[diffusion] CI: fix consistency check (#22251)

.github/workflows/diffusion-ci-gt-gen.yml

@@ -22,6 +22,10 @@ permissions:
   contents: write
   actions: read
 
+env:
+  SGLANG_IS_IN_CI: true
+  SGLANG_CUDA_COREDUMP: "1"
+
 jobs:
   multimodal-diffusion-gen-1gpu:
     if: github.repository == 'sgl-project/sglang'
@@ -40,6 +44,8 @@ jobs:
         run: bash scripts/ci/cuda/ci_install_dependency.sh diffusion
 
       - name: Generate outputs
+        env:
+          RUNAI_STREAMER_MEMORY_LIMIT: 0
         run: |
           cd python
           python -m sglang.multimodal_gen.test.scripts.gen_diffusion_ci_outputs \
@@ -56,6 +62,11 @@ jobs:
           path: python/diffusion-ci-outputs
           retention-days: 7
 
+      - name: Publish GT images to sglang-bot/sglang-ci-data
+        env:
+          GITHUB_TOKEN: ${{ secrets.GH_PAT_FOR_NIGHTLY_CI_DATA }}
+        run: python scripts/ci/utils/diffusion/publish_diffusion_gt.py --source-dir python/diffusion-ci-outputs
+
   multimodal-diffusion-gen-2gpu:
     if: github.repository == 'sgl-project/sglang'
     runs-on: 2-gpu-h100
@@ -73,6 +84,8 @@ jobs:
         run: bash scripts/ci/cuda/ci_install_dependency.sh diffusion
 
       - name: Generate outputs
+        env:
+          RUNAI_STREAMER_MEMORY_LIMIT: 0
         run: |
           cd python
           python -m sglang.multimodal_gen.test.scripts.gen_diffusion_ci_outputs \
@@ -89,27 +102,42 @@ jobs:
           path: python/diffusion-ci-outputs
           retention-days: 7
 
-  diffusion-ci-push:
-    needs: [multimodal-diffusion-gen-1gpu, multimodal-diffusion-gen-2gpu]
+      - name: Publish GT images to sglang-bot/sglang-ci-data
+        env:
+          GITHUB_TOKEN: ${{ secrets.GH_PAT_FOR_NIGHTLY_CI_DATA }}
+        run: python scripts/ci/utils/diffusion/publish_diffusion_gt.py --source-dir python/diffusion-ci-outputs
+
+  multimodal-diffusion-gen-b200:
     if: github.repository == 'sgl-project/sglang'
-    runs-on: ubuntu-latest
+    runs-on: 4-gpu-b200
+    timeout-minutes: 240
     steps:
       - name: Checkout code
         uses: actions/checkout@v4
-
-      - name: Download artifacts
-        uses: actions/download-artifact@v4
         with:
-          pattern: diffusion-gen-*
-          path: combined
-          merge-multiple: true
+          ref: ${{ inputs.ref || github.ref }}
 
-      - name: Collect image files
+      - name: Install dependencies
+        run: bash scripts/ci/cuda/ci_install_dependency.sh diffusion
+
+      - name: Generate outputs
+        env:
+          RUNAI_STREAMER_MEMORY_LIMIT: 0
         run: |
-          mkdir -p gt_images
-          find combined \( -name "*.png" -o -name "*.jpg" -o -name "*.jpeg" -o -name "*.webp" \) -type f -exec cp -f {} gt_images/ \;
+          cd python
+          python -m sglang.multimodal_gen.test.scripts.gen_diffusion_ci_outputs \
+            --suite 1-gpu-b200 \
+            --out-dir ./diffusion-ci-outputs \
+            ${{ inputs.case_ids != '' && format('--case-ids {0}', inputs.case_ids) || '' }}
+
+      - name: Upload artifact
+        uses: actions/upload-artifact@v4
+        with:
+          name: diffusion-gen-b200
+          path: python/diffusion-ci-outputs
+          retention-days: 7
 
       - name: Publish GT images to sglang-bot/sglang-ci-data
         env:
           GITHUB_TOKEN: ${{ secrets.GH_PAT_FOR_NIGHTLY_CI_DATA }}
-        run: python scripts/ci/utils/diffusion/publish_diffusion_gt.py --source-dir gt_images
+        run: python scripts/ci/utils/diffusion/publish_diffusion_gt.py --source-dir python/diffusion-ci-outputs