🧰 ezpz CLI⚓︎

Once installed, ezpz provides a CLI with a few useful utilities to help launch distributed PyTorch applications.

Explicitly, these are ezpz <command>:

• 🚀 ezpz launch: Launch commands with automatic job scheduler detection (PBS, Slurm)
  • 💯 ezpz test: Run simple distributed smoke test¹.
  • 📊 ezpz benchmark: Run all examples and generate a report
  • 📮 ezpz submit: Submit jobs to PBS (qsub) or SLURM (sbatch); generates job scripts automatically
• 📦 ezpz yeet: Distribute files (envs, models, datasets, etc.) to all worker nodes via parallel rsync
  • 🗜️ ezpz tar-env: Package current Python environment as a tarball
• 🩺 ezpz doctor: Health check your environment
• 💀 ezpz kill: Kill ezpz-launched python processes (local node or --all-nodes)

• 📝 ezpz.examples: Collection of distributed training examples (DDP, FSDP, ViT, FSDP+TP, diffusion, HF, HF Trainer, inference)

• ezpz --help

  To see the list of available commands, run:

  $ ezpz --help
  Usage: ezpz [OPTIONS] COMMAND [ARGS]...
  
  ezpz distributed utilities.
  
  Options:
  --version   Show the version and exit.
  -h, --help  Show this message and exit.
  
  Commands:
  benchmark  Run all ezpz examples sequentially and generate a report.
  doctor     Inspect the environment for ezpz launch readiness.
  kill       Kill ezpz-launched python processes (or any matching pattern).
  launch     Launch a command across the active scheduler.
  submit     Submit a job to the active scheduler (PBS/SLURM).
  tar-env    Create (or locate) a tarball for the current environment.
  test       Run the distributed smoke test.
  yeet       Distribute files (envs, models, datasets, etc.) to worker nodes.