synapse.ml.dl package

Submodules

synapse.ml.dl.DeepTextClassifier module

class synapse.ml.dl.DeepTextClassifier.DeepTextClassifier(checkpoint=None, additional_layers_to_train=3, num_classes=None, optimizer_name='adam', loss_name='cross_entropy', tokenizer=None, max_token_len=128, learning_rate=None, train_from_scratch=True, label_col='label', text_col='text', prediction_col='prediction', num_proc=None, backend=None, store=None, metrics=None, loss_weights=None, sample_weight_col=None, gradient_compression=None, input_shapes=None, validation=None, callbacks=None, batch_size=None, val_batch_size=None, epochs=None, verbose=1, random_seed=None, shuffle_buffer_size=None, partitions_per_process=None, run_id=None, train_minibatch_fn=None, train_steps_per_epoch=None, validation_steps_per_epoch=None, transformation_fn=None, transformation_edit_fields=None, transformation_removed_fields=None, train_reader_num_workers=None, trainer_args=None, val_reader_num_workers=None, reader_pool_type=None, label_shapes=None, inmemory_cache_all=False, num_gpus=None, logger=None, log_every_n_steps=50, data_module=None, loader_num_epochs=None, terminate_on_nan=False, profiler=None, debug_data_loader=False, train_async_data_loader_queue_size=None, val_async_data_loader_queue_size=None, use_gpu=True, mp_start_method=None)

Bases: TorchEstimator, TextPredictionParams

additional_layers_to_train = Param(parent='undefined', name='additional_layers_to_train', doc='number of last layers to fine tune for the model, should be larger or equal to 0. default to 3.')

checkpoint = Param(parent='undefined', name='checkpoint', doc='checkpoint of the deep text classifier')

getAdditionalLayersToTrain()

getCheckpoint()

getLearningRate()

getLossName()

getMaxTokenLen()

getNumClasses()

getOptimizerName()

getTokenizer()

getTrainFromScratch()

get_model_class()

learning_rate = Param(parent='undefined', name='learning_rate', doc='learning rate to be used for the optimizer')

loss_name = Param(parent='undefined', name='loss_name', doc='string representation of torch.nn.functional loss function for the underlying pytorch_lightning model, e.g. binary_cross_entropy')

max_token_len = Param(parent='undefined', name='max_token_len', doc='max_token_len')

num_classes = Param(parent='undefined', name='num_classes', doc='number of target classes')

optimizer_name = Param(parent='undefined', name='optimizer_name', doc='string representation of optimizer function for the underlying pytorch_lightning model')

setAdditionalLayersToTrain(value)

setCheckpoint(value)

setLearningRate(value)

setLossName(value)

setMaxTokenLen(value)

setNumClasses(value)

setOptimizerName(value)

setTokenizer(value)

setTrainFromScratch(value)

tokenizer = Param(parent='undefined', name='tokenizer', doc='tokenizer')

train_from_scratch = Param(parent='undefined', name='train_from_scratch', doc='whether to train the model from scratch or not, if set to False then param additional_layers_to_train need to be specified.')

synapse.ml.dl.DeepTextModel module

class synapse.ml.dl.DeepTextModel.DeepTextModel(history=None, model=None, input_shapes=None, optimizer=None, run_id=None, _metadata=None, loss=None, loss_constructors=None, checkpoint=None, tokenizer=None, max_token_len=128, label_col='label', text_col='text', prediction_col='prediction')

Bases: TorchModel, TextPredictionParams

checkpoint = Param(parent='undefined', name='checkpoint', doc='checkpoint of the deep text classifier')

getCheckpoint()

getMaxTokenLen()

getOptimizer()

getTokenizer()

get_prediction_fn()

Return a function to perdict output from Row.

max_token_len = Param(parent='undefined', name='max_token_len', doc='max_token_len')

setCheckpoint(value)

setMaxTokenLen(value)

setTokenizer(value)

tokenizer = Param(parent='undefined', name='tokenizer', doc='tokenizer')

synapse.ml.dl.DeepVisionClassifier module

class synapse.ml.dl.DeepVisionClassifier.DeepVisionClassifier(backbone=None, additional_layers_to_train=0, num_classes=None, optimizer_name='adam', loss_name='cross_entropy', dropout_aux=0.7, transform_fn=None, label_col='label', image_col='image', prediction_col='prediction', num_proc=None, backend=None, store=None, metrics=None, loss_weights=None, sample_weight_col=None, gradient_compression=None, input_shapes=None, validation=None, callbacks=None, batch_size=None, val_batch_size=None, epochs=None, verbose=1, random_seed=None, shuffle_buffer_size=None, partitions_per_process=None, run_id=None, train_minibatch_fn=None, train_steps_per_epoch=None, validation_steps_per_epoch=None, transformation_fn=None, train_reader_num_workers=None, trainer_args=None, val_reader_num_workers=None, reader_pool_type=None, label_shapes=None, inmemory_cache_all=False, num_gpus=None, logger=None, log_every_n_steps=50, data_module=None, loader_num_epochs=None, terminate_on_nan=False, profiler=None, debug_data_loader=False, train_async_data_loader_queue_size=None, val_async_data_loader_queue_size=None, use_gpu=True, mp_start_method=None)

Bases: TorchEstimator, VisionPredictionParams

additional_layers_to_train = Param(parent='undefined', name='additional_layers_to_train', doc='number of last layers to fine tune for the model, should be between 0 and 3')

backbone = Param(parent='undefined', name='backbone', doc='backbone of the deep vision classifier')

dropout_aux = Param(parent='undefined', name='dropout_aux', doc="numeric value that's applied to googlenet InceptionAux module's dropout layer only: probability of an element to be zeroed")

getAdditionalLayersToTrain()

getBackbone()

getDropoutAUX()

getLossName()

getNumClasses()

getOptimizerName()

getTransformFn()

get_model_class()

loss_name = Param(parent='undefined', name='loss_name', doc='string representation of torch.nn.functional loss function for the underlying pytorch_lightning model, e.g. binary_cross_entropy')

num_classes = Param(parent='undefined', name='num_classes', doc='number of target classes')

optimizer_name = Param(parent='undefined', name='optimizer_name', doc='string representation of optimizer function for the underlying pytorch_lightning model')

setAdditionalLayersToTrain(value)

setBackbone(value)

setDropoutAUX(value)

setLossName(value)

setNumClasses(value)

setOptimizerName(value)

setTransformFn(value)

transform_fn = Param(parent='undefined', name='transform_fn', doc='A composition of transforms used to transform and augnment the input image, should be of type torchvision.transforms.Compose')

synapse.ml.dl.DeepVisionModel module

class synapse.ml.dl.DeepVisionModel.DeepVisionModel(history=None, model=None, input_shapes=None, optimizer=None, run_id=None, _metadata=None, loss=None, loss_constructors=None, transform_fn=None, label_col='label', image_col='image', prediction_col='prediction')

Bases: TorchModel, VisionPredictionParams

getOptimizer()

getTransformFn()

getTransformationFn()

get_prediction_fn()

Return a function to perdict output from Row.

setTransformFn(value)

setTransformationFn(value)

transform_fn = Param(parent='undefined', name='transform_fn', doc='A composition of transforms used to transform and augnment the input image, should be of type torchvision.transforms.Compose')

synapse.ml.dl.LitDeepTextModel module

class synapse.ml.dl.LitDeepTextModel.LitDeepTextModel(checkpoint, text_col, label_col, num_labels, additional_layers_to_train, optimizer_name, loss_name, learning_rate=None, train_from_scratch=True)

Bases: LightningModule

configure_optimizers()

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.

Returns:

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.

• Tuple of dictionaries as described above, with an optional "frequency" key.

• 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 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.

# 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 LightningModule.

Note

The frequency value specified in a dict along with the optimizer key is an int corresponding to the number of sequential batches optimized with the specific optimizer. It should be given to none or to all of the optimizers. There is a difference between passing multiple optimizers in a list, and passing multiple optimizers in dictionaries with a frequency of 1:

• In the former case, all optimizers will operate on the given batch in each optimization step.

• In the latter, only one optimizer will operate on the given batch at every step.

This is different from the frequency value specified in the lr_scheduler_config mentioned above.

def configure_optimizers(self):
    optimizer_one = torch.optim.SGD(self.model.parameters(), lr=0.01)
    optimizer_two = torch.optim.SGD(self.model.parameters(), lr=0.01)
    return [
        {"optimizer": optimizer_one, "frequency": 5},
        {"optimizer": optimizer_two, "frequency": 10},
    ]

In this example, the first optimizer will be used for the first 5 steps, the second optimizer for the next 10 steps and that cycle will continue. If an LR scheduler is specified for an optimizer using the lr_scheduler key in the above dict, the scheduler will only be updated when its optimizer is being used.

