tqcq/llama.cpp
0
0
Fork 0
mirror of https://github.com/ggml-org/llama.cpp.git synced 2025-08-17 13:40:55 -04:00
Code Issues Packages Projects Releases Wiki Activity
Files
compilade/convert-prequant
llama.cpp/tests/CMakeLists.txt
Jonathan Graehl 5cdb27e091 finetune: SGD optimizer, more CLI args (#13873) 
* examples/finetune -opt SGD (stochastic gradient descent) memory opt

add unit tested GGML_OPT_OPTIMIZER_SGD to ggml - avoids allocating
m, v tensors.

support finetune.cpp arg -opt SGD (or sgd). (default adamw as before)

llama 3.2-1b-F32 result: observed 11gb gpu ram (41 sec/epoch)
when using SGD instead of 19gb (55 sec/epoch) using adamw.
(wikipedia 100 lines finetune)

(
using the same GPU memory, adamw can only do before OOM 512
batch/context, reaching:
train: [███████▉] data=0000140/0000140 loss=0.02575±0.00099 acc=99.52±0.03% t=00:00:47 ETA=00:00:00
val:   [███████▉] data=0000008/0000008 loss=4.76565±0.28810 acc=41.46±0.77% t=00:00:00 ETA=00:00:00

SGD is superior, though it converges slower, with max before OOM 1728
batch/context (esp see the better validation perf):
train: [███████▉] data=0000039/0000039 loss=0.00371±0.00010 acc=99.96±0.01% t=00:00:41 ETA=00:00:00
val:   [███████▉] data=0000003/0000003 loss=5.11406±0.76034 acc=48.01±0.69% t=00:00:01 ETA=00:00:00
)

note: when finetuning long enough (or w/ enough -lr),
validation accuracy *eventually* drops ('catastrophic forgetting')

-lr-half (halflife) option useful for SGD to avoid oscillation or
super slow underdamped learning (makes setting -lr more forgiving).
terminal -lr for now is set by lr-halvings i.e. if you want at most
1/8 the inital -lr you set -lr-halvings 3.

note: objective loss not directly comparable between adamw, sgd? -
check perplexity or accuracy or consider relative improvements
for convergence

new finetune args -wd 1e-9 to enable weight decay in sgd or adamw,
and max -epochs N (default 2 as before)

cache (1 - wd*alpha) in 'adamw' opt struct -
no noticeable perf benefit, disabled (still done
for new SGD though)

since opt. memory is pre-allocated, the ggml_opt_get_optimizer_params
would probably be able to change between SGD and AdamW with each epoch
but would need to use adamw for the first (unconfirmed - no cmdline arg
to set such a policy yet)

test-opt checks adamw as before and now sgd (except for a few disabled
tests for sgd only; probably just needs logging values and adding
alternate reference values);  tolerance on the 'regression'
test is broader for sgd (so we don't need many more epochs)

* Vulkan: Implement GGML_OP_OPT_STEP_SGD

* tests: Fix OPT_STEP_SGD test-backend-ops

* SGD op param store weight-decay and not 1-alpha*wd

* minor + cosmetic changes

* fix vulkan sgd

* try CI fix

---------

Co-authored-by: 0cc4m <picard12@live.de>
Co-authored-by: Johannes Gäßler <johannesg@5d6.de>
2025-08-14 12:03:57 +02:00

222 lines
9.4 KiB
CMake
Raw Permalink Blame History

llama_add_compile_flags()
function(llama_build source)
if (DEFINED LLAMA_TEST_NAME)
set(TEST_TARGET ${LLAMA_TEST_NAME})
else()
get_filename_component(TEST_TARGET ${source} NAME_WE)
endif()
add_executable(${TEST_TARGET} ${source})
target_link_libraries(${TEST_TARGET} PRIVATE common)
install(TARGETS ${TEST_TARGET} RUNTIME)
endfunction()
function(llama_test target)
include(CMakeParseArguments)
set(options)
set(oneValueArgs NAME LABEL WORKING_DIRECTORY)
set(multiValueArgs ARGS)
cmake_parse_arguments(LLAMA_TEST "${options}" "${oneValueArgs}" "${multiValueArgs}" ${ARGN})
if (NOT DEFINED LLAMA_TEST_LABEL)
set(LLAMA_TEST_LABEL "main")
endif()
if (NOT DEFINED LLAMA_TEST_WORKING_DIRECTORY)
set(LLAMA_TEST_WORKING_DIRECTORY .)
endif()
if (DEFINED LLAMA_TEST_NAME)
set(TEST_NAME ${LLAMA_TEST_NAME})
else()
set(TEST_NAME ${target})
endif()
set(TEST_TARGET ${target})
add_test(
NAME ${TEST_NAME}
WORKING_DIRECTORY ${LLAMA_TEST_WORKING_DIRECTORY}
COMMAND $<TARGET_FILE:${TEST_TARGET}>
${LLAMA_TEST_ARGS})
set_property(TEST ${TEST_NAME} PROPERTY LABELS ${LLAMA_TEST_LABEL})
endfunction()
function(llama_test_cmd target)
include(CMakeParseArguments)
set(options)
set(oneValueArgs NAME LABEL WORKING_DIRECTORY)
set(multiValueArgs ARGS)
cmake_parse_arguments(LLAMA_TEST "${options}" "${oneValueArgs}" "${multiValueArgs}" ${ARGN})
if (NOT DEFINED LLAMA_TEST_LABEL)
set(LLAMA_TEST_LABEL "main")
endif()
if (NOT DEFINED LLAMA_TEST_WORKING_DIRECTORY)
set(LLAMA_TEST_WORKING_DIRECTORY .)
endif()
if (DEFINED LLAMA_TEST_NAME)
set(TEST_NAME ${LLAMA_TEST_NAME})
else()
set(TEST_NAME ${target})
endif()
add_test(
NAME ${TEST_NAME}
WORKING_DIRECTORY ${LLAMA_TEST_WORKING_DIRECTORY}
COMMAND ${target}
${LLAMA_TEST_ARGS})
set_property(TEST ${TEST_NAME} PROPERTY LABELS ${LLAMA_TEST_LABEL})
endfunction()
# Builds and runs a test source file.
# Optional args:
# - NAME: name of the executable & test target (defaults to the source file name without extension)
# - LABEL: label for the test (defaults to main)
# - ARGS: arguments to pass to the test executable
# - WORKING_DIRECTORY
function(llama_build_and_test source)
include(CMakeParseArguments)
set(options)
set(oneValueArgs NAME LABEL WORKING_DIRECTORY)
set(multiValueArgs ARGS)
cmake_parse_arguments(LLAMA_TEST "${options}" "${oneValueArgs}" "${multiValueArgs}" ${ARGN})
if (NOT DEFINED LLAMA_TEST_LABEL)
set(LLAMA_TEST_LABEL "main")
endif()
if (NOT DEFINED LLAMA_TEST_WORKING_DIRECTORY)
set(LLAMA_TEST_WORKING_DIRECTORY .)
endif()
if (DEFINED LLAMA_TEST_NAME)
set(TEST_TARGET ${LLAMA_TEST_NAME})
else()
get_filename_component(TEST_TARGET ${source} NAME_WE)
endif()
add_executable(${TEST_TARGET} ${source} get-model.cpp)
install(TARGETS ${TEST_TARGET} RUNTIME)
target_link_libraries(${TEST_TARGET} PRIVATE common)
add_test(
NAME ${TEST_TARGET}
WORKING_DIRECTORY ${LLAMA_TEST_WORKING_DIRECTORY}
COMMAND $<TARGET_FILE:${TEST_TARGET}>
${LLAMA_TEST_ARGS})
set_property(TEST ${TEST_TARGET} PROPERTY LABELS ${LLAMA_TEST_LABEL})
