Dataset · Hugging Face

LogicStar/SWE-Star

Name: SWE-Star
Creator: LogicStar
Keywords: task_categories:text-generation, size_categories:100K<n<1M, format:parquet, modality:text, library:datasets, library:dask, library:polars, library:mlcroissant, region:us, code

SWE-Star Introduction SWE-Star is a family of language models based on the Qwen2.5-Coder family and trained on the SWE-Star dataset. The dataset contains approximately 250k agentic coding trajectories distilled from Devstral-2-Small using SWE-Smith tasks. The complete data generation, training, and evaluation pipeline is openly available in our GitHub repository, enabling anyone to reproduce our results. Additional details are available in our blog posts.… See the full description on the dataset page: https://huggingface.co/datasets/LogicStar/SWE-Star.

View on Hugging FaceReport an issue

Author / organization	LogicStar
Downloads	1.6K (as reported by Hugging Face)
Likes	2
Created	17 January 2026
Last updated at source	26 January 2026

code

Embed this entry

<iframe src="https://silytics.si/embed/item/c0565951-cfdb-4a52-a8c7-cdcaa4c45418" width="420" height="190" style="border:0" loading="lazy" title="Silytics entry"></iframe>

Free, no tracking. More about feeds and embeds.

Popularity, last 2 days

1,649 now (+87 since 2026-10-16). Counts as reported by the source, snapshotted daily by Silytics.

Connected in the knowledge graph

mradermacher/SWE-Star-7B-GGUFModel · used to train
mradermacher/SWE-Star-32B-i1-GGUFModel · used to train
mradermacher/SWE-Star-14B-i1-GGUFModel · used to train
mradermacher/SWE-Star-7B-i1-GGUFModel · used to train
mradermacher/SWE-Star-14B-GGUFModel · used to train

Explore this entry in the graph

More from LogicStar

LogicStar/SWE-SmithDataset

Silytics shows public metadata from Hugging Face and links to the original. Information may be incomplete or out of date. Last refreshed 17 October 2026. Licenses and terms of the original apply to the work itself. Not financial, legal or professional advice.