Why fine-tune LLMs? GPT-4o fine-tune for PERFECT FLUX Image Prompts | IndyDevDan | Podwise