Warning: Cannot resolve tags ['Ravdess_max_audio_size']

habla-liaa / ser-with-w2v2

Official implementation of INTERSPEECH 2021 paper 'Emotion Recognition from Speech Using Wav2vec 2.0 Embeddings'

126 stars 23 forks source link

Warning: Cannot resolve tags ['Ravdess_max_audio_size'] #7

Closed jiminbot20 closed 3 years ago

jiminbot20 commented 3 years ago

https://github.com/habla-liaa/ser-with-w2v2/blob/c9be8a9bc8c6c6969838e3e77c2fc8af10e27136/configs/datasets/ravdess.yaml#L1

when running shell file this warning occurs.

so 'Ravdess_max_audio_size' this variable needs to announce, I guess. unlikely to this code, in ser-with-w2v2/configs/datasets/iemocap_impro.yaml, 'IEMOCAP_max_audio_size' is written

https://github.com/habla-liaa/ser-with-w2v2/blob/c9be8a9bc8c6c6969838e3e77c2fc8af10e27136/configs/datasets/iemocap_impro.yaml#L10

mrpep commented 3 years ago

In the case of RAVDESS it is not needed to truncate the audios because the longest one is 6 seconds long, which is a lot less than 240k samples. In the other hand, some IEMOCAP audios are longer than 15 seconds (which corresponds to 240k samples).