Loading deepchem/models/losses.py +17 −1 Original line number Diff line number Diff line Loading @@ -192,7 +192,17 @@ class SparseSoftmaxCrossEntropy(Loss): def _create_pytorch_loss(self): import torch return torch.nn.CrossEntropyLoss(reduction='none') ce_loss = torch.nn.CrossEntropyLoss(reduction='mean') def loss(output, labels): # Convert (batch_size, tasks, classes) to (batch_size, classes, tasks) # CrossEntropyLoss only supports (batch_size, classes, tasks) # This is for API consistency if len(output.shape) == 3: output = output.permute(0, 2, 1) return ce_loss(output, labels.long()) return loss def _make_tf_shapes_consistent(output, labels): Loading Loading @@ -251,3 +261,9 @@ def _ensure_float(output, labels): if labels.dtype not in (tf.float32, tf.float64): labels = tf.cast(labels, tf.float32) return (output, labels) def _ensure_long(labels): """Make sure the outputs are Long types.""" labels = [val.long() for val in labels] return labels deepchem/models/tests/test_gat.py +9 −5 Original line number Diff line number Diff line Loading @@ -26,10 +26,10 @@ def test_gat_regression(): mode='regression', n_tasks=n_tasks, batch_size=4, learning_rate=0.001) # overfit test model.fit(dataset, nb_epoch=100) # GAT's convergence is a little slow model.fit(dataset, nb_epoch=300) scores = model.evaluate(dataset, [metric], transformers) # TODO: check this asseration is correct or not assert scores['mean_absolute_error'] < 1.0 assert scores['mean_absolute_error'] < 0.2 @unittest.skipIf(not has_pytorch_and_pyg, Loading @@ -43,9 +43,13 @@ def test_gat_classification(): # initialize models n_tasks = len(tasks) model = GATModel( mode='classification', n_tasks=n_tasks, batch_size=10, learning_rate=0.001) mode='classification', n_tasks=n_tasks, batch_size=10, learning_rate=0.001) # overfit test model.fit(dataset, nb_epoch=10) # GAT's convergence is a little slow model.fit(dataset, nb_epoch=150) scores = model.evaluate(dataset, [metric], transformers) assert scores['mean-roc_auc_score'] >= 0.9 deepchem/models/torch_models/gat.py +22 −21 Original line number Diff line number Diff line Loading @@ -52,11 +52,11 @@ class GAT(nn.Module): def __init__( self, in_node_dim: int = 39, hidden_node_dim: int = 64, heads: int = 4, hidden_node_dim: int = 32, heads: int = 1, dropout: float = 0.0, num_conv: int = 3, predictor_hidden_feats: int = 32, num_conv: int = 2, predictor_hidden_feats: int = 64, n_tasks: int = 1, mode: str = 'classification', n_classes: int = 2, Loading @@ -67,19 +67,19 @@ class GAT(nn.Module): in_node_dim: int, default 39 The length of the initial node feature vectors. The 39 is based on `MolGraphConvFeaturizer`. hidden_node_dim: int, default 64 hidden_node_dim: int, default 32 The length of the hidden node feature vectors. heads: int, default 4 heads: int, default 1 The number of multi-head-attentions. dropout: float, default 0.0 The dropout probability for each convolutional layer. num_conv: int, default 3 num_conv: int, default 2 The number of convolutional layers. predictor_hidden_feats: int, default 32 The size for hidden representations in the output MLP predictor, default to 32. predictor_hidden_feats: int, default 64 The size for hidden representations in the output MLP predictor, default to 64. n_tasks: int, default 1 The number of the output size, default to 1. mode: str, default 'regression' mode: str, default 'classification' The model type, 'classification' or 'regression'. n_classes: int, default 2 The number of classes to predict (only used in classification mode). Loading Loading @@ -131,7 +131,7 @@ class GAT(nn.Module): # pooling graph_feat = self.pooling(node_feat, data.batch) graph_feat = F.relu(self.fc(graph_feat)) graph_feat = F.leaky_relu(self.fc(graph_feat)) out = self.out(graph_feat) if