Elamur Research⌕
☰
home / papers / 2024 / Multilingual Arbitrage: Optimizing Data Pools to Accelerate Multilingual Progress

Multilingual Arbitrage: Optimizing Data Pools to Accelerate Multilingual Progress

Paper · 2024 · Ayomide Odumakinde, Daniel D'souza, Pat Verga et al.

Summary

Synthetic training data for LLMs usually comes from one "teacher" model. That approach can pass on the teacher's biases, contribute to model collapse, and fail in multilingual work, where no single model is strongest in every language. The authors propose multilingual arbitrage: each prompt is sent to whichever model in a pool of teachers is expected to answer it best, and the resulting completions are used to instruction fine-tune (IFT) a student model.

They compare three routing strategies across 15 languages using 9 multilingual teachers of roughly 8B parameters:

  • Fixed routing: a preset best model per language.
  • Reward-based routing: every teacher answers, and a Cohere reward model picks the best answer.
  • Learned routing: a router predicts the reward from the prompt alone.

Students are evaluated with GPT-4-Turbo as an LLM judge, on discriminative benchmarks, and through text statistics. Every arbitrage strategy beats every single teacher. Reward-based routing gives the largest gains, and the least-resourced languages benefit most. The work shows that drawing data strategically from a pool of models can outperform even the strongest individual teacher.

Key points

  • Setup: 15 languages and 9 teacher models (including Aya-23, Gemma-2, Llama-3). The headline win-rate comparison aggregates 7 languages, one of which is Ukrainian. Routing uses a proprietary Cohere reward model (May 2024), kept separate from the GPT-4-Turbo judge to avoid self-preference bias.

  • Generative win rates: Reward-based routing raises average win rates by 56.5%. The abstract frames this figure relative to the best single teacher. The contributions list instead reports 28.1% over the best single teacher.

  • Discriminative tasks: Reward-based routing improves on the base model by up to 3.27% (1.91% on average). Single teachers average only 0.98%.

  • Comparison with random routing:

    • Reward-based routing: +119.5%
    • Fixed routing: +76.8%
    • Learned routing: +40.6%

    Learned routing is far cheaper because it does not require completions from every teacher.

  • Text quality: Lexical diversity increases relative to single teachers (+6% with reward-based routing, +4.2% with learned routing). Output length also grows (+14.1% and +68.4% respectively).

  • Limitations: Safety was not evaluated, and only teachers of about 8B parameters were tested.

Our summary of the paper, in our own words — not its text. The work belongs to its authors; read it at the links above.

 Fields, topics and concepts

📚 Shelves 4The field of science this work belongs to (OpenAlex fields).Computer Science · 16,667 worksArtificial Intelligence · 16,194 worksArts and Humanities · 16,039 worksLanguage and Linguistics · 15,891 works

🧭 Topics 1Research subjects and languages the work is about.Multilingual & cross-lingual NLP · 828 works

Cite

Cite as Ayomide Odumakinde, Daniel D'souza, Pat Verga, Beyza Ermis and Sara Hooker. 2024. Multilingual Arbitrage: Optimizing Data Pools to Accelerate Multilingual Progress. arXiv:2408.14960.

The paper itself, not this page. BibTeX from the library's record of the work. DOI 10.48550/arxiv.2408.14960 · arXiv 2408.14960

@misc{odumakinde2024multilingual,
  title = {{Multilingual Arbitrage: Optimizing Data Pools to Accelerate Multilingual Progress}},
  author = {Ayomide Odumakinde, Daniel D'souza, Pat Verga, Beyza Ermis and Sara Hooker},
  year = {2024},
  doi = {10.48550/arxiv.2408.14960},
  eprint = {2408.14960},
  archivePrefix = {arXiv},
  url = {https://doi.org/10.48550/arxiv.2408.14960}
}

References

88 references, 35 of them with a card of ours. ↗ leads to the work, «find» searches for one we have no id for yet.

