After LWE installation we have in the folder "c:\Program files\SDL\SDL ETS\adapted-lps\archives" a number of files like EngFra_Adapted-xxxx_SRV_TNM_8_6_x_0.export. They are zip files and the total size of the 9 files is 20GB. Can we remove these zip files or move them to a different location? Can you explain what is their role and maybe how they should be handled?

If you trained your adapted Translation Engines deployed them, in theory,  you can delete the Adaptation project from LWE UI (especially test trained engines that nobody will ever deploy. You can delete those Adaptation projects from https:/[LWE_host]/adapted-lps/manual-adaptation). Fully trained LP which were deployed to Worker hosts will be preserved on the worker hosts, but the backups on the Controller hosts will be removed. If you still want to keep the Adapted LP project for future re-deployment, you can follow suggestions below


When an Adapted LP is deployed to a host, API Gateway will obtain the adapted LP from this adapted-lps\archives folder, and send it to the target host. The reason the Adapted LP is archived at the Controller is so that if a worker that had originally trained the Adapted LP is offline for whatever reason, the Controller still has a copy that can be used to deploy it to other workers (or to the Controller host itself)

If the files were moved and the user tried to deploy the adapted LP to a host, the deployment would fail since it wouldn’t find the needed archived adapted LP to deploy. In theory, you could manually move them out, and move them back when they need to deploy a specific Adapted LP, so API Gateway can find them. A better solution may be to map adapted-lps\archives (on the Controller) to a network share where you have more space, and the Controller would still have access to them.

If you 

The adaptation process is also requiring additional disk space to complete successfully. As of LWE 8.6.4 we are now recommending 100GB of disk space available on the machine where the training is taking place (it used to be 20GB). 

Please see the information in the Community https://community.rws.com/product-groups/linguistic-ai/edge/b/blogs/posts/adaptation-improvements-in-edge-8-6-4