pytorch

Pytorch

version

torch: 1.13.1

Prepare

poetry install

Run

poetry run python pytorch/distributed_example.py

CNN

Input x Filter (kernel) -> Feature Map (Overlapping application)

• Filter: The filter is smaller than the input data and the type of multiplication applied between a filter-sized patch of the input and the filter is a dot product. (will be learned by SGD to minimize the classification accuracy) Filter can capture feature. -> Multiple filters

• Feature Map: low-level features

• Multiple channels: color (red, green and blue). filter needs to have the same number of channels. 3x3 filter for 3 channles -> 3x3x3 (3-d)

• Multiple layers: extract features that are combination of lower-level features.

• https://machinelearningmastery.com/how-to-develop-a-convolutional-neural-network-from-scratch-for-mnist-handwritten-digit-classification/

• convolutional layer

References

• Neural Networks
• train_dist.py
• Pytorch distributed overview

Error

No module named '_lzma'

Traceback (most recent call last):
  File "/Users/nakamasato/repos/nakamasato/ml-training/pytorch/distributed_example.py", line 4, in <module>
    import lzma
  File "/Users/nakamasato/.pyenv/versions/3.9.0/lib/python3.9/lzma.py", line 27, in <module>
    from _lzma import *
ModuleNotFoundError: No module named '_lzma'

brew install xz