Skip to content

Commit 5277010

Browse files
Zeref996Zeref996
authored
Cinn conv 04 (#3028)
* Plt Support torch test, test=model * Plt Support torch test, test=model * plt update torch, test=model * fix plt torch, test=model * plt torch support, test=model * add plt torch case, test=model * add plt torch case, test=model * add plt torch case, test=model * add plt torch case, test=model * add plt torch case, test=model * add plt torch case, test=model * add plt torch case, test=model * add plt torch case, test=model * add cinn converge case * add cinn converge case * add cinn converge case * add cinn converge case * add cinn converge case
1 parent 47fee75 commit 5277010

24 files changed

+24
-0
lines changed

‎framework/e2e/cinn_converge/BlazeFace_CINN.sh

Lines changed: 1 addition & 0 deletions
Original file line numberDiff line numberDiff line change
@@ -0,0 +1 @@
1+
python main.py -c paddlex/configs/modules/face_detection/BlazeFace.yaml -o Global.mode=train -o Global.dataset_dir=../widerface_train -o Train.epochs_iters=1000 -o Train.batch_size=16 -o Train.learning_rate=0.001 -o Global.output='./output/face_detection/BlazeFace_CINN' -o Train.dy2st=True

‎framework/e2e/cinn_converge/BlazeFace_dy.sh

Lines changed: 1 addition & 0 deletions
Original file line numberDiff line numberDiff line change
@@ -0,0 +1 @@
1+
python main.py -c paddlex/configs/modules/face_detection/BlazeFace.yaml -o Global.mode=train -o Global.dataset_dir=../widerface_train -o Train.epochs_iters=1000 -o Train.batch_size=16 -o Train.learning_rate=0.001 -o Global.output='./output/face_detection/BlazeFace_dy'

‎framework/e2e/cinn_converge/Cascade-FasterRCNN-ResNet50-FPN_CINN.sh

Lines changed: 1 addition & 0 deletions
Original file line numberDiff line numberDiff line change
@@ -0,0 +1 @@
1+
python main.py -c paddlex/configs/modules/object_detection/Cascade-FasterRCNN-ResNet50-FPN.yaml -o Global.mode=train -o Global.dataset_dir=../coco -o Train.num_classes=80 -o Train.epochs_iters=12 -o Train.batch_size=1 -o Train.learning_rate=0.01 -o Global.device=gpu:0,1,2,3,4,5,6,7 -o Global.output='./output/object_detection/Cascade-FasterRCNN-ResNet50-FPN_CINN' -o Train.dy2st=True

‎framework/e2e/cinn_converge/Cascade-FasterRCNN-ResNet50-FPN_dy.sh

Lines changed: 1 addition & 0 deletions
Original file line numberDiff line numberDiff line change
@@ -0,0 +1 @@
1+
python main.py -c paddlex/configs/modules/object_detection/Cascade-FasterRCNN-ResNet50-FPN.yaml -o Global.mode=train -o Global.dataset_dir=../coco -o Train.num_classes=80 -o Train.epochs_iters=12 -o Train.batch_size=1 -o Train.learning_rate=0.01 -o Global.device=gpu:0,1,2,3,4,5,6,7 -o Global.output='./output/object_detection/Cascade-FasterRCNN-ResNet50-FPN_dy'

‎framework/e2e/cinn_converge/FCOS-ResNet50_CINN.sh

Lines changed: 1 addition & 0 deletions
Original file line numberDiff line numberDiff line change
@@ -0,0 +1 @@
1+
python main.py -c paddlex/configs/modules/object_detection/FCOS-ResNet50.yaml -o Global.mode=train -o Global.dataset_dir=../coco -o Train.num_classes=80 -o Train.epochs_iters=12 -o Train.batch_size=2 -o Train.learning_rate=0.01 -o Global.device=gpu:0,1,2,3,4,5,6,7 -o Global.output='./output/object_detection/FCOS-ResNet50_CINN' -o Train.dy2st=True