  1. Aakanksha, Arash Ahmadian, Beyza Ermis, Seraphina Goldfarb-Tarrant, Julia Kreutzer, Marzieh Fadaee, and Sara Hooker. The multilingual alignment prism: Aligning global and local preferences to reduce harm, 2024. URL https://arxiv.org/abs/2406.18682. ↗
  2. David Ifeoluwa Adelani, Jade Abbott, Graham Neubig, Daniel D’souza, Julia Kreutzer, Constantine Lignos, Chester Palen-Michel, Happy Buzaaba, Shruti Rijhwani, Sebastian Ruder, et al. Masakhaner: Named entity recognition for african languages. Transactions of the Association for Computational Linguist…
  3. Ahmed Alaa, Boris Van Breugel, Evgeny S Saveliev, and Mihaela van der Schaar. How faithful is your synthetic data? sample-level metrics for evaluating and auditing generative models. In International Conference on Machine Learning, pp. 290–306. PMLR, 2022. find
  4. Ateret Anaby-Tavor, Boaz Carmeli, Esther Goldbraich, Amir Kantor, George Kour, Segev Shlomov, Naama Tepper, and Naama Zwerdling. Not enough data? deep learning to the rescue!, 2019. URL https://arxiv.org/abs/1911.03118. — [arXiv:1911.03118](https://arxiv.org/abs/1911.03118) ↗
  5. Jonathan Anderson. Lix and rix: Variations on a little-known readability index. Journal of Reading, 26(6):490–496, 1983. ISSN 00224103. URL http://www.jstor.org/stable/40031755. ↗
  6. Jordi Armengol-Estapé et al. On the multilingual capabilities of very large-scale english language models. arXiv preprint arXiv:2108.13349, 2021. find
  7. Viraat Aryabumi, John Dang, Dwarak Talupuru, Saurabh Dash, David Cairuz, Hangyu Lin, Bharat Venkitesh, Madeline Smith, Kelly Marchisio, Sebastian Ruder, et al. Aya 23: Open weight releases to further multilingual progress. arXiv preprint arXiv:2405.15032, 2024.
  8. Hritik Bansal, John Dang, and Aditya Grover. Peering through preferences: Unraveling feedback acquisition for aligning large language models. arXiv preprint arXiv:2308.15812, 2023. — [arXiv:2308.15812](https://arxiv.org/abs/2308.15812) find
  9. Quentin Bertrand, Avishek Joey Bose, Alexandre Duplessis, Marco Jiralerspong, and Gauthier Gidel. On the stability of iterative retraining of generative models on their own data, 2024. find
  10. Martin Briesch, Dominik Sobania, and Franz Rothlauf. Large language models suffer from their own output: An analysis of the self-consuming training loop, 2023. find
  11. Alexander Bukharin and Tuo Zhao. Data diversity matters for robust instruction tuning. arXiv preprint arXiv:2311.14736, 2023. find
  12. Samuel Cahyawijaya, Holy Lovenia, Alham Fikri Aji, Genta Indra Winata, Bryan Wilie, Rahmad Mahendra, Christian Wibisono, Ade Romadhony, Karissa Vincentio, Fajri Koto, et al. Nusacrowd: Open source initiative for indonesian nlp resources. arXiv preprint arXiv:2212.09648, pp. 13745– 13818, July 2022.… find
  13. Lingjiao Chen, Matei Zaharia, and James Zou. Frugalgpt: How to use large language models while reducing cost and improving performance. arXiv preprint arXiv:2305.05176, 2023.
  14. Aakanksha Chowdhery et al. Palm: Scaling language modeling with pathways. arXiv preprint arXiv:2204.02311, 2022.
