CNN Experiments on muvr data

In [1]:

    

%matplotlib inline

import logging

logging.basicConfig(level=10)
logger = logging.getLogger()

import shutil
from os import remove
import cPickle as pkl
from os.path import expanduser, exists


    

In [2]:

    

# Dataset creation

import numpy as np
import math
import random
import csv
from neon.datasets.dataset import Dataset

class WorkoutDS(Dataset):
    # Number of features per example
    feature_count = None

    # Number of examples
    num_train_examples = None
    num_test_examples = None

    # Number of classes
    num_labels = None
    
    # Indicator if the data has been loaded yet
    initialized = False
    
    # Mapping of integer class labels to strings
    human_labels = {}
    
    def human_label_for(self, id):
        return self.human_labels[id]
    
    # Convert an integer representation to a one-hot-vector
    def as_one_hot(self, i, n):
        v = np.zeros(n)
        v[i] = 1
        return v
    
    # Convert an one-hot-vector to an integer representation
    def as_int_rep(self, oh):
        return np.where(oh == 1)[0][0]
    
    # Loads a label mapping from file. The file should contain a CSV tabel mapping integer labels to human readable
    # labels. Integer class labels should start with 1
    def load_label_mapping(self, filename):
        with open(expanduser(filename), 'rb') as csvfile:
            dialect = csv.Sniffer().sniff(csvfile.read(1024))
            csvfile.seek(0)
            csv_data = csv.reader(csvfile, dialect)
            next(csv_data, None)  # skip the headers
            label_mapping = {}
            for row in csv_data:
                # We need to offset the labels by one since counting starts at 0 in python...
                label_mapping[int(row[0]) - 1] = row[1]
        return label_mapping
                
    # Load examples from given CSV file. The dataset should already be splitted into test and train externally
    def load_examples(self, filename):
        with open(expanduser(filename), 'rb') as csvfile:
            dialect = csv.Sniffer().sniff(csvfile.read(1024))
            csvfile.seek(0)
            csv_data = csv.reader(csvfile, dialect)
            next(csv_data, None)  # skip the headers
            y = []
            X = []
            for row in csv_data:
                label = int(row[2]) - 1
                y.append(self.as_one_hot(label, self.num_labels))
                X.append(map(int, row[3:]))
                
            X = np.reshape(np.asarray(X, dtype = float), (len(X), len(X[0])))  
            y = np.reshape(np.asarray(y, dtype = float), (X.shape[0], self.num_labels))
            
            return X,y
    
    # Load label mapping and train / test data from disk.
    def initialize(self):
        logger.info("Loading DS from files...")
        self.human_labels = self.load_label_mapping(expanduser('~/data/labeled_exercise_data_f400_LABELS.csv'))
        self.num_labels = len(self.human_labels)
        
        X_train, y_train = self.load_examples(expanduser('~/data/labeled_exercise_data_f400_TRAIN.csv'))
        X_test, y_test = self.load_examples(expanduser('~/data/labeled_exercise_data_f400_TEST.csv'))
        
        self.num_train_examples = X_train.shape[0]
        self.num_test_examples = X_test.shape[0]
        self.feature_count = X_train.shape[1]
        self.X_train = X_train
        self.y_train = y_train
        self.X_test = X_test
        self.y_test = y_test
        self.initialized = True

    # Get the dataset ready for Neon training
    def load(self, **kwargs):
        if not self.initialized:
            self.initialize()

        # Assign training and test datasets
        # INFO: This assumes the data is already shuffeled! Make sure it is!
        self.inputs['train'] = self.X_train
        self.targets['train'] = self.y_train

        self.inputs['test'] =  self.X_test
        self.targets['test'] = self.y_test

        self.format()
        
dataset = WorkoutDS()
dataset.initialize()
print "Number of training examples:", dataset.num_train_examples
print "Number of test examples:", dataset.num_test_examples
print "Number of features:", dataset.feature_count
print "Number of labels:", dataset.num_labels


    

    




Number of training examples: 22579
Number of test examples: 5503
Number of features: 1200
Number of labels: 29

In [4]:

    

from ipy_table import *
from operator import itemgetter
import numpy as np

train_dist = np.reshape(np.transpose(np.sum(dataset.y_train, axis=0)), (dataset.num_labels,1))
test_dist = np.reshape(np.transpose(np.sum(dataset.y_test, axis=0)), (dataset.num_labels,1))

train_ratio = train_dist / dataset.num_train_examples
test_ratio = test_dist / dataset.num_test_examples

# Fiddle around to get it into table shape
table = np.hstack((np.zeros((dataset.num_labels,1), dtype=int), train_dist, train_ratio, test_dist, test_ratio))
table = np.vstack((np.zeros((1, 5), dtype=int), table)).tolist()

human_labels = map(dataset.human_label_for, range(0,dataset.num_labels))

for i,s in enumerate(human_labels):
    table[i + 1][0] = s
    
table.sort(lambda x,y: cmp(x[1], y[1]))

table[0][0] = ""
table[0][1] = "Train"
table[0][2] = "Train %"
table[0][3] = "Test"
table[0][4] = "Test %"

make_table(table)
set_global_style(float_format='%0.0f', align="center")
set_column_style(2, float_format='%0.2f%%')
set_column_style(4, float_format='%0.2f%%')
set_column_style(0, align="left")


    

    
Out[4]:




Train
Train %
Test
Test %
dumbbell-biceps-curl
0
0.00%
141
0.03%
rope-biceps-curl
74
0.00%
1
0.00%
barbell-pullup
136
0.01%
68
0.01%
straight-bar-triceps-extension
196
0.01%
1
0.00%
side-dips
257
0.01%
117
0.02%
alt-dumbbell-biceps-curl
319
0.01%
265
0.05%
dumbbell-row
348
0.02%
1
0.00%
barbell-biceps-curl
356
0.02%
55
0.01%
dumbbell-front-rise
379
0.02%
95
0.02%
lateral-raise
395
0.02%
57
0.01%
cable-deltoid-cross-overs
418
0.02%
312
0.06%
triceps-dips
430
0.02%
233
0.04%
dumbbell-side-rise
442
0.02%
0
0.00%
dumbbell-press
453
0.02%
412
0.07%
straight-bar-biceps-curl
528
0.02%
373
0.07%
rope-triceps-extension
621
0.03%
148
0.03%
barbell-press
775
0.03%
252
0.05%
lat-pulldown-straight
807
0.04%
172
0.03%
lat-pulldown-angled
828
0.04%
200
0.04%
deltoid-row
831
0.04%
0
0.00%
overhead-pull
907
0.04%
0
0.00%
angle-chest-press
950
0.04%
454
0.08%
leverage-high-row
1016
0.04%
0
0.00%
rowing
1175
0.05%
0
0.00%
twist
1350
0.06%
478
0.09%
cable-cross-overs
1374
0.06%
209
0.04%
pulldown-crunch
1760
0.08%
680
0.12%
dumbbell-chest-press
2413
0.11%
295
0.05%
dumbbell-chest-fly
3041
0.13%
484
0.09%

In [5]:

    

from matplotlib import pyplot, cm
from pylab import *

# Choose some random examples to plot from the training data
number_of_examples_to_plot = 3
plot_ids = np.random.random_integers(0, dataset.num_train_examples - 1, number_of_examples_to_plot)

print "Ids of plotted examples:",plot_ids

# Retrieve a human readable label given the idx of an example
def label_of_example(i):
    label_id = np.where(dataset.y_train[i] == 1)[0][0]
    return dataset.human_label_for(label_id)

figure(figsize=(20,10))
ax1 = subplot(311)
setp(ax1.get_xticklabels(), visible=False)
ax1.set_ylabel('X - Acceleration')

ax2 = subplot(312, sharex=ax1)
setp(ax2.get_xticklabels(), visible=False)
ax2.set_ylabel('Y - Acceleration')

ax3 = subplot(313, sharex=ax1)
ax3.set_ylabel('Z - Acceleration')

for i in plot_ids:
    c = np.random.random((3,))

    ax1.plot(range(0, dataset.feature_count / 3), dataset.X_train[i,0:400], '-o', c=c)
    ax2.plot(range(0, dataset.feature_count / 3), dataset.X_train[i,400:800], '-o', c=c)
    ax3.plot(range(0, dataset.feature_count / 3), dataset.X_train[i,800:1200], '-o', c=c)
    
legend(map(label_of_example, plot_ids))
suptitle('Feature values for the first three training examples', fontsize=16)
xlabel('Time')
show()


    

    




Ids of plotted examples: [ 4726  2349 13364]






    

In [55]:

    

from neon.backends import gen_backend
from neon.layers import *
from neon.models import MLP
from neon.transforms import RectLin, Tanh, Logistic, CrossEntropy
from neon.experiments import FitPredictErrorExperiment
from neon.params import val_init
from neon.util.persist import serialize

# General settings
max_epochs = 75
epoch_step_size = 1
batch_size = 30 # max(10, min(100, dataset.num_train_examples/10))
random_seed = 42 # Take your lucky number


# Storage director of the model and its snapshots
file_path = expanduser('~/data/workout-cnn/workout-cnn.prm')
#if exists(file_path):
#    remove(file_path)

# Captured errors for the different epochs
train_err = []
test_err = []

print 'Epochs: %d Batch-Size: %d' % (max_epochs, batch_size)

# Generate layers and a MLP model using the given settings
def model_gen(lrate, momentum_coef, num_epochs, batch_size):
    layers = []

    lrule = {
        'lr_params': {
            'learning_rate': lrate,
            'momentum_params': {
                'coef': momentum_coef,
                'type': 'constant'
            }},
        'type': 'gradient_descent_momentum'
    }

    weight_init = val_init.UniformValGen(low=-0.1, high=0.1)

    layers.append(DataLayer(
        nofm=3,
        ofmshape=[400,1],
        is_local=True
    ))
    
    layers.append(ConvLayer(
        name="cv_1",
        nofm=16,
        fshape = [5,1],
        stride = 1,
        lrule_init=lrule,
        weight_init= weight_init,
        activation=RectLin()
    ))

    layers.append(PoolingLayer(
        name="po_1",
        op="max",
        fshape=[2,1],
        stride=2,
    ))

    layers.append(ConvLayer(
        name="cv_2",
        nofm=32,
        fshape = [5,1],
        stride = 1,
        lrule_init=lrule,
        weight_init= weight_init,
        activation=RectLin()
    ))

    layers.append(PoolingLayer(
        name="po_2",
        op="max",
        fshape=[2,1],
        stride=2,
    ))
    
    layers.append(DropOutLayer(
            name="do_1",
            keep = 0.9
        )
    )

    layers.append(FCLayer(
        name="fc_1",
        nout=100,
        lrule_init=lrule,
        weight_init=weight_init,
        activation=RectLin()
    ))
    
    layers.append(DropOutLayer(
            name="do_2",
            keep = 0.9
        )
    )

    layers.append(FCLayer(
        name="fc_2",
        nout=dataset.num_labels,
        lrule_init=lrule,
        weight_init=weight_init,
        activation = Logistic()
    ))

    layers.append(CostLayer(
        name='cost',
        ref_layer=layers[0],
        cost=CrossEntropy()
    ))

    model = MLP(num_epochs=num_epochs, batch_size=batch_size, layers=layers, serialized_path=file_path)
    return model

# Set logging output...
for name in ["neon.util.persist", "neon.datasets.dataset", "neon.models.mlp"]:
    dslogger = logging.getLogger(name)
    dslogger.setLevel(20)

print "Starting training..."
for num_epochs in range(26,max_epochs+1, epoch_step_size):

    if num_epochs > 230:
        lrate = 0.0000003
    elif num_epochs > 60:
        lrate = 0.00001
    elif num_epochs > 40:
        lrate = 0.00003
    else:
        lrate = 0.0001

    # set up the model and experiment
    model = model_gen(lrate = lrate,
                      momentum_coef = 0.9,
                      num_epochs = num_epochs,
                      batch_size = batch_size)

    # Uncomment line below to run on CPU backend
    backend = gen_backend(rng_seed=random_seed)
    # Uncomment line below to run on GPU using cudanet backend
    # backend = gen_backend(rng_seed=0, gpu='cudanet')
    experiment = FitPredictErrorExperiment(model=model,
                                           backend=backend,
                                           dataset=dataset)

    # Run the training, and dump weights
    dest_path = expanduser('~/data/workout-cnn/workout-ep' + str(num_epochs) + '.prm')
    if num_epochs > 0:
        res = experiment.run()
        train_err.append(res['train']['MisclassPercentage_TOP_1'])
        test_err.append(res['test']['MisclassPercentage_TOP_1'])
        # Save the weights at this epoch
        shutil.copy2(file_path, dest_path)
        print "Finished epoch " + str(num_epochs)
    else:
        model.epochs_complete = 0
        serialize(model.get_params(), dest_path)

print "Finished training!"


