# Instruction Fine-tuning on Alpaca-52K (train)

Source: [Instruction Fine-tuning on Alpaca-52K (train)](<https://www.sota2.com/research/sota/instruction-fine-tuning-on-alpaca-52k-train>)

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

Data updated: 2026-02-26T05:35:50.012Z

Evaluation tracking training efficiency and resulting average performance metrics (Avg., Time, Memory) when fine-tuning models on the Alpaca-52K training dataset.

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 | Instruction Fine-tuning |
| Dataset (including any stated split) | Alpaca-52K (train) |
| Metric | Average Score |
| Unit | PERCENT |
| Direction | Higher 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

- [Average Score](<https://www.sota2.com/research/sota/instruction-fine-tuning-on-alpaca-52k-train.md?metric=Average%20Score>): Higher is better; unit: PERCENT.

- [Fine-tuning Time (h)](<https://www.sota2.com/research/sota/instruction-fine-tuning-on-alpaca-52k-train.md?metric=Fine-tuning%20Time%20(h)>): Lower is better; unit: TIME.

- [Memory Usage (GB)](<https://www.sota2.com/research/sota/instruction-fine-tuning-on-alpaca-52k-train.md?metric=Memory%20Usage%20(GB)>): Lower is better; unit: SIZE.

## Recorded results

| Method | Average Score | Result date | Report paper |
| --- | --- | --- | --- |
| Full-finetuning | 65.88 | Not recorded | 68fa78e91e209131cea7b2c5 |
| QLORA | 65.69 | Not recorded | 68fa78e91e209131cea7b2c5 |
| GSQ-Tuning | 65.39 | Not recorded | 68fa78e91e209131cea7b2c5 |
| LLaMA2-7B | 64.13 | Not recorded | 68fa78e91e209131cea7b2c5 |

### Full-finetuning

Average Score: 65.88. Result ID: 6995ae9a3badfaffe6adae9c.

[Reporting paper](<https://www.sota2.com/research/paper/68fa78e91e209131cea7b2c5.md>)

Recorded evaluation configuration: {"Backbone":"LLaMA2-7B","LLMs branch":"16-16-16","low-rank branch":"16-16-16"}. This record does not assert independent reproduction.

### QLORA

Average Score: 65.69. Result ID: 6995ae9a3badfaffe6adae9d.

[Reporting paper](<https://www.sota2.com/research/paper/68fa78e91e209131cea7b2c5.md>)

[Method origin paper](<https://www.sota2.com/research/paper/68fa6f6a1e209131cea456e1.md>)

[Implementation](<https://github.com/artidoro/qlora>)

Recorded evaluation configuration: {"Backbone":"LLaMA2-7B","LLMs branch":"4-16-16","low-rank branch":"16-16-16"}. This record does not assert independent reproduction.

### GSQ-Tuning

Average Score: 65.39. Result ID: 6995ae9a3badfaffe6adae9e.

[Reporting paper](<https://www.sota2.com/research/paper/68fa78e91e209131cea7b2c5.md>)

[Method origin paper](<https://www.sota2.com/research/paper/68fa78e91e209131cea7b2c5.md>)

Recorded evaluation configuration: {"Backbone":"LLaMA2-7B","LLMs branch":"4-6-6","low-rank branch":"6-6-6"}. This record does not assert independent reproduction.

### LLaMA2-7B

Average Score: 64.13. Result ID: 6995ae9a3badfaffe6adae9b.

[Reporting paper](<https://www.sota2.com/research/paper/68fa78e91e209131cea7b2c5.md>)

[Method origin paper](<https://www.sota2.com/research/paper/68fa701f1e209131cea492fe.md>)

[Implementation](<https://github.com/facebookresearch/llama>)

Recorded evaluation configuration: {"LLMs branch":"16-16-None","low-rank branch":"w/o"}. This record does not assert independent reproduction.

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