methane_super_emitters.model 

  1import lightning as L
  2import torchmetrics
  3import torch
  4import torch.nn as nn
  5
  6
  7class SuperEmitterLocator(L.LightningModule):
  8    def __init__(self, fields):
  9        self.fields = fields
 10        super().__init__()
 11        self.conv_layers = nn.Sequential(
 12            nn.Conv2d(len(self.fields), 16, kernel_size=3, padding=1),
 13            nn.ReLU(),
 14            nn.Conv2d(16, 32, kernel_size=3, padding=1),
 15            nn.ReLU(),
 16            nn.Conv2d(32, 64, kernel_size=3, padding=1),
 17            nn.ReLU(),
 18            nn.Conv2d(64, 128, kernel_size=3, padding=1),
 19            nn.ReLU(),
 20            nn.AdaptiveAvgPool2d((8, 8)),
 21        )
 22        self.fc_layers = nn.Sequential(
 23            nn.Linear(128 * 8 * 8, 1024),
 24            nn.ReLU(),
 25            nn.Linear(1024, 512),
 26            nn.ReLU(),
 27            nn.Linear(512, 32 * 32),  # Output layer
 28        )
 29
 30    def forward(self, x):
 31        out = self.conv_layers(x)
 32        out = out.view(out.size(0), -1)
 33        out = self.fc_layers(out)
 34        return out
 35
 36    def configure_optimizers(self):
 37        return torch.optim.Adam(self.parameters(), lr=1e-4)
 38
 39    def training_step(self, batch, batch_idx):
 40        x, y = batch
 41        y_hat = self(x)
 42        y_hat_flat = y_hat.view(y.shape[0], -1)
 43        y_indices = y.view(y.shape[0], -1).argmax(dim=1)
 44        loss = torch.nn.functional.cross_entropy(y_hat_flat, y_indices)
 45        self.log("train_loss", loss)
 46        return loss
 47
 48class SuperEmitterDetector(L.LightningModule):
 49    def __init__(self, fields, dropout=0.4, weight_decay=0.01, lr=1e-3):
 50        super().__init__()
 51        self.fields = fields
 52        self.dropout = dropout
 53        self.weight_decay = weight_decay
 54        self.lr = lr
 55
 56        self.conv_layers = nn.Sequential(
 57            nn.Conv2d(len(fields), 32, kernel_size=3, stride=1, padding=1),
 58            nn.ReLU(),
 59            nn.MaxPool2d(kernel_size=2, stride=2),
 60            nn.Conv2d(32, 64, kernel_size=3, stride=1, padding=1),
 61            nn.ReLU(),
 62            nn.MaxPool2d(kernel_size=2, stride=2),
 63            nn.Conv2d(64, 128, kernel_size=3, stride=1, padding=1),
 64            nn.ReLU(),
 65            nn.MaxPool2d(kernel_size=2, stride=2),
 66        )
 67
 68        self.fc_layers = nn.Sequential(
 69            nn.Flatten(),
 70            nn.Linear(128 * 4 * 4, 128),
 71            nn.ReLU(),
 72            nn.Dropout(self.dropout),
 73            nn.Linear(128, 1),
 74            nn.Sigmoid(),
 75        )
 76
 77    def forward(self, x):
 78        x = self.conv_layers(x)
 79        x = self.fc_layers(x)
 80        return x
 81
 82    def configure_optimizers(self):
 83        return torch.optim.Adam(self.parameters(), lr=self.lr, weight_decay=self.weight_decay)
 84
 85    def training_step(self, batch, batch_idx):
 86        images, labels = batch
 87        outputs = self(images).squeeze()
 88        loss = nn.functional.binary_cross_entropy(outputs, labels.float())
 89        acc = ((outputs > 0.5).int() == labels).float().mean()
 90        self.log("train_loss", loss)
 91        self.log("train_acc", acc)
 92        return loss
 93
 94    def validation_step(self, batch, batch_idx):
 95        images, labels = batch
 96        outputs = self(images).squeeze()
 97        loss = nn.functional.binary_cross_entropy(outputs, labels.float())
 98        acc = ((outputs > 0.5).int() == labels).float().mean()
 99        self.log("val_loss", loss)
100        self.log("val_acc", acc)
101        return loss
102
103    def test_step(self, batch, batch_idx):
104        images, labels = batch
105        outputs = self(images).squeeze()
106        acc = ((outputs > 0.5).int() == labels).float().mean()
107        self.log("test_acc", acc)
class SuperEmitterLocator(lightning.pytorch.core.module.LightningModule): 
 8class SuperEmitterLocator(L.LightningModule):
 9    def __init__(self, fields):
10        self.fields = fields
11        super().__init__()
12        self.conv_layers = nn.Sequential(
13            nn.Conv2d(len(self.fields), 16, kernel_size=3, padding=1),
14            nn.ReLU(),
15            nn.Conv2d(16, 32, kernel_size=3, padding=1),
16            nn.ReLU(),
17            nn.Conv2d(32, 64, kernel_size=3, padding=1),
18            nn.ReLU(),
19            nn.Conv2d(64, 128, kernel_size=3, padding=1),
20            nn.ReLU(),
21            nn.AdaptiveAvgPool2d((8, 8)),
22        )
23        self.fc_layers = nn.Sequential(
24            nn.Linear(128 * 8 * 8, 1024),
25            nn.ReLU(),
26            nn.Linear(1024, 512),
27            nn.ReLU(),
28            nn.Linear(512, 32 * 32),  # Output layer
29        )
30
31    def forward(self, x):
32        out = self.conv_layers(x)
33        out = out.view(out.size(0), -1)
34        out = self.fc_layers(out)
35        return out
36
37    def configure_optimizers(self):
38        return torch.optim.Adam(self.parameters(), lr=1e-4)
39
40    def training_step(self, batch, batch_idx):
41        x, y = batch
42        y_hat = self(x)
43        y_hat_flat = y_hat.view(y.shape[0], -1)
44        y_indices = y.view(y.shape[0], -1).argmax(dim=1)
45        loss = torch.nn.functional.cross_entropy(y_hat_flat, y_indices)
46        self.log("train_loss", loss)
47        return loss

