Skip to main content

Reactor runtime with public model API

Project description

Reactor Runtime

A Python runtime for building real-time video processing models. This runtime abstracts all the techincal implementations of real-time networking, allowing researchers and models developers to run their model focusing only on the ML code.

You can think of this similarly to the way you write Telegram/Discord applications or bots using SDKs. You don't have to worry about the networking and the protocols of the medium. Instead, you can put all your effort in writing your application code, which in this case is ML code.

Documentation: https://docs.reactor.inc/runtime/overview

Installation

pip install reactor-runtime

Development Setup

Proto dependencies are fetched from the private reactor-team/reactor-proto repo. The version is defined in pyproject.toml under [tool.reactor-proto].

# Set your GitHub token (required for private repo access)
export GH_TOKEN=your_github_token

# Install with proto dependencies
make install

# Or fetch protos separately
make proto

The proto wheel is downloaded to generated/ (gitignored). Once installed, proto types are available via:

from api import reactor_pb2
from api.types import api_types_pb2, base_pb2

Publishing

make publish

The publish script performs two key transformations before building:

1. Obfuscation

All files and directories prefixed with _ are stripped from the published package (except __init__.py and __pycache__). This allows you to keep private/internal code in the repository that won't be shipped to users.

Example: A _cloud/ directory or _internal.py file will exist in the repo for development but won't appear in the PyPI package.

2. Proto Vendoring

The script uses AST analysis to scan all remaining (public) source files and identifies which api.* modules are actually imported. Only those specific proto files are copied into the published package.

This means:

  • Users don't need access to the private reactor-proto repo
  • The published package is self-contained with proto types bundled
  • Only the protos actually used by public code are included, keeping the package minimal

Prerequisite: Run make proto before publishing to ensure proto types are installed locally for vendoring.

Reactor CLI

Docs: https://docs.reactor.inc/runtime/cli-reference

Coding Models

Guide: https://docs.reactor.inc/runtime/coding-models

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

reactor_runtime-1.1.0.tar.gz (53.8 kB view details)

Uploaded Source

Built Distribution

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

reactor_runtime-1.1.0-py3-none-any.whl (66.2 kB view details)

Uploaded Python 3

File details

Details for the file reactor_runtime-1.1.0.tar.gz.

File metadata

  • Download URL: reactor_runtime-1.1.0.tar.gz
  • Upload date:
  • Size: 53.8 kB
  • Tags: Source
  • Uploaded using Trusted Publishing? No
  • Uploaded via: twine/6.2.0 CPython/3.10.12

File hashes

Hashes for reactor_runtime-1.1.0.tar.gz
Algorithm Hash digest
SHA256 6520d13ef3b3b531d8ca2d8b2403431fe1403cccf405d4962b70d178f321618f
MD5 ad33b4e5a97ca89769b88fe1b618eeb6
BLAKE2b-256 b15a1690e450c569473a2dee9d0d51a4c3fa9d96f88872b4970f80be4b9f1322

See more details on using hashes here.

File details

Details for the file reactor_runtime-1.1.0-py3-none-any.whl.

File metadata

File hashes

Hashes for reactor_runtime-1.1.0-py3-none-any.whl
Algorithm Hash digest
SHA256 b1deb443b87cd8cc0a32e3d27a38ee6e482270d684b5effecbbaad632adaafff
MD5 45b2f35d8c4a1012a90e22931558b8cf
BLAKE2b-256 d57e36b33bc84230d49c7235bf9a1e1dd95d39c4bee8ea01fc210c886708a598

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