xyx361100238

Thanks for your reply! I'm sorry,still understand to generate decoder_input_ids & decoder_attention_mask,could you please give a example, or onnxruntime example with marian model!

got error： 


hello all： I use MARIANNMT have the same question： 1、train en-zh model according to examples/transformer 2、pre-process use jieba & bpe 3、done with train and test good 4、use convert_marian_to_pytorch.py to converted...

thanks to your all， it works! I got > [2022-08-26 07:08:09] [onnx] ONNX graph 'decode_first' written to model.onnx.decode_first.onnx [2022-08-26 07:08:09] [onnx] Exporting graph decode_next [2022-08-26 07:08:09] Error: Segmentation fault [2022-08-26...

@snukky have any plan to solve the error about “ Exporting graph decode_next”？ I located the problem, but I don't know how to modify it： > for (const auto& dss...

I'm not sure if this problem has been fixed. but I get three onnx model（encode_source、decode_first、decode_next）. Use tool [netron](https://netron.app/) found that decode_first outputs have no first_decoder_state_1&3&...  so modify [here](https://github.com/marian-nmt/marian/blob/6f6d484665dfb35fe1d85b49b6b9d29c2661fd15/src/onnx/expression_graph_onnx_exporter.cpp#:~:text=%7D-,for%20(const%20auto%26%20dss%20%3A%20extractStates(decodeFirstState)),-inputs.emplace_back) can...

It‘s not work in script marian_to_onnx_example.py： 

still need to fix null point： 

多谢，我可以先尝试采用http+Nebula实现；再请教可以推荐下http协议的C++库吗（或者开源），比自己写肯定要健壮、逻辑考虑也更全面些