Skip to content

P
PaddleDetection

PaddlePaddle / PaddleDetection
1 年多 前同步成功

通知 696
Star 11112
Fork 2696
P
PaddleDetection
提交 36d1bc7b 编写于 作者: L Liufang Sang 提交者： ceci3
浏览文件
操作

change to new quantization api in quantization demo (#127) 

* add slim quantization demo

* add quantization demo

* add skip quant

* fix details in doc

* fix details

* fix details

* fix details

* fix details

* fix details

* fix details

* add model zoo link in doc

* remove result in doc
上级 e5fe96e7
隐藏空白更改
内联 并排
Showing with 888 addition and 462 deletion
ppdet/modeling/anchor_heads/yolo_head.py
浏览文件 @ 36d1bc7b
......@@ -221,20 +221,22 @@ class YOLOv3Head(object):
# out channel number = mask_num * (5 + class_num)
num_filters = len(self.anchor_masks[i]) * (self.num_classes + 5)
block_out = fluid.layers.conv2d(
input=tip,
num_filters=num_filters,
filter_size=1,
stride=1,
padding=0,
act=None,
param_attr=ParamAttr(name=self.prefix_name +
"yolo_output.{}.conv.weights".format(i)),
bias_attr=ParamAttr(
regularizer=L2Decay(0.),
name=self.prefix_name +
"yolo_output.{}.conv.bias".format(i)))
outputs.append(block_out)
with fluid.name_scope('yolo_output'):
block_out = fluid.layers.conv2d(
input=tip,
num_filters=num_filters,
filter_size=1,
stride=1,
padding=0,
act=None,
param_attr=ParamAttr(
name=self.prefix_name +
"yolo_output.{}.conv.weights".format(i)),
bias_attr=ParamAttr(
regularizer=L2Decay(0.),
name=self.prefix_name +
"yolo_output.{}.conv.bias".format(i)))
outputs.append(block_out)
if i < len(blocks) - 1:
# do not perform upsample in the last detection_block
......
slim/quantization/README.md
浏览文件 @ 36d1bc7b
>运行该示例前请安装Paddle1.6或更高版本
>运行该示例前请安装Paddle1.6或更高版本和PaddleSlim
# 检测模型量化压缩示例
## 概述
该示例使用PaddleSlim提供的[量化压缩策略](https://github.com/PaddlePaddle/models/blob/develop/PaddleSlim/docs/tutorial.md#1-quantization-aware-training%E9%87%8F%E5%8C%96%E4%BB%8B%E7%BB%8D)对分类模型进行压缩。
该示例使用PaddleSlim提供的[量化压缩API](https://paddlepaddle.github.io/PaddleSlim/api/quantization_api/)对检测模型进行压缩。
在阅读该示例前，建议您先了解以下内容：
- [检测模型的常规训练方法](https://github.com/PaddlePaddle/PaddleDetection)
- [PaddleSlim使用文档](https://github.com/PaddlePaddle/models/blob/develop/PaddleSlim/docs/usage.md)
- [PaddleSlim使用文档](https://paddlepaddle.github.io/PaddleSlim/)
## 配置文件说明
## 安装PaddleSlim
可按照[PaddleSlim使用文档](https://paddlepaddle.github.io/PaddleSlim/)中的步骤安装PaddleSlim。
关于配置文件如何编写您可以参考：
- [PaddleSlim配置文件编写说明](https://github.com/PaddlePaddle/models/blob/develop/PaddleSlim/docs/usage.md#122-%E9%85%8D%E7%BD%AE%E6%96%87%E4%BB%B6%E7%9A%84%E4%BD%BF%E7%94%A8)
- [量化策略配置文件编写说明](https://github.com/PaddlePaddle/models/blob/develop/PaddleSlim/docs/usage.md#21-%E9%87%8F%E5%8C%96%E8%AE%AD%E7%BB%83)
## 训练
根据 [tools/train.py](../../tools/train.py) 编写压缩脚本train.py。脚本中量化的步骤如下。
### 定义量化配置
config = {
'weight_quantize_type': 'channel_wise_abs_max',
'activation_quantize_type': 'moving_average_abs_max',
'quantize_op_types': ['depthwise_conv2d', 'mul', 'conv2d'],
'not_quant_pattern': ['yolo_output']
}
其中save_out_nodes需要得到检测结果的Variable的名称，下面介绍如何确定save_out_nodes的参数
以MobileNet V1为例，可在compress.py中构建好网络之后，直接打印Variable得到Variable的名称信息。
代码示例：
如何配置以及含义请参考[PaddleSlim 量化API](https://paddlepaddle.github.io/PaddleSlim/api/quantization_api/)。
### 插入量化反量化OP
使用[PaddleSlim quant_aware API](https://paddlepaddle.github.io/PaddleSlim/api/quantization_api/#quant_aware)在Program中插入量化和反量化OP。
```
eval_keys, eval_values, eval_cls = parse_fetches(fetches, eval_prog,
extra_keys)
# print(eval_values)
train_prog = quant_aware(train_prog, place, config, for_test=False)
```
根据运行结果可看到Variable的名字为：`multiclass_nms_0.tmp_0`。
## 训练
根据 [tools/train.py](https://github.com/PaddlePaddle/PaddleDetection/tree/master/tools/train.py) 编写压缩脚本compress.py。
在该脚本中定义了Compressor对象，用于执行压缩任务。
### 关闭一些训练策略
通过`python compress.py --help`查看可配置参数，简述如下：
因为量化要对Program做修改，所以一些会修改Program的训练策略需要关闭。``sync_batch_norm`` 和量化多卡训练同时使用时会出错，原因暂不知，因此也需要将其关闭。
```
build_strategy.fuse_all_reduce_ops = False
build_strategy.sync_batch_norm = False
```
- config: 检测库的配置，其中配置了训练超参数、数据集信息等。
- slim_file: PaddleSlim的配置文件，参见[配置文件说明](#配置文件说明)。
### 开始训练
您可以通过运行以下命令运行该示例。
您可以通过运行以下命令运行该示例。(该示例是在COCO数据集上训练yolov3-mobilenetv1, 替换模型和数据集的方法和检测库类似，直接替换相应的配置文件即可)
step1: 设置gpu卡
```
export CUDA_VISIBLE_DEVICES=0
```
step2: 开始训练
使用PaddleDetection提供的配置文件在用8卡进行训练：
请在PaddleDetection根目录下运行。
```
python compress.py \
-s yolov3_mobilenet_v1_slim.yaml \
-c ../../configs/yolov3_mobilenet_v1_voc.yml \
-d "../../dataset/voc" \
-o max_iters=258 \
python slim/quantization/train.py \
--eval \
-c ./configs/yolov3_mobilenet_v1.yml \
-o max_iters=30000 \
save_dir=./output/mobilenetv1 \
LearningRate.base_lr=0.0001 \
LearningRate.schedulers="[!PiecewiseDecay {gamma: 0.1, milestones: [258, 516]}]" \
pretrain_weights=https://paddlemodels.bj.bcebos.com/object_detection/yolov3_mobilenet_v1_voc.tar \
YoloTrainFeed.batch_size=64
LearningRate.schedulers='[!PiecewiseDecay {gamma: 0.1, milestones: [10000]}]' \
pretrain_weights=https://paddlemodels.bj.bcebos.com/object_detection/yolov3_mobilenet_v1.tar
```
>通过命令行覆盖设置max_iters选项，因为PaddleDetection中训练是以`batch`为单位迭代的，并没有涉及`epoch`的概念，但是PaddleSlim需要知道当前训练进行到第几个`epoch`, 所以需要将`max_iters`设置为一个`epoch`内的`batch`的数量。
如果要调整训练卡数，需要调整配置文件`yolov3_mobilenet_v1_voc.yml`中的以下参数：
>通过命令行覆设置max_iters选项，因为量化的训练轮次比正常训练小很多，所以需要修改此选项。
如果要调整训练卡数，可根据需要调整配置文件`yolov3_mobilenet_v1_voc.yml`中的以下参数：
- **max_iters:** 一个`epoch`中batch的数量，需要设置为`total_num / batch_size`, 其中`total_num`为训练样本总数量，`batch_size`为多卡上总的batch size.