‎framework/e2e/cinn_converge/FCOS-ResNet50_dy.sh

Lines changed: 1 addition & 0 deletions
Original file line numberDiff line numberDiff line change
@@ -0,0 +1 @@
1+
python main.py -c paddlex/configs/modules/object_detection/FCOS-ResNet50.yaml -o Global.mode=train -o Global.dataset_dir=../coco -o Train.num_classes=80 -o Train.epochs_iters=12 -o Train.batch_size=2 -o Train.learning_rate=0.01 -o Global.device=gpu:0,1,2,3,4,5,6,7 -o Global.output='./output/object_detection/FCOS-ResNet50_dy'

‎framework/e2e/cinn_converge/FasterRCNN-ResNet50-vd-FPN_CINN.sh

Lines changed: 1 addition & 0 deletions
Original file line numberDiff line numberDiff line change
@@ -0,0 +1 @@
1+
python main.py -c paddlex/configs/modules/object_detection/FasterRCNN-ResNet50-vd-FPN.yaml -o Global.mode=train -o Global.dataset_dir=../coco -o Train.num_classes=80 -o Train.epochs_iters=12 -o Train.batch_size=1 -o Train.learning_rate=0.01 -o Global.device=gpu:0,1,2,3,4,5,6,7 -o Global.output='./output/object_detection/FasterRCNN-ResNet50-vd-FPN_CINN' -o Train.dy2st=True

‎framework/e2e/cinn_converge/FasterRCNN-ResNet50-vd-FPN_dy.sh

Lines changed: 1 addition & 0 deletions
Original file line numberDiff line numberDiff line change
@@ -0,0 +1 @@
1+
python main.py -c paddlex/configs/modules/object_detection/FasterRCNN-ResNet50-vd-FPN.yaml -o Global.mode=train -o Global.dataset_dir=../coco -o Train.num_classes=80 -o Train.epochs_iters=12 -o Train.batch_size=1 -o Train.learning_rate=0.01 -o Global.device=gpu:0,1,2,3,4,5,6,7 -o Global.output='./output/object_detection/FasterRCNN-ResNet50-vd-FPN_dy'

‎framework/e2e/cinn_converge/MobileFaceNet_CINN.sh

Lines changed: 1 addition & 0 deletions
Original file line numberDiff line numberDiff line change
@@ -0,0 +1 @@
1+
python main.py -c paddlex/configs/modules/face_feature/MobileFaceNet.yaml -o Global.mode=train -o Global.dataset_dir=../MS1M_v3 -o Train.epochs_iters=25 -o Train.batch_size=128 -o Train.learning_rate=4e-3 -o Global.device=gpu:0,1,2,3,4,5,6,7 -o Global.output='./output/face_detection/MobileFaceNet_CINN' -o Train.dy2st=True

‎framework/e2e/cinn_converge/MobileFaceNet_dy.sh

Lines changed: 1 addition & 0 deletions
Original file line numberDiff line numberDiff line change
@@ -0,0 +1 @@
1+
python main.py -c paddlex/configs/modules/face_feature/MobileFaceNet.yaml -o Global.mode=train -o Global.dataset_dir=../MS1M_v3 -o Train.epochs_iters=25 -o Train.batch_size=128 -o Train.learning_rate=4e-3 -o Global.device=gpu:0,1,2,3,4,5,6,7 -o Global.output='./output/face_detection/MobileFaceNet_dy'

‎framework/e2e/cinn_converge/PP-LCNet_x1_0_pedestrian_attribute_CINN.sh

Lines changed: 1 addition & 0 deletions
Original file line numberDiff line numberDiff line change
@@ -0,0 +1 @@
1+
python main.py -c paddlex/configs/modules/pedestrian_attribute_recognition/PP-LCNet_x1_0_pedestrian_attribute.yaml -o Global.mode=train -o Train.num_classes=26 -o epochs_iters=20 -o Train.batch_size=64 -o Train.learning_rate=0.01 -o Global.device=gpu:0,1,2,3 -o Global.dataset_dir=../xingrenfenxi -o Global.output='./output/pedestrian_attribute_recognition/PP-LCNet_x1_0_pedestrian_attribute_CINN' -o Train.dy2st=True

