Skip to content

Commit 8917767

Browse files
GalunidGalunid
committed
Merge branch 'master' into stablelm-support
2 parents a00bb06 + f28af0d commit 8917767

Some content is hidden

Large Commits have some content hidden by default. Use the searchbox below for content that may be hidden.

71 files changed

+8863
-7921
lines changed

‎.github/ISSUE_TEMPLATE/bug.md

Lines changed: 1 addition & 1 deletion
Original file line numberDiff line numberDiff line change
@@ -1,7 +1,7 @@
11
---
22
name: Bug template
33
about: Used to report bugs in llama.cpp
4-
labels: ["bug"]
4+
labels: ["bug-unconfirmed"]
55
assignees: ''
66

77
---

‎.gitignore

Lines changed: 2 additions & 1 deletion
Original file line numberDiff line numberDiff line change
@@ -15,6 +15,7 @@
1515
.DS_Store
1616
.build/
1717
.cache/
18+
.ccls-cache/
1819
.direnv/
1920
.envrc
2021
.swiftpm
@@ -64,7 +65,7 @@ models-mnt
6465
/parallel
6566
/train-text-from-scratch
6667
/vdot
67-
build-info.h
68+
/common/build-info.cpp
6869
arm_neon.h
6970
compile_commands.json
7071
CMakeSettings.json

‎CMakeLists.txt

Lines changed: 13 additions & 42 deletions
Original file line numberDiff line numberDiff line change
@@ -44,7 +44,7 @@ endif()
4444

4545
# general
4646
option(LLAMA_STATIC "llama: static link libraries" OFF)
47-
option(LLAMA_NATIVE "llama: enable -march=native flag" ON)
47+
option(LLAMA_NATIVE "llama: enable -march=native flag" OFF)
4848
option(LLAMA_LTO "llama: enable link time optimization" OFF)
4949

5050
# debug
@@ -82,6 +82,7 @@ set(LLAMA_BLAS_VENDOR "Generic" CACHE STRING "llama: BLAS library vendor")
8282
option(LLAMA_CUBLAS "llama: use CUDA" OFF)
8383
#option(LLAMA_CUDA_CUBLAS "llama: use cuBLAS for prompt processing" OFF)
8484
option(LLAMA_CUDA_FORCE_DMMV "llama: use dmmv instead of mmvq CUDA kernels" OFF)
85+
option(LLAMA_CUDA_FORCE_MMQ "llama: use mmq kernels instead of cuBLAS" OFF)
8586
set(LLAMA_CUDA_DMMV_X "32" CACHE STRING "llama: x stride for dmmv CUDA kernels")
8687
set(LLAMA_CUDA_MMV_Y "1" CACHE STRING "llama: y block size for mmv CUDA kernels")
8788
option(LLAMA_CUDA_F16 "llama: use 16 bit floats for some calculations" OFF)
@@ -93,46 +94,12 @@ option(LLAMA_CLBLAST "llama: use CLBlast"
9394
option(LLAMA_METAL "llama: use Metal" ${LLAMA_METAL_DEFAULT})
9495
option(LLAMA_METAL_NDEBUG "llama: disable Metal debugging" OFF)
9596
option(LLAMA_MPI "llama: use MPI" OFF)
96-
option(LLAMA_K_QUANTS "llama: use k-quants" ON)
9797
option(LLAMA_QKK_64 "llama: use super-block size of 64 for k-quants" OFF)
9898

9999
option(LLAMA_BUILD_TESTS "llama: build tests" ${LLAMA_STANDALONE})
100100
option(LLAMA_BUILD_EXAMPLES "llama: build examples" ${LLAMA_STANDALONE})
101101
option(LLAMA_BUILD_SERVER "llama: build server example" ON)
102102

