From da5e38218061bba151adcc37ede2ab513831320f Mon Sep 17 00:00:00 2001 From: lifangtian Date: Fri, 30 May 2025 08:37:30 +0000 Subject: [PATCH 01/35] =?UTF-8?q?=E6=9B=B4=E6=96=B0README=E4=B8=AD?= =?UTF-8?q?=E5=87=A0=E5=A4=84=E7=AC=94=E8=AF=AF?= MIME-Version: 1.0 Content-Type: text/plain; charset=UTF-8 Content-Transfer-Encoding: 8bit --- ACL_PyTorch/contrib/audio/wav2lip_ID100400/README.md | 12 ++++++------ 1 file changed, 6 insertions(+), 6 deletions(-) diff --git a/ACL_PyTorch/contrib/audio/wav2lip_ID100400/README.md b/ACL_PyTorch/contrib/audio/wav2lip_ID100400/README.md index 3211007e1c..84a7479100 100644 --- a/ACL_PyTorch/contrib/audio/wav2lip_ID100400/README.md +++ b/ACL_PyTorch/contrib/audio/wav2lip_ID100400/README.md @@ -142,17 +142,17 @@ 1. 获取权重文件。 - [wav2lip模型预训练pth权重文件](https://iiitaphyd-my.sharepoint.com/:u:/g/personal/radrabha_m_research_iiit_ac_in/Eb3LEzbfuKlJiR600lQWRxgBIY27JZg80f7V9jtMfbNDaQ?e=TBFBVW)，将获取的权重文件放在当前工作路径下。 + [wav2lip模型预训练pth权重文件](https://huggingface.co/numz/wav2lip_studio/blob/main/Wav2lip/wav2lip.pth)，将获取的权重文件放在当前工作路径下。 2. 导出onnx文件。 - 1. 使用pth2onnx.py脚本。 + 1. 使用wav2lip_pth2onnx.py脚本。 - 运行pth2onnx.py脚本。 + 运行wav2lip_pth2onnx.py脚本。 ``` batch_size=72 - python3 pth2onnx.py --checkpoint_path ./wav2lip.pth --onnx_dir ./ -batch_size ${batch_size} + python3 wav2lip_pth2onnx.py --checkpoint_path ./wav2lip.pth --onnx_dir ./ --batch_size ${batch_size} ``` 获得wav2lip.onnx文件。 @@ -186,7 +186,7 @@ 3. 执行ATC命令。 ``` - atc --model=./wav2lip_bs72.onnx --framework=5 --output=./wav2lip_bs72--input_format=ND --input_shape="input1:72,1,80,16;input2:72,6,96,96" --log=debug --soc_version=Ascend${chip_name} + atc --model=./wav2lip_bs72.onnx --framework=5 --output=./wav2lip_bs72 --input_format=ND --input_shape="input1:72,1,80,16;input2:72,6,96,96" --log=debug --soc_version=Ascend${chip_name} ``` - 参数说明： @@ -231,7 +231,7 @@ 调用wav2lip_postprocess.py脚本合成完整视频的文件。 ``` - python3 wav2lip_postprocess.py --om_pred mels_0.bin --frames ./inputs/frames.bin --coords ./inputs/coords.bin --outfile ./results/result_voice.mp4 --audio ./testdata/audio.mp3 + python3 wav2lip_postprocess.py --om_pred ${om_output_path} --frames ./inputs/frames.bin --coords ./inputs/coords.bin --outfile ./results/result_voice.mp4 --audio ./testdata/audio.mp3 ``` - 参数说明： -- Gitee From 67088123532f779c55c0f796f7b662ac33015827 Mon Sep 17 00:00:00 2001 From: lifangtian Date: Fri, 30 May 2025 08:39:31 +0000 Subject: [PATCH 02/35] =?UTF-8?q?=E4=BF=AE=E6=AD=A3mmdet.patch=E7=A7=BB?= =?UTF-8?q?=E5=8A=A8=E7=9A=84=E7=9B=AE=E7=9A=84=E8=B7=AF=E5=BE=84?= MIME-Version: 1.0 Content-Type: text/plain; charset=UTF-8 Content-Transfer-Encoding: 8bit --- ACL_PyTorch/built-in/cv/ViTDet_for_Pytorch/README.md | 6 +++--- 1 file changed, 3 insertions(+), 3 deletions(-) diff --git a/ACL_PyTorch/built-in/cv/ViTDet_for_Pytorch/README.md b/ACL_PyTorch/built-in/cv/ViTDet_for_Pytorch/README.md index a96e7e5c49..5a3c36024f 100755 --- a/ACL_PyTorch/built-in/cv/ViTDet_for_Pytorch/README.md +++ b/ACL_PyTorch/built-in/cv/ViTDet_for_Pytorch/README.md @@ -83,9 +83,9 @@ ViTDet使用公共数据集COCO进行推理 3. 转移文件位置 ``` - mv mmengine.patch mmengine/mmengine/ - mv mmdet.patch mmdetection/mmdet/ - mv infer.py mmdetection/ + cp mmengine.patch mmengine/mmengine/ + cp mmdet.patch mmdetection/ + cp infer.py mmdetection/ ``` 4. 更换当前路径并打补丁，修改完mmseg源码后进行安装 -- Gitee From 4b7d7d9e15795dc069a1fb3ddb27a2e410211d3d Mon Sep 17 00:00:00 2001 From: lifangtian Date: Fri, 30 May 2025 08:42:54 +0000 Subject: [PATCH 03/35] =?UTF-8?q?=E5=88=A0=E6=8E=89=E6=9C=AC=E4=B8=8D?= =?UTF-8?q?=E5=BA=94=E8=AF=A5=E5=AD=98=E5=9C=A8=E7=9A=84mmdet=E4=BE=9D?= =?UTF-8?q?=E8=B5=96?= MIME-Version: 1.0 Content-Type: text/plain; charset=UTF-8 Content-Transfer-Encoding: 8bit Signed-off-by: lifangtian --- .../cv/InternImage_detection_for_Pytorch/requirement.txt | 1 - 1 file changed, 1 deletion(-) diff --git a/ACL_PyTorch/built-in/cv/InternImage_detection_for_Pytorch/requirement.txt b/ACL_PyTorch/built-in/cv/InternImage_detection_for_Pytorch/requirement.txt index 8ddcce8ef9..bd2ca64041 100755 --- a/ACL_PyTorch/built-in/cv/InternImage_detection_for_Pytorch/requirement.txt +++ b/ACL_PyTorch/built-in/cv/InternImage_detection_for_Pytorch/requirement.txt @@ -4,7 +4,6 @@ tqdm torchvision==0.13.0 mmcv==2.1.0 timm==0.6.11 -mmdet==3.0.0 mmengine==0.10.6 opencv-python termcolor -- Gitee From 7635a0871dce92bac6bec60644f07a3eeabd108f Mon Sep 17 00:00:00 2001 From: lifangtian Date: Fri, 30 May 2025 08:43:52 +0000 Subject: [PATCH 04/35] =?UTF-8?q?=E4=BF=AE=E6=AD=A3=E4=BA=86mmdet=E6=BA=90?= =?UTF-8?q?=E7=A0=81=E4=BB=93=E7=9A=84git=E9=93=BE=E6=8E=A5?= MIME-Version: 1.0 Content-Type: text/plain; charset=UTF-8 Content-Transfer-Encoding: 8bit Signed-off-by: lifangtian --- .../built-in/cv/InternImage_detection_for_Pytorch/README.md | 2 +- 1 file changed, 1 insertion(+), 1 deletion(-) diff --git a/ACL_PyTorch/built-in/cv/InternImage_detection_for_Pytorch/README.md b/ACL_PyTorch/built-in/cv/InternImage_detection_for_Pytorch/README.md index 0247c47408..c88ab9d589 100755 --- a/ACL_PyTorch/built-in/cv/InternImage_detection_for_Pytorch/README.md +++ b/ACL_PyTorch/built-in/cv/InternImage_detection_for_Pytorch/README.md @@ -65,7 +65,7 @@ InternImage使用公共数据集COCO进行推理 2. 获取模型仓**InternImage**源码和依赖仓**mmdet**源码 ``` - git clone https://github.com/open-mmlab/mmdet.git + git clone https://github.com/open-mmlab/mmdetection.git git clone https://github.com/OpenGVLab/InternImage.git cd mmdetection git reset --hard cfd5d3a985b0249de009b67d04f37263e11cdf3d -- Gitee From 088febb13bc2c6ca13cf26fdbfba8c12a9ffb025 Mon Sep 17 00:00:00 2001 From: lifangtian Date: Fri, 30 May 2025 09:32:03 +0000 Subject: [PATCH 05/35] =?UTF-8?q?