YAML Metadata
Warning:
empty or missing yaml metadata in repo card
(https://huggingface.co/docs/hub/model-cards#model-card-metadata)
an attempt to uncensor the base delirium even further using new abliteration tools
Update: Unfortunately, this program was a big waste of time for ablating pre-existing finetunes like Delirium, and appears to only work on base models like GPT OSS.
54 hours spent running only to produce another safety maxxed, censored version. I will not be using Heretic for my ablations moving forward until the program is more reliable.
This is the tool I made v1 with and the one that seems to work best for finetunes: https://github.com/jim-plus/llm-abliteration/
Specifically, this version: https://github.com/jim-plus/llm-abliteration/archive/4f68fab37a2aa8f4f6d9d016c1977d16c25031b0.zip
(I tested the newest one with Refusal Purity and it is less stable, producing Chinese output)
- Downloads last month
- 2
Hardware compatibility
Log In
to view the estimation
4-bit
Inference Providers
NEW
This model isn't deployed by any Inference Provider.
🙋
Ask for provider support