A GPipe implementation in PyTorch

An I/O benchmark for deep Learning applications

Sayiir: simple, embeddable durable workflow engine in Rust with Node.js & Python bindings. Checkpoint-based recovery, no deterministic replay. A simplified alternative to Temporal, Restate, and Airflow.

Very-Low Overhead Checkpointing System

CLI to the Cedana Service

Extending DOLFINx with checkpointing functionality

Keras wrapper that autosaves what ModelCheckpoint cannot.

Execution runtime for intelligent agents with event-sourced, recoverable task orchestration.

Git-like branching, checkpointing, and comparison for AI agent execution paths. pip install agentgit

Weavegraph rust graph/agent/node

An operating system for autonomous research — from literature to manuscript inside a governed, checkpointed loop.

Zero-cost, crash-proof LLM pipeline orchestrator. Features disk-based checkpointing, free-tier routing, and structured output. (LangGraph / CrewAI alternative)

A Python package for performing memory-intensive computations in parallel using chunks and checkpointing.

[WIP] Debug TypeScript/JavaScript via TUI. Checkpoint functions, edit state, skip execution. Written in Rust 🦀

A Python package for checkpointing, saving, and loading objects.

Checkpoint and rewind Claude Code runs with repo and context recovery.

This FLINK project will consume streams from an azure event-hub and produce to a different event-hub ,and the config files for deploying the same in kubernetes

Cairn — recoverable long-horizon AI agents. A framework-agnostic reference implementation and benchmark for agent checkpointing, crash recovery, and re-grounding after context loss. Thesis: Checkpoints Are Compactions.

Perkunas AI Training Platform is a memory-aware model training and serving system for serious language model experimentation under tight hardware limits. It combines streaming training, rich telemetry, guarded recovery, checkpoint export, and OpenAI-compatible serving.

A lightweight checkpointing program written in C.