=E5=9C=A8README=E4=B8=AD=E5=A2=9E?= =?UTF-8?q?=E5=8A=A0=E8=8E=B7=E5=8F=96=E6=9C=AC=E4=BB=93=E6=BA=90=E7=A0=81?= =?UTF-8?q?=E6=AD=A5=E9=AA=A4?= MIME-Version: 1.0 Content-Type: text/plain; charset=UTF-8 Content-Transfer-Encoding: 8bit Signed-off-by: lifangtian --- ACL_PyTorch/built-in/cv/GLIP/README.md | 7 +++++++ 1 file changed, 7 insertions(+) diff --git a/ACL_PyTorch/built-in/cv/GLIP/README.md b/ACL_PyTorch/built-in/cv/GLIP/README.md index 9ac213ab4f..97fd8b2733 100644 --- a/ACL_PyTorch/built-in/cv/GLIP/README.md +++ b/ACL_PyTorch/built-in/cv/GLIP/README.md @@ -71,6 +71,13 @@ CLIP和ALIGN在大规模图像-文本对上进行跨模态对比学习，可以 ## 获取源码 +1. 获取本仓源码 + + ``` + git clone https://gitee.com/ascend/ModelZoo-PyTorch.git + cd ModelZoo-PyTorch/ACL_PyTorch/built-in/cv/GLIP + ``` + 1. 获取源码。 ``` -- Gitee From a5cd2aa8ffa6eea9c7c8eefb29895419a8fc333f Mon Sep 17 00:00:00 2001 From: lifangtian Date: Wed, 4 Jun 2025 02:10:37 +0000 Subject: [PATCH 06/35] =?UTF-8?q?=E4=BF=AE=E6=94=B9ACL=5FPyTorch/contrib/c?= =?UTF-8?q?v/image=5Fretrieval/BLIP/readme.md=EF=BC=9A=E5=88=A0=E9=99=A4?= =?UTF-8?q?=E4=BA=86=E6=9D=83=E9=87=8D=E6=96=87=E4=BB=B6=E9=93=BE=E6=8E=A5?= =?UTF-8?q?=E6=9C=80=E5=90=8E=E7=9A=84=E5=8F=A5=E5=8F=B7=EF=BC=9B=E8=A1=A5?= =?UTF-8?q?=E4=B8=8AONNX=E8=BD=ACOM=E7=9A=84=E7=AC=AC=E4=B8=89=E4=B8=AA?= =?UTF-8?q?=E5=91=BD=E4=BB=A4=E7=9A=84=E6=8D=A2=E8=A1=8C=E7=AC=A6?= MIME-Version: 1.0 Content-Type: text/plain; charset=UTF-8 Content-Transfer-Encoding: 8bit Signed-off-by: lifangtian --- ACL_PyTorch/contrib/cv/image_retrieval/BLIP/readme.md | 4 ++-- 1 file changed, 2 insertions(+), 2 deletions(-) diff --git a/ACL_PyTorch/contrib/cv/image_retrieval/BLIP/readme.md b/ACL_PyTorch/contrib/cv/image_retrieval/BLIP/readme.md index b002bfde7f..51d07f4420 100644 --- a/ACL_PyTorch/contrib/cv/image_retrieval/BLIP/readme.md +++ b/ACL_PyTorch/contrib/cv/image_retrieval/BLIP/readme.md @@ -173,7 +173,7 @@ BLIP模型为一种新的Vision-Language Pre-training框架，它可以灵活地 1. 获取权重文件。 - 训练权重链接为：https://storage.googleapis.com/sfr-vision-language-research/BLIP/models/model_base_retrieval_coco.pth。 + 训练权重链接为：https://storage.googleapis.com/sfr-vision-language-research/BLIP/models/model_base_retrieval_coco.pth 在`BLIP`工作目录下可通过以下命令获取训练权重并转为推理模型。 @@ -248,7 +248,7 @@ BLIP模型为一种新的Vision-Language Pre-training框架，它可以灵活地 --input_format=ND \ --input_shape="text_ids:${batchsize},35;text_atten_mask:${batchsize},35" \ --log=error \ - --soc_version=Ascend${chip_name} + --soc_version=Ascend${chip_name} \ --op_precision_mode=op_precision.ini ``` -- Gitee From 96bbee8ec96bf7964e7867e1f9d213fcd4fd85b9 Mon Sep 17 00:00:00 2001 From: lifangtian Date: Wed, 4 Jun 2025 03:12:23 +0000 Subject: [PATCH 07/35] =?UTF-8?q?=E4=BF=AE=E6=94=B9ACL=5FPyTorch/built-in/?= =?UTF-8?q?audio/SenseVoice/README=5Fonnx.md=EF=BC=9AONNX=E8=BD=ACOM?= =?UTF-8?q?=E5=91=BD=E4=BB=A4=E4=B8=AD=E4=BF=AE=E6=94=B9--soc=5Fversion?= =?UTF-8?q?=E7=9A=84=E5=80=BC=E4=B8=BA=E6=A0=87=E5=87=86=E6=A0=BC=E5=BC=8F?= MIME-Version: 1.0 Content-Type: text/plain; charset=UTF-8 Content-Transfer-Encoding: 8bit Signed-off-by: lifangtian --- ACL_PyTorch/built-in/audio/SenseVoice/README_onnx.md | 2 +- 1 file changed, 1 insertion(+), 1 deletion(-) diff --git a/ACL_PyTorch/built-in/audio/SenseVoice/README_onnx.md b/ACL_PyTorch/built-in/audio/SenseVoice/README_onnx.md index 7d73bdc267..35c8d0be1f 100755 --- a/ACL_PyTorch/built-in/audio/SenseVoice/README_onnx.md +++ b/ACL_PyTorch/built-in/audio/SenseVoice/README_onnx.md @@ -94,7 +94,7 @@ source /usr/local/Ascend/ascend-toolkit/set_env.sh ``` -atc --framework=5 --soc_version=${soc_version} --model ./SenseVoiceSmall/model_md.onnx --output SenseVoice --input_shape="speech:1,-1,560;speech_lengths:1;language:1;textnorm:1" +atc --framework=5 --soc_version=Ascend${soc_version} --model ./SenseVoiceSmall/model_md.onnx --output SenseVoice --input_shape="speech:1,-1,560;speech_lengths:1;language:1;textnorm:1" ``` 在当前目录下生成动态模型SenseVoice_{arch}.om -- Gitee From 35c8102dd54978027f2640952f2f3bf43c46e192 Mon Sep 17 00:00:00 2001 From: lifangtian Date: Wed, 4 Jun 2025 05:05:25 +0000 Subject: [PATCH 08/35] =?UTF-8?q?=E4=BF=AE=E6=94=B9ACL=5FPyTorch/built-in/?= =?UTF-8?q?foundation=5Fmodels/Chinese=5FCLIP/README.md=EF=BC=9A=E4=BF=AE?= =?UTF-8?q?=E6=94=B9=E5=9C=A8modelzoo=E4=BB=93=E9=87=8C=E7=9A=84cd?= =?UTF-8?q?=E8=B7=AF=E5=BE=84?= MIME-Version: 1.0 Content-Type: text/plain; charset=UTF-8 Content-Transfer-Encoding: 8bit Signed-off-by: lifangtian --- ACL_PyTorch/built-in/foundation_models/Chinese_CLIP/README.md | 2 +- 1 file changed, 1 insertion(+), 1 deletion(-) diff --git a/ACL_PyTorch/built-in/foundation_models/Chinese_CLIP/README.md b/ACL_PyTorch/built-in/foundation_models/Chinese_CLIP/README.md index 6de0c186b4..d1fb1595d5 100644 --- a/ACL_PyTorch/built-in/foundation_models/Chinese_CLIP/README.md +++ b/ACL_PyTorch/built-in/foundation_models/Chinese_CLIP/README.md @@ -73,7 +73,7 @@ Chinese_CLIP为CLIP模型的中文版本，使用大规模中文数据进行训 ```shell git clone https://gitee.com/ascend/ModelZoo-PyTorch.git - cd ModelZoo-PyTorch/ACL_PyTorch/built-in/foundation/Chinese-CLIP + cd ModelZoo-PyTorch/ACL_PyTorch/built-in/foundation_models/Chinese-CLIP ``` 2. 获取第三方源码。 -- Gitee From 8845c885a7e33ec2de24dffc42f8d05e0e02cb5c Mon Sep 17 00:00:00 2001 From: lifangtian Date: Wed, 4 Jun 2025 05:12:17 +0000 Subject: [PATCH 09/35] =?UTF-8?q?