- **YoloTrainFeed.batch_size:** 当使用DataLoader时，表示单张卡上的batch size; 当使用普通reader时，则表示多卡上的总的batch_size。batch_size受限于显存大小。
- **max_iters:** 训练的总轮次。
- **LeaningRate.base_lr:** 根据多卡的总`batch_size`调整`base_lr`，两者大小正相关，可以简单的按比例进行调整。
- **LearningRate.schedulers.PiecewiseDecay.milestones：** 请根据batch size的变化对其调整。
- **LearningRate.schedulers.PiecewiseDecay.LinearWarmup.steps：** 请根据batch size的变化对其进行调整。
以下为4卡训练示例，通过命令行覆盖`yolov3_mobilenet_v1_voc.yml`中的参数：
```
python compress.py \
-s yolov3_mobilenet_v1_slim.yaml \
-c ../../configs/yolov3_mobilenet_v1_voc.yml \
-d "../../dataset/voc" \
-o max_iters=258 \
LearningRate.base_lr=0.0001 \
LearningRate.schedulers="[!PiecewiseDecay {gamma: 0.1, milestones: [258, 516]}]" \
pretrain_weights=https://paddlemodels.bj.bcebos.com/object_detection/yolov3_mobilenet_v1_voc.tar \
YoloTrainFeed.batch_size=64
```
以下为2卡训练示例，受显存所制，单卡`batch_size`不变, 总`batch_size`减小，`base_lr`减小，一个epoch内batch数量增加，同时需要调整学习率相关参数，如下：
```
python compress.py \
-s yolov3_mobilenet_v1_slim.yaml \
-c ../../configs/yolov3_mobilenet_v1_voc.yml \
-d "../../dataset/voc" \
-o max_iters=516 \
LearningRate.base_lr=0.00005 \
LearningRate.schedulers="[!PiecewiseDecay {gamma: 0.1, milestones: [516, 1012]}]" \
pretrain_weights=https://paddlemodels.bj.bcebos.com/object_detection/yolov3_mobilenet_v1_voc.tar \
YoloTrainFeed.batch_size=32
```
通过`python compress.py --help`查看可配置参数。
通过`python ../../tools/configure.py ${option_name} help`查看如何通过命令行覆盖配置文件`yolov3_mobilenet_v1_voc.yml`中的参数。
通过`python slim/quantization/train.py --help`查看可配置参数。
通过`python .tools/configure.py ${option_name} help`查看如何通过命令行覆盖配置文件中的参数。
### 训练时的模型结构
这部分介绍来源于[量化low-level API介绍](https://github.com/PaddlePaddle/models/tree/develop/PaddleSlim/quant_low_level_api#1-%E9%87%8F%E5%8C%96%E8%AE%AD%E7%BB%83low-level-apis%E4%BB%8B%E7%BB%8D)。
PaddlePaddle框架中和量化相关的IrPass, 分别有QuantizationTransformPass、QuantizationFreezePass、ConvertToInt8Pass。在训练时，对网络应用了QuantizationTransformPass，作用是在网络中的conv2d、depthwise_conv2d、mul等算子的各个输入前插入连续的量化op和反量化op，并改变相应反向算子的某些输入。示例图如下：
[PaddleSlim 量化API](https://paddlepaddle.github.io/PaddleSlim/api/quantization_api/)文档中介绍了``paddleslim.quant.quant_aware``和``paddleslim.quant.convert``两个接口。
``paddleslim.quant.quant_aware`` 作用是在网络中的conv2d、depthwise_conv2d、mul等算子的各个输入前插入连续的量化op和反量化op，并改变相应反向算子的某些输入。示例图如下：
<p align="center">
<img src="./images/TransformPass.png" height=400 width=520 hspace='10'/> <br />
<strong>图1：应用QuantizationTransformPass后的结果</strong>
<strong>图1：应用 paddleslim.quant.quant_aware 后的结果</strong>
</p>
### 保存断点（checkpoint）
在脚本中使用保存checkpoint的代码为：
```
# insert quantize op in eval_prog
eval_prog = quant_aware(eval_prog, place, config, for_test=True)
checkpoint.save(exe, eval_prog, os.path.join(save_dir, save_name))
```
如果在配置文件中设置了`checkpoint_path`, 则在压缩任务执行过程中会自动保存断点，当任务异常中断时，
重启任务会自动从`checkpoint_path`路径下按数字顺序加载最新的checkpoint文件。如果不想让重启的任务从断点恢复，
需要修改配置文件中的`checkpoint_path`，或者将`checkpoint_path`路径下文件清空。
>注意：配置文件中的信息不会保存在断点中，重启前对配置文件的修改将会生效。
### 边训练边测试
### 保存评估和预测模型
在脚本中边训练边测试得到的测试精度是基于图1中的网络结构进行的。
如果在配置文件的量化策略中设置了`float_model_save_path`, `int8_model_save_path` 在训练结束后，会保存模型量化压缩之后用于预测的模型。接下来介绍这2种预测模型的区别。
## 评估
#### FP32模型
在介绍量化训练时的模型结构时介绍了PaddlePaddle框架中和量化相关的IrPass, 分别是QuantizationTransformPass、QuantizationFreezePass、ConvertToInt8Pass。FP32模型是在应用QuantizationFreezePass并删除eval_program中多余的operators之后，保存的模型。
### 最终评估模型
QuantizationFreezePass主要用于改变IrGraph中量化op和反量化op的顺序，即将类似图1中的量化op和反量化op顺序改变为图2中的布局。除此之外，QuantizationFreezePass还会将`conv2d`、`depthwise_conv2d`、`mul`等算子的权重离线量化为int8_t范围内的值(但数据类型仍为float32)，以减少预测过程中对权重的量化操作，示例如图2：
``paddleslim.quant.convert`` 主要用于改变Program中量化op和反量化op的顺序，即将类似图1中的量化op和反量化op顺序改变为图2中的布局。除此之外，``paddleslim.quant.convert`` 还会将`conv2d`、`depthwise_conv2d`、`mul`等算子参数变为量化后的int8_t范围内的值(但数据类型仍为float32)，示例如图2：
<p align="center">
<img src="./images/FreezePass.png" height=400 width=420 hspace='10'/> <br />
<strong>图2：应用QuantizationFreezePass后的结果</strong>
<strong>图2：paddleslim.quant.convert 后的结果</strong>
</p>
#### 8-bit模型
在对训练网络进行QuantizationFreezePass之后，执行ConvertToInt8Pass，
其主要目的是将执行完QuantizationFreezePass后输出的权重类型由`FP32`更改为`INT8`。换言之，用户可以选择将量化后的权重保存为float32类型（不执行ConvertToInt8Pass）或者int8_t类型（执行ConvertToInt8Pass），示例如图3：
<p align="center">
<img src="./images/ConvertToInt8Pass.png" height=400 width=400 hspace='10'/> <br />
<strong>图3：应用ConvertToInt8Pass后的结果</strong>
</p>
> 综上，可得在量化过程中有以下几种模型结构：
1. 原始模型
2. 经QuantizationTransformPass之后得到的适用于训练的量化模型结构，在${checkpoint_path}下保存的`eval_model`是这种结构，在训练过程中每个epoch结束时也使用这个网络结构进行评估，虽然这个模型结构不是最终想要的模型结构，但是每个epoch的评估结果可用来挑选模型。
3. 经QuantizationFreezePass之后得到的FP32模型结构，具体结构已在上面进行介绍。本文档中列出的数据集的评估结果是对FP32模型结构进行评估得到的结果。这种模型结构在训练过程中只会保存一次，也就是在量化配置文件中设置的`end_epoch`结束时进行保存，如果想将其他epoch的训练结果转化成FP32模型，可使用脚本 <a href='./freeze.py'>PaddleSlim/classification/quantization/freeze.py</a>进行转化，具体使用方法在[评估](#评估)中介绍。
4. 经ConvertToInt8Pass之后得到的8-bit模型结构，具体结构已在上面进行介绍。这种模型结构在训练过程中只会保存一次，也就是在量化配置文件中设置的`end_epoch`结束时进行保存，如果想将其他epoch的训练结果转化成8-bit模型，可使用脚本 <a href='./freeze.py'>slim/quantization/freeze.py</a>进行转化，具体使用方法在[评估](#评估)中介绍。
所以在调用 ``paddleslim.quant.convert`` 之后，才得到最终的量化模型。此模型可使用PaddleLite进行加载预测，可参见教程[Paddle-Lite如何加载运行量化模型](https://github.com/PaddlePaddle/Paddle-Lite/wiki/model_quantization)。
### 评估脚本
使用脚本[slim/quantization/eval.py](./eval.py)进行评估。
## 评估
### 每个epoch保存的评估模型
因为量化的最终模型只有在end_epoch时保存一次，不能保证保存的模型是最好的，因此
如果在配置文件中设置了`checkpoint_path`，则每个epoch会保存一个量化后的用于评估的模型，
该模型会保存在`${checkpoint_path}/${epoch_id}/eval_model/`路径下，包含`__model__`和`__params__`两个文件。
