Loading deepchem/models/tensorgraph/models/gan.py +40 −61 Original line number Diff line number Diff line Loading @@ -66,16 +66,13 @@ class GAN(TensorGraph): for shape in self.get_data_input_shapes(): self.data_inputs.append(layers.Feature(shape=shape)) self.conditional_inputs = [] self.noise_conditional_inputs = [] for shape in self.get_conditional_input_shapes(): self.conditional_inputs.append(layers.Feature(shape=shape)) self.noise_conditional_inputs.append(layers.Feature(shape=shape)) self.is_training = layers.Weights(shape=(None, 1)) # Create the generator. self.generator = self.create_generator(self.noise_input, self.noise_conditional_inputs) self.conditional_inputs) if not isinstance(self.generator, Sequence): raise ValueError('create_generator() must return a list of Layers') if len(self.generator) != len(self.data_inputs): Loading @@ -92,14 +89,18 @@ class GAN(TensorGraph): # Create the discriminator. self._discrim_data = [] self.discrim_train = self.create_discriminator(self.data_inputs, self.conditional_inputs) # Make a copy of the discriminator that takes the generator's output as # its input. replacements = {} for g, d in zip(self.generator, self.data_inputs): self._discrim_data.append(layers.Concat([g, d], axis=0)) self._discrim_conditional = [] for n, c in zip(self.noise_conditional_inputs, self.conditional_inputs): self._discrim_conditional.append(layers.Concat([n, c], axis=0)) self.discriminator = self.create_discriminator(self._discrim_data, self._discrim_conditional) replacements[d] = g for c in self.conditional_inputs: replacements[c] = c self.discrim_gen = self.discrim_train.copy(replacements, shared=True) # Make a list of all layers in the generator and discriminator. Loading @@ -113,14 +114,14 @@ class GAN(TensorGraph): for layer in self.generator: add_layers_to_set(layer, gen_layers) discrim_layers = set() add_layers_to_set(self.discriminator, discrim_layers) add_layers_to_set(self.discrim_train, discrim_layers) discrim_layers -= gen_layers # Create submodels for training the generator and discriminator. gen_loss = self.create_generator_loss(self.discriminator, self.is_training) discrim_loss = self.create_discriminator_loss(self.discriminator, self.is_training) gen_loss = self.create_generator_loss(self.discrim_gen) discrim_loss = self.create_discriminator_loss(self.discrim_train, self.discrim_gen) self.generator_submodel = self.create_submodel( layers=gen_layers, loss=gen_loss) self.discriminator_submodel = self.create_submodel( Loading Loading @@ -221,7 +222,7 @@ class GAN(TensorGraph): """ raise NotImplementedError("Subclasses must implement this.") def create_generator_loss(self, discrim_output, is_training): def create_generator_loss(self, discrim_output): """Create the loss function for the generator. The default implementation is appropriate for most cases. Subclasses can Loading @@ -230,21 +231,17 @@ class GAN(TensorGraph): Parameters ---------- discrim_output: Layer the discriminator's output layer, which computes the probability that each sample is training data. is_training: Layer outputs a set of flags indicating whether each sample is actually training data (1) or generated data (0). the output from the discriminator on a batch of generated data. This is its estimate of the probability that each sample is training data. Returns ------- A Layer object that outputs the loss function to use for optimizing the generator. """ prob = discrim_output + 1e-10 return -layers.ReduceMean(layers.Log(prob) * (1 - is_training)) return -layers.ReduceMean(layers.Log(discrim_output + 1e-10)) def create_discriminator_loss(self, discrim_output, is_training): def create_discriminator_loss(self, discrim_output_train, discrim_output_gen): """Create the loss function for the discriminator. The default implementation is appropriate for most cases. Subclasses can Loading @@ -252,20 +249,20 @@ class