Demo: Building a Gemma chatbot with Keras
summary
Shows how to load Gemma Instruct 7b from the Keras NLP collection / Kaggle Model Hub with a one‑liner, use turn‑by‑turn tagging and feed the full conversation as context, generate working Python (Sieve of Eratosthenes) that executes and returns primes to 1,000, and fine‑tune on a Pirate Speak dataset using LoRA (22M trainable params vs ~8.5B full) plus Keras model‑parallelism across eight TPUs on JAX, reloads the checkpoint and preserves the pirate voice.
AI-written summary · about
chapters (4)
related talks
09mUsing and fine-tuning Gemma 3
44mBuilding amazing Front-End Project with Gemma: A State-of-the-Art Open Model
39mFlutter Project with Gemma: A State-of-the-Art Open Model
10mAI for everyone with Gemma
10mBuilding intelligent agents with Gemma 3
39mBuilding Your Own Bot with GPT: Harnessing the Power of AI...
07mWelcome to the Gemmaverse
21mFrom 46% to 90%: Fine-Tuning Tiny LLMs for On-Device Agents