Search-VL-SFT-36K is a dataset for supervised fine-tuning of frontier multimodal search agents, created by OpenSearch-VL. The dataset was last updated on May 7, 2026. It likely contains data for training agents on multi-turn, fatal-aware tasks with visual tool use.
Use Cases
- Training multimodal search agents based on supervised fine-tuning recipes
- Developing agents capable of multi-turn interactions based on the described method
- Implementing fatal-aware agent training based on the GRPO framework mentioned
- Benchmarking visual tool use capabilities in agentic systems
Strengths
- Dataset is associated with a specific project, OpenSearch-VL, which suggests a research focus
- Last update timestamp is provided (2026-05-07 05:18:27)
Limitations
- Description metadata is limited; actual data quality requires manual inspection after download
- Column-level documentation is absent; field semantics must be inferred after download
- Row count is unknown, which may limit suitability assessment
Provenance
- Source
- OpenSearch-VL
- Collection Method
- Likely generated for supervised fine-tuning of multimodal agents
- Freshness
- Last updated 2026-05-07 05:18:27