ci(release-pypi): make release pipeline self-consistent (#2062)

.github/workflows/release-fake-tag.yml

@@ -30,7 +30,12 @@ jobs:
 
       - name: Create and push tag
         run: |
+          TAG=${{ steps.get_version.outputs.TAG }}
           git config user.name "ktransformers-bot"
           git config user.email "ktransformers-bot@users.noreply.github.com"
-          git tag ${{ steps.get_version.outputs.TAG }}
-          git push origin ${{ steps.get_version.outputs.TAG }}
+          if git ls-remote --tags --exit-code origin "refs/tags/${TAG}" > /dev/null 2>&1; then
+            echo "Tag ${TAG} already exists on origin, skipping."
+            exit 0
+          fi
+          git tag "${TAG}"
+          git push origin "${TAG}"

.github/workflows/release-pypi.yml

@@ -101,7 +101,7 @@ jobs:
         run: |
           apt-get update && apt-get install -y cmake libhwloc-dev pkg-config libnuma-dev
           python -m pip install --upgrade pip
-          pip install build wheel setuptools torch --index-url https://download.pytorch.org/whl/cu118
+          pip install build wheel setuptools torch --index-url https://download.pytorch.org/whl/cu128
 
       - name: Build kt-kernel wheel
         working-directory: kt-kernel
@@ -109,15 +109,15 @@ jobs:
           CPUINFER_BUILD_ALL_VARIANTS: '1'
           CPUINFER_ENABLE_CPPTRACE: '0'
           CPUINFER_USE_CUDA: '1'
-          CPUINFER_CUDA_ARCHS: '80;86;89;90'
+          CPUINFER_CUDA_ARCHS: '80;86;89;90;120'
           CPUINFER_CUDA_STATIC_RUNTIME: '1'
           CPUINFER_BUILD_TYPE: 'Release'
           CPUINFER_PARALLEL: '4'
           CPUINFER_FORCE_REBUILD: '1'
-          CUDA_HOME: '/usr/local/cuda-11.8'
+          CUDA_HOME: '/usr/local/cuda-12.8'
         run: |
           echo "Building kt-kernel with:"
-          echo "  - CUDA support (SM 80, 86, 89, 90)"
+          echo "  - CUDA support (SM 80, 86, 89, 90, 120)"
           echo "  - CPU multi-variant (AMX, AVX512, AVX2)"
           python -m build --wheel -v
 

.github/workflows/release-sglang-kt.yml

@@ -6,7 +6,6 @@ on:
       - main
     paths:
       - "third_party/sglang"
-      - "version.py"
   workflow_dispatch:
     inputs:
       test_pypi:

MANIFEST.in

@@ -0,0 +1 @@
+include version.py

setup.py

@@ -23,7 +23,7 @@ setup(
             "accelerate-kt==1.14.0.post1",
         ],
         "sglang": [
-            "sglang-kt==0.6.3",
+            f"sglang-kt=={_v}",
         ],
     },
 )