poutyne.framework¶
Model¶

class
poutyne.framework.
Model
(model, optimizer, loss_function, *, metrics=[])[source]¶ The Model class encapsulates a PyTorch module/network, a PyTorch optimizer, a loss function and metric functions. It allows the user to train a neural network without handcoding the epoch/step logic.
 Parameters
model (torch.nn.Module) – A PyTorch module.
optimizer (torch.optim.Optimizer) – Initialized PyTorch optimizer.
loss_function – Loss function. It can be any PyTorch loss layer or custom loss function. It can also be a string with the same name as a PyTorch loss function (either the functional or object name). The loss function must have the signature
loss_function(input, target)
whereinput
is the prediction of the network andtarget
is the ground truth.metrics (list) – List of functions with the same signature as the loss function. Each metric can be any PyTorch loss function. It can also be a string with the same name as a PyTorch loss function (either the functional or object name). ‘accuracy’ (or just ‘acc’) is also a valid metric. Each metric function is called on each batch of the optimization and on the validation batches at the end of the epoch. (Default value = [])

model
¶ The associated PyTorch module.
 Type
torch.nn.Module

optimizer
¶ The associated PyTorch optimizer.
 Type
torch.optim.Optimizer

loss_function
¶ The associated loss function.

metrics
¶ The associated metric functions.
 Type
list
Example
Using Numpy arrays (or tensors) dataset:
from poutyne.framework import Model import torch import numpy as np num_features = 20 num_classes = 5 # Our training dataset with 800 samples. num_train_samples = 800 train_x = np.random.randn(num_train_samples, num_features).astype('float32') train_y = np.random.randint(num_classes, size=num_train_samples).astype('int64') # Our validation dataset with 200 samples. num_valid_samples = 200 valid_x = np.random.randn(num_valid_samples, num_features).astype('float32') valid_y = np.random.randint(num_classes, size=num_valid_samples).astype('int64') pytorch_module = torch.nn.Linear(num_features, num_classes) # Our network # We create and optimize our model model = Model(pytorch_module, 'sgd', 'cross_entropy', metrics=['accuracy']) model.fit(train_x, train_y, validation_x=valid_x, validation_y=valid_y, epochs=5, batch_size=32)
Epoch 1/5 0.02s Step 25/25: loss: 1.719885, acc: 19.375000, val_loss: 1.667446, val_acc: 22.000000 Epoch 2/5 0.02s Step 25/25: loss: 1.705489, acc: 19.750000, val_loss: 1.660806, val_acc: 22.000000 Epoch 3/5 0.01s Step 25/25: loss: 1.692345, acc: 19.625000, val_loss: 1.655008, val_acc: 22.500000 ...
Using PyTorch DataLoader:
import torch from torch.utils.data import DataLoader, TensorDataset from poutyne.framework import Model num_features = 20 num_classes = 5 # Our training dataset with 800 samples. num_train_samples = 800 train_x = torch.rand(num_train_samples, num_features) train_y = torch.randint(num_classes, (num_train_samples,), dtype=torch.long) train_dataset = TensorDataset(train_x, train_y) train_generator = DataLoader(train_dataset, batch_size=32) # Our validation dataset with 200 samples. num_valid_samples = 200 valid_x = torch.rand(num_valid_samples, num_features) valid_y = torch.randint(num_classes, (num_valid_samples,), dtype=torch.long) valid_dataset = TensorDataset(valid_x, valid_y) valid_generator = DataLoader(valid_dataset, batch_size=32) pytorch_module = torch.nn.Linear(num_features, num_train_samples) model = Model(pytorch_module, 'sgd', 'cross_entropy', metrics=['accuracy']) model.fit_generator(train_generator, valid_generator, epochs=5)
Epoch 1/5 0.05s Step 25/25: loss: 6.752676, acc: 0.000000, val_loss: 6.575071, val_acc: 0.000000 Epoch 2/5 0.03s Step 25/25: loss: 6.454859, acc: 0.125000, val_loss: 6.279577, val_acc: 0.000000 Epoch 3/5 0.03s Step 25/25: loss: 6.158523, acc: 2.125000, val_loss: 5.985811, val_acc: 9.500000 ...

cpu
(*args, **kwargs)[source]¶ Tranfers the network on the CPU. The arguments are passed to the
torch.nn.Module.cpu()
method. Notice that the device is saved so that the batches can send to the right device before passing it to the network. Returns
self.

