Parameter description error: pnet_tune

[pbhfcycssjlmm]

In your description of the pnet_tune parameter：
pnet_tune
[False] tune the base/trunk network
[True] keep base/trunk frozen
However：
self.net = net_type(pretrained=not self.pnet_rand, requires_grad=self.pnet_tune)
Thus，the implementation in the code shows:
When pnet_tune=True, set by requires_grad=self.pnet_tune, it means that the weights of the base network will be updated (that is, participate in gradient calculation).
When pnet_tune=False, the weights of the base network will be frozen and will not participate in gradient updates.
So the correct comment should be:
pnet_tune=True: Indicates to "unfreeze" the base network, allowing its weights to be updated during training.
pnet_tune=False: means to "freeze" the base network and its weights will not be updated during training.