Notes on fastai Book Ch. 17
 A Neural Net from the Foundations
 Building a Neural Net Layer from Scratch
 The Forward and Backward Passes
 Conclusion
 References
import fastbook
fastbook.setup_book()
import inspect
def print_source(obj):
for line in inspect.getsource(obj).split("\n"):
print(line)
A Neural Net from the Foundations
Building a Neural Net Layer from Scratch
Modeling a Neuron
 a neuron receives a given number of inputs and has an internal weight for each of them
 the neuron sums the weighted inputs to produce an output and adds an inner bias
 \(out = \sum_{i=1}^{n}{x_{i}w_{i}+b}\), where \((x_{1},\ldots,x_{n})\) are inputs, \((w_{1},\ldots,w_{n})\) are the weights, and \(b\) is the bias
= sum([x*w for x,w in zip(inputs, weights)]) + bias output
 the output of the neuron is fed to a nonlinear function called an activation function
 the output of the nonlinear activation function is fed as input to another neuron
 Rectified Linear Unit (ReLU) activation function:
def relu(x): return x if x >= 0 else 0
 a deep learning model is build by stacking a lot of neurons in successive layers
 a linear layer: all inputs are linked to each neuron in the layer
 need to compute the dot product for each input and each neuron with a given weight
python sum([x*w for x,w in zip(input,weight)]
The output of a fully connected layer
 \(y_{i,j} = \sum_{k=1}^{n}{x_{i,k}w_{k,j}+b_{j}}\)
= sum([a*b for a,b in zip(x[i,:],w[j,:])]) + b[j] y[i,j]
= x @ w.t() + b y
x
: a matrix containing the inputs with a size ofbatch_size
byn_inputs
w
: a matrix containing the weights for the neurons with a size ofn_neurons
byn_inputs
b
: a vector containing the biases for the neurons with a size ofn_neurons
y
: the output of the fully connected layer of sizebatch_size
byn_neurons
@
: a matrix multiplicationw.t()
: the transpose matrix ofw
Matrix Multiplication from Scratch
 Need three nested loops
 for the row indices
 for the column indices
 for the inner sum
import torch
from torch import tensor
def matmul(a,b):
# Get the number of rows and columns for the two matrices
= a.shape
ar,ac = b.shape
br,bc # The number of columns in the first matrix need to be
# the same as the number of rows in the second matrix
assert ac==br
# Initialize the output matrix
= torch.zeros(ar, bc)
c # For each row in the first matrix
for i in range(ar):
# For each column in the second matrix
for j in range(bc):
# For each column in the first matrix
# Elementwise multiplication
# Sum the products
for k in range(ac): c[i,j] += a[i,k] * b[k,j]
return c
= torch.randn(5,28*28)
m1 = torch.randn(784,10) m2
Using nested forloops
%time t1=matmul(m1, m2)
CPU times: user 329 ms, sys: 0 ns, total: 329 ms
Wall time: 328 ms
%timeit n 20 t1=matmul(m1, m2)
325 ms ± 801 µs per loop (mean ± std. dev. of 7 runs, 20 loops each)
Note: Using loops is extremely inefficient!!! Avoid loops whenever possible.
Using PyTorch’s builtin matrix multiplication operator
 written in C++ to make it fast
 need to vectorize operations on tensors to take advantage of speed of PyTorch
 use elementwise arithmetic and broadcasting
%time t2=m1@m2
CPU times: user 190 µs, sys: 0 ns, total: 190 µs
Wall time: 132 µs
%timeit n 20 t2=m1@m2
The slowest run took 9.84 times longer than the fastest. This could mean that an intermediate result is being cached.
6.42 µs ± 8.4 µs per loop (mean ± std. dev. of 7 runs, 20 loops each)
Elementwise Arithmetic
 addition:
+
 subtraction:

