diff --git a/.github/workflows/_e2e_nightly_single_node.yaml b/.github/workflows/_e2e_nightly_single_node.yaml
index df91260f..bff4f0e2 100644
--- a/.github/workflows/_e2e_nightly_single_node.yaml
+++ b/.github/workflows/_e2e_nightly_single_node.yaml
@@ -57,6 +57,9 @@ jobs:
     timeout-minutes: 600
     container:
       image: ${{ inputs.image }}
+    env:
+      TRANSFORMERS_OFFLINE: 1
+      VLLM_USE_MODELSCOPE: True
     steps:
       - name: Check npu and CANN info
         run: |
diff --git a/.github/workflows/pr_tag_image_build_and_push.yaml b/.github/workflows/pr_tag_image_build_and_push.yaml
index 4991ae04..a417fb2c 100644
--- a/.github/workflows/pr_tag_image_build_and_push.yaml
+++ b/.github/workflows/pr_tag_image_build_and_push.yaml
@@ -26,7 +26,7 @@ on:
       - 'cmake/**'
       - 'CMakeLists.txt'
       - 'csrc/**'
-    types: [ labeled, synchronize ]
+    types: [ labeled ]
   push:
     # Publish image when tagging, the Dockerfile in tag will be build as tag image
     branches:
diff --git a/tests/e2e/conftest.py b/tests/e2e/conftest.py
index a8207e3f..b59ab7ce 100644
--- a/tests/e2e/conftest.py
+++ b/tests/e2e/conftest.py
@@ -162,7 +162,7 @@ class RemoteOpenAIServer:
         self.proxy_port = proxy_port
 
         self._start_server(model, vllm_serve_args, env_dict)
-        max_wait_seconds = max_wait_seconds or 1800
+        max_wait_seconds = max_wait_seconds or 2800
         if self.disaggregated_prefill:
             assert proxy_port is not None, "for disaggregated_prefill, proxy port must be provided"
             self._wait_for_server_pd(timeout=max_wait_seconds)