Wang Zhang

> @zw0610 can we discuss first what the use case is and the solution could look like? > One question I foresee is how do we handle the indexes. So...

> Even if the index is meaningless, we need to clarify the expectations to users. It's also useful for us to have a clear algorithm to manage the indexes so...

Sure. I shall update the elastic proposal in mpi-operator.

> Should we discuss it in the Training WG community meeting? > > /cc @andreyvelich @zw0610 Definitely. I would like also hear more user cases and requests about running generic...

ok, let me take this one.

I believe the training scripts should be separated from the training image, ideally. As code, training scripts are better organized and maintained on github/gitlab/etc. But either hosted on a server...

@MartinForReal Post-processing a model looks more like what we can do within the container. In the training script, we can use python api to upload the model to s3 or...

I don't think so. It looks the v1 API drops MPIDistribution from v1alpha2.

Just the G-Docs link or we shall markdownize and push it to this repo?

Carried unanimously.