=E4=BF=AE=E6=94=B9ACL=5FPyTorch/built-in/?= =?UTF-8?q?foundation=5Fmodels/Chinese=5FCLIP/README.md=EF=BC=9A=E6=9B=B4?= =?UTF-8?q?=E6=AD=A3=E5=9C=A8modelzoo=E4=BB=93=E9=87=8C=E7=9A=84cd?= =?UTF-8?q?=E8=B7=AF=E5=BE=84?= MIME-Version: 1.0 Content-Type: text/plain; charset=UTF-8 Content-Transfer-Encoding: 8bit Signed-off-by: lifangtian --- ACL_PyTorch/built-in/foundation_models/Chinese_CLIP/README.md | 2 +- 1 file changed, 1 insertion(+), 1 deletion(-) diff --git a/ACL_PyTorch/built-in/foundation_models/Chinese_CLIP/README.md b/ACL_PyTorch/built-in/foundation_models/Chinese_CLIP/README.md index d1fb1595d5..05538e637e 100644 --- a/ACL_PyTorch/built-in/foundation_models/Chinese_CLIP/README.md +++ b/ACL_PyTorch/built-in/foundation_models/Chinese_CLIP/README.md @@ -73,7 +73,7 @@ Chinese_CLIP为CLIP模型的中文版本，使用大规模中文数据进行训 ```shell git clone https://gitee.com/ascend/ModelZoo-PyTorch.git - cd ModelZoo-PyTorch/ACL_PyTorch/built-in/foundation_models/Chinese-CLIP + cd ModelZoo-PyTorch/ACL_PyTorch/built-in/foundation_models/Chinese_CLIP ``` 2. 获取第三方源码。 -- Gitee From beef95de0b36d4474d5395c292e3e7046762e10c Mon Sep 17 00:00:00 2001 From: lifangtian Date: Wed, 4 Jun 2025 05:14:57 +0000 Subject: [PATCH 10/35] =?UTF-8?q?=E4=BF=AE=E6=94=B9ACL=5FPyTorch/built-in/?= =?UTF-8?q?cv/GLIP/README.md=EF=BC=9A=E6=9B=B4=E6=AD=A3=E5=AE=89=E8=A3=85?= =?UTF-8?q?=E4=BE=9D=E8=B5=96=E5=91=BD=E4=BB=A4=E7=9A=84requirement.txt?= =?UTF-8?q?=E7=9A=84=E6=96=87=E4=BB=B6=E5=90=8D?= MIME-Version: 1.0 Content-Type: text/plain; charset=UTF-8 Content-Transfer-Encoding: 8bit Signed-off-by: lifangtian --- ACL_PyTorch/built-in/cv/GLIP/README.md | 2 +- 1 file changed, 1 insertion(+), 1 deletion(-) diff --git a/ACL_PyTorch/built-in/cv/GLIP/README.md b/ACL_PyTorch/built-in/cv/GLIP/README.md index 97fd8b2733..a450f96877 100644 --- a/ACL_PyTorch/built-in/cv/GLIP/README.md +++ b/ACL_PyTorch/built-in/cv/GLIP/README.md @@ -90,7 +90,7 @@ CLIP和ALIGN在大规模图像-文本对上进行跨模态对比学习，可以 2. 安装依赖。 ``` - pip3 install -r requirements.txt + pip3 install -r requirement.txt ``` 3. 打补丁。 -- Gitee From d255be8c660217d4423c3e88a0d828e3a88c02ed Mon Sep 17 00:00:00 2001 From: lifangtian Date: Wed, 4 Jun 2025 05:23:18 +0000 Subject: [PATCH 11/35] =?UTF-8?q?=E4=BF=AE=E6=94=B9ACL=5FPyTorch/built-in/?= =?UTF-8?q?cv/GLIP/requirement.txt=EF=BC=9A=E6=9B=B4=E6=AD=A3pycocotools?= =?UTF-8?q?=E7=9A=84=E7=89=88=E6=9C=AC=E5=8F=B7?= MIME-Version: 1.0 Content-Type: text/plain; charset=UTF-8 Content-Transfer-Encoding: 8bit Signed-off-by: lifangtian --- ACL_PyTorch/built-in/cv/GLIP/requirement.txt | 2 +- 1 file changed, 1 insertion(+), 1 deletion(-) diff --git a/ACL_PyTorch/built-in/cv/GLIP/requirement.txt b/ACL_PyTorch/built-in/cv/GLIP/requirement.txt index cfe91108db..15874dc5b6 100644 --- a/ACL_PyTorch/built-in/cv/GLIP/requirement.txt +++ b/ACL_PyTorch/built-in/cv/GLIP/requirement.txt @@ -1,7 +1,7 @@ onnx torch==1.13.1 tqdm==4.62.3 -pycocotools==2.07 +pycocotools==2.0.7 pyacl==1.0.0 torchvision==0.14.1 timm==0.6.13 -- Gitee From 69779bcd66d9cb997c5517cb2cbbe06359977e28 Mon Sep 17 00:00:00 2001 From: lifangtian Date: Wed, 4 Jun 2025 06:01:48 +0000 Subject: [PATCH 12/35] =?UTF-8?q?=E4=BF=AE=E6=94=B9PSENet=5FResNet50=5Fvd/?= =?UTF-8?q?README.md=EF=BC=9A=E8=A1=A5=E5=85=85=E6=95=B0=E6=8D=AE=E9=9B=86?= =?UTF-8?q?=E7=9A=84=E9=93=BE=E6=8E=A5=EF=BC=9B=E6=9B=B4=E6=AD=A3=E8=A7=A3?= =?UTF-8?q?=E5=8E=8B=E5=91=BD=E4=BB=A4=EF=BC=9B=E6=9B=BF=E6=8D=A2=E5=A4=B1?= =?UTF-8?q?=E6=95=88=E7=9A=84auto-optimizer=E9=93=BE=E6=8E=A5?= MIME-Version: 1.0 Content-Type: text/plain; charset=UTF-8 Content-Transfer-Encoding: 8bit Signed-off-by: lifangtian --- .../contrib/cv/detection/PSENet_ResNet50_vd/README.md | 6 +++--- 1 file changed, 3 insertions(+), 3 deletions(-) diff --git a/ACL_PyTorch/contrib/cv/detection/PSENet_ResNet50_vd/README.md b/ACL_PyTorch/contrib/cv/detection/PSENet_ResNet50_vd/README.md index 848f3daa92..62d7e61eb7 100644 --- a/ACL_PyTorch/contrib/cv/detection/PSENet_ResNet50_vd/README.md +++ b/ACL_PyTorch/contrib/cv/detection/PSENet_ResNet50_vd/README.md @@ -89,12 +89,12 @@ PSENet([Shape Robust Text Detection with Progressive Scale Expansion Network](ht 1. 获取原始数据集。（解压命令参考tar –xvf \*.tar与 unzip \*.zip） - IICDAR 2015 数据集包含1000张训练图像和500张测试图像。参考PaddleOCR数据集数据处理方式，ICDAR 2015 数据集可以点击链接进行下载，本模型需下载Test Set Images(43.3MB)。 + ICDAR 2015 数据集包含1000张训练图像和500张测试图像。参考PaddleOCR数据集数据处理方式，ICDAR 2015 数据集可以点击[链接](https://aistudio.baidu.com/datasetdetail/46088)进行下载，本模型需下载Test Set Images(43.3MB)。 将数据集`ch4_test_images.zip`放在工作目录下，通过以下命令创建`train_data/icdar2015/ch4_test_images`路径，并通过以下命令进行解压保存并获取标签文件。 ``` mkdir -p ./train_data/icdar2015/ch4_test_images/ - unzip -d ./train_data/icdar2015/ch4_test_images/ ch4_test_images.zip + unzip -d ch4_test_images.zip ./train_data/icdar2015/ch4_test_images/ wget -P ./train_data/icdar2015/ https://paddleocr.bj.bcebos.com/dataset/test_icdar2015_label.txt ``` @@ -166,7 +166,7 @@ PSENet([Shape Robust Text Detection with Progressive Scale Expansion Network](ht 运行后获得PSENet_ResNet50_vd_dybs.onnx文件。 2. 优化onnx模型。 - 请访问[auto-optimizer优化工具](https://gitee.com/ascend/msadvisor/tree/master/auto-optimizer)代码仓，根据readme文档进行工具安装。 + 请访问[auto-optimizer优化工具](https://gitee.com/ascend/msit/blob/master/msit/README.md)代码仓，根据readme文档进行benchmark和surgeon工具安装。 运行modify_onnx.py脚本优化onnx模型，优化点为：Resize算子按Paddle定义参数导出的onnx模型有精度问题，因此将按PyTorch定义重新构造Resize参数。 -- Gitee From 53a01a0ff2cd305673219dda5e7adf524128338e Mon Sep 17 00:00:00 2001 From: lifangtian Date: Wed, 4 Jun 2025 06:05:09 +0000 Subject: [PATCH 13/35] =?UTF-8?q?=E4=BF=AE=E6=94=B9ACL=5FPyTorch/contrib/c?= =?UTF-8?q?v/detection/PSENet=5FResNet50=5Fvd/requirements.txt=EF=BC=9A?= =?UTF-8?q?=E5=A2=9E=E5=8A=A0=E9=9C=80=E8=A6=81=E7=9A=84onnxsim=E4=BE=9D?= =?UTF-8?q?=E8=B5=96=EF=BC=9B=E9=99=8D=E7=BA=A7protobuf=E4=BB=A5=E8=B7=91?= =?UTF-8?q?=E9=80=9A?= MIME-Version: 1.0 Content-Type: text/plain; charset=UTF-8 Content-Transfer-Encoding: 8bit Signed-off-by: lifangtian --- .../contrib/cv/detection/PSENet_ResNet50_vd/requirements.txt | 4 +++- 1 file changed, 3 insertions(+), 1 deletion(-) diff --git a/ACL_PyTorch/contrib/cv/detection/PSENet_ResNet50_vd/requirements.txt b/ACL_PyTorch/contrib/cv/detection/PSENet_ResNet50_vd/requirements.txt index 40499b7911..04c3268a5f 100644 --- a/ACL_PyTorch/contrib/cv/detection/PSENet_ResNet50_vd/requirements.txt +++ b/ACL_PyTorch/contrib/cv/detection/PSENet_ResNet50_vd/requirements.txt @@ -19,4 +19,6 @@ lxml premailer openpyxl attrdict -sympy \ No newline at end of file +sympy +onnxsim +protobuf==3.20.3 \ No newline at end of file -- Gitee From 95e7a94395d32e44ad4bc545bbdced116502f0c5 Mon Sep 17 00:00:00 2001 From: lifangtian Date: Wed, 4 Jun 2025 06:32:27 +0000 Subject: [PATCH 14/35] =?UTF-8?q?=E4=BF=AE=E6=94=B9readme.md=EF=BC=9A?= =?UTF-8?q?=E5=AF=BC=E5=87=BAONNX=E5=91=BD=E4=BB=A4=E5=8A=A0=E5=85=A5?= =?UTF-8?q?=E7=BC=BA=E5=A4=B1=E7=9A=84=E6=8D=A2=E8=A1=8C=E7=AC=A6=EF=BC=9B?= =?UTF-8?q?=E7=A6=BB=E7=BA=BF=E6=8E=A8=E7=90=86=E5=92=8C=E6=80=A7=E8=83=BD?= =?UTF-8?q?=E6=8E=A8=E7=90=86=E7=9A=84=E5=91=BD=E4=BB=A4=E8=A1=A5=E5=85=85?= =?UTF-8?q?=E7=BC=BA=E5=A4=B1=E7=9A=84.om=E5=90=8E=E7=BC=80=EF=BC=9B?= =?UTF-8?q?=E8=AF=B4=E6=98=8E=E9=A1=BB=E6=89=8B=E5=8A=A8=E5=88=9B=E5=BB=BA?= =?UTF-8?q?=E7=9A=84=E8=B7=AF=E5=BE=84?= MIME-Version: 1.0 Content-Type: text/plain; charset=UTF-8 Content-Transfer-Encoding: 8bit Signed-off-by: lifangtian --- ACL_PyTorch/built-in/cv/FCENet_for_PyTorch/readme.md | 8 ++++---- 1 file changed, 4 insertions(+), 4 deletions(-) diff --git a/ACL_PyTorch/built-in/cv/FCENet_for_PyTorch/readme.md b/ACL_PyTorch/built-in/cv/FCENet_for_PyTorch/readme.md index 64b7196534..22b91e6c32 100644 --- a/ACL_PyTorch/built-in/cv/FCENet_for_PyTorch/readme.md +++ b/ACL_PyTorch/built-in/cv/FCENet_for_PyTorch/readme.md @@ -77,7 +77,7 @@ FCENet，使用傅里叶变换来得到文本的包围框，该方法在弯曲 ## 准备数据集 1. 获取原始数据集 - 本模型需要icdar2015数据集，数据集请参考开源代码仓方式获取。获取icdar2015数据集，放到mmocr的data文件夹内,放置顺序如下。 + 本模型需要icdar2015数据集，数据集请参考开源代码仓方式获取。获取icdar2015数据集，新建mmocr/data路径并将数据集放到该路径内内,放置顺序如下。 ``` ├── icdar2015 │ ├── imgs @@ -104,7 +104,7 @@ FCENet，使用傅里叶变换来得到文本的包围框，该方法在弯曲 然后执行执行以下命令生成 ONNX 模型： ```bash - python3 ./pytorch2onnx.py + python3 ./pytorch2onnx.py \ ./mmocr/configs/textdet/fcenet/fcenet_r50_fpn_1500e_icdar2015.py \ ./fcenet_r50_fpn_1500e_icdar2015_20211022-daefb6ed.pth \ det \ @@ -172,7 +172,7 @@ FCENet，使用傅里叶变换来得到文本的包围框，该方法在弯曲 该离线模型使用ais_infer作为推理工具，请参考[**安装文档**](https://gitee.com/ascend/tools/tree/master/ais-bench_workload/tool/ais_bench#%E4%B8%80%E9%94%AE%E5%AE%89%E8%A3%85)安装推理后端包aclruntime与推理前端包ais_bench。完成安装后，执行以下命令预处理后的数据进行推理。 ```bash python3 -m ais_bench \ - --model ./fcenet_bs${batch_size} \ + --model ./fcenet_bs${batch_size}.om \ --input ./preprocessed_imgs/ \ --output ./result \ --outfmt TXT \ @@ -192,7 +192,7 @@ FCENet，使用傅里叶变换来得到文本的包围框，该方法在弯曲 + 为了避免测试过程因持续时间太长而受到干扰，建议通过纯推理的方式进行性能测试。 + 使用吞吐率作为性能指标，单位为 fps，反映模型在单位时间（1秒）内处理的样本数。 ```bash - python3 -m ais_bench --model ./fcenet_bs${batch_size} --batchsize ${batch_size} + python3 -m ais_bench --model ./fcenet_bs${batch_size}.om --batchsize ${batch_size} ``` 执行完纯推理命令，程序会打印出与性能相关的指标，找到以关键字 **[INFO] throughput** 开头的一行，行尾的数字即为 OM 模型的吞吐率。 -- Gitee From 8c1ad3f504d1aa80e5c24e7014815e19e7b40935 Mon Sep 17 00:00:00 2001 From: lifangtian Date: Wed, 4 Jun 2025 07:01:21 +0000 Subject: [PATCH 15/35] =?UTF-8?q?update=20ACL=5FPyTorch/built-in/foundatio?= =?UTF-8?q?n=5Fmodels/stable=5Fdiffusion/README.md=EF=BC=9A=E5=A2=9E?= =?UTF-8?q?=E5=8A=A0=E8=8E=B7=E5=8F=96modelzoo=E6=BA=90=E7=A0=81=E7=8E=AF?= =?UTF-8?q?=E8=8A=82?= MIME-Version: 1.0 Content-Type: text/plain; charset=UTF-8 Content-Transfer-Encoding: 8bit Signed-off-by: lifangtian --- .../built-in/foundation_models/stable_diffusion/README.md | 6 ++++++ 1 file changed, 6 insertions(+) diff --git a/ACL_PyTorch/built-in/foundation_models/stable_diffusion/README.md b/ACL_PyTorch/built-in/foundation_models/stable_diffusion/README.md index e8f2fb832a..7b4f7d78e0 100755 --- a/ACL_PyTorch/built-in/foundation_models/stable_diffusion/README.md +++ b/ACL_PyTorch/built-in/foundation_models/stable_diffusion/README.md @@ -62,6 +62,12 @@ # 快速上手 ## 获取源码 +1. 获取本仓源码 + + ``` + git clone https://gitee.com/ascend/ModelZoo-PyTorch.git + cd ModelZoo-PyTorch/ACL_PyTorch/built-in/foundation_models/stable_diffusion + ``` 1. 安装依赖。 ```bash -- Gitee From 37f337f5becee606f443743671243989a6d61b64 Mon Sep 17 00:00:00 2001 From: lifangtian Date: Wed, 4 Jun 2025 07:12:17 +0000 Subject: [PATCH 16/35] =?UTF-8?q?update=20ACL=5FPyTorch/built-in/foundatio?= =?UTF-8?q?n=5Fmodels/stable=5Fdiffusionxl/README.md=EF=BC=9A=E5=A2=9E?= =?UTF-8?q?=E5=8A=A0=E8=8E=B7=E5=8F=96modelzoo=E6=BA=90=E7=A0=81=E7=8E=AF?= =?UTF-8?q?=E8=8A=82?