Patch release v5.6.2
Qwen 3.5 and 3.6 MoE (text-only) were broken when using with FP8. It should now work again with this 🫡
Full Changelog: v5.6.1...v5.6.2
Patch release v5.6.2
Transformers（HuggingFace）Releases / 4/24/2026
📰 NewsDeveloper Stack & InfrastructureTools & Practical Usage
Key Points
- Patch release v5.6.2 addresses a regression where Qwen 3.5 and 3.6 MoE (text-only) models were broken when used with FP8.
- The update restores compatibility so those models should work again under FP8.
- It includes fixes to configuration reading and improved error handling for kernels.
- The patch references a specific fix via Hugging Face Transformers PR #45610 and provides a link to the full changelog between v5.6.1 and v5.6.2.
Related Articles
Black Hat USA
AI Business
AI productivity tools 2026: top 10 tools for remote teams
Dev.to
How I Use GitHub Copilot + RapidForge to Generate Daily Stock Ideas
Dev.to
Big Tech firms are accelerating AI investments and integration, while regulators and companies focus on safety and responsible adoption.
Dev.to
Anthropic CVP Run 3 — Does Claude's Safety Stack Scale Down to Haiku 4.5?
Dev.to
