Make predictions in parallel ?

Is there any counter-indications that we could make predictions in parallel ? using joblib for instance.

Using notebooks implementation in [ Prediction.ipynb](https://github.com/euroargodev/OSnet-GulfStream/blob/main/Prediction.ipynb) I ended up with a trivial formulation like:

```python
for model in ensemble:
tmp_pred = model.predict(X)
```
where ``model`` is one instance returned by ``keras.models.load_model``

This is implemented here: https://github.com/euroargodev/OSnet-GulfStream/blob/df8486789fb86b3c5746b004ba3573b769350636/osnet/facade.py#L92

@loicbachelot @quai20 , any thoughts on this may be ?