Yan Li
Results
1
comments of
Yan Li
Maybe you can try: tritonserver xxx --model-load-thread-count=1. I encountered the same issue when starting model1: python backend, model2: vllm backend, and both models have EXECUTION_ENV_PATH.