cuda
(*args, **kwargs)[source]¶ Tranfers the network on the GPU. The arguments are passed to the
torch.nn.Module.cuda()
method. Notice that the device is saved so that the batches can send to the right device before passing it to the network. Returns
self.

evaluate
(x, y, *, batch_size=32, return_pred=False)[source]¶ Computes the loss and the metrics of the network on batches of samples and optionaly returns the predictions.
 Parameters
x (Union[Tensor, np.ndarray]) – Dataset.
y (Union[Tensor, np.ndarray]) – Dataset ground truths.
batch_size (int) – Number of samples given to the network at one time. (Default value = 32)
return_pred (bool, optional) – Whether to return the predictions for
x
. (Default value = False)
 Returns
Float
loss
if no metrics were specified andreturn_pred
is false.Otherwise, tuple
(loss, metrics)
ifreturn_pred
is false.metrics
is a Numpy array of sizen
, wheren
is the number of metrics ifn > 1
. Ifn == 1
, thenmetrics
is a float. Ifn == 0
, themetrics
is omitted.Tuple
(loss, metrics, pred_y)
ifreturn_pred
is true wherepred_y
is a Numpy array of the predictions.

evaluate_generator
(generator, *, steps=None, return_pred=False)[source]¶ Computes the loss and the metrics of the network on batches of samples and optionaly returns the predictions.
 Parameters
generator –
Generatorlike object for the dataset. The generator must yield a tuple
(x, y)
wherex
is a batch of the dataset andy
is the corresponding ground truths.y
should be a Tensor or a Numpy array with the first dimension being the batch size sincelen(y)
is taken as the batch size. The loss and the metrics are averaged using this batch size. Ify
is not a Tensor or a Numpy array, then a warning is raised and the “batch size” defaults to 1.See the
fit_generator()
method for details on the types of generators supported.steps (int, optional) – Number of iterations done on
generator
. (Defaults the number of steps needed to see the entire dataset)return_pred (bool, optional) – Whether to return the predictions for
x
. (Default value = False)
 Returns
Float
loss
if no metrics were specified andreturn_pred
is false.Otherwise, tuple
(loss, metrics)
ifreturn_pred
is false.metrics
is a Numpy array of sizen
, wheren
is the number of metrics ifn > 1
. Ifn == 1
, thenmetrics
is a float. Ifn == 0
, themetrics
is omitted.Tuple
(loss, metrics, pred_y)
ifreturn_pred
is true wherepred_y
is the list of the predictions of each batch with tensors converted into Numpy arrays.
Example
With no metrics:
model = Model(pytorch_module, optimizer, loss_function, metrics=[]) loss = model.evaluate_generator(test_generator)
With only one metric:
model = Model(pytorch_module, optimizer, loss_function, metrics=[my_metric_fn]) loss, my_metric = model.evaluate_generator(test_generator)
With only several metrics:
model = Model(pytorch_module, optimizer, loss_function, metrics=[my_metric1_fn, my_metric2_fn]) loss, (my_metric1, my_metric2) = model.evaluate_generator(test_generator)
With metrics and
return_pred
flag:model = Model(pytorch_module, optimizer, loss_function, metrics=[my_metric1_fn, my_metric2_fn]) loss, (my_metric1, my_metric2), pred_y = model.evaluate_generator( test_generator, return_pred=True )

evaluate_on_batch
(x, y, *, return_pred=False)[source]¶ Computes the loss and the metrics of the network on a single batch of samples and optionaly returns the predictions.
 Parameters
x (Union[Tensor, np.ndarray]) – Batch.
y (Union[Tensor, np.ndarray]) – Batch ground truths.
return_pred (bool, optional) – Whether to return the predictions for
x
. (Default value = False)
 Returns
Float
loss
if no metrics were specified andreturn_pred
is false.Otherwise, tuple
(loss, metrics)
ifreturn_pred
is false.metrics
is a Numpy array of sizen
, wheren
is the number of metrics ifn > 1
. Ifn == 1
, thenmetrics
is a float. Ifn == 0
, themetrics
is omitted.Tuple
(loss, metrics, pred_y)
ifreturn_pred
is true wherepred_y
is the predictions with tensors converted into Numpy arrays.