Hooks to be used in LightningModule.

SuperEmitterLocator(fields) 
 9    def __init__(self, fields):
10        self.fields = fields
11        super().__init__()
12        self.conv_layers = nn.Sequential(
13            nn.Conv2d(len(self.fields), 16, kernel_size=3, padding=1),
14            nn.ReLU(),
15            nn.Conv2d(16, 32, kernel_size=3, padding=1),
16            nn.ReLU(),
17            nn.Conv2d(32, 64, kernel_size=3, padding=1),
18            nn.ReLU(),
19            nn.Conv2d(64, 128, kernel_size=3, padding=1),
20            nn.ReLU(),
21            nn.AdaptiveAvgPool2d((8, 8)),
22        )
23        self.fc_layers = nn.Sequential(
24            nn.Linear(128 * 8 * 8, 1024),
25            nn.ReLU(),
26            nn.Linear(1024, 512),
27            nn.ReLU(),
28            nn.Linear(512, 32 * 32),  # Output layer
29        )

Attributes: prepare_data_per_node: If True, each LOCAL_RANK=0 will call prepare data. Otherwise only NODE_RANK=0, LOCAL_RANK=0 will prepare data. allow_zero_length_dataloader_with_multiple_devices: If True, dataloader with zero length within local rank is allowed. Default value is False.

fields
conv_layers
fc_layers
def forward(self, x): 
31    def forward(self, x):
32        out = self.conv_layers(x)
33        out = out.view(out.size(0), -1)
34        out = self.fc_layers(out)
35        return out

Same as torch.nn.Module.forward().

Args: args: Whatever you decide to pass into the forward method. *kwargs: Keyword arguments are also possible.

Return: Your model's output

def configure_optimizers(self): 
37    def configure_optimizers(self):
38        return torch.optim.Adam(self.parameters(), lr=1e-4)

Choose what optimizers and learning-rate schedulers to use in your optimization. Normally you'd need one. But in the case of GANs or similar you might have multiple. Optimization with multiple optimizers only works in the manual optimization mode.

