Fixes for the MLPerf inference nightly test failures

.github/workflows/test-amd-mlperf-inference-implementations.yml

@@ -12,7 +12,7 @@ jobs:
         fail-fast: false
         matrix:
           python-version: [ "3.12" ]
-          model: [ "llama2-70b" ]
+          model: [ "llama2-70b-99.9" ]
       steps:
       - name: Test MLPerf Inference AMD (build only) ${{ matrix.model }}
         run: |

.github/workflows/test-mlperf-inference-dlrm.yml

@@ -25,7 +25,7 @@ jobs:
         export CM_REPOS=$HOME/GH_CM
         python3 -m pip install cm4mlops
         cm pull repo
-        cm run script --tags=run-mlperf,inference,_performance-only --submitter="MLCommons" --model=dlrm-v2-99 --implementation=reference --backend=pytorch --category=datacenter --scenario=Offline --execution_mode=test --device=${{ matrix.device }} --docker --quiet --test_query_count=1 --target_qps=1 --docker_it=no --docker_cm_repo=gateoverflow@cm4mlops --adr.compiler.tags=gcc --hw_name=gh_action --docker_dt=yes  --results_dir=$HOME/gh_action_results --clean
+        cm run script --tags=run-mlperf,inference,_performance-only --adr.mlperf-implementation.tags=_branch.dev --adr.mlperf-implementation.version=custom --submitter="MLCommons" --model=dlrm-v2-99 --implementation=reference --backend=pytorch --category=datacenter --scenario=Offline --execution_mode=test --device=${{ matrix.device }} --docker --quiet --test_query_count=1 --target_qps=1 --docker_it=no --docker_cm_repo=gateoverflow@cm4mlops --adr.compiler.tags=gcc --hw_name=gh_action --docker_dt=yes  --results_dir=$HOME/gh_action_results --clean
 
   build_intel:
     if: github.repository_owner == 'gateoverflow_off'

.github/workflows/test-mlperf-inference-mixtral.yml

@@ -5,7 +5,7 @@ name: MLPerf inference MIXTRAL-8x7B
 
 on:
   schedule:
-    - cron: "30 03 * * *"   # 30th minute and 20th hour => 20:30 UTC => 2 AM IST 
+    - cron: "45 10 * * *"   # 30th minute and 20th hour => 20:30 UTC => 2 AM IST 
 
 jobs:
   build_reference:
@@ -24,7 +24,9 @@ jobs:
         source gh_action/bin/deactivate || python3 -m venv gh_action
         source gh_action/bin/activate
         export CM_REPOS=$HOME/GH_CM
-        python3 -m pip install cm4mlops
+        pip install cm4mlops
+        pip install "huggingface_hub[cli]"
+        huggingface-cli login --token ${{ secrets.HF_TOKEN }} --add-to-git-credential
         cm pull repo
         cm run script --tags=run-mlperf,inference,_submission,_short --submitter="MLCommons" --model=mixtral-8x7b --implementation=reference --batch_size=1 --backend=${{ matrix.backend }} --category=datacenter --scenario=Offline --execution_mode=test --device=${{ matrix.device }} --docker_it=no --docker_cm_repo=gateoverflow@cm4mlops --adr.compiler.tags=gcc --hw_name=gh_action --docker_dt=yes  --results_dir=$HOME/gh_action_results --submission_dir=$HOME/gh_action_submissions --docker --quiet --test_query_count=1 --target_qps=1 --clean --env.CM_MLPERF_MODEL_MIXTRAL_8X7B_DOWNLOAD_TO_HOST=yes --env.CM_MLPERF_DATASET_MIXTRAL_8X7B_DOWNLOAD_TO_HOST=yes
         cm run script --tags=push,github,mlperf,inference,submission --repo_url=https://github.com/gateoverflow/mlperf_inference_test_submissions_v5.0 --repo_branch=main --commit_message="Results from self hosted Github actions - GO-i9" --quiet --submission_dir=$HOME/gh_action_submissions