scotfang

What I did eventually is I expose looper() inside the internal recorder and call it when the audio ends. On Tue, Aug 3, 2021 at 11:11 PM Sahil Malik ***@***.***>...

@madkote I believe the reason zero-ed out wav-data produce spurious transcripts such as "And", "yes", "oh", "um" is due to the fact that the ASR model was not trained on...

> One possible reason is that you did not set finish signal when your VAD system triggers. please see https://github.com/wenet-e2e/wenet/blob/main/runtime/core/websocket/websocket_server.cc#L59 Hi Robin yes that would be a solution and I...

Actually, maybe I will submit a PR for implementing a "flush" signal for the websocket server

@robin1001 Are you sure that the current code supports flushing ASR outputs when signaling "end" to the websocket with chunk_size > 0? https://github.com/wenet-e2e/wenet/blob/main/runtime/core/decoder/torch_asr_decoder.cc#L106 Here we set state to DecodeState::kEndFeats via...

I implemented a working version of a "flush" signal that does not end the decoding thread, will submit a PR soon

Thanks, then I will start training it this week! Let’s collaborate! On Wed, Oct 27, 2021 at 7:33 PM Binbin Zhang ***@***.***> wrote: > We do not have the bandwidth...

Hi, we've finished training a streaming WenetSpeech model, here are the statistics below. Would you like collaborate to make this model public? Thanks! --Scot Staff NLP Researcher at Didi Chuxing...

I can evaluate the other results too, will post next week On Fri, Jan 14, 2022 at 7:34 PM Binbin Zhang ***@***.***> wrote: > Great, Any other result about the...

I don't think I have time to evaluate the model on these other metrics, do you have a server that we can copy the model to, if you want to...