refactor: rename llama-stable to llama-ggml (#1287)

* refactor: rename llama-stable to llama-ggml * Makefile: get sources in sources/ Signed-off-by: Ettore Di Giacinto <mudler@localai.io> * fixup path Signed-off-by: Ettore Di Giacinto <mudler@localai.io> * fixup sources Signed-off-by: Ettore Di Giacinto <mudler@localai.io> * fixups sd Signed-off-by: Ettore Di Giacinto <mudler@localai.io> * update SD * fixup * fixup: create piper libdir also when not built Signed-off-by: Ettore Di Giacinto <mudler@localai.io> * fix make target on linux test Signed-off-by: Ettore Di Giacinto <mudler@localai.io> --------- Signed-off-by: Ettore Di Giacinto <mudler@localai.io>
mudler · Nov 18, 2023 · 3c9544b · 3c9544b
1 parent 2f65671
commit 3c9544b
Show file tree

Hide file tree

Showing 8 changed files with 125 additions and 138 deletions.
diff --git a/.github/workflows/test.yml b/.github/workflows/test.yml
@@ -81,10 +81,9 @@ jobs:
           PATH=$PATH:/opt/conda/bin make -C backend/python/huggingface
 
           # Pre-build piper before we start tests in order to have shared libraries in place
-          make go-piper && \
-          GO_TAGS="tts" make -C go-piper piper.o && \
-          sudo cp -rfv go-piper/piper/build/pi/lib/. /usr/lib/ && \
-
+          make sources/go-piper && \
+          GO_TAGS="tts" make -C sources/go-piper piper.o && \
+          sudo cp -rfv sources/go-piper/piper/build/pi/lib/. /usr/lib/ && \
           # Pre-build stable diffusion before we install a newer version of abseil (not compatible with stablediffusion-ncn)
           GO_TAGS="stablediffusion tts" GRPC_BACKENDS=backend-assets/grpc/stablediffusion make build
 

diff --git a/.gitignore b/.gitignore
@@ -1,12 +1,5 @@
 # go-llama build artifacts
-go-llama
-go-llama-stable
-/gpt4all
-go-stable-diffusion
-go-piper
-/go-bert
-go-ggllm
-/piper
+/sources/
 __pycache__/
 *.a
 get-sources

diff --git a/Dockerfile b/Dockerfile
@@ -108,6 +108,11 @@ RUN if [ "${BUILD_GRPC}" = "true" ]; then \
 # Rebuild with defaults backends
 RUN make build
 
+RUN if [ ! -d "/build/sources/go-piper/piper/build/pi/lib/" ]; then \
+    mkdir -p /build/sources/go-piper/piper/build/pi/lib/ \
+    touch /build/sources/go-piper/piper/build/pi/lib/keep \
+    ; fi
+
 ###################################
 ###################################
 
@@ -145,7 +150,7 @@ RUN make prepare-sources
 COPY --from=builder /build/local-ai ./
 
 # Copy shared libraries for piper
-COPY --from=builder /build/go-piper/piper/build/pi/lib/* /usr/lib/
+COPY --from=builder /build/sources/go-piper/piper/build/pi/lib/* /usr/lib/
 
 # do not let stablediffusion rebuild (requires an older version of absl)
 COPY --from=builder /build/backend-assets/grpc/stablediffusion ./backend-assets/grpc/stablediffusion

diff --git a/Makefile b/Makefile
diff --git a/api/api_test.go b/api/api_test.go
@@ -301,7 +301,7 @@ var _ = Describe("API test", func() {
 				response := postModelApplyRequest("http://127.0.0.1:9090/models/apply", modelApplyRequest{
 					URL:       "github:go-skynet/model-gallery/openllama_3b.yaml",
 					Name:      "openllama_3b",
-					Overrides: map[string]interface{}{"backend": "llama-stable", "mmap": true, "f16": true, "context_size": 128},
+					Overrides: map[string]interface{}{"backend": "llama-ggml", "mmap": true, "f16": true, "context_size": 128},
 				})
 
 				Expect(response["uuid"]).ToNot(BeEmpty(), fmt.Sprint(response))

diff --git a/backend/go/llm/llama-stable/llama.go → backend/go/llm/llama-ggml/llama.go b/backend/go/llm/llama-stable/llama.go → backend/go/llm/llama-ggml/llama.go
diff --git a/backend/go/llm/llama-stable/main.go → backend/go/llm/llama-ggml/main.go b/backend/go/llm/llama-stable/main.go → backend/go/llm/llama-ggml/main.go
diff --git a/pkg/model/initializers.go b/pkg/model/initializers.go
@@ -15,8 +15,8 @@ import (
 )
 
 const (
-	LlamaBackend        = "llama"
-	LlamaStableBackend  = "llama-stable"
+	GoLlamaBackend      = "llama"
+	LlamaGGML           = "llama-ggml"
 	LLamaCPP            = "llama-cpp"
 	StarcoderBackend    = "starcoder"
 	GPTJBackend         = "gptj"
@@ -41,8 +41,8 @@ const (
 
 var AutoLoadBackends []string = []string{
 	LLamaCPP,
-	LlamaStableBackend,
-	LlamaBackend,
+	LlamaGGML,
+	GoLlamaBackend,
 	Gpt4All,
 	GPTNeoXBackend,
 	BertEmbeddingsBackend,