traingenerator | 🧙 A web app to generate template code for machine learning | Machine Learning library

what I would do is to use categorical cross entropy. In your generators change class_mode to 'categorical'. In model.compile make loss='categorical_crossentropy. Not sure this will fix it but it can't hurt. Could be when you use more images, perhaps there are some Na labels. Check your datafame for Na's.

I got this working, instead of a doubly-linked list, I just created a Global Dictionary and kept a flag of the trains for a signal ahead and a signal behind. In this way, every train would check the next signal, if it is red, it would decelerate upto the next signal. The link to the entire project could be found here -

https://github.com/abhiray92/train_simulation_simpy/blob/main/Simulation.ipynb

Ok, after spending two more days, trying to fix some really misleading errors, and getting python.exe to crash when it was finally working, I was able to convert my generator to tensorflow dataset.

I was unable to make it work with batches, and numpy.array is not accepted by Dataset because it is not a sequential in Dataset's world, and it is important to return a tuple, don't know how examples with "yield" and "return data, labels" works.

Generator

Before trying to reference _DICT_NAME verify it's been put in the general index (and thus became cross-referenceable!!). If using an autodoc directive without :undoc-members: option the dictionary won't be include if it doesn't have a comment. Variables can't have docstrings, so you need to use a comment #: before or """ immediately after, in the py for autodoc to pick it up.

Unless you are explicitly including the dictionary in the .rst with py:data:, using Domains and roles. Or, by including it in the .rst (using autodata or autoattribute, in case your automodule or autoclass doesn't pick it up and you're using directives of the autodoc extension.

Then your _DICT_NAME is also a private variable, so you need to check the conf.py settings to see if it's included like napoleon_include_private_with_doc = True, if you're using Sphinx Napoleon extension.

_DICT_NAME is rendered as if referenced, but no hyperlink.

First check that it shows up documented in the final rendering of you documentation be it HTML or other. If it is included and has a comment it can be then be cross-referenced and the link will be generated.

How to reference a dict with sphinx-autodoc?

Notice you aren't referencing with autodoc. Autodoc is extracting the docstring where you write its directive in your .rst. Your :data: cross-referencing is done by domain and roles. Although you can have referencing inside your docstring that is afterwards extracted by autodoc and finally rendered, altogether, by the Sphinx tool.

EDIT after OP added MCVE:

First problem I see is this, your dictionary _DEFAULT_TRAINGEN_CFG is in module _default_configs.py...

So, the cross-reference you wrote should be right :data:~deeptrain.util._default_configs._DEFAULT_TRAINGEN_CFG\....BUT, did you actually include this module in any .rst file? I don't think so, looking at deeptrain.util.rst the _default_configs.py module is not present in any automodule directive, so there's nothing to reference (that's why the hyperlink isn't generated)...

Three examples of cross-references:

A module level dictionary, dict_document.py

Your lambda is bound to the Button class it was passed into, which implicitly made the first parameter the Button object itself. The result was that the trainpath parameter, was actually a renamed btn instance of Button. The functions that were trying to use trainpath as a filepath string were confused and so threw the error.

If you want to keep the lambda, simply add self as the first parameter, and then ignore it: