Skip to content

Commit 180f64f

Browse files
shentanyueshentanyue
authored
[Cherry-pick][Doc] Update op and model list for HuaweiAscendNPU (#8947)
1 parent e945b2d commit 180f64f

File tree

2 files changed

+32
-26
lines changed

2 files changed

+32
-26
lines changed

‎docs/quick_start/support_hardware.md

Lines changed: 11 additions & 5 deletions
Original file line numberDiff line numberDiff line change
@@ -104,11 +104,17 @@ Paddle Lite 支持华为达芬奇架构麒麟 NPU，支持列表如下：
104104

105105
## 华为 (Huawei) 昇腾 NPU
106106

107-
Paddle Lite 已支持华为昇腾 NPU（Ascend310）在 X86 和 ARM 服务器上进行预测部署
108-
- 支持设备：
109-
* Ascend 310：Atlas 300I 推理卡（型号：3000/3010)
110-
* Atlas 200 DK 开发者套件
111-
* Atlas 800 推理服务器（型号：3000/3010）
107+
Paddle Lite 已支持华为昇腾 NPU（Ascend310、Ascend710和Ascend910）在 X86 和 ARM 服务器上进行预测部署
108+
- 支持芯片
109+
* Ascend 310 （CANN Version ≥ 3.3.0）
110+
* Ascend 710 （CANN Version ≥ 5.0.2.alpha005)
111+
* Ascend 910 （CANN Version ≥ 5.0.2.alpha005)
112+
- 已验证的支持设备
113+
* Atlas 300I 推理卡（型号：3000/3010)（CANN Version ≥ 3.3.0）
114+
* Atlas 200 DK 开发者套件（CANN Version ≥ 3.3.0）
115+
* Atlas 800 推理服务器（型号：3000/3010）（CANN Version ≥ 3.3.0）
116+
* Atlas 300I Pro（CANN Version ≥ 5.0.2.alpha005)
117+
* Atlas 300T 训练卡（CANN Version ≥ 5.0.2.alpha005)
112118

