Files
2026-07-13 12:40:42 +08:00

456 lines
16 KiB
Python

# Copyright (c) 2021 PaddlePaddle Authors. All Rights Reserved.
#
# Licensed under the Apache License, Version 2.0 (the "License");
# you may not use this file except in compliance with the License.
# You may obtain a copy of the License at
#
# http://www.apache.org/licenses/LICENSE-2.0
#
# Unless required by applicable law or agreed to in writing, software
# distributed under the License is distributed on an "AS IS" BASIS,
# WITHOUT WARRANTIES OR CONDITIONS OF ANY KIND, either express or implied.
# See the License for the specific language governing permissions and
# limitations under the License.
import errno
import os
import random
import unittest
import warnings
import numpy as np
import paddle
from paddle import base
from paddle.base import core
from paddle.base.core import AnalysisConfig, create_paddle_predictor
from paddle.base.framework import IrGraph
from paddle.static import Variable
from paddle.static.io import append_fetch_ops, prepend_feed_ops
from paddle.static.quantization import (
AddQuantDequantPass,
OutScaleForInferencePass,
OutScaleForTrainingPass,
QuantizationFreezePass,
QuantizationTransformPass,
)
class QuantDequantTest(unittest.TestCase):
def __init__(self, methodName='runTest'):
super().__init__(methodName)
paddle.enable_static()
self.main_program = paddle.static.Program()
self.startup_program = paddle.static.Program()
self.test_main_program = paddle.static.Program()
self.test_startup_program = paddle.static.Program()
self.feeds = None
self.fetch_list = None
self.enable_onednn = False
self.enable_onednn_bfloat16 = False
self.enable_trt = False
self.enable_tensorrt_varseqlen = True
self.trt_parameters = None
self.dynamic_shape_params = None
self.enable_lite = False
self.lite_parameters = None
self.path = "./inference_pass/" + self.__class__.__name__
self.data = None
self.label = None
self.result = None
np.random.seed(1)
random.seed(1)
# from Paddle release2.1
def _normalize_program(self, program, feed_vars, fetch_vars):
if not isinstance(program, paddle.static.Program):
raise TypeError(
f"program type must be `paddle.static.Program`, but received `{type(program)}`"
)
if not isinstance(feed_vars, list):
feed_vars = [feed_vars]
if not all(isinstance(v, Variable) for v in feed_vars):
raise TypeError(
"feed_vars type must be a Variable or a list of Variable."
)
if not isinstance(fetch_vars, list):
fetch_vars = [fetch_vars]
if not all(isinstance(v, Variable) for v in fetch_vars):
raise TypeError(
"fetch_vars type must be a Variable or a list of Variable."
)
# remind users to set auc_states to 0 if auc op were found.
for op in program.global_block().ops:
# clear device of Op
device_attr_name = (
core.op_proto_and_checker_maker.kOpDeviceAttrName()
)
op._set_attr(device_attr_name, "")
if op.type == 'auc':
warnings.warn(
"Be sure that you have set auc states to 0 "
"before saving inference model."
)
break
# serialize program
copy_program = program.clone()
global_block = copy_program.global_block()
remove_op_idx = []
for i, op in enumerate(global_block.ops):
op.desc.set_is_target(False)
if op.type == "feed" or op.type == "fetch":
remove_op_idx.append(i)
for idx in remove_op_idx[::-1]:
global_block._remove_op(idx)
copy_program.desc.flush()
feed_var_names = [var.name for var in feed_vars]
copy_program = copy_program._prune_with_input(
feeded_var_names=feed_var_names, targets=fetch_vars
)
copy_program = copy_program._inference_optimize(prune_read_op=True)
fetch_var_names = [var.name for var in fetch_vars]
prepend_feed_ops(copy_program, feed_var_names)
append_fetch_ops(copy_program, fetch_var_names)
copy_program.desc._set_version()
return copy_program
def _save_models(
self, dirname, feeded_var_names, target_vars, executor, program, scope
):
# save models as combined but sometimes params is null
# To adapt to this situation, the path needs to be adjusted to the old version format.
feeded_vars = []
for var in program.list_vars():
if var.name in feeded_var_names:
feeded_vars.append(var)
with paddle.static.scope_guard(scope):
paddle.static.io.save_inference_model(
dirname,
feeded_vars,
target_vars,
executor,
program=program,
clip_extra=True,
)
# if the param save is null
# replace model_path to old version
param_file = dirname + ".pdiparams"
if not os.path.exists(param_file):
model_path = dirname + ".pdmodel"
try:
save_dirname = os.path.normpath(dirname)
os.makedirs(save_dirname)
except OSError as e:
if e.errno != errno.EEXIST:
raise
model_path_old = os.path.join(save_dirname, "__model__")
if not os.path.exists(model_path_old):
os.rename(model_path, model_path_old)
def _get_paddle_outs(self, feed, fetch_list, executor, program, scope):
'''
Return PaddlePaddle outputs.
