Charlelie Laurent

> It would be nice to also allow just passing an initialized model itself. The downside is that the user would be required to determine the correct input and output...

> Is it the name field physicsnemo.ModelMetaData? This was what I expected. AFAIK the name does not matter that much, it's only used to infer a default filename when saving/loading...

Glad to know that it works. Let me know if you encounter other issues with the entrypoint mechanism. I'll also write some explanation about this in the docs, as many...

@luke-conibear thank you for reporting. We are aware of issues with CorrDiff checkpoints, and those will be addressed by #871 once it is merged. For the time being, you can...

@luke-conibear after discussion with @jialusui1102 it seems your problem is not due to checkpoints (downgrading to 1.0.1-rc should still fix your issue until we resolve this). Could please detail how...

@luke-conibear thank you for the details. > Generation has the error above. This was due to keeping AMP enabled in inference, which shouldn't be the case. It should be fixed...

@luke-conibear I am not able to reproduce the `RuntimeError` with the latest commit. To help me troubleshoot this, could you please: - Give me commit hash that you are using...

@luke-conibear thank you for the details, we will try to reproduce your error with the `generate.py`. > The above information is for non-patched diffusion, as I cannot get the patched...

> Yes, you're right about the patch shape. I used 32 and patched diffusion works. Great to know! We will update the log messages to more clearly explain why patching...