You can run nearly any code on Valohai, allowing it to handle a wide variety of tasks, although the level of effort required may differ. We provide an extensive suite of tools designed to support data science workloads, including data preprocessing, model training, data management, and other related use cases.
However, there are some use cases that Valohai does not automatically support:
Not for: Building Your Actual Model Logic:
Valohai does not offer AutoML capabilities or drag-and-drop interfaces for building predictive models. Users must provide their own program logic using their preferred programming languages, such as Python, R, or C++. Valohai supports all programming languages, frameworks, and development tools.
Not for: Data Acquisition:
We integrate with all major cloud-based binary data sources, and you should use those to ingest your data. Valohai itself does not provide features for acquiring new data samples. Once the data is available in AWS S3, Azure Store, Google Cloud Store, OpenStack Swift, or on a local mount, you can begin using Valohai.
Not for: Data Labeling:
Although Valohai workers do have an Internet connection, they cannot be used to reliably host web servers or other persistent services. This is intentional, as they are designed to be ephemeral. In theory, you could host a labeling service on Valohai Deployment, but none of the tools are specifically built for labeling. There are other excellent labeling tools available, such as Labelbox.