# meta-pytorch/monarch

PyTorch Single Controller

Repository: https://github.com/meta-pytorch/monarch
Canonical: https://ross.abutalabs.com/products/monarch
Homepage: https://meta-pytorch.org/monarch
Language: Rust
License: BSD-3-Clause
License Family: permissive
Last push: 2026-09-03T01:06:48+00:00

## Health v2 (maintenance only)
Score: 80/100 (v2, computed 2026-09-03T02:20:16.233290+00:00)
- activity 100, release rhythm 81, longevity 35
- inputs: {"age_days": 491, "days_push": 0, "days_rel": 49, "gap_med": 47.5, "n_releases_24m": 7}
- flags: none
- formula: round(0.45*activity + 0.35*rhythm + 0.20*longevity); archived -> min(score, 10)

## Adoption (not part of the score)
Stars 1073, forks 172 (observed 2026-09-03T02:15:14.910689+00:00)

## What it is
Monarch is a distributed programming framework for PyTorch built on scalable actor messaging, with actors grouped into meshes, supervision-tree fault tolerance, RDMA transfers, and distributed tensors. It exposes a simple Python API for spawning processes and actors across GPUs while being implemented in Rust.

## Use cases
- orchestrate distributed PyTorch training across multiple GPUs
- build fault-tolerant distributed ML pipelines with actor supervision trees
- broadcast messages to collections of remote actors
- perform point-to-point RDMA transfers of GPU or CPU memory
- work with tensors sharded across processes
- spawn one trainer process per GPU from Python

## When to choose
- you need fine-grained control over distributed PyTorch process and actor orchestration
- you want fault tolerance with supervision trees and fine-grained recovery
- you need cheap RDMA-based GPU/CPU memory transfers
- you prefer an imperative Python API over declarative launchers

## When to avoid
- you only need single-node single-GPU training
- you want a mature turnkey distributed trainer like torchrun or DeepSpeed
- you need Windows support or non-Linux GPU environments

## Facets
- artifact type: framework
- maturity: active
- function: concurrency, rpc, machine-learning, llm-training, streaming
- domain: machine-learning, microservices, deep-learning, gpu-computing
- platform: python, rust
- tags: pytorch, actor-model, distributed-training, rdma, supervision-trees, actor-messaging, distributed-tensors, linux, macos, gpu

## Member repositories
- meta-pytorch/monarch (main) score 80

## Provenance
- Observed fields: from GitHub, fetched 2026-09-03T02:15:14.910689+00:00.
- Health v2: computed from the inputs above; adoption is never an input.
- Inferred fields (summary, facets, guidance): AI-extracted, prompt v1, taxonomy v1, on 2026-08-30T06:53:50.848243+00:00, confidence not recorded.
  - readme: https://github.com/meta-pytorch/monarch (fetched 2026-09-03T02:15:14.910689+00:00, sha c5070e62ed51)
  - homepage: https://meta-pytorch.org/monarch (fetched 2026-08-29T12:55:57.418225+00:00, sha cfcf97dafeee)
  - site_page: https://meta-pytorch.org/monarch/stable (fetched 2026-08-29T12:55:57.421851+00:00, sha e78cc68ca738)
- Data as of 2026-08-30T08:39:29.467469+00:00.