"""
This module implements observers which are used to collect statistics about
the values observed during calibration (PTQ) or training (QAT).
"""
import re
import warnings
from abc import ABCMeta, abstractmethod
from collections import OrderedDict
from functools import partial
from typing import Any, List, Tuple, Optional, Dict, Union
import torch
import torch.nn as nn
from torch.ao.quantization.utils import check_min_max_valid, calculate_qmin_qmax
class _PartialWrapper(object):
def __init__(self, p):
self.p = p
self.callable_args = {}
def __call__(self, *args, **keywords):
# call each arg in callable_args and add them partial, then run with keywords
# skip if arg_name in keywords so its possible to overwrite
for arg_name in self.callable_args:
if arg_name not in keywords:
keywords = {**keywords, **{arg_name: self.callable_args[arg_name]()}}
return self.p(*args, **keywords)
def __repr__(self):
return self.p.__repr__() + self.callable_args.__repr__()
def with_args(self, **kwargs):
return _with_args(self, **kwargs)
def with_callable_args(self, **kwargs):
result = _PartialWrapper(p=self.p)
result.callable_args = {**self.callable_args, **kwargs}
return result
def _with_args(cls_or_self, **kwargs):
r"""Wrapper that allows creation of class factories.
This can be useful when there is a need to create classes with the same
constructor arguments, but different instances. Can be used in conjunction with
_callable_args
Example::
>>> Foo.with_args = classmethod(_with_args)
>>> foo_builder = Foo.with_args(a=3, b=4).with_args(answer=42)
>>> foo_instance1 = foo_builder()
>>> foo_instance2 = foo_builder()
>>> id(foo_instance1) == id(foo_instance2)
False
"""
r = _PartialWrapper(partial(cls_or_self, **kwargs))
return r
def _with_callable_args(cls_or_self, **kwargs):
r"""Wrapper that allows creation of class factories args that need to be
called at construction time.
This can be useful when there is a need to create classes with the same
constructor arguments, but different instances and those arguments should only
be calculated at construction time. Can be used in conjunction with _with_args
Example::
>>> Foo.with_callable_args = classmethod(_with_callable_args)
>>> Foo.with_args = classmethod(_with_args)
>>> foo_builder = Foo.with_callable_args(cur_time=get_time_func).with_args(name="dan")
>>> foo_instance1 = foo_builder()
>>> wait 50
>>> foo_instance2 = foo_builder()
>>> id(foo_instance1.creation_time) == id(foo_instance2.creation_time)
False
"""
r = _PartialWrapper(partial(cls_or_self))
return r.with_callable_args(**kwargs)
ABC: Any = ABCMeta(str("ABC"), (object,), {}) # compatible with Python 2 *and* 3:
[文档]class ObserverBase(ABC, nn.Module):
r"""Base observer Module.
Any observer implementation should derive from this class.
Concrete observers should follow the same API. In forward, they will update
the statistics of the observed Tensor. And they should provide a
`calculate_qparams` function that computes the quantization parameters given
the collected statistics.
Args:
dtype: Quantized data type
"""
def __init__(self, dtype):
super(ObserverBase, self).__init__()
self.dtype = dtype
@abstractmethod
def forward(self, x):
pass
@abstractmethod
def calculate_qparams(self, **kwargs):
pass
with_args = classmethod(_with_args)
with_callable_args = classmethod(_with_callable_args)
[文档]class _ObserverBase(ObserverBase):
r"""Internal common base for all qint/quint8 observers.
This base is for commonly used parameters used internally.
Users should use `~torch.ao.quantization.observer.ObserverBase` as a base class
for custom observers.
Args:
dtype: Quantized data type.
qscheme: Quantization scheme to be used.
reduce_range: Reduces the range of the quantized data type by 1 bit.
This is sometimes required to avoid instruction overflow.
quant_min: Minimum quantization value. If unspecified, it will follow the 8-bit setup.
quant_max: Maximum quantization value. If unspecified, it will follow the 8-bit setup.
.. warning::
:attr:`dtype` can only take ``torch.qint8`` or ``torch.quint8``.
.. warning::
:attr:`qscheme` can only take one of the following options:
- ``torch.per_tensor_affine``
- ``torch.per_tensor_symmetric``
- ``torch.per_channel_affine``
- ``torch.per_channel_symmetric``
"""
# Note: the version is shared by all observer types
#
# Version 1/None
# self
#
# Version 2 (base class only, does not include child class buffers)
# self
# |--- eps : Tensor
#
# Version 3
# for HistogramObserver only, changed the shape of uninitialized
# min_val and max_val buffers from torch.Size([0]) to torch.Size([])
# for PerChannelObservers, changed the name of the buffers from min_vals
# to min_val and from max_vals to max_val.
_version = 3
eps: torch.Tensor
def __init__(
self,
dtype=torch.quint8,
qscheme=torch.per_tensor_affine,
reduce_range=False,
quant_min=None,
quant_max=None,
factory_kwargs=None,
) -> None:
factory_kwargs = torch.nn.factory_kwargs(factory_kwargs)
super(_ObserverBase, self).__init__(dtype=dtype)
self.qscheme = qscheme
if reduce_range:
warnings.warn(
"Please use quant_min and quant_max to specify the range for observers. \
reduce_range will be deprecated in a future release of PyTorch."
)
self.reduce_range = reduce_range
self.register_buffer(
"eps", torch.tensor([torch.finfo(torch.float32).eps], **factory_kwargs)
)
assert self.qscheme in (
torch.per_tensor_affine,
torch.per_tensor_symmetric,
torch.per_channel_affine,
torch.per_channel_symmetric,
torch.per_channel_affine_float_qparams,
), "Default Observer only works for per_tensor_affine, \
per_tensor_symmetric, per_channel_affine, \
per_channel_symmetric and per_channel_float_qparams quantization scheme"
assert self.dtype in (
torch.qint8,
torch.quint8,
torch.quint4x2,
), "Default Observer only works for qint8, quint8 and quint4x2 data type"
self.has_customized_qrange = (quant_min is not None) and (quant_max is not None)
if self.has_customized_qrange:
self._validate_qmin_qmax(quant_min, quant_max)
self.quant_min, self.quant_max = \
calculate_qmin_qmax(quant_min, quant_max, self.has_customized_qrange, self.dtype, self.reduce_range)
def _load_from_state_dict(
self,
state_dict,
prefix,
local_metadata,
strict,
missing_keys,
unexpected_keys,
error_msgs,
):
version = local_metadata.get("version", None)
if version is None or version == 1:
# eps was moved to a buffer in version 2
eps = torch.tensor([torch.finfo(torch.float32).eps])
state_dict[prefix + "eps"] = eps
super(ObserverBase, self)._load_from_state_dict(
state_dict,
prefix,
local_metadata,
strict,
missing_keys,
unexpected_keys,
error_msgs,
)
@torch.jit.export
def _validate_qmin_qmax(self, quant_min: int, quant_max: int) -> None:
r"""Validates that the user-specified quantization range is properly initialized
and within the given bound supported by the observer dtype.
To accommodate lower-bit quantization with respect to the existing torch.qint8 and
torch.quint8 datatypes, the user can choose to use dynamic quantization range by passing
in a tuple of initial qmin and qmax values. One use case is these customized qmin and qmax
values are used to calculate static estimates of the scale and zero point for aggressive lower-bit
fake quantization. These estimates are compared against parameters learned through backpropagation.
The related literatures for scale and zero point via backpropagation are as follows:
Learned Step Size Quantization: https://openreview.net/pdf?id=rkgO66VKDS
Trained Quantization Thresholds: https://arxiv.org/pdf/1903.08066.pdf
"""
# The variable names are prefixed with "initial" because their values (qmin and qmax) might be adjusted
# based on whether quantization range is reduced and the datatype (signed/unsigned) used by the observer.
assert (
quant_min <= 0 <= quant_max
), "Used-specified quantization range must include 0."
assert (
quant_min < quant_max
), "qmin must be strictly less than qmax for user-specified quantization range."
@torch.jit.export
def _calculate_qparams(
self, min_val: torch.Tensor, max_val: torch.Tensor
) -> Tuple[torch.Tensor, torch.Tensor]:
r"""Calculates the quantization parameters, given min and max
value tensors. Works for both per tensor and per channel cases
Args:
min_val: Minimum values per channel
max_val: Maximum values per channel
Returns:
scales: Scales tensor of shape (#channels,)
zero_points: Zero points tensor of shape (#channels,)
"""
if not check_min_max_valid(min_val, max_val):
return torch.tensor([1.0], device=min_val.device.type), torch.tensor([0], device=min_val.device.type)
quant_min, quant_max = self.quant_min, self.quant_max
min_val_neg = torch.min(min_val, torch.zeros_like(min_val))
max_val_pos = torch.max(max_val, torch.zeros_like(max_val))
device = min_val_neg.device
scale = torch.ones(min_val_neg.size(), dtype=torch.float32, device=device)
zero_point = torch.zeros(min_val_neg.size(), dtype=torch.int64, device=device)
if (
self.qscheme == torch.per_tensor_symmetric
or self.qscheme == torch.per_channel_symmetric
):
max_val_pos = torch.max(-min_val_neg, max_val_pos)
scale = max_val_pos / (float(quant_max - quant_min) / 2)
scale = torch.max(scale, self.eps)
if self.dtype == torch.quint8:
if self.has_customized_qrange:
# When customized quantization range is used, down-rounded midpoint of the range is chosen.
zero_point = zero_point.new_full(
zero_point.size(), (quant_min + quant_max) // 2
)
else:
zero_point = zero_point.new_full(zero_point.size(), 128)
elif self.qscheme == torch.per_channel_affine_float_qparams:
scale = (max_val - min_val) / float(quant_max - quant_min)
scale = torch.where(scale > self.eps, scale, torch.ones_like(scale))
# We use the quantize function
# xq = Round(Xf * inv_scale + zero_point),
# setting zero_point to (-1 * min *inv_scale) we get
# Xq = Round((Xf - min) * inv_scale)
zero_point = -1 * min_val / scale
else:
scale = (max_val_pos - min_val_neg) / float(quant_max - quant_min)
scale = torch.max(scale, self.eps)
zero_point = quant_min - torch.round(min_val_neg / scale).to(torch.int)
zero_point = torch.clamp(zero_point, quant_min, quant_max)
# For scalar values, cast them to Tensors of size 1 to keep the shape
# consistent with default values in FakeQuantize.
if len(scale.shape) == 0:
# TODO: switch to scale.item() after adding JIT support
scale = torch.tensor([float(scale)], dtype=scale.dtype, device=device)
if len(zero_point.shape) == 0:
# TODO: switch to zero_point.item() after adding JIT support
zero_point = torch.tensor(
[int(zero_point)], dtype=zero_point.dtype, device=device
)
if self.qscheme == torch.per_channel_affine_float_qparams:
zero_point = torch.tensor(
[float(zero_point)], dtype=zero_point.dtype, device=device
)
return scale, zero_point
@torch.jit.export
def reset_min_max_vals(self):
raise NotImplementedError("Cannot reset min/max values in the given observer.")
[文档]class MinMaxObserver(_ObserverBase):
r"""Observer module for computing the quantization parameters based on the
running min and max values.
This observer uses the tensor min/max statistics to compute the quantization
parameters. The module records the running minimum and maximum of incoming
tensors, and uses this statistic to compute the quantization parameters.
Args:
dtype: Quantized data type
qscheme: Quantization scheme to be used
reduce_range: Reduces the range of the quantized data type by 1 bit
quant_min: Minimum quantization value. If unspecified, it will follow the 8-bit setup.
quant_max: Maximum quantization value. If unspecified, it will follow the 8-bit setup.
memoryless: Boolean that controls whether observer removes old data when a new input is seen.
This is most useful for simulating dynamic quantization, especially during QAT.
