Skip to main content

Anchor every W&B artifact to Bitcoin. Async, with reconciliation ledger.

Project description

umarise-wandb

Anchor W&B artifacts to Bitcoin. Zero-touch provenance for ML experiments.

pip install umarise-wandb

Quick start

import umarise_wandb

# Enable once. Every artifact is anchored automatically.
umarise_wandb.enable()

That's it. Every wandb.log_artifact() call now anchors the artifact to Bitcoin in the background.

Explicit anchoring

import wandb
import umarise_wandb

with wandb.init() as run:
    artifact = wandb.Artifact("my-model", type="model")
    artifact.add_file("model.pt")
    run.log_artifact(artifact)

    # Anchor explicitly
    results = umarise_wandb.anchor_artifact(artifact)
    # results[0]["origin_id"] -> UUID

Anchor a previously logged artifact

results = umarise_wandb.anchor_logged_artifact("my-model:v3")

What happens

  1. File is hashed locally (SHA-256). Bytes never leave your machine.
  2. Hash is submitted to Umarise Core API.
  3. Umarise anchors to Bitcoin via OpenTimestamps.
  4. origin_id is logged to your W&B run summary.
  5. Within ~2 hours, proof is confirmed on the Bitcoin blockchain.

W&B metadata written

Key Location Value
umarise/origin_id Run log UUID
umarise/anchored Run log true
umarise_anchored Run summary true
umarise_last_origin_id Run summary UUID

Environment

Set UMARISE_API_KEY or pass api_key= to any function.

export UMARISE_API_KEY=um_your_key

Design principle

umarise-wandb never breaks your training pipeline. If anchoring fails (network, API), it silently continues. Your experiment always completes.

Verify independently

The proof is yours. Verify without Umarise:

Links

License

Unlicense - Public Domain

Project details


Download files

Download the file for your platform. If you're not sure which to choose, learn more about installing packages.

Source Distribution

umarise_wandb-0.2.0.tar.gz (5.3 kB view details)

Uploaded Source

Built Distribution

If you're not sure about the file name format, learn more about wheel file names.

umarise_wandb-0.2.0-py3-none-any.whl (6.5 kB view details)

Uploaded Python 3

File details

Details for the file umarise_wandb-0.2.0.tar.gz.

File metadata

  • Download URL: umarise_wandb-0.2.0.tar.gz
  • Upload date:
  • Size: 5.3 kB
  • Tags: Source
  • Uploaded using Trusted Publishing? No
  • Uploaded via: twine/6.2.0 CPython/3.12.8

File hashes

Hashes for umarise_wandb-0.2.0.tar.gz
Algorithm Hash digest
SHA256 75a45eb325aeeef453b1b4e8e074fd97b1196de3bf34263b8a2937b7607c78c7
MD5 4aa3f0837cc06e14f8be48a2f368fdb8
BLAKE2b-256 e3201aec1e3c7279cabe00a5bb32bd8af8d1b7d564e72ceba3a58b14859fc87f

See more details on using hashes here.

File details

Details for the file umarise_wandb-0.2.0-py3-none-any.whl.

File metadata

  • Download URL: umarise_wandb-0.2.0-py3-none-any.whl
  • Upload date:
  • Size: 6.5 kB
  • Tags: Python 3
  • Uploaded using Trusted Publishing? No
  • Uploaded via: twine/6.2.0 CPython/3.12.8

File hashes

Hashes for umarise_wandb-0.2.0-py3-none-any.whl
Algorithm Hash digest
SHA256 afb37a36bb27a4f2f85c3c2c231db1fb541757443d3079c9baf292406e968041
MD5 c9ea9ec50ad5f0d73ad64eaa049015b1
BLAKE2b-256 451c47f38b68ef28c176a98729f04546608a26757d0f56cc605d71b090b48220

See more details on using hashes here.

Supported by

AWS Cloud computing and Security Sponsor Datadog Monitoring Depot Continuous Integration Fastly CDN Google Download Analytics Pingdom Monitoring Sentry Error logging StatusPage Status page