mirror of
https://github.com/saymrwulf/onnxruntime.git
synced 2026-07-14 18:12:05 +00:00
49 lines
No EOL
2.8 KiB
Markdown
49 lines
No EOL
2.8 KiB
Markdown
---
|
|
title: Accelerate TensorFlow
|
|
nav_order: 3
|
|
grand_parent: Tutorials
|
|
parent: Inferencing
|
|
---
|
|
# Accelerate TensorFlow model inferencing
|
|
{: .no_toc }
|
|
|
|
ONNX Runtime can accelerate inferencing times for TensorFlow, TFLite, and Keras models.
|
|
|
|
## Contents
|
|
{: .no_toc }
|
|
|
|
* TOC placeholder
|
|
{:toc}
|
|
|
|
## Get Started
|
|
* [End to end: Run TensorFlow models in ONNX Runtime](../tutorials/tf-get-started.md)
|
|
|
|
## Export model to ONNX
|
|
|
|
### TensorFlow
|
|
|
|
These examples use the [TensorFlow-ONNX converter](https://github.com/onnx/tensorflow-onnx), which supports TensorFlow 1, 2, Keras, and TFLite model formats.
|
|
|
|
* [TensorFlow: Object detection (efficentdet)](https://github.com/onnx/tensorflow-onnx/blob/master/tutorials/efficientdet.ipynb)
|
|
* [TensorFlow: Object detection (SSD Mobilenet)](https://github.com/onnx/tensorflow-onnx/blob/master/tutorials/ConvertingSSDMobilenetToONNX.ipynb)
|
|
* [TensorFlow: Image classification (efficientnet-edge)](https://github.com/onnx/tensorflow-onnx/blob/master/tutorials/efficientnet-edge.ipynb)
|
|
* [TensorFlow: Image classification (efficientnet-lite)](https://github.com/onnx/tensorflow-onnx/blob/master/tutorials/efficientnet-lite.ipynb)
|
|
* [TensorFlow: Natural Language Processing (BERT)](https://github.com/onnx/tensorflow-onnx/blob/master/tutorials/BertTutorial.ipynb)
|
|
|
|
### TFLite
|
|
* [TFLite: Image classifciation (mobiledet)](https://github.com/onnx/tensorflow-onnx/blob/master/tutorials/mobiledet-tflite.ipynb)
|
|
|
|
### Keras
|
|
Keras models can be converted using either the [tensorflow-onnx](https://github.com/onnx/tensorflow-onnx) or [Keras-ONNX converter](https://github.com/onnx/keras-onnx). The TensorFlow-ONNX converter supports newer opsets with more active support.
|
|
|
|
* [tf2onnx: Image classification (Resnet 50)](https://github.com/onnx/tensorflow-onnx/blob/master/tutorials/keras-resnet50.ipynb)
|
|
* [keras2onnx: Image classification (efficientnet)](https://github.com/onnx/keras-onnx/blob/master/tutorial/TensorFlow_Keras_EfficientNet.ipynb)
|
|
* [keras2onnx: Image classification (Densenet)](https://www.onnxruntime.ai/python/auto_examples/plot_dl_keras.html#sphx-glr-auto-examples-plot-dl-keras-py)
|
|
* [keras2onnx: Natural Language Processing (BERT)](https://github.com/microsoft/onnxruntime/tree/master/onnxruntime/python/tools/transformers/notebooks/Tensorflow_Keras_Bert-Squad_OnnxRuntime_CPU.ipynb)
|
|
* [keras2onnx: Handwritten Digit Recognition (MNIST)](https://github.com/onnx/keras-onnx/blob/master/tutorial/TensorFlow_Keras_MNIST.ipynb)
|
|
|
|
|
|
|
|
## Accelerate TensorFlow model inferencing
|
|
* [Accelerate BERT model on CPU](https://github.com/microsoft/onnxruntime/blob/master/onnxruntime/python/tools/transformers/notebooks/PyTorch_Bert-Squad_OnnxRuntime_CPU.ipynb)
|
|
* [Accelerate BERT model on GPU](https://github.com/microsoft/onnxruntime/blob/master/onnxruntime/python/tools/transformers/notebooks/PyTorch_Bert-Squad_OnnxRuntime_GPU.ipynb) |