From 41688ed2bd614aa087c3a58130588eceba1e3491 Mon Sep 17 00:00:00 2001 From: =?UTF-8?q?=E5=BD=AD=E4=B8=9A=E5=BA=86?= Date: Sat, 30 Oct 2021 02:33:31 +0000 Subject: [PATCH 1/3] =?UTF-8?q?update=20Ascend-PyTorch=E7=A6=BB=E7=BA=BF?= =?UTF-8?q?=E6=8E=A8=E7=90=86=E6=8C=87=E5=AF=BC/ONNX=E6=A8=A1=E5=9E=8B?= =?UTF-8?q?=E6=8E=A8=E7=90=86=E6=8C=87=E5=AF=BC/official/cv/ReID/ReID-stro?= =?UTF-8?q?ng-baseline/README.md.?= MIME-Version: 1.0 Content-Type: text/plain; charset=UTF-8 Content-Transfer-Encoding: 8bit --- .../official/cv/ReID/ReID-strong-baseline/README.md" | 4 ++-- 1 file changed, 2 insertions(+), 2 deletions(-) diff --git "a/Ascend-PyTorch\347\246\273\347\272\277\346\216\250\347\220\206\346\214\207\345\257\274/ONNX\346\250\241\345\236\213\346\216\250\347\220\206\346\214\207\345\257\274/official/cv/ReID/ReID-strong-baseline/README.md" "b/Ascend-PyTorch\347\246\273\347\272\277\346\216\250\347\220\206\346\214\207\345\257\274/ONNX\346\250\241\345\236\213\346\216\250\347\220\206\346\214\207\345\257\274/official/cv/ReID/ReID-strong-baseline/README.md" index 29173be..fdcccc0 100644 --- "a/Ascend-PyTorch\347\246\273\347\272\277\346\216\250\347\220\206\346\214\207\345\257\274/ONNX\346\250\241\345\236\213\346\216\250\347\220\206\346\214\207\345\257\274/official/cv/ReID/ReID-strong-baseline/README.md" +++ "b/Ascend-PyTorch\347\246\273\347\272\277\346\216\250\347\220\206\346\214\207\345\257\274/ONNX\346\250\241\345\236\213\346\216\250\347\220\206\346\214\207\345\257\274/official/cv/ReID/ReID-strong-baseline/README.md" @@ -24,11 +24,11 @@ cd .. 3.获取权重文件 -[market_resnet50_model_120_rank1_945.pth](https://drive.google.com/open?id=1hn0sXLZ5yJcxtmuY-ItQfYD7hBtHwt7A) +将权重文件[market_resnet50_model_120_rank1_945.pth](https://drive.google.com/open?id=1hn0sXLZ5yJcxtmuY-ItQfYD7hBtHwt7A)放到当前目录 ~~优先使用训练提供的权重文件,如果训练的权重文件网上能获则需给出网址,否则需要给出从哪获取权重文件。如果训练没有提供权重则使用开源代码仓基础配置的权重文件。需要给出权重文件名~~ 4.数据集 -[获取Market1501](http://www.liangzheng.org/Project/project_reid.html),并重命名为market1501 +[获取Market1501](http://www.liangzheng.org/Project/project_reid.html),并重命名为market1501,放到/root/datasets目录 5.[获取benchmark工具](https://gitee.com/ascend/cann-benchmark/tree/master/infer) 将benchmark.x86_64或benchmark.aarch64放到当前目录 -- Gitee From 53f90bd0a92711b22f6818d488f4d8544ab3a534 Mon Sep 17 00:00:00 2001 From: =?UTF-8?q?=E5=BD=AD=E4=B8=9A=E5=BA=86?= Date: Sat, 30 Oct 2021 02:35:11 +0000 Subject: [PATCH 2/3] =?UTF-8?q?update=20Ascend-PyTorch=E7=A6=BB=E7=BA=BF?= =?UTF-8?q?=E6=8E=A8=E7=90=86=E6=8C=87=E5=AF=BC/ONNX=E6=A8=A1=E5=9E=8B?= =?UTF-8?q?=E6=8E=A8=E7=90=86=E6=8C=87=E5=AF=BC/official/cv/ReID/ReID-stro?= =?UTF-8?q?ng-baseline/README.md.?= MIME-Version: 1.0 Content-Type: text/plain; charset=UTF-8 Content-Transfer-Encoding: 8bit --- .../official/cv/ReID/ReID-strong-baseline/README.md" | 4 ++-- 1 file changed, 2 insertions(+), 2 deletions(-) diff --git "a/Ascend-PyTorch\347\246\273\347\272\277\346\216\250\347\220\206\346\214\207\345\257\274/ONNX\346\250\241\345\236\213\346\216\250\347\220\206\346\214\207\345\257\274/official/cv/ReID/ReID-strong-baseline/README.md" "b/Ascend-PyTorch\347\246\273\347\272\277\346\216\250\347\220\206\346\214\207\345\257\274/ONNX\346\250\241\345\236\213\346\216\250\347\220\206\346\214\207\345\257\274/official/cv/ReID/ReID-strong-baseline/README.md" index