---
license: apache-2.0
task_categories:
- text-generation
language:
- ru
tags:
- medical
- russian
- instruction-tuning
- diagnosis
- healthcare
- gemma
- qwen
- icd
- ru
- icd-codes
pretty_name: Russian ICD Diagnosis Explanations
size_categories:
- n<1K
---

Dataset Name: ICD-10 (Russian ICD (МКБ-10) Diagnosis Explanations)   

Description:   
This dataset provides a simplified, citizen-friendly explanation for diagnoses from the МКБ (International Classification of Diseases) reference guide, specifically formatted for fine-tuning Large Language Models (LLMs) in Russian.    
It contains instruction-tuning pairs for two prominent model architectures: Google's Gemma and the Qwen family, using their respective chat templates.  
The goal of this dataset is to translate complex medical diagnosis codes and their official descriptions into clear, easy-to-understand language for a non-specialist.   
Language: Russian  
Source: МКБ справочник диагнозов  
Format: The dataset is provided as two JSONL files, pre-formatted for instruction tuning.  
ICD_gemma.jsonl: Formatted according to Gemma's chat template.  
ICD_qwen.jsonl: Formatted according to Qwen's chat template.  
Structure: Each example contains a diagnosis with its numerical ID and a simplified explanation intended for the general public.