Tulu 3 SFT Mix — LLM Instruction / SFT Dataset
AllenAI's carefully curated 939K SFT training mix that powers the Tulu 3 model family. Combines open-licensed instruction data from multiple domains (math, science, code, safety) with synthetic data generation, achieving performance competitive with frontier models.
Dataset Details
| Provider | allenai |
| Category | Instruction / SFT |
| Size | 939K Samples |
| License | ODC-BY |
| Downloads | 380k |
| Tags | Multi-domain, Curated Mix, AllenAI, Open License, 2024 |
from datasets import load_dataset
ds = load_dataset("allenai/tulu-3-sft-mixture")
Fine-tune with this dataset
Estimated VRAM to fine-tune with QLoRA (4-bit base model + LoRA adapters), using conservative defaults:
| 7B QLoRA | ~6GB VRAM |
| 13B QLoRA | ~10GB VRAM |
Check if your GPU can fine-tune this →
New to fine-tuning? Follow the step-by-step walkthrough: Fine-Tune Your First LLM in 1 Hour
Related datasets
- UltraFeedback — The default DPO preference set to run after any SFT pass
- SmolTalk 2 — Reproducing a complete modern post-training pipeline (mid-training → SFT → preference) for
- Python-Edu — Continued pretraining for Python code understanding
- OpenHermes 2.5 — The default general-purpose SFT mix for 7B-13B fine-tunes
Frequently asked questions
Can I use Tulu 3 SFT Mix commercially?
Yes — Tulu 3 SFT Mix is released under ODC-BY, a permissive license that allows commercial use, including training models you ship in a product. Check the dataset card for attribution requirements before release.
How much data does Tulu 3 SFT Mix contain, and do I need all of it?
Tulu 3 SFT Mix contains 939K Samples. You rarely need all of it: for style and format fine-tuning, a few hundred to a few thousand examples are enough — load a slice (e.g. split="train[:1000]") and scale up only if quality plateaus.
What is Tulu 3 SFT Mix best used for?
Reproducing a state-of-the-art fully open post-training recipe. It belongs to the Instruction / SFT section of our dataset hub, where you'll find alternatives and complementary sets.
← All datasets | Fine-Tuning Guide