  15. Hyung Won Chung, Le Hou, Shayne Longpre, Barret Zoph, Yi Tay, William Fedus, Yunxuan Li, Xuezhi Wang, Mostafa Dehghani, Siddhartha Brahma, et al. Scaling instruction-finetuned language models. arXiv preprint arXiv:2210.11416, 2022. find
  16. Alexis Conneau, Guillaume Lample, Ruty Rinott, Adina Williams, Samuel R Bowman, Holger Schwenk, and Veselin Stoyanov. Xnli: Evaluating cross-lingual sentence representations. arXiv preprint arXiv:1809.05053, 2018. find
  17. Marta R Costa-jussà, James Cross, Onur Çelebi, Maha Elbayad, Kenneth Heafield, Kevin Heffernan, Elahe Kalbassi, Janice Lam, Daniel Licht, Jean Maillard, et al. No language left behind: Scaling human-centered machine translation. arXiv preprint arXiv:2207.04672, 2022. find
  18. Viet Dac Lai, Chien Van Nguyen, Nghia Trung Ngo, Thuat Nguyen, Franck Dernoncourt, Ryan A Rossi, and Thien Huu Nguyen. Okapi: Instruction-tuned large language models in multiple languages with reinforcement learning from human feedback. arXiv e-prints, pp. arXiv–2307, 2023.
  19. John Dang, Arash Ahmadian, Kelly Marchisio, Julia Kreutzer, Ahmet Üstün, and Sara Hooker. Rlhf can speak many languages: Unlocking multilingual preference optimization for llms, 2024. URL https://arxiv.org/abs/2407.02552. ↗
  20. Elvis Dohmatob, Yunzhen Feng, and Julia Kempe. Model collapse demystified: The case of regression, 2024a. find
  21. Elvis Dohmatob, Yunzhen Feng, Pu Yang, Francois Charton, and Julia Kempe. A tale of tails: Model collapse as a change of scaling laws. arXiv preprint arXiv:2402.07043, 2024b. — [arXiv:2402.07043](https://arxiv.org/abs/2402.07043) find
  22. Abhimanyu Dubey, Abhinav Jauhri, Abhinav Pandey, Abhishek Kadian, Ahmad Al-Dahle, Aiesha Letman, Akhil Mathur, Alan Schelten, Amy Yang, Angela Fan, et al. The llama 3 herd of models. arXiv preprint arXiv:2407.21783, 2024.
  23. Angela Fan, Shruti Bhosale, Holger Schwenk, Zhiyi Ma, Ahmed El-Kishky, Siddharth Goyal, Mandeep Baines, Onur Celebi, Guillaume Wenzek, Vishrav Chaudhary, et al. Beyond english-centric multilingual machine translation. Journal of Machine Learning Research, 22(107):1–48, 2021. find
  24. Yunzhen Feng, Elvis Dohmatob, Pu Yang, Francois Charton, and Julia Kempe. Beyond model collapse: Scaling up with synthesized data requires reinforcement. arXiv preprint arXiv:2406.07515, 2024. — [arXiv:2406.07515](https://arxiv.org/abs/2406.07515) find
  25. Jiahui Gao, Renjie Pi, Yong Lin, Hang Xu, Jiacheng Ye, Zhiyong Wu, Weizhong Zhang, Xiaodan Liang, Zhenguo Li, and Lingpeng Kong. Self-guided noise-free data generation for efficient zero-shot learning, 2023. URL https://arxiv.org/abs/2205.12679. — [arXiv:2205.12679](https://arxiv.org/abs/2205.12679) ↗
  26. Matthias Gerstgrasser, Rylan Schaeffer, Apratim Dey, Rafael Rafailov, Henry Sleight, John Hughes, Tomasz Korbak, Rajashree Agrawal, Dhruv Pai, Andrey Gromov, Daniel A. Roberts, Diyi Yang, David L. Donoho, and Sanmi Koyejo. Is model collapse inevitable? breaking the curse of recursion by accumulating… find
