storage.md

copyright	years 2025, 2026
lastupdated	2026-05-12
keywords	instructlab, ai, project
subcollection	inference

{{site.data.keyword.attribute-definition-list}}

Creating storage locations for {{site.data.keyword.short_name}}

{: #storage}

Complete the following steps to create one or more {{site.data.keyword.cos_short}} buckets to store {{site.data.keyword.short_name}} resources in. {: shortdesc}

{{site.data.keyword.cos_short}} buckets are only required if you want to fine-tune or align models with your own data. If you use the inference feature with pre-trained models, you do not need to create a bucket. {: important}

Learn more about {{site.data.keyword.cos_short}}.

Creating an {{site.data.keyword.cos_short}} instance in the console for {{site.data.keyword.short_name}}

{: #storage-ui} {: ui}

1. Navigate to the {{site.data.keyword.cos_short}} in the console.
2. Click Create an instance.
3. Choose a plan.
4. Give your instance a name, select a resource group, and enter any tags that you want to use.
5. Click Create.

Creating an {{site.data.keyword.cos_short}} instance and bucket by using the CLI for {{site.data.keyword.short_name}}

{: #storage-cli} {: cli}

To log in: {{_include-segments/login.md}}

To create {{site.data.keyword.cos_short}} resources:

1. Run the following command to create an instance.

   ibmcloud resource service-instance-create <instance-name> cloud-object-storage <plan> global

   {: pre}

2. Create a new bucket and make a note of the bucket name for later.

   ibmcloud cos bucket-create --bucket <bucket-name> [--class <class-name>] [--ibm-service-instance-id <instance-id>] [--region REGION] [--output FORMAT]

   {: pre}

What's next?

{: #whats-next-storage}

After creating your project, review the following links for next steps.

• Assign users in your account access to your project and {{site.data.keyword.cos_short}}.
• Learn how taxonomies are structured.
• Prepare your taxonomy and upload it to {{site.data.keyword.cos_short}}.