0

I'm trying to run XGBoost example pipline in MiniKF installed in GCP through application manager. I have a free account with free credits and apparently the max number of CPUs for this account is 8. MiniKF was installed and other example pipelines ran without issues. However, when I try to run XGBoost pipeline I'm getting an error that quotas for 'CPUS' and 'CPUS_ALL_REGIONS' exceeded (12 CPUs are required).

Source code: https://github.com/kubeflow/pipelines/tree/0.1.40/samples/core/xgboost_training_cm

Is there any way to reduce the number of required CPUs in this XGBoost example pipeline?

Konstantin
  • 2,937
  • 10
  • 41
  • 58

1 Answers1

2

Your free trial credit applies to all Google Cloud resources, with the following exceptions:

You can't have more than 8 cores (or virtual CPUs) running at the same time.

You can't add GPUs to your VM instances.

You can't request a quota increase.

You can't create VM instances that are based on Windows Server images.

You must upgrade your account to perform any of the actions in the preceding list.

Naman Parekh
  • 176
  • 6
  • thank you for the reply, I'm aware of this. But is it possible to lower the requirements for the XGBoost example? Why does it need so many CPUs? – Konstantin Apr 14 '20 at 07:30
  • 1
    XGBoost is a powerful machine learning algorithm especially where speed and accuracy are concerned.We need to consider different parameters and their values to be specified while implementing an XGBoost model.The XGBoost model requires parameter tuning to improve and fully leverage its advantages over other algorithms.Computing different algorithms required high computation power which on hand required high core CPU or GPU to perform the task. – Naman Parekh Apr 15 '20 at 01:36
  • XGBoost can run on a single laptop with 2 cores. In my question I wanted to understand if and how it's possible to lower the requirements for this particular example to run. – Konstantin Apr 15 '20 at 11:35