2/6 Fine-tuned Solar LLM outperforms #GPT-4 in over 80% of tasks. Try it yourself!

→ View original post on X — @predibase