103-
#
104-
# Build info header
105-
#
106-
107-
# Generate initial build-info.h
108-
include(${CMAKE_CURRENT_SOURCE_DIR}/scripts/build-info.cmake)
109-
110-
if(EXISTS "${CMAKE_CURRENT_SOURCE_DIR}/.git")
111-
set(GIT_DIR "${CMAKE_CURRENT_SOURCE_DIR}/.git")
112-
113-
# Is git submodule
114-
if(NOT IS_DIRECTORY "${GIT_DIR}")
115-
file(READ ${GIT_DIR} REAL_GIT_DIR_LINK)
116-
string(REGEX REPLACE "gitdir: (.*)\n$" "\\1" REAL_GIT_DIR ${REAL_GIT_DIR_LINK})
117-
set(GIT_DIR "${CMAKE_CURRENT_SOURCE_DIR}/${REAL_GIT_DIR}")
118-
endif()
119-
120-
# Add a custom target for build-info.h
121-
add_custom_target(BUILD_INFO ALL DEPENDS "${CMAKE_CURRENT_SOURCE_DIR}/build-info.h")
122-
123-
# Add a custom command to rebuild build-info.h when .git/index changes
124-
add_custom_command(
125-
OUTPUT "${CMAKE_CURRENT_SOURCE_DIR}/build-info.h"
126-
COMMENT "Generating build details from Git"
127-
COMMAND ${CMAKE_COMMAND} -DMSVC=${MSVC} -DCMAKE_C_COMPILER_VERSION=${CMAKE_C_COMPILER_VERSION} -DCMAKE_C_COMPILER_ID=${CMAKE_C_COMPILER_ID} -DCMAKE_VS_PLATFORM_NAME=${CMAKE_VS_PLATFORM_NAME} -DCMAKE_C_COMPILER=${CMAKE_C_COMPILER} -P "${CMAKE_CURRENT_SOURCE_DIR}/scripts/build-info.cmake"
128-
WORKING_DIRECTORY ${CMAKE_CURRENT_SOURCE_DIR}
129-
DEPENDS "${GIT_DIR}/index"
130-
VERBATIM
131-
)
132-
else()
133-
message(WARNING "Git repository not found; to enable automatic generation of build info, make sure Git is installed and the project is a Git repository.")
134-
endif()
135-
136103
#
137104
# Compile flags
138105
#
@@ -277,13 +244,8 @@ if (LLAMA_BLAS)
277244
endif()
278245
endif()
279246

280-
if (LLAMA_K_QUANTS)
281-
set(GGML_HEADERS_EXTRA k_quants.h)
282-
set(GGML_SOURCES_EXTRA k_quants.c)
283-
add_compile_definitions(GGML_USE_K_QUANTS)
284-
if (LLAMA_QKK_64)
285-
add_compile_definitions(GGML_QKK_64)
286-
endif()
247+
if (LLAMA_QKK_64)
248+
add_compile_definitions(GGML_QKK_64)
287249
endif()
288250

