Note
Click here to download the full example code
Training an Image classifier¶
You will learn the basics of how to create an image classifier using the borch.nn package and fit it using the infer package.
Lets start of with importing what we need
import torch
from torch.utils.data import TensorDataset, DataLoader
import borch
from borch import infer, distributions
import torch.nn.functional as F
The module borch.nn provides implementations of neural network modules that are used
for deep probabilistic programming. It provides an interface almost identical to
the torch.nn modules and in many cases it is possible to just switch
from torch import nn
to
from borch import nn
Data¶
In this example we will use simulated data and not run the fitting until convergence,
but show how the model is set up and how one can construct the training loop.
We will just generate some random data, where data represent the image and
target is the class.
data = torch.randn(20, 1, 32, 32)
labels = torch.randperm(2).repeat(10)
data_set = TensorDataset(data, labels)
loader = DataLoader(data_set, batch_size=20)
Model¶
Lets set up the model.
In order to use infer and the borch to the fullest, we need to select a
a likelihood distribution. For classification the distributions.Categorical
is suitable.
class Net(borch.Module):
def __init__(self):
super(Net, self).__init__(posterior=borch.posterior.Automatic())
# 1 input image channel, 6 output channels, 5x5 square convolution
# kernel
self.conv1 = nn.Conv2d(1, 6, 5)
self.conv2 = nn.Conv2d(6, 16, 5)
# an affine operation: y = Wx + b
self.fc1 = nn.Linear(16 * 5 * 5, 120)
self.fc2 = nn.Linear(120, 2)
def forward(self, x):
# Max pooling over a (2, 2) window
x = F.max_pool2d(F.relu(self.conv1(x)), (2, 2))
# If the size is a square you can only specify a single number
x = F.max_pool2d(F.relu(self.conv2(x)), 2)
x = x.view(-1, self.num_flat_features(x))
x = F.relu(self.fc1(x))
x = self.fc2(x)
# Specifying the likelihood function
self.classification = distributions.Categorical(logits=x)
return self.classification
def num_flat_features(self, x):
size = x.size()[1:] # all dimensions except the batch dimension
num_features = 1
for s in size:
num_features *= s
return num_features
net = Net()
print(net)
Out:
Net(
(posterior): Automatic()
(prior): Module()
(observed): Observed()
(conv1): Conv2d(
1, 6, kernel_size=(5, 5), stride=(1, 1)
(posterior): Normal(
(weight): Normal:
posterior: Automatic()
prior: Module()
observed: Observed()
scale: Transform:
tensor([[[[0.0498, 0.0498, 0.0498, 0.0498, 0.0498],
[0.0498, 0.0498, 0.0498, 0.0498, 0.0498],
[0.0498, 0.0498, 0.0498, 0.0498, 0.0498],
[0.0498, 0.0498, 0.0498, 0.0498, 0.0498],
[0.0498, 0.0498, 0.0498, 0.0498, 0.0498]]],
[[[0.0498, 0.0498, 0.0498, 0.0498, 0.0498],
[0.0498, 0.0498, 0.0498, 0.0498, 0.0498],
[0.0498, 0.0498, 0.0498, 0.0498, 0.0498],
[0.0498, 0.0498, 0.0498, 0.0498, 0.0498],
[0.0498, 0.0498, 0.0498, 0.0498, 0.0498]]],
[[[0.0498, 0.0498, 0.0498, 0.0498, 0.0498],
[0.0498, 0.0498, 0.0498, 0.0498, 0.0498],
[0.0498, 0.0498, 0.0498, 0.0498, 0.0498],
[0.0498, 0.0498, 0.0498, 0.0498, 0.0498],
[0.0498, 0.0498, 0.0498, 0.0498, 0.0498]]],
[[[0.0498, 0.0498, 0.0498, 0.0498, 0.0498],
[0.0498, 0.0498, 0.0498, 0.0498, 0.0498],
[0.0498, 0.0498, 0.0498, 0.0498, 0.0498],
[0.0498, 0.0498, 0.0498, 0.0498, 0.0498],
[0.0498, 0.0498, 0.0498, 0.0498, 0.0498]]],
[[[0.0498, 0.0498, 0.0498, 0.0498, 0.0498],
[0.0498, 0.0498, 0.0498, 0.0498, 0.0498],
[0.0498, 0.0498, 0.0498, 0.0498, 0.0498],
[0.0498, 0.0498, 0.0498, 0.0498, 0.0498],
[0.0498, 0.0498, 0.0498, 0.0498, 0.0498]]],
[[[0.0498, 0.0498, 0.0498, 0.0498, 0.0498],
[0.0498, 0.0498, 0.0498, 0.0498, 0.0498],
[0.0498, 0.0498, 0.0498, 0.0498, 0.0498],
[0.0498, 0.0498, 0.0498, 0.0498, 0.0498],
[0.0498, 0.0498, 0.0498, 0.0498, 0.0498]]]], grad_fn=<ExpBackward0>)
loc: Parameter containing:
tensor([[[[-0.1927, 0.1399, -0.1354, -0.1729, -0.1509],
[-0.1643, 0.1672, 0.1231, 0.0760, 0.1983],
[-0.1877, 0.1498, 0.1842, -0.1852, -0.0757],
[-0.0739, -0.0365, 0.1345, 0.0660, -0.1410],
[ 0.0395, 0.1924, -0.0889, 0.1817, -0.1028]]],
[[[ 0.0161, -0.0283, 0.0567, 0.0753, 0.1422],
[ 0.1557, -0.0744, 0.1324, 0.1368, 0.1860],
[ 0.1084, -0.0066, -0.1755, 0.0982, -0.1038],
[ 0.0692, -0.1975, 0.1228, 0.1460, -0.1969],
[-0.0408, 0.0335, -0.1200, 0.0135, -0.0343]]],
[[[ 0.1090, -0.1523, -0.0839, -0.1336, 0.1845],
[-0.0021, -0.1854, -0.0692, 0.0818, 0.0268],
[ 0.0554, 0.0253, -0.0801, 0.0925, -0.1053],
[-0.0562, 0.0456, 0.1366, -0.1447, -0.1639],
[-0.0741, -0.1671, 0.1945, -0.1811, -0.1519]]],
[[[ 0.0950, 0.1374, 0.1735, 0.1682, 0.0029],
[ 0.0662, -0.0615, -0.1451, 0.1452, -0.1408],
[-0.1634, 0.0675, 0.1090, -0.1899, -0.0123],
[ 0.0842, -0.0821, 0.1183, -0.0658, 0.1601],
[-0.1688, 0.1212, -0.0177, -0.1106, 0.1500]]],
[[[-0.0711, 0.0887, 0.0110, -0.0312, 0.1235],
[ 0.1727, -0.1303, -0.1418, 0.0785, 0.0870],
[ 0.0772, 0.0361, -0.1826, -0.0933, 0.0847],
[ 0.0795, 0.1156, -0.1951, -0.1324, -0.1288],
[-0.1276, -0.1009, -0.1230, -0.1850, 0.0189]]],
[[[-0.0376, 0.1030, 0.1858, 0.1619, 0.1681],
[ 0.0571, -0.1018, -0.1616, 0.1122, 0.1009],
[ 0.1114, 0.0628, -0.0361, 0.0023, 0.0281],
[ 0.0725, 0.1015, -0.0013, 0.1214, -0.1805],
[ 0.1711, 0.1865, -0.1692, 0.0962, 0.0706]]]], requires_grad=True)
tensor: tensor([[[[-0.2157, 0.0700, -0.0908, -0.1634, -0.1610],
[-0.1348, 0.2526, 0.0959, -0.0100, 0.2411],
[-0.0937, 0.2073, 0.1696, -0.1477, -0.1274],
[-0.0645, 0.0310, 0.1207, 0.1160, -0.2358],
[ 0.1197, 0.2355, -0.0854, 0.1648, -0.1267]]],
[[[-0.0590, 0.0077, 0.0397, 0.0469, 0.1285],
[ 0.0683, -0.1121, 0.1573, 0.1348, 0.1819],
[ 0.1375, -0.0533, -0.2035, 0.1420, -0.1399],
[ 0.0028, -0.0762, 0.1180, 0.0648, -0.2013],
[-0.0807, 0.0769, -0.1508, -0.0319, -0.0546]]],
[[[ 0.0979, -0.1779, -0.0642, -0.0850, 0.1955],
[-0.0440, -0.2337, -0.0903, 0.0791, 0.0162],
[ 0.0858, -0.0131, -0.0561, 0.1656, -0.0313],
[-0.0810, 0.0303, 0.1008, -0.0657, -0.1811],
[-0.1711, -0.0730, 0.1278, -0.1982, -0.1405]]],
[[[ 0.0481, 0.1565, 0.2030, 0.1710, 0.0092],
[ 0.0199, -0.0892, -0.2083, 0.1719, -0.0671],
[-0.2068, 0.0169, 0.0941, -0.1937, 0.0171],
[ 0.0881, -0.0840, 0.0991, -0.0349, 0.1863],
[-0.2759, 0.0587, -0.0132, -0.1172, 0.2117]]],
[[[-0.0500, 0.0174, 0.0018, -0.0023, 0.1109],
[ 0.2215, -0.1161, -0.1446, 0.1037, 0.1574],
[ 0.1490, 0.0843, -0.2119, -0.1595, 0.1073],
[ 0.1108, 0.1599, -0.1039, -0.0907, -0.1746],
[-0.0869, -0.0473, -0.0885, -0.1409, 0.0183]]],
[[[-0.0328, 0.2260, 0.2178, 0.0954, 0.1086],
[ 0.0526, -0.1322, -0.1627, 0.1323, 0.1208],
[ 0.1250, 0.0943, 0.0322, 0.0166, 0.0406],
[-0.0006, 0.0562, -0.0055, 0.1553, -0.2777],
[ 0.2405, 0.2811, -0.2061, 0.1368, 0.0990]]]],
grad_fn=<AddBackward0>)
(bias): Normal:
posterior: Automatic()
prior: Module()
observed: Observed()
scale: Transform:
tensor([0.0498, 0.0498, 0.0498, 0.0498, 0.0498, 0.0498],
grad_fn=<ExpBackward0>)
loc: Parameter containing:
tensor([-0.1597, 0.1243, 0.0051, -0.0381, 0.0524, 0.0078],
requires_grad=True)
tensor: tensor([-0.1479, 0.0487, -0.0111, -0.0327, 0.0722, 0.0004],
grad_fn=<AddBackward0>)
)
(prior): Module(
(weight): Normal:
loc: tensor([[[[-0., 0., -0., -0., -0.],
[-0., 0., 0., 0., 0.],
[-0., 0., 0., -0., -0.],
[-0., -0., 0., 0., -0.],
[0., 0., -0., 0., -0.]]],
[[[0., -0., 0., 0., 0.],
[0., -0., 0., 0., 0.],
[0., -0., -0., 0., -0.],
[0., -0., 0., 0., -0.],
[-0., 0., -0., 0., -0.]]],
[[[0., -0., -0., -0., 0.],
[-0., -0., -0., 0., 0.],
[0., 0., -0., 0., -0.],
[-0., 0., 0., -0., -0.],
[-0., -0., 0., -0., -0.]]],
[[[0., 0., 0., 0., 0.],
[0., -0., -0., 0., -0.],
[-0., 0., 0., -0., -0.],
[0., -0., 0., -0., 0.],
[-0., 0., -0., -0., 0.]]],
[[[-0., 0., 0., -0., 0.],
[0., -0., -0., 0., 0.],
[0., 0., -0., -0., 0.],
[0., 0., -0., -0., -0.],
[-0., -0., -0., -0., 0.]]],
[[[-0., 0., 0., 0., 0.],
[0., -0., -0., 0., 0.],
[0., 0., -0., 0., 0.],
[0., 0., -0., 0., -0.],
[0., 0., -0., 0., 0.]]]])
scale: tensor([[[[0.2000, 0.2000, 0.2000, 0.2000, 0.2000],
[0.2000, 0.2000, 0.2000, 0.2000, 0.2000],
[0.2000, 0.2000, 0.2000, 0.2000, 0.2000],
[0.2000, 0.2000, 0.2000, 0.2000, 0.2000],
[0.2000, 0.2000, 0.2000, 0.2000, 0.2000]]],
[[[0.2000, 0.2000, 0.2000, 0.2000, 0.2000],
[0.2000, 0.2000, 0.2000, 0.2000, 0.2000],
[0.2000, 0.2000, 0.2000, 0.2000, 0.2000],
[0.2000, 0.2000, 0.2000, 0.2000, 0.2000],
[0.2000, 0.2000, 0.2000, 0.2000, 0.2000]]],
[[[0.2000, 0.2000, 0.2000, 0.2000, 0.2000],
[0.2000, 0.2000, 0.2000, 0.2000, 0.2000],
[0.2000, 0.2000, 0.2000, 0.2000, 0.2000],
[0.2000, 0.2000, 0.2000, 0.2000, 0.2000],
[0.2000, 0.2000, 0.2000, 0.2000, 0.2000]]],
[[[0.2000, 0.2000, 0.2000, 0.2000, 0.2000],
[0.2000, 0.2000, 0.2000, 0.2000, 0.2000],
[0.2000, 0.2000, 0.2000, 0.2000, 0.2000],
[0.2000, 0.2000, 0.2000, 0.2000, 0.2000],
[0.2000, 0.2000, 0.2000, 0.2000, 0.2000]]],
[[[0.2000, 0.2000, 0.2000, 0.2000, 0.2000],
[0.2000, 0.2000, 0.2000, 0.2000, 0.2000],
[0.2000, 0.2000, 0.2000, 0.2000, 0.2000],
[0.2000, 0.2000, 0.2000, 0.2000, 0.2000],
[0.2000, 0.2000, 0.2000, 0.2000, 0.2000]]],
[[[0.2000, 0.2000, 0.2000, 0.2000, 0.2000],
[0.2000, 0.2000, 0.2000, 0.2000, 0.2000],
[0.2000, 0.2000, 0.2000, 0.2000, 0.2000],
[0.2000, 0.2000, 0.2000, 0.2000, 0.2000],
[0.2000, 0.2000, 0.2000, 0.2000, 0.2000]]]])
posterior: Automatic()
prior: Module()
observed: Observed()
tensor: tensor([[[[-0.1927, 0.1399, -0.1354, -0.1729, -0.1509],
[-0.1643, 0.1672, 0.1231, 0.0760, 0.1983],
[-0.1877, 0.1498, 0.1842, -0.1852, -0.0757],
[-0.0739, -0.0365, 0.1345, 0.0660, -0.1410],
[ 0.0395, 0.1924, -0.0889, 0.1817, -0.1028]]],
[[[ 0.0161, -0.0283, 0.0567, 0.0753, 0.1422],
[ 0.1557, -0.0744, 0.1324, 0.1368, 0.1860],
[ 0.1084, -0.0066, -0.1755, 0.0982, -0.1038],
[ 0.0692, -0.1975, 0.1228, 0.1460, -0.1969],
[-0.0408, 0.0335, -0.1200, 0.0135, -0.0343]]],
[[[ 0.1090, -0.1523, -0.0839, -0.1336, 0.1845],
[-0.0021, -0.1854, -0.0692, 0.0818, 0.0268],
[ 0.0554, 0.0253, -0.0801, 0.0925, -0.1053],
[-0.0562, 0.0456, 0.1366, -0.1447, -0.1639],
[-0.0741, -0.1671, 0.1945, -0.1811, -0.1519]]],
[[[ 0.0950, 0.1374, 0.1735, 0.1682, 0.0029],
[ 0.0662, -0.0615, -0.1451, 0.1452, -0.1408],
[-0.1634, 0.0675, 0.1090, -0.1899, -0.0123],
[ 0.0842, -0.0821, 0.1183, -0.0658, 0.1601],
[-0.1688, 0.1212, -0.0177, -0.1106, 0.1500]]],
[[[-0.0711, 0.0887, 0.0110, -0.0312, 0.1235],
[ 0.1727, -0.1303, -0.1418, 0.0785, 0.0870],
[ 0.0772, 0.0361, -0.1826, -0.0933, 0.0847],
[ 0.0795, 0.1156, -0.1951, -0.1324, -0.1288],
[-0.1276, -0.1009, -0.1230, -0.1850, 0.0189]]],
[[[-0.0376, 0.1030, 0.1858, 0.1619, 0.1681],
[ 0.0571, -0.1018, -0.1616, 0.1122, 0.1009],
[ 0.1114, 0.0628, -0.0361, 0.0023, 0.0281],
[ 0.0725, 0.1015, -0.0013, 0.1214, -0.1805],
[ 0.1711, 0.1865, -0.1692, 0.0962, 0.0706]]]])
(bias): Normal:
loc: tensor([-0., 0., 0., -0., 0., 0.])
scale: tensor([0.4082, 0.4082, 0.4082, 0.4082, 0.4082, 0.4082])
posterior: Automatic()
prior: Module()
observed: Observed()
tensor: tensor([-0.1597, 0.1243, 0.0051, -0.0381, 0.0524, 0.0078])
)
(observed): Observed()
)
(conv2): Conv2d(
6, 16, kernel_size=(5, 5), stride=(1, 1)
(posterior): Normal(
(weight): Normal:
posterior: Automatic()
prior: Module()
observed: Observed()
scale: Transform:
tensor([[[[0.0498, 0.0498, 0.0498, 0.0498, 0.0498],
[0.0498, 0.0498, 0.0498, 0.0498, 0.0498],
[0.0498, 0.0498, 0.0498, 0.0498, 0.0498],
[0.0498, 0.0498, 0.0498, 0.0498, 0.0498],
[0.0498, 0.0498, 0.0498, 0.0498, 0.0498]],
[[0.0498, 0.0498, 0.0498, 0.0498, 0.0498],
[0.0498, 0.0498, 0.0498, 0.0498, 0.0498],
[0.0498, 0.0498, 0.0498, 0.0498, 0.0498],
[0.0498, 0.0498, 0.0498, 0.0498, 0.0498],
[0.0498, 0.0498, 0.0498, 0.0498, 0.0498]],
[[0.0498, 0.0498, 0.0498, 0.0498, 0.0498],
[0.0498, 0.0498, 0.0498, 0.0498, 0.0498],
[0.0498, 0.0498, 0.0498, 0.0498, 0.0498],
[0.0498, 0.0498, 0.0498, 0.0498, 0.0498],
[0.0498, 0.0498, 0.0498, 0.0498, 0.0498]],
[[0.0498, 0.0498, 0.0498, 0.0498, 0.0498],
[0.0498, 0.0498, 0.0498, 0.0498, 0.0498],
[0.0498, 0.0498, 0.0498, 0.0498, 0.0498],
[0.0498, 0.0498, 0.0498, 0.0498, 0.0498],
[0.0498, 0.0498, 0.0498, 0.0498, 0.0498]],
[[0.0498, 0.0498, 0.0498, 0.0498, 0.0498],
[0.0498, 0.0498, 0.0498, 0.0498, 0.0498],
[0.0498, 0.0498, 0.0498, 0.0498, 0.0498],
[0.0498, 0.0498, 0.0498, 0.0498, 0.0498],
[0.0498, 0.0498, 0.0498, 0.0498, 0.0498]],
[[0.0498, 0.0498, 0.0498, 0.0498, 0.0498],
[0.0498, 0.0498, 0.0498, 0.0498, 0.0498],
[0.0498, 0.0498, 0.0498, 0.0498, 0.0498],
[0.0498, 0.0498, 0.0498, 0.0498, 0.0498],
[0.0498, 0.0498, 0.0498, 0.0498, 0.0498]]],
[[[0.0498, 0.0498, 0.0498, 0.0498, 0.0498],
[0.0498, 0.0498, 0.0498, 0.0498, 0.0498],
[0.0498, 0.0498, 0.0498, 0.0498, 0.0498],
[0.0498, 0.0498, 0.0498, 0.0498, 0.0498],
[0.0498, 0.0498, 0.0498, 0.0498, 0.0498]],
[[0.0498, 0.0498, 0.0498, 0.0498, 0.0498],
[0.0498, 0.0498, 0.0498, 0.0498, 0.0498],
[0.0498, 0.0498, 0.0498, 0.0498, 0.0498],
[0.0498, 0.0498, 0.0498, 0.0498, 0.0498],
[0.0498, 0.0498, 0.0498, 0.0498, 0.0498]],
[[0.0498, 0.0498, 0.0498, 0.0498, 0.0498],
[0.0498, 0.0498, 0.0498, 0.0498, 0.0498],
[0.0498, 0.0498, 0.0498, 0.0498, 0.0498],
[0.0498, 0.0498, 0.0498, 0.0498, 0.0498],
[0.0498, 0.0498, 0.0498, 0.0498, 0.0498]],
[[0.0498, 0.0498, 0.0498, 0.0498, 0.0498],
[0.0498, 0.0498, 0.0498, 0.0498, 0.0498],
[0.0498, 0.0498, 0.0498, 0.0498, 0.0498],
[0.0498, 0.0498, 0.0498, 0.0498, 0.0498],
[0.0498, 0.0498, 0.0498, 0.0498, 0.0498]],
[[0.0498, 0.0498, 0.0498, 0.0498, 0.0498],
[0.0498, 0.0498, 0.0498, 0.0498, 0.0498],
[0.0498, 0.0498, 0.0498, 0.0498, 0.0498],
[0.0498, 0.0498, 0.0498, 0.0498, 0.0498],
[0.0498, 0.0498, 0.0498, 0.0498, 0.0498]],
[[0.0498, 0.0498, 0.0498, 0.0498, 0.0498],
[0.0498, 0.0498, 0.0498, 0.0498, 0.0498],
[0.0498, 0.0498, 0.0498, 0.0498, 0.0498],
[0.0498, 0.0498, 0.0498, 0.0498, 0.0498],
[0.0498, 0.0498, 0.0498, 0.0498, 0.0498]]],
[[[0.0498, 0.0498, 0.0498, 0.0498, 0.0498],
[0.0498, 0.0498, 0.0498, 0.0498, 0.0498],
[0.0498, 0.0498, 0.0498, 0.0498, 0.0498],
[0.0498, 0.0498, 0.0498, 0.0498, 0.0498],
[0.0498, 0.0498, 0.0498, 0.0498, 0.0498]],
[[0.0498, 0.0498, 0.0498, 0.0498, 0.0498],
[0.0498, 0.0498, 0.0498, 0.0498, 0.0498],
[0.0498, 0.0498, 0.0498, 0.0498, 0.0498],
[0.0498, 0.0498, 0.0498, 0.0498, 0.0498],
[0.0498, 0.0498, 0.0498, 0.0498, 0.0498]],
[[0.0498, 0.0498, 0.0498, 0.0498, 0.0498],
[0.0498, 0.0498, 0.0498, 0.0498, 0.0498],
[0.0498, 0.0498, 0.0498, 0.0498, 0.0498],
[0.0498, 0.0498, 0.0498, 0.0498, 0.0498],
[0.0498, 0.0498, 0.0498, 0.0498, 0.0498]],
[[0.0498, 0.0498, 0.0498, 0.0498, 0.0498],
[0.0498, 0.0498, 0.0498, 0.0498, 0.0498],
[0.0498, 0.0498, 0.0498, 0.0498, 0.0498],
[0.0498, 0.0498, 0.0498, 0.0498, 0.0498],
[0.0498, 0.0498, 0.0498, 0.0498, 0.0498]],
[[0.0498, 0.0498, 0.0498, 0.0498, 0.0498],
[0.0498, 0.0498, 0.0498, 0.0498, 0.0498],
[0.0498, 0.0498, 0.0498, 0.0498, 0.0498],
[0.0498, 0.0498, 0.0498, 0.0498, 0.0498],
[0.0498, 0.0498, 0.0498, 0.0498, 0.0498]],
