Name		Name	Last commit message	Last commit date
parent directory ..
license		license
supported_model_list		supported_model_list
README.md		README.md
create_ir_for_keras.ipynb		create_ir_for_keras.ipynb
create_ir_for_obj_det.ipynb		create_ir_for_obj_det.ipynb
create_ir_for_tfhub.ipynb		create_ir_for_tfhub.ipynb
ov-utils-arch.png		ov-utils-arch.png
ov_utils.py		ov_utils.py
remove-unused-conda-envs.sh		remove-unused-conda-envs.sh
requirements.txt		requirements.txt
sagemaker-instance-selection.png		sagemaker-instance-selection.png

README.md

OpenVINO™ Model Optimization in AWS Sagemaker

This demonstrates a OpenVINO utility to convert Keras or TFHub models or TF object detection models in AWS Sagemaker notebook instance.

Prerequisites

AWS Sagemaker notebook instance.
- Recommended instance type: ml.t3.xlarge
- Choose notebook-al2-v1 or notebook-al2-v2 as Platform identifier. al2 refers to Amazon Linux 2.

Instructions

Launch AWS Sagemaker Jupyer Notebook instance. See Prerequisites.
Upload the ov_utils.py and requirements.txt
Upload one of the jupyter notebook from this repo as per your requirement. See Examples.
Follow the instruction provided in Jupyter Notebook.

Examples

For Keras models, run create_ir_for_keras.ipynb. See Keras-SupportedModelList.md
For TFHub models, run create_ir_for_tfhub.ipynb. See TFHub-SupportedModelList.md
For Object detection models, run create_ir_for_obj_det.ipynb. See ObjDet-SupportedModelList.md