This document has instructions for running SSD-ResNet34 Int8 inference using Intel-optimized TensorFlow.
SSD-ResNet34 uses the COCO dataset for accuracy testing.
Download and preprocess the COCO validation images using the instructions here. After the script to convert the raw images to the TF records file completes, rename the tf_records file:
mv ${OUTPUT_DIR}/coco_val.record ${OUTPUT_DIR}/validation-00000-of-00001
Set the DATASET_DIR to the folder that has the validation-00000-of-00001
file when running the accuracy test. Note that the inference performance
test uses synthetic dataset.
| Script name | Description |
|---|---|
| int8_accuracy.sh | Tests accuracy using the COCO dataset in the TF Records format with an input size of 300x300. |
| int8_inference.sh | Run inference using synthetic data with an input size of 300x300 and outputs performance metrics. |
| int8_accuracy_1200.sh | Tests accuracy using the COCO dataset in the TF Records format with an input size of 1200x1200. |
| multi_instance_batch_inference_1200.sh | Uses numactl to run inference (batch_size=1) with an input size of 1200x1200 and one instance per socket. Waits for all instances to complete, then prints a summarized throughput value. |
| multi_instance_online_inference_1200.sh | Uses numactl to run inference (batch_size=1) with an input size of 1200x1200 and four cores per instance. Waits for all instances to complete, then prints a summarized throughput value. |
Setup your environment using the instructions below, depending on if you are using AI Kit:
| Setup using AI Kit on Linux | Setup without AI Kit on Linux | Setup without AI Kit on Windows |
|---|---|---|
|
To run using AI Kit on Linux you will need:
|
To run without AI Kit on Linux you will need:
|
To run without AI Kit on Windows you will need:
|
In addition to the libraries above, SSD-ResNet34 uses the
TensorFlow models and
TensorFlow benchmarks
repositories. Clone the repositories using the commit ids specified
below and set the TF_MODELS_DIR to point to the folder where the models
repository was cloned:
# Clone the TensorFlow models repo
git clone https://github.com/tensorflow/models.git tf_models
cd tf_models
git checkout f505cecde2d8ebf6fe15f40fb8bc350b2b1ed5dc
export TF_MODELS_DIR=$(pwd)
cd ..
# Clone the TensorFlow benchmarks repo
git clone --single-branch https://github.com/tensorflow/benchmarks.git ssd-resnet-benchmarks
cd ssd-resnet-benchmarks
git checkout 509b9d288937216ca7069f31cfb22aaa7db6a4a7
cd ..
Download the SSD-ResNet34 pretrained model for either the 300x300 or 1200x1200
input size, depending on which quickstart script you are
going to run. Set the PRETRAINED_MODEL environment variable for the path to the
pretrained model that you'll be using.
If you run on Windows, please use a browser to download the pretrained model using the link below.
For Linux, run:
# ssd-resnet34 300x300
wget https://storage.googleapis.com/intel-optimized-tensorflow/models/v1_8/ssd_resnet34_int8_bs1_pretrained_model.pb
export PRETRAINED_MODEL=$(pwd)/ssd_resnet34_int8_bs1_pretrained_model.pb
# ssd-resnet34 1200x1200
wget https://storage.googleapis.com/intel-optimized-tensorflow/models/v1_8/ssd_resnet34_int8_1200x1200_pretrained_model.pb
export PRETRAINED_MODEL=$(pwd)/ssd_resnet34_int8_1200x1200_pretrained_model.pb
After installing the prerequisites and cloning the models and benchmarks
repos, and downloading the pretrained model, set the required environment
variables. Set an environment variable for the path to an OUTPUT_DIR
where log files will be written. If the accuracy test is being run, then
also set the DATASET_DIR to point to the folder where the COCO dataset
validation-00000-of-00001 file is located. Once the environment variables
are set, you can then run a quickstart script from the
Model Zoo on either Linux or Windows.
# cd to your model zoo directory
cd models
# set environment variables
export DATASET_DIR=<directory with the validation-*-of-* files (for accuracy testing only)>
export TF_MODELS_DIR=<path to the TensorFlow Models repo>
export PRETRAINED_MODEL=<path to the 300x300 or 1200x1200 pretrained model pb file>
export OUTPUT_DIR=<directory where log files will be written>
./quickstart/object_detection/tensorflow/ssd-resnet34/inference/cpu/int8/<script name>.sh
Using cmd.exe, run:
# cd to your model zoo directory
cd models
set PRETRAINED_MODEL=<path to the 300x300 or 1200x1200 pretrained model pb file>
set DATASET_DIR=<directory with the validation-*-of-* files (for accuracy testing only)>
set OUTPUT_DIR=<directory where log files will be written>
set TF_MODELS_DIR=<path to the TensorFlow Models repo>
bash quickstart\object_detection\tensorflow\ssd-resnet34\inference\cpu\int8\<script name>.sh
Note: You may use
cygpathto convert the Windows paths to Unix paths before setting the environment variables. As an example, if the dataset location on Windows isD:\user\coco_dataset, convert the Windows path to Unix as shown:cygpath D:\user\coco_dataset /d/user/coco_datasetThen, set the
DATASET_DIRenvironment variableset DATASET_DIR=/d/user/coco_dataset.
- To run more advanced use cases, see the instructions here
for calling the
launch_benchmark.pyscript directly. - To run the model using docker, please see the oneContainer
workload container:
https://software.intel.com/content/www/us/en/develop/articles/containers/ssd-resnet34-int8-inference-tensorflow-container.html.