MrDragonFox's picture

MrDragonFox PRO

MrDragonFox

AI & ML interests

None yet

Recent Activity

Organizations

DeepGHS's profile picture SynthoCraft Ai's profile picture FoxEngineAi's profile picture Social Post Explorers's profile picture Mistral AI Game Jam's profile picture

MrDragonFox's activity

New activity in Jinsaryko/Elise about 6 hours ago

"borrowed" your dataset

#2 opened about 6 hours ago by
MrDragonFox
posted an update about 11 hours ago
view post
Post
541
did a small emotive classified test dataset for all the tts tuners out there

MrDragonFox/Elise

3h total mit - single speaker voice

dataset is a copy of an existing one just added the emotional tags over 1200 samples - should be good enough to test if emotional tags stick in your finetune
  • 1 reply
·
New activity in sesame/README 14 days ago

1b csm

#1 opened 14 days ago by
MrDragonFox
New activity in social-post-explorers/README 21 days ago

Post button doesn't appear.

7
#46 opened 22 days ago by
Blazgo
replied to Reality123b's post about 1 month ago
view reply

bro, im not that popular. its ok to do this ig?

also, no, im not saying openai's products are bad. nor im trying to like offend ANYONE or ANY company or organization. im just trying to promote my product

my problem isnt with you trying to promote a product - you asking if you should opensource something or not .. without showing anything working and state its not clickbait ..

show the goods man ..
otherwise noone cares - not how marketing works .. too many snakeoil sales men in that industry ... if you have something show it off .. if not .. well you know where this goes

replied to Reality123b's post about 1 month ago
view reply

the point you are mistaken here is that i dont care if he opensources it or not .. its the engagement farming for no reason while implying its not clickbait - but you do you

replied to Reality123b's post about 1 month ago
view reply

this is not clickbait .. if you want to opensource it .. you opensource it .. if not you dont .. its that simple - there are other approaches out there to this already

if that's not for farming what did you post it for ?

view reply

just limit vllm to 1 gpu and run the rest on a other one .. or use -gmu

view reply

8b repo empty and dataset empty too .. well its a little off from sota .... tbh glm4voice had better results - but its certainly a "ok" poc

gh repo empty / no paper