This repository contains 12 end-to-end demos and sample datasets for fine-tuning models on Azure AI Foundry. Use this repo to explore practical fine-tuning workflows and access ready-to-use data for your own projects.
New to fine-tuning? Start here:
- SFT_CNN_DailyMail - Best first demo! Learn Supervised Fine-Tuning with news summarization
- Sample_Datasets - Understand data formats for SFT, DPO, and RFT
Ready for advanced techniques?
- DPO_Intel_Orca - Direct Preference Optimization
- RFT_Countdown - Reinforcement Fine-Tuning
Explore end-to-end fine-tuning experiences in the Demos folder:
| Demo | Technique | Use Case | Difficulty |
|---|---|---|---|
| SFT_CNN_DailyMail | SFT | News summarization | ⭐ Beginner |
| SFT_PubMed_Summarization | SFT | Medical paper summarization | ⭐ Beginner |
| DPO_Intel_Orca | DPO | Preference optimization | ⭐⭐ Intermediate |
| RFT_Countdown | RFT | Math puzzle solving | ⭐⭐ Intermediate |
| DistillingSarcasm | Distillation | Knowledge transfer | ⭐⭐ Intermediate |
| Image_Breed_Classification_FT | Vision SFT | Dog breed classification | ⭐⭐ Intermediate |
| Image_FT_Chart_Analysis | Vision SFT | Chart understanding | ⭐⭐ Intermediate |
| Video_FT_Action_Recognition | Vision SFT | Video action detection | ⭐⭐⭐ Advanced |
| ZavaRetailAgent | SFT + RFT | Retail customer service agent | ⭐⭐⭐ Advanced |
| Agentic_RFT_PrivatePreview | RFT | Agentic workflows with tools | ⭐⭐⭐ Advanced |
| Evaluation | Evaluation | Multimodal model evaluation | ⭐⭐ Intermediate |
👉 See Demos/README.md for detailed descriptions of each demo.
Ready-to-use datasets for testing fine-tuning techniques in the Sample_Datasets folder:
| Technique | Dataset | Description |
|---|---|---|
| SFT | Text-GSM8K | Grade school math problems |
| SFT | Multimodal-chartqa | Chart interpretation |
| SFT | Tool-Calling | Function calling patterns |
| DPO | orca_dpo_pairs | Preference alignment |
| RFT | ClauseMatching | Legal contract analysis |
| RFT | MedMCQ | Medical Q&A |
👉 See Sample_Datasets/README.md for data format details and when to use each technique.
⚠️ Note: These datasets are for learning and experimentation only—not for production use. Training jobs may incur costs on your Azure subscription.
Before running any demo, ensure you have:
- Azure subscription with access to Azure AI Foundry
- Python 3.9+ installed
- Jupyter Notebook or VS Code with Jupyter extension
- Required Azure role assignments (see individual demo READMEs)
Each demo includes a requirements.txt and .env.template for setup.
👉 New here? See the Getting Started Guide for step-by-step setup instructions.
See CONTRIBUTING.md for guidelines on submitting issues and pull requests.
This project may contain trademarks or logos for projects, products, or services. Authorized use of Microsoft trademarks or logos is subject to and must follow Microsoft’s Trademark & Brand Guidelines. Use of Microsoft trademarks or logos in modified versions of this project must not cause confusion or imply Microsoft sponsorship. Any use of third-party trademarks or logos are subject to those third-party’s policies.