Medical R1 Distill Data: Supervised Fine-Tuning Dataset for Medical Reasoning

A distilled SFT dataset based on medical verifiable problems from HuatuoGPT-o1 and the Deepseek-R1 (Full Power Version) model. The dataset, created by FreedomIntelligence and last updated on 2025-02-22, is intended to help initialize models with reasoning chains from the R1 model.

Use Cases

• Initialize language models for medical QA based on the described distillation from Deepseek-R1.
• Train models to generate reasoning chains based on the dataset's stated purpose.
• Fine-tune models on verified medical problems as referenced in the description.
• Study knowledge distillation techniques in the medical domain based on the dataset's origin.

Strengths

• Dataset is derived from a specific, named model (Deepseek-R1) and a medical QA source (HuatuoGPT-o1).
• Last update date is precisely recorded as 2025-02-22 06:55:02.
• A Chinese-language version of the dataset is explicitly noted, suggesting multilingual potential.

Limitations

• Column-level documentation is absent; field semantics must be inferred after download.
• Row count, file formats, and dataset size are unknown, which may limit suitability assessment.
• License information is unknown, which could restrict usage.

Provenance

Source

FreedomIntelligence

Collection Method

Distilled from native Deepseek-R1 API requests based on medical verifiable problems from HuatuoGPT-o1.

Time Range

null

Freshness

Last updated 2025-02-22 06:55:02; freshness should be verified.

Geography

null