Given running min/max as :math:`x_\text{min}` and :math:`x_\text{max}`,
scale :math:`s` and zero point :math:`z` are computed as:
The running minimum/maximum :math:`x_\text{min/max}` is computed as:
.. math::
\begin{array}{ll}
x_\text{min} &= \begin{cases}
\min(X) & \text{if~}x_\text{min} = \text{None} \\
\min\left(x_\text{min}, \min(X)\right) & \text{otherwise}
\end{cases}\\
x_\text{max} &= \begin{cases}
\max(X) & \text{if~}x_\text{max} = \text{None} \\
\max\left(x_\text{max}, \max(X)\right) & \text{otherwise}
\end{cases}\\
\end{array}
where :math:`X` is the observed tensor.
The scale :math:`s` and zero point :math:`z` are then computed as:
.. math::
\begin{aligned}
\text{if Symmetric:}&\\
&s = 2 \max(|x_\text{min}|, x_\text{max}) /
\left( Q_\text{max} - Q_\text{min} \right) \\
&z = \begin{cases}
0 & \text{if dtype is qint8} \\
128 & \text{otherwise}
\end{cases}\\
\text{Otherwise:}&\\
&s = \left( x_\text{max} - x_\text{min} \right ) /
\left( Q_\text{max} - Q_\text{min} \right ) \\
&z = Q_\text{min} - \text{round}(x_\text{min} / s)
\end{aligned}
where :math:`Q_\text{min}` and :math:`Q_\text{max}` are the minimum and
maximum of the quantized data type.
.. warning:: :attr:`dtype` can only take ``torch.qint8`` or ``torch.quint8``.
.. note:: If the running minimum equals to the running maximum, the scale
and zero_point are set to 1.0 and 0.
"""
min_val: torch.Tensor
max_val: torch.Tensor
def __init__(
self,
dtype=torch.quint8,
qscheme=torch.per_tensor_affine,
reduce_range=False,
quant_min=None,
quant_max=None,
factory_kwargs=None,
memoryless=False,
) -> None:
# For x86 quantized kernels, we need to ensure that the vpmaddubsw
# instruction does not overflow. We allow for a reduce_range argument to
# observers that reduces the quantized range to (0,127) or (-64, 63).
# For more details see aten/src/ATen/native/quantized/cpu/qconv.cpp
# This is not an optimal choice for non x86 backends as it loses a bit
# of precision for activations.
super(MinMaxObserver, self).__init__(
dtype=dtype,
qscheme=qscheme,
reduce_range=reduce_range,
quant_min=quant_min,
quant_max=quant_max,
factory_kwargs=factory_kwargs,
)
self.memoryless = memoryless
factory_kwargs = torch.nn.factory_kwargs(factory_kwargs)
self.register_buffer("min_val", torch.tensor(float("inf"), **factory_kwargs))
self.register_buffer("max_val", torch.tensor(float("-inf"), **factory_kwargs))
if (
self.qscheme == torch.per_tensor_symmetric
and self.reduce_range
and self.dtype == torch.quint8
):
raise NotImplementedError(
"Cannot reduce range for symmetric \
quantization for quint8"
)
def forward(self, x_orig):
r"""Records the running minimum and maximum of ``x``."""
if x_orig.numel() == 0:
return x_orig
elif self.memoryless:
self.reset_min_max_vals()
x = x_orig.detach() # avoid keeping autograd tape
x = x.to(self.min_val.dtype)
min_val_cur, max_val_cur = torch.aminmax(x)
min_val = torch.min(min_val_cur, self.min_val)
max_val = torch.max(max_val_cur, self.max_val)
self.min_val.copy_(min_val)
self.max_val.copy_(max_val)
return x_orig
@torch.jit.export
def calculate_qparams(self):
r"""Calculates the quantization parameters."""
return self._calculate_qparams(self.min_val, self.max_val)
@torch.jit.export
def extra_repr(self):
return "min_val={}, max_val={}".format(self.min_val, self.max_val)
@torch.jit.export
def reset_min_max_vals(self):
"""Resets the min/max values."""
self.min_val.copy_(torch.tensor(float("inf")))
self.max_val.copy_(torch.tensor(float("-inf")))
[文档]class MovingAverageMinMaxObserver(MinMaxObserver):
r"""Observer module for computing the quantization parameters based on the
moving average of the min and max values.
This observer computes the quantization parameters based on the moving
averages of minimums and maximums of the incoming tensors. The module
records the average minimum and maximum of incoming tensors, and uses this
statistic to compute the quantization parameters.
Args:
averaging_constant: Averaging constant for min/max.
dtype: Quantized data type
qscheme: Quantization scheme to be used
reduce_range: Reduces the range of the quantized data type by 1 bit
quant_min: Minimum quantization value. If unspecified, it will follow the 8-bit setup.
quant_max: Maximum quantization value. If unspecified, it will follow the 8-bit setup.
The moving average min/max is computed as follows
.. math::
\begin{array}{ll}
x_\text{min} = \begin{cases}
\min(X) & \text{if~}x_\text{min} = \text{None} \\
(1 - c) x_\text{min} + c \min(X) & \text{otherwise}
\end{cases}\\
x_\text{max} = \begin{cases}
\max(X) & \text{if~}x_\text{max} = \text{None} \\
(1 - c) x_\text{max} + c \max(X) & \text{otherwise}
\end{cases}\\
\end{array}
where :math:`x_\text{min/max}` is the running average min/max, :math:`X` is
is the incoming tensor, and :math:`c` is the ``averaging_constant``.
The scale and zero point are then computed as in
:class:`~torch.ao.quantization.observer.MinMaxObserver`.
.. note:: Only works with ``torch.per_tensor_affine`` quantization scheme.
.. note:: If the running minimum equals to the running maximum, the scale
and zero_point are set to 1.0 and 0.
