openrunner-sdk 0.1.0

OpenRunner SDK - W&B-compatible ML experiment tracking client

OpenRunner SDK

Open-source, self-hosted ML experiment tracking — a drop-in replacement for Weights & Biases.

Install

pip install openrunner-sdk

Quick Start

export OPENRUNNER_API_KEY="or_your_key"
export OPENRUNNER_BASE_URL="https://your-server.com"

import openrunner

openrunner.init(project="my-project", config={"lr": 0.001})

for epoch in range(10):
    loss = train(epoch)
    openrunner.log({"loss": loss, "epoch": epoch})

openrunner.finish()

Migrating from W&B

import openrunner as wandb

wandb.init(project="my-project")
wandb.log({"loss": 0.5})
wandb.finish()

Features

• W&B-compatible API — init(), log(), finish(), config, summary
• Non-blocking — logging never slows down training
• Artifacts — version datasets, models, checkpoints with SHA-256 dedup
• Media — log images (Image()) and tables (Table())
• Offline mode — works without connectivity, sync later
• Framework integrations — PyTorch, HuggingFace, Lightning
• CLI — openrunner login, openrunner sync, openrunner ls
• System metrics — GPU/CPU/memory monitoring via psutil

License

MIT