 multiplication:
*
 division:
/
 greater than:
>
 less than:
<
 equal to:
==
Note: Both tensors need to have the same shape to perform elementwise arithmetic.
= tensor([10., 6, 4])
a = tensor([2., 8, 7])
b + b a
tensor([12., 14., 3.])
< b a
tensor([False, True, True])
Reduction Operators
 return tensors with only one element
all
: Tests if all elements evaluate toTrue
.sum
: Returns the sum of all elements in the tensor.mean
: Returns the mean value of all elements in the tensor.
# Check if every element in matrix a is less than the corresponding element in matrix b
< b).all(),
((a # Check if every element in matrix a is equal to the corresponding element in matrix b
==b).all()) (a
(tensor(False), tensor(False))
# Convert tensor to a plain python number or boolean
+ b).mean().item() (a
9.666666984558105
= tensor([[1., 2, 3], [4,5,6], [7,8,9]])
m *m m
tensor([[ 1., 4., 9.],
[16., 25., 36.],
[49., 64., 81.]])
# Attempt to perform elementwise arithmetic on tensors with different shapes
= tensor([[1., 2, 3], [4,5,6]])
n *n m

RuntimeError Traceback (most recent call last)
/tmp/ipykernel_38356/3763285369.py in <module>
1 n = tensor([[1., 2, 3], [4,5,6]])
> 2 m*n
RuntimeError: The size of tensor a (3) must match the size of tensor b (2) at nonsingleton dimension 0
# Replace the innermost forloop with elementwise arithmetic
def matmul(a,b):
= a.shape
ar,ac = b.shape
br,bc assert ac==br
= torch.zeros(ar, bc)
c for i in range(ar):
for j in range(bc): c[i,j] = (a[i] * b[:,j]).sum()
return c
%timeit n 20 t3 = matmul(m1,m2)
488 µs ± 159 µs per loop (mean ± std. dev. of 7 runs, 20 loops each)
Note: Just replacing one of the for loops with PyTorch elementwise arithmetic dramatically improved performance.
Broadcasting
 describes how tensors of different ranks are treated during arithmetic operations
 gives specific rules to codify when shapes are compatible when trying to do an elementwise operation, and how the tensor of the smaller shape is expanded to match the tensor of bigger shape
Broadcasting with a scalar
 the scalar is “virtually” expanded to the same shape as the tensor where every element contains the original scalar value
= tensor([10., 6, 4])
a > 0 a
tensor([ True, True, False])
Note: Broadcasting with a scalar is useful when normalizing a dataset by subtracting the mean and dividing by the standard deviation.
= tensor([[1., 2, 3], [4,5,6], [7,8,9]])
m  5) / 2.73 (m
tensor([[1.4652, 1.0989, 0.7326],
[0.3663, 0.0000, 0.3663],
[ 0.7326, 1.0989, 1.4652]])
Broadcasting a vector to a matrix
 the vector is virtually expanded to the same shape as the tensor, by duplicating the rows/columns as needed
 PyTorch uses the expand_as method to expand the vector to the same size as the higherrank tensor
 creates a new view on the existing vector tensor without allocating new memory
 It is only possible to broadcast a vector of size
n
by a matrix of sizem
byn
.
= tensor([10.,20,30])
c = tensor([[1., 2, 3], [4,5,6], [7,8,9]])
m m.shape,c.shape
(torch.Size([3, 3]), torch.Size([3]))
+ c m
tensor([[11., 22., 33.],
[14., 25., 36.],
[17., 28., 39.]])
c.expand_as(m)
tensor([[10., 20., 30.],
[10., 20., 30.],
[10., 20., 30.]])
help(torch.Tensor.expand_as)
Help on method_descriptor:
expand_as(...)
expand_as(other) > Tensor
Expand this tensor to the same size as :attr:`other`.
``self.expand_as(other)`` is equivalent to ``self.expand(other.size())``.
Please see :meth:`~Tensor.expand` for more information about ``expand``.
Args:
other (:class:`torch.Tensor`): The result tensor has the same size
as :attr:`other`.
help(torch.Tensor.expand)
Help on method_descriptor:
expand(...)
expand(*sizes) > Tensor
Returns a new view of the :attr:`self` tensor with singleton dimensions expanded
to a larger size.
Passing 1 as the size for a dimension means not changing the size of
that dimension.
Tensor can be also expanded to a larger number of dimensions, and the
new ones will be appended at the front. For the new dimensions, the
size cannot be set to 1.
Expanding a tensor does not allocate new memory, but only creates a
new view on the existing tensor where a dimension of size one is
expanded to a larger size by setting the ``stride`` to 0. Any dimension
of size 1 can be expanded to an arbitrary value without allocating new
memory.
Args:
*sizes (torch.Size or int...): the desired expanded size
.. warning::
More than one element of an expanded tensor may refer to a single
memory location. As a result, inplace operations (especially ones that
are vectorized) may result in incorrect behavior. If you need to write
to the tensors, please clone them first.
Example::
>>> x = torch.tensor([[1], [2], [3]])
>>> x.size()
torch.Size([3, 1])
>>> x.expand(3, 4)
tensor([[ 1, 1, 1, 1],
[ 2, 2, 2, 2],
[ 3, 3, 3, 3]])
>>> x.expand(1, 4) # 1 means not changing the size of that dimension
tensor([[ 1, 1, 1, 1],
[ 2, 2, 2, 2],
[ 3, 3, 3, 3]])
Note: Expanding the vector does not increase the amount of data stored.
= c.expand_as(m)
t t.storage()
10.0
20.0
30.0
[torch.FloatStorage of size 3]
help(torch.Tensor.storage)
Help on method_descriptor:
storage(...)
storage() > torch.Storage
Returns the underlying storage.
Note: PyTorch accomplishes this by giving the new dimension a stride of 0
* When PyTorch looks for the next row by adding the stride, it will stay at the same row
t.stride(), t.shape
((0, 1), torch.Size([3, 3]))
+ m c
tensor([[11., 22., 33.],
[14., 25., 36.],
[17., 28., 39.]])
= tensor([10.,20,30])
c = tensor([[1., 2, 3], [4,5,6]])
m +m c
tensor([[11., 22., 33.],
[14., 25., 36.]])
# Attempt to broadcast a vector with an incompatible matrix
= tensor([10.,20])
c = tensor([[1., 2, 3], [4,5,6]])
m +m c

RuntimeError Traceback (most recent call last)
/tmp/ipykernel_38356/3928136702.py in <module>
1 c = tensor([10.,20])
2 m = tensor([[1., 2, 3], [4,5,6]])
> 3 c+m
RuntimeError: The size of tensor a (2) must match the size of tensor b (3) at nonsingleton dimension 1
= tensor([10.,20,30])
c = tensor([[1., 2, 3], [4,5,6], [7,8,9]])
m # Expand the vector to broadcast across a different dimension
= c.unsqueeze(1)
c m.shape,c.shape, c
(torch.Size([3, 3]),
torch.Size([3, 1]),
tensor([[10.],
[20.],
[30.]]))
c.expand_as(m)
tensor([[10., 10., 10.],
[20., 20., 20.],
[30., 30., 30.]])
+m c
tensor([[11., 12., 13.],
[24., 25., 26.],
[37., 38., 39.]])
= c.expand_as(m)
t t.storage()
10.0
20.0
30.0
[torch.FloatStorage of size 3]
t.stride(), t.shape
((1, 0), torch.Size([3, 3]))
Note: By default, new broadcast dimensions are added at the beginning using c.unsqueeze(0)
behind the scenes.
= tensor([10.,20,30])
c 0).shape,c.unsqueeze(1).shape c.shape, c.unsqueeze(
(torch.Size([3]), torch.Size([1, 3]), torch.Size([3, 1]))
Note: The unsqueeze command can be replaced by None indexing.
None,:].shape,c[:,None].shape c.shape, c[
(torch.Size([3]), torch.Size([1, 3]), torch.Size([3, 1]))
Note: * You can omit training columns when indexing * ...
means all preceding dimensions
None].shape,c[...,None].shape c[
(torch.Size([1, 3]), torch.Size([3, 1]))
1) c,c.unsqueeze(
(tensor([10., 20., 30.]),
tensor([[10.],
[20.],
[30.]]))
# Replace the second for loop with broadcasting
def matmul(a,b):
= a.shape
ar,ac = b.shape
br,bc assert ac==br
= torch.zeros(ar, bc)
c for i in range(ar):
# c[i,j] = (a[i,:] * b[:,j]).sum() # previous
= (a[i ].unsqueeze(1) * b).sum(dim=0)
c[i] return c
1).shape m1.shape, m1.unsqueeze(
(torch.Size([5, 784]), torch.Size([5, 784, 1]))
0].unsqueeze(1).expand_as(m2).shape m1[
torch.Size([784, 10])
%timeit n 20 t4 = matmul(m1,m2)
414 µs ± 18 µs per loop (mean ± std. dev. of 7 runs, 20 loops each)
Note: Even faster still, though the improvement is not as dramatic.
Broadcasting rules
 when operating on two tensors, PyTorch compares their shapes elementwise
 starts with the trailing dimensions and works with its way backward
 adds
1
when it meets and empty dimension
 two dimensions are compatible when one of the following is true
 They are equal
 One of them is 1, in which case that dimension is broadcast to make it the same as the other
 arrays do not need to have the same number of dimensions
Einstein Summation
 a compact representation for combining products and sums in a general way
 \(ik,kj \rightarrow ij\)
 lefthand side represents the operands dimensions, separated by commas
 righthand side represents the result dimensions
 a practical way of expressing operations involving indexing and sum of products
Notaion Rules
 Repeated indices are implicitly summed over.
 Each index can appear at most twice in any term.
 Each term must contain identical nonrepeated indices.
