search

jaketae / storyteller

Multimodal AI Story Teller, built with Stable Diffusion, GPT, and neural text-to-speech
MIT License
482 stars 64 forks source link

[TODO] Add `speaker_device` #8

Closed jaketae closed 10 months ago

jaketae commented 1 year ago

Context

Currently, the StoryTellerConfig only allows users to specify devices for the painter (diffusion model) and writer (causal LM). This is largely because the Coqui TTS API doesn't appear to support model loading onto specific devices. See https://github.com/coqui-ai/TTS/issues/2282.

Solution

Once the referenced issue reaches resolution in upstream, make adjustments to the project to support more granular TTS model loading behavior.

stale[bot] commented 1 year ago

This issue has been automatically marked as stale because it has not had recent activity. It will be closed if no further activity occurs. Thank you for your contributions.

jaketae commented 11 months ago

Proposal at https://github.com/coqui-ai/TTS/pull/2855.

jaketae commented 10 months ago

Proposal merged upstream into TTS, and fix applied downstream via a1138ce.