Return: Any of these 6 options.

- **Single optimizer**.
- **List or Tuple** of optimizers.
- **Two lists** - The first list has multiple optimizers, and the second has multiple LR schedulers
  (or multiple ``lr_scheduler_config``).
- **Dictionary**, with an ``"optimizer"`` key, and (optionally) a ``"lr_scheduler"``
  key whose value is a single LR scheduler or ``lr_scheduler_config``.
- **None** - Fit will run without any optimizer.

The lr_scheduler_config is a dictionary which contains the scheduler and its associated configuration. The default configuration is shown below.

lr_scheduler_config = {
    # REQUIRED: The scheduler instance
    "scheduler": lr_scheduler,
    # The unit of the scheduler's step size, could also be 'step'.
    # 'epoch' updates the scheduler on epoch end whereas 'step'
    # updates it after a optimizer update.
    "interval": "epoch",
    # How many epochs/steps should pass between calls to
    # `scheduler.step()`. 1 corresponds to updating the learning
    # rate after every epoch/step.
    "frequency": 1,
    # Metric to monitor for schedulers like `ReduceLROnPlateau`
    "monitor": "val_loss",
    # If set to `True`, will enforce that the value specified 'monitor'
    # is available when the scheduler is updated, thus stopping
    # training if not found. If set to `False`, it will only produce a warning
    "strict": True,
    # If using the `LearningRateMonitor` callback to monitor the
    # learning rate progress, this keyword can be used to specify
    # a custom logged name
    "name": None,
}

When there are schedulers in which the .step() method is conditioned on a value, such as the torch.optim.lr_scheduler.ReduceLROnPlateau scheduler, Lightning requires that the lr_scheduler_config contains the keyword "monitor" set to the metric name that the scheduler should be conditioned on.

.. testcode::

# The ReduceLROnPlateau scheduler requires a monitor
def configure_optimizers(self):
    optimizer = Adam(...)
    return {
        "optimizer": optimizer,
        "lr_scheduler": {
            "scheduler": ReduceLROnPlateau(optimizer, ...),
            "monitor": "metric_to_track",
            "frequency": "indicates how often the metric is updated",
            # If "monitor" references validation metrics, then "frequency" should be set to a
            # multiple of "trainer.check_val_every_n_epoch".
        },
    }


# In the case of two optimizers, only one using the ReduceLROnPlateau scheduler
def configure_optimizers(self):
    optimizer1 = Adam(...)
    optimizer2 = SGD(...)
    scheduler1 = ReduceLROnPlateau(optimizer1, ...)
    scheduler2 = LambdaLR(optimizer2, ...)
    return (
        {
            "optimizer": optimizer1,
            "lr_scheduler": {
                "scheduler": scheduler1,
                "monitor": "metric_to_track",
            },
        },
        {"optimizer": optimizer2, "lr_scheduler": scheduler2},
    )

Metrics can be made available to monitor by simply logging it using self.log('metric_to_track', metric_val) in your ~lightning.pytorch.core.LightningModule.

Note: Some things to know:

- Lightning calls ``.backward()`` and ``.step()`` automatically in case of automatic optimization.
- If a learning rate scheduler is specified in ``configure_optimizers()`` with key
  ``"interval"`` (default "epoch") in the scheduler configuration, Lightning will call
  the scheduler's ``.step()`` method automatically in case of automatic optimization.
- If you use 16-bit precision (``precision=16``), Lightning will automatically handle the optimizer.
- If you use `torch.optim.LBFGS`, Lightning handles the closure function automatically for you.
- If you use multiple optimizers, you will have to switch to 'manual optimization' mode and step them
  yourself.
- If you need to control how often the optimizer steps, override the `optimizer_step()` hook.
def training_step(self, batch, batch_idx): 
40    def training_step(self, batch, batch_idx):
41        x, y = batch
42        y_hat = self(x)
43        y_hat_flat = y_hat.view(y.shape[0], -1)
44        y_indices = y.view(y.shape[0], -1).argmax(dim=1)
45        loss = torch.nn.functional.cross_entropy(y_hat_flat, y_indices)
46        self.log("train_loss", loss)
47        return loss

Here you compute and return the training loss and some additional metrics for e.g. the progress bar or logger.

