PaddlePaddle / PaddlePaddle/Paddle
Cannot finetune en_ppstructure_mobile_v2.0_SLANet_train with paddlepaddle-GPU
Open
@FrostML is already working on this.
Since May 28, 2024.
- Dominant language
- C++
- Stars
- 24.1k
- Forks
- 6k
- Avg merge
- 4d 17h
- Merged PRs (30d)
- 60
Description
问题描述 / Problem Description
Finetuning en_ppstructure_mobile_v2.0_SLANet_train with paddlepaddle-GPU fails with error.
Finetuning with the CPU package works just fine.
运行环境 / Runtime Environment
- OS: Windows 10, Windows 11
- Paddle: 2.6.0, 2.6.1
- PaddleOCR: 2.7.5
- CUDA/CUDNN: 11.7/8.9.7, 12.0/8.9.7
复现代码 / Reproduction Code
Dataset and annotation structure: PubTabNet
Loss function: SLALoss
完整报错 / Complete Error Message
(InvalidArgument) Currently, the type of tensor in slice indices only allows int32 and int64, please check the type of index tensor. (at ..\paddle/fluid/pybind/slice_utils.h:70)
可能解决方案 / Possible solutions
Finetuning with the CPU package works just fine.
附件 / Appendix
Contributor guide
First steps
- Read the whole issue, then the project's contributing guide.
- Comment on the issue to say you are picking it up — it saves two people doing the same work.
- Fork the repository and make your change on a branch.
- Open a pull request that references the issue number.
Assessment
This issue has not been assessed yet.