[docs] Pretrained model docs should explain how to specify cache dir and norm_layer
Nobody has claimed this yet.
- Dominant language
- Python
- Stars
- 17.9k
- Forks
- 7.3k
- Avg merge
- 1d 15h
- Merged PRs (30d)
- 13
Description
🐛 Describe the bug
https://pytorch.org/vision/stable/models.html?highlight=resnet18#torchvision.models.resnet18
should document:
-
how to set cache dir for downloaded models. many university systems have tight quota for home dir that prohibits clogging it with weights. it is explained at the very top of very long document (
TORCH_MODEL_ZOO) but it would be nice to duplicate it / link to this from every pretrained method -
how to set
norm_layer = torchvision.ops.misc.FrozenBatchNorm2dsince this is a very frequent need for fine-tuning -
how to replace stride with dilation for ResNet and to what layers it applies and what it can help achieving
Currently docs just specify **kwargs: Any which isn't very helpful
Versions
N/A
Contributor guide
First steps
- Read the whole issue, then the project's contributing guide.
- Comment on the issue to say you are picking it up — it saves two people doing the same work.
- Fork the repository and make your change on a branch.
- Open a pull request that references the issue number.
Research direction
Start with the linked torchvision.models.resnet18 documentation and trace the pretrained model method documentation behind it. Document how to configure the cache directory, use FrozenBatchNorm2d as norm_layer, and replace stride with dilation; done means the relevant **kwargs behavior and its use cases are explained for readers.
Written by the indexing model from the issue text.
Assessment
- Tech stack
- python, pytorch
- Domain
- computer-vision, documentation, machine-learning
- Issue type
- Documentation
- Difficulty
- 3/5
- Estimated time
- 1-2 days
- Activity status
- Stale
- Clarity
- Mostly clear
- Newbie friendliness
- 45/100