huggingface · Narsil · Jan 29, 2025 · Jan 29, 2025 · Jan 29, 2025 · Jan 29, 2025
diff --git a/.github/workflows/build.yaml b/.github/workflows/build.yaml
@@ -31,8 +31,10 @@ jobs:
         uses: actions/checkout@v3
 
       - id: set-matrix
+        env:
+          GITHUB_REF: ${{ github.ref }}
         run: |
-          branchName=$(echo '${{ github.ref }}' | sed 's,refs/heads/,,g')
+          branchName=$(echo $GITHUB_REF | sed 's,refs/heads/,,g')
           matrix=$(jq --arg branchName "$branchName" 'map(. | select((.runOn==$branchName) or (.runOn=="always")) )' .github/workflows/matrix.json)
           echo "{\"include\":$(echo $matrix)}"
           echo ::set-output name=matrix::{\"include\":$(echo $matrix)}\"
@@ -89,6 +91,15 @@ jobs:
           username: ${{ github.actor }}
           password: ${{ secrets.GITHUB_TOKEN }}
 
+      - name: configure aws credentials
+        id: aws-creds
+        uses: aws-actions/configure-aws-credentials@4fc4975a852c8cd99761e2de1f4ba73402e44dd9
+        with:
+          role-to-assume: ${{ secrets.AWS_ROLE_GITHUB_BUILDX_CACHE }}
+          role-duration-seconds: 7200
+          aws-region: us-east-1
+          output-credentials: true
+
       - name: Extract metadata (tags, labels) for Docker
         id: meta
         uses: docker/metadata-action@v5
@@ -107,23 +118,26 @@ jobs:
       - name: Build and push Docker image
         id: build-and-push
         uses: docker/build-push-action@v6
+        env:
+          DOCKER_BUILD_SUMMARY: false
         with:
           context: .
           file: ${{ matrix.dockerfile }}
           push: ${{ github.event_name != 'pull_request' }}
           platforms: 'linux/amd64'
           build-args: |
             SCCACHE_GHA_ENABLED=${{ matrix.sccache }}
-            ACTIONS_CACHE_URL=${{ env.ACTIONS_CACHE_URL }}
-            ACTIONS_RUNTIME_TOKEN=${{ env.ACTIONS_RUNTIME_TOKEN }}
             CUDA_COMPUTE_CAP=${{ matrix.cudaComputeCap }}
             GIT_SHA=${{ env.GITHUB_SHA }}
             DOCKER_LABEL=sha-${{ env.GITHUB_SHA_SHORT }}
             ${{matrix.extraBuildArgs}}
+          secrets: |
+            actions_cache_url=${{ env.ACTIONS_CACHE_URL }}
+            actions_runtime_token=${{ env.ACTIONS_RUNTIME_TOKEN }}
           tags: ${{ steps.meta.outputs.tags }}
           labels: ${{ steps.meta.outputs.labels }}
-          cache-from: type=s3,region=us-east-1,bucket=ci-docker-buildx-cache,name=text-embeddings-inference-cache-${{matrix.name}},access_key_id=${{ secrets.S3_CI_DOCKER_BUILDX_CACHE_ACCESS_KEY_ID }},secret_access_key=${{ secrets.S3_CI_DOCKER_BUILDX_CACHE_SECRET_ACCESS_KEY }},mode=max
-          cache-to: type=s3,region=us-east-1,bucket=ci-docker-buildx-cache,name=text-embeddings-inference-cache-${{matrix.name}},access_key_id=${{ secrets.S3_CI_DOCKER_BUILDX_CACHE_ACCESS_KEY_ID }},secret_access_key=${{ secrets.S3_CI_DOCKER_BUILDX_CACHE_SECRET_ACCESS_KEY }},mode=max
+          cache-from: type=s3,region=us-east-1,bucket=${{ vars.AWS_S3BUCKET_GITHUB_BUILDX_CACHE }},name=text-embeddings-inference-cache-${{matrix.name}},access_key_id=${{ steps.aws-creds.outputs.aws-access-key-id }},secret_access_key=${{ steps.aws-creds.outputs.aws-secret-access-key }},session_token=${{ steps.aws-creds.outputs.aws-session-token }},mode=max
+          cache-to: type=s3,region=us-east-1,bucket=${{ vars.AWS_S3BUCKET_GITHUB_BUILDX_CACHE }},name=text-embeddings-inference-cache-${{matrix.name}},access_key_id=${{ steps.aws-creds.outputs.aws-access-key-id }},secret_access_key=${{ steps.aws-creds.outputs.aws-secret-access-key }},session_token=${{ steps.aws-creds.outputs.aws-session-token }},mode=max
 
