XGBRegressor Early Stopping Rounds Working in Python

Question

I have been developing my model with XGBRegressor and I have a technical question about the early stopping rounds working. This is the code:

model = XGBRegressor(n_jobs=-1, n_estimators=1000, early_stopping_rounds=50, random_state=42)
model.fit(X_train, y_train, eval_set=[(X_train, y_train),(X_val,y_val)])

This is the output:

[0] validation_0-rmse:12740.82085   validation_1-rmse:14354.43509
[1] validation_0-rmse:9487.04070    validation_1-rmse:10616.13605
[2] validation_0-rmse:7283.04039    validation_1-rmse:8239.62752
[3] validation_0-rmse:5858.70801    validation_1-rmse:6670.77442
[4] validation_0-rmse:4965.48363    validation_1-rmse:5681.79305
[5] validation_0-rmse:4385.77247    validation_1-rmse:5110.99709
[6] validation_0-rmse:4043.46385    validation_1-rmse:4774.20268
[7] validation_0-rmse:3763.10815    validation_1-rmse:4635.01149
[8] validation_0-rmse:3591.72373    validation_1-rmse:4561.27221
[9] validation_0-rmse:3459.59007    validation_1-rmse:4520.31052
[10]    validation_0-rmse:3364.49960    validation_1-rmse:4493.90137
[11]    validation_0-rmse:3292.77090    validation_1-rmse:4488.63371
[12]    validation_0-rmse:3261.68385    validation_1-rmse:4477.47174
[13]    validation_0-rmse:3194.66407    validation_1-rmse:4479.46967
[14]    validation_0-rmse:3131.46617    validation_1-rmse:4483.96315
[15]    validation_0-rmse:3105.17158    validation_1-rmse:4496.17191
[16]    validation_0-rmse:2980.64759    validation_1-rmse:4513.39312
[17]    validation_0-rmse:2969.50538    validation_1-rmse:4515.67529
[18]    validation_0-rmse:2928.73793    validation_1-rmse:4514.08650
[19]    validation_0-rmse:2885.23440    validation_1-rmse:4512.02239
[20]    validation_0-rmse:2877.47452    validation_1-rmse:4511.70923
[21]    validation_0-rmse:2775.98275    validation_1-rmse:4541.79559
[22]    validation_0-rmse:2665.77462    validation_1-rmse:4555.71092
[23]    validation_0-rmse:2636.37427    validation_1-rmse:4549.20621
[24]    validation_0-rmse:2562.55110    validation_1-rmse:4556.94927
...
[58]    validation_0-rmse:1443.62609    validation_1-rmse:4736.20431
[59]    validation_0-rmse:1423.84305    validation_1-rmse:4746.66728
[60]    validation_0-rmse:1387.68330    validation_1-rmse:4747.30871
[61]    validation_0-rmse:1367.73335    validation_1-rmse:4747.11801

I have understood that the model has trained 62 trees and the best number of trees is 13, but if 13 is the optimal number why does it report the results until 62 if the RMSE is increasing?