def matmul(a,b): return torch.einsum('ik,kj>ij', a, b)
%timeit n 20 t5 = matmul(m1,m2)
The slowest run took 10.35 times longer than the fastest. This could mean that an intermediate result is being cached.
26.9 µs ± 37.3 µs per loop (mean ± std. dev. of 7 runs, 20 loops each)
Note: It is extremely fast even compared to the earlier broadcast implementation. * einsum is often the fastest way to do custom operations in PyTorch, without diving into C++ and CUDA * still not as fast as carefully optimizes CUDA code
Additional Einsum Notation Examples
= torch.randn(2, 2)
x print(x)
# Transpose
'ij>ji', x) torch.einsum(
tensor([[ 0.7541, 0.8633],
[ 2.2312, 0.0933]])
tensor([[ 0.7541, 2.2312],
[0.8633, 0.0933]])
= torch.randn(2, 2)
x = torch.randn(2, 2)
y = torch.randn(2, 2)
z print(x)
print(y)
print(z)
# Return a vector of size b where the kth coordinate is the sum of x[k,i] y[i,j] z[k,j]
'bi,ij,bj>b', x, y, z) torch.einsum(
tensor([[0.2458, 0.7571],
[ 0.0921, 0.5496]])
tensor([[1.2792, 0.0648],
[0.2263, 0.1153]])
tensor([[0.2433, 0.4558],
[ 0.8155, 0.5406]])
tensor([0.0711, 0.2349])
# trace
= torch.randn(2, 2)
x 'ii', x) x, torch.einsum(
(tensor([[ 1.4828, 0.7057],
[0.6288, 1.3791]]),
tensor(2.8619))
# diagonal
= torch.randn(2, 2)
x 'ii>i', x) x, torch.einsum(
(tensor([[1.0796, 1.1161],
[ 2.2944, 0.6225]]),
tensor([1.0796, 0.6225]))
# outer product
= torch.randn(3)
x = torch.randn(2)
y f"x: {x}", f"y: {y}", torch.einsum('i,j>ij', x, y)
('x: tensor([ 0.1439, 1.8456, 1.5355])',
'y: tensor([0.7276, 0.5566])',
tensor([[0.1047, 0.0801],
[ 1.3429, 1.0273],
[ 1.1172, 0.8547]]))
# batch matrix multiplication
= torch.randn(3,2,5)
As = torch.randn(3,5,4)
Bs 'bij,bjk>bik', As, Bs) torch.einsum(
tensor([[[ 1.9657, 0.5904, 2.8094, 2.2607],
[ 0.7610, 2.0402, 0.7331, 2.2257]],
[[1.5433, 2.9716, 1.3589, 0.1664],
[ 2.7327, 4.4207, 1.1955, 0.5618]],
[[1.7859, 0.8143, 0.8410, 0.2257],
[3.4942, 1.9947, 0.7098, 0.5964]]])
# with sublist format and ellipsis
0, 1], Bs, [..., 1, 2], [..., 0, 2]) torch.einsum(As, [...,
tensor([[[ 1.9657, 0.5904, 2.8094, 2.2607],
[ 0.7610, 2.0402, 0.7331, 2.2257]],
[[1.5433, 2.9716, 1.3589, 0.1664],
[ 2.7327, 4.4207, 1.1955, 0.5618]],
[[1.7859, 0.8143, 0.8410, 0.2257],
[3.4942, 1.9947, 0.7098, 0.5964]]])
# batch permute
= torch.randn(2, 3, 4, 5)
A '...ij>...ji', A).shape torch.einsum(
torch.Size([2, 3, 5, 4])
# equivalent to torch.nn.functional.bilinear
= torch.randn(3,5,4)
A = torch.randn(2,5)
l = torch.randn(2,4)
r 'bn,anm,bm>ba', l, A, r) torch.einsum(
tensor([[ 1.1410, 1.7888, 4.7315],
[ 3.8092, 3.0976, 2.2764]])
The Forward and Backward Passes
Defining and Initializing a Layer
# Linear layer
def lin(x, w, b): return x @ w + b
# Initialize random input values
= torch.randn(200, 100)
x # Initialize random target values
= torch.randn(200) y
# Initialize layer 1 weights
= torch.randn(100,50)
w1 # Initialize layer 1 biases
= torch.zeros(50)
b1 # Initialize layer 2 weights
