Bump vLLM version to v0.10.2rc2 (#2902)

### What this PR does / why we need it?

Upgrade vLLM version to 0.10.2rc2

### Does this PR introduce _any_ user-facing change?

Yes, image will use 0.10.2rc2 vLLM

### How was this patch tested?

- vLLM version: main
- vLLM main:
f17c075884

Signed-off-by: Yikun Jiang <yikunkero@gmail.com>

.github/workflows/accuracy_test.yaml

@@ -112,7 +112,7 @@ jobs:
         uses: actions/checkout@v4
         with:
           repository: vllm-project/vllm
-          ref: main
+          ref: v0.10.2rc2
           path: ./vllm-empty
 
       - name: Install vllm-project/vllm from source

.github/workflows/nightly_benchmarks.yaml

@@ -51,7 +51,7 @@ jobs:
     strategy:
       matrix:
         include:
-          - vllm_branch: main
+          - vllm_branch: v0.10.2rc2
             vllm_ascend_branch: main
             vllm_use_v1: 1
       max-parallel: 1

.github/workflows/vllm_ascend_dist.yaml

@@ -43,7 +43,7 @@ jobs:
     strategy:
       matrix:
         os: [linux-aarch64-a3-8]
-        vllm_version: [main]
+        vllm_version: [v0.10.2rc2]
     name: vLLM Ascend test
     runs-on: ${{ matrix.os }}
     container:

.github/workflows/vllm_ascend_test.yaml

@@ -82,7 +82,7 @@ jobs:
         VLLM_USE_MODELSCOPE: True
     strategy:
       matrix:
-        vllm_version: [main]
+        vllm_version: [v0.10.2rc2]
     steps:
       - name: Install packages
         run: |
@@ -138,7 +138,7 @@ jobs:
       max-parallel: 2
       matrix:
         os: [linux-aarch64-a2-1]
-        vllm_version: [main]
+        vllm_version: [v0.10.2rc2]
     name: singlecard e2e test - light
     runs-on: ${{ matrix.os }}
     container:
@@ -203,7 +203,7 @@ jobs:
       max-parallel: 2
       matrix:
         os: [linux-aarch64-a2-2]
-        vllm_version: [main]
+        vllm_version: [v0.10.2rc2]
     name: multicard e2e test - light
     runs-on: ${{ matrix.os }}
     container:

.github/workflows/vllm_ascend_test_310p.yaml

@@ -53,7 +53,7 @@ jobs:
       max-parallel: 2
       matrix:
         os: [linux-aarch64-310p-1, linux-aarch64-310p-4]
-        vllm_version: [main]
+        vllm_version: [v0.10.2rc2]
     name: 310p e2e test
     runs-on: ${{ matrix.os }}
     container:

.github/workflows/vllm_ascend_test_full.yaml

@@ -72,7 +72,7 @@ jobs:
       max-parallel: 2
       matrix:
         os: [linux-aarch64-a2-1]
-        vllm_version: [main]
+        vllm_version: [v0.10.2rc2]
     name: singlecard e2e test - full
     runs-on: ${{ matrix.os }}
     container:
@@ -156,7 +156,7 @@ jobs:
       max-parallel: 2
       matrix:
         os: [linux-aarch64-a2-2]
-        vllm_version: [main]
+        vllm_version: [v0.10.2rc2]
     name: multicard e2e test - full
     runs-on: ${{ matrix.os }}
     container:

Dockerfile

@@ -37,7 +37,7 @@ RUN pip config set global.index-url ${PIP_INDEX_URL}
 
 # Install vLLM
 ARG VLLM_REPO=https://github.com/vllm-project/vllm.git
-ARG VLLM_TAG=main
+ARG VLLM_TAG=v0.10.2rc2
 RUN git clone --depth 1 $VLLM_REPO --branch $VLLM_TAG /vllm-workspace/vllm
 # In x86, triton will be installed by vllm. But in Ascend, triton doesn't work correctly. we need to uninstall it.
 RUN VLLM_TARGET_DEVICE="empty" python3 -m pip install -v -e /vllm-workspace/vllm/ --extra-index https://download.pytorch.org/whl/cpu/ && \