  27. Robbie Gunning. The technique of clear writing. 1968. URL https://api.semanticscholar.org/CorpusID:145838278. ↗
  28. Yanzhu Guo, Guokan Shang, Michalis Vazirgiannis, and Chloé Clavel. The curious decline of linguistic diversity: Training language models on synthetic text, 2024. find
  29. Lasse Hansen, Ludvig Renbo Olsen, and Kenneth Enevoldsen. Textdescriptives: A python package for calculating a large variety of metrics from text. arXiv preprint arXiv:2301.02057, 2023. — [arXiv:2301.02057](https://arxiv.org/abs/2301.02057) find
  30. Geoffrey Hinton, Oriol Vinyals, and Jeff Dean. Distilling the knowledge in a neural network, 2015. URL https://arxiv.org/abs/1503.02531. ↗
  31. Srinivasan Iyer, Xi Victoria Lin, Ramakanth Pasunuru, Todor Mihaylov, Daniel Simig, Ping Yu, Kurt Shuster, Tianlu Wang, Qing Liu, Punit Singh Koura, Xian Li, Brian O’Horo, Gabriel Pereyra, Jeff Wang, Christopher Dewan, Asli Celikyilmaz, Luke Zettlemoyer, and Ves Stoyanov. Opt-iml: Scaling language m… — OPT-IML: Scaling Language Model Instruction Meta Learning through the Lens of Generalization
  32. Dongfu Jiang, Xiang Ren, and Bill Yuchen Lin. Llm-blender: Ensembling large language models with pairwise ranking and generative fusion. arXiv preprint arXiv:2306.02561, 2023.
  33. Pratik Joshi, Sebastin Santy, Amar Budhiraja, Kalika Bali, and Monojit Choudhury. The state and fate of linguistic diversity and inclusion in the nlp world. arXiv preprint arXiv:2004.09095, 2020.
  34. Jean Kaddour and Qi Liu. Text data augmentation in low-resource settings via fine-tuning of large language models. arXiv preprint arXiv:2310.01119, 2023. — [arXiv:2310.01119](https://arxiv.org/abs/2310.01119) find
  35. Tannon Kew, Florian Schottmann, and Rico Sennrich. Turning english-centric llms into polyglots: How much multilinguality is needed?, 2023. URL https://arxiv.org/abs/2312.12683. ↗
  36. Guneet Singh Kohli, Shantipriya Parida, Sambit Sekhar, Samirit Saha, Nipun B Nair, Parul Agarwal, Sonal Khosla, Kusumlata Patiyal, and Debasish Dhal. Building a llama2-finetuned llm for odia language utilizing domain knowledge instruction set, 2023. URL https://arxiv.org/abs/ 2312.12624. ↗
  37. Alina Kramchaninova and Arne Defauw. Synthetic data generation for multilingual domain-adaptable question answering systems. In Proceedings of the 23rd Annual Conference of the European Association for Machine Translation, pp. 151–160, 2022. find
  38. Nathan Lambert, Valentina Pyatkin, Jacob Morrison, LJ Miranda, Bill Yuchen Lin, Khyathi Chandu, Nouha Dziri, Sachin Kumar, Tom Zick, Yejin Choi, et al. Rewardbench: Evaluating reward models for language modeling. arXiv preprint arXiv:2403.13787, 2024. find
  39. Haonan Li, Fajri Koto, Minghao Wu, Alham Fikri Aji, and Timothy Baldwin. Bactrian-x: Multilingual replicable instruction-following models with low-rank adaptation. arXiv, abs/2305.15011, 2023.
  40. Haoran Li, Qingxiu Dong, Zhengyang Tang, Chaojun Wang, Xingxing Zhang, Haoyang Huang, Shaohan Huang, Xiaolong Huang, Zeqiang Huang, Dongdong Zhang, et al. Synthetic data (almost) from scratch: Generalized instruction tuning for language models. arXiv preprint arXiv:2402.13064, 2024. find
