MMNMT: Modularizing Multilingual Neural Machine Translation with Flexibly Assembled MoE and Dense Blocks (2023.emnlp-main)
Copied to clipboard
| Challenge: | Mixture-of-Experts (MoE) based sparse architectures are prone to overfitting on low-resource language translation. |
| Approach: | They propose a modularized MNMT framework that flexibly assembles dense and MoE-based sparse modules to achieve the best of both worlds. |
| Outcome: | The proposed framework outperforms existing models on low-resource language translation and zero-shot translation on benchmark datasets. |
Similar Papers
MoNMT: Modularly Leveraging Monolingual and Bilingual Knowledge for Neural Machine Translation (2024.lrec-main)
Copied to clipboard
| Challenge: | Existing models for multi-domain translation tasks only use monolingual data, whereas bilingual data is indispensable for improving the models. |
| Approach: | They propose a modular strategy that facilitates the cooperation of monolingual and bilingual knowledge in translation tasks by avoiding catastrophic forgetting. |
| Outcome: | The proposed model exhibits superior generalization and robustness over the conventional approach. |
Lego-MT: Learning Detachable Models for Massively Multilingual Machine Translation (2023.findings-acl)
Copied to clipboard
| Challenge: | Existing monolithic models for multilingual neural machine translation encounter parameter interference and inefficient inference for large models. |
| Approach: | They propose a detachable multi-way model that assigns each language to an individual branch . they use data from OPUS to build a translation benchmark covering 433 languages . |
| Outcome: | The proposed model outperforms existing models in OPUS and is faster than existing models. |
Modularized Multilingual NMT with Fine-grained Interlingua (2024.naacl-long)
Copied to clipboard
| Challenge: | Neural Machine Translation (MNMT) systems lack layer-sharing to generate interlingua features . however, layer-share structure does not guarantee explicit propagation of language-specific features to respective decoders. |
| Approach: | They propose to share top of language-specific encoder layers to enable interlingua features . their method demonstrates an improved average BLEU score by "+2.90" in En-to-Any directions . |
| Outcome: | The proposed approach improves the BLEU score by "+2.90" in En-to-Any directions and by "+1.06" in zero-shot translation. |
Multilingual Neural Machine Translation: Can Linguistic Hierarchies Help? (2021.findings-emnlp)
Copied to clipboard
| Challenge: | Multilingual Neural Machine Translation (MNMT) trains a single model that supports translation between multiple languages . transferring knowledge from a diverse set of languages degrades the translation performance due to negative transfer. |
| Approach: | They propose a hierarchical knowledge distillation approach to train multilingual models . they use typological features and phylogeny to overcome negative transfer issue . |
| Outcome: | The proposed approach avoids negative transfer effect by capitalising on language groups generated according to typological features and phylogeny of languages. |
Learn and Consolidate: Continual Adaptation for Zero-Shot and Multilingual Neural Machine Translation (2023.emnlp-main)
Copied to clipboard
| Challenge: | Existing multilingual neural machine translation models perform poorly on language pairs with no parallel corpus. |
| Approach: | They propose a two-stage approach that encourages original models to acquire language-agnostic multilingual representations from new data and preserves the model architecture without introducing parameters. |
| Outcome: | The proposed approach improves performance in translation directions where existing models are weak and mitigates degeneration in the well-performing translation directions, offering flexibility in the real-world scenario. |
ModularMoE: Fast LLM Customization with Parameter-Sharing Mixture-of-Experts for Low-Resource Settings (2026.findings-acl)
Copied to clipboard
Jiaxing Liu, Qi Qi, Haifeng Sun, Dunjun Li, Zirui Zhuang, Bo He, Xiang Yang, Cong Liu, Jianxin Liao, Jingyu Wang
| Challenge: | Large Language Models impose significant computational and storage burdens on personal devices . existing customization approaches incur excessive computational costs or lead to suboptimal performance . |
| Approach: | They propose a training framework that converts pre-trained LLMs into parameter-sharing MoE models for lightweight deployment. |
| Outcome: | The proposed training framework outperforms state-of-the-art training frameworks at the same sparsity level while delivering up to 2.71 inference speedup. |
Revisiting Modularized Multilingual NMT to Meet Industrial Demands (2020.emnlp-main)
Copied to clipboard
| Challenge: | Currently, the complete sharing of parameters for multilingual translation (1-1) is the most popular approach because of its compactness. |
| Approach: | They propose to use a multilingual neural machine translation model that only shares modules among the same languages as 1-1 to satisfy industrial requirements. |
| Outcome: | The proposed model can enjoy the benefits of multi-way training without the capacity bottleneck and low maintainability. |
Building Multilingual Machine Translation Systems That Serve Arbitrary XY Translations (2022.naacl-main)
Copied to clipboard
| Challenge: | Multilingual Neural Machine Translation (MNMT) systems are often limited to many-to-one directions and suffer from poor performance in one-to one directions. |
| Approach: | They propose to build multilingual machine translation systems that serve arbitrary X-Y directions while leveraging multilinguality with a two-stage training strategy of pretraining and finetuning. |
| Outcome: | The proposed system outperforms baseline bilingual models and pivot translation models in most directions without the need for architecture change or extra data collection. |
Unlocking Emergent Modularity in Large Language Models (2024.naacl-long)
Copied to clipboard
| Challenge: | Existing MNNs are explicit, with predefined modular architectures and individual modules expected to implement distinct functions. |
| Approach: | They propose to unlock emergent modularity in language models by fine-tuning them as Mixture-of-Experts (MoEs) EMoE is robust to various configurations and can scale up to Large Language Models . |
| Outcome: | The proposed models can be fine-tuned as Mixture-of-Expert (MoE) counterparts without introducing any extra parameters. |
mmT5: Modular Multilingual Pre-Training Solves Source Language Hallucinations (2023.findings-emnlp)
Copied to clipboard
| Challenge: | Recent large language models display surprising multilingual capabilities despite being pre-trained on English data. |
| Approach: | They propose a multilingual sequence-to-sequence model that disentangles language-specific information from language-agnostic information. |
| Outcome: | The proposed model outperforms existing models on representative natural language understanding and generation tasks in 40+ languages. |