# Text-to-image generation on MJHQ-30K (Category FID/CLIP Subset)

Source: [Text-to-image generation on MJHQ-30K (Category FID/CLIP Subset)](<https://www.sota2.com/research/sota/text-to-image-generation-on-mjhq-30k-category-fid-clip-subset>)

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

Data updated: 2026-06-09T08:23:30.071Z

Evaluation of Text-to-Image generation models on the MJHQ-30K dataset, focusing specifically on FID and CLIP scores calculated across the Plants, Food, Landscape, and People categories, along with overall scores.

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 | Text-to-image generation |
| Dataset (including any stated split) | MJHQ-30K |
| Metric | FID (Plants) |
| Unit | SCALAR |
| Direction | Lower is better |
| Results loaded | 33 of 33 |
| 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 (Plants)](<https://www.sota2.com/research/sota/text-to-image-generation-on-mjhq-30k-category-fid-clip-subset.md?metric=FID%20(Plants)>): Lower is better; unit: SCALAR.

- [CLIP Score (Plants)](<https://www.sota2.com/research/sota/text-to-image-generation-on-mjhq-30k-category-fid-clip-subset.md?metric=CLIP%20Score%20(Plants)>): Higher is better; unit: SCALAR.

- [FID (Food)](<https://www.sota2.com/research/sota/text-to-image-generation-on-mjhq-30k-category-fid-clip-subset.md?metric=FID%20(Food)>): Lower is better; unit: SCALAR.

- [CLIP Score (Food)](<https://www.sota2.com/research/sota/text-to-image-generation-on-mjhq-30k-category-fid-clip-subset.md?metric=CLIP%20Score%20(Food)>): Higher is better; unit: SCALAR.

- [FID (Landscape)](<https://www.sota2.com/research/sota/text-to-image-generation-on-mjhq-30k-category-fid-clip-subset.md?metric=FID%20(Landscape)>): Lower is better; unit: SCALAR.

- [CLIP Score (Landscape)](<https://www.sota2.com/research/sota/text-to-image-generation-on-mjhq-30k-category-fid-clip-subset.md?metric=CLIP%20Score%20(Landscape)>): Higher is better; unit: SCALAR.

- [FID (People)](<https://www.sota2.com/research/sota/text-to-image-generation-on-mjhq-30k-category-fid-clip-subset.md?metric=FID%20(People)>): Lower is better; unit: SCALAR.

- [CLIP Score (People)](<https://www.sota2.com/research/sota/text-to-image-generation-on-mjhq-30k-category-fid-clip-subset.md?metric=CLIP%20Score%20(People)>): Higher is better; unit: SCALAR.

- [FID (Overall)](<https://www.sota2.com/research/sota/text-to-image-generation-on-mjhq-30k-category-fid-clip-subset.md?metric=FID%20(Overall)>): Lower is better; unit: SCALAR.

- [CLIP Score (Overall)](<https://www.sota2.com/research/sota/text-to-image-generation-on-mjhq-30k-category-fid-clip-subset.md?metric=CLIP%20Score%20(Overall)>): Higher is better; unit: SCALAR.

## Recorded results

| Method | FID (Plants) | Result date | Report paper |
| --- | --- | --- | --- |
| HACK++ | 26.58 | Not recorded | 6a279e4890966ade919bf932 |
| HACK | 27.59 | Not recorded | 6a279e4890966ade919bf932 |
| HACK | 28.72 | Not recorded | 6a279e4890966ade919bf932 |
| Full | 28.81 | Not recorded | 6a279e4890966ade919bf932 |
| HACK++ | 28.82 | Not recorded | 6a279e4890966ade919bf932 |
| ScaleKV | 28.97 | Not recorded | 6a279e4890966ade919bf932 |
| H2O | 29.55 | Not recorded | 6a279e4890966ade919bf932 |
| Streaming | 29.7 | Not recorded | 6a279e4890966ade919bf932 |
| HACK++ | 30.09 | Not recorded | 6a279e4890966ade919bf932 |
| SnapKV | 30.23 | Not recorded | 6a279e4890966ade919bf932 |
| Full | 30.47 | Not recorded | 6a279e4890966ade919bf932 |
| ScaleKV | 30.57 | Not recorded | 6a279e4890966ade919bf932 |
| CAKE | 30.74 | Not recorded | 6a279e4890966ade919bf932 |
| HACK++ | 30.78 | Not recorded | 6a279e4890966ade919bf932 |
| SnapKV | 30.83 | Not recorded | 6a279e4890966ade919bf932 |
| Full | 30.96 | Not recorded | 6a279e4890966ade919bf932 |
| HACK++ | 31.39 | Not recorded | 6a279e4890966ade919bf932 |
| H2O | 31.49 | Not recorded | 6a279e4890966ade919bf932 |
| ScaleKV | 31.53 | Not recorded | 6a279e4890966ade919bf932 |
| HACK | 31.54 | Not recorded | 6a279e4890966ade919bf932 |
| CAKE | 31.67 | Not recorded | 6a279e4890966ade919bf932 |
| SnapKV | 31.71 | Not recorded | 6a279e4890966ade919bf932 |
| H2O | 31.72 | Not recorded | 6a279e4890966ade919bf932 |
| Streaming | 32.13 | Not recorded | 6a279e4890966ade919bf932 |
| LOOK-M | 32.29 | Not recorded | 6a279e4890966ade919bf932 |
| MEDA | 32.47 | Not recorded | 6a279e4890966ade919bf932 |
| LOOK-M | 32.53 | Not recorded | 6a279e4890966ade919bf932 |
| Streaming | 32.59 | Not recorded | 6a279e4890966ade919bf932 |
| CAKE | 32.75 | Not recorded | 6a279e4890966ade919bf932 |
| ScaleKV | 32.87 | Not recorded | 6a279e4890966ade919bf932 |
| MEDA | 33.48 | Not recorded | 6a279e4890966ade919bf932 |
| LOOK-M | 44.69 | Not recorded | 6a279e4890966ade919bf932 |
| MEDA | 47.12 | Not recorded | 6a279e4890966ade919bf932 |

### HACK++

FID (Plants): 26.58. Result ID: 6a27ccccfbc373592c247ed5.

[Reporting paper](<https://www.sota2.com/research/paper/6a279e4890966ade919bf932.md>)

[Method origin paper](<https://www.sota2.com/research/paper/6a279e4890966ade919bf932.md>)

Recorded evaluation configuration: {"Backbone":"HART","ηa":"30%","ηc":"20%","KV Size":"0.61 GB","Attn. TFLOPs":"7.39"}. This record does not assert independent reproduction.

### HACK

FID (Plants): 27.59. Result ID: 6a27ccccfbc373592c247ed4.

[Reporting paper](<https://www.sota2.com/research/paper/6a279e4890966ade919bf932.md>)

[Method origin paper](<https://www.sota2.com/research/paper/68fa77ec1e209131cea760c8.md>)

Recorded evaluation configuration: {"Backbone":"HART","ηa":"30%","ηc":"30%","KV Size":"0.84 GB","Attn. TFLOPs":"7.40"}. This record does not assert independent reproduction.

### HACK

FID (Plants): 28.72. Result ID: 6a27ccccfbc373592c247ec8.

[Reporting paper](<https://www.sota2.com/research/paper/6a279e4890966ade919bf932.md>)

[Method origin paper](<https://www.sota2.com/research/paper/68fa77ec1e209131cea760c8.md>)

Recorded evaluation configuration: {"Backbone":"Infinity-8B","ηa":"30%","ηc":"30%","KV Size":"5.06 GB","Attn. TFLOPs":"33.54"}. This record does not assert independent reproduction.

### Full

FID (Plants): 28.81. Result ID: 6a27ccccfbc373592c247ec1.

