Russian Medical Dialogues: 2,282 Patient-Bot Conversations

Russian-language dataset of 2,282 patient conversations with a medical bot. The training sample includes 2,053 conversations, and the test sample includes 229 conversations. It was authored by Mykes and last updated on July 15, 2024.

Use Cases

• Fine-tuning instruction-following models based on the provided prompt format.
• Training medical dialogue systems based on the user-ai message history (context).
• Evaluating AI response quality in a medical domain based on the assistant_answer field.
• Analyzing user query patterns in medical consultations based on the user_question field.

Strengths

• 2,282 total conversations provide a substantial text corpus.
• Explicitly split into 2,053 training and 229 test conversations for model development.
• Includes a ready prompt format adapted for fine-tuning with specific libraries.

Limitations

• Column-level documentation is absent; field semantics must be inferred after download.
• Row count for the full dataset is unknown, which may limit suitability assessment.
• Data may reflect bias inherent to the specific medical bot source.

Provenance

Source

huggingface

Collection Method

Conversations from a medical bot.

Time Range

null

Freshness

Last updated 2024-07-15 06:58:36.

Geography

null