Transformer + Simple Recurrent Unit

Closed weights ASAPP,Cornell University,Google,Princeton University 90M parameters September 2018

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
ASAPP,Cornell University,Google,Princeton University
Organisation type
Industry,Academia,Industry,Academia
Country
United States of America
Published
17 September 2018
Authors
Tao Lei, Yu Zhang, Sida I. Wang, Hui Dai, Yoav Artzi

What it does

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

Domain
Language
Task
Translation

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
90M

5-layer model, Table 3

Training data
112,500,000 tokens
Epochs
40

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
1.1 × 10¹⁹ FLOP

"We use a single NVIDIA Tesla V100 GPU for each model. The published results were obtained using 8 GPUs in parallel, which provide a large effective batch size during training. To approximate the setup, we update the model parameters every 5×5120 tokens and use 16,000 warm-up steps following OpenNMT suggestions. We train each model for 40 epochs (250,000 steps), and perform 3 independent trials for each model configuration. A single run takes about 3.5 days with a Tesla V100 GPU." 125 trillion …

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 V100
Chips used
8
Power draw
5.0 kW
Compute cost
$45

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
Closed — provider access only
Model access
Unreleased
Training code
Unreleased

repo, but no training code for translation: https://github.com/taolei87/sru

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

"We use the state-of-the-art Transformer model of Vaswani et al. (2017) as our base architecture... When SRU is incorporated into the architecture, both the 4-layer and 5-layer model outperform the Transformer base model"

Record confidence
Confident
Citations
306

Sources

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

Reference
Simple Recurrent Units for Highly Parallelizable Recurrence
Last updated
25 May 2026

What the numbers mean

Where it came from

Transformer + Simple Recurrent Unit was published by ASAPP,Cornell University,Google,Princeton University, in the country recorded as United States of America, during September 2018. The category the publisher falls under is industry,Academia,Industry,Academia.

It works in the domain of Language, and is recorded as performing the task of translation.

This is a closed model: the trained values stayed with whoever produced them, and there is no local version to run.

What went into building it

Training it took a computation budget of roughly 1.1 × 10¹⁹ FLOP, on hardware recorded as NVIDIA V100. That figure measures what producing the model cost, and has no bearing on how fast it answers.

It was trained on a corpus of about 112,500,000 tokens of text.

The reason it appears in this catalogue at all: sOTA improvement.

Answers

Transformer + Simple Recurrent Unit — common questions

01

Transformer + Simple Recurrent Unit— how many parameters does it have?

It has a parameter count of 90M. 5-layer model, Table 3. 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.

02

Transformer + Simple Recurrent Unit— who created it?

It was published by ASAPP,Cornell University,Google,Princeton University, based in United States of America, an organisation categorised as industry,Academia,Industry,Academia.

03

Transformer + Simple Recurrent Unit— when was it released?

It was published in September 2018. Capability per parameter has improved considerably since, so a newer model of the same size is often the better use of the same hardware.

04

Transformer + Simple Recurrent Unit— what is it used for?

It works in the domain of Language, and is recorded as handling the task of translation. 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.

05

Transformer + Simple Recurrent Unit— how much compute was used to train it?

Training consumed around 1.1 × 10¹⁹ FLOP, on hardware recorded as NVIDIA V100. 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.

06

Transformer + Simple Recurrent Unit— what GPU do I need to run it?

None. This 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.

07

Transformer + Simple Recurrent Unit— is it open source?

No. Its weights have not been published, so it exists only as a service controlled by its owner.

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.