Downloads · 30 days
44
13% of all-time downloads
Qwen3.8-27B-Uncensored-Compressed
SofiTesfay2010/Qwen3.8-27B-Uncensored-Compressed
Qwen3.8-27B-Uncensored-Compressed is a machine learning model from SofiTesfay2010. Use it for the machine learning task on the model card, and read the license before you ship it in a product. The card lists the license as apache-2.0.
Calibrated Sherry-X (1.45 bpw) dense quantization of orcarouter/Qwen3.8-27B-Uncensored. Calibration Source: 6block/Qwen3.8-27B-GGUF Saliency Partitioning: Top 1% outlier channels preserved in FP8 Base Core: Tencent ST…
All-time downloads
349
Public
Parameters
6.8B
21.5 GB on disk
Likes
0
Public
Repo makeup
Click a slice to open those files.
.safetensors7 GB · 100%
Parameter types
How the weights are stored.
U83.9B · 57%
Base models
Model card
From the Hugging Face model README
SofiTesfay2010/Qwen3.8-27B-Uncensored-Compressed
Calibrated Sherry-X (1.45 bpw) dense quantization of orcarouter/Qwen3.8-27B-Uncensored.
- Calibration Source:
6block/Qwen3.8-27B-GGUF - Saliency Partitioning: Top 1% outlier channels preserved in FP8
- Base Core: Tencent STQ1_0 WLS 3:4 ternary
- Residual Corrections: Rank-1 singular vectors in FP8
- Type
- qwen3_5
- License
- apache-2.0
- Created
- Sep 2, 2026
- Updated
- Sep 3, 2026