-
It seems that ```VisionTransformer``` doesn't support feature extraction of all outputs in the ```forward_features``` method. Only returning of the cls token or [cls_token, distillation_token] is avai…
-
用动态图的方式编写的MobileNetV3, 之后使用ProgramTranslator转换成静态图,用opt转换好的模型,在华为手机上,调用npu运行,不论输入什么数据,输出结果都是一样的(被转换成nb文件的静态图模型,在电脑上可以正常推理),请问可能是哪里出了问题?谢谢!
此外,使用官方提供的mobilenetv1静态图模型转换得到的nb文件在手机上可以正常推理。
版本号:
paddle…
-
Hi, I noticed that you have also provided results from FedADMM in your published work.
Can you provide solver scripts for that algorithm?
-
在rk平台上,使用量化后的mobilenetv3_db模型推理,设置线程数为2,能正常运行;设置线程数为4,出现malloc(): corrupted top size错误。
加载模型时,设置线程数为4:
std::shared_ptr loadModel(std::string model_file) {
MobileConfig config;
config.s…
-
Hi Tobias,
I'm trying to implement a segmentation model with mobilenetv3 (tensorflow mobilenetv3_large minimalistic) with a LR-ASPP segmentation head that I trained in python. When converting my mo…
-
# 【PaddlePaddle Hackathon】Paddle-Lite 任务合集
hi,大家好,非常高兴的告诉大家,首届 PaddlePaddle Hackathon 开始啦。PaddlePaddle Hackathon 是面向全球开发者的深度学习领域编程活动,鼓励开发者了解与参与 PaddlePaddle。本次共有四大方向(PaddlePaddle、Paddle Family、Padd…
-
请提供下述完整信息以便快速定位问题/Please provide the following information to quickly locate the problem
- 系统环境/System Environment:Win10
- 版本号/Version:Paddle: PaddleOCR: 问题相关组件/Related components:
- 运行指令/Comman…
-
问题: 训练模型时,在打印完日志‘During the training process, after the 0th iteration, an evaluation is run every 400 iterations’后,没有动静了(超过半个小时没有任何其他日志输出),需要如何排查问题。
- 系统环境/System Environment:windons 11
- 版本号/Ve…
-
anyone knows how to fix this issue or is it a problem in the training (even though the model is still training)?
i am getting this ppocr WARNING:
[2024/04/05 00:45:30] ppocr WARNING: The pretrain…
-
The code of backbone mobilenetv3 has register , but when I use the mobilenetv3, I have this error :
KeyError: "No object named 'build_mobilenetv3_backbone' found in 'BACKBONE' registry!"