fastreid.engine.hooks.EvalHook Class Reference

Inheritance diagram for fastreid.engine.hooks.EvalHook:

Public Member Functions
	__init__ (self, eval_period, eval_function)
	after_step (self)
	after_train (self)
Public Member Functions inherited from fastreid.engine.train_loop.HookBase
	before_train (self)
	before_step (self)

Protected Member Functions
	_do_eval (self)

Protected Attributes
	_period
	_func

Detailed Description

Run an evaluation function periodically, and at the end of training.
It is executed every ``eval_period`` iterations and after the last iteration.

Definition at line 292 of file hooks.py.

Constructor & Destructor Documentation

__init__()

fastreid.engine.hooks.EvalHook.__init__	(		self,
			eval_period,
			eval_function )

Args:
    eval_period (int): the period to run `eval_function`.
    eval_function (callable): a function which takes no arguments, and
        returns a nested dict of evaluation metrics.
Note:
    This hook must be enabled in all or none workers.
    If you would like only certain workers to perform evaluation,
    give other workers a no-op function (`eval_function=lambda: None`).

Definition at line 298 of file hooks.py.

    def __init__(self, eval_period, eval_function):
        """
        Args:
            eval_period (int): the period to run `eval_function`.
            eval_function (callable): a function which takes no arguments, and
                returns a nested dict of evaluation metrics.
        Note:
            This hook must be enabled in all or none workers.
            If you would like only certain workers to perform evaluation,
            give other workers a no-op function (`eval_function=lambda: None`).
        """
        self._period = eval_period
        self._func = eval_function
 

Member Function Documentation

_do_eval()

fastreid.engine.hooks.EvalHook._do_eval ( self )

protected

Definition at line 312 of file hooks.py.

    def _do_eval(self):
        results = self._func()
 
        if results:
            assert isinstance(
                results, dict
            ), "Eval function must return a dict. Got {} instead.".format(results)
 
            flattened_results = flatten_results_dict(results)
            for k, v in flattened_results.items():
                try:
                    v = float(v)
                except Exception:
                    raise ValueError(
                        "[EvalHook] eval_function should return a nested dict of float. "
                        "Got '{}: {}' instead.".format(k, v)
                    )
            self.trainer.storage.put_scalars(**flattened_results, smoothing_hint=False)
 
        # Remove extra memory cache of main process due to evaluation
        torch.cuda.empty_cache()
 

after_step()

fastreid.engine.hooks.EvalHook.after_step

(

self

)

Called after each iteration.

Reimplemented from fastreid.engine.train_loop.HookBase.

Definition at line 334 of file hooks.py.

    def after_step(self):
        next_iter = self.trainer.iter + 1
        is_final = next_iter == self.trainer.max_iter
        if is_final or (self._period > 0 and next_iter % self._period == 0):
            self._do_eval()
        # Evaluation may take different time among workers.
        # A barrier make them start the next iteration together.
        comm.synchronize()
 

after_train()

fastreid.engine.hooks.EvalHook.after_train

(

self

)

Called after the last iteration.

Reimplemented from fastreid.engine.train_loop.HookBase.

Definition at line 343 of file hooks.py.

    def after_train(self):
        # func is likely a closure that holds reference to the trainer
        # therefore we clean it to avoid circular reference in the end
        del self._func
 
 

Member Data Documentation

_func

fastreid.engine.hooks.EvalHook._func

protected

Definition at line 310 of file hooks.py.

_period

fastreid.engine.hooks.EvalHook._period

protected

Definition at line 309 of file hooks.py.

---

The documentation for this class was generated from the following file:

• smreid/fastreid/engine/hooks.py