DEV Community

Super Kai (Kazuya Ito)
Super Kai (Kazuya Ito)

Posted on • Edited on

MaxPool1d in PyTorch

Buy Me a Coffee

*Memos:

MaxPool1d() can get the 2D or 3D tensor of the one or more elements computed by 1D max pooling from the 2D or 3D tensor of one or more elements as shown below:

*Memos:

  • The 1st argument for initialization is kernel_size(Required-Type:int or tuple or list of int). *It must be 1 <= x.
  • The 2nd argument for initialization is stride(Optional-Default:None-Type:int or tuple or list of int): *Memos:
    • It must be 1 <= x.
    • If it's None, kernel_size is set.
  • The 3rd argument for initialization is padding(Optional-Default:0-Type:int or tuple or list of int). *It must be 0 <= x.
  • The 4th argument for initialization is dilation(Optional-Default:1-Type:int or tuple or list of int). *It must be 1 <= x.
  • The 5th argument for initialization is return_indices(Optional-Default:False-Type:bool).
  • The 6th argument for initialization is ceil_mode(Optional-Default:False-Type:bool).
  • The 1st argument is input(Required-Type:tensor of float): *Memos:
    • It must be the 2D or 3D tensor of one or more elements.
    • The tensor's requires_grad which is False by default is not set to True by MaxPool1d().
import torch
from torch import nn

tensor1 = torch.tensor([[8., -3., 0., 1., 5., -2.]])

tensor1.requires_grad
# False

maxpool1d = nn.MaxPool1d(kernel_size=1)
tensor2 = maxpool1d(input=tensor1)
tensor2
# tensor([[8., -3., 0., 1., 5., -2.]])

tensor2.requires_grad
# False

maxpool1d
# MaxPool1d(kernel_size=1, stride=1, padding=0, dilation=1, ceil_mode=False)

maxpool1d.kernel_size
# 1

maxpool1d.stride
# 1

maxpool1d.padding
# 0

maxpool1d.dilation
# 1

maxpool1d.return_indices
# False

maxpool1d.ceil_mode
# False

maxpool1d = nn.MaxPool1d(kernel_size=1, stride=None, padding=0, 
                         dilation=1, return_indices=False, ceil_mode=False)
maxpool1d(input=tensor1)
# tensor([[8., -3., 0., 1., 5., -2.]])

maxpool1d = nn.MaxPool1d(kernel_size=2, return_indices=True)
maxpool1d(input=tensor1)
# (tensor([[8., 1., 5.]]), tensor([[0, 3, 4]]))

maxpool1d = nn.MaxPool1d(kernel_size=3, return_indices=True)
maxpool1d(input=tensor1)
# (tensor([[8., 5.]]), tensor([[0, 4]]))

maxpool1d = nn.MaxPool1d(kernel_size=4, return_indices=True)
maxpool1d(input=tensor1)
# (tensor([[8.]]), tensor([[0]]))

maxpool1d = nn.MaxPool1d(kernel_size=5, return_indices=True)
maxpool1d(input=tensor1)
# (tensor([[8.]]), tensor([[0]]))

maxpool1d = nn.MaxPool1d(kernel_size=6, return_indices=True)
maxpool1d(input=tensor1)
# (tensor([[8.]]), tensor([[0]]))

my_tensor = torch.tensor([[8., -3., 0.],
                          [1., 5., -2.]])
maxpool1d = nn.MaxPool1d(kernel_size=1, return_indices=True)
maxpool1d(input=my_tensor)
# (tensor([[8., -3., 0.],
#          [1., 5., -2.]]),
#  tensor([[0, 1, 2],
#          [0, 1, 2]]))

maxpool1d = nn.MaxPool1d(kernel_size=2, return_indices=True)
maxpool1d(input=my_tensor)
# (tensor([[8.],
#          [5.]]),
#  tensor([[0],
#          [1]]))

maxpool1d = nn.MaxPool1d(kernel_size=3, return_indices=True)
maxpool1d(input=my_tensor)
# (tensor([[8.],
#          [5.]]),
#  tensor([[0],
#          [1]]))

my_tensor = torch.tensor([[8.], [-3.], [0.], [1.], [5.], [-2.]])

maxpool1d = nn.MaxPool1d(kernel_size=1, return_indices=True)
maxpool1d(input=my_tensor)
# (tensor([[8.], [-3.], [0.], [1.], [5.], [-2.]]),
#  tensor([[0], [0], [0], [0], [0], [0]]))

maxpool1d = nn.MaxPool1d(kernel_size=1, return_indices=True)
maxpool1d(input=my_tensor)
# (tensor([[8.], [-3.], [0.], [1.], [5.], [-2.]]),
#  tensor([[0], [0], [0], [0], [0], [0]]))

my_tensor = torch.tensor([[[8.], [-3.], [0.]],
                          [[1.], [5.], [-2.]]])
maxpool1d = nn.MaxPool1d(kernel_size=1, return_indices=True)
maxpool1d(input=my_tensor)
# (tensor([[[8.], [-3.], [0.]],
#          [[1.], [5.], [-2.]]]),
#  tensor([[[0], [0], [0]],
#          [[0], [0], [0]]]))

maxpool1d = nn.MaxPool1d(kernel_size=2, padding=1, return_indices=True)
maxpool1d(input=my_tensor)
# (tensor([[[8.], [-3.], [0.]],
#          [[1.], [5.], [-2.]]]),
#  tensor([[[0], [0], [0]],
#          [[0], [0], [0]]]))

maxpool1d = nn.MaxPool1d(kernel_size=3, padding=1, return_indices=True)
maxpool1d(input=my_tensor)
# (tensor([[[8.], [-3.], [0.]],
#          [[1.], [5.], [-2.]]]),
#  tensor([[[0], [0], [0]],
#          [[0], [0], [0]]]))
etc.
Enter fullscreen mode Exit fullscreen mode

Image of Timescale

🚀 pgai Vectorizer: SQLAlchemy and LiteLLM Make Vector Search Simple

We built pgai Vectorizer to simplify embedding management for AI applications—without needing a separate database or complex infrastructure. Since launch, developers have created over 3,000 vectorizers on Timescale Cloud, with many more self-hosted.

Read full post →

Top comments (0)

Billboard image

The Next Generation Developer Platform

Coherence is the first Platform-as-a-Service you can control. Unlike "black-box" platforms that are opinionated about the infra you can deploy, Coherence is powered by CNC, the open-source IaC framework, which offers limitless customization.

Learn more

👋 Kindness is contagious

Please leave a ❤️ or a friendly comment on this post if you found it helpful!

Okay