Skip to content

PrimeIntellect-ai/prime-iroh

Repository files navigation

Prime Intellect

PRIME-IROH: P2P Pipeline Parallel Communication


This library exposes a Python interface for reliable, asynchronous peer-to-peer communication built upon Iroh. The core classes exposed are:

  • Node: A class combining a single-peer sender/ receiver in one class, allowing to send to exactly one and receive from exactly one (potentially different) peer. The class allows for concurrent communication by opening multiple, consistent streams.
  • SendWork: A class representing the future of an asynchronous send operation, that can be awaited using a wait method.
  • RecvWork: A class representing the future of an asynchronous receive operation, that can be awaited using a wait method.

Because we are building on top of Iroh, we get many nice networking features out of the box. Most importantly, the library guarantees reliable P2P connections between nodes, trying to establish directions connections whenever possible, and falling back to NAT-hole punching and relaying when necessary. The API is mirroring the way asynchronous communication is handled in torch.distributed, i.e. exposing isend and irecv that return work objects that can be awaited using a wait method. This allows for a clean integration with the rest of the PyTorch ecosystem.

Installation

Quick Install: Run the following command for a quick install:

curl -sSL https://raw.githubusercontent.com/PrimeIntellect-ai/prime-iroh/main/install.sh | bash

Manual Install: First, install uv and cargo to build the project.

curl -LsSf https://astral.sh/uv/install.sh | sh
source $HOME/.local/bin/env
curl https://sh.rustup.rs -sSf | sh
source $HOME/.cargo/env

Then, clone the repository

git clone git@github.com:PrimeIntellect-ai/prime-iroh.git && cd prime-iroh

To build the Rust backend run cargo build, to build the Python bindings run uv sync. This will let you install prime-iroh as a Python package within the virtual environment.

uv run python -c "import prime_iroh"

Examples

You can find the basic usage examples in the rust/examples and python/examples directories showing unidirectional and bidirectional communication patterns in Rust and Python.

Run unidirectional communication example:

# Rust
cargo run --example unidirectional
# Python
uv run python python/examples/unidirectional.py

Run bidirectional communication example:

# Rust
cargo run --example bidirectional
# Python
uv run python python/examples/bidirectional.py

You can set the log level by setting the RUST_LOG environment variable. For example, to see info logs from the prime-iroh crate, set RUST_LOG=prime_iroh=info.

Tests

We include unit tests and integration tests for Rust and Python.

Rust Tests

Run full test suite (unit and integration tests):

cargo test

Note: To run the tests with verbose output, use cargo test -- --nocapture.

Run single unit test, e.g. tests in src/node.rs (this will pattern match the test function names):

cargo test node

Run single integration test, e.g. tests/connection.rs:

cargo test --test connection

Python Tests

Run full test suite (only integration tests are available for now):

uv run pytest

To run the tests with verbose output, use uv run pytest -s.

Run single test, e.g. python/tests/test_unidirectional.py:

uv run pytest python/tests/test_unidirectional.py

About

Asynchronous P2P communication backend for decentralized pipeline parallelism

Resources

License

Stars

Watchers

Forks

Packages

No packages published

Contributors 3

  •  
  •  
  •