ESM-GearNet

Closed weights Mila - Quebec AI (originally Montreal Institute for Learning Algorithms),University of Montreal / Université de Montréal,IBM Research,HEC Montreal,CIFAR AI Research 650M parameters May 2023

No estimate

No hardware requirements for this model

The weights for this model have not been published, so it cannot be downloaded or run on your own hardware at any size. It is reachable only through its provider, and no graphics card changes that.

On record

Full specification

Everything on record for this model. Most of it describes how it was trained rather than how it runs — useful context for judging how much work went into it, and how it compares with models built at a different scale.

Origin

Who built this model, where, and when it was published.

Organisation
Mila - Quebec AI (originally Montreal Institute for Learning Algorithms),University of Montreal / Université de Montréal,IBM Research,HEC Montreal,CIFAR AI Research
Organisation type
Academia,Academia,Industry,Academia,Research collective
Country
Canada, United States of America
Published
11 May 2023
Authors
Zuobai Zhang, Chuanrui Wang, Minghao Xu, Vijil Chenthamarakshan, Aurélie Lozano, Payel Das, Jian Tang

What it does

The problem areas the model was built for. A model can carry several of each.

Domain
Biology
Task
Proteins, Protein function prediction

Size

How large the model is and how much data it was trained on. Parameters are the figure that decides whether it fits on a given graphics card.

Parameters
650M
Training data
109,500,000 tokens

Calculating unique tokens seen in first epoch: Number of Proteins: 365,000 Average Residues per Protein: 300 Total Datapoints = 365,000 × 300 = 109,500,000 ≈ 1.1 × 10^8 tokens

Epochs
50

Training compute

The arithmetic performed to train the model, measured in floating-point operations. It is a measure of what the training run cost, not of how fast the finished model answers you.

Training compute
2.1 × 10¹⁹ FLOP

Using 6*N*D with 50 epochs: 6*650000000*110000001*50=2.145e+19

The training run

What it physically took to train: which chips, how many, for how long, and what that drew from the wall.

Training hardware
NVIDIA A100
Chips used
4
Power draw
3.2 kW

How it is classified

Labels the source dataset applies when tracking notable models, and how confident it is in the entry.

Record confidence
Likely
Citations
53

Sources

Where this record came from and when it was last checked.

Reference
A Systematic Study of Joint Representation Learning on Protein Sequences and Structures
Last updated
25 May 2026

What the numbers mean

About this model

ESM-GearNet was published by Mila - Quebec AI (originally Montreal Institute for Learning Algorithms),University of Montreal / Université de Montréal,IBM Research,HEC Montreal,CIFAR AI Research, in Canada, in May 2023. academia,Academia,Industry,Academia,Research collective is the category the publisher falls under.

It works in Biology, and is recorded as doing proteins, Protein function prediction.

Its weights were never published, so it can only be reached through its provider. No graphics card changes that.

Training and provenance

Training it took roughly 2.1 × 10¹⁹ FLOP of computation, on NVIDIA A100 — a measure of what producing the model cost, not of how fast it answers.

It was trained on about 109,500,000 tokens of text.

Answers

ESM-GearNet — common questions

01

How much compute was used to train ESM-GearNet?

Around 2.1 × 10¹⁹ FLOP, on NVIDIA A100. That measures what producing the model cost and says nothing about how quickly it answers once trained — inference speed comes from memory bandwidth, not from the training budget.

02

What GPU do I need to run ESM-GearNet?

None. ESM-GearNet is a closed model — its weights were never published, so it cannot be downloaded or run on your own hardware at any price. It is reachable only through its provider.

03

Is ESM-GearNet open source?

The licensing for ESM-GearNet was never recorded in our source data. We treat unstated licensing as closed, because an unrecorded licence is not one to rely on.

04

How many parameters does ESM-GearNet have?

ESM-GearNet has 650M parameters. That figure is the total, and it is what decides how much memory the model needs — roughly half a gigabyte per billion at the compression most people use.

05

Who created ESM-GearNet?

ESM-GearNet was published by Mila - Quebec AI (originally Montreal Institute for Learning Algorithms),University of Montreal / Université de Montréal,IBM Research,HEC Montreal,CIFAR AI Research, based in Canada, categorised as academia,Academia,Industry,Academia,Research collective.

06

When was ESM-GearNet released?

ESM-GearNet was published in May 2023. Capability per parameter has improved considerably since, so a newer model of the same size is often the better use of the same hardware.

07

What is ESM-GearNet used for?

ESM-GearNet works in Biology, and is recorded as handling proteins, Protein function prediction. These are the areas it was designed around; they describe intent rather than a hard boundary.

Source

Original publication

Record last updated 25 May 2026

The other direction

Looking at it from the other side?

This page starts from the model. If you already own a card and want to know everything it will run, start from the hardware instead.