Logan Adams

This should now be fixed by this PR: https://github.com/microsoft/DeepSpeed/pull/3236 Can you try again and let us know if that fixes this issue?

I was able to repro this using your Dockerfile, but I do see this error in the logs: ``` #8 439.8 g++ -pthread -B /opt/conda/envs/dev/compiler_compat -shared -Wl,--allow-shlib-undefined -Wl,-rpath,/opt/conda/envs/dev/lib -Wl,-rpath-link,/opt/conda/envs/dev/lib -L/opt/conda/envs/dev/lib...

Also we do not have compatibility with triton 2.0.0 yet, so you may want to try building with `pip install triton==1.0.0` to see if that resolves any issues as well.

Yes, it should be fine to just use the JIT ops, pre-compiling isn't necessary. I'm not sure why it errors out yet, but wanted to be able to unblock in...

This should be resolved if you can try with the latest master branch

@l241025097 - this looks like a different issue than @chenyaofo's above. Could you open a new ticket for this? Since yours seems to be building enough to at least run...

@chenyaofo - do you have cuda-toolkit installed? Also since you're building on a node that isn't the one you're running on, so the capabilities of that are being used for...

@chenyaofo - I tried making my own dockerfile to test this, and I'm able to get the below working. I'm not familiar with the needs of your system, but I...

@ldilov - could you comment what errors you were seeing and how this fixes them?

It looks like we're not up to date with the pre-releases from pydantic, but things do work with their latest stable release, 1.10.7. We will work on getting things to...