phybrain

can you provide more details? did you use model.cuda()

数据集是生成的

应该阔以吧，有问题找我

效果差不多吧，Joint CTC-Attention based End-to-End Speech Recognition using Multi-task Learning结合下不知道怎样

there is no lstm

yes ofcourse

用solo

> Please, provide a fully runnable test case, not just pictures. > […](#) > Message ID: ***@***.***> ```python from docxtpl import DocxTemplate, InlineImage tpl = DocxTemplate('./test.docx') context = { 'repeat_purchase_list':...

> I confirm the bug and I got workaround. > > I selected the table and changed the style to Normal ([test Normal style.docx](https://github.com/elapouya/python-docx-template/files/13909394/test.Normal.style.docx)). The output document was as excpected....