DeepSeek V4 Flash GGUF
Using MTP requires am17an/llama.cpp:dsv4-mtp until llama.cpp#25784 is merged.
Modellquelle
- Art
- Other
- Dateigröße
- 3,51–145 GB
- Basismodell
- deepseek-ai/DeepSeek-V4-Flash
- Version
- vRevision 4810cf3
- Aktualisiert
- 4. Okt. 2026
- Downloads
- 2957
- Quelle
- Hugging Face
Quellenbeschreibung
GGUF quantizations of DeepSeek-V4-Flash.
Using MTP requires am17an/llama.cpp:dsv4-mtp until llama.cpp#25784 is merged.
Quellen
1 Quelle
Quellen
1 QuelleVerifiziert 1. Sept.
Modellartefakte
5 Artefakte
Modellartefakte
5 ArtefakteDeepSeek-V4-Flash-3.86bpw-IQ3_S.gguf
gguf · 128 GB · SHA-256 46a1c2a740d8…c758 · Hugging Face
HerunterladenDeepSeek-V4-Flash-3.86bpw-Q3_K.gguf
gguf · 128 GB · SHA-256 6a1bb6ab90d5…1b26 · Hugging Face
HerunterladenDeepSeek-V4-Flash-MTP-3.93bpw.gguf
gguf · 3,51 GB · SHA-256 539e2e1fe0ac…a212 · Hugging Face
HerunterladenDeepSeek-V4-Flash-MTP-4.93bpw.gguf
gguf · 4,41 GB · SHA-256 d428e3eb8443…3e74 · Hugging Face
HerunterladenQuellenauszüge
2 Auszüge
Quellenauszüge
2 Auszüge--- base_model: - deepseek-ai/DeepSeek-V4-Flash --- GGUF quantizations of DeepSeek-V4-Flash. Using MTP requires [am17an/llama.cpp:dsv4-mtp](https://github.com/am17an/llama.cpp/tree/dsv4-mtp) until [llama.cpp#25784](https://github.com/ggml-org/llama.cpp/pull/25784) is merged.
Source context: 2957 downloads · 7 likes · Repo ddh0/DeepSeek-V4-Flash-GGUF
