Today, I worked on a document outlining the approach I will take for building a client's application and on using the Diart library for implementing real-time diarization in my project. I tested it using its WebSocket Client CLI that allows you to use your microphone to connect to the WebSocket Audio Source in the application that utilizes the Diart library, and it worked amazingly well.
However, when trying to work with it directly in the client, I wasn't able to get the transcriptions in the server. The audio couldn't be processed and came out distorted. I couldn't understand why, as the encoding process was similar to the one done in the CLI file, but I've asked the creator of the library why this is happening.
Instead of creating a new issue, I commented on his answer to the issue that inspired the WebSocket support, which is where I took the code for the client audio streaming.
Hoping to get this working tomorrow! This is the only thing limiting me right now :)
Should also note that I've finished reading Atomic Habits 2 days ago! Need to listen to some summaries and revise the rules to start implementing them in my life.
Happy coding everyone :)
Top comments (0)