def __init__(self, params, learning_rate=0.1, decay=0.9, momentum=0.0, epsilon=1e-10,
use_locking=False, centered=False, loss_scale=1.0, weight_decay=0.0):
super(RMSProp, self).__init__(learning_rate, params, weight_decay, loss_scale)
validator.check_value_type("decay", decay, [float], self.cls_name)
validator.check_non_negative_float(decay, "decay", self.cls_name)
validator.check_value_type("momentum", momentum, [float], self.cls_name)
validator.check_non_negative_float(momentum, "momentum", self.cls_name)
validator.check_value_type("epsilon", epsilon, [float], self.cls_name)
validator.check_positive_float(epsilon, "epsilon", self.cls_name)
validator.check_value_type("use_locking", use_locking, [bool], self.cls_name)
validator.check_value_type("centered", centered, [bool], self.cls_name)
self.centered = centered
if centered:
self.opt = P.ApplyCenteredRMSProp(use_locking)
self.mg = self.parameters.clone(prefix="mean_grad", init='zeros')
else:
self.opt = P.ApplyRMSProp(use_locking)
self.momentum = momentum
self.ms = self.parameters.clone(prefix="mean_square", init='ones')
self.moment = self.parameters.clone(prefix="moment", init='zeros')
self.hyper_map = C.HyperMap()
self.epsilon = epsilon
self.decay = decay
def __init__(self, lr, decay, momentum, epsilon):
super(NetRMSProp, self).__init__()
self.lr = lr
self.decay = decay
self.momentum = momentum
self.epsilon = epsilon
self.rms_opt = P.ApplyRMSProp()
def __init__(self, use_centered):
super(NetRMSProp, self).__init__()
self.use_centered = use_centered
if use_centered:
self.rms_opt = P.ApplyCenteredRMSProp()
else:
self.rms_opt = P.ApplyRMSProp()
def __init__(self,
params,
learning_rate=0.1,
decay=0.9,
momentum=0.0,
epsilon=1e-10,
use_locking=False,
centered=False,
loss_scale=1.0,
weight_decay=0.0,
decay_filter=lambda x: 'beta' not in x.name and 'gamma' not in
x.name):
super(RMSProp, self).__init__(learning_rate, params)
if isinstance(momentum, float) and momentum < 0.0:
raise ValueError(
"momentum should be at least 0.0, but got momentum {}".format(
momentum))
if decay < 0.0:
raise ValueError(
"decay should be at least 0.0, but got dampening {}".format(
decay))
self.decay = decay
self.epsilon = epsilon
validator.check_type("use_locking", use_locking, [bool])
validator.check_type("centered", centered, [bool])
self.centered = centered
if centered:
self.opt = P.ApplyCenteredRMSProp(use_locking)
self.mg = self.parameters.clone(prefix="mean_grad", init='zeros')
else:
self.opt = P.ApplyRMSProp(use_locking)
self.dynamic_lr = False
if not isinstance(learning_rate, float):
self.dynamic_lr = True
self.gather = P.GatherV2()
self.assignadd = P.AssignAdd()
self.global_step = Parameter(initializer(0, [1], mstype.int32),
name="global_step")
self.axis = 0
self.one = Tensor(1, mstype.int32)
self.momentum = momentum
self.ms = self.parameters.clone(prefix="mean_square", init='zeros')
self.moment = self.parameters.clone(prefix="moment", init='zeros')
self.hyper_map = C.HyperMap()
self.decay = decay
self.decay_tf = tuple(decay_filter(x) for x in self.parameters)
self.reciprocal_scale = 1.0 / loss_scale
self.weight_decay = weight_decay * loss_scale
def __init__(self, lr, decay, momentum, epsilon, var, g, mg, rms, mom):
super(NetRMSProp, self).__init__()
self.lr = lr
self.decay = decay
self.momentum = momentum
self.epsilon = epsilon
self.var = var
self.g = g
self.mg = mg
self.rms = rms
self.mom = mom
self.rms_opt = P.ApplyRMSProp()
def __init__(self,
params,
learning_rate=0.1,
decay=0.9,
momentum=0.0,
epsilon=1e-10,
use_locking=False,
centered=False,
loss_scale=1.0,
weight_decay=0.0,
decay_filter=lambda x: 'beta' not in x.name and 'gamma' not in
x.name):
super(RMSProp, self).__init__(learning_rate, params, weight_decay,
loss_scale, decay_filter)
if isinstance(momentum, float) and momentum < 0.0:
raise ValueError(
"momentum should be at least 0.0, but got momentum {}".format(
momentum))
if decay < 0.0:
raise ValueError(
"decay should be at least 0.0, but got dampening {}".format(
decay))
self.decay = decay
self.epsilon = epsilon
validator.check_value_type("use_locking", use_locking, [bool],
self.cls_name)
validator.check_value_type("centered", centered, [bool], self.cls_name)
self.centered = centered
if centered:
self.opt = P.ApplyCenteredRMSProp(use_locking)
self.mg = self.parameters.clone(prefix="mean_grad", init='zeros')
else:
self.opt = P.ApplyRMSProp(use_locking)
self.momentum = momentum
self.ms = self.parameters.clone(prefix="mean_square", init='zeros')
self.moment = self.parameters.clone(prefix="moment", init='zeros')
self.hyper_map = C.HyperMap()
self.decay = decay
def __init__(self,
params,
learning_rate=0.1,
decay=0.9,
momentum=0.0,
epsilon=1e-10,
use_locking=False,
centered=False,
loss_scale=1.0,
weight_decay=0.0,
decay_filter=lambda x: 'beta' not in x.name and 'gamma' not in
x.name):
super(RMSProp, self).__init__(learning_rate, params, weight_decay,
loss_scale, decay_filter)
validator.check_value_type("decay", decay, [float], self.cls_name)
validator.check_number_range("decay", decay, 0.0, float("inf"),
Rel.INC_LEFT, self.cls_name)
validator.check_value_type("momentum", momentum, [float],
self.cls_name)
validator.check_number_range("momentum", momentum, 0.0, float("inf"),
Rel.INC_LEFT, self.cls_name)
validator.check_value_type("epsilon", epsilon, [float], self.cls_name)
validator.check_number_range("epsilon", epsilon, 0.0, float("inf"),
Rel.INC_NEITHER, self.cls_name)
validator.check_value_type("use_locking", use_locking, [bool],
self.cls_name)
validator.check_value_type("centered", centered, [bool], self.cls_name)
self.centered = centered
if centered:
self.opt = P.ApplyCenteredRMSProp(use_locking)
self.mg = self.parameters.clone(prefix="mean_grad", init='zeros')
else:
self.opt = P.ApplyRMSProp(use_locking)
self.momentum = momentum
self.ms = self.parameters.clone(prefix="mean_square", init='zeros')
self.moment = self.parameters.clone(prefix="moment", init='zeros')
self.hyper_map = C.HyperMap()
self.epsilon = epsilon
self.decay = decay
'desc_const': [0],
'desc_inputs': [[3, 2]],
'desc_bprop': [[1, 2]]}),
('CumProd', {
'block': P.CumProd(),
'desc_const': [0],
'desc_inputs': [[3, 2]],
'desc_bprop': [[3, 2]]}),
('ApplyFtrl', {
'block': P.ApplyFtrl(),
'desc_const': [0.001, 0.0, 0.0, -0.5],
'desc_inputs': [[3, 3], [3, 3], [3, 3], [3, 3]],
'desc_bprop': [3, 3],
'skip': ['backward']}),
('ApplyRMSProp', {
'block': P.ApplyRMSProp(),
'desc_const': [0.9, 0.0, 1e-10, 0.001],
'desc_inputs': [[3, 3], [3, 3], [3, 3], [3, 3]],
'desc_bprop': [3, 3],
'skip': ['backward']}),
('ApplyCenteredRMSProp', {
'block': P.ApplyCenteredRMSProp(),
'desc_const': [0.9, 0.0, 1e-10, 0.001],
'desc_inputs': [[3, 3], [3, 3], [3, 3], [3, 3], [3, 3]],
'desc_bprop': [3, 3],
'skip': ['backward']}),
('L2Loss_1', {
'block': P.L2Loss(),
'desc_inputs': [Tensor(np.array([1, 2, 3, 4]), mstype.float16)],
'desc_bprop': []}),
('L2Loss_2', {
def __init__(self, var):
super(ApplyRMSPropNet, self).__init__()
self.apply_rms_prop = P.ApplyRMSProp()
self.var = Parameter(var, name="var")
