# machine-learning
- p
pi
06/02/2021, 8:28 PMan I only said that the alignment graph only shows alignment - p
pi
06/02/2021, 8:28 PM🤷 - s
SidPlays_144p
06/02/2021, 8:28 PMok - s
SuperLuig
06/02/2021, 8:40 PMi swear to christ if this voice somehow turns out decent - s
SuperLuig
06/02/2021, 8:43 PM122 epochs, waiting for the synthesis notebook to download stuff and i'll do a test - s
SuperLuig
06/02/2021, 8:45 PMbro WTF - s
SuperLuig
06/02/2021, 8:45 PM@User - s
SidPlays_144p
06/02/2021, 8:46 PMyou've got to be kidding me - s
SidPlays_144p
06/02/2021, 8:46 PMhow the hell - u
user
06/02/2021, 8:47 PMWith only 30 audio files - s
SidPlays_144p
06/02/2021, 8:47 PM🤯 - s
SuperLuig
06/02/2021, 8:48 PMnext test will be 200 epochs - u
user
06/02/2021, 8:49 PMWTF IT SOUNDS SO GOOD (Based on only 2 minutes of data) - s
SuperLuig
06/02/2021, 8:50 PMi swear to fuck if this turns out decent compared to my other models 😭 - s
SuperLuig
06/02/2021, 8:52 PMlike i'm gonna be pissed 😂 - u
user
06/02/2021, 8:52 PMWhy my model sounds better than your? - s
SuperLuig
06/02/2021, 8:52 PMi'm working with 25 lines and 1:20 of data - u
user
06/02/2021, 8:52 PMI'm mindblowed - s
SuperLuig
06/02/2021, 8:53 PMplus this is early on in the training - u
user
06/02/2021, 8:53 PMI'm working with 30 lines and 2:00 of data - u
user
06/02/2021, 8:53 PMAt 200 epochs and a validation loss of 0.06 - p
PUMPKINEATER
06/02/2021, 8:58 PMIs that Bill Gates? - u
user
06/02/2021, 8:58 PMExactly - s
SuperLuig
06/02/2021, 9:00 PMim pissed bro - s
SuperLuig
06/02/2021, 9:03 PMmeanwhile my snugglemagne voice with like 2:30 data sounds like ass, was it because i did too much training? same exact line - u
user
06/02/2021, 9:04 PMI dunno but it sounds fine - s
SuperLuig
06/02/2021, 9:04 PMcause i trained snugglemagne for over 4 hours - u
user
06/02/2021, 9:04 PMgive me the model and i will test it on hifi gan - s
SuperLuig
06/02/2021, 9:04 PM1-0bbORH1eSzEGka9qpopKbYPtCcX32aF @User - u
user
06/02/2021, 9:04 PMthx