* enable previous models * test mistral model * for test * run models separately * test all models * for test * revert the llm_performance_test.yaml
WeiguangHan