A community member is integrating the Command R+ model from Azure Marketplace into their systems using LlamaIndex, but they are unsure if LlamaIndex fully supports the model's capabilities, particularly in handling separate document inputs for RAG. They ask if LlamaIndex supports the Command R+ model and if there are any plans to update the support. Another community member responds that the Cohere LLM integration is being updated to handle this, and a timeline of "hopefully today" is provided. A link to a related GitHub pull request is also shared.
I'm currently integrating the Command R+ model from Azure Marketplace into our systems using LlamaIndex. I understand that there is some support for this model in LlamaIndex, but it might not be fully optimized to leverage all its capabilities. Specifically, the original Cohere client handles the "documents" field by adding chunks separately, which allows for internal optimizations by the model. However, it seems like in LlamaIndex, context needs to be added directly into the prompt.
Could someone confirm if LlamaIndex fully supports the Command R+ model, especially in terms of handling separate document inputs for RAG? If not, are there any plans to update this support or should I consider custom implementations to fully utilize the model's capabilities?
Example from the Cohere docs: co.chat(
model="command",
message="Where do the tallest penguins live?",
documents=[
{"title": "Tall penguins", "snippet": "Emperor penguins are the tallest."},
{"title": "Penguin habitats", "snippet": "Emperor penguins only live in Antarctica."},
{"title": "What are animals?", "snippet": "Animals are different from plants."}
])
Thanks for the update, @Logan M ! Do you have an estimated timeline for when the updated integration might be released? I was considering implementing a custom solution, but if it's just around the corner, I might wait for the official update. 😊
