saymrwulf/onnxruntime
1
0
Fork 0
mirror of https://github.com/saymrwulf/onnxruntime.git synced 2026-07-27 20:02:15 +00:00
Code Issues Projects Releases Packages Wiki Activity Actions

Modify the code to get correct ragne for symmetric quantization (#10170)

Browse source
This commit is contained in:
Chi Lo 2022-01-03 19:13:37 -08:00 committed by GitHub
parent 0c517112c4
commit c29397ad4f
No known key found for this signature in database
GPG key ID: 4AEE18F83AFDEB23
2 changed files with 6 additions and 6 deletions
Show stats Download patch file Download diff file Expand all files Collapse all files

2
onnxruntime/python/tools/quantization/onnx_quantizer.py
View file

@ -860,7 +860,7 @@ class ONNXQuantizer:
quantization_params = {}
for tensor_name in self.tensors_range.keys():
rmin, rmax = self.tensors_range[tensor_name]
qmin, qmax = get_qmin_qmax_for_qType(self.input_qType)
qmin, qmax = get_qmin_qmax_for_qType(self.input_qType, symmetric=self.is_activation_symmetric)
quantization_params[tensor_name] = compute_scale_zp(rmin, rmax,
qmin, qmax,

10
onnxruntime/python/tools/quantization/quant_utils.py
View file

@ -176,7 +176,7 @@ def quantize_data(data, qType, symmetric, reduce_range=False):
if len(data):
rmin = min(data)
rmax = max(data)
qmin, qmax = get_qmin_qmax_for_qType(qType, reduce_range, for_weight=True)
qmin, qmax = get_qmin_qmax_for_qType(qType, reduce_range, symmetric=symmetric)
zero_point, scale = compute_scale_zp(rmin, rmax, qmin, qmax, symmetric)
@ -184,7 +184,7 @@ def quantize_data(data, qType, symmetric, reduce_range=False):
return rmin, rmax, zero_point, scale, quantized_data
def get_qmin_qmax_for_qType(qType, reduce_range=False, for_weight=False):
def get_qmin_qmax_for_qType(qType, reduce_range=False, symmetric=False):
'''
Return qmin and qmax, the minimum and maximum value representable by the given qType
:parameter qType: onnx.onnx_pb.TensorProto.UINT8 or onnx.onnx_pb.TensorProto.UINT8
@ -193,7 +193,7 @@ def get_qmin_qmax_for_qType(qType, reduce_range=False, for_weight=False):
if qType == onnx_proto.TensorProto.UINT8:
(qmin, qmax) = (0,127) if reduce_range else (0,255)
elif qType == onnx_proto.TensorProto.INT8:
if for_weight:
if symmetric:
(qmin, qmax) = (-64,64) if reduce_range else (-127,127)
else:
(qmin, qmax) = (-64,64) if reduce_range else (-128,127)
@ -201,13 +201,13 @@ def get_qmin_qmax_for_qType(qType, reduce_range=False, for_weight=False):
raise ValueError("Unexpected data type {} requested. Only INT8 and UINT8 are supported.".format(qType))
return qmin, qmax
def get_qrange_for_qType(qType, reduce_range=False, for_weight=False):
def get_qrange_for_qType(qType, reduce_range=False, symmetric=False):
'''
Helper function to get the quantization range for a type.
parameter qType: quantization type.
return: quantization range.
'''
qmin, qmax = get_qmin_qmax_for_qType(qType, reduce_range, for_weight=for_weight)
qmin, qmax = get_qmin_qmax_for_qType(qType, reduce_range, symmetric=symmetric)
return qmax - qmin
class QuantizedInitializer: