Fractional Max-Pooling

Closed weights University of Warwick 27M parameters December 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
University of Warwick
Organisation type
Academia
Country
United Kingdom of Great Britain and Northern Ireland
Published
18 December 2014
Authors
Benjamin Graham

What it does

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

Domain
Vision
Task
Image classification

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
27M

27M weights in largest CIFAR-100 model

Training data
901,200 tokens
Epochs
250

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 × 10¹⁷ FLOP

For the 12M param model, training required "18 hours on a GeForce GTX 780". So would be somewhat larger for 27M. 4 TFLOPS * 18 * 3600 * 0.4 = 1e17

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 GeForce GTX 780
Wall-clock time
18 hours

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

"for instance, we improve on the state-of-the art for CIFAR-100 without even using dropout."

Record confidence
Likely
Citations
672

Sources

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

Reference
Fractional Max-Pooling
Last updated
28 November 2025

What the numbers mean

Where it came from

Fractional Max-Pooling was published by University of Warwick, in the country recorded as United Kingdom of Great Britain and Northern Ireland, during December 2014. The publishing organisation is categorised as academia.

It works in the domain of Vision, and is recorded as performing the task of image classification.

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

What went into building it

The training run consumed about 1 × 10¹⁷ FLOP, on hardware recorded as NVIDIA GeForce GTX 780. That figure measures what producing the model cost, and has no bearing on how fast it answers.

The training set ran to roughly 901,200 tokens of text.

It is tracked in the underlying dataset for one reason in particular: sOTA improvement.

Answers

Fractional Max-Pooling — common questions

01

Fractional Max-Pooling— 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.

02

Fractional Max-Pooling— how many parameters does it have?

It has a parameter count of 27M. 27M weights in largest CIFAR-100 model. 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.

03

Fractional Max-Pooling— who created it?

It was published by University of Warwick, based in United Kingdom of Great Britain and Northern Ireland, an organisation categorised as academia.

04

Fractional Max-Pooling— when was it released?

It was published in December 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.

05

Fractional Max-Pooling— what is it used for?

It works in the domain of Vision, and is recorded as handling the task of image classification. 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.

06

Fractional Max-Pooling— how much compute was used to train it?

Training consumed around 1 × 10¹⁷ FLOP, on hardware recorded as NVIDIA GeForce GTX 780. 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.

07

Fractional Max-Pooling— 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.

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.