Examples:

# most cases. no learning rate scheduler
def configure_optimizers(self):
    return Adam(self.parameters(), lr=1e-3)

# multiple optimizer case (e.g.: GAN)
def configure_optimizers(self):
    gen_opt = Adam(self.model_gen.parameters(), lr=0.01)
    dis_opt = Adam(self.model_dis.parameters(), lr=0.02)
    return gen_opt, dis_opt

# example with learning rate schedulers
def configure_optimizers(self):
    gen_opt = Adam(self.model_gen.parameters(), lr=0.01)
    dis_opt = Adam(self.model_dis.parameters(), lr=0.02)
    dis_sch = CosineAnnealing(dis_opt, T_max=10)
    return [gen_opt, dis_opt], [dis_sch]

# example with step-based learning rate schedulers
# each optimizer has its own scheduler
def configure_optimizers(self):
    gen_opt = Adam(self.model_gen.parameters(), lr=0.01)
    dis_opt = Adam(self.model_dis.parameters(), lr=0.02)
    gen_sch = {
        'scheduler': ExponentialLR(gen_opt, 0.99),
        'interval': 'step'  # called after each training step
    }
    dis_sch = CosineAnnealing(dis_opt, T_max=10) # called every epoch
    return [gen_opt, dis_opt], [gen_sch, dis_sch]

# example with optimizer frequencies
# see training procedure in `Improved Training of Wasserstein GANs`, Algorithm 1
# https://arxiv.org/abs/1704.00028
def configure_optimizers(self):
    gen_opt = Adam(self.model_gen.parameters(), lr=0.01)
    dis_opt = Adam(self.model_dis.parameters(), lr=0.02)
    n_critic = 5
    return (
        {'optimizer': dis_opt, 'frequency': n_critic},
        {'optimizer': gen_opt, 'frequency': 1}
    )

Note

Some things to know:

• Lightning calls .backward() and .step() on each optimizer and learning rate scheduler as needed.

• If you use 16-bit precision (precision=16), Lightning will automatically handle the optimizers.

• If you use multiple optimizers, training_step() will have an additional optimizer_idx parameter.

• If you use torch.optim.LBFGS, Lightning handles the closure function automatically for you.

• If you use multiple optimizers, gradients will be calculated only for the parameters of current optimizer at each training step.

• If you need to control how often those optimizers step or override the default .step() schedule, override the optimizer_step() hook.

forward(**inputs)

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

Parameters:

• *args – Whatever you decide to pass into the forward method.

• **kwargs – Keyword arguments are also possible.

Returns:

Your model’s output

test_step(batch, batch_idx)

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.

# the pseudocode for these calls
test_outs = []
for test_batch in test_data:
    out = test_step(test_batch)
    test_outs.append(out)
test_epoch_end(test_outs)

Parameters:

• batch (Tensor | (Tensor, …) | [Tensor, …]) – The output of your DataLoader. A tensor, tuple or list.

• batch_idx (int) – The index of this batch.

• dataloader_idx (int) – The index of the dataloader that produced this batch (only if multiple test dataloaders used).

Returns:

Any of.

• Any object or value

• None - Testing will 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):
    ...

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.

# CASE 2: multiple test dataloaders
def test_step(self, batch, batch_idx, dataloader_idx):
    # 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.

training_step(batch, batch_idx)

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

Parameters:

• batch (Tensor | (Tensor, …) | [Tensor, …]) – The output of your DataLoader. A tensor, tuple or list.

• batch_idx (int) – Integer displaying index of this batch

• optimizer_idx (int) – When using multiple optimizers, this argument will also be present.

• hiddens (Any) – Passed in if truncated_bptt_steps > 0.

Returns:

Any of.

• Tensor - The loss tensor

• dict - A dictionary. Can include any keys, but must include the key 'loss'

• None - Training will skip to the next batch. This is only for automatic optimization.

  This is not supported for multi-GPU, TPU, IPU, or DeepSpeed.

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

If you define multiple optimizers, this step will be called with an additional optimizer_idx parameter.

# Multiple optimizers (e.g.: GANs)
def training_step(self, batch, batch_idx, optimizer_idx):
    if optimizer_idx == 0:
        # do training_step with encoder
        ...
    if optimizer_idx == 1:
        # do training_step with decoder
        ...

If you add truncated back propagation through time you will also get an additional argument with the hidden states of the previous step.

