Refresh AI model setup (#13171)

Instead of initializing model at start time we do it at run time to be
able to swap model provider more easily.

Also introduce a third driver for openai-compatible providers, which
among other allows for local models with Ollama

packages/twenty-front/src/generated/graphql.ts

@@ -998,7 +998,8 @@ export enum MessageChannelVisibility {
 export enum ModelProvider {
   ANTHROPIC = 'ANTHROPIC',
   NONE = 'NONE',
-  OPENAI = 'OPENAI'
+  OPENAI = 'OPENAI',
+  OPENAI_COMPATIBLE = 'OPENAI_COMPATIBLE'
 }
 
 export type Mutation = {