Is text-davinci-003 GPT 3.0 and different from ChatGPT, GPT-3.5?

Question

Is text-davinci-003 GPT 3.0 and different from ChatGPT, GPT-3.5?

In my understanding, text-davinci-003 is GPT 3.0 and ChatGPT is GPT 3.5, and GPT 3.5 will be available on Azure OpenAI. Is my understanding correct?

Answer 1

Hello Hiroaki Machida

Thanks for reaching out to us. As you know, ChatGPT is fine-tuned from a model in the GPT-3.5 series, it is better suited for chatbot applications, which finished training in early 2022. You can learn more about the 3.5 series here. ChatGPT and GPT 3.5 were trained on an Azure AI supercomputing infrastructure.

As the description in OpenAI page, text-davinci-003 is recognized as GPT 3.5

GPT-3.5 series is a series of models that was trained on a blend of text and code from before Q4 2021. The following models are in the GPT-3.5 series:

1. code-davinci-002 is a base model, so good for pure code-completion tasks
2. text-davinci-002 is an InstructGPT model based on code-davinci-002
3. text-davinci-003 is an improvement on text-davinci-002

Reference link is here - https://platform.openai.com/docs/model-index-for-researchers/models-referred-to-as-gpt-3-5

I hope above information helps. Please let me know if you need more information, happy to help.

Regards,

Yutong

-Please kindly accept the answer if you feel helpful to support the community, thanks a lot.

Answer 2

Hey Hiroaki Machida,

I hope this email finds you well. I wanted to follow up on your previous question, although I understand some time has passed and you may no longer need assistance.
Recently, I came across a helpful resource that I believe would be beneficial to share with you.

In my own research of GPT-3 and ChatGPT, I found that utilizing both has been effective in achieving complex tasks to very simple tasks.
The website I discovered, https://learnprompting.org/docs/basics/chatbot_basics, provides detailed explanations, in-depth technical examples for maximizing the potential of ChatGPT specifically, but also applies to GPT-3.

~Cheers