chinese-speech-rec / README.md
wbbbbb's picture
Update README.md
7bc1f10
|
raw
history blame
152 Bytes
metadata
license: apache-2.0

使用seq-seq模型

encoder_id = "wbbbbb/wav2vec2-large-chinese-zh-cn"

decoder_id = "IDEA-CCNL/Randeng-BART-139M

wer=68.3