endfunction()
# build test-tokenizer-0 target once and add many tests
llama_build(test-tokenizer-0.cpp)
llama_test(test-tokenizer-0 NAME test-tokenizer-0-bert-bge ARGS ${PROJECT_SOURCE_DIR}/models/ggml-vocab-bert-bge.gguf)
llama_test(test-tokenizer-0 NAME test-tokenizer-0-command-r ARGS ${PROJECT_SOURCE_DIR}/models/ggml-vocab-command-r.gguf)
llama_test(test-tokenizer-0 NAME test-tokenizer-0-deepseek-coder ARGS ${PROJECT_SOURCE_DIR}/models/ggml-vocab-deepseek-coder.gguf)
llama_test(test-tokenizer-0 NAME test-tokenizer-0-deepseek-llm ARGS ${PROJECT_SOURCE_DIR}/models/ggml-vocab-deepseek-llm.gguf)
llama_test(test-tokenizer-0 NAME test-tokenizer-0-falcon ARGS ${PROJECT_SOURCE_DIR}/models/ggml-vocab-falcon.gguf)
llama_test(test-tokenizer-0 NAME test-tokenizer-0-gpt-2 ARGS ${PROJECT_SOURCE_DIR}/models/ggml-vocab-gpt-2.gguf)
llama_test(test-tokenizer-0 NAME test-tokenizer-0-llama-bpe ARGS ${PROJECT_SOURCE_DIR}/models/ggml-vocab-llama-bpe.gguf)
llama_test(test-tokenizer-0 NAME test-tokenizer-0-llama-spm ARGS ${PROJECT_SOURCE_DIR}/models/ggml-vocab-llama-spm.gguf)
llama_test(test-tokenizer-0 NAME test-tokenizer-0-mpt ARGS ${PROJECT_SOURCE_DIR}/models/ggml-vocab-mpt.gguf)
llama_test(test-tokenizer-0 NAME test-tokenizer-0-phi-3 ARGS ${PROJECT_SOURCE_DIR}/models/ggml-vocab-phi-3.gguf)
llama_test(test-tokenizer-0 NAME test-tokenizer-0-qwen2 ARGS ${PROJECT_SOURCE_DIR}/models/ggml-vocab-qwen2.gguf)
llama_test(test-tokenizer-0 NAME test-tokenizer-0-refact ARGS ${PROJECT_SOURCE_DIR}/models/ggml-vocab-refact.gguf)
llama_test(test-tokenizer-0 NAME test-tokenizer-0-starcoder ARGS ${PROJECT_SOURCE_DIR}/models/ggml-vocab-starcoder.gguf)
if (NOT WIN32)
llama_test_cmd(
${CMAKE_CURRENT_SOURCE_DIR}/test-tokenizers-repo.sh
NAME test-tokenizers-ggml-vocabs
WORKING_DIRECTORY ${CMAKE_RUNTIME_OUTPUT_DIRECTORY}
ARGS https://huggingface.co/ggml-org/vocabs ${PROJECT_SOURCE_DIR}/models/ggml-vocabs
)
endif()
if (LLAMA_LLGUIDANCE)
llama_build_and_test(test-grammar-llguidance.cpp ARGS ${PROJECT_SOURCE_DIR}/models/ggml-vocab-llama-bpe.gguf)
endif ()
if (NOT WIN32 OR NOT BUILD_SHARED_LIBS)
# these tests are disabled on Windows because they use internal functions not exported with LLAMA_API (when building with shared libraries)
llama_build_and_test(test-sampling.cpp)
llama_build_and_test(test-grammar-parser.cpp)
llama_build_and_test(test-grammar-integration.cpp)
llama_build_and_test(test-llama-grammar.cpp)
llama_build_and_test(test-chat.cpp)
# TODO: disabled on loongarch64 because the ggml-ci node lacks Python 3.8
if (NOT ${CMAKE_SYSTEM_PROCESSOR} MATCHES "loongarch64")
llama_build_and_test(test-json-schema-to-grammar.cpp WORKING_DIRECTORY ${PROJECT_SOURCE_DIR})
target_include_directories(test-json-schema-to-grammar PRIVATE ${PROJECT_SOURCE_DIR}/tools/server)
endif()
if (NOT GGML_BACKEND_DL)
llama_build(test-quantize-stats.cpp)
endif()
llama_build(test-gbnf-validator.cpp)
# build test-tokenizer-1-bpe target once and add many tests
llama_build(test-tokenizer-1-bpe.cpp)
