More Features (Model Adaptation)

If you have the quantization factor calculated by yourself and the original Caffe model, the model cannot be converted into an offline model adapted to the AI processor by using ATC. You need to use the function provided in this section to convert the model into the CANN quantization model format before using ATC to convert the CANN quantization model into an offline model adapted to the AI processor.

Adaptation Principles

The following figure shows the adaptation principles. The user implements the operations in blue, while those in gray are implemented by using the convert_model API in AMCT. Specifically, import the package to the source Caffe network inference code and call APIs where appropriate for model adaptation. For details about the adaptation example in this scenario, see "Model Adaptation" in Sample List.

Figure 1 Model adaptation principles

Example

Take the following steps to get started. Update the sample code based on your situation.
To reuse the following code for quantizing a different model, prepare the original model and build quantization factor record file based on user-defined quantization factors yourself.

Import the AMCT package and set the log level (see Setting Environment Variables for details).
1

import amct_caffe as amct

Set the run mode and target device.

AMCT runs on the CPU or GPU. To run the tool on the GPU, you first need to configure the Caffe run mode and target device before configuring AMCT's run mode. Since the target device has already been specified here, you do not need to configure the target device in the model inference function.

        
             if 'gpu':
    caffe.set_mod_gpu()
    caffe.set_device(gpu_id)
    amct.set_gpu_mode()
else:
    caffe.set_mode_cpu()

(Optional) Run inference on the original model in the Caffe environment based on the test dataset to validate the inference script and environment setup. (Update the sample code based on your situation.)

This step is recommended as it guarantees a properly functioning original model for inference with acceptable accuracy. You can use a subset from the test dataset to improve the efficiency.

        
             user_test_model(ori_model_file, ori_weights_file, test_data, test_iterations)

Call the convert_model API to perform model adaptation.

This API parses the original model into a graph, preprocesses the graph, parses the input quantization factor record file, inserts operators such as AscendQuant and AscendDequant into the modified graph based on the quantization factors, and then saves the quantized model.

         
              quant_model_path = './result/user_model'
record_file = './result/record.txt'
amct.convert_model(model_file=ori_model_file,
		   weights_file=ori_weights_file,
		   scale_offset_record_file=record_file,
		   save_path=quant_model_path)

(Optional) Run inference on the fake-quantized models fake_quant_model and fake_quant_weights in the Caffe environment based on the test dataset to test the accuracy. (Update the sample code based on your situation.)

Compare the accuracy of the fake-quantized model with that of the original model (see 3).

         
              fake_quant_model = './result/user_model_fake_quant_model.prototxt'
fake_quant_weights = './result/user_model_fake_quant_weights.caffemodel'
user_test_model(fake_quant_model, fake_quant_weights, test_data, test_iterations)

Parent topic: AMCT (Caffe)