ggml v0.28.0 Released: Sparse Attention and MoE Optimizations
The latest ggml release introduces GGML_OP_TOP_K_MASK for sparse attention, MoE optimizations for CUDA, and expanded ...
News, quick references and our own measurements for open-weight models you can run locally
The latest ggml release introduces GGML_OP_TOP_K_MASK for sparse attention, MoE optimizations for CUDA, and expanded ...