save_quant_retrain_model

Applicability

Product	Supported
Atlas A3 training series products/Atlas A3 inference series products	x
Atlas A2 training products/Atlas A2 inference products	x
Atlas 200I/500 A2 inference product	√
Atlas inference series products	√
Atlas training products	√

Description

Inserts operators such as AscendQuant and AscendDequant into the retrained model and generates a fake-quantized model for accuracy simulation and a deployable model.

Restrictions

None

Prototype

save_quant_retrain_model(retrained_model_file, retrained_weights_file, save_type, save_path, scale_offset_record_file = None, config_file = None)

Parameters

Parameter	Input/Output	Description
retrained_model_file	Input	Definition file (.prototxt) of the retrained Caffe model. A string. Restrictions: For layers for inference, the settings in LayerParameter in retrained_model_file must meet inference requirements. For example, use_global_stats of the BatchNorm layer must be set to 1.
retrained_weights_file	Input	Weight file (.caffemodel) of the retrained Caffe model. A string.
save_type	Input	Type of the model to be saved. Fakequant: a model for accuracy simulation. Deploy: a deployable model that can be used for inference on the Ascend AI Processor. Both: both models. A string.
save_path	Input	Model save path. Must include the prefix of the model name, for example, *./quantized_model/model**. A string.
scale_offset_record_file	Input	File for storing quantization factors. A string. (Initialized to None by default.)
config_file	Input	Quantization configuration file. A string. (Initialized to None by default.)

Returns

None

Example

from amct_caffe import amct
retrained_model_file =  './pre_model/retrained_resnet50.prototxt'
retrained_weights_file = './pre_model/resnet50_solver_iter_35000.caffemodel'
scale_offset_record_file = './record.txt'
# Insert this API, and save the quantized model to a .prototxt model file and a .caffemodel weight file. The following files can be found in the ./result directory: model_fake_quant_model.prototxt, model_fake_quant_weights.caffemodel, model_deploy_model.prototxt, and model_deploy_weights.prototxt.
amct.save_quant_retrain_model(retrained_model_file, retrained_weights_file, 'Both', './result/model', scale_offset_record_file, config_json_file)

Flush files:

A fake-quantized model file for accuracy simulation in the Caffe environment and its weight file, with names containing the fake_quant keyword.
A deployable model file and its weight file, with names containing the deploy keyword. The model can be deployed on the Ascend AI Processor after being converted by ATC.

Parent topic: QAT APIs