Alex Morehead

icon indicating a website link https://amorehead.github.io/ icon indicating an email [email protected]

icon company Lawrence Berkeley National Laboratory icon location Berkeley, California icon location Hopper Postdoctoral Fellow @BerkeleyLab. Prev: CS PhD @BioinfoMachineLearning Lab @ UMissouri; Research intern @Profluent-AI & @AbSciBio. Deep Learning & AI4Sci

Results 88 comments of Alex Morehead

Add learning rate scheduling support for `DeepSpeedStrategy`

@Borda, I've just fixed this test

Add learning rate scheduling support for `DeepSpeedStrategy`

@Borda, let's see if this latest commit of mine fixes it.

Add learning rate scheduling support for `DeepSpeedStrategy`

@Borda, may I ask for you to check the "Read the Docs" tests and why they are failing?

Add learning rate scheduling support for `DeepSpeedStrategy`

Thanks, @Borda!

Registered buffers not moved to correct device when using DeepSpeed Stage 3

Not yet.

How much time is required for making environment?

Mamba, in my experience, is the way to go.

latest version of rdkit not available

Yes, a new release would be greatly appreciated!

Issue with Cb atoms of threonines, perhaps other amino-acids?

@Toverkwark, thanks for the heads up. I've made a PR on your behalf at https://github.com/zrqiao/NeuralPLexer/pull/7.

pytest test/

I recommend referring to GitHub's CI pipeline to run these tests and inspect their outputs. For example: https://github.com/lucidrains/alphafold3-pytorch/actions/runs/10669195311/job/29570548592

size mismatch

I think the default decoder model configurations are out of sync with the ones we used during pretraining.