Spaces:
Running
Does this works on chinese dataset ?
It looks like that this model has been trained on LSVT. So, does this work for Chinese text ?
Hello, the models were trained on LSVT, but only on the subset containing Latin characters. Theoretically, the models should be able to support any character set. However, this is untested for large character sets such as those in Chinese text. You may try it by defining a new charset in configs/charset/, changing charset_test in configs/main.yaml, and training the models on Chinese datasets.
Hello, the models were trained on LSVT, but only on the subset containing Latin characters. Theoretically, the models should be able to support any character set. However, this is untested for large character sets such as those in Chinese text. You may try it by defining a new charset in
configs/charset/, changingcharset_testinconfigs/main.yaml, and training the models on Chinese datasets.
your models is amazzzzzing