Skip to content
体验新版
项目
组织
正在加载...
登录
切换导航
打开侧边栏
s920243400
PaddleDetection
提交
6409d062
P
PaddleDetection
项目概览
s920243400
/
PaddleDetection
与 Fork 源项目一致
Fork自
PaddlePaddle / PaddleDetection
通知
2
Star
0
Fork
0
代码
文件
提交
分支
Tags
贡献者
分支图
Diff
Issue
0
列表
看板
标记
里程碑
合并请求
0
Wiki
0
Wiki
分析
仓库
DevOps
项目成员
Pages
P
PaddleDetection
项目概览
项目概览
详情
发布
仓库
仓库
文件
提交
分支
标签
贡献者
分支图
比较
Issue
0
Issue
0
列表
看板
标记
里程碑
合并请求
0
合并请求
0
Pages
分析
分析
仓库分析
DevOps
Wiki
0
Wiki
成员
成员
收起侧边栏
关闭侧边栏
动态
分支图
创建新Issue
提交
Issue看板
未验证
提交
6409d062
编写于
8月 15, 2022
作者:
W
Wenyu
提交者:
GitHub
8月 15, 2022
浏览文件
操作
浏览文件
下载
电子邮件补丁
差异文件
update train config (#6633)
上级
113ff463
变更
7
隐藏空白更改
内联
并排
Showing
7 changed file
with
17 addition
and
134 deletion
+17
-134
deploy/auto_compression/configs/yolov5_s_qat_dis.yml
deploy/auto_compression/configs/yolov5_s_qat_dis.yml
+3
-5
deploy/auto_compression/configs/yolov6mt_reader.yml
deploy/auto_compression/configs/yolov6mt_reader.yml
+0
-26
deploy/auto_compression/configs/yolov6mt_s_qat_dis.yaml
deploy/auto_compression/configs/yolov6mt_s_qat_dis.yaml
+7
-9
deploy/auto_compression/configs/yolov7_l_qat_dis.yaml
deploy/auto_compression/configs/yolov7_l_qat_dis.yaml
+7
-9
deploy/auto_compression/configs/yolov7_reader.yml
deploy/auto_compression/configs/yolov7_reader.yml
+0
-26
deploy/auto_compression/configs/yolox_reader.yml
deploy/auto_compression/configs/yolox_reader.yml
+0
-26
deploy/auto_compression/configs/yolox_s_qat_dis.yaml
deploy/auto_compression/configs/yolox_s_qat_dis.yaml
+0
-33
未找到文件。
deploy/auto_compression/configs/yolov5_s_qat_dis.yml
浏览文件 @
6409d062
...
...
@@ -20,13 +20,11 @@ Quantization:
-
depthwise_conv2d
TrainConfig
:
train_iter
:
5
000
train_iter
:
3
000
eval_iter
:
1000
learning_rate
:
type
:
CosineAnnealingDecay
learning_rate
:
0.00003
T_max
:
6000
learning_rate
:
0.00001
optimizer_builder
:
optimizer
:
type
:
SGD
weight_decay
:
4.0e-05
target_metric
:
0.365
deploy/auto_compression/configs/yolov6mt_reader.yml
已删除
100644 → 0
浏览文件 @
113ff463
metric
:
COCO
num_classes
:
80
# Datset configuration
TrainDataset
:
!COCODataSet
image_dir
:
train2017
anno_path
:
annotations/instances_train2017.json
dataset_dir
:
dataset/coco/
EvalDataset
:
!COCODataSet
image_dir
:
val2017
anno_path
:
annotations/instances_val2017.json
dataset_dir
:
dataset/coco/
worker_num
:
0
# preprocess reader in test
EvalReader
:
sample_transforms
:
-
Decode
:
{}
-
Resize
:
{
target_size
:
[
640
,
640
],
keep_ratio
:
True
,
interp
:
1
}
-
Pad
:
{
size
:
[
640
,
640
],
fill_value
:
[
114.
,
114.
,
114.
]}
-
Permute
:
{}
batch_size
:
1
deploy/auto_compression/configs/yolov6mt_s_qat_dis.yaml
浏览文件 @
6409d062
Global
:
reader_config
:
configs/yolov
6mt
_reader.yml
reader_config
:
configs/yolov
5
_reader.yml
input_list
:
[
'
image'
,
'
scale_factor'
]
arch
:
YOLO
Evaluation
:
True
...
...
