Max Berrendorf
Max Berrendorf
* Support for non-square kernel-size/stride/padding/dilation * Automatic computation of output size from the other dimensions (if not explicitly set) * Support for custom colors * Ability to generate all steps...
A different try on #816
### Problem Statement Reloading the [`pykeen.pipeline.PipelineResult`](https://pykeen.readthedocs.io/en/latest/api/pykeen.pipeline.PipelineResult.html#pykeen.pipeline.PipelineResult) object from a directory seems to be frequently desired, cf., e.g., - https://github.com/pykeen/pykeen/discussions/1181#discussioncomment-4382887 - https://github.com/pykeen/pykeen/discussions/1189 ### Describe the solution you'd like Add a classmethod,...
For me, fixing the timm version seems to suffice to fix #242 locally. Other than that I kept the packages that are installed by `pip install`; for the package versions...
This is particularly useful when dealing with enums / `typing.Literal` as it generates normalization code which ensures the given type without having to resort to `typing.cast` and gives useful error...
## 🐛 Bug Calling nonzero for a tensor with more than INT_MAX elements leads to an error on GPU. On CPU there does not seem to be a problem. ##...
## Before submitting - [ ] Was this discussed/approved via a GitHub issue? (no need for typos and docs improvements) - [x] Did you create/update your **configuration file**? - [x]...
Fix https://github.com/pykeen/pykeen/issues/1384 Also, do not register an LR scheduler callback if we do not have a learning rate scheduler, instead of just doing nothing in the callback.
Extract epoch runtime estimates from all HPO trials as ``` (datetime_complete - datetime_start) / stopped_epoch ```
Closes #17