I replaced my tabbyml code assistant this week with ollama+continue.dev. But I’m having issues with speed. I think this is because I switched from code qwen 2.5B (ish) to Deepeek Coder 9B (ish) and I think I’m pushing the limits of my GPU. Maybe I’ll spend today sorting out which models I want to use and which computers I want to use them on so I dont run into this issue (I’ve got ollama on 2 computers with 3 GPUs shared between them, for a total of 24GB VRAM)

Floccus just syncs, so whatever hasnt been synced would just wait in your bookmarks until the server was available.