You signed in with another tab or window. Reload to refresh your session.You signed out in another tab or window. Reload to refresh your session.You switched accounts on another tab or window. Reload to refresh your session.Dismiss alert
A problem I observe is the models do not recognise more than two single-syllable words/characters in a sequence. The recogniser stop emitting tokens after the first two words or characters.
For example:
English: Speaking ONE ONE ONE is recognised only as ONE ONE. Speaking ONE over and over again produces no more tokens.
中文: Speaking 对对对对 is recognised only as 对对. Speaking more 对's produces no more tokens.
The problem goes away if single syllable words/characters are mixed with other single syllable words/characters. For example:
ONE TWO ONE TWO ONE TWO ONE
对的对的对的
The text was updated successfully, but these errors were encountered:
Given this happens across different datasets and training runs, is there something about the zipformer model architecture that intentionally limits output of many identical single-syllable word/character tokens?
Can it relate to the stateless transducer context length of 2? It doesn't
know how many copies of the symbol it has already emitted and guesses it
has already output them.
At one point we experimented with augmenting the context with a
repeat-count, but we didn't end up merging it.
I have been experimenting with the pre-trained zipformer streaming models:
A problem I observe is the models do not recognise more than two single-syllable words/characters in a sequence. The recogniser stop emitting tokens after the first two words or characters.
For example:
The problem goes away if single syllable words/characters are mixed with other single syllable words/characters. For example:
The text was updated successfully, but these errors were encountered: