A POST request to /free with: {"unload_models":true}
will unload models from vram.
A POST request to /free with: {"free_memory":true}
will unload models and free all cached data from the last run workflow.
29 KiB
29 KiB