Lesson 4
Raw notes
- ULMFit
- Language model from wikipedia (wikitext 103)
- Took pretrained model from wikipedia and ran a few epochs on IMDb
- fine-tuned this for a classifier
- https://www.kaggle.com/code/jhoward/getting-started-with-nlp-for-absolute-beginners.
- should learn numpy, pandas, matplotlib and pytorch.
- Python for Data Analysis book - O'Reilly
df.describe(include='object')
- think "what are some key features in this dataset?"
deberta-v3-small
- TODO: what is SIMD?
- Be careful using randomized validation sets.
- https://www.fast.ai/posts/2017-11-13-validation-sets.html
- TODO: what is cross-validation?
- https://www.fast.ai/posts/2019-09-24-metrics.html
Backlinks