289251
if (LLAMA_CUBLAS)
@@ -305,6 +267,9 @@ if (LLAMA_CUBLAS)
305267
if (LLAMA_CUDA_FORCE_DMMV)
306268
add_compile_definitions(GGML_CUDA_FORCE_DMMV)
307269
endif()
270+
if (LLAMA_CUDA_FORCE_MMQ)
271+
add_compile_definitions(GGML_CUDA_FORCE_MMQ)
272+
endif()
308273
add_compile_definitions(GGML_CUDA_DMMV_X=${LLAMA_CUDA_DMMV_X})
309274
add_compile_definitions(GGML_CUDA_MMV_Y=${LLAMA_CUDA_MMV_Y})
310275
if (DEFINED LLAMA_CUDA_DMMV_Y)
@@ -331,6 +296,7 @@ if (LLAMA_CUBLAS)
331296
set(CMAKE_CUDA_ARCHITECTURES "60;61;70") # needed for f16 CUDA intrinsics
332297
else()
333298
set(CMAKE_CUDA_ARCHITECTURES "52;61;70") # lowest CUDA 12 standard + lowest for integer intrinsics
299+
#set(CMAKE_CUDA_ARCHITECTURES "") # use this to compile much faster, but only F16 models work
334300
endif()
335301
endif()
336302
message(STATUS "Using CUDA architectures: ${CMAKE_CUDA_ARCHITECTURES}")
@@ -404,6 +370,9 @@ if (LLAMA_HIPBLAS)
404370
if (LLAMA_CUDA_FORCE_DMMV)
405371
target_compile_definitions(ggml-rocm PRIVATE GGML_CUDA_FORCE_DMMV)
406372
endif()
373+
if (LLAMA_CUDA_FORCE_MMQ)
374+
target_compile_definitions(ggml-rocm PRIVATE GGML_CUDA_FORCE_MMQ)
375+
endif()
407376
target_compile_definitions(ggml-rocm PRIVATE GGML_CUDA_DMMV_X=${LLAMA_CUDA_DMMV_X})
408377
target_compile_definitions(ggml-rocm PRIVATE GGML_CUDA_MMV_Y=${LLAMA_CUDA_MMV_Y})
409378
target_compile_definitions(ggml-rocm PRIVATE K_QUANTS_PER_ITERATION=${LLAMA_CUDA_KQUANTS_ITER})
@@ -665,6 +634,8 @@ add_library(ggml OBJECT
665634
ggml-alloc.h
666635
ggml-backend.c
667636
ggml-backend.h
637+
ggml-quants.c
638+
ggml-quants.h
668639
${GGML_SOURCES_CUDA} ${GGML_HEADERS_CUDA}
669640
${GGML_SOURCES_OPENCL} ${GGML_HEADERS_OPENCL}
670641
${GGML_SOURCES_METAL} ${GGML_HEADERS_METAL}

‎Makefile

Lines changed: 46 additions & 46 deletions
Original file line numberDiff line numberDiff line change
@@ -342,13 +342,9 @@ else
342342
MK_CXXFLAGS += -march=rv64gcv -mabi=lp64d
343343
endif
344344

345-
ifndef LLAMA_NO_K_QUANTS
346-
MK_CPPFLAGS += -DGGML_USE_K_QUANTS
347-
OBJS += k_quants.o
348345
ifdef LLAMA_QKK_64
349346
MK_CPPFLAGS += -DGGML_QKK_64
350347
endif
351-
endif
352348

353349
ifndef LLAMA_NO_ACCELERATE
354350
# Mac OS - include Accelerate framework.
@@ -365,7 +361,7 @@ ifdef LLAMA_MPI
365361
MK_CPPFLAGS += -DGGML_USE_MPI
366362
MK_CFLAGS += -Wno-cast-qual
367363
MK_CXXFLAGS += -Wno-cast-qual
368-
OBJS += ggml-mpi.o
364+
OBJS += ggml-mpi.o
369365
endif # LLAMA_MPI
370366