113119
### 华为 (Huawei) 昇腾 NPU 环境准备及示例程序
114120

‎docs/quick_start/support_model_list.md

Lines changed: 21 additions & 21 deletions
Original file line numberDiff line numberDiff line change
@@ -10,13 +10,13 @@
1010
| CV | 分类 | [MobileNetV2](https://paddlelite-demo.bj.bcebos.com/models/mobilenet_v2_fp32_224_fluid.tar.gz) | ARM, X86, GPU(OPENCL,METAL), HuaweiKirinNPU, KunlunxinXPU, HuaweiAscendNPU |
1111
| CV | 分类 | [MobileNetV3_large](https://paddle-inference-dist.bj.bcebos.com/AI-Rank/mobile/MobileNetV3_large_x1_0.tar.gz) | ARM, X86, GPU(OPENCL,METAL), HuaweiAscendNPU |
1212
| CV | 分类 | [MobileNetV3_small](https://paddle-inference-dist.bj.bcebos.com/AI-Rank/mobile/MobileNetV3_small_x1_0.tar.gz) | ARM, X86, GPU(OPENCL,METAL), HuaweiAscendNPU |
13-
| CV | 分类 | DPN68 | ARM, X86 |
13+
| CV | 分类 | DPN68 | ARM, X86, HuaweiAscendNPU |
1414
| CV | 分类 | AlexNet | ARM, X86, HuaweiAscendNPU |
15-
| CV | 分类 | DarkNet53 | ARM, X86 |
15+
| CV | 分类 | DarkNet53 | ARM, X86, HuaweiAscendNPU |
1616
| CV | 分类 | DenseNet121 | ARM, X86, HuaweiAscendNPU |
1717
| CV | 分类 | [EfficientNetB0](https://paddlelite-demo.bj.bcebos.com/models/EfficientNetB0.tar.gz) | ARM, X86, GPU(OPENCL), KunlunxinXPU, HuaweiAscendNPU |
18-
| CV | 分类 | GhostNet_x1_3 | ARM |
19-
| CV | 分类 | HRNet_W18_C | ARM, X86 |
18+
| CV | 分类 | GhostNet_x1_3 | ARM,HuaweiAscendNPU |
19+
| CV | 分类 | HRNet_W18_C | ARM, X86,HuaweiAscendNPU |
2020
| CV | 分类 | RegNetX_4GF | ARM, X86 |
2121
| CV | 分类 | Xception41 | ARM, X86 |
2222
| CV | 分类 | [ResNet18](https://paddlelite-demo.bj.bcebos.com/models/resnet18_fp32_224_fluid.tar.gz) | ARM, X86, GPU(OPENCL,METAL), HuaweiKirinNPU, RockchipNPU, KunlunxinXPU, HuaweiAscendNPU, VerisiliconTIMVX, AndroidNNAPI |
@@ -25,15 +25,15 @@
2525
| CV | 分类 | [ResNeXt50](https://paddlelite-demo.bj.bcebos.com/NNAdapter/models/PaddleClas/ResNeXt50_32x4d.tgz) | ARM, X86, HuaweiAscendNPU |
2626
| CV | 分类 | [MnasNet](https://paddlelite-demo.bj.bcebos.com/models/mnasnet_fp32_224_fluid.tar.gz)| ARM, HuaweiKirinNPU, HuaweiAscendNPU |
2727
| CV | 分类 | [SqueezeNet](https://paddlelite-demo.bj.bcebos.com/models/squeezenet_fp32_224_fluid.tar.gz) | ARM, HuaweiKirinNPU, KunlunxinXPU, HuaweiAscendNPU |
28-
| CV | 分类 | ShuffleNet | ARM |
28+
| CV | 分类 | ShuffleNet | ARM,HuaweiAscendNPU |
2929
| CV | 分类 | [ShufflenetV2](https://paddlelite-demo.bj.bcebos.com/models/shufflenetv2.tar.gz) | ARM, KunlunxinXPU, HuaweiAscendNPU |
3030
| CV | 分类 | [InceptionV3](https://paddlelite-demo.bj.bcebos.com/NNAdapter/models/PaddleClas/InceptionV3.tgz) | ARM, X86, HuaweiAscendNPU |
3131
| CV | 分类 | [InceptionV4](https://paddlelite-demo.bj.bcebos.com/models/inceptionv4.tar.gz) | ARM, X86, HuaweiKirinNPU, KunlunxinXPU, HuaweiAscendNPU |
3232
| CV | 分类 | VGG16 | ARM, X86, GPU(OPENCL), KunlunxinXPU, HuaweiAscendNPU |
3333
| CV | 分类 | VGG19 | ARM, X86, GPU(OPENCL,METAL), KunlunxinXPU, HuaweiAscendNPU|
3434
| CV | 分类 | GoogleNet | ARM, X86, KunlunxinXPU, HuaweiAscendNPU |
3535
