Continually pretrained on the International Corpus of English (18 varieties,
~20M tokens), then adapted via the explicit thread: dialect-specific SFT on
Multi-VALUE-transformed en-AU preference data, followed by GRPO with the
target-variety preference pairs.
This model is a fine-tuned version of
jordanpainter/diallm-qwen-sft-aus, trained using
TRL.
1from transformers import pipeline
2
3question = "If you had a time machine, but could only go to the past or the future once and never return, which would you choose and why?"
4generator = pipeline("text-generation", model="surrey-nlp/diallm-qwen-grpo-aus", device="cuda")
5output = generator([{"role": "user", "content": question}], max_new_tokens=128, return_full_text=False)[0]
6print(output["generated_text"])
This model was trained with GRPO.
1@article{painter2026diallm,
2 title = {DiaLLM: An Investigation into the Robustness-Generation Gap in English Dialect Adaptation},
3 author = {Painter, Jordan and Srirag, Dipankar and Kappiyath, Adarsh and Kanojia, Diptesh and Joshi, Aditya and Yin, Lu},
4 year = {2026},
5 eprint = {2607.07669},
6 archivePrefix = {arXiv}
7}
1@article{shao2024deepseekmath,
2 title = {{DeepSeekMath: Pushing the Limits of Mathematical Reasoning in Open Language Models}},
3 author = {Zhihong Shao and Peiyi Wang and Qihao Zhu and Runxin Xu and Junxiao Song and Mingchuan Zhang and Y. K. Li and Y. Wu and Daya Guo},
4 year = 2024,
5 eprint = {arXiv:2402.03300}
6}
1@software{vonwerra2020trl,
2 title = {{TRL: Transformers Reinforcement Learning}},
3 author = {von Werra, Leandro and Belkada, Younes and Tunstall, Lewis and Beeching, Edward and Thrush, Tristan and Lambert, Nathan and Huang, Shengyi and Rasul, Kashif and Gallouédec, Quentin},
4 license = {Apache-2.0},
5 url = {https://github.com/huggingface/trl},
6 year = {2020}
7}
