Weave llms.txt

llms.txt

# Weights & Biases (W&B) Weave

> Weave is a framework for tracking, experimenting with, evaluating, deploying, and improving LLM-based applications. It provides comprehensive tools for tracing LLM calls, monitoring application behavior, systematic prompt engineering, evaluation, and deployment of guardrails in production.

Weave is designed to support the entire lifecycle of LLM application development, from initial experimentation to production deployment. Key capabilities include:

- Tracing and monitoring of LLM interactions and application logic
- Systematic iteration on prompts, datasets, and models
- Experimentation through an LLM Playground
- Comprehensive evaluation tools with custom and pre-built scorers
- Production guardrails for content moderation and prompt safety
- Cost tracking and optimization
- Support for multiple LLM providers and local models

## Core Documentation

- [Getting Started Guide](https://weave-docs.wandb.ai/quickstart): Quick introduction to tracing LLMs with Weave
- [Python SDK Reference](https://weave-docs.wandb.ai/reference/python-sdk/weave/): Complete Python API documentation
- [TypeScript SDK Reference](https://weave-docs.wandb.ai/reference/typescript-sdk/weave/): Complete TypeScript API documentation
- [Service API Reference](https://weave-docs.wandb.ai/reference/service-api/call-start-call-start-post): REST API documentation for service integration

## Features

- [LLM Application Tracing](https://weave-docs.wandb.ai/guides/tracking/): Comprehensive guide to tracking and analyzing LLM interactions
- [Evaluation Guide](https://weave-docs.wandb.ai/guides/evaluation/scorers): Documentation on evaluation capabilities and scorer implementation
- [Prompts Management](https://weave-docs.wandb.ai/guides/core-types/prompts): Guide to working with prompts in Weave
- [Models Integration](https://weave-docs.wandb.ai/guides/core-types/models): Documentation on supported models and integration
- [Datasets Handling](https://weave-docs.wandb.ai/guides/core-types/datasets): Guide to working with datasets in Weave

## Integrations

- [LLM Providers Guide](https://weave-docs.wandb.ai/guides/integrations/): Documentation on supported LLM providers and integration steps
- [Local Models Guide](https://weave-docs.wandb.ai/guides/integrations/local_models): Guide to using local models with Weave
- [Framework Integrations](https://weave-docs.wandb.ai/guides/integrations/): Documentation on supported frameworks and integration steps

## Enterprise Features

- [Platform & Security](https://weave-docs.wandb.ai/guides/platform/): Enterprise platform features and security documentation
- [Self-Managed Deployment](https://weave-docs.wandb.ai/guides/platform/weave-self-managed): Guide to deploying Weave in self-managed environments

## Optional

- [Cookbooks](https://weave-docs.wandb.ai/reference/gen_notebooks/intro_notebook): Example notebooks and tutorials
- [Environment Variables](https://weave-docs.wandb.ai/guides/core-types/env-vars): Configuration and environment setup
- [Troubleshooting Guide](https://weave-docs.wandb.ai/guides/troubleshooting): Common issues and solutions
- [Tools & Utilities](https://weave-docs.wandb.ai/guides/tools/): Additional tools and utilities documentation