[Deps] Upgrade CUDA PyTorch stack to 2.13 (#28836)

Co-authored-by: Brayden Zhong <b8zhong@uwaterloo.ca>

.github/workflows/pr-test.yml

@@ -59,6 +59,8 @@ env:
   SGLANG_CUDA_COREDUMP: "1"
   SGLANG_JIT_DEEPGEMM_FAST_WARMUP: true
   SKIP_PR_TEST_HEALTH_CHECK: ${{ (inputs.skip_pr_test_health_check == true || inputs.test_parallel_dispatch == true || inputs.run_all_tests == true) && 'true' || 'false' }}
+  # TEMP: force DeepEP rebuilds during the initial Torch 2.13 rollout so every CI runner replaces its Torch 2.11 build; remove after the runner fleet has cycled.
+  FORCE_REBUILD_DEEPEP: '1'
   # Schedule / main-branch dispatch / workflow_call from main use refs/heads/main; PR events use refs/pull/*/merge
   PR_TEST_BYPASS_MAINTENANCE_ON_MAIN: ${{ github.ref == 'refs/heads/main' && 'true' || 'false' }}
   USE_VENV: false