'''
with paddle.static.scope_guard(scope):
outs = executor.run(
program=program,
feed=feed,
fetch_list=fetch_list,
return_numpy=True,
)
return outs
def _get_inference_outs(self, config):
'''
Return AnalysisPredictor outputs.
'''
predictor = create_paddle_predictor(config)
tensor_shapes = predictor.get_input_tensor_shape()
names = predictor.get_input_names()
for i, name in enumerate(names):
shape = tensor_shapes[name]
shape[0] = 1
tensor = predictor.get_input_tensor(name)
feed_data = list(self.feeds.values())[i]
tensor.copy_from_cpu(np.array(feed_data))
if type(feed_data) == base.DenseTensor:
tensor.set_lod(feed_data.lod())
predictor.zero_copy_run()
output_names = predictor.get_output_names()
outs = [
predictor.get_output_tensor(out_name).copy_to_cpu()
for out_name in output_names
]
return outs
def _get_analysis_config(
self, use_gpu=False, use_trt=False, use_onednn=False
):
'''
Return a new object of AnalysisConfig.
'''
# To adapt to save_inference_model
param_file = self.path + ".pdiparams"
if not os.path.exists(param_file):
config = AnalysisConfig(self.path)
else:
config = AnalysisConfig(
self.path + ".pdmodel", self.path + ".pdiparams"
)
config.disable_gpu()
config.disable_onednn()
config.switch_specify_input_names(True)
config.switch_ir_optim(True)
config.switch_use_feed_fetch_ops(False)
if use_gpu:
config.enable_use_gpu(100, 0)
if use_trt:
config.enable_tensorrt_engine(
self.trt_parameters.workspace_size,
self.trt_parameters.max_batch_size,
self.trt_parameters.min_subgraph_size,
self.trt_parameters.precision,
self.trt_parameters.use_static,
self.trt_parameters.use_calib_mode,
)
if self.dynamic_shape_params:
config.set_trt_dynamic_shape_info(
self.dynamic_shape_params.min_input_shape,
self.dynamic_shape_params.max_input_shape,
self.dynamic_shape_params.optim_input_shape,
self.dynamic_shape_params.disable_trt_plugin_fp16,
)
if self.enable_tensorrt_varseqlen:
config.enable_tensorrt_varseqlen()
elif use_onednn:
config.enable_onednn()
if self.enable_onednn_bfloat16:
config.enable_onednn_bfloat16()
print('config summary:', config.summary())
return config
def check_output_with_option(
self, use_gpu, atol=1e-5, flatten=False, quant=False, rtol=1e-5
):
'''
Check whether calculating on CPU and GPU, enable TensorRT
or disable TensorRT, enable ONEDNN or disable ONEDNN
are all the same.
'''
place = paddle.CUDAPlace(0) if use_gpu else paddle.CPUPlace()
executor = paddle.static.Executor(place)
scope = paddle.static.Scope()
device = "GPU" if use_gpu else "CPU"
with paddle.static.scope_guard(scope):
executor.run(self.startup_program)
executor.run(self.test_startup_program)
main_graph = IrGraph(core.Graph(self.main_program.desc), for_test=False)
test_graph = IrGraph(
core.Graph(self.test_main_program.desc), for_test=True
)
transform_pass = QuantizationTransformPass(
scope=scope,
place=place,
activation_quantize_type=self.activation_quantize_type,
weight_quantize_type=self.weight_quantize_type,
)
transform_pass.apply(main_graph)
transform_pass.apply(test_graph)
add_quant_dequant_pass = AddQuantDequantPass(scope=scope, place=place)
add_quant_dequant_pass.apply(main_graph)
add_quant_dequant_pass.apply(test_graph)
scale_training_pass = OutScaleForTrainingPass(scope=scope, place=place)
scale_training_pass.apply(main_graph)
build_strategy = paddle.static.BuildStrategy()
build_strategy.memory_optimize = False
build_strategy.enable_inplace = False
build_strategy.fuse_all_reduce_ops = False
binary = paddle.static.CompiledProgram(main_graph.graph)
iters = 10
batch_size = 1
train_reader = paddle.batch(
paddle.reader.shuffle(paddle.dataset.mnist.train(), buf_size=500),
batch_size=batch_size,
)
feeder = base.DataFeeder(feed_list=[self.data, self.label], place=place)
with paddle.static.scope_guard(scope):
for _ in range(iters):
data = next(train_reader())
loss_v = executor.run(
binary, feed=feeder.feed(data), fetch_list=[self.loss]
)
scale_inference_pass = OutScaleForInferencePass(scope=scope)
scale_inference_pass.apply(test_graph)
