`quark.onnx.equalization`#

Module Contents#

class quark.onnx.equalization.CLE_PAIR_TYPE#

Generic enumeration.

Derive from this class to define new enumerations.

class quark.onnx.equalization.Equalization(model: onnx.ModelProto, op_types_to_quantize: List[str], nodes_to_quantize: Optional[List[str]], nodes_to_exclude: Optional[List[str]])#: A class for layers equalization :param model: The ONNX model to be optimized. :type model: onnx.ModelProto :param op_types_to_quantize: A list of operation types to be quantized. :type op_types_to_quantize: list :param nodes_to_quantize: A list of node names to be quantized. :type nodes_to_quantize: list :param nodes_to_exclude: A list of node names to be excluded from quantization. :type nodes_to_exclude: list

quark.onnx.equalization.cle_transforms(model: onnx.ModelProto, op_types_to_quantize: List[str], nodes_to_quantize: List[str], nodes_to_exclude: List[str], cle_steps: int = -1, cle_balance_method: str = 'max', cle_weight_threshold: float = 0.5, cle_scale_append_bias: bool = True, cle_scale_use_threshold: bool = True, cle_total_layer_diff_threshold: float = 1.9e-07) → Any#: Equanlization transform models.