search

openai / guided-diffusion

MIT License
6.06k stars 807 forks source link

Missing optimizer checkpoints #16

Closed bahjat-kawar closed 2 years ago

bahjat-kawar commented 2 years ago

I tried training based on the pre-trained models that were published, but this requires the checkpoints for the AdamW optimizer.

Are you planning on publishing them soon?

Many thanks

unixpickle commented 2 years ago

I don't think we have plans to publish the optimizer checkpoints, but in the past I've found it possible to resume training when cold-starting the optimizer.