Hi All,
I know the Arc AI assistant is able to be set up with an embeddings server as of version 5. Do we have any guidance on doing so documented yet? Any basic steps outlined that we can take to set up the embeddings server/service correctly?
Thank you,
Nikita
Embeddings Model is used to query official TM1 documentation so the responses from LLM are more relevant.
There are 5 parameters related to setting it up. You can find the details in settings.sample file that Arc generates.
# AI Embeddings Server API Key - this is the API key for the AI Embeddings service. This is an optional parameter.
#aiembeddingsserviceapiKey: your-api-key
# AI Embeddings Service URL - this is the URL for the AI Embeddings service. This is an optional parameter.
#aiembeddingsserviceurl:
# Name of Large Language Model you want to use for AI Embeddings.
# For example: nomic-embed-text
# Note: if you change embeddings model, you need to delete content of docs directory within Arc
#aiembeddingsmodelname :
# AI Embeddings Service Provider - openai, ollama, huggingface, watsonx. Use openai for any openai compatible services.
#aiembeddingsserviceprovider:
# WatsonX project ID. This parameter is required when aiembeddingsserviceprovider is watsonx.
#aiembeddingswxprojectid: your-wx-project-id
If you run the embeddings model locally in ollama (it doesn’t require a lot of resources) you only need these two parameters: