SALMONN
SALMONN copied to clipboard
Published
20 hours ago •
bytedance
bytedance
SALMONN: Speech Audio Language Music Open Neural Network
Results
4
SALMONN issues
Sort by
recently updated
recently updated
newest added
Hello, I'm writing to inquire about the training data for the model, specifically for Task Level 3, which includes audio-based storytelling (Story) and speech audio co-reasoning (SAC) tasks. In your...
 ``` prompt = 'Please describe the audio.' prompt = [ cfg.config.model.prompt_template.format(" " + prompt.strip()) ] ``` 如果是 prompt = 'Please write down what your hear in the audio.' 则全都无法生成...
因为 Qwne-Audio 的 AAC 测试集用的的是 Clotho，你们用的是 AudioCaps，不知道在同一个测试集上两个模型哪个更好？Qwne-Audio 论文的 Figure 1 中说比 SALMONN 好，也不知道是否客观