= MIME-Version: 1.0 Content-Type: text/plain; charset=UTF-8 Content-Transfer-Encoding: 8bit Signed-off-by: lifangtian --- .../built-in/foundation_models/stable_diffusionxl/README.md | 6 ++++++ 1 file changed, 6 insertions(+) diff --git a/ACL_PyTorch/built-in/foundation_models/stable_diffusionxl/README.md b/ACL_PyTorch/built-in/foundation_models/stable_diffusionxl/README.md index 9ba1da8313..c0c6356c4c 100644 --- a/ACL_PyTorch/built-in/foundation_models/stable_diffusionxl/README.md +++ b/ACL_PyTorch/built-in/foundation_models/stable_diffusionxl/README.md @@ -60,6 +60,12 @@ # 快速上手 ## 获取源码 +1. 获取本仓源码 + + ``` + git clone https://gitee.com/ascend/ModelZoo-PyTorch.git + cd ModelZoo-PyTorch/ACL_PyTorch/built-in/foundation_models/stable_diffusionxl + ``` 1. 安装依赖。 ```bash -- Gitee From 65e486525b2267da96d6949de918e24c71d9d6df Mon Sep 17 00:00:00 2001 From: lifangtian Date: Wed, 4 Jun 2025 07:19:37 +0000 Subject: [PATCH 17/35] =?UTF-8?q?update=20ACL=5FPyTorch/built-in/foundatio?= =?UTF-8?q?n=5Fmodels/stable=5Fdiffusionxl=5Frefiner/README.md=EF=BC=9A?= =?UTF-8?q?=E5=A2=9E=E5=8A=A0=E8=8E=B7=E5=8F=96modelzoo=E6=BA=90=E7=A0=81?= =?UTF-8?q?=E7=8E=AF=E8=8A=82?= MIME-Version: 1.0 Content-Type: text/plain; charset=UTF-8 Content-Transfer-Encoding: 8bit Signed-off-by: lifangtian --- .../foundation_models/stable_diffusionxl_refiner/README.md | 6 ++++++ 1 file changed, 6 insertions(+) diff --git a/ACL_PyTorch/built-in/foundation_models/stable_diffusionxl_refiner/README.md b/ACL_PyTorch/built-in/foundation_models/stable_diffusionxl_refiner/README.md index 960633a6d1..a9bd9eed5f 100644 --- a/ACL_PyTorch/built-in/foundation_models/stable_diffusionxl_refiner/README.md +++ b/ACL_PyTorch/built-in/foundation_models/stable_diffusionxl_refiner/README.md @@ -58,6 +58,12 @@ # 快速上手 ## 获取源码 +1. 获取本仓源码 + + ``` + git clone https://gitee.com/ascend/ModelZoo-PyTorch.git + cd ModelZoo-PyTorch/ACL_PyTorch/built-in/foundation_models/stable_diffusionxl_refiner + ``` 1. 安装依赖。 ```bash -- Gitee From d14c275c2fcd73b11a031f132dadfd843d806b20 Mon Sep 17 00:00:00 2001 From: lifangtian Date: Wed, 4 Jun 2025 07:23:06 +0000 Subject: [PATCH 18/35] =?UTF-8?q?update=20ACL=5FPyTorch/contrib/cv/detecti?= =?UTF-8?q?on/FCENet/readme.md=EF=BC=9A=E5=A2=9E=E5=8A=A0=E8=8E=B7?= =?UTF-8?q?=E5=8F=96modelzoo=E4=BB=A3=E7=A0=81=E7=8E=AF=E8=8A=82?= MIME-Version: 1.0 Content-Type: text/plain; charset=UTF-8 Content-Transfer-Encoding: 8bit Signed-off-by: lifangtian --- ACL_PyTorch/contrib/cv/detection/FCENet/readme.md | 6 ++++++ 1 file changed, 6 insertions(+) diff --git a/ACL_PyTorch/contrib/cv/detection/FCENet/readme.md b/ACL_PyTorch/contrib/cv/detection/FCENet/readme.md index a5307dee69..b5ca284160 100644 --- a/ACL_PyTorch/contrib/cv/detection/FCENet/readme.md +++ b/ACL_PyTorch/contrib/cv/detection/FCENet/readme.md @@ -58,6 +58,12 @@ FCENet，使用傅里叶变换来得到文本的包围框，该方法在弯曲 # 快速上手 ## 安装 +- 获取本仓源码 + + ``` + git clone https://gitee.com/ascend/ModelZoo-PyTorch.git + cd ModelZoo-PyTorch/ACL_PyTorch/contrib/cv/detection/FCENet + ``` - 安装推理过程所需的依赖 ```bash -- Gitee From fc0b88220d7a31ddfe7325fb0e2128d31a16125f Mon Sep 17 00:00:00 2001 From: lifangtian Date: Wed, 4 Jun 2025 07:27:21 +0000 Subject: [PATCH 19/35] =?UTF-8?q?update=20ACL=5FPyTorch/contrib/cv/detecti?= =?UTF-8?q?on/PSENet=5FResNet50=5Fvd/README.md=EF=BC=9A=E5=A2=9E=E5=8A=A0?= =?UTF-8?q?=E8=8E=B7=E5=8F=96modelzoo=E6=BA=90=E7=A0=81=E7=8E=AF=E8=8A=82?= MIME-Version: 1.0 Content-Type: text/plain; charset=UTF-8 Content-Transfer-Encoding: 8bit Signed-off-by: lifangtian --- .../contrib/cv/detection/PSENet_ResNet50_vd/README.md | 6 ++++++ 1 file changed, 6 insertions(+) diff --git a/ACL_PyTorch/contrib/cv/detection/PSENet_ResNet50_vd/README.md b/ACL_PyTorch/contrib/cv/detection/PSENet_ResNet50_vd/README.md index 62d7e61eb7..4879cc8436 100644 --- a/ACL_PyTorch/contrib/cv/detection/PSENet_ResNet50_vd/README.md +++ b/ACL_PyTorch/contrib/cv/detection/PSENet_ResNet50_vd/README.md @@ -64,6 +64,12 @@ PSENet([Shape Robust Text Detection with Progressive Scale Expansion Network](ht # 快速上手 ## 获取源码 +1. 获取本仓源码 + + ``` + git clone https://gitee.com/ascend/ModelZoo-PyTorch.git + cd ModelZoo-PyTorch/ACL_PyTorch/contrib/cv/detection/PSENet_ResNet50_vd + ``` 1. 获取源码。 -- Gitee From 0318cbb13b5b87098878d5c62c5d3c72f643fff8 Mon Sep 17 00:00:00 2001 From: lifangtian Date: Wed, 4 Jun 2025 07:41:29 +0000 Subject: [PATCH 20/35] =?UTF-8?q?update=20ACL=5FPyTorch/built-in/cv/MGN=5F?= =?UTF-8?q?for=5FPytorch/README.md=EF=BC=9A=E5=A2=9E=E5=8A=A0=E8=8E=B7?= =?UTF-8?q?=E5=8F=96modelzoo=E4=BB=93=E6=BA=90=E7=A0=81=E7=8E=AF=E8=8A=82?= MIME-Version: 1.0 Content-Type: text/plain; charset=UTF-8 Content-Transfer-Encoding: 8bit Signed-off-by: lifangtian --- ACL_PyTorch/built-in/cv/MGN_for_Pytorch/README.md | 7 ++++++- 1 file changed, 6 insertions(+), 1 deletion(-) diff --git a/ACL_PyTorch/built-in/cv/MGN_for_Pytorch/README.md b/ACL_PyTorch/built-in/cv/MGN_for_Pytorch/README.md index 51831211ed..324b4a37e8 100644 --- a/ACL_PyTorch/built-in/cv/MGN_for_Pytorch/README.md +++ b/ACL_PyTorch/built-in/cv/MGN_for_Pytorch/README.md @@ -60,11 +60,16 @@ MGN网络是一种多分支深度网络架构的特征识别网络，由一个 # 快速上手 ## 获取源码 +1. 获取本仓源码 + + ``` + git clone https://gitee.com/ascend/ModelZoo-PyTorch.git + cd ModelZoo-PyTorch/ACL_PyTorch/built-in/cv/MGN_for_Pytorch + ``` 1. 获取源码。 ``` - cd ./ReId-MGN-master git clone https://github.com/GNAYUOHZ/ReID-MGN.git ./MGN patch -R MGN/data.py < module.patch ``` -- Gitee From 525539a0ff561b11252436ef92cf37bfdbcd0d63 Mon Sep 17 00:00:00 2001 From: lifangtian Date: Wed, 4 Jun 2025 07:44:20 +0000 Subject: [PATCH 21/35] =?UTF-8?q?update=20ACL=5FPyTorch/built-in/cv/SAM/RE?