This is GPT-OSS 20B, with reasoning from Claude Opus 4.6 distilled into it. This is based on a fully uncensored base (Jinx) and the dataset was filtered to remove refusals and standardised to Harmony format.

Downloads last month
13
Safetensors
Model size
21B params
Tensor type
F32
·
Inference Providers NEW
This model isn't deployed by any Inference Provider. 🙋 Ask for provider support

Model tree for blascotobasco/GPT-OSS-20B-Opus-Uncensored

Finetuned
(1)
this model
Quantizations
1 model