Watson Machine Learning plans and compute usage

Limits and defaults for retaining deployment jobs (Watson Machine Learning)

Watson Machine Learning places limits on the number of deployment jobs retained for each single deployment space. Importantly, with this update, none of your information will be lost, but your user experience when running deployment jobs might need to change.

If you exceed your limit, you will be unable to create new deployment jobs until you delete existing jobs or upgrade your plan. New automation will help you stay within the plan limits. By default, jobs metadata will be auto-delete after 30 days. You can override this value when creating a job. See Managing jobs.

CUH usage details

• Machine Learning compute usage is calculated by the number of capacity unit hours (CUH) consumed by an active machine learning instance.
• The rate of capacity units per hour consumed is determined by the computing requirements of your Machine Learning assets and models. For example, a model with a large, complex data set will consume more training resources than a model with a smaller, simpler data set. Note that scaling a deployment to support more concurrent users and requests also increases CUH consumption.
• Because there are so many variables that affect resource consumption for a deployment, we recommend you run tests on your models and deployments to analyze CUH consumption.
• Compute time is calculated to the millisecond. However, there is a one-minute minimum for each distinct operation. That is, a training run that takes 12 seconds is billed as one minute toward the capacity unit hour quota, while a training run that takes 83.555 seconds is billed exactly as calculated.
• The way that online deployments consume capacity units is based on framework. For some frameworks, CUH is charged for the number of hours the deployment asset is active in a deployment space. For example, SPSS models in online deployment mode that run 24 hours a day for seven days a week consume CUH and are charged for that period. There is no idle time for an active online deployment. For other frameworks, CUH is charged according to scoring duration. See the CUH consumption table for details on how CUH is calculated.

Determining whether a deployment is active

If a deployment does not receive scoring requests for a given duration, it is treated as inactive, or idle, and billing stops for all frameworks other than SPSS. The period used to determine when a deployment is inactive depends on the plan type.

CUH consumption by deployment and framework type

CUH consumption is calculated using these formulas:

These tables show the capacity units per hour calculation for machine learning environments, by usage type.

Capacity units per hour for training, evaluating, or scoring models

Capacity units per hour for AutoAI experiments

When data is joined for an AutoAI experiment, capacity units per hour are calculated differently as data is being joined. Once the data is joined, the CUH consumption reverts to the calculations for an experiment with a single data source.

Capacity units per hour for AutoAI experiments using a single data source:

Capacity units per hour for AutoAI experiments while joining data:

Note that consumption increases for each executor you add.

Capacity units per hour for Decision Optimization experiments

These plans apply to Decision Optimization experiments run in Watson Studio.

Capacity units per hour for Decision Optimization in Watson Machine Learning

These plans apply to Decision Optimization deployed and run from Watson Machine Learning.

For details on how resources are consumed, see Monitoring account resource usage

Track CUH consumption for machine learning in a notebook

To calculate capacity unit hours, use:

CP =  client.service_instance.get_details()
CUH = CUH["entity"]["usage"]["capacity_units"]["current"]/(3600*1000)
print(CUH)

For example:

'capacity_units': {'current': 19773430}

19773430/(3600*1000)

returns 5.49 CUH

For details, see the Service Instances section of the IBM Watson Machine Learning API documentation.