parameter-efficient fine-tuning

Parameter-Efficient Fine-Tuning (PEFT) is a technique used to adapt pre-trained models to new tasks with minimal changes to the model's parameters. This approach is particularly useful in scenarios where computational resources are limited or when it is desirable to maintain the original model's performance on the initial task.

Papers

Recently Added Most Hyped Most Active Needs Verification Most Verified

Showing 691–700 of 935 papers

Title	Date	Tasks	Status	Hype
Federated Fine-tuning of Large Language Models under Heterogeneous Tasks and Client Resources	Feb 18, 2024	Federated Learningparameter-efficient fine-tuning	—Unverified	0
LoRETTA: Low-Rank Economic Tensor-Train Adaptation for Ultra-Low-Parameter Fine-Tuning of Large Language Models	Feb 18, 2024	Multi-Task Learningparameter-efficient fine-tuning	CodeCode Available	1
GNNavi: Navigating the Information Flow in Large Language Models by Graph Neural Network	Feb 18, 2024	Graph Neural NetworkIn-Context Learning	CodeCode Available	0
TuneTables: Context Optimization for Scalable Prior-Data Fitted Networks	Feb 17, 2024	FairnessIn-Context Learning	CodeCode Available	1
Generalizability of Mixture of Domain-Specific Adapters from the Lens of Signed Weight Directions and its Application to Effective Model Pruning	Feb 16, 2024	Computational Efficiencyparameter-efficient fine-tuning	—Unverified	0
Efficiency at Scale: Investigating the Performance of Diminutive Language Models in Clinical Tasks	Feb 16, 2024	Decision Makingparameter-efficient fine-tuning	—Unverified	0
UNDIAL: Self-Distillation with Adjusted Logits for Robust Unlearning in Large Language Models	Feb 15, 2024	Natural Language Understandingparameter-efficient fine-tuning	CodeCode Available	1
Quantified Task Misalignment to Inform PEFT: An Exploration of Domain Generalization and Catastrophic Forgetting in CLIP	Feb 14, 2024	Domain Generalizationparameter-efficient fine-tuning	—Unverified	0
DoRA: Weight-Decomposed Low-Rank Adaptation	Feb 14, 2024	parameter-efficient fine-tuning	CodeCode Available	4
An Embarrassingly Simple Approach for LLM with Strong ASR Capacity	Feb 13, 2024	Automatic Speech RecognitionAutomatic Speech Recognition (ASR)	CodeCode Available	2

Show:10 25 50

← PrevPage 70 of 94Next →

All datasets BoolQ HellaSwag WinoGrande

Benchmark Results

#	Model	Metric	Claimed	Verified	Status
1	LLaMA2-7b	Accuracy (% )	82.63	—	Unverified
2	LLaMA2-7b	Accuracy (% )	82.63	—	Unverified
3	LLaMA2-7b	Accuracy (% )	81.93	—	Unverified
4	LLaMA2-7b	Accuracy (% )	80.28	—	Unverified

#	Model	Metric	Claimed	Verified	Status
1	LLaMA2-7b	Accuracy (% )	76.68	—	Unverified
2	LLaMA2-7b	Accuracy (% )	76.67	—	Unverified
3	LLaMA2-7b	Accuracy (% )	76.27	—	Unverified

#	Model	Metric	Claimed	Verified	Status
1	LLaMA2-7b	Accuracy (% )	70.8	—	Unverified
2	LLaMA2-7b	Accuracy (% )	70.09	—	Unverified
3	LLaMA2-7b	Accuracy (% )	69.85	—	Unverified