Hi everyone,
I have little knowledge about GPT family before but am really impressed by those API/models regarding their abilities on various tasks. However, when I read official introductions/documents of those models, I found a bit confused, especially about their differences and relationships between them.
- Does ChatGPT always refer to single product or a series of models fine tuned on GPT3.5 series models?
- Is ChatGPT = GPT 3.5 turbo? Is there any difference between them?
- Is GPT3.5 turbo trained with RLHF as well?
If you have any thoughts or suggestions, please leave a comment below. Thanks in advance for your input!