Describe the bug
I found that some names agruments in framework aren't consistent.
So for example:

class SupervisedRunner(Runner):
    """Runner for experiments with supervised model."""

    _experiment_fn: Callable = SupervisedExperiment

    def __init__(
        self,
        model: Model = None,
        device: Device = None,
        input_key: Any = "features", 
      

In our API docs we currently use

.. autosummary::
   Client
   Client.call_stack
   Client.cancel
    ...

To generate a table of Client methods at the top of the page. Later on we use

.. autoclass:: Client
   :members:

to display the docstrings for all the public methods on Client (here an example for

If enter_data() is called with the same train_path twice in a row and the data itself hasn't changed, a new Dataset does not need to be created.

We should add a column which stores some kind of hash of the actual data. When a Dataset would be created, if the metadata and data hash are exactly the same as an existing Dataset, nothing should be added to the ModelHub database and the existing

This could be example that uses these supported syntax and APIs:

https://github.com/couler-proj/couler/tree/d34a690/couler/core/syntax

It seems that the number of joining clients (not the num of computing clients) is fixed in fedml_api/data_preprocessing/**/data_loader and cannot be changed except CIFAR10 datasets.

Here I mean that it seems the total clients is decided by the datasets, rather the input from run_fedavg_distributed_pytorch.sh.

https://github.com/FedML-AI/FedML/blob/3d9fda8d149c95f25ec4898e31df76f035a33b5d/fed