Katherine Crowson

For DPM Adaptive I think you might want to decrease the per-step error tolerances atol and rtol, the defaults are kind of lax and don't reach the converged image. (Also...

> so with `sample_dpm_adaptive`… > > > make sure to pass in the Brownian tree > > first is Brownian, second is default: both ran for 51 iterations Since those...

It works for me, maybe you need to update pip and setuptools?

Oh I forgot to answer this! `K.sampling.log_likelihood()` returns positive log likelihood, not NLL, higher values indicate higher likelihood.

Also a footgun to be aware of: if you are evaluating log likelihood in a distributed environment you should *not* use a DDP wrapper on the denoiser's inner model because...

What sort of weird results? I think that should work, the problem I had was triggered by calling `accelerator.prepare()` on the `inner_model`.

I sum over the dimensions of each batch item when computing log likelihood so it is the log likelihood of the entire example, not per dimension, so it can be...

> But `x` should still be in batch shape right? So `x.shape == (batch_size, 3, 32, 32)` Yes. If you want per-dimension log likelihoods you need to divide the returned...

> Should I rather provide inner_model instead of model? `model` is correct, it won't know what to do with `inner_model`.

Since diffusion models operate on continuous data, the probability of sampling any given data point is not really defined, so for "log likelihood" we are evaluating a log probability density...