EN|CH

tools

Introduction

Ascend tools
Please go to the corresponding folders to get the tools according to your needs, or click the link below to select the tool you want.

Description

1. msame

   Model inference tool: Input the OM model and the input BIN file required by the model to output the model output data file.

2. img2bin

   BIN file generation tool: Generate input data required for model inference, saved in .bin format.

3. makesd

   makesd tool：card making tool package. Provide the card making function under Ubuntu.

4. configure_usb_ethernet
   USB virtual NIC connection script：Configure the IP address of the USB NIC.

5. pt2pb

   Tool for converting PyTorch models to TensorFlow PB models：Input the PyTorch weight parameters model and convert the model to the ONNX format and then to the PB format.

6. dnmetis

   Test tool for NPU inference accuracy and performance：Use Python to encapsulate AscendCL C++ APIs. Input the OM model, dataset images, and labels to perform model inference and output accuracy and performance data.

7. msquickcmp

   One-click full-process accuracy comparison tool：This tool applies to TensorFlow and ONNX models. Input the original model and the corresponding offline OM model to output the accuracy comparison result.

8. dockerimages
   Accuracy analysis tool：The toolkit provides common accuracy comparison functions. Currently, this tool is mainly used in TensorFlow training scenarios and provides interactive query and operation entries for dump data and graph information.

9. cann-benchmark_infer_scripts
   Model pre-processing and post-processing scripts corresponding to the cann-benchmark inference software：This tool contains the model processing scripts of the cann-benchmark inference tool, including the result parsing script and pre-processing and post-processing scripts. These scripts need to be used according to the cann-benchmark guide.

10. tfdbg_ascend
    Tensorflow2.x dump tool：This tool provides the data dump capability when Tensorflow2.x is running on the CPU/GPU platform.

11. intelligent_edge_tools
    pod_deploy_tool: This directory contains the tools which apply to intelligent edge equipment.

12. mindxedge_whitebox
    mindxedge_whitebox install tool: This directory contains the tools which is whitebox installation of the Atlas 500 edge equipment.