# Truncated back-propagation through time
def training_step(self, batch, batch_idx, hiddens):
    # hiddens are the hidden states from the previous truncated backprop step
    out, hiddens = self.lstm(data, hiddens)
    loss = ...
    return {"loss": loss, "hiddens": hiddens}

Note

The loss value shown in the progress bar is smoothed (averaged) over the last values, so it differs from the actual loss returned in train/validation step.

validation_epoch_end(outputs)

Called at the end of the validation epoch with the outputs of all validation steps.

# the pseudocode for these calls
val_outs = []
for val_batch in val_data:
    out = validation_step(val_batch)
    val_outs.append(out)
validation_epoch_end(val_outs)

Parameters:

outputs – List of outputs you defined in validation_step(), or if there are multiple dataloaders, a list containing a list of outputs for each dataloader.

Returns:

None

Note

If you didn’t define a validation_step(), this won’t be called.

Examples

With a single dataloader:

def validation_epoch_end(self, val_step_outputs):
    for out in val_step_outputs:
        ...

With multiple dataloaders, outputs will be a list of lists. The outer list contains one entry per dataloader, while the inner list contains the individual outputs of each validation step for that dataloader.

def validation_epoch_end(self, outputs):
    for dataloader_output_result in outputs:
        dataloader_outs = dataloader_output_result.dataloader_i_outputs

    self.log("final_metric", final_value)

validation_step(batch, batch_idx)

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.

# the pseudocode for these calls
val_outs = []
for val_batch in val_data:
    out = validation_step(val_batch)
    val_outs.append(out)
validation_epoch_end(val_outs)

Parameters:

• batch (Tensor | (Tensor, …) | [Tensor, …]) – The output of your DataLoader. A tensor, tuple or list.

• batch_idx (int) – The index of this batch

• dataloader_idx (int) – The index of the dataloader that produced this batch (only if multiple val dataloaders used)

Returns:

• Any object or value

• None - Validation will skip to the next batch

# pseudocode of order
val_outs = []
for val_batch in val_data:
    out = validation_step(val_batch)
    if defined("validation_step_end"):
        out = validation_step_end(out)
    val_outs.append(out)
val_outs = validation_epoch_end(val_outs)

# 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):
    ...

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.

# CASE 2: multiple validation dataloaders
def validation_step(self, batch, batch_idx, dataloader_idx):
    # 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.

synapse.ml.dl.LitDeepVisionModel module

class synapse.ml.dl.LitDeepVisionModel.LitDeepVisionModel(backbone, additional_layers_to_train, num_classes, input_shape, optimizer_name, loss_name, label_col, image_col, dropout_aux=0.7)

Bases: LightningModule

configure_optimizers()

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.

Returns:

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.

• Tuple of dictionaries as described above, with an optional "frequency" key.

• 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 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.

# 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 LightningModule.

Note

The frequency value specified in a dict along with the optimizer key is an int corresponding to the number of sequential batches optimized with the specific optimizer. It should be given to none or to all of the optimizers. There is a difference between passing multiple optimizers in a list, and passing multiple optimizers in dictionaries with a frequency of 1:

• In the former case, all optimizers will operate on the given batch in each optimization step.

• In the latter, only one optimizer will operate on the given batch at every step.

This is different from the frequency value specified in the lr_scheduler_config mentioned above.

def configure_optimizers(self):
    optimizer_one = torch.optim.SGD(self.model.parameters(), lr=0.01)
    optimizer_two = torch.optim.SGD(self.model.parameters(), lr=0.01)
    return [
        {"optimizer": optimizer_one, "frequency": 5},
        {"optimizer": optimizer_two, "frequency": 10},
    ]

In this example, the first optimizer will be used for the first 5 steps, the second optimizer for the next 10 steps and that cycle will continue. If an LR scheduler is specified for an optimizer using the lr_scheduler key in the above dict, the scheduler will only be updated when its optimizer is being used.

