[pull] master from mudler:master

.github/workflows/generate_grpc_cache.yaml

@@ -17,7 +17,7 @@ jobs:
         include:
           - grpc-base-image: ubuntu:22.04
             runs-on: 'ubuntu-latest'
-            platforms: 'linux/amd64'
+            platforms: 'linux/amd64,linux/arm64'
     runs-on: ${{matrix.runs-on}}
     steps:
       - name: Release space from worker

.github/workflows/image_build.yml

@@ -136,6 +136,7 @@ jobs:
 
       - name: Docker meta
         id: meta
+        if: github.event_name != 'pull_request'
         uses: docker/metadata-action@v5
         with:
           images: |
@@ -148,7 +149,20 @@ jobs:
           flavor: |
             latest=${{ inputs.tag-latest }}
             suffix=${{ inputs.tag-suffix }}
-
+      - name: Docker meta for PR
+        id: meta_pull_request
+        if: github.event_name == 'pull_request'
+        uses: docker/metadata-action@v5
+        with:
+          images: |
+            ttl.sh/localai-ci-pr-${{ github.event.number }}
+          tags: |
+            type=ref,event=branch
+            type=semver,pattern={{raw}}
+            type=sha
+          flavor: |
+            latest=${{ inputs.tag-latest }}
+            suffix=${{ inputs.tag-suffix }}
       - name: Docker meta AIO (quay.io)
         if: inputs.aio != ''
         id: meta_aio
@@ -202,6 +216,7 @@ jobs:
 
       - name: Build and push
         uses: docker/build-push-action@v5
+        if: github.event_name != 'pull_request'
         with:
           builder: ${{ steps.buildx.outputs.name }}
           # The build-args MUST be an EXACT match between the image cache and other workflow steps that want to use that cache.
@@ -226,7 +241,39 @@ jobs:
           push: ${{ github.event_name != 'pull_request' }}
           tags: ${{ steps.meta.outputs.tags }}
           labels: ${{ steps.meta.outputs.labels }}
-
+### Start testing image
+      - name: Build and push
+        uses: docker/build-push-action@v5
+        if: github.event_name == 'pull_request'
+        with:
+          builder: ${{ steps.buildx.outputs.name }}
+          # The build-args MUST be an EXACT match between the image cache and other workflow steps that want to use that cache.
+          # This means that even the MAKEFLAGS have to be an EXACT match.
+          # If the build-args are not an EXACT match, it will result in a cache miss, which will require GRPC to be built from scratch.
+          # This is why some build args like GRPC_VERSION and MAKEFLAGS are hardcoded
+          build-args: |
+            BUILD_TYPE=${{ inputs.build-type }}
+            CUDA_MAJOR_VERSION=${{ inputs.cuda-major-version }}
+            CUDA_MINOR_VERSION=${{ inputs.cuda-minor-version }}
+            FFMPEG=${{ inputs.ffmpeg }}
+            IMAGE_TYPE=${{ inputs.image-type }}
+            BASE_IMAGE=${{ inputs.base-image }}
+            GRPC_BASE_IMAGE=${{ inputs.grpc-base-image || inputs.base-image }}
+            GRPC_MAKEFLAGS=--jobs=4 --output-sync=target
+            GRPC_VERSION=v1.64.0
+            MAKEFLAGS=${{ inputs.makeflags }}
+          context: .
+          file: ./Dockerfile
+          cache-from: type=gha
+          platforms: ${{ inputs.platforms }}
+          push: true
+          tags: ${{ steps.meta_pull_request.outputs.tags }}
+          labels: ${{ steps.meta_pull_request.outputs.labels }}
+      - name: Testing image
+        if: github.event_name == 'pull_request'
+        run: |
+          echo "Image is available at ttl.sh/localai-ci-pr-${{ github.event.number }}:${{ steps.meta_pull_request.outputs.version }}" >> $GITHUB_STEP_SUMMARY
+## End testing image
       - name: Build and push AIO image
         if: inputs.aio != ''
         uses: docker/build-push-action@v5

Dockerfile

@@ -24,12 +24,9 @@ RUN apt-get update && \
         cmake \
         curl \
         git \
-        python3-pip \
-        python-is-python3 \
         unzip && \
     apt-get clean && \
-    rm -rf /var/lib/apt/lists/* && \
-    pip install --upgrade pip
+    rm -rf /var/lib/apt/lists/*
 
 # Install Go
 RUN curl -L -s https://go.dev/dl/go${GO_VERSION}.linux-${TARGETARCH}.tar.gz | tar -C /usr/local -xz
@@ -39,9 +36,6 @@ ENV PATH $PATH:/root/go/bin:/usr/local/go/bin
 RUN go install google.golang.org/protobuf/cmd/protoc-gen-go@latest && \
     go install google.golang.org/grpc/cmd/protoc-gen-go-grpc@latest
 
-# Install grpcio-tools (the version in 22.04 is too old)
-RUN pip install --user grpcio-tools
-
 COPY --chmod=644 custom-ca-certs/* /usr/local/share/ca-certificates/
 RUN update-ca-certificates
 
