Shayne Mei
Shayne Mei
Counting silence frames sounds like a good idea. Maybe we can make it more robust by adding a state machine with two states (one for silence, one for non-silence). Where...
Thanks for the suggestion. I have two other questions regarding using the endpoint information: 1. What is the plan for sending endpoint to the client? e.g. how will it be...
this seems to be obtaining the alignment in batches. Is it possible to obtain this alignment info in real-time while streaming?
@csukuangfj Just following up on this question: is it possible to obtain this alignment info in real-time, i.e. get alignment for each decoded segment while streaming?
Preferably in seconds. Either relative to the start of the segment or utterance should be fine
local batch decoding command: ```bash ./pruned_transducer_stateless5/decode.py \ --epoch 4 \ --avg 1 \ --simulate-streaming False \ --causal-convolution True \ --use-averaged-model False ``` local streaming decoding command: ```bash ./pruned_transducer_stateless5/decode.py \ --epoch...
Actually there isn't any s in sherpa based decoding for TEDLIUM
the utts in the two recogs.txt aren't in the same order, so I couldn't use `vimdiff`
> @shaynemei Did you use `decode-right-context=2` (the default value) in sherpa. If so, please try `decode-right-context=0`. We found that not all models can benefit from right context. @pkufool I reran...
@csukuangfj @danpovey @pkufool just following up on this issue. Is there anything else I should provide?