[PaddleV3] 模型修改进度跟踪

[megemini]

模型修改进度跟踪

框架	数量	paddle.fluid	算子缺失	模型问题	其他问题	整体修复占比
✅ PyTorch	33	33/33	13/13	4/4	2/2	100%
✅ ONNX	57	57/57	0/0	10/10	0/0	100%
TensorFlow	28	26/28	0/0	1/2	2/3	26/28=93%
✅ Caffe	20	20/20	0/0	1/1	0/0	100%

💠 PyTorch

序号	模型	paddle.fluid	算子缺失	模型问题	其他问题
A-1	ACG_UnitTest	✅ #1073	✅ #1073
A-2	AlexNet	✅ #1065		✅ #1058
A-3	BASNet	✅ #1065
A-4	BertForMaskedLM_dccuchile	✅ #1078	✅ #1078
A-5	BertModel_SpanBert	✅ #1078	✅ #1078
A-6	CamembertForQuestionAnswering	✅ #1078	✅ #1078
A-7	DBFace	✅ #1065
A-8	DPRContextEncoder	✅ #1078	✅ #1078
A-9	DeepLabv3_ResNet50	✅ #1075	✅ #1075
A-10	EDSR	✅ #1065
A-11	EasyOCR_detector	✅ #1084			✅ #1084 模型更新
A-12	EasyOCR_recognizer	✅ #1084			✅ #1084 模型更新
A-13	ElectraModel	✅ #1065
A-14	FCN_ResNet50	✅ #1075	✅ #1075
A-15	FlaubertModel	✅ #1065
A-16	GRU	✅ #1076	✅ #1076
A-17	InceptionV3	✅ #1065
A-18	MNasNet	✅ #1065
A-19	MiniFasNet	✅ #1079		✅ #1079
A-20	MobileNetV2	✅ #1065
A-21	Mobilestereonet	✅ #1073	✅ #1073
A-22	MockingBird	✅ #1080		✅ #1080
A-23	ResNet18	✅ #1065
A-24	ResNet18_2	✅ #1065
A-25	Roberta	✅ #1078	✅ #1078
A-26	Saicinpainting_LaMa	✅ #1073	✅ #1073
A-27	ShuffleNetV2	✅ #1065
A-28	SqueezeNet	✅ #1065
A-29	SwinTransformer	✅ #1079		✅ #1079
A-30	VGG16	✅ #1065
A-31	XLMRobertaForTokenClassification	✅ #1078	✅ #1078
A-32	mobilenetv3	✅ #1065
A-33	opadd	✅ #1073	✅ #1073

💠 ONNX

序号	模型	paddle.fluid	算子缺失	模型问题	其他问题
B-1	FaceNet	✅ #1067
B-2	GPEN	✅ #1067
B-3	IDG_detection_registration_p1	✅ #1067
B-4	IDG_detection_registration_p2	✅ #1067
B-5	IDG_keypoint_selection	✅ #1067
B-6	IDG_local_vis	✅ #1067
B-7	IDG_multi_modal_crop_cnn_model	✅ #1067
B-8	IDG_multi_modal_fully_cnn_model	✅ #1067
B-9	IDG_multi_modal_sequential_model	✅ #1067
B-10	IDG_test_process	✅ #1067
B-11	IDG_test_sim_feat_new	✅ #1067
B-12	MLPerf_ResNet50	✅ #1067
B-13	Poselandmarks	✅ #1067
B-14	Seg	✅ #1067
B-15	SwinTransformer	✅ #1082		✅ #1082
B-16	VIT_online	✅ #1067
B-17	alexnet	✅ #1067
B-18	basnet	✅ #1067
B-19	bert6	✅ #1082		✅ #1082
B-20	best_float32	✅ #1068
B-21	cifarnet	✅ #1068
B-22	circledet	✅ #1068
B-23	crnn_lstm	✅ #1068
B-24	densenet121	✅ #1068
B-25	efficientnet_b0	✅ #1068
B-26	face_cyclegan_34	✅ #1068
B-27	facedetector	✅ #1068
B-28	hand_landmark_ar	✅ #1068
B-29	inceptionv1	✅ #1068
B-30	inceptionv2	✅ #1068
B-31	mmdetection_fasterrcnn	✅ #1082		✅ #1082
B-32	mmdetection_fcos	✅ #1082		✅ #1082
B-33	mmdetection_fsaf	✅ #1082		✅ #1082
B-34	mmdetection_retina	✅ #1082		✅ #1082
B-35	mmdetection_ssd	✅ #1082		✅ #1082
B-36	mmdetection_yolov3	✅ #1082		✅ #1082
B-37	mnasnet_a1	✅ #1068
B-38	mobilefacenet	✅ #1082		✅ #1082
B-39	mobilenet_v2	✅ #1068
B-40	modelGen	✅ #1068
B-41	pnasnet_mobile	✅ #1068
B-42	resnet101	✅ #1068
B-43	resnet152	✅ #1068
B-44	resnet18	✅ #1068
B-45	resnet34	✅ #1068
B-46	resnet50	✅ #1068
B-47	ronin_2d_cp9	✅ #1068
B-48	shufflenet	✅ #1068
B-49	squeezenet	✅ #1068
B-50	transformer	✅ #1068
B-51	vgg11	✅ #1068
B-52	vgg11_bn	✅ #1068
B-53	vgg16_bn	✅ #1068
B-54	vgg19	✅ #1068
B-55	vgg19_bn	✅ #1068
B-56	yolov5s	✅ #1082		✅ #1082
B-57	yolov5s_fix_resize	✅ #1068

💠 TensorFlow

序号	模型	paddle.fluid	算子缺失	模型问题	其他问题
C-1	AlBert	✅ #1069			✅ 兼容 TensorFlow 新版本
C-2	EfficientNet	✅ #1089		✅ #1089
C-3	InceptionResNetV2	✅ #1069
C-4	InceptionV3	✅ #1069
C-5	InceptionV4	✅ #1069
C-6	KerasBert	✅ #1083			✅ #1083 读取输入
C-7	MNASNetA1	✅ #1069
C-8	MTCNN-ONet	✅ #1069
C-9	MTCNN-PNet	✅ #1069
C-10	MTCNN-RNet	✅ #1069
C-11	MobileNetV1	✅ #1069
C-12	MobileNetV2	✅ #1069
C-13	MobileNetV3_large	✅ #1069
C-14	ResNetV1	✅ #1070
C-15	ResNetV2	✅ #1070
C-16	RetinaFace	✅ #1070
C-17	ShuffleNet	✅ #1070
C-18	SqueezeNet	✅ #1070
C-19	ToyUNet-ToyUnet	🔵 #1083			🔵 #1083 读取输入
C-20	ToyUNet-UNet	✅ #1070
C-21	UNet-GanUNet	✅ #1070
C-22	UNet-UNet	✅ #1070
C-23	VGG16	✅ #1070
C-24	YOLOv3	✅ #1070
C-25	YOLOv3_darknet	✅ #1070
C-26	frozen	✅ #1070
C-27	frozen_MobileNetV1	✅ #1070
C-28	pix2pix	🔵 #1091		🔵 #1091

💠 Caffe

序号	模型	paddle.fluid	算子缺失	模型问题	其他问题
D-1	3class	✅ #1071
D-2	AlexNet	✅ #1071
D-3	FaceDetction	✅ #1071			✅ detection_output 兼容新版本
D-4	MNasNet	✅ #1071
D-5	MTCNN1	✅ #1071
D-6	MTCNN2	✅ #1071
D-7	MTCNN3	✅ #1071
D-8	MobileNetV1	✅ #1071
D-9	MobileNetV2	✅ #1071
D-10	ResNet18	✅ #1071
D-11	ResNet50	✅ #1071
D-12	SSD	✅ #1071
D-13	SegFlow	✅ #1071
D-14	ShuffleNet	✅ #1071
D-15	SqueezeNet	✅ #1071
D-16	TSM	✅ #1071
D-17	UNet	✅ #1071
D-18	VGG16	✅ #1071
D-19	VGG19	✅ #1071
D-20	caffemodel1	✅ #1071

修复 paddle.fluid ：

• [PaddleV3] 修复部分依赖 paddle.fluid 的 pytorch 模型 #1065
• [PaddleV3] 修复部分依赖 paddle.fluid 的 onnx 模型 part 1 #1067
• [PaddleV3] 修复部分依赖 paddle.fluid 的 onnx 模型 part 2 #1068
• [PaddleV3] 修复部分依赖 paddle.fluid 的 tensorflow 模型 part 1 #1069
• [PaddleV3] 修复部分依赖 paddle.fluid 的 tensorflow 模型 part 2 #1070
• [PaddleV3] 修复依赖 paddle.fluid 的 caffe 模型 #1071

关联 PR：#1060

🟠 存在问题

🔵 正在修复

✅ 已完成