"""
def __init__(
self,
averaging_constant=0.01,
dtype=torch.quint8,
qscheme=torch.per_tensor_affine,
reduce_range=False,
quant_min=None,
quant_max=None,
**kwargs
) -> None:
self.averaging_constant = averaging_constant
super(MovingAverageMinMaxObserver, self).__init__(
dtype=dtype,
qscheme=qscheme,
reduce_range=reduce_range,
quant_min=quant_min,
quant_max=quant_max,
**kwargs
)
def forward(self, x_orig):
if x_orig.numel() == 0:
return x_orig
x = x_orig.detach() # avoid keeping autograd tape
x = x.to(self.min_val.dtype)
min_val = self.min_val
max_val = self.max_val
if min_val == float("inf") and max_val == float("-inf"):
min_val, max_val = torch.aminmax(x)
else:
min_val_cur, max_val_cur = torch.aminmax(x)
min_val = min_val + self.averaging_constant * (min_val_cur - min_val)
max_val = max_val + self.averaging_constant * (max_val_cur - max_val)
self.min_val.copy_(min_val)
self.max_val.copy_(max_val)
return x_orig
[文档]class PerChannelMinMaxObserver(_ObserverBase):
r"""Observer module for computing the quantization parameters based on the
running per channel min and max values.
This observer uses the tensor min/max statistics to compute the per channel
quantization parameters. The module records the running minimum and maximum
of incoming tensors, and uses this statistic to compute the quantization
parameters.
Args:
ch_axis: Channel axis
dtype: Quantized data type
qscheme: Quantization scheme to be used
reduce_range: Reduces the range of the quantized data type by 1 bit
quant_min: Minimum quantization value. If unspecified, it will follow the 8-bit setup.
quant_max: Maximum quantization value. If unspecified, it will follow the 8-bit setup.
memoryless: Boolean that controls whether observer removes old data when a new input is seen.
This is most useful for simulating dynamic quantization, especially during QAT.
The quantization parameters are computed the same way as in
:class:`~torch.ao.quantization.observer.MinMaxObserver`, with the difference
that the running min/max values are stored per channel.
Scales and zero points are thus computed per channel as well.
.. note:: If the running minimum equals to the running maximum, the scales
and zero_points are set to 1.0 and 0.
"""
min_val: torch.Tensor
max_val: torch.Tensor
def __init__(
self,
ch_axis=0,
dtype=torch.quint8,
qscheme=torch.per_channel_affine,
reduce_range=False,
quant_min=None,
quant_max=None,
factory_kwargs=None,
memoryless=False,
) -> None:
super(PerChannelMinMaxObserver, self).__init__(
dtype=dtype,
qscheme=qscheme,
reduce_range=reduce_range,
quant_min=quant_min,
quant_max=quant_max,
factory_kwargs=factory_kwargs,
)
self.memoryless = memoryless
factory_kwargs = torch.nn.factory_kwargs(factory_kwargs)
self.ch_axis = ch_axis
self.register_buffer("min_val", torch.tensor([], **factory_kwargs))
self.register_buffer("max_val", torch.tensor([], **factory_kwargs))
if (
self.qscheme == torch.per_channel_symmetric
and self.reduce_range
and self.dtype == torch.quint8
):
raise NotImplementedError(
"Cannot reduce range for symmetric quantization for quint8"
)
def forward(self, x_orig):
return self._forward(x_orig)
def _forward(self, x_orig):
if x_orig.numel() == 0:
return x_orig
x = x_orig.detach() # avoid keeping autograd tape
min_val = self.min_val
max_val = self.max_val
x_dim = x.size()
new_axis_list = [i for i in range(len(x_dim))] # noqa: C416
new_axis_list[self.ch_axis] = 0
new_axis_list[0] = self.ch_axis
y = x.permute(new_axis_list)
# Need to match dtype of min/max because the updates to buffers
# are done in place and types need to match for comparisons
y = y.to(self.min_val.dtype)
y = torch.flatten(y, start_dim=1)
if min_val.numel() == 0 or max_val.numel() == 0 or self.memoryless:
min_val, max_val = torch.aminmax(y, dim=1)
else:
min_val_cur, max_val_cur = torch.aminmax(y, dim=1)
min_val = torch.min(min_val_cur, min_val)
max_val = torch.max(max_val_cur, max_val)
self.min_val.resize_(min_val.shape)
self.max_val.resize_(max_val.shape)
self.min_val.copy_(min_val)
self.max_val.copy_(max_val)
return x_orig
@torch.jit.export
def calculate_qparams(self):
return self._calculate_qparams(self.min_val, self.max_val)
def extra_repr(self):
return "min_val={}, max_val={}".format(self.min_val, self.max_val)
def _load_from_state_dict(
self,
state_dict: Union[Dict[str, torch.Tensor], Dict[str, torch.Tensor]],
prefix: str,
local_metadata: Dict[str, torch.Tensor],
strict: bool,
missing_keys: List[str],
unexpected_keys: List[str],
error_msgs: List[str],
):
version = local_metadata.get("version", None)
if version is None or version < 3:
local_state = ["min_vals", "max_vals"]
expected_min_name = "min_vals"
expected_max_name = "max_vals"
else:
local_state = ["min_val", "max_val"]
expected_min_name = "min_val"
expected_max_name = "max_val"
for name in local_state:
key = prefix + name
if key in state_dict:
val = state_dict[key]
# Custom handling to allow loading min_val or max_val
# of size N into uninitialized buffers of size 0. The
# buffers are resized here, and the values are copied in
# the default state_dict loading code of the parent.
if name == expected_min_name:
self.min_val.resize_(val.shape)
elif name == expected_max_name:
self.max_val.resize_(val.shape)
else:
warnings.warn("Observer load_from_state_dict got unexpected name {}".format(name))
# For torchscript module we need to update the attributes here since we do not
# call the `_load_from_state_dict` function defined module.py
if torch.jit.is_scripting():
if name == expected_min_name:
self.min_val.copy_(val)
elif name == expected_max_name:
self.max_val.copy_(val)
else:
warnings.warn("Observer load_from_state_dict got unexpected name {}".format(name))
elif strict:
missing_keys.append(key)
if not torch.jit.is_scripting():
super(PerChannelMinMaxObserver, self)._load_from_state_dict(
state_dict,
prefix,
local_metadata,
False,
missing_keys,
unexpected_keys,
error_msgs,
)
def _load_from_state_dict_script(
self,
state_dict: Union[Dict[str, torch.Tensor], Dict[str, torch.Tensor]],
prefix: str,
local_metadata: Dict[str, torch.Tensor],
strict: bool,
missing_keys: List[str],
unexpected_keys: List[str],
error_msgs: List[str],
):
self._load_from_state_dict(
state_dict,
prefix,
local_metadata,
strict,
missing_keys,
unexpected_keys,
error_msgs,
)
@torch.jit.export
def reset_min_max_vals(self):
"""Resets the min/max values."""
self.min_val = torch.tensor([])
self.max_val = torch.tensor([])
[文档]class MovingAveragePerChannelMinMaxObserver(PerChannelMinMaxObserver):
r"""Observer module for computing the quantization parameters based on the
running per channel min and max values.
