多语言狮城：基于Qwen3-ASR平衡微调的新加坡高效多语种语音识别系统

摘要

我们推出Polyglot-Lion系列紧凑型多语言自动语音识别（ASR）模型，专为新加坡多语言场景设计，涵盖英语、华语、泰米尔语和马来语。该模型通过仅使用公开语音语料库对Qwen3-ASR-0.6B和Qwen3-ASR-1.7B进行微调获得，采用平衡采样策略确保各语言训练语句数量均等，并刻意省略语言标签条件机制，使模型能够从音频中隐式识别语言。在涵盖四种目标语言的12个基准测试中，Polyglot-Lion-1.7B实现了14.85的平均错误率，与体积大6倍的MERaLiON-2-10B-ASR模型（14.32）性能相当，而训练成本仅需单张RTX PRO 6000 GPU耗费81美元，远低于128卡基准模型的18,862美元。推理吞吐量达到0.10秒/样本，较MERaLiON的2.02秒/样本提升约20倍。这些结果表明，对中等规模预训练模型进行语言平衡的微调，能以极低成本获得可投入实际应用的多语言ASR系统，其成本远低于大型专业系统。

English

We present Polyglot-Lion, a family of compact multilingual automatic speech recognition (ASR) models tailored for the linguistic landscape of Singapore, covering English, Mandarin, Tamil, and Malay. Our models are obtained by fine-tuning Qwen3-ASR-0.6B and Qwen3-ASR-1.7B exclusively on publicly available speech corpora, using a balanced sampling strategy that equalizes the number of training utterances per language and deliberately omits language-tag conditioning so that the model learns to identify languages implicitly from audio. On 12 benchmarks spanning the four target languages, Polyglot-Lion-1.7B achieves an average error rate of 14.85, competitive with MERaLiON-2-10B-ASR (14.32) - a model 6x larger - while incurring a training cost of \81 on a single RTX PRO 6000 GPU compared to 18,862 for the 128-GPU baseline. Inference throughput is approximately 20x faster than MERaLiON at 0.10 s/sample versus 2.02 s/sample. These results demonstrate that linguistically balanced fine-tuning of moderate-scale pretrained models can yield deployment-ready multilingual ASR at a fraction of the cost of larger specialist systems.

多语言狮城：基于Qwen3-ASR平衡微调的新加坡高效多语种语音识别系统

Polyglot-Lion: Efficient Multilingual ASR for Singapore via Balanced Fine-Tuning of Qwen3-ASR

摘要

Support