#Internal error on all completions from fine-tuned models, both on playground and through API

1 messages · Page 1 of 1 (latest)

silent aspen
#

Bug Report:

Steps to reproduce:
All of my fine-tuned GPT models (on davinci, curie, etc) have suddenly stopped working as of an hour or so ago.

Expected result:
A completion generated by the model.

Actual result:
No completions given, for any model. Just "Internal error". This is true whether I try from playground or from the API. However, the standard OpenAi models are still working as expected.

Additional information

Same thing happens from desktop, mobile, etc.is