  41. Xi Victoria Lin, Todor Mihaylov, Mikel Artetxe, Tianlu Wang, Shuohui Chen, Daniel Simig, Myle Ott, Naman Goyal, Shruti Bhosale, Jingfei Du, Ramakanth Pasunuru, Sam Shleifer, Punit Singh Koura, Vishrav Chaudhary, Brian O’Horo, Jeff Wang, Luke Zettlemoyer, Zornitsa Kozareva, Mona Diab, Veselin Stoyano… — Few-shot Learning with Multilingual Language Models
  42. Shayne Longpre, Le Hou, Tu Vu, Albert Webson, Hyung Won Chung, Yi Tay, Denny Zhou, Quoc V. Le, Barret Zoph, Jason Wei, and Adam Roberts. The flan collection: Designing data and methods for effective instruction tuning. arXiv, abs/2301.13688, 2023.
  43. Ilya Loshchilov and Frank Hutter. Decoupled weight decay regularization, 2019. URL https://arxiv.org/abs/1711.05101. ↗
  44. Keming Lu, Hongyi Yuan, Runji Lin, Junyang Lin, Zheng Yuan, Chang Zhou, and Jingren Zhou. Routing to the expert: Efficient reward-guided ensemble of large language models. In Kevin Duh, Helena Gomez, and Steven Bethard (eds.), Proceedings of the 2024 Conference of the North American Chapter of the A… find
  45. Haipeng Luo, Qingfeng Sun, Can Xu, Pu Zhao, Jianguang Lou, Chongyang Tao, Xiubo Geng, Qingwei Lin, Shifeng Chen, and Dongmei Zhang. Wizardmath: Empowering mathematical reasoning for large language models via reinforced evol-instruct. arXiv preprint arXiv:2308.09583, 2023a.
  46. Ziyang Luo, Can Xu, Pu Zhao, Qingfeng Sun, Xiubo Geng, Wenxiang Hu, Chongyang Tao, Jing Ma, Qingwei Lin, and Daxin Jiang. Wizardcoder: Empowering code large language models with evol-instruct. arXiv preprint arXiv:2306.08568, 2023b. find
  47. Lucie Charlotte Magister, Jonathan Mallinson, Jakub Adamek, Eric Malmi, and Aliaksei Severyn. Teaching small language models to reason, 2023.
  48. Philip M. McCarthy and Scott Jarvis. Mtld, vocd-d, and hd-d: A validation study of sophisticated approaches to lexical diversity assessment. Behavior Research Methods, 42:381–392, 2010. URL https://api.semanticscholar.org/CorpusID:42852342. ↗
  49. Sewon Min, Mike Lewis, Luke Zettlemoyer, and Hannaneh Hajishirzi. Metaicl: Learning to learn in context. arXiv preprint arXiv:2110.15943, pp. 2791–2809, July 2021. doi: 10.18653/v1/2022 .naacl-main.201. URL https://aclanthology.org/2022.naacl-main.201. ↗
  50. Jamshidbek Mirzakhalov, Anoop Babu, Duygu Ataman, Sherzod Kariev, Francis Tyers, Otabek Abduraufov, Mammad Hajili, Sardana Ivanova, Abror Khaytbaev, Antonio Laverghetta Jr, et al. A large-scale study of machine translation in turkic languages. In Proceedings of the 2021 Conference on Empirical Metho… find
  51. Swaroop Mishra, Daniel Khashabi, Chitta Baral, and Hannaneh Hajishirzi. Cross-task generalization via natural language crowdsourcing instructions. arXiv preprint arXiv:2104.08773, pp. 3470–3487, May 2021. doi: 10.18653/v1/2022.acl-long.244. URL https://aclanthology.org/2 022.acl-long.244. ↗
  52. Niklas Muennighoff, Qian Liu, Armel Zebaze, Qinkai Zheng, Binyuan Hui, Terry Yue Zhuo, Swayam Singh, Xiangru Tang, Leandro von Werra, and Shayne Longpre. Octopack: Instruction tuning code large language models. arXiv preprint arXiv:2308.07124, 2023a. find
