Vision-language models fail to load or abliterate
Multimodal models with vision wrappers such as Llava, Mistral3 and Qwen3-VL fail to load or crash during abliteration.
Tier II · Emerging Problems that are still building up.
4 public reports · first reported 2 December 2025 · last reported 13 December 2025
What's going wrong
Several users try to run the tool on vision-language checkpoints and get errors saying the configuration class is not recognised by the text-model loader.
AI-assisted: written with the help of AI from the reports. The idea report has the full summary.
In the idea report
- The full summary.
- The products the 4 reports are about, and how many came each month.
- Related problems.
No business has been found with this problem yet, so the report has no contacts. All sales are final — no refunds. Pricing.