trRosetta

Open weights Nankai University,University of Washington,Tianjin University,Harvard University August 2019

No estimate

No hardware requirements for this model

This model's weights are open, but no parameter count has been published for it. Every memory and speed figure starts from that number, so we would rather show nothing than a fabricated estimate.

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
Nankai University,University of Washington,Tianjin University,Harvard University
Organisation type
Academia,Academia,Academia,Academia
Country
China, United States of America
Published
22 August 2019
Authors
Jianyi Yanga, Ivan Anishchenko, Hahnbeom Parkb, Zhenling Peng, Sergey Ovchinnikov, David Baker

What it does

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

Domain
Biology
Task
Protein folding 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.

Training data
tokens

"resulting in a set of 16,047 proteinchains with the average length of 250 amino acids " 16047*250 = 4011750 "Each trainingepoch runs through the whole training set, and 100 epochs are performed intotal."

Epochs
100

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
3.8 × 10¹⁹ FLOP

32620000000000 FLOP / GPU / sec [Nvidia Titan RTX, FP16 assumed] * 1 GPU * 1080 hours [see training time notes] * 3600 sec / hour * 0.3 [assumed utilization] = 3.8047968e+19 FLOP

How it was established
Hardware

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 TITAN RTX
Chips used
1
Wall-clock time
1,080 hours (45 days)

"We train 5 networks with random 95/5%training/validation splits and use the average over the 5 networks as the finalprediction. Training a single network takes ∼9 d on one NVIDIA Titan RTX GPU" 9 days * 24 hours * 5 networks = 1080 hours

Power draw
316 W

Availability

Whether you can obtain the model and run it on your own hardware, which is what decides if any of the graphics-card figures on this page apply.

Weights
Open — downloadable
Model access
Open weights (unrestricted)
Training code
Open source

MIT license

How it is classified

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

Why it is tracked
SOTA improvement

"On benchmarktests on 13th Community-Wide Experiment on the Critical Assess-ment of Techniques for Protein Structure Prediction (CASP13)-and Continuous Automated Model Evaluation (CAMEO)-derivedsets, the method outperforms all previously described structure-prediction methods."

Record confidence
Confident

Sources

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

Reference
Improved protein structure prediction using predictedinterresidue orientations
Last updated
28 November 2025

What the numbers mean

Background

trRosetta was published by Nankai University,University of Washington,Tianjin University,Harvard University, in China, in August 2019. It comes out of academia,Academia,Academia,Academia.

It works in Biology, and is recorded as doing protein folding prediction.

The weights are published, so it can be downloaded and run on your own hardware indefinitely, offline, with no account attached.

Training and provenance

Producing it required around 3.8 × 10¹⁹ FLOP of arithmetic, on NVIDIA TITAN RTX, which is a statement about the training budget rather than about inference.

Its inclusion criterion is sOTA improvement.

Answers

trRosetta — common questions

01

Is trRosetta open source?

Its weights are published, so trRosetta can be downloaded and run on your own hardware. Note that open weights is not the same as open source in the full sense — it says nothing about the training data, the training code, or the commercial terms attached.

02

How many parameters does trRosetta have?

No parameter count has been published for trRosetta, which is why no memory or speed figure appears on this page.

03

Who created trRosetta?

trRosetta was published by Nankai University,University of Washington,Tianjin University,Harvard University, based in China, categorised as academia,Academia,Academia,Academia.

04

When was trRosetta released?

trRosetta was published in August 2019. Capability per parameter has improved considerably since, so a newer model of the same size is often the better use of the same hardware.

05

What is trRosetta used for?

trRosetta works in Biology, and is recorded as handling protein folding prediction. Models frequently carry more than one of each, and the tags describe purpose rather than capability limits.

06

Where can I download trRosetta?

The weights for trRosetta are published, though we do not hold a repository link for it. This site calculates hardware requirements rather than hosting model files.

07

How much compute was used to train trRosetta?

Around 3.8 × 10¹⁹ FLOP, on NVIDIA TITAN RTX. 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.

08

What GPU do I need to run trRosetta?

We cannot say. trRosetta has open weights, but no parameter count has been published for it, and every memory and speed calculation starts from that number. We would rather show nothing than a fabricated estimate.

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.