其中，`__model__`用于保存模型结构信息，`__params__`用于保存参数（parameters）信息。模型结构和训练时一样。
- 定义配置。使用和训练脚本中一样的量化配置，以得到和量化训练时同样的模型。
- 使用 ``paddleslim.quant.quant_aware`` 插入量化和反量化op。
- 使用 ``paddleslim.quant.convert`` 改变op顺序，得到最终量化模型进行评估。
如果不需要保存评估模型，可以在定义Compressor对象时，将`save_eval_model`选项设置为False（默认为True）。
评估命令：
脚本<a href="../eval.py">slim/eval.py</a>中为使用该模型在评估数据集上做评估的示例。
运行命令为：
```
python ../eval.py \
--model_path ${checkpoint_path}/${epoch_id}/eval_model/ \
--model_name __model__ \
--params_name __params__ \
-c ../../configs/yolov3_mobilenet_v1_voc.yml \
-d "../../dataset/voc"
python slim/quantization/eval.py -c ./configs/yolov3_mobilenet_v1.yml \
-o weights=./output/mobilenetv1/yolov3_mobilenet_v1/best_model
```
在评估之后，选取效果最好的epoch的模型，可使用脚本 <a href='./freeze.py'>slim/quantization/freeze.py</a>将该模型转化为以上介绍的2种模型：FP32模型，int8模型，需要配置的参数为：
## 导出模型
- model_path, 加载的模型路径，`为${checkpoint_path}/${epoch_id}/eval_model/`
- weight_quant_type 模型参数的量化方式，和配置文件中的类型保持一致
- save_path `FP32`, `8-bit` 模型的保存路径，分别为 `${save_path}/float/`, `${save_path}/int8/`
使用脚本[slim/quantization/export_model.py](./export_model.py)导出模型。
运行命令示例：
```
python freeze.py \
--model_path ${checkpoint_path}/${epoch_id}/eval_model/ \
--weight_quant_type ${weight_quant_type} \
--save_path ${any path you want} \
-c ../../configs/yolov3_mobilenet_v1_voc.yml \
-d "../../dataset/voc"
```
- 定义配置。使用和训练脚本中一样的量化配置，以得到和量化训练时同样的模型。
- 使用 ``paddleslim.quant.quant_aware`` 插入量化和反量化op。
- 使用 ``paddleslim.quant.convert`` 改变op顺序，得到最终量化模型进行评估。
导出模型命令：
### 最终评估模型
最终使用的评估模型是FP32模型，使用脚本<a href="../eval.py">slim/eval.py</a>中为使用该模型在评估数据集上做评估的示例。
运行命令为：
```
python ../eval.py \
--model_path ${float_model_path}
--model_name model \
--params_name weights \
-c ../../configs/yolov3_mobilenet_v1_voc.yml \
-d "../../dataset/voc"
python slim/quantization/export_model.py -c ./configs/yolov3_mobilenet_v1.yml --output_dir ${save path} \
-o weights=./output/mobilenetv1/yolov3_mobilenet_v1/best_model
```
## 预测
### python预测
FP32模型可直接使用原生PaddlePaddle Fluid预测方法进行预测。
在脚本<a href="../infer.py">slim/infer.py</a>中展示了如何使用fluid python API加载使用预测模型进行预测。
在脚本<a href="./infer.py">slim/quantization/infer.py</a>中展示了如何使用fluid python API加载使用预测模型进行预测。
运行命令示例:
```
python ../infer.py \
--model_path ${save_path}/float \
--model_name model \
--params_name weights \
-c ../../configs/yolov3_mobilenet_v1_voc.yml \
--infer_dir ../../demo
python slim/quantization/infer.py \
-c ./configs/yolov3_mobilenet_v1.yml \
--infer_dir ./demo \
-o weights=./output/mobilenetv1/yolov3_mobilenet_v1/best_model
```
### PaddleLite预测
FP32模型可使用PaddleLite进行加载预测，可参见教程[Paddle-Lite如何加载运行量化模型](https://github.com/PaddlePaddle/Paddle-Lite/wiki/model_quantization)
## 示例结果
>当前release的结果并非超参调优后的最好结果，仅做示例参考，后续我们会优化当前结果。
导出模型步骤中导出的FP32模型可使用PaddleLite进行加载预测，可参见教程[Paddle-Lite如何加载运行量化模型](https://github.com/PaddlePaddle/Paddle-Lite/wiki/model_quantization)
### MobileNetV1-YOLO-V3
| weight量化方式 | activation量化方式| Box ap |Paddle Fluid inference time(ms)| Paddle Lite inference time(ms)|
|---|---|---|---|---|
|baseline|- |76.2%|- |-|
|abs_max|abs_max|- |- |-|
|abs_max|moving_average_abs_max|- |- |-|
|channel_wise_abs_max|abs_max|- |- |-|
## 量化结果
## FAQ
slim/quantization/compress.py 已删除 100644 → 0
浏览文件 @ e5fe96e7
# Copyright (c) 2019 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.
from __future__ import absolute_import
from __future__ import division
from __future__ import print_function
import os
import time
import multiprocessing
import numpy as np
import datetime
from collections import deque
import sys
sys.path.append("../../")
from paddle.fluid.contrib.slim import Compressor
from paddle.fluid.framework import IrGraph
from paddle.fluid import core
def set_paddle_flags(**kwargs):
for key, value in kwargs.items():
if os.environ.get(key, None) is None:
os.environ[key] = str(value)
# NOTE(paddle-dev): All of these flags should be set before
# `import paddle`. Otherwise, it would not take any effect.
set_paddle_flags(
FLAGS_eager_delete_tensor_gb=0, # enable GC to save memory
)
from paddle import fluid
from ppdet.core.workspace import load_config, merge_config, create
from ppdet.data.data_feed import create_reader
from ppdet.utils.eval_utils import parse_fetches, eval_results
from ppdet.utils.stats import TrainingStats
from ppdet.utils.cli import ArgsParser, print_total_cfg
from ppdet.utils.check import check_gpu, check_version
import ppdet.utils.checkpoint as checkpoint
from ppdet.modeling.model_input import create_feed
import logging
FORMAT = '%(asctime)s-%(levelname)s: %(message)s'
logging.basicConfig(level=logging.INFO, format=FORMAT)
logger = logging.getLogger(__name__)
def eval_run(exe, compile_program, reader, keys, values, cls, test_feed, cfg):
"""
Run evaluation program, return program outputs.
"""
iter_id = 0
results = []
if len(cls) != 0:
values = []
for i in range(len(cls)):
_, accum_map = cls[i].get_map_var()
cls[i].reset(exe)
values.append(accum_map)
images_num = 0
start_time = time.time()
has_bbox = 'bbox' in keys
for data in reader():
data = test_feed.feed(data)
feed_data = {'image': data['image'], 'im_size': data['im_size']}
outs = exe.run(compile_program,
feed=feed_data,
fetch_list=[values[0]],
return_numpy=False)
if cfg.metric == 'VOC':
outs.append(data['gt_box'])
outs.append(data['gt_label'])
outs.append(data['is_difficult'])
elif cfg.metric == 'COCO':
outs.append(data['im_id'])
res = {
k: (np.array(v), v.recursive_sequence_lengths())
for k, v in zip(keys, outs)
}
results.append(res)
