def focal_loss(logits, labels, alpha, gamma=2, name='focal_loss'):
"""
Focal loss for multi classification
:param logits: A float32 tensor of shape [batch_size num_class].
:param labels: A int32 tensor of shape [batch_size, num_class] or [batch_size].
:param alpha: A 1D float32 tensor for focal loss alpha hyper-parameter
:param gamma: A scalar for focal loss gamma hyper-parameter.
Returns: A tensor of the same shape as `lables`
"""
if len(labels.shape) == 1:
labels = tf.one_hot(labels, logits.shape[-1])
else:
labels = labels
labels = tf.to_float(labels)
y_pred = tf.nn.softmax(logits, dim=-1)
L = -labels * tf.log(y_pred)
L *= alpha * ((1 - y_pred)**gamma)
loss = tf.reduce_sum(L)
if tf.executing_eagerly():
tf.contrib.summary.scalar(name, loss)
else:
tf.summary.scalar(name, loss)
return loss
def clip_gradients(self, grads_and_vars, clip_ratio, multitask=False):
"""Clip the gradients."""
is_zip_obj = False
if isinstance(grads_and_vars, zip):
grads_and_vars = list(grads_and_vars)
is_zip_obj = True
with tf.variable_scope('grad'):
for grad, var in grads_and_vars:
if grad is not None:
if tf.executing_eagerly():
tf.contrib.summary.histogram(var.name[:-2], grad)
else:
tf.summary.histogram(var.name[:-2], grad)
else:
logging.debug('%s gradient is None' % (var.name))
# not clip
if not clip_ratio:
if is_zip_obj:
grads, variables = zip(*grads_and_vars)
grads_and_vars = zip(grads, variables)
return grads_and_vars
if multitask:
grad_and_var_clipped, global_norm = tf.contrib.opt.clip_gradients_by_global_norm(
grads_and_vars, clip_ratio)
else:
gradients, variables = zip(*grads_and_vars)
clipped, global_norm = tf.clip_by_global_norm(gradients, clip_ratio)
grad_and_var_clipped = zip(clipped, variables)
if tf.executing_eagerly():
tf.contrib.summary.scalar('gradient/global_norm', global_norm)
else:
tf.summary.scalar('gradient/global_norm', global_norm)
return grad_and_var_clipped
def summary_writer(logdir,
graph=None,
max_queue=10,
flush_secs=120,
graph_def=None,
filename_suffix=None,
session=None,
name=None):
"""Summary writer."""
if tf.executing_eagerly():
return tf.contrib.summary.create_file_writer(
logdir,
max_queue=max_queue,
flush_millis=flush_secs * 1000,
filename_suffix=filename_suffix,
name=name)
return tf.summary.FileWriter(logdir, graph, max_queue, flush_secs,
graph_def, filename_suffix, session)
def test_fbank(self):
wav_path = str(Path(PACKAGE_OPS_DIR).joinpath('data/sm1_cln.wav'))
with self.cached_session(use_gpu=False, force_gpu=False):
read_wav = ReadWav.params().instantiate()
input_data, sample_rate = read_wav(wav_path)
config = {
'window_length': 0.025,
'output_type': 1,
'frame_length': 0.010,
'snip_edges': True
}
fbank = Fbank.params(config).instantiate()
fbank_test = fbank(input_data, sample_rate)
self.assertEqual(tf.rank(fbank_test).eval(), 3)
if tf.executing_eagerly():
print(fbank_test.numpy()[0:2, 0:6, 0])
else:
print(fbank_test.eval()[0:2, 0:6, 0])
def testSomeTFSymbols(self):
self.assertFalse(tf.executing_eagerly())
self.assertIsNotNone(tf.logging)
self.assertIsNotNone(tf.flags)
self.assertIs(tf.Defun, function.Defun)
def image(name, tensor, max_images=3):
"Image"
if tf.executing_eagerly():
tf.contrib.summary.image(name, tensor, max_images=max_images)
else:
tf.summary.image(name, tensor, max_outputs=max_images)
def audio(name, tensor, sample_rate, max_outputs=3):
"Audio"
if tf.executing_eagerly():
tf.contrib.summary.audio(name, tensor, sample_rate, max_outputs)
else:
tf.summary.audio(name, tensor, sample_rate, max_outputs)
def text(name, tensor):
"Text"
if tf.executing_eagerly():
tf.contrib.summary.text(name, tensor)
else:
tf.summary.text(name, tensor)
def histogram(name, values):
"Histogram"
if tf.executing_eagerly():
tf.contrib.summary.histogram(name, values)
else:
tf.summary.histogram(name, values)
def scalar(name, value): # pylint: redefined-outer-name
"Scalar"
if tf.executing_eagerly():
tf.contrib.summary.scalar(name, value)
else:
tf.summary.scalar(name, value)
def flush(writer=None, name=None):
"""Flush"""
if tf.executing_eagerly():
tf.contrib.summary.flush(writer, name)
else:
tf.summary.flush(writer, name) # pylint: disable=no-member