[[0.0498, 0.0498, 0.0498, 0.0498, 0.0498],
[0.0498, 0.0498, 0.0498, 0.0498, 0.0498],
[0.0498, 0.0498, 0.0498, 0.0498, 0.0498],
[0.0498, 0.0498, 0.0498, 0.0498, 0.0498],
[0.0498, 0.0498, 0.0498, 0.0498, 0.0498]]],
...,
[[[0.0498, 0.0498, 0.0498, 0.0498, 0.0498],
[0.0498, 0.0498, 0.0498, 0.0498, 0.0498],
[0.0498, 0.0498, 0.0498, 0.0498, 0.0498],
[0.0498, 0.0498, 0.0498, 0.0498, 0.0498],
[0.0498, 0.0498, 0.0498, 0.0498, 0.0498]],
[[0.0498, 0.0498, 0.0498, 0.0498, 0.0498],
[0.0498, 0.0498, 0.0498, 0.0498, 0.0498],
[0.0498, 0.0498, 0.0498, 0.0498, 0.0498],
[0.0498, 0.0498, 0.0498, 0.0498, 0.0498],
[0.0498, 0.0498, 0.0498, 0.0498, 0.0498]],
[[0.0498, 0.0498, 0.0498, 0.0498, 0.0498],
[0.0498, 0.0498, 0.0498, 0.0498, 0.0498],
[0.0498, 0.0498, 0.0498, 0.0498, 0.0498],
[0.0498, 0.0498, 0.0498, 0.0498, 0.0498],
[0.0498, 0.0498, 0.0498, 0.0498, 0.0498]],
[[0.0498, 0.0498, 0.0498, 0.0498, 0.0498],
[0.0498, 0.0498, 0.0498, 0.0498, 0.0498],
[0.0498, 0.0498, 0.0498, 0.0498, 0.0498],
[0.0498, 0.0498, 0.0498, 0.0498, 0.0498],
[0.0498, 0.0498, 0.0498, 0.0498, 0.0498]],
[[0.0498, 0.0498, 0.0498, 0.0498, 0.0498],
[0.0498, 0.0498, 0.0498, 0.0498, 0.0498],
[0.0498, 0.0498, 0.0498, 0.0498, 0.0498],
[0.0498, 0.0498, 0.0498, 0.0498, 0.0498],
[0.0498, 0.0498, 0.0498, 0.0498, 0.0498]],
[[0.0498, 0.0498, 0.0498, 0.0498, 0.0498],
[0.0498, 0.0498, 0.0498, 0.0498, 0.0498],
[0.0498, 0.0498, 0.0498, 0.0498, 0.0498],
[0.0498, 0.0498, 0.0498, 0.0498, 0.0498],
[0.0498, 0.0498, 0.0498, 0.0498, 0.0498]]],
[[[0.0498, 0.0498, 0.0498, 0.0498, 0.0498],
[0.0498, 0.0498, 0.0498, 0.0498, 0.0498],
[0.0498, 0.0498, 0.0498, 0.0498, 0.0498],
[0.0498, 0.0498, 0.0498, 0.0498, 0.0498],
[0.0498, 0.0498, 0.0498, 0.0498, 0.0498]],
[[0.0498, 0.0498, 0.0498, 0.0498, 0.0498],
[0.0498, 0.0498, 0.0498, 0.0498, 0.0498],
[0.0498, 0.0498, 0.0498, 0.0498, 0.0498],
[0.0498, 0.0498, 0.0498, 0.0498, 0.0498],
[0.0498, 0.0498, 0.0498, 0.0498, 0.0498]],
[[0.0498, 0.0498, 0.0498, 0.0498, 0.0498],
[0.0498, 0.0498, 0.0498, 0.0498, 0.0498],
[0.0498, 0.0498, 0.0498, 0.0498, 0.0498],
[0.0498, 0.0498, 0.0498, 0.0498, 0.0498],
[0.0498, 0.0498, 0.0498, 0.0498, 0.0498]],
[[0.0498, 0.0498, 0.0498, 0.0498, 0.0498],
[0.0498, 0.0498, 0.0498, 0.0498, 0.0498],
[0.0498, 0.0498, 0.0498, 0.0498, 0.0498],
[0.0498, 0.0498, 0.0498, 0.0498, 0.0498],
[0.0498, 0.0498, 0.0498, 0.0498, 0.0498]],
[[0.0498, 0.0498, 0.0498, 0.0498, 0.0498],
[0.0498, 0.0498, 0.0498, 0.0498, 0.0498],
[0.0498, 0.0498, 0.0498, 0.0498, 0.0498],
[0.0498, 0.0498, 0.0498, 0.0498, 0.0498],
[0.0498, 0.0498, 0.0498, 0.0498, 0.0498]],
[[0.0498, 0.0498, 0.0498, 0.0498, 0.0498],
[0.0498, 0.0498, 0.0498, 0.0498, 0.0498],
[0.0498, 0.0498, 0.0498, 0.0498, 0.0498],
[0.0498, 0.0498, 0.0498, 0.0498, 0.0498],
[0.0498, 0.0498, 0.0498, 0.0498, 0.0498]]],
[[[0.0498, 0.0498, 0.0498, 0.0498, 0.0498],
[0.0498, 0.0498, 0.0498, 0.0498, 0.0498],
[0.0498, 0.0498, 0.0498, 0.0498, 0.0498],
[0.0498, 0.0498, 0.0498, 0.0498, 0.0498],
[0.0498, 0.0498, 0.0498, 0.0498, 0.0498]],
[[0.0498, 0.0498, 0.0498, 0.0498, 0.0498],
[0.0498, 0.0498, 0.0498, 0.0498, 0.0498],
[0.0498, 0.0498, 0.0498, 0.0498, 0.0498],
[0.0498, 0.0498, 0.0498, 0.0498, 0.0498],
[0.0498, 0.0498, 0.0498, 0.0498, 0.0498]],
[[0.0498, 0.0498, 0.0498, 0.0498, 0.0498],
[0.0498, 0.0498, 0.0498, 0.0498, 0.0498],
[0.0498, 0.0498, 0.0498, 0.0498, 0.0498],
[0.0498, 0.0498, 0.0498, 0.0498, 0.0498],
[0.0498, 0.0498, 0.0498, 0.0498, 0.0498]],
[[0.0498, 0.0498, 0.0498, 0.0498, 0.0498],
[0.0498, 0.0498, 0.0498, 0.0498, 0.0498],
[0.0498, 0.0498, 0.0498, 0.0498, 0.0498],
[0.0498, 0.0498, 0.0498, 0.0498, 0.0498],
[0.0498, 0.0498, 0.0498, 0.0498, 0.0498]],
[[0.0498, 0.0498, 0.0498, 0.0498, 0.0498],
[0.0498, 0.0498, 0.0498, 0.0498, 0.0498],
[0.0498, 0.0498, 0.0498, 0.0498, 0.0498],
[0.0498, 0.0498, 0.0498, 0.0498, 0.0498],
[0.0498, 0.0498, 0.0498, 0.0498, 0.0498]],
[[0.0498, 0.0498, 0.0498, 0.0498, 0.0498],
[0.0498, 0.0498, 0.0498, 0.0498, 0.0498],
[0.0498, 0.0498, 0.0498, 0.0498, 0.0498],
[0.0498, 0.0498, 0.0498, 0.0498, 0.0498],
[0.0498, 0.0498, 0.0498, 0.0498, 0.0498]]]], grad_fn=<ExpBackward0>)
loc: Parameter containing:
tensor([[[[ 0.0228, -0.0484, 0.0637, -0.0249, -0.0302],
[ 0.0105, 0.0733, 0.0576, 0.0393, 0.0094],
[-0.0590, -0.0747, -0.0506, -0.0495, 0.0402],
[-0.0647, 0.0789, -0.0096, 0.0477, 0.0700],
[-0.0450, 0.0176, 0.0443, -0.0176, 0.0407]],
[[-0.0148, 0.0075, -0.0554, 0.0808, 0.0812],
[ 0.0636, -0.0776, 0.0617, 0.0497, 0.0760],
[-0.0350, 0.0633, 0.0318, 0.0562, 0.0742],
[-0.0229, -0.0075, 0.0338, 0.0709, -0.0382],
[ 0.0655, 0.0249, 0.0556, 0.0329, -0.0734]],
[[-0.0531, 0.0393, 0.0355, -0.0358, 0.0298],
[-0.0772, -0.0365, 0.0397, -0.0093, -0.0571],
[ 0.0087, 0.0256, 0.0568, 0.0123, 0.0098],
[ 0.0070, 0.0767, 0.0580, 0.0573, 0.0128],
[ 0.0476, -0.0090, 0.0190, 0.0597, -0.0108]],
[[-0.0097, -0.0755, 0.0251, 0.0349, 0.0423],
[ 0.0800, 0.0594, 0.0019, 0.0419, -0.0641],
[-0.0501, -0.0639, -0.0783, 0.0276, 0.0425],
[ 0.0793, -0.0161, 0.0031, 0.0700, -0.0077],
[ 0.0717, 0.0564, 0.0130, 0.0126, 0.0055]],
[[-0.0711, -0.0004, -0.0350, -0.0729, -0.0567],
[-0.0351, 0.0119, 0.0163, 0.0542, 0.0558],
[-0.0308, 0.0337, 0.0634, 0.0556, 0.0051],
[-0.0305, -0.0103, -0.0515, -0.0746, 0.0701],
[ 0.0014, -0.0249, 0.0724, 0.0282, -0.0773]],
[[ 0.0647, 0.0735, -0.0460, 0.0321, 0.0060],
[-0.0624, -0.0262, 0.0131, -0.0270, -0.0768],
[ 0.0053, 0.0301, 0.0262, -0.0012, 0.0230],
[ 0.0758, -0.0297, -0.0011, -0.0424, 0.0659],
[ 0.0155, -0.0027, -0.0785, -0.0715, 0.0603]]],
[[[-0.0680, 0.0027, -0.0140, 0.0451, 0.0738],
[-0.0323, -0.0084, -0.0281, 0.0522, -0.0210],
[-0.0525, -0.0399, -0.0072, -0.0577, 0.0280],
[-0.0374, -0.0055, 0.0143, -0.0258, 0.0404],
[-0.0804, 0.0563, -0.0096, -0.0485, -0.0052]],
[[ 0.0772, -0.0417, -0.0619, -0.0375, -0.0217],
[ 0.0312, -0.0186, -0.0225, 0.0433, 0.0327],
[ 0.0751, -0.0210, 0.0098, 0.0275, 0.0606],
[ 0.0308, -0.0484, -0.0063, 0.0799, -0.0184],
[-0.0195, 0.0014, -0.0263, -0.0651, -0.0203]],
[[-0.0441, -0.0229, -0.0547, 0.0813, -0.0195],
[-0.0343, 0.0006, 0.0058, 0.0113, 0.0668],
[-0.0674, 0.0276, -0.0072, 0.0360, -0.0707],
[-0.0468, 0.0240, -0.0658, 0.0561, 0.0396],
[-0.0704, -0.0553, 0.0372, -0.0310, 0.0525]],
[[-0.0081, -0.0563, 0.0125, 0.0304, 0.0486],
[-0.0515, 0.0539, -0.0165, 0.0263, 0.0241],
[-0.0189, 0.0789, -0.0254, 0.0502, 0.0336],
[-0.0288, -0.0308, 0.0397, -0.0238, -0.0365],
[ 0.0195, 0.0625, -0.0496, 0.0805, 0.0320]],
[[ 0.0451, 0.0478, -0.0347, -0.0325, 0.0203],
[-0.0460, -0.0406, 0.0251, 0.0666, 0.0718],
[-0.0109, -0.0145, 0.0313, 0.0103, 0.0525],
[-0.0299, -0.0280, -0.0036, 0.0186, 0.0350],
[ 0.0073, 0.0463, 0.0292, -0.0732, -0.0485]],
[[ 0.0059, -0.0394, 0.0745, 0.0484, 0.0053],
[-0.0341, -0.0371, 0.0608, 0.0058, -0.0537],
[-0.0605, -0.0791, -0.0253, 0.0762, 0.0762],
[-0.0059, 0.0153, 0.0580, 0.0198, -0.0173],
[-0.0174, 0.0578, -0.0368, -0.0736, -0.0631]]],
[[[ 0.0033, -0.0336, 0.0105, 0.0191, 0.0074],
[ 0.0714, -0.0613, -0.0475, -0.0452, -0.0098],
[-0.0617, -0.0515, 0.0353, -0.0375, 0.0360],
[ 0.0542, -0.0326, -0.0236, 0.0674, -0.0347],
[ 0.0394, 0.0746, -0.0467, 0.0522, 0.0424]],
[[ 0.0503, -0.0693, 0.0626, -0.0424, -0.0488],
[-0.0617, -0.0619, -0.0342, 0.0127, -0.0253],
[ 0.0319, 0.0270, -0.0089, 0.0175, -0.0308],
[-0.0551, 0.0107, 0.0212, 0.0026, 0.0019],
[ 0.0115, 0.0343, 0.0572, 0.0811, -0.0354]],
[[ 0.0632, -0.0197, 0.0072, -0.0372, 0.0581],
[-0.0492, -0.0602, -0.0201, 0.0465, -0.0482],
[ 0.0403, -0.0086, -0.0292, -0.0711, -0.0294],
[-0.0228, -0.0102, 0.0450, 0.0804, 0.0032],
[-0.0009, 0.0815, 0.0113, 0.0315, -0.0110]],
[[-0.0627, 0.0313, -0.0443, -0.0353, 0.0797],
[-0.0039, -0.0053, -0.0678, -0.0335, -0.0764],
[-0.0202, 0.0296, 0.0548, 0.0185, -0.0028],
[ 0.0044, 0.0496, 0.0734, -0.0418, -0.0792],
[ 0.0025, 0.0171, -0.0681, -0.0224, 0.0077]],
[[ 0.0424, 0.0476, 0.0079, -0.0167, 0.0523],
[-0.0335, 0.0003, -0.0560, -0.0448, -0.0173],
[ 0.0426, -0.0002, 0.0133, -0.0705, 0.0085],
[ 0.0724, 0.0664, 0.0455, 0.0461, -0.0607],
[-0.0546, 0.0696, -0.0658, 0.0364, -0.0562]],
[[-0.0390, 0.0704, 0.0786, 0.0033, 0.0295],
[-0.0145, -0.0198, 0.0638, 0.0332, -0.0197],
[ 0.0663, 0.0625, -0.0292, 0.0036, -0.0408],
[ 0.0201, 0.0786, -0.0615, -0.0266, -0.0627],
[-0.0323, 0.0509, -0.0516, -0.0624, -0.0025]]],
...,
[[[ 0.0777, 0.0133, -0.0427, 0.0800, -0.0246],
[-0.0430, 0.0370, 0.0550, 0.0002, -0.0453],
[-0.0782, 0.0747, -0.0467, 0.0130, -0.0217],
[ 0.0417, 0.0371, 0.0066, 0.0761, 0.0344],
[ 0.0590, 0.0188, -0.0681, 0.0579, 0.0579]],
[[ 0.0803, 0.0284, -0.0485, 0.0127, 0.0668],
[-0.0049, -0.0313, -0.0693, -0.0646, -0.0111],
[ 0.0460, 0.0529, 0.0564, -0.0790, -0.0583],
[-0.0102, 0.0005, 0.0437, 0.0308, -0.0437],
[ 0.0205, 0.0394, -0.0644, 0.0770, 0.0368]],
[[ 0.0157, -0.0298, -0.0073, 0.0653, -0.0469],
[-0.0527, 0.0319, -0.0647, 0.0084, -0.0241],
[ 0.0004, 0.0514, -0.0025, 0.0630, -0.0294],
[ 0.0689, 0.0191, 0.0546, 0.0365, -0.0539],
[-0.0045, -0.0745, 0.0544, -0.0686, 0.0816]],
[[ 0.0061, -0.0117, 0.0211, 0.0131, -0.0410],
[ 0.0168, -0.0675, -0.0497, 0.0285, -0.0718],
[-0.0401, -0.0345, 0.0003, -0.0376, -0.0554],
[ 0.0065, 0.0611, -0.0163, 0.0418, -0.0605],
[-0.0081, 0.0791, 0.0381, -0.0200, -0.0402]],
[[ 0.0350, 0.0182, -0.0750, 0.0471, 0.0045],
[ 0.0441, 0.0266, -0.0399, -0.0425, -0.0318],
[-0.0428, -0.0136, -0.0466, -0.0151, 0.0100],
[-0.0644, 0.0688, 0.0709, 0.0026, 0.0585],
[-0.0269, -0.0130, 0.0715, -0.0413, -0.0011]],
[[-0.0537, 0.0477, -0.0147, -0.0553, 0.0459],
[ 0.0619, 0.0005, -0.0194, 0.0200, 0.0296],
[-0.0205, -0.0382, 0.0277, 0.0322, -0.0201],
[-0.0251, -0.0238, -0.0758, 0.0534, 0.0487],
[-0.0023, 0.0328, -0.0237, 0.0328, -0.0399]]],
[[[ 0.0501, -0.0288, -0.0173, 0.0392, -0.0050],
[ 0.0156, 0.0807, 0.0459, -0.0056, 0.0346],
[-0.0486, 0.0271, 0.0530, -0.0753, -0.0678],
[ 0.0070, 0.0007, -0.0208, -0.0359, 0.0587],
[ 0.0728, -0.0304, 0.0334, -0.0355, 0.0323]],
[[-0.0734, 0.0403, 0.0763, 0.0570, -0.0033],
[-0.0664, 0.0105, -0.0366, 0.0042, 0.0546],
[-0.0745, 0.0639, -0.0417, -0.0033, 0.0495],
[-0.0550, 0.0292, -0.0198, -0.0241, 0.0367],
[ 0.0107, -0.0213, -0.0517, 0.0664, 0.0738]],
[[ 0.0123, 0.0034, -0.0046, 0.0620, -0.0356],
[-0.0580, -0.0810, 0.0235, 0.0637, -0.0323],
[ 0.0501, 0.0611, 0.0696, -0.0725, -0.0392],
[-0.0327, 0.0616, -0.0098, 0.0648, 0.0049],
[-0.0644, 0.0434, 0.0465, 0.0378, 0.0250]],
[[ 0.0816, 0.0159, 0.0255, 0.0219, -0.0652],
[ 0.0783, 0.0748, -0.0595, 0.0515, -0.0486],
[-0.0709, -0.0491, -0.0587, -0.0085, -0.0437],
[ 0.0395, -0.0117, 0.0683, 0.0806, -0.0066],
[-0.0332, -0.0257, -0.0023, -0.0359, 0.0064]],
[[-0.0328, -0.0616, -0.0107, -0.0231, -0.0393],
[ 0.0030, 0.0048, -0.0813, -0.0253, -0.0723],
[-0.0680, -0.0350, 0.0409, 0.0464, 0.0235],
[-0.0085, 0.0688, -0.0767, -0.0011, -0.0570],
[ 0.0553, -0.0721, 0.0039, -0.0811, -0.0608]],
[[ 0.0617, 0.0303, -0.0521, 0.0155, -0.0364],
[-0.0589, -0.0223, -0.0112, -0.0599, -0.0590],
[ 0.0729, 0.0326, 0.0761, -0.0415, -0.0048],
[ 0.0036, -0.0197, -0.0393, -0.0060, 0.0785],
[-0.0679, -0.0750, 0.0671, 0.0385, -0.0260]]],
[[[-0.0456, 0.0197, 0.0548, 0.0420, -0.0569],
[ 0.0518, -0.0172, -0.0758, -0.0328, 0.0196],
[-0.0712, -0.0446, 0.0593, -0.0403, -0.0250],
[-0.0142, -0.0058, -0.0283, 0.0783, 0.0075],
[ 0.0755, 0.0161, 0.0319, -0.0562, 0.0378]],
[[ 0.0572, 0.0773, 0.0243, 0.0638, -0.0472],
[-0.0081, 0.0225, 0.0298, -0.0442, -0.0075],
[-0.0814, 0.0369, -0.0680, -0.0471, 0.0187],
[ 0.0290, -0.0338, 0.0786, 0.0685, -0.0263],
[-0.0453, -0.0716, -0.0462, 0.0556, 0.0159]],
[[ 0.0359, -0.0511, 0.0707, -0.0696, 0.0407],
[-0.0717, 0.0521, 0.0813, 0.0335, -0.0515],
[-0.0295, -0.0124, -0.0406, -0.0247, 0.0162],
[-0.0252, 0.0105, 0.0624, -0.0701, 0.0153],
[-0.0490, 0.0815, 0.0331, 0.0130, -0.0174]],
[[-0.0801, -0.0390, -0.0246, 0.0187, 0.0752],
[-0.0654, -0.0242, 0.0666, 0.0303, -0.0114],
[ 0.0783, -0.0565, -0.0200, -0.0462, -0.0119],
[ 0.0788, 0.0656, 0.0623, 0.0350, 0.0254],
[-0.0227, 0.0380, -0.0172, 0.0293, -0.0065]],
[[-0.0086, -0.0572, 0.0217, -0.0286, 0.0476],
[ 0.0695, -0.0679, 0.0714, 0.0371, 0.0638],
[ 0.0099, -0.0652, -0.0545, -0.0068, 0.0805],
[-0.0506, -0.0737, -0.0110, -0.0198, 0.0047],
[-0.0288, 0.0730, 0.0794, -0.0033, 0.0242]],
[[-0.0616, -0.0632, -0.0110, -0.0658, -0.0470],
[ 0.0425, -0.0136, 0.0665, -0.0201, -0.0727],
[ 0.0189, 0.0189, 0.0641, -0.0384, 0.0180],
[ 0.0002, -0.0737, 0.0365, 0.0311, -0.0378],
[ 0.0789, 0.0037, -0.0582, 0.0148, 0.0323]]]], requires_grad=True)
tensor: tensor([[[[ 4.9998e-02, -9.8260e-02, 1.4967e-01, -4.4207e-03, -8.7300e-02],
[ 7.5572e-02, 4.8974e-02, 1.6032e-02, -1.6826e-02, -1.1976e-01],
[-6.9849e-02, -2.5324e-02, -5.0632e-02, -1.1284e-01, 4.8002e-02],
[-1.0279e-01, 7.0627e-03, 1.4877e-01, 8.8626e-02, 7.5717e-02],
[-5.4043e-02, 1.1418e-02, 6.2735e-02, -1.8255e-02, 5.9029e-03]],
[[-2.7467e-02, 2.2406e-02, -6.6786e-02, 8.7846e-02, 1.1628e-02],
[ 6.6246e-02, -2.0746e-02, 1.5433e-01, 4.1994e-02, 8.7076e-02],
[-4.2903e-02, 1.0213e-01, 7.2365e-02, 2.7206e-02, 1.3627e-01],
[ 6.4725e-03, -6.9502e-02, 8.2167e-02, 3.2131e-02, -4.3159e-02],
[ 1.0807e-01, -8.2906e-03, 1.3195e-01, 7.2410e-02, -5.3980e-02]],
[[-2.5737e-02, 1.0729e-01, 6.7881e-02, 1.2729e-02, 2.2233e-02],
[-5.4383e-02, -3.2342e-02, 1.1222e-01, 2.9954e-02, -5.6793e-02],
[ 6.1569e-02, -5.0857e-03, -7.2801e-03, 2.1612e-02, 5.6074e-02],
[ 8.1350e-02, 1.3603e-01, 6.5937e-02, 5.7080e-02, -3.3051e-02],
[ 9.2686e-02, 8.6042e-03, 5.0341e-02, 1.2102e-01, -4.2846e-02]],
[[ 3.1385e-02, -1.8880e-02, -4.3763e-02, 9.5897e-02, 2.5644e-03],
[ 1.0728e-01, 1.2668e-01, -2.4991e-02, 1.2733e-01, -1.1972e-02],
[-1.4936e-02, -8.9163e-02, -6.3422e-02, -5.8102e-02, 8.8998e-02],
[-1.2742e-03, 1.0116e-02, -1.3688e-03, 8.9744e-02, 3.2801e-02],
[ 5.3857e-02, 3.3760e-02, 9.7652e-02, -1.2658e-02, -4.1213e-02]],
[[-1.0549e-01, 1.1089e-02, 2.0750e-02, -8.4346e-02, -4.5414e-02],
[-7.8882e-02, 4.7766e-02, 3.7996e-02, -3.4279e-03, 1.0021e-01],
[-1.2175e-02, 8.0692e-02, 1.6572e-01, 1.4422e-01, -6.1436e-02],
[-6.2487e-02, -1.2147e-02, -5.6277e-02, -4.4224e-04, -7.0803e-03],
[-3.5251e-02, -6.1638e-02, 7.6515e-02, -4.3100e-02, -9.2633e-02]],
[[ 8.4949e-02, 1.0109e-01, -5.7688e-02, 7.8953e-02, 1.1039e-02],
[-1.8447e-03, 6.6093e-03, 5.5160e-02, -2.1061e-03, -7.6814e-02],
[-4.7536e-02, 5.4149e-03, 4.1168e-02, -3.2414e-03, -1.2026e-02],
[ 6.3581e-02, 2.1284e-03, 4.8441e-02, -2.8984e-02, 3.7449e-02],
[ 6.2353e-02, 1.7654e-02, 6.6143e-03, -2.6222e-02, 1.9975e-02]]],
[[[-9.0780e-03, 7.4168e-04, 5.0728e-02, 9.6808e-02, 4.4157e-02],
[-4.6064e-02, -4.7032e-02, -3.6292e-02, 1.0961e-01, -5.1720e-02],
[-2.7391e-02, -8.0438e-02, 2.3853e-02, -1.0298e-01, 1.1040e-02],
[-6.7849e-03, 1.1938e-03, 2.2602e-02, 5.5898e-02, 7.7367e-02],
[ 4.9093e-03, 8.2457e-02, -1.4296e-02, -3.9243e-02, 2.8431e-02]],
[[ 1.2188e-01, -9.1309e-02, -2.5300e-02, 3.8637e-02, -1.1966e-02],