       - name: Extract metadata (tags, labels) for Docker
         id: meta-grpc
@@ -145,6 +159,8 @@ jobs:
         id: build-and-push-grpc
         if: ${{ matrix.grpc }}
         uses: docker/build-push-action@v6
+        env:
+          DOCKER_BUILD_SUMMARY: false
         with:
           context: .
           target: grpc
@@ -153,12 +169,13 @@ jobs:
           platforms: 'linux/amd64'
           build-args: |
             SCCACHE_GHA_ENABLED=${{ matrix.sccache }}
-            ACTIONS_CACHE_URL=${{ env.ACTIONS_CACHE_URL }}
-            ACTIONS_RUNTIME_TOKEN=${{ env.ACTIONS_RUNTIME_TOKEN }}
             CUDA_COMPUTE_CAP=${{ matrix.cudaComputeCap }}
             GIT_SHA=${{ env.GITHUB_SHA }}
             DOCKER_LABEL=sha-${{ env.GITHUB_SHA_SHORT }}
             ${{matrix.extraBuildArgs}}
+          secrets: |
+            actions_cache_url=${{ env.ACTIONS_CACHE_URL }}
+            actions_runtime_token=${{ env.ACTIONS_RUNTIME_TOKEN }}
           tags: ${{ steps.meta-grpc.outputs.tags }}
           labels: ${{ steps.meta-grpc.outputs.labels }}
-          cache-from: type=s3,region=us-east-1,bucket=ci-docker-buildx-cache,name=text-embeddings-inference-cache-${{matrix.name}},access_key_id=${{ secrets.S3_CI_DOCKER_BUILDX_CACHE_ACCESS_KEY_ID }},secret_access_key=${{ secrets.S3_CI_DOCKER_BUILDX_CACHE_SECRET_ACCESS_KEY }},mode=max
+          cache-from: type=s3,region=us-east-1,bucket=${{ vars.AWS_S3BUCKET_GITHUB_BUILDX_CACHE }},name=text-embeddings-inference-cache-${{matrix.name}},access_key_id=${{ steps.aws-creds.outputs.aws-access-key-id }},secret_access_key=${{ steps.aws-creds.outputs.aws-secret-access-key }},session_token=${{ steps.aws-creds.outputs.aws-session-token }},mode=max
diff --git a/Dockerfile b/Dockerfile
@@ -24,8 +24,6 @@ ARG GIT_SHA
 ARG DOCKER_LABEL
 
 # sccache specific variables
-ARG ACTIONS_CACHE_URL
-ARG ACTIONS_RUNTIME_TOKEN
 ARG SCCACHE_GHA_ENABLED
 
 RUN wget -O- https://apt.repos.intel.com/intel-gpg-keys/GPG-PUB-KEY-INTEL-SW-PRODUCTS.PUB \
@@ -43,7 +41,9 @@ RUN echo "int mkl_serv_intel_cpu_true() {return 1;}" > fakeintel.c && \
 
 COPY --from=planner /usr/src/recipe.json recipe.json
 
-RUN cargo chef cook --release --features ort --features candle --features mkl-dynamic --no-default-features --recipe-path recipe.json && sccache -s
+RUN --mount=type=secret,id=actions_cache_url,env=ACTIONS_CACHE_URL \
+    --mount=type=secret,id=actions_runtime_token,env=ACTIONS_RUNTIME_TOKEN \
+     cargo chef cook --release --features ort --features candle --features mkl-dynamic --no-default-features --recipe-path recipe.json && sccache -s
 
