Qwen2.5-Omni fine-tuning use_audio_in_video #5202
Description
In Qwen2.5-Omni, use_audio_in_video can be enabled during inference. When fine-tuning in ms-swift, will this use_audio_in_video be enabled when using video data?
{{ message }}
In Qwen2.5-Omni, use_audio_in_video can be enabled during inference. When fine-tuning in ms-swift, will this use_audio_in_video be enabled when using video data?