[FEATURE] Support for Reciprocal Rank Fusion Retriever with Azure OpenAI's GPT-3.5 Turbo Instruct in Flowise

[seonghobae]

Describe the feature you'd like I am requesting support for implementing Reciprocal Rank Fusion Retriever using Azure OpenAI's GPT-3.5 Turbo Instruct model within Flowise. The current Azure OpenAI Node in Flowise does not list the GPT-3.5 Turbo Instruct model, which is essential for advanced retrieval-based tasks and has replaced the text-davinci-003 model.

Additional context The GPT-3.5 Turbo Instruct model is specifically designed to efficiently handle instruction-based prompts and deliver concise responses, which are key capabilities for the Reciprocal Rank Fusion Retriever methodology. Enabling Chat completion with this model would greatly improve the information retrieval process within Flowise, allowing users to develop more complex and effective AI-powered solutions.

Updating Flowise to include Azure OpenAI's GPT-3.5 Turbo Instruct model is critical to ensure that users can leverage the most up-to-date AI functionalities. This update would not only keep Flowise in line with Azure's evolving AI services but also expand the platform's capabilities in terms of retrieval accuracy and performance.

I believe that adding this support is crucial for users looking to utilize state-of-the-art AI models for sophisticated information retrieval and would appreciate your consideration in making this feature available.

Thank you for evaluating this feature request.

[HenryHengZJ]

its already available here - https://github.com/FlowiseAI/Flowise/tree/main/packages/components/nodes/retrievers/RRFRetriever

[seonghobae]

I know, however that's seems not work correctly with Azure Chat models.

[seonghobae]

I checked that the reciprocal rank fusion retriever runs correctly on Azure LLM with GPT3-instruct, not Azure Chat Modules with any other model.