    

    




INFO:neon.models.mlp:Layers:
	DataLayer layer: 3 x (400 x 1) nodes
	ConvLayer cv_1: 3 x (400 x 1) inputs, 16 x (396 x 1) nodes, RectLin act_fn
	PoolingLayer po_1: 16 x (396 x 1) inputs, 16 x (198 x 1) nodes, Linear act_fn
	ConvLayer cv_2: 16 x (198 x 1) inputs, 32 x (194 x 1) nodes, RectLin act_fn
	PoolingLayer po_2: 32 x (194 x 1) inputs, 32 x (97 x 1) nodes, Linear act_fn
	DropOutLayer do_1: 32 x (97 x 1) inputs, 32 x (97 x 1) nodes, Linear act_fn
	FCLayer fc_1: 3104 inputs, 100 nodes, RectLin act_fn
	DropOutLayer do_2: 100 inputs, 100 nodes, Linear act_fn
	FCLayer fc_2: 100 inputs, 29 nodes, Logistic act_fn
	CostLayer cost: 29 nodes, CrossEntropy cost_fn

WARNING:neon.datasets.dataset:Incompatible batch size. Discarding 13 samples...
WARNING:neon.datasets.dataset:Incompatible batch size. Discarding 19 samples...
WARNING:neon.datasets.dataset:Incompatible batch size. Discarding 13 samples...
WARNING:neon.datasets.dataset:Incompatible batch size. Discarding 19 samples...
WARNING:neon.util.persist:deserializing object from:  /Users/tombocklisch/data/workout-cnn/workout-cnn.prm
INFO:neon.models.mlp:commencing model fitting
INFO:neon.models.mlp:epoch: 26, training error: 0.18879
WARNING:neon.util.persist:serializing object to: /Users/tombocklisch/data/workout-cnn/workout-cnn.prm






    




Epochs: 75 Batch-Size: 30
Starting training...
Finished epoch 26





    




INFO:neon.models.mlp:Layers:
	DataLayer layer: 3 x (400 x 1) nodes
	ConvLayer cv_1: 3 x (400 x 1) inputs, 16 x (396 x 1) nodes, RectLin act_fn
	PoolingLayer po_1: 16 x (396 x 1) inputs, 16 x (198 x 1) nodes, Linear act_fn
	ConvLayer cv_2: 16 x (198 x 1) inputs, 32 x (194 x 1) nodes, RectLin act_fn
	PoolingLayer po_2: 32 x (194 x 1) inputs, 32 x (97 x 1) nodes, Linear act_fn
	DropOutLayer do_1: 32 x (97 x 1) inputs, 32 x (97 x 1) nodes, Linear act_fn
	FCLayer fc_1: 3104 inputs, 100 nodes, RectLin act_fn
	DropOutLayer do_2: 100 inputs, 100 nodes, Linear act_fn
	FCLayer fc_2: 100 inputs, 29 nodes, Logistic act_fn
	CostLayer cost: 29 nodes, CrossEntropy cost_fn

WARNING:neon.datasets.dataset:Incompatible batch size. Discarding 13 samples...
WARNING:neon.datasets.dataset:Incompatible batch size. Discarding 19 samples...
WARNING:neon.datasets.dataset:Incompatible batch size. Discarding 13 samples...
WARNING:neon.datasets.dataset:Incompatible batch size. Discarding 19 samples...
WARNING:neon.util.persist:deserializing object from:  /Users/tombocklisch/data/workout-cnn/workout-cnn.prm
INFO:neon.models.mlp:commencing model fitting
INFO:neon.models.mlp:epoch: 27, training error: 0.17689
WARNING:neon.util.persist:serializing object to: /Users/tombocklisch/data/workout-cnn/workout-cnn.prm






    




Finished epoch 27





    




INFO:neon.models.mlp:Layers:
	DataLayer layer: 3 x (400 x 1) nodes
	ConvLayer cv_1: 3 x (400 x 1) inputs, 16 x (396 x 1) nodes, RectLin act_fn
	PoolingLayer po_1: 16 x (396 x 1) inputs, 16 x (198 x 1) nodes, Linear act_fn
	ConvLayer cv_2: 16 x (198 x 1) inputs, 32 x (194 x 1) nodes, RectLin act_fn
	PoolingLayer po_2: 32 x (194 x 1) inputs, 32 x (97 x 1) nodes, Linear act_fn
	DropOutLayer do_1: 32 x (97 x 1) inputs, 32 x (97 x 1) nodes, Linear act_fn
	FCLayer fc_1: 3104 inputs, 100 nodes, RectLin act_fn
	DropOutLayer do_2: 100 inputs, 100 nodes, Linear act_fn
	FCLayer fc_2: 100 inputs, 29 nodes, Logistic act_fn
	CostLayer cost: 29 nodes, CrossEntropy cost_fn

WARNING:neon.datasets.dataset:Incompatible batch size. Discarding 13 samples...
WARNING:neon.datasets.dataset:Incompatible batch size. Discarding 19 samples...
WARNING:neon.datasets.dataset:Incompatible batch size. Discarding 13 samples...
WARNING:neon.datasets.dataset:Incompatible batch size. Discarding 19 samples...
WARNING:neon.util.persist:deserializing object from:  /Users/tombocklisch/data/workout-cnn/workout-cnn.prm
INFO:neon.models.mlp:commencing model fitting
INFO:neon.models.mlp:epoch: 28, training error: 0.16732
WARNING:neon.util.persist:serializing object to: /Users/tombocklisch/data/workout-cnn/workout-cnn.prm






    




Finished epoch 28





    




INFO:neon.models.mlp:Layers:
	DataLayer layer: 3 x (400 x 1) nodes
	ConvLayer cv_1: 3 x (400 x 1) inputs, 16 x (396 x 1) nodes, RectLin act_fn
	PoolingLayer po_1: 16 x (396 x 1) inputs, 16 x (198 x 1) nodes, Linear act_fn
	ConvLayer cv_2: 16 x (198 x 1) inputs, 32 x (194 x 1) nodes, RectLin act_fn
	PoolingLayer po_2: 32 x (194 x 1) inputs, 32 x (97 x 1) nodes, Linear act_fn
	DropOutLayer do_1: 32 x (97 x 1) inputs, 32 x (97 x 1) nodes, Linear act_fn
	FCLayer fc_1: 3104 inputs, 100 nodes, RectLin act_fn
	DropOutLayer do_2: 100 inputs, 100 nodes, Linear act_fn
	FCLayer fc_2: 100 inputs, 29 nodes, Logistic act_fn
	CostLayer cost: 29 nodes, CrossEntropy cost_fn

WARNING:neon.datasets.dataset:Incompatible batch size. Discarding 13 samples...
WARNING:neon.datasets.dataset:Incompatible batch size. Discarding 19 samples...
WARNING:neon.datasets.dataset:Incompatible batch size. Discarding 13 samples...
WARNING:neon.datasets.dataset:Incompatible batch size. Discarding 19 samples...
WARNING:neon.util.persist:deserializing object from:  /Users/tombocklisch/data/workout-cnn/workout-cnn.prm
INFO:neon.models.mlp:commencing model fitting
INFO:neon.models.mlp:epoch: 29, training error: 0.16821
WARNING:neon.util.persist:serializing object to: /Users/tombocklisch/data/workout-cnn/workout-cnn.prm






    




Finished epoch 29





    




INFO:neon.models.mlp:Layers:
	DataLayer layer: 3 x (400 x 1) nodes
	ConvLayer cv_1: 3 x (400 x 1) inputs, 16 x (396 x 1) nodes, RectLin act_fn
	PoolingLayer po_1: 16 x (396 x 1) inputs, 16 x (198 x 1) nodes, Linear act_fn
	ConvLayer cv_2: 16 x (198 x 1) inputs, 32 x (194 x 1) nodes, RectLin act_fn
	PoolingLayer po_2: 32 x (194 x 1) inputs, 32 x (97 x 1) nodes, Linear act_fn
	DropOutLayer do_1: 32 x (97 x 1) inputs, 32 x (97 x 1) nodes, Linear act_fn
	FCLayer fc_1: 3104 inputs, 100 nodes, RectLin act_fn
	DropOutLayer do_2: 100 inputs, 100 nodes, Linear act_fn
	FCLayer fc_2: 100 inputs, 29 nodes, Logistic act_fn
	CostLayer cost: 29 nodes, CrossEntropy cost_fn

WARNING:neon.datasets.dataset:Incompatible batch size. Discarding 13 samples...
WARNING:neon.datasets.dataset:Incompatible batch size. Discarding 19 samples...
WARNING:neon.datasets.dataset:Incompatible batch size. Discarding 13 samples...
WARNING:neon.datasets.dataset:Incompatible batch size. Discarding 19 samples...
WARNING:neon.util.persist:deserializing object from:  /Users/tombocklisch/data/workout-cnn/workout-cnn.prm
INFO:neon.models.mlp:commencing model fitting
INFO:neon.models.mlp:epoch: 30, training error: 0.15454
WARNING:neon.util.persist:serializing object to: /Users/tombocklisch/data/workout-cnn/workout-cnn.prm






    




Finished epoch 30





    




INFO:neon.models.mlp:Layers:
	DataLayer layer: 3 x (400 x 1) nodes
	ConvLayer cv_1: 3 x (400 x 1) inputs, 16 x (396 x 1) nodes, RectLin act_fn
	PoolingLayer po_1: 16 x (396 x 1) inputs, 16 x (198 x 1) nodes, Linear act_fn
	ConvLayer cv_2: 16 x (198 x 1) inputs, 32 x (194 x 1) nodes, RectLin act_fn
	PoolingLayer po_2: 32 x (194 x 1) inputs, 32 x (97 x 1) nodes, Linear act_fn
	DropOutLayer do_1: 32 x (97 x 1) inputs, 32 x (97 x 1) nodes, Linear act_fn
	FCLayer fc_1: 3104 inputs, 100 nodes, RectLin act_fn
	DropOutLayer do_2: 100 inputs, 100 nodes, Linear act_fn
	FCLayer fc_2: 100 inputs, 29 nodes, Logistic act_fn
	CostLayer cost: 29 nodes, CrossEntropy cost_fn

WARNING:neon.datasets.dataset:Incompatible batch size. Discarding 13 samples...
WARNING:neon.datasets.dataset:Incompatible batch size. Discarding 19 samples...
WARNING:neon.datasets.dataset:Incompatible batch size. Discarding 13 samples...
WARNING:neon.datasets.dataset:Incompatible batch size. Discarding 19 samples...
WARNING:neon.util.persist:deserializing object from:  /Users/tombocklisch/data/workout-cnn/workout-cnn.prm
INFO:neon.models.mlp:commencing model fitting
INFO:neon.models.mlp:epoch: 31, training error: 0.14773
WARNING:neon.util.persist:serializing object to: /Users/tombocklisch/data/workout-cnn/workout-cnn.prm






    




Finished epoch 31





    




INFO:neon.models.mlp:Layers:
	DataLayer layer: 3 x (400 x 1) nodes
	ConvLayer cv_1: 3 x (400 x 1) inputs, 16 x (396 x 1) nodes, RectLin act_fn
	PoolingLayer po_1: 16 x (396 x 1) inputs, 16 x (198 x 1) nodes, Linear act_fn
	ConvLayer cv_2: 16 x (198 x 1) inputs, 32 x (194 x 1) nodes, RectLin act_fn
	PoolingLayer po_2: 32 x (194 x 1) inputs, 32 x (97 x 1) nodes, Linear act_fn
	DropOutLayer do_1: 32 x (97 x 1) inputs, 32 x (97 x 1) nodes, Linear act_fn
	FCLayer fc_1: 3104 inputs, 100 nodes, RectLin act_fn
	DropOutLayer do_2: 100 inputs, 100 nodes, Linear act_fn
	FCLayer fc_2: 100 inputs, 29 nodes, Logistic act_fn
	CostLayer cost: 29 nodes, CrossEntropy cost_fn

WARNING:neon.datasets.dataset:Incompatible batch size. Discarding 13 samples...
WARNING:neon.datasets.dataset:Incompatible batch size. Discarding 19 samples...
WARNING:neon.datasets.dataset:Incompatible batch size. Discarding 13 samples...
WARNING:neon.datasets.dataset:Incompatible batch size. Discarding 19 samples...
WARNING:neon.util.persist:deserializing object from:  /Users/tombocklisch/data/workout-cnn/workout-cnn.prm
INFO:neon.models.mlp:commencing model fitting
INFO:neon.models.mlp:epoch: 32, training error: 0.14312
WARNING:neon.util.persist:serializing object to: /Users/tombocklisch/data/workout-cnn/workout-cnn.prm






    




Finished epoch 32





    




INFO:neon.models.mlp:Layers:
	DataLayer layer: 3 x (400 x 1) nodes
	ConvLayer cv_1: 3 x (400 x 1) inputs, 16 x (396 x 1) nodes, RectLin act_fn
	PoolingLayer po_1: 16 x (396 x 1) inputs, 16 x (198 x 1) nodes, Linear act_fn
	ConvLayer cv_2: 16 x (198 x 1) inputs, 32 x (194 x 1) nodes, RectLin act_fn
	PoolingLayer po_2: 32 x (194 x 1) inputs, 32 x (97 x 1) nodes, Linear act_fn
	DropOutLayer do_1: 32 x (97 x 1) inputs, 32 x (97 x 1) nodes, Linear act_fn
	FCLayer fc_1: 3104 inputs, 100 nodes, RectLin act_fn
	DropOutLayer do_2: 100 inputs, 100 nodes, Linear act_fn
	FCLayer fc_2: 100 inputs, 29 nodes, Logistic act_fn
	CostLayer cost: 29 nodes, CrossEntropy cost_fn

WARNING:neon.datasets.dataset:Incompatible batch size. Discarding 13 samples...
WARNING:neon.datasets.dataset:Incompatible batch size. Discarding 19 samples...
WARNING:neon.datasets.dataset:Incompatible batch size. Discarding 13 samples...
WARNING:neon.datasets.dataset:Incompatible batch size. Discarding 19 samples...
WARNING:neon.util.persist:deserializing object from:  /Users/tombocklisch/data/workout-cnn/workout-cnn.prm
INFO:neon.models.mlp:commencing model fitting
INFO:neon.models.mlp:epoch: 33, training error: 0.13829
WARNING:neon.util.persist:serializing object to: /Users/tombocklisch/data/workout-cnn/workout-cnn.prm






    




Finished epoch 33





    




INFO:neon.models.mlp:Layers:
	DataLayer layer: 3 x (400 x 1) nodes
	ConvLayer cv_1: 3 x (400 x 1) inputs, 16 x (396 x 1) nodes, RectLin act_fn
	PoolingLayer po_1: 16 x (396 x 1) inputs, 16 x (198 x 1) nodes, Linear act_fn
	ConvLayer cv_2: 16 x (198 x 1) inputs, 32 x (194 x 1) nodes, RectLin act_fn
	PoolingLayer po_2: 32 x (194 x 1) inputs, 32 x (97 x 1) nodes, Linear act_fn
	DropOutLayer do_1: 32 x (97 x 1) inputs, 32 x (97 x 1) nodes, Linear act_fn
	FCLayer fc_1: 3104 inputs, 100 nodes, RectLin act_fn
	DropOutLayer do_2: 100 inputs, 100 nodes, Linear act_fn
	FCLayer fc_2: 100 inputs, 29 nodes, Logistic act_fn
	CostLayer cost: 29 nodes, CrossEntropy cost_fn

