Exemplo n.º 1
0
def train_fast_rcnn(debug_output=False):
    if debug_output:
        print("Storing graphs and intermediate models to %s." % os.path.join(abs_path, "Output"))

    # Create the minibatch source
    minibatch_source = create_mb_source(image_height, image_width, num_channels,
                                        num_classes, num_rois, base_path, "train")

    # Input variables denoting features, rois and label data
    image_input = input_variable((num_channels, image_height, image_width))
    roi_input   = input_variable((num_rois, 4))
    label_input = input_variable((num_rois, num_classes))

    # define mapping from reader streams to network inputs
    input_map = {
        image_input: minibatch_source[features_stream_name],
        roi_input: minibatch_source[roi_stream_name],
        label_input: minibatch_source[label_stream_name]
    }

    # Instantiate the Fast R-CNN prediction model and loss function
    frcn_output = frcn_predictor(image_input, roi_input, num_classes)
    ce = cross_entropy_with_softmax(frcn_output, label_input, axis=1)
    pe = classification_error(frcn_output, label_input, axis=1)
    if debug_output:
        plot(frcn_output, os.path.join(abs_path, "Output", "graph_frcn.png"))

    # Set learning parameters
    l2_reg_weight = 0.0005
    lr_per_sample = [0.00001] * 10 + [0.000001] * 5 + [0.0000001]
    lr_schedule = learning_rate_schedule(lr_per_sample, unit=UnitType.sample)
    mm_schedule = momentum_as_time_constant_schedule(momentum_time_constant)

    # Instantiate the trainer object
    learner = momentum_sgd(frcn_output.parameters, lr_schedule, mm_schedule, l2_regularization_weight=l2_reg_weight)
    trainer = Trainer(frcn_output, (ce, pe), learner)

    # Get minibatches of images and perform model training
    print("Training Fast R-CNN model for %s epochs." % max_epochs)
    log_number_of_parameters(frcn_output)
    progress_printer = ProgressPrinter(tag='Training', num_epochs=max_epochs)
    for epoch in range(max_epochs):       # loop over epochs
        sample_count = 0
        while sample_count < epoch_size:  # loop over minibatches in the epoch
            data = minibatch_source.next_minibatch(min(mb_size, epoch_size-sample_count), input_map=input_map)
            trainer.train_minibatch(data)                                    # update model with it
            sample_count += trainer.previous_minibatch_sample_count          # count samples processed so far
            progress_printer.update_with_trainer(trainer, with_metric=True)  # log progress

        progress_printer.epoch_summary(with_metric=True)
        if debug_output:
            frcn_output.save(os.path.join(abs_path, "Output", "frcn_py_%s.model" % (epoch+1)))

    return frcn_output
Exemplo n.º 2
0
def train_model(base_model_file, feature_node_name, last_hidden_node_name,
                image_width, image_height, num_channels, num_classes, train_map_file,
                num_epochs, max_images=-1, freeze=False):
    epoch_size = sum(1 for line in open(train_map_file))
    if max_images > 0:
        epoch_size = min(epoch_size, max_images)

    # Create the minibatch source and input variables
    minibatch_source = create_mb_source(train_map_file, image_width, image_height, num_channels, num_classes)
    image_input = input_variable((num_channels, image_height, image_width))
    label_input = input_variable(num_classes)

    # Define mapping from reader streams to network inputs
    input_map = {
        image_input: minibatch_source[features_stream_name],
        label_input: minibatch_source[label_stream_name]
    }

    # Instantiate the transfer learning model and loss function
    tl_model = create_model(base_model_file, feature_node_name, last_hidden_node_name, num_classes, image_input, freeze)
    ce = cross_entropy_with_softmax(tl_model, label_input)
    pe = classification_error(tl_model, label_input)

    # Instantiate the trainer object
    lr_schedule = learning_rate_schedule(lr_per_mb, unit=UnitType.minibatch)
    mm_schedule = momentum_schedule(momentum_per_mb)
    learner = momentum_sgd(tl_model.parameters, lr_schedule, mm_schedule, l2_regularization_weight=l2_reg_weight)
    trainer = Trainer(tl_model, (ce, pe), learner)