fit
(x, y, validation_x=None, validation_y=None, *, batch_size=32, epochs=1000, steps_per_epoch=None, validation_steps=None, initial_epoch=1, verbose=True, callbacks=[])[source]¶ Trains the model on a dataset. This method creates generators and calls the
fit_generator
method. Parameters
x (Union[Tensor, np.ndarray]) – Training dataset.
y (Union[Tensor, np.ndarray]) – Ground truth.
validation_x (Union[Tensor, np.ndarray]) – Validation dataset. The validation datset is optional. (Default value = None)
validation_y (Union[Tensor, np.ndarray]) – Validation ground truth. (Default value = None)
batch_size (int) – Number of samples given to the network at one time. (Default value = 32)
epochs (int) – Number of times the entire training dataset is seen. (Default value = 1000)
steps_per_epoch (int, optional) – Number of batch used during one epoch. Obviously, using this argument may cause one epoch not to see the entire training dataset or see it multiple times. (Defaults the number of steps needed to see the entire training dataset)
validation_steps (int, optional) – Same as for
steps_per_epoch
but for the validation dataset. (Defaults tosteps_per_epoch
if provided or the number of steps needed to see the entire validation dataset)initial_epoch (int, optional) – Epoch at which to start training (useful for resuming a previous training run). (Default value = 1)
verbose (bool) – Whether to display the progress of the training. (Default value = True)
callbacks (list of poutyne.framework.Callback) – List of callbacks that will be called during training. (Default value = [])
 Returns
List of dict containing the history of each epoch.
Example
model = Model(pytorch_module, optimizer, loss_function) history = model.fit(train_x, train_y, validation_x=valid_x, validation_y=valid_y, epochs=num_epochs, batch_size=batch_size, verbose=False) print(*history, sep="\n")
{'epoch': 1, 'loss': 1.7198852968215943, 'time': 0.019999928001197986, 'acc': 19.375, 'val_loss': 1.6674459838867188, 'val_acc': 22.0} {'epoch': 2, 'loss': 1.7054892110824584, 'time': 0.015421080999658443, 'acc': 19.75, 'val_loss': 1.660806336402893, 'val_acc': 22.0} {'epoch': 3, 'loss': 1.6923445892333984, 'time': 0.01363091799794347, 'acc': 19.625, 'val_loss': 1.6550078630447387, 'val_acc': 22.5} ...

fit_generator
(train_generator, valid_generator=None, *, epochs=1000, steps_per_epoch=None, validation_steps=None, initial_epoch=1, verbose=True, callbacks=[])[source]¶ Trains the model on a dataset using a generator.
 Parameters
train_generator –
Generatorlike object for the training dataset. The generator must yield a tuple
(x, y)
wherex
is a batch of the training dataset andy
is the corresponding ground truths.y
should be a Tensor or a Numpy array with the first dimension being the batch size sincelen(y)
is taken as the batch size. The loss and the metrics are averaged using this batch size. Ify
is not a Tensor or a Numpy array, then a warning is raised and the “batch size” defaults to 1.If the generator does not have a method
__len__()
, either thesteps_per_epoch
argument must be provided, or the iterator returned raises a StopIteration exception at the end of the training dataset. PyTorch DataLoaders object do provide a__len__()
method.Before each epoch, the method
__iter__()
on the generator is called and the method__next__()
is called for each step on resulting object returned by__iter__()
. Notice that a call to__iter__()
on a generator made using the python keywordyield
returns the generator itself.valid_generator (optional) – Generatorlike object for the validation dataset. This generator is optional. The generator is used the same way as the generator
train_generator
. If the generator does not have a method__len__()
, either thevalidation_steps
or thesteps_per_epoch
argument must be provided or the iterator returned raises a StopIteration exception at the end of the validation dataset. (Default value = None)epochs (int) – Number of times the entire training dataset is seen. (Default value = 1000)
steps_per_epoch (int, optional) – Number of batch used during one epoch. Obviously, using this argument may cause one epoch not to see the entire training dataset or see it multiple times. (Defaults the number of steps needed to see the entire training dataset)
validation_steps (int, optional) – Same as for
steps_per_epoch
but for the validation dataset. (Defaults tosteps_per_epoch
if provided or the number of steps needed to see the entire validation dataset)initial_epoch (int, optional) – Epoch at which to start training (useful for resuming a previous training run). (Default value = 1)
verbose (bool) – Whether to display the progress of the training. (Default value = True)
callbacks (list of poutyne.framework.Callback) – List of callbacks that will be called during training. (Default value = [])
 Returns
List of dict containing the history of each epoch.
Example
model = Model(pytorch_module, optimizer, loss_function) history = model.fit_generator(train_generator, valid_generator, epochs=num_epochs, verbose=False) print(*history, sep="\n")
{'epoch': 1, 'loss': 1.7198852968215943, 'time': 0.019999928001197986, 'acc': 19.375, 'val_loss': 1.6674459838867188, 'val_acc': 22.0} {'epoch': 2, 'loss': 1.7054892110824584, 'time': 0.015421080999658443, 'acc': 19.75, 'val_loss': 1.660806336402893, 'val_acc': 22.0} {'epoch': 3, 'loss': 1.6923445892333984, 'time': 0.01363091799794347, 'acc': 19.625, 'val_loss': 1.6550078630447387, 'val_acc': 22.5} ...