Args: batch: The output of your data iterable, normally a ~torch.utils.data.DataLoader. batch_idx: The index of this batch. dataloader_idx: The index of the dataloader that produced this batch. (only if multiple dataloaders used)

Return: - ~torch.Tensor - The loss tensor - dict - A dictionary which can include any keys, but must include the key 'loss' in the case of automatic optimization. - None - In automatic optimization, this will skip to the next batch (but is not supported for multi-GPU, TPU, or DeepSpeed). For manual optimization, this has no special meaning, as returning the loss is not required.

In this step you'd normally do the forward pass and calculate the loss for a batch. You can also do fancier things like multiple forward passes or something model specific.

Example::

def training_step(self, batch, batch_idx):
    x, y, z = batch
    out = self.encoder(x)
    loss = self.loss(out, x)
    return loss

To use multiple optimizers, you can switch to 'manual optimization' and control their stepping:

def __init__(self):
    super().__init__()
    self.automatic_optimization = False


# Multiple optimizers (e.g.: GANs)
def training_step(self, batch, batch_idx):
    opt1, opt2 = self.optimizers()

    # do training_step with encoder
    ...
    opt1.step()
    # do training_step with decoder
    ...
    opt2.step()

Note: When accumulate_grad_batches > 1, the loss returned here will be automatically normalized by accumulate_grad_batches internally.

class SuperEmitterDetector(lightning.pytorch.core.module.LightningModule): 
 49class SuperEmitterDetector(L.LightningModule):
 50    def __init__(self, fields, dropout=0.4, weight_decay=0.01, lr=1e-3):
 51        super().__init__()
 52        self.fields = fields
 53        self.dropout = dropout
 54        self.weight_decay = weight_decay
 55        self.lr = lr
 56
 57        self.conv_layers = nn.Sequential(
 58            nn.Conv2d(len(fields), 32, kernel_size=3, stride=1, padding=1),
 59            nn.ReLU(),
 60            nn.MaxPool2d(kernel_size=2, stride=2),
 61            nn.Conv2d(32, 64, kernel_size=3, stride=1, padding=1),
 62            nn.ReLU(),
 63            nn.MaxPool2d(kernel_size=2, stride=2),
 64            nn.Conv2d(64, 128, kernel_size=3, stride=1, padding=1),
 65            nn.ReLU(),
 66            nn.MaxPool2d(kernel_size=2, stride=2),
 67        )
 68
 69        self.fc_layers = nn.Sequential(
 70            nn.Flatten(),
 71            nn.Linear(128 * 4 * 4, 128),
 72            nn.ReLU(),
 73            nn.Dropout(self.dropout),
 74            nn.Linear(128, 1),
 75            nn.Sigmoid(),
 76        )
 77
 78    def forward(self, x):
 79        x = self.conv_layers(x)
 80        x = self.fc_layers(x)
 81        return x
 82
 83    def configure_optimizers(self):
 84        return torch.optim.Adam(self.parameters(), lr=self.lr, weight_decay=self.weight_decay)
 85
 86    def training_step(self, batch, batch_idx):
 87        images, labels = batch
 88        outputs = self(images).squeeze()
 89        loss = nn.functional.binary_cross_entropy(outputs, labels.float())
 90        acc = ((outputs > 0.5).int() == labels).float().mean()
 91        self.log("train_loss", loss)
 92        self.log("train_acc", acc)
 93        return loss
 94
 95    def validation_step(self, batch, batch_idx):
 96        images, labels = batch
 97        outputs = self(images).squeeze()
 98        loss = nn.functional.binary_cross_entropy(outputs, labels.float())
 99        acc = ((outputs > 0.5).int() == labels).float().mean()
100        self.log("val_loss", loss)
101        self.log("val_acc", acc)
102        return loss
103
104    def test_step(self, batch, batch_idx):
105        images, labels = batch
106        outputs = self(images).squeeze()
107        acc = ((outputs > 0.5).int() == labels).float().mean()
108        self.log("test_acc", acc)

Hooks to be used in LightningModule.

