search

keunwoochoi / music-auto_tagging-keras

Music auto-tagging models and trained weights in keras/theano
MIT License
616 stars 142 forks source link

Fix quote mark in compact_cnn/prepare_audio.py #36

Open philkuz opened 6 years ago

keunwoochoi commented 6 years ago

Thanks for PR. On 499b366, seems like there are quire a lot of stuff that are not really required. I think requirements.txt should be minimal.

keunwoochoi commented 6 years ago

On e550412, thanks to spotting the error! But because this PR is mixed with the requirements.txt, I fixed the problem by myself by simply editting the code (then realised maybe it's not the best code of conduct to use git.. please understand :) ). So, how about we focus on getting a nice requirements.txt?

philkuz commented 6 years ago

whoops - didn't realize I had added requirements.txt to this - that was more for me haha. I'll close this for now since you already added the change. Also no worries about making the change w/o PR, I just want everything to work out of the box!

Btw working to port this over to the newest kapre and Keras versions. Is there anything you changed between the old Kapre and the most updated Kapre version? I'm able to replicate the results in main.py, but worried that I might be getting lucky with this result.

Any help would be appreciated!