Can I update my model using partial_fit after training my model using fit?

I have trained my model using MLPClassifier using fit method and saving in pickle object for the first time.

clf = MLPClassifier(hidden_layer_sizes=(50,50,5), max_iter=100, alpha=0.0001,
                     solver='sgd', verbose=10,  random_state=21,tol=0.000000001)
clf.fit(X_old, y_old)
joblib.dump(clf, 'saved_model_clf.pkl')

After few hours I am getting new data from online and I am updating my model using partial_fit .

clf_load = joblib.load('saved_model_clf.pkl')
y_new = to_categorical(y_new,num_classes=3) #I knew classes from training set
clf_load.partial_fit(X_new, y_new, classes=list(range(y_new.shape[1])))
joblib.dump(clf_load, 'saved_model_clf.pkl')

After partial_fit I'm predicting my results

clf_pred = joblib.load('saved_model_clf.pkl')
predictions = clf_pred.predict(df[0:1])

I found that my model is not learning on new incoming data while performing Prediction after the partial_fit

I am not sure where is the mistake

Case 1: I am doing mistake in saving and retrieving pickle object .

Case 2: My assumption of updating my model is wrong .

Can someone help me out to solve my problem please ?

Example dataset for clf.fit

X_old = {"TABLE":"XYZ","MODULE":"ABY", "MAX_TIME":647,"RESPONSE":"SUCCESS","SEVERITY":"Critical", "SERVICE":"ZC Service", "EVENT_NAME":"SendEmail"}

Y_old = {"Classification":"Ignore_Anyway"}

Example dataset for clf.partial_fit

X_new = {"TABLE":"XYZ","MODULE":"ABY", "MAX_TIME":647,"RESPONSE":"SUCCESS","SEVERITY":"Critical", "SERVICE":"ZC Service", "EVENT_NAME":"SendEmail"}

Y_new = {"Classification":"Not_Ignore"}