371367
ifdef LLAMA_OPENBLAS
@@ -382,7 +378,7 @@ endif # LLAMA_BLIS
382378
ifdef LLAMA_CUBLAS
383379
MK_CPPFLAGS += -DGGML_USE_CUBLAS -I/usr/local/cuda/include -I/opt/cuda/include -I$(CUDA_PATH)/targets/x86_64-linux/include
384380
MK_LDFLAGS += -lcublas -lculibos -lcudart -lcublasLt -lpthread -ldl -lrt -L/usr/local/cuda/lib64 -L/opt/cuda/lib64 -L$(CUDA_PATH)/targets/x86_64-linux/lib
385-
OBJS += ggml-cuda.o
381+
OBJS += ggml-cuda.o
386382
NVCCFLAGS = --forward-unknown-to-host-compiler -use_fast_math
387383
ifdef LLAMA_CUDA_NVCC
388384
NVCC = $(LLAMA_CUDA_NVCC)
@@ -397,6 +393,9 @@ endif # CUDA_DOCKER_ARCH
397393
ifdef LLAMA_CUDA_FORCE_DMMV
398394
NVCCFLAGS += -DGGML_CUDA_FORCE_DMMV
399395
endif # LLAMA_CUDA_FORCE_DMMV
396+
ifdef LLAMA_CUDA_FORCE_MMQ
397+
NVCCFLAGS += -DGGML_CUDA_FORCE_MMQ
398+
endif # LLAMA_CUDA_FORCE_MMQ
400399
ifdef LLAMA_CUDA_DMMV_X
401400
NVCCFLAGS += -DGGML_CUDA_DMMV_X=$(LLAMA_CUDA_DMMV_X)
402401
else
@@ -494,11 +493,6 @@ ggml-mpi.o: ggml-mpi.c ggml-mpi.h
494493
$(CC) $(CFLAGS) -c $< -o $@
495494
endif # LLAMA_MPI
496495

497-
ifndef LLAMA_NO_K_QUANTS
498-
k_quants.o: k_quants.c k_quants.h
499-
$(CC) $(CFLAGS) -c $< -o $@
500-
endif # LLAMA_NO_K_QUANTS
501-
502496
# combine build flags with cmdline overrides
503497
override CFLAGS := $(MK_CPPFLAGS) $(CPPFLAGS) $(MK_CFLAGS) $(CFLAGS)
504498
override CXXFLAGS := $(MK_CPPFLAGS) $(CPPFLAGS) $(MK_CXXFLAGS) $(CXXFLAGS)
@@ -539,13 +533,16 @@ ggml-alloc.o: ggml-alloc.c ggml.h ggml-alloc.h
539533
ggml-backend.o: ggml-backend.c ggml.h ggml-backend.h
540534
$(CC) $(CFLAGS) -c $< -o $@
541535

542-
OBJS += ggml-alloc.o ggml-backend.o
536+
ggml-quants.o: ggml-quants.c ggml.h ggml-quants.h
537+
$(CC) $(CFLAGS) -c $< -o $@
538+
539+
OBJS += ggml-alloc.o ggml-backend.o ggml-quants.o
543540

544541
llama.o: llama.cpp ggml.h ggml-alloc.h ggml-backend.h ggml-cuda.h ggml-metal.h llama.h
545542
$(CXX) $(CXXFLAGS) -c $< -o $@
546543

547-
COMMON_H_DEPS = common/common.h common/sampling.h build-info.h common/log.h
548-
COMMON_DEPS = $(COMMON_H_DEPS) common.o sampling.o grammar-parser.o
544+
COMMON_H_DEPS = common/common.h common/sampling.h common/log.h
545+
COMMON_DEPS = common.o sampling.o grammar-parser.o build-info.o
549546

550547
common.o: common/common.cpp $(COMMON_H_DEPS)
551548
$(CXX) $(CXXFLAGS) -c $< -o $@
@@ -566,46 +563,46 @@ libllama.so: llama.o ggml.o $(OBJS)
566563
$(CXX) $(CXXFLAGS) -shared -fPIC -o $@ $^ $(LDFLAGS)
567564

568565
clean:
569-
rm -vrf *.o tests/*.o *.so *.dll benchmark-matmult build-info.h *.dot $(COV_TARGETS) $(BUILD_TARGETS) $(TEST_TARGETS)
566+
rm -vrf *.o tests/*.o *.so *.dll benchmark-matmult common/build-info.cpp *.dot $(COV_TARGETS) $(BUILD_TARGETS) $(TEST_TARGETS)
570567

571568
#
572569
# Examples
573570
#
574571

575-
main: examples/main/main.cpp build-info.h ggml.o llama.o $(COMMON_DEPS) console.o grammar-parser.o $(OBJS)
572+
main: examples/main/main.cpp ggml.o llama.o $(COMMON_DEPS) console.o grammar-parser.o $(OBJS)
576573
$(CXX) $(CXXFLAGS) $(filter-out %.h,$^) -o $@ $(LDFLAGS)
577574
@echo
578575
@echo '==== Run ./main -h for help. ===='
579576
@echo
580577