 COPY backends backends
 COPY core core
@@ -53,7 +53,9 @@ COPY Cargo.lock ./
 
 FROM builder AS http-builder
 
-RUN cargo build --release --bin text-embeddings-router -F ort -F candle -F mkl-dynamic -F http --no-default-features && sccache -s
+RUN --mount=type=secret,id=actions_cache_url,env=ACTIONS_CACHE_URL \
+    --mount=type=secret,id=actions_runtime_token,env=ACTIONS_RUNTIME_TOKEN \
+    cargo build --release --bin text-embeddings-router -F ort -F candle -F mkl-dynamic -F http --no-default-features && sccache -s
 
 FROM builder AS grpc-builder
 
@@ -65,7 +67,9 @@ RUN PROTOC_ZIP=protoc-21.12-linux-x86_64.zip && \
 
 COPY proto proto
 
-RUN cargo build --release --bin text-embeddings-router -F grpc -F ort -F candle -F mkl-dynamic --no-default-features && sccache -s
+RUN --mount=type=secret,id=actions_cache_url,env=ACTIONS_CACHE_URL \
+    --mount=type=secret,id=actions_runtime_token,env=ACTIONS_RUNTIME_TOKEN \
+    cargo build --release --bin text-embeddings-router -F grpc -F ort -F candle -F mkl-dynamic --no-default-features && sccache -s
 
 FROM debian:bookworm-slim AS base
 

diff --git a/Dockerfile-cuda b/Dockerfile-cuda
@@ -41,13 +41,13 @@ ARG CARGO_BUILD_JOBS
 ARG CARGO_BUILD_INCREMENTAL
 
 # sccache specific variables
-ARG ACTIONS_CACHE_URL
-ARG ACTIONS_RUNTIME_TOKEN
 ARG SCCACHE_GHA_ENABLED
 
 WORKDIR /usr/src
 
-RUN if [ ${CUDA_COMPUTE_CAP} -ge 75 -a ${CUDA_COMPUTE_CAP} -lt 80 ]; \
+RUN --mount=type=secret,id=actions_cache_url,env=ACTIONS_CACHE_URL \
+    --mount=type=secret,id=actions_runtime_token,env=ACTIONS_RUNTIME_TOKEN \
+    if [ ${CUDA_COMPUTE_CAP} -ge 75 -a ${CUDA_COMPUTE_CAP} -lt 80 ]; \
     then  \
         nvprune --generate-code code=sm_${CUDA_COMPUTE_CAP} /usr/local/cuda/lib64/libcublas_static.a -o /usr/local/cuda/lib64/libcublas_static.a; \
     elif [ ${CUDA_COMPUTE_CAP} -ge 80 -a ${CUDA_COMPUTE_CAP} -lt 90 ]; \
@@ -62,7 +62,9 @@ RUN if [ ${CUDA_COMPUTE_CAP} -ge 75 -a ${CUDA_COMPUTE_CAP} -lt 80 ]; \
 
 COPY --from=planner /usr/src/recipe.json recipe.json
 
-RUN if [ ${CUDA_COMPUTE_CAP} -ge 75 -a ${CUDA_COMPUTE_CAP} -lt 80 ]; \
+RUN --mount=type=secret,id=actions_cache_url,env=ACTIONS_CACHE_URL \
+    --mount=type=secret,id=actions_runtime_token,env=ACTIONS_RUNTIME_TOKEN \
+    if [ ${CUDA_COMPUTE_CAP} -ge 75 -a ${CUDA_COMPUTE_CAP} -lt 80 ]; \
     then \
         cargo chef cook --release --features candle-cuda-turing --features static-linking --no-default-features --recipe-path recipe.json && sccache -s; \
     else \
@@ -77,7 +79,9 @@ COPY Cargo.lock ./
 
 FROM builder AS http-builder
 
-RUN if [ ${CUDA_COMPUTE_CAP} -ge 75 -a ${CUDA_COMPUTE_CAP} -lt 80 ]; \
+RUN --mount=type=secret,id=actions_cache_url,env=ACTIONS_CACHE_URL \
+    --mount=type=secret,id=actions_runtime_token,env=ACTIONS_RUNTIME_TOKEN \
+    if [ ${CUDA_COMPUTE_CAP} -ge 75 -a ${CUDA_COMPUTE_CAP} -lt 80 ]; \
     then \
         cargo build --release --bin text-embeddings-router -F candle-cuda-turing -F static-linking -F http --no-default-features && sccache -s; \
     else \
@@ -98,7 +102,9 @@ RUN PROTOC_ZIP=protoc-21.12-linux-x86_64.zip && \
 