GAN(TensorGraph): Parameters ---------- discrim_output: Layer the discriminator's output layer, which computes the probability that each sample is training data. is_training: Layer outputs a set of flags indicating whether each sample is actually training data (1) or generated data (0). discrim_output_train: Layer the output from the discriminator on a batch of generated data. This is its estimate of the probability that each sample is training data. discrim_output_gen: Layer the output from the discriminator on a batch of training data. This is its estimate of the probability that each sample is training data. Returns ------- A Layer object that outputs the loss function to use for optimizing the discriminator. """ training_data_loss = layers.Log(discrim_output + 1e-10) * is_training gen_data_loss = layers.Log(1 - discrim_output + 1e-10) * (1 - is_training) training_data_loss = layers.Log(discrim_output_train + 1e-10) gen_data_loss = layers.Log(1 - discrim_output_gen + 1e-10) return -layers.ReduceMean(training_data_loss + gen_data_loss) def fit_gan(self, Loading Loading @@ -317,26 +314,10 @@ class GAN(TensorGraph): global_step = self.global_step # Train the discriminator on training data. # Train the discriminator. feed_dict = dict(feed_dict) feed_dict[self.noise_input] = self.get_noise_batch(0) feed_dict[self.is_training] = np.ones((self.batch_size, 1)) self._set_empty_inputs(feed_dict, self.noise_conditional_inputs) discrim_error += self.fit_generator( [feed_dict], submodel=self.discriminator_submodel, checkpoint_interval=0) self.global_step = global_step # Train the discriminator on generated data. feed_dict[self.noise_input] = self.get_noise_batch(self.batch_size) feed_dict[self.is_training] = np.zeros((self.batch_size, 1)) for n, c in zip(self.noise_conditional_inputs, self.conditional_inputs): feed_dict[n] = feed_dict[c] self._set_empty_inputs(feed_dict, self.data_inputs) self._set_empty_inputs(feed_dict, self.conditional_inputs) discrim_error += self.fit_generator( [feed_dict], submodel=self.discriminator_submodel, Loading Loading @@ -419,7 +400,7 @@ class GAN(TensorGraph): noise_input = self.get_noise_batch(batch_size) batch = {} batch[self.noise_input] = noise_input for layer, value in zip(self.noise_conditional_inputs, conditional_inputs): for layer, value in zip(self.conditional_inputs, conditional_inputs): batch[layer] = value return self.predict_on_generator([batch]) Loading Loading @@ -484,26 +465,24 @@ class WGAN(GAN): super(WGAN, self).__init__(**kwargs) self.gradient_penalty = gradient_penalty def create_generator_loss(self, discrim_output, is_training): return layers.ReduceMean(discrim_output * (1 - is_training)) def create_generator_loss(self, discrim_output): return layers.ReduceMean(discrim_output) def create_discriminator_loss(self, discrim_output, is_training): training_data_loss = discrim_output * is_training gen_data_loss = -discrim_output * (1 - is_training) gradient_penalty = GradientPenaltyLayer(discrim_output, self) return gradient_penalty + layers.ReduceMean(training_data_loss + gen_data_loss) def create_discriminator_loss(self, discrim_output_train, discrim_output_gen): gradient_penalty = GradientPenaltyLayer(discrim_output_train, self) return gradient_penalty + layers.ReduceMean(discrim_output_train - discrim_output_gen) class GradientPenaltyLayer(layers.Layer): """Implements the gradient penalty loss term for WGANs.""" def __init__(self, discrim_output, gan): super(GradientPenaltyLayer, self).__init__(discrim_output) def __init__(self, discrim_output_train, gan): super(GradientPenaltyLayer, self).__init__([discrim_output_train]) self.gan = gan def create_tensor(self, in_layers=None, set_tensors=True, **kwargs): gradients = tf.gradients(self.in_layers[0], self.gan._discrim_data) gradients = tf.gradients(self.in_layers[0], self.gan.data_inputs) norm2 = 0.0 for g in gradients: g2 = tf.square(g) Loading Loading