SuperEmitterDetector(fields, dropout=0.4, weight_decay=0.01, lr=0.001) 
50    def __init__(self, fields, dropout=0.4, weight_decay=0.01, lr=1e-3):
51        super().__init__()
52        self.fields = fields
53        self.dropout = dropout
54        self.weight_decay = weight_decay
55        self.lr = lr
56
57        self.conv_layers = nn.Sequential(
58            nn.Conv2d(len(fields), 32, kernel_size=3, stride=1, padding=1),
59            nn.ReLU(),
60            nn.MaxPool2d(kernel_size=2, stride=2),
61            nn.Conv2d(32, 64, kernel_size=3, stride=1, padding=1),
62            nn.ReLU(),
63            nn.MaxPool2d(kernel_size=2, stride=2),
64            nn.Conv2d(64, 128, kernel_size=3, stride=1, padding=1),
65            nn.ReLU(),
66            nn.MaxPool2d(kernel_size=2, stride=2),
67        )
68
69        self.fc_layers = nn.Sequential(
70            nn.Flatten(),
71            nn.Linear(128 * 4 * 4, 128),
72            nn.ReLU(),
73            nn.Dropout(self.dropout),
74            nn.Linear(128, 1),
75            nn.Sigmoid(),
76        )

Attributes: prepare_data_per_node: If True, each LOCAL_RANK=0 will call prepare data. Otherwise only NODE_RANK=0, LOCAL_RANK=0 will prepare data. allow_zero_length_dataloader_with_multiple_devices: If True, dataloader with zero length within local rank is allowed. Default value is False.

fields
dropout
weight_decay
lr
conv_layers
fc_layers
def forward(self, x): 
78    def forward(self, x):
79        x = self.conv_layers(x)
80        x = self.fc_layers(x)
81        return x

Same as torch.nn.Module.forward().

Args: args: Whatever you decide to pass into the forward method. *kwargs: Keyword arguments are also possible.

Return: Your model's output

def configure_optimizers(self): 
83    def configure_optimizers(self):
84        return torch.optim.Adam(self.parameters(), lr=self.lr, weight_decay=self.weight_decay)

Choose what optimizers and learning-rate schedulers to use in your optimization. Normally you'd need one. But in the case of GANs or similar you might have multiple. Optimization with multiple optimizers only works in the manual optimization mode.

Return: Any of these 6 options.

- **Single optimizer**.
- **List or Tuple** of optimizers.
- **Two lists** - The first list has multiple optimizers, and the second has multiple LR schedulers
  (or multiple ``lr_scheduler_config``).
- **Dictionary**, with an ``"optimizer"`` key, and (optionally) a ``"lr_scheduler"``
  key whose value is a single LR scheduler or ``lr_scheduler_config``.
- **None** - Fit will run without any optimizer.

The lr_scheduler_config is a dictionary which contains the scheduler and its associated configuration. The default configuration is shown below.

lr_scheduler_config = {
    # REQUIRED: The scheduler instance
    "scheduler": lr_scheduler,
    # The unit of the scheduler's step size, could also be 'step'.
    # 'epoch' updates the scheduler on epoch end whereas 'step'
    # updates it after a optimizer update.
    "interval": "epoch",
    # How many epochs/steps should pass between calls to
    # `scheduler.step()`. 1 corresponds to updating the learning
    # rate after every epoch/step.
    "frequency": 1,
    # Metric to monitor for schedulers like `ReduceLROnPlateau`
    "monitor": "val_loss",
    # If set to `True`, will enforce that the value specified 'monitor'
    # is available when the scheduler is updated, thus stopping
    # training if not found. If set to `False`, it will only produce a warning
    "strict": True,
    # If using the `LearningRateMonitor` callback to monitor the
    # learning rate progress, this keyword can be used to specify
    # a custom logged name
    "name": None,
}

When there are schedulers in which the .step() method is conditioned on a value, such as the torch.optim.lr_scheduler.ReduceLROnPlateau scheduler, Lightning requires that the lr_scheduler_config contains the keyword "monitor" set to the metric name that the scheduler should be conditioned on.

