AI is happening. The future is updating.DeepSeek-V4.1-Flash-GGUF
GGUF conversions and mixed-precision quantizations of deepseek-ai/DeepSeek-V4.1-Flash for the experimental DeepSeek-V4.1 runtime in smalinin/llama.cpp, branch `my_build_deepseek41`. [! **ONLY_for_CUDA** ] This is **DeepSeek-V4.1-Flash** (`DeepseekV41ForCausalLM`), not DeepSeek-V4-Flash-0731. The architecture includes sparse attention, hyper-connections, MoE layers, and two very large n-gram Engram lookup tables. > [!IMPORTANT] > These files require the linked `my_build_deepseek41` branch. Compatibility > with stock upstream `llama.cpp` is not claimed.
GGUF conversions and mixed-precision quantizations of deepseek-ai/DeepSeek-V4.1-Flash for the experimental DeepSeek-V4.1 runtime in smalinin/llama.cpp, branch `my_build_deepseek41`. [! **ONLY_for_CUDA** ] This is **DeepSeek-V4.1-Flash** (`DeepseekV41ForCausalLM`), not DeepSeek-V4-Flash-0731. The architecture includes sparse attention, hyper-connections, MoE layers, and two very large n-gram Engram lookup tables. > [!IMPORTANT] > These files require the linked `my_build_deepseek41` branch. Compatibility > with stock upstream `llama.cpp` is not claimed.