:orphan:

Examples
================

- :doc:`ping_pong.py </generated/examples/ping_pong>`: Demonstrates the basics of Monarch's Actor/endpoint API with a ping-pong communication example
- :doc:`crawler.py </generated/examples/crawler>`: Demonstrates Monarch's actor API and many-to-one communications with a web crawler example
- :doc:`spmd_ddp.py </generated/examples/ddp/spmd_ddp>`: Shows how to run PyTorch's Distributed Data Parallel (DDP) using SPMDActor
- :doc:`Interactive SPMD Job </generated/examples/ddp/spmd_job>`: Shows how to use ``serve()`` and ``run_spmd()`` for interactive SPMD training with job caching and debugging
- :doc:`kubernetes_ddp.py </generated/examples/ddp/kubernetes_ddp>`: Extends the DDP example to run on Kubernetes using MonarchMesh CRD and operator
- :doc:`grpo_actor.py </generated/examples/grpo_actor>`: Implements a distributed PPO-like reinforcement learning algorithm using the Monarch actor framework
- :doc:`grpo_yield.py </generated/examples/grpo_yield>`: Same GRPO task as ``grpo_actor.py``, wired with direct port messaging instead of queue actors and ``RDMABuffer``; each actor's ``run`` endpoint is structured like a Python generator that yields values to, and receives values from, the next actor in the ring
- :doc:`kubernetes_grpo.py </generated/examples/grpo/kubernetes_grpo>`: Extends the GRPO example to run on Kubernetes using MonarchMesh CRDs, fine-tuning the open-source Qwen3.5-0.8B model on the GSM8K math dataset
- :doc:`distributed_tensors.py </generated/examples/distributed_tensors>`: Shows how to dispatch tensors and tensor level operations to a distributed mesh of workers and GPUs
- :doc:`debugging.py </generated/examples/debugging>`: Shows how to use the Monarch debugger to debug a distributed program
- :doc:`otel_collector.py </generated/examples/otel_collector>`: Exports Monarch metrics and logs to an OpenTelemetry Collector deployed on Kubernetes, with Grafana for visualization
- `Multinode Slurm Tutorial <https://docs.pytorch.org/tutorials/intermediate/monarch_distributed_tutorial.html>`_: Multinode distributed training tutorial using Monarch and Slurm to run an SPMD training job.
- `Running on Kubernetes using Skypilot <https://github.com/meta-pytorch/monarch/tree/main/examples/skypilot>`_: Run Monarch on Kubernetes and cloud VMs via SkyPilot.

.. toctree::
   :hidden:

   /generated/examples/ping_pong
   /generated/examples/crawler
   /generated/examples/ddp/spmd_ddp
   /generated/examples/ddp/spmd_job
   /generated/examples/ddp/kubernetes_ddp
   /generated/examples/grpo_actor
   /generated/examples/grpo_yield
   /generated/examples/grpo/kubernetes_grpo
   /generated/examples/distributed_tensors
   /generated/examples/debugging
   /generated/examples/otel_collector
   Multinode Slurm Tutorial <https://docs.pytorch.org/tutorials/intermediate/monarch_distributed_tutorial.html>