Fraternal dropout + AWD-LSTM 3-layer (WT2)

Closed weights Jagiellonian University,Mila - Quebec AI (originally Montreal Institute for Learning Algorithms),University of Montreal / Université de Montréal 34M parameters October 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
Jagiellonian University,Mila - Quebec AI (originally Montreal Institute for Learning Algorithms),University of Montreal / Université de Montréal
Organisation type
Academia,Academia,Academia
Country
Poland, Canada
Published
31 October 2017
Authors
Konrad Zolna, Devansh Arpit, Dendi Suhubdy, Yoshua Bengio

What it does

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

Domain
Language
Task
Language modeling

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
34M

34M (Table 2)

Training data
2,000,000 tokens

code suggests 750 epochs https://github.com/kondiz/fraternal-dropout/blob/WT2/main.py

Epochs
750

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
3.1 × 10¹⁷ FLOP

6 FLOP / token / parameter * 34000000 parameters * 2000000 tokens * 750 epochs = 3.06e+17 FLOP _________________ In the Algorithmic Progress paper, the compute was estimated to be 9.85 × 10¹⁶ FLOP, assuming 520 epochs reported for the PTB dataset.

How it was established
Operation counting

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
Open source

BSD-3 license: https://github.com/kondiz/fraternal-dropout

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 evaluate our model and achieve state-of-the-art results in sequence modeling tasks on two benchmark datasets – Penn Treebank and Wikitext-2"

Record confidence
Likely
Citations
55
Benchmark data
Fraternal dropout + AWD-LSTM 3-layer (WT2)

Sources

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

Reference
Fraternal Dropout
Last updated
11 February 2026

What the numbers mean

What this model is

Fraternal dropout + AWD-LSTM 3-layer (WT2) was published by Jagiellonian University,Mila - Quebec AI (originally Montreal Institute for Learning Algorithms),University of Montreal / Université de Montréal, in the country recorded as Poland, during October 2017. The publishing organisation is categorised as academia,Academia,Academia.

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

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

How it was trained

Training it took a computation budget of roughly 3.1 × 10¹⁷ FLOP. That figure measures what producing the model cost, and has no bearing on how fast it answers.

Training consumed a corpus of around 2,000,000 tokens of text.

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

Answers

Fraternal dropout + AWD-LSTM 3-layer (WT2) — common questions

01

Fraternal dropout + AWD-LSTM 3-layer (WT2)— how much compute was used to train it?

Training consumed around 3.1 × 10¹⁷ FLOP. 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

Fraternal dropout + AWD-LSTM 3-layer (WT2)— 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.

03

Fraternal dropout + AWD-LSTM 3-layer (WT2)— is it open source?

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

04

Fraternal dropout + AWD-LSTM 3-layer (WT2)— how many parameters does it have?

It has a parameter count of 34M. 34M (Table 2). 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

Fraternal dropout + AWD-LSTM 3-layer (WT2)— who created it?

It was published by Jagiellonian University,Mila - Quebec AI (originally Montreal Institute for Learning Algorithms),University of Montreal / Université de Montréal, based in Poland, an organisation categorised as academia,Academia,Academia.

06

Fraternal dropout + AWD-LSTM 3-layer (WT2)— when was it released?

It was published in October 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.

07

Fraternal dropout + AWD-LSTM 3-layer (WT2)— what is it used for?

It works in the domain of Language, and is recorded as handling the task of language modeling. These are the areas it was designed around; they describe intent rather than a hard boundary.

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.