if iter_id % 100 == 0:
logger.info('Test iter {}'.format(iter_id))
iter_id += 1
images_num += len(res['bbox'][1][0]) if has_bbox else 1
logger.info('Test finish iter {}'.format(iter_id))
end_time = time.time()
fps = images_num / (end_time - start_time)
if has_bbox:
logger.info('Total number of images: {}, inference time: {} fps.'.
format(images_num, fps))
else:
logger.info('Total iteration: {}, inference time: {} batch/s.'.format(
images_num, fps))
return results
def main():
cfg = load_config(FLAGS.config)
if 'architecture' in cfg:
main_arch = cfg.architecture
else:
raise ValueError("'architecture' not specified in config file.")
merge_config(FLAGS.opt)
if 'log_iter' not in cfg:
cfg.log_iter = 20
# check if set use_gpu=True in paddlepaddle cpu version
check_gpu(cfg.use_gpu)
# print_total_cfg(cfg)
#check_version()
if cfg.use_gpu:
devices_num = fluid.core.get_cuda_device_count()
else:
devices_num = int(
os.environ.get('CPU_NUM', multiprocessing.cpu_count()))
if 'train_feed' not in cfg:
train_feed = create(main_arch + 'TrainFeed')
else:
train_feed = create(cfg.train_feed)
if 'eval_feed' not in cfg:
eval_feed = create(main_arch + 'EvalFeed')
else:
eval_feed = create(cfg.eval_feed)
place = fluid.CUDAPlace(0) if cfg.use_gpu else fluid.CPUPlace()
exe = fluid.Executor(place)
lr_builder = create('LearningRate')
optim_builder = create('OptimizerBuilder')
# build program
startup_prog = fluid.Program()
train_prog = fluid.Program()
with fluid.program_guard(train_prog, startup_prog):
with fluid.unique_name.guard():
model = create(main_arch)
_, feed_vars = create_feed(train_feed, True)
train_fetches = model.train(feed_vars)
loss = train_fetches['loss']
lr = lr_builder()
optimizer = optim_builder(lr)
optimizer.minimize(loss)
train_reader = create_reader(train_feed, cfg.max_iters, FLAGS.dataset_dir)
# parse train fetches
train_keys, train_values, _ = parse_fetches(train_fetches)
train_values.append(lr)
train_fetch_list = []
for k, v in zip(train_keys, train_values):
train_fetch_list.append((k, v))
print("train_fetch_list: {}".format(train_fetch_list))
eval_prog = fluid.Program()
with fluid.program_guard(eval_prog, startup_prog):
with fluid.unique_name.guard():
model = create(main_arch)
_, test_feed_vars = create_feed(eval_feed, True)
fetches = model.eval(test_feed_vars)
eval_prog = eval_prog.clone(True)
eval_reader = create_reader(eval_feed, args_path=FLAGS.dataset_dir)
#eval_pyreader.decorate_sample_list_generator(eval_reader, place)
test_data_feed = fluid.DataFeeder(test_feed_vars.values(), place)
# parse eval fetches
extra_keys = []
if cfg.metric == 'COCO':
extra_keys = ['im_info', 'im_id', 'im_shape']
if cfg.metric == 'VOC':
extra_keys = ['gt_box', 'gt_label', 'is_difficult']
eval_keys, eval_values, eval_cls = parse_fetches(fetches, eval_prog,
extra_keys)
# print(eval_values)
eval_fetch_list = []
for k, v in zip(eval_keys, eval_values):
eval_fetch_list.append((k, v))
exe.run(startup_prog)
start_iter = 0
checkpoint.load_params(exe, train_prog, cfg.pretrain_weights)
best_box_ap_list = []
def eval_func(program, scope):
#place = fluid.CPUPlace()
#exe = fluid.Executor(place)
results = eval_run(exe, program, eval_reader, eval_keys, eval_values,
eval_cls, test_data_feed, cfg)
resolution = None
if 'mask' in results[0]:
resolution = model.mask_head.resolution
box_ap_stats = eval_results(results, eval_feed, cfg.metric,
cfg.num_classes, resolution, False,
FLAGS.output_eval)
if len(best_box_ap_list) == 0:
best_box_ap_list.append(box_ap_stats[0])
elif box_ap_stats[0] > best_box_ap_list[0]:
best_box_ap_list[0] = box_ap_stats[0]
logger.info("Best test box ap: {}".format(best_box_ap_list[0]))
return best_box_ap_list[0]
test_feed = [('image', test_feed_vars['image'].name),
('im_size', test_feed_vars['im_size'].name)]
com = Compressor(
place,
fluid.global_scope(),
train_prog,
train_reader=train_reader,
train_feed_list=[(key, value.name) for key, value in feed_vars.items()],
train_fetch_list=train_fetch_list,
eval_program=eval_prog,
eval_reader=eval_reader,
eval_feed_list=test_feed,
eval_func={'map': eval_func},
eval_fetch_list=[eval_fetch_list[0]],
prune_infer_model=[["image", "im_size"], ["multiclass_nms_0.tmp_0"]],
train_optimizer=None)
com.config(FLAGS.slim_file)
com.run()
if __name__ == '__main__':
parser = ArgsParser()
parser.add_argument(
"-s",
"--slim_file",
default=None,
type=str,
help="Config file of PaddleSlim.")
parser.add_argument(
"--output_eval",
default=None,
type=str,
help="Evaluation directory, default is current directory.")
parser.add_argument(
"-d",
"--dataset_dir",
default=None,
type=str,
help="Dataset path, same as DataFeed.dataset.dataset_dir")
FLAGS = parser.parse_args()
main()
slim/quantization/eval.py 0 → 100644
浏览文件 @ 36d1bc7b
# Copyright (c) 2019 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.
from __future__ import absolute_import
from __future__ import division
from __future__ import print_function
import os
import sys
import paddle.fluid as fluid
from ppdet.utils.eval_utils import parse_fetches, eval_run, eval_results, json_eval_results
import ppdet.utils.checkpoint as checkpoint
from ppdet.utils.check import check_gpu, check_version
from ppdet.data.reader import create_reader
from ppdet.core.workspace import load_config, merge_config, create
from ppdet.utils.cli import ArgsParser
import logging
FORMAT = '%(asctime)s-%(levelname)s: %(message)s'
logging.basicConfig(level=logging.INFO, format=FORMAT)
