ggml-webgpu: replace f32 with kv_type and q_type#23372

reeselevine merged 1 commit into

ggml-org:masterggml-org/llama.cpp:masterfrom

noumena-labs:webgpu/flash-attn-f16noumena-labs/llama.cpp:webgpu/flash-attn-f16Copy head branch name to clipboard