‎framework/e2e/cinn_converge/PP-LCNet_x1_0_pedestrian_attribute_dy.sh

Lines changed: 1 addition & 0 deletions
Original file line numberDiff line numberDiff line change
@@ -0,0 +1 @@
1+
python main.py -c paddlex/configs/modules/pedestrian_attribute_recognition/PP-LCNet_x1_0_pedestrian_attribute.yaml -o Global.mode=train -o Train.num_classes=26 -o epochs_iters=20 -o Train.batch_size=64 -o Train.learning_rate=0.01 -o Global.device=gpu:0,1,2,3 -o Global.dataset_dir=../xingrenfenxi -o Global.output='./output/pedestrian_attribute_recognition/PP-LCNet_x1_0_pedestrian_attribute_dy'

‎framework/e2e/cinn_converge/PP-TinyPose128x96_CINN.sh

Lines changed: 1 addition & 0 deletions
Original file line numberDiff line numberDiff line change
@@ -0,0 +1 @@
1+
python main.py -c paddlex/configs/modules/human_keypoint_detection/PP-TinyPose128x96.yaml -o Global.mode=train -o Global.dataset_dir=../coco -o Train.epochs_iters=420 -o Train.batch_size=512 -o Train.learning_rate=0.008 -o Global.device=gpu:0,1,2,3,4,5,6,7 -o Global.output='./output/human_keypoint_detection/PP-TinyPose128x96_CINN' -o Train.dy2st=True

‎framework/e2e/cinn_converge/PP-TinyPose128x96_dy.sh

Lines changed: 1 addition & 0 deletions
Original file line numberDiff line numberDiff line change
@@ -0,0 +1 @@
1+
python main.py -c paddlex/configs/modules/human_keypoint_detection/PP-TinyPose128x96.yaml -o Global.mode=train -o Global.dataset_dir=../coco -o Train.epochs_iters=420 -o Train.batch_size=512 -o Train.learning_rate=0.008 -o Global.device=gpu:0,1,2,3,4,5,6,7 -o Global.output='./output/human_keypoint_detection/PP-TinyPose128x96_dy'

‎framework/e2e/cinn_converge/PP-YOLOE_plus-S_face_CINN.sh

Lines changed: 1 addition & 0 deletions
Original file line numberDiff line numberDiff line change
@@ -0,0 +1 @@
1+
python main.py -c paddlex/configs/modules/face_detection/PP-YOLOE_plus-S_face.yaml -o Global.mode=train -o Global.dataset_dir=../widerface_train -o Train.epochs_iters=80 -o Train.batch_size=8 -o Train.learning_rate=0.001 -o Global.device=gpu:0,1,2,3,4,5,6,7 -o Global.output='./output/face_detection/PP-YOLOE_plus-S_face_CINN' -o Train.dy2st=True

‎framework/e2e/cinn_converge/PP-YOLOE_plus-S_face_dy.sh

Lines changed: 1 addition & 0 deletions
Original file line numberDiff line numberDiff line change
@@ -0,0 +1 @@
1+
python main.py -c paddlex/configs/modules/face_detection/PP-YOLOE_plus-S_face.yaml -o Global.mode=train -o Global.dataset_dir=../widerface_train -o Train.epochs_iters=80 -o Train.batch_size=8 -o Train.learning_rate=0.001 -o Global.device=gpu:0,1,2,3,4,5,6,7 -o Global.output='./output/face_detection/PP-YOLOE_plus-S_face_dy'

‎framework/e2e/cinn_converge/ResNet50_ML_CINN.sh

Lines changed: 1 addition & 0 deletions
Original file line numberDiff line numberDiff line change
@@ -0,0 +1 @@
1+
python main.py -c paddlex/configs/modules/image_multilabel_classification/ResNet50_ML.yaml -o Global.mode=train -o Train.num_classes=80 -o epochs_iters=40 -o Train.batch_size=32 -o Train.learning_rate=0.00005 -o Global.device=gpu:0,1,2,3,4,5,6,7 -o Global.dataset_dir=../coco_ml -o Global.output='./output/image_multilabel_classification/ResNet50_ML_CINN' -o Train.dy2st=True