@@ -85,10 +79,16 @@ RUN apt-get update && \
     apt-get install -y --no-install-recommends \
         espeak-ng \
         espeak \
+        python3-pip \
+        python-is-python3 \
         python3-dev \
         python3-venv && \
     apt-get clean && \
-    rm -rf /var/lib/apt/lists/*
+    rm -rf /var/lib/apt/lists/* && \
+    pip install --upgrade pip
+
+# Install grpcio-tools (the version in 22.04 is too old)
+RUN pip install --user grpcio-tools
 
 ###################################
 ###################################

Makefile

@@ -5,7 +5,7 @@ BINARY_NAME=local-ai
 
 # llama.cpp versions
 GOLLAMA_STABLE_VERSION?=2b57a8ae43e4699d3dc5d1496a1ccd42922993be
-CPPLLAMA_VERSION?=dff451cfa1f297348751ce6b538670e1ae9a7d5b
+CPPLLAMA_VERSION?=10b1e4587670feba2c7730a645accf8234873113
 
 # gpt4all version
 GPT4ALL_REPO?=https://github.com/nomic-ai/gpt4all
@@ -16,7 +16,7 @@ RWKV_REPO?=https://github.com/donomii/go-rwkv.cpp
 RWKV_VERSION?=661e7ae26d442f5cfebd2a0881b44e8c55949ec6
 
 # whisper.cpp version
-WHISPER_CPP_VERSION?=a7dc2aab16822b80a6491b0bd4bbf4900404a8a0
+WHISPER_CPP_VERSION?=c7b6988678779901d02ceba1a8212d2c9908956e
 
 # bert.cpp version
 BERT_VERSION?=710044b124545415f555e4260d16b146c725a6e4
@@ -112,7 +112,7 @@ ifeq ($(BUILD_TYPE),hipblas)
 	# llama-ggml has no hipblas support, so override it here.
 	export STABLE_BUILD_TYPE=
 	export WHISPER_HIPBLAS=1
-	GPU_TARGETS ?= gfx900,gfx90a,gfx1030,gfx1031,gfx1100
+	GPU_TARGETS ?= gfx900,gfx906,gfx908,gfx940,gfx941,gfx942,gfx90a,gfx1030,gfx1031,gfx1100,gfx1101
 	AMDGPU_TARGETS ?= "$(GPU_TARGETS)"
 	CMAKE_ARGS+=-DLLAMA_HIPBLAS=ON -DAMDGPU_TARGETS="$(AMDGPU_TARGETS)" -DGPU_TARGETS="$(GPU_TARGETS)"
 	CGO_LDFLAGS += -O3 --rtlib=compiler-rt -unwindlib=libgcc -lhipblas -lrocblas --hip-link -L${ROCM_HOME}/lib/llvm/lib

gallery/index.yaml

@@ -779,6 +779,21 @@
     - filename: Tess-2.0-Llama-3-8B-Q4_K_M.gguf
       sha256: 3b5fbd6c59d7d38205ab81970c0227c74693eb480acf20d8c2f211f62e3ca5f6
       uri: huggingface://bartowski/Tess-2.0-Llama-3-8B-GGUF/Tess-2.0-Llama-3-8B-Q4_K_M.gguf
+- !!merge <<: *llama3
+  name: "llama3-iterative-dpo-final"
+  urls:
+    - https://huggingface.co/bartowski/LLaMA3-iterative-DPO-final-GGUF
+    - https://huggingface.co/RLHFlow/LLaMA3-iterative-DPO-final
+  description: |
+     From model card:
+      We release an unofficial checkpoint of a state-of-the-art instruct model of its class, LLaMA3-iterative-DPO-final. On all three widely-used instruct model benchmarks: Alpaca-Eval-V2, MT-Bench, Chat-Arena-Hard, our model outperforms all models of similar size (e.g., LLaMA-3-8B-it), most large open-sourced models (e.g., Mixtral-8x7B-it), and strong proprietary models (e.g., GPT-3.5-turbo-0613). The model is trained with open-sourced datasets without any additional human-/GPT4-labeling.
+  overrides:
+    parameters:
+      model: LLaMA3-iterative-DPO-final-Q4_K_M.gguf
+  files:
+    - filename: LLaMA3-iterative-DPO-final-Q4_K_M.gguf
+      sha256: 480703ff85af337e1db2a9d9a678a3ac8ca0802e366b14d9c59b81d3fc689da8
+      uri: huggingface://bartowski/LLaMA3-iterative-DPO-final-GGUF/LLaMA3-iterative-DPO-final-Q4_K_M.gguf
 - &dolphin
   name: "dolphin-2.9-llama3-8b"
   url: "github:mudler/LocalAI/gallery/hermes-2-pro-mistral.yaml@master"
@@ -1248,7 +1263,7 @@
   files:
     - filename: minicpm-llama3-Q4_K_M.gguf
       sha256: 010ec3ba94cb5ad2d9c8f95f46f01c6d80f83deab9df0a0831334ea45afff3e2
-      uri: huggingface://openbmb/MiniCPM-Llama3-V-2_5-gguf/minicpm-llama3-Q4_K_M.gguf
+      uri: huggingface://openbmb/MiniCPM-Llama3-V-2_5-gguf/ggml-model-Q4_K_M.gguf
     - filename: minicpm-llama3-mmproj-f16.gguf
       sha256: 391d11736c3cd24a90417c47b0c88975e86918fcddb1b00494c4d715b08af13e
       uri: huggingface://openbmb/MiniCPM-Llama3-V-2_5-gguf/mmproj-model-f16.gguf