logger = logging.getLogger(__name__)
# import paddleslim
from paddleslim.quant import quant_aware, convert
def main():
"""
Main evaluate function
"""
cfg = load_config(FLAGS.config)
if 'architecture' in cfg:
main_arch = cfg.architecture
else:
raise ValueError("'architecture' not specified in config file.")
merge_config(FLAGS.opt)
# check if set use_gpu=True in paddlepaddle cpu version
check_gpu(cfg.use_gpu)
# check if paddlepaddle version is satisfied
check_version()
# define executor
place = fluid.CUDAPlace(0) if cfg.use_gpu else fluid.CPUPlace()
exe = fluid.Executor(place)
# build program
model = create(main_arch)
startup_prog = fluid.Program()
eval_prog = fluid.Program()
with fluid.program_guard(eval_prog, startup_prog):
with fluid.unique_name.guard():
inputs_def = cfg['EvalReader']['inputs_def']
test_feed_vars, loader = model.build_inputs(**inputs_def)
test_fetches = model.eval(test_feed_vars)
eval_prog = eval_prog.clone(True)
reader = create_reader(cfg.EvalReader)
loader.set_sample_list_generator(reader, place)
# eval already exists json file
if FLAGS.json_eval:
logger.info(
"In json_eval mode, PaddleDetection will evaluate json files in "
"output_eval directly. And proposal.json, bbox.json and mask.json "
"will be detected by default.")
json_eval_results(
cfg.metric, json_directory=FLAGS.output_eval, dataset=dataset)
return
assert cfg.metric != 'OID', "eval process of OID dataset \
is not supported."
if cfg.metric == "WIDERFACE":
raise ValueError("metric type {} does not support in tools/eval.py, "
"please use tools/face_eval.py".format(cfg.metric))
assert cfg.metric in ['COCO', 'VOC'], \
"unknown metric type {}".format(cfg.metric)
extra_keys = []
if cfg.metric == 'COCO':
extra_keys = ['im_info', 'im_id', 'im_shape']
if cfg.metric == 'VOC':
extra_keys = ['gt_bbox', 'gt_class', 'is_difficult']
keys, values, cls = parse_fetches(test_fetches, eval_prog, extra_keys)
# whether output bbox is normalized in model output layer
is_bbox_normalized = False
if hasattr(model, 'is_bbox_normalized') and \
callable(model.is_bbox_normalized):
is_bbox_normalized = model.is_bbox_normalized()
dataset = cfg['EvalReader']['dataset']
sub_eval_prog = None
sub_keys = None
sub_values = None
not_quant_pattern = []
if FLAGS.not_quant_pattern:
not_quant_pattern = FLAGS.not_quant_pattern
config = {
'weight_quantize_type': 'channel_wise_abs_max',
'activation_quantize_type': 'moving_average_abs_max',
'quantize_op_types': ['depthwise_conv2d', 'mul', 'conv2d'],
'not_quant_pattern': not_quant_pattern
}
eval_prog = quant_aware(eval_prog, place, config, for_test=True)
# load model
exe.run(startup_prog)
if 'weights' in cfg:
checkpoint.load_params(exe, eval_prog, cfg.weights)
eval_prog = convert(eval_prog, place, config, save_int8=False)
compile_program = fluid.compiler.CompiledProgram(
eval_prog).with_data_parallel()
results = eval_run(exe, compile_program, loader, keys, values, cls, cfg,
sub_eval_prog, sub_keys, sub_values)
# evaluation
resolution = None
if 'mask' in results[0]:
resolution = model.mask_head.resolution
# if map_type not set, use default 11point, only use in VOC eval
map_type = cfg.map_type if 'map_type' in cfg else '11point'
eval_results(
results,
cfg.metric,
cfg.num_classes,
resolution,
is_bbox_normalized,
FLAGS.output_eval,
map_type,
dataset=dataset)
if __name__ == '__main__':
parser = ArgsParser()
parser.add_argument(
"--json_eval",
action='store_true',
default=False,
help="Whether to re eval with already exists bbox.json or mask.json")
parser.add_argument(
"-f",
"--output_eval",
default=None,
type=str,
help="Evaluation file directory, default is current directory.")
parser.add_argument(
"--not_quant_pattern",
nargs='+',
type=str,
help="Layers which name_scope contains string in not_quant_pattern will not be quantized"
)
FLAGS = parser.parse_args()
main()
slim/quantization/export_model.py 0 → 100644
浏览文件 @ 36d1bc7b
# Copyright (c) 2019 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.
from __future__ import absolute_import
from __future__ import division
from __future__ import print_function
import os
import sys
from paddle import fluid
from ppdet.core.workspace import load_config, merge_config, create
from ppdet.modeling.model_input import create_feed
from ppdet.utils.cli import ArgsParser
import ppdet.utils.checkpoint as checkpoint
from tools.export_model import prune_feed_vars
import logging
FORMAT = '%(asctime)s-%(levelname)s: %(message)s'
logging.basicConfig(level=logging.INFO, format=FORMAT)
logger = logging.getLogger(__name__)
from paddleslim.quant import quant_aware, convert
def save_infer_model(save_dir, exe, feed_vars, test_fetches, infer_prog):
feed_var_names = [var.name for var in feed_vars.values()]
target_vars = list(test_fetches.values())
feed_var_names = prune_feed_vars(feed_var_names, target_vars, infer_prog)
logger.info("Export inference model to {}, input: {}, output: "
"{}...".format(save_dir, feed_var_names,
[str(var.name) for var in target_vars]))
fluid.io.save_inference_model(
save_dir,
feeded_var_names=feed_var_names,
target_vars=target_vars,
executor=exe,
main_program=infer_prog,
params_filename="__params__")
def main():
cfg = load_config(FLAGS.config)
if 'architecture' in cfg:
main_arch = cfg.architecture
else:
raise ValueError("'architecture' not specified in config file.")
merge_config(FLAGS.opt)
# Use CPU for exporting inference model instead of GPU
place = fluid.CPUPlace()
exe = fluid.Executor(place)