Examples:

# most cases. no learning rate scheduler
def configure_optimizers(self):
    return Adam(self.parameters(), lr=1e-3)

# multiple optimizer case (e.g.: GAN)
def configure_optimizers(self):
    gen_opt = Adam(self.model_gen.parameters(), lr=0.01)
    dis_opt = Adam(self.model_dis.parameters(), lr=0.02)
    return gen_opt, dis_opt

# example with learning rate schedulers
def configure_optimizers(self):
    gen_opt = Adam(self.model_gen.parameters(), lr=0.01)
    dis_opt = Adam(self.model_dis.parameters(), lr=0.02)
    dis_sch = CosineAnnealing(dis_opt, T_max=10)
    return [gen_opt, dis_opt], [dis_sch]

# example with step-based learning rate schedulers
# each optimizer has its own scheduler
def configure_optimizers(self):
    gen_opt = Adam(self.model_gen.parameters(), lr=0.01)
    dis_opt = Adam(self.model_dis.parameters(), lr=0.02)
    gen_sch = {
        'scheduler': ExponentialLR(gen_opt, 0.99),
        'interval': 'step'  # called after each training step
    }
    dis_sch = CosineAnnealing(dis_opt, T_max=10) # called every epoch
    return [gen_opt, dis_opt], [gen_sch, dis_sch]

# example with optimizer frequencies
# see training procedure in `Improved Training of Wasserstein GANs`, Algorithm 1
# https://arxiv.org/abs/1704.00028
def configure_optimizers(self):
    gen_opt = Adam(self.model_gen.parameters(), lr=0.01)
    dis_opt = Adam(self.model_dis.parameters(), lr=0.02)
    n_critic = 5
    return (
        {'optimizer': dis_opt, 'frequency': n_critic},
        {'optimizer': gen_opt, 'frequency': 1}
    )

Note

Some things to know:

• Lightning calls .backward() and .step() on each optimizer and learning rate scheduler as needed.

• If you use 16-bit precision (precision=16), Lightning will automatically handle the optimizers.

• If you use multiple optimizers, training_step() will have an additional optimizer_idx parameter.

• If you use torch.optim.LBFGS, Lightning handles the closure function automatically for you.

• If you use multiple optimizers, gradients will be calculated only for the parameters of current optimizer at each training step.

• If you need to control how often those optimizers step or override the default .step() schedule, override the optimizer_step() hook.

forward(x)

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

Parameters:

• *args – Whatever you decide to pass into the forward method.

• **kwargs – Keyword arguments are also possible.

Returns:

Your model’s output

test_step(batch, batch_idx)

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.

# the pseudocode for these calls
test_outs = []
for test_batch in test_data:
    out = test_step(test_batch)
    test_outs.append(out)
test_epoch_end(test_outs)

Parameters:

• batch (Tensor | (Tensor, …) | [Tensor, …]) – The output of your DataLoader. A tensor, tuple or list.

• batch_idx (int) – The index of this batch.

• dataloader_idx (int) – The index of the dataloader that produced this batch (only if multiple test dataloaders used).

Returns:

Any of.

• Any object or value

• None - Testing will 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):
    ...

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.

# CASE 2: multiple test dataloaders
def test_step(self, batch, batch_idx, dataloader_idx):
    # 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.

training_step(batch, batch_idx)

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

Parameters:

• batch (Tensor | (Tensor, …) | [Tensor, …]) – The output of your DataLoader. A tensor, tuple or list.

• batch_idx (int) – Integer displaying index of this batch

• optimizer_idx (int) – When using multiple optimizers, this argument will also be present.

• hiddens (Any) – Passed in if truncated_bptt_steps > 0.

Returns:

Any of.

• Tensor - The loss tensor

• dict - A dictionary. Can include any keys, but must include the key 'loss'

• None - Training will skip to the next batch. This is only for automatic optimization.

  This is not supported for multi-GPU, TPU, IPU, or DeepSpeed.

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

If you define multiple optimizers, this step will be called with an additional optimizer_idx parameter.

# Multiple optimizers (e.g.: GANs)
def training_step(self, batch, batch_idx, optimizer_idx):
    if optimizer_idx == 0:
        # do training_step with encoder
        ...
    if optimizer_idx == 1:
        # do training_step with decoder
        ...

If you add truncated back propagation through time you will also get an additional argument with the hidden states of the previous step.

# Truncated back-propagation through time
def training_step(self, batch, batch_idx, hiddens):
    # hiddens are the hidden states from the previous truncated backprop step
    out, hiddens = self.lstm(data, hiddens)
    loss = ...
    return {"loss": loss, "hiddens": hiddens}

Note

The loss value shown in the progress bar is smoothed (averaged) over the last values, so it differs from the actual loss returned in train/validation step.

validation_epoch_end(outputs)

Called at the end of the validation epoch with the outputs of all validation steps.

