ConvS2S (ensemble of 8 models)

Closed weights Meta AI July 2017

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
Meta AI
Organisation type
Industry
Country
United States of America
Published
25 July 2017
Authors
Jonas Gehring, Michael Auli, David Grangier, Denis Yarats, Yann N. Dauphin

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.

Training data
1,183,333,333 tokens

2.8M + 4.5M + 35.5M + 3.8M = 46.6M We consider three major WMT translation tasks as well as a text summarization task. WMT’16 English-Romanian. We use the same data and pre-processing as Sennrich et al. (2016b) but remove sentences with more than 175 words. This results in 2.8M sentence pairs for training and we evaluate on newstest2016.2 WMT’14 English-German. We use the same setup as Luong et al. (2015) which comprises 4.5M sentence pairs for training and we test on newstest2014. WMT’14 En…

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
5.6 × 10¹⁹ FLOP

All models are implemented in Torch (Collobert et al., 2011) and trained on a single Nvidia M40 GPU except for WMT’14 English-French for which we use a multi-GPU setup on a single machine. We train on up to eight GPUs synchronously by maintaining copies of the model on each card and split the batch so that each worker computes 1/8-th of the gradients; at the end we sum the gradients via Nvidia NCCL. 1. English-Romanian: "Training took between 6 and 7.5 days on a single GPU." 7 days * 24 * 3600 …

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 M40

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
Highly cited,SOTA improvement

"We achieve a new state of the art on several public translation benchmark data sets. On the WMT’16 EnglishRomanian task we outperform the previous best result by 1.9 BLEU, on WMT’14 English-French translation we improve over the LSTM model of Wu et al. (2016) by 1.6 BLEU in a comparable setting, and on WMT’14 EnglishGerman translation we ouperform the same model by 0.5 BLEU"

Record confidence
Likely

Sources

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

Reference
Convolutional Sequence to Sequence Learning
Last updated
11 February 2026

What the numbers mean

What this model is

ConvS2S (ensemble of 8 models) was published by Meta AI, in the country recorded as United States of America, during July 2017. The publishing organisation is categorised as industry.

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.

Training and provenance

Training it took a computation budget of roughly 5.6 × 10¹⁹ FLOP, on hardware recorded as NVIDIA M40. That figure measures what producing the model cost, and has no bearing on how fast it answers.

The training set ran to roughly 1,183,333,333 tokens of text.

It is tracked in the underlying dataset for one reason in particular: highly cited,SOTA improvement.

Answers

ConvS2S (ensemble of 8 models) — common questions

01

ConvS2S (ensemble of 8 models)— when was it released?

It was published in July 2017. 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

ConvS2S (ensemble of 8 models)— what is it used for?

It works in the domain of Language, and is recorded as handling the task of translation. Models frequently carry more than one of each, and the tags describe purpose rather than capability limits.

03

ConvS2S (ensemble of 8 models)— how much compute was used to train it?

Training consumed around 5.6 × 10¹⁹ FLOP, on hardware recorded as NVIDIA M40. 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.

04

ConvS2S (ensemble of 8 models)— 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.

05

ConvS2S (ensemble of 8 models)— is it open source?

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

06

ConvS2S (ensemble of 8 models)— how many parameters does it have?

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

07

ConvS2S (ensemble of 8 models)— who created it?

It was published by Meta AI, based in United States of America, an organisation categorised as industry.

Source

Original publication

Record last updated 11 February 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.