Qwen3-ASR#
Qwen3-ASR is an audio speech recognition model from the Qwen3 family. Megatron Bridge supports checkpoint conversion and audio inference through the Qwen3-ASR bridge.
Supported Variants#
Qwen3-ASR-1.7B: https://huggingface.co/Qwen/Qwen3-ASR-1.7B
Architecture Notes#
Uses the
Qwen3ASRForConditionalGenerationarchitecture on the Hugging Face side.The Megatron implementation includes a Qwen3-ASR thinker model, transformer config, and ASR-specific RoPE support.
Example conversion and round-trip validation use
--trust-remote-code.
Examples#
For checkpoint import/export, round-trip validation, and audio inference commands, see the Qwen3-ASR examples README.