# Finetuning

> Fine-tune models on Microsoft Foundry using SFT (supervised), DPO (preference), or RFT (reinforcement with graders). Covers dataset preparation, training job submission, deployment, and evaluation. USE FOR: fine-tune, SFT, DPO, RFT, training data, grader, distillation, fine-tuned model, training job, large file upload, calibrate grader, deploy fine-tuned model, evaluate fine-tuned model. DO NOT USE FOR: general model deployment without fine-tuning (use deploy-model), agent creation (use agents), prompt optimization without training (use prompt-optimizer).

Source: https://skillsagentes.com/skills/microsoft/azure-skills/finetuning
Repository: https://github.com/microsoft/azure-skills
Author: microsoft
License: MIT
Updated: hace 7 días
Context cost: 140 tok installed, 1.4k tok once triggered, 47.2k tok with every bundled file
Bundle: 34 files, 184 KB
Permissions requested: none declared

## Install

```bash
npx -y skills add microsoft/azure-skills --skill finetuning --agent claude-code
```

## What it does

- Hace fine-tuning de modelos en Microsoft Foundry con tres métodos: SFT (supervisado), DPO (por preferencia) y RFT (por refuerzo con graders).
- Cubre el ciclo entero: preparación y validación del dataset, envío y monitorización del trabajo de entrenamiento, despliegue y evaluación.
- Ayuda a elegir entre los tres tipos de entrenamiento y a calibrar graders y umbrales de aprobado para RFT.
- Incluye destilación, datos sintéticos y subida de archivos grandes.

## Use it when

- Se quiere hacer fine-tuning de un modelo, preparar datos de entrenamiento o enviar y diagnosticar trabajos de entrenamiento.
- Se quiere desplegar o evaluar un modelo ya ajustado.

## Don't bother when

- Para desplegar un modelo sin fine-tuning, que va a `deploy-model`.
- Para crear agentes, o para optimizar prompts sin entrenar.

## What triggers it

- "haz fine-tuning de este modelo"
- "¿SFT, DPO o RFT?"
- "prepara mis datos de entrenamiento"
- "despliega el modelo ajustado"

## Before you install

- El frontmatter declara versión `0.0.0-placeholder`, así que es un sub-skill de Foundry en estado inicial.
- Environment: API_KEY, AZURE_AI_PROJECT_ENDPOINT, AZURE_COGSERVICES_ACCOUNT, AZURE_OPENAI_API_KEY, AZURE_OPENAI_ENDPOINT, AZURE_RESOURCE_GROUP, AZURE_SUBSCRIPTION_ID, AZ_CLI_PATH, BASE_URL, ENDPOINT, OPENAI_BASE_URL

## Files

- SKILL.md — 5 KB
- references/agentic-rft.md — 3 KB
- references/dataset-formats.md — 4 KB
- references/deployment.md — 3 KB
- references/evaluation.md — 6 KB
- references/grader-design.md — 3 KB
- references/hyperparameters.md — 3 KB
- references/large-file-uploads.md — 927 B
- references/platform-gotchas.md — 2 KB
- references/reward-hacking.md — 3 KB
- references/training-curves.md — 4 KB
- references/training-types.md — 3 KB
- references/vision-fine-tuning.md — 4 KB
- scripts/calibrate_grader.py — 9 KB
- scripts/check_training.py — 8 KB
- scripts/cleanup.py — 9 KB
- scripts/common.py — 8 KB
- scripts/convert_dataset.py — 11 KB
- scripts/deploy_model.py — 9 KB
- scripts/evaluate_model.py — 11 KB
- scripts/generate_distillation_data.py — 10 KB
- scripts/monitor_training.py — 6 KB
- scripts/score_dataset.py — 8 KB
- scripts/submit_training.py — 11 KB
- scripts/validate/__init__.py — 305 B
- scripts/validate/data_stats.py — 6 KB
- scripts/validate/validate_dpo.py — 4 KB
- scripts/validate/validate_rft.py — 9 KB
- scripts/validate/validate_sft.py — 4 KB
- workflows/dataset-creation.md — 3 KB
- workflows/diagnose-poor-results.md — 2 KB
- workflows/full-pipeline.md — 3 KB
- workflows/iterative-training.md — 3 KB
- workflows/quickstart.md — 4 KB

## SKILL.md

Reproduced verbatim from microsoft/azure-skills under MIT. This section is the upstream document and is in English.

# Fine-Tuning on Microsoft Foundry

Fine-tune models using SFT (supervised), DPO (preference), or RFT (reinforcement with graders). Covers dataset prep, training, deployment, and evaluation.