[ 7.2569e-02, 3.6117e-02, -5.2285e-02, 5.4047e-02, 4.2884e-03],
[ 3.4672e-02, -7.6760e-04, -2.4154e-02, -3.2986e-02, 6.2598e-02],
[ 3.8618e-02, 5.3779e-02, 5.0678e-02, 2.6810e-02, -1.5022e-02],
[ 5.0676e-02, 9.1586e-02, 1.7718e-02, -4.8524e-02, -2.0987e-02]],
[[ 1.8777e-02, -3.6430e-03, -5.1505e-02, 2.5505e-02, -9.2066e-02],
[-8.3652e-02, 3.8952e-02, 5.1191e-02, 1.6772e-02, 2.0135e-01],
[-8.1641e-02, 7.3640e-02, 7.7579e-02, 1.5251e-02, -2.0884e-02],
[-1.3105e-01, 1.5262e-02, -5.6016e-02, -5.0386e-02, 7.2419e-02],
[-1.2733e-01, -3.3231e-02, 8.8184e-02, 6.6656e-02, 6.6930e-02]],
[[ 4.0774e-02, -1.3535e-01, 1.3731e-02, 2.4058e-02, 1.0693e-02],
[-5.5981e-02, 7.2441e-02, -3.5613e-03, -1.2205e-02, 4.7171e-02],
[-1.1775e-02, 2.7184e-02, -7.1557e-02, 1.0347e-01, 5.8896e-03],
[ 2.0610e-02, -8.3879e-02, -7.2425e-02, 2.1839e-02, -4.3554e-02],
[ 6.8138e-02, 3.1565e-02, -1.4528e-01, 1.7535e-01, 3.2564e-02]],
[[ 1.8692e-02, 3.7771e-02, 5.5205e-02, -1.3458e-02, 4.8685e-02],
[ 2.7660e-02, -1.7240e-02, -4.8733e-02, 9.4653e-02, 5.5204e-02],
[ 1.2270e-01, 6.1749e-03, 5.6012e-02, 1.0850e-01, 9.6627e-02],
[-6.4985e-02, -5.4565e-02, 7.3250e-02, 4.8888e-03, 1.9920e-02],
[ 3.1991e-02, 1.5198e-02, -2.6509e-02, -1.4226e-01, -1.3095e-01]],
[[ 1.2591e-02, -5.1089e-02, 7.1670e-02, 9.0602e-02, -2.2682e-02],
[-1.1734e-01, -1.1738e-01, 2.7109e-02, -1.0356e-01, -3.0563e-02],
[-1.0117e-01, -1.4831e-01, 7.1119e-02, 2.7001e-02, 1.9252e-01],
[ 1.5424e-02, 1.2361e-01, 1.2723e-01, 2.6140e-02, -1.6384e-02],
[-4.4956e-02, 4.8875e-02, -3.7166e-02, -1.5698e-02, -5.9875e-02]]],
[[[-4.0060e-02, -7.5376e-02, 9.0522e-02, -1.3705e-02, 5.0330e-02],
[ 1.2410e-01, -6.8057e-04, -1.6394e-02, -4.6147e-02, 7.4656e-03],
[-2.4582e-02, -4.8235e-02, 1.1284e-02, -2.6863e-02, 4.7994e-02],
[ 1.7968e-02, 1.4713e-02, -4.4666e-02, 8.4956e-02, -1.1188e-01],
[ 6.1636e-02, 5.7068e-02, -7.3509e-02, 1.2213e-02, 7.9951e-02]],
[[ 1.3779e-02, -5.6942e-02, 6.5871e-02, -9.4319e-02, -1.0116e-01],
[-8.5877e-02, -1.4712e-01, 2.4106e-02, 4.5554e-02, -1.7046e-01],
[ 7.1870e-02, 8.6280e-02, 2.3830e-02, 1.2538e-01, -6.0586e-02],
[-5.0712e-02, -5.6716e-02, 2.2811e-02, 4.6917e-03, 2.1000e-02],
[ 1.4774e-02, -3.2031e-02, 1.3210e-01, 1.1982e-01, -8.9682e-02]],
[[ 1.6284e-01, -1.0248e-01, 2.2695e-02, -5.2900e-02, 6.1035e-02],
[ 2.0388e-02, -6.5181e-02, -1.0114e-01, 4.6024e-02, -4.4779e-02],
[-3.0573e-04, -1.1372e-02, -3.3065e-02, 2.6070e-02, -5.5357e-02],
[-1.0599e-02, -5.1236e-02, 9.6448e-02, 7.5023e-02, -7.7667e-02],
[-7.2784e-02, 1.3918e-01, -1.4998e-02, 4.7092e-02, 5.4265e-02]],
[[-3.7319e-02, 3.8862e-02, -6.2182e-02, 6.0732e-02, 1.1152e-01],
[-6.8488e-02, -4.2621e-02, -1.0851e-01, 3.8274e-02, -2.8389e-02],
[-2.4795e-03, 6.5596e-02, -4.4869e-03, 4.4756e-02, -6.8641e-02],
[-1.9861e-02, 2.4411e-02, 1.5822e-01, -8.8753e-02, -4.8671e-02],
[-3.9062e-02, -4.8847e-02, -7.2248e-02, -4.5459e-04, 2.7606e-02]],
[[-3.5982e-03, 6.8191e-02, 9.6101e-02, 5.5497e-02, 7.6344e-02],
[-1.9417e-02, -1.5556e-01, -1.9490e-02, 5.0801e-02, -3.7765e-02],
[-1.1910e-02, -9.1878e-02, 4.2141e-02, -9.9299e-02, -5.6762e-02],
[ 7.9792e-02, 1.8574e-01, 4.4123e-02, -6.9109e-03, -6.1165e-02],
[-1.2580e-01, 3.9366e-02, -1.0186e-01, 3.6060e-02, -6.7405e-02]],
[[-6.8059e-04, 7.9329e-02, 4.2778e-02, 4.9050e-03, -3.9437e-02],
[-3.1596e-02, 5.7778e-02, 1.1185e-01, 4.8292e-02, -4.3491e-02],
[ 9.8594e-02, 6.9557e-02, 1.0506e-02, -1.4311e-01, -7.4874e-02],
[ 4.5525e-02, 1.2678e-01, -5.5705e-04, -3.9545e-02, -5.2146e-02],
[-6.7104e-02, 2.1735e-02, -2.9231e-02, 1.9563e-03, 3.3551e-02]]],
...,
[[[ 7.9785e-03, 9.6521e-02, -1.3083e-02, 8.3893e-02, -2.8526e-02],
[-1.5271e-01, 6.7064e-02, -1.3156e-02, 1.5758e-02, -1.6052e-02],
[-2.0791e-02, 1.1361e-01, -1.4968e-01, 3.5289e-02, -4.0983e-02],
[-3.4127e-02, 7.5340e-02, 2.4503e-02, 1.1301e-01, 1.6968e-01],
[-1.2136e-01, 1.4160e-02, -1.2052e-01, 3.4367e-02, 4.9111e-02]],
[[ 9.0677e-02, -5.5500e-02, -9.1093e-02, 4.9036e-02, 4.8991e-02],
[ 8.5656e-02, -1.3216e-01, -6.3505e-02, -3.2264e-02, 7.5033e-03],
[ 1.0007e-01, 3.5466e-02, 5.0721e-02, -5.8828e-02, -1.1460e-01],
[-3.6455e-02, 1.1234e-02, 2.6212e-02, -8.2686e-02, -6.2735e-02],
[ 2.6178e-02, -3.6884e-03, -1.3710e-01, 1.6569e-02, 2.8620e-02]],
[[-5.0623e-02, -4.4185e-02, 4.6047e-02, 3.9004e-02, -7.5805e-03],
[-4.6561e-02, 4.3014e-02, 1.0764e-02, -4.7813e-03, -4.7869e-02],
[-1.0108e-01, -1.7958e-03, 1.0872e-02, 1.2845e-01, -1.5940e-02],
[ 4.3679e-02, 7.6629e-02, 2.1028e-02, 5.5007e-02, -7.4609e-02],
[ 9.3900e-02, -8.2633e-02, 2.7914e-02, 2.5687e-02, -2.3510e-02]],
[[ 7.7326e-03, 1.1611e-02, 2.9614e-02, 1.0284e-02, 2.9104e-02],
[ 3.0220e-02, -8.0378e-02, -2.7949e-02, 2.2446e-02, 8.0459e-03],
[ 6.5281e-02, -5.7804e-02, 8.3168e-02, 5.4988e-02, -6.6446e-02],
[-7.8295e-02, 3.3191e-02, -9.8877e-02, 1.0647e-01, -3.7064e-03],
[ 9.2358e-02, 1.5699e-01, 7.2006e-02, -3.4939e-02, -2.7969e-02]],
[[ 4.8673e-02, -1.7830e-02, -7.4592e-02, 1.0337e-01, 4.3368e-02],
[ 3.7064e-02, 5.9052e-02, -6.1762e-02, 5.2548e-03, 1.1638e-02],
[-3.2117e-02, -1.3252e-01, 5.2480e-02, -5.5170e-02, -7.9090e-03],
[-9.8518e-02, 9.4210e-02, 1.4956e-02, 6.6029e-02, 3.8304e-02],
[-3.7070e-02, -7.0037e-02, 1.9132e-01, -4.1338e-02, -2.5241e-02]],
[[-4.6659e-02, 4.2209e-03, -4.2241e-02, 1.9754e-02, 7.7949e-02],
[ 6.5216e-02, 2.0466e-02, 1.5637e-02, 3.5587e-04, 6.2204e-02],
[-1.2770e-01, 1.4639e-03, 5.8013e-02, 3.8882e-02, -3.6296e-02],
[-5.8678e-02, 4.7554e-02, -4.9292e-02, 8.2163e-02, -2.8183e-02],
[ 3.1189e-02, -4.4150e-02, -6.2355e-02, 4.8382e-02, 4.5446e-02]]],
[[[ 9.5645e-02, -1.0301e-01, 4.6273e-02, 7.3074e-02, -6.2531e-03],
[-1.1468e-02, -1.7400e-02, 3.8642e-03, 1.0186e-01, 6.0620e-02],
[-3.4799e-03, 9.2991e-02, 1.3765e-01, -1.7145e-01, -8.0353e-02],
[ 4.2481e-02, 4.6607e-02, 1.6294e-02, 3.1393e-02, 2.1176e-01],
[ 1.1362e-01, 8.9815e-03, 6.5340e-02, -4.6683e-02, 6.6436e-02]],
[[-4.0611e-02, 7.9461e-02, 1.3987e-01, 7.9224e-02, 1.3151e-02],
[ 5.1524e-03, -1.9823e-02, -2.6767e-02, 7.3505e-02, 7.4152e-02],
[-9.5245e-02, 5.9597e-02, 3.4118e-02, -3.6963e-02, -1.3391e-02],
[-2.4452e-02, -4.9684e-03, -7.2417e-02, -2.5737e-02, 9.2375e-03],
[ 5.6024e-02, -2.5624e-02, -2.9548e-02, 1.0268e-01, 6.6207e-02]],
[[-2.8858e-02, -1.6285e-03, 2.3269e-03, 5.0814e-02, -1.1348e-01],
[-1.0934e-01, -5.2677e-02, 5.9965e-02, -5.4806e-02, 1.5476e-02],
[ 1.4079e-01, -1.4167e-03, 3.5588e-02, -8.0859e-02, -3.6366e-02],
[ 1.9536e-02, 1.0368e-01, -7.7792e-02, 1.5817e-01, 5.5855e-02],
[-8.3879e-02, 4.9135e-02, 4.2749e-02, 4.6981e-02, -3.3361e-02]],
[[ 8.6765e-02, -7.7088e-03, -4.8883e-02, 3.6100e-02, -6.2219e-02],
[ 4.5523e-02, 5.6251e-02, -5.2107e-02, 1.2930e-01, -5.1350e-02],
[ 6.5248e-03, -1.0238e-01, -6.8020e-02, -1.1667e-01, -1.9077e-02],
[ 6.1764e-02, -1.6211e-02, 1.3773e-01, 9.2896e-02, -5.1193e-02],
[-3.0970e-02, 3.7340e-02, 3.6440e-02, 2.1100e-02, 1.5138e-03]],
[[-3.9962e-02, -1.1714e-01, 5.2935e-02, -9.2095e-02, 5.2764e-02],
[ 5.3439e-02, 8.3302e-02, -1.9304e-02, 3.3452e-02, -1.0472e-01],
[ 5.2374e-03, -7.3624e-02, -2.3803e-02, 5.9410e-02, 2.6254e-02],
[ 4.3904e-02, 7.5527e-02, -5.5117e-02, -3.0872e-02, -7.4420e-02],
[ 5.9820e-02, -1.4020e-01, -3.1974e-02, -1.3241e-01, -8.4654e-02]],
[[ 2.9300e-02, 8.5097e-02, -4.9078e-02, 1.3321e-02, 2.6851e-02],
[-9.7055e-02, 1.2338e-01, -5.9273e-02, -1.1743e-01, -6.2792e-02],
[-8.2991e-03, 3.5198e-02, 1.1417e-01, -7.3450e-02, 1.5150e-02],
[ 2.6375e-02, 2.8174e-03, -3.1688e-02, -6.8431e-02, 1.1147e-01],
[-2.8619e-02, -8.2053e-03, 5.5010e-02, 1.2504e-01, -1.0463e-03]]],
[[[-1.1390e-01, 1.1178e-01, 2.0661e-02, 7.5986e-02, -3.5510e-02],
[ 6.9884e-02, 7.9008e-03, -3.1892e-02, -6.2933e-02, -1.8630e-02],
[-1.6785e-01, -1.3450e-01, 9.2579e-02, -9.7790e-02, 5.4691e-02],
[-4.4550e-02, -8.8956e-03, -5.8851e-02, 1.7682e-01, -4.1232e-02],
[ 1.1308e-01, -1.0594e-01, 8.4209e-02, -9.3546e-02, 4.6904e-02]],
[[ 8.0923e-03, 5.9262e-02, -3.0235e-02, 7.7266e-02, -4.6861e-02],
[ 9.1123e-03, 1.6089e-02, 2.7371e-02, 2.4351e-02, 4.7679e-03],
[-1.2141e-01, 7.4713e-02, -1.1535e-01, 2.6721e-02, 1.0784e-01],
[ 9.8689e-02, -6.4300e-02, 1.3670e-01, 7.1682e-02, -3.5123e-02],
[-1.6633e-02, -1.0464e-01, -4.0829e-02, 5.4382e-02, 7.7213e-03]],
[[ 3.5077e-02, 4.2135e-02, 3.0175e-02, -4.3647e-02, 1.5406e-01],
[-1.6523e-02, 4.2702e-02, 3.7327e-03, -2.4782e-02, -1.9466e-02],
[-9.2885e-05, 4.0268e-02, -2.1415e-02, -2.9881e-02, -5.3303e-02],
[-2.6457e-02, 7.5620e-03, 2.6807e-02, -1.3104e-01, 1.0279e-02],
[-5.7137e-02, -4.6889e-02, -5.5574e-02, -9.9751e-02, -3.5737e-02]],
[[-9.9814e-02, -9.5435e-02, -2.6537e-02, 3.1237e-02, 1.5104e-02],
[-9.9790e-02, 2.9716e-03, 5.0457e-02, -8.9010e-02, -3.9197e-02],
[ 3.2547e-02, 6.0830e-03, -2.2157e-02, -7.0033e-02, -1.0172e-01],
[ 4.9774e-02, 5.4871e-02, 5.6303e-02, 4.8631e-02, -3.4986e-02],
[ 1.2138e-02, -2.2211e-02, -3.3582e-02, -1.7598e-02, -1.0158e-02]],
[[-7.2373e-03, -4.2170e-02, 5.9605e-02, -8.4486e-02, 6.3807e-02],
[ 1.0146e-01, -4.4409e-02, 6.9227e-02, 2.7258e-02, -4.4921e-02],
[ 2.0162e-02, -1.3525e-01, -3.3158e-02, -4.9422e-02, 6.1210e-02],
[-6.3034e-02, 8.9230e-02, 2.8164e-02, -1.8634e-02, 1.1193e-02],
[-1.5343e-01, 5.8297e-02, 4.0476e-02, -8.3992e-03, -2.8954e-02]],
[[-1.4177e-02, -1.4059e-01, 3.4488e-02, -1.1158e-01, 2.7263e-02],
[ 9.5575e-02, -1.5587e-03, 6.7742e-02, -9.6203e-03, -5.8986e-02],
[-1.8375e-03, -4.2862e-02, 7.6257e-02, 5.3290e-02, -5.9313e-02],
[-1.1387e-02, -1.2381e-01, 2.8557e-02, 1.3121e-02, 1.9561e-02],
[ 9.1249e-02, -3.3576e-03, -3.5484e-02, 2.8714e-02, -7.3947e-03]]]],
grad_fn=<AddBackward0>)
(bias): Normal:
posterior: Automatic()
prior: Module()
observed: Observed()
scale: Transform:
tensor([0.0498, 0.0498, 0.0498, 0.0498, 0.0498, 0.0498, 0.0498, 0.0498, 0.0498,
0.0498, 0.0498, 0.0498, 0.0498, 0.0498, 0.0498, 0.0498],
grad_fn=<ExpBackward0>)
loc: Parameter containing:
tensor([-3.9858e-02, -6.2551e-02, -1.4050e-02, 5.5231e-02, 1.7201e-02,
-2.6817e-02, 2.8532e-02, -2.6857e-02, -2.8279e-02, 9.3339e-04,
-5.8307e-02, -6.2624e-02, 7.1998e-05, 1.1212e-02, -2.0352e-02,
-5.8423e-02], requires_grad=True)
tensor: tensor([-0.0379, -0.0506, 0.0362, 0.0304, -0.0709, -0.0483, -0.0155, -0.0590,
-0.0637, 0.0148, -0.0911, -0.1465, 0.0204, 0.0746, -0.0178, -0.1370],
grad_fn=<AddBackward0>)
)
(prior): Module(
(weight): Normal:
loc: tensor([[[[0., -0., 0., -0., -0.],
[0., 0., 0., 0., 0.],
[-0., -0., -0., -0., 0.],
[-0., 0., -0., 0., 0.],
[-0., 0., 0., -0., 0.]],
[[-0., 0., -0., 0., 0.],
[0., -0., 0., 0., 0.],
[-0., 0., 0., 0., 0.],
[-0., -0., 0., 0., -0.],
[0., 0., 0., 0., -0.]],
[[-0., 0., 0., -0., 0.],
[-0., -0., 0., -0., -0.],
[0., 0., 0., 0., 0.],
[0., 0., 0., 0., 0.],
[0., -0., 0., 0., -0.]],
[[-0., -0., 0., 0., 0.],
[0., 0., 0., 0., -0.],
[-0., -0., -0., 0., 0.],
[0., -0., 0., 0., -0.],
[0., 0., 0., 0., 0.]],
[[-0., -0., -0., -0., -0.],
[-0., 0., 0., 0., 0.],
[-0., 0., 0., 0., 0.],
[-0., -0., -0., -0., 0.],
[0., -0., 0., 0., -0.]],
[[0., 0., -0., 0., 0.],
[-0., -0., 0., -0., -0.],
[0., 0., 0., -0., 0.],
[0., -0., -0., -0., 0.],
[0., -0., -0., -0., 0.]]],
[[[-0., 0., -0., 0., 0.],
[-0., -0., -0., 0., -0.],
[-0., -0., -0., -0., 0.],
[-0., -0., 0., -0., 0.],
[-0., 0., -0., -0., -0.]],
[[0., -0., -0., -0., -0.],
[0., -0., -0., 0., 0.],
[0., -0., 0., 0., 0.],
[0., -0., -0., 0., -0.],
[-0., 0., -0., -0., -0.]],
[[-0., -0., -0., 0., -0.],
[-0., 0., 0., 0., 0.],
[-0., 0., -0., 0., -0.],
[-0., 0., -0., 0., 0.],
[-0., -0., 0., -0., 0.]],
[[-0., -0., 0., 0., 0.],
[-0., 0., -0., 0., 0.],
[-0., 0., -0., 0., 0.],
[-0., -0., 0., -0., -0.],
[0., 0., -0., 0., 0.]],
[[0., 0., -0., -0., 0.],
[-0., -0., 0., 0., 0.],
[-0., -0., 0., 0., 0.],
[-0., -0., -0., 0., 0.],
[0., 0., 0., -0., -0.]],
[[0., -0., 0., 0., 0.],
[-0., -0., 0., 0., -0.],
[-0., -0., -0., 0., 0.],
[-0., 0., 0., 0., -0.],
[-0., 0., -0., -0., -0.]]],
[[[0., -0., 0., 0., 0.],
[0., -0., -0., -0., -0.],
[-0., -0., 0., -0., 0.],
[0., -0., -0., 0., -0.],
[0., 0., -0., 0., 0.]],
[[0., -0., 0., -0., -0.],
[-0., -0., -0., 0., -0.],
[0., 0., -0., 0., -0.],
[-0., 0., 0., 0., 0.],
[0., 0., 0., 0., -0.]],
[[0., -0., 0., -0., 0.],
[-0., -0., -0., 0., -0.],
[0., -0., -0., -0., -0.],
[-0., -0., 0., 0., 0.],
[-0., 0., 0., 0., -0.]],
[[-0., 0., -0., -0., 0.],
[-0., -0., -0., -0., -0.],
[-0., 0., 0., 0., -0.],
[0., 0., 0., -0., -0.],
[0., 0., -0., -0., 0.]],
[[0., 0., 0., -0., 0.],
[-0., 0., -0., -0., -0.],
[0., -0., 0., -0., 0.],
[0., 0., 0., 0., -0.],
[-0., 0., -0., 0., -0.]],
[[-0., 0., 0., 0., 0.],
[-0., -0., 0., 0., -0.],
[0., 0., -0., 0., -0.],
[0., 0., -0., -0., -0.],
[-0., 0., -0., -0., -0.]]],
...,
[[[0., 0., -0., 0., -0.],
[-0., 0., 0., 0., -0.],
[-0., 0., -0., 0., -0.],
[0., 0., 0., 0., 0.],
[0., 0., -0., 0., 0.]],
[[0., 0., -0., 0., 0.],
[-0., -0., -0., -0., -0.],
[0., 0., 0., -0., -0.],
[-0., 0., 0., 0., -0.],
[0., 0., -0., 0., 0.]],
[[0., -0., -0., 0., -0.],
[-0., 0., -0., 0., -0.],
[0., 0., -0., 0., -0.],
[0., 0., 0., 0., -0.],
[-0., -0., 0., -0., 0.]],
[[0., -0., 0., 0., -0.],
[0., -0., -0., 0., -0.],
[-0., -0., 0., -0., -0.],
[0., 0., -0., 0., -0.],
[-0., 0., 0., -0., -0.]],
[[0., 0., -0., 0., 0.],
[0., 0., -0., -0., -0.],
[-0., -0., -0., -0., 0.],
[-0., 0., 0., 0., 0.],
[-0., -0., 0., -0., -0.]],
[[-0., 0., -0., -0., 0.],
[0., 0., -0., 0., 0.],
[-0., -0., 0., 0., -0.],
[-0., -0., -0., 0., 0.],
[-0., 0., -0., 0., -0.]]],
[[[0., -0., -0., 0., -0.],
[0., 0., 0., -0., 0.],
[-0., 0., 0., -0., -0.],
[0., 0., -0., -0., 0.],
[0., -0., 0., -0., 0.]],
[[-0., 0., 0., 0., -0.],
[-0., 0., -0., 0., 0.],
[-0., 0., -0., -0., 0.],
[-0., 0., -0., -0., 0.],
[0., -0., -0., 0., 0.]],
[[0., 0., -0., 0., -0.],
[-0., -0., 0., 0., -0.],
[0., 0., 0., -0., -0.],
[-0., 0., -0., 0., 0.],
[-0., 0., 0., 0., 0.]],
[[0., 0., 0., 0., -0.],
[0., 0., -0., 0., -0.],
[-0., -0., -0., -0., -0.],
[0., -0., 0., 0., -0.],
[-0., -0., -0., -0., 0.]],
[[-0., -0., -0., -0., -0.],
[0., 0., -0., -0., -0.],
[-0., -0., 0., 0., 0.],
[-0., 0., -0., -0., -0.],
[0., -0., 0., -0., -0.]],
[[0., 0., -0., 0., -0.],
[-0., -0., -0., -0., -0.],
[0., 0., 0., -0., -0.],
[0., -0., -0., -0., 0.],
[-0., -0., 0., 0., -0.]]],
[[[-0., 0., 0., 0., -0.],
[0., -0., -0., -0., 0.],
[-0., -0., 0., -0., -0.],
[-0., -0., -0., 0., 0.],
[0., 0., 0., -0., 0.]],
[[0., 0., 0., 0., -0.],
[-0., 0., 0., -0., -0.],
[-0., 0., -0., -0., 0.],
[0., -0., 0., 0., -0.],
[-0., -0., -0., 0., 0.]],
[[0., -0., 0., -0., 0.],
[-0., 0., 0., 0., -0.],
[-0., -0., -0., -0., 0.],
[-0., 0., 0., -0., 0.],
[-0., 0., 0., 0., -0.]],
[[-0., -0., -0., 0., 0.],
[-0., -0., 0., 0., -0.],
[0., -0., -0., -0., -0.],
[0., 0., 0., 0., 0.],
[-0., 0., -0., 0., -0.]],
[[-0., -0., 0., -0., 0.],
[0., -0., 0., 0., 0.],
[0., -0., -0., -0., 0.],
[-0., -0., -0., -0., 0.],
[-0., 0., 0., -0., 0.]],
[[-0., -0., -0., -0., -0.],
[0., -0., 0., -0., -0.],
[0., 0., 0., -0., 0.],
[0., -0., 0., 0., -0.],
[0., 0., -0., 0., 0.]]]])
