how to get the text description for music data

[CrossLee1]

Which LLM do you use to get the text description for the music data, qwen2-audio?

Could you provide the prompt?

Thanks~

[iris2c]

Which LLM do you use to get the text description for the music data, qwen2-audio?

Could you provide the prompt?

Thanks~

GPT-4o

[CrossLee1]

GPT-4o 可以处理音频数据么，还是说先用qwen-audio 对音频打caption，然后再用 gpt-4o refine？

[iris2c]

refine？

Our dataset includes tags, therefore we use GPT-4o to generate captions from tags.