| CV | 检测 | [SSD-MobileNetV1](https://paddlelite-demo.bj.bcebos.com/models/ssd_mobilenet_v1_pascalvoc_fp32_300_fluid.tar.gz) | ARM, HuaweiKirinNPU*, HuaweiAscendNPU*, VerisiliconTIMVX, AndroidNNAPI |
36-
| CV | 检测 | [SSD-MobileNetV3-large](https://paddle-inference-dist.bj.bcebos.com/AI-Rank/mobile/ssdlite_mobilenet_v3_large.tar.gz) | ARM, X86, GPU(OPENCL,METAL) |
36+
| CV | 检测 | [SSD-MobileNetV3-large](https://paddle-inference-dist.bj.bcebos.com/AI-Rank/mobile/ssdlite_mobilenet_v3_large.tar.gz) | ARM, X86, GPU(OPENCL,METAL),HuaweiAscendNPU* |
3737
| CV | 检测 | [SSD-VGG16](https://paddlelite-demo.bj.bcebos.com/NNAdapter/models/PaddleDetection/ssd_vgg16_300_240e_voc.tgz) | ARM, X86, HuaweiAscendNPU* |
3838
| CV | 检测 | [YOLOv3-DarkNet53](https://paddlelite-demo.bj.bcebos.com/NNAdapter/models/PaddleDetection/yolov3_darknet53_270e_coco.tgz) | ARM, X86, HuaweiAscendNPU* , NVIDIATensorRT|
3939
| CV | 检测 | [YOLOv3-MobileNetV1](https://paddlelite-demo.bj.bcebos.com/NNAdapter/models/PaddleDetection/yolov3_mobilenet_v1_270e_coco.tgz) | ARM, X86, HuaweiAscendNPU* |
@@ -42,31 +42,31 @@
4242
| CV | 检测 | [YOLOv4](https://paddlelite-demo.bj.bcebos.com/NNAdapter/models/PaddleDetection/yolov4_cspdarknet.tgz) | ARM, X86, HuaweiAscendNPU* |
4343
| CV | 检测 | Faster RCNN | ARM |
4444
| CV | 检测 | [Mask RCNN*](https://github.com/PaddlePaddle/PaddleDetection/blob/release/0.4/docs/MODEL_ZOO_cn.md) | ARM |
45-
| CV | 检测 | ppyolo_2x | ARM |
45+
| CV | 检测 | ppyolo_2x | ARM,HuaweiAscendNPU* |
4646
| CV | 检测 | solov2_r50_fpn_1x | ARM |
47-
| CV | OCR | ch_ppocr_mobile_v2.0_cls_infer | ARM, X86, GPU(OPENCL) |
47+
| CV | OCR | ch_ppocr_mobile_v2.0_cls_infer | ARM, X86, GPU(OPENCL),HuaweiAscendNPU |
4848
| CV | OCR | [ch_ppocr_mobile_v2.0_det_infer](https://paddlelite-demo.bj.bcebos.com/NNAdapter/models/PaddleOCR/ch_ppocr_mobile_v2.0_det_infer.tgz) | ARM, X86, GPU(OPENCL), HuaweiAscendNPU |
4949
| CV | OCR | [ch_ppocr_mobile_v2.0_rec_infer](https://paddlelite-demo.bj.bcebos.com/NNAdapter/models/PaddleOCR/ch_ppocr_mobile_v2.0_rec_infer.tgz) | ARM, X86, GPU(OPENCL), HuaweiAscendNPU* |
50-
| CV | OCR | ch_ppocr_server_v2.0_rec_infer | ARM |
51-
| CV | OCR | CRNN | ARM |
52-
| CV | OCR | DB | ARM, GPU(OPENCL) |
50+
| CV | OCR | ch_ppocr_server_v2.0_rec_infer | ARM,HuaweiAscendNPU* |
51+
| CV | OCR | CRNN | ARM,HuaweiAscendNPU |
52+
| CV | OCR | DB | ARM, GPU(OPENCL),HuaweiAscendNPU |
5353
| CV | OCR | [OCR-Attention](https://paddle-inference-dist.bj.bcebos.com/ocr_attention.tar.gz) | ARM |
5454
| CV | REG | inference_dnn | ARM, GPU(OPENCL) |
55-
| CV | 分割 | [Deeplabv3](https://paddlelite-demo.bj.bcebos.com/models/deeplab_mobilenet_fp32_fluid.tar.gz) | ARM, GPU(OPENCL) |
56-
| CV | 分割 | [UNet](https://paddlelite-demo.bj.bcebos.com/models/Unet.zip) | ARM, GPU(OPENCL) |
57-
| CV | 分割 | bisenet | ARM, GPU(OPENCL) |
55+
| CV | 分割 | [Deeplabv3](https://paddlelite-demo.bj.bcebos.com/models/deeplab_mobilenet_fp32_fluid.tar.gz) | ARM, GPU(OPENCL), HuaweiAscendNPU |
56+
| CV | 分割 | [UNet](https://paddlelite-demo.bj.bcebos.com/models/Unet.zip) | ARM, GPU(OPENCL), HuaweiAscendNPU |
