mirror of https://github.com/kubeflow/examples.git
34 lines
1.1 KiB
Markdown
34 lines
1.1 KiB
Markdown
# Setup
|
|
|
|
## Deploying Kubeflow to Google Cloud Platform
|
|
This example requires a running Kubeflow environment (v0.5.0). The easiest way to setup a Kubeflow environment is by using the [Deployment UI](https://www.kubeflow.org/docs/gke/deploy/deploy-ui/).
|
|
|
|
## Set environment variables
|
|
|
|
Create the following environment variables, follow the [documenation](https://cloud.google.com/resource-manager/docs/creating-managing-projects#identifying_projects) to get the project id :
|
|
|
|
```bash
|
|
export BUCKET=your-bucket-name
|
|
export PROJECT_ID=your-gcp-project-id
|
|
```
|
|
|
|
## Create bucket
|
|
Create a bucket that will contain everything required for our Kubeflow pipeline.
|
|
|
|
```bash
|
|
gsutil mb -c regional -l us-east1 gs://${BUCKET}
|
|
```
|
|
|
|
## Clone this repository
|
|
Clone the following repository, which contains everything needed for this example.
|
|
|
|
```bash
|
|
git clone https://github.com/kubeflow/examples.git
|
|
```
|
|
|
|
Open a Terminal and navigate to the folder `/examples/named-entity-recognition/`.
|
|
|
|
*Next*: [Build the pipeline components](step-2-build-components.md)
|
|
|
|
*Previous*: [Index](../README.md)
|