Introduction to Pytorch
Scott Ladenheim, PhD
help@[Link]
Please check
your name on
the attendance
sheet
• The course materials can be found at:
[Link]
• If you have not already done so, download the Jupyter
notebook file [Link]
Course materials
• On the SCC copy the notebook to your home directory using
this terminal command:
cp /project/scv/examples/machine_learning/pytorch/[Link] ~
• You can also download the slides
[Link]
Jupyter notebooks using SCC OnDemand
If you have an SCC account
• Open a web browser
• Navigate to [Link]
• Login using BU id and password
• Select Jupyter Notebook from the Interactive Apps tab
• Fill in the request with the following information
• Modules: miniconda/23.1.0 academic-ml
• Pre-Launch Command: conda activate spring-2024-pyt
• Number of hours: 2
• Number of cores: 1
• Click the blue Launch button
Jupyter notebooks using SCC OnDemand
If you do NOT have an SCC account
• Open a web browser
• Navigate to: [Link]
• Login using
• ID tuta# (# is the number on your desk)
• Password is LittleMartha# (# is the number on your desk)
• Select Jupyter Notebook from the Interactive Apps tab
• Fill in the request with the following information
• Modules: miniconda/23.1.0 academic-ml
• Pre-Launch Command: conda activate spring-2024-pyt
• Number of hours: 2
• Number of cores: 1
• Click the blue Launch button
Tutorial outline
Pytorch neural network
CPU vs GPU
Autograd
Navigate to
[Link]
After the
tutorial
Tutorial feedback survey