Note
Click here to download the full example code
Train, convert and predict with ONNX Runtime¶
This example demonstrates an end to end scenario starting with the training of a machine learned model to its use in its converted from.
Train a logistic regression¶
The first step consists in retrieving the iris datset.
from sklearn.datasets import load_iris
iris = load_iris()
X, y = iris.data, iris.target
from sklearn.model_selection import train_test_split
X_train, X_test, y_train, y_test = train_test_split(X, y)
Then we fit a model.
from sklearn.linear_model import LogisticRegression
clr = LogisticRegression()
clr.fit(X_train, y_train)
We compute the prediction on the test set and we show the confusion matrix.
from sklearn.metrics import confusion_matrix
pred = clr.predict(X_test)
print(confusion_matrix(y_test, pred))
[[13 0 0]
[ 0 11 2]
[ 0 0 12]]
Conversion to ONNX format¶
We use module sklearn-onnx to convert the model into ONNX format.
from skl2onnx import convert_sklearn
from skl2onnx.common.data_types import FloatTensorType
initial_type = [("float_input", FloatTensorType([None, 4]))]
onx = convert_sklearn(clr, initial_types=initial_type)
with open("logreg_iris.onnx", "wb") as f:
f.write(onx.SerializeToString())
We load the model with ONNX Runtime and look at its input and output.
import onnxruntime as rt
sess = rt.InferenceSession("logreg_iris.onnx", providers=rt.get_available_providers())
print("input name='{}' and shape={}".format(sess.get_inputs()[0].name, sess.get_inputs()[0].shape))
print("output name='{}' and shape={}".format(sess.get_outputs()[0].name, sess.get_outputs()[0].shape))
input name='float_input' and shape=[None, 4]
output name='output_label' and shape=[None]
We compute the predictions.
input_name = sess.get_inputs()[0].name
label_name = sess.get_outputs()[0].name
import numpy
pred_onx = sess.run([label_name], {input_name: X_test.astype(numpy.float32)})[0]
print(confusion_matrix(pred, pred_onx))
[[13 0 0]
[ 0 11 0]
[ 0 0 14]]
The prediction are perfectly identical.
Probabilities¶
Probabilities are needed to compute other relevant metrics such as the ROC Curve. Let’s see how to get them first with scikit-learn.
prob_sklearn = clr.predict_proba(X_test)
print(prob_sklearn[:3])
[[9.60244512e-02 8.88242228e-01 1.57333213e-02]
[2.71799658e-02 9.10234192e-01 6.25858422e-02]
[9.45106278e-01 5.48934527e-02 2.69394014e-07]]
And then with ONNX Runtime. The probabilies appear to be
prob_name = sess.get_outputs()[1].name
prob_rt = sess.run([prob_name], {input_name: X_test.astype(numpy.float32)})[0]
import pprint
pprint.pprint(prob_rt[0:3])
[{0: 0.09602457284927368, 1: 0.8882421851158142, 2: 0.015733299776911736},
{0: 0.027180003002285957, 1: 0.9102342128753662, 2: 0.06258579343557358},
{0: 0.9451063275337219, 1: 0.05489342659711838, 2: 2.693937801723223e-07}]
Let’s benchmark.
from timeit import Timer
def speed(inst, number=10, repeat=20):
timer = Timer(inst, globals=globals())
raw = numpy.array(timer.repeat(repeat, number=number))
ave = raw.sum() / len(raw) / number
mi, ma = raw.min() / number, raw.max() / number
print("Average %1.3g min=%1.3g max=%1.3g" % (ave, mi, ma))
return ave
print("Execution time for clr.predict")
speed("clr.predict(X_test)")
print("Execution time for ONNX Runtime")
speed("sess.run([label_name], {input_name: X_test.astype(numpy.float32)})[0]")
Execution time for clr.predict
Average 8.66e-05 min=7.56e-05 max=0.000104
Execution time for ONNX Runtime
Average 2.74e-05 min=2.42e-05 max=3.36e-05
2.7374654999903216e-05
Let’s benchmark a scenario similar to what a webservice experiences: the model has to do one prediction at a time as opposed to a batch of prediction.
def loop(X_test, fct, n=None):
nrow = X_test.shape[0]
if n is None:
n = nrow
for i in range(0, n):
im = i % nrow
fct(X_test[im : im + 1])
print("Execution time for clr.predict")
speed("loop(X_test, clr.predict, 100)")
def sess_predict(x):
return sess.run([label_name], {input_name: x.astype(numpy.float32)})[0]
print("Execution time for sess_predict")
speed("loop(X_test, sess_predict, 100)")
Execution time for clr.predict
Average 0.00817 min=0.00748 max=0.00911
Execution time for sess_predict
Average 0.00152 min=0.00143 max=0.00175
0.0015233170999999857
Let’s do the same for the probabilities.
print("Execution time for predict_proba")
speed("loop(X_test, clr.predict_proba, 100)")
def sess_predict_proba(x):
return sess.run([prob_name], {input_name: x.astype(numpy.float32)})[0]
print("Execution time for sess_predict_proba")
speed("loop(X_test, sess_predict_proba, 100)")
Execution time for predict_proba
Average 0.0115 min=0.011 max=0.0136
Execution time for sess_predict_proba
Average 0.0016 min=0.00147 max=0.00227
0.0015962060200003236
This second comparison is better as ONNX Runtime, in this experience, computes the label and the probabilities in every case.
Benchmark with RandomForest¶
We first train and save a model in ONNX format.
from sklearn.ensemble import RandomForestClassifier
rf = RandomForestClassifier()
rf.fit(X_train, y_train)
initial_type = [("float_input", FloatTensorType([1, 4]))]
onx = convert_sklearn(rf, initial_types=initial_type)
with open("rf_iris.onnx", "wb") as f:
f.write(onx.SerializeToString())
We compare.
sess = rt.InferenceSession("rf_iris.onnx", providers=rt.get_available_providers())
def sess_predict_proba_rf(x):
return sess.run([prob_name], {input_name: x.astype(numpy.float32)})[0]
print("Execution time for predict_proba")
speed("loop(X_test, rf.predict_proba, 100)")
print("Execution time for sess_predict_proba")
speed("loop(X_test, sess_predict_proba_rf, 100)")
Execution time for predict_proba
Average 1.31 min=1.29 max=1.33
Execution time for sess_predict_proba
Average 0.0022 min=0.00198 max=0.00278
0.002198638819999985
Let’s see with different number of trees.
measures = []
for n_trees in range(5, 51, 5):
print(n_trees)
rf = RandomForestClassifier(n_estimators=n_trees)
rf.fit(X_train, y_train)
initial_type = [("float_input", FloatTensorType([1, 4]))]
onx = convert_sklearn(rf, initial_types=initial_type)
with open("rf_iris_%d.onnx" % n_trees, "wb") as f:
f.write(onx.SerializeToString())
sess = rt.InferenceSession("rf_iris_%d.onnx" % n_trees, providers=rt.get_available_providers())
def sess_predict_proba_loop(x):
return sess.run([prob_name], {input_name: x.astype(numpy.float32)})[0]
tsk = speed("loop(X_test, rf.predict_proba, 100)", number=5, repeat=5)
trt = speed("loop(X_test, sess_predict_proba_loop, 100)", number=5, repeat=5)
measures.append({"n_trees": n_trees, "sklearn": tsk, "rt": trt})
from pandas import DataFrame
df = DataFrame(measures)
ax = df.plot(x="n_trees", y="sklearn", label="scikit-learn", c="blue", logy=True)
df.plot(x="n_trees", y="rt", label="onnxruntime", ax=ax, c="green", logy=True)
ax.set_xlabel("Number of trees")
ax.set_ylabel("Prediction time (s)")
ax.set_title("Speed comparison between scikit-learn and ONNX Runtime\nFor a random forest on Iris dataset")
ax.legend()
5
Average 0.0984 min=0.0967 max=0.1
Average 0.00149 min=0.00143 max=0.00152
10
Average 0.162 min=0.158 max=0.164
Average 0.00153 min=0.0014 max=0.00172
15
Average 0.229 min=0.225 max=0.235
Average 0.00157 min=0.00151 max=0.00161
20
Average 0.287 min=0.283 max=0.291
Average 0.00161 min=0.00149 max=0.00184
25
Average 0.35 min=0.349 max=0.352
Average 0.00164 min=0.0015 max=0.00184
30
Average 0.414 min=0.411 max=0.422
Average 0.00165 min=0.0014 max=0.00175
35
Average 0.475 min=0.471 max=0.483
Average 0.00171 min=0.00159 max=0.00178
40
Average 0.532 min=0.527 max=0.536
Average 0.00176 min=0.00165 max=0.00198
45
Average 0.598 min=0.593 max=0.604
Average 0.00179 min=0.00171 max=0.00191
50
Average 0.669 min=0.663 max=0.677
Average 0.00186 min=0.00167 max=0.00213
<matplotlib.legend.Legend object at 0x7f551da12d00>
Total running time of the script: ( 6 minutes 3.597 seconds)
