# Mathematical Question Answering on MetaMathQA (final)

Source: [Mathematical Question Answering on MetaMathQA (final)](<https://www.sota2.com/research/sota/mathematical-question-answering-on-metamathqa-final>)

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

Data updated: 2026-04-08T07:50:18.263Z

Evaluation of model performance on the final test set of the MetaMathQA dataset for Mathematical Question Answering, measured using Accuracy.

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 | Mathematical Question Answering |
| Dataset (including any stated split) | MetaMathQA (final) |
| Metric | 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

- [Accuracy](<https://www.sota2.com/research/sota/mathematical-question-answering-on-metamathqa-final.md?metric=Accuracy>): Higher is better; unit: PERCENT.

## Recorded results

| Method | Accuracy | Result date | Report paper |
| --- | --- | --- | --- |
| MegaTrain | 92.52 | Not recorded | 69d5e1247d0172140ec75b40 |
| ZeRO-3 Offload | 92.41 | Not recorded | 69d5e1247d0172140ec75b40 |
| ZeRO Infinity | 92.36 | Not recorded | 69d5e1247d0172140ec75b40 |
| MegaTrain | 88.99 | Not recorded | 69d5e1247d0172140ec75b40 |
| ZeRO Infinity | 88.97 | Not recorded | 69d5e1247d0172140ec75b40 |
| ZeRO-3 Offload | 88.93 | Not recorded | 69d5e1247d0172140ec75b40 |
| PyTorch Native | 88.91 | Not recorded | 69d5e1247d0172140ec75b40 |
| Baseline | 37.58 | Not recorded | 69d5e1247d0172140ec75b40 |
| Baseline | 33.47 | Not recorded | 69d5e1247d0172140ec75b40 |

### MegaTrain

Accuracy: 92.52. Result ID: 69d608b047f34ce924be5786.

[Reporting paper](<https://www.sota2.com/research/paper/69d5e1247d0172140ec75b40.md>)

[Method origin paper](<https://www.sota2.com/research/paper/69d5e1247d0172140ec75b40.md>)

[Implementation](<https://github.com/DLYuanGod/MegaTrain>)

Recorded evaluation configuration: {"Model Scale":"14B"}. This record does not assert independent reproduction.

### ZeRO-3 Offload

Accuracy: 92.41. Result ID: 69d608b047f34ce924be5784.

[Reporting paper](<https://www.sota2.com/research/paper/69d5e1247d0172140ec75b40.md>)

[Method origin paper](<https://www.sota2.com/research/paper/68fa678f1e209131cea17034.md>)

Recorded evaluation configuration: {"Model Scale":"14B"}. This record does not assert independent reproduction.

### ZeRO Infinity

Accuracy: 92.36. Result ID: 69d608b047f34ce924be5785.

[Reporting paper](<https://www.sota2.com/research/paper/69d5e1247d0172140ec75b40.md>)

[Method origin paper](<https://www.sota2.com/research/paper/68fa69cd1e209131cea24271.md>)

Recorded evaluation configuration: {"Model Scale":"14B"}. This record does not assert independent reproduction.

### MegaTrain

Accuracy: 88.99. Result ID: 69d608b047f34ce924be5782.

[Reporting paper](<https://www.sota2.com/research/paper/69d5e1247d0172140ec75b40.md>)

[Method origin paper](<https://www.sota2.com/research/paper/69d5e1247d0172140ec75b40.md>)

[Implementation](<https://github.com/DLYuanGod/MegaTrain>)

Recorded evaluation configuration: {"Model Scale":"7B"}. This record does not assert independent reproduction.

### ZeRO Infinity

Accuracy: 88.97. Result ID: 69d608b047f34ce924be5780.

[Reporting paper](<https://www.sota2.com/research/paper/69d5e1247d0172140ec75b40.md>)

[Method origin paper](<https://www.sota2.com/research/paper/68fa69cd1e209131cea24271.md>)

Recorded evaluation configuration: {"Model Scale":"7B"}. This record does not assert independent reproduction.

### ZeRO-3 Offload

Accuracy: 88.93. Result ID: 69d608b047f34ce924be577f.

[Reporting paper](<https://www.sota2.com/research/paper/69d5e1247d0172140ec75b40.md>)

[Method origin paper](<https://www.sota2.com/research/paper/68fa678f1e209131cea17034.md>)

Recorded evaluation configuration: {"Model Scale":"7B"}. This record does not assert independent reproduction.

### PyTorch Native

Accuracy: 88.91. Result ID: 69d608b047f34ce924be5781.

[Reporting paper](<https://www.sota2.com/research/paper/69d5e1247d0172140ec75b40.md>)

Recorded evaluation configuration: {"Model Scale":"7B"}. This record does not assert independent reproduction.

### Baseline

Accuracy: 37.58. Result ID: 69d608b047f34ce924be5783.

[Reporting paper](<https://www.sota2.com/research/paper/69d5e1247d0172140ec75b40.md>)

Recorded evaluation configuration: {"Model Scale":"14B"}. This record does not assert independent reproduction.

### Baseline

Accuracy: 33.47. Result ID: 69d608b047f34ce924be577e.

[Reporting paper](<https://www.sota2.com/research/paper/69d5e1247d0172140ec75b40.md>)

Recorded evaluation configuration: {"Model Scale":"7B"}. This record does not assert independent reproduction.

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