def test_maxout_layer():
random_state = np.random.RandomState(42)
graph = OrderedDict()
X_sym, y_sym = add_datasets_to_graph([X, y], ["X", "y"], graph)
single_o = maxout_layer([X_sym], graph, 'single', proj_dim=5,
random_state=random_state)
concat_o = maxout_layer([X_sym, y_sym], graph, 'concat', proj_dim=5,
random_state=random_state)
# Check that strict mode raises an error if repeated
assert_raises(AssertionError, maxout_layer, [X_sym], graph, 'concat')
f = theano.function([X_sym, y_sym], [single_o, concat_o],
mode="FAST_COMPILE")
single, concat = f(X, y)
valid_indices = data["valid_indices"]
minibatch_size = 10
X_mb, X_mb_mask = make_masked_minibatch(X, slice(0, minibatch_size))
y_mb, y_mb_mask = make_masked_minibatch(y, slice(0, minibatch_size))
n_hid = 500
n_out = vocab_size + 1
datasets_list = [X_mb, X_mb_mask, y_mb, y_mb_mask]
names_list = ["X", "X_mask", "y", "y_mask"]
X_sym, X_mask_sym, y_sym, y_mask_sym = add_datasets_to_graph(
datasets_list, names_list, graph)
l1 = maxout_layer([X_sym], graph, 'l1', n_hid, random_state=random_state)
h = bidirectional_gru_recurrent_layer([l1], X_mask_sym, n_hid, graph,
'l1_rec', random_state=random_state)
l2 = maxout_layer([h], graph, 'l2', n_hid, random_state=random_state)
y_pred = softmax_layer([l2], graph, 'softmax', n_out, random_state=random_state)
cost = log_ctc_cost(y_sym, y_mask_sym, y_pred, X_mask_sym).mean()
params, grads = get_params_and_grads(graph, cost)
opt = adadelta(params)
updates = opt.updates(params, grads)
checkpoint_dict = {}
fit_function = theano.function([X_sym, X_mask_sym, y_sym, y_mask_sym], [cost],
updates=updates)
valid_indices = data["valid_indices"]
minibatch_size = 10
X_mb, X_mb_mask = make_masked_minibatch(X, slice(0, minibatch_size))
y_mb, y_mb_mask = make_masked_minibatch(y, slice(0, minibatch_size))
n_hid = 500
n_out = vocab_size + 1
datasets_list = [X_mb, X_mb_mask, y_mb, y_mb_mask]
names_list = ["X", "X_mask", "y", "y_mask"]
X_sym, X_mask_sym, y_sym, y_mask_sym = add_datasets_to_graph(
datasets_list, names_list, graph)
l1 = maxout_layer([X_sym], graph, 'l1', n_hid, random_state=random_state)
h = bidirectional_gru_recurrent_layer([l1],
X_mask_sym,
n_hid,
graph,
'l1_rec',
random_state=random_state)
l2 = maxout_layer([h], graph, 'l2', n_hid, random_state=random_state)
y_pred = softmax_layer([l2],
graph,
'softmax',
n_out,
random_state=random_state)
cost = log_ctc_cost(y_sym, y_mask_sym, y_pred, X_mask_sym).mean()
params, grads = get_params_and_grads(graph, cost)