581-
infill: examples/infill/infill.cpp build-info.h ggml.o llama.o $(COMMON_DEPS) console.o grammar-parser.o $(OBJS)
578+
infill: examples/infill/infill.cpp ggml.o llama.o $(COMMON_DEPS) console.o grammar-parser.o $(OBJS)
582579
$(CXX) $(CXXFLAGS) $(filter-out %.h,$^) -o $@ $(LDFLAGS)
583580

584-
simple: examples/simple/simple.cpp build-info.h ggml.o llama.o $(COMMON_DEPS) $(OBJS)
581+
simple: examples/simple/simple.cpp ggml.o llama.o $(COMMON_DEPS) $(OBJS)
585582
$(CXX) $(CXXFLAGS) $(filter-out %.h,$^) -o $@ $(LDFLAGS)
586583

587-
batched: examples/batched/batched.cpp build-info.h ggml.o llama.o $(COMMON_DEPS) $(OBJS)
584+
batched: examples/batched/batched.cpp ggml.o llama.o $(COMMON_DEPS) $(OBJS)
588585
$(CXX) $(CXXFLAGS) $(filter-out %.h,$^) -o $@ $(LDFLAGS)
589586

590-
batched-bench: examples/batched-bench/batched-bench.cpp build-info.h ggml.o llama.o common.o $(OBJS)
587+
batched-bench: examples/batched-bench/batched-bench.cpp build-info.o ggml.o llama.o common.o $(OBJS)
591588
$(CXX) $(CXXFLAGS) $(filter-out %.h,$^) -o $@ $(LDFLAGS)
592589

593-
quantize: examples/quantize/quantize.cpp build-info.h ggml.o llama.o $(OBJS)
590+
quantize: examples/quantize/quantize.cpp build-info.o ggml.o llama.o $(OBJS)
594591
$(CXX) $(CXXFLAGS) $(filter-out %.h,$^) -o $@ $(LDFLAGS)
595592

596-
quantize-stats: examples/quantize-stats/quantize-stats.cpp build-info.h ggml.o llama.o $(OBJS)
593+
quantize-stats: examples/quantize-stats/quantize-stats.cpp build-info.o ggml.o llama.o $(OBJS)
597594
$(CXX) $(CXXFLAGS) $(filter-out %.h,$^) -o $@ $(LDFLAGS)
598595

599-
perplexity: examples/perplexity/perplexity.cpp build-info.h ggml.o llama.o $(COMMON_DEPS) $(OBJS)
596+
perplexity: examples/perplexity/perplexity.cpp ggml.o llama.o $(COMMON_DEPS) $(OBJS)
600597
$(CXX) $(CXXFLAGS) $(filter-out %.h,$^) -o $@ $(LDFLAGS)
601598

602-
embedding: examples/embedding/embedding.cpp build-info.h ggml.o llama.o $(COMMON_DEPS) $(OBJS)
599+
embedding: examples/embedding/embedding.cpp ggml.o llama.o $(COMMON_DEPS) $(OBJS)
603600
$(CXX) $(CXXFLAGS) $(filter-out %.h,$^) -o $@ $(LDFLAGS)
604601

605-
save-load-state: examples/save-load-state/save-load-state.cpp build-info.h ggml.o llama.o $(COMMON_DEPS) $(OBJS)
602+
save-load-state: examples/save-load-state/save-load-state.cpp ggml.o llama.o $(COMMON_DEPS) $(OBJS)
606603
$(CXX) $(CXXFLAGS) $(filter-out %.h,$^) -o $@ $(LDFLAGS)
607604

