]> git.djapps.eu Git - pkg/ggml/sources/ggml/commit
ggml : reduce hash table reset cost (llama/8698)
authorslaren <redacted>
Sat, 27 Jul 2024 02:41:55 +0000 (04:41 +0200)
committerGeorgi Gerganov <redacted>
Sat, 27 Jul 2024 15:26:12 +0000 (18:26 +0300)
commit8a7be58d9f6c79c79025a3e396a14b8dc9669a5f
treefd8fe686be87495696b3048650eaf8626a9abf87
parent482abf92174d9748dcd23ab1932ae778a45651f6
ggml : reduce hash table reset cost (llama/8698)

* ggml : reduce hash table reset cost

* fix unreachable code warnings after GGML_ASSERT(false)

* GGML_ASSERT(false) -> GGML_ABORT("fatal error")

* GGML_ABORT use format string
35 files changed:
ggml/src/ggml-cann.cpp
include/ggml.h
src/ggml-alloc.c
src/ggml-backend.c
src/ggml-blas.cpp
src/ggml-cuda.cu
src/ggml-cuda/argsort.cu
src/ggml-cuda/binbcast.cu
src/ggml-cuda/common.cuh
src/ggml-cuda/cpy.cu
src/ggml-cuda/dmmv.cu
src/ggml-cuda/fattn-common.cuh
src/ggml-cuda/fattn-tile-f16.cu
src/ggml-cuda/fattn-tile-f32.cu
src/ggml-cuda/fattn.cu
src/ggml-cuda/getrows.cu
src/ggml-cuda/mmq.cu
src/ggml-cuda/mmq.cuh
src/ggml-cuda/mmvq.cu
src/ggml-cuda/quantize.cu
src/ggml-cuda/rope.cu
src/ggml-impl.h
src/ggml-kompute.cpp
src/ggml-metal.m
src/ggml-quants.c
src/ggml-sycl.cpp
src/ggml-sycl/common.hpp
src/ggml-sycl/dmmv.cpp
src/ggml-sycl/dpct/helper.hpp
src/ggml-sycl/mmq.cpp
src/ggml-sycl/mmvq.cpp
src/ggml-sycl/rope.cpp
src/ggml-vulkan.cpp
src/ggml.c
tests/test-backend-ops.cpp