Skip to main content

Multi-agent system for HED annotation generation and validation

Project description

HEDit

PyPI version Tests

Convert natural language event descriptions into valid HED (Hierarchical Event Descriptors) annotations.

Part of the Annotation Garden Initiative.

Installation

# Default (lightweight API client, ~100MB)
pip install hedit

# Standalone mode (run locally without backend, ~2GB)
pip install hedit[standalone]

Quick Start

# Configure your OpenRouter API key (https://openrouter.ai)
hedit init --api-key sk-or-v1-xxx

# Generate HED annotation from text
hedit annotate "participant pressed the left button"

# Generate HED from an image
hedit annotate-image stimulus.png

# Validate a HED string
hedit validate "Sensory-event, Visual-presentation"

Commands

Command Description
hedit init Configure API key and preferences
hedit annotate "text" Convert natural language to HED
hedit annotate-image <file> Generate HED from image
hedit validate "HED-string" Validate HED annotation
hedit health Check service status
hedit config show Display configuration

Options

hedit annotate "text" -o json          # JSON output for scripting
hedit annotate "text" --schema 8.3.0   # Specific HED schema version
hedit annotate "text" --standalone     # Run locally (requires hedit[standalone])

How It Works

HEDit uses a multi-agent system (LangGraph) with feedback loops:

  1. Annotation Agent - Generates initial HED tags
  2. Validation Agent - Checks syntax and tag validity
  3. Evaluation Agent - Assesses faithfulness to input
  4. Assessment Agent - Identifies missing elements

Annotations are automatically refined until validation passes.

Links

License

MIT

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

hedit-0.7.5.dev1.tar.gz (157.0 kB view details)

Uploaded Source

Built Distribution

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

hedit-0.7.5.dev1-py3-none-any.whl (124.8 kB view details)

Uploaded Python 3

File details

Details for the file hedit-0.7.5.dev1.tar.gz.

File metadata

  • Download URL: hedit-0.7.5.dev1.tar.gz
  • Upload date:
  • Size: 157.0 kB
  • Tags: Source
  • Uploaded using Trusted Publishing? Yes
  • Uploaded via: twine/6.1.0 CPython/3.13.7

File hashes

Hashes for hedit-0.7.5.dev1.tar.gz
Algorithm Hash digest
SHA256 49a927b3441087cbafd50d6e531e8bf1f8201391519edc78b4e3d06a9e8ed81d
MD5 0d6994a5132b21d5824de6837d19176a
BLAKE2b-256 71ea3b291f979b6690d34dc3db19f1e868d3d5aae57e0cbb06c1bca528ea90b8

See more details on using hashes here.

Provenance

The following attestation bundles were made for hedit-0.7.5.dev1.tar.gz:

Publisher: publish.yml on Annotation-Garden/HEDit

Attestations: Values shown here reflect the state when the release was signed and may no longer be current.

File details

Details for the file hedit-0.7.5.dev1-py3-none-any.whl.

File metadata

  • Download URL: hedit-0.7.5.dev1-py3-none-any.whl
  • Upload date:
  • Size: 124.8 kB
  • Tags: Python 3
  • Uploaded using Trusted Publishing? Yes
  • Uploaded via: twine/6.1.0 CPython/3.13.7

File hashes

Hashes for hedit-0.7.5.dev1-py3-none-any.whl
Algorithm Hash digest
SHA256 88be5f2a6dfc48dbd2782b749b412cb06539ddc30aa10b652bf9e3096e27c3f4
MD5 8dff78164dcdd8e212a909812ae6c36c
BLAKE2b-256 584a5ea9b3304619ca71af6c1e4a5ff10d42268b302765e7d584ef6827a86ab1

See more details on using hashes here.

Provenance

The following attestation bundles were made for hedit-0.7.5.dev1-py3-none-any.whl:

Publisher: publish.yml on Annotation-Garden/HEDit

Attestations: Values shown here reflect the state when the release was signed and may no longer be current.

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