608-
server: examples/server/server.cpp examples/server/httplib.h examples/server/json.hpp examples/server/index.html.hpp examples/server/index.js.hpp examples/server/completion.js.hpp examples/llava/clip.cpp examples/llava/clip.h common/stb_image.h build-info.h ggml.o llama.o $(COMMON_DEPS) grammar-parser.o $(OBJS)
605+
server: examples/server/server.cpp examples/server/httplib.h examples/server/json.hpp examples/server/index.html.hpp examples/server/index.js.hpp examples/server/completion.js.hpp examples/llava/clip.cpp examples/llava/clip.h common/stb_image.h ggml.o llama.o $(COMMON_DEPS) grammar-parser.o $(OBJS)
609606
$(CXX) $(CXXFLAGS) -Iexamples/server $(filter-out %.h,$(filter-out %.hpp,$^)) -o $@ $(LDFLAGS) $(LWINSOCK2) -Wno-cast-qual
610607

611608
gguf: examples/gguf/gguf.cpp ggml.o llama.o $(OBJS)
@@ -617,7 +614,7 @@ train-text-from-scratch: examples/train-text-from-scratch/train-text-from-scratc
617614
convert-llama2c-to-ggml: examples/convert-llama2c-to-ggml/convert-llama2c-to-ggml.cpp ggml.o llama.o $(OBJS)
618615
$(CXX) $(CXXFLAGS) $(filter-out %.h,$^) -o $@ $(LDFLAGS)
619616

620-
llama-bench: examples/llama-bench/llama-bench.cpp build-info.h ggml.o llama.o $(COMMON_DEPS) $(OBJS)
617+
llama-bench: examples/llama-bench/llama-bench.cpp ggml.o llama.o $(COMMON_DEPS) $(OBJS)
621618
$(CXX) $(CXXFLAGS) $(filter-out %.h,$^) -o $@ $(LDFLAGS)
622619

623620
llava: examples/llava/llava.cpp examples/llava/llava-utils.h examples/llava/clip.cpp examples/llava/clip.h common/stb_image.h ggml.o llama.o $(COMMON_DEPS) $(OBJS)
@@ -626,19 +623,19 @@ llava: examples/llava/llava.cpp examples/llava/llava-utils.h examples/llava/clip
626623
baby-llama: examples/baby-llama/baby-llama.cpp ggml.o llama.o $(COMMON_DEPS) train.o $(OBJS)
627624
$(CXX) $(CXXFLAGS) $(filter-out %.h,$^) -o $@ $(LDFLAGS)
628625

629-
beam-search: examples/beam-search/beam-search.cpp build-info.h ggml.o llama.o $(COMMON_DEPS) $(OBJS)
626+
beam-search: examples/beam-search/beam-search.cpp ggml.o llama.o $(COMMON_DEPS) $(OBJS)
630627
$(CXX) $(CXXFLAGS) $(filter-out %.h,$^) -o $@ $(LDFLAGS)
631628

632-
finetune: examples/finetune/finetune.cpp build-info.h ggml.o llama.o $(COMMON_DEPS) train.o $(OBJS)
629+
finetune: examples/finetune/finetune.cpp ggml.o llama.o $(COMMON_DEPS) train.o $(OBJS)
633630
$(CXX) $(CXXFLAGS) $(filter-out %.h,$^) -o $@ $(LDFLAGS)
634631

635-
export-lora: examples/export-lora/export-lora.cpp build-info.h ggml.o llama.o $(COMMON_DEPS) $(OBJS)
632+
export-lora: examples/export-lora/export-lora.cpp ggml.o llama.o $(COMMON_DEPS) $(OBJS)
636633
$(CXX) $(CXXFLAGS) $(filter-out %.h,$^) -o $@ $(LDFLAGS)
637634