‎framework/e2e/cinn_converge/ResNet50_ML_dy.sh

Lines changed: 1 addition & 0 deletions
Original file line numberDiff line numberDiff line change
@@ -0,0 +1 @@
1+
python main.py -c paddlex/configs/modules/image_multilabel_classification/ResNet50_ML.yaml -o Global.mode=train -o Train.num_classes=80 -o epochs_iters=40 -o Train.batch_size=32 -o Train.learning_rate=0.00005 -o Global.device=gpu:0,1,2,3,4,5,6,7 -o Global.dataset_dir=../coco_ml -o Global.output='./output/image_multilabel_classification/ResNet50_ML_dy'

‎framework/e2e/cinn_converge/YOLOX-S_CINN.sh

Lines changed: 1 addition & 0 deletions
Original file line numberDiff line numberDiff line change
@@ -0,0 +1 @@
1+
python main.py -c paddlex/configs/modules/object_detection/YOLOX-S.yaml -o Global.mode=train -o Global.dataset_dir=../coco -o Train.num_classes=80 -o Train.epochs_iters=300 -o Train.batch_size=8 -o Train.learning_rate=0.01 -o Global.device=gpu:0,1,2,3,4,5,6,7 -o Global.output='./output/object_detection/YOLOX-S_CINN' -o Train.dy2st=True

‎framework/e2e/cinn_converge/YOLOX-S_dy.sh

Lines changed: 1 addition & 0 deletions
Original file line numberDiff line numberDiff line change
@@ -0,0 +1 @@
1+
python main.py -c paddlex/configs/modules/object_detection/YOLOX-S.yaml -o Global.mode=train -o Global.dataset_dir=../coco -o Train.num_classes=80 -o Train.epochs_iters=300 -o Train.batch_size=8 -o Train.learning_rate=0.01 -o Global.device=gpu:0,1,2,3,4,5,6,7 -o Global.output='./output/object_detection/YOLOX-S_dy'

‎framework/e2e/cinn_converge/YOLOv3-DarkNet53_CINN.sh

Lines changed: 1 addition & 0 deletions
Original file line numberDiff line numberDiff line change
@@ -0,0 +1 @@
1+
python main.py -c paddlex/configs/modules/object_detection/YOLOv3-DarkNet53.yaml -o Global.mode=train -o Global.dataset_dir=../coco -o Train.num_classes=80 -o Train.epochs_iters=270 -o Train.batch_size=8 -o Train.learning_rate=0.001 -o Global.device=gpu:0,1,2,3,4,5,6,7 -o Global.output='./output/object_detection/YOLOv3-DarkNet53_CINN' -o Train.dy2st=True

‎framework/e2e/cinn_converge/YOLOv3-DarkNet53_dy.sh

Lines changed: 1 addition & 0 deletions
Original file line numberDiff line numberDiff line change
@@ -0,0 +1 @@
1+
python main.py -c paddlex/configs/modules/object_detection/YOLOv3-DarkNet53.yaml -o Global.mode=train -o Global.dataset_dir=../coco -o Train.num_classes=80 -o Train.epochs_iters=270 -o Train.batch_size=8 -o Train.learning_rate=0.001 -o Global.device=gpu:0,1,2,3,4,5,6,7 -o Global.output='./output/object_detection/YOLOv3-DarkNet53_dy'

‎framework/e2e/cinn_converge/YOWO_CINN.sh

Lines changed: 1 addition & 0 deletions
Original file line numberDiff line numberDiff line change
@@ -0,0 +1 @@
1+
python main.py -c paddlex/configs/modules/video_detection/YOWO.yaml -o Global.mode=train -o Global.dataset_dir=../yowo_dataset/ -o Global.output='./output/video_detection/YOWO_CINN' -o Train.dy2st=True

‎framework/e2e/cinn_converge/YOWO_dy.sh

Lines changed: 1 addition & 0 deletions
Original file line numberDiff line numberDiff line change
@@ -0,0 +1 @@
1+
python main.py -c paddlex/configs/modules/video_detection/YOWO.yaml -o Global.mode=train -o Global.dataset_dir=../yowo_dataset/ -o Global.output='./output/video_detection/YOWO_dy'

0 commit comments

Comments
 (0)