VAI_C Usage - 1.2 English

Vitis AI User Guide (UG1414)

Document ID
UG1414
Release Date
2020-07-21
Version
1.2 English

The corresponding Vitis AI compiler for Caffe and TensorFlow framework are vai_c_caffe and vai_c_tensorflow across cloud-to-edge DPU. The common options for VAI_C are illustrated in the following table.

Table 1. VAI_C Common Options for Cloud and Edge DPU
Parameters Description
--arch DPU architecture configuration file for VAI_C compiler in JSON format. It contains the dedicated options for cloud and edge DPU during compilation.
--prototxt Path of Caffe prototxt file for the compiler vai_c_caffe. This option is only required while compiling the quantized Caffe model generated by vai_q_caffe.
--caffemodel Path of Caffe caffemodel file for the compiler vai_c_caffe. This option is only required while compiling the quantized Caffe model generated by vai_q_caffe.
--frozen_pb Path of TensorFlow frozen protobuf file for the compiler vai_c_tensorflow. This option is only required the quantized TensorFlow model generated by vai_q_tensorflow.
--output_dir Path of output directory of vai_c_caffe and vai_c_tensorflow after compilation process.
--net_name Name of DPU kernel for network model after compiled by VAI_C.
--options The list for the extra options for cloud or edge DPU in the format of 'key':'value'. If there are multiple options to be specified, they are separated by ‘,’, and if the extra option has no value, an empty string must be provided. For example:

--options "{'cpu_arch':'arm32', 'dcf':'/home/edge-dpu/zynq7020.dcf', 'save_kernel':''}"

Note: For arguments specified with “--options”, they have the highest priorities and will override the values specified in other places. For example, specifying ‘dcf’ with “--options” will replace the value specified in JSON file.