for multi-turn, agentic workloads.

• Continued Pretraining

  • Continue to pretrain Gemma on TPU

• Responsible AI

  • Use LLM Comparator to compare Gemma with another LLM (i.e., Llama)