scale: tensor([[[[0.0816, 0.0816, 0.0816, 0.0816, 0.0816],
[0.0816, 0.0816, 0.0816, 0.0816, 0.0816],
[0.0816, 0.0816, 0.0816, 0.0816, 0.0816],
[0.0816, 0.0816, 0.0816, 0.0816, 0.0816],
[0.0816, 0.0816, 0.0816, 0.0816, 0.0816]],
[[0.0816, 0.0816, 0.0816, 0.0816, 0.0816],
[0.0816, 0.0816, 0.0816, 0.0816, 0.0816],
[0.0816, 0.0816, 0.0816, 0.0816, 0.0816],
[0.0816, 0.0816, 0.0816, 0.0816, 0.0816],
[0.0816, 0.0816, 0.0816, 0.0816, 0.0816]],
[[0.0816, 0.0816, 0.0816, 0.0816, 0.0816],
[0.0816, 0.0816, 0.0816, 0.0816, 0.0816],
[0.0816, 0.0816, 0.0816, 0.0816, 0.0816],
[0.0816, 0.0816, 0.0816, 0.0816, 0.0816],
[0.0816, 0.0816, 0.0816, 0.0816, 0.0816]],
[[0.0816, 0.0816, 0.0816, 0.0816, 0.0816],
[0.0816, 0.0816, 0.0816, 0.0816, 0.0816],
[0.0816, 0.0816, 0.0816, 0.0816, 0.0816],
[0.0816, 0.0816, 0.0816, 0.0816, 0.0816],
[0.0816, 0.0816, 0.0816, 0.0816, 0.0816]],
[[0.0816, 0.0816, 0.0816, 0.0816, 0.0816],
[0.0816, 0.0816, 0.0816, 0.0816, 0.0816],
[0.0816, 0.0816, 0.0816, 0.0816, 0.0816],
[0.0816, 0.0816, 0.0816, 0.0816, 0.0816],
[0.0816, 0.0816, 0.0816, 0.0816, 0.0816]],
[[0.0816, 0.0816, 0.0816, 0.0816, 0.0816],
[0.0816, 0.0816, 0.0816, 0.0816, 0.0816],
[0.0816, 0.0816, 0.0816, 0.0816, 0.0816],
[0.0816, 0.0816, 0.0816, 0.0816, 0.0816],
[0.0816, 0.0816, 0.0816, 0.0816, 0.0816]]],
[[[0.0816, 0.0816, 0.0816, 0.0816, 0.0816],
[0.0816, 0.0816, 0.0816, 0.0816, 0.0816],
[0.0816, 0.0816, 0.0816, 0.0816, 0.0816],
[0.0816, 0.0816, 0.0816, 0.0816, 0.0816],
[0.0816, 0.0816, 0.0816, 0.0816, 0.0816]],
[[0.0816, 0.0816, 0.0816, 0.0816, 0.0816],
[0.0816, 0.0816, 0.0816, 0.0816, 0.0816],
[0.0816, 0.0816, 0.0816, 0.0816, 0.0816],
[0.0816, 0.0816, 0.0816, 0.0816, 0.0816],
[0.0816, 0.0816, 0.0816, 0.0816, 0.0816]],
[[0.0816, 0.0816, 0.0816, 0.0816, 0.0816],
[0.0816, 0.0816, 0.0816, 0.0816, 0.0816],
[0.0816, 0.0816, 0.0816, 0.0816, 0.0816],
[0.0816, 0.0816, 0.0816, 0.0816, 0.0816],
[0.0816, 0.0816, 0.0816, 0.0816, 0.0816]],
[[0.0816, 0.0816, 0.0816, 0.0816, 0.0816],
[0.0816, 0.0816, 0.0816, 0.0816, 0.0816],
[0.0816, 0.0816, 0.0816, 0.0816, 0.0816],
[0.0816, 0.0816, 0.0816, 0.0816, 0.0816],
[0.0816, 0.0816, 0.0816, 0.0816, 0.0816]],
[[0.0816, 0.0816, 0.0816, 0.0816, 0.0816],
[0.0816, 0.0816, 0.0816, 0.0816, 0.0816],
[0.0816, 0.0816, 0.0816, 0.0816, 0.0816],
[0.0816, 0.0816, 0.0816, 0.0816, 0.0816],
[0.0816, 0.0816, 0.0816, 0.0816, 0.0816]],
[[0.0816, 0.0816, 0.0816, 0.0816, 0.0816],
[0.0816, 0.0816, 0.0816, 0.0816, 0.0816],
[0.0816, 0.0816, 0.0816, 0.0816, 0.0816],
[0.0816, 0.0816, 0.0816, 0.0816, 0.0816],
[0.0816, 0.0816, 0.0816, 0.0816, 0.0816]]],
[[[0.0816, 0.0816, 0.0816, 0.0816, 0.0816],
[0.0816, 0.0816, 0.0816, 0.0816, 0.0816],
[0.0816, 0.0816, 0.0816, 0.0816, 0.0816],
[0.0816, 0.0816, 0.0816, 0.0816, 0.0816],
[0.0816, 0.0816, 0.0816, 0.0816, 0.0816]],
[[0.0816, 0.0816, 0.0816, 0.0816, 0.0816],
[0.0816, 0.0816, 0.0816, 0.0816, 0.0816],
[0.0816, 0.0816, 0.0816, 0.0816, 0.0816],
[0.0816, 0.0816, 0.0816, 0.0816, 0.0816],
[0.0816, 0.0816, 0.0816, 0.0816, 0.0816]],
[[0.0816, 0.0816, 0.0816, 0.0816, 0.0816],
[0.0816, 0.0816, 0.0816, 0.0816, 0.0816],
[0.0816, 0.0816, 0.0816, 0.0816, 0.0816],
[0.0816, 0.0816, 0.0816, 0.0816, 0.0816],
[0.0816, 0.0816, 0.0816, 0.0816, 0.0816]],
[[0.0816, 0.0816, 0.0816, 0.0816, 0.0816],
[0.0816, 0.0816, 0.0816, 0.0816, 0.0816],
[0.0816, 0.0816, 0.0816, 0.0816, 0.0816],
[0.0816, 0.0816, 0.0816, 0.0816, 0.0816],
[0.0816, 0.0816, 0.0816, 0.0816, 0.0816]],
[[0.0816, 0.0816, 0.0816, 0.0816, 0.0816],
[0.0816, 0.0816, 0.0816, 0.0816, 0.0816],
[0.0816, 0.0816, 0.0816, 0.0816, 0.0816],
[0.0816, 0.0816, 0.0816, 0.0816, 0.0816],
[0.0816, 0.0816, 0.0816, 0.0816, 0.0816]],
[[0.0816, 0.0816, 0.0816, 0.0816, 0.0816],
[0.0816, 0.0816, 0.0816, 0.0816, 0.0816],
[0.0816, 0.0816, 0.0816, 0.0816, 0.0816],
[0.0816, 0.0816, 0.0816, 0.0816, 0.0816],
[0.0816, 0.0816, 0.0816, 0.0816, 0.0816]]],
...,
[[[0.0816, 0.0816, 0.0816, 0.0816, 0.0816],
[0.0816, 0.0816, 0.0816, 0.0816, 0.0816],
[0.0816, 0.0816, 0.0816, 0.0816, 0.0816],
[0.0816, 0.0816, 0.0816, 0.0816, 0.0816],
[0.0816, 0.0816, 0.0816, 0.0816, 0.0816]],
[[0.0816, 0.0816, 0.0816, 0.0816, 0.0816],
[0.0816, 0.0816, 0.0816, 0.0816, 0.0816],
[0.0816, 0.0816, 0.0816, 0.0816, 0.0816],
[0.0816, 0.0816, 0.0816, 0.0816, 0.0816],
[0.0816, 0.0816, 0.0816, 0.0816, 0.0816]],
[[0.0816, 0.0816, 0.0816, 0.0816, 0.0816],
[0.0816, 0.0816, 0.0816, 0.0816, 0.0816],
[0.0816, 0.0816, 0.0816, 0.0816, 0.0816],
[0.0816, 0.0816, 0.0816, 0.0816, 0.0816],
[0.0816, 0.0816, 0.0816, 0.0816, 0.0816]],
[[0.0816, 0.0816, 0.0816, 0.0816, 0.0816],
[0.0816, 0.0816, 0.0816, 0.0816, 0.0816],
[0.0816, 0.0816, 0.0816, 0.0816, 0.0816],
[0.0816, 0.0816, 0.0816, 0.0816, 0.0816],
[0.0816, 0.0816, 0.0816, 0.0816, 0.0816]],
[[0.0816, 0.0816, 0.0816, 0.0816, 0.0816],
[0.0816, 0.0816, 0.0816, 0.0816, 0.0816],
[0.0816, 0.0816, 0.0816, 0.0816, 0.0816],
[0.0816, 0.0816, 0.0816, 0.0816, 0.0816],
[0.0816, 0.0816, 0.0816, 0.0816, 0.0816]],
[[0.0816, 0.0816, 0.0816, 0.0816, 0.0816],
[0.0816, 0.0816, 0.0816, 0.0816, 0.0816],
[0.0816, 0.0816, 0.0816, 0.0816, 0.0816],
[0.0816, 0.0816, 0.0816, 0.0816, 0.0816],
[0.0816, 0.0816, 0.0816, 0.0816, 0.0816]]],
[[[0.0816, 0.0816, 0.0816, 0.0816, 0.0816],
[0.0816, 0.0816, 0.0816, 0.0816, 0.0816],
[0.0816, 0.0816, 0.0816, 0.0816, 0.0816],
[0.0816, 0.0816, 0.0816, 0.0816, 0.0816],
[0.0816, 0.0816, 0.0816, 0.0816, 0.0816]],
[[0.0816, 0.0816, 0.0816, 0.0816, 0.0816],
[0.0816, 0.0816, 0.0816, 0.0816, 0.0816],
[0.0816, 0.0816, 0.0816, 0.0816, 0.0816],
[0.0816, 0.0816, 0.0816, 0.0816, 0.0816],
[0.0816, 0.0816, 0.0816, 0.0816, 0.0816]],
[[0.0816, 0.0816, 0.0816, 0.0816, 0.0816],
[0.0816, 0.0816, 0.0816, 0.0816, 0.0816],
[0.0816, 0.0816, 0.0816, 0.0816, 0.0816],
[0.0816, 0.0816, 0.0816, 0.0816, 0.0816],
[0.0816, 0.0816, 0.0816, 0.0816, 0.0816]],
[[0.0816, 0.0816, 0.0816, 0.0816, 0.0816],
[0.0816, 0.0816, 0.0816, 0.0816, 0.0816],
[0.0816, 0.0816, 0.0816, 0.0816, 0.0816],
[0.0816, 0.0816, 0.0816, 0.0816, 0.0816],
[0.0816, 0.0816, 0.0816, 0.0816, 0.0816]],
[[0.0816, 0.0816, 0.0816, 0.0816, 0.0816],
[0.0816, 0.0816, 0.0816, 0.0816, 0.0816],
[0.0816, 0.0816, 0.0816, 0.0816, 0.0816],
[0.0816, 0.0816, 0.0816, 0.0816, 0.0816],
[0.0816, 0.0816, 0.0816, 0.0816, 0.0816]],
[[0.0816, 0.0816, 0.0816, 0.0816, 0.0816],
[0.0816, 0.0816, 0.0816, 0.0816, 0.0816],
[0.0816, 0.0816, 0.0816, 0.0816, 0.0816],
[0.0816, 0.0816, 0.0816, 0.0816, 0.0816],
[0.0816, 0.0816, 0.0816, 0.0816, 0.0816]]],
[[[0.0816, 0.0816, 0.0816, 0.0816, 0.0816],
[0.0816, 0.0816, 0.0816, 0.0816, 0.0816],
[0.0816, 0.0816, 0.0816, 0.0816, 0.0816],
[0.0816, 0.0816, 0.0816, 0.0816, 0.0816],
[0.0816, 0.0816, 0.0816, 0.0816, 0.0816]],
[[0.0816, 0.0816, 0.0816, 0.0816, 0.0816],
[0.0816, 0.0816, 0.0816, 0.0816, 0.0816],
[0.0816, 0.0816, 0.0816, 0.0816, 0.0816],
[0.0816, 0.0816, 0.0816, 0.0816, 0.0816],
[0.0816, 0.0816, 0.0816, 0.0816, 0.0816]],
[[0.0816, 0.0816, 0.0816, 0.0816, 0.0816],
[0.0816, 0.0816, 0.0816, 0.0816, 0.0816],
[0.0816, 0.0816, 0.0816, 0.0816, 0.0816],
[0.0816, 0.0816, 0.0816, 0.0816, 0.0816],
[0.0816, 0.0816, 0.0816, 0.0816, 0.0816]],
[[0.0816, 0.0816, 0.0816, 0.0816, 0.0816],
[0.0816, 0.0816, 0.0816, 0.0816, 0.0816],
[0.0816, 0.0816, 0.0816, 0.0816, 0.0816],
[0.0816, 0.0816, 0.0816, 0.0816, 0.0816],
[0.0816, 0.0816, 0.0816, 0.0816, 0.0816]],
[[0.0816, 0.0816, 0.0816, 0.0816, 0.0816],
[0.0816, 0.0816, 0.0816, 0.0816, 0.0816],
[0.0816, 0.0816, 0.0816, 0.0816, 0.0816],
[0.0816, 0.0816, 0.0816, 0.0816, 0.0816],
[0.0816, 0.0816, 0.0816, 0.0816, 0.0816]],
[[0.0816, 0.0816, 0.0816, 0.0816, 0.0816],
[0.0816, 0.0816, 0.0816, 0.0816, 0.0816],
[0.0816, 0.0816, 0.0816, 0.0816, 0.0816],
[0.0816, 0.0816, 0.0816, 0.0816, 0.0816],
[0.0816, 0.0816, 0.0816, 0.0816, 0.0816]]]])
posterior: Automatic()
prior: Module()
observed: Observed()
tensor: tensor([[[[ 0.0228, -0.0484, 0.0637, -0.0249, -0.0302],
[ 0.0105, 0.0733, 0.0576, 0.0393, 0.0094],
[-0.0590, -0.0747, -0.0506, -0.0495, 0.0402],
[-0.0647, 0.0789, -0.0096, 0.0477, 0.0700],
[-0.0450, 0.0176, 0.0443, -0.0176, 0.0407]],
[[-0.0148, 0.0075, -0.0554, 0.0808, 0.0812],
[ 0.0636, -0.0776, 0.0617, 0.0497, 0.0760],
[-0.0350, 0.0633, 0.0318, 0.0562, 0.0742],
[-0.0229, -0.0075, 0.0338, 0.0709, -0.0382],
[ 0.0655, 0.0249, 0.0556, 0.0329, -0.0734]],
[[-0.0531, 0.0393, 0.0355, -0.0358, 0.0298],
[-0.0772, -0.0365, 0.0397, -0.0093, -0.0571],
[ 0.0087, 0.0256, 0.0568, 0.0123, 0.0098],
[ 0.0070, 0.0767, 0.0580, 0.0573, 0.0128],
[ 0.0476, -0.0090, 0.0190, 0.0597, -0.0108]],
[[-0.0097, -0.0755, 0.0251, 0.0349, 0.0423],
[ 0.0800, 0.0594, 0.0019, 0.0419, -0.0641],
[-0.0501, -0.0639, -0.0783, 0.0276, 0.0425],
[ 0.0793, -0.0161, 0.0031, 0.0700, -0.0077],
[ 0.0717, 0.0564, 0.0130, 0.0126, 0.0055]],
[[-0.0711, -0.0004, -0.0350, -0.0729, -0.0567],
[-0.0351, 0.0119, 0.0163, 0.0542, 0.0558],
[-0.0308, 0.0337, 0.0634, 0.0556, 0.0051],
[-0.0305, -0.0103, -0.0515, -0.0746, 0.0701],
[ 0.0014, -0.0249, 0.0724, 0.0282, -0.0773]],
[[ 0.0647, 0.0735, -0.0460, 0.0321, 0.0060],
[-0.0624, -0.0262, 0.0131, -0.0270, -0.0768],
[ 0.0053, 0.0301, 0.0262, -0.0012, 0.0230],
[ 0.0758, -0.0297, -0.0011, -0.0424, 0.0659],
[ 0.0155, -0.0027, -0.0785, -0.0715, 0.0603]]],
[[[-0.0680, 0.0027, -0.0140, 0.0451, 0.0738],
[-0.0323, -0.0084, -0.0281, 0.0522, -0.0210],
[-0.0525, -0.0399, -0.0072, -0.0577, 0.0280],
[-0.0374, -0.0055, 0.0143, -0.0258, 0.0404],
[-0.0804, 0.0563, -0.0096, -0.0485, -0.0052]],
[[ 0.0772, -0.0417, -0.0619, -0.0375, -0.0217],
[ 0.0312, -0.0186, -0.0225, 0.0433, 0.0327],
[ 0.0751, -0.0210, 0.0098, 0.0275, 0.0606],
[ 0.0308, -0.0484, -0.0063, 0.0799, -0.0184],
[-0.0195, 0.0014, -0.0263, -0.0651, -0.0203]],
[[-0.0441, -0.0229, -0.0547, 0.0813, -0.0195],
[-0.0343, 0.0006, 0.0058, 0.0113, 0.0668],
[-0.0674, 0.0276, -0.0072, 0.0360, -0.0707],
[-0.0468, 0.0240, -0.0658, 0.0561, 0.0396],
[-0.0704, -0.0553, 0.0372, -0.0310, 0.0525]],
[[-0.0081, -0.0563, 0.0125, 0.0304, 0.0486],
[-0.0515, 0.0539, -0.0165, 0.0263, 0.0241],
[-0.0189, 0.0789, -0.0254, 0.0502, 0.0336],
[-0.0288, -0.0308, 0.0397, -0.0238, -0.0365],
[ 0.0195, 0.0625, -0.0496, 0.0805, 0.0320]],
[[ 0.0451, 0.0478, -0.0347, -0.0325, 0.0203],
[-0.0460, -0.0406, 0.0251, 0.0666, 0.0718],
[-0.0109, -0.0145, 0.0313, 0.0103, 0.0525],
[-0.0299, -0.0280, -0.0036, 0.0186, 0.0350],
[ 0.0073, 0.0463, 0.0292, -0.0732, -0.0485]],
[[ 0.0059, -0.0394, 0.0745, 0.0484, 0.0053],
[-0.0341, -0.0371, 0.0608, 0.0058, -0.0537],
[-0.0605, -0.0791, -0.0253, 0.0762, 0.0762],
[-0.0059, 0.0153, 0.0580, 0.0198, -0.0173],
[-0.0174, 0.0578, -0.0368, -0.0736, -0.0631]]],
[[[ 0.0033, -0.0336, 0.0105, 0.0191, 0.0074],
[ 0.0714, -0.0613, -0.0475, -0.0452, -0.0098],
[-0.0617, -0.0515, 0.0353, -0.0375, 0.0360],
[ 0.0542, -0.0326, -0.0236, 0.0674, -0.0347],
[ 0.0394, 0.0746, -0.0467, 0.0522, 0.0424]],
[[ 0.0503, -0.0693, 0.0626, -0.0424, -0.0488],
[-0.0617, -0.0619, -0.0342, 0.0127, -0.0253],
[ 0.0319, 0.0270, -0.0089, 0.0175, -0.0308],
[-0.0551, 0.0107, 0.0212, 0.0026, 0.0019],
[ 0.0115, 0.0343, 0.0572, 0.0811, -0.0354]],
[[ 0.0632, -0.0197, 0.0072, -0.0372, 0.0581],
[-0.0492, -0.0602, -0.0201, 0.0465, -0.0482],
[ 0.0403, -0.0086, -0.0292, -0.0711, -0.0294],
[-0.0228, -0.0102, 0.0450, 0.0804, 0.0032],
[-0.0009, 0.0815, 0.0113, 0.0315, -0.0110]],
[[-0.0627, 0.0313, -0.0443, -0.0353, 0.0797],
[-0.0039, -0.0053, -0.0678, -0.0335, -0.0764],
[-0.0202, 0.0296, 0.0548, 0.0185, -0.0028],
[ 0.0044, 0.0496, 0.0734, -0.0418, -0.0792],
[ 0.0025, 0.0171, -0.0681, -0.0224, 0.0077]],
[[ 0.0424, 0.0476, 0.0079, -0.0167, 0.0523],
[-0.0335, 0.0003, -0.0560, -0.0448, -0.0173],
[ 0.0426, -0.0002, 0.0133, -0.0705, 0.0085],
[ 0.0724, 0.0664, 0.0455, 0.0461, -0.0607],
[-0.0546, 0.0696, -0.0658, 0.0364, -0.0562]],
[[-0.0390, 0.0704, 0.0786, 0.0033, 0.0295],
[-0.0145, -0.0198, 0.0638, 0.0332, -0.0197],
[ 0.0663, 0.0625, -0.0292, 0.0036, -0.0408],
[ 0.0201, 0.0786, -0.0615, -0.0266, -0.0627],
[-0.0323, 0.0509, -0.0516, -0.0624, -0.0025]]],
...,
[[[ 0.0777, 0.0133, -0.0427, 0.0800, -0.0246],
[-0.0430, 0.0370, 0.0550, 0.0002, -0.0453],
[-0.0782, 0.0747, -0.0467, 0.0130, -0.0217],
[ 0.0417, 0.0371, 0.0066, 0.0761, 0.0344],
[ 0.0590, 0.0188, -0.0681, 0.0579, 0.0579]],
[[ 0.0803, 0.0284, -0.0485, 0.0127, 0.0668],
[-0.0049, -0.0313, -0.0693, -0.0646, -0.0111],
[ 0.0460, 0.0529, 0.0564, -0.0790, -0.0583],
[-0.0102, 0.0005, 0.0437, 0.0308, -0.0437],
[ 0.0205, 0.0394, -0.0644, 0.0770, 0.0368]],
[[ 0.0157, -0.0298, -0.0073, 0.0653, -0.0469],
[-0.0527, 0.0319, -0.0647, 0.0084, -0.0241],
[ 0.0004, 0.0514, -0.0025, 0.0630, -0.0294],
[ 0.0689, 0.0191, 0.0546, 0.0365, -0.0539],
[-0.0045, -0.0745, 0.0544, -0.0686, 0.0816]],
[[ 0.0061, -0.0117, 0.0211, 0.0131, -0.0410],
[ 0.0168, -0.0675, -0.0497, 0.0285, -0.0718],
[-0.0401, -0.0345, 0.0003, -0.0376, -0.0554],
[ 0.0065, 0.0611, -0.0163, 0.0418, -0.0605],
[-0.0081, 0.0791, 0.0381, -0.0200, -0.0402]],
[[ 0.0350, 0.0182, -0.0750, 0.0471, 0.0045],
[ 0.0441, 0.0266, -0.0399, -0.0425, -0.0318],
[-0.0428, -0.0136, -0.0466, -0.0151, 0.0100],
[-0.0644, 0.0688, 0.0709, 0.0026, 0.0585],
[-0.0269, -0.0130, 0.0715, -0.0413, -0.0011]],
[[-0.0537, 0.0477, -0.0147, -0.0553, 0.0459],
[ 0.0619, 0.0005, -0.0194, 0.0200, 0.0296],
[-0.0205, -0.0382, 0.0277, 0.0322, -0.0201],
[-0.0251, -0.0238, -0.0758, 0.0534, 0.0487],
[-0.0023, 0.0328, -0.0237, 0.0328, -0.0399]]],
[[[ 0.0501, -0.0288, -0.0173, 0.0392, -0.0050],
[ 0.0156, 0.0807, 0.0459, -0.0056, 0.0346],
[-0.0486, 0.0271, 0.0530, -0.0753, -0.0678],
[ 0.0070, 0.0007, -0.0208, -0.0359, 0.0587],
[ 0.0728, -0.0304, 0.0334, -0.0355, 0.0323]],
[[-0.0734, 0.0403, 0.0763, 0.0570, -0.0033],
[-0.0664, 0.0105, -0.0366, 0.0042, 0.0546],
[-0.0745, 0.0639, -0.0417, -0.0033, 0.0495],
[-0.0550, 0.0292, -0.0198, -0.0241, 0.0367],
[ 0.0107, -0.0213, -0.0517, 0.0664, 0.0738]],
[[ 0.0123, 0.0034, -0.0046, 0.0620, -0.0356],
[-0.0580, -0.0810, 0.0235, 0.0637, -0.0323],
[ 0.0501, 0.0611, 0.0696, -0.0725, -0.0392],
[-0.0327, 0.0616, -0.0098, 0.0648, 0.0049],
[-0.0644, 0.0434, 0.0465, 0.0378, 0.0250]],
[[ 0.0816, 0.0159, 0.0255, 0.0219, -0.0652],
[ 0.0783, 0.0748, -0.0595, 0.0515, -0.0486],
[-0.0709, -0.0491, -0.0587, -0.0085, -0.0437],
[ 0.0395, -0.0117, 0.0683, 0.0806, -0.0066],
[-0.0332, -0.0257, -0.0023, -0.0359, 0.0064]],
[[-0.0328, -0.0616, -0.0107, -0.0231, -0.0393],
[ 0.0030, 0.0048, -0.0813, -0.0253, -0.0723],
[-0.0680, -0.0350, 0.0409, 0.0464, 0.0235],
[-0.0085, 0.0688, -0.0767, -0.0011, -0.0570],
[ 0.0553, -0.0721, 0.0039, -0.0811, -0.0608]],
[[ 0.0617, 0.0303, -0.0521, 0.0155, -0.0364],
[-0.0589, -0.0223, -0.0112, -0.0599, -0.0590],
[ 0.0729, 0.0326, 0.0761, -0.0415, -0.0048],
[ 0.0036, -0.0197, -0.0393, -0.0060, 0.0785],
[-0.0679, -0.0750, 0.0671, 0.0385, -0.0260]]],
[[[-0.0456, 0.0197, 0.0548, 0.0420, -0.0569],
[ 0.0518, -0.0172, -0.0758, -0.0328, 0.0196],
[-0.0712, -0.0446, 0.0593, -0.0403, -0.0250],
[-0.0142, -0.0058, -0.0283, 0.0783, 0.0075],
[ 0.0755, 0.0161, 0.0319, -0.0562, 0.0378]],
[[ 0.0572, 0.0773, 0.0243, 0.0638, -0.0472],
[-0.0081, 0.0225, 0.0298, -0.0442, -0.0075],
[-0.0814, 0.0369, -0.0680, -0.0471, 0.0187],
[ 0.0290, -0.0338, 0.0786, 0.0685, -0.0263],
[-0.0453, -0.0716, -0.0462, 0.0556, 0.0159]],
[[ 0.0359, -0.0511, 0.0707, -0.0696, 0.0407],
[-0.0717, 0.0521, 0.0813, 0.0335, -0.0515],
[-0.0295, -0.0124, -0.0406, -0.0247, 0.0162],
[-0.0252, 0.0105, 0.0624, -0.0701, 0.0153],
[-0.0490, 0.0815, 0.0331, 0.0130, -0.0174]],
[[-0.0801, -0.0390, -0.0246, 0.0187, 0.0752],
[-0.0654, -0.0242, 0.0666, 0.0303, -0.0114],
[ 0.0783, -0.0565, -0.0200, -0.0462, -0.0119],
[ 0.0788, 0.0656, 0.0623, 0.0350, 0.0254],
[-0.0227, 0.0380, -0.0172, 0.0293, -0.0065]],
[[-0.0086, -0.0572, 0.0217, -0.0286, 0.0476],
[ 0.0695, -0.0679, 0.0714, 0.0371, 0.0638],
[ 0.0099, -0.0652, -0.0545, -0.0068, 0.0805],
[-0.0506, -0.0737, -0.0110, -0.0198, 0.0047],
[-0.0288, 0.0730, 0.0794, -0.0033, 0.0242]],
[[-0.0616, -0.0632, -0.0110, -0.0658, -0.0470],
[ 0.0425, -0.0136, 0.0665, -0.0201, -0.0727],
[ 0.0189, 0.0189, 0.0641, -0.0384, 0.0180],
[ 0.0002, -0.0737, 0.0365, 0.0311, -0.0378],
[ 0.0789, 0.0037, -0.0582, 0.0148, 0.0323]]]])
(bias): Normal:
loc: tensor([-0., -0., -0., 0., 0., -0., 0., -0., -0., 0., -0., -0., 0., 0., -0., -0.])
scale: tensor([0.2500, 0.2500, 0.2500, 0.2500, 0.2500, 0.2500, 0.2500, 0.2500, 0.2500,
0.2500, 0.2500, 0.2500, 0.2500, 0.2500, 0.2500, 0.2500])
posterior: Automatic()
prior: Module()
observed: Observed()
tensor: tensor([-3.9858e-02, -6.2551e-02, -1.4050e-02, 5.5231e-02, 1.7201e-02,
-2.6817e-02, 2.8532e-02, -2.6857e-02, -2.8279e-02, 9.3339e-04,
-5.8307e-02, -6.2624e-02, 7.1998e-05, 1.1212e-02, -2.0352e-02,
-5.8423e-02])
)
(observed): Observed()
)
(fc1): Linear(
in_features=400, out_features=120, bias=True
(posterior): Normal(
(weight): Normal:
posterior: Automatic()
prior: Module()
observed: Observed()
scale: Transform:
tensor([[0.0498, 0.0498, 0.0498, ..., 0.0498, 0.0498, 0.0498],
[0.0498, 0.0498, 0.0498, ..., 0.0498, 0.0498, 0.0498],
[0.0498, 0.0498, 0.0498, ..., 0.0498, 0.0498, 0.0498],
...,
[0.0498, 0.0498, 0.0498, ..., 0.0498, 0.0498, 0.0498],
[0.0498, 0.0498, 0.0498, ..., 0.0498, 0.0498, 0.0498],
[0.0498, 0.0498, 0.0498, ..., 0.0498, 0.0498, 0.0498]],
grad_fn=<ExpBackward0>)
loc: Parameter containing:
tensor([[ 0.0337, -0.0484, 0.0253, ..., -0.0286, -0.0485, -0.0091],
[ 0.0469, 0.0292, 0.0258, ..., 0.0047, -0.0409, 0.0409],
[ 0.0367, 0.0313, 0.0040, ..., 0.0234, -0.0487, -0.0428],
...,
[-0.0139, -0.0203, 0.0175, ..., -0.0324, -0.0387, 0.0258],
[ 0.0096, 0.0293, 0.0120, ..., 0.0264, 0.0297, 0.0347],
[-0.0126, -0.0436, 0.0311, ..., -0.0195, 0.0352, 0.0191]],
requires_grad=True)
tensor: tensor([[ 0.0732, -0.0255, -0.0578, ..., -0.0882, -0.0816, -0.0213],
[ 0.1146, 0.0466, -0.0112, ..., 0.0476, 0.0129, 0.0057],
[ 0.0800, -0.0415, -0.0323, ..., 0.0578, -0.0181, -0.0136],
...,
[-0.0417, -0.0255, 0.0027, ..., -0.0518, -0.0193, 0.0991],
[-0.0082, 0.0043, 0.0350, ..., 0.0453, 0.0136, 0.0155],
[ 0.0230, -0.0360, -0.0075, ..., -0.0218, 0.0702, 0.0851]],
grad_fn=<AddBackward0>)
(bias): Normal:
posterior: Automatic()
prior: Module()
observed: Observed()
scale: Transform:
tensor([0.0498, 0.0498, 0.0498, 0.0498, 0.0498, 0.0498, 0.0498, 0.0498, 0.0498,
0.0498, 0.0498, 0.0498, 0.0498, 0.0498, 0.0498, 0.0498, 0.0498, 0.0498,
0.0498, 0.0498, 0.0498, 0.0498, 0.0498, 0.0498, 0.0498, 0.0498, 0.0498,
0.0498, 0.0498, 0.0498, 0.0498, 0.0498, 0.0498, 0.0498, 0.0498, 0.0498,
0.0498, 0.0498, 0.0498, 0.0498, 0.0498, 0.0498, 0.0498, 0.0498, 0.0498,
0.0498, 0.0498, 0.0498, 0.0498, 0.0498, 0.0498, 0.0498, 0.0498, 0.0498,
0.0498, 0.0498, 0.0498, 0.0498, 0.0498, 0.0498, 0.0498, 0.0498, 0.0498,
0.0498, 0.0498, 0.0498, 0.0498, 0.0498, 0.0498, 0.0498, 0.0498, 0.0498,
0.0498, 0.0498, 0.0498, 0.0498, 0.0498, 0.0498, 0.0498, 0.0498, 0.0498,
0.0498, 0.0498, 0.0498, 0.0498, 0.0498, 0.0498, 0.0498, 0.0498, 0.0498,
0.0498, 0.0498, 0.0498, 0.0498, 0.0498, 0.0498, 0.0498, 0.0498, 0.0498,
0.0498, 0.0498, 0.0498, 0.0498, 0.0498, 0.0498, 0.0498, 0.0498, 0.0498,
0.0498, 0.0498, 0.0498, 0.0498, 0.0498, 0.0498, 0.0498, 0.0498, 0.0498,
0.0498, 0.0498, 0.0498], grad_fn=<ExpBackward0>)
loc: Parameter containing:
tensor([ 0.0227, -0.0239, -0.0015, 0.0243, 0.0179, 0.0240, 0.0462, 0.0295,
-0.0409, -0.0290, -0.0125, -0.0128, -0.0085, 0.0326, -0.0219, 0.0423,
0.0489, -0.0179, -0.0246, -0.0301, 0.0485, -0.0077, 0.0185, 0.0481,
-0.0032, -0.0149, -0.0211, -0.0451, 0.0282, -0.0297, 0.0339, -0.0216,
-0.0034, -0.0181, -0.0157, 0.0106, -0.0460, -0.0273, -0.0288, -0.0371,
-0.0023, -0.0038, 0.0074, 0.0185, -0.0443, 0.0172, -0.0478, 0.0460,
0.0363, 0.0282, -0.0387, -0.0109, -0.0231, 0.0013, -0.0196, 0.0272,
0.0049, 0.0265, -0.0020, -0.0435, 0.0185, -0.0403, 0.0289, 0.0379,
-0.0112, -0.0080, -0.0427, 0.0491, -0.0431, -0.0402, -0.0102, -0.0105,
-0.0474, -0.0193, -0.0236, -0.0411, 0.0166, 0.0335, -0.0161, -0.0324,
-0.0196, 0.0304, -0.0400, 0.0024, 0.0160, 0.0193, 0.0080, -0.0252,
0.0398, -0.0498, 0.0386, 0.0138, 0.0152, 0.0196, 0.0355, -0.0123,
-0.0179, 0.0390, 0.0361, -0.0140, -0.0484, 0.0458, 0.0205, -0.0043,
-0.0300, 0.0102, -0.0160, -0.0108, -0.0162, 0.0330, 0.0324, -0.0429,
0.0008, 0.0134, 0.0364, -0.0246, 0.0498, 0.0140, -0.0339, 0.0392],
requires_grad=True)
tensor: tensor([ 1.1447e-03, -5.1819e-02, 1.6640e-02, 6.5499e-02, -8.0907e-02,
-4.8519e-02, 5.6677e-02, -4.8956e-02, -7.7639e-03, -5.7330e-03,
-7.1541e-02, -8.8738e-03, 3.3044e-03, 3.5168e-02, -5.3030e-06,
1.1116e-01, 5.3860e-02, 1.8072e-02, -4.6139e-02, -4.0794e-02,
2.5714e-02, -9.4141e-03, -4.2178e-04, -2.5774e-02, -7.7567e-02,
-8.2163e-02, 2.3858e-02, -8.7571e-02, 6.2566e-02, -8.3659e-02,
1.3654e-01, -4.0236e-03, 9.7985e-02, 9.0484e-02, 5.9550e-02,
2.4722e-02, -3.9010e-02, -2.2623e-02, 5.7798e-02, -3.9112e-02,
-3.8512e-02, 7.7203e-02, -3.1446e-02, -2.6466e-02, -4.2992e-02,
3.5493e-03, -2.0020e-02, 5.2569e-03, -2.8117e-02, 7.0782e-02,
9.1462e-04, -4.1758e-02, -1.9813e-03, 3.4395e-02, 6.1325e-02,
-1.3560e-02, -8.4248e-02, -4.5469e-02, -1.5659e-02, -1.4572e-02,
-1.8009e-02, -1.7755e-02, 5.8105e-03, 2.0910e-02, 5.4390e-03,
-5.3495e-02, -1.9020e-02, 6.4683e-02, -5.4073e-02, -3.9146e-02,
5.6359e-02, -2.8833e-02, -1.3513e-02, 4.1304e-02, -8.1053e-02,
-4.9787e-02, -8.1154e-03, -6.9614e-02, -8.6335e-02, -8.2642e-02,
1.1239e-02, -2.9094e-02, -7.9066e-02, 8.6259e-02, 2.8128e-02,
-1.7309e-02, -3.9473e-02, 3.1097e-03, 6.6451e-02, 2.0752e-02,
6.5958e-02, 1.3649e-02, 1.2680e-01, 1.0900e-02, -2.6149e-02,
5.0733e-02, 4.0548e-02, 7.8464e-02, 4.5867e-02, 3.9911e-02,
-5.9929e-02, 3.6826e-02, -5.9146e-02, -3.4829e-02, -8.1041e-02,
5.6606e-02, -2.2417e-02, -4.2888e-03, -3.5870e-02, 2.8688e-02,
1.1745e-01, 1.3533e-03, 4.2071e-02, -3.4773e-02, 1.8792e-02,
-3.4782e-02, 7.3807e-02, -2.7561e-02, -1.0201e-01, 9.2736e-02],
grad_fn=<AddBackward0>)
)
(prior): Module(
(weight): Normal:
loc: tensor([[0., -0., 0., ..., -0., -0., -0.],
[0., 0., 0., ..., 0., -0., 0.],
[0., 0., 0., ..., 0., -0., -0.],
...,
[-0., -0., 0., ..., -0., -0., 0.],
[0., 0., 0., ..., 0., 0., 0.],
[-0., -0., 0., ..., -0., 0., 0.]])