# Freeze graph for inference, but the weight of fc/conv is still float type.
freeze_pass = QuantizationFreezePass(
scope=scope,
place=place,
weight_quantize_type=self.weight_quantize_type,
)
freeze_pass.apply(test_graph)
self.main_program = test_graph.to_program()
with paddle.static.scope_guard(scope):
self.main_program = self._normalize_program(
self.main_program, self.data, self.fetch_list
)
self._save_models(
self.path,
list(self.feeds.keys()),
self.fetch_list,
executor,
self.main_program,
scope,
)
paddle_outs = self._get_paddle_outs(
self.feeds, self.fetch_list, executor, self.main_program, scope
)
inference_outs = self._get_inference_outs(
self._get_analysis_config(use_gpu=use_gpu)
)
# Check whether the results calculated on CPU and on GPU are the same.
self.assertTrue(
len(paddle_outs) == len(inference_outs),
f"The number of outputs is different between inference and training forward at {device}",
)
for out, inference_out in zip(paddle_outs, inference_outs):
paddle_out = np.array(out)
if flatten:
paddle_out = paddle_out.flatten()
inference_out = inference_out.flatten()
np.testing.assert_allclose(
paddle_out,
inference_out,
rtol=1e-05,
atol=atol,
err_msg=f'Output has diff between inference and training forward at {device} ',
)
# Check whether the trt results and the GPU results are the same.
if use_gpu and self.enable_trt:
tensorrt_outputs = self._get_inference_outs(
self._get_analysis_config(
use_gpu=use_gpu, use_trt=self.enable_trt
)
)
if self.trt_parameters.use_static:
# deserialize
tensorrt_outputs = self._get_inference_outs(
self._get_analysis_config(
use_gpu=use_gpu, use_trt=self.enable_trt
)
)
self.assertTrue(
len(tensorrt_outputs) == len(paddle_outs),
"The number of outputs is different between GPU and TensorRT. ",
)
for paddle_out, tensorrt_output in zip(
paddle_outs, tensorrt_outputs
):
paddle_out = np.array(paddle_out)
if flatten:
paddle_out = paddle_out.flatten()
tensorrt_output = tensorrt_output.flatten()
np.testing.assert_allclose(
paddle_out,
tensorrt_output,
rtol=rtol,
atol=atol,
err_msg='Output has diff between GPU and TensorRT. ',
)
# Check whether the onednn results and the CPU results are the same.
if (not use_gpu) and self.enable_onednn:
onednn_outputs = self._get_inference_outs(
self._get_analysis_config(
use_gpu=use_gpu, use_onednn=self.enable_onednn
)
)
self.assertTrue(
len(paddle_outs) == len(onednn_outputs),
"The number of outputs is different between CPU and ONEDNN. ",
)
if self.enable_onednn_bfloat16:
atol = 0.01
for paddle_out, onednn_output in zip(paddle_outs, onednn_outputs):
np.testing.assert_allclose(
np.array(paddle_out),
onednn_output,
rtol=1e-05,
atol=atol,
err_msg='Output has diff between CPU and ONEDNN. ',
)
class TensorRTParam:
'''
Prepare TensorRT subgraph engine parameters.
'''
def __init__(
self,
workspace_size,
max_batch_size,
min_subgraph_size,
precision,
use_static,
use_calib_mode,
):
self.workspace_size = workspace_size
self.max_batch_size = max_batch_size
self.min_subgraph_size = min_subgraph_size
self.precision = precision
self.use_static = use_static
self.use_calib_mode = use_calib_mode
class DynamicShapeParam:
'''
Prepare TensorRT subgraph engine dynamic shape parameters.
'''
def __init__(
self,
min_input_shape,
max_input_shape,
optim_input_shape,
disable_trt_plugin_fp16,
):
self.min_input_shape = min_input_shape
self.max_input_shape = max_input_shape
self.optim_input_shape = optim_input_shape
self.disable_trt_plugin_fp16 = disable_trt_plugin_fp16
def quant_dequant(self):
place = paddle.CPUPlace()
exe = paddle.static.Executor(place)
scope = paddle.static.Scope()