    # Get minibatches of images and perform model training
    print("Training transfer learning model for {0} epochs (epoch_size = {1}).".format(num_epochs, epoch_size))
    log_number_of_parameters(tl_model)
    progress_printer = ProgressPrinter(tag='Training', num_epochs=num_epochs)
    for epoch in range(num_epochs):       # loop over epochs
        sample_count = 0
        while sample_count < epoch_size:  # loop over minibatches in the epoch
            data = minibatch_source.next_minibatch(min(mb_size, epoch_size-sample_count), input_map=input_map)
            trainer.train_minibatch(data)                                    # update model with it
            sample_count += trainer.previous_minibatch_sample_count          # count samples processed so far
            progress_printer.update_with_trainer(trainer, with_metric=True)  # log progress
            if sample_count % (100 * mb_size) == 0:
                print ("Processed {0} samples".format(sample_count))

        progress_printer.epoch_summary(with_metric=True)

    return tl_model
Exemplo n.º 3
0
def main():
    # Ensure we always get the same amount of randomness
    np.random.seed(0)

    global minibatch_size, skip_window

    if len(sys.argv) < 2:
        print(
            'Insufficient number of arguments. For running the example case, run: $ python word2vec.py runexample'
        )
        exit(1)
    filename = sys.argv[1]
    process_text(filename)

    inp, label, trainer = train(emb_size, vocab_size)
    pp = ProgressPrinter(50)
    for _epoch in range(num_epochs):
        i = 0
        while curr_epoch == _epoch:
            features, labels = generate_batch(minibatch_size, skip_window)
            features = get_one_hot(features)
            labels = get_one_hot(labels)

            trainer.train_minibatch({inp: features, label: labels})
            pp.update_with_trainer(trainer)
            i += 1
            if i % 200 == 0:
                print('Saving Embeddings..')
                with open(embpickle, 'wb') as handle:
                    pickle.dump(embeddings.value, handle)

        pp.epoch_summary()

    test_features, test_labels = generate_batch(minibatch_size, skip_window)
    test_features = get_one_hot(test_features)
    test_labels = get_one_hot(test_labels)

    avg_error = trainer.test_minibatch({
        inp: test_features,
        label: test_labels
    })
    print('Avg. Error on Test Set: ', avg_error)
Exemplo n.º 4
0
def ffnet():
    input_dim = 2
    num_output_classes = 2
    num_hidden_layers = 2
    hidden_layers_dim = 50

    # Input variables denoting the features and label data
    input = input_variable((input_dim), np.float32)
    label = input_variable((num_output_classes), np.float32)

    # Instantiate the feedforward classification model
    netout = fully_connected_classifier_net(input, num_output_classes,
                                            hidden_layers_dim,
                                            num_hidden_layers, sigmoid)

    ce = cross_entropy_with_softmax(netout, label)
    pe = classification_error(netout, label)

    # Instantiate the trainer object to drive the model training
    trainer = Trainer(netout, ce, pe, [sgd(netout.parameters, lr=0.005)])

    # Get minibatches of training data and perform model training
    minibatch_size = 25

    pp = ProgressPrinter(128)
    for i in range(1024):
        features, labels = generate_random_data(minibatch_size, input_dim,
                                                num_output_classes)
        # Specify the mapping of input variables in the model to actual
        # minibatch data to be trained with
        trainer.train_minibatch({input: features, label: labels})
        pp.update_with_trainer(trainer)
    pp.epoch_summary()
    test_features, test_labels = generate_random_data(minibatch_size,
                                                      input_dim,
                                                      num_output_classes)
    avg_error = trainer.test_minibatch({
        input: test_features,
        label: test_labels
    })
    return avg_error
Exemplo n.º 5
0
def ffnet():
    input_dim = 2
    num_output_classes = 2
    num_hidden_layers = 2
    hidden_layers_dim = 50

