speech_recognition/FunASR
1
0
Fork 0
mirror of https://github.com/modelscope/FunASR synced 2025-09-15 14:48:36 +08:00
Code Issues Actions Packages Projects Releases Wiki Activity
85cabd30a9
FunASR/egs/aishell/branchformer
History
jmwang66 18b6080701
update (#779)
2023-07-25 17:03:53 +08:00
..
conf update 2023-07-13 11:08:58 +08:00
local update 2023-07-05 20:36:27 +08:00
path.sh update 2023-07-05 20:36:27 +08:00
README.md update (#779) 2023-07-25 17:03:53 +08:00
run.sh update 2023-07-13 11:08:58 +08:00
utils update 2023-07-05 20:36:27 +08:00

README.md

Branchformer Result

Training Config

  • Feature info: using raw speech, extracting 80 dims fbank online, global cmvn, speed perturb(0.9, 1.0, 1.1), specaugment
  • Train info: lr 0.001, batch_size 10000, 4 gpu(Tesla V100), acc_grad 1, 180 epochs
  • Train config: conf/train_asr_branchformer.yaml
  • LM config: LM was not used

Results (CER)

  • Decode config: conf/decode_asr_transformer.yaml (ctc weight:0.4)
testset CER(%)
dev 4.15
test 4.51