search

cywang97 / StreamingTransformer

Apache License 2.0
271 stars 42 forks source link

Joint CTC-triggered attention decoding algorithm code #9

Closed yjiangling closed 3 years ago

yjiangling commented 3 years ago

Hi, did you finish the implement of Joint CTC-triggered attention decoding algorithm?

I complete the Streaming Transformer model train and finished the train with TensorFlow, but when I tried to finish the Joint CTC-triggered attention decoding algorithm, I was stucked in the CTCPREFIX algorithm, is there any good ideal to complete it?

Papers: [1]. STREAMING AUTOMATIC SPEECH RECOGNITION WITH THE TRANSFORMER MODEL [2]. Streaming End-to-End Speech Recognition with Joint CTC-Attention Based Models

Gqwert123 commented 3 years ago

I also meet this problem. That means how can I use the trained model to inference the utt ?

cywang97 commented 3 years ago

Hi, did you finish the implement of Joint CTC-triggered attention decoding algorithm?

I complete the Streaming Transformer model train and finished the train with TensorFlow, but when I tried to finish the Joint CTC-triggered attention decoding algorithm, I was stucked in the CTCPREFIX algorithm, is there any good ideal to complete it?

Papers: [1]. STREAMING AUTOMATIC SPEECH RECOGNITION WITH THE TRANSFORMER MODEL [2]. Streaming End-to-End Speech Recognition with Joint CTC-Attention Based Models

The function prefix_recognize is joint CTC-triggered attention decoding algorithm. You can use decode.sh for decoding. I didn't understand your question clearly. Do you implement streaming Transformer with Tensorflow? And what do you mean by stucked in the CTCPREFIX algorithm?

yjiangling commented 3 years ago

Hi, did you finish the implement of Joint CTC-triggered attention decoding algorithm? I complete the Streaming Transformer model train and finished the train with TensorFlow, but when I tried to finish the Joint CTC-triggered attention decoding algorithm, I was stucked in the CTCPREFIX algorithm, is there any good ideal to complete it? Papers: [1]. STREAMING AUTOMATIC SPEECH RECOGNITION WITH THE TRANSFORMER MODEL [2]. Streaming End-to-End Speech Recognition with Joint CTC-Attention Based Models

The function prefix_recognize is joint CTC-triggered attention decoding algorithm. You can use decode.sh for decoding. I didn't understand your question clearly. Do you implement streaming Transformer with Tensorflow? And what do you mean by stucked in the CTCPREFIX algorithm?

Thanks a lot for your time. Yes, I implement Streaming Transformer with TensorFlow and I'm trying to implement the joint CTC-Triggered Attention algorithm with TensorFlow. But compared with PyTorch, It seems that it's more difficult to implement it whith TensorFlow.

yjiangling commented 3 years ago

Hi, did you finish the implement of Joint CTC-triggered attention decoding algorithm? I complete the Streaming Transformer model train and finished the train with TensorFlow, but when I tried to finish the Joint CTC-triggered attention decoding algorithm, I was stucked in the CTCPREFIX algorithm, is there any good ideal to complete it? Papers: [1]. STREAMING AUTOMATIC SPEECH RECOGNITION WITH THE TRANSFORMER MODEL [2]. Streaming End-to-End Speech Recognition with Joint CTC-Attention Based Models

The function prefix_recognize is joint CTC-triggered attention decoding algorithm. You can use decode.sh for decoding. I didn't understand your question clearly. Do you implement streaming Transformer with Tensorflow? And what do you mean by stucked in the CTCPREFIX algorithm?

Sorry, could you please tell me the location of the function prefix_recognize ? I really can not find it. Thanks a lot.

cywang97 commented 3 years ago

Hi, did you finish the implement of Joint CTC-triggered attention decoding algorithm? I complete the Streaming Transformer model train and finished the train with TensorFlow, but when I tried to finish the Joint CTC-triggered attention decoding algorithm, I was stucked in the CTCPREFIX algorithm, is there any good ideal to complete it? Papers: [1]. STREAMING AUTOMATIC SPEECH RECOGNITION WITH THE TRANSFORMER MODEL [2]. Streaming End-to-End Speech Recognition with Joint CTC-Attention Based Models

The function prefix_recognize is joint CTC-triggered attention decoding algorithm. You can use decode.sh for decoding. I didn't understand your question clearly. Do you implement streaming Transformer with Tensorflow? And what do you mean by stucked in the CTCPREFIX algorithm?

Sorry, could you please tell me the location of the function prefix_recognize ? I really can not find it. Thanks a lot.

It's in https://github.com/cywang97/StreamingTransformer/blob/master/espnet/nets/pytorch_backend/streaming_transformer.py

yjiangling commented 3 years ago

Hi, did you finish the implement of Joint CTC-triggered attention decoding algorithm? I complete the Streaming Transformer model train and finished the train with TensorFlow, but when I tried to finish the Joint CTC-triggered attention decoding algorithm, I was stucked in the CTCPREFIX algorithm, is there any good ideal to complete it? Papers: [1]. STREAMING AUTOMATIC SPEECH RECOGNITION WITH THE TRANSFORMER MODEL [2]. Streaming End-to-End Speech Recognition with Joint CTC-Attention Based Models

The function prefix_recognize is joint CTC-triggered attention decoding algorithm. You can use decode.sh for decoding. I didn't understand your question clearly. Do you implement streaming Transformer with Tensorflow? And what do you mean by stucked in the CTCPREFIX algorithm?

Sorry, could you please tell me the location of the function prefix_recognize ? I really can not find it. Thanks a lot.

It's in https://github.com/cywang97/StreamingTransformer/blob/master/espnet/nets/pytorch_backend/streaming_transformer.py

Ok，thank you very much for your help!