********************
Machine Learning
********************
Sitara Machine Learning toolkit brings machine learning to the edge by enabling 
machine learning inference on all Sitara devices (Arm only, Arm + specialized hardware 
accelerators). It is provided as part of TI's Processor SDK Linux, free to download 
and use. Sitara machine learning today consists of TI Deep Learning (TIDL), Neo-AI-DLR, 
TVM runtime, TensorFlow Lite, Arm NN, and RNN library.

.. figure:: ../images/Sitara_machine_learning_stack_diagram.png
    :width: 1120px
    :align: center
    :height: 576px

    Sitara Machine Learnining Offering

.. ifconfig:: CONFIG_part_family in ('General_family')

        .. rubric:: `TI Deep Learning (TIDL) <Foundational_Components/Machine_Learning/tidl.html>`__

        * Accelerates deep learning inference on C66x DSP cores and/or on Embedded Vision Engine (EVE) subsystems.
        * Available on AM57x device only.
        * Supports CNN at the moment, and imports Caffe, ONNX, and Tensorflow models.

.. rubric:: `Neo-AI Deep Learning Runtime (DLR) <Foundational_Components/Machine_Learning/neo.html>`__

* Neo-AI-DLR is a new open source machine learning runtime for on-device inference.
* Supports Keras, Tensorflow, TFLite, GluonCV, MXNet, Pytorch, ONNX, and XGBoost models optimized 
  automatically by `Amazon SageMaker Neo <https://aws.amazon.com/sagemaker/neo/>`__ 
  or `TVM compiler <https://github.com/apache/incubator-tvm/>`__.
* Supports all Cortex-A ARM cores (AM3x, AM4x, AM5x, AM6x Sitara devices).
* On AM5729 and AM5749 devices, uses TIDL to accelerate supported models automatically.

.. rubric:: `TVM Runtime <Foundational_Components/Machine_Learning/tvm.html>`__

* Open source deep learning runtime for on-device inference, supporting models compiled by 
  `TVM compiler <https://github.com/apache/incubator-tvm/>`__.
* Available on all Cortex-A ARM cores (AM3x, AM4x, AM5x, AM6x Sitara devices).

.. rubric:: `TensorFlow Lite <Foundational_Components/Machine_Learning/tflite.html>`__

* Open source deep learning runtime for on-device inference.
* Runs on all Cortex-A ARM cores (AM3x, AM4x, AM5x, AM6x Sitara devices).
* Imports Tensorflow Lite models.
* Uses TIDL import tool to create TIDL offloadable Tensorflow Lite models,
  which can be executed via Tensorflow Lite runtime with TIDL acceleration on AM5729 and AM5749 devices.

.. rubric:: `Arm NN <Foundational_Components/Machine_Learning/armnn.html>`__

* Open source inference engine available from Arm.
* Runs on all Cortex-A ARM cores (AM3x, AM4x, AM5x, AM6x Sitara devices).
* Imports Caffe, ONNX, TensorFlow, and TensorFlow Lite models.

.. rubric:: `RNN Library <Examples_and_Demos_Application_Demos.html#deploy-and-integration>`__

* Provides Long Short-Term Memory (LSTM) and fully connected layers in a standalone library to allow for rapid prototyping of inference applications that require Recurrent Neural Networks.
* Runs on all Cortex-A ARM cores (AM3x, AM4x, AM5x, AM6x Sitara devices).
* Integrated into TI's Processor SDK Linux in an OOB demo for `Predictive Maintenance <Examples_and_Demos_Application_Demos.html#predictive-maintenance-demo>`__.

.. toctree::
   :hidden:

   Foundational_Components/Machine_Learning/tidl
   Foundational_Components/Machine_Learning/neo
   Foundational_Components/Machine_Learning/tvm
   Foundational_Components/Machine_Learning/tflite
   Foundational_Components/Machine_Learning/armnn