MoFO: Momentum-Filtered Optimizer for Mitigating Forgetting in LLM Fine-Tuning
- URL: http://arxiv.org/abs/2407.20999v2
- Date: Wed, 31 Jul 2024 17:56:03 GMT
- Title: MoFO: Momentum-Filtered Optimizer for Mitigating Forgetting in LLM Fine-Tuning
- Authors: Yupeng Chen, Senmiao Wang, Zhihang Lin, Zeyu Qin, Yushun Zhang, Tian Ding, Ruoyu Sun,
- Abstract summary: During fine-tuning, large language models (LLMs) may forget the knowledge acquired in the pre-training stage, leading to a decline in general capabilities.
We propose a new fine-tuning algorithm termed Momentum-Filtered algorithm (MoFO)
MoFO achieves similar fine-tuning performance while keeping parameters closer to the pre-trained model.
- Score: 11.174544614042984
- License: http://creativecommons.org/licenses/by/4.0/
- Abstract: Recently, large language models (LLMs) have demonstrated remarkable capabilities in a wide range of tasks. Typically, an LLM is pre-trained on large corpora and subsequently fine-tuned on task-specific datasets. However, during fine-tuning, LLMs may forget the knowledge acquired in the pre-training stage, leading to a decline in general capabilities. To address this issue, we propose a new fine-tuning algorithm termed Momentum-Filtered Optimizer (MoFO). The key idea of MoFO is to iteratively select and update the model parameters with the largest momentum magnitudes. Compared to full-parameter training, MoFO achieves similar fine-tuning performance while keeping parameters closer to the pre-trained model, thereby mitigating knowledge forgetting. Unlike most existing methods for forgetting mitigation, MoFO combines the following two advantages. First, MoFO does not require access to pre-training data. This makes MoFO particularly suitable for fine-tuning scenarios where pre-training data is unavailable, such as fine-tuning checkpoint-only open-source LLMs. Second, MoFO does not alter the original loss function. This could avoid impairing the model performance on the fine-tuning tasks. We validate MoFO through rigorous convergence analysis and extensive experiments, demonstrating its superiority over existing methods in mitigating forgetting and enhancing fine-tuning performance.
Related papers
- Predicting Emergent Capabilities by Finetuning [98.9684114851891]
We find that finetuning language models can shift the point in scaling at which emergence occurs towards less capable models.
We validate this approach using four standard NLP benchmarks.
We find that, in some cases, we can accurately predict whether models trained with up to 4x more compute have emerged.
arXiv Detail & Related papers (2024-11-25T01:48:09Z) - LoRA Unlearns More and Retains More (Student Abstract) [0.0]
PruneLoRA reduces the need for large-scale parameter updates by applying low-rank updates to the model.
We leverage LoRA to selectively modify a subset of the pruned model's parameters, thereby reducing the computational cost, memory requirements and improving the model's ability to retain performance on the remaining classes.
arXiv Detail & Related papers (2024-11-16T16:47:57Z) - Pruning Foundation Models for High Accuracy without Retraining [48.256389781305415]
It is challenging to deploy foundation models or large language models (LLMs) due to their massive parameters and computations.
Post-training pruning methods are proposed to prune LLMs in one-shot without retraining.
Our experiments demonstrate the superior performance of the proposed methods in comparison to SOTA baselines.
arXiv Detail & Related papers (2024-10-21T01:23:34Z) - Pandora's White-Box: Precise Training Data Detection and Extraction in Large Language Models [4.081098869497239]
We develop state-of-the-art privacy attacks against Large Language Models (LLMs)
New membership inference attacks (MIAs) against pretrained LLMs perform hundreds of times better than baseline attacks.
In fine-tuning, we find that a simple attack based on the ratio of the loss between the base and fine-tuned models is able to achieve near-perfect MIA performance.
arXiv Detail & Related papers (2024-02-26T20:41:50Z) - Dynamic Sparse No Training: Training-Free Fine-tuning for Sparse LLMs [67.38165028487242]
We introduce Dynamic Sparse No Training (DSnoT), a training-free fine-tuning approach to fine-tune large language models (LLMs)
Inspired by the Dynamic Sparse Training, DSnoT minimizes the reconstruction error between the dense and sparse LLMs.
Our paper offers fresh insights into how to fine-tune sparse LLMs in an efficient training-free manner and open new venues to scale the great potential of sparsity to LLMs.
arXiv Detail & Related papers (2023-10-13T07:38:52Z) - Distilling Step-by-Step! Outperforming Larger Language Models with Less
Training Data and Smaller Model Sizes [91.58845026796149]
We introduce Distilling step-by-step, a new mechanism that trains small models that outperform large language models.
We present three findings across 4 NLP benchmarks.
arXiv Detail & Related papers (2023-05-03T17:50:56Z) - nanoLM: an Affordable LLM Pre-training Benchmark via Accurate Loss Prediction across Scales [65.01417261415833]
We present an approach to predict the pre-training loss based on our observations that Maximal Update Parametrization (muP) enables accurate fitting of scaling laws.
With around 14% of the one-time pre-training cost, we can accurately forecast the loss for models up to 52B.
Our goal with nanoLM is to empower researchers with limited resources to reach meaningful conclusions on large models.
arXiv Detail & Related papers (2023-04-14T00:45:01Z) - Model-Agnostic Multitask Fine-tuning for Few-shot Vision-Language
Transfer Learning [59.38343286807997]
We propose Model-Agnostic Multitask Fine-tuning (MAMF) for vision-language models on unseen tasks.
Compared with model-agnostic meta-learning (MAML), MAMF discards the bi-level optimization and uses only first-order gradients.
We show that MAMF consistently outperforms the classical fine-tuning method for few-shot transfer learning on five benchmark datasets.
arXiv Detail & Related papers (2022-03-09T17:26:53Z)
This list is automatically generated from the titles and abstracts of the papers in this site.
This site does not guarantee the quality of this site (including all information) and is not responsible for any consequences.