WARNING:neon.datasets.dataset:Incompatible batch size. Discarding 13 samples...
WARNING:neon.datasets.dataset:Incompatible batch size. Discarding 19 samples...
WARNING:neon.datasets.dataset:Incompatible batch size. Discarding 13 samples...
WARNING:neon.datasets.dataset:Incompatible batch size. Discarding 19 samples...
WARNING:neon.util.persist:deserializing object from:  /Users/tombocklisch/data/workout-cnn/workout-cnn.prm
INFO:neon.models.mlp:commencing model fitting
INFO:neon.models.mlp:epoch: 34, training error: 0.13373
WARNING:neon.util.persist:serializing object to: /Users/tombocklisch/data/workout-cnn/workout-cnn.prm






    




Finished epoch 34





    




INFO:neon.models.mlp:Layers:
	DataLayer layer: 3 x (400 x 1) nodes
	ConvLayer cv_1: 3 x (400 x 1) inputs, 16 x (396 x 1) nodes, RectLin act_fn
	PoolingLayer po_1: 16 x (396 x 1) inputs, 16 x (198 x 1) nodes, Linear act_fn
	ConvLayer cv_2: 16 x (198 x 1) inputs, 32 x (194 x 1) nodes, RectLin act_fn
	PoolingLayer po_2: 32 x (194 x 1) inputs, 32 x (97 x 1) nodes, Linear act_fn
	DropOutLayer do_1: 32 x (97 x 1) inputs, 32 x (97 x 1) nodes, Linear act_fn
	FCLayer fc_1: 3104 inputs, 100 nodes, RectLin act_fn
	DropOutLayer do_2: 100 inputs, 100 nodes, Linear act_fn
	FCLayer fc_2: 100 inputs, 29 nodes, Logistic act_fn
	CostLayer cost: 29 nodes, CrossEntropy cost_fn

WARNING:neon.datasets.dataset:Incompatible batch size. Discarding 13 samples...
WARNING:neon.datasets.dataset:Incompatible batch size. Discarding 19 samples...
WARNING:neon.datasets.dataset:Incompatible batch size. Discarding 13 samples...
WARNING:neon.datasets.dataset:Incompatible batch size. Discarding 19 samples...
WARNING:neon.util.persist:deserializing object from:  /Users/tombocklisch/data/workout-cnn/workout-cnn.prm
INFO:neon.models.mlp:commencing model fitting
INFO:neon.models.mlp:epoch: 35, training error: 0.12945
WARNING:neon.util.persist:serializing object to: /Users/tombocklisch/data/workout-cnn/workout-cnn.prm






    




Finished epoch 35





    




INFO:neon.models.mlp:Layers:
	DataLayer layer: 3 x (400 x 1) nodes
	ConvLayer cv_1: 3 x (400 x 1) inputs, 16 x (396 x 1) nodes, RectLin act_fn
	PoolingLayer po_1: 16 x (396 x 1) inputs, 16 x (198 x 1) nodes, Linear act_fn
	ConvLayer cv_2: 16 x (198 x 1) inputs, 32 x (194 x 1) nodes, RectLin act_fn
	PoolingLayer po_2: 32 x (194 x 1) inputs, 32 x (97 x 1) nodes, Linear act_fn
	DropOutLayer do_1: 32 x (97 x 1) inputs, 32 x (97 x 1) nodes, Linear act_fn
	FCLayer fc_1: 3104 inputs, 100 nodes, RectLin act_fn
	DropOutLayer do_2: 100 inputs, 100 nodes, Linear act_fn
	FCLayer fc_2: 100 inputs, 29 nodes, Logistic act_fn
	CostLayer cost: 29 nodes, CrossEntropy cost_fn

WARNING:neon.datasets.dataset:Incompatible batch size. Discarding 13 samples...
WARNING:neon.datasets.dataset:Incompatible batch size. Discarding 19 samples...
WARNING:neon.datasets.dataset:Incompatible batch size. Discarding 13 samples...
WARNING:neon.datasets.dataset:Incompatible batch size. Discarding 19 samples...
WARNING:neon.util.persist:deserializing object from:  /Users/tombocklisch/data/workout-cnn/workout-cnn.prm
INFO:neon.models.mlp:commencing model fitting
INFO:neon.models.mlp:epoch: 36, training error: 0.11935
WARNING:neon.util.persist:serializing object to: /Users/tombocklisch/data/workout-cnn/workout-cnn.prm






    




Finished epoch 36





    




INFO:neon.models.mlp:Layers:
	DataLayer layer: 3 x (400 x 1) nodes
	ConvLayer cv_1: 3 x (400 x 1) inputs, 16 x (396 x 1) nodes, RectLin act_fn
	PoolingLayer po_1: 16 x (396 x 1) inputs, 16 x (198 x 1) nodes, Linear act_fn
	ConvLayer cv_2: 16 x (198 x 1) inputs, 32 x (194 x 1) nodes, RectLin act_fn
	PoolingLayer po_2: 32 x (194 x 1) inputs, 32 x (97 x 1) nodes, Linear act_fn
	DropOutLayer do_1: 32 x (97 x 1) inputs, 32 x (97 x 1) nodes, Linear act_fn
	FCLayer fc_1: 3104 inputs, 100 nodes, RectLin act_fn
	DropOutLayer do_2: 100 inputs, 100 nodes, Linear act_fn
	FCLayer fc_2: 100 inputs, 29 nodes, Logistic act_fn
	CostLayer cost: 29 nodes, CrossEntropy cost_fn

WARNING:neon.datasets.dataset:Incompatible batch size. Discarding 13 samples...
WARNING:neon.datasets.dataset:Incompatible batch size. Discarding 19 samples...
WARNING:neon.datasets.dataset:Incompatible batch size. Discarding 13 samples...
WARNING:neon.datasets.dataset:Incompatible batch size. Discarding 19 samples...
WARNING:neon.util.persist:deserializing object from:  /Users/tombocklisch/data/workout-cnn/workout-cnn.prm
INFO:neon.models.mlp:commencing model fitting
INFO:neon.models.mlp:epoch: 37, training error: 0.11506
WARNING:neon.util.persist:serializing object to: /Users/tombocklisch/data/workout-cnn/workout-cnn.prm






    




Finished epoch 37





    




INFO:neon.models.mlp:Layers:
	DataLayer layer: 3 x (400 x 1) nodes
	ConvLayer cv_1: 3 x (400 x 1) inputs, 16 x (396 x 1) nodes, RectLin act_fn
	PoolingLayer po_1: 16 x (396 x 1) inputs, 16 x (198 x 1) nodes, Linear act_fn
	ConvLayer cv_2: 16 x (198 x 1) inputs, 32 x (194 x 1) nodes, RectLin act_fn
	PoolingLayer po_2: 32 x (194 x 1) inputs, 32 x (97 x 1) nodes, Linear act_fn
	DropOutLayer do_1: 32 x (97 x 1) inputs, 32 x (97 x 1) nodes, Linear act_fn
	FCLayer fc_1: 3104 inputs, 100 nodes, RectLin act_fn
	DropOutLayer do_2: 100 inputs, 100 nodes, Linear act_fn
	FCLayer fc_2: 100 inputs, 29 nodes, Logistic act_fn
	CostLayer cost: 29 nodes, CrossEntropy cost_fn

WARNING:neon.datasets.dataset:Incompatible batch size. Discarding 13 samples...
WARNING:neon.datasets.dataset:Incompatible batch size. Discarding 19 samples...
WARNING:neon.datasets.dataset:Incompatible batch size. Discarding 13 samples...
WARNING:neon.datasets.dataset:Incompatible batch size. Discarding 19 samples...
WARNING:neon.util.persist:deserializing object from:  /Users/tombocklisch/data/workout-cnn/workout-cnn.prm
INFO:neon.models.mlp:commencing model fitting
INFO:neon.models.mlp:epoch: 38, training error: 0.11072
WARNING:neon.util.persist:serializing object to: /Users/tombocklisch/data/workout-cnn/workout-cnn.prm






    




Finished epoch 38





    




INFO:neon.models.mlp:Layers:
	DataLayer layer: 3 x (400 x 1) nodes
	ConvLayer cv_1: 3 x (400 x 1) inputs, 16 x (396 x 1) nodes, RectLin act_fn
	PoolingLayer po_1: 16 x (396 x 1) inputs, 16 x (198 x 1) nodes, Linear act_fn
	ConvLayer cv_2: 16 x (198 x 1) inputs, 32 x (194 x 1) nodes, RectLin act_fn
	PoolingLayer po_2: 32 x (194 x 1) inputs, 32 x (97 x 1) nodes, Linear act_fn
	DropOutLayer do_1: 32 x (97 x 1) inputs, 32 x (97 x 1) nodes, Linear act_fn
	FCLayer fc_1: 3104 inputs, 100 nodes, RectLin act_fn
	DropOutLayer do_2: 100 inputs, 100 nodes, Linear act_fn
	FCLayer fc_2: 100 inputs, 29 nodes, Logistic act_fn
	CostLayer cost: 29 nodes, CrossEntropy cost_fn

WARNING:neon.datasets.dataset:Incompatible batch size. Discarding 13 samples...
WARNING:neon.datasets.dataset:Incompatible batch size. Discarding 19 samples...
WARNING:neon.datasets.dataset:Incompatible batch size. Discarding 13 samples...
WARNING:neon.datasets.dataset:Incompatible batch size. Discarding 19 samples...
WARNING:neon.util.persist:deserializing object from:  /Users/tombocklisch/data/workout-cnn/workout-cnn.prm
INFO:neon.models.mlp:commencing model fitting
INFO:neon.models.mlp:epoch: 39, training error: 0.10636
WARNING:neon.util.persist:serializing object to: /Users/tombocklisch/data/workout-cnn/workout-cnn.prm






    




Finished epoch 39





    




INFO:neon.models.mlp:Layers:
	DataLayer layer: 3 x (400 x 1) nodes
	ConvLayer cv_1: 3 x (400 x 1) inputs, 16 x (396 x 1) nodes, RectLin act_fn
	PoolingLayer po_1: 16 x (396 x 1) inputs, 16 x (198 x 1) nodes, Linear act_fn
	ConvLayer cv_2: 16 x (198 x 1) inputs, 32 x (194 x 1) nodes, RectLin act_fn
	PoolingLayer po_2: 32 x (194 x 1) inputs, 32 x (97 x 1) nodes, Linear act_fn
	DropOutLayer do_1: 32 x (97 x 1) inputs, 32 x (97 x 1) nodes, Linear act_fn
	FCLayer fc_1: 3104 inputs, 100 nodes, RectLin act_fn
	DropOutLayer do_2: 100 inputs, 100 nodes, Linear act_fn
	FCLayer fc_2: 100 inputs, 29 nodes, Logistic act_fn
	CostLayer cost: 29 nodes, CrossEntropy cost_fn

WARNING:neon.datasets.dataset:Incompatible batch size. Discarding 13 samples...
WARNING:neon.datasets.dataset:Incompatible batch size. Discarding 19 samples...
WARNING:neon.datasets.dataset:Incompatible batch size. Discarding 13 samples...
WARNING:neon.datasets.dataset:Incompatible batch size. Discarding 19 samples...
WARNING:neon.util.persist:deserializing object from:  /Users/tombocklisch/data/workout-cnn/workout-cnn.prm
INFO:neon.models.mlp:commencing model fitting
INFO:neon.models.mlp:epoch: 40, training error: 0.10155
WARNING:neon.util.persist:serializing object to: /Users/tombocklisch/data/workout-cnn/workout-cnn.prm






    




Finished epoch 40





    




INFO:neon.models.mlp:Layers:
	DataLayer layer: 3 x (400 x 1) nodes
	ConvLayer cv_1: 3 x (400 x 1) inputs, 16 x (396 x 1) nodes, RectLin act_fn
	PoolingLayer po_1: 16 x (396 x 1) inputs, 16 x (198 x 1) nodes, Linear act_fn
	ConvLayer cv_2: 16 x (198 x 1) inputs, 32 x (194 x 1) nodes, RectLin act_fn
	PoolingLayer po_2: 32 x (194 x 1) inputs, 32 x (97 x 1) nodes, Linear act_fn
	DropOutLayer do_1: 32 x (97 x 1) inputs, 32 x (97 x 1) nodes, Linear act_fn
	FCLayer fc_1: 3104 inputs, 100 nodes, RectLin act_fn
	DropOutLayer do_2: 100 inputs, 100 nodes, Linear act_fn
	FCLayer fc_2: 100 inputs, 29 nodes, Logistic act_fn
	CostLayer cost: 29 nodes, CrossEntropy cost_fn

WARNING:neon.datasets.dataset:Incompatible batch size. Discarding 13 samples...
WARNING:neon.datasets.dataset:Incompatible batch size. Discarding 19 samples...
WARNING:neon.datasets.dataset:Incompatible batch size. Discarding 13 samples...
WARNING:neon.datasets.dataset:Incompatible batch size. Discarding 19 samples...
WARNING:neon.util.persist:deserializing object from:  /Users/tombocklisch/data/workout-cnn/workout-cnn.prm
INFO:neon.models.mlp:commencing model fitting
INFO:neon.models.mlp:epoch: 41, training error: 0.08609
WARNING:neon.util.persist:serializing object to: /Users/tombocklisch/data/workout-cnn/workout-cnn.prm






    




Finished epoch 41





    




