Downloads · 30 days
551K
54% of all-time downloads
philbert440/Qwen3.8-27B-Uncensored-Cyber-GGUF
Qwen3.8-27B-Uncensored-Cyber-GGUF is a image-text-to-text model from philbert440. Use it for the image-text-to-text 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.
llama.cpp GGUF quants of Qwen3.8-27B-Uncensored-Cyber (v2 recipe), with the vision projector (mmproj-) for image input and the MTP head (mtp-) for speculative decoding.
Downloads · 30 days
551K
54% of all-time downloads
All-time downloads
1M
Public
Repo size
264 GB
Likes
80
Public
Click a slice to open those files.
.gguf105 GB · 100%
From the Hugging Face model README
llama.cpp GGUF quants of Qwen3.8-27B-Uncensored-Cyber (v2 recipe), with the vision projector
(mmproj-*) for image input and the MTP head (mtp-*) for speculative decoding.
| file | notes |
|---|---|
*-Q8_0 / *-Q6_K / *-Q5_K_M / *-Q4_K_M | weight quants (Q4_K_M smallest, Q8_0 ≈ lossless) |
mmproj-*-{BF16,Q8_0} | vision projector (--mmproj for image input) |
mtp-*-{BF16,Q8_0,Q4_0} | MTP speculative-decoding head (optional) |
Full BF16 is not shipped as a single GGUF (exceeds the 50GB per-file limit) — use Q8_0 (≈lossless) or the
bf16 safetensors repo.
| cyber-open ↑ | confab ↓ | factual ↑ | gsm8k ↑ | degen ↓ | |
|---|---|---|---|---|---|
| Cyber v2 (this line) | 100/100 | 0.867 | 1.00 | 0.80 | 0.00 |
| previous Cyber build | 93/100 | 1.00 | 0.933 | 0.825 | 0.00 |
Uncensored / de-refused, tuned to fully answer cyber and offensive-security questions. Use responsibly and in compliance with applicable law.