Zeming Lin ✓

@ebetica

Co-founder at EvolutionaryScale, now Biohub. ESM3 / ESMFold / PyTorch. Unsupervised learner but sometimes still gets a few rewards. Views are my own.

3 455Followers

536Following

2 557Posts total

18.4KViews on collected posts

Ultimi post

Zeming Lin ✓@ebetica17 Sep

@ebetica @nvidia ESMFold 50x faster than AlphaFold on proteins it has never seen. Stanford Steve is already building a wrapper that charges $49/mo for the same API endpoint. 124 views · 2 likes · 2 reposts · 1 replies Open on X →

Zeming Lin ✓@ebetica16 Sep

Tagging all the people that contributed to this: @faustomille, @imathur16, @KilianFatras from Biohub, Greg Zynda from @nvidia, @carrigmat from @huggingface and all the incredible reviewers who made the 600+ comments on the PR in transformers :D 305 views · 6 likes · 2 reposts · 0 replies Open on X →

Zeming Lin ✓@ebetica16 Sep

Spoilers: we have even better kernels on the way. Look forward soon to even faster folding kernels we worked on developing in the last few months. Check it out! Github: https://t.co/p9FiupLS6m Huggingface: https://t.co/7LdjDt2adO 445 views · 9 likes · 3 reposts · 2 replies Open on X →

Zeming Lin ✓@ebetica16 Sep

Note that existing users _must_ upgrade, or your installations will be broken!! Unfortunately we had to do this to properly integrate with Huggingface. 367 views · 3 likes · 1 reposts · 1 replies Open on X →

Zeming Lin ✓@ebetica16 Sep

Our fused kernels result in almost a 10x improvement in folding throughput. These were released without fanfare soon after release but as of v3.4.1 we officially support these kernels. We also now officially support FoldCP, allowing you to fold very large protein complexes. http

878 views · 23 likes · 3 reposts · 2 replies Open on X →

Zeming Lin ✓@ebetica16 Sep

Install it with `pip install esm` or `pip install transformers`! The huggingface version only supports protein modalities so we encourage you to use it directly through `esm`. What's even more fun is that with mlx support you can fold on your laptop! https://t.co/JnuYhbprne http

743 views · 12 likes · 1 reposts · 1 replies Open on X →

Zeming Lin ✓@ebetica16 Sep

The ESMC and ESMFold2 models are now on Huggingface! You can now install it directly from PyPi or from huggingface/transformers v5.16.0. As a part of this, we officially are releasing support for our fused Triton kernels and FoldCP in partnership with @nvidia! https://t.co/OVuVCJ

15.5K views · 217 likes · 30 reposts · 2 replies Open on X →

Rispetto ad account della stessa dimensione

7 post degli ultimi 90 giorni, accanto alla fascia di under 10K follower. raggiunge meno persone dei pari, ma le coinvolge molto di più.

Visualizzazioni mediane445questo account4 617mediana per under 10K

Copertura, %12.88%questo account277.07%mediana per under 10K

Interazione, %2.62%questo account1.41%mediana per under 10K

Metrica	Questo account	Mediana per under 10K	Rapporto
Visualizzazioni mediane per post	445	4 617	0.10×
Copertura (visualizzazioni ÷ follower)	12.88%	2.8× audience	0.05×
Tasso di interazione	2.62%	1.41%	1.86×

Altri account di questa fascia → Confronta con un altro account → Come sono costruiti questi parametri →

Growth & engagement

How the posts we collected actually performed: views and reaction rate post by post, what the audience did with them, and where the follower count goes.

Views per post

15.5K16 Sep

743

878

367

445

305

12417 Sep

Last 7 collected posts, oldest on the left. The scale is logarithmic: one post can outrun the rest a hundred times over.

Engagement rate per post

1.63%16 Sep

1.88%

3.19%

1.36%

3.15%

2.62%

4.03%17 Sep

Reactions — likes, reposts, replies and quotes — divided by views. Median for under 10K accounts is 1.41%.

What the audience does

Likes61.1%272 in total

Reposts9.4%42 in total

Replies2.0%9 in total

Quotes0.9%4 in total

Bookmarks26.5%118 in total

Share of every reaction we collected for this account. Replies mean argument, reposts mean endorsement, bookmarks mean the post was worth keeping.

The follower curve appears once this account has two daily snapshots — we take one a day, and this one is on its first.

Ultimi post

Rispetto ad account della stessa dimensione

Growth & engagement

Views per post

Engagement rate per post

What the audience does

Account simili