# flopscope

> NumPy-compatible math primitives with FLOP counting

## Getting Started

- [Install flopscope and verify setup](https://aicrowd.github.io/flopscope/docs/getting-started/installation)
- [Run your first FLOP-counted computation](https://aicrowd.github.io/flopscope/docs/getting-started/quickstart)
- [Everything you need to compete within a FLOP budget](https://aicrowd.github.io/flopscope/docs/getting-started/competition)

## Guides

- [Migrate existing NumPy code to flopscope](https://aicrowd.github.io/flopscope/docs/guides/migrate-from-numpy)
- [Einsum patterns, costs, and optimization](https://aicrowd.github.io/flopscope/docs/guides/einsum)
- [Reduce FLOP costs with symmetric tensors](https://aicrowd.github.io/flopscope/docs/guides/symmetry)
- [Linear algebra operations and costs](https://aicrowd.github.io/flopscope/docs/guides/linalg)
- [FFT operations, real vs complex, and costs](https://aicrowd.github.io/flopscope/docs/guides/fft)
- [Estimate costs and diagnose budget overruns](https://aicrowd.github.io/flopscope/docs/guides/budget-planning)

## Understanding flopscope

- [How Flopscope wraps NumPy to count every FLOP](https://aicrowd.github.io/flopscope/docs/understanding/how-flopscope-works)
- [Analytical FLOP cost formulas and conventions](https://aicrowd.github.io/flopscope/docs/understanding/flop-counting-model)
- [Free vs counted vs blocked operations](https://aicrowd.github.io/flopscope/docs/understanding/operation-categories)
- [Subgraph symmetry detection algorithm deep dive](https://aicrowd.github.io/flopscope/docs/understanding/symmetry-detection)

## API Reference

- [Interactive searchable reference for all 508 operations](https://aicrowd.github.io/flopscope/docs/api)
- [Machine-readable resources and rules for AI coding assistants](https://aicrowd.github.io/flopscope/docs/api/for-agents)

## Infrastructure

- [Client-server architecture for competition evaluation](https://aicrowd.github.io/flopscope/docs/infrastructure/client-server)
- [Running flopscope with Docker](https://aicrowd.github.io/flopscope/docs/infrastructure/docker)

## Development

- [Repository layout, local workflows, and test suite](https://aicrowd.github.io/flopscope/docs/development/contributing)
- [Empirical weight calibration methodology](https://aicrowd.github.io/flopscope/docs/development/calibration)

## Changelog

- [Release history and breaking changes](https://aicrowd.github.io/flopscope/docs/changelog)