OpenSeeker Search Agent Training Examples

11,700 training examples were used to fine-tune the Qwen3-30B-A3B-Thinking-2507 model for search agent tasks. The dataset, created by the OpenSeeker project, was last updated in March 2026 to democratize frontier search capabilities.

Use Cases

• Fine-tuning large language models for search tasks using the 11.7K training examples.
• Benchmarking search agent performance on tasks like BrowseComp-ZH using the provided evaluation data.
• Studying the impact of open-sourced training data on model performance for question-answering agents.

Strengths

• Contains 11,700 training examples for model fine-tuning.
• Achieved a benchmark score of 48.4 on BrowseComp-ZH.

Limitations

• Dataset size of 11.7K examples is relatively small for large-scale model training.
• Specific column structure and data features are not publicly documented.

Provenance

Source

OpenSeeker project.

Collection Method

Created for fine-tuning a Qwen3-30B-A3B-Thinking-2507 model.

Freshness

Last updated March 2026.