    # Input variables denoting the features and label data
    input = input_variable((input_dim), np.float32)
    label = input_variable((num_output_classes), np.float32)

    # Instantiate the feedforward classification model
    netout = fully_connected_classifier_net(
        input, num_output_classes, hidden_layers_dim, num_hidden_layers, sigmoid)

    ce = cross_entropy_with_softmax(netout, label)
    pe = classification_error(netout, label)

    lr_per_minibatch=learning_rate_schedule(0.5, UnitType.minibatch)
    # Instantiate the trainer object to drive the model training
    trainer = Trainer(netout, ce, pe, sgd(netout.parameters, lr=lr_per_minibatch))

    # Get minibatches of training data and perform model training
    minibatch_size = 25

    pp = ProgressPrinter(128)
    for i in range(1024):
        features, labels = generate_random_data(
            minibatch_size, input_dim, num_output_classes)
        # Specify the mapping of input variables in the model to actual
        # minibatch data to be trained with
        trainer.train_minibatch({input: features, label: labels})
        pp.update_with_trainer(trainer)
    pp.epoch_summary()
    test_features, test_labels = generate_random_data(
        minibatch_size, input_dim, num_output_classes)
    avg_error = trainer.test_minibatch(
        {input: test_features, label: test_labels})
    return avg_error
Exemplo n.º 6
0
        if origlabels[t, 0] < vocab_size and origlabels[t, 0] >= 0:
            labels[t, origlabels[t, 0]] = 1.0
    return labels
#Testing & training
build_dataset()
    inp, label, trainer = train(emb_size, vocab_size)
    print('Model Creation Done.')
    pp = ProgressPrinter(50)
    for _epoch in range(num_epochs):
        i = 0
        while curr_epoch == _epoch:
            features, labels = generate_batch(minibatch_size, skip_window)
            features = get_one_hot(features)
            labels = get_one_hot(labels)

            trainer.train_minibatch({inp: features, label: labels})
            pp.update_with_trainer(trainer)
            i += 1
            if i % 200 == 0:
                print('Saving Embeddings..')
                with open(embpickle, 'wb') as handle:
                    pickle.dump(embeddings.value, handle)

        pp.epoch_summary()

    test_features, test_labels = generate_batch(minibatch_size, skip_window)
    test_features = get_one_hot(test_features)
    test_labels = get_one_hot(test_labels)

    avg_error = trainer.test_minibatch({inp: test_features, label: test_labels})
    print('Avg. Error on Test Set: ', avg_error)
Exemplo n.º 7
0
def train_model(base_model_file,
                feature_node_name,
                last_hidden_node_name,
                image_width,
                image_height,
                num_channels,
                num_classes,
                train_map_file,
                num_epochs,
                max_images=-1,
                freeze=False):
    epoch_size = sum(1 for line in open(train_map_file))
    if max_images > 0:
        epoch_size = min(epoch_size, max_images)

    # Create the minibatch source and input variables
    minibatch_source = create_mb_source(train_map_file, image_width,
                                        image_height, num_channels,
                                        num_classes)
    image_input = input_variable((num_channels, image_height, image_width))
    label_input = input_variable(num_classes)

    # Define mapping from reader streams to network inputs
    input_map = {
        image_input: minibatch_source[features_stream_name],
        label_input: minibatch_source[label_stream_name]
    }

    # Instantiate the transfer learning model and loss function
    tl_model = create_model(base_model_file, feature_node_name,
                            last_hidden_node_name, num_classes, image_input,
                            freeze)
    ce = cross_entropy_with_softmax(tl_model, label_input)
    pe = classification_error(tl_model, label_input)

    # Instantiate the trainer object
    lr_schedule = learning_rate_schedule(lr_per_mb, unit=UnitType.minibatch)
    mm_schedule = momentum_schedule(momentum_per_mb)
    learner = momentum_sgd(tl_model.parameters,
                           lr_schedule,
                           mm_schedule,
                           l2_regularization_weight=l2_reg_weight)
    trainer = Trainer(tl_model, (ce, pe), learner)

