Sklearn OneVsRestClassifier Giving Multiple positive outputs

Question

I have trained a model RandomForestModel in Sklearn for an multiclass classification task. I came across the OneVsRestModel to train for this task. Given below is a function I use to train and generate predictions which is just working fine. However, when I save and reload my trained model in some other file like any other trained model - my predictions do not make any sense. I expect one-hot outputs from it, but the model returns things like [0,1,1,0,0] etc. N_Classes = 5

Any hints on what I am missing or doing wrong?

def applyModel(clfr, X_, y_, xt, yt):
     classifier = OneVsRestClassifier(globals()[clfr]())
     trainedClf = classifier.fit(X_, y_)
     y_score = trainedClf.predict(xt)
     scor = trainedClf.score(xt, yt)*100
     return y_score, yt, scor, trainedClf