AR-LDM

Closed weights Alibaba,University of Waterloo,Vector Institute 1.5B parameters November 2022

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
Alibaba,University of Waterloo,Vector Institute
Organisation type
Industry,Academia,Academia
Country
China, Canada
Published
20 November 2022
Authors
Xichen Pan, Pengda Qin, Yuhong Li, Hui Xue, Wenhu Chen

What it does

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

Domain
Image generation
Task
Text-to-image
Base model
Stable Diffusion (LDM-KL-8-G)

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
1.5B

Table 1

Training data
tokens

PororoSV, FlintstonesSV and VIST. All storytelling datasets, sizes would be possible to look up.

Epochs
50

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
5.1 × 10²⁰ FLOP

8 NVIDIA A100 GPUs for 8 days

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 A100
Wall-clock time
194 hours (8.1 days)

8 NVIDIA A100 GPUs for 8 days

Compute cost
$746

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 (non-commercial)

no weights, no license. training and inference code here: https://github.com/xichenpan/ARLDM

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

The first latent diffusion model for coherent visual story synthesizing. "Quantitative results show that AR-LDM achieves SoTA FID scores on PororoSV, FlintstonesSV, and the newly introduced challenging dataset VIST containing natural images"

Record confidence
Confident
Citations
86

Sources

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

Reference
Synthesizing Coherent Story with Auto-Regressive Latent Diffusion Models
Last updated
25 May 2026

What the numbers mean

Background

AR-LDM was published by Alibaba,University of Waterloo,Vector Institute, in the country recorded as China, during November 2022. The category the publisher falls under is industry,Academia,Academia.

It works in the domain of Image generation, and is recorded as performing the task of text-to-image.

Rather than being trained from scratch, it is derived from Stable Diffusion (LDM-KL-8-G). Most models at this scale are adapted from an existing base rather than built from nothing.

This is a closed model: the trained values stayed with whoever produced them, and there is no local version to run.

Training and provenance

Producing it required arithmetic totalling around 5.1 × 10²⁰ FLOP, on hardware recorded as NVIDIA A100. That figure measures what producing the model cost, and has no bearing on how fast it answers.

Its inclusion criterion: sOTA improvement.

Answers

AR-LDM — common questions

01

AR-LDM— who created it?

It was published by Alibaba,University of Waterloo,Vector Institute, based in China, an organisation categorised as industry,Academia,Academia.

02

AR-LDM— when was it released?

It was published in November 2022. 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

AR-LDM— what is it used for?

It works in the domain of Image generation, and is recorded as handling the task of text-to-image. These are the areas it was designed around; they describe intent rather than a hard boundary.

04

AR-LDM— how much compute was used to train it?

Training consumed around 5.1 × 10²⁰ FLOP, on hardware recorded as NVIDIA A100. 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

AR-LDM— 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.

06

AR-LDM— is it open source?

No. Its weights have not been published, so it exists only as a service controlled by its owner.

07

AR-LDM— how many parameters does it have?

It has a parameter count of 1.5B. Table 1. 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.