Simplified Configuration File for Auto Channel Pruning Search

Find the basic_info.proto file in /amct_tensorflow/proto/basic_info.proto under the AMCT installation directory. The file content is as follows:

Message	Required	Type	Parameter	Description
AutoMixedPrecisionConfig	-	-	-	AMCT simplified configuration for automatic mixed precision search. The current version does not support this feature.
	optional	float	compress_ratio	Compression ratio. The computation amount of all quantizable layers is used as a reference compression multiple.
	repeated	QuantBitLimit	quant_bit_limit	Quantization bit width search range of some layers.
	optional	string	ptq_cfg	Simplified PTQ configuration file, which is used to obtain the quantization factors under the quantization bit widths of INT4 and INT8 during calibration. If this parameter is not set, the default PTQ configuration is used. Currently, only INT8 quantization is supported.
	optional	int64	test_iteration	Number of batches of the dump data. The data is used to measure the quantization impact and computation amount. The data volume should be representative.
	optional	string	override_qat_cfg	Simplified configuration file for QAT. The output of the automatic mixed precision search overwrites the bit width of the layer, and other parameters remain unchanged. If this parameter is not set, the simplified quantization aware training configuration file (in .proto format) is used to generate a .cfg configuration file with quantization bit width information.
AutoChannelPruneConfig	-	-	-	AMCT simplified configuration for auto channel pruning search
	required	float	compress_ratio	Compression ratio. The computation amount of all quantizable layers is used as a reference compression multiple.
	optional	bool	ascend_optimized	Whether to perform adaptation to Ascend platforms. If the pruned model is to be deployed on Ascend AI Processor, you are advised to set this parameter to true.
	optional	float	max_prune_ratio	Maximum sparsity rate of a single layer, which is the maximum sparsity rate in the sparsity configuration output by the API. The default value is 1.
	optional	int64	test_iteration	Batch number of the input test data.
	optional	string	override_prune_cfg	Simplified configuration file for sparsity of a specified channel. Only the skip and override configurations can be included. The configured layer uses the specified configuration and will not be overridden by the automatic channel sparsity search API.
QuantBitLimit	-	-	-	Quantization bit width search range of some layers.
	optional	string	layer_name	Layer name.
	repeated	DataType	data_range	Quantization bit width range.
DataType	-	-	-	Quantization bit width range. Enumeration Types Currently, only INT8 quantization is supported.
	-	-	FLOAT	Floating point, not quantized.
	-	-	INT8	INT8 quantization
	-	-	INT4	INT4 quantization

The following is an example of the simplified configuration file (amc.cfg) for auto channel pruning search:

compress_ratio: 1.5
ascend_optimized: true
max_prune_ratio: 0.8
test_iteration: 1
override_prune_cfg: 'your/path/to/override_channel_prune.cfg'

Parent topic: See Also