A 11,705-example ShareGPT instruction-tuning dataset for training marketing and growth strategist LLMs. Each example is a multi-turn conversation with a fixed system persona, a realistic growth/marketing user question, and a structured assistant response.

Provenance

This dataset was distilled from DeepSeek V4 Pro (accounts/fireworks/models/deepseek-v4-pro) via the Fireworks AI API. Source chunks were extracted from materials in the… See the full description on the dataset page: https://huggingface.co/datasets/opestro/marketing-11k-valid.