Maciej (MJ) Mikulski

Sure that it can be limiting. But if it just a warning, it should be fine. Anyway, if loading data by `import` is going to be deprecated, then it may...

I guess that floydhub is not planning any further development. It's a pity, because it was a nice minimalist approach to the GPU cloud computing.

The problem that inspired this [questions](https://stackoverflow.com/q/75003385/8788960) was type-annotating a decorator that takes a function and returns a function with an attribute added (see [here](https://github.com/mjmikulski/horology/blob/master/horology/timed_decorator.py#L115)) in my hobby project [horology](https://github.com/mjmikulski/horology). Mypy...

Thanks for the clarification.

Thank you for the answer. How many GPUs did you use? AFAIK this was executed on a single node with 8 GPUs (H100).

### Additional info (as requested by @tfogal) If the line `model = thunder.jit(model)` in the above repro script is replaced with the following lines: ```py from thunder.dynamo import ThunderCompiler executors...

``` return torch._C._nn._upsample_bicubic2d_aa( NotImplementedError: The operator 'aten::_upsample_bicubic2d_aa.out' is not currently implemented for the MPS device. If you want this op to be considered for addition please comment on https://github.com/pytorch/pytorch/issues/141287 and...

How can `--no-lock` be used when the env is specified in yaml file? Following fails: ```sh $ conda env create --no-lock -f environment/environment.yaml usage: conda [-h] [-v] [--no-plugins] [-V] COMMAND...

Thanks for the hint. It worked with `CONDA_NO_LOCK`.