用小模型实现新加坡多语言语音识别，成本低效果好。
Polyglot-Lion: Efficient Multilingual ASR for Singapore via Balanced Fine-Tuning of Qwen3-ASR
- 用均衡语料微调小模型，不依赖语言标签自动识别语种。
- 1.7B模型平均错误率14.85，仅用81美元训练成本。
- 推理速度比大模型快20倍，适合资源有限场景。
我们提出Polyglot-Lion，一套针对新加坡多语言环境（英语、华语、泰米尔语、马来语）的紧凑型多语言语音识别模型。通过在公开语音语料上对Qwen3-ASR-0.6B和Qwen3-ASR-1.7B进行微调，采用每语言等量采样策略，并去除语言标签条件，使模型从音频中隐式学习语言识别能力。在覆盖四种目标语言的12个基准上，Polyglot-Lion-1.7B平均错误率为14.85，媲美6倍更大的MERaLiON-2-10B-ASR（14.32），但训练成本仅为81美元（单张RTX PRO 6000 GPU），相较128卡基线的18,862美元大幅降低。推理吞吐量达0.10秒/样本，较MERaLiON的2.02秒/样本提升约20倍。结果表明，对中等规模预训练模型进行语言均衡微调，可低成本实现可部署的多语言语音识别。
原文摘要 · Abstract (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.
Thank you to arXiv for use of its open access interoperability. PaperDance 不是 arXiv 官方产品；中文卡片由大模型生成，请以原文为准。