CodonTransformer

Closed weights Vector Institute,University of Toronto,Université Paris Cité 89.6M parameters September 2024

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
Vector Institute,University of Toronto,Université Paris Cité
Organisation type
Academia,Academia,Academia
Country
Canada, France
Published
13 September 2024
Authors
Adibvafa Fallahpour, Vincent Gureghian, Guillaume J. Filion, Ariel B. Lindner, Amir Pandi

What it does

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

Domain
Biology
Task
Codon optimization

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
89.6M

"total number of parameters to 89.6 million."

Training data
45,053,865 tokens

1,001,197 sequences × 300 tokens/sequence = 300359100

Epochs
20

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 V100
Chips used
16
Power draw
9.5 kW

How it is classified

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

Record confidence
Confident

Sources

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

Reference
CodonTransformer: a multispecies codon optimizer using context-aware neural networks
Last updated
28 November 2025

What the numbers mean

Background

CodonTransformer was published by Vector Institute,University of Toronto,Université Paris Cité, in Canada, in September 2024. The organisation is categorised as academia,Academia,Academia.

It works in Biology, and is recorded as doing codon optimization.

Because the weights are not available, none of the hardware figures elsewhere on this site apply to it.

Training and provenance

The training set ran to roughly 45,053,865 tokens.

Answers

CodonTransformer — common questions

01

When was CodonTransformer released?

CodonTransformer was published in September 2024. Capability per parameter has improved considerably since, so a newer model of the same size is often the better use of the same hardware.

02

What is CodonTransformer used for?

CodonTransformer works in Biology, and is recorded as handling codon optimization. A model can carry several of each, so these are the areas it was built for rather than a limit on what it will attempt.

03

What GPU do I need to run CodonTransformer?

None. CodonTransformer 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.

04

Is CodonTransformer open source?

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

05

How many parameters does CodonTransformer have?

CodonTransformer has 89.6M parameters. "total number of parameters to 89.6 million.". 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.

06

Who created CodonTransformer?

CodonTransformer was published by Vector Institute,University of Toronto,Université Paris Cité, based in Canada, categorised as academia,Academia,Academia.

Source

Original publication

Record last updated 28 November 2025

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.