model = create(main_arch)
startup_prog = fluid.Program()
infer_prog = fluid.Program()
with fluid.program_guard(infer_prog, startup_prog):
with fluid.unique_name.guard():
inputs_def = cfg['TestReader']['inputs_def']
inputs_def['use_dataloader'] = False
feed_vars, _ = model.build_inputs(**inputs_def)
test_fetches = model.test(feed_vars)
infer_prog = infer_prog.clone(True)
not_quant_pattern = []
if FLAGS.not_quant_pattern:
not_quant_pattern = FLAGS.not_quant_pattern
config = {
'weight_quantize_type': 'channel_wise_abs_max',
'activation_quantize_type': 'moving_average_abs_max',
'quantize_op_types': ['depthwise_conv2d', 'mul', 'conv2d'],
'not_quant_pattern': not_quant_pattern
}
infer_prog = quant_aware(infer_prog, place, config, for_test=True)
exe.run(startup_prog)
checkpoint.load_params(exe, infer_prog, cfg.weights)
infer_prog, int8_program = convert(
infer_prog, place, config, save_int8=True)
save_infer_model(
os.path.join(FLAGS.output_dir, 'float'), exe, feed_vars, test_fetches,
infer_prog)
save_infer_model(
os.path.join(FLAGS.output_dir, 'int'), exe, feed_vars, test_fetches,
int8_program)
if __name__ == '__main__':
parser = ArgsParser()
parser.add_argument(
"--output_dir",
type=str,
default="output",
help="Directory for storing the output model files.")
parser.add_argument(
"--not_quant_pattern",
nargs='+',
type=str,
help="Layers which name_scope contains string in not_quant_pattern will not be quantized"
)
FLAGS = parser.parse_args()
main()
slim/quantization/infer.py 0 → 100644
浏览文件 @ 36d1bc7b
# Copyright (c) 2019 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.
from __future__ import absolute_import
from __future__ import division
from __future__ import print_function
import os
import glob
import sys
import numpy as np
from PIL import Image
from paddle import fluid
from ppdet.core.workspace import load_config, merge_config, create
from ppdet.utils.eval_utils import parse_fetches
from ppdet.utils.cli import ArgsParser
from ppdet.utils.check import check_gpu, check_version
from ppdet.utils.visualizer import visualize_results
import ppdet.utils.checkpoint as checkpoint
from ppdet.data.reader import create_reader
from tools.infer import get_test_images, get_save_image_name
import logging
FORMAT = '%(asctime)s-%(levelname)s: %(message)s'
logging.basicConfig(level=logging.INFO, format=FORMAT)
logger = logging.getLogger(__name__)
from paddleslim.quant import quant_aware, convert
def main():
cfg = load_config(FLAGS.config)
if 'architecture' in cfg:
main_arch = cfg.architecture
else:
raise ValueError("'architecture' not specified in config file.")
merge_config(FLAGS.opt)
# check if set use_gpu=True in paddlepaddle cpu version
check_gpu(cfg.use_gpu)
# check if paddlepaddle version is satisfied
check_version()
dataset = cfg.TestReader['dataset']
test_images = get_test_images(FLAGS.infer_dir, FLAGS.infer_img)
dataset.set_images(test_images)
place = fluid.CUDAPlace(0) if cfg.use_gpu else fluid.CPUPlace()
exe = fluid.Executor(place)
model = create(main_arch)
startup_prog = fluid.Program()
infer_prog = fluid.Program()
with fluid.program_guard(infer_prog, startup_prog):
with fluid.unique_name.guard():
inputs_def = cfg['TestReader']['inputs_def']
feed_vars, loader = model.build_inputs(**inputs_def)
test_fetches = model.test(feed_vars)
infer_prog = infer_prog.clone(True)
reader = create_reader(cfg.TestReader)
loader.set_sample_list_generator(reader, place)
not_quant_pattern = []
if FLAGS.not_quant_pattern:
not_quant_pattern = FLAGS.not_quant_pattern
config = {
'weight_quantize_type': 'channel_wise_abs_max',
'activation_quantize_type': 'moving_average_abs_max',
'quantize_op_types': ['depthwise_conv2d', 'mul', 'conv2d'],
'not_quant_pattern': not_quant_pattern
}
infer_prog = quant_aware(infer_prog, place, config, for_test=True)
exe.run(startup_prog)
if cfg.weights:
checkpoint.load_params(exe, infer_prog, cfg.weights)
infer_prog = convert(infer_prog, place, config, save_int8=False)
# parse infer fetches
assert cfg.metric in ['COCO', 'VOC', 'OID', 'WIDERFACE'], \
"unknown metric type {}".format(cfg.metric)
extra_keys = []
if cfg['metric'] in ['COCO', 'OID']:
extra_keys = ['im_info', 'im_id', 'im_shape']
if cfg['metric'] == 'VOC' or cfg['metric'] == 'WIDERFACE':
extra_keys = ['im_id', 'im_shape']
keys, values, _ = parse_fetches(test_fetches, infer_prog, extra_keys)
# parse dataset category
if cfg.metric == 'COCO':
from ppdet.utils.coco_eval import bbox2out, mask2out, get_category_info
if cfg.metric == 'OID':
from ppdet.utils.oid_eval import bbox2out, get_category_info
if cfg.metric == "VOC":
from ppdet.utils.voc_eval import bbox2out, get_category_info
if cfg.metric == "WIDERFACE":
from ppdet.utils.widerface_eval_utils import bbox2out, get_category_info
anno_file = dataset.get_anno()
with_background = dataset.with_background
use_default_label = dataset.use_default_label
clsid2catid, catid2name = get_category_info(anno_file, with_background,
use_default_label)
# whether output bbox is normalized in model output layer
is_bbox_normalized = False
if hasattr(model, 'is_bbox_normalized') and \
callable(model.is_bbox_normalized):
is_bbox_normalized = model.is_bbox_normalized()
imid2path = dataset.get_imid2path()
iter_id = 0
try:
loader.start()
while True:
outs = exe.run(infer_prog, fetch_list=values, return_numpy=False)
res = {
k: (np.array(v), v.recursive_sequence_lengths())
for k, v in zip(keys, outs)
}
logger.info('Infer iter {}'.format(iter_id))