638-
speculative: examples/speculative/speculative.cpp build-info.h ggml.o llama.o $(COMMON_DEPS) grammar-parser.o $(OBJS)
635+
speculative: examples/speculative/speculative.cpp ggml.o llama.o $(COMMON_DEPS) grammar-parser.o $(OBJS)
639636
$(CXX) $(CXXFLAGS) $(filter-out %.h,$^) -o $@ $(LDFLAGS)
640637

641-
parallel: examples/parallel/parallel.cpp build-info.h ggml.o llama.o $(COMMON_DEPS) $(OBJS)
638+
parallel: examples/parallel/parallel.cpp ggml.o llama.o $(COMMON_DEPS) $(OBJS)
642639
$(CXX) $(CXXFLAGS) $(filter-out %.h,$^) -o $@ $(LDFLAGS)
643640

644641
ifdef LLAMA_METAL
@@ -651,21 +648,24 @@ swift: examples/batched.swift
651648
(cd examples/batched.swift; make build)
652649
endif
653650

654-
build-info.h: $(wildcard .git/index) scripts/build-info.sh
651+
common/build-info.cpp: $(wildcard .git/index) scripts/build-info.sh
655652
@sh scripts/build-info.sh $(CC) > $@.tmp
656653
@if ! cmp -s $@.tmp $@; then \
657654
mv $@.tmp $@; \
658655
else \
659656
rm $@.tmp; \
660657
fi
661658

659+
build-info.o: common/build-info.cpp
660+
$(CXX) $(CXXFLAGS) -c $(filter-out %.h,$^) -o $@
661+
662662
#
663663
# Tests
664664
#
665665

666666
tests: $(TEST_TARGETS)
667667

668-
benchmark-matmult: examples/benchmark/benchmark-matmult.cpp build-info.h ggml.o $(OBJS)
668+
benchmark-matmult: examples/benchmark/benchmark-matmult.cpp build-info.o ggml.o $(OBJS)
669669
$(CXX) $(CXXFLAGS) $(filter-out %.h,$^) -o $@ $(LDFLAGS)
670670

671671
run-benchmark-matmult: benchmark-matmult
@@ -679,40 +679,40 @@ vdot: pocs/vdot/vdot.cpp ggml.o $(OBJS)
679679
q8dot: pocs/vdot/q8dot.cpp ggml.o $(OBJS)
680680
$(CXX) $(CXXFLAGS) $^ -o $@ $(LDFLAGS)
681681

682-
tests/test-llama-grammar: tests/test-llama-grammar.cpp build-info.h ggml.o $(COMMON_DEPS) grammar-parser.o $(OBJS)
682+
tests/test-llama-grammar: tests/test-llama-grammar.cpp ggml.o $(COMMON_DEPS) grammar-parser.o $(OBJS)
683683
$(CXX) $(CXXFLAGS) $(filter-out %.h,$^) -o $@ $(LDFLAGS)
684684

685-
tests/test-grammar-parser: tests/test-grammar-parser.cpp build-info.h ggml.o llama.o $(COMMON_DEPS) grammar-parser.o $(OBJS)
685+
tests/test-grammar-parser: tests/test-grammar-parser.cpp ggml.o llama.o $(COMMON_DEPS) grammar-parser.o $(OBJS)
686686
$(CXX) $(CXXFLAGS) $(filter-out %.h,$^) -o $@ $(LDFLAGS)
687687

688-
tests/test-double-float: tests/test-double-float.cpp build-info.h ggml.o llama.o $(COMMON_DEPS) $(OBJS)
688+
tests/test-double-float: tests/test-double-float.cpp ggml.o llama.o $(COMMON_DEPS) $(OBJS)
689689
$(CXX) $(CXXFLAGS) $(filter-out %.h,$^) -o $@ $(LDFLAGS)
690690

691-
tests/test-grad0: tests/test-grad0.cpp build-info.h ggml.o llama.o $(COMMON_DEPS) $(OBJS)
691+
tests/test-grad0: tests/test-grad0.cpp ggml.o llama.o $(COMMON_DEPS) $(OBJS)
692692
$(CXX) $(CXXFLAGS) $(filter-out %.h,$^) -o $@ $(LDFLAGS)
693693

