Benchmark
FLEURS
language
speech-to-text
audio
多语言
Few-shot Learning Evaluation of Universal Representations of Speech - a parallel speech dataset in 102 languages built on FLoRes-101 with approximately 12 hours of speech supervision per language for tasks including ASR, speech language identification, translation and retrieval
语言EN
满分100
参评模型5
模型排名
| 名次 | 模型 | 机构 | 分数 | 来源 |
|---|---|---|---|---|
| 1 | Gemini 1.5 Flash 8B | 86.4 | 来源 ↗ | |
| 2 | Gemini 1.5 Flash | 9.6 | 来源 ↗ | |
| 3 | Gemini 1.5 Pro | 6.7 | 来源 ↗ | |
| 4 | Gemini 1.0 Pro | 6.4 | 来源 ↗ | |
| 5 | Qwen2.5-Omni-7B | Alibaba Cloud / Qwen Team | 4.1 | 来源 ↗ |