sabersabersaber / transformer_ocr Goto Github PK

我在训练Mj ST两个数据集时，遇到了RuntimeError: expected device cuda:1 but got device cuda:0报错请问该如何修改呢？Original Traceback (most recent call last):
File "/home/will/anaconda3/envs/transocr/lib/python3.6/site-packages/torch/nn/parallel/parallel_apply.py", line 60, in _worker
output = module(*input, **kwargs)
File "/home/will/anaconda3/envs/transocr/lib/python3.6/site-packages/torch/nn/modules/module.py", line 541, in call
result = self.forward(*input, **kwargs)
File "/home/will/transformer/codes/transformer_OCR-main/model.py", line 129, in forward
prediction = self.Prediction(contextual_feature.contiguous(), text, is_train)
File "/home/will/anaconda3/envs/transocr/lib/python3.6/site-packages/torch/nn/modules/module.py", line 541, in call
result = self.forward(*input, **kwargs)
File "/home/will/transformer/codes/transformer_OCR-main/modules/transformer_util.py", line 376, in forward
enc_src = self.encoder(src, src_mask)
File "/home/will/anaconda3/envs/transocr/lib/python3.6/site-packages/torch/nn/modules/module.py", line 541, in call
result = self.forward(*input, **kwargs)
File "/home/will/transformer/codes/transformer_OCR-main/modules/transformer_util.py", line 49, in forward
src = self.dropout(((src) * self.scale) + self.pos_embedding(pos))
RuntimeError: expected device cuda:1 but got device cuda:0

请问作者有和论文的结果比较吗？

你好！
感谢你的开源代码。我跑了你的开源代码，但是平均准确率好像达不到你说的85%，请问你在训练过程中加了什么trick吗？

assert len(datasets) > 0, 'datasets should not be an empty iterable'  # type: ignore

AssertionError: datasets should not be an empty iterable

数据格式是