Danielle

Sin2pi

AI & ML interests

Cognitive Scientist. Cognitive science and Mathematics USC Genetics UCDavis

Recent Activity

updated a model 35 minutes ago
Sin2pi/whisper_experimental
updated a dataset 8 days ago
Sin2pi/improved_collator
updated a dataset 19 days ago
Sin2pi/code_for_data
View all activity

Organizations

Sin2pi's activity

replied to sanchit-gandhi's post about 1 month ago
view reply

You might consider including some type of subjective testing measure. The metrics you use are often misleading when it comes to quality and usability. Your tensorboard metrics show your looking for the answer to the right question at least.. but you have to test each model with your own ears to understand what's not working.. which is almost all of it.

New activity in openai/whisper-large-v3-turbo about 2 months ago

Just for inference?

1
#20 opened about 2 months ago by Sin2pi