dearwolf/LocalAI
1
0
Fork 0
mirror of https://github.com/mudler/LocalAI.git synced 2025-05-31 07:54:59 +00:00
Code Issues Projects Releases Packages Wiki Activity Actions 36

chore(llama-ggml): drop deprecated backend (#4775)

Browse source 
The GGML format is now dead, since in the next version of LocalAI we
already bring many breaking compatibility changes, taking the occasion
also to drop ggml support (pre-gguf).

Signed-off-by: Ettore Di Giacinto <mudler@localai.io>
This commit is contained in:
Ettore Di Giacinto 2025-02-06 18:36:23 +01:00 • committed by GitHub
parent 8d45670e41
commit 7f90ff7aec
No known key found for this signature in database
GPG key ID: B5690EEEBB952194
6 changed files with 7 additions and 348 deletions
Download patch file Download diff file Expand all files Collapse all files

38
Makefile
View file

@ -6,8 +6,6 @@ BINARY_NAME=local-ai
DETECT_LIBS?=true
# llama.cpp versions
GOLLAMA_REPO?=https://github.com/go-skynet/go-llama.cpp
GOLLAMA_VERSION?=2b57a8ae43e4699d3dc5d1496a1ccd42922993be
CPPLLAMA_VERSION?=d774ab3acc4fee41fbed6dbfc192b57d5f79f34b
# whisper.cpp version
@ -151,7 +149,6 @@ ifeq ($(BUILD_TYPE),hipblas)
LD_LIBRARY_PATH ?= /opt/rocm/lib:/opt/rocm/llvm/lib
export CXX=$(ROCM_HOME)/llvm/bin/clang++
export CC=$(ROCM_HOME)/llvm/bin/clang
# llama-ggml has no hipblas support, so override it here.
export STABLE_BUILD_TYPE=
export GGML_HIP=1
GPU_TARGETS ?= gfx900,gfx906,gfx908,gfx940,gfx941,gfx942,gfx90a,gfx1030,gfx1031,gfx1100,gfx1101
@ -188,7 +185,6 @@ ALL_GRPC_BACKENDS+=backend-assets/grpc/llama-cpp-avx
ALL_GRPC_BACKENDS+=backend-assets/grpc/llama-cpp-avx2
ALL_GRPC_BACKENDS+=backend-assets/grpc/llama-cpp-avx512
ALL_GRPC_BACKENDS+=backend-assets/grpc/llama-cpp-fallback
ALL_GRPC_BACKENDS+=backend-assets/grpc/llama-ggml
ALL_GRPC_BACKENDS+=backend-assets/grpc/llama-cpp-grpc
ALL_GRPC_BACKENDS+=backend-assets/util/llama-cpp-rpc-server
ALL_GRPC_BACKENDS+=backend-assets/grpc/whisper
@ -222,19 +218,6 @@ endif
all: help
## go-llama.cpp
sources/go-llama.cpp:
mkdir -p sources/go-llama.cpp
cd sources/go-llama.cpp && \
git init && \
git remote add origin $(GOLLAMA_REPO) && \
git fetch origin && \
git checkout $(GOLLAMA_VERSION) && \
git submodule update --init --recursive --depth 1 --single-branch
sources/go-llama.cpp/libbinding.a: sources/go-llama.cpp
$(MAKE) -C sources/go-llama.cpp BUILD_TYPE=$(STABLE_BUILD_TYPE) libbinding.a
## bark.cpp
sources/bark.cpp:
git clone --recursive $(BARKCPP_REPO) sources/bark.cpp && \
@ -310,19 +293,17 @@ sources/whisper.cpp:
sources/whisper.cpp/libwhisper.a: sources/whisper.cpp
cd sources/whisper.cpp && $(MAKE) libwhisper.a libggml.a
get-sources: sources/go-llama.cpp sources/go-piper sources/stablediffusion-ggml.cpp sources/bark.cpp sources/whisper.cpp backend/cpp/llama/llama.cpp
get-sources: sources/go-piper sources/stablediffusion-ggml.cpp sources/bark.cpp sources/whisper.cpp backend/cpp/llama/llama.cpp
replace:
$(GOCMD) mod edit -replace github.com/ggerganov/whisper.cpp=$(CURDIR)/sources/whisper.cpp
$(GOCMD) mod edit -replace github.com/ggerganov/whisper.cpp/bindings/go=$(CURDIR)/sources/whisper.cpp/bindings/go
$(GOCMD) mod edit -replace github.com/mudler/go-piper=$(CURDIR)/sources/go-piper
$(GOCMD) mod edit -replace github.com/go-skynet/go-llama.cpp=$(CURDIR)/sources/go-llama.cpp
dropreplace:
$(GOCMD) mod edit -dropreplace github.com/ggerganov/whisper.cpp
$(GOCMD) mod edit -dropreplace github.com/ggerganov/whisper.cpp/bindings/go
$(GOCMD) mod edit -dropreplace github.com/mudler/go-piper