INFO:neon.models.mlp:Layers:
	DataLayer layer: 3 x (400 x 1) nodes
	ConvLayer cv_1: 3 x (400 x 1) inputs, 16 x (396 x 1) nodes, RectLin act_fn
	PoolingLayer po_1: 16 x (396 x 1) inputs, 16 x (198 x 1) nodes, Linear act_fn
	ConvLayer cv_2: 16 x (198 x 1) inputs, 32 x (194 x 1) nodes, RectLin act_fn
	PoolingLayer po_2: 32 x (194 x 1) inputs, 32 x (97 x 1) nodes, Linear act_fn
	DropOutLayer do_1: 32 x (97 x 1) inputs, 32 x (97 x 1) nodes, Linear act_fn
	FCLayer fc_1: 3104 inputs, 100 nodes, RectLin act_fn
	DropOutLayer do_2: 100 inputs, 100 nodes, Linear act_fn
	FCLayer fc_2: 100 inputs, 29 nodes, Logistic act_fn
	CostLayer cost: 29 nodes, CrossEntropy cost_fn

WARNING:neon.datasets.dataset:Incompatible batch size. Discarding 13 samples...
WARNING:neon.datasets.dataset:Incompatible batch size. Discarding 19 samples...
WARNING:neon.datasets.dataset:Incompatible batch size. Discarding 13 samples...
WARNING:neon.datasets.dataset:Incompatible batch size. Discarding 19 samples...
WARNING:neon.util.persist:deserializing object from:  /Users/tombocklisch/data/workout-cnn/workout-cnn.prm
INFO:neon.models.mlp:commencing model fitting
INFO:neon.models.mlp:epoch: 42, training error: 0.08162
WARNING:neon.util.persist:serializing object to: /Users/tombocklisch/data/workout-cnn/workout-cnn.prm






    




Finished epoch 42





    




INFO:neon.models.mlp:Layers:
	DataLayer layer: 3 x (400 x 1) nodes
	ConvLayer cv_1: 3 x (400 x 1) inputs, 16 x (396 x 1) nodes, RectLin act_fn
	PoolingLayer po_1: 16 x (396 x 1) inputs, 16 x (198 x 1) nodes, Linear act_fn
	ConvLayer cv_2: 16 x (198 x 1) inputs, 32 x (194 x 1) nodes, RectLin act_fn
	PoolingLayer po_2: 32 x (194 x 1) inputs, 32 x (97 x 1) nodes, Linear act_fn
	DropOutLayer do_1: 32 x (97 x 1) inputs, 32 x (97 x 1) nodes, Linear act_fn
	FCLayer fc_1: 3104 inputs, 100 nodes, RectLin act_fn
	DropOutLayer do_2: 100 inputs, 100 nodes, Linear act_fn
	FCLayer fc_2: 100 inputs, 29 nodes, Logistic act_fn
	CostLayer cost: 29 nodes, CrossEntropy cost_fn

WARNING:neon.datasets.dataset:Incompatible batch size. Discarding 13 samples...
WARNING:neon.datasets.dataset:Incompatible batch size. Discarding 19 samples...
WARNING:neon.datasets.dataset:Incompatible batch size. Discarding 13 samples...
WARNING:neon.datasets.dataset:Incompatible batch size. Discarding 19 samples...
WARNING:neon.util.persist:deserializing object from:  /Users/tombocklisch/data/workout-cnn/workout-cnn.prm
INFO:neon.models.mlp:commencing model fitting
INFO:neon.models.mlp:epoch: 43, training error: 0.07943
WARNING:neon.util.persist:serializing object to: /Users/tombocklisch/data/workout-cnn/workout-cnn.prm






    




Finished epoch 43





    




INFO:neon.models.mlp:Layers:
	DataLayer layer: 3 x (400 x 1) nodes
	ConvLayer cv_1: 3 x (400 x 1) inputs, 16 x (396 x 1) nodes, RectLin act_fn
	PoolingLayer po_1: 16 x (396 x 1) inputs, 16 x (198 x 1) nodes, Linear act_fn
	ConvLayer cv_2: 16 x (198 x 1) inputs, 32 x (194 x 1) nodes, RectLin act_fn
	PoolingLayer po_2: 32 x (194 x 1) inputs, 32 x (97 x 1) nodes, Linear act_fn
	DropOutLayer do_1: 32 x (97 x 1) inputs, 32 x (97 x 1) nodes, Linear act_fn
	FCLayer fc_1: 3104 inputs, 100 nodes, RectLin act_fn
	DropOutLayer do_2: 100 inputs, 100 nodes, Linear act_fn
	FCLayer fc_2: 100 inputs, 29 nodes, Logistic act_fn
	CostLayer cost: 29 nodes, CrossEntropy cost_fn

WARNING:neon.datasets.dataset:Incompatible batch size. Discarding 13 samples...
WARNING:neon.datasets.dataset:Incompatible batch size. Discarding 19 samples...
WARNING:neon.datasets.dataset:Incompatible batch size. Discarding 13 samples...
WARNING:neon.datasets.dataset:Incompatible batch size. Discarding 19 samples...
WARNING:neon.util.persist:deserializing object from:  /Users/tombocklisch/data/workout-cnn/workout-cnn.prm
INFO:neon.models.mlp:commencing model fitting
INFO:neon.models.mlp:epoch: 44, training error: 0.07751
WARNING:neon.util.persist:serializing object to: /Users/tombocklisch/data/workout-cnn/workout-cnn.prm






    




Finished epoch 44





    




INFO:neon.models.mlp:Layers:
	DataLayer layer: 3 x (400 x 1) nodes
	ConvLayer cv_1: 3 x (400 x 1) inputs, 16 x (396 x 1) nodes, RectLin act_fn
	PoolingLayer po_1: 16 x (396 x 1) inputs, 16 x (198 x 1) nodes, Linear act_fn
	ConvLayer cv_2: 16 x (198 x 1) inputs, 32 x (194 x 1) nodes, RectLin act_fn
	PoolingLayer po_2: 32 x (194 x 1) inputs, 32 x (97 x 1) nodes, Linear act_fn
	DropOutLayer do_1: 32 x (97 x 1) inputs, 32 x (97 x 1) nodes, Linear act_fn
	FCLayer fc_1: 3104 inputs, 100 nodes, RectLin act_fn
	DropOutLayer do_2: 100 inputs, 100 nodes, Linear act_fn
	FCLayer fc_2: 100 inputs, 29 nodes, Logistic act_fn
	CostLayer cost: 29 nodes, CrossEntropy cost_fn

WARNING:neon.datasets.dataset:Incompatible batch size. Discarding 13 samples...
WARNING:neon.datasets.dataset:Incompatible batch size. Discarding 19 samples...
WARNING:neon.datasets.dataset:Incompatible batch size. Discarding 13 samples...
WARNING:neon.datasets.dataset:Incompatible batch size. Discarding 19 samples...
WARNING:neon.util.persist:deserializing object from:  /Users/tombocklisch/data/workout-cnn/workout-cnn.prm
INFO:neon.models.mlp:commencing model fitting
INFO:neon.models.mlp:epoch: 45, training error: 0.07616
WARNING:neon.util.persist:serializing object to: /Users/tombocklisch/data/workout-cnn/workout-cnn.prm






    




Finished epoch 45





    




INFO:neon.models.mlp:Layers:
	DataLayer layer: 3 x (400 x 1) nodes
	ConvLayer cv_1: 3 x (400 x 1) inputs, 16 x (396 x 1) nodes, RectLin act_fn
	PoolingLayer po_1: 16 x (396 x 1) inputs, 16 x (198 x 1) nodes, Linear act_fn
	ConvLayer cv_2: 16 x (198 x 1) inputs, 32 x (194 x 1) nodes, RectLin act_fn
	PoolingLayer po_2: 32 x (194 x 1) inputs, 32 x (97 x 1) nodes, Linear act_fn
	DropOutLayer do_1: 32 x (97 x 1) inputs, 32 x (97 x 1) nodes, Linear act_fn
	FCLayer fc_1: 3104 inputs, 100 nodes, RectLin act_fn
	DropOutLayer do_2: 100 inputs, 100 nodes, Linear act_fn
	FCLayer fc_2: 100 inputs, 29 nodes, Logistic act_fn
	CostLayer cost: 29 nodes, CrossEntropy cost_fn

WARNING:neon.datasets.dataset:Incompatible batch size. Discarding 13 samples...
WARNING:neon.datasets.dataset:Incompatible batch size. Discarding 19 samples...
WARNING:neon.datasets.dataset:Incompatible batch size. Discarding 13 samples...
WARNING:neon.datasets.dataset:Incompatible batch size. Discarding 19 samples...
WARNING:neon.util.persist:deserializing object from:  /Users/tombocklisch/data/workout-cnn/workout-cnn.prm
INFO:neon.models.mlp:commencing model fitting
INFO:neon.models.mlp:epoch: 46, training error: 0.07479
WARNING:neon.util.persist:serializing object to: /Users/tombocklisch/data/workout-cnn/workout-cnn.prm






    




Finished epoch 46





    




INFO:neon.models.mlp:Layers:
	DataLayer layer: 3 x (400 x 1) nodes
	ConvLayer cv_1: 3 x (400 x 1) inputs, 16 x (396 x 1) nodes, RectLin act_fn
	PoolingLayer po_1: 16 x (396 x 1) inputs, 16 x (198 x 1) nodes, Linear act_fn
	ConvLayer cv_2: 16 x (198 x 1) inputs, 32 x (194 x 1) nodes, RectLin act_fn
	PoolingLayer po_2: 32 x (194 x 1) inputs, 32 x (97 x 1) nodes, Linear act_fn
	DropOutLayer do_1: 32 x (97 x 1) inputs, 32 x (97 x 1) nodes, Linear act_fn
	FCLayer fc_1: 3104 inputs, 100 nodes, RectLin act_fn
	DropOutLayer do_2: 100 inputs, 100 nodes, Linear act_fn
	FCLayer fc_2: 100 inputs, 29 nodes, Logistic act_fn
	CostLayer cost: 29 nodes, CrossEntropy cost_fn

WARNING:neon.datasets.dataset:Incompatible batch size. Discarding 13 samples...
WARNING:neon.datasets.dataset:Incompatible batch size. Discarding 19 samples...
WARNING:neon.datasets.dataset:Incompatible batch size. Discarding 13 samples...
WARNING:neon.datasets.dataset:Incompatible batch size. Discarding 19 samples...
WARNING:neon.util.persist:deserializing object from:  /Users/tombocklisch/data/workout-cnn/workout-cnn.prm
INFO:neon.models.mlp:commencing model fitting
INFO:neon.models.mlp:epoch: 47, training error: 0.07343
WARNING:neon.util.persist:serializing object to: /Users/tombocklisch/data/workout-cnn/workout-cnn.prm






    




Finished epoch 47





    




INFO:neon.models.mlp:Layers:
	DataLayer layer: 3 x (400 x 1) nodes
	ConvLayer cv_1: 3 x (400 x 1) inputs, 16 x (396 x 1) nodes, RectLin act_fn
	PoolingLayer po_1: 16 x (396 x 1) inputs, 16 x (198 x 1) nodes, Linear act_fn
	ConvLayer cv_2: 16 x (198 x 1) inputs, 32 x (194 x 1) nodes, RectLin act_fn
	PoolingLayer po_2: 32 x (194 x 1) inputs, 32 x (97 x 1) nodes, Linear act_fn
	DropOutLayer do_1: 32 x (97 x 1) inputs, 32 x (97 x 1) nodes, Linear act_fn
	FCLayer fc_1: 3104 inputs, 100 nodes, RectLin act_fn
	DropOutLayer do_2: 100 inputs, 100 nodes, Linear act_fn
	FCLayer fc_2: 100 inputs, 29 nodes, Logistic act_fn
	CostLayer cost: 29 nodes, CrossEntropy cost_fn

WARNING:neon.datasets.dataset:Incompatible batch size. Discarding 13 samples...
WARNING:neon.datasets.dataset:Incompatible batch size. Discarding 19 samples...
WARNING:neon.datasets.dataset:Incompatible batch size. Discarding 13 samples...
WARNING:neon.datasets.dataset:Incompatible batch size. Discarding 19 samples...
WARNING:neon.util.persist:deserializing object from:  /Users/tombocklisch/data/workout-cnn/workout-cnn.prm
INFO:neon.models.mlp:commencing model fitting
INFO:neon.models.mlp:epoch: 48, training error: 0.07225
WARNING:neon.util.persist:serializing object to: /Users/tombocklisch/data/workout-cnn/workout-cnn.prm






    




Finished epoch 48





    




INFO:neon.models.mlp:Layers:
	DataLayer layer: 3 x (400 x 1) nodes
	ConvLayer cv_1: 3 x (400 x 1) inputs, 16 x (396 x 1) nodes, RectLin act_fn
	PoolingLayer po_1: 16 x (396 x 1) inputs, 16 x (198 x 1) nodes, Linear act_fn
	ConvLayer cv_2: 16 x (198 x 1) inputs, 32 x (194 x 1) nodes, RectLin act_fn
	PoolingLayer po_2: 32 x (194 x 1) inputs, 32 x (97 x 1) nodes, Linear act_fn
	DropOutLayer do_1: 32 x (97 x 1) inputs, 32 x (97 x 1) nodes, Linear act_fn
	FCLayer fc_1: 3104 inputs, 100 nodes, RectLin act_fn
	DropOutLayer do_2: 100 inputs, 100 nodes, Linear act_fn
	FCLayer fc_2: 100 inputs, 29 nodes, Logistic act_fn
	CostLayer cost: 29 nodes, CrossEntropy cost_fn

WARNING:neon.datasets.dataset:Incompatible batch size. Discarding 13 samples...
WARNING:neon.datasets.dataset:Incompatible batch size. Discarding 19 samples...
WARNING:neon.datasets.dataset:Incompatible batch size. Discarding 13 samples...
WARNING:neon.datasets.dataset:Incompatible batch size. Discarding 19 samples...
WARNING:neon.util.persist:deserializing object from:  /Users/tombocklisch/data/workout-cnn/workout-cnn.prm
INFO:neon.models.mlp:commencing model fitting
INFO:neon.models.mlp:epoch: 49, training error: 0.07110
WARNING:neon.util.persist:serializing object to: /Users/tombocklisch/data/workout-cnn/workout-cnn.prm






