mkurman/med-synth-questions-gemma-3-27b-deepseek-v4-flash
Med Synth Questions (Gemma-3 + DeepSeek V4 Flash) Synthetic reasoning traces and answers for medical questions from openmed-community/med-synth-questions-gemma-3-27b-it. Each record contains a medical question with SYNTH-style reasoning and a generated answer by DeepSeek V4 Flash. Dataset Summary 29,148 records (2 dupes + 3,410 incomplete/truncated removed from 32,560 source) 29,148 reasoning turns (99.2% format compliance) Average 1,591 chars per reasoning trace… See the full description on the dataset page: https://huggingface.co/datasets/mkurman/med-synth-questions-gemma-3-27b-deepseek-v4-flash.
Med Synth Questions (Gemma-3 + DeepSeek V4 Flash)
Synthetic reasoning traces and answers for medical questions from openmed-community/med-synth-questions-gemma-3-27b-it. Each record contains a medical question with SYNTH-style reasoning and a generated answer by DeepSeek V4 Flash.
Dataset Summary
- 29,148 records (2 dupes + 3,410 incomplete/truncated removed from 32,560 source)
- 29,148 reasoning turns (99.2% format compliance)
- Average 1,591 chars per reasoning trace
- Incomplete records (cut off mid-sentence, likely hit max_tokens) filtered out
Provider
Filtering
Records where the assistant content did not end with terminal punctuation (., !, ?, ), ], }, ", ', >) were removed. These were predominantly the longest responses that hit the max_tokens cutoff mid-sentence.
Data Format
{
"messages": [
{"role": "user", "content": "Medical question..."},
{"role": "assistant", "content": "Complete answer ending with proper punctuation.", "reasoning_content": "● Question: ... → ..."}
],
"source": "hf:openmed-community/med-synth-questions-gemma-3-27b-it",
"source_index": 500,
"target": "both",
"protocol": "ollama",
"model": "deepseek-v4-flash:cloud",
"num_conversation_turns": 2,
"num_generated_turns": 1,
"generated_at": "2026-07-22T..."
}Validation
Disclaimer
This resource is provided for research and educational use. It is not a source of medical advice. Always consult qualified professionals and follow relevant laws, ethics, and platform/model terms. Use responsibly.