Dockerfile.310p

@@ -37,7 +37,7 @@ RUN pip config set global.index-url ${PIP_INDEX_URL}
 
 # Install vLLM
 ARG VLLM_REPO=https://github.com/vllm-project/vllm.git
-ARG VLLM_TAG=main
+ARG VLLM_TAG=v0.10.2rc2
 RUN git clone --depth 1 $VLLM_REPO --branch $VLLM_TAG /vllm-workspace/vllm
 # In x86, triton will be installed by vllm. But in Ascend, triton doesn't work correctly. we need to uninstall it.
 RUN VLLM_TARGET_DEVICE="empty" python3 -m pip install -v -e /vllm-workspace/vllm/ --extra-index https://download.pytorch.org/whl/cpu/ && \

Dockerfile.310p.openEuler

@@ -34,7 +34,7 @@ COPY . /vllm-workspace/vllm-ascend/
 
 # Install vLLM
 ARG VLLM_REPO=https://github.com/vllm-project/vllm.git
-ARG VLLM_TAG=main
+ARG VLLM_TAG=v0.10.2rc2
 
 RUN git clone --depth 1 $VLLM_REPO --branch $VLLM_TAG /vllm-workspace/vllm
 # In x86, triton will be installed by vllm. But in Ascend, triton doesn't work correctly. we need to uninstall it.

Dockerfile.a3

@@ -37,7 +37,7 @@ RUN pip config set global.index-url ${PIP_INDEX_URL}
 
 # Install vLLM
 ARG VLLM_REPO=https://github.com/vllm-project/vllm.git
-ARG VLLM_TAG=main
+ARG VLLM_TAG=v0.10.2rc2
 RUN git clone --depth 1 $VLLM_REPO --branch $VLLM_TAG /vllm-workspace/vllm
 # In x86, triton will be installed by vllm. But in Ascend, triton doesn't work correctly. we need to uninstall it.
 RUN VLLM_TARGET_DEVICE="empty" python3 -m pip install -v -e /vllm-workspace/vllm/ --extra-index https://download.pytorch.org/whl/cpu/ && \

Dockerfile.a3.openEuler

@@ -34,7 +34,7 @@ COPY . /vllm-workspace/vllm-ascend/
 
 # Install vLLM
 ARG VLLM_REPO=https://github.com/vllm-project/vllm.git
-ARG VLLM_TAG=main
+ARG VLLM_TAG=v0.10.2rc2
 
 RUN git clone --depth 1 $VLLM_REPO --branch $VLLM_TAG /vllm-workspace/vllm
 # In x86, triton will be installed by vllm. But in Ascend, triton doesn't work correctly. we need to uninstall it.

Dockerfile.openEuler

@@ -34,7 +34,7 @@ COPY . /vllm-workspace/vllm-ascend/
 
 # Install vLLM
 ARG VLLM_REPO=https://github.com/vllm-project/vllm.git
-ARG VLLM_TAG=main
+ARG VLLM_TAG=v0.10.2rc2
 
 RUN git clone --depth 1 $VLLM_REPO --branch $VLLM_TAG /vllm-workspace/vllm
 # In x86, triton will be installed by vllm. But in Ascend, triton doesn't work correctly. we need to uninstall it.

docs/source/conf.py

@@ -77,7 +77,7 @@ myst_substitutions = {
     # CANN image tag
     'cann_image_tag': "8.2.rc1-910b-ubuntu22.04-py3.11",
     # vllm version in ci
-    'ci_vllm_version': 'main',
+    'ci_vllm_version': 'v0.10.2rc2',
 }
 
 # Add any paths that contain templates here, relative to this directory.