 COPY proto proto
 
-RUN if [ ${CUDA_COMPUTE_CAP} -ge 75 -a ${CUDA_COMPUTE_CAP} -lt 80 ]; \
+RUN --mount=type=secret,id=actions_cache_url,env=ACTIONS_CACHE_URL \
+    --mount=type=secret,id=actions_runtime_token,env=ACTIONS_RUNTIME_TOKEN \
+    if [ ${CUDA_COMPUTE_CAP} -ge 75 -a ${CUDA_COMPUTE_CAP} -lt 80 ]; \
     then \
         cargo build --release --bin text-embeddings-router -F candle-cuda-turing -F static-linking -F grpc --no-default-features && sccache -s; \
     else \

diff --git a/Dockerfile-cuda-all b/Dockerfile-cuda-all
@@ -36,8 +36,6 @@ ARG DOCKER_LABEL
 ARG VERTEX="false"
 
 # sccache specific variables
-ARG ACTIONS_CACHE_URL
-ARG ACTIONS_RUNTIME_TOKEN
 ARG SCCACHE_GHA_ENABLED
 
 # Limit parallelism
@@ -49,28 +47,36 @@ WORKDIR /usr/src
 
 COPY --from=planner /usr/src/recipe.json recipe.json
 
-RUN if [ $VERTEX = "true" ]; \
+RUN --mount=type=secret,id=actions_cache_url,env=ACTIONS_CACHE_URL \
+    --mount=type=secret,id=actions_runtime_token,env=ACTIONS_RUNTIME_TOKEN \
+    if [ $VERTEX = "true" ]; \
     then \
       cargo chef cook --release --features google --recipe-path recipe.json && sccache -s; \
     else \
       cargo chef cook --release --recipe-path recipe.json && sccache -s; \
     fi;
 
-RUN if [ $VERTEX = "true" ]; \
+RUN --mount=type=secret,id=actions_cache_url,env=ACTIONS_CACHE_URL \
+    --mount=type=secret,id=actions_runtime_token,env=ACTIONS_RUNTIME_TOKEN \
+    if [ $VERTEX = "true" ]; \
     then \
       CUDA_COMPUTE_CAP=75 cargo chef cook --release --features google --features candle-cuda-turing --recipe-path recipe.json && sccache -s; \
     else \
       CUDA_COMPUTE_CAP=75 cargo chef cook --release --features candle-cuda-turing --recipe-path recipe.json && sccache -s; \
     fi;
 
-RUN if [ $VERTEX = "true" ]; \
+RUN --mount=type=secret,id=actions_cache_url,env=ACTIONS_CACHE_URL \
+    --mount=type=secret,id=actions_runtime_token,env=ACTIONS_RUNTIME_TOKEN \
+    if [ $VERTEX = "true" ]; \
     then \
       CUDA_COMPUTE_CAP=80 cargo chef cook --release --features google --features candle-cuda --recipe-path recipe.json && sccache -s; \
     else \
       CUDA_COMPUTE_CAP=80 cargo chef cook --release --features candle-cuda --recipe-path recipe.json && sccache -s; \
     fi;
 
-RUN if [ $VERTEX = "true" ]; \
+RUN --mount=type=secret,id=actions_cache_url,env=ACTIONS_CACHE_URL \
+    --mount=type=secret,id=actions_runtime_token,env=ACTIONS_RUNTIME_TOKEN \
+    if [ $VERTEX = "true" ]; \
     then \
       CUDA_COMPUTE_CAP=90 cargo chef cook --release --features google --features candle-cuda --recipe-path recipe.json && sccache -s; \
     else \
@@ -83,7 +89,9 @@ COPY router router
 COPY Cargo.toml ./
 COPY Cargo.lock ./
 
