阿里千问开源语音识别模型Qwen3-ASR

Core Viewpoint - Alibaba's Qwen has officially open-sourced the Qwen3-ASR series of models, which are powerful speech recognition models designed to support multiple languages and dialects [1] Group 1: Model Details - The Qwen3-ASR series includes two comprehensive speech recognition models: Qwen3-ASR-1.7B and Qwen3-ASR-0.6B [1] - Additionally, the series features an innovative speech forced alignment model, Qwen3-ForcedAligner-0.6B [1] Group 2: Language Support - The Qwen3-ASR models support speech recognition and language identification for 52 languages and dialects [1]