284

When should I use .eval()? I understand it is supposed to allow me to "evaluate my model". How do I turn it back off for training?

Example training code using .eval().

Improve this question
4

5 Answers 5

Reset to default
389

model.eval() is a kind of switch for some specific layers/parts of the model that behave differently during training and inference (evaluating) time. For example, Dropouts Layers, BatchNorm Layers etc. You need to turn them off during model evaluation, and .eval() will do it for you. In addition, the common practice for evaluating/validation is using torch.no_grad() in pair with model.eval() to turn off gradients computation:

# evaluate model:
model.eval()

with torch.no_grad():
    ...
    out_data = model(data)
    ...

BUT, don't forget to turn back to training mode after eval step:

# training step
...
model.train()
...
Improve this answer
11
  • 33
    torch.no_grad() is a context manager, so you should use it in a form of with torch.no_grad():, that guarantees when leaving with ... block model will turn on gradients computations automatically
    – trsvchn
    Commented Feb 1, 2020 at 17:18
  • 29
    so, model.train() and model.eval() have effect only on Layers, not on gradients, by default grad comp is switch on, but using context manager torch.no_grad() during evaluation allows you easily turn off and then autimatically turn on gradients comp at the end
    – trsvchn
    Commented Feb 1, 2020 at 17:26
  • 10
    Why do we need to turn off grad comp on Eval?
    – shtse8
    Commented Aug 13, 2020 at 12:44
  • 22
    @shtse8 we don't compute or use gradients during evaluation, so turning off the autograd will speed up execution and will reduce memory usage
    – trsvchn
    Commented Oct 2, 2020 at 0:27
  • 3
    @NagabhushanSN yes! They work recursively, it looks like this for .eval(): for module in self.children(): module.train(False) and for module in self.children(): module.train(True) for .train()
    – trsvchn
    Commented Jan 24, 2021 at 13:59
91
model.train() model.eval()
Sets model in training mode:

• normalisation layers1 use per-batch statistics
• activates Dropout layers2		 Sets model in evaluation (inference) mode:

• normalisation layers use running statistics
• de-activates Dropout layers
Equivalent to model.train(False).

You can turn off evaluation mode by running model.train(). You should use it when running your model as an inference engine - i.e. when testing, validating, and predicting (though practically it will make no difference if your model does not include any of the differently behaving layers).

  1. e.g. BatchNorm, InstanceNorm
  2. This includes sub-modules of RNN modules etc.
Improve this answer
0
21

model.eval is a method of torch.nn.Module:

eval()

Sets the module in evaluation mode.

This has any effect only on certain modules. See documentations of particular modules for details of their behaviors in training/evaluation mode, if they are affected, e.g. Dropout, BatchNorm, etc.

This is equivalent with self.train(False).

The opposite method is model.train explained nicely by Umang Gupta.

Improve this answer
2
3

An extra addition to the above answers:

I recently started working with Pytorch-lightning, which wraps much of the boilerplate in the training-validation-testing pipelines.

Among other things, it makes model.eval() and model.train() near redundant by allowing the train_step and validation_step callbacks which wrap the eval and train so you never forget to.

Improve this answer
1
  • 2
    sorry I saw delete not elaborate. a bit dyslectic.to the question: Lightning handles the train/test loop for you, and you only have to define train_step and val_step and so on. the model.eval() and model.train() are done in he background, and you don't have to worry about them. I recommend you watch some of their videos, it is a well worth 30 minute investment.
    – Gulzar
    Commented Mar 2, 2021 at 10:30
0

model.eval()

GFG has very clear explanation about it.

  • sets the PyTorch model to evaluation mode, disabling operations like dropout, useful for inference and testing.
  • This method plays a pivotal role in ensuring consistent and reliable model behaviour during inference and testing.
Improve this answer

Not the answer you're looking for? Browse other questions tagged or ask your own question.