True-Regularization+Finetune+Dynamic-Eval

Closed weights Mobvoi,Williams College 7M parameters April 2019

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
Mobvoi,Williams College
Organisation type
Industry,Academia
Country
China, United States of America
Published
8 April 2019
Authors
Yangyang Shi, Mei-Yuh Hwang, Xin Lei, Haoyu Sheng

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
7M
Training data
929,000 tokens

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
Unreleased

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

"In the first experiment, the student model achieves state-of-the-art perplexity results on the Penn Treebank dataset [1] with a model size one third of that of the previously published best model"

Record confidence
Confident
Citations
26
Benchmark data
True-Regularization+Finetune+Dynamic-Eval

Sources

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

Reference
Knowledge Distillation For Recurrent Neural Network Language Modeling With Trust Regularization
Last updated
25 May 2026

What the numbers mean

Where it came from

True-Regularization+Finetune+Dynamic-Eval was published by Mobvoi,Williams College, in China, in April 2019. It comes out of industry,Academia.

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

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

What went into building it

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

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

Answers

True-Regularization+Finetune+Dynamic-Eval — common questions

01

Who created True-Regularization+Finetune+Dynamic-Eval?

True-Regularization+Finetune+Dynamic-Eval was published by Mobvoi,Williams College, based in China, categorised as industry,Academia.

02

When was True-Regularization+Finetune+Dynamic-Eval released?

True-Regularization+Finetune+Dynamic-Eval was published in April 2019. 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 True-Regularization+Finetune+Dynamic-Eval used for?

True-Regularization+Finetune+Dynamic-Eval works in Language, and is recorded as handling language modeling. These are the areas it was designed around; they describe intent rather than a hard boundary.

04

What GPU do I need to run True-Regularization+Finetune+Dynamic-Eval?

None. True-Regularization+Finetune+Dynamic-Eval 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 True-Regularization+Finetune+Dynamic-Eval open source?

No. True-Regularization+Finetune+Dynamic-Eval has not had its weights published, so it exists only as a service controlled by its owner.

06

How many parameters does True-Regularization+Finetune+Dynamic-Eval have?

True-Regularization+Finetune+Dynamic-Eval has 7M 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.

Source

Original publication

Record last updated 25 May 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.