Gcloud ML-Engine Prection Error OOM 429

问题

I'm getting the following error when trying to use gcloud ml-engine predict

ERROR: (gcloud.ml-engine.predict) HTTP request failed. Response: {
  "error": {
    "code": 429,
    "message": "Prediction server is out of memory, possibly because model size is too big.",
    "status": "RESOURCE_EXHAUSTED"
  }
}

My model size is 151 mb, I'm also using Tensorflow version 1.4 that does not requiere variables folder. When performing prediction it uses over 2gb. I'm using a modified version of inception.

回答1:

Currently, the machines used for prediction have only 2 GB of RAM. We are working on bringing machines with more RAM to the service.

That said, "vanilla" inception models are usually about the same size on disk as you are reporting but tend to easily fit in 2 GB of RAM. Is the explosion of RAM expected given the changes that you've made?

来源：https://stackoverflow.com/questions/49820507/gcloud-ml-engine-prection-error-oom-429