qwe2.5-vl-72B to abliterate
#1
by
testai111
- opened
Hi,
Can you abliterate 72B version of qwen2.5-vl too? It is supported by Ollama now an it will be very interesting to use it for purging closedai products.
We don't actually abliterate models, we mostly just quant them. The original model is by huihui-ai, you should ask them for an abliteration (and then can ask us to quant it, if we don't pick it up automatically). I think huihui-ai take requests.