57+
| CV | 分割 | bisenet | ARM, GPU(OPENCL),HuaweiAscendNPU |
5858
| CV | 分割 | fastscnn | ARM, GPU(OPENCL) |
59-
| CV | 分割 | bisenet_v2 | ARM, GPU(OPENCL) |
59+
| CV | 分割 | bisenet_v2 | ARM, GPU(OPENCL),HuaweiAscendNPU |
6060
| CV | 关键点 | [HigherHRNet](https://paddlelite-demo.bj.bcebos.com/NNAdapter/models/PaddleDetection/higherhrnet_hrnet_w32_640.tgz) | ARM,X86,HuaweiAscendNPU |
6161
| CV | 关键点 | [HRNet](https://paddlelite-demo.bj.bcebos.com/NNAdapter/models/PaddleDetection/hrnet_w32_384x288.tgz) | ARM,X86,HuaweiAscendNPU |
6262
| CV | 人脸 | [FaceDetection](https://paddlelite-demo.bj.bcebos.com/models/facedetection_fp32_240_430_fluid.tar.gz) | ARM |
63-
| CV | 人脸 | [FaceBoxes*](https://github.com/PaddlePaddle/PaddleDetection/blob/release/0.4/docs/featured_model/FACE_DETECTION.md#FaceBoxes) | ARM, GPU(OPENCL) |
64-
| CV | 人脸 | [BlazeFace*](https://github.com/PaddlePaddle/PaddleDetection/blob/release/0.4/docs/featured_model/FACE_DETECTION.md#BlazeFace) | ARM |
63+
| CV | 人脸 | [FaceBoxes*](https://github.com/PaddlePaddle/PaddleDetection/blob/release/0.4/docs/featured_model/FACE_DETECTION.md#FaceBoxes) | ARM, GPU(OPENCL), HuaweiAscendNPU |
64+
| CV | 人脸 | [BlazeFace*](https://github.com/PaddlePaddle/PaddleDetection/blob/release/0.4/docs/featured_model/FACE_DETECTION.md#BlazeFace) | ARM,HuaweiAscendNPU |
6565
| CV | 人脸 | [MTCNN](https://paddlelite-demo.bj.bcebos.com/models/mtcnn.zip) | ARM, GPU(OPENCL) |
66-
| NLP | 机器翻译 | [Transformer*](https://github.com/PaddlePaddle/models/tree/release/1.8/PaddleNLP/machine_translation/transformer) | ARM,HuaweiKirinNPU* |
67-
| NLP | 机器翻译 | [BERT](https://paddle-inference-dist.bj.bcebos.com/PaddleLite/models_and_data_for_unittests/bert.tar.gz) | KunlunxinXPU |
68-
| NLP | 语义表示 | [ERNIE](https://paddle-inference-dist.bj.bcebos.com/PaddleLite/models_and_data_for_unittests/ernie.tar.gz) | KunlunxinXPU |
69-
| NLP | 语义理解 | [ERNIE-TINY](https://paddlelite-demo.bj.bcebos.com/NNAdapter/models/PaddleNLP/ernie_tiny.tgz) | ARM,KunlunxinXPU,HuaweiAscendNPU* |
66+
| NLP | 机器翻译 | [Transformer*](https://github.com/PaddlePaddle/models/tree/release/1.8/PaddleNLP/machine_translation/transformer) | ARM,HuaweiKirinNPU*,HuaweiAscendNPU* |
67+
| NLP | 机器翻译 | [BERT](https://paddle-inference-dist.bj.bcebos.com/PaddleLite/models_and_data_for_unittests/bert.tar.gz) | KunlunxinXPU,HuaweiAscendNPU |
68+
| NLP | 语义表示 | [ERNIE](https://paddle-inference-dist.bj.bcebos.com/PaddleLite/models_and_data_for_unittests/ernie.tar.gz) | KunlunxinXPU,HuaweiAscendNPU |
69+
| NLP | 语义理解 | [ERNIE-TINY](https://paddlelite-demo.bj.bcebos.com/NNAdapter/models/PaddleNLP/ernie_tiny.tgz) | ARM,KunlunxinXPU,HuaweiAscendNPU |
7070
| GAN | 风格转换 | [CycleGAN*](https://github.com/PaddlePaddle/models/tree/release/1.7/PaddleCV/gan/cycle_gan) | HuaweiKirinNPU |
7171
| GAN | 超分辨率 | [ESRGAN](https://paddlelite-demo.bj.bcebos.com/NNAdapter/models/PaddleGAN/esrgan_psnr_x4_div2k.tgz) |ARM,X86,HuaweiAscendNPU|
7272

0 commit comments

Comments
 (0)