= torch.randn(50,1)
w2 # Initialize layer 2 biases
= torch.zeros(1) b2
# Get a batch of hidden state
= lin(x, w1, b1)
l1 l1.shape
torch.Size([200, 50])
l1.mean(), l1.std()
(tensor(0.0385), tensor(10.0544))
Note: Having with activations with a high standard deviation is a problem since the values can scale to numbers that can’t be represented by a computer by the end of the model.
= torch.randn(200, 100)
x for i in range(50): x = x @ torch.randn(100,100)
0:5,0:5] x[
tensor([[nan, nan, nan, nan, nan],
[nan, nan, nan, nan, nan],
[nan, nan, nan, nan, nan],
[nan, nan, nan, nan, nan],
[nan, nan, nan, nan, nan]])
Note: Having activations that are too small can cause all the activations at the end of the model to go to zero.
= torch.randn(200, 100)
x for i in range(50): x = x @ (torch.randn(100,100) * 0.01)
0:5,0:5] x[
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.]])
Note: Need to scale the weight matrices so the standard deviation of the activations stays at 1 * Understanding the difficulty of training deep feedforward neural networks * the right scale for a given layer is \(\frac{1}{\sqrt{n_{in}}}\), where \(n_{in}"\) represents the number of inputs.
Note: For a layer with 100 inputs, \(\frac{1}{\sqrt{100}}=0.1\)
= torch.randn(200, 100)
x for i in range(50): x = x @ (torch.randn(100,100) * 0.1)
0:5,0:5] x[
tensor([[1.7695, 0.5923, 0.3357, 0.7702, 0.8877],
[ 0.6093, 0.8594, 0.5679, 0.4050, 0.2279],
[ 0.4312, 0.0497, 0.1795, 0.3184, 1.7031],
[0.7370, 0.0251, 0.8574, 0.6826, 2.0615],
[0.2335, 0.0042, 0.1503, 0.2087, 0.0405]])
x.std()
tensor(1.0150)
Note: Even a slight variation from \(0.1\) will dramatically change the values
# Redefine inputs and targets
= torch.randn(200, 100)
x = torch.randn(200) y
from math import sqrt
# Scale the weights based on the number of inputs to the layers
= torch.randn(100,50) / sqrt(100)
w1 = torch.zeros(50)
b1 = torch.randn(50,1) / sqrt(50)
w2 = torch.zeros(1) b2
= lin(x, w1, b1)
l1 l1.mean(),l1.std()
(tensor(0.0062), tensor(1.0231))
# Define nonlinear activation function
def relu(x): return x.clamp_min(0.)
= relu(l1)
l2 l2.mean(),l2.std()
(tensor(0.3758), tensor(0.6150))
Note: The activation function ruined the mean and standard deviation. * The \(\frac{1}{\sqrt{n_{in}}}\) weight initialization method used not account for the ReLU function.
= torch.randn(200, 100)
x for i in range(50): x = relu(x @ (torch.randn(100,100) * 0.1))
0:5,0:5] x[
tensor([[1.2172e08, 0.0000e+00, 0.0000e+00, 7.1241e09, 5.9308e09],
[1.9647e08, 0.0000e+00, 0.0000e+00, 9.2189e09, 7.1026e09],
[1.8266e08, 0.0000e+00, 0.0000e+00, 1.1150e08, 7.0774e09],
[1.8673e08, 0.0000e+00, 0.0000e+00, 1.0574e08, 7.3020e09],
[2.1829e08, 0.0000e+00, 0.0000e+00, 1.1662e08, 1.0466e08]])
Delving Deep into Rectifiers: Surpassing HumanLevel Performance on ImageNet Classification
 the article that introduced ResNet
 Introduced Kaiming initialization:
