From 7cbc46ae6be6df71b33eb1def9e428cab56126c6 Mon Sep 17 00:00:00 2001 From: lifangtian Date: Fri, 20 Jun 2025 07:08:00 +0000 Subject: [PATCH] =?UTF-8?q?update=20ACL=5FPyTorch/built-in/cv/SAM/README.m?= =?UTF-8?q?d=EF=BC=9A=E4=BF=AE=E6=AD=A3=E4=BA=86=E6=89=93patch=E5=91=BD?= =?UTF-8?q?=E4=BB=A4=E7=9A=84=E4=BD=8D=E7=BD=AE=E3=80=82=E5=A2=9E=E5=8A=A0?= =?UTF-8?q?=E5=AF=B9=E7=A6=BB=E7=BA=BF=E6=8E=A8=E7=90=86=E9=98=B6=E6=AE=B5?= =?UTF-8?q?=E5=8F=AF=E8=83=BD=E4=BA=A7=E7=94=9F=E7=9A=84=E9=94=99=E8=AF=AF?= =?UTF-8?q?=E7=9A=84=E8=A7=A3=E5=86=B3=E5=8A=9E=E6=B3=95?= 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 | 10 ++++++++-- 1 file changed, 8 insertions(+), 2 deletions(-) diff --git a/ACL_PyTorch/built-in/cv/SAM/README.md b/ACL_PyTorch/built-in/cv/SAM/README.md index d5a94032a9..7fee7963ca 100644 --- a/ACL_PyTorch/built-in/cv/SAM/README.md +++ b/ACL_PyTorch/built-in/cv/SAM/README.md @@ -78,9 +78,9 @@ 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 +patch -p2 < segment_anything_diff.patch pip3 install -e . cd .. -patch -p1 < segment_anything_diff.patch ``` ### 3.2 安装依赖。 @@ -236,7 +236,7 @@ python3 encoder_onnx_modify.py \ ### 3.5 推理验证 -1. 端到端推理。 +1. 端到端推理。成功执行下述命令后会在save-path参数指定的目录生成离线推理的结果。 ```bash python3 sam_end2end_infer.py \ @@ -256,6 +256,12 @@ python3 encoder_onnx_modify.py \ - decoder-model-path：decoder模型路径。 - input-point：分割目标上的坐标（坐标数量根据带分割目标物体分割效果确定，不同的图片数据，不同的待分割目标坐标不一样） - device-id：NPU卡ID + + + ``` + 若出现类似"OSError: .../python3.10/site-packages/torch/lib/libgomp-d22c30c5.so.1: cannot allocate memory in static TLS block"错误，可以使用export LD_PRELOAD=$LD_PRELOAD:<报错路径>解决。 + 例子：export LD_PRELOAD=$LD_PRELOAD:/root/anaconda3/envs/sam/lib/python3.10/site-packages/torch/lib/libgomp-d22c30c5.so.1 + ``` 在线模型推理结果： -- Gitee