reasonwang/skillgym-sft
SkillGym SFT trajectories 19,053 successful (reward 1.0) agent trajectories on skill-requiring tasks, one JSONL file per agent harness. file harness trajectories agentfly.jsonl AgentFly 5,936 miniswe.jsonl mini-swe-agent 6,423 terminus-2.jsonl terminus-2 6,694 Record fields conversation (list of messages), harness, model (teacher model), task, skill (owner/repo/skill), task_profile (hard_procedural, hard_abductive, hard_constraint_satisfaction… See the full description on the dataset page: https://huggingface.co/datasets/reasonwang/skillgym-sft.
| Author / organization | reasonwang |
|---|---|
| Downloads | 224 (as reported by Hugging Face) |
| Likes | 3 |
| Created | 1 October 2026 |
| Last updated at source | 1 October 2026 |
Popularity, last 3 days
224 now (+13 since 2026-10-16). Counts as reported by the source, snapshotted daily by Silytics.
Connected in the knowledge graph
- mradermacher/SkillGym-Qwen3.5-27B-GGUFModel · used to train
- mradermacher/SkillGym-Qwen3.5-9B-GGUFModel · used to train
- mradermacher/SkillGym-Qwen3.5-122B-A10B-GGUFModel · used to train
- reasonwang/SkillGym-Qwen3.5-4BModel · used to train
- reasonwang/SkillGym-Qwen3.5-9BModel · used to train
- reasonwang/SkillGym-Qwen3.5-122B-A10BModel · used to train
- mradermacher/SkillGym-Qwen3.5-4B-GGUFModel · used to train
More from reasonwang
- reasonwang/skillgym-testDataset
- reasonwang/ToolGen-DatasetsDataset
Silytics shows public metadata from Hugging Face and links to the original. Information may be incomplete or out of date. Last refreshed 18 October 2026. Licenses and terms of the original apply to the work itself. Not financial, legal or professional advice.