 \(\sqrt{\frac{2}{n_{in}}}\), where \(n_{in}\) is the number of inputs of our model
= torch.randn(200, 100)
x for i in range(50): x = relu(x @ (torch.randn(100,100) * sqrt(2/100)))
0:5,0:5] x[
tensor([[0.0000, 0.0000, 0.1001, 0.0358, 0.0000],
[0.0000, 0.0000, 0.1612, 0.0164, 0.0000],
[0.0000, 0.0000, 0.0000, 0.0000, 0.0000],
[0.0000, 0.0000, 0.1764, 0.0000, 0.0000],
[0.0000, 0.0000, 0.1331, 0.0000, 0.0000]])
= torch.randn(200, 100)
x = torch.randn(200) y
= torch.randn(100,50) * sqrt(2 / 100)
w1 = torch.zeros(50)
b1 = torch.randn(50,1) * sqrt(2 / 50)
w2 = torch.zeros(1) b2
= lin(x, w1, b1)
l1 = relu(l1)
l2 l2.mean(), l2.std()
(tensor(0.5720), tensor(0.8259))
def model(x):
= lin(x, w1, b1)
l1 = relu(l1)
l2 = lin(l2, w2, b2)
l3 return l3
= model(x)
out out.shape
torch.Size([200, 1])
# Squeeze the output to get rid of the trailing dimension
def mse(output, targ): return (output.squeeze(1)  targ).pow(2).mean()
= mse(out, y) loss
Gradients and the Backward Pass
 the gradients are computed in the backward pass using the chain rule from calculus
 chain rule: \((g \circ f)'(x) = g'(f(x)) f'(x)\)
 our loss if a big composition of different functions
= mse(out,y) = mse(lin(l2, w2, b2), y) loss
 chain rule: \[\frac{\text{d} loss}{\text{d} b_{2}} = \frac{\text{d} loss}{\text{d} out} \times \frac{\text{d} out}{\text{d} b_{2}} = \frac{\text{d}}{\text{d} out} mse(out, y) \times \frac{\text{d}}{\text{d} b_{2}} lin(l_{2}, w_{2}, b_{2})\]
 To compute all the gradients we need for the update, we need to begin from the output of the model and work our way backward, one layer after the other.
 We can automate this process by having each function we implemented provided its backward step
Gradient of the loss function
 undo the squeeze in the mse function
 calculate the derivative of \(x^{2}\): \(2x\)
 calculate the derivative of the mean: \(\frac{1}{n}\) where \(n\) is the number of elements in the input
def mse_grad(inp, targ):
# grad of loss with respect to output of previous layer
= 2. * (inp.squeeze()  targ).unsqueeze(1) / inp.shape[0] inp.g
Gradient of the ReLU activation function
def relu_grad(inp, out):
# grad of relu with respect to input activations
= (inp>0).float() * out.g inp.g
Gradient of a linear layer
def lin_grad(inp, out, w, b):
# grad of matmul with respect to input
= out.g @ w.t()
inp.g = inp.t() @ out.g
w.g = out.g.sum(0) b.g
SymPy
 a library for symbolic computation that is extremely useful when working with calculus
 Symbolic computation deals with the computation of mathematical objects symbolically
 the mathematical objects are represented exactly, not approximately, and mathematical expressions with unevaluated variables are left in symbolic form
from sympy import symbols,diff
= symbols('sx sy')
sx,sy # Calculate the derivative of sx**2
**2, sx)
diff(sx
2*sx
Define Forward and Backward Pass
def forward_and_backward(inp, targ):
# forward pass:
= inp @ w1 + b1
l1 = relu(l1)
