Hi everyone! Each semester I lecture at the University of West Florida through one of our clubs on fastai. Normally I keep it offline but this year I have decided to live stream it along with have a dedicated megathread to discussing the course, asking questions, and networking together! This semester’s version will be using the version 2 of the fastai library (thank you to @Jeremy and @sgugger for all the amazing work you have put in!). This course is designed to be intro friendly (it’s geared towards Undergraduate students in terms of pre-reqs). Now that the intro is done, onto the important bits! Below I have detailed key dates and a syllabus. I am working on the material as we speak but in the meantime take a look at my Practical Deep Learning for Coders 2.0 repository for some v2 codebase to get caught up on v2 if you are unfamiliar and don’t want to wait!
Videos:
Block 1: Computer Vision
Lesson 1, Part 1
Lesson 1, Part 2
Lesson 2
Lesson 3
Lesson 4
Lesson 5
Lesson 6
Lesson 7
Block 2: Tabular Neural Networks
Lesson 1
Lesson 2
Lesson 3
Lesson 4
Block 3: Natural Language Processing
Lesson 1
Lesson 2 (pending)
Lesson 3 (pending)
(Colab Links) Notebooks we have covered:
01_Pets
01_Custom
02_MNIST
02_SGD
02_Deployment
03_Multi_Label
03_Unknown_Labels
03_Cross_Validation
03_Internal_API_Walkthrough
04_Segmentation
04_ImageWoof
04_DataBlock_Summary
05_Style_transfer
05_Inference_Server
05_EfficientNet_and_Custom_Weights.ipynb
06_Scalar_Regression
06_Keypoint_Regression
06_Hybridizing_Models
06_Object_Detection
06_Multimodal_Head_and_Kaggle
07_Super_Resolution
07_Siamese
07_Audio
07_Binary_Segmentation
A walk with fastai2
How is this different from Practical Deep Learning for Coders?
This course focuses on a subject-to-subject basis, exploring the datablock API to it’s full extent and applying various techniques that are not taught in either course (such as feature importance and k-fold validation).
What do I need to take part?
You need a Google account and a Paperspace account. Both will utilize their free notebooks. We will mostly be working out of Google Colaboratory for all except Natural Language Processing, as Paperspace is nicer for that with data persistence.
When is this?
This course will be running from January 15th until the end of April. This will coincide with Jeremy’s run for Practical Deep Learning and I highly recommend doing both (if it winds up being online available shortly afterwards and or live-streamed). The livestreams will be from 5pm to 7:30pm Central Standard Time on Saturdays.
How is this structured?
We will do lecture for one hour to an hour 15 minutes, with the rest of the time dedicated to debugging and working through the lecture material together along with individual project time.
How do I make the most out of this?
Spend an hour or two a day going through the notebooks and playing around with everything, learning how everything works together. And also get yourself a mini-project to do! If you can’t come up with one yourself, we can all brainstorm together to find a few!
Schedule
Here is the overall schedule. The format different than Jeremy’s course in the sense of we will move from datatype to datatype, starting with Computer Vision and ending with NLP.
this schedule is subject to change
BLOCKS:
- Block 1: Computer Vision
- Block 2: Tabular Neural Networks
- Block 3: Natural Language Processing
Here is the overall schedule broken down by week:
This schedule is subject to change
Block 1 (January 15th - March 4th):
- Lesson 1: PETs and Custom Datasets (a warm introduction to the DataBlock API)
- Lesson 2: Image Classification Models from Scratch, Stochastic Gradient Descent, Deployment, Exploring the Documentation and Source Code
- Lesson 3: Multi-Label Classification, Dealing with Unknown Labels, and K-Fold Validation
- Lesson 4: Image Segmentation, Weighted Loss Functions, State-of-the-Art in Computer Vision
- Lesson 5: Style Transfer,
nbdev
, and Deployment - Lesson 6: Keypoint Regression and Object Detection, More Pose
- Lesson 7: Image Generation, Audio, Other DataBlocks
Block 2 (March 18th - April 8th):
- Lesson 1: Pandas Workshop and Tabular Classification, SHAP
- Lesson 2: Feature Engineering and Tabular Regression, Permutation Importance
- Lesson 3: Bayesian Optimization, Cross-Validation, and Labeled Test Sets
- Lesson 4: TabNet, DeepGBM
BLOCK 3 (April 15th - May 6th):
- Lesson 1: Introduction to NLP and the LSTM
- Lesson 2: Full Sentiment Classification, Tokenizers, and Ensembling
- Lesson 3: Other State-of-the-Art NLP Models
- Lesson 4: Multi-Lingual Data, DeViSe
Closing notes
This will be my first time live-streaming so this will be an experiment for everyone but I have high hopes that this will turn out to be a successful study group with your help! Please use this thread for any questions and starting discussions about this material, we’re all learning fastai (and especially the second version) together! I will update this post with youtube links to the livestream, as well as post on this thread as well. Looking forward to seeing everyone next month!!!
(Also minor PSA, this is in no way for any credit whatsoever. I am just an undergraduate student wanting to help others learn how to use this amazing library to its fullest potential. Instead of worrying about credit, try using what you’ve learned into a project or two and some blogs, this provides evidence you know the material much better than a slip of paper can in some cases )