fairseq
fairseq copied to clipboard
[MMS ASR] The doc for calc WER/CER can be unclear
As the doc says:
$ ls /path/to/manifest dev.tsv dev.wrd dev.ltr dev.uid
# dev.tsv each line contains
$ cat dev.ltr t h i s | i s | o n e | t h i s | i s | t w o |
$ cat dev.wrd this is one this is two
$ cat dev.uid audio_1 audio_2
There are several points that can lead to misunderstanding. Is the /path/to/audio_1 indicates a directory or a wav file? What does the number of samples mean?