Home › Datasets › Fitness & Nutrition
Fitness & Nutrition Fine-Tuning Dataset
Fitness, nutrition, health, wellness
- 716 instruction-response pairs
- Format JSONL
- Domain fitness
- Dataset $19 one-time
$19 for the dataset. Training is priced separately by model, starting at $47.
What is in this dataset
716 instruction-response pairs covering fitness, nutrition, health, wellness. 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 new to working out and want to start with a full-body routine. What exercises should I include, and how do I structure it?
A beginner full-body workout should target all major muscle groups with compound movements—exercises that work multiple joints and muscles at once. A solid template includes: 1) a lower-body push (e.g., bodyweight squats), 2) a lower-body pull or hinge (e.g., glute bridges), 3) an upper-body push (e.g., push-ups or incline push-ups), 4) an upper-body pull (e.g., rows using a resistance band or dumbbell), and 5) a core stability exercise (e.g., planks). Perform 2–3 sets of 8–12 reps per exercise, resting 60–90 seconds between sets. Aim for 2–3 non-consecutive days per week, like Monday, Wednesday, and Friday. Always warm up with 5 minutes of light cardio and dynamic stretches, and cool down…
Can you explain how to do a proper push-up? I'm worried about hurting my shoulders.
A proper push-up starts with good setup: place your hands slightly wider than shoulder-width apart, with fingers pointing forward or slightly outward. Your body should form a straight line from head to heels—engage your core and glutes to avoid sagging or piking your hips. As you lower, keep your elbows at about a 45-degree angle to your torso (not flared out to 90 degrees), and lower your chest toward the floor until your elbows are at 90 degrees or your chest lightly touches the ground. Push through your palms to return to the start, exhaling on the effort. Common mistakes include letting your head drop or craning your neck—keep your gaze a few inches in front of your hands. If you feel…
I want to create a weekly workout split that balances strength and cardio. I can exercise 4 days a week. What would you recommend?
A balanced 4-day split could be: Day 1—Upper Body Strength (e.g., push-ups, rows, overhead press), Day 2—Lower Body Strength (e.g., squats, lunges, deadlifts), Day 3—Active Recovery or Cardio (e.g., 30 minutes of brisk walking, cycling, or swimming at moderate intensity), Day 4—Full Body or Hybrid (e.g., kettlebell swings, step-ups, and a short HIIT circuit like 20 seconds of work, 40 seconds rest for 10 rounds). Space strength days with at least 48 hours between them for the same muscle groups—for example, do Upper Body on Monday, Lower Body on Tuesday, Cardio on Thursday, and Full Body on Saturday. Each strength session should include 5–6 exercises, 3 sets of 8–12 reps, with 60–90 second…
Who this is for
Training and nutrition apps, with consult-a-professional notes throughout.
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:
- Qwen — Qwen3.5 (27B, 9B, 4B), Qwen3, Qwen2.5, Qwen2.5-Coder
- Gemma — Gemma 4, Gemma 3, Gemma 2
- Phi — Phi-4, Phi-3
- DeepSeek — R1 Distill (70B, 32B, 14B)
- Mistral and Llama families
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
- Pick this dataset and a base model.
- Rebiha runs the job on a managed GPU. You set up nothing.
- 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
- SHTF: Emergency Preparedness & Survival — 878 pairs
- Business & Finance — 956 pairs
- Legal DIY — 821 pairs
- E-commerce & Product Descriptions — 948 pairs