llama.cpp/ggml
Jeff Bolz e6d65fb02d
vulkan: support arbitrary KV dimension in flash attention (#16160)
The "Clamp" spec constant is already based on whether KV is a multiple of Bc,
so use that to control whether bounds checking is performed. Add bounds checking
to the scalar and coopmat1 paths. Coopmat2 didn't need any changes (the K/V
tensors are already optionally clamped, nothing else needed to be changed).
2025-09-27 22:43:39 +02:00
..
cmake ggml: Skip backend library linking code when GGML_BACKEND_DL=ON (#15094) 2025-08-07 13:45:41 +02:00
include llama: print memory breakdown on exit (#15860) 2025-09-24 16:53:48 +02:00
src vulkan: support arbitrary KV dimension in flash attention (#16160) 2025-09-27 22:43:39 +02:00
.gitignore vulkan : cmake integration (#8119) 2024-07-13 18:12:39 +02:00
CMakeLists.txt common : use cpp-httplib as a cURL alternative for downloads (#16185) 2025-09-26 14:12:19 +03:00