Adaptive Input Transformer + RD

Closed weights Microsoft Research Asia,Soochow University 247M parameters June 2021

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
Microsoft Research Asia,Soochow University
Organisation type
Industry,Academia
Country
China, Taiwan
Published
28 June 2021
Authors
Xiaobo Liang, Lijun Wu, Juntao Li, Yue Wang, Qi Meng, Tao Qin, Wei Chen, Min Zhang, Tie-Yan Liu

What it does

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

Domain
Language
Task
Language modeling
Numerical format
FP16

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
247M
Training data
103,000,000 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
8.6 × 10¹⁹ FLOP

"We train Transformer model for 50k steps and Adaptive Input Transformer for 286k steps" assuming 1 step took 1 second: 125000000000000 FLOP / sec [assumed precision: bf16] * 286000 seconds * 8 GPUs * 0.3 [assumed utilization] = 8.58e+19 FLOP

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 V100
Chips used
8
Wall-clock time
79 hours

" The training is on 8 Tesla V100 GPU cards" 286k steps ~ [assumption] 286000 seconds = 79.4 hours

Power draw
4.9 kW

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

train and inference code for translation models, MIT license. no weights https://github.com/dropreg/R-Drop/blob/main/fairseq_src/examples/translation_rdrop/README.md

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 particular, it yields substantial improvements when applied to fine-tune large-scale pre-trained models, e.g., ViT, RoBERTa-large, and BART, and achieves state-of-the-art (SOTA) performances with the vanilla Transformer model " I don't see absolute SOTA claimes for this model

Record confidence
Likely
Citations
538
Benchmark data
Adaptive Input Transformer + RD

Sources

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

Reference
R-Drop: Regularized Dropout for Neural Networks
Last updated
25 May 2026

What the numbers mean

Background

Adaptive Input Transformer + RD was published by Microsoft Research Asia,Soochow University, in China, in June 2021. It comes out of industry,Academia.

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

The training run consumed about 8.6 × 10¹⁹ FLOP, on NVIDIA V100. That figure describes the cost of creating it and has no bearing on how quickly it generates text.

It was trained on about 103,000,000 tokens of text.

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

Answers

Adaptive Input Transformer + RD — common questions

01

When was Adaptive Input Transformer + RD released?

Adaptive Input Transformer + RD was published in June 2021. 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 Adaptive Input Transformer + RD used for?

Adaptive Input Transformer + RD works in Language, and is recorded as handling language modeling. 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.

03

How much compute was used to train Adaptive Input Transformer + RD?

Around 8.6 × 10¹⁹ FLOP, on NVIDIA V100. 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 Adaptive Input Transformer + RD?

None. Adaptive Input Transformer + RD 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 Adaptive Input Transformer + RD open source?

No. Adaptive Input Transformer + RD has not had its weights published, so it exists only as a service controlled by its owner.

06

How many parameters does Adaptive Input Transformer + RD have?

Adaptive Input Transformer + RD has 247M 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.

07

Who created Adaptive Input Transformer + RD?

Adaptive Input Transformer + RD was published by Microsoft Research Asia,Soochow University, based in China, categorised as industry,Academia.

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.