  53. Niklas Muennighoff, Thomas Wang, Lintang Sutawika, Adam Roberts, Stella Biderman, Teven Le Scao, M Saiful Bari, Sheng Shen, Zheng Xin Yong, Hailey Schoelkopf, Xiangru Tang, Dragomir Radev, Alham Fikri Aji, Khalid Almubarak, Samuel Albanie, Zaid Alyafeai, Albert Webson, Edward Raff, and Colin Raffel.… — Crosslingual Generalization through Multitask Finetuning
  54. Long Ouyang, Jeffrey Wu, Xu Jiang, Diogo Almeida, Carroll Wainwright, Pamela Mishkin, Chong Zhang, Sandhini Agarwal, Katarina Slama, Alex Ray, et al. Training language models to follow instructions with human feedback. Advances in Neural Information Processing Systems, 35:27730– 27744, 2022.
  55. Edoardo Maria Ponti, Goran Glavaš, Olga Majewska, Qianchu Liu, Ivan Vulić, and Anna Korhonen. Xcopa: A multilingual dataset for causal commonsense reasoning. arXiv preprint arXiv:2005.00333, 2020.
  56. Victor Sanh, Albert Webson, Colin Raffel, Stephen H. Bach, Lintang Sutawika, Zaid Alyafeai, Antoine Chaffin, Arnaud Stiegler, Teven Le Scao, Arun Raja, et al. Multitask prompted training enables zero-shot task generalization. ICLR 2022, 2021. URL https://arxiv.org/abs/2110.0 8207. ↗
  57. Teven Le Scao, Angela Fan, Christopher Akiki, Ellie Pavlick, Suzana Ilić, Daniel Hesslow, Roman Castagné, Alexandra Sasha Luccioni, François Yvon, Matthias Gallé, et al. Bloom: A 176b-parameter open-access multilingual language model. arXiv preprint arXiv:2211.05100, 2022. find
  58. Lucas Shen. LexicalRichness: A small module to compute textual lexical richness, 2022. URL https://github.com/LSYS/lexicalrichness. ↗
  59. Luísa Shimabucoro, Sebastian Ruder, Julia Kreutzer, Marzieh Fadaee, and Sara Hooker. Llm see, llm do: Guiding data generation to target non-differentiable objectives, 2024. URL https://arxiv.org/abs/2407.01490. — [arXiv:2407.01490](https://arxiv.org/abs/2407.01490) ↗
  60. Oleh Shliazhko, Alena Fenogenova, Maria Tikhonova, Vladislav Mikhailov, Anastasia Kozlova, and Tatiana Shavrina. mgpt: Few-shot learners go multilingual. arXiv preprint arXiv:2204.07580, 2022.
  61. Tal Shnitzer, Anthony Ou, Mírian Silva, Kate Soule, Yuekai Sun, Justin Solomon, Neil Thompson, and Mikhail Yurochkin. Large language model routing with benchmark datasets. arXiv preprint arXiv:2309.15789, 2023.
  62. Ilia Shumailov, Zakhar Shumaylov, Yiren Zhao, Yarin Gal, Nicolas Papernot, and Ross Anderson. The curse of recursion: Training on generated data makes models forget. arXiv preprint arXiv:2305.17493, 2023.
  63. Ilia Shumailov, Zakhar Shumaylov, Yiren Zhao, Yarin Gal, Nicolas Papernot, and Ross Anderson. The curse of recursion: Training on generated data makes models forget, 2024.
  64. AI Singapore. Sea-lion (southeast asian languages in one network): A family of large language models for southeast asia. — SEA-LION: Southeast Asian Languages in One Network https://github.com/aisingapore/sealion, 2024. ↗
