Drop the sample_weight entry from GMM.fit's docstring

[VenishPaneliya]

What this fixes

GMM.fit has the signature fit(self, X, y=None), but its docstring documents a third argument:

sample_weight : array-like, shape (n_samples,)
    Per-sample weights. Rescale C per sample. Higher weights
    force the classifier to put more emphasis on these points.

There's no **kwargs to absorb it, so following the documentation fails:

>>> GMM().fit(X, sample_weight=w)
TypeError: GMM.fit() got an unexpected keyword argument 'sample_weight'

Two things say this is a stray copy rather than a missing feature:

• The wording comes from OCSVM.fit, which really does take sample_weight (fit(self, X, y=None, sample_weight=None, **params)). "Rescale C per sample" refers to the SVM penalty parameter, which has no counterpart in a Gaussian mixture — and a GMM is not a classifier.
• It couldn't be forwarded even if we wanted to: sklearn.mixture.GaussianMixture.fit is fit(self, X, y) with no sample_weight.

So the only correct reconciliation is removing the documentation. OCSVM is untouched — it documents the argument correctly because it accepts it.

GMM and OCSVM are the only two detectors in pyod/models/ whose fit docstring mentions sample_weight; OCSVM is the correct one.

Testing

Documentation-only, so no behaviour to test:

• pyod/test/test_gmm.py: 14 passed
• verified OCSVM().fit(X, sample_weight=w) still works and GMM.fit's signature is unchanged
• flake8 reports the same single pre-existing finding as an unmodified checkout