From 8ea6c2d43394ce5f8dd4bc136f897e2df7a63d07 Mon Sep 17 00:00:00 2001 From: Tracy Narine Date: Wed, 17 Jan 2024 13:14:28 +0000 Subject: IVGCVSW-8202 Update documents after TF 2.15 update * Updated .md files with new version number Signed-off-by: Tracy Narine Change-Id: I351d141838b95aca29f3ebe43f7e2f9944ec3417 --- BuildGuideAndroidNDK.md | 2 +- delegate/DelegateQuickStartGuide.md | 10 +++++----- samples/ObjectDetection/Readme.md | 8 ++++---- 3 files changed, 10 insertions(+), 10 deletions(-) diff --git a/BuildGuideAndroidNDK.md b/BuildGuideAndroidNDK.md index eea70b3e49..1c1864db70 100644 --- a/BuildGuideAndroidNDK.md +++ b/BuildGuideAndroidNDK.md @@ -170,7 +170,7 @@ First clone Tensorflow manually and check out the version Arm NN was tested with cd $WORKING_DIR git clone https://github.com/tensorflow/tensorflow.git cd tensorflow -git fetch && git checkout v2.14.0 +git fetch && git checkout v2.15.0 ``` Or use the script that Arm NN provides: ```bash diff --git a/delegate/DelegateQuickStartGuide.md b/delegate/DelegateQuickStartGuide.md index d6fc487f2b..9e4d3e9106 100644 --- a/delegate/DelegateQuickStartGuide.md +++ b/delegate/DelegateQuickStartGuide.md @@ -36,11 +36,11 @@ print(output_data) # Prepare the environment Pre-requisites: - * Dynamically build Arm NN Delegate library or download the Arm NN binaries (built with a particular SHA of Tensorflow v2.14.0, which is 4dacf3f368eb7965e9b5c3bbdd5193986081c3b2) + * Dynamically build Arm NN Delegate library or download the Arm NN binaries (built with a particular SHA of Tensorflow v2.15.0, which is 6887368d6d46223f460358323c4b76d61d1558a8) * python3 (Depends on TfLite version) * virtualenv * numpy (Depends on TfLite version) - * tflite_runtime (v2.14.0 currently available) + * tflite_runtime (v2.15.0 currently available) If you haven't built the delegate yet then take a look at the [build guide](./BuildGuideNative.md). Otherwise, you can download the binaries [here](https://github.com/ARM-software/armnn/releases/). Set the following environment variable to the location of the .so binary files: @@ -50,7 +50,7 @@ export LD_LIBRARY_PATH= We recommend creating a virtual environment for this tutorial. For the following code to work python3 is needed. Please also check the documentation of the TfLite version you want to use. There might be additional prerequisites for the python -version. We will use Tensorflow Lite 2.14.0 for this guide. +version. We will use Tensorflow Lite 2.15.0 for this guide. ```bash # Install python3 (We ended up with python3.5.3) and virtualenv sudo apt-get install python3-pip @@ -74,10 +74,10 @@ mobile and embedded devices. The TfLite [website](https://www.tensorflow.org/lite/guide/python) shows you two methods to download the `tflite_runtime` package. In our experience, the use of the pip command works for most systems including debian. However, if you're using an older version of Tensorflow, you may need to build the pip package from source. You can find more information [here](https://github.com/tensorflow/tensorflow/blob/master/tensorflow/lite/tools/pip_package/README.md). -But in our case, with Tensorflow Lite 2.14.0, we can install through: +But in our case, with Tensorflow Lite 2.15.0, we can install through: ``` -pip3 install --extra-index-url https://google-coral.github.io/py-repo/ tflite_runtime==2.14.0 +pip3 install --extra-index-url https://google-coral.github.io/py-repo/ tflite_runtime==2.15.0 ``` Your virtual environment is now all setup. Copy the final python script into a python file e.g. diff --git a/samples/ObjectDetection/Readme.md b/samples/ObjectDetection/Readme.md index 0cc41eb785..29d627c0b2 100644 --- a/samples/ObjectDetection/Readme.md +++ b/samples/ObjectDetection/Readme.md @@ -20,7 +20,7 @@ with detections shown in bounding boxes, class labels and confidence. This example utilizes OpenCV functions to capture and output video data. 1. Public Arm NN C++ API is provided by Arm NN library. 2. For Delegate file mode following dependencies exist: -2.1 Tensorflow version 2.14 +2.1 Tensorflow version 2.15 2.2 Flatbuffers version 23.5.26 2.3 Arm NN delegate library @@ -97,7 +97,7 @@ Please see [find_opencv.cmake](./cmake/find_opencv.cmake) for implementation det ### Tensorflow Lite (Needed only in delegate file mode) -This application uses [Tensorflow Lite)](https://www.tensorflow.org/) version 2.14 for demonstrating use of 'armnnDelegate'. +This application uses [Tensorflow Lite)](https://www.tensorflow.org/) version 2.15 for demonstrating use of 'armnnDelegate'. armnnDelegate is a library for accelerating certain TensorFlow Lite operators on Arm hardware by providing the TensorFlow Lite interpreter with an alternative implementation of the operators via its delegation mechanism. You may clone and build Tensorflow lite and provide the path to its root and output library directories through the cmake @@ -106,13 +106,13 @@ For implementation details see the scripts FindTfLite.cmake and FindTfLiteSrc.cm The application links with the Tensorflow lite library libtensorflow-lite.a -#### Download and build Tensorflow Lite version. We currently use Tf 2.14 for the Cmake build. +#### Download and build Tensorflow Lite version. We currently use Tf 2.15 for the Cmake build. Example for Tensorflow Lite native compilation ```commandline sudo apt install build-essential git clone https://github.com/tensorflow/tensorflow.git cd tensorflow/tensorflow -git checkout v2.14.0 +git checkout v2.15.0 mkdir build && cd build cmake ../lite -DTFLITE_ENABLE_XNNPACK=OFF make -- cgit v1.2.1