self.mode == 'regression': Loading @@ -140,7 +140,7 @@ class GAT(nn.Module): logits = out.view(-1, self.n_tasks, self.n_classes) # for n_tasks == 1 case logits = torch.squeeze(logits) proba = F.softmax(logits) proba = F.softmax(logits, dim=-1) return proba, logits Loading Loading @@ -177,11 +177,11 @@ class GATModel(TorchModel): def __init__(self, in_node_dim: int = 39, hidden_node_dim: int = 64, heads: int = 4, hidden_node_dim: int = 32, heads: int = 1, dropout: float = 0.0, num_conv: int = 3, predictor_hidden_feats: int = 32, num_conv: int = 2, predictor_hidden_feats: int = 64, n_tasks: int = 1, mode: str = 'regression', n_classes: int = 2, Loading @@ -194,16 +194,16 @@ class GATModel(TorchModel): in_node_dim: int, default 39 The length of the initial node feature vectors. The 39 is based on `MolGraphConvFeaturizer`. hidden_node_dim: int, default 64 hidden_node_dim: int, default 32 The length of the hidden node feature vectors. heads: int, default 4 heads: int, default 1 The number of multi-head-attentions. dropout: float, default 0.0 The dropout probability for each convolutional layer. num_conv: int, default 3 num_conv: int, default 2 The number of convolutional layers. predictor_hidden_feats: int, default 32 The size for hidden representations in the output MLP predictor, default to 32. predictor_hidden_feats: int, default 64 The size for hidden representations in the output MLP predictor, default to 64. n_tasks: int, default 1 The number of the output size, default to 1. mode: str, default 'regression' Loading Loading @@ -249,6 +249,7 @@ class GATModel(TorchModel): inputs, labels, weights = batch pyg_graphs = [graph.to_pyg_graph() for graph in inputs[0]] inputs = Batch.from_data_list(pyg_graphs) inputs = inputs.to(self.device) _, labels, weights = super(GATModel, self)._prepare_batch(([], labels, weights)) return inputs, labels, weights Loading
deepchem/models/losses.py +17 −1 Original line number Diff line number Diff line Loading @@ -192,7 +192,17 @@ class SparseSoftmaxCrossEntropy(Loss): def _create_pytorch_loss(self): import torch return torch.nn.CrossEntropyLoss(reduction='none') ce_loss = torch.nn.CrossEntropyLoss(reduction='mean') def loss(output, labels): # Convert (batch_size, tasks, classes) to (batch_size, classes, tasks) # CrossEntropyLoss only supports (batch_size, classes, tasks) # This is for API consistency if len(output.shape) == 3: output = output.permute(0, 2, 1) return ce_loss(output, labels.long()) return loss def _make_tf_shapes_consistent(output, labels): Loading Loading @@ -251,3 +261,9 @@ def _ensure_float(output, labels): if labels.dtype not in (tf.float32, tf.float64): labels = tf.cast(labels, tf.float32) return (output, labels) def _ensure_long(labels): """Make sure the outputs are Long types.""" labels = [val.long() for val in labels] return labels
deepchem/models/tests/test_gat.py +9 −5 Original line number Diff line number Diff line Loading @@ -26,10 +26,10 @@ def test_gat_regression(): mode='regression', n_tasks=n_tasks, batch_size=4, learning_rate=0.001) # overfit test model.fit(dataset, nb_epoch=100) # GAT's convergence is a little slow model.fit(dataset, nb_epoch=300) scores = model.evaluate(dataset, [metric], transformers) # TODO: check this asseration is correct or not assert scores['mean_absolute_error'] < 1.0 assert scores['mean_absolute_error'] < 0.2 @unittest.skipIf(not has_pytorch_and_pyg, Loading @@ -43,9 +43,13 @@ def test_gat_classification(): # initialize models n_tasks = len(tasks) model = GATModel( mode='classification', n_tasks=n_tasks, batch_size=10, learning_rate=0.001) mode='classification', n_tasks=n_tasks, batch_size=10, learning_rate=0.001) # overfit test model.fit(dataset, nb_epoch=10) # GAT's convergence is a little slow model.fit(dataset, nb_epoch=150) scores = model.evaluate(dataset, [metric], transformers) assert scores['mean-roc_auc_score'] >= 0.9
