# machine-learning
- m
mega b
06/22/2021, 1:59 AMUpdate: its really fast at transcribing - m
mega b
06/22/2021, 2:00 AMand somewhat accurate to say the least - m
mega b
06/22/2021, 2:00 AMbetter than the options we have now - j
Justin
06/22/2021, 2:46 AMfixed it - m
mega b
06/22/2021, 2:46 AM? - m
mega b
06/22/2021, 2:46 AMoh yea i did - m
mega b
06/22/2021, 2:46 AMim not sharing it around right now - j
Justin
06/22/2021, 2:48 AMinput image - j
Justin
06/22/2021, 2:48 AMIt just morphed the features excluding the paint - j
Justin
06/22/2021, 2:49 AMI mean I did too - m
mega b
06/22/2021, 2:49 AM? - m
mega b
06/22/2021, 2:49 AMcool - u
Uberduck Agent
06/22/2021, 2:49 AMUpdated! - m
mega b
06/22/2021, 2:55 AMTranscription works like an absolute charm! - m
mega b
06/22/2021, 6:48 AMtest of the auto transcription colab - m
mega b
06/22/2021, 6:49 AMno editing of the transcription, automatically split audio into 12 second peices so it sounds pretty bad - m
mega b
06/22/2021, 6:49 AMbut its great 👍 - g
Gosmokeless28
06/22/2021, 7:38 AMNote to self: Epochs: 101 or 201 Batch size: 14 or 7 - j
jens
06/22/2021, 8:03 AMYeah I had that too when I tried the joker - j
Justin
06/22/2021, 8:03 AMnoice! - u
(Edd) bruh moment
06/22/2021, 1:00 PMstroke at the end - u
(Edd) bruh moment
06/22/2021, 1:37 PMI'm laughinggg - p
pi
06/22/2021, 2:36 PMthe cadence is very weird though - p
pi
06/22/2021, 2:36 PMdo you have an idea why? - p
PUMPKINEATER
06/22/2021, 2:37 PMDamn, I need that shit lol. - j
jmcb
06/22/2021, 3:51 PMhas anyone tried training a model with a transcript that says a bunch of things but then every audio file is just the same thing over and over again - j
jmcb
06/22/2021, 3:51 PMso that whatever you try to get it to synthesise it just says the same thign - j
Justin
06/22/2021, 3:51 PMOh nooo - j
Justin
06/22/2021, 3:51 PMDon't do that