Try to extract the best open source speech recognition model parameters and transplant them to wenet wenet 中已经支持 此项目废弃 TODO Paraformer, wenet-e2e/wenet#2067