Sign In

Dataster Documentation

Dataster helps you build Generative AI applications with better accuracy and lower latency.
Back to documentation

Bring your own Amazon Bedrock model

Dataster supports many Amazon Bedrock models both for completions and embeddings. The full list of models can be found on our pricing page here. Dataster provides off-the-shelf access to each supported model, which is shared across several users. However, users can choose to bring their own Amazon Bedrock model into Dataster for isolated and private usage, eliminating potential competition for resources with other users. The process of bringing your own Amazon Bedrock model into Dataster is straightforward.

Prerequisites

  1. A Dataster account.
  2. An existing and publicly accessible model in an AWS account.
  3. An access key and a secret access key for the AWS account with the necessary permissions to invoke the model.

Step 1: Navigate to the LLM Catalog

  1. Navigate to the LLM catalog by clicking "LLM" in the left navigation pane.

Step 2: Add an LLM

  1. Click BYO LLM.
  2. Fill in a unique name for the LLM across the account.
  3. Set the provider to Amazon Bedrock.
  4. Use the drop-down menu to indicate which model is deployed.
  5. Fill in the AWS region where the model is accessible.
  6. Fill in the access key and secret access key.


Add an AOAI Model

Step 3: View the Model

  1. Return to the LLM Catalog.
  2. The model appears in the order it was created.
  3. Optionally filter the models.

Conclusion

You have successfully brought your own Amazon Bedrock model in Dataster. This model is now available for use case evaluation or can be combined with a vector store for Retrieval-Augmented Generation (RAG).


If you encounter any issues or need further assistance, please contact our support team at support@dataster.com.