Login / Signup
Baby Llama: knowledge distillation from an ensemble of teachers trained on a small dataset with no performance penalty.
Inar Timiryasov
Jean-Loup Tastet
Published in:
CoRR (2023)
Keyphrases
</>
domain knowledge
learning perl
expert systems
training set
knowledge management
knowledge base
objective function
concept maps
authoring tool
manually labeled
young students
learning process
knowledge representation
training data
effective learning
mathematical concepts
life long
databases