LLaVA-MoLE: Sparse Mixture of LoRA Experts for Mitigating Data Conflicts in Instruction Finetuning MLLMs

• URL: http://arxiv.org/abs/2401.16160v2
• Date: Tue, 30 Jan 2024 15:44:58 GMT
• Title: LLaVA-MoLE: Sparse Mixture of LoRA Experts for Mitigating Data Conflicts in Instruction Finetuning MLLMs
• Authors: Shaoxiang Chen, Zequn Jie, Lin Ma
• Abstract summary: We propose an efficient Mixture of Experts (MoE) design for instruction finetuning MLLMs. Extensive experiments proved that LLaVA-MoLE effectively mitigates the data conflict issue when mixing multiple distinct instruction datasets. LLaVA-MoLE can even outperform the plain-LoRA baseline trained with twice the samples.
• Score: 29.96139552754377
• License: http://creativecommons.org/licenses/by/4.0/
• Abstract: Instruction finetuning on a variety of image-text instruction data is the key to obtaining a versatile Multimodal Large Language Model (MLLM), and different configurations of the instruction data can lead to finetuned models with different capabilities. However, we have discovered that data conflicts are inevitable when mixing instruction data from distinct domains, which can result in performance drops for tasks of a specific domain. To address this issue, we propose to apply an efficient Mixture of Experts (MoE) design, which is a sparse Mixture of LoRA Experts (MoLE) for instruction finetuning MLLMs. Within the Transformer layers, we extend the popular Low-Rank Adaption (LoRA) method by creating a set of LoRA experts specifically for the MLP layer, and route each token to the top-1 expert based on a routing function, allowing adaptive choices for tokens from different domains. Since the LoRA experts are sparsely activated, the training and inference cost are kept roughly constant compared to the original LoRA method. By replacing the plain-LoRA of LLaVA-1.5 with our MoE design, our final model is named LLaVA-MoLE. Extensive experiments proved that LLaVA-MoLE effectively mitigates the data conflict issue when mixing multiple distinct instruction datasets with various configurations, and achieves consistent performance gains over the strong plain-LoRA baselines. Most importantly, on the mixed datasets, LLaVA-MoLE can even outperform the plain-LoRA baseline trained with twice the samples.

Related papers

• Retrieval-Augmented Mixture of LoRA Experts for Uploadable Machine Learning [57.36978335727009]
  Low-Rank Adaptation (LoRA) offers an efficient way to fine-tune large language models (LLMs) In this paper, we propose a framework that adaptively retrieves and composes multiple LoRAs based on input prompts.
  arXiv  Detail & Related papers  (2024-06-24T05:24:41Z)
• Mixture of LoRA Experts [87.50120181861362]
  This paper introduces the Mixture of LoRA Experts (MoLE) approach, which harnesses hierarchical control and unfettered branch selection. The MoLE approach achieves superior LoRA fusion performance in comparison to direct arithmetic merging.
  arXiv  Detail & Related papers  (2024-04-21T11:59:53Z)
• Mixture-of-LoRAs: An Efficient Multitask Tuning for Large Language Models [7.966452497550907]
  We propose the Mixture-of-LoRAs (MoA) architecture for multi-task learning with large language models (LLMs) Multiple domain-specific LoRA modules can be aligned with the expert design principles observed in Mixture-of-Experts (MoE) Each LoRA model can be iteratively adapted to a new domain, allowing for quick domain-specific adaptation.
  arXiv  Detail & Related papers  (2024-03-06T03:33:48Z)
• Multimodal Instruction Tuning with Conditional Mixture of LoRA [54.65520214291653]
  This paper introduces a novel approach that integrates multimodal instruction tuning with Low-Rank Adaption (LoRA) It innovates upon LoRA by dynamically constructing low-rank adaptation matrices tailored to the unique demands of each input instance. Experimental results on various multimodal evaluation datasets indicate that MixLoRA not only outperforms the conventional LoRA with the same or even higher ranks.
  arXiv  Detail & Related papers  (2024-02-24T20:15:31Z)
• LoRA-Flow: Dynamic LoRA Fusion for Large Language Models in Generative Tasks [72.88244322513039]
  LoRA employs lightweight modules to customize large language models (LLMs) for each downstream task or domain. We propose LoRA-Flow, which utilizes dynamic weights to adjust the impact of different LoRAs. Experiments across six generative tasks demonstrate that our method consistently outperforms baselines with task-level fusion weights.
  arXiv  Detail & Related papers  (2024-02-18T04:41:25Z)
• LoraRetriever: Input-Aware LoRA Retrieval and Composition for Mixed Tasks in the Wild [76.67343971195267]
  Low-Rank Adaptation (LoRA) provides an efficient solution for fine-tuning large language models (LLM) LoraRetriever is a retrieve-then-compose framework that adaptively retrieves and composes multiple LoRAs according to the input prompts. Experimental results indicate that LoraRetriever consistently outperforms the baselines.
  arXiv  Detail & Related papers  (2024-02-15T15:02:46Z)
• MultiLoRA: Democratizing LoRA for Better Multi-Task Learning [20.750808913757396]
  LoRA achieves remarkable resource efficiency and comparable performance when adapting LLMs for specific tasks. LoRA is dominated by a small number of top singular vectors while fine-tuning decomposes into a set of less important unitary transforms. We propose MultiLoRA for better multi-task adaptation by reducing the dominance of top singular vectors observed in LoRA.
  arXiv  Detail & Related papers  (2023-11-20T02:59:18Z)
• Octavius: Mitigating Task Interference in MLLMs via LoRA-MoE [85.76186554492543]
  Large Language Models (LLMs) can extend their zero-shot capabilities to multimodal learning through instruction tuning. negative conflicts and interference may have a worse impact on performance. We propose a novel framework, called Octavius, for comprehensive studies and experimentation on multimodal learning with MLLMs.
  arXiv  Detail & Related papers  (2023-11-05T15:48:29Z)
• NOLA: Compressing LoRA using Linear Combination of Random Basis [22.76088132446952]
  We introduce NOLA, which overcomes the rank one lower bound present in LoRA. NOLA performs as well as LoRA models with much fewer number of parameters compared to LoRA with rank one, the best compression LoRA can archive.
  arXiv  Detail & Related papers  (2023-10-04T03:30:24Z)

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.