@@ -13,21 +13,19 @@ Distillation:
loss
:
soft_label
Quantization
:
use_pact
:
true
activation_quantize_type
:
'
moving_average_abs_max'
quantize_op_types
:
-
conv2d
-
depthwise_conv2d
TrainConfig
:
train_iter
:
5
000
train_iter
:
8
000
eval_iter
:
1000
learning_rate
:
type
:
CosineAnnealingDecay
learning_rate
:
type
:
CosineAnnealingDecay
learning_rate
:
0.00003
T_max
:
6
000
T_max
:
8
000
optimizer_builder
:
optimizer
:
optimizer
:
type
:
SGD
weight_decay
:
4.0e-05
weight_decay
:
0.00004
deploy/auto_compression/configs/yolov7_l_qat_dis.yaml
浏览文件 @
6409d062
Global
:
reader_config
:
configs/yolov
7
_reader.yml
reader_config
:
configs/yolov
5
_reader.yml
input_list
:
[
'
image'
,
'
scale_factor'
]
arch
:
YOLO
Evaluation
:
True
...
...
@@ -13,21 +13,19 @@ Distillation:
loss
:
soft_label
Quantization
:
use_pact
:
true
activation_quantize_type
:
'
moving_average_abs_max'
quantize_op_types
:
-
conv2d
-
depthwise_conv2d
TrainConfig
:
train_iter
:
5
000
train_iter
:
8
000
eval_iter
:
1000
learning_rate
:
type
:
CosineAnnealingDecay
learning_rate
:
type
:
CosineAnnealingDecay
learning_rate
:
0.00003
T_max
:
6
000
T_max
:
8
000
optimizer_builder
:
optimizer
:
optimizer
:
type
:
SGD
weight_decay
:
4.0e-05
weight_decay
:
0.00004
\ No newline at end of file
deploy/auto_compression/configs/yolov7_reader.yml
已删除
100644 → 0
浏览文件 @
113ff463
metric
:
COCO
num_classes
:
80
# Datset configuration
TrainDataset
:
!COCODataSet
image_dir
:
train2017
anno_path
:
annotations/instances_train2017.json
dataset_dir
:
dataset/coco/
EvalDataset
:
!COCODataSet
image_dir
:
val2017
anno_path
:
annotations/instances_val2017.json
dataset_dir
:
dataset/coco/
worker_num
:
0
# preprocess reader in test
EvalReader
:
sample_transforms
:
-
Decode
:
{}
-
Resize
:
{
target_size
:
[
640
,
640
],
keep_ratio
:
True
,
interp
:
1
}
-
Pad
:
{
size
:
[
640
,
640
],
fill_value
:
[
114.
,
114.
,
114.
]}
-
Permute
:
{}
batch_size
:
1
deploy/auto_compression/configs/yolox_reader.yml
已删除
100644 → 0
浏览文件 @
113ff463
metric
:
COCO
num_classes
:
80
# Datset configuration
TrainDataset
:
!COCODataSet
image_dir
:
train2017
anno_path
:
annotations/instances_train2017.json
dataset_dir
:
dataset/coco/
EvalDataset
:
!COCODataSet
image_dir
:
val2017
anno_path
:
annotations/instances_val2017.json
dataset_dir
:
dataset/coco/
worker_num
:
0
# preprocess reader in test
TestReader
:
sample_transforms
:
-
Decode
:
{}
-
Resize
:
{
target_size
:
[
640
,
640
],
keep_ratio
:
True
,
interp
:
1
}
-
Pad
:
{
size
:
[
640
,
640
],
fill_value
:
[
114.
,
114.
,
114.
]}
-
Permute
:
{}
batch_size
:
1
deploy/auto_compression/configs/yolox_s_qat_dis.yaml
已删除
100644 → 0
浏览文件 @
113ff463
Global
:
reader_config
:
configs/yolox_reader.yml
input_list
:
[
'
image'
,
'
scale_factor'
]
arch
:
YOLO
Evaluation
:
True
model_dir
:
./yolox_s_300e_coco
model_filename
:
model.pdmodel
params_filename
:
model.pdiparams
Distillation
:
alpha
:
1.0
loss
:
soft_label
Quantization
:
use_pact
:
true
activation_quantize_type
:
'
moving_average_abs_max'
quantize_op_types
:
-
conv2d
-
depthwise_conv2d
TrainConfig
:
train_iter
:
5000
eval_iter
:
1000
learning_rate
:
type
:
CosineAnnealingDecay
learning_rate
:
0.00003
T_max
:
6000
optimizer_builder
:
optimizer
:
type
:
SGD
weight_decay
:
4.0e-05
编辑
预览
Markdown
is supported
0%
请重试
或
添加新附件
.
添加附件
取消
You are about to add
0
people
to the discussion. Proceed with caution.
先完成此消息的编辑!
取消
想要评论请
注册
或
登录