Eastbay ML will be hosting a series of meetups as we learn pytorch together. Presenters will be volunteers from the group and are not domain experts but will lead the discussion as we learn together. The meetups assume familiarity with python and some exposure to deep learning concepts. We hope to use this series as a tool for exploring implemenation details of popular architectures and methods.
- Suitable both to people new to deep learning and people with experience on other frameworks
- Use colab notebooks for hands on learning
- Notebooks should be precanned, but exercises require some coding in class
- Attendees should not require work outside of the meetup
- Sessions recorded, or alternate videos for people that miss a meetup
- Volunteer presenters guide each topic
- Use existing content where available. Presenters are not expected to be authors
- Presenters are not expected to spend more that a couple hours in preparation
| Week | Topic | Presenter | Notebook |
|---|---|---|---|
| 1/30 | Intro,Tensors, GPU,Autograd, nn-model, SGD | Roger | lesson 1 |
| 2/13 | Torchvision,CNN image classification, Resnet, | Jerry | lesson 2 |
| 2/27 | Data Utilities; Optimizers - Adam, Regularization | Shalom | |
| 3/12 | visualization and tensorboard | Jerry | |
| 3/26 | CNN object detection | Praveen | |
| 4/9 | transfer learning | Selly | |
| 4/23 | RNNs and time series | Dev | |
| 5/7 | NLP | Dev | |
| 5/21 | Parallel and Distributed Training | Jeremy | |
| 6/4 | Mobile/embedded Platform - Quantization, computational efficiency and deploy to edge | Selly | |
| 6/18 | RL | Selly | |
| 7/2 | ACKTR | Roger | |
| 7/16 | GAN / VAE | Shalom | |
| 7/30 | Nextgen CNN - Mobilnet, EfficientNet, | Patrick |
A series of notebooks from our pytorch tutorial series