Qwen3.8-27B HauhauCS Aggressive Q8_K_P — deployment mirror
This repository is a public, single-weight deployment mirror maintained by yiliankexi for convenient model-cache deployment. It contains the unmodified Q8_K_P GGUF published by HauhauCS, based on Qwen / Alibaba Cloud's Qwen3.8-27B.
No weights have been retrained, re-quantized, renamed, or otherwise modified by this mirror's maintainer. This is only a deployment mirror, not a new model release, and it is not an official release by Qwen or HauhauCS.
本仓库仅为公开部署镜像,保留 HauhauCS 发布的原始 Q8_K_P 权重;镜像维护者未重新训练、量化或修改权重。
File and provenance
| Field | Value |
|---|---|
| Weight | Qwen3.8-27B-Uncensored-HauhauCS-Aggressive-Q8_K_P.gguf |
| Size | 31457990784 bytes (approximately 31.46 GB) |
| SHA-256 | 4e7735df4d1e2ec721f2551f531b815702a2f89123238c564797eda4b0304bc2 |
| Source revision | 993a5971fda8f30dd1b7eb2654792ba4415c7460 |
| License | Apache-2.0 |
The SHA-256 and byte size are recorded from Hugging Face's LFS metadata. The weight is copied on Hugging Face's servers; it is not downloaded to and re-uploaded from the maintainer's computer. Full provenance is in provenance.json.
Original authors and sources
- Variant and quantization author: HauhauCS.
- Original Aggressive MTP GGUF release.
- Original source model card, preserved without modification.
- Base model author: Qwen / Alibaba Cloud.
- Original base model.
Deployment scope
Only the single Q8_K_P text-model weight is mirrored. The vision projector and the separate HauhauCS FastMTP draft sidecar/runtime patch are not included. Runtime requirements and model behavior remain those documented by the original authors. This mirror has not itself been benchmarked or loaded for inference.
License and notices
The original model and this unchanged weight are distributed under Apache-2.0. The full upstream LICENSE, including Alibaba Cloud's copyright notice, and NOTICE attribution are retained. Original legal files are also archived under upstream/.