Q8_0 GGUF quantization of dalatexcoder/Rice-Cracker-Qwen3.5-0.8B-Abliterated-Base.
Method
The standard convert_hf_to_gguf.py produces broken GGUFs for this model due to differences between plain Q8_0 quantization and the imatrix-guided quantization used by the base Qwen3.5-0.8B GGUF. This was created by binary-splicing the abliterated weight deltas into unsloth/Qwen3.5-0.8B-GGUF (Q8_0), overwriting only the 212 tensors modified by abliteration while preserving unsloth's imatrix-quantized data for the remaining 108 unchanged tensors.
Usage
Load in LM Studio or any llama.cpp-compatible runtime as a standard Qwen3.5 model.
- Downloads last month
- 131
Hardware compatibility
Log In to add your hardware
8-bit
Inference Providers NEW
This model isn't deployed by any Inference Provider. 🙋 Ask for provider support