# End-to-end single-step decoding on DeepSeek-R1-Distill-LLaMA-8B 32K Context

Source: [End-to-end single-step decoding on DeepSeek-R1-Distill-LLaMA-8B 32K Context](<https://www.sota2.com/research/sota/end-to-end-single-step-decoding-on-deepseek-r1-distill-llama-8b-32k-context>)

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

Data updated: 2026-04-29T07:01:36.114Z

Evaluation of the end-to-end single-step decoding latency for the DeepSeek-R1-Distill-LLaMA-8B model configured with a 32K context window.

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 | End-to-end single-step decoding |
| Dataset (including any stated split) | DeepSeek-R1-Distill-LLaMA-8B 32K Context |
| Metric | Latency (ms) |
| 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

- [Latency (ms)](<https://www.sota2.com/research/sota/end-to-end-single-step-decoding-on-deepseek-r1-distill-llama-8b-32k-context.md?metric=Latency%20(ms)>): Lower is better; unit: TIME.

## Recorded results

| Method | Latency (ms) | Result date | Report paper |
| --- | --- | --- | --- |
| LessIsMore | 23.4 | Not recorded | 68fa7a801e209131cea83853 |
| Quest | 24.4 | Not recorded | 68fa7a801e209131cea83853 |
| TidalDecode | 24.7 | Not recorded | 68fa7a801e209131cea83853 |
| Full Attention | 28.4 | Not recorded | 68fa7a801e209131cea83853 |

### LessIsMore

Latency (ms): 23.4. Result ID: 69f1ab4447f34ce924d79e36.

[Reporting paper](<https://www.sota2.com/research/paper/68fa7a801e209131cea83853.md>)

[Method origin paper](<https://www.sota2.com/research/paper/68fa7a801e209131cea83853.md>)

Recorded evaluation configuration: {"Token Budget":"2K","Serving Stack":"SGLang + FlashInfer","Hardware":"NVIDIA A5000 GPU"}. This record does not assert independent reproduction.

### Quest

Latency (ms): 24.4. Result ID: 69f1ab4447f34ce924d79e38.

[Reporting paper](<https://www.sota2.com/research/paper/68fa7a801e209131cea83853.md>)

[Method origin paper](<https://www.sota2.com/research/paper/68fa74a51e209131cea62e6e.md>)

Recorded evaluation configuration: {"Token Budget":"2K","Serving Stack":"SGLang + FlashInfer","Hardware":"NVIDIA A5000 GPU"}. This record does not assert independent reproduction.

### TidalDecode

Latency (ms): 24.7. Result ID: 69f1ab4447f34ce924d79e37.

[Reporting paper](<https://www.sota2.com/research/paper/68fa7a801e209131cea83853.md>)

[Method origin paper](<https://www.sota2.com/research/paper/68fa75161e209131cea65a64.md>)

Recorded evaluation configuration: {"Token Budget":"2K","Serving Stack":"SGLang + FlashInfer","Hardware":"NVIDIA A5000 GPU"}. This record does not assert independent reproduction.

### Full Attention

Latency (ms): 28.4. Result ID: 69f1ab4447f34ce924d79e39.

[Reporting paper](<https://www.sota2.com/research/paper/68fa7a801e209131cea83853.md>)

Recorded evaluation configuration: {"Token Budget":"2K","Serving Stack":"SGLang + FlashInfer","Hardware":"NVIDIA A5000 GPU"}. This record does not assert independent reproduction.

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