Problem with pose segmentation (slow? stuck?)

[SusanL82]

Hello team VAME,

First off, my apologies for posting yet another issue.
I was able to run the VAME pipeline on the provided test video and on a video of my own, but I am running into trouble when I try to process an entire experiment worth of videos. I'm not sure if this is because I have too many videos, or because something else is wrong.
Basically, model training and evaluation run fine.
When I then run pose segmentation:

1. Spyder displays progress bars for latents embedding for each video
2. a results folder is created for each video
3. I get the message that the same settings are applied for all animals for 60 clusters
4. I get a message that HMM is used

and then nothing. There is no error message, but as far as I can tell nothing else happens either (nothing is written in the VAME project folder and there are no 'busy' sounds coming from either the PC or the harddrive with my project on it). I gave my computer 2 days to produce something and then I killed the script.

I realise that this is a very vague problem, but I am really not sure where to start looking for a solution. Is there perhaps a maximum number of videos per project? Or do I need a certain amount of free space on my anaconda path? Am I just being too impatient? Something else?

Susan