支持模型¶
目前,Paddle Lite 已严格验证 52 个模型的精度和性能。对视觉类模型做到了充分的支持,覆盖分类、检测和定位,也包含了特色的 OCR 模型的支持。对 NLP 模型也做到了广泛支持,包含翻译、语义表达等等。
除了已严格验证的模型,Paddle Lite 对其他 CV 和 NLP 模型也可以做到大概率支持。
类别 | 类别细分 | 模型 | 支持平台 |
---|---|---|---|
CV | 分类 | MobileNetV1 | ARM, X86, GPU(OPENCL,METAL), HuaweiKirinNPU, RockchipNPU, MediatekAPU, BaiduXPU, HuaweiAscendNPU |
CV | 分类 | MobileNetV2 | ARM, X86, GPU(OPENCL,METAL), HuaweiKirinNPU, BaiduXPU, HuaweiAscendNPU |
CV | 分类 | MobileNetV3_large | ARM, X86, GPU(OPENCL,METAL), HuaweiAscendNPU |
CV | 分类 | MobileNetV3_small | ARM, X86, GPU(OPENCL,METAL), HuaweiAscendNPU |
CV | 分类 | DPN68 | ARM, X86 |
CV | 分类 | AlexNet | ARM, X86, HuaweiAscendNPU |
CV | 分类 | DarkNet53 | ARM, X86 |
CV | 分类 | DenseNet121 | ARM, X86, HuaweiAscendNPU |
CV | 分类 | EfficientNetB0 | ARM, X86, GPU(OPENCL), BaiduXPU, HuaweiAscendNPU |
CV | 分类 | GhostNet_x1_3 | ARM |
CV | 分类 | HRNet_W18_C | ARM, X86 |
CV | 分类 | RegNetX_4GF | ARM, X86 |
CV | 分类 | Xception41 | ARM, X86 |
CV | 分类 | ResNet18 | ARM, X86, GPU(OPENCL,METAL), HuaweiKirinNPU, RockchipNPU, BaiduXPU, HuaweiAscendNPU |
CV | 分类 | ResNet50 | ARM, X86, GPU(OPENCL,METAL), HuaweiKirinNPU, RockchipNPU, BaiduXPU, HuaweiAscendNPU |
CV | 分类 | ResNet101 | ARM, X86, HuaweiKirinNPU, RockchipNPU, BaiduXPU, HuaweiAscendNPU |
CV | 分类 | ResNeXt50 | ARM, X86, HuaweiAscendNPU |
CV | 分类 | MnasNet | ARM, HuaweiKirinNPU, HuaweiAscendNPU |
CV | 分类 | SqueezeNet | ARM, HuaweiKirinNPU, BaiduXPU, HuaweiAscendNPU |
CV | 分类 | ShuffleNet | ARM |
CV | 分类 | ShufflenetV2 | ARM, BaiduXPU, HuaweiAscendNPU |
CV | 分类 | InceptionV3 | ARM, X86, HuaweiAscendNPU |
CV | 分类 | InceptionV4 | ARM, X86, HuaweiKirinNPU, BaiduXPU, HuaweiAscendNPU |
CV | 分类 | VGG16 | ARM, X86, GPU(OPENCL), BaiduXPU, HuaweiAscendNPU |
CV | 分类 | VGG19 | ARM, X86, GPU(OPENCL,METAL), BaiduXPU, HuaweiAscendNPU |
CV | 分类 | GoogleNet | ARM, X86, BaiduXPU, HuaweiAscendNPU |
CV | 检测 | SSD-MobileNetV1 | ARM, HuaweiKirinNPU, HuaweiAscendNPU |
CV | 检测 | SSD-MobileNetV3-large | ARM, X86, GPU(OPENCL,METAL) |
CV | 检测 | SSD-VGG16 | ARM, X86, HuaweiAscendNPU* |
CV | 检测 | YOLOv3-DarkNet53 | ARM, X86, HuaweiAscendNPU* |
CV | 检测 | YOLOv3-MobileNetV1 | ARM, X86, HuaweiAscendNPU* |
CV | 检测 | YOLOv3-MobileNetV3 | ARM, X86, HuaweiAscendNPU* |
CV | 检测 | yolov3_r50vd_dcn | ARM, HuaweiKirinNPU, HuaweiAscendNPU |
CV | 检测 | YOLOv4 | ARM, X86, HuaweiAscendNPU* |
CV | 检测 | Faster RCNN | ARM |
CV | 检测 | Mask RCNN* | ARM |
CV | 检测 | ppyolo_2x | ARM |
CV | 检测 | solov2_r50_fpn_1x | ARM |
CV | OCR | ch_ppocr_mobile_v2.0_cls_infer | ARM, X86, GPU(OPENCL) |
CV | OCR | ch_ppocr_mobile_v2.0_det_infer | ARM, X86, GPU(OPENCL), HuaweiAscendNPU |
CV | OCR | ch_ppocr_mobile_v2.0_rec_infer | ARM, X86, GPU(OPENCL), HuaweiAscendNPU* |
CV | OCR | ch_ppocr_server_v2.0_rec_infer | ARM |
CV | OCR | CRNN | ARM |
CV | OCR | DB | ARM, GPU(OPENCL) |
CV | OCR | OCR-Attention | ARM |
CV | REG | inference_dnn | ARM, GPU(OPENCL) |
CV | 分割 | Deeplabv3 | ARM, GPU(OPENCL) |
CV | 分割 | UNet | ARM, GPU(OPENCL) |
CV | 分割 | bisenet | ARM, GPU(OPENCL) |
CV | 分割 | fastscnn | ARM, GPU(OPENCL) |
CV | 分割 | bisenet_v2 | ARM, GPU(OPENCL) |
CV | 关键点 | HigherHRNet | ARM,X86,HuaweiAscendNPU |
CV | 关键点 | HRNet | ARM,X86,HuaweiAscendNPU |
CV | 人脸 | FaceDetection | ARM |
CV | 人脸 | FaceBoxes* | ARM, GPU(OPENCL) |
CV | 人脸 | BlazeFace* | ARM |
CV | 人脸 | MTCNN | ARM, GPU(OPENCL) |
NLP | 机器翻译 | Transformer* | ARM,HuaweiKirinNPU* |
NLP | 机器翻译 | BERT | BaiduXPU |
NLP | 语义表示 | ERNIE | BaiduXPU |
NLP | 语义理解 | ERNIE-TINY | ARM,BaiduXPU,HuaweiAscendNPU* |
GAN | 风格转换 | CycleGAN* | HuaweiKirinNPU |
GAN | 超分辨率 | ESRGAN | ARM,X86,HuaweiAscendNPU |
注意:
模型列表中 * 代表该模型链接来自 PaddlePaddle/models ,否则为推理模型的下载链接
支持平台列表中 HuaweiKirinNPU* 代表 ARM + HuaweiKirinNPU 异构计算,否则为 HuaweiKirinNPU 计算
支持平台列表中 HuaweiAscendNPU* 代表 X86 或 ARM+HuaweiAscendNPU 异构计算,否则为 HuaweiAscendNPU 计算
寻找更多的可支持的模型还可以转至 PaddlePaddle/models , PaddleHub , PaddleOCR , PaddleDetection