scale: tensor([[0.0500, 0.0500, 0.0500, ..., 0.0500, 0.0500, 0.0500],
[0.0500, 0.0500, 0.0500, ..., 0.0500, 0.0500, 0.0500],
[0.0500, 0.0500, 0.0500, ..., 0.0500, 0.0500, 0.0500],
...,
[0.0500, 0.0500, 0.0500, ..., 0.0500, 0.0500, 0.0500],
[0.0500, 0.0500, 0.0500, ..., 0.0500, 0.0500, 0.0500],
[0.0500, 0.0500, 0.0500, ..., 0.0500, 0.0500, 0.0500]])
posterior: Automatic()
prior: Module()
observed: Observed()
tensor: tensor([[ 0.0337, -0.0484, 0.0253, ..., -0.0286, -0.0485, -0.0091],
[ 0.0469, 0.0292, 0.0258, ..., 0.0047, -0.0409, 0.0409],
[ 0.0367, 0.0313, 0.0040, ..., 0.0234, -0.0487, -0.0428],
...,
[-0.0139, -0.0203, 0.0175, ..., -0.0324, -0.0387, 0.0258],
[ 0.0096, 0.0293, 0.0120, ..., 0.0264, 0.0297, 0.0347],
[-0.0126, -0.0436, 0.0311, ..., -0.0195, 0.0352, 0.0191]])
(bias): Normal:
loc: tensor([0., -0., -0., 0., 0., 0., 0., 0., -0., -0., -0., -0., -0., 0., -0., 0., 0., -0., -0., -0., 0., -0., 0., 0.,
-0., -0., -0., -0., 0., -0., 0., -0., -0., -0., -0., 0., -0., -0., -0., -0., -0., -0., 0., 0., -0., 0., -0., 0.,
0., 0., -0., -0., -0., 0., -0., 0., 0., 0., -0., -0., 0., -0., 0., 0., -0., -0., -0., 0., -0., -0., -0., -0.,
-0., -0., -0., -0., 0., 0., -0., -0., -0., 0., -0., 0., 0., 0., 0., -0., 0., -0., 0., 0., 0., 0., 0., -0.,
-0., 0., 0., -0., -0., 0., 0., -0., -0., 0., -0., -0., -0., 0., 0., -0., 0., 0., 0., -0., 0., 0., -0., 0.])
scale: tensor([0.0913, 0.0913, 0.0913, 0.0913, 0.0913, 0.0913, 0.0913, 0.0913, 0.0913,
0.0913, 0.0913, 0.0913, 0.0913, 0.0913, 0.0913, 0.0913, 0.0913, 0.0913,
0.0913, 0.0913, 0.0913, 0.0913, 0.0913, 0.0913, 0.0913, 0.0913, 0.0913,
0.0913, 0.0913, 0.0913, 0.0913, 0.0913, 0.0913, 0.0913, 0.0913, 0.0913,
0.0913, 0.0913, 0.0913, 0.0913, 0.0913, 0.0913, 0.0913, 0.0913, 0.0913,
0.0913, 0.0913, 0.0913, 0.0913, 0.0913, 0.0913, 0.0913, 0.0913, 0.0913,
0.0913, 0.0913, 0.0913, 0.0913, 0.0913, 0.0913, 0.0913, 0.0913, 0.0913,
0.0913, 0.0913, 0.0913, 0.0913, 0.0913, 0.0913, 0.0913, 0.0913, 0.0913,
0.0913, 0.0913, 0.0913, 0.0913, 0.0913, 0.0913, 0.0913, 0.0913, 0.0913,
0.0913, 0.0913, 0.0913, 0.0913, 0.0913, 0.0913, 0.0913, 0.0913, 0.0913,
0.0913, 0.0913, 0.0913, 0.0913, 0.0913, 0.0913, 0.0913, 0.0913, 0.0913,
0.0913, 0.0913, 0.0913, 0.0913, 0.0913, 0.0913, 0.0913, 0.0913, 0.0913,
0.0913, 0.0913, 0.0913, 0.0913, 0.0913, 0.0913, 0.0913, 0.0913, 0.0913,
0.0913, 0.0913, 0.0913])
posterior: Automatic()
prior: Module()
observed: Observed()
tensor: tensor([ 0.0227, -0.0239, -0.0015, 0.0243, 0.0179, 0.0240, 0.0462, 0.0295,
-0.0409, -0.0290, -0.0125, -0.0128, -0.0085, 0.0326, -0.0219, 0.0423,
0.0489, -0.0179, -0.0246, -0.0301, 0.0485, -0.0077, 0.0185, 0.0481,
-0.0032, -0.0149, -0.0211, -0.0451, 0.0282, -0.0297, 0.0339, -0.0216,
-0.0034, -0.0181, -0.0157, 0.0106, -0.0460, -0.0273, -0.0288, -0.0371,
-0.0023, -0.0038, 0.0074, 0.0185, -0.0443, 0.0172, -0.0478, 0.0460,
0.0363, 0.0282, -0.0387, -0.0109, -0.0231, 0.0013, -0.0196, 0.0272,
0.0049, 0.0265, -0.0020, -0.0435, 0.0185, -0.0403, 0.0289, 0.0379,
-0.0112, -0.0080, -0.0427, 0.0491, -0.0431, -0.0402, -0.0102, -0.0105,
-0.0474, -0.0193, -0.0236, -0.0411, 0.0166, 0.0335, -0.0161, -0.0324,
-0.0196, 0.0304, -0.0400, 0.0024, 0.0160, 0.0193, 0.0080, -0.0252,
0.0398, -0.0498, 0.0386, 0.0138, 0.0152, 0.0196, 0.0355, -0.0123,
-0.0179, 0.0390, 0.0361, -0.0140, -0.0484, 0.0458, 0.0205, -0.0043,
-0.0300, 0.0102, -0.0160, -0.0108, -0.0162, 0.0330, 0.0324, -0.0429,
0.0008, 0.0134, 0.0364, -0.0246, 0.0498, 0.0140, -0.0339, 0.0392])
)
(observed): Observed()
)
(fc2): Linear(
in_features=120, out_features=2, bias=True
(posterior): Normal(
(weight): Normal:
posterior: Automatic()
prior: Module()
observed: Observed()
scale: Transform:
tensor([[0.0498, 0.0498, 0.0498, 0.0498, 0.0498, 0.0498, 0.0498, 0.0498, 0.0498,
0.0498, 0.0498, 0.0498, 0.0498, 0.0498, 0.0498, 0.0498, 0.0498, 0.0498,
0.0498, 0.0498, 0.0498, 0.0498, 0.0498, 0.0498, 0.0498, 0.0498, 0.0498,
0.0498, 0.0498, 0.0498, 0.0498, 0.0498, 0.0498, 0.0498, 0.0498, 0.0498,
0.0498, 0.0498, 0.0498, 0.0498, 0.0498, 0.0498, 0.0498, 0.0498, 0.0498,
0.0498, 0.0498, 0.0498, 0.0498, 0.0498, 0.0498, 0.0498, 0.0498, 0.0498,
0.0498, 0.0498, 0.0498, 0.0498, 0.0498, 0.0498, 0.0498, 0.0498, 0.0498,
0.0498, 0.0498, 0.0498, 0.0498, 0.0498, 0.0498, 0.0498, 0.0498, 0.0498,
0.0498, 0.0498, 0.0498, 0.0498, 0.0498, 0.0498, 0.0498, 0.0498, 0.0498,
0.0498, 0.0498, 0.0498, 0.0498, 0.0498, 0.0498, 0.0498, 0.0498, 0.0498,
0.0498, 0.0498, 0.0498, 0.0498, 0.0498, 0.0498, 0.0498, 0.0498, 0.0498,
0.0498, 0.0498, 0.0498, 0.0498, 0.0498, 0.0498, 0.0498, 0.0498, 0.0498,
0.0498, 0.0498, 0.0498, 0.0498, 0.0498, 0.0498, 0.0498, 0.0498, 0.0498,
0.0498, 0.0498, 0.0498],
[0.0498, 0.0498, 0.0498, 0.0498, 0.0498, 0.0498, 0.0498, 0.0498, 0.0498,
0.0498, 0.0498, 0.0498, 0.0498, 0.0498, 0.0498, 0.0498, 0.0498, 0.0498,
0.0498, 0.0498, 0.0498, 0.0498, 0.0498, 0.0498, 0.0498, 0.0498, 0.0498,
0.0498, 0.0498, 0.0498, 0.0498, 0.0498, 0.0498, 0.0498, 0.0498, 0.0498,
0.0498, 0.0498, 0.0498, 0.0498, 0.0498, 0.0498, 0.0498, 0.0498, 0.0498,
0.0498, 0.0498, 0.0498, 0.0498, 0.0498, 0.0498, 0.0498, 0.0498, 0.0498,
0.0498, 0.0498, 0.0498, 0.0498, 0.0498, 0.0498, 0.0498, 0.0498, 0.0498,
0.0498, 0.0498, 0.0498, 0.0498, 0.0498, 0.0498, 0.0498, 0.0498, 0.0498,
0.0498, 0.0498, 0.0498, 0.0498, 0.0498, 0.0498, 0.0498, 0.0498, 0.0498,
0.0498, 0.0498, 0.0498, 0.0498, 0.0498, 0.0498, 0.0498, 0.0498, 0.0498,
0.0498, 0.0498, 0.0498, 0.0498, 0.0498, 0.0498, 0.0498, 0.0498, 0.0498,
0.0498, 0.0498, 0.0498, 0.0498, 0.0498, 0.0498, 0.0498, 0.0498, 0.0498,
0.0498, 0.0498, 0.0498, 0.0498, 0.0498, 0.0498, 0.0498, 0.0498, 0.0498,
0.0498, 0.0498, 0.0498]], grad_fn=<ExpBackward0>)
loc: Parameter containing:
tensor([[ 0.0552, -0.0814, -0.0594, -0.0877, 0.0260, 0.0764, 0.0603, -0.0690,
-0.0617, 0.0816, 0.0810, 0.0872, -0.0560, 0.0101, 0.0851, -0.0210,
-0.0042, 0.0030, 0.0784, -0.0529, 0.0262, 0.0469, 0.0635, 0.0703,
-0.0497, -0.0707, -0.0536, -0.0544, 0.0331, -0.0068, 0.0086, -0.0108,
-0.0055, -0.0076, 0.0049, -0.0805, 0.0471, 0.0136, -0.0250, -0.0345,
0.0100, -0.0639, 0.0511, 0.0064, -0.0230, 0.0270, -0.0317, -0.0259,
-0.0307, -0.0506, -0.0617, -0.0262, -0.0688, -0.0900, 0.0328, 0.0166,
-0.0635, 0.0072, 0.0536, -0.0793, -0.0726, 0.0021, -0.0797, 0.0413,
0.0666, 0.0185, -0.0274, -0.0572, -0.0811, -0.0831, -0.0043, -0.0517,
-0.0639, 0.0666, -0.0211, -0.0832, -0.0887, 0.0568, -0.0423, 0.0110,
-0.0736, 0.0800, 0.0822, 0.0823, 0.0557, 0.0646, -0.0729, 0.0857,
0.0337, -0.0079, 0.0632, -0.0813, -0.0178, -0.0147, 0.0018, 0.0151,
0.0909, -0.0895, 0.0524, -0.0362, -0.0328, 0.0500, 0.0494, -0.0450,
-0.0204, -0.0412, 0.0766, -0.0161, -0.0584, -0.0680, 0.0278, 0.0007,
-0.0566, 0.0467, 0.0536, -0.0230, 0.0731, 0.0413, -0.0785, -0.0119],
[ 0.0685, -0.0193, 0.0604, 0.0138, 0.0828, 0.0634, -0.0749, 0.0419,
-0.0212, -0.0736, -0.0009, 0.0376, -0.0012, 0.0102, -0.0813, -0.0153,
-0.0003, -0.0116, 0.0483, -0.0689, -0.0361, -0.0136, 0.0256, -0.0330,
0.0639, 0.0103, -0.0673, 0.0759, 0.0751, 0.0812, 0.0010, -0.0302,
0.0461, -0.0861, -0.0432, -0.0070, 0.0077, -0.0529, 0.0748, 0.0328,
-0.0610, 0.0524, 0.0129, 0.0665, 0.0886, -0.0200, 0.0524, 0.0696,
-0.0629, -0.0878, 0.0708, 0.0556, -0.0741, 0.0161, -0.0897, 0.0735,
0.0793, -0.0354, 0.0309, -0.0521, 0.0006, 0.0265, -0.0274, 0.0792,
0.0860, -0.0430, -0.0282, 0.0335, -0.0274, 0.0322, 0.0616, -0.0157,
-0.0142, 0.0187, 0.0102, -0.0078, 0.0554, -0.0854, -0.0591, 0.0875,
-0.0630, -0.0741, -0.0793, 0.0149, 0.0818, -0.0127, -0.0881, 0.0015,
-0.0594, 0.0065, -0.0806, 0.0295, -0.0144, 0.0879, 0.0663, 0.0900,
0.0258, -0.0155, 0.0731, -0.0102, -0.0611, -0.0473, -0.0538, 0.0004,
-0.0415, -0.0457, -0.0481, -0.0759, 0.0621, -0.0188, 0.0160, 0.0484,
-0.0819, -0.0031, 0.0558, 0.0735, 0.0219, -0.0744, -0.0153, -0.0397]],
requires_grad=True)
tensor: tensor([[ 0.0155, -0.1381, -0.0459, -0.1173, 0.0346, 0.0912, 0.0721, -0.0664,
0.0061, 0.0576, 0.1026, 0.0628, -0.0534, -0.0191, 0.0756, -0.0014,
0.0800, 0.0257, 0.0461, -0.0446, -0.0176, 0.1071, 0.1174, -0.0118,
-0.0709, -0.1369, -0.0317, -0.0239, 0.0649, 0.0462, -0.0075, 0.0183,
0.0289, 0.0655, 0.0824, -0.0964, 0.1179, 0.0729, -0.0761, 0.0014,
0.1068, -0.0576, 0.0095, -0.0066, -0.0316, -0.0846, 0.0223, -0.0321,
-0.1153, -0.0251, 0.0454, 0.0507, -0.1412, -0.1396, 0.0252, 0.0494,
-0.1182, -0.0545, 0.0342, -0.1580, -0.0585, 0.0391, -0.1098, 0.0674,
0.0748, 0.0277, -0.1225, -0.0614, -0.0111, -0.0965, -0.0249, -0.0885,
-0.0442, 0.0160, 0.0275, 0.0262, -0.1012, -0.0106, -0.1147, -0.0267,
-0.0619, 0.1101, 0.0542, 0.0399, 0.0702, 0.0073, -0.0637, 0.0863,
0.0255, 0.0294, 0.0963, -0.1078, 0.0166, -0.0889, -0.0432, 0.0194,
0.1233, -0.1050, 0.0582, -0.0849, -0.0124, 0.0670, -0.0239, -0.0239,
-0.0879, -0.0510, 0.0547, 0.0301, -0.1166, -0.1147, -0.0102, 0.0029,
-0.0347, 0.1045, 0.0452, -0.0121, 0.1167, 0.0179, -0.0259, -0.0612],
[ 0.0958, -0.0319, 0.0668, 0.0619, 0.1232, 0.0888, -0.0047, 0.1079,
-0.0665, -0.0406, -0.1337, 0.0833, 0.0152, 0.0016, -0.1006, -0.0306,
-0.0050, -0.0641, -0.0259, -0.0763, -0.0317, -0.0147, 0.0674, -0.0595,
0.0330, -0.0659, -0.0305, 0.0856, 0.0609, 0.0719, -0.0262, -0.0806,
0.1531, -0.1720, 0.0591, -0.0311, 0.0744, -0.0318, 0.0379, -0.0081,
-0.0684, 0.0417, 0.0077, 0.0655, 0.0082, -0.1232, -0.0634, 0.1215,
-0.0478, -0.1038, 0.0839, 0.1241, -0.0877, -0.0262, -0.0386, 0.1171,
0.1270, -0.0735, 0.0073, -0.0233, -0.0936, 0.0021, -0.0512, 0.0938,
0.0540, -0.1652, -0.0778, 0.0516, 0.0366, 0.0138, 0.0603, -0.0893,
0.0044, 0.0672, 0.0058, 0.0546, -0.0217, -0.0781, -0.0321, 0.0029,
-0.0905, -0.1680, -0.0874, -0.0058, 0.1265, -0.0179, -0.0598, 0.0163,
-0.0342, 0.0565, -0.0393, 0.0613, -0.0383, 0.0677, 0.1313, 0.1180,
0.0259, -0.0481, 0.1126, 0.0305, -0.0701, -0.0525, -0.0936, 0.0749,
-0.0364, -0.0313, -0.0123, 0.0132, 0.0251, 0.0448, -0.0225, 0.1201,
-0.1090, -0.0413, 0.1151, 0.1027, 0.0813, -0.0764, -0.0066, -0.0852]],
grad_fn=<AddBackward0>)
(bias): Normal:
posterior: Automatic()
prior: Module()
observed: Observed()
scale: Transform:
tensor([0.0498, 0.0498], grad_fn=<ExpBackward0>)
loc: Parameter containing:
tensor([ 0.0430, -0.0439], requires_grad=True)
tensor: tensor([ 0.1120, -0.0376], grad_fn=<AddBackward0>)
)
(prior): Module(
(weight): Normal:
loc: tensor([[0., -0., -0., -0., 0., 0., 0., -0., -0., 0., 0., 0., -0., 0., 0., -0., -0., 0., 0., -0., 0., 0., 0., 0.,
-0., -0., -0., -0., 0., -0., 0., -0., -0., -0., 0., -0., 0., 0., -0., -0., 0., -0., 0., 0., -0., 0., -0., -0.,
-0., -0., -0., -0., -0., -0., 0., 0., -0., 0., 0., -0., -0., 0., -0., 0., 0., 0., -0., -0., -0., -0., -0., -0.,
-0., 0., -0., -0., -0., 0., -0., 0., -0., 0., 0., 0., 0., 0., -0., 0., 0., -0., 0., -0., -0., -0., 0., 0.,
0., -0., 0., -0., -0., 0., 0., -0., -0., -0., 0., -0., -0., -0., 0., 0., -0., 0., 0., -0., 0., 0., -0., -0.],
[0., -0., 0., 0., 0., 0., -0., 0., -0., -0., -0., 0., -0., 0., -0., -0., -0., -0., 0., -0., -0., -0., 0., -0.,
0., 0., -0., 0., 0., 0., 0., -0., 0., -0., -0., -0., 0., -0., 0., 0., -0., 0., 0., 0., 0., -0., 0., 0.,
-0., -0., 0., 0., -0., 0., -0., 0., 0., -0., 0., -0., 0., 0., -0., 0., 0., -0., -0., 0., -0., 0., 0., -0.,
-0., 0., 0., -0., 0., -0., -0., 0., -0., -0., -0., 0., 0., -0., -0., 0., -0., 0., -0., 0., -0., 0., 0., 0.,
0., -0., 0., -0., -0., -0., -0., 0., -0., -0., -0., -0., 0., -0., 0., 0., -0., -0., 0., 0., 0., -0., -0., -0.]])
scale: tensor([[0.0913, 0.0913, 0.0913, 0.0913, 0.0913, 0.0913, 0.0913, 0.0913, 0.0913,
0.0913, 0.0913, 0.0913, 0.0913, 0.0913, 0.0913, 0.0913, 0.0913, 0.0913,
0.0913, 0.0913, 0.0913, 0.0913, 0.0913, 0.0913, 0.0913, 0.0913, 0.0913,
0.0913, 0.0913, 0.0913, 0.0913, 0.0913, 0.0913, 0.0913, 0.0913, 0.0913,
0.0913, 0.0913, 0.0913, 0.0913, 0.0913, 0.0913, 0.0913, 0.0913, 0.0913,
0.0913, 0.0913, 0.0913, 0.0913, 0.0913, 0.0913, 0.0913, 0.0913, 0.0913,
0.0913, 0.0913, 0.0913, 0.0913, 0.0913, 0.0913, 0.0913, 0.0913, 0.0913,
0.0913, 0.0913, 0.0913, 0.0913, 0.0913, 0.0913, 0.0913, 0.0913, 0.0913,
0.0913, 0.0913, 0.0913, 0.0913, 0.0913, 0.0913, 0.0913, 0.0913, 0.0913,
0.0913, 0.0913, 0.0913, 0.0913, 0.0913, 0.0913, 0.0913, 0.0913, 0.0913,
0.0913, 0.0913, 0.0913, 0.0913, 0.0913, 0.0913, 0.0913, 0.0913, 0.0913,
0.0913, 0.0913, 0.0913, 0.0913, 0.0913, 0.0913, 0.0913, 0.0913, 0.0913,
0.0913, 0.0913, 0.0913, 0.0913, 0.0913, 0.0913, 0.0913, 0.0913, 0.0913,
0.0913, 0.0913, 0.0913],
[0.0913, 0.0913, 0.0913, 0.0913, 0.0913, 0.0913, 0.0913, 0.0913, 0.0913,
0.0913, 0.0913, 0.0913, 0.0913, 0.0913, 0.0913, 0.0913, 0.0913, 0.0913,
0.0913, 0.0913, 0.0913, 0.0913, 0.0913, 0.0913, 0.0913, 0.0913, 0.0913,
0.0913, 0.0913, 0.0913, 0.0913, 0.0913, 0.0913, 0.0913, 0.0913, 0.0913,
0.0913, 0.0913, 0.0913, 0.0913, 0.0913, 0.0913, 0.0913, 0.0913, 0.0913,
0.0913, 0.0913, 0.0913, 0.0913, 0.0913, 0.0913, 0.0913, 0.0913, 0.0913,
0.0913, 0.0913, 0.0913, 0.0913, 0.0913, 0.0913, 0.0913, 0.0913, 0.0913,
0.0913, 0.0913, 0.0913, 0.0913, 0.0913, 0.0913, 0.0913, 0.0913, 0.0913,
0.0913, 0.0913, 0.0913, 0.0913, 0.0913, 0.0913, 0.0913, 0.0913, 0.0913,
0.0913, 0.0913, 0.0913, 0.0913, 0.0913, 0.0913, 0.0913, 0.0913, 0.0913,
0.0913, 0.0913, 0.0913, 0.0913, 0.0913, 0.0913, 0.0913, 0.0913, 0.0913,
0.0913, 0.0913, 0.0913, 0.0913, 0.0913, 0.0913, 0.0913, 0.0913, 0.0913,
0.0913, 0.0913, 0.0913, 0.0913, 0.0913, 0.0913, 0.0913, 0.0913, 0.0913,
0.0913, 0.0913, 0.0913]])
posterior: Automatic()
prior: Module()
observed: Observed()
tensor: tensor([[ 0.0552, -0.0814, -0.0594, -0.0877, 0.0260, 0.0764, 0.0603, -0.0690,
-0.0617, 0.0816, 0.0810, 0.0872, -0.0560, 0.0101, 0.0851, -0.0210,
-0.0042, 0.0030, 0.0784, -0.0529, 0.0262, 0.0469, 0.0635, 0.0703,
-0.0497, -0.0707, -0.0536, -0.0544, 0.0331, -0.0068, 0.0086, -0.0108,
-0.0055, -0.0076, 0.0049, -0.0805, 0.0471, 0.0136, -0.0250, -0.0345,
0.0100, -0.0639, 0.0511, 0.0064, -0.0230, 0.0270, -0.0317, -0.0259,
-0.0307, -0.0506, -0.0617, -0.0262, -0.0688, -0.0900, 0.0328, 0.0166,
-0.0635, 0.0072, 0.0536, -0.0793, -0.0726, 0.0021, -0.0797, 0.0413,
0.0666, 0.0185, -0.0274, -0.0572, -0.0811, -0.0831, -0.0043, -0.0517,
-0.0639, 0.0666, -0.0211, -0.0832, -0.0887, 0.0568, -0.0423, 0.0110,
-0.0736, 0.0800, 0.0822, 0.0823, 0.0557, 0.0646, -0.0729, 0.0857,
0.0337, -0.0079, 0.0632, -0.0813, -0.0178, -0.0147, 0.0018, 0.0151,
0.0909, -0.0895, 0.0524, -0.0362, -0.0328, 0.0500, 0.0494, -0.0450,
-0.0204, -0.0412, 0.0766, -0.0161, -0.0584, -0.0680, 0.0278, 0.0007,
-0.0566, 0.0467, 0.0536, -0.0230, 0.0731, 0.0413, -0.0785, -0.0119],
[ 0.0685, -0.0193, 0.0604, 0.0138, 0.0828, 0.0634, -0.0749, 0.0419,
-0.0212, -0.0736, -0.0009, 0.0376, -0.0012, 0.0102, -0.0813, -0.0153,
-0.0003, -0.0116, 0.0483, -0.0689, -0.0361, -0.0136, 0.0256, -0.0330,
0.0639, 0.0103, -0.0673, 0.0759, 0.0751, 0.0812, 0.0010, -0.0302,
0.0461, -0.0861, -0.0432, -0.0070, 0.0077, -0.0529, 0.0748, 0.0328,
-0.0610, 0.0524, 0.0129, 0.0665, 0.0886, -0.0200, 0.0524, 0.0696,
-0.0629, -0.0878, 0.0708, 0.0556, -0.0741, 0.0161, -0.0897, 0.0735,
0.0793, -0.0354, 0.0309, -0.0521, 0.0006, 0.0265, -0.0274, 0.0792,
0.0860, -0.0430, -0.0282, 0.0335, -0.0274, 0.0322, 0.0616, -0.0157,
-0.0142, 0.0187, 0.0102, -0.0078, 0.0554, -0.0854, -0.0591, 0.0875,
-0.0630, -0.0741, -0.0793, 0.0149, 0.0818, -0.0127, -0.0881, 0.0015,
-0.0594, 0.0065, -0.0806, 0.0295, -0.0144, 0.0879, 0.0663, 0.0900,
0.0258, -0.0155, 0.0731, -0.0102, -0.0611, -0.0473, -0.0538, 0.0004,
-0.0415, -0.0457, -0.0481, -0.0759, 0.0621, -0.0188, 0.0160, 0.0484,
-0.0819, -0.0031, 0.0558, 0.0735, 0.0219, -0.0744, -0.0153, -0.0397]])
(bias): Normal:
loc: tensor([0., -0.])
