AI Paper Atlas

Research map / Large language models / Language, safety and interpretability

LLM-Pruner: On the Structural Pruning of Large Language Models

Xinyin Ma, Gongfan Fang, Xinchao Wang

NeurIPS 2023 · Large language models › Language, safety and interpretability · approach: Pruning and circuits in LLMs

Read the paperCheck your idea against 137,217 papers

AI-extracted. The measured effects below were read by AI from the paper's full text (no summary yet). Numbers and claims can be wrong or missing context, so check the paper before you cite or build on it.

Measured effects reported in the paper

What was changedWhereMeasuredEffect
post-train pruned model with LoRA on AlpacaLLaMA-7B 20% pruned (Element2) · 7 zero-shot tasksaverage accuracy: 56.82 -> 60.07Improved
use block strategy (Element1) instead of channel strategy (ablation)LLaMA-7B 20% pruned no tune · 7 zero-shot tasksaverage accuracy: 45.38 -> 56.69Improved
use dependency-aware group pruning instead of ignoring dependencies (ablation)LLaMA-7B 20% Element1 no tune · 7 zero-shot tasksaverage accuracy: 38.32 -> 56.69Improved
use dependency-aware pruning with tuning (ablation)LLaMA-7B 20% Element1 · WikiText2perplexity: 13307.46 -> 17.58Improved
use gradient-based element importance instead of L2 magnitude (ablation)LLaMA-7B 20% no tune · 7 zero-shot tasksaverage accuracy: 42.65 -> 56.69Improved
use gradient-based importance instead of random pruning (ablation)LLaMA-7B 20% no tune · 7 zero-shot tasksaverage accuracy: 52.69 -> 56.69Improved
second-order Element2 instead of first-order Element1 (ablation)LLaMA-7B 20% w/ tune · 7 zero-shot tasksaverage accuracy: 59.23 -> 60.07No clear effect
element-level importance for ChatGLM-6B instead of vector-level (ablation)ChatGLM-6B 10% no tune · 6 zero-shot tasksaverage accuracy: 46.16 -> 36.45Hurt
local pruning instead of global pruning (ablation)LLaMA-7B 20% Element1 · 7 zero-shot tasksaverage accuracy: 56.49 -> 56.69Improved
recover with 2.59M samples instead of 50k (ablation)LLaMA-5.4B · 7 zero-shot tasksaverage accuracy: 59.23 -> 62.36Improved
combine pruning with LLM.int8() quantizationLLaMA-5.4B · 7 zero-shot tasksaverage accuracy: 62.36 -> 62.30No clear effect
longer LoRA post-training steps (ablation)LLaMA-7B 20% Element1 · 7 zero-shot tasksaverage accuracy: 59.24 -> 59.68 -> 59.32Mixed

Later papers that compared with this method

3 later accepted papers report LLM-Pruner as a baseline in their results tables; 3 of them report a better number on most of the results they share with it.

Matched by method name in results tables read by AI. Authors' own tables tend to favour the newer method, and a comparison can be missed when a method is renamed.

Closest papers by meaning

Writing a paper in this area? scime draft is a free LaTeX editor with AI included, real-time co-authors and a check of your references and venue rules before you submit. It also imports OpenAI Prism projects.

Open scime draft

Get the next scime Dispatch and a monthly field update

Free. We send a confirmation link first; every email has a one-click unsubscribe. Privacy