    




Finished epoch 49





    




INFO:neon.models.mlp:Layers:
	DataLayer layer: 3 x (400 x 1) nodes
	ConvLayer cv_1: 3 x (400 x 1) inputs, 16 x (396 x 1) nodes, RectLin act_fn
	PoolingLayer po_1: 16 x (396 x 1) inputs, 16 x (198 x 1) nodes, Linear act_fn
	ConvLayer cv_2: 16 x (198 x 1) inputs, 32 x (194 x 1) nodes, RectLin act_fn
	PoolingLayer po_2: 32 x (194 x 1) inputs, 32 x (97 x 1) nodes, Linear act_fn
	DropOutLayer do_1: 32 x (97 x 1) inputs, 32 x (97 x 1) nodes, Linear act_fn
	FCLayer fc_1: 3104 inputs, 100 nodes, RectLin act_fn
	DropOutLayer do_2: 100 inputs, 100 nodes, Linear act_fn
	FCLayer fc_2: 100 inputs, 29 nodes, Logistic act_fn
	CostLayer cost: 29 nodes, CrossEntropy cost_fn

WARNING:neon.datasets.dataset:Incompatible batch size. Discarding 13 samples...
WARNING:neon.datasets.dataset:Incompatible batch size. Discarding 19 samples...
WARNING:neon.datasets.dataset:Incompatible batch size. Discarding 13 samples...
WARNING:neon.datasets.dataset:Incompatible batch size. Discarding 19 samples...
WARNING:neon.util.persist:deserializing object from:  /Users/tombocklisch/data/workout-cnn/workout-cnn.prm
INFO:neon.models.mlp:commencing model fitting
INFO:neon.models.mlp:epoch: 50, training error: 0.07009
WARNING:neon.util.persist:serializing object to: /Users/tombocklisch/data/workout-cnn/workout-cnn.prm






    




Finished epoch 50





    




INFO:neon.models.mlp:Layers:
	DataLayer layer: 3 x (400 x 1) nodes
	ConvLayer cv_1: 3 x (400 x 1) inputs, 16 x (396 x 1) nodes, RectLin act_fn
	PoolingLayer po_1: 16 x (396 x 1) inputs, 16 x (198 x 1) nodes, Linear act_fn
	ConvLayer cv_2: 16 x (198 x 1) inputs, 32 x (194 x 1) nodes, RectLin act_fn
	PoolingLayer po_2: 32 x (194 x 1) inputs, 32 x (97 x 1) nodes, Linear act_fn
	DropOutLayer do_1: 32 x (97 x 1) inputs, 32 x (97 x 1) nodes, Linear act_fn
	FCLayer fc_1: 3104 inputs, 100 nodes, RectLin act_fn
	DropOutLayer do_2: 100 inputs, 100 nodes, Linear act_fn
	FCLayer fc_2: 100 inputs, 29 nodes, Logistic act_fn
	CostLayer cost: 29 nodes, CrossEntropy cost_fn

WARNING:neon.datasets.dataset:Incompatible batch size. Discarding 13 samples...
WARNING:neon.datasets.dataset:Incompatible batch size. Discarding 19 samples...
WARNING:neon.datasets.dataset:Incompatible batch size. Discarding 13 samples...
WARNING:neon.datasets.dataset:Incompatible batch size. Discarding 19 samples...
WARNING:neon.util.persist:deserializing object from:  /Users/tombocklisch/data/workout-cnn/workout-cnn.prm
INFO:neon.models.mlp:commencing model fitting
INFO:neon.models.mlp:epoch: 51, training error: 0.06924
WARNING:neon.util.persist:serializing object to: /Users/tombocklisch/data/workout-cnn/workout-cnn.prm






    




Finished epoch 51





    




INFO:neon.models.mlp:Layers:
	DataLayer layer: 3 x (400 x 1) nodes
	ConvLayer cv_1: 3 x (400 x 1) inputs, 16 x (396 x 1) nodes, RectLin act_fn
	PoolingLayer po_1: 16 x (396 x 1) inputs, 16 x (198 x 1) nodes, Linear act_fn
	ConvLayer cv_2: 16 x (198 x 1) inputs, 32 x (194 x 1) nodes, RectLin act_fn
	PoolingLayer po_2: 32 x (194 x 1) inputs, 32 x (97 x 1) nodes, Linear act_fn
	DropOutLayer do_1: 32 x (97 x 1) inputs, 32 x (97 x 1) nodes, Linear act_fn
	FCLayer fc_1: 3104 inputs, 100 nodes, RectLin act_fn
	DropOutLayer do_2: 100 inputs, 100 nodes, Linear act_fn
	FCLayer fc_2: 100 inputs, 29 nodes, Logistic act_fn
	CostLayer cost: 29 nodes, CrossEntropy cost_fn

WARNING:neon.datasets.dataset:Incompatible batch size. Discarding 13 samples...
WARNING:neon.datasets.dataset:Incompatible batch size. Discarding 19 samples...
WARNING:neon.datasets.dataset:Incompatible batch size. Discarding 13 samples...
WARNING:neon.datasets.dataset:Incompatible batch size. Discarding 19 samples...
WARNING:neon.util.persist:deserializing object from:  /Users/tombocklisch/data/workout-cnn/workout-cnn.prm
INFO:neon.models.mlp:commencing model fitting
INFO:neon.models.mlp:epoch: 52, training error: 0.06819
WARNING:neon.util.persist:serializing object to: /Users/tombocklisch/data/workout-cnn/workout-cnn.prm






    




Finished epoch 52





    




INFO:neon.models.mlp:Layers:
	DataLayer layer: 3 x (400 x 1) nodes
	ConvLayer cv_1: 3 x (400 x 1) inputs, 16 x (396 x 1) nodes, RectLin act_fn
	PoolingLayer po_1: 16 x (396 x 1) inputs, 16 x (198 x 1) nodes, Linear act_fn
	ConvLayer cv_2: 16 x (198 x 1) inputs, 32 x (194 x 1) nodes, RectLin act_fn
	PoolingLayer po_2: 32 x (194 x 1) inputs, 32 x (97 x 1) nodes, Linear act_fn
	DropOutLayer do_1: 32 x (97 x 1) inputs, 32 x (97 x 1) nodes, Linear act_fn
	FCLayer fc_1: 3104 inputs, 100 nodes, RectLin act_fn
	DropOutLayer do_2: 100 inputs, 100 nodes, Linear act_fn
	FCLayer fc_2: 100 inputs, 29 nodes, Logistic act_fn
	CostLayer cost: 29 nodes, CrossEntropy cost_fn

WARNING:neon.datasets.dataset:Incompatible batch size. Discarding 13 samples...
WARNING:neon.datasets.dataset:Incompatible batch size. Discarding 19 samples...
WARNING:neon.datasets.dataset:Incompatible batch size. Discarding 13 samples...
WARNING:neon.datasets.dataset:Incompatible batch size. Discarding 19 samples...
WARNING:neon.util.persist:deserializing object from:  /Users/tombocklisch/data/workout-cnn/workout-cnn.prm
INFO:neon.models.mlp:commencing model fitting
INFO:neon.models.mlp:epoch: 53, training error: 0.06720
WARNING:neon.util.persist:serializing object to: /Users/tombocklisch/data/workout-cnn/workout-cnn.prm






    




Finished epoch 53





    




INFO:neon.models.mlp:Layers:
	DataLayer layer: 3 x (400 x 1) nodes
	ConvLayer cv_1: 3 x (400 x 1) inputs, 16 x (396 x 1) nodes, RectLin act_fn
	PoolingLayer po_1: 16 x (396 x 1) inputs, 16 x (198 x 1) nodes, Linear act_fn
	ConvLayer cv_2: 16 x (198 x 1) inputs, 32 x (194 x 1) nodes, RectLin act_fn
	PoolingLayer po_2: 32 x (194 x 1) inputs, 32 x (97 x 1) nodes, Linear act_fn
	DropOutLayer do_1: 32 x (97 x 1) inputs, 32 x (97 x 1) nodes, Linear act_fn
	FCLayer fc_1: 3104 inputs, 100 nodes, RectLin act_fn
	DropOutLayer do_2: 100 inputs, 100 nodes, Linear act_fn
	FCLayer fc_2: 100 inputs, 29 nodes, Logistic act_fn
	CostLayer cost: 29 nodes, CrossEntropy cost_fn

WARNING:neon.datasets.dataset:Incompatible batch size. Discarding 13 samples...
WARNING:neon.datasets.dataset:Incompatible batch size. Discarding 19 samples...
WARNING:neon.datasets.dataset:Incompatible batch size. Discarding 13 samples...
WARNING:neon.datasets.dataset:Incompatible batch size. Discarding 19 samples...
WARNING:neon.util.persist:deserializing object from:  /Users/tombocklisch/data/workout-cnn/workout-cnn.prm
INFO:neon.models.mlp:commencing model fitting
INFO:neon.models.mlp:epoch: 54, training error: 0.06627
WARNING:neon.util.persist:serializing object to: /Users/tombocklisch/data/workout-cnn/workout-cnn.prm






    




Finished epoch 54





    




INFO:neon.models.mlp:Layers:
	DataLayer layer: 3 x (400 x 1) nodes
	ConvLayer cv_1: 3 x (400 x 1) inputs, 16 x (396 x 1) nodes, RectLin act_fn
	PoolingLayer po_1: 16 x (396 x 1) inputs, 16 x (198 x 1) nodes, Linear act_fn
	ConvLayer cv_2: 16 x (198 x 1) inputs, 32 x (194 x 1) nodes, RectLin act_fn
	PoolingLayer po_2: 32 x (194 x 1) inputs, 32 x (97 x 1) nodes, Linear act_fn
	DropOutLayer do_1: 32 x (97 x 1) inputs, 32 x (97 x 1) nodes, Linear act_fn
	FCLayer fc_1: 3104 inputs, 100 nodes, RectLin act_fn
	DropOutLayer do_2: 100 inputs, 100 nodes, Linear act_fn
	FCLayer fc_2: 100 inputs, 29 nodes, Logistic act_fn
	CostLayer cost: 29 nodes, CrossEntropy cost_fn

WARNING:neon.datasets.dataset:Incompatible batch size. Discarding 13 samples...
WARNING:neon.datasets.dataset:Incompatible batch size. Discarding 19 samples...
WARNING:neon.datasets.dataset:Incompatible batch size. Discarding 13 samples...
WARNING:neon.datasets.dataset:Incompatible batch size. Discarding 19 samples...
WARNING:neon.util.persist:deserializing object from:  /Users/tombocklisch/data/workout-cnn/workout-cnn.prm
INFO:neon.models.mlp:commencing model fitting
INFO:neon.models.mlp:epoch: 55, training error: 0.06550
WARNING:neon.util.persist:serializing object to: /Users/tombocklisch/data/workout-cnn/workout-cnn.prm






    




Finished epoch 55





    




INFO:neon.models.mlp:Layers:
	DataLayer layer: 3 x (400 x 1) nodes
	ConvLayer cv_1: 3 x (400 x 1) inputs, 16 x (396 x 1) nodes, RectLin act_fn
	PoolingLayer po_1: 16 x (396 x 1) inputs, 16 x (198 x 1) nodes, Linear act_fn
	ConvLayer cv_2: 16 x (198 x 1) inputs, 32 x (194 x 1) nodes, RectLin act_fn
	PoolingLayer po_2: 32 x (194 x 1) inputs, 32 x (97 x 1) nodes, Linear act_fn
	DropOutLayer do_1: 32 x (97 x 1) inputs, 32 x (97 x 1) nodes, Linear act_fn
	FCLayer fc_1: 3104 inputs, 100 nodes, RectLin act_fn
	DropOutLayer do_2: 100 inputs, 100 nodes, Linear act_fn
	FCLayer fc_2: 100 inputs, 29 nodes, Logistic act_fn
	CostLayer cost: 29 nodes, CrossEntropy cost_fn

WARNING:neon.datasets.dataset:Incompatible batch size. Discarding 13 samples...
WARNING:neon.datasets.dataset:Incompatible batch size. Discarding 19 samples...
WARNING:neon.datasets.dataset:Incompatible batch size. Discarding 13 samples...
WARNING:neon.datasets.dataset:Incompatible batch size. Discarding 19 samples...
WARNING:neon.util.persist:deserializing object from:  /Users/tombocklisch/data/workout-cnn/workout-cnn.prm
INFO:neon.models.mlp:commencing model fitting
INFO:neon.models.mlp:epoch: 56, training error: 0.06437
WARNING:neon.util.persist:serializing object to: /Users/tombocklisch/data/workout-cnn/workout-cnn.prm






    




Finished epoch 56





    




INFO:neon.models.mlp:Layers:
	DataLayer layer: 3 x (400 x 1) nodes
	ConvLayer cv_1: 3 x (400 x 1) inputs, 16 x (396 x 1) nodes, RectLin act_fn
	PoolingLayer po_1: 16 x (396 x 1) inputs, 16 x (198 x 1) nodes, Linear act_fn
	ConvLayer cv_2: 16 x (198 x 1) inputs, 32 x (194 x 1) nodes, RectLin act_fn
	PoolingLayer po_2: 32 x (194 x 1) inputs, 32 x (97 x 1) nodes, Linear act_fn
	DropOutLayer do_1: 32 x (97 x 1) inputs, 32 x (97 x 1) nodes, Linear act_fn
	FCLayer fc_1: 3104 inputs, 100 nodes, RectLin act_fn
	DropOutLayer do_2: 100 inputs, 100 nodes, Linear act_fn
	FCLayer fc_2: 100 inputs, 29 nodes, Logistic act_fn
	CostLayer cost: 29 nodes, CrossEntropy cost_fn

