Lightweight Transducer Based on Frame-Level Criterion

Open paper
Benchmark Model Rank Results
speech-recognition-on-aishell-1Lightweight Transducer With LM#6Word Error Rate (WER): 4.03Params(M): 45.3
speech-recognition-on-aishell-1Lightweight Transducer#9Word Error Rate (WER): 4.31Params(M): 45.3

Rank counts only results with a code link.