RNN LM

Closed weights Johns Hopkins University 70.3M parameters September 2010

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
Johns Hopkins University
Organisation type
Academia
Country
United States of America
Published
26 September 2010
Authors
Tomas Mikolov, M. Karafiát, L. Burget, J. Černocký, S. Khudanpur

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
70.3M

This database entry refers to the 3xRNN rows in Table 2 (static and dynamic likely use the same model ensemble, but allow the model weights to update once when testing the dynamic version). I assume the 3xRNN represents interpolation between the three largest models shown explicitly (RNN 250/5, RNN 250/2, and RNN 400/10). This seems likely, since smaller models do considerably worse on their own. In the following colab notebook, I estimate vocabulary sizes for the NYT Gigaword data at around 5…

Training data
6,400,000 tokens

As it is very time consuming to train RNN LM on large data, we have used only up to 6.4M words for training RNN models.

Epochs
1

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
5.4 × 10¹⁶ FLOP

"Convergence is usually achieved after 10-20 epochs." Training was done over a 6.4M subset of the NYT section of English Gigaword. 6 * 70,265,000 * 20 * 6.4M = 5.396e16

How it was established
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.

Wall-clock time
504 hours (21 days)

"it takes several weeks to train the most complex models." Rough guess, 3 weeks = 504 hours Assuming these models trained at the same time on different machines.

How it is classified

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

Frontier model
Yes
Why it is tracked
Highly cited
Record confidence
Speculative
Citations
6,038

Sources

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

Reference
Recurrent neural network based language model
Last updated
1 January 2026

What the numbers mean

Where it came from

RNN LM was published by Johns Hopkins University, in United States of America, in September 2010. academia is the category the publisher falls under.

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

Its weights were never published, so it can only be reached through its provider. No graphics card changes that.

How it was trained

Producing it required around 5.4 × 10¹⁶ FLOP of arithmetic, which is a statement about the training budget rather than about inference.

Around 6,400,000 tokens went into training it.

Its inclusion criterion is highly cited.

Answers

RNN LM — common questions

01

How much compute was used to train RNN LM?

Around 5.4 × 10¹⁶ 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

What GPU do I need to run RNN LM?

None. RNN LM 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

Is RNN LM open source?

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

04

How many parameters does RNN LM have?

RNN LM has 70.3M parameters. This database entry refers to the 3xRNN rows in Table 2 (static and dynamic likely use the same model ensemble, but allow the model weights to update once when testing the dynamic version). I assume the 3xRNN represents interpolation between the three largest models shown explicitly (RNN 250/5, RNN 250/2, and RNN 400/10). This seems likely, since smaller models do considerably worse on their own. In the following colab notebook, I estimate vocabulary sizes for the NYT Gigaword data at around 54.4k, 41.4k, and 27.6k for merge thresholds of 2, 5, and 10, respectively: https://colab.research.google.com/drive/1K5qH0EqXtFwTLESNtp4oelCM28GpGXt6#scrollTo=tedUkbgklNJ3 So the total number of parameters in each constituent model is: - RNN 250/2: (250 + 54.4k) * 250 + (250 * 54.4k) = 27,262,500 - RNN 250/5: (250 + 41.4k) * 250 + (250 * 41.4k) = 20,762,500 - RNN 400/10: (400 + 27.6k) * 400 + (400 * 27.6k) = 22,240,000 In total: 70,265,000 parameters. 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

Who created RNN LM?

RNN LM was published by Johns Hopkins University, based in United States of America, categorised as academia.

06

When was RNN LM released?

RNN LM was published in September 2010. 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

What is RNN LM used for?

RNN LM 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.

Source

Original publication

Record last updated 1 January 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.