Models supported by the Bedrock RetrieveAndGenerate API?

0

In following the instructions to create a RAG-backed search experience I get the error below if I don't use a Claude model. Is Claude the only model supported by the RetrieveAndGenerate API?

So, this works:

    const command = new RetrieveAndGenerateCommand({
        input: {
            text: input
        },
        retrieveAndGenerateConfiguration: {
            type: 'KNOWLEDGE_BASE',
            knowledgeBaseConfiguration: {
                knowledgeBaseId: kbId,
                  modelArn: 'arn:aws:bedrock:us-east-1::foundation-model/anthropic.claude-v2'
              }
        }
    });

But this..

    const command = new RetrieveAndGenerateCommand({
        input: {
            text: input
        },
        retrieveAndGenerateConfiguration: {
            type: 'KNOWLEDGE_BASE',
            knowledgeBaseConfiguration: {
                knowledgeBaseId: kbId,
                   modelArn: 'arn:aws:bedrock:us-east-1::foundation-model/meta.llama2-70b-chat-v1'
              }
        }
    });

...throws the following error:

 model arn provided is not supported. Please check your configuration and retry the request.

I have access to both models.

Expected?

If so, are there plans to support other LLMs in this API?

2 Risposte
0

Currently Claude V1 and V2 are only supported for Knowledgebase. Other models are on the roadmap but they will need to support Chain of Thought using reasoning techniques like ReAct which not all models support. The following blog post provides some more background detail. Agents for Amazon Bedrock announcement

AWS
ESPERTO
con risposta 5 mesi fa
0

The Amazon Bedrock User Guide Supported regions and models page contains the most recent information. As of 6-MAR-2023

You can use the following models to generate responses after retrieving information from knowledge bases:

  • Anthropic Claude Instant v1
  • Anthropic Claude v2.0
  • Anthropic Claude v2.1
profile pictureAWS
con risposta 3 mesi fa

Accesso non effettuato. Accedi per postare una risposta.

Una buona risposta soddisfa chiaramente la domanda, fornisce un feedback costruttivo e incoraggia la crescita professionale del richiedente.

Linee guida per rispondere alle domande