codeman008

Results 9 comments of codeman008

somebody can help me????Thanks very much!

谢谢大佬 ,我理解错了,我以为能文件夹批量检测

Have you solved this problem?I am currently studying such a project. May I communicate with you?

I have meet the same problem,have you solve it?

> > 骨干网络换成resnet50就可以了 以pytorch自带的ResNet为例 > > ``` > > def _forward_impl(self, x: Tensor) -> Tensor: > > # See note [TorchScript super()] > > x = self.conv1(x) > > x...

sorry, i solve it ! and i want to know how to quantitative training model? thank you so much

以上问题已经解决 主要是paddlepaddle相关的版本不对,重新安装版本之后,能加载数据但是报错: desc: 100%|███████████████████████████████████████████████████████████████████████████████████████████████████████████████████████████████████████████████████████████████████████| 132/132 [00:00

> 同[huggingface/transformers#24612 (comment)](https://github.com/huggingface/transformers/issues/24612#issuecomment-1618929505), 应该是你的transformers版本不对, [SLOW_TO_FAST_CONVERTERS mapping](https://github.com/huggingface/transformers/blob/f4e4b4d0e2dc248433e808594f7595292037d891/src/transformers/convert_slow_tokenizer.py#L1230)里边连'LayoutXLMTokenizer', 'LayoutLMv3Tokenizer'都没有 使用高一点的版本呢,我尝试了一下4.18.0,4.26.1都没问题 感谢大佬的解答,我将transformers版本重置到transformers==4.33.1问题解决,但报了一个其他的错误 ![image](https://github.com/yongzhuo/layoutlmv3-layoutxlm-chinese/assets/30332825/3c6ce041-eb47-4bf3-b733-971989a9c7a8) 不知您之前是否遇到过,对于您提交的项目,数据是不用在做操作对吧

![image](https://github.com/yongzhuo/layoutlmv3-layoutxlm-chinese/assets/30332825/990e77f5-2771-4fab-a8c7-a4a4b1edcae6) 感谢大佬的解答,已经成功跑起来了,想请问如果想训练CDLA数据集:https://github.com/buptlihang/CDLA,构造数据集按您提供的代码是否可以?