diff --git a/.github/workflows/benchmark.yml b/.github/workflows/benchmark.yml
index 265971b24a..cf5b117aad 100644
--- a/.github/workflows/benchmark.yml
+++ b/.github/workflows/benchmark.yml
@@ -299,6 +299,10 @@ jobs:
       run: NV=1 DEFAULT_FLOAT=HALF BENCHMARK=10 BS=256 GPUS=1 MODEL=resnet python3 examples/mlperf/model_train.py | tee train_resnet_one_gpu.txt
     - name: Run 10 MLPerf ResNet50 training steps (6 gpu)
       run: NV=1 CAPTURE_PROCESS_REPLAY=0 DEFAULT_FLOAT=HALF BENCHMARK=10 BS=1536 GPUS=6 MODEL=resnet python3 examples/mlperf/model_train.py | tee train_resnet.txt
+    - name: Run 10 MLPerf Bert training steps (6 gpu)
+      # TODO: remove DISABLE_DROPOUT once dropout is fixed
+      # TODO: remove BERT_LAYERS once scheduler is fast
+      run: NV=1 CAPTURE_PROCESS_REPLAY=0 DEFAULT_FLOAT=HALF BENCHMARK=10 BS=66 GPUS=6 DISABLE_DROPOUT=1 BERT_LAYERS=2 MODEL=bert python3 examples/mlperf/model_train.py | tee train_bert.txt
     - uses: actions/upload-artifact@v4
       with:
         name: Speed (NVIDIA Training)
@@ -309,9 +313,10 @@ jobs:
           train_cifar_bf16.txt
           train_cifar_wino.txt
           train_cifar_one_gpu.txt
+          train_cifar_six_gpu.txt
           train_resnet.txt
           train_resnet_one_gpu.txt
-          train_cifar_six_gpu.txt
+          train_bert.txt
     - name: Run process replay tests
       run: cp test/external/process_replay/process_replay.py ./process_replay.py && git fetch origin master && git -c advice.detachedHead=false checkout origin/master && PYTHONPATH=. python3 process_replay.py
 
@@ -492,6 +497,10 @@ jobs:
       run: AMD=1 DEFAULT_FLOAT=HALF BENCHMARK=10 BS=256 GPUS=1 MODEL=resnet python3 examples/mlperf/model_train.py | tee train_resnet_one_gpu.txt
     - name: Run 10 MLPerf ResNet50 training steps (6 gpu)
       run: AMD=1 CAPTURE_PROCESS_REPLAY=0 DEFAULT_FLOAT=HALF BENCHMARK=10 BS=1536 GPUS=6 MODEL=resnet python3 examples/mlperf/model_train.py | tee train_resnet.txt
+    - name: Run 10 MLPerf Bert training steps (6 gpu)
+      # TODO: remove DISABLE_DROPOUT once dropout is fixed
+      # TODO: remove BERT_LAYERS once scheduler is fast
+      run: AMD=1 CAPTURE_PROCESS_REPLAY=0 DEFAULT_FLOAT=HALF BENCHMARK=10 BS=66 GPUS=6 DISABLE_DROPOUT=1 BERT_LAYERS=2 MODEL=bert python3 examples/mlperf/model_train.py | tee train_bert.txt
     - uses: actions/upload-artifact@v4
       with:
         name: Speed (AMD Training)
@@ -502,9 +511,10 @@ jobs:
           train_cifar_bf16.txt
           train_cifar_wino.txt
           train_cifar_one_gpu.txt
+          train_cifar_six_gpu.txt
           train_resnet.txt
           train_resnet_one_gpu.txt
-          train_cifar_six_gpu.txt
+          train_bert.txt
     - name: Run process replay tests
       run: cp test/external/process_replay/process_replay.py ./process_replay.py && git fetch origin master && git -c advice.detachedHead=false checkout origin/master && PYTHONPATH=. python3 process_replay.py