WizardLM Evol Instruct 70k
Uses the 'Evol-Instruct' method to progressively rewrite simple instructions into increasingly complex and diverse ones using GPT-4. Results in a dataset where ~70k examples span from basic to graduate-level difficulty. WizardLM 13B outperformed Vicuna 13B on complex reasoning tasks using this data alone.
from datasets import load_dataset
ds = load_dataset("WizardLM/WizardLM_evol_instruct_70k")Preview a sample row
{
"instruction": "Develop a detailed multi-step plan for how a newly established non-profit could build a sustainable funding model, incorporating grant writing, corporate partnerships, individual donors, and earned revenue streams.",
"output": "Building a sustainable funding model requires diversification across four pillars..."
}Fine-tune with this dataset
Estimated VRAM to fine-tune with QLoRA (4-bit base model + LoRA adapters), using conservative defaults:
New to fine-tuning? Follow the step-by-step walkthrough: Fine-Tune Your First LLM in 1 Hour
Frequently asked questions
Can I use WizardLM Evol Instruct 70k commercially?
Yes — WizardLM Evol Instruct 70k is released under Apache 2.0, 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 WizardLM Evol Instruct 70k contain, and do I need all of it?
WizardLM Evol Instruct 70k contains 70k Rows. 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 WizardLM Evol Instruct 70k best used for?
Boosting complex-instruction handling with Evol-Instruct data. It belongs to the Instruction / SFT section of our dataset hub, where you'll find alternatives and complementary sets.