Uninformative error message when using sim.compile on a network with no probed outputs

[xchoo]

If you use the sim.compile method on a NengoDL model with no probed outputs, the following error message is thrown:

Traceback (most recent call last):
  File ".\test_dl_probe_out.py", line 15, in <module>
    sim.compile(
  File "d:\xchoo\git\nengo\nengo\utils\magic.py", line 181, in __call__
    return self.wrapper(self.__wrapped__, self.instance, args, kwargs)
  File "d:\xchoo\git\nengo-dl\nengo_dl\simulator.py", line 67, in require_open
    return wrapped(*args, **kwargs)
  File "d:\xchoo\git\nengo\nengo\utils\magic.py", line 181, in __call__
    return self.wrapper(self.__wrapped__, self.instance, args, kwargs)
  File "d:\xchoo\git\nengo-dl\nengo_dl\simulator.py", line 50, in with_self
    output = wrapped(*args, **kwargs)
  File "d:\xchoo\git\nengo-dl\nengo_dl\simulator.py", line 801, in compile
    metrics = self._standardize_data(
  File "d:\xchoo\git\nengo-dl\nengo_dl\simulator.py", line 1766, in _standardize_data
    f"Number of elements ({len(data)}) in "
IndexError: list index out of range

Minimum working example:

import nengo
import nengo_dl
import numpy as np
import tensorflow as tf

with nengo.Network() as model:
    inp = nengo.Node(np.ones(10))
    x = nengo_dl.Layer(tf.keras.layers.Dense(10))(inp)

    out = nengo.Probe(x)  # Comment out this line for error to be thrown

with nengo_dl.Simulator(model, minibatch_size=50) as sim:
    sim.compile(
        metrics=["accuracy"],
    )

While I acknowledge that the probe is logically necessary when constructing the NengoDL network, if you accidentally omit the probes, the error message thrown by the compilation step should be more informative.