Flab-Pruner applies unified vocabulary, layer, and FFN pruning with a KL objective to Code LLMs, cutting 22% of parameters while retaining about 97% of code-generation performance and recovering the rest via LoRA post-training.
Analyzing redundancy in pretrained transformer models, in: Proceedings of the 2020 Conference on Empirical Methods in Natural Language Processing (EMNLP), pp
1 Pith paper cite this work. Polarity classification is still indexing.
1
Pith paper citing it
citation-role summary
background 1
citation-polarity summary
fields
cs.SE 1years
2024 1verdicts
CONDITIONAL 1roles
background 1polarities
unclear 1representative citing papers
citing papers explorer
-
Less is More: Towards Green Code Large Language Models via Unified Structural Pruning
Flab-Pruner applies unified vocabulary, layer, and FFN pruning with a KL objective to Code LLMs, cutting 22% of parameters while retaining about 97% of code-generation performance and recovering the rest via LoRA post-training.