David Marx

NB: as of this writing, testing has basically been limited to making sure the modified notebook here runs. Also be advised: the notebook in this branch currently clones from this...

Oh yeah also a general tip: `pip install -e` isn't always supported. Better to just use `pip install`. I think that's why taming or CLIP wasn't being installed correctly, for...

@GraphtyLove NB: this PR was suggested 5 months ago and has been stale since then. If this refactor is something you're interested in incorporating, it will take some additional work...

fastapi might be a good option, basically get this for free with the pydantic models

oh lol, linking #97

i might tinker with this refactor

I think a way that we could rethink this would be as "sensitivity" rather than "importance". In other words, if we were to define "feature importance" as an answer to...

this is normal, you only need the weights from the CLIP text encoder. This warning is telling you the vision encoder weights aren't being used, which is deliberate.

implementations for those samplers can be found here: https://github.com/crowsonkb/k-diffusion

Getting same issue, with and without the `--use-dataset-reader` argument. Using google colab for training as detailed [here](https://github.com/allenai/allennlp/issues/3103#issuecomment-521754848). Trying to run the predict command as follows: ``` !echo '{"text":"the hip bone...