template.runner.image_classification package

Submodules

template.runner.image_classification.evaluate module

template.runner.image_classification.evaluate.evaluate(data_loader, model, criterion, writer, epoch, logging_label, no_cuda=False, log_interval=20, **kwargs)

The evaluation routine

Parameters

• data_loader (torch.utils.data.DataLoader) – The dataloader of the evaluation set

• model (torch.nn.module) – The network model being used

• criterion (torch.nn.loss) – The loss function used to compute the loss of the model

• writer (tensorboardX.writer.SummaryWriter) – The tensorboard writer object. Used to log values on file for the tensorboard visualization.

• epoch (int) – Number of the epoch (for logging purposes)

• logging_label (string) – Label for logging purposes. Typically ‘test’ or ‘valid’. Its prepended to the logging output path and messages.

• no_cuda (boolean) – Specifies whether the GPU should be used or not. A value of ‘True’ means the CPU will be used.

• log_interval (int) – Interval limiting the logging of mini-batches. Default value of 10.

Returns

top1.avg – Accuracy of the model of the evaluated split

Return type

float

template.runner.image_classification.image_classification module

This file is the template for the boilerplate of train/test of a DNN for image classification

There are a lot of parameter which can be specified to modify the behaviour and they should be used instead of hard-coding stuff.

class template.runner.image_classification.image_classification.ImageClassification

Bases: object

classmethod prepare(model_name, **kwargs)

Loads and prepares the data, the optimizer and the criterion

Parameters

• model_name (str) – Name of the model. Used for loading the model.

• kwargs (dict) – Any additional arguments.

Returns

• model (DataParallel) – The model to train

• num_classes (int) – How many different classes there are in our problem. Used for loading the model.

• best_value (float) – Best value of the model so far. Non-zero only in case of –resume being used

• train_loader (torch.utils.data.dataloader.DataLoader) – Training dataloader

• val_loader (torch.utils.data.dataloader.DataLoader) – Validation dataloader

• test_loader (torch.utils.data.dataloader.DataLoader) – Test set dataloader

• optimizer (torch.optim) – Optimizer to use during training, e.g. SGD

• criterion (torch.nn.modules.loss) – Loss function to use, e.g. cross-entropy

classmethod single_run(**kwargs)

This is the main routine where train(), validate() and test() are called.

Returns

• train_value (ndarray[floats] of size (1, epochs)) – Accuracy values for train split

• val_value (ndarray[floats] of size (1, `epochs`+1)) – Accuracy values for validation split

• test_value (float) – Accuracy value for test split

classmethod test_routine(model_name, num_classes, criterion, epochs, current_log_folder, writer, **kwargs)

Load the best model according to the validation score (early stopping) and runs the test routine.

Parameters

• model_name (str) – name of the model. Used for loading the model.

• num_classes (int) – How many different classes there are in our problem. Used for loading the model.

• criterion (torch.nn.modules.loss) – Loss function to use, e.g. cross-entropy

• epochs (int) – After how many epochs are we testing

• current_log_folder (string) – Path to where logs/checkpoints are saved

• writer (Tensorboard.SummaryWriter) – Responsible for writing logs in Tensorboard compatible format.

• kwargs (dict) – Any additional arguments.

Returns

test_value – Accuracy value for test split

Return type

float

classmethod train_routine(best_value, decay_lr, validation_interval, start_epoch, epochs, checkpoint_all_epochs, current_log_folder, **kwargs)

Performs the training and validatation routines

Parameters

• best_value (float) – Best value of the model so far. Non-zero only in case of –resume being used

• decay_lr (boolean) – Decay the lr flag

• validation_interval (int) – Run evaluation on validation set every N epochs

• start_epoch (int) – Int to initialize the starting epoch. Non-zero only in case of –resume being used

• epochs (int) – Number of epochs to train

• checkpoint_all_epochs (bool) – Save checkpoint at each epoch

• current_log_folder (string) – Path to where logs/checkpoints are saved

• kwargs (dict) – Any additional arguments.

Returns

• train_value (ndarray[floats] of size (1, epochs)) – Accuracy values for train split

• val_value (ndarray[floats] of size (1, `epochs`+1)) – Accuracy values for validation split

template.runner.image_classification.train module

template.runner.image_classification.train.train(train_loader, model, criterion, optimizer, writer, epoch, no_cuda=False, log_interval=25, **kwargs)

Training routine

Parameters

• train_loader (torch.utils.data.DataLoader) – The dataloader of the train set.

