feat: 3-mode selfhosted refactoring (--gpu, --cpu, --hosted) + audio token auth fallback (#896)

* fix: local processing instead of http server for cpu

* add fallback token if service worker doesnt work

* chore: rename processors to keep processor pattern up to date and allow other processors to be createed and used with env vars
This commit is contained in:
Juan Diego García
2026-03-04 16:31:08 -05:00
committed by GitHub
parent 4235ab4293
commit a682846645
34 changed files with 2640 additions and 172 deletions

View File

@@ -89,11 +89,11 @@ LLM_CONTEXT_WINDOW=16000
## =======================================================
## Audio Padding
##
## backends: local (in-process PyAV), modal (HTTP API client)
## Default is "local" — no external service needed.
## backends: pyav (in-process PyAV), modal (HTTP API client)
## Default is "pyav" — no external service needed.
## Set to "modal" when using Modal.com or self-hosted gpu/self_hosted/ container.
## =======================================================
#PADDING_BACKEND=local
#PADDING_BACKEND=pyav
#PADDING_BACKEND=modal
#PADDING_URL=https://xxxxx--reflector-padding-web.modal.run
#PADDING_MODAL_API_KEY=xxxxx
@@ -101,8 +101,8 @@ LLM_CONTEXT_WINDOW=16000
## =======================================================
## Diarization
##
## Only available on modal
## To allow diarization, you need to expose expose the files to be dowloded by the pipeline
## backends: modal (HTTP API), pyannote (in-process pyannote.audio)
## To allow diarization, you need to expose expose the files to be downloaded by the pipeline
## =======================================================
DIARIZATION_ENABLED=false
DIARIZATION_BACKEND=modal