Not able to enable the feature rag_pipeline_feature_enabled in ml plugin

Plugins Machine Learning
1

Hi All,

I am not able to enable the RAG feature. I have run this below API successfully

PUT /_cluster/settings
{
  "persistent": {"plugins.ml_commons.rag_pipeline_feature_enabled": "true"}
}

After deploying my remote LLM model successfully ( I am using hugging face model inference endpoints). Which is working well with _predict API and giving successful response

I am getting below error

{
    "error": {
        "root_cause": [
            {
                "type": "m_l_exception",
                "reason": "plugins.ml_commons.rag_pipeline_feature_enabled is not enabled."
            }
        ],
        "type": "m_l_exception",
        "reason": "plugins.ml_commons.rag_pipeline_feature_enabled is not enabled."
    },
    "status": 500
}

Below is my request

PUT /_search/pipeline/rag_pipeline

{
  "response_processors": [
    {
      "retrieval_augmented_generation": {
        "tag": "movie bot",
        "description": "Demo pipeline Using HF",
        "model_id": "Pbh0p4wB9_MXHuOCuY",
        "context_field_list": ["text"],
        "system_prompt": "You are a helpful assistant",
        "user_instructions": "Generate a concise and informative answer in less than 100 words for the given question"
      }
    }
  ]
}

Thanks

Ram Da

2

What version of OpenSearch are you using and how did you install it? E.g. Docker on your laptop?

When you ran the PUT command, did you get a response showing “acknowledged: true”? Can you run /_cluster/settings to confirm the feature flag is in fact set to “true”?

You can also try setting it to “false” and then setting it back to “true”.

2 Likes
3

This worked for me , by setting it to “false” and back to “true”

4

This topic was automatically closed 60 days after the last reply. New replies are no longer allowed.