# Dataset Condensation on CIFAR-100 (train)

Source: [Dataset Condensation on CIFAR-100 (train)](<https://www.sota2.com/research/sota/dataset-condensation-on-cifar-100-train>)

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

Data updated: 2026-02-26T06:07:34.524Z

Evaluation of Dataset Condensation performance on the training split of the CIFAR-100 dataset, focusing on training efficiency metrics like speed and memory usage.

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 | Dataset Condensation |
| Dataset (including any stated split) | CIFAR-100 (train) |
| Metric | Training Speed (s/iter) |
| Unit | TIME |
| Direction | Lower is better |
| Results loaded | 8 of 8 |
| 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

- [Training Speed (s/iter)](<https://www.sota2.com/research/sota/dataset-condensation-on-cifar-100-train.md?metric=Training%20Speed%20(s%2Fiter)>): Lower is better; unit: TIME.

- [Peak GPU Memory (GB)](<https://www.sota2.com/research/sota/dataset-condensation-on-cifar-100-train.md?metric=Peak%20GPU%20Memory%20(GB)>): Lower is better; unit: SIZE.

## Recorded results

| Method | Training Speed (s/iter) | Result date | Report paper |
| --- | --- | --- | --- |
| NCFM w/o ψ | 0.73 | Not recorded | 68fa77281e209131cea71713 |
| NCFM w/o ψ | 0.96 | Not recorded | 68fa77281e209131cea71713 |
| NCFM | 1.33 | Not recorded | 68fa77281e209131cea71713 |
| NCFM | 1.36 | Not recorded | 68fa77281e209131cea71713 |
| FTD | 1.68 | Not recorded | 68fa77281e209131cea71713 |
| MTT | 1.92 | Not recorded | 68fa77281e209131cea71713 |
| TESLA | 5.71 | Not recorded | 68fa77281e209131cea71713 |
| TESLA | 28.24 | Not recorded | 68fa77281e209131cea71713 |

### NCFM w/o ψ

Training Speed (s/iter): 0.73. Result ID: 69946b593a6c3148491d1b78.

[Reporting paper](<https://www.sota2.com/research/paper/68fa77281e209131cea71713.md>)

[Method origin paper](<https://www.sota2.com/research/paper/68fa77281e209131cea71713.md>)

Recorded evaluation configuration: {"IPC":10,"Sampling Network (psi)":false,"Hardware":"NVIDIA A100 80G"}. This record does not assert independent reproduction.

### NCFM w/o ψ

Training Speed (s/iter): 0.96. Result ID: 69946b593a6c3148491d1b79.

[Reporting paper](<https://www.sota2.com/research/paper/68fa77281e209131cea71713.md>)

[Method origin paper](<https://www.sota2.com/research/paper/68fa77281e209131cea71713.md>)

Recorded evaluation configuration: {"IPC":50,"Sampling Network (psi)":false,"Hardware":"NVIDIA A100 80G"}. This record does not assert independent reproduction.

### NCFM

Training Speed (s/iter): 1.33. Result ID: 69946b593a6c3148491d1b7a.

[Reporting paper](<https://www.sota2.com/research/paper/68fa77281e209131cea71713.md>)

[Method origin paper](<https://www.sota2.com/research/paper/68fa77281e209131cea71713.md>)

Recorded evaluation configuration: {"IPC":10,"Sampling Network (psi)":true,"Hardware":"NVIDIA A100 80G"}. This record does not assert independent reproduction.

### NCFM

Training Speed (s/iter): 1.36. Result ID: 69946b593a6c3148491d1b7b.

[Reporting paper](<https://www.sota2.com/research/paper/68fa77281e209131cea71713.md>)

[Method origin paper](<https://www.sota2.com/research/paper/68fa77281e209131cea71713.md>)

Recorded evaluation configuration: {"IPC":50,"Sampling Network (psi)":true,"Hardware":"NVIDIA A100 80G"}. This record does not assert independent reproduction.

### FTD

Training Speed (s/iter): 1.68. Result ID: 69946b593a6c3148491d1b75.

[Reporting paper](<https://www.sota2.com/research/paper/68fa77281e209131cea71713.md>)

[Method origin paper](<https://www.sota2.com/research/paper/68fa6ee81e209131cea42cd5.md>)

[Implementation](<https://github.com/AngusDujw/FTD-distillation>)

Recorded evaluation configuration: {"IPC":10,"Hardware":"NVIDIA A100 80G"}. This record does not assert independent reproduction.

### MTT

Training Speed (s/iter): 1.92. Result ID: 69946b593a6c3148491d1b74.

[Reporting paper](<https://www.sota2.com/research/paper/68fa77281e209131cea71713.md>)

[Method origin paper](<https://www.sota2.com/research/paper/68fa6c0d1e209131cea31ba0.md>)

Recorded evaluation configuration: {"IPC":10,"Hardware":"NVIDIA A100 80G"}. This record does not assert independent reproduction.

### TESLA

Training Speed (s/iter): 5.71. Result ID: 69946b593a6c3148491d1b76.

[Reporting paper](<https://www.sota2.com/research/paper/68fa77281e209131cea71713.md>)

[Method origin paper](<https://www.sota2.com/research/paper/68fa71111e209131cea4f081.md>)

Recorded evaluation configuration: {"IPC":10,"Hardware":"NVIDIA A100 80G"}. This record does not assert independent reproduction.

### TESLA

Training Speed (s/iter): 28.24. Result ID: 69946b593a6c3148491d1b77.

[Reporting paper](<https://www.sota2.com/research/paper/68fa77281e209131cea71713.md>)

[Method origin paper](<https://www.sota2.com/research/paper/68fa71111e209131cea4f081.md>)

Recorded evaluation configuration: {"IPC":50,"Hardware":"NVIDIA A100 80G"}. This record does not assert independent reproduction.

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