Hi,
Having to do a few docker down and ups, I can see the initialisation / start up process before I can access the UI, takes quite a bit of time. I noticed initially it was downloading gigs of data with the canary model, and then, I tried to introduce some cache into the compose to stop that, but either way it takes a long time. Is that expected ?
Here is the sequence based in the docker logs (via Dozzle) - is this normal
Setting up custom user with UID=99, GID=100...
Group with GID 100 already exists, using it
Setting up data directories...
=== Setup Complete ===
Switching to user appuser (UID=99, GID=100) and starting application...
time=21:29:00
level=INFO
msg=Starting Scriberr
version=dev
[+] Loading configuration
time=21:29:00
level=INFO
msg=Registering adapters with environment path
whisperx_env=/app/whisperx-env
time=21:29:00
level=INFO
msg=Adapter registration complete
[+] Connecting to database
[+] Setting up authentication
[+] Initializing SSE broadcaster
[+] Initializing repositories
[+] Initializing services
[+] Initializing transcription service
[+] Initializing transcription service
[+] Preparing Python environment
time=21:30:38
level=INFO
msg=Initializing unified transcription service
time=21:30:38
level=INFO
msg=Initializing registered models in parallel...
time=21:30:38
level=INFO
msg=Preparing NVIDIA Sortformer environment
env_path=/app/whisperx-env/parakeet
time=21:30:38
level=INFO
msg=Preparing NVIDIA Parakeet environment
env_path=/app/whisperx-env/parakeet
time=21:30:38
level=INFO
msg=Preparing WhisperX environment
env_path=/app/whisperx-env
time=21:30:38
level=INFO
msg=Preparing PyAnnote environment
env_path=/app/whisperx-env/pyannote
time=21:30:38
level=INFO
msg=Preparing NVIDIA Canary environment
env_path=/app/whisperx-env/parakeet
time=21:30:38
level=INFO
msg=transcription model initialized
model_id=openai_whisper
time=21:41:22
level=INFO
msg=WhisperX environment already ready
time=21:41:22
level=INFO
msg=transcription model initialized
model_id=whisperx
time=21:43:04
level=INFO
msg=PyAnnote already available in environment
time=21:43:04
level=INFO
msg=diarization model initialized
model_id=pyannote
time=21:51:31
level=INFO
msg=Parakeet environment already ready
time=21:51:31
level=INFO
msg=transcription model initialized
model_id=parakeet
time=21:51:31
level=INFO
msg=Canary environment already ready
time=21:51:31
level=INFO
msg=transcription model initialized
model_id=canary
time=21:51:31
level=INFO
msg=Sortformer environment already ready
time=21:51:31
level=INFO
msg=diarization model initialized
model_id=sortformer
time=21:51:31
level=INFO
msg=Model initialization completed
time=21:51:31
level=INFO
msg=Unified transcription service initialized successfully
[+] Initializing quick transcription service
[+] Starting background processing
time=21:51:31
level=INFO
msg=Scriberr is ready
url=http://0.0.0.0:8080/
Hi,
Having to do a few docker down and ups, I can see the initialisation / start up process before I can access the UI, takes quite a bit of time. I noticed initially it was downloading gigs of data with the canary model, and then, I tried to introduce some cache into the compose to stop that, but either way it takes a long time. Is that expected ?
Here is the sequence based in the docker logs (via Dozzle) - is this normal