Tagged articles
Parameter-Efficient Fine-Tuning
12 articles · Page 1 of 1
ETH Zurich postdoc Bingcong Li will join HKU as a tenure-track assistant professor, focusing on efficient scaling of large language models through architecture design and optimization, including width/depth scaling, LoRA fine-tuning, MoE efficiency, and zero-order optimization methods like DPZero and RefLoRA.
Architecture DesignEfficient TrainingLoRA
0 likes · 6 min read

A systematic study of reinforcement‑learning fine‑tuning for large language models reveals that RL gains are highly concentrated in a few middle Transformer layers, and training just one such layer can match or even exceed full‑parameter RL performance across multiple models, tasks, and algorithms.
Model TrainingParameter-Efficient Fine-Tuninglayer contribution
0 likes · 17 min read

Microsoft Research introduces SkillOpt, a method that freezes large‑model weights and instead trains a natural‑language skill document as the sole learnable parameter, using a rollout‑reflect‑edit‑gate loop, achieving optimal results across 52 benchmark‑model‑environment combinations and demonstrating strong transferability.
LLM AgentsParameter-Efficient Fine-TuningSkillOpt
0 likes · 9 min read

The SHINE hypernetwork can turn arbitrary text into LoRA parameters for a large language model with just one forward pass, internalizing the knowledge for multi‑turn dialogue, achieving efficiency and scaling comparable to in‑context methods while outperforming traditional fine‑tuning baselines.
LoRAParameter-Efficient Fine-Tuninghypernetwork
0 likes · 17 min read

The paper introduces CSMF, a cascaded selective‑mask fine‑tuning framework for dual‑tower embedding‑based retrieval that sequentially optimizes exposure, click and conversion goals, incorporates cumulative percentile pruning and adaptive margin loss, and achieves measurable gains in ad revenue and CTR without increasing model size or latency.
Parameter-Efficient Fine-Tuningcascaded selective maskdual-tower
0 likes · 15 min read

TinyLoRA, a Meta‑proposed method that fine‑tunes Qwen2.5‑7B with only 13 trainable parameters (26 bytes), achieves 91% accuracy on GSM8K under reinforcement learning, revealing that ultra‑low‑parameter RL can rival full‑scale supervised fine‑tuning.
GSM8KParameter-Efficient Fine-TuningQwen2.5
0 likes · 7 min read

This article surveys recent advances in neural network reprogrammability, presenting a unified framework that categorizes model reprogramming, prompt tuning, prompt instruction, and in‑context learning, highlights the shift from parameter‑centric to reprogrammability‑centric adaptation, and provides efficiency analyses, taxonomy, and practical case studies.
Neural Network ReprogrammabilityParameter-Efficient Fine-TuningPrompt Tuning
0 likes · 16 min read

This article explains how large language models are first pre‑trained on massive text corpora and then fine‑tuned with smaller, task‑specific datasets, covering the fine‑tuning process, types such as full‑parameter and PEFT, practical benefits, real‑world analogies, and key challenges like data quality and catastrophic forgetting.
AI Product ManagementParameter-Efficient Fine-Tuningfine-tuning
0 likes · 6 min read

This article explains the mathematical basis of LoRA, compares it with QLoRA, Prompt Tuning, Prefix Tuning and P‑tuning, shows practical PyTorch implementations, and provides mixed‑precision training tips so readers can choose the most memory‑efficient fine‑tuning method for their large language models.
LoRAParameter-Efficient Fine-TuningPrompt Tuning
0 likes · 17 min read

This article provides a comprehensive guide to customizing pre‑trained large language models through fine‑tuning techniques—including parameter‑efficient methods, data preparation, knowledge injection, and robust evaluation—offering practical steps, best practices, and domain‑specific considerations for achieving superior task performance.
LLM fine-tuningParameter-Efficient Fine-Tuningdata preparation
0 likes · 18 min read


AntTechOct 29, 2024 · Artificial Intelligence
This announcement introduces three Ant Group papers accepted at EMNLP 2024—Mixture‑of‑Modules for dynamic Transformer assembly, a plug‑and‑play visual reasoning framework built via data synthesis, and a layer‑wise importance‑aware efficient fine‑tuning method for large language models—highlighting their innovations and upcoming live presentations.
AI researchEMNLP 2024Parameter-Efficient Fine-Tuning
0 likes · 6 min read

This article introduces parameter‑efficient fine‑tuning (PEFT) techniques—including additive adapters, soft‑prompt methods, selection‑based BitFit, and re‑parameterization approaches like LoRA and AdaLoRA—explains their architectures, experimental results, and provides end‑to‑end code for fine‑tuning ChatGLM2‑6B on a Chinese medical QA dataset.
AdaLoRAAdapterLoRA
0 likes · 22 min read