= =?UTF-8?q?ADME.md=EF=BC=9A=E5=A2=9E=E5=8A=A0=E8=8E=B7=E5=8F=96modelzoo?= =?UTF-8?q?=E6=BA=90=E7=A0=81=E4=BB=93=E7=9A=84=E7=8E=AF=E8=8A=82?= MIME-Version: 1.0 Content-Type: text/plain; charset=UTF-8 Content-Transfer-Encoding: 8bit Signed-off-by: lifangtian --- ACL_PyTorch/built-in/cv/SAM/README.md | 8 ++++++++ 1 file changed, 8 insertions(+) diff --git a/ACL_PyTorch/built-in/cv/SAM/README.md b/ACL_PyTorch/built-in/cv/SAM/README.md index 56d5fb23ab..ecc12cc473 100644 --- a/ACL_PyTorch/built-in/cv/SAM/README.md +++ b/ACL_PyTorch/built-in/cv/SAM/README.md @@ -69,10 +69,18 @@ SAM 首先会自动分割图像中的所有内容，但是如果你需要分割 | 说明：Atlas 300I Duo 推理卡请以CANN版本选择实际固件与驱动版本。 | \ | \ | ## 3. 快速上手 +1. 获取本仓源码 + + ``` + git clone https://gitee.com/ascend/ModelZoo-PyTorch.git + cd ModelZoo-PyTorch/ACL_PyTorch/built-in/cv/GLIP + ``` ### 3.1 获取源码 ``` +git clone https://gitee.com/ascend/ModelZoo-PyTorch.git +cd ModelZoo-PyTorch/ACL_PyTorch/built-in/cv/SAM git clone https://github.com/facebookresearch/segment-anything.git cd segment-anything git reset --hard 6fdee8f2727f4506cfbbe553e23b895e27956588 -- Gitee From 9dac1630982facab3b87be6bff99c575c2c64ad0 Mon Sep 17 00:00:00 2001 From: lifangtian Date: Wed, 4 Jun 2025 07:45:05 +0000 Subject: [PATCH 22/35] =?UTF-8?q?update=20ACL=5FPyTorch/built-in/cv/SAM/RE?= =?UTF-8?q?ADME.md=EF=BC=9A=E5=88=A0=E9=99=A4=E8=8D=89=E7=A8=BF=E9=81=97?= =?UTF-8?q?=E7=95=99?= MIME-Version: 1.0 Content-Type: text/plain; charset=UTF-8 Content-Transfer-Encoding: 8bit Signed-off-by: lifangtian --- ACL_PyTorch/built-in/cv/SAM/README.md | 6 ------ 1 file changed, 6 deletions(-) diff --git a/ACL_PyTorch/built-in/cv/SAM/README.md b/ACL_PyTorch/built-in/cv/SAM/README.md index ecc12cc473..d5a94032a9 100644 --- a/ACL_PyTorch/built-in/cv/SAM/README.md +++ b/ACL_PyTorch/built-in/cv/SAM/README.md @@ -69,12 +69,6 @@ SAM 首先会自动分割图像中的所有内容，但是如果你需要分割 | 说明：Atlas 300I Duo 推理卡请以CANN版本选择实际固件与驱动版本。 | \ | \ | ## 3. 快速上手 -1. 获取本仓源码 - - ``` - git clone https://gitee.com/ascend/ModelZoo-PyTorch.git - cd ModelZoo-PyTorch/ACL_PyTorch/built-in/cv/GLIP - ``` ### 3.1 获取源码 -- Gitee From 957b270fe2cf2ac96f155f03fdbd370a62642cc0 Mon Sep 17 00:00:00 2001 From: lifangtian Date: Wed, 4 Jun 2025 07:50:31 +0000 Subject: [PATCH 23/35] =?UTF-8?q?update=20ACL=5FPyTorch/built-in/foundatio?= =?UTF-8?q?n=5Fmodels/DiT/README.md=EF=BC=9A=E6=96=B0=E5=A2=9E=E8=8E=B7?= =?UTF-8?q?=E5=8F=96modelzoo=E4=BB=A3=E7=A0=81=E4=BB=93=E7=8E=AF=E8=8A=82?= MIME-Version: 1.0 Content-Type: text/plain; charset=UTF-8 Content-Transfer-Encoding: 8bit Signed-off-by: lifangtian --- ACL_PyTorch/built-in/foundation_models/DiT/README.md | 6 ++++++ 1 file changed, 6 insertions(+) diff --git a/ACL_PyTorch/built-in/foundation_models/DiT/README.md b/ACL_PyTorch/built-in/foundation_models/DiT/README.md index 2850b7f8bd..aa230fd5a7 100644 --- a/ACL_PyTorch/built-in/foundation_models/DiT/README.md +++ b/ACL_PyTorch/built-in/foundation_models/DiT/README.md @@ -71,6 +71,12 @@ latent_size = image_size // 8 # 快速上手 ## 获取源码 +1. 获取本仓源码 + + ``` + git clone https://gitee.com/ascend/ModelZoo-PyTorch.git + cd ModelZoo-PyTorch/ACL_PyTorch/built-in/foundation_models/DiT + ``` 1. 获取源码。 -- Gitee From df8bb4d5c77d0290fad80aa8622cd0dc00e94744 Mon Sep 17 00:00:00 2001 From: lifangtian Date: Wed, 4 Jun 2025 07:53:12 +0000 Subject: [PATCH 24/35] =?UTF-8?q?update=20ACL=5FPyTorch/contrib/cv/image?= =?UTF-8?q?=5Fretrieval/BLIP/readme.md=EF=BC=9A=E6=96=B0=E5=A2=9E=E8=8E=B7?= =?UTF-8?q?=E5=8F=96modelzoo=E4=BB=A3=E7=A0=81=E4=BB=93=E7=8E=AF=E8=8A=82?= MIME-Version: 1.0 Content-Type: text/plain; charset=UTF-8 Content-Transfer-Encoding: 8bit Signed-off-by: lifangtian --- ACL_PyTorch/contrib/cv/image_retrieval/BLIP/readme.md | 6 ++++++ 1 file changed, 6 insertions(+) diff --git a/ACL_PyTorch/contrib/cv/image_retrieval/BLIP/readme.md b/ACL_PyTorch/contrib/cv/image_retrieval/BLIP/readme.md index 51d07f4420..71de34dc57 100644 --- a/ACL_PyTorch/contrib/cv/image_retrieval/BLIP/readme.md +++ b/ACL_PyTorch/contrib/cv/image_retrieval/BLIP/readme.md @@ -109,6 +109,12 @@ BLIP模型为一种新的Vision-Language Pre-training框架，它可以灵活地 # 快速上手 ## 获取源码 +1. 获取本仓源码 + + ``` + git clone https://gitee.com/ascend/ModelZoo-PyTorch.git + cd ModelZoo-PyTorch/ACL_PyTorch/built-in/cv/image_retrieval/BLIP + ``` 1. 获取源码。 -- Gitee From 74b21f09f713bf98d07eb8fb4886a768b1a57e12 Mon Sep 17 00:00:00 2001 From: lifangtian Date: Wed, 4 Jun 2025 07:54:27 +0000 Subject: [PATCH 25/35] =?UTF-8?q?update=20ACL=5FPyTorch/contrib/cv/image?= =?UTF-8?q?=5Fretrieval/BLIP/readme.md=EF=BC=9A=E6=96=B0=E5=A2=9E=E8=8E=B7?= =?UTF-8?q?=E5=8F=96modelzoo=E4=BB=A3=E7=A0=81=E4=BB=93=E7=8E=AF=E8=8A=82?= MIME-Version: 1.0 Content-Type: text/plain; charset=UTF-8 Content-Transfer-Encoding: 8bit Signed-off-by: lifangtian --- ACL_PyTorch/contrib/cv/image_retrieval/BLIP/readme.md | 2 +- 1 file changed, 1 insertion(+), 1 deletion(-) diff --git a/ACL_PyTorch/contrib/cv/image_retrieval/BLIP/readme.md b/ACL_PyTorch/contrib/cv/image_retrieval/BLIP/readme.md index 71de34dc57..2e0e9cb47c 100644 --- a/ACL_PyTorch/contrib/cv/image_retrieval/BLIP/readme.md +++ b/ACL_PyTorch/contrib/cv/image_retrieval/BLIP/readme.md @@ -113,7 +113,7 @@ BLIP模型为一种新的Vision-Language Pre-training框架，它可以灵活地 ``` git clone https://gitee.com/ascend/ModelZoo-PyTorch.git - cd ModelZoo-PyTorch/ACL_PyTorch/built-in/cv/image_retrieval/BLIP + cd ModelZoo-PyTorch/ACL_PyTorch/contrib/cv/image_retrieval/BLIP ``` 1. 获取源码。 -- Gitee From 80512ff85b5e5673476e48a1fe91409eb163446d Mon Sep 17 00:00:00 2001 From: lifangtian Date: Wed, 4 Jun 2025 07:57:47 +0000 Subject: [PATCH 26/35] =?UTF-8?q?update=20ACL=5FPyTorch/contrib/audio/wav2?= =?UTF-8?q?lip=5FID100400/README.md=EF=BC=9A=E6=96=B0=E5=A2=9E=E8=8E=B7?= =?UTF-8?q?=E5=8F=96modelzoo=E4=BB=A3=E7=A0=81=E4=BB=93=E7=8E=AF=E8=8A=82?= MIME-Version: 1.0 Content-Type: text/plain; charset=UTF-8 Content-Transfer-Encoding: 8bit Signed-off-by: lifangtian --- ACL_PyTorch/contrib/audio/wav2lip_ID100400/README.md | 6 ++++++ 1 file changed, 6 insertions(+) diff --git a/ACL_PyTorch/contrib/audio/wav2lip_ID100400/README.md b/ACL_PyTorch/contrib/audio/wav2lip_ID100400/README.md index 84a7479100..bfe51c10d1 100644 --- a/ACL_PyTorch/contrib/audio/wav2lip_ID100400/README.md +++ b/ACL_PyTorch/contrib/audio/wav2lip_ID100400/README.md @@ -68,6 +68,12 @@ # 快速上手 ## 获取源码 +1. 获取本仓源码 + + ``` + git clone https://gitee.com/ascend/ModelZoo-PyTorch.git + cd ModelZoo-PyTorch/ACL_PyTorch/contrib/audio/wav2lip_ID100400 + ``` 1. 获取源码。 -- Gitee From 1966028d423e85c3886d3261464ca33765df1b51 Mon Sep 17 00:00:00 2001 From: lifangtian Date: Wed, 4 Jun 2025 08:00:05 +0000 Subject: [PATCH 27/35] =?UTF-8?q?update=20ACL=5FPyTorch/built-in/embedding?= =?UTF-8?q?/jina-embeddings-v2-base-code/README.md=EF=BC=9A=E6=96=B0?= =?UTF-8?q?=E5=A2=9E=E8=8E=B7=E5=8F=96modelzoo=E4=BB=A3=E7=A0=81=E4=BB=93?= =?UTF-8?q?=E7=8E=AF=E8=8A=82?= MIME-Version: 1.0 Content-Type: text/plain; charset=UTF-8 Content-Transfer-Encoding: 8bit Signed-off-by: lifangtian --- .../embedding/jina-embeddings-v2-base-code/README.md | 6 ++++++ 1 file changed, 6 insertions(+) diff --git a/ACL_PyTorch/built-in/embedding/jina-embeddings-v2-base-code/README.md b/ACL_PyTorch/built-in/embedding/jina-embeddings-v2-base-code/README.md index 2a4d24e7f8..a36416d312 100644 --- a/ACL_PyTorch/built-in/embedding/jina-embeddings-v2-base-code/README.md +++ b/ACL_PyTorch/built-in/embedding/jina-embeddings-v2-base-code/README.md @@ -38,6 +38,12 @@ # 快速上手 ## 获取源码 +1. 获取本仓源码 + + ``` + git clone https://gitee.com/ascend/ModelZoo-PyTorch.git + cd ModelZoo-PyTorch/ACL_PyTorch/built-in/embedding/jina-embeddings-v2-base-code + ``` 1. 获取开源模型源码和权重（可选） > 如果您的设备可以方便的直接从hugging-hub下载权重和代码，则不需要执行这一步 -- Gitee From 564eefd0f39aeead812e682c75809785cb01b306 Mon Sep 17 00:00:00 2001 From: lifangtian Date: Wed, 4 Jun 2025 08:01:48 +0000 Subject: [PATCH 28/35] =?UTF-8?q?update=20ACL=5FPyTorch/built-in/audio/Cos?= =?UTF-8?q?yVoice/README.md=EF=BC=9A=E6=96=B0=E5=A2=9E=E8=8E=B7=E5=8F=96mo?= =?UTF-8?q?delzoo=E4=BB=A3=E7=A0=81=E4=BB=93=E7=8E=AF=E8=8A=82?= MIME-Version: 1.0 Content-Type: text/plain; charset=UTF-8 Content-Transfer-Encoding: 8bit Signed-off-by: lifangtian --- ACL_PyTorch/built-in/audio/CosyVoice/README.md | 6 ++++++ 1 file changed, 6 insertions(+) diff --git a/ACL_PyTorch/built-in/audio/CosyVoice/README.md b/ACL_PyTorch/built-in/audio/CosyVoice/README.md index 1265ca94a6..a835df476a 100755 --- a/ACL_PyTorch/built-in/audio/CosyVoice/README.md +++ b/ACL_PyTorch/built-in/audio/CosyVoice/README.md @@ -39,6 +39,12 @@ # 快速上手 ## 获取源码 +1. 获取本仓源码 + + ``` + git clone https://gitee.com/ascend/ModelZoo-PyTorch.git + cd ModelZoo-PyTorch/ACL_PyTorch/built-in/audio/CosyVoice + ``` 1. 获取`PyTorch`源码 ``` -- Gitee From 2945288b174bc5ed8e627b9ea3d7ca8dd29f6872 Mon Sep 17 00:00:00 2001 From: lifangtian Date: Wed, 4 Jun 2025 08:03:11 +0000 Subject: [PATCH 29/35] =?UTF-8?q?update=20ACL=5FPyTorch/built-in/audio/Sen?= =?UTF-8?q?seVoice/README=5Fonnx.md=EF=BC=9A=E6=96=B0=E5=A2=9E=E8=8E=B7?= =?UTF-8?q?=E5=8F=96modelzoo=E4=BB=A3=E7=A0=81=E4=BB=93=E7=8E=AF=E8=8A=82?= MIME-Version: 1.0 Content-Type: text/plain; charset=UTF-8 Content-Transfer-Encoding: 8bit Signed-off-by: lifangtian --- ACL_PyTorch/built-in/audio/SenseVoice/README_onnx.md | 6 ++++++ 1 file changed, 6 insertions(+) diff --git a/ACL_PyTorch/built-in/audio/SenseVoice/README_onnx.md b/ACL_PyTorch/built-in/audio/SenseVoice/README_onnx.md index 35c8d0be1f..0ec56a6047 100755 --- a/ACL_PyTorch/built-in/audio/SenseVoice/README_onnx.md +++ b/ACL_PyTorch/built-in/audio/SenseVoice/README_onnx.md @@ -40,6 +40,12 @@ SenseVoice作为一款专注于高精度多语言语音识别的模型，其独 # 快速上手 ## 获取源码 +1. 获取本仓源码 + + ``` + git clone https://gitee.com/ascend/ModelZoo-PyTorch.git + cd ModelZoo-PyTorch/ACL_PyTorch/built-in/audio/SenseVoice + ``` 1. 获取`Pytorch`源码 ``` -- Gitee From ca731fd5f0e26b3739a60a32d6532a0f359cb86a Mon Sep 17 00:00:00 2001 From: lifangtian Date: Wed, 4 Jun 2025 08:07:10 +0000 Subject: [PATCH 30/35] =?UTF-8?q?update=20ACL=5FPyTorch/built-in/audio/Par?= =?UTF-8?q?aformer/README.md=EF=BC=9A=E6=96=B0=E5=A2=9E=E8=8E=B7=E5=8F=96m?= =?UTF-8?q?odelzoo=E4=BB=A3=E7=A0=81=E4=BB=93=E7=8E=AF=E8=8A=82?= MIME-Version: 1.0 Content-Type: text/plain; charset=UTF-8 Content-Transfer-Encoding: 8bit Signed-off-by: lifangtian --- ACL_PyTorch/built-in/audio/Paraformer/README.md | 8 +++++++- 1 file changed, 7 insertions(+), 1 deletion(-) diff --git a/ACL_PyTorch/built-in/audio/Paraformer/README.md b/ACL_PyTorch/built-in/audio/Paraformer/README.md index 4808977d00..e7c0d7b618 100644 --- a/ACL_PyTorch/built-in/audio/Paraformer/README.md +++ b/ACL_PyTorch/built-in/audio/Paraformer/README.md @@ -36,10 +36,16 @@ Paraformer是阿里达摩院语音团队提出的一种高效的非自回归端 # 快速上手 ## 获取源码 +1. 获取本仓源码 + + ``` + git clone https://gitee.com/ascend/ModelZoo-PyTorch.git + cd ModelZoo-PyTorch/ACL_PyTorch/built-in/audio/Paraformer + ``` 1. 安装依赖 ```bash - pip3 install -r ../requirements.txt + pip3 install -r requirements.txt ``` 2. 