scale: tensor([0.7071, 0.7071])
posterior: Automatic()
prior: Module()
observed: Observed()
tensor: tensor([ 0.0430, -0.0439])
)
(observed): Observed()
)
)
Fit the model¶
Finally we can set up the training loop
optim = torch.optim.Adam(net.parameters())
for i in range(1):
for data, target in loader:
net.observe(classification=target)
borch.sample(net)
net(data)
loss = infer.vi_loss(**borch.pq_to_infer(net), kl_scaling=1 / len(loader))
loss.backward()
optim.step()
optim.zero_grad()
Now we can check the accuracy, Note that one should stop condtioning on the target by setting net.observe(None)
net.observe(None)
tot_acc = 0
with torch.no_grad():
for i, (data, target) in enumerate(loader):
borch.sample(net)
out = net(data)
acc = float((target == out).sum().float() / target.shape[0]) * 100
tot_acc += acc
tot_acc /= i + 1
print(tot_acc)
Out:
64.99999761581421
the accuracy is basically random, this is due to the fact that we are fitting white noise so it to be expected.
But in case you have trouble getting higher accuracy you should consider running for more epochs, setting up an augmentation pipeline (see: the data loading tutorial) and changing your posterior. The posterior can be changed using
net.apply(borch.set_posteriors(borch.posterior.Automatic))
Out:
Net(
(posterior): Automatic()
(prior): Module(
(classification): Categorical:
logits: tensor([[0.4066, 0.9223],
[0.3959, 0.5796],
[0.2345, 1.0376],
[0.4777, 0.7572],
[0.3595, 0.5583],
[0.5373, 0.4655],
[0.3681, 0.9276],
[0.4100, 0.7398],
[0.4180, 0.5060],
[0.4787, 0.8265],
[0.5473, 0.8474],
[0.5303, 0.7285],
[0.4437, 0.7785],
[0.3840, 0.7335],
[0.6166, 0.8975],
[0.1428, 0.6157],
[0.2379, 0.7260],
[0.3817, 0.7180],
[0.5401, 0.6208],
[0.3308, 0.8017]])
posterior: Automatic()
prior: Module()
observed: Observed()
tensor: tensor([])
)
(observed): Observed()
(conv1): Conv2d(
1, 6, kernel_size=(5, 5), stride=(1, 1)
(posterior): Automatic(
(bias): Normal:
posterior: Automatic()
prior: Module()
observed: Observed()
scale: Transform:
tensor([0.4082, 0.4082, 0.4082, 0.4082, 0.4082, 0.4082],
grad_fn=<ExpBackward0>)
loc: Parameter containing:
tensor([-0., 0., 0., -0., 0., 0.], requires_grad=True)
tensor: tensor([-0.1332, -0.5036, 0.1559, -0.2012, -0.6483, 0.5744],
grad_fn=<AddBackward0>)
)
(prior): Module(
(weight): Normal:
loc: tensor([[[[-0., 0., -0., -0., -0.],
[-0., 0., 0., 0., 0.],
[-0., 0., 0., -0., -0.],
[-0., -0., 0., 0., -0.],
[0., 0., -0., 0., -0.]]],
[[[0., -0., 0., 0., 0.],
[0., -0., 0., 0., 0.],
[0., -0., -0., 0., -0.],
[0., -0., 0., 0., -0.],
[-0., 0., -0., 0., -0.]]],
[[[0., -0., -0., -0., 0.],
[-0., -0., -0., 0., 0.],
[0., 0., -0., 0., -0.],
[-0., 0., 0., -0., -0.],
[-0., -0., 0., -0., -0.]]],
[[[0., 0., 0., 0., 0.],
[0., -0., -0., 0., -0.],
[-0., 0., 0., -0., -0.],
[0., -0., 0., -0., 0.],
[-0., 0., -0., -0., 0.]]],
[[[-0., 0., 0., -0., 0.],
[0., -0., -0., 0., 0.],
[0., 0., -0., -0., 0.],
[0., 0., -0., -0., -0.],
[-0., -0., -0., -0., 0.]]],
[[[-0., 0., 0., 0., 0.],
[0., -0., -0., 0., 0.],
[0., 0., -0., 0., 0.],
[0., 0., -0., 0., -0.],
[0., 0., -0., 0., 0.]]]])
scale: tensor([[[[0.2000, 0.2000, 0.2000, 0.2000, 0.2000],
[0.2000, 0.2000, 0.2000, 0.2000, 0.2000],
[0.2000, 0.2000, 0.2000, 0.2000, 0.2000],
[0.2000, 0.2000, 0.2000, 0.2000, 0.2000],
[0.2000, 0.2000, 0.2000, 0.2000, 0.2000]]],
[[[0.2000, 0.2000, 0.2000, 0.2000, 0.2000],
[0.2000, 0.2000, 0.2000, 0.2000, 0.2000],
[0.2000, 0.2000, 0.2000, 0.2000, 0.2000],
[0.2000, 0.2000, 0.2000, 0.2000, 0.2000],
[0.2000, 0.2000, 0.2000, 0.2000, 0.2000]]],
[[[0.2000, 0.2000, 0.2000, 0.2000, 0.2000],
[0.2000, 0.2000, 0.2000, 0.2000, 0.2000],
[0.2000, 0.2000, 0.2000, 0.2000, 0.2000],
[0.2000, 0.2000, 0.2000, 0.2000, 0.2000],
[0.2000, 0.2000, 0.2000, 0.2000, 0.2000]]],
[[[0.2000, 0.2000, 0.2000, 0.2000, 0.2000],
[0.2000, 0.2000, 0.2000, 0.2000, 0.2000],
[0.2000, 0.2000, 0.2000, 0.2000, 0.2000],
[0.2000, 0.2000, 0.2000, 0.2000, 0.2000],
[0.2000, 0.2000, 0.2000, 0.2000, 0.2000]]],
[[[0.2000, 0.2000, 0.2000, 0.2000, 0.2000],
[0.2000, 0.2000, 0.2000, 0.2000, 0.2000],
[0.2000, 0.2000, 0.2000, 0.2000, 0.2000],
[0.2000, 0.2000, 0.2000, 0.2000, 0.2000],
[0.2000, 0.2000, 0.2000, 0.2000, 0.2000]]],
[[[0.2000, 0.2000, 0.2000, 0.2000, 0.2000],
[0.2000, 0.2000, 0.2000, 0.2000, 0.2000],
[0.2000, 0.2000, 0.2000, 0.2000, 0.2000],
[0.2000, 0.2000, 0.2000, 0.2000, 0.2000],
[0.2000, 0.2000, 0.2000, 0.2000, 0.2000]]]])
posterior: Automatic()
prior: Module()
observed: Observed()
tensor: tensor([[[[-0.1927, 0.1399, -0.1354, -0.1729, -0.1509],
[-0.1643, 0.1672, 0.1231, 0.0760, 0.1983],
[-0.1877, 0.1498, 0.1842, -0.1852, -0.0757],
[-0.0739, -0.0365, 0.1345, 0.0660, -0.1410],
[ 0.0395, 0.1924, -0.0889, 0.1817, -0.1028]]],
[[[ 0.0161, -0.0283, 0.0567, 0.0753, 0.1422],
[ 0.1557, -0.0744, 0.1324, 0.1368, 0.1860],
[ 0.1084, -0.0066, -0.1755, 0.0982, -0.1038],
[ 0.0692, -0.1975, 0.1228, 0.1460, -0.1969],
[-0.0408, 0.0335, -0.1200, 0.0135, -0.0343]]],
[[[ 0.1090, -0.1523, -0.0839, -0.1336, 0.1845],
[-0.0021, -0.1854, -0.0692, 0.0818, 0.0268],
[ 0.0554, 0.0253, -0.0801, 0.0925, -0.1053],
[-0.0562, 0.0456, 0.1366, -0.1447, -0.1639],
[-0.0741, -0.1671, 0.1945, -0.1811, -0.1519]]],
[[[ 0.0950, 0.1374, 0.1735, 0.1682, 0.0029],
[ 0.0662, -0.0615, -0.1451, 0.1452, -0.1408],
[-0.1634, 0.0675, 0.1090, -0.1899, -0.0123],
[ 0.0842, -0.0821, 0.1183, -0.0658, 0.1601],
[-0.1688, 0.1212, -0.0177, -0.1106, 0.1500]]],
[[[-0.0711, 0.0887, 0.0110, -0.0312, 0.1235],
[ 0.1727, -0.1303, -0.1418, 0.0785, 0.0870],
[ 0.0772, 0.0361, -0.1826, -0.0933, 0.0847],
[ 0.0795, 0.1156, -0.1951, -0.1324, -0.1288],
[-0.1276, -0.1009, -0.1230, -0.1850, 0.0189]]],
[[[-0.0376, 0.1030, 0.1858, 0.1619, 0.1681],
[ 0.0571, -0.1018, -0.1616, 0.1122, 0.1009],
[ 0.1114, 0.0628, -0.0361, 0.0023, 0.0281],
[ 0.0725, 0.1015, -0.0013, 0.1214, -0.1805],
[ 0.1711, 0.1865, -0.1692, 0.0962, 0.0706]]]])
(bias): Normal:
loc: tensor([-0., 0., 0., -0., 0., 0.])
scale: tensor([0.4082, 0.4082, 0.4082, 0.4082, 0.4082, 0.4082])
posterior: Automatic()
prior: Module()
observed: Observed()
tensor: tensor([-0.1597, 0.1243, 0.0051, -0.0381, 0.0524, 0.0078])
)
(observed): Observed()
)
(conv2): Conv2d(
6, 16, kernel_size=(5, 5), stride=(1, 1)
(posterior): Automatic(
(bias): Normal:
posterior: Automatic()
prior: Module()
observed: Observed()
scale: Transform:
tensor([0.2500, 0.2500, 0.2500, 0.2500, 0.2500, 0.2500, 0.2500, 0.2500, 0.2500,
0.2500, 0.2500, 0.2500, 0.2500, 0.2500, 0.2500, 0.2500],
grad_fn=<ExpBackward0>)
loc: Parameter containing:
tensor([-0., -0., -0., 0., 0., -0., 0., -0., -0., 0., -0., -0., 0., 0., -0., -0.],
requires_grad=True)
tensor: tensor([ 0.3275, -0.0609, 0.1431, -0.5142, 0.4180, -0.0330, -0.0895, 0.5254,
0.2556, 0.1222, -0.0713, 0.4351, 0.2290, -0.3585, -0.0342, 0.0564],
grad_fn=<AddBackward0>)
)
(prior): Module(
(weight): Normal:
loc: tensor([[[[0., -0., 0., -0., -0.],
[0., 0., 0., 0., 0.],
[-0., -0., -0., -0., 0.],
[-0., 0., -0., 0., 0.],
[-0., 0., 0., -0., 0.]],
[[-0., 0., -0., 0., 0.],
[0., -0., 0., 0., 0.],
[-0., 0., 0., 0., 0.],
[-0., -0., 0., 0., -0.],
[0., 0., 0., 0., -0.]],
[[-0., 0., 0., -0., 0.],
[-0., -0., 0., -0., -0.],
[0., 0., 0., 0., 0.],
[0., 0., 0., 0., 0.],
[0., -0., 0., 0., -0.]],
[[-0., -0., 0., 0., 0.],
[0., 0., 0., 0., -0.],
[-0., -0., -0., 0., 0.],
[0., -0., 0., 0., -0.],
[0., 0., 0., 0., 0.]],
[[-0., -0., -0., -0., -0.],
[-0., 0., 0., 0., 0.],
[-0., 0., 0., 0., 0.],
[-0., -0., -0., -0., 0.],
[0., -0., 0., 0., -0.]],
[[0., 0., -0., 0., 0.],
[-0., -0., 0., -0., -0.],
[0., 0., 0., -0., 0.],
[0., -0., -0., -0., 0.],
[0., -0., -0., -0., 0.]]],
[[[-0., 0., -0., 0., 0.],
[-0., -0., -0., 0., -0.],
[-0., -0., -0., -0., 0.],
[-0., -0., 0., -0., 0.],
[-0., 0., -0., -0., -0.]],
[[0., -0., -0., -0., -0.],
[0., -0., -0., 0., 0.],
[0., -0., 0., 0., 0.],
[0., -0., -0., 0., -0.],
[-0., 0., -0., -0., -0.]],
[[-0., -0., -0., 0., -0.],
[-0., 0., 0., 0., 0.],
[-0., 0., -0., 0., -0.],
[-0., 0., -0., 0., 0.],
[-0., -0., 0., -0., 0.]],
[[-0., -0., 0., 0., 0.],
[-0., 0., -0., 0., 0.],
[-0., 0., -0., 0., 0.],
[-0., -0., 0., -0., -0.],
[0., 0., -0., 0., 0.]],
[[0., 0., -0., -0., 0.],
[-0., -0., 0., 0., 0.],
[-0., -0., 0., 0., 0.],
[-0., -0., -0., 0., 0.],
[0., 0., 0., -0., -0.]],
[[0., -0., 0., 0., 0.],
[-0., -0., 0., 0., -0.],
[-0., -0., -0., 0., 0.],
[-0., 0., 0., 0., -0.],
[-0., 0., -0., -0., -0.]]],
[[[0., -0., 0., 0., 0.],
[0., -0., -0., -0., -0.],
[-0., -0., 0., -0., 0.],
[0., -0., -0., 0., -0.],
[0., 0., -0., 0., 0.]],
[[0., -0., 0., -0., -0.],
[-0., -0., -0., 0., -0.],
[0., 0., -0., 0., -0.],
[-0., 0., 0., 0., 0.],
[0., 0., 0., 0., -0.]],
[[0., -0., 0., -0., 0.],
[-0., -0., -0., 0., -0.],
[0., -0., -0., -0., -0.],
[-0., -0., 0., 0., 0.],
[-0., 0., 0., 0., -0.]],
[[-0., 0., -0., -0., 0.],
[-0., -0., -0., -0., -0.],
[-0., 0., 0., 0., -0.],
[0., 0., 0., -0., -0.],
[0., 0., -0., -0., 0.]],
[[0., 0., 0., -0., 0.],
[-0., 0., -0., -0., -0.],
[0., -0., 0., -0., 0.],
[0., 0., 0., 0., -0.],
[-0., 0., -0., 0., -0.]],
[[-0., 0., 0., 0., 0.],
[-0., -0., 0., 0., -0.],
[0., 0., -0., 0., -0.],
[0., 0., -0., -0., -0.],
[-0., 0., -0., -0., -0.]]],
...,
[[[0., 0., -0., 0., -0.],
[-0., 0., 0., 0., -0.],
[-0., 0., -0., 0., -0.],
[0., 0., 0., 0., 0.],
[0., 0., -0., 0., 0.]],
[[0., 0., -0., 0., 0.],
[-0., -0., -0., -0., -0.],
[0., 0., 0., -0., -0.],
[-0., 0., 0., 0., -0.],
[0., 0., -0., 0., 0.]],
[[0., -0., -0., 0., -0.],
[-0., 0., -0., 0., -0.],
[0., 0., -0., 0., -0.],
[0., 0., 0., 0., -0.],
[-0., -0., 0., -0., 0.]],
[[0., -0., 0., 0., -0.],
[0., -0., -0., 0., -0.],
[-0., -0., 0., -0., -0.],
[0., 0., -0., 0., -0.],
[-0., 0., 0., -0., -0.]],
[[0., 0., -0., 0., 0.],
[0., 0., -0., -0., -0.],
[-0., -0., -0., -0., 0.],
[-0., 0., 0., 0., 0.],
[-0., -0., 0., -0., -0.]],
[[-0., 0., -0., -0., 0.],
[0., 0., -0., 0., 0.],
[-0., -0., 0., 0., -0.],
[-0., -0., -0., 0., 0.],
[-0., 0., -0., 0., -0.]]],
[[[0., -0., -0., 0., -0.],
[0., 0., 0., -0., 0.],
[-0., 0., 0., -0., -0.],
[0., 0., -0., -0., 0.],
[0., -0., 0., -0., 0.]],
[[-0., 0., 0., 0., -0.],
[-0., 0., -0., 0., 0.],
[-0., 0., -0., -0., 0.],
[-0., 0., -0., -0., 0.],
[0., -0., -0., 0., 0.]],
[[0., 0., -0., 0., -0.],
[-0., -0., 0., 0., -0.],
[0., 0., 0., -0., -0.],
[-0., 0., -0., 0., 0.],
[-0., 0., 0., 0., 0.]],
[[0., 0., 0., 0., -0.],
[0., 0., -0., 0., -0.],
[-0., -0., -0., -0., -0.],
[0., -0., 0., 0., -0.],
[-0., -0., -0., -0., 0.]],
[[-0., -0., -0., -0., -0.],
[0., 0., -0., -0., -0.],
[-0., -0., 0., 0., 0.],
[-0., 0., -0., -0., -0.],
[0., -0., 0., -0., -0.]],
[[0., 0., -0., 0., -0.],
[-0., -0., -0., -0., -0.],
[0., 0., 0., -0., -0.],
[0., -0., -0., -0., 0.],
[-0., -0., 0., 0., -0.]]],
[[[-0., 0., 0., 0., -0.],
[0., -0., -0., -0., 0.],
[-0., -0., 0., -0., -0.],
[-0., -0., -0., 0., 0.],
[0., 0., 0., -0., 0.]],
[[0., 0., 0., 0., -0.],
[-0., 0., 0., -0., -0.],
[-0., 0., -0., -0., 0.],
[0., -0., 0., 0., -0.],
[-0., -0., -0., 0., 0.]],
[[0., -0., 0., -0., 0.],
[-0., 0., 0., 0., -0.],
[-0., -0., -0., -0., 0.],
[-0., 0., 0., -0., 0.],
[-0., 0., 0., 0., -0.]],
[[-0., -0., -0., 0., 0.],
[-0., -0., 0., 0., -0.],
[0., -0., -0., -0., -0.],
[0., 0., 0., 0., 0.],
[-0., 0., -0., 0., -0.]],
[[-0., -0., 0., -0., 0.],
[0., -0., 0., 0., 0.],
[0., -0., -0., -0., 0.],
[-0., -0., -0., -0., 0.],
[-0., 0., 0., -0., 0.]],
[[-0., -0., -0., -0., -0.],
[0., -0., 0., -0., -0.],
[0., 0., 0., -0., 0.],
[0., -0., 0., 0., -0.],
[0., 0., -0., 0., 0.]]]])
