Hands-On: Continued Pretraining Teaches Qwen New Domains

A hands-on guide walks through using Unsloth to perform continued pretraining on Qwen 3 4B, teaching it to understand a fictional city as a demonstration of adapting models to new domains.

2026-08-21 ~ 2026-08-21 · 2 related posts

• Hands-on: Teaching a local LLM reasoning in new domains via Continued Pretraining with Unsloth — funJS · 2026-08-21
• Teaching Qwen 3 4B a new domain via continued pretraining — funJS · 2026-08-21