DeepSeek-V4 - a 6block Collection

6block 's Collections

DeepSeek-V4

updated 6 days ago

Sub-4-bit GGUF quants of DeepSeek-V4: Flash-0731 (seven PPL-tested tiers + DSpark draft) and Pro-0813 (1.57T, factory FP4 master).

6block/DeepSeek-V4-Flash-0731-GGUF

Text Generation • 284B • Updated 23 days ago • 3.05k • 2

Note GGUF quants for llama.cpp
6block/DeepSeek-V4-Pro-0813-GGUF

Text Generation • 1.6T • Updated 6 days ago • 4.27k • 2

Note 1.57T MoE, sub-4-bit tiers only (factory FP4 master)