Hi everyone, this is an ML Based Feedback Analysis Software, feel free to have a look, meanwhile, I'm writing up the docs.

I have changed a bunch of things for the docker-specific branch. [Other changes like changes to model or core logic are not mentioned here]

For instance all of it's dependencies are localized; meaning, before we were supposed to download whisper independently and add the path to voice-2-text.sh. However under the whisper/models folder, I have added a script that automatically downloads the whisper model for you.

Along with the model you also need the executable. I noticed the executable was ~1.5 megabytes so I pushed it to the repo as well under the whisper folder. So no headache of building it as such.

Of course, I have modified all the paths to be relative instead, this way would prove much better and intuitive. I call this methodology general relativity because of the usage of relative paths.

TODO

• The file_watcher already runs, however I am not so sure if the models are being loaded. this is bizarre to me as of right now but I haven't looked into it much anyway.

• Cowj needs to be taken care of, I didn't touch it.

• I don't remember what was next. Oh well. I would love if you would try building the image though.

If you would like to contact me, then feel free to drop a message here or on Linkedin.

Referals, Donations, Job Applications, Internships, Consultancy, anything at all, if you have anything in mind, I'm open to anything, anytime, anyday. THANK YOU FOR VISITING!