make.py

import numpy as np
from keras.models import Sequential, model_from_json
from keras.layers import Dense, Dropout, Activation
from keras.optimizers import SGD, Adam

def fizzbuzz(i):
    if   i % 15 == 0: return np.array([0, 0, 0, 1])
    elif i % 5  == 0: return np.array([0, 0, 1, 0])
    elif i % 3  == 0: return np.array([0, 1, 0, 0])
    else:             return np.array([1, 0, 0, 0])

def bin(i, num_digits):
    return np.array([i >> d & 1 for d in range(num_digits)])

NUM_DIGITS = 7
trX = np.array([bin(i, NUM_DIGITS) for i in range(1, 101)])
trY = np.array([fizzbuzz(i) for i in range(1, 101)])
model = Sequential()
model.add(Dense(64, input_dim = 7))
model.add(Activation('tanh'))
model.add(Dense(4, input_dim = 64))
model.add(Activation('softmax'))
model.compile(loss = 'categorical_crossentropy', optimizer = 'adam', metrics = ['accuracy'])
model.fit(trX, trY, epochs = 3600, batch_size = 64)
import onnxmltools
onnx_model = onnxmltools.convert_keras(model, target_opset=0) 
onnxmltools.utils.save_model(onnx_model, 'fizzbuzz.onnx')