i want to send results of colour your caption to client through websocket i don't know how.

[jarvisx17]

# Chain of operations to apply on the stream of microphone audio
source.stream.pipe(
    # Format audio stream to sliding windows of 5s with a step of `500ms`
    dops.rearrange_audio_stream(
        config.duration, config.step, config.sample_rate
    ),
    # Wait until a batch is full
    # The output is a list of audio chunks
    ops.buffer_with_count(count=batch_size),
    # Obtain diarization prediction
    # The output is a list of pairs `(diarization, audio chunk)`
    `ops.map(dia),`
    # Concatenate 500ms predictions/chunks to form a single 2s chunk
    ops.map(concat),
    # Ignore this chunk if it does not contain speech
    ops.filter(lambda ann_wav: ann_wav[0].get_timeline().duration() > 0),
    # Obtain speaker-aware transcriptions
    # The output is a list of pairs `(speaker: int, caption: str)`
    ops.starmap(asr),
    # Color transcriptions according to the speaker
    # The output is plain text with color references for rich
    ops.map(colorize_transcription),
).subscribe(
    on_next=rich.print,  # print colored text
    on_error=lambda _: traceback.print_exc()  # print stacktrace if error
)

print("Listening...")
# source.send()
source.read()

[jarvisx17]

in source.send() which message parameter I should send so that i vcan get the transcription?

[juanmc2005]

Hi @jarvisx17, keep in mind that source.send() is synchronous, while the rx pipeline is asynchronous. This means that you need to "react" with source.send() whenever a new transcription block is available. This can be done in the on_next function of subscribe. In my example I used rich.print but you could also use source.send.

Btw, this is also something I have been working on. Check out the SpeakerAwareTranscription pipeline class in PR #147. You can install from branch feat/spk-asr to try it out (currently a work in progress).

I'm planning to release this feature as part of v0.8. Keep posted!

[jarvisx17]

Can you help me how i can install that specific branch and use it with diart.client

[juanmc2005]

Sure:

# Create environment
conda create -n diart python=3.8
conda activate diart

# Install dependencies
conda install portaudio pysoundfile ffmpeg -c conda-forge

# Clone repo and move to branch
git clone https://github.com/juanmc2005/diart.git
cd diart
git fetch
git checkout feat/spk-asr

# Install from source
pip install .

[juanmc2005]

Then you can run:

diart.serve 0.0.0.0 --port 7007 --pipeline SpeakerAwareTranscription --whisper small --language en

and in another console/machine:

diart.client microphone --host <server-address> --port 7007

You should see speaker-labeled transcriptions printed by the client (e.g. "[speaker0] Hello [speaker1] Hi!")

In python, you can import the pipeline as:

from diart.pipelines import SpeakerAwareTranscription, SpeakerAwareTranscriptionConfig