fdcccc0..90b0df0 100644 --- "a/Ascend-PyTorch\347\246\273\347\272\277\346\216\250\347\220\206\346\214\207\345\257\274/ONNX\346\250\241\345\236\213\346\216\250\347\220\206\346\214\207\345\257\274/official/cv/ReID/ReID-strong-baseline/README.md" +++ "b/Ascend-PyTorch\347\246\273\347\272\277\346\216\250\347\220\206\346\214\207\345\257\274/ONNX\346\250\241\345\236\213\346\216\250\347\220\206\346\214\207\345\257\274/official/cv/ReID/ReID-strong-baseline/README.md" @@ -24,14 +24,14 @@ cd .. 3.获取权重文件 -将权重文件[market_resnet50_model_120_rank1_945.pth](https://drive.google.com/open?id=1hn0sXLZ5yJcxtmuY-ItQfYD7hBtHwt7A)放到当前目录 +将权重文件[market_resnet50_model_120_rank1_945.pth](https://drive.google.com/open?id=1hn0sXLZ5yJcxtmuY-ItQfYD7hBtHwt7A)放到当前工作目录 ~~优先使用训练提供的权重文件,如果训练的权重文件网上能获则需给出网址,否则需要给出从哪获取权重文件。如果训练没有提供权重则使用开源代码仓基础配置的权重文件。需要给出权重文件名~~ 4.数据集 [获取Market1501](http://www.liangzheng.org/Project/project_reid.html),并重命名为market1501,放到/root/datasets目录 5.[获取benchmark工具](https://gitee.com/ascend/cann-benchmark/tree/master/infer) -将benchmark.x86_64或benchmark.aarch64放到当前目录 +将benchmark.x86_64或benchmark.aarch64放到当前工作目录 ## 2 离线推理 -- Gitee From 54cd29fdcae25b366422b9b9388534e41d3bf17f Mon Sep 17 00:00:00 2001 From: =?UTF-8?q?=E5=BD=AD=E4=B8=9A=E5=BA=86?= Date: Sat, 30 Oct 2021 02:38:01 +0000 Subject: [PATCH 3/3] =?UTF-8?q?update=20Ascend-PyTorch=E7=A6=BB=E7=BA=BF?= =?UTF-8?q?=E6=8E=A8=E7=90=86=E6=8C=87=E5=AF=BC/PyTorch=E7=A6=BB=E7=BA=BF?= =?UTF-8?q?=E6=8E=A8=E7=90=86-=E4=BA=A4=E4=BB=98=E6=A0=87=E5=87=86?= =?UTF-8?q?=E4=B8=8E=E4=BA=A4=E4=BB=98=E4=BB=B6.md.?= MIME-Version: 1.0 Content-Type: text/plain; charset=UTF-8 Content-Transfer-Encoding: 8bit --- ...207\206\344\270\216\344\272\244\344\273\230\344\273\266.md" | 3 ++- 1 file changed, 2 insertions(+), 1 deletion(-) diff --git "a/Ascend-PyTorch\347\246\273\347\272\277\346\216\250\347\220\206\346\214\207\345\257\274/PyTorch\347\246\273\347\272\277\346\216\250\347\220\206-\344\272\244\344\273\230\346\240\207\345\207\206\344\270\216\344\272\244\344\273\230\344\273\266.md" "b/Ascend-PyTorch\347\246\273\347\272\277\346\216\250\347\220\206\346\214\207\345\257\274/PyTorch\347\246\273\347\272\277\346\216\250\347\220\206-\344\272\244\344\273\230\346\240\207\345\207\206\344\270\216\344\272\244\344\273\230\344\273\266.md" index 6541d8d..ccbb77c 100644 --- "a/Ascend-PyTorch\347\246\273\347\272\277\346\216\250\347\220\206\346\214\207\345\257\274/PyTorch\347\246\273\347\272\277\346\216\250\347\220\206-\344\272\244\344\273\230\346\240\207\345\207\206\344\270\216\344\272\244\344\273\230\344\273\266.md" +++ "b/Ascend-PyTorch\347\246\273\347\272\277\346\216\250\347\220\206\346\214\207\345\257\274/PyTorch\347\246\273\347\272\277\346\216\250\347\220\206-\344\272\244\344\273\230\346\240\207\345\207\206\344\270\216\344\272\244\344\273\230\344\273\266.md" @@ -169,7 +169,8 @@ 2.如果是无法规避的算子缺陷导致性能不达标,这里需要简要添加性能不达标的原因与解决方案。 3.如果onnx因包含自定义算子不支持推理,需要说明性能是在基准环境测的在线推理 4.如果模型不支持batch 16,也需要说明一下 - 5.模型推理中遇到的主要问题点等其它特殊情况也可以在此备注 + 5.如果onnx不支持动态batch,也需要说明一下 + 6.模型推理中遇到的主要问题点等其它特殊情况也可以在此备注 自验报告 # 第x次验收测试 -- Gitee