deepchem/models/torch_models/gat.py +22 −21 Original line number Diff line number Diff line Loading @@ -52,11 +52,11 @@ class GAT(nn.Module): def __init__( self, in_node_dim: int = 39, hidden_node_dim: int = 64, heads: int = 4, hidden_node_dim: int = 32, heads: int = 1, dropout: float = 0.0, num_conv: int = 3, predictor_hidden_feats: int = 32, num_conv: int = 2, predictor_hidden_feats: int = 64, n_tasks: int = 1, mode: str = 'classification', n_classes: int = 2, Loading @@ -67,19 +67,19 @@ class GAT(nn.Module): in_node_dim: int, default 39 The length of the initial node feature vectors. The 39 is based on `MolGraphConvFeaturizer`. hidden_node_dim: int, default 64 hidden_node_dim: int, default 32 The length of the hidden node feature vectors. heads: int, default 4 heads: int, default 1 The number of multi-head-attentions. dropout: float, default 0.0 The dropout probability for each convolutional layer. num_conv: int, default 3 num_conv: int, default 2 The number of convolutional layers. predictor_hidden_feats: int, default 32 The size for hidden representations in the output MLP predictor, default to 32. predictor_hidden_feats: int, default 64 The size for hidden representations in the output MLP predictor, default to 64. n_tasks: int, default 1 The number of the output size, default to 1. mode: str, default 'regression' mode: str, default 'classification' The model type, 'classification' or 'regression'. n_classes: int, default 2 The number of classes to predict (only used in classification mode). Loading Loading @@ -131,7 +131,7 @@ class GAT(nn.Module): # pooling graph_feat = self.pooling(node_feat, data.batch) graph_feat = F.relu(self.fc(graph_feat)) graph_feat = F.leaky_relu(self.fc(graph_feat)) out = self.out(graph_feat) if self.mode == 'regression': Loading @@ -140,7 +140,7 @@ class GAT(nn.Module): logits = out.view(-1, self.n_tasks, self.n_classes) # for n_tasks == 1 case logits = torch.squeeze(logits) proba = F.softmax(logits) proba = F.softmax(logits, dim=-1) return proba, logits Loading Loading @@ -177,11 +177,11 @@ class GATModel(TorchModel): def __init__(self, in_node_dim: int = 39, hidden_node_dim: int = 64, heads: int = 4, hidden_node_dim: int = 32, heads: int = 1, dropout: float = 0.0, num_conv: int = 3, predictor_hidden_feats: int = 32, num_conv: int = 2, predictor_hidden_feats: int = 64, n_tasks: int = 1, mode: str = 'regression', n_classes: int = 2, Loading @@ -194,16 +194,16 @@ class GATModel(TorchModel): in_node_dim: int, default 39 The length of the initial node feature vectors. The 39 is based on `MolGraphConvFeaturizer`. hidden_node_dim: int, default 64 hidden_node_dim: int, default 32 The length of the hidden node feature vectors. heads: int, default 4 heads: int, default 1 The number of multi-head-attentions. dropout: float, default 0.0 The dropout probability for each convolutional layer. num_conv: int, default 3 num_conv: int, default 2 The number of convolutional layers. predictor_hidden_feats: int, default 32 The size for hidden representations in the output MLP predictor, default to 32. predictor_hidden_feats: int, default 64 The size for hidden representations in the output MLP predictor, default to 64. n_tasks: int, default 1 The number of the output size, default to 1. mode: str, default 'regression' Loading Loading @@ -249,6 +249,7 @@ class GATModel(TorchModel): inputs, labels, weights = batch pyg_graphs = [graph.to_pyg_graph() for graph in inputs[0]] inputs = Batch.from_data_list(pyg_graphs) inputs = inputs.to(self.device) _, labels, weights = super(GATModel, self)._prepare_batch(([], labels, weights)) return inputs, labels, weights
