flybird11111

Results 68 comments of flybird11111

[BUG]: Randomizer Raise Error When running on the CPU

Thank you for your contribution

sh train_rm.sh bug

Hi, can you pull the latest code? It seems that the naive strategy has been removed.

[BUG]: The placement policies between Bert example and the stable ColAI are not compatible

Thank you, we will fix this issue as soon as possible.

No module named 'colossalai.kernel.op_builder' Can not run resnet example

Can you downgrade cuda to 12.1?

No module named 'colossalai.kernel.op_builder' Can not run resnet example

I haven't reproduced your issue.This looks like you are missing some files. https://colossalai.org/docs/get_started/installation, can you try to install colossalai with the main branch of the repository.

No module named 'colossalai.kernel.op_builder' Can not run resnet example

Hi, could you please post the command you ran?

No module named 'colossalai.kernel.op_builder' Can not run resnet example

Hi, it seems something wrong with python 3.11, we will fix it soon, can you change the python version to python 3.10, It works in my environment.

No module named 'colossalai.kernel.op_builder' Can not run resnet example

export PATH=/usr/local/cuda/bin:${PATH}, how about adding it to ~/.bashrc

[BUG]: Assertion error when forward passing `nn.Embedding` without gradient.

Hi, this seems to be incompatible with the Gemini strategy. Indeed, encountering this issue with a frozen embedding table.

[shardformer] fix pipeline forward error if custom layer distribution is used

Thank you for your contribution.