Lightning-AI/pytorch-lightning · error · MisconfigurationException

Received multiple values for {', '.join(duplicated_plugin_ke

Error message

Received multiple values for {', '.join(duplicated_plugin_key)} flags in `plugins`. Expected one value for each type at most.

What it means

Each plugin type may appear at most once in Trainer(plugins=[...]). The connector counts instances per type and raises when any type key has more than one entry. This prevents ambiguity about which plugin instance should win.

Source

Thrown at src/lightning/pytorch/trainer/connectors/accelerator_connector.py:254

                    self._cluster_environment_flag = plugin
                    plugins_flags_types[ClusterEnvironment.__name__] += 1
                elif isinstance(plugin, LayerSync):
                    if sync_batchnorm and not isinstance(plugin, TorchSyncBatchNorm):
                        raise MisconfigurationException(
                            f"You set `Trainer(sync_batchnorm=True)` and provided a `{plugin.__class__.__name__}`"
                            " plugin, but this is not allowed. Choose one or the other."
                        )
                    self._layer_sync = plugin
                    plugins_flags_types[TorchSyncBatchNorm.__name__] += 1
                else:
                    raise MisconfigurationException(
                        f"Found invalid type for plugin {plugin}. Expected one of: Precision, "
                        "CheckpointIO, ClusterEnvironment, or LayerSync."
                    )

            duplicated_plugin_key = [k for k, v in plugins_flags_types.items() if v > 1]
            if duplicated_plugin_key:
                raise MisconfigurationException(
                    f"Received multiple values for {', '.join(duplicated_plugin_key)} flags in `plugins`."
                    " Expected one value for each type at most."
                )

            if plugins_flags_types.get(Precision.__name__) and precision_flag is not None:
                raise ValueError(
                    f"Received both `precision={precision_flag}` and `plugins={self._precision_plugin_flag}`."
                    f" Choose one."
                )

        self._precision_flag = "32-true" if precision_flag is None else precision_flag

        # handle the case when the user passes in a strategy instance which has an accelerator, precision,
        # checkpoint io or cluster env set up
        # TODO: improve the error messages below
        if self._strategy_flag and isinstance(self._strategy_flag, Strategy):
            if self._strategy_flag._accelerator:
                if self._accelerator_flag != "auto":

View on GitHub (pinned to 9fed5c27d2)

Solutions

1. Inspect the plugins list and keep exactly one instance of the duplicated type
2. If you need combined behavior, wrap the two plugins in a single composite/custom plugin subclassing the same base

Example fix

# before
trainer = Trainer(plugins=[Precision16(), MixedPrecision('bf16-mixed')])
# after
trainer = Trainer(plugins=[MixedPrecision('bf16-mixed')])

from collections import Counter
from lightning.pytorch.plugins import Precision, CheckpointIO, ClusterEnvironment, LayerSync
_KEYS = {Precision: "Precision", CheckpointIO: "CheckpointIO", ClusterEnvironment: "ClusterEnvironment", LayerSync: "LayerSync"}
counts = Counter(_KEYS[type(p)] for p in plugins)
assert all(v == 1 for v in counts.values()), f"duplicate plugins: {counts}"

def has_duplicate_plugin_types(plugins: list) -> bool:
    seen = set()
    for p in plugins:
        k = type(p).__name__
        if k in seen:
            return True
        seen.add(k)
    return False

When it happens

Trigger: Trainer(plugins=[PrecisionPluginA(), PrecisionPluginB()]) or two CheckpointIO/ClusterEnvironment/LayerSync instances in the same plugins list.

Common situations: Combining shared code snippets or copy-pasted configs that each add their own precision plugin; passing both a LayersSync/TorchSyncBatchNorm and enabling sync_batchnorm style setups multiple times.

Related errors

• You set `Trainer(sync_batchnorm=True)` and provided a `{plug
• Found invalid type for plugin {plugin}. Expected one of: Pre
• Device should be CPU, got {device} instead.
• `devices` selected with `CPUAccelerator` should be an int >
• Device should be CUDA, got {device} instead.

AI-assisted analysis of Lightning-AI/pytorch-lightning@9fed5c27d2 (2026-08-28). Data as JSON: /api/errors/2e7a72a6c2b4ee8c. Report an issue: GitHub.