# Image Classification on ImageNet-1K 1.0 (val) (Top-1 Acc and Training Time)

Source: [Image Classification on ImageNet-1K 1.0 (val) (Top-1 Acc and Training Time)](<https://www.sota2.com/research/sota/image-classification-on-imagenet-1k-1-0-val-top-1-acc-and-training-time>)

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

Data updated: 2026-02-26T06:07:03.389Z

Standard evaluation of Image Classification performance on the ImageNet-1K validation set (v1.0), reporting Top-1 Accuracy (including an EMA variant) alongside the total Training time required.

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 | Image Classification |
| Dataset (including any stated split) | ImageNet-1K 1.0 (val) |
| Metric | Top-1 Accuracy |
| Unit | PERCENT |
| Direction | Higher is better |
| Results loaded | 9 of 9 |
| 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

- [Top-1 Accuracy](<https://www.sota2.com/research/sota/image-classification-on-imagenet-1k-1-0-val-top-1-acc-and-training-time.md?metric=Top-1%20Accuracy>): Higher is better; unit: PERCENT.

- [Top-1 Accuracy (EMA)](<https://www.sota2.com/research/sota/image-classification-on-imagenet-1k-1-0-val-top-1-acc-and-training-time.md?metric=Top-1%20Accuracy%20(EMA)>): Higher is better; unit: PERCENT.

- [Training Time (GPU hours)](<https://www.sota2.com/research/sota/image-classification-on-imagenet-1k-1-0-val-top-1-acc-and-training-time.md?metric=Training%20Time%20(GPU%20hours)>): Lower is better; unit: TIME.

## Recorded results

| Method | Top-1 Accuracy | Result date | Report paper |
| --- | --- | --- | --- |
| Baseline | 81.8 | Not recorded | 68fa72be1e209131cea58525 |
| ToE r1=0.5 | 81.6 | Not recorded | 68fa72be1e209131cea58525 |
| NetworkExpansion 6→12 | 81 | Not recorded | 68fa72be1e209131cea58525 |
| StackBERT | 80.8 | Not recorded | 68fa72be1e209131cea58525 |
| ToE r1=0.5 | 72.6 | Not recorded | 68fa72be1e209131cea58525 |
| Baseline | 72.2 | Not recorded | 68fa72be1e209131cea58525 |
| ToMe | 71.7 | Not recorded | 68fa72be1e209131cea58525 |
| NetworkExpansion 6→12 | 70.3 | Not recorded | 68fa72be1e209131cea58525 |
| S2ViTE | 70.1 | Not recorded | 68fa72be1e209131cea58525 |

### Baseline

Top-1 Accuracy: 81.8. Result ID: 699276663a6c3148490f1fa2.

[Reporting paper](<https://www.sota2.com/research/paper/68fa72be1e209131cea58525.md>)

[Method origin paper](<https://www.sota2.com/research/paper/68fa6a1d1e209131cea25fd2.md>)

[Implementation](<https://github.com/zihangJiang/TokenLabeling>)

Recorded evaluation configuration: {"Model":"DeiT-Base","Hyper-parameter consistency":true,"Architecture consistency":true,"Strategy consistency":true}. This record does not assert independent reproduction.

### ToE r1=0.5

Top-1 Accuracy: 81.6. Result ID: 699276663a6c3148490f1fa5.

[Reporting paper](<https://www.sota2.com/research/paper/68fa72be1e209131cea58525.md>)

[Method origin paper](<https://www.sota2.com/research/paper/68fa72be1e209131cea58525.md>)

[Implementation](<https://github.com/Osilly/TokenExpansion>)

Recorded evaluation configuration: {"Model":"DeiT-Base","Hyper-parameter consistency":true,"Architecture consistency":true,"Strategy consistency":true}. This record does not assert independent reproduction.

### NetworkExpansion 6→12

Top-1 Accuracy: 81. Result ID: 699276663a6c3148490f1fa4.

[Reporting paper](<https://www.sota2.com/research/paper/68fa72be1e209131cea58525.md>)

Recorded evaluation configuration: {"Model":"DeiT-Base","Hyper-parameter consistency":true,"Architecture consistency":false,"Strategy consistency":false}. This record does not assert independent reproduction.

### StackBERT

Top-1 Accuracy: 80.8. Result ID: 699276663a6c3148490f1fa3.

[Reporting paper](<https://www.sota2.com/research/paper/68fa72be1e209131cea58525.md>)

Recorded evaluation configuration: {"Model":"DeiT-Base","Hyper-parameter consistency":false,"Architecture consistency":true,"Strategy consistency":false}. This record does not assert independent reproduction.

### ToE r1=0.5

Top-1 Accuracy: 72.6. Result ID: 699276663a6c3148490f1fa1.

[Reporting paper](<https://www.sota2.com/research/paper/68fa72be1e209131cea58525.md>)

[Method origin paper](<https://www.sota2.com/research/paper/68fa72be1e209131cea58525.md>)

[Implementation](<https://github.com/Osilly/TokenExpansion>)

Recorded evaluation configuration: {"Model":"DeiT-Tiny","Hyper-parameter consistency":true,"Architecture consistency":true,"Strategy consistency":true}. This record does not assert independent reproduction.

### Baseline

Top-1 Accuracy: 72.2. Result ID: 699276663a6c3148490f1f9d.

[Reporting paper](<https://www.sota2.com/research/paper/68fa72be1e209131cea58525.md>)

[Method origin paper](<https://www.sota2.com/research/paper/68fa6a1d1e209131cea25fd2.md>)

[Implementation](<https://github.com/zihangJiang/TokenLabeling>)

Recorded evaluation configuration: {"Model":"DeiT-Tiny","Hyper-parameter consistency":true,"Architecture consistency":true,"Strategy consistency":true}. This record does not assert independent reproduction.

### ToMe

Top-1 Accuracy: 71.7. Result ID: 699276663a6c3148490f1f9f.

[Reporting paper](<https://www.sota2.com/research/paper/68fa72be1e209131cea58525.md>)

[Method origin paper](<https://www.sota2.com/research/paper/68fa6ea91e209131cea4130c.md>)

[Implementation](<http://github.com/facebookresearch/ToMe>)

Recorded evaluation configuration: {"Model":"DeiT-Tiny","Hyper-parameter consistency":false,"Architecture consistency":true,"Strategy consistency":false}. This record does not assert independent reproduction.

### NetworkExpansion 6→12

Top-1 Accuracy: 70.3. Result ID: 699276663a6c3148490f1fa0.

[Reporting paper](<https://www.sota2.com/research/paper/68fa72be1e209131cea58525.md>)

Recorded evaluation configuration: {"Model":"DeiT-Tiny","Hyper-parameter consistency":true,"Architecture consistency":false,"Strategy consistency":false}. This record does not assert independent reproduction.

### S2ViTE

Top-1 Accuracy: 70.1. Result ID: 699276663a6c3148490f1f9e.

[Reporting paper](<https://www.sota2.com/research/paper/68fa72be1e209131cea58525.md>)

[Method origin paper](<https://www.sota2.com/research/paper/68fa6b191e209131cea2c2a7.md>)

[Implementation](<https://github.com/VITA-Group/SViTE>)

Recorded evaluation configuration: {"Model":"DeiT-Tiny","Training epochs":600,"Hyper-parameter consistency":false,"Architecture consistency":false,"Strategy consistency":false}. This record does not assert independent reproduction.

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