ollama/enable-lto.patch
Loren Burkholder a4111a1692 Accepting request 1169791 from home:rrahl0:branches:science:machinelearning
- Update to version 0.1.32:
  * scale graph based on gpu count
  * Support unicode characters in model path (#3681)
  * darwin: no partial offloading if required memory greater than system
  * update llama.cpp submodule to `7593639` (#3665)
  * fix padding in decode
  * Revert "cmd: provide feedback if OLLAMA_MODELS is set on non-serve command (#3470)" (#3662)
  * Added Solar example at README.md (#3610)
  * Update langchainjs.md (#2030)
  * Added MindsDB information (#3595)
  * examples: add more Go examples using the API (#3599)
  * Update modelfile.md
  * Add llama2 / torch models for `ollama create` (#3607)
  * Terminate subprocess if receiving `SIGINT` or `SIGTERM` signals while model is loading (#3653)
  * app: gracefully shut down `ollama serve` on windows (#3641)
  * types/model: add path helpers (#3619)
  * update llama.cpp submodule to `4bd0f93` (#3627)
  * types/model: make ParseName variants less confusing (#3617)
  * types/model: remove (*Digest).Scan and Digest.Value (#3605)
  * Fix rocm deps with new subprocess paths
  * mixtral mem
  * Revert "types/model: remove (*Digest).Scan and Digest.Value (#3589)"
  * types/model: remove (*Digest).Scan and Digest.Value (#3589)
  * types/model: remove DisplayLong (#3587)
  * types/model: remove MarshalText/UnmarshalText from Digest (#3586)
  * types/model: init with Name and Digest types (#3541)
  * server: provide helpful workaround hint when stalling on pull (#3584)
  * partial offloading
  * refactor tensor query
  * api: start adding documentation to package api (#2878)

OBS-URL: https://build.opensuse.org/request/show/1169791
OBS-URL: https://build.opensuse.org/package/show/science:machinelearning/ollama?expand=0&rev=9
2024-04-23 12:00:24 +00:00

31 lines
1.8 KiB
Diff

diff -rub ollama/llm/generate/gen_linux.sh ollama-patched/llm/generate/gen_linux.sh
--- ollama/llm/generate/gen_linux.sh 2024-04-23 04:40:58.246062467 +0200
+++ ollama-patched/llm/generate/gen_linux.sh 2024-04-23 04:37:36.432294889 +0200
@@ -51,7 +51,7 @@
export CUDACXX=$(command -v nvcc)
fi
fi
-COMMON_CMAKE_DEFS="-DCMAKE_POSITION_INDEPENDENT_CODE=on -DLLAMA_NATIVE=off -DLLAMA_AVX=on -DLLAMA_AVX2=off -DLLAMA_AVX512=off -DLLAMA_FMA=off -DLLAMA_F16C=off"
+COMMON_CMAKE_DEFS="-DCMAKE_POSITION_INDEPENDENT_CODE=on -DLLAMA_LTO=on -DCMAKE_BUILD_TYPE=Release -DLLAMA_NATIVE=off -DLLAMA_AVX=on -DLLAMA_AVX2=off -DLLAMA_AVX512=off -DLLAMA_FMA=off -DLLAMA_F16C=off"
source $(dirname $0)/gen_common.sh
init_vars
git_module_setup
@@ -77,7 +77,7 @@
if [ -n "${OLLAMA_CUSTOM_CPU_DEFS}" ]; then
init_vars
echo "OLLAMA_CUSTOM_CPU_DEFS=\"${OLLAMA_CUSTOM_CPU_DEFS}\""
- CMAKE_DEFS="${OLLAMA_CUSTOM_CPU_DEFS} -DCMAKE_POSITION_INDEPENDENT_CODE=on ${CMAKE_DEFS}"
+ CMAKE_DEFS="${OLLAMA_CUSTOM_CPU_DEFS} -DCMAKE_POSITION_INDEPENDENT_CODE=on -DLLAMA_LTO=on -DCMAKE_BUILD_TYPE=Release ${CMAKE_DEFS}"
BUILD_DIR="../build/linux/${ARCH}/cpu"
echo "Building custom CPU"
build
@@ -93,7 +93,7 @@
# -DLLAMA_AVX512_VBMI -- 2018 Intel Cannon Lake
# -DLLAMA_AVX512_VNNI -- 2021 Intel Alder Lake
- COMMON_CPU_DEFS="-DCMAKE_POSITION_INDEPENDENT_CODE=on -DLLAMA_NATIVE=off"
+ COMMON_CPU_DEFS="-DCMAKE_POSITION_INDEPENDENT_CODE=on -DLLAMA_LTO=on -DCMAKE_BUILD_TYPE=Release -DLLAMA_NATIVE=off"
if [ -z "${OLLAMA_CPU_TARGET}" -o "${OLLAMA_CPU_TARGET}" = "cpu" ]; then
#
# CPU first for the default library, set up as lowest common denominator for maximum compatibility (including Rosetta)