aws / sagemaker-inference-toolkit

Serve machine learning models within a 🐳 Docker container using 🧠 Amazon SageMaker.
Apache License 2.0
379 stars 82 forks source link

model-server archives model in /opt/ml/model #17

Closed ericangelokim closed 4 years ago

ericangelokim commented 4 years ago

Hello,

Is it an expected behavior that in multi model, the container would be run with one model in the single model dir /opt/ml/model? I noticed that the model server starts by running the model-archive in this directory. I wasn't aware that this was a requirement for starting MME in a container