Distribute and run LLMs with a single file.
C++ cross-platformggufllama-cpplocal-ailocal-inferencelocal-llmopen-source-aisingle-file-executablespeech-to-text
Release llamafile v0.8.12
Justine Tunneya73ea134aa09
diff --git a/llamafile/BUILD.mk b/llamafile/BUILD.mkindex 0be2be02..0e599dbb 100644--- a/llamafile/BUILD.mk+++ b/llamafile/BUILD.mk@@ -136,7 +136,7 @@ o/$(MODE)/llamafile/tinyblas_cpu_sgemm_amd_fma.o \o/$(MODE)/llamafile/tinyblas_cpu_sgemm_amd_zen4.o \o/$(MODE)/llamafile/tinyblas_cpu_sgemm_arm80.o \o/$(MODE)/llamafile/tinyblas_cpu_sgemm_arm82.o: \- private CCFLAGS += -O3+ private CCFLAGS += -O3 -fopenmp################################################################################# testing@@ -163,6 +163,9 @@ o/$(MODE)/llamafile/sgemm_vecdot_test: \o/$(MODE)/llamafile/sgemm_vecdot_test.o \o/$(MODE)/llama.cpp/llama.cpp.a+o/$(MODE)/llamafile/sgemm_vecdot_test: \+ private LDFLAGS += -fopenmp+o/$(MODE)/llamafile/%.o: llamafile/%.cu llamafile/BUILD.mk@mkdir -p $(@D)build/cudacc -fPIE -g -O3 -march=native -ffast-math --use_fast_math -c -o $@ $<diff --git a/llamafile/version.h b/llamafile/version.hindex b4793c92..cab6b5e6 100644--- a/llamafile/version.h+++ b/llamafile/version.h@@ -2,7 +2,7 @@#define LLAMAFILE_MAJOR 0#define LLAMAFILE_MINOR 8-#define LLAMAFILE_PATCH 11+#define LLAMAFILE_PATCH 12#define LLAMAFILE_VERSION \(100000000 * LLAMAFILE_MAJOR + 1000000 * LLAMAFILE_MINOR + LLAMAFILE_PATCH)