Why doesn't MLPclassifier work on my data?

I want to predict the products which a person will buy, by looking at the products which they bought earlier.

my dataframe has 'overall', 'reviewerID', 'asin' , 'brand'.

overall - rating of the product
reviewerID - ID of the reviewer
asin - ID of the product
brand - brand name

My code is as follows:

import pandas as pd 
import numpy as np
first = pd.read_pickle("customers_more_than_one_product.pkl")
data = pd.read_pickle("customers_more_than_one_product.pkl")

data['reviewerID'] = data["reviewerID"].rank(method='dense').astype(int)
data['asin'] = data["asin"].rank(method='dense').astype(int)
data['brand'] = data["brand"].rank(method='dense').astype(int)
data = data.drop(["reviewTime","reviewText","title"],1)

My dataframe is this.

I assigned 3 columns which are 'overall', 'reviewerID', 'brand' to X, and 1 column which is 'asin' to y below:

X= data["reviewerID"].values
x1=data["overall"].values
x2=data["brand"].values
X=np.vstack(X, x1,x2).T

y= data["asin"].values

  

I splitted data such that randomly selected 70% tuples are used for training while 30% tuples are used for testing.

from sklearn.model_selection import train_test_split
X_train,X_test , y_train , y_test = train_test_split(X, y, test_size=0.3, random_state=1)

I used Sklearn library for MLP Classsifier.

from sklearn.neural_network import MLPClassifier
import time


training_time_avg=0
error_cost_avg=0
i=0

I applied single-layer perceptron network solver to handle my data 10 times. Also I tried to fit my training datas and calculate the time by using 'time':

while i < 10:
    clf= MLPClassifier(hidden_layer_sizes=(), max_iter=100)

    start = time.time()
    clf.fit(X_train, y_train)
    stop = time.time()

    accuracy=clf.score(X_test, y_test)
    error_rate = 1 - accuracy

    training_time_avg = training_time_avg + (stop - start)
    error_cost_avg=error_cost_avg + error_rate

    i=i+1

I printed average training time and error cost my result.

print("max_iter:",100) 
print("\nTraining Time Average (in ms): ",(training_time_avg/10))
print("Error Average (cost): ", (error_cost_avg/10))

However, my code keeps run, although I got no errors when I run the code. It's like endless loop. I checked Variable Explorer screen on Spyder, everything worked until clf.fit(X_train, y_train) line.

Is there anyone who can help me to fix this problem? Or using which method can I reach prediction of the products which a person will buy, by looking at the products which they bought earlier?