def mvnpdf(data, means, covs):
'''
Compute multivariate normal log pdf
Parameters
----------
Returns
-------
'''
logdets = [np.log(np.linalg.det(c)) for c in covs]
ichol_sigmas = [inv(chol(c)) for c in covs]
packed_params = util.pack_params(means, ichol_sigmas, logdets)
packed_data = util.pad_data(data)
return testmod.mvn_call(packed_data, packed_params, data.shape[1])
def mvnpdf(data, means, covs):
'''
Compute multivariate normal log pdf
Parameters
----------
Returns
-------
'''
logdets = [np.log(np.linalg.det(c)) for c in covs]
ichol_sigmas = [inv(chol(c)) for c in covs]
packed_params = util.pack_params(means, ichol_sigmas, logdets)
packed_data = util.pad_data(data)
return testmod.mvn_call(packed_data, packed_params,
data.shape[1])
MAX_WORD_LENGTH = config.word_length
wl = MAX_WORD_LENGTH
train_char, sl, wl = gen_data(texts, 0, wl, index_char)
val_char, sl, wl = gen_data(val_texts, sl, wl, index_char)
test_char, sl, wl = gen_data(test_texts, sl, wl, index_char)
MAX_SEQUENCE_LENGTH = sl
if MAX_SEQUENCE_LENGTH % 2 == 1:
MAX_SEQUENCE_LENGTH += 1
print(MAX_WORD_LENGTH)
print(MAX_SEQUENCE_LENGTH)
train_data_char = pad_data(train_char, MAX_SEQUENCE_LENGTH, MAX_WORD_LENGTH)
val_data_char = pad_data(val_char, MAX_SEQUENCE_LENGTH, MAX_WORD_LENGTH)
test_data_char = pad_data(test_char, MAX_SEQUENCE_LENGTH, MAX_WORD_LENGTH)
# print(np.shape(train_char))
labels = pad_label(labels, MAX_SEQUENCE_LENGTH)
val_labels = pad_label(val_labels, MAX_SEQUENCE_LENGTH)
test_labels = pad_label(test_labels, MAX_SEQUENCE_LENGTH)
num_chars = len(index_char)
model_char = Sequential()
# embedding shape (None/word num in a sent, max_char_num_in_word, char_emb_dim)
model_char.add(
Embedding(input_dim=num_chars,
print cpu_speed / gpu_speed
if __name__ == '__main__':
testmod.set_device(0)
n = 1e3
k = 16
data = randn(n, k).astype(np.float32)
mean = randn(k)
cov = np.array(util.random_cov(k), dtype=np.float32)
j = 32
padded_data = util.pad_data(data)
chol_sigma = chol(cov)
ichol_sigma = L.inv(chol_sigma)
logdet = np.log(np.linalg.det(cov))
means = (mean, ) * j
covs = (ichol_sigma, ) * j
logdets = (logdet, ) * j
packed_params = util.pack_params(means, covs, logdets)
cpu_func = lambda: testmod.cpu_mvnpdf(padded_data, packed_params, k
).squeeze()
gpu_func = lambda: testmod._mvnpdf(padded_data, packed_params, k).squeeze()
print 'GPU speed: %.3f' % (gpu_speed * 1000)
print cpu_speed / gpu_speed
if __name__ == '__main__':
testmod.set_device(0)
n = 1e3
k = 16
data = randn(n, k).astype(np.float32)
mean = randn(k)
cov = np.array(util.random_cov(k), dtype=np.float32)
j = 32
padded_data = util.pad_data(data)
chol_sigma = chol(cov)
ichol_sigma = L.inv(chol_sigma)
logdet = np.log(np.linalg.det(cov))
means = (mean,) * j
covs = (ichol_sigma,) * j
logdets = (logdet,) * j
packed_params = util.pack_params(means, covs, logdets)
cpu_func = lambda: testmod.cpu_mvnpdf(padded_data, packed_params, k).squeeze()
gpu_func = lambda: testmod._mvnpdf(padded_data, packed_params, k).squeeze()
print cpu_func()