Skip to main content
Weights & Biases (W&B) Weave is a framework for tracking, experimenting with, evaluating, deploying, and improving LLM-based applications. Designed for flexibility and scalability, Weave supports every stage of your LLM application development workflow.
The SambaNova Weave integration setup is detailed in the weave integration repository.

Prerequisites

Setup

1

Install required packages

2

Set your API keys

SAMBANOVA_URL is required by the notebook’s OpenAI SDK connector. Without it, the OpenAI client falls back to its own default endpoint and your SambaNova key is rejected.
3

Download and run the example notebook

Download the Weave@SambaNova notebook and launch it with Jupyter:
The notebook demonstrates three connectors:
  • SambaNova LangChain chat object
  • LiteLLM
  • OpenAI SDK

Troubleshooting

  • Verify the key is exported: echo "${SAMBANOVA_API_KEY:0:10}..."
  • Regenerate your key in the SambaCloud portal
  • Re-export and restart your terminal or notebook kernel
  • Verify the key is exported: echo "${WANDB_API_KEY:0:10}..."
  • Retrieve your key from the W&B settings page
  • Re-export and restart your terminal or notebook kernel
  • Install Jupyter: pip install jupyter
  • Or run the notebook with: pip install jupyterlab && jupyter lab