MLN-ASR

Closed weights McGill University 10K parameters August 1988

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
McGill University
Organisation type
Academia
Country
Canada
Published
1 August 1988
Authors
Renato De Mori, Yoshua Bengio, Régis Cardin

What it does

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

Domain
Speech
Task
Speech recognition (ASR)

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
10K

“For an MLN of about 10,000 links, the time was 115 CPU msecs for the recognition of a spoken letter and 317 msecs for the learning of a spoken letter on the SUN 4/280. A 20% reduction was obtained on the VAX 8650”

Training data
12,600 tokens

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 × 10⁸ FLOP

“For an MLN of about 10,000 links, the time was 115 CPU msecs for the recognition of a spoken letter and 317 msecs for the learning of a spoken letter on the SUN 4/280. A 20% reduction was obtained on the VAX 8650”, “Learning and recognition were performed on a VAX 8650.” Dataset: 70*10*2=1400 (Train) 10*10*2=200 (Test) “The ten words of the El set were pronounced twice by 80 speakers (40 males and 40 females)” “The data from 70 speakers were used as a training set while the data from the remain…

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.

Wall-clock time
0 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
Historical significance
Record confidence
Confident

Sources

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

Reference
Data-Driven Execution of Multi-Layered Networks for Automatic Speech Recognition
Last updated
28 November 2025

What the numbers mean

About this model

MLN-ASR was published by McGill University, in Canada, in August 1988. The organisation is categorised as academia.

It works in Speech, and is recorded as doing speech recognition (ASR).

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 roughly 3 × 10⁸ FLOP of computation — a measure of what producing the model cost, not of how fast it answers.

The training set ran to roughly 12,600 tokens.

The reason it appears in this catalogue at all is historical significance.

Answers

MLN-ASR — common questions

01

Who created MLN-ASR?

MLN-ASR was published by McGill University, based in Canada, categorised as academia.

02

When was MLN-ASR released?

MLN-ASR was published in August 1988. Capability per parameter has improved considerably since, so a newer model of the same size is often the better use of the same hardware.

03

What is MLN-ASR used for?

MLN-ASR works in Speech, and is recorded as handling speech recognition (ASR). 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.

04

How much compute was used to train MLN-ASR?

Around 3 × 10⁸ 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.

05

What GPU do I need to run MLN-ASR?

None. MLN-ASR 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.

06

Is MLN-ASR open source?

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

07

How many parameters does MLN-ASR have?

MLN-ASR has 10K parameters. “For an MLN of about 10,000 links, the time was 115 CPU msecs for the recognition of a spoken letter and 317 msecs for the learning of a spoken letter on the SUN 4/280. A 20% reduction was obtained on the VAX 8650”. 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.

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.