-
-
Notifications
You must be signed in to change notification settings - Fork 620
QS labs internship 2023 project ideas
🙌🏼 Mentor(s): Victor Fomin
🧰 Skills needed:
- Good knowledge of Python and Git
- Basic understanding how to write/hack in Javascript
- Intermediate knowledge of PyTorch
- Beginner knowledge of PyTorch-Ignite API and codebase
- Already trained neural networks with PyTorch
- Willing to maintain AI-related open-source project
- Curiosity and motivation to learn new technical things
We are providing to PyTorch-Ignite's community a code-generator tool, a web application built with Vue.js 3 to quickly produce quick-start python code for common training tasks in deep learning. Code is using PyTorch framework and PyTorch-Ignite library and it can be configured in the UI. This tools can be helpful to start working on a task without rewriting everything from scratch: Kaggle competition, client prototype project, etc.
The idea is to provide more valuable templates: object detection, RL, knowledge distillation, active learning, few-shot learning, unsupervised learning, diffusion etc
In addition to that we need to improve the current state of the templates, the application and the its code.
We are thinking of the following work plan:
(easy) Review all existing templates and improve their readmes and the learning curve
(hard) Work on new features:
- "Use wget with download link", (https://github.com/pytorch-ignite/code-generator/issues/95)
- Other optional configuration management (python-fire, hydra, etc), https://github.com/pytorch-ignite/code-generator/issues/102
- Configuration storage for reproducilbilty, https://github.com/pytorch-ignite/code-generator/issues/36
(medium) Add a reinforcement learning template based on [pytorch-ignite's example](https://github.com/pytorch/ignite/tree/master/examples/reinforcement_learning)
(medium) Add an object detection template
(hard) Add a template on one of the following topics: knowledge distillation, active learning, few-shot learning, unsupervised learning, diffusion
(hard) Explore integrations with infrastructure tools: e.g. Nebari. Work on a prototype functionality like "Open in Nebari" (similar to existing)
We would expect all easy and at least 4 medium items (hard item == 2 medium) from the above list to be done for this project. In addition, it would be nice to have a short blog post communicating about the work done.
PyTorch-Ignite presented to you with love by PyTorch community