Rebiha

Home › Datasets › Social Media Management & Content

Social Media Management & Content Fine-Tuning Dataset

Content strategy, community management, and social media advice.

Fine-tune a model on this dataset

$29 for the dataset. Training is priced separately by model, starting at $47.

What is in this dataset

741 instruction-response pairs covering content strategy, community management, and social media advice. Each pair is a realistic question with a full written answer, in the JSONL format used for supervised fine-tuning. No synthetic filler, no duplicated rows.

Example pairs

Real rows from the dataset, shown in full.

I'm a small business owner posting a photo of our new eco-friendly product packaging. How do I write a caption that doesn't sound salesy but still gets people to care?

Focus on the story behind the change, not the product itself. Start with a relatable pain point: 'We spent months trying to figure out how to ship your orders without drowning the planet in plastic.' Then pivot to the solution with a specific detail: 'This new wrap dissolves in hot water, so you can toss it in your sink instead of a landfill.' End with a micro-commitment: 'Tag a friend who’s helped you go greener—we’ll send you both a sample pack.' Avoid emoji overload; one well-placed 🌍 at the end is enough. Use line breaks for scannability: pain point, solution, call to action.

I run a lifestyle brand and I'm struggling to plan my content for the week. How do I create a weekly content calendar that actually works without overcomplicating it?

Start with a simple four-pillar system: one educational post (e.g., '3 ways to style this scarf'), one behind-the-scenes (e.g., your packing process), one user-generated content (repost a customer photo with credit), and one sales or announcement (e.g., a new drop). Block each pillar on specific days: educational on Monday, BTS on Wednesday, UGC on Friday, sales on Saturday. Every Sunday, batch-write captions and schedule them in your tool. Leave one 'wildcard' slot for trending audio or a real-time moment. The key is consistency over perfection—stick to the pillars for 4 weeks, then audit which got the most saves and shares.

I got a nasty comment on my brand's latest post calling our product overpriced junk. What's a professional way to respond without feeding the troll?

First, assess intent: if it's a genuine complaint, reply publicly with empathy, then move to DMs. If it's a clear troll, one calm public response is enough—don't engage further. Example: 'Hey [name], we hear you. Pricing is tough because we use [specific material or process] that costs more but lasts longer. We'd love to understand your experience better—could you DM us so we can make it right?' This shows other readers you're responsive and transparent, without arguing. Never delete the comment unless it violates platform rules, as that can escalate. If the tone is aggressive but the point is valid, acknowledge the emotion: 'We appreciate you being…

Who this is for

Social media managers and content-planning tools.

Which models this works with

The dataset is plain JSONL, so it works with any instruction-tuned open model. On Rebiha you can train it directly on:

Training uses LoRA, or QLoRA on a 4-bit base for larger models. Your base model's weights are never modified — training produces a separate adapter.

How it works

  1. Pick this dataset and a base model.
  2. Rebiha runs the job on a managed GPU. You set up nothing.
  3. Download your model: a ready-to-run GGUF, plus developer assets (adapter, tokenizer, configs).

Fine-tuning or RAG?

Use RAG when the model needs facts that change — your prices, your documents, today's inventory. Use fine-tuning when you want consistent style and structure without prompting for it every time. This dataset teaches the shape of a good answer in this domain, not facts to look up.

Related datasets

Start fine-tuning