Large regularized LSTM

Closed weights New York University (NYU),Google Brain 66M parameters September 2014

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
New York University (NYU),Google Brain
Organisation type
Academia,Industry
Country
United States of America
Published
8 September 2014
Authors
Wojciech Zaremba, Ilya Sutskever, Oriol Vinyals

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
66M

"The large LSTM has 1500 units per layer and its parameters are initialized uniformly in [−0.04, 0.04]" 66M according to Table 3 of https://arxiv.org/pdf/1611.01462

Training data
929,000 tokens

" We train the model for 55 epochs with a learning rate of 1"

Epochs
55

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
4.3 × 10¹⁶ FLOP

3520000000000 FLOP / s/ GPU [tesla k20c assumed, the paper says just "Nvidia k20"] * 1 GPU * 24 hours [could be less, they report "the entire day"] * 3600 s / hour * 0.3 [assumed utilization] = 9.12384e+16 FLOP 6 FLOP / parameter / token * 66000000 parameters * 929000 tokens * 55 epochs = 2.023362e+16 FLOP sqrt(2.023362e+16*9.12384e+16) = 4.2966069e+16

How it was established
Hardware,Operation counting

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 Tesla K20c
Chips used
1
Wall-clock time
24 hours

"Training this network takes an entire day on an NVIDIA K20 GPU."

Power draw
264 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
Closed — provider access only
Model access
Unreleased
Training code
Open source

Apache: https://github.com/wojzaremba/lstm train: https://github.com/wojzaremba/lstm/blob/master/main.lua

How it is classified

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

Record confidence
Confident
Citations
3,224
Benchmark data
Large regularized LSTM

Sources

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

Reference
Recurrent Neural Network Regularization
Last updated
28 November 2025

What the numbers mean

Where it came from

Large regularized LSTM was published by New York University (NYU),Google Brain, in United States of America, in September 2014. The organisation is categorised as academia,Industry.

It works in Language, and is recorded as doing language modeling.

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

Producing it required around 4.3 × 10¹⁶ FLOP of arithmetic, on NVIDIA Tesla K20c, which is a statement about the training budget rather than about inference.

It was trained on about 929,000 tokens of text.

Answers

Large regularized LSTM — common questions

01

When was Large regularized LSTM released?

Large regularized LSTM was published in September 2014. 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 Large regularized LSTM used for?

Large regularized LSTM works in Language, and is recorded as handling language modeling. Models frequently carry more than one of each, and the tags describe purpose rather than capability limits.

03

How much compute was used to train Large regularized LSTM?

Around 4.3 × 10¹⁶ FLOP, on NVIDIA Tesla K20c. 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

What GPU do I need to run Large regularized LSTM?

None. Large regularized LSTM 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

Is Large regularized LSTM open source?

No. Large regularized LSTM has not had its weights published, so it exists only as a service controlled by its owner.

06

How many parameters does Large regularized LSTM have?

Large regularized LSTM has 66M parameters. "The large LSTM has 1500 units per layer and its parameters are initialized uniformly in [−0.04, 0.04]" 66M according to Table 3 of https://arxiv.org/pdf/1611.01462. 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.

07

Who created Large regularized LSTM?

Large regularized LSTM was published by New York University (NYU),Google Brain, based in United States of America, categorised as academia,Industry.

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.