This is the code base for the paper [BERxiT: Early Exiting for BERT with Better Fine-Tuning and Extension to Regression].
The main features of castorini/berxit are: Early Exit Decoding.
Open-source alternatives to castorini/berxit include: autoliuweijie/fastbert — The score code of FastBERT (ACL2020). pan-x-c/ee-llm — EE-LLM is a framework for large-scale training and inference of early-exit (EE) large language models (LLMs).
EE-LLM is a framework for large-scale training and inference of early-exit (EE) large language models (LLMs).