WARNING:neon.datasets.dataset:Incompatible batch size. Discarding 13 samples...
WARNING:neon.datasets.dataset:Incompatible batch size. Discarding 19 samples...
WARNING:neon.datasets.dataset:Incompatible batch size. Discarding 13 samples...
WARNING:neon.datasets.dataset:Incompatible batch size. Discarding 19 samples...
WARNING:neon.util.persist:deserializing object from:  /Users/tombocklisch/data/workout-cnn/workout-cnn.prm
INFO:neon.models.mlp:commencing model fitting
INFO:neon.models.mlp:epoch: 57, training error: 0.06377
WARNING:neon.util.persist:serializing object to: /Users/tombocklisch/data/workout-cnn/workout-cnn.prm






    




Finished epoch 57





    




INFO:neon.models.mlp:Layers:
	DataLayer layer: 3 x (400 x 1) nodes
	ConvLayer cv_1: 3 x (400 x 1) inputs, 16 x (396 x 1) nodes, RectLin act_fn
	PoolingLayer po_1: 16 x (396 x 1) inputs, 16 x (198 x 1) nodes, Linear act_fn
	ConvLayer cv_2: 16 x (198 x 1) inputs, 32 x (194 x 1) nodes, RectLin act_fn
	PoolingLayer po_2: 32 x (194 x 1) inputs, 32 x (97 x 1) nodes, Linear act_fn
	DropOutLayer do_1: 32 x (97 x 1) inputs, 32 x (97 x 1) nodes, Linear act_fn
	FCLayer fc_1: 3104 inputs, 100 nodes, RectLin act_fn
	DropOutLayer do_2: 100 inputs, 100 nodes, Linear act_fn
	FCLayer fc_2: 100 inputs, 29 nodes, Logistic act_fn
	CostLayer cost: 29 nodes, CrossEntropy cost_fn

WARNING:neon.datasets.dataset:Incompatible batch size. Discarding 13 samples...
WARNING:neon.datasets.dataset:Incompatible batch size. Discarding 19 samples...
WARNING:neon.datasets.dataset:Incompatible batch size. Discarding 13 samples...
WARNING:neon.datasets.dataset:Incompatible batch size. Discarding 19 samples...
WARNING:neon.util.persist:deserializing object from:  /Users/tombocklisch/data/workout-cnn/workout-cnn.prm
INFO:neon.models.mlp:commencing model fitting
INFO:neon.models.mlp:epoch: 58, training error: 0.06272
WARNING:neon.util.persist:serializing object to: /Users/tombocklisch/data/workout-cnn/workout-cnn.prm






    




Finished epoch 58





    




INFO:neon.models.mlp:Layers:
	DataLayer layer: 3 x (400 x 1) nodes
	ConvLayer cv_1: 3 x (400 x 1) inputs, 16 x (396 x 1) nodes, RectLin act_fn
	PoolingLayer po_1: 16 x (396 x 1) inputs, 16 x (198 x 1) nodes, Linear act_fn
	ConvLayer cv_2: 16 x (198 x 1) inputs, 32 x (194 x 1) nodes, RectLin act_fn
	PoolingLayer po_2: 32 x (194 x 1) inputs, 32 x (97 x 1) nodes, Linear act_fn
	DropOutLayer do_1: 32 x (97 x 1) inputs, 32 x (97 x 1) nodes, Linear act_fn
	FCLayer fc_1: 3104 inputs, 100 nodes, RectLin act_fn
	DropOutLayer do_2: 100 inputs, 100 nodes, Linear act_fn
	FCLayer fc_2: 100 inputs, 29 nodes, Logistic act_fn
	CostLayer cost: 29 nodes, CrossEntropy cost_fn

WARNING:neon.datasets.dataset:Incompatible batch size. Discarding 13 samples...
WARNING:neon.datasets.dataset:Incompatible batch size. Discarding 19 samples...
WARNING:neon.datasets.dataset:Incompatible batch size. Discarding 13 samples...
WARNING:neon.datasets.dataset:Incompatible batch size. Discarding 19 samples...
WARNING:neon.util.persist:deserializing object from:  /Users/tombocklisch/data/workout-cnn/workout-cnn.prm
INFO:neon.models.mlp:commencing model fitting
INFO:neon.models.mlp:epoch: 59, training error: 0.06190
WARNING:neon.util.persist:serializing object to: /Users/tombocklisch/data/workout-cnn/workout-cnn.prm






    




Finished epoch 59





    




INFO:neon.models.mlp:Layers:
	DataLayer layer: 3 x (400 x 1) nodes
	ConvLayer cv_1: 3 x (400 x 1) inputs, 16 x (396 x 1) nodes, RectLin act_fn
	PoolingLayer po_1: 16 x (396 x 1) inputs, 16 x (198 x 1) nodes, Linear act_fn
	ConvLayer cv_2: 16 x (198 x 1) inputs, 32 x (194 x 1) nodes, RectLin act_fn
	PoolingLayer po_2: 32 x (194 x 1) inputs, 32 x (97 x 1) nodes, Linear act_fn
	DropOutLayer do_1: 32 x (97 x 1) inputs, 32 x (97 x 1) nodes, Linear act_fn
	FCLayer fc_1: 3104 inputs, 100 nodes, RectLin act_fn
	DropOutLayer do_2: 100 inputs, 100 nodes, Linear act_fn
	FCLayer fc_2: 100 inputs, 29 nodes, Logistic act_fn
	CostLayer cost: 29 nodes, CrossEntropy cost_fn

WARNING:neon.datasets.dataset:Incompatible batch size. Discarding 13 samples...
WARNING:neon.datasets.dataset:Incompatible batch size. Discarding 19 samples...
WARNING:neon.datasets.dataset:Incompatible batch size. Discarding 13 samples...
WARNING:neon.datasets.dataset:Incompatible batch size. Discarding 19 samples...
WARNING:neon.util.persist:deserializing object from:  /Users/tombocklisch/data/workout-cnn/workout-cnn.prm
INFO:neon.models.mlp:commencing model fitting
INFO:neon.models.mlp:epoch: 60, training error: 0.06101
WARNING:neon.util.persist:serializing object to: /Users/tombocklisch/data/workout-cnn/workout-cnn.prm






    




Finished epoch 60





    




INFO:neon.models.mlp:Layers:
	DataLayer layer: 3 x (400 x 1) nodes
	ConvLayer cv_1: 3 x (400 x 1) inputs, 16 x (396 x 1) nodes, RectLin act_fn
	PoolingLayer po_1: 16 x (396 x 1) inputs, 16 x (198 x 1) nodes, Linear act_fn
	ConvLayer cv_2: 16 x (198 x 1) inputs, 32 x (194 x 1) nodes, RectLin act_fn
	PoolingLayer po_2: 32 x (194 x 1) inputs, 32 x (97 x 1) nodes, Linear act_fn
	DropOutLayer do_1: 32 x (97 x 1) inputs, 32 x (97 x 1) nodes, Linear act_fn
	FCLayer fc_1: 3104 inputs, 100 nodes, RectLin act_fn
	DropOutLayer do_2: 100 inputs, 100 nodes, Linear act_fn
	FCLayer fc_2: 100 inputs, 29 nodes, Logistic act_fn
	CostLayer cost: 29 nodes, CrossEntropy cost_fn

WARNING:neon.datasets.dataset:Incompatible batch size. Discarding 13 samples...
WARNING:neon.datasets.dataset:Incompatible batch size. Discarding 19 samples...
WARNING:neon.datasets.dataset:Incompatible batch size. Discarding 13 samples...
WARNING:neon.datasets.dataset:Incompatible batch size. Discarding 19 samples...
WARNING:neon.util.persist:deserializing object from:  /Users/tombocklisch/data/workout-cnn/workout-cnn.prm
INFO:neon.models.mlp:commencing model fitting
INFO:neon.models.mlp:epoch: 61, training error: 0.05833
WARNING:neon.util.persist:serializing object to: /Users/tombocklisch/data/workout-cnn/workout-cnn.prm






    




Finished epoch 61





    




INFO:neon.models.mlp:Layers:
	DataLayer layer: 3 x (400 x 1) nodes
	ConvLayer cv_1: 3 x (400 x 1) inputs, 16 x (396 x 1) nodes, RectLin act_fn
	PoolingLayer po_1: 16 x (396 x 1) inputs, 16 x (198 x 1) nodes, Linear act_fn
	ConvLayer cv_2: 16 x (198 x 1) inputs, 32 x (194 x 1) nodes, RectLin act_fn
	PoolingLayer po_2: 32 x (194 x 1) inputs, 32 x (97 x 1) nodes, Linear act_fn
	DropOutLayer do_1: 32 x (97 x 1) inputs, 32 x (97 x 1) nodes, Linear act_fn
	FCLayer fc_1: 3104 inputs, 100 nodes, RectLin act_fn
	DropOutLayer do_2: 100 inputs, 100 nodes, Linear act_fn
	FCLayer fc_2: 100 inputs, 29 nodes, Logistic act_fn
	CostLayer cost: 29 nodes, CrossEntropy cost_fn

WARNING:neon.datasets.dataset:Incompatible batch size. Discarding 13 samples...
WARNING:neon.datasets.dataset:Incompatible batch size. Discarding 19 samples...
WARNING:neon.datasets.dataset:Incompatible batch size. Discarding 13 samples...
WARNING:neon.datasets.dataset:Incompatible batch size. Discarding 19 samples...
WARNING:neon.util.persist:deserializing object from:  /Users/tombocklisch/data/workout-cnn/workout-cnn.prm
INFO:neon.models.mlp:commencing model fitting
INFO:neon.models.mlp:epoch: 62, training error: 0.05705
WARNING:neon.util.persist:serializing object to: /Users/tombocklisch/data/workout-cnn/workout-cnn.prm






    




Finished epoch 62





    




INFO:neon.models.mlp:Layers:
	DataLayer layer: 3 x (400 x 1) nodes
	ConvLayer cv_1: 3 x (400 x 1) inputs, 16 x (396 x 1) nodes, RectLin act_fn
	PoolingLayer po_1: 16 x (396 x 1) inputs, 16 x (198 x 1) nodes, Linear act_fn
	ConvLayer cv_2: 16 x (198 x 1) inputs, 32 x (194 x 1) nodes, RectLin act_fn
	PoolingLayer po_2: 32 x (194 x 1) inputs, 32 x (97 x 1) nodes, Linear act_fn
	DropOutLayer do_1: 32 x (97 x 1) inputs, 32 x (97 x 1) nodes, Linear act_fn
	FCLayer fc_1: 3104 inputs, 100 nodes, RectLin act_fn
	DropOutLayer do_2: 100 inputs, 100 nodes, Linear act_fn
	FCLayer fc_2: 100 inputs, 29 nodes, Logistic act_fn
	CostLayer cost: 29 nodes, CrossEntropy cost_fn

WARNING:neon.datasets.dataset:Incompatible batch size. Discarding 13 samples...
WARNING:neon.datasets.dataset:Incompatible batch size. Discarding 19 samples...
WARNING:neon.datasets.dataset:Incompatible batch size. Discarding 13 samples...
WARNING:neon.datasets.dataset:Incompatible batch size. Discarding 19 samples...
WARNING:neon.util.persist:deserializing object from:  /Users/tombocklisch/data/workout-cnn/workout-cnn.prm
INFO:neon.models.mlp:commencing model fitting
INFO:neon.models.mlp:epoch: 63, training error: 0.05655
WARNING:neon.util.persist:serializing object to: /Users/tombocklisch/data/workout-cnn/workout-cnn.prm






    




Finished epoch 63





    




INFO:neon.models.mlp:Layers:
	DataLayer layer: 3 x (400 x 1) nodes
	ConvLayer cv_1: 3 x (400 x 1) inputs, 16 x (396 x 1) nodes, RectLin act_fn
	PoolingLayer po_1: 16 x (396 x 1) inputs, 16 x (198 x 1) nodes, Linear act_fn
	ConvLayer cv_2: 16 x (198 x 1) inputs, 32 x (194 x 1) nodes, RectLin act_fn
	PoolingLayer po_2: 32 x (194 x 1) inputs, 32 x (97 x 1) nodes, Linear act_fn
	DropOutLayer do_1: 32 x (97 x 1) inputs, 32 x (97 x 1) nodes, Linear act_fn
	FCLayer fc_1: 3104 inputs, 100 nodes, RectLin act_fn
	DropOutLayer do_2: 100 inputs, 100 nodes, Linear act_fn
	FCLayer fc_2: 100 inputs, 29 nodes, Logistic act_fn
	CostLayer cost: 29 nodes, CrossEntropy cost_fn

WARNING:neon.datasets.dataset:Incompatible batch size. Discarding 13 samples...
WARNING:neon.datasets.dataset:Incompatible batch size. Discarding 19 samples...
WARNING:neon.datasets.dataset:Incompatible batch size. Discarding 13 samples...
WARNING:neon.datasets.dataset:Incompatible batch size. Discarding 19 samples...
WARNING:neon.util.persist:deserializing object from:  /Users/tombocklisch/data/workout-cnn/workout-cnn.prm
INFO:neon.models.mlp:commencing model fitting
INFO:neon.models.mlp:epoch: 64, training error: 0.05618
WARNING:neon.util.persist:serializing object to: /Users/tombocklisch/data/workout-cnn/workout-cnn.prm






    




Finished epoch 64





    




INFO:neon.models.mlp:Layers:
	DataLayer layer: 3 x (400 x 1) nodes
	ConvLayer cv_1: 3 x (400 x 1) inputs, 16 x (396 x 1) nodes, RectLin act_fn
	PoolingLayer po_1: 16 x (396 x 1) inputs, 16 x (198 x 1) nodes, Linear act_fn
	ConvLayer cv_2: 16 x (198 x 1) inputs, 32 x (194 x 1) nodes, RectLin act_fn
	PoolingLayer po_2: 32 x (194 x 1) inputs, 32 x (97 x 1) nodes, Linear act_fn
	DropOutLayer do_1: 32 x (97 x 1) inputs, 32 x (97 x 1) nodes, Linear act_fn
	FCLayer fc_1: 3104 inputs, 100 nodes, RectLin act_fn
	DropOutLayer do_2: 100 inputs, 100 nodes, Linear act_fn
	FCLayer fc_2: 100 inputs, 29 nodes, Logistic act_fn
	CostLayer cost: 29 nodes, CrossEntropy cost_fn

WARNING:neon.datasets.dataset:Incompatible batch size. Discarding 13 samples...
WARNING:neon.datasets.dataset:Incompatible batch size. Discarding 19 samples...
WARNING:neon.datasets.dataset:Incompatible batch size. Discarding 13 samples...
WARNING:neon.datasets.dataset:Incompatible batch size. Discarding 19 samples...
WARNING:neon.util.persist:deserializing object from:  /Users/tombocklisch/data/workout-cnn/workout-cnn.prm
INFO:neon.models.mlp:commencing model fitting
INFO:neon.models.mlp:epoch: 65, training error: 0.05582
WARNING:neon.util.persist:serializing object to: /Users/tombocklisch/data/workout-cnn/workout-cnn.prm






    




Finished epoch 65





    




INFO:neon.models.mlp:Layers:
	DataLayer layer: 3 x (400 x 1) nodes
	ConvLayer cv_1: 3 x (400 x 1) inputs, 16 x (396 x 1) nodes, RectLin act_fn
	PoolingLayer po_1: 16 x (396 x 1) inputs, 16 x (198 x 1) nodes, Linear act_fn
	ConvLayer cv_2: 16 x (198 x 1) inputs, 32 x (194 x 1) nodes, RectLin act_fn
	PoolingLayer po_2: 32 x (194 x 1) inputs, 32 x (97 x 1) nodes, Linear act_fn
	DropOutLayer do_1: 32 x (97 x 1) inputs, 32 x (97 x 1) nodes, Linear act_fn
	FCLayer fc_1: 3104 inputs, 100 nodes, RectLin act_fn
	DropOutLayer do_2: 100 inputs, 100 nodes, Linear act_fn
	FCLayer fc_2: 100 inputs, 29 nodes, Logistic act_fn
	CostLayer cost: 29 nodes, CrossEntropy cost_fn

WARNING:neon.datasets.dataset:Incompatible batch size. Discarding 13 samples...
WARNING:neon.datasets.dataset:Incompatible batch size. Discarding 19 samples...
WARNING:neon.datasets.dataset:Incompatible batch size. Discarding 13 samples...
WARNING:neon.datasets.dataset:Incompatible batch size. Discarding 19 samples...
WARNING:neon.util.persist:deserializing object from:  /Users/tombocklisch/data/workout-cnn/workout-cnn.prm
INFO:neon.models.mlp:commencing model fitting
INFO:neon.models.mlp:epoch: 66, training error: 0.05549
WARNING:neon.util.persist:serializing object to: /Users/tombocklisch/data/workout-cnn/workout-cnn.prm






    




Finished epoch 66





    




INFO:neon.models.mlp:Layers:
	DataLayer layer: 3 x (400 x 1) nodes
	ConvLayer cv_1: 3 x (400 x 1) inputs, 16 x (396 x 1) nodes, RectLin act_fn
	PoolingLayer po_1: 16 x (396 x 1) inputs, 16 x (198 x 1) nodes, Linear act_fn
	ConvLayer cv_2: 16 x (198 x 1) inputs, 32 x (194 x 1) nodes, RectLin act_fn
	PoolingLayer po_2: 32 x (194 x 1) inputs, 32 x (97 x 1) nodes, Linear act_fn
	DropOutLayer do_1: 32 x (97 x 1) inputs, 32 x (97 x 1) nodes, Linear act_fn
	FCLayer fc_1: 3104 inputs, 100 nodes, RectLin act_fn
	DropOutLayer do_2: 100 inputs, 100 nodes, Linear act_fn
	FCLayer fc_2: 100 inputs, 29 nodes, Logistic act_fn
	CostLayer cost: 29 nodes, CrossEntropy cost_fn

WARNING:neon.datasets.dataset:Incompatible batch size. Discarding 13 samples...
WARNING:neon.datasets.dataset:Incompatible batch size. Discarding 19 samples...
WARNING:neon.datasets.dataset:Incompatible batch size. Discarding 13 samples...
WARNING:neon.datasets.dataset:Incompatible batch size. Discarding 19 samples...
WARNING:neon.util.persist:deserializing object from:  /Users/tombocklisch/data/workout-cnn/workout-cnn.prm
INFO:neon.models.mlp:commencing model fitting
INFO:neon.models.mlp:epoch: 67, training error: 0.05523
WARNING:neon.util.persist:serializing object to: /Users/tombocklisch/data/workout-cnn/workout-cnn.prm






    




Finished epoch 67





    




INFO:neon.models.mlp:Layers:
	DataLayer layer: 3 x (400 x 1) nodes
	ConvLayer cv_1: 3 x (400 x 1) inputs, 16 x (396 x 1) nodes, RectLin act_fn
	PoolingLayer po_1: 16 x (396 x 1) inputs, 16 x (198 x 1) nodes, Linear act_fn
	ConvLayer cv_2: 16 x (198 x 1) inputs, 32 x (194 x 1) nodes, RectLin act_fn
	PoolingLayer po_2: 32 x (194 x 1) inputs, 32 x (97 x 1) nodes, Linear act_fn
	DropOutLayer do_1: 32 x (97 x 1) inputs, 32 x (97 x 1) nodes, Linear act_fn
	FCLayer fc_1: 3104 inputs, 100 nodes, RectLin act_fn
	DropOutLayer do_2: 100 inputs, 100 nodes, Linear act_fn
	FCLayer fc_2: 100 inputs, 29 nodes, Logistic act_fn
	CostLayer cost: 29 nodes, CrossEntropy cost_fn

WARNING:neon.datasets.dataset:Incompatible batch size. Discarding 13 samples...
WARNING:neon.datasets.dataset:Incompatible batch size. Discarding 19 samples...
WARNING:neon.datasets.dataset:Incompatible batch size. Discarding 13 samples...
WARNING:neon.datasets.dataset:Incompatible batch size. Discarding 19 samples...
WARNING:neon.util.persist:deserializing object from:  /Users/tombocklisch/data/workout-cnn/workout-cnn.prm
INFO:neon.models.mlp:commencing model fitting
INFO:neon.models.mlp:epoch: 68, training error: 0.05488
WARNING:neon.util.persist:serializing object to: /Users/tombocklisch/data/workout-cnn/workout-cnn.prm






    




Finished epoch 68





    




INFO:neon.models.mlp:Layers:
	DataLayer layer: 3 x (400 x 1) nodes
	ConvLayer cv_1: 3 x (400 x 1) inputs, 16 x (396 x 1) nodes, RectLin act_fn
	PoolingLayer po_1: 16 x (396 x 1) inputs, 16 x (198 x 1) nodes, Linear act_fn
	ConvLayer cv_2: 16 x (198 x 1) inputs, 32 x (194 x 1) nodes, RectLin act_fn
	PoolingLayer po_2: 32 x (194 x 1) inputs, 32 x (97 x 1) nodes, Linear act_fn
	DropOutLayer do_1: 32 x (97 x 1) inputs, 32 x (97 x 1) nodes, Linear act_fn
	FCLayer fc_1: 3104 inputs, 100 nodes, RectLin act_fn
	DropOutLayer do_2: 100 inputs, 100 nodes, Linear act_fn
	FCLayer fc_2: 100 inputs, 29 nodes, Logistic act_fn
	CostLayer cost: 29 nodes, CrossEntropy cost_fn

WARNING:neon.datasets.dataset:Incompatible batch size. Discarding 13 samples...
WARNING:neon.datasets.dataset:Incompatible batch size. Discarding 19 samples...
WARNING:neon.datasets.dataset:Incompatible batch size. Discarding 13 samples...
WARNING:neon.datasets.dataset:Incompatible batch size. Discarding 19 samples...
WARNING:neon.util.persist:deserializing object from:  /Users/tombocklisch/data/workout-cnn/workout-cnn.prm
INFO:neon.models.mlp:commencing model fitting
INFO:neon.models.mlp:epoch: 69, training error: 0.05459
WARNING:neon.util.persist:serializing object to: /Users/tombocklisch/data/workout-cnn/workout-cnn.prm






    




Finished epoch 69





    




INFO:neon.models.mlp:Layers:
	DataLayer layer: 3 x (400 x 1) nodes
	ConvLayer cv_1: 3 x (400 x 1) inputs, 16 x (396 x 1) nodes, RectLin act_fn
	PoolingLayer po_1: 16 x (396 x 1) inputs, 16 x (198 x 1) nodes, Linear act_fn
	ConvLayer cv_2: 16 x (198 x 1) inputs, 32 x (194 x 1) nodes, RectLin act_fn
	PoolingLayer po_2: 32 x (194 x 1) inputs, 32 x (97 x 1) nodes, Linear act_fn
	DropOutLayer do_1: 32 x (97 x 1) inputs, 32 x (97 x 1) nodes, Linear act_fn
	FCLayer fc_1: 3104 inputs, 100 nodes, RectLin act_fn
	DropOutLayer do_2: 100 inputs, 100 nodes, Linear act_fn
	FCLayer fc_2: 100 inputs, 29 nodes, Logistic act_fn
	CostLayer cost: 29 nodes, CrossEntropy cost_fn

WARNING:neon.datasets.dataset:Incompatible batch size. Discarding 13 samples...
WARNING:neon.datasets.dataset:Incompatible batch size. Discarding 19 samples...
WARNING:neon.datasets.dataset:Incompatible batch size. Discarding 13 samples...
WARNING:neon.datasets.dataset:Incompatible batch size. Discarding 19 samples...
WARNING:neon.util.persist:deserializing object from:  /Users/tombocklisch/data/workout-cnn/workout-cnn.prm
INFO:neon.models.mlp:commencing model fitting
INFO:neon.models.mlp:epoch: 70, training error: 0.05430
WARNING:neon.util.persist:serializing object to: /Users/tombocklisch/data/workout-cnn/workout-cnn.prm






    




Finished epoch 70





    




INFO:neon.models.mlp:Layers:
	DataLayer layer: 3 x (400 x 1) nodes
	ConvLayer cv_1: 3 x (400 x 1) inputs, 16 x (396 x 1) nodes, RectLin act_fn
	PoolingLayer po_1: 16 x (396 x 1) inputs, 16 x (198 x 1) nodes, Linear act_fn
	ConvLayer cv_2: 16 x (198 x 1) inputs, 32 x (194 x 1) nodes, RectLin act_fn
	PoolingLayer po_2: 32 x (194 x 1) inputs, 32 x (97 x 1) nodes, Linear act_fn
	DropOutLayer do_1: 32 x (97 x 1) inputs, 32 x (97 x 1) nodes, Linear act_fn
	FCLayer fc_1: 3104 inputs, 100 nodes, RectLin act_fn
	DropOutLayer do_2: 100 inputs, 100 nodes, Linear act_fn
	FCLayer fc_2: 100 inputs, 29 nodes, Logistic act_fn
	CostLayer cost: 29 nodes, CrossEntropy cost_fn

WARNING:neon.datasets.dataset:Incompatible batch size. Discarding 13 samples...
WARNING:neon.datasets.dataset:Incompatible batch size. Discarding 19 samples...
WARNING:neon.datasets.dataset:Incompatible batch size. Discarding 13 samples...
WARNING:neon.datasets.dataset:Incompatible batch size. Discarding 19 samples...
WARNING:neon.util.persist:deserializing object from:  /Users/tombocklisch/data/workout-cnn/workout-cnn.prm
INFO:neon.models.mlp:commencing model fitting
INFO:neon.models.mlp:epoch: 71, training error: 0.05402
WARNING:neon.util.persist:serializing object to: /Users/tombocklisch/data/workout-cnn/workout-cnn.prm






    




Finished epoch 71





    




INFO:neon.models.mlp:Layers:
	DataLayer layer: 3 x (400 x 1) nodes
	ConvLayer cv_1: 3 x (400 x 1) inputs, 16 x (396 x 1) nodes, RectLin act_fn
	PoolingLayer po_1: 16 x (396 x 1) inputs, 16 x (198 x 1) nodes, Linear act_fn
	ConvLayer cv_2: 16 x (198 x 1) inputs, 32 x (194 x 1) nodes, RectLin act_fn
	PoolingLayer po_2: 32 x (194 x 1) inputs, 32 x (97 x 1) nodes, Linear act_fn
	DropOutLayer do_1: 32 x (97 x 1) inputs, 32 x (97 x 1) nodes, Linear act_fn
	FCLayer fc_1: 3104 inputs, 100 nodes, RectLin act_fn
	DropOutLayer do_2: 100 inputs, 100 nodes, Linear act_fn
	FCLayer fc_2: 100 inputs, 29 nodes, Logistic act_fn
	CostLayer cost: 29 nodes, CrossEntropy cost_fn

WARNING:neon.datasets.dataset:Incompatible batch size. Discarding 13 samples...
WARNING:neon.datasets.dataset:Incompatible batch size. Discarding 19 samples...
WARNING:neon.datasets.dataset:Incompatible batch size. Discarding 13 samples...
WARNING:neon.datasets.dataset:Incompatible batch size. Discarding 19 samples...
WARNING:neon.util.persist:deserializing object from:  /Users/tombocklisch/data/workout-cnn/workout-cnn.prm
INFO:neon.models.mlp:commencing model fitting
INFO:neon.models.mlp:epoch: 72, training error: 0.05376
WARNING:neon.util.persist:serializing object to: /Users/tombocklisch/data/workout-cnn/workout-cnn.prm






    




Finished epoch 72





    




INFO:neon.models.mlp:Layers:
	DataLayer layer: 3 x (400 x 1) nodes
	ConvLayer cv_1: 3 x (400 x 1) inputs, 16 x (396 x 1) nodes, RectLin act_fn
	PoolingLayer po_1: 16 x (396 x 1) inputs, 16 x (198 x 1) nodes, Linear act_fn
	ConvLayer cv_2: 16 x (198 x 1) inputs, 32 x (194 x 1) nodes, RectLin act_fn
	PoolingLayer po_2: 32 x (194 x 1) inputs, 32 x (97 x 1) nodes, Linear act_fn
	DropOutLayer do_1: 32 x (97 x 1) inputs, 32 x (97 x 1) nodes, Linear act_fn
	FCLayer fc_1: 3104 inputs, 100 nodes, RectLin act_fn
	DropOutLayer do_2: 100 inputs, 100 nodes, Linear act_fn
	FCLayer fc_2: 100 inputs, 29 nodes, Logistic act_fn
	CostLayer cost: 29 nodes, CrossEntropy cost_fn

WARNING:neon.datasets.dataset:Incompatible batch size. Discarding 13 samples...
WARNING:neon.datasets.dataset:Incompatible batch size. Discarding 19 samples...
WARNING:neon.datasets.dataset:Incompatible batch size. Discarding 13 samples...
WARNING:neon.datasets.dataset:Incompatible batch size. Discarding 19 samples...
WARNING:neon.util.persist:deserializing object from:  /Users/tombocklisch/data/workout-cnn/workout-cnn.prm
INFO:neon.models.mlp:commencing model fitting
INFO:neon.models.mlp:epoch: 73, training error: 0.05347
WARNING:neon.util.persist:serializing object to: /Users/tombocklisch/data/workout-cnn/workout-cnn.prm






    




Finished epoch 73





    




INFO:neon.models.mlp:Layers:
	DataLayer layer: 3 x (400 x 1) nodes
	ConvLayer cv_1: 3 x (400 x 1) inputs, 16 x (396 x 1) nodes, RectLin act_fn
	PoolingLayer po_1: 16 x (396 x 1) inputs, 16 x (198 x 1) nodes, Linear act_fn
	ConvLayer cv_2: 16 x (198 x 1) inputs, 32 x (194 x 1) nodes, RectLin act_fn
	PoolingLayer po_2: 32 x (194 x 1) inputs, 32 x (97 x 1) nodes, Linear act_fn
	DropOutLayer do_1: 32 x (97 x 1) inputs, 32 x (97 x 1) nodes, Linear act_fn
	FCLayer fc_1: 3104 inputs, 100 nodes, RectLin act_fn
	DropOutLayer do_2: 100 inputs, 100 nodes, Linear act_fn
	FCLayer fc_2: 100 inputs, 29 nodes, Logistic act_fn
	CostLayer cost: 29 nodes, CrossEntropy cost_fn

WARNING:neon.datasets.dataset:Incompatible batch size. Discarding 13 samples...
WARNING:neon.datasets.dataset:Incompatible batch size. Discarding 19 samples...
WARNING:neon.datasets.dataset:Incompatible batch size. Discarding 13 samples...
WARNING:neon.datasets.dataset:Incompatible batch size. Discarding 19 samples...
WARNING:neon.util.persist:deserializing object from:  /Users/tombocklisch/data/workout-cnn/workout-cnn.prm
INFO:neon.models.mlp:commencing model fitting
INFO:neon.models.mlp:epoch: 74, training error: 0.05319
WARNING:neon.util.persist:serializing object to: /Users/tombocklisch/data/workout-cnn/workout-cnn.prm






    




Finished epoch 74





    




INFO:neon.models.mlp:Layers:
	DataLayer layer: 3 x (400 x 1) nodes
	ConvLayer cv_1: 3 x (400 x 1) inputs, 16 x (396 x 1) nodes, RectLin act_fn
	PoolingLayer po_1: 16 x (396 x 1) inputs, 16 x (198 x 1) nodes, Linear act_fn
	ConvLayer cv_2: 16 x (198 x 1) inputs, 32 x (194 x 1) nodes, RectLin act_fn
	PoolingLayer po_2: 32 x (194 x 1) inputs, 32 x (97 x 1) nodes, Linear act_fn
	DropOutLayer do_1: 32 x (97 x 1) inputs, 32 x (97 x 1) nodes, Linear act_fn
	FCLayer fc_1: 3104 inputs, 100 nodes, RectLin act_fn
	DropOutLayer do_2: 100 inputs, 100 nodes, Linear act_fn
	FCLayer fc_2: 100 inputs, 29 nodes, Logistic act_fn
	CostLayer cost: 29 nodes, CrossEntropy cost_fn

WARNING:neon.datasets.dataset:Incompatible batch size. Discarding 13 samples...
WARNING:neon.datasets.dataset:Incompatible batch size. Discarding 19 samples...
WARNING:neon.datasets.dataset:Incompatible batch size. Discarding 13 samples...
WARNING:neon.datasets.dataset:Incompatible batch size. Discarding 19 samples...
WARNING:neon.util.persist:deserializing object from:  /Users/tombocklisch/data/workout-cnn/workout-cnn.prm
INFO:neon.models.mlp:commencing model fitting
INFO:neon.models.mlp:epoch: 75, training error: 0.05296
WARNING:neon.util.persist:serializing object to: /Users/tombocklisch/data/workout-cnn/workout-cnn.prm






    




Finished epoch 75
Finished training!

In [56]:

    

import numpy as np
import math
from matplotlib import pyplot, cm
from pylab import *
from IPython.html import widgets
from IPython.html.widgets import interact

def closestSqrt(i):
    N = int(math.sqrt(i))
    while True:
        M = int(i / N)
        if N * M == i:
            return N, M
        N -= 1
        
def plot_filters(**kwargs):
    n = kwargs['n']
    layer_name = kwargs['layer']
    dest_path = expanduser('~/data/workout-cnn/workout-ep' + str(n) + '.prm')
    params = pkl.load(open(dest_path, 'r'))

    wts = params[layer_name]['weights']

    nrows, ncols = closestSqrt(wts.shape[0])
    fr, fc = closestSqrt(wts.shape[1])
    
    fi = 0

    W = np.zeros((fr*nrows, fc*ncols))
    for row, col in [(row, col) for row in range(nrows) for col in range(ncols)]:
        W[fr*row:fr*(row+1):,fc*col:fc*(col+1)] = wts[fi].reshape(fr,fc)
        fi = fi + 1

    matshow(W, cmap=cm.gray)
    title('Visualizing weights of '+layer_name+' in epoch ' + str(n) )
    show()
    
layer_names = map(lambda l: l[1].name+"_"+str(l[0]), filter(lambda l: l[1].has_params, enumerate(model.layers)))

_i = interact(plot_filters,
              layer=widgets.widget_selection.ToggleButtons(options = layer_names),
              n=widgets.IntSliderWidget(description='epochs',
                                        min=0, max=max_epochs, value=0, step=epoch_step_size))


    

In [58]:

    

print "Lowest test error: %0.2f%%" % np.min(test_err)
print "Lowest train error: %0.2f%%" % np.min(train_err)

pyplot.plot(range(epoch_step_size*26, max_epochs+1, epoch_step_size), train_err, linewidth=3, label='train')
pyplot.plot(range(epoch_step_size*26, max_epochs+1, epoch_step_size), test_err, linewidth=3, label='test')
pyplot.grid()
pyplot.legend()
pyplot.xlabel("epoch")
pyplot.ylabel("error %")
pyplot.show()


    

    




Lowest test error: 9.80%
Lowest train error: 1.24%






    

In [54]:

    

from sklearn.metrics import confusion_matrix
from ipy_table import *

# confusion_matrix(y_true, y_pred)
predicted, actual = model.predict_fullset(dataset, "test")

y_pred = np.argmax(predicted.asnumpyarray(), axis = 0) 
y_true = np.argmax(actual.asnumpyarray(), axis = 0) 

confusion_mat = confusion_matrix(y_true, y_pred, range(0,dataset.num_labels))

# Fiddle around with cm to get it into table shape
confusion_mat = vstack((np.zeros((1,dataset.num_labels), dtype=int), confusion_mat))
confusion_mat = hstack((np.zeros((dataset.num_labels + 1, 1), dtype=int), confusion_mat))

table = confusion_mat.tolist()

human_labels = map(dataset.human_label_for, range(0,dataset.num_labels))

for i,s in enumerate(human_labels):
    table[0][i+1] = s
    table[i+1][0] = s

table[0][0] = "actual \ predicted"

mt = make_table(table)
set_row_style(0, color='lightGray', rotate = "315deg")
set_column_style(0, color='lightGray')
set_global_style(align='center')

for i in range(1, dataset.num_labels + 1):
    for j in range(1, dataset.num_labels + 1):
        if i == j:
            set_cell_style(i,j, color='lightGreen', width = 80)
        elif table[i][j] > 20:
            set_cell_style(i,j, color='Pink')
        elif table[i][j] > 0:
            set_cell_style(i,j, color='lightYellow')
mt


    

    
Out[54]:




actual \&nbsppredicted
dumbbell-chest-press
dumbbell-chest-fly
barbell-press
lat-pulldown-straight
twist
angle-chest-press
lat-pulldown-angled
cable-cross-overs
dumbbell-press
lateral-raise
side-dips
deltoid-row
leverage-high-row
pulldown-crunch
dumbbell-row
dumbbell-front-rise
triceps-dips
rope-triceps-extension
straight-bar-biceps-curl
cable-deltoid-cross-overs
rope-biceps-curl
straight-bar-triceps-extension
alt-dumbbell-biceps-curl
barbell-biceps-curl
dumbbell-side-rise
overhead-pull
barbell-pullup
rowing
dumbbell-biceps-curl
dumbbell-chest-press
294
0
0
0
0
0
0
0
0
0
0
0
0
0
0
0
0
0
0
0
0
0
0
0
0
0
0
0
0
dumbbell-chest-fly
0
484
0
0
0
0
0
0
0
0
0
0
0
0
0
0
0
0
0
0
0
0
0
0
0
0
0
0
0
barbell-press
0
0
249
0
0
0
0
0
0
2
0
0
0
0
0
0
0
0
0
0
0
0
0
0
0
0
0
0
0
lat-pulldown-straight
0
0
0
153
0
0
0
0
0
0
0
0
19
0
0
0
0
0
0
0
0
0
0
0
0
0
0
0
0
twist
0
0
0
0
467
0
0
0
0
0
0
0
0
0
0
0
0
0
9
0
0
0
0
0
1
0
0
0
0
angle-chest-press
0
0
0
0
0
451
0
0
0
0
0
0
0
0
0
0
0
0
0
0
0
0
0
0
0
0
0
0
0
lat-pulldown-angled
0
0
0
0
0
0
200
0
0
0
0
0
0
0
0
0
0
0
0
0
0
0
0
0
0
0
0
0
0
cable-cross-overs
0
0
0
0
0
0
0
208
0
0
0
0
0
0
0
0
0
0
0
0
0
0
0
0
0
0
0
0
0
dumbbell-press
137
0
23
0
0
0
0
0
152
0
0
0
0
0
2
39
0
0
0
0
0
0
0
0
56
2
0
0
0
lateral-raise
0
0
0
0
0
0
0
0
0
57
0
0
0
0
0
0
0
0
0
0
0
0
0
0
0
0
0
0
0
side-dips
0
0
0
0
0
0
0
0
0
0
117
0
0
0
0
0
0
0
0
0
0
0
0
0
0
0
0
0
0
deltoid-row
0
0
0
0
0
0
0
0
0
0
0
0
0
0
0
0
0
0
0
0
0
0
0
0
0
0
0
0
0
leverage-high-row
0
0
0
0
0
0
0
0
0
0
0
0
0
0
0
0
0
0
0
0
0
0
0
0
0
0
0
0
0
pulldown-crunch
0
0
0
0
0
0
0
0
0
0
0
0
0
679
0
0
0
0
0
0
0
0
0
0
0
0
0
0
0
dumbbell-row
0
0
0
0
0
0
0
0
0
0
0
0
0
0
1
0
0
0
0
0
0
0
0
0
0
0
0
0
0
dumbbell-front-rise
0
0
0
0
5
1
0
0
0
0
0
0
0
0
0
70
15
0
0
0
0
0
0
0
3
0
0
0
0
triceps-dips
0
0
0
0
0
0
0
0
0
0
0
0
0
0
0
0
233
0
0
0
0
0
0
0
0
0
0
0
0
rope-triceps-extension
0
0
0
0
0
0
0
0
0
0
0
0
0
0
0
0
0
148
0
0
0
0
0
0
0
0
0
0
0
straight-bar-biceps-curl
0
1
0
0
0
0
0
0
0
0
0
0
0
0
0
0
0
0
371
0
0
0
0
0
0
0
0
0
0
cable-deltoid-cross-overs
0
0
0
0
0
0
0
0
0
3
0
0
0
0
91
0
0
0
0
218
0
0
0
0
0
0
0
0
0
rope-biceps-curl
0
0
0
0
0
0
0
0
0
0
0
0
0
0
0
0
0
0
0
0
1
0
0
0
0
0
0
0
0
straight-bar-triceps-extension
0
0
0
0
0
0
0
0
0
0
0
0
0
0
0
0
0
0
0
0
0
1
0
0
0
0
0
0
0
alt-dumbbell-biceps-curl
0
0
0
0
0
0
0
0
0
0
0
0
0
0
0
0
0
0
0
0
0
0
261
3
0
0
0
0
0
barbell-biceps-curl
0
0
0
0
0
0
0
0
0
0
0
0
0
0
0
0
0
0
0
0
0
0
0
55
0
0
0
0
0
dumbbell-side-rise
0
0
0
0
0
0
0
0
0
0
0
0
0
0
0
0
0
0
0
0
0
0
0
0
0
0
0
0
0
overhead-pull
0
0
0
0
0
0
0
0
0
0
0
0
0
0
0
0
0
0
0
0
0
0
0
0
0
0
0
0
0
barbell-pullup
0
0
0
0
0
0
0
15
0
0
0
0
0
0
0
1
0
1
0
0
0
0
0
0
8
0
43
0
0
rowing
0
0
0
0
0
0
0
0
0
0
0
0
0
0
0
0
0
0
0
0
0
0
0
0
0
0
0
0
0
dumbbell-biceps-curl
0
11
0
0
23
0
0
0
0
0
0
0
0
0
2
0
0
0
58
0
0
0
0
45
1
0
0
0
0

In [ ]: