ConceptAI & ML
Curriculum-guided layer scaling
A pretraining approach where a model starts at a smaller size and easier data, gradually adding layers throughout training until reaching the final model size on hardest data.
Know first
Named as a prerequisite by the lectures that define this. The number is how many say so. A grey one has no page yet.
Language model pretraining1Layer stacking1
Lecturers
1Connections
1The lectures that define it or assume it, and the concepts next to it.