A high-quality conversational dataset for fine-tuning LLMs on the pydantic-deepagents library.
Format: Conversational (JSONL with messages array)
Entries: 1,125 training examples
Language: Python (pydantic-ai framework)
Purpose: Code generation and agent creation patterns
Each entry follows the conversational format:
{
"messages": [
{"role": "system", "content": "You are an expert… See the full description on the dataset page:
https://huggingface.co/datasets/georgep4181/pydanticai-DeepAgents.