This observer uses the tensor min/max statistics to compute the per channel
quantization parameters. The module records the running minimum and maximum
of incoming tensors, and uses this statistic to compute the quantization
parameters.
Args:
averaging_constant: Averaging constant for min/max.
ch_axis: Channel axis
dtype: Quantized data type
qscheme: Quantization scheme to be used
reduce_range: Reduces the range of the quantized data type by 1 bit
quant_min: Minimum quantization value. If unspecified, it will follow the 8-bit setup.
quant_max: Maximum quantization value. If unspecified, it will follow the 8-bit setup.
The quantization parameters are computed the same way as in
:class:`~torch.ao.quantization.observer.MovingAverageMinMaxObserver`, with the
difference that the running min/max values are stored per channel.
Scales and zero points are thus computed per channel as well.
.. note:: If the running minimum equals to the running maximum, the scales
and zero_points are set to 1.0 and 0.
"""
def __init__(
self,
averaging_constant=0.01,
ch_axis=0,
dtype=torch.quint8,
qscheme=torch.per_channel_affine,
reduce_range=False,
quant_min=None,
quant_max=None,
**kwargs
) -> None:
super(MovingAveragePerChannelMinMaxObserver, self).__init__(
ch_axis=ch_axis,
dtype=dtype,
qscheme=qscheme,
reduce_range=reduce_range,
quant_min=quant_min,
quant_max=quant_max,
**kwargs
)
self.averaging_constant = averaging_constant
def forward(self, x_orig):
if x_orig.numel() == 0:
return x_orig
x = x_orig.detach() # avoid keeping autograd tape
x = x.to(self.min_val.dtype)
min_val = self.min_val
max_val = self.max_val
x_dim = x.size()
new_axis_list = [i for i in range(len(x_dim))] # noqa: C416
new_axis_list[self.ch_axis] = 0
new_axis_list[0] = self.ch_axis
y = x.permute(new_axis_list)
y = torch.flatten(y, start_dim=1)
if min_val.numel() == 0 or max_val.numel() == 0:
min_val, max_val = torch.aminmax(y, dim=1)
else:
min_val_cur, max_val_cur = torch.aminmax(y, dim=1)
min_val = min_val + self.averaging_constant * (min_val_cur - min_val)
max_val = max_val + self.averaging_constant * (max_val_cur - max_val)
self.min_val.resize_(min_val.shape)
self.max_val.resize_(max_val.shape)
self.min_val.copy_(min_val)
self.max_val.copy_(max_val)
return x_orig
[文档]class HistogramObserver(_ObserverBase):
r"""
The module records the running histogram of tensor values along with
min/max values. ``calculate_qparams`` will calculate scale and zero_point.
Args:
bins: Number of bins to use for the histogram
upsample_rate: Factor by which the histograms are upsampled, this is
used to interpolate histograms with varying ranges across observations
dtype: Quantized data type
qscheme: Quantization scheme to be used
reduce_range: Reduces the range of the quantized data type by 1 bit
The scale and zero point are computed as follows:
1. Create the histogram of the incoming inputs.
The histogram is computed continuously, and the ranges per bin change
with every new tensor observed.
2. Search the distribution in the histogram for optimal min/max values.
The search for the min/max values ensures the minimization of the
quantization error with respect to the floating point model.
3. Compute the scale and zero point the same way as in the
:class:`~torch.ao.quantization.MinMaxObserver`
"""
histogram: torch.Tensor
min_val: torch.Tensor
max_val: torch.Tensor
def __init__(
self,
bins: int = 2048,
upsample_rate: int = 128,
dtype: torch.dtype = torch.quint8,
qscheme=torch.per_tensor_affine,
reduce_range=False,
quant_min=None,
quant_max=None,
factory_kwargs=None,
) -> None:
# bins: The number of bins used for histogram calculation.
super(HistogramObserver, self).__init__(
dtype=dtype,
qscheme=qscheme,
reduce_range=reduce_range,
quant_min=quant_min,
quant_max=quant_max,
factory_kwargs=factory_kwargs,
)
factory_kwargs = torch.nn.factory_kwargs(factory_kwargs)
self.bins = bins
self.register_buffer("histogram", torch.zeros(self.bins, **factory_kwargs))
self.register_buffer("min_val", torch.tensor(float("inf"), **factory_kwargs))
self.register_buffer("max_val", torch.tensor(float("-inf"), **factory_kwargs))
self.dst_nbins = 2 ** torch.iinfo(self.dtype).bits
self.upsample_rate = upsample_rate
def _get_norm(
self, delta_begin: torch.Tensor, delta_end: torch.Tensor, density: torch.Tensor
) -> torch.Tensor:
r"""
Compute the norm of the values uniformaly distributed between
delta_begin and delta_end.
Currently only L2 norm is supported.
norm = density * (integral_{begin, end} x^2)
= density * (end^3 - begin^3) / 3
"""
norm = (
delta_end * delta_end * delta_end - delta_begin * delta_begin * delta_begin
) / 3
return density * norm
def _compute_quantization_error(self, next_start_bin: int, next_end_bin: int):
r"""
Compute the quantization error if we use start_bin to end_bin as the
min and max to do the quantization.