[Reporting paper](<https://www.sota2.com/research/paper/6a279e4890966ade919bf932.md>)

Recorded evaluation configuration: {"Backbone":"Infinity-8B","ηa":"100%","ηc":"100%","KV Size":"16.86 GB","Attn. TFLOPs":"78.54"}. This record does not assert independent reproduction.

### HACK++

FID (Plants): 28.82. Result ID: 6a27ccccfbc373592c247ecb.

[Reporting paper](<https://www.sota2.com/research/paper/6a279e4890966ade919bf932.md>)

[Method origin paper](<https://www.sota2.com/research/paper/6a279e4890966ade919bf932.md>)

Recorded evaluation configuration: {"Backbone":"Infinity-8B","ηa":"30%","ηc":"10%","KV Size":"1.69 GB","Attn. TFLOPs":"31.53"}. This record does not assert independent reproduction.

### ScaleKV

FID (Plants): 28.97. Result ID: 6a27ccccfbc373592c247ec9.

[Reporting paper](<https://www.sota2.com/research/paper/6a279e4890966ade919bf932.md>)

[Method origin paper](<https://www.sota2.com/research/paper/68fa78c11e209131cea7a50a.md>)

Recorded evaluation configuration: {"Backbone":"Infinity-8B","ηc":"10%","KV Size":"1.69 GB","Attn. TFLOPs":"41.98"}. This record does not assert independent reproduction.

### H2O

FID (Plants): 29.55. Result ID: 6a27ccccfbc373592c247ec3.

[Reporting paper](<https://www.sota2.com/research/paper/6a279e4890966ade919bf932.md>)

[Method origin paper](<https://www.sota2.com/research/paper/68fa718c1e209131cea51ead.md>)

[Implementation](<https://github.com/FMInference/H2O>)

Recorded evaluation configuration: {"Backbone":"Infinity-8B","ηa":"30%","ηc":"30%","KV Size":"5.06 GB","Attn. TFLOPs":"33.55"}. This record does not assert independent reproduction.

### Streaming

FID (Plants): 29.7. Result ID: 6a27ccccfbc373592c247ec2.

[Reporting paper](<https://www.sota2.com/research/paper/6a279e4890966ade919bf932.md>)

[Method origin paper](<https://www.sota2.com/research/paper/68fa72d71e209131cea58d79.md>)

Recorded evaluation configuration: {"Backbone":"Infinity-8B","ηa":"30%","ηc":"30%","KV Size":"5.06 GB","Attn. TFLOPs":"33.55"}. This record does not assert independent reproduction.

### HACK++

FID (Plants): 30.09. Result ID: 6a27ccccfbc373592c247ecc.

[Reporting paper](<https://www.sota2.com/research/paper/6a279e4890966ade919bf932.md>)

[Method origin paper](<https://www.sota2.com/research/paper/6a279e4890966ade919bf932.md>)

Recorded evaluation configuration: {"Backbone":"Infinity-8B","ηa":"30%","ηc":"1%","KV Size":"0.17 GB","Attn. TFLOPs":"26.50"}. This record does not assert independent reproduction.

### SnapKV

FID (Plants): 30.23. Result ID: 6a27ccccfbc373592c247ec4.

[Reporting paper](<https://www.sota2.com/research/paper/6a279e4890966ade919bf932.md>)

[Method origin paper](<https://www.sota2.com/research/paper/68fa73b21e209131cea5df89.md>)

[Implementation](<https://github.com/FasterDecoding/SnapKV>)

Recorded evaluation configuration: {"Backbone":"Infinity-8B","ηa":"30%","ηc":"30%","KV Size":"5.06 GB","Attn. TFLOPs":"33.55"}. This record does not assert independent reproduction.

### Full

FID (Plants): 30.47. Result ID: 6a27ccccfbc373592c247ecd.

[Reporting paper](<https://www.sota2.com/research/paper/6a279e4890966ade919bf932.md>)

Recorded evaluation configuration: {"Backbone":"HART","ηa":"100%","ηc":"100%","KV Size":"2.68 GB","Attn. TFLOPs":"17.66"}. This record does not assert independent reproduction.

### ScaleKV

FID (Plants): 30.57. Result ID: 6a27ccccfbc373592c247eca.

[Reporting paper](<https://www.sota2.com/research/paper/6a279e4890966ade919bf932.md>)

[Method origin paper](<https://www.sota2.com/research/paper/68fa78c11e209131cea7a50a.md>)

Recorded evaluation configuration: {"Backbone":"Infinity-8B","ηc":"1%","KV Size":"0.17 GB","Attn. TFLOPs":"31.40"}. This record does not assert independent reproduction.

### CAKE

FID (Plants): 30.74. Result ID: 6a27ccccfbc373592c247ec6.

[Reporting paper](<https://www.sota2.com/research/paper/6a279e4890966ade919bf932.md>)

[Method origin paper](<https://www.sota2.com/research/paper/68fa77741e209131cea73352.md>)

Recorded evaluation configuration: {"Backbone":"Infinity-8B","ηa":"30%","ηc":"30%","KV Size":"5.06 GB","Attn. TFLOPs":"33.55"}. This record does not assert independent reproduction.

### HACK++

FID (Plants): 30.78. Result ID: 6a27ccccfbc373592c247ebf.

[Reporting paper](<https://www.sota2.com/research/paper/6a279e4890966ade919bf932.md>)

[Method origin paper](<https://www.sota2.com/research/paper/6a279e4890966ade919bf932.md>)

Recorded evaluation configuration: {"Backbone":"Infinity-2B","ηa":"30%","ηc":"10%","KV Size":"0.77 GB","Attn. TFLOPs":"14.32"}. This record does not assert independent reproduction.

### SnapKV

FID (Plants): 30.83. Result ID: 6a27ccccfbc373592c247eb8.

[Reporting paper](<https://www.sota2.com/research/paper/6a279e4890966ade919bf932.md>)

[Method origin paper](<https://www.sota2.com/research/paper/68fa73b21e209131cea5df89.md>)

[Implementation](<https://github.com/FasterDecoding/SnapKV>)

Recorded evaluation configuration: {"Backbone":"Infinity-2B","ηa":"30%","ηc":"30%","KV Size":"2.31 GB","Attn. TFLOPs":"15.34"}. This record does not assert independent reproduction.

### Full

FID (Plants): 30.96. Result ID: 6a27ccccfbc373592c247eb5.

[Reporting paper](<https://www.sota2.com/research/paper/6a279e4890966ade919bf932.md>)

Recorded evaluation configuration: {"Backbone":"Infinity-2B","ηa":"100%","ηc":"100%","KV Size":"7.71 GB","Attn. TFLOPs":"35.91"}. This record does not assert independent reproduction.

### HACK++

FID (Plants): 31.39. Result ID: 6a27ccccfbc373592c247ec0.

[Reporting paper](<https://www.sota2.com/research/paper/6a279e4890966ade919bf932.md>)

[Method origin paper](<https://www.sota2.com/research/paper/6a279e4890966ade919bf932.md>)

Recorded evaluation configuration: {"Backbone":"Infinity-2B","ηa":"30%","ηc":"1%","KV Size":"0.08 GB","Attn. TFLOPs":"12.11"}. This record does not assert independent reproduction.

### H2O

FID (Plants): 31.49. Result ID: 6a27ccccfbc373592c247ecf.

[Reporting paper](<https://www.sota2.com/research/paper/6a279e4890966ade919bf932.md>)

[Method origin paper](<https://www.sota2.com/research/paper/68fa718c1e209131cea51ead.md>)

[Implementation](<https://github.com/FMInference/H2O>)

Recorded evaluation configuration: {"Backbone":"HART","ηa":"30%","ηc":"30%","KV Size":"0.80 GB","Attn. TFLOPs":"7.40"}. This record does not assert independent reproduction.

### ScaleKV

FID (Plants): 31.53. Result ID: 6a27ccccfbc373592c247ebd.

[Reporting paper](<https://www.sota2.com/research/paper/6a279e4890966ade919bf932.md>)

[Method origin paper](<https://www.sota2.com/research/paper/68fa78c11e209131cea7a50a.md>)

Recorded evaluation configuration: {"Backbone":"Infinity-2B","ηc":"10%","KV Size":"0.77 GB","Attn. TFLOPs":"19.19"}. This record does not assert independent reproduction.

### HACK

FID (Plants): 31.54. Result ID: 6a27ccccfbc373592c247ebc.

[Reporting paper](<https://www.sota2.com/research/paper/6a279e4890966ade919bf932.md>)

[Method origin paper](<https://www.sota2.com/research/paper/68fa77ec1e209131cea760c8.md>)

Recorded evaluation configuration: {"Backbone":"Infinity-2B","ηa":"30%","ηc":"30%","KV Size":"2.31 GB","Attn. TFLOPs":"15.34"}. This record does not assert independent reproduction.

### CAKE

FID (Plants): 31.67. Result ID: 6a27ccccfbc373592c247eba.

[Reporting paper](<https://www.sota2.com/research/paper/6a279e4890966ade919bf932.md>)

[Method origin paper](<https://www.sota2.com/research/paper/68fa77741e209131cea73352.md>)

Recorded evaluation configuration: {"Backbone":"Infinity-2B","ηa":"30%","ηc":"30%","KV Size":"2.31 GB","Attn. TFLOPs":"15.34"}. This record does not assert independent reproduction.

### SnapKV

FID (Plants): 31.71. Result ID: 6a27ccccfbc373592c247ed0.

[Reporting paper](<https://www.sota2.com/research/paper/6a279e4890966ade919bf932.md>)

[Method origin paper](<https://www.sota2.com/research/paper/68fa73b21e209131cea5df89.md>)

[Implementation](<https://github.com/FasterDecoding/SnapKV>)

Recorded evaluation configuration: {"Backbone":"HART","ηa":"30%","ηc":"30%","KV Size":"0.80 GB","Attn. TFLOPs":"7.40"}. This record does not assert independent reproduction.

### H2O

FID (Plants): 31.72. Result ID: 6a27ccccfbc373592c247eb7.

[Reporting paper](<https://www.sota2.com/research/paper/6a279e4890966ade919bf932.md>)

[Method origin paper](<https://www.sota2.com/research/paper/68fa718c1e209131cea51ead.md>)

[Implementation](<https://github.com/FMInference/H2O>)

Recorded evaluation configuration: {"Backbone":"Infinity-2B","ηa":"30%","ηc":"30%","KV Size":"2.31 GB","Attn. TFLOPs":"15.34"}. This record does not assert independent reproduction.

### Streaming

FID (Plants): 32.13. Result ID: 6a27ccccfbc373592c247ece.

[Reporting paper](<https://www.sota2.com/research/paper/6a279e4890966ade919bf932.md>)

[Method origin paper](<https://www.sota2.com/research/paper/68fa72d71e209131cea58d79.md>)

Recorded evaluation configuration: {"Backbone":"HART","ηa":"30%","ηc":"30%","KV Size":"0.80 GB","Attn. TFLOPs":"7.40"}. This record does not assert independent reproduction.

### LOOK-M

FID (Plants): 32.29. Result ID: 6a27ccccfbc373592c247ec5.

[Reporting paper](<https://www.sota2.com/research/paper/6a279e4890966ade919bf932.md>)

[Method origin paper](<https://www.sota2.com/research/paper/68fa73d31e209131cea5e5a8.md>)

Recorded evaluation configuration: {"Backbone":"Infinity-8B","ηa":"30%","ηc":"30%","KV Size":"5.06 GB","Attn. TFLOPs":"33.55"}. This record does not assert independent reproduction.

### MEDA

FID (Plants): 32.47. Result ID: 6a27ccccfbc373592c247ebb.

