Downloads · 30 days
23
19% of all-time downloads
lkjiop8/cybersentinel-v5-phoenix-27b-MTP-Q8_0-GGUF-v5-3
cybersentinel-v5-phoenix-27b-MTP-Q8_0-GGUF-v5-3 is a machine learning model from lkjiop8. 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.
CyberSentinel v5-3 Phoenix 27B,含 MTP (Multi-Token Prediction) 头权重,支持 llama.cpp b9180+ 的原生 MTP 投机解码加速。
Downloads · 30 days
23
19% of all-time downloads
All-time downloads
123
Public
Repo size
29 GB
Likes
0
Public
Click a slice to open those files.
.gguf29 GB · 100%
From the Hugging Face model README
CyberSentinel v5-3 Phoenix 27B,含 MTP (Multi-Token Prediction) 头权重,支持 llama.cpp b9180+ 的原生 MTP 投机解码加速。
./llama-server \
-m CyberSentinel-v5-3-MTP-Q8_0.gguf \
--spec-type draft-mtp \
--spec-draft-n-max 3 \
--jinja \
--reasoning auto --reasoning-format deepseek --reasoning-budget -1 \
-ngl 99 -fa on -c 131072 --parallel 4 \
-ctk q8_0 -ctv q8_0
预期加速:约 2× 相对于无 MTP 的同款模型。