What's in the paper?
- A short overview about Transformer NLP models including what they are, why they work so well, and best practices for training modern transformers.
- What type of GPU gets the right results? We trained 3 different transformers of different sizes on different GPU types, see the results.
- System recommendations: Are you ready to train? We have some of the best NLP training systems in the industry, for all budgets.
- Resources: A curated list of recommended readings for further learning about NLP and transformer models.