Update benchmarks.yml

Author: juliagmt-google

.github/workflows/benchmarks.yml

@@ -92,13 +92,13 @@ jobs:
         continue-on-error: true
         run: nvidia-smi
 
-      # - name: Configure XLA
-      #   working-directory: xla
-      #   run: ./configure.py --backend CUDA --nccl
+      - name: Configure XLA
+        working-directory: xla
+        run: ./configure.py --backend CUDA --nccl
 
-      # - name: Set TF_CPP_MAX_VLOG_LEVEL
-      #   working-directory: xla
-      #   run: echo "TF_CPP_MAX_VLOG_LEVEL=1" >> $GITHUB_ENV
+      - name: Set TF_CPP_MAX_VLOG_LEVEL
+        working-directory: xla
+        run: echo "TF_CPP_MAX_VLOG_LEVEL=1" >> $GITHUB_ENV
 
       - name: Build run_hlo_module with GPU
         working-directory: xla