Reference-Free Post-Training of Open Large Language Models for Multilingual Machine Translation

• URL: http://arxiv.org/abs/2608.10812v2
• Date: Wed, 12 Aug 2026 10:06:17 GMT
• Title: Reference-Free Post-Training of Open Large Language Models for Multilingual Machine Translation
• Abstract summary: We study reference-free post-training for multilingual machine translation with open large language models.<n>Across 46 languages, the resulting models consistently improve translation quality over their SFT counterparts.<n>We further investigate on-policy distillation (OPD) and find that it reaches, but does not surpass, the quality frontier achieved by RL with checkpoint checkpoints.
• Score: 16.262185435621078
• License: http://arxiv.org/licenses/nonexclusive-distrib/1.0/
• Abstract: We study reference-free post-training for multilingual machine translation with open large language models. Starting from the supervised-finetuned MiLMMT-46-v0.1 models, we apply Group Relative Policy Optimization (GRPO) with a reward that averages two reference-free quality estimation models and is gated by language identification. We then linearly interpolate the supervised fine-tuning (SFT) and reinforcement learning (RL) model checkpoints to obtain MiLMMT-46-v1.0. Across 46 languages, the resulting models consistently improve translation quality over their SFT counterparts, outperform strong recent open baselines, including Seed-X, HY-MT2, and TranslateGemma, and achieve leading reference-free scores against evaluated proprietary systems such as Google Translate, Gemini 3 Pro, and GPT-5. We further investigate on-policy distillation (OPD) and find that it reaches, but does not surpass, the quality frontier achieved by RL with checkpoint interpolation. We release the models and code to facilitate future research.

Related papers

• Scaling Model and Data for Multilingual Machine Translation with Open Large Language Models [38.626788948945844]
  We present a study of open large language models (LLMs) for multilingual machine translation (MT) across a range of languages.<n>We develop MiLMMT-46, which achieves top-tier multilingual translation performance across 46 languages.
  arXiv  Detail & Related papers  (2026-02-12T13:56:02Z)
• Beyond English: Toward Inclusive and Scalable Multilingual Machine Translation with LLMs [41.53385639669034]
  Large language models have significantly advanced Multilingual Machine Translation (MMT), yet the broad language coverage, consistent translation quality, and English-centric bias remain open challenges.<n>We introduce textbfLMT, a suite of textbfLarge-scale textbfMultilingual textbfTranslation models centered on both Chinese and English.<n>LMT achieves SOTA performance among models of comparable language coverage, with our 4B model surpassing the much larger Aya-13B and NLLB-54B models by a substantial margin.
  arXiv  Detail & Related papers  (2025-11-10T11:54:53Z)
• Seed-X: Building Strong Multilingual Translation LLM with 7B Parameters [53.59868121093848]
  We introduce Seed-X, a family of open-source language models (LLMs) with 7B parameter size.<n>The base model is pre-trained on a diverse, high-quality dataset encompassing both monolingual and bilingual content across 28 languages.<n>The instruct model is then finetuned to translate by Chain-of-Thought (CoT) reasoning and further enhanced through reinforcement learning (RL) to achieve better generalization across diverse language pairs.
  arXiv  Detail & Related papers  (2025-07-18T03:19:43Z)
• Checklist Engineering Empowers Multilingual LLM Judges [12.64438771302935]
  Checklist Engineering based LLM-as-a-Judge (CE-Judge) is a training-free framework that uses checklist intuition for multilingual evaluation with an open-source model.<n>Our method generally surpasses the baselines and performs on par with the GPT-4o model.
  arXiv  Detail & Related papers  (2025-07-09T12:03:06Z)
• Trans-Zero: Self-Play Incentivizes Large Language Models for Multilingual Translation Without Parallel Data [64.4458540273004]
  We propose a self-play framework that leverages only monolingual data and the intrinsic multilingual knowledge of Large Language Models (LLMs)<n>Experiments demonstrate that this approach not only matches the performance of models trained on large-scale parallel data but also excels in non-English translation directions.
  arXiv  Detail & Related papers  (2025-04-20T16:20:30Z)
• M-Prometheus: A Suite of Open Multilingual LLM Judges [65.63858815169924]
  We introduce M-Prometheus, a suite of open-weight LLM judges that can provide both direct assessment and pairwise comparison feedback on multilingual outputs.<n>M-Prometheus models outperform state-of-the-art open LLM judges on multilingual reward benchmarks spanning more than 20 languages, as well as on literary machine translation (MT) evaluation covering 4 language pairs.
  arXiv  Detail & Related papers  (2025-04-07T11:37:26Z)
• Expanding Performance Boundaries of Open-Source Multimodal Models with Model, Data, and Test-Time Scaling [191.7830199016589]
  We introduce InternVL 2.5, an advanced multimodal large language model (MLLM) series that builds upon InternVL 2.0.<n>InternVL 2.5 exhibits competitive performance, rivaling leading commercial models such as GPT-4o and Claude-3.5-Sonnet.<n>We hope this model contributes to the open-source community by setting new standards for developing and applying multimodal AI systems.
  arXiv  Detail & Related papers  (2024-12-06T18:57:08Z)
• Towards Zero-Shot Multimodal Machine Translation [64.9141931372384]
  We propose a method to bypass the need for fully supervised data to train multimodal machine translation systems.<n>Our method, called ZeroMMT, consists in adapting a strong text-only machine translation (MT) model by training it on a mixture of two objectives.<n>To prove that our method generalizes to languages with no fully supervised training data available, we extend the CoMMuTE evaluation dataset to three new languages: Arabic, Russian and Chinese.
  arXiv  Detail & Related papers  (2024-07-18T15:20:31Z)
• Machine Translation for Ge'ez Language [0.0]
  Machine translation for low-resource languages such as Ge'ez faces challenges such as out-of-vocabulary words, domain mismatches, and lack of labeled training data. We develop a multilingual neural machine translation (MNMT) model based on languages relatedness. We also experiment with using GPT-3.5, a state-of-the-art LLM, for few-shot translation with fuzzy matches.
  arXiv  Detail & Related papers  (2023-11-24T14:55:23Z)
• Pre-training Multilingual Neural Machine Translation by Leveraging Alignment Information [72.2412707779571]
  mRASP is an approach to pre-train a universal multilingual neural machine translation model. We carry out experiments on 42 translation directions across a diverse setting, including low, medium, rich resource, and as well as transferring to exotic language pairs.
  arXiv  Detail & Related papers  (2020-10-07T03:57:54Z)

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.