Yuekai Zhang
> Another way is removing the output of padding zeros in the emission. You can compute the mapping from waveform length to emission frame length, and remove the frames by...
Also, for the empty result, if I do torch.softmax(emission), then check every column, I found that blank probs are always more than 0.95 for every frame. So my guess the...
Thanks. I see. The WAR could be always set wav_lens = torch.tensor(wavforms.shape[1]).repeat(batch_size). Then using @nateanl 's method removing from emission side by calculating real emission_lengths manually. Correct me if I...
> The engine plan file is not compatible with this version of TensorRT, expecting library version 9.2.0.5 got 9.3.0.1, please rebuild. @StephennFernandes Seems you build engines and run engines in...
> @yuekaizhang i got it working thanks a ton for your assistance. also. noticed that we cannot do inference for longer audio files. beyond 30s @StephennFernandes Since whisper could only...
> I got same issue. but it work properly. > > * Error env: > windows (ubuntu 20.04) > worksation ( intel xeon gold 6246 / rtx 3090 ) >...
> When i ran client.py, i got errror message: `tritonclient.utils.InferenceServerException: [StatusCode.INTERNAL] in ensemble 'whisper', Failed to process the request(s) for model instance 'scorer_0', message: AssertionError: ` How to fix? I...
> @csukuangfj The code for this client is in https://github.com/k2-fsa/sherpa/tree/master/triton/client/client.py > > After my test, if you do not use multi-process mode, that is, send data to the server with...
Have you tried our official recipe first? https://github.com/k2-fsa/sherpa/blob/master/triton/scripts/build_librispeech_pruned_transducer_stateless3_streaming.sh and https://github.com/k2-fsa/sherpa/blob/master/triton/scripts/build_wenetspeech_pruned_transducer_stateless5_streaming.sh. We currently support reworked conformer streaming ASR for en and zh.
Not yet. We will try to support this in a month. Or you're welcome to contribute a script or a model_repo to support it e.g. build_librispeech_zipformer2_streaming.sh