# TODO: disabled due to slowness
#llama_test(test-tokenizer-1-bpe NAME test-tokenizer-1-aquila ARGS ${PROJECT_SOURCE_DIR}/models/ggml-vocab-aquila.gguf)
#llama_test(test-tokenizer-1-bpe NAME test-tokenizer-1-falcon ARGS ${PROJECT_SOURCE_DIR}/models/ggml-vocab-falcon.gguf)
#llama_test(test-tokenizer-1-bpe NAME test-tokenizer-1-gpt-2 ARGS ${PROJECT_SOURCE_DIR}/models/ggml-vocab-gpt-2.gguf)
#llama_test(test-tokenizer-1-bpe NAME test-tokenizer-1-gpt-neox ARGS ${PROJECT_SOURCE_DIR}/models/ggml-vocab-gpt-neox.gguf)
#llama_test(test-tokenizer-1-bpe NAME test-tokenizer-1-llama-bpe ARGS ${PROJECT_SOURCE_DIR}/models/ggml-vocab-llama-bpe.gguf --ignore-merges)
#llama_test(test-tokenizer-1-bpe NAME test-tokenizer-1-mpt ARGS ${PROJECT_SOURCE_DIR}/models/ggml-vocab-mpt.gguf)
#llama_test(test-tokenizer-1-bpe NAME test-tokenizer-1-refact ARGS ${PROJECT_SOURCE_DIR}/models/ggml-vocab-refact.gguf)
#llama_test(test-tokenizer-1-bpe NAME test-tokenizer-1-starcoder ARGS ${PROJECT_SOURCE_DIR}/models/ggml-vocab-starcoder.gguf)
# build test-tokenizer-1-spm target once and add many tests
llama_build(test-tokenizer-1-spm.cpp)
llama_test(test-tokenizer-1-spm NAME test-tokenizer-1-llama-spm ARGS ${PROJECT_SOURCE_DIR}/models/ggml-vocab-llama-spm.gguf)
#llama_test(test-tokenizer-1-spm NAME test-tokenizer-1-baichuan ARGS ${PROJECT_SOURCE_DIR}/models/ggml-vocab-baichuan.gguf)
# llama_build_and_test(test-double-float.cpp) # SLOW
endif()
llama_build_and_test(test-chat-parser.cpp)
llama_build_and_test(test-chat-template.cpp)
llama_build_and_test(test-json-partial.cpp)
llama_build_and_test(test-log.cpp)
llama_build_and_test(test-regex-partial.cpp)
llama_build_and_test(test-thread-safety.cpp ARGS -hf ggml-org/models -hff tinyllamas/stories15M-q4_0.gguf -ngl 99 -p "The meaning of life is" -n 128 -c 256 -ub 32 -np 4 -t 2)
# this fails on windows (github hosted runner) due to curl DLL not found (exit code 0xc0000135)
if (NOT WIN32)
llama_build_and_test(test-arg-parser.cpp)
endif()
if (NOT LLAMA_SANITIZE_ADDRESS)
# TODO: repair known memory leaks
llama_build_and_test(test-opt.cpp)
endif()
llama_build_and_test(test-gguf.cpp)
llama_build_and_test(test-backend-ops.cpp)
llama_build_and_test(test-model-load-cancel.cpp LABEL "model")
llama_build_and_test(test-autorelease.cpp LABEL "model")
if (NOT GGML_BACKEND_DL)
# these tests use the backends directly and cannot be built with dynamic loading
llama_build_and_test(test-barrier.cpp)
llama_build_and_test(test-quantize-fns.cpp)
llama_build_and_test(test-quantize-perf.cpp)
llama_build_and_test(test-rope.cpp)
endif()
# libmtmd
set(LLAMA_TEST_NAME test-mtmd-c-api)
llama_build_and_test(test-mtmd-c-api.c)
target_link_libraries(${LLAMA_TEST_NAME} PRIVATE mtmd)
# dummy executable - not installed
get_filename_component(TEST_TARGET test-c.c NAME_WE)
add_executable(${TEST_TARGET} test-c.c)
target_link_libraries(${TEST_TARGET} PRIVATE llama)
Reference in New Issue View Git Blame Copy Permalink