mirror of
https://github.com/saymrwulf/onnxruntime.git
synced 2026-05-15 20:50:42 +00:00
95 lines
2.6 KiB
Python
95 lines
2.6 KiB
Python
# Copyright (c) Microsoft Corporation. All rights reserved.
|
|
# Licensed under the MIT License.
|
|
|
|
"""
|
|
|
|
.. _l-example-backend-api-tensorflow:
|
|
|
|
ONNX Runtime for Keras
|
|
======================
|
|
|
|
The following demonstrates how to compute the predictions
|
|
of a pretrained deep learning model obtained from
|
|
`keras <https://keras.io/>`_
|
|
with *onnxruntime*. The conversion requires
|
|
`keras <https://keras.io/>`_,
|
|
`tensorflow <https://www.tensorflow.org/>`_,
|
|
`keras-onnx <https://github.com/onnx/keras-onnx/>`_,
|
|
`onnxmltools <https://pypi.org/project/onnxmltools/>`_
|
|
but then only *onnxruntime* is required
|
|
to compute the predictions.
|
|
"""
|
|
import os
|
|
if not os.path.exists('dense121.onnx'):
|
|
from keras.applications.densenet import DenseNet121
|
|
model = DenseNet121(include_top=True, weights='imagenet')
|
|
|
|
from keras2onnx import convert_keras
|
|
onx = convert_keras(model, 'dense121.onnx')
|
|
onx.ir_version = 6
|
|
with open("dense121.onnx", "wb") as f:
|
|
f.write(onx.SerializeToString())
|
|
|
|
##################################
|
|
# Let's load an image (source: wikipedia).
|
|
|
|
from keras.preprocessing.image import array_to_img, img_to_array, load_img
|
|
img = load_img('Sannosawa1.jpg')
|
|
ximg = img_to_array(img)
|
|
|
|
import matplotlib.pyplot as plt
|
|
plt.imshow(ximg / 255)
|
|
plt.axis('off')
|
|
|
|
#############################################
|
|
# Let's load the model with onnxruntime.
|
|
import onnxruntime as rt
|
|
from onnxruntime.capi.onnxruntime_pybind11_state import InvalidGraph
|
|
|
|
try:
|
|
sess = rt.InferenceSession('dense121.onnx')
|
|
ok = True
|
|
except (InvalidGraph, TypeError, RuntimeError) as e:
|
|
# Probably a mismatch between onnxruntime and onnx version.
|
|
print(e)
|
|
ok = False
|
|
|
|
if ok:
|
|
print("The model expects input shape:", sess.get_inputs()[0].shape)
|
|
print("image shape:", ximg.shape)
|
|
|
|
#######################################
|
|
# Let's resize the image.
|
|
|
|
if ok:
|
|
from skimage.transform import resize
|
|
import numpy
|
|
|
|
ximg224 = resize(ximg / 255, (224, 224, 3), anti_aliasing=True)
|
|
ximg = ximg224[numpy.newaxis, :, :, :]
|
|
ximg = ximg.astype(numpy.float32)
|
|
|
|
print("new shape:", ximg.shape)
|
|
|
|
##################################
|
|
# Let's compute the output.
|
|
|
|
if ok:
|
|
input_name = sess.get_inputs()[0].name
|
|
res = sess.run(None, {input_name: ximg})
|
|
prob = res[0]
|
|
print(prob.ravel()[:10]) # Too big to be displayed.
|
|
|
|
|
|
##################################
|
|
# Let's get more comprehensive results.
|
|
|
|
if ok:
|
|
from keras.applications.densenet import decode_predictions
|
|
decoded = decode_predictions(prob)
|
|
|
|
import pandas
|
|
df = pandas.DataFrame(decoded[0], columns=["class_id", "name", "P"])
|
|
print(df)
|
|
|
|
|