logistic Regression on a arrays of various sizes

Question

In an audio classification exercise, I am trying to do a logistic regression where my X_train is a list. In fact, the list contains 100 elements and each element is an array of size: N1 But after doing the fit, I have the error of "setting an array element with a sequence. The requested array has an inhomogeneous shape after 1 dimensions. The detected shape was (100,) + inhomogeneous part." which I imagine is because of various size of the elements of my list. My question is: how can I equalize the size of the elements? For example, making my list a list of arrays of size 5001 ?

mfcc_all = []
labels_all = []

for i in range (100):
    entry = entries.loc[i,:]
    mfcc = load_mfcc(entry, _msdi_path)
    mfcc_all.append(mfcc)
    y = get_label(entry)
    labels_all.append(y)

logReg = LogisticRegression()
logReg.fit(mfcc_all, labels_all)