"""
bin_width = (self.max_val.item() - self.min_val.item()) / self.bins
dst_bin_width = bin_width * (next_end_bin - next_start_bin + 1) / self.dst_nbins
if dst_bin_width == 0.0:
return 0.0
src_bin = torch.arange(self.bins, device=self.histogram.device)
# distances from the beginning of first dst_bin to the beginning and
# end of src_bin
src_bin_begin = (src_bin - next_start_bin) * bin_width
src_bin_end = src_bin_begin + bin_width
# which dst_bins the beginning and end of src_bin belong to?
dst_bin_of_begin = torch.clamp(
torch.div(src_bin_begin, dst_bin_width, rounding_mode='floor'), 0, self.dst_nbins - 1
)
dst_bin_of_begin_center = (dst_bin_of_begin + 0.5) * dst_bin_width
dst_bin_of_end = torch.clamp(
torch.div(src_bin_end, dst_bin_width, rounding_mode='floor'), 0, self.dst_nbins - 1
)
dst_bin_of_end_center = (dst_bin_of_end + 0.5) * dst_bin_width
density = self.histogram / bin_width
norm = torch.zeros(self.bins, device=self.histogram.device)
delta_begin = src_bin_begin - dst_bin_of_begin_center
delta_end = dst_bin_width / 2
norm += self._get_norm(delta_begin,
torch.ones(self.bins, device=self.histogram.device) * delta_end,
density)
norm += (dst_bin_of_end - dst_bin_of_begin - 1) * self._get_norm(
torch.tensor(-dst_bin_width / 2), torch.tensor(dst_bin_width / 2), density
)
dst_bin_of_end_center = dst_bin_of_end * dst_bin_width + dst_bin_width / 2
delta_begin = -dst_bin_width / 2
delta_end = src_bin_end - dst_bin_of_end_center
norm += self._get_norm(torch.tensor(delta_begin), delta_end, density)
return norm.sum().item()
def _non_linear_param_search(self) -> Tuple[torch.Tensor, torch.Tensor]:
r"""Non-linear parameter search.
An approximation for L2 error minimization for selecting min/max.
By selecting new min/max, we filter out outliers in input distribution.
This follows the implementation of NormMinimization::NonlinearQuantizationParamsSearch in
caffe2/quantization/server/norm_minimization.cc
"""
assert self.histogram.size()[0] == self.bins, "bins mistmatch"
bin_width = (self.max_val - self.min_val) / self.bins
# cumulative sum
total = torch.sum(self.histogram).item()
cSum = torch.cumsum(self.histogram, dim=0)
stepsize = 1e-5 # granularity
alpha = 0.0 # lower bound
beta = 1.0 # upper bound
start_bin = 0
end_bin = self.bins - 1
norm_min = float("inf")
while alpha < beta:
# Find the next step
next_alpha = alpha + stepsize
next_beta = beta - stepsize
# find the left and right bins between the quantile bounds
l = start_bin
r = end_bin
while l < end_bin and cSum[l] < next_alpha * total:
l = l + 1
while r > start_bin and cSum[r] > next_beta * total:
r = r - 1
# decide the next move
next_start_bin = start_bin
next_end_bin = end_bin
if (l - start_bin) > (end_bin - r):
# move the start bin
next_start_bin = l
alpha = next_alpha
else:
# move the end bin
next_end_bin = r
beta = next_beta
if next_start_bin == start_bin and next_end_bin == end_bin:
continue
# calculate the quantization error using next_start_bin and next_end_bin
norm = self._compute_quantization_error(next_start_bin, next_end_bin)
if norm > norm_min:
break
norm_min = norm
start_bin = next_start_bin
end_bin = next_end_bin
new_min = self.min_val + bin_width * start_bin
new_max = self.min_val + bin_width * (end_bin + 1)
return new_min, new_max
def _adjust_min_max(
self, combined_min: torch.Tensor, combined_max: torch.Tensor, upsample_rate: int
) -> Tuple[torch.Tensor, torch.Tensor, int, int]:
# We ensure that:
# (combined_max - combined_min)/(downsample_rate*Nbins) = (max - min)/(upsample_rate*Nbins)
# This allows us to have a common grid of resolution s, where we can align
# the input histogram
# start_idx maps min_val to the histogram bin index.
hist_bin_width = (self.max_val - self.min_val) / (self.bins * upsample_rate)
downsample_rate = int(
torch.ceil(
(combined_max - combined_min) / (self.bins * hist_bin_width)
).item()
)
e = downsample_rate * (self.bins * hist_bin_width) - (
combined_max - combined_min
)
# Relax only the max, not the min, so that for one sided distributions, min stays at zero
combined_max = combined_max + e
combined_min = combined_min
start_idx = int(
torch.round((self.min_val - combined_min) / hist_bin_width).item()
)
return combined_min, combined_max, downsample_rate, start_idx
def _combine_histograms(
self,
orig_hist: torch.Tensor,
new_hist: torch.Tensor,
upsample_rate: int,
downsample_rate: int,
start_idx: int,
Nbins: int,
) -> torch.Tensor:
# First up-sample the histogram with new data by a factor of L
# This creates an approximate probability density thats piecwise constant
upsampled_histogram = new_hist.repeat_interleave(upsample_rate)
# Now insert the upsampled histogram into the output
# histogram, which is initialized with zeros.
# The offset at which the histogram is introduced is determined
# by the start index as the output histogram can cover a wider range
histogram_with_output_range = torch.zeros(
(Nbins * downsample_rate), device=orig_hist.device
)
histogram_with_output_range[
start_idx : Nbins * upsample_rate + start_idx
] = upsampled_histogram
# Compute integral histogram, double precision is needed to ensure
# that there are no overflows
integral_histogram = torch.cumsum(
histogram_with_output_range, 0, dtype=torch.double
)[downsample_rate - 1 :: downsample_rate]
# Finally perform interpolation
shifted_integral_histogram = torch.zeros((Nbins), device=orig_hist.device)
shifted_integral_histogram[1:Nbins] = integral_histogram[0:-1]
interpolated_histogram = (
integral_histogram - shifted_integral_histogram
) / upsample_rate
orig_hist = orig_hist + interpolated_histogram.to(torch.float)
return orig_hist
def forward(self, x_orig: torch.Tensor) -> torch.Tensor:
if x_orig.numel() == 0:
return x_orig
x = x_orig.detach()
min_val = self.min_val
max_val = self.max_val
same_values = min_val.item() == max_val.item()
is_uninitialized = min_val == float("inf") and max_val == float("-inf")
if is_uninitialized or same_values:
min_val, max_val = torch.aminmax(x)
self.min_val.resize_(min_val.shape)
self.min_val.copy_(min_val)
self.max_val.resize_(max_val.shape)
self.max_val.copy_(max_val)
assert (
min_val.numel() == 1 and max_val.numel() == 1
), "histogram min/max values must be scalar."
