search

SWivid / F5-TTS

Official code for "F5-TTS: A Fairytaler that Fakes Fluent and Faithful Speech with Flow Matching"
https://arxiv.org/abs/2410.06885
MIT License
5.26k stars 532 forks source link

Ability to specify finetuned model in the Gradio app #232

Open danielw97 opened 3 hours ago

danielw97 commented 3 hours ago

Hello, I've come to this repo recently, so apologies if I've missed something here. Things sound great, although I'm currently doing some finetuning tests from the original f5tts checkpoint with the gradio finetuning app. One feature that would be great is an ability to specify a custom checkpoint in the gradio_app.py if possible, maybe in the advanced settings? That way a finetuned checkpoint can be used more easily. If this feature is already available, do please let me know and thank you again for your work so far on this. Thanks in advance.

SWivid commented 2 hours ago

https://github.com/SWivid/F5-TTS/blob/32c3ee77017d728321ffaa5d10e9f9f6cd44f20d/gradio_app.py#L48 Could temporarily change str(cached_path("hf://SWivid/F5-TTS/F5TTS_Base/model_1200000.safetensors") to `.

Maybe someone to add an on-the-fly selection feature to gradio, welcome PR