#Release Introduction

The PTQ conversion sample release includes the model quantization sample package samples/onnx_ptq and the original models used for compilation under the model_zoo/mapper.

#Model Quantization Sample Package: onnx_ptq

After the original model and dataset for all samples are acquired, the contents of the model conversion sample package are shown below:

onnx_ptq/
├── common                                # This path contains the PTQ sample package generic script and data 
│   ├── calibration_data                 
│   ├── data -> /data/horizon_j6/data     # For storing soft links of the public datasets
│   ├── model_zoo -> ../../model_zoo      # For storing soft links of the model releases
│   ├── python
│   ├── test_data
│   └── tools
├── examples                              
│   ├── classification                    # Classification model examples
│   ├── detection                         # Detection model examples
│   └── segmentation                      # Segmentation model samples
└── tools
    ├── data_preprocess.py                # Data preprocessing script
    └── requirements.txt                  # Required dependencies for sample run

Main contents in the sample package:

Common Algorithm Model Samples are those in the classification/, detection/, and segmentation/ folders.

The main goal of these algorithm model samples is to help you:

• experience the model conversion process.
• experience the accuracy evaluation of model conversion.
• experience the model conversion results.

#Model Release: model_zoo

The model_zoo contains two paths: mapper and runtime.

• The mapper path contains the models (floating point models in ONNX or Caffe format) to be used for model conversion.

• The runtime path contains the hbm models that you will use for embedded runtime development.

The next section will introduce you the algorithm model samples.