.. testcode::

# The ReduceLROnPlateau scheduler requires a monitor
def configure_optimizers(self):
    optimizer = Adam(...)
    return {
        "optimizer": optimizer,
        "lr_scheduler": {
            "scheduler": ReduceLROnPlateau(optimizer, ...),
            "monitor": "metric_to_track",
            "frequency": "indicates how often the metric is updated",
            # If "monitor" references validation metrics, then "frequency" should be set to a
            # multiple of "trainer.check_val_every_n_epoch".
        },
    }


# In the case of two optimizers, only one using the ReduceLROnPlateau scheduler
def configure_optimizers(self):
    optimizer1 = Adam(...)
    optimizer2 = SGD(...)
    scheduler1 = ReduceLROnPlateau(optimizer1, ...)
    scheduler2 = LambdaLR(optimizer2, ...)
    return (
        {
            "optimizer": optimizer1,
            "lr_scheduler": {
                "scheduler": scheduler1,
                "monitor": "metric_to_track",
            },
        },
        {"optimizer": optimizer2, "lr_scheduler": scheduler2},
    )

Metrics can be made available to monitor by simply logging it using self.log('metric_to_track', metric_val) in your ~lightning.pytorch.core.LightningModule.

Note: Some things to know:

- Lightning calls ``.backward()`` and ``.step()`` automatically in case of automatic optimization.
- If a learning rate scheduler is specified in ``configure_optimizers()`` with key
  ``"interval"`` (default "epoch") in the scheduler configuration, Lightning will call
  the scheduler's ``.step()`` method automatically in case of automatic optimization.
- If you use 16-bit precision (``precision=16``), Lightning will automatically handle the optimizer.
- If you use `torch.optim.LBFGS`, Lightning handles the closure function automatically for you.
- If you use multiple optimizers, you will have to switch to 'manual optimization' mode and step them
  yourself.
- If you need to control how often the optimizer steps, override the `optimizer_step()` hook.
def training_step(self, batch, batch_idx): 
86    def training_step(self, batch, batch_idx):
87        images, labels = batch
88        outputs = self(images).squeeze()
89        loss = nn.functional.binary_cross_entropy(outputs, labels.float())
90        acc = ((outputs > 0.5).int() == labels).float().mean()
91        self.log("train_loss", loss)
92        self.log("train_acc", acc)
93        return loss

Here you compute and return the training loss and some additional metrics for e.g. the progress bar or logger.

Args: batch: The output of your data iterable, normally a ~torch.utils.data.DataLoader. batch_idx: The index of this batch. dataloader_idx: The index of the dataloader that produced this batch. (only if multiple dataloaders used)

Return: - ~torch.Tensor - The loss tensor - dict - A dictionary which can include any keys, but must include the key 'loss' in the case of automatic optimization. - None - In automatic optimization, this will skip to the next batch (but is not supported for multi-GPU, TPU, or DeepSpeed). For manual optimization, this has no special meaning, as returning the loss is not required.

In this step you'd normally do the forward pass and calculate the loss for a batch. You can also do fancier things like multiple forward passes or something model specific.

Example::

def training_step(self, batch, batch_idx):
    x, y, z = batch
    out = self.encoder(x)
    loss = self.loss(out, x)
    return loss

To use multiple optimizers, you can switch to 'manual optimization' and control their stepping:

def __init__(self):
    super().__init__()
    self.automatic_optimization = False


# Multiple optimizers (e.g.: GANs)
def training_step(self, batch, batch_idx):
    opt1, opt2 = self.optimizers()

    # do training_step with encoder
    ...
    opt1.step()
    # do training_step with decoder
    ...
    opt2.step()

Note: When accumulate_grad_batches > 1, the loss returned here will be automatically normalized by accumulate_grad_batches internally.

def validation_step(self, batch, batch_idx): 
 95    def validation_step(self, batch, batch_idx):
 96        images, labels = batch
 97        outputs = self(images).squeeze()
 98        loss = nn.functional.binary_cross_entropy(outputs, labels.float())
 99        acc = ((outputs > 0.5).int() == labels).float().mean()
100        self.log("val_loss", loss)
101        self.log("val_acc", acc)
102        return loss

Operates on a single batch of data from the validation set. In this step you'd might generate examples or calculate anything of interest like accuracy.