l2 = l2 @ w2 + b2
out # we don't actually need the loss in backward!
= mse(out, targ)
loss
# backward pass:
mse_grad(out, targ)
lin_grad(l2, out, w2, b2)
relu_grad(l1, l2) lin_grad(inp, l1, w1, b1)
Refactoring the Model
 define classes for each function that include their own forward and backward pass functions
class Relu():
def __call__(self, inp):
self.inp = inp
self.out = inp.clamp_min(0.)
return self.out
def backward(self): self.inp.g = (self.inp>0).float() * self.out.g
class Lin():
def __init__(self, w, b): self.w,self.b = w,b
def __call__(self, inp):
self.inp = inp
self.out = inp@self.w + self.b
return self.out
def backward(self):
self.inp.g = self.out.g @ self.w.t()
self.w.g = self.inp.t() @ self.out.g
self.b.g = self.out.g.sum(0)
class Mse():
def __call__(self, inp, targ):
self.inp = inp
self.targ = targ
self.out = (inp.squeeze()  targ).pow(2).mean()
return self.out
def backward(self):
= (self.inp.squeeze()self.targ).unsqueeze(1)
x self.inp.g = 2.*x/self.targ.shape[0]
class Model():
def __init__(self, w1, b1, w2, b2):
self.layers = [Lin(w1,b1), Relu(), Lin(w2,b2)]
self.loss = Mse()
def __call__(self, x, targ):
for l in self.layers: x = l(x)
return self.loss(x, targ)
def backward(self):
self.loss.backward()
for l in reversed(self.layers): l.backward()
= Model(w1, b1, w2, b2) model
= model(x, y) loss
model.backward()
Going to PyTorch
# Define a base class for all functions in the model
class LayerFunction():
def __call__(self, *args):
self.args = args
self.out = self.forward(*args)
return self.out
def forward(self): raise Exception('not implemented')
def bwd(self): raise Exception('not implemented')
def backward(self): self.bwd(self.out, *self.args)
class Relu(LayerFunction):
def forward(self, inp): return inp.clamp_min(0.)
def bwd(self, out, inp): inp.g = (inp>0).float() * out.g
class Lin(LayerFunction):
def __init__(self, w, b): self.w,self.b = w,b
def forward(self, inp): return inp@self.w + self.b
def bwd(self, out, inp):
= out.g @ self.w.t()
inp.g self.w.g = inp.t() @ self.out.g
self.b.g = out.g.sum(0)
class Mse(LayerFunction):
def forward (self, inp, targ): return (inp.squeeze()  targ).pow(2).mean()
def bwd(self, out, inp, targ):
= 2*(inp.squeeze()targ).unsqueeze(1) / targ.shape[0] inp.g
torch.autograd.Function
 In PyTorch, each basic function we need to differentiate is written as a torch.autograd.Function that has a forward and backward method
from torch.autograd import Function
class MyRelu(Function):
# Performs the operation
@staticmethod
def forward(ctx, i):
= i.clamp_min(0.)
result
ctx.save_for_backward(i)return result
# Defines a formula for differentiating the operation with
# backward mode automatic differentiation
@staticmethod
def backward(ctx, grad_output):
= ctx.saved_tensors
i, return grad_output * (i>0).float()
help(staticmethod)
Help on class staticmethod in module builtins:
class staticmethod(object)
 staticmethod(function) > method

