feat: self-hosted gpu api (#636)

* Self-hosted gpu api

* Refactor self-hosted api

* Rename model api tests

* Use lifespan instead of startup event

* Fix self hosted imports

* Add newlines

* Add response models

* Move gpu dir to the root

* Add project description

* Refactor lifespan

* Update env var names for model api tests

* Preload diarizarion service

* Refactor uploaded file paths
This commit is contained in:
2025-09-17 18:52:03 +02:00
committed by GitHub
parent fa049e8d06
commit ab859d65a6
30 changed files with 4020 additions and 16 deletions

View File

@@ -0,0 +1,19 @@
[project]
name = "reflector-gpu"
version = "0.1.0"
description = "Self-hosted GPU service for speech transcription, diarization, and translation via FastAPI."
readme = "README.md"
requires-python = ">=3.12"
dependencies = [
"fastapi[standard]>=0.116.1",
"uvicorn[standard]>=0.30.0",
"torch>=2.3.0",
"faster-whisper>=1.1.0",
"librosa==0.10.1",
"numpy<2",
"silero-vad==5.1.0",
"transformers>=4.35.0",
"sentencepiece",
"pyannote.audio==3.1.0",
"torchaudio>=2.3.0",
]