    # Get minibatches of images and perform model training
    print(
        "Training transfer learning model for {0} epochs (epoch_size = {1}).".
        format(num_epochs, epoch_size))
    log_number_of_parameters(tl_model)
    progress_printer = ProgressPrinter(tag='Training', num_epochs=num_epochs)
    for epoch in range(num_epochs):  # loop over epochs
        sample_count = 0
        while sample_count < epoch_size:  # loop over minibatches in the epoch
            data = minibatch_source.next_minibatch(min(
                mb_size, epoch_size - sample_count),
                                                   input_map=input_map)
            trainer.train_minibatch(data)  # update model with it
            sample_count += trainer.previous_minibatch_sample_count  # count samples processed so far
            progress_printer.update_with_trainer(
                trainer, with_metric=True)  # log progress
            if sample_count % (100 * mb_size) == 0:
                print("Processed {0} samples".format(sample_count))

        progress_printer.epoch_summary(with_metric=True)

    return tl_model
Exemplo n.º 8
0
def train(train_reader, valid_reader, vocab, i2w, s2smodel, max_epochs, epoch_size):

    # Note: We would like to set the signature of 's2smodel' (s2smodel.update_signature()), but that will cause
    # an error since the training criterion uses a reduced sequence axis for the labels.
    # This is because it removes the initial <s> symbol. Hence, we must leave the model
    # with unspecified input shapes and axes.

    # create the training wrapper for the s2smodel, as well as the criterion function
    model_train = create_model_train(s2smodel)
    criterion = create_criterion_function(model_train)

    # also wire in a greedy decoder so that we can properly log progress on a validation example
    # This is not used for the actual training process.
    model_greedy = create_model_greedy(s2smodel)

    # This does not need to be done in training generally though
    # Instantiate the trainer object to drive the model training
    minibatch_size = 72
    lr = 0.001 if use_attention else 0.005   # TODO: can we use the same value for both?
    learner = adam_sgd(model_train.parameters,
                       lr       = learning_rate_schedule([lr]*2+[lr/2]*3+[lr/4], UnitType.sample, epoch_size),
                       momentum = momentum_as_time_constant_schedule(1100),
                       gradient_clipping_threshold_per_sample=2.3,
                       gradient_clipping_with_truncation=True)
    trainer = Trainer(None, criterion, learner)

    # Get minibatches of sequences to train with and perform model training
    total_samples = 0
    mbs = 0
    eval_freq = 100

    # print out some useful training information
    log_number_of_parameters(model_train) ; print()
    progress_printer = ProgressPrinter(freq=30, tag='Training')
    #progress_printer = ProgressPrinter(freq=30, tag='Training', log_to_file=model_path_stem + ".log") # use this to log to file

    sparse_to_dense = create_sparse_to_dense(input_vocab_dim)

    for epoch in range(max_epochs):
        print("Saving model to '%s'" % model_path(epoch))
        s2smodel.save(model_path(epoch))

        while total_samples < (epoch+1) * epoch_size:
            # get next minibatch of training data
            mb_train = train_reader.next_minibatch(minibatch_size)
            #trainer.train_minibatch(mb_train[train_reader.streams.features], mb_train[train_reader.streams.labels])
            trainer.train_minibatch({criterion.arguments[0]: mb_train[train_reader.streams.features], criterion.arguments[1]: mb_train[train_reader.streams.labels]})

            progress_printer.update_with_trainer(trainer, with_metric=True) # log progress

            # every N MBs evaluate on a test sequence to visually show how we're doing
            if mbs % eval_freq == 0:
                mb_valid = valid_reader.next_minibatch(1)

                # run an eval on the decoder output model (i.e. don't use the groundtruth)
                e = model_greedy(mb_valid[valid_reader.streams.features])
                print(format_sequences(sparse_to_dense(mb_valid[valid_reader.streams.features]), i2w))
                print("->")
                print(format_sequences(e, i2w))