## When to Use

Use this sub-skill when the user asks about:
- Fine-tuning a model (SFT, DPO, or RFT)
- Preparing, validating, or formatting training data
- Submitting, monitoring, or diagnosing training jobs
- Calibrating graders or pass thresholds for RFT
- Deploying or evaluating a fine-tuned model
- Choosing between training types (SFT vs DPO vs RFT)
- Distillation, synthetic data generation, or dataset quality scoring
- Large file uploads for training data
- Cleaning up fine-tuning resources (files, deployments)

**Do NOT use for:** General model deployment without fine-tuning (use deploy-model), agent creation (use agents), prompt optimization without training (use prompt-optimizer).

## Workflows

| Stage | Guide |
|-------|-------|
| **Quick start** | [workflows/quickstart.md](workflows/quickstart.md) |
| **Full pipeline** | [workflows/full-pipeline.md](workflows/full-pipeline.md) |
| **Create data** | [workflows/dataset-creation.md](workflows/dataset-creation.md) |
| **Iterate** | [workflows/iterative-training.md](workflows/iterative-training.md) |
| **Diagnose** | [workflows/diagnose-poor-results.md](workflows/diagnose-poor-results.md) |

## References

| Topic | File |
|-------|------|
| SFT vs DPO vs RFT | [references/training-types.md](references/training-types.md) |
| Hyperparameters | [references/hyperparameters.md](references/hyperparameters.md) |
| Data formats | [references/dataset-formats.md](references/dataset-formats.md) |
| Grader design (RFT) | [references/grader-design.md](references/grader-design.md) |
| Reward hacking | [references/reward-hacking.md](references/reward-hacking.md) |
| Agentic RFT (tools) | [references/agentic-rft.md](references/agentic-rft.md) |
| Deployment | [references/deployment.md](references/deployment.md) |
| Training curves | [references/training-curves.md](references/training-curves.md) |
| Evaluation | [references/evaluation.md](references/evaluation.md) |
| Vision fine-tuning | [references/vision-fine-tuning.md](references/vision-fine-tuning.md) |
| Large file uploads | [references/large-file-uploads.md](references/large-file-uploads.md) |
| Platform gotchas | [references/platform-gotchas.md](references/platform-gotchas.md) |

## Scripts

| Script | Purpose |
|--------|---------|
| `scripts/submit_training.py` | Submit SFT/DPO/RFT jobs |
| `scripts/monitor_training.py` | Poll job until completion |
| `scripts/calibrate_grader.py` | Find optimal RFT pass_threshold |
| `scripts/check_training.py` | Analyze curves, list checkpoints |
| `scripts/deploy_model.py` | Deploy via ARM REST API |
| `scripts/evaluate_model.py` | LLM judge evaluation |
| `scripts/convert_dataset.py` | Convert between SFT/DPO/RFT formats |
| `scripts/generate_distillation_data.py` | Generate synthetic training data |
| `scripts/score_dataset.py` | Quality scoring on training data |
| `scripts/cleanup.py` | Delete old files and deployments |
| `scripts/validate/` | Data validators (SFT, DPO, RFT) + stats |

## Rules

1. **Always baseline first** — evaluate the base model before fine-tuning
2. **Validate data** before submitting — run `scripts/validate/validate_sft.py`
3. **Calibrate RFT graders** — target 25-50% failure rate on the base model
4. **Evaluate checkpoints** — don't blindly deploy the final one
5. **Measure token cost** alongside accuracy when comparing models

## Quick Reference

| Task | Command |
|------|---------|
| Validate SFT data | `python scripts/validate/validate_sft.py data.jsonl` |
| Submit SFT job | `python scripts/submit_training.py --model gpt-4.1-mini --training-file train.jsonl --validation-file val.jsonl --type sft` |
| Monitor job | `python scripts/monitor_training.py --job-id ftjob-xxx` |
| Analyze curves | `python scripts/check_training.py --job-id ftjob-xxx` |
| Deploy model | `python scripts/deploy_model.py --model-id ft:gpt-4.1-mini:... --name my-eval` |
| Evaluate model | `python scripts/evaluate_model.py --deployment-name my-eval --test-file test.jsonl` |

## Error Handling

| Error | Cause | Fix |
|-------|-------|-----|
| "API version not supported" | Older `openai` SDK on `/v1/` endpoint | Upgrade to `openai>=1.0` |
| "does not support fine-tuning with Standard TrainingType" | OSS model needs `globalStandard` | Use `--use-rest` flag or script auto-falls back |
| Job stuck in post-training eval | Under-provisioned tool endpoint (RFT) | Scale to S2+, enable Always On |
| "DeploymentNotReady" after ARM succeeds | ARM/data-plane race condition | Delete and recreate deployment, wait 5 min |
| Content safety block at deployment | PII-dense training data | Remove problematic document types |

---

Skills Agentes — https://skillsagentes.com/skills/microsoft/azure-skills/finetuning