trajep comments

Results 7 comments of


                                            trajep

Can not export Huggingface GPT-2 to ONNX

Same issue for me. Anyone can help?

will it work on python 3

There is the answer [issues 35](https://github.com/aditya-grover/node2vec/issues/35).

Refactor/Pydantify nebula checkpoint engine config

> @trajepl if you are able to, can you take a look over this PR? LGTM, better to wait for Nebula local tests, then to release. Wei Luo will help...

ONNX files for T5 model with text2text-generation-with-past task do not work

Met the same issue, any progress?

ONNX files for T5 model with text2text-generation-with-past task do not work

Seems it is related with this PR. https://github.com/huggingface/optimum/pull/1257 It worked when I removed the merged decoder.

Cannot load LLaMA3 with transformers==4.57.1 on non-cuda accelerators using tensor parallelism (tp_plan="auto")

Hmm, it does not work. @Cyrilvallez Check this: This shallow copy(`a.data=a.data.to(xxx)`) fails because non-cuda device is not treated [as compatible with CPU in PyTorch’s TensorImpl logic. CUDA has legacy compatibility,...