Skip to content

v2.6.2

Latest
Latest

Choose a tag to compare

View all tags
@shubhadeepd shubhadeepd released this 20 Aug 08:22
v2.6.2
f20716d
This commit was created on GitHub.com and signed with GitHub’s verified signature.
GPG key ID: B5690EEEBB952194
Verified
Learn about vigilant mode.

This release updates NVIDIA-hosted cloud endpoint defaults for embedding and reranking documentation and configuration. It does not change the default self-hosted/on-prem model choices or update any RAG Blueprint application container images, Helm chart versions, or Python library package versions.

Highlights

This release includes the following key updates:

  • Default cloud embedding endpoint updated: NVIDIA-hosted cloud examples and environment overlays now use nvidia/llama-nemotron-embed-vl-1b-v2 with the API Catalog base URL https://integrate.api.nvidia.com/v1.
  • Default cloud reranker endpoint updated: NVIDIA-hosted cloud examples and environment overlays now use nvidia/llama-nemotron-rerank-vl-1b-v2 with an empty reranker server URL so the NVIDIA API Catalog client uses the default hosted reranking endpoint.
  • Updated Docker Compose environment overlays, variables.env, cloud deployment documentation, text-only ingest guidance, Python client examples, and notebook examples to align with the hosted embedding and reranker defaults.
  • Kept on-prem defaults unchanged: self-hosted embeddings continue to use the VLM embedding service, while the self-hosted reranker remains nvidia/llama-nemotron-rerank-1b-v2.
Assets 2
Loading