ERNIE-ViLG 2.0
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
- Baidu,Wuhan University of Science and Technology
- Organisation type
- Industry,Academia
- Country
- China
- Published
- 28 March 2023
- Authors
- Zhida Feng, Zhenyu Zhang, Xintong Yu, Yewei Fang, Lanxin Li, Xuyi Chen, Yuxiang Lu, Jiaxiang Liu, Weichong Yin, Shikun Feng, Yu Sun, Li Chen, Hao Tian, Hua Wu, Haifeng Wang
What it does
The problem areas the model was built for. A model can carry several of each.
- Domain
- Image generation
- Task
- Image generation
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
- 24B
- Training data
- 11,141,120,000,000 tokens
"we train ERNIEViLG 2.0 and scale up the model size to 24B parameters."
"The training data consists of 170M image-text pairs, including publicly available English datasets like LAION [28] and a series of internal Chinese datasets. The image autoencoder is trained on the same set. For images with English captions, we translate them with Baidu Translate API3 to get the Chinese version."
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
- 4.7 × 10²² FLOP
- How it was established
- Hardware
312000000000000 FLOP / GPU / sec [A100 reported, bf16 assumed]* 432 hours * 3600 sec / hour * 320 GPUs * 0.3 [assumed utilization] = 4.658135e+22 FLOP
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
- Chips used
- 320
- Wall-clock time
- 432 hours (18 days)
- Power draw
- 255.2 kW
"We train ERNIE-ViLG 2.0 on 320 Tesla A100 GPUs for 18 days." 18*24=432
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
- API access
- Training code
- Unreleased
https://wenxin.baidu.com/ernie-vilg
How it is classified
Labels the source dataset applies when tracking notable models, and how confident it is in the entry.
- Record confidence
- Confident
Sources
Where this record came from and when it was last checked.
- Reference
- ERNIE-ViLG 2.0: Improving Text-to-Image Diffusion Model with Knowledge-Enhanced Mixture-of-Denoising-Experts
- Last updated
- 28 November 2025
What the numbers mean
Background
ERNIE-ViLG 2.0 was published by Baidu,Wuhan University of Science and Technology, in the country recorded as China, during March 2023. It comes out of an organisation categorised as industry,Academia.
It works in the domain of Image generation, and is recorded as performing the task of image generation.
Its weights were never published, so it can only be reached through its provider. No graphics card changes that.
Training and provenance
The training run consumed about 4.7 × 10²² FLOP, on hardware recorded as NVIDIA A100. That figure measures what producing the model cost, and has no bearing on how fast it answers.
Training consumed a corpus of around 11,141,120,000,000 tokens of text.
Answers
ERNIE-ViLG 2.0 — common questions
ERNIE-ViLG 2.0— how many parameters does it have?
It has a parameter count of 24B. "we train ERNIEViLG 2.0 and scale up the model size to 24B 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.
ERNIE-ViLG 2.0— who created it?
It was published by Baidu,Wuhan University of Science and Technology, based in China, an organisation categorised as industry,Academia.
ERNIE-ViLG 2.0— when was it released?
It was published in March 2023. Capability per parameter has improved considerably since, so a newer model of the same size is often the better use of the same hardware.
ERNIE-ViLG 2.0— what is it used for?
It works in the domain of Image generation, and is recorded as handling the task of image generation. These are the areas it was designed around; they describe intent rather than a hard boundary.
ERNIE-ViLG 2.0— how much compute was used to train it?
Training consumed around 4.7 × 10²² 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.
ERNIE-ViLG 2.0— 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.
ERNIE-ViLG 2.0— is it open source?
No. Its weights have not been published, so it exists only as a service controlled by its owner.
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.