Qwen3.6-27B-Heretic2-Uncensored-Finetune-Thinking-mlx-4Bit
Dracoform/Qwen3.6-27B-Heretic2-Uncensored-Finetune-Thinking-mlx-4Bit
finetuneMLXSAFETENSORS
Published metrics
- KL divergence
- 0.047
- Refusals
- 4/100
- Original refusals
- 99/100
- Heretic version
- —
Inherited from the parent conversion — source
Matched snippet
| KL divergence | 0.0469 | | Refusals | 4/100 |
Provenance
- Canonical model
- Qwen/Qwen3.6-27B
- Declared parent(s)
- DavidAU/Qwen3.6-27B-Heretic2-Uncensored-Finetune-Thinking
- Classification
- card identifies this as a finetune of a Heretic model
- Created / updated
- 2026-06-19 / 2026-06-19
- Downloads
- 1.3K all time · 27 last 30 days
- License
- apache-2.0
Refusal rate 4%. Metrics are self-reported by the publisher; the directory does not re-run evaluations.