我们从Python开源项目中，提取了以下32个代码示例，用于说明如何使用chainer.functions.softplus()。
def to_function(self): if self.nonlinearity.lower() == "clipped_relu": return clipped_relu() if self.nonlinearity.lower() == "crelu": return crelu() if self.nonlinearity.lower() == "elu": return elu() if self.nonlinearity.lower() == "hard_sigmoid": return hard_sigmoid() if self.nonlinearity.lower() == "leaky_relu": return leaky_relu() if self.nonlinearity.lower() == "relu": return relu() if self.nonlinearity.lower() == "sigmoid": return sigmoid() if self.nonlinearity.lower() == "softmax": return softmax() if self.nonlinearity.lower() == "softplus": return softplus() if self.nonlinearity.lower() == "tanh": return tanh() if self.nonlinearity.lower() == "bst": return bst() raise NotImplementedError()
def to_function(self): if self.nonlinearity.lower() == "clipped_relu": return clipped_relu() if self.nonlinearity.lower() == "crelu": return crelu() if self.nonlinearity.lower() == "elu": return elu() if self.nonlinearity.lower() == "hard_sigmoid": return hard_sigmoid() if self.nonlinearity.lower() == "leaky_relu": return leaky_relu() if self.nonlinearity.lower() == "relu": return relu() if self.nonlinearity.lower() == "sigmoid": return sigmoid() if self.nonlinearity.lower() == "softmax": return softmax() if self.nonlinearity.lower() == "softplus": return softplus() if self.nonlinearity.lower() == "tanh": return tanh() raise NotImplementedError()
def loss_func_dcgan_dis_real(y_real): return F.sum(F.softplus(-y_real)) / np.prod(y_real.data.shape)
def loss_func_dcgan_dis_fake(y_fake): return F.sum(F.softplus(y_fake)) / np.prod(y_fake.data.shape)
def loss_sigmoid_cross_entropy_with_logits(x, t): return F.average(x - x*t + F.softplus(-x))# / x.data.shape[0]
def compute_mean_and_var(self, x): h = x for layer in self.hidden_layers: h = self.nonlinearity(layer(h)) mean = self.mean_layer(h) if self.bound_mean: mean = bound_by_tanh(mean, self.min_action, self.max_action) var = F.broadcast_to(F.softplus(self.var_layer(h)), mean.shape) + \ self.min_var return mean, var
def __call__(self, x): mean = self.hidden_layers(x) var = F.broadcast_to( F.softplus(self.var_param), mean.shape) return distribution.GaussianDistribution(mean, var)
def compute_mean_and_var(self, x): # mean = self.mean_layer(x) mean = F.tanh(self.mean_layer(x)) * 2.0 var = F.softplus(self.var_layer(x)) return mean, var
def compute_mean_and_var(self, x): # mean = self.mean_layer(x) mean = F.tanh(self.mean_layer(x)) * 2.0 var = F.softplus(F.broadcast_to(self.var_layer(x), mean.data.shape)) return mean, var
def __call__(self, x): return functions.softplus(x, self.beta)
def check_forward(self, x_data): x = chainer.Variable(x_data) y = functions.softplus(x, beta=self.beta) x_value = cuda.to_cpu(x_data) y_exp = numpy.log(1 + numpy.exp(self.beta * x_value)) / self.beta self.assertEqual(y.data.dtype, self.dtype) gradient_check.assert_allclose( y_exp, y.data, **self.check_forward_options)
def __init__(self, use_cudnn=True): self._function = "softplus" self.use_cudnn = use_cudnn
def __call__(self, x): return F.softplus(x, self.use_cudnn)
def bernoulli_nll_keepbatch(self, x, y): nll = F.softplus(y) - x * y return F.sum(nll, axis=1)
