Lightning-AI/pytorch-lightning · error · ValueError

`num_nodes` must be a positive integer, but got {num_nodes}.

Error message

`num_nodes` must be a positive integer, but got {num_nodes}.

What it means

Trainer validates that num_nodes is a positive integer (>=1). Anything else — zero, negatives, floats, strings — raises ValueError during connector init. This is plain input validation before any distributed setup.

Source

Thrown at src/lightning/pytorch/trainer/connectors/accelerator_connector.py:314

                if self._strategy_flag.parallel_devices[0].type == "cpu":
                    if self._accelerator_flag and self._accelerator_flag not in ("auto", "cpu"):
                        raise MisconfigurationException(
                            f"CPU parallel_devices set through {self._strategy_flag.__class__.__name__} class,"
                            f" but accelerator set to {self._accelerator_flag}, please choose one device type"
                        )
                    self._accelerator_flag = "cpu"
                if self._strategy_flag.parallel_devices[0].type == "cuda":
                    if self._accelerator_flag and self._accelerator_flag not in ("auto", "cuda", "gpu"):
                        raise MisconfigurationException(
                            f"GPU parallel_devices set through {self._strategy_flag.__class__.__name__} class,"
                            f" but accelerator set to {self._accelerator_flag}, please choose one device type"
                        )
                    self._accelerator_flag = "cuda"
                self._parallel_devices = self._strategy_flag.parallel_devices

    def _check_device_config_and_set_final_flags(self, devices: Union[list[int], str, int], num_nodes: int) -> None:
        if not isinstance(num_nodes, int) or num_nodes < 1:
            raise ValueError(f"`num_nodes` must be a positive integer, but got {num_nodes}.")

        self._num_nodes_flag = num_nodes
        self._devices_flag = devices

        if self._devices_flag in ([], 0, "0"):
            accelerator_name = (
                self._accelerator_flag.__class__.__qualname__
                if isinstance(self._accelerator_flag, Accelerator)
                else self._accelerator_flag
            )
            raise MisconfigurationException(
                f"`Trainer(devices={self._devices_flag!r})` value is not a valid input"
                f" using {accelerator_name} accelerator."
            )

    @staticmethod
    def _choose_auto_accelerator() -> str:
        """Choose the accelerator type (str) based on availability."""

View on GitHub (pinned to 9fed5c27d2)

Solutions

1. Coerce and validate: int(num_nodes) with a >=1 guard before constructing Trainer
2. Default to 1 on single-node runs instead of computing 0

Example fix

# before
trainer = Trainer(num_nodes=int(os.environ.get("WORLD_SIZE", 0)))
# after
num_nodes = max(1, int(os.environ.get("WORLD_SIZE", 1)))
trainer = Trainer(num_nodes=num_nodes)

num_nodes = int(num_nodes)
if num_nodes < 1:
    raise ValueError(f"num_nodes must be >= 1, got {num_nodes}")
trainer = Trainer(num_nodes=num_nodes)

def valid_num_nodes(n) -> bool:
    return isinstance(n, int) and not isinstance(n, bool) and n >= 1

When it happens

Trigger: Trainer(num_nodes=0), Trainer(num_nodes=-1), Trainer(num_nodes=2.0), or num_nodes sourced from an unparsed env var/config string like '2'.

Common situations: num_nodes computed from SLURM/环境 variables as strings or floats; arithmetic that can yield 0 on single-node runs.

Understand the failure class

Background: "Must be a positive integer", "Invalid value", "Unsupported": the invalid-argument-value error family, when a library rejects the value you pass — this error's family across 35 libraries.

Related errors

• Device should be CPU, got {device} instead.
• `devices` selected with `CPUAccelerator` should be an int >
• Device should be CUDA, got {device} instead.
• You requested to find {num_devices} devices but there are no
• You requested to find {num_devices} devices but this machine

AI-assisted analysis of Lightning-AI/pytorch-lightning@9fed5c27d2 (2026-08-28). Data as JSON: /api/errors/28fd5e77877fe18c. Report an issue: GitHub.