scale: tensor([[[[0.0816, 0.0816, 0.0816, 0.0816, 0.0816],
[0.0816, 0.0816, 0.0816, 0.0816, 0.0816],
[0.0816, 0.0816, 0.0816, 0.0816, 0.0816],
[0.0816, 0.0816, 0.0816, 0.0816, 0.0816],
[0.0816, 0.0816, 0.0816, 0.0816, 0.0816]],
[[0.0816, 0.0816, 0.0816, 0.0816, 0.0816],
[0.0816, 0.0816, 0.0816, 0.0816, 0.0816],
[0.0816, 0.0816, 0.0816, 0.0816, 0.0816],
[0.0816, 0.0816, 0.0816, 0.0816, 0.0816],
[0.0816, 0.0816, 0.0816, 0.0816, 0.0816]],
[[0.0816, 0.0816, 0.0816, 0.0816, 0.0816],
[0.0816, 0.0816, 0.0816, 0.0816, 0.0816],
[0.0816, 0.0816, 0.0816, 0.0816, 0.0816],
[0.0816, 0.0816, 0.0816, 0.0816, 0.0816],
[0.0816, 0.0816, 0.0816, 0.0816, 0.0816]],
[[0.0816, 0.0816, 0.0816, 0.0816, 0.0816],
[0.0816, 0.0816, 0.0816, 0.0816, 0.0816],
[0.0816, 0.0816, 0.0816, 0.0816, 0.0816],
[0.0816, 0.0816, 0.0816, 0.0816, 0.0816],
[0.0816, 0.0816, 0.0816, 0.0816, 0.0816]],
[[0.0816, 0.0816, 0.0816, 0.0816, 0.0816],
[0.0816, 0.0816, 0.0816, 0.0816, 0.0816],
[0.0816, 0.0816, 0.0816, 0.0816, 0.0816],
[0.0816, 0.0816, 0.0816, 0.0816, 0.0816],
[0.0816, 0.0816, 0.0816, 0.0816, 0.0816]],
[[0.0816, 0.0816, 0.0816, 0.0816, 0.0816],
[0.0816, 0.0816, 0.0816, 0.0816, 0.0816],
[0.0816, 0.0816, 0.0816, 0.0816, 0.0816],
[0.0816, 0.0816, 0.0816, 0.0816, 0.0816],
[0.0816, 0.0816, 0.0816, 0.0816, 0.0816]]],
[[[0.0816, 0.0816, 0.0816, 0.0816, 0.0816],
[0.0816, 0.0816, 0.0816, 0.0816, 0.0816],
[0.0816, 0.0816, 0.0816, 0.0816, 0.0816],
[0.0816, 0.0816, 0.0816, 0.0816, 0.0816],
[0.0816, 0.0816, 0.0816, 0.0816, 0.0816]],
[[0.0816, 0.0816, 0.0816, 0.0816, 0.0816],
[0.0816, 0.0816, 0.0816, 0.0816, 0.0816],
[0.0816, 0.0816, 0.0816, 0.0816, 0.0816],
[0.0816, 0.0816, 0.0816, 0.0816, 0.0816],
[0.0816, 0.0816, 0.0816, 0.0816, 0.0816]],
[[0.0816, 0.0816, 0.0816, 0.0816, 0.0816],
[0.0816, 0.0816, 0.0816, 0.0816, 0.0816],
[0.0816, 0.0816, 0.0816, 0.0816, 0.0816],
[0.0816, 0.0816, 0.0816, 0.0816, 0.0816],
[0.0816, 0.0816, 0.0816, 0.0816, 0.0816]],
[[0.0816, 0.0816, 0.0816, 0.0816, 0.0816],
[0.0816, 0.0816, 0.0816, 0.0816, 0.0816],
[0.0816, 0.0816, 0.0816, 0.0816, 0.0816],
[0.0816, 0.0816, 0.0816, 0.0816, 0.0816],
[0.0816, 0.0816, 0.0816, 0.0816, 0.0816]],
[[0.0816, 0.0816, 0.0816, 0.0816, 0.0816],
[0.0816, 0.0816, 0.0816, 0.0816, 0.0816],
[0.0816, 0.0816, 0.0816, 0.0816, 0.0816],
[0.0816, 0.0816, 0.0816, 0.0816, 0.0816],
[0.0816, 0.0816, 0.0816, 0.0816, 0.0816]],
[[0.0816, 0.0816, 0.0816, 0.0816, 0.0816],
[0.0816, 0.0816, 0.0816, 0.0816, 0.0816],
[0.0816, 0.0816, 0.0816, 0.0816, 0.0816],
[0.0816, 0.0816, 0.0816, 0.0816, 0.0816],
[0.0816, 0.0816, 0.0816, 0.0816, 0.0816]]],
[[[0.0816, 0.0816, 0.0816, 0.0816, 0.0816],
[0.0816, 0.0816, 0.0816, 0.0816, 0.0816],
[0.0816, 0.0816, 0.0816, 0.0816, 0.0816],
[0.0816, 0.0816, 0.0816, 0.0816, 0.0816],
[0.0816, 0.0816, 0.0816, 0.0816, 0.0816]],
[[0.0816, 0.0816, 0.0816, 0.0816, 0.0816],
[0.0816, 0.0816, 0.0816, 0.0816, 0.0816],
[0.0816, 0.0816, 0.0816, 0.0816, 0.0816],
[0.0816, 0.0816, 0.0816, 0.0816, 0.0816],
[0.0816, 0.0816, 0.0816, 0.0816, 0.0816]],
[[0.0816, 0.0816, 0.0816, 0.0816, 0.0816],
[0.0816, 0.0816, 0.0816, 0.0816, 0.0816],
[0.0816, 0.0816, 0.0816, 0.0816, 0.0816],
[0.0816, 0.0816, 0.0816, 0.0816, 0.0816],
[0.0816, 0.0816, 0.0816, 0.0816, 0.0816]],
[[0.0816, 0.0816, 0.0816, 0.0816, 0.0816],
[0.0816, 0.0816, 0.0816, 0.0816, 0.0816],
[0.0816, 0.0816, 0.0816, 0.0816, 0.0816],
[0.0816, 0.0816, 0.0816, 0.0816, 0.0816],
[0.0816, 0.0816, 0.0816, 0.0816, 0.0816]],
[[0.0816, 0.0816, 0.0816, 0.0816, 0.0816],
[0.0816, 0.0816, 0.0816, 0.0816, 0.0816],
[0.0816, 0.0816, 0.0816, 0.0816, 0.0816],
[0.0816, 0.0816, 0.0816, 0.0816, 0.0816],
[0.0816, 0.0816, 0.0816, 0.0816, 0.0816]],
[[0.0816, 0.0816, 0.0816, 0.0816, 0.0816],
[0.0816, 0.0816, 0.0816, 0.0816, 0.0816],
[0.0816, 0.0816, 0.0816, 0.0816, 0.0816],
[0.0816, 0.0816, 0.0816, 0.0816, 0.0816],
[0.0816, 0.0816, 0.0816, 0.0816, 0.0816]]],
...,
[[[0.0816, 0.0816, 0.0816, 0.0816, 0.0816],
[0.0816, 0.0816, 0.0816, 0.0816, 0.0816],
[0.0816, 0.0816, 0.0816, 0.0816, 0.0816],
[0.0816, 0.0816, 0.0816, 0.0816, 0.0816],
[0.0816, 0.0816, 0.0816, 0.0816, 0.0816]],
[[0.0816, 0.0816, 0.0816, 0.0816, 0.0816],
[0.0816, 0.0816, 0.0816, 0.0816, 0.0816],
[0.0816, 0.0816, 0.0816, 0.0816, 0.0816],
[0.0816, 0.0816, 0.0816, 0.0816, 0.0816],
[0.0816, 0.0816, 0.0816, 0.0816, 0.0816]],
[[0.0816, 0.0816, 0.0816, 0.0816, 0.0816],
[0.0816, 0.0816, 0.0816, 0.0816, 0.0816],
[0.0816, 0.0816, 0.0816, 0.0816, 0.0816],
[0.0816, 0.0816, 0.0816, 0.0816, 0.0816],
[0.0816, 0.0816, 0.0816, 0.0816, 0.0816]],
[[0.0816, 0.0816, 0.0816, 0.0816, 0.0816],
[0.0816, 0.0816, 0.0816, 0.0816, 0.0816],
[0.0816, 0.0816, 0.0816, 0.0816, 0.0816],
[0.0816, 0.0816, 0.0816, 0.0816, 0.0816],
[0.0816, 0.0816, 0.0816, 0.0816, 0.0816]],
[[0.0816, 0.0816, 0.0816, 0.0816, 0.0816],
[0.0816, 0.0816, 0.0816, 0.0816, 0.0816],
[0.0816, 0.0816, 0.0816, 0.0816, 0.0816],
[0.0816, 0.0816, 0.0816, 0.0816, 0.0816],
[0.0816, 0.0816, 0.0816, 0.0816, 0.0816]],
[[0.0816, 0.0816, 0.0816, 0.0816, 0.0816],
[0.0816, 0.0816, 0.0816, 0.0816, 0.0816],
[0.0816, 0.0816, 0.0816, 0.0816, 0.0816],
[0.0816, 0.0816, 0.0816, 0.0816, 0.0816],
[0.0816, 0.0816, 0.0816, 0.0816, 0.0816]]],
[[[0.0816, 0.0816, 0.0816, 0.0816, 0.0816],
[0.0816, 0.0816, 0.0816, 0.0816, 0.0816],
[0.0816, 0.0816, 0.0816, 0.0816, 0.0816],
[0.0816, 0.0816, 0.0816, 0.0816, 0.0816],
[0.0816, 0.0816, 0.0816, 0.0816, 0.0816]],
[[0.0816, 0.0816, 0.0816, 0.0816, 0.0816],
[0.0816, 0.0816, 0.0816, 0.0816, 0.0816],
[0.0816, 0.0816, 0.0816, 0.0816, 0.0816],
[0.0816, 0.0816, 0.0816, 0.0816, 0.0816],
[0.0816, 0.0816, 0.0816, 0.0816, 0.0816]],
[[0.0816, 0.0816, 0.0816, 0.0816, 0.0816],
[0.0816, 0.0816, 0.0816, 0.0816, 0.0816],
[0.0816, 0.0816, 0.0816, 0.0816, 0.0816],
[0.0816, 0.0816, 0.0816, 0.0816, 0.0816],
[0.0816, 0.0816, 0.0816, 0.0816, 0.0816]],
[[0.0816, 0.0816, 0.0816, 0.0816, 0.0816],
[0.0816, 0.0816, 0.0816, 0.0816, 0.0816],
[0.0816, 0.0816, 0.0816, 0.0816, 0.0816],
[0.0816, 0.0816, 0.0816, 0.0816, 0.0816],
[0.0816, 0.0816, 0.0816, 0.0816, 0.0816]],
[[0.0816, 0.0816, 0.0816, 0.0816, 0.0816],
[0.0816, 0.0816, 0.0816, 0.0816, 0.0816],
[0.0816, 0.0816, 0.0816, 0.0816, 0.0816],
[0.0816, 0.0816, 0.0816, 0.0816, 0.0816],
[0.0816, 0.0816, 0.0816, 0.0816, 0.0816]],
[[0.0816, 0.0816, 0.0816, 0.0816, 0.0816],
[0.0816, 0.0816, 0.0816, 0.0816, 0.0816],
[0.0816, 0.0816, 0.0816, 0.0816, 0.0816],
[0.0816, 0.0816, 0.0816, 0.0816, 0.0816],
[0.0816, 0.0816, 0.0816, 0.0816, 0.0816]]],
[[[0.0816, 0.0816, 0.0816, 0.0816, 0.0816],
[0.0816, 0.0816, 0.0816, 0.0816, 0.0816],
[0.0816, 0.0816, 0.0816, 0.0816, 0.0816],
[0.0816, 0.0816, 0.0816, 0.0816, 0.0816],
[0.0816, 0.0816, 0.0816, 0.0816, 0.0816]],
[[0.0816, 0.0816, 0.0816, 0.0816, 0.0816],
[0.0816, 0.0816, 0.0816, 0.0816, 0.0816],
[0.0816, 0.0816, 0.0816, 0.0816, 0.0816],
[0.0816, 0.0816, 0.0816, 0.0816, 0.0816],
[0.0816, 0.0816, 0.0816, 0.0816, 0.0816]],
[[0.0816, 0.0816, 0.0816, 0.0816, 0.0816],
[0.0816, 0.0816, 0.0816, 0.0816, 0.0816],
[0.0816, 0.0816, 0.0816, 0.0816, 0.0816],
[0.0816, 0.0816, 0.0816, 0.0816, 0.0816],
[0.0816, 0.0816, 0.0816, 0.0816, 0.0816]],
[[0.0816, 0.0816, 0.0816, 0.0816, 0.0816],
[0.0816, 0.0816, 0.0816, 0.0816, 0.0816],
[0.0816, 0.0816, 0.0816, 0.0816, 0.0816],
[0.0816, 0.0816, 0.0816, 0.0816, 0.0816],
[0.0816, 0.0816, 0.0816, 0.0816, 0.0816]],
[[0.0816, 0.0816, 0.0816, 0.0816, 0.0816],
[0.0816, 0.0816, 0.0816, 0.0816, 0.0816],
[0.0816, 0.0816, 0.0816, 0.0816, 0.0816],
[0.0816, 0.0816, 0.0816, 0.0816, 0.0816],
[0.0816, 0.0816, 0.0816, 0.0816, 0.0816]],
[[0.0816, 0.0816, 0.0816, 0.0816, 0.0816],
[0.0816, 0.0816, 0.0816, 0.0816, 0.0816],
[0.0816, 0.0816, 0.0816, 0.0816, 0.0816],
[0.0816, 0.0816, 0.0816, 0.0816, 0.0816],
[0.0816, 0.0816, 0.0816, 0.0816, 0.0816]]]])
posterior: Automatic()
prior: Module()
observed: Observed()
tensor: tensor([[[[ 0.0228, -0.0484, 0.0637, -0.0249, -0.0302],
[ 0.0105, 0.0733, 0.0576, 0.0393, 0.0094],
[-0.0590, -0.0747, -0.0506, -0.0495, 0.0402],
[-0.0647, 0.0789, -0.0096, 0.0477, 0.0700],
[-0.0450, 0.0176, 0.0443, -0.0176, 0.0407]],
[[-0.0148, 0.0075, -0.0554, 0.0808, 0.0812],
[ 0.0636, -0.0776, 0.0617, 0.0497, 0.0760],
[-0.0350, 0.0633, 0.0318, 0.0562, 0.0742],
[-0.0229, -0.0075, 0.0338, 0.0709, -0.0382],
[ 0.0655, 0.0249, 0.0556, 0.0329, -0.0734]],
[[-0.0531, 0.0393, 0.0355, -0.0358, 0.0298],
[-0.0772, -0.0365, 0.0397, -0.0093, -0.0571],
[ 0.0087, 0.0256, 0.0568, 0.0123, 0.0098],
[ 0.0070, 0.0767, 0.0580, 0.0573, 0.0128],
[ 0.0476, -0.0090, 0.0190, 0.0597, -0.0108]],
[[-0.0097, -0.0755, 0.0251, 0.0349, 0.0423],
[ 0.0800, 0.0594, 0.0019, 0.0419, -0.0641],
[-0.0501, -0.0639, -0.0783, 0.0276, 0.0425],
[ 0.0793, -0.0161, 0.0031, 0.0700, -0.0077],
[ 0.0717, 0.0564, 0.0130, 0.0126, 0.0055]],
[[-0.0711, -0.0004, -0.0350, -0.0729, -0.0567],
[-0.0351, 0.0119, 0.0163, 0.0542, 0.0558],
[-0.0308, 0.0337, 0.0634, 0.0556, 0.0051],
[-0.0305, -0.0103, -0.0515, -0.0746, 0.0701],
[ 0.0014, -0.0249, 0.0724, 0.0282, -0.0773]],
[[ 0.0647, 0.0735, -0.0460, 0.0321, 0.0060],
[-0.0624, -0.0262, 0.0131, -0.0270, -0.0768],
[ 0.0053, 0.0301, 0.0262, -0.0012, 0.0230],
[ 0.0758, -0.0297, -0.0011, -0.0424, 0.0659],
[ 0.0155, -0.0027, -0.0785, -0.0715, 0.0603]]],
[[[-0.0680, 0.0027, -0.0140, 0.0451, 0.0738],
[-0.0323, -0.0084, -0.0281, 0.0522, -0.0210],
[-0.0525, -0.0399, -0.0072, -0.0577, 0.0280],
[-0.0374, -0.0055, 0.0143, -0.0258, 0.0404],
[-0.0804, 0.0563, -0.0096, -0.0485, -0.0052]],
[[ 0.0772, -0.0417, -0.0619, -0.0375, -0.0217],
[ 0.0312, -0.0186, -0.0225, 0.0433, 0.0327],
[ 0.0751, -0.0210, 0.0098, 0.0275, 0.0606],
[ 0.0308, -0.0484, -0.0063, 0.0799, -0.0184],
[-0.0195, 0.0014, -0.0263, -0.0651, -0.0203]],
[[-0.0441, -0.0229, -0.0547, 0.0813, -0.0195],
[-0.0343, 0.0006, 0.0058, 0.0113, 0.0668],
[-0.0674, 0.0276, -0.0072, 0.0360, -0.0707],
[-0.0468, 0.0240, -0.0658, 0.0561, 0.0396],
[-0.0704, -0.0553, 0.0372, -0.0310, 0.0525]],
[[-0.0081, -0.0563, 0.0125, 0.0304, 0.0486],
[-0.0515, 0.0539, -0.0165, 0.0263, 0.0241],
[-0.0189, 0.0789, -0.0254, 0.0502, 0.0336],
[-0.0288, -0.0308, 0.0397, -0.0238, -0.0365],
[ 0.0195, 0.0625, -0.0496, 0.0805, 0.0320]],
[[ 0.0451, 0.0478, -0.0347, -0.0325, 0.0203],
[-0.0460, -0.0406, 0.0251, 0.0666, 0.0718],
[-0.0109, -0.0145, 0.0313, 0.0103, 0.0525],
[-0.0299, -0.0280, -0.0036, 0.0186, 0.0350],
[ 0.0073, 0.0463, 0.0292, -0.0732, -0.0485]],
[[ 0.0059, -0.0394, 0.0745, 0.0484, 0.0053],
[-0.0341, -0.0371, 0.0608, 0.0058, -0.0537],
[-0.0605, -0.0791, -0.0253, 0.0762, 0.0762],
[-0.0059, 0.0153, 0.0580, 0.0198, -0.0173],
[-0.0174, 0.0578, -0.0368, -0.0736, -0.0631]]],
[[[ 0.0033, -0.0336, 0.0105, 0.0191, 0.0074],
[ 0.0714, -0.0613, -0.0475, -0.0452, -0.0098],
[-0.0617, -0.0515, 0.0353, -0.0375, 0.0360],
[ 0.0542, -0.0326, -0.0236, 0.0674, -0.0347],
[ 0.0394, 0.0746, -0.0467, 0.0522, 0.0424]],
[[ 0.0503, -0.0693, 0.0626, -0.0424, -0.0488],
[-0.0617, -0.0619, -0.0342, 0.0127, -0.0253],
[ 0.0319, 0.0270, -0.0089, 0.0175, -0.0308],
[-0.0551, 0.0107, 0.0212, 0.0026, 0.0019],
[ 0.0115, 0.0343, 0.0572, 0.0811, -0.0354]],
[[ 0.0632, -0.0197, 0.0072, -0.0372, 0.0581],
[-0.0492, -0.0602, -0.0201, 0.0465, -0.0482],
[ 0.0403, -0.0086, -0.0292, -0.0711, -0.0294],
[-0.0228, -0.0102, 0.0450, 0.0804, 0.0032],
[-0.0009, 0.0815, 0.0113, 0.0315, -0.0110]],
[[-0.0627, 0.0313, -0.0443, -0.0353, 0.0797],
[-0.0039, -0.0053, -0.0678, -0.0335, -0.0764],
[-0.0202, 0.0296, 0.0548, 0.0185, -0.0028],
[ 0.0044, 0.0496, 0.0734, -0.0418, -0.0792],
[ 0.0025, 0.0171, -0.0681, -0.0224, 0.0077]],
[[ 0.0424, 0.0476, 0.0079, -0.0167, 0.0523],
[-0.0335, 0.0003, -0.0560, -0.0448, -0.0173],
[ 0.0426, -0.0002, 0.0133, -0.0705, 0.0085],
[ 0.0724, 0.0664, 0.0455, 0.0461, -0.0607],
[-0.0546, 0.0696, -0.0658, 0.0364, -0.0562]],
[[-0.0390, 0.0704, 0.0786, 0.0033, 0.0295],
[-0.0145, -0.0198, 0.0638, 0.0332, -0.0197],
[ 0.0663, 0.0625, -0.0292, 0.0036, -0.0408],
[ 0.0201, 0.0786, -0.0615, -0.0266, -0.0627],
[-0.0323, 0.0509, -0.0516, -0.0624, -0.0025]]],
...,
[[[ 0.0777, 0.0133, -0.0427, 0.0800, -0.0246],
[-0.0430, 0.0370, 0.0550, 0.0002, -0.0453],
[-0.0782, 0.0747, -0.0467, 0.0130, -0.0217],
[ 0.0417, 0.0371, 0.0066, 0.0761, 0.0344],
[ 0.0590, 0.0188, -0.0681, 0.0579, 0.0579]],
[[ 0.0803, 0.0284, -0.0485, 0.0127, 0.0668],
[-0.0049, -0.0313, -0.0693, -0.0646, -0.0111],
[ 0.0460, 0.0529, 0.0564, -0.0790, -0.0583],
[-0.0102, 0.0005, 0.0437, 0.0308, -0.0437],
[ 0.0205, 0.0394, -0.0644, 0.0770, 0.0368]],
[[ 0.0157, -0.0298, -0.0073, 0.0653, -0.0469],
[-0.0527, 0.0319, -0.0647, 0.0084, -0.0241],
[ 0.0004, 0.0514, -0.0025, 0.0630, -0.0294],
[ 0.0689, 0.0191, 0.0546, 0.0365, -0.0539],
[-0.0045, -0.0745, 0.0544, -0.0686, 0.0816]],
[[ 0.0061, -0.0117, 0.0211, 0.0131, -0.0410],
[ 0.0168, -0.0675, -0.0497, 0.0285, -0.0718],
[-0.0401, -0.0345, 0.0003, -0.0376, -0.0554],
[ 0.0065, 0.0611, -0.0163, 0.0418, -0.0605],
[-0.0081, 0.0791, 0.0381, -0.0200, -0.0402]],
[[ 0.0350, 0.0182, -0.0750, 0.0471, 0.0045],
[ 0.0441, 0.0266, -0.0399, -0.0425, -0.0318],
[-0.0428, -0.0136, -0.0466, -0.0151, 0.0100],
[-0.0644, 0.0688, 0.0709, 0.0026, 0.0585],
[-0.0269, -0.0130, 0.0715, -0.0413, -0.0011]],
[[-0.0537, 0.0477, -0.0147, -0.0553, 0.0459],
[ 0.0619, 0.0005, -0.0194, 0.0200, 0.0296],
[-0.0205, -0.0382, 0.0277, 0.0322, -0.0201],
[-0.0251, -0.0238, -0.0758, 0.0534, 0.0487],
[-0.0023, 0.0328, -0.0237, 0.0328, -0.0399]]],
[[[ 0.0501, -0.0288, -0.0173, 0.0392, -0.0050],
[ 0.0156, 0.0807, 0.0459, -0.0056, 0.0346],
[-0.0486, 0.0271, 0.0530, -0.0753, -0.0678],
[ 0.0070, 0.0007, -0.0208, -0.0359, 0.0587],
[ 0.0728, -0.0304, 0.0334, -0.0355, 0.0323]],
[[-0.0734, 0.0403, 0.0763, 0.0570, -0.0033],
[-0.0664, 0.0105, -0.0366, 0.0042, 0.0546],
[-0.0745, 0.0639, -0.0417, -0.0033, 0.0495],
[-0.0550, 0.0292, -0.0198, -0.0241, 0.0367],
[ 0.0107, -0.0213, -0.0517, 0.0664, 0.0738]],
[[ 0.0123, 0.0034, -0.0046, 0.0620, -0.0356],
[-0.0580, -0.0810, 0.0235, 0.0637, -0.0323],
[ 0.0501, 0.0611, 0.0696, -0.0725, -0.0392],
[-0.0327, 0.0616, -0.0098, 0.0648, 0.0049],
[-0.0644, 0.0434, 0.0465, 0.0378, 0.0250]],
[[ 0.0816, 0.0159, 0.0255, 0.0219, -0.0652],
[ 0.0783, 0.0748, -0.0595, 0.0515, -0.0486],
[-0.0709, -0.0491, -0.0587, -0.0085, -0.0437],
[ 0.0395, -0.0117, 0.0683, 0.0806, -0.0066],
[-0.0332, -0.0257, -0.0023, -0.0359, 0.0064]],
[[-0.0328, -0.0616, -0.0107, -0.0231, -0.0393],
[ 0.0030, 0.0048, -0.0813, -0.0253, -0.0723],
[-0.0680, -0.0350, 0.0409, 0.0464, 0.0235],
[-0.0085, 0.0688, -0.0767, -0.0011, -0.0570],
[ 0.0553, -0.0721, 0.0039, -0.0811, -0.0608]],
[[ 0.0617, 0.0303, -0.0521, 0.0155, -0.0364],
[-0.0589, -0.0223, -0.0112, -0.0599, -0.0590],
[ 0.0729, 0.0326, 0.0761, -0.0415, -0.0048],
[ 0.0036, -0.0197, -0.0393, -0.0060, 0.0785],
[-0.0679, -0.0750, 0.0671, 0.0385, -0.0260]]],
[[[-0.0456, 0.0197, 0.0548, 0.0420, -0.0569],
[ 0.0518, -0.0172, -0.0758, -0.0328, 0.0196],
[-0.0712, -0.0446, 0.0593, -0.0403, -0.0250],
[-0.0142, -0.0058, -0.0283, 0.0783, 0.0075],
[ 0.0755, 0.0161, 0.0319, -0.0562, 0.0378]],
[[ 0.0572, 0.0773, 0.0243, 0.0638, -0.0472],
[-0.0081, 0.0225, 0.0298, -0.0442, -0.0075],
[-0.0814, 0.0369, -0.0680, -0.0471, 0.0187],
[ 0.0290, -0.0338, 0.0786, 0.0685, -0.0263],
[-0.0453, -0.0716, -0.0462, 0.0556, 0.0159]],
[[ 0.0359, -0.0511, 0.0707, -0.0696, 0.0407],
[-0.0717, 0.0521, 0.0813, 0.0335, -0.0515],
[-0.0295, -0.0124, -0.0406, -0.0247, 0.0162],
[-0.0252, 0.0105, 0.0624, -0.0701, 0.0153],
[-0.0490, 0.0815, 0.0331, 0.0130, -0.0174]],
[[-0.0801, -0.0390, -0.0246, 0.0187, 0.0752],
[-0.0654, -0.0242, 0.0666, 0.0303, -0.0114],
[ 0.0783, -0.0565, -0.0200, -0.0462, -0.0119],
[ 0.0788, 0.0656, 0.0623, 0.0350, 0.0254],
[-0.0227, 0.0380, -0.0172, 0.0293, -0.0065]],
[[-0.0086, -0.0572, 0.0217, -0.0286, 0.0476],
[ 0.0695, -0.0679, 0.0714, 0.0371, 0.0638],
[ 0.0099, -0.0652, -0.0545, -0.0068, 0.0805],
[-0.0506, -0.0737, -0.0110, -0.0198, 0.0047],
[-0.0288, 0.0730, 0.0794, -0.0033, 0.0242]],
[[-0.0616, -0.0632, -0.0110, -0.0658, -0.0470],
[ 0.0425, -0.0136, 0.0665, -0.0201, -0.0727],
[ 0.0189, 0.0189, 0.0641, -0.0384, 0.0180],
[ 0.0002, -0.0737, 0.0365, 0.0311, -0.0378],
[ 0.0789, 0.0037, -0.0582, 0.0148, 0.0323]]]])
(bias): Normal:
loc: tensor([-0., -0., -0., 0., 0., -0., 0., -0., -0., 0., -0., -0., 0., 0., -0., -0.])
scale: tensor([0.2500, 0.2500, 0.2500, 0.2500, 0.2500, 0.2500, 0.2500, 0.2500, 0.2500,
0.2500, 0.2500, 0.2500, 0.2500, 0.2500, 0.2500, 0.2500])
posterior: Automatic()
prior: Module()
observed: Observed()
tensor: tensor([-3.9858e-02, -6.2551e-02, -1.4050e-02, 5.5231e-02, 1.7201e-02,
-2.6817e-02, 2.8532e-02, -2.6857e-02, -2.8279e-02, 9.3339e-04,
-5.8307e-02, -6.2624e-02, 7.1998e-05, 1.1212e-02, -2.0352e-02,
-5.8423e-02])
)
(observed): Observed()
)
(fc1): Linear(
in_features=400, out_features=120, bias=True
(posterior): Automatic(
(bias): Normal:
posterior: Automatic()
prior: Module()
observed: Observed()
scale: Transform:
tensor([0.0913, 0.0913, 0.0913, 0.0913, 0.0913, 0.0913, 0.0913, 0.0913, 0.0913,
0.0913, 0.0913, 0.0913, 0.0913, 0.0913, 0.0913, 0.0913, 0.0913, 0.0913,
0.0913, 0.0913, 0.0913, 0.0913, 0.0913, 0.0913, 0.0913, 0.0913, 0.0913,
0.0913, 0.0913, 0.0913, 0.0913, 0.0913, 0.0913, 0.0913, 0.0913, 0.0913,
0.0913, 0.0913, 0.0913, 0.0913, 0.0913, 0.0913, 0.0913, 0.0913, 0.0913,
0.0913, 0.0913, 0.0913, 0.0913, 0.0913, 0.0913, 0.0913, 0.0913, 0.0913,
0.0913, 0.0913, 0.0913, 0.0913, 0.0913, 0.0913, 0.0913, 0.0913, 0.0913,
0.0913, 0.0913, 0.0913, 0.0913, 0.0913, 0.0913, 0.0913, 0.0913, 0.0913,
0.0913, 0.0913, 0.0913, 0.0913, 0.0913, 0.0913, 0.0913, 0.0913, 0.0913,
0.0913, 0.0913, 0.0913, 0.0913, 0.0913, 0.0913, 0.0913, 0.0913, 0.0913,
0.0913, 0.0913, 0.0913, 0.0913, 0.0913, 0.0913, 0.0913, 0.0913, 0.0913,
0.0913, 0.0913, 0.0913, 0.0913, 0.0913, 0.0913, 0.0913, 0.0913, 0.0913,
0.0913, 0.0913, 0.0913, 0.0913, 0.0913, 0.0913, 0.0913, 0.0913, 0.0913,
0.0913, 0.0913, 0.0913], grad_fn=<ExpBackward0>)
loc: Parameter containing:
tensor([0., -0., -0., 0., 0., 0., 0., 0., -0., -0., -0., -0., -0., 0., -0., 0., 0., -0., -0., -0., 0., -0., 0., 0.,
-0., -0., -0., -0., 0., -0., 0., -0., -0., -0., -0., 0., -0., -0., -0., -0., -0., -0., 0., 0., -0., 0., -0., 0.,
0., 0., -0., -0., -0., 0., -0., 0., 0., 0., -0., -0., 0., -0., 0., 0., -0., -0., -0., 0., -0., -0., -0., -0.,
-0., -0., -0., -0., 0., 0., -0., -0., -0., 0., -0., 0., 0., 0., 0., -0., 0., -0., 0., 0., 0., 0., 0., -0.,
-0., 0., 0., -0., -0., 0., 0., -0., -0., 0., -0., -0., -0., 0., 0., -0., 0., 0., 0., -0., 0., 0., -0., 0.],
requires_grad=True)
tensor: tensor([-0.0551, -0.0427, -0.0985, -0.0071, 0.0669, -0.0145, -0.1063, 0.0711,
0.0262, -0.0387, -0.0015, -0.0757, 0.0055, -0.0922, -0.1753, -0.0271,
-0.0179, 0.1418, 0.0163, -0.0432, -0.0603, 0.0581, 0.0135, -0.1230,
0.0199, 0.0857, 0.1111, -0.0646, -0.0754, 0.0303, 0.0280, 0.0519,
0.1128, -0.1034, 0.0422, -0.0920, -0.0389, 0.0287, -0.1931, 0.0913,
-0.1261, 0.0266, 0.1357, -0.0125, 0.0096, 0.0660, -0.0504, 0.0232,
0.1478, 0.0955, -0.0457, -0.2297, 0.0295, -0.0228, 0.0511, 0.0642,
-0.2107, 0.0015, 0.1223, 0.0589, 0.0315, 0.0577, -0.0832, -0.1663,
-0.0293, -0.0221, 0.0139, -0.0254, 0.1410, -0.0805, 0.0196, 0.1361,
0.1635, 0.0949, 0.0235, 0.0668, 0.0409, -0.0313, -0.1028, 0.0487,
0.2261, -0.0205, 0.0892, 0.1026, -0.0642, 0.0138, 0.1022, -0.0512,
0.1710, 0.1139, 0.1485, 0.0546, -0.0279, 0.0331, -0.0193, 0.0144,
0.0365, -0.0167, -0.0164, 0.1642, -0.0179, -0.0689, 0.0483, 0.1306,
-0.0528, 0.0161, 0.0134, 0.1056, -0.0197, -0.0272, 0.0707, 0.2466,
0.1049, -0.0313, -0.0176, 0.2206, 0.0065, -0.1687, 0.0019, 0.0558],
grad_fn=<AddBackward0>)
)
(prior): Module(
(weight): Normal:
loc: tensor([[0., -0., 0., ..., -0., -0., -0.],
[0., 0., 0., ..., 0., -0., 0.],
[0., 0., 0., ..., 0., -0., -0.],
...,
[-0., -0., 0., ..., -0., -0., 0.],
[0., 0., 0., ..., 0., 0., 0.],
[-0., -0., 0., ..., -0., 0., 0.]])