iter_id += 1
bbox_results = None
mask_results = None
if 'bbox' in res:
bbox_results = bbox2out([res], clsid2catid, is_bbox_normalized)
if 'mask' in res:
mask_results = mask2out([res], clsid2catid,
model.mask_head.resolution)
# visualize result
im_ids = res['im_id'][0]
for im_id in im_ids:
image_path = imid2path[int(im_id)]
image = Image.open(image_path).convert('RGB')
image = visualize_results(image,
int(im_id), catid2name,
FLAGS.draw_threshold, bbox_results,
mask_results)
save_name = get_save_image_name(FLAGS.output_dir, image_path)
logger.info("Detection bbox results save in {}".format(
save_name))
image.save(save_name, quality=95)
except (StopIteration, fluid.core.EOFException):
loader.reset()
if __name__ == '__main__':
parser = ArgsParser()
parser.add_argument(
"--infer_dir",
type=str,
default=None,
help="Directory for images to perform inference on.")
parser.add_argument(
"--infer_img",
type=str,
default=None,
help="Image path, has higher priority over --infer_dir")
parser.add_argument(
"--output_dir",
type=str,
default="output",
help="Directory for storing the output visualization files.")
parser.add_argument(
"--draw_threshold",
type=float,
default=0.5,
help="Threshold to reserve the result for visualization.")
parser.add_argument(
"--not_quant_pattern",
nargs='+',
type=str,
help="Layers which name_scope contains string in not_quant_pattern will not be quantized"
)
FLAGS = parser.parse_args()
main()
slim/quantization/train.py 0 → 100644
浏览文件 @ 36d1bc7b
# Copyright (c) 2019 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.
from __future__ import absolute_import
from __future__ import division
from __future__ import print_function
import os
import sys
import time
import numpy as np
import datetime
from collections import deque
from paddle import fluid
from ppdet.core.workspace import load_config, merge_config, create
from ppdet.data.reader import create_reader
from ppdet.utils.cli import print_total_cfg
from ppdet.utils import dist_utils
from ppdet.utils.eval_utils import parse_fetches, eval_run, eval_results
from ppdet.utils.stats import TrainingStats
from ppdet.utils.cli import ArgsParser
from ppdet.utils.check import check_gpu, check_version
import ppdet.utils.checkpoint as checkpoint
from paddleslim.quant import quant_aware, convert
import logging
FORMAT = '%(asctime)s-%(levelname)s: %(message)s'
logging.basicConfig(level=logging.INFO, format=FORMAT)
logger = logging.getLogger(__name__)
def main():
env = os.environ
FLAGS.dist = 'PADDLE_TRAINER_ID' in env and 'PADDLE_TRAINERS_NUM' in env
if FLAGS.dist:
trainer_id = int(env['PADDLE_TRAINER_ID'])
import random
local_seed = (99 + trainer_id)
random.seed(local_seed)
np.random.seed(local_seed)
cfg = load_config(FLAGS.config)
if 'architecture' in cfg:
main_arch = cfg.architecture
else:
raise ValueError("'architecture' not specified in config file.")
merge_config(FLAGS.opt)
if 'log_iter' not in cfg:
cfg.log_iter = 20
# check if set use_gpu=True in paddlepaddle cpu version
check_gpu(cfg.use_gpu)
# check if paddlepaddle version is satisfied
check_version()
if not FLAGS.dist or trainer_id == 0:
print_total_cfg(cfg)
if cfg.use_gpu:
devices_num = fluid.core.get_cuda_device_count()
else:
devices_num = int(os.environ.get('CPU_NUM', 1))
if 'FLAGS_selected_gpus' in env:
device_id = int(env['FLAGS_selected_gpus'])
else:
device_id = 0
place = fluid.CUDAPlace(device_id) if cfg.use_gpu else fluid.CPUPlace()
exe = fluid.Executor(place)
lr_builder = create('LearningRate')
optim_builder = create('OptimizerBuilder')
# build program
startup_prog = fluid.Program()
train_prog = fluid.Program()
with fluid.program_guard(train_prog, startup_prog):
with fluid.unique_name.guard():
model = create(main_arch)
inputs_def = cfg['TrainReader']['inputs_def']
feed_vars, train_loader = model.build_inputs(**inputs_def)
train_fetches = model.train(feed_vars)
loss = train_fetches['loss']
lr = lr_builder()
optimizer = optim_builder(lr)
optimizer.minimize(loss)
# parse train fetches
train_keys, train_values, _ = parse_fetches(train_fetches)
train_values.append(lr)
if FLAGS.eval:
eval_prog = fluid.Program()
with fluid.program_guard(eval_prog, startup_prog):
with fluid.unique_name.guard():
model = create(main_arch)
inputs_def = cfg['EvalReader']['inputs_def']
feed_vars, eval_loader = model.build_inputs(**inputs_def)
fetches = model.eval(feed_vars)
eval_prog = eval_prog.clone(True)
eval_reader = create_reader(cfg.EvalReader)
eval_loader.set_sample_list_generator(eval_reader, place)
# parse eval fetches
extra_keys = []
if cfg.metric == 'COCO':
extra_keys = ['im_info', 'im_id', 'im_shape']
if cfg.metric == 'VOC':
extra_keys = ['gt_bbox', 'gt_class', 'is_difficult']
if cfg.metric == 'WIDERFACE':
extra_keys = ['im_id', 'im_shape', 'gt_bbox']
eval_keys, eval_values, eval_cls = parse_fetches(fetches, eval_prog,
extra_keys)
# compile program for multi-devices
build_strategy = fluid.BuildStrategy()
build_strategy.fuse_all_optimizer_ops = False
build_strategy.fuse_elewise_add_act_ops = True
build_strategy.fuse_all_reduce_ops = False
# only enable sync_bn in multi GPU devices
sync_bn = getattr(model.backbone, 'norm_type', None) == 'sync_bn'
sync_bn = False
build_strategy.sync_batch_norm = sync_bn and devices_num > 1 \
and cfg.use_gpu
exec_strategy = fluid.ExecutionStrategy()
