[BUG] - Incorrect list rendering in Google Colab tutorials

[ribhuji]

Add Link

1. https://docs.pytorch.org/tutorials/advanced/usb_semisup_learn.html
2. https://docs.pytorch.org/tutorials/beginner/basics/autogradqs_tutorial.html
3. https://docs.pytorch.org/tutorials/beginner/basics/data_tutorial.html
4. https://docs.pytorch.org/tutorials/beginner/blitz/neural_networks_tutorial.html
5. https://docs.pytorch.org/tutorials/beginner/nn_tutorial.html
6. https://docs.pytorch.org/tutorials/intermediate/optimizer_step_in_backward_tutorial.html
7. https://docs.pytorch.org/tutorials/recipes/recipes/profiler_recipe.html
8. https://docs.pytorch.org/tutorials/recipes/recipes/timer_quick_start.html
9. https://docs.pytorch.org/tutorials/recipes/torch_compiler_set_stance_tutorial.html
10. https://docs.pytorch.org/tutorials/unstable/gpu_direct_storage.html

Describe the bug

Description

The list-rendering problem previously reported in #3939 also occurs in several other tutorials.

On the PyTorch Tutorials HTML pages, the affected lists render correctly. However, after opening the generated notebooks through Run in Google Colab, the same content is not rendered as a proper list.

Steps to reproduce

1. Open one of the tutorial pages listed below.
2. Select Run in Google Colab.
3. Navigate to the affected section.
4. Compare the list in Google Colab with the corresponding list on the tutorial HTML page.

Affected tutorials

1. USB Semi-Supervised Learning

Tutorial: https://docs.pytorch.org/tutorials/advanced/usb_semisup_learn.html

Affected content: the list of functions imported from semilearn in the Use USB to Train FreeMatch/SoftMatch on CIFAR-10 with only 40 labels section.

2. Automatic Differentiation with torch.autograd

Tutorial: https://docs.pytorch.org/tutorials/beginner/basics/autogradqs_tutorial.html

Affected content:

• the list in the note explaining gradient availability; and
• the list of reasons for disabling gradient tracking.

3. Datasets & DataLoaders

Tutorial: https://docs.pytorch.org/tutorials/beginner/basics/data_tutorial.html

Affected content: the FashionMNIST parameter list in the Loading a Dataset section.

4. Neural Networks

Tutorial: https://docs.pytorch.org/tutorials/beginner/blitz/neural_networks_tutorial.html

Affected content: the lists under Recap, At this point, we covered, and Still Left.

5. What is torch.nn really?

Tutorial: https://docs.pytorch.org/tutorials/beginner/nn_tutorial.html

Affected content: the list of assumptions in the Wrapping DataLoader section.

6. Optimizer Step in Backward

Tutorial: https://docs.pytorch.org/tutorials/intermediate/optimizer_step_in_backward_tutorial.html

Affected content: the numbered list following Several major observations.

7. PyTorch Profiler

Tutorial: https://docs.pytorch.org/tutorials/recipes/recipes/profiler_recipe.html

Affected content: the nested activity types under the activities profiler parameter in the Using profiler to analyze execution time section.

8. Timer Quick Start

Tutorial: https://docs.pytorch.org/tutorials/recipes/recipes/timer_quick_start.html

Affected content: the Contents list near the beginning of the tutorial.

9. Changing the Compilation Stance

Tutorial: https://docs.pytorch.org/tutorials/recipes/torch_compiler_set_stance_tutorial.html

Affected content: the list following Other stances include.

10. GPU Direct Storage

Tutorial: https://docs.pytorch.org/tutorials/unstable/gpu_direct_storage.html

Affected content: the list following The steps involved in the process are as follows.

Expected result

Lists in the generated Google Colab notebooks should preserve the same structure as the corresponding lists on the PyTorch Tutorials HTML pages:

• each item should appear on a separate line;
• bullet and numbered lists should retain their markers;
• nested items should retain their hierarchy; and
• surrounding prose should remain separate from the list.

Describe your environment

Google Colab