mirror of
https://github.com/mudler/LocalAI.git
synced 2026-07-11 00:38:53 -04:00
94bdc825dc
C++/ggml transcription + speaker diarization + timestamps backend. Purego dlopens libmoss-transcribe.so (ggml statically linked) from moss-transcribe.cpp and serves offline AudioTranscription, parsing the [start][Sxx]text[end] output into segments with nanosecond timestamps. Adds the importer (surfaces in GET /backends/known), backend-matrix (Linux + Darwin/metal), backend/index.yaml, and a gallery entry (default q5_k GGUF from mudler/moss-transcribe.cpp-gguf). Local L0 smoke (go build + go test ./... = 16 pass, golangci-lint 0 issues) passed against the real libmoss-transcribe.so. The pre-commit coverage gate (full pkg/core + tests/e2e) could not run in the authoring sandbox (no live models, port 9090 held); CI must enforce it before merge. Assisted-by: Claude:claude-opus-4-8 golangci-lint Signed-off-by: Ettore Di Giacinto <mudler@localai.io> Co-authored-by: Ettore Di Giacinto <mudler@localai.io>
99 lines
4.0 KiB
Makefile
99 lines
4.0 KiB
Makefile
# moss-transcribe-cpp backend Makefile.
|
|
#
|
|
# Upstream pin lives below as MOSS_VERSION?=1fbe88c8ebf0134ed2c49b49345102752500fbe5
|
|
# (.github/bump_deps.sh) can find and update it - matches the
|
|
# whisper.cpp / parakeet-cpp / ds4 convention.
|
|
#
|
|
# Local dev shortcut: if you already have an out-of-tree moss-transcribe.cpp
|
|
# build, you can symlink the .so + header into this directory and skip the
|
|
# clone/cmake steps entirely, e.g.:
|
|
#
|
|
# ln -sf /path/to/moss-transcribe.cpp/build-shared/libmoss-transcribe.so .
|
|
# ln -sf /path/to/moss-transcribe.cpp/include/moss_transcribe_capi.h .
|
|
# go build -o moss-transcribe-cpp-grpc .
|
|
#
|
|
# That's what the L0 smoke test uses. The default target below does the proper
|
|
# clone-at-pin + cmake build so CI doesn't need a side-checkout.
|
|
|
|
MOSS_VERSION?=1fbe88c8ebf0134ed2c49b49345102752500fbe5
|
|
MOSS_REPO?=https://github.com/mudler/moss-transcribe.cpp
|
|
|
|
GOCMD?=go
|
|
GO_TAGS?=
|
|
JOBS?=$(shell nproc 2>/dev/null || sysctl -n hw.ncpu 2>/dev/null || echo 4)
|
|
|
|
BUILD_TYPE?=
|
|
NATIVE?=false
|
|
|
|
# Build ggml statically into libmoss-transcribe.so (PIC) so the shared lib is
|
|
# self-contained: dlopen needs no libggml*.so alongside it, only system libs
|
|
# (libstdc++/libgomp/libc) that the runtime image already provides. MT_SHARED
|
|
# flips ggml to a static PIC build (see the moss-transcribe.cpp CMakeLists).
|
|
CMAKE_ARGS?=-DCMAKE_BUILD_TYPE=Release -DMT_SHARED=ON -DMT_BUILD_CLI=OFF -DBUILD_SHARED_LIBS=OFF -DCMAKE_POSITION_INDEPENDENT_CODE=ON
|
|
|
|
ifeq ($(NATIVE),false)