  65. Shivalika Singh, Freddie Vargus, Daniel Dsouza, Börje F. Karlsson, Abinaya Mahendiran, Wei-Yin Ko, Herumb Shandilya, Jay Patel, Deividas Mataciunas, Laura OMahony, Mike Zhang, Ramith Hettiarachchi, Joseph Wilson, Marina Machado, Luisa Souza Moura, Dominik Krzemiński, Hakimeh Fadaei, Irem Ergün, Ifeo… — Aya Dataset: An Open-Access Collection for Multilingual Instruction Tuning
  66. Rohan Taori, Ishaan Gulrajani, Tianyi Zhang, Yann Dubois, Xuechen Li, Carlos Guestrin, Percy Liang, and Tatsunori B. Hashimoto. Stanford alpaca: An instruction-following llama model. https://github.com/tatsu-lab/stanford_alpaca, 2023a. ↗
  67. Rohan Taori, Ishaan Gulrajani, Tianyi Zhang, Yann Dubois, Xuechen Li, Carlos Guestrin, Percy Liang, and Tatsunori B Hashimoto. Stanford alpaca: An instruction-following llama model. 2023b. find
  68. Gemma Team, Thomas Mesnard, Cassidy Hardin, Robert Dadashi, Surya Bhupatiraju, Shreya Pathak, Laurent Sifre, Morgane Rivière, Mihir Sanjay Kale, Juliette Love, et al. Gemma: Open models based on gemini research and technology. arXiv preprint arXiv:2403.08295, 2024. find
  69. NLLB Team, Marta R Costa-jussà, James Cross, Onur Çelebi, Maha Elbayad, Kenneth Heafield, Kevin Heffernan, Elahe Kalbassi, Janice Lam, Daniel Licht, et al. No language left behind: Scaling human-centered machine translation (2022). URL https://arxiv.org/abs/2207.04672, 2022. ↗
  70. Atula Tejaswi, Nilesh Gupta, and Eunsol Choi. Exploring design choices for building language-specific llms, 2024. URL https://arxiv.org/abs/2406.14670. ↗
  71. Jeanine Treffers-Daller, Patrick Parslow, and Shirley Williams. Back to Basics: How Measures of Lexical Diversity Can Help Discriminate between CEFR Levels. Applied Linguistics, 39(3):302– 327, 04 2016. ISSN 0142-6001. doi: 10.1093/applin/amw009. URL https://doi.org/10.1093/ applin/amw009. — [doi:10.1093/applin/amw009](https://doi.org/10.1093/applin/amw009) find
  72. Lewis Tunstall, Edward Beeching, Nathan Lambert, Nazneen Rajani, Kashif Rasul, Younes Belkada, Shengyi Huang, Leandro von Werra, Clémentine Fourrier, Nathan Habib, et al. Zephyr: Direct distillation of lm alignment. arXiv preprint arXiv:2310.16944, 2023. find
  73. Ahmet Üstün, Viraat Aryabumi, Zheng-Xin Yong, Wei-Yin Ko, Daniel D’souza, Gbemileke Onilude, Neel Bhandari, Shivalika Singh, Hui-Lee Ooi, Amr Kayid, et al. Aya model: An instruction finetuned open-access multilingual language model. arXiv preprint arXiv:2402.07827, 2024.
  74. Pat Verga, Sebastian Hofstatter, Sophia Althammer, Yixuan Su, Aleksandra Piktus, Arkady Arkhangorodsky, Minjie Xu, Naomi White, and Patrick Lewis. Replacing judges with juries: Evaluating llm generations with a panel of diverse models. arXiv preprint arXiv:2404.18796, 2024.
  75. Hongyi Wang, Felipe Maia Polo, Yuekai Sun, Souvik Kundu, Eric Xing, and Mikhail Yurochkin. Fusing models with complementary expertise. arXiv preprint arXiv:2310.01542, 2023a. — [arXiv:2310.01542](https://arxiv.org/abs/2310.01542) find
  76. Yizhong Wang, Yeganeh Kordi, Swaroop Mishra, Alisa Liu, Noah A Smith, Daniel Khashabi, and Hannaneh Hajishirzi. Self-instruct: Aligning language models with self-generated instructions. arXiv preprint arXiv:2212.10560, 2022. find