deepchem/models/tensorgraph/models/gan.py +40 −61 Original line number Diff line number Diff line Loading @@ -66,16 +66,13 @@ class GAN(TensorGraph): for shape in self.get_data_input_shapes(): self.data_inputs.append(layers.Feature(shape=shape)) self.conditional_inputs = [] self.noise_conditional_inputs = [] for shape in self.get_conditional_input_shapes(): self.conditional_inputs.append(layers.Feature(shape=shape)) self.noise_conditional_inputs.append(layers.Feature(shape=shape)) self.is_training = layers.Weights(shape=(None, 1)) # Create the generator. self.generator = self.create_generator(self.noise_input, self.noise_conditional_inputs) self.conditional_inputs) if not isinstance(self.generator, Sequence): raise ValueError('create_generator() must return a list of Layers') if len(self.generator) != len(self.data_inputs): Loading @@ -92,14 +89,18 @@ class GAN(TensorGraph): # Create the discriminator. self._discrim_data = [] self.discrim_train = self.create_discriminator(self.data_inputs, self.conditional_inputs) # Make a copy of the discriminator that takes the generator's output as # its input. replacements = {} for g, d in zip(self.generator, self.data_inputs): self._discrim_data.append(layers.Concat([g, d], axis=0)) self._discrim_conditional = [] for n, c in zip(self.noise_conditional_inputs, self.conditional_inputs): self._discrim_conditional.append(layers.Concat([n, c], axis=0)) self.discriminator = self.create_discriminator(self._discrim_data, self._discrim_conditional) replacements[d] = g for c in self.conditional_inputs: replacements[c] = c self.discrim_gen = self.discrim_train.copy(replacements, shared=True) # Make a list of all layers in the generator and discriminator. Loading @@ -113,14 +114,14 @@ class GAN(TensorGraph): for layer in self.generator: add_layers_to_set(layer, gen_layers) discrim_layers = set() add_layers_to_set(self.discriminator, discrim_layers) add_layers_to_set(self.discrim_train, discrim_layers) discrim_layers -= gen_layers # Create submodels for training the generator and discriminator. gen_loss = self.create_generator_loss(self.discriminator, self.is_training) discrim_loss = self.create_discriminator_loss(self.discriminator, self.is_training) gen_loss = self.create_generator_loss(self.discrim_gen) discrim_loss = self.create_discriminator_loss(self.discrim_train, self.discrim_gen) self.generator_submodel = self.create_submodel( layers=gen_layers, loss=gen_loss) self.discriminator_submodel = self.create_submodel( Loading Loading @@ -221,7 +222,7 @@ class GAN(TensorGraph): """ raise NotImplementedError("Subclasses must implement this.") def create_generator_loss(self, discrim_output, is_training): def create_generator_loss(self, discrim_output): """Create the loss function for the generator. The default implementation is appropriate for most cases. Subclasses can Loading @@ -230,21 +231,17 @@ class GAN(TensorGraph): Parameters ---------- discrim_output: Layer the discriminator's output layer, which computes the probability that each sample is training data. is_training: Layer outputs a set of flags indicating whether each sample is actually training data (1) or generated data (0). the output from the discriminator on a batch of generated data. This is its estimate of the probability that each sample is training data. Returns ------- A Layer object that outputs the loss function to use for optimizing the generator. """ prob = discrim_output + 1e-10 return -layers.ReduceMean(layers.Log(prob) * (1 - is_training)) return -layers.ReduceMean(layers.Log(discrim_output + 1e-10)) def create_discriminator_loss(self, discrim_output, is_training): def create_discriminator_loss(self, discrim_output_train, discrim_output_gen): """Create the loss function for the discriminator. The default implementation is appropriate for most cases. Subclasses can Loading @@ -252,20 +249,20 @@ class