SUBLLM: A Novel Efficient Architecture with Token Sequence Subsampling for LLM

Add code
Jun 03, 2024

Share this with someone who'll enjoy it:

View paper onarxiv icon

Share this with someone who'll enjoy it: