IstroSec/ThinkingCap-Qwen3.8-27B-abliterated
Adapter vs MuXodious/Qwen3.8-27B-absolute-heresy-LoRA. MuXodious/Qwen3.8-27B-absolute-heresy-LoRA is a LoRA adapter, not full weights: rank 32, on down_proj, o_proj, out_proj in 64 layers (128 adapted matrices). It modifies IstroSec/ThinkingCap-Qwen3.8-27B-abliterated at load time.
7 derivatives checked
- IstroSec/ThinkingCap-Qwen3.8-27B-abliterated-GGUF: Quantization
- mradermacher/ThinkingCap-Qwen3.8-27B-abliterated-i1-GGUF: Quantization
- mradermacher/ThinkingCap-Qwen3.8-27B-abliterated-GGUF: Quantization
- groxaxo/ThinkingCap-Qwen3.8-27B-abliterated-UD-Q4_K_XL-GGUF: Quantization
- DoktorMincs/ThinkingCap-Qwen3.8-27B-abliterated-W4A16-AutoRound: Quantization
- IstroSec/ThinkingCap-Qwen3.8-27B-abliterated-NVFP4A16: Quantization
- IstroSec/ThinkingCap-Qwen3.8-27B-abliterated-FP8-DYNAMIC: Quantization