# the pseudocode for these calls
val_outs = []
for val_batch in val_data:
    out = validation_step(val_batch)
    val_outs.append(out)
validation_epoch_end(val_outs)

Parameters:

outputs – List of outputs you defined in validation_step(), or if there are multiple dataloaders, a list containing a list of outputs for each dataloader.

Returns:

None

Note

If you didn’t define a validation_step(), this won’t be called.

Examples

With a single dataloader:

def validation_epoch_end(self, val_step_outputs):
    for out in val_step_outputs:
        ...

With multiple dataloaders, outputs will be a list of lists. The outer list contains one entry per dataloader, while the inner list contains the individual outputs of each validation step for that dataloader.

def validation_epoch_end(self, outputs):
    for dataloader_output_result in outputs:
        dataloader_outs = dataloader_output_result.dataloader_i_outputs

    self.log("final_metric", final_value)

validation_step(batch, batch_idx)

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.

# the pseudocode for these calls
val_outs = []
for val_batch in val_data:
    out = validation_step(val_batch)
    val_outs.append(out)
validation_epoch_end(val_outs)

Parameters:

• batch (Tensor | (Tensor, …) | [Tensor, …]) – The output of your DataLoader. A tensor, tuple or list.

• batch_idx (int) – The index of this batch

• dataloader_idx (int) – The index of the dataloader that produced this batch (only if multiple val dataloaders used)

Returns:

• Any object or value

• None - Validation will skip to the next batch

# pseudocode of order
val_outs = []
for val_batch in val_data:
    out = validation_step(val_batch)
    if defined("validation_step_end"):
        out = validation_step_end(out)
    val_outs.append(out)
val_outs = validation_epoch_end(val_outs)

# 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):
    ...

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.

# CASE 2: multiple validation dataloaders
def validation_step(self, batch, batch_idx, dataloader_idx):
    # 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.

synapse.ml.dl.PredictionParams module

class synapse.ml.dl.PredictionParams.HasImageColParam

Bases: Params

getImageCol()

Gets the value of image_col or its default value.

image_col = Param(parent='undefined', name='image_col', doc='image column name.')

setImageCol(value)

Sets the value of image_col.

class synapse.ml.dl.PredictionParams.HasLabelColParam

Bases: Params

getLabelCol()

Gets the value of label_col or its default value.

label_col = Param(parent='undefined', name='label_col', doc='label column name.')

setLabelCol(value)

Sets the value of label_col.

class synapse.ml.dl.PredictionParams.HasPredictionColParam

Bases: Params

getPredictionCol()

Gets the value of prediction_col or its default value.

prediction_col = Param(parent='undefined', name='prediction_col', doc='prediction column name.')

setPredictionCol(value)

Sets the value of prediction_col.

class synapse.ml.dl.PredictionParams.HasTextColParam

Bases: Params

getTextCol()

Gets the value of text_col or its default value.

setTextCol(value)

Sets the value of text_col.

text_col = Param(parent='undefined', name='text_col', doc='text column name.')

class synapse.ml.dl.PredictionParams.TextPredictionParams

Bases: HasLabelColParam, HasTextColParam, HasPredictionColParam

class synapse.ml.dl.PredictionParams.VisionPredictionParams

Bases: HasLabelColParam, HasImageColParam, HasPredictionColParam

synapse.ml.dl.utils module

synapse.ml.dl.utils.get_or_create_backend(backend, num_proc, verbose, use_gpu)

synapse.ml.dl.utils.keywords_catch(func)

A decorator that forces keyword arguments in the wrapped method and saves actual input keyword arguments in _kwargs.

Notes

Should only be used to wrap a method where first arg is self

Module contents

SynapseML is an ecosystem of tools aimed towards expanding the distributed computing framework Apache Spark in several new directions. SynapseML adds many deep learning and data science tools to the Spark ecosystem, including seamless integration of Spark Machine Learning pipelines with Microsoft Cognitive Toolkit (CNTK), LightGBM and OpenCV. These tools enable powerful and highly-scalable predictive and analytical models for a variety of datasources.

SynapseML also brings new networking capabilities to the Spark Ecosystem. With the HTTP on Spark project, users can embed any web service into their SparkML models. In this vein, SynapseML provides easy to use SparkML transformers for a wide variety of Microsoft Cognitive Services. For production grade deployment, the Spark Serving project enables high throughput, sub-millisecond latency web services, backed by your Spark cluster.

SynapseML requires Scala 2.12, Spark 3.0+, and Python 3.6+.