GLM Cybersecurity FP8 Fine-tune With Refusals Removed Trends on Hugging Face

dealignai · hf · 2026-09-06

dealignai/GLM-5.3-CYBERSECURITY-FP8, an FP8-quantized GLM MoE (DSA) text-generation model, is trending on Hugging Face. It is a domain-specific cybersecurity fine-tune tagged abliterated and refusal-removed, targeting offensive security, red-teaming and pentesting workflows.

Original post →

More from Models

Models channel →