Home | Notifications | New Note | Local | Federated | Search | Logout

Note Detail

ganbold@ganbold@hiradai.space (2026-10-03 21:51:58)
an attempt few weeks ago to run the TTS inference through GPU but its not complete, so its still "leaking" to CPU (running different process on both computing)

I want to fix it, but idk how many tokens am I going to burn, and I don't have the luxury to put an attention to this

I am cursing deadline
---Attachments---
image: https://hiradai.space/files/9d86a2a8-a551-4eef-a983-2763f6fcc07f.webp