v0.1.8

• Better memory management in model manager
• multi-model deployment
• smart resource approximator for model memory allocation and kv cache allocation
• added completions endpoint on server
• fixed url opening through markdown
• added status bar footer
• added system prompt editor
• fix UI layouting
• added auto-scroll
• added feature to add your own model easily