🧰 `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 test
🩺 ezpz doctor: Health check your environment
📝 ezpz.examples: Collection of distributed training examples
- Distributed Training Examples
  
  See the Examples page for full details.
  - test: Simplest DDP training loop
    ezpz launch python3 -m ezpz.examples.test
  - fsdp: FSDP for memory-efficient training
    ezpz launch python3 -m ezpz.examples.fsdp
  - vit: Vision Transformer with FSDP + optional torch.compile
    ezpz launch python3 -m ezpz.examples.vit
  - fsdp_tp: 2D parallelism (FSDP + Tensor Parallel)
    ezpz launch python3 -m ezpz.examples.fsdp_tp
  - diffusion: Diffusion model training with FSDP
    ezpz launch python3 -m ezpz.examples.diffusion
  - hf: Fine-tune causal LM with explicit training loop (Accelerate + FSDP)
    ezpz launch python3 -m ezpz.examples.hf
  - hf_trainer: Hugging Face Trainer integration
    ezpz launch python3 -m ezpz.examples.hf_trainer
Experimental
- 📦 ezpz tar-env: Package current Python environment as a tarball
- 🚀 ezpz yeet-env: Broadcast environment tarball to all worker nodes via MPI

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:
doctor    Inspect the environment for ezpz launch readiness.
launch    Launch a command across the active scheduler.
test      Run the distributed smoke test.

🧰 ezpz CLI⚓︎

🧰 `ezpz` CLI⚓︎