# machine-learning
- j
JacobLenstar
08/13/2021, 4:48 PMi'm trying to train Saddam Hussen from south park but he keeps turning out terrible - i
iGerman
08/13/2021, 4:49 PMhere we go, papanomaly model - i
iGerman
08/13/2021, 4:52 PMfor anyone wondering how i transcribed 200+ audio files in under 1.5 hours - i
iGerman
08/13/2021, 4:53 PMi used this to get the "rough" transcriptions https://github.com/raryelcostasouza/pyTranscriber/releases/tag/v1.5-stable - i
iGerman
08/13/2021, 4:53 PMthen i used this to time them better and correct mistakes and add punctuation (not that hard honestly) https://github.com/SubtitleEdit/subtitleedit/releases - i
iGerman
08/13/2021, 4:54 PMthen i used `@smoke oui’d#0001`'s https://github.com/HexCodeFFF/aidatasetfromvideo to put convert the video with subtitles into a dataset - i
iGerman
08/13/2021, 4:54 PMand then i regex-added
to the beginning of every line of list.txt so that the notebook understands itwavs/ - i
iGerman
08/13/2021, 4:54 PMand then just train like usual - i
iGerman
08/13/2021, 5:00 PM10 minutes of data, wish me luck haha - h
hecko
08/13/2021, 5:24 PMi mean i could see myself transcribing 10 minutes of data in that time clipping though not so much - a
AbeFroman
08/13/2021, 5:34 PMLast question now that my training is complete: where do i find the drive id? - a
AbeFroman
08/13/2021, 5:35 PMOr what file/folder do I want the drive id of? - h
hecko
08/13/2021, 5:39 PMin your google drive you should find a file called whatever right-click it,
, set toget link
then from the url above that select the gibberish between two slashes e.g.anyone with the link1-6brLs0jbejLxzu6jg2ifb9XVVQN2eN5 - a
AbeFroman
08/13/2021, 5:39 PMWhich file? - a
AbeFroman
08/13/2021, 5:39 PMthere are a lot in the /output/ - a
AbeFroman
08/13/2021, 5:40 PMthe .zip? - i
iGerman
08/13/2021, 5:41 PM1.5 hours from finding vod link to training model basically - h
hecko
08/13/2021, 5:41 PMoh you did talknet? yeah the zip - a
AbeFroman
08/13/2021, 5:41 PMGreat! - a
AbeFroman
08/13/2021, 5:41 PMHow come I can't upload refrence audio? All it says is that there are none - h
hecko
08/13/2021, 5:42 PMleft sidebar, 📁 icon - h
hecko
08/13/2021, 5:42 PMyou need to upload them there and rescan - a
AbeFroman
08/13/2021, 5:42 PMoops - a
AbeFroman
08/13/2021, 5:42 PMduh - i
iGerman
08/13/2021, 5:42 PMany way to synthesize vocals on your own pc? i wanna test my model while it's training on colab but i cant switch notebooks since i only get one - h
hecko
08/13/2021, 5:44 PM - i
iGerman
08/13/2021, 5:44 PMseems like a lot of storage and hassle - h
hecko
08/13/2021, 5:44 PMyeah - i
iGerman
08/13/2021, 5:45 PMi checked, this doesn't work unfortunately https://github.com/lokkelvin2/tacotron2-tts-GUI - h
hecko
08/13/2021, 5:45 PMplus i doubt it'd even work on windows