paper-with-me

Papers

Constrained Edge AI Deployment: Fine-Tuning vs Distillation for LLM Compression

2025-05-13 · Jacob Sander, David Moe, Achraf Cohen, Brent Venable, Venkat Dasari, Brian Jalaian

Modern foundational models are often compressed via a combination of structured pruning and re-training to meet the strict compute, memory, and connectivity constraints of edge deployments. While state-of-the-art pruning schemes target the entire Transformer, we adopt a simple, layer-wise L2-norm pruning on only the MLP blocks as a fixed baseline. Our focus is not on achieving maximal compression, but on isolating the impact of the re-training loss function: (i) Fine-tuning with Cross- Entropy (L2PFT), which requires labeled data, versus (ii) Self-Distillation with KL-divergence, which leverages only teacher logits (no labels) (L2PSD). We evaluate both pipelines on the OLMo2- 7B-SFT model for CommonsenseQA suitable for intermittent or denied connectivity scenarios typical of edge networks. Under identical pruning schedules, KL-based distillation matches or exceeds CE fine-tuning in test accuracy, demonstrating that, even with a basic MLP-only pruning, the choice of loss function materially affects compressed model recovery in resource-constrained environments.

📄 PDF Abstract BibTeX arXiv:2505.18166

Code (0)

등록된 구현이 없습니다.

Methods 이 논문이 사용한 방법론

Attention 설명 없음
Linear Layer A Linear Layer is a projection $\mathbf{XW + b}$.
Layer Normalization Unlike batch normalization, Layer Normalization directly estimates the normalization statistics from the summed inputs…
BPE Byte Pair Encoding, or BPE, is a subword segmentation algorithm that encodes rare and unknown words as sequences of subword units. The intuition is that various word…
Residual Connection 설명 없음
Dense Connections Dense Connections, or Fully Connected Connections, are a type of layer in a deep neural network that use a linear operation where every input is connected to every output…
ADOPT Please enter a description about the method here
Pruning 설명 없음

Similar Papers 제목 키워드 기반

Advancing Model Refinement: Muon-Optimized Distillation and Quantization for LLM Deployment

2026-01-14 · Jacob Sander, Brian Jalaian, Venkat R. Dasari arxiv

Large Language Models (LLMs) enable advanced natural language processing but face deployment challenges on resource-constrained edge devices due to high computational, memory, and energy demands. Optimizing these models …

Hyperparameter OptimizationKnowledge Distillation

Efficient Fine-Tuning and Concept Suppression for Pruned Diffusion Models

2024-12-19 · CVPR 2025 1 · Reza Shirkavand, Peiran Yu, Shangqian Gao, Gowthami Somepalli 외

Recent advances in diffusion generative models have yielded remarkable progress. While the quality of generated content continues to improve, these models have grown considerably in size and complexity. This increasing c…

Bilevel OptimizationKnowledge DistillationStyle Transfer

Winning Big with Small Models: Knowledge Distillation vs. Self-Training for Reducing Hallucination in QA Agents

2025-02-26 · Ashley Lewis, Michael White, Jing Liu, Toshiaki Koike-Akino 외

The deployment of Large Language Models (LLMs) in customer support is constrained by hallucination-generating false information-and the high cost of proprietary models. To address these challenges, we propose a retrieval…

HallucinationKnowledge DistillationQuestion AnsweringRetrieval

Distilling Safe LLM Systems via Soft Prompts for On Device Settings

2026-06-08 · Motasem Alfarra, Cristina Pinneri, Dana Kianfar, Mohammed Almousa 외 arxiv

Deploying safe large language models (LLMs) on resource-constrained edge devices presents a critical challenge: while dual-model systems combining LLMs with guard models provide effective safety guarantees, their substan…

parameter-efficient fine-tuning

One-Step Knowledge Distillation and Fine-Tuning in Using Large Pre-Trained Self-Supervised Learning Models for Speaker Verification

2023-05-27 · Jungwoo Heo, Chan-yeong Lim, Ju-ho Kim, Hyun-seo Shin 외

The application of speech self-supervised learning (SSL) models has achieved remarkable performance in speaker verification (SV). However, there is a computational cost hurdle in employing them, which makes development a…

Knowledge DistillationSelf-Supervised LearningSpeaker Verification