|
|
CMAKE_ARGS+=-DGGML_NATIVE=OFF
|
|
endif
|
|
|
|
# moss-transcribe.cpp gates its GGML backends behind MT_GGML_* options and does
|
|
# set(GGML_<be> ${MT_GGML_<be>} CACHE BOOL "" FORCE), so a bare -DGGML_CUDA=ON is
|
|
# overwritten back to OFF and the build silently falls back to CPU. Forward the
|
|
# MT_GGML_* options instead (openblas is not gated, so -DGGML_BLAS passes through).
|
|
ifeq ($(BUILD_TYPE),cublas)
|
|
CMAKE_ARGS+=-DMT_GGML_CUDA=ON -DGGML_CUDA_GRAPHS=ON
|
|
else ifeq ($(BUILD_TYPE),openblas)
|
|
CMAKE_ARGS+=-DGGML_BLAS=ON -DGGML_BLAS_VENDOR=OpenBLAS
|
|
else ifeq ($(BUILD_TYPE),hipblas)
|
|
CMAKE_ARGS+=-DMT_GGML_HIP=ON
|
|
else ifeq ($(BUILD_TYPE),vulkan)
|
|
CMAKE_ARGS+=-DMT_GGML_VULKAN=ON
|
|
else ifeq ($(BUILD_TYPE),metal)
|
|
CMAKE_ARGS+=-DMT_GGML_METAL=ON
|
|
endif
|
|
|
|
.PHONY: moss-transcribe-cpp-grpc package build clean purge test all
|
|
|
|
all: moss-transcribe-cpp-grpc
|
|
|
|
# Clone the upstream moss-transcribe.cpp source at the pinned commit. Directory
|
|
# acts as the target so make only re-clones when missing. After a MOSS_VERSION
|
|
# bump, run 'make purge && make' to refetch.
|
|
sources/moss-transcribe.cpp:
|
|
mkdir -p sources/moss-transcribe.cpp
|
|
cd sources/moss-transcribe.cpp && \
|
|
git init -q && \
|
|
git remote add origin $(MOSS_REPO) && \
|
|
git fetch --depth 1 origin $(MOSS_VERSION) && \
|
|
git checkout FETCH_HEAD && \
|
|
git submodule update --init --recursive --depth 1 --single-branch
|
|
|
|
# Build the shared lib + header out-of-tree, then stage them next to the Go
|
|
# sources so purego.Dlopen("libmoss-transcribe.so") and the cgo-less build both
|
|
# pick them up.
|
|
libmoss-transcribe.so: sources/moss-transcribe.cpp
|
|
cmake -B sources/moss-transcribe.cpp/build-shared -S sources/moss-transcribe.cpp $(CMAKE_ARGS)
|
|
cmake --build sources/moss-transcribe.cpp/build-shared --config Release -j$(JOBS)
|
|
cp -fv sources/moss-transcribe.cpp/build-shared/libmoss-transcribe.so* ./ 2>/dev/null || true
|
|
cp -fv sources/moss-transcribe.cpp/build-shared/libmoss-transcribe.dylib ./ 2>/dev/null || true
|
|
cp -fv sources/moss-transcribe.cpp/include/moss_transcribe_capi.h ./
|
|
|
|
moss-transcribe-cpp-grpc: libmoss-transcribe.so main.go gomosstranscribecpp.go segments.go
|
|
CGO_ENABLED=0 $(GOCMD) build -tags "$(GO_TAGS)" -o moss-transcribe-cpp-grpc .
|
|
|
|
package: moss-transcribe-cpp-grpc
|
|
bash package.sh
|
|
|
|
build: package
|
|
|
|
# Test target. The model-backed smoke test is gated on
|
|
# MOSS_BACKEND_TEST_MODEL + MOSS_BACKEND_TEST_WAV; without them that spec
|
|
# auto-skips, leaving the pure-Go parser/unit tests.
|
|
test:
|
|
LD_LIBRARY_PATH=$(CURDIR):$$LD_LIBRARY_PATH $(GOCMD) test ./... -count=1
|
|
|
|
clean: purge
|
|
rm -rf libmoss-transcribe.so* moss_transcribe_capi.h package moss-transcribe-cpp-grpc
|
|
|
|
purge:
|
|
rm -rf sources/moss-transcribe.cpp
|