-
**PaddleDetection team appreciate any suggestion or problem you delivered~**
ppyoloe中df_loss中，pred_dist.shape==[n,4,17]与target_left.shape==[n,4]使用F.cross_entropy进行计算，是不是有问题？
## Checklist:
1. 查找历史…
-
## 描述问题/Describe the bug
使用paddleDetection fastercnn 训练自己的数据集，tool/train.py 训练能够正常完成 tool/eval.py 也能正常验证
但是tool/infer.py 测试数据时需要分配大量显存导致显存空间不足无法预测。
执行tools/eval.py能够正常执行log如下：
[12/28 20:23:3…
-
环境：aistudio， paddle2.1.2
运行命令：
!python /home/aistudio/PaddleSeg/deploy/python/infer.py \
--config /home/aistudio/output/deploy.yaml \
--image_path /home/aistudio/infer/images \
--save…
-
**PaddleDetection team appreciate any suggestion or problem you delivered~**
## Checklist:
1. 查找历史相关issue寻求解答/I have searched related issues but cannot get the expected help.
2. 翻阅[FAQ](https:/…
-
**PaddleDetection team appreciate any suggestion or problem you delivered~**
## Checklist:
1. 查找历史相关issue寻求解答/I have searched related issues but cannot get the expected help.
2. 翻阅[FAQ](https:/…
-
**PaddleDetection team appreciate any suggestion or problem you delivered~**
## Checklist:
1. 查找历史相关issue寻求解答/I have searched related issues but cannot get the expected help.
2. 翻阅[FAQ](https:/…
-
**PaddleDetection team appreciate any suggestion or problem you delivered~**
## Checklist:
1. 查找历史相关issue寻求解答/I have searched related issues but cannot get the expected help.
2. 翻阅[FAQ](https:/…
-
## 描述问题/Describe the bug
国产化环境使用XPU训练，训练几步后报错:
ioctl() fail, (705) Device memory not enough
batch_size设为1仍然不能正常训练。
xpu_smi大部分时间显示memory在4000 / 8064MB以内。
## 复现/Reproduction
python tools/train.py -c y…
-
**PaddleDetection team appreciate any suggestion or problem you delivered~**
## Checklist:
1. 查找历史相关issue寻求解答/I have searched related issues but cannot get the expected help.
2. 翻阅[FAQ](https:/…
-
**PaddleDetection team appreciate any suggestion or problem you delivered~**
## Checklist:
1. 查找历史相关issue寻求解答/I have searched related issues but cannot get the expected help.
2. 翻阅[FAQ](https:/…