 Convert a function to be a static method.

 A static method does not receive an implicit first argument.
 To declare a static method, use this idiom:

 class C:
 @staticmethod
 def f(arg1, arg2, ...):
 ...

 It can be called either on the class (e.g. C.f()) or on an instance
 (e.g. C().f()). Both the class and the instance are ignored, and
 neither is passed implicitly as the first argument to the method.

 Static methods in Python are similar to those found in Java or C++.
 For a more advanced concept, see the classmethod builtin.

 Methods defined here:

 __get__(self, instance, owner, /)
 Return an attribute of instance, which is of type owner.

 __init__(self, /, *args, **kwargs)
 Initialize self. See help(type(self)) for accurate signature.

 
 Static methods defined here:

 __new__(*args, **kwargs) from builtins.type
 Create and return a new object. See help(type) for accurate signature.

 
 Data descriptors defined here:

 __dict__

 __func__

 __isabstractmethod__
torch.nn.Module
 the base structure for all models in PyTorch
Implementation Steps 1. Make sure the superclass __init__
is called first when you initialize it. 2. Define any parameters of the model as attributes with nn.Parameter
. 3. Define a forward function that returns the output of your model.
import torch.nn as nn
class LinearLayer(nn.Module):
def __init__(self, n_in, n_out):
super().__init__()
self.weight = nn.Parameter(torch.randn(n_out, n_in) * sqrt(2/n_in))
self.bias = nn.Parameter(torch.zeros(n_out))
def forward(self, x): return x @ self.weight.t() + self.bias
= LinearLayer(10,2)
lin = lin.parameters()
p1,p2 p1.shape,p2.shape
(torch.Size([2, 10]), torch.Size([2]))
class Model(nn.Module):
def __init__(self, n_in, nh, n_out):
super().__init__()
self.layers = nn.Sequential(
nn.Linear(n_in,nh), nn.ReLU(), nn.Linear(nh,n_out))self.loss = mse
def forward(self, x, targ): return self.loss(self.layers(x).squeeze(), targ)
Note: fsatai provides its own variant of Module that is identical to nn.Module
, but automatically calls super().__init__()
.
class Model(Module):
def __init__(self, n_in, nh, n_out):
self.layers = nn.Sequential(
nn.Linear(n_in,nh), nn.ReLU(), nn.Linear(nh,n_out))self.loss = mse
def forward(self, x, targ): return self.loss(self.layers(x).squeeze(), targ)
Conclusion
 A neural net is a bunch of matrix multiplications with nonlinearities in between
 Vectorize and take advantage of techniques such as elementwise arithmetic and broadcasting when possible
 Two tensors are broadcastable if the dimensions starting from the end and going backward match
 May need to add dimensions of size one to make tensors broadcastable
 Properly initializing a neural net is crucial to get training started
 Use Kaiming initialization when using ReLU
 The backward pass is the chain rule applied multiple times, computing the gradient from the model output and going back, one layer at a time
References
Previous: Notes on fastai Book Ch. 16