# iteration number when CompiledProgram tries to drop local execution scopes.
# Set it to be 1 to save memory usages, so that unused variables in
# local execution scopes can be deleted after each iteration.
exec_strategy.num_iteration_per_drop_scope = 1
if FLAGS.dist:
dist_utils.prepare_for_multi_process(exe, build_strategy, startup_prog,
train_prog)
exec_strategy.num_threads = 1
exe.run(startup_prog)
not_quant_pattern = []
if FLAGS.not_quant_pattern:
not_quant_pattern = FLAGS.not_quant_pattern
config = {
'weight_quantize_type': 'channel_wise_abs_max',
'activation_quantize_type': 'moving_average_abs_max',
'quantize_op_types': ['depthwise_conv2d', 'mul', 'conv2d'],
'not_quant_pattern': not_quant_pattern
}
ignore_params = cfg.finetune_exclude_pretrained_params \
if 'finetune_exclude_pretrained_params' in cfg else []
fuse_bn = getattr(model.backbone, 'norm_type', None) == 'affine_channel'
if FLAGS.resume_checkpoint:
checkpoint.load_checkpoint(exe, train_prog, FLAGS.resume_checkpoint)
start_iter = checkpoint.global_step()
elif cfg.pretrain_weights and fuse_bn and not ignore_params:
checkpoint.load_and_fusebn(exe, train_prog, cfg.pretrain_weights)
elif cfg.pretrain_weights:
checkpoint.load_params(
exe, train_prog, cfg.pretrain_weights, ignore_params=ignore_params)
# insert quantize op in train_prog, return type is CompiledProgram
train_prog = quant_aware(train_prog, place, config, for_test=False)
compiled_train_prog = train_prog.with_data_parallel(
loss_name=loss.name,
build_strategy=build_strategy,
exec_strategy=exec_strategy)
if FLAGS.eval:
# insert quantize op in eval_prog
eval_prog = quant_aware(eval_prog, place, config, for_test=True)
compiled_eval_prog = fluid.compiler.CompiledProgram(eval_prog)
start_iter = 0
train_reader = create_reader(cfg.TrainReader,
(cfg.max_iters - start_iter) * devices_num)
train_loader.set_sample_list_generator(train_reader, place)
# whether output bbox is normalized in model output layer
is_bbox_normalized = False
if hasattr(model, 'is_bbox_normalized') and \
callable(model.is_bbox_normalized):
is_bbox_normalized = model.is_bbox_normalized()
# if map_type not set, use default 11point, only use in VOC eval
map_type = cfg.map_type if 'map_type' in cfg else '11point'
train_stats = TrainingStats(cfg.log_smooth_window, train_keys)
train_loader.start()
start_time = time.time()
end_time = time.time()
cfg_name = os.path.basename(FLAGS.config).split('.')[0]
save_dir = os.path.join(cfg.save_dir, cfg_name)
time_stat = deque(maxlen=cfg.log_smooth_window)
best_box_ap_list = [0.0, 0] #[map, iter]
for it in range(start_iter, cfg.max_iters):
start_time = end_time
end_time = time.time()
time_stat.append(end_time - start_time)
time_cost = np.mean(time_stat)
eta_sec = (cfg.max_iters - it) * time_cost
eta = str(datetime.timedelta(seconds=int(eta_sec)))
outs = exe.run(compiled_train_prog, fetch_list=train_values)
stats = {k: np.array(v).mean() for k, v in zip(train_keys, outs[:-1])}
train_stats.update(stats)
logs = train_stats.log()
if it % cfg.log_iter == 0 and (not FLAGS.dist or trainer_id == 0):
strs = 'iter: {}, lr: {:.6f}, {}, time: {:.3f}, eta: {}'.format(
it, np.mean(outs[-1]), logs, time_cost, eta)
logger.info(strs)
if (it > 0 and it % cfg.snapshot_iter == 0 or it == cfg.max_iters - 1) \
and (not FLAGS.dist or trainer_id == 0):
save_name = str(it) if it != cfg.max_iters - 1 else "model_final"
checkpoint.save(exe, eval_prog, os.path.join(save_dir, save_name))
if FLAGS.eval:
# evaluation
results = eval_run(exe, compiled_eval_prog, eval_loader,
eval_keys, eval_values, eval_cls)
resolution = None
if 'mask' in results[0]:
resolution = model.mask_head.resolution
box_ap_stats = eval_results(
results, cfg.metric, cfg.num_classes, resolution,
is_bbox_normalized, FLAGS.output_eval, map_type,
cfg['EvalReader']['dataset'])
if box_ap_stats[0] > best_box_ap_list[0]:
best_box_ap_list[0] = box_ap_stats[0]
best_box_ap_list[1] = it
checkpoint.save(exe, eval_prog,
os.path.join(save_dir, "best_model"))
logger.info("Best test box ap: {}, in iter: {}".format(
best_box_ap_list[0], best_box_ap_list[1]))
train_loader.reset()
if __name__ == '__main__':
parser = ArgsParser()
parser.add_argument(
"-r",
"--resume_checkpoint",
default=None,
type=str,
help="Checkpoint path for resuming training.")
parser.add_argument(
"--loss_scale",
default=8.,
type=float,
help="Mixed precision training loss scale.")
parser.add_argument(
"--eval",
action='store_true',
default=False,
help="Whether to perform evaluation in train")
parser.add_argument(
"--output_eval",
default=None,
type=str,
help="Evaluation directory, default is current directory.")
parser.add_argument(
"--not_quant_pattern",
nargs='+',
type=str,
help="Layers which name_scope contains string in not_quant_pattern will not be quantized"
)
FLAGS = parser.parse_args()
main()
slim/quantization/yolov3_mobilenet_v1_slim.yaml 已删除 100644 → 0
浏览文件 @ e5fe96e7
version: 1.0
strategies:
quantization_strategy:
class: 'QuantizationStrategy'
start_epoch: 0
end_epoch: 4
float_model_save_path: './output/yolov3/float'
mobile_model_save_path: './output/yolov3/mobile'
int8_model_save_path: './output/yolov3/int8'
weight_bits: 8
activation_bits: 8
weight_quantize_type: 'abs_max'
activation_quantize_type: 'moving_average_abs_max'
save_in_nodes: ['image', 'im_size']
save_out_nodes: ['multiclass_nms_0.tmp_0']
compressor:
epoch: 5
checkpoint_path: './checkpoints/yolov3/'
strategies:
- quantization_strategy
Markdown is supported
0% 或 .
You are about to add 0 people to the discussion. Proceed with caution.
先完成此消息的编辑！
想要评论请 注册 或