  77. Yizhong Wang, Hamish Ivison, Pradeep Dasigi, Jack Hessel, Tushar Khot, Khyathi Raghavi Chandu, David Wadden, Kelsey MacMillan, Noah A Smith, Iz Beltagy, et al. How far can camels go? exploring the state of instruction tuning on open resources. arXiv preprint arXiv:2306.04751, 2023b.
  78. Jason Wei, Maarten Bosma, Vincent Y Zhao, Kelvin Guu, Adams Wei Yu, Brian Lester, Nan Du, Andrew M Dai, and Quoc V Le. Finetuned language models are zero-shot learners. arXiv preprint arXiv:2109.01652, 2021.
  79. Jason Wei, Maarten Bosma, Vincent Y. Zhao, Kelvin Guu, Adams Wei Yu, Brian Lester, Nan Du, Andrew M. Dai, and Quoc V. Le. Finetuned language models are zero-shot learners, 2022. URL https://arxiv.org/abs/2109.01652. ↗
  80. Can Xu, Qingfeng Sun, Kai Zheng, Xiubo Geng, Pu Zhao, Jiazhan Feng, Chongyang Tao, and Daxin Jiang. Wizardlm: Empowering large language models to follow complex instructions. arXiv preprint arXiv:2304.12244, 2023.
  81. Linting Xue, Noah Constant, Adam Roberts, Mihir Kale, Rami Al-Rfou, Aditya Siddhant, Aditya Barua, and Colin Raffel. mt5: A massively multilingual pre-trained text-to-text transformer. arXiv preprint arXiv:2010.11934, 2020.
  82. Shweta Yadav. Assessing the efficacy of synthetic data for enhancing machine translation models in low resource domains. In International Conference on Big Data Analytics, pp. 122–132. Springer, 2023. find
  83. An Yang, Baosong Yang, Binyuan Hui, Bo Zheng, Bowen Yu, Chang Zhou, Chengpeng Li, Chengyuan Li, Dayiheng Liu, Fei Huang, Guanting Dong, Haoran Wei, Huan Lin, Jialong Tang, Jialin Wang, Jian Yang, Jianhong Tu, Jianwei Zhang, Jianxin Ma, Jianxin Yang, Jin Xu, Jingren Zhou, Jinze Bai, Jinzheng He, Juny… find
  84. Dylan Zhang, Justin Wang, and Francois Charton. Instruction diversity drives generalization to unseen tasks. arXiv preprint arXiv:2402.10891, 2024. — [arXiv:2402.10891](https://arxiv.org/abs/2402.10891) find
  85. Susan Zhang, Stephen Roller, Naman Goyal, Mikel Artetxe, Moya Chen, Shuohui Chen, Christopher Dewan, Mona Diab, Xian Li, Xi Victoria Lin, et al. Opt: Open pre-trained transformer language models. arXiv preprint arXiv:2205.01068, 2022.
  86. Zhihan Zhang, Shuohang Wang, Wenhao Yu, Yichong Xu, Dan Iter, Qingkai Zeng, Yang Liu, Chenguang Zhu, and Meng Jiang. Auto-instruct: Automatic instruction generation and ranking for black-box language models. arXiv preprint arXiv:2310.13127, 2023. find
  87. Chunting Zhou, Pengfei Liu, Puxin Xu, Srini Iyer, Jiao Sun, Yuning Mao, Xuezhe Ma, Avia Efrat, Ping Yu, Lili Yu, et al. Lima: Less is more for alignment. arXiv preprint arXiv:2305.11206, 2023. find
  88. Terry Yue Zhuo, Armel Zebaze, Nitchakarn Suppattarachai, Leandro von Werra, Harm de Vries, Qian Liu, and Niklas Muennighoff. Astraios: Parameter-efficient instruction tuning code large language models. arXiv preprint arXiv:2401.00788, 2024.

Updated

↑ back to top

This page in Markdown → https://md.research.elamur.ai/free/en/papers/odumakinde-2024-multilingual-arbitrage-optimizing-data-pools/