Merci d'avoir envoyé votre demande ! Un membre de notre équipe vous contactera sous peu.
Merci d'avoir envoyé votre réservation ! Un membre de notre équipe vous contactera sous peu.
Plan du cours
Introduction to Devstral and Mistral Models
- Overview of Mistral’s open-source models
- Apache-2.0 licensing and enterprise adoption
- Devstral’s role in coding and agentic workflows
Self-Hosting Mistral and Devstral Models
- Environment preparation and infrastructure choices
- Containerization and deployment with Docker/Kubernetes
- Scaling considerations for production use
Fine-Tuning Techniques
- Supervised fine-tuning vs parameter-efficient tuning
- Dataset preparation and cleaning
- Domain-specific customization examples
Model Ops and Versioning
- Best practices for model lifecycle management
- Model versioning and rollback strategies
- CI/CD pipelines for ML models
Governance and Compliance
- Security considerations for open-source deployment
- Monitoring and auditability in enterprise contexts
- Compliance frameworks and responsible AI practices
Monitoring and Observability
- Tracking model drift and accuracy degradation
- Instrumentation for inference performance
- Alerting and response workflows
Case Studies and Best Practices
- Industry use cases of Mistral and Devstral adoption
- Balancing cost, performance, and control
- Lessons learned from open-source Model Ops
Summary and Next Steps
Pré requis
- An understanding of machine learning workflows
- Experience with Python-based ML frameworks
- Familiarity with containerization and deployment environments
Audience
- ML engineers
- Data platform teams
- Research engineers
14 Heures