def __init__(self): self.image_width = 28 self.image_height = 28 self.ndim_x = 28 * 28 self.ndim_y = 10 self.ndim_z = 50 # True : y = f(BN(Wx + b)) # False: y = f(W*BN(x) + b) self.batchnorm_before_activation = True # gaussianmarg | gaussian self.type_pz = "gaussianmarg" self.type_qz = "gaussianmarg" self.encoder_xy_z_hidden_units = [500] self.encoder_xy_z_activation_function = "softplus" self.encoder_xy_z_apply_dropout = False self.encoder_xy_z_apply_batchnorm = True self.encoder_xy_z_apply_batchnorm_to_input = True self.encoder_x_y_hidden_units = [500] self.encoder_x_y_activation_function = "softplus" self.encoder_x_y_apply_dropout = False self.encoder_x_y_apply_batchnorm = True self.encoder_x_y_apply_batchnorm_to_input = True self.decoder_hidden_units = [500] self.decoder_activation_function = "softplus" self.decoder_apply_dropout = False self.decoder_apply_batchnorm = True self.decoder_apply_batchnorm_to_input = True self.gpu_enabled = True self.learning_rate = 0.0003 self.gradient_momentum = 0.9 self.gradient_clipping = 5.0
def __init__(self, **layers): super(SoftmaxEncoder, self).__init__(**layers) self.activation_function = "softplus" self.apply_batchnorm_to_input = True self.apply_batchnorm = True self.apply_dropout = False self.batchnorm_before_activation = True
def __init__(self, **layers): super(GaussianEncoder, self).__init__(**layers) self.activation_function = "softplus" self.apply_batchnorm_to_input = True self.apply_batchnorm = True self.apply_dropout = False self.batchnorm_before_activation = True
def __init__(self): self.image_width = 28 self.image_height = 28 self.ndim_x = 28 * 28 self.ndim_z = 100 self.batchnorm_before_activation = True # gaussianmarg | gaussian # We recommend you to use "gaussianmarg" when decoder is gaussian. self.type_pz = "gaussianmarg" self.type_qz = "gaussianmarg" # e.g. # ndim_x (input) -> 2000 -> 1000 -> 100 (output) # encoder_hidden_units = [2000, 1000] self.encoder_hidden_units = [600, 600] self.encoder_activation_function = "softplus" self.encoder_apply_dropout = True self.encoder_apply_batchnorm = True self.encoder_apply_batchnorm_to_input = True self.decoder_hidden_units = [600, 600] self.decoder_activation_function = "softplus" self.decoder_apply_dropout = True self.decoder_apply_batchnorm = True self.decoder_apply_batchnorm_to_input = True self.gpu_enabled = True self.learning_rate = 0.0003 self.gradient_momentum = 0.9 self.gradient_clipping = 1.0
def __init__(self, **layers): super(Encoder, self).__init__(**layers) self.activation_function = "softplus" self.apply_batchnorm_to_input = True self.apply_batchnorm = True self.apply_dropout = True self.batchnorm_before_activation = True
def __init__(self, **layers): super(BernoulliDecoder, self).__init__(**layers) self.activation_function = "softplus" self.apply_batchnorm_to_input = True self.apply_batchnorm = True self.apply_dropout = True self.batchnorm_before_activation = True
def loss_cnn(self, cnn, x_out, dst, dis_out, lam1=100, lam2=1): loss_rec = lam1 * ( F.mean_absolute_error(x_out, dst) ) batchsize,_,w,h = dis_out.data.shape loss_adv = lam2 * F.sum( F.softplus(-dis_out) ) / batchsize / w / h loss = loss_rec + loss_adv chainer.report({'loss': loss,"loss_rec":loss_rec, 'loss_adv': loss_adv }, cnn) return loss
def loss_dis(self, dis, dis_real, dis_fake): batchsize,_,w,h = dis_real.data.shape L1 = (2+np.random.rand()) * F.sum(F.softplus(-dis_real)) / batchsize / w / h L2 = (2+np.random.rand()) * F.sum(F.softplus(dis_fake)) / batchsize / w / h loss = L1 + L2 chainer.report({'loss': loss}, dis) return loss
