Using Multiple Training Data with TPOT

If I have multiple training sets (i.e., `X_train_1`, `Y_train_1`, `X_train_2`, `Y_train_2`), is it possible to use TPOT to fit the same model parameters to multiple training sets? I'm using Dask distributed for parallel model building.

In the documentation, for a single training set, I should do:

from sklearn.externals import joblib
import distributed.joblib
from dask.distributed import Client

# connect to the cluster
client = Client('schedueler-address')

# create the estimator normally
estimator = TPOTClassifier(n_jobs=-1)

# perform the fit in this context manager
with joblib.parallel_backend("dask"):
estimator.fit(X_train_1, Y_train_1)

However, it isn't clear how I could also fit `X_train_2` and `Y_train_2` in parallel with `X_train_1` and `Y_train_1` without having to wait for `X_train_1` to complete.