GAN(TensorGraph): Parameters ---------- discrim_output: Layer the discriminator's output layer, which computes the probability that each sample is training data. is_training: Layer outputs a set of flags indicating whether each sample is actually training data (1) or generated data (0). discrim_output_train: Layer the output from the discriminator on a batch of generated data. This is its estimate of the probability that each sample is training data. discrim_output_gen: Layer the output from the discriminator on a batch of training data. This is its estimate of the probability that each sample is training data. Returns ------- A Layer object that outputs the loss function to use for optimizing the discriminator. """ training_data_loss = layers.Log(discrim_output + 1e-10) * is_training gen_data_loss = layers.Log(1 - discrim_output + 1e-10) * (1 - is_training) training_data_loss = layers.Log(discrim_output_train + 1e-10) gen_data_loss = layers.Log(1 - discrim_output_gen + 1e-10) return -layers.ReduceMean(training_data_loss + gen_data_loss) def fit_gan(self, Loading Loading @@ -317,26 +314,10 @@ class GAN(TensorGraph): global_step = self.global_step # Train the discriminator on training data. # Train the discriminator. feed_dict = dict(feed_dict) feed_dict[self.noise_input] = self.get_noise_batch(0) feed_dict[self.is_training] = np.ones((self.batch_size, 1)) self._set_empty_inputs(feed_dict, self.noise_conditional_inputs) discrim_error += self.fit_generator( [feed_dict], submodel=self.discriminator_submodel, checkpoint_interval=0) self.global_step = global_step # Train the discriminator on generated data. feed_dict[self.noise_input] = self.get_noise_batch(self.batch_size) feed_dict[self.is_training] = np.zeros((self.batch_size, 1)) for n, c in zip(self.noise_conditional_inputs, self.conditional_inputs): feed_dict[n] = feed_dict[c] self._set_empty_inputs(feed_dict, self.data_inputs) self._set_empty_inputs(feed_dict, self.conditional_inputs) discrim_error += self.fit_generator( [feed_dict], submodel=self.discriminator_submodel, Loading Loading @@ -419,7 +400,7 @@ class GAN(TensorGraph): noise_input = self.get_noise_batch(batch_size) batch = {} batch[self.noise_input] = noise_input for layer, value in zip(self.noise_conditional_inputs, conditional_inputs): for layer, value in zip(self.conditional_inputs, conditional_inputs): batch[layer] = value return self.predict_on_generator([batch]) Loading Loading @@ -484,26 +465,24 @@ class WGAN(GAN): super(WGAN, self).__init__(**kwargs) self.gradient_penalty = gradient_penalty def create_generator_loss(self, discrim_output, is_training): return layers.ReduceMean(discrim_output * (1 - is_training)) def create_generator_loss(self, discrim_output): return layers.ReduceMean(discrim_output) def create_discriminator_loss(self, discrim_output, is_training): training_data_loss = discrim_output * is_training gen_data_loss = -discrim_output * (1 - is_training) gradient_penalty = GradientPenaltyLayer(discrim_output, self) return gradient_penalty + layers.ReduceMean(training_data_loss + gen_data_loss) def create_discriminator_loss(self, discrim_output_train, discrim_output_gen): gradient_penalty = GradientPenaltyLayer(discrim_output_train, self) return gradient_penalty + layers.ReduceMean(discrim_output_train - discrim_output_gen) class GradientPenaltyLayer(layers.Layer): """Implements the gradient penalty loss term for WGANs.""" def __init__(self, discrim_output, gan): super(GradientPenaltyLayer, self).__init__(discrim_output) def __init__(self, discrim_output_train, gan): super(GradientPenaltyLayer, self).__init__([discrim_output_train]) self.gan = gan def create_tensor(self, in_layers=None, set_tensors=True, **kwargs): gradients = tf.gradients(self.in_layers[0], self.gan._discrim_data) gradients = tf.gradients(self.in_layers[0], self.gan.data_inputs) norm2 = 0.0 for g in gradients: g2 = tf.square(g) Loading
