llama.cpp/ggml/src/ggml-metal
도로로도로또 945bf10627
metal : add MoE kernel specialization for ne20=5 (#18667)
Add template specialization for kernel_mul_mm_id_map0 with ne20=5
to support models using 5 active experts (e.g., VAETKI).
2026-01-08 12:37:45 +02:00
..
CMakeLists.txt metal : refactor + optimize v2 (#15995) 2025-09-17 20:38:12 +03:00
ggml-metal-common.cpp metal : fix loop bound in ggml_mem_ranges (#16412) 2025-10-03 19:18:56 +03:00
ggml-metal-common.h metal : refactor + optimize v2 (#15995) 2025-09-17 20:38:12 +03:00
ggml-metal-context.h metal : refactor + optimize v2 (#15995) 2025-09-17 20:38:12 +03:00
ggml-metal-context.m metal : add residency sets keep-alive heartbeat (#17766) 2025-12-05 19:38:54 +02:00
ggml-metal-device.cpp metal : add count_equal op (#18314) 2025-12-31 10:39:48 +02:00
ggml-metal-device.h ggml: add env var GGML_OP_OFFLOAD_MIN_BATCH (#18535) 2026-01-08 11:03:21 +02:00
ggml-metal-device.m ggml: add env var GGML_OP_OFFLOAD_MIN_BATCH (#18535) 2026-01-08 11:03:21 +02:00
ggml-metal-impl.h metal : add count_equal op (#18314) 2025-12-31 10:39:48 +02:00
ggml-metal-ops.cpp metal : adjust extra size for FA buffer to avoid reallocations (#18545) 2026-01-02 19:02:18 +02:00
ggml-metal-ops.h metal : add count_equal op (#18314) 2025-12-31 10:39:48 +02:00
ggml-metal.cpp ggml: add env var GGML_OP_OFFLOAD_MIN_BATCH (#18535) 2026-01-08 11:03:21 +02:00
ggml-metal.metal metal : add MoE kernel specialization for ne20=5 (#18667) 2026-01-08 12:37:45 +02:00