# Copyright (c) 2023 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 math import unittest from test_trt_explicit_quantization_model import TestExplicitQuantizationModel import paddle class ResNet: def __init__(self, layers=50, prefix_name=''): self.layers = layers self.prefix_name = prefix_name def net(self, input, class_dim=1000, conv1_name='conv1', fc_name=None): layers = self.layers prefix_name = ( self.prefix_name if self.prefix_name == '' else self.prefix_name + '_' ) supported_layers = [34, 50, 101, 152] assert layers in supported_layers, ( f"supported layers are {supported_layers} but input layer is {layers}" ) if layers == 34 or layers == 50: depth = [3, 4, 6, 3] elif layers == 101: depth = [3, 4, 23, 3] elif layers == 152: depth = [3, 8, 36, 3] num_filters = [64, 128, 256, 512] conv = self.conv_bn_layer( input=input, num_filters=64, filter_size=7, stride=2, act='relu', name=prefix_name + conv1_name, ) if layers >= 50: for block in range(len(depth)): for i in range(depth[block]): if layers in [101, 152] and block == 2: if i == 0: conv_name = "res" + str(block + 2) + "a" else: conv_name = "res" + str(block + 2) + "b" + str(i) else: conv_name = "res" + str(block + 2) + chr(97 + i) conv_name = prefix_name + conv_name conv = self.bottleneck_block( input=conv, num_filters=num_filters[block], stride=2 if i == 0 and block != 0 else 1, name=conv_name, ) pool = paddle.nn.functional.adaptive_avg_pool2d(conv, 1) pool = paddle.reshape(pool, [-1, 2048]) stdv = 1.0 / math.sqrt(pool.shape[1] * 1.0) fc_name = fc_name if fc_name is None else prefix_name + fc_name out = paddle.static.nn.fc( pool, class_dim, activation='softmax', name=fc_name, weight_attr=paddle.ParamAttr( initializer=paddle.nn.initializer.Uniform(-stdv, stdv) ), ) else: for block in range(len(depth)): for i in range(depth[block]): conv_name = "res" + str(block + 2) + chr(97 + i) conv_name = prefix_name + conv_name conv = self.basic_block( input=conv, num_filters=num_filters[block], stride=2 if i == 0 and block != 0 else 1, is_first=block == i == 0, name=conv_name, ) pool = paddle.nn.functional.adaptive_avg_pool2d(conv, 1) stdv = 1.0 / math.sqrt(pool.shape[1] * 1.0) fc_name = fc_name if fc_name is None else prefix_name + fc_name out = paddle.static.nn.fc( pool, class_dim, name=fc_name, weight_attr=paddle.ParamAttr( initializer=paddle.nn.initializer.Uniform(-stdv, stdv) ), ) return out def conv_bn_layer( self, input, num_filters, filter_size, stride=1, groups=1, act=None, name=None, ): conv = paddle.static.nn.conv2d( input=input, num_filters=num_filters, filter_size=filter_size, stride=stride, padding=(filter_size - 1) // 2, groups=groups, act=None, param_attr=paddle.ParamAttr(name=name + "_weights"), bias_attr=False, name=name + '.conv2d.output.1', ) if self.prefix_name == '': if name == "conv1": bn_name = "bn_" + name else: bn_name = "bn" + name[3:] else: if name.split("_")[1] == "conv1": bn_name = name.split("_", 1)[0] + "_bn_" + name.split("_", 1)[1] else: bn_name = ( name.split("_", 1)[0] + "_bn" + name.split("_", 1)[1][3:] ) return paddle.static.nn.batch_norm( input=conv, act=act, name=bn_name + '.output.1', param_attr=paddle.ParamAttr(name=bn_name + '_scale'), bias_attr=paddle.ParamAttr(bn_name + '_offset'), moving_mean_name=bn_name + '_mean', moving_variance_name=bn_name + '_variance', ) def shortcut(self, input, ch_out, stride, is_first, name): ch_in = input.shape[1] if ch_in != ch_out or stride != 1 or is_first: return self.conv_bn_layer(input, ch_out, 1, stride, name=name) else: return input def bottleneck_block(self, input, num_filters, stride, name): conv0 = self.conv_bn_layer( input=input, num_filters=num_filters, filter_size=1, act='relu', name=name + "_branch2a", ) conv1 = self.conv_bn_layer( input=conv0, num_filters=num_filters, filter_size=3, stride=stride, act='relu', name=name + "_branch2b", ) conv2 = self.conv_bn_layer( input=conv1, num_filters=num_filters * 4, filter_size=1, act=None, name=name + "_branch2c", ) short = self.shortcut( input, num_filters * 4, stride, is_first=False, name=name + "_branch1", ) out = paddle.add(x=short, y=conv2, name=name + ".add.output.5") return paddle.nn.functional.relu(out) def basic_block(self, input, num_filters, stride, is_first, name): conv0 = self.conv_bn_layer( input=input, num_filters=num_filters, filter_size=3, act='relu', stride=stride, name=name + "_branch2a", ) conv1 = self.conv_bn_layer( input=conv0, num_filters=num_filters, filter_size=3, act=None, name=name + "_branch2b", ) short = self.shortcut( input, num_filters, stride, is_first, name=name + "_branch1" ) out = paddle.add(x=short, y=conv1) return paddle.nn.functional.relu(out) @unittest.skipIf( paddle.inference.get_trt_compile_version() < (8, 5, 1), "Quantization axis is consistent with Paddle after TRT 8.5.2.", ) class TestExplicitQuantizationResNet( TestExplicitQuantizationModel, unittest.TestCase ): def build_model(self): model = ResNet(layers=50, prefix_name='') return model if __name__ == '__main__': unittest.main()