Args: batch: The output of your data iterable, normally a ~torch.utils.data.DataLoader. batch_idx: The index of this batch. dataloader_idx: The index of the dataloader that produced this batch. (only if multiple dataloaders used)

Return: - ~torch.Tensor - The loss tensor - dict - A dictionary. Can include any keys, but must include the key 'loss'. - None - Skip to the next batch.

# if you have one val dataloader:
def validation_step(self, batch, batch_idx): ...


# if you have multiple val dataloaders:
def validation_step(self, batch, batch_idx, dataloader_idx=0): ...

Examples::

# CASE 1: A single validation dataset
def validation_step(self, batch, batch_idx):
    x, y = batch

    # implement your own
    out = self(x)
    loss = self.loss(out, y)

    # log 6 example images
    # or generated text... or whatever
    sample_imgs = x[:6]
    grid = torchvision.utils.make_grid(sample_imgs)
    self.logger.experiment.add_image('example_images', grid, 0)

    # calculate acc
    labels_hat = torch.argmax(out, dim=1)
    val_acc = torch.sum(y == labels_hat).item() / (len(y) * 1.0)

    # log the outputs!
    self.log_dict({'val_loss': loss, 'val_acc': val_acc})

If you pass in multiple val dataloaders, validation_step() will have an additional argument. We recommend setting the default value of 0 so that you can quickly switch between single and multiple dataloaders.

# CASE 2: multiple validation dataloaders
def validation_step(self, batch, batch_idx, dataloader_idx=0):
    # dataloader_idx tells you which dataset this is.
    ...

Note: If you don't need to validate you don't need to implement this method.

Note: When the validation_step() is called, the model has been put in eval mode and PyTorch gradients have been disabled. At the end of validation, the model goes back to training mode and gradients are enabled.

def test_step(self, batch, batch_idx): 
104    def test_step(self, batch, batch_idx):
105        images, labels = batch
106        outputs = self(images).squeeze()
107        acc = ((outputs > 0.5).int() == labels).float().mean()
108        self.log("test_acc", acc)

Operates on a single batch of data from the test set. In this step you'd normally generate examples or calculate anything of interest such as accuracy.

Args: batch: The output of your data iterable, normally a ~torch.utils.data.DataLoader. batch_idx: The index of this batch. dataloader_idx: The index of the dataloader that produced this batch. (only if multiple dataloaders used)

Return: - ~torch.Tensor - The loss tensor - dict - A dictionary. Can include any keys, but must include the key 'loss'. - None - Skip to the next batch.

# if you have one test dataloader:
def test_step(self, batch, batch_idx): ...


# if you have multiple test dataloaders:
def test_step(self, batch, batch_idx, dataloader_idx=0): ...

Examples::

# CASE 1: A single test dataset
def test_step(self, batch, batch_idx):
    x, y = batch

    # implement your own
    out = self(x)
    loss = self.loss(out, y)

    # log 6 example images
    # or generated text... or whatever
    sample_imgs = x[:6]
    grid = torchvision.utils.make_grid(sample_imgs)
    self.logger.experiment.add_image('example_images', grid, 0)

    # calculate acc
    labels_hat = torch.argmax(out, dim=1)
    test_acc = torch.sum(y == labels_hat).item() / (len(y) * 1.0)

    # log the outputs!
    self.log_dict({'test_loss': loss, 'test_acc': test_acc})

If you pass in multiple test dataloaders, test_step() will have an additional argument. We recommend setting the default value of 0 so that you can quickly switch between single and multiple dataloaders.

# CASE 2: multiple test dataloaders
def test_step(self, batch, batch_idx, dataloader_idx=0):
    # dataloader_idx tells you which dataset this is.
    ...

Note: If you don't need to test you don't need to implement this method.

Note: When the test_step() is called, the model has been put in eval mode and PyTorch gradients have been disabled. At the end of the test epoch, the model goes back to training mode and gradients are enabled.