from typing import List
import torch
from torch import nn
from torch_geometric.nn.inits import zeros
from greatx.nn.layers import Sequential, activations
from greatx.utils import wrapper
[docs]class MLP(nn.Module):
r"""Implementation of Multi-layer Perceptron (MLP) or
Feed-forward Neural Network (FNN).
Parameters
----------
in_channels : int,
the input dimensions of model
out_channels : int,
the output dimensions of model
hids : List[int], optional
the number of hidden units for each hidden layer, by default [16]
acts : List[str], optional
the activation function for each hidden layer, by default ['relu']
dropout : float, optional
the dropout ratio of model, by default 0.5
bias : bool, optional
whether to use bias in the layers, by default True
bn: bool, optional
whether to use :class:`BatchNorm1d` after the Linear layer,
by default False
Examples
--------
>>> # MLP with one hidden layer
>>> model = MLP(100, 10)
>>> # MLP with two hidden layers
>>> model = MLP(100, 10, hids=[32, 16], acts=['relu', 'elu'])
>>> # MLP with two hidden layers, without first activation
>>> model = MLP(100, 10, hids=[32, 16], acts=[None, 'relu'])
>>> # MLP with deep architectures, each layer has elu activation
>>> model = MLP(100, 10, hids=[16]*8, acts=['elu'])
See also
--------
:class:`greatx.nn.models.supervised.LogisticRegression`
"""
@wrapper
def __init__(self, in_channels: int, out_channels: int,
hids: List[int] = [16], acts: List[str] = ['relu'],
dropout: float = 0.5, bias: bool = True, bn: bool = False):
super().__init__()
lin = []
for hid, act in zip(hids, acts):
lin.append(nn.Linear(in_channels, hid, bias=bias))
if bn:
lin.append(nn.BatchNorm1d(hid))
lin.append(activations.get(act))
lin.append(nn.Dropout(dropout))
in_channels = hid
lin.append(nn.Linear(in_channels, out_channels, bias=bias))
self.lin = Sequential(*lin)
[docs] def reset_parameters(self):
self.lin.reset_parameters()
[docs] def forward(self, x, *args, **kwargs):
""""""
return self.lin(x)
[docs]class LogisticRegression(nn.Module):
r"""Simple logistic regression model for
self-supervised/unsupervised learning.
Parameters
----------
in_channels : int,
the input dimensions of model
out_channels : int,
the output dimensions of model
bias : bool, optional
whether to use bias in the layers, by default True
See Examples below.
Examples
--------
>>> # LogisticRegression without hidden layer
>>> model = LogisticRegression(100, 10)
See also
--------
:class:`greatx.nn.models.supervised.MLP`
"""
def __init__(self, in_channels: int, out_channels: int, bias: bool = True):
super().__init__()
self.lin = nn.Linear(in_channels, out_channels, bias=bias)
[docs] def reset_parameters(self):
torch.nn.init.xavier_uniform_(self.lin.weight.data)
zeros(self.lin.bias)
[docs] def forward(self, x, *args, **kwargs):
""""""
return self.lin(x)