Update README.md

README.md

@@ -11,9 +11,9 @@ pipeline_tag: automatic-speech-recognition
 
 ## ESPnet2 ASR model 
 
-### `espnet/aishell2_transducer`
+### `espnet/tedlium3`
 
-This model was trained by jctian98 using aishell2 recipe in [espnet](https://github.com/espnet/espnet/).
+This model was trained by Dongwei Jiang using tedlium3 recipe in [espnet](https://github.com/espnet/espnet/).
 
 ### Demo: How to use in ESPnet2
 
@@ -40,21 +40,21 @@ cd egs2/tedlium3/asr1
 
 |dataset|Snt|Wrd|Corr|Sub|Del|Ins|Err|S.Err|
 |---|---|---|---|---|---|---|---|---|
-|decode_lm_lm_train_lm_en_bpe500_valid.loss.ave_asr_model_valid.acc.ave/dev|466|14671|94.0|2.7|3.3|0.7|6.6|65.9|
+|decode_lm_lm_train_lm_en_bpe500_valid.loss.ave_asr_model_valid.acc.ave/dev|507|17783|93.2|3.2|3.5|1.0|7.8|68.6|
 |decode_lm_lm_train_lm_en_bpe500_valid.loss.ave_asr_model_valid.acc.ave/test|1155|27500|93.9|2.7|3.4|0.7|6.8|61.1|
 
 ### CER
 
 |dataset|Snt|Wrd|Corr|Sub|Del|Ins|Err|S.Err|
 |---|---|---|---|---|---|---|---|---|
-|decode_lm_lm_train_lm_en_bpe500_valid.loss.ave_asr_model_valid.acc.ave/dev|466|78259|96.6|0.6|2.8|0.6|4.0|65.9|
+|decode_lm_lm_train_lm_en_bpe500_valid.loss.ave_asr_model_valid.acc.ave/dev|507|95429|96.2|0.7|3.1|0.9|4.7|68.6|
 |decode_lm_lm_train_lm_en_bpe500_valid.loss.ave_asr_model_valid.acc.ave/test|1155|145066|96.6|0.6|2.8|0.6|4.1|61.1|
 
 ### TER
 
 |dataset|Snt|Wrd|Corr|Sub|Del|Ins|Err|S.Err|
 |---|---|---|---|---|---|---|---|---|
-|decode_lm_lm_train_lm_en_bpe500_valid.loss.ave_asr_model_valid.acc.ave/dev|466|29364|95.5|1.9|2.7|0.5|5.1|65.9|
+|decode_lm_lm_train_lm_en_bpe500_valid.loss.ave_asr_model_valid.acc.ave/dev|507|36002|95.0|2.2|2.8|0.9|5.8|68.6|
 |decode_lm_lm_train_lm_en_bpe500_valid.loss.ave_asr_model_valid.acc.ave/test|1155|54206|95.5|1.7|2.7|0.6|5.1|61.1|