[Reporting paper](<https://www.sota2.com/research/paper/6a279e4890966ade919bf932.md>)

[Method origin paper](<https://www.sota2.com/research/paper/68fa77611e209131cea72d66.md>)

Recorded evaluation configuration: {"Backbone":"Infinity-2B","ηa":"30%","ηc":"30%","KV Size":"2.31 GB","Attn. TFLOPs":"15.34"}. This record does not assert independent reproduction.

### LOOK-M

FID (Plants): 32.53. Result ID: 6a27ccccfbc373592c247eb9.

[Reporting paper](<https://www.sota2.com/research/paper/6a279e4890966ade919bf932.md>)

[Method origin paper](<https://www.sota2.com/research/paper/68fa73d31e209131cea5e5a8.md>)

Recorded evaluation configuration: {"Backbone":"Infinity-2B","ηa":"30%","ηc":"30%","KV Size":"2.31 GB","Attn. TFLOPs":"15.34"}. This record does not assert independent reproduction.

### Streaming

FID (Plants): 32.59. Result ID: 6a27ccccfbc373592c247eb6.

[Reporting paper](<https://www.sota2.com/research/paper/6a279e4890966ade919bf932.md>)

[Method origin paper](<https://www.sota2.com/research/paper/68fa72d71e209131cea58d79.md>)

Recorded evaluation configuration: {"Backbone":"Infinity-2B","ηa":"30%","ηc":"30%","KV Size":"2.31 GB","Attn. TFLOPs":"15.34"}. This record does not assert independent reproduction.

### CAKE

FID (Plants): 32.75. Result ID: 6a27ccccfbc373592c247ed2.

[Reporting paper](<https://www.sota2.com/research/paper/6a279e4890966ade919bf932.md>)

[Method origin paper](<https://www.sota2.com/research/paper/68fa77741e209131cea73352.md>)

Recorded evaluation configuration: {"Backbone":"HART","ηa":"30%","ηc":"30%","KV Size":"0.80 GB","Attn. TFLOPs":"7.40"}. This record does not assert independent reproduction.

### ScaleKV

FID (Plants): 32.87. Result ID: 6a27ccccfbc373592c247ebe.

[Reporting paper](<https://www.sota2.com/research/paper/6a279e4890966ade919bf932.md>)

[Method origin paper](<https://www.sota2.com/research/paper/68fa78c11e209131cea7a50a.md>)

Recorded evaluation configuration: {"Backbone":"Infinity-2B","ηc":"1%","KV Size":"0.08 GB","Attn. TFLOPs":"14.35"}. This record does not assert independent reproduction.

### MEDA

FID (Plants): 33.48. Result ID: 6a27ccccfbc373592c247ec7.

[Reporting paper](<https://www.sota2.com/research/paper/6a279e4890966ade919bf932.md>)

[Method origin paper](<https://www.sota2.com/research/paper/68fa77611e209131cea72d66.md>)

Recorded evaluation configuration: {"Backbone":"Infinity-8B","ηa":"30%","ηc":"30%","KV Size":"5.06 GB","Attn. TFLOPs":"33.55"}. This record does not assert independent reproduction.

### LOOK-M

FID (Plants): 44.69. Result ID: 6a27ccccfbc373592c247ed1.

[Reporting paper](<https://www.sota2.com/research/paper/6a279e4890966ade919bf932.md>)

[Method origin paper](<https://www.sota2.com/research/paper/68fa73d31e209131cea5e5a8.md>)

Recorded evaluation configuration: {"Backbone":"HART","ηa":"30%","ηc":"30%","KV Size":"0.80 GB","Attn. TFLOPs":"7.40"}. This record does not assert independent reproduction.

### MEDA

FID (Plants): 47.12. Result ID: 6a27ccccfbc373592c247ed3.

[Reporting paper](<https://www.sota2.com/research/paper/6a279e4890966ade919bf932.md>)

[Method origin paper](<https://www.sota2.com/research/paper/68fa77611e209131cea72d66.md>)

Recorded evaluation configuration: {"Backbone":"HART","ηa":"30%","ηc":"30%","KV Size":"0.80 GB","Attn. TFLOPs":"7.40"}. This record does not assert independent reproduction.

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