# Class-conditional image generation on ImageNet 1k 512x512 (train)

Source: [Class-conditional image generation on ImageNet 1k 512x512 (train)](<https://www.sota2.com/research/sota/class-conditional-image-generation-on-imagenet-1k-512x512-train>)

Collection index: [llms.txt](<https://www.sota2.com/research/llms.txt>)

Data updated: 2026-02-26T05:44:30.822Z

Evaluation of Class-Conditional Image Generation models on the training split of the ImageNet-1K dataset rendered at 512x512 resolution, focusing on FID scores, including a variant measured without classifier-free guidance (CFG).

A research result applies to its stated dataset, split, metric and evaluation configuration. Compare like-for-like settings, preserve the metric's unit and direction, and distinguish paper-reported results from independent reproduction. The best result recorded here is not a claim of exhaustive, current global coverage.

| Scope | Recorded value |
| --- | --- |
| Task | Class-conditional image generation |
| Dataset (including any stated split) | ImageNet 1k 512x512 (train) |
| Metric | FID |
| Unit | SCALAR |
| Direction | Lower is better |
| Results loaded | 7 of 7 |
| Coverage | All results currently returned for this benchmark |

Evaluation settings may differ between results. The original papers and recorded configurations are needed to establish comparability. A shared metric name alone does not establish an identical protocol.

## Metrics

- [FID](<https://www.sota2.com/research/sota/class-conditional-image-generation-on-imagenet-1k-512x512-train.md?metric=FID>): Lower is better; unit: SCALAR.

- [FID (w/o CFG)](<https://www.sota2.com/research/sota/class-conditional-image-generation-on-imagenet-1k-512x512-train.md?metric=FID%20(w%2Fo%20CFG)>): Lower is better; unit: SCALAR.

## Recorded results

| Method | FID | Result date | Report paper |
| --- | --- | --- | --- |
| DiMR-XL/3R | 2.89 | Not recorded | 68fa75de1e209131cea6a565 |
| DiT-XL/2 | 3.04 | Not recorded | 68fa75de1e209131cea6a565 |
| DiMR-XL/3R | 3.23 | Not recorded | 68fa75de1e209131cea6a565 |
| ADM-U | 3.85 | Not recorded | 68fa75de1e209131cea6a565 |
| U-ViT-H/4 | 4.05 | Not recorded | 68fa75de1e209131cea6a565 |
| U-ViT-L/4 | 4.67 | Not recorded | 68fa75de1e209131cea6a565 |
| ADM | 7.72 | Not recorded | 68fa75de1e209131cea6a565 |

### DiMR-XL/3R

FID: 2.89. Result ID: 6994042d3a6c3148491a6126.

[Reporting paper](<https://www.sota2.com/research/paper/68fa75de1e209131cea6a565.md>)

[Method origin paper](<https://www.sota2.com/research/paper/68fa75de1e209131cea6a565.md>)

[Implementation](<https://qihao067.github.io/projects/DiMR>)

Recorded evaluation configuration: {"Epoch":800,"#Params":"525M","Gflops":206}. This record does not assert independent reproduction.

### DiT-XL/2

FID: 3.04. Result ID: 6994042d3a6c3148491a6124.

[Reporting paper](<https://www.sota2.com/research/paper/68fa75de1e209131cea6a565.md>)

[Method origin paper](<https://www.sota2.com/research/paper/68fa6ea91e209131cea41328.md>)

Recorded evaluation configuration: {"Epoch":599,"#Params":"675M","Gflops":525}. This record does not assert independent reproduction.

### DiMR-XL/3R

FID: 3.23. Result ID: 6994042d3a6c3148491a6125.

[Reporting paper](<https://www.sota2.com/research/paper/68fa75de1e209131cea6a565.md>)

[Method origin paper](<https://www.sota2.com/research/paper/68fa75de1e209131cea6a565.md>)

[Implementation](<https://qihao067.github.io/projects/DiMR>)

Recorded evaluation configuration: {"Epoch":400,"#Params":"525M","Gflops":206}. This record does not assert independent reproduction.

### ADM-U

FID: 3.85. Result ID: 6994042d3a6c3148491a6121.

[Reporting paper](<https://www.sota2.com/research/paper/68fa75de1e209131cea6a565.md>)

[Method origin paper](<https://www.sota2.com/research/paper/68fa6a0e1e209131cea25dba.md>)

[Implementation](<https://github.com/openai/guided-diffusion>)

Recorded evaluation configuration: {"Epoch":1081,"#Params":"731M","Gflops":2813}. This record does not assert independent reproduction.

### U-ViT-H/4

FID: 4.05. Result ID: 6994042d3a6c3148491a6123.

[Reporting paper](<https://www.sota2.com/research/paper/68fa75de1e209131cea6a565.md>)

[Method origin paper](<https://www.sota2.com/research/paper/68fa6ee71e209131cea42caa.md>)

[Implementation](<https://github.com/baofff/U-ViT>)

Recorded evaluation configuration: {"Epoch":400,"#Params":"501M","Gflops":133}. This record does not assert independent reproduction.

### U-ViT-L/4

FID: 4.67. Result ID: 6994042d3a6c3148491a6122.

[Reporting paper](<https://www.sota2.com/research/paper/68fa75de1e209131cea6a565.md>)

[Method origin paper](<https://www.sota2.com/research/paper/68fa6ee71e209131cea42caa.md>)

[Implementation](<https://github.com/baofff/U-ViT>)

Recorded evaluation configuration: {"Epoch":400,"#Params":"287M","Gflops":77}. This record does not assert independent reproduction.

### ADM

FID: 7.72. Result ID: 6994042d3a6c3148491a6120.

[Reporting paper](<https://www.sota2.com/research/paper/68fa75de1e209131cea6a565.md>)

[Method origin paper](<https://www.sota2.com/research/paper/68fa6a0e1e209131cea25dba.md>)

[Implementation](<https://github.com/openai/guided-diffusion>)

Recorded evaluation configuration: {"#Params":"422M"}. This record does not assert independent reproduction.

Page 1 of 1. This is an index page, not the full collection.