Original training data

[mirecta]

Is there an original train dataset for czech language ? For train from scratch ?

Thank you

[a-n-lundgren]

Hi @mirecta,

Take a look at Hugging Face datasets for text to speach in Czech. I guess you will have the same challenge as me who is looking for good Swedish datasets. If the datasets at Hugging Face only contain a small set in your language you can try to combine data from multiple datasets. Might need some programming and/or scripting to get it all in the LJ Speech format.