get_weight_copies
()[source]¶ Returns a dictionary containing copies of the parameters of the network.

get_weights
()[source]¶ Returns a dictionary containing the parameters of the network. The tensors are just references to the parameters. To get copies of the weights, see the
get_weight_copies()
method.

load_optimizer_state
(f)[source]¶ Loads the optimizer state saved using the
torch.save()
method or thesave_optimizer_state()
method of this class. Parameters
f – Filelike object (has to implement fileno that returns a file descriptor) or string containing a file name.

load_weights
(f)[source]¶ Loads the weights saved using the
torch.save()
method or thesave_weights()
method of this class. Contrary totorch.load()
, the weights are not transfered to the device from which they were saved from. In other words, the PyTorch module will stay on the same device it already is on. Parameters
f – Filelike object (has to implement fileno that returns a file descriptor) or string containing a file name.

predict
(x, *, batch_size=32)[source]¶ Returns the predictions of the network given a dataset
x
, where the tensors are converted into Numpy arrays. Parameters
x (Union[Tensor, np.ndarray]) – Dataset for which to predict.
batch_size (int) – Number of samples given to the network at one time. (Default value = 32)
 Returns
Numpy arrays of the predictions.

predict_generator
(generator, *, steps=None)[source]¶ Returns the predictions of the network given batches of samples
x
, where the tensors are converted into Numpy arrays. generator: Generatorlike object for the dataset. The generator must
yield a batch of samples. See the
fit_generator()
method for details on the types of generators supported. steps (int, optional): Number of iterations done on
generator
. (Defaults the number of steps needed to see the entire dataset)
 Returns
List of the predictions of each batch with tensors converted into Numpy arrays.

predict_on_batch
(x)[source]¶ Returns the predictions of the network given a batch
x
, where the tensors are converted into Numpy arrays. Parameters
x (Union[Tensor, np.ndarray]) – Batch for which to predict.
 Returns
The predictions with tensors converted into Numpy arrays.

save_optimizer_state
(f)[source]¶ Saves the state of the current optimizer.
 Parameters
f – Filelike object (has to implement fileno that returns a file descriptor) or string containing a file name.

save_weights
(f)[source]¶ Saves the weights of the current network.
 Parameters
f – Filelike object (has to implement fileno that returns a file descriptor) or string containing a file name.

set_weights
(weights)[source]¶ Modifies the weights of the network with the given weights.
 Parameters
weights (dict) – Weights returned by either
get_weights()
orget_weight_copies()
.

to
(device)[source]¶ Tranfers the network on the specified device. The device is saved so that the batches can send to the right device before passing it to the network.
 Parameters
device (torch.device) – The device to which the network is sent.
 Returns
self.

train_on_batch
(x, y, *, return_pred=False)[source]¶ Trains the model for the batch
(x, y)
and computes the loss and the metrics, and optionaly returns the predictions. Parameters
x – Batch.
y – Batch ground truths.
return_pred (bool, optional) – Whether to return the predictions for
x
. (Default value = False)
 Returns
Float
loss
if no metrics were specified andreturn_pred
is false.Otherwise, tuple
(loss, metrics)
ifreturn_pred
is false.metrics
is a Numpy array of sizen
, wheren
is the number of metrics ifn > 1
. Ifn == 1
, thenmetrics
is a float. Ifn == 0
, themetrics
is omitted.Tuple
(loss, metrics, pred_y)
ifreturn_pred
is true wherepred_y
is the predictions with tensors converted into Numpy arrays.