Merged changes from main branch (4e03bc74) · Commits · 钟慕尧 / deepchem

deepchem/feat/tests/test_one_hot_featurizer.py

+1 −1

Original line number	Diff line number	Diff line
		@@ -6,7 +6,7 @@ from deepchem.feat import OneHotFeaturizer
		from deepchem.feat.molecule_featurizers.one_hot_featurizer import ZINC_CHARSET


		class TestOneHotFeaturizert(unittest.TestCase):
		class TestOneHotFeaturizer(unittest.TestCase):
		"""
		Test OneHotFeaturizer.
		"""

deepchem/models/init.py

+1 −3

Original line number	Diff line number	Diff line
		@@ -7,9 +7,6 @@ from deepchem.models.keras_model import KerasModel
		from deepchem.models.multitask import SingletaskToMultitask
		from deepchem.models.callbacks import ValidationCallback

		from deepchem.models.fcnet import MultitaskRegressor
		from deepchem.models.fcnet import MultitaskClassifier
		from deepchem.models.fcnet import MultitaskFitTransformRegressor
		from deepchem.models.IRV import MultitaskIRVClassifier
		from deepchem.models.robust_multitask import RobustMultitaskClassifier
		from deepchem.models.robust_multitask import RobustMultitaskRegressor
		@@ -38,6 +35,7 @@ try:
		from deepchem.models.torch_models import GCN, GCNModel
		from deepchem.models.torch_models import LCNN, LCNNModel
		from deepchem.models.torch_models import Pagtn, PagtnModel
		from deepchem.models.fcnet import MultitaskRegressor, MultitaskClassifier, MultitaskFitTransformRegressor
		except ModuleNotFoundError:
		pass

deepchem/models/atomic_conv.py

+2 −2

Original line number	Diff line number	Diff line
		@@ -13,7 +13,7 @@ try:
		except:
		from collections import Sequence as SequenceCollection
		from typing import Sequence, Union
		from deepchem.utils.typing import KerasActivationFn, LossFn, OneOrMany
		from deepchem.utils.typing import ActivationFn, LossFn, OneOrMany
		from deepchem.utils.data_utils import load_from_disk, save_to_disk

		logger = logging.getLogger(__name__)
		@@ -56,7 +56,7 @@ class AtomicConvModel(KerasModel):
		weight_decay_penalty: float = 0.0,
		weight_decay_penalty_type: str = "l2",
		dropouts: OneOrMany[float] = 0.5,
		activation_fns: OneOrMany[KerasActivationFn] = tf.nn.relu,
		activation_fns: OneOrMany[ActivationFn] = tf.nn.relu,
		residual: bool = False,
		learning_rate=0.001,
		**kwargs) -> None:

deepchem/models/graph_models.py

+29 −30

Original line number	Diff line number	Diff line
		@@ -8,7 +8,7 @@ import numpy as np
		import tensorflow as tf

		from typing import List, Union, Tuple, Iterable, Dict, Optional
		from deepchem.utils.typing import OneOrMany, LossFn, KerasActivationFn
		from deepchem.utils.typing import OneOrMany, LossFn, ActivationFn
		from deepchem.data import Dataset, NumpyDataset, pad_features
		from deepchem.feat.graph_features import ConvMolFeaturizer
		from deepchem.feat.mol_graphs import ConvMol
		@@ -81,8 +81,7 @@ class WeaveModel(KerasModel):

		"""

		def __init__(
		self,
		def __init__(self,
		n_tasks: int,
		n_atom_feat: OneOrMany[int] = 75,
		n_pair_feat: OneOrMany[int] = 14,
		@@ -96,8 +95,8 @@ class WeaveModel(KerasModel):
		weight_decay_penalty: float = 0.0,
		weight_decay_penalty_type: str = "l2",
		dropouts: OneOrMany[float] = 0.25,
		final_conv_activation_fn: Optional[KerasActivationFn] = tf.nn.tanh,
		activation_fns: OneOrMany[KerasActivationFn] = tf.nn.relu,
		final_conv_activation_fn: Optional[ActivationFn] = tf.nn.tanh,
		activation_fns: OneOrMany[ActivationFn] = tf.nn.relu,
		batch_normalize: bool = True,
		batch_normalize_kwargs: Dict = {
		"renorm": True,
		@@ -151,7 +150,7 @@ class WeaveModel(KerasModel):
		The dropout probablity to use for each fully connected layer. The length of this list
		should equal len(layer_sizes). Alternatively this may be a single value
		instead of a list, in which case the same value is used for every layer.
		final_conv_activation_fn: Optional[KerasActivationFn] (default `tf.nn.tanh`)
		final_conv_activation_fn: Optional[ActivationFn] (default `tf.nn.tanh`)
		The Tensorflow activation funcntion to apply to the final
		convolution at the end of the weave convolutions. If `None`, then no
		activate is applied (hence linear).

deepchem/models/losses.py

+20 −4

Original line number	Diff line number	Diff line
		@@ -38,7 +38,12 @@ class L1Loss(Loss):

		def _create_pytorch_loss(self):
		import torch
		return torch.nn.L1Loss(reduction='none')

		def loss(output, labels):
		output, labels = _make_pytorch_shapes_consistent(output, labels)
		return torch.nn.functional.l1_loss(output, labels, reduction='none')

		return loss


		class HuberLoss(Loss):
		@@ -55,7 +60,13 @@ class HuberLoss(Loss):

		def _create_pytorch_loss(self):
		import torch
		return torch.nn.SmoothL1Loss(reduction='none')

		def loss(output, labels):
		output, labels = _make_pytorch_shapes_consistent(output, labels)
		return torch.nn.functional.smooth_l1_loss(
		output, labels, reduction='none')

		return loss


		class L2Loss(Loss):
		@@ -69,7 +80,12 @@ class L2Loss(Loss):

		def _create_pytorch_loss(self):
		import torch
		return torch.nn.MSELoss(reduction='none')

		def loss(output, labels):
		output, labels = _make_pytorch_shapes_consistent(output, labels)
		return torch.nn.functional.mse_loss(output, labels, reduction='none')

		return loss


		class HingeLoss(Loss):
		@@ -229,7 +245,7 @@ class SoftmaxCrossEntropy(Loss):

		def _create_pytorch_loss(self):
		import torch
		ls = torch.nn.LogSoftmax(dim=1)
		ls = torch.nn.LogSoftmax(dim=-1)

		def loss(output, labels):
		output, labels = _make_pytorch_shapes_consistent(output, labels)

Admin message