Adding Cloud Providers

A cloud provider object is a conduit to access LLMs, embeddings models, or vector stores that are served by Aizen or by a non-Aizen vendor, such as OpenAI. To use an LLM, embeddings model, or vector store served on Aizen, you must add an Aizen cloud provider. To use an LLM or embeddings model served by a vendor, such as OpenAI, you must add an OpenAI cloud provider.

To add a cloud provider, follow these steps:

1. Log in to the Aizen Jupyter console. See Using the Aizen Jupyter Console.

2. Create an ML project if you have not already done so or set the current working project.

   create project <project name>

   or

   set project <project name>

3. Add the cloud provider by running the add cloudprovider command:

   add cloudprovider <cloud provider name>,<provider type>,<instance type>,{<credential file name>,<credential file section>|<base url>}

4. You must provide a unique provider name, and specify the provider type and instance type. The provider type can be aizen or openai. The instance type can be either llm for LLMs and embeddings models in Aizen or in OpenAI or vectorstore for vector stores in Aizen. If you plan to use both LLMs and vector stores, you must add cloud providers for LLMs or embeddings and cloud providers for vector stores.

   • If the provider type is aizen, you must specify the base URL of the provider. Use the status llm command to get the base URL for LLMs and embeddings models, or use the status vectorstore command to get the base URL for vector stores. Each unique base URL requires a unique cloud provider name.

   • If the provider type is openai, specify the OpenAI credentials.

   The add cloudprovider command supports other instance types, such as gpu. The GPU type is for resource providers configured by the configure resource command.

5. List the cloud providers by running the list cloudproviders command:

   list cloudproviders
   list cloudprovider <provider name>