获取模型仓源码 -- Gitee From 8f7024c7725b1524d5e781be8422b58799519b82 Mon Sep 17 00:00:00 2001 From: lifangtian Date: Wed, 4 Jun 2025 08:10:36 +0000 Subject: [PATCH 31/35] =?UTF-8?q?update=20ACL=5FPyTorch/built-in/embedding?= =?UTF-8?q?/bge-m3/README.md=EF=BC=9A=E6=96=B0=E5=A2=9E=E8=8E=B7=E5=8F=96m?= =?UTF-8?q?odelzoo=E4=BB=A3=E7=A0=81=E4=BB=93=E7=8E=AF=E8=8A=82?= MIME-Version: 1.0 Content-Type: text/plain; charset=UTF-8 Content-Transfer-Encoding: 8bit Signed-off-by: lifangtian --- ACL_PyTorch/built-in/embedding/bge-m3/README.md | 7 +++++++ 1 file changed, 7 insertions(+) diff --git a/ACL_PyTorch/built-in/embedding/bge-m3/README.md b/ACL_PyTorch/built-in/embedding/bge-m3/README.md index e139c5ebc2..6290df26b8 100644 --- a/ACL_PyTorch/built-in/embedding/bge-m3/README.md +++ b/ACL_PyTorch/built-in/embedding/bge-m3/README.md @@ -29,6 +29,13 @@ # 快速上手 ## 获取源码 +1. 获取本仓源码 + + ``` + git clone https://gitee.com/ascend/ModelZoo-PyTorch.git + cd ModelZoo-PyTorch/ACL_PyTorch/built-in/embedding/bge-m + ``` + 1. 获取开源模型源码和权重（可选） > 如果您的设备可以方便的直接从hugging-hub下载权重和代码，则不需要执行这一步 ``` -- Gitee From f8d22dc67b85dc02e47ca5fb839146453ed4bb33 Mon Sep 17 00:00:00 2001 From: lifangtian Date: Wed, 4 Jun 2025 08:17:40 +0000 Subject: [PATCH 32/35] =?UTF-8?q?update=20ACL=5FPyTorch/built-in/embedding?= =?UTF-8?q?/bge-reranker-v2-m3/README.md=EF=BC=9A=E6=96=B0=E5=A2=9E?= =?UTF-8?q?=E8=8E=B7=E5=8F=96modelzoo=E4=BB=A3=E7=A0=81=E4=BB=93=E7=8E=AF?= =?UTF-8?q?=E8=8A=82?= MIME-Version: 1.0 Content-Type: text/plain; charset=UTF-8 Content-Transfer-Encoding: 8bit Signed-off-by: lifangtian --- .../built-in/embedding/bge-reranker-v2-m3/README.md | 7 +++++++ 1 file changed, 7 insertions(+) diff --git a/ACL_PyTorch/built-in/embedding/bge-reranker-v2-m3/README.md b/ACL_PyTorch/built-in/embedding/bge-reranker-v2-m3/README.md index e554011c5a..c215e5ceb8 100644 --- a/ACL_PyTorch/built-in/embedding/bge-reranker-v2-m3/README.md +++ b/ACL_PyTorch/built-in/embedding/bge-reranker-v2-m3/README.md @@ -29,6 +29,13 @@ # 快速上手 ## 获取源码 +1. 获取本仓源码 + + ``` + git clone https://gitee.com/ascend/ModelZoo-PyTorch.git + cd ModelZoo-PyTorch/ACL_PyTorch/built-in/embedding/bge-reranker-v2-m3 + ``` + 1. 获取开源模型源码和权重（可选） > 如果您的设备可以方便的直接从hugging-hub下载权重和代码，则不需要执行这一步 ``` -- Gitee From ba12eeca794c326bb76a58be7c0ba35e050e9c0c Mon Sep 17 00:00:00 2001 From: lifangtian Date: Wed, 4 Jun 2025 08:22:18 +0000 Subject: [PATCH 33/35] =?UTF-8?q?update=20ACL=5FPyTorch/built-in/embedding?= =?UTF-8?q?/jina-embeddings-v2-base-zh/README.md=EF=BC=9A=E6=96=B0?= =?UTF-8?q?=E5=A2=9E=E8=8E=B7=E5=8F=96modelzoo=E4=BB=A3=E7=A0=81=E4=BB=93?= =?UTF-8?q?=E7=8E=AF=E8=8A=82?= MIME-Version: 1.0 Content-Type: text/plain; charset=UTF-8 Content-Transfer-Encoding: 8bit Signed-off-by: lifangtian --- .../built-in/embedding/jina-embeddings-v2-base-zh/README.md | 6 ++++++ 1 file changed, 6 insertions(+) diff --git a/ACL_PyTorch/built-in/embedding/jina-embeddings-v2-base-zh/README.md b/ACL_PyTorch/built-in/embedding/jina-embeddings-v2-base-zh/README.md index 0e91701736..ccd3826fc5 100644 --- a/ACL_PyTorch/built-in/embedding/jina-embeddings-v2-base-zh/README.md +++ b/ACL_PyTorch/built-in/embedding/jina-embeddings-v2-base-zh/README.md @@ -38,6 +38,12 @@ # 快速上手 ## 获取源码 +1. 获取本仓源码 + + ``` + git clone https://gitee.com/ascend/ModelZoo-PyTorch.git + cd ModelZoo-PyTorch/ACL_PyTorch/built-in/embedding/jina-embeddings-v2-base-zh + ``` 1. 获取开源模型源码和权重（可选） > 如果您的设备可以方便的直接从hugging-hub下载权重和代码，则不需要执行这一步 -- Gitee From b6b140c6552d1dd245fb38fe7db6ffb34682b42c Mon Sep 17 00:00:00 2001 From: lifangtian Date: Wed, 4 Jun 2025 08:24:02 +0000 Subject: [PATCH 34/35] =?UTF-8?q?update=20ACL=5FPyTorch/built-in/cv/MuseTa?= =?UTF-8?q?lk/README.md=EF=BC=9A=E6=96=B0=E5=A2=9E=E8=8E=B7=E5=8F=96modelz?= =?UTF-8?q?oo=E4=BB=A3=E7=A0=81=E4=BB=93=E7=8E=AF=E8=8A=82?= MIME-Version: 1.0 Content-Type: text/plain; charset=UTF-8 Content-Transfer-Encoding: 8bit Signed-off-by: lifangtian --- ACL_PyTorch/built-in/cv/MuseTalk/README.md | 6 ++++++ 1 file changed, 6 insertions(+) diff --git a/ACL_PyTorch/built-in/cv/MuseTalk/README.md b/ACL_PyTorch/built-in/cv/MuseTalk/README.md index 61f16295c2..9f813493bb 100644 --- a/ACL_PyTorch/built-in/cv/MuseTalk/README.md +++ b/ACL_PyTorch/built-in/cv/MuseTalk/README.md @@ -35,6 +35,12 @@ # 快速上手 ## 获取源码 +1. 获取本仓源码 + + ``` + git clone https://gitee.com/ascend/ModelZoo-PyTorch.git + cd ModelZoo-PyTorch/ACL_PyTorch/built-in/cv/MuseTalk + ``` 1. 获取开源模型源码 ``` -- Gitee From 3b17e073ddbce3b831b8548fa98a24ccf00ce6ee Mon Sep 17 00:00:00 2001 From: lifangtian Date: Wed, 4 Jun 2025 08:25:20 +0000 Subject: [PATCH 35/35] =?UTF-8?q?update=20ACL=5FPyTorch/built-in/cv/Ground?= =?UTF-8?q?ingDINO/README.md=EF=BC=9A=E6=96=B0=E5=A2=9E=E8=8E=B7=E5=8F=96m?= =?UTF-8?q?odelzoo=E4=BB=A3=E7=A0=81=E4=BB=93=E7=8E=AF=E8=8A=82?= MIME-Version: 1.0 Content-Type: text/plain; charset=UTF-8 Content-Transfer-Encoding: 8bit Signed-off-by: lifangtian --- ACL_PyTorch/built-in/cv/GroundingDINO/README.md | 6 ++++++ 1 file changed, 6 insertions(+) diff --git a/ACL_PyTorch/built-in/cv/GroundingDINO/README.md b/ACL_PyTorch/built-in/cv/GroundingDINO/README.md index 89260cab1f..a912428a95 100644 --- a/ACL_PyTorch/built-in/cv/GroundingDINO/README.md +++ b/ACL_PyTorch/built-in/cv/GroundingDINO/README.md @@ -39,6 +39,12 @@ # 快速上手 ## 获取源码 +1. 获取本仓源码 + + ``` + git clone https://gitee.com/ascend/ModelZoo-PyTorch.git + cd ModelZoo-PyTorch/ACL_PyTorch/built-in/cv/GroundingDINO + ``` 1. 获取开源模型源码 ``` -- Gitee