scale: tensor([[0.0500, 0.0500, 0.0500, ..., 0.0500, 0.0500, 0.0500],
[0.0500, 0.0500, 0.0500, ..., 0.0500, 0.0500, 0.0500],
[0.0500, 0.0500, 0.0500, ..., 0.0500, 0.0500, 0.0500],
...,
[0.0500, 0.0500, 0.0500, ..., 0.0500, 0.0500, 0.0500],
[0.0500, 0.0500, 0.0500, ..., 0.0500, 0.0500, 0.0500],
[0.0500, 0.0500, 0.0500, ..., 0.0500, 0.0500, 0.0500]])
posterior: Automatic()
prior: Module()
observed: Observed()
tensor: tensor([[ 0.0337, -0.0484, 0.0253, ..., -0.0286, -0.0485, -0.0091],
[ 0.0469, 0.0292, 0.0258, ..., 0.0047, -0.0409, 0.0409],
[ 0.0367, 0.0313, 0.0040, ..., 0.0234, -0.0487, -0.0428],
...,
[-0.0139, -0.0203, 0.0175, ..., -0.0324, -0.0387, 0.0258],
[ 0.0096, 0.0293, 0.0120, ..., 0.0264, 0.0297, 0.0347],
[-0.0126, -0.0436, 0.0311, ..., -0.0195, 0.0352, 0.0191]])
(bias): Normal:
loc: tensor([0., -0., -0., 0., 0., 0., 0., 0., -0., -0., -0., -0., -0., 0., -0., 0., 0., -0., -0., -0., 0., -0., 0., 0.,
-0., -0., -0., -0., 0., -0., 0., -0., -0., -0., -0., 0., -0., -0., -0., -0., -0., -0., 0., 0., -0., 0., -0., 0.,
0., 0., -0., -0., -0., 0., -0., 0., 0., 0., -0., -0., 0., -0., 0., 0., -0., -0., -0., 0., -0., -0., -0., -0.,
-0., -0., -0., -0., 0., 0., -0., -0., -0., 0., -0., 0., 0., 0., 0., -0., 0., -0., 0., 0., 0., 0., 0., -0.,
-0., 0., 0., -0., -0., 0., 0., -0., -0., 0., -0., -0., -0., 0., 0., -0., 0., 0., 0., -0., 0., 0., -0., 0.])
scale: tensor([0.0913, 0.0913, 0.0913, 0.0913, 0.0913, 0.0913, 0.0913, 0.0913, 0.0913,
0.0913, 0.0913, 0.0913, 0.0913, 0.0913, 0.0913, 0.0913, 0.0913, 0.0913,
0.0913, 0.0913, 0.0913, 0.0913, 0.0913, 0.0913, 0.0913, 0.0913, 0.0913,
0.0913, 0.0913, 0.0913, 0.0913, 0.0913, 0.0913, 0.0913, 0.0913, 0.0913,
0.0913, 0.0913, 0.0913, 0.0913, 0.0913, 0.0913, 0.0913, 0.0913, 0.0913,
0.0913, 0.0913, 0.0913, 0.0913, 0.0913, 0.0913, 0.0913, 0.0913, 0.0913,
0.0913, 0.0913, 0.0913, 0.0913, 0.0913, 0.0913, 0.0913, 0.0913, 0.0913,
0.0913, 0.0913, 0.0913, 0.0913, 0.0913, 0.0913, 0.0913, 0.0913, 0.0913,
0.0913, 0.0913, 0.0913, 0.0913, 0.0913, 0.0913, 0.0913, 0.0913, 0.0913,
0.0913, 0.0913, 0.0913, 0.0913, 0.0913, 0.0913, 0.0913, 0.0913, 0.0913,
0.0913, 0.0913, 0.0913, 0.0913, 0.0913, 0.0913, 0.0913, 0.0913, 0.0913,
0.0913, 0.0913, 0.0913, 0.0913, 0.0913, 0.0913, 0.0913, 0.0913, 0.0913,
0.0913, 0.0913, 0.0913, 0.0913, 0.0913, 0.0913, 0.0913, 0.0913, 0.0913,
0.0913, 0.0913, 0.0913])
posterior: Automatic()
prior: Module()
observed: Observed()
tensor: tensor([ 0.0227, -0.0239, -0.0015, 0.0243, 0.0179, 0.0240, 0.0462, 0.0295,
-0.0409, -0.0290, -0.0125, -0.0128, -0.0085, 0.0326, -0.0219, 0.0423,
0.0489, -0.0179, -0.0246, -0.0301, 0.0485, -0.0077, 0.0185, 0.0481,
-0.0032, -0.0149, -0.0211, -0.0451, 0.0282, -0.0297, 0.0339, -0.0216,
-0.0034, -0.0181, -0.0157, 0.0106, -0.0460, -0.0273, -0.0288, -0.0371,
-0.0023, -0.0038, 0.0074, 0.0185, -0.0443, 0.0172, -0.0478, 0.0460,
0.0363, 0.0282, -0.0387, -0.0109, -0.0231, 0.0013, -0.0196, 0.0272,
0.0049, 0.0265, -0.0020, -0.0435, 0.0185, -0.0403, 0.0289, 0.0379,
-0.0112, -0.0080, -0.0427, 0.0491, -0.0431, -0.0402, -0.0102, -0.0105,
-0.0474, -0.0193, -0.0236, -0.0411, 0.0166, 0.0335, -0.0161, -0.0324,
-0.0196, 0.0304, -0.0400, 0.0024, 0.0160, 0.0193, 0.0080, -0.0252,
0.0398, -0.0498, 0.0386, 0.0138, 0.0152, 0.0196, 0.0355, -0.0123,
-0.0179, 0.0390, 0.0361, -0.0140, -0.0484, 0.0458, 0.0205, -0.0043,
-0.0300, 0.0102, -0.0160, -0.0108, -0.0162, 0.0330, 0.0324, -0.0429,
0.0008, 0.0134, 0.0364, -0.0246, 0.0498, 0.0140, -0.0339, 0.0392])
)
(observed): Observed()
)
(fc2): Linear(
in_features=120, out_features=2, bias=True
(posterior): Automatic(
(bias): Normal:
posterior: Automatic()
prior: Module()
observed: Observed()
scale: Transform:
tensor([0.7071, 0.7071], grad_fn=<ExpBackward0>)
loc: Parameter containing:
tensor([0., -0.], requires_grad=True)
tensor: tensor([-0.8370, 0.1714], grad_fn=<AddBackward0>)
)
(prior): Module(
(weight): Normal:
loc: tensor([[0., -0., -0., -0., 0., 0., 0., -0., -0., 0., 0., 0., -0., 0., 0., -0., -0., 0., 0., -0., 0., 0., 0., 0.,
-0., -0., -0., -0., 0., -0., 0., -0., -0., -0., 0., -0., 0., 0., -0., -0., 0., -0., 0., 0., -0., 0., -0., -0.,
-0., -0., -0., -0., -0., -0., 0., 0., -0., 0., 0., -0., -0., 0., -0., 0., 0., 0., -0., -0., -0., -0., -0., -0.,
-0., 0., -0., -0., -0., 0., -0., 0., -0., 0., 0., 0., 0., 0., -0., 0., 0., -0., 0., -0., -0., -0., 0., 0.,
0., -0., 0., -0., -0., 0., 0., -0., -0., -0., 0., -0., -0., -0., 0., 0., -0., 0., 0., -0., 0., 0., -0., -0.],
[0., -0., 0., 0., 0., 0., -0., 0., -0., -0., -0., 0., -0., 0., -0., -0., -0., -0., 0., -0., -0., -0., 0., -0.,
0., 0., -0., 0., 0., 0., 0., -0., 0., -0., -0., -0., 0., -0., 0., 0., -0., 0., 0., 0., 0., -0., 0., 0.,
-0., -0., 0., 0., -0., 0., -0., 0., 0., -0., 0., -0., 0., 0., -0., 0., 0., -0., -0., 0., -0., 0., 0., -0.,
-0., 0., 0., -0., 0., -0., -0., 0., -0., -0., -0., 0., 0., -0., -0., 0., -0., 0., -0., 0., -0., 0., 0., 0.,
0., -0., 0., -0., -0., -0., -0., 0., -0., -0., -0., -0., 0., -0., 0., 0., -0., -0., 0., 0., 0., -0., -0., -0.]])
scale: tensor([[0.0913, 0.0913, 0.0913, 0.0913, 0.0913, 0.0913, 0.0913, 0.0913, 0.0913,
0.0913, 0.0913, 0.0913, 0.0913, 0.0913, 0.0913, 0.0913, 0.0913, 0.0913,
0.0913, 0.0913, 0.0913, 0.0913, 0.0913, 0.0913, 0.0913, 0.0913, 0.0913,
0.0913, 0.0913, 0.0913, 0.0913, 0.0913, 0.0913, 0.0913, 0.0913, 0.0913,
0.0913, 0.0913, 0.0913, 0.0913, 0.0913, 0.0913, 0.0913, 0.0913, 0.0913,
0.0913, 0.0913, 0.0913, 0.0913, 0.0913, 0.0913, 0.0913, 0.0913, 0.0913,
0.0913, 0.0913, 0.0913, 0.0913, 0.0913, 0.0913, 0.0913, 0.0913, 0.0913,
0.0913, 0.0913, 0.0913, 0.0913, 0.0913, 0.0913, 0.0913, 0.0913, 0.0913,
0.0913, 0.0913, 0.0913, 0.0913, 0.0913, 0.0913, 0.0913, 0.0913, 0.0913,
0.0913, 0.0913, 0.0913, 0.0913, 0.0913, 0.0913, 0.0913, 0.0913, 0.0913,
0.0913, 0.0913, 0.0913, 0.0913, 0.0913, 0.0913, 0.0913, 0.0913, 0.0913,
0.0913, 0.0913, 0.0913, 0.0913, 0.0913, 0.0913, 0.0913, 0.0913, 0.0913,
0.0913, 0.0913, 0.0913, 0.0913, 0.0913, 0.0913, 0.0913, 0.0913, 0.0913,
0.0913, 0.0913, 0.0913],
[0.0913, 0.0913, 0.0913, 0.0913, 0.0913, 0.0913, 0.0913, 0.0913, 0.0913,
0.0913, 0.0913, 0.0913, 0.0913, 0.0913, 0.0913, 0.0913, 0.0913, 0.0913,
0.0913, 0.0913, 0.0913, 0.0913, 0.0913, 0.0913, 0.0913, 0.0913, 0.0913,
0.0913, 0.0913, 0.0913, 0.0913, 0.0913, 0.0913, 0.0913, 0.0913, 0.0913,
0.0913, 0.0913, 0.0913, 0.0913, 0.0913, 0.0913, 0.0913, 0.0913, 0.0913,
0.0913, 0.0913, 0.0913, 0.0913, 0.0913, 0.0913, 0.0913, 0.0913, 0.0913,
0.0913, 0.0913, 0.0913, 0.0913, 0.0913, 0.0913, 0.0913, 0.0913, 0.0913,
0.0913, 0.0913, 0.0913, 0.0913, 0.0913, 0.0913, 0.0913, 0.0913, 0.0913,
0.0913, 0.0913, 0.0913, 0.0913, 0.0913, 0.0913, 0.0913, 0.0913, 0.0913,
0.0913, 0.0913, 0.0913, 0.0913, 0.0913, 0.0913, 0.0913, 0.0913, 0.0913,
0.0913, 0.0913, 0.0913, 0.0913, 0.0913, 0.0913, 0.0913, 0.0913, 0.0913,
0.0913, 0.0913, 0.0913, 0.0913, 0.0913, 0.0913, 0.0913, 0.0913, 0.0913,
0.0913, 0.0913, 0.0913, 0.0913, 0.0913, 0.0913, 0.0913, 0.0913, 0.0913,
0.0913, 0.0913, 0.0913]])
posterior: Automatic()
prior: Module()
observed: Observed()
tensor: tensor([[ 0.0552, -0.0814, -0.0594, -0.0877, 0.0260, 0.0764, 0.0603, -0.0690,
-0.0617, 0.0816, 0.0810, 0.0872, -0.0560, 0.0101, 0.0851, -0.0210,
-0.0042, 0.0030, 0.0784, -0.0529, 0.0262, 0.0469, 0.0635, 0.0703,
-0.0497, -0.0707, -0.0536, -0.0544, 0.0331, -0.0068, 0.0086, -0.0108,
-0.0055, -0.0076, 0.0049, -0.0805, 0.0471, 0.0136, -0.0250, -0.0345,
0.0100, -0.0639, 0.0511, 0.0064, -0.0230, 0.0270, -0.0317, -0.0259,
-0.0307, -0.0506, -0.0617, -0.0262, -0.0688, -0.0900, 0.0328, 0.0166,
-0.0635, 0.0072, 0.0536, -0.0793, -0.0726, 0.0021, -0.0797, 0.0413,
0.0666, 0.0185, -0.0274, -0.0572, -0.0811, -0.0831, -0.0043, -0.0517,
-0.0639, 0.0666, -0.0211, -0.0832, -0.0887, 0.0568, -0.0423, 0.0110,
-0.0736, 0.0800, 0.0822, 0.0823, 0.0557, 0.0646, -0.0729, 0.0857,
0.0337, -0.0079, 0.0632, -0.0813, -0.0178, -0.0147, 0.0018, 0.0151,
0.0909, -0.0895, 0.0524, -0.0362, -0.0328, 0.0500, 0.0494, -0.0450,
-0.0204, -0.0412, 0.0766, -0.0161, -0.0584, -0.0680, 0.0278, 0.0007,
-0.0566, 0.0467, 0.0536, -0.0230, 0.0731, 0.0413, -0.0785, -0.0119],
[ 0.0685, -0.0193, 0.0604, 0.0138, 0.0828, 0.0634, -0.0749, 0.0419,
-0.0212, -0.0736, -0.0009, 0.0376, -0.0012, 0.0102, -0.0813, -0.0153,
-0.0003, -0.0116, 0.0483, -0.0689, -0.0361, -0.0136, 0.0256, -0.0330,
0.0639, 0.0103, -0.0673, 0.0759, 0.0751, 0.0812, 0.0010, -0.0302,
0.0461, -0.0861, -0.0432, -0.0070, 0.0077, -0.0529, 0.0748, 0.0328,
-0.0610, 0.0524, 0.0129, 0.0665, 0.0886, -0.0200, 0.0524, 0.0696,
-0.0629, -0.0878, 0.0708, 0.0556, -0.0741, 0.0161, -0.0897, 0.0735,
0.0793, -0.0354, 0.0309, -0.0521, 0.0006, 0.0265, -0.0274, 0.0792,
0.0860, -0.0430, -0.0282, 0.0335, -0.0274, 0.0322, 0.0616, -0.0157,
-0.0142, 0.0187, 0.0102, -0.0078, 0.0554, -0.0854, -0.0591, 0.0875,
-0.0630, -0.0741, -0.0793, 0.0149, 0.0818, -0.0127, -0.0881, 0.0015,
-0.0594, 0.0065, -0.0806, 0.0295, -0.0144, 0.0879, 0.0663, 0.0900,
0.0258, -0.0155, 0.0731, -0.0102, -0.0611, -0.0473, -0.0538, 0.0004,
-0.0415, -0.0457, -0.0481, -0.0759, 0.0621, -0.0188, 0.0160, 0.0484,
-0.0819, -0.0031, 0.0558, 0.0735, 0.0219, -0.0744, -0.0153, -0.0397]])
(bias): Normal:
loc: tensor([0., -0.])
scale: tensor([0.7071, 0.7071])
posterior: Automatic()
prior: Module()
observed: Observed()
tensor: tensor([ 0.0430, -0.0439])
)
(observed): Observed()
)
)
One can also set the posterior when one creates the module
nn.Linear(10, 10, posterior=borch.posterior.Normal(log_scale=-3))
Out:
Linear(
in_features=10, out_features=10, bias=True
(posterior): Normal(
(weight): Normal:
posterior: Automatic()
prior: Module()
observed: Observed()
scale: Transform:
tensor([[0.0498, 0.0498, 0.0498, 0.0498, 0.0498, 0.0498, 0.0498, 0.0498, 0.0498,
0.0498],
[0.0498, 0.0498, 0.0498, 0.0498, 0.0498, 0.0498, 0.0498, 0.0498, 0.0498,
0.0498],
[0.0498, 0.0498, 0.0498, 0.0498, 0.0498, 0.0498, 0.0498, 0.0498, 0.0498,
0.0498],
[0.0498, 0.0498, 0.0498, 0.0498, 0.0498, 0.0498, 0.0498, 0.0498, 0.0498,
0.0498],
[0.0498, 0.0498, 0.0498, 0.0498, 0.0498, 0.0498, 0.0498, 0.0498, 0.0498,
0.0498],
[0.0498, 0.0498, 0.0498, 0.0498, 0.0498, 0.0498, 0.0498, 0.0498, 0.0498,
0.0498],
[0.0498, 0.0498, 0.0498, 0.0498, 0.0498, 0.0498, 0.0498, 0.0498, 0.0498,
0.0498],
[0.0498, 0.0498, 0.0498, 0.0498, 0.0498, 0.0498, 0.0498, 0.0498, 0.0498,
0.0498],
[0.0498, 0.0498, 0.0498, 0.0498, 0.0498, 0.0498, 0.0498, 0.0498, 0.0498,
0.0498],
[0.0498, 0.0498, 0.0498, 0.0498, 0.0498, 0.0498, 0.0498, 0.0498, 0.0498,
0.0498]], grad_fn=<ExpBackward0>)
loc: Parameter containing:
tensor([[0., -0., 0., 0., 0., 0., -0., -0., -0., -0.],
[0., -0., 0., 0., -0., 0., 0., 0., -0., 0.],
[0., -0., -0., -0., 0., -0., 0., 0., 0., 0.],
[0., -0., 0., 0., 0., -0., 0., 0., 0., -0.],
[-0., 0., 0., 0., 0., -0., -0., 0., -0., -0.],
[0., 0., -0., 0., -0., 0., 0., -0., -0., -0.],
[0., -0., 0., 0., 0., -0., -0., -0., 0., 0.],
[-0., -0., -0., 0., -0., 0., 0., -0., -0., -0.],
[0., 0., 0., 0., 0., -0., -0., 0., 0., 0.],
[-0., -0., 0., 0., 0., -0., 0., -0., -0., -0.]], requires_grad=True)
tensor: tensor([[-1.7528e-02, 4.8843e-02, -4.4339e-03, 2.3555e-03, 9.4015e-02,
-3.7064e-02, -7.4909e-02, 3.4505e-02, -8.3809e-02, -2.4774e-02],
[ 3.2571e-02, -4.0296e-02, -3.0393e-02, -4.4918e-02, 9.7954e-05,
5.2636e-02, -8.8278e-03, 3.9501e-02, 6.7966e-02, -2.8009e-02],
[-4.1282e-02, -6.1663e-02, 4.1791e-02, 1.3551e-02, 1.2286e-04,
-7.7635e-02, -4.3325e-03, 6.4397e-03, 1.2994e-02, 1.6747e-02],
[-2.2902e-02, 1.1335e-02, 3.2603e-02, -9.1162e-02, -1.1745e-02,
1.8527e-02, 1.5697e-02, -2.6060e-02, -5.6732e-02, -1.4607e-02],
[-1.3031e-03, 1.1841e-02, 9.7400e-02, -1.0226e-01, -6.9819e-02,
-5.1301e-03, 1.1550e-02, -3.9957e-03, 1.5816e-02, -1.1080e-02],
[-3.9156e-02, -6.4962e-02, 1.4351e-02, -2.4294e-02, -1.5933e-02,
5.4919e-02, -9.4059e-02, 4.6750e-02, -1.4138e-02, 1.1048e-02],
[-6.7411e-02, 3.9682e-02, -2.6654e-02, -1.5400e-02, 1.3210e-02,
-1.7574e-02, -9.3503e-02, 9.6496e-02, 1.4712e-02, -3.1450e-02],
[ 6.4931e-02, 3.1781e-02, -2.1908e-02, -7.8900e-02, 2.1049e-02,
-5.4553e-02, 4.7148e-02, 3.0550e-03, -8.0232e-02, 7.5974e-02],
[-4.0818e-02, -3.0395e-02, -7.0800e-02, 3.9479e-02, 6.9372e-02,
1.2751e-04, 1.8037e-02, -6.9233e-02, -6.7140e-02, 6.6097e-03],
[ 3.6661e-02, -9.0289e-02, -3.2585e-02, -7.5074e-02, -2.3557e-02,
-4.3972e-02, -6.3312e-03, -9.9403e-03, 3.9945e-02, 2.5346e-02]],
grad_fn=<AddBackward0>)
(bias): Normal:
posterior: Automatic()
prior: Module()
observed: Observed()
scale: Transform:
tensor([0.0498, 0.0498, 0.0498, 0.0498, 0.0498, 0.0498, 0.0498, 0.0498, 0.0498,
0.0498], grad_fn=<ExpBackward0>)
loc: Parameter containing:
tensor([-0., 0., -0., -0., -0., -0., 0., 0., 0., -0.], requires_grad=True)
tensor: tensor([-0.0109, 0.0057, -0.0320, -0.0700, -0.0560, 0.0503, -0.0742, -0.1038,
-0.0172, -0.0039], grad_fn=<AddBackward0>)
)
(prior): Module(
(weight): Normal:
loc: tensor([[0., -0., 0., 0., 0., 0., -0., -0., -0., -0.],
[0., -0., 0., 0., -0., 0., 0., 0., -0., 0.],
[0., -0., -0., -0., 0., -0., 0., 0., 0., 0.],
[0., -0., 0., 0., 0., -0., 0., 0., 0., -0.],
[-0., 0., 0., 0., 0., -0., -0., 0., -0., -0.],
[0., 0., -0., 0., -0., 0., 0., -0., -0., -0.],
[0., -0., 0., 0., 0., -0., -0., -0., 0., 0.],
[-0., -0., -0., 0., -0., 0., 0., -0., -0., -0.],
[0., 0., 0., 0., 0., -0., -0., 0., 0., 0.],
[-0., -0., 0., 0., 0., -0., 0., -0., -0., -0.]])
scale: tensor([[0.3162, 0.3162, 0.3162, 0.3162, 0.3162, 0.3162, 0.3162, 0.3162, 0.3162,
0.3162],
[0.3162, 0.3162, 0.3162, 0.3162, 0.3162, 0.3162, 0.3162, 0.3162, 0.3162,
0.3162],
[0.3162, 0.3162, 0.3162, 0.3162, 0.3162, 0.3162, 0.3162, 0.3162, 0.3162,
0.3162],
[0.3162, 0.3162, 0.3162, 0.3162, 0.3162, 0.3162, 0.3162, 0.3162, 0.3162,
0.3162],
[0.3162, 0.3162, 0.3162, 0.3162, 0.3162, 0.3162, 0.3162, 0.3162, 0.3162,
0.3162],
[0.3162, 0.3162, 0.3162, 0.3162, 0.3162, 0.3162, 0.3162, 0.3162, 0.3162,
0.3162],
[0.3162, 0.3162, 0.3162, 0.3162, 0.3162, 0.3162, 0.3162, 0.3162, 0.3162,
0.3162],
[0.3162, 0.3162, 0.3162, 0.3162, 0.3162, 0.3162, 0.3162, 0.3162, 0.3162,
0.3162],
[0.3162, 0.3162, 0.3162, 0.3162, 0.3162, 0.3162, 0.3162, 0.3162, 0.3162,
0.3162],
[0.3162, 0.3162, 0.3162, 0.3162, 0.3162, 0.3162, 0.3162, 0.3162, 0.3162,
0.3162]])
posterior: Automatic()
prior: Module()
observed: Observed()
tensor: tensor([[ 0.1768, -0.0492, 0.0335, 0.2470, 0.2241, 0.2600, -0.0639, -0.0142,
-0.1177, -0.2890],
[ 0.0065, -0.1177, 0.1118, 0.3136, -0.1325, 0.2423, 0.1595, 0.0413,
-0.2334, 0.1953],
[ 0.0350, -0.2088, -0.3060, -0.2115, 0.0493, -0.0122, 0.1962, 0.0292,
0.1152, 0.1241],
[ 0.0738, -0.0087, 0.2692, 0.3067, 0.0663, -0.1652, 0.0351, 0.0828,
0.0910, -0.1150],
[-0.2252, 0.2810, 0.0442, 0.0448, 0.3011, -0.0721, -0.0053, 0.0346,
-0.2713, -0.1514],
[ 0.1203, 0.0920, -0.1798, 0.2911, -0.1815, 0.0126, 0.0843, -0.2110,
-0.2804, -0.1401],
[ 0.1194, -0.1758, 0.1153, 0.1420, 0.0124, -0.0216, -0.1932, -0.0692,
0.0776, 0.2156],
[-0.0955, -0.1280, -0.2442, 0.2149, -0.1078, 0.2254, 0.0832, -0.2987,
-0.0894, -0.0538],
[ 0.0035, 0.1922, 0.1726, 0.1055, 0.0854, -0.1505, -0.2445, 0.0725,
0.1223, 0.1330],
[-0.1026, -0.2841, 0.2254, 0.2714, 0.1426, -0.2921, 0.0770, -0.0995,
-0.2960, -0.2719]])
(bias): Normal:
loc: tensor([-0., 0., -0., -0., -0., -0., 0., 0., 0., -0.])
scale: tensor([0.3162, 0.3162, 0.3162, 0.3162, 0.3162, 0.3162, 0.3162, 0.3162, 0.3162,
0.3162])
posterior: Automatic()
prior: Module()
observed: Observed()
tensor: tensor([-0.1022, 0.2749, -0.0217, -0.0210, -0.0487, -0.2076, 0.0649, 0.1562,
0.2436, -0.1976])
)
(observed): Observed()
)
See the borch.posterior documentation for other posteriors and what parameters
you can set. Note that all posteriors does not work with all parameters but you can
have different posteriors for the different borch.Module’s in your network.
Exercises¶
Use what you have learned to train an image classifier for MNIST, you should achieve an accuracy larger than 98 %. Note: you can access MNST using
torchvision.datasets.MNIST.Fit the same model architecture with normal torch and compare the likelihood with the borch network, What are the differences and why?
Port the model to CIFAR and see how you can improve the accuracy.
Show how the Categorical distribution is related to the cross entropy loss function that is commonly used in frequentest deep learning.
Total running time of the script: ( 0 minutes 0.226 seconds)