Pipelines and Blocks – Accessing Pre-built Pipelines
More articles in Accessing and Manipulating Data
You want to execute pre-built pipelines. These pipelines have been built by you, or available in the marketplace.
You need multiple capabilities to be able to access pre-built pipelines
- List available pipelines filtered on descriptions, tags, authors etc.
- Review the pipeline documentation
- Import and use a pipeline
- # list all pipelines training MNIST razor.pipelines().filter(contains("description:","MNIST")) # list all blocks from marketplace split strings razor.marketplace().pipelines().filter(contains("description:","MNIST")) # list all blocks written by an author razor.pipelines().filter(equals("author","sandeep")) # install a block from the marketplace to local p = razor.marketplace().pipelines("7891232945") p.download() # describe a local block razor.pipelines("7891232945").describe()
Pre-built pipelines are available for all types of use cases that you might need, from structure data ML algorithms to . Once you familiarize yourself with the structure, you can put rapidly put together a pipeline and address a wide variety of use cases simply and effectively.
It is also possible to build your own inventory of pipelines that are shared across the organization. The following quick video provides an introduction to providing your own pre-built pipelines.