• model (torch.nn.module) – The network model being used.

• criterion (torch.nn.loss) – The loss function used to compute the loss of the model.

• optimizer (torch.optim) – The optimizer used to perform the weight update.

• writer (tensorboardX.writer.SummaryWriter) – The tensorboard writer object. Used to log values on file for the tensorboard visualization.

• epoch (int) – Number of the epoch (for logging purposes).

• no_cuda (boolean) – Specifies whether the GPU should be used or not. A value of ‘True’ means the CPU will be used.

• log_interval (int) – Interval limiting the logging of mini-batches. Default value of 10.

Returns

top1.avg – Accuracy of the model of the evaluated split

Return type

float

template.runner.image_classification.train.train_one_mini_batch(model, criterion, optimizer, input, target, loss_meter, acc_meter)

This routing train the model passed as parameter for one mini-batch

Parameters

• model (torch.nn.module) – The network model being used.

• criterion (torch.nn.loss) – The loss function used to compute the loss of the model.

• optimizer (torch.optim) – The optimizer used to perform the weight update.

• input (torch.autograd.Variable) – The input data for the mini-batch

• target (torch.autograd.Variable) – The target data (labels) for the mini-batch

• loss_meter (AverageMeter) – Tracker for the overall loss

• acc_meter (AverageMeter) – Tracker for the overall accuracy

Returns

• acc (float) – Accuracy for this mini-batch

• loss (float) – Loss for this mini-batch

Module contents

class template.runner.image_classification.ImageClassification

Bases: object

classmethod prepare(model_name, **kwargs)

Loads and prepares the data, the optimizer and the criterion

Parameters

• model_name (str) – Name of the model. Used for loading the model.

• kwargs (dict) – Any additional arguments.

Returns

• model (DataParallel) – The model to train

• num_classes (int) – How many different classes there are in our problem. Used for loading the model.

• best_value (float) – Best value of the model so far. Non-zero only in case of –resume being used

• train_loader (torch.utils.data.dataloader.DataLoader) – Training dataloader

• val_loader (torch.utils.data.dataloader.DataLoader) – Validation dataloader

• test_loader (torch.utils.data.dataloader.DataLoader) – Test set dataloader

• optimizer (torch.optim) – Optimizer to use during training, e.g. SGD

• criterion (torch.nn.modules.loss) – Loss function to use, e.g. cross-entropy

classmethod single_run(**kwargs)

This is the main routine where train(), validate() and test() are called.

Returns

• train_value (ndarray[floats] of size (1, epochs)) – Accuracy values for train split

• val_value (ndarray[floats] of size (1, `epochs`+1)) – Accuracy values for validation split

• test_value (float) – Accuracy value for test split

classmethod test_routine(model_name, num_classes, criterion, epochs, current_log_folder, writer, **kwargs)

Load the best model according to the validation score (early stopping) and runs the test routine.

Parameters

• model_name (str) – name of the model. Used for loading the model.

• num_classes (int) – How many different classes there are in our problem. Used for loading the model.

• criterion (torch.nn.modules.loss) – Loss function to use, e.g. cross-entropy

• epochs (int) – After how many epochs are we testing

• current_log_folder (string) – Path to where logs/checkpoints are saved

• writer (Tensorboard.SummaryWriter) – Responsible for writing logs in Tensorboard compatible format.

• kwargs (dict) – Any additional arguments.

Returns

test_value – Accuracy value for test split

Return type

float

classmethod train_routine(best_value, decay_lr, validation_interval, start_epoch, epochs, checkpoint_all_epochs, current_log_folder, **kwargs)

Performs the training and validatation routines

Parameters

• best_value (float) – Best value of the model so far. Non-zero only in case of –resume being used

• decay_lr (boolean) – Decay the lr flag

• validation_interval (int) – Run evaluation on validation set every N epochs

• start_epoch (int) – Int to initialize the starting epoch. Non-zero only in case of –resume being used

• epochs (int) – Number of epochs to train

• checkpoint_all_epochs (bool) – Save checkpoint at each epoch

• current_log_folder (string) – Path to where logs/checkpoints are saved

• kwargs (dict) – Any additional arguments.

Returns

• train_value (ndarray[floats] of size (1, epochs)) – Accuracy values for train split

• val_value (ndarray[floats] of size (1, `epochs`+1)) – Accuracy values for validation split