$(GOCMD) mod edit -dropreplace github.com/go-skynet/go-llama.cpp
prepare-sources: get-sources replace
$(GOCMD) mod download
@ -330,7 +311,6 @@ prepare-sources: get-sources replace
## GENERIC
rebuild: ## Rebuilds the project
$(GOCMD) clean -cache
$(MAKE) -C sources/go-llama.cpp clean
$(MAKE) -C sources/whisper.cpp clean
$(MAKE) -C sources/go-piper clean
$(MAKE) build
@ -434,7 +414,7 @@ run: prepare ## run local-ai
test-models/testmodel.ggml:
mkdir test-models
mkdir test-dir
wget -q https://huggingface.co/TheBloke/orca_mini_3B-GGML/resolve/main/orca-mini-3b.ggmlv3.q4_0.bin -O test-models/testmodel.ggml
wget -q https://huggingface.co/RichardErkhov/Qwen_-_Qwen2-1.5B-Instruct-gguf/resolve/main/Qwen2-1.5B-Instruct.Q2_K.gguf -O test-models/testmodel.ggml
wget -q https://huggingface.co/ggerganov/whisper.cpp/resolve/main/ggml-base.en.bin -O test-models/whisper-en
wget -q https://huggingface.co/mudler/all-MiniLM-L6-v2/resolve/main/ggml-model-q4_0.bin -O test-models/bert
wget -q https://cdn.openai.com/whisper/draft-20220913a/micro-machines.wav -O test-dir/audio.wav
@ -449,8 +429,7 @@ test: prepare test-models/testmodel.ggml grpcs
export GO_TAGS="tts debug"
$(MAKE) prepare-test
HUGGINGFACE_GRPC=$(abspath ./)/backend/python/transformers/run.sh TEST_DIR=$(abspath ./)/test-dir/ FIXTURES=$(abspath ./)/tests/fixtures CONFIG_FILE=$(abspath ./)/test-models/config.yaml MODELS_PATH=$(abspath ./)/test-models \
$(GOCMD) run github.com/onsi/ginkgo/v2/ginkgo --label-filter="!llama && !llama-gguf" --flake-attempts $(TEST_FLAKES) --fail-fast -v -r $(TEST_PATHS)
$(MAKE) test-llama
$(GOCMD) run github.com/onsi/ginkgo/v2/ginkgo --label-filter="!llama-gguf" --flake-attempts $(TEST_FLAKES) --fail-fast -v -r $(TEST_PATHS)
$(MAKE) test-llama-gguf
$(MAKE) test-tts
$(MAKE) test-stablediffusion
@ -479,10 +458,6 @@ teardown-e2e:
rm -rf $(TEST_DIR) || true
docker stop $$(docker ps -q --filter ancestor=localai-tests)
test-llama: prepare-test
TEST_DIR=$(abspath ./)/test-dir/ FIXTURES=$(abspath ./)/tests/fixtures CONFIG_FILE=$(abspath ./)/test-models/config.yaml MODELS_PATH=$(abspath ./)/test-models \
$(GOCMD) run github.com/onsi/ginkgo/v2/ginkgo --label-filter="llama" --flake-attempts $(TEST_FLAKES) -v -r $(TEST_PATHS)
test-llama-gguf: prepare-test
TEST_DIR=$(abspath ./)/test-dir/ FIXTURES=$(abspath ./)/tests/fixtures CONFIG_FILE=$(abspath ./)/test-models/config.yaml MODELS_PATH=$(abspath ./)/test-models \
$(GOCMD) run github.com/onsi/ginkgo/v2/ginkgo --label-filter="llama-gguf" --flake-attempts $(TEST_FLAKES) -v -r $(TEST_PATHS)
@ -760,13 +735,6 @@ backend-assets/util/llama-cpp-rpc-server: backend-assets/grpc/llama-cpp-grpc
mkdir -p backend-assets/util/
cp -rf backend/cpp/llama-grpc/llama.cpp/build/bin/rpc-server backend-assets/util/llama-cpp-rpc-server
backend-assets/grpc/llama-ggml: sources/go-llama.cpp sources/go-llama.cpp/libbinding.a backend-assets/grpc
CGO_LDFLAGS="$(CGO_LDFLAGS)" C_INCLUDE_PATH=$(CURDIR)/sources/go-llama.cpp LIBRARY_PATH=$(CURDIR)/sources/go-llama.cpp \
$(GOCMD) build -ldflags "$(LD_FLAGS)" -tags "$(GO_TAGS)" -o backend-assets/grpc/llama-ggml ./backend/go/llm/llama-ggml/
ifneq ($(UPX),)
$(UPX) backend-assets/grpc/llama-ggml
endif
backend-assets/grpc/bark-cpp: backend/go/bark/libbark.a backend-assets/grpc
CGO_LDFLAGS="$(CGO_LDFLAGS)" C_INCLUDE_PATH=$(CURDIR)/backend/go/bark/ LIBRARY_PATH=$(CURDIR)/backend/go/bark/ \
$(GOCMD) build -ldflags "$(LD_FLAGS)" -tags "$(GO_TAGS)" -o backend-assets/grpc/bark-cpp ./backend/go/bark/