Microsoft Azure - Weights & Biases Documentation

Weights & Biases (W&B) Weave integrates with Microsoft Azure OpenAI services, helping teams optimize their Azure AI applications. This page shows you how to start tracking calls to Azure OpenAI with Weave and points to additional resources for more advanced workflows.

For the latest tutorials, visit Weights & Biases on Microsoft Azure.

Get started

To start tracking Azure OpenAI calls with Weave, decorate the functions you want to track with weave.op. Once decorated, Weave automatically captures inputs, outputs, and metadata for each call to the function.

@weave.op()
def call_azure_chat(model_id: str, messages: list, max_tokens: int = 1000, temperature: float = 0.5):
    response = client.chat.completions.create(
        model=model_id,
        messages=messages,
        max_tokens=max_tokens,
        temperature=temperature
    )
    return {"status": "success", "response": response.choices[0].message.content}