- Fully supported Granite language and code models
- Model IP indemnification for Granite models
- InstructLab LLM alignment tooling for scalable and accessible model fine-tuning
- Optimized, bootable model runtime instances
- Gen AI package dependencies and software drivers for AI hardware
- Synthetic data generation
- Containerized AI workloads
- Pre-installed specialized toolkits / drivers / optimizations designed for deep learning (NVIDIA’s TensorRT, cuDNN, TensorFlow-optimized libraries, etc.)
- Direct access / integration into cloud-native AI tools / ecosystems