694-
tests/test-opt: tests/test-opt.cpp build-info.h ggml.o llama.o $(COMMON_DEPS) $(OBJS)
694+
tests/test-opt: tests/test-opt.cpp ggml.o llama.o $(COMMON_DEPS) $(OBJS)
695695
$(CXX) $(CXXFLAGS) $(filter-out %.h,$^) -o $@ $(LDFLAGS)
696696

697-
tests/test-quantize-fns: tests/test-quantize-fns.cpp build-info.h ggml.o llama.o $(COMMON_DEPS) $(OBJS)
697+
tests/test-quantize-fns: tests/test-quantize-fns.cpp ggml.o llama.o $(COMMON_DEPS) $(OBJS)
698698
$(CXX) $(CXXFLAGS) $(filter-out %.h,$^) -o $@ $(LDFLAGS)
699699

700-
tests/test-quantize-perf: tests/test-quantize-perf.cpp build-info.h ggml.o llama.o $(COMMON_DEPS) $(OBJS)
700+
tests/test-quantize-perf: tests/test-quantize-perf.cpp ggml.o llama.o $(COMMON_DEPS) $(OBJS)
701701
$(CXX) $(CXXFLAGS) $(filter-out %.h,$^) -o $@ $(LDFLAGS)
702702

703-
tests/test-sampling: tests/test-sampling.cpp build-info.h ggml.o llama.o $(COMMON_DEPS) $(OBJS)
703+
tests/test-sampling: tests/test-sampling.cpp ggml.o llama.o $(COMMON_DEPS) $(OBJS)
704704
$(CXX) $(CXXFLAGS) $(filter-out %.h,$^) -o $@ $(LDFLAGS)
705705

706-
tests/test-tokenizer-0-falcon: tests/test-tokenizer-0-falcon.cpp build-info.h ggml.o llama.o $(COMMON_DEPS) $(OBJS)
706+
tests/test-tokenizer-0-falcon: tests/test-tokenizer-0-falcon.cpp ggml.o llama.o $(COMMON_DEPS) $(OBJS)
707707
$(CXX) $(CXXFLAGS) $(filter-out %.h,$^) -o $@ $(LDFLAGS)
708708

709-
tests/test-tokenizer-0-llama: tests/test-tokenizer-0-llama.cpp build-info.h ggml.o llama.o $(COMMON_DEPS) $(OBJS)
709+
tests/test-tokenizer-0-llama: tests/test-tokenizer-0-llama.cpp ggml.o llama.o $(COMMON_DEPS) $(OBJS)
710710
$(CXX) $(CXXFLAGS) $(filter-out %.h,$^) -o $@ $(LDFLAGS)
711711

712-
tests/test-tokenizer-1-bpe: tests/test-tokenizer-1-bpe.cpp build-info.h ggml.o llama.o $(COMMON_DEPS) $(OBJS)
712+
tests/test-tokenizer-1-bpe: tests/test-tokenizer-1-bpe.cpp ggml.o llama.o $(COMMON_DEPS) $(OBJS)
713713
$(CXX) $(CXXFLAGS) $(filter-out %.h,$^) -o $@ $(LDFLAGS)
714714

715-
tests/test-tokenizer-1-llama: tests/test-tokenizer-1-llama.cpp build-info.h ggml.o llama.o $(COMMON_DEPS) $(OBJS)
715+
tests/test-tokenizer-1-llama: tests/test-tokenizer-1-llama.cpp ggml.o llama.o $(COMMON_DEPS) $(OBJS)
716716
$(CXX) $(CXXFLAGS) $(filter-out %.h,$^) -o $@ $(LDFLAGS)
717717

718718
tests/test-c.o: tests/test-c.c llama.h

0 commit comments

Comments
 (0)