Kuairand Recexplain: Recommendation Explanation SFT/DPO and Semantic ID Data

Kuairand Recexplain SFT/DPO Data provides derived post-training data and Semantic ID assets for fine-tuning a large language model for recommendation explanation tasks. The dataset is associated with the LoRA adapter plumliu/qwen35-9b-kuairand-recexplain-lora and was last updated on June 22, 2026. Author plumliu created this resource to support work on explainable recommendation systems.

Use Cases

• Fine-tuning language models for recommendation explanation generation based on the described SFT/DPO data.
• Training or evaluating models for explainable AI in recommendation systems using the provided semantic ID assets.
• Researching instruction tuning methods for recommendation tasks based on the project's described focus.

Strengths

• Dataset is directly linked to a specific, released LoRA adapter (plumliu/qwen35-9b-kuairand-recexplain-lora), providing a clear application context.
• Includes both post-training data (SFT/DPO) and semantic embedding assets, suggesting a multi-component resource.
• Last update timestamp is precisely recorded as 2026-06-22 14:41:43.

Limitations

• Description metadata is limited; actual data quality, structure, and content require manual inspection after download.
• Column-level documentation is absent; field semantics and data format must be inferred after download.
• Row count, file formats, and license information are unknown, which may limit suitability assessment.

Provenance

Source

huggingface

Collection Method

Derived post-training data, likely from the KuaiRand recommendation platform.

Freshness

Last updated 2026-06-22 14:41:43; freshness should be verified.