Max Berrendorf

Results 61 issues of Max Berrendorf

[pathlib](https://docs.python.org/3/library/pathlib.html#module-pathlib) offers a high-level API for working with paths. It also allows for clearer type annotations than just `str`. This issues proposes to use pathlib wherever possible. For ease-of-use, users...

enhancement

related to https://github.com/mberr/torch-ppr/pull/18

Since `ClassResolver.from_subclasses` excludes classes from private modules by default, and the name of the module from which a Python script is invoked is replaced by `__main__`, classes defined there are...

##### Summary This PR adds a simple data collector to extract information about the queue in [slurm](https://slurm.schedmd.com/). It is based on parsing the output of [`squeue`](https://slurm.schedmd.com/squeue.html), and currently only extracts...

area/docs
area/web
area/collectors
area/build
collectors/python.d

### Problem Statement By having the interaction module and the implementation of the interaction function split across two files, the code becomes less readable. It also spreads documentation across the...

enhancement

This PR merges AutoSF's functional interaction with the state-ful module. It is also intended to serve as an example on how to do https://github.com/pykeen/pykeen/issues/1102 .

This PR merges ComplEx' functional interaction with the (functional) module. It also de-duplicates information between the ComplEx model class and the interaction, and updates the docstring. It is also intended...

### Problem Statement [Lacroix et al., 2018](http://imagine.enpc.fr/~obozinsg/papers/lacroix2018canonical.pdf) introduce a new training objective $$ l_{i,j,k}(X) = - X_{i,j,k} + \log \left( \sum_{k'} \exp(X_{i,j,k′}) \right) - X_{k,j+P,i} + \log \left( \sum_{i'} \exp...

enhancement
⚽ Training Loop

Cherry pick from #752 to reduce diff, as well as consistent renaming across all files. *note*: the large number of changed files is partially caused by the experimental configurations With...