-RUN if [ $VERTEX = "true" ]; \
+RUN --mount=type=secret,id=actions_cache_url,env=ACTIONS_CACHE_URL \
+    --mount=type=secret,id=actions_runtime_token,env=ACTIONS_RUNTIME_TOKEN \
+    if [ $VERTEX = "true" ]; \
     then \
         CUDA_COMPUTE_CAP=75 cargo build --release --bin text-embeddings-router -F candle-cuda-turing -F google  && sccache -s; \
     else \
@@ -92,7 +100,9 @@ RUN if [ $VERTEX = "true" ]; \
 
 RUN mv /usr/src/target/release/text-embeddings-router /usr/src/target/release/text-embeddings-router-75
 
-RUN if [ $VERTEX = "true" ]; \
+RUN --mount=type=secret,id=actions_cache_url,env=ACTIONS_CACHE_URL \
+    --mount=type=secret,id=actions_runtime_token,env=ACTIONS_RUNTIME_TOKEN \
+    if [ $VERTEX = "true" ]; \
     then \
         CUDA_COMPUTE_CAP=80 cargo build --release --bin text-embeddings-router -F candle-cuda -F google  && sccache -s; \
     else \
@@ -101,7 +111,9 @@ RUN if [ $VERTEX = "true" ]; \
 
 RUN mv /usr/src/target/release/text-embeddings-router /usr/src/target/release/text-embeddings-router-80
 
-RUN if [ $VERTEX = "true" ]; \
+RUN --mount=type=secret,id=actions_cache_url,env=ACTIONS_CACHE_URL \
+    --mount=type=secret,id=actions_runtime_token,env=ACTIONS_RUNTIME_TOKEN \
+    if [ $VERTEX = "true" ]; \
     then \
         CUDA_COMPUTE_CAP=90 cargo build --release --bin text-embeddings-router -F candle-cuda -F google  && sccache -s; \
     else \

diff --git a/Dockerfile-intel b/Dockerfile-intel
@@ -24,13 +24,13 @@ ARG GIT_SHA
 ARG DOCKER_LABEL
 
 # sccache specific variables
-ARG ACTIONS_CACHE_URL
-ARG ACTIONS_RUNTIME_TOKEN
 ARG SCCACHE_GHA_ENABLED
 
 COPY --from=planner /usr/src/recipe.json recipe.json
 
-RUN cargo chef cook --release --features python --no-default-features --recipe-path recipe.json && sccache -s
+RUN --mount=type=secret,id=actions_cache_url,env=ACTIONS_CACHE_URL \
+    --mount=type=secret,id=actions_runtime_token,env=ACTIONS_RUNTIME_TOKEN \
+    cargo chef cook --release --features python --no-default-features --recipe-path recipe.json && sccache -s
 
 COPY backends backends
 COPY core core
@@ -46,13 +46,17 @@ RUN PROTOC_ZIP=protoc-21.12-linux-x86_64.zip && \
 
 FROM builder as http-builder
 
-RUN cargo build --release --bin text-embeddings-router -F python -F http --no-default-features && sccache -s
+RUN --mount=type=secret,id=actions_cache_url,env=ACTIONS_CACHE_URL \
+    --mount=type=secret,id=actions_runtime_token,env=ACTIONS_RUNTIME_TOKEN \
+    cargo build --release --bin text-embeddings-router -F python -F http --no-default-features && sccache -s
 
 FROM builder as grpc-builder
 
 COPY proto proto
 
-RUN cargo build --release --bin text-embeddings-router -F grpc -F python --no-default-features && sccache -s
+RUN --mount=type=secret,id=actions_cache_url,env=ACTIONS_CACHE_URL \
+    --mount=type=secret,id=actions_runtime_token,env=ACTIONS_RUNTIME_TOKEN \
+    cargo build --release --bin text-embeddings-router -F grpc -F python --no-default-features && sccache -s
 
 FROM intel/intel-optimized-pytorch:2.4.0-pip-base AS cpu
 ENV HUGGINGFACE_HUB_CACHE=/data \

diff --git a/backends/proto/embed.proto b/backends/proto/embed.proto
@@ -37,4 +37,4 @@ message Score {
 
 message PredictResponse {
     repeated Score scores = 1;
-}
+}
-Original file line number
+Diff line change
@@ Expand Up / @@ -37,4 +37,4 @@ message Score { @@
     message PredictResponse {
         repeated Score scores = 1;
-    }
+    }