Eduardo Blancas
Eduardo Blancas
New users find confusing that Ploomber generates an output `.ipynb` from tasks, this is expected as `.ipynb` is considered a format for "source code" rather and an output format. We...
most users get started with the YAML API (i.e. `pipeline.yaml`), however, sometimes the project grows in complexity beyond what they expected and they start to hit the limits. at that...
If using grid, Ploomber currently adds a suffix to all products but if the user has any placeholders (eg [[parameter]]) it is not necessary and we can remove it
validate the content of task resources before pushing it to the cloud. Example, if there isn't a task with name `some-name`, the following should throw an error before uploading the...
implement some default values: ```yaml task_resources: __default: vcpus: 4 memory: 8192 gpu: 1 ``` and wildcard: ```yaml task_resources: fit-__: vcpus: 4 memory: 8192 gpu: 1% ```
The output of `ploomber build` hasn't changed in a while. There's an opportunity here to improve it to help users better track the status of the execution. This is especially...
Users are having a hard time finding some of our features. Adding Algolia to the docs helps with discovery but we can do more. I think we should make suggestions...
We currently don't have full examples for AWS and Google Cloud. This forces users to go around the docs to assemble their pipelines. Since consuming data from a warehouse, transforming...
when a script is in a subdirectory (say `scripts/fit.py`) and we pair notebooks in a certain folder (say `ploomber nb --pair notebooks`), the notebooks are created relative to the scripts...