torch.histc(
x, self.bins, min=int(min_val), max=int(max_val), out=self.histogram
)
else:
new_min, new_max = torch.aminmax(x)
combined_min = torch.min(new_min, min_val)
combined_max = torch.max(new_max, max_val)
# combine the existing histogram and new histogram into 1 histogram
# We do this by first upsampling the histogram to a dense grid
# and then downsampling the histogram efficiently
(
combined_min,
combined_max,
downsample_rate,
start_idx,
) = self._adjust_min_max(combined_min, combined_max, self.upsample_rate)
assert (
combined_min.numel() == 1 and combined_max.numel() == 1
), "histogram min/max values must be scalar."
combined_histogram = torch.histc(
x, self.bins, min=int(combined_min), max=int(combined_max)
)
if combined_min == min_val and combined_max == max_val:
combined_histogram += self.histogram
else:
combined_histogram = self._combine_histograms(
combined_histogram,
self.histogram,
self.upsample_rate,
downsample_rate,
start_idx,
self.bins,
)
self.histogram.detach_().resize_(combined_histogram.shape)
self.histogram.copy_(combined_histogram)
self.min_val.detach_().resize_(combined_min.shape)
self.min_val.copy_(combined_min)
self.max_val.detach_().resize_(combined_max.shape)
self.max_val.copy_(combined_max)
return x_orig
@torch.jit.export
def calculate_qparams(self):
is_uninitialized = self.min_val == float("inf") and self.max_val == float(
"-inf"
)
if is_uninitialized:
warnings.warn(
"must run observer before calling calculate_qparams.\
Returning default scale and zero point "
)
return torch.tensor([1.0], device=self.min_val.device.type), torch.tensor([0], device=self.min_val.device.type)
assert self.bins == len(self.histogram), (
"The number of bins in histogram should be equal to the number of bins "
"supplied while making this observer"
)
new_min, new_max = self._non_linear_param_search()
return self._calculate_qparams(new_min, new_max)
def _save_to_state_dict(self, destination, prefix, keep_vars):
super(HistogramObserver, self)._save_to_state_dict(
destination, prefix, keep_vars
)
destination[prefix + "min_val"] = self.min_val
destination[prefix + "max_val"] = self.max_val
def _load_from_state_dict(
self,
state_dict,
prefix,
local_metadata,
strict,
missing_keys,
unexpected_keys,
error_msgs,
):
version = local_metadata.get("version", None)
if version is None or version < 3:
# if min_val and max_val are not initialized, update their shape
# to account for the differences between v2 and v3
min_val_name, max_val_name = prefix + "min_val", prefix + "max_val"
if min_val_name in state_dict:
if state_dict[min_val_name].shape == torch.Size([0]):
state_dict[min_val_name] = torch.tensor(float("inf"))
if max_val_name in state_dict:
if state_dict[max_val_name].shape == torch.Size([0]):
state_dict[max_val_name] = torch.tensor(float("-inf"))
local_state = ["min_val", "max_val"]
for name in local_state:
key = prefix + name
if key in state_dict:
val = state_dict[key]
setattr(self, name, val)
elif strict:
missing_keys.append(key)
super(HistogramObserver, self)._load_from_state_dict(
state_dict,
prefix,
local_metadata,
strict,
missing_keys,
unexpected_keys,
error_msgs,
)
[文档]class FixedQParamsObserver(ObserverBase):
r"""
Observer that simulates quantize and dequantize with fixed
quantization parameters in training time. Only per tensor
quantization is supported.
Args:
`scale` (float): fixed scale for the observer
`zero_point` (int): fixed zero point for the observer
`dtype`, `qscheme`, `quant_min`, `quant_max`
"""
scale: torch.Tensor
zero_point: torch.Tensor
def __init__(self,
scale,
zero_point,
dtype=torch.quint8,
qscheme=torch.per_tensor_affine,
quant_min=0,
quant_max=255):
super(FixedQParamsObserver, self).__init__(dtype=dtype)
self.quant_min = quant_min
self.quant_max = quant_max
self.register_buffer('scale', torch.tensor([scale], dtype=torch.float))
self.register_buffer('zero_point', torch.tensor([zero_point], dtype=torch.int))
self.dtype = dtype
self.qscheme = qscheme
def forward(self, X):
return X
@torch.jit.export
def calculate_qparams(self):
return self.scale, self.zero_point
[文档]class PlaceholderObserver(ObserverBase):
r"""
Observer that doesn't do anything and just passes its configuration to the
quantized module's ``.from_float()``.
Can be used for quantization to float16 which doesn't require determining
ranges.
Args:
dtype: Quantized data type
custom_op_name: (temporary) specify this observer for an operator that doesn't require any observation
(Can be used in Graph Mode Passes for special case ops).
"""
def __init__(
self, dtype=torch.float32, custom_op_name="", compute_dtype=None
) -> None:
super(PlaceholderObserver, self).__init__(dtype=dtype)
# dtype of input of the target operator, e.g. for dynamic quantization
# ops, the dtype will be float32
self.dtype = dtype
self.custom_op = custom_op_name
# used for configuration of computation type for dynamic quantization
if compute_dtype:
self.compute_dtype = compute_dtype
def forward(self, x):
return x
@torch.jit.export
def calculate_qparams(self):
raise Exception(
"calculate_qparams should not be called for PlaceholderObserver"
)
[文档]class RecordingObserver(_ObserverBase):
r"""
The module is mainly for debug and records the tensor values during runtime.
