Remove BaseStochasticGradient step method

Closes #5138

Author: michaelosthege

RELEASE-NOTES.md

@@ -15,7 +15,6 @@ We plan to get these working again, but at this point their inner workings have
 - Timeseries distributions (see [#4642](https://github.com/pymc-devs/pymc/issues/4642))
 - Nested Mixture distributions (see [#5533](https://github.com/pymc-devs/pymc/issues/5533))
 - Elliptical slice sampling (see [#5137](https://github.com/pymc-devs/pymc/issues/5137))
-- `BaseStochasticGradient` (see [#5138](https://github.com/pymc-devs/pymc/issues/5138))
 - `pm.sample_posterior_predictive_w` (see [#4807](https://github.com/pymc-devs/pymc/issues/4807))
 - Partially observed Multivariate distributions (see [#5260](https://github.com/pymc-devs/pymc/issues/5260))
 
@@ -28,6 +27,7 @@ Also check out the [milestones](https://github.com/pymc-devs/pymc/milestones) fo
 All of the above apply to:
 
 - BART was removed [#5566](https://github.com/pymc-devs/pymc/pull/5566). It is now available from [pymc-experimental](https://github.com/pymc-devs/pymc-experimental)
+- `BaseStochasticGradient` was removed (see [#5630](https://github.com/pymc-devs/pymc/pull/5630))
 - ⚠ The library is now named, installed and imported as "pymc". For example: `pip install pymc`.
 - ⚠ Theano-PyMC has been replaced with Aesara, so all external references to `theano`, `tt`, and `pymc3.theanof` need to be replaced with `aesara`, `at`, and `pymc.aesaraf` (see [4471](https://github.com/pymc-devs/pymc/pull/4471)).
 - `pm.Distribution(...).logp(x)` is now `pm.logp(pm.Distribution(...), x)`.