Alex Morehead

Hi, @v-shaoningli. I'm glad to see you've found this work useful so far. As you said, `mmcif_parsing` relies initially on `Biopython` to parse the mmCIF input files' metadata, after which...

What's weird is, I've wrapped the `DetermineBonds` function (in Python) within a `timeout_decorator` of 3 seconds that should raise a `TimeoutException` if the function as a whole takes longer than...

Here's an updated full example of what my input pipeline for this function looks like. I've identified that, by trimming out atoms from the end of the NAP molecule, I...

Hey, all. Was this issue ever revisited?

*Another respectful bump

@senarvi, while testing out your callback in a local codebase of mine, I discovered an edge case that should be simple to address. Namely, when one is using a `LightningModule`'s...

Also, I've filled in the `EMAWeightAveraging` example into what may be a nice default callback configuration for users. Feel free to have a look: ```python import torch from torch.optim.swa_utils import...

@senarvi, you are right! I just noticed that the model will not be sharded unless Lightning's private method `lightning.pytorch.trainer.call._call_configure_model` is called instead of `pl_module.configure_model`. However, I don't think this will...

@senarvi, for the time being, it seems like the [reference implementation](https://github.com/pytorch/pytorch/blob/7b94d934313320eff91236540d37021d62bc8984/test/distributed/fsdp/test_fsdp_mixed_precision.py#L1353) mentioned in [this GitHub issue](https://github.com/pytorch/pytorch/issues/117742) is a standardized (PyTorch-tested) way of running EMA weight updates with FSDP. At first...

@lantiga, apologies. Just now getting to fixing this pull request up. I've updated the `docs/source-fabric/api/fabric_methods.rst` file. Are there any other relevant docs I've missed? I believe I've already updated all...