Args:
dtype: Quantized data type
qscheme: Quantization scheme to be used
reduce_range: Reduces the range of the quantized data type by 1 bit
"""
__annotations__ = {"tensor_val": List[Optional[torch.Tensor]]}
def __init__(self, **kwargs):
super(RecordingObserver, self).__init__(**kwargs)
self.tensor_val = []
def forward(self, x):
self.tensor_val.append(x.clone())
return x
@torch.jit.export
def calculate_qparams(self):
raise Exception("calculate_qparams should not be called for RecordingObserver")
@torch.jit.export
def get_tensor_value(self):
return self.tensor_val
[文档]class NoopObserver(ObserverBase):
r"""
Observer that doesn't do anything and just passes its configuration to the
quantized module's ``.from_float()``.
Primarily used for quantization to float16 which doesn't require determining
ranges.
Args:
dtype: Quantized data type
custom_op_name: (temporary) specify this observer for an operator that doesn't require any observation
(Can be used in Graph Mode Passes for special case ops).
"""
def __init__(self, dtype=torch.float16, custom_op_name="") -> None:
super(NoopObserver, self).__init__(dtype=dtype)
self.dtype = dtype
self.custom_op = custom_op_name
def forward(self, x):
return x
@torch.jit.export
def calculate_qparams(self):
raise Exception("calculate_qparams should not be called for NoopObserver")
def _is_observer_script_module(mod, obs_type_name):
"""Returns true if given mod is an instance of Observer script module."""
if isinstance(mod, torch.jit.RecursiveScriptModule):
# qualified name looks like '__torch__.torch.ao.quantization.observer.___torch_mangle_2.MinMaxObserver'
suffix = mod._c.qualified_name.split(".", 1)[1]
name = re.sub(r"\.___torch_mangle_\d+", "", suffix)
return obs_type_name in name
return False
def _is_activation_post_process(module):
return (
isinstance(module, torch.ao.quantization.ObserverBase)
or isinstance(module, torch.ao.quantization.FakeQuantize)
or _is_observer_script_module(module, "quantization.observer")
)
def _is_per_channel_script_obs_instance(module):
if isinstance(module, torch.jit.RecursiveScriptModule):
return _is_observer_script_module(
module, "quantization.observer.PerChannelMinMaxObserver"
) or _is_observer_script_module(
module, "quantization.observer.MovingAveragePerChannelMinMaxObserver"
)
return False
def get_observer_state_dict(mod):
r"""
Returns the state dict corresponding to the observer stats.
Traverse the model state_dict and extract out the stats.
"""
od = OrderedDict()
if isinstance(mod, torch.jit.RecursiveScriptModule):
for k, v in mod.state_dict().items():
if "observer" in k:
od[k] = v
else:
# path for GraphModule and nn.Module (eager mode)
for k, v in mod.state_dict().items():
if "activation_post_process" in k:
od[k] = v
od._metadata = mod.state_dict()._metadata # type: ignore[attr-defined]
return od
def load_observer_state_dict(mod, obs_dict):
r"""
Given input model and a state_dict containing model observer stats,
load the stats back into the model. The observer state_dict can be saved
using torch.ao.quantization.get_observer_state_dict
"""
missing_keys: List[str] = []
unexpected_keys: List[str] = []
for name, module in mod.named_modules():
prefix = name + "."
if _is_activation_post_process(module):
if _is_per_channel_script_obs_instance(module):
# For per-channel observers we need to call a custom load_from_state_dict to resize the tensor.
# However this is not called when the module is scripted and we end up calling the default one in module.py
module._load_from_state_dict_script(
obs_dict, prefix, {}, True, missing_keys, unexpected_keys, []
)
else:
module._load_from_state_dict(
obs_dict, prefix, {}, False, missing_keys, unexpected_keys, []
)
for k in missing_keys:
if "observer" in k or "activation_post_process" in k:
raise Exception("Missing keys for observer {} in state_dict".format(k))
for k in unexpected_keys:
if "observer" in k or "activation_post_process" in k:
raise Exception("Unexpected keys for observer {} in state_dict".format(k))
# Restrict activations to be in the range (0,127)
default_observer = MinMaxObserver.with_args(quant_min=0, quant_max=127)
"""
Default observer for static quantization, usually used for debugging.
"""
default_placeholder_observer = PlaceholderObserver
"""
Default placeholder observer, usually used for quantization to torch.float16.
"""
default_debug_observer = RecordingObserver
"""
Default debug-only observer.
"""
default_weight_observer = MinMaxObserver.with_args(
dtype=torch.qint8, qscheme=torch.per_tensor_symmetric
)
"""
Default weight observer.
"""
default_histogram_observer = HistogramObserver.with_args(quant_min=0, quant_max=127)
"""
Default histogram observer, usually used for PTQ.
"""
default_per_channel_weight_observer = PerChannelMinMaxObserver.with_args(
dtype=torch.qint8, qscheme=torch.per_channel_symmetric
)
"""
Default per-channel weight observer, usually used on backends where per-channel
weight quantization is supported, such as `fbgemm`.
"""
default_dynamic_quant_observer = PlaceholderObserver.with_args(
dtype=torch.float, compute_dtype=torch.quint8
)
"""
Default observer for dynamic quantization.
"""
default_float_qparams_observer = PerChannelMinMaxObserver.with_args(
dtype=torch.quint8, qscheme=torch.per_channel_affine_float_qparams, ch_axis=0
)
"""
Default observer for a floating point zero-point.
"""
default_float_qparams_observer_4bit = PerChannelMinMaxObserver.with_args(
dtype=torch.quint4x2, qscheme=torch.per_channel_affine_float_qparams, ch_axis=0
)
"""
Default observer for a floating point zero-point and 4 bit activations.
"""
# TODO(future PR): remove these defaults and enforce activation functions
# to explicitly specify their output range
default_symmetric_fixed_qparams_observer = FixedQParamsObserver.with_args(
scale=2.0 / 256.0, zero_point=128, dtype=torch.quint8, quant_min=0, quant_max=255)
default_affine_fixed_qparams_observer = FixedQParamsObserver.with_args(
scale=1.0 / 256.0, zero_point=0, dtype=torch.quint8, quant_min=0, quant_max=255)
"""
Default observers for fixed qparams operations.
"""
default_reuse_input_observer = ReuseInputObserver
"""
Default observer for operators like reshape that reuses the observer of input to
the operator
"""