Google has introduced gemma-trainer, an agent-assisted "skill" designed to make fine-tuning its Gemma 4 open-weight models accessible to a much broader range of developers. The tool lets coding agents configure training runs, manage execution, and evaluate results, lowering the barrier to customizing models that previously required substantial machine-learning expertise.
Continue reading
The rest of this article is for AI News Blitz readers. Choose an option below to keep reading.
Already purchased? Sign in✓ Signed in — this article isn’t included in your current plan.