                # debugging attention
                if use_attention:
                    debug_attention(model_greedy, mb_valid[valid_reader.streams.features])

            total_samples += mb_train[train_reader.streams.labels].num_samples
            mbs += 1

        # log a summary of the stats for the epoch
        progress_printer.epoch_summary(with_metric=True)

    # done: save the final model
    print("Saving final model to '%s'" % model_path(max_epochs))
    s2smodel.save(model_path(max_epochs))
    print("%d epochs complete." % max_epochs)
Exemplo n.º 9
0
def train(train_reader, valid_reader, vocab, i2w, s2smodel, max_epochs,
          epoch_size):

    # Note: We would like to set the signature of 's2smodel' (s2smodel.update_signature()), but that will cause
    # an error since the training criterion uses a reduced sequence axis for the labels.
    # This is because it removes the initial <s> symbol. Hence, we must leave the model
    # with unspecified input shapes and axes.

    # create the training wrapper for the s2smodel, as well as the criterion function
    model_train = create_model_train(s2smodel)
    criterion = create_criterion_function(model_train)

    # also wire in a greedy decoder so that we can properly log progress on a validation example
    # This is not used for the actual training process.
    model_greedy = create_model_greedy(s2smodel)

    # This does not need to be done in training generally though
    # Instantiate the trainer object to drive the model training
    minibatch_size = 72
    lr = 0.001 if use_attention else 0.005  # TODO: can we use the same value for both?
    learner = adam_sgd(
        model_train.parameters,
        lr=learning_rate_schedule([lr] * 2 + [lr / 2] * 3 + [lr / 4],
                                  UnitType.sample, epoch_size),
        momentum=momentum_as_time_constant_schedule(1100),
        gradient_clipping_threshold_per_sample=2.3,
        gradient_clipping_with_truncation=True)
    trainer = Trainer(None, criterion, learner)

    # Get minibatches of sequences to train with and perform model training
    total_samples = 0
    mbs = 0
    eval_freq = 100

    # print out some useful training information
    log_number_of_parameters(model_train)
    print()
    progress_printer = ProgressPrinter(freq=30, tag='Training')
    #progress_printer = ProgressPrinter(freq=30, tag='Training', log_to_file=model_path_stem + ".log") # use this to log to file

    sparse_to_dense = create_sparse_to_dense(input_vocab_dim)

    for epoch in range(max_epochs):
        print("Saving model to '%s'" % model_path(epoch))
        s2smodel.save(model_path(epoch))

        while total_samples < (epoch + 1) * epoch_size:
            # get next minibatch of training data
            mb_train = train_reader.next_minibatch(minibatch_size)
            #trainer.train_minibatch(mb_train[train_reader.streams.features], mb_train[train_reader.streams.labels])
            trainer.train_minibatch({
                criterion.arguments[0]:
                mb_train[train_reader.streams.features],
                criterion.arguments[1]:
                mb_train[train_reader.streams.labels]
            })

            progress_printer.update_with_trainer(
                trainer, with_metric=True)  # log progress

            # every N MBs evaluate on a test sequence to visually show how we're doing
            if mbs % eval_freq == 0:
                mb_valid = valid_reader.next_minibatch(1)

                # run an eval on the decoder output model (i.e. don't use the groundtruth)
                e = model_greedy(mb_valid[valid_reader.streams.features])
                print(
                    format_sequences(
                        sparse_to_dense(
                            mb_valid[valid_reader.streams.features]), i2w))
                print("->")
                print(format_sequences(e, i2w))

                # debugging attention
                if use_attention:
                    debug_attention(model_greedy,
                                    mb_valid[valid_reader.streams.features])

            total_samples += mb_train[train_reader.streams.labels].num_samples
            mbs += 1

        # log a summary of the stats for the epoch
        progress_printer.epoch_summary(with_metric=True)

    # done: save the final model
    print("Saving final model to '%s'" % model_path(max_epochs))
    s2smodel.save(model_path(max_epochs))
    print("%d epochs complete." % max_epochs)