# talknet-support
- u
{K EY1} (Kei)
10/20/2022, 11:57 PMlike it doesn't error out - g
Gosmokeless28
10/21/2022, 1:12 AMYes, that's right - u
{K EY1} (Kei)
10/21/2022, 1:36 AMOk cool Thanks - u
{K EY1} (Kei)
10/21/2022, 7:28 PManyone know why this happened in step 6 of the talknet notebook in #841437191073955920 ? - u
{K EY1} (Kei)
10/21/2022, 7:30 PMi'm rerunning the step, maybe it was a fluke? - u
{K EY1} (Kei)
10/21/2022, 8:16 PMhappened again at the same spot - u
{K EY1} (Kei)
10/21/2022, 8:25 PM@Justin do you know what this error means? - m
moomoo
10/21/2022, 11:15 PMI want to ask, if 30 minutes of data set is a enough to give a decent model? and if I'm making a model of a korean singer can I transcribe the korean lyrics according to their romanization and get good results or would it fail?? - u
{K EY1} (Kei)
10/21/2022, 11:22 PM30 minutes is good enough You coild try the korean thing but it might not work - m
moomoo
10/21/2022, 11:40 PMokay, thank you ♥️ - l
Lexi (delulu posts on the daily)
10/22/2022, 12:30 AMleme guess blackpink or kpop related - m
moomoo
10/22/2022, 12:31 AMyeah, it's BTS - m
moomoo
10/22/2022, 12:31 AMwanted to give jin some lines in their older songs - m
moomoo
10/22/2022, 12:35 AM+ I want to try Arabic too, but thought about trying with korean first. I don't want to get the dataset of om kalthoum's discography if using non english languages won't give good results 😭😭 - j
Justin
10/22/2022, 2:06 AMtry running the cell again - j
Justin
10/22/2022, 2:07 AMoh iSee you already tried doing that hah! - j
Justin
10/22/2022, 2:08 AMmm - u
{K EY1} (Kei)
10/22/2022, 2:10 AMShould i try again in a different runtime - l
LeonardoC7 (JimmyTrpcev)
10/22/2022, 10:49 AMi submitted the Poopy Bear voice on UberDuck site - l
LeonardoC7 (JimmyTrpcev)
10/22/2022, 10:49 AMthat i made with talknet - l
LeonardoC7 (JimmyTrpcev)
10/22/2022, 10:51 AMthis doesn't have a category - m
moomoo
10/22/2022, 2:10 PMin the controllable collab, it says model can't be downloaded when I click generate, can someone tell me what's the problem?? - g
Gosmokeless28
10/22/2022, 2:31 PMWhy did you put
at the beginning of the bot command?/tts_ - m
moomoo
10/22/2022, 10:44 PMI made my first model ever finally, but it's a little robotic, so if I duplicated the wav files because I can't find anymore resources for the dataset, would this enhance the quality? - u
{K EY1} (Kei)
10/22/2022, 10:44 PMIt doesn't change anything - m
moomoo
10/22/2022, 10:52 PMokay, thank you - l
LeonardoC7 (JimmyTrpcev)
10/22/2022, 10:53 PMOops i didn't know how it was - d
dn2rafael
10/26/2022, 3:27 PMhey, I was training my model using the Talknet Training notebook yesterday and I had to stop it at one point. Luckily, I still have the checkpoints and stuff saved in my drive, now what cells do I need to execute to get back to training the HiFi GAN? - s
Sonic2022_mario
10/26/2022, 7:28 PMHow Do I Make A Dataset - u
{K EY1} (Kei)
10/26/2022, 7:43 PMSame way you make a tacotron dataset, check the tutorial in #841422801965416538