Applications using the stable model names for base GPT-3 models (ada, babbage, curie, davinci) will routinely be upgraded to the brand new models listed above on January 4, 2024. The brand new models can even be accessible in the approaching weeks for early testing by specifying the next model names in API calls: ada-002, babbage-002, curie-002, davinci-002.

Developers using other older completion models (comparable to text-davinci-003) might want to manually upgrade their integration by January 4, 2024 by specifying gpt-3.5-turbo-instruct within the “model” parameter of their API requests. gpt-3.5-turbo-instruct is an InstructGPT-style model, trained similarly to text-davinci-003. This recent model is a drop-in alternative within the Completions API and might be available in the approaching weeks for early testing.

Developers wishing to proceed using their fine-tuned models beyond January 4, 2024 might want to fine-tune replacements atop the brand new base GPT-3 models (ada-002, babbage-002, curie-002, davinci-002), or newer models (gpt-3.5-turbo, gpt-4). Once this feature is on the market later this 12 months, we are going to give priority access to GPT-3.5 Turbo and GPT-4 fine-tuning to users who previously fine-tuned older models. We acknowledge that migrating off of models which are fine-tuned on your individual data is difficult. We might be providing support to users who previously fine-tuned models to make this transition as smooth as possible.

In the approaching weeks, we are going to reach out to developers who’ve recently used these older models, and can provide more information once the brand new completion models are ready for early testing.