9 ms·
Not all use cases need GPT-4 level performance. I'd argue that even LLaMA-7B is quite limited. Also, new and improved models are being released all the time.
by Tepix 4y ago
Not all use cases need GPT-4 level performance. I'd argue that even LLaMA-7B is quite limited. Also, new and improved models are being released all the time.