Fabian Zills
running the Node updates the `plots:` key in the `dvc.yaml`
For Projects with many groups it might be better to split these files into group-specific files
https://github.com/zincware/ZnTrack/blob/ac3938b80d571e40eb9ba10f7e0c7a8f4285c4c3/zntrack/core/node.py#L63-L69 This could be used as a check for everything that should be stored to disk. Something like ```python fig = ... if self.state.eager: fig.show() else: # triggered via `dvc...
something like: ```python class MyModel(zntrack.Node): dataset: dict = zntrack.import("https://huggingface.co/datasets/fka/awesome-chatgpt-prompts", import: bool = True) ``` - use the `import` flag, to define whether the file should be added to the repo...
Test loading `zntrack.params_path` within `with use_tmp_path`
when using `zntrack list` it might produce nodes, such as `ML1_ASEMD+checker_list+0` which aren't actual nodes but only `zntrack.deps` but are listed because `zntrack.json` is used instead of the `dvc.repo.Repo` to...
Something like `zntrack.outs_path(keep=False)` but how to handle it?
```python modifier = ... for _ in range: with projct: node(x=modifer) modifer.param += 1 # this will only use the latest one for every Node ```