# Pretraining on Llama pretraining 124M

Source: [Pretraining on Llama pretraining 124M](<https://www.sota2.com/research/sota/pretraining-on-llama-pretraining-124m>)

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

Data updated: 2026-05-22T08:57:22.376Z

Evaluation of resource efficiency and throughput metrics (iteration time, throughput) during the pretraining phase for the Llama 124M model variant.

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 | Pretraining |
| Dataset (including any stated split) | Llama pretraining 124M |
| Metric | Iteration Time (s) |
| Unit | TIME |
| Direction | Lower is better |
| Results loaded | 4 of 4 |
| 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

- [Iteration Time (s)](<https://www.sota2.com/research/sota/pretraining-on-llama-pretraining-124m.md?metric=Iteration%20Time%20(s)>): Lower is better; unit: TIME.

- [Relative Time](<https://www.sota2.com/research/sota/pretraining-on-llama-pretraining-124m.md?metric=Relative%20Time>): Lower is better; unit: SCALAR.

- [Throughput (tokens/s)](<https://www.sota2.com/research/sota/pretraining-on-llama-pretraining-124m.md?metric=Throughput%20(tokens%2Fs)>): Higher is better; unit: SCALAR.

## Recorded results

| Method | Iteration Time (s) | Result date | Report paper |
| --- | --- | --- | --- |
| AdamW | 0.674 | Not recorded | 6a0fe30af984739ed6739d6f |
| SF-AdamW | 0.684 | Not recorded | 6a0fe30af984739ed6739d6f |
| Muon | 0.706 | Not recorded | 6a0fe30af984739ed6739d6f |
| AMUSE | 0.719 | Not recorded | 6a0fe30af984739ed6739d6f |

### AdamW

Iteration Time (s): 0.674. Result ID: 6a101a68fbc373592c094347.

[Reporting paper](<https://www.sota2.com/research/paper/6a0fe30af984739ed6739d6f.md>)

[Method origin paper](<https://www.sota2.com/research/paper/68fa55761e209131ce9946db.md>)

Recorded evaluation configuration: Not recorded. This record does not assert independent reproduction.

### SF-AdamW

Iteration Time (s): 0.684. Result ID: 6a101a68fbc373592c094348.

[Reporting paper](<https://www.sota2.com/research/paper/6a0fe30af984739ed6739d6f.md>)

[Method origin paper](<https://www.sota2.com/research/paper/68fa5c811e209131ce9d176a.md>)

Recorded evaluation configuration: Not recorded. This record does not assert independent reproduction.

### Muon

Iteration Time (s): 0.706. Result ID: 6a101a68fbc373592c094349.

[Reporting paper](<https://www.sota2.com/research/paper/6a0fe30af984739ed6739d6f.md>)

Recorded evaluation configuration: Not recorded. This record does not assert independent reproduction.

### AMUSE

Iteration Time (s): 0.719. Result ID: 6a101a68fbc373592c09434a.

[Reporting paper](<https://www.sota2.com/research/paper/6a0fe30af984739ed6739d6f.md>)

[Method origin paper](<https://www.sota2.com/research/paper/6a0fe30af984739ed6739d6f.md>)

Recorded evaluation configuration: Not recorded. This record does not assert independent reproduction.

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