Comparing Parameter Efficient Finetuning Techniques (PEFT) using Datamodels
Name
chamdal_harshalc_MEng_eecs_2024_thesis.pdf
Description
Thesis PDF
Size
669.03 KB
Format
Adobe PDF
Checksum (MD5)
8b53867d61b406cc417d37fb416777ac
Author(s)
Chamdal, Harshal
Advisor(s)
Mądry, Aleksander
Date Issued
May 2024
Publisher
Massachusetts Institute of Technology
Abstract
Advances in machine learning, particularly through algorithmic innovations and large datasets, have led to models with hundreds of billions of parameters. Deploying these models is challenging and costly, especially due to the extensive finetuning required. Parameter-efficient finetuning techniques (PEFT) have been proposed to address this issue by significantly reducing the number of trainable parameters, achieving comparable results to full-parameter finetuning. Despite widespread adoption, PEFT methods are often used interchangeably without considering their qualitative differences and performance under various data distributions. This thesis extensively compares three PEFT methods: LoRA, BitFit, and (IA)³, using the ModelDiff framework to identify and apply data interventions. Our analysis reveals that the performance of these methods varies widely with different interventions, with BitFit showing the most variance, while LoRA and (IA)³ demonstrate greater resilience. This study informs the selection and optimization of PEFT techniques based on specific NLP task requirements, balancing performance, computational efficiency, and robustness to text variations.
MIT Department
Massachusetts Institute of Technology. Department of Electrical Engineering and Computer Science
Terms of Use
In Copyright - Educational Use Permitted
Copyright retained by author(s)
Persistent DSpace Link
